| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 750, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004, | |
| "grad_norm": 6.988942623138428, | |
| "learning_rate": 1.3333333333333336e-07, | |
| "loss": 1.2845, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.008, | |
| "grad_norm": 6.956341743469238, | |
| "learning_rate": 2.666666666666667e-07, | |
| "loss": 1.2808, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.012, | |
| "grad_norm": 6.937110900878906, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 1.2611, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.016, | |
| "grad_norm": 6.7839250564575195, | |
| "learning_rate": 5.333333333333335e-07, | |
| "loss": 1.2732, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 6.882997989654541, | |
| "learning_rate": 6.666666666666667e-07, | |
| "loss": 1.2664, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.024, | |
| "grad_norm": 6.942823886871338, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 1.268, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.028, | |
| "grad_norm": 6.420243740081787, | |
| "learning_rate": 9.333333333333334e-07, | |
| "loss": 1.2577, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 6.184328556060791, | |
| "learning_rate": 1.066666666666667e-06, | |
| "loss": 1.2439, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.036, | |
| "grad_norm": 6.088969707489014, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 1.2432, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 4.597912788391113, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 1.1761, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.044, | |
| "grad_norm": 4.545207500457764, | |
| "learning_rate": 1.4666666666666669e-06, | |
| "loss": 1.1907, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 4.477672100067139, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 1.1929, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.052, | |
| "grad_norm": 2.8524184226989746, | |
| "learning_rate": 1.7333333333333336e-06, | |
| "loss": 1.1881, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.056, | |
| "grad_norm": 2.8725757598876953, | |
| "learning_rate": 1.8666666666666669e-06, | |
| "loss": 1.1643, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.601120948791504, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.1545, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 2.457542657852173, | |
| "learning_rate": 2.133333333333334e-06, | |
| "loss": 1.1581, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.068, | |
| "grad_norm": 2.867591381072998, | |
| "learning_rate": 2.266666666666667e-06, | |
| "loss": 1.1464, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.072, | |
| "grad_norm": 3.6930630207061768, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 1.1216, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.076, | |
| "grad_norm": 3.9270849227905273, | |
| "learning_rate": 2.5333333333333338e-06, | |
| "loss": 1.1291, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.724717140197754, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 1.1498, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.084, | |
| "grad_norm": 3.2976064682006836, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 1.1392, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.088, | |
| "grad_norm": 2.683112382888794, | |
| "learning_rate": 2.9333333333333338e-06, | |
| "loss": 1.0711, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.092, | |
| "grad_norm": 2.0106770992279053, | |
| "learning_rate": 3.066666666666667e-06, | |
| "loss": 1.0696, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 1.7597943544387817, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 1.0776, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.7341334819793701, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.0791, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.104, | |
| "grad_norm": 1.6312559843063354, | |
| "learning_rate": 3.4666666666666672e-06, | |
| "loss": 1.0203, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.108, | |
| "grad_norm": 1.4349817037582397, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 1.0466, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.112, | |
| "grad_norm": 1.2373616695404053, | |
| "learning_rate": 3.7333333333333337e-06, | |
| "loss": 1.075, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.116, | |
| "grad_norm": 1.2284890413284302, | |
| "learning_rate": 3.866666666666667e-06, | |
| "loss": 1.0127, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.9730978608131409, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.0196, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.124, | |
| "grad_norm": 0.9298821687698364, | |
| "learning_rate": 4.133333333333333e-06, | |
| "loss": 1.0085, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 0.9621558785438538, | |
| "learning_rate": 4.266666666666668e-06, | |
| "loss": 0.974, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.132, | |
| "grad_norm": 1.032867193222046, | |
| "learning_rate": 4.4e-06, | |
| "loss": 1.0066, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.136, | |
| "grad_norm": 0.9056654572486877, | |
| "learning_rate": 4.533333333333334e-06, | |
| "loss": 1.007, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.8527935147285461, | |
| "learning_rate": 4.666666666666667e-06, | |
| "loss": 0.9797, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 0.7919298410415649, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 0.944, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.148, | |
| "grad_norm": 0.8432771563529968, | |
| "learning_rate": 4.933333333333334e-06, | |
| "loss": 0.9859, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.152, | |
| "grad_norm": 0.9072924256324768, | |
| "learning_rate": 5.0666666666666676e-06, | |
| "loss": 0.9217, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.156, | |
| "grad_norm": 0.9000473618507385, | |
| "learning_rate": 5.2e-06, | |
| "loss": 1.0214, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.8418648838996887, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 0.9908, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.164, | |
| "grad_norm": 0.8197311758995056, | |
| "learning_rate": 5.466666666666667e-06, | |
| "loss": 0.9556, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.168, | |
| "grad_norm": 0.7588638663291931, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 0.9895, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.172, | |
| "grad_norm": 0.780891478061676, | |
| "learning_rate": 5.733333333333334e-06, | |
| "loss": 0.9436, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.176, | |
| "grad_norm": 0.8387202620506287, | |
| "learning_rate": 5.8666666666666675e-06, | |
| "loss": 0.9396, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.7959762811660767, | |
| "learning_rate": 6e-06, | |
| "loss": 0.9163, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.184, | |
| "grad_norm": 0.7720610499382019, | |
| "learning_rate": 6.133333333333334e-06, | |
| "loss": 0.9423, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.188, | |
| "grad_norm": 0.8092529773712158, | |
| "learning_rate": 6.266666666666668e-06, | |
| "loss": 0.9291, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 0.6679025888442993, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 0.9342, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.196, | |
| "grad_norm": 0.6735404133796692, | |
| "learning_rate": 6.533333333333334e-06, | |
| "loss": 0.9692, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.6841786503791809, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.9217, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.204, | |
| "grad_norm": 0.6928501725196838, | |
| "learning_rate": 6.800000000000001e-06, | |
| "loss": 0.9415, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.208, | |
| "grad_norm": 0.6858928203582764, | |
| "learning_rate": 6.9333333333333344e-06, | |
| "loss": 0.9221, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.212, | |
| "grad_norm": 0.6531127691268921, | |
| "learning_rate": 7.066666666666667e-06, | |
| "loss": 0.9241, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.216, | |
| "grad_norm": 0.7553710341453552, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 0.938, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.8223641514778137, | |
| "learning_rate": 7.333333333333333e-06, | |
| "loss": 0.9446, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.224, | |
| "grad_norm": 0.7550656795501709, | |
| "learning_rate": 7.4666666666666675e-06, | |
| "loss": 0.9678, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.228, | |
| "grad_norm": 0.8396725058555603, | |
| "learning_rate": 7.600000000000001e-06, | |
| "loss": 0.9881, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.232, | |
| "grad_norm": 0.7557951807975769, | |
| "learning_rate": 7.733333333333334e-06, | |
| "loss": 0.8889, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.236, | |
| "grad_norm": 1.0014610290527344, | |
| "learning_rate": 7.866666666666667e-06, | |
| "loss": 0.9487, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.796205997467041, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.9578, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.244, | |
| "grad_norm": 0.7757744789123535, | |
| "learning_rate": 8.133333333333334e-06, | |
| "loss": 0.8946, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.248, | |
| "grad_norm": 0.8658983707427979, | |
| "learning_rate": 8.266666666666667e-06, | |
| "loss": 0.9212, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.252, | |
| "grad_norm": 0.7084074020385742, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 0.9059, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.7214030623435974, | |
| "learning_rate": 8.533333333333335e-06, | |
| "loss": 0.9252, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.6902719736099243, | |
| "learning_rate": 8.666666666666668e-06, | |
| "loss": 0.9111, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.264, | |
| "grad_norm": 0.8375234603881836, | |
| "learning_rate": 8.8e-06, | |
| "loss": 0.9215, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.268, | |
| "grad_norm": 0.6496827602386475, | |
| "learning_rate": 8.933333333333333e-06, | |
| "loss": 0.9181, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.272, | |
| "grad_norm": 0.7157948613166809, | |
| "learning_rate": 9.066666666666667e-06, | |
| "loss": 0.9102, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.276, | |
| "grad_norm": 0.6876959800720215, | |
| "learning_rate": 9.200000000000002e-06, | |
| "loss": 0.9628, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.8131017684936523, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 0.9093, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.284, | |
| "grad_norm": 0.7312742471694946, | |
| "learning_rate": 9.466666666666667e-06, | |
| "loss": 0.9196, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 0.6563302874565125, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 0.9103, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.292, | |
| "grad_norm": 0.6630368828773499, | |
| "learning_rate": 9.733333333333334e-06, | |
| "loss": 0.912, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.296, | |
| "grad_norm": 0.5970990657806396, | |
| "learning_rate": 9.866666666666668e-06, | |
| "loss": 0.9267, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.676497220993042, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9362, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.304, | |
| "grad_norm": 0.9087813496589661, | |
| "learning_rate": 9.999945845889795e-06, | |
| "loss": 0.9044, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.308, | |
| "grad_norm": 0.6283800601959229, | |
| "learning_rate": 9.999783384732242e-06, | |
| "loss": 0.9154, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.312, | |
| "grad_norm": 0.7198421359062195, | |
| "learning_rate": 9.999512620046523e-06, | |
| "loss": 0.9165, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.316, | |
| "grad_norm": 0.7074084877967834, | |
| "learning_rate": 9.99913355769784e-06, | |
| "loss": 0.888, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.7027278542518616, | |
| "learning_rate": 9.99864620589731e-06, | |
| "loss": 0.893, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.324, | |
| "grad_norm": 0.819341778755188, | |
| "learning_rate": 9.998050575201772e-06, | |
| "loss": 0.9172, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.328, | |
| "grad_norm": 0.6771552562713623, | |
| "learning_rate": 9.99734667851357e-06, | |
| "loss": 0.92, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.332, | |
| "grad_norm": 0.9496891498565674, | |
| "learning_rate": 9.99653453108026e-06, | |
| "loss": 0.9024, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 0.6092897653579712, | |
| "learning_rate": 9.995614150494293e-06, | |
| "loss": 0.8922, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.8747417330741882, | |
| "learning_rate": 9.994585556692624e-06, | |
| "loss": 0.8498, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.344, | |
| "grad_norm": 0.788902759552002, | |
| "learning_rate": 9.993448771956285e-06, | |
| "loss": 0.8773, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.348, | |
| "grad_norm": 0.8757255673408508, | |
| "learning_rate": 9.992203820909906e-06, | |
| "loss": 0.8891, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.352, | |
| "grad_norm": 0.8785430788993835, | |
| "learning_rate": 9.99085073052117e-06, | |
| "loss": 0.8881, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.356, | |
| "grad_norm": 0.7009172439575195, | |
| "learning_rate": 9.989389530100242e-06, | |
| "loss": 0.8782, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.8035963177680969, | |
| "learning_rate": 9.987820251299121e-06, | |
| "loss": 0.8983, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.364, | |
| "grad_norm": 0.6838966012001038, | |
| "learning_rate": 9.986142928110972e-06, | |
| "loss": 0.9172, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.368, | |
| "grad_norm": 0.8380683660507202, | |
| "learning_rate": 9.984357596869369e-06, | |
| "loss": 0.9262, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.372, | |
| "grad_norm": 0.6786786913871765, | |
| "learning_rate": 9.982464296247523e-06, | |
| "loss": 0.8963, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.376, | |
| "grad_norm": 0.7757643461227417, | |
| "learning_rate": 9.980463067257437e-06, | |
| "loss": 0.8983, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.647390604019165, | |
| "learning_rate": 9.978353953249023e-06, | |
| "loss": 0.9304, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.7196109294891357, | |
| "learning_rate": 9.976136999909156e-06, | |
| "loss": 0.8958, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.388, | |
| "grad_norm": 0.7190807461738586, | |
| "learning_rate": 9.973812255260692e-06, | |
| "loss": 0.8795, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.392, | |
| "grad_norm": 0.738398551940918, | |
| "learning_rate": 9.971379769661422e-06, | |
| "loss": 0.9067, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.396, | |
| "grad_norm": 0.767160177230835, | |
| "learning_rate": 9.968839595802982e-06, | |
| "loss": 0.9028, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.5966362953186035, | |
| "learning_rate": 9.966191788709716e-06, | |
| "loss": 0.8738, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.404, | |
| "grad_norm": 0.7598825693130493, | |
| "learning_rate": 9.963436405737476e-06, | |
| "loss": 0.9183, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.408, | |
| "grad_norm": 0.7260900735855103, | |
| "learning_rate": 9.960573506572391e-06, | |
| "loss": 0.8845, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.412, | |
| "grad_norm": 0.7225805521011353, | |
| "learning_rate": 9.957603153229559e-06, | |
| "loss": 0.8923, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.416, | |
| "grad_norm": 0.7357990145683289, | |
| "learning_rate": 9.95452541005172e-06, | |
| "loss": 0.9223, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.6201126575469971, | |
| "learning_rate": 9.951340343707852e-06, | |
| "loss": 0.9089, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.424, | |
| "grad_norm": 0.7525957226753235, | |
| "learning_rate": 9.948048023191728e-06, | |
| "loss": 0.9297, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.428, | |
| "grad_norm": 0.6546491384506226, | |
| "learning_rate": 9.94464851982042e-06, | |
| "loss": 0.8841, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 0.6633215546607971, | |
| "learning_rate": 9.941141907232766e-06, | |
| "loss": 0.8413, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.436, | |
| "grad_norm": 0.7645735740661621, | |
| "learning_rate": 9.937528261387753e-06, | |
| "loss": 0.8973, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.6612620949745178, | |
| "learning_rate": 9.933807660562898e-06, | |
| "loss": 0.9139, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.444, | |
| "grad_norm": 0.6729806661605835, | |
| "learning_rate": 9.929980185352525e-06, | |
| "loss": 0.8575, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.5767675638198853, | |
| "learning_rate": 9.926045918666045e-06, | |
| "loss": 0.8975, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.452, | |
| "grad_norm": 0.5622053146362305, | |
| "learning_rate": 9.92200494572614e-06, | |
| "loss": 0.8744, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.456, | |
| "grad_norm": 0.5797794461250305, | |
| "learning_rate": 9.91785735406693e-06, | |
| "loss": 0.9071, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.6670066714286804, | |
| "learning_rate": 9.913603233532067e-06, | |
| "loss": 0.8851, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.464, | |
| "grad_norm": 0.5439545512199402, | |
| "learning_rate": 9.909242676272797e-06, | |
| "loss": 0.8859, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.468, | |
| "grad_norm": 0.7104266881942749, | |
| "learning_rate": 9.904775776745959e-06, | |
| "loss": 0.9192, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.472, | |
| "grad_norm": 0.6157470345497131, | |
| "learning_rate": 9.90020263171194e-06, | |
| "loss": 0.924, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.476, | |
| "grad_norm": 0.6168134212493896, | |
| "learning_rate": 9.89552334023258e-06, | |
| "loss": 0.9017, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.6102380752563477, | |
| "learning_rate": 9.890738003669029e-06, | |
| "loss": 0.885, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.484, | |
| "grad_norm": 0.6744838953018188, | |
| "learning_rate": 9.88584672567954e-06, | |
| "loss": 0.8535, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.488, | |
| "grad_norm": 0.6119304895401001, | |
| "learning_rate": 9.880849612217238e-06, | |
| "loss": 0.9136, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.492, | |
| "grad_norm": 0.5440135598182678, | |
| "learning_rate": 9.875746771527817e-06, | |
| "loss": 0.891, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.496, | |
| "grad_norm": 0.6523988842964172, | |
| "learning_rate": 9.870538314147194e-06, | |
| "loss": 0.8793, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.7202532291412354, | |
| "learning_rate": 9.86522435289912e-06, | |
| "loss": 0.9114, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.504, | |
| "grad_norm": 0.7352282404899597, | |
| "learning_rate": 9.859805002892733e-06, | |
| "loss": 0.925, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.508, | |
| "grad_norm": 0.7045993804931641, | |
| "learning_rate": 9.85428038152006e-06, | |
| "loss": 0.9076, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.6108631491661072, | |
| "learning_rate": 9.84865060845349e-06, | |
| "loss": 0.8609, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.516, | |
| "grad_norm": 0.5544862747192383, | |
| "learning_rate": 9.842915805643156e-06, | |
| "loss": 0.8873, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.6903448104858398, | |
| "learning_rate": 9.83707609731432e-06, | |
| "loss": 0.8593, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.524, | |
| "grad_norm": 0.634729266166687, | |
| "learning_rate": 9.831131609964664e-06, | |
| "loss": 0.8927, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 0.5572838187217712, | |
| "learning_rate": 9.825082472361558e-06, | |
| "loss": 0.8913, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.532, | |
| "grad_norm": 0.690453827381134, | |
| "learning_rate": 9.818928815539266e-06, | |
| "loss": 0.8991, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.536, | |
| "grad_norm": 0.6307182908058167, | |
| "learning_rate": 9.812670772796113e-06, | |
| "loss": 0.8736, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.6966649889945984, | |
| "learning_rate": 9.806308479691595e-06, | |
| "loss": 0.9092, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.544, | |
| "grad_norm": 0.6075841188430786, | |
| "learning_rate": 9.799842074043438e-06, | |
| "loss": 0.8942, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.548, | |
| "grad_norm": 0.7378610968589783, | |
| "learning_rate": 9.793271695924621e-06, | |
| "loss": 0.9019, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.552, | |
| "grad_norm": 0.5990914106369019, | |
| "learning_rate": 9.786597487660336e-06, | |
| "loss": 0.8997, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.556, | |
| "grad_norm": 0.6251294016838074, | |
| "learning_rate": 9.779819593824909e-06, | |
| "loss": 0.902, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.6688668727874756, | |
| "learning_rate": 9.77293816123866e-06, | |
| "loss": 0.9324, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.564, | |
| "grad_norm": 0.5460559725761414, | |
| "learning_rate": 9.765953338964736e-06, | |
| "loss": 0.9258, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.568, | |
| "grad_norm": 0.639219343662262, | |
| "learning_rate": 9.75886527830587e-06, | |
| "loss": 0.9035, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.572, | |
| "grad_norm": 0.5945067405700684, | |
| "learning_rate": 9.751674132801106e-06, | |
| "loss": 0.8789, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.5545593500137329, | |
| "learning_rate": 9.744380058222483e-06, | |
| "loss": 0.848, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.7910006642341614, | |
| "learning_rate": 9.736983212571646e-06, | |
| "loss": 0.9007, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.584, | |
| "grad_norm": 0.617174506187439, | |
| "learning_rate": 9.729483756076436e-06, | |
| "loss": 0.8592, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.588, | |
| "grad_norm": 0.7085884809494019, | |
| "learning_rate": 9.721881851187406e-06, | |
| "loss": 0.8595, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.592, | |
| "grad_norm": 0.7189193367958069, | |
| "learning_rate": 9.714177662574316e-06, | |
| "loss": 0.8635, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.596, | |
| "grad_norm": 0.7401798367500305, | |
| "learning_rate": 9.70637135712256e-06, | |
| "loss": 0.9157, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.7053728103637695, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.8875, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.604, | |
| "grad_norm": 0.6575140953063965, | |
| "learning_rate": 9.690453074301035e-06, | |
| "loss": 0.8598, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.608, | |
| "grad_norm": 0.6678581237792969, | |
| "learning_rate": 9.682341441747446e-06, | |
| "loss": 0.8933, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.612, | |
| "grad_norm": 0.6995222568511963, | |
| "learning_rate": 9.674128381980073e-06, | |
| "loss": 0.8675, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.616, | |
| "grad_norm": 0.6304057240486145, | |
| "learning_rate": 9.665814072907293e-06, | |
| "loss": 0.8784, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.6518959403038025, | |
| "learning_rate": 9.657398694630713e-06, | |
| "loss": 0.8861, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 0.7137309312820435, | |
| "learning_rate": 9.648882429441258e-06, | |
| "loss": 0.8722, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.628, | |
| "grad_norm": 0.6826603412628174, | |
| "learning_rate": 9.640265461815235e-06, | |
| "loss": 0.9008, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.632, | |
| "grad_norm": 0.5501178503036499, | |
| "learning_rate": 9.63154797841033e-06, | |
| "loss": 0.8751, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.636, | |
| "grad_norm": 0.5834782719612122, | |
| "learning_rate": 9.622730168061568e-06, | |
| "loss": 0.8597, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.5569977164268494, | |
| "learning_rate": 9.613812221777212e-06, | |
| "loss": 0.872, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.644, | |
| "grad_norm": 0.619074821472168, | |
| "learning_rate": 9.604794332734647e-06, | |
| "loss": 0.8832, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.648, | |
| "grad_norm": 0.5631194710731506, | |
| "learning_rate": 9.595676696276173e-06, | |
| "loss": 0.8768, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.652, | |
| "grad_norm": 0.6887290477752686, | |
| "learning_rate": 9.586459509904786e-06, | |
| "loss": 0.8677, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.656, | |
| "grad_norm": 0.6014657616615295, | |
| "learning_rate": 9.577142973279896e-06, | |
| "loss": 0.8617, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.5356835722923279, | |
| "learning_rate": 9.567727288213005e-06, | |
| "loss": 0.8855, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.664, | |
| "grad_norm": 0.7122846841812134, | |
| "learning_rate": 9.55821265866333e-06, | |
| "loss": 0.8652, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.668, | |
| "grad_norm": 0.5801354646682739, | |
| "learning_rate": 9.548599290733393e-06, | |
| "loss": 0.923, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 0.6620292663574219, | |
| "learning_rate": 9.538887392664544e-06, | |
| "loss": 0.9191, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.676, | |
| "grad_norm": 0.6333494186401367, | |
| "learning_rate": 9.529077174832466e-06, | |
| "loss": 0.8987, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.7241418957710266, | |
| "learning_rate": 9.519168849742603e-06, | |
| "loss": 0.8746, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.684, | |
| "grad_norm": 0.6281126737594604, | |
| "learning_rate": 9.50916263202557e-06, | |
| "loss": 0.8844, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.688, | |
| "grad_norm": 0.7443065047264099, | |
| "learning_rate": 9.499058738432492e-06, | |
| "loss": 0.8763, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.692, | |
| "grad_norm": 0.6006516814231873, | |
| "learning_rate": 9.488857387830315e-06, | |
| "loss": 0.8783, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.696, | |
| "grad_norm": 0.6572588682174683, | |
| "learning_rate": 9.478558801197065e-06, | |
| "loss": 0.8808, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.628270149230957, | |
| "learning_rate": 9.468163201617063e-06, | |
| "loss": 0.8312, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.6484922170639038, | |
| "learning_rate": 9.457670814276083e-06, | |
| "loss": 0.8979, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.708, | |
| "grad_norm": 0.6248320937156677, | |
| "learning_rate": 9.44708186645649e-06, | |
| "loss": 0.8708, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.712, | |
| "grad_norm": 0.8069347739219666, | |
| "learning_rate": 9.436396587532297e-06, | |
| "loss": 0.9053, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.716, | |
| "grad_norm": 0.5655907392501831, | |
| "learning_rate": 9.425615208964217e-06, | |
| "loss": 0.8701, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.9169421195983887, | |
| "learning_rate": 9.414737964294636e-06, | |
| "loss": 0.9063, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.724, | |
| "grad_norm": 0.6653777956962585, | |
| "learning_rate": 9.403765089142554e-06, | |
| "loss": 0.8534, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.728, | |
| "grad_norm": 0.7143352627754211, | |
| "learning_rate": 9.392696821198488e-06, | |
| "loss": 0.868, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.732, | |
| "grad_norm": 0.7472746968269348, | |
| "learning_rate": 9.381533400219319e-06, | |
| "loss": 0.8942, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.736, | |
| "grad_norm": 0.6738988757133484, | |
| "learning_rate": 9.370275068023097e-06, | |
| "loss": 0.9089, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.7050944566726685, | |
| "learning_rate": 9.358922068483813e-06, | |
| "loss": 0.8677, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.744, | |
| "grad_norm": 0.7963171601295471, | |
| "learning_rate": 9.347474647526095e-06, | |
| "loss": 0.8855, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.748, | |
| "grad_norm": 0.8421769738197327, | |
| "learning_rate": 9.335933053119906e-06, | |
| "loss": 0.8683, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.752, | |
| "grad_norm": 0.6927816867828369, | |
| "learning_rate": 9.324297535275156e-06, | |
| "loss": 0.8607, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.756, | |
| "grad_norm": 0.7820339202880859, | |
| "learning_rate": 9.312568346036288e-06, | |
| "loss": 0.8911, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.6646762490272522, | |
| "learning_rate": 9.30074573947683e-06, | |
| "loss": 0.8981, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.764, | |
| "grad_norm": 0.6774222254753113, | |
| "learning_rate": 9.288829971693869e-06, | |
| "loss": 0.8827, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.5905522108078003, | |
| "learning_rate": 9.276821300802535e-06, | |
| "loss": 0.8589, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.772, | |
| "grad_norm": 0.7953365445137024, | |
| "learning_rate": 9.264719986930376e-06, | |
| "loss": 0.8921, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.776, | |
| "grad_norm": 0.6584376692771912, | |
| "learning_rate": 9.25252629221175e-06, | |
| "loss": 0.8886, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.6534374952316284, | |
| "learning_rate": 9.24024048078213e-06, | |
| "loss": 0.891, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.784, | |
| "grad_norm": 0.6690033078193665, | |
| "learning_rate": 9.227862818772392e-06, | |
| "loss": 0.8743, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.788, | |
| "grad_norm": 0.6820501089096069, | |
| "learning_rate": 9.215393574303043e-06, | |
| "loss": 0.8985, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.792, | |
| "grad_norm": 0.7419943809509277, | |
| "learning_rate": 9.202833017478421e-06, | |
| "loss": 0.8848, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.796, | |
| "grad_norm": 0.6446091532707214, | |
| "learning_rate": 9.190181420380838e-06, | |
| "loss": 0.8926, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.715246856212616, | |
| "learning_rate": 9.177439057064684e-06, | |
| "loss": 0.8825, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.804, | |
| "grad_norm": 0.7427367568016052, | |
| "learning_rate": 9.164606203550498e-06, | |
| "loss": 0.8927, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.808, | |
| "grad_norm": 0.4903748333454132, | |
| "learning_rate": 9.151683137818989e-06, | |
| "loss": 0.8901, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.812, | |
| "grad_norm": 0.580417811870575, | |
| "learning_rate": 9.138670139805004e-06, | |
| "loss": 0.8637, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 0.610196590423584, | |
| "learning_rate": 9.125567491391476e-06, | |
| "loss": 0.8796, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.6373223066329956, | |
| "learning_rate": 9.112375476403313e-06, | |
| "loss": 0.8838, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.824, | |
| "grad_norm": 0.5853657722473145, | |
| "learning_rate": 9.099094380601244e-06, | |
| "loss": 0.8806, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.828, | |
| "grad_norm": 0.49319297075271606, | |
| "learning_rate": 9.085724491675642e-06, | |
| "loss": 0.8677, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.6091550588607788, | |
| "learning_rate": 9.072266099240286e-06, | |
| "loss": 0.8656, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.836, | |
| "grad_norm": 0.5975654125213623, | |
| "learning_rate": 9.058719494826076e-06, | |
| "loss": 0.9051, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.7250429391860962, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.8553, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.844, | |
| "grad_norm": 0.6404184103012085, | |
| "learning_rate": 9.031362825732456e-06, | |
| "loss": 0.838, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.848, | |
| "grad_norm": 0.636917233467102, | |
| "learning_rate": 9.017553353643479e-06, | |
| "loss": 0.8485, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.852, | |
| "grad_norm": 0.7165372371673584, | |
| "learning_rate": 9.003656854743667e-06, | |
| "loss": 0.8685, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.856, | |
| "grad_norm": 0.6287906765937805, | |
| "learning_rate": 8.989673630054044e-06, | |
| "loss": 0.8838, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.6897153258323669, | |
| "learning_rate": 8.97560398247424e-06, | |
| "loss": 0.8941, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.7562617659568787, | |
| "learning_rate": 8.961448216775955e-06, | |
| "loss": 0.8749, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.868, | |
| "grad_norm": 0.5686541795730591, | |
| "learning_rate": 8.947206639596346e-06, | |
| "loss": 0.8698, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.872, | |
| "grad_norm": 0.7385779619216919, | |
| "learning_rate": 8.932879559431392e-06, | |
| "loss": 0.8785, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.876, | |
| "grad_norm": 0.6453441977500916, | |
| "learning_rate": 8.9184672866292e-06, | |
| "loss": 0.8544, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.8072611093521118, | |
| "learning_rate": 8.903970133383297e-06, | |
| "loss": 0.8718, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.884, | |
| "grad_norm": 0.6873461008071899, | |
| "learning_rate": 8.889388413725857e-06, | |
| "loss": 0.8845, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.888, | |
| "grad_norm": 0.7314361929893494, | |
| "learning_rate": 8.874722443520898e-06, | |
| "loss": 0.8734, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.892, | |
| "grad_norm": 0.582042932510376, | |
| "learning_rate": 8.859972540457451e-06, | |
| "loss": 0.8626, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.5957586169242859, | |
| "learning_rate": 8.845139024042664e-06, | |
| "loss": 0.8909, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.684726357460022, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.8861, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.904, | |
| "grad_norm": 0.6183290481567383, | |
| "learning_rate": 8.815222438236726e-06, | |
| "loss": 0.8782, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.908, | |
| "grad_norm": 0.6130846738815308, | |
| "learning_rate": 8.800140016888009e-06, | |
| "loss": 0.9471, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 0.6433553099632263, | |
| "learning_rate": 8.784975278258783e-06, | |
| "loss": 0.8722, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.916, | |
| "grad_norm": 0.5252291560173035, | |
| "learning_rate": 8.769728550842217e-06, | |
| "loss": 0.8919, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.6188841462135315, | |
| "learning_rate": 8.754400164907496e-06, | |
| "loss": 0.8262, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.924, | |
| "grad_norm": 0.6740955114364624, | |
| "learning_rate": 8.73899045249266e-06, | |
| "loss": 0.8602, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.928, | |
| "grad_norm": 0.5842002630233765, | |
| "learning_rate": 8.723499747397415e-06, | |
| "loss": 0.8629, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.932, | |
| "grad_norm": 0.6343997120857239, | |
| "learning_rate": 8.707928385175898e-06, | |
| "loss": 0.8843, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.936, | |
| "grad_norm": 0.5856566429138184, | |
| "learning_rate": 8.692276703129421e-06, | |
| "loss": 0.8469, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.5288295745849609, | |
| "learning_rate": 8.676545040299145e-06, | |
| "loss": 0.8523, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.944, | |
| "grad_norm": 0.5988165736198425, | |
| "learning_rate": 8.660733737458751e-06, | |
| "loss": 0.8734, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.948, | |
| "grad_norm": 0.5942241549491882, | |
| "learning_rate": 8.644843137107058e-06, | |
| "loss": 0.872, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.952, | |
| "grad_norm": 0.6277519464492798, | |
| "learning_rate": 8.628873583460593e-06, | |
| "loss": 0.8441, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.956, | |
| "grad_norm": 0.6779516339302063, | |
| "learning_rate": 8.61282542244614e-06, | |
| "loss": 0.8928, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.6148802638053894, | |
| "learning_rate": 8.596699001693257e-06, | |
| "loss": 0.8329, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.964, | |
| "grad_norm": 0.6112699508666992, | |
| "learning_rate": 8.580494670526725e-06, | |
| "loss": 0.892, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.968, | |
| "grad_norm": 0.6572790741920471, | |
| "learning_rate": 8.564212779959003e-06, | |
| "loss": 0.9021, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.972, | |
| "grad_norm": 0.6486849784851074, | |
| "learning_rate": 8.547853682682605e-06, | |
| "loss": 0.8855, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.976, | |
| "grad_norm": 0.5732429027557373, | |
| "learning_rate": 8.531417733062476e-06, | |
| "loss": 0.8609, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.6210253834724426, | |
| "learning_rate": 8.51490528712831e-06, | |
| "loss": 0.8396, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.984, | |
| "grad_norm": 0.5373111963272095, | |
| "learning_rate": 8.498316702566828e-06, | |
| "loss": 0.8544, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.988, | |
| "grad_norm": 0.5937831401824951, | |
| "learning_rate": 8.481652338714048e-06, | |
| "loss": 0.8283, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.992, | |
| "grad_norm": 0.7062596082687378, | |
| "learning_rate": 8.464912556547486e-06, | |
| "loss": 0.8407, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.996, | |
| "grad_norm": 0.7154653668403625, | |
| "learning_rate": 8.44809771867835e-06, | |
| "loss": 0.8549, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.734148383140564, | |
| "learning_rate": 8.43120818934367e-06, | |
| "loss": 0.87, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.004, | |
| "grad_norm": 0.6793926358222961, | |
| "learning_rate": 8.414244334398418e-06, | |
| "loss": 0.7984, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.008, | |
| "grad_norm": 0.6782513856887817, | |
| "learning_rate": 8.397206521307584e-06, | |
| "loss": 0.8299, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.012, | |
| "grad_norm": 0.7735963463783264, | |
| "learning_rate": 8.380095119138209e-06, | |
| "loss": 0.8501, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.016, | |
| "grad_norm": 0.7407426238059998, | |
| "learning_rate": 8.362910498551402e-06, | |
| "loss": 0.8452, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.7997364401817322, | |
| "learning_rate": 8.345653031794292e-06, | |
| "loss": 0.8553, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.7399711608886719, | |
| "learning_rate": 8.328323092691985e-06, | |
| "loss": 0.7941, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.028, | |
| "grad_norm": 0.6705906987190247, | |
| "learning_rate": 8.310921056639451e-06, | |
| "loss": 0.8211, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.032, | |
| "grad_norm": 0.6934743523597717, | |
| "learning_rate": 8.293447300593402e-06, | |
| "loss": 0.8224, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.036, | |
| "grad_norm": 0.611541211605072, | |
| "learning_rate": 8.275902203064125e-06, | |
| "loss": 0.8435, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.7381828427314758, | |
| "learning_rate": 8.258286144107277e-06, | |
| "loss": 0.8238, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.044, | |
| "grad_norm": 0.6333577036857605, | |
| "learning_rate": 8.240599505315656e-06, | |
| "loss": 0.8142, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.048, | |
| "grad_norm": 0.7947782874107361, | |
| "learning_rate": 8.222842669810936e-06, | |
| "loss": 0.8364, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.052, | |
| "grad_norm": 0.6043962836265564, | |
| "learning_rate": 8.205016022235368e-06, | |
| "loss": 0.8271, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.056, | |
| "grad_norm": 0.6960867047309875, | |
| "learning_rate": 8.18711994874345e-06, | |
| "loss": 0.8077, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 0.60722815990448, | |
| "learning_rate": 8.16915483699355e-06, | |
| "loss": 0.7997, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.064, | |
| "grad_norm": 0.6178349852561951, | |
| "learning_rate": 8.151121076139534e-06, | |
| "loss": 0.8704, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.068, | |
| "grad_norm": 0.6908637881278992, | |
| "learning_rate": 8.133019056822303e-06, | |
| "loss": 0.8778, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.072, | |
| "grad_norm": 0.5272776484489441, | |
| "learning_rate": 8.11484917116136e-06, | |
| "loss": 0.8283, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.076, | |
| "grad_norm": 0.5782445073127747, | |
| "learning_rate": 8.096611812746302e-06, | |
| "loss": 0.8078, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.6523330211639404, | |
| "learning_rate": 8.078307376628292e-06, | |
| "loss": 0.8318, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.084, | |
| "grad_norm": 0.4992999732494354, | |
| "learning_rate": 8.059936259311514e-06, | |
| "loss": 0.7898, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.6073350310325623, | |
| "learning_rate": 8.041498858744572e-06, | |
| "loss": 0.8512, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.092, | |
| "grad_norm": 0.6151973605155945, | |
| "learning_rate": 8.022995574311876e-06, | |
| "loss": 0.8189, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.096, | |
| "grad_norm": 0.5909215211868286, | |
| "learning_rate": 8.004426806824985e-06, | |
| "loss": 0.7754, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.6663960814476013, | |
| "learning_rate": 7.985792958513932e-06, | |
| "loss": 0.7863, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.104, | |
| "grad_norm": 0.5585909485816956, | |
| "learning_rate": 7.967094433018508e-06, | |
| "loss": 0.8159, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.108, | |
| "grad_norm": 0.6069695949554443, | |
| "learning_rate": 7.948331635379517e-06, | |
| "loss": 0.8046, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.112, | |
| "grad_norm": 0.6288580894470215, | |
| "learning_rate": 7.929504972030003e-06, | |
| "loss": 0.8191, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.116, | |
| "grad_norm": 0.5330884456634521, | |
| "learning_rate": 7.910614850786448e-06, | |
| "loss": 0.8353, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.5828813314437866, | |
| "learning_rate": 7.891661680839932e-06, | |
| "loss": 0.8404, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.124, | |
| "grad_norm": 0.6435533165931702, | |
| "learning_rate": 7.872645872747281e-06, | |
| "loss": 0.8151, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.1280000000000001, | |
| "grad_norm": 0.7021773457527161, | |
| "learning_rate": 7.85356783842216e-06, | |
| "loss": 0.8314, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.1320000000000001, | |
| "grad_norm": 0.5960518717765808, | |
| "learning_rate": 7.834427991126155e-06, | |
| "loss": 0.7978, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.1360000000000001, | |
| "grad_norm": 0.6122227311134338, | |
| "learning_rate": 7.815226745459831e-06, | |
| "loss": 0.8245, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.1400000000000001, | |
| "grad_norm": 0.5207979679107666, | |
| "learning_rate": 7.795964517353734e-06, | |
| "loss": 0.8185, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.144, | |
| "grad_norm": 0.5806198716163635, | |
| "learning_rate": 7.776641724059398e-06, | |
| "loss": 0.8114, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.148, | |
| "grad_norm": 0.6067046523094177, | |
| "learning_rate": 7.757258784140286e-06, | |
| "loss": 0.8145, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.5425034165382385, | |
| "learning_rate": 7.737816117462752e-06, | |
| "loss": 0.8004, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.156, | |
| "grad_norm": 0.5887053608894348, | |
| "learning_rate": 7.718314145186918e-06, | |
| "loss": 0.8432, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 0.5984364151954651, | |
| "learning_rate": 7.698753289757565e-06, | |
| "loss": 0.8205, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.164, | |
| "grad_norm": 0.6469537615776062, | |
| "learning_rate": 7.679133974894984e-06, | |
| "loss": 0.82, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.168, | |
| "grad_norm": 0.5600078701972961, | |
| "learning_rate": 7.65945662558579e-06, | |
| "loss": 0.8316, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.172, | |
| "grad_norm": 0.5654756426811218, | |
| "learning_rate": 7.639721668073718e-06, | |
| "loss": 0.8174, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.176, | |
| "grad_norm": 0.7125182151794434, | |
| "learning_rate": 7.619929529850397e-06, | |
| "loss": 0.813, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 0.5197665095329285, | |
| "learning_rate": 7.600080639646077e-06, | |
| "loss": 0.8239, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.184, | |
| "grad_norm": 0.6301279664039612, | |
| "learning_rate": 7.580175427420358e-06, | |
| "loss": 0.8434, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.188, | |
| "grad_norm": 0.5137627124786377, | |
| "learning_rate": 7.560214324352858e-06, | |
| "loss": 0.8076, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.192, | |
| "grad_norm": 0.6159291863441467, | |
| "learning_rate": 7.54019776283389e-06, | |
| "loss": 0.8301, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.196, | |
| "grad_norm": 0.68797367811203, | |
| "learning_rate": 7.520126176455084e-06, | |
| "loss": 0.8249, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.5145540237426758, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.7956, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.204, | |
| "grad_norm": 0.5432578921318054, | |
| "learning_rate": 7.479819669434712e-06, | |
| "loss": 0.8577, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.208, | |
| "grad_norm": 0.5872198343276978, | |
| "learning_rate": 7.459585621898353e-06, | |
| "loss": 0.8661, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.212, | |
| "grad_norm": 0.5940935015678406, | |
| "learning_rate": 7.4392982956936644e-06, | |
| "loss": 0.8543, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.5254611372947693, | |
| "learning_rate": 7.418958130277483e-06, | |
| "loss": 0.8148, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 0.5751111507415771, | |
| "learning_rate": 7.398565566251232e-06, | |
| "loss": 0.8162, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.224, | |
| "grad_norm": 0.5236274003982544, | |
| "learning_rate": 7.378121045351378e-06, | |
| "loss": 0.8242, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.228, | |
| "grad_norm": 0.5233936309814453, | |
| "learning_rate": 7.357625010439853e-06, | |
| "loss": 0.8429, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.232, | |
| "grad_norm": 0.6502784490585327, | |
| "learning_rate": 7.337077905494472e-06, | |
| "loss": 0.8413, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.236, | |
| "grad_norm": 0.5654633641242981, | |
| "learning_rate": 7.31648017559931e-06, | |
| "loss": 0.8431, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.6566516757011414, | |
| "learning_rate": 7.295832266935059e-06, | |
| "loss": 0.8634, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.244, | |
| "grad_norm": 0.6114216446876526, | |
| "learning_rate": 7.275134626769369e-06, | |
| "loss": 0.8387, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.248, | |
| "grad_norm": 0.668125331401825, | |
| "learning_rate": 7.254387703447154e-06, | |
| "loss": 0.8087, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.252, | |
| "grad_norm": 0.6646181344985962, | |
| "learning_rate": 7.233591946380884e-06, | |
| "loss": 0.8138, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.256, | |
| "grad_norm": 0.5681068301200867, | |
| "learning_rate": 7.212747806040845e-06, | |
| "loss": 0.8258, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 0.6887611746788025, | |
| "learning_rate": 7.191855733945388e-06, | |
| "loss": 0.8444, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.264, | |
| "grad_norm": 0.707757294178009, | |
| "learning_rate": 7.170916182651141e-06, | |
| "loss": 0.8155, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.268, | |
| "grad_norm": 0.6170247793197632, | |
| "learning_rate": 7.149929605743214e-06, | |
| "loss": 0.7777, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.272, | |
| "grad_norm": 0.9348410367965698, | |
| "learning_rate": 7.128896457825364e-06, | |
| "loss": 0.7944, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.276, | |
| "grad_norm": 0.6176925301551819, | |
| "learning_rate": 7.107817194510157e-06, | |
| "loss": 0.8503, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.6812437176704407, | |
| "learning_rate": 7.08669227240909e-06, | |
| "loss": 0.8266, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.284, | |
| "grad_norm": 0.7825397253036499, | |
| "learning_rate": 7.06552214912271e-06, | |
| "loss": 0.8107, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.288, | |
| "grad_norm": 0.583035409450531, | |
| "learning_rate": 7.04430728323069e-06, | |
| "loss": 0.7889, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.292, | |
| "grad_norm": 0.8215271830558777, | |
| "learning_rate": 7.023048134281907e-06, | |
| "loss": 0.814, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.296, | |
| "grad_norm": 0.6114867329597473, | |
| "learning_rate": 7.0017451627844765e-06, | |
| "loss": 0.8144, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.6182892322540283, | |
| "learning_rate": 6.980398830195785e-06, | |
| "loss": 0.816, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.304, | |
| "grad_norm": 0.7972235679626465, | |
| "learning_rate": 6.959009598912493e-06, | |
| "loss": 0.8439, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.308, | |
| "grad_norm": 0.5884739756584167, | |
| "learning_rate": 6.9375779322605154e-06, | |
| "loss": 0.8244, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.312, | |
| "grad_norm": 0.81697678565979, | |
| "learning_rate": 6.916104294484988e-06, | |
| "loss": 0.851, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.316, | |
| "grad_norm": 0.6550260782241821, | |
| "learning_rate": 6.8945891507402075e-06, | |
| "loss": 0.7992, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.7166432738304138, | |
| "learning_rate": 6.873032967079562e-06, | |
| "loss": 0.8443, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.324, | |
| "grad_norm": 0.6055835485458374, | |
| "learning_rate": 6.851436210445427e-06, | |
| "loss": 0.8277, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.328, | |
| "grad_norm": 0.5842556953430176, | |
| "learning_rate": 6.829799348659061e-06, | |
| "loss": 0.8135, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.332, | |
| "grad_norm": 0.7298783659934998, | |
| "learning_rate": 6.808122850410461e-06, | |
| "loss": 0.8288, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.336, | |
| "grad_norm": 0.5614907145500183, | |
| "learning_rate": 6.7864071852482205e-06, | |
| "loss": 0.8283, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 0.7044088840484619, | |
| "learning_rate": 6.7646528235693445e-06, | |
| "loss": 0.8464, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.7158640027046204, | |
| "learning_rate": 6.7428602366090764e-06, | |
| "loss": 0.8347, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.3479999999999999, | |
| "grad_norm": 0.5172532200813293, | |
| "learning_rate": 6.721029896430678e-06, | |
| "loss": 0.8384, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.3519999999999999, | |
| "grad_norm": 0.6013163924217224, | |
| "learning_rate": 6.699162275915208e-06, | |
| "loss": 0.8151, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.3559999999999999, | |
| "grad_norm": 0.5935965180397034, | |
| "learning_rate": 6.677257848751276e-06, | |
| "loss": 0.832, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.3599999999999999, | |
| "grad_norm": 0.5422365069389343, | |
| "learning_rate": 6.655317089424791e-06, | |
| "loss": 0.7989, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.3639999999999999, | |
| "grad_norm": 0.5724878907203674, | |
| "learning_rate": 6.633340473208673e-06, | |
| "loss": 0.8106, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.3679999999999999, | |
| "grad_norm": 0.5668215155601501, | |
| "learning_rate": 6.611328476152557e-06, | |
| "loss": 0.8197, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.3719999999999999, | |
| "grad_norm": 0.6239653825759888, | |
| "learning_rate": 6.58928157507249e-06, | |
| "loss": 0.836, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.376, | |
| "grad_norm": 0.561103105545044, | |
| "learning_rate": 6.567200247540599e-06, | |
| "loss": 0.8207, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 0.5621249079704285, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.8452, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.384, | |
| "grad_norm": 0.5491905808448792, | |
| "learning_rate": 6.522936227128139e-06, | |
| "loss": 0.7999, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.388, | |
| "grad_norm": 0.5302982330322266, | |
| "learning_rate": 6.500754493079029e-06, | |
| "loss": 0.8387, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.392, | |
| "grad_norm": 0.49233949184417725, | |
| "learning_rate": 6.4785402502202345e-06, | |
| "loss": 0.8127, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.396, | |
| "grad_norm": 0.5672960877418518, | |
| "learning_rate": 6.456293979748778e-06, | |
| "loss": 0.8312, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.5462888479232788, | |
| "learning_rate": 6.434016163555452e-06, | |
| "loss": 0.818, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.404, | |
| "grad_norm": 0.5679849982261658, | |
| "learning_rate": 6.411707284214384e-06, | |
| "loss": 0.8109, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.6677523255348206, | |
| "learning_rate": 6.389367824972575e-06, | |
| "loss": 0.8329, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.412, | |
| "grad_norm": 0.5302258729934692, | |
| "learning_rate": 6.366998269739442e-06, | |
| "loss": 0.8209, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.416, | |
| "grad_norm": 0.6249000430107117, | |
| "learning_rate": 6.344599103076329e-06, | |
| "loss": 0.8423, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 0.6532145142555237, | |
| "learning_rate": 6.322170810186013e-06, | |
| "loss": 0.8315, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.424, | |
| "grad_norm": 0.5602342486381531, | |
| "learning_rate": 6.299713876902188e-06, | |
| "loss": 0.8397, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.428, | |
| "grad_norm": 0.7034306526184082, | |
| "learning_rate": 6.277228789678953e-06, | |
| "loss": 0.8165, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.432, | |
| "grad_norm": 0.6753026843070984, | |
| "learning_rate": 6.254716035580264e-06, | |
| "loss": 0.8237, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.436, | |
| "grad_norm": 0.6752943396568298, | |
| "learning_rate": 6.23217610226939e-06, | |
| "loss": 0.8345, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.6669530272483826, | |
| "learning_rate": 6.209609477998339e-06, | |
| "loss": 0.7928, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.444, | |
| "grad_norm": 0.667064368724823, | |
| "learning_rate": 6.187016651597299e-06, | |
| "loss": 0.8073, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.448, | |
| "grad_norm": 0.7478964924812317, | |
| "learning_rate": 6.16439811246403e-06, | |
| "loss": 0.8672, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.452, | |
| "grad_norm": 0.5698713064193726, | |
| "learning_rate": 6.141754350553279e-06, | |
| "loss": 0.8226, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.456, | |
| "grad_norm": 0.6351299285888672, | |
| "learning_rate": 6.119085856366158e-06, | |
| "loss": 0.7964, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 0.697109043598175, | |
| "learning_rate": 6.0963931209395165e-06, | |
| "loss": 0.7774, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.464, | |
| "grad_norm": 0.6185441017150879, | |
| "learning_rate": 6.073676635835317e-06, | |
| "loss": 0.8087, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.468, | |
| "grad_norm": 0.552899181842804, | |
| "learning_rate": 6.05093689312997e-06, | |
| "loss": 0.8159, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.6541306376457214, | |
| "learning_rate": 6.028174385403693e-06, | |
| "loss": 0.8259, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.476, | |
| "grad_norm": 0.5495116114616394, | |
| "learning_rate": 6.005389605729824e-06, | |
| "loss": 0.8156, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.6034694910049438, | |
| "learning_rate": 5.982583047664151e-06, | |
| "loss": 0.8131, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.484, | |
| "grad_norm": 0.6365712881088257, | |
| "learning_rate": 5.9597552052342174e-06, | |
| "loss": 0.8305, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.488, | |
| "grad_norm": 0.5371845960617065, | |
| "learning_rate": 5.936906572928625e-06, | |
| "loss": 0.8154, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.492, | |
| "grad_norm": 0.6355674266815186, | |
| "learning_rate": 5.914037645686308e-06, | |
| "loss": 0.8203, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.496, | |
| "grad_norm": 0.5485637187957764, | |
| "learning_rate": 5.891148918885834e-06, | |
| "loss": 0.8417, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.5740164518356323, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.8497, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.504, | |
| "grad_norm": 0.5761863589286804, | |
| "learning_rate": 5.84531405025837e-06, | |
| "loss": 0.7931, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.508, | |
| "grad_norm": 0.6561781167984009, | |
| "learning_rate": 5.8223689012899945e-06, | |
| "loss": 0.7777, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.512, | |
| "grad_norm": 0.5368021726608276, | |
| "learning_rate": 5.799405938459175e-06, | |
| "loss": 0.8256, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.516, | |
| "grad_norm": 0.5972307324409485, | |
| "learning_rate": 5.776425659181438e-06, | |
| "loss": 0.8315, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.4860232174396515, | |
| "learning_rate": 5.753428561247416e-06, | |
| "loss": 0.8276, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.524, | |
| "grad_norm": 0.5551620721817017, | |
| "learning_rate": 5.730415142812059e-06, | |
| "loss": 0.8376, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.528, | |
| "grad_norm": 0.6248589158058167, | |
| "learning_rate": 5.707385902383845e-06, | |
| "loss": 0.8344, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.532, | |
| "grad_norm": 0.519512951374054, | |
| "learning_rate": 5.684341338813986e-06, | |
| "loss": 0.8016, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.5413280725479126, | |
| "learning_rate": 5.661281951285613e-06, | |
| "loss": 0.8216, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 0.5940418243408203, | |
| "learning_rate": 5.638208239302975e-06, | |
| "loss": 0.8302, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.544, | |
| "grad_norm": 0.5316104888916016, | |
| "learning_rate": 5.615120702680604e-06, | |
| "loss": 0.8124, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.548, | |
| "grad_norm": 0.5062645673751831, | |
| "learning_rate": 5.592019841532507e-06, | |
| "loss": 0.7866, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.552, | |
| "grad_norm": 0.5303630828857422, | |
| "learning_rate": 5.568906156261309e-06, | |
| "loss": 0.8453, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.556, | |
| "grad_norm": 0.5888771414756775, | |
| "learning_rate": 5.54578014754744e-06, | |
| "loss": 0.8188, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.5190044641494751, | |
| "learning_rate": 5.522642316338268e-06, | |
| "loss": 0.8008, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.564, | |
| "grad_norm": 0.537983238697052, | |
| "learning_rate": 5.499493163837258e-06, | |
| "loss": 0.876, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.568, | |
| "grad_norm": 0.5991531610488892, | |
| "learning_rate": 5.476333191493108e-06, | |
| "loss": 0.7961, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.572, | |
| "grad_norm": 0.48126405477523804, | |
| "learning_rate": 5.453162900988902e-06, | |
| "loss": 0.8384, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.576, | |
| "grad_norm": 0.5848698616027832, | |
| "learning_rate": 5.429982794231221e-06, | |
| "loss": 0.8287, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 0.5644991397857666, | |
| "learning_rate": 5.406793373339292e-06, | |
| "loss": 0.825, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.584, | |
| "grad_norm": 0.49802395701408386, | |
| "learning_rate": 5.383595140634093e-06, | |
| "loss": 0.8195, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.588, | |
| "grad_norm": 0.5119688510894775, | |
| "learning_rate": 5.360388598627487e-06, | |
| "loss": 0.7859, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.592, | |
| "grad_norm": 0.5061742067337036, | |
| "learning_rate": 5.337174250011326e-06, | |
| "loss": 0.7739, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.596, | |
| "grad_norm": 0.5984253287315369, | |
| "learning_rate": 5.3139525976465675e-06, | |
| "loss": 0.8241, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.5524809956550598, | |
| "learning_rate": 5.290724144552379e-06, | |
| "loss": 0.8186, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.604, | |
| "grad_norm": 0.5304140448570251, | |
| "learning_rate": 5.267489393895247e-06, | |
| "loss": 0.7989, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.608, | |
| "grad_norm": 0.5568251609802246, | |
| "learning_rate": 5.244248848978067e-06, | |
| "loss": 0.8459, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.612, | |
| "grad_norm": 0.47173407673835754, | |
| "learning_rate": 5.221003013229253e-06, | |
| "loss": 0.8308, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.616, | |
| "grad_norm": 0.49636995792388916, | |
| "learning_rate": 5.197752390191827e-06, | |
| "loss": 0.8204, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 0.5288768410682678, | |
| "learning_rate": 5.174497483512506e-06, | |
| "loss": 0.8598, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.624, | |
| "grad_norm": 0.4819859564304352, | |
| "learning_rate": 5.151238796930804e-06, | |
| "loss": 0.8271, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.6280000000000001, | |
| "grad_norm": 0.5618477463722229, | |
| "learning_rate": 5.127976834268112e-06, | |
| "loss": 0.8019, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.6320000000000001, | |
| "grad_norm": 0.5314104557037354, | |
| "learning_rate": 5.1047120994167855e-06, | |
| "loss": 0.8342, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.6360000000000001, | |
| "grad_norm": 0.47630175948143005, | |
| "learning_rate": 5.081445096329229e-06, | |
| "loss": 0.8294, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.6400000000000001, | |
| "grad_norm": 0.5262562036514282, | |
| "learning_rate": 5.0581763290069865e-06, | |
| "loss": 0.797, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.6440000000000001, | |
| "grad_norm": 0.5042218565940857, | |
| "learning_rate": 5.034906301489808e-06, | |
| "loss": 0.7843, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.6480000000000001, | |
| "grad_norm": 0.6031314730644226, | |
| "learning_rate": 5.011635517844753e-06, | |
| "loss": 0.7805, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.6520000000000001, | |
| "grad_norm": 0.5020240545272827, | |
| "learning_rate": 4.988364482155249e-06, | |
| "loss": 0.8101, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.6560000000000001, | |
| "grad_norm": 0.4847033619880676, | |
| "learning_rate": 4.965093698510192e-06, | |
| "loss": 0.7977, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.6600000000000001, | |
| "grad_norm": 0.5217180252075195, | |
| "learning_rate": 4.941823670993016e-06, | |
| "loss": 0.8385, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.4937134385108948, | |
| "learning_rate": 4.9185549036707715e-06, | |
| "loss": 0.8213, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.6680000000000001, | |
| "grad_norm": 0.5219268202781677, | |
| "learning_rate": 4.895287900583216e-06, | |
| "loss": 0.8416, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.6720000000000002, | |
| "grad_norm": 0.6660354137420654, | |
| "learning_rate": 4.87202316573189e-06, | |
| "loss": 0.809, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.6760000000000002, | |
| "grad_norm": 0.5231005549430847, | |
| "learning_rate": 4.8487612030691975e-06, | |
| "loss": 0.8451, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 0.49823904037475586, | |
| "learning_rate": 4.825502516487497e-06, | |
| "loss": 0.8033, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.6840000000000002, | |
| "grad_norm": 0.5597706437110901, | |
| "learning_rate": 4.802247609808175e-06, | |
| "loss": 0.7785, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.688, | |
| "grad_norm": 0.5253022313117981, | |
| "learning_rate": 4.778996986770747e-06, | |
| "loss": 0.8273, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.692, | |
| "grad_norm": 0.4566793739795685, | |
| "learning_rate": 4.755751151021934e-06, | |
| "loss": 0.8136, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.696, | |
| "grad_norm": 0.5758362412452698, | |
| "learning_rate": 4.732510606104754e-06, | |
| "loss": 0.7944, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.5128145813941956, | |
| "learning_rate": 4.7092758554476215e-06, | |
| "loss": 0.7787, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.704, | |
| "grad_norm": 0.4326518177986145, | |
| "learning_rate": 4.686047402353433e-06, | |
| "loss": 0.8449, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.708, | |
| "grad_norm": 0.5038052201271057, | |
| "learning_rate": 4.662825749988675e-06, | |
| "loss": 0.8656, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.712, | |
| "grad_norm": 0.48481109738349915, | |
| "learning_rate": 4.639611401372514e-06, | |
| "loss": 0.8263, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.716, | |
| "grad_norm": 0.45356905460357666, | |
| "learning_rate": 4.6164048593659076e-06, | |
| "loss": 0.7882, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.5066087245941162, | |
| "learning_rate": 4.59320662666071e-06, | |
| "loss": 0.8271, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.724, | |
| "grad_norm": 0.5313457250595093, | |
| "learning_rate": 4.570017205768779e-06, | |
| "loss": 0.8182, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.5594462752342224, | |
| "learning_rate": 4.546837099011101e-06, | |
| "loss": 0.7873, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.732, | |
| "grad_norm": 0.4426927864551544, | |
| "learning_rate": 4.523666808506893e-06, | |
| "loss": 0.8012, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.736, | |
| "grad_norm": 0.5021507740020752, | |
| "learning_rate": 4.500506836162746e-06, | |
| "loss": 0.8149, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 0.46885576844215393, | |
| "learning_rate": 4.477357683661734e-06, | |
| "loss": 0.8252, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.744, | |
| "grad_norm": 0.4938061833381653, | |
| "learning_rate": 4.45421985245256e-06, | |
| "loss": 0.7868, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.748, | |
| "grad_norm": 0.4414080083370209, | |
| "learning_rate": 4.431093843738693e-06, | |
| "loss": 0.8138, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.752, | |
| "grad_norm": 0.4636024236679077, | |
| "learning_rate": 4.4079801584674955e-06, | |
| "loss": 0.8376, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.756, | |
| "grad_norm": 0.46044057607650757, | |
| "learning_rate": 4.384879297319398e-06, | |
| "loss": 0.8347, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.45228543877601624, | |
| "learning_rate": 4.361791760697027e-06, | |
| "loss": 0.8501, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.764, | |
| "grad_norm": 0.468711793422699, | |
| "learning_rate": 4.3387180487143875e-06, | |
| "loss": 0.8194, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.768, | |
| "grad_norm": 0.515216052532196, | |
| "learning_rate": 4.315658661186016e-06, | |
| "loss": 0.8352, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.772, | |
| "grad_norm": 0.5516295433044434, | |
| "learning_rate": 4.2926140976161555e-06, | |
| "loss": 0.8033, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.776, | |
| "grad_norm": 0.4367079734802246, | |
| "learning_rate": 4.269584857187942e-06, | |
| "loss": 0.7962, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 0.502176821231842, | |
| "learning_rate": 4.246571438752585e-06, | |
| "loss": 0.8152, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.784, | |
| "grad_norm": 0.5594773888587952, | |
| "learning_rate": 4.2235743408185635e-06, | |
| "loss": 0.8351, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.788, | |
| "grad_norm": 0.5325698852539062, | |
| "learning_rate": 4.200594061540827e-06, | |
| "loss": 0.8292, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.48320385813713074, | |
| "learning_rate": 4.1776310987100054e-06, | |
| "loss": 0.8349, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.796, | |
| "grad_norm": 0.6167989373207092, | |
| "learning_rate": 4.154685949741631e-06, | |
| "loss": 0.841, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.5394158363342285, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.8478, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.804, | |
| "grad_norm": 0.5120943188667297, | |
| "learning_rate": 4.108851081114169e-06, | |
| "loss": 0.842, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.808, | |
| "grad_norm": 0.5258692502975464, | |
| "learning_rate": 4.0859623543136935e-06, | |
| "loss": 0.8085, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.812, | |
| "grad_norm": 0.5280582308769226, | |
| "learning_rate": 4.063093427071376e-06, | |
| "loss": 0.8226, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.8159999999999998, | |
| "grad_norm": 0.5288624167442322, | |
| "learning_rate": 4.040244794765783e-06, | |
| "loss": 0.7907, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.8199999999999998, | |
| "grad_norm": 0.5390572547912598, | |
| "learning_rate": 4.017416952335849e-06, | |
| "loss": 0.8148, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.8239999999999998, | |
| "grad_norm": 0.5915282964706421, | |
| "learning_rate": 3.994610394270178e-06, | |
| "loss": 0.8446, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.8279999999999998, | |
| "grad_norm": 0.5086544752120972, | |
| "learning_rate": 3.971825614596308e-06, | |
| "loss": 0.8151, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.8319999999999999, | |
| "grad_norm": 0.5306235551834106, | |
| "learning_rate": 3.949063106870031e-06, | |
| "loss": 0.8365, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.8359999999999999, | |
| "grad_norm": 0.4914201498031616, | |
| "learning_rate": 3.926323364164684e-06, | |
| "loss": 0.8226, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.8399999999999999, | |
| "grad_norm": 0.5558998584747314, | |
| "learning_rate": 3.903606879060483e-06, | |
| "loss": 0.7785, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.8439999999999999, | |
| "grad_norm": 0.5253325700759888, | |
| "learning_rate": 3.880914143633844e-06, | |
| "loss": 0.7751, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.8479999999999999, | |
| "grad_norm": 0.539868175983429, | |
| "learning_rate": 3.8582456494467214e-06, | |
| "loss": 0.8252, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.8519999999999999, | |
| "grad_norm": 0.5210921168327332, | |
| "learning_rate": 3.835601887535971e-06, | |
| "loss": 0.782, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.43055686354637146, | |
| "learning_rate": 3.812983348402703e-06, | |
| "loss": 0.7884, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.8599999999999999, | |
| "grad_norm": 0.4651299715042114, | |
| "learning_rate": 3.790390522001662e-06, | |
| "loss": 0.8269, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.8639999999999999, | |
| "grad_norm": 0.516089141368866, | |
| "learning_rate": 3.767823897730612e-06, | |
| "loss": 0.8175, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.8679999999999999, | |
| "grad_norm": 0.4905009865760803, | |
| "learning_rate": 3.745283964419736e-06, | |
| "loss": 0.8265, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.8719999999999999, | |
| "grad_norm": 0.4264768958091736, | |
| "learning_rate": 3.7227712103210485e-06, | |
| "loss": 0.8422, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.876, | |
| "grad_norm": 0.46796920895576477, | |
| "learning_rate": 3.700286123097814e-06, | |
| "loss": 0.8218, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.49499964714050293, | |
| "learning_rate": 3.6778291898139907e-06, | |
| "loss": 0.7906, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.884, | |
| "grad_norm": 0.4863964021205902, | |
| "learning_rate": 3.655400896923672e-06, | |
| "loss": 0.7864, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.888, | |
| "grad_norm": 0.49399423599243164, | |
| "learning_rate": 3.633001730260558e-06, | |
| "loss": 0.8228, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.892, | |
| "grad_norm": 0.4784768521785736, | |
| "learning_rate": 3.6106321750274275e-06, | |
| "loss": 0.816, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.896, | |
| "grad_norm": 0.4161972403526306, | |
| "learning_rate": 3.5882927157856175e-06, | |
| "loss": 0.8663, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.4483587145805359, | |
| "learning_rate": 3.5659838364445505e-06, | |
| "loss": 0.8209, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.904, | |
| "grad_norm": 0.47628068923950195, | |
| "learning_rate": 3.543706020251223e-06, | |
| "loss": 0.8279, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.908, | |
| "grad_norm": 0.4492845833301544, | |
| "learning_rate": 3.521459749779769e-06, | |
| "loss": 0.8156, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.912, | |
| "grad_norm": 0.463316410779953, | |
| "learning_rate": 3.4992455069209717e-06, | |
| "loss": 0.8152, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.916, | |
| "grad_norm": 0.4651520252227783, | |
| "learning_rate": 3.4770637728718608e-06, | |
| "loss": 0.832, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.4673428535461426, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.8177, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.924, | |
| "grad_norm": 0.49237164855003357, | |
| "learning_rate": 3.4327997524594026e-06, | |
| "loss": 0.8401, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.928, | |
| "grad_norm": 0.47739100456237793, | |
| "learning_rate": 3.4107184249275114e-06, | |
| "loss": 0.8577, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.932, | |
| "grad_norm": 0.48048797249794006, | |
| "learning_rate": 3.3886715238474454e-06, | |
| "loss": 0.8171, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.936, | |
| "grad_norm": 0.45767322182655334, | |
| "learning_rate": 3.3666595267913293e-06, | |
| "loss": 0.8573, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 0.4932113289833069, | |
| "learning_rate": 3.3446829105752103e-06, | |
| "loss": 0.8174, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.944, | |
| "grad_norm": 0.4785451292991638, | |
| "learning_rate": 3.322742151248726e-06, | |
| "loss": 0.8119, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.948, | |
| "grad_norm": 0.5046617388725281, | |
| "learning_rate": 3.3008377240847955e-06, | |
| "loss": 0.8051, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.952, | |
| "grad_norm": 0.5119112730026245, | |
| "learning_rate": 3.2789701035693242e-06, | |
| "loss": 0.8547, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.956, | |
| "grad_norm": 0.4656376838684082, | |
| "learning_rate": 3.2571397633909252e-06, | |
| "loss": 0.8522, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 0.5343756675720215, | |
| "learning_rate": 3.2353471764306567e-06, | |
| "loss": 0.8417, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.964, | |
| "grad_norm": 0.47745269536972046, | |
| "learning_rate": 3.2135928147517803e-06, | |
| "loss": 0.7664, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.968, | |
| "grad_norm": 0.4895247220993042, | |
| "learning_rate": 3.1918771495895395e-06, | |
| "loss": 0.8248, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.972, | |
| "grad_norm": 0.4918672740459442, | |
| "learning_rate": 3.1702006513409393e-06, | |
| "loss": 0.8189, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 1.976, | |
| "grad_norm": 0.46551230549812317, | |
| "learning_rate": 3.148563789554575e-06, | |
| "loss": 0.8193, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 0.5140730738639832, | |
| "learning_rate": 3.12696703292044e-06, | |
| "loss": 0.8297, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.4619219899177551, | |
| "learning_rate": 3.105410849259796e-06, | |
| "loss": 0.7984, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 1.988, | |
| "grad_norm": 0.49016299843788147, | |
| "learning_rate": 3.0838957055150136e-06, | |
| "loss": 0.8072, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 1.992, | |
| "grad_norm": 0.4515412747859955, | |
| "learning_rate": 3.0624220677394854e-06, | |
| "loss": 0.7919, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 1.996, | |
| "grad_norm": 0.4848708510398865, | |
| "learning_rate": 3.040990401087508e-06, | |
| "loss": 0.8406, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.4492388665676117, | |
| "learning_rate": 3.019601169804216e-06, | |
| "loss": 0.7982, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.004, | |
| "grad_norm": 0.4607851207256317, | |
| "learning_rate": 2.9982548372155264e-06, | |
| "loss": 0.7587, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 2.008, | |
| "grad_norm": 0.4731832444667816, | |
| "learning_rate": 2.9769518657180953e-06, | |
| "loss": 0.7893, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 2.012, | |
| "grad_norm": 0.4613538682460785, | |
| "learning_rate": 2.9556927167693107e-06, | |
| "loss": 0.782, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 2.016, | |
| "grad_norm": 0.4291563630104065, | |
| "learning_rate": 2.934477850877292e-06, | |
| "loss": 0.7916, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 0.41530513763427734, | |
| "learning_rate": 2.9133077275909112e-06, | |
| "loss": 0.7631, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 2.024, | |
| "grad_norm": 0.39784783124923706, | |
| "learning_rate": 2.892182805489846e-06, | |
| "loss": 0.7966, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 2.028, | |
| "grad_norm": 0.5020673274993896, | |
| "learning_rate": 2.871103542174637e-06, | |
| "loss": 0.81, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 2.032, | |
| "grad_norm": 0.5143479704856873, | |
| "learning_rate": 2.8500703942567874e-06, | |
| "loss": 0.7654, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 2.036, | |
| "grad_norm": 0.455581396818161, | |
| "learning_rate": 2.82908381734886e-06, | |
| "loss": 0.7834, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 0.44526249170303345, | |
| "learning_rate": 2.8081442660546126e-06, | |
| "loss": 0.8032, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.044, | |
| "grad_norm": 0.4740583896636963, | |
| "learning_rate": 2.7872521939591556e-06, | |
| "loss": 0.7803, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.4611470401287079, | |
| "learning_rate": 2.7664080536191178e-06, | |
| "loss": 0.7928, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 2.052, | |
| "grad_norm": 0.5031675696372986, | |
| "learning_rate": 2.7456122965528475e-06, | |
| "loss": 0.7821, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 2.056, | |
| "grad_norm": 0.4245532155036926, | |
| "learning_rate": 2.724865373230632e-06, | |
| "loss": 0.7755, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 0.48005154728889465, | |
| "learning_rate": 2.7041677330649408e-06, | |
| "loss": 0.7647, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 2.064, | |
| "grad_norm": 0.4330499768257141, | |
| "learning_rate": 2.683519824400693e-06, | |
| "loss": 0.7806, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 2.068, | |
| "grad_norm": 0.4752224087715149, | |
| "learning_rate": 2.662922094505529e-06, | |
| "loss": 0.7562, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 2.072, | |
| "grad_norm": 0.4366036057472229, | |
| "learning_rate": 2.6423749895601494e-06, | |
| "loss": 0.7736, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 2.076, | |
| "grad_norm": 0.4761313796043396, | |
| "learning_rate": 2.6218789546486235e-06, | |
| "loss": 0.7694, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.41412127017974854, | |
| "learning_rate": 2.601434433748771e-06, | |
| "loss": 0.7603, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.084, | |
| "grad_norm": 0.4283531606197357, | |
| "learning_rate": 2.581041869722519e-06, | |
| "loss": 0.8048, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 2.088, | |
| "grad_norm": 0.48063892126083374, | |
| "learning_rate": 2.560701704306336e-06, | |
| "loss": 0.754, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 2.092, | |
| "grad_norm": 0.47472384572029114, | |
| "learning_rate": 2.540414378101647e-06, | |
| "loss": 0.7941, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 2.096, | |
| "grad_norm": 0.43496283888816833, | |
| "learning_rate": 2.52018033056529e-06, | |
| "loss": 0.7982, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 0.46004152297973633, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.7704, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.104, | |
| "grad_norm": 0.4385717809200287, | |
| "learning_rate": 2.4798738235449164e-06, | |
| "loss": 0.7626, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 2.108, | |
| "grad_norm": 0.45860159397125244, | |
| "learning_rate": 2.4598022371661113e-06, | |
| "loss": 0.7781, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.4268695116043091, | |
| "learning_rate": 2.4397856756471435e-06, | |
| "loss": 0.7615, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.116, | |
| "grad_norm": 0.533136785030365, | |
| "learning_rate": 2.4198245725796427e-06, | |
| "loss": 0.781, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 0.4675978422164917, | |
| "learning_rate": 2.3999193603539234e-06, | |
| "loss": 0.7935, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.124, | |
| "grad_norm": 0.4748784899711609, | |
| "learning_rate": 2.380070470149605e-06, | |
| "loss": 0.7558, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 2.128, | |
| "grad_norm": 0.4731289744377136, | |
| "learning_rate": 2.3602783319262847e-06, | |
| "loss": 0.7801, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 2.132, | |
| "grad_norm": 0.4706327021121979, | |
| "learning_rate": 2.340543374414212e-06, | |
| "loss": 0.7615, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 2.136, | |
| "grad_norm": 0.5207027196884155, | |
| "learning_rate": 2.320866025105016e-06, | |
| "loss": 0.7831, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 0.41023868322372437, | |
| "learning_rate": 2.3012467102424373e-06, | |
| "loss": 0.7642, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 2.144, | |
| "grad_norm": 0.4167621433734894, | |
| "learning_rate": 2.2816858548130837e-06, | |
| "loss": 0.7502, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 2.148, | |
| "grad_norm": 0.4176595211029053, | |
| "learning_rate": 2.2621838825372496e-06, | |
| "loss": 0.7878, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 2.152, | |
| "grad_norm": 0.4613254964351654, | |
| "learning_rate": 2.2427412158597133e-06, | |
| "loss": 0.8084, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 2.156, | |
| "grad_norm": 0.4380951225757599, | |
| "learning_rate": 2.2233582759406065e-06, | |
| "loss": 0.7675, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.43744876980781555, | |
| "learning_rate": 2.204035482646267e-06, | |
| "loss": 0.7903, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.164, | |
| "grad_norm": 0.42547735571861267, | |
| "learning_rate": 2.184773254540169e-06, | |
| "loss": 0.7992, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 2.168, | |
| "grad_norm": 0.48103204369544983, | |
| "learning_rate": 2.165572008873845e-06, | |
| "loss": 0.7768, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 2.172, | |
| "grad_norm": 0.5394415855407715, | |
| "learning_rate": 2.146432161577842e-06, | |
| "loss": 0.7797, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.5370150208473206, | |
| "learning_rate": 2.12735412725272e-06, | |
| "loss": 0.7717, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 0.434938907623291, | |
| "learning_rate": 2.1083383191600676e-06, | |
| "loss": 0.752, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 2.184, | |
| "grad_norm": 0.4815244674682617, | |
| "learning_rate": 2.0893851492135536e-06, | |
| "loss": 0.7791, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 2.188, | |
| "grad_norm": 0.4812857210636139, | |
| "learning_rate": 2.0704950279699986e-06, | |
| "loss": 0.7951, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 2.192, | |
| "grad_norm": 0.49643102288246155, | |
| "learning_rate": 2.0516683646204836e-06, | |
| "loss": 0.7639, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 2.196, | |
| "grad_norm": 0.4618609845638275, | |
| "learning_rate": 2.0329055669814936e-06, | |
| "loss": 0.7593, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 0.41718965768814087, | |
| "learning_rate": 2.0142070414860704e-06, | |
| "loss": 0.8132, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.204, | |
| "grad_norm": 0.44279876351356506, | |
| "learning_rate": 1.9955731931750182e-06, | |
| "loss": 0.7773, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 2.208, | |
| "grad_norm": 0.4497832953929901, | |
| "learning_rate": 1.977004425688126e-06, | |
| "loss": 0.7781, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 2.212, | |
| "grad_norm": 0.45165765285491943, | |
| "learning_rate": 1.958501141255427e-06, | |
| "loss": 0.8189, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 2.216, | |
| "grad_norm": 0.482389897108078, | |
| "learning_rate": 1.9400637406884875e-06, | |
| "loss": 0.7639, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 0.4482229948043823, | |
| "learning_rate": 1.9216926233717087e-06, | |
| "loss": 0.767, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 2.224, | |
| "grad_norm": 0.4738655388355255, | |
| "learning_rate": 1.9033881872537009e-06, | |
| "loss": 0.7658, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 2.228, | |
| "grad_norm": 0.43093428015708923, | |
| "learning_rate": 1.88515082883864e-06, | |
| "loss": 0.7824, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 2.232, | |
| "grad_norm": 0.4479147791862488, | |
| "learning_rate": 1.8669809431776991e-06, | |
| "loss": 0.8217, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 2.2359999999999998, | |
| "grad_norm": 0.5110467076301575, | |
| "learning_rate": 1.8488789238604676e-06, | |
| "loss": 0.8179, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.4612603783607483, | |
| "learning_rate": 1.8308451630064484e-06, | |
| "loss": 0.8095, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.2439999999999998, | |
| "grad_norm": 0.4464091360569, | |
| "learning_rate": 1.8128800512565514e-06, | |
| "loss": 0.7891, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 2.248, | |
| "grad_norm": 0.43865087628364563, | |
| "learning_rate": 1.7949839777646327e-06, | |
| "loss": 0.805, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 2.252, | |
| "grad_norm": 0.4769790768623352, | |
| "learning_rate": 1.7771573301890666e-06, | |
| "loss": 0.7491, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 2.2560000000000002, | |
| "grad_norm": 0.42606115341186523, | |
| "learning_rate": 1.7594004946843458e-06, | |
| "loss": 0.7802, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 0.46678170561790466, | |
| "learning_rate": 1.7417138558927244e-06, | |
| "loss": 0.7608, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 2.2640000000000002, | |
| "grad_norm": 0.4411817789077759, | |
| "learning_rate": 1.7240977969358757e-06, | |
| "loss": 0.7541, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 2.268, | |
| "grad_norm": 0.463142991065979, | |
| "learning_rate": 1.7065526994065973e-06, | |
| "loss": 0.7967, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 2.2720000000000002, | |
| "grad_norm": 0.49492907524108887, | |
| "learning_rate": 1.6890789433605508e-06, | |
| "loss": 0.7817, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 2.276, | |
| "grad_norm": 0.48016270995140076, | |
| "learning_rate": 1.671676907308018e-06, | |
| "loss": 0.7586, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 2.2800000000000002, | |
| "grad_norm": 0.4515690803527832, | |
| "learning_rate": 1.6543469682057105e-06, | |
| "loss": 0.7612, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.284, | |
| "grad_norm": 0.4577561020851135, | |
| "learning_rate": 1.6370895014486e-06, | |
| "loss": 0.7828, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 2.288, | |
| "grad_norm": 0.47520819306373596, | |
| "learning_rate": 1.6199048808617896e-06, | |
| "loss": 0.7579, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 2.292, | |
| "grad_norm": 0.46000567078590393, | |
| "learning_rate": 1.6027934786924187e-06, | |
| "loss": 0.7288, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 2.296, | |
| "grad_norm": 0.46916693449020386, | |
| "learning_rate": 1.5857556656015837e-06, | |
| "loss": 0.7974, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 0.42662757635116577, | |
| "learning_rate": 1.5687918106563326e-06, | |
| "loss": 0.8026, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.4421575665473938, | |
| "learning_rate": 1.551902281321651e-06, | |
| "loss": 0.7705, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 2.308, | |
| "grad_norm": 0.40408822894096375, | |
| "learning_rate": 1.5350874434525142e-06, | |
| "loss": 0.79, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 2.312, | |
| "grad_norm": 0.46466535329818726, | |
| "learning_rate": 1.5183476612859538e-06, | |
| "loss": 0.7796, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 2.316, | |
| "grad_norm": 0.5628266930580139, | |
| "learning_rate": 1.5016832974331725e-06, | |
| "loss": 0.7863, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 0.41899147629737854, | |
| "learning_rate": 1.4850947128716914e-06, | |
| "loss": 0.7793, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.324, | |
| "grad_norm": 0.415682852268219, | |
| "learning_rate": 1.4685822669375239e-06, | |
| "loss": 0.7962, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 2.328, | |
| "grad_norm": 0.3926596939563751, | |
| "learning_rate": 1.4521463173173966e-06, | |
| "loss": 0.7752, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 2.332, | |
| "grad_norm": 0.4485792815685272, | |
| "learning_rate": 1.4357872200409988e-06, | |
| "loss": 0.7563, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 2.336, | |
| "grad_norm": 0.5413661599159241, | |
| "learning_rate": 1.4195053294732757e-06, | |
| "loss": 0.7666, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 0.4145437479019165, | |
| "learning_rate": 1.4033009983067454e-06, | |
| "loss": 0.8055, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.344, | |
| "grad_norm": 0.38938337564468384, | |
| "learning_rate": 1.3871745775538598e-06, | |
| "loss": 0.7531, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 2.348, | |
| "grad_norm": 0.4246900975704193, | |
| "learning_rate": 1.371126416539409e-06, | |
| "loss": 0.7978, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 2.352, | |
| "grad_norm": 0.48616209626197815, | |
| "learning_rate": 1.3551568628929434e-06, | |
| "loss": 0.7487, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 2.356, | |
| "grad_norm": 0.43307557702064514, | |
| "learning_rate": 1.339266262541249e-06, | |
| "loss": 0.7962, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 0.3837439715862274, | |
| "learning_rate": 1.3234549597008572e-06, | |
| "loss": 0.7901, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.364, | |
| "grad_norm": 0.4052961468696594, | |
| "learning_rate": 1.3077232968705805e-06, | |
| "loss": 0.7982, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.4140758216381073, | |
| "learning_rate": 1.2920716148241036e-06, | |
| "loss": 0.7654, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 2.372, | |
| "grad_norm": 0.3948996067047119, | |
| "learning_rate": 1.2765002526025871e-06, | |
| "loss": 0.7895, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 2.376, | |
| "grad_norm": 0.3942357003688812, | |
| "learning_rate": 1.2610095475073415e-06, | |
| "loss": 0.7466, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 0.4024730324745178, | |
| "learning_rate": 1.2455998350925042e-06, | |
| "loss": 0.7965, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.384, | |
| "grad_norm": 0.4260917603969574, | |
| "learning_rate": 1.2302714491577834e-06, | |
| "loss": 0.7675, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 2.388, | |
| "grad_norm": 0.42180994153022766, | |
| "learning_rate": 1.2150247217412186e-06, | |
| "loss": 0.7693, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 2.392, | |
| "grad_norm": 0.3940885663032532, | |
| "learning_rate": 1.1998599831119912e-06, | |
| "loss": 0.7664, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 2.396, | |
| "grad_norm": 0.4478139281272888, | |
| "learning_rate": 1.1847775617632746e-06, | |
| "loss": 0.7766, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.464181512594223, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.779, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.404, | |
| "grad_norm": 0.38157689571380615, | |
| "learning_rate": 1.1548609759573375e-06, | |
| "loss": 0.7929, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 2.408, | |
| "grad_norm": 0.4406698942184448, | |
| "learning_rate": 1.1400274595425499e-06, | |
| "loss": 0.7581, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 2.412, | |
| "grad_norm": 0.39558476209640503, | |
| "learning_rate": 1.1252775564791023e-06, | |
| "loss": 0.7649, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 2.416, | |
| "grad_norm": 0.4835960865020752, | |
| "learning_rate": 1.1106115862741457e-06, | |
| "loss": 0.7938, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 0.4387381076812744, | |
| "learning_rate": 1.096029866616704e-06, | |
| "loss": 0.7637, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.424, | |
| "grad_norm": 0.43130138516426086, | |
| "learning_rate": 1.0815327133708015e-06, | |
| "loss": 0.7722, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 2.428, | |
| "grad_norm": 0.4822551906108856, | |
| "learning_rate": 1.0671204405686108e-06, | |
| "loss": 0.7867, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.40164196491241455, | |
| "learning_rate": 1.052793360403655e-06, | |
| "loss": 0.8015, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 2.436, | |
| "grad_norm": 0.3969416320323944, | |
| "learning_rate": 1.0385517832240472e-06, | |
| "loss": 0.7635, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 0.3942321836948395, | |
| "learning_rate": 1.0243960175257605e-06, | |
| "loss": 0.8132, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.444, | |
| "grad_norm": 0.4158990979194641, | |
| "learning_rate": 1.010326369945957e-06, | |
| "loss": 0.7571, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 2.448, | |
| "grad_norm": 0.46401944756507874, | |
| "learning_rate": 9.963431452563331e-07, | |
| "loss": 0.7628, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 2.452, | |
| "grad_norm": 0.425135999917984, | |
| "learning_rate": 9.824466463565246e-07, | |
| "loss": 0.787, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 2.456, | |
| "grad_norm": 0.45811745524406433, | |
| "learning_rate": 9.686371742675443e-07, | |
| "loss": 0.7998, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 0.41915279626846313, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.7981, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.464, | |
| "grad_norm": 0.4088100790977478, | |
| "learning_rate": 9.412805051739266e-07, | |
| "loss": 0.7642, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 2.468, | |
| "grad_norm": 0.4364009499549866, | |
| "learning_rate": 9.277339007597158e-07, | |
| "loss": 0.784, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 2.472, | |
| "grad_norm": 0.38804692029953003, | |
| "learning_rate": 9.142755083243577e-07, | |
| "loss": 0.7681, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 2.476, | |
| "grad_norm": 0.38228940963745117, | |
| "learning_rate": 9.009056193987569e-07, | |
| "loss": 0.7979, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 0.38130438327789307, | |
| "learning_rate": 8.876245235966884e-07, | |
| "loss": 0.7802, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.484, | |
| "grad_norm": 0.4364047944545746, | |
| "learning_rate": 8.744325086085248e-07, | |
| "loss": 0.7582, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 2.488, | |
| "grad_norm": 0.3930681049823761, | |
| "learning_rate": 8.613298601949971e-07, | |
| "loss": 0.7943, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 2.492, | |
| "grad_norm": 0.406017005443573, | |
| "learning_rate": 8.483168621810133e-07, | |
| "loss": 0.799, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.4273027181625366, | |
| "learning_rate": 8.353937964495029e-07, | |
| "loss": 0.747, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.4582826793193817, | |
| "learning_rate": 8.225609429353187e-07, | |
| "loss": 0.772, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.504, | |
| "grad_norm": 0.4531767666339874, | |
| "learning_rate": 8.098185796191632e-07, | |
| "loss": 0.7802, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 2.508, | |
| "grad_norm": 0.4155261814594269, | |
| "learning_rate": 7.971669825215789e-07, | |
| "loss": 0.7883, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 2.512, | |
| "grad_norm": 0.4580751657485962, | |
| "learning_rate": 7.846064256969571e-07, | |
| "loss": 0.7879, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 2.516, | |
| "grad_norm": 0.4104625880718231, | |
| "learning_rate": 7.72137181227608e-07, | |
| "loss": 0.797, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 0.40306803584098816, | |
| "learning_rate": 7.597595192178702e-07, | |
| "loss": 0.7693, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.524, | |
| "grad_norm": 0.44633400440216064, | |
| "learning_rate": 7.47473707788251e-07, | |
| "loss": 0.7791, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 2.528, | |
| "grad_norm": 0.40079960227012634, | |
| "learning_rate": 7.352800130696253e-07, | |
| "loss": 0.7548, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 2.532, | |
| "grad_norm": 0.41967469453811646, | |
| "learning_rate": 7.23178699197467e-07, | |
| "loss": 0.7598, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 2.536, | |
| "grad_norm": 0.3938508927822113, | |
| "learning_rate": 7.111700283061318e-07, | |
| "loss": 0.781, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 0.4104136824607849, | |
| "learning_rate": 6.992542605231739e-07, | |
| "loss": 0.7823, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.544, | |
| "grad_norm": 0.411960244178772, | |
| "learning_rate": 6.874316539637127e-07, | |
| "loss": 0.7728, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.548, | |
| "grad_norm": 0.41505640745162964, | |
| "learning_rate": 6.757024647248456e-07, | |
| "loss": 0.8157, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 2.552, | |
| "grad_norm": 0.38173797726631165, | |
| "learning_rate": 6.640669468800947e-07, | |
| "loss": 0.7474, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 2.556, | |
| "grad_norm": 0.42165470123291016, | |
| "learning_rate": 6.52525352473905e-07, | |
| "loss": 0.7731, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.4714242219924927, | |
| "learning_rate": 6.410779315161885e-07, | |
| "loss": 0.75, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.564, | |
| "grad_norm": 0.41340571641921997, | |
| "learning_rate": 6.297249319769016e-07, | |
| "loss": 0.755, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 2.568, | |
| "grad_norm": 0.43762728571891785, | |
| "learning_rate": 6.184665997806832e-07, | |
| "loss": 0.7378, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.572, | |
| "grad_norm": 0.41566336154937744, | |
| "learning_rate": 6.073031788015133e-07, | |
| "loss": 0.7588, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 2.576, | |
| "grad_norm": 0.3761170506477356, | |
| "learning_rate": 5.962349108574478e-07, | |
| "loss": 0.7837, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 0.4188023805618286, | |
| "learning_rate": 5.852620357053651e-07, | |
| "loss": 0.7922, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.584, | |
| "grad_norm": 0.39329683780670166, | |
| "learning_rate": 5.743847910357836e-07, | |
| "loss": 0.7545, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 2.588, | |
| "grad_norm": 0.3915480375289917, | |
| "learning_rate": 5.636034124677043e-07, | |
| "loss": 0.8032, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 2.592, | |
| "grad_norm": 0.4099782705307007, | |
| "learning_rate": 5.529181335435124e-07, | |
| "loss": 0.781, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 2.596, | |
| "grad_norm": 0.3995542824268341, | |
| "learning_rate": 5.423291857239177e-07, | |
| "loss": 0.7701, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.4187732934951782, | |
| "learning_rate": 5.318367983829393e-07, | |
| "loss": 0.7526, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.604, | |
| "grad_norm": 0.4345127046108246, | |
| "learning_rate": 5.214411988029355e-07, | |
| "loss": 0.7531, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 2.608, | |
| "grad_norm": 0.3876914381980896, | |
| "learning_rate": 5.111426121696866e-07, | |
| "loss": 0.7993, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 2.612, | |
| "grad_norm": 0.44101542234420776, | |
| "learning_rate": 5.009412615675102e-07, | |
| "loss": 0.8017, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 2.616, | |
| "grad_norm": 0.36027947068214417, | |
| "learning_rate": 4.908373679744316e-07, | |
| "loss": 0.7407, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 0.38491421937942505, | |
| "learning_rate": 4.808311502573976e-07, | |
| "loss": 0.8001, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.41484707593917847, | |
| "learning_rate": 4.709228251675357e-07, | |
| "loss": 0.8007, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 2.628, | |
| "grad_norm": 0.3889584541320801, | |
| "learning_rate": 4.6111260733545714e-07, | |
| "loss": 0.8044, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 2.632, | |
| "grad_norm": 0.3918435275554657, | |
| "learning_rate": 4.514007092666084e-07, | |
| "loss": 0.7936, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 2.636, | |
| "grad_norm": 0.4111655652523041, | |
| "learning_rate": 4.417873413366702e-07, | |
| "loss": 0.7905, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.4380398988723755, | |
| "learning_rate": 4.322727117869951e-07, | |
| "loss": 0.781, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.644, | |
| "grad_norm": 0.4324621856212616, | |
| "learning_rate": 4.228570267201049e-07, | |
| "loss": 0.822, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 2.648, | |
| "grad_norm": 0.409640371799469, | |
| "learning_rate": 4.1354049009521504e-07, | |
| "loss": 0.7982, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 2.652, | |
| "grad_norm": 0.4422895014286041, | |
| "learning_rate": 4.043233037238281e-07, | |
| "loss": 0.7581, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 2.656, | |
| "grad_norm": 0.39165276288986206, | |
| "learning_rate": 3.9520566726535367e-07, | |
| "loss": 0.8131, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 0.37336215376853943, | |
| "learning_rate": 3.8618777822278854e-07, | |
| "loss": 0.7423, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.664, | |
| "grad_norm": 0.46584251523017883, | |
| "learning_rate": 3.772698319384349e-07, | |
| "loss": 0.727, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 2.668, | |
| "grad_norm": 0.39146313071250916, | |
| "learning_rate": 3.684520215896703e-07, | |
| "loss": 0.7862, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 2.672, | |
| "grad_norm": 0.39932945370674133, | |
| "learning_rate": 3.597345381847656e-07, | |
| "loss": 0.7574, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 2.676, | |
| "grad_norm": 0.40844547748565674, | |
| "learning_rate": 3.511175705587433e-07, | |
| "loss": 0.7754, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 0.4052230715751648, | |
| "learning_rate": 3.426013053692878e-07, | |
| "loss": 0.7575, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.684, | |
| "grad_norm": 0.40574780106544495, | |
| "learning_rate": 3.341859270927067e-07, | |
| "loss": 0.7761, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.3910069465637207, | |
| "learning_rate": 3.258716180199278e-07, | |
| "loss": 0.7723, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 2.692, | |
| "grad_norm": 0.3847978115081787, | |
| "learning_rate": 3.1765855825255543e-07, | |
| "loss": 0.7849, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 2.6959999999999997, | |
| "grad_norm": 0.3743051290512085, | |
| "learning_rate": 3.0954692569896585e-07, | |
| "loss": 0.7825, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 0.43019863963127136, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.7791, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.7039999999999997, | |
| "grad_norm": 0.42280012369155884, | |
| "learning_rate": 2.9362864287744266e-07, | |
| "loss": 0.7682, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 2.708, | |
| "grad_norm": 0.4100990891456604, | |
| "learning_rate": 2.858223374256841e-07, | |
| "loss": 0.7816, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 2.7119999999999997, | |
| "grad_norm": 0.4392656981945038, | |
| "learning_rate": 2.7811814881259503e-07, | |
| "loss": 0.7645, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 2.716, | |
| "grad_norm": 0.37788864970207214, | |
| "learning_rate": 2.705162439235648e-07, | |
| "loss": 0.7793, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 2.7199999999999998, | |
| "grad_norm": 0.3875264823436737, | |
| "learning_rate": 2.63016787428354e-07, | |
| "loss": 0.7777, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.724, | |
| "grad_norm": 0.41191229224205017, | |
| "learning_rate": 2.556199417775174e-07, | |
| "loss": 0.8064, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 2.7279999999999998, | |
| "grad_norm": 0.3729294538497925, | |
| "learning_rate": 2.483258671988942e-07, | |
| "loss": 0.7664, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 2.732, | |
| "grad_norm": 0.39030614495277405, | |
| "learning_rate": 2.4113472169413176e-07, | |
| "loss": 0.768, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 2.7359999999999998, | |
| "grad_norm": 0.3516885042190552, | |
| "learning_rate": 2.3404666103526542e-07, | |
| "loss": 0.7609, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 0.38836079835891724, | |
| "learning_rate": 2.2706183876134047e-07, | |
| "loss": 0.8086, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 2.7439999999999998, | |
| "grad_norm": 0.40275564789772034, | |
| "learning_rate": 2.2018040617509174e-07, | |
| "loss": 0.8176, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 2.748, | |
| "grad_norm": 0.3835076689720154, | |
| "learning_rate": 2.134025123396638e-07, | |
| "loss": 0.7824, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.3846163749694824, | |
| "learning_rate": 2.0672830407537925e-07, | |
| "loss": 0.775, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 2.7560000000000002, | |
| "grad_norm": 0.4121825397014618, | |
| "learning_rate": 2.0015792595656225e-07, | |
| "loss": 0.8086, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 0.39317595958709717, | |
| "learning_rate": 1.9369152030840553e-07, | |
| "loss": 0.7897, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.7640000000000002, | |
| "grad_norm": 0.4010584354400635, | |
| "learning_rate": 1.873292272038868e-07, | |
| "loss": 0.7258, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 2.768, | |
| "grad_norm": 0.39328062534332275, | |
| "learning_rate": 1.8107118446073492e-07, | |
| "loss": 0.7727, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 2.7720000000000002, | |
| "grad_norm": 0.4069308042526245, | |
| "learning_rate": 1.7491752763844294e-07, | |
| "loss": 0.7364, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 2.776, | |
| "grad_norm": 0.4272211790084839, | |
| "learning_rate": 1.688683900353366e-07, | |
| "loss": 0.769, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 2.7800000000000002, | |
| "grad_norm": 0.3933575749397278, | |
| "learning_rate": 1.6292390268568103e-07, | |
| "loss": 0.8167, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 2.784, | |
| "grad_norm": 0.416881799697876, | |
| "learning_rate": 1.5708419435684463e-07, | |
| "loss": 0.7854, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 2.7880000000000003, | |
| "grad_norm": 0.3927464187145233, | |
| "learning_rate": 1.5134939154651196e-07, | |
| "loss": 0.7785, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 2.792, | |
| "grad_norm": 0.3979874551296234, | |
| "learning_rate": 1.4571961847993977e-07, | |
| "loss": 0.782, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 2.7960000000000003, | |
| "grad_norm": 0.4113554060459137, | |
| "learning_rate": 1.4019499710726913e-07, | |
| "loss": 0.776, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.40728065371513367, | |
| "learning_rate": 1.3477564710088097e-07, | |
| "loss": 0.7813, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.8040000000000003, | |
| "grad_norm": 0.37924978137016296, | |
| "learning_rate": 1.294616858528064e-07, | |
| "loss": 0.7608, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 2.808, | |
| "grad_norm": 0.4243476986885071, | |
| "learning_rate": 1.2425322847218368e-07, | |
| "loss": 0.7679, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 2.8120000000000003, | |
| "grad_norm": 0.385475754737854, | |
| "learning_rate": 1.1915038778276212e-07, | |
| "loss": 0.7511, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.40648695826530457, | |
| "learning_rate": 1.1415327432046041e-07, | |
| "loss": 0.7406, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 0.39061108231544495, | |
| "learning_rate": 1.0926199633097156e-07, | |
| "loss": 0.8138, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 2.824, | |
| "grad_norm": 0.4108492434024811, | |
| "learning_rate": 1.044766597674196e-07, | |
| "loss": 0.7478, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 2.828, | |
| "grad_norm": 0.3769419193267822, | |
| "learning_rate": 9.979736828806096e-08, | |
| "loss": 0.7825, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 2.832, | |
| "grad_norm": 0.39844822883605957, | |
| "learning_rate": 9.522422325404234e-08, | |
| "loss": 0.8049, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 2.836, | |
| "grad_norm": 0.39169108867645264, | |
| "learning_rate": 9.075732372720414e-08, | |
| "loss": 0.7697, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 0.3846679925918579, | |
| "learning_rate": 8.639676646793382e-08, | |
| "loss": 0.7875, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.844, | |
| "grad_norm": 0.3852086365222931, | |
| "learning_rate": 8.214264593307097e-08, | |
| "loss": 0.7587, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 2.848, | |
| "grad_norm": 0.43485692143440247, | |
| "learning_rate": 7.799505427386001e-08, | |
| "loss": 0.7964, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 2.852, | |
| "grad_norm": 0.3984343111515045, | |
| "learning_rate": 7.395408133395509e-08, | |
| "loss": 0.7781, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 2.856, | |
| "grad_norm": 0.4015841782093048, | |
| "learning_rate": 7.001981464747565e-08, | |
| "loss": 0.7413, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 0.3932044804096222, | |
| "learning_rate": 6.61923394371039e-08, | |
| "loss": 0.7592, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 2.864, | |
| "grad_norm": 0.38348260521888733, | |
| "learning_rate": 6.247173861224753e-08, | |
| "loss": 0.7833, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 2.868, | |
| "grad_norm": 0.39616912603378296, | |
| "learning_rate": 5.8858092767236084e-08, | |
| "loss": 0.8241, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 2.872, | |
| "grad_norm": 0.39276373386383057, | |
| "learning_rate": 5.535148017958014e-08, | |
| "loss": 0.775, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 2.876, | |
| "grad_norm": 0.3859685957431793, | |
| "learning_rate": 5.19519768082738e-08, | |
| "loss": 0.7998, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.3932885229587555, | |
| "learning_rate": 4.865965629214819e-08, | |
| "loss": 0.728, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.884, | |
| "grad_norm": 0.3808436989784241, | |
| "learning_rate": 4.5474589948280026e-08, | |
| "loss": 0.7968, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 2.888, | |
| "grad_norm": 0.39818060398101807, | |
| "learning_rate": 4.2396846770441644e-08, | |
| "loss": 0.798, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 2.892, | |
| "grad_norm": 0.41582825779914856, | |
| "learning_rate": 3.9426493427611177e-08, | |
| "loss": 0.7847, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 2.896, | |
| "grad_norm": 0.38983744382858276, | |
| "learning_rate": 3.65635942625242e-08, | |
| "loss": 0.8021, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.38485726714134216, | |
| "learning_rate": 3.3808211290284886e-08, | |
| "loss": 0.8147, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.904, | |
| "grad_norm": 0.39613547921180725, | |
| "learning_rate": 3.1160404197018155e-08, | |
| "loss": 0.794, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 2.908, | |
| "grad_norm": 0.40311679244041443, | |
| "learning_rate": 2.8620230338578526e-08, | |
| "loss": 0.7908, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 2.912, | |
| "grad_norm": 0.3574402630329132, | |
| "learning_rate": 2.6187744739308297e-08, | |
| "loss": 0.8137, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 2.916, | |
| "grad_norm": 0.3904540538787842, | |
| "learning_rate": 2.386300009084408e-08, | |
| "loss": 0.764, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 0.3986371159553528, | |
| "learning_rate": 2.1646046750978255e-08, | |
| "loss": 0.7881, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.924, | |
| "grad_norm": 0.39145928621292114, | |
| "learning_rate": 1.953693274256374e-08, | |
| "loss": 0.764, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 2.928, | |
| "grad_norm": 0.38288819789886475, | |
| "learning_rate": 1.753570375247815e-08, | |
| "loss": 0.7774, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 2.932, | |
| "grad_norm": 0.3810698986053467, | |
| "learning_rate": 1.5642403130632367e-08, | |
| "loss": 0.7491, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 2.936, | |
| "grad_norm": 0.3975498080253601, | |
| "learning_rate": 1.3857071889029073e-08, | |
| "loss": 0.7755, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 0.38917872309684753, | |
| "learning_rate": 1.2179748700879013e-08, | |
| "loss": 0.7924, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.38839098811149597, | |
| "learning_rate": 1.0610469899760001e-08, | |
| "loss": 0.8068, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 2.948, | |
| "grad_norm": 0.38586539030075073, | |
| "learning_rate": 9.14926947883088e-09, | |
| "loss": 0.7816, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 2.952, | |
| "grad_norm": 0.4118897020816803, | |
| "learning_rate": 7.796179090094891e-09, | |
| "loss": 0.7837, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 2.956, | |
| "grad_norm": 0.36527642607688904, | |
| "learning_rate": 6.551228043715218e-09, | |
| "loss": 0.8002, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 0.3902311623096466, | |
| "learning_rate": 5.414443307377171e-09, | |
| "loss": 0.7759, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.964, | |
| "grad_norm": 0.37341415882110596, | |
| "learning_rate": 4.385849505708084e-09, | |
| "loss": 0.7821, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 2.968, | |
| "grad_norm": 0.40262770652770996, | |
| "learning_rate": 3.4654689197405335e-09, | |
| "loss": 0.7754, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 2.972, | |
| "grad_norm": 0.37969642877578735, | |
| "learning_rate": 2.6533214864310485e-09, | |
| "loss": 0.7755, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 2.976, | |
| "grad_norm": 0.3812060058116913, | |
| "learning_rate": 1.9494247982282386e-09, | |
| "loss": 0.7779, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "grad_norm": 0.3999031186103821, | |
| "learning_rate": 1.3537941026914302e-09, | |
| "loss": 0.7995, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 2.984, | |
| "grad_norm": 0.36233609914779663, | |
| "learning_rate": 8.664423021614854e-10, | |
| "loss": 0.7679, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 2.988, | |
| "grad_norm": 0.4313134253025055, | |
| "learning_rate": 4.87379953478806e-10, | |
| "loss": 0.8057, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 2.992, | |
| "grad_norm": 0.3804738223552704, | |
| "learning_rate": 2.1661526775795804e-10, | |
| "loss": 0.7728, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 2.996, | |
| "grad_norm": 0.410442054271698, | |
| "learning_rate": 5.4154110206150465e-11, | |
| "loss": 0.7773, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.37331917881965637, | |
| "learning_rate": 0.0, | |
| "loss": 0.7804, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 750, | |
| "total_flos": 750914578743296.0, | |
| "train_loss": 0.842607304652532, | |
| "train_runtime": 42164.2983, | |
| "train_samples_per_second": 1.708, | |
| "train_steps_per_second": 0.018 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 750, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 750914578743296.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |