| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.642857142857143, | |
| "eval_steps": 500, | |
| "global_step": 236, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01680672268907563, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 2e-05, | |
| "loss": 2.9247, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03361344537815126, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 4e-05, | |
| "loss": 2.8013, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05042016806722689, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 6e-05, | |
| "loss": 2.9511, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.06722689075630252, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 8e-05, | |
| "loss": 2.91, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.08403361344537816, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 0.0001, | |
| "loss": 2.766, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.10084033613445378, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 0.00012, | |
| "loss": 2.6274, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.11764705882352941, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 0.00014, | |
| "loss": 2.3908, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.13445378151260504, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00016, | |
| "loss": 2.544, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.15126050420168066, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00018, | |
| "loss": 2.4026, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.16806722689075632, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.0002, | |
| "loss": 2.4927, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.18487394957983194, | |
| "grad_norm": 0.2177734375, | |
| "learning_rate": 0.00019999033847063811, | |
| "loss": 2.4668, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.20168067226890757, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.00019996135574945544, | |
| "loss": 2.3493, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.2184873949579832, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 0.00019991305743680013, | |
| "loss": 2.4891, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 0.0001998454528653836, | |
| "loss": 2.3903, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.25210084033613445, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 0.00019975855509847686, | |
| "loss": 2.3877, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2689075630252101, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.00019965238092738643, | |
| "loss": 2.4276, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 0.00019952695086820975, | |
| "loss": 2.3165, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3025210084033613, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 0.0001993822891578708, | |
| "loss": 2.3638, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.31932773109243695, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.0001992184237494368, | |
| "loss": 2.2576, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.33613445378151263, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 0.0001990353863067169, | |
| "loss": 2.4158, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.35294117647058826, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 0.0001988332121981436, | |
| "loss": 2.2791, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.3697478991596639, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 0.00019861194048993863, | |
| "loss": 2.4497, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.3865546218487395, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 0.0001983716139385641, | |
| "loss": 2.4217, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.40336134453781514, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 0.0001981122789824607, | |
| "loss": 2.3775, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.42016806722689076, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 0.00019783398573307428, | |
| "loss": 2.3257, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.4369747899159664, | |
| "grad_norm": 0.2265625, | |
| "learning_rate": 0.00019753678796517282, | |
| "loss": 2.3262, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.453781512605042, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 0.00019722074310645553, | |
| "loss": 2.3464, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 0.00019688591222645607, | |
| "loss": 2.4651, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.48739495798319327, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.000196532360024742, | |
| "loss": 2.2998, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5042016806722689, | |
| "grad_norm": 0.2138671875, | |
| "learning_rate": 0.0001961601548184129, | |
| "loss": 2.3678, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5210084033613446, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.00019576936852889936, | |
| "loss": 2.285, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5378151260504201, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.00019536007666806556, | |
| "loss": 2.316, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5546218487394958, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 0.0001949323583236181, | |
| "loss": 2.3368, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0001944862961438239, | |
| "loss": 2.2433, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5882352941176471, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 0.00019402197632153992, | |
| "loss": 2.4387, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.6050420168067226, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.00019353948857755803, | |
| "loss": 2.3809, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.6218487394957983, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 0.00019303892614326836, | |
| "loss": 2.2351, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.6386554621848739, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.00019252038574264405, | |
| "loss": 2.231, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6554621848739496, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 0.00019198396757355118, | |
| "loss": 2.3298, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.6722689075630253, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.00019142977528838762, | |
| "loss": 2.3444, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6890756302521008, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 0.00019085791597405404, | |
| "loss": 2.2819, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 0.00019026850013126157, | |
| "loss": 2.2556, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.7226890756302521, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 0.00018966164165317966, | |
| "loss": 2.3512, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.7394957983193278, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 0.00018903745780342839, | |
| "loss": 2.242, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7563025210084033, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 0.0001883960691934196, | |
| "loss": 2.2857, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.773109243697479, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.00018773759975905098, | |
| "loss": 2.3035, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.7899159663865546, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 0.00018706217673675811, | |
| "loss": 2.3832, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.8067226890756303, | |
| "grad_norm": 0.2333984375, | |
| "learning_rate": 0.0001863699306389282, | |
| "loss": 2.2734, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.8235294117647058, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 0.00018566099522868119, | |
| "loss": 2.3121, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.8403361344537815, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 0.00018493550749402278, | |
| "loss": 2.3118, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 0.00018419360762137395, | |
| "loss": 2.2884, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.8739495798319328, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 0.00018343543896848273, | |
| "loss": 2.2719, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.8907563025210085, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 0.00018266114803672318, | |
| "loss": 2.275, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.907563025210084, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.00018187088444278674, | |
| "loss": 2.2271, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.9243697478991597, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.00018106480088977172, | |
| "loss": 2.2866, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 0.00018024305313767646, | |
| "loss": 2.2417, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.957983193277311, | |
| "grad_norm": 0.2333984375, | |
| "learning_rate": 0.00017940579997330165, | |
| "loss": 2.327, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.9747899159663865, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 0.00017855320317956784, | |
| "loss": 2.2002, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.9915966386554622, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 0.00017768542750425426, | |
| "loss": 2.3351, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.0084033613445378, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 0.0001768026406281642, | |
| "loss": 2.2484, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0252100840336134, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 0.00017590501313272415, | |
| "loss": 2.3226, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.0420168067226891, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 0.00017499271846702213, | |
| "loss": 2.3126, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.0588235294117647, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 0.00017406593291429217, | |
| "loss": 2.3496, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.0756302521008403, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 0.00017312483555785086, | |
| "loss": 2.3221, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.0924369747899159, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 0.00017216960824649303, | |
| "loss": 2.2812, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.0084033613445378, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 0.00017120043555935298, | |
| "loss": 2.1682, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.0252100840336134, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.0001702175047702382, | |
| "loss": 2.1584, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.0420168067226891, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 0.00016922100581144228, | |
| "loss": 2.1734, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.0588235294117647, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 0.00016821113123704424, | |
| "loss": 2.3534, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.0756302521008403, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00016718807618570106, | |
| "loss": 2.2248, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.092436974789916, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 0.00016615203834294119, | |
| "loss": 2.2215, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.1092436974789917, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00016510321790296525, | |
| "loss": 2.2602, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.1260504201680672, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 0.00016404181752996289, | |
| "loss": 2.2532, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.00016296804231895142, | |
| "loss": 2.2066, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.1596638655462184, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00016188209975614542, | |
| "loss": 2.1352, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.00016078419967886402, | |
| "loss": 2.256, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.1932773109243697, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 0.00015967455423498387, | |
| "loss": 2.2871, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.2100840336134453, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 0.00015855337784194577, | |
| "loss": 2.2594, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.226890756302521, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 0.00015742088714532247, | |
| "loss": 2.2358, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.2436974789915967, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 0.00015627730097695638, | |
| "loss": 2.2143, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.2605042016806722, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00015512284031267437, | |
| "loss": 2.1646, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.2773109243697478, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 0.00015395772822958845, | |
| "loss": 2.1203, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.2941176470588236, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00015278218986299074, | |
| "loss": 2.0921, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.3109243697478992, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.0001515964523628501, | |
| "loss": 2.1823, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.3277310924369747, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.00015040074484992, | |
| "loss": 2.1761, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.3445378151260505, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00014919529837146528, | |
| "loss": 2.1554, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.361344537815126, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 0.00014798034585661695, | |
| "loss": 2.222, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.3781512605042017, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 0.0001467561220713628, | |
| "loss": 2.0991, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.3949579831932772, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.0001455228635731839, | |
| "loss": 2.2474, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.4117647058823528, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.00014428080866534396, | |
| "loss": 2.049, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 0.00014303019735084226, | |
| "loss": 2.2173, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.4453781512605042, | |
| "grad_norm": 0.2265625, | |
| "learning_rate": 0.00014177127128603745, | |
| "loss": 2.1714, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.46218487394958, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.0001405042737339524, | |
| "loss": 2.1857, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.4789915966386555, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 0.0001392294495172681, | |
| "loss": 2.2088, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.495798319327731, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 0.00013794704497101655, | |
| "loss": 2.0494, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.5126050420168067, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.0001366573078949813, | |
| "loss": 2.2291, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.5294117647058822, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 0.00013536048750581494, | |
| "loss": 2.1513, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.5462184873949578, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.00013405683438888282, | |
| "loss": 2.2065, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.5630252100840336, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00013274660044984224, | |
| "loss": 2.1891, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.5798319327731094, | |
| "grad_norm": 0.220703125, | |
| "learning_rate": 0.00013143003886596669, | |
| "loss": 2.1263, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.596638655462185, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0001301074040372242, | |
| "loss": 2.1815, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.6134453781512605, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00012877895153711935, | |
| "loss": 2.1458, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.6302521008403361, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 0.0001274449380633089, | |
| "loss": 2.2204, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.6470588235294117, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 0.00012610562138799978, | |
| "loss": 2.2575, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.6638655462184873, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00012476126030813963, | |
| "loss": 2.1376, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.680672268907563, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.0001234121145954094, | |
| "loss": 2.1891, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.6974789915966386, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 0.0001220584449460274, | |
| "loss": 2.2351, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00012070051293037492, | |
| "loss": 2.2202, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.73109243697479, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 0.00011933858094245281, | |
| "loss": 2.1824, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.7478991596638656, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.00011797291214917881, | |
| "loss": 2.1288, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.7647058823529411, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 0.00011660377043953588, | |
| "loss": 2.1954, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.7815126050420167, | |
| "grad_norm": 0.2275390625, | |
| "learning_rate": 0.0001152314203735805, | |
| "loss": 2.224, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.7983193277310925, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 0.0001138561271313219, | |
| "loss": 2.2759, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.815126050420168, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00011247815646148087, | |
| "loss": 2.0882, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.8319327731092439, | |
| "grad_norm": 0.2294921875, | |
| "learning_rate": 0.00011109777463013915, | |
| "loss": 2.0767, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.8487394957983194, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.0001097152483692886, | |
| "loss": 2.2182, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.865546218487395, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 0.00010833084482529048, | |
| "loss": 2.1451, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.8823529411764706, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 0.00010694483150725458, | |
| "loss": 2.1774, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.8991596638655461, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.00010555747623534831, | |
| "loss": 2.1748, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.9159663865546217, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00010416904708904548, | |
| "loss": 2.1665, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.9327731092436975, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.00010277981235532541, | |
| "loss": 2.1585, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.949579831932773, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 0.00010139004047683151, | |
| "loss": 2.1784, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.9663865546218489, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.0001, | |
| "loss": 2.2399, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.9831932773109244, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 9.860995952316851e-05, | |
| "loss": 2.2187, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.2294921875, | |
| "learning_rate": 9.722018764467461e-05, | |
| "loss": 2.2007, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.0168067226890756, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 9.583095291095453e-05, | |
| "loss": 2.2312, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.033613445378151, | |
| "grad_norm": 0.220703125, | |
| "learning_rate": 9.444252376465171e-05, | |
| "loss": 2.2207, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.0504201680672267, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 9.305516849274541e-05, | |
| "loss": 2.2537, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.0672268907563023, | |
| "grad_norm": 0.2265625, | |
| "learning_rate": 9.166915517470953e-05, | |
| "loss": 2.2015, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.084033613445378, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 9.028475163071141e-05, | |
| "loss": 2.2012, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.100840336134454, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 8.890222536986085e-05, | |
| "loss": 2.1247, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.008403361344538, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 8.752184353851916e-05, | |
| "loss": 2.1723, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.0252100840336134, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 8.614387286867814e-05, | |
| "loss": 2.083, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.042016806722689, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 8.47685796264195e-05, | |
| "loss": 2.014, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.0588235294117645, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 8.339622956046417e-05, | |
| "loss": 2.0522, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.0756302521008405, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 8.202708785082121e-05, | |
| "loss": 2.0255, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.092436974789916, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 8.066141905754723e-05, | |
| "loss": 2.0615, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.1092436974789917, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 7.929948706962508e-05, | |
| "loss": 2.0913, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.1260504201680672, | |
| "grad_norm": 0.25, | |
| "learning_rate": 7.794155505397261e-05, | |
| "loss": 2.0726, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 7.658788540459062e-05, | |
| "loss": 2.0546, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.1596638655462184, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 7.523873969186039e-05, | |
| "loss": 2.0894, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.176470588235294, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 7.389437861200024e-05, | |
| "loss": 1.9735, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.19327731092437, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 7.25550619366911e-05, | |
| "loss": 2.0254, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.2100840336134455, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 7.122104846288064e-05, | |
| "loss": 2.0967, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.226890756302521, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 6.989259596277582e-05, | |
| "loss": 2.0115, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.2436974789915967, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 6.85699611340333e-05, | |
| "loss": 2.1694, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.2605042016806722, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 6.725339955015777e-05, | |
| "loss": 2.0343, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.277310924369748, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 6.594316561111724e-05, | |
| "loss": 2.0645, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.2941176470588234, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 6.46395124941851e-05, | |
| "loss": 2.0, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.310924369747899, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 6.334269210501875e-05, | |
| "loss": 2.0821, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.327731092436975, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 6.205295502898348e-05, | |
| "loss": 2.031, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.3445378151260505, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 6.0770550482731924e-05, | |
| "loss": 2.0395, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.361344537815126, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 5.9495726266047605e-05, | |
| "loss": 2.1277, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.3781512605042017, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 5.8228728713962543e-05, | |
| "loss": 2.0844, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.3949579831932772, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 5.696980264915777e-05, | |
| "loss": 2.0771, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.411764705882353, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 5.571919133465605e-05, | |
| "loss": 2.0715, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.4285714285714284, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 5.447713642681612e-05, | |
| "loss": 2.0176, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.4453781512605044, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 5.324387792863719e-05, | |
| "loss": 2.0616, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.46218487394958, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 5.201965414338308e-05, | |
| "loss": 2.1075, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.4789915966386555, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 5.080470162853472e-05, | |
| "loss": 2.1146, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.495798319327731, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 4.959925515008002e-05, | |
| "loss": 2.0432, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.5126050420168067, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 4.840354763714991e-05, | |
| "loss": 2.0304, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.5294117647058822, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 4.7217810137009274e-05, | |
| "loss": 2.128, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.546218487394958, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 4.604227177041156e-05, | |
| "loss": 2.1879, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.5630252100840334, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 4.487715968732568e-05, | |
| "loss": 2.0896, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.5798319327731094, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 4.372269902304363e-05, | |
| "loss": 2.1105, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.596638655462185, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 4.257911285467754e-05, | |
| "loss": 2.0194, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.6134453781512605, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 4.144662215805426e-05, | |
| "loss": 2.1294, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.630252100840336, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 4.0325445765016145e-05, | |
| "loss": 2.0898, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.6470588235294117, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 3.921580032113602e-05, | |
| "loss": 2.0419, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.6638655462184873, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 3.8117900243854595e-05, | |
| "loss": 2.0747, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.6806722689075633, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 3.7031957681048604e-05, | |
| "loss": 2.1093, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.697478991596639, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 3.595818247003713e-05, | |
| "loss": 2.0736, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.7142857142857144, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 3.489678209703475e-05, | |
| "loss": 2.1207, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.73109243697479, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 3.3847961657058845e-05, | |
| "loss": 2.1366, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.7478991596638656, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 3.281192381429894e-05, | |
| "loss": 2.1676, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.764705882352941, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 3.178886876295578e-05, | |
| "loss": 2.0748, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.7815126050420167, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 3.077899418855772e-05, | |
| "loss": 2.0757, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.7983193277310923, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 2.9782495229761808e-05, | |
| "loss": 2.106, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.815126050420168, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 2.879956444064703e-05, | |
| "loss": 2.0526, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.831932773109244, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 2.783039175350699e-05, | |
| "loss": 2.1291, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.8487394957983194, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 2.6875164442149147e-05, | |
| "loss": 2.116, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.865546218487395, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 2.5934067085707834e-05, | |
| "loss": 2.0595, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.8823529411764706, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 2.500728153297788e-05, | |
| "loss": 2.0402, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.899159663865546, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 2.409498686727587e-05, | |
| "loss": 2.1448, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.9159663865546217, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 2.3197359371835802e-05, | |
| "loss": 2.0672, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.9327731092436977, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 2.2314572495745746e-05, | |
| "loss": 2.1085, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.9495798319327733, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 2.1446796820432167e-05, | |
| "loss": 2.0861, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.966386554621849, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 2.0594200026698363e-05, | |
| "loss": 2.1058, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.9831932773109244, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 1.9756946862323535e-05, | |
| "loss": 2.1034, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 1.8935199110228275e-05, | |
| "loss": 2.112, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 3.0168067226890756, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 1.8129115557213262e-05, | |
| "loss": 2.0627, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 3.033613445378151, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 1.7338851963276825e-05, | |
| "loss": 2.0705, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 3.0504201680672267, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 1.656456103151728e-05, | |
| "loss": 2.0702, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 3.0672268907563023, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 1.580639237862608e-05, | |
| "loss": 2.0607, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 3.084033613445378, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 1.5064492505977234e-05, | |
| "loss": 2.0568, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 3.100840336134454, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 1.433900477131882e-05, | |
| "loss": 2.0892, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 3.004201680672269, | |
| "grad_norm": 0.68359375, | |
| "learning_rate": 1.363006936107183e-05, | |
| "loss": 2.092, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 3.0210084033613445, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 1.29378232632419e-05, | |
| "loss": 2.0288, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 3.03781512605042, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 1.2262400240949023e-05, | |
| "loss": 2.0081, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.0546218487394956, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 1.1603930806580444e-05, | |
| "loss": 2.1497, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 3.0714285714285716, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 1.0962542196571634e-05, | |
| "loss": 2.0169, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 3.088235294117647, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 1.0338358346820353e-05, | |
| "loss": 2.0305, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 3.1050420168067228, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 9.731499868738447e-06, | |
| "loss": 2.1016, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 3.1218487394957983, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 9.142084025945984e-06, | |
| "loss": 2.0196, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 3.138655462184874, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 8.570224711612385e-06, | |
| "loss": 2.0644, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 3.1554621848739495, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 8.016032426448817e-06, | |
| "loss": 2.0463, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 3.172268907563025, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 7.479614257355971e-06, | |
| "loss": 2.043, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 3.189075630252101, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 6.961073856731648e-06, | |
| "loss": 2.0055, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 3.2058823529411766, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 6.460511422441984e-06, | |
| "loss": 2.0479, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.222689075630252, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 5.978023678460099e-06, | |
| "loss": 2.0462, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 3.2394957983193278, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 5.5137038561761115e-06, | |
| "loss": 1.9518, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 3.2563025210084033, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 5.067641676381918e-06, | |
| "loss": 2.0281, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 3.273109243697479, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 4.639923331934471e-06, | |
| "loss": 2.0816, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 3.2899159663865545, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 4.230631471100655e-06, | |
| "loss": 1.9686, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 3.30672268907563, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 3.839845181587098e-06, | |
| "loss": 2.0643, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 3.323529411764706, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 3.467639975257997e-06, | |
| "loss": 1.9867, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 3.3403361344537816, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 3.1140877735439387e-06, | |
| "loss": 1.9972, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 3.357142857142857, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 2.7792568935444796e-06, | |
| "loss": 2.0394, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 3.3739495798319328, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 2.4632120348272003e-06, | |
| "loss": 1.945, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.3907563025210083, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 2.166014266925731e-06, | |
| "loss": 2.0512, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 3.407563025210084, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 1.88772101753929e-06, | |
| "loss": 1.991, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 3.4243697478991595, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 1.6283860614358936e-06, | |
| "loss": 1.9694, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 3.4411764705882355, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 1.3880595100613792e-06, | |
| "loss": 2.0378, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 3.457983193277311, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 1.1667878018564171e-06, | |
| "loss": 1.9876, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 3.4747899159663866, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 9.64613693283123e-07, | |
| "loss": 2.0086, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 3.491596638655462, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 7.815762505632096e-07, | |
| "loss": 1.9774, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 3.508403361344538, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 6.177108421292266e-07, | |
| "loss": 2.0522, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 3.5252100840336134, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 4.7304913179025965e-07, | |
| "loss": 2.0401, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 3.542016806722689, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 3.4761907261356976e-07, | |
| "loss": 2.1176, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.5588235294117645, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 2.414449015231357e-07, | |
| "loss": 2.0368, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 3.57563025210084, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 1.545471346164007e-07, | |
| "loss": 2.0378, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 3.592436974789916, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 8.694256319987659e-08, | |
| "loss": 2.0836, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 3.6092436974789917, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 3.8644250544594975e-08, | |
| "loss": 1.9831, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.6260504201680672, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 9.661529361892907e-09, | |
| "loss": 2.014, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.642857142857143, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.0, | |
| "loss": 1.9648, | |
| "step": 236 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 236, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 30, | |
| "total_flos": 3.519304308086538e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |