| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 676, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0014792899408284023, | |
| "grad_norm": 19.051719665527344, | |
| "learning_rate": 0.0002, | |
| "loss": 11.4654, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0029585798816568047, | |
| "grad_norm": 23.727689743041992, | |
| "learning_rate": 0.00019970414201183435, | |
| "loss": 9.7769, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.004437869822485207, | |
| "grad_norm": 8.161388397216797, | |
| "learning_rate": 0.00019940828402366864, | |
| "loss": 8.0434, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.005917159763313609, | |
| "grad_norm": 7.546419620513916, | |
| "learning_rate": 0.00019911242603550298, | |
| "loss": 7.4566, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0073964497041420114, | |
| "grad_norm": 5.024906158447266, | |
| "learning_rate": 0.0001988165680473373, | |
| "loss": 7.1297, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.008875739644970414, | |
| "grad_norm": 4.575231075286865, | |
| "learning_rate": 0.0001985207100591716, | |
| "loss": 6.7507, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.010355029585798817, | |
| "grad_norm": 3.720844268798828, | |
| "learning_rate": 0.00019822485207100592, | |
| "loss": 6.7162, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.011834319526627219, | |
| "grad_norm": 3.82381534576416, | |
| "learning_rate": 0.00019792899408284023, | |
| "loss": 6.4728, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.013313609467455622, | |
| "grad_norm": 3.960883140563965, | |
| "learning_rate": 0.00019763313609467458, | |
| "loss": 6.0401, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.014792899408284023, | |
| "grad_norm": 11.000751495361328, | |
| "learning_rate": 0.0001973372781065089, | |
| "loss": 6.2245, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.016272189349112426, | |
| "grad_norm": 6.994083404541016, | |
| "learning_rate": 0.0001970414201183432, | |
| "loss": 5.9574, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.01775147928994083, | |
| "grad_norm": 5.17557430267334, | |
| "learning_rate": 0.00019674556213017752, | |
| "loss": 5.8088, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.019230769230769232, | |
| "grad_norm": 5.766794681549072, | |
| "learning_rate": 0.00019644970414201186, | |
| "loss": 5.8526, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.020710059171597635, | |
| "grad_norm": 8.630630493164062, | |
| "learning_rate": 0.00019615384615384615, | |
| "loss": 5.7402, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.022189349112426034, | |
| "grad_norm": 8.114038467407227, | |
| "learning_rate": 0.00019585798816568049, | |
| "loss": 5.657, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.023668639053254437, | |
| "grad_norm": 4.028295516967773, | |
| "learning_rate": 0.0001955621301775148, | |
| "loss": 5.6914, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.02514792899408284, | |
| "grad_norm": 4.497015953063965, | |
| "learning_rate": 0.00019526627218934911, | |
| "loss": 5.5357, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.026627218934911243, | |
| "grad_norm": 6.016404628753662, | |
| "learning_rate": 0.00019497041420118345, | |
| "loss": 5.8588, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.028106508875739646, | |
| "grad_norm": 3.9681546688079834, | |
| "learning_rate": 0.00019467455621301777, | |
| "loss": 5.6267, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.029585798816568046, | |
| "grad_norm": 5.046078681945801, | |
| "learning_rate": 0.00019437869822485208, | |
| "loss": 5.8643, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03106508875739645, | |
| "grad_norm": 5.350937366485596, | |
| "learning_rate": 0.0001940828402366864, | |
| "loss": 5.4822, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.03254437869822485, | |
| "grad_norm": 5.343432426452637, | |
| "learning_rate": 0.0001937869822485207, | |
| "loss": 5.5491, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.034023668639053255, | |
| "grad_norm": 5.352399826049805, | |
| "learning_rate": 0.00019349112426035502, | |
| "loss": 5.4912, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.03550295857988166, | |
| "grad_norm": 4.353022575378418, | |
| "learning_rate": 0.00019319526627218937, | |
| "loss": 5.4095, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.03698224852071006, | |
| "grad_norm": 5.44649076461792, | |
| "learning_rate": 0.00019289940828402368, | |
| "loss": 5.6072, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.038461538461538464, | |
| "grad_norm": 3.686049699783325, | |
| "learning_rate": 0.000192603550295858, | |
| "loss": 5.4506, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.03994082840236687, | |
| "grad_norm": 5.867905139923096, | |
| "learning_rate": 0.00019230769230769233, | |
| "loss": 5.0424, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.04142011834319527, | |
| "grad_norm": 4.650610446929932, | |
| "learning_rate": 0.00019201183431952662, | |
| "loss": 5.4751, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.042899408284023666, | |
| "grad_norm": 7.5463361740112305, | |
| "learning_rate": 0.00019171597633136096, | |
| "loss": 5.1818, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.04437869822485207, | |
| "grad_norm": 4.3337297439575195, | |
| "learning_rate": 0.00019142011834319528, | |
| "loss": 5.2133, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.04585798816568047, | |
| "grad_norm": 3.1711843013763428, | |
| "learning_rate": 0.0001911242603550296, | |
| "loss": 5.552, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.047337278106508875, | |
| "grad_norm": 3.5800342559814453, | |
| "learning_rate": 0.00019082840236686393, | |
| "loss": 5.4874, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.04881656804733728, | |
| "grad_norm": 3.3150734901428223, | |
| "learning_rate": 0.00019053254437869822, | |
| "loss": 5.5885, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.05029585798816568, | |
| "grad_norm": 4.20947790145874, | |
| "learning_rate": 0.00019023668639053256, | |
| "loss": 5.156, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.051775147928994084, | |
| "grad_norm": 8.529911041259766, | |
| "learning_rate": 0.00018994082840236687, | |
| "loss": 4.8622, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.05325443786982249, | |
| "grad_norm": 4.638500690460205, | |
| "learning_rate": 0.0001896449704142012, | |
| "loss": 5.4578, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.05473372781065089, | |
| "grad_norm": 4.114462375640869, | |
| "learning_rate": 0.0001893491124260355, | |
| "loss": 5.596, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.05621301775147929, | |
| "grad_norm": 3.748403310775757, | |
| "learning_rate": 0.00018905325443786984, | |
| "loss": 5.5348, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.057692307692307696, | |
| "grad_norm": 3.6923162937164307, | |
| "learning_rate": 0.00018875739644970416, | |
| "loss": 5.4819, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.05917159763313609, | |
| "grad_norm": 23.383670806884766, | |
| "learning_rate": 0.00018846153846153847, | |
| "loss": 5.0785, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.060650887573964495, | |
| "grad_norm": 3.344937324523926, | |
| "learning_rate": 0.00018816568047337278, | |
| "loss": 5.1104, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.0621301775147929, | |
| "grad_norm": 5.313603401184082, | |
| "learning_rate": 0.0001878698224852071, | |
| "loss": 5.5287, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.06360946745562131, | |
| "grad_norm": 4.41381311416626, | |
| "learning_rate": 0.00018757396449704144, | |
| "loss": 5.3752, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.0650887573964497, | |
| "grad_norm": 3.2099368572235107, | |
| "learning_rate": 0.00018727810650887573, | |
| "loss": 5.3049, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.06656804733727811, | |
| "grad_norm": 3.733710765838623, | |
| "learning_rate": 0.00018698224852071007, | |
| "loss": 5.1356, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.06804733727810651, | |
| "grad_norm": 3.271888494491577, | |
| "learning_rate": 0.0001866863905325444, | |
| "loss": 5.1147, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.0695266272189349, | |
| "grad_norm": 6.020442008972168, | |
| "learning_rate": 0.0001863905325443787, | |
| "loss": 5.7782, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.07100591715976332, | |
| "grad_norm": 3.2710890769958496, | |
| "learning_rate": 0.00018609467455621304, | |
| "loss": 5.3381, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.07248520710059171, | |
| "grad_norm": 3.6689212322235107, | |
| "learning_rate": 0.00018579881656804735, | |
| "loss": 5.0549, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.07396449704142012, | |
| "grad_norm": 3.1567306518554688, | |
| "learning_rate": 0.00018550295857988166, | |
| "loss": 5.0467, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07544378698224852, | |
| "grad_norm": 4.436924934387207, | |
| "learning_rate": 0.00018520710059171598, | |
| "loss": 5.1077, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 3.6421241760253906, | |
| "learning_rate": 0.00018491124260355032, | |
| "loss": 5.2728, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.07840236686390532, | |
| "grad_norm": 3.36057186126709, | |
| "learning_rate": 0.00018461538461538463, | |
| "loss": 4.8919, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.07988165680473373, | |
| "grad_norm": 3.2132999897003174, | |
| "learning_rate": 0.00018431952662721895, | |
| "loss": 5.1664, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.08136094674556213, | |
| "grad_norm": 3.4055607318878174, | |
| "learning_rate": 0.00018402366863905326, | |
| "loss": 5.193, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.08284023668639054, | |
| "grad_norm": 6.030857086181641, | |
| "learning_rate": 0.00018372781065088757, | |
| "loss": 5.1533, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.08431952662721894, | |
| "grad_norm": 3.724673271179199, | |
| "learning_rate": 0.00018343195266272192, | |
| "loss": 5.3483, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.08579881656804733, | |
| "grad_norm": 2.424447536468506, | |
| "learning_rate": 0.0001831360946745562, | |
| "loss": 5.2737, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.08727810650887574, | |
| "grad_norm": 4.9499430656433105, | |
| "learning_rate": 0.00018284023668639054, | |
| "loss": 4.7834, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.08875739644970414, | |
| "grad_norm": 2.8640456199645996, | |
| "learning_rate": 0.00018254437869822488, | |
| "loss": 5.204, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.09023668639053255, | |
| "grad_norm": 4.299962043762207, | |
| "learning_rate": 0.00018224852071005917, | |
| "loss": 4.9461, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.09171597633136094, | |
| "grad_norm": 3.068300724029541, | |
| "learning_rate": 0.0001819526627218935, | |
| "loss": 5.1019, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.09319526627218935, | |
| "grad_norm": 3.216639995574951, | |
| "learning_rate": 0.00018165680473372783, | |
| "loss": 5.43, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.09467455621301775, | |
| "grad_norm": 4.38277530670166, | |
| "learning_rate": 0.00018136094674556214, | |
| "loss": 5.3557, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.09615384615384616, | |
| "grad_norm": 2.8199565410614014, | |
| "learning_rate": 0.00018106508875739645, | |
| "loss": 5.4996, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.09763313609467456, | |
| "grad_norm": 4.008239269256592, | |
| "learning_rate": 0.00018076923076923077, | |
| "loss": 5.1045, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.09911242603550297, | |
| "grad_norm": 3.61327862739563, | |
| "learning_rate": 0.00018047337278106508, | |
| "loss": 5.2507, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.10059171597633136, | |
| "grad_norm": 4.783330917358398, | |
| "learning_rate": 0.00018017751479289942, | |
| "loss": 5.0782, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.10207100591715976, | |
| "grad_norm": 3.5379090309143066, | |
| "learning_rate": 0.00017988165680473374, | |
| "loss": 5.3468, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.10355029585798817, | |
| "grad_norm": 3.339912176132202, | |
| "learning_rate": 0.00017958579881656805, | |
| "loss": 5.2642, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.10502958579881656, | |
| "grad_norm": 3.801499605178833, | |
| "learning_rate": 0.0001792899408284024, | |
| "loss": 5.3676, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.10650887573964497, | |
| "grad_norm": 5.259742259979248, | |
| "learning_rate": 0.00017899408284023668, | |
| "loss": 5.0667, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.10798816568047337, | |
| "grad_norm": 3.887979507446289, | |
| "learning_rate": 0.00017869822485207102, | |
| "loss": 5.3508, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.10946745562130178, | |
| "grad_norm": 2.761725425720215, | |
| "learning_rate": 0.00017840236686390533, | |
| "loss": 5.3571, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.11094674556213018, | |
| "grad_norm": 4.331815719604492, | |
| "learning_rate": 0.00017810650887573965, | |
| "loss": 4.9632, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.11242603550295859, | |
| "grad_norm": 3.6547133922576904, | |
| "learning_rate": 0.000177810650887574, | |
| "loss": 5.1787, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.11390532544378698, | |
| "grad_norm": 3.22377347946167, | |
| "learning_rate": 0.00017751479289940828, | |
| "loss": 5.3842, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.11538461538461539, | |
| "grad_norm": 6.254403591156006, | |
| "learning_rate": 0.00017721893491124262, | |
| "loss": 5.1528, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.11686390532544379, | |
| "grad_norm": 3.523214101791382, | |
| "learning_rate": 0.00017692307692307693, | |
| "loss": 5.3192, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.11834319526627218, | |
| "grad_norm": 3.261544704437256, | |
| "learning_rate": 0.00017662721893491124, | |
| "loss": 5.0825, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.11982248520710059, | |
| "grad_norm": 3.994359016418457, | |
| "learning_rate": 0.00017633136094674556, | |
| "loss": 5.1593, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.12130177514792899, | |
| "grad_norm": 2.5975234508514404, | |
| "learning_rate": 0.0001760355029585799, | |
| "loss": 5.4471, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.1227810650887574, | |
| "grad_norm": 5.125651836395264, | |
| "learning_rate": 0.0001757396449704142, | |
| "loss": 4.8965, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.1242603550295858, | |
| "grad_norm": 4.02432918548584, | |
| "learning_rate": 0.00017544378698224853, | |
| "loss": 5.1502, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.1257396449704142, | |
| "grad_norm": 4.148054599761963, | |
| "learning_rate": 0.00017514792899408287, | |
| "loss": 4.9383, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.12721893491124261, | |
| "grad_norm": 4.669896125793457, | |
| "learning_rate": 0.00017485207100591716, | |
| "loss": 5.1252, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.128698224852071, | |
| "grad_norm": 3.3551809787750244, | |
| "learning_rate": 0.0001745562130177515, | |
| "loss": 4.8864, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.1301775147928994, | |
| "grad_norm": 4.218849182128906, | |
| "learning_rate": 0.0001742603550295858, | |
| "loss": 5.963, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.13165680473372782, | |
| "grad_norm": 3.2162094116210938, | |
| "learning_rate": 0.00017396449704142012, | |
| "loss": 5.0602, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.13313609467455623, | |
| "grad_norm": 3.9395463466644287, | |
| "learning_rate": 0.00017366863905325447, | |
| "loss": 5.1657, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1346153846153846, | |
| "grad_norm": 3.0270235538482666, | |
| "learning_rate": 0.00017337278106508875, | |
| "loss": 4.956, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.13609467455621302, | |
| "grad_norm": 4.6423726081848145, | |
| "learning_rate": 0.0001730769230769231, | |
| "loss": 5.2184, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.13757396449704143, | |
| "grad_norm": 3.070094347000122, | |
| "learning_rate": 0.0001727810650887574, | |
| "loss": 5.3557, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.1390532544378698, | |
| "grad_norm": 4.294172286987305, | |
| "learning_rate": 0.00017248520710059172, | |
| "loss": 4.9325, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.14053254437869822, | |
| "grad_norm": 2.961416721343994, | |
| "learning_rate": 0.00017218934911242603, | |
| "loss": 5.0524, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.14201183431952663, | |
| "grad_norm": 2.807966470718384, | |
| "learning_rate": 0.00017189349112426038, | |
| "loss": 4.9914, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.14349112426035504, | |
| "grad_norm": 4.755406379699707, | |
| "learning_rate": 0.0001715976331360947, | |
| "loss": 5.295, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.14497041420118342, | |
| "grad_norm": 2.498399257659912, | |
| "learning_rate": 0.000171301775147929, | |
| "loss": 5.3439, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.14644970414201183, | |
| "grad_norm": 3.2769603729248047, | |
| "learning_rate": 0.00017100591715976332, | |
| "loss": 5.0286, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.14792899408284024, | |
| "grad_norm": 3.3242433071136475, | |
| "learning_rate": 0.00017071005917159763, | |
| "loss": 5.2284, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.14940828402366865, | |
| "grad_norm": 3.9057908058166504, | |
| "learning_rate": 0.00017041420118343197, | |
| "loss": 5.1046, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.15088757396449703, | |
| "grad_norm": 2.3905794620513916, | |
| "learning_rate": 0.00017011834319526626, | |
| "loss": 4.9429, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.15236686390532544, | |
| "grad_norm": 2.85048246383667, | |
| "learning_rate": 0.0001698224852071006, | |
| "loss": 5.1951, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 3.160964012145996, | |
| "learning_rate": 0.00016952662721893494, | |
| "loss": 5.0117, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.15532544378698224, | |
| "grad_norm": 6.133135795593262, | |
| "learning_rate": 0.00016923076923076923, | |
| "loss": 5.3851, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.15680473372781065, | |
| "grad_norm": 4.323611259460449, | |
| "learning_rate": 0.00016893491124260357, | |
| "loss": 4.9845, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.15828402366863906, | |
| "grad_norm": 3.989495038986206, | |
| "learning_rate": 0.00016863905325443788, | |
| "loss": 5.0833, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.15976331360946747, | |
| "grad_norm": 2.9743452072143555, | |
| "learning_rate": 0.0001683431952662722, | |
| "loss": 5.4684, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.16124260355029585, | |
| "grad_norm": 3.270275592803955, | |
| "learning_rate": 0.0001680473372781065, | |
| "loss": 4.861, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.16272189349112426, | |
| "grad_norm": 3.555215358734131, | |
| "learning_rate": 0.00016775147928994083, | |
| "loss": 5.0709, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.16420118343195267, | |
| "grad_norm": 4.3059306144714355, | |
| "learning_rate": 0.00016745562130177514, | |
| "loss": 5.5366, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.16568047337278108, | |
| "grad_norm": 3.379615068435669, | |
| "learning_rate": 0.00016715976331360948, | |
| "loss": 4.9129, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.16715976331360946, | |
| "grad_norm": 5.602575302124023, | |
| "learning_rate": 0.0001668639053254438, | |
| "loss": 4.8603, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.16863905325443787, | |
| "grad_norm": 2.965013265609741, | |
| "learning_rate": 0.0001665680473372781, | |
| "loss": 5.2722, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.17011834319526628, | |
| "grad_norm": 2.8539206981658936, | |
| "learning_rate": 0.00016627218934911245, | |
| "loss": 5.217, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.17159763313609466, | |
| "grad_norm": 2.861959457397461, | |
| "learning_rate": 0.00016597633136094674, | |
| "loss": 4.9753, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.17307692307692307, | |
| "grad_norm": 3.451565742492676, | |
| "learning_rate": 0.00016568047337278108, | |
| "loss": 4.9347, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.17455621301775148, | |
| "grad_norm": 2.7960565090179443, | |
| "learning_rate": 0.0001653846153846154, | |
| "loss": 5.2664, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.1760355029585799, | |
| "grad_norm": 3.4076054096221924, | |
| "learning_rate": 0.0001650887573964497, | |
| "loss": 4.8754, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.17751479289940827, | |
| "grad_norm": 3.4976394176483154, | |
| "learning_rate": 0.00016479289940828405, | |
| "loss": 5.0685, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.17899408284023668, | |
| "grad_norm": 2.982083320617676, | |
| "learning_rate": 0.00016449704142011836, | |
| "loss": 5.2266, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.1804733727810651, | |
| "grad_norm": 3.0358190536499023, | |
| "learning_rate": 0.00016420118343195267, | |
| "loss": 5.1363, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.1819526627218935, | |
| "grad_norm": 3.6040701866149902, | |
| "learning_rate": 0.000163905325443787, | |
| "loss": 4.8832, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.1834319526627219, | |
| "grad_norm": 4.876418590545654, | |
| "learning_rate": 0.0001636094674556213, | |
| "loss": 4.7069, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.1849112426035503, | |
| "grad_norm": 6.96181058883667, | |
| "learning_rate": 0.00016331360946745562, | |
| "loss": 4.6315, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.1863905325443787, | |
| "grad_norm": 2.648008108139038, | |
| "learning_rate": 0.00016301775147928996, | |
| "loss": 5.3977, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.1878698224852071, | |
| "grad_norm": 3.0898566246032715, | |
| "learning_rate": 0.00016272189349112427, | |
| "loss": 4.9533, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.1893491124260355, | |
| "grad_norm": 3.143807888031006, | |
| "learning_rate": 0.00016242603550295858, | |
| "loss": 5.3733, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.1908284023668639, | |
| "grad_norm": 3.230045795440674, | |
| "learning_rate": 0.00016213017751479293, | |
| "loss": 4.8937, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.19230769230769232, | |
| "grad_norm": 3.5977890491485596, | |
| "learning_rate": 0.0001618343195266272, | |
| "loss": 5.0298, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.1937869822485207, | |
| "grad_norm": 3.7372870445251465, | |
| "learning_rate": 0.00016153846153846155, | |
| "loss": 4.8582, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.1952662721893491, | |
| "grad_norm": 3.4929022789001465, | |
| "learning_rate": 0.00016124260355029587, | |
| "loss": 5.043, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.19674556213017752, | |
| "grad_norm": 3.389169931411743, | |
| "learning_rate": 0.00016094674556213018, | |
| "loss": 5.196, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.19822485207100593, | |
| "grad_norm": 5.478149890899658, | |
| "learning_rate": 0.00016065088757396452, | |
| "loss": 4.8719, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.1997041420118343, | |
| "grad_norm": 2.303262233734131, | |
| "learning_rate": 0.0001603550295857988, | |
| "loss": 5.1638, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.20118343195266272, | |
| "grad_norm": 3.297595739364624, | |
| "learning_rate": 0.00016005917159763315, | |
| "loss": 4.8086, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.20266272189349113, | |
| "grad_norm": 3.509148120880127, | |
| "learning_rate": 0.00015976331360946746, | |
| "loss": 4.8932, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.20414201183431951, | |
| "grad_norm": 3.4905240535736084, | |
| "learning_rate": 0.00015946745562130178, | |
| "loss": 5.0621, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.20562130177514792, | |
| "grad_norm": 3.4399490356445312, | |
| "learning_rate": 0.0001591715976331361, | |
| "loss": 5.1894, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.20710059171597633, | |
| "grad_norm": 3.0734777450561523, | |
| "learning_rate": 0.00015887573964497043, | |
| "loss": 4.9211, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.20857988165680474, | |
| "grad_norm": 3.179959297180176, | |
| "learning_rate": 0.00015857988165680475, | |
| "loss": 5.2447, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.21005917159763313, | |
| "grad_norm": 4.863833427429199, | |
| "learning_rate": 0.00015828402366863906, | |
| "loss": 5.0381, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.21153846153846154, | |
| "grad_norm": 5.0287556648254395, | |
| "learning_rate": 0.00015798816568047337, | |
| "loss": 4.7826, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.21301775147928995, | |
| "grad_norm": 2.806042194366455, | |
| "learning_rate": 0.0001576923076923077, | |
| "loss": 5.1359, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.21449704142011836, | |
| "grad_norm": 3.897864818572998, | |
| "learning_rate": 0.00015739644970414203, | |
| "loss": 5.2164, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.21597633136094674, | |
| "grad_norm": 3.212893009185791, | |
| "learning_rate": 0.00015710059171597634, | |
| "loss": 4.9873, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.21745562130177515, | |
| "grad_norm": 3.277351140975952, | |
| "learning_rate": 0.00015680473372781066, | |
| "loss": 4.9393, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.21893491124260356, | |
| "grad_norm": 3.687052011489868, | |
| "learning_rate": 0.000156508875739645, | |
| "loss": 5.0331, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.22041420118343194, | |
| "grad_norm": 2.938491106033325, | |
| "learning_rate": 0.00015621301775147929, | |
| "loss": 4.9259, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.22189349112426035, | |
| "grad_norm": 3.9487521648406982, | |
| "learning_rate": 0.00015591715976331363, | |
| "loss": 4.7096, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.22337278106508876, | |
| "grad_norm": 3.2990753650665283, | |
| "learning_rate": 0.00015562130177514794, | |
| "loss": 5.2393, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.22485207100591717, | |
| "grad_norm": 4.3339762687683105, | |
| "learning_rate": 0.00015532544378698225, | |
| "loss": 4.6888, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.22633136094674555, | |
| "grad_norm": 2.981172561645508, | |
| "learning_rate": 0.00015502958579881657, | |
| "loss": 5.1174, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.22781065088757396, | |
| "grad_norm": 3.2095184326171875, | |
| "learning_rate": 0.0001547337278106509, | |
| "loss": 5.5025, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.22928994082840237, | |
| "grad_norm": 3.3522627353668213, | |
| "learning_rate": 0.0001544378698224852, | |
| "loss": 5.0102, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 3.347395658493042, | |
| "learning_rate": 0.00015414201183431954, | |
| "loss": 5.3176, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.23224852071005916, | |
| "grad_norm": 4.909267902374268, | |
| "learning_rate": 0.00015384615384615385, | |
| "loss": 4.7972, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.23372781065088757, | |
| "grad_norm": 3.8333449363708496, | |
| "learning_rate": 0.00015355029585798817, | |
| "loss": 4.829, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.23520710059171598, | |
| "grad_norm": 4.241271495819092, | |
| "learning_rate": 0.0001532544378698225, | |
| "loss": 4.9222, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.23668639053254437, | |
| "grad_norm": 2.662621021270752, | |
| "learning_rate": 0.0001529585798816568, | |
| "loss": 4.8007, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.23816568047337278, | |
| "grad_norm": 4.020458221435547, | |
| "learning_rate": 0.00015266272189349113, | |
| "loss": 5.0829, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.23964497041420119, | |
| "grad_norm": 3.640207052230835, | |
| "learning_rate": 0.00015236686390532545, | |
| "loss": 5.0856, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.2411242603550296, | |
| "grad_norm": 3.3714780807495117, | |
| "learning_rate": 0.00015207100591715976, | |
| "loss": 5.0783, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.24260355029585798, | |
| "grad_norm": 3.97548508644104, | |
| "learning_rate": 0.0001517751479289941, | |
| "loss": 4.8682, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.2440828402366864, | |
| "grad_norm": 2.2095658779144287, | |
| "learning_rate": 0.00015147928994082842, | |
| "loss": 5.2263, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.2455621301775148, | |
| "grad_norm": 3.80568265914917, | |
| "learning_rate": 0.00015118343195266273, | |
| "loss": 4.7762, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.2470414201183432, | |
| "grad_norm": 3.4083971977233887, | |
| "learning_rate": 0.00015088757396449705, | |
| "loss": 5.2287, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.2485207100591716, | |
| "grad_norm": 5.997743606567383, | |
| "learning_rate": 0.00015059171597633136, | |
| "loss": 4.7403, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 3.7599215507507324, | |
| "learning_rate": 0.00015029585798816567, | |
| "loss": 4.7957, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.2514792899408284, | |
| "grad_norm": 2.9996368885040283, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 5.0695, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2529585798816568, | |
| "grad_norm": 3.03244686126709, | |
| "learning_rate": 0.00014970414201183433, | |
| "loss": 4.7598, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.25443786982248523, | |
| "grad_norm": 2.692690372467041, | |
| "learning_rate": 0.00014940828402366864, | |
| "loss": 5.17, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.2559171597633136, | |
| "grad_norm": 4.609525203704834, | |
| "learning_rate": 0.00014911242603550298, | |
| "loss": 4.8048, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.257396449704142, | |
| "grad_norm": 4.3488264083862305, | |
| "learning_rate": 0.00014881656804733727, | |
| "loss": 5.0132, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.2588757396449704, | |
| "grad_norm": 4.0892558097839355, | |
| "learning_rate": 0.0001485207100591716, | |
| "loss": 5.0826, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.2603550295857988, | |
| "grad_norm": 3.7176010608673096, | |
| "learning_rate": 0.00014822485207100592, | |
| "loss": 5.0371, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.2618343195266272, | |
| "grad_norm": 2.8590428829193115, | |
| "learning_rate": 0.00014792899408284024, | |
| "loss": 5.201, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.26331360946745563, | |
| "grad_norm": 4.960610866546631, | |
| "learning_rate": 0.00014763313609467458, | |
| "loss": 4.7447, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.26479289940828404, | |
| "grad_norm": 4.061326503753662, | |
| "learning_rate": 0.0001473372781065089, | |
| "loss": 5.2911, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.26627218934911245, | |
| "grad_norm": 3.747265338897705, | |
| "learning_rate": 0.0001470414201183432, | |
| "loss": 4.8774, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2677514792899408, | |
| "grad_norm": 3.293882369995117, | |
| "learning_rate": 0.00014674556213017752, | |
| "loss": 5.1072, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.2692307692307692, | |
| "grad_norm": 3.514312505722046, | |
| "learning_rate": 0.00014644970414201184, | |
| "loss": 4.8875, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.27071005917159763, | |
| "grad_norm": 3.192707061767578, | |
| "learning_rate": 0.00014615384615384615, | |
| "loss": 4.7942, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.27218934911242604, | |
| "grad_norm": 4.239992141723633, | |
| "learning_rate": 0.0001458579881656805, | |
| "loss": 4.8851, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.27366863905325445, | |
| "grad_norm": 3.4382450580596924, | |
| "learning_rate": 0.0001455621301775148, | |
| "loss": 5.0304, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.27514792899408286, | |
| "grad_norm": 3.2028238773345947, | |
| "learning_rate": 0.00014526627218934912, | |
| "loss": 4.947, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.27662721893491127, | |
| "grad_norm": 3.0747134685516357, | |
| "learning_rate": 0.00014497041420118346, | |
| "loss": 4.9979, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.2781065088757396, | |
| "grad_norm": 3.7282187938690186, | |
| "learning_rate": 0.00014467455621301775, | |
| "loss": 4.8038, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.27958579881656803, | |
| "grad_norm": 3.0893115997314453, | |
| "learning_rate": 0.0001443786982248521, | |
| "loss": 5.1064, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.28106508875739644, | |
| "grad_norm": 4.98429012298584, | |
| "learning_rate": 0.0001440828402366864, | |
| "loss": 5.7315, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.28254437869822485, | |
| "grad_norm": 3.111341714859009, | |
| "learning_rate": 0.00014378698224852072, | |
| "loss": 5.0315, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.28402366863905326, | |
| "grad_norm": 3.6870968341827393, | |
| "learning_rate": 0.00014349112426035503, | |
| "loss": 5.0105, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.28550295857988167, | |
| "grad_norm": 4.709840774536133, | |
| "learning_rate": 0.00014319526627218934, | |
| "loss": 5.0156, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.2869822485207101, | |
| "grad_norm": 2.8306596279144287, | |
| "learning_rate": 0.00014289940828402368, | |
| "loss": 4.9832, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.28846153846153844, | |
| "grad_norm": 4.066086292266846, | |
| "learning_rate": 0.000142603550295858, | |
| "loss": 4.886, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.28994082840236685, | |
| "grad_norm": 3.7111706733703613, | |
| "learning_rate": 0.0001423076923076923, | |
| "loss": 4.868, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.29142011834319526, | |
| "grad_norm": 2.9184136390686035, | |
| "learning_rate": 0.00014201183431952663, | |
| "loss": 5.1351, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.29289940828402367, | |
| "grad_norm": 3.7855818271636963, | |
| "learning_rate": 0.00014171597633136097, | |
| "loss": 4.6282, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.2943786982248521, | |
| "grad_norm": 2.7705442905426025, | |
| "learning_rate": 0.00014142011834319525, | |
| "loss": 4.9413, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.2958579881656805, | |
| "grad_norm": 2.315896511077881, | |
| "learning_rate": 0.0001411242603550296, | |
| "loss": 5.1848, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2973372781065089, | |
| "grad_norm": 2.548875331878662, | |
| "learning_rate": 0.0001408284023668639, | |
| "loss": 5.1358, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.2988165680473373, | |
| "grad_norm": 2.73306941986084, | |
| "learning_rate": 0.00014053254437869822, | |
| "loss": 4.9356, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.30029585798816566, | |
| "grad_norm": 5.324806213378906, | |
| "learning_rate": 0.00014023668639053256, | |
| "loss": 4.8917, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.30177514792899407, | |
| "grad_norm": 3.943690299987793, | |
| "learning_rate": 0.00013994082840236685, | |
| "loss": 4.8406, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.3032544378698225, | |
| "grad_norm": 3.848397970199585, | |
| "learning_rate": 0.0001396449704142012, | |
| "loss": 4.7171, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.3047337278106509, | |
| "grad_norm": 3.5584487915039062, | |
| "learning_rate": 0.0001393491124260355, | |
| "loss": 4.8052, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.3062130177514793, | |
| "grad_norm": 3.0798633098602295, | |
| "learning_rate": 0.00013905325443786982, | |
| "loss": 4.959, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 4.075172424316406, | |
| "learning_rate": 0.00013875739644970416, | |
| "loss": 4.6734, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.3091715976331361, | |
| "grad_norm": 2.977128505706787, | |
| "learning_rate": 0.00013846153846153847, | |
| "loss": 5.128, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.3106508875739645, | |
| "grad_norm": 2.536226987838745, | |
| "learning_rate": 0.0001381656804733728, | |
| "loss": 4.9083, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3121301775147929, | |
| "grad_norm": 2.7597198486328125, | |
| "learning_rate": 0.0001378698224852071, | |
| "loss": 5.0091, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.3136094674556213, | |
| "grad_norm": 4.09480619430542, | |
| "learning_rate": 0.00013757396449704144, | |
| "loss": 4.6064, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.3150887573964497, | |
| "grad_norm": 4.707205772399902, | |
| "learning_rate": 0.00013727810650887573, | |
| "loss": 5.1816, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.3165680473372781, | |
| "grad_norm": 3.8236589431762695, | |
| "learning_rate": 0.00013698224852071007, | |
| "loss": 4.8127, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.3180473372781065, | |
| "grad_norm": 3.736663341522217, | |
| "learning_rate": 0.00013668639053254439, | |
| "loss": 5.0851, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.31952662721893493, | |
| "grad_norm": 3.5251119136810303, | |
| "learning_rate": 0.0001363905325443787, | |
| "loss": 4.7787, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.3210059171597633, | |
| "grad_norm": 4.254045009613037, | |
| "learning_rate": 0.00013609467455621304, | |
| "loss": 4.6852, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.3224852071005917, | |
| "grad_norm": 2.4598679542541504, | |
| "learning_rate": 0.00013579881656804733, | |
| "loss": 5.1052, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.3239644970414201, | |
| "grad_norm": 3.000013828277588, | |
| "learning_rate": 0.00013550295857988167, | |
| "loss": 5.092, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.3254437869822485, | |
| "grad_norm": 3.0184524059295654, | |
| "learning_rate": 0.00013520710059171598, | |
| "loss": 5.1709, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3269230769230769, | |
| "grad_norm": 4.693943977355957, | |
| "learning_rate": 0.0001349112426035503, | |
| "loss": 5.0312, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.32840236686390534, | |
| "grad_norm": 4.594445705413818, | |
| "learning_rate": 0.00013461538461538464, | |
| "loss": 4.884, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.32988165680473375, | |
| "grad_norm": 2.1385953426361084, | |
| "learning_rate": 0.00013431952662721895, | |
| "loss": 5.1498, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.33136094674556216, | |
| "grad_norm": 3.80926251411438, | |
| "learning_rate": 0.00013402366863905326, | |
| "loss": 4.7281, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.3328402366863905, | |
| "grad_norm": 2.7746124267578125, | |
| "learning_rate": 0.00013372781065088758, | |
| "loss": 5.1611, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.3343195266272189, | |
| "grad_norm": 4.454751014709473, | |
| "learning_rate": 0.0001334319526627219, | |
| "loss": 4.6656, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.33579881656804733, | |
| "grad_norm": 3.0681076049804688, | |
| "learning_rate": 0.0001331360946745562, | |
| "loss": 4.947, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.33727810650887574, | |
| "grad_norm": 4.087713718414307, | |
| "learning_rate": 0.00013284023668639055, | |
| "loss": 4.659, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.33875739644970415, | |
| "grad_norm": 2.7272188663482666, | |
| "learning_rate": 0.00013254437869822486, | |
| "loss": 4.875, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.34023668639053256, | |
| "grad_norm": 4.372305393218994, | |
| "learning_rate": 0.00013224852071005918, | |
| "loss": 5.0376, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.34171597633136097, | |
| "grad_norm": 2.4269564151763916, | |
| "learning_rate": 0.00013195266272189352, | |
| "loss": 4.865, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.3431952662721893, | |
| "grad_norm": 2.970414876937866, | |
| "learning_rate": 0.0001316568047337278, | |
| "loss": 5.1939, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.34467455621301774, | |
| "grad_norm": 2.476336717605591, | |
| "learning_rate": 0.00013136094674556214, | |
| "loss": 4.8974, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.34615384615384615, | |
| "grad_norm": 2.5719239711761475, | |
| "learning_rate": 0.00013106508875739646, | |
| "loss": 4.7806, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.34763313609467456, | |
| "grad_norm": 3.170459747314453, | |
| "learning_rate": 0.00013076923076923077, | |
| "loss": 4.841, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.34911242603550297, | |
| "grad_norm": 3.061983108520508, | |
| "learning_rate": 0.00013047337278106509, | |
| "loss": 5.1059, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.3505917159763314, | |
| "grad_norm": 3.1553242206573486, | |
| "learning_rate": 0.0001301775147928994, | |
| "loss": 4.8559, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.3520710059171598, | |
| "grad_norm": 4.066651344299316, | |
| "learning_rate": 0.00012988165680473374, | |
| "loss": 4.8189, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.35355029585798814, | |
| "grad_norm": 2.3812334537506104, | |
| "learning_rate": 0.00012958579881656806, | |
| "loss": 4.8804, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.35502958579881655, | |
| "grad_norm": 4.674489974975586, | |
| "learning_rate": 0.00012928994082840237, | |
| "loss": 5.0399, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.35650887573964496, | |
| "grad_norm": 3.953132390975952, | |
| "learning_rate": 0.00012899408284023668, | |
| "loss": 5.3266, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.35798816568047337, | |
| "grad_norm": 3.108064651489258, | |
| "learning_rate": 0.00012869822485207102, | |
| "loss": 5.032, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.3594674556213018, | |
| "grad_norm": 3.2447121143341064, | |
| "learning_rate": 0.0001284023668639053, | |
| "loss": 4.9344, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.3609467455621302, | |
| "grad_norm": 6.096102714538574, | |
| "learning_rate": 0.00012810650887573965, | |
| "loss": 4.7905, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.3624260355029586, | |
| "grad_norm": 3.0096826553344727, | |
| "learning_rate": 0.00012781065088757397, | |
| "loss": 4.663, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.363905325443787, | |
| "grad_norm": 3.5750575065612793, | |
| "learning_rate": 0.00012751479289940828, | |
| "loss": 4.7274, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.36538461538461536, | |
| "grad_norm": 2.610539674758911, | |
| "learning_rate": 0.00012721893491124262, | |
| "loss": 4.9819, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.3668639053254438, | |
| "grad_norm": 2.911921977996826, | |
| "learning_rate": 0.00012692307692307693, | |
| "loss": 4.5621, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.3683431952662722, | |
| "grad_norm": 3.698395252227783, | |
| "learning_rate": 0.00012662721893491125, | |
| "loss": 4.8973, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.3698224852071006, | |
| "grad_norm": 2.5374014377593994, | |
| "learning_rate": 0.00012633136094674556, | |
| "loss": 5.0435, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.371301775147929, | |
| "grad_norm": 2.8798906803131104, | |
| "learning_rate": 0.00012603550295857988, | |
| "loss": 4.928, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.3727810650887574, | |
| "grad_norm": 3.2215588092803955, | |
| "learning_rate": 0.00012573964497041422, | |
| "loss": 5.0017, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.3742603550295858, | |
| "grad_norm": 3.9945900440216064, | |
| "learning_rate": 0.00012544378698224853, | |
| "loss": 4.9283, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.3757396449704142, | |
| "grad_norm": 3.116191864013672, | |
| "learning_rate": 0.00012514792899408285, | |
| "loss": 5.1576, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.3772189349112426, | |
| "grad_norm": 4.140561103820801, | |
| "learning_rate": 0.00012485207100591716, | |
| "loss": 4.6843, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.378698224852071, | |
| "grad_norm": 3.4369871616363525, | |
| "learning_rate": 0.0001245562130177515, | |
| "loss": 4.7908, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.3801775147928994, | |
| "grad_norm": 3.218079090118408, | |
| "learning_rate": 0.0001242603550295858, | |
| "loss": 5.1624, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.3816568047337278, | |
| "grad_norm": 2.8588600158691406, | |
| "learning_rate": 0.00012396449704142013, | |
| "loss": 4.9308, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.3831360946745562, | |
| "grad_norm": 3.2073476314544678, | |
| "learning_rate": 0.00012366863905325444, | |
| "loss": 5.0278, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 3.6556711196899414, | |
| "learning_rate": 0.00012337278106508876, | |
| "loss": 4.6337, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.386094674556213, | |
| "grad_norm": 3.5764808654785156, | |
| "learning_rate": 0.0001230769230769231, | |
| "loss": 4.7867, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.3875739644970414, | |
| "grad_norm": 2.8595478534698486, | |
| "learning_rate": 0.00012278106508875738, | |
| "loss": 4.7535, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.3890532544378698, | |
| "grad_norm": 2.653027296066284, | |
| "learning_rate": 0.00012248520710059173, | |
| "loss": 5.0196, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.3905325443786982, | |
| "grad_norm": 5.582137584686279, | |
| "learning_rate": 0.00012218934911242604, | |
| "loss": 5.0398, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.39201183431952663, | |
| "grad_norm": 2.5141005516052246, | |
| "learning_rate": 0.00012189349112426037, | |
| "loss": 4.7242, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.39349112426035504, | |
| "grad_norm": 2.708251476287842, | |
| "learning_rate": 0.0001215976331360947, | |
| "loss": 5.2731, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.39497041420118345, | |
| "grad_norm": 3.10066294670105, | |
| "learning_rate": 0.000121301775147929, | |
| "loss": 5.0398, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.39644970414201186, | |
| "grad_norm": 3.866253137588501, | |
| "learning_rate": 0.00012100591715976332, | |
| "loss": 4.7469, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.3979289940828402, | |
| "grad_norm": 1.983142614364624, | |
| "learning_rate": 0.00012071005917159764, | |
| "loss": 5.1332, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.3994082840236686, | |
| "grad_norm": 4.262152671813965, | |
| "learning_rate": 0.00012041420118343196, | |
| "loss": 4.8862, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.40088757396449703, | |
| "grad_norm": 3.2552340030670166, | |
| "learning_rate": 0.00012011834319526626, | |
| "loss": 4.7106, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.40236686390532544, | |
| "grad_norm": 3.415714740753174, | |
| "learning_rate": 0.0001198224852071006, | |
| "loss": 4.7708, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.40384615384615385, | |
| "grad_norm": 3.799057960510254, | |
| "learning_rate": 0.00011952662721893493, | |
| "loss": 4.8023, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.40532544378698226, | |
| "grad_norm": 4.48440408706665, | |
| "learning_rate": 0.00011923076923076923, | |
| "loss": 4.6143, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.4068047337278107, | |
| "grad_norm": 2.993614435195923, | |
| "learning_rate": 0.00011893491124260356, | |
| "loss": 5.0484, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.40828402366863903, | |
| "grad_norm": 3.1117966175079346, | |
| "learning_rate": 0.00011863905325443787, | |
| "loss": 5.0049, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.40976331360946744, | |
| "grad_norm": 2.981684446334839, | |
| "learning_rate": 0.0001183431952662722, | |
| "loss": 4.9651, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.41124260355029585, | |
| "grad_norm": 4.644701957702637, | |
| "learning_rate": 0.0001180473372781065, | |
| "loss": 5.1983, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.41272189349112426, | |
| "grad_norm": 3.9957146644592285, | |
| "learning_rate": 0.00011775147928994083, | |
| "loss": 4.7752, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.41420118343195267, | |
| "grad_norm": 2.156813144683838, | |
| "learning_rate": 0.00011745562130177514, | |
| "loss": 4.9941, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.4156804733727811, | |
| "grad_norm": 3.0435688495635986, | |
| "learning_rate": 0.00011715976331360947, | |
| "loss": 5.1542, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.4171597633136095, | |
| "grad_norm": 2.6798031330108643, | |
| "learning_rate": 0.0001168639053254438, | |
| "loss": 5.0989, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.41863905325443784, | |
| "grad_norm": 4.525575637817383, | |
| "learning_rate": 0.00011656804733727811, | |
| "loss": 4.8101, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.42011834319526625, | |
| "grad_norm": 2.770167827606201, | |
| "learning_rate": 0.00011627218934911244, | |
| "loss": 4.8678, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.42159763313609466, | |
| "grad_norm": 3.7138099670410156, | |
| "learning_rate": 0.00011597633136094674, | |
| "loss": 4.5939, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.4230769230769231, | |
| "grad_norm": 2.398003339767456, | |
| "learning_rate": 0.00011568047337278107, | |
| "loss": 4.8854, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.4245562130177515, | |
| "grad_norm": 2.4538278579711914, | |
| "learning_rate": 0.00011538461538461538, | |
| "loss": 4.8346, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.4260355029585799, | |
| "grad_norm": 2.757596969604492, | |
| "learning_rate": 0.00011508875739644971, | |
| "loss": 4.8026, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.4275147928994083, | |
| "grad_norm": 2.84289813041687, | |
| "learning_rate": 0.00011479289940828404, | |
| "loss": 5.1102, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.4289940828402367, | |
| "grad_norm": 3.8550050258636475, | |
| "learning_rate": 0.00011449704142011835, | |
| "loss": 4.7836, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.43047337278106507, | |
| "grad_norm": 2.811088800430298, | |
| "learning_rate": 0.00011420118343195268, | |
| "loss": 4.7735, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.4319526627218935, | |
| "grad_norm": 3.845970392227173, | |
| "learning_rate": 0.00011390532544378698, | |
| "loss": 4.5021, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.4334319526627219, | |
| "grad_norm": 2.977811574935913, | |
| "learning_rate": 0.0001136094674556213, | |
| "loss": 5.0106, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.4349112426035503, | |
| "grad_norm": 3.136218309402466, | |
| "learning_rate": 0.00011331360946745562, | |
| "loss": 4.9972, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.4363905325443787, | |
| "grad_norm": 3.3354506492614746, | |
| "learning_rate": 0.00011301775147928995, | |
| "loss": 4.7461, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.4378698224852071, | |
| "grad_norm": 2.847717761993408, | |
| "learning_rate": 0.00011272189349112428, | |
| "loss": 5.0909, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.4393491124260355, | |
| "grad_norm": 3.3575217723846436, | |
| "learning_rate": 0.00011242603550295858, | |
| "loss": 4.6807, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.4408284023668639, | |
| "grad_norm": 4.795094966888428, | |
| "learning_rate": 0.00011213017751479292, | |
| "loss": 4.6449, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.4423076923076923, | |
| "grad_norm": 3.1832058429718018, | |
| "learning_rate": 0.00011183431952662722, | |
| "loss": 5.1715, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.4437869822485207, | |
| "grad_norm": 2.93605899810791, | |
| "learning_rate": 0.00011153846153846154, | |
| "loss": 4.743, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4452662721893491, | |
| "grad_norm": 3.169785737991333, | |
| "learning_rate": 0.00011124260355029586, | |
| "loss": 4.8851, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.4467455621301775, | |
| "grad_norm": 3.1830434799194336, | |
| "learning_rate": 0.00011094674556213019, | |
| "loss": 4.6635, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.44822485207100593, | |
| "grad_norm": 3.0651752948760986, | |
| "learning_rate": 0.00011065088757396451, | |
| "loss": 4.9519, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.44970414201183434, | |
| "grad_norm": 2.6180307865142822, | |
| "learning_rate": 0.00011035502958579881, | |
| "loss": 4.8693, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.4511834319526627, | |
| "grad_norm": 2.6694605350494385, | |
| "learning_rate": 0.00011005917159763315, | |
| "loss": 4.9804, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.4526627218934911, | |
| "grad_norm": 4.2127299308776855, | |
| "learning_rate": 0.00010976331360946746, | |
| "loss": 4.9376, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.4541420118343195, | |
| "grad_norm": 3.186117172241211, | |
| "learning_rate": 0.00010946745562130178, | |
| "loss": 4.9848, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.4556213017751479, | |
| "grad_norm": 5.589068412780762, | |
| "learning_rate": 0.0001091715976331361, | |
| "loss": 4.603, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.45710059171597633, | |
| "grad_norm": 3.5685253143310547, | |
| "learning_rate": 0.00010887573964497042, | |
| "loss": 4.6498, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.45857988165680474, | |
| "grad_norm": 3.711947441101074, | |
| "learning_rate": 0.00010857988165680475, | |
| "loss": 4.8806, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.46005917159763315, | |
| "grad_norm": 3.018235683441162, | |
| "learning_rate": 0.00010828402366863905, | |
| "loss": 4.9364, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 3.832859754562378, | |
| "learning_rate": 0.00010798816568047338, | |
| "loss": 4.5133, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.4630177514792899, | |
| "grad_norm": 3.1703977584838867, | |
| "learning_rate": 0.0001076923076923077, | |
| "loss": 4.9799, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.46449704142011833, | |
| "grad_norm": 3.1763527393341064, | |
| "learning_rate": 0.00010739644970414202, | |
| "loss": 5.1171, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.46597633136094674, | |
| "grad_norm": 3.3300058841705322, | |
| "learning_rate": 0.00010710059171597633, | |
| "loss": 5.0527, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.46745562130177515, | |
| "grad_norm": 4.077115535736084, | |
| "learning_rate": 0.00010680473372781066, | |
| "loss": 5.1883, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.46893491124260356, | |
| "grad_norm": 2.77453875541687, | |
| "learning_rate": 0.00010650887573964499, | |
| "loss": 4.7195, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.47041420118343197, | |
| "grad_norm": 3.7889904975891113, | |
| "learning_rate": 0.00010621301775147929, | |
| "loss": 4.6898, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.4718934911242604, | |
| "grad_norm": 3.072606325149536, | |
| "learning_rate": 0.00010591715976331362, | |
| "loss": 4.6968, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.47337278106508873, | |
| "grad_norm": 3.511852741241455, | |
| "learning_rate": 0.00010562130177514793, | |
| "loss": 5.2093, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.47485207100591714, | |
| "grad_norm": 3.1733694076538086, | |
| "learning_rate": 0.00010532544378698226, | |
| "loss": 4.708, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.47633136094674555, | |
| "grad_norm": 3.6563074588775635, | |
| "learning_rate": 0.00010502958579881656, | |
| "loss": 4.9192, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.47781065088757396, | |
| "grad_norm": 3.6182966232299805, | |
| "learning_rate": 0.0001047337278106509, | |
| "loss": 4.4543, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.47928994082840237, | |
| "grad_norm": 2.6363067626953125, | |
| "learning_rate": 0.0001044378698224852, | |
| "loss": 4.8366, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.4807692307692308, | |
| "grad_norm": 3.9731101989746094, | |
| "learning_rate": 0.00010414201183431953, | |
| "loss": 4.3, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.4822485207100592, | |
| "grad_norm": 3.4638442993164062, | |
| "learning_rate": 0.00010384615384615386, | |
| "loss": 4.7362, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.48372781065088755, | |
| "grad_norm": 3.5232033729553223, | |
| "learning_rate": 0.00010355029585798817, | |
| "loss": 4.6581, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.48520710059171596, | |
| "grad_norm": 3.3872711658477783, | |
| "learning_rate": 0.0001032544378698225, | |
| "loss": 5.115, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.48668639053254437, | |
| "grad_norm": 4.918847560882568, | |
| "learning_rate": 0.0001029585798816568, | |
| "loss": 4.7896, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.4881656804733728, | |
| "grad_norm": 2.2880265712738037, | |
| "learning_rate": 0.00010266272189349113, | |
| "loss": 4.9476, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.4896449704142012, | |
| "grad_norm": 2.6146740913391113, | |
| "learning_rate": 0.00010236686390532544, | |
| "loss": 4.8704, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.4911242603550296, | |
| "grad_norm": 4.03132438659668, | |
| "learning_rate": 0.00010207100591715977, | |
| "loss": 4.6685, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.492603550295858, | |
| "grad_norm": 2.9792962074279785, | |
| "learning_rate": 0.0001017751479289941, | |
| "loss": 4.6436, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.4940828402366864, | |
| "grad_norm": 4.724271774291992, | |
| "learning_rate": 0.00010147928994082841, | |
| "loss": 4.7177, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.49556213017751477, | |
| "grad_norm": 2.644087076187134, | |
| "learning_rate": 0.00010118343195266274, | |
| "loss": 4.8767, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.4970414201183432, | |
| "grad_norm": 3.115257740020752, | |
| "learning_rate": 0.00010088757396449704, | |
| "loss": 4.789, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.4985207100591716, | |
| "grad_norm": 2.866811752319336, | |
| "learning_rate": 0.00010059171597633136, | |
| "loss": 4.7059, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 2.982800245285034, | |
| "learning_rate": 0.00010029585798816568, | |
| "loss": 4.9884, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.5014792899408284, | |
| "grad_norm": 5.207017421722412, | |
| "learning_rate": 0.0001, | |
| "loss": 4.7667, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.5029585798816568, | |
| "grad_norm": 4.413080215454102, | |
| "learning_rate": 9.970414201183432e-05, | |
| "loss": 4.4587, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5044378698224852, | |
| "grad_norm": 3.2763004302978516, | |
| "learning_rate": 9.940828402366865e-05, | |
| "loss": 4.8882, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.5059171597633136, | |
| "grad_norm": 4.111000061035156, | |
| "learning_rate": 9.911242603550296e-05, | |
| "loss": 4.9992, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.507396449704142, | |
| "grad_norm": 3.3207952976226807, | |
| "learning_rate": 9.881656804733729e-05, | |
| "loss": 4.6362, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.5088757396449705, | |
| "grad_norm": 4.325817108154297, | |
| "learning_rate": 9.85207100591716e-05, | |
| "loss": 4.3929, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.5103550295857988, | |
| "grad_norm": 2.3893871307373047, | |
| "learning_rate": 9.822485207100593e-05, | |
| "loss": 4.7993, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.5118343195266272, | |
| "grad_norm": 2.4651639461517334, | |
| "learning_rate": 9.792899408284024e-05, | |
| "loss": 4.871, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.5133136094674556, | |
| "grad_norm": 3.43371319770813, | |
| "learning_rate": 9.763313609467456e-05, | |
| "loss": 4.831, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.514792899408284, | |
| "grad_norm": 4.255308151245117, | |
| "learning_rate": 9.733727810650888e-05, | |
| "loss": 4.6105, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.5162721893491125, | |
| "grad_norm": 2.4571831226348877, | |
| "learning_rate": 9.70414201183432e-05, | |
| "loss": 4.5852, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.5177514792899408, | |
| "grad_norm": 3.8601267337799072, | |
| "learning_rate": 9.674556213017751e-05, | |
| "loss": 5.0089, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5192307692307693, | |
| "grad_norm": 4.343974590301514, | |
| "learning_rate": 9.644970414201184e-05, | |
| "loss": 4.585, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.5207100591715976, | |
| "grad_norm": 3.3244740962982178, | |
| "learning_rate": 9.615384615384617e-05, | |
| "loss": 4.6961, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.522189349112426, | |
| "grad_norm": 3.011272668838501, | |
| "learning_rate": 9.585798816568048e-05, | |
| "loss": 4.7542, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.5236686390532544, | |
| "grad_norm": 4.02368688583374, | |
| "learning_rate": 9.55621301775148e-05, | |
| "loss": 4.605, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.5251479289940828, | |
| "grad_norm": 2.5176520347595215, | |
| "learning_rate": 9.526627218934911e-05, | |
| "loss": 5.057, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.5266272189349113, | |
| "grad_norm": 3.331259250640869, | |
| "learning_rate": 9.497041420118344e-05, | |
| "loss": 4.8883, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.5281065088757396, | |
| "grad_norm": 4.451882362365723, | |
| "learning_rate": 9.467455621301775e-05, | |
| "loss": 3.7867, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.5295857988165681, | |
| "grad_norm": 3.5634541511535645, | |
| "learning_rate": 9.437869822485208e-05, | |
| "loss": 4.6808, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.5310650887573964, | |
| "grad_norm": 4.832805156707764, | |
| "learning_rate": 9.408284023668639e-05, | |
| "loss": 4.8233, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.5325443786982249, | |
| "grad_norm": 3.726456642150879, | |
| "learning_rate": 9.378698224852072e-05, | |
| "loss": 5.0317, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5340236686390533, | |
| "grad_norm": 5.337479114532471, | |
| "learning_rate": 9.349112426035503e-05, | |
| "loss": 4.701, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.5355029585798816, | |
| "grad_norm": 3.692549228668213, | |
| "learning_rate": 9.319526627218935e-05, | |
| "loss": 4.8653, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.5369822485207101, | |
| "grad_norm": 3.613229990005493, | |
| "learning_rate": 9.289940828402367e-05, | |
| "loss": 4.8845, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 3.317173957824707, | |
| "learning_rate": 9.260355029585799e-05, | |
| "loss": 4.9675, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.5399408284023669, | |
| "grad_norm": 5.250802516937256, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 4.7967, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.5414201183431953, | |
| "grad_norm": 2.8899145126342773, | |
| "learning_rate": 9.201183431952663e-05, | |
| "loss": 4.7554, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.5428994082840237, | |
| "grad_norm": 2.3002591133117676, | |
| "learning_rate": 9.171597633136096e-05, | |
| "loss": 4.9671, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.5443786982248521, | |
| "grad_norm": 2.672611951828003, | |
| "learning_rate": 9.142011834319527e-05, | |
| "loss": 4.9408, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.5458579881656804, | |
| "grad_norm": 2.823984146118164, | |
| "learning_rate": 9.112426035502959e-05, | |
| "loss": 4.6108, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.5473372781065089, | |
| "grad_norm": 3.269995927810669, | |
| "learning_rate": 9.082840236686391e-05, | |
| "loss": 4.6865, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5488165680473372, | |
| "grad_norm": 2.842123031616211, | |
| "learning_rate": 9.053254437869823e-05, | |
| "loss": 5.0082, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.5502958579881657, | |
| "grad_norm": 3.5125341415405273, | |
| "learning_rate": 9.023668639053254e-05, | |
| "loss": 4.6579, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.5517751479289941, | |
| "grad_norm": 2.400810718536377, | |
| "learning_rate": 8.994082840236687e-05, | |
| "loss": 4.9984, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.5532544378698225, | |
| "grad_norm": 2.6020796298980713, | |
| "learning_rate": 8.96449704142012e-05, | |
| "loss": 4.6243, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.5547337278106509, | |
| "grad_norm": 2.897449493408203, | |
| "learning_rate": 8.934911242603551e-05, | |
| "loss": 4.9226, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.5562130177514792, | |
| "grad_norm": 3.834491491317749, | |
| "learning_rate": 8.905325443786982e-05, | |
| "loss": 4.8789, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.5576923076923077, | |
| "grad_norm": 2.6195783615112305, | |
| "learning_rate": 8.875739644970414e-05, | |
| "loss": 4.9757, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.5591715976331361, | |
| "grad_norm": 2.357797861099243, | |
| "learning_rate": 8.846153846153847e-05, | |
| "loss": 4.8308, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.5606508875739645, | |
| "grad_norm": 4.5268659591674805, | |
| "learning_rate": 8.816568047337278e-05, | |
| "loss": 5.017, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.5621301775147929, | |
| "grad_norm": 3.124450445175171, | |
| "learning_rate": 8.78698224852071e-05, | |
| "loss": 4.902, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5636094674556213, | |
| "grad_norm": 3.822741985321045, | |
| "learning_rate": 8.757396449704143e-05, | |
| "loss": 4.7571, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.5650887573964497, | |
| "grad_norm": 3.3395605087280273, | |
| "learning_rate": 8.727810650887575e-05, | |
| "loss": 4.3808, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.5665680473372781, | |
| "grad_norm": 3.0946218967437744, | |
| "learning_rate": 8.698224852071006e-05, | |
| "loss": 4.853, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.5680473372781065, | |
| "grad_norm": 2.6763172149658203, | |
| "learning_rate": 8.668639053254438e-05, | |
| "loss": 4.7149, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.5695266272189349, | |
| "grad_norm": 3.2569401264190674, | |
| "learning_rate": 8.63905325443787e-05, | |
| "loss": 4.7693, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.5710059171597633, | |
| "grad_norm": 2.8756442070007324, | |
| "learning_rate": 8.609467455621302e-05, | |
| "loss": 4.737, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.5724852071005917, | |
| "grad_norm": 3.0808393955230713, | |
| "learning_rate": 8.579881656804734e-05, | |
| "loss": 5.1146, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.5739644970414202, | |
| "grad_norm": 2.7643120288848877, | |
| "learning_rate": 8.550295857988166e-05, | |
| "loss": 4.9102, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.5754437869822485, | |
| "grad_norm": 4.045496940612793, | |
| "learning_rate": 8.520710059171599e-05, | |
| "loss": 4.4583, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.5769230769230769, | |
| "grad_norm": 2.765378952026367, | |
| "learning_rate": 8.49112426035503e-05, | |
| "loss": 4.6736, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5784023668639053, | |
| "grad_norm": 2.4997928142547607, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 4.9013, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.5798816568047337, | |
| "grad_norm": 6.506393909454346, | |
| "learning_rate": 8.431952662721894e-05, | |
| "loss": 4.4212, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.5813609467455622, | |
| "grad_norm": 3.8824899196624756, | |
| "learning_rate": 8.402366863905326e-05, | |
| "loss": 4.8729, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.5828402366863905, | |
| "grad_norm": 2.4025533199310303, | |
| "learning_rate": 8.372781065088757e-05, | |
| "loss": 4.6856, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.584319526627219, | |
| "grad_norm": 2.9260175228118896, | |
| "learning_rate": 8.34319526627219e-05, | |
| "loss": 4.6581, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.5857988165680473, | |
| "grad_norm": 2.606157064437866, | |
| "learning_rate": 8.313609467455622e-05, | |
| "loss": 4.7248, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.5872781065088757, | |
| "grad_norm": 3.9847888946533203, | |
| "learning_rate": 8.284023668639054e-05, | |
| "loss": 4.4983, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.5887573964497042, | |
| "grad_norm": 4.147398471832275, | |
| "learning_rate": 8.254437869822485e-05, | |
| "loss": 4.6558, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.5902366863905325, | |
| "grad_norm": 3.062237501144409, | |
| "learning_rate": 8.224852071005918e-05, | |
| "loss": 4.6806, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.591715976331361, | |
| "grad_norm": 2.655552625656128, | |
| "learning_rate": 8.19526627218935e-05, | |
| "loss": 4.7624, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5931952662721893, | |
| "grad_norm": 2.5713272094726562, | |
| "learning_rate": 8.165680473372781e-05, | |
| "loss": 4.4758, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.5946745562130178, | |
| "grad_norm": 4.519266605377197, | |
| "learning_rate": 8.136094674556214e-05, | |
| "loss": 4.1983, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.5961538461538461, | |
| "grad_norm": 2.5204124450683594, | |
| "learning_rate": 8.106508875739646e-05, | |
| "loss": 4.9982, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.5976331360946746, | |
| "grad_norm": 2.2742385864257812, | |
| "learning_rate": 8.076923076923078e-05, | |
| "loss": 4.8676, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.599112426035503, | |
| "grad_norm": 3.3117566108703613, | |
| "learning_rate": 8.047337278106509e-05, | |
| "loss": 4.5992, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.6005917159763313, | |
| "grad_norm": 2.826481819152832, | |
| "learning_rate": 8.01775147928994e-05, | |
| "loss": 4.8464, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.6020710059171598, | |
| "grad_norm": 2.4774935245513916, | |
| "learning_rate": 7.988165680473373e-05, | |
| "loss": 4.6491, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.6035502958579881, | |
| "grad_norm": 3.357351779937744, | |
| "learning_rate": 7.958579881656805e-05, | |
| "loss": 4.8258, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.6050295857988166, | |
| "grad_norm": 3.904404640197754, | |
| "learning_rate": 7.928994082840237e-05, | |
| "loss": 5.0641, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.606508875739645, | |
| "grad_norm": 3.978461742401123, | |
| "learning_rate": 7.899408284023669e-05, | |
| "loss": 4.8184, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6079881656804734, | |
| "grad_norm": 2.7608234882354736, | |
| "learning_rate": 7.869822485207101e-05, | |
| "loss": 5.0468, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.6094674556213018, | |
| "grad_norm": 4.422279357910156, | |
| "learning_rate": 7.840236686390533e-05, | |
| "loss": 4.6623, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.6109467455621301, | |
| "grad_norm": 2.7389438152313232, | |
| "learning_rate": 7.810650887573964e-05, | |
| "loss": 4.6685, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.6124260355029586, | |
| "grad_norm": 3.2453866004943848, | |
| "learning_rate": 7.781065088757397e-05, | |
| "loss": 4.796, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.613905325443787, | |
| "grad_norm": 4.576845645904541, | |
| "learning_rate": 7.751479289940828e-05, | |
| "loss": 5.1346, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 3.6460392475128174, | |
| "learning_rate": 7.72189349112426e-05, | |
| "loss": 4.3795, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.6168639053254438, | |
| "grad_norm": 3.7447457313537598, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 4.6371, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.6183431952662722, | |
| "grad_norm": 5.934640407562256, | |
| "learning_rate": 7.662721893491125e-05, | |
| "loss": 4.3938, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.6198224852071006, | |
| "grad_norm": 2.61598801612854, | |
| "learning_rate": 7.633136094674557e-05, | |
| "loss": 4.6006, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.621301775147929, | |
| "grad_norm": 2.915923833847046, | |
| "learning_rate": 7.603550295857988e-05, | |
| "loss": 4.5953, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6227810650887574, | |
| "grad_norm": 3.000753402709961, | |
| "learning_rate": 7.573964497041421e-05, | |
| "loss": 4.7139, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.6242603550295858, | |
| "grad_norm": 2.5539729595184326, | |
| "learning_rate": 7.544378698224852e-05, | |
| "loss": 5.232, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.6257396449704142, | |
| "grad_norm": 3.208082675933838, | |
| "learning_rate": 7.514792899408284e-05, | |
| "loss": 4.6739, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.6272189349112426, | |
| "grad_norm": 4.210474491119385, | |
| "learning_rate": 7.485207100591716e-05, | |
| "loss": 4.7495, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.628698224852071, | |
| "grad_norm": 3.1150636672973633, | |
| "learning_rate": 7.455621301775149e-05, | |
| "loss": 4.4906, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.6301775147928994, | |
| "grad_norm": 4.412407398223877, | |
| "learning_rate": 7.42603550295858e-05, | |
| "loss": 4.0822, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.6316568047337278, | |
| "grad_norm": 2.090398073196411, | |
| "learning_rate": 7.396449704142012e-05, | |
| "loss": 4.8458, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.6331360946745562, | |
| "grad_norm": 2.350902795791626, | |
| "learning_rate": 7.366863905325445e-05, | |
| "loss": 4.761, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.6346153846153846, | |
| "grad_norm": 3.2042136192321777, | |
| "learning_rate": 7.337278106508876e-05, | |
| "loss": 4.796, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.636094674556213, | |
| "grad_norm": 3.3418521881103516, | |
| "learning_rate": 7.307692307692307e-05, | |
| "loss": 4.7222, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6375739644970414, | |
| "grad_norm": 3.0160398483276367, | |
| "learning_rate": 7.27810650887574e-05, | |
| "loss": 4.4557, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.6390532544378699, | |
| "grad_norm": 2.5732924938201904, | |
| "learning_rate": 7.248520710059173e-05, | |
| "loss": 4.6779, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.6405325443786982, | |
| "grad_norm": 2.399313449859619, | |
| "learning_rate": 7.218934911242604e-05, | |
| "loss": 4.5695, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.6420118343195266, | |
| "grad_norm": 2.9960498809814453, | |
| "learning_rate": 7.189349112426036e-05, | |
| "loss": 4.8253, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.643491124260355, | |
| "grad_norm": 2.1950154304504395, | |
| "learning_rate": 7.159763313609467e-05, | |
| "loss": 4.492, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.6449704142011834, | |
| "grad_norm": 3.8510971069335938, | |
| "learning_rate": 7.1301775147929e-05, | |
| "loss": 4.6542, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.6464497041420119, | |
| "grad_norm": 3.028498649597168, | |
| "learning_rate": 7.100591715976331e-05, | |
| "loss": 4.5609, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.6479289940828402, | |
| "grad_norm": 3.7498788833618164, | |
| "learning_rate": 7.071005917159763e-05, | |
| "loss": 4.6621, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.6494082840236687, | |
| "grad_norm": 3.1128766536712646, | |
| "learning_rate": 7.041420118343195e-05, | |
| "loss": 5.0167, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.650887573964497, | |
| "grad_norm": 3.3049442768096924, | |
| "learning_rate": 7.011834319526628e-05, | |
| "loss": 4.5484, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6523668639053254, | |
| "grad_norm": 2.101262331008911, | |
| "learning_rate": 6.98224852071006e-05, | |
| "loss": 4.7794, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.6538461538461539, | |
| "grad_norm": 3.5438482761383057, | |
| "learning_rate": 6.952662721893491e-05, | |
| "loss": 4.9424, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.6553254437869822, | |
| "grad_norm": 4.34566593170166, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 5.0576, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.6568047337278107, | |
| "grad_norm": 2.580064296722412, | |
| "learning_rate": 6.893491124260355e-05, | |
| "loss": 4.7086, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.658284023668639, | |
| "grad_norm": 4.950741291046143, | |
| "learning_rate": 6.863905325443787e-05, | |
| "loss": 4.877, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.6597633136094675, | |
| "grad_norm": 4.225985527038574, | |
| "learning_rate": 6.834319526627219e-05, | |
| "loss": 4.7599, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.6612426035502958, | |
| "grad_norm": 2.690343141555786, | |
| "learning_rate": 6.804733727810652e-05, | |
| "loss": 4.604, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.6627218934911243, | |
| "grad_norm": 2.512183904647827, | |
| "learning_rate": 6.775147928994083e-05, | |
| "loss": 5.0561, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.6642011834319527, | |
| "grad_norm": 4.449777126312256, | |
| "learning_rate": 6.745562130177515e-05, | |
| "loss": 4.8173, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.665680473372781, | |
| "grad_norm": 3.311457633972168, | |
| "learning_rate": 6.715976331360948e-05, | |
| "loss": 4.3368, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6671597633136095, | |
| "grad_norm": 3.376208543777466, | |
| "learning_rate": 6.686390532544379e-05, | |
| "loss": 4.5102, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.6686390532544378, | |
| "grad_norm": 5.265395641326904, | |
| "learning_rate": 6.65680473372781e-05, | |
| "loss": 5.0219, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.6701183431952663, | |
| "grad_norm": 2.1007046699523926, | |
| "learning_rate": 6.627218934911243e-05, | |
| "loss": 5.1141, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.6715976331360947, | |
| "grad_norm": 4.058145046234131, | |
| "learning_rate": 6.597633136094676e-05, | |
| "loss": 4.422, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.6730769230769231, | |
| "grad_norm": 3.761136770248413, | |
| "learning_rate": 6.568047337278107e-05, | |
| "loss": 4.6039, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.6745562130177515, | |
| "grad_norm": 3.9088778495788574, | |
| "learning_rate": 6.538461538461539e-05, | |
| "loss": 4.5297, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.6760355029585798, | |
| "grad_norm": 3.4910857677459717, | |
| "learning_rate": 6.50887573964497e-05, | |
| "loss": 4.6043, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.6775147928994083, | |
| "grad_norm": 2.457960844039917, | |
| "learning_rate": 6.479289940828403e-05, | |
| "loss": 4.83, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.6789940828402367, | |
| "grad_norm": 3.0204050540924072, | |
| "learning_rate": 6.449704142011834e-05, | |
| "loss": 4.6305, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.6804733727810651, | |
| "grad_norm": 2.5136165618896484, | |
| "learning_rate": 6.420118343195266e-05, | |
| "loss": 4.79, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6819526627218935, | |
| "grad_norm": 3.7918813228607178, | |
| "learning_rate": 6.390532544378698e-05, | |
| "loss": 5.0849, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.6834319526627219, | |
| "grad_norm": 3.5558226108551025, | |
| "learning_rate": 6.360946745562131e-05, | |
| "loss": 4.6819, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.6849112426035503, | |
| "grad_norm": 3.061944007873535, | |
| "learning_rate": 6.331360946745562e-05, | |
| "loss": 4.7948, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.6863905325443787, | |
| "grad_norm": 2.5074996948242188, | |
| "learning_rate": 6.301775147928994e-05, | |
| "loss": 4.6494, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.6878698224852071, | |
| "grad_norm": 3.643146276473999, | |
| "learning_rate": 6.272189349112427e-05, | |
| "loss": 4.221, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.6893491124260355, | |
| "grad_norm": 3.808267831802368, | |
| "learning_rate": 6.242603550295858e-05, | |
| "loss": 5.0687, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.6908284023668639, | |
| "grad_norm": 2.975130558013916, | |
| "learning_rate": 6.21301775147929e-05, | |
| "loss": 4.7507, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 3.661033868789673, | |
| "learning_rate": 6.183431952662722e-05, | |
| "loss": 4.4896, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.6937869822485208, | |
| "grad_norm": 3.4433846473693848, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 5.1373, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.6952662721893491, | |
| "grad_norm": 2.975217342376709, | |
| "learning_rate": 6.124260355029586e-05, | |
| "loss": 4.5473, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.6967455621301775, | |
| "grad_norm": 3.5943000316619873, | |
| "learning_rate": 6.094674556213018e-05, | |
| "loss": 4.5376, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.6982248520710059, | |
| "grad_norm": 2.9393863677978516, | |
| "learning_rate": 6.06508875739645e-05, | |
| "loss": 4.6031, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.6997041420118343, | |
| "grad_norm": 4.864683151245117, | |
| "learning_rate": 6.035502958579882e-05, | |
| "loss": 4.7268, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.7011834319526628, | |
| "grad_norm": 2.9984517097473145, | |
| "learning_rate": 6.005917159763313e-05, | |
| "loss": 4.9315, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.7026627218934911, | |
| "grad_norm": 3.4088656902313232, | |
| "learning_rate": 5.9763313609467466e-05, | |
| "loss": 4.4509, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.7041420118343196, | |
| "grad_norm": 2.966920852661133, | |
| "learning_rate": 5.946745562130178e-05, | |
| "loss": 4.8591, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.7056213017751479, | |
| "grad_norm": 3.441845178604126, | |
| "learning_rate": 5.91715976331361e-05, | |
| "loss": 4.6297, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.7071005917159763, | |
| "grad_norm": 3.430345296859741, | |
| "learning_rate": 5.8875739644970415e-05, | |
| "loss": 4.7492, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.7085798816568047, | |
| "grad_norm": 3.0503721237182617, | |
| "learning_rate": 5.8579881656804736e-05, | |
| "loss": 4.4038, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.7100591715976331, | |
| "grad_norm": 3.72664213180542, | |
| "learning_rate": 5.8284023668639056e-05, | |
| "loss": 4.5192, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7115384615384616, | |
| "grad_norm": 2.554041862487793, | |
| "learning_rate": 5.798816568047337e-05, | |
| "loss": 4.6383, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.7130177514792899, | |
| "grad_norm": 2.4178683757781982, | |
| "learning_rate": 5.769230769230769e-05, | |
| "loss": 4.706, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.7144970414201184, | |
| "grad_norm": 2.1827499866485596, | |
| "learning_rate": 5.739644970414202e-05, | |
| "loss": 4.6598, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.7159763313609467, | |
| "grad_norm": 3.2616827487945557, | |
| "learning_rate": 5.710059171597634e-05, | |
| "loss": 4.1314, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.7174556213017751, | |
| "grad_norm": 4.249692440032959, | |
| "learning_rate": 5.680473372781065e-05, | |
| "loss": 4.311, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.7189349112426036, | |
| "grad_norm": 3.310980796813965, | |
| "learning_rate": 5.6508875739644974e-05, | |
| "loss": 4.5443, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.7204142011834319, | |
| "grad_norm": 3.357168436050415, | |
| "learning_rate": 5.621301775147929e-05, | |
| "loss": 4.3564, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.7218934911242604, | |
| "grad_norm": 4.302040100097656, | |
| "learning_rate": 5.591715976331361e-05, | |
| "loss": 4.4839, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.7233727810650887, | |
| "grad_norm": 3.120237112045288, | |
| "learning_rate": 5.562130177514793e-05, | |
| "loss": 4.335, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.7248520710059172, | |
| "grad_norm": 3.0357606410980225, | |
| "learning_rate": 5.532544378698226e-05, | |
| "loss": 4.6092, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7263313609467456, | |
| "grad_norm": 3.9619128704071045, | |
| "learning_rate": 5.502958579881658e-05, | |
| "loss": 4.5274, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.727810650887574, | |
| "grad_norm": 3.0399205684661865, | |
| "learning_rate": 5.473372781065089e-05, | |
| "loss": 4.3014, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.7292899408284024, | |
| "grad_norm": 2.8128726482391357, | |
| "learning_rate": 5.443786982248521e-05, | |
| "loss": 4.6141, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.7307692307692307, | |
| "grad_norm": 4.735112190246582, | |
| "learning_rate": 5.4142011834319526e-05, | |
| "loss": 4.8622, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.7322485207100592, | |
| "grad_norm": 5.150043487548828, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 4.9381, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.7337278106508875, | |
| "grad_norm": 3.8757338523864746, | |
| "learning_rate": 5.355029585798817e-05, | |
| "loss": 4.6672, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.735207100591716, | |
| "grad_norm": 3.1939332485198975, | |
| "learning_rate": 5.3254437869822495e-05, | |
| "loss": 4.2007, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.7366863905325444, | |
| "grad_norm": 2.566030979156494, | |
| "learning_rate": 5.295857988165681e-05, | |
| "loss": 4.8573, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.7381656804733728, | |
| "grad_norm": 4.250690460205078, | |
| "learning_rate": 5.266272189349113e-05, | |
| "loss": 5.2471, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.7396449704142012, | |
| "grad_norm": 3.5470082759857178, | |
| "learning_rate": 5.236686390532545e-05, | |
| "loss": 5.0067, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7411242603550295, | |
| "grad_norm": 3.6443684101104736, | |
| "learning_rate": 5.2071005917159764e-05, | |
| "loss": 4.5634, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.742603550295858, | |
| "grad_norm": 3.061985492706299, | |
| "learning_rate": 5.1775147928994085e-05, | |
| "loss": 4.5597, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.7440828402366864, | |
| "grad_norm": 4.4746623039245605, | |
| "learning_rate": 5.14792899408284e-05, | |
| "loss": 4.298, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.7455621301775148, | |
| "grad_norm": 3.3475050926208496, | |
| "learning_rate": 5.118343195266272e-05, | |
| "loss": 4.733, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.7470414201183432, | |
| "grad_norm": 2.60097074508667, | |
| "learning_rate": 5.088757396449705e-05, | |
| "loss": 4.5727, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.7485207100591716, | |
| "grad_norm": 3.4892466068267822, | |
| "learning_rate": 5.059171597633137e-05, | |
| "loss": 4.8583, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 3.30724835395813, | |
| "learning_rate": 5.029585798816568e-05, | |
| "loss": 4.3811, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.7514792899408284, | |
| "grad_norm": 2.1683871746063232, | |
| "learning_rate": 5e-05, | |
| "loss": 4.5991, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.7529585798816568, | |
| "grad_norm": 2.1917285919189453, | |
| "learning_rate": 4.970414201183432e-05, | |
| "loss": 4.8409, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.7544378698224852, | |
| "grad_norm": 3.028266668319702, | |
| "learning_rate": 4.9408284023668644e-05, | |
| "loss": 4.5322, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.7559171597633136, | |
| "grad_norm": 3.031081438064575, | |
| "learning_rate": 4.9112426035502965e-05, | |
| "loss": 4.7952, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.757396449704142, | |
| "grad_norm": 3.272369861602783, | |
| "learning_rate": 4.881656804733728e-05, | |
| "loss": 4.3832, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.7588757396449705, | |
| "grad_norm": 4.823290824890137, | |
| "learning_rate": 4.85207100591716e-05, | |
| "loss": 4.6106, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.7603550295857988, | |
| "grad_norm": 2.7338478565216064, | |
| "learning_rate": 4.822485207100592e-05, | |
| "loss": 4.5112, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.7618343195266272, | |
| "grad_norm": 2.4823102951049805, | |
| "learning_rate": 4.792899408284024e-05, | |
| "loss": 4.6146, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.7633136094674556, | |
| "grad_norm": 3.007741689682007, | |
| "learning_rate": 4.7633136094674555e-05, | |
| "loss": 4.6655, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.764792899408284, | |
| "grad_norm": 4.28762674331665, | |
| "learning_rate": 4.7337278106508875e-05, | |
| "loss": 4.3793, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.7662721893491125, | |
| "grad_norm": 3.1191656589508057, | |
| "learning_rate": 4.7041420118343196e-05, | |
| "loss": 4.4016, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.7677514792899408, | |
| "grad_norm": 3.467637062072754, | |
| "learning_rate": 4.674556213017752e-05, | |
| "loss": 4.7173, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 3.6230361461639404, | |
| "learning_rate": 4.644970414201184e-05, | |
| "loss": 4.5995, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.7707100591715976, | |
| "grad_norm": 3.44942045211792, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 4.6219, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.772189349112426, | |
| "grad_norm": 3.29604172706604, | |
| "learning_rate": 4.585798816568048e-05, | |
| "loss": 4.094, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.7736686390532544, | |
| "grad_norm": 2.802762031555176, | |
| "learning_rate": 4.556213017751479e-05, | |
| "loss": 4.7087, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.7751479289940828, | |
| "grad_norm": 4.227675437927246, | |
| "learning_rate": 4.5266272189349114e-05, | |
| "loss": 4.3078, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.7766272189349113, | |
| "grad_norm": 3.0525543689727783, | |
| "learning_rate": 4.4970414201183434e-05, | |
| "loss": 4.7471, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.7781065088757396, | |
| "grad_norm": 3.044851064682007, | |
| "learning_rate": 4.4674556213017755e-05, | |
| "loss": 4.9371, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.7795857988165681, | |
| "grad_norm": 3.847062826156616, | |
| "learning_rate": 4.437869822485207e-05, | |
| "loss": 4.3949, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.7810650887573964, | |
| "grad_norm": 2.669889211654663, | |
| "learning_rate": 4.408284023668639e-05, | |
| "loss": 4.5557, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.7825443786982249, | |
| "grad_norm": 2.970334768295288, | |
| "learning_rate": 4.378698224852072e-05, | |
| "loss": 4.5767, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.7840236686390533, | |
| "grad_norm": 2.5080573558807373, | |
| "learning_rate": 4.349112426035503e-05, | |
| "loss": 4.5653, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.7855029585798816, | |
| "grad_norm": 2.9842259883880615, | |
| "learning_rate": 4.319526627218935e-05, | |
| "loss": 4.7996, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.7869822485207101, | |
| "grad_norm": 3.3907594680786133, | |
| "learning_rate": 4.289940828402367e-05, | |
| "loss": 4.9064, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.7884615384615384, | |
| "grad_norm": 4.384812355041504, | |
| "learning_rate": 4.260355029585799e-05, | |
| "loss": 4.2071, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.7899408284023669, | |
| "grad_norm": 2.564908266067505, | |
| "learning_rate": 4.230769230769231e-05, | |
| "loss": 4.5635, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.7914201183431953, | |
| "grad_norm": 4.1615753173828125, | |
| "learning_rate": 4.201183431952663e-05, | |
| "loss": 4.3675, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.7928994082840237, | |
| "grad_norm": 3.0495238304138184, | |
| "learning_rate": 4.171597633136095e-05, | |
| "loss": 4.3898, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.7943786982248521, | |
| "grad_norm": 4.264047145843506, | |
| "learning_rate": 4.142011834319527e-05, | |
| "loss": 4.6589, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.7958579881656804, | |
| "grad_norm": 2.76383113861084, | |
| "learning_rate": 4.112426035502959e-05, | |
| "loss": 4.6825, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.7973372781065089, | |
| "grad_norm": 4.031206130981445, | |
| "learning_rate": 4.0828402366863904e-05, | |
| "loss": 5.0627, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.7988165680473372, | |
| "grad_norm": 4.084919452667236, | |
| "learning_rate": 4.053254437869823e-05, | |
| "loss": 4.1003, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8002958579881657, | |
| "grad_norm": 2.7242236137390137, | |
| "learning_rate": 4.0236686390532545e-05, | |
| "loss": 5.074, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.8017751479289941, | |
| "grad_norm": 3.6945011615753174, | |
| "learning_rate": 3.9940828402366866e-05, | |
| "loss": 4.3444, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.8032544378698225, | |
| "grad_norm": 2.436042308807373, | |
| "learning_rate": 3.964497041420119e-05, | |
| "loss": 4.4678, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.8047337278106509, | |
| "grad_norm": 3.4951744079589844, | |
| "learning_rate": 3.934911242603551e-05, | |
| "loss": 4.4405, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.8062130177514792, | |
| "grad_norm": 3.391136646270752, | |
| "learning_rate": 3.905325443786982e-05, | |
| "loss": 4.7685, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.8076923076923077, | |
| "grad_norm": 3.002776861190796, | |
| "learning_rate": 3.875739644970414e-05, | |
| "loss": 4.754, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.8091715976331361, | |
| "grad_norm": 2.6767425537109375, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 4.8214, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.8106508875739645, | |
| "grad_norm": 4.293323516845703, | |
| "learning_rate": 3.8165680473372784e-05, | |
| "loss": 4.7054, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.8121301775147929, | |
| "grad_norm": 3.0231809616088867, | |
| "learning_rate": 3.7869822485207104e-05, | |
| "loss": 4.3928, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.8136094674556213, | |
| "grad_norm": 2.637932300567627, | |
| "learning_rate": 3.757396449704142e-05, | |
| "loss": 4.6446, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8150887573964497, | |
| "grad_norm": 6.3277974128723145, | |
| "learning_rate": 3.7278106508875746e-05, | |
| "loss": 4.3953, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.8165680473372781, | |
| "grad_norm": 2.680094003677368, | |
| "learning_rate": 3.698224852071006e-05, | |
| "loss": 4.3732, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.8180473372781065, | |
| "grad_norm": 1.9856054782867432, | |
| "learning_rate": 3.668639053254438e-05, | |
| "loss": 4.6402, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.8195266272189349, | |
| "grad_norm": 2.212771415710449, | |
| "learning_rate": 3.63905325443787e-05, | |
| "loss": 4.7445, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.8210059171597633, | |
| "grad_norm": 2.5814425945281982, | |
| "learning_rate": 3.609467455621302e-05, | |
| "loss": 4.5684, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.8224852071005917, | |
| "grad_norm": 3.313979387283325, | |
| "learning_rate": 3.5798816568047336e-05, | |
| "loss": 4.6471, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.8239644970414202, | |
| "grad_norm": 2.2291247844696045, | |
| "learning_rate": 3.5502958579881656e-05, | |
| "loss": 4.77, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.8254437869822485, | |
| "grad_norm": 3.7490108013153076, | |
| "learning_rate": 3.520710059171598e-05, | |
| "loss": 4.8479, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.8269230769230769, | |
| "grad_norm": 2.6882386207580566, | |
| "learning_rate": 3.49112426035503e-05, | |
| "loss": 5.0244, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.8284023668639053, | |
| "grad_norm": 3.1564576625823975, | |
| "learning_rate": 3.461538461538462e-05, | |
| "loss": 4.6782, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.8298816568047337, | |
| "grad_norm": 2.728583335876465, | |
| "learning_rate": 3.431952662721893e-05, | |
| "loss": 4.8741, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.8313609467455622, | |
| "grad_norm": 3.116046190261841, | |
| "learning_rate": 3.402366863905326e-05, | |
| "loss": 4.6757, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.8328402366863905, | |
| "grad_norm": 3.0283756256103516, | |
| "learning_rate": 3.3727810650887574e-05, | |
| "loss": 4.7774, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.834319526627219, | |
| "grad_norm": 4.741537094116211, | |
| "learning_rate": 3.3431952662721895e-05, | |
| "loss": 5.1097, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.8357988165680473, | |
| "grad_norm": 3.0617265701293945, | |
| "learning_rate": 3.3136094674556215e-05, | |
| "loss": 4.6326, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.8372781065088757, | |
| "grad_norm": 3.294005870819092, | |
| "learning_rate": 3.2840236686390536e-05, | |
| "loss": 4.4754, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.8387573964497042, | |
| "grad_norm": 3.0990583896636963, | |
| "learning_rate": 3.254437869822485e-05, | |
| "loss": 5.0362, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.8402366863905325, | |
| "grad_norm": 3.523688316345215, | |
| "learning_rate": 3.224852071005917e-05, | |
| "loss": 5.065, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.841715976331361, | |
| "grad_norm": 2.907668113708496, | |
| "learning_rate": 3.195266272189349e-05, | |
| "loss": 4.6707, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.8431952662721893, | |
| "grad_norm": 2.66795015335083, | |
| "learning_rate": 3.165680473372781e-05, | |
| "loss": 4.7274, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.8446745562130178, | |
| "grad_norm": 3.600651502609253, | |
| "learning_rate": 3.136094674556213e-05, | |
| "loss": 4.7253, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 2.160475730895996, | |
| "learning_rate": 3.106508875739645e-05, | |
| "loss": 4.7706, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.8476331360946746, | |
| "grad_norm": 2.354764223098755, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 4.4115, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.849112426035503, | |
| "grad_norm": 3.3605852127075195, | |
| "learning_rate": 3.047337278106509e-05, | |
| "loss": 4.5752, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.8505917159763313, | |
| "grad_norm": 2.4779610633850098, | |
| "learning_rate": 3.017751479289941e-05, | |
| "loss": 4.4164, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.8520710059171598, | |
| "grad_norm": 4.133167266845703, | |
| "learning_rate": 2.9881656804733733e-05, | |
| "loss": 4.7699, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.8535502958579881, | |
| "grad_norm": 3.904437303543091, | |
| "learning_rate": 2.958579881656805e-05, | |
| "loss": 4.8485, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.8550295857988166, | |
| "grad_norm": 2.1680612564086914, | |
| "learning_rate": 2.9289940828402368e-05, | |
| "loss": 4.5462, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.856508875739645, | |
| "grad_norm": 2.6322898864746094, | |
| "learning_rate": 2.8994082840236685e-05, | |
| "loss": 4.5086, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.8579881656804734, | |
| "grad_norm": 3.004564046859741, | |
| "learning_rate": 2.869822485207101e-05, | |
| "loss": 4.2941, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.8594674556213018, | |
| "grad_norm": 3.2100727558135986, | |
| "learning_rate": 2.8402366863905327e-05, | |
| "loss": 4.918, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.8609467455621301, | |
| "grad_norm": 3.243868112564087, | |
| "learning_rate": 2.8106508875739644e-05, | |
| "loss": 4.8079, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.8624260355029586, | |
| "grad_norm": 1.9000004529953003, | |
| "learning_rate": 2.7810650887573965e-05, | |
| "loss": 4.6606, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.863905325443787, | |
| "grad_norm": 4.0184502601623535, | |
| "learning_rate": 2.751479289940829e-05, | |
| "loss": 4.206, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.8653846153846154, | |
| "grad_norm": 3.3465614318847656, | |
| "learning_rate": 2.7218934911242606e-05, | |
| "loss": 4.6692, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.8668639053254438, | |
| "grad_norm": 3.0921947956085205, | |
| "learning_rate": 2.6923076923076923e-05, | |
| "loss": 4.3901, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.8683431952662722, | |
| "grad_norm": 4.194738864898682, | |
| "learning_rate": 2.6627218934911247e-05, | |
| "loss": 4.6715, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.8698224852071006, | |
| "grad_norm": 4.712646484375, | |
| "learning_rate": 2.6331360946745565e-05, | |
| "loss": 4.2221, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.871301775147929, | |
| "grad_norm": 4.114252090454102, | |
| "learning_rate": 2.6035502958579882e-05, | |
| "loss": 4.3563, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.8727810650887574, | |
| "grad_norm": 3.0066916942596436, | |
| "learning_rate": 2.57396449704142e-05, | |
| "loss": 4.4031, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.8742603550295858, | |
| "grad_norm": 4.690349102020264, | |
| "learning_rate": 2.5443786982248524e-05, | |
| "loss": 4.2049, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.8757396449704142, | |
| "grad_norm": 2.729142904281616, | |
| "learning_rate": 2.514792899408284e-05, | |
| "loss": 4.3712, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.8772189349112426, | |
| "grad_norm": 3.8181116580963135, | |
| "learning_rate": 2.485207100591716e-05, | |
| "loss": 4.8115, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.878698224852071, | |
| "grad_norm": 4.084876537322998, | |
| "learning_rate": 2.4556213017751482e-05, | |
| "loss": 4.778, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.8801775147928994, | |
| "grad_norm": 2.857482671737671, | |
| "learning_rate": 2.42603550295858e-05, | |
| "loss": 4.3799, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.8816568047337278, | |
| "grad_norm": 4.242053985595703, | |
| "learning_rate": 2.396449704142012e-05, | |
| "loss": 4.3937, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.8831360946745562, | |
| "grad_norm": 2.85383677482605, | |
| "learning_rate": 2.3668639053254438e-05, | |
| "loss": 4.1794, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.8846153846153846, | |
| "grad_norm": 2.333695411682129, | |
| "learning_rate": 2.337278106508876e-05, | |
| "loss": 4.6517, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.886094674556213, | |
| "grad_norm": 2.6515889167785645, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 4.4109, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.8875739644970414, | |
| "grad_norm": 3.1488544940948486, | |
| "learning_rate": 2.2781065088757396e-05, | |
| "loss": 4.4439, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.8890532544378699, | |
| "grad_norm": 2.4748241901397705, | |
| "learning_rate": 2.2485207100591717e-05, | |
| "loss": 4.4223, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.8905325443786982, | |
| "grad_norm": 2.4698967933654785, | |
| "learning_rate": 2.2189349112426034e-05, | |
| "loss": 4.5374, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.8920118343195266, | |
| "grad_norm": 4.023700714111328, | |
| "learning_rate": 2.189349112426036e-05, | |
| "loss": 4.6281, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.893491124260355, | |
| "grad_norm": 2.4827523231506348, | |
| "learning_rate": 2.1597633136094676e-05, | |
| "loss": 4.6307, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.8949704142011834, | |
| "grad_norm": 3.7092514038085938, | |
| "learning_rate": 2.1301775147928997e-05, | |
| "loss": 4.9738, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.8964497041420119, | |
| "grad_norm": 3.7633941173553467, | |
| "learning_rate": 2.1005917159763314e-05, | |
| "loss": 4.3193, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.8979289940828402, | |
| "grad_norm": 3.2077150344848633, | |
| "learning_rate": 2.0710059171597635e-05, | |
| "loss": 4.2583, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.8994082840236687, | |
| "grad_norm": 4.071720600128174, | |
| "learning_rate": 2.0414201183431952e-05, | |
| "loss": 4.3272, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.900887573964497, | |
| "grad_norm": 3.028972864151001, | |
| "learning_rate": 2.0118343195266273e-05, | |
| "loss": 4.9391, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.9023668639053254, | |
| "grad_norm": 2.4915459156036377, | |
| "learning_rate": 1.9822485207100593e-05, | |
| "loss": 4.5273, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.9038461538461539, | |
| "grad_norm": 3.3177757263183594, | |
| "learning_rate": 1.952662721893491e-05, | |
| "loss": 4.6075, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.9053254437869822, | |
| "grad_norm": 2.9359259605407715, | |
| "learning_rate": 1.923076923076923e-05, | |
| "loss": 4.4901, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.9068047337278107, | |
| "grad_norm": 3.027765989303589, | |
| "learning_rate": 1.8934911242603552e-05, | |
| "loss": 4.5698, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.908284023668639, | |
| "grad_norm": 3.5717012882232666, | |
| "learning_rate": 1.8639053254437873e-05, | |
| "loss": 4.1682, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.9097633136094675, | |
| "grad_norm": 3.882838487625122, | |
| "learning_rate": 1.834319526627219e-05, | |
| "loss": 4.551, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.9112426035502958, | |
| "grad_norm": 4.922489166259766, | |
| "learning_rate": 1.804733727810651e-05, | |
| "loss": 4.3742, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.9127218934911243, | |
| "grad_norm": 2.828356981277466, | |
| "learning_rate": 1.7751479289940828e-05, | |
| "loss": 4.8154, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.9142011834319527, | |
| "grad_norm": 2.8683178424835205, | |
| "learning_rate": 1.745562130177515e-05, | |
| "loss": 4.512, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.915680473372781, | |
| "grad_norm": 2.6329286098480225, | |
| "learning_rate": 1.7159763313609466e-05, | |
| "loss": 4.414, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.9171597633136095, | |
| "grad_norm": 4.237077236175537, | |
| "learning_rate": 1.6863905325443787e-05, | |
| "loss": 4.7751, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.9186390532544378, | |
| "grad_norm": 2.7668392658233643, | |
| "learning_rate": 1.6568047337278108e-05, | |
| "loss": 4.4033, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.9201183431952663, | |
| "grad_norm": 3.169762134552002, | |
| "learning_rate": 1.6272189349112425e-05, | |
| "loss": 4.5085, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.9215976331360947, | |
| "grad_norm": 4.470832824707031, | |
| "learning_rate": 1.5976331360946746e-05, | |
| "loss": 4.3483, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 2.081108808517456, | |
| "learning_rate": 1.5680473372781066e-05, | |
| "loss": 4.7229, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.9245562130177515, | |
| "grad_norm": 2.9792003631591797, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 4.4489, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.9260355029585798, | |
| "grad_norm": 3.471381902694702, | |
| "learning_rate": 1.5088757396449705e-05, | |
| "loss": 4.3747, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.9275147928994083, | |
| "grad_norm": 2.503862142562866, | |
| "learning_rate": 1.4792899408284025e-05, | |
| "loss": 4.3829, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.9289940828402367, | |
| "grad_norm": 3.348893165588379, | |
| "learning_rate": 1.4497041420118343e-05, | |
| "loss": 4.5536, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.9304733727810651, | |
| "grad_norm": 2.7751362323760986, | |
| "learning_rate": 1.4201183431952663e-05, | |
| "loss": 4.8366, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.9319526627218935, | |
| "grad_norm": 3.739258289337158, | |
| "learning_rate": 1.3905325443786982e-05, | |
| "loss": 5.409, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9334319526627219, | |
| "grad_norm": 2.438932180404663, | |
| "learning_rate": 1.3609467455621303e-05, | |
| "loss": 5.0738, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.9349112426035503, | |
| "grad_norm": 3.888420820236206, | |
| "learning_rate": 1.3313609467455624e-05, | |
| "loss": 4.4, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.9363905325443787, | |
| "grad_norm": 3.376814842224121, | |
| "learning_rate": 1.3017751479289941e-05, | |
| "loss": 4.7996, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.9378698224852071, | |
| "grad_norm": 2.792365550994873, | |
| "learning_rate": 1.2721893491124262e-05, | |
| "loss": 4.5187, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.9393491124260355, | |
| "grad_norm": 2.4097423553466797, | |
| "learning_rate": 1.242603550295858e-05, | |
| "loss": 4.6622, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.9408284023668639, | |
| "grad_norm": 2.6368215084075928, | |
| "learning_rate": 1.21301775147929e-05, | |
| "loss": 4.6149, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.9423076923076923, | |
| "grad_norm": 3.4069018363952637, | |
| "learning_rate": 1.1834319526627219e-05, | |
| "loss": 4.1836, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.9437869822485208, | |
| "grad_norm": 3.1973531246185303, | |
| "learning_rate": 1.153846153846154e-05, | |
| "loss": 4.5128, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.9452662721893491, | |
| "grad_norm": 3.2625622749328613, | |
| "learning_rate": 1.1242603550295859e-05, | |
| "loss": 4.4204, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.9467455621301775, | |
| "grad_norm": 4.215187072753906, | |
| "learning_rate": 1.094674556213018e-05, | |
| "loss": 4.8363, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.9482248520710059, | |
| "grad_norm": 3.0874249935150146, | |
| "learning_rate": 1.0650887573964498e-05, | |
| "loss": 4.5997, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.9497041420118343, | |
| "grad_norm": 3.4833452701568604, | |
| "learning_rate": 1.0355029585798817e-05, | |
| "loss": 4.6514, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.9511834319526628, | |
| "grad_norm": 2.5931572914123535, | |
| "learning_rate": 1.0059171597633136e-05, | |
| "loss": 4.5665, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.9526627218934911, | |
| "grad_norm": 4.655201435089111, | |
| "learning_rate": 9.763313609467455e-06, | |
| "loss": 4.442, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.9541420118343196, | |
| "grad_norm": 4.815556526184082, | |
| "learning_rate": 9.467455621301776e-06, | |
| "loss": 4.4401, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.9556213017751479, | |
| "grad_norm": 2.570767641067505, | |
| "learning_rate": 9.171597633136095e-06, | |
| "loss": 4.4435, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.9571005917159763, | |
| "grad_norm": 3.6513237953186035, | |
| "learning_rate": 8.875739644970414e-06, | |
| "loss": 4.1544, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.9585798816568047, | |
| "grad_norm": 2.992757797241211, | |
| "learning_rate": 8.579881656804733e-06, | |
| "loss": 4.5563, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.9600591715976331, | |
| "grad_norm": 3.3865644931793213, | |
| "learning_rate": 8.284023668639054e-06, | |
| "loss": 4.5594, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.9615384615384616, | |
| "grad_norm": 3.6119096279144287, | |
| "learning_rate": 7.988165680473373e-06, | |
| "loss": 4.6004, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.9630177514792899, | |
| "grad_norm": 2.800621509552002, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 4.6929, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.9644970414201184, | |
| "grad_norm": 3.412346601486206, | |
| "learning_rate": 7.396449704142013e-06, | |
| "loss": 4.5545, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.9659763313609467, | |
| "grad_norm": 3.1978702545166016, | |
| "learning_rate": 7.100591715976332e-06, | |
| "loss": 4.8381, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.9674556213017751, | |
| "grad_norm": 2.437225341796875, | |
| "learning_rate": 6.8047337278106515e-06, | |
| "loss": 4.8636, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.9689349112426036, | |
| "grad_norm": 2.7770652770996094, | |
| "learning_rate": 6.5088757396449705e-06, | |
| "loss": 4.4207, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.9704142011834319, | |
| "grad_norm": 4.442773342132568, | |
| "learning_rate": 6.21301775147929e-06, | |
| "loss": 4.2258, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.9718934911242604, | |
| "grad_norm": 3.711191177368164, | |
| "learning_rate": 5.917159763313609e-06, | |
| "loss": 4.4003, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.9733727810650887, | |
| "grad_norm": 3.1590025424957275, | |
| "learning_rate": 5.621301775147929e-06, | |
| "loss": 4.499, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.9748520710059172, | |
| "grad_norm": 2.811072587966919, | |
| "learning_rate": 5.325443786982249e-06, | |
| "loss": 4.9061, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.9763313609467456, | |
| "grad_norm": 3.6592109203338623, | |
| "learning_rate": 5.029585798816568e-06, | |
| "loss": 4.1664, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.977810650887574, | |
| "grad_norm": 2.923866033554077, | |
| "learning_rate": 4.733727810650888e-06, | |
| "loss": 4.5202, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.9792899408284024, | |
| "grad_norm": 2.50335431098938, | |
| "learning_rate": 4.437869822485207e-06, | |
| "loss": 4.5732, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.9807692307692307, | |
| "grad_norm": 2.5192694664001465, | |
| "learning_rate": 4.142011834319527e-06, | |
| "loss": 5.3358, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.9822485207100592, | |
| "grad_norm": 2.9927144050598145, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 4.7768, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.9837278106508875, | |
| "grad_norm": 2.4072325229644775, | |
| "learning_rate": 3.550295857988166e-06, | |
| "loss": 4.7046, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.985207100591716, | |
| "grad_norm": 3.98689866065979, | |
| "learning_rate": 3.2544378698224853e-06, | |
| "loss": 4.1346, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.9866863905325444, | |
| "grad_norm": 4.064346790313721, | |
| "learning_rate": 2.9585798816568047e-06, | |
| "loss": 4.5276, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.9881656804733728, | |
| "grad_norm": 3.102790117263794, | |
| "learning_rate": 2.6627218934911246e-06, | |
| "loss": 4.5344, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.9896449704142012, | |
| "grad_norm": 3.3671998977661133, | |
| "learning_rate": 2.366863905325444e-06, | |
| "loss": 4.3674, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.9911242603550295, | |
| "grad_norm": 4.128759384155273, | |
| "learning_rate": 2.0710059171597635e-06, | |
| "loss": 4.7371, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.992603550295858, | |
| "grad_norm": 2.914989709854126, | |
| "learning_rate": 1.775147928994083e-06, | |
| "loss": 4.5296, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.9940828402366864, | |
| "grad_norm": 2.991652727127075, | |
| "learning_rate": 1.4792899408284024e-06, | |
| "loss": 4.6279, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.9955621301775148, | |
| "grad_norm": 3.1215224266052246, | |
| "learning_rate": 1.183431952662722e-06, | |
| "loss": 4.3423, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.9970414201183432, | |
| "grad_norm": 4.466518878936768, | |
| "learning_rate": 8.875739644970415e-07, | |
| "loss": 4.348, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.9985207100591716, | |
| "grad_norm": 3.410566568374634, | |
| "learning_rate": 5.91715976331361e-07, | |
| "loss": 4.4988, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.770328044891357, | |
| "learning_rate": 2.958579881656805e-07, | |
| "loss": 4.2684, | |
| "step": 676 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 676, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 5000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 719100446989824.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |