| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9942299156679982, | |
| "eval_steps": 500, | |
| "global_step": 140, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.5339343547821045, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4406, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.4306606948375702, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4472, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.48949554562568665, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4724, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.7450588941574097, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4314, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.39597198367118835, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4576, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.4372778832912445, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4573, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.5514771938323975, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4342, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.6987231969833374, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4239, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.5473514795303345, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4599, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.5179272890090942, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4462, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.5821484923362732, | |
| "learning_rate": 1e-05, | |
| "loss": 0.435, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.5863294005393982, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4272, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.4016607403755188, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4491, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.4974324703216553, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4542, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.5260458588600159, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4387, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.7002240419387817, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4274, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.3600125014781952, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4564, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.4479730725288391, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4284, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.539320170879364, | |
| "learning_rate": 1e-05, | |
| "loss": 0.44, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.8680666089057922, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4301, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.41423720121383667, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4466, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.441376268863678, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4447, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.5229262709617615, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4344, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.7713359594345093, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4375, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.4245271384716034, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4618, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.4831457734107971, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4403, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.58194899559021, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4229, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.7558773159980774, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3974, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.39629215002059937, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4616, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.49573493003845215, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4417, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.5668038129806519, | |
| "learning_rate": 1e-05, | |
| "loss": 0.43, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.7516561150550842, | |
| "learning_rate": 1e-05, | |
| "loss": 0.414, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 1.1010799407958984, | |
| "learning_rate": 1e-05, | |
| "loss": 892.876, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.4981920123100281, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4428, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.6654168963432312, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4234, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.9636304378509521, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4185, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.39848971366882324, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4517, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.5225277543067932, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4459, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.5011440515518188, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4403, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.9465950727462769, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4103, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.43049702048301697, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4553, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.48856088519096375, | |
| "learning_rate": 1e-05, | |
| "loss": 0.449, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.6367886066436768, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4405, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.6144514679908752, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4364, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.4697144031524658, | |
| "learning_rate": 1e-05, | |
| "loss": 0.451, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.5866898894309998, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4387, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.5469757914543152, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4464, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.5702504515647888, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4237, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.4027068614959717, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4608, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.4725416302680969, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4564, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.5751639008522034, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4406, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 1.0833802223205566, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4157, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.44081103801727295, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4618, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.46274518966674805, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4391, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.48259931802749634, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4456, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.7039265632629395, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4201, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.4099649488925934, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4373, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.47295379638671875, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4475, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.5771341919898987, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4205, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.7415468692779541, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4287, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.5697640776634216, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4351, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.494108110666275, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4356, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.7063843011856079, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4101, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.9190882444381714, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4008, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.6143580675125122, | |
| "learning_rate": 1e-05, | |
| "loss": 1077.2749, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.6259483098983765, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4455, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.6354689002037048, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4249, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.8926694393157959, | |
| "learning_rate": 1e-05, | |
| "loss": 0.428, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.37563061714172363, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4522, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.4575391113758087, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4601, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.5881872177124023, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4284, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 1.1317410469055176, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4023, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.5603578090667725, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4573, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.6214807629585266, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4535, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.6941476464271545, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4351, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.8727200031280518, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4248, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.6732370853424072, | |
| "learning_rate": 1e-05, | |
| "loss": 0.45, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.7978621125221252, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4417, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.7042059898376465, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3983, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 1.2186613082885742, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3991, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.5385881066322327, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4516, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.6898678541183472, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4246, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.8373177647590637, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4199, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.8186657428741455, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4425, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.8698004484176636, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4432, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 1.0647398233413696, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4308, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.8065099120140076, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4369, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 1.0753732919692993, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4055, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.4561431407928467, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4482, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.5589205026626587, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4517, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.7920203804969788, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4107, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 1.1868764162063599, | |
| "learning_rate": 1e-05, | |
| "loss": 0.406, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.5007905960083008, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4582, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.588740348815918, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4285, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.7123900651931763, | |
| "learning_rate": 1e-05, | |
| "loss": 0.445, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.7591605186462402, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4111, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.6483467817306519, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4353, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.6339621543884277, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4167, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.9539307951927185, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4129, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.9566183090209961, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4122, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.45071572065353394, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4367, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.6243826746940613, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4517, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 1.0642685890197754, | |
| "learning_rate": 1e-05, | |
| "loss": 0.418, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.8378849029541016, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3948, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.5254150629043579, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4566, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.677213728427887, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4484, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 1.118269920349121, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4523, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 1.260345458984375, | |
| "learning_rate": 1e-05, | |
| "loss": 0.408, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.7494204044342041, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4275, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.6795545220375061, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4505, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.8415333032608032, | |
| "learning_rate": 1e-05, | |
| "loss": 0.442, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.064477562904358, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4019, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.6557851433753967, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4534, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.7865703701972961, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4339, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.7419100403785706, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4364, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 1.0382827520370483, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4028, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.9644023776054382, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4383, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.859690248966217, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4327, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.8699947595596313, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4275, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 1.2335870265960693, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3891, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.506551206111908, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4568, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 2.2713043689727783, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4419, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.6728506684303284, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4447, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.8511173725128174, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3946, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.5495295524597168, | |
| "learning_rate": 1e-05, | |
| "loss": 1092.1152, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.9345693588256836, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4477, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.6886237263679504, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4283, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.6619666814804077, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4212, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.534345805644989, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4368, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.7200596332550049, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4117, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.6517333388328552, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4215, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.6811349391937256, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4057, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.4850296974182129, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4407, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.5540636777877808, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4077, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.4562750458717346, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4208, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.8182024359703064, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4151, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.39840784668922424, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4633, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.5493362545967102, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4378, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.6024540066719055, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4087, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.6966805458068848, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4299, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "step": 140, | |
| "total_flos": 190411138334720.0, | |
| "train_loss": 22.297974420232432, | |
| "train_runtime": 30487.3439, | |
| "train_samples_per_second": 0.074, | |
| "train_steps_per_second": 0.005 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 140, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 30, | |
| "total_flos": 190411138334720.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |