| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.7523048555623848, | |
| "eval_steps": 204, | |
| "global_step": 612, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.001229256299938537, | |
| "grad_norm": 0.19411148130893707, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.1612, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.001229256299938537, | |
| "eval_loss": 2.1468453407287598, | |
| "eval_runtime": 66.4837, | |
| "eval_samples_per_second": 10.303, | |
| "eval_steps_per_second": 5.159, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.002458512599877074, | |
| "grad_norm": 0.2264145463705063, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.4401, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0036877688998156115, | |
| "grad_norm": 0.2364473193883896, | |
| "learning_rate": 6e-06, | |
| "loss": 1.4676, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.004917025199754148, | |
| "grad_norm": 0.24018821120262146, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.3851, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.006146281499692686, | |
| "grad_norm": 0.23238497972488403, | |
| "learning_rate": 1e-05, | |
| "loss": 1.213, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.007375537799631223, | |
| "grad_norm": 0.24634625017642975, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.2627, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.008604794099569761, | |
| "grad_norm": 0.26495596766471863, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 1.3908, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.009834050399508297, | |
| "grad_norm": 0.2719455361366272, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.3814, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.011063306699446834, | |
| "grad_norm": 0.26454323530197144, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.2438, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.012292562999385371, | |
| "grad_norm": 0.3004608750343323, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3694, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.013521819299323909, | |
| "grad_norm": 0.3035408854484558, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 1.4792, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.014751075599262446, | |
| "grad_norm": 0.4270775020122528, | |
| "learning_rate": 2.4e-05, | |
| "loss": 1.1673, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.015980331899200985, | |
| "grad_norm": 0.4388391971588135, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 1.5171, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.017209588199139522, | |
| "grad_norm": 0.7133700847625732, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 1.0732, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.01843884449907806, | |
| "grad_norm": 1.026840329170227, | |
| "learning_rate": 3e-05, | |
| "loss": 1.1705, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.019668100799016593, | |
| "grad_norm": 0.7934454679489136, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 1.3509, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.02089735709895513, | |
| "grad_norm": 0.8138520121574402, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 1.181, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.022126613398893668, | |
| "grad_norm": 1.7830528020858765, | |
| "learning_rate": 3.6e-05, | |
| "loss": 2.1836, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.023355869698832205, | |
| "grad_norm": 10.527496337890625, | |
| "learning_rate": 3.8e-05, | |
| "loss": 3.3749, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.024585125998770743, | |
| "grad_norm": 6.364173889160156, | |
| "learning_rate": 4e-05, | |
| "loss": 2.9191, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02581438229870928, | |
| "grad_norm": 7.087876796722412, | |
| "learning_rate": 4.2e-05, | |
| "loss": 3.0788, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.027043638598647817, | |
| "grad_norm": 5.370169639587402, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 2.7809, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.028272894898586354, | |
| "grad_norm": 4.118806838989258, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 2.7475, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.02950215119852489, | |
| "grad_norm": 4.46057653427124, | |
| "learning_rate": 4.8e-05, | |
| "loss": 2.6906, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.03073140749846343, | |
| "grad_norm": 3.8601913452148438, | |
| "learning_rate": 5e-05, | |
| "loss": 2.652, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.03196066379840197, | |
| "grad_norm": 0.19972144067287445, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 1.1845, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.03318992009834051, | |
| "grad_norm": 0.21230019629001617, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 1.2875, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.034419176398279044, | |
| "grad_norm": 0.22694356739521027, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.1829, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.03564843269821758, | |
| "grad_norm": 0.2587474584579468, | |
| "learning_rate": 5.8e-05, | |
| "loss": 1.3909, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.03687768899815612, | |
| "grad_norm": 0.24409259855747223, | |
| "learning_rate": 6e-05, | |
| "loss": 1.1671, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03810694529809465, | |
| "grad_norm": 0.26323097944259644, | |
| "learning_rate": 6.2e-05, | |
| "loss": 1.2055, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.03933620159803319, | |
| "grad_norm": 0.2842409908771515, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.3051, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.040565457897971724, | |
| "grad_norm": 0.32476744055747986, | |
| "learning_rate": 6.6e-05, | |
| "loss": 1.0828, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.04179471419791026, | |
| "grad_norm": 0.32893380522727966, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 1.3896, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0430239704978488, | |
| "grad_norm": 0.3359004855155945, | |
| "learning_rate": 7e-05, | |
| "loss": 1.3713, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.044253226797787336, | |
| "grad_norm": 0.5471646189689636, | |
| "learning_rate": 7.2e-05, | |
| "loss": 1.2572, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.04548248309772587, | |
| "grad_norm": 0.5404387712478638, | |
| "learning_rate": 7.4e-05, | |
| "loss": 1.2321, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.04671173939766441, | |
| "grad_norm": 1.0199828147888184, | |
| "learning_rate": 7.6e-05, | |
| "loss": 0.8434, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.04794099569760295, | |
| "grad_norm": 1.5890088081359863, | |
| "learning_rate": 7.800000000000001e-05, | |
| "loss": 0.7781, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.049170251997541485, | |
| "grad_norm": 0.7897126078605652, | |
| "learning_rate": 8e-05, | |
| "loss": 1.1982, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.05039950829748002, | |
| "grad_norm": 0.7874982953071594, | |
| "learning_rate": 8.2e-05, | |
| "loss": 1.3432, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.05162876459741856, | |
| "grad_norm": 1.3902230262756348, | |
| "learning_rate": 8.4e-05, | |
| "loss": 1.7817, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.0528580208973571, | |
| "grad_norm": 4.006369590759277, | |
| "learning_rate": 8.6e-05, | |
| "loss": 1.8832, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.054087277197295634, | |
| "grad_norm": 7.2500996589660645, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 1.3256, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.05531653349723417, | |
| "grad_norm": 5.088122844696045, | |
| "learning_rate": 9e-05, | |
| "loss": 1.4272, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.05654578979717271, | |
| "grad_norm": 2.9680252075195312, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 1.8063, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.057775046097111246, | |
| "grad_norm": 3.4886820316314697, | |
| "learning_rate": 9.4e-05, | |
| "loss": 1.7842, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.05900430239704978, | |
| "grad_norm": 2.635120153427124, | |
| "learning_rate": 9.6e-05, | |
| "loss": 1.7775, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.06023355869698832, | |
| "grad_norm": 2.7715940475463867, | |
| "learning_rate": 9.8e-05, | |
| "loss": 1.6795, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.06146281499692686, | |
| "grad_norm": 4.598182678222656, | |
| "learning_rate": 0.0001, | |
| "loss": 2.0141, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0626920712968654, | |
| "grad_norm": 0.4595154821872711, | |
| "learning_rate": 9.999957617159031e-05, | |
| "loss": 1.1302, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.06392132759680394, | |
| "grad_norm": 0.3996050953865051, | |
| "learning_rate": 9.999830469354645e-05, | |
| "loss": 1.3499, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.06515058389674247, | |
| "grad_norm": 0.4248620867729187, | |
| "learning_rate": 9.999618558742398e-05, | |
| "loss": 1.4393, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.06637984019668101, | |
| "grad_norm": 0.37063130736351013, | |
| "learning_rate": 9.999321888914836e-05, | |
| "loss": 1.4761, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.06760909649661954, | |
| "grad_norm": 0.3327302038669586, | |
| "learning_rate": 9.998940464901447e-05, | |
| "loss": 1.1365, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.06883835279655809, | |
| "grad_norm": 0.3424387276172638, | |
| "learning_rate": 9.998474293168562e-05, | |
| "loss": 1.2037, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.07006760909649662, | |
| "grad_norm": 0.34453633427619934, | |
| "learning_rate": 9.997923381619256e-05, | |
| "loss": 0.9586, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.07129686539643516, | |
| "grad_norm": 0.3327544033527374, | |
| "learning_rate": 9.997287739593206e-05, | |
| "loss": 1.3026, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.0725261216963737, | |
| "grad_norm": 0.33542299270629883, | |
| "learning_rate": 9.996567377866537e-05, | |
| "loss": 1.1601, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.07375537799631224, | |
| "grad_norm": 0.5743572115898132, | |
| "learning_rate": 9.99576230865164e-05, | |
| "loss": 1.3892, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.07498463429625077, | |
| "grad_norm": 0.4624180495738983, | |
| "learning_rate": 9.994872545596966e-05, | |
| "loss": 1.2519, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.0762138905961893, | |
| "grad_norm": 0.6259918808937073, | |
| "learning_rate": 9.993898103786786e-05, | |
| "loss": 1.315, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.07744314689612784, | |
| "grad_norm": 0.6186118125915527, | |
| "learning_rate": 9.992838999740947e-05, | |
| "loss": 0.877, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.07867240319606637, | |
| "grad_norm": 0.6788893342018127, | |
| "learning_rate": 9.991695251414583e-05, | |
| "loss": 0.886, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.07990165949600492, | |
| "grad_norm": 0.7688488960266113, | |
| "learning_rate": 9.990466878197817e-05, | |
| "loss": 0.7427, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.08113091579594345, | |
| "grad_norm": 0.6739158630371094, | |
| "learning_rate": 9.989153900915427e-05, | |
| "loss": 1.091, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.08236017209588199, | |
| "grad_norm": 1.0515763759613037, | |
| "learning_rate": 9.987756341826493e-05, | |
| "loss": 1.4195, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.08358942839582052, | |
| "grad_norm": 2.324380397796631, | |
| "learning_rate": 9.98627422462403e-05, | |
| "loss": 1.8108, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.08481868469575907, | |
| "grad_norm": 4.131134510040283, | |
| "learning_rate": 9.98470757443457e-05, | |
| "loss": 1.2769, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.0860479409956976, | |
| "grad_norm": 6.158152103424072, | |
| "learning_rate": 9.983056417817747e-05, | |
| "loss": 1.609, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.08727719729563614, | |
| "grad_norm": 2.710057020187378, | |
| "learning_rate": 9.981320782765846e-05, | |
| "loss": 1.6382, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.08850645359557467, | |
| "grad_norm": 2.729590654373169, | |
| "learning_rate": 9.979500698703323e-05, | |
| "loss": 1.8179, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.08973570989551322, | |
| "grad_norm": 2.1861114501953125, | |
| "learning_rate": 9.977596196486314e-05, | |
| "loss": 1.7416, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.09096496619545175, | |
| "grad_norm": 2.614532947540283, | |
| "learning_rate": 9.975607308402101e-05, | |
| "loss": 1.8413, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.09219422249539029, | |
| "grad_norm": 3.3295183181762695, | |
| "learning_rate": 9.973534068168579e-05, | |
| "loss": 2.1946, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.09342347879532882, | |
| "grad_norm": 0.3009834885597229, | |
| "learning_rate": 9.97137651093367e-05, | |
| "loss": 1.1058, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.09465273509526737, | |
| "grad_norm": 0.2889084815979004, | |
| "learning_rate": 9.969134673274738e-05, | |
| "loss": 1.0812, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.0958819913952059, | |
| "grad_norm": 0.26639047265052795, | |
| "learning_rate": 9.966808593197959e-05, | |
| "loss": 1.2787, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.09711124769514444, | |
| "grad_norm": 0.2839871048927307, | |
| "learning_rate": 9.964398310137688e-05, | |
| "loss": 1.2314, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.09834050399508297, | |
| "grad_norm": 0.29856863617897034, | |
| "learning_rate": 9.961903864955783e-05, | |
| "loss": 1.1781, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.09956976029502151, | |
| "grad_norm": 0.3113296329975128, | |
| "learning_rate": 9.959325299940914e-05, | |
| "loss": 1.1297, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.10079901659496004, | |
| "grad_norm": 0.3259466290473938, | |
| "learning_rate": 9.956662658807842e-05, | |
| "loss": 1.3892, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.10202827289489859, | |
| "grad_norm": 0.3366626501083374, | |
| "learning_rate": 9.95391598669669e-05, | |
| "loss": 1.1833, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.10325752919483712, | |
| "grad_norm": 0.3032483458518982, | |
| "learning_rate": 9.95108533017216e-05, | |
| "loss": 1.1729, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.10448678549477566, | |
| "grad_norm": 0.4028280973434448, | |
| "learning_rate": 9.948170737222762e-05, | |
| "loss": 1.1019, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1057160417947142, | |
| "grad_norm": 0.3796052932739258, | |
| "learning_rate": 9.945172257259986e-05, | |
| "loss": 1.3822, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.10694529809465274, | |
| "grad_norm": 0.3956368565559387, | |
| "learning_rate": 9.942089941117472e-05, | |
| "loss": 1.2101, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.10817455439459127, | |
| "grad_norm": 0.5040555596351624, | |
| "learning_rate": 9.938923841050147e-05, | |
| "loss": 1.059, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.10940381069452981, | |
| "grad_norm": 0.7209507822990417, | |
| "learning_rate": 9.935674010733336e-05, | |
| "loss": 0.9387, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.11063306699446834, | |
| "grad_norm": 0.6711410284042358, | |
| "learning_rate": 9.932340505261855e-05, | |
| "loss": 0.9325, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.11186232329440689, | |
| "grad_norm": 0.670559823513031, | |
| "learning_rate": 9.928923381149078e-05, | |
| "loss": 1.1188, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.11309157959434542, | |
| "grad_norm": 1.4009896516799927, | |
| "learning_rate": 9.925422696325975e-05, | |
| "loss": 1.4021, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.11432083589428396, | |
| "grad_norm": 2.7449545860290527, | |
| "learning_rate": 9.921838510140135e-05, | |
| "loss": 1.7181, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.11555009219422249, | |
| "grad_norm": 3.5462844371795654, | |
| "learning_rate": 9.918170883354755e-05, | |
| "loss": 1.4934, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.11677934849416104, | |
| "grad_norm": 3.204674005508423, | |
| "learning_rate": 9.914419878147611e-05, | |
| "loss": 1.2952, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.11800860479409957, | |
| "grad_norm": 2.583436965942383, | |
| "learning_rate": 9.910585558110006e-05, | |
| "loss": 1.418, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.11923786109403811, | |
| "grad_norm": 3.0214803218841553, | |
| "learning_rate": 9.906667988245692e-05, | |
| "loss": 1.8579, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.12046711739397664, | |
| "grad_norm": 2.359790325164795, | |
| "learning_rate": 9.902667234969764e-05, | |
| "loss": 1.2705, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.12169637369391519, | |
| "grad_norm": 2.093607187271118, | |
| "learning_rate": 9.898583366107538e-05, | |
| "loss": 1.4241, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.12292562999385372, | |
| "grad_norm": 2.613720655441284, | |
| "learning_rate": 9.8944164508934e-05, | |
| "loss": 1.7558, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.12415488629379226, | |
| "grad_norm": 0.29464319348335266, | |
| "learning_rate": 9.890166559969631e-05, | |
| "loss": 1.1966, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.1253841425937308, | |
| "grad_norm": 0.27224430441856384, | |
| "learning_rate": 9.885833765385212e-05, | |
| "loss": 1.3172, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.12661339889366932, | |
| "grad_norm": 0.2738960385322571, | |
| "learning_rate": 9.881418140594603e-05, | |
| "loss": 1.2875, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.12784265519360788, | |
| "grad_norm": 0.274746298789978, | |
| "learning_rate": 9.876919760456492e-05, | |
| "loss": 1.3156, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.1290719114935464, | |
| "grad_norm": 0.3050672113895416, | |
| "learning_rate": 9.872338701232526e-05, | |
| "loss": 1.2426, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.13030116779348494, | |
| "grad_norm": 0.2726648449897766, | |
| "learning_rate": 9.867675040586034e-05, | |
| "loss": 1.1997, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.13153042409342347, | |
| "grad_norm": 0.2615199685096741, | |
| "learning_rate": 9.862928857580687e-05, | |
| "loss": 1.1518, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.13275968039336203, | |
| "grad_norm": 0.27568066120147705, | |
| "learning_rate": 9.858100232679175e-05, | |
| "loss": 0.9874, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.13398893669330056, | |
| "grad_norm": 0.29168951511383057, | |
| "learning_rate": 9.853189247741833e-05, | |
| "loss": 1.2147, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.1352181929932391, | |
| "grad_norm": 0.30630671977996826, | |
| "learning_rate": 9.848195986025257e-05, | |
| "loss": 1.2474, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.13644744929317762, | |
| "grad_norm": 0.3246194124221802, | |
| "learning_rate": 9.843120532180896e-05, | |
| "loss": 1.1839, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.13767670559311618, | |
| "grad_norm": 0.34899017214775085, | |
| "learning_rate": 9.837962972253612e-05, | |
| "loss": 1.2389, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.1389059618930547, | |
| "grad_norm": 0.3848627805709839, | |
| "learning_rate": 9.83272339368022e-05, | |
| "loss": 1.1833, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.14013521819299324, | |
| "grad_norm": 0.4109489917755127, | |
| "learning_rate": 9.827401885288013e-05, | |
| "loss": 1.1026, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.14136447449293177, | |
| "grad_norm": 0.6600728034973145, | |
| "learning_rate": 9.821998537293245e-05, | |
| "loss": 1.4073, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.14259373079287033, | |
| "grad_norm": 0.5556017756462097, | |
| "learning_rate": 9.816513441299613e-05, | |
| "loss": 0.6878, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.14382298709280886, | |
| "grad_norm": 0.5937761068344116, | |
| "learning_rate": 9.810946690296698e-05, | |
| "loss": 0.7988, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.1450522433927474, | |
| "grad_norm": 0.6892157196998596, | |
| "learning_rate": 9.80529837865839e-05, | |
| "loss": 1.2152, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.14628149969268592, | |
| "grad_norm": 1.1046031713485718, | |
| "learning_rate": 9.799568602141283e-05, | |
| "loss": 1.4396, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.14751075599262448, | |
| "grad_norm": 3.366898536682129, | |
| "learning_rate": 9.793757457883062e-05, | |
| "loss": 1.6062, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.148740012292563, | |
| "grad_norm": 4.46527624130249, | |
| "learning_rate": 9.787865044400848e-05, | |
| "loss": 1.041, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.14996926859250154, | |
| "grad_norm": 3.8992013931274414, | |
| "learning_rate": 9.781891461589531e-05, | |
| "loss": 1.6166, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.15119852489244007, | |
| "grad_norm": 2.6794042587280273, | |
| "learning_rate": 9.775836810720074e-05, | |
| "loss": 1.5444, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.1524277811923786, | |
| "grad_norm": 2.1487152576446533, | |
| "learning_rate": 9.769701194437799e-05, | |
| "loss": 1.4051, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.15365703749231716, | |
| "grad_norm": 2.6264848709106445, | |
| "learning_rate": 9.763484716760649e-05, | |
| "loss": 1.7286, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.15488629379225569, | |
| "grad_norm": 0.2960408329963684, | |
| "learning_rate": 9.757187483077413e-05, | |
| "loss": 1.1932, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.15611555009219422, | |
| "grad_norm": 0.2633897364139557, | |
| "learning_rate": 9.750809600145954e-05, | |
| "loss": 1.2997, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.15734480639213275, | |
| "grad_norm": 0.2459549605846405, | |
| "learning_rate": 9.744351176091393e-05, | |
| "loss": 1.0985, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.1585740626920713, | |
| "grad_norm": 0.30462849140167236, | |
| "learning_rate": 9.737812320404271e-05, | |
| "loss": 1.4303, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.15980331899200984, | |
| "grad_norm": 0.27317526936531067, | |
| "learning_rate": 9.731193143938704e-05, | |
| "loss": 1.224, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.16103257529194837, | |
| "grad_norm": 0.26538556814193726, | |
| "learning_rate": 9.724493758910491e-05, | |
| "loss": 1.2667, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.1622618315918869, | |
| "grad_norm": 0.28112831711769104, | |
| "learning_rate": 9.71771427889522e-05, | |
| "loss": 1.1212, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.16349108789182545, | |
| "grad_norm": 0.2989320755004883, | |
| "learning_rate": 9.71085481882634e-05, | |
| "loss": 1.0484, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.16472034419176398, | |
| "grad_norm": 0.2814895212650299, | |
| "learning_rate": 9.703915494993215e-05, | |
| "loss": 0.7544, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.16594960049170251, | |
| "grad_norm": 0.3104398846626282, | |
| "learning_rate": 9.696896425039146e-05, | |
| "loss": 1.0323, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.16717885679164105, | |
| "grad_norm": 0.4948181211948395, | |
| "learning_rate": 9.689797727959387e-05, | |
| "loss": 1.2073, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.1684081130915796, | |
| "grad_norm": 0.4018343985080719, | |
| "learning_rate": 9.682619524099112e-05, | |
| "loss": 1.2409, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.16963736939151813, | |
| "grad_norm": 0.5637558102607727, | |
| "learning_rate": 9.675361935151395e-05, | |
| "loss": 1.3184, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.17086662569145666, | |
| "grad_norm": 0.7405252456665039, | |
| "learning_rate": 9.66802508415513e-05, | |
| "loss": 1.0983, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.1720958819913952, | |
| "grad_norm": 0.6686736345291138, | |
| "learning_rate": 9.660609095492952e-05, | |
| "loss": 1.0025, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.17332513829133375, | |
| "grad_norm": 0.7121345400810242, | |
| "learning_rate": 9.653114094889127e-05, | |
| "loss": 0.9337, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.17455439459127228, | |
| "grad_norm": 1.06205153465271, | |
| "learning_rate": 9.645540209407425e-05, | |
| "loss": 1.2931, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.1757836508912108, | |
| "grad_norm": 2.3874034881591797, | |
| "learning_rate": 9.637887567448959e-05, | |
| "loss": 1.5124, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.17701290719114934, | |
| "grad_norm": 2.6609811782836914, | |
| "learning_rate": 9.630156298750011e-05, | |
| "loss": 1.4161, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.1782421634910879, | |
| "grad_norm": 2.413705587387085, | |
| "learning_rate": 9.622346534379833e-05, | |
| "loss": 1.2768, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.17947141979102643, | |
| "grad_norm": 2.920910120010376, | |
| "learning_rate": 9.614458406738427e-05, | |
| "loss": 1.0866, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.18070067609096496, | |
| "grad_norm": 2.389439582824707, | |
| "learning_rate": 9.606492049554297e-05, | |
| "loss": 1.4862, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.1819299323909035, | |
| "grad_norm": 2.03515887260437, | |
| "learning_rate": 9.598447597882181e-05, | |
| "loss": 1.3503, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.18315918869084205, | |
| "grad_norm": 2.016889810562134, | |
| "learning_rate": 9.590325188100768e-05, | |
| "loss": 1.2565, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.18438844499078058, | |
| "grad_norm": 2.1591711044311523, | |
| "learning_rate": 9.582124957910375e-05, | |
| "loss": 1.1261, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1856177012907191, | |
| "grad_norm": 0.2707172632217407, | |
| "learning_rate": 9.573847046330628e-05, | |
| "loss": 1.1045, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.18684695759065764, | |
| "grad_norm": 0.25980842113494873, | |
| "learning_rate": 9.565491593698086e-05, | |
| "loss": 1.274, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.1880762138905962, | |
| "grad_norm": 0.25503602623939514, | |
| "learning_rate": 9.55705874166388e-05, | |
| "loss": 1.0971, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.18930547019053473, | |
| "grad_norm": 0.27756351232528687, | |
| "learning_rate": 9.548548633191299e-05, | |
| "loss": 1.215, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.19053472649047326, | |
| "grad_norm": 0.2732703387737274, | |
| "learning_rate": 9.539961412553375e-05, | |
| "loss": 1.1326, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.1917639827904118, | |
| "grad_norm": 0.28855475783348083, | |
| "learning_rate": 9.531297225330429e-05, | |
| "loss": 1.2862, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.19299323909035035, | |
| "grad_norm": 0.3158769905567169, | |
| "learning_rate": 9.522556218407608e-05, | |
| "loss": 1.2254, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.19422249539028888, | |
| "grad_norm": 0.30355289578437805, | |
| "learning_rate": 9.513738539972394e-05, | |
| "loss": 1.062, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.1954517516902274, | |
| "grad_norm": 0.3448358178138733, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 0.9856, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.19668100799016594, | |
| "grad_norm": 0.3306958079338074, | |
| "learning_rate": 9.495873767811305e-05, | |
| "loss": 1.2696, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1979102642901045, | |
| "grad_norm": 0.4231187105178833, | |
| "learning_rate": 9.486826976949345e-05, | |
| "loss": 1.1711, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.19913952059004303, | |
| "grad_norm": 0.5289990901947021, | |
| "learning_rate": 9.477704120297697e-05, | |
| "loss": 1.4088, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.20036877688998156, | |
| "grad_norm": 0.5111967921257019, | |
| "learning_rate": 9.468505352517394e-05, | |
| "loss": 1.1683, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.2015980331899201, | |
| "grad_norm": 0.7477207779884338, | |
| "learning_rate": 9.459230829556401e-05, | |
| "loss": 0.995, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.20282728948985865, | |
| "grad_norm": 0.7836649417877197, | |
| "learning_rate": 9.449880708646971e-05, | |
| "loss": 0.8027, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.20405654578979718, | |
| "grad_norm": 0.6803653240203857, | |
| "learning_rate": 9.440455148302977e-05, | |
| "loss": 0.9725, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.2052858020897357, | |
| "grad_norm": 0.8779723048210144, | |
| "learning_rate": 9.430954308317233e-05, | |
| "loss": 1.1995, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.20651505838967424, | |
| "grad_norm": 1.3584879636764526, | |
| "learning_rate": 9.421378349758769e-05, | |
| "loss": 1.4558, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.2077443146896128, | |
| "grad_norm": 2.1976521015167236, | |
| "learning_rate": 9.411727434970121e-05, | |
| "loss": 1.0717, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.20897357098955133, | |
| "grad_norm": 3.9302353858947754, | |
| "learning_rate": 9.402001727564565e-05, | |
| "loss": 1.5138, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.21020282728948986, | |
| "grad_norm": 3.9594686031341553, | |
| "learning_rate": 9.392201392423342e-05, | |
| "loss": 1.4295, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.2114320835894284, | |
| "grad_norm": 3.2994837760925293, | |
| "learning_rate": 9.382326595692868e-05, | |
| "loss": 1.8676, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.21266133988936695, | |
| "grad_norm": 2.219341993331909, | |
| "learning_rate": 9.372377504781924e-05, | |
| "loss": 1.3185, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.21389059618930548, | |
| "grad_norm": 2.3389649391174316, | |
| "learning_rate": 9.362354288358803e-05, | |
| "loss": 0.9969, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.215119852489244, | |
| "grad_norm": 3.8493995666503906, | |
| "learning_rate": 9.35225711634846e-05, | |
| "loss": 1.2903, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.21634910878918254, | |
| "grad_norm": 0.24931700527668, | |
| "learning_rate": 9.34208615992963e-05, | |
| "loss": 1.051, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.2175783650891211, | |
| "grad_norm": 0.2944095730781555, | |
| "learning_rate": 9.331841591531922e-05, | |
| "loss": 1.3364, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.21880762138905963, | |
| "grad_norm": 0.26118403673171997, | |
| "learning_rate": 9.321523584832905e-05, | |
| "loss": 1.1487, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.22003687768899816, | |
| "grad_norm": 0.29458168148994446, | |
| "learning_rate": 9.311132314755149e-05, | |
| "loss": 1.365, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.2212661339889367, | |
| "grad_norm": 0.2739919424057007, | |
| "learning_rate": 9.300667957463278e-05, | |
| "loss": 1.2595, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.22249539028887522, | |
| "grad_norm": 0.25647538900375366, | |
| "learning_rate": 9.290130690360965e-05, | |
| "loss": 0.9865, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.22372464658881377, | |
| "grad_norm": 0.27343517541885376, | |
| "learning_rate": 9.279520692087938e-05, | |
| "loss": 1.1263, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.2249539028887523, | |
| "grad_norm": 0.3220975697040558, | |
| "learning_rate": 9.268838142516943e-05, | |
| "loss": 1.3404, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.22618315918869084, | |
| "grad_norm": 0.3012546896934509, | |
| "learning_rate": 9.258083222750703e-05, | |
| "loss": 0.934, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.22741241548862937, | |
| "grad_norm": 0.3433031439781189, | |
| "learning_rate": 9.247256115118835e-05, | |
| "loss": 1.1895, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.22864167178856792, | |
| "grad_norm": 0.3515290915966034, | |
| "learning_rate": 9.236357003174775e-05, | |
| "loss": 1.3236, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.22987092808850645, | |
| "grad_norm": 0.4033795893192291, | |
| "learning_rate": 9.225386071692654e-05, | |
| "loss": 1.2089, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.23110018438844498, | |
| "grad_norm": 0.42729562520980835, | |
| "learning_rate": 9.214343506664168e-05, | |
| "loss": 1.1346, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.23232944068838352, | |
| "grad_norm": 0.6692906618118286, | |
| "learning_rate": 9.203229495295429e-05, | |
| "loss": 1.0211, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.23355869698832207, | |
| "grad_norm": 0.6882857084274292, | |
| "learning_rate": 9.192044226003789e-05, | |
| "loss": 0.8235, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2347879532882606, | |
| "grad_norm": 0.6821665167808533, | |
| "learning_rate": 9.18078788841464e-05, | |
| "loss": 0.8171, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.23601720958819913, | |
| "grad_norm": 0.7368921041488647, | |
| "learning_rate": 9.169460673358212e-05, | |
| "loss": 0.9993, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.23724646588813766, | |
| "grad_norm": 0.9759008884429932, | |
| "learning_rate": 9.158062772866325e-05, | |
| "loss": 1.2029, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.23847572218807622, | |
| "grad_norm": 2.167100667953491, | |
| "learning_rate": 9.146594380169143e-05, | |
| "loss": 1.1393, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.23970497848801475, | |
| "grad_norm": 2.76292085647583, | |
| "learning_rate": 9.135055689691888e-05, | |
| "loss": 0.946, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.24093423478795328, | |
| "grad_norm": 3.504427671432495, | |
| "learning_rate": 9.123446897051555e-05, | |
| "loss": 1.7001, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.2421634910878918, | |
| "grad_norm": 2.606448173522949, | |
| "learning_rate": 9.111768199053588e-05, | |
| "loss": 1.6293, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.24339274738783037, | |
| "grad_norm": 2.1803855895996094, | |
| "learning_rate": 9.100019793688549e-05, | |
| "loss": 1.2392, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.2446220036877689, | |
| "grad_norm": 2.3470633029937744, | |
| "learning_rate": 9.088201880128755e-05, | |
| "loss": 1.0844, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.24585125998770743, | |
| "grad_norm": 2.47255802154541, | |
| "learning_rate": 9.076314658724906e-05, | |
| "loss": 1.19, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.24708051628764596, | |
| "grad_norm": 0.2115241140127182, | |
| "learning_rate": 9.064358331002691e-05, | |
| "loss": 0.9038, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.24830977258758452, | |
| "grad_norm": 0.2693980038166046, | |
| "learning_rate": 9.05233309965936e-05, | |
| "loss": 1.0014, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.24953902888752305, | |
| "grad_norm": 0.28890225291252136, | |
| "learning_rate": 9.040239168560303e-05, | |
| "loss": 1.1698, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.2507682851874616, | |
| "grad_norm": 0.27143335342407227, | |
| "learning_rate": 9.028076742735583e-05, | |
| "loss": 1.1856, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.2507682851874616, | |
| "eval_loss": 1.0315037965774536, | |
| "eval_runtime": 65.4064, | |
| "eval_samples_per_second": 10.473, | |
| "eval_steps_per_second": 5.244, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.2519975414874001, | |
| "grad_norm": 0.3105545938014984, | |
| "learning_rate": 9.015846028376462e-05, | |
| "loss": 1.2827, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.25322679778733864, | |
| "grad_norm": 0.2826372981071472, | |
| "learning_rate": 9.00354723283191e-05, | |
| "loss": 1.1159, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.2544560540872772, | |
| "grad_norm": 0.2823708951473236, | |
| "learning_rate": 8.991180564605086e-05, | |
| "loss": 1.0368, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.25568531038721576, | |
| "grad_norm": 0.28265297412872314, | |
| "learning_rate": 8.978746233349802e-05, | |
| "loss": 1.1583, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.2569145666871543, | |
| "grad_norm": 0.3202212452888489, | |
| "learning_rate": 8.966244449866973e-05, | |
| "loss": 1.2069, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.2581438229870928, | |
| "grad_norm": 0.30576291680336, | |
| "learning_rate": 8.953675426101038e-05, | |
| "loss": 1.1588, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.25937307928703135, | |
| "grad_norm": 0.3853960633277893, | |
| "learning_rate": 8.941039375136371e-05, | |
| "loss": 1.1947, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.2606023355869699, | |
| "grad_norm": 0.4404067099094391, | |
| "learning_rate": 8.928336511193669e-05, | |
| "loss": 1.0786, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.2618315918869084, | |
| "grad_norm": 0.422333300113678, | |
| "learning_rate": 8.915567049626315e-05, | |
| "loss": 1.1454, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.26306084818684694, | |
| "grad_norm": 0.5277565121650696, | |
| "learning_rate": 8.902731206916734e-05, | |
| "loss": 0.7775, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.26429010448678547, | |
| "grad_norm": 0.7032243609428406, | |
| "learning_rate": 8.889829200672719e-05, | |
| "loss": 0.5771, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.26551936078672406, | |
| "grad_norm": 0.6663339734077454, | |
| "learning_rate": 8.876861249623739e-05, | |
| "loss": 0.616, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.2667486170866626, | |
| "grad_norm": 0.8129518628120422, | |
| "learning_rate": 8.863827573617238e-05, | |
| "loss": 1.1483, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.2679778733866011, | |
| "grad_norm": 1.0273211002349854, | |
| "learning_rate": 8.850728393614902e-05, | |
| "loss": 1.2066, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.26920712968653965, | |
| "grad_norm": 1.5424954891204834, | |
| "learning_rate": 8.837563931688919e-05, | |
| "loss": 1.247, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.2704363859864782, | |
| "grad_norm": 2.9167752265930176, | |
| "learning_rate": 8.824334411018204e-05, | |
| "loss": 1.3413, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.2716656422864167, | |
| "grad_norm": 5.498292446136475, | |
| "learning_rate": 8.811040055884629e-05, | |
| "loss": 1.0072, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.27289489858635524, | |
| "grad_norm": 3.1687686443328857, | |
| "learning_rate": 8.797681091669206e-05, | |
| "loss": 1.3309, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.27412415488629377, | |
| "grad_norm": 2.760160446166992, | |
| "learning_rate": 8.784257744848279e-05, | |
| "loss": 1.5268, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.27535341118623236, | |
| "grad_norm": 2.3323326110839844, | |
| "learning_rate": 8.770770242989679e-05, | |
| "loss": 1.27, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.2765826674861709, | |
| "grad_norm": 2.150510549545288, | |
| "learning_rate": 8.75721881474886e-05, | |
| "loss": 1.0602, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.2778119237861094, | |
| "grad_norm": 0.23049846291542053, | |
| "learning_rate": 8.743603689865039e-05, | |
| "loss": 1.0067, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.27904118008604795, | |
| "grad_norm": 0.2650708556175232, | |
| "learning_rate": 8.729925099157281e-05, | |
| "loss": 1.1932, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.2802704363859865, | |
| "grad_norm": 0.2723963260650635, | |
| "learning_rate": 8.7161832745206e-05, | |
| "loss": 1.2495, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.281499692685925, | |
| "grad_norm": 0.26627010107040405, | |
| "learning_rate": 8.702378448922026e-05, | |
| "loss": 1.2837, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.28272894898586354, | |
| "grad_norm": 0.2728361189365387, | |
| "learning_rate": 8.688510856396648e-05, | |
| "loss": 1.2969, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.28395820528580207, | |
| "grad_norm": 0.26788559556007385, | |
| "learning_rate": 8.674580732043656e-05, | |
| "loss": 1.0944, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.28518746158574065, | |
| "grad_norm": 0.3129604160785675, | |
| "learning_rate": 8.660588312022344e-05, | |
| "loss": 1.3591, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.2864167178856792, | |
| "grad_norm": 0.32250627875328064, | |
| "learning_rate": 8.646533833548119e-05, | |
| "loss": 1.1469, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.2876459741856177, | |
| "grad_norm": 0.32614386081695557, | |
| "learning_rate": 8.632417534888473e-05, | |
| "loss": 1.3551, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.28887523048555624, | |
| "grad_norm": 0.3620636463165283, | |
| "learning_rate": 8.61823965535894e-05, | |
| "loss": 1.1427, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.2901044867854948, | |
| "grad_norm": 0.39082473516464233, | |
| "learning_rate": 8.604000435319047e-05, | |
| "loss": 1.0041, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.2913337430854333, | |
| "grad_norm": 0.3823097050189972, | |
| "learning_rate": 8.589700116168232e-05, | |
| "loss": 1.1756, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.29256299938537184, | |
| "grad_norm": 0.5359341502189636, | |
| "learning_rate": 8.575338940341757e-05, | |
| "loss": 1.1814, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.29379225568531037, | |
| "grad_norm": 0.6902546286582947, | |
| "learning_rate": 8.560917151306593e-05, | |
| "loss": 0.9253, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.29502151198524895, | |
| "grad_norm": 0.7236252427101135, | |
| "learning_rate": 8.5464349935573e-05, | |
| "loss": 0.6398, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2962507682851875, | |
| "grad_norm": 0.7172759175300598, | |
| "learning_rate": 8.53189271261187e-05, | |
| "loss": 0.9061, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.297480024585126, | |
| "grad_norm": 0.7999723553657532, | |
| "learning_rate": 8.517290555007578e-05, | |
| "loss": 1.0691, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.29870928088506454, | |
| "grad_norm": 1.235872745513916, | |
| "learning_rate": 8.502628768296788e-05, | |
| "loss": 1.5235, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.2999385371850031, | |
| "grad_norm": 1.9676207304000854, | |
| "learning_rate": 8.487907601042777e-05, | |
| "loss": 1.5859, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.3011677934849416, | |
| "grad_norm": 3.5035860538482666, | |
| "learning_rate": 8.473127302815496e-05, | |
| "loss": 1.1743, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.30239704978488013, | |
| "grad_norm": 4.519472599029541, | |
| "learning_rate": 8.458288124187359e-05, | |
| "loss": 0.7165, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.30362630608481866, | |
| "grad_norm": 2.3718838691711426, | |
| "learning_rate": 8.443390316728987e-05, | |
| "loss": 1.1449, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.3048555623847572, | |
| "grad_norm": 2.1668829917907715, | |
| "learning_rate": 8.428434133004937e-05, | |
| "loss": 1.0383, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.3060848186846958, | |
| "grad_norm": 3.2350733280181885, | |
| "learning_rate": 8.413419826569435e-05, | |
| "loss": 1.2341, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.3073140749846343, | |
| "grad_norm": 2.3541886806488037, | |
| "learning_rate": 8.398347651962064e-05, | |
| "loss": 1.0355, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.30854333128457284, | |
| "grad_norm": 0.2730487883090973, | |
| "learning_rate": 8.383217864703456e-05, | |
| "loss": 1.2813, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.30977258758451137, | |
| "grad_norm": 0.2517383098602295, | |
| "learning_rate": 8.36803072129096e-05, | |
| "loss": 1.1793, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.3110018438844499, | |
| "grad_norm": 0.28486472368240356, | |
| "learning_rate": 8.352786479194288e-05, | |
| "loss": 1.4065, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.31223110018438843, | |
| "grad_norm": 0.3247184455394745, | |
| "learning_rate": 8.337485396851155e-05, | |
| "loss": 1.4863, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.31346035648432696, | |
| "grad_norm": 0.26896461844444275, | |
| "learning_rate": 8.322127733662897e-05, | |
| "loss": 1.1373, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.3146896127842655, | |
| "grad_norm": 0.29333245754241943, | |
| "learning_rate": 8.306713749990072e-05, | |
| "loss": 1.0615, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.3159188690842041, | |
| "grad_norm": 0.2958793640136719, | |
| "learning_rate": 8.291243707148048e-05, | |
| "loss": 0.9392, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.3171481253841426, | |
| "grad_norm": 0.3320540487766266, | |
| "learning_rate": 8.275717867402575e-05, | |
| "loss": 1.2935, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.31837738168408114, | |
| "grad_norm": 0.3567339777946472, | |
| "learning_rate": 8.260136493965326e-05, | |
| "loss": 1.0954, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.31960663798401967, | |
| "grad_norm": 0.38393881916999817, | |
| "learning_rate": 8.244499850989452e-05, | |
| "loss": 1.045, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.3208358942839582, | |
| "grad_norm": 0.41993001103401184, | |
| "learning_rate": 8.228808203565095e-05, | |
| "loss": 1.2225, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.32206515058389673, | |
| "grad_norm": 0.6547941565513611, | |
| "learning_rate": 8.213061817714893e-05, | |
| "loss": 0.9286, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.32329440688383526, | |
| "grad_norm": 0.7117279767990112, | |
| "learning_rate": 8.197260960389474e-05, | |
| "loss": 0.5088, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.3245236631837738, | |
| "grad_norm": 0.7041743993759155, | |
| "learning_rate": 8.181405899462926e-05, | |
| "loss": 0.8899, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.3257529194837124, | |
| "grad_norm": 0.7142787575721741, | |
| "learning_rate": 8.16549690372826e-05, | |
| "loss": 0.7447, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.3269821757836509, | |
| "grad_norm": 0.8879908323287964, | |
| "learning_rate": 8.14953424289285e-05, | |
| "loss": 1.2607, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.32821143208358944, | |
| "grad_norm": 0.9387282133102417, | |
| "learning_rate": 8.133518187573862e-05, | |
| "loss": 1.1611, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.32944068838352797, | |
| "grad_norm": 1.4039078950881958, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 0.9947, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.3306699446834665, | |
| "grad_norm": 3.3686740398406982, | |
| "learning_rate": 8.101326980475237e-05, | |
| "loss": 1.0783, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.33189920098340503, | |
| "grad_norm": 2.8384785652160645, | |
| "learning_rate": 8.085152374437525e-05, | |
| "loss": 0.9008, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.33312845728334356, | |
| "grad_norm": 2.453441619873047, | |
| "learning_rate": 8.06892546539083e-05, | |
| "loss": 0.5504, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.3343577135832821, | |
| "grad_norm": 2.592667579650879, | |
| "learning_rate": 8.052646528432158e-05, | |
| "loss": 0.7489, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.3355869698832207, | |
| "grad_norm": 1.9753395318984985, | |
| "learning_rate": 8.036315839540545e-05, | |
| "loss": 0.9747, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.3368162261831592, | |
| "grad_norm": 3.042698860168457, | |
| "learning_rate": 8.019933675572389e-05, | |
| "loss": 1.6841, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.33804548248309774, | |
| "grad_norm": 2.4343316555023193, | |
| "learning_rate": 8.00350031425675e-05, | |
| "loss": 0.869, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.33927473878303627, | |
| "grad_norm": 0.2026144415140152, | |
| "learning_rate": 7.98701603419064e-05, | |
| "loss": 0.8867, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.3405039950829748, | |
| "grad_norm": 0.24370141327381134, | |
| "learning_rate": 7.970481114834312e-05, | |
| "loss": 1.3135, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.34173325138291333, | |
| "grad_norm": 0.22894087433815002, | |
| "learning_rate": 7.953895836506508e-05, | |
| "loss": 1.0986, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.34296250768285186, | |
| "grad_norm": 0.2533970773220062, | |
| "learning_rate": 7.937260480379712e-05, | |
| "loss": 1.1821, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.3441917639827904, | |
| "grad_norm": 0.25789350271224976, | |
| "learning_rate": 7.920575328475385e-05, | |
| "loss": 1.1414, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.345421020282729, | |
| "grad_norm": 0.28820541501045227, | |
| "learning_rate": 7.903840663659186e-05, | |
| "loss": 1.3332, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.3466502765826675, | |
| "grad_norm": 0.28611505031585693, | |
| "learning_rate": 7.887056769636165e-05, | |
| "loss": 1.0901, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.34787953288260604, | |
| "grad_norm": 0.28022873401641846, | |
| "learning_rate": 7.870223930945972e-05, | |
| "loss": 0.8461, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.34910878918254457, | |
| "grad_norm": 0.3246136009693146, | |
| "learning_rate": 7.853342432958013e-05, | |
| "loss": 0.9325, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.3503380454824831, | |
| "grad_norm": 0.3149406611919403, | |
| "learning_rate": 7.836412561866629e-05, | |
| "loss": 1.013, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.3515673017824216, | |
| "grad_norm": 0.3745490610599518, | |
| "learning_rate": 7.819434604686228e-05, | |
| "loss": 1.2624, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.35279655808236016, | |
| "grad_norm": 0.4822925329208374, | |
| "learning_rate": 7.802408849246442e-05, | |
| "loss": 1.2424, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.3540258143822987, | |
| "grad_norm": 0.6210641264915466, | |
| "learning_rate": 7.785335584187219e-05, | |
| "loss": 1.2527, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.3552550706822373, | |
| "grad_norm": 0.6488444805145264, | |
| "learning_rate": 7.768215098953952e-05, | |
| "loss": 0.7986, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.3564843269821758, | |
| "grad_norm": 0.760388195514679, | |
| "learning_rate": 7.751047683792561e-05, | |
| "loss": 1.0136, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.35771358328211433, | |
| "grad_norm": 0.7666548490524292, | |
| "learning_rate": 7.73383362974458e-05, | |
| "loss": 0.8205, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.35894283958205286, | |
| "grad_norm": 0.7492078542709351, | |
| "learning_rate": 7.71657322864221e-05, | |
| "loss": 0.9254, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.3601720958819914, | |
| "grad_norm": 0.9061193466186523, | |
| "learning_rate": 7.699266773103389e-05, | |
| "loss": 1.3013, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.3614013521819299, | |
| "grad_norm": 2.1404013633728027, | |
| "learning_rate": 7.681914556526817e-05, | |
| "loss": 1.5957, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.36263060848186845, | |
| "grad_norm": 2.647864580154419, | |
| "learning_rate": 7.664516873086987e-05, | |
| "loss": 1.1658, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.363859864781807, | |
| "grad_norm": 3.0906460285186768, | |
| "learning_rate": 7.647074017729202e-05, | |
| "loss": 1.1344, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.36508912108174557, | |
| "grad_norm": 2.2348814010620117, | |
| "learning_rate": 7.629586286164565e-05, | |
| "loss": 0.8813, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.3663183773816841, | |
| "grad_norm": 2.937446117401123, | |
| "learning_rate": 7.612053974864976e-05, | |
| "loss": 1.0414, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.36754763368162263, | |
| "grad_norm": 2.5343546867370605, | |
| "learning_rate": 7.594477381058098e-05, | |
| "loss": 1.1847, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.36877688998156116, | |
| "grad_norm": 2.8971638679504395, | |
| "learning_rate": 7.576856802722325e-05, | |
| "loss": 0.9029, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3700061462814997, | |
| "grad_norm": 0.1982557773590088, | |
| "learning_rate": 7.559192538581722e-05, | |
| "loss": 0.9314, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.3712354025814382, | |
| "grad_norm": 0.24721381068229675, | |
| "learning_rate": 7.541484888100974e-05, | |
| "loss": 1.2432, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.37246465888137675, | |
| "grad_norm": 0.24999506771564484, | |
| "learning_rate": 7.523734151480289e-05, | |
| "loss": 1.285, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.3736939151813153, | |
| "grad_norm": 0.267764151096344, | |
| "learning_rate": 7.505940629650326e-05, | |
| "loss": 1.198, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.3749231714812538, | |
| "grad_norm": 0.26003679633140564, | |
| "learning_rate": 7.488104624267091e-05, | |
| "loss": 1.2001, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.3761524277811924, | |
| "grad_norm": 0.28197526931762695, | |
| "learning_rate": 7.470226437706813e-05, | |
| "loss": 1.1687, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.37738168408113093, | |
| "grad_norm": 0.29367661476135254, | |
| "learning_rate": 7.452306373060829e-05, | |
| "loss": 1.211, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.37861094038106946, | |
| "grad_norm": 0.2982727885246277, | |
| "learning_rate": 7.434344734130437e-05, | |
| "loss": 1.151, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.379840196681008, | |
| "grad_norm": 0.3283758759498596, | |
| "learning_rate": 7.416341825421754e-05, | |
| "loss": 0.9875, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.3810694529809465, | |
| "grad_norm": 0.32420334219932556, | |
| "learning_rate": 7.398297952140544e-05, | |
| "loss": 1.0796, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.38229870928088505, | |
| "grad_norm": 0.4046980142593384, | |
| "learning_rate": 7.380213420187055e-05, | |
| "loss": 1.1158, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.3835279655808236, | |
| "grad_norm": 0.391736775636673, | |
| "learning_rate": 7.36208853615082e-05, | |
| "loss": 1.1682, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.3847572218807621, | |
| "grad_norm": 0.6027556657791138, | |
| "learning_rate": 7.343923607305471e-05, | |
| "loss": 1.0696, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.3859864781807007, | |
| "grad_norm": 0.6483603119850159, | |
| "learning_rate": 7.325718941603527e-05, | |
| "loss": 0.7843, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.38721573448063923, | |
| "grad_norm": 0.6711483001708984, | |
| "learning_rate": 7.307474847671168e-05, | |
| "loss": 0.7247, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.38844499078057776, | |
| "grad_norm": 0.7372632026672363, | |
| "learning_rate": 7.289191634803003e-05, | |
| "loss": 1.0535, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.3896742470805163, | |
| "grad_norm": 0.7427420020103455, | |
| "learning_rate": 7.270869612956835e-05, | |
| "loss": 1.0563, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.3909035033804548, | |
| "grad_norm": 2.6449501514434814, | |
| "learning_rate": 7.252509092748401e-05, | |
| "loss": 1.3099, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.39213275968039335, | |
| "grad_norm": 3.1938464641571045, | |
| "learning_rate": 7.234110385446103e-05, | |
| "loss": 1.1728, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.3933620159803319, | |
| "grad_norm": 2.584103584289551, | |
| "learning_rate": 7.215673802965734e-05, | |
| "loss": 0.792, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.3945912722802704, | |
| "grad_norm": 2.358025074005127, | |
| "learning_rate": 7.197199657865195e-05, | |
| "loss": 1.0462, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.395820528580209, | |
| "grad_norm": 2.9621617794036865, | |
| "learning_rate": 7.178688263339184e-05, | |
| "loss": 1.4222, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.3970497848801475, | |
| "grad_norm": 2.5362660884857178, | |
| "learning_rate": 7.160139933213898e-05, | |
| "loss": 1.1527, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.39827904118008606, | |
| "grad_norm": 2.4901375770568848, | |
| "learning_rate": 7.141554981941709e-05, | |
| "loss": 1.1712, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.3995082974800246, | |
| "grad_norm": 2.9214236736297607, | |
| "learning_rate": 7.12293372459583e-05, | |
| "loss": 1.1977, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.4007375537799631, | |
| "grad_norm": 0.24753543734550476, | |
| "learning_rate": 7.104276476864974e-05, | |
| "loss": 1.2176, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.40196681007990165, | |
| "grad_norm": 0.25986090302467346, | |
| "learning_rate": 7.085583555048008e-05, | |
| "loss": 1.2854, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.4031960663798402, | |
| "grad_norm": 0.2640175521373749, | |
| "learning_rate": 7.066855276048587e-05, | |
| "loss": 1.2204, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.4044253226797787, | |
| "grad_norm": 0.2603614330291748, | |
| "learning_rate": 7.048091957369776e-05, | |
| "loss": 1.2621, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.4056545789797173, | |
| "grad_norm": 0.2921195924282074, | |
| "learning_rate": 7.029293917108678e-05, | |
| "loss": 1.281, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.4068838352796558, | |
| "grad_norm": 0.2984941899776459, | |
| "learning_rate": 7.010461473951033e-05, | |
| "loss": 1.071, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.40811309157959436, | |
| "grad_norm": 0.31219175457954407, | |
| "learning_rate": 6.991594947165818e-05, | |
| "loss": 1.3161, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.4093423478795329, | |
| "grad_norm": 0.31329602003097534, | |
| "learning_rate": 6.972694656599834e-05, | |
| "loss": 0.9854, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.4105716041794714, | |
| "grad_norm": 0.3356671929359436, | |
| "learning_rate": 6.953760922672286e-05, | |
| "loss": 1.02, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.41180086047940995, | |
| "grad_norm": 0.3843994438648224, | |
| "learning_rate": 6.934794066369348e-05, | |
| "loss": 1.2173, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.4130301167793485, | |
| "grad_norm": 0.45338544249534607, | |
| "learning_rate": 6.915794409238718e-05, | |
| "loss": 1.3614, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.414259373079287, | |
| "grad_norm": 0.4857298731803894, | |
| "learning_rate": 6.896762273384178e-05, | |
| "loss": 1.0175, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.4154886293792256, | |
| "grad_norm": 0.6512896418571472, | |
| "learning_rate": 6.877697981460125e-05, | |
| "loss": 0.6555, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.4167178856791641, | |
| "grad_norm": 0.6744720935821533, | |
| "learning_rate": 6.858601856666094e-05, | |
| "loss": 0.6057, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.41794714197910265, | |
| "grad_norm": 0.6527014374732971, | |
| "learning_rate": 6.839474222741299e-05, | |
| "loss": 0.9116, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.4191763982790412, | |
| "grad_norm": 0.6935631036758423, | |
| "learning_rate": 6.820315403959123e-05, | |
| "loss": 0.9876, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.4204056545789797, | |
| "grad_norm": 0.6856899261474609, | |
| "learning_rate": 6.801125725121636e-05, | |
| "loss": 0.9591, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.42163491087891825, | |
| "grad_norm": 1.2577812671661377, | |
| "learning_rate": 6.781905511554079e-05, | |
| "loss": 1.3174, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.4228641671788568, | |
| "grad_norm": 2.421950578689575, | |
| "learning_rate": 6.762655089099353e-05, | |
| "loss": 1.6442, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.4240934234787953, | |
| "grad_norm": 2.6432454586029053, | |
| "learning_rate": 6.743374784112501e-05, | |
| "loss": 1.0468, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.4253226797787339, | |
| "grad_norm": 2.7061827182769775, | |
| "learning_rate": 6.724064923455155e-05, | |
| "loss": 1.1526, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.4265519360786724, | |
| "grad_norm": 2.466057777404785, | |
| "learning_rate": 6.704725834490024e-05, | |
| "loss": 1.1463, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.42778119237861095, | |
| "grad_norm": 2.753512144088745, | |
| "learning_rate": 6.685357845075315e-05, | |
| "loss": 0.9492, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.4290104486785495, | |
| "grad_norm": 2.76118803024292, | |
| "learning_rate": 6.665961283559197e-05, | |
| "loss": 0.8543, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.430239704978488, | |
| "grad_norm": 2.295574426651001, | |
| "learning_rate": 6.646536478774222e-05, | |
| "loss": 0.9564, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.43146896127842654, | |
| "grad_norm": 0.22731368243694305, | |
| "learning_rate": 6.627083760031754e-05, | |
| "loss": 0.9719, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.4326982175783651, | |
| "grad_norm": 0.20097078382968903, | |
| "learning_rate": 6.60760345711639e-05, | |
| "loss": 1.0094, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.4339274738783036, | |
| "grad_norm": 0.23321934044361115, | |
| "learning_rate": 6.58809590028036e-05, | |
| "loss": 1.101, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.4351567301782422, | |
| "grad_norm": 0.27995625138282776, | |
| "learning_rate": 6.568561420237935e-05, | |
| "loss": 1.3545, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.4363859864781807, | |
| "grad_norm": 0.259082168340683, | |
| "learning_rate": 6.54900034815982e-05, | |
| "loss": 1.1598, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.43761524277811925, | |
| "grad_norm": 0.2688703238964081, | |
| "learning_rate": 6.52941301566754e-05, | |
| "loss": 1.1141, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.4388444990780578, | |
| "grad_norm": 0.34018442034721375, | |
| "learning_rate": 6.50979975482781e-05, | |
| "loss": 1.2811, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.4400737553779963, | |
| "grad_norm": 0.2925175130367279, | |
| "learning_rate": 6.490160898146918e-05, | |
| "loss": 0.9025, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.44130301167793484, | |
| "grad_norm": 0.30208972096443176, | |
| "learning_rate": 6.470496778565082e-05, | |
| "loss": 1.0301, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.4425322679778734, | |
| "grad_norm": 0.3110770285129547, | |
| "learning_rate": 6.4508077294508e-05, | |
| "loss": 1.0911, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.4437615242778119, | |
| "grad_norm": 0.426252543926239, | |
| "learning_rate": 6.431094084595209e-05, | |
| "loss": 1.1214, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.44499078057775043, | |
| "grad_norm": 0.4019356966018677, | |
| "learning_rate": 6.411356178206419e-05, | |
| "loss": 1.3063, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.446220036877689, | |
| "grad_norm": 0.4622703194618225, | |
| "learning_rate": 6.391594344903848e-05, | |
| "loss": 1.1208, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.44744929317762755, | |
| "grad_norm": 0.5752270817756653, | |
| "learning_rate": 6.371808919712549e-05, | |
| "loss": 0.9653, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.4486785494775661, | |
| "grad_norm": 0.6309720277786255, | |
| "learning_rate": 6.35200023805754e-05, | |
| "loss": 0.5664, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.4499078057775046, | |
| "grad_norm": 0.612684965133667, | |
| "learning_rate": 6.332168635758097e-05, | |
| "loss": 1.0443, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.45113706207744314, | |
| "grad_norm": 0.6797056794166565, | |
| "learning_rate": 6.31231444902208e-05, | |
| "loss": 0.8389, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.45236631837738167, | |
| "grad_norm": 1.222960352897644, | |
| "learning_rate": 6.292438014440227e-05, | |
| "loss": 1.4688, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.4535955746773202, | |
| "grad_norm": 2.9443516731262207, | |
| "learning_rate": 6.272539668980441e-05, | |
| "loss": 1.0079, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.45482483097725873, | |
| "grad_norm": 3.0168612003326416, | |
| "learning_rate": 6.252619749982089e-05, | |
| "loss": 0.9232, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.4560540872771973, | |
| "grad_norm": 1.9470983743667603, | |
| "learning_rate": 6.232678595150275e-05, | |
| "loss": 0.8126, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.45728334357713585, | |
| "grad_norm": 2.4769911766052246, | |
| "learning_rate": 6.212716542550112e-05, | |
| "loss": 0.7786, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.4585125998770744, | |
| "grad_norm": 2.849158525466919, | |
| "learning_rate": 6.192733930601005e-05, | |
| "loss": 1.1914, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.4597418561770129, | |
| "grad_norm": 2.6154119968414307, | |
| "learning_rate": 6.172731098070899e-05, | |
| "loss": 0.9171, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.46097111247695144, | |
| "grad_norm": 3.5901479721069336, | |
| "learning_rate": 6.152708384070541e-05, | |
| "loss": 1.1269, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.46220036877688997, | |
| "grad_norm": 0.23536159098148346, | |
| "learning_rate": 6.132666128047732e-05, | |
| "loss": 0.8768, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.4634296250768285, | |
| "grad_norm": 0.24834086000919342, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 1.0644, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.46465888137676703, | |
| "grad_norm": 0.3041445016860962, | |
| "learning_rate": 6.0925243493767016e-05, | |
| "loss": 1.2779, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.4658881376767056, | |
| "grad_norm": 0.3158765137195587, | |
| "learning_rate": 6.0724255072575275e-05, | |
| "loss": 1.352, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.46711739397664415, | |
| "grad_norm": 0.2845201790332794, | |
| "learning_rate": 6.0523084841624635e-05, | |
| "loss": 1.2567, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.4683466502765827, | |
| "grad_norm": 0.2909673750400543, | |
| "learning_rate": 6.0321736211381464e-05, | |
| "loss": 1.1735, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.4695759065765212, | |
| "grad_norm": 0.2946690022945404, | |
| "learning_rate": 6.0120212595336545e-05, | |
| "loss": 1.1514, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.47080516287645974, | |
| "grad_norm": 0.302846223115921, | |
| "learning_rate": 5.9918517409947215e-05, | |
| "loss": 1.0621, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.47203441917639827, | |
| "grad_norm": 0.3197241425514221, | |
| "learning_rate": 5.971665407457948e-05, | |
| "loss": 1.0299, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.4732636754763368, | |
| "grad_norm": 0.342777281999588, | |
| "learning_rate": 5.951462601144998e-05, | |
| "loss": 1.0858, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.47449293177627533, | |
| "grad_norm": 0.3554008901119232, | |
| "learning_rate": 5.931243664556803e-05, | |
| "loss": 1.1441, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.4757221880762139, | |
| "grad_norm": 0.36057665944099426, | |
| "learning_rate": 5.9110089404677524e-05, | |
| "loss": 1.1836, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.47695144437615244, | |
| "grad_norm": 0.5004509091377258, | |
| "learning_rate": 5.890758771919884e-05, | |
| "loss": 1.4109, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.478180700676091, | |
| "grad_norm": 0.5211744904518127, | |
| "learning_rate": 5.8704935022170684e-05, | |
| "loss": 1.0097, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.4794099569760295, | |
| "grad_norm": 0.7474620938301086, | |
| "learning_rate": 5.8502134749191816e-05, | |
| "loss": 0.8777, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.48063921327596804, | |
| "grad_norm": 0.7044636011123657, | |
| "learning_rate": 5.8299190338362996e-05, | |
| "loss": 0.9007, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.48186846957590657, | |
| "grad_norm": 0.6484948992729187, | |
| "learning_rate": 5.8096105230228435e-05, | |
| "loss": 0.8261, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.4830977258758451, | |
| "grad_norm": 0.672816812992096, | |
| "learning_rate": 5.78928828677177e-05, | |
| "loss": 1.0531, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.4843269821757836, | |
| "grad_norm": 1.1637803316116333, | |
| "learning_rate": 5.768952669608724e-05, | |
| "loss": 1.1586, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.4855562384757222, | |
| "grad_norm": 3.1862003803253174, | |
| "learning_rate": 5.748604016286192e-05, | |
| "loss": 1.6232, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.48678549477566074, | |
| "grad_norm": 3.3833253383636475, | |
| "learning_rate": 5.728242671777672e-05, | |
| "loss": 1.0918, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.4880147510755993, | |
| "grad_norm": 3.116319417953491, | |
| "learning_rate": 5.707868981271815e-05, | |
| "loss": 0.8615, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.4892440073755378, | |
| "grad_norm": 2.5967965126037598, | |
| "learning_rate": 5.687483290166573e-05, | |
| "loss": 0.8579, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.49047326367547633, | |
| "grad_norm": 3.7683048248291016, | |
| "learning_rate": 5.6670859440633486e-05, | |
| "loss": 1.0777, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.49170251997541486, | |
| "grad_norm": 3.182317018508911, | |
| "learning_rate": 5.646677288761132e-05, | |
| "loss": 0.8932, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.4929317762753534, | |
| "grad_norm": 0.19372360408306122, | |
| "learning_rate": 5.6262576702506406e-05, | |
| "loss": 0.8516, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.4941610325752919, | |
| "grad_norm": 0.223338320851326, | |
| "learning_rate": 5.6058274347084504e-05, | |
| "loss": 1.1287, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.4953902888752305, | |
| "grad_norm": 0.23367400467395782, | |
| "learning_rate": 5.585386928491134e-05, | |
| "loss": 1.1128, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.49661954517516904, | |
| "grad_norm": 0.2717371881008148, | |
| "learning_rate": 5.5649364981293786e-05, | |
| "loss": 1.2813, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.49784880147510757, | |
| "grad_norm": 0.25909724831581116, | |
| "learning_rate": 5.54447649032212e-05, | |
| "loss": 1.2149, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.4990780577750461, | |
| "grad_norm": 0.25411197543144226, | |
| "learning_rate": 5.5240072519306606e-05, | |
| "loss": 1.0679, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.5003073140749846, | |
| "grad_norm": 0.2962128520011902, | |
| "learning_rate": 5.503529129972792e-05, | |
| "loss": 1.1156, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.5015365703749232, | |
| "grad_norm": 0.29252949357032776, | |
| "learning_rate": 5.483042471616908e-05, | |
| "loss": 1.125, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.5015365703749232, | |
| "eval_loss": 0.9695894122123718, | |
| "eval_runtime": 65.3254, | |
| "eval_samples_per_second": 10.486, | |
| "eval_steps_per_second": 5.251, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.5027658266748617, | |
| "grad_norm": 0.31012120842933655, | |
| "learning_rate": 5.4625476241761196e-05, | |
| "loss": 1.1491, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.5039950829748002, | |
| "grad_norm": 0.3182609975337982, | |
| "learning_rate": 5.442044935102375e-05, | |
| "loss": 1.0786, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.5052243392747388, | |
| "grad_norm": 0.35413217544555664, | |
| "learning_rate": 5.421534751980556e-05, | |
| "loss": 1.2194, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.5064535955746773, | |
| "grad_norm": 0.39505264163017273, | |
| "learning_rate": 5.401017422522594e-05, | |
| "loss": 1.0296, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.5076828518746158, | |
| "grad_norm": 0.4267808496952057, | |
| "learning_rate": 5.380493294561573e-05, | |
| "loss": 1.0906, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.5089121081745543, | |
| "grad_norm": 0.5495923161506653, | |
| "learning_rate": 5.359962716045835e-05, | |
| "loss": 1.2171, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.510141364474493, | |
| "grad_norm": 0.6385165452957153, | |
| "learning_rate": 5.3394260350330796e-05, | |
| "loss": 0.8179, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.5113706207744315, | |
| "grad_norm": 0.7211401462554932, | |
| "learning_rate": 5.318883599684456e-05, | |
| "loss": 0.7624, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.51259987707437, | |
| "grad_norm": 0.800375759601593, | |
| "learning_rate": 5.298335758258678e-05, | |
| "loss": 0.9597, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.5138291333743086, | |
| "grad_norm": 0.9896557927131653, | |
| "learning_rate": 5.2777828591060984e-05, | |
| "loss": 1.12, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.5150583896742471, | |
| "grad_norm": 2.869112968444824, | |
| "learning_rate": 5.257225250662823e-05, | |
| "loss": 1.2992, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.5162876459741856, | |
| "grad_norm": 3.394543409347534, | |
| "learning_rate": 5.236663281444791e-05, | |
| "loss": 1.298, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.5175169022741242, | |
| "grad_norm": 3.3043065071105957, | |
| "learning_rate": 5.21609730004187e-05, | |
| "loss": 1.1924, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.5187461585740627, | |
| "grad_norm": 2.603189706802368, | |
| "learning_rate": 5.1955276551119495e-05, | |
| "loss": 1.2155, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.5199754148740012, | |
| "grad_norm": 3.018608808517456, | |
| "learning_rate": 5.174954695375023e-05, | |
| "loss": 1.2001, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.5212046711739398, | |
| "grad_norm": 1.8929078578948975, | |
| "learning_rate": 5.154378769607286e-05, | |
| "loss": 0.7124, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.5224339274738783, | |
| "grad_norm": 2.4811208248138428, | |
| "learning_rate": 5.1338002266352106e-05, | |
| "loss": 0.8491, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.5236631837738168, | |
| "grad_norm": 0.22651559114456177, | |
| "learning_rate": 5.113219415329645e-05, | |
| "loss": 0.9983, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.5248924400737554, | |
| "grad_norm": 0.2817627787590027, | |
| "learning_rate": 5.0926366845998904e-05, | |
| "loss": 1.1314, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.5261216963736939, | |
| "grad_norm": 0.27301520109176636, | |
| "learning_rate": 5.072052383387786e-05, | |
| "loss": 1.2722, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.5273509526736324, | |
| "grad_norm": 0.2770383059978485, | |
| "learning_rate": 5.0514668606618e-05, | |
| "loss": 1.1654, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.5285802089735709, | |
| "grad_norm": 0.3147628903388977, | |
| "learning_rate": 5.0308804654111056e-05, | |
| "loss": 1.0896, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.5298094652735095, | |
| "grad_norm": 0.305074542760849, | |
| "learning_rate": 5.01029354663967e-05, | |
| "loss": 1.1754, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.5310387215734481, | |
| "grad_norm": 0.3240547776222229, | |
| "learning_rate": 4.9897064533603315e-05, | |
| "loss": 1.2413, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.5322679778733866, | |
| "grad_norm": 0.34143832325935364, | |
| "learning_rate": 4.9691195345888956e-05, | |
| "loss": 1.1504, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.5334972341733252, | |
| "grad_norm": 0.3945367634296417, | |
| "learning_rate": 4.948533139338202e-05, | |
| "loss": 1.2914, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.5347264904732637, | |
| "grad_norm": 0.371367871761322, | |
| "learning_rate": 4.927947616612215e-05, | |
| "loss": 1.3269, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.5359557467732022, | |
| "grad_norm": 0.4348413646221161, | |
| "learning_rate": 4.90736331540011e-05, | |
| "loss": 1.2061, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.5371850030731408, | |
| "grad_norm": 0.5112064480781555, | |
| "learning_rate": 4.886780584670356e-05, | |
| "loss": 1.2283, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.5384142593730793, | |
| "grad_norm": 0.6409791707992554, | |
| "learning_rate": 4.866199773364789e-05, | |
| "loss": 0.7663, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.5396435156730178, | |
| "grad_norm": 0.7076222896575928, | |
| "learning_rate": 4.845621230392716e-05, | |
| "loss": 0.8073, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.5408727719729564, | |
| "grad_norm": 0.6545782089233398, | |
| "learning_rate": 4.825045304624978e-05, | |
| "loss": 0.762, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.5421020282728949, | |
| "grad_norm": 0.6853161454200745, | |
| "learning_rate": 4.804472344888052e-05, | |
| "loss": 0.9905, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.5433312845728334, | |
| "grad_norm": 0.6758946776390076, | |
| "learning_rate": 4.7839026999581296e-05, | |
| "loss": 0.9353, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.544560540872772, | |
| "grad_norm": 1.1472845077514648, | |
| "learning_rate": 4.7633367185552095e-05, | |
| "loss": 0.9796, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.5457897971727105, | |
| "grad_norm": 2.543483018875122, | |
| "learning_rate": 4.742774749337179e-05, | |
| "loss": 1.5897, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.547019053472649, | |
| "grad_norm": 3.029282331466675, | |
| "learning_rate": 4.7222171408939034e-05, | |
| "loss": 1.2527, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.5482483097725875, | |
| "grad_norm": 3.1332318782806396, | |
| "learning_rate": 4.701664241741323e-05, | |
| "loss": 1.1093, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.5494775660725261, | |
| "grad_norm": 2.407163381576538, | |
| "learning_rate": 4.681116400315544e-05, | |
| "loss": 0.75, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.5507068223724647, | |
| "grad_norm": 2.622605085372925, | |
| "learning_rate": 4.6605739649669236e-05, | |
| "loss": 0.8924, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.5519360786724032, | |
| "grad_norm": 2.5338971614837646, | |
| "learning_rate": 4.640037283954165e-05, | |
| "loss": 0.681, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.5531653349723418, | |
| "grad_norm": 2.4635331630706787, | |
| "learning_rate": 4.619506705438428e-05, | |
| "loss": 0.7866, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.5543945912722803, | |
| "grad_norm": 0.18258269131183624, | |
| "learning_rate": 4.598982577477408e-05, | |
| "loss": 0.8069, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.5556238475722188, | |
| "grad_norm": 0.23453976213932037, | |
| "learning_rate": 4.578465248019445e-05, | |
| "loss": 1.1606, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.5568531038721574, | |
| "grad_norm": 0.25334998965263367, | |
| "learning_rate": 4.557955064897626e-05, | |
| "loss": 1.2323, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.5580823601720959, | |
| "grad_norm": 0.2757891118526459, | |
| "learning_rate": 4.537452375823881e-05, | |
| "loss": 1.4173, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.5593116164720344, | |
| "grad_norm": 0.2755081355571747, | |
| "learning_rate": 4.5169575283830936e-05, | |
| "loss": 1.2255, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.560540872771973, | |
| "grad_norm": 0.29173940420150757, | |
| "learning_rate": 4.496470870027209e-05, | |
| "loss": 1.4528, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.5617701290719115, | |
| "grad_norm": 0.27483510971069336, | |
| "learning_rate": 4.475992748069339e-05, | |
| "loss": 0.7703, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.56299938537185, | |
| "grad_norm": 0.3180837333202362, | |
| "learning_rate": 4.455523509677882e-05, | |
| "loss": 1.025, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.5642286416717885, | |
| "grad_norm": 0.3103027939796448, | |
| "learning_rate": 4.435063501870622e-05, | |
| "loss": 1.0503, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.5654578979717271, | |
| "grad_norm": 0.3448195457458496, | |
| "learning_rate": 4.4146130715088676e-05, | |
| "loss": 1.0366, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.5666871542716656, | |
| "grad_norm": 0.3892152011394501, | |
| "learning_rate": 4.3941725652915494e-05, | |
| "loss": 1.1834, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.5679164105716041, | |
| "grad_norm": 0.4770027697086334, | |
| "learning_rate": 4.373742329749362e-05, | |
| "loss": 1.4524, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.5691456668715427, | |
| "grad_norm": 0.5856226682662964, | |
| "learning_rate": 4.3533227112388694e-05, | |
| "loss": 1.2538, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.5703749231714813, | |
| "grad_norm": 0.6203077435493469, | |
| "learning_rate": 4.332914055936653e-05, | |
| "loss": 0.6276, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.5716041794714198, | |
| "grad_norm": 0.6294959187507629, | |
| "learning_rate": 4.3125167098334286e-05, | |
| "loss": 0.642, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.5728334357713584, | |
| "grad_norm": 0.7135624885559082, | |
| "learning_rate": 4.2921310187281864e-05, | |
| "loss": 1.1486, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.5740626920712969, | |
| "grad_norm": 0.7505180835723877, | |
| "learning_rate": 4.27175732822233e-05, | |
| "loss": 0.9109, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.5752919483712354, | |
| "grad_norm": 1.0719517469406128, | |
| "learning_rate": 4.251395983713809e-05, | |
| "loss": 1.2909, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.576521204671174, | |
| "grad_norm": 2.5531132221221924, | |
| "learning_rate": 4.231047330391278e-05, | |
| "loss": 1.1723, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.5777504609711125, | |
| "grad_norm": 3.104837656021118, | |
| "learning_rate": 4.21071171322823e-05, | |
| "loss": 1.3296, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.578979717271051, | |
| "grad_norm": 3.239119052886963, | |
| "learning_rate": 4.190389476977156e-05, | |
| "loss": 0.8768, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.5802089735709896, | |
| "grad_norm": 2.681823968887329, | |
| "learning_rate": 4.170080966163702e-05, | |
| "loss": 0.8809, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.5814382298709281, | |
| "grad_norm": 2.557533025741577, | |
| "learning_rate": 4.149786525080819e-05, | |
| "loss": 0.8249, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.5826674861708666, | |
| "grad_norm": 2.4591336250305176, | |
| "learning_rate": 4.1295064977829334e-05, | |
| "loss": 0.7384, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.5838967424708051, | |
| "grad_norm": 2.414395809173584, | |
| "learning_rate": 4.109241228080115e-05, | |
| "loss": 0.7307, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.5851259987707437, | |
| "grad_norm": 0.2034798115491867, | |
| "learning_rate": 4.088991059532248e-05, | |
| "loss": 0.8326, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.5863552550706822, | |
| "grad_norm": 0.24682076275348663, | |
| "learning_rate": 4.0687563354431984e-05, | |
| "loss": 1.3382, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.5875845113706207, | |
| "grad_norm": 0.2430254966020584, | |
| "learning_rate": 4.048537398855003e-05, | |
| "loss": 1.2718, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.5888137676705593, | |
| "grad_norm": 0.27249017357826233, | |
| "learning_rate": 4.028334592542054e-05, | |
| "loss": 1.2328, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.5900430239704979, | |
| "grad_norm": 0.28996631503105164, | |
| "learning_rate": 4.008148259005279e-05, | |
| "loss": 1.3963, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.5912722802704364, | |
| "grad_norm": 0.28658372163772583, | |
| "learning_rate": 3.9879787404663474e-05, | |
| "loss": 1.2326, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.592501536570375, | |
| "grad_norm": 0.3246002495288849, | |
| "learning_rate": 3.967826378861854e-05, | |
| "loss": 1.1071, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.5937307928703135, | |
| "grad_norm": 0.33610400557518005, | |
| "learning_rate": 3.947691515837537e-05, | |
| "loss": 1.1798, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.594960049170252, | |
| "grad_norm": 0.35427945852279663, | |
| "learning_rate": 3.927574492742473e-05, | |
| "loss": 1.3277, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.5961893054701906, | |
| "grad_norm": 0.3942146599292755, | |
| "learning_rate": 3.907475650623299e-05, | |
| "loss": 1.0499, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.5974185617701291, | |
| "grad_norm": 0.36504921317100525, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 1.1876, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.5986478180700676, | |
| "grad_norm": 0.45135965943336487, | |
| "learning_rate": 3.867333871952269e-05, | |
| "loss": 1.1775, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.5998770743700061, | |
| "grad_norm": 0.6297984719276428, | |
| "learning_rate": 3.84729161592946e-05, | |
| "loss": 0.9756, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.6011063306699447, | |
| "grad_norm": 0.6099340915679932, | |
| "learning_rate": 3.827268901929102e-05, | |
| "loss": 0.7669, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.6023355869698832, | |
| "grad_norm": 0.7176752686500549, | |
| "learning_rate": 3.8072660693989967e-05, | |
| "loss": 0.5387, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.6035648432698217, | |
| "grad_norm": 0.6262489557266235, | |
| "learning_rate": 3.78728345744989e-05, | |
| "loss": 0.9306, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.6047940995697603, | |
| "grad_norm": 1.0194954872131348, | |
| "learning_rate": 3.767321404849727e-05, | |
| "loss": 1.1677, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.6060233558696988, | |
| "grad_norm": 1.6236090660095215, | |
| "learning_rate": 3.7473802500179114e-05, | |
| "loss": 1.0458, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.6072526121696373, | |
| "grad_norm": 2.5714359283447266, | |
| "learning_rate": 3.727460331019559e-05, | |
| "loss": 1.2519, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.6084818684695759, | |
| "grad_norm": 2.4615423679351807, | |
| "learning_rate": 3.7075619855597744e-05, | |
| "loss": 0.5842, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.6097111247695144, | |
| "grad_norm": 3.4131109714508057, | |
| "learning_rate": 3.6876855509779206e-05, | |
| "loss": 1.3071, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.610940381069453, | |
| "grad_norm": 3.461987257003784, | |
| "learning_rate": 3.667831364241904e-05, | |
| "loss": 1.1426, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.6121696373693916, | |
| "grad_norm": 2.462989330291748, | |
| "learning_rate": 3.6479997619424605e-05, | |
| "loss": 0.8421, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.6133988936693301, | |
| "grad_norm": 2.423410654067993, | |
| "learning_rate": 3.628191080287451e-05, | |
| "loss": 0.7845, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.6146281499692686, | |
| "grad_norm": 2.5741212368011475, | |
| "learning_rate": 3.608405655096154e-05, | |
| "loss": 0.6916, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6158574062692072, | |
| "grad_norm": 0.2071269154548645, | |
| "learning_rate": 3.588643821793582e-05, | |
| "loss": 0.9648, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.6170866625691457, | |
| "grad_norm": 0.23868902027606964, | |
| "learning_rate": 3.5689059154047915e-05, | |
| "loss": 1.261, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.6183159188690842, | |
| "grad_norm": 0.2465473711490631, | |
| "learning_rate": 3.5491922705492e-05, | |
| "loss": 1.2999, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.6195451751690227, | |
| "grad_norm": 0.25372591614723206, | |
| "learning_rate": 3.5295032214349196e-05, | |
| "loss": 1.3633, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.6207744314689613, | |
| "grad_norm": 0.24923810362815857, | |
| "learning_rate": 3.5098391018530816e-05, | |
| "loss": 1.1408, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.6220036877688998, | |
| "grad_norm": 0.2594156563282013, | |
| "learning_rate": 3.4902002451721916e-05, | |
| "loss": 1.158, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.6232329440688383, | |
| "grad_norm": 0.28343746066093445, | |
| "learning_rate": 3.4705869843324614e-05, | |
| "loss": 1.1717, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.6244622003687769, | |
| "grad_norm": 0.29203376173973083, | |
| "learning_rate": 3.450999651840179e-05, | |
| "loss": 1.0693, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.6256914566687154, | |
| "grad_norm": 0.3236950933933258, | |
| "learning_rate": 3.431438579762066e-05, | |
| "loss": 1.0098, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.6269207129686539, | |
| "grad_norm": 0.3463621139526367, | |
| "learning_rate": 3.411904099719642e-05, | |
| "loss": 1.1123, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.6281499692685925, | |
| "grad_norm": 0.4081440269947052, | |
| "learning_rate": 3.3923965428836105e-05, | |
| "loss": 1.1303, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.629379225568531, | |
| "grad_norm": 0.38460928201675415, | |
| "learning_rate": 3.3729162399682456e-05, | |
| "loss": 1.0858, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.6306084818684696, | |
| "grad_norm": 0.474970281124115, | |
| "learning_rate": 3.35346352122578e-05, | |
| "loss": 1.1176, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.6318377381684082, | |
| "grad_norm": 0.7059311270713806, | |
| "learning_rate": 3.3340387164408046e-05, | |
| "loss": 1.1786, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.6330669944683467, | |
| "grad_norm": 0.7394426465034485, | |
| "learning_rate": 3.314642154924686e-05, | |
| "loss": 0.7836, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.6342962507682852, | |
| "grad_norm": 0.766044020652771, | |
| "learning_rate": 3.295274165509979e-05, | |
| "loss": 0.8919, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.6355255070682237, | |
| "grad_norm": 0.7920165061950684, | |
| "learning_rate": 3.275935076544845e-05, | |
| "loss": 1.0439, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.6367547633681623, | |
| "grad_norm": 1.224188208580017, | |
| "learning_rate": 3.256625215887502e-05, | |
| "loss": 1.3103, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.6379840196681008, | |
| "grad_norm": 2.0128793716430664, | |
| "learning_rate": 3.237344910900648e-05, | |
| "loss": 1.0804, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.6392132759680393, | |
| "grad_norm": 3.370530366897583, | |
| "learning_rate": 3.218094488445923e-05, | |
| "loss": 0.5978, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.6404425322679779, | |
| "grad_norm": 3.095421314239502, | |
| "learning_rate": 3.198874274878365e-05, | |
| "loss": 0.847, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.6416717885679164, | |
| "grad_norm": 3.4100868701934814, | |
| "learning_rate": 3.179684596040878e-05, | |
| "loss": 0.8489, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.6429010448678549, | |
| "grad_norm": 2.608668088912964, | |
| "learning_rate": 3.1605257772587035e-05, | |
| "loss": 0.8378, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.6441303011677935, | |
| "grad_norm": 2.300096273422241, | |
| "learning_rate": 3.141398143333907e-05, | |
| "loss": 0.7825, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.645359557467732, | |
| "grad_norm": 2.652704954147339, | |
| "learning_rate": 3.1223020185398765e-05, | |
| "loss": 0.6505, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.6465888137676705, | |
| "grad_norm": 0.19123272597789764, | |
| "learning_rate": 3.103237726615822e-05, | |
| "loss": 0.8317, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.647818070067609, | |
| "grad_norm": 0.22544066607952118, | |
| "learning_rate": 3.084205590761284e-05, | |
| "loss": 1.1174, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.6490473263675476, | |
| "grad_norm": 0.25559771060943604, | |
| "learning_rate": 3.065205933630655e-05, | |
| "loss": 1.258, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.6502765826674862, | |
| "grad_norm": 0.2453518509864807, | |
| "learning_rate": 3.0462390773277154e-05, | |
| "loss": 1.1197, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.6515058389674248, | |
| "grad_norm": 0.26897767186164856, | |
| "learning_rate": 3.0273053434001662e-05, | |
| "loss": 1.2231, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.6527350952673633, | |
| "grad_norm": 0.2956353425979614, | |
| "learning_rate": 3.0084050528341824e-05, | |
| "loss": 0.9809, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.6539643515673018, | |
| "grad_norm": 0.3043264150619507, | |
| "learning_rate": 2.989538526048968e-05, | |
| "loss": 1.2625, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.6551936078672403, | |
| "grad_norm": 0.3115137219429016, | |
| "learning_rate": 2.9707060828913225e-05, | |
| "loss": 1.1544, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.6564228641671789, | |
| "grad_norm": 0.29051673412323, | |
| "learning_rate": 2.9519080426302238e-05, | |
| "loss": 0.8664, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.6576521204671174, | |
| "grad_norm": 0.31704720854759216, | |
| "learning_rate": 2.933144723951414e-05, | |
| "loss": 0.8553, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.6588813767670559, | |
| "grad_norm": 0.3343563675880432, | |
| "learning_rate": 2.9144164449519917e-05, | |
| "loss": 1.042, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.6601106330669945, | |
| "grad_norm": 0.411081999540329, | |
| "learning_rate": 2.895723523135028e-05, | |
| "loss": 1.2587, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.661339889366933, | |
| "grad_norm": 0.4190918505191803, | |
| "learning_rate": 2.877066275404172e-05, | |
| "loss": 1.1546, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.6625691456668715, | |
| "grad_norm": 0.5606351494789124, | |
| "learning_rate": 2.8584450180582912e-05, | |
| "loss": 0.9889, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.6637984019668101, | |
| "grad_norm": 0.6339558959007263, | |
| "learning_rate": 2.839860066786103e-05, | |
| "loss": 0.5278, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.6650276582667486, | |
| "grad_norm": 0.6629421710968018, | |
| "learning_rate": 2.8213117366608188e-05, | |
| "loss": 0.752, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.6662569145666871, | |
| "grad_norm": 0.7566842436790466, | |
| "learning_rate": 2.802800342134807e-05, | |
| "loss": 1.0395, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.6674861708666256, | |
| "grad_norm": 1.169986605644226, | |
| "learning_rate": 2.784326197034266e-05, | |
| "loss": 1.1942, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.6687154271665642, | |
| "grad_norm": 2.78249192237854, | |
| "learning_rate": 2.7658896145538983e-05, | |
| "loss": 1.1694, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.6699446834665027, | |
| "grad_norm": 2.5933189392089844, | |
| "learning_rate": 2.7474909072515993e-05, | |
| "loss": 0.9325, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.6711739397664414, | |
| "grad_norm": 2.37711238861084, | |
| "learning_rate": 2.7291303870431662e-05, | |
| "loss": 0.9246, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.6724031960663799, | |
| "grad_norm": 3.3568241596221924, | |
| "learning_rate": 2.710808365197e-05, | |
| "loss": 0.9492, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.6736324523663184, | |
| "grad_norm": 2.975985527038574, | |
| "learning_rate": 2.6925251523288346e-05, | |
| "loss": 0.8018, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.6748617086662569, | |
| "grad_norm": 2.78823184967041, | |
| "learning_rate": 2.674281058396473e-05, | |
| "loss": 0.843, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.6760909649661955, | |
| "grad_norm": 4.179988384246826, | |
| "learning_rate": 2.6560763926945275e-05, | |
| "loss": 1.0288, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.677320221266134, | |
| "grad_norm": 0.21164661645889282, | |
| "learning_rate": 2.6379114638491807e-05, | |
| "loss": 0.9468, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.6785494775660725, | |
| "grad_norm": 0.2241944521665573, | |
| "learning_rate": 2.6197865798129462e-05, | |
| "loss": 1.1578, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.6797787338660111, | |
| "grad_norm": 0.25557973980903625, | |
| "learning_rate": 2.601702047859455e-05, | |
| "loss": 1.2432, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.6810079901659496, | |
| "grad_norm": 0.2733787000179291, | |
| "learning_rate": 2.5836581745782475e-05, | |
| "loss": 1.2624, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.6822372464658881, | |
| "grad_norm": 0.29645049571990967, | |
| "learning_rate": 2.5656552658695642e-05, | |
| "loss": 1.2544, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.6834665027658267, | |
| "grad_norm": 0.29211127758026123, | |
| "learning_rate": 2.5476936269391726e-05, | |
| "loss": 1.0548, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.6846957590657652, | |
| "grad_norm": 0.3013113737106323, | |
| "learning_rate": 2.5297735622931874e-05, | |
| "loss": 1.0941, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.6859250153657037, | |
| "grad_norm": 0.33897635340690613, | |
| "learning_rate": 2.5118953757329088e-05, | |
| "loss": 1.0855, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.6871542716656422, | |
| "grad_norm": 0.36230790615081787, | |
| "learning_rate": 2.494059370349673e-05, | |
| "loss": 1.0817, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.6883835279655808, | |
| "grad_norm": 0.36360612511634827, | |
| "learning_rate": 2.4762658485197123e-05, | |
| "loss": 1.0868, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.6896127842655193, | |
| "grad_norm": 0.45407038927078247, | |
| "learning_rate": 2.4585151118990286e-05, | |
| "loss": 1.1186, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.690842040565458, | |
| "grad_norm": 0.5940976142883301, | |
| "learning_rate": 2.4408074614182773e-05, | |
| "loss": 1.1448, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.6920712968653965, | |
| "grad_norm": 0.6550815105438232, | |
| "learning_rate": 2.4231431972776758e-05, | |
| "loss": 0.931, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.693300553165335, | |
| "grad_norm": 0.6394757032394409, | |
| "learning_rate": 2.4055226189419018e-05, | |
| "loss": 0.8543, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.6945298094652735, | |
| "grad_norm": 0.6903315186500549, | |
| "learning_rate": 2.3879460251350255e-05, | |
| "loss": 0.8152, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.6957590657652121, | |
| "grad_norm": 0.9499403238296509, | |
| "learning_rate": 2.3704137138354355e-05, | |
| "loss": 1.2532, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.6969883220651506, | |
| "grad_norm": 1.2395949363708496, | |
| "learning_rate": 2.3529259822708e-05, | |
| "loss": 1.3153, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.6982175783650891, | |
| "grad_norm": 2.5209338665008545, | |
| "learning_rate": 2.3354831269130133e-05, | |
| "loss": 1.4484, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.6994468346650277, | |
| "grad_norm": 3.072936773300171, | |
| "learning_rate": 2.318085443473185e-05, | |
| "loss": 0.9566, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.7006760909649662, | |
| "grad_norm": 2.40030574798584, | |
| "learning_rate": 2.300733226896612e-05, | |
| "loss": 0.6958, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.7019053472649047, | |
| "grad_norm": 3.1356310844421387, | |
| "learning_rate": 2.2834267713577904e-05, | |
| "loss": 1.261, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.7031346035648433, | |
| "grad_norm": 2.5506911277770996, | |
| "learning_rate": 2.2661663702554208e-05, | |
| "loss": 0.7489, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.7043638598647818, | |
| "grad_norm": 2.605987548828125, | |
| "learning_rate": 2.2489523162074393e-05, | |
| "loss": 0.8642, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.7055931161647203, | |
| "grad_norm": 2.3878705501556396, | |
| "learning_rate": 2.2317849010460507e-05, | |
| "loss": 0.559, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.7068223724646588, | |
| "grad_norm": 4.309600353240967, | |
| "learning_rate": 2.2146644158127827e-05, | |
| "loss": 1.3898, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.7080516287645974, | |
| "grad_norm": 0.2340036779642105, | |
| "learning_rate": 2.197591150753559e-05, | |
| "loss": 1.1237, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.7092808850645359, | |
| "grad_norm": 0.2404404878616333, | |
| "learning_rate": 2.1805653953137707e-05, | |
| "loss": 1.2235, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.7105101413644745, | |
| "grad_norm": 0.2546350657939911, | |
| "learning_rate": 2.1635874381333714e-05, | |
| "loss": 1.1931, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.7117393976644131, | |
| "grad_norm": 0.28081899881362915, | |
| "learning_rate": 2.1466575670419876e-05, | |
| "loss": 1.1851, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.7129686539643516, | |
| "grad_norm": 0.27532532811164856, | |
| "learning_rate": 2.1297760690540302e-05, | |
| "loss": 1.0733, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.7141979102642901, | |
| "grad_norm": 0.26407289505004883, | |
| "learning_rate": 2.1129432303638352e-05, | |
| "loss": 1.0863, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.7154271665642287, | |
| "grad_norm": 0.31387853622436523, | |
| "learning_rate": 2.0961593363408156e-05, | |
| "loss": 1.1808, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.7166564228641672, | |
| "grad_norm": 0.3309866487979889, | |
| "learning_rate": 2.079424671524616e-05, | |
| "loss": 1.0741, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.7178856791641057, | |
| "grad_norm": 0.300689697265625, | |
| "learning_rate": 2.0627395196202898e-05, | |
| "loss": 0.9483, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.7191149354640443, | |
| "grad_norm": 0.3478316366672516, | |
| "learning_rate": 2.046104163493493e-05, | |
| "loss": 1.144, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.7203441917639828, | |
| "grad_norm": 0.44858890771865845, | |
| "learning_rate": 2.0295188851656892e-05, | |
| "loss": 1.1384, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.7215734480639213, | |
| "grad_norm": 0.4991127550601959, | |
| "learning_rate": 2.0129839658093607e-05, | |
| "loss": 1.3527, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.7228027043638598, | |
| "grad_norm": 0.5499400496482849, | |
| "learning_rate": 1.996499685743254e-05, | |
| "loss": 1.2116, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.7240319606637984, | |
| "grad_norm": 0.6611432433128357, | |
| "learning_rate": 1.980066324427613e-05, | |
| "loss": 0.5055, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.7252612169637369, | |
| "grad_norm": 0.7008240222930908, | |
| "learning_rate": 1.9636841604594557e-05, | |
| "loss": 0.7072, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.7264904732636754, | |
| "grad_norm": 0.7243725061416626, | |
| "learning_rate": 1.9473534715678427e-05, | |
| "loss": 0.8031, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.727719729563614, | |
| "grad_norm": 0.8577756285667419, | |
| "learning_rate": 1.9310745346091714e-05, | |
| "loss": 1.1383, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.7289489858635525, | |
| "grad_norm": 1.2171465158462524, | |
| "learning_rate": 1.9148476255624764e-05, | |
| "loss": 0.9733, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.7301782421634911, | |
| "grad_norm": 3.537208318710327, | |
| "learning_rate": 1.898673019524764e-05, | |
| "loss": 1.1197, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.7314074984634297, | |
| "grad_norm": 2.420771360397339, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 0.7744, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.7326367547633682, | |
| "grad_norm": 3.1013336181640625, | |
| "learning_rate": 1.8664818124261374e-05, | |
| "loss": 0.7931, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.7338660110633067, | |
| "grad_norm": 3.5393359661102295, | |
| "learning_rate": 1.8504657571071515e-05, | |
| "loss": 1.0527, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.7350952673632453, | |
| "grad_norm": 2.782148838043213, | |
| "learning_rate": 1.8345030962717407e-05, | |
| "loss": 0.8034, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.7363245236631838, | |
| "grad_norm": 2.12009596824646, | |
| "learning_rate": 1.8185941005370745e-05, | |
| "loss": 0.4618, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.7375537799631223, | |
| "grad_norm": 3.6186935901641846, | |
| "learning_rate": 1.802739039610527e-05, | |
| "loss": 0.793, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.7387830362630609, | |
| "grad_norm": 0.22188998758792877, | |
| "learning_rate": 1.786938182285107e-05, | |
| "loss": 1.1901, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.7400122925629994, | |
| "grad_norm": 0.2288677990436554, | |
| "learning_rate": 1.7711917964349062e-05, | |
| "loss": 1.1755, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.7412415488629379, | |
| "grad_norm": 0.25471025705337524, | |
| "learning_rate": 1.7555001490105488e-05, | |
| "loss": 1.2616, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.7424708051628764, | |
| "grad_norm": 0.23840656876564026, | |
| "learning_rate": 1.7398635060346746e-05, | |
| "loss": 1.0882, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.743700061462815, | |
| "grad_norm": 0.2534352242946625, | |
| "learning_rate": 1.7242821325974258e-05, | |
| "loss": 1.1328, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.7449293177627535, | |
| "grad_norm": 0.279732346534729, | |
| "learning_rate": 1.7087562928519514e-05, | |
| "loss": 1.1703, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.746158574062692, | |
| "grad_norm": 0.3084731698036194, | |
| "learning_rate": 1.69328625000993e-05, | |
| "loss": 1.1646, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.7473878303626306, | |
| "grad_norm": 0.3052552044391632, | |
| "learning_rate": 1.6778722663371053e-05, | |
| "loss": 1.1481, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.7486170866625691, | |
| "grad_norm": 0.3502384126186371, | |
| "learning_rate": 1.662514603148847e-05, | |
| "loss": 1.0333, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.7498463429625076, | |
| "grad_norm": 0.3393804728984833, | |
| "learning_rate": 1.6472135208057126e-05, | |
| "loss": 1.0056, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.7510755992624463, | |
| "grad_norm": 0.3970140814781189, | |
| "learning_rate": 1.631969278709041e-05, | |
| "loss": 1.1957, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.7523048555623848, | |
| "grad_norm": 0.4345152974128723, | |
| "learning_rate": 1.616782135296544e-05, | |
| "loss": 1.0145, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.7523048555623848, | |
| "eval_loss": 0.9307632446289062, | |
| "eval_runtime": 64.7426, | |
| "eval_samples_per_second": 10.58, | |
| "eval_steps_per_second": 5.298, | |
| "step": 612 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 813, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 204, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.261491430967214e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |