| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.2507682851874616, | |
| "eval_steps": 204, | |
| "global_step": 204, | |
| "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 | |
| } | |
| ], | |
| "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": 4.204971437116621e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |