| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.983792544570502, | |
| "eval_steps": 500, | |
| "global_step": 3080, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03241491085899514, | |
| "grad_norm": 1.712692141532898, | |
| "learning_rate": 0.00019999479806942977, | |
| "loss": 2.6861, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06482982171799027, | |
| "grad_norm": 2.658978223800659, | |
| "learning_rate": 0.00019997919281892067, | |
| "loss": 2.2023, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09724473257698542, | |
| "grad_norm": 1.8215035200119019, | |
| "learning_rate": 0.0001999531858720213, | |
| "loss": 1.8589, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12965964343598055, | |
| "grad_norm": 1.7040213346481323, | |
| "learning_rate": 0.0001999167799344583, | |
| "loss": 1.7118, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1620745542949757, | |
| "grad_norm": 1.647182583808899, | |
| "learning_rate": 0.00019986997879385487, | |
| "loss": 1.5532, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.19448946515397084, | |
| "grad_norm": 1.7023558616638184, | |
| "learning_rate": 0.0001998127873193367, | |
| "loss": 1.4578, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.22690437601296595, | |
| "grad_norm": 1.5219024419784546, | |
| "learning_rate": 0.00019974521146102537, | |
| "loss": 1.4007, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2593192868719611, | |
| "grad_norm": 1.5780329704284668, | |
| "learning_rate": 0.00019966725824941932, | |
| "loss": 1.409, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2917341977309562, | |
| "grad_norm": 1.7620179653167725, | |
| "learning_rate": 0.00019957893579466252, | |
| "loss": 1.3303, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3241491085899514, | |
| "grad_norm": 2.03952693939209, | |
| "learning_rate": 0.00019948025328570042, | |
| "loss": 1.2894, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3565640194489465, | |
| "grad_norm": 1.8262405395507812, | |
| "learning_rate": 0.00019937122098932428, | |
| "loss": 1.2291, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3889789303079417, | |
| "grad_norm": 1.7981160879135132, | |
| "learning_rate": 0.00019925185024910277, | |
| "loss": 1.2562, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4213938411669368, | |
| "grad_norm": 1.7449265718460083, | |
| "learning_rate": 0.000199122153484202, | |
| "loss": 1.2019, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4538087520259319, | |
| "grad_norm": 1.8147671222686768, | |
| "learning_rate": 0.0001989821441880933, | |
| "loss": 1.1528, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.4862236628849271, | |
| "grad_norm": 2.1661763191223145, | |
| "learning_rate": 0.00019883183692714936, | |
| "loss": 1.148, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5186385737439222, | |
| "grad_norm": 2.311131477355957, | |
| "learning_rate": 0.0001986712473391289, | |
| "loss": 1.1565, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5510534846029174, | |
| "grad_norm": 1.8314425945281982, | |
| "learning_rate": 0.00019850039213154973, | |
| "loss": 1.1258, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5834683954619124, | |
| "grad_norm": 2.260610580444336, | |
| "learning_rate": 0.0001983192890799503, | |
| "loss": 1.0491, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6158833063209076, | |
| "grad_norm": 2.0741446018218994, | |
| "learning_rate": 0.00019812795702604073, | |
| "loss": 1.0044, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6482982171799028, | |
| "grad_norm": 2.3104991912841797, | |
| "learning_rate": 0.00019792641587574212, | |
| "loss": 1.0391, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6807131280388979, | |
| "grad_norm": 2.024132251739502, | |
| "learning_rate": 0.00019771468659711595, | |
| "loss": 1.0802, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.713128038897893, | |
| "grad_norm": 2.667066812515259, | |
| "learning_rate": 0.00019749279121818235, | |
| "loss": 1.0215, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7455429497568882, | |
| "grad_norm": 2.127154588699341, | |
| "learning_rate": 0.00019726075282462845, | |
| "loss": 1.0016, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7779578606158833, | |
| "grad_norm": 2.398350238800049, | |
| "learning_rate": 0.00019701859555740648, | |
| "loss": 1.0767, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8103727714748784, | |
| "grad_norm": 2.4408252239227295, | |
| "learning_rate": 0.00019676634461022224, | |
| "loss": 0.969, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8427876823338736, | |
| "grad_norm": 2.225031614303589, | |
| "learning_rate": 0.000196504026226914, | |
| "loss": 1.0099, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8752025931928687, | |
| "grad_norm": 2.4684712886810303, | |
| "learning_rate": 0.00019623166769872213, | |
| "loss": 0.9523, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9076175040518638, | |
| "grad_norm": 2.4853105545043945, | |
| "learning_rate": 0.00019594929736144976, | |
| "loss": 0.8981, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.940032414910859, | |
| "grad_norm": 2.3052029609680176, | |
| "learning_rate": 0.00019565694459251458, | |
| "loss": 0.9157, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9724473257698542, | |
| "grad_norm": 2.227735996246338, | |
| "learning_rate": 0.00019535463980789277, | |
| "loss": 0.9085, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0048622366288493, | |
| "grad_norm": 2.3589160442352295, | |
| "learning_rate": 0.00019504241445895436, | |
| "loss": 0.9731, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0372771474878444, | |
| "grad_norm": 2.2362122535705566, | |
| "learning_rate": 0.000194720301029191, | |
| "loss": 0.8788, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.0696920583468394, | |
| "grad_norm": 2.2762033939361572, | |
| "learning_rate": 0.00019438833303083678, | |
| "loss": 0.8274, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.1021069692058347, | |
| "grad_norm": 2.9542107582092285, | |
| "learning_rate": 0.00019404654500138117, | |
| "loss": 0.8729, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.1345218800648298, | |
| "grad_norm": 2.541868209838867, | |
| "learning_rate": 0.0001936949724999762, | |
| "loss": 0.834, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.1669367909238249, | |
| "grad_norm": 2.314488649368286, | |
| "learning_rate": 0.0001933336521037367, | |
| "loss": 0.8432, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.1993517017828201, | |
| "grad_norm": 2.504655361175537, | |
| "learning_rate": 0.00019296262140393498, | |
| "loss": 0.8251, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.2317666126418152, | |
| "grad_norm": 2.568459987640381, | |
| "learning_rate": 0.0001925819190020898, | |
| "loss": 0.8152, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.2641815235008105, | |
| "grad_norm": 2.7527027130126953, | |
| "learning_rate": 0.0001921915845059505, | |
| "loss": 0.8599, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.2965964343598055, | |
| "grad_norm": 3.213120222091675, | |
| "learning_rate": 0.00019179165852537596, | |
| "loss": 0.7986, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.3290113452188006, | |
| "grad_norm": 2.718676805496216, | |
| "learning_rate": 0.00019138218266810986, | |
| "loss": 0.7604, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.3614262560777957, | |
| "grad_norm": 2.3581483364105225, | |
| "learning_rate": 0.00019096319953545185, | |
| "loss": 0.8096, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.393841166936791, | |
| "grad_norm": 2.966271162033081, | |
| "learning_rate": 0.0001905347527178252, | |
| "loss": 0.873, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.426256077795786, | |
| "grad_norm": 2.8818719387054443, | |
| "learning_rate": 0.0001900968867902419, | |
| "loss": 0.78, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.4586709886547813, | |
| "grad_norm": 3.3322315216064453, | |
| "learning_rate": 0.00018964964730766508, | |
| "loss": 0.8353, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.4910858995137763, | |
| "grad_norm": 2.844998359680176, | |
| "learning_rate": 0.0001891930808002694, | |
| "loss": 0.7845, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.5235008103727714, | |
| "grad_norm": 2.704843044281006, | |
| "learning_rate": 0.00018872723476860034, | |
| "loss": 0.7735, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.5559157212317665, | |
| "grad_norm": 2.935382604598999, | |
| "learning_rate": 0.00018825215767863214, | |
| "loss": 0.8104, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.5883306320907618, | |
| "grad_norm": 4.604203701019287, | |
| "learning_rate": 0.00018776789895672558, | |
| "loss": 0.7568, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.620745542949757, | |
| "grad_norm": 2.889993190765381, | |
| "learning_rate": 0.00018727450898448563, | |
| "loss": 0.7854, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.653160453808752, | |
| "grad_norm": 2.4442269802093506, | |
| "learning_rate": 0.00018677203909351988, | |
| "loss": 0.7134, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.6855753646677472, | |
| "grad_norm": 3.6301393508911133, | |
| "learning_rate": 0.00018626054156009806, | |
| "loss": 0.7339, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.7179902755267422, | |
| "grad_norm": 2.4832422733306885, | |
| "learning_rate": 0.00018574006959971333, | |
| "loss": 0.7377, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.7504051863857373, | |
| "grad_norm": 2.5708158016204834, | |
| "learning_rate": 0.00018521067736154568, | |
| "loss": 0.6838, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.7828200972447326, | |
| "grad_norm": 2.2001380920410156, | |
| "learning_rate": 0.00018467241992282843, | |
| "loss": 0.7134, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.8152350081037278, | |
| "grad_norm": 2.669135332107544, | |
| "learning_rate": 0.00018412535328311814, | |
| "loss": 0.7912, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.847649918962723, | |
| "grad_norm": 2.5101096630096436, | |
| "learning_rate": 0.0001835695343584683, | |
| "loss": 0.7279, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.880064829821718, | |
| "grad_norm": 1.9649478197097778, | |
| "learning_rate": 0.00018300502097550806, | |
| "loss": 0.6177, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.912479740680713, | |
| "grad_norm": 2.6189944744110107, | |
| "learning_rate": 0.00018243187186542593, | |
| "loss": 0.6625, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.9448946515397083, | |
| "grad_norm": 2.7569029331207275, | |
| "learning_rate": 0.00018185014665785936, | |
| "loss": 0.6726, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.9773095623987034, | |
| "grad_norm": 2.448319673538208, | |
| "learning_rate": 0.00018125990587469123, | |
| "loss": 0.7116, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.0097244732576987, | |
| "grad_norm": 2.4071240425109863, | |
| "learning_rate": 0.000180661210923753, | |
| "loss": 0.7076, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.0421393841166937, | |
| "grad_norm": 2.3127925395965576, | |
| "learning_rate": 0.00018005412409243606, | |
| "loss": 0.658, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.0745542949756888, | |
| "grad_norm": 2.6991357803344727, | |
| "learning_rate": 0.00017943870854121124, | |
| "loss": 0.6366, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.106969205834684, | |
| "grad_norm": 2.266465902328491, | |
| "learning_rate": 0.00017881502829705808, | |
| "loss": 0.6914, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.139384116693679, | |
| "grad_norm": 2.4977452754974365, | |
| "learning_rate": 0.000178183148246803, | |
| "loss": 0.6819, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.1717990275526744, | |
| "grad_norm": 2.3938634395599365, | |
| "learning_rate": 0.00017754313413036906, | |
| "loss": 0.6578, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.2042139384116695, | |
| "grad_norm": 2.4295880794525146, | |
| "learning_rate": 0.0001768950525339362, | |
| "loss": 0.6468, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.2366288492706645, | |
| "grad_norm": 2.621041774749756, | |
| "learning_rate": 0.00017623897088301385, | |
| "loss": 0.684, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.2690437601296596, | |
| "grad_norm": 1.9489623308181763, | |
| "learning_rate": 0.00017557495743542585, | |
| "loss": 0.5617, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.3014586709886546, | |
| "grad_norm": 2.1973514556884766, | |
| "learning_rate": 0.00017490308127420928, | |
| "loss": 0.6397, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.3338735818476497, | |
| "grad_norm": 3.3451037406921387, | |
| "learning_rate": 0.000174223412300427, | |
| "loss": 0.6761, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.366288492706645, | |
| "grad_norm": 3.31162691116333, | |
| "learning_rate": 0.00017353602122589527, | |
| "loss": 0.7305, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.3987034035656403, | |
| "grad_norm": 2.1021759510040283, | |
| "learning_rate": 0.00017284097956582692, | |
| "loss": 0.699, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.4311183144246353, | |
| "grad_norm": 2.290239095687866, | |
| "learning_rate": 0.0001721383596313912, | |
| "loss": 0.7112, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.4635332252836304, | |
| "grad_norm": 2.1187663078308105, | |
| "learning_rate": 0.00017142823452219038, | |
| "loss": 0.6473, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.4959481361426255, | |
| "grad_norm": 2.947108745574951, | |
| "learning_rate": 0.00017071067811865476, | |
| "loss": 0.6325, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.528363047001621, | |
| "grad_norm": 2.2677927017211914, | |
| "learning_rate": 0.00016998576507435618, | |
| "loss": 0.6988, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.560777957860616, | |
| "grad_norm": 3.2524213790893555, | |
| "learning_rate": 0.00016925357080824118, | |
| "loss": 0.6584, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.593192868719611, | |
| "grad_norm": 2.626284122467041, | |
| "learning_rate": 0.00016851417149678444, | |
| "loss": 0.5591, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.625607779578606, | |
| "grad_norm": 2.319129705429077, | |
| "learning_rate": 0.0001677676440660636, | |
| "loss": 0.5546, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.658022690437601, | |
| "grad_norm": 2.2486627101898193, | |
| "learning_rate": 0.00016701406618375596, | |
| "loss": 0.6935, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.6904376012965967, | |
| "grad_norm": 2.283766508102417, | |
| "learning_rate": 0.00016625351625105796, | |
| "loss": 0.673, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.7228525121555913, | |
| "grad_norm": 2.7028627395629883, | |
| "learning_rate": 0.00016548607339452853, | |
| "loss": 0.6226, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.755267423014587, | |
| "grad_norm": 2.0403945446014404, | |
| "learning_rate": 0.00016471181745785672, | |
| "loss": 0.6318, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.787682333873582, | |
| "grad_norm": 2.255915641784668, | |
| "learning_rate": 0.00016393082899355516, | |
| "loss": 0.5682, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.820097244732577, | |
| "grad_norm": 2.118745803833008, | |
| "learning_rate": 0.0001631431892545791, | |
| "loss": 0.6582, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.852512155591572, | |
| "grad_norm": 2.0190374851226807, | |
| "learning_rate": 0.00016234898018587337, | |
| "loss": 0.5768, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.884927066450567, | |
| "grad_norm": 2.165712356567383, | |
| "learning_rate": 0.00016154828441584655, | |
| "loss": 0.5919, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.9173419773095626, | |
| "grad_norm": 2.0873501300811768, | |
| "learning_rate": 0.00016074118524777477, | |
| "loss": 0.6416, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.9497568881685576, | |
| "grad_norm": 2.1629409790039062, | |
| "learning_rate": 0.0001599277666511347, | |
| "loss": 0.6438, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.9821717990275527, | |
| "grad_norm": 2.325366973876953, | |
| "learning_rate": 0.00015910811325286768, | |
| "loss": 0.6756, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.0145867098865478, | |
| "grad_norm": 2.101053237915039, | |
| "learning_rate": 0.00015828231032857503, | |
| "loss": 0.5999, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.047001620745543, | |
| "grad_norm": 1.8169053792953491, | |
| "learning_rate": 0.00015745044379364634, | |
| "loss": 0.6478, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.079416531604538, | |
| "grad_norm": 2.284700632095337, | |
| "learning_rate": 0.00015661260019432077, | |
| "loss": 0.6163, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.1118314424635334, | |
| "grad_norm": 3.3290014266967773, | |
| "learning_rate": 0.00015576886669868296, | |
| "loss": 0.5751, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.1442463533225284, | |
| "grad_norm": 2.246464252471924, | |
| "learning_rate": 0.0001549193310875942, | |
| "loss": 0.5481, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.1766612641815235, | |
| "grad_norm": 2.6920738220214844, | |
| "learning_rate": 0.00015406408174555976, | |
| "loss": 0.587, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.2090761750405186, | |
| "grad_norm": 2.41670560836792, | |
| "learning_rate": 0.00015320320765153367, | |
| "loss": 0.5998, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.2414910858995136, | |
| "grad_norm": 1.9720252752304077, | |
| "learning_rate": 0.00015233679836966122, | |
| "loss": 0.5327, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.2739059967585087, | |
| "grad_norm": 1.99226975440979, | |
| "learning_rate": 0.000151464944039961, | |
| "loss": 0.5932, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.306320907617504, | |
| "grad_norm": 2.752708911895752, | |
| "learning_rate": 0.00015058773536894685, | |
| "loss": 0.5989, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.3387358184764993, | |
| "grad_norm": 1.9078640937805176, | |
| "learning_rate": 0.00014970526362019079, | |
| "loss": 0.5447, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.3711507293354943, | |
| "grad_norm": 2.3850903511047363, | |
| "learning_rate": 0.00014881762060482814, | |
| "loss": 0.5739, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.4035656401944894, | |
| "grad_norm": 1.8733508586883545, | |
| "learning_rate": 0.0001479248986720057, | |
| "loss": 0.5407, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.4359805510534844, | |
| "grad_norm": 1.924999713897705, | |
| "learning_rate": 0.0001470271906992737, | |
| "loss": 0.576, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 3.46839546191248, | |
| "grad_norm": 3.5489633083343506, | |
| "learning_rate": 0.00014612459008292306, | |
| "loss": 0.5886, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 3.500810372771475, | |
| "grad_norm": 2.7815005779266357, | |
| "learning_rate": 0.00014521719072826858, | |
| "loss": 0.5426, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 3.53322528363047, | |
| "grad_norm": 1.678726315498352, | |
| "learning_rate": 0.00014430508703987914, | |
| "loss": 0.5983, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 3.565640194489465, | |
| "grad_norm": 2.03645658493042, | |
| "learning_rate": 0.00014338837391175582, | |
| "loss": 0.6045, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.59805510534846, | |
| "grad_norm": 2.302565813064575, | |
| "learning_rate": 0.00014246714671745965, | |
| "loss": 0.5739, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 3.6304700162074557, | |
| "grad_norm": 1.8924729824066162, | |
| "learning_rate": 0.00014154150130018866, | |
| "loss": 0.5296, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 3.6628849270664503, | |
| "grad_norm": 1.8624053001403809, | |
| "learning_rate": 0.00014061153396280674, | |
| "loss": 0.5886, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 3.695299837925446, | |
| "grad_norm": 2.516749143600464, | |
| "learning_rate": 0.00013967734145782425, | |
| "loss": 0.6285, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 3.727714748784441, | |
| "grad_norm": 2.6221561431884766, | |
| "learning_rate": 0.0001387390209773323, | |
| "loss": 0.6025, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.760129659643436, | |
| "grad_norm": 1.7727645635604858, | |
| "learning_rate": 0.00013779667014289065, | |
| "loss": 0.604, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 3.792544570502431, | |
| "grad_norm": 1.945691704750061, | |
| "learning_rate": 0.00013685038699537164, | |
| "loss": 0.6304, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.824959481361426, | |
| "grad_norm": 2.1147408485412598, | |
| "learning_rate": 0.00013590026998475986, | |
| "loss": 0.5754, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 3.8573743922204216, | |
| "grad_norm": 2.109344959259033, | |
| "learning_rate": 0.00013494641795990986, | |
| "loss": 0.6092, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 3.8897893030794166, | |
| "grad_norm": 1.9908583164215088, | |
| "learning_rate": 0.00013398893015826167, | |
| "loss": 0.5375, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.9222042139384117, | |
| "grad_norm": 2.164598226547241, | |
| "learning_rate": 0.00013302790619551674, | |
| "loss": 0.5625, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 3.9546191247974067, | |
| "grad_norm": 2.0762100219726562, | |
| "learning_rate": 0.00013206344605527355, | |
| "loss": 0.6212, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 3.987034035656402, | |
| "grad_norm": 2.2591452598571777, | |
| "learning_rate": 0.00013109565007862596, | |
| "loss": 0.5491, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.019448946515397, | |
| "grad_norm": 2.087096691131592, | |
| "learning_rate": 0.00013012461895372344, | |
| "loss": 0.5869, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 4.051863857374392, | |
| "grad_norm": 1.8744010925292969, | |
| "learning_rate": 0.00012915045370529585, | |
| "loss": 0.5585, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.084278768233387, | |
| "grad_norm": 2.3939051628112793, | |
| "learning_rate": 0.00012817325568414297, | |
| "loss": 0.5298, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 4.116693679092383, | |
| "grad_norm": 2.6422526836395264, | |
| "learning_rate": 0.00012719312655658994, | |
| "loss": 0.5765, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 4.1491085899513775, | |
| "grad_norm": 2.3070068359375, | |
| "learning_rate": 0.00012621016829391022, | |
| "loss": 0.5601, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 4.181523500810373, | |
| "grad_norm": 2.8217501640319824, | |
| "learning_rate": 0.0001252244831617165, | |
| "loss": 0.555, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 4.213938411669368, | |
| "grad_norm": 2.306428909301758, | |
| "learning_rate": 0.00012423617370932127, | |
| "loss": 0.546, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.246353322528363, | |
| "grad_norm": 2.691399097442627, | |
| "learning_rate": 0.00012324534275906748, | |
| "loss": 0.5794, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 4.278768233387358, | |
| "grad_norm": 2.576984167098999, | |
| "learning_rate": 0.00012225209339563145, | |
| "loss": 0.4677, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 4.311183144246353, | |
| "grad_norm": 2.3105177879333496, | |
| "learning_rate": 0.00012125652895529766, | |
| "loss": 0.4888, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 4.343598055105349, | |
| "grad_norm": 2.361992359161377, | |
| "learning_rate": 0.0001202587530152081, | |
| "loss": 0.5242, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 4.376012965964343, | |
| "grad_norm": 2.0497052669525146, | |
| "learning_rate": 0.00011925886938258605, | |
| "loss": 0.5341, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 4.408427876823339, | |
| "grad_norm": 2.1436045169830322, | |
| "learning_rate": 0.00011825698208393619, | |
| "loss": 0.565, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 4.4408427876823335, | |
| "grad_norm": 2.7684664726257324, | |
| "learning_rate": 0.00011725319535422188, | |
| "loss": 0.5257, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 4.473257698541329, | |
| "grad_norm": 2.2785606384277344, | |
| "learning_rate": 0.00011624761362602061, | |
| "loss": 0.5249, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 4.5056726094003245, | |
| "grad_norm": 4.94090461730957, | |
| "learning_rate": 0.00011524034151865903, | |
| "loss": 0.5505, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 4.538087520259319, | |
| "grad_norm": 1.984014868736267, | |
| "learning_rate": 0.00011423148382732853, | |
| "loss": 0.5028, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.570502431118315, | |
| "grad_norm": 2.671797275543213, | |
| "learning_rate": 0.00011322114551218239, | |
| "loss": 0.5099, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 4.602917341977309, | |
| "grad_norm": 2.390010356903076, | |
| "learning_rate": 0.000112209431687416, | |
| "loss": 0.5825, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 4.635332252836305, | |
| "grad_norm": 2.2715370655059814, | |
| "learning_rate": 0.00011119644761033078, | |
| "loss": 0.4849, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 4.667747163695299, | |
| "grad_norm": 1.9864623546600342, | |
| "learning_rate": 0.00011018229867038356, | |
| "loss": 0.5466, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 4.700162074554295, | |
| "grad_norm": 3.156597137451172, | |
| "learning_rate": 0.00010916709037822173, | |
| "loss": 0.5345, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 4.73257698541329, | |
| "grad_norm": 1.721481204032898, | |
| "learning_rate": 0.00010815092835470633, | |
| "loss": 0.5449, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 4.764991896272285, | |
| "grad_norm": 1.9270528554916382, | |
| "learning_rate": 0.00010713391831992323, | |
| "loss": 0.5303, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 4.7974068071312805, | |
| "grad_norm": 2.3868727684020996, | |
| "learning_rate": 0.00010611616608218429, | |
| "loss": 0.5057, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 4.829821717990275, | |
| "grad_norm": 1.9875102043151855, | |
| "learning_rate": 0.00010509777752701899, | |
| "loss": 0.5823, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 4.862236628849271, | |
| "grad_norm": 3.3236424922943115, | |
| "learning_rate": 0.00010407885860615859, | |
| "loss": 0.5184, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.894651539708266, | |
| "grad_norm": 2.4154036045074463, | |
| "learning_rate": 0.00010305951532651261, | |
| "loss": 0.5869, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 4.927066450567261, | |
| "grad_norm": 2.291398286819458, | |
| "learning_rate": 0.00010203985373914056, | |
| "loss": 0.4688, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 4.959481361426256, | |
| "grad_norm": 1.7252575159072876, | |
| "learning_rate": 0.00010101997992821797, | |
| "loss": 0.5315, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 4.991896272285251, | |
| "grad_norm": 2.263150453567505, | |
| "learning_rate": 0.0001, | |
| "loss": 0.5146, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 5.024311183144246, | |
| "grad_norm": 2.434597969055176, | |
| "learning_rate": 9.898002007178204e-05, | |
| "loss": 0.5458, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 5.056726094003242, | |
| "grad_norm": 2.3639023303985596, | |
| "learning_rate": 9.79601462608595e-05, | |
| "loss": 0.4089, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 5.0891410048622365, | |
| "grad_norm": 1.8895375728607178, | |
| "learning_rate": 9.69404846734874e-05, | |
| "loss": 0.5381, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 5.121555915721232, | |
| "grad_norm": 2.34358549118042, | |
| "learning_rate": 9.592114139384145e-05, | |
| "loss": 0.5385, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 5.153970826580227, | |
| "grad_norm": 2.613858699798584, | |
| "learning_rate": 9.490222247298099e-05, | |
| "loss": 0.5411, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 5.186385737439222, | |
| "grad_norm": 2.525398015975952, | |
| "learning_rate": 9.388383391781575e-05, | |
| "loss": 0.5645, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 5.218800648298217, | |
| "grad_norm": 2.6018924713134766, | |
| "learning_rate": 9.286608168007678e-05, | |
| "loss": 0.4934, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 5.251215559157212, | |
| "grad_norm": 2.232938051223755, | |
| "learning_rate": 9.184907164529368e-05, | |
| "loss": 0.4636, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 5.283630470016208, | |
| "grad_norm": 1.9651650190353394, | |
| "learning_rate": 9.083290962177828e-05, | |
| "loss": 0.5732, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 5.316045380875202, | |
| "grad_norm": 2.057870626449585, | |
| "learning_rate": 8.981770132961649e-05, | |
| "loss": 0.5266, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 5.348460291734198, | |
| "grad_norm": 2.1037161350250244, | |
| "learning_rate": 8.880355238966923e-05, | |
| "loss": 0.4874, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 5.3808752025931925, | |
| "grad_norm": 2.544857978820801, | |
| "learning_rate": 8.779056831258402e-05, | |
| "loss": 0.4599, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 5.413290113452188, | |
| "grad_norm": 1.9760535955429077, | |
| "learning_rate": 8.67788544878176e-05, | |
| "loss": 0.4675, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 5.4457050243111835, | |
| "grad_norm": 1.8703639507293701, | |
| "learning_rate": 8.57685161726715e-05, | |
| "loss": 0.4413, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 5.478119935170178, | |
| "grad_norm": 2.0244393348693848, | |
| "learning_rate": 8.475965848134099e-05, | |
| "loss": 0.4954, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 5.510534846029174, | |
| "grad_norm": 1.9399715662002563, | |
| "learning_rate": 8.375238637397942e-05, | |
| "loss": 0.4709, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 5.542949756888168, | |
| "grad_norm": 2.892331600189209, | |
| "learning_rate": 8.274680464577816e-05, | |
| "loss": 0.5021, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 5.575364667747164, | |
| "grad_norm": 1.9682655334472656, | |
| "learning_rate": 8.174301791606385e-05, | |
| "loss": 0.5648, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 5.607779578606159, | |
| "grad_norm": 2.52781343460083, | |
| "learning_rate": 8.074113061741397e-05, | |
| "loss": 0.5293, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 5.640194489465154, | |
| "grad_norm": 2.3522939682006836, | |
| "learning_rate": 7.974124698479192e-05, | |
| "loss": 0.5401, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 5.672609400324149, | |
| "grad_norm": 2.7281885147094727, | |
| "learning_rate": 7.874347104470234e-05, | |
| "loss": 0.4837, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 5.705024311183144, | |
| "grad_norm": 2.070852279663086, | |
| "learning_rate": 7.774790660436858e-05, | |
| "loss": 0.5035, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 5.7374392220421395, | |
| "grad_norm": 2.5109684467315674, | |
| "learning_rate": 7.675465724093254e-05, | |
| "loss": 0.4815, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 5.769854132901134, | |
| "grad_norm": 2.1801116466522217, | |
| "learning_rate": 7.576382629067877e-05, | |
| "loss": 0.497, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 5.80226904376013, | |
| "grad_norm": 2.372771978378296, | |
| "learning_rate": 7.47755168382835e-05, | |
| "loss": 0.499, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 5.834683954619125, | |
| "grad_norm": 2.1925764083862305, | |
| "learning_rate": 7.378983170608982e-05, | |
| "loss": 0.4702, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 5.86709886547812, | |
| "grad_norm": 2.0080742835998535, | |
| "learning_rate": 7.280687344341007e-05, | |
| "loss": 0.4886, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 5.899513776337115, | |
| "grad_norm": 2.1518988609313965, | |
| "learning_rate": 7.182674431585704e-05, | |
| "loss": 0.5052, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 5.93192868719611, | |
| "grad_norm": 2.1942789554595947, | |
| "learning_rate": 7.084954629470417e-05, | |
| "loss": 0.4922, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 5.964343598055105, | |
| "grad_norm": 2.380929470062256, | |
| "learning_rate": 6.98753810462766e-05, | |
| "loss": 0.4879, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 5.9967585089141, | |
| "grad_norm": 2.1748461723327637, | |
| "learning_rate": 6.890434992137407e-05, | |
| "loss": 0.4609, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 6.0291734197730955, | |
| "grad_norm": 1.9654253721237183, | |
| "learning_rate": 6.793655394472644e-05, | |
| "loss": 0.4869, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 6.061588330632091, | |
| "grad_norm": 2.2075066566467285, | |
| "learning_rate": 6.697209380448333e-05, | |
| "loss": 0.467, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 6.094003241491086, | |
| "grad_norm": 1.5888482332229614, | |
| "learning_rate": 6.601106984173835e-05, | |
| "loss": 0.4614, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 6.126418152350081, | |
| "grad_norm": 2.0989603996276855, | |
| "learning_rate": 6.505358204009017e-05, | |
| "loss": 0.4693, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 6.158833063209076, | |
| "grad_norm": 2.3755545616149902, | |
| "learning_rate": 6.409973001524012e-05, | |
| "loss": 0.4883, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 6.191247974068071, | |
| "grad_norm": 2.5981929302215576, | |
| "learning_rate": 6.314961300462841e-05, | |
| "loss": 0.4616, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 6.223662884927067, | |
| "grad_norm": 2.172053575515747, | |
| "learning_rate": 6.220332985710936e-05, | |
| "loss": 0.471, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 6.256077795786061, | |
| "grad_norm": 2.0949530601501465, | |
| "learning_rate": 6.126097902266772e-05, | |
| "loss": 0.405, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 6.288492706645057, | |
| "grad_norm": 2.166477680206299, | |
| "learning_rate": 6.0322658542175736e-05, | |
| "loss": 0.4973, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 6.3209076175040515, | |
| "grad_norm": 1.9600187540054321, | |
| "learning_rate": 5.938846603719332e-05, | |
| "loss": 0.4483, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 6.353322528363047, | |
| "grad_norm": 2.055723190307617, | |
| "learning_rate": 5.845849869981137e-05, | |
| "loss": 0.4557, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 6.3857374392220425, | |
| "grad_norm": 1.9242348670959473, | |
| "learning_rate": 5.7532853282540387e-05, | |
| "loss": 0.5454, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 6.418152350081037, | |
| "grad_norm": 2.1381163597106934, | |
| "learning_rate": 5.6611626088244194e-05, | |
| "loss": 0.4838, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 6.450567260940033, | |
| "grad_norm": 2.3222544193267822, | |
| "learning_rate": 5.569491296012095e-05, | |
| "loss": 0.5139, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 6.482982171799027, | |
| "grad_norm": 2.2901763916015625, | |
| "learning_rate": 5.478280927173145e-05, | |
| "loss": 0.4359, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.515397082658023, | |
| "grad_norm": 2.2004926204681396, | |
| "learning_rate": 5.387540991707697e-05, | |
| "loss": 0.5119, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 6.547811993517017, | |
| "grad_norm": 2.1431353092193604, | |
| "learning_rate": 5.297280930072632e-05, | |
| "loss": 0.4609, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 6.580226904376013, | |
| "grad_norm": 2.191962718963623, | |
| "learning_rate": 5.207510132799436e-05, | |
| "loss": 0.4566, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 6.612641815235008, | |
| "grad_norm": 1.8917332887649536, | |
| "learning_rate": 5.11823793951719e-05, | |
| "loss": 0.4692, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 6.645056726094003, | |
| "grad_norm": 2.5211122035980225, | |
| "learning_rate": 5.029473637980926e-05, | |
| "loss": 0.4943, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 6.6774716369529985, | |
| "grad_norm": 2.163224935531616, | |
| "learning_rate": 4.9412264631053216e-05, | |
| "loss": 0.4562, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 6.709886547811994, | |
| "grad_norm": 1.8480360507965088, | |
| "learning_rate": 4.853505596003905e-05, | |
| "loss": 0.4499, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 6.742301458670989, | |
| "grad_norm": 2.506883382797241, | |
| "learning_rate": 4.7663201630338816e-05, | |
| "loss": 0.4378, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 6.774716369529984, | |
| "grad_norm": 3.5621259212493896, | |
| "learning_rate": 4.6796792348466356e-05, | |
| "loss": 0.4866, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 6.807131280388979, | |
| "grad_norm": 2.3629701137542725, | |
| "learning_rate": 4.593591825444028e-05, | |
| "loss": 0.4785, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 6.839546191247974, | |
| "grad_norm": 2.753488779067993, | |
| "learning_rate": 4.5080668912405855e-05, | |
| "loss": 0.5154, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 6.871961102106969, | |
| "grad_norm": 2.4580469131469727, | |
| "learning_rate": 4.423113330131707e-05, | |
| "loss": 0.4569, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 6.904376012965964, | |
| "grad_norm": 2.252030849456787, | |
| "learning_rate": 4.3387399805679255e-05, | |
| "loss": 0.5148, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 6.93679092382496, | |
| "grad_norm": 2.605240821838379, | |
| "learning_rate": 4.25495562063537e-05, | |
| "loss": 0.4743, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 6.9692058346839545, | |
| "grad_norm": 2.0974061489105225, | |
| "learning_rate": 4.1717689671425e-05, | |
| "loss": 0.4211, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 7.00162074554295, | |
| "grad_norm": 1.8956321477890015, | |
| "learning_rate": 4.089188674713236e-05, | |
| "loss": 0.4759, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 7.034035656401945, | |
| "grad_norm": 2.391812801361084, | |
| "learning_rate": 4.007223334886531e-05, | |
| "loss": 0.4802, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 7.06645056726094, | |
| "grad_norm": 2.2890560626983643, | |
| "learning_rate": 3.9258814752225284e-05, | |
| "loss": 0.4934, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 7.098865478119935, | |
| "grad_norm": 1.9036931991577148, | |
| "learning_rate": 3.845171558415348e-05, | |
| "loss": 0.4442, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 7.13128038897893, | |
| "grad_norm": 2.3117735385894775, | |
| "learning_rate": 3.7651019814126654e-05, | |
| "loss": 0.4885, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 7.163695299837926, | |
| "grad_norm": 2.3671298027038574, | |
| "learning_rate": 3.6856810745420886e-05, | |
| "loss": 0.4552, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 7.19611021069692, | |
| "grad_norm": 2.2790403366088867, | |
| "learning_rate": 3.606917100644488e-05, | |
| "loss": 0.4461, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 7.228525121555916, | |
| "grad_norm": 3.0118699073791504, | |
| "learning_rate": 3.528818254214329e-05, | |
| "loss": 0.4328, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 7.2609400324149105, | |
| "grad_norm": 2.2282354831695557, | |
| "learning_rate": 3.45139266054715e-05, | |
| "loss": 0.4658, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 7.293354943273906, | |
| "grad_norm": 2.078871965408325, | |
| "learning_rate": 3.374648374894204e-05, | |
| "loss": 0.3856, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 7.3257698541329015, | |
| "grad_norm": 2.222506523132324, | |
| "learning_rate": 3.298593381624406e-05, | |
| "loss": 0.4336, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 7.358184764991896, | |
| "grad_norm": 2.130702257156372, | |
| "learning_rate": 3.223235593393642e-05, | |
| "loss": 0.4888, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 7.390599675850892, | |
| "grad_norm": 1.9949439764022827, | |
| "learning_rate": 3.1485828503215585e-05, | |
| "loss": 0.4748, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 7.423014586709886, | |
| "grad_norm": 2.212193727493286, | |
| "learning_rate": 3.074642919175883e-05, | |
| "loss": 0.4602, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 7.455429497568882, | |
| "grad_norm": 2.4536526203155518, | |
| "learning_rate": 3.0014234925643837e-05, | |
| "loss": 0.4509, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 7.487844408427877, | |
| "grad_norm": 1.9900058507919312, | |
| "learning_rate": 2.9289321881345254e-05, | |
| "loss": 0.3902, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 7.520259319286872, | |
| "grad_norm": 2.0711722373962402, | |
| "learning_rate": 2.8571765477809643e-05, | |
| "loss": 0.4508, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 7.552674230145867, | |
| "grad_norm": 2.640794515609741, | |
| "learning_rate": 2.7861640368608844e-05, | |
| "loss": 0.4944, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 7.585089141004862, | |
| "grad_norm": 2.4325428009033203, | |
| "learning_rate": 2.71590204341731e-05, | |
| "loss": 0.4438, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 7.6175040518638575, | |
| "grad_norm": 2.28466534614563, | |
| "learning_rate": 2.6463978774104758e-05, | |
| "loss": 0.4577, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 7.649918962722852, | |
| "grad_norm": 2.5928237438201904, | |
| "learning_rate": 2.5776587699573006e-05, | |
| "loss": 0.5176, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 7.682333873581848, | |
| "grad_norm": 1.911382794380188, | |
| "learning_rate": 2.509691872579075e-05, | |
| "loss": 0.4663, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 7.714748784440843, | |
| "grad_norm": 2.9931087493896484, | |
| "learning_rate": 2.4425042564574184e-05, | |
| "loss": 0.4284, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 7.747163695299838, | |
| "grad_norm": 2.6645960807800293, | |
| "learning_rate": 2.3761029116986178e-05, | |
| "loss": 0.472, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 7.779578606158833, | |
| "grad_norm": 2.0710065364837646, | |
| "learning_rate": 2.3104947466063787e-05, | |
| "loss": 0.4391, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 7.811993517017828, | |
| "grad_norm": 2.8465991020202637, | |
| "learning_rate": 2.2456865869630972e-05, | |
| "loss": 0.4844, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 7.844408427876823, | |
| "grad_norm": 2.3676860332489014, | |
| "learning_rate": 2.181685175319702e-05, | |
| "loss": 0.4028, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 7.876823338735819, | |
| "grad_norm": 2.0618326663970947, | |
| "learning_rate": 2.118497170294195e-05, | |
| "loss": 0.5056, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 7.9092382495948135, | |
| "grad_norm": 2.6422433853149414, | |
| "learning_rate": 2.0561291458788733e-05, | |
| "loss": 0.4602, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 7.941653160453809, | |
| "grad_norm": 2.1969330310821533, | |
| "learning_rate": 1.994587590756397e-05, | |
| "loss": 0.4256, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 7.974068071312804, | |
| "grad_norm": 2.574162721633911, | |
| "learning_rate": 1.9338789076247e-05, | |
| "loss": 0.4586, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 8.006482982171798, | |
| "grad_norm": 1.6469364166259766, | |
| "learning_rate": 1.874009412530877e-05, | |
| "loss": 0.419, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 8.038897893030795, | |
| "grad_norm": 2.7968332767486572, | |
| "learning_rate": 1.8149853342140645e-05, | |
| "loss": 0.4858, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 8.07131280388979, | |
| "grad_norm": 2.0327930450439453, | |
| "learning_rate": 1.7568128134574113e-05, | |
| "loss": 0.447, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 8.103727714748784, | |
| "grad_norm": 1.877231478691101, | |
| "learning_rate": 1.6994979024491942e-05, | |
| "loss": 0.4012, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 8.13614262560778, | |
| "grad_norm": 2.364046573638916, | |
| "learning_rate": 1.64304656415317e-05, | |
| "loss": 0.419, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 8.168557536466775, | |
| "grad_norm": 2.0480690002441406, | |
| "learning_rate": 1.587464671688187e-05, | |
| "loss": 0.4063, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 8.20097244732577, | |
| "grad_norm": 1.9457817077636719, | |
| "learning_rate": 1.5327580077171587e-05, | |
| "loss": 0.389, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 8.233387358184766, | |
| "grad_norm": 2.2139978408813477, | |
| "learning_rate": 1.4789322638454351e-05, | |
| "loss": 0.4014, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 8.26580226904376, | |
| "grad_norm": 2.7978177070617676, | |
| "learning_rate": 1.4259930400286669e-05, | |
| "loss": 0.4783, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 8.298217179902755, | |
| "grad_norm": 2.2762577533721924, | |
| "learning_rate": 1.373945843990192e-05, | |
| "loss": 0.4984, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 8.33063209076175, | |
| "grad_norm": 2.136845827102661, | |
| "learning_rate": 1.322796090648013e-05, | |
| "loss": 0.4487, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 8.363047001620746, | |
| "grad_norm": 2.696369171142578, | |
| "learning_rate": 1.272549101551438e-05, | |
| "loss": 0.4836, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 8.39546191247974, | |
| "grad_norm": 1.9966158866882324, | |
| "learning_rate": 1.2232101043274436e-05, | |
| "loss": 0.4004, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 8.427876823338735, | |
| "grad_norm": 2.289769411087036, | |
| "learning_rate": 1.1747842321367886e-05, | |
| "loss": 0.4693, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 8.460291734197732, | |
| "grad_norm": 1.8652838468551636, | |
| "learning_rate": 1.1272765231399685e-05, | |
| "loss": 0.4859, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 8.492706645056726, | |
| "grad_norm": 2.4249916076660156, | |
| "learning_rate": 1.0806919199730615e-05, | |
| "loss": 0.4189, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 8.525121555915721, | |
| "grad_norm": 2.178057909011841, | |
| "learning_rate": 1.035035269233493e-05, | |
| "loss": 0.4055, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 8.557536466774716, | |
| "grad_norm": 2.8611860275268555, | |
| "learning_rate": 9.903113209758096e-06, | |
| "loss": 0.4782, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 8.589951377633712, | |
| "grad_norm": 1.872373342514038, | |
| "learning_rate": 9.465247282174805e-06, | |
| "loss": 0.4262, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 8.622366288492707, | |
| "grad_norm": 2.744319200515747, | |
| "learning_rate": 9.036800464548157e-06, | |
| "loss": 0.4203, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 8.654781199351701, | |
| "grad_norm": 2.082002639770508, | |
| "learning_rate": 8.617817331890154e-06, | |
| "loss": 0.449, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 8.687196110210698, | |
| "grad_norm": 1.7450381517410278, | |
| "learning_rate": 8.208341474624071e-06, | |
| "loss": 0.4141, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 8.719611021069692, | |
| "grad_norm": 1.9353231191635132, | |
| "learning_rate": 7.808415494049514e-06, | |
| "loss": 0.4344, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 8.752025931928687, | |
| "grad_norm": 2.474977493286133, | |
| "learning_rate": 7.4180809979102036e-06, | |
| "loss": 0.4689, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 8.784440842787681, | |
| "grad_norm": 2.5193610191345215, | |
| "learning_rate": 7.0373785960650475e-06, | |
| "loss": 0.4791, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 8.816855753646678, | |
| "grad_norm": 2.5343973636627197, | |
| "learning_rate": 6.666347896263325e-06, | |
| "loss": 0.4609, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 8.849270664505672, | |
| "grad_norm": 2.6066677570343018, | |
| "learning_rate": 6.3050275000238414e-06, | |
| "loss": 0.4913, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 8.881685575364667, | |
| "grad_norm": 1.9997366666793823, | |
| "learning_rate": 5.953454998618857e-06, | |
| "loss": 0.4407, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 8.914100486223663, | |
| "grad_norm": 2.1540727615356445, | |
| "learning_rate": 5.611666969163243e-06, | |
| "loss": 0.4356, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 8.946515397082658, | |
| "grad_norm": 2.627286195755005, | |
| "learning_rate": 5.27969897080901e-06, | |
| "loss": 0.3782, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 8.978930307941653, | |
| "grad_norm": 2.116828680038452, | |
| "learning_rate": 4.957585541045684e-06, | |
| "loss": 0.4608, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 9.011345218800649, | |
| "grad_norm": 2.2772679328918457, | |
| "learning_rate": 4.6453601921072395e-06, | |
| "loss": 0.4667, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 9.043760129659644, | |
| "grad_norm": 1.7504632472991943, | |
| "learning_rate": 4.34305540748543e-06, | |
| "loss": 0.4035, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 9.076175040518638, | |
| "grad_norm": 1.660288691520691, | |
| "learning_rate": 4.050702638550275e-06, | |
| "loss": 0.4033, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 9.108589951377633, | |
| "grad_norm": 2.4989752769470215, | |
| "learning_rate": 3.768332301277866e-06, | |
| "loss": 0.4518, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 9.14100486223663, | |
| "grad_norm": 1.8193674087524414, | |
| "learning_rate": 3.495973773086014e-06, | |
| "loss": 0.3902, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 9.173419773095624, | |
| "grad_norm": 2.869234323501587, | |
| "learning_rate": 3.233655389777801e-06, | |
| "loss": 0.4842, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 9.205834683954619, | |
| "grad_norm": 2.684445381164551, | |
| "learning_rate": 2.9814044425935606e-06, | |
| "loss": 0.4263, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 9.238249594813615, | |
| "grad_norm": 2.5223307609558105, | |
| "learning_rate": 2.739247175371562e-06, | |
| "loss": 0.4493, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 9.27066450567261, | |
| "grad_norm": 2.3767752647399902, | |
| "learning_rate": 2.5072087818176382e-06, | |
| "loss": 0.4369, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 9.303079416531604, | |
| "grad_norm": 2.1580522060394287, | |
| "learning_rate": 2.2853134028840594e-06, | |
| "loss": 0.4093, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 9.335494327390599, | |
| "grad_norm": 1.7261040210723877, | |
| "learning_rate": 2.073584124257899e-06, | |
| "loss": 0.451, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 9.367909238249595, | |
| "grad_norm": 1.6289855241775513, | |
| "learning_rate": 1.8720429739592982e-06, | |
| "loss": 0.4305, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 9.40032414910859, | |
| "grad_norm": 2.0105481147766113, | |
| "learning_rate": 1.6807109200496995e-06, | |
| "loss": 0.4455, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 9.432739059967584, | |
| "grad_norm": 2.8889999389648438, | |
| "learning_rate": 1.4996078684503144e-06, | |
| "loss": 0.4451, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 9.46515397082658, | |
| "grad_norm": 2.474945306777954, | |
| "learning_rate": 1.3287526608711131e-06, | |
| "loss": 0.4945, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 9.497568881685575, | |
| "grad_norm": 1.9426692724227905, | |
| "learning_rate": 1.1681630728506699e-06, | |
| "loss": 0.4767, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 9.52998379254457, | |
| "grad_norm": 2.104893445968628, | |
| "learning_rate": 1.0178558119067315e-06, | |
| "loss": 0.4088, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 9.562398703403566, | |
| "grad_norm": 2.283007860183716, | |
| "learning_rate": 8.778465157979976e-07, | |
| "loss": 0.3957, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 9.594813614262561, | |
| "grad_norm": 2.111274480819702, | |
| "learning_rate": 7.481497508972312e-07, | |
| "loss": 0.4488, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 9.627228525121556, | |
| "grad_norm": 1.921383023262024, | |
| "learning_rate": 6.287790106757396e-07, | |
| "loss": 0.4685, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 9.65964343598055, | |
| "grad_norm": 2.4498233795166016, | |
| "learning_rate": 5.19746714299596e-07, | |
| "loss": 0.4298, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 9.692058346839547, | |
| "grad_norm": 2.4924449920654297, | |
| "learning_rate": 4.210642053375069e-07, | |
| "loss": 0.4393, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 9.724473257698541, | |
| "grad_norm": 2.780961513519287, | |
| "learning_rate": 3.3274175058067846e-07, | |
| "loss": 0.4444, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 9.756888168557536, | |
| "grad_norm": 2.037053108215332, | |
| "learning_rate": 2.547885389746485e-07, | |
| "loss": 0.4866, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 9.789303079416532, | |
| "grad_norm": 1.8418095111846924, | |
| "learning_rate": 1.8721268066330676e-07, | |
| "loss": 0.3989, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 9.821717990275527, | |
| "grad_norm": 2.2461559772491455, | |
| "learning_rate": 1.300212061451367e-07, | |
| "loss": 0.4047, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 9.854132901134522, | |
| "grad_norm": 1.7511504888534546, | |
| "learning_rate": 8.322006554171146e-08, | |
| "loss": 0.4171, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 9.886547811993516, | |
| "grad_norm": 2.4576804637908936, | |
| "learning_rate": 4.6814127978722644e-08, | |
| "loss": 0.4155, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 9.918962722852513, | |
| "grad_norm": 2.030487298965454, | |
| "learning_rate": 2.080718107935198e-08, | |
| "loss": 0.4397, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 9.951377633711507, | |
| "grad_norm": 2.6254327297210693, | |
| "learning_rate": 5.201930570242208e-09, | |
| "loss": 0.4338, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 9.983792544570502, | |
| "grad_norm": 2.197205066680908, | |
| "learning_rate": 0.0, | |
| "loss": 0.4586, | |
| "step": 3080 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3080, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.87794451980288e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |