| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.06078042060051055, | |
| "eval_steps": 375, | |
| "global_step": 375, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00016208112160136148, | |
| "grad_norm": 0.32745468616485596, | |
| "learning_rate": 2e-05, | |
| "loss": 3.9495, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00032416224320272295, | |
| "grad_norm": 0.3682727813720703, | |
| "learning_rate": 4e-05, | |
| "loss": 4.3546, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00048624336480408443, | |
| "grad_norm": 0.3981589674949646, | |
| "learning_rate": 6e-05, | |
| "loss": 4.6652, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0006483244864054459, | |
| "grad_norm": 0.3928574323654175, | |
| "learning_rate": 8e-05, | |
| "loss": 4.4432, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0008104056080068074, | |
| "grad_norm": 0.4036743640899658, | |
| "learning_rate": 0.0001, | |
| "loss": 4.3653, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0009724867296081689, | |
| "grad_norm": 0.48630794882774353, | |
| "learning_rate": 0.00012, | |
| "loss": 4.258, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0011345678512095303, | |
| "grad_norm": 0.5550399422645569, | |
| "learning_rate": 0.00014, | |
| "loss": 4.3999, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0012966489728108918, | |
| "grad_norm": 0.6449398994445801, | |
| "learning_rate": 0.00016, | |
| "loss": 4.4376, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0014587300944122533, | |
| "grad_norm": 0.6104982495307922, | |
| "learning_rate": 0.00018, | |
| "loss": 4.3146, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0016208112160136148, | |
| "grad_norm": 0.509041965007782, | |
| "learning_rate": 0.0002, | |
| "loss": 3.9339, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0017828923376149762, | |
| "grad_norm": 0.6153712272644043, | |
| "learning_rate": 0.00019999977772170748, | |
| "loss": 3.9375, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0019449734592163377, | |
| "grad_norm": 0.8663574457168579, | |
| "learning_rate": 0.00019999911088781805, | |
| "loss": 3.9669, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.002107054580817699, | |
| "grad_norm": 1.0472509860992432, | |
| "learning_rate": 0.0001999979995012962, | |
| "loss": 3.9359, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0022691357024190607, | |
| "grad_norm": 1.0902113914489746, | |
| "learning_rate": 0.00019999644356708261, | |
| "loss": 3.7759, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.002431216824020422, | |
| "grad_norm": 0.8225506544113159, | |
| "learning_rate": 0.00019999444309209432, | |
| "loss": 3.8277, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0025932979456217836, | |
| "grad_norm": 0.6225056052207947, | |
| "learning_rate": 0.0001999919980852246, | |
| "loss": 3.7358, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.002755379067223145, | |
| "grad_norm": 0.7404794692993164, | |
| "learning_rate": 0.00019998910855734288, | |
| "loss": 3.8308, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0029174601888245066, | |
| "grad_norm": 1.1214481592178345, | |
| "learning_rate": 0.0001999857745212947, | |
| "loss": 3.8571, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.003079541310425868, | |
| "grad_norm": 1.0956681966781616, | |
| "learning_rate": 0.00019998199599190178, | |
| "loss": 3.6167, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0032416224320272295, | |
| "grad_norm": 1.0617244243621826, | |
| "learning_rate": 0.0001999777729859618, | |
| "loss": 3.6155, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.003403703553628591, | |
| "grad_norm": 0.872593104839325, | |
| "learning_rate": 0.00019997310552224846, | |
| "loss": 3.5173, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0035657846752299525, | |
| "grad_norm": 0.6966594457626343, | |
| "learning_rate": 0.00019996799362151122, | |
| "loss": 3.4923, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.003727865796831314, | |
| "grad_norm": 0.8658596277236938, | |
| "learning_rate": 0.00019996243730647538, | |
| "loss": 3.5702, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0038899469184326754, | |
| "grad_norm": 0.9274637699127197, | |
| "learning_rate": 0.00019995643660184191, | |
| "loss": 3.2378, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.004052028040034037, | |
| "grad_norm": 0.8372480273246765, | |
| "learning_rate": 0.00019994999153428737, | |
| "loss": 3.4719, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.004214109161635398, | |
| "grad_norm": 0.8468143939971924, | |
| "learning_rate": 0.00019994310213246368, | |
| "loss": 3.4335, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.00437619028323676, | |
| "grad_norm": 0.6658130884170532, | |
| "learning_rate": 0.00019993576842699816, | |
| "loss": 3.3862, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.004538271404838121, | |
| "grad_norm": 0.9544421434402466, | |
| "learning_rate": 0.0001999279904504933, | |
| "loss": 3.4167, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.004700352526439483, | |
| "grad_norm": 1.2153635025024414, | |
| "learning_rate": 0.00019991976823752653, | |
| "loss": 3.4523, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.004862433648040844, | |
| "grad_norm": 1.0208470821380615, | |
| "learning_rate": 0.00019991110182465032, | |
| "loss": 3.2573, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.005024514769642206, | |
| "grad_norm": 0.7645734548568726, | |
| "learning_rate": 0.00019990199125039174, | |
| "loss": 3.2797, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.005186595891243567, | |
| "grad_norm": 0.7118749022483826, | |
| "learning_rate": 0.00019989243655525247, | |
| "loss": 3.2598, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.005348677012844929, | |
| "grad_norm": 1.0173993110656738, | |
| "learning_rate": 0.00019988243778170853, | |
| "loss": 3.2678, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.00551075813444629, | |
| "grad_norm": 1.3724397420883179, | |
| "learning_rate": 0.0001998719949742101, | |
| "loss": 3.2351, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.005672839256047652, | |
| "grad_norm": 1.1269340515136719, | |
| "learning_rate": 0.0001998611081791814, | |
| "loss": 3.2311, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.005834920377649013, | |
| "grad_norm": 0.7900427579879761, | |
| "learning_rate": 0.00019984977744502038, | |
| "loss": 3.1872, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.005997001499250375, | |
| "grad_norm": 0.996651828289032, | |
| "learning_rate": 0.00019983800282209857, | |
| "loss": 3.272, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.006159082620851736, | |
| "grad_norm": 1.2565242052078247, | |
| "learning_rate": 0.00019982578436276082, | |
| "loss": 3.2472, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.006321163742453098, | |
| "grad_norm": 1.2510484457015991, | |
| "learning_rate": 0.00019981312212132512, | |
| "loss": 3.4406, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.006483244864054459, | |
| "grad_norm": 0.872576117515564, | |
| "learning_rate": 0.00019980001615408228, | |
| "loss": 3.2457, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0066453259856558205, | |
| "grad_norm": 1.4087495803833008, | |
| "learning_rate": 0.00019978646651929572, | |
| "loss": 3.6152, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.006807407107257182, | |
| "grad_norm": 1.4746192693710327, | |
| "learning_rate": 0.00019977247327720128, | |
| "loss": 3.4374, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.0069694882288585435, | |
| "grad_norm": 1.278622031211853, | |
| "learning_rate": 0.0001997580364900068, | |
| "loss": 3.3347, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.007131569350459905, | |
| "grad_norm": 1.1743550300598145, | |
| "learning_rate": 0.000199743156221892, | |
| "loss": 3.4706, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.007293650472061266, | |
| "grad_norm": 1.4559597969055176, | |
| "learning_rate": 0.00019972783253900808, | |
| "loss": 3.3314, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.007455731593662628, | |
| "grad_norm": 1.3842203617095947, | |
| "learning_rate": 0.00019971206550947748, | |
| "loss": 3.477, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.007617812715263989, | |
| "grad_norm": 1.3693876266479492, | |
| "learning_rate": 0.00019969585520339354, | |
| "loss": 3.4538, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.007779893836865351, | |
| "grad_norm": 1.9597094058990479, | |
| "learning_rate": 0.0001996792016928203, | |
| "loss": 3.5043, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.007941974958466713, | |
| "grad_norm": 1.8300191164016724, | |
| "learning_rate": 0.00019966210505179197, | |
| "loss": 3.7964, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.008104056080068074, | |
| "grad_norm": 2.7875523567199707, | |
| "learning_rate": 0.00019964456535631286, | |
| "loss": 4.3156, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.008266137201669436, | |
| "grad_norm": 1.066572666168213, | |
| "learning_rate": 0.0001996265826843568, | |
| "loss": 2.8682, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.008428218323270797, | |
| "grad_norm": 1.0621248483657837, | |
| "learning_rate": 0.00019960815711586696, | |
| "loss": 3.0677, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.008590299444872159, | |
| "grad_norm": 0.9207605123519897, | |
| "learning_rate": 0.00019958928873275539, | |
| "loss": 2.8688, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.00875238056647352, | |
| "grad_norm": 0.7903249263763428, | |
| "learning_rate": 0.00019956997761890277, | |
| "loss": 2.7778, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.008914461688074882, | |
| "grad_norm": 0.8444027900695801, | |
| "learning_rate": 0.00019955022386015792, | |
| "loss": 2.6653, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.009076542809676243, | |
| "grad_norm": 0.8025749921798706, | |
| "learning_rate": 0.00019953002754433743, | |
| "loss": 2.6123, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.009238623931277605, | |
| "grad_norm": 0.7850587964057922, | |
| "learning_rate": 0.00019950938876122542, | |
| "loss": 2.7537, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.009400705052878966, | |
| "grad_norm": 0.7457495927810669, | |
| "learning_rate": 0.00019948830760257291, | |
| "loss": 2.7807, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.009562786174480328, | |
| "grad_norm": 0.7835890650749207, | |
| "learning_rate": 0.0001994667841620976, | |
| "loss": 2.6427, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.009724867296081689, | |
| "grad_norm": 0.8000785708427429, | |
| "learning_rate": 0.00019944481853548335, | |
| "loss": 2.5705, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.009886948417683051, | |
| "grad_norm": 1.0432358980178833, | |
| "learning_rate": 0.00019942241082037982, | |
| "loss": 2.8235, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.010049029539284412, | |
| "grad_norm": 0.7705084681510925, | |
| "learning_rate": 0.00019939956111640197, | |
| "loss": 2.5888, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.010211110660885774, | |
| "grad_norm": 0.8200328350067139, | |
| "learning_rate": 0.00019937626952512964, | |
| "loss": 2.7773, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.010373191782487134, | |
| "grad_norm": 0.6848300099372864, | |
| "learning_rate": 0.0001993525361501072, | |
| "loss": 2.7091, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.010535272904088497, | |
| "grad_norm": 0.7549343705177307, | |
| "learning_rate": 0.00019932836109684286, | |
| "loss": 2.4635, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.010697354025689857, | |
| "grad_norm": 0.7910590171813965, | |
| "learning_rate": 0.00019930374447280845, | |
| "loss": 2.6742, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.01085943514729122, | |
| "grad_norm": 0.8571573495864868, | |
| "learning_rate": 0.00019927868638743875, | |
| "loss": 2.6027, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.01102151626889258, | |
| "grad_norm": 0.7997406721115112, | |
| "learning_rate": 0.0001992531869521312, | |
| "loss": 2.6535, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.011183597390493943, | |
| "grad_norm": 0.8311846852302551, | |
| "learning_rate": 0.00019922724628024515, | |
| "loss": 2.743, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.011345678512095303, | |
| "grad_norm": 0.7518640756607056, | |
| "learning_rate": 0.0001992008644871016, | |
| "loss": 2.5151, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.011507759633696666, | |
| "grad_norm": 0.890500545501709, | |
| "learning_rate": 0.00019917404168998256, | |
| "loss": 2.7083, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.011669840755298026, | |
| "grad_norm": 0.7954429984092712, | |
| "learning_rate": 0.0001991467780081305, | |
| "loss": 2.6477, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.011831921876899389, | |
| "grad_norm": 0.7375217080116272, | |
| "learning_rate": 0.00019911907356274795, | |
| "loss": 2.4982, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.01199400299850075, | |
| "grad_norm": 0.7944385409355164, | |
| "learning_rate": 0.00019909092847699683, | |
| "loss": 2.6131, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.012156084120102112, | |
| "grad_norm": 0.881007194519043, | |
| "learning_rate": 0.00019906234287599798, | |
| "loss": 2.7245, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.012318165241703472, | |
| "grad_norm": 0.8365350961685181, | |
| "learning_rate": 0.00019903331688683057, | |
| "loss": 2.6156, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.012480246363304835, | |
| "grad_norm": 0.8571576476097107, | |
| "learning_rate": 0.00019900385063853154, | |
| "loss": 2.5997, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.012642327484906195, | |
| "grad_norm": 0.8759616017341614, | |
| "learning_rate": 0.00019897394426209505, | |
| "loss": 2.5906, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.012804408606507557, | |
| "grad_norm": 0.8624269962310791, | |
| "learning_rate": 0.00019894359789047187, | |
| "loss": 2.7459, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.012966489728108918, | |
| "grad_norm": 0.9585336446762085, | |
| "learning_rate": 0.00019891281165856873, | |
| "loss": 2.4964, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.01312857084971028, | |
| "grad_norm": 0.9535956978797913, | |
| "learning_rate": 0.00019888158570324795, | |
| "loss": 2.6338, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.013290651971311641, | |
| "grad_norm": 0.8365891575813293, | |
| "learning_rate": 0.0001988499201633265, | |
| "loss": 2.6502, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.013452733092913003, | |
| "grad_norm": 1.0698297023773193, | |
| "learning_rate": 0.00019881781517957562, | |
| "loss": 2.6211, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.013614814214514364, | |
| "grad_norm": 0.9466975331306458, | |
| "learning_rate": 0.0001987852708947202, | |
| "loss": 2.69, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.013776895336115726, | |
| "grad_norm": 1.2203270196914673, | |
| "learning_rate": 0.00019875228745343794, | |
| "loss": 2.741, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.013938976457717087, | |
| "grad_norm": 1.1437451839447021, | |
| "learning_rate": 0.0001987188650023589, | |
| "loss": 2.8166, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.01410105757931845, | |
| "grad_norm": 1.4375919103622437, | |
| "learning_rate": 0.0001986850036900648, | |
| "loss": 2.6542, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.01426313870091981, | |
| "grad_norm": 1.3663667440414429, | |
| "learning_rate": 0.00019865070366708836, | |
| "loss": 2.597, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.014425219822521172, | |
| "grad_norm": 1.2260905504226685, | |
| "learning_rate": 0.00019861596508591255, | |
| "loss": 2.9777, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.014587300944122533, | |
| "grad_norm": 1.7002184391021729, | |
| "learning_rate": 0.00019858078810097002, | |
| "loss": 2.6876, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.014749382065723895, | |
| "grad_norm": 1.382377028465271, | |
| "learning_rate": 0.00019854517286864245, | |
| "loss": 2.9024, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.014911463187325256, | |
| "grad_norm": 1.7291725873947144, | |
| "learning_rate": 0.0001985091195472596, | |
| "loss": 2.8279, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.015073544308926618, | |
| "grad_norm": 1.690996766090393, | |
| "learning_rate": 0.0001984726282970989, | |
| "loss": 2.9042, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.015235625430527979, | |
| "grad_norm": 1.3773317337036133, | |
| "learning_rate": 0.0001984356992803847, | |
| "loss": 2.8591, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.015397706552129341, | |
| "grad_norm": 2.045140027999878, | |
| "learning_rate": 0.00019839833266128724, | |
| "loss": 3.0208, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.015559787673730702, | |
| "grad_norm": 1.5940877199172974, | |
| "learning_rate": 0.00019836052860592237, | |
| "loss": 3.0601, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.015721868795332064, | |
| "grad_norm": 1.720646858215332, | |
| "learning_rate": 0.0001983222872823505, | |
| "loss": 2.9597, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.015883949916933426, | |
| "grad_norm": 2.037874460220337, | |
| "learning_rate": 0.00019828360886057594, | |
| "loss": 3.1805, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.016046031038534785, | |
| "grad_norm": 2.515458822250366, | |
| "learning_rate": 0.00019824449351254616, | |
| "loss": 3.2132, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.016208112160136148, | |
| "grad_norm": 3.2635490894317627, | |
| "learning_rate": 0.00019820494141215104, | |
| "loss": 3.7978, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.01637019328173751, | |
| "grad_norm": 1.3721284866333008, | |
| "learning_rate": 0.000198164952735222, | |
| "loss": 2.7209, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.016532274403338872, | |
| "grad_norm": 0.8304929733276367, | |
| "learning_rate": 0.00019812452765953135, | |
| "loss": 2.4077, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.01669435552494023, | |
| "grad_norm": 1.0842708349227905, | |
| "learning_rate": 0.00019808366636479147, | |
| "loss": 2.4252, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.016856436646541594, | |
| "grad_norm": 0.8582240343093872, | |
| "learning_rate": 0.00019804236903265388, | |
| "loss": 2.2418, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.017018517768142956, | |
| "grad_norm": 0.8859397172927856, | |
| "learning_rate": 0.00019800063584670863, | |
| "loss": 2.3411, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.017180598889744318, | |
| "grad_norm": 0.9900501370429993, | |
| "learning_rate": 0.00019795846699248332, | |
| "loss": 2.4639, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.017342680011345677, | |
| "grad_norm": 1.003010869026184, | |
| "learning_rate": 0.00019791586265744237, | |
| "loss": 2.5822, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.01750476113294704, | |
| "grad_norm": 0.8580072522163391, | |
| "learning_rate": 0.00019787282303098617, | |
| "loss": 2.3763, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.017666842254548402, | |
| "grad_norm": 0.9203416109085083, | |
| "learning_rate": 0.0001978293483044502, | |
| "loss": 2.416, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.017828923376149764, | |
| "grad_norm": 0.7686523199081421, | |
| "learning_rate": 0.00019778543867110426, | |
| "loss": 2.3347, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.017991004497751123, | |
| "grad_norm": 0.7696153521537781, | |
| "learning_rate": 0.00019774109432615147, | |
| "loss": 2.1202, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.018153085619352485, | |
| "grad_norm": 0.8893393278121948, | |
| "learning_rate": 0.00019769631546672756, | |
| "loss": 2.4122, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.018315166740953848, | |
| "grad_norm": 0.9954109787940979, | |
| "learning_rate": 0.00019765110229189988, | |
| "loss": 2.3672, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.01847724786255521, | |
| "grad_norm": 0.9527397751808167, | |
| "learning_rate": 0.00019760545500266657, | |
| "loss": 2.3611, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.01863932898415657, | |
| "grad_norm": 0.9039053320884705, | |
| "learning_rate": 0.00019755937380195568, | |
| "loss": 2.3882, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.01880141010575793, | |
| "grad_norm": 0.8039470911026001, | |
| "learning_rate": 0.00019751285889462423, | |
| "loss": 2.3428, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.018963491227359294, | |
| "grad_norm": 0.8728412389755249, | |
| "learning_rate": 0.0001974659104874573, | |
| "loss": 2.4066, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.019125572348960656, | |
| "grad_norm": 0.847098708152771, | |
| "learning_rate": 0.0001974185287891671, | |
| "loss": 2.465, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.019287653470562015, | |
| "grad_norm": 0.8857790231704712, | |
| "learning_rate": 0.0001973707140103921, | |
| "loss": 2.4859, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.019449734592163377, | |
| "grad_norm": 0.9171362519264221, | |
| "learning_rate": 0.00019732246636369605, | |
| "loss": 2.3492, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.01961181571376474, | |
| "grad_norm": 0.8632153868675232, | |
| "learning_rate": 0.00019727378606356703, | |
| "loss": 2.3529, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.019773896835366102, | |
| "grad_norm": 1.0254154205322266, | |
| "learning_rate": 0.00019722467332641656, | |
| "loss": 2.4594, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.01993597795696746, | |
| "grad_norm": 0.9391173720359802, | |
| "learning_rate": 0.00019717512837057855, | |
| "loss": 2.1394, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.020098059078568823, | |
| "grad_norm": 1.1696561574935913, | |
| "learning_rate": 0.0001971251514163083, | |
| "loss": 2.4457, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.020260140200170185, | |
| "grad_norm": 1.1914767026901245, | |
| "learning_rate": 0.0001970747426857817, | |
| "loss": 2.4778, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.020422221321771548, | |
| "grad_norm": 1.1912813186645508, | |
| "learning_rate": 0.00019702390240309404, | |
| "loss": 2.5327, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.020584302443372907, | |
| "grad_norm": 1.272786021232605, | |
| "learning_rate": 0.0001969726307942592, | |
| "loss": 2.5287, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.02074638356497427, | |
| "grad_norm": 0.9330551028251648, | |
| "learning_rate": 0.00019692092808720846, | |
| "loss": 2.5395, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.02090846468657563, | |
| "grad_norm": 1.370786428451538, | |
| "learning_rate": 0.0001968687945117896, | |
| "loss": 2.4154, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.021070545808176994, | |
| "grad_norm": 1.041229009628296, | |
| "learning_rate": 0.00019681623029976588, | |
| "loss": 2.435, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.021232626929778353, | |
| "grad_norm": 1.0356900691986084, | |
| "learning_rate": 0.00019676323568481498, | |
| "loss": 2.603, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.021394708051379715, | |
| "grad_norm": 1.4157955646514893, | |
| "learning_rate": 0.00019670981090252792, | |
| "loss": 2.4694, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.021556789172981077, | |
| "grad_norm": 1.2365660667419434, | |
| "learning_rate": 0.00019665595619040808, | |
| "loss": 2.4311, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.02171887029458244, | |
| "grad_norm": 1.5523958206176758, | |
| "learning_rate": 0.0001966016717878702, | |
| "loss": 2.5472, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.0218809514161838, | |
| "grad_norm": 1.3933662176132202, | |
| "learning_rate": 0.00019654695793623907, | |
| "loss": 2.5602, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.02204303253778516, | |
| "grad_norm": 1.3399230241775513, | |
| "learning_rate": 0.0001964918148787488, | |
| "loss": 2.3793, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.022205113659386523, | |
| "grad_norm": 1.1035332679748535, | |
| "learning_rate": 0.00019643624286054144, | |
| "loss": 2.6046, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.022367194780987885, | |
| "grad_norm": 1.1267286539077759, | |
| "learning_rate": 0.00019638024212866606, | |
| "loss": 2.5393, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.022529275902589244, | |
| "grad_norm": 1.1763005256652832, | |
| "learning_rate": 0.0001963238129320776, | |
| "loss": 2.5834, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.022691357024190607, | |
| "grad_norm": 1.3848681449890137, | |
| "learning_rate": 0.00019626695552163578, | |
| "loss": 2.8858, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.02285343814579197, | |
| "grad_norm": 1.1526983976364136, | |
| "learning_rate": 0.00019620967015010395, | |
| "loss": 2.6724, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.02301551926739333, | |
| "grad_norm": 1.6136480569839478, | |
| "learning_rate": 0.00019615195707214803, | |
| "loss": 2.7808, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.02317760038899469, | |
| "grad_norm": 1.360098123550415, | |
| "learning_rate": 0.0001960938165443353, | |
| "loss": 2.923, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.023339681510596053, | |
| "grad_norm": 1.4082938432693481, | |
| "learning_rate": 0.00019603524882513327, | |
| "loss": 2.744, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.023501762632197415, | |
| "grad_norm": 1.5163816213607788, | |
| "learning_rate": 0.0001959762541749086, | |
| "loss": 2.8498, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.023663843753798777, | |
| "grad_norm": 1.6904058456420898, | |
| "learning_rate": 0.00019591683285592593, | |
| "loss": 2.8353, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.023825924875400136, | |
| "grad_norm": 1.7553011178970337, | |
| "learning_rate": 0.00019585698513234663, | |
| "loss": 2.8577, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.0239880059970015, | |
| "grad_norm": 2.2253310680389404, | |
| "learning_rate": 0.0001957967112702277, | |
| "loss": 3.1318, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.02415008711860286, | |
| "grad_norm": 2.3773317337036133, | |
| "learning_rate": 0.00019573601153752052, | |
| "loss": 3.1695, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.024312168240204223, | |
| "grad_norm": 3.46358585357666, | |
| "learning_rate": 0.00019567488620406983, | |
| "loss": 3.66, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.024474249361805582, | |
| "grad_norm": 1.1002161502838135, | |
| "learning_rate": 0.00019561333554161224, | |
| "loss": 2.5787, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.024636330483406944, | |
| "grad_norm": 0.8970770835876465, | |
| "learning_rate": 0.0001955513598237753, | |
| "loss": 2.3198, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.024798411605008307, | |
| "grad_norm": 0.9427694082260132, | |
| "learning_rate": 0.00019548895932607621, | |
| "loss": 2.4208, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.02496049272660967, | |
| "grad_norm": 0.936225414276123, | |
| "learning_rate": 0.00019542613432592038, | |
| "loss": 2.3963, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.025122573848211028, | |
| "grad_norm": 0.8925445675849915, | |
| "learning_rate": 0.00019536288510260056, | |
| "loss": 2.2791, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.02528465496981239, | |
| "grad_norm": 1.1286380290985107, | |
| "learning_rate": 0.00019529921193729534, | |
| "loss": 2.3948, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.025446736091413753, | |
| "grad_norm": 0.8829410672187805, | |
| "learning_rate": 0.00019523511511306793, | |
| "loss": 2.3171, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.025608817213015115, | |
| "grad_norm": 0.9467886686325073, | |
| "learning_rate": 0.000195170594914865, | |
| "loss": 2.2482, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.025770898334616474, | |
| "grad_norm": 1.098661184310913, | |
| "learning_rate": 0.00019510565162951537, | |
| "loss": 2.2541, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.025932979456217836, | |
| "grad_norm": 1.0854690074920654, | |
| "learning_rate": 0.00019504028554572864, | |
| "loss": 2.261, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.0260950605778192, | |
| "grad_norm": 0.887253999710083, | |
| "learning_rate": 0.00019497449695409408, | |
| "loss": 2.3605, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.02625714169942056, | |
| "grad_norm": 0.8699051737785339, | |
| "learning_rate": 0.00019490828614707916, | |
| "loss": 2.1781, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.02641922282102192, | |
| "grad_norm": 0.9914195537567139, | |
| "learning_rate": 0.00019484165341902845, | |
| "loss": 2.3487, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.026581303942623282, | |
| "grad_norm": 0.7977816462516785, | |
| "learning_rate": 0.00019477459906616206, | |
| "loss": 2.2499, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.026743385064224644, | |
| "grad_norm": 1.0885521173477173, | |
| "learning_rate": 0.00019470712338657458, | |
| "loss": 2.3353, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.026905466185826007, | |
| "grad_norm": 0.9307372570037842, | |
| "learning_rate": 0.0001946392266802336, | |
| "loss": 2.349, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.02706754730742737, | |
| "grad_norm": 0.8721451163291931, | |
| "learning_rate": 0.0001945709092489783, | |
| "loss": 2.1844, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.027229628429028728, | |
| "grad_norm": 0.8243699669837952, | |
| "learning_rate": 0.00019450217139651844, | |
| "loss": 2.3944, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.02739170955063009, | |
| "grad_norm": 0.8237115144729614, | |
| "learning_rate": 0.0001944330134284326, | |
| "loss": 2.3017, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.027553790672231453, | |
| "grad_norm": 1.0215017795562744, | |
| "learning_rate": 0.00019436343565216711, | |
| "loss": 2.377, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.027715871793832815, | |
| "grad_norm": 1.0807178020477295, | |
| "learning_rate": 0.00019429343837703455, | |
| "loss": 2.3705, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.027877952915434174, | |
| "grad_norm": 0.8931332230567932, | |
| "learning_rate": 0.0001942230219142124, | |
| "loss": 2.3758, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.028040034037035536, | |
| "grad_norm": 0.9059082865715027, | |
| "learning_rate": 0.0001941521865767417, | |
| "loss": 2.3395, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.0282021151586369, | |
| "grad_norm": 0.8843547701835632, | |
| "learning_rate": 0.0001940809326795256, | |
| "loss": 2.3348, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.02836419628023826, | |
| "grad_norm": 0.9769134521484375, | |
| "learning_rate": 0.000194009260539328, | |
| "loss": 2.3773, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.02852627740183962, | |
| "grad_norm": 0.9252240657806396, | |
| "learning_rate": 0.0001939371704747721, | |
| "loss": 2.2362, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.028688358523440982, | |
| "grad_norm": 0.9405879974365234, | |
| "learning_rate": 0.00019386466280633906, | |
| "loss": 2.3078, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.028850439645042344, | |
| "grad_norm": 0.9854640960693359, | |
| "learning_rate": 0.00019379173785636646, | |
| "loss": 2.4224, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.029012520766643707, | |
| "grad_norm": 1.0185924768447876, | |
| "learning_rate": 0.000193718395949047, | |
| "loss": 2.3796, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.029174601888245066, | |
| "grad_norm": 1.1267231702804565, | |
| "learning_rate": 0.00019364463741042694, | |
| "loss": 2.5936, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.029336683009846428, | |
| "grad_norm": 1.0917437076568604, | |
| "learning_rate": 0.00019357046256840473, | |
| "loss": 2.436, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.02949876413144779, | |
| "grad_norm": 1.4547231197357178, | |
| "learning_rate": 0.00019349587175272948, | |
| "loss": 2.4356, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.029660845253049153, | |
| "grad_norm": 1.2778880596160889, | |
| "learning_rate": 0.0001934208652949996, | |
| "loss": 2.3618, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.02982292637465051, | |
| "grad_norm": 1.2359460592269897, | |
| "learning_rate": 0.00019334544352866127, | |
| "loss": 2.2503, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.029985007496251874, | |
| "grad_norm": 1.6187139749526978, | |
| "learning_rate": 0.00019326960678900688, | |
| "loss": 2.5739, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.030147088617853236, | |
| "grad_norm": 1.077547311782837, | |
| "learning_rate": 0.00019319335541317361, | |
| "loss": 2.5456, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.0303091697394546, | |
| "grad_norm": 1.2906684875488281, | |
| "learning_rate": 0.00019311668974014208, | |
| "loss": 2.4993, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.030471250861055958, | |
| "grad_norm": 1.446529507637024, | |
| "learning_rate": 0.00019303961011073447, | |
| "loss": 2.4235, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.03063333198265732, | |
| "grad_norm": 1.2975163459777832, | |
| "learning_rate": 0.00019296211686761346, | |
| "loss": 2.6774, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.030795413104258682, | |
| "grad_norm": 1.6122586727142334, | |
| "learning_rate": 0.00019288421035528028, | |
| "loss": 2.5858, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.030957494225860045, | |
| "grad_norm": 1.3004416227340698, | |
| "learning_rate": 0.00019280589092007352, | |
| "loss": 2.6111, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.031119575347461403, | |
| "grad_norm": 1.5591881275177002, | |
| "learning_rate": 0.00019272715891016735, | |
| "loss": 2.7386, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.031281656469062766, | |
| "grad_norm": 1.6340643167495728, | |
| "learning_rate": 0.00019264801467557007, | |
| "loss": 2.628, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.03144373759066413, | |
| "grad_norm": 1.4968132972717285, | |
| "learning_rate": 0.00019256845856812266, | |
| "loss": 2.8749, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.03160581871226549, | |
| "grad_norm": 1.6335184574127197, | |
| "learning_rate": 0.000192488490941497, | |
| "loss": 2.6994, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.03176789983386685, | |
| "grad_norm": 1.96530020236969, | |
| "learning_rate": 0.00019240811215119448, | |
| "loss": 2.7515, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.031929980955468215, | |
| "grad_norm": 1.7235907316207886, | |
| "learning_rate": 0.00019232732255454422, | |
| "loss": 2.7286, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.03209206207706957, | |
| "grad_norm": 2.007321834564209, | |
| "learning_rate": 0.00019224612251070175, | |
| "loss": 2.8945, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.03225414319867093, | |
| "grad_norm": 2.7299695014953613, | |
| "learning_rate": 0.0001921645123806472, | |
| "loss": 3.2715, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.032416224320272295, | |
| "grad_norm": 3.4719176292419434, | |
| "learning_rate": 0.0001920824925271838, | |
| "loss": 3.5109, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03257830544187366, | |
| "grad_norm": 1.11237370967865, | |
| "learning_rate": 0.0001920000633149362, | |
| "loss": 2.4583, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.03274038656347502, | |
| "grad_norm": 1.248706579208374, | |
| "learning_rate": 0.00019191722511034884, | |
| "loss": 2.1951, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.03290246768507638, | |
| "grad_norm": 1.2184851169586182, | |
| "learning_rate": 0.00019183397828168448, | |
| "loss": 2.1751, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.033064548806677745, | |
| "grad_norm": 1.3734302520751953, | |
| "learning_rate": 0.00019175032319902234, | |
| "loss": 2.2204, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.03322662992827911, | |
| "grad_norm": 1.0059350728988647, | |
| "learning_rate": 0.00019166626023425662, | |
| "loss": 2.2202, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.03338871104988046, | |
| "grad_norm": 0.9674267172813416, | |
| "learning_rate": 0.00019158178976109476, | |
| "loss": 2.2254, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.033550792171481825, | |
| "grad_norm": 1.008048415184021, | |
| "learning_rate": 0.0001914969121550558, | |
| "loss": 2.3269, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.03371287329308319, | |
| "grad_norm": 0.890281081199646, | |
| "learning_rate": 0.00019141162779346874, | |
| "loss": 2.0735, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.03387495441468455, | |
| "grad_norm": 0.901919424533844, | |
| "learning_rate": 0.00019132593705547082, | |
| "loss": 2.2881, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.03403703553628591, | |
| "grad_norm": 1.0526825189590454, | |
| "learning_rate": 0.00019123984032200586, | |
| "loss": 2.1952, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.034199116657887274, | |
| "grad_norm": 0.7686760425567627, | |
| "learning_rate": 0.00019115333797582254, | |
| "loss": 2.1207, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.034361197779488636, | |
| "grad_norm": 1.027368426322937, | |
| "learning_rate": 0.00019106643040147278, | |
| "loss": 2.1631, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.03452327890109, | |
| "grad_norm": 1.0761582851409912, | |
| "learning_rate": 0.00019097911798530987, | |
| "loss": 2.2614, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.034685360022691354, | |
| "grad_norm": 0.8369147777557373, | |
| "learning_rate": 0.00019089140111548696, | |
| "loss": 2.1758, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.034847441144292716, | |
| "grad_norm": 0.8233863115310669, | |
| "learning_rate": 0.00019080328018195513, | |
| "loss": 2.1106, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.03500952226589408, | |
| "grad_norm": 1.033157229423523, | |
| "learning_rate": 0.0001907147555764618, | |
| "loss": 2.2292, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.03517160338749544, | |
| "grad_norm": 0.8740055561065674, | |
| "learning_rate": 0.00019062582769254895, | |
| "loss": 2.1657, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.035333684509096804, | |
| "grad_norm": 1.2374004125595093, | |
| "learning_rate": 0.00019053649692555135, | |
| "loss": 2.2105, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.035495765630698166, | |
| "grad_norm": 0.9455053806304932, | |
| "learning_rate": 0.00019044676367259476, | |
| "loss": 2.2436, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.03565784675229953, | |
| "grad_norm": 1.1991380453109741, | |
| "learning_rate": 0.00019035662833259432, | |
| "loss": 2.2907, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.03581992787390089, | |
| "grad_norm": 1.0990439653396606, | |
| "learning_rate": 0.00019026609130625257, | |
| "loss": 2.0842, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.035982008995502246, | |
| "grad_norm": 0.9472472667694092, | |
| "learning_rate": 0.00019017515299605788, | |
| "loss": 2.189, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.03614409011710361, | |
| "grad_norm": 1.0272729396820068, | |
| "learning_rate": 0.00019008381380628247, | |
| "loss": 2.1383, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.03630617123870497, | |
| "grad_norm": 1.0076351165771484, | |
| "learning_rate": 0.00018999207414298067, | |
| "loss": 2.2219, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.03646825236030633, | |
| "grad_norm": 0.9722249507904053, | |
| "learning_rate": 0.00018989993441398726, | |
| "loss": 2.2792, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.036630333481907695, | |
| "grad_norm": 0.9026159048080444, | |
| "learning_rate": 0.00018980739502891546, | |
| "loss": 2.2208, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.03679241460350906, | |
| "grad_norm": 1.2710939645767212, | |
| "learning_rate": 0.0001897144563991552, | |
| "loss": 2.1999, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.03695449572511042, | |
| "grad_norm": 1.0243679285049438, | |
| "learning_rate": 0.00018962111893787128, | |
| "loss": 2.3907, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.03711657684671178, | |
| "grad_norm": 1.4581419229507446, | |
| "learning_rate": 0.00018952738306000151, | |
| "loss": 2.3261, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.03727865796831314, | |
| "grad_norm": 1.0644052028656006, | |
| "learning_rate": 0.00018943324918225494, | |
| "loss": 2.2427, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.0374407390899145, | |
| "grad_norm": 0.9349438548088074, | |
| "learning_rate": 0.0001893387177231099, | |
| "loss": 2.2227, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.03760282021151586, | |
| "grad_norm": 1.7584396600723267, | |
| "learning_rate": 0.0001892437891028122, | |
| "loss": 2.5222, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.037764901333117225, | |
| "grad_norm": 1.1246528625488281, | |
| "learning_rate": 0.0001891484637433733, | |
| "loss": 2.3478, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.03792698245471859, | |
| "grad_norm": 1.7504336833953857, | |
| "learning_rate": 0.00018905274206856837, | |
| "loss": 2.4176, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.03808906357631995, | |
| "grad_norm": 1.7333769798278809, | |
| "learning_rate": 0.00018895662450393438, | |
| "loss": 2.3627, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.03825114469792131, | |
| "grad_norm": 1.1291919946670532, | |
| "learning_rate": 0.00018886011147676833, | |
| "loss": 2.2326, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.038413225819522674, | |
| "grad_norm": 2.3500661849975586, | |
| "learning_rate": 0.00018876320341612522, | |
| "loss": 2.3338, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.03857530694112403, | |
| "grad_norm": 1.9575247764587402, | |
| "learning_rate": 0.00018866590075281624, | |
| "loss": 2.4039, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.03873738806272539, | |
| "grad_norm": 1.1364452838897705, | |
| "learning_rate": 0.00018856820391940674, | |
| "loss": 2.4962, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.038899469184326754, | |
| "grad_norm": 1.8569177389144897, | |
| "learning_rate": 0.00018847011335021449, | |
| "loss": 2.5976, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.03906155030592812, | |
| "grad_norm": 1.4362958669662476, | |
| "learning_rate": 0.00018837162948130752, | |
| "loss": 2.5504, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.03922363142752948, | |
| "grad_norm": 1.2726794481277466, | |
| "learning_rate": 0.00018827275275050233, | |
| "loss": 2.5475, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.03938571254913084, | |
| "grad_norm": 1.394042730331421, | |
| "learning_rate": 0.00018817348359736203, | |
| "loss": 2.6246, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.039547793670732204, | |
| "grad_norm": 1.6995958089828491, | |
| "learning_rate": 0.00018807382246319412, | |
| "loss": 2.6082, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.039709874792333566, | |
| "grad_norm": 1.566512942314148, | |
| "learning_rate": 0.00018797376979104872, | |
| "loss": 2.6337, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.03987195591393492, | |
| "grad_norm": 1.7819331884384155, | |
| "learning_rate": 0.00018787332602571662, | |
| "loss": 2.6822, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.040034037035536284, | |
| "grad_norm": 1.922326922416687, | |
| "learning_rate": 0.00018777249161372713, | |
| "loss": 2.7644, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.040196118157137646, | |
| "grad_norm": 2.2240724563598633, | |
| "learning_rate": 0.00018767126700334634, | |
| "loss": 3.0321, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.04035819927873901, | |
| "grad_norm": 2.2767648696899414, | |
| "learning_rate": 0.0001875696526445749, | |
| "loss": 2.767, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.04052028040034037, | |
| "grad_norm": 3.148064613342285, | |
| "learning_rate": 0.0001874676489891461, | |
| "loss": 3.5881, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.04068236152194173, | |
| "grad_norm": 1.428133487701416, | |
| "learning_rate": 0.00018736525649052394, | |
| "loss": 2.3483, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.040844442643543095, | |
| "grad_norm": 1.0642366409301758, | |
| "learning_rate": 0.00018726247560390099, | |
| "loss": 2.3377, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.04100652376514446, | |
| "grad_norm": 1.277927279472351, | |
| "learning_rate": 0.00018715930678619644, | |
| "loss": 2.3101, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.04116860488674581, | |
| "grad_norm": 1.297062873840332, | |
| "learning_rate": 0.00018705575049605413, | |
| "loss": 2.1857, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.041330686008347176, | |
| "grad_norm": 1.0266833305358887, | |
| "learning_rate": 0.00018695180719384029, | |
| "loss": 2.2796, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.04149276712994854, | |
| "grad_norm": 1.329106330871582, | |
| "learning_rate": 0.00018684747734164177, | |
| "loss": 2.097, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.0416548482515499, | |
| "grad_norm": 1.1051158905029297, | |
| "learning_rate": 0.00018674276140326376, | |
| "loss": 2.1601, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.04181692937315126, | |
| "grad_norm": 1.1395552158355713, | |
| "learning_rate": 0.00018663765984422786, | |
| "loss": 2.2773, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.041979010494752625, | |
| "grad_norm": 1.0176094770431519, | |
| "learning_rate": 0.00018653217313177004, | |
| "loss": 2.1027, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.04214109161635399, | |
| "grad_norm": 1.410070538520813, | |
| "learning_rate": 0.00018642630173483832, | |
| "loss": 2.2349, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.04230317273795535, | |
| "grad_norm": 1.0913872718811035, | |
| "learning_rate": 0.00018632004612409103, | |
| "loss": 2.1952, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.042465253859556705, | |
| "grad_norm": 1.0407989025115967, | |
| "learning_rate": 0.00018621340677189453, | |
| "loss": 2.0673, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.04262733498115807, | |
| "grad_norm": 0.8925177454948425, | |
| "learning_rate": 0.00018610638415232097, | |
| "loss": 2.0325, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.04278941610275943, | |
| "grad_norm": 1.2974311113357544, | |
| "learning_rate": 0.00018599897874114652, | |
| "loss": 2.2241, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.04295149722436079, | |
| "grad_norm": 1.0207972526550293, | |
| "learning_rate": 0.00018589119101584898, | |
| "loss": 2.1243, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.043113578345962154, | |
| "grad_norm": 0.9445371627807617, | |
| "learning_rate": 0.00018578302145560584, | |
| "loss": 2.1291, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.04327565946756352, | |
| "grad_norm": 1.0071361064910889, | |
| "learning_rate": 0.00018567447054129195, | |
| "loss": 2.0553, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.04343774058916488, | |
| "grad_norm": 0.9603691101074219, | |
| "learning_rate": 0.00018556553875547754, | |
| "loss": 2.1355, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.04359982171076624, | |
| "grad_norm": 0.9493493437767029, | |
| "learning_rate": 0.00018545622658242607, | |
| "loss": 2.2426, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.0437619028323676, | |
| "grad_norm": 1.3570773601531982, | |
| "learning_rate": 0.00018534653450809197, | |
| "loss": 2.17, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.04392398395396896, | |
| "grad_norm": 1.1473950147628784, | |
| "learning_rate": 0.00018523646302011867, | |
| "loss": 2.1311, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.04408606507557032, | |
| "grad_norm": 0.9925466775894165, | |
| "learning_rate": 0.00018512601260783606, | |
| "loss": 2.0384, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.044248146197171684, | |
| "grad_norm": 1.0609676837921143, | |
| "learning_rate": 0.00018501518376225887, | |
| "loss": 2.2505, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.044410227318773046, | |
| "grad_norm": 0.9563316702842712, | |
| "learning_rate": 0.00018490397697608395, | |
| "loss": 2.0907, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.04457230844037441, | |
| "grad_norm": 0.998549222946167, | |
| "learning_rate": 0.0001847923927436884, | |
| "loss": 2.1951, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.04473438956197577, | |
| "grad_norm": 0.9470577836036682, | |
| "learning_rate": 0.00018468043156112728, | |
| "loss": 2.1417, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.04489647068357713, | |
| "grad_norm": 1.134791374206543, | |
| "learning_rate": 0.0001845680939261314, | |
| "loss": 2.1117, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.04505855180517849, | |
| "grad_norm": 1.16586434841156, | |
| "learning_rate": 0.00018445538033810515, | |
| "loss": 2.2197, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.04522063292677985, | |
| "grad_norm": 1.1667602062225342, | |
| "learning_rate": 0.00018434229129812418, | |
| "loss": 2.3319, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.04538271404838121, | |
| "grad_norm": 1.3442655801773071, | |
| "learning_rate": 0.0001842288273089332, | |
| "loss": 2.1906, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.045544795169982576, | |
| "grad_norm": 1.522715449333191, | |
| "learning_rate": 0.00018411498887494396, | |
| "loss": 2.2284, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.04570687629158394, | |
| "grad_norm": 1.083176851272583, | |
| "learning_rate": 0.00018400077650223263, | |
| "loss": 2.4033, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.0458689574131853, | |
| "grad_norm": 1.6272213459014893, | |
| "learning_rate": 0.0001838861906985379, | |
| "loss": 2.4303, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.04603103853478666, | |
| "grad_norm": 1.0966017246246338, | |
| "learning_rate": 0.00018377123197325842, | |
| "loss": 2.4413, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.046193119656388025, | |
| "grad_norm": 1.427307367324829, | |
| "learning_rate": 0.00018365590083745085, | |
| "loss": 2.2448, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.04635520077798938, | |
| "grad_norm": 1.1614744663238525, | |
| "learning_rate": 0.00018354019780382735, | |
| "loss": 2.3916, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.04651728189959074, | |
| "grad_norm": 1.4718098640441895, | |
| "learning_rate": 0.0001834241233867533, | |
| "loss": 2.3877, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.046679363021192105, | |
| "grad_norm": 1.2141605615615845, | |
| "learning_rate": 0.00018330767810224524, | |
| "loss": 2.4906, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.04684144414279347, | |
| "grad_norm": 1.296593427658081, | |
| "learning_rate": 0.0001831908624679683, | |
| "loss": 2.4217, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.04700352526439483, | |
| "grad_norm": 1.238234519958496, | |
| "learning_rate": 0.0001830736770032341, | |
| "loss": 2.4124, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.04716560638599619, | |
| "grad_norm": 1.3287264108657837, | |
| "learning_rate": 0.0001829561222289984, | |
| "loss": 2.5919, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.047327687507597554, | |
| "grad_norm": 1.575232744216919, | |
| "learning_rate": 0.00018283819866785853, | |
| "loss": 2.6511, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.04748976862919892, | |
| "grad_norm": 1.5590267181396484, | |
| "learning_rate": 0.0001827199068440516, | |
| "loss": 2.3753, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.04765184975080027, | |
| "grad_norm": 1.3798424005508423, | |
| "learning_rate": 0.00018260124728345162, | |
| "loss": 2.6057, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.047813930872401635, | |
| "grad_norm": 1.6804996728897095, | |
| "learning_rate": 0.00018248222051356754, | |
| "loss": 2.7774, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.047976011994003, | |
| "grad_norm": 1.7314913272857666, | |
| "learning_rate": 0.00018236282706354063, | |
| "loss": 2.603, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.04813809311560436, | |
| "grad_norm": 1.8370392322540283, | |
| "learning_rate": 0.00018224306746414238, | |
| "loss": 2.6394, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.04830017423720572, | |
| "grad_norm": 2.3413655757904053, | |
| "learning_rate": 0.00018212294224777197, | |
| "loss": 2.8592, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.048462255358807084, | |
| "grad_norm": 3.0159950256347656, | |
| "learning_rate": 0.00018200245194845399, | |
| "loss": 3.157, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.048624336480408446, | |
| "grad_norm": 3.672441005706787, | |
| "learning_rate": 0.00018188159710183594, | |
| "loss": 3.2499, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.04878641760200981, | |
| "grad_norm": 1.427798867225647, | |
| "learning_rate": 0.000181760378245186, | |
| "loss": 2.4282, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.048948498723611164, | |
| "grad_norm": 1.2800281047821045, | |
| "learning_rate": 0.00018163879591739067, | |
| "loss": 2.1978, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.049110579845212526, | |
| "grad_norm": 0.8815104961395264, | |
| "learning_rate": 0.0001815168506589521, | |
| "loss": 2.2418, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.04927266096681389, | |
| "grad_norm": 1.0522781610488892, | |
| "learning_rate": 0.000181394543011986, | |
| "loss": 2.0928, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.04943474208841525, | |
| "grad_norm": 1.028746485710144, | |
| "learning_rate": 0.00018127187352021907, | |
| "loss": 2.11, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.04959682321001661, | |
| "grad_norm": 0.9874324798583984, | |
| "learning_rate": 0.0001811488427289866, | |
| "loss": 2.0513, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.049758904331617976, | |
| "grad_norm": 0.8898067474365234, | |
| "learning_rate": 0.00018102545118523007, | |
| "loss": 2.0351, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.04992098545321934, | |
| "grad_norm": 0.834274411201477, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 1.9415, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.0500830665748207, | |
| "grad_norm": 0.928421676158905, | |
| "learning_rate": 0.00018077758803592718, | |
| "loss": 2.1434, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.050245147696422056, | |
| "grad_norm": 0.8722582459449768, | |
| "learning_rate": 0.00018065311753227273, | |
| "loss": 2.0717, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.05040722881802342, | |
| "grad_norm": 0.7962062358856201, | |
| "learning_rate": 0.0001805282884798732, | |
| "loss": 2.0131, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.05056930993962478, | |
| "grad_norm": 0.9616008400917053, | |
| "learning_rate": 0.00018040310143366446, | |
| "loss": 2.0606, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.05073139106122614, | |
| "grad_norm": 0.811673641204834, | |
| "learning_rate": 0.00018027755695017368, | |
| "loss": 2.0475, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.050893472182827505, | |
| "grad_norm": 0.8326351642608643, | |
| "learning_rate": 0.00018015165558751717, | |
| "loss": 2.2088, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.05105555330442887, | |
| "grad_norm": 0.9125694632530212, | |
| "learning_rate": 0.00018002539790539773, | |
| "loss": 1.91, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.05121763442603023, | |
| "grad_norm": 1.0221319198608398, | |
| "learning_rate": 0.00017989878446510215, | |
| "loss": 2.1462, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.05137971554763159, | |
| "grad_norm": 0.8795247077941895, | |
| "learning_rate": 0.00017977181582949888, | |
| "loss": 2.1541, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.05154179666923295, | |
| "grad_norm": 0.8685728311538696, | |
| "learning_rate": 0.0001796444925630353, | |
| "loss": 2.1562, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.05170387779083431, | |
| "grad_norm": 1.1004655361175537, | |
| "learning_rate": 0.00017951681523173542, | |
| "loss": 2.1231, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.05186595891243567, | |
| "grad_norm": 0.9287633299827576, | |
| "learning_rate": 0.0001793887844031972, | |
| "loss": 2.0784, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.052028040034037035, | |
| "grad_norm": 1.0385360717773438, | |
| "learning_rate": 0.00017926040064659014, | |
| "loss": 2.086, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.0521901211556384, | |
| "grad_norm": 0.8999946117401123, | |
| "learning_rate": 0.0001791316645326526, | |
| "loss": 2.0312, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.05235220227723976, | |
| "grad_norm": 1.0374078750610352, | |
| "learning_rate": 0.00017900257663368963, | |
| "loss": 2.1407, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.05251428339884112, | |
| "grad_norm": 1.061728835105896, | |
| "learning_rate": 0.0001788731375235698, | |
| "loss": 2.1352, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.052676364520442484, | |
| "grad_norm": 1.033845067024231, | |
| "learning_rate": 0.00017874334777772327, | |
| "loss": 2.2804, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.05283844564204384, | |
| "grad_norm": 1.2489559650421143, | |
| "learning_rate": 0.00017861320797313892, | |
| "loss": 2.1746, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.0530005267636452, | |
| "grad_norm": 0.988294243812561, | |
| "learning_rate": 0.0001784827186883618, | |
| "loss": 2.1422, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.053162607885246564, | |
| "grad_norm": 1.2897940874099731, | |
| "learning_rate": 0.00017835188050349064, | |
| "loss": 2.0962, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.053324689006847927, | |
| "grad_norm": 1.1834853887557983, | |
| "learning_rate": 0.00017822069400017516, | |
| "loss": 2.2816, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.05348677012844929, | |
| "grad_norm": 1.1888794898986816, | |
| "learning_rate": 0.00017808915976161362, | |
| "loss": 2.0314, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.05364885125005065, | |
| "grad_norm": 1.5024569034576416, | |
| "learning_rate": 0.00017795727837255015, | |
| "loss": 2.1153, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.053810932371652014, | |
| "grad_norm": 1.130858302116394, | |
| "learning_rate": 0.00017782505041927216, | |
| "loss": 2.3989, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.053973013493253376, | |
| "grad_norm": 1.5125319957733154, | |
| "learning_rate": 0.00017769247648960774, | |
| "loss": 2.1236, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.05413509461485474, | |
| "grad_norm": 1.5707167387008667, | |
| "learning_rate": 0.00017755955717292296, | |
| "loss": 2.1928, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.054297175736456094, | |
| "grad_norm": 1.1016305685043335, | |
| "learning_rate": 0.00017742629306011944, | |
| "loss": 2.2474, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.054459256858057456, | |
| "grad_norm": 1.2245185375213623, | |
| "learning_rate": 0.00017729268474363154, | |
| "loss": 2.2744, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.05462133797965882, | |
| "grad_norm": 1.4685486555099487, | |
| "learning_rate": 0.0001771587328174239, | |
| "loss": 2.3349, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.05478341910126018, | |
| "grad_norm": 1.1773123741149902, | |
| "learning_rate": 0.0001770244378769885, | |
| "loss": 2.3933, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.05494550022286154, | |
| "grad_norm": 1.5868879556655884, | |
| "learning_rate": 0.0001768898005193425, | |
| "loss": 2.4543, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.055107581344462905, | |
| "grad_norm": 1.4568393230438232, | |
| "learning_rate": 0.000176754821343025, | |
| "loss": 2.3851, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.05526966246606427, | |
| "grad_norm": 1.3486446142196655, | |
| "learning_rate": 0.0001766195009480949, | |
| "loss": 2.4321, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.05543174358766563, | |
| "grad_norm": 1.5296896696090698, | |
| "learning_rate": 0.0001764838399361279, | |
| "loss": 2.5294, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.055593824709266985, | |
| "grad_norm": 1.6126346588134766, | |
| "learning_rate": 0.00017634783891021393, | |
| "loss": 2.6512, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.05575590583086835, | |
| "grad_norm": 1.5726284980773926, | |
| "learning_rate": 0.00017621149847495458, | |
| "loss": 2.5767, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.05591798695246971, | |
| "grad_norm": 2.3036086559295654, | |
| "learning_rate": 0.00017607481923646016, | |
| "loss": 2.6289, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.05608006807407107, | |
| "grad_norm": 1.6342629194259644, | |
| "learning_rate": 0.0001759378018023473, | |
| "loss": 2.6854, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.056242149195672435, | |
| "grad_norm": 1.9900621175765991, | |
| "learning_rate": 0.00017580044678173592, | |
| "loss": 2.9069, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.0564042303172738, | |
| "grad_norm": 2.33109450340271, | |
| "learning_rate": 0.00017566275478524693, | |
| "loss": 2.5896, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.05656631143887516, | |
| "grad_norm": 2.7142515182495117, | |
| "learning_rate": 0.0001755247264249991, | |
| "loss": 3.0566, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.05672839256047652, | |
| "grad_norm": 3.5597503185272217, | |
| "learning_rate": 0.0001753863623146066, | |
| "loss": 3.4787, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.05689047368207788, | |
| "grad_norm": 1.2078644037246704, | |
| "learning_rate": 0.00017524766306917618, | |
| "loss": 2.2079, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.05705255480367924, | |
| "grad_norm": 0.9320293068885803, | |
| "learning_rate": 0.0001751086293053045, | |
| "loss": 2.2482, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.0572146359252806, | |
| "grad_norm": 1.6044763326644897, | |
| "learning_rate": 0.0001749692616410753, | |
| "loss": 2.1207, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.057376717046881964, | |
| "grad_norm": 0.9895527362823486, | |
| "learning_rate": 0.00017482956069605668, | |
| "loss": 2.1629, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.05753879816848333, | |
| "grad_norm": 0.9529826641082764, | |
| "learning_rate": 0.00017468952709129846, | |
| "loss": 2.1446, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.05770087929008469, | |
| "grad_norm": 1.0963599681854248, | |
| "learning_rate": 0.00017454916144932922, | |
| "loss": 2.0365, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.05786296041168605, | |
| "grad_norm": 0.8234672546386719, | |
| "learning_rate": 0.0001744084643941536, | |
| "loss": 2.0966, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.058025041533287414, | |
| "grad_norm": 0.9184057712554932, | |
| "learning_rate": 0.00017426743655124974, | |
| "loss": 2.0621, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.05818712265488877, | |
| "grad_norm": 0.7979874014854431, | |
| "learning_rate": 0.0001741260785475661, | |
| "loss": 2.0949, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.05834920377649013, | |
| "grad_norm": 1.0196961164474487, | |
| "learning_rate": 0.00017398439101151905, | |
| "loss": 2.0579, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.058511284898091494, | |
| "grad_norm": 0.9489096403121948, | |
| "learning_rate": 0.00017384237457298987, | |
| "loss": 2.2265, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.058673366019692856, | |
| "grad_norm": 0.8029397130012512, | |
| "learning_rate": 0.00017370002986332193, | |
| "loss": 2.087, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.05883544714129422, | |
| "grad_norm": 0.9402124285697937, | |
| "learning_rate": 0.00017355735751531807, | |
| "loss": 2.1656, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.05899752826289558, | |
| "grad_norm": 1.070616364479065, | |
| "learning_rate": 0.00017341435816323756, | |
| "loss": 2.2847, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.05915960938449694, | |
| "grad_norm": 0.8646097183227539, | |
| "learning_rate": 0.00017327103244279348, | |
| "loss": 2.0102, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.059321690506098305, | |
| "grad_norm": 0.9388591647148132, | |
| "learning_rate": 0.00017312738099114973, | |
| "loss": 2.1441, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.05948377162769966, | |
| "grad_norm": 0.9204132556915283, | |
| "learning_rate": 0.00017298340444691835, | |
| "loss": 2.0562, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.05964585274930102, | |
| "grad_norm": 0.9248064160346985, | |
| "learning_rate": 0.00017283910345015647, | |
| "loss": 2.0438, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.059807933870902386, | |
| "grad_norm": 1.0187243223190308, | |
| "learning_rate": 0.0001726944786423637, | |
| "loss": 2.1806, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.05997001499250375, | |
| "grad_norm": 0.9070429801940918, | |
| "learning_rate": 0.00017254953066647913, | |
| "loss": 2.0144, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.06013209611410511, | |
| "grad_norm": 1.2213033437728882, | |
| "learning_rate": 0.00017240426016687863, | |
| "loss": 2.0817, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.06029417723570647, | |
| "grad_norm": 0.9484167695045471, | |
| "learning_rate": 0.00017225866778937165, | |
| "loss": 2.0152, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.060456258357307835, | |
| "grad_norm": 1.308074712753296, | |
| "learning_rate": 0.00017211275418119876, | |
| "loss": 2.1363, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.0606183394789092, | |
| "grad_norm": 1.1076298952102661, | |
| "learning_rate": 0.0001719665199910285, | |
| "loss": 2.0591, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.06078042060051055, | |
| "grad_norm": 1.1952176094055176, | |
| "learning_rate": 0.00017181996586895454, | |
| "loss": 2.0732, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.06078042060051055, | |
| "eval_loss": 2.2569665908813477, | |
| "eval_runtime": 615.1133, | |
| "eval_samples_per_second": 16.894, | |
| "eval_steps_per_second": 8.447, | |
| "step": 375 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 375, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.750698952097792e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |