| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 506, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.003952569169960474, | |
| "grad_norm": 16.020900110158806, | |
| "learning_rate": 9.999903631006022e-06, | |
| "loss": 1.0657, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007905138339920948, | |
| "grad_norm": 9.896813279978835, | |
| "learning_rate": 9.999614527738882e-06, | |
| "loss": 0.8817, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.011857707509881422, | |
| "grad_norm": 4.710814039433313, | |
| "learning_rate": 9.99913270134281e-06, | |
| "loss": 0.6038, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.015810276679841896, | |
| "grad_norm": 3.213510727342975, | |
| "learning_rate": 9.998458170391065e-06, | |
| "loss": 0.5555, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.019762845849802372, | |
| "grad_norm": 3.399601233005852, | |
| "learning_rate": 9.99759096088519e-06, | |
| "loss": 0.4486, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.023715415019762844, | |
| "grad_norm": 3.2249204890399112, | |
| "learning_rate": 9.996531106254027e-06, | |
| "loss": 0.4616, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.02766798418972332, | |
| "grad_norm": 2.5076762498803196, | |
| "learning_rate": 9.995278647352428e-06, | |
| "loss": 0.4705, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.03162055335968379, | |
| "grad_norm": 2.865637105499682, | |
| "learning_rate": 9.993833632459675e-06, | |
| "loss": 0.4065, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03557312252964427, | |
| "grad_norm": 2.182471312496633, | |
| "learning_rate": 9.99219611727762e-06, | |
| "loss": 0.3791, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.039525691699604744, | |
| "grad_norm": 1.8817059063630837, | |
| "learning_rate": 9.990366164928538e-06, | |
| "loss": 0.3968, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.043478260869565216, | |
| "grad_norm": 1.7437964277471723, | |
| "learning_rate": 9.988343845952697e-06, | |
| "loss": 0.3593, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.04743083003952569, | |
| "grad_norm": 1.718975131845364, | |
| "learning_rate": 9.986129238305635e-06, | |
| "loss": 0.3241, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.05138339920948617, | |
| "grad_norm": 1.7519764781357152, | |
| "learning_rate": 9.983722427355157e-06, | |
| "loss": 0.3514, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.05533596837944664, | |
| "grad_norm": 1.8585607479109172, | |
| "learning_rate": 9.98112350587804e-06, | |
| "loss": 0.3307, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.05928853754940711, | |
| "grad_norm": 1.648833246586736, | |
| "learning_rate": 9.978332574056468e-06, | |
| "loss": 0.3382, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06324110671936758, | |
| "grad_norm": 1.5144376829290096, | |
| "learning_rate": 9.975349739474156e-06, | |
| "loss": 0.3025, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.06719367588932806, | |
| "grad_norm": 1.4722946130746835, | |
| "learning_rate": 9.972175117112208e-06, | |
| "loss": 0.3291, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.07114624505928854, | |
| "grad_norm": 1.5590402414570852, | |
| "learning_rate": 9.968808829344692e-06, | |
| "loss": 0.318, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.07509881422924901, | |
| "grad_norm": 1.529364514445946, | |
| "learning_rate": 9.965251005933915e-06, | |
| "loss": 0.3029, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.07905138339920949, | |
| "grad_norm": 1.5355343280276874, | |
| "learning_rate": 9.961501784025423e-06, | |
| "loss": 0.3075, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08300395256916997, | |
| "grad_norm": 1.4284315641862246, | |
| "learning_rate": 9.95756130814271e-06, | |
| "loss": 0.2733, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.08695652173913043, | |
| "grad_norm": 1.529934818238827, | |
| "learning_rate": 9.953429730181653e-06, | |
| "loss": 0.3319, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 1.5594666788219782, | |
| "learning_rate": 9.949107209404664e-06, | |
| "loss": 0.2998, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.09486166007905138, | |
| "grad_norm": 1.43783031647119, | |
| "learning_rate": 9.94459391243453e-06, | |
| "loss": 0.2856, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.09881422924901186, | |
| "grad_norm": 1.6552583911804946, | |
| "learning_rate": 9.939890013248006e-06, | |
| "loss": 0.3287, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.10276679841897234, | |
| "grad_norm": 1.4553289738630346, | |
| "learning_rate": 9.934995693169104e-06, | |
| "loss": 0.2726, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.1067193675889328, | |
| "grad_norm": 1.7057032170116595, | |
| "learning_rate": 9.929911140862109e-06, | |
| "loss": 0.3412, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.11067193675889328, | |
| "grad_norm": 1.4032529471627393, | |
| "learning_rate": 9.924636552324296e-06, | |
| "loss": 0.2484, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.11462450592885376, | |
| "grad_norm": 1.6007146129563619, | |
| "learning_rate": 9.919172130878378e-06, | |
| "loss": 0.3265, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.11857707509881422, | |
| "grad_norm": 1.4391496848114302, | |
| "learning_rate": 9.913518087164678e-06, | |
| "loss": 0.285, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1225296442687747, | |
| "grad_norm": 1.4062720696044961, | |
| "learning_rate": 9.907674639132995e-06, | |
| "loss": 0.2548, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.12648221343873517, | |
| "grad_norm": 1.410271063639066, | |
| "learning_rate": 9.901642012034214e-06, | |
| "loss": 0.2508, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.13043478260869565, | |
| "grad_norm": 1.5527348593856665, | |
| "learning_rate": 9.895420438411616e-06, | |
| "loss": 0.2976, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.13438735177865613, | |
| "grad_norm": 1.321537975619614, | |
| "learning_rate": 9.889010158091917e-06, | |
| "loss": 0.2572, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1383399209486166, | |
| "grad_norm": 1.421996806107464, | |
| "learning_rate": 9.882411418176023e-06, | |
| "loss": 0.2706, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.1422924901185771, | |
| "grad_norm": 1.3572216549078873, | |
| "learning_rate": 9.875624473029508e-06, | |
| "loss": 0.2545, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.14624505928853754, | |
| "grad_norm": 1.2631690700776463, | |
| "learning_rate": 9.8686495842728e-06, | |
| "loss": 0.2625, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.15019762845849802, | |
| "grad_norm": 1.2621066835028363, | |
| "learning_rate": 9.861487020771103e-06, | |
| "loss": 0.2591, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.1541501976284585, | |
| "grad_norm": 1.2949235488783764, | |
| "learning_rate": 9.854137058624034e-06, | |
| "loss": 0.2309, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.15810276679841898, | |
| "grad_norm": 1.3003116841403821, | |
| "learning_rate": 9.846599981154975e-06, | |
| "loss": 0.2326, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.16205533596837945, | |
| "grad_norm": 1.3768166039897276, | |
| "learning_rate": 9.838876078900158e-06, | |
| "loss": 0.2397, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.16600790513833993, | |
| "grad_norm": 1.3538215396932514, | |
| "learning_rate": 9.830965649597455e-06, | |
| "loss": 0.252, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.16996047430830039, | |
| "grad_norm": 1.4163162387664199, | |
| "learning_rate": 9.822868998174914e-06, | |
| "loss": 0.2448, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 1.4257156730091618, | |
| "learning_rate": 9.814586436738998e-06, | |
| "loss": 0.2371, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.17786561264822134, | |
| "grad_norm": 1.4437391682465417, | |
| "learning_rate": 9.806118284562547e-06, | |
| "loss": 0.2907, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 1.391249530608691, | |
| "learning_rate": 9.797464868072489e-06, | |
| "loss": 0.2683, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1857707509881423, | |
| "grad_norm": 1.4016200934541518, | |
| "learning_rate": 9.788626520837235e-06, | |
| "loss": 0.2473, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.18972332015810275, | |
| "grad_norm": 1.279328106626107, | |
| "learning_rate": 9.779603583553842e-06, | |
| "loss": 0.2341, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.19367588932806323, | |
| "grad_norm": 1.3404807365252567, | |
| "learning_rate": 9.770396404034863e-06, | |
| "loss": 0.2675, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.1976284584980237, | |
| "grad_norm": 1.4585279577064512, | |
| "learning_rate": 9.76100533719495e-06, | |
| "loss": 0.2685, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2015810276679842, | |
| "grad_norm": 1.4424775598833863, | |
| "learning_rate": 9.75143074503717e-06, | |
| "loss": 0.2311, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.20553359683794467, | |
| "grad_norm": 1.3955125763500507, | |
| "learning_rate": 9.741672996639046e-06, | |
| "loss": 0.2627, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.20948616600790515, | |
| "grad_norm": 1.4097242793234939, | |
| "learning_rate": 9.731732468138338e-06, | |
| "loss": 0.2512, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.2134387351778656, | |
| "grad_norm": 1.2779409261860988, | |
| "learning_rate": 9.72160954271854e-06, | |
| "loss": 0.248, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.21739130434782608, | |
| "grad_norm": 1.3115141733721187, | |
| "learning_rate": 9.711304610594104e-06, | |
| "loss": 0.2548, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.22134387351778656, | |
| "grad_norm": 1.3040054071646883, | |
| "learning_rate": 9.700818068995407e-06, | |
| "loss": 0.2115, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.22529644268774704, | |
| "grad_norm": 1.3228671690231326, | |
| "learning_rate": 9.69015032215344e-06, | |
| "loss": 0.2372, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.22924901185770752, | |
| "grad_norm": 1.2810295113715335, | |
| "learning_rate": 9.679301781284209e-06, | |
| "loss": 0.2586, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.233201581027668, | |
| "grad_norm": 1.2300547258655599, | |
| "learning_rate": 9.668272864572904e-06, | |
| "loss": 0.2122, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.23715415019762845, | |
| "grad_norm": 1.331764793327734, | |
| "learning_rate": 9.65706399715777e-06, | |
| "loss": 0.2294, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.24110671936758893, | |
| "grad_norm": 1.245135656410948, | |
| "learning_rate": 9.645675611113715e-06, | |
| "loss": 0.2234, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.2450592885375494, | |
| "grad_norm": 1.2821048117267706, | |
| "learning_rate": 9.634108145435665e-06, | |
| "loss": 0.2111, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.2490118577075099, | |
| "grad_norm": 1.3953990850010947, | |
| "learning_rate": 9.62236204602163e-06, | |
| "loss": 0.2423, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.25296442687747034, | |
| "grad_norm": 1.3305442952896998, | |
| "learning_rate": 9.610437765655522e-06, | |
| "loss": 0.2169, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.25691699604743085, | |
| "grad_norm": 1.3427264285897607, | |
| "learning_rate": 9.598335763989703e-06, | |
| "loss": 0.2387, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.2608695652173913, | |
| "grad_norm": 1.2462397253811373, | |
| "learning_rate": 9.586056507527266e-06, | |
| "loss": 0.2002, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.2648221343873518, | |
| "grad_norm": 1.4106601403861017, | |
| "learning_rate": 9.573600469604044e-06, | |
| "loss": 0.2345, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.26877470355731226, | |
| "grad_norm": 1.4558970695841007, | |
| "learning_rate": 9.560968130370376e-06, | |
| "loss": 0.257, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 1.2869342015693788, | |
| "learning_rate": 9.548159976772593e-06, | |
| "loss": 0.2272, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2766798418972332, | |
| "grad_norm": 1.2457392708337578, | |
| "learning_rate": 9.535176502534242e-06, | |
| "loss": 0.2011, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.28063241106719367, | |
| "grad_norm": 1.2710277485877495, | |
| "learning_rate": 9.522018208137066e-06, | |
| "loss": 0.2198, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.2845849802371542, | |
| "grad_norm": 1.4429390388818002, | |
| "learning_rate": 9.508685600801704e-06, | |
| "loss": 0.2277, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.2885375494071146, | |
| "grad_norm": 1.2389843654949855, | |
| "learning_rate": 9.495179194468135e-06, | |
| "loss": 0.2269, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.2924901185770751, | |
| "grad_norm": 1.2313451113400058, | |
| "learning_rate": 9.481499509775878e-06, | |
| "loss": 0.2229, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.2964426877470356, | |
| "grad_norm": 1.2005736870957262, | |
| "learning_rate": 9.467647074043911e-06, | |
| "loss": 0.2132, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.30039525691699603, | |
| "grad_norm": 1.3356666012024943, | |
| "learning_rate": 9.453622421250353e-06, | |
| "loss": 0.2261, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.30434782608695654, | |
| "grad_norm": 1.307750115235254, | |
| "learning_rate": 9.439426092011877e-06, | |
| "loss": 0.2186, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.308300395256917, | |
| "grad_norm": 1.3023411955103703, | |
| "learning_rate": 9.42505863356287e-06, | |
| "loss": 0.2241, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.31225296442687744, | |
| "grad_norm": 1.2340243980259775, | |
| "learning_rate": 9.410520599734338e-06, | |
| "loss": 0.2195, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.31620553359683795, | |
| "grad_norm": 1.2136232872238946, | |
| "learning_rate": 9.395812550932559e-06, | |
| "loss": 0.2105, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.3201581027667984, | |
| "grad_norm": 1.3027913426674749, | |
| "learning_rate": 9.38093505411748e-06, | |
| "loss": 0.2112, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.3241106719367589, | |
| "grad_norm": 1.2355979390019485, | |
| "learning_rate": 9.365888682780862e-06, | |
| "loss": 0.2041, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.32806324110671936, | |
| "grad_norm": 1.3255611172622812, | |
| "learning_rate": 9.35067401692417e-06, | |
| "loss": 0.2133, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.33201581027667987, | |
| "grad_norm": 1.2110730151497104, | |
| "learning_rate": 9.335291643036221e-06, | |
| "loss": 0.1937, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.3359683794466403, | |
| "grad_norm": 1.2302820012512985, | |
| "learning_rate": 9.319742154070578e-06, | |
| "loss": 0.2127, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.33992094861660077, | |
| "grad_norm": 1.396952411441183, | |
| "learning_rate": 9.30402614942268e-06, | |
| "loss": 0.2498, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.3438735177865613, | |
| "grad_norm": 1.4467617558469186, | |
| "learning_rate": 9.288144234906753e-06, | |
| "loss": 0.2582, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 1.2724563579817645, | |
| "learning_rate": 9.272097022732444e-06, | |
| "loss": 0.2094, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.35177865612648224, | |
| "grad_norm": 1.374464706651341, | |
| "learning_rate": 9.255885131481231e-06, | |
| "loss": 0.2333, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.3557312252964427, | |
| "grad_norm": 1.1876568411175543, | |
| "learning_rate": 9.239509186082574e-06, | |
| "loss": 0.1893, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.35968379446640314, | |
| "grad_norm": 1.3268240601065355, | |
| "learning_rate": 9.222969817789829e-06, | |
| "loss": 0.2124, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 1.2681330364031638, | |
| "learning_rate": 9.206267664155906e-06, | |
| "loss": 0.1973, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.3675889328063241, | |
| "grad_norm": 1.2255166053795126, | |
| "learning_rate": 9.189403369008704e-06, | |
| "loss": 0.224, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3715415019762846, | |
| "grad_norm": 1.3042156326130103, | |
| "learning_rate": 9.172377582426286e-06, | |
| "loss": 0.2366, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.37549407114624506, | |
| "grad_norm": 1.2467295856925376, | |
| "learning_rate": 9.155190960711822e-06, | |
| "loss": 0.2307, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.3794466403162055, | |
| "grad_norm": 1.3438520348323209, | |
| "learning_rate": 9.137844166368289e-06, | |
| "loss": 0.2358, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.383399209486166, | |
| "grad_norm": 1.195811544532563, | |
| "learning_rate": 9.120337868072933e-06, | |
| "loss": 0.1998, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.38735177865612647, | |
| "grad_norm": 1.1483129999113106, | |
| "learning_rate": 9.1026727406515e-06, | |
| "loss": 0.1992, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.391304347826087, | |
| "grad_norm": 1.2903993706253556, | |
| "learning_rate": 9.08484946505221e-06, | |
| "loss": 0.2103, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.3952569169960474, | |
| "grad_norm": 1.3546933181924383, | |
| "learning_rate": 9.066868728319522e-06, | |
| "loss": 0.2431, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.39920948616600793, | |
| "grad_norm": 1.2376669346579956, | |
| "learning_rate": 9.048731223567636e-06, | |
| "loss": 0.2112, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.4031620553359684, | |
| "grad_norm": 1.1433249084583228, | |
| "learning_rate": 9.03043764995379e-06, | |
| "loss": 0.187, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.40711462450592883, | |
| "grad_norm": 1.2494316975005797, | |
| "learning_rate": 9.011988712651295e-06, | |
| "loss": 0.2254, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.41106719367588934, | |
| "grad_norm": 1.164269222531622, | |
| "learning_rate": 8.993385122822364e-06, | |
| "loss": 0.196, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.4150197628458498, | |
| "grad_norm": 1.2168120947010408, | |
| "learning_rate": 8.974627597590693e-06, | |
| "loss": 0.1871, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.4189723320158103, | |
| "grad_norm": 1.2068782855519822, | |
| "learning_rate": 8.955716860013812e-06, | |
| "loss": 0.2061, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.42292490118577075, | |
| "grad_norm": 1.217200175520014, | |
| "learning_rate": 8.936653639055225e-06, | |
| "loss": 0.2128, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.4268774703557312, | |
| "grad_norm": 1.1978052117951064, | |
| "learning_rate": 8.917438669556307e-06, | |
| "loss": 0.2154, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.4308300395256917, | |
| "grad_norm": 1.2353347282860097, | |
| "learning_rate": 8.898072692207964e-06, | |
| "loss": 0.2166, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.43478260869565216, | |
| "grad_norm": 1.2843635498058075, | |
| "learning_rate": 8.8785564535221e-06, | |
| "loss": 0.222, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.43873517786561267, | |
| "grad_norm": 1.2399015935864328, | |
| "learning_rate": 8.85889070580283e-06, | |
| "loss": 0.2106, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.4426877470355731, | |
| "grad_norm": 1.2553929591861368, | |
| "learning_rate": 8.839076207117485e-06, | |
| "loss": 0.2428, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.44664031620553357, | |
| "grad_norm": 1.344840565207265, | |
| "learning_rate": 8.819113721267385e-06, | |
| "loss": 0.231, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.4505928853754941, | |
| "grad_norm": 1.1925173566212948, | |
| "learning_rate": 8.7990040177584e-06, | |
| "loss": 0.1862, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 1.2960852273499728, | |
| "learning_rate": 8.778747871771293e-06, | |
| "loss": 0.2015, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.45849802371541504, | |
| "grad_norm": 1.222878498466299, | |
| "learning_rate": 8.758346064131824e-06, | |
| "loss": 0.2153, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.4624505928853755, | |
| "grad_norm": 1.274954724070636, | |
| "learning_rate": 8.737799381280667e-06, | |
| "loss": 0.2027, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.466403162055336, | |
| "grad_norm": 1.387737355454323, | |
| "learning_rate": 8.717108615243081e-06, | |
| "loss": 0.2221, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.47035573122529645, | |
| "grad_norm": 1.3426117595660954, | |
| "learning_rate": 8.696274563598395e-06, | |
| "loss": 0.2255, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.4743083003952569, | |
| "grad_norm": 1.119139769942221, | |
| "learning_rate": 8.675298029449241e-06, | |
| "loss": 0.18, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4782608695652174, | |
| "grad_norm": 1.2709986698183513, | |
| "learning_rate": 8.65417982139062e-06, | |
| "loss": 0.2204, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.48221343873517786, | |
| "grad_norm": 1.3155178470853532, | |
| "learning_rate": 8.63292075347872e-06, | |
| "loss": 0.2098, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.48616600790513836, | |
| "grad_norm": 1.1164093498148737, | |
| "learning_rate": 8.611521645199532e-06, | |
| "loss": 0.1699, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.4901185770750988, | |
| "grad_norm": 1.2232423322705634, | |
| "learning_rate": 8.589983321437271e-06, | |
| "loss": 0.2048, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.49407114624505927, | |
| "grad_norm": 1.19896747867504, | |
| "learning_rate": 8.568306612442579e-06, | |
| "loss": 0.1818, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4980237154150198, | |
| "grad_norm": 1.1935132661062275, | |
| "learning_rate": 8.546492353800504e-06, | |
| "loss": 0.1905, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.5019762845849802, | |
| "grad_norm": 1.3129073388913475, | |
| "learning_rate": 8.524541386398318e-06, | |
| "loss": 0.2249, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.5059288537549407, | |
| "grad_norm": 1.128105255244932, | |
| "learning_rate": 8.502454556393071e-06, | |
| "loss": 0.1853, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.5098814229249012, | |
| "grad_norm": 1.275277474434624, | |
| "learning_rate": 8.480232715179004e-06, | |
| "loss": 0.2268, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.5138339920948617, | |
| "grad_norm": 1.199574317058949, | |
| "learning_rate": 8.457876719354708e-06, | |
| "loss": 0.2159, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5177865612648221, | |
| "grad_norm": 1.333440075825727, | |
| "learning_rate": 8.435387430690114e-06, | |
| "loss": 0.2071, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.5217391304347826, | |
| "grad_norm": 1.2298826627346486, | |
| "learning_rate": 8.412765716093273e-06, | |
| "loss": 0.1978, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.525691699604743, | |
| "grad_norm": 1.1487447920512157, | |
| "learning_rate": 8.390012447576931e-06, | |
| "loss": 0.1681, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.5296442687747036, | |
| "grad_norm": 1.1422378679889824, | |
| "learning_rate": 8.367128502224931e-06, | |
| "loss": 0.1799, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.5335968379446641, | |
| "grad_norm": 1.1642741860659866, | |
| "learning_rate": 8.344114762158391e-06, | |
| "loss": 0.1945, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.5375494071146245, | |
| "grad_norm": 1.3124275732270296, | |
| "learning_rate": 8.320972114501698e-06, | |
| "loss": 0.1994, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.541501976284585, | |
| "grad_norm": 1.1914086199917326, | |
| "learning_rate": 8.297701451348324e-06, | |
| "loss": 0.1982, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 1.251119288035372, | |
| "learning_rate": 8.274303669726427e-06, | |
| "loss": 0.1961, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.549407114624506, | |
| "grad_norm": 1.1421929686539185, | |
| "learning_rate": 8.250779671564277e-06, | |
| "loss": 0.205, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5533596837944664, | |
| "grad_norm": 1.26120508447205, | |
| "learning_rate": 8.22713036365549e-06, | |
| "loss": 0.2056, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5573122529644269, | |
| "grad_norm": 1.1982016744926482, | |
| "learning_rate": 8.20335665762407e-06, | |
| "loss": 0.2152, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5612648221343873, | |
| "grad_norm": 1.2176314854159234, | |
| "learning_rate": 8.179459469889269e-06, | |
| "loss": 0.2154, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5652173913043478, | |
| "grad_norm": 1.2473624074495893, | |
| "learning_rate": 8.155439721630265e-06, | |
| "loss": 0.2175, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.5691699604743083, | |
| "grad_norm": 1.1267383357984246, | |
| "learning_rate": 8.131298338750648e-06, | |
| "loss": 0.1892, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5731225296442688, | |
| "grad_norm": 1.182358551787801, | |
| "learning_rate": 8.10703625184273e-06, | |
| "loss": 0.1972, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5770750988142292, | |
| "grad_norm": 1.2568983709583992, | |
| "learning_rate": 8.082654396151676e-06, | |
| "loss": 0.2156, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.5810276679841897, | |
| "grad_norm": 1.3251313569923084, | |
| "learning_rate": 8.058153711539444e-06, | |
| "loss": 0.2089, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5849802371541502, | |
| "grad_norm": 1.0726734543721594, | |
| "learning_rate": 8.03353514244857e-06, | |
| "loss": 0.1809, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5889328063241107, | |
| "grad_norm": 1.217655834438009, | |
| "learning_rate": 8.008799637865741e-06, | |
| "loss": 0.2023, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.5928853754940712, | |
| "grad_norm": 1.0772407801697832, | |
| "learning_rate": 7.983948151285242e-06, | |
| "loss": 0.1796, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5968379446640316, | |
| "grad_norm": 1.2413074180785202, | |
| "learning_rate": 7.958981640672173e-06, | |
| "loss": 0.1977, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.6007905138339921, | |
| "grad_norm": 1.0852808186316445, | |
| "learning_rate": 7.933901068425539e-06, | |
| "loss": 0.1731, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.6047430830039525, | |
| "grad_norm": 1.299528282255904, | |
| "learning_rate": 7.908707401341146e-06, | |
| "loss": 0.2216, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.6086956521739131, | |
| "grad_norm": 1.185556918204072, | |
| "learning_rate": 7.883401610574338e-06, | |
| "loss": 0.1908, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.6126482213438735, | |
| "grad_norm": 1.304948505750968, | |
| "learning_rate": 7.857984671602547e-06, | |
| "loss": 0.2008, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.616600790513834, | |
| "grad_norm": 1.1900402122684852, | |
| "learning_rate": 7.832457564187715e-06, | |
| "loss": 0.1706, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.6205533596837944, | |
| "grad_norm": 1.210872241558616, | |
| "learning_rate": 7.806821272338504e-06, | |
| "loss": 0.2054, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.6245059288537549, | |
| "grad_norm": 1.1791680850500765, | |
| "learning_rate": 7.781076784272377e-06, | |
| "loss": 0.1911, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.6284584980237155, | |
| "grad_norm": 1.2064915376432837, | |
| "learning_rate": 7.755225092377498e-06, | |
| "loss": 0.2123, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.6324110671936759, | |
| "grad_norm": 1.1788766410507692, | |
| "learning_rate": 7.729267193174483e-06, | |
| "loss": 0.2166, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6363636363636364, | |
| "grad_norm": 1.2252698894797731, | |
| "learning_rate": 7.703204087277989e-06, | |
| "loss": 0.1919, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.6403162055335968, | |
| "grad_norm": 1.349649897365813, | |
| "learning_rate": 7.67703677935813e-06, | |
| "loss": 0.2178, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.6442687747035574, | |
| "grad_norm": 1.1345917440027748, | |
| "learning_rate": 7.650766278101762e-06, | |
| "loss": 0.197, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.6482213438735178, | |
| "grad_norm": 1.287989390837344, | |
| "learning_rate": 7.624393596173598e-06, | |
| "loss": 0.2381, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.6521739130434783, | |
| "grad_norm": 1.1299739332721368, | |
| "learning_rate": 7.597919750177168e-06, | |
| "loss": 0.1808, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.6561264822134387, | |
| "grad_norm": 1.2891189521599526, | |
| "learning_rate": 7.5713457606156335e-06, | |
| "loss": 0.219, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.6600790513833992, | |
| "grad_norm": 1.2643945396009584, | |
| "learning_rate": 7.5446726518524505e-06, | |
| "loss": 0.2049, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.6640316205533597, | |
| "grad_norm": 1.2293140206996043, | |
| "learning_rate": 7.51790145207188e-06, | |
| "loss": 0.2103, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6679841897233202, | |
| "grad_norm": 1.178935960357507, | |
| "learning_rate": 7.4910331932393634e-06, | |
| "loss": 0.199, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6719367588932806, | |
| "grad_norm": 1.2052463331102627, | |
| "learning_rate": 7.464068911061726e-06, | |
| "loss": 0.2096, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6758893280632411, | |
| "grad_norm": 1.1862206001747997, | |
| "learning_rate": 7.437009644947268e-06, | |
| "loss": 0.1987, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.6798418972332015, | |
| "grad_norm": 1.256150663623612, | |
| "learning_rate": 7.40985643796569e-06, | |
| "loss": 0.2198, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.6837944664031621, | |
| "grad_norm": 1.1442878472683295, | |
| "learning_rate": 7.382610336807887e-06, | |
| "loss": 0.1735, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.6877470355731226, | |
| "grad_norm": 1.21934881533331, | |
| "learning_rate": 7.355272391745605e-06, | |
| "loss": 0.201, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.691699604743083, | |
| "grad_norm": 1.2085810780811668, | |
| "learning_rate": 7.327843656590948e-06, | |
| "loss": 0.1997, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 1.1706763346917641, | |
| "learning_rate": 7.300325188655762e-06, | |
| "loss": 0.175, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.6996047430830039, | |
| "grad_norm": 1.3103080637398228, | |
| "learning_rate": 7.2727180487108725e-06, | |
| "loss": 0.2316, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.7035573122529645, | |
| "grad_norm": 1.2414175503064024, | |
| "learning_rate": 7.245023300945203e-06, | |
| "loss": 0.2086, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.7075098814229249, | |
| "grad_norm": 1.1028313396028102, | |
| "learning_rate": 7.217242012924747e-06, | |
| "loss": 0.1614, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.7114624505928854, | |
| "grad_norm": 1.2545708120381063, | |
| "learning_rate": 7.189375255551413e-06, | |
| "loss": 0.2129, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7154150197628458, | |
| "grad_norm": 1.2208082871634096, | |
| "learning_rate": 7.161424103021752e-06, | |
| "loss": 0.186, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.7193675889328063, | |
| "grad_norm": 1.1856323959588346, | |
| "learning_rate": 7.133389632785543e-06, | |
| "loss": 0.18, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.7233201581027668, | |
| "grad_norm": 1.0491684795158043, | |
| "learning_rate": 7.1052729255042645e-06, | |
| "loss": 0.1738, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 1.1602063012667307, | |
| "learning_rate": 7.0770750650094335e-06, | |
| "loss": 0.1888, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.7312252964426877, | |
| "grad_norm": 1.168233348357676, | |
| "learning_rate": 7.048797138260829e-06, | |
| "loss": 0.1938, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.7351778656126482, | |
| "grad_norm": 1.0998419207012584, | |
| "learning_rate": 7.020440235304593e-06, | |
| "loss": 0.1541, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.7391304347826086, | |
| "grad_norm": 1.2094474362406284, | |
| "learning_rate": 6.9920054492312086e-06, | |
| "loss": 0.1863, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.7430830039525692, | |
| "grad_norm": 1.1862926984636035, | |
| "learning_rate": 6.963493876133367e-06, | |
| "loss": 0.1991, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.7470355731225297, | |
| "grad_norm": 1.309657217121461, | |
| "learning_rate": 6.934906615063716e-06, | |
| "loss": 0.2202, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.7509881422924901, | |
| "grad_norm": 1.1164109935733424, | |
| "learning_rate": 6.90624476799249e-06, | |
| "loss": 0.1755, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.7549407114624506, | |
| "grad_norm": 1.0725104657393945, | |
| "learning_rate": 6.8775094397650375e-06, | |
| "loss": 0.1669, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.758893280632411, | |
| "grad_norm": 1.2352816410215464, | |
| "learning_rate": 6.8487017380592266e-06, | |
| "loss": 0.1962, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.7628458498023716, | |
| "grad_norm": 1.145478161105048, | |
| "learning_rate": 6.81982277334275e-06, | |
| "loss": 0.1864, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.766798418972332, | |
| "grad_norm": 1.2512415214036399, | |
| "learning_rate": 6.790873658830321e-06, | |
| "loss": 0.1972, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.7707509881422925, | |
| "grad_norm": 1.2466180949277024, | |
| "learning_rate": 6.761855510440752e-06, | |
| "loss": 0.1817, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7747035573122529, | |
| "grad_norm": 1.181804911384916, | |
| "learning_rate": 6.732769446753954e-06, | |
| "loss": 0.175, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.7786561264822134, | |
| "grad_norm": 1.2910997699602191, | |
| "learning_rate": 6.703616588967804e-06, | |
| "loss": 0.2146, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.782608695652174, | |
| "grad_norm": 1.2156034432115361, | |
| "learning_rate": 6.674398060854931e-06, | |
| "loss": 0.2065, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.7865612648221344, | |
| "grad_norm": 1.2324700294150777, | |
| "learning_rate": 6.645114988719401e-06, | |
| "loss": 0.218, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.7905138339920948, | |
| "grad_norm": 1.1788814940788603, | |
| "learning_rate": 6.615768501353297e-06, | |
| "loss": 0.1911, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7944664031620553, | |
| "grad_norm": 1.239782727484998, | |
| "learning_rate": 6.5863597299932e-06, | |
| "loss": 0.2033, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.7984189723320159, | |
| "grad_norm": 1.2328166491787746, | |
| "learning_rate": 6.5568898082765945e-06, | |
| "loss": 0.1885, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.8023715415019763, | |
| "grad_norm": 1.2257569520585552, | |
| "learning_rate": 6.527359872198166e-06, | |
| "loss": 0.2087, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.8063241106719368, | |
| "grad_norm": 1.090893060994598, | |
| "learning_rate": 6.497771060066008e-06, | |
| "loss": 0.1673, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.8102766798418972, | |
| "grad_norm": 1.262115835491651, | |
| "learning_rate": 6.468124512457743e-06, | |
| "loss": 0.2173, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.8142292490118577, | |
| "grad_norm": 1.1404830620435313, | |
| "learning_rate": 6.4384213721765565e-06, | |
| "loss": 0.1821, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.8181818181818182, | |
| "grad_norm": 1.0789980915654802, | |
| "learning_rate": 6.408662784207149e-06, | |
| "loss": 0.1809, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.8221343873517787, | |
| "grad_norm": 1.1275632208098325, | |
| "learning_rate": 6.378849895671594e-06, | |
| "loss": 0.1807, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.8260869565217391, | |
| "grad_norm": 1.1869280328722958, | |
| "learning_rate": 6.348983855785122e-06, | |
| "loss": 0.193, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.8300395256916996, | |
| "grad_norm": 1.1369831660790293, | |
| "learning_rate": 6.3190658158118205e-06, | |
| "loss": 0.1736, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.83399209486166, | |
| "grad_norm": 1.099689140622585, | |
| "learning_rate": 6.289096929020254e-06, | |
| "loss": 0.1737, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.8379446640316206, | |
| "grad_norm": 1.1149596155598573, | |
| "learning_rate": 6.25907835063901e-06, | |
| "loss": 0.185, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.841897233201581, | |
| "grad_norm": 1.2181136196244295, | |
| "learning_rate": 6.229011237812172e-06, | |
| "loss": 0.2183, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.8458498023715415, | |
| "grad_norm": 1.1288870761802783, | |
| "learning_rate": 6.1988967495547016e-06, | |
| "loss": 0.1705, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.849802371541502, | |
| "grad_norm": 1.2184909621033004, | |
| "learning_rate": 6.168736046707777e-06, | |
| "loss": 0.2087, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.8537549407114624, | |
| "grad_norm": 1.0707635750160447, | |
| "learning_rate": 6.138530291894033e-06, | |
| "loss": 0.1599, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.857707509881423, | |
| "grad_norm": 1.15244187620798, | |
| "learning_rate": 6.108280649472751e-06, | |
| "loss": 0.1977, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.8616600790513834, | |
| "grad_norm": 1.140414407474285, | |
| "learning_rate": 6.0779882854949745e-06, | |
| "loss": 0.1916, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.8656126482213439, | |
| "grad_norm": 1.168184854563085, | |
| "learning_rate": 6.047654367658563e-06, | |
| "loss": 0.1877, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 1.095382559762059, | |
| "learning_rate": 6.0172800652631706e-06, | |
| "loss": 0.1941, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.8735177865612648, | |
| "grad_norm": 1.170558717112789, | |
| "learning_rate": 5.986866549165185e-06, | |
| "loss": 0.1841, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.8774703557312253, | |
| "grad_norm": 1.1273692690940986, | |
| "learning_rate": 5.9564149917325845e-06, | |
| "loss": 0.1776, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.8814229249011858, | |
| "grad_norm": 1.1615247705497285, | |
| "learning_rate": 5.925926566799754e-06, | |
| "loss": 0.1923, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.8853754940711462, | |
| "grad_norm": 1.1310915029362631, | |
| "learning_rate": 5.895402449622226e-06, | |
| "loss": 0.1772, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.8893280632411067, | |
| "grad_norm": 1.168875863428828, | |
| "learning_rate": 5.864843816831388e-06, | |
| "loss": 0.1844, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.8932806324110671, | |
| "grad_norm": 1.1674668972919495, | |
| "learning_rate": 5.8342518463891195e-06, | |
| "loss": 0.1632, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.8972332015810277, | |
| "grad_norm": 1.2880944650610078, | |
| "learning_rate": 5.803627717542386e-06, | |
| "loss": 0.1961, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.9011857707509882, | |
| "grad_norm": 1.224311023303702, | |
| "learning_rate": 5.7729726107777855e-06, | |
| "loss": 0.1884, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.9051383399209486, | |
| "grad_norm": 1.1555445601349188, | |
| "learning_rate": 5.742287707776034e-06, | |
| "loss": 0.1746, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 1.137507280031818, | |
| "learning_rate": 5.711574191366427e-06, | |
| "loss": 0.1767, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.9130434782608695, | |
| "grad_norm": 1.1300350500871745, | |
| "learning_rate": 5.680833245481234e-06, | |
| "loss": 0.1748, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.9169960474308301, | |
| "grad_norm": 1.061749424236297, | |
| "learning_rate": 5.650066055110067e-06, | |
| "loss": 0.1682, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.9209486166007905, | |
| "grad_norm": 1.270168772191801, | |
| "learning_rate": 5.6192738062542e-06, | |
| "loss": 0.1883, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.924901185770751, | |
| "grad_norm": 1.0546305674656924, | |
| "learning_rate": 5.588457685880851e-06, | |
| "loss": 0.1634, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.9288537549407114, | |
| "grad_norm": 1.1513270673921783, | |
| "learning_rate": 5.557618881877428e-06, | |
| "loss": 0.1745, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.932806324110672, | |
| "grad_norm": 1.4466685740671457, | |
| "learning_rate": 5.526758583005736e-06, | |
| "loss": 0.2153, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.9367588932806324, | |
| "grad_norm": 1.2211664220988006, | |
| "learning_rate": 5.495877978856159e-06, | |
| "loss": 0.1962, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.9407114624505929, | |
| "grad_norm": 1.1382037688163558, | |
| "learning_rate": 5.464978259801797e-06, | |
| "loss": 0.1765, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.9446640316205533, | |
| "grad_norm": 1.1600221011599767, | |
| "learning_rate": 5.4340606169525915e-06, | |
| "loss": 0.1948, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.9486166007905138, | |
| "grad_norm": 1.2329095686972222, | |
| "learning_rate": 5.40312624210939e-06, | |
| "loss": 0.1958, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.9525691699604744, | |
| "grad_norm": 1.1077759614799594, | |
| "learning_rate": 5.372176327718029e-06, | |
| "loss": 0.1772, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.9565217391304348, | |
| "grad_norm": 1.3525530316416365, | |
| "learning_rate": 5.341212066823356e-06, | |
| "loss": 0.2068, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.9604743083003953, | |
| "grad_norm": 1.1641895345672653, | |
| "learning_rate": 5.3102346530232365e-06, | |
| "loss": 0.1983, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.9644268774703557, | |
| "grad_norm": 1.210471873881558, | |
| "learning_rate": 5.2792452804225535e-06, | |
| "loss": 0.203, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.9683794466403162, | |
| "grad_norm": 1.1895101061042292, | |
| "learning_rate": 5.248245143587172e-06, | |
| "loss": 0.1838, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.9723320158102767, | |
| "grad_norm": 1.132321900931639, | |
| "learning_rate": 5.2172354374978905e-06, | |
| "loss": 0.1843, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.9762845849802372, | |
| "grad_norm": 1.2024875975928633, | |
| "learning_rate": 5.186217357504382e-06, | |
| "loss": 0.1971, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.9802371541501976, | |
| "grad_norm": 1.1385111787681224, | |
| "learning_rate": 5.155192099279113e-06, | |
| "loss": 0.1581, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.9841897233201581, | |
| "grad_norm": 1.1812364752105893, | |
| "learning_rate": 5.124160858771252e-06, | |
| "loss": 0.1868, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.9881422924901185, | |
| "grad_norm": 1.0663746142171262, | |
| "learning_rate": 5.093124832160569e-06, | |
| "loss": 0.1702, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.9920948616600791, | |
| "grad_norm": 1.0637107177598129, | |
| "learning_rate": 5.06208521581133e-06, | |
| "loss": 0.1758, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.9960474308300395, | |
| "grad_norm": 1.082027772728129, | |
| "learning_rate": 5.0310432062261764e-06, | |
| "loss": 0.1718, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.2004136292445025, | |
| "learning_rate": 5e-06, | |
| "loss": 0.18, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.0039525691699605, | |
| "grad_norm": 1.1438358979908518, | |
| "learning_rate": 4.968956793773825e-06, | |
| "loss": 0.128, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.007905138339921, | |
| "grad_norm": 1.0815724819973256, | |
| "learning_rate": 4.9379147841886715e-06, | |
| "loss": 0.1374, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.0118577075098814, | |
| "grad_norm": 0.9971041298616835, | |
| "learning_rate": 4.906875167839433e-06, | |
| "loss": 0.1295, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.0158102766798418, | |
| "grad_norm": 1.0834575346736253, | |
| "learning_rate": 4.875839141228751e-06, | |
| "loss": 0.1435, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.0197628458498025, | |
| "grad_norm": 1.068216857230524, | |
| "learning_rate": 4.844807900720888e-06, | |
| "loss": 0.1369, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.023715415019763, | |
| "grad_norm": 1.1436718795229075, | |
| "learning_rate": 4.813782642495618e-06, | |
| "loss": 0.1299, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.0276679841897234, | |
| "grad_norm": 1.2041205262898764, | |
| "learning_rate": 4.78276456250211e-06, | |
| "loss": 0.1393, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.0316205533596838, | |
| "grad_norm": 1.3153176210311313, | |
| "learning_rate": 4.75175485641283e-06, | |
| "loss": 0.1587, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.0355731225296443, | |
| "grad_norm": 1.2818237661945875, | |
| "learning_rate": 4.720754719577448e-06, | |
| "loss": 0.1417, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.0395256916996047, | |
| "grad_norm": 1.346590041013075, | |
| "learning_rate": 4.689765346976765e-06, | |
| "loss": 0.1414, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.0434782608695652, | |
| "grad_norm": 1.396103108426267, | |
| "learning_rate": 4.6587879331766465e-06, | |
| "loss": 0.1541, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.0474308300395256, | |
| "grad_norm": 1.3369506628795504, | |
| "learning_rate": 4.627823672281972e-06, | |
| "loss": 0.1566, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.051383399209486, | |
| "grad_norm": 1.2097827541662982, | |
| "learning_rate": 4.596873757890612e-06, | |
| "loss": 0.1381, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.0553359683794465, | |
| "grad_norm": 1.1170762868644442, | |
| "learning_rate": 4.565939383047411e-06, | |
| "loss": 0.1135, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.0592885375494072, | |
| "grad_norm": 1.2233155007588437, | |
| "learning_rate": 4.535021740198202e-06, | |
| "loss": 0.1609, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.0632411067193677, | |
| "grad_norm": 1.2914900494723907, | |
| "learning_rate": 4.504122021143842e-06, | |
| "loss": 0.1454, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.0671936758893281, | |
| "grad_norm": 1.0833968374959027, | |
| "learning_rate": 4.473241416994265e-06, | |
| "loss": 0.1374, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.0711462450592886, | |
| "grad_norm": 1.130910587654628, | |
| "learning_rate": 4.442381118122573e-06, | |
| "loss": 0.1249, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.075098814229249, | |
| "grad_norm": 1.0290406864536354, | |
| "learning_rate": 4.41154231411915e-06, | |
| "loss": 0.1179, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.0790513833992095, | |
| "grad_norm": 1.1351038290482456, | |
| "learning_rate": 4.3807261937458005e-06, | |
| "loss": 0.1446, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.08300395256917, | |
| "grad_norm": 0.9979191858289315, | |
| "learning_rate": 4.349933944889934e-06, | |
| "loss": 0.1085, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.0869565217391304, | |
| "grad_norm": 1.0703887978514581, | |
| "learning_rate": 4.319166754518768e-06, | |
| "loss": 0.1277, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 1.1475934247053179, | |
| "learning_rate": 4.2884258086335755e-06, | |
| "loss": 0.1394, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.0948616600790513, | |
| "grad_norm": 1.1836305673827001, | |
| "learning_rate": 4.257712292223967e-06, | |
| "loss": 0.1342, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.098814229249012, | |
| "grad_norm": 1.2046134166270397, | |
| "learning_rate": 4.227027389222215e-06, | |
| "loss": 0.1344, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.1027667984189724, | |
| "grad_norm": 1.035753316225185, | |
| "learning_rate": 4.196372282457614e-06, | |
| "loss": 0.1127, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.1067193675889329, | |
| "grad_norm": 1.1282169817707273, | |
| "learning_rate": 4.165748153610881e-06, | |
| "loss": 0.1293, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.1106719367588933, | |
| "grad_norm": 1.1372474070307015, | |
| "learning_rate": 4.1351561831686136e-06, | |
| "loss": 0.1332, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.1146245059288538, | |
| "grad_norm": 1.1594526273221435, | |
| "learning_rate": 4.104597550377776e-06, | |
| "loss": 0.1263, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.1185770750988142, | |
| "grad_norm": 1.15152490259451, | |
| "learning_rate": 4.074073433200249e-06, | |
| "loss": 0.1285, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.1225296442687747, | |
| "grad_norm": 1.1540429682296736, | |
| "learning_rate": 4.043585008267418e-06, | |
| "loss": 0.118, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.1264822134387351, | |
| "grad_norm": 1.2879322008831273, | |
| "learning_rate": 4.013133450834818e-06, | |
| "loss": 0.1532, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.1304347826086956, | |
| "grad_norm": 1.021378919711874, | |
| "learning_rate": 3.982719934736832e-06, | |
| "loss": 0.0988, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.1343873517786562, | |
| "grad_norm": 1.1026842967832002, | |
| "learning_rate": 3.95234563234144e-06, | |
| "loss": 0.1148, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.1383399209486167, | |
| "grad_norm": 1.1221443001758493, | |
| "learning_rate": 3.9220117145050254e-06, | |
| "loss": 0.1346, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.1422924901185771, | |
| "grad_norm": 1.179825748135588, | |
| "learning_rate": 3.89171935052725e-06, | |
| "loss": 0.132, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.1462450592885376, | |
| "grad_norm": 1.1145447568885478, | |
| "learning_rate": 3.861469708105969e-06, | |
| "loss": 0.1262, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.150197628458498, | |
| "grad_norm": 1.1805243512710089, | |
| "learning_rate": 3.831263953292225e-06, | |
| "loss": 0.1464, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.1541501976284585, | |
| "grad_norm": 1.1752099012387707, | |
| "learning_rate": 3.8011032504453e-06, | |
| "loss": 0.1303, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.158102766798419, | |
| "grad_norm": 1.1724153479966961, | |
| "learning_rate": 3.7709887621878305e-06, | |
| "loss": 0.1237, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.1620553359683794, | |
| "grad_norm": 1.1312088107686393, | |
| "learning_rate": 3.740921649360991e-06, | |
| "loss": 0.1329, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.1660079051383399, | |
| "grad_norm": 1.2140114398756852, | |
| "learning_rate": 3.710903070979749e-06, | |
| "loss": 0.146, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.1699604743083003, | |
| "grad_norm": 1.177316021638641, | |
| "learning_rate": 3.680934184188182e-06, | |
| "loss": 0.1369, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.1739130434782608, | |
| "grad_norm": 1.0958487706218825, | |
| "learning_rate": 3.6510161442148783e-06, | |
| "loss": 0.1224, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.1778656126482214, | |
| "grad_norm": 1.0391815446060055, | |
| "learning_rate": 3.621150104328407e-06, | |
| "loss": 0.1248, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.1818181818181819, | |
| "grad_norm": 1.0701427010593552, | |
| "learning_rate": 3.5913372157928515e-06, | |
| "loss": 0.1224, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.1857707509881423, | |
| "grad_norm": 1.1848224795154656, | |
| "learning_rate": 3.5615786278234443e-06, | |
| "loss": 0.1491, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.1897233201581028, | |
| "grad_norm": 1.1762535991099512, | |
| "learning_rate": 3.5318754875422588e-06, | |
| "loss": 0.1347, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.1936758893280632, | |
| "grad_norm": 1.0206443853029172, | |
| "learning_rate": 3.5022289399339933e-06, | |
| "loss": 0.1022, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.1976284584980237, | |
| "grad_norm": 1.0357149112196293, | |
| "learning_rate": 3.4726401278018353e-06, | |
| "loss": 0.1134, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.2015810276679841, | |
| "grad_norm": 1.156607546118926, | |
| "learning_rate": 3.443110191723407e-06, | |
| "loss": 0.127, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.2055335968379446, | |
| "grad_norm": 1.1452610773145862, | |
| "learning_rate": 3.4136402700068034e-06, | |
| "loss": 0.1242, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.2094861660079053, | |
| "grad_norm": 1.137660713683979, | |
| "learning_rate": 3.384231498646706e-06, | |
| "loss": 0.1302, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.2134387351778657, | |
| "grad_norm": 1.3751968299311201, | |
| "learning_rate": 3.3548850112805985e-06, | |
| "loss": 0.1345, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.2173913043478262, | |
| "grad_norm": 1.1820163169421647, | |
| "learning_rate": 3.3256019391450696e-06, | |
| "loss": 0.124, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.2213438735177866, | |
| "grad_norm": 1.254836523224082, | |
| "learning_rate": 3.296383411032198e-06, | |
| "loss": 0.1592, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.225296442687747, | |
| "grad_norm": 1.2092831479934754, | |
| "learning_rate": 3.267230553246047e-06, | |
| "loss": 0.1374, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.2292490118577075, | |
| "grad_norm": 1.3298953359466184, | |
| "learning_rate": 3.2381444895592483e-06, | |
| "loss": 0.1599, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.233201581027668, | |
| "grad_norm": 1.15714090782168, | |
| "learning_rate": 3.209126341169681e-06, | |
| "loss": 0.1178, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.2371541501976284, | |
| "grad_norm": 1.1321197195627186, | |
| "learning_rate": 3.180177226657251e-06, | |
| "loss": 0.1324, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.2411067193675889, | |
| "grad_norm": 1.1427780558349572, | |
| "learning_rate": 3.151298261940775e-06, | |
| "loss": 0.1354, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.2450592885375493, | |
| "grad_norm": 1.1792578676088221, | |
| "learning_rate": 3.122490560234964e-06, | |
| "loss": 0.1316, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.2490118577075098, | |
| "grad_norm": 1.173771729377104, | |
| "learning_rate": 3.0937552320075116e-06, | |
| "loss": 0.1237, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.2529644268774702, | |
| "grad_norm": 1.2142503247487184, | |
| "learning_rate": 3.065093384936285e-06, | |
| "loss": 0.1385, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.256916996047431, | |
| "grad_norm": 1.2169127173091283, | |
| "learning_rate": 3.0365061238666336e-06, | |
| "loss": 0.1442, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.2608695652173914, | |
| "grad_norm": 1.3055869261396447, | |
| "learning_rate": 3.007994550768793e-06, | |
| "loss": 0.136, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.2648221343873518, | |
| "grad_norm": 1.2043193143210125, | |
| "learning_rate": 2.979559764695409e-06, | |
| "loss": 0.1447, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.2687747035573123, | |
| "grad_norm": 1.1233412613298097, | |
| "learning_rate": 2.951202861739173e-06, | |
| "loss": 0.1282, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 1.0606810788891305, | |
| "learning_rate": 2.9229249349905686e-06, | |
| "loss": 0.1212, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.2766798418972332, | |
| "grad_norm": 1.0460756961226252, | |
| "learning_rate": 2.8947270744957385e-06, | |
| "loss": 0.1173, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.2806324110671936, | |
| "grad_norm": 1.1897624975044134, | |
| "learning_rate": 2.8666103672144597e-06, | |
| "loss": 0.1407, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.2845849802371543, | |
| "grad_norm": 1.1942264596916479, | |
| "learning_rate": 2.8385758969782507e-06, | |
| "loss": 0.1286, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.2885375494071147, | |
| "grad_norm": 1.1341245622744474, | |
| "learning_rate": 2.810624744448588e-06, | |
| "loss": 0.128, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.2924901185770752, | |
| "grad_norm": 1.165964008709128, | |
| "learning_rate": 2.7827579870752542e-06, | |
| "loss": 0.136, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.2964426877470356, | |
| "grad_norm": 1.1358629236681395, | |
| "learning_rate": 2.7549766990547973e-06, | |
| "loss": 0.1238, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.300395256916996, | |
| "grad_norm": 1.144473644830131, | |
| "learning_rate": 2.727281951289128e-06, | |
| "loss": 0.1333, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.3043478260869565, | |
| "grad_norm": 1.1761434599454856, | |
| "learning_rate": 2.6996748113442397e-06, | |
| "loss": 0.1223, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.308300395256917, | |
| "grad_norm": 1.1788613585620644, | |
| "learning_rate": 2.672156343409053e-06, | |
| "loss": 0.1349, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.3122529644268774, | |
| "grad_norm": 1.1047408820558797, | |
| "learning_rate": 2.644727608254396e-06, | |
| "loss": 0.1215, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.316205533596838, | |
| "grad_norm": 1.1915355492120698, | |
| "learning_rate": 2.6173896631921134e-06, | |
| "loss": 0.1405, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.3201581027667983, | |
| "grad_norm": 1.1396002338772575, | |
| "learning_rate": 2.590143562034312e-06, | |
| "loss": 0.1297, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.3241106719367588, | |
| "grad_norm": 0.9928956463128177, | |
| "learning_rate": 2.5629903550527343e-06, | |
| "loss": 0.1032, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.3280632411067192, | |
| "grad_norm": 1.1497535091692899, | |
| "learning_rate": 2.535931088938274e-06, | |
| "loss": 0.1342, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.33201581027668, | |
| "grad_norm": 1.1209952848082338, | |
| "learning_rate": 2.5089668067606365e-06, | |
| "loss": 0.1185, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.3359683794466404, | |
| "grad_norm": 1.0839706555945572, | |
| "learning_rate": 2.4820985479281184e-06, | |
| "loss": 0.1139, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.3399209486166008, | |
| "grad_norm": 1.1440761155145645, | |
| "learning_rate": 2.45532734814755e-06, | |
| "loss": 0.134, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.3438735177865613, | |
| "grad_norm": 1.082616882918915, | |
| "learning_rate": 2.4286542393843665e-06, | |
| "loss": 0.1281, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.3478260869565217, | |
| "grad_norm": 1.17500675790559, | |
| "learning_rate": 2.4020802498228333e-06, | |
| "loss": 0.1432, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.3517786561264822, | |
| "grad_norm": 1.1182114019109513, | |
| "learning_rate": 2.3756064038264033e-06, | |
| "loss": 0.134, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.3557312252964426, | |
| "grad_norm": 1.1566712806713801, | |
| "learning_rate": 2.3492337218982396e-06, | |
| "loss": 0.1285, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.359683794466403, | |
| "grad_norm": 1.0164245424014764, | |
| "learning_rate": 2.3229632206418727e-06, | |
| "loss": 0.1055, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.3636363636363638, | |
| "grad_norm": 1.1202273103032814, | |
| "learning_rate": 2.296795912722014e-06, | |
| "loss": 0.1164, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.3675889328063242, | |
| "grad_norm": 1.2025482829810825, | |
| "learning_rate": 2.270732806825517e-06, | |
| "loss": 0.1413, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.3715415019762847, | |
| "grad_norm": 1.1369886395664646, | |
| "learning_rate": 2.244774907622504e-06, | |
| "loss": 0.1312, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.3754940711462451, | |
| "grad_norm": 1.1150013437618065, | |
| "learning_rate": 2.2189232157276247e-06, | |
| "loss": 0.1379, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.3794466403162056, | |
| "grad_norm": 1.086272977870767, | |
| "learning_rate": 2.1931787276614968e-06, | |
| "loss": 0.1178, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.383399209486166, | |
| "grad_norm": 1.1640308203458913, | |
| "learning_rate": 2.167542435812286e-06, | |
| "loss": 0.1356, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.3873517786561265, | |
| "grad_norm": 1.1121490307629227, | |
| "learning_rate": 2.142015328397454e-06, | |
| "loss": 0.1237, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 1.1574528436812335, | |
| "learning_rate": 2.1165983894256647e-06, | |
| "loss": 0.1255, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.3952569169960474, | |
| "grad_norm": 1.361980674009939, | |
| "learning_rate": 2.0912925986588547e-06, | |
| "loss": 0.1389, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.3992094861660078, | |
| "grad_norm": 1.051150691395027, | |
| "learning_rate": 2.0660989315744624e-06, | |
| "loss": 0.1168, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.4031620553359683, | |
| "grad_norm": 1.1382181091816, | |
| "learning_rate": 2.0410183593278287e-06, | |
| "loss": 0.1367, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.4071146245059287, | |
| "grad_norm": 1.127194102052322, | |
| "learning_rate": 2.016051848714758e-06, | |
| "loss": 0.131, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.4110671936758894, | |
| "grad_norm": 1.3646197569162608, | |
| "learning_rate": 1.991200362134258e-06, | |
| "loss": 0.1668, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.4150197628458498, | |
| "grad_norm": 1.179231150635698, | |
| "learning_rate": 1.9664648575514316e-06, | |
| "loss": 0.1309, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.4189723320158103, | |
| "grad_norm": 1.1478284079295227, | |
| "learning_rate": 1.9418462884605555e-06, | |
| "loss": 0.1407, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.4229249011857708, | |
| "grad_norm": 0.966600715786351, | |
| "learning_rate": 1.9173456038483244e-06, | |
| "loss": 0.0989, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.4268774703557312, | |
| "grad_norm": 1.089318761705495, | |
| "learning_rate": 1.8929637481572715e-06, | |
| "loss": 0.1233, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.4308300395256917, | |
| "grad_norm": 1.1956595892466026, | |
| "learning_rate": 1.8687016612493542e-06, | |
| "loss": 0.1266, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.434782608695652, | |
| "grad_norm": 1.199815800783593, | |
| "learning_rate": 1.8445602783697375e-06, | |
| "loss": 0.1383, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.4387351778656128, | |
| "grad_norm": 1.0596031990819328, | |
| "learning_rate": 1.8205405301107343e-06, | |
| "loss": 0.1183, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.4426877470355732, | |
| "grad_norm": 1.1937639817958505, | |
| "learning_rate": 1.7966433423759327e-06, | |
| "loss": 0.1452, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.4466403162055337, | |
| "grad_norm": 1.0906265845240057, | |
| "learning_rate": 1.772869636344512e-06, | |
| "loss": 0.111, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.4505928853754941, | |
| "grad_norm": 1.2231633736793504, | |
| "learning_rate": 1.7492203284357245e-06, | |
| "loss": 0.1362, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 1.0619383271235998, | |
| "learning_rate": 1.7256963302735752e-06, | |
| "loss": 0.1202, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.458498023715415, | |
| "grad_norm": 1.0900241430901978, | |
| "learning_rate": 1.702298548651678e-06, | |
| "loss": 0.1162, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.4624505928853755, | |
| "grad_norm": 1.1512284006660594, | |
| "learning_rate": 1.6790278854983033e-06, | |
| "loss": 0.1322, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.466403162055336, | |
| "grad_norm": 1.1329200558876649, | |
| "learning_rate": 1.6558852378416113e-06, | |
| "loss": 0.1362, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.4703557312252964, | |
| "grad_norm": 1.0410869979854105, | |
| "learning_rate": 1.6328714977750698e-06, | |
| "loss": 0.1239, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.4743083003952568, | |
| "grad_norm": 1.2545052203915814, | |
| "learning_rate": 1.6099875524230707e-06, | |
| "loss": 0.1476, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.4782608695652173, | |
| "grad_norm": 1.2003692839751088, | |
| "learning_rate": 1.5872342839067305e-06, | |
| "loss": 0.1381, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.4822134387351777, | |
| "grad_norm": 1.1739098836348705, | |
| "learning_rate": 1.5646125693098863e-06, | |
| "loss": 0.1295, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.4861660079051384, | |
| "grad_norm": 1.1535530949440382, | |
| "learning_rate": 1.542123280645292e-06, | |
| "loss": 0.1234, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.4901185770750989, | |
| "grad_norm": 1.1168208484387032, | |
| "learning_rate": 1.519767284820996e-06, | |
| "loss": 0.1152, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.4940711462450593, | |
| "grad_norm": 1.0340813124452886, | |
| "learning_rate": 1.4975454436069292e-06, | |
| "loss": 0.1114, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.4980237154150198, | |
| "grad_norm": 1.117157676559404, | |
| "learning_rate": 1.4754586136016841e-06, | |
| "loss": 0.127, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.5019762845849802, | |
| "grad_norm": 1.1411823176863023, | |
| "learning_rate": 1.4535076461994974e-06, | |
| "loss": 0.1243, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.5059288537549407, | |
| "grad_norm": 1.0074369184939824, | |
| "learning_rate": 1.431693387557424e-06, | |
| "loss": 0.1007, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.5098814229249014, | |
| "grad_norm": 1.1674952879116636, | |
| "learning_rate": 1.4100166785627301e-06, | |
| "loss": 0.1201, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.5138339920948618, | |
| "grad_norm": 1.0610397050784564, | |
| "learning_rate": 1.3884783548004704e-06, | |
| "loss": 0.1208, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.5177865612648223, | |
| "grad_norm": 1.1703142856070294, | |
| "learning_rate": 1.3670792465212828e-06, | |
| "loss": 0.131, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.5217391304347827, | |
| "grad_norm": 1.24844816636085, | |
| "learning_rate": 1.3458201786093795e-06, | |
| "loss": 0.1426, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.5256916996047432, | |
| "grad_norm": 1.0289708643260822, | |
| "learning_rate": 1.3247019705507596e-06, | |
| "loss": 0.107, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.5296442687747036, | |
| "grad_norm": 1.0411982481948332, | |
| "learning_rate": 1.3037254364016068e-06, | |
| "loss": 0.1106, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.533596837944664, | |
| "grad_norm": 1.0887512215886415, | |
| "learning_rate": 1.2828913847569185e-06, | |
| "loss": 0.126, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.5375494071146245, | |
| "grad_norm": 1.1038917846986824, | |
| "learning_rate": 1.2622006187193348e-06, | |
| "loss": 0.1169, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.541501976284585, | |
| "grad_norm": 1.1650275815326414, | |
| "learning_rate": 1.2416539358681772e-06, | |
| "loss": 0.1243, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.5454545454545454, | |
| "grad_norm": 1.126326384634314, | |
| "learning_rate": 1.2212521282287093e-06, | |
| "loss": 0.129, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.5494071146245059, | |
| "grad_norm": 1.129180074421009, | |
| "learning_rate": 1.2009959822416012e-06, | |
| "loss": 0.1382, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.5533596837944663, | |
| "grad_norm": 1.164913976767265, | |
| "learning_rate": 1.1808862787326176e-06, | |
| "loss": 0.1217, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.5573122529644268, | |
| "grad_norm": 1.1842645447436517, | |
| "learning_rate": 1.1609237928825174e-06, | |
| "loss": 0.1392, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.5612648221343872, | |
| "grad_norm": 1.127758214888184, | |
| "learning_rate": 1.1411092941971702e-06, | |
| "loss": 0.1334, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.5652173913043477, | |
| "grad_norm": 1.0268028668010976, | |
| "learning_rate": 1.1214435464779006e-06, | |
| "loss": 0.1092, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.5691699604743083, | |
| "grad_norm": 1.1298748698800432, | |
| "learning_rate": 1.1019273077920366e-06, | |
| "loss": 0.1199, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.5731225296442688, | |
| "grad_norm": 1.151983299642192, | |
| "learning_rate": 1.0825613304436938e-06, | |
| "loss": 0.1214, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.5770750988142292, | |
| "grad_norm": 1.056563475272345, | |
| "learning_rate": 1.0633463609447753e-06, | |
| "loss": 0.1137, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.5810276679841897, | |
| "grad_norm": 1.0401127604187286, | |
| "learning_rate": 1.0442831399861903e-06, | |
| "loss": 0.1116, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.5849802371541502, | |
| "grad_norm": 1.137297843112467, | |
| "learning_rate": 1.0253724024093103e-06, | |
| "loss": 0.1259, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.5889328063241108, | |
| "grad_norm": 1.2731646482391143, | |
| "learning_rate": 1.006614877177638e-06, | |
| "loss": 0.1398, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.5928853754940713, | |
| "grad_norm": 1.2031755255565708, | |
| "learning_rate": 9.880112873487068e-07, | |
| "loss": 0.1435, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.5968379446640317, | |
| "grad_norm": 1.0673733506453118, | |
| "learning_rate": 9.695623500462114e-07, | |
| "loss": 0.1213, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.6007905138339922, | |
| "grad_norm": 1.1540390417696156, | |
| "learning_rate": 9.512687764323647e-07, | |
| "loss": 0.1309, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.6047430830039526, | |
| "grad_norm": 1.1719558092176683, | |
| "learning_rate": 9.331312716804791e-07, | |
| "loss": 0.1353, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.608695652173913, | |
| "grad_norm": 1.1328079202728492, | |
| "learning_rate": 9.151505349477901e-07, | |
| "loss": 0.1126, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.6126482213438735, | |
| "grad_norm": 1.1668524122475088, | |
| "learning_rate": 8.973272593485011e-07, | |
| "loss": 0.1232, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.616600790513834, | |
| "grad_norm": 1.1765999211223719, | |
| "learning_rate": 8.796621319270676e-07, | |
| "loss": 0.1381, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.6205533596837944, | |
| "grad_norm": 1.1580704897137293, | |
| "learning_rate": 8.621558336317132e-07, | |
| "loss": 0.125, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.6245059288537549, | |
| "grad_norm": 1.2221405236185294, | |
| "learning_rate": 8.448090392881797e-07, | |
| "loss": 0.1434, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.6284584980237153, | |
| "grad_norm": 1.1484711907465028, | |
| "learning_rate": 8.276224175737152e-07, | |
| "loss": 0.1341, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.6324110671936758, | |
| "grad_norm": 1.0668615052922243, | |
| "learning_rate": 8.105966309912966e-07, | |
| "loss": 0.119, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.6363636363636362, | |
| "grad_norm": 1.1708621908812695, | |
| "learning_rate": 7.937323358440935e-07, | |
| "loss": 0.1365, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.6403162055335967, | |
| "grad_norm": 1.0037003731577279, | |
| "learning_rate": 7.770301822101712e-07, | |
| "loss": 0.1056, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.6442687747035574, | |
| "grad_norm": 1.0648681490157104, | |
| "learning_rate": 7.604908139174255e-07, | |
| "loss": 0.1098, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.6482213438735178, | |
| "grad_norm": 1.1995736066849585, | |
| "learning_rate": 7.441148685187694e-07, | |
| "loss": 0.1341, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.6521739130434783, | |
| "grad_norm": 1.1848136469443982, | |
| "learning_rate": 7.279029772675572e-07, | |
| "loss": 0.1336, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.6561264822134387, | |
| "grad_norm": 1.1795612477078212, | |
| "learning_rate": 7.11855765093249e-07, | |
| "loss": 0.1379, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.6600790513833992, | |
| "grad_norm": 1.244369267153096, | |
| "learning_rate": 6.959738505773211e-07, | |
| "loss": 0.1373, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.6640316205533598, | |
| "grad_norm": 1.148156272898155, | |
| "learning_rate": 6.802578459294235e-07, | |
| "loss": 0.1274, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.6679841897233203, | |
| "grad_norm": 1.1618407598954121, | |
| "learning_rate": 6.647083569637797e-07, | |
| "loss": 0.1208, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.6719367588932808, | |
| "grad_norm": 1.1228370521334565, | |
| "learning_rate": 6.493259830758325e-07, | |
| "loss": 0.1181, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.6758893280632412, | |
| "grad_norm": 1.1659039027502425, | |
| "learning_rate": 6.341113172191399e-07, | |
| "loss": 0.1251, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.6798418972332017, | |
| "grad_norm": 1.1667891678872329, | |
| "learning_rate": 6.190649458825204e-07, | |
| "loss": 0.1373, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.683794466403162, | |
| "grad_norm": 1.1054656728972119, | |
| "learning_rate": 6.041874490674416e-07, | |
| "loss": 0.1183, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.6877470355731226, | |
| "grad_norm": 1.1318157363661536, | |
| "learning_rate": 5.894794002656628e-07, | |
| "loss": 0.1276, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.691699604743083, | |
| "grad_norm": 1.1427617467212088, | |
| "learning_rate": 5.749413664371312e-07, | |
| "loss": 0.1314, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.6956521739130435, | |
| "grad_norm": 1.123370168432176, | |
| "learning_rate": 5.60573907988124e-07, | |
| "loss": 0.1172, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.699604743083004, | |
| "grad_norm": 1.2640707026191205, | |
| "learning_rate": 5.463775787496484e-07, | |
| "loss": 0.1345, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.7035573122529644, | |
| "grad_norm": 1.2270465900901621, | |
| "learning_rate": 5.323529259560911e-07, | |
| "loss": 0.1337, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.7075098814229248, | |
| "grad_norm": 1.0714382150531283, | |
| "learning_rate": 5.185004902241241e-07, | |
| "loss": 0.1125, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.7114624505928853, | |
| "grad_norm": 1.0825327963083162, | |
| "learning_rate": 5.04820805531866e-07, | |
| "loss": 0.1059, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.7154150197628457, | |
| "grad_norm": 1.0596003222945678, | |
| "learning_rate": 4.91314399198296e-07, | |
| "loss": 0.0999, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.7193675889328062, | |
| "grad_norm": 1.1990608827420846, | |
| "learning_rate": 4.779817918629326e-07, | |
| "loss": 0.1257, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.7233201581027668, | |
| "grad_norm": 1.0887979080170394, | |
| "learning_rate": 4.6482349746575783e-07, | |
| "loss": 0.1236, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.7272727272727273, | |
| "grad_norm": 1.0412689170684666, | |
| "learning_rate": 4.5184002322740784e-07, | |
| "loss": 0.1066, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.7312252964426877, | |
| "grad_norm": 1.0802724381125226, | |
| "learning_rate": 4.390318696296247e-07, | |
| "loss": 0.1149, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.7351778656126482, | |
| "grad_norm": 1.1551936194230803, | |
| "learning_rate": 4.2639953039595725e-07, | |
| "loss": 0.135, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 1.2135177448970054, | |
| "learning_rate": 4.139434924727359e-07, | |
| "loss": 0.1349, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.7430830039525693, | |
| "grad_norm": 1.0507122616854452, | |
| "learning_rate": 4.0166423601029735e-07, | |
| "loss": 0.1157, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.7470355731225298, | |
| "grad_norm": 1.1753401932840621, | |
| "learning_rate": 3.8956223434447936e-07, | |
| "loss": 0.136, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.7509881422924902, | |
| "grad_norm": 1.1404040824302084, | |
| "learning_rate": 3.776379539783709e-07, | |
| "loss": 0.1223, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.7549407114624507, | |
| "grad_norm": 1.2777386537799145, | |
| "learning_rate": 3.658918545643353e-07, | |
| "loss": 0.1485, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.7588932806324111, | |
| "grad_norm": 1.1413712412237624, | |
| "learning_rate": 3.543243888862841e-07, | |
| "loss": 0.1212, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.7628458498023716, | |
| "grad_norm": 1.0872960731651602, | |
| "learning_rate": 3.429360028422307e-07, | |
| "loss": 0.1193, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.766798418972332, | |
| "grad_norm": 1.0637737003753145, | |
| "learning_rate": 3.317271354270968e-07, | |
| "loss": 0.1133, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.7707509881422925, | |
| "grad_norm": 1.1840091720746755, | |
| "learning_rate": 3.2069821871579255e-07, | |
| "loss": 0.1333, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.774703557312253, | |
| "grad_norm": 1.1255230443713273, | |
| "learning_rate": 3.098496778465621e-07, | |
| "loss": 0.1181, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.7786561264822134, | |
| "grad_norm": 1.0775241906777941, | |
| "learning_rate": 2.991819310045929e-07, | |
| "loss": 0.1189, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.7826086956521738, | |
| "grad_norm": 1.0583221665971685, | |
| "learning_rate": 2.88695389405898e-07, | |
| "loss": 0.1103, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.7865612648221343, | |
| "grad_norm": 1.1242814725320212, | |
| "learning_rate": 2.783904572814622e-07, | |
| "loss": 0.1285, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.7905138339920947, | |
| "grad_norm": 1.1779401797110727, | |
| "learning_rate": 2.682675318616618e-07, | |
| "loss": 0.1294, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.7944664031620552, | |
| "grad_norm": 1.0859491545785036, | |
| "learning_rate": 2.583270033609536e-07, | |
| "loss": 0.1123, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.7984189723320159, | |
| "grad_norm": 1.0763948870282039, | |
| "learning_rate": 2.4856925496283045e-07, | |
| "loss": 0.1104, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.8023715415019763, | |
| "grad_norm": 1.0887108595666282, | |
| "learning_rate": 2.3899466280504936e-07, | |
| "loss": 0.1126, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.8063241106719368, | |
| "grad_norm": 1.1717149496208563, | |
| "learning_rate": 2.2960359596513714e-07, | |
| "loss": 0.1351, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.8102766798418972, | |
| "grad_norm": 1.229561030964994, | |
| "learning_rate": 2.203964164461597e-07, | |
| "loss": 0.1369, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.8142292490118577, | |
| "grad_norm": 1.1633472228135062, | |
| "learning_rate": 2.113734791627664e-07, | |
| "loss": 0.1291, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 1.2035274009151151, | |
| "learning_rate": 2.0253513192751374e-07, | |
| "loss": 0.1469, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.8221343873517788, | |
| "grad_norm": 1.1836270092696015, | |
| "learning_rate": 1.9388171543745394e-07, | |
| "loss": 0.1379, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.8260869565217392, | |
| "grad_norm": 1.0758870256725546, | |
| "learning_rate": 1.8541356326100436e-07, | |
| "loss": 0.1117, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.8300395256916997, | |
| "grad_norm": 1.1793736035903457, | |
| "learning_rate": 1.7713100182508604e-07, | |
| "loss": 0.1398, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.8339920948616601, | |
| "grad_norm": 1.0801974661120426, | |
| "learning_rate": 1.6903435040254545e-07, | |
| "loss": 0.1128, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.8379446640316206, | |
| "grad_norm": 1.07339663507577, | |
| "learning_rate": 1.6112392109984386e-07, | |
| "loss": 0.1099, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.841897233201581, | |
| "grad_norm": 1.1120046775263792, | |
| "learning_rate": 1.5340001884502577e-07, | |
| "loss": 0.1262, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.8458498023715415, | |
| "grad_norm": 1.0831243721534856, | |
| "learning_rate": 1.4586294137596768e-07, | |
| "loss": 0.1193, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.849802371541502, | |
| "grad_norm": 1.0472977237699574, | |
| "learning_rate": 1.385129792288986e-07, | |
| "loss": 0.1041, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.8537549407114624, | |
| "grad_norm": 1.1482729228107225, | |
| "learning_rate": 1.313504157272022e-07, | |
| "loss": 0.1301, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.8577075098814229, | |
| "grad_norm": 1.1397420372115987, | |
| "learning_rate": 1.2437552697049327e-07, | |
| "loss": 0.1289, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.8616600790513833, | |
| "grad_norm": 1.1569650321244938, | |
| "learning_rate": 1.1758858182397692e-07, | |
| "loss": 0.1283, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.8656126482213438, | |
| "grad_norm": 1.1694788716855629, | |
| "learning_rate": 1.1098984190808403e-07, | |
| "loss": 0.1335, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.8695652173913042, | |
| "grad_norm": 1.0086960578539177, | |
| "learning_rate": 1.0457956158838545e-07, | |
| "loss": 0.0996, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.8735177865612647, | |
| "grad_norm": 1.0308075869001883, | |
| "learning_rate": 9.835798796578755e-08, | |
| "loss": 0.1153, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.8774703557312253, | |
| "grad_norm": 1.170279753009744, | |
| "learning_rate": 9.232536086700605e-08, | |
| "loss": 0.1304, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.8814229249011858, | |
| "grad_norm": 1.0870287190560486, | |
| "learning_rate": 8.648191283532337e-08, | |
| "loss": 0.1213, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.8853754940711462, | |
| "grad_norm": 1.1165962048020976, | |
| "learning_rate": 8.082786912162243e-08, | |
| "loss": 0.1161, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.8893280632411067, | |
| "grad_norm": 1.1494916541958282, | |
| "learning_rate": 7.536344767570536e-08, | |
| "loss": 0.1141, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.8932806324110671, | |
| "grad_norm": 1.0640073229737435, | |
| "learning_rate": 7.008885913789066e-08, | |
| "loss": 0.1228, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.8972332015810278, | |
| "grad_norm": 1.1870546140040668, | |
| "learning_rate": 6.500430683089532e-08, | |
| "loss": 0.1303, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.9011857707509883, | |
| "grad_norm": 1.173615256902204, | |
| "learning_rate": 6.010998675199554e-08, | |
| "loss": 0.1338, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.9051383399209487, | |
| "grad_norm": 1.1958445869792897, | |
| "learning_rate": 5.5406087565471054e-08, | |
| "loss": 0.1405, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.9090909090909092, | |
| "grad_norm": 1.1122683154170514, | |
| "learning_rate": 5.089279059533658e-08, | |
| "loss": 0.1281, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.9130434782608696, | |
| "grad_norm": 1.0958024070105987, | |
| "learning_rate": 4.657026981834623e-08, | |
| "loss": 0.12, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.91699604743083, | |
| "grad_norm": 1.1024888588089012, | |
| "learning_rate": 4.2438691857292215e-08, | |
| "loss": 0.1144, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.9209486166007905, | |
| "grad_norm": 1.123264444499075, | |
| "learning_rate": 3.849821597457892e-08, | |
| "loss": 0.1347, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.924901185770751, | |
| "grad_norm": 1.0765658056683685, | |
| "learning_rate": 3.474899406608501e-08, | |
| "loss": 0.1129, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.9288537549407114, | |
| "grad_norm": 1.148979953908725, | |
| "learning_rate": 3.119117065530808e-08, | |
| "loss": 0.123, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.9328063241106719, | |
| "grad_norm": 1.2501363554379483, | |
| "learning_rate": 2.7824882887793058e-08, | |
| "loss": 0.1538, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.9367588932806323, | |
| "grad_norm": 1.196112067372065, | |
| "learning_rate": 2.4650260525846404e-08, | |
| "loss": 0.1416, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.9407114624505928, | |
| "grad_norm": 1.1128332896345188, | |
| "learning_rate": 2.1667425943532884e-08, | |
| "loss": 0.1147, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.9446640316205532, | |
| "grad_norm": 1.1016376226716715, | |
| "learning_rate": 1.8876494121959908e-08, | |
| "loss": 0.1289, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.9486166007905137, | |
| "grad_norm": 1.135611367054567, | |
| "learning_rate": 1.627757264484442e-08, | |
| "loss": 0.1294, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 1.9525691699604744, | |
| "grad_norm": 1.1398251434326172, | |
| "learning_rate": 1.387076169436563e-08, | |
| "loss": 0.1262, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 1.9565217391304348, | |
| "grad_norm": 1.0846986101599616, | |
| "learning_rate": 1.1656154047303691e-08, | |
| "loss": 0.1126, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.9604743083003953, | |
| "grad_norm": 1.2370763808268233, | |
| "learning_rate": 9.633835071463094e-09, | |
| "loss": 0.1485, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 1.9644268774703557, | |
| "grad_norm": 1.1863239612132057, | |
| "learning_rate": 7.803882722381417e-09, | |
| "loss": 0.1293, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 1.9683794466403162, | |
| "grad_norm": 1.2282535432480546, | |
| "learning_rate": 6.166367540325624e-09, | |
| "loss": 0.1428, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 1.9723320158102768, | |
| "grad_norm": 1.0276030484998475, | |
| "learning_rate": 4.721352647572564e-09, | |
| "loss": 0.1129, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 1.9762845849802373, | |
| "grad_norm": 1.2463780542335487, | |
| "learning_rate": 3.4688937459737004e-09, | |
| "loss": 0.1394, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.9762845849802373, | |
| "eval_loss": 0.16894643008708954, | |
| "eval_runtime": 3.7534, | |
| "eval_samples_per_second": 5.595, | |
| "eval_steps_per_second": 1.599, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.9802371541501977, | |
| "grad_norm": 1.149104137588847, | |
| "learning_rate": 2.4090391148112734e-09, | |
| "loss": 0.126, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 1.9841897233201582, | |
| "grad_norm": 1.0644722994856588, | |
| "learning_rate": 1.5418296089358964e-09, | |
| "loss": 0.1161, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.9881422924901186, | |
| "grad_norm": 1.1078146584096278, | |
| "learning_rate": 8.672986571894859e-10, | |
| "loss": 0.1159, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 1.992094861660079, | |
| "grad_norm": 1.0306240076981967, | |
| "learning_rate": 3.854722611201789e-10, | |
| "loss": 0.1143, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 1.9960474308300395, | |
| "grad_norm": 1.150741799871838, | |
| "learning_rate": 9.636899397813537e-11, | |
| "loss": 0.1237, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.0328104166457073, | |
| "learning_rate": 0.0, | |
| "loss": 0.1028, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 506, | |
| "total_flos": 6736634707968.0, | |
| "train_loss": 0.1787011340083812, | |
| "train_runtime": 1032.8943, | |
| "train_samples_per_second": 3.913, | |
| "train_steps_per_second": 0.49 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 506, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 70000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6736634707968.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |