| { | |
| "best_metric": 0.9264618754386902, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-100", | |
| "epoch": 3.0, | |
| "eval_steps": 50, | |
| "global_step": 141, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02127659574468085, | |
| "grad_norm": 14.213656425476074, | |
| "learning_rate": 1.1200000000000001e-05, | |
| "loss": 4.4616, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02127659574468085, | |
| "eval_loss": 3.8836147785186768, | |
| "eval_runtime": 3.7874, | |
| "eval_samples_per_second": 167.135, | |
| "eval_steps_per_second": 5.281, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0425531914893617, | |
| "grad_norm": 7.082957744598389, | |
| "learning_rate": 2.2400000000000002e-05, | |
| "loss": 3.8433, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.06382978723404255, | |
| "grad_norm": 5.922336578369141, | |
| "learning_rate": 3.36e-05, | |
| "loss": 3.5573, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0851063829787234, | |
| "grad_norm": 5.772343158721924, | |
| "learning_rate": 4.4800000000000005e-05, | |
| "loss": 3.5562, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.10638297872340426, | |
| "grad_norm": 6.365505695343018, | |
| "learning_rate": 5.6e-05, | |
| "loss": 3.1005, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.1276595744680851, | |
| "grad_norm": 7.662172794342041, | |
| "learning_rate": 6.72e-05, | |
| "loss": 2.6307, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.14893617021276595, | |
| "grad_norm": 7.855373382568359, | |
| "learning_rate": 7.84e-05, | |
| "loss": 2.1008, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1702127659574468, | |
| "grad_norm": 5.307789325714111, | |
| "learning_rate": 8.960000000000001e-05, | |
| "loss": 1.8141, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.19148936170212766, | |
| "grad_norm": 2.865940570831299, | |
| "learning_rate": 0.0001008, | |
| "loss": 1.8787, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.2127659574468085, | |
| "grad_norm": 1.4380123615264893, | |
| "learning_rate": 0.000112, | |
| "loss": 1.4918, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.23404255319148937, | |
| "grad_norm": 1.5105291604995728, | |
| "learning_rate": 0.00011198389746057678, | |
| "loss": 1.4308, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.2553191489361702, | |
| "grad_norm": 1.0955153703689575, | |
| "learning_rate": 0.0001119355991027277, | |
| "loss": 1.1133, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.2765957446808511, | |
| "grad_norm": 8.190502166748047, | |
| "learning_rate": 0.00011185513270238886, | |
| "loss": 0.7666, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2978723404255319, | |
| "grad_norm": 12.864703178405762, | |
| "learning_rate": 0.00011174254453503828, | |
| "loss": 0.6613, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.3191489361702128, | |
| "grad_norm": 2.111825466156006, | |
| "learning_rate": 0.00011159789934908318, | |
| "loss": 1.6219, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.3404255319148936, | |
| "grad_norm": 1.2697428464889526, | |
| "learning_rate": 0.00011142128032862395, | |
| "loss": 1.4151, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.3617021276595745, | |
| "grad_norm": 0.5486758351325989, | |
| "learning_rate": 0.00011121278904561553, | |
| "loss": 1.2228, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3829787234042553, | |
| "grad_norm": 0.6482474207878113, | |
| "learning_rate": 0.0001109725454014545, | |
| "loss": 1.1537, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.40425531914893614, | |
| "grad_norm": 1.133864402770996, | |
| "learning_rate": 0.00011070068755802486, | |
| "loss": 0.7857, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.425531914893617, | |
| "grad_norm": 0.4991963803768158, | |
| "learning_rate": 0.00011039737185824234, | |
| "loss": 0.1432, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.44680851063829785, | |
| "grad_norm": 0.7302205562591553, | |
| "learning_rate": 0.00011006277273614316, | |
| "loss": 1.3804, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.46808510638297873, | |
| "grad_norm": 0.6448777914047241, | |
| "learning_rate": 0.00010969708261656854, | |
| "loss": 1.4363, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.48936170212765956, | |
| "grad_norm": 0.6198523044586182, | |
| "learning_rate": 0.00010930051180450303, | |
| "loss": 1.2917, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.5106382978723404, | |
| "grad_norm": 0.40335163474082947, | |
| "learning_rate": 0.00010887328836413005, | |
| "loss": 1.2051, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.5319148936170213, | |
| "grad_norm": 0.5545335412025452, | |
| "learning_rate": 0.00010841565798767423, | |
| "loss": 0.9898, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.5531914893617021, | |
| "grad_norm": 1.7060941457748413, | |
| "learning_rate": 0.00010792788385410628, | |
| "loss": 0.0614, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.574468085106383, | |
| "grad_norm": 0.7902534008026123, | |
| "learning_rate": 0.00010741024647779101, | |
| "loss": 1.128, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.5957446808510638, | |
| "grad_norm": 0.44602009654045105, | |
| "learning_rate": 0.00010686304354716622, | |
| "loss": 1.4464, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.6170212765957447, | |
| "grad_norm": 0.3824542164802551, | |
| "learning_rate": 0.00010628658975354491, | |
| "loss": 1.2856, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.6382978723404256, | |
| "grad_norm": 0.42575424909591675, | |
| "learning_rate": 0.00010568121661013911, | |
| "loss": 1.134, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6595744680851063, | |
| "grad_norm": 0.506963312625885, | |
| "learning_rate": 0.00010504727226140981, | |
| "loss": 0.9792, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.6808510638297872, | |
| "grad_norm": 0.46332627534866333, | |
| "learning_rate": 0.00010438512128285228, | |
| "loss": 0.3159, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.7021276595744681, | |
| "grad_norm": 0.4355514943599701, | |
| "learning_rate": 0.00010369514447133208, | |
| "loss": 0.8945, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.723404255319149, | |
| "grad_norm": 0.4772055447101593, | |
| "learning_rate": 0.0001029777386260924, | |
| "loss": 1.4357, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.7446808510638298, | |
| "grad_norm": 0.3987302780151367, | |
| "learning_rate": 0.00010223331632055843, | |
| "loss": 1.3066, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.7659574468085106, | |
| "grad_norm": 0.37436139583587646, | |
| "learning_rate": 0.00010146230566507025, | |
| "loss": 1.1248, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.7872340425531915, | |
| "grad_norm": 0.4493124783039093, | |
| "learning_rate": 0.00010066515006068056, | |
| "loss": 1.0106, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.8085106382978723, | |
| "grad_norm": 0.8829596042633057, | |
| "learning_rate": 9.984230794415887e-05, | |
| "loss": 0.4913, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.8297872340425532, | |
| "grad_norm": 0.5549487471580505, | |
| "learning_rate": 9.899425252434878e-05, | |
| "loss": 0.6699, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.851063829787234, | |
| "grad_norm": 0.4452535808086395, | |
| "learning_rate": 9.812147151002993e-05, | |
| "loss": 1.4244, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.8723404255319149, | |
| "grad_norm": 0.35962700843811035, | |
| "learning_rate": 9.722446682944128e-05, | |
| "loss": 1.2756, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.8936170212765957, | |
| "grad_norm": 0.3556835949420929, | |
| "learning_rate": 9.630375434162683e-05, | |
| "loss": 1.1866, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.9148936170212766, | |
| "grad_norm": 0.34359875321388245, | |
| "learning_rate": 9.53598635397699e-05, | |
| "loss": 1.0799, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.9361702127659575, | |
| "grad_norm": 0.5090391039848328, | |
| "learning_rate": 9.43933372466865e-05, | |
| "loss": 0.7165, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.9574468085106383, | |
| "grad_norm": 0.42997509241104126, | |
| "learning_rate": 9.340473130265294e-05, | |
| "loss": 1.0119, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.9787234042553191, | |
| "grad_norm": 0.3640575408935547, | |
| "learning_rate": 9.239461424574742e-05, | |
| "loss": 1.2632, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.3492790460586548, | |
| "learning_rate": 9.136356698488885e-05, | |
| "loss": 0.9953, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.0212765957446808, | |
| "grad_norm": 0.3192073702812195, | |
| "learning_rate": 9.031218246576172e-05, | |
| "loss": 0.21, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.0425531914893618, | |
| "grad_norm": 0.434736043214798, | |
| "learning_rate": 8.924106532981847e-05, | |
| "loss": 1.1353, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.0638297872340425, | |
| "grad_norm": 0.36542627215385437, | |
| "learning_rate": 8.815083156655581e-05, | |
| "loss": 1.3234, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.0638297872340425, | |
| "eval_loss": 0.9605554938316345, | |
| "eval_runtime": 3.8841, | |
| "eval_samples_per_second": 162.971, | |
| "eval_steps_per_second": 5.149, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.0851063829787233, | |
| "grad_norm": 0.3345516622066498, | |
| "learning_rate": 8.704210815926495e-05, | |
| "loss": 1.1487, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.1063829787234043, | |
| "grad_norm": 0.39309221506118774, | |
| "learning_rate": 8.59155327244593e-05, | |
| "loss": 1.0582, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.127659574468085, | |
| "grad_norm": 0.5331593751907349, | |
| "learning_rate": 8.477175314518714e-05, | |
| "loss": 0.8832, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.148936170212766, | |
| "grad_norm": 0.8910200595855713, | |
| "learning_rate": 8.361142719844015e-05, | |
| "loss": 0.2747, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.1702127659574468, | |
| "grad_norm": 0.35534268617630005, | |
| "learning_rate": 8.243522217687193e-05, | |
| "loss": 0.7217, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.1914893617021276, | |
| "grad_norm": 0.44460806250572205, | |
| "learning_rate": 8.124381450504426e-05, | |
| "loss": 1.3437, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.2127659574468086, | |
| "grad_norm": 0.37915581464767456, | |
| "learning_rate": 8.00378893504216e-05, | |
| "loss": 1.2378, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.2340425531914894, | |
| "grad_norm": 0.3455619513988495, | |
| "learning_rate": 7.881814022933765e-05, | |
| "loss": 1.1035, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.2553191489361701, | |
| "grad_norm": 0.39131176471710205, | |
| "learning_rate": 7.758526860816059e-05, | |
| "loss": 0.9421, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.2765957446808511, | |
| "grad_norm": 0.3872637450695038, | |
| "learning_rate": 7.633998349988623e-05, | |
| "loss": 0.4359, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.297872340425532, | |
| "grad_norm": 0.3637562692165375, | |
| "learning_rate": 7.508300105639138e-05, | |
| "loss": 0.498, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.3191489361702127, | |
| "grad_norm": 0.4627838730812073, | |
| "learning_rate": 7.381504415658137e-05, | |
| "loss": 1.3362, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.3404255319148937, | |
| "grad_norm": 0.40288859605789185, | |
| "learning_rate": 7.253684199066931e-05, | |
| "loss": 1.2646, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.3617021276595744, | |
| "grad_norm": 0.3752979636192322, | |
| "learning_rate": 7.124912964082547e-05, | |
| "loss": 1.0772, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.3829787234042552, | |
| "grad_norm": 0.4062426686286926, | |
| "learning_rate": 6.995264765843836e-05, | |
| "loss": 1.0008, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.4042553191489362, | |
| "grad_norm": 0.4707888662815094, | |
| "learning_rate": 6.86481416382306e-05, | |
| "loss": 0.6325, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.425531914893617, | |
| "grad_norm": 0.28356093168258667, | |
| "learning_rate": 6.733636178947425e-05, | |
| "loss": 0.2474, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.4468085106382977, | |
| "grad_norm": 0.44394785165786743, | |
| "learning_rate": 6.601806250455254e-05, | |
| "loss": 1.2399, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.4680851063829787, | |
| "grad_norm": 0.3906940519809723, | |
| "learning_rate": 6.469400192511568e-05, | |
| "loss": 1.2571, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.4893617021276595, | |
| "grad_norm": 0.40481507778167725, | |
| "learning_rate": 6.33649415060808e-05, | |
| "loss": 1.1472, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.5106382978723403, | |
| "grad_norm": 0.3902008533477783, | |
| "learning_rate": 6.203164557772622e-05, | |
| "loss": 1.0084, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.5319148936170213, | |
| "grad_norm": 0.47971320152282715, | |
| "learning_rate": 6.069488090613228e-05, | |
| "loss": 0.8643, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.5531914893617023, | |
| "grad_norm": 0.35613465309143066, | |
| "learning_rate": 5.935541625222126e-05, | |
| "loss": 0.072, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.574468085106383, | |
| "grad_norm": 0.4089168310165405, | |
| "learning_rate": 5.801402192965016e-05, | |
| "loss": 1.0347, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.5957446808510638, | |
| "grad_norm": 0.4405708909034729, | |
| "learning_rate": 5.667146936181042e-05, | |
| "loss": 1.2854, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.6170212765957448, | |
| "grad_norm": 0.38711073994636536, | |
| "learning_rate": 5.53285306381896e-05, | |
| "loss": 1.2008, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.6382978723404256, | |
| "grad_norm": 0.3995616137981415, | |
| "learning_rate": 5.398597807034986e-05, | |
| "loss": 1.0411, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.6595744680851063, | |
| "grad_norm": 0.4823427200317383, | |
| "learning_rate": 5.2644583747778746e-05, | |
| "loss": 0.8773, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.6808510638297873, | |
| "grad_norm": 0.27052074670791626, | |
| "learning_rate": 5.130511909386772e-05, | |
| "loss": 0.2128, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.702127659574468, | |
| "grad_norm": 0.34951841831207275, | |
| "learning_rate": 4.996835442227378e-05, | |
| "loss": 0.6826, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.7234042553191489, | |
| "grad_norm": 0.44672834873199463, | |
| "learning_rate": 4.863505849391921e-05, | |
| "loss": 1.3443, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.7446808510638299, | |
| "grad_norm": 0.37677329778671265, | |
| "learning_rate": 4.7305998074884325e-05, | |
| "loss": 1.1532, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.7659574468085106, | |
| "grad_norm": 0.3660067319869995, | |
| "learning_rate": 4.598193749544746e-05, | |
| "loss": 1.0569, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.7872340425531914, | |
| "grad_norm": 0.4302046597003937, | |
| "learning_rate": 4.466363821052573e-05, | |
| "loss": 0.9091, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.8085106382978724, | |
| "grad_norm": 0.3698153793811798, | |
| "learning_rate": 4.335185836176942e-05, | |
| "loss": 0.4112, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.8297872340425532, | |
| "grad_norm": 0.27242642641067505, | |
| "learning_rate": 4.2047352341561654e-05, | |
| "loss": 0.4239, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.851063829787234, | |
| "grad_norm": 0.463748574256897, | |
| "learning_rate": 4.0750870359174544e-05, | |
| "loss": 1.3113, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.872340425531915, | |
| "grad_norm": 0.3822973966598511, | |
| "learning_rate": 3.946315800933069e-05, | |
| "loss": 1.2033, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.8936170212765957, | |
| "grad_norm": 0.3759160041809082, | |
| "learning_rate": 3.8184955843418635e-05, | |
| "loss": 1.1545, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.9148936170212765, | |
| "grad_norm": 0.3939042091369629, | |
| "learning_rate": 3.691699894360862e-05, | |
| "loss": 0.9759, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.9361702127659575, | |
| "grad_norm": 0.463613361120224, | |
| "learning_rate": 3.5660016500113756e-05, | |
| "loss": 0.6793, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.9574468085106385, | |
| "grad_norm": 0.39532244205474854, | |
| "learning_rate": 3.441473139183941e-05, | |
| "loss": 1.0509, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.978723404255319, | |
| "grad_norm": 0.37965285778045654, | |
| "learning_rate": 3.3181859770662366e-05, | |
| "loss": 1.1558, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.4093703627586365, | |
| "learning_rate": 3.196211064957841e-05, | |
| "loss": 0.9204, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.021276595744681, | |
| "grad_norm": 0.11672822386026382, | |
| "learning_rate": 3.075618549495574e-05, | |
| "loss": 0.0652, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.0425531914893615, | |
| "grad_norm": 0.4452710747718811, | |
| "learning_rate": 2.9564777823128087e-05, | |
| "loss": 1.0477, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.0638297872340425, | |
| "grad_norm": 0.4406222999095917, | |
| "learning_rate": 2.8388572801559853e-05, | |
| "loss": 1.2642, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.0851063829787235, | |
| "grad_norm": 0.3580534756183624, | |
| "learning_rate": 2.7228246854812867e-05, | |
| "loss": 1.0798, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 2.106382978723404, | |
| "grad_norm": 0.3822851777076721, | |
| "learning_rate": 2.60844672755407e-05, | |
| "loss": 1.0079, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 2.127659574468085, | |
| "grad_norm": 0.4569602608680725, | |
| "learning_rate": 2.4957891840735056e-05, | |
| "loss": 0.8454, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.127659574468085, | |
| "eval_loss": 0.9264618754386902, | |
| "eval_runtime": 3.7503, | |
| "eval_samples_per_second": 168.787, | |
| "eval_steps_per_second": 5.333, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.148936170212766, | |
| "grad_norm": 0.26321274042129517, | |
| "learning_rate": 2.384916843344419e-05, | |
| "loss": 0.2123, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 2.1702127659574466, | |
| "grad_norm": 0.36982113122940063, | |
| "learning_rate": 2.275893467018154e-05, | |
| "loss": 0.7511, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 2.1914893617021276, | |
| "grad_norm": 0.4143376648426056, | |
| "learning_rate": 2.1687817534238292e-05, | |
| "loss": 1.2891, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 2.2127659574468086, | |
| "grad_norm": 0.3768226206302643, | |
| "learning_rate": 2.0636433015111154e-05, | |
| "loss": 1.127, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 2.2340425531914896, | |
| "grad_norm": 0.3875614106655121, | |
| "learning_rate": 1.9605385754252593e-05, | |
| "loss": 0.9696, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.25531914893617, | |
| "grad_norm": 0.4567450284957886, | |
| "learning_rate": 1.8595268697347047e-05, | |
| "loss": 0.8626, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 2.276595744680851, | |
| "grad_norm": 0.43263593316078186, | |
| "learning_rate": 1.76066627533135e-05, | |
| "loss": 0.417, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 2.297872340425532, | |
| "grad_norm": 0.2629643380641937, | |
| "learning_rate": 1.664013646023009e-05, | |
| "loss": 0.3379, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 2.3191489361702127, | |
| "grad_norm": 0.47793322801589966, | |
| "learning_rate": 1.5696245658373157e-05, | |
| "loss": 1.1889, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 2.3404255319148937, | |
| "grad_norm": 0.4138708710670471, | |
| "learning_rate": 1.4775533170558723e-05, | |
| "loss": 1.1401, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.3617021276595747, | |
| "grad_norm": 0.4172796905040741, | |
| "learning_rate": 1.3878528489970085e-05, | |
| "loss": 1.0396, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 2.382978723404255, | |
| "grad_norm": 0.4409525692462921, | |
| "learning_rate": 1.3005747475651238e-05, | |
| "loss": 0.9267, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.404255319148936, | |
| "grad_norm": 0.49562060832977295, | |
| "learning_rate": 1.2157692055841128e-05, | |
| "loss": 0.6103, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.425531914893617, | |
| "grad_norm": 0.16902895271778107, | |
| "learning_rate": 1.1334849939319436e-05, | |
| "loss": 0.1601, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.4468085106382977, | |
| "grad_norm": 0.5262559652328491, | |
| "learning_rate": 1.0537694334929756e-05, | |
| "loss": 1.2164, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.4680851063829787, | |
| "grad_norm": 0.45780274271965027, | |
| "learning_rate": 9.766683679441566e-06, | |
| "loss": 1.2017, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.4893617021276597, | |
| "grad_norm": 0.4124818444252014, | |
| "learning_rate": 9.022261373907599e-06, | |
| "loss": 1.0688, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.5106382978723403, | |
| "grad_norm": 0.41770249605178833, | |
| "learning_rate": 8.304855528667915e-06, | |
| "loss": 0.893, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.5319148936170213, | |
| "grad_norm": 0.5359745025634766, | |
| "learning_rate": 7.614878717147731e-06, | |
| "loss": 0.734, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.5531914893617023, | |
| "grad_norm": 0.12150562554597855, | |
| "learning_rate": 6.952727738590198e-06, | |
| "loss": 0.0908, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.574468085106383, | |
| "grad_norm": 0.45060330629348755, | |
| "learning_rate": 6.318783389860888e-06, | |
| "loss": 0.9969, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.595744680851064, | |
| "grad_norm": 0.4552464485168457, | |
| "learning_rate": 5.7134102464550925e-06, | |
| "loss": 1.2133, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.617021276595745, | |
| "grad_norm": 0.41424861550331116, | |
| "learning_rate": 5.136956452833776e-06, | |
| "loss": 1.0531, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.6382978723404253, | |
| "grad_norm": 0.4196318984031677, | |
| "learning_rate": 4.589753522209003e-06, | |
| "loss": 0.9811, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.6595744680851063, | |
| "grad_norm": 0.4966506063938141, | |
| "learning_rate": 4.072116145893723e-06, | |
| "loss": 0.8532, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.6808510638297873, | |
| "grad_norm": 0.2934734523296356, | |
| "learning_rate": 3.584342012325771e-06, | |
| "loss": 0.1873, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.702127659574468, | |
| "grad_norm": 0.3809848725795746, | |
| "learning_rate": 3.126711635869966e-06, | |
| "loss": 0.6348, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.723404255319149, | |
| "grad_norm": 0.463334858417511, | |
| "learning_rate": 2.699488195496971e-06, | |
| "loss": 1.2586, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.74468085106383, | |
| "grad_norm": 0.43426749110221863, | |
| "learning_rate": 2.3029173834314634e-06, | |
| "loss": 1.1442, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.7659574468085104, | |
| "grad_norm": 0.4165303111076355, | |
| "learning_rate": 1.9372272638568494e-06, | |
| "loss": 1.0423, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.7872340425531914, | |
| "grad_norm": 0.48008161783218384, | |
| "learning_rate": 1.6026281417576689e-06, | |
| "loss": 0.8429, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.8085106382978724, | |
| "grad_norm": 0.45738592743873596, | |
| "learning_rate": 1.299312441975153e-06, | |
| "loss": 0.3767, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.829787234042553, | |
| "grad_norm": 0.350779265165329, | |
| "learning_rate": 1.0274545985455078e-06, | |
| "loss": 0.5226, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.851063829787234, | |
| "grad_norm": 0.49393415451049805, | |
| "learning_rate": 7.872109543844799e-07, | |
| "loss": 1.2395, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.872340425531915, | |
| "grad_norm": 0.4229400157928467, | |
| "learning_rate": 5.787196713760618e-07, | |
| "loss": 1.1422, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.8936170212765955, | |
| "grad_norm": 0.42691662907600403, | |
| "learning_rate": 4.021006509168048e-07, | |
| "loss": 1.0496, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.9148936170212765, | |
| "grad_norm": 0.4528738260269165, | |
| "learning_rate": 2.574554649617209e-07, | |
| "loss": 0.905, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.9361702127659575, | |
| "grad_norm": 0.4881468713283539, | |
| "learning_rate": 1.4486729761113447e-07, | |
| "loss": 0.5696, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.9574468085106385, | |
| "grad_norm": 0.412494957447052, | |
| "learning_rate": 6.440089727230269e-08, | |
| "loss": 0.7729, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.978723404255319, | |
| "grad_norm": 0.4168694317340851, | |
| "learning_rate": 1.6102539423217266e-08, | |
| "loss": 1.1272, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.48817628622055054, | |
| "learning_rate": 0.0, | |
| "loss": 0.8255, | |
| "step": 141 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 141, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.466930952990884e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |