| { | |
| "best_metric": 0.8657492399215698, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 3.0038910505836576, | |
| "eval_steps": 50, | |
| "global_step": 193, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01556420233463035, | |
| "grad_norm": 0.7498895525932312, | |
| "learning_rate": 1e-05, | |
| "loss": 1.0317, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01556420233463035, | |
| "eval_loss": 1.197574496269226, | |
| "eval_runtime": 2.5784, | |
| "eval_samples_per_second": 41.886, | |
| "eval_steps_per_second": 10.471, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0311284046692607, | |
| "grad_norm": 0.8710477352142334, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1157, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.04669260700389105, | |
| "grad_norm": 0.9946465492248535, | |
| "learning_rate": 3e-05, | |
| "loss": 1.131, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0622568093385214, | |
| "grad_norm": 0.8549498915672302, | |
| "learning_rate": 4e-05, | |
| "loss": 1.0619, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.07782101167315175, | |
| "grad_norm": 0.6751075387001038, | |
| "learning_rate": 5e-05, | |
| "loss": 1.0552, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0933852140077821, | |
| "grad_norm": 0.6019833087921143, | |
| "learning_rate": 6e-05, | |
| "loss": 1.1261, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.10894941634241245, | |
| "grad_norm": 0.5454035997390747, | |
| "learning_rate": 7e-05, | |
| "loss": 1.0568, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1245136186770428, | |
| "grad_norm": 0.4805472791194916, | |
| "learning_rate": 8e-05, | |
| "loss": 0.9972, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.14007782101167315, | |
| "grad_norm": 0.5375916957855225, | |
| "learning_rate": 9e-05, | |
| "loss": 1.0284, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1556420233463035, | |
| "grad_norm": 0.553347110748291, | |
| "learning_rate": 0.0001, | |
| "loss": 1.0405, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.17120622568093385, | |
| "grad_norm": 0.5552372336387634, | |
| "learning_rate": 9.999263238525136e-05, | |
| "loss": 1.0279, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1867704280155642, | |
| "grad_norm": 0.547298014163971, | |
| "learning_rate": 9.997053171227526e-05, | |
| "loss": 0.9419, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.20233463035019456, | |
| "grad_norm": 0.48389917612075806, | |
| "learning_rate": 9.993370449424153e-05, | |
| "loss": 0.9592, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2178988326848249, | |
| "grad_norm": 0.4673521816730499, | |
| "learning_rate": 9.988216158430033e-05, | |
| "loss": 0.938, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.23346303501945526, | |
| "grad_norm": 0.4670703113079071, | |
| "learning_rate": 9.981591817238378e-05, | |
| "loss": 0.8948, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2490272373540856, | |
| "grad_norm": 0.5665939450263977, | |
| "learning_rate": 9.973499378072945e-05, | |
| "loss": 1.0737, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.26459143968871596, | |
| "grad_norm": 0.4406505525112152, | |
| "learning_rate": 9.963941225812701e-05, | |
| "loss": 0.9687, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2801556420233463, | |
| "grad_norm": 0.406009316444397, | |
| "learning_rate": 9.952920177288986e-05, | |
| "loss": 0.9306, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.29571984435797666, | |
| "grad_norm": 0.3831186294555664, | |
| "learning_rate": 9.940439480455386e-05, | |
| "loss": 0.9469, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.311284046692607, | |
| "grad_norm": 0.3777288794517517, | |
| "learning_rate": 9.926502813430545e-05, | |
| "loss": 0.9571, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.32684824902723736, | |
| "grad_norm": 0.37088674306869507, | |
| "learning_rate": 9.911114283414205e-05, | |
| "loss": 0.9663, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.3424124513618677, | |
| "grad_norm": 0.39943259954452515, | |
| "learning_rate": 9.89427842547679e-05, | |
| "loss": 0.9503, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.35797665369649806, | |
| "grad_norm": 0.3599907457828522, | |
| "learning_rate": 9.876000201222912e-05, | |
| "loss": 0.9655, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3735408560311284, | |
| "grad_norm": 0.36868664622306824, | |
| "learning_rate": 9.856284997329158e-05, | |
| "loss": 0.9281, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.38910505836575876, | |
| "grad_norm": 0.34976479411125183, | |
| "learning_rate": 9.835138623956603e-05, | |
| "loss": 0.9323, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.4046692607003891, | |
| "grad_norm": 0.3601384162902832, | |
| "learning_rate": 9.812567313038542e-05, | |
| "loss": 0.8767, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.42023346303501946, | |
| "grad_norm": 0.3567204773426056, | |
| "learning_rate": 9.788577716443902e-05, | |
| "loss": 0.8901, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.4357976653696498, | |
| "grad_norm": 0.35787394642829895, | |
| "learning_rate": 9.763176904016913e-05, | |
| "loss": 0.8908, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.45136186770428016, | |
| "grad_norm": 0.3854931890964508, | |
| "learning_rate": 9.736372361493584e-05, | |
| "loss": 0.9265, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.4669260700389105, | |
| "grad_norm": 0.3999091386795044, | |
| "learning_rate": 9.708171988295631e-05, | |
| "loss": 0.9161, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.48249027237354086, | |
| "grad_norm": 0.3725212514400482, | |
| "learning_rate": 9.678584095202468e-05, | |
| "loss": 0.8742, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4980544747081712, | |
| "grad_norm": 0.4123667776584625, | |
| "learning_rate": 9.647617401902002e-05, | |
| "loss": 0.9402, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5136186770428015, | |
| "grad_norm": 0.31651416420936584, | |
| "learning_rate": 9.61528103442088e-05, | |
| "loss": 0.8945, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5291828793774319, | |
| "grad_norm": 0.3038080334663391, | |
| "learning_rate": 9.581584522435024e-05, | |
| "loss": 0.8982, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5447470817120622, | |
| "grad_norm": 0.3260709047317505, | |
| "learning_rate": 9.546537796461179e-05, | |
| "loss": 0.9635, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5603112840466926, | |
| "grad_norm": 0.31259122490882874, | |
| "learning_rate": 9.510151184930354e-05, | |
| "loss": 0.8647, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.5758754863813229, | |
| "grad_norm": 0.316936731338501, | |
| "learning_rate": 9.472435411143978e-05, | |
| "loss": 0.8711, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.5914396887159533, | |
| "grad_norm": 0.3249667286872864, | |
| "learning_rate": 9.433401590113701e-05, | |
| "loss": 0.9385, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6070038910505836, | |
| "grad_norm": 0.3141433894634247, | |
| "learning_rate": 9.393061225285743e-05, | |
| "loss": 0.8555, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.622568093385214, | |
| "grad_norm": 0.31393998861312866, | |
| "learning_rate": 9.351426205150777e-05, | |
| "loss": 0.91, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6381322957198443, | |
| "grad_norm": 0.3161945343017578, | |
| "learning_rate": 9.308508799740341e-05, | |
| "loss": 0.9217, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.6536964980544747, | |
| "grad_norm": 0.3333125114440918, | |
| "learning_rate": 9.2643216570108e-05, | |
| "loss": 0.9125, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.669260700389105, | |
| "grad_norm": 0.3434793949127197, | |
| "learning_rate": 9.218877799115928e-05, | |
| "loss": 0.8761, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.6848249027237354, | |
| "grad_norm": 0.35263577103614807, | |
| "learning_rate": 9.172190618569236e-05, | |
| "loss": 0.9013, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7003891050583657, | |
| "grad_norm": 0.3485749363899231, | |
| "learning_rate": 9.124273874297122e-05, | |
| "loss": 0.8326, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.7159533073929961, | |
| "grad_norm": 0.3562600612640381, | |
| "learning_rate": 9.075141687584057e-05, | |
| "loss": 0.8795, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.7315175097276264, | |
| "grad_norm": 0.3698873519897461, | |
| "learning_rate": 9.024808537910981e-05, | |
| "loss": 0.914, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.7470817120622568, | |
| "grad_norm": 0.3896391689777374, | |
| "learning_rate": 8.973289258688125e-05, | |
| "loss": 0.8028, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.7626459143968871, | |
| "grad_norm": 0.2838418483734131, | |
| "learning_rate": 8.920599032883554e-05, | |
| "loss": 0.8242, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.7782101167315175, | |
| "grad_norm": 0.32417213916778564, | |
| "learning_rate": 8.86675338854865e-05, | |
| "loss": 0.9819, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.7782101167315175, | |
| "eval_loss": 0.8939120173454285, | |
| "eval_runtime": 2.6016, | |
| "eval_samples_per_second": 41.512, | |
| "eval_steps_per_second": 10.378, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.7937743190661478, | |
| "grad_norm": 0.3049919009208679, | |
| "learning_rate": 8.811768194241952e-05, | |
| "loss": 0.8823, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.8093385214007782, | |
| "grad_norm": 0.3248423635959625, | |
| "learning_rate": 8.755659654352599e-05, | |
| "loss": 0.9283, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.8249027237354085, | |
| "grad_norm": 0.3247882127761841, | |
| "learning_rate": 8.698444304324835e-05, | |
| "loss": 0.9166, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.8404669260700389, | |
| "grad_norm": 0.30053266882896423, | |
| "learning_rate": 8.640139005784924e-05, | |
| "loss": 0.858, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.8560311284046692, | |
| "grad_norm": 0.33963730931282043, | |
| "learning_rate": 8.580760941571967e-05, | |
| "loss": 0.8852, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.8715953307392996, | |
| "grad_norm": 0.31807518005371094, | |
| "learning_rate": 8.520327610674029e-05, | |
| "loss": 0.9407, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.8871595330739299, | |
| "grad_norm": 0.3238030970096588, | |
| "learning_rate": 8.458856823071111e-05, | |
| "loss": 0.8734, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.9027237354085603, | |
| "grad_norm": 0.3245297372341156, | |
| "learning_rate": 8.396366694486466e-05, | |
| "loss": 0.8825, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.9182879377431906, | |
| "grad_norm": 0.3227308988571167, | |
| "learning_rate": 8.332875641047817e-05, | |
| "loss": 0.8962, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.933852140077821, | |
| "grad_norm": 0.3529612123966217, | |
| "learning_rate": 8.26840237386003e-05, | |
| "loss": 0.9568, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.9494163424124513, | |
| "grad_norm": 0.36405235528945923, | |
| "learning_rate": 8.202965893490878e-05, | |
| "loss": 0.897, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.9649805447470817, | |
| "grad_norm": 0.3392234444618225, | |
| "learning_rate": 8.13658548437147e-05, | |
| "loss": 0.8032, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.980544747081712, | |
| "grad_norm": 0.3424031436443329, | |
| "learning_rate": 8.06928070911306e-05, | |
| "loss": 0.8227, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.9961089494163424, | |
| "grad_norm": 0.38254794478416443, | |
| "learning_rate": 8.001071402741842e-05, | |
| "loss": 0.9149, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.0116731517509727, | |
| "grad_norm": 0.9797287583351135, | |
| "learning_rate": 7.931977666853479e-05, | |
| "loss": 1.5691, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.027237354085603, | |
| "grad_norm": 0.2772282361984253, | |
| "learning_rate": 7.862019863689074e-05, | |
| "loss": 0.8003, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.0428015564202335, | |
| "grad_norm": 0.285792738199234, | |
| "learning_rate": 7.791218610134323e-05, | |
| "loss": 0.8685, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.0583657587548638, | |
| "grad_norm": 0.32011640071868896, | |
| "learning_rate": 7.719594771643623e-05, | |
| "loss": 0.9241, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.0739299610894941, | |
| "grad_norm": 0.30448997020721436, | |
| "learning_rate": 7.647169456090925e-05, | |
| "loss": 0.8244, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.0894941634241244, | |
| "grad_norm": 0.3131871819496155, | |
| "learning_rate": 7.573964007549155e-05, | |
| "loss": 0.8418, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.105058365758755, | |
| "grad_norm": 0.3209502398967743, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.7923, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.1206225680933852, | |
| "grad_norm": 0.3268880844116211, | |
| "learning_rate": 7.425299230975981e-05, | |
| "loss": 0.8179, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.1361867704280155, | |
| "grad_norm": 0.3202899396419525, | |
| "learning_rate": 7.3498837151366e-05, | |
| "loss": 0.8361, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.1517509727626458, | |
| "grad_norm": 0.33722591400146484, | |
| "learning_rate": 7.27377567778053e-05, | |
| "loss": 0.8352, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.1673151750972763, | |
| "grad_norm": 0.3361928164958954, | |
| "learning_rate": 7.196997548295708e-05, | |
| "loss": 0.8771, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.1828793774319066, | |
| "grad_norm": 0.3383100628852844, | |
| "learning_rate": 7.119571953549305e-05, | |
| "loss": 0.8353, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.198443579766537, | |
| "grad_norm": 0.342599481344223, | |
| "learning_rate": 7.041521711219467e-05, | |
| "loss": 0.7851, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.2140077821011672, | |
| "grad_norm": 0.3692112863063812, | |
| "learning_rate": 6.96286982307086e-05, | |
| "loss": 0.7754, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.2295719844357977, | |
| "grad_norm": 0.3618566691875458, | |
| "learning_rate": 6.883639468175927e-05, | |
| "loss": 0.7592, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.245136186770428, | |
| "grad_norm": 0.3885677456855774, | |
| "learning_rate": 6.803853996083917e-05, | |
| "loss": 0.8669, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.2607003891050583, | |
| "grad_norm": 0.3042435646057129, | |
| "learning_rate": 6.72353691993967e-05, | |
| "loss": 0.8014, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.2762645914396886, | |
| "grad_norm": 0.2926236093044281, | |
| "learning_rate": 6.642711909554174e-05, | |
| "loss": 0.757, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.2918287937743191, | |
| "grad_norm": 0.315763384103775, | |
| "learning_rate": 6.561402784428974e-05, | |
| "loss": 0.8411, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.3073929961089494, | |
| "grad_norm": 0.3296045958995819, | |
| "learning_rate": 6.479633506736446e-05, | |
| "loss": 0.8441, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.3229571984435797, | |
| "grad_norm": 0.3326404392719269, | |
| "learning_rate": 6.397428174258047e-05, | |
| "loss": 0.8187, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.3385214007782102, | |
| "grad_norm": 0.32586634159088135, | |
| "learning_rate": 6.314811013282574e-05, | |
| "loss": 0.8349, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.3540856031128405, | |
| "grad_norm": 0.33808082342147827, | |
| "learning_rate": 6.231806371466574e-05, | |
| "loss": 0.848, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.3696498054474708, | |
| "grad_norm": 0.3542570173740387, | |
| "learning_rate": 6.14843871065898e-05, | |
| "loss": 0.8243, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.3852140077821011, | |
| "grad_norm": 0.356636643409729, | |
| "learning_rate": 6.064732599692079e-05, | |
| "loss": 0.8619, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.4007782101167314, | |
| "grad_norm": 0.3576579689979553, | |
| "learning_rate": 5.980712707140985e-05, | |
| "loss": 0.7986, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.416342412451362, | |
| "grad_norm": 0.36819812655448914, | |
| "learning_rate": 5.896403794053679e-05, | |
| "loss": 0.8189, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.4319066147859922, | |
| "grad_norm": 0.3678414821624756, | |
| "learning_rate": 5.8118307066538193e-05, | |
| "loss": 0.7622, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.4474708171206225, | |
| "grad_norm": 0.38239309191703796, | |
| "learning_rate": 5.7270183690184495e-05, | |
| "loss": 0.8324, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.463035019455253, | |
| "grad_norm": 0.36592909693717957, | |
| "learning_rate": 5.6419917757327555e-05, | |
| "loss": 0.7254, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.4785992217898833, | |
| "grad_norm": 0.3889016807079315, | |
| "learning_rate": 5.5567759845240444e-05, | |
| "loss": 0.7894, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.4941634241245136, | |
| "grad_norm": 0.4369194507598877, | |
| "learning_rate": 5.4713961088771226e-05, | |
| "loss": 0.8497, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.509727626459144, | |
| "grad_norm": 0.3298012614250183, | |
| "learning_rate": 5.385877310633233e-05, | |
| "loss": 0.7555, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.5252918287937742, | |
| "grad_norm": 0.3367776572704315, | |
| "learning_rate": 5.300244792574742e-05, | |
| "loss": 0.8044, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.5408560311284045, | |
| "grad_norm": 0.3444443941116333, | |
| "learning_rate": 5.214523790997773e-05, | |
| "loss": 0.8537, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.556420233463035, | |
| "grad_norm": 0.33900946378707886, | |
| "learning_rate": 5.128739568274944e-05, | |
| "loss": 0.7601, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.556420233463035, | |
| "eval_loss": 0.8714919686317444, | |
| "eval_runtime": 2.6058, | |
| "eval_samples_per_second": 41.446, | |
| "eval_steps_per_second": 10.362, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5719844357976653, | |
| "grad_norm": 0.3435685634613037, | |
| "learning_rate": 5.0429174054104355e-05, | |
| "loss": 0.8486, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.5875486381322959, | |
| "grad_norm": 0.3485165536403656, | |
| "learning_rate": 4.9570825945895656e-05, | |
| "loss": 0.8179, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.6031128404669261, | |
| "grad_norm": 0.3704763352870941, | |
| "learning_rate": 4.8712604317250576e-05, | |
| "loss": 0.9108, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.6186770428015564, | |
| "grad_norm": 0.33625274896621704, | |
| "learning_rate": 4.785476209002227e-05, | |
| "loss": 0.8262, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.6342412451361867, | |
| "grad_norm": 0.3515668511390686, | |
| "learning_rate": 4.699755207425259e-05, | |
| "loss": 0.8336, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.649805447470817, | |
| "grad_norm": 0.3408714830875397, | |
| "learning_rate": 4.6141226893667684e-05, | |
| "loss": 0.714, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.6653696498054473, | |
| "grad_norm": 0.36845165491104126, | |
| "learning_rate": 4.5286038911228785e-05, | |
| "loss": 0.8147, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.6809338521400778, | |
| "grad_norm": 0.3738980293273926, | |
| "learning_rate": 4.443224015475956e-05, | |
| "loss": 0.8297, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.6964980544747081, | |
| "grad_norm": 0.38755711913108826, | |
| "learning_rate": 4.358008224267245e-05, | |
| "loss": 0.8268, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.7120622568093387, | |
| "grad_norm": 0.367796927690506, | |
| "learning_rate": 4.272981630981551e-05, | |
| "loss": 0.7379, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.727626459143969, | |
| "grad_norm": 0.3794459104537964, | |
| "learning_rate": 4.188169293346183e-05, | |
| "loss": 0.7374, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.7431906614785992, | |
| "grad_norm": 0.4307495653629303, | |
| "learning_rate": 4.103596205946323e-05, | |
| "loss": 0.7783, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.7587548638132295, | |
| "grad_norm": 0.3288561701774597, | |
| "learning_rate": 4.019287292859016e-05, | |
| "loss": 0.8254, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.7743190661478598, | |
| "grad_norm": 0.3198075592517853, | |
| "learning_rate": 3.9352674003079225e-05, | |
| "loss": 0.8722, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.7898832684824901, | |
| "grad_norm": 0.3411862850189209, | |
| "learning_rate": 3.851561289341023e-05, | |
| "loss": 0.8044, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.8054474708171206, | |
| "grad_norm": 0.33337950706481934, | |
| "learning_rate": 3.768193628533427e-05, | |
| "loss": 0.7877, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.821011673151751, | |
| "grad_norm": 0.34444868564605713, | |
| "learning_rate": 3.6851889867174275e-05, | |
| "loss": 0.8396, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.8365758754863815, | |
| "grad_norm": 0.35829517245292664, | |
| "learning_rate": 3.602571825741953e-05, | |
| "loss": 0.7998, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.8521400778210118, | |
| "grad_norm": 0.33558353781700134, | |
| "learning_rate": 3.520366493263554e-05, | |
| "loss": 0.7706, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.867704280155642, | |
| "grad_norm": 0.3696330487728119, | |
| "learning_rate": 3.438597215571027e-05, | |
| "loss": 0.9026, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.8832684824902723, | |
| "grad_norm": 0.34265345335006714, | |
| "learning_rate": 3.357288090445827e-05, | |
| "loss": 0.8013, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.8988326848249026, | |
| "grad_norm": 0.3573516309261322, | |
| "learning_rate": 3.2764630800603314e-05, | |
| "loss": 0.8148, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.914396887159533, | |
| "grad_norm": 0.36973556876182556, | |
| "learning_rate": 3.196146003916084e-05, | |
| "loss": 0.7784, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.9299610894941635, | |
| "grad_norm": 0.37476617097854614, | |
| "learning_rate": 3.116360531824074e-05, | |
| "loss": 0.8075, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.9455252918287937, | |
| "grad_norm": 0.36869266629219055, | |
| "learning_rate": 3.0371301769291417e-05, | |
| "loss": 0.7631, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.9610894941634243, | |
| "grad_norm": 0.38179001212120056, | |
| "learning_rate": 2.9584782887805328e-05, | |
| "loss": 0.7691, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.9766536964980546, | |
| "grad_norm": 0.4227118492126465, | |
| "learning_rate": 2.8804280464506973e-05, | |
| "loss": 0.7856, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.9922178988326849, | |
| "grad_norm": 0.46422451734542847, | |
| "learning_rate": 2.8030024517042907e-05, | |
| "loss": 0.8267, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.007782101167315, | |
| "grad_norm": 1.0904223918914795, | |
| "learning_rate": 2.726224322219473e-05, | |
| "loss": 1.539, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.0233463035019454, | |
| "grad_norm": 0.2821767330169678, | |
| "learning_rate": 2.650116284863402e-05, | |
| "loss": 0.6871, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.0389105058365757, | |
| "grad_norm": 0.3239165246486664, | |
| "learning_rate": 2.5747007690240198e-05, | |
| "loss": 0.8398, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.054474708171206, | |
| "grad_norm": 0.33111441135406494, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 0.8076, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.0700389105058368, | |
| "grad_norm": 0.36252424120903015, | |
| "learning_rate": 2.426035992450848e-05, | |
| "loss": 0.8682, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.085603112840467, | |
| "grad_norm": 0.3458457589149475, | |
| "learning_rate": 2.3528305439090743e-05, | |
| "loss": 0.7667, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.1011673151750974, | |
| "grad_norm": 0.3382876217365265, | |
| "learning_rate": 2.280405228356377e-05, | |
| "loss": 0.7525, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.1167315175097277, | |
| "grad_norm": 0.34494414925575256, | |
| "learning_rate": 2.2087813898656774e-05, | |
| "loss": 0.7819, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.132295719844358, | |
| "grad_norm": 0.3596420884132385, | |
| "learning_rate": 2.137980136310926e-05, | |
| "loss": 0.7824, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.1478599221789882, | |
| "grad_norm": 0.35264426469802856, | |
| "learning_rate": 2.068022333146522e-05, | |
| "loss": 0.7032, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.1634241245136185, | |
| "grad_norm": 0.3882901072502136, | |
| "learning_rate": 1.9989285972581595e-05, | |
| "loss": 0.826, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.178988326848249, | |
| "grad_norm": 0.38036638498306274, | |
| "learning_rate": 1.9307192908869397e-05, | |
| "loss": 0.7353, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.1945525291828796, | |
| "grad_norm": 0.3903452157974243, | |
| "learning_rate": 1.863414515628531e-05, | |
| "loss": 0.712, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.21011673151751, | |
| "grad_norm": 0.4027177095413208, | |
| "learning_rate": 1.7970341065091245e-05, | |
| "loss": 0.7076, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.22568093385214, | |
| "grad_norm": 0.4014608561992645, | |
| "learning_rate": 1.7315976261399696e-05, | |
| "loss": 0.6695, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.2412451361867705, | |
| "grad_norm": 0.43404000997543335, | |
| "learning_rate": 1.667124358952184e-05, | |
| "loss": 0.7546, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.2568093385214008, | |
| "grad_norm": 0.37586840987205505, | |
| "learning_rate": 1.6036333055135344e-05, | |
| "loss": 0.6992, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.272373540856031, | |
| "grad_norm": 0.34168022871017456, | |
| "learning_rate": 1.541143176928891e-05, | |
| "loss": 0.8211, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.2879377431906613, | |
| "grad_norm": 0.33489376306533813, | |
| "learning_rate": 1.4796723893259712e-05, | |
| "loss": 0.7855, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.3035019455252916, | |
| "grad_norm": 0.3305031359195709, | |
| "learning_rate": 1.4192390584280346e-05, | |
| "loss": 0.6823, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.319066147859922, | |
| "grad_norm": 0.3635585606098175, | |
| "learning_rate": 1.3598609942150765e-05, | |
| "loss": 0.7662, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.3346303501945527, | |
| "grad_norm": 0.34550437331199646, | |
| "learning_rate": 1.3015556956751669e-05, | |
| "loss": 0.7758, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.3346303501945527, | |
| "eval_loss": 0.8657492399215698, | |
| "eval_runtime": 2.6057, | |
| "eval_samples_per_second": 41.447, | |
| "eval_steps_per_second": 10.362, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.350194552529183, | |
| "grad_norm": 0.374606192111969, | |
| "learning_rate": 1.2443403456474017e-05, | |
| "loss": 0.7851, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.3657587548638133, | |
| "grad_norm": 0.3662430942058563, | |
| "learning_rate": 1.1882318057580489e-05, | |
| "loss": 0.74, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.3813229571984436, | |
| "grad_norm": 0.3952215909957886, | |
| "learning_rate": 1.1332466114513512e-05, | |
| "loss": 0.8131, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.396887159533074, | |
| "grad_norm": 0.38979917764663696, | |
| "learning_rate": 1.0794009671164484e-05, | |
| "loss": 0.821, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.412451361867704, | |
| "grad_norm": 0.37041887640953064, | |
| "learning_rate": 1.0267107413118742e-05, | |
| "loss": 0.7174, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.4280155642023344, | |
| "grad_norm": 0.41075101494789124, | |
| "learning_rate": 9.751914620890206e-06, | |
| "loss": 0.7549, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.443579766536965, | |
| "grad_norm": 0.4010712206363678, | |
| "learning_rate": 9.248583124159438e-06, | |
| "loss": 0.7279, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.4591439688715955, | |
| "grad_norm": 0.40612390637397766, | |
| "learning_rate": 8.757261257028777e-06, | |
| "loss": 0.7278, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.4747081712062258, | |
| "grad_norm": 0.4142809510231018, | |
| "learning_rate": 8.278093814307637e-06, | |
| "loss": 0.7059, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.490272373540856, | |
| "grad_norm": 0.46315258741378784, | |
| "learning_rate": 7.81122200884072e-06, | |
| "loss": 0.7573, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.5058365758754864, | |
| "grad_norm": 0.38741129636764526, | |
| "learning_rate": 7.356783429892023e-06, | |
| "loss": 0.7816, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.5214007782101167, | |
| "grad_norm": 0.31710007786750793, | |
| "learning_rate": 6.9149120025965905e-06, | |
| "loss": 0.6611, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.536964980544747, | |
| "grad_norm": 0.3693569004535675, | |
| "learning_rate": 6.4857379484922375e-06, | |
| "loss": 0.8636, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.5525291828793772, | |
| "grad_norm": 0.34923169016838074, | |
| "learning_rate": 6.069387747142591e-06, | |
| "loss": 0.7413, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.5680933852140075, | |
| "grad_norm": 0.3575231432914734, | |
| "learning_rate": 5.665984098862992e-06, | |
| "loss": 0.7553, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.5836575875486383, | |
| "grad_norm": 0.3520582318305969, | |
| "learning_rate": 5.275645888560232e-06, | |
| "loss": 0.7394, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.5992217898832686, | |
| "grad_norm": 0.37276095151901245, | |
| "learning_rate": 4.898488150696467e-06, | |
| "loss": 0.7774, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.614785992217899, | |
| "grad_norm": 0.3675203323364258, | |
| "learning_rate": 4.534622035388214e-06, | |
| "loss": 0.7305, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.630350194552529, | |
| "grad_norm": 0.3742014467716217, | |
| "learning_rate": 4.184154775649768e-06, | |
| "loss": 0.7991, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.6459143968871595, | |
| "grad_norm": 0.38438355922698975, | |
| "learning_rate": 3.8471896557912e-06, | |
| "loss": 0.7519, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.6614785992217898, | |
| "grad_norm": 0.41580691933631897, | |
| "learning_rate": 3.523825980979989e-06, | |
| "loss": 0.7663, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.6770428015564205, | |
| "grad_norm": 0.4046590030193329, | |
| "learning_rate": 3.2141590479753236e-06, | |
| "loss": 0.7113, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.692607003891051, | |
| "grad_norm": 0.4197351038455963, | |
| "learning_rate": 2.918280117043709e-06, | |
| "loss": 0.8028, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.708171206225681, | |
| "grad_norm": 0.4047519862651825, | |
| "learning_rate": 2.636276385064157e-06, | |
| "loss": 0.7148, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.7237354085603114, | |
| "grad_norm": 0.42242541909217834, | |
| "learning_rate": 2.3682309598308747e-06, | |
| "loss": 0.6966, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.7392996108949417, | |
| "grad_norm": 0.4608907103538513, | |
| "learning_rate": 2.114222835560986e-06, | |
| "loss": 0.7255, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.754863813229572, | |
| "grad_norm": 0.4060788154602051, | |
| "learning_rate": 1.8743268696145954e-06, | |
| "loss": 0.833, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.7704280155642023, | |
| "grad_norm": 0.33689865469932556, | |
| "learning_rate": 1.6486137604339813e-06, | |
| "loss": 0.7972, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.7859922178988326, | |
| "grad_norm": 0.3548411726951599, | |
| "learning_rate": 1.4371500267084338e-06, | |
| "loss": 0.7801, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.801556420233463, | |
| "grad_norm": 0.3427428603172302, | |
| "learning_rate": 1.2399979877708745e-06, | |
| "loss": 0.7001, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.817120622568093, | |
| "grad_norm": 0.36466020345687866, | |
| "learning_rate": 1.0572157452321097e-06, | |
| "loss": 0.7898, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.832684824902724, | |
| "grad_norm": 0.39266034960746765, | |
| "learning_rate": 8.888571658579703e-07, | |
| "loss": 0.7809, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.848249027237354, | |
| "grad_norm": 0.3834577202796936, | |
| "learning_rate": 7.349718656945504e-07, | |
| "loss": 0.7889, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.8638132295719845, | |
| "grad_norm": 0.3879198133945465, | |
| "learning_rate": 5.956051954461472e-07, | |
| "loss": 0.8306, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.8793774319066148, | |
| "grad_norm": 0.36868515610694885, | |
| "learning_rate": 4.7079822711015296e-07, | |
| "loss": 0.7332, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.894941634241245, | |
| "grad_norm": 0.38267624378204346, | |
| "learning_rate": 3.605877418729975e-07, | |
| "loss": 0.6865, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.9105058365758754, | |
| "grad_norm": 0.39616188406944275, | |
| "learning_rate": 2.6500621927054715e-07, | |
| "loss": 0.7302, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.926070038910506, | |
| "grad_norm": 0.4041755199432373, | |
| "learning_rate": 1.840818276162226e-07, | |
| "loss": 0.8152, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.9416342412451364, | |
| "grad_norm": 0.4070306122303009, | |
| "learning_rate": 1.1783841569968367e-07, | |
| "loss": 0.6876, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.9571984435797667, | |
| "grad_norm": 0.42624837160110474, | |
| "learning_rate": 6.629550575847354e-08, | |
| "loss": 0.7026, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.972762645914397, | |
| "grad_norm": 0.4421103298664093, | |
| "learning_rate": 2.946828772473764e-08, | |
| "loss": 0.7546, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.9883268482490273, | |
| "grad_norm": 0.4374568462371826, | |
| "learning_rate": 7.36761474865455e-09, | |
| "loss": 0.7218, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 3.0038910505836576, | |
| "grad_norm": 0.8102526068687439, | |
| "learning_rate": 0.0, | |
| "loss": 1.2001, | |
| "step": 193 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 193, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "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": 4.879575712530432e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |