| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 7.0, | |
| "eval_steps": 500, | |
| "global_step": 231, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.030303030303030304, | |
| "grad_norm": 6.273081003245219, | |
| "learning_rate": 8.333333333333333e-07, | |
| "loss": 0.9871, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.06060606060606061, | |
| "grad_norm": 5.920272498116613, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 0.9621, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 6.527587701666643, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.0461, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.12121212121212122, | |
| "grad_norm": 5.485565230783434, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.9545, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.15151515151515152, | |
| "grad_norm": 4.5688405868061395, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 0.9528, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 2.4988225200461174, | |
| "learning_rate": 5e-06, | |
| "loss": 0.9396, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.21212121212121213, | |
| "grad_norm": 1.9499384996938154, | |
| "learning_rate": 5.833333333333334e-06, | |
| "loss": 0.8889, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 4.098668152082479, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.9113, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 4.7823409618301485, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.9217, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.30303030303030304, | |
| "grad_norm": 4.386674093027094, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.8726, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 3.660053614123903, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 0.8059, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 3.3495829853528085, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7948, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.3939393939393939, | |
| "grad_norm": 2.6099413548694645, | |
| "learning_rate": 1.0833333333333334e-05, | |
| "loss": 0.7697, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.42424242424242425, | |
| "grad_norm": 2.344473444788401, | |
| "learning_rate": 1.1666666666666668e-05, | |
| "loss": 0.8359, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 2.433078681243702, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.8041, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 2.1982815992161546, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.7724, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.5151515151515151, | |
| "grad_norm": 1.971097850598862, | |
| "learning_rate": 1.416666666666667e-05, | |
| "loss": 0.7489, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 1.537352649584359, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.7809, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.5757575757575758, | |
| "grad_norm": 1.7534318440205705, | |
| "learning_rate": 1.5833333333333333e-05, | |
| "loss": 0.7585, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 1.5257604889750747, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.7338, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.6363636363636364, | |
| "grad_norm": 1.2156861254422033, | |
| "learning_rate": 1.7500000000000002e-05, | |
| "loss": 0.7144, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 1.0022213926401629, | |
| "learning_rate": 1.8333333333333333e-05, | |
| "loss": 0.7033, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.696969696969697, | |
| "grad_norm": 1.3722005956092598, | |
| "learning_rate": 1.916666666666667e-05, | |
| "loss": 0.74, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 1.004972615933361, | |
| "learning_rate": 2e-05, | |
| "loss": 0.6999, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.7575757575757576, | |
| "grad_norm": 0.8944076946792522, | |
| "learning_rate": 1.999884834944106e-05, | |
| "loss": 0.6946, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.7878787878787878, | |
| "grad_norm": 0.9309582976970029, | |
| "learning_rate": 1.9995393663024054e-05, | |
| "loss": 0.6714, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.8181818181818182, | |
| "grad_norm": 0.7049709053016852, | |
| "learning_rate": 1.9989636736467278e-05, | |
| "loss": 0.6711, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.8484848484848485, | |
| "grad_norm": 0.8845278918815375, | |
| "learning_rate": 1.9981578895764272e-05, | |
| "loss": 0.6607, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.8787878787878788, | |
| "grad_norm": 0.7199384893002656, | |
| "learning_rate": 1.9971221996878395e-05, | |
| "loss": 0.6703, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.850328608798665, | |
| "learning_rate": 1.9958568425315316e-05, | |
| "loss": 0.7279, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.9393939393939394, | |
| "grad_norm": 0.7718255750261337, | |
| "learning_rate": 1.9943621095573588e-05, | |
| "loss": 0.7034, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.9696969696969697, | |
| "grad_norm": 0.6158021448652374, | |
| "learning_rate": 1.9926383450473344e-05, | |
| "loss": 0.6627, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.7596403929636277, | |
| "learning_rate": 1.9906859460363307e-05, | |
| "loss": 0.6935, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.0303030303030303, | |
| "grad_norm": 0.6828218430032198, | |
| "learning_rate": 1.9885053622206305e-05, | |
| "loss": 0.6222, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.0606060606060606, | |
| "grad_norm": 0.6263620556255787, | |
| "learning_rate": 1.986097095854347e-05, | |
| "loss": 0.6132, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.6478619722649033, | |
| "learning_rate": 1.9834617016337424e-05, | |
| "loss": 0.6237, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.121212121212121, | |
| "grad_norm": 0.7870991549717925, | |
| "learning_rate": 1.9805997865694616e-05, | |
| "loss": 0.614, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.1515151515151516, | |
| "grad_norm": 0.6355469614847318, | |
| "learning_rate": 1.9775120098467212e-05, | |
| "loss": 0.613, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.1818181818181819, | |
| "grad_norm": 0.7079185439053011, | |
| "learning_rate": 1.9741990826734793e-05, | |
| "loss": 0.6125, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.2121212121212122, | |
| "grad_norm": 0.7751596186983533, | |
| "learning_rate": 1.970661768116622e-05, | |
| "loss": 0.6294, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.2424242424242424, | |
| "grad_norm": 0.6776826382537969, | |
| "learning_rate": 1.9669008809262064e-05, | |
| "loss": 0.6168, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 0.808029514242149, | |
| "learning_rate": 1.9629172873477995e-05, | |
| "loss": 0.6163, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.303030303030303, | |
| "grad_norm": 0.5971954728088179, | |
| "learning_rate": 1.9587119049229558e-05, | |
| "loss": 0.5695, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.5766041259872118, | |
| "learning_rate": 1.954285702277879e-05, | |
| "loss": 0.554, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.3636363636363638, | |
| "grad_norm": 0.7949827100541759, | |
| "learning_rate": 1.9496396989003195e-05, | |
| "loss": 0.5969, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.393939393939394, | |
| "grad_norm": 0.6243455449888042, | |
| "learning_rate": 1.944774964904754e-05, | |
| "loss": 0.606, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.4242424242424243, | |
| "grad_norm": 0.5635873197816895, | |
| "learning_rate": 1.9396926207859085e-05, | |
| "loss": 0.5778, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.7289841184480138, | |
| "learning_rate": 1.9343938371606714e-05, | |
| "loss": 0.5715, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.4848484848484849, | |
| "grad_norm": 0.7064687729202578, | |
| "learning_rate": 1.9288798344984673e-05, | |
| "loss": 0.584, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.5151515151515151, | |
| "grad_norm": 0.537009482679704, | |
| "learning_rate": 1.9231518828401458e-05, | |
| "loss": 0.5671, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.5454545454545454, | |
| "grad_norm": 0.6392807465897785, | |
| "learning_rate": 1.917211301505453e-05, | |
| "loss": 0.5454, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.5757575757575757, | |
| "grad_norm": 0.6312528720614113, | |
| "learning_rate": 1.911059458789152e-05, | |
| "loss": 0.6145, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.606060606060606, | |
| "grad_norm": 0.609027768464515, | |
| "learning_rate": 1.9046977716458627e-05, | |
| "loss": 0.5505, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.6363636363636362, | |
| "grad_norm": 0.5817024281526302, | |
| "learning_rate": 1.8981277053636963e-05, | |
| "loss": 0.5741, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.6144828079386417, | |
| "learning_rate": 1.891350773226754e-05, | |
| "loss": 0.5645, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.696969696969697, | |
| "grad_norm": 0.5846959748827001, | |
| "learning_rate": 1.8843685361665724e-05, | |
| "loss": 0.5908, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.7272727272727273, | |
| "grad_norm": 0.5128702503897558, | |
| "learning_rate": 1.8771826024025944e-05, | |
| "loss": 0.5662, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.7575757575757576, | |
| "grad_norm": 0.6172166900891981, | |
| "learning_rate": 1.8697946270717468e-05, | |
| "loss": 0.5841, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.7878787878787878, | |
| "grad_norm": 0.5841540579991137, | |
| "learning_rate": 1.8622063118472135e-05, | |
| "loss": 0.5718, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.6388703303215946, | |
| "learning_rate": 1.8544194045464888e-05, | |
| "loss": 0.601, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.8484848484848486, | |
| "grad_norm": 0.6286964494131108, | |
| "learning_rate": 1.8464356987288012e-05, | |
| "loss": 0.5705, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.878787878787879, | |
| "grad_norm": 0.6572301406697294, | |
| "learning_rate": 1.8382570332820045e-05, | |
| "loss": 0.6215, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.9090909090909092, | |
| "grad_norm": 0.5712505713768377, | |
| "learning_rate": 1.8298852919990254e-05, | |
| "loss": 0.5704, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.9393939393939394, | |
| "grad_norm": 0.5967512654867877, | |
| "learning_rate": 1.821322403143969e-05, | |
| "loss": 0.5842, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.9696969696969697, | |
| "grad_norm": 0.5304423473677742, | |
| "learning_rate": 1.812570339007983e-05, | |
| "loss": 0.5915, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.5603841035863913, | |
| "learning_rate": 1.8036311154549783e-05, | |
| "loss": 0.5925, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.0303030303030303, | |
| "grad_norm": 0.6943534466423801, | |
| "learning_rate": 1.7945067914573147e-05, | |
| "loss": 0.4651, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.0606060606060606, | |
| "grad_norm": 0.5583784807940949, | |
| "learning_rate": 1.7851994686215592e-05, | |
| "loss": 0.4738, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.090909090909091, | |
| "grad_norm": 0.7232539580663242, | |
| "learning_rate": 1.77571129070442e-05, | |
| "loss": 0.4636, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.121212121212121, | |
| "grad_norm": 0.74672924088183, | |
| "learning_rate": 1.766044443118978e-05, | |
| "loss": 0.4484, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.1515151515151514, | |
| "grad_norm": 0.6260869309416016, | |
| "learning_rate": 1.7562011524313187e-05, | |
| "loss": 0.4521, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.6515446391010243, | |
| "learning_rate": 1.7461836858476858e-05, | |
| "loss": 0.45, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.212121212121212, | |
| "grad_norm": 0.7494472655063361, | |
| "learning_rate": 1.7359943506922775e-05, | |
| "loss": 0.4732, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.242424242424242, | |
| "grad_norm": 0.6199544391366648, | |
| "learning_rate": 1.725635493875799e-05, | |
| "loss": 0.4337, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.2727272727272725, | |
| "grad_norm": 0.6090503441507851, | |
| "learning_rate": 1.7151095013548996e-05, | |
| "loss": 0.4603, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.303030303030303, | |
| "grad_norm": 0.6430114361878058, | |
| "learning_rate": 1.7044187975826126e-05, | |
| "loss": 0.4671, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.3333333333333335, | |
| "grad_norm": 0.49318470415863913, | |
| "learning_rate": 1.693565844949933e-05, | |
| "loss": 0.4333, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.3636363636363638, | |
| "grad_norm": 0.6101231354456229, | |
| "learning_rate": 1.6825531432186545e-05, | |
| "loss": 0.4518, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.393939393939394, | |
| "grad_norm": 1.1631063366479188, | |
| "learning_rate": 1.671383228945597e-05, | |
| "loss": 0.4308, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.4242424242424243, | |
| "grad_norm": 0.5929379937219628, | |
| "learning_rate": 1.6600586748983642e-05, | |
| "loss": 0.4465, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.4545454545454546, | |
| "grad_norm": 0.5896567252240413, | |
| "learning_rate": 1.648582089462756e-05, | |
| "loss": 0.4588, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.484848484848485, | |
| "grad_norm": 0.5707583046071192, | |
| "learning_rate": 1.6369561160419783e-05, | |
| "loss": 0.4654, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.515151515151515, | |
| "grad_norm": 0.58097034664572, | |
| "learning_rate": 1.625183432447789e-05, | |
| "loss": 0.4737, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.4894992630573922, | |
| "learning_rate": 1.6132667502837164e-05, | |
| "loss": 0.4337, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.5757575757575757, | |
| "grad_norm": 0.602056644273567, | |
| "learning_rate": 1.6012088143204953e-05, | |
| "loss": 0.4601, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.606060606060606, | |
| "grad_norm": 0.5549027244080226, | |
| "learning_rate": 1.589012401863864e-05, | |
| "loss": 0.4353, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 2.6363636363636362, | |
| "grad_norm": 0.49762724437545114, | |
| "learning_rate": 1.5766803221148676e-05, | |
| "loss": 0.4451, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.5754996524706565, | |
| "learning_rate": 1.5642154155228124e-05, | |
| "loss": 0.4647, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 2.6969696969696972, | |
| "grad_norm": 0.5465553803793626, | |
| "learning_rate": 1.5516205531310272e-05, | |
| "loss": 0.4494, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 0.5367324837942568, | |
| "learning_rate": 1.538898635915576e-05, | |
| "loss": 0.4568, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.757575757575758, | |
| "grad_norm": 0.5106760397623434, | |
| "learning_rate": 1.526052594117071e-05, | |
| "loss": 0.4463, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 2.787878787878788, | |
| "grad_norm": 0.5199784457999436, | |
| "learning_rate": 1.513085386565758e-05, | |
| "loss": 0.4595, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 2.8181818181818183, | |
| "grad_norm": 0.4585035646613187, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.424, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.8484848484848486, | |
| "grad_norm": 0.5084917957307276, | |
| "learning_rate": 1.4867994483783485e-05, | |
| "loss": 0.4311, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.878787878787879, | |
| "grad_norm": 0.5045556066929575, | |
| "learning_rate": 1.4734867721853341e-05, | |
| "loss": 0.4622, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.5611379812719008, | |
| "learning_rate": 1.4600650377311523e-05, | |
| "loss": 0.461, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.9393939393939394, | |
| "grad_norm": 0.4864406389088855, | |
| "learning_rate": 1.4465373364454001e-05, | |
| "loss": 0.4555, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.9696969696969697, | |
| "grad_norm": 0.5672721310004961, | |
| "learning_rate": 1.4329067841650274e-05, | |
| "loss": 0.4434, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.4545468633078514, | |
| "learning_rate": 1.4191765204166643e-05, | |
| "loss": 0.4218, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.0303030303030303, | |
| "grad_norm": 0.7643361870827023, | |
| "learning_rate": 1.4053497076934948e-05, | |
| "loss": 0.3341, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.0606060606060606, | |
| "grad_norm": 0.5380843322386185, | |
| "learning_rate": 1.3914295307268396e-05, | |
| "loss": 0.3457, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 3.090909090909091, | |
| "grad_norm": 0.8495898983486152, | |
| "learning_rate": 1.3774191957526144e-05, | |
| "loss": 0.3365, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 3.121212121212121, | |
| "grad_norm": 0.8238297887149431, | |
| "learning_rate": 1.3633219297728415e-05, | |
| "loss": 0.3489, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 3.1515151515151514, | |
| "grad_norm": 0.6688708370220503, | |
| "learning_rate": 1.3491409798123687e-05, | |
| "loss": 0.3523, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 3.1818181818181817, | |
| "grad_norm": 0.6734311736209873, | |
| "learning_rate": 1.3348796121709862e-05, | |
| "loss": 0.3416, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 3.212121212121212, | |
| "grad_norm": 0.5420767973640096, | |
| "learning_rate": 1.3205411116710973e-05, | |
| "loss": 0.341, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 3.242424242424242, | |
| "grad_norm": 0.5714403388816376, | |
| "learning_rate": 1.3061287809011243e-05, | |
| "loss": 0.3578, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.5878564048028506, | |
| "learning_rate": 1.291645939454825e-05, | |
| "loss": 0.3387, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 3.303030303030303, | |
| "grad_norm": 0.5783270076763397, | |
| "learning_rate": 1.277095923166689e-05, | |
| "loss": 0.3228, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 0.56339440839081, | |
| "learning_rate": 1.2624820833435939e-05, | |
| "loss": 0.3359, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 3.3636363636363638, | |
| "grad_norm": 0.5635584050608189, | |
| "learning_rate": 1.2478077859929e-05, | |
| "loss": 0.3066, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 3.393939393939394, | |
| "grad_norm": 0.5833609440720666, | |
| "learning_rate": 1.2330764110471567e-05, | |
| "loss": 0.346, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 3.4242424242424243, | |
| "grad_norm": 0.5361388733521907, | |
| "learning_rate": 1.2182913515856016e-05, | |
| "loss": 0.3375, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 3.4545454545454546, | |
| "grad_norm": 0.5610524512691069, | |
| "learning_rate": 1.2034560130526341e-05, | |
| "loss": 0.3359, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 3.484848484848485, | |
| "grad_norm": 0.4896809632141169, | |
| "learning_rate": 1.1885738124734359e-05, | |
| "loss": 0.3224, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 3.515151515151515, | |
| "grad_norm": 0.5243166297806552, | |
| "learning_rate": 1.1736481776669307e-05, | |
| "loss": 0.3143, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 3.5454545454545454, | |
| "grad_norm": 0.5175262587060634, | |
| "learning_rate": 1.1586825464562515e-05, | |
| "loss": 0.3399, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 3.5757575757575757, | |
| "grad_norm": 0.5088277721523784, | |
| "learning_rate": 1.1436803658769082e-05, | |
| "loss": 0.3333, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 3.606060606060606, | |
| "grad_norm": 0.5448004846894714, | |
| "learning_rate": 1.1286450913828313e-05, | |
| "loss": 0.3317, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.48338448013631813, | |
| "learning_rate": 1.113580186050475e-05, | |
| "loss": 0.3207, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.6666666666666665, | |
| "grad_norm": 0.5074831149839798, | |
| "learning_rate": 1.0984891197811686e-05, | |
| "loss": 0.3347, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 3.6969696969696972, | |
| "grad_norm": 0.5118069172780434, | |
| "learning_rate": 1.0833753685018935e-05, | |
| "loss": 0.3328, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 3.7272727272727275, | |
| "grad_norm": 0.4724263712758371, | |
| "learning_rate": 1.0682424133646712e-05, | |
| "loss": 0.3514, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 3.757575757575758, | |
| "grad_norm": 0.5171786295597968, | |
| "learning_rate": 1.0530937399447496e-05, | |
| "loss": 0.32, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 3.787878787878788, | |
| "grad_norm": 0.47499666370126736, | |
| "learning_rate": 1.0379328374377715e-05, | |
| "loss": 0.311, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 3.8181818181818183, | |
| "grad_norm": 0.48756533172554267, | |
| "learning_rate": 1.0227631978561057e-05, | |
| "loss": 0.3429, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 3.8484848484848486, | |
| "grad_norm": 0.5562080574708145, | |
| "learning_rate": 1.0075883152245334e-05, | |
| "loss": 0.3411, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 3.878787878787879, | |
| "grad_norm": 0.4674385217266003, | |
| "learning_rate": 9.92411684775467e-06, | |
| "loss": 0.3246, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 3.909090909090909, | |
| "grad_norm": 0.5064363403234642, | |
| "learning_rate": 9.772368021438943e-06, | |
| "loss": 0.3435, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 3.9393939393939394, | |
| "grad_norm": 0.4568842796750717, | |
| "learning_rate": 9.620671625622287e-06, | |
| "loss": 0.3271, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 3.9696969696969697, | |
| "grad_norm": 0.4374908689494871, | |
| "learning_rate": 9.469062600552509e-06, | |
| "loss": 0.3117, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.43345441567131227, | |
| "learning_rate": 9.317575866353293e-06, | |
| "loss": 0.3093, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 4.03030303030303, | |
| "grad_norm": 0.6788823582743904, | |
| "learning_rate": 9.166246314981066e-06, | |
| "loss": 0.2535, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 4.0606060606060606, | |
| "grad_norm": 0.5077486602256049, | |
| "learning_rate": 9.015108802188314e-06, | |
| "loss": 0.252, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 4.090909090909091, | |
| "grad_norm": 0.4808160627565754, | |
| "learning_rate": 8.86419813949525e-06, | |
| "loss": 0.2476, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 4.121212121212121, | |
| "grad_norm": 0.7873249154626186, | |
| "learning_rate": 8.71354908617169e-06, | |
| "loss": 0.2346, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 4.151515151515151, | |
| "grad_norm": 0.6902922144982812, | |
| "learning_rate": 8.56319634123092e-06, | |
| "loss": 0.2378, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 4.181818181818182, | |
| "grad_norm": 0.499612327327111, | |
| "learning_rate": 8.413174535437486e-06, | |
| "loss": 0.2348, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 4.212121212121212, | |
| "grad_norm": 0.5763022592376869, | |
| "learning_rate": 8.263518223330698e-06, | |
| "loss": 0.2097, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 4.242424242424242, | |
| "grad_norm": 0.5522419924253761, | |
| "learning_rate": 8.114261875265643e-06, | |
| "loss": 0.2254, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 4.2727272727272725, | |
| "grad_norm": 0.5909281296551671, | |
| "learning_rate": 7.965439869473664e-06, | |
| "loss": 0.2591, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 4.303030303030303, | |
| "grad_norm": 0.501430641597351, | |
| "learning_rate": 7.817086484143987e-06, | |
| "loss": 0.2232, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 4.333333333333333, | |
| "grad_norm": 0.5361210286942784, | |
| "learning_rate": 7.669235889528436e-06, | |
| "loss": 0.2387, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.5930364997151729, | |
| "learning_rate": 7.521922140071003e-06, | |
| "loss": 0.2357, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 4.393939393939394, | |
| "grad_norm": 0.5161657743899781, | |
| "learning_rate": 7.375179166564062e-06, | |
| "loss": 0.2267, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 4.424242424242424, | |
| "grad_norm": 0.4703990125421506, | |
| "learning_rate": 7.2290407683331154e-06, | |
| "loss": 0.2251, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 4.454545454545454, | |
| "grad_norm": 0.554596466384478, | |
| "learning_rate": 7.0835406054517505e-06, | |
| "loss": 0.2383, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 4.484848484848484, | |
| "grad_norm": 0.47203250276916336, | |
| "learning_rate": 6.93871219098876e-06, | |
| "loss": 0.2259, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 4.515151515151516, | |
| "grad_norm": 0.5851715969114882, | |
| "learning_rate": 6.79458888328903e-06, | |
| "loss": 0.2258, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 4.545454545454545, | |
| "grad_norm": 0.5087145971318606, | |
| "learning_rate": 6.651203878290139e-06, | |
| "loss": 0.2492, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 4.575757575757576, | |
| "grad_norm": 0.4752705124774156, | |
| "learning_rate": 6.508590201876317e-06, | |
| "loss": 0.2426, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 4.606060606060606, | |
| "grad_norm": 0.4842557860781673, | |
| "learning_rate": 6.366780702271589e-06, | |
| "loss": 0.2356, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 4.636363636363637, | |
| "grad_norm": 0.4930551696316793, | |
| "learning_rate": 6.225808042473857e-06, | |
| "loss": 0.2473, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 4.666666666666667, | |
| "grad_norm": 0.4566892934845832, | |
| "learning_rate": 6.085704692731609e-06, | |
| "loss": 0.2449, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 4.696969696969697, | |
| "grad_norm": 0.4738031298591034, | |
| "learning_rate": 5.946502923065054e-06, | |
| "loss": 0.2363, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.46202057693972615, | |
| "learning_rate": 5.8082347958333625e-06, | |
| "loss": 0.2212, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 4.757575757575758, | |
| "grad_norm": 0.4612201856123231, | |
| "learning_rate": 5.670932158349732e-06, | |
| "loss": 0.2218, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 4.787878787878788, | |
| "grad_norm": 0.4526319659556159, | |
| "learning_rate": 5.534626635546e-06, | |
| "loss": 0.2474, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 4.818181818181818, | |
| "grad_norm": 0.4079769127223875, | |
| "learning_rate": 5.399349622688479e-06, | |
| "loss": 0.234, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 4.848484848484849, | |
| "grad_norm": 0.459446307016369, | |
| "learning_rate": 5.2651322781466606e-06, | |
| "loss": 0.217, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.878787878787879, | |
| "grad_norm": 0.4442272973402734, | |
| "learning_rate": 5.132005516216512e-06, | |
| "loss": 0.228, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "grad_norm": 0.42350118178243185, | |
| "learning_rate": 5.000000000000003e-06, | |
| "loss": 0.2254, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 4.9393939393939394, | |
| "grad_norm": 0.421419507634738, | |
| "learning_rate": 4.869146134342426e-06, | |
| "loss": 0.2238, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 4.96969696969697, | |
| "grad_norm": 0.4146660231102621, | |
| "learning_rate": 4.739474058829288e-06, | |
| "loss": 0.2291, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.42798276421076487, | |
| "learning_rate": 4.611013640844245e-06, | |
| "loss": 0.2428, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 5.03030303030303, | |
| "grad_norm": 0.7135551727580502, | |
| "learning_rate": 4.483794468689728e-06, | |
| "loss": 0.1837, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 5.0606060606060606, | |
| "grad_norm": 0.5954421974028176, | |
| "learning_rate": 4.357845844771881e-06, | |
| "loss": 0.187, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.42936088158795754, | |
| "learning_rate": 4.2331967788513295e-06, | |
| "loss": 0.1874, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 5.121212121212121, | |
| "grad_norm": 0.4567103523932135, | |
| "learning_rate": 4.109875981361363e-06, | |
| "loss": 0.1661, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 5.151515151515151, | |
| "grad_norm": 0.6028377796283919, | |
| "learning_rate": 3.987911856795047e-06, | |
| "loss": 0.1644, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 5.181818181818182, | |
| "grad_norm": 0.6851291027441531, | |
| "learning_rate": 3.867332497162836e-06, | |
| "loss": 0.186, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 5.212121212121212, | |
| "grad_norm": 0.5551187801583767, | |
| "learning_rate": 3.748165675522113e-06, | |
| "loss": 0.1646, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 5.242424242424242, | |
| "grad_norm": 0.5019692144623259, | |
| "learning_rate": 3.630438839580217e-06, | |
| "loss": 0.1787, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 5.2727272727272725, | |
| "grad_norm": 0.43540057450718517, | |
| "learning_rate": 3.5141791053724405e-06, | |
| "loss": 0.1622, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 5.303030303030303, | |
| "grad_norm": 0.4502288241855601, | |
| "learning_rate": 3.399413251016359e-06, | |
| "loss": 0.183, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 5.333333333333333, | |
| "grad_norm": 0.48061803401108444, | |
| "learning_rate": 3.2861677105440335e-06, | |
| "loss": 0.1734, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 5.363636363636363, | |
| "grad_norm": 0.46474401588938824, | |
| "learning_rate": 3.174468567813461e-06, | |
| "loss": 0.1828, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 5.393939393939394, | |
| "grad_norm": 0.41209812489824965, | |
| "learning_rate": 3.0643415505006733e-06, | |
| "loss": 0.178, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 5.424242424242424, | |
| "grad_norm": 0.40781589521589334, | |
| "learning_rate": 2.9558120241738786e-06, | |
| "loss": 0.1659, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.40077637178054026, | |
| "learning_rate": 2.8489049864510053e-06, | |
| "loss": 0.1688, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 5.484848484848484, | |
| "grad_norm": 0.4087833881453937, | |
| "learning_rate": 2.7436450612420098e-06, | |
| "loss": 0.167, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 5.515151515151516, | |
| "grad_norm": 0.4338992637714084, | |
| "learning_rate": 2.640056493077231e-06, | |
| "loss": 0.1695, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 5.545454545454545, | |
| "grad_norm": 0.42076419038226126, | |
| "learning_rate": 2.5381631415231455e-06, | |
| "loss": 0.1689, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 5.575757575757576, | |
| "grad_norm": 0.4107700093047128, | |
| "learning_rate": 2.4379884756868167e-06, | |
| "loss": 0.1622, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 5.606060606060606, | |
| "grad_norm": 0.4149841568513108, | |
| "learning_rate": 2.339555568810221e-06, | |
| "loss": 0.156, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 5.636363636363637, | |
| "grad_norm": 0.38268790150794213, | |
| "learning_rate": 2.2428870929558012e-06, | |
| "loss": 0.1571, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 5.666666666666667, | |
| "grad_norm": 0.40358171363521395, | |
| "learning_rate": 2.1480053137844115e-06, | |
| "loss": 0.1612, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 5.696969696969697, | |
| "grad_norm": 0.37962062472536845, | |
| "learning_rate": 2.054932085426856e-06, | |
| "loss": 0.1725, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 5.7272727272727275, | |
| "grad_norm": 0.4185374915330779, | |
| "learning_rate": 1.963688845450218e-06, | |
| "loss": 0.1611, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 5.757575757575758, | |
| "grad_norm": 0.3900456177603315, | |
| "learning_rate": 1.8742966099201699e-06, | |
| "loss": 0.1773, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 5.787878787878788, | |
| "grad_norm": 0.39681669269427383, | |
| "learning_rate": 1.7867759685603115e-06, | |
| "loss": 0.1555, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 0.39537183092245015, | |
| "learning_rate": 1.7011470800097496e-06, | |
| "loss": 0.1655, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 5.848484848484849, | |
| "grad_norm": 0.4613954121183705, | |
| "learning_rate": 1.6174296671799571e-06, | |
| "loss": 0.1835, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 5.878787878787879, | |
| "grad_norm": 0.36858437882918565, | |
| "learning_rate": 1.5356430127119915e-06, | |
| "loss": 0.1788, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 5.909090909090909, | |
| "grad_norm": 0.3880701998652866, | |
| "learning_rate": 1.4558059545351144e-06, | |
| "loss": 0.1679, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 5.9393939393939394, | |
| "grad_norm": 0.3789364742507397, | |
| "learning_rate": 1.3779368815278648e-06, | |
| "loss": 0.1707, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 5.96969696969697, | |
| "grad_norm": 0.36147645918290583, | |
| "learning_rate": 1.302053729282533e-06, | |
| "loss": 0.16, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.40947695535213685, | |
| "learning_rate": 1.2281739759740575e-06, | |
| "loss": 0.1615, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 6.03030303030303, | |
| "grad_norm": 0.4509774085082487, | |
| "learning_rate": 1.156314638334277e-06, | |
| "loss": 0.1424, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 6.0606060606060606, | |
| "grad_norm": 0.45252151181427464, | |
| "learning_rate": 1.086492267732462e-06, | |
| "loss": 0.1347, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 6.090909090909091, | |
| "grad_norm": 0.40400885923731233, | |
| "learning_rate": 1.01872294636304e-06, | |
| "loss": 0.1333, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 6.121212121212121, | |
| "grad_norm": 0.39349345706418215, | |
| "learning_rate": 9.530222835413739e-07, | |
| "loss": 0.1467, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 6.151515151515151, | |
| "grad_norm": 0.3612201567683752, | |
| "learning_rate": 8.894054121084839e-07, | |
| "loss": 0.1524, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.3480530668992422, | |
| "learning_rate": 8.278869849454718e-07, | |
| "loss": 0.1437, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 6.212121212121212, | |
| "grad_norm": 0.3343067967543719, | |
| "learning_rate": 7.684811715985429e-07, | |
| "loss": 0.1455, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 6.242424242424242, | |
| "grad_norm": 0.3786584180768872, | |
| "learning_rate": 7.1120165501533e-07, | |
| "loss": 0.1453, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 6.2727272727272725, | |
| "grad_norm": 0.3741756558808332, | |
| "learning_rate": 6.560616283932897e-07, | |
| "loss": 0.1183, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 6.303030303030303, | |
| "grad_norm": 0.4577134449188471, | |
| "learning_rate": 6.030737921409169e-07, | |
| "loss": 0.1559, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 6.333333333333333, | |
| "grad_norm": 0.41321176408469956, | |
| "learning_rate": 5.522503509524591e-07, | |
| "loss": 0.1356, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 6.363636363636363, | |
| "grad_norm": 0.4249157693127469, | |
| "learning_rate": 5.036030109968082e-07, | |
| "loss": 0.1317, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 6.393939393939394, | |
| "grad_norm": 0.36135312990476776, | |
| "learning_rate": 4.5714297722121105e-07, | |
| "loss": 0.145, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 6.424242424242424, | |
| "grad_norm": 0.3637087942242107, | |
| "learning_rate": 4.128809507704445e-07, | |
| "loss": 0.1441, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 6.454545454545454, | |
| "grad_norm": 0.3664052621305879, | |
| "learning_rate": 3.708271265220087e-07, | |
| "loss": 0.1293, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 6.484848484848484, | |
| "grad_norm": 0.3513959882631917, | |
| "learning_rate": 3.309911907379393e-07, | |
| "loss": 0.1494, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 6.515151515151516, | |
| "grad_norm": 0.37988680146261766, | |
| "learning_rate": 2.9338231883378365e-07, | |
| "loss": 0.1541, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.3510483154141182, | |
| "learning_rate": 2.5800917326521013e-07, | |
| "loss": 0.1391, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 6.575757575757576, | |
| "grad_norm": 0.34230875514426157, | |
| "learning_rate": 2.248799015327907e-07, | |
| "loss": 0.1466, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 6.606060606060606, | |
| "grad_norm": 0.37049229007031076, | |
| "learning_rate": 1.9400213430538773e-07, | |
| "loss": 0.1341, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 6.636363636363637, | |
| "grad_norm": 0.3346587086128842, | |
| "learning_rate": 1.6538298366257975e-07, | |
| "loss": 0.1464, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 0.3574807896784062, | |
| "learning_rate": 1.3902904145653094e-07, | |
| "loss": 0.1452, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 6.696969696969697, | |
| "grad_norm": 0.38095155601900865, | |
| "learning_rate": 1.1494637779369766e-07, | |
| "loss": 0.1435, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 6.7272727272727275, | |
| "grad_norm": 0.3791801173687394, | |
| "learning_rate": 9.314053963669245e-08, | |
| "loss": 0.159, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 6.757575757575758, | |
| "grad_norm": 0.35188904285544703, | |
| "learning_rate": 7.361654952665608e-08, | |
| "loss": 0.1404, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 6.787878787878788, | |
| "grad_norm": 0.35697879614024036, | |
| "learning_rate": 5.637890442641403e-08, | |
| "loss": 0.1265, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 6.818181818181818, | |
| "grad_norm": 0.3541289855500666, | |
| "learning_rate": 4.143157468468717e-08, | |
| "loss": 0.145, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 6.848484848484849, | |
| "grad_norm": 0.3462278060765447, | |
| "learning_rate": 2.8778003121607834e-08, | |
| "loss": 0.1595, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 6.878787878787879, | |
| "grad_norm": 0.3606296743470051, | |
| "learning_rate": 1.8421104235727406e-08, | |
| "loss": 0.1439, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.3647370882548279, | |
| "learning_rate": 1.0363263532724433e-08, | |
| "loss": 0.1241, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 6.9393939393939394, | |
| "grad_norm": 0.34811139100231836, | |
| "learning_rate": 4.606336975948589e-09, | |
| "loss": 0.1448, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 6.96969696969697, | |
| "grad_norm": 0.32933481010553317, | |
| "learning_rate": 1.1516505589381777e-09, | |
| "loss": 0.147, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.3493372290493467, | |
| "learning_rate": 0.0, | |
| "loss": 0.137, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "step": 231, | |
| "total_flos": 3.354167586660024e+17, | |
| "train_loss": 0.3882459873309383, | |
| "train_runtime": 6538.9235, | |
| "train_samples_per_second": 3.383, | |
| "train_steps_per_second": 0.035 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 231, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.354167586660024e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |