| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.850632911392405, | |
| "eval_steps": 500, | |
| "global_step": 224, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.030379746835443037, | |
| "grad_norm": 7.0615418785296304, | |
| "learning_rate": 8.695652173913044e-07, | |
| "loss": 1.0776, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.060759493670886074, | |
| "grad_norm": 7.031434280233566, | |
| "learning_rate": 1.7391304347826088e-06, | |
| "loss": 1.0618, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.09113924050632911, | |
| "grad_norm": 6.900740322782977, | |
| "learning_rate": 2.6086956521739132e-06, | |
| "loss": 1.0648, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.12151898734177215, | |
| "grad_norm": 6.237291849041829, | |
| "learning_rate": 3.4782608695652175e-06, | |
| "loss": 1.0266, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.1518987341772152, | |
| "grad_norm": 4.867661832847373, | |
| "learning_rate": 4.347826086956522e-06, | |
| "loss": 1.0005, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.18227848101265823, | |
| "grad_norm": 2.9290580966219886, | |
| "learning_rate": 5.2173913043478265e-06, | |
| "loss": 0.9763, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.21265822784810126, | |
| "grad_norm": 2.7461883169716166, | |
| "learning_rate": 6.086956521739132e-06, | |
| "loss": 0.9694, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.2430379746835443, | |
| "grad_norm": 4.328544147340575, | |
| "learning_rate": 6.956521739130435e-06, | |
| "loss": 0.9277, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.27341772151898736, | |
| "grad_norm": 4.521945151424538, | |
| "learning_rate": 7.82608695652174e-06, | |
| "loss": 0.9406, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 3.9079266559768144, | |
| "learning_rate": 8.695652173913044e-06, | |
| "loss": 0.9226, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3341772151898734, | |
| "grad_norm": 3.6395220631355647, | |
| "learning_rate": 9.565217391304349e-06, | |
| "loss": 0.894, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.36455696202531646, | |
| "grad_norm": 2.9449357151850117, | |
| "learning_rate": 1.0434782608695653e-05, | |
| "loss": 0.8802, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.3949367088607595, | |
| "grad_norm": 2.1328440517315914, | |
| "learning_rate": 1.1304347826086957e-05, | |
| "loss": 0.857, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.4253164556962025, | |
| "grad_norm": 1.7576195263816612, | |
| "learning_rate": 1.2173913043478263e-05, | |
| "loss": 0.838, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.45569620253164556, | |
| "grad_norm": 1.8892473762350221, | |
| "learning_rate": 1.3043478260869566e-05, | |
| "loss": 0.8134, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.4860759493670886, | |
| "grad_norm": 1.3920077129906296, | |
| "learning_rate": 1.391304347826087e-05, | |
| "loss": 0.8001, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.5164556962025316, | |
| "grad_norm": 1.1872393850497394, | |
| "learning_rate": 1.4782608695652174e-05, | |
| "loss": 0.7915, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.5468354430379747, | |
| "grad_norm": 1.410459649546826, | |
| "learning_rate": 1.565217391304348e-05, | |
| "loss": 0.7718, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.5772151898734177, | |
| "grad_norm": 1.0659729068231212, | |
| "learning_rate": 1.6521739130434785e-05, | |
| "loss": 0.7686, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 1.1803502134894528, | |
| "learning_rate": 1.739130434782609e-05, | |
| "loss": 0.752, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.6379746835443038, | |
| "grad_norm": 1.1916862494591993, | |
| "learning_rate": 1.8260869565217393e-05, | |
| "loss": 0.7461, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.6683544303797468, | |
| "grad_norm": 0.9913950639976256, | |
| "learning_rate": 1.9130434782608697e-05, | |
| "loss": 0.7363, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.6987341772151898, | |
| "grad_norm": 0.9384699419296254, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7365, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.7291139240506329, | |
| "grad_norm": 1.014369524017478, | |
| "learning_rate": 1.999877856940653e-05, | |
| "loss": 0.7327, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.759493670886076, | |
| "grad_norm": 1.007967715634519, | |
| "learning_rate": 1.999511457600466e-05, | |
| "loss": 0.7416, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.789873417721519, | |
| "grad_norm": 0.8916124844925031, | |
| "learning_rate": 1.9989008914857115e-05, | |
| "loss": 0.7196, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.8202531645569621, | |
| "grad_norm": 0.7317460107570771, | |
| "learning_rate": 1.998046307749216e-05, | |
| "loss": 0.7145, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.850632911392405, | |
| "grad_norm": 0.8049550373425254, | |
| "learning_rate": 1.9969479151539238e-05, | |
| "loss": 0.7056, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.8810126582278481, | |
| "grad_norm": 0.7575137043533016, | |
| "learning_rate": 1.9956059820218982e-05, | |
| "loss": 0.7055, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 0.7409708248685565, | |
| "learning_rate": 1.9940208361687762e-05, | |
| "loss": 0.7232, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.9417721518987342, | |
| "grad_norm": 0.7197472665530662, | |
| "learning_rate": 1.9921928648236855e-05, | |
| "loss": 0.7091, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.9721518987341772, | |
| "grad_norm": 0.7197756931186233, | |
| "learning_rate": 1.990122514534651e-05, | |
| "loss": 0.709, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.010126582278481, | |
| "grad_norm": 0.5926789947879165, | |
| "learning_rate": 1.9878102910595097e-05, | |
| "loss": 0.6872, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.040506329113924, | |
| "grad_norm": 0.6105105708579948, | |
| "learning_rate": 1.985256759242359e-05, | |
| "loss": 0.6668, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.070886075949367, | |
| "grad_norm": 0.7850734281573082, | |
| "learning_rate": 1.982462542875576e-05, | |
| "loss": 0.6862, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.1012658227848102, | |
| "grad_norm": 0.6023359613588858, | |
| "learning_rate": 1.979428324547432e-05, | |
| "loss": 0.6604, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.1316455696202532, | |
| "grad_norm": 0.8744364748314618, | |
| "learning_rate": 1.9761548454753455e-05, | |
| "loss": 0.6688, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.1620253164556962, | |
| "grad_norm": 0.623501389791168, | |
| "learning_rate": 1.972642905324813e-05, | |
| "loss": 0.6582, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.1924050632911392, | |
| "grad_norm": 0.7009295264201532, | |
| "learning_rate": 1.9688933620140638e-05, | |
| "loss": 0.6718, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.2227848101265824, | |
| "grad_norm": 0.5093178677347182, | |
| "learning_rate": 1.96490713150448e-05, | |
| "loss": 0.6488, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.2531645569620253, | |
| "grad_norm": 0.5802069964623212, | |
| "learning_rate": 1.9606851875768404e-05, | |
| "loss": 0.6432, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.2835443037974683, | |
| "grad_norm": 0.5542872509749827, | |
| "learning_rate": 1.956228561593441e-05, | |
| "loss": 0.6481, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.3139240506329113, | |
| "grad_norm": 0.6193122090800646, | |
| "learning_rate": 1.9515383422461457e-05, | |
| "loss": 0.651, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.3443037974683545, | |
| "grad_norm": 0.4896883922136707, | |
| "learning_rate": 1.9466156752904344e-05, | |
| "loss": 0.648, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.3746835443037975, | |
| "grad_norm": 0.5789102466015484, | |
| "learning_rate": 1.9414617632655114e-05, | |
| "loss": 0.641, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.4050632911392404, | |
| "grad_norm": 0.5872778984939839, | |
| "learning_rate": 1.9360778652005416e-05, | |
| "loss": 0.6506, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.4354430379746836, | |
| "grad_norm": 0.5380497381909899, | |
| "learning_rate": 1.9304652963070868e-05, | |
| "loss": 0.6356, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.4658227848101266, | |
| "grad_norm": 0.5252839118698206, | |
| "learning_rate": 1.9246254276578175e-05, | |
| "loss": 0.6266, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.4962025316455696, | |
| "grad_norm": 0.5602822733250735, | |
| "learning_rate": 1.9185596858515797e-05, | |
| "loss": 0.6413, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.5265822784810128, | |
| "grad_norm": 0.5835837551243547, | |
| "learning_rate": 1.9122695526648968e-05, | |
| "loss": 0.6602, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.5569620253164556, | |
| "grad_norm": 0.5917500673359287, | |
| "learning_rate": 1.905756564689991e-05, | |
| "loss": 0.6456, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.5873417721518988, | |
| "grad_norm": 0.6910314203022653, | |
| "learning_rate": 1.8990223129594146e-05, | |
| "loss": 0.6278, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.6177215189873417, | |
| "grad_norm": 0.6420529676042861, | |
| "learning_rate": 1.8920684425573865e-05, | |
| "loss": 0.645, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.6481012658227847, | |
| "grad_norm": 0.7008480199873627, | |
| "learning_rate": 1.884896652217917e-05, | |
| "loss": 0.6369, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.678481012658228, | |
| "grad_norm": 0.6358696901772899, | |
| "learning_rate": 1.877508693909831e-05, | |
| "loss": 0.6216, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.7088607594936709, | |
| "grad_norm": 0.5282776593930177, | |
| "learning_rate": 1.8699063724087905e-05, | |
| "loss": 0.629, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.7392405063291139, | |
| "grad_norm": 0.6664146772298517, | |
| "learning_rate": 1.862091544856407e-05, | |
| "loss": 0.6251, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.769620253164557, | |
| "grad_norm": 0.5607847151990235, | |
| "learning_rate": 1.854066120306571e-05, | |
| "loss": 0.6191, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.5591150963867303, | |
| "learning_rate": 1.8458320592590976e-05, | |
| "loss": 0.6276, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.830379746835443, | |
| "grad_norm": 0.4743388621074495, | |
| "learning_rate": 1.837391373180801e-05, | |
| "loss": 0.62, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.8607594936708862, | |
| "grad_norm": 0.6426054204999431, | |
| "learning_rate": 1.8287461240141217e-05, | |
| "loss": 0.6287, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.891139240506329, | |
| "grad_norm": 0.5774861483694179, | |
| "learning_rate": 1.8198984236734246e-05, | |
| "loss": 0.6397, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.9215189873417722, | |
| "grad_norm": 0.5747896534668795, | |
| "learning_rate": 1.8108504335290852e-05, | |
| "loss": 0.6151, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.9518987341772152, | |
| "grad_norm": 0.730974313149879, | |
| "learning_rate": 1.8016043638794975e-05, | |
| "loss": 0.6275, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.9822784810126581, | |
| "grad_norm": 0.5475199878731721, | |
| "learning_rate": 1.7921624734111292e-05, | |
| "loss": 0.629, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.020253164556962, | |
| "grad_norm": 0.847409072947711, | |
| "learning_rate": 1.7825270686467567e-05, | |
| "loss": 0.5857, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.050632911392405, | |
| "grad_norm": 0.5407154896674116, | |
| "learning_rate": 1.7727005033820117e-05, | |
| "loss": 0.5681, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.081012658227848, | |
| "grad_norm": 0.818013284078951, | |
| "learning_rate": 1.762685178110382e-05, | |
| "loss": 0.5719, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.1113924050632913, | |
| "grad_norm": 0.6926650075090995, | |
| "learning_rate": 1.752483539436807e-05, | |
| "loss": 0.553, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.141772151898734, | |
| "grad_norm": 0.6639396402156127, | |
| "learning_rate": 1.7420980794800013e-05, | |
| "loss": 0.5707, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.1721518987341772, | |
| "grad_norm": 0.6774236998586597, | |
| "learning_rate": 1.731531335263669e-05, | |
| "loss": 0.5813, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 0.64013541137373, | |
| "learning_rate": 1.720785888096743e-05, | |
| "loss": 0.5767, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.232911392405063, | |
| "grad_norm": 0.6629462705529504, | |
| "learning_rate": 1.7098643629428035e-05, | |
| "loss": 0.5688, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.2632911392405064, | |
| "grad_norm": 0.6342907413552974, | |
| "learning_rate": 1.698769427778842e-05, | |
| "loss": 0.5627, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.293670886075949, | |
| "grad_norm": 0.626757162523397, | |
| "learning_rate": 1.687503792943506e-05, | |
| "loss": 0.5603, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.3240506329113924, | |
| "grad_norm": 0.6575665400172498, | |
| "learning_rate": 1.6760702104750046e-05, | |
| "loss": 0.5621, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.3544303797468356, | |
| "grad_norm": 0.6523794654328163, | |
| "learning_rate": 1.664471473438822e-05, | |
| "loss": 0.5612, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.3848101265822783, | |
| "grad_norm": 0.6478399779634784, | |
| "learning_rate": 1.6527104152454096e-05, | |
| "loss": 0.5575, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.4151898734177215, | |
| "grad_norm": 0.8759804495730426, | |
| "learning_rate": 1.6407899089580263e-05, | |
| "loss": 0.5528, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.4455696202531647, | |
| "grad_norm": 0.7117100436867281, | |
| "learning_rate": 1.628712866590885e-05, | |
| "loss": 0.5469, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.4759493670886075, | |
| "grad_norm": 0.7378885656373552, | |
| "learning_rate": 1.6164822383977912e-05, | |
| "loss": 0.5675, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 0.7957303838326724, | |
| "learning_rate": 1.604101012151436e-05, | |
| "loss": 0.5811, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.536708860759494, | |
| "grad_norm": 0.7573955891024758, | |
| "learning_rate": 1.5915722124135227e-05, | |
| "loss": 0.5497, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.5670886075949366, | |
| "grad_norm": 0.6708953227921186, | |
| "learning_rate": 1.5788988997959115e-05, | |
| "loss": 0.5544, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.59746835443038, | |
| "grad_norm": 0.850252331118449, | |
| "learning_rate": 1.5660841702129533e-05, | |
| "loss": 0.5622, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.6278481012658226, | |
| "grad_norm": 0.4676631825402915, | |
| "learning_rate": 1.5531311541251995e-05, | |
| "loss": 0.5485, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 2.6582278481012658, | |
| "grad_norm": 0.8444812709318328, | |
| "learning_rate": 1.540043015774676e-05, | |
| "loss": 0.547, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 2.688607594936709, | |
| "grad_norm": 0.5582278352975758, | |
| "learning_rate": 1.5268229524119007e-05, | |
| "loss": 0.5709, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 2.7189873417721517, | |
| "grad_norm": 0.5799848790857646, | |
| "learning_rate": 1.513474193514842e-05, | |
| "loss": 0.5706, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 2.749367088607595, | |
| "grad_norm": 0.5620679955414621, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.5672, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.779746835443038, | |
| "grad_norm": 0.5671006611265023, | |
| "learning_rate": 1.4864036634258112e-05, | |
| "loss": 0.5863, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 0.5686950439151207, | |
| "learning_rate": 1.4726885051885654e-05, | |
| "loss": 0.5691, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 2.840506329113924, | |
| "grad_norm": 0.47892905200146035, | |
| "learning_rate": 1.4588578757110359e-05, | |
| "loss": 0.5701, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.8708860759493673, | |
| "grad_norm": 0.5790354621980266, | |
| "learning_rate": 1.4449151536240167e-05, | |
| "loss": 0.5558, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.90126582278481, | |
| "grad_norm": 0.501662826949052, | |
| "learning_rate": 1.4308637449409705e-05, | |
| "loss": 0.5672, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.9316455696202532, | |
| "grad_norm": 3.4950046508673807, | |
| "learning_rate": 1.4167070822259868e-05, | |
| "loss": 0.5614, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.962025316455696, | |
| "grad_norm": 3.3438469861172364, | |
| "learning_rate": 1.402448623755254e-05, | |
| "loss": 0.5393, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.992405063291139, | |
| "grad_norm": 0.6859495868588724, | |
| "learning_rate": 1.3880918526722497e-05, | |
| "loss": 0.5532, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.030379746835443, | |
| "grad_norm": 0.7521360747329994, | |
| "learning_rate": 1.3736402761368597e-05, | |
| "loss": 0.5193, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.060759493670886, | |
| "grad_norm": 0.6576809722331366, | |
| "learning_rate": 1.3590974244686248e-05, | |
| "loss": 0.4932, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.091139240506329, | |
| "grad_norm": 0.7942760946436969, | |
| "learning_rate": 1.344466850284333e-05, | |
| "loss": 0.5025, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 3.1215189873417724, | |
| "grad_norm": 0.7170996802661206, | |
| "learning_rate": 1.3297521276301666e-05, | |
| "loss": 0.5014, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 3.151898734177215, | |
| "grad_norm": 0.6576747862254556, | |
| "learning_rate": 1.3149568511086104e-05, | |
| "loss": 0.502, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 3.1822784810126583, | |
| "grad_norm": 0.6599875667844219, | |
| "learning_rate": 1.300084635000341e-05, | |
| "loss": 0.5195, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 3.212658227848101, | |
| "grad_norm": 0.5439541226868704, | |
| "learning_rate": 1.2851391123813075e-05, | |
| "loss": 0.505, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 3.2430379746835443, | |
| "grad_norm": 0.6316175841130693, | |
| "learning_rate": 1.2701239342352223e-05, | |
| "loss": 0.5103, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 3.2734177215189875, | |
| "grad_norm": 0.5837063641790433, | |
| "learning_rate": 1.2550427685616767e-05, | |
| "loss": 0.5124, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 3.3037974683544302, | |
| "grad_norm": 0.6994289458844141, | |
| "learning_rate": 1.239899299480098e-05, | |
| "loss": 0.5106, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 3.3341772151898734, | |
| "grad_norm": 0.6457846026508504, | |
| "learning_rate": 1.2246972263297718e-05, | |
| "loss": 0.5037, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 3.3645569620253166, | |
| "grad_norm": 0.6070849280063823, | |
| "learning_rate": 1.2094402627661447e-05, | |
| "loss": 0.5228, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 3.3949367088607594, | |
| "grad_norm": 0.625501181676829, | |
| "learning_rate": 1.1941321358536278e-05, | |
| "loss": 0.5273, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 3.4253164556962026, | |
| "grad_norm": 0.602055007307549, | |
| "learning_rate": 1.1787765851551296e-05, | |
| "loss": 0.4798, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 3.4556962025316453, | |
| "grad_norm": 0.49853804980255195, | |
| "learning_rate": 1.1633773618185302e-05, | |
| "loss": 0.507, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 3.4860759493670885, | |
| "grad_norm": 0.6231379080242243, | |
| "learning_rate": 1.14793822766033e-05, | |
| "loss": 0.478, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 3.5164556962025317, | |
| "grad_norm": 0.48924371668798267, | |
| "learning_rate": 1.132462954246688e-05, | |
| "loss": 0.5019, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 3.546835443037975, | |
| "grad_norm": 0.5955862004601938, | |
| "learning_rate": 1.1169553219720828e-05, | |
| "loss": 0.4937, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 3.5772151898734177, | |
| "grad_norm": 0.49509931576942456, | |
| "learning_rate": 1.1014191191358118e-05, | |
| "loss": 0.5107, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 3.607594936708861, | |
| "grad_norm": 0.44896078822877394, | |
| "learning_rate": 1.085858141016566e-05, | |
| "loss": 0.4958, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 3.6379746835443036, | |
| "grad_norm": 0.49392910720016003, | |
| "learning_rate": 1.070276188945293e-05, | |
| "loss": 0.5018, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 3.668354430379747, | |
| "grad_norm": 0.38365341221057153, | |
| "learning_rate": 1.0546770693765859e-05, | |
| "loss": 0.5047, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.69873417721519, | |
| "grad_norm": 0.4188829547619208, | |
| "learning_rate": 1.0390645929588197e-05, | |
| "loss": 0.4935, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 3.729113924050633, | |
| "grad_norm": 0.4277619227419128, | |
| "learning_rate": 1.0234425736032607e-05, | |
| "loss": 0.4941, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 3.759493670886076, | |
| "grad_norm": 0.397738634192115, | |
| "learning_rate": 1.007814827552384e-05, | |
| "loss": 0.5013, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 3.7898734177215188, | |
| "grad_norm": 0.4650378596323948, | |
| "learning_rate": 9.92185172447616e-06, | |
| "loss": 0.5118, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 3.820253164556962, | |
| "grad_norm": 0.4428572274275131, | |
| "learning_rate": 9.765574263967397e-06, | |
| "loss": 0.4981, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 3.850632911392405, | |
| "grad_norm": 0.5605146355587542, | |
| "learning_rate": 9.609354070411807e-06, | |
| "loss": 0.4737, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 3.8810126582278484, | |
| "grad_norm": 0.4252528572786787, | |
| "learning_rate": 9.453229306234143e-06, | |
| "loss": 0.4813, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 3.911392405063291, | |
| "grad_norm": 0.4420897920210455, | |
| "learning_rate": 9.297238110547075e-06, | |
| "loss": 0.5146, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 3.9417721518987343, | |
| "grad_norm": 0.42303157380834855, | |
| "learning_rate": 9.14141858983434e-06, | |
| "loss": 0.4689, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 3.972151898734177, | |
| "grad_norm": 0.4120609905793025, | |
| "learning_rate": 8.985808808641883e-06, | |
| "loss": 0.5002, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 4.010126582278481, | |
| "grad_norm": 0.44182601746536293, | |
| "learning_rate": 8.830446780279175e-06, | |
| "loss": 0.5119, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 4.040506329113924, | |
| "grad_norm": 0.5089833058435673, | |
| "learning_rate": 8.675370457533122e-06, | |
| "loss": 0.4309, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 4.0708860759493675, | |
| "grad_norm": 0.48165083481136944, | |
| "learning_rate": 8.520617723396702e-06, | |
| "loss": 0.4589, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 4.10126582278481, | |
| "grad_norm": 0.5618872632126407, | |
| "learning_rate": 8.366226381814698e-06, | |
| "loss": 0.4355, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 4.131645569620253, | |
| "grad_norm": 0.42356324010728974, | |
| "learning_rate": 8.212234148448708e-06, | |
| "loss": 0.4588, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 4.162025316455696, | |
| "grad_norm": 0.49508953245119475, | |
| "learning_rate": 8.058678641463724e-06, | |
| "loss": 0.4725, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 4.192405063291139, | |
| "grad_norm": 0.3923564426051679, | |
| "learning_rate": 7.905597372338558e-06, | |
| "loss": 0.4474, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 4.222784810126583, | |
| "grad_norm": 0.4301745327088533, | |
| "learning_rate": 7.753027736702283e-06, | |
| "loss": 0.4406, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 4.253164556962025, | |
| "grad_norm": 0.4714740782880516, | |
| "learning_rate": 7.601007005199022e-06, | |
| "loss": 0.4465, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 4.283544303797468, | |
| "grad_norm": 0.4425577671413435, | |
| "learning_rate": 7.449572314383237e-06, | |
| "loss": 0.4566, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 4.313924050632911, | |
| "grad_norm": 0.3849039481457668, | |
| "learning_rate": 7.298760657647779e-06, | |
| "loss": 0.4483, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 4.3443037974683545, | |
| "grad_norm": 0.41365363233224023, | |
| "learning_rate": 7.148608876186931e-06, | |
| "loss": 0.4386, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 4.374683544303798, | |
| "grad_norm": 0.4178017025345433, | |
| "learning_rate": 6.999153649996595e-06, | |
| "loss": 0.4569, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 4.405063291139241, | |
| "grad_norm": 0.39367373984053783, | |
| "learning_rate": 6.8504314889138956e-06, | |
| "loss": 0.4397, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 4.435443037974683, | |
| "grad_norm": 0.39628796802515664, | |
| "learning_rate": 6.702478723698336e-06, | |
| "loss": 0.4523, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 4.465822784810126, | |
| "grad_norm": 0.40668096801195014, | |
| "learning_rate": 6.555331497156671e-06, | |
| "loss": 0.4826, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 4.49620253164557, | |
| "grad_norm": 0.4167902385322104, | |
| "learning_rate": 6.4090257553137566e-06, | |
| "loss": 0.449, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 4.526582278481013, | |
| "grad_norm": 0.40634157550268557, | |
| "learning_rate": 6.263597238631405e-06, | |
| "loss": 0.455, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 4.556962025316456, | |
| "grad_norm": 0.357518536806457, | |
| "learning_rate": 6.119081473277502e-06, | |
| "loss": 0.4441, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 4.587341772151898, | |
| "grad_norm": 0.3491865395204641, | |
| "learning_rate": 5.975513762447465e-06, | |
| "loss": 0.4114, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 4.6177215189873415, | |
| "grad_norm": 0.4050579553722853, | |
| "learning_rate": 5.832929177740134e-06, | |
| "loss": 0.4199, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 4.648101265822785, | |
| "grad_norm": 0.3916631906846261, | |
| "learning_rate": 5.6913625505902966e-06, | |
| "loss": 0.4605, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 4.678481012658228, | |
| "grad_norm": 0.38490848485694573, | |
| "learning_rate": 5.550848463759835e-06, | |
| "loss": 0.444, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 4.708860759493671, | |
| "grad_norm": 0.374864639240743, | |
| "learning_rate": 5.411421242889643e-06, | |
| "loss": 0.448, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 4.739240506329114, | |
| "grad_norm": 0.3917812536676905, | |
| "learning_rate": 5.273114948114346e-06, | |
| "loss": 0.4696, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 4.769620253164557, | |
| "grad_norm": 0.3823077908754771, | |
| "learning_rate": 5.135963365741892e-06, | |
| "loss": 0.4594, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.36813688549395657, | |
| "learning_rate": 5.000000000000003e-06, | |
| "loss": 0.465, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 4.830379746835443, | |
| "grad_norm": 0.40011519702438747, | |
| "learning_rate": 4.865258064851579e-06, | |
| "loss": 0.4637, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 4.860759493670886, | |
| "grad_norm": 0.3436544082342494, | |
| "learning_rate": 4.731770475880995e-06, | |
| "loss": 0.4663, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 4.891139240506329, | |
| "grad_norm": 0.34360454730891277, | |
| "learning_rate": 4.599569842253244e-06, | |
| "loss": 0.4594, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.921518987341772, | |
| "grad_norm": 0.3700652364382893, | |
| "learning_rate": 4.468688458748006e-06, | |
| "loss": 0.4197, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 4.951898734177215, | |
| "grad_norm": 0.371826825349328, | |
| "learning_rate": 4.339158297870469e-06, | |
| "loss": 0.4841, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 4.982278481012658, | |
| "grad_norm": 0.35445702313966165, | |
| "learning_rate": 4.211011002040885e-06, | |
| "loss": 0.4298, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 5.020253164556962, | |
| "grad_norm": 0.40399266704685155, | |
| "learning_rate": 4.084277875864776e-06, | |
| "loss": 0.4481, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 5.050632911392405, | |
| "grad_norm": 0.4740934136180896, | |
| "learning_rate": 3.958989878485644e-06, | |
| "loss": 0.4047, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 5.0810126582278485, | |
| "grad_norm": 0.32680870921385735, | |
| "learning_rate": 3.83517761602209e-06, | |
| "loss": 0.386, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 5.111392405063291, | |
| "grad_norm": 0.42299081072789346, | |
| "learning_rate": 3.712871334091154e-06, | |
| "loss": 0.441, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 5.141772151898734, | |
| "grad_norm": 0.5604785771360025, | |
| "learning_rate": 3.592100910419738e-06, | |
| "loss": 0.4259, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 5.172151898734177, | |
| "grad_norm": 0.34838811231973854, | |
| "learning_rate": 3.4728958475459052e-06, | |
| "loss": 0.4243, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 5.2025316455696204, | |
| "grad_norm": 0.4098355590589843, | |
| "learning_rate": 3.355285265611784e-06, | |
| "loss": 0.4534, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 5.232911392405064, | |
| "grad_norm": 0.40050297271934493, | |
| "learning_rate": 3.2392978952499553e-06, | |
| "loss": 0.4323, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 5.263291139240506, | |
| "grad_norm": 0.3896489746805148, | |
| "learning_rate": 3.1249620705649417e-06, | |
| "loss": 0.3991, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 5.293670886075949, | |
| "grad_norm": 0.3410028925336058, | |
| "learning_rate": 3.0123057222115835e-06, | |
| "loss": 0.4253, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 5.324050632911392, | |
| "grad_norm": 0.35525336326312473, | |
| "learning_rate": 2.9013563705719673e-06, | |
| "loss": 0.4269, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 5.3544303797468356, | |
| "grad_norm": 0.38868130412224144, | |
| "learning_rate": 2.7921411190325753e-06, | |
| "loss": 0.4099, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 5.384810126582279, | |
| "grad_norm": 0.3790599699590579, | |
| "learning_rate": 2.6846866473633126e-06, | |
| "loss": 0.4395, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 5.415189873417722, | |
| "grad_norm": 0.3474226546140671, | |
| "learning_rate": 2.579019205199992e-06, | |
| "loss": 0.3995, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 5.445569620253164, | |
| "grad_norm": 0.31650392331797883, | |
| "learning_rate": 2.4751646056319334e-06, | |
| "loss": 0.4358, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 5.4759493670886075, | |
| "grad_norm": 0.31845326576263716, | |
| "learning_rate": 2.373148218896182e-06, | |
| "loss": 0.4456, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 5.506329113924051, | |
| "grad_norm": 0.367102520267415, | |
| "learning_rate": 2.2729949661798876e-06, | |
| "loss": 0.3943, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 5.536708860759494, | |
| "grad_norm": 0.3417614714382495, | |
| "learning_rate": 2.174729313532433e-06, | |
| "loss": 0.4191, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 5.567088607594937, | |
| "grad_norm": 0.3515069834690961, | |
| "learning_rate": 2.078375265888707e-06, | |
| "loss": 0.3655, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 5.597468354430379, | |
| "grad_norm": 0.2988245398228493, | |
| "learning_rate": 1.9839563612050273e-06, | |
| "loss": 0.3824, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 5.627848101265823, | |
| "grad_norm": 0.3246033190500736, | |
| "learning_rate": 1.8914956647091497e-06, | |
| "loss": 0.4008, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 5.658227848101266, | |
| "grad_norm": 0.3273898529922116, | |
| "learning_rate": 1.8010157632657544e-06, | |
| "loss": 0.4063, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 5.688607594936709, | |
| "grad_norm": 0.3486645039001246, | |
| "learning_rate": 1.7125387598587862e-06, | |
| "loss": 0.3852, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 5.718987341772152, | |
| "grad_norm": 0.32532001109704317, | |
| "learning_rate": 1.6260862681919965e-06, | |
| "loss": 0.4146, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 5.749367088607595, | |
| "grad_norm": 0.30830008513478885, | |
| "learning_rate": 1.5416794074090258e-06, | |
| "loss": 0.3808, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 5.779746835443038, | |
| "grad_norm": 0.3282318164089578, | |
| "learning_rate": 1.459338796934293e-06, | |
| "loss": 0.4047, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 5.810126582278481, | |
| "grad_norm": 0.28795053482465843, | |
| "learning_rate": 1.3790845514359363e-06, | |
| "loss": 0.4157, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 5.840506329113924, | |
| "grad_norm": 0.3023610731098569, | |
| "learning_rate": 1.300936275912098e-06, | |
| "loss": 0.3977, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 5.870886075949367, | |
| "grad_norm": 0.2961959609038373, | |
| "learning_rate": 1.224913060901688e-06, | |
| "loss": 0.4201, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 5.9012658227848105, | |
| "grad_norm": 0.2850581846328519, | |
| "learning_rate": 1.1510334778208332e-06, | |
| "loss": 0.4363, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 5.931645569620253, | |
| "grad_norm": 0.3106810269371866, | |
| "learning_rate": 1.0793155744261352e-06, | |
| "loss": 0.4455, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 5.962025316455696, | |
| "grad_norm": 0.28156915746358596, | |
| "learning_rate": 1.0097768704058542e-06, | |
| "loss": 0.4538, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 5.992405063291139, | |
| "grad_norm": 0.29361360200944137, | |
| "learning_rate": 9.424343531000968e-07, | |
| "loss": 0.4218, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 6.030379746835443, | |
| "grad_norm": 0.3212717521002966, | |
| "learning_rate": 8.773044733510338e-07, | |
| "loss": 0.3992, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 6.060759493670886, | |
| "grad_norm": 0.2997665370706395, | |
| "learning_rate": 8.144031414842012e-07, | |
| "loss": 0.4266, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 6.091139240506329, | |
| "grad_norm": 0.2804862733482462, | |
| "learning_rate": 7.537457234218271e-07, | |
| "loss": 0.4318, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 6.121518987341772, | |
| "grad_norm": 0.2619466344637825, | |
| "learning_rate": 6.953470369291349e-07, | |
| "loss": 0.3855, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 6.151898734177215, | |
| "grad_norm": 0.2854071711419194, | |
| "learning_rate": 6.392213479945852e-07, | |
| "loss": 0.3982, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 6.182278481012658, | |
| "grad_norm": 0.27289043298395305, | |
| "learning_rate": 5.853823673448877e-07, | |
| "loss": 0.3997, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 6.2126582278481015, | |
| "grad_norm": 0.27605498623386876, | |
| "learning_rate": 5.33843247095659e-07, | |
| "loss": 0.4078, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 6.243037974683545, | |
| "grad_norm": 0.30000471255697597, | |
| "learning_rate": 4.846165775385459e-07, | |
| "loss": 0.4125, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 6.273417721518987, | |
| "grad_norm": 0.27160370915625476, | |
| "learning_rate": 4.3771438406559173e-07, | |
| "loss": 0.3987, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 6.30379746835443, | |
| "grad_norm": 0.29737724099989066, | |
| "learning_rate": 3.931481242315993e-07, | |
| "loss": 0.3957, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 6.334177215189873, | |
| "grad_norm": 0.2810346634829443, | |
| "learning_rate": 3.5092868495520294e-07, | |
| "loss": 0.4177, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 6.364556962025317, | |
| "grad_norm": 0.2625025357055661, | |
| "learning_rate": 3.110663798593616e-07, | |
| "loss": 0.3776, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 6.39493670886076, | |
| "grad_norm": 0.2697091654844699, | |
| "learning_rate": 2.735709467518699e-07, | |
| "loss": 0.3958, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 6.425316455696202, | |
| "grad_norm": 0.27391077969336497, | |
| "learning_rate": 2.384515452465475e-07, | |
| "loss": 0.4096, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 6.455696202531645, | |
| "grad_norm": 0.25145925030805255, | |
| "learning_rate": 2.0571675452567997e-07, | |
| "loss": 0.4089, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 6.4860759493670885, | |
| "grad_norm": 0.25183514894064907, | |
| "learning_rate": 1.7537457124423896e-07, | |
| "loss": 0.4003, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 6.516455696202532, | |
| "grad_norm": 0.2551574925821408, | |
| "learning_rate": 1.474324075764111e-07, | |
| "loss": 0.393, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 6.546835443037975, | |
| "grad_norm": 0.2620509858849196, | |
| "learning_rate": 1.2189708940490653e-07, | |
| "loss": 0.4313, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 6.577215189873417, | |
| "grad_norm": 0.29065497157612546, | |
| "learning_rate": 9.877485465349057e-08, | |
| "loss": 0.4221, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 6.6075949367088604, | |
| "grad_norm": 0.2743875887530047, | |
| "learning_rate": 7.807135176314707e-08, | |
| "loss": 0.3746, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 6.637974683544304, | |
| "grad_norm": 0.2870484872009282, | |
| "learning_rate": 5.979163831223988e-08, | |
| "loss": 0.3625, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 6.668354430379747, | |
| "grad_norm": 0.2554830495931789, | |
| "learning_rate": 4.394017978101905e-08, | |
| "loss": 0.4116, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 6.69873417721519, | |
| "grad_norm": 0.26444347681445124, | |
| "learning_rate": 3.0520848460765525e-08, | |
| "loss": 0.3864, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 6.729113924050633, | |
| "grad_norm": 0.25798469506071947, | |
| "learning_rate": 1.9536922507841227e-08, | |
| "loss": 0.4204, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 6.759493670886076, | |
| "grad_norm": 0.2557364276643878, | |
| "learning_rate": 1.099108514288627e-08, | |
| "loss": 0.3968, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 6.789873417721519, | |
| "grad_norm": 0.264462317402514, | |
| "learning_rate": 4.885423995341088e-09, | |
| "loss": 0.4067, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 6.820253164556962, | |
| "grad_norm": 0.2634380681550479, | |
| "learning_rate": 1.2214305934699078e-09, | |
| "loss": 0.3949, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 6.850632911392405, | |
| "grad_norm": 0.2518720301251443, | |
| "learning_rate": 0.0, | |
| "loss": 0.4155, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 6.850632911392405, | |
| "step": 224, | |
| "total_flos": 8.169716519157105e+17, | |
| "train_loss": 0.5474886966070959, | |
| "train_runtime": 22562.2646, | |
| "train_samples_per_second": 0.98, | |
| "train_steps_per_second": 0.01 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 224, | |
| "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": 8.169716519157105e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |