| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9967051070840198, | |
| "eval_steps": 500, | |
| "global_step": 606, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0032948929159802307, | |
| "grad_norm": 12.442080297711913, | |
| "learning_rate": 1.639344262295082e-07, | |
| "loss": 0.383, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006589785831960461, | |
| "grad_norm": 13.440096974039598, | |
| "learning_rate": 3.278688524590164e-07, | |
| "loss": 0.3663, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.009884678747940691, | |
| "grad_norm": 13.796855251219549, | |
| "learning_rate": 4.918032786885246e-07, | |
| "loss": 0.3663, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.013179571663920923, | |
| "grad_norm": 12.70572090942759, | |
| "learning_rate": 6.557377049180328e-07, | |
| "loss": 0.3061, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.016474464579901153, | |
| "grad_norm": 11.468076158597158, | |
| "learning_rate": 8.196721311475409e-07, | |
| "loss": 0.3387, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.019769357495881382, | |
| "grad_norm": 10.302737632617346, | |
| "learning_rate": 9.836065573770493e-07, | |
| "loss": 0.2814, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.023064250411861616, | |
| "grad_norm": 9.415064505565756, | |
| "learning_rate": 1.1475409836065575e-06, | |
| "loss": 0.2815, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.026359143327841845, | |
| "grad_norm": 8.420756675130125, | |
| "learning_rate": 1.3114754098360657e-06, | |
| "loss": 0.2483, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.029654036243822075, | |
| "grad_norm": 5.932373113068906, | |
| "learning_rate": 1.4754098360655739e-06, | |
| "loss": 0.1744, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.032948929159802305, | |
| "grad_norm": 4.602925255551132, | |
| "learning_rate": 1.6393442622950819e-06, | |
| "loss": 0.1821, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.036243822075782535, | |
| "grad_norm": 4.1437512962542025, | |
| "learning_rate": 1.8032786885245903e-06, | |
| "loss": 0.1557, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.039538714991762765, | |
| "grad_norm": 4.1008407646409495, | |
| "learning_rate": 1.9672131147540985e-06, | |
| "loss": 0.1901, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.042833607907743, | |
| "grad_norm": 3.2895218176449177, | |
| "learning_rate": 2.1311475409836067e-06, | |
| "loss": 0.1695, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.04612850082372323, | |
| "grad_norm": 2.578808113814672, | |
| "learning_rate": 2.295081967213115e-06, | |
| "loss": 0.1431, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.04942339373970346, | |
| "grad_norm": 2.0925168309074804, | |
| "learning_rate": 2.459016393442623e-06, | |
| "loss": 0.1459, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.05271828665568369, | |
| "grad_norm": 5.776302012227753, | |
| "learning_rate": 2.6229508196721314e-06, | |
| "loss": 0.2116, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.05601317957166392, | |
| "grad_norm": 4.863535909815622, | |
| "learning_rate": 2.786885245901639e-06, | |
| "loss": 0.157, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.05930807248764415, | |
| "grad_norm": 2.709665323729833, | |
| "learning_rate": 2.9508196721311478e-06, | |
| "loss": 0.1463, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.06260296540362438, | |
| "grad_norm": 2.911144968708877, | |
| "learning_rate": 3.114754098360656e-06, | |
| "loss": 0.1283, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.06589785831960461, | |
| "grad_norm": 2.4985115549367873, | |
| "learning_rate": 3.2786885245901638e-06, | |
| "loss": 0.13, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06919275123558484, | |
| "grad_norm": 2.158826062926261, | |
| "learning_rate": 3.4426229508196724e-06, | |
| "loss": 0.1135, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.07248764415156507, | |
| "grad_norm": 1.8839820680217123, | |
| "learning_rate": 3.6065573770491806e-06, | |
| "loss": 0.1196, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0757825370675453, | |
| "grad_norm": 2.208880925801888, | |
| "learning_rate": 3.7704918032786884e-06, | |
| "loss": 0.1642, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.07907742998352553, | |
| "grad_norm": 1.948969438238695, | |
| "learning_rate": 3.934426229508197e-06, | |
| "loss": 0.1057, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.08237232289950576, | |
| "grad_norm": 1.5855531759235064, | |
| "learning_rate": 4.098360655737705e-06, | |
| "loss": 0.0908, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.085667215815486, | |
| "grad_norm": 1.7484734055235995, | |
| "learning_rate": 4.2622950819672135e-06, | |
| "loss": 0.1088, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.08896210873146623, | |
| "grad_norm": 1.7763409124600038, | |
| "learning_rate": 4.426229508196722e-06, | |
| "loss": 0.1264, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.09225700164744646, | |
| "grad_norm": 1.943799751354267, | |
| "learning_rate": 4.59016393442623e-06, | |
| "loss": 0.1204, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.09555189456342669, | |
| "grad_norm": 2.0024112586646723, | |
| "learning_rate": 4.754098360655738e-06, | |
| "loss": 0.132, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.09884678747940692, | |
| "grad_norm": 1.927718117469908, | |
| "learning_rate": 4.918032786885246e-06, | |
| "loss": 0.0988, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.10214168039538715, | |
| "grad_norm": 1.8175857487104838, | |
| "learning_rate": 5.0819672131147545e-06, | |
| "loss": 0.0997, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.10543657331136738, | |
| "grad_norm": 1.1821292523186457, | |
| "learning_rate": 5.245901639344263e-06, | |
| "loss": 0.0874, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.10873146622734761, | |
| "grad_norm": 1.6925074350020268, | |
| "learning_rate": 5.409836065573772e-06, | |
| "loss": 0.115, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.11202635914332784, | |
| "grad_norm": 1.4637368762611331, | |
| "learning_rate": 5.573770491803278e-06, | |
| "loss": 0.0916, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.11532125205930807, | |
| "grad_norm": 1.4814671153620174, | |
| "learning_rate": 5.737704918032787e-06, | |
| "loss": 0.0853, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.1186161449752883, | |
| "grad_norm": 1.243594463126339, | |
| "learning_rate": 5.9016393442622956e-06, | |
| "loss": 0.0903, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.12191103789126853, | |
| "grad_norm": 1.2713537957193175, | |
| "learning_rate": 6.065573770491804e-06, | |
| "loss": 0.1271, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.12520593080724876, | |
| "grad_norm": 2.2092835366807893, | |
| "learning_rate": 6.229508196721312e-06, | |
| "loss": 0.1137, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.128500823723229, | |
| "grad_norm": 1.1846569097593065, | |
| "learning_rate": 6.393442622950821e-06, | |
| "loss": 0.0714, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.13179571663920922, | |
| "grad_norm": 3.001321928490275, | |
| "learning_rate": 6.5573770491803276e-06, | |
| "loss": 0.1129, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.13509060955518945, | |
| "grad_norm": 1.8758843455974739, | |
| "learning_rate": 6.721311475409837e-06, | |
| "loss": 0.1021, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.13838550247116968, | |
| "grad_norm": 2.993102960488022, | |
| "learning_rate": 6.885245901639345e-06, | |
| "loss": 0.1155, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.1416803953871499, | |
| "grad_norm": 1.8441651337946723, | |
| "learning_rate": 7.049180327868853e-06, | |
| "loss": 0.1004, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.14497528830313014, | |
| "grad_norm": 1.495517808358825, | |
| "learning_rate": 7.213114754098361e-06, | |
| "loss": 0.0868, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.14827018121911037, | |
| "grad_norm": 1.8953155374136303, | |
| "learning_rate": 7.3770491803278695e-06, | |
| "loss": 0.1371, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1515650741350906, | |
| "grad_norm": 1.702133714045992, | |
| "learning_rate": 7.540983606557377e-06, | |
| "loss": 0.1597, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.15485996705107083, | |
| "grad_norm": 1.0196937438072402, | |
| "learning_rate": 7.704918032786886e-06, | |
| "loss": 0.0892, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.15815485996705106, | |
| "grad_norm": 1.2171514922994324, | |
| "learning_rate": 7.868852459016394e-06, | |
| "loss": 0.0778, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.1614497528830313, | |
| "grad_norm": 1.0680772396455713, | |
| "learning_rate": 8.032786885245902e-06, | |
| "loss": 0.084, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.16474464579901152, | |
| "grad_norm": 2.9252478676320557, | |
| "learning_rate": 8.19672131147541e-06, | |
| "loss": 0.1275, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.16803953871499178, | |
| "grad_norm": 2.422192715088673, | |
| "learning_rate": 8.360655737704919e-06, | |
| "loss": 0.1473, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.171334431630972, | |
| "grad_norm": 1.8804479896519768, | |
| "learning_rate": 8.524590163934427e-06, | |
| "loss": 0.1036, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.17462932454695224, | |
| "grad_norm": 1.568191135105415, | |
| "learning_rate": 8.688524590163935e-06, | |
| "loss": 0.1109, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.17792421746293247, | |
| "grad_norm": 1.7822383034585774, | |
| "learning_rate": 8.852459016393443e-06, | |
| "loss": 0.1138, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.1812191103789127, | |
| "grad_norm": 2.39128847516323, | |
| "learning_rate": 9.016393442622952e-06, | |
| "loss": 0.0995, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.18451400329489293, | |
| "grad_norm": 2.005472201286874, | |
| "learning_rate": 9.18032786885246e-06, | |
| "loss": 0.1071, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.18780889621087316, | |
| "grad_norm": 1.771963789348757, | |
| "learning_rate": 9.344262295081968e-06, | |
| "loss": 0.1194, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.19110378912685339, | |
| "grad_norm": 1.9883587282426103, | |
| "learning_rate": 9.508196721311476e-06, | |
| "loss": 0.1263, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.19439868204283361, | |
| "grad_norm": 1.869337708289752, | |
| "learning_rate": 9.672131147540984e-06, | |
| "loss": 0.0906, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.19769357495881384, | |
| "grad_norm": 1.95284840649176, | |
| "learning_rate": 9.836065573770493e-06, | |
| "loss": 0.0941, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.20098846787479407, | |
| "grad_norm": 0.9408728760921459, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0892, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.2042833607907743, | |
| "grad_norm": 1.4827130661563888, | |
| "learning_rate": 9.999916929744365e-06, | |
| "loss": 0.0897, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.20757825370675453, | |
| "grad_norm": 1.7548705917176584, | |
| "learning_rate": 9.999667721737726e-06, | |
| "loss": 0.1052, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.21087314662273476, | |
| "grad_norm": 1.5160957230364678, | |
| "learning_rate": 9.999252384260794e-06, | |
| "loss": 0.0959, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.214168039538715, | |
| "grad_norm": 1.4961421846399192, | |
| "learning_rate": 9.998670931114443e-06, | |
| "loss": 0.0738, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.21746293245469522, | |
| "grad_norm": 1.6241097546909007, | |
| "learning_rate": 9.997923381619257e-06, | |
| "loss": 0.1057, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.22075782537067545, | |
| "grad_norm": 1.3354496262353763, | |
| "learning_rate": 9.99700976061489e-06, | |
| "loss": 0.0951, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.22405271828665568, | |
| "grad_norm": 1.0847564491580965, | |
| "learning_rate": 9.99593009845923e-06, | |
| "loss": 0.0863, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.2273476112026359, | |
| "grad_norm": 1.6871231023520186, | |
| "learning_rate": 9.994684431027407e-06, | |
| "loss": 0.0804, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.23064250411861614, | |
| "grad_norm": 1.428280535176863, | |
| "learning_rate": 9.99327279971058e-06, | |
| "loss": 0.0865, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.23393739703459637, | |
| "grad_norm": 1.1179352637786575, | |
| "learning_rate": 9.991695251414584e-06, | |
| "loss": 0.0734, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.2372322899505766, | |
| "grad_norm": 1.2033588488413391, | |
| "learning_rate": 9.989951838558352e-06, | |
| "loss": 0.105, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.24052718286655683, | |
| "grad_norm": 1.465169988779563, | |
| "learning_rate": 9.988042619072185e-06, | |
| "loss": 0.0729, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.24382207578253706, | |
| "grad_norm": 0.8979589269956051, | |
| "learning_rate": 9.985967656395823e-06, | |
| "loss": 0.0802, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.2471169686985173, | |
| "grad_norm": 1.733084591549866, | |
| "learning_rate": 9.98372701947634e-06, | |
| "loss": 0.1105, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2504118616144975, | |
| "grad_norm": 1.3264470296840676, | |
| "learning_rate": 9.981320782765847e-06, | |
| "loss": 0.0994, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.25370675453047775, | |
| "grad_norm": 1.6544247928428517, | |
| "learning_rate": 9.978749026219023e-06, | |
| "loss": 0.0729, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.257001647446458, | |
| "grad_norm": 1.7379037158314299, | |
| "learning_rate": 9.976011835290457e-06, | |
| "loss": 0.1084, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.2602965403624382, | |
| "grad_norm": 2.1898271902712465, | |
| "learning_rate": 9.973109300931813e-06, | |
| "loss": 0.14, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.26359143327841844, | |
| "grad_norm": 1.5180418313244188, | |
| "learning_rate": 9.970041519588797e-06, | |
| "loss": 0.1032, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.26688632619439867, | |
| "grad_norm": 0.9058107350669203, | |
| "learning_rate": 9.966808593197959e-06, | |
| "loss": 0.0659, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.2701812191103789, | |
| "grad_norm": 1.6740173453146032, | |
| "learning_rate": 9.963410629183311e-06, | |
| "loss": 0.0903, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.27347611202635913, | |
| "grad_norm": 1.6927879596763102, | |
| "learning_rate": 9.959847740452746e-06, | |
| "loss": 0.1011, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.27677100494233936, | |
| "grad_norm": 1.1858716785831125, | |
| "learning_rate": 9.956120045394297e-06, | |
| "loss": 0.1001, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.2800658978583196, | |
| "grad_norm": 1.7472793459338325, | |
| "learning_rate": 9.952227667872197e-06, | |
| "loss": 0.0999, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.2833607907742998, | |
| "grad_norm": 1.3355726739748197, | |
| "learning_rate": 9.948170737222763e-06, | |
| "loss": 0.0869, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.28665568369028005, | |
| "grad_norm": 1.6554862377061288, | |
| "learning_rate": 9.943949388250102e-06, | |
| "loss": 0.0956, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.2899505766062603, | |
| "grad_norm": 1.4959889750287383, | |
| "learning_rate": 9.939563761221628e-06, | |
| "loss": 0.1011, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.2932454695222405, | |
| "grad_norm": 1.90960683038771, | |
| "learning_rate": 9.935014001863405e-06, | |
| "loss": 0.086, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.29654036243822074, | |
| "grad_norm": 1.33889912510083, | |
| "learning_rate": 9.930300261355305e-06, | |
| "loss": 0.0884, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.29983525535420097, | |
| "grad_norm": 1.8364942381661888, | |
| "learning_rate": 9.925422696325976e-06, | |
| "loss": 0.1198, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.3031301482701812, | |
| "grad_norm": 0.9853906443215683, | |
| "learning_rate": 9.920381468847648e-06, | |
| "loss": 0.0805, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.30642504118616143, | |
| "grad_norm": 1.802022784884664, | |
| "learning_rate": 9.915176746430746e-06, | |
| "loss": 0.1, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.30971993410214166, | |
| "grad_norm": 1.6019896189177425, | |
| "learning_rate": 9.909808702018315e-06, | |
| "loss": 0.1063, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.3130148270181219, | |
| "grad_norm": 1.7270522226909808, | |
| "learning_rate": 9.904277513980285e-06, | |
| "loss": 0.1009, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.3163097199341021, | |
| "grad_norm": 1.355594119054743, | |
| "learning_rate": 9.898583366107539e-06, | |
| "loss": 0.0875, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.31960461285008235, | |
| "grad_norm": 1.8076099955595957, | |
| "learning_rate": 9.892726447605803e-06, | |
| "loss": 0.1236, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3228995057660626, | |
| "grad_norm": 1.6099906623254512, | |
| "learning_rate": 9.886706953089364e-06, | |
| "loss": 0.0873, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.3261943986820428, | |
| "grad_norm": 1.1035445355215792, | |
| "learning_rate": 9.880525082574604e-06, | |
| "loss": 0.0869, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.32948929159802304, | |
| "grad_norm": 1.4670801213842743, | |
| "learning_rate": 9.874181041473344e-06, | |
| "loss": 0.1076, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.33278418451400327, | |
| "grad_norm": 1.1434779708887954, | |
| "learning_rate": 9.867675040586035e-06, | |
| "loss": 0.0987, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.33607907742998355, | |
| "grad_norm": 1.2329688966692105, | |
| "learning_rate": 9.861007296094736e-06, | |
| "loss": 0.0685, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.3393739703459638, | |
| "grad_norm": 1.33973829264815, | |
| "learning_rate": 9.854178029555945e-06, | |
| "loss": 0.0926, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.342668863261944, | |
| "grad_norm": 2.4442395017358565, | |
| "learning_rate": 9.847187467893228e-06, | |
| "loss": 0.0942, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.34596375617792424, | |
| "grad_norm": 1.5545572705565789, | |
| "learning_rate": 9.840035843389684e-06, | |
| "loss": 0.0849, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.34925864909390447, | |
| "grad_norm": 0.8361181058922856, | |
| "learning_rate": 9.832723393680222e-06, | |
| "loss": 0.0678, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.3525535420098847, | |
| "grad_norm": 1.1875767399598742, | |
| "learning_rate": 9.825250361743667e-06, | |
| "loss": 0.0922, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.35584843492586493, | |
| "grad_norm": 1.018855526493989, | |
| "learning_rate": 9.817616995894694e-06, | |
| "loss": 0.0893, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.35914332784184516, | |
| "grad_norm": 0.8793948272396893, | |
| "learning_rate": 9.809823549775559e-06, | |
| "loss": 0.0816, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.3624382207578254, | |
| "grad_norm": 1.1843731418180148, | |
| "learning_rate": 9.801870282347686e-06, | |
| "loss": 0.0815, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3657331136738056, | |
| "grad_norm": 1.3513059654506845, | |
| "learning_rate": 9.793757457883062e-06, | |
| "loss": 0.0838, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.36902800658978585, | |
| "grad_norm": 0.9877986179117915, | |
| "learning_rate": 9.785485345955446e-06, | |
| "loss": 0.0873, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.3723228995057661, | |
| "grad_norm": 0.8970818631280407, | |
| "learning_rate": 9.777054221431418e-06, | |
| "loss": 0.0611, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.3756177924217463, | |
| "grad_norm": 0.9226544421763598, | |
| "learning_rate": 9.768464364461248e-06, | |
| "loss": 0.078, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.37891268533772654, | |
| "grad_norm": 1.4766754423858683, | |
| "learning_rate": 9.75971606046958e-06, | |
| "loss": 0.0858, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.38220757825370677, | |
| "grad_norm": 1.320312770948565, | |
| "learning_rate": 9.750809600145955e-06, | |
| "loss": 0.0872, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.385502471169687, | |
| "grad_norm": 0.9205622576015271, | |
| "learning_rate": 9.741745279435144e-06, | |
| "loss": 0.0855, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.38879736408566723, | |
| "grad_norm": 2.1130956623266384, | |
| "learning_rate": 9.732523399527328e-06, | |
| "loss": 0.0869, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.39209225700164746, | |
| "grad_norm": 1.346773242530782, | |
| "learning_rate": 9.723144266848073e-06, | |
| "loss": 0.0891, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.3953871499176277, | |
| "grad_norm": 0.954418724615336, | |
| "learning_rate": 9.713608193048156e-06, | |
| "loss": 0.0927, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3986820428336079, | |
| "grad_norm": 1.1613672266850283, | |
| "learning_rate": 9.703915494993215e-06, | |
| "loss": 0.0946, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.40197693574958815, | |
| "grad_norm": 0.9668677562255413, | |
| "learning_rate": 9.694066494753211e-06, | |
| "loss": 0.0828, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.4052718286655684, | |
| "grad_norm": 1.1556457477946604, | |
| "learning_rate": 9.684061519591734e-06, | |
| "loss": 0.0926, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.4085667215815486, | |
| "grad_norm": 1.0348616873807939, | |
| "learning_rate": 9.673900901955118e-06, | |
| "loss": 0.0942, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.41186161449752884, | |
| "grad_norm": 0.954839683053947, | |
| "learning_rate": 9.663584979461407e-06, | |
| "loss": 0.0841, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.41515650741350907, | |
| "grad_norm": 0.8951987570639582, | |
| "learning_rate": 9.653114094889128e-06, | |
| "loss": 0.082, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.4184514003294893, | |
| "grad_norm": 0.5582851260265275, | |
| "learning_rate": 9.642488596165903e-06, | |
| "loss": 0.0579, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.42174629324546953, | |
| "grad_norm": 0.7282031291743011, | |
| "learning_rate": 9.631708836356893e-06, | |
| "loss": 0.0686, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.42504118616144976, | |
| "grad_norm": 0.7341380700784487, | |
| "learning_rate": 9.620775173653055e-06, | |
| "loss": 0.0581, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.42833607907743, | |
| "grad_norm": 1.0426311808706392, | |
| "learning_rate": 9.609687971359254e-06, | |
| "loss": 0.0863, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4316309719934102, | |
| "grad_norm": 0.7221932362532579, | |
| "learning_rate": 9.598447597882181e-06, | |
| "loss": 0.0904, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.43492586490939045, | |
| "grad_norm": 1.057905146517346, | |
| "learning_rate": 9.587054426718117e-06, | |
| "loss": 0.087, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.4382207578253707, | |
| "grad_norm": 1.0654844742045295, | |
| "learning_rate": 9.575508836440516e-06, | |
| "loss": 0.0833, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.4415156507413509, | |
| "grad_norm": 0.9708392842496616, | |
| "learning_rate": 9.563811210687433e-06, | |
| "loss": 0.07, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.44481054365733114, | |
| "grad_norm": 0.869842645454385, | |
| "learning_rate": 9.551961938148772e-06, | |
| "loss": 0.0798, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.44810543657331137, | |
| "grad_norm": 0.9291669588809491, | |
| "learning_rate": 9.539961412553375e-06, | |
| "loss": 0.0717, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.4514003294892916, | |
| "grad_norm": 0.971502019921878, | |
| "learning_rate": 9.52781003265593e-06, | |
| "loss": 0.0936, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.4546952224052718, | |
| "grad_norm": 0.7797238947376094, | |
| "learning_rate": 9.515508202223735e-06, | |
| "loss": 0.0711, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.45799011532125206, | |
| "grad_norm": 1.0743726394776476, | |
| "learning_rate": 9.503056330023267e-06, | |
| "loss": 0.0755, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.4612850082372323, | |
| "grad_norm": 1.1954773475892864, | |
| "learning_rate": 9.490454829806609e-06, | |
| "loss": 0.1304, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.4645799011532125, | |
| "grad_norm": 0.5763939890921029, | |
| "learning_rate": 9.477704120297698e-06, | |
| "loss": 0.0614, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.46787479406919275, | |
| "grad_norm": 0.9472151466304063, | |
| "learning_rate": 9.464804625178414e-06, | |
| "loss": 0.0712, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.471169686985173, | |
| "grad_norm": 0.886844022382936, | |
| "learning_rate": 9.4517567730745e-06, | |
| "loss": 0.0797, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.4744645799011532, | |
| "grad_norm": 1.0134166748412918, | |
| "learning_rate": 9.438560997541319e-06, | |
| "loss": 0.0899, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.47775947281713343, | |
| "grad_norm": 0.8070690603660651, | |
| "learning_rate": 9.425217737049452e-06, | |
| "loss": 0.0826, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.48105436573311366, | |
| "grad_norm": 0.8619994488550511, | |
| "learning_rate": 9.411727434970121e-06, | |
| "loss": 0.086, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.4843492586490939, | |
| "grad_norm": 0.8778816857460797, | |
| "learning_rate": 9.398090539560465e-06, | |
| "loss": 0.0854, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.4876441515650741, | |
| "grad_norm": 1.1118672558694775, | |
| "learning_rate": 9.384307503948637e-06, | |
| "loss": 0.1105, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.49093904448105435, | |
| "grad_norm": 1.032735807082786, | |
| "learning_rate": 9.370378786118755e-06, | |
| "loss": 0.0783, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.4942339373970346, | |
| "grad_norm": 1.0495708384737894, | |
| "learning_rate": 9.356304848895676e-06, | |
| "loss": 0.0815, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.4975288303130148, | |
| "grad_norm": 1.0945068947939716, | |
| "learning_rate": 9.342086159929629e-06, | |
| "loss": 0.0875, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.500823723228995, | |
| "grad_norm": 1.1286751183737302, | |
| "learning_rate": 9.327723191680666e-06, | |
| "loss": 0.0545, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.5041186161449753, | |
| "grad_norm": 0.9183247855730471, | |
| "learning_rate": 9.31321642140296e-06, | |
| "loss": 0.0757, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.5074135090609555, | |
| "grad_norm": 1.0502035656993673, | |
| "learning_rate": 9.29856633112896e-06, | |
| "loss": 0.0809, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.5107084019769358, | |
| "grad_norm": 0.8738749178267629, | |
| "learning_rate": 9.283773407653363e-06, | |
| "loss": 0.0562, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.514003294892916, | |
| "grad_norm": 1.0813438666782247, | |
| "learning_rate": 9.268838142516943e-06, | |
| "loss": 0.085, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.5172981878088962, | |
| "grad_norm": 1.0115574109758954, | |
| "learning_rate": 9.253761031990218e-06, | |
| "loss": 0.0749, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.5205930807248764, | |
| "grad_norm": 1.375042365212545, | |
| "learning_rate": 9.238542577056957e-06, | |
| "loss": 0.078, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.5238879736408567, | |
| "grad_norm": 1.59130187611513, | |
| "learning_rate": 9.223183283397538e-06, | |
| "loss": 0.1029, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.5271828665568369, | |
| "grad_norm": 1.1395752837101527, | |
| "learning_rate": 9.20768366137214e-06, | |
| "loss": 0.1128, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5304777594728172, | |
| "grad_norm": 0.8079597092244131, | |
| "learning_rate": 9.19204422600379e-06, | |
| "loss": 0.0527, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.5337726523887973, | |
| "grad_norm": 1.0336684369348528, | |
| "learning_rate": 9.176265496961242e-06, | |
| "loss": 0.0828, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.5370675453047776, | |
| "grad_norm": 0.8656528491730779, | |
| "learning_rate": 9.160347998541722e-06, | |
| "loss": 0.0704, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.5403624382207578, | |
| "grad_norm": 2.086614798853126, | |
| "learning_rate": 9.144292259653493e-06, | |
| "loss": 0.104, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.5436573311367381, | |
| "grad_norm": 1.5779000987891205, | |
| "learning_rate": 9.128098813798291e-06, | |
| "loss": 0.0996, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5469522240527183, | |
| "grad_norm": 0.7283818376854391, | |
| "learning_rate": 9.111768199053588e-06, | |
| "loss": 0.0621, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.5502471169686985, | |
| "grad_norm": 1.2155979859224575, | |
| "learning_rate": 9.095300958054722e-06, | |
| "loss": 0.0653, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.5535420098846787, | |
| "grad_norm": 1.3669964434695867, | |
| "learning_rate": 9.078697637976861e-06, | |
| "loss": 0.1071, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.556836902800659, | |
| "grad_norm": 0.659337598200629, | |
| "learning_rate": 9.061958790516821e-06, | |
| "loss": 0.101, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.5601317957166392, | |
| "grad_norm": 3.064428601730586, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.0631, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5634266886326195, | |
| "grad_norm": 1.8617437169334994, | |
| "learning_rate": 9.028076742735583e-06, | |
| "loss": 0.1062, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.5667215815485996, | |
| "grad_norm": 1.005975190266642, | |
| "learning_rate": 9.010934668250533e-06, | |
| "loss": 0.0706, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.5700164744645799, | |
| "grad_norm": 1.2704133524125742, | |
| "learning_rate": 8.993659318018191e-06, | |
| "loss": 0.1047, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.5733113673805601, | |
| "grad_norm": 1.5091840035688024, | |
| "learning_rate": 8.976251266065663e-06, | |
| "loss": 0.0915, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.5766062602965404, | |
| "grad_norm": 0.9983867645520093, | |
| "learning_rate": 8.958711090829477e-06, | |
| "loss": 0.0868, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.5799011532125206, | |
| "grad_norm": 0.9695191782522918, | |
| "learning_rate": 8.94103937513637e-06, | |
| "loss": 0.0782, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.5831960461285008, | |
| "grad_norm": 1.485821220772562, | |
| "learning_rate": 8.923236706183923e-06, | |
| "loss": 0.088, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.586490939044481, | |
| "grad_norm": 0.8778171297271887, | |
| "learning_rate": 8.905303675521031e-06, | |
| "loss": 0.0675, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.5897858319604613, | |
| "grad_norm": 1.4148551976101613, | |
| "learning_rate": 8.887240879028276e-06, | |
| "loss": 0.0968, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.5930807248764415, | |
| "grad_norm": 1.2313668213255908, | |
| "learning_rate": 8.869048916898109e-06, | |
| "loss": 0.0885, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.5963756177924218, | |
| "grad_norm": 0.7935552297095103, | |
| "learning_rate": 8.850728393614903e-06, | |
| "loss": 0.0919, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.5996705107084019, | |
| "grad_norm": 0.6219191194067725, | |
| "learning_rate": 8.832279917934881e-06, | |
| "loss": 0.0495, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.6029654036243822, | |
| "grad_norm": 0.891323405042387, | |
| "learning_rate": 8.813704102865881e-06, | |
| "loss": 0.1036, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.6062602965403624, | |
| "grad_norm": 0.7672391951678563, | |
| "learning_rate": 8.795001565646983e-06, | |
| "loss": 0.0728, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.6095551894563427, | |
| "grad_norm": 0.9915439152928615, | |
| "learning_rate": 8.776172927728008e-06, | |
| "loss": 0.0744, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6128500823723229, | |
| "grad_norm": 0.7415850366120278, | |
| "learning_rate": 8.75721881474886e-06, | |
| "loss": 0.0999, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.6161449752883031, | |
| "grad_norm": 0.7803751476377428, | |
| "learning_rate": 8.738139856518746e-06, | |
| "loss": 0.084, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.6194398682042833, | |
| "grad_norm": 0.6444369328111929, | |
| "learning_rate": 8.718936686995239e-06, | |
| "loss": 0.0632, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.6227347611202636, | |
| "grad_norm": 1.063336491675898, | |
| "learning_rate": 8.699609944263219e-06, | |
| "loss": 0.0854, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.6260296540362438, | |
| "grad_norm": 0.5254825539939502, | |
| "learning_rate": 8.680160270513671e-06, | |
| "loss": 0.0658, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6293245469522241, | |
| "grad_norm": 2.1957036152343097, | |
| "learning_rate": 8.660588312022345e-06, | |
| "loss": 0.0767, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.6326194398682042, | |
| "grad_norm": 1.1425043272398887, | |
| "learning_rate": 8.640894719128274e-06, | |
| "loss": 0.1092, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.6359143327841845, | |
| "grad_norm": 0.6811077233527263, | |
| "learning_rate": 8.621080146212181e-06, | |
| "loss": 0.0552, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.6392092257001647, | |
| "grad_norm": 1.0217179933628129, | |
| "learning_rate": 8.601145251674718e-06, | |
| "loss": 0.0749, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.642504118616145, | |
| "grad_norm": 0.9919309113126284, | |
| "learning_rate": 8.581090697914602e-06, | |
| "loss": 0.0929, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.6457990115321252, | |
| "grad_norm": 1.178486792691301, | |
| "learning_rate": 8.560917151306594e-06, | |
| "loss": 0.1023, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.6490939044481054, | |
| "grad_norm": 1.155779967707836, | |
| "learning_rate": 8.540625282179364e-06, | |
| "loss": 0.0821, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.6523887973640856, | |
| "grad_norm": 1.0278193202776953, | |
| "learning_rate": 8.520215764793214e-06, | |
| "loss": 0.0739, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.6556836902800659, | |
| "grad_norm": 5.872968813903652, | |
| "learning_rate": 8.499689277317675e-06, | |
| "loss": 0.0763, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.6589785831960461, | |
| "grad_norm": 1.4949053388183082, | |
| "learning_rate": 8.479046501808971e-06, | |
| "loss": 0.0696, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6622734761120264, | |
| "grad_norm": 0.9342591652212858, | |
| "learning_rate": 8.45828812418736e-06, | |
| "loss": 0.0629, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.6655683690280065, | |
| "grad_norm": 0.5142904539906094, | |
| "learning_rate": 8.437414834214333e-06, | |
| "loss": 0.0653, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.6688632619439868, | |
| "grad_norm": 1.3243894037794444, | |
| "learning_rate": 8.416427325469705e-06, | |
| "loss": 0.1095, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.6721581548599671, | |
| "grad_norm": 1.3555237832265656, | |
| "learning_rate": 8.395326295328562e-06, | |
| "loss": 0.1028, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.6754530477759473, | |
| "grad_norm": 0.818855723195901, | |
| "learning_rate": 8.374112444938094e-06, | |
| "loss": 0.088, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.6787479406919276, | |
| "grad_norm": 0.8852074571699485, | |
| "learning_rate": 8.352786479194288e-06, | |
| "loss": 0.0526, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.6820428336079077, | |
| "grad_norm": 1.372137747701387, | |
| "learning_rate": 8.331349106718515e-06, | |
| "loss": 0.0957, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.685337726523888, | |
| "grad_norm": 0.8727917319312107, | |
| "learning_rate": 8.309801039833978e-06, | |
| "loss": 0.0895, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.6886326194398682, | |
| "grad_norm": 1.0509761315988109, | |
| "learning_rate": 8.28814299454205e-06, | |
| "loss": 0.0996, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.6919275123558485, | |
| "grad_norm": 0.8839477803791086, | |
| "learning_rate": 8.266375690498475e-06, | |
| "loss": 0.0865, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.6952224052718287, | |
| "grad_norm": 0.7064553470448757, | |
| "learning_rate": 8.244499850989453e-06, | |
| "loss": 0.0728, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.6985172981878089, | |
| "grad_norm": 0.9694201547278781, | |
| "learning_rate": 8.22251620290762e-06, | |
| "loss": 0.0581, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.7018121911037891, | |
| "grad_norm": 0.724783690835583, | |
| "learning_rate": 8.20042547672788e-06, | |
| "loss": 0.0828, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.7051070840197694, | |
| "grad_norm": 0.7137797204098316, | |
| "learning_rate": 8.178228406483145e-06, | |
| "loss": 0.0707, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.7084019769357496, | |
| "grad_norm": 0.4606523745916078, | |
| "learning_rate": 8.15592572973993e-06, | |
| "loss": 0.044, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.7116968698517299, | |
| "grad_norm": 0.6640092591585782, | |
| "learning_rate": 8.133518187573864e-06, | |
| "loss": 0.0561, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.71499176276771, | |
| "grad_norm": 0.8170937626114954, | |
| "learning_rate": 8.111006524545043e-06, | |
| "loss": 0.0823, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.7182866556836903, | |
| "grad_norm": 0.6372182616249538, | |
| "learning_rate": 8.088391488673313e-06, | |
| "loss": 0.066, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.7215815485996705, | |
| "grad_norm": 0.6375067936631406, | |
| "learning_rate": 8.065673831413396e-06, | |
| "loss": 0.0506, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.7248764415156508, | |
| "grad_norm": 0.6423309999365212, | |
| "learning_rate": 8.042854307629932e-06, | |
| "loss": 0.0629, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.728171334431631, | |
| "grad_norm": 0.752582677343103, | |
| "learning_rate": 8.019933675572389e-06, | |
| "loss": 0.0722, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.7314662273476112, | |
| "grad_norm": 0.8112402585396246, | |
| "learning_rate": 7.996912696849873e-06, | |
| "loss": 0.0842, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.7347611202635914, | |
| "grad_norm": 0.6919487336036387, | |
| "learning_rate": 7.97379213640582e-06, | |
| "loss": 0.0684, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.7380560131795717, | |
| "grad_norm": 0.7434807539917022, | |
| "learning_rate": 7.950572762492577e-06, | |
| "loss": 0.0682, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.7413509060955519, | |
| "grad_norm": 0.6848656249105235, | |
| "learning_rate": 7.927255346645872e-06, | |
| "loss": 0.0546, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.7446457990115322, | |
| "grad_norm": 0.7826479624770332, | |
| "learning_rate": 7.903840663659186e-06, | |
| "loss": 0.0684, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.7479406919275123, | |
| "grad_norm": 0.6927518800734283, | |
| "learning_rate": 7.880329491557996e-06, | |
| "loss": 0.079, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.7512355848434926, | |
| "grad_norm": 0.8763045243203113, | |
| "learning_rate": 7.856722611573938e-06, | |
| "loss": 0.1068, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.7545304777594728, | |
| "grad_norm": 0.8300681217056403, | |
| "learning_rate": 7.83302080811883e-06, | |
| "loss": 0.0667, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.7578253706754531, | |
| "grad_norm": 0.5437395859594083, | |
| "learning_rate": 7.809224868758621e-06, | |
| "loss": 0.0671, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7611202635914333, | |
| "grad_norm": 0.7134167417475868, | |
| "learning_rate": 7.78533558418722e-06, | |
| "loss": 0.079, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.7644151565074135, | |
| "grad_norm": 0.8367162252369527, | |
| "learning_rate": 7.761353748200213e-06, | |
| "loss": 0.075, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.7677100494233937, | |
| "grad_norm": 0.6993381735068975, | |
| "learning_rate": 7.737280157668503e-06, | |
| "loss": 0.0665, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.771004942339374, | |
| "grad_norm": 0.644489745443266, | |
| "learning_rate": 7.713115612511815e-06, | |
| "loss": 0.0704, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.7742998352553542, | |
| "grad_norm": 0.6337392482963783, | |
| "learning_rate": 7.688860915672129e-06, | |
| "loss": 0.0487, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.7775947281713345, | |
| "grad_norm": 0.4306575759208823, | |
| "learning_rate": 7.664516873086987e-06, | |
| "loss": 0.0498, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.7808896210873146, | |
| "grad_norm": 0.6371209076121114, | |
| "learning_rate": 7.640084293662731e-06, | |
| "loss": 0.0581, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.7841845140032949, | |
| "grad_norm": 0.809205628205596, | |
| "learning_rate": 7.615563989247604e-06, | |
| "loss": 0.0886, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.7874794069192751, | |
| "grad_norm": 0.6807826450879982, | |
| "learning_rate": 7.590956774604791e-06, | |
| "loss": 0.0824, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.7907742998352554, | |
| "grad_norm": 0.9092838195300236, | |
| "learning_rate": 7.566263467385335e-06, | |
| "loss": 0.0703, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.7940691927512356, | |
| "grad_norm": 0.736565350841279, | |
| "learning_rate": 7.541484888100974e-06, | |
| "loss": 0.0695, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.7973640856672158, | |
| "grad_norm": 0.7220288466907268, | |
| "learning_rate": 7.516621860096873e-06, | |
| "loss": 0.0707, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.800658978583196, | |
| "grad_norm": 0.6829838831547227, | |
| "learning_rate": 7.491675209524272e-06, | |
| "loss": 0.0666, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.8039538714991763, | |
| "grad_norm": 0.8226949141177504, | |
| "learning_rate": 7.466645765313023e-06, | |
| "loss": 0.0752, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.8072487644151565, | |
| "grad_norm": 0.5909405519820083, | |
| "learning_rate": 7.4415343591440604e-06, | |
| "loss": 0.0582, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.8105436573311368, | |
| "grad_norm": 0.7318765147109815, | |
| "learning_rate": 7.416341825421755e-06, | |
| "loss": 0.078, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.8138385502471169, | |
| "grad_norm": 0.7063912838195767, | |
| "learning_rate": 7.391069001246193e-06, | |
| "loss": 0.0868, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.8171334431630972, | |
| "grad_norm": 0.6799477779267012, | |
| "learning_rate": 7.365716726385361e-06, | |
| "loss": 0.0681, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.8204283360790774, | |
| "grad_norm": 0.8516971338664023, | |
| "learning_rate": 7.3402858432472416e-06, | |
| "loss": 0.0761, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.8237232289950577, | |
| "grad_norm": 0.8051104503646311, | |
| "learning_rate": 7.3147771968518175e-06, | |
| "loss": 0.077, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8270181219110379, | |
| "grad_norm": 0.8417638265928152, | |
| "learning_rate": 7.289191634803002e-06, | |
| "loss": 0.0721, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.8303130148270181, | |
| "grad_norm": 0.9280576906426667, | |
| "learning_rate": 7.263530007260466e-06, | |
| "loss": 0.0839, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.8336079077429983, | |
| "grad_norm": 0.8205604877193189, | |
| "learning_rate": 7.2377931669113934e-06, | |
| "loss": 0.084, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.8369028006589786, | |
| "grad_norm": 0.7347246169190605, | |
| "learning_rate": 7.211981968942147e-06, | |
| "loss": 0.0508, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.8401976935749588, | |
| "grad_norm": 0.7727540137134915, | |
| "learning_rate": 7.186097271009852e-06, | |
| "loss": 0.0504, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.8434925864909391, | |
| "grad_norm": 0.6116838823458901, | |
| "learning_rate": 7.160139933213899e-06, | |
| "loss": 0.0533, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.8467874794069192, | |
| "grad_norm": 0.8518782068127816, | |
| "learning_rate": 7.134110818067361e-06, | |
| "loss": 0.0775, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.8500823723228995, | |
| "grad_norm": 0.9449160515812749, | |
| "learning_rate": 7.1080107904683405e-06, | |
| "loss": 0.0721, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.8533772652388797, | |
| "grad_norm": 0.6964873142430633, | |
| "learning_rate": 7.08184071767122e-06, | |
| "loss": 0.0673, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.85667215815486, | |
| "grad_norm": 0.768104304709271, | |
| "learning_rate": 7.0556014692578554e-06, | |
| "loss": 0.0749, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8599670510708401, | |
| "grad_norm": 0.7599189113700034, | |
| "learning_rate": 7.029293917108678e-06, | |
| "loss": 0.0684, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.8632619439868204, | |
| "grad_norm": 0.777387517223909, | |
| "learning_rate": 7.0029189353737195e-06, | |
| "loss": 0.0656, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.8665568369028006, | |
| "grad_norm": 0.7045793540209936, | |
| "learning_rate": 6.9764774004435685e-06, | |
| "loss": 0.0619, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.8698517298187809, | |
| "grad_norm": 0.6234760268316166, | |
| "learning_rate": 6.949970190920255e-06, | |
| "loss": 0.0708, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.8731466227347611, | |
| "grad_norm": 0.7124980322892176, | |
| "learning_rate": 6.9233981875880416e-06, | |
| "loss": 0.0521, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.8764415156507414, | |
| "grad_norm": 0.8490902000387839, | |
| "learning_rate": 6.896762273384179e-06, | |
| "loss": 0.0632, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.8797364085667215, | |
| "grad_norm": 0.6944201065528963, | |
| "learning_rate": 6.870063333369543e-06, | |
| "loss": 0.0716, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.8830313014827018, | |
| "grad_norm": 0.758349126043532, | |
| "learning_rate": 6.8433022546992444e-06, | |
| "loss": 0.0596, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.886326194398682, | |
| "grad_norm": 1.2664444257431744, | |
| "learning_rate": 6.81647992659314e-06, | |
| "loss": 0.0628, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.8896210873146623, | |
| "grad_norm": 0.8379324844684077, | |
| "learning_rate": 6.789597240306295e-06, | |
| "loss": 0.0674, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.8929159802306426, | |
| "grad_norm": 0.8462600835900949, | |
| "learning_rate": 6.762655089099353e-06, | |
| "loss": 0.0659, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.8962108731466227, | |
| "grad_norm": 0.9094387161179498, | |
| "learning_rate": 6.735654368208875e-06, | |
| "loss": 0.0623, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.899505766062603, | |
| "grad_norm": 0.7877875224066865, | |
| "learning_rate": 6.7085959748175685e-06, | |
| "loss": 0.0696, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.9028006589785832, | |
| "grad_norm": 0.6514864513423558, | |
| "learning_rate": 6.681480808024503e-06, | |
| "loss": 0.0766, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.9060955518945635, | |
| "grad_norm": 1.148236164352365, | |
| "learning_rate": 6.654309768815208e-06, | |
| "loss": 0.0903, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.9093904448105437, | |
| "grad_norm": 0.7078109102899715, | |
| "learning_rate": 6.627083760031755e-06, | |
| "loss": 0.0607, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.9126853377265239, | |
| "grad_norm": 0.613094345393223, | |
| "learning_rate": 6.599803686342748e-06, | |
| "loss": 0.0655, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.9159802306425041, | |
| "grad_norm": 0.6642339763695972, | |
| "learning_rate": 6.572470454213266e-06, | |
| "loss": 0.0731, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.9192751235584844, | |
| "grad_norm": 0.6971630112819691, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.0473, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.9225700164744646, | |
| "grad_norm": 0.7592858638911076, | |
| "learning_rate": 6.517648149294774e-06, | |
| "loss": 0.0581, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9258649093904449, | |
| "grad_norm": 0.7189143571066544, | |
| "learning_rate": 6.490160898146919e-06, | |
| "loss": 0.0733, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.929159802306425, | |
| "grad_norm": 0.8305599945381572, | |
| "learning_rate": 6.4626241317803665e-06, | |
| "loss": 0.0807, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.9324546952224053, | |
| "grad_norm": 0.8787944618632045, | |
| "learning_rate": 6.4350387651896025e-06, | |
| "loss": 0.0648, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.9357495881383855, | |
| "grad_norm": 0.649270561331511, | |
| "learning_rate": 6.407405714984011e-06, | |
| "loss": 0.0921, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.9390444810543658, | |
| "grad_norm": 0.9873611661857511, | |
| "learning_rate": 6.379725899357408e-06, | |
| "loss": 0.0847, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.942339373970346, | |
| "grad_norm": 0.8338719043181901, | |
| "learning_rate": 6.3520002380575395e-06, | |
| "loss": 0.0673, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.9456342668863262, | |
| "grad_norm": 0.8390156519820746, | |
| "learning_rate": 6.324229652355513e-06, | |
| "loss": 0.0626, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.9489291598023064, | |
| "grad_norm": 0.7197773939188823, | |
| "learning_rate": 6.29641506501519e-06, | |
| "loss": 0.0864, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.9522240527182867, | |
| "grad_norm": 0.942984980454084, | |
| "learning_rate": 6.2685574002625235e-06, | |
| "loss": 0.0686, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.9555189456342669, | |
| "grad_norm": 0.9649936636393807, | |
| "learning_rate": 6.2406575837548455e-06, | |
| "loss": 0.0599, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9588138385502472, | |
| "grad_norm": 0.6889881534410974, | |
| "learning_rate": 6.212716542550112e-06, | |
| "loss": 0.101, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.9621087314662273, | |
| "grad_norm": 0.9632795509211302, | |
| "learning_rate": 6.184735205076097e-06, | |
| "loss": 0.0773, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.9654036243822076, | |
| "grad_norm": 1.0400767819370376, | |
| "learning_rate": 6.156714501099544e-06, | |
| "loss": 0.0638, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.9686985172981878, | |
| "grad_norm": 1.0147243725605253, | |
| "learning_rate": 6.1286553616952705e-06, | |
| "loss": 0.0593, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.9719934102141681, | |
| "grad_norm": 0.6613193470791487, | |
| "learning_rate": 6.100558719215228e-06, | |
| "loss": 0.0632, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.9752883031301482, | |
| "grad_norm": 1.0408938474730054, | |
| "learning_rate": 6.072425507257528e-06, | |
| "loss": 0.0876, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.9785831960461285, | |
| "grad_norm": 0.712701647042842, | |
| "learning_rate": 6.044256660635412e-06, | |
| "loss": 0.0733, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.9818780889621087, | |
| "grad_norm": 0.6397491114376809, | |
| "learning_rate": 6.016053115346197e-06, | |
| "loss": 0.0561, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.985172981878089, | |
| "grad_norm": 0.7191102659386986, | |
| "learning_rate": 5.987815808540169e-06, | |
| "loss": 0.0791, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.9884678747940692, | |
| "grad_norm": 0.4709712102337363, | |
| "learning_rate": 5.959545678489447e-06, | |
| "loss": 0.0475, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.9917627677100495, | |
| "grad_norm": 0.8715274588578796, | |
| "learning_rate": 5.931243664556803e-06, | |
| "loss": 0.0771, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.9950576606260296, | |
| "grad_norm": 0.7017524340447387, | |
| "learning_rate": 5.902910707164449e-06, | |
| "loss": 0.0712, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.9983525535420099, | |
| "grad_norm": 0.7619744594259967, | |
| "learning_rate": 5.874547747762792e-06, | |
| "loss": 0.0585, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.9983525535420099, | |
| "eval_loss": 0.07007648050785065, | |
| "eval_runtime": 143.0638, | |
| "eval_samples_per_second": 35.683, | |
| "eval_steps_per_second": 1.118, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.00164744645799, | |
| "grad_norm": 0.5917774055195716, | |
| "learning_rate": 5.8461557287991455e-06, | |
| "loss": 0.0686, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.0049423393739703, | |
| "grad_norm": 0.4973762275932349, | |
| "learning_rate": 5.81773559368642e-06, | |
| "loss": 0.0524, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.0082372322899507, | |
| "grad_norm": 0.5021975231329254, | |
| "learning_rate": 5.7892882867717705e-06, | |
| "loss": 0.0577, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.0115321252059308, | |
| "grad_norm": 0.704352626743678, | |
| "learning_rate": 5.7608147533052194e-06, | |
| "loss": 0.0509, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.014827018121911, | |
| "grad_norm": 1.189723828759097, | |
| "learning_rate": 5.732315939408251e-06, | |
| "loss": 0.0815, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.0181219110378912, | |
| "grad_norm": 0.6036027009145574, | |
| "learning_rate": 5.703792792042363e-06, | |
| "loss": 0.0556, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.0214168039538716, | |
| "grad_norm": 0.5342904909103813, | |
| "learning_rate": 5.675246258977617e-06, | |
| "loss": 0.0487, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0247116968698518, | |
| "grad_norm": 0.46763620767148034, | |
| "learning_rate": 5.646677288761132e-06, | |
| "loss": 0.0491, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.028006589785832, | |
| "grad_norm": 0.5696375911949768, | |
| "learning_rate": 5.618086830685569e-06, | |
| "loss": 0.047, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.031301482701812, | |
| "grad_norm": 0.38600791899244996, | |
| "learning_rate": 5.589475834757595e-06, | |
| "loss": 0.032, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.0345963756177925, | |
| "grad_norm": 1.0072710877393638, | |
| "learning_rate": 5.560845251666307e-06, | |
| "loss": 0.063, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.0378912685337727, | |
| "grad_norm": 0.663725882779124, | |
| "learning_rate": 5.532196032751647e-06, | |
| "loss": 0.0563, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.0411861614497528, | |
| "grad_norm": 0.6135177621912624, | |
| "learning_rate": 5.503529129972792e-06, | |
| "loss": 0.0514, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.044481054365733, | |
| "grad_norm": 0.7549455934476204, | |
| "learning_rate": 5.474845495876518e-06, | |
| "loss": 0.0563, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.0477759472817134, | |
| "grad_norm": 0.8244910748727189, | |
| "learning_rate": 5.4461460835655535e-06, | |
| "loss": 0.0804, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.0510708401976936, | |
| "grad_norm": 0.604488079236042, | |
| "learning_rate": 5.417431846666903e-06, | |
| "loss": 0.0679, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.0543657331136738, | |
| "grad_norm": 0.5136265587955748, | |
| "learning_rate": 5.388703739300167e-06, | |
| "loss": 0.0388, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.057660626029654, | |
| "grad_norm": 0.6007478171198604, | |
| "learning_rate": 5.359962716045836e-06, | |
| "loss": 0.0632, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.0609555189456343, | |
| "grad_norm": 0.4928892879154173, | |
| "learning_rate": 5.331209731913568e-06, | |
| "loss": 0.058, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.0642504118616145, | |
| "grad_norm": 0.5300520318408385, | |
| "learning_rate": 5.30244574231046e-06, | |
| "loss": 0.0528, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.0675453047775947, | |
| "grad_norm": 0.49159187140329286, | |
| "learning_rate": 5.273671703009301e-06, | |
| "loss": 0.046, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.0708401976935749, | |
| "grad_norm": 0.804620211006138, | |
| "learning_rate": 5.2448885701168094e-06, | |
| "loss": 0.0601, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.0741350906095553, | |
| "grad_norm": 0.4648011852930538, | |
| "learning_rate": 5.21609730004187e-06, | |
| "loss": 0.0438, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.0774299835255354, | |
| "grad_norm": 0.5362596735899865, | |
| "learning_rate": 5.187298849463748e-06, | |
| "loss": 0.0507, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.0807248764415156, | |
| "grad_norm": 0.5443586783585722, | |
| "learning_rate": 5.158494175300304e-06, | |
| "loss": 0.053, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.084019769357496, | |
| "grad_norm": 0.6076056192307563, | |
| "learning_rate": 5.129684234676195e-06, | |
| "loss": 0.0594, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.0873146622734762, | |
| "grad_norm": 0.9033252357763137, | |
| "learning_rate": 5.100869984891077e-06, | |
| "loss": 0.06, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.0906095551894563, | |
| "grad_norm": 0.454480847306655, | |
| "learning_rate": 5.072052383387787e-06, | |
| "loss": 0.0424, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.0939044481054365, | |
| "grad_norm": 0.46517988927206794, | |
| "learning_rate": 5.043232387720532e-06, | |
| "loss": 0.0443, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.0971993410214167, | |
| "grad_norm": 0.4148720401510593, | |
| "learning_rate": 5.014410955523079e-06, | |
| "loss": 0.0387, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.100494233937397, | |
| "grad_norm": 0.5146539821704307, | |
| "learning_rate": 4.9855890444769226e-06, | |
| "loss": 0.0563, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.1037891268533773, | |
| "grad_norm": 0.5267211782218569, | |
| "learning_rate": 4.956767612279468e-06, | |
| "loss": 0.044, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.1070840197693574, | |
| "grad_norm": 0.5731696810590752, | |
| "learning_rate": 4.927947616612216e-06, | |
| "loss": 0.0469, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.1103789126853378, | |
| "grad_norm": 0.4606767989043497, | |
| "learning_rate": 4.899130015108923e-06, | |
| "loss": 0.0556, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.113673805601318, | |
| "grad_norm": 0.5591348812226693, | |
| "learning_rate": 4.8703157653238055e-06, | |
| "loss": 0.0526, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.1169686985172982, | |
| "grad_norm": 0.5103079438074868, | |
| "learning_rate": 4.841505824699697e-06, | |
| "loss": 0.0651, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.1202635914332784, | |
| "grad_norm": 0.6163138349681117, | |
| "learning_rate": 4.812701150536254e-06, | |
| "loss": 0.0509, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.1235584843492585, | |
| "grad_norm": 0.4842115475256147, | |
| "learning_rate": 4.78390269995813e-06, | |
| "loss": 0.035, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.126853377265239, | |
| "grad_norm": 0.4047877822645327, | |
| "learning_rate": 4.755111429883191e-06, | |
| "loss": 0.0342, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.130148270181219, | |
| "grad_norm": 0.5782935405242332, | |
| "learning_rate": 4.726328296990699e-06, | |
| "loss": 0.0416, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.1334431630971993, | |
| "grad_norm": 0.5846524401590787, | |
| "learning_rate": 4.697554257689541e-06, | |
| "loss": 0.0419, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.1367380560131797, | |
| "grad_norm": 0.5096985328650335, | |
| "learning_rate": 4.668790268086432e-06, | |
| "loss": 0.044, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.1400329489291599, | |
| "grad_norm": 0.5796683420196656, | |
| "learning_rate": 4.640037283954165e-06, | |
| "loss": 0.0634, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.14332784184514, | |
| "grad_norm": 0.5897186824110954, | |
| "learning_rate": 4.611296260699833e-06, | |
| "loss": 0.0511, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.1466227347611202, | |
| "grad_norm": 0.5043407904517478, | |
| "learning_rate": 4.582568153333098e-06, | |
| "loss": 0.0474, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.1499176276771004, | |
| "grad_norm": 0.49203813884361564, | |
| "learning_rate": 4.553853916434448e-06, | |
| "loss": 0.0399, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.1532125205930808, | |
| "grad_norm": 0.6380533145833258, | |
| "learning_rate": 4.525154504123483e-06, | |
| "loss": 0.0628, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.156507413509061, | |
| "grad_norm": 0.6307435685302706, | |
| "learning_rate": 4.496470870027209e-06, | |
| "loss": 0.0544, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.1598023064250411, | |
| "grad_norm": 0.58051661483701, | |
| "learning_rate": 4.467803967248354e-06, | |
| "loss": 0.0549, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.1630971993410215, | |
| "grad_norm": 0.45506940053593953, | |
| "learning_rate": 4.439154748333695e-06, | |
| "loss": 0.0455, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.1663920922570017, | |
| "grad_norm": 0.4477960561383021, | |
| "learning_rate": 4.410524165242407e-06, | |
| "loss": 0.0417, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.1696869851729819, | |
| "grad_norm": 0.5024790404868378, | |
| "learning_rate": 4.381913169314432e-06, | |
| "loss": 0.0483, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.172981878088962, | |
| "grad_norm": 0.43352510094853813, | |
| "learning_rate": 4.3533227112388694e-06, | |
| "loss": 0.0381, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.1762767710049424, | |
| "grad_norm": 0.8015757322992388, | |
| "learning_rate": 4.324753741022383e-06, | |
| "loss": 0.0589, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.1795716639209226, | |
| "grad_norm": 0.554923192898479, | |
| "learning_rate": 4.296207207957638e-06, | |
| "loss": 0.0469, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.1828665568369028, | |
| "grad_norm": 0.4540612599730088, | |
| "learning_rate": 4.26768406059175e-06, | |
| "loss": 0.0469, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.186161449752883, | |
| "grad_norm": 0.4977016265485015, | |
| "learning_rate": 4.239185246694781e-06, | |
| "loss": 0.0486, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.1894563426688634, | |
| "grad_norm": 0.5773178206107633, | |
| "learning_rate": 4.21071171322823e-06, | |
| "loss": 0.0588, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.1927512355848435, | |
| "grad_norm": 0.5714806332591411, | |
| "learning_rate": 4.182264406313582e-06, | |
| "loss": 0.0473, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.1960461285008237, | |
| "grad_norm": 0.5399317568380463, | |
| "learning_rate": 4.1538442712008545e-06, | |
| "loss": 0.0515, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.1993410214168039, | |
| "grad_norm": 0.5077736606662918, | |
| "learning_rate": 4.12545225223721e-06, | |
| "loss": 0.0473, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.2026359143327843, | |
| "grad_norm": 0.65833510309246, | |
| "learning_rate": 4.097089292835551e-06, | |
| "loss": 0.0574, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.2059308072487644, | |
| "grad_norm": 0.5750314764693017, | |
| "learning_rate": 4.0687563354431986e-06, | |
| "loss": 0.033, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.2092257001647446, | |
| "grad_norm": 0.6672168173906087, | |
| "learning_rate": 4.040454321510554e-06, | |
| "loss": 0.0507, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.2125205930807248, | |
| "grad_norm": 0.46572043828398524, | |
| "learning_rate": 4.012184191459832e-06, | |
| "loss": 0.0448, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.2158154859967052, | |
| "grad_norm": 0.5294456067061011, | |
| "learning_rate": 3.983946884653804e-06, | |
| "loss": 0.0421, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.2191103789126854, | |
| "grad_norm": 0.7181848630920071, | |
| "learning_rate": 3.95574333936459e-06, | |
| "loss": 0.0609, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.2224052718286655, | |
| "grad_norm": 0.4872681980462519, | |
| "learning_rate": 3.927574492742473e-06, | |
| "loss": 0.0332, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.2257001647446457, | |
| "grad_norm": 0.5978073219647344, | |
| "learning_rate": 3.899441280784773e-06, | |
| "loss": 0.0557, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.2289950576606261, | |
| "grad_norm": 0.49268040219816195, | |
| "learning_rate": 3.8713446383047295e-06, | |
| "loss": 0.0539, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.2322899505766063, | |
| "grad_norm": 0.553488767277818, | |
| "learning_rate": 3.843285498900457e-06, | |
| "loss": 0.0438, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.2355848434925865, | |
| "grad_norm": 0.5769809240481462, | |
| "learning_rate": 3.815264794923903e-06, | |
| "loss": 0.0438, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.2388797364085666, | |
| "grad_norm": 0.4680099999633115, | |
| "learning_rate": 3.7872834574498894e-06, | |
| "loss": 0.0391, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.242174629324547, | |
| "grad_norm": 0.4990397184455205, | |
| "learning_rate": 3.7593424162451553e-06, | |
| "loss": 0.0513, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.2454695222405272, | |
| "grad_norm": 0.5670279278262034, | |
| "learning_rate": 3.731442599737478e-06, | |
| "loss": 0.0611, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.2487644151565074, | |
| "grad_norm": 0.4178810778744549, | |
| "learning_rate": 3.70358493498481e-06, | |
| "loss": 0.0461, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.2520593080724876, | |
| "grad_norm": 0.5498450231361147, | |
| "learning_rate": 3.6757703476444885e-06, | |
| "loss": 0.0372, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.255354200988468, | |
| "grad_norm": 0.45367014770072983, | |
| "learning_rate": 3.6479997619424605e-06, | |
| "loss": 0.0423, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.2586490939044481, | |
| "grad_norm": 0.4294200611194709, | |
| "learning_rate": 3.620274100642593e-06, | |
| "loss": 0.0552, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.2619439868204283, | |
| "grad_norm": 0.6276700882265509, | |
| "learning_rate": 3.5925942850159895e-06, | |
| "loss": 0.0659, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.2652388797364087, | |
| "grad_norm": 0.7113783547292587, | |
| "learning_rate": 3.564961234810399e-06, | |
| "loss": 0.067, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.2685337726523889, | |
| "grad_norm": 0.6367177743488461, | |
| "learning_rate": 3.5373758682196347e-06, | |
| "loss": 0.0626, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.271828665568369, | |
| "grad_norm": 0.6068919065481327, | |
| "learning_rate": 3.509839101853082e-06, | |
| "loss": 0.0546, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.2751235584843492, | |
| "grad_norm": 0.742600911574775, | |
| "learning_rate": 3.4823518507052277e-06, | |
| "loss": 0.061, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.2784184514003294, | |
| "grad_norm": 0.4142179254874713, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.0405, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.2817133443163098, | |
| "grad_norm": 0.7545310044049625, | |
| "learning_rate": 3.427529545786736e-06, | |
| "loss": 0.055, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.28500823723229, | |
| "grad_norm": 0.5556958907162003, | |
| "learning_rate": 3.400196313657253e-06, | |
| "loss": 0.0469, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.2883031301482701, | |
| "grad_norm": 0.5082442265119059, | |
| "learning_rate": 3.372916239968246e-06, | |
| "loss": 0.048, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.2915980230642505, | |
| "grad_norm": 0.5683891171997948, | |
| "learning_rate": 3.345690231184794e-06, | |
| "loss": 0.0413, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.2948929159802307, | |
| "grad_norm": 0.6720011431709395, | |
| "learning_rate": 3.318519191975499e-06, | |
| "loss": 0.0604, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.2981878088962109, | |
| "grad_norm": 0.5633138961258451, | |
| "learning_rate": 3.291404025182432e-06, | |
| "loss": 0.065, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.301482701812191, | |
| "grad_norm": 0.7184125904469478, | |
| "learning_rate": 3.264345631791127e-06, | |
| "loss": 0.0653, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.3047775947281712, | |
| "grad_norm": 0.6249487221408845, | |
| "learning_rate": 3.2373449109006476e-06, | |
| "loss": 0.0476, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.3080724876441516, | |
| "grad_norm": 0.8032427870564648, | |
| "learning_rate": 3.210402759693706e-06, | |
| "loss": 0.0731, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.3113673805601318, | |
| "grad_norm": 0.56609919677685, | |
| "learning_rate": 3.1835200734068604e-06, | |
| "loss": 0.0484, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.314662273476112, | |
| "grad_norm": 0.6259799233731378, | |
| "learning_rate": 3.1566977453007564e-06, | |
| "loss": 0.0621, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.3179571663920924, | |
| "grad_norm": 0.5571126132586377, | |
| "learning_rate": 3.1299366666304586e-06, | |
| "loss": 0.0536, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.3212520593080725, | |
| "grad_norm": 0.7177292111816991, | |
| "learning_rate": 3.103237726615822e-06, | |
| "loss": 0.0634, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.3245469522240527, | |
| "grad_norm": 0.4655411649268851, | |
| "learning_rate": 3.076601812411959e-06, | |
| "loss": 0.0386, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.327841845140033, | |
| "grad_norm": 0.5171630532203868, | |
| "learning_rate": 3.0500298090797465e-06, | |
| "loss": 0.0483, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.331136738056013, | |
| "grad_norm": 0.624563937797765, | |
| "learning_rate": 3.0235225995564323e-06, | |
| "loss": 0.0556, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.3344316309719935, | |
| "grad_norm": 0.5622033908392009, | |
| "learning_rate": 2.9970810646262805e-06, | |
| "loss": 0.0478, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.3377265238879736, | |
| "grad_norm": 0.4858535947041361, | |
| "learning_rate": 2.9707060828913226e-06, | |
| "loss": 0.0478, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.3410214168039538, | |
| "grad_norm": 0.40196887575712115, | |
| "learning_rate": 2.944398530742144e-06, | |
| "loss": 0.0462, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.3443163097199342, | |
| "grad_norm": 0.5650818660979695, | |
| "learning_rate": 2.9181592823287807e-06, | |
| "loss": 0.0655, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.3476112026359144, | |
| "grad_norm": 0.5475272072919456, | |
| "learning_rate": 2.8919892095316616e-06, | |
| "loss": 0.0519, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.3509060955518946, | |
| "grad_norm": 0.6655092099152591, | |
| "learning_rate": 2.865889181932639e-06, | |
| "loss": 0.0416, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.3542009884678747, | |
| "grad_norm": 0.8536644782226072, | |
| "learning_rate": 2.8398600667861032e-06, | |
| "loss": 0.0669, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.357495881383855, | |
| "grad_norm": 0.48323829394508544, | |
| "learning_rate": 2.813902728990149e-06, | |
| "loss": 0.0367, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.3607907742998353, | |
| "grad_norm": 0.6949809437868909, | |
| "learning_rate": 2.7880180310578546e-06, | |
| "loss": 0.0523, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.3640856672158155, | |
| "grad_norm": 0.43770382604271224, | |
| "learning_rate": 2.762206833088608e-06, | |
| "loss": 0.0527, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.3673805601317957, | |
| "grad_norm": 0.5081547664603686, | |
| "learning_rate": 2.7364699927395355e-06, | |
| "loss": 0.0613, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.370675453047776, | |
| "grad_norm": 0.6539834762026684, | |
| "learning_rate": 2.710808365197e-06, | |
| "loss": 0.0555, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.3739703459637562, | |
| "grad_norm": 0.39905302203795334, | |
| "learning_rate": 2.6852228031481837e-06, | |
| "loss": 0.0408, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.3772652388797364, | |
| "grad_norm": 0.45958447904532335, | |
| "learning_rate": 2.6597141567527614e-06, | |
| "loss": 0.0503, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.3805601317957166, | |
| "grad_norm": 0.4995326285015215, | |
| "learning_rate": 2.6342832736146403e-06, | |
| "loss": 0.0605, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.3838550247116967, | |
| "grad_norm": 0.4884609849725302, | |
| "learning_rate": 2.608930998753809e-06, | |
| "loss": 0.0602, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.3871499176276771, | |
| "grad_norm": 0.48846990238806653, | |
| "learning_rate": 2.5836581745782474e-06, | |
| "loss": 0.0429, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.3904448105436573, | |
| "grad_norm": 0.5807005137969414, | |
| "learning_rate": 2.558465640855943e-06, | |
| "loss": 0.052, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.3937397034596375, | |
| "grad_norm": 0.3856666040192254, | |
| "learning_rate": 2.533354234686979e-06, | |
| "loss": 0.0422, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.3970345963756179, | |
| "grad_norm": 0.3701363763937253, | |
| "learning_rate": 2.508324790475731e-06, | |
| "loss": 0.0449, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.400329489291598, | |
| "grad_norm": 0.5283156125790535, | |
| "learning_rate": 2.4833781399031275e-06, | |
| "loss": 0.0583, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.4036243822075782, | |
| "grad_norm": 0.5143083208475716, | |
| "learning_rate": 2.4585151118990286e-06, | |
| "loss": 0.0582, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.4069192751235584, | |
| "grad_norm": 0.4580082823859306, | |
| "learning_rate": 2.433736532614666e-06, | |
| "loss": 0.0503, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.4102141680395386, | |
| "grad_norm": 0.4733586276806861, | |
| "learning_rate": 2.4090432253952113e-06, | |
| "loss": 0.0595, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.413509060955519, | |
| "grad_norm": 0.46027613089003067, | |
| "learning_rate": 2.3844360107523973e-06, | |
| "loss": 0.0334, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.4168039538714992, | |
| "grad_norm": 0.618163403358967, | |
| "learning_rate": 2.3599157063372712e-06, | |
| "loss": 0.0505, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.4200988467874793, | |
| "grad_norm": 0.5692914543756001, | |
| "learning_rate": 2.3354831269130133e-06, | |
| "loss": 0.047, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.4233937397034597, | |
| "grad_norm": 0.5569933619176715, | |
| "learning_rate": 2.3111390843278743e-06, | |
| "loss": 0.0506, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.42668863261944, | |
| "grad_norm": 0.4384099356121434, | |
| "learning_rate": 2.2868843874881856e-06, | |
| "loss": 0.0453, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.42998352553542, | |
| "grad_norm": 0.5320473584418453, | |
| "learning_rate": 2.2627198423314988e-06, | |
| "loss": 0.0547, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.4332784184514002, | |
| "grad_norm": 0.4951776816150561, | |
| "learning_rate": 2.238646251799787e-06, | |
| "loss": 0.0517, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.4365733113673804, | |
| "grad_norm": 0.5305051346570233, | |
| "learning_rate": 2.2146644158127827e-06, | |
| "loss": 0.0508, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.4398682042833608, | |
| "grad_norm": 0.48235120417487776, | |
| "learning_rate": 2.1907751312413793e-06, | |
| "loss": 0.0498, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.443163097199341, | |
| "grad_norm": 0.7575565682766872, | |
| "learning_rate": 2.1669791918811724e-06, | |
| "loss": 0.0482, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.4464579901153214, | |
| "grad_norm": 0.6122464829305898, | |
| "learning_rate": 2.1432773884260627e-06, | |
| "loss": 0.0661, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.4497528830313016, | |
| "grad_norm": 0.49382428143445756, | |
| "learning_rate": 2.119670508442004e-06, | |
| "loss": 0.0372, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.4530477759472817, | |
| "grad_norm": 0.6113296705934868, | |
| "learning_rate": 2.0961593363408154e-06, | |
| "loss": 0.0489, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.456342668863262, | |
| "grad_norm": 0.4764803472849658, | |
| "learning_rate": 2.0727446533541302e-06, | |
| "loss": 0.0426, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.459637561779242, | |
| "grad_norm": 0.5321931460957434, | |
| "learning_rate": 2.0494272375074247e-06, | |
| "loss": 0.0428, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.4629324546952225, | |
| "grad_norm": 0.43368533141343174, | |
| "learning_rate": 2.0262078635941818e-06, | |
| "loss": 0.0377, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.4662273476112027, | |
| "grad_norm": 0.5227900476116077, | |
| "learning_rate": 2.0030873031501274e-06, | |
| "loss": 0.048, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.4695222405271828, | |
| "grad_norm": 0.40044438580877817, | |
| "learning_rate": 1.980066324427613e-06, | |
| "loss": 0.0367, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.4728171334431632, | |
| "grad_norm": 0.42569057497544066, | |
| "learning_rate": 1.9571456923700696e-06, | |
| "loss": 0.0485, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.4761120263591434, | |
| "grad_norm": 0.5011955876540544, | |
| "learning_rate": 1.9343261685866054e-06, | |
| "loss": 0.0684, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.4794069192751236, | |
| "grad_norm": 0.5257059685422952, | |
| "learning_rate": 1.911608511326688e-06, | |
| "loss": 0.0469, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.4827018121911038, | |
| "grad_norm": 0.5330212717649231, | |
| "learning_rate": 1.8889934754549583e-06, | |
| "loss": 0.0615, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.485996705107084, | |
| "grad_norm": 0.4377288880184422, | |
| "learning_rate": 1.8664818124261375e-06, | |
| "loss": 0.04, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.4892915980230643, | |
| "grad_norm": 0.4821221712040424, | |
| "learning_rate": 1.8440742702600706e-06, | |
| "loss": 0.0496, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.4925864909390445, | |
| "grad_norm": 0.42358079608202237, | |
| "learning_rate": 1.8217715935168562e-06, | |
| "loss": 0.0446, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.4958813838550247, | |
| "grad_norm": 0.6521628225316723, | |
| "learning_rate": 1.7995745232721207e-06, | |
| "loss": 0.0665, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.499176276771005, | |
| "grad_norm": 0.5512352891912379, | |
| "learning_rate": 1.777483797092381e-06, | |
| "loss": 0.0527, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.5024711696869852, | |
| "grad_norm": 0.4132207895248971, | |
| "learning_rate": 1.755500149010549e-06, | |
| "loss": 0.0369, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.5057660626029654, | |
| "grad_norm": 0.5452488198197322, | |
| "learning_rate": 1.7336243095015271e-06, | |
| "loss": 0.0457, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.5090609555189456, | |
| "grad_norm": 1.8013972479013802, | |
| "learning_rate": 1.7118570054579508e-06, | |
| "loss": 0.0788, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.5123558484349258, | |
| "grad_norm": 0.6158971711077378, | |
| "learning_rate": 1.6901989601660224e-06, | |
| "loss": 0.0577, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.515650741350906, | |
| "grad_norm": 0.6402888520963839, | |
| "learning_rate": 1.6686508932814871e-06, | |
| "loss": 0.0426, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.5189456342668863, | |
| "grad_norm": 0.5815365915637473, | |
| "learning_rate": 1.6472135208057128e-06, | |
| "loss": 0.0526, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.5222405271828665, | |
| "grad_norm": 0.5219074399966507, | |
| "learning_rate": 1.625887555061907e-06, | |
| "loss": 0.0428, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.525535420098847, | |
| "grad_norm": 0.5007230705662209, | |
| "learning_rate": 1.6046737046714366e-06, | |
| "loss": 0.0386, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.528830313014827, | |
| "grad_norm": 0.492364769802372, | |
| "learning_rate": 1.5835726745302953e-06, | |
| "loss": 0.0364, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.5321252059308073, | |
| "grad_norm": 0.5652000556154251, | |
| "learning_rate": 1.5625851657856666e-06, | |
| "loss": 0.0546, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.5354200988467874, | |
| "grad_norm": 0.5993897339775979, | |
| "learning_rate": 1.5417118758126408e-06, | |
| "loss": 0.0579, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.5387149917627676, | |
| "grad_norm": 1.3251444571487765, | |
| "learning_rate": 1.520953498191028e-06, | |
| "loss": 0.0747, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.5420098846787478, | |
| "grad_norm": 0.440371155414081, | |
| "learning_rate": 1.5003107226823255e-06, | |
| "loss": 0.0495, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.5453047775947282, | |
| "grad_norm": 0.5274460518323345, | |
| "learning_rate": 1.479784235206786e-06, | |
| "loss": 0.0457, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.5485996705107083, | |
| "grad_norm": 0.4509159507608483, | |
| "learning_rate": 1.459374717820637e-06, | |
| "loss": 0.0441, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.5518945634266887, | |
| "grad_norm": 0.5787329784185842, | |
| "learning_rate": 1.439082848693406e-06, | |
| "loss": 0.0455, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.555189456342669, | |
| "grad_norm": 0.9428413760935695, | |
| "learning_rate": 1.4189093020853989e-06, | |
| "loss": 0.0635, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.558484349258649, | |
| "grad_norm": 0.5810607886116554, | |
| "learning_rate": 1.3988547483252812e-06, | |
| "loss": 0.0591, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.5617792421746293, | |
| "grad_norm": 0.5621586581333317, | |
| "learning_rate": 1.3789198537878202e-06, | |
| "loss": 0.048, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.5650741350906094, | |
| "grad_norm": 0.614476271893997, | |
| "learning_rate": 1.3591052808717258e-06, | |
| "loss": 0.0574, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.5683690280065898, | |
| "grad_norm": 0.4366280882804736, | |
| "learning_rate": 1.339411687977657e-06, | |
| "loss": 0.0387, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.57166392092257, | |
| "grad_norm": 1.654229386447125, | |
| "learning_rate": 1.3198397294863285e-06, | |
| "loss": 0.0525, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.5749588138385504, | |
| "grad_norm": 0.5124984935464315, | |
| "learning_rate": 1.3003900557367816e-06, | |
| "loss": 0.0586, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.5782537067545306, | |
| "grad_norm": 0.5039902746309534, | |
| "learning_rate": 1.281063313004761e-06, | |
| "loss": 0.0409, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.5815485996705108, | |
| "grad_norm": 0.4453799136874429, | |
| "learning_rate": 1.261860143481255e-06, | |
| "loss": 0.0437, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.584843492586491, | |
| "grad_norm": 0.44736265726220936, | |
| "learning_rate": 1.2427811852511396e-06, | |
| "loss": 0.05, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.588138385502471, | |
| "grad_norm": 0.5751552043472024, | |
| "learning_rate": 1.223827072271993e-06, | |
| "loss": 0.0513, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.5914332784184513, | |
| "grad_norm": 0.4854076213664054, | |
| "learning_rate": 1.204998434353018e-06, | |
| "loss": 0.0434, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.5947281713344317, | |
| "grad_norm": 0.5304616858985192, | |
| "learning_rate": 1.1862958971341199e-06, | |
| "loss": 0.0537, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.5980230642504119, | |
| "grad_norm": 0.5357970833666896, | |
| "learning_rate": 1.1677200820651197e-06, | |
| "loss": 0.049, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.6013179571663922, | |
| "grad_norm": 0.6703644083736745, | |
| "learning_rate": 1.1492716063850973e-06, | |
| "loss": 0.0553, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.6046128500823724, | |
| "grad_norm": 0.5104289346948437, | |
| "learning_rate": 1.1309510831018927e-06, | |
| "loss": 0.0484, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.6079077429983526, | |
| "grad_norm": 0.791691486031595, | |
| "learning_rate": 1.112759120971723e-06, | |
| "loss": 0.0516, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.6112026359143328, | |
| "grad_norm": 0.5044446696201748, | |
| "learning_rate": 1.09469632447897e-06, | |
| "loss": 0.0412, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.614497528830313, | |
| "grad_norm": 0.44744812374789733, | |
| "learning_rate": 1.0767632938160787e-06, | |
| "loss": 0.0441, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.6177924217462931, | |
| "grad_norm": 0.689652904031741, | |
| "learning_rate": 1.0589606248636291e-06, | |
| "loss": 0.0468, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.6210873146622735, | |
| "grad_norm": 0.4644587386351254, | |
| "learning_rate": 1.0412889091705242e-06, | |
| "loss": 0.0356, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.6243822075782537, | |
| "grad_norm": 0.6499107202235256, | |
| "learning_rate": 1.0237487339343382e-06, | |
| "loss": 0.0574, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 1.627677100494234, | |
| "grad_norm": 0.4542177117918383, | |
| "learning_rate": 1.0063406819818106e-06, | |
| "loss": 0.0443, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 1.6309719934102143, | |
| "grad_norm": 0.6343789726555299, | |
| "learning_rate": 9.890653317494686e-07, | |
| "loss": 0.0524, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.6342668863261944, | |
| "grad_norm": 0.4208852075289343, | |
| "learning_rate": 9.719232572644189e-07, | |
| "loss": 0.0407, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 1.6375617792421746, | |
| "grad_norm": 0.45018417664569393, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.0382, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 1.6408566721581548, | |
| "grad_norm": 0.4664038740894182, | |
| "learning_rate": 9.380412094831809e-07, | |
| "loss": 0.0413, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 1.644151565074135, | |
| "grad_norm": 0.5658115763517576, | |
| "learning_rate": 9.213023620231404e-07, | |
| "loss": 0.055, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 1.6474464579901154, | |
| "grad_norm": 0.42505631586549236, | |
| "learning_rate": 9.046990419452795e-07, | |
| "loss": 0.0374, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.6507413509060955, | |
| "grad_norm": 0.5554370313066022, | |
| "learning_rate": 8.882318009464124e-07, | |
| "loss": 0.0758, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 1.654036243822076, | |
| "grad_norm": 0.6820183163387327, | |
| "learning_rate": 8.719011862017108e-07, | |
| "loss": 0.067, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.657331136738056, | |
| "grad_norm": 0.8815741831945997, | |
| "learning_rate": 8.557077403465069e-07, | |
| "loss": 0.0635, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 1.6606260296540363, | |
| "grad_norm": 0.6530261534927284, | |
| "learning_rate": 8.396520014582798e-07, | |
| "loss": 0.0564, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 1.6639209225700164, | |
| "grad_norm": 0.5563910425802013, | |
| "learning_rate": 8.237345030387589e-07, | |
| "loss": 0.0568, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.6672158154859966, | |
| "grad_norm": 1.6616538016948608, | |
| "learning_rate": 8.079557739962129e-07, | |
| "loss": 0.0433, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 1.6705107084019768, | |
| "grad_norm": 0.4729743527848457, | |
| "learning_rate": 7.923163386278615e-07, | |
| "loss": 0.0477, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 1.6738056013179572, | |
| "grad_norm": 0.484207261501026, | |
| "learning_rate": 7.768167166024637e-07, | |
| "loss": 0.0393, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 1.6771004942339374, | |
| "grad_norm": 0.4347790202564516, | |
| "learning_rate": 7.614574229430432e-07, | |
| "loss": 0.0348, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 1.6803953871499178, | |
| "grad_norm": 0.5159230901740568, | |
| "learning_rate": 7.462389680097831e-07, | |
| "loss": 0.0511, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.683690280065898, | |
| "grad_norm": 0.7846302974584749, | |
| "learning_rate": 7.31161857483057e-07, | |
| "loss": 0.0428, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 1.6869851729818781, | |
| "grad_norm": 0.5154541689981792, | |
| "learning_rate": 7.162265923466383e-07, | |
| "loss": 0.0481, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 1.6902800658978583, | |
| "grad_norm": 0.5103769113667321, | |
| "learning_rate": 7.014336688710411e-07, | |
| "loss": 0.0559, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 1.6935749588138385, | |
| "grad_norm": 0.5601518054326986, | |
| "learning_rate": 6.867835785970417e-07, | |
| "loss": 0.0383, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 1.6968698517298186, | |
| "grad_norm": 0.452690076025677, | |
| "learning_rate": 6.722768083193354e-07, | |
| "loss": 0.0393, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.700164744645799, | |
| "grad_norm": 0.46134612749678455, | |
| "learning_rate": 6.579138400703716e-07, | |
| "loss": 0.0515, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 1.7034596375617792, | |
| "grad_norm": 0.41746324511751276, | |
| "learning_rate": 6.436951511043243e-07, | |
| "loss": 0.0445, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 1.7067545304777596, | |
| "grad_norm": 0.45014253067830906, | |
| "learning_rate": 6.296212138812474e-07, | |
| "loss": 0.0438, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 1.7100494233937398, | |
| "grad_norm": 0.49892377023307155, | |
| "learning_rate": 6.156924960513638e-07, | |
| "loss": 0.0452, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 1.71334431630972, | |
| "grad_norm": 0.47105725038366275, | |
| "learning_rate": 6.019094604395359e-07, | |
| "loss": 0.054, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.7166392092257001, | |
| "grad_norm": 0.4259013974972623, | |
| "learning_rate": 5.882725650298787e-07, | |
| "loss": 0.0377, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 1.7199341021416803, | |
| "grad_norm": 0.7122055487087868, | |
| "learning_rate": 5.747822629505484e-07, | |
| "loss": 0.0606, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 1.7232289950576605, | |
| "grad_norm": 0.6463022317948001, | |
| "learning_rate": 5.614390024586808e-07, | |
| "loss": 0.0948, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 1.7265238879736409, | |
| "grad_norm": 0.6557315175026269, | |
| "learning_rate": 5.482432269255011e-07, | |
| "loss": 0.0594, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 1.729818780889621, | |
| "grad_norm": 0.48159732485479767, | |
| "learning_rate": 5.351953748215872e-07, | |
| "loss": 0.0562, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.7331136738056014, | |
| "grad_norm": 0.5429610524246544, | |
| "learning_rate": 5.222958797023036e-07, | |
| "loss": 0.0469, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 1.7364085667215816, | |
| "grad_norm": 0.4703329992162841, | |
| "learning_rate": 5.095451701933923e-07, | |
| "loss": 0.0495, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.7397034596375618, | |
| "grad_norm": 0.4436763771779832, | |
| "learning_rate": 4.969436699767344e-07, | |
| "loss": 0.0354, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.742998352553542, | |
| "grad_norm": 0.5025190325700492, | |
| "learning_rate": 4.844917977762653e-07, | |
| "loss": 0.056, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.7462932454695221, | |
| "grad_norm": 0.661761525862607, | |
| "learning_rate": 4.721899673440694e-07, | |
| "loss": 0.0436, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.7495881383855023, | |
| "grad_norm": 0.5762138308981635, | |
| "learning_rate": 4.6003858744662564e-07, | |
| "loss": 0.0552, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 1.7528830313014827, | |
| "grad_norm": 0.5387051088420545, | |
| "learning_rate": 4.4803806185122866e-07, | |
| "loss": 0.0479, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 1.7561779242174629, | |
| "grad_norm": 0.537079178923195, | |
| "learning_rate": 4.361887893125677e-07, | |
| "loss": 0.0565, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.7594728171334433, | |
| "grad_norm": 0.5780295144664594, | |
| "learning_rate": 4.244911635594856e-07, | |
| "loss": 0.0555, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 1.7627677100494235, | |
| "grad_norm": 0.6971174595241891, | |
| "learning_rate": 4.1294557328188376e-07, | |
| "loss": 0.0468, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.7660626029654036, | |
| "grad_norm": 0.5714544842706962, | |
| "learning_rate": 4.0155240211781966e-07, | |
| "loss": 0.069, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 1.7693574958813838, | |
| "grad_norm": 0.5560933792076089, | |
| "learning_rate": 3.9031202864074634e-07, | |
| "loss": 0.0526, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 1.772652388797364, | |
| "grad_norm": 0.5076535062447541, | |
| "learning_rate": 3.7922482634694667e-07, | |
| "loss": 0.0495, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 1.7759472817133442, | |
| "grad_norm": 0.4733107950145234, | |
| "learning_rate": 3.6829116364310914e-07, | |
| "loss": 0.048, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 1.7792421746293245, | |
| "grad_norm": 0.5645137566585913, | |
| "learning_rate": 3.575114038340977e-07, | |
| "loss": 0.0503, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.782537067545305, | |
| "grad_norm": 0.5224355605212421, | |
| "learning_rate": 3.4688590511087304e-07, | |
| "loss": 0.0553, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 1.7858319604612851, | |
| "grad_norm": 0.495152224806279, | |
| "learning_rate": 3.3641502053859355e-07, | |
| "loss": 0.0304, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 1.7891268533772653, | |
| "grad_norm": 0.40973303317244536, | |
| "learning_rate": 3.2609909804488195e-07, | |
| "loss": 0.0313, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 1.7924217462932455, | |
| "grad_norm": 0.47058046777642437, | |
| "learning_rate": 3.159384804082666e-07, | |
| "loss": 0.0526, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 1.7957166392092256, | |
| "grad_norm": 0.4838577514212764, | |
| "learning_rate": 3.0593350524678823e-07, | |
| "loss": 0.0371, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.7990115321252058, | |
| "grad_norm": 0.6690590772761237, | |
| "learning_rate": 2.9608450500678566e-07, | |
| "loss": 0.0604, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.8023064250411862, | |
| "grad_norm": 0.4109056760035354, | |
| "learning_rate": 2.863918069518451e-07, | |
| "loss": 0.0331, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 1.8056013179571664, | |
| "grad_norm": 0.7159007380510568, | |
| "learning_rate": 2.7685573315192895e-07, | |
| "loss": 0.0721, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 1.8088962108731468, | |
| "grad_norm": 0.4865674890840018, | |
| "learning_rate": 2.67476600472672e-07, | |
| "loss": 0.0451, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 1.812191103789127, | |
| "grad_norm": 0.5942738860891101, | |
| "learning_rate": 2.5825472056485556e-07, | |
| "loss": 0.062, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.8154859967051071, | |
| "grad_norm": 0.5357536121944303, | |
| "learning_rate": 2.4919039985404626e-07, | |
| "loss": 0.0609, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 1.8187808896210873, | |
| "grad_norm": 0.3866866163849165, | |
| "learning_rate": 2.4028393953042074e-07, | |
| "loss": 0.0296, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 1.8220757825370675, | |
| "grad_norm": 0.49750309911350266, | |
| "learning_rate": 2.315356355387527e-07, | |
| "loss": 0.0444, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 1.8253706754530477, | |
| "grad_norm": 0.6734726196583376, | |
| "learning_rate": 2.2294577856858236e-07, | |
| "loss": 0.0552, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 1.828665568369028, | |
| "grad_norm": 0.4599661443293077, | |
| "learning_rate": 2.1451465404455473e-07, | |
| "loss": 0.041, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.8319604612850082, | |
| "grad_norm": 0.691277736971953, | |
| "learning_rate": 2.0624254211693894e-07, | |
| "loss": 0.061, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 1.8352553542009886, | |
| "grad_norm": 0.42780093096680555, | |
| "learning_rate": 1.9812971765231394e-07, | |
| "loss": 0.0364, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 1.8385502471169688, | |
| "grad_norm": 0.578589972444535, | |
| "learning_rate": 1.901764502244424e-07, | |
| "loss": 0.0613, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 1.841845140032949, | |
| "grad_norm": 0.4763862013898226, | |
| "learning_rate": 1.823830041053065e-07, | |
| "loss": 0.0437, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 1.8451400329489291, | |
| "grad_norm": 0.45718093901396817, | |
| "learning_rate": 1.7474963825633185e-07, | |
| "loss": 0.043, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.8484349258649093, | |
| "grad_norm": 0.6366482024659451, | |
| "learning_rate": 1.6727660631977894e-07, | |
| "loss": 0.0556, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 1.8517298187808895, | |
| "grad_norm": 0.7471772626827173, | |
| "learning_rate": 1.5996415661031662e-07, | |
| "loss": 0.0425, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 1.8550247116968699, | |
| "grad_norm": 0.3521582825319659, | |
| "learning_rate": 1.528125321067725e-07, | |
| "loss": 0.0382, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 1.85831960461285, | |
| "grad_norm": 0.5973061078363581, | |
| "learning_rate": 1.4582197044405556e-07, | |
| "loss": 0.0509, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 1.8616144975288305, | |
| "grad_norm": 0.5093523377380892, | |
| "learning_rate": 1.389927039052652e-07, | |
| "loss": 0.0444, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.8649093904448106, | |
| "grad_norm": 0.44171165477963154, | |
| "learning_rate": 1.323249594139664e-07, | |
| "loss": 0.0468, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 1.8682042833607908, | |
| "grad_norm": 0.44675757705445573, | |
| "learning_rate": 1.2581895852665671e-07, | |
| "loss": 0.0374, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 1.871499176276771, | |
| "grad_norm": 0.6020514933209717, | |
| "learning_rate": 1.1947491742539841e-07, | |
| "loss": 0.0503, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 1.8747940691927512, | |
| "grad_norm": 0.42586058680868216, | |
| "learning_rate": 1.1329304691063692e-07, | |
| "loss": 0.0392, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 1.8780889621087313, | |
| "grad_norm": 0.4423611658945744, | |
| "learning_rate": 1.0727355239419868e-07, | |
| "loss": 0.0469, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.8813838550247117, | |
| "grad_norm": 0.424883733072018, | |
| "learning_rate": 1.014166338924627e-07, | |
| "loss": 0.0475, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 1.884678747940692, | |
| "grad_norm": 0.508275043696402, | |
| "learning_rate": 9.572248601971646e-08, | |
| "loss": 0.0588, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 1.8879736408566723, | |
| "grad_norm": 0.532268589710946, | |
| "learning_rate": 9.019129798168658e-08, | |
| "loss": 0.0413, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 1.8912685337726525, | |
| "grad_norm": 0.49075007985093444, | |
| "learning_rate": 8.482325356925614e-08, | |
| "loss": 0.0438, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 1.8945634266886326, | |
| "grad_norm": 0.6405783999237776, | |
| "learning_rate": 7.96185311523523e-08, | |
| "loss": 0.0517, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.8978583196046128, | |
| "grad_norm": 0.4374824821968711, | |
| "learning_rate": 7.45773036740255e-08, | |
| "loss": 0.0594, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 1.901153212520593, | |
| "grad_norm": 0.4881691605292657, | |
| "learning_rate": 6.969973864469626e-08, | |
| "loss": 0.0478, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 1.9044481054365732, | |
| "grad_norm": 0.4169109199296669, | |
| "learning_rate": 6.498599813659524e-08, | |
| "loss": 0.0329, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 1.9077429983525536, | |
| "grad_norm": 0.7757102581030492, | |
| "learning_rate": 6.043623877837301e-08, | |
| "loss": 0.0459, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 1.9110378912685337, | |
| "grad_norm": 0.6871632648014142, | |
| "learning_rate": 5.6050611749899896e-08, | |
| "loss": 0.0499, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.9143327841845141, | |
| "grad_norm": 0.4909555169664079, | |
| "learning_rate": 5.182926277723821e-08, | |
| "loss": 0.0439, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 1.9176276771004943, | |
| "grad_norm": 0.5085351517406403, | |
| "learning_rate": 4.777233212780396e-08, | |
| "loss": 0.043, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 1.9209225700164745, | |
| "grad_norm": 0.4133261269232907, | |
| "learning_rate": 4.387995460570282e-08, | |
| "loss": 0.0422, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 1.9242174629324547, | |
| "grad_norm": 0.3457486280213094, | |
| "learning_rate": 4.015225954725421e-08, | |
| "loss": 0.0302, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 1.9275123558484348, | |
| "grad_norm": 0.5306123017488412, | |
| "learning_rate": 3.658937081669034e-08, | |
| "loss": 0.0347, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.930807248764415, | |
| "grad_norm": 0.6170841106215647, | |
| "learning_rate": 3.3191406802041693e-08, | |
| "loss": 0.0427, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 1.9341021416803954, | |
| "grad_norm": 0.45743745755641063, | |
| "learning_rate": 2.9958480411204086e-08, | |
| "loss": 0.0487, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 1.9373970345963756, | |
| "grad_norm": 0.7985516650647139, | |
| "learning_rate": 2.6890699068187197e-08, | |
| "loss": 0.0598, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 1.940691927512356, | |
| "grad_norm": 0.969435591211712, | |
| "learning_rate": 2.3988164709542462e-08, | |
| "loss": 0.046, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 1.9439868204283361, | |
| "grad_norm": 0.5530499902423203, | |
| "learning_rate": 2.1250973780977957e-08, | |
| "loss": 0.0626, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.9472817133443163, | |
| "grad_norm": 0.4726855774076432, | |
| "learning_rate": 1.8679217234154335e-08, | |
| "loss": 0.0442, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 1.9505766062602965, | |
| "grad_norm": 0.5910043256950622, | |
| "learning_rate": 1.627298052366111e-08, | |
| "loss": 0.0533, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 1.9538714991762767, | |
| "grad_norm": 0.37046514170506584, | |
| "learning_rate": 1.4032343604177267e-08, | |
| "loss": 0.0436, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 1.9571663920922568, | |
| "grad_norm": 0.45020576348906904, | |
| "learning_rate": 1.1957380927816176e-08, | |
| "loss": 0.0392, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 1.9604612850082372, | |
| "grad_norm": 0.4456283814258553, | |
| "learning_rate": 1.0048161441649217e-08, | |
| "loss": 0.0464, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.9637561779242174, | |
| "grad_norm": 0.5191136185767354, | |
| "learning_rate": 8.304748585417077e-09, | |
| "loss": 0.0432, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 1.9670510708401978, | |
| "grad_norm": 0.7541935440652175, | |
| "learning_rate": 6.72720028942031e-09, | |
| "loss": 0.0417, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 1.970345963756178, | |
| "grad_norm": 0.4930059149169314, | |
| "learning_rate": 5.315568972594775e-09, | |
| "loss": 0.0522, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 1.9736408566721582, | |
| "grad_norm": 0.43182626877243907, | |
| "learning_rate": 4.0699015407702495e-09, | |
| "loss": 0.0426, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 1.9769357495881383, | |
| "grad_norm": 0.5518177594295087, | |
| "learning_rate": 2.990239385112226e-09, | |
| "loss": 0.0565, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.9802306425041185, | |
| "grad_norm": 0.5188892218788691, | |
| "learning_rate": 2.076618380744133e-09, | |
| "loss": 0.0684, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 1.9835255354200987, | |
| "grad_norm": 0.43199397069665996, | |
| "learning_rate": 1.3290688855588374e-09, | |
| "loss": 0.0396, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 1.986820428336079, | |
| "grad_norm": 0.4470921544911916, | |
| "learning_rate": 7.476157392072303e-10, | |
| "loss": 0.0385, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 1.9901153212520593, | |
| "grad_norm": 0.5332848840635471, | |
| "learning_rate": 3.322782622738885e-10, | |
| "loss": 0.0585, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 1.9934102141680397, | |
| "grad_norm": 0.6284366325020901, | |
| "learning_rate": 8.307025563536464e-11, | |
| "loss": 0.0517, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.9967051070840198, | |
| "grad_norm": 0.6056246452486423, | |
| "learning_rate": 0.0, | |
| "loss": 0.048, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 1.9967051070840198, | |
| "eval_loss": 0.0587012954056263, | |
| "eval_runtime": 144.0131, | |
| "eval_samples_per_second": 35.448, | |
| "eval_steps_per_second": 1.111, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 1.9967051070840198, | |
| "step": 606, | |
| "total_flos": 1.811911707237417e+17, | |
| "train_loss": 0.0716345354195426, | |
| "train_runtime": 6724.9995, | |
| "train_samples_per_second": 11.537, | |
| "train_steps_per_second": 0.09 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 606, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "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": 1.811911707237417e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |