| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9997505612372163, | |
| "eval_steps": 500, | |
| "global_step": 1002, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0009977550511349464, | |
| "grad_norm": 3.95422100343339e+17, | |
| "learning_rate": 1.9801980198019803e-07, | |
| "loss": 1.1312, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.004988775255674732, | |
| "grad_norm": 2182.5525466022195, | |
| "learning_rate": 9.900990099009902e-07, | |
| "loss": 1.1202, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.009977550511349464, | |
| "grad_norm": 10.696306402986602, | |
| "learning_rate": 1.9801980198019803e-06, | |
| "loss": 1.0843, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.014966325767024195, | |
| "grad_norm": 4.8421359026925295, | |
| "learning_rate": 2.9702970297029703e-06, | |
| "loss": 1.024, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.01995510102269893, | |
| "grad_norm": 1.4664994361442427, | |
| "learning_rate": 3.960396039603961e-06, | |
| "loss": 0.9479, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.024943876278373658, | |
| "grad_norm": 1.1625196109582225, | |
| "learning_rate": 4.950495049504951e-06, | |
| "loss": 0.8935, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.02993265153404839, | |
| "grad_norm": 1.2801604519342376, | |
| "learning_rate": 5.940594059405941e-06, | |
| "loss": 0.871, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.034921426789723126, | |
| "grad_norm": 2.033559906503939, | |
| "learning_rate": 6.930693069306931e-06, | |
| "loss": 0.8425, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.03991020204539786, | |
| "grad_norm": 1.2072354289878107, | |
| "learning_rate": 7.920792079207921e-06, | |
| "loss": 0.8193, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04489897730107259, | |
| "grad_norm": 1.3137258088382902, | |
| "learning_rate": 8.910891089108911e-06, | |
| "loss": 0.8013, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.049887752556747315, | |
| "grad_norm": 1.1213863664000594, | |
| "learning_rate": 9.900990099009901e-06, | |
| "loss": 0.7906, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05487652781242205, | |
| "grad_norm": 0.9999895583902438, | |
| "learning_rate": 1.0891089108910893e-05, | |
| "loss": 0.7642, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.05986530306809678, | |
| "grad_norm": 1.3377086334649673, | |
| "learning_rate": 1.1881188118811881e-05, | |
| "loss": 0.7495, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06485407832377152, | |
| "grad_norm": 1.0583607289478394, | |
| "learning_rate": 1.2871287128712873e-05, | |
| "loss": 0.7328, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.06984285357944625, | |
| "grad_norm": 1.3493827534349543, | |
| "learning_rate": 1.3861386138613861e-05, | |
| "loss": 0.7383, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.07483162883512098, | |
| "grad_norm": 1.2261995345556986, | |
| "learning_rate": 1.4851485148514853e-05, | |
| "loss": 0.7281, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.07982040409079572, | |
| "grad_norm": 1.3328058553211537, | |
| "learning_rate": 1.5841584158415843e-05, | |
| "loss": 0.7236, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.08480917934647045, | |
| "grad_norm": 1.1145579124084846, | |
| "learning_rate": 1.683168316831683e-05, | |
| "loss": 0.7255, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.08979795460214518, | |
| "grad_norm": 0.9193198348331784, | |
| "learning_rate": 1.7821782178217823e-05, | |
| "loss": 0.7181, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.0947867298578199, | |
| "grad_norm": 0.9946484577080871, | |
| "learning_rate": 1.881188118811881e-05, | |
| "loss": 0.7103, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.09977550511349463, | |
| "grad_norm": 1.041681712316516, | |
| "learning_rate": 1.9801980198019803e-05, | |
| "loss": 0.7007, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.10476428036916936, | |
| "grad_norm": 2.164939078852322, | |
| "learning_rate": 1.9999027402586235e-05, | |
| "loss": 0.7064, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.1097530556248441, | |
| "grad_norm": 2.7326170629817335, | |
| "learning_rate": 1.9995076549835638e-05, | |
| "loss": 0.7129, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.11474183088051883, | |
| "grad_norm": 0.958937879949106, | |
| "learning_rate": 1.9988087854284224e-05, | |
| "loss": 0.6984, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.11973060613619356, | |
| "grad_norm": 1.1161507946755316, | |
| "learning_rate": 1.997806344003363e-05, | |
| "loss": 0.7025, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.12471938139186829, | |
| "grad_norm": 0.9783646600814512, | |
| "learning_rate": 1.996500635384337e-05, | |
| "loss": 0.6918, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.12970815664754304, | |
| "grad_norm": 0.8060832238745128, | |
| "learning_rate": 1.994892056420485e-05, | |
| "loss": 0.6842, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.13469693190321777, | |
| "grad_norm": 0.8669120420326126, | |
| "learning_rate": 1.992981096013517e-05, | |
| "loss": 0.6806, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.1396857071588925, | |
| "grad_norm": 0.8551160872182698, | |
| "learning_rate": 1.990768334969122e-05, | |
| "loss": 0.6908, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.14467448241456723, | |
| "grad_norm": 0.6818189908779505, | |
| "learning_rate": 1.9882544458204386e-05, | |
| "loss": 0.6888, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.14966325767024197, | |
| "grad_norm": 0.7649549705168317, | |
| "learning_rate": 1.9854401926236518e-05, | |
| "loss": 0.6867, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1546520329259167, | |
| "grad_norm": 0.8499401555713652, | |
| "learning_rate": 1.9823264307257683e-05, | |
| "loss": 0.6707, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.15964080818159143, | |
| "grad_norm": 0.7166408203324516, | |
| "learning_rate": 1.9789141065046495e-05, | |
| "loss": 0.676, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.16462958343726616, | |
| "grad_norm": 0.6182479498389858, | |
| "learning_rate": 1.9752042570813733e-05, | |
| "loss": 0.6738, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.1696183586929409, | |
| "grad_norm": 0.6690357471493384, | |
| "learning_rate": 1.9711980100050196e-05, | |
| "loss": 0.6672, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.17460713394861563, | |
| "grad_norm": 0.5822226665191199, | |
| "learning_rate": 1.966896582909968e-05, | |
| "loss": 0.6736, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.17959590920429036, | |
| "grad_norm": 0.7406095161368872, | |
| "learning_rate": 1.962301283145819e-05, | |
| "loss": 0.6761, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.18458468445996506, | |
| "grad_norm": 0.7296907738291923, | |
| "learning_rate": 1.957413507380046e-05, | |
| "loss": 0.6678, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.1895734597156398, | |
| "grad_norm": 0.8518753276369776, | |
| "learning_rate": 1.952234741173499e-05, | |
| "loss": 0.6733, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.19456223497131453, | |
| "grad_norm": 0.5554147873181055, | |
| "learning_rate": 1.946766558528895e-05, | |
| "loss": 0.6621, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.19955101022698926, | |
| "grad_norm": 0.6525658967181038, | |
| "learning_rate": 1.941010621412422e-05, | |
| "loss": 0.6649, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.204539785482664, | |
| "grad_norm": 0.5780920597815026, | |
| "learning_rate": 1.9349686792486143e-05, | |
| "loss": 0.657, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.20952856073833873, | |
| "grad_norm": 0.5578703671497164, | |
| "learning_rate": 1.9286425683886403e-05, | |
| "loss": 0.6687, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.21451733599401346, | |
| "grad_norm": 0.72305301104437, | |
| "learning_rate": 1.9220342115521746e-05, | |
| "loss": 0.6624, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.2195061112496882, | |
| "grad_norm": 0.6165020865206677, | |
| "learning_rate": 1.9151456172430186e-05, | |
| "loss": 0.6532, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.22449488650536292, | |
| "grad_norm": 0.5606109731701377, | |
| "learning_rate": 1.9079788791386468e-05, | |
| "loss": 0.6562, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.22948366176103766, | |
| "grad_norm": 0.6353820290325317, | |
| "learning_rate": 1.9005361754538677e-05, | |
| "loss": 0.6671, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.2344724370167124, | |
| "grad_norm": 0.5815877946035953, | |
| "learning_rate": 1.8928197682787914e-05, | |
| "loss": 0.6593, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.23946121227238712, | |
| "grad_norm": 0.6348680697801418, | |
| "learning_rate": 1.8848320028913017e-05, | |
| "loss": 0.6507, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.24444998752806185, | |
| "grad_norm": 0.5695451117286802, | |
| "learning_rate": 1.8765753070442486e-05, | |
| "loss": 0.6523, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.24943876278373658, | |
| "grad_norm": 0.5168318177147343, | |
| "learning_rate": 1.868052190227571e-05, | |
| "loss": 0.6591, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2544275380394113, | |
| "grad_norm": 0.7348069026040828, | |
| "learning_rate": 1.859265242905577e-05, | |
| "loss": 0.6541, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.2594163132950861, | |
| "grad_norm": 0.5329392930102724, | |
| "learning_rate": 1.8502171357296144e-05, | |
| "loss": 0.6589, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.2644050885507608, | |
| "grad_norm": 0.6628783134030882, | |
| "learning_rate": 1.84091061872637e-05, | |
| "loss": 0.6443, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.26939386380643554, | |
| "grad_norm": 0.6959729166297904, | |
| "learning_rate": 1.8313485204620428e-05, | |
| "loss": 0.6459, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.27438263906211025, | |
| "grad_norm": 0.5848246972834032, | |
| "learning_rate": 1.821533747182645e-05, | |
| "loss": 0.6606, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.279371414317785, | |
| "grad_norm": 0.6475331258568309, | |
| "learning_rate": 1.811469281930698e-05, | |
| "loss": 0.656, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.2843601895734597, | |
| "grad_norm": 0.5357406950163816, | |
| "learning_rate": 1.8011581836385828e-05, | |
| "loss": 0.6474, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.28934896482913447, | |
| "grad_norm": 0.5973939212545811, | |
| "learning_rate": 1.790603586198827e-05, | |
| "loss": 0.6376, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.2943377400848092, | |
| "grad_norm": 0.5096252662581786, | |
| "learning_rate": 1.7798086975116096e-05, | |
| "loss": 0.6487, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.29932651534048393, | |
| "grad_norm": 0.545088543038122, | |
| "learning_rate": 1.7687767985097695e-05, | |
| "loss": 0.6526, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.30431529059615864, | |
| "grad_norm": 0.707465419807657, | |
| "learning_rate": 1.7575112421616203e-05, | |
| "loss": 0.6465, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.3093040658518334, | |
| "grad_norm": 0.520929461518716, | |
| "learning_rate": 1.7460154524518688e-05, | |
| "loss": 0.6346, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.3142928411075081, | |
| "grad_norm": 0.5724374980096262, | |
| "learning_rate": 1.73429292334095e-05, | |
| "loss": 0.6533, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.31928161636318286, | |
| "grad_norm": 0.545092535159253, | |
| "learning_rate": 1.722347217703094e-05, | |
| "loss": 0.6437, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.32427039161885757, | |
| "grad_norm": 0.49712776699676936, | |
| "learning_rate": 1.710181966243447e-05, | |
| "loss": 0.6373, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.3292591668745323, | |
| "grad_norm": 0.5463641929477563, | |
| "learning_rate": 1.6978008663945794e-05, | |
| "loss": 0.6496, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.33424794213020703, | |
| "grad_norm": 0.5263242364290428, | |
| "learning_rate": 1.6852076811927066e-05, | |
| "loss": 0.6369, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.3392367173858818, | |
| "grad_norm": 0.5461638677630304, | |
| "learning_rate": 1.672406238133978e-05, | |
| "loss": 0.639, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.3442254926415565, | |
| "grad_norm": 0.5328798983853884, | |
| "learning_rate": 1.6594004280111697e-05, | |
| "loss": 0.6497, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.34921426789723126, | |
| "grad_norm": 0.5449564175014876, | |
| "learning_rate": 1.6461942037311406e-05, | |
| "loss": 0.64, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.35420304315290596, | |
| "grad_norm": 0.4924182140259062, | |
| "learning_rate": 1.6327915791134107e-05, | |
| "loss": 0.6396, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.3591918184085807, | |
| "grad_norm": 0.6190239635709287, | |
| "learning_rate": 1.6191966276702235e-05, | |
| "loss": 0.6377, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3641805936642554, | |
| "grad_norm": 0.6147520812137072, | |
| "learning_rate": 1.6054134813684697e-05, | |
| "loss": 0.6375, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.36916936891993013, | |
| "grad_norm": 0.5215763519986214, | |
| "learning_rate": 1.5914463293738402e-05, | |
| "loss": 0.6368, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3741581441756049, | |
| "grad_norm": 0.6214647949635035, | |
| "learning_rate": 1.5772994167775986e-05, | |
| "loss": 0.6303, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.3791469194312796, | |
| "grad_norm": 0.6365612833821749, | |
| "learning_rate": 1.5629770433063523e-05, | |
| "loss": 0.6244, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.38413569468695435, | |
| "grad_norm": 0.594264835461608, | |
| "learning_rate": 1.5484835620152198e-05, | |
| "loss": 0.6323, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.38912446994262906, | |
| "grad_norm": 0.6107791643380025, | |
| "learning_rate": 1.533823377964791e-05, | |
| "loss": 0.6298, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.3941132451983038, | |
| "grad_norm": 0.6001005800935538, | |
| "learning_rate": 1.5190009468822782e-05, | |
| "loss": 0.63, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.3991020204539785, | |
| "grad_norm": 0.5176320349977592, | |
| "learning_rate": 1.5040207738072714e-05, | |
| "loss": 0.6296, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.4040907957096533, | |
| "grad_norm": 0.5530769815557066, | |
| "learning_rate": 1.4888874117225013e-05, | |
| "loss": 0.6202, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.409079570965328, | |
| "grad_norm": 0.4750983813132654, | |
| "learning_rate": 1.4736054601700361e-05, | |
| "loss": 0.6339, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.41406834622100275, | |
| "grad_norm": 0.5280260411948464, | |
| "learning_rate": 1.4581795638533227e-05, | |
| "loss": 0.6244, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.41905712147667745, | |
| "grad_norm": 0.49444720544930004, | |
| "learning_rate": 1.4426144112255057e-05, | |
| "loss": 0.6226, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.4240458967323522, | |
| "grad_norm": 0.5416758041687645, | |
| "learning_rate": 1.426914733064444e-05, | |
| "loss": 0.6281, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.4290346719880269, | |
| "grad_norm": 0.4913059879733367, | |
| "learning_rate": 1.4110853010348717e-05, | |
| "loss": 0.6327, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.4340234472437017, | |
| "grad_norm": 0.5475336882998988, | |
| "learning_rate": 1.3951309262381231e-05, | |
| "loss": 0.6319, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.4390122224993764, | |
| "grad_norm": 0.5887136176843448, | |
| "learning_rate": 1.3790564577498791e-05, | |
| "loss": 0.6323, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.44400099775505114, | |
| "grad_norm": 0.53325274747872, | |
| "learning_rate": 1.3628667811463654e-05, | |
| "loss": 0.6165, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.44898977301072585, | |
| "grad_norm": 0.5538441977213863, | |
| "learning_rate": 1.3465668170194633e-05, | |
| "loss": 0.6259, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.4539785482664006, | |
| "grad_norm": 0.516406002882252, | |
| "learning_rate": 1.330161519481172e-05, | |
| "loss": 0.6251, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.4589673235220753, | |
| "grad_norm": 0.5294430423934866, | |
| "learning_rate": 1.3136558746578888e-05, | |
| "loss": 0.6269, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.46395609877775007, | |
| "grad_norm": 0.5548195323966518, | |
| "learning_rate": 1.2970548991749538e-05, | |
| "loss": 0.6239, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.4689448740334248, | |
| "grad_norm": 0.48735736648704486, | |
| "learning_rate": 1.2803636386319288e-05, | |
| "loss": 0.62, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.47393364928909953, | |
| "grad_norm": 0.5093962136183301, | |
| "learning_rate": 1.2635871660690677e-05, | |
| "loss": 0.6259, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.47892242454477424, | |
| "grad_norm": 0.5220605413877938, | |
| "learning_rate": 1.2467305804254472e-05, | |
| "loss": 0.6233, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.483911199800449, | |
| "grad_norm": 0.46987756163402217, | |
| "learning_rate": 1.2297990049892274e-05, | |
| "loss": 0.6224, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.4888999750561237, | |
| "grad_norm": 0.4851271226507692, | |
| "learning_rate": 1.2127975858405096e-05, | |
| "loss": 0.6248, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.49388875031179846, | |
| "grad_norm": 0.49812035008570954, | |
| "learning_rate": 1.1957314902872686e-05, | |
| "loss": 0.6162, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.49887752556747317, | |
| "grad_norm": 0.4543789239256326, | |
| "learning_rate": 1.178605905294832e-05, | |
| "loss": 0.6191, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5038663008231479, | |
| "grad_norm": 0.49450863626445246, | |
| "learning_rate": 1.1614260359093869e-05, | |
| "loss": 0.6298, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.5088550760788226, | |
| "grad_norm": 0.46204592956818197, | |
| "learning_rate": 1.144197103675988e-05, | |
| "loss": 0.6108, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.5138438513344974, | |
| "grad_norm": 0.46446807942496315, | |
| "learning_rate": 1.1269243450515537e-05, | |
| "loss": 0.6255, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.5188326265901722, | |
| "grad_norm": 0.4523344385336006, | |
| "learning_rate": 1.1096130098133296e-05, | |
| "loss": 0.621, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.5238214018458468, | |
| "grad_norm": 0.5281185337565003, | |
| "learning_rate": 1.092268359463302e-05, | |
| "loss": 0.6181, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.5288101771015216, | |
| "grad_norm": 0.5592357254092617, | |
| "learning_rate": 1.0748956656290512e-05, | |
| "loss": 0.625, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.5337989523571963, | |
| "grad_norm": 0.49985683194766767, | |
| "learning_rate": 1.057500208461522e-05, | |
| "loss": 0.6088, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.5387877276128711, | |
| "grad_norm": 0.5348402703139689, | |
| "learning_rate": 1.0400872750302095e-05, | |
| "loss": 0.6215, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5437765028685457, | |
| "grad_norm": 0.5016453093230498, | |
| "learning_rate": 1.0226621577162377e-05, | |
| "loss": 0.6067, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.5487652781242205, | |
| "grad_norm": 0.5033968932851582, | |
| "learning_rate": 1.005230152603826e-05, | |
| "loss": 0.6056, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5537540533798953, | |
| "grad_norm": 0.4779633495557308, | |
| "learning_rate": 9.877965578706286e-06, | |
| "loss": 0.6158, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.55874282863557, | |
| "grad_norm": 0.507768651916945, | |
| "learning_rate": 9.703666721774403e-06, | |
| "loss": 0.6168, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.5637316038912447, | |
| "grad_norm": 0.4975220405187006, | |
| "learning_rate": 9.52945793057753e-06, | |
| "loss": 0.6133, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.5687203791469194, | |
| "grad_norm": 0.567581970238524, | |
| "learning_rate": 9.355392153076541e-06, | |
| "loss": 0.6153, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.5737091544025942, | |
| "grad_norm": 0.45994820207954157, | |
| "learning_rate": 9.18152229376561e-06, | |
| "loss": 0.6075, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.5786979296582689, | |
| "grad_norm": 0.4776147923803636, | |
| "learning_rate": 9.007901197592722e-06, | |
| "loss": 0.6083, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.5836867049139436, | |
| "grad_norm": 0.4458298111691603, | |
| "learning_rate": 8.834581633898307e-06, | |
| "loss": 0.6151, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.5886754801696183, | |
| "grad_norm": 0.5015975267018272, | |
| "learning_rate": 8.661616280376846e-06, | |
| "loss": 0.6083, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.5936642554252931, | |
| "grad_norm": 0.44736317209890847, | |
| "learning_rate": 8.489057707066335e-06, | |
| "loss": 0.6077, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.5986530306809679, | |
| "grad_norm": 0.4542312795928054, | |
| "learning_rate": 8.316958360370462e-06, | |
| "loss": 0.6089, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6036418059366425, | |
| "grad_norm": 0.4669024055915673, | |
| "learning_rate": 8.145370547118374e-06, | |
| "loss": 0.614, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.6086305811923173, | |
| "grad_norm": 0.460625150490194, | |
| "learning_rate": 7.974346418666854e-06, | |
| "loss": 0.6097, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.613619356447992, | |
| "grad_norm": 0.45169077334743357, | |
| "learning_rate": 7.803937955049743e-06, | |
| "loss": 0.6134, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.6186081317036668, | |
| "grad_norm": 0.5045442729073779, | |
| "learning_rate": 7.634196949179472e-06, | |
| "loss": 0.6056, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.6235969069593414, | |
| "grad_norm": 0.4976507270581154, | |
| "learning_rate": 7.465174991105405e-06, | |
| "loss": 0.6087, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.6285856822150162, | |
| "grad_norm": 0.47413589578954074, | |
| "learning_rate": 7.296923452333908e-06, | |
| "loss": 0.6073, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.633574457470691, | |
| "grad_norm": 0.4468358705510705, | |
| "learning_rate": 7.129493470214775e-06, | |
| "loss": 0.6065, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.6385632327263657, | |
| "grad_norm": 0.44813379558060285, | |
| "learning_rate": 6.962935932398862e-06, | |
| "loss": 0.5989, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.6435520079820404, | |
| "grad_norm": 0.4267618712107622, | |
| "learning_rate": 6.797301461371626e-06, | |
| "loss": 0.5981, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.6485407832377151, | |
| "grad_norm": 0.44491038507852154, | |
| "learning_rate": 6.632640399067197e-06, | |
| "loss": 0.602, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.6535295584933899, | |
| "grad_norm": 0.455548462410925, | |
| "learning_rate": 6.469002791567792e-06, | |
| "loss": 0.6077, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.6585183337490647, | |
| "grad_norm": 0.4655649758885719, | |
| "learning_rate": 6.306438373892985e-06, | |
| "loss": 0.6027, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.6635071090047393, | |
| "grad_norm": 0.4603705069002276, | |
| "learning_rate": 6.144996554883556e-06, | |
| "loss": 0.6072, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.6684958842604141, | |
| "grad_norm": 0.42809636369496656, | |
| "learning_rate": 5.98472640218449e-06, | |
| "loss": 0.5984, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.6734846595160888, | |
| "grad_norm": 0.4311787554066561, | |
| "learning_rate": 5.825676627331614e-06, | |
| "loss": 0.5997, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.6784734347717636, | |
| "grad_norm": 0.45825795981326356, | |
| "learning_rate": 5.667895570946554e-06, | |
| "loss": 0.6034, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.6834622100274382, | |
| "grad_norm": 0.4167893564717864, | |
| "learning_rate": 5.5114311880443374e-06, | |
| "loss": 0.5975, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.688450985283113, | |
| "grad_norm": 0.46394354368013807, | |
| "learning_rate": 5.356331033458276e-06, | |
| "loss": 0.6065, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.6934397605387878, | |
| "grad_norm": 0.45003494825662715, | |
| "learning_rate": 5.202642247386409e-06, | |
| "loss": 0.6052, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.6984285357944625, | |
| "grad_norm": 0.40702863232292363, | |
| "learning_rate": 5.0504115410640105e-06, | |
| "loss": 0.5985, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7034173110501372, | |
| "grad_norm": 0.42165174105900677, | |
| "learning_rate": 4.899685182566472e-06, | |
| "loss": 0.5917, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.7084060863058119, | |
| "grad_norm": 0.5050037064515478, | |
| "learning_rate": 4.7505089827468335e-06, | |
| "loss": 0.5959, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.7133948615614867, | |
| "grad_norm": 0.4558754163109762, | |
| "learning_rate": 4.602928281312351e-06, | |
| "loss": 0.5933, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.7183836368171614, | |
| "grad_norm": 0.4071864016802162, | |
| "learning_rate": 4.456987933044185e-06, | |
| "loss": 0.5992, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.7233724120728361, | |
| "grad_norm": 0.4248472014437075, | |
| "learning_rate": 4.3127322941645385e-06, | |
| "loss": 0.5937, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.7283611873285109, | |
| "grad_norm": 0.40980252115955373, | |
| "learning_rate": 4.170205208855281e-06, | |
| "loss": 0.5968, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.7333499625841856, | |
| "grad_norm": 0.4052569143394039, | |
| "learning_rate": 4.029449995932213e-06, | |
| "loss": 0.5926, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.7383387378398603, | |
| "grad_norm": 0.41360853365570666, | |
| "learning_rate": 3.890509435679026e-06, | |
| "loss": 0.6021, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.743327513095535, | |
| "grad_norm": 0.41835345554744635, | |
| "learning_rate": 3.7534257568448995e-06, | |
| "loss": 0.5952, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.7483162883512098, | |
| "grad_norm": 0.401256370852048, | |
| "learning_rate": 3.6182406238097745e-06, | |
| "loss": 0.5972, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.7533050636068845, | |
| "grad_norm": 0.4174244244132666, | |
| "learning_rate": 3.484995123921112e-06, | |
| "loss": 0.5945, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.7582938388625592, | |
| "grad_norm": 0.4081805968439954, | |
| "learning_rate": 3.353729755006081e-06, | |
| "loss": 0.5952, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.763282614118234, | |
| "grad_norm": 0.40126195224404465, | |
| "learning_rate": 3.2244844130628684e-06, | |
| "loss": 0.5869, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.7682713893739087, | |
| "grad_norm": 0.4252743360297237, | |
| "learning_rate": 3.0972983801349464e-06, | |
| "loss": 0.6057, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.7732601646295835, | |
| "grad_norm": 0.40849879013758245, | |
| "learning_rate": 2.9722103123719324e-06, | |
| "loss": 0.5987, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.7782489398852581, | |
| "grad_norm": 0.4124888296503082, | |
| "learning_rate": 2.849258228280656e-06, | |
| "loss": 0.6048, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.7832377151409329, | |
| "grad_norm": 0.408929858016356, | |
| "learning_rate": 2.728479497170066e-06, | |
| "loss": 0.591, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.7882264903966076, | |
| "grad_norm": 0.40717905524554926, | |
| "learning_rate": 2.6099108277934105e-06, | |
| "loss": 0.5957, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.7932152656522824, | |
| "grad_norm": 0.40906715914325326, | |
| "learning_rate": 2.4935882571912107e-06, | |
| "loss": 0.585, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.798204040907957, | |
| "grad_norm": 0.39973439105002123, | |
| "learning_rate": 2.379547139738392e-06, | |
| "loss": 0.5881, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.8031928161636318, | |
| "grad_norm": 0.39566194107090014, | |
| "learning_rate": 2.267822136398864e-06, | |
| "loss": 0.5948, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.8081815914193066, | |
| "grad_norm": 0.3943529028329431, | |
| "learning_rate": 2.15844720419091e-06, | |
| "loss": 0.5928, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.8131703666749813, | |
| "grad_norm": 0.396743086451483, | |
| "learning_rate": 2.0514555858664663e-06, | |
| "loss": 0.5955, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.818159141930656, | |
| "grad_norm": 0.4158031464312053, | |
| "learning_rate": 1.9468797998075494e-06, | |
| "loss": 0.5938, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.8231479171863307, | |
| "grad_norm": 0.4159354667890639, | |
| "learning_rate": 1.844751630142797e-06, | |
| "loss": 0.5811, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.8281366924420055, | |
| "grad_norm": 0.38601272105412726, | |
| "learning_rate": 1.7451021170871974e-06, | |
| "loss": 0.5933, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.8331254676976803, | |
| "grad_norm": 0.3986644463931031, | |
| "learning_rate": 1.6479615475079291e-06, | |
| "loss": 0.5892, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.8381142429533549, | |
| "grad_norm": 0.40080909281458676, | |
| "learning_rate": 1.5533594457191326e-06, | |
| "loss": 0.5898, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.8431030182090297, | |
| "grad_norm": 0.3942697489480343, | |
| "learning_rate": 1.4613245645084894e-06, | |
| "loss": 0.5863, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.8480917934647044, | |
| "grad_norm": 0.39162896724772583, | |
| "learning_rate": 1.3718848763982596e-06, | |
| "loss": 0.5963, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.8530805687203792, | |
| "grad_norm": 0.38744891578031804, | |
| "learning_rate": 1.2850675651434962e-06, | |
| "loss": 0.5931, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.8580693439760538, | |
| "grad_norm": 0.3757895616733075, | |
| "learning_rate": 1.2008990174699685e-06, | |
| "loss": 0.5958, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.8630581192317286, | |
| "grad_norm": 0.38105622554372953, | |
| "learning_rate": 1.1194048150543457e-06, | |
| "loss": 0.5928, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.8680468944874034, | |
| "grad_norm": 0.38949843229532266, | |
| "learning_rate": 1.0406097267490644e-06, | |
| "loss": 0.5894, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.8730356697430781, | |
| "grad_norm": 0.39133896634308507, | |
| "learning_rate": 9.645377010542212e-07, | |
| "loss": 0.5893, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.8780244449987528, | |
| "grad_norm": 0.4015625907458179, | |
| "learning_rate": 8.91211858838823e-07, | |
| "loss": 0.5982, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.8830132202544275, | |
| "grad_norm": 0.4047912248494648, | |
| "learning_rate": 8.206544863135612e-07, | |
| "loss": 0.5865, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.8880019955101023, | |
| "grad_norm": 0.38503511137561774, | |
| "learning_rate": 7.528870282572864e-07, | |
| "loss": 0.5831, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.892990770765777, | |
| "grad_norm": 0.387759644652555, | |
| "learning_rate": 6.879300814992007e-07, | |
| "loss": 0.5985, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.8979795460214517, | |
| "grad_norm": 0.39050325202063846, | |
| "learning_rate": 6.258033886587911e-07, | |
| "loss": 0.5881, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9029683212771265, | |
| "grad_norm": 0.3870890888051975, | |
| "learning_rate": 5.66525832145377e-07, | |
| "loss": 0.5945, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.9079570965328012, | |
| "grad_norm": 0.3856245616892951, | |
| "learning_rate": 5.101154284191035e-07, | |
| "loss": 0.5929, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.912945871788476, | |
| "grad_norm": 0.3910109970295147, | |
| "learning_rate": 4.5658932251512856e-07, | |
| "loss": 0.6021, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.9179346470441506, | |
| "grad_norm": 0.3676846272782438, | |
| "learning_rate": 4.059637828326657e-07, | |
| "loss": 0.5878, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.9229234222998254, | |
| "grad_norm": 0.3819934536854667, | |
| "learning_rate": 3.5825419619046176e-07, | |
| "loss": 0.5936, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.9279121975555001, | |
| "grad_norm": 0.3875995828243716, | |
| "learning_rate": 3.1347506315023036e-07, | |
| "loss": 0.6038, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.9329009728111749, | |
| "grad_norm": 0.38002441832217876, | |
| "learning_rate": 2.716399936094294e-07, | |
| "loss": 0.5931, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.9378897480668495, | |
| "grad_norm": 0.3848197669462124, | |
| "learning_rate": 2.327617026647533e-07, | |
| "loss": 0.595, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.9428785233225243, | |
| "grad_norm": 0.38930339170989675, | |
| "learning_rate": 1.968520067475921e-07, | |
| "loss": 0.5938, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.9478672985781991, | |
| "grad_norm": 0.3797874082101493, | |
| "learning_rate": 1.6392182003260427e-07, | |
| "loss": 0.5942, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.9528560738338738, | |
| "grad_norm": 0.3803276424025686, | |
| "learning_rate": 1.3398115112054243e-07, | |
| "loss": 0.5876, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.9578448490895485, | |
| "grad_norm": 0.3765573493263119, | |
| "learning_rate": 1.070390999962867e-07, | |
| "loss": 0.5922, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.9628336243452232, | |
| "grad_norm": 0.3849948624629748, | |
| "learning_rate": 8.31038552630603e-08, | |
| "loss": 0.5945, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.967822399600898, | |
| "grad_norm": 0.38879243314742995, | |
| "learning_rate": 6.218269165363166e-08, | |
| "loss": 0.5954, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.9728111748565728, | |
| "grad_norm": 0.38003968946404176, | |
| "learning_rate": 4.42819678192774e-08, | |
| "loss": 0.5913, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.9777999501122474, | |
| "grad_norm": 0.3869545513042188, | |
| "learning_rate": 2.9407124397169418e-08, | |
| "loss": 0.594, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.9827887253679222, | |
| "grad_norm": 0.3851047915663657, | |
| "learning_rate": 1.7562682356786488e-08, | |
| "loss": 0.5925, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.9877775006235969, | |
| "grad_norm": 0.38621262692876246, | |
| "learning_rate": 8.752241625831215e-09, | |
| "loss": 0.6004, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.9927662758792717, | |
| "grad_norm": 0.3798867531020366, | |
| "learning_rate": 2.978479996098571e-09, | |
| "loss": 0.5909, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.9977550511349463, | |
| "grad_norm": 0.38274768291794226, | |
| "learning_rate": 2.4315230959359726e-10, | |
| "loss": 0.5939, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.9997505612372163, | |
| "eval_loss": 0.5914663672447205, | |
| "eval_runtime": 139.1149, | |
| "eval_samples_per_second": 48.528, | |
| "eval_steps_per_second": 1.517, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 0.9997505612372163, | |
| "step": 1002, | |
| "total_flos": 838984280309760.0, | |
| "train_loss": 0.6422406448099666, | |
| "train_runtime": 10627.6191, | |
| "train_samples_per_second": 12.071, | |
| "train_steps_per_second": 0.094 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1002, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 838984280309760.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |