|
{ |
|
"best_metric": 0.5068374276161194, |
|
"best_model_checkpoint": "/home/datta0/models/lora_final/Meta-Llama-3-8B_metamath_default/checkpoint-611", |
|
"epoch": 0.9898744430943702, |
|
"eval_steps": 13, |
|
"global_step": 611, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.0016200891049007696, |
|
"grad_norm": 21.83199119567871, |
|
"learning_rate": 2.3076923076923076e-05, |
|
"loss": 1.2178, |
|
"step": 1 |
|
}, |
|
{ |
|
"epoch": 0.011340623734305387, |
|
"grad_norm": 4.216019153594971, |
|
"learning_rate": 0.00016153846153846153, |
|
"loss": 0.8666, |
|
"step": 7 |
|
}, |
|
{ |
|
"epoch": 0.021061158363710003, |
|
"eval_loss": 0.7509176135063171, |
|
"eval_runtime": 12.1244, |
|
"eval_samples_per_second": 41.239, |
|
"eval_steps_per_second": 5.196, |
|
"step": 13 |
|
}, |
|
{ |
|
"epoch": 0.022681247468610773, |
|
"grad_norm": 4.796965599060059, |
|
"learning_rate": 0.0002999979709808197, |
|
"loss": 0.6998, |
|
"step": 14 |
|
}, |
|
{ |
|
"epoch": 0.03402187120291616, |
|
"grad_norm": 5.564920425415039, |
|
"learning_rate": 0.0002998701612152596, |
|
"loss": 0.6952, |
|
"step": 21 |
|
}, |
|
{ |
|
"epoch": 0.042122316727420006, |
|
"eval_loss": 0.736819326877594, |
|
"eval_runtime": 102.3178, |
|
"eval_samples_per_second": 4.887, |
|
"eval_steps_per_second": 0.616, |
|
"step": 26 |
|
}, |
|
{ |
|
"epoch": 0.04536249493722155, |
|
"grad_norm": 7.243711948394775, |
|
"learning_rate": 0.0002995437011859465, |
|
"loss": 0.7117, |
|
"step": 28 |
|
}, |
|
{ |
|
"epoch": 0.056703118671526935, |
|
"grad_norm": 5.288980960845947, |
|
"learning_rate": 0.00029901902360990936, |
|
"loss": 0.7079, |
|
"step": 35 |
|
}, |
|
{ |
|
"epoch": 0.06318347509113001, |
|
"eval_loss": 0.7195977568626404, |
|
"eval_runtime": 102.5692, |
|
"eval_samples_per_second": 4.875, |
|
"eval_steps_per_second": 0.614, |
|
"step": 39 |
|
}, |
|
{ |
|
"epoch": 0.06804374240583232, |
|
"grad_norm": 5.120002746582031, |
|
"learning_rate": 0.00029829682393805085, |
|
"loss": 0.6827, |
|
"step": 42 |
|
}, |
|
{ |
|
"epoch": 0.0793843661401377, |
|
"grad_norm": 8.668340682983398, |
|
"learning_rate": 0.0002973780594333385, |
|
"loss": 0.6922, |
|
"step": 49 |
|
}, |
|
{ |
|
"epoch": 0.08424463345484001, |
|
"eval_loss": 0.7065702080726624, |
|
"eval_runtime": 12.1388, |
|
"eval_samples_per_second": 41.19, |
|
"eval_steps_per_second": 5.19, |
|
"step": 52 |
|
}, |
|
{ |
|
"epoch": 0.0907249898744431, |
|
"grad_norm": 4.637055397033691, |
|
"learning_rate": 0.00029626394790197025, |
|
"loss": 0.6824, |
|
"step": 56 |
|
}, |
|
{ |
|
"epoch": 0.10206561360874848, |
|
"grad_norm": 5.630373477935791, |
|
"learning_rate": 0.00029495596607919305, |
|
"loss": 0.6565, |
|
"step": 63 |
|
}, |
|
{ |
|
"epoch": 0.10530579181855002, |
|
"eval_loss": 0.7073802351951599, |
|
"eval_runtime": 103.2813, |
|
"eval_samples_per_second": 4.841, |
|
"eval_steps_per_second": 0.61, |
|
"step": 65 |
|
}, |
|
{ |
|
"epoch": 0.11340623734305387, |
|
"grad_norm": 6.744913578033447, |
|
"learning_rate": 0.00029345584767191685, |
|
"loss": 0.6785, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 0.12474686107735926, |
|
"grad_norm": 5.90676736831665, |
|
"learning_rate": 0.0002917655810607161, |
|
"loss": 0.6791, |
|
"step": 77 |
|
}, |
|
{ |
|
"epoch": 0.12636695018226002, |
|
"eval_loss": 0.7263020873069763, |
|
"eval_runtime": 93.1665, |
|
"eval_samples_per_second": 5.367, |
|
"eval_steps_per_second": 0.676, |
|
"step": 78 |
|
}, |
|
{ |
|
"epoch": 0.13608748481166463, |
|
"grad_norm": 7.519783020019531, |
|
"learning_rate": 0.0002898874066642667, |
|
"loss": 0.6764, |
|
"step": 84 |
|
}, |
|
{ |
|
"epoch": 0.14742810854597002, |
|
"grad_norm": 7.202000617980957, |
|
"learning_rate": 0.00028782381396971003, |
|
"loss": 0.6858, |
|
"step": 91 |
|
}, |
|
{ |
|
"epoch": 0.14742810854597002, |
|
"eval_loss": 0.7018572092056274, |
|
"eval_runtime": 12.1782, |
|
"eval_samples_per_second": 41.057, |
|
"eval_steps_per_second": 5.173, |
|
"step": 91 |
|
}, |
|
{ |
|
"epoch": 0.1587687322802754, |
|
"grad_norm": 5.747766971588135, |
|
"learning_rate": 0.00028557753823288173, |
|
"loss": 0.6693, |
|
"step": 98 |
|
}, |
|
{ |
|
"epoch": 0.16848926690968002, |
|
"eval_loss": 0.6925504207611084, |
|
"eval_runtime": 108.0515, |
|
"eval_samples_per_second": 4.627, |
|
"eval_steps_per_second": 0.583, |
|
"step": 104 |
|
}, |
|
{ |
|
"epoch": 0.1701093560145808, |
|
"grad_norm": 5.340470790863037, |
|
"learning_rate": 0.0002831515568527781, |
|
"loss": 0.6859, |
|
"step": 105 |
|
}, |
|
{ |
|
"epoch": 0.1814499797488862, |
|
"grad_norm": 6.736198902130127, |
|
"learning_rate": 0.00028054908542506627, |
|
"loss": 0.6503, |
|
"step": 112 |
|
}, |
|
{ |
|
"epoch": 0.18955042527339003, |
|
"eval_loss": 0.6922364234924316, |
|
"eval_runtime": 96.4494, |
|
"eval_samples_per_second": 5.184, |
|
"eval_steps_per_second": 0.653, |
|
"step": 117 |
|
}, |
|
{ |
|
"epoch": 0.19279060348319157, |
|
"grad_norm": 5.258207321166992, |
|
"learning_rate": 0.00027777357347986823, |
|
"loss": 0.6626, |
|
"step": 119 |
|
}, |
|
{ |
|
"epoch": 0.20413122721749696, |
|
"grad_norm": 7.422248840332031, |
|
"learning_rate": 0.00027482869990946986, |
|
"loss": 0.6488, |
|
"step": 126 |
|
}, |
|
{ |
|
"epoch": 0.21061158363710003, |
|
"eval_loss": 0.692547619342804, |
|
"eval_runtime": 12.1559, |
|
"eval_samples_per_second": 41.132, |
|
"eval_steps_per_second": 5.183, |
|
"step": 130 |
|
}, |
|
{ |
|
"epoch": 0.21547185095180235, |
|
"grad_norm": 4.835718631744385, |
|
"learning_rate": 0.0002717183680920135, |
|
"loss": 0.641, |
|
"step": 133 |
|
}, |
|
{ |
|
"epoch": 0.22681247468610774, |
|
"grad_norm": 6.634873390197754, |
|
"learning_rate": 0.00026844670071763906, |
|
"loss": 0.6505, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 0.23167274200081003, |
|
"eval_loss": 0.6843672394752502, |
|
"eval_runtime": 110.1028, |
|
"eval_samples_per_second": 4.541, |
|
"eval_steps_per_second": 0.572, |
|
"step": 143 |
|
}, |
|
{ |
|
"epoch": 0.23815309842041313, |
|
"grad_norm": 5.616636276245117, |
|
"learning_rate": 0.00026501803432393037, |
|
"loss": 0.6686, |
|
"step": 147 |
|
}, |
|
{ |
|
"epoch": 0.24949372215471852, |
|
"grad_norm": 4.444299221038818, |
|
"learning_rate": 0.00026143691354791145, |
|
"loss": 0.6533, |
|
"step": 154 |
|
}, |
|
{ |
|
"epoch": 0.25273390036452004, |
|
"eval_loss": 0.6842171549797058, |
|
"eval_runtime": 104.1575, |
|
"eval_samples_per_second": 4.8, |
|
"eval_steps_per_second": 0.605, |
|
"step": 156 |
|
}, |
|
{ |
|
"epoch": 0.2608343458890239, |
|
"grad_norm": 4.952290058135986, |
|
"learning_rate": 0.00025770808510220956, |
|
"loss": 0.6518, |
|
"step": 161 |
|
}, |
|
{ |
|
"epoch": 0.27217496962332927, |
|
"grad_norm": 5.576895236968994, |
|
"learning_rate": 0.00025383649148337105, |
|
"loss": 0.6505, |
|
"step": 168 |
|
}, |
|
{ |
|
"epoch": 0.27379505872823007, |
|
"eval_loss": 0.6708831787109375, |
|
"eval_runtime": 12.1814, |
|
"eval_samples_per_second": 41.046, |
|
"eval_steps_per_second": 5.172, |
|
"step": 169 |
|
}, |
|
{ |
|
"epoch": 0.28351559335763465, |
|
"grad_norm": 4.597885608673096, |
|
"learning_rate": 0.0002498272644206695, |
|
"loss": 0.6479, |
|
"step": 175 |
|
}, |
|
{ |
|
"epoch": 0.29485621709194004, |
|
"grad_norm": 4.833500862121582, |
|
"learning_rate": 0.0002456857180740884, |
|
"loss": 0.6456, |
|
"step": 182 |
|
}, |
|
{ |
|
"epoch": 0.29485621709194004, |
|
"eval_loss": 0.6661036014556885, |
|
"eval_runtime": 108.3108, |
|
"eval_samples_per_second": 4.616, |
|
"eval_steps_per_second": 0.582, |
|
"step": 182 |
|
}, |
|
{ |
|
"epoch": 0.30619684082624543, |
|
"grad_norm": 5.517384052276611, |
|
"learning_rate": 0.0002414173419904956, |
|
"loss": 0.6307, |
|
"step": 189 |
|
}, |
|
{ |
|
"epoch": 0.3159173754556501, |
|
"eval_loss": 0.6699193120002747, |
|
"eval_runtime": 105.864, |
|
"eval_samples_per_second": 4.723, |
|
"eval_steps_per_second": 0.595, |
|
"step": 195 |
|
}, |
|
{ |
|
"epoch": 0.3175374645605508, |
|
"grad_norm": 5.475388526916504, |
|
"learning_rate": 0.00023702779382734566, |
|
"loss": 0.6403, |
|
"step": 196 |
|
}, |
|
{ |
|
"epoch": 0.3288780882948562, |
|
"grad_norm": 6.005538463592529, |
|
"learning_rate": 0.0002325228918535541, |
|
"loss": 0.6144, |
|
"step": 203 |
|
}, |
|
{ |
|
"epoch": 0.33697853381936005, |
|
"eval_loss": 0.6628842949867249, |
|
"eval_runtime": 12.1654, |
|
"eval_samples_per_second": 41.1, |
|
"eval_steps_per_second": 5.179, |
|
"step": 208 |
|
}, |
|
{ |
|
"epoch": 0.3402187120291616, |
|
"grad_norm": 5.392065048217773, |
|
"learning_rate": 0.00022790860723748442, |
|
"loss": 0.6437, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 0.351559335763467, |
|
"grad_norm": 4.369307041168213, |
|
"learning_rate": 0.00022319105613226921, |
|
"loss": 0.6286, |
|
"step": 217 |
|
}, |
|
{ |
|
"epoch": 0.3580396921830701, |
|
"eval_loss": 0.6546506285667419, |
|
"eval_runtime": 106.5209, |
|
"eval_samples_per_second": 4.694, |
|
"eval_steps_per_second": 0.591, |
|
"step": 221 |
|
}, |
|
{ |
|
"epoch": 0.3628999594977724, |
|
"grad_norm": 5.212913990020752, |
|
"learning_rate": 0.00021837649156895706, |
|
"loss": 0.6325, |
|
"step": 224 |
|
}, |
|
{ |
|
"epoch": 0.37424058323207776, |
|
"grad_norm": 4.210490703582764, |
|
"learning_rate": 0.00021347129516822945, |
|
"loss": 0.6261, |
|
"step": 231 |
|
}, |
|
{ |
|
"epoch": 0.37910085054678005, |
|
"eval_loss": 0.6468992829322815, |
|
"eval_runtime": 108.4304, |
|
"eval_samples_per_second": 4.611, |
|
"eval_steps_per_second": 0.581, |
|
"step": 234 |
|
}, |
|
{ |
|
"epoch": 0.38558120696638315, |
|
"grad_norm": 4.121553421020508, |
|
"learning_rate": 0.00020848196868167505, |
|
"loss": 0.6326, |
|
"step": 238 |
|
}, |
|
{ |
|
"epoch": 0.39692183070068854, |
|
"grad_norm": 4.0263142585754395, |
|
"learning_rate": 0.000203415125373832, |
|
"loss": 0.6365, |
|
"step": 245 |
|
}, |
|
{ |
|
"epoch": 0.4001620089104901, |
|
"eval_loss": 0.6482229232788086, |
|
"eval_runtime": 12.1869, |
|
"eval_samples_per_second": 41.028, |
|
"eval_steps_per_second": 5.169, |
|
"step": 247 |
|
}, |
|
{ |
|
"epoch": 0.4082624544349939, |
|
"grad_norm": 4.285489559173584, |
|
"learning_rate": 0.00019827748125642242, |
|
"loss": 0.6276, |
|
"step": 252 |
|
}, |
|
{ |
|
"epoch": 0.4196030781692993, |
|
"grad_norm": 4.923107147216797, |
|
"learning_rate": 0.0001930758461863965, |
|
"loss": 0.6108, |
|
"step": 259 |
|
}, |
|
{ |
|
"epoch": 0.42122316727420006, |
|
"eval_loss": 0.6427775621414185, |
|
"eval_runtime": 102.1449, |
|
"eval_samples_per_second": 4.895, |
|
"eval_steps_per_second": 0.617, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 0.4309437019036047, |
|
"grad_norm": 5.982186317443848, |
|
"learning_rate": 0.0001878171148395872, |
|
"loss": 0.6149, |
|
"step": 266 |
|
}, |
|
{ |
|
"epoch": 0.4422843256379101, |
|
"grad_norm": 3.5820586681365967, |
|
"learning_rate": 0.00018250825757193848, |
|
"loss": 0.6207, |
|
"step": 273 |
|
}, |
|
{ |
|
"epoch": 0.4422843256379101, |
|
"eval_loss": 0.6321972608566284, |
|
"eval_runtime": 102.3465, |
|
"eval_samples_per_second": 4.885, |
|
"eval_steps_per_second": 0.616, |
|
"step": 273 |
|
}, |
|
{ |
|
"epoch": 0.4536249493722155, |
|
"grad_norm": 4.147037029266357, |
|
"learning_rate": 0.0001771563111804211, |
|
"loss": 0.6219, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 0.46334548400162007, |
|
"eval_loss": 0.6264978051185608, |
|
"eval_runtime": 12.203, |
|
"eval_samples_per_second": 40.974, |
|
"eval_steps_per_second": 5.163, |
|
"step": 286 |
|
}, |
|
{ |
|
"epoch": 0.46496557310652087, |
|
"grad_norm": 3.318942070007324, |
|
"learning_rate": 0.0001717683695758819, |
|
"loss": 0.602, |
|
"step": 287 |
|
}, |
|
{ |
|
"epoch": 0.47630619684082626, |
|
"grad_norm": 3.9704108238220215, |
|
"learning_rate": 0.00016635157438018983, |
|
"loss": 0.6133, |
|
"step": 294 |
|
}, |
|
{ |
|
"epoch": 0.4844066423653301, |
|
"eval_loss": 0.6212669610977173, |
|
"eval_runtime": 96.1686, |
|
"eval_samples_per_second": 5.199, |
|
"eval_steps_per_second": 0.655, |
|
"step": 299 |
|
}, |
|
{ |
|
"epoch": 0.48764682057513165, |
|
"grad_norm": 4.209397792816162, |
|
"learning_rate": 0.0001609131054601416, |
|
"loss": 0.6142, |
|
"step": 301 |
|
}, |
|
{ |
|
"epoch": 0.49898744430943703, |
|
"grad_norm": 3.506843328475952, |
|
"learning_rate": 0.00015546017141067432, |
|
"loss": 0.5944, |
|
"step": 308 |
|
}, |
|
{ |
|
"epoch": 0.5054678007290401, |
|
"eval_loss": 0.6138330101966858, |
|
"eval_runtime": 96.8265, |
|
"eval_samples_per_second": 5.164, |
|
"eval_steps_per_second": 0.651, |
|
"step": 312 |
|
}, |
|
{ |
|
"epoch": 0.5103280680437424, |
|
"grad_norm": 3.397951364517212, |
|
"learning_rate": 0.00015, |
|
"loss": 0.6082, |
|
"step": 315 |
|
}, |
|
{ |
|
"epoch": 0.5216686917780478, |
|
"grad_norm": 3.766294002532959, |
|
"learning_rate": 0.0001445398285893257, |
|
"loss": 0.5871, |
|
"step": 322 |
|
}, |
|
{ |
|
"epoch": 0.5265289590927501, |
|
"eval_loss": 0.6034436225891113, |
|
"eval_runtime": 12.2193, |
|
"eval_samples_per_second": 40.919, |
|
"eval_steps_per_second": 5.156, |
|
"step": 325 |
|
}, |
|
{ |
|
"epoch": 0.5330093155123532, |
|
"grad_norm": 3.9341065883636475, |
|
"learning_rate": 0.0001390868945398584, |
|
"loss": 0.588, |
|
"step": 329 |
|
}, |
|
{ |
|
"epoch": 0.5443499392466585, |
|
"grad_norm": 3.7946388721466064, |
|
"learning_rate": 0.00013364842561981014, |
|
"loss": 0.5827, |
|
"step": 336 |
|
}, |
|
{ |
|
"epoch": 0.5475901174564601, |
|
"eval_loss": 0.6012772917747498, |
|
"eval_runtime": 100.8024, |
|
"eval_samples_per_second": 4.96, |
|
"eval_steps_per_second": 0.625, |
|
"step": 338 |
|
}, |
|
{ |
|
"epoch": 0.555690562980964, |
|
"grad_norm": 2.895064353942871, |
|
"learning_rate": 0.00012823163042411807, |
|
"loss": 0.5728, |
|
"step": 343 |
|
}, |
|
{ |
|
"epoch": 0.5670311867152693, |
|
"grad_norm": 3.9632740020751953, |
|
"learning_rate": 0.0001228436888195789, |
|
"loss": 0.5714, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 0.56865127582017, |
|
"eval_loss": 0.5922607183456421, |
|
"eval_runtime": 95.7391, |
|
"eval_samples_per_second": 5.223, |
|
"eval_steps_per_second": 0.658, |
|
"step": 351 |
|
}, |
|
{ |
|
"epoch": 0.5783718104495748, |
|
"grad_norm": 3.257824420928955, |
|
"learning_rate": 0.00011749174242806152, |
|
"loss": 0.5824, |
|
"step": 357 |
|
}, |
|
{ |
|
"epoch": 0.5897124341838801, |
|
"grad_norm": 4.506229877471924, |
|
"learning_rate": 0.00011218288516041279, |
|
"loss": 0.5512, |
|
"step": 364 |
|
}, |
|
{ |
|
"epoch": 0.5897124341838801, |
|
"eval_loss": 0.5848901867866516, |
|
"eval_runtime": 12.2285, |
|
"eval_samples_per_second": 40.888, |
|
"eval_steps_per_second": 5.152, |
|
"step": 364 |
|
}, |
|
{ |
|
"epoch": 0.6010530579181855, |
|
"grad_norm": 3.2441561222076416, |
|
"learning_rate": 0.00010692415381360349, |
|
"loss": 0.5636, |
|
"step": 371 |
|
}, |
|
{ |
|
"epoch": 0.6107735925475901, |
|
"eval_loss": 0.5755201578140259, |
|
"eval_runtime": 106.0764, |
|
"eval_samples_per_second": 4.714, |
|
"eval_steps_per_second": 0.594, |
|
"step": 377 |
|
}, |
|
{ |
|
"epoch": 0.6123936816524909, |
|
"grad_norm": 3.45365834236145, |
|
"learning_rate": 0.00010172251874357757, |
|
"loss": 0.5535, |
|
"step": 378 |
|
}, |
|
{ |
|
"epoch": 0.6237343053867963, |
|
"grad_norm": 3.905564785003662, |
|
"learning_rate": 9.658487462616794e-05, |
|
"loss": 0.5564, |
|
"step": 385 |
|
}, |
|
{ |
|
"epoch": 0.6318347509113001, |
|
"eval_loss": 0.5684311389923096, |
|
"eval_runtime": 96.7227, |
|
"eval_samples_per_second": 5.169, |
|
"eval_steps_per_second": 0.651, |
|
"step": 390 |
|
}, |
|
{ |
|
"epoch": 0.6350749291211016, |
|
"grad_norm": 2.8492631912231445, |
|
"learning_rate": 9.151803131832493e-05, |
|
"loss": 0.5624, |
|
"step": 392 |
|
}, |
|
{ |
|
"epoch": 0.6464155528554071, |
|
"grad_norm": 3.4206936359405518, |
|
"learning_rate": 8.652870483177049e-05, |
|
"loss": 0.5444, |
|
"step": 399 |
|
}, |
|
{ |
|
"epoch": 0.6528959092750102, |
|
"eval_loss": 0.5647120475769043, |
|
"eval_runtime": 12.2178, |
|
"eval_samples_per_second": 40.924, |
|
"eval_steps_per_second": 5.156, |
|
"step": 403 |
|
}, |
|
{ |
|
"epoch": 0.6577561765897124, |
|
"grad_norm": 3.9881739616394043, |
|
"learning_rate": 8.162350843104291e-05, |
|
"loss": 0.5499, |
|
"step": 406 |
|
}, |
|
{ |
|
"epoch": 0.6690968003240179, |
|
"grad_norm": 3.396562337875366, |
|
"learning_rate": 7.680894386773072e-05, |
|
"loss": 0.5431, |
|
"step": 413 |
|
}, |
|
{ |
|
"epoch": 0.6739570676387201, |
|
"eval_loss": 0.5582295060157776, |
|
"eval_runtime": 108.0987, |
|
"eval_samples_per_second": 4.625, |
|
"eval_steps_per_second": 0.583, |
|
"step": 416 |
|
}, |
|
{ |
|
"epoch": 0.6804374240583232, |
|
"grad_norm": 2.720857858657837, |
|
"learning_rate": 7.209139276251558e-05, |
|
"loss": 0.5387, |
|
"step": 420 |
|
}, |
|
{ |
|
"epoch": 0.6917780477926286, |
|
"grad_norm": 2.7699272632598877, |
|
"learning_rate": 6.747710814644589e-05, |
|
"loss": 0.5311, |
|
"step": 427 |
|
}, |
|
{ |
|
"epoch": 0.6950182260024301, |
|
"eval_loss": 0.5533111691474915, |
|
"eval_runtime": 97.6016, |
|
"eval_samples_per_second": 5.123, |
|
"eval_steps_per_second": 0.645, |
|
"step": 429 |
|
}, |
|
{ |
|
"epoch": 0.703118671526934, |
|
"grad_norm": 3.2543938159942627, |
|
"learning_rate": 6.297220617265435e-05, |
|
"loss": 0.5219, |
|
"step": 434 |
|
}, |
|
{ |
|
"epoch": 0.7144592952612394, |
|
"grad_norm": 3.016479015350342, |
|
"learning_rate": 5.858265800950438e-05, |
|
"loss": 0.5323, |
|
"step": 441 |
|
}, |
|
{ |
|
"epoch": 0.7160793843661402, |
|
"eval_loss": 0.545806884765625, |
|
"eval_runtime": 12.2002, |
|
"eval_samples_per_second": 40.983, |
|
"eval_steps_per_second": 5.164, |
|
"step": 442 |
|
}, |
|
{ |
|
"epoch": 0.7257999189955447, |
|
"grad_norm": 3.142430067062378, |
|
"learning_rate": 5.4314281925911634e-05, |
|
"loss": 0.5278, |
|
"step": 448 |
|
}, |
|
{ |
|
"epoch": 0.7371405427298502, |
|
"grad_norm": 2.5312345027923584, |
|
"learning_rate": 5.0172735579330526e-05, |
|
"loss": 0.5172, |
|
"step": 455 |
|
}, |
|
{ |
|
"epoch": 0.7371405427298502, |
|
"eval_loss": 0.53862464427948, |
|
"eval_runtime": 109.318, |
|
"eval_samples_per_second": 4.574, |
|
"eval_steps_per_second": 0.576, |
|
"step": 455 |
|
}, |
|
{ |
|
"epoch": 0.7484811664641555, |
|
"grad_norm": 3.1494243144989014, |
|
"learning_rate": 4.616350851662895e-05, |
|
"loss": 0.5113, |
|
"step": 462 |
|
}, |
|
{ |
|
"epoch": 0.7582017010935601, |
|
"eval_loss": 0.5340880751609802, |
|
"eval_runtime": 98.7219, |
|
"eval_samples_per_second": 5.065, |
|
"eval_steps_per_second": 0.638, |
|
"step": 468 |
|
}, |
|
{ |
|
"epoch": 0.759821790198461, |
|
"grad_norm": 3.0264575481414795, |
|
"learning_rate": 4.229191489779047e-05, |
|
"loss": 0.535, |
|
"step": 469 |
|
}, |
|
{ |
|
"epoch": 0.7711624139327663, |
|
"grad_norm": 3.485691547393799, |
|
"learning_rate": 3.8563086452088506e-05, |
|
"loss": 0.4989, |
|
"step": 476 |
|
}, |
|
{ |
|
"epoch": 0.7792628594572701, |
|
"eval_loss": 0.5295507907867432, |
|
"eval_runtime": 12.2147, |
|
"eval_samples_per_second": 40.934, |
|
"eval_steps_per_second": 5.158, |
|
"step": 481 |
|
}, |
|
{ |
|
"epoch": 0.7825030376670717, |
|
"grad_norm": 2.9964497089385986, |
|
"learning_rate": 3.498196567606959e-05, |
|
"loss": 0.5217, |
|
"step": 483 |
|
}, |
|
{ |
|
"epoch": 0.7938436614013771, |
|
"grad_norm": 2.854008913040161, |
|
"learning_rate": 3.1553299282360966e-05, |
|
"loss": 0.4929, |
|
"step": 490 |
|
}, |
|
{ |
|
"epoch": 0.8003240178209802, |
|
"eval_loss": 0.5264267325401306, |
|
"eval_runtime": 107.6368, |
|
"eval_samples_per_second": 4.645, |
|
"eval_steps_per_second": 0.585, |
|
"step": 494 |
|
}, |
|
{ |
|
"epoch": 0.8051842851356824, |
|
"grad_norm": 2.8643412590026855, |
|
"learning_rate": 2.828163190798644e-05, |
|
"loss": 0.508, |
|
"step": 497 |
|
}, |
|
{ |
|
"epoch": 0.8165249088699879, |
|
"grad_norm": 2.737133502960205, |
|
"learning_rate": 2.5171300090530106e-05, |
|
"loss": 0.5266, |
|
"step": 504 |
|
}, |
|
{ |
|
"epoch": 0.8213851761846902, |
|
"eval_loss": 0.5213577747344971, |
|
"eval_runtime": 107.3148, |
|
"eval_samples_per_second": 4.659, |
|
"eval_steps_per_second": 0.587, |
|
"step": 507 |
|
}, |
|
{ |
|
"epoch": 0.8278655326042932, |
|
"grad_norm": 2.559361457824707, |
|
"learning_rate": 2.2226426520131734e-05, |
|
"loss": 0.5014, |
|
"step": 511 |
|
}, |
|
{ |
|
"epoch": 0.8392061563385986, |
|
"grad_norm": 3.189368486404419, |
|
"learning_rate": 1.9450914574933725e-05, |
|
"loss": 0.5075, |
|
"step": 518 |
|
}, |
|
{ |
|
"epoch": 0.8424463345484001, |
|
"eval_loss": 0.518385112285614, |
|
"eval_runtime": 12.2167, |
|
"eval_samples_per_second": 40.928, |
|
"eval_steps_per_second": 5.157, |
|
"step": 520 |
|
}, |
|
{ |
|
"epoch": 0.850546780072904, |
|
"grad_norm": 2.4905502796173096, |
|
"learning_rate": 1.6848443147221828e-05, |
|
"loss": 0.4834, |
|
"step": 525 |
|
}, |
|
{ |
|
"epoch": 0.8618874038072094, |
|
"grad_norm": 2.5532329082489014, |
|
"learning_rate": 1.4422461767118233e-05, |
|
"loss": 0.4917, |
|
"step": 532 |
|
}, |
|
{ |
|
"epoch": 0.8635074929121102, |
|
"eval_loss": 0.515020489692688, |
|
"eval_runtime": 102.4557, |
|
"eval_samples_per_second": 4.88, |
|
"eval_steps_per_second": 0.615, |
|
"step": 533 |
|
}, |
|
{ |
|
"epoch": 0.8732280275415147, |
|
"grad_norm": 2.1145200729370117, |
|
"learning_rate": 1.2176186030289936e-05, |
|
"loss": 0.4893, |
|
"step": 539 |
|
}, |
|
{ |
|
"epoch": 0.8845686512758202, |
|
"grad_norm": 3.142739772796631, |
|
"learning_rate": 1.011259333573326e-05, |
|
"loss": 0.5078, |
|
"step": 546 |
|
}, |
|
{ |
|
"epoch": 0.8845686512758202, |
|
"eval_loss": 0.5123878121376038, |
|
"eval_runtime": 107.0251, |
|
"eval_samples_per_second": 4.672, |
|
"eval_steps_per_second": 0.589, |
|
"step": 546 |
|
}, |
|
{ |
|
"epoch": 0.8959092750101255, |
|
"grad_norm": 2.60658860206604, |
|
"learning_rate": 8.234418939283866e-06, |
|
"loss": 0.4897, |
|
"step": 553 |
|
}, |
|
{ |
|
"epoch": 0.9056298096395302, |
|
"eval_loss": 0.509852945804596, |
|
"eval_runtime": 12.2385, |
|
"eval_samples_per_second": 40.855, |
|
"eval_steps_per_second": 5.148, |
|
"step": 559 |
|
}, |
|
{ |
|
"epoch": 0.907249898744431, |
|
"grad_norm": 3.024414300918579, |
|
"learning_rate": 6.544152328083152e-06, |
|
"loss": 0.5, |
|
"step": 560 |
|
}, |
|
{ |
|
"epoch": 0.9185905224787363, |
|
"grad_norm": 3.256274700164795, |
|
"learning_rate": 5.044033920806933e-06, |
|
"loss": 0.4879, |
|
"step": 567 |
|
}, |
|
{ |
|
"epoch": 0.9266909680032401, |
|
"eval_loss": 0.5081230998039246, |
|
"eval_runtime": 101.7485, |
|
"eval_samples_per_second": 4.914, |
|
"eval_steps_per_second": 0.619, |
|
"step": 572 |
|
}, |
|
{ |
|
"epoch": 0.9299311462130417, |
|
"grad_norm": 3.299356698989868, |
|
"learning_rate": 3.7360520980297514e-06, |
|
"loss": 0.4923, |
|
"step": 574 |
|
}, |
|
{ |
|
"epoch": 0.9412717699473471, |
|
"grad_norm": 2.953251361846924, |
|
"learning_rate": 2.6219405666614402e-06, |
|
"loss": 0.5007, |
|
"step": 581 |
|
}, |
|
{ |
|
"epoch": 0.9477521263669502, |
|
"eval_loss": 0.5073325634002686, |
|
"eval_runtime": 101.2425, |
|
"eval_samples_per_second": 4.939, |
|
"eval_steps_per_second": 0.622, |
|
"step": 585 |
|
}, |
|
{ |
|
"epoch": 0.9526123936816525, |
|
"grad_norm": 2.408773183822632, |
|
"learning_rate": 1.7031760619491353e-06, |
|
"loss": 0.4757, |
|
"step": 588 |
|
}, |
|
{ |
|
"epoch": 0.9639530174159578, |
|
"grad_norm": 2.3097264766693115, |
|
"learning_rate": 9.809763900905875e-07, |
|
"loss": 0.4979, |
|
"step": 595 |
|
}, |
|
{ |
|
"epoch": 0.9688132847306602, |
|
"eval_loss": 0.5070714950561523, |
|
"eval_runtime": 12.2216, |
|
"eval_samples_per_second": 40.911, |
|
"eval_steps_per_second": 5.155, |
|
"step": 598 |
|
}, |
|
{ |
|
"epoch": 0.9752936411502633, |
|
"grad_norm": 2.2127528190612793, |
|
"learning_rate": 4.562988140535073e-07, |
|
"loss": 0.4892, |
|
"step": 602 |
|
}, |
|
{ |
|
"epoch": 0.9866342648845686, |
|
"grad_norm": 2.4536213874816895, |
|
"learning_rate": 1.298387847403437e-07, |
|
"loss": 0.4991, |
|
"step": 609 |
|
}, |
|
{ |
|
"epoch": 0.9898744430943702, |
|
"eval_loss": 0.5068374276161194, |
|
"eval_runtime": 97.1562, |
|
"eval_samples_per_second": 5.146, |
|
"eval_steps_per_second": 0.648, |
|
"step": 611 |
|
} |
|
], |
|
"logging_steps": 7, |
|
"max_steps": 617, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 1, |
|
"save_steps": 13, |
|
"stateful_callbacks": { |
|
"TrainerControl": { |
|
"args": { |
|
"should_epoch_stop": false, |
|
"should_evaluate": false, |
|
"should_log": false, |
|
"should_save": true, |
|
"should_training_stop": false |
|
}, |
|
"attributes": {} |
|
} |
|
}, |
|
"total_flos": 7.394413366880502e+17, |
|
"train_batch_size": 8, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|