LEIA/LEIA-multilingual
08
1{2 "best_metric": 0.7091297032684482,3 "best_model_checkpoint": "transformers/checkpoint/vent/linear_probe_ft_cont/vent_bernice_lr1e-5_embdrop0.1_labelsmoothing0.1_42_jointsupcl/checkpoint-11500",4 "epoch": 0.46387802024928404,5 "global_step": 11500,6 "is_hyper_param_search": false,7 "is_local_process_zero": true,8 "is_world_process_zero": true,9 "log_history": [10 {11 "epoch": 0.02,12 "learning_rate": 0.0005,13 "loss": 1.224,14 "step": 50015 },16 {17 "epoch": 0.04,18 "learning_rate": 0.0005,19 "loss": 1.1917,20 "step": 100021 },22 {23 "epoch": 0.04,24 "step": 1000,25 "total_flos": 1.6839561117696e+16,26 "train_loss": 1.2078435668945313,27 "train_runtime": 247.7258,28 "train_samples_per_second": 1033.4,29 "train_steps_per_second": 4.03730 },31 {32 "epoch": 0.06,33 "learning_rate": 1e-05,34 "loss": 4.8868,35 "step": 150036 },37 {38 "epoch": 0.06,39 "eval_accuracy": 0.6693328022956848,40 "eval_loss": 4.761493682861328,41 "eval_macroPrec": 0.673714797769901,42 "eval_macroRec": 0.6665857187883608,43 "eval_macrof1": 0.6679247692492374,44 "eval_runtime": 599.7459,45 "eval_samples_per_second": 1175.733,46 "eval_steps_per_second": 4.594,47 "step": 150048 },49 {50 "epoch": 0.08,51 "learning_rate": 1e-05,52 "loss": 4.7865,53 "step": 200054 },55 {56 "epoch": 0.08,57 "eval_accuracy": 0.6801448464393616,58 "eval_loss": 4.736236572265625,59 "eval_macroPrec": 0.6789197866299151,60 "eval_macroRec": 0.6820194214808827,61 "eval_macrof1": 0.6792291630889113,62 "eval_runtime": 578.3287,63 "eval_samples_per_second": 1219.274,64 "eval_steps_per_second": 4.764,65 "step": 200066 },67 {68 "epoch": 0.1,69 "learning_rate": 1e-05,70 "loss": 4.7608,71 "step": 250072 },73 {74 "epoch": 0.1,75 "eval_accuracy": 0.6877347826957703,76 "eval_loss": 4.717987537384033,77 "eval_macroPrec": 0.6935462320329414,78 "eval_macroRec": 0.684167600736026,79 "eval_macrof1": 0.6858003553695454,80 "eval_runtime": 577.8607,81 "eval_samples_per_second": 1220.261,82 "eval_steps_per_second": 4.768,83 "step": 250084 },85 {86 "epoch": 0.12,87 "learning_rate": 1e-05,88 "loss": 4.7529,89 "step": 300090 },91 {92 "epoch": 0.12,93 "eval_accuracy": 0.6885204315185547,94 "eval_loss": 4.716394901275635,95 "eval_macroPrec": 0.6948738613915024,96 "eval_macroRec": 0.6856825977848897,97 "eval_macrof1": 0.6876887565361411,98 "eval_runtime": 577.0236,99 "eval_samples_per_second": 1222.032,100 "eval_steps_per_second": 4.775,101 "step": 3000102 },103 {104 "epoch": 0.14,105 "learning_rate": 1e-05,106 "loss": 4.7382,107 "step": 3500108 },109 {110 "epoch": 0.14,111 "eval_accuracy": 0.695059597492218,112 "eval_loss": 4.704948902130127,113 "eval_macroPrec": 0.7011306349125166,114 "eval_macroRec": 0.6888804113847884,115 "eval_macrof1": 0.6931494236874167,116 "eval_runtime": 576.8932,117 "eval_samples_per_second": 1222.308,118 "eval_steps_per_second": 4.776,119 "step": 3500120 },121 {122 "epoch": 0.16,123 "learning_rate": 1e-05,124 "loss": 4.7348,125 "step": 4000126 },127 {128 "epoch": 0.16,129 "eval_accuracy": 0.6983993053436279,130 "eval_loss": 4.708113193511963,131 "eval_macroPrec": 0.7063721040316608,132 "eval_macroRec": 0.6894054446460803,133 "eval_macrof1": 0.6957023053182533,134 "eval_runtime": 577.0472,135 "eval_samples_per_second": 1221.982,136 "eval_steps_per_second": 4.774,137 "step": 4000138 },139 {140 "epoch": 0.18,141 "learning_rate": 1e-05,142 "loss": 4.7271,143 "step": 4500144 },145 {146 "epoch": 0.18,147 "eval_accuracy": 0.6965146064758301,148 "eval_loss": 4.703989505767822,149 "eval_macroPrec": 0.7016966733736484,150 "eval_macroRec": 0.6930968979067849,151 "eval_macrof1": 0.6956711319634581,152 "eval_runtime": 575.3624,153 "eval_samples_per_second": 1225.56,154 "eval_steps_per_second": 4.788,155 "step": 4500156 },157 {158 "epoch": 0.2,159 "learning_rate": 1e-05,160 "loss": 4.7259,161 "step": 5000162 },163 {164 "epoch": 0.2,165 "eval_accuracy": 0.69874107837677,166 "eval_loss": 4.702021598815918,167 "eval_macroPrec": 0.7016042945042669,168 "eval_macroRec": 0.6974226584847779,169 "eval_macrof1": 0.6986616770403948,170 "eval_runtime": 575.2314,171 "eval_samples_per_second": 1225.839,172 "eval_steps_per_second": 4.789,173 "step": 5000174 },175 {176 "epoch": 0.22,177 "learning_rate": 1e-05,178 "loss": 4.7197,179 "step": 5500180 },181 {182 "epoch": 0.22,183 "eval_accuracy": 0.6974264979362488,184 "eval_loss": 4.707930564880371,185 "eval_macroPrec": 0.7003383131143052,186 "eval_macroRec": 0.6976402031572404,187 "eval_macrof1": 0.6962975286974858,188 "eval_runtime": 575.3071,189 "eval_samples_per_second": 1225.677,190 "eval_steps_per_second": 4.789,191 "step": 5500192 },193 {194 "epoch": 0.24,195 "learning_rate": 1e-05,196 "loss": 4.7135,197 "step": 6000198 },199 {200 "epoch": 0.24,201 "eval_accuracy": 0.7048065662384033,202 "eval_loss": 4.693523406982422,203 "eval_macroPrec": 0.7054175910959026,204 "eval_macroRec": 0.7021117345379546,205 "eval_macrof1": 0.7034792144842983,206 "eval_runtime": 575.4054,207 "eval_samples_per_second": 1225.468,208 "eval_steps_per_second": 4.788,209 "step": 6000210 },211 {212 "epoch": 0.26,213 "learning_rate": 1e-05,214 "loss": 4.7118,215 "step": 6500216 },217 {218 "epoch": 0.26,219 "eval_accuracy": 0.7045952677726746,220 "eval_loss": 4.700965881347656,221 "eval_macroPrec": 0.7048673550484386,222 "eval_macroRec": 0.7022521214081285,223 "eval_macrof1": 0.703096187006955,224 "eval_runtime": 575.3149,225 "eval_samples_per_second": 1225.661,226 "eval_steps_per_second": 4.789,227 "step": 6500228 },229 {230 "epoch": 0.28,231 "learning_rate": 1e-05,232 "loss": 4.7101,233 "step": 7000234 },235 {236 "epoch": 0.28,237 "eval_accuracy": 0.7059183716773987,238 "eval_loss": 4.694024562835693,239 "eval_macroPrec": 0.7049601649874102,240 "eval_macroRec": 0.7057936031010754,241 "eval_macrof1": 0.7049956470156717,242 "eval_runtime": 575.6186,243 "eval_samples_per_second": 1225.014,244 "eval_steps_per_second": 4.786,245 "step": 7000246 },247 {248 "epoch": 0.3,249 "learning_rate": 1e-05,250 "loss": 4.7035,251 "step": 7500252 },253 {254 "epoch": 0.3,255 "eval_accuracy": 0.7056205868721008,256 "eval_loss": 4.702441692352295,257 "eval_macroPrec": 0.7058376660421668,258 "eval_macroRec": 0.7052063553135387,259 "eval_macrof1": 0.7050265253043888,260 "eval_runtime": 575.0614,261 "eval_samples_per_second": 1226.201,262 "eval_steps_per_second": 4.791,263 "step": 7500264 },265 {266 "epoch": 0.32,267 "learning_rate": 1e-05,268 "loss": 4.7036,269 "step": 8000270 },271 {272 "epoch": 0.32,273 "eval_accuracy": 0.7070202827453613,274 "eval_loss": 4.692503452301025,275 "eval_macroPrec": 0.7096606504418442,276 "eval_macroRec": 0.7034349307228963,277 "eval_macrof1": 0.7056030837342271,278 "eval_runtime": 575.3402,279 "eval_samples_per_second": 1225.607,280 "eval_steps_per_second": 4.788,281 "step": 8000282 },283 {284 "epoch": 0.34,285 "learning_rate": 1e-05,286 "loss": 4.7041,287 "step": 8500288 },289 {290 "epoch": 0.34,291 "eval_accuracy": 0.7082498669624329,292 "eval_loss": 4.698132514953613,293 "eval_macroPrec": 0.7126236486682623,294 "eval_macroRec": 0.7034168237683918,295 "eval_macrof1": 0.7064635115147275,296 "eval_runtime": 575.5597,297 "eval_samples_per_second": 1225.14,298 "eval_steps_per_second": 4.787,299 "step": 8500300 },301 {302 "epoch": 0.36,303 "learning_rate": 1e-05,304 "loss": 4.7032,305 "step": 9000306 },307 {308 "epoch": 0.36,309 "eval_accuracy": 0.7080697417259216,310 "eval_loss": 4.700057029724121,311 "eval_macroPrec": 0.7085243650622768,312 "eval_macroRec": 0.7073288346726845,313 "eval_macrof1": 0.707453549137527,314 "eval_runtime": 575.3159,315 "eval_samples_per_second": 1225.659,316 "eval_steps_per_second": 4.789,317 "step": 9000318 },319 {320 "epoch": 0.38,321 "learning_rate": 1e-05,322 "loss": 4.6971,323 "step": 9500324 },325 {326 "epoch": 0.38,327 "eval_accuracy": 0.7096211910247803,328 "eval_loss": 4.695491790771484,329 "eval_macroPrec": 0.7093534257679457,330 "eval_macroRec": 0.7089873902352126,331 "eval_macrof1": 0.7089231628132355,332 "eval_runtime": 575.2922,333 "eval_samples_per_second": 1225.709,334 "eval_steps_per_second": 4.789,335 "step": 9500336 },337 {338 "epoch": 0.4,339 "learning_rate": 1e-05,340 "loss": 4.697,341 "step": 10000342 },343 {344 "epoch": 0.4,345 "eval_accuracy": 0.7091531753540039,346 "eval_loss": 4.69527530670166,347 "eval_macroPrec": 0.7115081483863327,348 "eval_macroRec": 0.7068199357358054,349 "eval_macrof1": 0.7085571548254068,350 "eval_runtime": 575.5537,351 "eval_samples_per_second": 1225.152,352 "eval_steps_per_second": 4.787,353 "step": 10000354 },355 {356 "epoch": 0.42,357 "learning_rate": 1e-05,358 "loss": 4.6962,359 "step": 10500360 },361 {362 "epoch": 0.42,363 "eval_accuracy": 0.7073847651481628,364 "eval_loss": 4.711122989654541,365 "eval_macroPrec": 0.7123855412743447,366 "eval_macroRec": 0.7031976032088301,367 "eval_macrof1": 0.7063874239625263,368 "eval_runtime": 575.4008,369 "eval_samples_per_second": 1225.478,370 "eval_steps_per_second": 4.788,371 "step": 10500372 },373 {374 "epoch": 0.44,375 "learning_rate": 1e-05,376 "loss": 4.6967,377 "step": 11000378 },379 {380 "epoch": 0.44,381 "eval_accuracy": 0.7083250284194946,382 "eval_loss": 4.702260494232178,383 "eval_macroPrec": 0.7106193512409639,384 "eval_macroRec": 0.707312062312635,385 "eval_macrof1": 0.7078044182223,386 "eval_runtime": 576.2666,387 "eval_samples_per_second": 1223.637,388 "eval_steps_per_second": 4.781,389 "step": 11000390 },391 {392 "epoch": 0.46,393 "learning_rate": 1e-05,394 "loss": 4.6933,395 "step": 11500396 },397 {398 "epoch": 0.46,399 "eval_accuracy": 0.7103700041770935,400 "eval_loss": 4.70102596282959,401 "eval_macroPrec": 0.7089688659300091,402 "eval_macroRec": 0.7097678048602494,403 "eval_macrof1": 0.7091297032684482,404 "eval_runtime": 570.7047,405 "eval_samples_per_second": 1235.562,406 "eval_steps_per_second": 4.827,407 "step": 11500408 }409 ],410 "max_steps": 123955,411 "num_train_epochs": 5,412 "total_flos": 1.93654952853504e+17,413 "trial_name": null,414 "trial_params": null415}416 