|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 1.0, |
|
"eval_steps": 500, |
|
"global_step": 2410, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.0004149377593360996, |
|
"grad_norm": 26.93341562765044, |
|
"learning_rate": 4.1493775933609963e-08, |
|
"loss": 1.4725, |
|
"step": 1 |
|
}, |
|
{ |
|
"epoch": 0.002074688796680498, |
|
"grad_norm": 25.297316212761707, |
|
"learning_rate": 2.074688796680498e-07, |
|
"loss": 1.4346, |
|
"step": 5 |
|
}, |
|
{ |
|
"epoch": 0.004149377593360996, |
|
"grad_norm": 16.240075815340298, |
|
"learning_rate": 4.149377593360996e-07, |
|
"loss": 1.3793, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 0.006224066390041493, |
|
"grad_norm": 8.414236179504735, |
|
"learning_rate": 6.224066390041494e-07, |
|
"loss": 1.274, |
|
"step": 15 |
|
}, |
|
{ |
|
"epoch": 0.008298755186721992, |
|
"grad_norm": 10.510339353255072, |
|
"learning_rate": 8.298755186721992e-07, |
|
"loss": 1.1601, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 0.01037344398340249, |
|
"grad_norm": 4.437200517809562, |
|
"learning_rate": 1.037344398340249e-06, |
|
"loss": 1.0375, |
|
"step": 25 |
|
}, |
|
{ |
|
"epoch": 0.012448132780082987, |
|
"grad_norm": 3.4287551512904613, |
|
"learning_rate": 1.2448132780082988e-06, |
|
"loss": 0.9974, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 0.014522821576763486, |
|
"grad_norm": 3.4049944593465016, |
|
"learning_rate": 1.4522821576763488e-06, |
|
"loss": 0.9596, |
|
"step": 35 |
|
}, |
|
{ |
|
"epoch": 0.016597510373443983, |
|
"grad_norm": 3.0107447624660275, |
|
"learning_rate": 1.6597510373443984e-06, |
|
"loss": 0.938, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 0.01867219917012448, |
|
"grad_norm": 3.0215484563628694, |
|
"learning_rate": 1.8672199170124482e-06, |
|
"loss": 0.9083, |
|
"step": 45 |
|
}, |
|
{ |
|
"epoch": 0.02074688796680498, |
|
"grad_norm": 2.973401033499103, |
|
"learning_rate": 2.074688796680498e-06, |
|
"loss": 0.9175, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.022821576763485476, |
|
"grad_norm": 3.1166975342093104, |
|
"learning_rate": 2.282157676348548e-06, |
|
"loss": 0.9051, |
|
"step": 55 |
|
}, |
|
{ |
|
"epoch": 0.024896265560165973, |
|
"grad_norm": 3.1493583912363667, |
|
"learning_rate": 2.4896265560165977e-06, |
|
"loss": 0.8867, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 0.026970954356846474, |
|
"grad_norm": 2.971810139692584, |
|
"learning_rate": 2.6970954356846475e-06, |
|
"loss": 0.9024, |
|
"step": 65 |
|
}, |
|
{ |
|
"epoch": 0.029045643153526972, |
|
"grad_norm": 2.9673454675974535, |
|
"learning_rate": 2.9045643153526977e-06, |
|
"loss": 0.8855, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 0.03112033195020747, |
|
"grad_norm": 3.195994891124804, |
|
"learning_rate": 3.112033195020747e-06, |
|
"loss": 0.8873, |
|
"step": 75 |
|
}, |
|
{ |
|
"epoch": 0.03319502074688797, |
|
"grad_norm": 3.0341908061093603, |
|
"learning_rate": 3.319502074688797e-06, |
|
"loss": 0.8785, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 0.035269709543568464, |
|
"grad_norm": 3.0888190331536958, |
|
"learning_rate": 3.526970954356847e-06, |
|
"loss": 0.8686, |
|
"step": 85 |
|
}, |
|
{ |
|
"epoch": 0.03734439834024896, |
|
"grad_norm": 3.069834595940876, |
|
"learning_rate": 3.7344398340248965e-06, |
|
"loss": 0.8795, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 0.03941908713692946, |
|
"grad_norm": 3.07342880372721, |
|
"learning_rate": 3.941908713692946e-06, |
|
"loss": 0.8543, |
|
"step": 95 |
|
}, |
|
{ |
|
"epoch": 0.04149377593360996, |
|
"grad_norm": 3.149595655831201, |
|
"learning_rate": 4.149377593360996e-06, |
|
"loss": 0.8516, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.043568464730290454, |
|
"grad_norm": 3.1588136250157293, |
|
"learning_rate": 4.356846473029046e-06, |
|
"loss": 0.861, |
|
"step": 105 |
|
}, |
|
{ |
|
"epoch": 0.04564315352697095, |
|
"grad_norm": 3.2472498773631298, |
|
"learning_rate": 4.564315352697096e-06, |
|
"loss": 0.8626, |
|
"step": 110 |
|
}, |
|
{ |
|
"epoch": 0.04771784232365145, |
|
"grad_norm": 3.0627212516038003, |
|
"learning_rate": 4.771784232365146e-06, |
|
"loss": 0.8585, |
|
"step": 115 |
|
}, |
|
{ |
|
"epoch": 0.04979253112033195, |
|
"grad_norm": 3.3866567736302415, |
|
"learning_rate": 4.979253112033195e-06, |
|
"loss": 0.8546, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 0.05186721991701245, |
|
"grad_norm": 3.045817807536523, |
|
"learning_rate": 5.1867219917012455e-06, |
|
"loss": 0.8579, |
|
"step": 125 |
|
}, |
|
{ |
|
"epoch": 0.05394190871369295, |
|
"grad_norm": 2.9392093861675868, |
|
"learning_rate": 5.394190871369295e-06, |
|
"loss": 0.8282, |
|
"step": 130 |
|
}, |
|
{ |
|
"epoch": 0.056016597510373446, |
|
"grad_norm": 3.310150868761337, |
|
"learning_rate": 5.601659751037345e-06, |
|
"loss": 0.846, |
|
"step": 135 |
|
}, |
|
{ |
|
"epoch": 0.058091286307053944, |
|
"grad_norm": 3.03717223043934, |
|
"learning_rate": 5.809128630705395e-06, |
|
"loss": 0.8459, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 0.06016597510373444, |
|
"grad_norm": 3.028995697830279, |
|
"learning_rate": 6.016597510373444e-06, |
|
"loss": 0.8488, |
|
"step": 145 |
|
}, |
|
{ |
|
"epoch": 0.06224066390041494, |
|
"grad_norm": 3.2031060057481935, |
|
"learning_rate": 6.224066390041494e-06, |
|
"loss": 0.8323, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 0.06431535269709543, |
|
"grad_norm": 3.3375024554537545, |
|
"learning_rate": 6.431535269709544e-06, |
|
"loss": 0.8365, |
|
"step": 155 |
|
}, |
|
{ |
|
"epoch": 0.06639004149377593, |
|
"grad_norm": 3.348800008787665, |
|
"learning_rate": 6.639004149377594e-06, |
|
"loss": 0.8337, |
|
"step": 160 |
|
}, |
|
{ |
|
"epoch": 0.06846473029045644, |
|
"grad_norm": 3.052961732749455, |
|
"learning_rate": 6.846473029045644e-06, |
|
"loss": 0.8315, |
|
"step": 165 |
|
}, |
|
{ |
|
"epoch": 0.07053941908713693, |
|
"grad_norm": 3.1121690085579607, |
|
"learning_rate": 7.053941908713694e-06, |
|
"loss": 0.8401, |
|
"step": 170 |
|
}, |
|
{ |
|
"epoch": 0.07261410788381743, |
|
"grad_norm": 3.242023378743259, |
|
"learning_rate": 7.261410788381743e-06, |
|
"loss": 0.8158, |
|
"step": 175 |
|
}, |
|
{ |
|
"epoch": 0.07468879668049792, |
|
"grad_norm": 3.075162049471092, |
|
"learning_rate": 7.468879668049793e-06, |
|
"loss": 0.8347, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 0.07676348547717843, |
|
"grad_norm": 2.9264920323994255, |
|
"learning_rate": 7.676348547717844e-06, |
|
"loss": 0.8088, |
|
"step": 185 |
|
}, |
|
{ |
|
"epoch": 0.07883817427385892, |
|
"grad_norm": 2.948569896440745, |
|
"learning_rate": 7.883817427385892e-06, |
|
"loss": 0.8218, |
|
"step": 190 |
|
}, |
|
{ |
|
"epoch": 0.08091286307053942, |
|
"grad_norm": 3.3517457188955224, |
|
"learning_rate": 8.091286307053943e-06, |
|
"loss": 0.8147, |
|
"step": 195 |
|
}, |
|
{ |
|
"epoch": 0.08298755186721991, |
|
"grad_norm": 3.220462454017356, |
|
"learning_rate": 8.298755186721992e-06, |
|
"loss": 0.8349, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.08506224066390042, |
|
"grad_norm": 3.1111446024937006, |
|
"learning_rate": 8.506224066390042e-06, |
|
"loss": 0.8057, |
|
"step": 205 |
|
}, |
|
{ |
|
"epoch": 0.08713692946058091, |
|
"grad_norm": 2.892565595009811, |
|
"learning_rate": 8.713692946058093e-06, |
|
"loss": 0.8294, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 0.08921161825726141, |
|
"grad_norm": 3.235067221557126, |
|
"learning_rate": 8.921161825726142e-06, |
|
"loss": 0.7982, |
|
"step": 215 |
|
}, |
|
{ |
|
"epoch": 0.0912863070539419, |
|
"grad_norm": 3.1287542161799013, |
|
"learning_rate": 9.128630705394191e-06, |
|
"loss": 0.7973, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 0.09336099585062241, |
|
"grad_norm": 2.9450638727706315, |
|
"learning_rate": 9.33609958506224e-06, |
|
"loss": 0.8147, |
|
"step": 225 |
|
}, |
|
{ |
|
"epoch": 0.0954356846473029, |
|
"grad_norm": 3.170060713175617, |
|
"learning_rate": 9.543568464730292e-06, |
|
"loss": 0.8161, |
|
"step": 230 |
|
}, |
|
{ |
|
"epoch": 0.0975103734439834, |
|
"grad_norm": 3.0907201923890146, |
|
"learning_rate": 9.751037344398341e-06, |
|
"loss": 0.8163, |
|
"step": 235 |
|
}, |
|
{ |
|
"epoch": 0.0995850622406639, |
|
"grad_norm": 3.102859260824346, |
|
"learning_rate": 9.95850622406639e-06, |
|
"loss": 0.7918, |
|
"step": 240 |
|
}, |
|
{ |
|
"epoch": 0.1016597510373444, |
|
"grad_norm": 2.9868921661921517, |
|
"learning_rate": 9.999916085034977e-06, |
|
"loss": 0.8164, |
|
"step": 245 |
|
}, |
|
{ |
|
"epoch": 0.1037344398340249, |
|
"grad_norm": 2.8504855656123937, |
|
"learning_rate": 9.999575185316994e-06, |
|
"loss": 0.8105, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 0.10580912863070539, |
|
"grad_norm": 2.969587333276916, |
|
"learning_rate": 9.998972074026074e-06, |
|
"loss": 0.8055, |
|
"step": 255 |
|
}, |
|
{ |
|
"epoch": 0.1078838174273859, |
|
"grad_norm": 2.9939024309770157, |
|
"learning_rate": 9.998106782793455e-06, |
|
"loss": 0.7973, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 0.10995850622406639, |
|
"grad_norm": 2.997723701099069, |
|
"learning_rate": 9.996979357000869e-06, |
|
"loss": 0.7941, |
|
"step": 265 |
|
}, |
|
{ |
|
"epoch": 0.11203319502074689, |
|
"grad_norm": 2.979776887568956, |
|
"learning_rate": 9.995589855778159e-06, |
|
"loss": 0.7897, |
|
"step": 270 |
|
}, |
|
{ |
|
"epoch": 0.11410788381742738, |
|
"grad_norm": 2.7677213383011443, |
|
"learning_rate": 9.993938352000174e-06, |
|
"loss": 0.7962, |
|
"step": 275 |
|
}, |
|
{ |
|
"epoch": 0.11618257261410789, |
|
"grad_norm": 2.936072860527269, |
|
"learning_rate": 9.992024932282955e-06, |
|
"loss": 0.7934, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 0.11825726141078838, |
|
"grad_norm": 2.7901975755448003, |
|
"learning_rate": 9.989849696979188e-06, |
|
"loss": 0.8062, |
|
"step": 285 |
|
}, |
|
{ |
|
"epoch": 0.12033195020746888, |
|
"grad_norm": 2.8991854152652974, |
|
"learning_rate": 9.987412760172939e-06, |
|
"loss": 0.8007, |
|
"step": 290 |
|
}, |
|
{ |
|
"epoch": 0.12240663900414937, |
|
"grad_norm": 3.0373899188891267, |
|
"learning_rate": 9.984714249673676e-06, |
|
"loss": 0.7941, |
|
"step": 295 |
|
}, |
|
{ |
|
"epoch": 0.12448132780082988, |
|
"grad_norm": 2.8040271211308596, |
|
"learning_rate": 9.981754307009556e-06, |
|
"loss": 0.8004, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.12655601659751037, |
|
"grad_norm": 2.8585104734793263, |
|
"learning_rate": 9.978533087420015e-06, |
|
"loss": 0.8029, |
|
"step": 305 |
|
}, |
|
{ |
|
"epoch": 0.12863070539419086, |
|
"grad_norm": 2.87456284995761, |
|
"learning_rate": 9.97505075984762e-06, |
|
"loss": 0.7807, |
|
"step": 310 |
|
}, |
|
{ |
|
"epoch": 0.13070539419087138, |
|
"grad_norm": 3.056578905955286, |
|
"learning_rate": 9.971307506929202e-06, |
|
"loss": 0.7795, |
|
"step": 315 |
|
}, |
|
{ |
|
"epoch": 0.13278008298755187, |
|
"grad_norm": 2.722917451156625, |
|
"learning_rate": 9.967303524986294e-06, |
|
"loss": 0.7873, |
|
"step": 320 |
|
}, |
|
{ |
|
"epoch": 0.13485477178423236, |
|
"grad_norm": 2.9180827249941013, |
|
"learning_rate": 9.963039024014811e-06, |
|
"loss": 0.7861, |
|
"step": 325 |
|
}, |
|
{ |
|
"epoch": 0.13692946058091288, |
|
"grad_norm": 2.73431405256196, |
|
"learning_rate": 9.958514227674065e-06, |
|
"loss": 0.766, |
|
"step": 330 |
|
}, |
|
{ |
|
"epoch": 0.13900414937759337, |
|
"grad_norm": 2.9085046091623137, |
|
"learning_rate": 9.953729373275008e-06, |
|
"loss": 0.7731, |
|
"step": 335 |
|
}, |
|
{ |
|
"epoch": 0.14107883817427386, |
|
"grad_norm": 2.8708373658306523, |
|
"learning_rate": 9.9486847117678e-06, |
|
"loss": 0.7594, |
|
"step": 340 |
|
}, |
|
{ |
|
"epoch": 0.14315352697095435, |
|
"grad_norm": 3.0046135483136696, |
|
"learning_rate": 9.943380507728647e-06, |
|
"loss": 0.7618, |
|
"step": 345 |
|
}, |
|
{ |
|
"epoch": 0.14522821576763487, |
|
"grad_norm": 2.8484743596508486, |
|
"learning_rate": 9.937817039345924e-06, |
|
"loss": 0.7604, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 0.14730290456431536, |
|
"grad_norm": 2.72605173131386, |
|
"learning_rate": 9.931994598405576e-06, |
|
"loss": 0.7564, |
|
"step": 355 |
|
}, |
|
{ |
|
"epoch": 0.14937759336099585, |
|
"grad_norm": 2.909791550347346, |
|
"learning_rate": 9.925913490275834e-06, |
|
"loss": 0.7512, |
|
"step": 360 |
|
}, |
|
{ |
|
"epoch": 0.15145228215767634, |
|
"grad_norm": 2.819547837925891, |
|
"learning_rate": 9.919574033891175e-06, |
|
"loss": 0.7457, |
|
"step": 365 |
|
}, |
|
{ |
|
"epoch": 0.15352697095435686, |
|
"grad_norm": 2.7924198276169854, |
|
"learning_rate": 9.912976561735617e-06, |
|
"loss": 0.7414, |
|
"step": 370 |
|
}, |
|
{ |
|
"epoch": 0.15560165975103735, |
|
"grad_norm": 2.832091750839748, |
|
"learning_rate": 9.906121419825269e-06, |
|
"loss": 0.7363, |
|
"step": 375 |
|
}, |
|
{ |
|
"epoch": 0.15767634854771784, |
|
"grad_norm": 2.829243284643042, |
|
"learning_rate": 9.899008967690185e-06, |
|
"loss": 0.7326, |
|
"step": 380 |
|
}, |
|
{ |
|
"epoch": 0.15975103734439833, |
|
"grad_norm": 2.826535782313376, |
|
"learning_rate": 9.891639578355511e-06, |
|
"loss": 0.7538, |
|
"step": 385 |
|
}, |
|
{ |
|
"epoch": 0.16182572614107885, |
|
"grad_norm": 2.882936305924773, |
|
"learning_rate": 9.88401363832192e-06, |
|
"loss": 0.7302, |
|
"step": 390 |
|
}, |
|
{ |
|
"epoch": 0.16390041493775934, |
|
"grad_norm": 2.906899687666024, |
|
"learning_rate": 9.87613154754534e-06, |
|
"loss": 0.7129, |
|
"step": 395 |
|
}, |
|
{ |
|
"epoch": 0.16597510373443983, |
|
"grad_norm": 2.7967270291757944, |
|
"learning_rate": 9.867993719415974e-06, |
|
"loss": 0.7416, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.16804979253112035, |
|
"grad_norm": 2.8084997358792023, |
|
"learning_rate": 9.859600580736632e-06, |
|
"loss": 0.7353, |
|
"step": 405 |
|
}, |
|
{ |
|
"epoch": 0.17012448132780084, |
|
"grad_norm": 2.680436779833713, |
|
"learning_rate": 9.850952571700332e-06, |
|
"loss": 0.7376, |
|
"step": 410 |
|
}, |
|
{ |
|
"epoch": 0.17219917012448133, |
|
"grad_norm": 2.8865976677763907, |
|
"learning_rate": 9.842050145867219e-06, |
|
"loss": 0.7262, |
|
"step": 415 |
|
}, |
|
{ |
|
"epoch": 0.17427385892116182, |
|
"grad_norm": 2.8914725126052416, |
|
"learning_rate": 9.832893770140778e-06, |
|
"loss": 0.7161, |
|
"step": 420 |
|
}, |
|
{ |
|
"epoch": 0.17634854771784234, |
|
"grad_norm": 2.7982408695630725, |
|
"learning_rate": 9.823483924743348e-06, |
|
"loss": 0.7283, |
|
"step": 425 |
|
}, |
|
{ |
|
"epoch": 0.17842323651452283, |
|
"grad_norm": 2.6688029764570604, |
|
"learning_rate": 9.813821103190932e-06, |
|
"loss": 0.7192, |
|
"step": 430 |
|
}, |
|
{ |
|
"epoch": 0.18049792531120332, |
|
"grad_norm": 2.8287346081752327, |
|
"learning_rate": 9.803905812267317e-06, |
|
"loss": 0.724, |
|
"step": 435 |
|
}, |
|
{ |
|
"epoch": 0.1825726141078838, |
|
"grad_norm": 2.744322820321148, |
|
"learning_rate": 9.793738571997488e-06, |
|
"loss": 0.7204, |
|
"step": 440 |
|
}, |
|
{ |
|
"epoch": 0.18464730290456433, |
|
"grad_norm": 2.736399026435038, |
|
"learning_rate": 9.783319915620365e-06, |
|
"loss": 0.7265, |
|
"step": 445 |
|
}, |
|
{ |
|
"epoch": 0.18672199170124482, |
|
"grad_norm": 2.8039795008875887, |
|
"learning_rate": 9.772650389560829e-06, |
|
"loss": 0.7217, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 0.1887966804979253, |
|
"grad_norm": 2.6831409383313893, |
|
"learning_rate": 9.761730553401067e-06, |
|
"loss": 0.7016, |
|
"step": 455 |
|
}, |
|
{ |
|
"epoch": 0.1908713692946058, |
|
"grad_norm": 2.7717760950966492, |
|
"learning_rate": 9.750560979851222e-06, |
|
"loss": 0.728, |
|
"step": 460 |
|
}, |
|
{ |
|
"epoch": 0.19294605809128632, |
|
"grad_norm": 2.614242945562297, |
|
"learning_rate": 9.739142254719351e-06, |
|
"loss": 0.7025, |
|
"step": 465 |
|
}, |
|
{ |
|
"epoch": 0.1950207468879668, |
|
"grad_norm": 2.649844888470243, |
|
"learning_rate": 9.727474976880718e-06, |
|
"loss": 0.7069, |
|
"step": 470 |
|
}, |
|
{ |
|
"epoch": 0.1970954356846473, |
|
"grad_norm": 2.800760446265028, |
|
"learning_rate": 9.715559758246363e-06, |
|
"loss": 0.6901, |
|
"step": 475 |
|
}, |
|
{ |
|
"epoch": 0.1991701244813278, |
|
"grad_norm": 2.7922242993251265, |
|
"learning_rate": 9.703397223731028e-06, |
|
"loss": 0.6913, |
|
"step": 480 |
|
}, |
|
{ |
|
"epoch": 0.2012448132780083, |
|
"grad_norm": 2.907407187432472, |
|
"learning_rate": 9.690988011220367e-06, |
|
"loss": 0.7121, |
|
"step": 485 |
|
}, |
|
{ |
|
"epoch": 0.2033195020746888, |
|
"grad_norm": 2.626218495773325, |
|
"learning_rate": 9.678332771537506e-06, |
|
"loss": 0.7026, |
|
"step": 490 |
|
}, |
|
{ |
|
"epoch": 0.2053941908713693, |
|
"grad_norm": 2.8404176936123338, |
|
"learning_rate": 9.665432168408895e-06, |
|
"loss": 0.7113, |
|
"step": 495 |
|
}, |
|
{ |
|
"epoch": 0.2074688796680498, |
|
"grad_norm": 2.786382253827723, |
|
"learning_rate": 9.652286878429508e-06, |
|
"loss": 0.6852, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.2095435684647303, |
|
"grad_norm": 2.7800940175136515, |
|
"learning_rate": 9.638897591027355e-06, |
|
"loss": 0.6917, |
|
"step": 505 |
|
}, |
|
{ |
|
"epoch": 0.21161825726141079, |
|
"grad_norm": 2.710182741542463, |
|
"learning_rate": 9.625265008427317e-06, |
|
"loss": 0.6903, |
|
"step": 510 |
|
}, |
|
{ |
|
"epoch": 0.21369294605809128, |
|
"grad_norm": 3.0976476780998827, |
|
"learning_rate": 9.61138984561433e-06, |
|
"loss": 0.6776, |
|
"step": 515 |
|
}, |
|
{ |
|
"epoch": 0.2157676348547718, |
|
"grad_norm": 2.6986195603682015, |
|
"learning_rate": 9.597272830295877e-06, |
|
"loss": 0.6759, |
|
"step": 520 |
|
}, |
|
{ |
|
"epoch": 0.21784232365145229, |
|
"grad_norm": 2.8093612594263493, |
|
"learning_rate": 9.582914702863816e-06, |
|
"loss": 0.691, |
|
"step": 525 |
|
}, |
|
{ |
|
"epoch": 0.21991701244813278, |
|
"grad_norm": 2.668102059117624, |
|
"learning_rate": 9.568316216355569e-06, |
|
"loss": 0.6658, |
|
"step": 530 |
|
}, |
|
{ |
|
"epoch": 0.22199170124481327, |
|
"grad_norm": 2.968496552447308, |
|
"learning_rate": 9.553478136414606e-06, |
|
"loss": 0.6585, |
|
"step": 535 |
|
}, |
|
{ |
|
"epoch": 0.22406639004149378, |
|
"grad_norm": 2.788495237514456, |
|
"learning_rate": 9.538401241250302e-06, |
|
"loss": 0.6761, |
|
"step": 540 |
|
}, |
|
{ |
|
"epoch": 0.22614107883817428, |
|
"grad_norm": 2.692620475621199, |
|
"learning_rate": 9.523086321597123e-06, |
|
"loss": 0.6862, |
|
"step": 545 |
|
}, |
|
{ |
|
"epoch": 0.22821576763485477, |
|
"grad_norm": 2.8519734853628274, |
|
"learning_rate": 9.507534180673142e-06, |
|
"loss": 0.6765, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 0.23029045643153526, |
|
"grad_norm": 2.6424845030364934, |
|
"learning_rate": 9.49174563413793e-06, |
|
"loss": 0.693, |
|
"step": 555 |
|
}, |
|
{ |
|
"epoch": 0.23236514522821577, |
|
"grad_norm": 2.68422130623634, |
|
"learning_rate": 9.475721510049765e-06, |
|
"loss": 0.6549, |
|
"step": 560 |
|
}, |
|
{ |
|
"epoch": 0.23443983402489627, |
|
"grad_norm": 2.745364732974406, |
|
"learning_rate": 9.459462648822209e-06, |
|
"loss": 0.6762, |
|
"step": 565 |
|
}, |
|
{ |
|
"epoch": 0.23651452282157676, |
|
"grad_norm": 2.7138771122564496, |
|
"learning_rate": 9.442969903180021e-06, |
|
"loss": 0.6617, |
|
"step": 570 |
|
}, |
|
{ |
|
"epoch": 0.23858921161825727, |
|
"grad_norm": 2.594975020618961, |
|
"learning_rate": 9.426244138114456e-06, |
|
"loss": 0.6644, |
|
"step": 575 |
|
}, |
|
{ |
|
"epoch": 0.24066390041493776, |
|
"grad_norm": 2.6711485289021217, |
|
"learning_rate": 9.409286230837876e-06, |
|
"loss": 0.6515, |
|
"step": 580 |
|
}, |
|
{ |
|
"epoch": 0.24273858921161826, |
|
"grad_norm": 3.157835949894035, |
|
"learning_rate": 9.39209707073775e-06, |
|
"loss": 0.6552, |
|
"step": 585 |
|
}, |
|
{ |
|
"epoch": 0.24481327800829875, |
|
"grad_norm": 2.7728392273294427, |
|
"learning_rate": 9.37467755933002e-06, |
|
"loss": 0.6547, |
|
"step": 590 |
|
}, |
|
{ |
|
"epoch": 0.24688796680497926, |
|
"grad_norm": 2.6326098618284726, |
|
"learning_rate": 9.357028610211802e-06, |
|
"loss": 0.6581, |
|
"step": 595 |
|
}, |
|
{ |
|
"epoch": 0.24896265560165975, |
|
"grad_norm": 2.645526952797483, |
|
"learning_rate": 9.339151149013483e-06, |
|
"loss": 0.66, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.25103734439834025, |
|
"grad_norm": 2.695303011335832, |
|
"learning_rate": 9.32104611335017e-06, |
|
"loss": 0.6494, |
|
"step": 605 |
|
}, |
|
{ |
|
"epoch": 0.25311203319502074, |
|
"grad_norm": 2.595062725441063, |
|
"learning_rate": 9.302714452772515e-06, |
|
"loss": 0.6443, |
|
"step": 610 |
|
}, |
|
{ |
|
"epoch": 0.2551867219917012, |
|
"grad_norm": 2.6458460885695168, |
|
"learning_rate": 9.284157128716916e-06, |
|
"loss": 0.6514, |
|
"step": 615 |
|
}, |
|
{ |
|
"epoch": 0.2572614107883817, |
|
"grad_norm": 2.5786064674401583, |
|
"learning_rate": 9.265375114455091e-06, |
|
"loss": 0.6405, |
|
"step": 620 |
|
}, |
|
{ |
|
"epoch": 0.25933609958506226, |
|
"grad_norm": 2.6907850585932027, |
|
"learning_rate": 9.246369395043033e-06, |
|
"loss": 0.6476, |
|
"step": 625 |
|
}, |
|
{ |
|
"epoch": 0.26141078838174275, |
|
"grad_norm": 2.6953268245247557, |
|
"learning_rate": 9.227140967269348e-06, |
|
"loss": 0.6414, |
|
"step": 630 |
|
}, |
|
{ |
|
"epoch": 0.26348547717842324, |
|
"grad_norm": 2.6657263650658427, |
|
"learning_rate": 9.20769083960298e-06, |
|
"loss": 0.6514, |
|
"step": 635 |
|
}, |
|
{ |
|
"epoch": 0.26556016597510373, |
|
"grad_norm": 2.5956487125743615, |
|
"learning_rate": 9.188020032140308e-06, |
|
"loss": 0.6494, |
|
"step": 640 |
|
}, |
|
{ |
|
"epoch": 0.2676348547717842, |
|
"grad_norm": 2.6461882359542095, |
|
"learning_rate": 9.168129576551665e-06, |
|
"loss": 0.6388, |
|
"step": 645 |
|
}, |
|
{ |
|
"epoch": 0.2697095435684647, |
|
"grad_norm": 2.734469941118851, |
|
"learning_rate": 9.148020516027207e-06, |
|
"loss": 0.639, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 0.2717842323651452, |
|
"grad_norm": 2.85311061995995, |
|
"learning_rate": 9.127693905222223e-06, |
|
"loss": 0.6239, |
|
"step": 655 |
|
}, |
|
{ |
|
"epoch": 0.27385892116182575, |
|
"grad_norm": 2.905457532313008, |
|
"learning_rate": 9.107150810201805e-06, |
|
"loss": 0.6332, |
|
"step": 660 |
|
}, |
|
{ |
|
"epoch": 0.27593360995850624, |
|
"grad_norm": 2.866240537811393, |
|
"learning_rate": 9.086392308384946e-06, |
|
"loss": 0.6402, |
|
"step": 665 |
|
}, |
|
{ |
|
"epoch": 0.27800829875518673, |
|
"grad_norm": 2.6848783944743717, |
|
"learning_rate": 9.065419488488029e-06, |
|
"loss": 0.6198, |
|
"step": 670 |
|
}, |
|
{ |
|
"epoch": 0.2800829875518672, |
|
"grad_norm": 2.806882576661356, |
|
"learning_rate": 9.044233450467728e-06, |
|
"loss": 0.6329, |
|
"step": 675 |
|
}, |
|
{ |
|
"epoch": 0.2821576763485477, |
|
"grad_norm": 2.6697354108572457, |
|
"learning_rate": 9.022835305463322e-06, |
|
"loss": 0.628, |
|
"step": 680 |
|
}, |
|
{ |
|
"epoch": 0.2842323651452282, |
|
"grad_norm": 2.692739103753555, |
|
"learning_rate": 9.001226175738409e-06, |
|
"loss": 0.6285, |
|
"step": 685 |
|
}, |
|
{ |
|
"epoch": 0.2863070539419087, |
|
"grad_norm": 2.6356169772477887, |
|
"learning_rate": 8.979407194622062e-06, |
|
"loss": 0.6029, |
|
"step": 690 |
|
}, |
|
{ |
|
"epoch": 0.2883817427385892, |
|
"grad_norm": 2.684304580451599, |
|
"learning_rate": 8.95737950644938e-06, |
|
"loss": 0.6163, |
|
"step": 695 |
|
}, |
|
{ |
|
"epoch": 0.29045643153526973, |
|
"grad_norm": 2.665591903369743, |
|
"learning_rate": 8.93514426650147e-06, |
|
"loss": 0.6036, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 0.2925311203319502, |
|
"grad_norm": 2.5849646219326394, |
|
"learning_rate": 8.912702640944862e-06, |
|
"loss": 0.63, |
|
"step": 705 |
|
}, |
|
{ |
|
"epoch": 0.2946058091286307, |
|
"grad_norm": 2.642637231105898, |
|
"learning_rate": 8.89005580677034e-06, |
|
"loss": 0.6142, |
|
"step": 710 |
|
}, |
|
{ |
|
"epoch": 0.2966804979253112, |
|
"grad_norm": 2.8131632244224543, |
|
"learning_rate": 8.867204951731227e-06, |
|
"loss": 0.6177, |
|
"step": 715 |
|
}, |
|
{ |
|
"epoch": 0.2987551867219917, |
|
"grad_norm": 2.7483245438817794, |
|
"learning_rate": 8.844151274281074e-06, |
|
"loss": 0.6194, |
|
"step": 720 |
|
}, |
|
{ |
|
"epoch": 0.3008298755186722, |
|
"grad_norm": 3.0664428917760675, |
|
"learning_rate": 8.820895983510813e-06, |
|
"loss": 0.6186, |
|
"step": 725 |
|
}, |
|
{ |
|
"epoch": 0.3029045643153527, |
|
"grad_norm": 2.688656442588083, |
|
"learning_rate": 8.797440299085344e-06, |
|
"loss": 0.5988, |
|
"step": 730 |
|
}, |
|
{ |
|
"epoch": 0.3049792531120332, |
|
"grad_norm": 2.7050533397545258, |
|
"learning_rate": 8.77378545117957e-06, |
|
"loss": 0.616, |
|
"step": 735 |
|
}, |
|
{ |
|
"epoch": 0.3070539419087137, |
|
"grad_norm": 2.693473576801362, |
|
"learning_rate": 8.74993268041387e-06, |
|
"loss": 0.5999, |
|
"step": 740 |
|
}, |
|
{ |
|
"epoch": 0.3091286307053942, |
|
"grad_norm": 2.6799339600570766, |
|
"learning_rate": 8.725883237789046e-06, |
|
"loss": 0.5903, |
|
"step": 745 |
|
}, |
|
{ |
|
"epoch": 0.3112033195020747, |
|
"grad_norm": 2.5430481996060528, |
|
"learning_rate": 8.701638384620694e-06, |
|
"loss": 0.5856, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 0.3132780082987552, |
|
"grad_norm": 2.672587466894312, |
|
"learning_rate": 8.677199392473068e-06, |
|
"loss": 0.6042, |
|
"step": 755 |
|
}, |
|
{ |
|
"epoch": 0.3153526970954357, |
|
"grad_norm": 2.7790737972229484, |
|
"learning_rate": 8.652567543092385e-06, |
|
"loss": 0.5917, |
|
"step": 760 |
|
}, |
|
{ |
|
"epoch": 0.31742738589211617, |
|
"grad_norm": 2.7954825670577677, |
|
"learning_rate": 8.627744128339599e-06, |
|
"loss": 0.5939, |
|
"step": 765 |
|
}, |
|
{ |
|
"epoch": 0.31950207468879666, |
|
"grad_norm": 2.662658318367537, |
|
"learning_rate": 8.602730450122648e-06, |
|
"loss": 0.6011, |
|
"step": 770 |
|
}, |
|
{ |
|
"epoch": 0.3215767634854772, |
|
"grad_norm": 2.6270972350688884, |
|
"learning_rate": 8.577527820328176e-06, |
|
"loss": 0.5859, |
|
"step": 775 |
|
}, |
|
{ |
|
"epoch": 0.3236514522821577, |
|
"grad_norm": 2.5997336791640144, |
|
"learning_rate": 8.552137560752728e-06, |
|
"loss": 0.5695, |
|
"step": 780 |
|
}, |
|
{ |
|
"epoch": 0.3257261410788382, |
|
"grad_norm": 2.6129156491647096, |
|
"learning_rate": 8.526561003033424e-06, |
|
"loss": 0.5991, |
|
"step": 785 |
|
}, |
|
{ |
|
"epoch": 0.3278008298755187, |
|
"grad_norm": 2.641434925698528, |
|
"learning_rate": 8.50079948857812e-06, |
|
"loss": 0.5756, |
|
"step": 790 |
|
}, |
|
{ |
|
"epoch": 0.32987551867219916, |
|
"grad_norm": 2.744754568258112, |
|
"learning_rate": 8.474854368495055e-06, |
|
"loss": 0.5916, |
|
"step": 795 |
|
}, |
|
{ |
|
"epoch": 0.33195020746887965, |
|
"grad_norm": 2.8742812662542767, |
|
"learning_rate": 8.44872700352199e-06, |
|
"loss": 0.5858, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 0.33402489626556015, |
|
"grad_norm": 2.675688772837824, |
|
"learning_rate": 8.422418763954841e-06, |
|
"loss": 0.5858, |
|
"step": 805 |
|
}, |
|
{ |
|
"epoch": 0.3360995850622407, |
|
"grad_norm": 2.6454009935375202, |
|
"learning_rate": 8.395931029575817e-06, |
|
"loss": 0.5793, |
|
"step": 810 |
|
}, |
|
{ |
|
"epoch": 0.3381742738589212, |
|
"grad_norm": 2.5799174291219855, |
|
"learning_rate": 8.369265189581048e-06, |
|
"loss": 0.5577, |
|
"step": 815 |
|
}, |
|
{ |
|
"epoch": 0.34024896265560167, |
|
"grad_norm": 2.6266021968279025, |
|
"learning_rate": 8.342422642507727e-06, |
|
"loss": 0.5908, |
|
"step": 820 |
|
}, |
|
{ |
|
"epoch": 0.34232365145228216, |
|
"grad_norm": 2.6270567170702104, |
|
"learning_rate": 8.31540479616076e-06, |
|
"loss": 0.5583, |
|
"step": 825 |
|
}, |
|
{ |
|
"epoch": 0.34439834024896265, |
|
"grad_norm": 2.6605567528471283, |
|
"learning_rate": 8.288213067538936e-06, |
|
"loss": 0.5792, |
|
"step": 830 |
|
}, |
|
{ |
|
"epoch": 0.34647302904564314, |
|
"grad_norm": 2.6648839547096936, |
|
"learning_rate": 8.260848882760616e-06, |
|
"loss": 0.5758, |
|
"step": 835 |
|
}, |
|
{ |
|
"epoch": 0.34854771784232363, |
|
"grad_norm": 2.4985116541737487, |
|
"learning_rate": 8.233313676988917e-06, |
|
"loss": 0.5589, |
|
"step": 840 |
|
}, |
|
{ |
|
"epoch": 0.3506224066390041, |
|
"grad_norm": 2.5295015470513014, |
|
"learning_rate": 8.205608894356461e-06, |
|
"loss": 0.5747, |
|
"step": 845 |
|
}, |
|
{ |
|
"epoch": 0.35269709543568467, |
|
"grad_norm": 2.6313000253018317, |
|
"learning_rate": 8.177735987889628e-06, |
|
"loss": 0.5794, |
|
"step": 850 |
|
}, |
|
{ |
|
"epoch": 0.35477178423236516, |
|
"grad_norm": 2.841463850249757, |
|
"learning_rate": 8.149696419432352e-06, |
|
"loss": 0.5662, |
|
"step": 855 |
|
}, |
|
{ |
|
"epoch": 0.35684647302904565, |
|
"grad_norm": 2.499217630200596, |
|
"learning_rate": 8.121491659569442e-06, |
|
"loss": 0.5489, |
|
"step": 860 |
|
}, |
|
{ |
|
"epoch": 0.35892116182572614, |
|
"grad_norm": 2.624500775254657, |
|
"learning_rate": 8.093123187549475e-06, |
|
"loss": 0.5529, |
|
"step": 865 |
|
}, |
|
{ |
|
"epoch": 0.36099585062240663, |
|
"grad_norm": 2.891141879312641, |
|
"learning_rate": 8.064592491207193e-06, |
|
"loss": 0.5615, |
|
"step": 870 |
|
}, |
|
{ |
|
"epoch": 0.3630705394190871, |
|
"grad_norm": 2.985654722661823, |
|
"learning_rate": 8.035901066885486e-06, |
|
"loss": 0.5727, |
|
"step": 875 |
|
}, |
|
{ |
|
"epoch": 0.3651452282157676, |
|
"grad_norm": 2.7907723136875995, |
|
"learning_rate": 8.007050419356898e-06, |
|
"loss": 0.5608, |
|
"step": 880 |
|
}, |
|
{ |
|
"epoch": 0.36721991701244816, |
|
"grad_norm": 2.6416506986010813, |
|
"learning_rate": 7.978042061744728e-06, |
|
"loss": 0.5557, |
|
"step": 885 |
|
}, |
|
{ |
|
"epoch": 0.36929460580912865, |
|
"grad_norm": 2.5258498715580067, |
|
"learning_rate": 7.94887751544365e-06, |
|
"loss": 0.5534, |
|
"step": 890 |
|
}, |
|
{ |
|
"epoch": 0.37136929460580914, |
|
"grad_norm": 2.561386861258123, |
|
"learning_rate": 7.919558310039937e-06, |
|
"loss": 0.556, |
|
"step": 895 |
|
}, |
|
{ |
|
"epoch": 0.37344398340248963, |
|
"grad_norm": 2.727627068843935, |
|
"learning_rate": 7.890085983231225e-06, |
|
"loss": 0.5462, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 0.3755186721991701, |
|
"grad_norm": 2.5718069262603533, |
|
"learning_rate": 7.860462080745884e-06, |
|
"loss": 0.5399, |
|
"step": 905 |
|
}, |
|
{ |
|
"epoch": 0.3775933609958506, |
|
"grad_norm": 2.5380387267872746, |
|
"learning_rate": 7.830688156261927e-06, |
|
"loss": 0.5502, |
|
"step": 910 |
|
}, |
|
{ |
|
"epoch": 0.3796680497925311, |
|
"grad_norm": 2.6214419026204565, |
|
"learning_rate": 7.800765771325546e-06, |
|
"loss": 0.5417, |
|
"step": 915 |
|
}, |
|
{ |
|
"epoch": 0.3817427385892116, |
|
"grad_norm": 2.712261230434538, |
|
"learning_rate": 7.7706964952692e-06, |
|
"loss": 0.545, |
|
"step": 920 |
|
}, |
|
{ |
|
"epoch": 0.38381742738589214, |
|
"grad_norm": 2.5420503220632797, |
|
"learning_rate": 7.740481905129307e-06, |
|
"loss": 0.5333, |
|
"step": 925 |
|
}, |
|
{ |
|
"epoch": 0.38589211618257263, |
|
"grad_norm": 2.6001135316476893, |
|
"learning_rate": 7.710123585563552e-06, |
|
"loss": 0.532, |
|
"step": 930 |
|
}, |
|
{ |
|
"epoch": 0.3879668049792531, |
|
"grad_norm": 2.720274250658824, |
|
"learning_rate": 7.679623128767754e-06, |
|
"loss": 0.5466, |
|
"step": 935 |
|
}, |
|
{ |
|
"epoch": 0.3900414937759336, |
|
"grad_norm": 2.751465451077315, |
|
"learning_rate": 7.648982134392378e-06, |
|
"loss": 0.5411, |
|
"step": 940 |
|
}, |
|
{ |
|
"epoch": 0.3921161825726141, |
|
"grad_norm": 2.5435730720837864, |
|
"learning_rate": 7.618202209458623e-06, |
|
"loss": 0.5225, |
|
"step": 945 |
|
}, |
|
{ |
|
"epoch": 0.3941908713692946, |
|
"grad_norm": 2.5155581478521456, |
|
"learning_rate": 7.587284968274155e-06, |
|
"loss": 0.5361, |
|
"step": 950 |
|
}, |
|
{ |
|
"epoch": 0.3962655601659751, |
|
"grad_norm": 2.6500904333286384, |
|
"learning_rate": 7.556232032348429e-06, |
|
"loss": 0.5323, |
|
"step": 955 |
|
}, |
|
{ |
|
"epoch": 0.3983402489626556, |
|
"grad_norm": 2.6832927074173947, |
|
"learning_rate": 7.5250450303076526e-06, |
|
"loss": 0.5324, |
|
"step": 960 |
|
}, |
|
{ |
|
"epoch": 0.4004149377593361, |
|
"grad_norm": 2.5794687904843934, |
|
"learning_rate": 7.49372559780937e-06, |
|
"loss": 0.5401, |
|
"step": 965 |
|
}, |
|
{ |
|
"epoch": 0.4024896265560166, |
|
"grad_norm": 2.779973776708556, |
|
"learning_rate": 7.462275377456671e-06, |
|
"loss": 0.5353, |
|
"step": 970 |
|
}, |
|
{ |
|
"epoch": 0.4045643153526971, |
|
"grad_norm": 2.510096649942157, |
|
"learning_rate": 7.430696018712049e-06, |
|
"loss": 0.5178, |
|
"step": 975 |
|
}, |
|
{ |
|
"epoch": 0.4066390041493776, |
|
"grad_norm": 2.544609714596366, |
|
"learning_rate": 7.398989177810889e-06, |
|
"loss": 0.5305, |
|
"step": 980 |
|
}, |
|
{ |
|
"epoch": 0.4087136929460581, |
|
"grad_norm": 2.6358450407419394, |
|
"learning_rate": 7.3671565176746025e-06, |
|
"loss": 0.5083, |
|
"step": 985 |
|
}, |
|
{ |
|
"epoch": 0.4107883817427386, |
|
"grad_norm": 2.5104688131658115, |
|
"learning_rate": 7.335199707823415e-06, |
|
"loss": 0.5213, |
|
"step": 990 |
|
}, |
|
{ |
|
"epoch": 0.41286307053941906, |
|
"grad_norm": 2.534767895856582, |
|
"learning_rate": 7.3031204242888e-06, |
|
"loss": 0.5003, |
|
"step": 995 |
|
}, |
|
{ |
|
"epoch": 0.4149377593360996, |
|
"grad_norm": 2.734749407560919, |
|
"learning_rate": 7.270920349525584e-06, |
|
"loss": 0.5179, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 0.4170124481327801, |
|
"grad_norm": 2.404837852320902, |
|
"learning_rate": 7.238601172323701e-06, |
|
"loss": 0.5231, |
|
"step": 1005 |
|
}, |
|
{ |
|
"epoch": 0.4190871369294606, |
|
"grad_norm": 2.655040755780098, |
|
"learning_rate": 7.206164587719627e-06, |
|
"loss": 0.5191, |
|
"step": 1010 |
|
}, |
|
{ |
|
"epoch": 0.4211618257261411, |
|
"grad_norm": 2.537938752408207, |
|
"learning_rate": 7.173612296907473e-06, |
|
"loss": 0.5206, |
|
"step": 1015 |
|
}, |
|
{ |
|
"epoch": 0.42323651452282157, |
|
"grad_norm": 2.6276182145014135, |
|
"learning_rate": 7.1409460071497675e-06, |
|
"loss": 0.5294, |
|
"step": 1020 |
|
}, |
|
{ |
|
"epoch": 0.42531120331950206, |
|
"grad_norm": 2.543796942084407, |
|
"learning_rate": 7.108167431687917e-06, |
|
"loss": 0.5166, |
|
"step": 1025 |
|
}, |
|
{ |
|
"epoch": 0.42738589211618255, |
|
"grad_norm": 2.6672869217839468, |
|
"learning_rate": 7.075278289652349e-06, |
|
"loss": 0.5225, |
|
"step": 1030 |
|
}, |
|
{ |
|
"epoch": 0.42946058091286304, |
|
"grad_norm": 2.531083972850949, |
|
"learning_rate": 7.042280305972354e-06, |
|
"loss": 0.5142, |
|
"step": 1035 |
|
}, |
|
{ |
|
"epoch": 0.4315352697095436, |
|
"grad_norm": 2.4747366529454693, |
|
"learning_rate": 7.009175211285611e-06, |
|
"loss": 0.5141, |
|
"step": 1040 |
|
}, |
|
{ |
|
"epoch": 0.4336099585062241, |
|
"grad_norm": 2.493745172465378, |
|
"learning_rate": 6.975964741847427e-06, |
|
"loss": 0.5241, |
|
"step": 1045 |
|
}, |
|
{ |
|
"epoch": 0.43568464730290457, |
|
"grad_norm": 2.59288402554468, |
|
"learning_rate": 6.942650639439678e-06, |
|
"loss": 0.5127, |
|
"step": 1050 |
|
}, |
|
{ |
|
"epoch": 0.43775933609958506, |
|
"grad_norm": 2.5743234047985952, |
|
"learning_rate": 6.9092346512794475e-06, |
|
"loss": 0.5112, |
|
"step": 1055 |
|
}, |
|
{ |
|
"epoch": 0.43983402489626555, |
|
"grad_norm": 2.46036118970127, |
|
"learning_rate": 6.875718529927404e-06, |
|
"loss": 0.4881, |
|
"step": 1060 |
|
}, |
|
{ |
|
"epoch": 0.44190871369294604, |
|
"grad_norm": 2.582868746430813, |
|
"learning_rate": 6.8421040331958745e-06, |
|
"loss": 0.5072, |
|
"step": 1065 |
|
}, |
|
{ |
|
"epoch": 0.44398340248962653, |
|
"grad_norm": 2.5161140611104056, |
|
"learning_rate": 6.808392924056659e-06, |
|
"loss": 0.5068, |
|
"step": 1070 |
|
}, |
|
{ |
|
"epoch": 0.4460580912863071, |
|
"grad_norm": 2.616608707294859, |
|
"learning_rate": 6.774586970548567e-06, |
|
"loss": 0.5132, |
|
"step": 1075 |
|
}, |
|
{ |
|
"epoch": 0.44813278008298757, |
|
"grad_norm": 2.4704454065037424, |
|
"learning_rate": 6.7406879456846875e-06, |
|
"loss": 0.4942, |
|
"step": 1080 |
|
}, |
|
{ |
|
"epoch": 0.45020746887966806, |
|
"grad_norm": 2.5654508809058805, |
|
"learning_rate": 6.7066976273594e-06, |
|
"loss": 0.4914, |
|
"step": 1085 |
|
}, |
|
{ |
|
"epoch": 0.45228215767634855, |
|
"grad_norm": 2.5642213246435017, |
|
"learning_rate": 6.672617798255135e-06, |
|
"loss": 0.5043, |
|
"step": 1090 |
|
}, |
|
{ |
|
"epoch": 0.45435684647302904, |
|
"grad_norm": 2.3759068788938538, |
|
"learning_rate": 6.63845024574887e-06, |
|
"loss": 0.4744, |
|
"step": 1095 |
|
}, |
|
{ |
|
"epoch": 0.45643153526970953, |
|
"grad_norm": 2.4485238164491, |
|
"learning_rate": 6.604196761818395e-06, |
|
"loss": 0.4992, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 0.45850622406639, |
|
"grad_norm": 2.481384678496845, |
|
"learning_rate": 6.5698591429483286e-06, |
|
"loss": 0.5016, |
|
"step": 1105 |
|
}, |
|
{ |
|
"epoch": 0.4605809128630705, |
|
"grad_norm": 2.571902891821277, |
|
"learning_rate": 6.535439190035884e-06, |
|
"loss": 0.4893, |
|
"step": 1110 |
|
}, |
|
{ |
|
"epoch": 0.46265560165975106, |
|
"grad_norm": 2.559158463935003, |
|
"learning_rate": 6.5009387082964405e-06, |
|
"loss": 0.4991, |
|
"step": 1115 |
|
}, |
|
{ |
|
"epoch": 0.46473029045643155, |
|
"grad_norm": 2.538459567637443, |
|
"learning_rate": 6.466359507168849e-06, |
|
"loss": 0.4953, |
|
"step": 1120 |
|
}, |
|
{ |
|
"epoch": 0.46680497925311204, |
|
"grad_norm": 2.5267222175872623, |
|
"learning_rate": 6.431703400220541e-06, |
|
"loss": 0.482, |
|
"step": 1125 |
|
}, |
|
{ |
|
"epoch": 0.46887966804979253, |
|
"grad_norm": 2.38064786704353, |
|
"learning_rate": 6.396972205052407e-06, |
|
"loss": 0.4839, |
|
"step": 1130 |
|
}, |
|
{ |
|
"epoch": 0.470954356846473, |
|
"grad_norm": 2.5624442145243935, |
|
"learning_rate": 6.362167743203474e-06, |
|
"loss": 0.4975, |
|
"step": 1135 |
|
}, |
|
{ |
|
"epoch": 0.4730290456431535, |
|
"grad_norm": 2.4427415841202014, |
|
"learning_rate": 6.327291840055365e-06, |
|
"loss": 0.4759, |
|
"step": 1140 |
|
}, |
|
{ |
|
"epoch": 0.475103734439834, |
|
"grad_norm": 2.5750114799610095, |
|
"learning_rate": 6.292346324736578e-06, |
|
"loss": 0.4931, |
|
"step": 1145 |
|
}, |
|
{ |
|
"epoch": 0.47717842323651455, |
|
"grad_norm": 2.368089990602912, |
|
"learning_rate": 6.2573330300265375e-06, |
|
"loss": 0.4943, |
|
"step": 1150 |
|
}, |
|
{ |
|
"epoch": 0.47925311203319504, |
|
"grad_norm": 2.430038445579993, |
|
"learning_rate": 6.22225379225948e-06, |
|
"loss": 0.4776, |
|
"step": 1155 |
|
}, |
|
{ |
|
"epoch": 0.48132780082987553, |
|
"grad_norm": 2.655610966122537, |
|
"learning_rate": 6.18711045122814e-06, |
|
"loss": 0.4931, |
|
"step": 1160 |
|
}, |
|
{ |
|
"epoch": 0.483402489626556, |
|
"grad_norm": 2.561229036359029, |
|
"learning_rate": 6.151904850087265e-06, |
|
"loss": 0.4853, |
|
"step": 1165 |
|
}, |
|
{ |
|
"epoch": 0.4854771784232365, |
|
"grad_norm": 2.5716333011695487, |
|
"learning_rate": 6.116638835256943e-06, |
|
"loss": 0.4662, |
|
"step": 1170 |
|
}, |
|
{ |
|
"epoch": 0.487551867219917, |
|
"grad_norm": 2.6260675883477123, |
|
"learning_rate": 6.081314256325762e-06, |
|
"loss": 0.4771, |
|
"step": 1175 |
|
}, |
|
{ |
|
"epoch": 0.4896265560165975, |
|
"grad_norm": 2.464680850201926, |
|
"learning_rate": 6.045932965953813e-06, |
|
"loss": 0.4819, |
|
"step": 1180 |
|
}, |
|
{ |
|
"epoch": 0.491701244813278, |
|
"grad_norm": 2.4612043275661852, |
|
"learning_rate": 6.010496819775518e-06, |
|
"loss": 0.4744, |
|
"step": 1185 |
|
}, |
|
{ |
|
"epoch": 0.49377593360995853, |
|
"grad_norm": 2.4847373767230994, |
|
"learning_rate": 5.9750076763023025e-06, |
|
"loss": 0.4895, |
|
"step": 1190 |
|
}, |
|
{ |
|
"epoch": 0.495850622406639, |
|
"grad_norm": 2.324351303455581, |
|
"learning_rate": 5.939467396825137e-06, |
|
"loss": 0.4771, |
|
"step": 1195 |
|
}, |
|
{ |
|
"epoch": 0.4979253112033195, |
|
"grad_norm": 2.4411119886082546, |
|
"learning_rate": 5.903877845316906e-06, |
|
"loss": 0.4698, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 0.5, |
|
"grad_norm": 2.5240319324749887, |
|
"learning_rate": 5.8682408883346535e-06, |
|
"loss": 0.4705, |
|
"step": 1205 |
|
}, |
|
{ |
|
"epoch": 0.5020746887966805, |
|
"grad_norm": 2.6132898743051123, |
|
"learning_rate": 5.832558394921688e-06, |
|
"loss": 0.4557, |
|
"step": 1210 |
|
}, |
|
{ |
|
"epoch": 0.504149377593361, |
|
"grad_norm": 2.484591361229376, |
|
"learning_rate": 5.796832236509556e-06, |
|
"loss": 0.4753, |
|
"step": 1215 |
|
}, |
|
{ |
|
"epoch": 0.5062240663900415, |
|
"grad_norm": 2.392319441698671, |
|
"learning_rate": 5.761064286819895e-06, |
|
"loss": 0.458, |
|
"step": 1220 |
|
}, |
|
{ |
|
"epoch": 0.508298755186722, |
|
"grad_norm": 2.6251874596416966, |
|
"learning_rate": 5.725256421766158e-06, |
|
"loss": 0.4809, |
|
"step": 1225 |
|
}, |
|
{ |
|
"epoch": 0.5103734439834025, |
|
"grad_norm": 2.4584477311068125, |
|
"learning_rate": 5.689410519355226e-06, |
|
"loss": 0.4709, |
|
"step": 1230 |
|
}, |
|
{ |
|
"epoch": 0.5124481327800829, |
|
"grad_norm": 2.3560580563552858, |
|
"learning_rate": 5.653528459588925e-06, |
|
"loss": 0.4659, |
|
"step": 1235 |
|
}, |
|
{ |
|
"epoch": 0.5145228215767634, |
|
"grad_norm": 2.485258794734814, |
|
"learning_rate": 5.617612124365411e-06, |
|
"loss": 0.4647, |
|
"step": 1240 |
|
}, |
|
{ |
|
"epoch": 0.516597510373444, |
|
"grad_norm": 2.4250973675636938, |
|
"learning_rate": 5.5816633973804766e-06, |
|
"loss": 0.4721, |
|
"step": 1245 |
|
}, |
|
{ |
|
"epoch": 0.5186721991701245, |
|
"grad_norm": 2.4772730090302515, |
|
"learning_rate": 5.545684164028764e-06, |
|
"loss": 0.4558, |
|
"step": 1250 |
|
}, |
|
{ |
|
"epoch": 0.520746887966805, |
|
"grad_norm": 2.6817558433388053, |
|
"learning_rate": 5.509676311304869e-06, |
|
"loss": 0.4632, |
|
"step": 1255 |
|
}, |
|
{ |
|
"epoch": 0.5228215767634855, |
|
"grad_norm": 2.482278609637873, |
|
"learning_rate": 5.4736417277043865e-06, |
|
"loss": 0.4662, |
|
"step": 1260 |
|
}, |
|
{ |
|
"epoch": 0.524896265560166, |
|
"grad_norm": 2.4538360481145873, |
|
"learning_rate": 5.4375823031248545e-06, |
|
"loss": 0.4542, |
|
"step": 1265 |
|
}, |
|
{ |
|
"epoch": 0.5269709543568465, |
|
"grad_norm": 2.3974115486101693, |
|
"learning_rate": 5.401499928766644e-06, |
|
"loss": 0.4688, |
|
"step": 1270 |
|
}, |
|
{ |
|
"epoch": 0.529045643153527, |
|
"grad_norm": 2.35868507934765, |
|
"learning_rate": 5.365396497033763e-06, |
|
"loss": 0.4528, |
|
"step": 1275 |
|
}, |
|
{ |
|
"epoch": 0.5311203319502075, |
|
"grad_norm": 2.350069783189846, |
|
"learning_rate": 5.32927390143462e-06, |
|
"loss": 0.4606, |
|
"step": 1280 |
|
}, |
|
{ |
|
"epoch": 0.533195020746888, |
|
"grad_norm": 2.48472920397625, |
|
"learning_rate": 5.293134036482697e-06, |
|
"loss": 0.443, |
|
"step": 1285 |
|
}, |
|
{ |
|
"epoch": 0.5352697095435685, |
|
"grad_norm": 2.4861989102755, |
|
"learning_rate": 5.256978797597202e-06, |
|
"loss": 0.4639, |
|
"step": 1290 |
|
}, |
|
{ |
|
"epoch": 0.5373443983402489, |
|
"grad_norm": 2.4356508616340116, |
|
"learning_rate": 5.220810081003656e-06, |
|
"loss": 0.4453, |
|
"step": 1295 |
|
}, |
|
{ |
|
"epoch": 0.5394190871369294, |
|
"grad_norm": 2.7333647570132213, |
|
"learning_rate": 5.184629783634441e-06, |
|
"loss": 0.4575, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 0.5414937759336099, |
|
"grad_norm": 2.4190653155415967, |
|
"learning_rate": 5.1484398030293135e-06, |
|
"loss": 0.4522, |
|
"step": 1305 |
|
}, |
|
{ |
|
"epoch": 0.5435684647302904, |
|
"grad_norm": 2.4062268029705636, |
|
"learning_rate": 5.112242037235885e-06, |
|
"loss": 0.4517, |
|
"step": 1310 |
|
}, |
|
{ |
|
"epoch": 0.5456431535269709, |
|
"grad_norm": 2.474917237517535, |
|
"learning_rate": 5.076038384710077e-06, |
|
"loss": 0.456, |
|
"step": 1315 |
|
}, |
|
{ |
|
"epoch": 0.5477178423236515, |
|
"grad_norm": 2.5820885338481077, |
|
"learning_rate": 5.039830744216548e-06, |
|
"loss": 0.4575, |
|
"step": 1320 |
|
}, |
|
{ |
|
"epoch": 0.549792531120332, |
|
"grad_norm": 2.5020192978566365, |
|
"learning_rate": 5.003621014729113e-06, |
|
"loss": 0.4468, |
|
"step": 1325 |
|
}, |
|
{ |
|
"epoch": 0.5518672199170125, |
|
"grad_norm": 2.4844474809833224, |
|
"learning_rate": 4.967411095331149e-06, |
|
"loss": 0.4426, |
|
"step": 1330 |
|
}, |
|
{ |
|
"epoch": 0.553941908713693, |
|
"grad_norm": 2.567949384188064, |
|
"learning_rate": 4.931202885115994e-06, |
|
"loss": 0.4491, |
|
"step": 1335 |
|
}, |
|
{ |
|
"epoch": 0.5560165975103735, |
|
"grad_norm": 2.3971814860863128, |
|
"learning_rate": 4.894998283087341e-06, |
|
"loss": 0.4521, |
|
"step": 1340 |
|
}, |
|
{ |
|
"epoch": 0.558091286307054, |
|
"grad_norm": 2.6050382832392502, |
|
"learning_rate": 4.858799188059651e-06, |
|
"loss": 0.4402, |
|
"step": 1345 |
|
}, |
|
{ |
|
"epoch": 0.5601659751037344, |
|
"grad_norm": 2.504067986804623, |
|
"learning_rate": 4.822607498558555e-06, |
|
"loss": 0.4309, |
|
"step": 1350 |
|
}, |
|
{ |
|
"epoch": 0.5622406639004149, |
|
"grad_norm": 2.436795227940985, |
|
"learning_rate": 4.786425112721293e-06, |
|
"loss": 0.4432, |
|
"step": 1355 |
|
}, |
|
{ |
|
"epoch": 0.5643153526970954, |
|
"grad_norm": 2.463086006536084, |
|
"learning_rate": 4.75025392819715e-06, |
|
"loss": 0.4547, |
|
"step": 1360 |
|
}, |
|
{ |
|
"epoch": 0.5663900414937759, |
|
"grad_norm": 2.526031011478867, |
|
"learning_rate": 4.714095842047952e-06, |
|
"loss": 0.4382, |
|
"step": 1365 |
|
}, |
|
{ |
|
"epoch": 0.5684647302904564, |
|
"grad_norm": 2.484312790550504, |
|
"learning_rate": 4.677952750648544e-06, |
|
"loss": 0.4395, |
|
"step": 1370 |
|
}, |
|
{ |
|
"epoch": 0.5705394190871369, |
|
"grad_norm": 2.51470713708296, |
|
"learning_rate": 4.641826549587352e-06, |
|
"loss": 0.4324, |
|
"step": 1375 |
|
}, |
|
{ |
|
"epoch": 0.5726141078838174, |
|
"grad_norm": 2.328972994785194, |
|
"learning_rate": 4.605719133566955e-06, |
|
"loss": 0.4409, |
|
"step": 1380 |
|
}, |
|
{ |
|
"epoch": 0.5746887966804979, |
|
"grad_norm": 2.467684453460364, |
|
"learning_rate": 4.56963239630472e-06, |
|
"loss": 0.4502, |
|
"step": 1385 |
|
}, |
|
{ |
|
"epoch": 0.5767634854771784, |
|
"grad_norm": 2.5795480839970093, |
|
"learning_rate": 4.533568230433477e-06, |
|
"loss": 0.4417, |
|
"step": 1390 |
|
}, |
|
{ |
|
"epoch": 0.578838174273859, |
|
"grad_norm": 2.3301068759580907, |
|
"learning_rate": 4.497528527402262e-06, |
|
"loss": 0.4361, |
|
"step": 1395 |
|
}, |
|
{ |
|
"epoch": 0.5809128630705395, |
|
"grad_norm": 2.581867988962586, |
|
"learning_rate": 4.461515177377113e-06, |
|
"loss": 0.4398, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 0.58298755186722, |
|
"grad_norm": 2.437392550100252, |
|
"learning_rate": 4.42553006914194e-06, |
|
"loss": 0.4234, |
|
"step": 1405 |
|
}, |
|
{ |
|
"epoch": 0.5850622406639004, |
|
"grad_norm": 2.43460113329211, |
|
"learning_rate": 4.3895750899994566e-06, |
|
"loss": 0.4299, |
|
"step": 1410 |
|
}, |
|
{ |
|
"epoch": 0.5871369294605809, |
|
"grad_norm": 2.4540600474475016, |
|
"learning_rate": 4.353652125672208e-06, |
|
"loss": 0.429, |
|
"step": 1415 |
|
}, |
|
{ |
|
"epoch": 0.5892116182572614, |
|
"grad_norm": 2.402133768218254, |
|
"learning_rate": 4.317763060203665e-06, |
|
"loss": 0.4195, |
|
"step": 1420 |
|
}, |
|
{ |
|
"epoch": 0.5912863070539419, |
|
"grad_norm": 2.3040455325782103, |
|
"learning_rate": 4.281909775859411e-06, |
|
"loss": 0.4142, |
|
"step": 1425 |
|
}, |
|
{ |
|
"epoch": 0.5933609958506224, |
|
"grad_norm": 2.437892690340681, |
|
"learning_rate": 4.246094153028426e-06, |
|
"loss": 0.4376, |
|
"step": 1430 |
|
}, |
|
{ |
|
"epoch": 0.5954356846473029, |
|
"grad_norm": 2.4883366883550067, |
|
"learning_rate": 4.210318070124465e-06, |
|
"loss": 0.4343, |
|
"step": 1435 |
|
}, |
|
{ |
|
"epoch": 0.5975103734439834, |
|
"grad_norm": 2.623855396529678, |
|
"learning_rate": 4.1745834034875435e-06, |
|
"loss": 0.4316, |
|
"step": 1440 |
|
}, |
|
{ |
|
"epoch": 0.5995850622406639, |
|
"grad_norm": 2.4818254566920186, |
|
"learning_rate": 4.138892027285525e-06, |
|
"loss": 0.4201, |
|
"step": 1445 |
|
}, |
|
{ |
|
"epoch": 0.6016597510373444, |
|
"grad_norm": 2.487523042972382, |
|
"learning_rate": 4.10324581341583e-06, |
|
"loss": 0.4159, |
|
"step": 1450 |
|
}, |
|
{ |
|
"epoch": 0.6037344398340249, |
|
"grad_norm": 2.4142763091450186, |
|
"learning_rate": 4.067646631407259e-06, |
|
"loss": 0.4171, |
|
"step": 1455 |
|
}, |
|
{ |
|
"epoch": 0.6058091286307054, |
|
"grad_norm": 2.535593504431872, |
|
"learning_rate": 4.0320963483219485e-06, |
|
"loss": 0.4239, |
|
"step": 1460 |
|
}, |
|
{ |
|
"epoch": 0.6078838174273858, |
|
"grad_norm": 2.3293103348557094, |
|
"learning_rate": 3.996596828657437e-06, |
|
"loss": 0.423, |
|
"step": 1465 |
|
}, |
|
{ |
|
"epoch": 0.6099585062240664, |
|
"grad_norm": 2.3875069760256045, |
|
"learning_rate": 3.961149934248893e-06, |
|
"loss": 0.426, |
|
"step": 1470 |
|
}, |
|
{ |
|
"epoch": 0.6120331950207469, |
|
"grad_norm": 2.473197291626687, |
|
"learning_rate": 3.925757524171455e-06, |
|
"loss": 0.427, |
|
"step": 1475 |
|
}, |
|
{ |
|
"epoch": 0.6141078838174274, |
|
"grad_norm": 2.7741410864382754, |
|
"learning_rate": 3.8904214546427355e-06, |
|
"loss": 0.4122, |
|
"step": 1480 |
|
}, |
|
{ |
|
"epoch": 0.6161825726141079, |
|
"grad_norm": 2.4047571270961345, |
|
"learning_rate": 3.855143578925468e-06, |
|
"loss": 0.4131, |
|
"step": 1485 |
|
}, |
|
{ |
|
"epoch": 0.6182572614107884, |
|
"grad_norm": 2.423122390796353, |
|
"learning_rate": 3.819925747230309e-06, |
|
"loss": 0.4234, |
|
"step": 1490 |
|
}, |
|
{ |
|
"epoch": 0.6203319502074689, |
|
"grad_norm": 2.625870115353325, |
|
"learning_rate": 3.7847698066187975e-06, |
|
"loss": 0.4169, |
|
"step": 1495 |
|
}, |
|
{ |
|
"epoch": 0.6224066390041494, |
|
"grad_norm": 2.3339760463576047, |
|
"learning_rate": 3.749677600906489e-06, |
|
"loss": 0.4167, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 0.6244813278008299, |
|
"grad_norm": 2.381759807149871, |
|
"learning_rate": 3.7146509705662453e-06, |
|
"loss": 0.4127, |
|
"step": 1505 |
|
}, |
|
{ |
|
"epoch": 0.6265560165975104, |
|
"grad_norm": 2.49037216210766, |
|
"learning_rate": 3.6796917526317153e-06, |
|
"loss": 0.4196, |
|
"step": 1510 |
|
}, |
|
{ |
|
"epoch": 0.6286307053941909, |
|
"grad_norm": 2.3307430355853813, |
|
"learning_rate": 3.6448017806009804e-06, |
|
"loss": 0.4032, |
|
"step": 1515 |
|
}, |
|
{ |
|
"epoch": 0.6307053941908713, |
|
"grad_norm": 2.459076699312393, |
|
"learning_rate": 3.609982884340402e-06, |
|
"loss": 0.4191, |
|
"step": 1520 |
|
}, |
|
{ |
|
"epoch": 0.6327800829875518, |
|
"grad_norm": 2.349992533584889, |
|
"learning_rate": 3.575236889988646e-06, |
|
"loss": 0.4096, |
|
"step": 1525 |
|
}, |
|
{ |
|
"epoch": 0.6348547717842323, |
|
"grad_norm": 2.505421167742011, |
|
"learning_rate": 3.540565619860906e-06, |
|
"loss": 0.4149, |
|
"step": 1530 |
|
}, |
|
{ |
|
"epoch": 0.6369294605809128, |
|
"grad_norm": 2.3662854321031994, |
|
"learning_rate": 3.5059708923533354e-06, |
|
"loss": 0.4092, |
|
"step": 1535 |
|
}, |
|
{ |
|
"epoch": 0.6390041493775933, |
|
"grad_norm": 2.33810078459921, |
|
"learning_rate": 3.4714545218476727e-06, |
|
"loss": 0.4022, |
|
"step": 1540 |
|
}, |
|
{ |
|
"epoch": 0.6410788381742739, |
|
"grad_norm": 2.292723665647368, |
|
"learning_rate": 3.437018318616084e-06, |
|
"loss": 0.3924, |
|
"step": 1545 |
|
}, |
|
{ |
|
"epoch": 0.6431535269709544, |
|
"grad_norm": 2.3059685019935343, |
|
"learning_rate": 3.4026640887262196e-06, |
|
"loss": 0.4062, |
|
"step": 1550 |
|
}, |
|
{ |
|
"epoch": 0.6452282157676349, |
|
"grad_norm": 2.4576342501679966, |
|
"learning_rate": 3.3683936339464957e-06, |
|
"loss": 0.408, |
|
"step": 1555 |
|
}, |
|
{ |
|
"epoch": 0.6473029045643154, |
|
"grad_norm": 2.2902363586226624, |
|
"learning_rate": 3.334208751651593e-06, |
|
"loss": 0.4123, |
|
"step": 1560 |
|
}, |
|
{ |
|
"epoch": 0.6493775933609959, |
|
"grad_norm": 2.2743147574529328, |
|
"learning_rate": 3.300111234728191e-06, |
|
"loss": 0.4097, |
|
"step": 1565 |
|
}, |
|
{ |
|
"epoch": 0.6514522821576764, |
|
"grad_norm": 2.4569598537485327, |
|
"learning_rate": 3.2661028714809405e-06, |
|
"loss": 0.4012, |
|
"step": 1570 |
|
}, |
|
{ |
|
"epoch": 0.6535269709543569, |
|
"grad_norm": 2.28227869578094, |
|
"learning_rate": 3.2321854455386657e-06, |
|
"loss": 0.4008, |
|
"step": 1575 |
|
}, |
|
{ |
|
"epoch": 0.6556016597510373, |
|
"grad_norm": 2.4708407304872533, |
|
"learning_rate": 3.198360735760827e-06, |
|
"loss": 0.3873, |
|
"step": 1580 |
|
}, |
|
{ |
|
"epoch": 0.6576763485477178, |
|
"grad_norm": 2.2707165698258103, |
|
"learning_rate": 3.1646305161442183e-06, |
|
"loss": 0.4059, |
|
"step": 1585 |
|
}, |
|
{ |
|
"epoch": 0.6597510373443983, |
|
"grad_norm": 2.2036975566545993, |
|
"learning_rate": 3.1309965557299303e-06, |
|
"loss": 0.3878, |
|
"step": 1590 |
|
}, |
|
{ |
|
"epoch": 0.6618257261410788, |
|
"grad_norm": 2.4960520314367627, |
|
"learning_rate": 3.097460618510571e-06, |
|
"loss": 0.3885, |
|
"step": 1595 |
|
}, |
|
{ |
|
"epoch": 0.6639004149377593, |
|
"grad_norm": 2.2804809511371755, |
|
"learning_rate": 3.064024463337747e-06, |
|
"loss": 0.4002, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 0.6659751037344398, |
|
"grad_norm": 2.6325297095462767, |
|
"learning_rate": 3.0306898438298184e-06, |
|
"loss": 0.4003, |
|
"step": 1605 |
|
}, |
|
{ |
|
"epoch": 0.6680497925311203, |
|
"grad_norm": 2.366691569211348, |
|
"learning_rate": 2.997458508279928e-06, |
|
"loss": 0.3941, |
|
"step": 1610 |
|
}, |
|
{ |
|
"epoch": 0.6701244813278008, |
|
"grad_norm": 2.2056640704322494, |
|
"learning_rate": 2.964332199564309e-06, |
|
"loss": 0.3966, |
|
"step": 1615 |
|
}, |
|
{ |
|
"epoch": 0.6721991701244814, |
|
"grad_norm": 2.423642247816294, |
|
"learning_rate": 2.9313126550508762e-06, |
|
"loss": 0.396, |
|
"step": 1620 |
|
}, |
|
{ |
|
"epoch": 0.6742738589211619, |
|
"grad_norm": 2.433390053325066, |
|
"learning_rate": 2.8984016065081073e-06, |
|
"loss": 0.4023, |
|
"step": 1625 |
|
}, |
|
{ |
|
"epoch": 0.6763485477178424, |
|
"grad_norm": 2.357419717448147, |
|
"learning_rate": 2.865600780014216e-06, |
|
"loss": 0.4088, |
|
"step": 1630 |
|
}, |
|
{ |
|
"epoch": 0.6784232365145229, |
|
"grad_norm": 2.245544606902691, |
|
"learning_rate": 2.8329118958666236e-06, |
|
"loss": 0.3858, |
|
"step": 1635 |
|
}, |
|
{ |
|
"epoch": 0.6804979253112033, |
|
"grad_norm": 2.439298195310241, |
|
"learning_rate": 2.80033666849174e-06, |
|
"loss": 0.3986, |
|
"step": 1640 |
|
}, |
|
{ |
|
"epoch": 0.6825726141078838, |
|
"grad_norm": 2.4100345874144518, |
|
"learning_rate": 2.7678768063550454e-06, |
|
"loss": 0.3946, |
|
"step": 1645 |
|
}, |
|
{ |
|
"epoch": 0.6846473029045643, |
|
"grad_norm": 2.3584382118069973, |
|
"learning_rate": 2.735534011871479e-06, |
|
"loss": 0.378, |
|
"step": 1650 |
|
}, |
|
{ |
|
"epoch": 0.6867219917012448, |
|
"grad_norm": 2.5751863977070903, |
|
"learning_rate": 2.7033099813161696e-06, |
|
"loss": 0.3901, |
|
"step": 1655 |
|
}, |
|
{ |
|
"epoch": 0.6887966804979253, |
|
"grad_norm": 2.2159484402883827, |
|
"learning_rate": 2.6712064047354515e-06, |
|
"loss": 0.393, |
|
"step": 1660 |
|
}, |
|
{ |
|
"epoch": 0.6908713692946058, |
|
"grad_norm": 2.3569350609642257, |
|
"learning_rate": 2.6392249658582454e-06, |
|
"loss": 0.3818, |
|
"step": 1665 |
|
}, |
|
{ |
|
"epoch": 0.6929460580912863, |
|
"grad_norm": 2.476100739673621, |
|
"learning_rate": 2.607367342007738e-06, |
|
"loss": 0.387, |
|
"step": 1670 |
|
}, |
|
{ |
|
"epoch": 0.6950207468879668, |
|
"grad_norm": 2.394265629557479, |
|
"learning_rate": 2.5756352040134193e-06, |
|
"loss": 0.3893, |
|
"step": 1675 |
|
}, |
|
{ |
|
"epoch": 0.6970954356846473, |
|
"grad_norm": 2.4572385167590234, |
|
"learning_rate": 2.5440302161234542e-06, |
|
"loss": 0.3962, |
|
"step": 1680 |
|
}, |
|
{ |
|
"epoch": 0.6991701244813278, |
|
"grad_norm": 2.3166108869604387, |
|
"learning_rate": 2.5125540359173893e-06, |
|
"loss": 0.3935, |
|
"step": 1685 |
|
}, |
|
{ |
|
"epoch": 0.7012448132780082, |
|
"grad_norm": 2.446525027779214, |
|
"learning_rate": 2.481208314219233e-06, |
|
"loss": 0.3882, |
|
"step": 1690 |
|
}, |
|
{ |
|
"epoch": 0.7033195020746889, |
|
"grad_norm": 2.284958027056727, |
|
"learning_rate": 2.449994695010856e-06, |
|
"loss": 0.3758, |
|
"step": 1695 |
|
}, |
|
{ |
|
"epoch": 0.7053941908713693, |
|
"grad_norm": 2.516170194826004, |
|
"learning_rate": 2.4189148153457875e-06, |
|
"loss": 0.3882, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 0.7074688796680498, |
|
"grad_norm": 2.3397843827362914, |
|
"learning_rate": 2.387970305263349e-06, |
|
"loss": 0.3887, |
|
"step": 1705 |
|
}, |
|
{ |
|
"epoch": 0.7095435684647303, |
|
"grad_norm": 2.3903105786705123, |
|
"learning_rate": 2.3571627877031596e-06, |
|
"loss": 0.3785, |
|
"step": 1710 |
|
}, |
|
{ |
|
"epoch": 0.7116182572614108, |
|
"grad_norm": 2.4388910192687523, |
|
"learning_rate": 2.326493878420028e-06, |
|
"loss": 0.3704, |
|
"step": 1715 |
|
}, |
|
{ |
|
"epoch": 0.7136929460580913, |
|
"grad_norm": 2.3725455021448245, |
|
"learning_rate": 2.295965185899205e-06, |
|
"loss": 0.3778, |
|
"step": 1720 |
|
}, |
|
{ |
|
"epoch": 0.7157676348547718, |
|
"grad_norm": 2.3541520314694946, |
|
"learning_rate": 2.265578311272021e-06, |
|
"loss": 0.3827, |
|
"step": 1725 |
|
}, |
|
{ |
|
"epoch": 0.7178423236514523, |
|
"grad_norm": 2.318059450682999, |
|
"learning_rate": 2.2353348482319233e-06, |
|
"loss": 0.3641, |
|
"step": 1730 |
|
}, |
|
{ |
|
"epoch": 0.7199170124481328, |
|
"grad_norm": 2.303956633945153, |
|
"learning_rate": 2.2052363829508776e-06, |
|
"loss": 0.3804, |
|
"step": 1735 |
|
}, |
|
{ |
|
"epoch": 0.7219917012448133, |
|
"grad_norm": 2.3145838554705547, |
|
"learning_rate": 2.1752844939961926e-06, |
|
"loss": 0.3747, |
|
"step": 1740 |
|
}, |
|
{ |
|
"epoch": 0.7240663900414938, |
|
"grad_norm": 2.2586817786943603, |
|
"learning_rate": 2.1454807522477128e-06, |
|
"loss": 0.3722, |
|
"step": 1745 |
|
}, |
|
{ |
|
"epoch": 0.7261410788381742, |
|
"grad_norm": 2.4038449021090256, |
|
"learning_rate": 2.1158267208154497e-06, |
|
"loss": 0.3807, |
|
"step": 1750 |
|
}, |
|
{ |
|
"epoch": 0.7282157676348547, |
|
"grad_norm": 2.2154594010619166, |
|
"learning_rate": 2.0863239549575865e-06, |
|
"loss": 0.3763, |
|
"step": 1755 |
|
}, |
|
{ |
|
"epoch": 0.7302904564315352, |
|
"grad_norm": 2.5107957004825976, |
|
"learning_rate": 2.0569740019989136e-06, |
|
"loss": 0.3752, |
|
"step": 1760 |
|
}, |
|
{ |
|
"epoch": 0.7323651452282157, |
|
"grad_norm": 2.350354806134385, |
|
"learning_rate": 2.0277784012496865e-06, |
|
"loss": 0.3819, |
|
"step": 1765 |
|
}, |
|
{ |
|
"epoch": 0.7344398340248963, |
|
"grad_norm": 2.4464933220936125, |
|
"learning_rate": 1.998738683924875e-06, |
|
"loss": 0.3628, |
|
"step": 1770 |
|
}, |
|
{ |
|
"epoch": 0.7365145228215768, |
|
"grad_norm": 2.5270841137858064, |
|
"learning_rate": 1.9698563730638794e-06, |
|
"loss": 0.3754, |
|
"step": 1775 |
|
}, |
|
{ |
|
"epoch": 0.7385892116182573, |
|
"grad_norm": 2.250774099933747, |
|
"learning_rate": 1.9411329834506286e-06, |
|
"loss": 0.3667, |
|
"step": 1780 |
|
}, |
|
{ |
|
"epoch": 0.7406639004149378, |
|
"grad_norm": 2.3407235003652804, |
|
"learning_rate": 1.9125700215341476e-06, |
|
"loss": 0.3735, |
|
"step": 1785 |
|
}, |
|
{ |
|
"epoch": 0.7427385892116183, |
|
"grad_norm": 2.3446167395429347, |
|
"learning_rate": 1.8841689853495516e-06, |
|
"loss": 0.3723, |
|
"step": 1790 |
|
}, |
|
{ |
|
"epoch": 0.7448132780082988, |
|
"grad_norm": 2.3350133034877656, |
|
"learning_rate": 1.8559313644394677e-06, |
|
"loss": 0.3785, |
|
"step": 1795 |
|
}, |
|
{ |
|
"epoch": 0.7468879668049793, |
|
"grad_norm": 2.37500798404668, |
|
"learning_rate": 1.827858639775925e-06, |
|
"loss": 0.3786, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 0.7489626556016598, |
|
"grad_norm": 2.4392836582028945, |
|
"learning_rate": 1.7999522836826744e-06, |
|
"loss": 0.3736, |
|
"step": 1805 |
|
}, |
|
{ |
|
"epoch": 0.7510373443983402, |
|
"grad_norm": 2.408718217237952, |
|
"learning_rate": 1.7722137597579698e-06, |
|
"loss": 0.3648, |
|
"step": 1810 |
|
}, |
|
{ |
|
"epoch": 0.7531120331950207, |
|
"grad_norm": 2.4393541244678216, |
|
"learning_rate": 1.744644522797817e-06, |
|
"loss": 0.3556, |
|
"step": 1815 |
|
}, |
|
{ |
|
"epoch": 0.7551867219917012, |
|
"grad_norm": 2.360665467065917, |
|
"learning_rate": 1.7172460187196588e-06, |
|
"loss": 0.3754, |
|
"step": 1820 |
|
}, |
|
{ |
|
"epoch": 0.7572614107883817, |
|
"grad_norm": 2.3486896399291, |
|
"learning_rate": 1.6900196844865575e-06, |
|
"loss": 0.3615, |
|
"step": 1825 |
|
}, |
|
{ |
|
"epoch": 0.7593360995850622, |
|
"grad_norm": 2.330649947928342, |
|
"learning_rate": 1.6629669480318166e-06, |
|
"loss": 0.3763, |
|
"step": 1830 |
|
}, |
|
{ |
|
"epoch": 0.7614107883817427, |
|
"grad_norm": 2.3447118121017287, |
|
"learning_rate": 1.6360892281841007e-06, |
|
"loss": 0.3639, |
|
"step": 1835 |
|
}, |
|
{ |
|
"epoch": 0.7634854771784232, |
|
"grad_norm": 2.507491155826583, |
|
"learning_rate": 1.609387934593019e-06, |
|
"loss": 0.3779, |
|
"step": 1840 |
|
}, |
|
{ |
|
"epoch": 0.7655601659751037, |
|
"grad_norm": 2.288411757086649, |
|
"learning_rate": 1.5828644676551892e-06, |
|
"loss": 0.366, |
|
"step": 1845 |
|
}, |
|
{ |
|
"epoch": 0.7676348547717843, |
|
"grad_norm": 2.3840080652247826, |
|
"learning_rate": 1.5565202184408e-06, |
|
"loss": 0.3589, |
|
"step": 1850 |
|
}, |
|
{ |
|
"epoch": 0.7697095435684648, |
|
"grad_norm": 2.4900677150816364, |
|
"learning_rate": 1.5303565686206452e-06, |
|
"loss": 0.3683, |
|
"step": 1855 |
|
}, |
|
{ |
|
"epoch": 0.7717842323651453, |
|
"grad_norm": 2.5276222084732067, |
|
"learning_rate": 1.5043748903936672e-06, |
|
"loss": 0.3725, |
|
"step": 1860 |
|
}, |
|
{ |
|
"epoch": 0.7738589211618258, |
|
"grad_norm": 2.406387332540463, |
|
"learning_rate": 1.4785765464149836e-06, |
|
"loss": 0.3511, |
|
"step": 1865 |
|
}, |
|
{ |
|
"epoch": 0.7759336099585062, |
|
"grad_norm": 2.409678473099456, |
|
"learning_rate": 1.4529628897244214e-06, |
|
"loss": 0.3637, |
|
"step": 1870 |
|
}, |
|
{ |
|
"epoch": 0.7780082987551867, |
|
"grad_norm": 2.372236015081896, |
|
"learning_rate": 1.4275352636755613e-06, |
|
"loss": 0.3693, |
|
"step": 1875 |
|
}, |
|
{ |
|
"epoch": 0.7800829875518672, |
|
"grad_norm": 2.3018141960934013, |
|
"learning_rate": 1.4022950018652703e-06, |
|
"loss": 0.3557, |
|
"step": 1880 |
|
}, |
|
{ |
|
"epoch": 0.7821576763485477, |
|
"grad_norm": 2.3440334623732375, |
|
"learning_rate": 1.3772434280637737e-06, |
|
"loss": 0.3669, |
|
"step": 1885 |
|
}, |
|
{ |
|
"epoch": 0.7842323651452282, |
|
"grad_norm": 2.4523078746497755, |
|
"learning_rate": 1.3523818561452145e-06, |
|
"loss": 0.3646, |
|
"step": 1890 |
|
}, |
|
{ |
|
"epoch": 0.7863070539419087, |
|
"grad_norm": 2.4219074003538346, |
|
"learning_rate": 1.3277115900187515e-06, |
|
"loss": 0.3577, |
|
"step": 1895 |
|
}, |
|
{ |
|
"epoch": 0.7883817427385892, |
|
"grad_norm": 2.8682882199437842, |
|
"learning_rate": 1.3032339235601749e-06, |
|
"loss": 0.3474, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 0.7904564315352697, |
|
"grad_norm": 2.329298210972444, |
|
"learning_rate": 1.2789501405440397e-06, |
|
"loss": 0.3725, |
|
"step": 1905 |
|
}, |
|
{ |
|
"epoch": 0.7925311203319502, |
|
"grad_norm": 2.365427473253902, |
|
"learning_rate": 1.254861514576348e-06, |
|
"loss": 0.3705, |
|
"step": 1910 |
|
}, |
|
{ |
|
"epoch": 0.7946058091286307, |
|
"grad_norm": 2.449442819984942, |
|
"learning_rate": 1.2309693090277392e-06, |
|
"loss": 0.3498, |
|
"step": 1915 |
|
}, |
|
{ |
|
"epoch": 0.7966804979253111, |
|
"grad_norm": 2.2309766163591362, |
|
"learning_rate": 1.207274776967235e-06, |
|
"loss": 0.3497, |
|
"step": 1920 |
|
}, |
|
{ |
|
"epoch": 0.7987551867219918, |
|
"grad_norm": 2.5332149638146606, |
|
"learning_rate": 1.183779161096526e-06, |
|
"loss": 0.3533, |
|
"step": 1925 |
|
}, |
|
{ |
|
"epoch": 0.8008298755186722, |
|
"grad_norm": 2.464873395759963, |
|
"learning_rate": 1.1604836936847852e-06, |
|
"loss": 0.3669, |
|
"step": 1930 |
|
}, |
|
{ |
|
"epoch": 0.8029045643153527, |
|
"grad_norm": 2.353844278985767, |
|
"learning_rate": 1.1373895965040522e-06, |
|
"loss": 0.3531, |
|
"step": 1935 |
|
}, |
|
{ |
|
"epoch": 0.8049792531120332, |
|
"grad_norm": 2.478326080734339, |
|
"learning_rate": 1.1144980807651413e-06, |
|
"loss": 0.3469, |
|
"step": 1940 |
|
}, |
|
{ |
|
"epoch": 0.8070539419087137, |
|
"grad_norm": 2.3657798765264837, |
|
"learning_rate": 1.0918103470541297e-06, |
|
"loss": 0.3434, |
|
"step": 1945 |
|
}, |
|
{ |
|
"epoch": 0.8091286307053942, |
|
"grad_norm": 2.4345052899894517, |
|
"learning_rate": 1.069327585269383e-06, |
|
"loss": 0.3575, |
|
"step": 1950 |
|
}, |
|
{ |
|
"epoch": 0.8112033195020747, |
|
"grad_norm": 2.335213511709306, |
|
"learning_rate": 1.047050974559149e-06, |
|
"loss": 0.3542, |
|
"step": 1955 |
|
}, |
|
{ |
|
"epoch": 0.8132780082987552, |
|
"grad_norm": 2.3666035829182164, |
|
"learning_rate": 1.024981683259723e-06, |
|
"loss": 0.3484, |
|
"step": 1960 |
|
}, |
|
{ |
|
"epoch": 0.8153526970954357, |
|
"grad_norm": 2.461171014362592, |
|
"learning_rate": 1.0031208688341609e-06, |
|
"loss": 0.3484, |
|
"step": 1965 |
|
}, |
|
{ |
|
"epoch": 0.8174273858921162, |
|
"grad_norm": 2.5462843981961223, |
|
"learning_rate": 9.814696778115834e-07, |
|
"loss": 0.3466, |
|
"step": 1970 |
|
}, |
|
{ |
|
"epoch": 0.8195020746887967, |
|
"grad_norm": 2.3821519258481705, |
|
"learning_rate": 9.600292457270416e-07, |
|
"loss": 0.3502, |
|
"step": 1975 |
|
}, |
|
{ |
|
"epoch": 0.8215767634854771, |
|
"grad_norm": 2.397078021612525, |
|
"learning_rate": 9.388006970619557e-07, |
|
"loss": 0.3518, |
|
"step": 1980 |
|
}, |
|
{ |
|
"epoch": 0.8236514522821576, |
|
"grad_norm": 2.3345715923241945, |
|
"learning_rate": 9.177851451851511e-07, |
|
"loss": 0.3483, |
|
"step": 1985 |
|
}, |
|
{ |
|
"epoch": 0.8257261410788381, |
|
"grad_norm": 2.363653449360418, |
|
"learning_rate": 8.969836922944536e-07, |
|
"loss": 0.3414, |
|
"step": 1990 |
|
}, |
|
{ |
|
"epoch": 0.8278008298755186, |
|
"grad_norm": 2.3624129572192616, |
|
"learning_rate": 8.76397429358895e-07, |
|
"loss": 0.359, |
|
"step": 1995 |
|
}, |
|
{ |
|
"epoch": 0.8298755186721992, |
|
"grad_norm": 2.314328813522686, |
|
"learning_rate": 8.560274360614829e-07, |
|
"loss": 0.3385, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 0.8319502074688797, |
|
"grad_norm": 2.199340475549847, |
|
"learning_rate": 8.358747807425827e-07, |
|
"loss": 0.3539, |
|
"step": 2005 |
|
}, |
|
{ |
|
"epoch": 0.8340248962655602, |
|
"grad_norm": 2.4399680516899633, |
|
"learning_rate": 8.159405203438875e-07, |
|
"loss": 0.3518, |
|
"step": 2010 |
|
}, |
|
{ |
|
"epoch": 0.8360995850622407, |
|
"grad_norm": 2.4262479785068733, |
|
"learning_rate": 7.962257003529777e-07, |
|
"loss": 0.344, |
|
"step": 2015 |
|
}, |
|
{ |
|
"epoch": 0.8381742738589212, |
|
"grad_norm": 2.3545302681837987, |
|
"learning_rate": 7.767313547484979e-07, |
|
"loss": 0.3487, |
|
"step": 2020 |
|
}, |
|
{ |
|
"epoch": 0.8402489626556017, |
|
"grad_norm": 2.3064171274116747, |
|
"learning_rate": 7.574585059459171e-07, |
|
"loss": 0.3447, |
|
"step": 2025 |
|
}, |
|
{ |
|
"epoch": 0.8423236514522822, |
|
"grad_norm": 2.5487816431519477, |
|
"learning_rate": 7.384081647439184e-07, |
|
"loss": 0.356, |
|
"step": 2030 |
|
}, |
|
{ |
|
"epoch": 0.8443983402489627, |
|
"grad_norm": 2.459116013365039, |
|
"learning_rate": 7.195813302713761e-07, |
|
"loss": 0.3375, |
|
"step": 2035 |
|
}, |
|
{ |
|
"epoch": 0.8464730290456431, |
|
"grad_norm": 2.4240198038099905, |
|
"learning_rate": 7.009789899349567e-07, |
|
"loss": 0.3427, |
|
"step": 2040 |
|
}, |
|
{ |
|
"epoch": 0.8485477178423236, |
|
"grad_norm": 2.339613272043201, |
|
"learning_rate": 6.826021193673416e-07, |
|
"loss": 0.3481, |
|
"step": 2045 |
|
}, |
|
{ |
|
"epoch": 0.8506224066390041, |
|
"grad_norm": 2.4864055871608484, |
|
"learning_rate": 6.644516823760439e-07, |
|
"loss": 0.3488, |
|
"step": 2050 |
|
}, |
|
{ |
|
"epoch": 0.8526970954356846, |
|
"grad_norm": 2.3920126614464254, |
|
"learning_rate": 6.465286308928676e-07, |
|
"loss": 0.3427, |
|
"step": 2055 |
|
}, |
|
{ |
|
"epoch": 0.8547717842323651, |
|
"grad_norm": 2.432956486348477, |
|
"learning_rate": 6.28833904923985e-07, |
|
"loss": 0.3522, |
|
"step": 2060 |
|
}, |
|
{ |
|
"epoch": 0.8568464730290456, |
|
"grad_norm": 2.4012672614414243, |
|
"learning_rate": 6.113684325006286e-07, |
|
"loss": 0.3426, |
|
"step": 2065 |
|
}, |
|
{ |
|
"epoch": 0.8589211618257261, |
|
"grad_norm": 2.4925509857922656, |
|
"learning_rate": 5.941331296304254e-07, |
|
"loss": 0.3428, |
|
"step": 2070 |
|
}, |
|
{ |
|
"epoch": 0.8609958506224067, |
|
"grad_norm": 2.2832723407803606, |
|
"learning_rate": 5.771289002493502e-07, |
|
"loss": 0.3472, |
|
"step": 2075 |
|
}, |
|
{ |
|
"epoch": 0.8630705394190872, |
|
"grad_norm": 2.3874716687155564, |
|
"learning_rate": 5.603566361743229e-07, |
|
"loss": 0.3413, |
|
"step": 2080 |
|
}, |
|
{ |
|
"epoch": 0.8651452282157677, |
|
"grad_norm": 2.351570855990709, |
|
"learning_rate": 5.438172170564288e-07, |
|
"loss": 0.3322, |
|
"step": 2085 |
|
}, |
|
{ |
|
"epoch": 0.8672199170124482, |
|
"grad_norm": 2.5008832392383207, |
|
"learning_rate": 5.275115103347884e-07, |
|
"loss": 0.3477, |
|
"step": 2090 |
|
}, |
|
{ |
|
"epoch": 0.8692946058091287, |
|
"grad_norm": 2.3216651986422088, |
|
"learning_rate": 5.114403711910631e-07, |
|
"loss": 0.336, |
|
"step": 2095 |
|
}, |
|
{ |
|
"epoch": 0.8713692946058091, |
|
"grad_norm": 2.326035363037846, |
|
"learning_rate": 4.956046425045991e-07, |
|
"loss": 0.3439, |
|
"step": 2100 |
|
}, |
|
{ |
|
"epoch": 0.8734439834024896, |
|
"grad_norm": 2.4613685736167557, |
|
"learning_rate": 4.80005154808228e-07, |
|
"loss": 0.3468, |
|
"step": 2105 |
|
}, |
|
{ |
|
"epoch": 0.8755186721991701, |
|
"grad_norm": 2.3791168252802537, |
|
"learning_rate": 4.646427262447034e-07, |
|
"loss": 0.3572, |
|
"step": 2110 |
|
}, |
|
{ |
|
"epoch": 0.8775933609958506, |
|
"grad_norm": 2.437202584127699, |
|
"learning_rate": 4.49518162523791e-07, |
|
"loss": 0.3462, |
|
"step": 2115 |
|
}, |
|
{ |
|
"epoch": 0.8796680497925311, |
|
"grad_norm": 2.4764470732059096, |
|
"learning_rate": 4.34632256880016e-07, |
|
"loss": 0.3445, |
|
"step": 2120 |
|
}, |
|
{ |
|
"epoch": 0.8817427385892116, |
|
"grad_norm": 2.3787585521250945, |
|
"learning_rate": 4.1998579003105553e-07, |
|
"loss": 0.3401, |
|
"step": 2125 |
|
}, |
|
{ |
|
"epoch": 0.8838174273858921, |
|
"grad_norm": 2.3948308594804346, |
|
"learning_rate": 4.0557953013679873e-07, |
|
"loss": 0.3536, |
|
"step": 2130 |
|
}, |
|
{ |
|
"epoch": 0.8858921161825726, |
|
"grad_norm": 2.3397934739178363, |
|
"learning_rate": 3.914142327590531e-07, |
|
"loss": 0.3334, |
|
"step": 2135 |
|
}, |
|
{ |
|
"epoch": 0.8879668049792531, |
|
"grad_norm": 2.45922644953553, |
|
"learning_rate": 3.7749064082191976e-07, |
|
"loss": 0.3349, |
|
"step": 2140 |
|
}, |
|
{ |
|
"epoch": 0.8900414937759336, |
|
"grad_norm": 2.413840835433163, |
|
"learning_rate": 3.6380948457283293e-07, |
|
"loss": 0.3315, |
|
"step": 2145 |
|
}, |
|
{ |
|
"epoch": 0.8921161825726142, |
|
"grad_norm": 2.347524977176241, |
|
"learning_rate": 3.5037148154425494e-07, |
|
"loss": 0.346, |
|
"step": 2150 |
|
}, |
|
{ |
|
"epoch": 0.8941908713692946, |
|
"grad_norm": 2.47677069324449, |
|
"learning_rate": 3.3717733651604967e-07, |
|
"loss": 0.3415, |
|
"step": 2155 |
|
}, |
|
{ |
|
"epoch": 0.8962655601659751, |
|
"grad_norm": 2.397745011270569, |
|
"learning_rate": 3.242277414785128e-07, |
|
"loss": 0.3384, |
|
"step": 2160 |
|
}, |
|
{ |
|
"epoch": 0.8983402489626556, |
|
"grad_norm": 2.415367296859196, |
|
"learning_rate": 3.1152337559608725e-07, |
|
"loss": 0.3387, |
|
"step": 2165 |
|
}, |
|
{ |
|
"epoch": 0.9004149377593361, |
|
"grad_norm": 2.23801113438236, |
|
"learning_rate": 2.990649051717348e-07, |
|
"loss": 0.3458, |
|
"step": 2170 |
|
}, |
|
{ |
|
"epoch": 0.9024896265560166, |
|
"grad_norm": 2.4178418849940413, |
|
"learning_rate": 2.868529836119954e-07, |
|
"loss": 0.3347, |
|
"step": 2175 |
|
}, |
|
{ |
|
"epoch": 0.9045643153526971, |
|
"grad_norm": 2.5082956753743204, |
|
"learning_rate": 2.7488825139272037e-07, |
|
"loss": 0.3368, |
|
"step": 2180 |
|
}, |
|
{ |
|
"epoch": 0.9066390041493776, |
|
"grad_norm": 2.2379502275244194, |
|
"learning_rate": 2.631713360254734e-07, |
|
"loss": 0.3436, |
|
"step": 2185 |
|
}, |
|
{ |
|
"epoch": 0.9087136929460581, |
|
"grad_norm": 2.424932946521832, |
|
"learning_rate": 2.5170285202462895e-07, |
|
"loss": 0.3446, |
|
"step": 2190 |
|
}, |
|
{ |
|
"epoch": 0.9107883817427386, |
|
"grad_norm": 2.4571108685084107, |
|
"learning_rate": 2.4048340087513665e-07, |
|
"loss": 0.336, |
|
"step": 2195 |
|
}, |
|
{ |
|
"epoch": 0.9128630705394191, |
|
"grad_norm": 2.3387656990162395, |
|
"learning_rate": 2.295135710009755e-07, |
|
"loss": 0.3414, |
|
"step": 2200 |
|
}, |
|
{ |
|
"epoch": 0.9149377593360996, |
|
"grad_norm": 2.3291759526847113, |
|
"learning_rate": 2.1879393773429903e-07, |
|
"loss": 0.3442, |
|
"step": 2205 |
|
}, |
|
{ |
|
"epoch": 0.91701244813278, |
|
"grad_norm": 2.365416524223355, |
|
"learning_rate": 2.0832506328525237e-07, |
|
"loss": 0.3317, |
|
"step": 2210 |
|
}, |
|
{ |
|
"epoch": 0.9190871369294605, |
|
"grad_norm": 2.384106710657764, |
|
"learning_rate": 1.9810749671249353e-07, |
|
"loss": 0.347, |
|
"step": 2215 |
|
}, |
|
{ |
|
"epoch": 0.921161825726141, |
|
"grad_norm": 2.380728030255305, |
|
"learning_rate": 1.8814177389439304e-07, |
|
"loss": 0.3466, |
|
"step": 2220 |
|
}, |
|
{ |
|
"epoch": 0.9232365145228216, |
|
"grad_norm": 2.297466874753593, |
|
"learning_rate": 1.784284175009282e-07, |
|
"loss": 0.3368, |
|
"step": 2225 |
|
}, |
|
{ |
|
"epoch": 0.9253112033195021, |
|
"grad_norm": 2.4935459102531685, |
|
"learning_rate": 1.68967936966275e-07, |
|
"loss": 0.3404, |
|
"step": 2230 |
|
}, |
|
{ |
|
"epoch": 0.9273858921161826, |
|
"grad_norm": 2.36033858236717, |
|
"learning_rate": 1.5976082846208552e-07, |
|
"loss": 0.3499, |
|
"step": 2235 |
|
}, |
|
{ |
|
"epoch": 0.9294605809128631, |
|
"grad_norm": 2.4730950951949353, |
|
"learning_rate": 1.5080757487146823e-07, |
|
"loss": 0.3327, |
|
"step": 2240 |
|
}, |
|
{ |
|
"epoch": 0.9315352697095436, |
|
"grad_norm": 2.4704731647631255, |
|
"learning_rate": 1.4210864576365891e-07, |
|
"loss": 0.338, |
|
"step": 2245 |
|
}, |
|
{ |
|
"epoch": 0.9336099585062241, |
|
"grad_norm": 2.3890389217669723, |
|
"learning_rate": 1.3366449736939845e-07, |
|
"loss": 0.3375, |
|
"step": 2250 |
|
}, |
|
{ |
|
"epoch": 0.9356846473029046, |
|
"grad_norm": 2.6765870984768663, |
|
"learning_rate": 1.254755725570006e-07, |
|
"loss": 0.3504, |
|
"step": 2255 |
|
}, |
|
{ |
|
"epoch": 0.9377593360995851, |
|
"grad_norm": 2.3147038405622333, |
|
"learning_rate": 1.1754230080912588e-07, |
|
"loss": 0.3267, |
|
"step": 2260 |
|
}, |
|
{ |
|
"epoch": 0.9398340248962656, |
|
"grad_norm": 2.559635037749535, |
|
"learning_rate": 1.0986509820025915e-07, |
|
"loss": 0.3502, |
|
"step": 2265 |
|
}, |
|
{ |
|
"epoch": 0.941908713692946, |
|
"grad_norm": 2.378006403702019, |
|
"learning_rate": 1.0244436737488373e-07, |
|
"loss": 0.3496, |
|
"step": 2270 |
|
}, |
|
{ |
|
"epoch": 0.9439834024896265, |
|
"grad_norm": 2.4406461268285216, |
|
"learning_rate": 9.528049752636714e-08, |
|
"loss": 0.3292, |
|
"step": 2275 |
|
}, |
|
{ |
|
"epoch": 0.946058091286307, |
|
"grad_norm": 2.3994648318804632, |
|
"learning_rate": 8.837386437654861e-08, |
|
"loss": 0.3369, |
|
"step": 2280 |
|
}, |
|
{ |
|
"epoch": 0.9481327800829875, |
|
"grad_norm": 2.3387687484550583, |
|
"learning_rate": 8.172483015603139e-08, |
|
"loss": 0.3445, |
|
"step": 2285 |
|
}, |
|
{ |
|
"epoch": 0.950207468879668, |
|
"grad_norm": 2.3329266437152243, |
|
"learning_rate": 7.533374358518974e-08, |
|
"loss": 0.3374, |
|
"step": 2290 |
|
}, |
|
{ |
|
"epoch": 0.9522821576763485, |
|
"grad_norm": 2.344823680655971, |
|
"learning_rate": 6.920093985587351e-08, |
|
"loss": 0.3369, |
|
"step": 2295 |
|
}, |
|
{ |
|
"epoch": 0.9543568464730291, |
|
"grad_norm": 2.297823156508915, |
|
"learning_rate": 6.332674061383492e-08, |
|
"loss": 0.3387, |
|
"step": 2300 |
|
}, |
|
{ |
|
"epoch": 0.9564315352697096, |
|
"grad_norm": 2.436614135753504, |
|
"learning_rate": 5.7711453941852736e-08, |
|
"loss": 0.334, |
|
"step": 2305 |
|
}, |
|
{ |
|
"epoch": 0.9585062240663901, |
|
"grad_norm": 2.2863118140565564, |
|
"learning_rate": 5.23553743435784e-08, |
|
"loss": 0.3284, |
|
"step": 2310 |
|
}, |
|
{ |
|
"epoch": 0.9605809128630706, |
|
"grad_norm": 2.1820719536802504, |
|
"learning_rate": 4.7258782728092366e-08, |
|
"loss": 0.3293, |
|
"step": 2315 |
|
}, |
|
{ |
|
"epoch": 0.9626556016597511, |
|
"grad_norm": 2.333916190818195, |
|
"learning_rate": 4.2421946395164174e-08, |
|
"loss": 0.33, |
|
"step": 2320 |
|
}, |
|
{ |
|
"epoch": 0.9647302904564315, |
|
"grad_norm": 2.4479781364843096, |
|
"learning_rate": 3.784511902124033e-08, |
|
"loss": 0.3491, |
|
"step": 2325 |
|
}, |
|
{ |
|
"epoch": 0.966804979253112, |
|
"grad_norm": 2.4249579257495566, |
|
"learning_rate": 3.352854064613553e-08, |
|
"loss": 0.345, |
|
"step": 2330 |
|
}, |
|
{ |
|
"epoch": 0.9688796680497925, |
|
"grad_norm": 2.2853865711169785, |
|
"learning_rate": 2.9472437660446605e-08, |
|
"loss": 0.3358, |
|
"step": 2335 |
|
}, |
|
{ |
|
"epoch": 0.970954356846473, |
|
"grad_norm": 2.4792048595371647, |
|
"learning_rate": 2.56770227936759e-08, |
|
"loss": 0.3365, |
|
"step": 2340 |
|
}, |
|
{ |
|
"epoch": 0.9730290456431535, |
|
"grad_norm": 2.38571031272437, |
|
"learning_rate": 2.2142495103075757e-08, |
|
"loss": 0.3426, |
|
"step": 2345 |
|
}, |
|
{ |
|
"epoch": 0.975103734439834, |
|
"grad_norm": 2.3533072019330095, |
|
"learning_rate": 1.8869039963210766e-08, |
|
"loss": 0.3384, |
|
"step": 2350 |
|
}, |
|
{ |
|
"epoch": 0.9771784232365145, |
|
"grad_norm": 2.42964006589875, |
|
"learning_rate": 1.585682905623054e-08, |
|
"loss": 0.3406, |
|
"step": 2355 |
|
}, |
|
{ |
|
"epoch": 0.979253112033195, |
|
"grad_norm": 2.2589268313436968, |
|
"learning_rate": 1.3106020362869675e-08, |
|
"loss": 0.3379, |
|
"step": 2360 |
|
}, |
|
{ |
|
"epoch": 0.9813278008298755, |
|
"grad_norm": 2.4389590802442203, |
|
"learning_rate": 1.0616758154161633e-08, |
|
"loss": 0.342, |
|
"step": 2365 |
|
}, |
|
{ |
|
"epoch": 0.983402489626556, |
|
"grad_norm": 2.387697736294216, |
|
"learning_rate": 8.38917298386921e-09, |
|
"loss": 0.3414, |
|
"step": 2370 |
|
}, |
|
{ |
|
"epoch": 0.9854771784232366, |
|
"grad_norm": 2.2923054120317836, |
|
"learning_rate": 6.423381681641694e-09, |
|
"loss": 0.3416, |
|
"step": 2375 |
|
}, |
|
{ |
|
"epoch": 0.9875518672199171, |
|
"grad_norm": 2.341070250308901, |
|
"learning_rate": 4.719487346884211e-09, |
|
"loss": 0.3353, |
|
"step": 2380 |
|
}, |
|
{ |
|
"epoch": 0.9896265560165975, |
|
"grad_norm": 2.390677320345825, |
|
"learning_rate": 3.277579343352599e-09, |
|
"loss": 0.3273, |
|
"step": 2385 |
|
}, |
|
{ |
|
"epoch": 0.991701244813278, |
|
"grad_norm": 2.3301588985719905, |
|
"learning_rate": 2.097733294464943e-09, |
|
"loss": 0.355, |
|
"step": 2390 |
|
}, |
|
{ |
|
"epoch": 0.9937759336099585, |
|
"grad_norm": 2.423326896617218, |
|
"learning_rate": 1.1800110793358521e-09, |
|
"loss": 0.3402, |
|
"step": 2395 |
|
}, |
|
{ |
|
"epoch": 0.995850622406639, |
|
"grad_norm": 2.3567177848437986, |
|
"learning_rate": 5.244608295323916e-10, |
|
"loss": 0.3361, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 0.9979253112033195, |
|
"grad_norm": 2.4643918613730165, |
|
"learning_rate": 1.3111692654776963e-10, |
|
"loss": 0.3175, |
|
"step": 2405 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"grad_norm": 2.3790989221416656, |
|
"learning_rate": 0.0, |
|
"loss": 0.3387, |
|
"step": 2410 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_runtime": 3.4735, |
|
"eval_samples_per_second": 2.879, |
|
"eval_steps_per_second": 0.864, |
|
"step": 2410 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"step": 2410, |
|
"total_flos": 252302485094400.0, |
|
"train_loss": 0.0, |
|
"train_runtime": 0.0084, |
|
"train_samples_per_second": 4572412.844, |
|
"train_steps_per_second": 285835.105 |
|
} |
|
], |
|
"logging_steps": 5, |
|
"max_steps": 2410, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 1, |
|
"save_steps": 100, |
|
"stateful_callbacks": { |
|
"TrainerControl": { |
|
"args": { |
|
"should_epoch_stop": false, |
|
"should_evaluate": false, |
|
"should_log": false, |
|
"should_save": true, |
|
"should_training_stop": true |
|
}, |
|
"attributes": {} |
|
} |
|
}, |
|
"total_flos": 252302485094400.0, |
|
"train_batch_size": 4, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|