| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 1857, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0005385029617662897, | |
| "grad_norm": 22.7613713160104, | |
| "learning_rate": 5.376344086021506e-08, | |
| "loss": 1.3472, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0026925148088314485, | |
| "grad_norm": 22.37683485347051, | |
| "learning_rate": 2.688172043010753e-07, | |
| "loss": 1.334, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.005385029617662897, | |
| "grad_norm": 16.198717013287386, | |
| "learning_rate": 5.376344086021506e-07, | |
| "loss": 1.3098, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.008077544426494346, | |
| "grad_norm": 12.105460048613201, | |
| "learning_rate": 8.064516129032258e-07, | |
| "loss": 1.1643, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.010770059235325794, | |
| "grad_norm": 7.514844224376465, | |
| "learning_rate": 1.0752688172043011e-06, | |
| "loss": 1.0474, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.013462574044157244, | |
| "grad_norm": 3.8632002795574367, | |
| "learning_rate": 1.3440860215053765e-06, | |
| "loss": 0.9317, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.01615508885298869, | |
| "grad_norm": 3.400445342173989, | |
| "learning_rate": 1.6129032258064516e-06, | |
| "loss": 0.8826, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.01884760366182014, | |
| "grad_norm": 2.9898366351100547, | |
| "learning_rate": 1.881720430107527e-06, | |
| "loss": 0.8635, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.021540118470651588, | |
| "grad_norm": 2.8585969196090373, | |
| "learning_rate": 2.1505376344086023e-06, | |
| "loss": 0.8577, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.024232633279483037, | |
| "grad_norm": 3.0122772432907277, | |
| "learning_rate": 2.4193548387096776e-06, | |
| "loss": 0.8418, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.026925148088314487, | |
| "grad_norm": 3.0903177647257336, | |
| "learning_rate": 2.688172043010753e-06, | |
| "loss": 0.8311, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.029617662897145933, | |
| "grad_norm": 3.0109701116495216, | |
| "learning_rate": 2.9569892473118283e-06, | |
| "loss": 0.8072, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.03231017770597738, | |
| "grad_norm": 2.9583394058653876, | |
| "learning_rate": 3.225806451612903e-06, | |
| "loss": 0.7965, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03500269251480883, | |
| "grad_norm": 3.1460179996258146, | |
| "learning_rate": 3.494623655913979e-06, | |
| "loss": 0.8118, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.03769520732364028, | |
| "grad_norm": 3.376763980257778, | |
| "learning_rate": 3.763440860215054e-06, | |
| "loss": 0.7851, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.04038772213247173, | |
| "grad_norm": 2.9673425200438013, | |
| "learning_rate": 4.032258064516129e-06, | |
| "loss": 0.7957, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.043080236941303175, | |
| "grad_norm": 3.0128704990376267, | |
| "learning_rate": 4.3010752688172045e-06, | |
| "loss": 0.7751, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.04577275175013463, | |
| "grad_norm": 3.3210132634127656, | |
| "learning_rate": 4.56989247311828e-06, | |
| "loss": 0.7689, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.048465266558966075, | |
| "grad_norm": 3.175041603297429, | |
| "learning_rate": 4.838709677419355e-06, | |
| "loss": 0.7685, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.05115778136779752, | |
| "grad_norm": 3.1666439879487838, | |
| "learning_rate": 5.1075268817204305e-06, | |
| "loss": 0.7642, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.053850296176628974, | |
| "grad_norm": 3.050039288877326, | |
| "learning_rate": 5.376344086021506e-06, | |
| "loss": 0.75, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.05654281098546042, | |
| "grad_norm": 2.9472819168277455, | |
| "learning_rate": 5.645161290322582e-06, | |
| "loss": 0.7437, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.05923532579429187, | |
| "grad_norm": 3.1420868913301185, | |
| "learning_rate": 5.9139784946236566e-06, | |
| "loss": 0.7317, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.06192784060312332, | |
| "grad_norm": 3.2485163047441423, | |
| "learning_rate": 6.182795698924732e-06, | |
| "loss": 0.7392, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.06462035541195477, | |
| "grad_norm": 3.4854700395087552, | |
| "learning_rate": 6.451612903225806e-06, | |
| "loss": 0.7441, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.06731287022078622, | |
| "grad_norm": 2.9119403460101148, | |
| "learning_rate": 6.720430107526882e-06, | |
| "loss": 0.7216, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.07000538502961766, | |
| "grad_norm": 2.956926563215588, | |
| "learning_rate": 6.989247311827958e-06, | |
| "loss": 0.7283, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.07269789983844911, | |
| "grad_norm": 3.256266278148656, | |
| "learning_rate": 7.258064516129033e-06, | |
| "loss": 0.7296, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.07539041464728057, | |
| "grad_norm": 2.8692943178965185, | |
| "learning_rate": 7.526881720430108e-06, | |
| "loss": 0.7202, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.078082929456112, | |
| "grad_norm": 2.8804161059666935, | |
| "learning_rate": 7.795698924731183e-06, | |
| "loss": 0.7195, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.08077544426494346, | |
| "grad_norm": 3.0006510283335848, | |
| "learning_rate": 8.064516129032258e-06, | |
| "loss": 0.721, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.08346795907377491, | |
| "grad_norm": 2.978282671191466, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.7151, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.08616047388260635, | |
| "grad_norm": 2.832558314575903, | |
| "learning_rate": 8.602150537634409e-06, | |
| "loss": 0.712, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.0888529886914378, | |
| "grad_norm": 2.9438520239067265, | |
| "learning_rate": 8.870967741935484e-06, | |
| "loss": 0.7161, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.09154550350026926, | |
| "grad_norm": 3.165458696824085, | |
| "learning_rate": 9.13978494623656e-06, | |
| "loss": 0.7146, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.0942380183091007, | |
| "grad_norm": 3.372895429211009, | |
| "learning_rate": 9.408602150537635e-06, | |
| "loss": 0.7008, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.09693053311793215, | |
| "grad_norm": 2.9822482771935785, | |
| "learning_rate": 9.67741935483871e-06, | |
| "loss": 0.702, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.0996230479267636, | |
| "grad_norm": 2.7297175029131555, | |
| "learning_rate": 9.946236559139786e-06, | |
| "loss": 0.701, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.10231556273559504, | |
| "grad_norm": 2.797512257800779, | |
| "learning_rate": 9.999858614526654e-06, | |
| "loss": 0.7073, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.1050080775444265, | |
| "grad_norm": 2.7494639915648516, | |
| "learning_rate": 9.999284249745143e-06, | |
| "loss": 0.6922, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.10770059235325795, | |
| "grad_norm": 2.8999873055676404, | |
| "learning_rate": 9.998268119778471e-06, | |
| "loss": 0.6998, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.11039310716208939, | |
| "grad_norm": 2.709033463381412, | |
| "learning_rate": 9.996810314417666e-06, | |
| "loss": 0.7017, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.11308562197092084, | |
| "grad_norm": 2.6497064978122498, | |
| "learning_rate": 9.9949109624827e-06, | |
| "loss": 0.697, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.1157781367797523, | |
| "grad_norm": 2.850561924743422, | |
| "learning_rate": 9.992570231811108e-06, | |
| "loss": 0.6937, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.11847065158858373, | |
| "grad_norm": 2.7049716604581953, | |
| "learning_rate": 9.989788329243163e-06, | |
| "loss": 0.7057, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.12116316639741519, | |
| "grad_norm": 2.5917851309791673, | |
| "learning_rate": 9.986565500603598e-06, | |
| "loss": 0.6956, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.12385568120624664, | |
| "grad_norm": 2.6711850256792258, | |
| "learning_rate": 9.982902030679875e-06, | |
| "loss": 0.7109, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.12654819601507808, | |
| "grad_norm": 2.6766440274545995, | |
| "learning_rate": 9.978798243197031e-06, | |
| "loss": 0.6909, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.12924071082390953, | |
| "grad_norm": 2.8552384364271775, | |
| "learning_rate": 9.974254500789065e-06, | |
| "loss": 0.6879, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.13193322563274099, | |
| "grad_norm": 2.5756385146469882, | |
| "learning_rate": 9.969271204966895e-06, | |
| "loss": 0.6802, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.13462574044157244, | |
| "grad_norm": 2.717537618462663, | |
| "learning_rate": 9.963848796082873e-06, | |
| "loss": 0.6921, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.13731825525040386, | |
| "grad_norm": 2.5131980662996756, | |
| "learning_rate": 9.95798775329189e-06, | |
| "loss": 0.689, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.14001077005923532, | |
| "grad_norm": 2.51836898271677, | |
| "learning_rate": 9.951688594509012e-06, | |
| "loss": 0.6793, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.14270328486806677, | |
| "grad_norm": 2.6980273572649676, | |
| "learning_rate": 9.944951876363731e-06, | |
| "loss": 0.6834, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.14539579967689822, | |
| "grad_norm": 2.646134452091696, | |
| "learning_rate": 9.937778194150771e-06, | |
| "loss": 0.6767, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.14808831448572968, | |
| "grad_norm": 2.5534608799808627, | |
| "learning_rate": 9.930168181777486e-06, | |
| "loss": 0.6907, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.15078082929456113, | |
| "grad_norm": 2.5908475307902656, | |
| "learning_rate": 9.922122511707843e-06, | |
| "loss": 0.6861, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.15347334410339256, | |
| "grad_norm": 2.5765130409804695, | |
| "learning_rate": 9.913641894903006e-06, | |
| "loss": 0.6791, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.156165858912224, | |
| "grad_norm": 2.741146958946548, | |
| "learning_rate": 9.904727080758495e-06, | |
| "loss": 0.6874, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.15885837372105546, | |
| "grad_norm": 2.8381749428470595, | |
| "learning_rate": 9.895378857037989e-06, | |
| "loss": 0.676, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.16155088852988692, | |
| "grad_norm": 2.7204348331914447, | |
| "learning_rate": 9.885598049803693e-06, | |
| "loss": 0.673, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.16424340333871837, | |
| "grad_norm": 2.6441432061304178, | |
| "learning_rate": 9.875385523343353e-06, | |
| "loss": 0.671, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.16693591814754982, | |
| "grad_norm": 2.5486118216678277, | |
| "learning_rate": 9.86474218009388e-06, | |
| "loss": 0.6622, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.16962843295638125, | |
| "grad_norm": 2.448370238657585, | |
| "learning_rate": 9.853668960561611e-06, | |
| "loss": 0.6848, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.1723209477652127, | |
| "grad_norm": 2.5840299111800955, | |
| "learning_rate": 9.84216684323919e-06, | |
| "loss": 0.6568, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.17501346257404415, | |
| "grad_norm": 2.57305271825157, | |
| "learning_rate": 9.830236844519114e-06, | |
| "loss": 0.66, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.1777059773828756, | |
| "grad_norm": 2.53266567006045, | |
| "learning_rate": 9.81788001860391e-06, | |
| "loss": 0.662, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.18039849219170706, | |
| "grad_norm": 2.4978159844537458, | |
| "learning_rate": 9.805097457412981e-06, | |
| "loss": 0.6509, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.1830910070005385, | |
| "grad_norm": 2.533762872533261, | |
| "learning_rate": 9.791890290486127e-06, | |
| "loss": 0.6336, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.18578352180936994, | |
| "grad_norm": 2.5709499517182888, | |
| "learning_rate": 9.77825968488372e-06, | |
| "loss": 0.6672, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.1884760366182014, | |
| "grad_norm": 2.5155435537189432, | |
| "learning_rate": 9.76420684508358e-06, | |
| "loss": 0.6587, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.19116855142703285, | |
| "grad_norm": 2.3629608947394214, | |
| "learning_rate": 9.74973301287455e-06, | |
| "loss": 0.6382, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.1938610662358643, | |
| "grad_norm": 2.423886375936418, | |
| "learning_rate": 9.734839467246744e-06, | |
| "loss": 0.6399, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.19655358104469575, | |
| "grad_norm": 2.4151592067478713, | |
| "learning_rate": 9.719527524278552e-06, | |
| "loss": 0.6567, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.1992460958535272, | |
| "grad_norm": 2.339311457038808, | |
| "learning_rate": 9.703798537020327e-06, | |
| "loss": 0.6329, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.20193861066235863, | |
| "grad_norm": 2.401955911603551, | |
| "learning_rate": 9.687653895374824e-06, | |
| "loss": 0.6443, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.20463112547119008, | |
| "grad_norm": 2.349530919733858, | |
| "learning_rate": 9.671095025974388e-06, | |
| "loss": 0.6262, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.20732364028002154, | |
| "grad_norm": 2.5574771679537, | |
| "learning_rate": 9.654123392054884e-06, | |
| "loss": 0.6493, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.210016155088853, | |
| "grad_norm": 2.5279142423497443, | |
| "learning_rate": 9.636740493326398e-06, | |
| "loss": 0.6458, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.21270866989768444, | |
| "grad_norm": 2.407565936871044, | |
| "learning_rate": 9.618947865840705e-06, | |
| "loss": 0.6356, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.2154011847065159, | |
| "grad_norm": 2.50737386034651, | |
| "learning_rate": 9.600747081855553e-06, | |
| "loss": 0.6505, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.21809369951534732, | |
| "grad_norm": 2.492501588167292, | |
| "learning_rate": 9.582139749695713e-06, | |
| "loss": 0.625, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.22078621432417878, | |
| "grad_norm": 2.3085166355816322, | |
| "learning_rate": 9.563127513610868e-06, | |
| "loss": 0.6531, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.22347872913301023, | |
| "grad_norm": 2.3663625111521553, | |
| "learning_rate": 9.543712053630314e-06, | |
| "loss": 0.6185, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.22617124394184168, | |
| "grad_norm": 2.362289550568963, | |
| "learning_rate": 9.523895085414503e-06, | |
| "loss": 0.6151, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.22886375875067314, | |
| "grad_norm": 2.4285748744254394, | |
| "learning_rate": 9.503678360103433e-06, | |
| "loss": 0.6195, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.2315562735595046, | |
| "grad_norm": 2.320249188264771, | |
| "learning_rate": 9.483063664161922e-06, | |
| "loss": 0.6217, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.23424878836833601, | |
| "grad_norm": 2.4068743254230953, | |
| "learning_rate": 9.462052819221726e-06, | |
| "loss": 0.6269, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.23694130317716747, | |
| "grad_norm": 2.443692426314357, | |
| "learning_rate": 9.440647681920587e-06, | |
| "loss": 0.6414, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.23963381798599892, | |
| "grad_norm": 2.376216504990232, | |
| "learning_rate": 9.418850143738158e-06, | |
| "loss": 0.6349, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.24232633279483037, | |
| "grad_norm": 2.4206153603989566, | |
| "learning_rate": 9.396662130828869e-06, | |
| "loss": 0.6059, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.24501884760366183, | |
| "grad_norm": 2.4419798835688074, | |
| "learning_rate": 9.374085603851709e-06, | |
| "loss": 0.6091, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.24771136241249328, | |
| "grad_norm": 2.3263506335508817, | |
| "learning_rate": 9.351122557796987e-06, | |
| "loss": 0.6188, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.25040387722132473, | |
| "grad_norm": 2.533382037846899, | |
| "learning_rate": 9.327775021810037e-06, | |
| "loss": 0.6145, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.25309639203015616, | |
| "grad_norm": 2.678311932146126, | |
| "learning_rate": 9.304045059011905e-06, | |
| "loss": 0.5881, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.25578890683898764, | |
| "grad_norm": 2.5127273745887466, | |
| "learning_rate": 9.279934766317045e-06, | |
| "loss": 0.621, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.25848142164781907, | |
| "grad_norm": 2.303317205670643, | |
| "learning_rate": 9.255446274248023e-06, | |
| "loss": 0.6036, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.2611739364566505, | |
| "grad_norm": 2.4449467721888376, | |
| "learning_rate": 9.230581746747257e-06, | |
| "loss": 0.6035, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.26386645126548197, | |
| "grad_norm": 2.4020663292502165, | |
| "learning_rate": 9.205343380985793e-06, | |
| "loss": 0.5939, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.2665589660743134, | |
| "grad_norm": 2.526561720865294, | |
| "learning_rate": 9.179733407169146e-06, | |
| "loss": 0.6, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.2692514808831449, | |
| "grad_norm": 2.4642667203514783, | |
| "learning_rate": 9.153754088340237e-06, | |
| "loss": 0.6143, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.2719439956919763, | |
| "grad_norm": 2.546297508506232, | |
| "learning_rate": 9.12740772017941e-06, | |
| "loss": 0.6006, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.27463651050080773, | |
| "grad_norm": 2.332663269345773, | |
| "learning_rate": 9.10069663080158e-06, | |
| "loss": 0.5894, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.2773290253096392, | |
| "grad_norm": 2.414500443877397, | |
| "learning_rate": 9.073623180550496e-06, | |
| "loss": 0.5978, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.28002154011847064, | |
| "grad_norm": 2.618196788357666, | |
| "learning_rate": 9.046189761790179e-06, | |
| "loss": 0.5914, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.2827140549273021, | |
| "grad_norm": 2.4760448307330964, | |
| "learning_rate": 9.018398798693512e-06, | |
| "loss": 0.6094, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.28540656973613354, | |
| "grad_norm": 2.403492095012741, | |
| "learning_rate": 8.99025274702803e-06, | |
| "loss": 0.5792, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.288099084544965, | |
| "grad_norm": 2.424533931438387, | |
| "learning_rate": 8.961754093938917e-06, | |
| "loss": 0.5813, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.29079159935379645, | |
| "grad_norm": 2.4025980246295555, | |
| "learning_rate": 8.932905357729213e-06, | |
| "loss": 0.5747, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.2934841141626279, | |
| "grad_norm": 2.4566763739595343, | |
| "learning_rate": 8.903709087637307e-06, | |
| "loss": 0.6003, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.29617662897145935, | |
| "grad_norm": 2.307834128807205, | |
| "learning_rate": 8.874167863611647e-06, | |
| "loss": 0.5838, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.2988691437802908, | |
| "grad_norm": 2.352469567943909, | |
| "learning_rate": 8.844284296082776e-06, | |
| "loss": 0.5845, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.30156165858912226, | |
| "grad_norm": 2.3341023442033895, | |
| "learning_rate": 8.81406102573266e-06, | |
| "loss": 0.5689, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.3042541733979537, | |
| "grad_norm": 2.4151388881341322, | |
| "learning_rate": 8.783500723261333e-06, | |
| "loss": 0.5793, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.3069466882067851, | |
| "grad_norm": 2.352881741468682, | |
| "learning_rate": 8.752606089150903e-06, | |
| "loss": 0.5872, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.3096392030156166, | |
| "grad_norm": 2.2929845371154665, | |
| "learning_rate": 8.721379853426928e-06, | |
| "loss": 0.5827, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.312331717824448, | |
| "grad_norm": 2.351649375339974, | |
| "learning_rate": 8.689824775417164e-06, | |
| "loss": 0.5704, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.3150242326332795, | |
| "grad_norm": 2.3536783488717425, | |
| "learning_rate": 8.657943643507747e-06, | |
| "loss": 0.5916, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.3177167474421109, | |
| "grad_norm": 2.4858567387352783, | |
| "learning_rate": 8.625739274896785e-06, | |
| "loss": 0.5679, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.3204092622509424, | |
| "grad_norm": 2.5420824976461613, | |
| "learning_rate": 8.593214515345426e-06, | |
| "loss": 0.5747, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.32310177705977383, | |
| "grad_norm": 2.411084856535756, | |
| "learning_rate": 8.560372238926372e-06, | |
| "loss": 0.5818, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.32579429186860526, | |
| "grad_norm": 2.493451871387303, | |
| "learning_rate": 8.527215347769932e-06, | |
| "loss": 0.5709, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.32848680667743674, | |
| "grad_norm": 2.2974929968587388, | |
| "learning_rate": 8.493746771807559e-06, | |
| "loss": 0.578, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.33117932148626816, | |
| "grad_norm": 2.368635502736958, | |
| "learning_rate": 8.459969468512943e-06, | |
| "loss": 0.5507, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.33387183629509964, | |
| "grad_norm": 2.5093780273624384, | |
| "learning_rate": 8.425886422640684e-06, | |
| "loss": 0.5574, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.33656435110393107, | |
| "grad_norm": 2.3784956807998254, | |
| "learning_rate": 8.391500645962527e-06, | |
| "loss": 0.5629, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.3392568659127625, | |
| "grad_norm": 2.5705211303648987, | |
| "learning_rate": 8.356815177001243e-06, | |
| "loss": 0.5563, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.341949380721594, | |
| "grad_norm": 2.296265064615366, | |
| "learning_rate": 8.321833080762109e-06, | |
| "loss": 0.5673, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.3446418955304254, | |
| "grad_norm": 2.468686941019328, | |
| "learning_rate": 8.286557448462074e-06, | |
| "loss": 0.5615, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.3473344103392569, | |
| "grad_norm": 2.304658882065399, | |
| "learning_rate": 8.250991397256609e-06, | |
| "loss": 0.5716, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.3500269251480883, | |
| "grad_norm": 2.4328147271748968, | |
| "learning_rate": 8.215138069964248e-06, | |
| "loss": 0.5497, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.3527194399569198, | |
| "grad_norm": 2.5638444164193626, | |
| "learning_rate": 8.179000634788872e-06, | |
| "loss": 0.5634, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.3554119547657512, | |
| "grad_norm": 2.304000568970403, | |
| "learning_rate": 8.142582285039753e-06, | |
| "loss": 0.5552, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.35810446957458264, | |
| "grad_norm": 2.312118564540539, | |
| "learning_rate": 8.105886238849369e-06, | |
| "loss": 0.5665, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.3607969843834141, | |
| "grad_norm": 2.564926504630545, | |
| "learning_rate": 8.068915738889038e-06, | |
| "loss": 0.5432, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.36348949919224555, | |
| "grad_norm": 2.552416968513296, | |
| "learning_rate": 8.031674052082372e-06, | |
| "loss": 0.5418, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.366182014001077, | |
| "grad_norm": 2.465276318426562, | |
| "learning_rate": 7.994164469316596e-06, | |
| "loss": 0.5438, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.36887452880990845, | |
| "grad_norm": 2.3946509022495728, | |
| "learning_rate": 7.956390305151744e-06, | |
| "loss": 0.5399, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.3715670436187399, | |
| "grad_norm": 2.5006069528524266, | |
| "learning_rate": 7.918354897527767e-06, | |
| "loss": 0.5556, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.37425955842757136, | |
| "grad_norm": 2.4646542978549633, | |
| "learning_rate": 7.880061607469574e-06, | |
| "loss": 0.5465, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.3769520732364028, | |
| "grad_norm": 2.464786801902601, | |
| "learning_rate": 7.841513818790035e-06, | |
| "loss": 0.5412, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.37964458804523427, | |
| "grad_norm": 2.470119700617401, | |
| "learning_rate": 7.80271493779096e-06, | |
| "loss": 0.5518, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.3823371028540657, | |
| "grad_norm": 2.487807822690958, | |
| "learning_rate": 7.763668392962107e-06, | |
| "loss": 0.539, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.3850296176628972, | |
| "grad_norm": 2.3449768254321812, | |
| "learning_rate": 7.724377634678222e-06, | |
| "loss": 0.5367, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.3877221324717286, | |
| "grad_norm": 2.342137329316273, | |
| "learning_rate": 7.684846134894133e-06, | |
| "loss": 0.5361, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.39041464728056, | |
| "grad_norm": 2.3478632374387236, | |
| "learning_rate": 7.645077386837963e-06, | |
| "loss": 0.5377, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.3931071620893915, | |
| "grad_norm": 2.311127470477535, | |
| "learning_rate": 7.605074904702438e-06, | |
| "loss": 0.5326, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.39579967689822293, | |
| "grad_norm": 2.5257265487257317, | |
| "learning_rate": 7.5648422233343564e-06, | |
| "loss": 0.5253, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.3984921917070544, | |
| "grad_norm": 2.2883296027896862, | |
| "learning_rate": 7.524382897922229e-06, | |
| "loss": 0.5198, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.40118470651588584, | |
| "grad_norm": 2.2938224585559994, | |
| "learning_rate": 7.483700503682127e-06, | |
| "loss": 0.5179, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.40387722132471726, | |
| "grad_norm": 2.351764729142054, | |
| "learning_rate": 7.442798635541749e-06, | |
| "loss": 0.5245, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.40656973613354874, | |
| "grad_norm": 2.2446519389708057, | |
| "learning_rate": 7.401680907822756e-06, | |
| "loss": 0.5356, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.40926225094238017, | |
| "grad_norm": 2.463243359027525, | |
| "learning_rate": 7.360350953921386e-06, | |
| "loss": 0.5261, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.41195476575121165, | |
| "grad_norm": 2.1930439197029634, | |
| "learning_rate": 7.3188124259873946e-06, | |
| "loss": 0.5401, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.4146472805600431, | |
| "grad_norm": 2.37252598970325, | |
| "learning_rate": 7.2770689946013215e-06, | |
| "loss": 0.5047, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.41733979536887456, | |
| "grad_norm": 2.284338792322131, | |
| "learning_rate": 7.235124348450143e-06, | |
| "loss": 0.5121, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.420032310177706, | |
| "grad_norm": 2.253488965644508, | |
| "learning_rate": 7.192982194001312e-06, | |
| "loss": 0.5029, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.4227248249865374, | |
| "grad_norm": 2.258949760855827, | |
| "learning_rate": 7.150646255175242e-06, | |
| "loss": 0.5282, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.4254173397953689, | |
| "grad_norm": 2.2735662795523317, | |
| "learning_rate": 7.108120273016236e-06, | |
| "loss": 0.514, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.4281098546042003, | |
| "grad_norm": 2.2850305304777865, | |
| "learning_rate": 7.065408005361902e-06, | |
| "loss": 0.5046, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.4308023694130318, | |
| "grad_norm": 2.2431623733294925, | |
| "learning_rate": 7.022513226511098e-06, | |
| "loss": 0.5108, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.4334948842218632, | |
| "grad_norm": 2.3702355406435958, | |
| "learning_rate": 6.9794397268904094e-06, | |
| "loss": 0.5039, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.43618739903069464, | |
| "grad_norm": 2.424641619769436, | |
| "learning_rate": 6.9361913127192026e-06, | |
| "loss": 0.5135, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.4388799138395261, | |
| "grad_norm": 2.417084290942502, | |
| "learning_rate": 6.892771805673289e-06, | |
| "loss": 0.5008, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.44157242864835755, | |
| "grad_norm": 2.483155905595343, | |
| "learning_rate": 6.8491850425472216e-06, | |
| "loss": 0.5061, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.44426494345718903, | |
| "grad_norm": 2.3647966120276624, | |
| "learning_rate": 6.805434874915249e-06, | |
| "loss": 0.5116, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.44695745826602046, | |
| "grad_norm": 2.3941067808468235, | |
| "learning_rate": 6.7615251687909714e-06, | |
| "loss": 0.5115, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.44964997307485194, | |
| "grad_norm": 2.431013960473848, | |
| "learning_rate": 6.7174598042857186e-06, | |
| "loss": 0.4884, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.45234248788368336, | |
| "grad_norm": 2.3101487072378575, | |
| "learning_rate": 6.67324267526568e-06, | |
| "loss": 0.4952, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.4550350026925148, | |
| "grad_norm": 2.252689751264616, | |
| "learning_rate": 6.62887768900782e-06, | |
| "loss": 0.4908, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.45772751750134627, | |
| "grad_norm": 2.308986864251815, | |
| "learning_rate": 6.58436876585461e-06, | |
| "loss": 0.5115, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.4604200323101777, | |
| "grad_norm": 2.3582282125809333, | |
| "learning_rate": 6.539719838867604e-06, | |
| "loss": 0.5037, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.4631125471190092, | |
| "grad_norm": 2.7549506575377065, | |
| "learning_rate": 6.494934853479889e-06, | |
| "loss": 0.4939, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.4658050619278406, | |
| "grad_norm": 3.0601934459715916, | |
| "learning_rate": 6.450017767147447e-06, | |
| "loss": 0.4988, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.46849757673667203, | |
| "grad_norm": 2.3645028092509914, | |
| "learning_rate": 6.404972548999453e-06, | |
| "loss": 0.4962, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.4711900915455035, | |
| "grad_norm": 2.345655002717708, | |
| "learning_rate": 6.3598031794875355e-06, | |
| "loss": 0.4993, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.47388260635433493, | |
| "grad_norm": 2.4224055568308858, | |
| "learning_rate": 6.314513650034044e-06, | |
| "loss": 0.4798, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.4765751211631664, | |
| "grad_norm": 2.3397132476622047, | |
| "learning_rate": 6.269107962679344e-06, | |
| "loss": 0.4879, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.47926763597199784, | |
| "grad_norm": 2.4522897916842914, | |
| "learning_rate": 6.223590129728171e-06, | |
| "loss": 0.4765, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.4819601507808293, | |
| "grad_norm": 2.3108192463540593, | |
| "learning_rate": 6.177964173395087e-06, | |
| "loss": 0.4864, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.48465266558966075, | |
| "grad_norm": 2.267641247075609, | |
| "learning_rate": 6.1322341254490515e-06, | |
| "loss": 0.4836, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.4873451803984922, | |
| "grad_norm": 2.2555810036335497, | |
| "learning_rate": 6.086404026857146e-06, | |
| "loss": 0.4842, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.49003769520732365, | |
| "grad_norm": 2.327632792501213, | |
| "learning_rate": 6.040477927427502e-06, | |
| "loss": 0.4704, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.4927302100161551, | |
| "grad_norm": 2.2698670809887185, | |
| "learning_rate": 5.994459885451423e-06, | |
| "loss": 0.4749, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.49542272482498656, | |
| "grad_norm": 2.4725183020536847, | |
| "learning_rate": 5.948353967344786e-06, | |
| "loss": 0.4713, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.498115239633818, | |
| "grad_norm": 2.3182514849471993, | |
| "learning_rate": 5.902164247288701e-06, | |
| "loss": 0.4591, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.5008077544426495, | |
| "grad_norm": 2.2450126172776543, | |
| "learning_rate": 5.855894806869493e-06, | |
| "loss": 0.4706, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.5035002692514808, | |
| "grad_norm": 2.4095070380073333, | |
| "learning_rate": 5.809549734718031e-06, | |
| "loss": 0.4706, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.5061927840603123, | |
| "grad_norm": 2.2933657857094905, | |
| "learning_rate": 5.763133126148438e-06, | |
| "loss": 0.4665, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.5088852988691438, | |
| "grad_norm": 2.3063316226503123, | |
| "learning_rate": 5.716649082796199e-06, | |
| "loss": 0.4625, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.5115778136779753, | |
| "grad_norm": 2.3233885177577966, | |
| "learning_rate": 5.670101712255716e-06, | |
| "loss": 0.4688, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.5142703284868066, | |
| "grad_norm": 2.211725156600061, | |
| "learning_rate": 5.6234951277173524e-06, | |
| "loss": 0.4543, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.5169628432956381, | |
| "grad_norm": 2.4389466884168147, | |
| "learning_rate": 5.576833447603943e-06, | |
| "loss": 0.4703, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.5196553581044696, | |
| "grad_norm": 2.3368172021818725, | |
| "learning_rate": 5.530120795206894e-06, | |
| "loss": 0.4655, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.522347872913301, | |
| "grad_norm": 2.2744905601737804, | |
| "learning_rate": 5.483361298321803e-06, | |
| "loss": 0.4598, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.5250403877221325, | |
| "grad_norm": 2.2417520684111354, | |
| "learning_rate": 5.4365590888837175e-06, | |
| "loss": 0.4542, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.5277329025309639, | |
| "grad_norm": 2.260056934844272, | |
| "learning_rate": 5.389718302602008e-06, | |
| "loss": 0.4786, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.5304254173397954, | |
| "grad_norm": 2.2937595883241, | |
| "learning_rate": 5.342843078594916e-06, | |
| "loss": 0.4649, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.5331179321486268, | |
| "grad_norm": 2.298350623188229, | |
| "learning_rate": 5.295937559023794e-06, | |
| "loss": 0.4605, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.5358104469574583, | |
| "grad_norm": 2.2861153437802915, | |
| "learning_rate": 5.2490058887270815e-06, | |
| "loss": 0.4529, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.5385029617662898, | |
| "grad_norm": 2.2661516039174288, | |
| "learning_rate": 5.202052214854048e-06, | |
| "loss": 0.448, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5411954765751211, | |
| "grad_norm": 2.287756336798117, | |
| "learning_rate": 5.1550806864983235e-06, | |
| "loss": 0.4577, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.5438879913839526, | |
| "grad_norm": 2.21952469191807, | |
| "learning_rate": 5.108095454331261e-06, | |
| "loss": 0.4401, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.5465805061927841, | |
| "grad_norm": 2.289575301200996, | |
| "learning_rate": 5.061100670235161e-06, | |
| "loss": 0.4504, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.5492730210016155, | |
| "grad_norm": 2.254586229300791, | |
| "learning_rate": 5.014100486936383e-06, | |
| "loss": 0.4454, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.5519655358104469, | |
| "grad_norm": 2.137160999612689, | |
| "learning_rate": 4.967099057638397e-06, | |
| "loss": 0.4366, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.5546580506192784, | |
| "grad_norm": 2.241158588636164, | |
| "learning_rate": 4.92010053565477e-06, | |
| "loss": 0.442, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.5573505654281099, | |
| "grad_norm": 2.218608478823859, | |
| "learning_rate": 4.873109074042169e-06, | |
| "loss": 0.4346, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.5600430802369413, | |
| "grad_norm": 2.43524119080364, | |
| "learning_rate": 4.826128825233362e-06, | |
| "loss": 0.4428, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.5627355950457728, | |
| "grad_norm": 2.353557854902857, | |
| "learning_rate": 4.77916394067029e-06, | |
| "loss": 0.4414, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.5654281098546042, | |
| "grad_norm": 2.2938687493350423, | |
| "learning_rate": 4.732218570437224e-06, | |
| "loss": 0.4568, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.5681206246634356, | |
| "grad_norm": 2.3216946351133405, | |
| "learning_rate": 4.6852968628940345e-06, | |
| "loss": 0.4397, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.5708131394722671, | |
| "grad_norm": 2.392290169077891, | |
| "learning_rate": 4.638402964309626e-06, | |
| "loss": 0.4323, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.5735056542810986, | |
| "grad_norm": 2.259967562074028, | |
| "learning_rate": 4.591541018495543e-06, | |
| "loss": 0.4427, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.57619816908993, | |
| "grad_norm": 2.4321895943132597, | |
| "learning_rate": 4.5447151664398e-06, | |
| "loss": 0.44, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.5788906838987614, | |
| "grad_norm": 2.350780966144005, | |
| "learning_rate": 4.497929545940968e-06, | |
| "loss": 0.44, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.5815831987075929, | |
| "grad_norm": 2.327312584804818, | |
| "learning_rate": 4.4511882912425214e-06, | |
| "loss": 0.4366, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.5842757135164244, | |
| "grad_norm": 2.294345174675599, | |
| "learning_rate": 4.404495532667525e-06, | |
| "loss": 0.4383, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.5869682283252557, | |
| "grad_norm": 2.2174354559341767, | |
| "learning_rate": 4.357855396253648e-06, | |
| "loss": 0.4412, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.5896607431340872, | |
| "grad_norm": 2.350816369766935, | |
| "learning_rate": 4.311272003388569e-06, | |
| "loss": 0.442, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.5923532579429187, | |
| "grad_norm": 2.3824872125165286, | |
| "learning_rate": 4.2647494704457775e-06, | |
| "loss": 0.4349, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.5950457727517501, | |
| "grad_norm": 2.2923371304598605, | |
| "learning_rate": 4.218291908420842e-06, | |
| "loss": 0.4286, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.5977382875605816, | |
| "grad_norm": 2.346360268676029, | |
| "learning_rate": 4.171903422568128e-06, | |
| "loss": 0.4336, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.600430802369413, | |
| "grad_norm": 2.218093217821692, | |
| "learning_rate": 4.1255881120380335e-06, | |
| "loss": 0.446, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.6031233171782445, | |
| "grad_norm": 2.276362681621089, | |
| "learning_rate": 4.079350069514779e-06, | |
| "loss": 0.4407, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.6058158319870759, | |
| "grad_norm": 2.3056261390539934, | |
| "learning_rate": 4.033193380854733e-06, | |
| "loss": 0.4289, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.6085083467959074, | |
| "grad_norm": 2.1989394458411686, | |
| "learning_rate": 3.9871221247253835e-06, | |
| "loss": 0.4175, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.6112008616047389, | |
| "grad_norm": 2.3562705886952915, | |
| "learning_rate": 3.9411403722449075e-06, | |
| "loss": 0.4259, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.6138933764135702, | |
| "grad_norm": 2.322065883191611, | |
| "learning_rate": 3.895252186622433e-06, | |
| "loss": 0.418, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.6165858912224017, | |
| "grad_norm": 2.315390737203203, | |
| "learning_rate": 3.849461622798993e-06, | |
| "loss": 0.4174, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.6192784060312332, | |
| "grad_norm": 2.136385235781553, | |
| "learning_rate": 3.8037727270891927e-06, | |
| "loss": 0.4182, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.6219709208400647, | |
| "grad_norm": 2.2676350954914257, | |
| "learning_rate": 3.758189536823673e-06, | |
| "loss": 0.4162, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.624663435648896, | |
| "grad_norm": 2.3081143159586164, | |
| "learning_rate": 3.71271607999234e-06, | |
| "loss": 0.415, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.6273559504577275, | |
| "grad_norm": 2.1196383763132784, | |
| "learning_rate": 3.6673563748884268e-06, | |
| "loss": 0.4225, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.630048465266559, | |
| "grad_norm": 2.3577148406371284, | |
| "learning_rate": 3.6221144297534178e-06, | |
| "loss": 0.41, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.6327409800753904, | |
| "grad_norm": 2.332635685786891, | |
| "learning_rate": 3.5769942424228578e-06, | |
| "loss": 0.4137, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.6354334948842219, | |
| "grad_norm": 2.254772419320123, | |
| "learning_rate": 3.531999799973079e-06, | |
| "loss": 0.429, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.6381260096930533, | |
| "grad_norm": 2.1991733799648907, | |
| "learning_rate": 3.4871350783688795e-06, | |
| "loss": 0.3971, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.6408185245018848, | |
| "grad_norm": 2.2671939485505916, | |
| "learning_rate": 3.4424040421121903e-06, | |
| "loss": 0.4163, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.6435110393107162, | |
| "grad_norm": 2.258087169479648, | |
| "learning_rate": 3.397810643891741e-06, | |
| "loss": 0.4059, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.6462035541195477, | |
| "grad_norm": 2.3543023457270325, | |
| "learning_rate": 3.35335882423378e-06, | |
| "loss": 0.4224, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.6488960689283791, | |
| "grad_norm": 2.146955492636297, | |
| "learning_rate": 3.309052511153873e-06, | |
| "loss": 0.4087, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 0.6515885837372105, | |
| "grad_norm": 2.238937583269941, | |
| "learning_rate": 3.2648956198097946e-06, | |
| "loss": 0.4053, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.654281098546042, | |
| "grad_norm": 2.190819239320771, | |
| "learning_rate": 3.2208920521555677e-06, | |
| "loss": 0.3945, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.6569736133548735, | |
| "grad_norm": 2.1827434556639442, | |
| "learning_rate": 3.177045696596658e-06, | |
| "loss": 0.4089, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.6596661281637048, | |
| "grad_norm": 2.128197496062037, | |
| "learning_rate": 3.133360427646378e-06, | |
| "loss": 0.3943, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.6623586429725363, | |
| "grad_norm": 2.258181737280764, | |
| "learning_rate": 3.089840105583516e-06, | |
| "loss": 0.4014, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.6650511577813678, | |
| "grad_norm": 2.2263287080503202, | |
| "learning_rate": 3.0464885761112083e-06, | |
| "loss": 0.4085, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 0.6677436725901993, | |
| "grad_norm": 2.1158495070811467, | |
| "learning_rate": 3.0033096700171218e-06, | |
| "loss": 0.3919, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.6704361873990307, | |
| "grad_norm": 2.1418686648492415, | |
| "learning_rate": 2.960307202834941e-06, | |
| "loss": 0.408, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.6731287022078621, | |
| "grad_norm": 2.13102582996017, | |
| "learning_rate": 2.9174849745072003e-06, | |
| "loss": 0.3941, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.6758212170166936, | |
| "grad_norm": 2.1255705801818277, | |
| "learning_rate": 2.874846769049509e-06, | |
| "loss": 0.4043, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 0.678513731825525, | |
| "grad_norm": 2.314046111156198, | |
| "learning_rate": 2.8323963542161665e-06, | |
| "loss": 0.4049, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.6812062466343565, | |
| "grad_norm": 2.247094999606519, | |
| "learning_rate": 2.7901374811672256e-06, | |
| "loss": 0.3924, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 0.683898761443188, | |
| "grad_norm": 2.1331620184210958, | |
| "learning_rate": 2.748073884137016e-06, | |
| "loss": 0.3942, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.6865912762520194, | |
| "grad_norm": 2.2729136222517257, | |
| "learning_rate": 2.7062092801041717e-06, | |
| "loss": 0.39, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.6892837910608508, | |
| "grad_norm": 2.1267931341347968, | |
| "learning_rate": 2.664547368463171e-06, | |
| "loss": 0.3833, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.6919763058696823, | |
| "grad_norm": 2.1728485155426838, | |
| "learning_rate": 2.6230918306974378e-06, | |
| "loss": 0.3929, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 0.6946688206785138, | |
| "grad_norm": 2.247179993428309, | |
| "learning_rate": 2.581846330054034e-06, | |
| "loss": 0.3915, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.6973613354873451, | |
| "grad_norm": 2.146951309924855, | |
| "learning_rate": 2.5408145112199424e-06, | |
| "loss": 0.3894, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 0.7000538502961766, | |
| "grad_norm": 2.1704461453183628, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.3942, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.7027463651050081, | |
| "grad_norm": 2.164674877249693, | |
| "learning_rate": 2.4594064029965197e-06, | |
| "loss": 0.3883, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 0.7054388799138396, | |
| "grad_norm": 2.2000767508999175, | |
| "learning_rate": 2.4190373072905687e-06, | |
| "loss": 0.3884, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.708131394722671, | |
| "grad_norm": 2.3684557762781546, | |
| "learning_rate": 2.378896280125006e-06, | |
| "loss": 0.3864, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 0.7108239095315024, | |
| "grad_norm": 1.98524822133492, | |
| "learning_rate": 2.3389868685892573e-06, | |
| "loss": 0.3704, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.7135164243403339, | |
| "grad_norm": 2.19288217749482, | |
| "learning_rate": 2.2993125993058727e-06, | |
| "loss": 0.3747, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.7162089391491653, | |
| "grad_norm": 2.1617200756921497, | |
| "learning_rate": 2.2598769781189004e-06, | |
| "loss": 0.3865, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.7189014539579968, | |
| "grad_norm": 2.132621198575595, | |
| "learning_rate": 2.2206834897840814e-06, | |
| "loss": 0.3871, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 0.7215939687668282, | |
| "grad_norm": 2.145671425646657, | |
| "learning_rate": 2.1817355976609212e-06, | |
| "loss": 0.3865, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.7242864835756596, | |
| "grad_norm": 2.0530895422503868, | |
| "learning_rate": 2.14303674340665e-06, | |
| "loss": 0.3673, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 0.7269789983844911, | |
| "grad_norm": 2.2838748871806054, | |
| "learning_rate": 2.1045903466720915e-06, | |
| "loss": 0.38, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.7296715131933226, | |
| "grad_norm": 2.2502198899358348, | |
| "learning_rate": 2.0663998047994906e-06, | |
| "loss": 0.3874, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 0.732364028002154, | |
| "grad_norm": 2.128622996026954, | |
| "learning_rate": 2.0284684925223006e-06, | |
| "loss": 0.3775, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.7350565428109854, | |
| "grad_norm": 2.219696821235303, | |
| "learning_rate": 1.990799761666975e-06, | |
| "loss": 0.3798, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 0.7377490576198169, | |
| "grad_norm": 2.197304231151606, | |
| "learning_rate": 1.953396940856776e-06, | |
| "loss": 0.3724, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.7404415724286484, | |
| "grad_norm": 2.2465171659133674, | |
| "learning_rate": 1.9162633352176484e-06, | |
| "loss": 0.3831, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.7431340872374798, | |
| "grad_norm": 2.1996595546449367, | |
| "learning_rate": 1.8794022260861483e-06, | |
| "loss": 0.383, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.7458266020463112, | |
| "grad_norm": 2.1017548100685417, | |
| "learning_rate": 1.8428168707194922e-06, | |
| "loss": 0.3765, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 0.7485191168551427, | |
| "grad_norm": 2.064327514284528, | |
| "learning_rate": 1.8065105020077262e-06, | |
| "loss": 0.3694, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.7512116316639742, | |
| "grad_norm": 2.0962178314477016, | |
| "learning_rate": 1.7704863281880496e-06, | |
| "loss": 0.3866, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 0.7539041464728056, | |
| "grad_norm": 2.124206708166897, | |
| "learning_rate": 1.7347475325613134e-06, | |
| "loss": 0.3741, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.756596661281637, | |
| "grad_norm": 2.379442181939827, | |
| "learning_rate": 1.6992972732107322e-06, | |
| "loss": 0.3762, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 0.7592891760904685, | |
| "grad_norm": 2.2436516667976933, | |
| "learning_rate": 1.6641386827228107e-06, | |
| "loss": 0.3668, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.7619816908992999, | |
| "grad_norm": 2.106915257521231, | |
| "learning_rate": 1.6292748679105407e-06, | |
| "loss": 0.3715, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 0.7646742057081314, | |
| "grad_norm": 2.0865052999042546, | |
| "learning_rate": 1.594708909538854e-06, | |
| "loss": 0.3798, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.7673667205169629, | |
| "grad_norm": 2.2132376460253753, | |
| "learning_rate": 1.5604438620523932e-06, | |
| "loss": 0.3641, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.7700592353257943, | |
| "grad_norm": 2.101010632458687, | |
| "learning_rate": 1.5264827533056058e-06, | |
| "loss": 0.3736, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.7727517501346257, | |
| "grad_norm": 2.0835331975727573, | |
| "learning_rate": 1.492828584295184e-06, | |
| "loss": 0.3569, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 0.7754442649434572, | |
| "grad_norm": 2.265759304633531, | |
| "learning_rate": 1.4594843288948773e-06, | |
| "loss": 0.3662, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.7781367797522887, | |
| "grad_norm": 2.2173463946322323, | |
| "learning_rate": 1.4264529335927114e-06, | |
| "loss": 0.37, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 0.78082929456112, | |
| "grad_norm": 2.1518226151098125, | |
| "learning_rate": 1.3937373172306056e-06, | |
| "loss": 0.3638, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.7835218093699515, | |
| "grad_norm": 2.055682580111425, | |
| "learning_rate": 1.361340370746464e-06, | |
| "loss": 0.3702, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 0.786214324178783, | |
| "grad_norm": 2.0945805072566586, | |
| "learning_rate": 1.3292649569187072e-06, | |
| "loss": 0.3641, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.7889068389876144, | |
| "grad_norm": 2.099143771457234, | |
| "learning_rate": 1.2975139101132999e-06, | |
| "loss": 0.3647, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 0.7915993537964459, | |
| "grad_norm": 2.2165194526856378, | |
| "learning_rate": 1.2660900360332927e-06, | |
| "loss": 0.3715, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.7942918686052773, | |
| "grad_norm": 2.244374868499319, | |
| "learning_rate": 1.2349961114708936e-06, | |
| "loss": 0.3623, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.7969843834141088, | |
| "grad_norm": 2.186870000073866, | |
| "learning_rate": 1.2042348840620955e-06, | |
| "loss": 0.3729, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.7996768982229402, | |
| "grad_norm": 2.5050253183494244, | |
| "learning_rate": 1.1738090720438782e-06, | |
| "loss": 0.3576, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 0.8023694130317717, | |
| "grad_norm": 2.0739438241919594, | |
| "learning_rate": 1.1437213640140105e-06, | |
| "loss": 0.3617, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.8050619278406032, | |
| "grad_norm": 2.0754012982092265, | |
| "learning_rate": 1.113974418693471e-06, | |
| "loss": 0.3595, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 0.8077544426494345, | |
| "grad_norm": 2.3081364092115133, | |
| "learning_rate": 1.0845708646915054e-06, | |
| "loss": 0.361, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.810446957458266, | |
| "grad_norm": 2.0032280348669067, | |
| "learning_rate": 1.055513300273352e-06, | |
| "loss": 0.3578, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 0.8131394722670975, | |
| "grad_norm": 2.0937736954570365, | |
| "learning_rate": 1.026804293130641e-06, | |
| "loss": 0.3616, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.815831987075929, | |
| "grad_norm": 2.1070535051869443, | |
| "learning_rate": 9.984463801544992e-07, | |
| "loss": 0.3535, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 0.8185245018847603, | |
| "grad_norm": 2.3516801013906097, | |
| "learning_rate": 9.704420672113774e-07, | |
| "loss": 0.3536, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.8212170166935918, | |
| "grad_norm": 1.9740143572115736, | |
| "learning_rate": 9.42793828921616e-07, | |
| "loss": 0.3608, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.8239095315024233, | |
| "grad_norm": 2.0979215240658027, | |
| "learning_rate": 9.15504108440774e-07, | |
| "loss": 0.3736, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.8266020463112547, | |
| "grad_norm": 2.0284760942560323, | |
| "learning_rate": 8.885753172437389e-07, | |
| "loss": 0.3508, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 0.8292945611200861, | |
| "grad_norm": 2.0211247506598395, | |
| "learning_rate": 8.620098349116329e-07, | |
| "loss": 0.352, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.8319870759289176, | |
| "grad_norm": 2.3428781214119985, | |
| "learning_rate": 8.358100089215426e-07, | |
| "loss": 0.3603, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 0.8346795907377491, | |
| "grad_norm": 2.109207498019035, | |
| "learning_rate": 8.099781544390795e-07, | |
| "loss": 0.3629, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.8373721055465805, | |
| "grad_norm": 2.100615618738898, | |
| "learning_rate": 7.845165541138011e-07, | |
| "loss": 0.3556, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 0.840064620355412, | |
| "grad_norm": 2.0303755145271243, | |
| "learning_rate": 7.594274578775007e-07, | |
| "loss": 0.3555, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.8427571351642434, | |
| "grad_norm": 2.0652917821688104, | |
| "learning_rate": 7.347130827453941e-07, | |
| "loss": 0.3588, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 0.8454496499730748, | |
| "grad_norm": 1.9882512618837487, | |
| "learning_rate": 7.103756126202071e-07, | |
| "loss": 0.3613, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.8481421647819063, | |
| "grad_norm": 2.067179222751156, | |
| "learning_rate": 6.864171980991985e-07, | |
| "loss": 0.3453, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.8508346795907378, | |
| "grad_norm": 2.093050443455704, | |
| "learning_rate": 6.628399562841159e-07, | |
| "loss": 0.3548, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.8535271943995691, | |
| "grad_norm": 2.1033534008065073, | |
| "learning_rate": 6.396459705941204e-07, | |
| "loss": 0.3471, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 0.8562197092084006, | |
| "grad_norm": 1.9167623533757014, | |
| "learning_rate": 6.168372905816822e-07, | |
| "loss": 0.3477, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.8589122240172321, | |
| "grad_norm": 2.1430265742267105, | |
| "learning_rate": 5.944159317514709e-07, | |
| "loss": 0.3517, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 0.8616047388260636, | |
| "grad_norm": 2.1988162518545686, | |
| "learning_rate": 5.723838753822525e-07, | |
| "loss": 0.3689, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.864297253634895, | |
| "grad_norm": 2.2356747531202807, | |
| "learning_rate": 5.507430683518161e-07, | |
| "loss": 0.3449, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 0.8669897684437264, | |
| "grad_norm": 2.1308682594997324, | |
| "learning_rate": 5.294954229649329e-07, | |
| "loss": 0.3507, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.8696822832525579, | |
| "grad_norm": 2.0875409392190107, | |
| "learning_rate": 5.086428167843755e-07, | |
| "loss": 0.3476, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 0.8723747980613893, | |
| "grad_norm": 2.147966222762087, | |
| "learning_rate": 4.881870924650062e-07, | |
| "loss": 0.342, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.8750673128702208, | |
| "grad_norm": 2.0556458342996464, | |
| "learning_rate": 4.681300575909492e-07, | |
| "loss": 0.3471, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.8777598276790523, | |
| "grad_norm": 2.015550654050713, | |
| "learning_rate": 4.4847348451586226e-07, | |
| "loss": 0.3478, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.8804523424878837, | |
| "grad_norm": 1.998136493936924, | |
| "learning_rate": 4.2921911020631926e-07, | |
| "loss": 0.3503, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 0.8831448572967151, | |
| "grad_norm": 1.8537901854708771, | |
| "learning_rate": 4.1036863608832856e-07, | |
| "loss": 0.3468, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.8858373721055466, | |
| "grad_norm": 2.080626726929853, | |
| "learning_rate": 3.919237278969773e-07, | |
| "loss": 0.3445, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 0.8885298869143781, | |
| "grad_norm": 1.8500187551625011, | |
| "learning_rate": 3.7388601552924066e-07, | |
| "loss": 0.3515, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.8912224017232094, | |
| "grad_norm": 1.9930943811147626, | |
| "learning_rate": 3.56257092899957e-07, | |
| "loss": 0.3389, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 0.8939149165320409, | |
| "grad_norm": 2.053675256803709, | |
| "learning_rate": 3.390385178009781e-07, | |
| "loss": 0.3534, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.8966074313408724, | |
| "grad_norm": 2.0176018019009225, | |
| "learning_rate": 3.222318117635143e-07, | |
| "loss": 0.3391, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 0.8992999461497039, | |
| "grad_norm": 2.165084875616374, | |
| "learning_rate": 3.058384599236847e-07, | |
| "loss": 0.3392, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.9019924609585352, | |
| "grad_norm": 1.8300243137706005, | |
| "learning_rate": 2.8985991089127916e-07, | |
| "loss": 0.3519, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.9046849757673667, | |
| "grad_norm": 1.9039488415439678, | |
| "learning_rate": 2.7429757662175316e-07, | |
| "loss": 0.3362, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.9073774905761982, | |
| "grad_norm": 2.1498883952955805, | |
| "learning_rate": 2.5915283229146106e-07, | |
| "loss": 0.3485, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 0.9100700053850296, | |
| "grad_norm": 1.9985654494313574, | |
| "learning_rate": 2.44427016176132e-07, | |
| "loss": 0.349, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.9127625201938611, | |
| "grad_norm": 2.2813576484209896, | |
| "learning_rate": 2.301214295326193e-07, | |
| "loss": 0.3462, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 0.9154550350026925, | |
| "grad_norm": 2.0155307395335385, | |
| "learning_rate": 2.162373364839071e-07, | |
| "loss": 0.334, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.9181475498115239, | |
| "grad_norm": 1.865462382587576, | |
| "learning_rate": 2.0277596390741005e-07, | |
| "loss": 0.3465, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 0.9208400646203554, | |
| "grad_norm": 1.9778403222139225, | |
| "learning_rate": 1.8973850132655957e-07, | |
| "loss": 0.3399, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.9235325794291869, | |
| "grad_norm": 2.1982054104516986, | |
| "learning_rate": 1.7712610080568738e-07, | |
| "loss": 0.3471, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 0.9262250942380184, | |
| "grad_norm": 2.0362535026984396, | |
| "learning_rate": 1.649398768482241e-07, | |
| "loss": 0.35, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.9289176090468497, | |
| "grad_norm": 2.0586480283901714, | |
| "learning_rate": 1.5318090629821757e-07, | |
| "loss": 0.3414, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.9316101238556812, | |
| "grad_norm": 2.0363932956107, | |
| "learning_rate": 1.4185022824517392e-07, | |
| "loss": 0.3432, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.9343026386645127, | |
| "grad_norm": 2.0787927070430943, | |
| "learning_rate": 1.3094884393223872e-07, | |
| "loss": 0.3491, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 0.9369951534733441, | |
| "grad_norm": 1.9680040636502156, | |
| "learning_rate": 1.2047771666772124e-07, | |
| "loss": 0.3452, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.9396876682821755, | |
| "grad_norm": 1.8963738671804522, | |
| "learning_rate": 1.1043777173997239e-07, | |
| "loss": 0.3437, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 0.942380183091007, | |
| "grad_norm": 2.074849031085683, | |
| "learning_rate": 1.0082989633561746e-07, | |
| "loss": 0.3399, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.9450726978998385, | |
| "grad_norm": 2.005988363818133, | |
| "learning_rate": 9.165493946116432e-08, | |
| "loss": 0.341, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 0.9477652127086699, | |
| "grad_norm": 1.9739191876783688, | |
| "learning_rate": 8.29137118679757e-08, | |
| "loss": 0.3505, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.9504577275175013, | |
| "grad_norm": 2.0544856800909863, | |
| "learning_rate": 7.460698598063032e-08, | |
| "loss": 0.3504, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 0.9531502423263328, | |
| "grad_norm": 2.0266695028121604, | |
| "learning_rate": 6.673549582866368e-08, | |
| "loss": 0.3437, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.9558427571351642, | |
| "grad_norm": 1.9158086758881532, | |
| "learning_rate": 5.929993698170877e-08, | |
| "loss": 0.3407, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.9585352719439957, | |
| "grad_norm": 2.060993161808353, | |
| "learning_rate": 5.2300966488026406e-08, | |
| "loss": 0.3514, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.9612277867528272, | |
| "grad_norm": 2.2300563421538735, | |
| "learning_rate": 4.573920281645161e-08, | |
| "loss": 0.3329, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 0.9639203015616586, | |
| "grad_norm": 1.9501346199167315, | |
| "learning_rate": 3.96152258017346e-08, | |
| "loss": 0.3373, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.96661281637049, | |
| "grad_norm": 2.078047814961423, | |
| "learning_rate": 3.3929576593310644e-08, | |
| "loss": 0.3498, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 0.9693053311793215, | |
| "grad_norm": 1.9588663874547265, | |
| "learning_rate": 2.868275760747441e-08, | |
| "loss": 0.3351, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.971997845988153, | |
| "grad_norm": 2.0555012024055563, | |
| "learning_rate": 2.387523248298773e-08, | |
| "loss": 0.3438, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 0.9746903607969843, | |
| "grad_norm": 2.02985190947748, | |
| "learning_rate": 1.9507426040110134e-08, | |
| "loss": 0.3371, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.9773828756058158, | |
| "grad_norm": 1.9411470270835758, | |
| "learning_rate": 1.557972424305665e-08, | |
| "loss": 0.3423, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 0.9800753904146473, | |
| "grad_norm": 2.1196737223933773, | |
| "learning_rate": 1.209247416589232e-08, | |
| "loss": 0.344, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.9827679052234787, | |
| "grad_norm": 2.051817684549473, | |
| "learning_rate": 9.045983961865622e-09, | |
| "loss": 0.3568, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 0.9854604200323102, | |
| "grad_norm": 2.0189362221244957, | |
| "learning_rate": 6.440522836174135e-09, | |
| "loss": 0.3428, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.9881529348411416, | |
| "grad_norm": 1.9669363709037755, | |
| "learning_rate": 4.27632102218023e-09, | |
| "loss": 0.3344, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 0.9908454496499731, | |
| "grad_norm": 1.905202601388718, | |
| "learning_rate": 2.5535697610618026e-09, | |
| "loss": 0.3398, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.9935379644588045, | |
| "grad_norm": 1.9833936251009352, | |
| "learning_rate": 1.2724212849180019e-09, | |
| "loss": 0.3458, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 0.996230479267636, | |
| "grad_norm": 1.8216154780128988, | |
| "learning_rate": 4.329888033127727e-10, | |
| "loss": 0.3413, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.9989229940764675, | |
| "grad_norm": 2.0143818367459474, | |
| "learning_rate": 3.5346493273968886e-11, | |
| "loss": 0.3454, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_runtime": 3.4477, | |
| "eval_samples_per_second": 2.9, | |
| "eval_steps_per_second": 0.87, | |
| "step": 1857 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1857, | |
| "total_flos": 194409010298880.0, | |
| "train_loss": 0.5086988852623655, | |
| "train_runtime": 16625.9643, | |
| "train_samples_per_second": 1.787, | |
| "train_steps_per_second": 0.112 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1857, | |
| "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": 194409010298880.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |