{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 2414, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00041425020712510354, "grad_norm": 25.016763725925124, "learning_rate": 4.132231404958678e-08, "loss": 1.397, "step": 1 }, { "epoch": 0.002071251035625518, "grad_norm": 23.435372936086015, "learning_rate": 2.066115702479339e-07, "loss": 1.4246, "step": 5 }, { "epoch": 0.004142502071251036, "grad_norm": 15.926196367492746, "learning_rate": 4.132231404958678e-07, "loss": 1.3815, "step": 10 }, { "epoch": 0.006213753106876553, "grad_norm": 8.285490402242075, "learning_rate": 6.198347107438018e-07, "loss": 1.2646, "step": 15 }, { "epoch": 0.008285004142502071, "grad_norm": 9.681774945849943, "learning_rate": 8.264462809917356e-07, "loss": 1.1507, "step": 20 }, { "epoch": 0.010356255178127589, "grad_norm": 5.230792061371047, "learning_rate": 1.0330578512396695e-06, "loss": 1.0394, "step": 25 }, { "epoch": 0.012427506213753107, "grad_norm": 3.3262682696488306, "learning_rate": 1.2396694214876035e-06, "loss": 0.9704, "step": 30 }, { "epoch": 0.014498757249378625, "grad_norm": 3.4392074654116844, "learning_rate": 1.4462809917355372e-06, "loss": 0.9451, "step": 35 }, { "epoch": 0.016570008285004142, "grad_norm": 3.180988818969295, "learning_rate": 1.6528925619834712e-06, "loss": 0.9337, "step": 40 }, { "epoch": 0.018641259320629662, "grad_norm": 3.0735536911016488, "learning_rate": 1.859504132231405e-06, "loss": 0.9215, "step": 45 }, { "epoch": 0.020712510356255178, "grad_norm": 3.0462786353403875, "learning_rate": 2.066115702479339e-06, "loss": 0.9062, "step": 50 }, { "epoch": 0.022783761391880698, "grad_norm": 3.0730134752808147, "learning_rate": 2.2727272727272728e-06, "loss": 0.8842, "step": 55 }, { "epoch": 0.024855012427506214, "grad_norm": 3.3155266163853256, "learning_rate": 2.479338842975207e-06, "loss": 0.8857, "step": 60 }, { "epoch": 0.026926263463131733, "grad_norm": 3.039384074859877, "learning_rate": 2.6859504132231405e-06, "loss": 0.8877, "step": 65 }, { "epoch": 0.02899751449875725, "grad_norm": 3.20996125258494, "learning_rate": 2.8925619834710743e-06, "loss": 0.8757, "step": 70 }, { "epoch": 0.03106876553438277, "grad_norm": 3.1874442173865187, "learning_rate": 3.0991735537190086e-06, "loss": 0.8764, "step": 75 }, { "epoch": 0.033140016570008285, "grad_norm": 3.1164439730882765, "learning_rate": 3.3057851239669424e-06, "loss": 0.8808, "step": 80 }, { "epoch": 0.035211267605633804, "grad_norm": 3.0968653743341643, "learning_rate": 3.5123966942148763e-06, "loss": 0.8586, "step": 85 }, { "epoch": 0.037282518641259324, "grad_norm": 3.0053612558440497, "learning_rate": 3.71900826446281e-06, "loss": 0.8649, "step": 90 }, { "epoch": 0.03935376967688484, "grad_norm": 3.0612961239540524, "learning_rate": 3.925619834710744e-06, "loss": 0.8569, "step": 95 }, { "epoch": 0.041425020712510356, "grad_norm": 3.236061113233546, "learning_rate": 4.132231404958678e-06, "loss": 0.8504, "step": 100 }, { "epoch": 0.043496271748135876, "grad_norm": 3.1025374820048786, "learning_rate": 4.338842975206612e-06, "loss": 0.866, "step": 105 }, { "epoch": 0.045567522783761395, "grad_norm": 3.20993811542058, "learning_rate": 4.5454545454545455e-06, "loss": 0.8629, "step": 110 }, { "epoch": 0.04763877381938691, "grad_norm": 3.2374421513486196, "learning_rate": 4.75206611570248e-06, "loss": 0.8355, "step": 115 }, { "epoch": 0.04971002485501243, "grad_norm": 3.177183609067753, "learning_rate": 4.958677685950414e-06, "loss": 0.834, "step": 120 }, { "epoch": 0.05178127589063795, "grad_norm": 3.2498395569826326, "learning_rate": 5.165289256198347e-06, "loss": 0.8426, "step": 125 }, { "epoch": 0.053852526926263466, "grad_norm": 2.8921933174613157, "learning_rate": 5.371900826446281e-06, "loss": 0.839, "step": 130 }, { "epoch": 0.05592377796188898, "grad_norm": 3.2052052013820957, "learning_rate": 5.578512396694216e-06, "loss": 0.835, "step": 135 }, { "epoch": 0.0579950289975145, "grad_norm": 3.1882028747145403, "learning_rate": 5.785123966942149e-06, "loss": 0.8275, "step": 140 }, { "epoch": 0.06006628003314002, "grad_norm": 2.972427085962632, "learning_rate": 5.991735537190083e-06, "loss": 0.8205, "step": 145 }, { "epoch": 0.06213753106876554, "grad_norm": 3.214555174889974, "learning_rate": 6.198347107438017e-06, "loss": 0.8484, "step": 150 }, { "epoch": 0.06420878210439106, "grad_norm": 3.03373676144583, "learning_rate": 6.404958677685951e-06, "loss": 0.8377, "step": 155 }, { "epoch": 0.06628003314001657, "grad_norm": 3.0487535010694287, "learning_rate": 6.611570247933885e-06, "loss": 0.8343, "step": 160 }, { "epoch": 0.06835128417564208, "grad_norm": 3.256256317132895, "learning_rate": 6.818181818181818e-06, "loss": 0.8225, "step": 165 }, { "epoch": 0.07042253521126761, "grad_norm": 3.2258894202166815, "learning_rate": 7.0247933884297525e-06, "loss": 0.8087, "step": 170 }, { "epoch": 0.07249378624689312, "grad_norm": 3.077835207004385, "learning_rate": 7.231404958677687e-06, "loss": 0.8236, "step": 175 }, { "epoch": 0.07456503728251865, "grad_norm": 3.0625511189530705, "learning_rate": 7.43801652892562e-06, "loss": 0.8193, "step": 180 }, { "epoch": 0.07663628831814416, "grad_norm": 3.1102405842286522, "learning_rate": 7.644628099173555e-06, "loss": 0.8158, "step": 185 }, { "epoch": 0.07870753935376967, "grad_norm": 2.966000673356083, "learning_rate": 7.851239669421489e-06, "loss": 0.8173, "step": 190 }, { "epoch": 0.0807787903893952, "grad_norm": 3.0276973675209886, "learning_rate": 8.057851239669421e-06, "loss": 0.8042, "step": 195 }, { "epoch": 0.08285004142502071, "grad_norm": 3.085225304665788, "learning_rate": 8.264462809917356e-06, "loss": 0.8215, "step": 200 }, { "epoch": 0.08492129246064622, "grad_norm": 3.1936879115581096, "learning_rate": 8.47107438016529e-06, "loss": 0.8154, "step": 205 }, { "epoch": 0.08699254349627175, "grad_norm": 3.1414612904880603, "learning_rate": 8.677685950413224e-06, "loss": 0.8093, "step": 210 }, { "epoch": 0.08906379453189726, "grad_norm": 3.0130083817372624, "learning_rate": 8.884297520661158e-06, "loss": 0.8092, "step": 215 }, { "epoch": 0.09113504556752279, "grad_norm": 3.1005428851433607, "learning_rate": 9.090909090909091e-06, "loss": 0.8041, "step": 220 }, { "epoch": 0.0932062966031483, "grad_norm": 3.371007490865208, "learning_rate": 9.297520661157025e-06, "loss": 0.8224, "step": 225 }, { "epoch": 0.09527754763877382, "grad_norm": 3.2498503309557423, "learning_rate": 9.50413223140496e-06, "loss": 0.8086, "step": 230 }, { "epoch": 0.09734879867439934, "grad_norm": 2.919113050075341, "learning_rate": 9.710743801652894e-06, "loss": 0.7894, "step": 235 }, { "epoch": 0.09942004971002485, "grad_norm": 3.2170577951401893, "learning_rate": 9.917355371900828e-06, "loss": 0.8012, "step": 240 }, { "epoch": 0.10149130074565037, "grad_norm": 3.0595629094580055, "learning_rate": 9.999952928077044e-06, "loss": 0.7978, "step": 245 }, { "epoch": 0.1035625517812759, "grad_norm": 3.4530658794769664, "learning_rate": 9.999665269535307e-06, "loss": 0.8063, "step": 250 }, { "epoch": 0.1056338028169014, "grad_norm": 3.260759624069405, "learning_rate": 9.99911611854702e-06, "loss": 0.7854, "step": 255 }, { "epoch": 0.10770505385252693, "grad_norm": 3.272272731222918, "learning_rate": 9.998305503833872e-06, "loss": 0.7929, "step": 260 }, { "epoch": 0.10977630488815245, "grad_norm": 2.948051026674754, "learning_rate": 9.997233467792626e-06, "loss": 0.8019, "step": 265 }, { "epoch": 0.11184755592377796, "grad_norm": 3.02127711528984, "learning_rate": 9.995900066492902e-06, "loss": 0.8161, "step": 270 }, { "epoch": 0.11391880695940348, "grad_norm": 3.2292340670893243, "learning_rate": 9.994305369674242e-06, "loss": 0.7907, "step": 275 }, { "epoch": 0.115990057995029, "grad_norm": 2.9232386794035263, "learning_rate": 9.992449460742464e-06, "loss": 0.769, "step": 280 }, { "epoch": 0.11806130903065451, "grad_norm": 3.097523266953478, "learning_rate": 9.9903324367653e-06, "loss": 0.7767, "step": 285 }, { "epoch": 0.12013256006628004, "grad_norm": 2.8968960992375545, "learning_rate": 9.98795440846732e-06, "loss": 0.7977, "step": 290 }, { "epoch": 0.12220381110190555, "grad_norm": 2.9926682310599526, "learning_rate": 9.985315500224135e-06, "loss": 0.7874, "step": 295 }, { "epoch": 0.12427506213753108, "grad_norm": 2.776553567732678, "learning_rate": 9.982415850055902e-06, "loss": 0.7815, "step": 300 }, { "epoch": 0.1263463131731566, "grad_norm": 2.842205852918582, "learning_rate": 9.979255609620095e-06, "loss": 0.7597, "step": 305 }, { "epoch": 0.12841756420878211, "grad_norm": 3.0039444948066705, "learning_rate": 9.975834944203581e-06, "loss": 0.7599, "step": 310 }, { "epoch": 0.1304888152444076, "grad_norm": 2.8629202197596655, "learning_rate": 9.972154032713973e-06, "loss": 0.7659, "step": 315 }, { "epoch": 0.13256006628003314, "grad_norm": 2.877720215673019, "learning_rate": 9.968213067670265e-06, "loss": 0.7554, "step": 320 }, { "epoch": 0.13463131731565867, "grad_norm": 3.1605926782105764, "learning_rate": 9.964012255192776e-06, "loss": 0.7571, "step": 325 }, { "epoch": 0.13670256835128416, "grad_norm": 2.9404503322981648, "learning_rate": 9.959551814992364e-06, "loss": 0.7689, "step": 330 }, { "epoch": 0.1387738193869097, "grad_norm": 2.908375316005171, "learning_rate": 9.954831980358928e-06, "loss": 0.7614, "step": 335 }, { "epoch": 0.14084507042253522, "grad_norm": 2.7295236804133256, "learning_rate": 9.949852998149217e-06, "loss": 0.7678, "step": 340 }, { "epoch": 0.14291632145816072, "grad_norm": 3.027964894628488, "learning_rate": 9.944615128773911e-06, "loss": 0.7461, "step": 345 }, { "epoch": 0.14498757249378624, "grad_norm": 2.78594869593756, "learning_rate": 9.939118646184007e-06, "loss": 0.7445, "step": 350 }, { "epoch": 0.14705882352941177, "grad_norm": 2.836807951697266, "learning_rate": 9.933363837856485e-06, "loss": 0.7509, "step": 355 }, { "epoch": 0.1491300745650373, "grad_norm": 2.796078575429126, "learning_rate": 9.927351004779275e-06, "loss": 0.7376, "step": 360 }, { "epoch": 0.1512013256006628, "grad_norm": 3.028666823225622, "learning_rate": 9.921080461435522e-06, "loss": 0.7406, "step": 365 }, { "epoch": 0.15327257663628832, "grad_norm": 2.9470450041585234, "learning_rate": 9.914552535787122e-06, "loss": 0.7438, "step": 370 }, { "epoch": 0.15534382767191385, "grad_norm": 2.9733251290513554, "learning_rate": 9.90776756925758e-06, "loss": 0.7263, "step": 375 }, { "epoch": 0.15741507870753935, "grad_norm": 2.9855931226882855, "learning_rate": 9.900725916714157e-06, "loss": 0.735, "step": 380 }, { "epoch": 0.15948632974316487, "grad_norm": 3.4466776345203196, "learning_rate": 9.893427946449297e-06, "loss": 0.7442, "step": 385 }, { "epoch": 0.1615575807787904, "grad_norm": 2.7719407562241924, "learning_rate": 9.885874040161373e-06, "loss": 0.7211, "step": 390 }, { "epoch": 0.1636288318144159, "grad_norm": 2.9606877506272866, "learning_rate": 9.878064592934723e-06, "loss": 0.7175, "step": 395 }, { "epoch": 0.16570008285004142, "grad_norm": 2.705855551144559, "learning_rate": 9.87000001321898e-06, "loss": 0.7322, "step": 400 }, { "epoch": 0.16777133388566695, "grad_norm": 2.8094965870567195, "learning_rate": 9.86168072280772e-06, "loss": 0.7285, "step": 405 }, { "epoch": 0.16984258492129245, "grad_norm": 2.9187862973419465, "learning_rate": 9.853107156816393e-06, "loss": 0.7383, "step": 410 }, { "epoch": 0.17191383595691798, "grad_norm": 2.865758799428203, "learning_rate": 9.844279763659566e-06, "loss": 0.7318, "step": 415 }, { "epoch": 0.1739850869925435, "grad_norm": 2.751989580564356, "learning_rate": 9.835199005027477e-06, "loss": 0.7064, "step": 420 }, { "epoch": 0.176056338028169, "grad_norm": 2.8605799848391418, "learning_rate": 9.825865355861878e-06, "loss": 0.7313, "step": 425 }, { "epoch": 0.17812758906379453, "grad_norm": 2.682790962335214, "learning_rate": 9.816279304331202e-06, "loss": 0.7202, "step": 430 }, { "epoch": 0.18019884009942005, "grad_norm": 2.8122615918935594, "learning_rate": 9.806441351805025e-06, "loss": 0.7043, "step": 435 }, { "epoch": 0.18227009113504558, "grad_norm": 2.818728755171135, "learning_rate": 9.79635201282785e-06, "loss": 0.6963, "step": 440 }, { "epoch": 0.18434134217067108, "grad_norm": 2.9010708033703674, "learning_rate": 9.786011815092193e-06, "loss": 0.7175, "step": 445 }, { "epoch": 0.1864125932062966, "grad_norm": 2.6424852748903476, "learning_rate": 9.775421299410977e-06, "loss": 0.7156, "step": 450 }, { "epoch": 0.18848384424192213, "grad_norm": 2.7138618939624704, "learning_rate": 9.764581019689255e-06, "loss": 0.7027, "step": 455 }, { "epoch": 0.19055509527754763, "grad_norm": 2.8793826106519806, "learning_rate": 9.753491542895237e-06, "loss": 0.7158, "step": 460 }, { "epoch": 0.19262634631317316, "grad_norm": 2.783907178104797, "learning_rate": 9.742153449030639e-06, "loss": 0.7116, "step": 465 }, { "epoch": 0.19469759734879868, "grad_norm": 2.8753178441704406, "learning_rate": 9.730567331100333e-06, "loss": 0.7061, "step": 470 }, { "epoch": 0.19676884838442418, "grad_norm": 2.764850128693421, "learning_rate": 9.71873379508136e-06, "loss": 0.7069, "step": 475 }, { "epoch": 0.1988400994200497, "grad_norm": 3.1364863935974867, "learning_rate": 9.706653459891207e-06, "loss": 0.7192, "step": 480 }, { "epoch": 0.20091135045567524, "grad_norm": 2.830123093657544, "learning_rate": 9.694326957355452e-06, "loss": 0.6847, "step": 485 }, { "epoch": 0.20298260149130073, "grad_norm": 2.751751032731495, "learning_rate": 9.681754932174719e-06, "loss": 0.7033, "step": 490 }, { "epoch": 0.20505385252692626, "grad_norm": 2.7814835120527692, "learning_rate": 9.668938041890952e-06, "loss": 0.6831, "step": 495 }, { "epoch": 0.2071251035625518, "grad_norm": 2.7232422415442565, "learning_rate": 9.655876956853025e-06, "loss": 0.6797, "step": 500 }, { "epoch": 0.2091963545981773, "grad_norm": 2.690514918222078, "learning_rate": 9.64257236018169e-06, "loss": 0.6757, "step": 505 }, { "epoch": 0.2112676056338028, "grad_norm": 2.80973699674189, "learning_rate": 9.629024947733836e-06, "loss": 0.6787, "step": 510 }, { "epoch": 0.21333885666942834, "grad_norm": 2.9306486927031843, "learning_rate": 9.615235428066106e-06, "loss": 0.6868, "step": 515 }, { "epoch": 0.21541010770505387, "grad_norm": 2.650892915091397, "learning_rate": 9.601204522397826e-06, "loss": 0.6614, "step": 520 }, { "epoch": 0.21748135874067936, "grad_norm": 2.8844201092197435, "learning_rate": 9.586932964573298e-06, "loss": 0.6645, "step": 525 }, { "epoch": 0.2195526097763049, "grad_norm": 2.768154495837564, "learning_rate": 9.572421501023403e-06, "loss": 0.693, "step": 530 }, { "epoch": 0.22162386081193042, "grad_norm": 2.9966519058001815, "learning_rate": 9.557670890726576e-06, "loss": 0.674, "step": 535 }, { "epoch": 0.22369511184755592, "grad_norm": 3.067720386259704, "learning_rate": 9.5426819051691e-06, "loss": 0.6677, "step": 540 }, { "epoch": 0.22576636288318144, "grad_norm": 2.7847522404872542, "learning_rate": 9.527455328304756e-06, "loss": 0.6525, "step": 545 }, { "epoch": 0.22783761391880697, "grad_norm": 2.8421747999303744, "learning_rate": 9.511991956513828e-06, "loss": 0.6563, "step": 550 }, { "epoch": 0.22990886495443247, "grad_norm": 3.4652140294515266, "learning_rate": 9.496292598561445e-06, "loss": 0.6554, "step": 555 }, { "epoch": 0.231980115990058, "grad_norm": 3.0639024030581044, "learning_rate": 9.480358075555278e-06, "loss": 0.6626, "step": 560 }, { "epoch": 0.23405136702568352, "grad_norm": 2.685285704166162, "learning_rate": 9.464189220902603e-06, "loss": 0.6594, "step": 565 }, { "epoch": 0.23612261806130902, "grad_norm": 2.776081914672314, "learning_rate": 9.447786880266706e-06, "loss": 0.6503, "step": 570 }, { "epoch": 0.23819386909693455, "grad_norm": 3.042568605346591, "learning_rate": 9.431151911522656e-06, "loss": 0.6499, "step": 575 }, { "epoch": 0.24026512013256007, "grad_norm": 2.809161641721612, "learning_rate": 9.414285184712432e-06, "loss": 0.6525, "step": 580 }, { "epoch": 0.24233637116818557, "grad_norm": 2.7149883053122696, "learning_rate": 9.397187581999424e-06, "loss": 0.6489, "step": 585 }, { "epoch": 0.2444076222038111, "grad_norm": 2.7980428823017647, "learning_rate": 9.37985999762229e-06, "loss": 0.6597, "step": 590 }, { "epoch": 0.24647887323943662, "grad_norm": 2.639884812296948, "learning_rate": 9.362303337848188e-06, "loss": 0.6409, "step": 595 }, { "epoch": 0.24855012427506215, "grad_norm": 2.6103559930486044, "learning_rate": 9.344518520925377e-06, "loss": 0.6447, "step": 600 }, { "epoch": 0.2506213753106877, "grad_norm": 2.576929638356832, "learning_rate": 9.326506477035179e-06, "loss": 0.6451, "step": 605 }, { "epoch": 0.2526926263463132, "grad_norm": 2.7868011782667295, "learning_rate": 9.308268148243355e-06, "loss": 0.6498, "step": 610 }, { "epoch": 0.2547638773819387, "grad_norm": 2.6682477424551787, "learning_rate": 9.289804488450805e-06, "loss": 0.6559, "step": 615 }, { "epoch": 0.25683512841756423, "grad_norm": 2.960808461467563, "learning_rate": 9.271116463343692e-06, "loss": 0.6294, "step": 620 }, { "epoch": 0.25890637945318973, "grad_norm": 2.4997823490108715, "learning_rate": 9.25220505034293e-06, "loss": 0.6293, "step": 625 }, { "epoch": 0.2609776304888152, "grad_norm": 2.771537343410323, "learning_rate": 9.23307123855307e-06, "loss": 0.6275, "step": 630 }, { "epoch": 0.2630488815244408, "grad_norm": 2.5277700591972554, "learning_rate": 9.213716028710558e-06, "loss": 0.6311, "step": 635 }, { "epoch": 0.2651201325600663, "grad_norm": 2.7629185844938258, "learning_rate": 9.194140433131397e-06, "loss": 0.6268, "step": 640 }, { "epoch": 0.2671913835956918, "grad_norm": 2.701080417350094, "learning_rate": 9.174345475658208e-06, "loss": 0.6355, "step": 645 }, { "epoch": 0.26926263463131733, "grad_norm": 2.8198778014585644, "learning_rate": 9.154332191606671e-06, "loss": 0.6403, "step": 650 }, { "epoch": 0.27133388566694283, "grad_norm": 2.6039227822358537, "learning_rate": 9.134101627711384e-06, "loss": 0.6173, "step": 655 }, { "epoch": 0.27340513670256833, "grad_norm": 2.6757707127664343, "learning_rate": 9.113654842071114e-06, "loss": 0.6154, "step": 660 }, { "epoch": 0.2754763877381939, "grad_norm": 2.8276883897664336, "learning_rate": 9.092992904093451e-06, "loss": 0.6235, "step": 665 }, { "epoch": 0.2775476387738194, "grad_norm": 2.732324381662033, "learning_rate": 9.072116894438885e-06, "loss": 0.6216, "step": 670 }, { "epoch": 0.2796188898094449, "grad_norm": 2.7010101991119266, "learning_rate": 9.051027904964279e-06, "loss": 0.6101, "step": 675 }, { "epoch": 0.28169014084507044, "grad_norm": 2.672617421633942, "learning_rate": 9.029727038665765e-06, "loss": 0.6069, "step": 680 }, { "epoch": 0.28376139188069593, "grad_norm": 2.622602908111673, "learning_rate": 9.008215409621053e-06, "loss": 0.6197, "step": 685 }, { "epoch": 0.28583264291632143, "grad_norm": 2.710910806083558, "learning_rate": 8.986494142931168e-06, "loss": 0.6009, "step": 690 }, { "epoch": 0.287903893951947, "grad_norm": 2.653359612000082, "learning_rate": 8.964564374661597e-06, "loss": 0.6197, "step": 695 }, { "epoch": 0.2899751449875725, "grad_norm": 2.6724368567225376, "learning_rate": 8.94242725178288e-06, "loss": 0.6261, "step": 700 }, { "epoch": 0.292046396023198, "grad_norm": 2.7936657328237082, "learning_rate": 8.920083932110608e-06, "loss": 0.6031, "step": 705 }, { "epoch": 0.29411764705882354, "grad_norm": 2.668934522152384, "learning_rate": 8.89753558424488e-06, "loss": 0.6108, "step": 710 }, { "epoch": 0.29618889809444904, "grad_norm": 2.681633601569966, "learning_rate": 8.874783387509181e-06, "loss": 0.6049, "step": 715 }, { "epoch": 0.2982601491300746, "grad_norm": 2.5791868398872615, "learning_rate": 8.851828531888692e-06, "loss": 0.5964, "step": 720 }, { "epoch": 0.3003314001657001, "grad_norm": 2.7181507753767837, "learning_rate": 8.828672217968055e-06, "loss": 0.5952, "step": 725 }, { "epoch": 0.3024026512013256, "grad_norm": 2.7487732021020586, "learning_rate": 8.805315656868587e-06, "loss": 0.586, "step": 730 }, { "epoch": 0.30447390223695114, "grad_norm": 2.590564729483526, "learning_rate": 8.781760070184933e-06, "loss": 0.5892, "step": 735 }, { "epoch": 0.30654515327257664, "grad_norm": 8.095535038257605, "learning_rate": 8.75800668992117e-06, "loss": 0.5959, "step": 740 }, { "epoch": 0.30861640430820214, "grad_norm": 2.7294575476189924, "learning_rate": 8.734056758426367e-06, "loss": 0.5987, "step": 745 }, { "epoch": 0.3106876553438277, "grad_norm": 3.7285643615617174, "learning_rate": 8.709911528329623e-06, "loss": 0.6058, "step": 750 }, { "epoch": 0.3127589063794532, "grad_norm": 2.669796273888425, "learning_rate": 8.685572262474538e-06, "loss": 0.5785, "step": 755 }, { "epoch": 0.3148301574150787, "grad_norm": 2.9058866092425086, "learning_rate": 8.661040233853166e-06, "loss": 0.5899, "step": 760 }, { "epoch": 0.31690140845070425, "grad_norm": 2.8116261113212313, "learning_rate": 8.636316725539445e-06, "loss": 0.5771, "step": 765 }, { "epoch": 0.31897265948632975, "grad_norm": 2.8272995698160446, "learning_rate": 8.611403030622074e-06, "loss": 0.6032, "step": 770 }, { "epoch": 0.32104391052195524, "grad_norm": 3.260834765824046, "learning_rate": 8.586300452136895e-06, "loss": 0.5602, "step": 775 }, { "epoch": 0.3231151615575808, "grad_norm": 3.069267338732482, "learning_rate": 8.561010302998734e-06, "loss": 0.5604, "step": 780 }, { "epoch": 0.3251864125932063, "grad_norm": 2.5620662778656875, "learning_rate": 8.535533905932739e-06, "loss": 0.5785, "step": 785 }, { "epoch": 0.3272576636288318, "grad_norm": 2.7895108805109516, "learning_rate": 8.509872593405189e-06, "loss": 0.567, "step": 790 }, { "epoch": 0.32932891466445735, "grad_norm": 2.6181934693524447, "learning_rate": 8.484027707553818e-06, "loss": 0.5738, "step": 795 }, { "epoch": 0.33140016570008285, "grad_norm": 2.7501255893706014, "learning_rate": 8.458000600117604e-06, "loss": 0.5736, "step": 800 }, { "epoch": 0.33347141673570835, "grad_norm": 2.901195432495842, "learning_rate": 8.43179263236608e-06, "loss": 0.5612, "step": 805 }, { "epoch": 0.3355426677713339, "grad_norm": 2.785675517621372, "learning_rate": 8.40540517502813e-06, "loss": 0.5619, "step": 810 }, { "epoch": 0.3376139188069594, "grad_norm": 2.6828110658134743, "learning_rate": 8.378839608220304e-06, "loss": 0.5715, "step": 815 }, { "epoch": 0.3396851698425849, "grad_norm": 2.734513692698524, "learning_rate": 8.35209732137463e-06, "loss": 0.5561, "step": 820 }, { "epoch": 0.34175642087821045, "grad_norm": 2.6246880168238356, "learning_rate": 8.32517971316595e-06, "loss": 0.5677, "step": 825 }, { "epoch": 0.34382767191383595, "grad_norm": 2.634645569128113, "learning_rate": 8.298088191438753e-06, "loss": 0.5569, "step": 830 }, { "epoch": 0.34589892294946145, "grad_norm": 2.7121817599047517, "learning_rate": 8.270824173133563e-06, "loss": 0.5619, "step": 835 }, { "epoch": 0.347970173985087, "grad_norm": 2.614258516133416, "learning_rate": 8.243389084212808e-06, "loss": 0.567, "step": 840 }, { "epoch": 0.3500414250207125, "grad_norm": 2.480398234333055, "learning_rate": 8.215784359586257e-06, "loss": 0.5583, "step": 845 }, { "epoch": 0.352112676056338, "grad_norm": 2.8213935260355583, "learning_rate": 8.188011443035962e-06, "loss": 0.5725, "step": 850 }, { "epoch": 0.35418392709196356, "grad_norm": 2.645272354373542, "learning_rate": 8.160071787140742e-06, "loss": 0.5511, "step": 855 }, { "epoch": 0.35625517812758906, "grad_norm": 2.6181625341732317, "learning_rate": 8.131966853200226e-06, "loss": 0.5466, "step": 860 }, { "epoch": 0.35832642916321455, "grad_norm": 2.681437305490622, "learning_rate": 8.103698111158405e-06, "loss": 0.5545, "step": 865 }, { "epoch": 0.3603976801988401, "grad_norm": 2.630632612800765, "learning_rate": 8.075267039526764e-06, "loss": 0.536, "step": 870 }, { "epoch": 0.3624689312344656, "grad_norm": 2.6249086598741607, "learning_rate": 8.046675125306948e-06, "loss": 0.5605, "step": 875 }, { "epoch": 0.36454018227009116, "grad_norm": 4.674979453530624, "learning_rate": 8.017923863912989e-06, "loss": 0.5585, "step": 880 }, { "epoch": 0.36661143330571666, "grad_norm": 2.62162006940138, "learning_rate": 7.989014759093095e-06, "loss": 0.5453, "step": 885 }, { "epoch": 0.36868268434134216, "grad_norm": 2.6726086284207464, "learning_rate": 7.959949322850994e-06, "loss": 0.5591, "step": 890 }, { "epoch": 0.3707539353769677, "grad_norm": 2.544795844734608, "learning_rate": 7.930729075366867e-06, "loss": 0.5436, "step": 895 }, { "epoch": 0.3728251864125932, "grad_norm": 2.7090278211076466, "learning_rate": 7.901355544917827e-06, "loss": 0.533, "step": 900 }, { "epoch": 0.3748964374482187, "grad_norm": 2.642998247145757, "learning_rate": 7.87183026779799e-06, "loss": 0.5258, "step": 905 }, { "epoch": 0.37696768848384427, "grad_norm": 2.6452103374642983, "learning_rate": 7.842154788238124e-06, "loss": 0.5255, "step": 910 }, { "epoch": 0.37903893951946976, "grad_norm": 2.512168376439407, "learning_rate": 7.812330658324884e-06, "loss": 0.5296, "step": 915 }, { "epoch": 0.38111019055509526, "grad_norm": 2.4671558460813903, "learning_rate": 7.782359437919644e-06, "loss": 0.5371, "step": 920 }, { "epoch": 0.3831814415907208, "grad_norm": 2.528637016335179, "learning_rate": 7.75224269457689e-06, "loss": 0.5363, "step": 925 }, { "epoch": 0.3852526926263463, "grad_norm": 2.4988082614489677, "learning_rate": 7.721982003462255e-06, "loss": 0.5457, "step": 930 }, { "epoch": 0.3873239436619718, "grad_norm": 2.7063020232523605, "learning_rate": 7.691578947270122e-06, "loss": 0.5553, "step": 935 }, { "epoch": 0.38939519469759737, "grad_norm": 2.580731521965071, "learning_rate": 7.661035116140856e-06, "loss": 0.5393, "step": 940 }, { "epoch": 0.39146644573322287, "grad_norm": 2.6125128680259966, "learning_rate": 7.63035210757763e-06, "loss": 0.537, "step": 945 }, { "epoch": 0.39353769676884837, "grad_norm": 2.88009530336612, "learning_rate": 7.599531526362873e-06, "loss": 0.5102, "step": 950 }, { "epoch": 0.3956089478044739, "grad_norm": 2.6917058823917572, "learning_rate": 7.568574984474335e-06, "loss": 0.5444, "step": 955 }, { "epoch": 0.3976801988400994, "grad_norm": 2.528915826914381, "learning_rate": 7.537484101000787e-06, "loss": 0.5139, "step": 960 }, { "epoch": 0.3997514498757249, "grad_norm": 2.7742452091080545, "learning_rate": 7.506260502057325e-06, "loss": 0.5228, "step": 965 }, { "epoch": 0.40182270091135047, "grad_norm": 2.735824512413685, "learning_rate": 7.474905820700334e-06, "loss": 0.5076, "step": 970 }, { "epoch": 0.40389395194697597, "grad_norm": 2.504632839470433, "learning_rate": 7.443421696842066e-06, "loss": 0.5113, "step": 975 }, { "epoch": 0.40596520298260147, "grad_norm": 2.6189990731985953, "learning_rate": 7.411809777164873e-06, "loss": 0.5236, "step": 980 }, { "epoch": 0.408036454018227, "grad_norm": 2.7202816114646433, "learning_rate": 7.380071715035089e-06, "loss": 0.4957, "step": 985 }, { "epoch": 0.4101077050538525, "grad_norm": 2.544603815528399, "learning_rate": 7.3482091704165405e-06, "loss": 0.5149, "step": 990 }, { "epoch": 0.412178956089478, "grad_norm": 2.653808287858057, "learning_rate": 7.316223809783745e-06, "loss": 0.5318, "step": 995 }, { "epoch": 0.4142502071251036, "grad_norm": 2.4975583018053347, "learning_rate": 7.284117306034733e-06, "loss": 0.502, "step": 1000 }, { "epoch": 0.4163214581607291, "grad_norm": 2.6068164134263307, "learning_rate": 7.2518913384035685e-06, "loss": 0.5219, "step": 1005 }, { "epoch": 0.4183927091963546, "grad_norm": 2.5592134499430936, "learning_rate": 7.219547592372512e-06, "loss": 0.5232, "step": 1010 }, { "epoch": 0.4204639602319801, "grad_norm": 2.55902486919466, "learning_rate": 7.187087759583869e-06, "loss": 0.5105, "step": 1015 }, { "epoch": 0.4225352112676056, "grad_norm": 2.6747339014176115, "learning_rate": 7.15451353775151e-06, "loss": 0.5155, "step": 1020 }, { "epoch": 0.4246064623032311, "grad_norm": 2.7651158146577623, "learning_rate": 7.121826630572084e-06, "loss": 0.5031, "step": 1025 }, { "epoch": 0.4266777133388567, "grad_norm": 2.599835665850454, "learning_rate": 7.089028747635908e-06, "loss": 0.49, "step": 1030 }, { "epoch": 0.4287489643744822, "grad_norm": 2.376877884439919, "learning_rate": 7.056121604337554e-06, "loss": 0.5145, "step": 1035 }, { "epoch": 0.43082021541010773, "grad_norm": 2.6499811378123086, "learning_rate": 7.023106921786118e-06, "loss": 0.5026, "step": 1040 }, { "epoch": 0.43289146644573323, "grad_norm": 2.4416021270543258, "learning_rate": 6.9899864267152275e-06, "loss": 0.5104, "step": 1045 }, { "epoch": 0.43496271748135873, "grad_norm": 2.65952972663692, "learning_rate": 6.956761851392706e-06, "loss": 0.4891, "step": 1050 }, { "epoch": 0.4370339685169843, "grad_norm": 2.7372649420510915, "learning_rate": 6.9234349335299835e-06, "loss": 0.4982, "step": 1055 }, { "epoch": 0.4391052195526098, "grad_norm": 2.6195757411144784, "learning_rate": 6.890007416191209e-06, "loss": 0.495, "step": 1060 }, { "epoch": 0.4411764705882353, "grad_norm": 2.786670325821884, "learning_rate": 6.8564810477020835e-06, "loss": 0.5064, "step": 1065 }, { "epoch": 0.44324772162386084, "grad_norm": 2.5878694595493896, "learning_rate": 6.822857581558423e-06, "loss": 0.5099, "step": 1070 }, { "epoch": 0.44531897265948633, "grad_norm": 2.4447379993288982, "learning_rate": 6.789138776334441e-06, "loss": 0.4898, "step": 1075 }, { "epoch": 0.44739022369511183, "grad_norm": 2.6631225395828553, "learning_rate": 6.7553263955907755e-06, "loss": 0.4931, "step": 1080 }, { "epoch": 0.4494614747307374, "grad_norm": 2.654211485645013, "learning_rate": 6.721422207782249e-06, "loss": 0.4993, "step": 1085 }, { "epoch": 0.4515327257663629, "grad_norm": 2.600239674778423, "learning_rate": 6.687427986165379e-06, "loss": 0.5037, "step": 1090 }, { "epoch": 0.4536039768019884, "grad_norm": 2.54732788792508, "learning_rate": 6.653345508705629e-06, "loss": 0.4753, "step": 1095 }, { "epoch": 0.45567522783761394, "grad_norm": 2.7822986021593463, "learning_rate": 6.6191765579844205e-06, "loss": 0.4917, "step": 1100 }, { "epoch": 0.45774647887323944, "grad_norm": 2.939248938320084, "learning_rate": 6.584922921105894e-06, "loss": 0.4883, "step": 1105 }, { "epoch": 0.45981772990886494, "grad_norm": 2.423586271173749, "learning_rate": 6.550586389603451e-06, "loss": 0.491, "step": 1110 }, { "epoch": 0.4618889809444905, "grad_norm": 2.4513704540315606, "learning_rate": 6.5161687593460395e-06, "loss": 0.4825, "step": 1115 }, { "epoch": 0.463960231980116, "grad_norm": 2.6796215648493966, "learning_rate": 6.481671830444243e-06, "loss": 0.488, "step": 1120 }, { "epoch": 0.4660314830157415, "grad_norm": 2.545881520256298, "learning_rate": 6.447097407156114e-06, "loss": 0.5015, "step": 1125 }, { "epoch": 0.46810273405136704, "grad_norm": 3.1915012933342424, "learning_rate": 6.412447297792818e-06, "loss": 0.4819, "step": 1130 }, { "epoch": 0.47017398508699254, "grad_norm": 2.553235577004568, "learning_rate": 6.377723314624057e-06, "loss": 0.4738, "step": 1135 }, { "epoch": 0.47224523612261804, "grad_norm": 2.5439392265675713, "learning_rate": 6.3429272737832726e-06, "loss": 0.4813, "step": 1140 }, { "epoch": 0.4743164871582436, "grad_norm": 2.6091491623909193, "learning_rate": 6.308060995172673e-06, "loss": 0.4715, "step": 1145 }, { "epoch": 0.4763877381938691, "grad_norm": 2.5315606363649366, "learning_rate": 6.273126302368037e-06, "loss": 0.4634, "step": 1150 }, { "epoch": 0.4784589892294946, "grad_norm": 2.6403560356162497, "learning_rate": 6.238125022523343e-06, "loss": 0.4708, "step": 1155 }, { "epoch": 0.48053024026512015, "grad_norm": 2.657576962556168, "learning_rate": 6.203058986275207e-06, "loss": 0.4704, "step": 1160 }, { "epoch": 0.48260149130074564, "grad_norm": 3.279911405391214, "learning_rate": 6.1679300276471285e-06, "loss": 0.4703, "step": 1165 }, { "epoch": 0.48467274233637114, "grad_norm": 2.763561883210254, "learning_rate": 6.132739983953579e-06, "loss": 0.4678, "step": 1170 }, { "epoch": 0.4867439933719967, "grad_norm": 2.586169770999439, "learning_rate": 6.097490695703896e-06, "loss": 0.4808, "step": 1175 }, { "epoch": 0.4888152444076222, "grad_norm": 2.5949893218236317, "learning_rate": 6.062184006506027e-06, "loss": 0.4676, "step": 1180 }, { "epoch": 0.4908864954432477, "grad_norm": 2.4910197230584887, "learning_rate": 6.026821762970102e-06, "loss": 0.4715, "step": 1185 }, { "epoch": 0.49295774647887325, "grad_norm": 2.649434397581146, "learning_rate": 5.991405814611855e-06, "loss": 0.4512, "step": 1190 }, { "epoch": 0.49502899751449875, "grad_norm": 2.551201712225206, "learning_rate": 5.955938013755888e-06, "loss": 0.4823, "step": 1195 }, { "epoch": 0.4971002485501243, "grad_norm": 2.5799023691296306, "learning_rate": 5.920420215438794e-06, "loss": 0.4628, "step": 1200 }, { "epoch": 0.4991714995857498, "grad_norm": 2.5504083573066745, "learning_rate": 5.8848542773121285e-06, "loss": 0.4523, "step": 1205 }, { "epoch": 0.5012427506213754, "grad_norm": 2.480187119312849, "learning_rate": 5.849242059545259e-06, "loss": 0.4512, "step": 1210 }, { "epoch": 0.5033140016570008, "grad_norm": 2.4039163996261994, "learning_rate": 5.81358542472807e-06, "loss": 0.4579, "step": 1215 }, { "epoch": 0.5053852526926264, "grad_norm": 2.786434938537425, "learning_rate": 5.777886237773542e-06, "loss": 0.4541, "step": 1220 }, { "epoch": 0.5074565037282519, "grad_norm": 2.53888078337496, "learning_rate": 5.742146365820223e-06, "loss": 0.452, "step": 1225 }, { "epoch": 0.5095277547638773, "grad_norm": 2.425357941164124, "learning_rate": 5.706367678134562e-06, "loss": 0.4534, "step": 1230 }, { "epoch": 0.5115990057995029, "grad_norm": 2.5846634023226613, "learning_rate": 5.670552046013151e-06, "loss": 0.4571, "step": 1235 }, { "epoch": 0.5136702568351285, "grad_norm": 2.617133045234817, "learning_rate": 5.634701342684852e-06, "loss": 0.4678, "step": 1240 }, { "epoch": 0.5157415078707539, "grad_norm": 2.613871639924252, "learning_rate": 5.598817443212813e-06, "loss": 0.4512, "step": 1245 }, { "epoch": 0.5178127589063795, "grad_norm": 2.5058094367984105, "learning_rate": 5.562902224396416e-06, "loss": 0.4509, "step": 1250 }, { "epoch": 0.519884009942005, "grad_norm": 2.6389072877354076, "learning_rate": 5.526957564673098e-06, "loss": 0.4604, "step": 1255 }, { "epoch": 0.5219552609776305, "grad_norm": 2.672563749030521, "learning_rate": 5.49098534402012e-06, "loss": 0.4519, "step": 1260 }, { "epoch": 0.524026512013256, "grad_norm": 2.5230301620718896, "learning_rate": 5.454987443856235e-06, "loss": 0.4609, "step": 1265 }, { "epoch": 0.5260977630488816, "grad_norm": 2.4317534037121646, "learning_rate": 5.418965746943281e-06, "loss": 0.4549, "step": 1270 }, { "epoch": 0.528169014084507, "grad_norm": 2.408256454364234, "learning_rate": 5.3829221372877175e-06, "loss": 0.4471, "step": 1275 }, { "epoch": 0.5302402651201326, "grad_norm": 2.486322304924638, "learning_rate": 5.34685850004208e-06, "loss": 0.4411, "step": 1280 }, { "epoch": 0.5323115161557581, "grad_norm": 2.443407072113582, "learning_rate": 5.310776721406392e-06, "loss": 0.4394, "step": 1285 }, { "epoch": 0.5343827671913836, "grad_norm": 2.480647417041202, "learning_rate": 5.2746786885295034e-06, "loss": 0.4338, "step": 1290 }, { "epoch": 0.5364540182270091, "grad_norm": 2.5522614079112627, "learning_rate": 5.238566289410396e-06, "loss": 0.4348, "step": 1295 }, { "epoch": 0.5385252692626347, "grad_norm": 2.4598209355181173, "learning_rate": 5.2024414127994325e-06, "loss": 0.4409, "step": 1300 }, { "epoch": 0.5405965202982601, "grad_norm": 2.5418070276004725, "learning_rate": 5.166305948099574e-06, "loss": 0.4317, "step": 1305 }, { "epoch": 0.5426677713338857, "grad_norm": 2.6143541261220067, "learning_rate": 5.13016178526756e-06, "loss": 0.4303, "step": 1310 }, { "epoch": 0.5447390223695112, "grad_norm": 2.515422020642716, "learning_rate": 5.094010814715062e-06, "loss": 0.4485, "step": 1315 }, { "epoch": 0.5468102734051367, "grad_norm": 2.5928748688853025, "learning_rate": 5.057854927209804e-06, "loss": 0.4322, "step": 1320 }, { "epoch": 0.5488815244407622, "grad_norm": 2.5029232379490436, "learning_rate": 5.0216960137766805e-06, "loss": 0.4289, "step": 1325 }, { "epoch": 0.5509527754763878, "grad_norm": 2.443158305252529, "learning_rate": 4.985535965598843e-06, "loss": 0.4312, "step": 1330 }, { "epoch": 0.5530240265120132, "grad_norm": 2.408569855897696, "learning_rate": 4.949376673918802e-06, "loss": 0.4283, "step": 1335 }, { "epoch": 0.5550952775476388, "grad_norm": 2.410755925506855, "learning_rate": 4.913220029939491e-06, "loss": 0.4288, "step": 1340 }, { "epoch": 0.5571665285832643, "grad_norm": 2.478534791183418, "learning_rate": 4.877067924725368e-06, "loss": 0.4332, "step": 1345 }, { "epoch": 0.5592377796188898, "grad_norm": 2.4240926015008206, "learning_rate": 4.840922249103506e-06, "loss": 0.4354, "step": 1350 }, { "epoch": 0.5613090306545153, "grad_norm": 2.5501733813586007, "learning_rate": 4.804784893564697e-06, "loss": 0.4375, "step": 1355 }, { "epoch": 0.5633802816901409, "grad_norm": 2.554955579898062, "learning_rate": 4.7686577481645745e-06, "loss": 0.418, "step": 1360 }, { "epoch": 0.5654515327257663, "grad_norm": 2.459375213074571, "learning_rate": 4.732542702424759e-06, "loss": 0.4243, "step": 1365 }, { "epoch": 0.5675227837613919, "grad_norm": 2.397494158089266, "learning_rate": 4.696441645234042e-06, "loss": 0.4204, "step": 1370 }, { "epoch": 0.5695940347970174, "grad_norm": 2.4345369407733526, "learning_rate": 4.660356464749578e-06, "loss": 0.4263, "step": 1375 }, { "epoch": 0.5716652858326429, "grad_norm": 2.454149233410318, "learning_rate": 4.624289048298147e-06, "loss": 0.4356, "step": 1380 }, { "epoch": 0.5737365368682684, "grad_norm": 2.365793643326068, "learning_rate": 4.588241282277428e-06, "loss": 0.4218, "step": 1385 }, { "epoch": 0.575807787903894, "grad_norm": 2.49267921877343, "learning_rate": 4.55221505205734e-06, "loss": 0.4147, "step": 1390 }, { "epoch": 0.5778790389395194, "grad_norm": 2.47045962295073, "learning_rate": 4.516212241881448e-06, "loss": 0.419, "step": 1395 }, { "epoch": 0.579950289975145, "grad_norm": 2.333976066950849, "learning_rate": 4.480234734768393e-06, "loss": 0.4258, "step": 1400 }, { "epoch": 0.5820215410107705, "grad_norm": 2.4480903616402574, "learning_rate": 4.444284412413418e-06, "loss": 0.4228, "step": 1405 }, { "epoch": 0.584092792046396, "grad_norm": 2.73948339117454, "learning_rate": 4.408363155089952e-06, "loss": 0.4182, "step": 1410 }, { "epoch": 0.5861640430820215, "grad_norm": 2.5318399376905574, "learning_rate": 4.3724728415512585e-06, "loss": 0.4243, "step": 1415 }, { "epoch": 0.5882352941176471, "grad_norm": 2.349988056487565, "learning_rate": 4.3366153489321855e-06, "loss": 0.4088, "step": 1420 }, { "epoch": 0.5903065451532725, "grad_norm": 2.5814054127223547, "learning_rate": 4.30079255265098e-06, "loss": 0.427, "step": 1425 }, { "epoch": 0.5923777961888981, "grad_norm": 2.5464682543387647, "learning_rate": 4.265006326311199e-06, "loss": 0.4184, "step": 1430 }, { "epoch": 0.5944490472245236, "grad_norm": 2.6981909077259654, "learning_rate": 4.229258541603723e-06, "loss": 0.4112, "step": 1435 }, { "epoch": 0.5965202982601492, "grad_norm": 2.530642985109849, "learning_rate": 4.1935510682088545e-06, "loss": 0.4161, "step": 1440 }, { "epoch": 0.5985915492957746, "grad_norm": 2.3991994956531477, "learning_rate": 4.157885773698535e-06, "loss": 0.4057, "step": 1445 }, { "epoch": 0.6006628003314002, "grad_norm": 2.508735547313183, "learning_rate": 4.122264523438668e-06, "loss": 0.4179, "step": 1450 }, { "epoch": 0.6027340513670257, "grad_norm": 2.461104449052784, "learning_rate": 4.086689180491554e-06, "loss": 0.4026, "step": 1455 }, { "epoch": 0.6048053024026512, "grad_norm": 2.474286356254044, "learning_rate": 4.051161605518453e-06, "loss": 0.4077, "step": 1460 }, { "epoch": 0.6068765534382767, "grad_norm": 2.4043276638514133, "learning_rate": 4.015683656682255e-06, "loss": 0.4113, "step": 1465 }, { "epoch": 0.6089478044739023, "grad_norm": 2.35632471405803, "learning_rate": 3.980257189550316e-06, "loss": 0.4046, "step": 1470 }, { "epoch": 0.6110190555095277, "grad_norm": 2.536709571304453, "learning_rate": 3.94488405699739e-06, "loss": 0.4203, "step": 1475 }, { "epoch": 0.6130903065451533, "grad_norm": 2.5149137277006166, "learning_rate": 3.909566109108727e-06, "loss": 0.4174, "step": 1480 }, { "epoch": 0.6151615575807788, "grad_norm": 2.5385856803177584, "learning_rate": 3.874305193083313e-06, "loss": 0.4002, "step": 1485 }, { "epoch": 0.6172328086164043, "grad_norm": 2.349589477854346, "learning_rate": 3.839103153137247e-06, "loss": 0.402, "step": 1490 }, { "epoch": 0.6193040596520298, "grad_norm": 2.3946918231711636, "learning_rate": 3.803961830407297e-06, "loss": 0.4061, "step": 1495 }, { "epoch": 0.6213753106876554, "grad_norm": 2.544602612840506, "learning_rate": 3.768883062854598e-06, "loss": 0.4025, "step": 1500 }, { "epoch": 0.6234465617232808, "grad_norm": 2.383657763129309, "learning_rate": 3.7338686851685267e-06, "loss": 0.4028, "step": 1505 }, { "epoch": 0.6255178127589064, "grad_norm": 2.5515450906768247, "learning_rate": 3.6989205286707398e-06, "loss": 0.3963, "step": 1510 }, { "epoch": 0.6275890637945319, "grad_norm": 2.570785968389103, "learning_rate": 3.664040421219393e-06, "loss": 0.392, "step": 1515 }, { "epoch": 0.6296603148301574, "grad_norm": 2.415155890596637, "learning_rate": 3.6292301871135425e-06, "loss": 0.402, "step": 1520 }, { "epoch": 0.6317315658657829, "grad_norm": 2.2700735692867338, "learning_rate": 3.59449164699773e-06, "loss": 0.4027, "step": 1525 }, { "epoch": 0.6338028169014085, "grad_norm": 2.503259239290081, "learning_rate": 3.55982661776676e-06, "loss": 0.3979, "step": 1530 }, { "epoch": 0.6358740679370339, "grad_norm": 2.3776461351972533, "learning_rate": 3.5252369124706697e-06, "loss": 0.399, "step": 1535 }, { "epoch": 0.6379453189726595, "grad_norm": 2.4407919104876408, "learning_rate": 3.4907243402199013e-06, "loss": 0.3848, "step": 1540 }, { "epoch": 0.640016570008285, "grad_norm": 2.4235423953455584, "learning_rate": 3.4562907060906908e-06, "loss": 0.3942, "step": 1545 }, { "epoch": 0.6420878210439105, "grad_norm": 2.334813016710133, "learning_rate": 3.4219378110306523e-06, "loss": 0.3832, "step": 1550 }, { "epoch": 0.644159072079536, "grad_norm": 2.4475057109815483, "learning_rate": 3.3876674517645815e-06, "loss": 0.3846, "step": 1555 }, { "epoch": 0.6462303231151616, "grad_norm": 2.326795339596567, "learning_rate": 3.353481420700495e-06, "loss": 0.3798, "step": 1560 }, { "epoch": 0.648301574150787, "grad_norm": 2.435390022677688, "learning_rate": 3.319381505835868e-06, "loss": 0.3941, "step": 1565 }, { "epoch": 0.6503728251864126, "grad_norm": 2.4536094043181036, "learning_rate": 3.285369490664133e-06, "loss": 0.3821, "step": 1570 }, { "epoch": 0.6524440762220381, "grad_norm": 2.504683129191115, "learning_rate": 3.251447154081394e-06, "loss": 0.3898, "step": 1575 }, { "epoch": 0.6545153272576636, "grad_norm": 2.3956689805469686, "learning_rate": 3.2176162702933816e-06, "loss": 0.3962, "step": 1580 }, { "epoch": 0.6565865782932891, "grad_norm": 2.378453578818355, "learning_rate": 3.183878608722669e-06, "loss": 0.3821, "step": 1585 }, { "epoch": 0.6586578293289147, "grad_norm": 2.5459599177310683, "learning_rate": 3.150235933916115e-06, "loss": 0.3769, "step": 1590 }, { "epoch": 0.6607290803645401, "grad_norm": 2.5400034610229802, "learning_rate": 3.1166900054525873e-06, "loss": 0.395, "step": 1595 }, { "epoch": 0.6628003314001657, "grad_norm": 2.454940256127563, "learning_rate": 3.0832425778509235e-06, "loss": 0.3912, "step": 1600 }, { "epoch": 0.6648715824357913, "grad_norm": 2.5362804306603226, "learning_rate": 3.049895400478174e-06, "loss": 0.3715, "step": 1605 }, { "epoch": 0.6669428334714167, "grad_norm": 2.41081895664832, "learning_rate": 3.0166502174581012e-06, "loss": 0.397, "step": 1610 }, { "epoch": 0.6690140845070423, "grad_norm": 2.2563258575491463, "learning_rate": 2.983508767579956e-06, "loss": 0.3818, "step": 1615 }, { "epoch": 0.6710853355426678, "grad_norm": 2.376842003953282, "learning_rate": 2.950472784207544e-06, "loss": 0.3846, "step": 1620 }, { "epoch": 0.6731565865782932, "grad_norm": 2.3664949016960346, "learning_rate": 2.917543995188562e-06, "loss": 0.3781, "step": 1625 }, { "epoch": 0.6752278376139188, "grad_norm": 2.3849302735848115, "learning_rate": 2.8847241227642255e-06, "loss": 0.3958, "step": 1630 }, { "epoch": 0.6772990886495444, "grad_norm": 2.5344791694793787, "learning_rate": 2.852014883479198e-06, "loss": 0.3826, "step": 1635 }, { "epoch": 0.6793703396851698, "grad_norm": 2.4818949599326454, "learning_rate": 2.819417988091814e-06, "loss": 0.3792, "step": 1640 }, { "epoch": 0.6814415907207954, "grad_norm": 2.435060675968546, "learning_rate": 2.786935141484586e-06, "loss": 0.3757, "step": 1645 }, { "epoch": 0.6835128417564209, "grad_norm": 2.419041362796777, "learning_rate": 2.754568042575061e-06, "loss": 0.3582, "step": 1650 }, { "epoch": 0.6855840927920464, "grad_norm": 2.307149287337768, "learning_rate": 2.7223183842269442e-06, "loss": 0.3654, "step": 1655 }, { "epoch": 0.6876553438276719, "grad_norm": 2.6864534476540385, "learning_rate": 2.6901878531615677e-06, "loss": 0.3748, "step": 1660 }, { "epoch": 0.6897265948632975, "grad_norm": 2.2618648344542422, "learning_rate": 2.658178129869672e-06, "loss": 0.3712, "step": 1665 }, { "epoch": 0.6917978458989229, "grad_norm": 2.536865044716639, "learning_rate": 2.6262908885235046e-06, "loss": 0.3839, "step": 1670 }, { "epoch": 0.6938690969345485, "grad_norm": 2.364529571592883, "learning_rate": 2.594527796889265e-06, "loss": 0.3769, "step": 1675 }, { "epoch": 0.695940347970174, "grad_norm": 2.4511493617139157, "learning_rate": 2.5628905162398797e-06, "loss": 0.3775, "step": 1680 }, { "epoch": 0.6980115990057995, "grad_norm": 2.286811524822981, "learning_rate": 2.531380701268108e-06, "loss": 0.3691, "step": 1685 }, { "epoch": 0.700082850041425, "grad_norm": 2.374881319089612, "learning_rate": 2.5000000000000015e-06, "loss": 0.3682, "step": 1690 }, { "epoch": 0.7021541010770506, "grad_norm": 2.3297735306333567, "learning_rate": 2.4687500537087027e-06, "loss": 0.3692, "step": 1695 }, { "epoch": 0.704225352112676, "grad_norm": 2.411714180522495, "learning_rate": 2.4376324968286154e-06, "loss": 0.3686, "step": 1700 }, { "epoch": 0.7062966031483016, "grad_norm": 2.5922730167326993, "learning_rate": 2.40664895686991e-06, "loss": 0.3751, "step": 1705 }, { "epoch": 0.7083678541839271, "grad_norm": 2.4677915696013595, "learning_rate": 2.375801054333409e-06, "loss": 0.3659, "step": 1710 }, { "epoch": 0.7104391052195526, "grad_norm": 2.243293986744168, "learning_rate": 2.345090402625822e-06, "loss": 0.3624, "step": 1715 }, { "epoch": 0.7125103562551781, "grad_norm": 2.450116391199614, "learning_rate": 2.3145186079753685e-06, "loss": 0.3669, "step": 1720 }, { "epoch": 0.7145816072908037, "grad_norm": 2.474541853299616, "learning_rate": 2.2840872693477694e-06, "loss": 0.3664, "step": 1725 }, { "epoch": 0.7166528583264291, "grad_norm": 2.3872001686516406, "learning_rate": 2.253797978362617e-06, "loss": 0.361, "step": 1730 }, { "epoch": 0.7187241093620547, "grad_norm": 2.2943599415418556, "learning_rate": 2.2236523192101264e-06, "loss": 0.3719, "step": 1735 }, { "epoch": 0.7207953603976802, "grad_norm": 2.21150040192589, "learning_rate": 2.193651868568285e-06, "loss": 0.3663, "step": 1740 }, { "epoch": 0.7228666114333057, "grad_norm": 2.4319334394881555, "learning_rate": 2.16379819552038e-06, "loss": 0.3658, "step": 1745 }, { "epoch": 0.7249378624689312, "grad_norm": 2.5322580117212823, "learning_rate": 2.1340928614729445e-06, "loss": 0.3559, "step": 1750 }, { "epoch": 0.7270091135045568, "grad_norm": 2.3279134325134025, "learning_rate": 2.1045374200740863e-06, "loss": 0.3529, "step": 1755 }, { "epoch": 0.7290803645401823, "grad_norm": 2.5739897139034458, "learning_rate": 2.075133417132223e-06, "loss": 0.3555, "step": 1760 }, { "epoch": 0.7311516155758078, "grad_norm": 2.353018197097101, "learning_rate": 2.045882390535248e-06, "loss": 0.343, "step": 1765 }, { "epoch": 0.7332228666114333, "grad_norm": 2.491663965193569, "learning_rate": 2.016785870170079e-06, "loss": 0.359, "step": 1770 }, { "epoch": 0.7352941176470589, "grad_norm": 2.3908242354152947, "learning_rate": 1.987845377842656e-06, "loss": 0.3579, "step": 1775 }, { "epoch": 0.7373653686826843, "grad_norm": 2.5104066455677994, "learning_rate": 1.9590624271983406e-06, "loss": 0.3699, "step": 1780 }, { "epoch": 0.7394366197183099, "grad_norm": 2.4861661251321476, "learning_rate": 1.9304385236427505e-06, "loss": 0.3557, "step": 1785 }, { "epoch": 0.7415078707539354, "grad_norm": 2.4253505879037736, "learning_rate": 1.9019751642630252e-06, "loss": 0.3623, "step": 1790 }, { "epoch": 0.7435791217895609, "grad_norm": 2.340083931791682, "learning_rate": 1.8736738377495196e-06, "loss": 0.343, "step": 1795 }, { "epoch": 0.7456503728251864, "grad_norm": 2.6661015705638285, "learning_rate": 1.8455360243179537e-06, "loss": 0.3598, "step": 1800 }, { "epoch": 0.747721623860812, "grad_norm": 2.1646178929171276, "learning_rate": 1.8175631956319823e-06, "loss": 0.3638, "step": 1805 }, { "epoch": 0.7497928748964374, "grad_norm": 2.228011432034891, "learning_rate": 1.7897568147262323e-06, "loss": 0.3499, "step": 1810 }, { "epoch": 0.751864125932063, "grad_norm": 2.4740496369496214, "learning_rate": 1.7621183359297817e-06, "loss": 0.3525, "step": 1815 }, { "epoch": 0.7539353769676885, "grad_norm": 2.358825941979143, "learning_rate": 1.7346492047900897e-06, "loss": 0.3437, "step": 1820 }, { "epoch": 0.756006628003314, "grad_norm": 2.337603535008447, "learning_rate": 1.7073508579973996e-06, "loss": 0.3637, "step": 1825 }, { "epoch": 0.7580778790389395, "grad_norm": 2.360415808362112, "learning_rate": 1.6802247233095914e-06, "loss": 0.347, "step": 1830 }, { "epoch": 0.7601491300745651, "grad_norm": 2.6763904086938, "learning_rate": 1.6532722194775108e-06, "loss": 0.3463, "step": 1835 }, { "epoch": 0.7622203811101905, "grad_norm": 2.3514585924650526, "learning_rate": 1.626494756170765e-06, "loss": 0.3484, "step": 1840 }, { "epoch": 0.7642916321458161, "grad_norm": 2.42524484482098, "learning_rate": 1.5998937339039889e-06, "loss": 0.3487, "step": 1845 }, { "epoch": 0.7663628831814416, "grad_norm": 2.58087105557009, "learning_rate": 1.5734705439636017e-06, "loss": 0.3549, "step": 1850 }, { "epoch": 0.7684341342170671, "grad_norm": 2.394771197223568, "learning_rate": 1.5472265683350397e-06, "loss": 0.3559, "step": 1855 }, { "epoch": 0.7705053852526926, "grad_norm": 2.3582167282060658, "learning_rate": 1.5211631796304721e-06, "loss": 0.3584, "step": 1860 }, { "epoch": 0.7725766362883182, "grad_norm": 2.331092724299011, "learning_rate": 1.495281741017016e-06, "loss": 0.3466, "step": 1865 }, { "epoch": 0.7746478873239436, "grad_norm": 2.28617703940716, "learning_rate": 1.46958360614543e-06, "loss": 0.3517, "step": 1870 }, { "epoch": 0.7767191383595692, "grad_norm": 2.3966104590327237, "learning_rate": 1.4440701190793278e-06, "loss": 0.3537, "step": 1875 }, { "epoch": 0.7787903893951947, "grad_norm": 2.4201249487304213, "learning_rate": 1.4187426142248723e-06, "loss": 0.3403, "step": 1880 }, { "epoch": 0.7808616404308202, "grad_norm": 2.3902702727318914, "learning_rate": 1.3936024162609897e-06, "loss": 0.3505, "step": 1885 }, { "epoch": 0.7829328914664457, "grad_norm": 2.2302767559192125, "learning_rate": 1.3686508400700787e-06, "loss": 0.3348, "step": 1890 }, { "epoch": 0.7850041425020713, "grad_norm": 2.4319942079194075, "learning_rate": 1.3438891906692447e-06, "loss": 0.3421, "step": 1895 }, { "epoch": 0.7870753935376967, "grad_norm": 2.375091350674846, "learning_rate": 1.3193187631420462e-06, "loss": 0.3348, "step": 1900 }, { "epoch": 0.7891466445733223, "grad_norm": 2.4301957256777964, "learning_rate": 1.2949408425707566e-06, "loss": 0.3548, "step": 1905 }, { "epoch": 0.7912178956089478, "grad_norm": 2.398834908595337, "learning_rate": 1.2707567039691505e-06, "loss": 0.3344, "step": 1910 }, { "epoch": 0.7932891466445733, "grad_norm": 2.3505842004513706, "learning_rate": 1.2467676122158224e-06, "loss": 0.3417, "step": 1915 }, { "epoch": 0.7953603976801988, "grad_norm": 2.4517225523020483, "learning_rate": 1.222974821988024e-06, "loss": 0.3414, "step": 1920 }, { "epoch": 0.7974316487158244, "grad_norm": 2.3881967478357162, "learning_rate": 1.1993795776960498e-06, "loss": 0.3414, "step": 1925 }, { "epoch": 0.7995028997514498, "grad_norm": 2.6306627816170356, "learning_rate": 1.1759831134181504e-06, "loss": 0.342, "step": 1930 }, { "epoch": 0.8015741507870754, "grad_norm": 2.4222984296737895, "learning_rate": 1.1527866528359805e-06, "loss": 0.3402, "step": 1935 }, { "epoch": 0.8036454018227009, "grad_norm": 2.3762377591772945, "learning_rate": 1.1297914091706086e-06, "loss": 0.3379, "step": 1940 }, { "epoch": 0.8057166528583264, "grad_norm": 2.417870771557216, "learning_rate": 1.1069985851190524e-06, "loss": 0.3343, "step": 1945 }, { "epoch": 0.8077879038939519, "grad_norm": 2.4001550264199243, "learning_rate": 1.0844093727913868e-06, "loss": 0.3324, "step": 1950 }, { "epoch": 0.8098591549295775, "grad_norm": 2.412548311604175, "learning_rate": 1.062024953648384e-06, "loss": 0.3411, "step": 1955 }, { "epoch": 0.8119304059652029, "grad_norm": 2.43743502490611, "learning_rate": 1.039846498439727e-06, "loss": 0.3285, "step": 1960 }, { "epoch": 0.8140016570008285, "grad_norm": 2.2943521986587094, "learning_rate": 1.0178751671427755e-06, "loss": 0.3377, "step": 1965 }, { "epoch": 0.816072908036454, "grad_norm": 2.3843142630142293, "learning_rate": 9.961121089018933e-07, "loss": 0.3404, "step": 1970 }, { "epoch": 0.8181441590720795, "grad_norm": 2.6382759934475475, "learning_rate": 9.745584619683524e-07, "loss": 0.3353, "step": 1975 }, { "epoch": 0.820215410107705, "grad_norm": 2.3722231153320266, "learning_rate": 9.532153536407923e-07, "loss": 0.3319, "step": 1980 }, { "epoch": 0.8222866611433306, "grad_norm": 2.3133068132446972, "learning_rate": 9.320839002062682e-07, "loss": 0.3382, "step": 1985 }, { "epoch": 0.824357912178956, "grad_norm": 2.2756315531006033, "learning_rate": 9.111652068818621e-07, "loss": 0.3289, "step": 1990 }, { "epoch": 0.8264291632145816, "grad_norm": 2.4605656863599425, "learning_rate": 8.904603677568785e-07, "loss": 0.3377, "step": 1995 }, { "epoch": 0.8285004142502072, "grad_norm": 2.27699874953015, "learning_rate": 8.699704657356195e-07, "loss": 0.3211, "step": 2000 }, { "epoch": 0.8305716652858326, "grad_norm": 2.3894658068590116, "learning_rate": 8.496965724807516e-07, "loss": 0.3367, "step": 2005 }, { "epoch": 0.8326429163214581, "grad_norm": 2.2445746992679996, "learning_rate": 8.296397483572515e-07, "loss": 0.3389, "step": 2010 }, { "epoch": 0.8347141673570837, "grad_norm": 2.381830368968363, "learning_rate": 8.098010423769503e-07, "loss": 0.3262, "step": 2015 }, { "epoch": 0.8367854183927091, "grad_norm": 2.426499782170603, "learning_rate": 7.901814921436624e-07, "loss": 0.3272, "step": 2020 }, { "epoch": 0.8388566694283347, "grad_norm": 2.490160647108125, "learning_rate": 7.70782123798921e-07, "loss": 0.3308, "step": 2025 }, { "epoch": 0.8409279204639603, "grad_norm": 2.535040500341068, "learning_rate": 7.516039519683105e-07, "loss": 0.3473, "step": 2030 }, { "epoch": 0.8429991714995857, "grad_norm": 2.23382584266138, "learning_rate": 7.326479797083963e-07, "loss": 0.3302, "step": 2035 }, { "epoch": 0.8450704225352113, "grad_norm": 2.4396469263041713, "learning_rate": 7.139151984542636e-07, "loss": 0.3298, "step": 2040 }, { "epoch": 0.8471416735708368, "grad_norm": 2.2842616169616248, "learning_rate": 6.954065879676653e-07, "loss": 0.3334, "step": 2045 }, { "epoch": 0.8492129246064622, "grad_norm": 2.419243879506469, "learning_rate": 6.771231162857722e-07, "loss": 0.3275, "step": 2050 }, { "epoch": 0.8512841756420878, "grad_norm": 2.4323047115332064, "learning_rate": 6.590657396705525e-07, "loss": 0.3245, "step": 2055 }, { "epoch": 0.8533554266777134, "grad_norm": 2.339208863990624, "learning_rate": 6.412354025587509e-07, "loss": 0.3209, "step": 2060 }, { "epoch": 0.8554266777133389, "grad_norm": 2.360114851629984, "learning_rate": 6.236330375124921e-07, "loss": 0.3243, "step": 2065 }, { "epoch": 0.8574979287489644, "grad_norm": 2.5130965816548194, "learning_rate": 6.062595651705111e-07, "loss": 0.3383, "step": 2070 }, { "epoch": 0.8595691797845899, "grad_norm": 2.3368416170990263, "learning_rate": 5.891158941999959e-07, "loss": 0.3269, "step": 2075 }, { "epoch": 0.8616404308202155, "grad_norm": 2.330701427888861, "learning_rate": 5.722029212490666e-07, "loss": 0.3236, "step": 2080 }, { "epoch": 0.8637116818558409, "grad_norm": 2.540475224729907, "learning_rate": 5.555215308998779e-07, "loss": 0.3285, "step": 2085 }, { "epoch": 0.8657829328914665, "grad_norm": 2.5600806865761374, "learning_rate": 5.390725956223531e-07, "loss": 0.3211, "step": 2090 }, { "epoch": 0.867854183927092, "grad_norm": 2.34829263603315, "learning_rate": 5.22856975728554e-07, "loss": 0.3297, "step": 2095 }, { "epoch": 0.8699254349627175, "grad_norm": 2.442824739766999, "learning_rate": 5.068755193276798e-07, "loss": 0.3369, "step": 2100 }, { "epoch": 0.871996685998343, "grad_norm": 2.392129300326281, "learning_rate": 4.911290622817161e-07, "loss": 0.3299, "step": 2105 }, { "epoch": 0.8740679370339686, "grad_norm": 2.311534168637044, "learning_rate": 4.756184281617121e-07, "loss": 0.3179, "step": 2110 }, { "epoch": 0.876139188069594, "grad_norm": 2.2917487645499137, "learning_rate": 4.6034442820471037e-07, "loss": 0.3285, "step": 2115 }, { "epoch": 0.8782104391052196, "grad_norm": 2.500541636074517, "learning_rate": 4.4530786127131575e-07, "loss": 0.3273, "step": 2120 }, { "epoch": 0.8802816901408451, "grad_norm": 2.2822677762583177, "learning_rate": 4.305095138039106e-07, "loss": 0.3157, "step": 2125 }, { "epoch": 0.8823529411764706, "grad_norm": 2.3134578397681467, "learning_rate": 4.159501597855287e-07, "loss": 0.3262, "step": 2130 }, { "epoch": 0.8844241922120961, "grad_norm": 2.4825534943370546, "learning_rate": 4.0163056069936757e-07, "loss": 0.3149, "step": 2135 }, { "epoch": 0.8864954432477217, "grad_norm": 2.403007178722133, "learning_rate": 3.8755146548896784e-07, "loss": 0.3217, "step": 2140 }, { "epoch": 0.8885666942833471, "grad_norm": 2.3873149152261344, "learning_rate": 3.737136105190337e-07, "loss": 0.3138, "step": 2145 }, { "epoch": 0.8906379453189727, "grad_norm": 2.49980767807061, "learning_rate": 3.6011771953693044e-07, "loss": 0.3237, "step": 2150 }, { "epoch": 0.8927091963545982, "grad_norm": 2.2351626566232716, "learning_rate": 3.4676450363481937e-07, "loss": 0.3237, "step": 2155 }, { "epoch": 0.8947804473902237, "grad_norm": 2.4257911808908688, "learning_rate": 3.336546612124758e-07, "loss": 0.3135, "step": 2160 }, { "epoch": 0.8968516984258492, "grad_norm": 2.519765653654668, "learning_rate": 3.20788877940757e-07, "loss": 0.3243, "step": 2165 }, { "epoch": 0.8989229494614748, "grad_norm": 2.3858511648749676, "learning_rate": 3.081678267257404e-07, "loss": 0.3208, "step": 2170 }, { "epoch": 0.9009942004971002, "grad_norm": 2.4861439475389093, "learning_rate": 2.9579216767352815e-07, "loss": 0.3197, "step": 2175 }, { "epoch": 0.9030654515327258, "grad_norm": 2.4294619704816958, "learning_rate": 2.836625480557265e-07, "loss": 0.3228, "step": 2180 }, { "epoch": 0.9051367025683513, "grad_norm": 2.3277516458933496, "learning_rate": 2.7177960227558863e-07, "loss": 0.3339, "step": 2185 }, { "epoch": 0.9072079536039768, "grad_norm": 2.3773432515045534, "learning_rate": 2.601439518348331e-07, "loss": 0.3124, "step": 2190 }, { "epoch": 0.9092792046396023, "grad_norm": 2.5339708541978716, "learning_rate": 2.487562053011422e-07, "loss": 0.326, "step": 2195 }, { "epoch": 0.9113504556752279, "grad_norm": 2.481716967996545, "learning_rate": 2.376169582763288e-07, "loss": 0.3151, "step": 2200 }, { "epoch": 0.9134217067108533, "grad_norm": 2.2686406418223313, "learning_rate": 2.2672679336518789e-07, "loss": 0.312, "step": 2205 }, { "epoch": 0.9154929577464789, "grad_norm": 2.5173675055564897, "learning_rate": 2.1608628014502364e-07, "loss": 0.3053, "step": 2210 }, { "epoch": 0.9175642087821044, "grad_norm": 2.481673571543988, "learning_rate": 2.0569597513586004e-07, "loss": 0.3238, "step": 2215 }, { "epoch": 0.9196354598177299, "grad_norm": 2.5531731176476873, "learning_rate": 1.955564217713335e-07, "loss": 0.3188, "step": 2220 }, { "epoch": 0.9217067108533554, "grad_norm": 2.5276339569374087, "learning_rate": 1.8566815037026897e-07, "loss": 0.3118, "step": 2225 }, { "epoch": 0.923777961888981, "grad_norm": 2.252154625231819, "learning_rate": 1.7603167810894662e-07, "loss": 0.3265, "step": 2230 }, { "epoch": 0.9258492129246064, "grad_norm": 2.273522902304356, "learning_rate": 1.6664750899404892e-07, "loss": 0.3098, "step": 2235 }, { "epoch": 0.927920463960232, "grad_norm": 2.50675768705392, "learning_rate": 1.5751613383630128e-07, "loss": 0.3204, "step": 2240 }, { "epoch": 0.9299917149958575, "grad_norm": 2.407392282132766, "learning_rate": 1.4863803022480362e-07, "loss": 0.3275, "step": 2245 }, { "epoch": 0.932062966031483, "grad_norm": 2.436345135350679, "learning_rate": 1.4001366250204762e-07, "loss": 0.3235, "step": 2250 }, { "epoch": 0.9341342170671085, "grad_norm": 2.3678545927429338, "learning_rate": 1.3164348173963392e-07, "loss": 0.3174, "step": 2255 }, { "epoch": 0.9362054681027341, "grad_norm": 2.427336823758693, "learning_rate": 1.235279257146804e-07, "loss": 0.3202, "step": 2260 }, { "epoch": 0.9382767191383595, "grad_norm": 2.4196877754077564, "learning_rate": 1.1566741888692168e-07, "loss": 0.3181, "step": 2265 }, { "epoch": 0.9403479701739851, "grad_norm": 2.4954501100817796, "learning_rate": 1.080623723765134e-07, "loss": 0.3183, "step": 2270 }, { "epoch": 0.9424192212096106, "grad_norm": 2.3208587306459516, "learning_rate": 1.0071318394252849e-07, "loss": 0.3099, "step": 2275 }, { "epoch": 0.9444904722452361, "grad_norm": 2.2876493966364118, "learning_rate": 9.362023796215036e-08, "loss": 0.3216, "step": 2280 }, { "epoch": 0.9465617232808616, "grad_norm": 2.515768505794023, "learning_rate": 8.678390541057512e-08, "loss": 0.3278, "step": 2285 }, { "epoch": 0.9486329743164872, "grad_norm": 2.3999993199901466, "learning_rate": 8.020454384160437e-08, "loss": 0.313, "step": 2290 }, { "epoch": 0.9507042253521126, "grad_norm": 2.298443535586414, "learning_rate": 7.388249736894615e-08, "loss": 0.3111, "step": 2295 }, { "epoch": 0.9527754763877382, "grad_norm": 2.315063848172232, "learning_rate": 6.78180966482156e-08, "loss": 0.322, "step": 2300 }, { "epoch": 0.9548467274233637, "grad_norm": 2.403741400386069, "learning_rate": 6.201165885964311e-08, "loss": 0.3085, "step": 2305 }, { "epoch": 0.9569179784589892, "grad_norm": 2.3919137237639507, "learning_rate": 5.646348769148491e-08, "loss": 0.3206, "step": 2310 }, { "epoch": 0.9589892294946147, "grad_norm": 2.47499264680871, "learning_rate": 5.117387332413737e-08, "loss": 0.3132, "step": 2315 }, { "epoch": 0.9610604805302403, "grad_norm": 2.338185949687448, "learning_rate": 4.6143092414961396e-08, "loss": 0.3168, "step": 2320 }, { "epoch": 0.9631317315658657, "grad_norm": 2.479851438198001, "learning_rate": 4.1371408083815126e-08, "loss": 0.309, "step": 2325 }, { "epoch": 0.9652029826014913, "grad_norm": 2.349586877794245, "learning_rate": 3.685906989928656e-08, "loss": 0.3126, "step": 2330 }, { "epoch": 0.9672742336371168, "grad_norm": 2.3451220158746806, "learning_rate": 3.2606313865646276e-08, "loss": 0.3216, "step": 2335 }, { "epoch": 0.9693454846727423, "grad_norm": 2.4750945398377895, "learning_rate": 2.861336241050061e-08, "loss": 0.3188, "step": 2340 }, { "epoch": 0.9714167357083678, "grad_norm": 2.31406090105744, "learning_rate": 2.488042437315985e-08, "loss": 0.3151, "step": 2345 }, { "epoch": 0.9734879867439934, "grad_norm": 2.608847916655168, "learning_rate": 2.1407694993714755e-08, "loss": 0.315, "step": 2350 }, { "epoch": 0.9755592377796188, "grad_norm": 2.6290475819113537, "learning_rate": 1.8195355902824174e-08, "loss": 0.3198, "step": 2355 }, { "epoch": 0.9776304888152444, "grad_norm": 2.501907594320039, "learning_rate": 1.5243575112218744e-08, "loss": 0.3262, "step": 2360 }, { "epoch": 0.97970173985087, "grad_norm": 2.5324795778435463, "learning_rate": 1.2552507005909042e-08, "loss": 0.3257, "step": 2365 }, { "epoch": 0.9817729908864954, "grad_norm": 2.5414315083037846, "learning_rate": 1.0122292332114814e-08, "loss": 0.3201, "step": 2370 }, { "epoch": 0.9838442419221209, "grad_norm": 2.4138191146737173, "learning_rate": 7.953058195900864e-09, "loss": 0.315, "step": 2375 }, { "epoch": 0.9859154929577465, "grad_norm": 2.516201320443761, "learning_rate": 6.044918052531268e-09, "loss": 0.3331, "step": 2380 }, { "epoch": 0.987986743993372, "grad_norm": 2.269236864889141, "learning_rate": 4.397971701533554e-09, "loss": 0.3202, "step": 2385 }, { "epoch": 0.9900579950289975, "grad_norm": 2.40160196051819, "learning_rate": 3.0123052814812203e-09, "loss": 0.3129, "step": 2390 }, { "epoch": 0.992129246064623, "grad_norm": 2.5061197371508785, "learning_rate": 1.887991265486222e-09, "loss": 0.332, "step": 2395 }, { "epoch": 0.9942004971002486, "grad_norm": 2.4868832937938166, "learning_rate": 1.025088457409229e-09, "loss": 0.3238, "step": 2400 }, { "epoch": 0.996271748135874, "grad_norm": 2.3062986700535184, "learning_rate": 4.2364198878597216e-10, "loss": 0.3151, "step": 2405 }, { "epoch": 0.9983429991714996, "grad_norm": 2.4684579842087415, "learning_rate": 8.368331646302353e-11, "loss": 0.324, "step": 2410 }, { "epoch": 1.0, "eval_runtime": 2.7207, "eval_samples_per_second": 3.676, "eval_steps_per_second": 1.103, "step": 2414 }, { "epoch": 1.0, "step": 2414, "total_flos": 252721244405760.0, "train_loss": 0.0, "train_runtime": 0.0084, "train_samples_per_second": 4618577.431, "train_steps_per_second": 288668.563 } ], "logging_steps": 5, "max_steps": 2414, "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": 252721244405760.0, "train_batch_size": 4, "trial_name": null, "trial_params": null }