{ "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": 25.544971119297156, "learning_rate": 4.132231404958678e-08, "loss": 1.4202, "step": 1 }, { "epoch": 0.002071251035625518, "grad_norm": 22.908596652075445, "learning_rate": 2.066115702479339e-07, "loss": 1.3991, "step": 5 }, { "epoch": 0.004142502071251036, "grad_norm": 14.46167637817398, "learning_rate": 4.132231404958678e-07, "loss": 1.3444, "step": 10 }, { "epoch": 0.006213753106876553, "grad_norm": 8.290495106293157, "learning_rate": 6.198347107438018e-07, "loss": 1.265, "step": 15 }, { "epoch": 0.008285004142502071, "grad_norm": 10.380639172364994, "learning_rate": 8.264462809917356e-07, "loss": 1.1339, "step": 20 }, { "epoch": 0.010356255178127589, "grad_norm": 4.6120455640556655, "learning_rate": 1.0330578512396695e-06, "loss": 1.0227, "step": 25 }, { "epoch": 0.012427506213753107, "grad_norm": 3.474075737289564, "learning_rate": 1.2396694214876035e-06, "loss": 0.9686, "step": 30 }, { "epoch": 0.014498757249378625, "grad_norm": 3.5169761983961307, "learning_rate": 1.4462809917355372e-06, "loss": 0.9698, "step": 35 }, { "epoch": 0.016570008285004142, "grad_norm": 3.080371238649772, "learning_rate": 1.6528925619834712e-06, "loss": 0.9282, "step": 40 }, { "epoch": 0.018641259320629662, "grad_norm": 3.1171379571192817, "learning_rate": 1.859504132231405e-06, "loss": 0.9337, "step": 45 }, { "epoch": 0.020712510356255178, "grad_norm": 3.073597957607947, "learning_rate": 2.066115702479339e-06, "loss": 0.9055, "step": 50 }, { "epoch": 0.022783761391880698, "grad_norm": 2.9703230058655525, "learning_rate": 2.2727272727272728e-06, "loss": 0.8945, "step": 55 }, { "epoch": 0.024855012427506214, "grad_norm": 3.1846969380023857, "learning_rate": 2.479338842975207e-06, "loss": 0.8822, "step": 60 }, { "epoch": 0.026926263463131733, "grad_norm": 3.0811606787156673, "learning_rate": 2.6859504132231405e-06, "loss": 0.8753, "step": 65 }, { "epoch": 0.02899751449875725, "grad_norm": 2.991431842056114, "learning_rate": 2.8925619834710743e-06, "loss": 0.8717, "step": 70 }, { "epoch": 0.03106876553438277, "grad_norm": 2.9935603364172456, "learning_rate": 3.0991735537190086e-06, "loss": 0.8799, "step": 75 }, { "epoch": 0.033140016570008285, "grad_norm": 3.1118618030596426, "learning_rate": 3.3057851239669424e-06, "loss": 0.8764, "step": 80 }, { "epoch": 0.035211267605633804, "grad_norm": 2.944950169684341, "learning_rate": 3.5123966942148763e-06, "loss": 0.8653, "step": 85 }, { "epoch": 0.037282518641259324, "grad_norm": 2.9739085526435427, "learning_rate": 3.71900826446281e-06, "loss": 0.8549, "step": 90 }, { "epoch": 0.03935376967688484, "grad_norm": 3.1253110494762013, "learning_rate": 3.925619834710744e-06, "loss": 0.8704, "step": 95 }, { "epoch": 0.041425020712510356, "grad_norm": 3.242980841012618, "learning_rate": 4.132231404958678e-06, "loss": 0.8588, "step": 100 }, { "epoch": 0.043496271748135876, "grad_norm": 3.0122534126779295, "learning_rate": 4.338842975206612e-06, "loss": 0.837, "step": 105 }, { "epoch": 0.045567522783761395, "grad_norm": 3.1069186883719895, "learning_rate": 4.5454545454545455e-06, "loss": 0.8595, "step": 110 }, { "epoch": 0.04763877381938691, "grad_norm": 2.9563700149903323, "learning_rate": 4.75206611570248e-06, "loss": 0.8473, "step": 115 }, { "epoch": 0.04971002485501243, "grad_norm": 3.3541971707561156, "learning_rate": 4.958677685950414e-06, "loss": 0.8354, "step": 120 }, { "epoch": 0.05178127589063795, "grad_norm": 3.175924731149988, "learning_rate": 5.165289256198347e-06, "loss": 0.8586, "step": 125 }, { "epoch": 0.053852526926263466, "grad_norm": 2.995462162831613, "learning_rate": 5.371900826446281e-06, "loss": 0.8341, "step": 130 }, { "epoch": 0.05592377796188898, "grad_norm": 3.3766169021162122, "learning_rate": 5.578512396694216e-06, "loss": 0.8193, "step": 135 }, { "epoch": 0.0579950289975145, "grad_norm": 2.9810267959097985, "learning_rate": 5.785123966942149e-06, "loss": 0.8392, "step": 140 }, { "epoch": 0.06006628003314002, "grad_norm": 3.096657680618283, "learning_rate": 5.991735537190083e-06, "loss": 0.8312, "step": 145 }, { "epoch": 0.06213753106876554, "grad_norm": 3.298036046188238, "learning_rate": 6.198347107438017e-06, "loss": 0.8317, "step": 150 }, { "epoch": 0.06420878210439106, "grad_norm": 3.1212455612648675, "learning_rate": 6.404958677685951e-06, "loss": 0.8197, "step": 155 }, { "epoch": 0.06628003314001657, "grad_norm": 3.0402863673109364, "learning_rate": 6.611570247933885e-06, "loss": 0.8275, "step": 160 }, { "epoch": 0.06835128417564208, "grad_norm": 3.5022361283638324, "learning_rate": 6.818181818181818e-06, "loss": 0.8053, "step": 165 }, { "epoch": 0.07042253521126761, "grad_norm": 3.348710798098812, "learning_rate": 7.0247933884297525e-06, "loss": 0.8189, "step": 170 }, { "epoch": 0.07249378624689312, "grad_norm": 3.4458320445449315, "learning_rate": 7.231404958677687e-06, "loss": 0.8226, "step": 175 }, { "epoch": 0.07456503728251865, "grad_norm": 3.303113534217444, "learning_rate": 7.43801652892562e-06, "loss": 0.8163, "step": 180 }, { "epoch": 0.07663628831814416, "grad_norm": 3.343520607614814, "learning_rate": 7.644628099173555e-06, "loss": 0.8304, "step": 185 }, { "epoch": 0.07870753935376967, "grad_norm": 3.1056946041337166, "learning_rate": 7.851239669421489e-06, "loss": 0.8094, "step": 190 }, { "epoch": 0.0807787903893952, "grad_norm": 3.183371153299341, "learning_rate": 8.057851239669421e-06, "loss": 0.7949, "step": 195 }, { "epoch": 0.08285004142502071, "grad_norm": 3.269984586984269, "learning_rate": 8.264462809917356e-06, "loss": 0.806, "step": 200 }, { "epoch": 0.08492129246064622, "grad_norm": 3.335022249388234, "learning_rate": 8.47107438016529e-06, "loss": 0.8078, "step": 205 }, { "epoch": 0.08699254349627175, "grad_norm": 2.9606812615195053, "learning_rate": 8.677685950413224e-06, "loss": 0.8209, "step": 210 }, { "epoch": 0.08906379453189726, "grad_norm": 3.202967921746539, "learning_rate": 8.884297520661158e-06, "loss": 0.8187, "step": 215 }, { "epoch": 0.09113504556752279, "grad_norm": 3.204189894993486, "learning_rate": 9.090909090909091e-06, "loss": 0.794, "step": 220 }, { "epoch": 0.0932062966031483, "grad_norm": 2.954713853298521, "learning_rate": 9.297520661157025e-06, "loss": 0.7962, "step": 225 }, { "epoch": 0.09527754763877382, "grad_norm": 3.0875834441491667, "learning_rate": 9.50413223140496e-06, "loss": 0.7996, "step": 230 }, { "epoch": 0.09734879867439934, "grad_norm": 2.9139649747682577, "learning_rate": 9.710743801652894e-06, "loss": 0.8072, "step": 235 }, { "epoch": 0.09942004971002485, "grad_norm": 3.041063558792154, "learning_rate": 9.917355371900828e-06, "loss": 0.782, "step": 240 }, { "epoch": 0.10149130074565037, "grad_norm": 3.1128633848066163, "learning_rate": 9.999952928077044e-06, "loss": 0.816, "step": 245 }, { "epoch": 0.1035625517812759, "grad_norm": 3.066590654725202, "learning_rate": 9.999665269535307e-06, "loss": 0.8032, "step": 250 }, { "epoch": 0.1056338028169014, "grad_norm": 2.91105191628418, "learning_rate": 9.99911611854702e-06, "loss": 0.791, "step": 255 }, { "epoch": 0.10770505385252693, "grad_norm": 3.006318406918519, "learning_rate": 9.998305503833872e-06, "loss": 0.8038, "step": 260 }, { "epoch": 0.10977630488815245, "grad_norm": 3.0115800135705566, "learning_rate": 9.997233467792626e-06, "loss": 0.7912, "step": 265 }, { "epoch": 0.11184755592377796, "grad_norm": 3.023844776872081, "learning_rate": 9.995900066492902e-06, "loss": 0.792, "step": 270 }, { "epoch": 0.11391880695940348, "grad_norm": 3.1715416482126915, "learning_rate": 9.994305369674242e-06, "loss": 0.7991, "step": 275 }, { "epoch": 0.115990057995029, "grad_norm": 2.9257225672853746, "learning_rate": 9.992449460742464e-06, "loss": 0.7807, "step": 280 }, { "epoch": 0.11806130903065451, "grad_norm": 2.7664836018212893, "learning_rate": 9.9903324367653e-06, "loss": 0.7856, "step": 285 }, { "epoch": 0.12013256006628004, "grad_norm": 2.9468391906330247, "learning_rate": 9.98795440846732e-06, "loss": 0.7722, "step": 290 }, { "epoch": 0.12220381110190555, "grad_norm": 2.8733137592595286, "learning_rate": 9.985315500224135e-06, "loss": 0.7735, "step": 295 }, { "epoch": 0.12427506213753108, "grad_norm": 2.854643690941792, "learning_rate": 9.982415850055902e-06, "loss": 0.7698, "step": 300 }, { "epoch": 0.1263463131731566, "grad_norm": 2.863511338124913, "learning_rate": 9.979255609620095e-06, "loss": 0.7917, "step": 305 }, { "epoch": 0.12841756420878211, "grad_norm": 2.9944118362700904, "learning_rate": 9.975834944203581e-06, "loss": 0.7695, "step": 310 }, { "epoch": 0.1304888152444076, "grad_norm": 2.771403108971041, "learning_rate": 9.972154032713973e-06, "loss": 0.7753, "step": 315 }, { "epoch": 0.13256006628003314, "grad_norm": 3.2277842078089374, "learning_rate": 9.968213067670265e-06, "loss": 0.7613, "step": 320 }, { "epoch": 0.13463131731565867, "grad_norm": 2.925991890930421, "learning_rate": 9.964012255192776e-06, "loss": 0.7776, "step": 325 }, { "epoch": 0.13670256835128416, "grad_norm": 2.87449247681849, "learning_rate": 9.959551814992364e-06, "loss": 0.7782, "step": 330 }, { "epoch": 0.1387738193869097, "grad_norm": 3.190659210015262, "learning_rate": 9.954831980358928e-06, "loss": 0.7767, "step": 335 }, { "epoch": 0.14084507042253522, "grad_norm": 2.9321464206149495, "learning_rate": 9.949852998149217e-06, "loss": 0.7405, "step": 340 }, { "epoch": 0.14291632145816072, "grad_norm": 3.226386690882447, "learning_rate": 9.944615128773911e-06, "loss": 0.7511, "step": 345 }, { "epoch": 0.14498757249378624, "grad_norm": 3.0453214708306042, "learning_rate": 9.939118646184007e-06, "loss": 0.7564, "step": 350 }, { "epoch": 0.14705882352941177, "grad_norm": 2.8510835653406694, "learning_rate": 9.933363837856485e-06, "loss": 0.7534, "step": 355 }, { "epoch": 0.1491300745650373, "grad_norm": 2.9653747140222166, "learning_rate": 9.927351004779275e-06, "loss": 0.7333, "step": 360 }, { "epoch": 0.1512013256006628, "grad_norm": 2.812708687715769, "learning_rate": 9.921080461435522e-06, "loss": 0.7539, "step": 365 }, { "epoch": 0.15327257663628832, "grad_norm": 2.915870749598374, "learning_rate": 9.914552535787122e-06, "loss": 0.7373, "step": 370 }, { "epoch": 0.15534382767191385, "grad_norm": 2.8076111518882, "learning_rate": 9.90776756925758e-06, "loss": 0.7441, "step": 375 }, { "epoch": 0.15741507870753935, "grad_norm": 2.8608289315852526, "learning_rate": 9.900725916714157e-06, "loss": 0.7222, "step": 380 }, { "epoch": 0.15948632974316487, "grad_norm": 2.8612454321586243, "learning_rate": 9.893427946449297e-06, "loss": 0.7117, "step": 385 }, { "epoch": 0.1615575807787904, "grad_norm": 2.953283089910919, "learning_rate": 9.885874040161373e-06, "loss": 0.7427, "step": 390 }, { "epoch": 0.1636288318144159, "grad_norm": 2.7815731663847174, "learning_rate": 9.878064592934723e-06, "loss": 0.7395, "step": 395 }, { "epoch": 0.16570008285004142, "grad_norm": 2.7419741730795795, "learning_rate": 9.87000001321898e-06, "loss": 0.7296, "step": 400 }, { "epoch": 0.16777133388566695, "grad_norm": 2.8743326993853646, "learning_rate": 9.86168072280772e-06, "loss": 0.7516, "step": 405 }, { "epoch": 0.16984258492129245, "grad_norm": 2.6739468010673417, "learning_rate": 9.853107156816393e-06, "loss": 0.7212, "step": 410 }, { "epoch": 0.17191383595691798, "grad_norm": 2.843192826013082, "learning_rate": 9.844279763659566e-06, "loss": 0.7117, "step": 415 }, { "epoch": 0.1739850869925435, "grad_norm": 3.2338486647651994, "learning_rate": 9.835199005027477e-06, "loss": 0.7096, "step": 420 }, { "epoch": 0.176056338028169, "grad_norm": 2.8399723159571644, "learning_rate": 9.825865355861878e-06, "loss": 0.7088, "step": 425 }, { "epoch": 0.17812758906379453, "grad_norm": 2.8047724008761517, "learning_rate": 9.816279304331202e-06, "loss": 0.7051, "step": 430 }, { "epoch": 0.18019884009942005, "grad_norm": 2.965002963831138, "learning_rate": 9.806441351805025e-06, "loss": 0.7144, "step": 435 }, { "epoch": 0.18227009113504558, "grad_norm": 3.17550957312684, "learning_rate": 9.79635201282785e-06, "loss": 0.7034, "step": 440 }, { "epoch": 0.18434134217067108, "grad_norm": 2.8366101693679258, "learning_rate": 9.786011815092193e-06, "loss": 0.7142, "step": 445 }, { "epoch": 0.1864125932062966, "grad_norm": 2.8130134851461652, "learning_rate": 9.775421299410977e-06, "loss": 0.6979, "step": 450 }, { "epoch": 0.18848384424192213, "grad_norm": 2.64038641528435, "learning_rate": 9.764581019689255e-06, "loss": 0.7019, "step": 455 }, { "epoch": 0.19055509527754763, "grad_norm": 2.6453379760100435, "learning_rate": 9.753491542895237e-06, "loss": 0.6933, "step": 460 }, { "epoch": 0.19262634631317316, "grad_norm": 2.9918621516440327, "learning_rate": 9.742153449030639e-06, "loss": 0.696, "step": 465 }, { "epoch": 0.19469759734879868, "grad_norm": 2.7984841744334785, "learning_rate": 9.730567331100333e-06, "loss": 0.7109, "step": 470 }, { "epoch": 0.19676884838442418, "grad_norm": 2.8102695071499166, "learning_rate": 9.71873379508136e-06, "loss": 0.6933, "step": 475 }, { "epoch": 0.1988400994200497, "grad_norm": 2.94185661750108, "learning_rate": 9.706653459891207e-06, "loss": 0.6929, "step": 480 }, { "epoch": 0.20091135045567524, "grad_norm": 2.8883338708822968, "learning_rate": 9.694326957355452e-06, "loss": 0.7139, "step": 485 }, { "epoch": 0.20298260149130073, "grad_norm": 2.7899074917015745, "learning_rate": 9.681754932174719e-06, "loss": 0.6957, "step": 490 }, { "epoch": 0.20505385252692626, "grad_norm": 2.8237012982336833, "learning_rate": 9.668938041890952e-06, "loss": 0.692, "step": 495 }, { "epoch": 0.2071251035625518, "grad_norm": 2.743263934485085, "learning_rate": 9.655876956853025e-06, "loss": 0.7125, "step": 500 }, { "epoch": 0.2091963545981773, "grad_norm": 2.7780277306316723, "learning_rate": 9.64257236018169e-06, "loss": 0.6769, "step": 505 }, { "epoch": 0.2112676056338028, "grad_norm": 2.794978739860878, "learning_rate": 9.629024947733836e-06, "loss": 0.6666, "step": 510 }, { "epoch": 0.21333885666942834, "grad_norm": 2.745567236605668, "learning_rate": 9.615235428066106e-06, "loss": 0.6934, "step": 515 }, { "epoch": 0.21541010770505387, "grad_norm": 2.8531534339381097, "learning_rate": 9.601204522397826e-06, "loss": 0.6711, "step": 520 }, { "epoch": 0.21748135874067936, "grad_norm": 2.675335493590183, "learning_rate": 9.586932964573298e-06, "loss": 0.6532, "step": 525 }, { "epoch": 0.2195526097763049, "grad_norm": 2.669778389943462, "learning_rate": 9.572421501023403e-06, "loss": 0.6633, "step": 530 }, { "epoch": 0.22162386081193042, "grad_norm": 2.7963996187147915, "learning_rate": 9.557670890726576e-06, "loss": 0.6568, "step": 535 }, { "epoch": 0.22369511184755592, "grad_norm": 2.597018418041498, "learning_rate": 9.5426819051691e-06, "loss": 0.6696, "step": 540 }, { "epoch": 0.22576636288318144, "grad_norm": 2.7931222599950556, "learning_rate": 9.527455328304756e-06, "loss": 0.6633, "step": 545 }, { "epoch": 0.22783761391880697, "grad_norm": 2.889773068112702, "learning_rate": 9.511991956513828e-06, "loss": 0.6597, "step": 550 }, { "epoch": 0.22990886495443247, "grad_norm": 2.9057842019365223, "learning_rate": 9.496292598561445e-06, "loss": 0.6557, "step": 555 }, { "epoch": 0.231980115990058, "grad_norm": 2.738062408705336, "learning_rate": 9.480358075555278e-06, "loss": 0.6588, "step": 560 }, { "epoch": 0.23405136702568352, "grad_norm": 2.7070986484262907, "learning_rate": 9.464189220902603e-06, "loss": 0.6521, "step": 565 }, { "epoch": 0.23612261806130902, "grad_norm": 2.659861286263656, "learning_rate": 9.447786880266706e-06, "loss": 0.6657, "step": 570 }, { "epoch": 0.23819386909693455, "grad_norm": 2.7250348611867445, "learning_rate": 9.431151911522656e-06, "loss": 0.6726, "step": 575 }, { "epoch": 0.24026512013256007, "grad_norm": 2.842184575287412, "learning_rate": 9.414285184712432e-06, "loss": 0.6365, "step": 580 }, { "epoch": 0.24233637116818557, "grad_norm": 2.67728938333087, "learning_rate": 9.397187581999424e-06, "loss": 0.6423, "step": 585 }, { "epoch": 0.2444076222038111, "grad_norm": 2.6563429501337787, "learning_rate": 9.37985999762229e-06, "loss": 0.6551, "step": 590 }, { "epoch": 0.24647887323943662, "grad_norm": 2.7799367526121577, "learning_rate": 9.362303337848188e-06, "loss": 0.6421, "step": 595 }, { "epoch": 0.24855012427506215, "grad_norm": 2.805960537561276, "learning_rate": 9.344518520925377e-06, "loss": 0.6698, "step": 600 }, { "epoch": 0.2506213753106877, "grad_norm": 2.624600321091577, "learning_rate": 9.326506477035179e-06, "loss": 0.6401, "step": 605 }, { "epoch": 0.2526926263463132, "grad_norm": 2.8754259673720166, "learning_rate": 9.308268148243355e-06, "loss": 0.6398, "step": 610 }, { "epoch": 0.2547638773819387, "grad_norm": 2.5919249587146864, "learning_rate": 9.289804488450805e-06, "loss": 0.6246, "step": 615 }, { "epoch": 0.25683512841756423, "grad_norm": 2.8063112887664983, "learning_rate": 9.271116463343692e-06, "loss": 0.639, "step": 620 }, { "epoch": 0.25890637945318973, "grad_norm": 2.756087197880415, "learning_rate": 9.25220505034293e-06, "loss": 0.6345, "step": 625 }, { "epoch": 0.2609776304888152, "grad_norm": 2.730342779651488, "learning_rate": 9.23307123855307e-06, "loss": 0.6387, "step": 630 }, { "epoch": 0.2630488815244408, "grad_norm": 2.556975883297732, "learning_rate": 9.213716028710558e-06, "loss": 0.6048, "step": 635 }, { "epoch": 0.2651201325600663, "grad_norm": 2.757235655234476, "learning_rate": 9.194140433131397e-06, "loss": 0.6108, "step": 640 }, { "epoch": 0.2671913835956918, "grad_norm": 2.6914526094035294, "learning_rate": 9.174345475658208e-06, "loss": 0.6312, "step": 645 }, { "epoch": 0.26926263463131733, "grad_norm": 2.8844859146955004, "learning_rate": 9.154332191606671e-06, "loss": 0.6149, "step": 650 }, { "epoch": 0.27133388566694283, "grad_norm": 2.6643656724584526, "learning_rate": 9.134101627711384e-06, "loss": 0.6266, "step": 655 }, { "epoch": 0.27340513670256833, "grad_norm": 2.706291314368935, "learning_rate": 9.113654842071114e-06, "loss": 0.6036, "step": 660 }, { "epoch": 0.2754763877381939, "grad_norm": 2.7891110506251486, "learning_rate": 9.092992904093451e-06, "loss": 0.6234, "step": 665 }, { "epoch": 0.2775476387738194, "grad_norm": 2.5789918392866835, "learning_rate": 9.072116894438885e-06, "loss": 0.6059, "step": 670 }, { "epoch": 0.2796188898094449, "grad_norm": 2.7333577490037917, "learning_rate": 9.051027904964279e-06, "loss": 0.622, "step": 675 }, { "epoch": 0.28169014084507044, "grad_norm": 3.0491907339310784, "learning_rate": 9.029727038665765e-06, "loss": 0.6224, "step": 680 }, { "epoch": 0.28376139188069593, "grad_norm": 2.804892423129682, "learning_rate": 9.008215409621053e-06, "loss": 0.6079, "step": 685 }, { "epoch": 0.28583264291632143, "grad_norm": 2.6893102664034725, "learning_rate": 8.986494142931168e-06, "loss": 0.5979, "step": 690 }, { "epoch": 0.287903893951947, "grad_norm": 2.69250215446606, "learning_rate": 8.964564374661597e-06, "loss": 0.601, "step": 695 }, { "epoch": 0.2899751449875725, "grad_norm": 2.745549147408966, "learning_rate": 8.94242725178288e-06, "loss": 0.6047, "step": 700 }, { "epoch": 0.292046396023198, "grad_norm": 2.814513963081923, "learning_rate": 8.920083932110608e-06, "loss": 0.5816, "step": 705 }, { "epoch": 0.29411764705882354, "grad_norm": 2.9268622951553978, "learning_rate": 8.89753558424488e-06, "loss": 0.6131, "step": 710 }, { "epoch": 0.29618889809444904, "grad_norm": 2.8629295406095427, "learning_rate": 8.874783387509181e-06, "loss": 0.5825, "step": 715 }, { "epoch": 0.2982601491300746, "grad_norm": 2.862631696140764, "learning_rate": 8.851828531888692e-06, "loss": 0.586, "step": 720 }, { "epoch": 0.3003314001657001, "grad_norm": 2.6450707993957407, "learning_rate": 8.828672217968055e-06, "loss": 0.6044, "step": 725 }, { "epoch": 0.3024026512013256, "grad_norm": 2.7777839624343894, "learning_rate": 8.805315656868587e-06, "loss": 0.5887, "step": 730 }, { "epoch": 0.30447390223695114, "grad_norm": 3.079711451209696, "learning_rate": 8.781760070184933e-06, "loss": 0.5967, "step": 735 }, { "epoch": 0.30654515327257664, "grad_norm": 2.6470021775768733, "learning_rate": 8.75800668992117e-06, "loss": 0.6012, "step": 740 }, { "epoch": 0.30861640430820214, "grad_norm": 2.632158539097956, "learning_rate": 8.734056758426367e-06, "loss": 0.5807, "step": 745 }, { "epoch": 0.3106876553438277, "grad_norm": 2.7108175778655594, "learning_rate": 8.709911528329623e-06, "loss": 0.5874, "step": 750 }, { "epoch": 0.3127589063794532, "grad_norm": 2.676873121568127, "learning_rate": 8.685572262474538e-06, "loss": 0.5902, "step": 755 }, { "epoch": 0.3148301574150787, "grad_norm": 2.7375620980248745, "learning_rate": 8.661040233853166e-06, "loss": 0.5559, "step": 760 }, { "epoch": 0.31690140845070425, "grad_norm": 2.8490016296800196, "learning_rate": 8.636316725539445e-06, "loss": 0.5809, "step": 765 }, { "epoch": 0.31897265948632975, "grad_norm": 2.699127585749091, "learning_rate": 8.611403030622074e-06, "loss": 0.5934, "step": 770 }, { "epoch": 0.32104391052195524, "grad_norm": 2.7937924462914463, "learning_rate": 8.586300452136895e-06, "loss": 0.5809, "step": 775 }, { "epoch": 0.3231151615575808, "grad_norm": 2.769306106741451, "learning_rate": 8.561010302998734e-06, "loss": 0.5936, "step": 780 }, { "epoch": 0.3251864125932063, "grad_norm": 2.7727931886537327, "learning_rate": 8.535533905932739e-06, "loss": 0.5877, "step": 785 }, { "epoch": 0.3272576636288318, "grad_norm": 2.7160030705029463, "learning_rate": 8.509872593405189e-06, "loss": 0.577, "step": 790 }, { "epoch": 0.32932891466445735, "grad_norm": 2.8193068873263147, "learning_rate": 8.484027707553818e-06, "loss": 0.5849, "step": 795 }, { "epoch": 0.33140016570008285, "grad_norm": 2.6677224808649624, "learning_rate": 8.458000600117604e-06, "loss": 0.5769, "step": 800 }, { "epoch": 0.33347141673570835, "grad_norm": 2.7583639483208295, "learning_rate": 8.43179263236608e-06, "loss": 0.5692, "step": 805 }, { "epoch": 0.3355426677713339, "grad_norm": 2.6729706959933357, "learning_rate": 8.40540517502813e-06, "loss": 0.5698, "step": 810 }, { "epoch": 0.3376139188069594, "grad_norm": 2.641297931390912, "learning_rate": 8.378839608220304e-06, "loss": 0.5688, "step": 815 }, { "epoch": 0.3396851698425849, "grad_norm": 2.569908946691123, "learning_rate": 8.35209732137463e-06, "loss": 0.564, "step": 820 }, { "epoch": 0.34175642087821045, "grad_norm": 2.7384094120020728, "learning_rate": 8.32517971316595e-06, "loss": 0.5629, "step": 825 }, { "epoch": 0.34382767191383595, "grad_norm": 2.703316718621585, "learning_rate": 8.298088191438753e-06, "loss": 0.5651, "step": 830 }, { "epoch": 0.34589892294946145, "grad_norm": 2.660556164602584, "learning_rate": 8.270824173133563e-06, "loss": 0.5609, "step": 835 }, { "epoch": 0.347970173985087, "grad_norm": 2.60747703062317, "learning_rate": 8.243389084212808e-06, "loss": 0.5513, "step": 840 }, { "epoch": 0.3500414250207125, "grad_norm": 2.717999035175483, "learning_rate": 8.215784359586257e-06, "loss": 0.5576, "step": 845 }, { "epoch": 0.352112676056338, "grad_norm": 2.6267818958939544, "learning_rate": 8.188011443035962e-06, "loss": 0.5685, "step": 850 }, { "epoch": 0.35418392709196356, "grad_norm": 2.6776314235505687, "learning_rate": 8.160071787140742e-06, "loss": 0.5645, "step": 855 }, { "epoch": 0.35625517812758906, "grad_norm": 2.6505501839520194, "learning_rate": 8.131966853200226e-06, "loss": 0.5397, "step": 860 }, { "epoch": 0.35832642916321455, "grad_norm": 2.840835300889455, "learning_rate": 8.103698111158405e-06, "loss": 0.5545, "step": 865 }, { "epoch": 0.3603976801988401, "grad_norm": 2.6348466544480234, "learning_rate": 8.075267039526764e-06, "loss": 0.5413, "step": 870 }, { "epoch": 0.3624689312344656, "grad_norm": 2.7064544291801, "learning_rate": 8.046675125306948e-06, "loss": 0.5324, "step": 875 }, { "epoch": 0.36454018227009116, "grad_norm": 2.533264278841598, "learning_rate": 8.017923863912989e-06, "loss": 0.5439, "step": 880 }, { "epoch": 0.36661143330571666, "grad_norm": 2.5698740147958636, "learning_rate": 7.989014759093095e-06, "loss": 0.5234, "step": 885 }, { "epoch": 0.36868268434134216, "grad_norm": 2.650820323691549, "learning_rate": 7.959949322850994e-06, "loss": 0.5368, "step": 890 }, { "epoch": 0.3707539353769677, "grad_norm": 2.5026566853575574, "learning_rate": 7.930729075366867e-06, "loss": 0.5303, "step": 895 }, { "epoch": 0.3728251864125932, "grad_norm": 2.606393560750939, "learning_rate": 7.901355544917827e-06, "loss": 0.5349, "step": 900 }, { "epoch": 0.3748964374482187, "grad_norm": 2.7177209606874837, "learning_rate": 7.87183026779799e-06, "loss": 0.5322, "step": 905 }, { "epoch": 0.37696768848384427, "grad_norm": 2.7325337591856482, "learning_rate": 7.842154788238124e-06, "loss": 0.5411, "step": 910 }, { "epoch": 0.37903893951946976, "grad_norm": 2.633934908338642, "learning_rate": 7.812330658324884e-06, "loss": 0.5452, "step": 915 }, { "epoch": 0.38111019055509526, "grad_norm": 2.5378522460092197, "learning_rate": 7.782359437919644e-06, "loss": 0.5312, "step": 920 }, { "epoch": 0.3831814415907208, "grad_norm": 2.594725922812132, "learning_rate": 7.75224269457689e-06, "loss": 0.5349, "step": 925 }, { "epoch": 0.3852526926263463, "grad_norm": 2.5768563073476822, "learning_rate": 7.721982003462255e-06, "loss": 0.5426, "step": 930 }, { "epoch": 0.3873239436619718, "grad_norm": 2.7238722942400884, "learning_rate": 7.691578947270122e-06, "loss": 0.5307, "step": 935 }, { "epoch": 0.38939519469759737, "grad_norm": 2.511458813651525, "learning_rate": 7.661035116140856e-06, "loss": 0.5147, "step": 940 }, { "epoch": 0.39146644573322287, "grad_norm": 2.7270220375577066, "learning_rate": 7.63035210757763e-06, "loss": 0.5052, "step": 945 }, { "epoch": 0.39353769676884837, "grad_norm": 2.623071761450853, "learning_rate": 7.599531526362873e-06, "loss": 0.525, "step": 950 }, { "epoch": 0.3956089478044739, "grad_norm": 2.6264824183942213, "learning_rate": 7.568574984474335e-06, "loss": 0.5224, "step": 955 }, { "epoch": 0.3976801988400994, "grad_norm": 2.6012679357411086, "learning_rate": 7.537484101000787e-06, "loss": 0.5255, "step": 960 }, { "epoch": 0.3997514498757249, "grad_norm": 2.55160475771275, "learning_rate": 7.506260502057325e-06, "loss": 0.5309, "step": 965 }, { "epoch": 0.40182270091135047, "grad_norm": 2.4346011440951627, "learning_rate": 7.474905820700334e-06, "loss": 0.5064, "step": 970 }, { "epoch": 0.40389395194697597, "grad_norm": 2.5979458980564867, "learning_rate": 7.443421696842066e-06, "loss": 0.5094, "step": 975 }, { "epoch": 0.40596520298260147, "grad_norm": 2.7888112093300643, "learning_rate": 7.411809777164873e-06, "loss": 0.5183, "step": 980 }, { "epoch": 0.408036454018227, "grad_norm": 2.4885203243420735, "learning_rate": 7.380071715035089e-06, "loss": 0.5249, "step": 985 }, { "epoch": 0.4101077050538525, "grad_norm": 2.7714914857659965, "learning_rate": 7.3482091704165405e-06, "loss": 0.5073, "step": 990 }, { "epoch": 0.412178956089478, "grad_norm": 2.614607486361334, "learning_rate": 7.316223809783745e-06, "loss": 0.5253, "step": 995 }, { "epoch": 0.4142502071251036, "grad_norm": 2.5290816828485054, "learning_rate": 7.284117306034733e-06, "loss": 0.498, "step": 1000 }, { "epoch": 0.4163214581607291, "grad_norm": 2.658163333493523, "learning_rate": 7.2518913384035685e-06, "loss": 0.517, "step": 1005 }, { "epoch": 0.4183927091963546, "grad_norm": 2.635643720493481, "learning_rate": 7.219547592372512e-06, "loss": 0.4942, "step": 1010 }, { "epoch": 0.4204639602319801, "grad_norm": 2.4911628142287903, "learning_rate": 7.187087759583869e-06, "loss": 0.4977, "step": 1015 }, { "epoch": 0.4225352112676056, "grad_norm": 2.5975015336888148, "learning_rate": 7.15451353775151e-06, "loss": 0.5091, "step": 1020 }, { "epoch": 0.4246064623032311, "grad_norm": 2.5595125829959224, "learning_rate": 7.121826630572084e-06, "loss": 0.4985, "step": 1025 }, { "epoch": 0.4266777133388567, "grad_norm": 2.5325553784547874, "learning_rate": 7.089028747635908e-06, "loss": 0.4966, "step": 1030 }, { "epoch": 0.4287489643744822, "grad_norm": 2.52826031642588, "learning_rate": 7.056121604337554e-06, "loss": 0.4943, "step": 1035 }, { "epoch": 0.43082021541010773, "grad_norm": 2.427352999890359, "learning_rate": 7.023106921786118e-06, "loss": 0.5069, "step": 1040 }, { "epoch": 0.43289146644573323, "grad_norm": 2.775128162589827, "learning_rate": 6.9899864267152275e-06, "loss": 0.4971, "step": 1045 }, { "epoch": 0.43496271748135873, "grad_norm": 2.461227676382275, "learning_rate": 6.956761851392706e-06, "loss": 0.5014, "step": 1050 }, { "epoch": 0.4370339685169843, "grad_norm": 2.5269545040942916, "learning_rate": 6.9234349335299835e-06, "loss": 0.503, "step": 1055 }, { "epoch": 0.4391052195526098, "grad_norm": 2.394453478310078, "learning_rate": 6.890007416191209e-06, "loss": 0.4973, "step": 1060 }, { "epoch": 0.4411764705882353, "grad_norm": 2.403311053474268, "learning_rate": 6.8564810477020835e-06, "loss": 0.4625, "step": 1065 }, { "epoch": 0.44324772162386084, "grad_norm": 2.547625977515838, "learning_rate": 6.822857581558423e-06, "loss": 0.4977, "step": 1070 }, { "epoch": 0.44531897265948633, "grad_norm": 2.4609332754827564, "learning_rate": 6.789138776334441e-06, "loss": 0.4956, "step": 1075 }, { "epoch": 0.44739022369511183, "grad_norm": 2.4858053793498525, "learning_rate": 6.7553263955907755e-06, "loss": 0.4941, "step": 1080 }, { "epoch": 0.4494614747307374, "grad_norm": 2.4015768436682654, "learning_rate": 6.721422207782249e-06, "loss": 0.4776, "step": 1085 }, { "epoch": 0.4515327257663629, "grad_norm": 2.863324693290168, "learning_rate": 6.687427986165379e-06, "loss": 0.478, "step": 1090 }, { "epoch": 0.4536039768019884, "grad_norm": 2.471162662938083, "learning_rate": 6.653345508705629e-06, "loss": 0.4878, "step": 1095 }, { "epoch": 0.45567522783761394, "grad_norm": 2.4277728219037087, "learning_rate": 6.6191765579844205e-06, "loss": 0.4805, "step": 1100 }, { "epoch": 0.45774647887323944, "grad_norm": 2.678570884207627, "learning_rate": 6.584922921105894e-06, "loss": 0.4781, "step": 1105 }, { "epoch": 0.45981772990886494, "grad_norm": 2.532089285295406, "learning_rate": 6.550586389603451e-06, "loss": 0.4687, "step": 1110 }, { "epoch": 0.4618889809444905, "grad_norm": 2.5649088008217804, "learning_rate": 6.5161687593460395e-06, "loss": 0.4692, "step": 1115 }, { "epoch": 0.463960231980116, "grad_norm": 2.4747899161662548, "learning_rate": 6.481671830444243e-06, "loss": 0.4904, "step": 1120 }, { "epoch": 0.4660314830157415, "grad_norm": 2.529769284750395, "learning_rate": 6.447097407156114e-06, "loss": 0.4777, "step": 1125 }, { "epoch": 0.46810273405136704, "grad_norm": 2.4796272753405266, "learning_rate": 6.412447297792818e-06, "loss": 0.4712, "step": 1130 }, { "epoch": 0.47017398508699254, "grad_norm": 2.486729070834996, "learning_rate": 6.377723314624057e-06, "loss": 0.4733, "step": 1135 }, { "epoch": 0.47224523612261804, "grad_norm": 2.447720009959368, "learning_rate": 6.3429272737832726e-06, "loss": 0.4817, "step": 1140 }, { "epoch": 0.4743164871582436, "grad_norm": 2.475066169419796, "learning_rate": 6.308060995172673e-06, "loss": 0.4446, "step": 1145 }, { "epoch": 0.4763877381938691, "grad_norm": 2.604837459035885, "learning_rate": 6.273126302368037e-06, "loss": 0.4527, "step": 1150 }, { "epoch": 0.4784589892294946, "grad_norm": 2.626085677583465, "learning_rate": 6.238125022523343e-06, "loss": 0.4731, "step": 1155 }, { "epoch": 0.48053024026512015, "grad_norm": 2.489403917845554, "learning_rate": 6.203058986275207e-06, "loss": 0.4693, "step": 1160 }, { "epoch": 0.48260149130074564, "grad_norm": 2.5408918650775423, "learning_rate": 6.1679300276471285e-06, "loss": 0.4703, "step": 1165 }, { "epoch": 0.48467274233637114, "grad_norm": 2.5741432676541023, "learning_rate": 6.132739983953579e-06, "loss": 0.4574, "step": 1170 }, { "epoch": 0.4867439933719967, "grad_norm": 2.5055987629878875, "learning_rate": 6.097490695703896e-06, "loss": 0.4598, "step": 1175 }, { "epoch": 0.4888152444076222, "grad_norm": 2.599659752996103, "learning_rate": 6.062184006506027e-06, "loss": 0.4616, "step": 1180 }, { "epoch": 0.4908864954432477, "grad_norm": 2.5766536661712345, "learning_rate": 6.026821762970102e-06, "loss": 0.4655, "step": 1185 }, { "epoch": 0.49295774647887325, "grad_norm": 2.576398426328721, "learning_rate": 5.991405814611855e-06, "loss": 0.455, "step": 1190 }, { "epoch": 0.49502899751449875, "grad_norm": 2.378065054333025, "learning_rate": 5.955938013755888e-06, "loss": 0.4657, "step": 1195 }, { "epoch": 0.4971002485501243, "grad_norm": 2.5777328764366843, "learning_rate": 5.920420215438794e-06, "loss": 0.4536, "step": 1200 }, { "epoch": 0.4991714995857498, "grad_norm": 2.590032178484324, "learning_rate": 5.8848542773121285e-06, "loss": 0.4709, "step": 1205 }, { "epoch": 0.5012427506213754, "grad_norm": 2.6721053688454908, "learning_rate": 5.849242059545259e-06, "loss": 0.4536, "step": 1210 }, { "epoch": 0.5033140016570008, "grad_norm": 2.492216069957813, "learning_rate": 5.81358542472807e-06, "loss": 0.4565, "step": 1215 }, { "epoch": 0.5053852526926264, "grad_norm": 2.5630596324380224, "learning_rate": 5.777886237773542e-06, "loss": 0.4701, "step": 1220 }, { "epoch": 0.5074565037282519, "grad_norm": 2.507245052161205, "learning_rate": 5.742146365820223e-06, "loss": 0.453, "step": 1225 }, { "epoch": 0.5095277547638773, "grad_norm": 2.4042777020925645, "learning_rate": 5.706367678134562e-06, "loss": 0.453, "step": 1230 }, { "epoch": 0.5115990057995029, "grad_norm": 2.5265503464847328, "learning_rate": 5.670552046013151e-06, "loss": 0.4502, "step": 1235 }, { "epoch": 0.5136702568351285, "grad_norm": 2.8403722966962284, "learning_rate": 5.634701342684852e-06, "loss": 0.451, "step": 1240 }, { "epoch": 0.5157415078707539, "grad_norm": 2.5437149212702392, "learning_rate": 5.598817443212813e-06, "loss": 0.4357, "step": 1245 }, { "epoch": 0.5178127589063795, "grad_norm": 2.5209421401735272, "learning_rate": 5.562902224396416e-06, "loss": 0.4476, "step": 1250 }, { "epoch": 0.519884009942005, "grad_norm": 2.406037647219463, "learning_rate": 5.526957564673098e-06, "loss": 0.438, "step": 1255 }, { "epoch": 0.5219552609776305, "grad_norm": 2.44986117662171, "learning_rate": 5.49098534402012e-06, "loss": 0.4353, "step": 1260 }, { "epoch": 0.524026512013256, "grad_norm": 2.496524011645464, "learning_rate": 5.454987443856235e-06, "loss": 0.4415, "step": 1265 }, { "epoch": 0.5260977630488816, "grad_norm": 2.4598099313435675, "learning_rate": 5.418965746943281e-06, "loss": 0.427, "step": 1270 }, { "epoch": 0.528169014084507, "grad_norm": 2.51587918880957, "learning_rate": 5.3829221372877175e-06, "loss": 0.4432, "step": 1275 }, { "epoch": 0.5302402651201326, "grad_norm": 2.385525721474367, "learning_rate": 5.34685850004208e-06, "loss": 0.446, "step": 1280 }, { "epoch": 0.5323115161557581, "grad_norm": 2.6024656561083237, "learning_rate": 5.310776721406392e-06, "loss": 0.4383, "step": 1285 }, { "epoch": 0.5343827671913836, "grad_norm": 2.478361664190618, "learning_rate": 5.2746786885295034e-06, "loss": 0.4469, "step": 1290 }, { "epoch": 0.5364540182270091, "grad_norm": 2.462559560068187, "learning_rate": 5.238566289410396e-06, "loss": 0.4445, "step": 1295 }, { "epoch": 0.5385252692626347, "grad_norm": 2.5652360752702705, "learning_rate": 5.2024414127994325e-06, "loss": 0.4268, "step": 1300 }, { "epoch": 0.5405965202982601, "grad_norm": 2.479268879286496, "learning_rate": 5.166305948099574e-06, "loss": 0.4411, "step": 1305 }, { "epoch": 0.5426677713338857, "grad_norm": 2.4814080548859536, "learning_rate": 5.13016178526756e-06, "loss": 0.4337, "step": 1310 }, { "epoch": 0.5447390223695112, "grad_norm": 2.518581213893244, "learning_rate": 5.094010814715062e-06, "loss": 0.4361, "step": 1315 }, { "epoch": 0.5468102734051367, "grad_norm": 2.375407800118248, "learning_rate": 5.057854927209804e-06, "loss": 0.4333, "step": 1320 }, { "epoch": 0.5488815244407622, "grad_norm": 2.4443582316359582, "learning_rate": 5.0216960137766805e-06, "loss": 0.4253, "step": 1325 }, { "epoch": 0.5509527754763878, "grad_norm": 2.5441411002094783, "learning_rate": 4.985535965598843e-06, "loss": 0.4305, "step": 1330 }, { "epoch": 0.5530240265120132, "grad_norm": 2.4305764050134955, "learning_rate": 4.949376673918802e-06, "loss": 0.4248, "step": 1335 }, { "epoch": 0.5550952775476388, "grad_norm": 2.5334727900548994, "learning_rate": 4.913220029939491e-06, "loss": 0.4141, "step": 1340 }, { "epoch": 0.5571665285832643, "grad_norm": 2.5234820724710683, "learning_rate": 4.877067924725368e-06, "loss": 0.4332, "step": 1345 }, { "epoch": 0.5592377796188898, "grad_norm": 2.4799505423568946, "learning_rate": 4.840922249103506e-06, "loss": 0.4318, "step": 1350 }, { "epoch": 0.5613090306545153, "grad_norm": 2.5121729317010164, "learning_rate": 4.804784893564697e-06, "loss": 0.4389, "step": 1355 }, { "epoch": 0.5633802816901409, "grad_norm": 2.3947951693271303, "learning_rate": 4.7686577481645745e-06, "loss": 0.4143, "step": 1360 }, { "epoch": 0.5654515327257663, "grad_norm": 2.5337735975104194, "learning_rate": 4.732542702424759e-06, "loss": 0.4289, "step": 1365 }, { "epoch": 0.5675227837613919, "grad_norm": 2.434769812531381, "learning_rate": 4.696441645234042e-06, "loss": 0.4236, "step": 1370 }, { "epoch": 0.5695940347970174, "grad_norm": 2.5167542766488804, "learning_rate": 4.660356464749578e-06, "loss": 0.4306, "step": 1375 }, { "epoch": 0.5716652858326429, "grad_norm": 2.3974566625882336, "learning_rate": 4.624289048298147e-06, "loss": 0.4114, "step": 1380 }, { "epoch": 0.5737365368682684, "grad_norm": 2.385989686167348, "learning_rate": 4.588241282277428e-06, "loss": 0.4222, "step": 1385 }, { "epoch": 0.575807787903894, "grad_norm": 2.3379950890042136, "learning_rate": 4.55221505205734e-06, "loss": 0.4167, "step": 1390 }, { "epoch": 0.5778790389395194, "grad_norm": 2.6245882848438544, "learning_rate": 4.516212241881448e-06, "loss": 0.4096, "step": 1395 }, { "epoch": 0.579950289975145, "grad_norm": 2.283991444400329, "learning_rate": 4.480234734768393e-06, "loss": 0.4165, "step": 1400 }, { "epoch": 0.5820215410107705, "grad_norm": 2.498405431139293, "learning_rate": 4.444284412413418e-06, "loss": 0.4138, "step": 1405 }, { "epoch": 0.584092792046396, "grad_norm": 2.355256842651903, "learning_rate": 4.408363155089952e-06, "loss": 0.4086, "step": 1410 }, { "epoch": 0.5861640430820215, "grad_norm": 2.511272069011922, "learning_rate": 4.3724728415512585e-06, "loss": 0.4173, "step": 1415 }, { "epoch": 0.5882352941176471, "grad_norm": 2.5460297743905187, "learning_rate": 4.3366153489321855e-06, "loss": 0.4049, "step": 1420 }, { "epoch": 0.5903065451532725, "grad_norm": 2.5260852357481927, "learning_rate": 4.30079255265098e-06, "loss": 0.4194, "step": 1425 }, { "epoch": 0.5923777961888981, "grad_norm": 2.3738308604652167, "learning_rate": 4.265006326311199e-06, "loss": 0.4124, "step": 1430 }, { "epoch": 0.5944490472245236, "grad_norm": 2.370218652591744, "learning_rate": 4.229258541603723e-06, "loss": 0.3977, "step": 1435 }, { "epoch": 0.5965202982601492, "grad_norm": 2.284788920612559, "learning_rate": 4.1935510682088545e-06, "loss": 0.4098, "step": 1440 }, { "epoch": 0.5985915492957746, "grad_norm": 2.3819155019816303, "learning_rate": 4.157885773698535e-06, "loss": 0.4007, "step": 1445 }, { "epoch": 0.6006628003314002, "grad_norm": 2.439446100087243, "learning_rate": 4.122264523438668e-06, "loss": 0.4002, "step": 1450 }, { "epoch": 0.6027340513670257, "grad_norm": 2.3813825718286883, "learning_rate": 4.086689180491554e-06, "loss": 0.3994, "step": 1455 }, { "epoch": 0.6048053024026512, "grad_norm": 2.408250839202308, "learning_rate": 4.051161605518453e-06, "loss": 0.3975, "step": 1460 }, { "epoch": 0.6068765534382767, "grad_norm": 2.452576823642211, "learning_rate": 4.015683656682255e-06, "loss": 0.406, "step": 1465 }, { "epoch": 0.6089478044739023, "grad_norm": 2.384778355857051, "learning_rate": 3.980257189550316e-06, "loss": 0.3991, "step": 1470 }, { "epoch": 0.6110190555095277, "grad_norm": 2.288031466247923, "learning_rate": 3.94488405699739e-06, "loss": 0.3841, "step": 1475 }, { "epoch": 0.6130903065451533, "grad_norm": 2.4325130154405263, "learning_rate": 3.909566109108727e-06, "loss": 0.4052, "step": 1480 }, { "epoch": 0.6151615575807788, "grad_norm": 2.4621406129863526, "learning_rate": 3.874305193083313e-06, "loss": 0.4064, "step": 1485 }, { "epoch": 0.6172328086164043, "grad_norm": 2.4148001556880123, "learning_rate": 3.839103153137247e-06, "loss": 0.4013, "step": 1490 }, { "epoch": 0.6193040596520298, "grad_norm": 2.527488147311755, "learning_rate": 3.803961830407297e-06, "loss": 0.4141, "step": 1495 }, { "epoch": 0.6213753106876554, "grad_norm": 2.4140596278075095, "learning_rate": 3.768883062854598e-06, "loss": 0.3958, "step": 1500 }, { "epoch": 0.6234465617232808, "grad_norm": 2.3671251798358925, "learning_rate": 3.7338686851685267e-06, "loss": 0.3954, "step": 1505 }, { "epoch": 0.6255178127589064, "grad_norm": 2.3443935871907984, "learning_rate": 3.6989205286707398e-06, "loss": 0.3883, "step": 1510 }, { "epoch": 0.6275890637945319, "grad_norm": 2.4179094736002735, "learning_rate": 3.664040421219393e-06, "loss": 0.3984, "step": 1515 }, { "epoch": 0.6296603148301574, "grad_norm": 2.3700384431761803, "learning_rate": 3.6292301871135425e-06, "loss": 0.4016, "step": 1520 }, { "epoch": 0.6317315658657829, "grad_norm": 2.330878967673718, "learning_rate": 3.59449164699773e-06, "loss": 0.3874, "step": 1525 }, { "epoch": 0.6338028169014085, "grad_norm": 2.375441813687718, "learning_rate": 3.55982661776676e-06, "loss": 0.3873, "step": 1530 }, { "epoch": 0.6358740679370339, "grad_norm": 2.432663080893473, "learning_rate": 3.5252369124706697e-06, "loss": 0.3826, "step": 1535 }, { "epoch": 0.6379453189726595, "grad_norm": 2.4211109265989346, "learning_rate": 3.4907243402199013e-06, "loss": 0.3917, "step": 1540 }, { "epoch": 0.640016570008285, "grad_norm": 2.383500236539082, "learning_rate": 3.4562907060906908e-06, "loss": 0.3871, "step": 1545 }, { "epoch": 0.6420878210439105, "grad_norm": 2.5020986029985353, "learning_rate": 3.4219378110306523e-06, "loss": 0.3843, "step": 1550 }, { "epoch": 0.644159072079536, "grad_norm": 2.411568393268961, "learning_rate": 3.3876674517645815e-06, "loss": 0.3851, "step": 1555 }, { "epoch": 0.6462303231151616, "grad_norm": 2.2871927774448113, "learning_rate": 3.353481420700495e-06, "loss": 0.3832, "step": 1560 }, { "epoch": 0.648301574150787, "grad_norm": 2.487402344766119, "learning_rate": 3.319381505835868e-06, "loss": 0.3808, "step": 1565 }, { "epoch": 0.6503728251864126, "grad_norm": 2.575422487090241, "learning_rate": 3.285369490664133e-06, "loss": 0.3937, "step": 1570 }, { "epoch": 0.6524440762220381, "grad_norm": 2.4411912546035794, "learning_rate": 3.251447154081394e-06, "loss": 0.3814, "step": 1575 }, { "epoch": 0.6545153272576636, "grad_norm": 2.278355028515059, "learning_rate": 3.2176162702933816e-06, "loss": 0.3805, "step": 1580 }, { "epoch": 0.6565865782932891, "grad_norm": 2.3844617654490508, "learning_rate": 3.183878608722669e-06, "loss": 0.381, "step": 1585 }, { "epoch": 0.6586578293289147, "grad_norm": 2.329561758304698, "learning_rate": 3.150235933916115e-06, "loss": 0.387, "step": 1590 }, { "epoch": 0.6607290803645401, "grad_norm": 2.4156074527478046, "learning_rate": 3.1166900054525873e-06, "loss": 0.3994, "step": 1595 }, { "epoch": 0.6628003314001657, "grad_norm": 2.3783657924844515, "learning_rate": 3.0832425778509235e-06, "loss": 0.3779, "step": 1600 }, { "epoch": 0.6648715824357913, "grad_norm": 2.4716803217245733, "learning_rate": 3.049895400478174e-06, "loss": 0.3732, "step": 1605 }, { "epoch": 0.6669428334714167, "grad_norm": 2.3458142528440136, "learning_rate": 3.0166502174581012e-06, "loss": 0.3752, "step": 1610 }, { "epoch": 0.6690140845070423, "grad_norm": 2.4198975838792585, "learning_rate": 2.983508767579956e-06, "loss": 0.3661, "step": 1615 }, { "epoch": 0.6710853355426678, "grad_norm": 2.372657886753349, "learning_rate": 2.950472784207544e-06, "loss": 0.3752, "step": 1620 }, { "epoch": 0.6731565865782932, "grad_norm": 2.281721990191394, "learning_rate": 2.917543995188562e-06, "loss": 0.3787, "step": 1625 }, { "epoch": 0.6752278376139188, "grad_norm": 2.491592795734995, "learning_rate": 2.8847241227642255e-06, "loss": 0.3835, "step": 1630 }, { "epoch": 0.6772990886495444, "grad_norm": 2.4831276366154054, "learning_rate": 2.852014883479198e-06, "loss": 0.3832, "step": 1635 }, { "epoch": 0.6793703396851698, "grad_norm": 2.403780522647955, "learning_rate": 2.819417988091814e-06, "loss": 0.384, "step": 1640 }, { "epoch": 0.6814415907207954, "grad_norm": 2.465619590474164, "learning_rate": 2.786935141484586e-06, "loss": 0.3749, "step": 1645 }, { "epoch": 0.6835128417564209, "grad_norm": 2.4202342922381668, "learning_rate": 2.754568042575061e-06, "loss": 0.3634, "step": 1650 }, { "epoch": 0.6855840927920464, "grad_norm": 2.4313702462550553, "learning_rate": 2.7223183842269442e-06, "loss": 0.3613, "step": 1655 }, { "epoch": 0.6876553438276719, "grad_norm": 2.3522602311047587, "learning_rate": 2.6901878531615677e-06, "loss": 0.377, "step": 1660 }, { "epoch": 0.6897265948632975, "grad_norm": 2.4828984837987518, "learning_rate": 2.658178129869672e-06, "loss": 0.3733, "step": 1665 }, { "epoch": 0.6917978458989229, "grad_norm": 2.447859748941827, "learning_rate": 2.6262908885235046e-06, "loss": 0.3756, "step": 1670 }, { "epoch": 0.6938690969345485, "grad_norm": 2.4042907104977544, "learning_rate": 2.594527796889265e-06, "loss": 0.3656, "step": 1675 }, { "epoch": 0.695940347970174, "grad_norm": 2.2665438118554233, "learning_rate": 2.5628905162398797e-06, "loss": 0.3714, "step": 1680 }, { "epoch": 0.6980115990057995, "grad_norm": 2.363177421824413, "learning_rate": 2.531380701268108e-06, "loss": 0.3638, "step": 1685 }, { "epoch": 0.700082850041425, "grad_norm": 2.203051752988435, "learning_rate": 2.5000000000000015e-06, "loss": 0.3644, "step": 1690 }, { "epoch": 0.7021541010770506, "grad_norm": 2.2614655636331973, "learning_rate": 2.4687500537087027e-06, "loss": 0.3664, "step": 1695 }, { "epoch": 0.704225352112676, "grad_norm": 2.312353684904146, "learning_rate": 2.4376324968286154e-06, "loss": 0.3749, "step": 1700 }, { "epoch": 0.7062966031483016, "grad_norm": 2.302063654985313, "learning_rate": 2.40664895686991e-06, "loss": 0.3531, "step": 1705 }, { "epoch": 0.7083678541839271, "grad_norm": 2.3982616088942987, "learning_rate": 2.375801054333409e-06, "loss": 0.3657, "step": 1710 }, { "epoch": 0.7104391052195526, "grad_norm": 2.3723480581164127, "learning_rate": 2.345090402625822e-06, "loss": 0.3621, "step": 1715 }, { "epoch": 0.7125103562551781, "grad_norm": 2.2668673481971755, "learning_rate": 2.3145186079753685e-06, "loss": 0.353, "step": 1720 }, { "epoch": 0.7145816072908037, "grad_norm": 2.331916407183293, "learning_rate": 2.2840872693477694e-06, "loss": 0.3446, "step": 1725 }, { "epoch": 0.7166528583264291, "grad_norm": 2.4260189666734897, "learning_rate": 2.253797978362617e-06, "loss": 0.3714, "step": 1730 }, { "epoch": 0.7187241093620547, "grad_norm": 2.772035063389101, "learning_rate": 2.2236523192101264e-06, "loss": 0.3756, "step": 1735 }, { "epoch": 0.7207953603976802, "grad_norm": 2.3445465173771303, "learning_rate": 2.193651868568285e-06, "loss": 0.3598, "step": 1740 }, { "epoch": 0.7228666114333057, "grad_norm": 2.4507276447056823, "learning_rate": 2.16379819552038e-06, "loss": 0.3534, "step": 1745 }, { "epoch": 0.7249378624689312, "grad_norm": 2.3286932117618897, "learning_rate": 2.1340928614729445e-06, "loss": 0.3676, "step": 1750 }, { "epoch": 0.7270091135045568, "grad_norm": 2.4278124699427432, "learning_rate": 2.1045374200740863e-06, "loss": 0.3478, "step": 1755 }, { "epoch": 0.7290803645401823, "grad_norm": 2.338964151182083, "learning_rate": 2.075133417132223e-06, "loss": 0.3497, "step": 1760 }, { "epoch": 0.7311516155758078, "grad_norm": 2.2860327377758516, "learning_rate": 2.045882390535248e-06, "loss": 0.3461, "step": 1765 }, { "epoch": 0.7332228666114333, "grad_norm": 2.440050304414306, "learning_rate": 2.016785870170079e-06, "loss": 0.3487, "step": 1770 }, { "epoch": 0.7352941176470589, "grad_norm": 2.336585306226817, "learning_rate": 1.987845377842656e-06, "loss": 0.3404, "step": 1775 }, { "epoch": 0.7373653686826843, "grad_norm": 2.4239291315507843, "learning_rate": 1.9590624271983406e-06, "loss": 0.3572, "step": 1780 }, { "epoch": 0.7394366197183099, "grad_norm": 2.2655747278638008, "learning_rate": 1.9304385236427505e-06, "loss": 0.3534, "step": 1785 }, { "epoch": 0.7415078707539354, "grad_norm": 2.411369709306088, "learning_rate": 1.9019751642630252e-06, "loss": 0.3588, "step": 1790 }, { "epoch": 0.7435791217895609, "grad_norm": 2.5661063715212515, "learning_rate": 1.8736738377495196e-06, "loss": 0.3707, "step": 1795 }, { "epoch": 0.7456503728251864, "grad_norm": 2.271714584404189, "learning_rate": 1.8455360243179537e-06, "loss": 0.3461, "step": 1800 }, { "epoch": 0.747721623860812, "grad_norm": 2.398528583226255, "learning_rate": 1.8175631956319823e-06, "loss": 0.357, "step": 1805 }, { "epoch": 0.7497928748964374, "grad_norm": 2.39827636942671, "learning_rate": 1.7897568147262323e-06, "loss": 0.3424, "step": 1810 }, { "epoch": 0.751864125932063, "grad_norm": 2.2477982364673523, "learning_rate": 1.7621183359297817e-06, "loss": 0.3443, "step": 1815 }, { "epoch": 0.7539353769676885, "grad_norm": 2.468469299919912, "learning_rate": 1.7346492047900897e-06, "loss": 0.3498, "step": 1820 }, { "epoch": 0.756006628003314, "grad_norm": 2.448144890103532, "learning_rate": 1.7073508579973996e-06, "loss": 0.3426, "step": 1825 }, { "epoch": 0.7580778790389395, "grad_norm": 2.168562497980907, "learning_rate": 1.6802247233095914e-06, "loss": 0.3449, "step": 1830 }, { "epoch": 0.7601491300745651, "grad_norm": 2.368438576336641, "learning_rate": 1.6532722194775108e-06, "loss": 0.3531, "step": 1835 }, { "epoch": 0.7622203811101905, "grad_norm": 2.3241440671031377, "learning_rate": 1.626494756170765e-06, "loss": 0.3553, "step": 1840 }, { "epoch": 0.7642916321458161, "grad_norm": 2.310085520479401, "learning_rate": 1.5998937339039889e-06, "loss": 0.3392, "step": 1845 }, { "epoch": 0.7663628831814416, "grad_norm": 2.2581792463298997, "learning_rate": 1.5734705439636017e-06, "loss": 0.3512, "step": 1850 }, { "epoch": 0.7684341342170671, "grad_norm": 2.439537688679, "learning_rate": 1.5472265683350397e-06, "loss": 0.3421, "step": 1855 }, { "epoch": 0.7705053852526926, "grad_norm": 2.5094842908311072, "learning_rate": 1.5211631796304721e-06, "loss": 0.3633, "step": 1860 }, { "epoch": 0.7725766362883182, "grad_norm": 2.265608557062427, "learning_rate": 1.495281741017016e-06, "loss": 0.333, "step": 1865 }, { "epoch": 0.7746478873239436, "grad_norm": 2.300719293992617, "learning_rate": 1.46958360614543e-06, "loss": 0.3469, "step": 1870 }, { "epoch": 0.7767191383595692, "grad_norm": 2.2656950757005574, "learning_rate": 1.4440701190793278e-06, "loss": 0.3516, "step": 1875 }, { "epoch": 0.7787903893951947, "grad_norm": 2.4327514419777065, "learning_rate": 1.4187426142248723e-06, "loss": 0.3377, "step": 1880 }, { "epoch": 0.7808616404308202, "grad_norm": 2.3520087841436053, "learning_rate": 1.3936024162609897e-06, "loss": 0.3472, "step": 1885 }, { "epoch": 0.7829328914664457, "grad_norm": 2.732996885339388, "learning_rate": 1.3686508400700787e-06, "loss": 0.351, "step": 1890 }, { "epoch": 0.7850041425020713, "grad_norm": 2.250710944434201, "learning_rate": 1.3438891906692447e-06, "loss": 0.3456, "step": 1895 }, { "epoch": 0.7870753935376967, "grad_norm": 2.2717407369401763, "learning_rate": 1.3193187631420462e-06, "loss": 0.3363, "step": 1900 }, { "epoch": 0.7891466445733223, "grad_norm": 2.581162442428822, "learning_rate": 1.2949408425707566e-06, "loss": 0.3509, "step": 1905 }, { "epoch": 0.7912178956089478, "grad_norm": 2.451318834794777, "learning_rate": 1.2707567039691505e-06, "loss": 0.3496, "step": 1910 }, { "epoch": 0.7932891466445733, "grad_norm": 2.3990332386613775, "learning_rate": 1.2467676122158224e-06, "loss": 0.3351, "step": 1915 }, { "epoch": 0.7953603976801988, "grad_norm": 2.495492531110974, "learning_rate": 1.222974821988024e-06, "loss": 0.3408, "step": 1920 }, { "epoch": 0.7974316487158244, "grad_norm": 2.3769169276951443, "learning_rate": 1.1993795776960498e-06, "loss": 0.3295, "step": 1925 }, { "epoch": 0.7995028997514498, "grad_norm": 2.4871113176142283, "learning_rate": 1.1759831134181504e-06, "loss": 0.3422, "step": 1930 }, { "epoch": 0.8015741507870754, "grad_norm": 2.3429331132949325, "learning_rate": 1.1527866528359805e-06, "loss": 0.3291, "step": 1935 }, { "epoch": 0.8036454018227009, "grad_norm": 2.450086674629505, "learning_rate": 1.1297914091706086e-06, "loss": 0.3377, "step": 1940 }, { "epoch": 0.8057166528583264, "grad_norm": 2.3124784250203807, "learning_rate": 1.1069985851190524e-06, "loss": 0.3397, "step": 1945 }, { "epoch": 0.8077879038939519, "grad_norm": 2.5121429598446836, "learning_rate": 1.0844093727913868e-06, "loss": 0.3297, "step": 1950 }, { "epoch": 0.8098591549295775, "grad_norm": 2.3886654298111156, "learning_rate": 1.062024953648384e-06, "loss": 0.3472, "step": 1955 }, { "epoch": 0.8119304059652029, "grad_norm": 2.416515049214574, "learning_rate": 1.039846498439727e-06, "loss": 0.3333, "step": 1960 }, { "epoch": 0.8140016570008285, "grad_norm": 2.292484990998401, "learning_rate": 1.0178751671427755e-06, "loss": 0.3354, "step": 1965 }, { "epoch": 0.816072908036454, "grad_norm": 2.3922313945952465, "learning_rate": 9.961121089018933e-07, "loss": 0.3257, "step": 1970 }, { "epoch": 0.8181441590720795, "grad_norm": 2.3117852804307124, "learning_rate": 9.745584619683524e-07, "loss": 0.3262, "step": 1975 }, { "epoch": 0.820215410107705, "grad_norm": 2.3782012557831567, "learning_rate": 9.532153536407923e-07, "loss": 0.3381, "step": 1980 }, { "epoch": 0.8222866611433306, "grad_norm": 2.316778024028577, "learning_rate": 9.320839002062682e-07, "loss": 0.329, "step": 1985 }, { "epoch": 0.824357912178956, "grad_norm": 2.4579677816004715, "learning_rate": 9.111652068818621e-07, "loss": 0.3365, "step": 1990 }, { "epoch": 0.8264291632145816, "grad_norm": 2.2909173606716555, "learning_rate": 8.904603677568785e-07, "loss": 0.327, "step": 1995 }, { "epoch": 0.8285004142502072, "grad_norm": 2.378065910884157, "learning_rate": 8.699704657356195e-07, "loss": 0.3272, "step": 2000 }, { "epoch": 0.8305716652858326, "grad_norm": 2.3648030957840187, "learning_rate": 8.496965724807516e-07, "loss": 0.3339, "step": 2005 }, { "epoch": 0.8326429163214581, "grad_norm": 2.136214101086001, "learning_rate": 8.296397483572515e-07, "loss": 0.3196, "step": 2010 }, { "epoch": 0.8347141673570837, "grad_norm": 2.4656930931044965, "learning_rate": 8.098010423769503e-07, "loss": 0.35, "step": 2015 }, { "epoch": 0.8367854183927091, "grad_norm": 2.3398867724259977, "learning_rate": 7.901814921436624e-07, "loss": 0.3326, "step": 2020 }, { "epoch": 0.8388566694283347, "grad_norm": 2.351803701234294, "learning_rate": 7.70782123798921e-07, "loss": 0.3273, "step": 2025 }, { "epoch": 0.8409279204639603, "grad_norm": 2.2776835231240184, "learning_rate": 7.516039519683105e-07, "loss": 0.3238, "step": 2030 }, { "epoch": 0.8429991714995857, "grad_norm": 2.3606143376947135, "learning_rate": 7.326479797083963e-07, "loss": 0.3297, "step": 2035 }, { "epoch": 0.8450704225352113, "grad_norm": 2.3999583975189056, "learning_rate": 7.139151984542636e-07, "loss": 0.3155, "step": 2040 }, { "epoch": 0.8471416735708368, "grad_norm": 2.324664898231089, "learning_rate": 6.954065879676653e-07, "loss": 0.327, "step": 2045 }, { "epoch": 0.8492129246064622, "grad_norm": 2.2630117680123103, "learning_rate": 6.771231162857722e-07, "loss": 0.3258, "step": 2050 }, { "epoch": 0.8512841756420878, "grad_norm": 2.351884420298746, "learning_rate": 6.590657396705525e-07, "loss": 0.3256, "step": 2055 }, { "epoch": 0.8533554266777134, "grad_norm": 2.4325965447816813, "learning_rate": 6.412354025587509e-07, "loss": 0.3337, "step": 2060 }, { "epoch": 0.8554266777133389, "grad_norm": 2.3977440470960145, "learning_rate": 6.236330375124921e-07, "loss": 0.3358, "step": 2065 }, { "epoch": 0.8574979287489644, "grad_norm": 2.3326291708257902, "learning_rate": 6.062595651705111e-07, "loss": 0.3308, "step": 2070 }, { "epoch": 0.8595691797845899, "grad_norm": 2.335144843157432, "learning_rate": 5.891158941999959e-07, "loss": 0.3301, "step": 2075 }, { "epoch": 0.8616404308202155, "grad_norm": 2.2635540786964685, "learning_rate": 5.722029212490666e-07, "loss": 0.3185, "step": 2080 }, { "epoch": 0.8637116818558409, "grad_norm": 2.4276016208630473, "learning_rate": 5.555215308998779e-07, "loss": 0.327, "step": 2085 }, { "epoch": 0.8657829328914665, "grad_norm": 2.4699900638628742, "learning_rate": 5.390725956223531e-07, "loss": 0.3287, "step": 2090 }, { "epoch": 0.867854183927092, "grad_norm": 2.2549133406841695, "learning_rate": 5.22856975728554e-07, "loss": 0.3157, "step": 2095 }, { "epoch": 0.8699254349627175, "grad_norm": 2.245156912707329, "learning_rate": 5.068755193276798e-07, "loss": 0.326, "step": 2100 }, { "epoch": 0.871996685998343, "grad_norm": 2.356212546488791, "learning_rate": 4.911290622817161e-07, "loss": 0.317, "step": 2105 }, { "epoch": 0.8740679370339686, "grad_norm": 2.2771498825878513, "learning_rate": 4.756184281617121e-07, "loss": 0.3255, "step": 2110 }, { "epoch": 0.876139188069594, "grad_norm": 2.375932171841295, "learning_rate": 4.6034442820471037e-07, "loss": 0.3198, "step": 2115 }, { "epoch": 0.8782104391052196, "grad_norm": 2.5597319859630034, "learning_rate": 4.4530786127131575e-07, "loss": 0.3224, "step": 2120 }, { "epoch": 0.8802816901408451, "grad_norm": 2.481498730243582, "learning_rate": 4.305095138039106e-07, "loss": 0.3319, "step": 2125 }, { "epoch": 0.8823529411764706, "grad_norm": 2.3294677529663983, "learning_rate": 4.159501597855287e-07, "loss": 0.3219, "step": 2130 }, { "epoch": 0.8844241922120961, "grad_norm": 2.281056688111297, "learning_rate": 4.0163056069936757e-07, "loss": 0.3133, "step": 2135 }, { "epoch": 0.8864954432477217, "grad_norm": 2.3568722518604073, "learning_rate": 3.8755146548896784e-07, "loss": 0.3318, "step": 2140 }, { "epoch": 0.8885666942833471, "grad_norm": 2.2630442301266873, "learning_rate": 3.737136105190337e-07, "loss": 0.325, "step": 2145 }, { "epoch": 0.8906379453189727, "grad_norm": 2.369821699455466, "learning_rate": 3.6011771953693044e-07, "loss": 0.3359, "step": 2150 }, { "epoch": 0.8927091963545982, "grad_norm": 2.547108380324384, "learning_rate": 3.4676450363481937e-07, "loss": 0.3162, "step": 2155 }, { "epoch": 0.8947804473902237, "grad_norm": 2.548525193875049, "learning_rate": 3.336546612124758e-07, "loss": 0.3241, "step": 2160 }, { "epoch": 0.8968516984258492, "grad_norm": 2.3650184441515374, "learning_rate": 3.20788877940757e-07, "loss": 0.3201, "step": 2165 }, { "epoch": 0.8989229494614748, "grad_norm": 2.4354180705208734, "learning_rate": 3.081678267257404e-07, "loss": 0.3232, "step": 2170 }, { "epoch": 0.9009942004971002, "grad_norm": 2.437804136816326, "learning_rate": 2.9579216767352815e-07, "loss": 0.3278, "step": 2175 }, { "epoch": 0.9030654515327258, "grad_norm": 2.32676030586252, "learning_rate": 2.836625480557265e-07, "loss": 0.3229, "step": 2180 }, { "epoch": 0.9051367025683513, "grad_norm": 2.420221090023951, "learning_rate": 2.7177960227558863e-07, "loss": 0.3286, "step": 2185 }, { "epoch": 0.9072079536039768, "grad_norm": 2.361797437348497, "learning_rate": 2.601439518348331e-07, "loss": 0.3281, "step": 2190 }, { "epoch": 0.9092792046396023, "grad_norm": 2.3489153533889593, "learning_rate": 2.487562053011422e-07, "loss": 0.3214, "step": 2195 }, { "epoch": 0.9113504556752279, "grad_norm": 2.56111688526178, "learning_rate": 2.376169582763288e-07, "loss": 0.3147, "step": 2200 }, { "epoch": 0.9134217067108533, "grad_norm": 2.57814653530945, "learning_rate": 2.2672679336518789e-07, "loss": 0.3171, "step": 2205 }, { "epoch": 0.9154929577464789, "grad_norm": 2.3494738988744723, "learning_rate": 2.1608628014502364e-07, "loss": 0.3205, "step": 2210 }, { "epoch": 0.9175642087821044, "grad_norm": 2.3096668943633927, "learning_rate": 2.0569597513586004e-07, "loss": 0.3119, "step": 2215 }, { "epoch": 0.9196354598177299, "grad_norm": 2.3426579322038714, "learning_rate": 1.955564217713335e-07, "loss": 0.3155, "step": 2220 }, { "epoch": 0.9217067108533554, "grad_norm": 2.2566958158189987, "learning_rate": 1.8566815037026897e-07, "loss": 0.3207, "step": 2225 }, { "epoch": 0.923777961888981, "grad_norm": 2.359224561499725, "learning_rate": 1.7603167810894662e-07, "loss": 0.3182, "step": 2230 }, { "epoch": 0.9258492129246064, "grad_norm": 2.226966607924682, "learning_rate": 1.6664750899404892e-07, "loss": 0.3129, "step": 2235 }, { "epoch": 0.927920463960232, "grad_norm": 2.3591702073510485, "learning_rate": 1.5751613383630128e-07, "loss": 0.3187, "step": 2240 }, { "epoch": 0.9299917149958575, "grad_norm": 2.3992320019355753, "learning_rate": 1.4863803022480362e-07, "loss": 0.3305, "step": 2245 }, { "epoch": 0.932062966031483, "grad_norm": 2.445578832040922, "learning_rate": 1.4001366250204762e-07, "loss": 0.3154, "step": 2250 }, { "epoch": 0.9341342170671085, "grad_norm": 2.4939555171455114, "learning_rate": 1.3164348173963392e-07, "loss": 0.3231, "step": 2255 }, { "epoch": 0.9362054681027341, "grad_norm": 2.1840069471253005, "learning_rate": 1.235279257146804e-07, "loss": 0.3113, "step": 2260 }, { "epoch": 0.9382767191383595, "grad_norm": 2.3098200854659523, "learning_rate": 1.1566741888692168e-07, "loss": 0.3041, "step": 2265 }, { "epoch": 0.9403479701739851, "grad_norm": 2.539664992169192, "learning_rate": 1.080623723765134e-07, "loss": 0.3093, "step": 2270 }, { "epoch": 0.9424192212096106, "grad_norm": 2.4935115385278124, "learning_rate": 1.0071318394252849e-07, "loss": 0.3181, "step": 2275 }, { "epoch": 0.9444904722452361, "grad_norm": 2.3940149584011237, "learning_rate": 9.362023796215036e-08, "loss": 0.3228, "step": 2280 }, { "epoch": 0.9465617232808616, "grad_norm": 2.487298613154887, "learning_rate": 8.678390541057512e-08, "loss": 0.3326, "step": 2285 }, { "epoch": 0.9486329743164872, "grad_norm": 2.3469443684683293, "learning_rate": 8.020454384160437e-08, "loss": 0.3143, "step": 2290 }, { "epoch": 0.9507042253521126, "grad_norm": 2.2606661120914264, "learning_rate": 7.388249736894615e-08, "loss": 0.3115, "step": 2295 }, { "epoch": 0.9527754763877382, "grad_norm": 2.174724810527737, "learning_rate": 6.78180966482156e-08, "loss": 0.3115, "step": 2300 }, { "epoch": 0.9548467274233637, "grad_norm": 2.430189671520879, "learning_rate": 6.201165885964311e-08, "loss": 0.3162, "step": 2305 }, { "epoch": 0.9569179784589892, "grad_norm": 2.431002564643125, "learning_rate": 5.646348769148491e-08, "loss": 0.3315, "step": 2310 }, { "epoch": 0.9589892294946147, "grad_norm": 2.4561298588521736, "learning_rate": 5.117387332413737e-08, "loss": 0.3214, "step": 2315 }, { "epoch": 0.9610604805302403, "grad_norm": 2.29137591324708, "learning_rate": 4.6143092414961396e-08, "loss": 0.3171, "step": 2320 }, { "epoch": 0.9631317315658657, "grad_norm": 2.2798572393263625, "learning_rate": 4.1371408083815126e-08, "loss": 0.3179, "step": 2325 }, { "epoch": 0.9652029826014913, "grad_norm": 2.5201896418594862, "learning_rate": 3.685906989928656e-08, "loss": 0.3243, "step": 2330 }, { "epoch": 0.9672742336371168, "grad_norm": 2.3567114541229355, "learning_rate": 3.2606313865646276e-08, "loss": 0.3233, "step": 2335 }, { "epoch": 0.9693454846727423, "grad_norm": 2.4845883074404824, "learning_rate": 2.861336241050061e-08, "loss": 0.3171, "step": 2340 }, { "epoch": 0.9714167357083678, "grad_norm": 2.383209324193614, "learning_rate": 2.488042437315985e-08, "loss": 0.3141, "step": 2345 }, { "epoch": 0.9734879867439934, "grad_norm": 2.5226130703660035, "learning_rate": 2.1407694993714755e-08, "loss": 0.3259, "step": 2350 }, { "epoch": 0.9755592377796188, "grad_norm": 2.370723863389497, "learning_rate": 1.8195355902824174e-08, "loss": 0.3124, "step": 2355 }, { "epoch": 0.9776304888152444, "grad_norm": 2.4294371458952604, "learning_rate": 1.5243575112218744e-08, "loss": 0.3216, "step": 2360 }, { "epoch": 0.97970173985087, "grad_norm": 2.3857766427686946, "learning_rate": 1.2552507005909042e-08, "loss": 0.313, "step": 2365 }, { "epoch": 0.9817729908864954, "grad_norm": 2.451896147327658, "learning_rate": 1.0122292332114814e-08, "loss": 0.3222, "step": 2370 }, { "epoch": 0.9838442419221209, "grad_norm": 2.538668369610046, "learning_rate": 7.953058195900864e-09, "loss": 0.3244, "step": 2375 }, { "epoch": 0.9859154929577465, "grad_norm": 2.407872701486615, "learning_rate": 6.044918052531268e-09, "loss": 0.3145, "step": 2380 }, { "epoch": 0.987986743993372, "grad_norm": 2.2756913351567443, "learning_rate": 4.397971701533554e-09, "loss": 0.3225, "step": 2385 }, { "epoch": 0.9900579950289975, "grad_norm": 2.3323337356994656, "learning_rate": 3.0123052814812203e-09, "loss": 0.3193, "step": 2390 }, { "epoch": 0.992129246064623, "grad_norm": 2.360225047221282, "learning_rate": 1.887991265486222e-09, "loss": 0.3173, "step": 2395 }, { "epoch": 0.9942004971002486, "grad_norm": 2.4651589444064737, "learning_rate": 1.025088457409229e-09, "loss": 0.3196, "step": 2400 }, { "epoch": 0.996271748135874, "grad_norm": 2.2576161492102496, "learning_rate": 4.2364198878597216e-10, "loss": 0.316, "step": 2405 }, { "epoch": 0.9983429991714996, "grad_norm": 2.4279743469778388, "learning_rate": 8.368331646302353e-11, "loss": 0.3179, "step": 2410 }, { "epoch": 1.0, "eval_runtime": 3.3314, "eval_samples_per_second": 3.002, "eval_steps_per_second": 0.901, "step": 2414 }, { "epoch": 1.0, "step": 2414, "total_flos": 252721244405760.0, "train_loss": 0.5125990820108807, "train_runtime": 21458.7367, "train_samples_per_second": 1.8, "train_steps_per_second": 0.112 } ], "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 }