{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 2459, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0020339672531272247, "grad_norm": 1.4257310628890991, "learning_rate": 2.601626016260163e-06, "loss": 2.917485809326172, "step": 5 }, { "epoch": 0.004067934506254449, "grad_norm": 1.46852707862854, "learning_rate": 5.853658536585366e-06, "loss": 2.8873306274414063, "step": 10 }, { "epoch": 0.006101901759381674, "grad_norm": 1.0875977277755737, "learning_rate": 9.10569105691057e-06, "loss": 2.862717628479004, "step": 15 }, { "epoch": 0.008135869012508899, "grad_norm": 0.6434921026229858, "learning_rate": 1.2357723577235774e-05, "loss": 2.760717010498047, "step": 20 }, { "epoch": 0.010169836265636124, "grad_norm": 0.3681577742099762, "learning_rate": 1.5609756097560978e-05, "loss": 2.6175188064575194, "step": 25 }, { "epoch": 0.012203803518763347, "grad_norm": 0.2874605357646942, "learning_rate": 1.886178861788618e-05, "loss": 2.6090604782104494, "step": 30 }, { "epoch": 0.014237770771890573, "grad_norm": 0.2970505654811859, "learning_rate": 2.2113821138211385e-05, "loss": 2.48167724609375, "step": 35 }, { "epoch": 0.016271738025017798, "grad_norm": 0.29987362027168274, "learning_rate": 2.536585365853659e-05, "loss": 2.3590694427490235, "step": 40 }, { "epoch": 0.018305705278145023, "grad_norm": 0.29447707533836365, "learning_rate": 2.8617886178861792e-05, "loss": 2.3665092468261717, "step": 45 }, { "epoch": 0.020339672531272248, "grad_norm": 0.23196953535079956, "learning_rate": 3.186991869918699e-05, "loss": 2.2919033050537108, "step": 50 }, { "epoch": 0.02237363978439947, "grad_norm": 0.2012770175933838, "learning_rate": 3.51219512195122e-05, "loss": 2.228291893005371, "step": 55 }, { "epoch": 0.024407607037526695, "grad_norm": 0.22747980058193207, "learning_rate": 3.83739837398374e-05, "loss": 2.187364959716797, "step": 60 }, { "epoch": 0.02644157429065392, "grad_norm": 0.1936585158109665, "learning_rate": 4.162601626016261e-05, "loss": 2.2040355682373045, "step": 65 }, { "epoch": 0.028475541543781145, "grad_norm": 0.21204930543899536, "learning_rate": 4.4878048780487814e-05, "loss": 2.0758914947509766, "step": 70 }, { "epoch": 0.03050950879690837, "grad_norm": 0.25752541422843933, "learning_rate": 4.8130081300813014e-05, "loss": 2.074118804931641, "step": 75 }, { "epoch": 0.032543476050035595, "grad_norm": 0.24572649598121643, "learning_rate": 5.138211382113822e-05, "loss": 2.0441198348999023, "step": 80 }, { "epoch": 0.03457744330316282, "grad_norm": 0.21846380829811096, "learning_rate": 5.463414634146342e-05, "loss": 1.934136199951172, "step": 85 }, { "epoch": 0.036611410556290046, "grad_norm": 0.2201363742351532, "learning_rate": 5.788617886178863e-05, "loss": 1.922892189025879, "step": 90 }, { "epoch": 0.03864537780941727, "grad_norm": 0.23432138562202454, "learning_rate": 6.113821138211384e-05, "loss": 1.8857765197753906, "step": 95 }, { "epoch": 0.040679345062544496, "grad_norm": 0.23306460678577423, "learning_rate": 6.439024390243903e-05, "loss": 1.9038379669189454, "step": 100 }, { "epoch": 0.042713312315671714, "grad_norm": 0.22993463277816772, "learning_rate": 6.764227642276424e-05, "loss": 1.8566299438476563, "step": 105 }, { "epoch": 0.04474727956879894, "grad_norm": 0.22449328005313873, "learning_rate": 7.089430894308944e-05, "loss": 1.8021886825561524, "step": 110 }, { "epoch": 0.046781246821926165, "grad_norm": 0.2615789771080017, "learning_rate": 7.414634146341465e-05, "loss": 1.7697391510009766, "step": 115 }, { "epoch": 0.04881521407505339, "grad_norm": 0.21783658862113953, "learning_rate": 7.739837398373984e-05, "loss": 1.8027568817138673, "step": 120 }, { "epoch": 0.050849181328180615, "grad_norm": 0.241287499666214, "learning_rate": 7.999996382704413e-05, "loss": 1.785761260986328, "step": 125 }, { "epoch": 0.05288314858130784, "grad_norm": 0.21905580163002014, "learning_rate": 7.999869778045771e-05, "loss": 1.8103389739990234, "step": 130 }, { "epoch": 0.054917115834435065, "grad_norm": 0.24265141785144806, "learning_rate": 7.99956231515007e-05, "loss": 1.7369571685791017, "step": 135 }, { "epoch": 0.05695108308756229, "grad_norm": 0.2645692229270935, "learning_rate": 7.999074007919565e-05, "loss": 1.7425952911376954, "step": 140 }, { "epoch": 0.058985050340689515, "grad_norm": 0.22949549555778503, "learning_rate": 7.998404878433566e-05, "loss": 1.7011558532714843, "step": 145 }, { "epoch": 0.06101901759381674, "grad_norm": 0.28005731105804443, "learning_rate": 7.997554956947454e-05, "loss": 1.6993677139282226, "step": 150 }, { "epoch": 0.06305298484694397, "grad_norm": 0.25947996973991394, "learning_rate": 7.996524281891309e-05, "loss": 1.752249526977539, "step": 155 }, { "epoch": 0.06508695210007119, "grad_norm": 0.27306196093559265, "learning_rate": 7.995312899868164e-05, "loss": 1.6812099456787108, "step": 160 }, { "epoch": 0.06712091935319842, "grad_norm": 0.2487000674009323, "learning_rate": 7.993920865651903e-05, "loss": 1.6785842895507812, "step": 165 }, { "epoch": 0.06915488660632564, "grad_norm": 0.2866049110889435, "learning_rate": 7.992348242184795e-05, "loss": 1.6411857604980469, "step": 170 }, { "epoch": 0.07118885385945287, "grad_norm": 0.2541375458240509, "learning_rate": 7.99059510057463e-05, "loss": 1.6829517364501954, "step": 175 }, { "epoch": 0.07322282111258009, "grad_norm": 0.2908504009246826, "learning_rate": 7.988661520091513e-05, "loss": 1.660489273071289, "step": 180 }, { "epoch": 0.07525678836570732, "grad_norm": 0.30211013555526733, "learning_rate": 7.986547588164276e-05, "loss": 1.6549396514892578, "step": 185 }, { "epoch": 0.07729075561883454, "grad_norm": 0.28774863481521606, "learning_rate": 7.984253400376539e-05, "loss": 1.6788974761962892, "step": 190 }, { "epoch": 0.07932472287196177, "grad_norm": 0.2667076587677002, "learning_rate": 7.981779060462364e-05, "loss": 1.6379096984863282, "step": 195 }, { "epoch": 0.08135869012508899, "grad_norm": 0.31970497965812683, "learning_rate": 7.979124680301582e-05, "loss": 1.605440902709961, "step": 200 }, { "epoch": 0.08339265737821622, "grad_norm": 0.2745591998100281, "learning_rate": 7.976290379914728e-05, "loss": 1.6268569946289062, "step": 205 }, { "epoch": 0.08542662463134343, "grad_norm": 0.2973490357398987, "learning_rate": 7.973276287457622e-05, "loss": 1.5992630004882813, "step": 210 }, { "epoch": 0.08746059188447065, "grad_norm": 0.28332599997520447, "learning_rate": 7.97008253921556e-05, "loss": 1.6371658325195313, "step": 215 }, { "epoch": 0.08949455913759788, "grad_norm": 0.2501751482486725, "learning_rate": 7.96670927959716e-05, "loss": 1.6693359375, "step": 220 }, { "epoch": 0.0915285263907251, "grad_norm": 0.3090277910232544, "learning_rate": 7.963156661127839e-05, "loss": 1.575566005706787, "step": 225 }, { "epoch": 0.09356249364385233, "grad_norm": 0.2466857135295868, "learning_rate": 7.959424844442902e-05, "loss": 1.6048229217529297, "step": 230 }, { "epoch": 0.09559646089697955, "grad_norm": 0.2916584610939026, "learning_rate": 7.95551399828029e-05, "loss": 1.5845436096191405, "step": 235 }, { "epoch": 0.09763042815010678, "grad_norm": 0.23102006316184998, "learning_rate": 7.951424299472945e-05, "loss": 1.5969648361206055, "step": 240 }, { "epoch": 0.099664395403234, "grad_norm": 0.261629581451416, "learning_rate": 7.947155932940817e-05, "loss": 1.5344152450561523, "step": 245 }, { "epoch": 0.10169836265636123, "grad_norm": 0.2802213430404663, "learning_rate": 7.9427090916825e-05, "loss": 1.6227588653564453, "step": 250 }, { "epoch": 0.10373232990948845, "grad_norm": 0.276521772146225, "learning_rate": 7.938083976766508e-05, "loss": 1.5408702850341798, "step": 255 }, { "epoch": 0.10576629716261568, "grad_norm": 0.255205363035202, "learning_rate": 7.933280797322181e-05, "loss": 1.576058292388916, "step": 260 }, { "epoch": 0.1078002644157429, "grad_norm": 0.25489553809165955, "learning_rate": 7.928299770530232e-05, "loss": 1.5977807998657227, "step": 265 }, { "epoch": 0.10983423166887013, "grad_norm": 0.253703773021698, "learning_rate": 7.923141121612922e-05, "loss": 1.6158340454101563, "step": 270 }, { "epoch": 0.11186819892199736, "grad_norm": 0.2618684470653534, "learning_rate": 7.917805083823886e-05, "loss": 1.56042537689209, "step": 275 }, { "epoch": 0.11390216617512458, "grad_norm": 0.2836095988750458, "learning_rate": 7.912291898437572e-05, "loss": 1.6143115997314452, "step": 280 }, { "epoch": 0.1159361334282518, "grad_norm": 0.3135998547077179, "learning_rate": 7.906601814738345e-05, "loss": 1.5314339637756347, "step": 285 }, { "epoch": 0.11797010068137903, "grad_norm": 0.28871971368789673, "learning_rate": 7.900735090009208e-05, "loss": 1.5611291885375977, "step": 290 }, { "epoch": 0.12000406793450626, "grad_norm": 0.264883816242218, "learning_rate": 7.894691989520168e-05, "loss": 1.58094482421875, "step": 295 }, { "epoch": 0.12203803518763348, "grad_norm": 0.25624939799308777, "learning_rate": 7.888472786516246e-05, "loss": 1.5583287239074708, "step": 300 }, { "epoch": 0.1240720024407607, "grad_norm": 0.27313604950904846, "learning_rate": 7.882077762205122e-05, "loss": 1.5159448623657226, "step": 305 }, { "epoch": 0.12610596969388793, "grad_norm": 0.25955769419670105, "learning_rate": 7.875507205744412e-05, "loss": 1.5097437858581544, "step": 310 }, { "epoch": 0.12813993694701514, "grad_norm": 0.3122912347316742, "learning_rate": 7.868761414228604e-05, "loss": 1.5064099311828614, "step": 315 }, { "epoch": 0.13017390420014238, "grad_norm": 0.2565915584564209, "learning_rate": 7.861840692675622e-05, "loss": 1.581326389312744, "step": 320 }, { "epoch": 0.1322078714532696, "grad_norm": 0.32710567116737366, "learning_rate": 7.854745354013022e-05, "loss": 1.5986787796020507, "step": 325 }, { "epoch": 0.13424183870639683, "grad_norm": 0.2665398120880127, "learning_rate": 7.847475719063862e-05, "loss": 1.52600154876709, "step": 330 }, { "epoch": 0.13627580595952404, "grad_norm": 0.2865259349346161, "learning_rate": 7.840032116532184e-05, "loss": 1.5572259902954102, "step": 335 }, { "epoch": 0.13830977321265128, "grad_norm": 0.34650203585624695, "learning_rate": 7.832414882988153e-05, "loss": 1.4841368675231934, "step": 340 }, { "epoch": 0.1403437404657785, "grad_norm": 0.2880828082561493, "learning_rate": 7.824624362852839e-05, "loss": 1.4806913375854491, "step": 345 }, { "epoch": 0.14237770771890573, "grad_norm": 0.27728167176246643, "learning_rate": 7.816660908382642e-05, "loss": 1.5396782875061035, "step": 350 }, { "epoch": 0.14441167497203294, "grad_norm": 0.3270809054374695, "learning_rate": 7.808524879653372e-05, "loss": 1.506382656097412, "step": 355 }, { "epoch": 0.14644564222516018, "grad_norm": 0.33753281831741333, "learning_rate": 7.800216644543956e-05, "loss": 1.5199095726013183, "step": 360 }, { "epoch": 0.1484796094782874, "grad_norm": 0.293161541223526, "learning_rate": 7.791736578719814e-05, "loss": 1.5270258903503418, "step": 365 }, { "epoch": 0.15051357673141463, "grad_norm": 0.34172746539115906, "learning_rate": 7.783085065615869e-05, "loss": 1.5228445053100585, "step": 370 }, { "epoch": 0.15254754398454184, "grad_norm": 0.29226797819137573, "learning_rate": 7.774262496419205e-05, "loss": 1.5537534713745118, "step": 375 }, { "epoch": 0.15458151123766908, "grad_norm": 0.3251171112060547, "learning_rate": 7.765269270051389e-05, "loss": 1.4998278617858887, "step": 380 }, { "epoch": 0.1566154784907963, "grad_norm": 0.28418686985969543, "learning_rate": 7.756105793150426e-05, "loss": 1.5408721923828126, "step": 385 }, { "epoch": 0.15864944574392353, "grad_norm": 0.30530861020088196, "learning_rate": 7.746772480052373e-05, "loss": 1.5005302429199219, "step": 390 }, { "epoch": 0.16068341299705075, "grad_norm": 0.2712857723236084, "learning_rate": 7.737269752772608e-05, "loss": 1.4546459197998047, "step": 395 }, { "epoch": 0.16271738025017798, "grad_norm": 0.28193557262420654, "learning_rate": 7.727598040986746e-05, "loss": 1.513768482208252, "step": 400 }, { "epoch": 0.1647513475033052, "grad_norm": 0.2774534821510315, "learning_rate": 7.717757782011207e-05, "loss": 1.47432861328125, "step": 405 }, { "epoch": 0.16678531475643243, "grad_norm": 0.30369889736175537, "learning_rate": 7.707749420783454e-05, "loss": 1.482215690612793, "step": 410 }, { "epoch": 0.16881928200955965, "grad_norm": 0.3143354058265686, "learning_rate": 7.697573409841854e-05, "loss": 1.4836478233337402, "step": 415 }, { "epoch": 0.17085324926268686, "grad_norm": 0.2981318533420563, "learning_rate": 7.687230209305242e-05, "loss": 1.5305231094360352, "step": 420 }, { "epoch": 0.1728872165158141, "grad_norm": 0.2954687476158142, "learning_rate": 7.676720286852096e-05, "loss": 1.527289581298828, "step": 425 }, { "epoch": 0.1749211837689413, "grad_norm": 0.31553202867507935, "learning_rate": 7.666044117699397e-05, "loss": 1.4683035850524901, "step": 430 }, { "epoch": 0.17695515102206855, "grad_norm": 0.2957124710083008, "learning_rate": 7.655202184581145e-05, "loss": 1.4824369430541993, "step": 435 }, { "epoch": 0.17898911827519576, "grad_norm": 0.3332107365131378, "learning_rate": 7.644194977726527e-05, "loss": 1.480666446685791, "step": 440 }, { "epoch": 0.181023085528323, "grad_norm": 0.31734415888786316, "learning_rate": 7.633022994837751e-05, "loss": 1.4788445472717284, "step": 445 }, { "epoch": 0.1830570527814502, "grad_norm": 0.29531416296958923, "learning_rate": 7.621686741067548e-05, "loss": 1.5645590782165528, "step": 450 }, { "epoch": 0.18509102003457745, "grad_norm": 0.31266340613365173, "learning_rate": 7.610186728996319e-05, "loss": 1.5170171737670899, "step": 455 }, { "epoch": 0.18712498728770466, "grad_norm": 0.2995188236236572, "learning_rate": 7.598523478608974e-05, "loss": 1.506637954711914, "step": 460 }, { "epoch": 0.1891589545408319, "grad_norm": 0.28108903765678406, "learning_rate": 7.586697517271406e-05, "loss": 1.4750299453735352, "step": 465 }, { "epoch": 0.1911929217939591, "grad_norm": 0.281231552362442, "learning_rate": 7.574709379706651e-05, "loss": 1.5297168731689452, "step": 470 }, { "epoch": 0.19322688904708635, "grad_norm": 0.3016623556613922, "learning_rate": 7.562559607970714e-05, "loss": 1.4363483428955077, "step": 475 }, { "epoch": 0.19526085630021356, "grad_norm": 0.2738240659236908, "learning_rate": 7.550248751428054e-05, "loss": 1.4974472045898437, "step": 480 }, { "epoch": 0.1972948235533408, "grad_norm": 0.2762449085712433, "learning_rate": 7.537777366726745e-05, "loss": 1.4656939506530762, "step": 485 }, { "epoch": 0.199328790806468, "grad_norm": 0.3260698914527893, "learning_rate": 7.52514601777331e-05, "loss": 1.4498090744018555, "step": 490 }, { "epoch": 0.20136275805959525, "grad_norm": 0.30235323309898376, "learning_rate": 7.512355275707218e-05, "loss": 1.454343318939209, "step": 495 }, { "epoch": 0.20339672531272246, "grad_norm": 0.2750537097454071, "learning_rate": 7.49940571887506e-05, "loss": 1.4856033325195312, "step": 500 }, { "epoch": 0.2054306925658497, "grad_norm": 0.26810595393180847, "learning_rate": 7.486297932804404e-05, "loss": 1.5180526733398438, "step": 505 }, { "epoch": 0.2074646598189769, "grad_norm": 0.34094545245170593, "learning_rate": 7.473032510177316e-05, "loss": 1.5161296844482421, "step": 510 }, { "epoch": 0.20949862707210415, "grad_norm": 0.29126617312431335, "learning_rate": 7.459610050803561e-05, "loss": 1.553421974182129, "step": 515 }, { "epoch": 0.21153259432523136, "grad_norm": 0.33887821435928345, "learning_rate": 7.446031161593481e-05, "loss": 1.4819551467895509, "step": 520 }, { "epoch": 0.2135665615783586, "grad_norm": 0.3397113084793091, "learning_rate": 7.432296456530557e-05, "loss": 1.4011404037475585, "step": 525 }, { "epoch": 0.2156005288314858, "grad_norm": 0.33551251888275146, "learning_rate": 7.418406556643642e-05, "loss": 1.4781397819519042, "step": 530 }, { "epoch": 0.21763449608461305, "grad_norm": 0.34282976388931274, "learning_rate": 7.404362089978883e-05, "loss": 1.5161099433898926, "step": 535 }, { "epoch": 0.21966846333774026, "grad_norm": 0.3225310742855072, "learning_rate": 7.390163691571325e-05, "loss": 1.463518238067627, "step": 540 }, { "epoch": 0.2217024305908675, "grad_norm": 0.25978654623031616, "learning_rate": 7.375812003416195e-05, "loss": 1.4573867797851563, "step": 545 }, { "epoch": 0.2237363978439947, "grad_norm": 0.3284743130207062, "learning_rate": 7.361307674439873e-05, "loss": 1.4482636451721191, "step": 550 }, { "epoch": 0.22577036509712195, "grad_norm": 0.27696937322616577, "learning_rate": 7.346651360470555e-05, "loss": 1.4380542755126953, "step": 555 }, { "epoch": 0.22780433235024916, "grad_norm": 0.27538761496543884, "learning_rate": 7.331843724208589e-05, "loss": 1.4844141006469727, "step": 560 }, { "epoch": 0.22983829960337637, "grad_norm": 0.3249717354774475, "learning_rate": 7.316885435196518e-05, "loss": 1.4259913444519043, "step": 565 }, { "epoch": 0.2318722668565036, "grad_norm": 0.3172062635421753, "learning_rate": 7.30177716978881e-05, "loss": 1.4604599952697754, "step": 570 }, { "epoch": 0.23390623410963082, "grad_norm": 0.25277936458587646, "learning_rate": 7.286519611121262e-05, "loss": 1.5142675399780274, "step": 575 }, { "epoch": 0.23594020136275806, "grad_norm": 0.30122074484825134, "learning_rate": 7.271113449080131e-05, "loss": 1.4454142570495605, "step": 580 }, { "epoch": 0.23797416861588527, "grad_norm": 0.29546478390693665, "learning_rate": 7.255559380270915e-05, "loss": 1.4152225494384765, "step": 585 }, { "epoch": 0.2400081358690125, "grad_norm": 0.25515016913414, "learning_rate": 7.239858107986883e-05, "loss": 1.4481670379638671, "step": 590 }, { "epoch": 0.24204210312213972, "grad_norm": 0.36043909192085266, "learning_rate": 7.224010342177246e-05, "loss": 1.3873469352722168, "step": 595 }, { "epoch": 0.24407607037526696, "grad_norm": 0.28646320104599, "learning_rate": 7.208016799415083e-05, "loss": 1.4386262893676758, "step": 600 }, { "epoch": 0.24611003762839417, "grad_norm": 0.30333590507507324, "learning_rate": 7.191878202864926e-05, "loss": 1.4411699295043945, "step": 605 }, { "epoch": 0.2481440048815214, "grad_norm": 0.3224600851535797, "learning_rate": 7.175595282250058e-05, "loss": 1.4494745254516601, "step": 610 }, { "epoch": 0.25017797213464865, "grad_norm": 0.2596285939216614, "learning_rate": 7.159168773819529e-05, "loss": 1.42305269241333, "step": 615 }, { "epoch": 0.25221193938777586, "grad_norm": 0.2846454977989197, "learning_rate": 7.142599420314857e-05, "loss": 1.4384309768676757, "step": 620 }, { "epoch": 0.2542459066409031, "grad_norm": 0.3560178875923157, "learning_rate": 7.12588797093645e-05, "loss": 1.4937101364135743, "step": 625 }, { "epoch": 0.2562798738940303, "grad_norm": 0.3016885221004486, "learning_rate": 7.109035181309728e-05, "loss": 1.4564781188964844, "step": 630 }, { "epoch": 0.25831384114715755, "grad_norm": 0.2757645547389984, "learning_rate": 7.092041813450948e-05, "loss": 1.4405845642089843, "step": 635 }, { "epoch": 0.26034780840028476, "grad_norm": 0.2569405734539032, "learning_rate": 7.074908635732768e-05, "loss": 1.4014852523803711, "step": 640 }, { "epoch": 0.262381775653412, "grad_norm": 0.3015080690383911, "learning_rate": 7.057636422849487e-05, "loss": 1.433372688293457, "step": 645 }, { "epoch": 0.2644157429065392, "grad_norm": 0.29660722613334656, "learning_rate": 7.040225955782022e-05, "loss": 1.4342848777770996, "step": 650 }, { "epoch": 0.26644971015966645, "grad_norm": 0.33201417326927185, "learning_rate": 7.022678021762598e-05, "loss": 1.4227266311645508, "step": 655 }, { "epoch": 0.26848367741279366, "grad_norm": 0.2671425938606262, "learning_rate": 7.004993414239147e-05, "loss": 1.4589752197265624, "step": 660 }, { "epoch": 0.2705176446659209, "grad_norm": 0.26866036653518677, "learning_rate": 6.98717293283944e-05, "loss": 1.4873647689819336, "step": 665 }, { "epoch": 0.2725516119190481, "grad_norm": 0.28285905718803406, "learning_rate": 6.969217383334917e-05, "loss": 1.459207534790039, "step": 670 }, { "epoch": 0.27458557917217535, "grad_norm": 0.2887929081916809, "learning_rate": 6.951127577604269e-05, "loss": 1.4669806480407714, "step": 675 }, { "epoch": 0.27661954642530256, "grad_norm": 0.2804192006587982, "learning_rate": 6.932904333596718e-05, "loss": 1.4024450302124023, "step": 680 }, { "epoch": 0.2786535136784298, "grad_norm": 0.2690044343471527, "learning_rate": 6.914548475295033e-05, "loss": 1.4585531234741211, "step": 685 }, { "epoch": 0.280687480931557, "grad_norm": 0.3106100261211395, "learning_rate": 6.89606083267828e-05, "loss": 1.4643203735351562, "step": 690 }, { "epoch": 0.28272144818468425, "grad_norm": 0.2849403917789459, "learning_rate": 6.877442241684288e-05, "loss": 1.5139555931091309, "step": 695 }, { "epoch": 0.28475541543781147, "grad_norm": 0.28812873363494873, "learning_rate": 6.858693544171845e-05, "loss": 1.4307933807373048, "step": 700 }, { "epoch": 0.2867893826909387, "grad_norm": 0.25747042894363403, "learning_rate": 6.839815587882647e-05, "loss": 1.4572304725646972, "step": 705 }, { "epoch": 0.2888233499440659, "grad_norm": 0.25883084535598755, "learning_rate": 6.820809226402957e-05, "loss": 1.4297290802001954, "step": 710 }, { "epoch": 0.2908573171971931, "grad_norm": 0.316441148519516, "learning_rate": 6.801675319125007e-05, "loss": 1.4357834815979005, "step": 715 }, { "epoch": 0.29289128445032037, "grad_norm": 0.33814454078674316, "learning_rate": 6.782414731208154e-05, "loss": 1.4652581214904785, "step": 720 }, { "epoch": 0.2949252517034476, "grad_norm": 0.285017192363739, "learning_rate": 6.763028333539737e-05, "loss": 1.395161247253418, "step": 725 }, { "epoch": 0.2969592189565748, "grad_norm": 0.33924534916877747, "learning_rate": 6.743517002695724e-05, "loss": 1.3853025436401367, "step": 730 }, { "epoch": 0.298993186209702, "grad_norm": 0.2676454782485962, "learning_rate": 6.723881620901059e-05, "loss": 1.4198028564453125, "step": 735 }, { "epoch": 0.30102715346282927, "grad_norm": 0.3764224946498871, "learning_rate": 6.70412307598978e-05, "loss": 1.4417477607727052, "step": 740 }, { "epoch": 0.3030611207159565, "grad_norm": 0.27534952759742737, "learning_rate": 6.684242261364874e-05, "loss": 1.4394853591918946, "step": 745 }, { "epoch": 0.3050950879690837, "grad_norm": 0.30524176359176636, "learning_rate": 6.664240075957875e-05, "loss": 1.4469685554504395, "step": 750 }, { "epoch": 0.3071290552222109, "grad_norm": 0.27213791012763977, "learning_rate": 6.644117424188224e-05, "loss": 1.4358532905578614, "step": 755 }, { "epoch": 0.30916302247533817, "grad_norm": 0.280111700296402, "learning_rate": 6.623875215922375e-05, "loss": 1.406216526031494, "step": 760 }, { "epoch": 0.3111969897284654, "grad_norm": 0.25174808502197266, "learning_rate": 6.603514366432649e-05, "loss": 1.4507400512695312, "step": 765 }, { "epoch": 0.3132309569815926, "grad_norm": 0.3225403130054474, "learning_rate": 6.583035796355853e-05, "loss": 1.4420936584472657, "step": 770 }, { "epoch": 0.3152649242347198, "grad_norm": 0.2946665585041046, "learning_rate": 6.562440431651655e-05, "loss": 1.3917129516601563, "step": 775 }, { "epoch": 0.31729889148784707, "grad_norm": 0.28828397393226624, "learning_rate": 6.54172920356071e-05, "loss": 1.3753639221191407, "step": 780 }, { "epoch": 0.3193328587409743, "grad_norm": 0.3176410496234894, "learning_rate": 6.520903048562554e-05, "loss": 1.4257610321044922, "step": 785 }, { "epoch": 0.3213668259941015, "grad_norm": 0.2822883725166321, "learning_rate": 6.499962908333262e-05, "loss": 1.3781893730163575, "step": 790 }, { "epoch": 0.3234007932472287, "grad_norm": 0.2906973361968994, "learning_rate": 6.478909729702873e-05, "loss": 1.4165437698364258, "step": 795 }, { "epoch": 0.32543476050035597, "grad_norm": 0.2236020267009735, "learning_rate": 6.457744464612568e-05, "loss": 1.397031307220459, "step": 800 }, { "epoch": 0.3274687277534832, "grad_norm": 0.2636953592300415, "learning_rate": 6.436468070071636e-05, "loss": 1.3988968849182128, "step": 805 }, { "epoch": 0.3295026950066104, "grad_norm": 0.2582594156265259, "learning_rate": 6.4150815081142e-05, "loss": 1.3823265075683593, "step": 810 }, { "epoch": 0.3315366622597376, "grad_norm": 0.2765803933143616, "learning_rate": 6.393585745755714e-05, "loss": 1.350075912475586, "step": 815 }, { "epoch": 0.33357062951286487, "grad_norm": 0.3021979033946991, "learning_rate": 6.371981754949241e-05, "loss": 1.4218207359313966, "step": 820 }, { "epoch": 0.3356045967659921, "grad_norm": 0.2982991933822632, "learning_rate": 6.350270512541506e-05, "loss": 1.4619415283203125, "step": 825 }, { "epoch": 0.3376385640191193, "grad_norm": 0.3263569176197052, "learning_rate": 6.328453000228726e-05, "loss": 1.374096393585205, "step": 830 }, { "epoch": 0.3396725312722465, "grad_norm": 0.24030761420726776, "learning_rate": 6.306530204512218e-05, "loss": 1.3717924118041993, "step": 835 }, { "epoch": 0.3417064985253737, "grad_norm": 0.30486997961997986, "learning_rate": 6.2845031166538e-05, "loss": 1.4392295837402345, "step": 840 }, { "epoch": 0.343740465778501, "grad_norm": 0.28287866711616516, "learning_rate": 6.26237273263097e-05, "loss": 1.4080901145935059, "step": 845 }, { "epoch": 0.3457744330316282, "grad_norm": 0.2588256299495697, "learning_rate": 6.240140053091859e-05, "loss": 1.3927085876464844, "step": 850 }, { "epoch": 0.3478084002847554, "grad_norm": 0.32203900814056396, "learning_rate": 6.217806083310002e-05, "loss": 1.4236017227172852, "step": 855 }, { "epoch": 0.3498423675378826, "grad_norm": 0.36687156558036804, "learning_rate": 6.195371833138878e-05, "loss": 1.3992155075073243, "step": 860 }, { "epoch": 0.3518763347910099, "grad_norm": 0.3180311322212219, "learning_rate": 6.172838316966242e-05, "loss": 1.3512189865112305, "step": 865 }, { "epoch": 0.3539103020441371, "grad_norm": 0.2805245518684387, "learning_rate": 6.15020655366827e-05, "loss": 1.3968843460083007, "step": 870 }, { "epoch": 0.3559442692972643, "grad_norm": 0.26564761996269226, "learning_rate": 6.127477566563475e-05, "loss": 1.4331690788269043, "step": 875 }, { "epoch": 0.3579782365503915, "grad_norm": 0.2855769395828247, "learning_rate": 6.104652383366453e-05, "loss": 1.3854801177978515, "step": 880 }, { "epoch": 0.3600122038035188, "grad_norm": 0.3231807351112366, "learning_rate": 6.0817320361414e-05, "loss": 1.3894742965698241, "step": 885 }, { "epoch": 0.362046171056646, "grad_norm": 0.27855047583580017, "learning_rate": 6.058717561255453e-05, "loss": 1.4054418563842774, "step": 890 }, { "epoch": 0.3640801383097732, "grad_norm": 0.34973031282424927, "learning_rate": 6.035609999331831e-05, "loss": 1.449208641052246, "step": 895 }, { "epoch": 0.3661141055629004, "grad_norm": 0.3263750672340393, "learning_rate": 6.012410395202774e-05, "loss": 1.357381820678711, "step": 900 }, { "epoch": 0.3681480728160277, "grad_norm": 0.3052593767642975, "learning_rate": 5.9891197978623106e-05, "loss": 1.420388412475586, "step": 905 }, { "epoch": 0.3701820400691549, "grad_norm": 0.2666208744049072, "learning_rate": 5.9657392604188175e-05, "loss": 1.378359031677246, "step": 910 }, { "epoch": 0.3722160073222821, "grad_norm": 0.2978176772594452, "learning_rate": 5.9422698400474085e-05, "loss": 1.3628035545349122, "step": 915 }, { "epoch": 0.3742499745754093, "grad_norm": 0.2530423104763031, "learning_rate": 5.918712597942129e-05, "loss": 1.3725166320800781, "step": 920 }, { "epoch": 0.3762839418285366, "grad_norm": 0.2587495744228363, "learning_rate": 5.895068599267976e-05, "loss": 1.4014698028564454, "step": 925 }, { "epoch": 0.3783179090816638, "grad_norm": 0.27560991048812866, "learning_rate": 5.871338913112732e-05, "loss": 1.392547035217285, "step": 930 }, { "epoch": 0.380351876334791, "grad_norm": 0.3113063871860504, "learning_rate": 5.847524612438631e-05, "loss": 1.4030019760131835, "step": 935 }, { "epoch": 0.3823858435879182, "grad_norm": 0.26230597496032715, "learning_rate": 5.823626774033836e-05, "loss": 1.4046947479248046, "step": 940 }, { "epoch": 0.3844198108410455, "grad_norm": 0.3309447765350342, "learning_rate": 5.799646478463756e-05, "loss": 1.4129743576049805, "step": 945 }, { "epoch": 0.3864537780941727, "grad_norm": 0.29528191685676575, "learning_rate": 5.775584810022187e-05, "loss": 1.3438188552856445, "step": 950 }, { "epoch": 0.3884877453472999, "grad_norm": 0.2803042232990265, "learning_rate": 5.751442856682281e-05, "loss": 1.411970329284668, "step": 955 }, { "epoch": 0.3905217126004271, "grad_norm": 0.2928633689880371, "learning_rate": 5.7272217100473574e-05, "loss": 1.338977336883545, "step": 960 }, { "epoch": 0.3925556798535544, "grad_norm": 0.2996228039264679, "learning_rate": 5.702922465301538e-05, "loss": 1.406255340576172, "step": 965 }, { "epoch": 0.3945896471066816, "grad_norm": 0.3038686215877533, "learning_rate": 5.6785462211602366e-05, "loss": 1.387649917602539, "step": 970 }, { "epoch": 0.3966236143598088, "grad_norm": 0.27196818590164185, "learning_rate": 5.6540940798204685e-05, "loss": 1.4394011497497559, "step": 975 }, { "epoch": 0.398657581612936, "grad_norm": 0.3314681947231293, "learning_rate": 5.629567146911023e-05, "loss": 1.362769603729248, "step": 980 }, { "epoch": 0.40069154886606323, "grad_norm": 0.2561464309692383, "learning_rate": 5.604966531442467e-05, "loss": 1.376676368713379, "step": 985 }, { "epoch": 0.4027255161191905, "grad_norm": 0.2741257846355438, "learning_rate": 5.580293345756998e-05, "loss": 1.3746231079101563, "step": 990 }, { "epoch": 0.4047594833723177, "grad_norm": 0.277779757976532, "learning_rate": 5.555548705478153e-05, "loss": 1.4261016845703125, "step": 995 }, { "epoch": 0.4067934506254449, "grad_norm": 0.3017802834510803, "learning_rate": 5.5307337294603595e-05, "loss": 1.4316215515136719, "step": 1000 }, { "epoch": 0.40882741787857213, "grad_norm": 0.28466475009918213, "learning_rate": 5.5058495397383516e-05, "loss": 1.356630802154541, "step": 1005 }, { "epoch": 0.4108613851316994, "grad_norm": 0.2867753803730011, "learning_rate": 5.48089726147643e-05, "loss": 1.3727928161621095, "step": 1010 }, { "epoch": 0.4128953523848266, "grad_norm": 0.28801488876342773, "learning_rate": 5.4558780229175886e-05, "loss": 1.3452640533447267, "step": 1015 }, { "epoch": 0.4149293196379538, "grad_norm": 0.2714615762233734, "learning_rate": 5.430792955332503e-05, "loss": 1.3697635650634765, "step": 1020 }, { "epoch": 0.41696328689108103, "grad_norm": 0.37546607851982117, "learning_rate": 5.4056431929683716e-05, "loss": 1.329300308227539, "step": 1025 }, { "epoch": 0.4189972541442083, "grad_norm": 0.3115745186805725, "learning_rate": 5.3804298729976386e-05, "loss": 1.3818323135375976, "step": 1030 }, { "epoch": 0.4210312213973355, "grad_norm": 0.2840063273906708, "learning_rate": 5.355154135466567e-05, "loss": 1.329262924194336, "step": 1035 }, { "epoch": 0.4230651886504627, "grad_norm": 0.2970770299434662, "learning_rate": 5.329817123243698e-05, "loss": 1.3751975059509278, "step": 1040 }, { "epoch": 0.42509915590358993, "grad_norm": 0.29037079215049744, "learning_rate": 5.304419981968169e-05, "loss": 1.3338433265686036, "step": 1045 }, { "epoch": 0.4271331231567172, "grad_norm": 0.3004801571369171, "learning_rate": 5.278963859997912e-05, "loss": 1.3847236633300781, "step": 1050 }, { "epoch": 0.4291670904098444, "grad_norm": 0.28835079073905945, "learning_rate": 5.253449908357738e-05, "loss": 1.382228946685791, "step": 1055 }, { "epoch": 0.4312010576629716, "grad_norm": 0.3049657940864563, "learning_rate": 5.227879280687281e-05, "loss": 1.3938045501708984, "step": 1060 }, { "epoch": 0.43323502491609883, "grad_norm": 0.3073517382144928, "learning_rate": 5.202253133188842e-05, "loss": 1.353543472290039, "step": 1065 }, { "epoch": 0.4352689921692261, "grad_norm": 0.3502001464366913, "learning_rate": 5.176572624575109e-05, "loss": 1.3763172149658203, "step": 1070 }, { "epoch": 0.4373029594223533, "grad_norm": 0.30016398429870605, "learning_rate": 5.150838916016763e-05, "loss": 1.3476577758789063, "step": 1075 }, { "epoch": 0.4393369266754805, "grad_norm": 0.2778322398662567, "learning_rate": 5.125053171089974e-05, "loss": 1.366950225830078, "step": 1080 }, { "epoch": 0.44137089392860773, "grad_norm": 0.30253079533576965, "learning_rate": 5.099216555723794e-05, "loss": 1.4146586418151856, "step": 1085 }, { "epoch": 0.443404861181735, "grad_norm": 0.29914391040802, "learning_rate": 5.0733302381474295e-05, "loss": 1.3904467582702638, "step": 1090 }, { "epoch": 0.4454388284348622, "grad_norm": 0.32875391840934753, "learning_rate": 5.047395388837429e-05, "loss": 1.3523862838745118, "step": 1095 }, { "epoch": 0.4474727956879894, "grad_norm": 0.26864272356033325, "learning_rate": 5.021413180464748e-05, "loss": 1.36580228805542, "step": 1100 }, { "epoch": 0.44950676294111663, "grad_norm": 0.2538329064846039, "learning_rate": 4.995384787841735e-05, "loss": 1.3824766159057618, "step": 1105 }, { "epoch": 0.4515407301942439, "grad_norm": 0.3180592358112335, "learning_rate": 4.9693113878690054e-05, "loss": 1.3753103256225585, "step": 1110 }, { "epoch": 0.4535746974473711, "grad_norm": 0.3057761788368225, "learning_rate": 4.943194159482226e-05, "loss": 1.363571548461914, "step": 1115 }, { "epoch": 0.4556086647004983, "grad_norm": 0.31448543071746826, "learning_rate": 4.917034283598816e-05, "loss": 1.4244548797607421, "step": 1120 }, { "epoch": 0.45764263195362553, "grad_norm": 0.29453349113464355, "learning_rate": 4.8908329430645394e-05, "loss": 1.36822566986084, "step": 1125 }, { "epoch": 0.45967659920675275, "grad_norm": 0.2871814966201782, "learning_rate": 4.8645913226000307e-05, "loss": 1.3477168083190918, "step": 1130 }, { "epoch": 0.46171056645988, "grad_norm": 0.2801770567893982, "learning_rate": 4.8383106087472216e-05, "loss": 1.325541114807129, "step": 1135 }, { "epoch": 0.4637445337130072, "grad_norm": 0.2973332405090332, "learning_rate": 4.8119919898156934e-05, "loss": 1.37321195602417, "step": 1140 }, { "epoch": 0.46577850096613443, "grad_norm": 0.26172134280204773, "learning_rate": 4.785636655828945e-05, "loss": 1.4330161094665528, "step": 1145 }, { "epoch": 0.46781246821926165, "grad_norm": 0.27954164147377014, "learning_rate": 4.759245798470577e-05, "loss": 1.3601627349853516, "step": 1150 }, { "epoch": 0.4698464354723889, "grad_norm": 0.271559476852417, "learning_rate": 4.732820611030428e-05, "loss": 1.3042434692382812, "step": 1155 }, { "epoch": 0.4718804027255161, "grad_norm": 0.25448518991470337, "learning_rate": 4.706362288350595e-05, "loss": 1.3800363540649414, "step": 1160 }, { "epoch": 0.47391436997864334, "grad_norm": 0.3514825701713562, "learning_rate": 4.679872026771427e-05, "loss": 1.372146987915039, "step": 1165 }, { "epoch": 0.47594833723177055, "grad_norm": 0.33299720287323, "learning_rate": 4.653351024077421e-05, "loss": 1.3028112411499024, "step": 1170 }, { "epoch": 0.4779823044848978, "grad_norm": 0.26740890741348267, "learning_rate": 4.6268004794430625e-05, "loss": 1.335719394683838, "step": 1175 }, { "epoch": 0.480016271738025, "grad_norm": 0.3404020667076111, "learning_rate": 4.600221593378611e-05, "loss": 1.3231941223144532, "step": 1180 }, { "epoch": 0.48205023899115224, "grad_norm": 0.33737000823020935, "learning_rate": 4.573615567675809e-05, "loss": 1.3836843490600585, "step": 1185 }, { "epoch": 0.48408420624427945, "grad_norm": 0.29234227538108826, "learning_rate": 4.546983605353548e-05, "loss": 1.340735626220703, "step": 1190 }, { "epoch": 0.4861181734974067, "grad_norm": 0.2995792627334595, "learning_rate": 4.520326910603471e-05, "loss": 1.3517415046691894, "step": 1195 }, { "epoch": 0.4881521407505339, "grad_norm": 0.30948907136917114, "learning_rate": 4.493646688735519e-05, "loss": 1.366987133026123, "step": 1200 }, { "epoch": 0.49018610800366114, "grad_norm": 0.3263540267944336, "learning_rate": 4.466944146123441e-05, "loss": 1.3390093803405763, "step": 1205 }, { "epoch": 0.49222007525678835, "grad_norm": 0.29587090015411377, "learning_rate": 4.44022049015024e-05, "loss": 1.3350526809692382, "step": 1210 }, { "epoch": 0.4942540425099156, "grad_norm": 0.328902006149292, "learning_rate": 4.413476929153579e-05, "loss": 1.3469122886657714, "step": 1215 }, { "epoch": 0.4962880097630428, "grad_norm": 0.34572383761405945, "learning_rate": 4.38671467237115e-05, "loss": 1.3268512725830077, "step": 1220 }, { "epoch": 0.49832197701617004, "grad_norm": 0.33818185329437256, "learning_rate": 4.3599349298859924e-05, "loss": 1.3095382690429687, "step": 1225 }, { "epoch": 0.5003559442692973, "grad_norm": 0.2669960856437683, "learning_rate": 4.333138912571782e-05, "loss": 1.3039896965026856, "step": 1230 }, { "epoch": 0.5023899115224245, "grad_norm": 0.29393133521080017, "learning_rate": 4.306327832038074e-05, "loss": 1.3419631958007812, "step": 1235 }, { "epoch": 0.5044238787755517, "grad_norm": 0.2801229655742645, "learning_rate": 4.279502900575526e-05, "loss": 1.3102739334106446, "step": 1240 }, { "epoch": 0.5064578460286789, "grad_norm": 0.29053130745887756, "learning_rate": 4.2526653311010805e-05, "loss": 1.3475839614868164, "step": 1245 }, { "epoch": 0.5084918132818061, "grad_norm": 0.3196859657764435, "learning_rate": 4.2258163371031155e-05, "loss": 1.308444881439209, "step": 1250 }, { "epoch": 0.5105257805349334, "grad_norm": 0.2578422725200653, "learning_rate": 4.1989571325865856e-05, "loss": 1.2768720626831054, "step": 1255 }, { "epoch": 0.5125597477880606, "grad_norm": 0.2650304138660431, "learning_rate": 4.1720889320181224e-05, "loss": 1.3345602035522461, "step": 1260 }, { "epoch": 0.5145937150411878, "grad_norm": 0.35498908162117004, "learning_rate": 4.145212950271123e-05, "loss": 1.3324790954589845, "step": 1265 }, { "epoch": 0.5166276822943151, "grad_norm": 0.2839304804801941, "learning_rate": 4.118330402570821e-05, "loss": 1.3553616523742675, "step": 1270 }, { "epoch": 0.5186616495474423, "grad_norm": 0.29432836174964905, "learning_rate": 4.091442504439332e-05, "loss": 1.3277172088623046, "step": 1275 }, { "epoch": 0.5206956168005695, "grad_norm": 0.2678625285625458, "learning_rate": 4.0645504716407016e-05, "loss": 1.3551125526428223, "step": 1280 }, { "epoch": 0.5227295840536967, "grad_norm": 0.2659880220890045, "learning_rate": 4.037655520125925e-05, "loss": 1.367404842376709, "step": 1285 }, { "epoch": 0.524763551306824, "grad_norm": 0.27753204107284546, "learning_rate": 4.0107588659779715e-05, "loss": 1.3482036590576172, "step": 1290 }, { "epoch": 0.5267975185599512, "grad_norm": 0.26163339614868164, "learning_rate": 3.983861725356798e-05, "loss": 1.3400317192077638, "step": 1295 }, { "epoch": 0.5288314858130784, "grad_norm": 0.276053786277771, "learning_rate": 3.9569653144443546e-05, "loss": 1.302187728881836, "step": 1300 }, { "epoch": 0.5308654530662056, "grad_norm": 0.3655849099159241, "learning_rate": 3.930070849389601e-05, "loss": 1.3204415321350098, "step": 1305 }, { "epoch": 0.5328994203193329, "grad_norm": 0.29762181639671326, "learning_rate": 3.90317954625351e-05, "loss": 1.368598175048828, "step": 1310 }, { "epoch": 0.5349333875724601, "grad_norm": 0.2925015389919281, "learning_rate": 3.8762926209540846e-05, "loss": 1.2881315231323243, "step": 1315 }, { "epoch": 0.5369673548255873, "grad_norm": 0.3502486050128937, "learning_rate": 3.849411289211383e-05, "loss": 1.3247400283813477, "step": 1320 }, { "epoch": 0.5390013220787145, "grad_norm": 0.2803286015987396, "learning_rate": 3.822536766492539e-05, "loss": 1.345424747467041, "step": 1325 }, { "epoch": 0.5410352893318418, "grad_norm": 0.28980642557144165, "learning_rate": 3.795670267956818e-05, "loss": 1.3474702835083008, "step": 1330 }, { "epoch": 0.543069256584969, "grad_norm": 0.3616182506084442, "learning_rate": 3.768813008400652e-05, "loss": 1.3432271957397461, "step": 1335 }, { "epoch": 0.5451032238380962, "grad_norm": 0.27098384499549866, "learning_rate": 3.7419662022027356e-05, "loss": 1.3439754486083983, "step": 1340 }, { "epoch": 0.5471371910912234, "grad_norm": 0.34370723366737366, "learning_rate": 3.715131063269094e-05, "loss": 1.3055066108703612, "step": 1345 }, { "epoch": 0.5491711583443507, "grad_norm": 0.3249366581439972, "learning_rate": 3.688308804978212e-05, "loss": 1.336754035949707, "step": 1350 }, { "epoch": 0.5512051255974779, "grad_norm": 0.281678706407547, "learning_rate": 3.661500640126157e-05, "loss": 1.305647850036621, "step": 1355 }, { "epoch": 0.5532390928506051, "grad_norm": 0.3009491264820099, "learning_rate": 3.634707780871752e-05, "loss": 1.328620147705078, "step": 1360 }, { "epoch": 0.5552730601037323, "grad_norm": 0.28976067900657654, "learning_rate": 3.6079314386817575e-05, "loss": 1.4197721481323242, "step": 1365 }, { "epoch": 0.5573070273568596, "grad_norm": 0.3061009645462036, "learning_rate": 3.581172824276102e-05, "loss": 1.3280505180358886, "step": 1370 }, { "epoch": 0.5593409946099868, "grad_norm": 0.31219398975372314, "learning_rate": 3.554433147573132e-05, "loss": 1.3654787063598632, "step": 1375 }, { "epoch": 0.561374961863114, "grad_norm": 0.3715786933898926, "learning_rate": 3.527713617634904e-05, "loss": 1.3954607009887696, "step": 1380 }, { "epoch": 0.5634089291162412, "grad_norm": 0.28110235929489136, "learning_rate": 3.5010154426125227e-05, "loss": 1.3955727577209474, "step": 1385 }, { "epoch": 0.5654428963693685, "grad_norm": 0.31595325469970703, "learning_rate": 3.4743398296915024e-05, "loss": 1.3604469299316406, "step": 1390 }, { "epoch": 0.5674768636224957, "grad_norm": 0.29918357729911804, "learning_rate": 3.447687985037194e-05, "loss": 1.3830037117004395, "step": 1395 }, { "epoch": 0.5695108308756229, "grad_norm": 0.28876686096191406, "learning_rate": 3.421061113740237e-05, "loss": 1.3515007019042968, "step": 1400 }, { "epoch": 0.5715447981287501, "grad_norm": 0.2939162254333496, "learning_rate": 3.3944604197620814e-05, "loss": 1.3486581802368165, "step": 1405 }, { "epoch": 0.5735787653818774, "grad_norm": 0.2977907061576843, "learning_rate": 3.367887105880533e-05, "loss": 1.3130126953125, "step": 1410 }, { "epoch": 0.5756127326350046, "grad_norm": 0.3273273706436157, "learning_rate": 3.3413423736353894e-05, "loss": 1.35062894821167, "step": 1415 }, { "epoch": 0.5776466998881318, "grad_norm": 0.2974117696285248, "learning_rate": 3.31482742327409e-05, "loss": 1.320713710784912, "step": 1420 }, { "epoch": 0.579680667141259, "grad_norm": 0.38478466868400574, "learning_rate": 3.288343453697463e-05, "loss": 1.425486946105957, "step": 1425 }, { "epoch": 0.5817146343943862, "grad_norm": 0.3534519374370575, "learning_rate": 3.261891662405505e-05, "loss": 1.3442761421203613, "step": 1430 }, { "epoch": 0.5837486016475135, "grad_norm": 0.3142675459384918, "learning_rate": 3.2354732454432375e-05, "loss": 1.3268475532531738, "step": 1435 }, { "epoch": 0.5857825689006407, "grad_norm": 0.3035406172275543, "learning_rate": 3.2090893973466305e-05, "loss": 1.3226852416992188, "step": 1440 }, { "epoch": 0.5878165361537679, "grad_norm": 0.3025301992893219, "learning_rate": 3.1827413110885844e-05, "loss": 1.3266612052917481, "step": 1445 }, { "epoch": 0.5898505034068952, "grad_norm": 0.28104445338249207, "learning_rate": 3.156430178024992e-05, "loss": 1.3382318496704102, "step": 1450 }, { "epoch": 0.5918844706600224, "grad_norm": 0.287640780210495, "learning_rate": 3.130157187840869e-05, "loss": 1.3086122512817382, "step": 1455 }, { "epoch": 0.5939184379131496, "grad_norm": 0.28183332085609436, "learning_rate": 3.103923528496562e-05, "loss": 1.3180087089538575, "step": 1460 }, { "epoch": 0.5959524051662768, "grad_norm": 0.2993577718734741, "learning_rate": 3.0777303861740335e-05, "loss": 1.3392560958862305, "step": 1465 }, { "epoch": 0.597986372419404, "grad_norm": 0.2900375723838806, "learning_rate": 3.0515789452232236e-05, "loss": 1.3448850631713867, "step": 1470 }, { "epoch": 0.6000203396725313, "grad_norm": 0.24633945524692535, "learning_rate": 3.0254703881085093e-05, "loss": 1.3059326171875, "step": 1475 }, { "epoch": 0.6020543069256585, "grad_norm": 0.263874888420105, "learning_rate": 2.9994058953552233e-05, "loss": 1.3294736862182617, "step": 1480 }, { "epoch": 0.6040882741787857, "grad_norm": 0.2907504439353943, "learning_rate": 2.9733866454962867e-05, "loss": 1.3658993721008301, "step": 1485 }, { "epoch": 0.606122241431913, "grad_norm": 0.34029877185821533, "learning_rate": 2.947413815018914e-05, "loss": 1.3477883338928223, "step": 1490 }, { "epoch": 0.6081562086850402, "grad_norm": 0.27803581953048706, "learning_rate": 2.921488578311423e-05, "loss": 1.2762548446655273, "step": 1495 }, { "epoch": 0.6101901759381674, "grad_norm": 0.3703933358192444, "learning_rate": 2.8956121076101244e-05, "loss": 1.3101913452148437, "step": 1500 }, { "epoch": 0.6122241431912946, "grad_norm": 0.3121544122695923, "learning_rate": 2.8697855729463326e-05, "loss": 1.3178982734680176, "step": 1505 }, { "epoch": 0.6142581104444218, "grad_norm": 0.3321402072906494, "learning_rate": 2.844010142093441e-05, "loss": 1.31915340423584, "step": 1510 }, { "epoch": 0.6162920776975491, "grad_norm": 0.29758766293525696, "learning_rate": 2.8182869805141435e-05, "loss": 1.3031368255615234, "step": 1515 }, { "epoch": 0.6183260449506763, "grad_norm": 0.2962036728858948, "learning_rate": 2.7926172513077144e-05, "loss": 1.3243389129638672, "step": 1520 }, { "epoch": 0.6203600122038035, "grad_norm": 0.2997724413871765, "learning_rate": 2.7670021151574356e-05, "loss": 1.3242579460144044, "step": 1525 }, { "epoch": 0.6223939794569308, "grad_norm": 0.31056100130081177, "learning_rate": 2.741442730278105e-05, "loss": 1.3283409118652343, "step": 1530 }, { "epoch": 0.624427946710058, "grad_norm": 0.26714909076690674, "learning_rate": 2.715940252363666e-05, "loss": 1.2941314697265625, "step": 1535 }, { "epoch": 0.6264619139631852, "grad_norm": 0.4038776755332947, "learning_rate": 2.6904958345349633e-05, "loss": 1.3778356552124023, "step": 1540 }, { "epoch": 0.6284958812163124, "grad_norm": 0.2689177095890045, "learning_rate": 2.665110627287584e-05, "loss": 1.3075853347778321, "step": 1545 }, { "epoch": 0.6305298484694396, "grad_norm": 0.33493635058403015, "learning_rate": 2.6397857784398584e-05, "loss": 1.3377172470092773, "step": 1550 }, { "epoch": 0.6325638157225668, "grad_norm": 0.3054926097393036, "learning_rate": 2.6145224330809413e-05, "loss": 1.3471275329589845, "step": 1555 }, { "epoch": 0.6345977829756941, "grad_norm": 0.33510857820510864, "learning_rate": 2.5893217335190485e-05, "loss": 1.327333354949951, "step": 1560 }, { "epoch": 0.6366317502288213, "grad_norm": 0.3082568049430847, "learning_rate": 2.564184819229799e-05, "loss": 1.3183286666870118, "step": 1565 }, { "epoch": 0.6386657174819486, "grad_norm": 0.3351670503616333, "learning_rate": 2.5391128268046973e-05, "loss": 1.3455984115600585, "step": 1570 }, { "epoch": 0.6406996847350758, "grad_norm": 0.29731276631355286, "learning_rate": 2.514106889899735e-05, "loss": 1.2923025131225585, "step": 1575 }, { "epoch": 0.642733651988203, "grad_norm": 0.32418668270111084, "learning_rate": 2.48916813918414e-05, "loss": 1.2798898696899415, "step": 1580 }, { "epoch": 0.6447676192413302, "grad_norm": 0.320596307516098, "learning_rate": 2.4642977022892466e-05, "loss": 1.3829521179199218, "step": 1585 }, { "epoch": 0.6468015864944574, "grad_norm": 0.3012886643409729, "learning_rate": 2.439496703757504e-05, "loss": 1.2997875213623047, "step": 1590 }, { "epoch": 0.6488355537475846, "grad_norm": 0.30694982409477234, "learning_rate": 2.4147662649916416e-05, "loss": 1.3377013206481934, "step": 1595 }, { "epoch": 0.6508695210007119, "grad_norm": 0.3202068507671356, "learning_rate": 2.3901075042039484e-05, "loss": 1.2985017776489258, "step": 1600 }, { "epoch": 0.6529034882538391, "grad_norm": 0.3469294607639313, "learning_rate": 2.3655215363657283e-05, "loss": 1.3118343353271484, "step": 1605 }, { "epoch": 0.6549374555069664, "grad_norm": 0.2770400643348694, "learning_rate": 2.3410094731568684e-05, "loss": 1.3099807739257812, "step": 1610 }, { "epoch": 0.6569714227600936, "grad_norm": 0.2902584373950958, "learning_rate": 2.3165724229155887e-05, "loss": 1.3072982788085938, "step": 1615 }, { "epoch": 0.6590053900132208, "grad_norm": 0.32597291469573975, "learning_rate": 2.2922114905883167e-05, "loss": 1.3408931732177733, "step": 1620 }, { "epoch": 0.661039357266348, "grad_norm": 0.3000136613845825, "learning_rate": 2.2679277776797323e-05, "loss": 1.3196088790893554, "step": 1625 }, { "epoch": 0.6630733245194752, "grad_norm": 0.2830285429954529, "learning_rate": 2.243722382202954e-05, "loss": 1.326588249206543, "step": 1630 }, { "epoch": 0.6651072917726024, "grad_norm": 0.30097711086273193, "learning_rate": 2.2195963986299053e-05, "loss": 1.3054583549499512, "step": 1635 }, { "epoch": 0.6671412590257297, "grad_norm": 0.28644073009490967, "learning_rate": 2.195550917841815e-05, "loss": 1.3511393547058106, "step": 1640 }, { "epoch": 0.669175226278857, "grad_norm": 0.33317920565605164, "learning_rate": 2.171587027079892e-05, "loss": 1.3012124061584474, "step": 1645 }, { "epoch": 0.6712091935319842, "grad_norm": 0.3151117265224457, "learning_rate": 2.147705809896172e-05, "loss": 1.345674705505371, "step": 1650 }, { "epoch": 0.6732431607851114, "grad_norm": 0.34405115246772766, "learning_rate": 2.1239083461045178e-05, "loss": 1.3370431900024413, "step": 1655 }, { "epoch": 0.6752771280382386, "grad_norm": 0.29694926738739014, "learning_rate": 2.1001957117318002e-05, "loss": 1.3077101707458496, "step": 1660 }, { "epoch": 0.6773110952913658, "grad_norm": 0.33181169629096985, "learning_rate": 2.0765689789692317e-05, "loss": 1.3055456161499024, "step": 1665 }, { "epoch": 0.679345062544493, "grad_norm": 0.31455737352371216, "learning_rate": 2.053029216123906e-05, "loss": 1.3169405937194825, "step": 1670 }, { "epoch": 0.6813790297976202, "grad_norm": 0.27947914600372314, "learning_rate": 2.0295774875704726e-05, "loss": 1.3500905990600587, "step": 1675 }, { "epoch": 0.6834129970507474, "grad_norm": 0.31668201088905334, "learning_rate": 2.006214853703025e-05, "loss": 1.31624116897583, "step": 1680 }, { "epoch": 0.6854469643038748, "grad_norm": 0.2817995846271515, "learning_rate": 1.9829423708871465e-05, "loss": 1.2531603813171386, "step": 1685 }, { "epoch": 0.687480931557002, "grad_norm": 0.3473471999168396, "learning_rate": 1.9597610914121498e-05, "loss": 1.3449520111083983, "step": 1690 }, { "epoch": 0.6895148988101292, "grad_norm": 0.31141623854637146, "learning_rate": 1.9366720634434935e-05, "loss": 1.3509068489074707, "step": 1695 }, { "epoch": 0.6915488660632564, "grad_norm": 0.30905452370643616, "learning_rate": 1.913676330975385e-05, "loss": 1.3226234436035156, "step": 1700 }, { "epoch": 0.6935828333163836, "grad_norm": 0.3230130970478058, "learning_rate": 1.890774933783589e-05, "loss": 1.342616653442383, "step": 1705 }, { "epoch": 0.6956168005695108, "grad_norm": 0.32048484683036804, "learning_rate": 1.8679689073783925e-05, "loss": 1.3299571990966796, "step": 1710 }, { "epoch": 0.697650767822638, "grad_norm": 0.26584529876708984, "learning_rate": 1.845259282957802e-05, "loss": 1.3601978302001954, "step": 1715 }, { "epoch": 0.6996847350757652, "grad_norm": 0.3089277148246765, "learning_rate": 1.8226470873609043e-05, "loss": 1.341663646697998, "step": 1720 }, { "epoch": 0.7017187023288926, "grad_norm": 0.30121877789497375, "learning_rate": 1.800133343021446e-05, "loss": 1.339681911468506, "step": 1725 }, { "epoch": 0.7037526695820198, "grad_norm": 0.3219095468521118, "learning_rate": 1.7777190679215923e-05, "loss": 1.3493562698364259, "step": 1730 }, { "epoch": 0.705786636835147, "grad_norm": 0.3001890778541565, "learning_rate": 1.755405275545907e-05, "loss": 1.3034308433532715, "step": 1735 }, { "epoch": 0.7078206040882742, "grad_norm": 0.34980499744415283, "learning_rate": 1.73319297483553e-05, "loss": 1.300498104095459, "step": 1740 }, { "epoch": 0.7098545713414014, "grad_norm": 0.3548882007598877, "learning_rate": 1.7110831701425406e-05, "loss": 1.3058424949645997, "step": 1745 }, { "epoch": 0.7118885385945286, "grad_norm": 0.2676161527633667, "learning_rate": 1.6890768611845603e-05, "loss": 1.315950584411621, "step": 1750 }, { "epoch": 0.7139225058476558, "grad_norm": 0.31819629669189453, "learning_rate": 1.6671750429995444e-05, "loss": 1.3172121047973633, "step": 1755 }, { "epoch": 0.715956473100783, "grad_norm": 0.3689669966697693, "learning_rate": 1.645378705900791e-05, "loss": 1.311483097076416, "step": 1760 }, { "epoch": 0.7179904403539104, "grad_norm": 0.34183305501937866, "learning_rate": 1.623688835432155e-05, "loss": 1.2912078857421876, "step": 1765 }, { "epoch": 0.7200244076070376, "grad_norm": 0.2638580799102783, "learning_rate": 1.6021064123235057e-05, "loss": 1.32799072265625, "step": 1770 }, { "epoch": 0.7220583748601648, "grad_norm": 0.34436461329460144, "learning_rate": 1.5806324124463564e-05, "loss": 1.2865323066711425, "step": 1775 }, { "epoch": 0.724092342113292, "grad_norm": 0.3081623613834381, "learning_rate": 1.5592678067697594e-05, "loss": 1.3356817245483399, "step": 1780 }, { "epoch": 0.7261263093664192, "grad_norm": 0.29722127318382263, "learning_rate": 1.538013561316393e-05, "loss": 1.329073143005371, "step": 1785 }, { "epoch": 0.7281602766195464, "grad_norm": 0.29349419474601746, "learning_rate": 1.5168706371188852e-05, "loss": 1.360109519958496, "step": 1790 }, { "epoch": 0.7301942438726736, "grad_norm": 0.2984657883644104, "learning_rate": 1.4958399901763598e-05, "loss": 1.3421735763549805, "step": 1795 }, { "epoch": 0.7322282111258008, "grad_norm": 0.37349897623062134, "learning_rate": 1.4749225714112024e-05, "loss": 1.286893653869629, "step": 1800 }, { "epoch": 0.7342621783789282, "grad_norm": 0.2944484353065491, "learning_rate": 1.4541193266260805e-05, "loss": 1.3260204315185546, "step": 1805 }, { "epoch": 0.7362961456320554, "grad_norm": 0.3129271864891052, "learning_rate": 1.4334311964611569e-05, "loss": 1.3124813079833983, "step": 1810 }, { "epoch": 0.7383301128851826, "grad_norm": 0.33595311641693115, "learning_rate": 1.4128591163515739e-05, "loss": 1.315912628173828, "step": 1815 }, { "epoch": 0.7403640801383098, "grad_norm": 0.3355139493942261, "learning_rate": 1.3924040164851497e-05, "loss": 1.301335334777832, "step": 1820 }, { "epoch": 0.742398047391437, "grad_norm": 0.38431426882743835, "learning_rate": 1.3720668217603215e-05, "loss": 1.3113818168640137, "step": 1825 }, { "epoch": 0.7444320146445642, "grad_norm": 0.3036423921585083, "learning_rate": 1.3518484517443194e-05, "loss": 1.3164487838745118, "step": 1830 }, { "epoch": 0.7464659818976914, "grad_norm": 0.35051092505455017, "learning_rate": 1.331749820631595e-05, "loss": 1.3247812271118165, "step": 1835 }, { "epoch": 0.7484999491508186, "grad_norm": 0.3836432099342346, "learning_rate": 1.3117718372024859e-05, "loss": 1.3123231887817384, "step": 1840 }, { "epoch": 0.7505339164039458, "grad_norm": 0.30517545342445374, "learning_rate": 1.2919154047821114e-05, "loss": 1.3329618453979493, "step": 1845 }, { "epoch": 0.7525678836570732, "grad_norm": 0.30319541692733765, "learning_rate": 1.2721814211995427e-05, "loss": 1.3476465225219727, "step": 1850 }, { "epoch": 0.7546018509102004, "grad_norm": 0.2818050682544708, "learning_rate": 1.2525707787471993e-05, "loss": 1.2673000335693358, "step": 1855 }, { "epoch": 0.7566358181633276, "grad_norm": 0.31418922543525696, "learning_rate": 1.2330843641405057e-05, "loss": 1.2989997863769531, "step": 1860 }, { "epoch": 0.7586697854164548, "grad_norm": 0.3015255630016327, "learning_rate": 1.2137230584777898e-05, "loss": 1.3390537261962892, "step": 1865 }, { "epoch": 0.760703752669582, "grad_norm": 0.2891969382762909, "learning_rate": 1.1944877372004599e-05, "loss": 1.2631784439086915, "step": 1870 }, { "epoch": 0.7627377199227092, "grad_norm": 0.3059086203575134, "learning_rate": 1.1753792700534014e-05, "loss": 1.3097925186157227, "step": 1875 }, { "epoch": 0.7647716871758364, "grad_norm": 0.3744591474533081, "learning_rate": 1.1563985210456634e-05, "loss": 1.276920700073242, "step": 1880 }, { "epoch": 0.7668056544289636, "grad_norm": 0.3360139727592468, "learning_rate": 1.1375463484113878e-05, "loss": 1.28460693359375, "step": 1885 }, { "epoch": 0.768839621682091, "grad_norm": 0.3096720278263092, "learning_rate": 1.1188236045710017e-05, "loss": 1.3360540390014648, "step": 1890 }, { "epoch": 0.7708735889352182, "grad_norm": 0.29982519149780273, "learning_rate": 1.1002311360926776e-05, "loss": 1.363850212097168, "step": 1895 }, { "epoch": 0.7729075561883454, "grad_norm": 0.33204522728919983, "learning_rate": 1.0817697836540493e-05, "loss": 1.300833511352539, "step": 1900 }, { "epoch": 0.7749415234414726, "grad_norm": 0.3104506731033325, "learning_rate": 1.063440382004211e-05, "loss": 1.2700294494628905, "step": 1905 }, { "epoch": 0.7769754906945998, "grad_norm": 0.31560346484184265, "learning_rate": 1.045243759925957e-05, "loss": 1.3110705375671388, "step": 1910 }, { "epoch": 0.779009457947727, "grad_norm": 0.2993502616882324, "learning_rate": 1.0271807401983222e-05, "loss": 1.287836742401123, "step": 1915 }, { "epoch": 0.7810434252008542, "grad_norm": 0.3384788930416107, "learning_rate": 1.0092521395593722e-05, "loss": 1.3011297225952148, "step": 1920 }, { "epoch": 0.7830773924539814, "grad_norm": 0.33607131242752075, "learning_rate": 9.914587686692756e-06, "loss": 1.3346430778503418, "step": 1925 }, { "epoch": 0.7851113597071088, "grad_norm": 0.5029808282852173, "learning_rate": 9.738014320736448e-06, "loss": 1.3240625381469726, "step": 1930 }, { "epoch": 0.787145326960236, "grad_norm": 0.2745057940483093, "learning_rate": 9.562809281671654e-06, "loss": 1.2865424156188965, "step": 1935 }, { "epoch": 0.7891792942133632, "grad_norm": 0.34032368659973145, "learning_rate": 9.388980491574901e-06, "loss": 1.3326154708862306, "step": 1940 }, { "epoch": 0.7912132614664904, "grad_norm": 0.25044453144073486, "learning_rate": 9.216535810294207e-06, "loss": 1.312424373626709, "step": 1945 }, { "epoch": 0.7932472287196176, "grad_norm": 0.2655031979084015, "learning_rate": 9.045483035093668e-06, "loss": 1.343673038482666, "step": 1950 }, { "epoch": 0.7952811959727448, "grad_norm": 0.3135182857513428, "learning_rate": 8.875829900300928e-06, "loss": 1.3705928802490235, "step": 1955 }, { "epoch": 0.797315163225872, "grad_norm": 0.31982138752937317, "learning_rate": 8.707584076957443e-06, "loss": 1.3401841163635253, "step": 1960 }, { "epoch": 0.7993491304789992, "grad_norm": 0.3781191408634186, "learning_rate": 8.540753172471615e-06, "loss": 1.3395721435546875, "step": 1965 }, { "epoch": 0.8013830977321265, "grad_norm": 0.27083179354667664, "learning_rate": 8.375344730274846e-06, "loss": 1.3188928604125976, "step": 1970 }, { "epoch": 0.8034170649852538, "grad_norm": 0.3123554289340973, "learning_rate": 8.21136622948043e-06, "loss": 1.3410266876220702, "step": 1975 }, { "epoch": 0.805451032238381, "grad_norm": 0.3166046738624573, "learning_rate": 8.048825084545413e-06, "loss": 1.3043322563171387, "step": 1980 }, { "epoch": 0.8074849994915082, "grad_norm": 0.2929735481739044, "learning_rate": 7.887728644935256e-06, "loss": 1.3204127311706544, "step": 1985 }, { "epoch": 0.8095189667446354, "grad_norm": 0.337723970413208, "learning_rate": 7.728084194791648e-06, "loss": 1.2446714401245118, "step": 1990 }, { "epoch": 0.8115529339977626, "grad_norm": 0.3172164559364319, "learning_rate": 7.569898952603019e-06, "loss": 1.2812080383300781, "step": 1995 }, { "epoch": 0.8135869012508898, "grad_norm": 0.26954787969589233, "learning_rate": 7.413180070878225e-06, "loss": 1.3087306022644043, "step": 2000 }, { "epoch": 0.815620868504017, "grad_norm": 0.3072230815887451, "learning_rate": 7.257934635823165e-06, "loss": 1.3327550888061523, "step": 2005 }, { "epoch": 0.8176548357571443, "grad_norm": 0.3195462226867676, "learning_rate": 7.104169667020251e-06, "loss": 1.3385271072387694, "step": 2010 }, { "epoch": 0.8196888030102716, "grad_norm": 0.2930026054382324, "learning_rate": 6.951892117111141e-06, "loss": 1.2699304580688477, "step": 2015 }, { "epoch": 0.8217227702633988, "grad_norm": 0.3249114751815796, "learning_rate": 6.801108871482287e-06, "loss": 1.239455223083496, "step": 2020 }, { "epoch": 0.823756737516526, "grad_norm": 0.2966950535774231, "learning_rate": 6.651826747953664e-06, "loss": 1.302536964416504, "step": 2025 }, { "epoch": 0.8257907047696532, "grad_norm": 0.3190607726573944, "learning_rate": 6.504052496470415e-06, "loss": 1.3334075927734375, "step": 2030 }, { "epoch": 0.8278246720227804, "grad_norm": 0.28147175908088684, "learning_rate": 6.357792798797726e-06, "loss": 1.2913707733154296, "step": 2035 }, { "epoch": 0.8298586392759076, "grad_norm": 0.2634967863559723, "learning_rate": 6.213054268218672e-06, "loss": 1.3377763748168945, "step": 2040 }, { "epoch": 0.8318926065290349, "grad_norm": 0.31263190507888794, "learning_rate": 6.069843449235176e-06, "loss": 1.3046329498291016, "step": 2045 }, { "epoch": 0.8339265737821621, "grad_norm": 0.31119638681411743, "learning_rate": 5.9281668172721115e-06, "loss": 1.3355186462402344, "step": 2050 }, { "epoch": 0.8359605410352894, "grad_norm": 0.3515127897262573, "learning_rate": 5.7880307783845055e-06, "loss": 1.2901551246643066, "step": 2055 }, { "epoch": 0.8379945082884166, "grad_norm": 0.36530110239982605, "learning_rate": 5.6494416689678855e-06, "loss": 1.3465386390686036, "step": 2060 }, { "epoch": 0.8400284755415438, "grad_norm": 0.28482675552368164, "learning_rate": 5.5124057554717395e-06, "loss": 1.3183640480041503, "step": 2065 }, { "epoch": 0.842062442794671, "grad_norm": 0.2931864857673645, "learning_rate": 5.376929234116235e-06, "loss": 1.2733064651489259, "step": 2070 }, { "epoch": 0.8440964100477982, "grad_norm": 0.350818008184433, "learning_rate": 5.243018230611996e-06, "loss": 1.3183445930480957, "step": 2075 }, { "epoch": 0.8461303773009254, "grad_norm": 0.2808331549167633, "learning_rate": 5.110678799883162e-06, "loss": 1.316090965270996, "step": 2080 }, { "epoch": 0.8481643445540527, "grad_norm": 0.3167017698287964, "learning_rate": 4.979916925793542e-06, "loss": 1.342299747467041, "step": 2085 }, { "epoch": 0.8501983118071799, "grad_norm": 0.3244977295398712, "learning_rate": 4.850738520876164e-06, "loss": 1.2839815139770507, "step": 2090 }, { "epoch": 0.8522322790603071, "grad_norm": 0.3639869689941406, "learning_rate": 4.723149426065803e-06, "loss": 1.300691509246826, "step": 2095 }, { "epoch": 0.8542662463134344, "grad_norm": 0.38750073313713074, "learning_rate": 4.597155410434972e-06, "loss": 1.2385212898254394, "step": 2100 }, { "epoch": 0.8563002135665616, "grad_norm": 0.2886674106121063, "learning_rate": 4.472762170933021e-06, "loss": 1.3366718292236328, "step": 2105 }, { "epoch": 0.8583341808196888, "grad_norm": 0.3020629584789276, "learning_rate": 4.349975332128562e-06, "loss": 1.2933056831359864, "step": 2110 }, { "epoch": 0.860368148072816, "grad_norm": 0.29996567964553833, "learning_rate": 4.22880044595515e-06, "loss": 1.2919550895690919, "step": 2115 }, { "epoch": 0.8624021153259432, "grad_norm": 0.30315011739730835, "learning_rate": 4.109242991460205e-06, "loss": 1.2908141136169433, "step": 2120 }, { "epoch": 0.8644360825790705, "grad_norm": 0.31712713837623596, "learning_rate": 3.991308374557363e-06, "loss": 1.3030555725097657, "step": 2125 }, { "epoch": 0.8664700498321977, "grad_norm": 0.31594905257225037, "learning_rate": 3.875001927781932e-06, "loss": 1.2869560241699218, "step": 2130 }, { "epoch": 0.8685040170853249, "grad_norm": 0.2830830216407776, "learning_rate": 3.760328910049853e-06, "loss": 1.3083958625793457, "step": 2135 }, { "epoch": 0.8705379843384522, "grad_norm": 0.31977933645248413, "learning_rate": 3.6472945064198827e-06, "loss": 1.297631072998047, "step": 2140 }, { "epoch": 0.8725719515915794, "grad_norm": 0.308992475271225, "learning_rate": 3.5359038278591683e-06, "loss": 1.2782371520996094, "step": 2145 }, { "epoch": 0.8746059188447066, "grad_norm": 0.33491194248199463, "learning_rate": 3.426161911012087e-06, "loss": 1.3169782638549805, "step": 2150 }, { "epoch": 0.8766398860978338, "grad_norm": 0.322786420583725, "learning_rate": 3.3180737179726097e-06, "loss": 1.2556682586669923, "step": 2155 }, { "epoch": 0.878673853350961, "grad_norm": 0.2721087336540222, "learning_rate": 3.2116441360598593e-06, "loss": 1.278902816772461, "step": 2160 }, { "epoch": 0.8807078206040883, "grad_norm": 0.3361051380634308, "learning_rate": 3.1068779775971268e-06, "loss": 1.2502001762390136, "step": 2165 }, { "epoch": 0.8827417878572155, "grad_norm": 0.29746389389038086, "learning_rate": 3.0037799796943123e-06, "loss": 1.3131372451782226, "step": 2170 }, { "epoch": 0.8847757551103427, "grad_norm": 0.27195727825164795, "learning_rate": 2.9023548040337266e-06, "loss": 1.267716884613037, "step": 2175 }, { "epoch": 0.88680972236347, "grad_norm": 0.33992087841033936, "learning_rate": 2.802607036659284e-06, "loss": 1.3220355987548829, "step": 2180 }, { "epoch": 0.8888436896165972, "grad_norm": 0.3313898742198944, "learning_rate": 2.7045411877691365e-06, "loss": 1.2927869796752929, "step": 2185 }, { "epoch": 0.8908776568697244, "grad_norm": 0.2934510409832001, "learning_rate": 2.608161691511808e-06, "loss": 1.3552505493164062, "step": 2190 }, { "epoch": 0.8929116241228516, "grad_norm": 0.27263787388801575, "learning_rate": 2.51347290578559e-06, "loss": 1.3077566146850585, "step": 2195 }, { "epoch": 0.8949455913759788, "grad_norm": 0.32106345891952515, "learning_rate": 2.420479112041596e-06, "loss": 1.3361632347106933, "step": 2200 }, { "epoch": 0.896979558629106, "grad_norm": 0.303587406873703, "learning_rate": 2.3291845150901128e-06, "loss": 1.2889724731445313, "step": 2205 }, { "epoch": 0.8990135258822333, "grad_norm": 0.3349932134151459, "learning_rate": 2.2395932429104892e-06, "loss": 1.277025318145752, "step": 2210 }, { "epoch": 0.9010474931353605, "grad_norm": 0.36167261004447937, "learning_rate": 2.1517093464644967e-06, "loss": 1.2718290328979491, "step": 2215 }, { "epoch": 0.9030814603884878, "grad_norm": 0.3638148307800293, "learning_rate": 2.065536799513135e-06, "loss": 1.285135555267334, "step": 2220 }, { "epoch": 0.905115427641615, "grad_norm": 0.3448006212711334, "learning_rate": 1.981079498437004e-06, "loss": 1.356567668914795, "step": 2225 }, { "epoch": 0.9071493948947422, "grad_norm": 0.3148203492164612, "learning_rate": 1.8983412620600594e-06, "loss": 1.2908525466918945, "step": 2230 }, { "epoch": 0.9091833621478694, "grad_norm": 0.26251962780952454, "learning_rate": 1.8173258314769837e-06, "loss": 1.3202886581420898, "step": 2235 }, { "epoch": 0.9112173294009966, "grad_norm": 0.34208986163139343, "learning_rate": 1.7380368698840256e-06, "loss": 1.257115364074707, "step": 2240 }, { "epoch": 0.9132512966541239, "grad_norm": 0.3172779679298401, "learning_rate": 1.6604779624133626e-06, "loss": 1.3262968063354492, "step": 2245 }, { "epoch": 0.9152852639072511, "grad_norm": 0.3330855369567871, "learning_rate": 1.5846526159709653e-06, "loss": 1.2560693740844726, "step": 2250 }, { "epoch": 0.9173192311603783, "grad_norm": 0.36517953872680664, "learning_rate": 1.5105642590780778e-06, "loss": 1.2578863143920898, "step": 2255 }, { "epoch": 0.9193531984135055, "grad_norm": 0.2745852768421173, "learning_rate": 1.4382162417161748e-06, "loss": 1.2556609153747558, "step": 2260 }, { "epoch": 0.9213871656666328, "grad_norm": 0.34077954292297363, "learning_rate": 1.3676118351754598e-06, "loss": 1.268846321105957, "step": 2265 }, { "epoch": 0.92342113291976, "grad_norm": 0.31594374775886536, "learning_rate": 1.2987542319069913e-06, "loss": 1.3389588356018067, "step": 2270 }, { "epoch": 0.9254551001728872, "grad_norm": 0.2979227900505066, "learning_rate": 1.23164654537832e-06, "loss": 1.3274189949035644, "step": 2275 }, { "epoch": 0.9274890674260144, "grad_norm": 0.3171345591545105, "learning_rate": 1.166291809932698e-06, "loss": 1.29821834564209, "step": 2280 }, { "epoch": 0.9295230346791417, "grad_norm": 0.3295547068119049, "learning_rate": 1.1026929806518738e-06, "loss": 1.287254524230957, "step": 2285 }, { "epoch": 0.9315570019322689, "grad_norm": 0.35802721977233887, "learning_rate": 1.040852933222527e-06, "loss": 1.307052993774414, "step": 2290 }, { "epoch": 0.9335909691853961, "grad_norm": 0.29962143301963806, "learning_rate": 9.8077446380616e-07, "loss": 1.3314196586608886, "step": 2295 }, { "epoch": 0.9356249364385233, "grad_norm": 0.3351229727268219, "learning_rate": 9.224602889127321e-07, "loss": 1.311518955230713, "step": 2300 }, { "epoch": 0.9376589036916506, "grad_norm": 0.2955498993396759, "learning_rate": 8.65913045277802e-07, "loss": 1.2884058952331543, "step": 2305 }, { "epoch": 0.9396928709447778, "grad_norm": 0.3082011342048645, "learning_rate": 8.111352897432989e-07, "loss": 1.2431337356567382, "step": 2310 }, { "epoch": 0.941726838197905, "grad_norm": 0.3545486032962799, "learning_rate": 7.581294991419441e-07, "loss": 1.2952088356018066, "step": 2315 }, { "epoch": 0.9437608054510322, "grad_norm": 0.32563820481300354, "learning_rate": 7.068980701852113e-07, "loss": 1.319763469696045, "step": 2320 }, { "epoch": 0.9457947727041595, "grad_norm": 0.3194541335105896, "learning_rate": 6.574433193550089e-07, "loss": 1.276276206970215, "step": 2325 }, { "epoch": 0.9478287399572867, "grad_norm": 0.26381540298461914, "learning_rate": 6.097674827988842e-07, "loss": 1.33615140914917, "step": 2330 }, { "epoch": 0.9498627072104139, "grad_norm": 0.3122498393058777, "learning_rate": 5.638727162289659e-07, "loss": 1.3045156478881836, "step": 2335 }, { "epoch": 0.9518966744635411, "grad_norm": 0.29193928837776184, "learning_rate": 5.197610948244469e-07, "loss": 1.239583396911621, "step": 2340 }, { "epoch": 0.9539306417166684, "grad_norm": 0.32634857296943665, "learning_rate": 4.774346131377927e-07, "loss": 1.3298480987548829, "step": 2345 }, { "epoch": 0.9559646089697956, "grad_norm": 0.33866873383522034, "learning_rate": 4.368951850045111e-07, "loss": 1.2732733726501464, "step": 2350 }, { "epoch": 0.9579985762229228, "grad_norm": 0.30094456672668457, "learning_rate": 3.9814464345665717e-07, "loss": 1.3632909774780273, "step": 2355 }, { "epoch": 0.96003254347605, "grad_norm": 0.28664594888687134, "learning_rate": 3.611847406399349e-07, "loss": 1.3255720138549805, "step": 2360 }, { "epoch": 0.9620665107291773, "grad_norm": 0.30658918619155884, "learning_rate": 3.260171477344587e-07, "loss": 1.2762239456176758, "step": 2365 }, { "epoch": 0.9641004779823045, "grad_norm": 0.2883574962615967, "learning_rate": 2.9264345487921786e-07, "loss": 1.2771632194519043, "step": 2370 }, { "epoch": 0.9661344452354317, "grad_norm": 0.2709819972515106, "learning_rate": 2.6106517110015217e-07, "loss": 1.249297523498535, "step": 2375 }, { "epoch": 0.9681684124885589, "grad_norm": 0.3463249206542969, "learning_rate": 2.3128372424192635e-07, "loss": 1.2934127807617188, "step": 2380 }, { "epoch": 0.9702023797416861, "grad_norm": 0.28999075293540955, "learning_rate": 2.0330046090336842e-07, "loss": 1.261418342590332, "step": 2385 }, { "epoch": 0.9722363469948134, "grad_norm": 0.3243755102157593, "learning_rate": 1.771166463765983e-07, "loss": 1.2737961769104005, "step": 2390 }, { "epoch": 0.9742703142479406, "grad_norm": 0.3598092496395111, "learning_rate": 1.5273346458979376e-07, "loss": 1.2843939781188964, "step": 2395 }, { "epoch": 0.9763042815010678, "grad_norm": 0.3956236243247986, "learning_rate": 1.301520180536553e-07, "loss": 1.3170789718627929, "step": 2400 }, { "epoch": 0.9783382487541951, "grad_norm": 0.3373051583766937, "learning_rate": 1.0937332781158383e-07, "loss": 1.2935243606567384, "step": 2405 }, { "epoch": 0.9803722160073223, "grad_norm": 0.29248279333114624, "learning_rate": 9.039833339348658e-08, "loss": 1.2881365776062013, "step": 2410 }, { "epoch": 0.9824061832604495, "grad_norm": 0.31586557626724243, "learning_rate": 7.322789277330433e-08, "loss": 1.3256579399108888, "step": 2415 }, { "epoch": 0.9844401505135767, "grad_norm": 0.35788750648498535, "learning_rate": 5.7862782330215804e-08, "loss": 1.2509141921997071, "step": 2420 }, { "epoch": 0.9864741177667039, "grad_norm": 0.3229547142982483, "learning_rate": 4.43036968135413e-08, "loss": 1.2915958404541015, "step": 2425 }, { "epoch": 0.9885080850198312, "grad_norm": 0.30491015315055847, "learning_rate": 3.255124931131004e-08, "loss": 1.3207027435302734, "step": 2430 }, { "epoch": 0.9905420522729584, "grad_norm": 0.3059276342391968, "learning_rate": 2.260597122255792e-08, "loss": 1.3477352142333985, "step": 2435 }, { "epoch": 0.9925760195260857, "grad_norm": 0.28954240679740906, "learning_rate": 1.4468312233288929e-08, "loss": 1.309218406677246, "step": 2440 }, { "epoch": 0.9946099867792129, "grad_norm": 0.34162813425064087, "learning_rate": 8.138640296149192e-09, "loss": 1.3401227951049806, "step": 2445 }, { "epoch": 0.9966439540323401, "grad_norm": 0.2948080897331238, "learning_rate": 3.6172416137825184e-09, "loss": 1.3475109100341798, "step": 2450 }, { "epoch": 0.9986779212854673, "grad_norm": 0.2868635058403015, "learning_rate": 9.043206258940729e-10, "loss": 1.307164764404297, "step": 2455 } ], "logging_steps": 5, "max_steps": 2459, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 3.0061186646645146e+17, "train_batch_size": 1, "trial_name": null, "trial_params": null }