{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.23539970133662894, "eval_steps": 500, "global_step": 8000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "grad_norm": 31.245906829833984, "learning_rate": 5e-06, "loss": 12.4767, "step": 5 }, { "epoch": 0.0, "grad_norm": 29.32766342163086, "learning_rate": 1e-05, "loss": 12.1908, "step": 10 }, { "epoch": 0.0, "grad_norm": 31.548030853271484, "learning_rate": 1.5e-05, "loss": 11.4907, "step": 15 }, { "epoch": 0.0, "grad_norm": 28.47829818725586, "learning_rate": 2e-05, "loss": 10.5741, "step": 20 }, { "epoch": 0.0, "grad_norm": 33.89305877685547, "learning_rate": 2.5e-05, "loss": 9.6283, "step": 25 }, { "epoch": 0.0, "grad_norm": 13.979351997375488, "learning_rate": 3e-05, "loss": 8.833, "step": 30 }, { "epoch": 0.0, "grad_norm": 20.524765014648438, "learning_rate": 3.5e-05, "loss": 8.3122, "step": 35 }, { "epoch": 0.0, "grad_norm": 10.864826202392578, "learning_rate": 4e-05, "loss": 8.0788, "step": 40 }, { "epoch": 0.0, "grad_norm": 7.724523544311523, "learning_rate": 4.5e-05, "loss": 7.935, "step": 45 }, { "epoch": 0.0, "grad_norm": 8.686882019042969, "learning_rate": 5e-05, "loss": 7.8363, "step": 50 }, { "epoch": 0.0, "grad_norm": 8.194204330444336, "learning_rate": 5.500000000000001e-05, "loss": 7.7852, "step": 55 }, { "epoch": 0.0, "grad_norm": 6.38210391998291, "learning_rate": 6e-05, "loss": 7.673, "step": 60 }, { "epoch": 0.0, "grad_norm": 6.843475818634033, "learning_rate": 6.500000000000001e-05, "loss": 7.5376, "step": 65 }, { "epoch": 0.0, "grad_norm": 7.340760707855225, "learning_rate": 7e-05, "loss": 7.4279, "step": 70 }, { "epoch": 0.0, "grad_norm": 7.592554092407227, "learning_rate": 7.500000000000001e-05, "loss": 7.4571, "step": 75 }, { "epoch": 0.0, "grad_norm": 6.165151119232178, "learning_rate": 8e-05, "loss": 7.1802, "step": 80 }, { "epoch": 0.0, "grad_norm": 5.665134429931641, "learning_rate": 8.5e-05, "loss": 7.2924, "step": 85 }, { "epoch": 0.0, "grad_norm": 7.772270202636719, "learning_rate": 9e-05, "loss": 7.1999, "step": 90 }, { "epoch": 0.0, "grad_norm": 7.570541858673096, "learning_rate": 9.5e-05, "loss": 7.0424, "step": 95 }, { "epoch": 0.0, "grad_norm": 9.979948043823242, "learning_rate": 0.0001, "loss": 6.9661, "step": 100 }, { "epoch": 0.0, "grad_norm": 9.006866455078125, "learning_rate": 9.999999462732701e-05, "loss": 6.968, "step": 105 }, { "epoch": 0.0, "grad_norm": 9.186387062072754, "learning_rate": 9.999997850930916e-05, "loss": 6.7222, "step": 110 }, { "epoch": 0.0, "grad_norm": 10.189663887023926, "learning_rate": 9.999995164594991e-05, "loss": 6.7098, "step": 115 }, { "epoch": 0.0, "grad_norm": 7.710358142852783, "learning_rate": 9.999991403725507e-05, "loss": 6.7266, "step": 120 }, { "epoch": 0.0, "grad_norm": 5.905910491943359, "learning_rate": 9.99998656832327e-05, "loss": 6.6801, "step": 125 }, { "epoch": 0.0, "grad_norm": 10.010580062866211, "learning_rate": 9.999980658389317e-05, "loss": 6.5847, "step": 130 }, { "epoch": 0.0, "grad_norm": 9.009326934814453, "learning_rate": 9.999973673924921e-05, "loss": 6.4997, "step": 135 }, { "epoch": 0.0, "grad_norm": 8.097753524780273, "learning_rate": 9.999965614931583e-05, "loss": 6.4689, "step": 140 }, { "epoch": 0.0, "grad_norm": 11.300614356994629, "learning_rate": 9.999956481411035e-05, "loss": 6.3991, "step": 145 }, { "epoch": 0.0, "grad_norm": 7.934354782104492, "learning_rate": 9.999946273365238e-05, "loss": 6.2501, "step": 150 }, { "epoch": 0.0, "grad_norm": 11.2971830368042, "learning_rate": 9.999934990796387e-05, "loss": 6.2159, "step": 155 }, { "epoch": 0.0, "grad_norm": 7.320553779602051, "learning_rate": 9.999922633706906e-05, "loss": 6.1304, "step": 160 }, { "epoch": 0.0, "grad_norm": 5.86361026763916, "learning_rate": 9.999909202099453e-05, "loss": 6.1929, "step": 165 }, { "epoch": 0.01, "grad_norm": 10.322848320007324, "learning_rate": 9.999894695976912e-05, "loss": 6.0066, "step": 170 }, { "epoch": 0.01, "grad_norm": 12.492103576660156, "learning_rate": 9.9998791153424e-05, "loss": 6.0208, "step": 175 }, { "epoch": 0.01, "grad_norm": 8.456936836242676, "learning_rate": 9.999862460199267e-05, "loss": 5.9207, "step": 180 }, { "epoch": 0.01, "grad_norm": 5.34593391418457, "learning_rate": 9.999844730551092e-05, "loss": 5.8864, "step": 185 }, { "epoch": 0.01, "grad_norm": 9.129504203796387, "learning_rate": 9.999825926401686e-05, "loss": 5.817, "step": 190 }, { "epoch": 0.01, "grad_norm": 6.712257385253906, "learning_rate": 9.999806047755088e-05, "loss": 5.9447, "step": 195 }, { "epoch": 0.01, "grad_norm": 6.016158103942871, "learning_rate": 9.999785094615573e-05, "loss": 5.8714, "step": 200 }, { "epoch": 0.01, "grad_norm": 5.835657119750977, "learning_rate": 9.99976306698764e-05, "loss": 5.7875, "step": 205 }, { "epoch": 0.01, "grad_norm": 4.937032699584961, "learning_rate": 9.999739964876026e-05, "loss": 5.7525, "step": 210 }, { "epoch": 0.01, "grad_norm": 3.235369920730591, "learning_rate": 9.999715788285695e-05, "loss": 5.752, "step": 215 }, { "epoch": 0.01, "grad_norm": 4.9037346839904785, "learning_rate": 9.999690537221844e-05, "loss": 5.5506, "step": 220 }, { "epoch": 0.01, "grad_norm": 7.131804466247559, "learning_rate": 9.999664211689897e-05, "loss": 5.7098, "step": 225 }, { "epoch": 0.01, "grad_norm": 4.668625831604004, "learning_rate": 9.999636811695513e-05, "loss": 5.6286, "step": 230 }, { "epoch": 0.01, "grad_norm": 5.116443157196045, "learning_rate": 9.99960833724458e-05, "loss": 5.5792, "step": 235 }, { "epoch": 0.01, "grad_norm": 7.184804439544678, "learning_rate": 9.999578788343218e-05, "loss": 5.5886, "step": 240 }, { "epoch": 0.01, "grad_norm": 6.27599573135376, "learning_rate": 9.999548164997778e-05, "loss": 5.5753, "step": 245 }, { "epoch": 0.01, "grad_norm": 6.488445281982422, "learning_rate": 9.999516467214839e-05, "loss": 5.523, "step": 250 }, { "epoch": 0.01, "grad_norm": 5.085388660430908, "learning_rate": 9.999483695001214e-05, "loss": 5.4937, "step": 255 }, { "epoch": 0.01, "grad_norm": 5.797093868255615, "learning_rate": 9.999449848363946e-05, "loss": 5.4659, "step": 260 }, { "epoch": 0.01, "grad_norm": 5.1919732093811035, "learning_rate": 9.999414927310309e-05, "loss": 5.3514, "step": 265 }, { "epoch": 0.01, "grad_norm": 7.658706188201904, "learning_rate": 9.99937893184781e-05, "loss": 5.5521, "step": 270 }, { "epoch": 0.01, "grad_norm": 3.8249082565307617, "learning_rate": 9.99934186198418e-05, "loss": 5.278, "step": 275 }, { "epoch": 0.01, "grad_norm": 5.147459506988525, "learning_rate": 9.999303717727389e-05, "loss": 5.2519, "step": 280 }, { "epoch": 0.01, "grad_norm": 4.760087013244629, "learning_rate": 9.999264499085633e-05, "loss": 5.4345, "step": 285 }, { "epoch": 0.01, "grad_norm": 4.443284034729004, "learning_rate": 9.999224206067341e-05, "loss": 5.2506, "step": 290 }, { "epoch": 0.01, "grad_norm": 5.569728374481201, "learning_rate": 9.999182838681171e-05, "loss": 5.2378, "step": 295 }, { "epoch": 0.01, "grad_norm": 4.8742356300354, "learning_rate": 9.999140396936016e-05, "loss": 5.1968, "step": 300 }, { "epoch": 0.01, "grad_norm": 3.438101053237915, "learning_rate": 9.999096880840994e-05, "loss": 5.2749, "step": 305 }, { "epoch": 0.01, "grad_norm": 5.271472930908203, "learning_rate": 9.999052290405459e-05, "loss": 5.1478, "step": 310 }, { "epoch": 0.01, "grad_norm": 4.731297016143799, "learning_rate": 9.999006625638994e-05, "loss": 5.2142, "step": 315 }, { "epoch": 0.01, "grad_norm": 3.0178442001342773, "learning_rate": 9.998959886551411e-05, "loss": 5.199, "step": 320 }, { "epoch": 0.01, "grad_norm": 4.84102725982666, "learning_rate": 9.998912073152755e-05, "loss": 5.2397, "step": 325 }, { "epoch": 0.01, "grad_norm": 5.11923885345459, "learning_rate": 9.998863185453302e-05, "loss": 5.1836, "step": 330 }, { "epoch": 0.01, "grad_norm": 3.0714895725250244, "learning_rate": 9.998813223463557e-05, "loss": 5.1681, "step": 335 }, { "epoch": 0.01, "grad_norm": 4.088279724121094, "learning_rate": 9.998762187194258e-05, "loss": 5.0567, "step": 340 }, { "epoch": 0.01, "grad_norm": 3.813769578933716, "learning_rate": 9.998710076656373e-05, "loss": 5.0646, "step": 345 }, { "epoch": 0.01, "grad_norm": 5.132027626037598, "learning_rate": 9.998656891861102e-05, "loss": 5.1077, "step": 350 }, { "epoch": 0.01, "grad_norm": 3.61967396736145, "learning_rate": 9.998602632819874e-05, "loss": 5.1387, "step": 355 }, { "epoch": 0.01, "grad_norm": 3.9906136989593506, "learning_rate": 9.998547299544348e-05, "loss": 5.0749, "step": 360 }, { "epoch": 0.01, "grad_norm": 3.384110689163208, "learning_rate": 9.99849089204642e-05, "loss": 5.0254, "step": 365 }, { "epoch": 0.01, "grad_norm": 3.2073588371276855, "learning_rate": 9.998433410338206e-05, "loss": 5.1226, "step": 370 }, { "epoch": 0.01, "grad_norm": 2.6449954509735107, "learning_rate": 9.998374854432064e-05, "loss": 5.0906, "step": 375 }, { "epoch": 0.01, "grad_norm": 3.6864638328552246, "learning_rate": 9.998315224340577e-05, "loss": 5.0479, "step": 380 }, { "epoch": 0.01, "grad_norm": 2.5346381664276123, "learning_rate": 9.99825452007656e-05, "loss": 4.9982, "step": 385 }, { "epoch": 0.01, "grad_norm": 2.4365124702453613, "learning_rate": 9.998192741653057e-05, "loss": 5.0591, "step": 390 }, { "epoch": 0.01, "grad_norm": 2.493992805480957, "learning_rate": 9.998129889083346e-05, "loss": 4.9182, "step": 395 }, { "epoch": 0.01, "grad_norm": 2.805351734161377, "learning_rate": 9.998065962380936e-05, "loss": 5.0363, "step": 400 }, { "epoch": 0.01, "grad_norm": 3.115131139755249, "learning_rate": 9.998000961559562e-05, "loss": 5.0244, "step": 405 }, { "epoch": 0.01, "grad_norm": 2.8879072666168213, "learning_rate": 9.997934886633195e-05, "loss": 5.056, "step": 410 }, { "epoch": 0.01, "grad_norm": 3.224872589111328, "learning_rate": 9.997867737616035e-05, "loss": 5.0275, "step": 415 }, { "epoch": 0.01, "grad_norm": 3.734250068664551, "learning_rate": 9.997799514522513e-05, "loss": 4.9506, "step": 420 }, { "epoch": 0.01, "grad_norm": 3.154108762741089, "learning_rate": 9.99773021736729e-05, "loss": 4.8751, "step": 425 }, { "epoch": 0.01, "grad_norm": 2.869624137878418, "learning_rate": 9.99765984616526e-05, "loss": 5.0141, "step": 430 }, { "epoch": 0.01, "grad_norm": 2.4207420349121094, "learning_rate": 9.997588400931542e-05, "loss": 4.8798, "step": 435 }, { "epoch": 0.01, "grad_norm": 2.6371402740478516, "learning_rate": 9.997515881681496e-05, "loss": 4.936, "step": 440 }, { "epoch": 0.01, "grad_norm": 2.990797281265259, "learning_rate": 9.9974422884307e-05, "loss": 4.887, "step": 445 }, { "epoch": 0.01, "grad_norm": 3.5174179077148438, "learning_rate": 9.997367621194977e-05, "loss": 4.9114, "step": 450 }, { "epoch": 0.01, "grad_norm": 3.26190185546875, "learning_rate": 9.99729187999037e-05, "loss": 5.0168, "step": 455 }, { "epoch": 0.01, "grad_norm": 3.5801522731781006, "learning_rate": 9.997215064833155e-05, "loss": 4.8883, "step": 460 }, { "epoch": 0.01, "grad_norm": 3.067654848098755, "learning_rate": 9.99713717573984e-05, "loss": 4.6865, "step": 465 }, { "epoch": 0.01, "grad_norm": 3.806952714920044, "learning_rate": 9.997058212727169e-05, "loss": 4.7972, "step": 470 }, { "epoch": 0.01, "grad_norm": 3.139887809753418, "learning_rate": 9.996978175812105e-05, "loss": 4.7952, "step": 475 }, { "epoch": 0.01, "grad_norm": 2.5752458572387695, "learning_rate": 9.996897065011852e-05, "loss": 4.9118, "step": 480 }, { "epoch": 0.01, "grad_norm": 2.8559348583221436, "learning_rate": 9.996814880343842e-05, "loss": 4.7684, "step": 485 }, { "epoch": 0.01, "grad_norm": 2.4169933795928955, "learning_rate": 9.996731621825736e-05, "loss": 4.7103, "step": 490 }, { "epoch": 0.01, "grad_norm": 2.2148659229278564, "learning_rate": 9.996647289475426e-05, "loss": 4.7409, "step": 495 }, { "epoch": 0.01, "grad_norm": 2.9424631595611572, "learning_rate": 9.996561883311036e-05, "loss": 4.8819, "step": 500 }, { "epoch": 0.01, "grad_norm": 2.1137075424194336, "learning_rate": 9.996475403350921e-05, "loss": 4.8252, "step": 505 }, { "epoch": 0.02, "grad_norm": 2.317061185836792, "learning_rate": 9.996387849613667e-05, "loss": 4.7179, "step": 510 }, { "epoch": 0.02, "grad_norm": 2.3563737869262695, "learning_rate": 9.996299222118086e-05, "loss": 4.7741, "step": 515 }, { "epoch": 0.02, "grad_norm": 1.9870095252990723, "learning_rate": 9.99620952088323e-05, "loss": 4.8856, "step": 520 }, { "epoch": 0.02, "grad_norm": 2.4353551864624023, "learning_rate": 9.996118745928373e-05, "loss": 4.7092, "step": 525 }, { "epoch": 0.02, "grad_norm": 2.119088888168335, "learning_rate": 9.996026897273024e-05, "loss": 4.7475, "step": 530 }, { "epoch": 0.02, "grad_norm": 2.2172281742095947, "learning_rate": 9.995933974936922e-05, "loss": 4.8118, "step": 535 }, { "epoch": 0.02, "grad_norm": 2.8021445274353027, "learning_rate": 9.995839978940036e-05, "loss": 4.7787, "step": 540 }, { "epoch": 0.02, "grad_norm": 2.1373543739318848, "learning_rate": 9.995744909302567e-05, "loss": 4.7965, "step": 545 }, { "epoch": 0.02, "grad_norm": 2.0478668212890625, "learning_rate": 9.995648766044947e-05, "loss": 4.6403, "step": 550 }, { "epoch": 0.02, "grad_norm": 3.0331385135650635, "learning_rate": 9.995551549187836e-05, "loss": 4.7013, "step": 555 }, { "epoch": 0.02, "grad_norm": 3.018932342529297, "learning_rate": 9.995453258752127e-05, "loss": 4.6625, "step": 560 }, { "epoch": 0.02, "grad_norm": 2.7461764812469482, "learning_rate": 9.995353894758945e-05, "loss": 4.7154, "step": 565 }, { "epoch": 0.02, "grad_norm": 2.719439744949341, "learning_rate": 9.995253457229643e-05, "loss": 4.7638, "step": 570 }, { "epoch": 0.02, "grad_norm": 2.4424612522125244, "learning_rate": 9.995151946185804e-05, "loss": 4.6799, "step": 575 }, { "epoch": 0.02, "grad_norm": 3.1325790882110596, "learning_rate": 9.995049361649247e-05, "loss": 4.6011, "step": 580 }, { "epoch": 0.02, "grad_norm": 2.51981258392334, "learning_rate": 9.994945703642013e-05, "loss": 4.6994, "step": 585 }, { "epoch": 0.02, "grad_norm": 2.1592016220092773, "learning_rate": 9.994840972186386e-05, "loss": 4.6314, "step": 590 }, { "epoch": 0.02, "grad_norm": 2.0830729007720947, "learning_rate": 9.994735167304866e-05, "loss": 4.6222, "step": 595 }, { "epoch": 0.02, "grad_norm": 2.048530340194702, "learning_rate": 9.994628289020195e-05, "loss": 4.6577, "step": 600 }, { "epoch": 0.02, "grad_norm": 2.5059728622436523, "learning_rate": 9.994520337355344e-05, "loss": 4.5979, "step": 605 }, { "epoch": 0.02, "grad_norm": 2.245636224746704, "learning_rate": 9.994411312333507e-05, "loss": 4.5582, "step": 610 }, { "epoch": 0.02, "grad_norm": 2.2596888542175293, "learning_rate": 9.99430121397812e-05, "loss": 4.6809, "step": 615 }, { "epoch": 0.02, "grad_norm": 2.633591651916504, "learning_rate": 9.994190042312839e-05, "loss": 4.6444, "step": 620 }, { "epoch": 0.02, "grad_norm": 2.562018871307373, "learning_rate": 9.994077797361559e-05, "loss": 4.5521, "step": 625 }, { "epoch": 0.02, "grad_norm": 3.829125165939331, "learning_rate": 9.9939644791484e-05, "loss": 4.6366, "step": 630 }, { "epoch": 0.02, "grad_norm": 3.4246346950531006, "learning_rate": 9.993850087697716e-05, "loss": 4.6108, "step": 635 }, { "epoch": 0.02, "grad_norm": 2.107767343521118, "learning_rate": 9.99373462303409e-05, "loss": 4.6407, "step": 640 }, { "epoch": 0.02, "grad_norm": 1.9003762006759644, "learning_rate": 9.993618085182337e-05, "loss": 4.4926, "step": 645 }, { "epoch": 0.02, "grad_norm": 2.0748512744903564, "learning_rate": 9.993500474167501e-05, "loss": 4.4894, "step": 650 }, { "epoch": 0.02, "grad_norm": 2.4961984157562256, "learning_rate": 9.99338179001486e-05, "loss": 4.4764, "step": 655 }, { "epoch": 0.02, "grad_norm": 1.9952843189239502, "learning_rate": 9.993262032749914e-05, "loss": 4.5934, "step": 660 }, { "epoch": 0.02, "grad_norm": 2.1803219318389893, "learning_rate": 9.993141202398407e-05, "loss": 4.5009, "step": 665 }, { "epoch": 0.02, "grad_norm": 2.8017783164978027, "learning_rate": 9.993019298986301e-05, "loss": 4.634, "step": 670 }, { "epoch": 0.02, "grad_norm": 2.9518401622772217, "learning_rate": 9.992896322539796e-05, "loss": 4.6488, "step": 675 }, { "epoch": 0.02, "grad_norm": 2.1056196689605713, "learning_rate": 9.992772273085319e-05, "loss": 4.5225, "step": 680 }, { "epoch": 0.02, "grad_norm": 2.0276081562042236, "learning_rate": 9.99264715064953e-05, "loss": 4.4888, "step": 685 }, { "epoch": 0.02, "grad_norm": 2.4653189182281494, "learning_rate": 9.992520955259319e-05, "loss": 4.4529, "step": 690 }, { "epoch": 0.02, "grad_norm": 2.1601784229278564, "learning_rate": 9.992393686941808e-05, "loss": 4.4694, "step": 695 }, { "epoch": 0.02, "grad_norm": 1.974210262298584, "learning_rate": 9.992265345724344e-05, "loss": 4.5081, "step": 700 }, { "epoch": 0.02, "grad_norm": 2.7300446033477783, "learning_rate": 9.99213593163451e-05, "loss": 4.5084, "step": 705 }, { "epoch": 0.02, "grad_norm": 2.2425975799560547, "learning_rate": 9.992005444700121e-05, "loss": 4.4816, "step": 710 }, { "epoch": 0.02, "grad_norm": 2.3210794925689697, "learning_rate": 9.991873884949214e-05, "loss": 4.6199, "step": 715 }, { "epoch": 0.02, "grad_norm": 1.6861833333969116, "learning_rate": 9.991741252410067e-05, "loss": 4.4707, "step": 720 }, { "epoch": 0.02, "grad_norm": 1.6325076818466187, "learning_rate": 9.99160754711118e-05, "loss": 4.4422, "step": 725 }, { "epoch": 0.02, "grad_norm": 1.855193853378296, "learning_rate": 9.99147276908129e-05, "loss": 4.2949, "step": 730 }, { "epoch": 0.02, "grad_norm": 1.7147743701934814, "learning_rate": 9.991336918349361e-05, "loss": 4.4258, "step": 735 }, { "epoch": 0.02, "grad_norm": 1.6372383832931519, "learning_rate": 9.991199994944586e-05, "loss": 4.5394, "step": 740 }, { "epoch": 0.02, "grad_norm": 1.7629388570785522, "learning_rate": 9.991061998896393e-05, "loss": 4.5262, "step": 745 }, { "epoch": 0.02, "grad_norm": 1.6172864437103271, "learning_rate": 9.99092293023444e-05, "loss": 4.4303, "step": 750 }, { "epoch": 0.02, "grad_norm": 1.5990750789642334, "learning_rate": 9.990782788988611e-05, "loss": 4.4771, "step": 755 }, { "epoch": 0.02, "grad_norm": 1.6229430437088013, "learning_rate": 9.990641575189022e-05, "loss": 4.3809, "step": 760 }, { "epoch": 0.02, "grad_norm": 3.014019250869751, "learning_rate": 9.990499288866024e-05, "loss": 4.4121, "step": 765 }, { "epoch": 0.02, "grad_norm": 1.9543707370758057, "learning_rate": 9.990355930050194e-05, "loss": 4.297, "step": 770 }, { "epoch": 0.02, "grad_norm": 1.8294355869293213, "learning_rate": 9.990211498772341e-05, "loss": 4.4067, "step": 775 }, { "epoch": 0.02, "grad_norm": 1.819288969039917, "learning_rate": 9.990065995063506e-05, "loss": 4.3356, "step": 780 }, { "epoch": 0.02, "grad_norm": 1.575514316558838, "learning_rate": 9.989919418954956e-05, "loss": 4.3854, "step": 785 }, { "epoch": 0.02, "grad_norm": 1.6439075469970703, "learning_rate": 9.989771770478191e-05, "loss": 4.4376, "step": 790 }, { "epoch": 0.02, "grad_norm": 1.3715301752090454, "learning_rate": 9.989623049664944e-05, "loss": 4.353, "step": 795 }, { "epoch": 0.02, "grad_norm": 1.5874420404434204, "learning_rate": 9.989473256547175e-05, "loss": 4.3793, "step": 800 }, { "epoch": 0.02, "grad_norm": 1.5349891185760498, "learning_rate": 9.989322391157076e-05, "loss": 4.4114, "step": 805 }, { "epoch": 0.02, "grad_norm": 1.7908573150634766, "learning_rate": 9.989170453527068e-05, "loss": 4.3693, "step": 810 }, { "epoch": 0.02, "grad_norm": 1.5562800168991089, "learning_rate": 9.989017443689805e-05, "loss": 4.373, "step": 815 }, { "epoch": 0.02, "grad_norm": 1.8784722089767456, "learning_rate": 9.988863361678169e-05, "loss": 4.2695, "step": 820 }, { "epoch": 0.02, "grad_norm": 1.671901822090149, "learning_rate": 9.988708207525272e-05, "loss": 4.398, "step": 825 }, { "epoch": 0.02, "grad_norm": 1.609654426574707, "learning_rate": 9.98855198126446e-05, "loss": 4.3956, "step": 830 }, { "epoch": 0.02, "grad_norm": 1.6533249616622925, "learning_rate": 9.988394682929305e-05, "loss": 4.4109, "step": 835 }, { "epoch": 0.02, "grad_norm": 2.0188634395599365, "learning_rate": 9.988236312553614e-05, "loss": 4.3442, "step": 840 }, { "epoch": 0.02, "grad_norm": 2.0584287643432617, "learning_rate": 9.988076870171419e-05, "loss": 4.3139, "step": 845 }, { "epoch": 0.03, "grad_norm": 3.633040189743042, "learning_rate": 9.987916355816988e-05, "loss": 4.3409, "step": 850 }, { "epoch": 0.03, "grad_norm": 1.5486620664596558, "learning_rate": 9.987754769524814e-05, "loss": 4.4065, "step": 855 }, { "epoch": 0.03, "grad_norm": 1.6065951585769653, "learning_rate": 9.987592111329626e-05, "loss": 4.3377, "step": 860 }, { "epoch": 0.03, "grad_norm": 1.597908616065979, "learning_rate": 9.987428381266377e-05, "loss": 4.2969, "step": 865 }, { "epoch": 0.03, "grad_norm": 1.5601239204406738, "learning_rate": 9.987263579370257e-05, "loss": 4.2445, "step": 870 }, { "epoch": 0.03, "grad_norm": 2.14276385307312, "learning_rate": 9.987097705676681e-05, "loss": 4.3588, "step": 875 }, { "epoch": 0.03, "grad_norm": 1.5181288719177246, "learning_rate": 9.986930760221296e-05, "loss": 4.3168, "step": 880 }, { "epoch": 0.03, "grad_norm": 1.8659099340438843, "learning_rate": 9.986762743039983e-05, "loss": 4.3861, "step": 885 }, { "epoch": 0.03, "grad_norm": 1.551437258720398, "learning_rate": 9.986593654168846e-05, "loss": 4.2599, "step": 890 }, { "epoch": 0.03, "grad_norm": 1.6148712635040283, "learning_rate": 9.986423493644225e-05, "loss": 4.2671, "step": 895 }, { "epoch": 0.03, "grad_norm": 1.6241729259490967, "learning_rate": 9.98625226150269e-05, "loss": 4.2183, "step": 900 }, { "epoch": 0.03, "grad_norm": 1.511135220527649, "learning_rate": 9.986079957781038e-05, "loss": 4.2817, "step": 905 }, { "epoch": 0.03, "grad_norm": 1.3672266006469727, "learning_rate": 9.9859065825163e-05, "loss": 4.3296, "step": 910 }, { "epoch": 0.03, "grad_norm": 1.6353356838226318, "learning_rate": 9.985732135745733e-05, "loss": 4.3095, "step": 915 }, { "epoch": 0.03, "grad_norm": 1.289083480834961, "learning_rate": 9.985556617506828e-05, "loss": 4.1558, "step": 920 }, { "epoch": 0.03, "grad_norm": 1.6122968196868896, "learning_rate": 9.985380027837308e-05, "loss": 4.3289, "step": 925 }, { "epoch": 0.03, "grad_norm": 1.7070633172988892, "learning_rate": 9.98520236677512e-05, "loss": 4.1968, "step": 930 }, { "epoch": 0.03, "grad_norm": 1.2219825983047485, "learning_rate": 9.985023634358443e-05, "loss": 4.2167, "step": 935 }, { "epoch": 0.03, "grad_norm": 1.2288941144943237, "learning_rate": 9.984843830625691e-05, "loss": 4.2001, "step": 940 }, { "epoch": 0.03, "grad_norm": 1.4346648454666138, "learning_rate": 9.984662955615507e-05, "loss": 4.3222, "step": 945 }, { "epoch": 0.03, "grad_norm": 1.2870702743530273, "learning_rate": 9.984481009366758e-05, "loss": 4.2046, "step": 950 }, { "epoch": 0.03, "grad_norm": 1.3178324699401855, "learning_rate": 9.984297991918548e-05, "loss": 4.2475, "step": 955 }, { "epoch": 0.03, "grad_norm": 1.2238595485687256, "learning_rate": 9.984113903310206e-05, "loss": 4.2774, "step": 960 }, { "epoch": 0.03, "grad_norm": 1.6214929819107056, "learning_rate": 9.983928743581296e-05, "loss": 4.2287, "step": 965 }, { "epoch": 0.03, "grad_norm": 1.9230448007583618, "learning_rate": 9.983742512771611e-05, "loss": 4.2831, "step": 970 }, { "epoch": 0.03, "grad_norm": 1.8009263277053833, "learning_rate": 9.983555210921171e-05, "loss": 4.2093, "step": 975 }, { "epoch": 0.03, "grad_norm": 1.610308289527893, "learning_rate": 9.983366838070231e-05, "loss": 4.1312, "step": 980 }, { "epoch": 0.03, "grad_norm": 1.35236656665802, "learning_rate": 9.983177394259271e-05, "loss": 4.1411, "step": 985 }, { "epoch": 0.03, "grad_norm": 1.4383875131607056, "learning_rate": 9.982986879529006e-05, "loss": 4.2745, "step": 990 }, { "epoch": 0.03, "grad_norm": 1.357028603553772, "learning_rate": 9.982795293920377e-05, "loss": 4.1077, "step": 995 }, { "epoch": 0.03, "grad_norm": 1.6876486539840698, "learning_rate": 9.982602637474558e-05, "loss": 4.2628, "step": 1000 }, { "epoch": 0.03, "grad_norm": 1.2707496881484985, "learning_rate": 9.982408910232952e-05, "loss": 4.2908, "step": 1005 }, { "epoch": 0.03, "grad_norm": 1.766549825668335, "learning_rate": 9.982214112237193e-05, "loss": 4.2629, "step": 1010 }, { "epoch": 0.03, "grad_norm": 1.3197702169418335, "learning_rate": 9.982018243529146e-05, "loss": 4.1586, "step": 1015 }, { "epoch": 0.03, "grad_norm": 1.725451946258545, "learning_rate": 9.981821304150901e-05, "loss": 4.2295, "step": 1020 }, { "epoch": 0.03, "grad_norm": 1.5005009174346924, "learning_rate": 9.981623294144782e-05, "loss": 4.1455, "step": 1025 }, { "epoch": 0.03, "grad_norm": 1.4445676803588867, "learning_rate": 9.981424213553344e-05, "loss": 4.2151, "step": 1030 }, { "epoch": 0.03, "grad_norm": 1.2467132806777954, "learning_rate": 9.981224062419372e-05, "loss": 4.0965, "step": 1035 }, { "epoch": 0.03, "grad_norm": 1.387924075126648, "learning_rate": 9.981022840785877e-05, "loss": 4.254, "step": 1040 }, { "epoch": 0.03, "grad_norm": 1.3359142541885376, "learning_rate": 9.980820548696106e-05, "loss": 4.127, "step": 1045 }, { "epoch": 0.03, "grad_norm": 1.3878377676010132, "learning_rate": 9.98061718619353e-05, "loss": 4.0996, "step": 1050 }, { "epoch": 0.03, "grad_norm": 1.213212013244629, "learning_rate": 9.980412753321856e-05, "loss": 4.2301, "step": 1055 }, { "epoch": 0.03, "grad_norm": 1.3506423234939575, "learning_rate": 9.980207250125015e-05, "loss": 4.1602, "step": 1060 }, { "epoch": 0.03, "grad_norm": 1.2402514219284058, "learning_rate": 9.980000676647174e-05, "loss": 4.2488, "step": 1065 }, { "epoch": 0.03, "grad_norm": 1.0766572952270508, "learning_rate": 9.979793032932725e-05, "loss": 4.1083, "step": 1070 }, { "epoch": 0.03, "grad_norm": 1.2045143842697144, "learning_rate": 9.979584319026292e-05, "loss": 4.0277, "step": 1075 }, { "epoch": 0.03, "grad_norm": 1.2845330238342285, "learning_rate": 9.97937453497273e-05, "loss": 4.2403, "step": 1080 }, { "epoch": 0.03, "grad_norm": 1.2786062955856323, "learning_rate": 9.979163680817124e-05, "loss": 4.1247, "step": 1085 }, { "epoch": 0.03, "grad_norm": 1.1196534633636475, "learning_rate": 9.978951756604786e-05, "loss": 4.1587, "step": 1090 }, { "epoch": 0.03, "grad_norm": 1.4554393291473389, "learning_rate": 9.978738762381262e-05, "loss": 4.1086, "step": 1095 }, { "epoch": 0.03, "grad_norm": 1.421023964881897, "learning_rate": 9.978524698192322e-05, "loss": 4.1772, "step": 1100 }, { "epoch": 0.03, "grad_norm": 1.163741111755371, "learning_rate": 9.978309564083975e-05, "loss": 4.1544, "step": 1105 }, { "epoch": 0.03, "grad_norm": 1.2630544900894165, "learning_rate": 9.97809336010245e-05, "loss": 4.2158, "step": 1110 }, { "epoch": 0.03, "grad_norm": 1.0878833532333374, "learning_rate": 9.977876086294216e-05, "loss": 4.141, "step": 1115 }, { "epoch": 0.03, "grad_norm": 1.014878273010254, "learning_rate": 9.977657742705962e-05, "loss": 4.0646, "step": 1120 }, { "epoch": 0.03, "grad_norm": 1.4150738716125488, "learning_rate": 9.977438329384614e-05, "loss": 4.0066, "step": 1125 }, { "epoch": 0.03, "grad_norm": 1.2574440240859985, "learning_rate": 9.977217846377324e-05, "loss": 4.0581, "step": 1130 }, { "epoch": 0.03, "grad_norm": 1.117933988571167, "learning_rate": 9.976996293731476e-05, "loss": 4.0028, "step": 1135 }, { "epoch": 0.03, "grad_norm": 1.0256320238113403, "learning_rate": 9.976773671494684e-05, "loss": 4.0564, "step": 1140 }, { "epoch": 0.03, "grad_norm": 1.0857431888580322, "learning_rate": 9.976549979714791e-05, "loss": 4.1129, "step": 1145 }, { "epoch": 0.03, "grad_norm": 1.1032960414886475, "learning_rate": 9.976325218439868e-05, "loss": 4.0331, "step": 1150 }, { "epoch": 0.03, "grad_norm": 1.1655795574188232, "learning_rate": 9.976099387718218e-05, "loss": 4.0366, "step": 1155 }, { "epoch": 0.03, "grad_norm": 1.1679975986480713, "learning_rate": 9.975872487598377e-05, "loss": 4.0897, "step": 1160 }, { "epoch": 0.03, "grad_norm": 0.9969255328178406, "learning_rate": 9.975644518129104e-05, "loss": 4.0432, "step": 1165 }, { "epoch": 0.03, "grad_norm": 1.207139253616333, "learning_rate": 9.975415479359394e-05, "loss": 4.0284, "step": 1170 }, { "epoch": 0.03, "grad_norm": 1.0415829420089722, "learning_rate": 9.975185371338464e-05, "loss": 4.1151, "step": 1175 }, { "epoch": 0.03, "grad_norm": 1.1046624183654785, "learning_rate": 9.97495419411577e-05, "loss": 4.0904, "step": 1180 }, { "epoch": 0.03, "grad_norm": 1.2057961225509644, "learning_rate": 9.974721947740995e-05, "loss": 4.0402, "step": 1185 }, { "epoch": 0.04, "grad_norm": 0.9968929886817932, "learning_rate": 9.974488632264046e-05, "loss": 4.0314, "step": 1190 }, { "epoch": 0.04, "grad_norm": 1.0434792041778564, "learning_rate": 9.974254247735066e-05, "loss": 4.0828, "step": 1195 }, { "epoch": 0.04, "grad_norm": 1.128531575202942, "learning_rate": 9.974018794204426e-05, "loss": 3.9792, "step": 1200 }, { "epoch": 0.04, "grad_norm": 1.1299066543579102, "learning_rate": 9.973782271722727e-05, "loss": 3.9566, "step": 1205 }, { "epoch": 0.04, "grad_norm": 1.3382987976074219, "learning_rate": 9.973544680340799e-05, "loss": 4.0369, "step": 1210 }, { "epoch": 0.04, "grad_norm": 1.2568211555480957, "learning_rate": 9.973306020109701e-05, "loss": 3.9994, "step": 1215 }, { "epoch": 0.04, "grad_norm": 1.229506492614746, "learning_rate": 9.973066291080724e-05, "loss": 3.9346, "step": 1220 }, { "epoch": 0.04, "grad_norm": 1.1474674940109253, "learning_rate": 9.972825493305387e-05, "loss": 4.0818, "step": 1225 }, { "epoch": 0.04, "grad_norm": 1.087825894355774, "learning_rate": 9.97258362683544e-05, "loss": 4.0879, "step": 1230 }, { "epoch": 0.04, "grad_norm": 1.0302047729492188, "learning_rate": 9.972340691722858e-05, "loss": 4.0404, "step": 1235 }, { "epoch": 0.04, "grad_norm": 1.020920991897583, "learning_rate": 9.972096688019856e-05, "loss": 4.0275, "step": 1240 }, { "epoch": 0.04, "grad_norm": 1.1431856155395508, "learning_rate": 9.971851615778868e-05, "loss": 3.9545, "step": 1245 }, { "epoch": 0.04, "grad_norm": 1.059474229812622, "learning_rate": 9.971605475052561e-05, "loss": 3.9888, "step": 1250 }, { "epoch": 0.04, "grad_norm": 1.101056456565857, "learning_rate": 9.971358265893834e-05, "loss": 3.8715, "step": 1255 }, { "epoch": 0.04, "grad_norm": 1.0579644441604614, "learning_rate": 9.971109988355814e-05, "loss": 4.0248, "step": 1260 }, { "epoch": 0.04, "grad_norm": 1.0268296003341675, "learning_rate": 9.970860642491854e-05, "loss": 4.023, "step": 1265 }, { "epoch": 0.04, "grad_norm": 1.031416416168213, "learning_rate": 9.970610228355548e-05, "loss": 3.9692, "step": 1270 }, { "epoch": 0.04, "grad_norm": 1.1259690523147583, "learning_rate": 9.970358746000704e-05, "loss": 3.9095, "step": 1275 }, { "epoch": 0.04, "grad_norm": 1.0979198217391968, "learning_rate": 9.97010619548137e-05, "loss": 3.9574, "step": 1280 }, { "epoch": 0.04, "grad_norm": 1.0067107677459717, "learning_rate": 9.969852576851822e-05, "loss": 3.9153, "step": 1285 }, { "epoch": 0.04, "grad_norm": 1.2386833429336548, "learning_rate": 9.969597890166565e-05, "loss": 4.0187, "step": 1290 }, { "epoch": 0.04, "grad_norm": 1.0346835851669312, "learning_rate": 9.969342135480328e-05, "loss": 3.978, "step": 1295 }, { "epoch": 0.04, "grad_norm": 0.9244931936264038, "learning_rate": 9.969085312848081e-05, "loss": 3.8581, "step": 1300 }, { "epoch": 0.04, "grad_norm": 1.132784128189087, "learning_rate": 9.968827422325012e-05, "loss": 4.0105, "step": 1305 }, { "epoch": 0.04, "grad_norm": 1.025078535079956, "learning_rate": 9.968568463966545e-05, "loss": 3.8805, "step": 1310 }, { "epoch": 0.04, "grad_norm": 1.0972689390182495, "learning_rate": 9.968308437828333e-05, "loss": 3.9369, "step": 1315 }, { "epoch": 0.04, "grad_norm": 1.0510038137435913, "learning_rate": 9.968047343966256e-05, "loss": 3.965, "step": 1320 }, { "epoch": 0.04, "grad_norm": 1.0470211505889893, "learning_rate": 9.967785182436426e-05, "loss": 3.8825, "step": 1325 }, { "epoch": 0.04, "grad_norm": 1.0741597414016724, "learning_rate": 9.967521953295185e-05, "loss": 3.9229, "step": 1330 }, { "epoch": 0.04, "grad_norm": 2.712617874145508, "learning_rate": 9.967257656599099e-05, "loss": 4.0451, "step": 1335 }, { "epoch": 0.04, "grad_norm": 1.095920205116272, "learning_rate": 9.966992292404967e-05, "loss": 3.8371, "step": 1340 }, { "epoch": 0.04, "grad_norm": 1.0120892524719238, "learning_rate": 9.966725860769821e-05, "loss": 4.0108, "step": 1345 }, { "epoch": 0.04, "grad_norm": 1.257179617881775, "learning_rate": 9.966458361750916e-05, "loss": 3.8193, "step": 1350 }, { "epoch": 0.04, "grad_norm": 1.1157857179641724, "learning_rate": 9.966189795405743e-05, "loss": 3.829, "step": 1355 }, { "epoch": 0.04, "grad_norm": 0.970812201499939, "learning_rate": 9.965920161792016e-05, "loss": 3.8401, "step": 1360 }, { "epoch": 0.04, "grad_norm": 0.954301118850708, "learning_rate": 9.96564946096768e-05, "loss": 3.8101, "step": 1365 }, { "epoch": 0.04, "grad_norm": 0.88419508934021, "learning_rate": 9.965377692990914e-05, "loss": 3.8407, "step": 1370 }, { "epoch": 0.04, "grad_norm": 1.128898024559021, "learning_rate": 9.965104857920121e-05, "loss": 3.8151, "step": 1375 }, { "epoch": 0.04, "grad_norm": 1.091166615486145, "learning_rate": 9.964830955813935e-05, "loss": 3.8387, "step": 1380 }, { "epoch": 0.04, "grad_norm": 0.9677389860153198, "learning_rate": 9.964555986731219e-05, "loss": 3.8692, "step": 1385 }, { "epoch": 0.04, "grad_norm": 1.0396766662597656, "learning_rate": 9.964279950731066e-05, "loss": 3.844, "step": 1390 }, { "epoch": 0.04, "grad_norm": 0.9413672685623169, "learning_rate": 9.9640028478728e-05, "loss": 3.7964, "step": 1395 }, { "epoch": 0.04, "grad_norm": 0.9077847003936768, "learning_rate": 9.963724678215968e-05, "loss": 3.8673, "step": 1400 }, { "epoch": 0.04, "grad_norm": 1.1119506359100342, "learning_rate": 9.963445441820356e-05, "loss": 3.8552, "step": 1405 }, { "epoch": 0.04, "grad_norm": 1.0764799118041992, "learning_rate": 9.963165138745969e-05, "loss": 3.7871, "step": 1410 }, { "epoch": 0.04, "grad_norm": 0.9036362767219543, "learning_rate": 9.962883769053051e-05, "loss": 3.8696, "step": 1415 }, { "epoch": 0.04, "grad_norm": 0.9879042506217957, "learning_rate": 9.962601332802065e-05, "loss": 3.8011, "step": 1420 }, { "epoch": 0.04, "grad_norm": 0.9978138208389282, "learning_rate": 9.962317830053711e-05, "loss": 3.8722, "step": 1425 }, { "epoch": 0.04, "grad_norm": 0.9543709754943848, "learning_rate": 9.962033260868916e-05, "loss": 3.868, "step": 1430 }, { "epoch": 0.04, "grad_norm": 0.9377373456954956, "learning_rate": 9.961747625308838e-05, "loss": 3.8539, "step": 1435 }, { "epoch": 0.04, "grad_norm": 0.9829211831092834, "learning_rate": 9.961460923434858e-05, "loss": 3.8096, "step": 1440 }, { "epoch": 0.04, "grad_norm": 0.8982797861099243, "learning_rate": 9.961173155308591e-05, "loss": 3.7751, "step": 1445 }, { "epoch": 0.04, "grad_norm": 0.9109345078468323, "learning_rate": 9.960884320991881e-05, "loss": 3.918, "step": 1450 }, { "epoch": 0.04, "grad_norm": 1.1583112478256226, "learning_rate": 9.960594420546802e-05, "loss": 3.8505, "step": 1455 }, { "epoch": 0.04, "grad_norm": 1.2373203039169312, "learning_rate": 9.960303454035654e-05, "loss": 3.8644, "step": 1460 }, { "epoch": 0.04, "grad_norm": 0.8303922414779663, "learning_rate": 9.960011421520969e-05, "loss": 3.8997, "step": 1465 }, { "epoch": 0.04, "grad_norm": 0.9642992615699768, "learning_rate": 9.959718323065503e-05, "loss": 3.8469, "step": 1470 }, { "epoch": 0.04, "grad_norm": 0.8670044541358948, "learning_rate": 9.959424158732252e-05, "loss": 3.8587, "step": 1475 }, { "epoch": 0.04, "grad_norm": 0.9006289839744568, "learning_rate": 9.959128928584426e-05, "loss": 3.8379, "step": 1480 }, { "epoch": 0.04, "grad_norm": 0.8690353631973267, "learning_rate": 9.958832632685478e-05, "loss": 3.7249, "step": 1485 }, { "epoch": 0.04, "grad_norm": 0.8903810977935791, "learning_rate": 9.95853527109908e-05, "loss": 3.7599, "step": 1490 }, { "epoch": 0.04, "grad_norm": 0.8518982529640198, "learning_rate": 9.95823684388914e-05, "loss": 3.8433, "step": 1495 }, { "epoch": 0.04, "grad_norm": 0.861099898815155, "learning_rate": 9.95793735111979e-05, "loss": 3.7059, "step": 1500 }, { "epoch": 0.04, "grad_norm": 0.9849363565444946, "learning_rate": 9.957636792855394e-05, "loss": 3.6851, "step": 1505 }, { "epoch": 0.04, "grad_norm": 0.9486384987831116, "learning_rate": 9.957335169160545e-05, "loss": 3.6519, "step": 1510 }, { "epoch": 0.04, "grad_norm": 1.0623517036437988, "learning_rate": 9.957032480100061e-05, "loss": 3.7927, "step": 1515 }, { "epoch": 0.04, "grad_norm": 0.8179353475570679, "learning_rate": 9.956728725738996e-05, "loss": 3.6934, "step": 1520 }, { "epoch": 0.04, "grad_norm": 0.8650668859481812, "learning_rate": 9.956423906142626e-05, "loss": 3.7657, "step": 1525 }, { "epoch": 0.05, "grad_norm": 0.8139018416404724, "learning_rate": 9.956118021376461e-05, "loss": 3.7672, "step": 1530 }, { "epoch": 0.05, "grad_norm": 0.8629039525985718, "learning_rate": 9.955811071506236e-05, "loss": 3.7398, "step": 1535 }, { "epoch": 0.05, "grad_norm": 0.8906451463699341, "learning_rate": 9.955503056597917e-05, "loss": 3.678, "step": 1540 }, { "epoch": 0.05, "grad_norm": 0.8075469136238098, "learning_rate": 9.955193976717698e-05, "loss": 3.7796, "step": 1545 }, { "epoch": 0.05, "grad_norm": 0.9265354871749878, "learning_rate": 9.954883831932003e-05, "loss": 3.7281, "step": 1550 }, { "epoch": 0.05, "grad_norm": 0.883394718170166, "learning_rate": 9.954572622307485e-05, "loss": 3.7534, "step": 1555 }, { "epoch": 0.05, "grad_norm": 0.8667304515838623, "learning_rate": 9.954260347911025e-05, "loss": 3.7888, "step": 1560 }, { "epoch": 0.05, "grad_norm": 0.8533546924591064, "learning_rate": 9.953947008809731e-05, "loss": 3.8181, "step": 1565 }, { "epoch": 0.05, "grad_norm": 0.8444490432739258, "learning_rate": 9.953632605070944e-05, "loss": 3.7513, "step": 1570 }, { "epoch": 0.05, "grad_norm": 0.8287249803543091, "learning_rate": 9.95331713676223e-05, "loss": 3.7853, "step": 1575 }, { "epoch": 0.05, "grad_norm": 1.1331722736358643, "learning_rate": 9.953000603951386e-05, "loss": 3.6451, "step": 1580 }, { "epoch": 0.05, "grad_norm": 1.0632293224334717, "learning_rate": 9.952683006706438e-05, "loss": 3.6961, "step": 1585 }, { "epoch": 0.05, "grad_norm": 0.8901936411857605, "learning_rate": 9.952364345095638e-05, "loss": 3.7488, "step": 1590 }, { "epoch": 0.05, "grad_norm": 1.0943388938903809, "learning_rate": 9.95204461918747e-05, "loss": 3.7775, "step": 1595 }, { "epoch": 0.05, "grad_norm": 0.8000125288963318, "learning_rate": 9.951723829050646e-05, "loss": 3.746, "step": 1600 }, { "epoch": 0.05, "grad_norm": 0.970822274684906, "learning_rate": 9.951401974754101e-05, "loss": 3.6389, "step": 1605 }, { "epoch": 0.05, "grad_norm": 0.932509183883667, "learning_rate": 9.951079056367013e-05, "loss": 3.7031, "step": 1610 }, { "epoch": 0.05, "grad_norm": 0.9065966010093689, "learning_rate": 9.95075507395877e-05, "loss": 3.7625, "step": 1615 }, { "epoch": 0.05, "grad_norm": 0.8580222725868225, "learning_rate": 9.950430027599003e-05, "loss": 3.6984, "step": 1620 }, { "epoch": 0.05, "grad_norm": 0.8689137101173401, "learning_rate": 9.950103917357566e-05, "loss": 3.6544, "step": 1625 }, { "epoch": 0.05, "grad_norm": 0.8698611855506897, "learning_rate": 9.949776743304542e-05, "loss": 3.6441, "step": 1630 }, { "epoch": 0.05, "grad_norm": 0.8408386707305908, "learning_rate": 9.949448505510243e-05, "loss": 3.7137, "step": 1635 }, { "epoch": 0.05, "grad_norm": 1.0276784896850586, "learning_rate": 9.94911920404521e-05, "loss": 3.667, "step": 1640 }, { "epoch": 0.05, "grad_norm": 0.8747833371162415, "learning_rate": 9.94878883898021e-05, "loss": 3.7531, "step": 1645 }, { "epoch": 0.05, "grad_norm": 0.8138440847396851, "learning_rate": 9.948457410386244e-05, "loss": 3.642, "step": 1650 }, { "epoch": 0.05, "grad_norm": 0.8010612726211548, "learning_rate": 9.948124918334538e-05, "loss": 3.7169, "step": 1655 }, { "epoch": 0.05, "grad_norm": 0.8289493322372437, "learning_rate": 9.947791362896543e-05, "loss": 3.777, "step": 1660 }, { "epoch": 0.05, "grad_norm": 0.8454409837722778, "learning_rate": 9.947456744143947e-05, "loss": 3.5626, "step": 1665 }, { "epoch": 0.05, "grad_norm": 0.8723094463348389, "learning_rate": 9.947121062148658e-05, "loss": 3.6989, "step": 1670 }, { "epoch": 0.05, "grad_norm": 0.8869099020957947, "learning_rate": 9.94678431698282e-05, "loss": 3.7136, "step": 1675 }, { "epoch": 0.05, "grad_norm": 0.8179925084114075, "learning_rate": 9.946446508718801e-05, "loss": 3.617, "step": 1680 }, { "epoch": 0.05, "grad_norm": 0.8121179938316345, "learning_rate": 9.946107637429196e-05, "loss": 3.6315, "step": 1685 }, { "epoch": 0.05, "grad_norm": 0.9434096217155457, "learning_rate": 9.945767703186832e-05, "loss": 3.636, "step": 1690 }, { "epoch": 0.05, "grad_norm": 0.8432795405387878, "learning_rate": 9.945426706064764e-05, "loss": 3.6857, "step": 1695 }, { "epoch": 0.05, "grad_norm": 4.524977207183838, "learning_rate": 9.945084646136275e-05, "loss": 3.573, "step": 1700 }, { "epoch": 0.05, "grad_norm": 0.8250172734260559, "learning_rate": 9.944741523474874e-05, "loss": 3.5982, "step": 1705 }, { "epoch": 0.05, "grad_norm": 0.7767294645309448, "learning_rate": 9.944397338154301e-05, "loss": 3.5719, "step": 1710 }, { "epoch": 0.05, "grad_norm": 0.8832421898841858, "learning_rate": 9.944052090248526e-05, "loss": 3.6062, "step": 1715 }, { "epoch": 0.05, "grad_norm": 0.8160373568534851, "learning_rate": 9.943705779831742e-05, "loss": 3.7329, "step": 1720 }, { "epoch": 0.05, "grad_norm": 0.7346221208572388, "learning_rate": 9.943358406978377e-05, "loss": 3.6443, "step": 1725 }, { "epoch": 0.05, "grad_norm": 0.8048697113990784, "learning_rate": 9.943009971763082e-05, "loss": 3.6307, "step": 1730 }, { "epoch": 0.05, "grad_norm": 0.8407803177833557, "learning_rate": 9.942660474260737e-05, "loss": 3.6248, "step": 1735 }, { "epoch": 0.05, "grad_norm": 0.9443001747131348, "learning_rate": 9.942309914546453e-05, "loss": 3.5912, "step": 1740 }, { "epoch": 0.05, "grad_norm": 0.8309940695762634, "learning_rate": 9.941958292695565e-05, "loss": 3.6327, "step": 1745 }, { "epoch": 0.05, "grad_norm": 0.7845703363418579, "learning_rate": 9.941605608783644e-05, "loss": 3.6408, "step": 1750 }, { "epoch": 0.05, "grad_norm": 0.735937237739563, "learning_rate": 9.94125186288648e-05, "loss": 3.6223, "step": 1755 }, { "epoch": 0.05, "grad_norm": 0.8459873199462891, "learning_rate": 9.940897055080098e-05, "loss": 3.643, "step": 1760 }, { "epoch": 0.05, "grad_norm": 0.8891588449478149, "learning_rate": 9.940541185440746e-05, "loss": 3.6811, "step": 1765 }, { "epoch": 0.05, "grad_norm": 0.8388894200325012, "learning_rate": 9.940184254044904e-05, "loss": 3.6801, "step": 1770 }, { "epoch": 0.05, "grad_norm": 0.9111271500587463, "learning_rate": 9.939826260969282e-05, "loss": 3.6182, "step": 1775 }, { "epoch": 0.05, "grad_norm": 0.9713955521583557, "learning_rate": 9.93946720629081e-05, "loss": 3.6693, "step": 1780 }, { "epoch": 0.05, "grad_norm": 0.9383721351623535, "learning_rate": 9.939107090086654e-05, "loss": 3.6122, "step": 1785 }, { "epoch": 0.05, "grad_norm": 0.8370587825775146, "learning_rate": 9.938745912434208e-05, "loss": 3.6417, "step": 1790 }, { "epoch": 0.05, "grad_norm": 0.8760790824890137, "learning_rate": 9.938383673411087e-05, "loss": 3.5442, "step": 1795 }, { "epoch": 0.05, "grad_norm": 0.7888710498809814, "learning_rate": 9.938020373095141e-05, "loss": 3.6032, "step": 1800 }, { "epoch": 0.05, "grad_norm": 0.7731709480285645, "learning_rate": 9.937656011564445e-05, "loss": 3.6378, "step": 1805 }, { "epoch": 0.05, "grad_norm": 0.7636626362800598, "learning_rate": 9.937290588897302e-05, "loss": 3.5193, "step": 1810 }, { "epoch": 0.05, "grad_norm": 0.8181933760643005, "learning_rate": 9.936924105172248e-05, "loss": 3.6707, "step": 1815 }, { "epoch": 0.05, "grad_norm": 0.7306434512138367, "learning_rate": 9.936556560468037e-05, "loss": 3.7169, "step": 1820 }, { "epoch": 0.05, "grad_norm": 0.754530668258667, "learning_rate": 9.93618795486366e-05, "loss": 3.6341, "step": 1825 }, { "epoch": 0.05, "grad_norm": 1.3126890659332275, "learning_rate": 9.935818288438334e-05, "loss": 3.6786, "step": 1830 }, { "epoch": 0.05, "grad_norm": 0.7640548944473267, "learning_rate": 9.935447561271501e-05, "loss": 3.5773, "step": 1835 }, { "epoch": 0.05, "grad_norm": 0.7683709859848022, "learning_rate": 9.935075773442835e-05, "loss": 3.5992, "step": 1840 }, { "epoch": 0.05, "grad_norm": 0.7417837381362915, "learning_rate": 9.934702925032232e-05, "loss": 3.5879, "step": 1845 }, { "epoch": 0.05, "grad_norm": 0.7797106504440308, "learning_rate": 9.934329016119824e-05, "loss": 3.5904, "step": 1850 }, { "epoch": 0.05, "grad_norm": 0.9057390689849854, "learning_rate": 9.933954046785963e-05, "loss": 3.5383, "step": 1855 }, { "epoch": 0.05, "grad_norm": 0.9141639471054077, "learning_rate": 9.933578017111233e-05, "loss": 3.6554, "step": 1860 }, { "epoch": 0.05, "grad_norm": 0.8078970909118652, "learning_rate": 9.933200927176449e-05, "loss": 3.4188, "step": 1865 }, { "epoch": 0.06, "grad_norm": 0.7535152435302734, "learning_rate": 9.932822777062646e-05, "loss": 3.5499, "step": 1870 }, { "epoch": 0.06, "grad_norm": 0.8829749822616577, "learning_rate": 9.932443566851092e-05, "loss": 3.6093, "step": 1875 }, { "epoch": 0.06, "grad_norm": 0.7857192754745483, "learning_rate": 9.932063296623284e-05, "loss": 3.624, "step": 1880 }, { "epoch": 0.06, "grad_norm": 0.700707733631134, "learning_rate": 9.931681966460944e-05, "loss": 3.546, "step": 1885 }, { "epoch": 0.06, "grad_norm": 0.7530134916305542, "learning_rate": 9.93129957644602e-05, "loss": 3.5148, "step": 1890 }, { "epoch": 0.06, "grad_norm": 0.7049728035926819, "learning_rate": 9.930916126660691e-05, "loss": 3.5535, "step": 1895 }, { "epoch": 0.06, "grad_norm": 0.7310341596603394, "learning_rate": 9.930531617187367e-05, "loss": 3.5879, "step": 1900 }, { "epoch": 0.06, "grad_norm": 0.7308007478713989, "learning_rate": 9.930146048108678e-05, "loss": 3.5512, "step": 1905 }, { "epoch": 0.06, "grad_norm": 0.7546719312667847, "learning_rate": 9.929759419507487e-05, "loss": 3.5411, "step": 1910 }, { "epoch": 0.06, "grad_norm": 0.7326837182044983, "learning_rate": 9.92937173146688e-05, "loss": 3.5791, "step": 1915 }, { "epoch": 0.06, "grad_norm": 0.7149990797042847, "learning_rate": 9.928982984070179e-05, "loss": 3.4825, "step": 1920 }, { "epoch": 0.06, "grad_norm": 0.7729883790016174, "learning_rate": 9.928593177400925e-05, "loss": 3.4977, "step": 1925 }, { "epoch": 0.06, "grad_norm": 0.8264164924621582, "learning_rate": 9.928202311542889e-05, "loss": 3.6036, "step": 1930 }, { "epoch": 0.06, "grad_norm": 0.8064064979553223, "learning_rate": 9.927810386580075e-05, "loss": 3.5402, "step": 1935 }, { "epoch": 0.06, "grad_norm": 0.732982873916626, "learning_rate": 9.927417402596707e-05, "loss": 3.596, "step": 1940 }, { "epoch": 0.06, "grad_norm": 0.795852541923523, "learning_rate": 9.927023359677241e-05, "loss": 3.5162, "step": 1945 }, { "epoch": 0.06, "grad_norm": 0.6934268474578857, "learning_rate": 9.92662825790636e-05, "loss": 3.5333, "step": 1950 }, { "epoch": 0.06, "grad_norm": 0.8366135954856873, "learning_rate": 9.926232097368974e-05, "loss": 3.5683, "step": 1955 }, { "epoch": 0.06, "grad_norm": 0.7585029006004333, "learning_rate": 9.925834878150219e-05, "loss": 3.4268, "step": 1960 }, { "epoch": 0.06, "grad_norm": 0.7670012712478638, "learning_rate": 9.925436600335464e-05, "loss": 3.503, "step": 1965 }, { "epoch": 0.06, "grad_norm": 1.521193027496338, "learning_rate": 9.925037264010298e-05, "loss": 3.5176, "step": 1970 }, { "epoch": 0.06, "grad_norm": 0.8860667943954468, "learning_rate": 9.924636869260542e-05, "loss": 3.503, "step": 1975 }, { "epoch": 0.06, "grad_norm": 0.7409327030181885, "learning_rate": 9.924235416172243e-05, "loss": 3.5686, "step": 1980 }, { "epoch": 0.06, "grad_norm": 0.7880022525787354, "learning_rate": 9.923832904831679e-05, "loss": 3.4567, "step": 1985 }, { "epoch": 0.06, "grad_norm": 0.7202027440071106, "learning_rate": 9.923429335325349e-05, "loss": 3.4427, "step": 1990 }, { "epoch": 0.06, "grad_norm": 0.8414711356163025, "learning_rate": 9.923024707739985e-05, "loss": 3.5183, "step": 1995 }, { "epoch": 0.06, "grad_norm": 0.8426960706710815, "learning_rate": 9.922619022162545e-05, "loss": 3.5227, "step": 2000 }, { "epoch": 0.06, "grad_norm": 0.8294447064399719, "learning_rate": 9.92221227868021e-05, "loss": 3.486, "step": 2005 }, { "epoch": 0.06, "grad_norm": 0.7031897902488708, "learning_rate": 9.921804477380394e-05, "loss": 3.5868, "step": 2010 }, { "epoch": 0.06, "grad_norm": 0.8516455888748169, "learning_rate": 9.921395618350739e-05, "loss": 3.4631, "step": 2015 }, { "epoch": 0.06, "grad_norm": 0.7513537406921387, "learning_rate": 9.920985701679106e-05, "loss": 3.5133, "step": 2020 }, { "epoch": 0.06, "grad_norm": 0.7153122425079346, "learning_rate": 9.920574727453595e-05, "loss": 3.5687, "step": 2025 }, { "epoch": 0.06, "grad_norm": 0.674989640712738, "learning_rate": 9.920162695762522e-05, "loss": 3.5351, "step": 2030 }, { "epoch": 0.06, "grad_norm": 0.7642735838890076, "learning_rate": 9.919749606694439e-05, "loss": 3.5423, "step": 2035 }, { "epoch": 0.06, "grad_norm": 0.8540580868721008, "learning_rate": 9.919335460338122e-05, "loss": 3.57, "step": 2040 }, { "epoch": 0.06, "grad_norm": 0.7173014283180237, "learning_rate": 9.918920256782572e-05, "loss": 3.4689, "step": 2045 }, { "epoch": 0.06, "grad_norm": 0.7232670783996582, "learning_rate": 9.918503996117016e-05, "loss": 3.528, "step": 2050 }, { "epoch": 0.06, "grad_norm": 0.814834713935852, "learning_rate": 9.918086678430917e-05, "loss": 3.5305, "step": 2055 }, { "epoch": 0.06, "grad_norm": 0.7525804042816162, "learning_rate": 9.917668303813959e-05, "loss": 3.532, "step": 2060 }, { "epoch": 0.06, "grad_norm": 0.8478850722312927, "learning_rate": 9.917248872356051e-05, "loss": 3.5093, "step": 2065 }, { "epoch": 0.06, "grad_norm": 0.7819585800170898, "learning_rate": 9.916828384147331e-05, "loss": 3.4774, "step": 2070 }, { "epoch": 0.06, "grad_norm": 0.8738789558410645, "learning_rate": 9.916406839278167e-05, "loss": 3.495, "step": 2075 }, { "epoch": 0.06, "grad_norm": 0.815098762512207, "learning_rate": 9.915984237839152e-05, "loss": 3.434, "step": 2080 }, { "epoch": 0.06, "grad_norm": 0.7201285362243652, "learning_rate": 9.915560579921106e-05, "loss": 3.4578, "step": 2085 }, { "epoch": 0.06, "grad_norm": 0.7269396185874939, "learning_rate": 9.915135865615074e-05, "loss": 3.4465, "step": 2090 }, { "epoch": 0.06, "grad_norm": 0.6888795495033264, "learning_rate": 9.914710095012332e-05, "loss": 3.4218, "step": 2095 }, { "epoch": 0.06, "grad_norm": 0.702541172504425, "learning_rate": 9.91428326820438e-05, "loss": 3.4611, "step": 2100 }, { "epoch": 0.06, "grad_norm": 0.7156362533569336, "learning_rate": 9.913855385282947e-05, "loss": 3.3565, "step": 2105 }, { "epoch": 0.06, "grad_norm": 0.7372973561286926, "learning_rate": 9.913426446339986e-05, "loss": 3.5238, "step": 2110 }, { "epoch": 0.06, "grad_norm": 0.9288309812545776, "learning_rate": 9.912996451467682e-05, "loss": 3.4588, "step": 2115 }, { "epoch": 0.06, "grad_norm": 0.7342353463172913, "learning_rate": 9.912565400758443e-05, "loss": 3.3814, "step": 2120 }, { "epoch": 0.06, "grad_norm": 0.71045982837677, "learning_rate": 9.912133294304901e-05, "loss": 3.5011, "step": 2125 }, { "epoch": 0.06, "grad_norm": 0.7333769798278809, "learning_rate": 9.911700132199924e-05, "loss": 3.505, "step": 2130 }, { "epoch": 0.06, "grad_norm": 0.7125977277755737, "learning_rate": 9.911265914536598e-05, "loss": 3.4928, "step": 2135 }, { "epoch": 0.06, "grad_norm": 0.708410382270813, "learning_rate": 9.910830641408242e-05, "loss": 3.4546, "step": 2140 }, { "epoch": 0.06, "grad_norm": 0.7392338514328003, "learning_rate": 9.910394312908397e-05, "loss": 3.5723, "step": 2145 }, { "epoch": 0.06, "grad_norm": 0.7267410159111023, "learning_rate": 9.909956929130834e-05, "loss": 3.4009, "step": 2150 }, { "epoch": 0.06, "grad_norm": 0.8049618601799011, "learning_rate": 9.90951849016955e-05, "loss": 3.398, "step": 2155 }, { "epoch": 0.06, "grad_norm": 0.8601283431053162, "learning_rate": 9.909078996118768e-05, "loss": 3.4611, "step": 2160 }, { "epoch": 0.06, "grad_norm": 0.7945659160614014, "learning_rate": 9.908638447072939e-05, "loss": 3.5444, "step": 2165 }, { "epoch": 0.06, "grad_norm": 0.8157386779785156, "learning_rate": 9.90819684312674e-05, "loss": 3.5183, "step": 2170 }, { "epoch": 0.06, "grad_norm": 0.7712591886520386, "learning_rate": 9.907754184375074e-05, "loss": 3.4777, "step": 2175 }, { "epoch": 0.06, "grad_norm": 0.7726550102233887, "learning_rate": 9.90731047091307e-05, "loss": 3.4551, "step": 2180 }, { "epoch": 0.06, "grad_norm": 0.7134284973144531, "learning_rate": 9.906865702836088e-05, "loss": 3.4422, "step": 2185 }, { "epoch": 0.06, "grad_norm": 0.8338950276374817, "learning_rate": 9.90641988023971e-05, "loss": 3.4534, "step": 2190 }, { "epoch": 0.06, "grad_norm": 0.7876210808753967, "learning_rate": 9.905973003219748e-05, "loss": 3.4421, "step": 2195 }, { "epoch": 0.06, "grad_norm": 0.7938791513442993, "learning_rate": 9.905525071872237e-05, "loss": 3.568, "step": 2200 }, { "epoch": 0.06, "grad_norm": 0.670796811580658, "learning_rate": 9.905076086293443e-05, "loss": 3.5264, "step": 2205 }, { "epoch": 0.07, "grad_norm": 0.7210372686386108, "learning_rate": 9.904626046579852e-05, "loss": 3.5041, "step": 2210 }, { "epoch": 0.07, "grad_norm": 0.6619365811347961, "learning_rate": 9.904174952828186e-05, "loss": 3.4729, "step": 2215 }, { "epoch": 0.07, "grad_norm": 0.6500024199485779, "learning_rate": 9.903722805135385e-05, "loss": 3.4561, "step": 2220 }, { "epoch": 0.07, "grad_norm": 0.6735613942146301, "learning_rate": 9.903269603598617e-05, "loss": 3.435, "step": 2225 }, { "epoch": 0.07, "grad_norm": 0.7042248249053955, "learning_rate": 9.902815348315282e-05, "loss": 3.4837, "step": 2230 }, { "epoch": 0.07, "grad_norm": 0.6572805643081665, "learning_rate": 9.902360039383e-05, "loss": 3.4449, "step": 2235 }, { "epoch": 0.07, "grad_norm": 0.6619617938995361, "learning_rate": 9.901903676899621e-05, "loss": 3.3625, "step": 2240 }, { "epoch": 0.07, "grad_norm": 0.6633297204971313, "learning_rate": 9.901446260963221e-05, "loss": 3.4752, "step": 2245 }, { "epoch": 0.07, "grad_norm": 0.7962461709976196, "learning_rate": 9.9009877916721e-05, "loss": 3.5551, "step": 2250 }, { "epoch": 0.07, "grad_norm": 0.7498233914375305, "learning_rate": 9.900528269124787e-05, "loss": 3.373, "step": 2255 }, { "epoch": 0.07, "grad_norm": 0.7017486691474915, "learning_rate": 9.900067693420039e-05, "loss": 3.3791, "step": 2260 }, { "epoch": 0.07, "grad_norm": 0.6685238480567932, "learning_rate": 9.899606064656832e-05, "loss": 3.4144, "step": 2265 }, { "epoch": 0.07, "grad_norm": 0.6954371333122253, "learning_rate": 9.899143382934379e-05, "loss": 3.3559, "step": 2270 }, { "epoch": 0.07, "grad_norm": 0.7096899747848511, "learning_rate": 9.898679648352108e-05, "loss": 3.4722, "step": 2275 }, { "epoch": 0.07, "grad_norm": 0.7208156585693359, "learning_rate": 9.898214861009683e-05, "loss": 3.3507, "step": 2280 }, { "epoch": 0.07, "grad_norm": 0.8103269934654236, "learning_rate": 9.897749021006988e-05, "loss": 3.4038, "step": 2285 }, { "epoch": 0.07, "grad_norm": 0.7369759678840637, "learning_rate": 9.897282128444135e-05, "loss": 3.4423, "step": 2290 }, { "epoch": 0.07, "grad_norm": 0.7848606705665588, "learning_rate": 9.896814183421464e-05, "loss": 3.5073, "step": 2295 }, { "epoch": 0.07, "grad_norm": 0.8676486015319824, "learning_rate": 9.896345186039538e-05, "loss": 3.3529, "step": 2300 }, { "epoch": 0.07, "grad_norm": 2.0192975997924805, "learning_rate": 9.895875136399149e-05, "loss": 3.4561, "step": 2305 }, { "epoch": 0.07, "grad_norm": 0.6903976202011108, "learning_rate": 9.895404034601313e-05, "loss": 3.4027, "step": 2310 }, { "epoch": 0.07, "grad_norm": 0.7228520512580872, "learning_rate": 9.894931880747273e-05, "loss": 3.3848, "step": 2315 }, { "epoch": 0.07, "grad_norm": 0.6471682190895081, "learning_rate": 9.894458674938499e-05, "loss": 3.3961, "step": 2320 }, { "epoch": 0.07, "grad_norm": 0.7371661067008972, "learning_rate": 9.893984417276686e-05, "loss": 3.414, "step": 2325 }, { "epoch": 0.07, "grad_norm": 0.8883764743804932, "learning_rate": 9.893509107863754e-05, "loss": 3.5027, "step": 2330 }, { "epoch": 0.07, "grad_norm": 0.651530385017395, "learning_rate": 9.893032746801854e-05, "loss": 3.3923, "step": 2335 }, { "epoch": 0.07, "grad_norm": 0.7196444869041443, "learning_rate": 9.892555334193353e-05, "loss": 3.398, "step": 2340 }, { "epoch": 0.07, "grad_norm": 0.6983844041824341, "learning_rate": 9.892076870140857e-05, "loss": 3.4249, "step": 2345 }, { "epoch": 0.07, "grad_norm": 0.6999300122261047, "learning_rate": 9.891597354747186e-05, "loss": 3.3823, "step": 2350 }, { "epoch": 0.07, "grad_norm": 0.7112782001495361, "learning_rate": 9.891116788115393e-05, "loss": 3.376, "step": 2355 }, { "epoch": 0.07, "grad_norm": 0.7380182147026062, "learning_rate": 9.890635170348756e-05, "loss": 3.3478, "step": 2360 }, { "epoch": 0.07, "grad_norm": 0.7203754782676697, "learning_rate": 9.890152501550777e-05, "loss": 3.4098, "step": 2365 }, { "epoch": 0.07, "grad_norm": 0.6570917963981628, "learning_rate": 9.889668781825185e-05, "loss": 3.4262, "step": 2370 }, { "epoch": 0.07, "grad_norm": 0.7178522944450378, "learning_rate": 9.889184011275936e-05, "loss": 3.3426, "step": 2375 }, { "epoch": 0.07, "grad_norm": 0.6552417278289795, "learning_rate": 9.888698190007209e-05, "loss": 3.4666, "step": 2380 }, { "epoch": 0.07, "grad_norm": 0.696057140827179, "learning_rate": 9.88821131812341e-05, "loss": 3.3817, "step": 2385 }, { "epoch": 0.07, "grad_norm": 0.660396158695221, "learning_rate": 9.887723395729174e-05, "loss": 3.4626, "step": 2390 }, { "epoch": 0.07, "grad_norm": 0.6552888751029968, "learning_rate": 9.887234422929355e-05, "loss": 3.3101, "step": 2395 }, { "epoch": 0.07, "grad_norm": 0.6441607475280762, "learning_rate": 9.886744399829039e-05, "loss": 3.4529, "step": 2400 }, { "epoch": 0.07, "grad_norm": 0.6489095687866211, "learning_rate": 9.886253326533535e-05, "loss": 3.3907, "step": 2405 }, { "epoch": 0.07, "grad_norm": 0.6902307868003845, "learning_rate": 9.885761203148378e-05, "loss": 3.3348, "step": 2410 }, { "epoch": 0.07, "grad_norm": 0.6929933428764343, "learning_rate": 9.885268029779327e-05, "loss": 3.3661, "step": 2415 }, { "epoch": 0.07, "grad_norm": 0.7540335059165955, "learning_rate": 9.884773806532373e-05, "loss": 3.3276, "step": 2420 }, { "epoch": 0.07, "grad_norm": 0.7650929093360901, "learning_rate": 9.884278533513723e-05, "loss": 3.3864, "step": 2425 }, { "epoch": 0.07, "grad_norm": 0.6789025068283081, "learning_rate": 9.883782210829816e-05, "loss": 3.478, "step": 2430 }, { "epoch": 0.07, "grad_norm": 0.6747824549674988, "learning_rate": 9.883284838587317e-05, "loss": 3.3207, "step": 2435 }, { "epoch": 0.07, "grad_norm": 0.7538822293281555, "learning_rate": 9.882786416893112e-05, "loss": 3.3759, "step": 2440 }, { "epoch": 0.07, "grad_norm": 0.6605286002159119, "learning_rate": 9.882286945854319e-05, "loss": 3.4137, "step": 2445 }, { "epoch": 0.07, "grad_norm": 0.7063916325569153, "learning_rate": 9.881786425578274e-05, "loss": 3.4272, "step": 2450 }, { "epoch": 0.07, "grad_norm": 0.6728413701057434, "learning_rate": 9.881284856172545e-05, "loss": 3.4269, "step": 2455 }, { "epoch": 0.07, "grad_norm": 0.8078657388687134, "learning_rate": 9.880782237744921e-05, "loss": 3.4545, "step": 2460 }, { "epoch": 0.07, "grad_norm": 0.8296356201171875, "learning_rate": 9.880278570403419e-05, "loss": 3.4093, "step": 2465 }, { "epoch": 0.07, "grad_norm": 0.6820523142814636, "learning_rate": 9.87977385425628e-05, "loss": 3.4441, "step": 2470 }, { "epoch": 0.07, "grad_norm": 0.7303041219711304, "learning_rate": 9.87926808941197e-05, "loss": 3.3292, "step": 2475 }, { "epoch": 0.07, "grad_norm": 0.797197699546814, "learning_rate": 9.878761275979184e-05, "loss": 3.359, "step": 2480 }, { "epoch": 0.07, "grad_norm": 0.639845609664917, "learning_rate": 9.878253414066839e-05, "loss": 3.2871, "step": 2485 }, { "epoch": 0.07, "grad_norm": 0.7427608370780945, "learning_rate": 9.877744503784076e-05, "loss": 3.3827, "step": 2490 }, { "epoch": 0.07, "grad_norm": 0.6702534556388855, "learning_rate": 9.877234545240265e-05, "loss": 3.3562, "step": 2495 }, { "epoch": 0.07, "grad_norm": 0.6945351958274841, "learning_rate": 9.876723538545001e-05, "loss": 3.3781, "step": 2500 }, { "epoch": 0.07, "grad_norm": 0.7000056505203247, "learning_rate": 9.8762114838081e-05, "loss": 3.3585, "step": 2505 }, { "epoch": 0.07, "grad_norm": 0.6774452924728394, "learning_rate": 9.875698381139607e-05, "loss": 3.4451, "step": 2510 }, { "epoch": 0.07, "grad_norm": 0.6585217118263245, "learning_rate": 9.875184230649792e-05, "loss": 3.3785, "step": 2515 }, { "epoch": 0.07, "grad_norm": 0.6287395358085632, "learning_rate": 9.874669032449149e-05, "loss": 3.3182, "step": 2520 }, { "epoch": 0.07, "grad_norm": 0.6221347451210022, "learning_rate": 9.874152786648399e-05, "loss": 3.2285, "step": 2525 }, { "epoch": 0.07, "grad_norm": 0.6900961399078369, "learning_rate": 9.873635493358484e-05, "loss": 3.3228, "step": 2530 }, { "epoch": 0.07, "grad_norm": 0.6223905682563782, "learning_rate": 9.873117152690576e-05, "loss": 3.2808, "step": 2535 }, { "epoch": 0.07, "grad_norm": 0.670589029788971, "learning_rate": 9.872597764756071e-05, "loss": 3.3886, "step": 2540 }, { "epoch": 0.07, "grad_norm": 0.6326797008514404, "learning_rate": 9.872077329666587e-05, "loss": 3.4215, "step": 2545 }, { "epoch": 0.08, "grad_norm": 0.6296000480651855, "learning_rate": 9.871555847533968e-05, "loss": 3.298, "step": 2550 }, { "epoch": 0.08, "grad_norm": 0.6420964598655701, "learning_rate": 9.871033318470289e-05, "loss": 3.3498, "step": 2555 }, { "epoch": 0.08, "grad_norm": 0.6330773830413818, "learning_rate": 9.870509742587838e-05, "loss": 3.2644, "step": 2560 }, { "epoch": 0.08, "grad_norm": 0.6644457578659058, "learning_rate": 9.869985119999143e-05, "loss": 3.2862, "step": 2565 }, { "epoch": 0.08, "grad_norm": 0.7229827642440796, "learning_rate": 9.869459450816942e-05, "loss": 3.3475, "step": 2570 }, { "epoch": 0.08, "grad_norm": 0.6522821187973022, "learning_rate": 9.86893273515421e-05, "loss": 3.3489, "step": 2575 }, { "epoch": 0.08, "grad_norm": 0.632811427116394, "learning_rate": 9.868404973124139e-05, "loss": 3.3371, "step": 2580 }, { "epoch": 0.08, "grad_norm": 0.6673424243927002, "learning_rate": 9.86787616484015e-05, "loss": 3.3707, "step": 2585 }, { "epoch": 0.08, "grad_norm": 0.6768385171890259, "learning_rate": 9.867346310415886e-05, "loss": 3.3406, "step": 2590 }, { "epoch": 0.08, "grad_norm": 0.6486900448799133, "learning_rate": 9.866815409965219e-05, "loss": 3.3321, "step": 2595 }, { "epoch": 0.08, "grad_norm": 0.6569724678993225, "learning_rate": 9.86628346360224e-05, "loss": 3.327, "step": 2600 }, { "epoch": 0.08, "grad_norm": 0.6284388899803162, "learning_rate": 9.865750471441271e-05, "loss": 3.3931, "step": 2605 }, { "epoch": 0.08, "grad_norm": 0.7590078711509705, "learning_rate": 9.865216433596853e-05, "loss": 3.3176, "step": 2610 }, { "epoch": 0.08, "grad_norm": 0.6513044238090515, "learning_rate": 9.864681350183758e-05, "loss": 3.4031, "step": 2615 }, { "epoch": 0.08, "grad_norm": 0.6849138140678406, "learning_rate": 9.864145221316975e-05, "loss": 3.4945, "step": 2620 }, { "epoch": 0.08, "grad_norm": 0.7085199952125549, "learning_rate": 9.863608047111725e-05, "loss": 3.4191, "step": 2625 }, { "epoch": 0.08, "grad_norm": 0.6591835021972656, "learning_rate": 9.863069827683446e-05, "loss": 3.3912, "step": 2630 }, { "epoch": 0.08, "grad_norm": 0.6397624611854553, "learning_rate": 9.862530563147812e-05, "loss": 3.381, "step": 2635 }, { "epoch": 0.08, "grad_norm": 0.6810360550880432, "learning_rate": 9.861990253620709e-05, "loss": 3.1939, "step": 2640 }, { "epoch": 0.08, "grad_norm": 0.6820381879806519, "learning_rate": 9.861448899218255e-05, "loss": 3.3499, "step": 2645 }, { "epoch": 0.08, "grad_norm": 0.6606658697128296, "learning_rate": 9.860906500056792e-05, "loss": 3.2597, "step": 2650 }, { "epoch": 0.08, "grad_norm": 1.1574220657348633, "learning_rate": 9.860363056252883e-05, "loss": 3.4242, "step": 2655 }, { "epoch": 0.08, "grad_norm": 0.5940909385681152, "learning_rate": 9.859818567923321e-05, "loss": 3.3207, "step": 2660 }, { "epoch": 0.08, "grad_norm": 0.7738911509513855, "learning_rate": 9.859273035185118e-05, "loss": 3.3424, "step": 2665 }, { "epoch": 0.08, "grad_norm": 0.6553605794906616, "learning_rate": 9.858726458155512e-05, "loss": 3.3888, "step": 2670 }, { "epoch": 0.08, "grad_norm": 0.8117573261260986, "learning_rate": 9.858178836951967e-05, "loss": 3.2456, "step": 2675 }, { "epoch": 0.08, "grad_norm": 0.7590681314468384, "learning_rate": 9.857630171692174e-05, "loss": 3.355, "step": 2680 }, { "epoch": 0.08, "grad_norm": 0.7324884533882141, "learning_rate": 9.857080462494039e-05, "loss": 3.299, "step": 2685 }, { "epoch": 0.08, "grad_norm": 0.6585342288017273, "learning_rate": 9.856529709475703e-05, "loss": 3.3112, "step": 2690 }, { "epoch": 0.08, "grad_norm": 0.6684226393699646, "learning_rate": 9.855977912755524e-05, "loss": 3.3928, "step": 2695 }, { "epoch": 0.08, "grad_norm": 0.7311132550239563, "learning_rate": 9.855425072452086e-05, "loss": 3.2507, "step": 2700 }, { "epoch": 0.08, "grad_norm": 0.5805177092552185, "learning_rate": 9.854871188684204e-05, "loss": 3.1977, "step": 2705 }, { "epoch": 0.08, "grad_norm": 0.678617537021637, "learning_rate": 9.854316261570905e-05, "loss": 3.2952, "step": 2710 }, { "epoch": 0.08, "grad_norm": 0.5892093181610107, "learning_rate": 9.85376029123145e-05, "loss": 3.3137, "step": 2715 }, { "epoch": 0.08, "grad_norm": 0.6289623379707336, "learning_rate": 9.853203277785318e-05, "loss": 3.4457, "step": 2720 }, { "epoch": 0.08, "grad_norm": 0.6970494389533997, "learning_rate": 9.85264522135222e-05, "loss": 3.3068, "step": 2725 }, { "epoch": 0.08, "grad_norm": 0.6420297622680664, "learning_rate": 9.852086122052083e-05, "loss": 3.3335, "step": 2730 }, { "epoch": 0.08, "grad_norm": 0.6377905607223511, "learning_rate": 9.85152598000506e-05, "loss": 3.3212, "step": 2735 }, { "epoch": 0.08, "grad_norm": 0.6411159634590149, "learning_rate": 9.850964795331531e-05, "loss": 3.4157, "step": 2740 }, { "epoch": 0.08, "grad_norm": 0.634062647819519, "learning_rate": 9.8504025681521e-05, "loss": 3.3188, "step": 2745 }, { "epoch": 0.08, "grad_norm": 0.6769227385520935, "learning_rate": 9.84983929858759e-05, "loss": 3.2857, "step": 2750 }, { "epoch": 0.08, "grad_norm": 0.8588881492614746, "learning_rate": 9.849274986759054e-05, "loss": 3.2272, "step": 2755 }, { "epoch": 0.08, "grad_norm": 0.6263747215270996, "learning_rate": 9.848709632787768e-05, "loss": 3.3856, "step": 2760 }, { "epoch": 0.08, "grad_norm": 0.6576184630393982, "learning_rate": 9.848143236795225e-05, "loss": 3.2868, "step": 2765 }, { "epoch": 0.08, "grad_norm": 0.6202845573425293, "learning_rate": 9.847575798903152e-05, "loss": 3.3234, "step": 2770 }, { "epoch": 0.08, "grad_norm": 0.6550593376159668, "learning_rate": 9.847007319233496e-05, "loss": 3.388, "step": 2775 }, { "epoch": 0.08, "grad_norm": 0.6809086799621582, "learning_rate": 9.846437797908424e-05, "loss": 3.4014, "step": 2780 }, { "epoch": 0.08, "grad_norm": 0.7005331516265869, "learning_rate": 9.845867235050333e-05, "loss": 3.2363, "step": 2785 }, { "epoch": 0.08, "grad_norm": 0.7052482962608337, "learning_rate": 9.845295630781837e-05, "loss": 3.3794, "step": 2790 }, { "epoch": 0.08, "grad_norm": 0.6093045473098755, "learning_rate": 9.844722985225783e-05, "loss": 3.2313, "step": 2795 }, { "epoch": 0.08, "grad_norm": 0.6210280656814575, "learning_rate": 9.844149298505232e-05, "loss": 3.2822, "step": 2800 }, { "epoch": 0.08, "grad_norm": 0.644005298614502, "learning_rate": 9.843574570743474e-05, "loss": 3.316, "step": 2805 }, { "epoch": 0.08, "grad_norm": 0.6252589225769043, "learning_rate": 9.842998802064025e-05, "loss": 3.2607, "step": 2810 }, { "epoch": 0.08, "grad_norm": 0.6432593464851379, "learning_rate": 9.842421992590619e-05, "loss": 3.3735, "step": 2815 }, { "epoch": 0.08, "grad_norm": 0.7279803156852722, "learning_rate": 9.841844142447215e-05, "loss": 3.3069, "step": 2820 }, { "epoch": 0.08, "grad_norm": 0.6755356192588806, "learning_rate": 9.841265251758e-05, "loss": 3.2942, "step": 2825 }, { "epoch": 0.08, "grad_norm": 0.6164819002151489, "learning_rate": 9.840685320647381e-05, "loss": 3.3637, "step": 2830 }, { "epoch": 0.08, "grad_norm": 0.6049258708953857, "learning_rate": 9.840104349239988e-05, "loss": 3.3222, "step": 2835 }, { "epoch": 0.08, "grad_norm": 0.6181380748748779, "learning_rate": 9.839522337660677e-05, "loss": 3.313, "step": 2840 }, { "epoch": 0.08, "grad_norm": 0.6054235100746155, "learning_rate": 9.838939286034526e-05, "loss": 3.2657, "step": 2845 }, { "epoch": 0.08, "grad_norm": 0.6516208648681641, "learning_rate": 9.838355194486836e-05, "loss": 3.2984, "step": 2850 }, { "epoch": 0.08, "grad_norm": 0.6166696548461914, "learning_rate": 9.837770063143133e-05, "loss": 3.2742, "step": 2855 }, { "epoch": 0.08, "grad_norm": 0.6297792196273804, "learning_rate": 9.837183892129164e-05, "loss": 3.2653, "step": 2860 }, { "epoch": 0.08, "grad_norm": 0.6807376742362976, "learning_rate": 9.836596681570906e-05, "loss": 3.3364, "step": 2865 }, { "epoch": 0.08, "grad_norm": 0.5804564952850342, "learning_rate": 9.83600843159455e-05, "loss": 3.3164, "step": 2870 }, { "epoch": 0.08, "grad_norm": 0.6153858304023743, "learning_rate": 9.835419142326517e-05, "loss": 3.3168, "step": 2875 }, { "epoch": 0.08, "grad_norm": 0.6111676692962646, "learning_rate": 9.834828813893448e-05, "loss": 3.3129, "step": 2880 }, { "epoch": 0.08, "grad_norm": 0.6354882717132568, "learning_rate": 9.83423744642221e-05, "loss": 3.2187, "step": 2885 }, { "epoch": 0.09, "grad_norm": 0.647480845451355, "learning_rate": 9.833645040039892e-05, "loss": 3.2614, "step": 2890 }, { "epoch": 0.09, "grad_norm": 0.6097538471221924, "learning_rate": 9.833051594873805e-05, "loss": 3.3242, "step": 2895 }, { "epoch": 0.09, "grad_norm": 0.6406658887863159, "learning_rate": 9.832457111051485e-05, "loss": 3.4, "step": 2900 }, { "epoch": 0.09, "grad_norm": 0.5902734398841858, "learning_rate": 9.831861588700692e-05, "loss": 3.2714, "step": 2905 }, { "epoch": 0.09, "grad_norm": 0.5859482288360596, "learning_rate": 9.831265027949406e-05, "loss": 3.1442, "step": 2910 }, { "epoch": 0.09, "grad_norm": 0.6079094409942627, "learning_rate": 9.830667428925833e-05, "loss": 3.3367, "step": 2915 }, { "epoch": 0.09, "grad_norm": 0.643936812877655, "learning_rate": 9.8300687917584e-05, "loss": 3.3446, "step": 2920 }, { "epoch": 0.09, "grad_norm": 0.6310209631919861, "learning_rate": 9.82946911657576e-05, "loss": 3.2836, "step": 2925 }, { "epoch": 0.09, "grad_norm": 0.6102332472801208, "learning_rate": 9.828868403506788e-05, "loss": 3.2548, "step": 2930 }, { "epoch": 0.09, "grad_norm": 0.6653298735618591, "learning_rate": 9.828266652680577e-05, "loss": 3.1943, "step": 2935 }, { "epoch": 0.09, "grad_norm": 0.599534273147583, "learning_rate": 9.827663864226453e-05, "loss": 3.3385, "step": 2940 }, { "epoch": 0.09, "grad_norm": 0.6124973297119141, "learning_rate": 9.827060038273956e-05, "loss": 3.2328, "step": 2945 }, { "epoch": 0.09, "grad_norm": 0.5937116146087646, "learning_rate": 9.826455174952852e-05, "loss": 3.2258, "step": 2950 }, { "epoch": 0.09, "grad_norm": 0.6210477352142334, "learning_rate": 9.825849274393133e-05, "loss": 3.3952, "step": 2955 }, { "epoch": 0.09, "grad_norm": 0.6480196118354797, "learning_rate": 9.825242336725009e-05, "loss": 3.333, "step": 2960 }, { "epoch": 0.09, "grad_norm": 0.6186568140983582, "learning_rate": 9.824634362078915e-05, "loss": 3.2746, "step": 2965 }, { "epoch": 0.09, "grad_norm": 0.6681870818138123, "learning_rate": 9.824025350585511e-05, "loss": 3.3285, "step": 2970 }, { "epoch": 0.09, "grad_norm": 0.5900624990463257, "learning_rate": 9.823415302375676e-05, "loss": 3.3052, "step": 2975 }, { "epoch": 0.09, "grad_norm": 0.5735652446746826, "learning_rate": 9.822804217580515e-05, "loss": 3.3126, "step": 2980 }, { "epoch": 0.09, "grad_norm": 0.5781747102737427, "learning_rate": 9.822192096331351e-05, "loss": 3.2822, "step": 2985 }, { "epoch": 0.09, "grad_norm": 0.6086410284042358, "learning_rate": 9.82157893875974e-05, "loss": 3.3214, "step": 2990 }, { "epoch": 0.09, "grad_norm": 0.6496842503547668, "learning_rate": 9.820964744997445e-05, "loss": 3.194, "step": 2995 }, { "epoch": 0.09, "grad_norm": 0.6699143052101135, "learning_rate": 9.820349515176466e-05, "loss": 3.3178, "step": 3000 }, { "epoch": 0.09, "grad_norm": 0.6290653347969055, "learning_rate": 9.819733249429019e-05, "loss": 3.2486, "step": 3005 }, { "epoch": 0.09, "grad_norm": 0.6164487600326538, "learning_rate": 9.819115947887545e-05, "loss": 3.2754, "step": 3010 }, { "epoch": 0.09, "grad_norm": 0.6195471882820129, "learning_rate": 9.818497610684705e-05, "loss": 3.2304, "step": 3015 }, { "epoch": 0.09, "grad_norm": 0.6471094489097595, "learning_rate": 9.817878237953383e-05, "loss": 3.2438, "step": 3020 }, { "epoch": 0.09, "grad_norm": 0.6307451128959656, "learning_rate": 9.817257829826689e-05, "loss": 3.2893, "step": 3025 }, { "epoch": 0.09, "grad_norm": 0.6299492120742798, "learning_rate": 9.81663638643795e-05, "loss": 3.2391, "step": 3030 }, { "epoch": 0.09, "grad_norm": 0.6102019548416138, "learning_rate": 9.81601390792072e-05, "loss": 3.298, "step": 3035 }, { "epoch": 0.09, "grad_norm": 0.5936606526374817, "learning_rate": 9.815390394408776e-05, "loss": 3.1512, "step": 3040 }, { "epoch": 0.09, "grad_norm": 0.6099717617034912, "learning_rate": 9.814765846036111e-05, "loss": 3.1716, "step": 3045 }, { "epoch": 0.09, "grad_norm": 0.606975257396698, "learning_rate": 9.81414026293695e-05, "loss": 3.3141, "step": 3050 }, { "epoch": 0.09, "grad_norm": 0.5904830694198608, "learning_rate": 9.813513645245729e-05, "loss": 3.2687, "step": 3055 }, { "epoch": 0.09, "grad_norm": 0.5931333303451538, "learning_rate": 9.812885993097118e-05, "loss": 3.3064, "step": 3060 }, { "epoch": 0.09, "grad_norm": 0.6358606815338135, "learning_rate": 9.812257306626002e-05, "loss": 3.2601, "step": 3065 }, { "epoch": 0.09, "grad_norm": 0.6355937719345093, "learning_rate": 9.811627585967487e-05, "loss": 3.2344, "step": 3070 }, { "epoch": 0.09, "grad_norm": 0.6403692364692688, "learning_rate": 9.810996831256909e-05, "loss": 3.2417, "step": 3075 }, { "epoch": 0.09, "grad_norm": 0.6127886176109314, "learning_rate": 9.81036504262982e-05, "loss": 3.2897, "step": 3080 }, { "epoch": 0.09, "grad_norm": 0.5933552980422974, "learning_rate": 9.809732220221995e-05, "loss": 3.2357, "step": 3085 }, { "epoch": 0.09, "grad_norm": 0.5836284756660461, "learning_rate": 9.80909836416943e-05, "loss": 3.2875, "step": 3090 }, { "epoch": 0.09, "grad_norm": 0.6071823835372925, "learning_rate": 9.80846347460835e-05, "loss": 3.3336, "step": 3095 }, { "epoch": 0.09, "grad_norm": 0.6008365154266357, "learning_rate": 9.807827551675192e-05, "loss": 3.225, "step": 3100 }, { "epoch": 0.09, "grad_norm": 0.593515932559967, "learning_rate": 9.807190595506623e-05, "loss": 3.2699, "step": 3105 }, { "epoch": 0.09, "grad_norm": 0.6305873990058899, "learning_rate": 9.80655260623953e-05, "loss": 3.2992, "step": 3110 }, { "epoch": 0.09, "grad_norm": 0.6164395213127136, "learning_rate": 9.805913584011019e-05, "loss": 3.3866, "step": 3115 }, { "epoch": 0.09, "grad_norm": 0.5860743522644043, "learning_rate": 9.805273528958423e-05, "loss": 3.3116, "step": 3120 }, { "epoch": 0.09, "grad_norm": 0.7481256127357483, "learning_rate": 9.804632441219293e-05, "loss": 3.2524, "step": 3125 }, { "epoch": 0.09, "grad_norm": 0.594959557056427, "learning_rate": 9.803990320931402e-05, "loss": 3.3215, "step": 3130 }, { "epoch": 0.09, "grad_norm": 0.609107255935669, "learning_rate": 9.803347168232746e-05, "loss": 3.2532, "step": 3135 }, { "epoch": 0.09, "grad_norm": 0.7000026106834412, "learning_rate": 9.802702983261545e-05, "loss": 3.1524, "step": 3140 }, { "epoch": 0.09, "grad_norm": 0.5761329531669617, "learning_rate": 9.802057766156238e-05, "loss": 3.2812, "step": 3145 }, { "epoch": 0.09, "grad_norm": 0.593431830406189, "learning_rate": 9.801411517055488e-05, "loss": 3.242, "step": 3150 }, { "epoch": 0.09, "grad_norm": 0.5649806261062622, "learning_rate": 9.800764236098175e-05, "loss": 3.2737, "step": 3155 }, { "epoch": 0.09, "grad_norm": 0.5709115862846375, "learning_rate": 9.800115923423405e-05, "loss": 3.2472, "step": 3160 }, { "epoch": 0.09, "grad_norm": 0.5728706121444702, "learning_rate": 9.799466579170508e-05, "loss": 3.2799, "step": 3165 }, { "epoch": 0.09, "grad_norm": 0.6307482123374939, "learning_rate": 9.798816203479029e-05, "loss": 3.2892, "step": 3170 }, { "epoch": 0.09, "grad_norm": 0.5783202052116394, "learning_rate": 9.79816479648874e-05, "loss": 3.2645, "step": 3175 }, { "epoch": 0.09, "grad_norm": 0.5890851020812988, "learning_rate": 9.797512358339633e-05, "loss": 3.1806, "step": 3180 }, { "epoch": 0.09, "grad_norm": 0.5921275615692139, "learning_rate": 9.796858889171921e-05, "loss": 3.1968, "step": 3185 }, { "epoch": 0.09, "grad_norm": 0.6137367486953735, "learning_rate": 9.79620438912604e-05, "loss": 3.286, "step": 3190 }, { "epoch": 0.09, "grad_norm": 0.5878112316131592, "learning_rate": 9.795548858342645e-05, "loss": 3.1325, "step": 3195 }, { "epoch": 0.09, "grad_norm": 0.5992056727409363, "learning_rate": 9.794892296962615e-05, "loss": 3.2567, "step": 3200 }, { "epoch": 0.09, "grad_norm": 0.5811440348625183, "learning_rate": 9.794234705127047e-05, "loss": 3.2528, "step": 3205 }, { "epoch": 0.09, "grad_norm": 0.6312845945358276, "learning_rate": 9.793576082977267e-05, "loss": 3.2037, "step": 3210 }, { "epoch": 0.09, "grad_norm": 0.6497148275375366, "learning_rate": 9.792916430654815e-05, "loss": 3.3002, "step": 3215 }, { "epoch": 0.09, "grad_norm": 0.5811987519264221, "learning_rate": 9.792255748301453e-05, "loss": 3.2023, "step": 3220 }, { "epoch": 0.09, "grad_norm": 0.697776734828949, "learning_rate": 9.791594036059168e-05, "loss": 3.3298, "step": 3225 }, { "epoch": 0.1, "grad_norm": 0.5982942581176758, "learning_rate": 9.790931294070169e-05, "loss": 3.2838, "step": 3230 }, { "epoch": 0.1, "grad_norm": 0.6204692721366882, "learning_rate": 9.79026752247688e-05, "loss": 3.2377, "step": 3235 }, { "epoch": 0.1, "grad_norm": 0.5880468487739563, "learning_rate": 9.78960272142195e-05, "loss": 3.2309, "step": 3240 }, { "epoch": 0.1, "grad_norm": 0.5789374709129333, "learning_rate": 9.788936891048251e-05, "loss": 3.2496, "step": 3245 }, { "epoch": 0.1, "grad_norm": 0.5692088007926941, "learning_rate": 9.788270031498876e-05, "loss": 3.184, "step": 3250 }, { "epoch": 0.1, "grad_norm": 0.5535498261451721, "learning_rate": 9.787602142917135e-05, "loss": 3.2218, "step": 3255 }, { "epoch": 0.1, "grad_norm": 0.6147019267082214, "learning_rate": 9.786933225446563e-05, "loss": 3.2589, "step": 3260 }, { "epoch": 0.1, "grad_norm": 0.5895227789878845, "learning_rate": 9.786263279230915e-05, "loss": 3.1811, "step": 3265 }, { "epoch": 0.1, "grad_norm": 0.5798574686050415, "learning_rate": 9.785592304414167e-05, "loss": 3.2071, "step": 3270 }, { "epoch": 0.1, "grad_norm": 0.6958770155906677, "learning_rate": 9.784920301140516e-05, "loss": 3.2166, "step": 3275 }, { "epoch": 0.1, "grad_norm": 0.5738431811332703, "learning_rate": 9.78424726955438e-05, "loss": 3.2225, "step": 3280 }, { "epoch": 0.1, "grad_norm": 0.5915527939796448, "learning_rate": 9.783573209800398e-05, "loss": 3.2916, "step": 3285 }, { "epoch": 0.1, "grad_norm": 0.567115843296051, "learning_rate": 9.782898122023432e-05, "loss": 3.1545, "step": 3290 }, { "epoch": 0.1, "grad_norm": 0.5832808613777161, "learning_rate": 9.78222200636856e-05, "loss": 3.1565, "step": 3295 }, { "epoch": 0.1, "grad_norm": 0.5933319330215454, "learning_rate": 9.781544862981087e-05, "loss": 3.3251, "step": 3300 }, { "epoch": 0.1, "grad_norm": 0.5824843645095825, "learning_rate": 9.780866692006534e-05, "loss": 3.2173, "step": 3305 }, { "epoch": 0.1, "grad_norm": 0.5954950451850891, "learning_rate": 9.780187493590645e-05, "loss": 3.1761, "step": 3310 }, { "epoch": 0.1, "grad_norm": 0.5867497324943542, "learning_rate": 9.779507267879385e-05, "loss": 3.2406, "step": 3315 }, { "epoch": 0.1, "grad_norm": 0.5827246904373169, "learning_rate": 9.778826015018936e-05, "loss": 3.2133, "step": 3320 }, { "epoch": 0.1, "grad_norm": 0.57917320728302, "learning_rate": 9.778143735155709e-05, "loss": 3.1058, "step": 3325 }, { "epoch": 0.1, "grad_norm": 0.5814988017082214, "learning_rate": 9.777460428436327e-05, "loss": 3.2101, "step": 3330 }, { "epoch": 0.1, "grad_norm": 0.5712803602218628, "learning_rate": 9.776776095007638e-05, "loss": 3.2651, "step": 3335 }, { "epoch": 0.1, "grad_norm": 0.5587366223335266, "learning_rate": 9.776090735016711e-05, "loss": 3.2956, "step": 3340 }, { "epoch": 0.1, "grad_norm": 0.5617526769638062, "learning_rate": 9.775404348610835e-05, "loss": 3.0944, "step": 3345 }, { "epoch": 0.1, "grad_norm": 0.6102294325828552, "learning_rate": 9.774716935937518e-05, "loss": 3.2336, "step": 3350 }, { "epoch": 0.1, "grad_norm": 0.5824145078659058, "learning_rate": 9.77402849714449e-05, "loss": 3.2351, "step": 3355 }, { "epoch": 0.1, "grad_norm": 0.6034987568855286, "learning_rate": 9.773339032379701e-05, "loss": 3.1675, "step": 3360 }, { "epoch": 0.1, "grad_norm": 0.5860341787338257, "learning_rate": 9.772648541791323e-05, "loss": 3.2245, "step": 3365 }, { "epoch": 0.1, "grad_norm": 0.594593346118927, "learning_rate": 9.771957025527745e-05, "loss": 3.1567, "step": 3370 }, { "epoch": 0.1, "grad_norm": 0.5829384922981262, "learning_rate": 9.771264483737581e-05, "loss": 3.2071, "step": 3375 }, { "epoch": 0.1, "grad_norm": 0.575003445148468, "learning_rate": 9.770570916569661e-05, "loss": 3.1994, "step": 3380 }, { "epoch": 0.1, "grad_norm": 0.6086431741714478, "learning_rate": 9.76987632417304e-05, "loss": 3.251, "step": 3385 }, { "epoch": 0.1, "grad_norm": 0.618727445602417, "learning_rate": 9.769180706696988e-05, "loss": 3.2128, "step": 3390 }, { "epoch": 0.1, "grad_norm": 0.5472085475921631, "learning_rate": 9.768484064290997e-05, "loss": 3.1623, "step": 3395 }, { "epoch": 0.1, "grad_norm": 0.587695300579071, "learning_rate": 9.767786397104786e-05, "loss": 3.1755, "step": 3400 }, { "epoch": 0.1, "grad_norm": 0.5886109471321106, "learning_rate": 9.767087705288283e-05, "loss": 3.2578, "step": 3405 }, { "epoch": 0.1, "grad_norm": 0.576249897480011, "learning_rate": 9.766387988991642e-05, "loss": 3.2334, "step": 3410 }, { "epoch": 0.1, "grad_norm": 0.5888702273368835, "learning_rate": 9.765687248365241e-05, "loss": 3.3252, "step": 3415 }, { "epoch": 0.1, "grad_norm": 0.6603434085845947, "learning_rate": 9.76498548355967e-05, "loss": 3.209, "step": 3420 }, { "epoch": 0.1, "grad_norm": 0.5563474297523499, "learning_rate": 9.764282694725744e-05, "loss": 3.2188, "step": 3425 }, { "epoch": 0.1, "grad_norm": 0.5820965766906738, "learning_rate": 9.7635788820145e-05, "loss": 3.1713, "step": 3430 }, { "epoch": 0.1, "grad_norm": 0.6104404926300049, "learning_rate": 9.762874045577189e-05, "loss": 3.1631, "step": 3435 }, { "epoch": 0.1, "grad_norm": 0.6003999710083008, "learning_rate": 9.762168185565284e-05, "loss": 3.2135, "step": 3440 }, { "epoch": 0.1, "grad_norm": 0.605941116809845, "learning_rate": 9.761461302130484e-05, "loss": 3.2193, "step": 3445 }, { "epoch": 0.1, "grad_norm": 0.5936506986618042, "learning_rate": 9.760753395424699e-05, "loss": 3.167, "step": 3450 }, { "epoch": 0.1, "grad_norm": 0.6127610206604004, "learning_rate": 9.760044465600065e-05, "loss": 3.2707, "step": 3455 }, { "epoch": 0.1, "grad_norm": 0.6188499927520752, "learning_rate": 9.759334512808935e-05, "loss": 3.1171, "step": 3460 }, { "epoch": 0.1, "grad_norm": 0.5784873366355896, "learning_rate": 9.758623537203883e-05, "loss": 3.2251, "step": 3465 }, { "epoch": 0.1, "grad_norm": 0.5823073387145996, "learning_rate": 9.757911538937706e-05, "loss": 3.2228, "step": 3470 }, { "epoch": 0.1, "grad_norm": 0.6331013441085815, "learning_rate": 9.757198518163411e-05, "loss": 3.2274, "step": 3475 }, { "epoch": 0.1, "grad_norm": 0.6748613715171814, "learning_rate": 9.756484475034235e-05, "loss": 3.1411, "step": 3480 }, { "epoch": 0.1, "grad_norm": 0.8411778211593628, "learning_rate": 9.75576940970363e-05, "loss": 3.1937, "step": 3485 }, { "epoch": 0.1, "grad_norm": 0.7323165535926819, "learning_rate": 9.755053322325269e-05, "loss": 3.1974, "step": 3490 }, { "epoch": 0.1, "grad_norm": 0.6047230958938599, "learning_rate": 9.754336213053043e-05, "loss": 3.156, "step": 3495 }, { "epoch": 0.1, "grad_norm": 0.5602149963378906, "learning_rate": 9.753618082041066e-05, "loss": 3.1847, "step": 3500 }, { "epoch": 0.1, "grad_norm": 0.6350635886192322, "learning_rate": 9.752898929443666e-05, "loss": 3.2073, "step": 3505 }, { "epoch": 0.1, "grad_norm": 0.5672016739845276, "learning_rate": 9.752178755415396e-05, "loss": 3.2028, "step": 3510 }, { "epoch": 0.1, "grad_norm": 0.5844739079475403, "learning_rate": 9.751457560111026e-05, "loss": 3.1413, "step": 3515 }, { "epoch": 0.1, "grad_norm": 0.5495454668998718, "learning_rate": 9.750735343685547e-05, "loss": 3.1419, "step": 3520 }, { "epoch": 0.1, "grad_norm": 0.5942600965499878, "learning_rate": 9.750012106294167e-05, "loss": 3.1924, "step": 3525 }, { "epoch": 0.1, "grad_norm": 0.5801374912261963, "learning_rate": 9.749287848092315e-05, "loss": 3.1642, "step": 3530 }, { "epoch": 0.1, "grad_norm": 0.6421699523925781, "learning_rate": 9.748562569235637e-05, "loss": 3.129, "step": 3535 }, { "epoch": 0.1, "grad_norm": 0.5806105732917786, "learning_rate": 9.747836269880003e-05, "loss": 3.2763, "step": 3540 }, { "epoch": 0.1, "grad_norm": 0.5928403735160828, "learning_rate": 9.7471089501815e-05, "loss": 3.2326, "step": 3545 }, { "epoch": 0.1, "grad_norm": 0.5585184097290039, "learning_rate": 9.746380610296435e-05, "loss": 3.2498, "step": 3550 }, { "epoch": 0.1, "grad_norm": 0.5656123161315918, "learning_rate": 9.74565125038133e-05, "loss": 3.1275, "step": 3555 }, { "epoch": 0.1, "grad_norm": 0.5340902209281921, "learning_rate": 9.744920870592931e-05, "loss": 3.1717, "step": 3560 }, { "epoch": 0.1, "grad_norm": 0.5904857516288757, "learning_rate": 9.7441894710882e-05, "loss": 3.1464, "step": 3565 }, { "epoch": 0.11, "grad_norm": 0.572156548500061, "learning_rate": 9.743457052024323e-05, "loss": 3.1729, "step": 3570 }, { "epoch": 0.11, "grad_norm": 0.5723282694816589, "learning_rate": 9.742723613558702e-05, "loss": 3.2608, "step": 3575 }, { "epoch": 0.11, "grad_norm": 0.5760853886604309, "learning_rate": 9.741989155848953e-05, "loss": 3.2187, "step": 3580 }, { "epoch": 0.11, "grad_norm": 0.57720947265625, "learning_rate": 9.741253679052923e-05, "loss": 3.1138, "step": 3585 }, { "epoch": 0.11, "grad_norm": 0.5696415305137634, "learning_rate": 9.740517183328666e-05, "loss": 3.1509, "step": 3590 }, { "epoch": 0.11, "grad_norm": 0.5563202500343323, "learning_rate": 9.739779668834461e-05, "loss": 3.1489, "step": 3595 }, { "epoch": 0.11, "grad_norm": 0.5492807626724243, "learning_rate": 9.739041135728806e-05, "loss": 3.1095, "step": 3600 }, { "epoch": 0.11, "grad_norm": 0.6879315972328186, "learning_rate": 9.738301584170417e-05, "loss": 3.2344, "step": 3605 }, { "epoch": 0.11, "grad_norm": 0.5783190131187439, "learning_rate": 9.737561014318228e-05, "loss": 3.2432, "step": 3610 }, { "epoch": 0.11, "grad_norm": 0.5431708693504333, "learning_rate": 9.736819426331392e-05, "loss": 3.2781, "step": 3615 }, { "epoch": 0.11, "grad_norm": 0.5643943548202515, "learning_rate": 9.736076820369283e-05, "loss": 3.1039, "step": 3620 }, { "epoch": 0.11, "grad_norm": 0.6564369201660156, "learning_rate": 9.735333196591492e-05, "loss": 3.204, "step": 3625 }, { "epoch": 0.11, "grad_norm": 0.5608118176460266, "learning_rate": 9.734588555157826e-05, "loss": 3.1887, "step": 3630 }, { "epoch": 0.11, "grad_norm": 0.5692112445831299, "learning_rate": 9.733842896228318e-05, "loss": 3.2229, "step": 3635 }, { "epoch": 0.11, "grad_norm": 0.5543650388717651, "learning_rate": 9.733096219963211e-05, "loss": 3.2483, "step": 3640 }, { "epoch": 0.11, "grad_norm": 0.580712616443634, "learning_rate": 9.732348526522974e-05, "loss": 3.0888, "step": 3645 }, { "epoch": 0.11, "grad_norm": 0.5978821516036987, "learning_rate": 9.73159981606829e-05, "loss": 3.219, "step": 3650 }, { "epoch": 0.11, "grad_norm": 0.572778582572937, "learning_rate": 9.730850088760064e-05, "loss": 3.2503, "step": 3655 }, { "epoch": 0.11, "grad_norm": 0.6753056049346924, "learning_rate": 9.730099344759413e-05, "loss": 3.1068, "step": 3660 }, { "epoch": 0.11, "grad_norm": 0.5945046544075012, "learning_rate": 9.729347584227684e-05, "loss": 3.1648, "step": 3665 }, { "epoch": 0.11, "grad_norm": 0.612129807472229, "learning_rate": 9.72859480732643e-05, "loss": 3.2152, "step": 3670 }, { "epoch": 0.11, "grad_norm": 0.6208339929580688, "learning_rate": 9.727841014217429e-05, "loss": 3.1891, "step": 3675 }, { "epoch": 0.11, "grad_norm": 0.594271183013916, "learning_rate": 9.727086205062676e-05, "loss": 3.1894, "step": 3680 }, { "epoch": 0.11, "grad_norm": 0.5892277956008911, "learning_rate": 9.72633038002439e-05, "loss": 3.2237, "step": 3685 }, { "epoch": 0.11, "grad_norm": 0.5302879214286804, "learning_rate": 9.725573539264994e-05, "loss": 3.2354, "step": 3690 }, { "epoch": 0.11, "grad_norm": 0.6049067378044128, "learning_rate": 9.724815682947146e-05, "loss": 3.1805, "step": 3695 }, { "epoch": 0.11, "grad_norm": 0.5331502556800842, "learning_rate": 9.72405681123371e-05, "loss": 3.1122, "step": 3700 }, { "epoch": 0.11, "grad_norm": 0.548711895942688, "learning_rate": 9.723296924287775e-05, "loss": 3.1791, "step": 3705 }, { "epoch": 0.11, "grad_norm": 0.5474268198013306, "learning_rate": 9.722536022272646e-05, "loss": 3.137, "step": 3710 }, { "epoch": 0.11, "grad_norm": 0.6364389657974243, "learning_rate": 9.721774105351845e-05, "loss": 3.1343, "step": 3715 }, { "epoch": 0.11, "grad_norm": 0.5584818124771118, "learning_rate": 9.721011173689114e-05, "loss": 3.1289, "step": 3720 }, { "epoch": 0.11, "grad_norm": 0.5809286832809448, "learning_rate": 9.720247227448411e-05, "loss": 3.2068, "step": 3725 }, { "epoch": 0.11, "grad_norm": 0.5665468573570251, "learning_rate": 9.719482266793916e-05, "loss": 3.2744, "step": 3730 }, { "epoch": 0.11, "grad_norm": 0.5712750554084778, "learning_rate": 9.718716291890022e-05, "loss": 3.1839, "step": 3735 }, { "epoch": 0.11, "grad_norm": 0.5766240954399109, "learning_rate": 9.717949302901343e-05, "loss": 3.1865, "step": 3740 }, { "epoch": 0.11, "grad_norm": 0.5582440495491028, "learning_rate": 9.71718129999271e-05, "loss": 3.1678, "step": 3745 }, { "epoch": 0.11, "grad_norm": 0.5307667255401611, "learning_rate": 9.716412283329172e-05, "loss": 3.1222, "step": 3750 }, { "epoch": 0.11, "grad_norm": 0.5278261303901672, "learning_rate": 9.715642253075998e-05, "loss": 3.1663, "step": 3755 }, { "epoch": 0.11, "grad_norm": 0.589076817035675, "learning_rate": 9.714871209398671e-05, "loss": 3.1325, "step": 3760 }, { "epoch": 0.11, "grad_norm": 0.5442926287651062, "learning_rate": 9.714099152462892e-05, "loss": 3.0989, "step": 3765 }, { "epoch": 0.11, "grad_norm": 0.536267876625061, "learning_rate": 9.713326082434584e-05, "loss": 3.1665, "step": 3770 }, { "epoch": 0.11, "grad_norm": 0.5515673160552979, "learning_rate": 9.712551999479884e-05, "loss": 3.1539, "step": 3775 }, { "epoch": 0.11, "grad_norm": 0.5835931897163391, "learning_rate": 9.71177690376515e-05, "loss": 3.1027, "step": 3780 }, { "epoch": 0.11, "grad_norm": 0.5274897217750549, "learning_rate": 9.711000795456951e-05, "loss": 3.1109, "step": 3785 }, { "epoch": 0.11, "grad_norm": 0.5360762476921082, "learning_rate": 9.710223674722083e-05, "loss": 3.1565, "step": 3790 }, { "epoch": 0.11, "grad_norm": 0.549220860004425, "learning_rate": 9.709445541727551e-05, "loss": 3.0812, "step": 3795 }, { "epoch": 0.11, "grad_norm": 0.5517001748085022, "learning_rate": 9.708666396640582e-05, "loss": 3.1122, "step": 3800 }, { "epoch": 0.11, "grad_norm": 0.5295144319534302, "learning_rate": 9.707886239628621e-05, "loss": 3.2091, "step": 3805 }, { "epoch": 0.11, "grad_norm": 0.5762179493904114, "learning_rate": 9.707105070859326e-05, "loss": 3.1404, "step": 3810 }, { "epoch": 0.11, "grad_norm": 0.5425832271575928, "learning_rate": 9.706322890500579e-05, "loss": 3.1535, "step": 3815 }, { "epoch": 0.11, "grad_norm": 0.5557460784912109, "learning_rate": 9.705539698720476e-05, "loss": 3.1723, "step": 3820 }, { "epoch": 0.11, "grad_norm": 0.5599187016487122, "learning_rate": 9.704755495687328e-05, "loss": 3.1785, "step": 3825 }, { "epoch": 0.11, "grad_norm": 0.5282739996910095, "learning_rate": 9.703970281569668e-05, "loss": 3.1275, "step": 3830 }, { "epoch": 0.11, "grad_norm": 0.6188558340072632, "learning_rate": 9.70318405653624e-05, "loss": 3.1211, "step": 3835 }, { "epoch": 0.11, "grad_norm": 0.5646204352378845, "learning_rate": 9.702396820756014e-05, "loss": 3.1653, "step": 3840 }, { "epoch": 0.11, "grad_norm": 0.5301668643951416, "learning_rate": 9.701608574398169e-05, "loss": 3.1047, "step": 3845 }, { "epoch": 0.11, "grad_norm": 0.554869532585144, "learning_rate": 9.700819317632107e-05, "loss": 3.2204, "step": 3850 }, { "epoch": 0.11, "grad_norm": 0.5582436323165894, "learning_rate": 9.700029050627443e-05, "loss": 3.1913, "step": 3855 }, { "epoch": 0.11, "grad_norm": 0.5721524953842163, "learning_rate": 9.699237773554012e-05, "loss": 3.1435, "step": 3860 }, { "epoch": 0.11, "grad_norm": 0.5872426629066467, "learning_rate": 9.698445486581864e-05, "loss": 3.1047, "step": 3865 }, { "epoch": 0.11, "grad_norm": 0.5753690004348755, "learning_rate": 9.697652189881266e-05, "loss": 3.1755, "step": 3870 }, { "epoch": 0.11, "grad_norm": 0.5664029121398926, "learning_rate": 9.696857883622706e-05, "loss": 3.2622, "step": 3875 }, { "epoch": 0.11, "grad_norm": 0.5424938797950745, "learning_rate": 9.696062567976885e-05, "loss": 3.2178, "step": 3880 }, { "epoch": 0.11, "grad_norm": 0.5263423323631287, "learning_rate": 9.695266243114718e-05, "loss": 3.245, "step": 3885 }, { "epoch": 0.11, "grad_norm": 0.5440256595611572, "learning_rate": 9.694468909207345e-05, "loss": 3.0809, "step": 3890 }, { "epoch": 0.11, "grad_norm": 0.517082154750824, "learning_rate": 9.693670566426119e-05, "loss": 3.075, "step": 3895 }, { "epoch": 0.11, "grad_norm": 0.5196235775947571, "learning_rate": 9.692871214942608e-05, "loss": 3.1222, "step": 3900 }, { "epoch": 0.11, "grad_norm": 0.550031840801239, "learning_rate": 9.692070854928596e-05, "loss": 3.2557, "step": 3905 }, { "epoch": 0.12, "grad_norm": 0.5820319056510925, "learning_rate": 9.691269486556088e-05, "loss": 3.1866, "step": 3910 }, { "epoch": 0.12, "grad_norm": 0.5211647748947144, "learning_rate": 9.690467109997303e-05, "loss": 3.1548, "step": 3915 }, { "epoch": 0.12, "grad_norm": 0.6654908657073975, "learning_rate": 9.68966372542468e-05, "loss": 3.2227, "step": 3920 }, { "epoch": 0.12, "grad_norm": 0.5360572934150696, "learning_rate": 9.688859333010866e-05, "loss": 3.1898, "step": 3925 }, { "epoch": 0.12, "grad_norm": 0.5261736512184143, "learning_rate": 9.688053932928737e-05, "loss": 3.0622, "step": 3930 }, { "epoch": 0.12, "grad_norm": 0.5612882971763611, "learning_rate": 9.687247525351375e-05, "loss": 3.1681, "step": 3935 }, { "epoch": 0.12, "grad_norm": 0.5498191714286804, "learning_rate": 9.686440110452084e-05, "loss": 3.1017, "step": 3940 }, { "epoch": 0.12, "grad_norm": 0.5876068472862244, "learning_rate": 9.685631688404384e-05, "loss": 3.2005, "step": 3945 }, { "epoch": 0.12, "grad_norm": 0.6026333570480347, "learning_rate": 9.684822259382008e-05, "loss": 3.2159, "step": 3950 }, { "epoch": 0.12, "grad_norm": 0.5814046859741211, "learning_rate": 9.684011823558909e-05, "loss": 3.1778, "step": 3955 }, { "epoch": 0.12, "grad_norm": 0.5522966384887695, "learning_rate": 9.683200381109254e-05, "loss": 3.2166, "step": 3960 }, { "epoch": 0.12, "grad_norm": 0.5432887673377991, "learning_rate": 9.68238793220743e-05, "loss": 3.0406, "step": 3965 }, { "epoch": 0.12, "grad_norm": 0.5634650588035583, "learning_rate": 9.681574477028039e-05, "loss": 3.074, "step": 3970 }, { "epoch": 0.12, "grad_norm": 0.5388665795326233, "learning_rate": 9.680760015745894e-05, "loss": 3.2079, "step": 3975 }, { "epoch": 0.12, "grad_norm": 0.5822561979293823, "learning_rate": 9.67994454853603e-05, "loss": 3.0622, "step": 3980 }, { "epoch": 0.12, "grad_norm": 0.5853989720344543, "learning_rate": 9.679128075573698e-05, "loss": 3.1689, "step": 3985 }, { "epoch": 0.12, "grad_norm": 0.7572869658470154, "learning_rate": 9.678310597034364e-05, "loss": 3.191, "step": 3990 }, { "epoch": 0.12, "grad_norm": 0.6983926296234131, "learning_rate": 9.677492113093706e-05, "loss": 3.0335, "step": 3995 }, { "epoch": 0.12, "grad_norm": 0.7749678492546082, "learning_rate": 9.676672623927625e-05, "loss": 3.1444, "step": 4000 }, { "epoch": 0.12, "grad_norm": 0.5738393068313599, "learning_rate": 9.675852129712234e-05, "loss": 3.1518, "step": 4005 }, { "epoch": 0.12, "grad_norm": 0.5945497751235962, "learning_rate": 9.675030630623864e-05, "loss": 3.1776, "step": 4010 }, { "epoch": 0.12, "grad_norm": 0.5662328600883484, "learning_rate": 9.67420812683906e-05, "loss": 3.2445, "step": 4015 }, { "epoch": 0.12, "grad_norm": 0.5791964530944824, "learning_rate": 9.673384618534582e-05, "loss": 3.0957, "step": 4020 }, { "epoch": 0.12, "grad_norm": 0.5653262138366699, "learning_rate": 9.672560105887409e-05, "loss": 3.0927, "step": 4025 }, { "epoch": 0.12, "grad_norm": 0.5353372097015381, "learning_rate": 9.671734589074736e-05, "loss": 3.1841, "step": 4030 }, { "epoch": 0.12, "grad_norm": 0.568449079990387, "learning_rate": 9.670908068273972e-05, "loss": 3.1276, "step": 4035 }, { "epoch": 0.12, "grad_norm": 0.5734129548072815, "learning_rate": 9.67008054366274e-05, "loss": 3.0176, "step": 4040 }, { "epoch": 0.12, "grad_norm": 0.6094769239425659, "learning_rate": 9.669252015418882e-05, "loss": 3.1718, "step": 4045 }, { "epoch": 0.12, "grad_norm": 0.5928288102149963, "learning_rate": 9.668422483720456e-05, "loss": 3.1693, "step": 4050 }, { "epoch": 0.12, "grad_norm": 0.5851165652275085, "learning_rate": 9.66759194874573e-05, "loss": 3.2032, "step": 4055 }, { "epoch": 0.12, "grad_norm": 0.533561646938324, "learning_rate": 9.666760410673195e-05, "loss": 3.1815, "step": 4060 }, { "epoch": 0.12, "grad_norm": 0.5719203948974609, "learning_rate": 9.665927869681553e-05, "loss": 3.2044, "step": 4065 }, { "epoch": 0.12, "grad_norm": 0.6013598442077637, "learning_rate": 9.665094325949723e-05, "loss": 3.1605, "step": 4070 }, { "epoch": 0.12, "grad_norm": 0.5698485374450684, "learning_rate": 9.664259779656842e-05, "loss": 3.0578, "step": 4075 }, { "epoch": 0.12, "grad_norm": 0.5933488011360168, "learning_rate": 9.663424230982254e-05, "loss": 3.0937, "step": 4080 }, { "epoch": 0.12, "grad_norm": 0.5215538740158081, "learning_rate": 9.66258768010553e-05, "loss": 3.1554, "step": 4085 }, { "epoch": 0.12, "grad_norm": 0.5413704514503479, "learning_rate": 9.661750127206445e-05, "loss": 3.0711, "step": 4090 }, { "epoch": 0.12, "grad_norm": 0.5175780057907104, "learning_rate": 9.660911572465e-05, "loss": 3.1479, "step": 4095 }, { "epoch": 0.12, "grad_norm": 0.5335823893547058, "learning_rate": 9.660072016061406e-05, "loss": 3.1558, "step": 4100 }, { "epoch": 0.12, "grad_norm": 0.5246891379356384, "learning_rate": 9.659231458176085e-05, "loss": 3.1747, "step": 4105 }, { "epoch": 0.12, "grad_norm": 0.5273491740226746, "learning_rate": 9.658389898989684e-05, "loss": 3.054, "step": 4110 }, { "epoch": 0.12, "grad_norm": 0.5592836141586304, "learning_rate": 9.657547338683055e-05, "loss": 3.1963, "step": 4115 }, { "epoch": 0.12, "grad_norm": 0.5795887112617493, "learning_rate": 9.656703777437273e-05, "loss": 3.1115, "step": 4120 }, { "epoch": 0.12, "grad_norm": 0.6332036256790161, "learning_rate": 9.655859215433625e-05, "loss": 3.1483, "step": 4125 }, { "epoch": 0.12, "grad_norm": 0.5233336091041565, "learning_rate": 9.655013652853612e-05, "loss": 3.1268, "step": 4130 }, { "epoch": 0.12, "grad_norm": 0.5246163010597229, "learning_rate": 9.654167089878952e-05, "loss": 3.1211, "step": 4135 }, { "epoch": 0.12, "grad_norm": 0.5878833532333374, "learning_rate": 9.653319526691577e-05, "loss": 3.0233, "step": 4140 }, { "epoch": 0.12, "grad_norm": 0.579940676689148, "learning_rate": 9.652470963473635e-05, "loss": 3.0819, "step": 4145 }, { "epoch": 0.12, "grad_norm": 0.5590419173240662, "learning_rate": 9.651621400407487e-05, "loss": 3.1047, "step": 4150 }, { "epoch": 0.12, "grad_norm": 0.5084700584411621, "learning_rate": 9.650770837675712e-05, "loss": 3.109, "step": 4155 }, { "epoch": 0.12, "grad_norm": 0.5264461040496826, "learning_rate": 9.649919275461099e-05, "loss": 3.1755, "step": 4160 }, { "epoch": 0.12, "grad_norm": 0.5388926267623901, "learning_rate": 9.649066713946655e-05, "loss": 3.0902, "step": 4165 }, { "epoch": 0.12, "grad_norm": 0.5323820114135742, "learning_rate": 9.648213153315603e-05, "loss": 3.1382, "step": 4170 }, { "epoch": 0.12, "grad_norm": 0.5286802053451538, "learning_rate": 9.647358593751378e-05, "loss": 3.0628, "step": 4175 }, { "epoch": 0.12, "grad_norm": 0.5955986380577087, "learning_rate": 9.646503035437633e-05, "loss": 3.0602, "step": 4180 }, { "epoch": 0.12, "grad_norm": 0.5460296869277954, "learning_rate": 9.64564647855823e-05, "loss": 3.0903, "step": 4185 }, { "epoch": 0.12, "grad_norm": 0.5450156927108765, "learning_rate": 9.644788923297253e-05, "loss": 3.101, "step": 4190 }, { "epoch": 0.12, "grad_norm": 0.5772224068641663, "learning_rate": 9.64393036983899e-05, "loss": 3.1046, "step": 4195 }, { "epoch": 0.12, "grad_norm": 0.5209029316902161, "learning_rate": 9.643070818367957e-05, "loss": 3.1296, "step": 4200 }, { "epoch": 0.12, "grad_norm": 0.5578837394714355, "learning_rate": 9.642210269068874e-05, "loss": 3.0736, "step": 4205 }, { "epoch": 0.12, "grad_norm": 0.5143368244171143, "learning_rate": 9.641348722126679e-05, "loss": 3.0821, "step": 4210 }, { "epoch": 0.12, "grad_norm": 0.5601150989532471, "learning_rate": 9.640486177726527e-05, "loss": 3.1234, "step": 4215 }, { "epoch": 0.12, "grad_norm": 0.6001995801925659, "learning_rate": 9.639622636053781e-05, "loss": 3.0494, "step": 4220 }, { "epoch": 0.12, "grad_norm": 0.6016066670417786, "learning_rate": 9.638758097294024e-05, "loss": 3.0418, "step": 4225 }, { "epoch": 0.12, "grad_norm": 0.5626603960990906, "learning_rate": 9.637892561633054e-05, "loss": 3.1311, "step": 4230 }, { "epoch": 0.12, "grad_norm": 0.5416538119316101, "learning_rate": 9.637026029256874e-05, "loss": 3.1109, "step": 4235 }, { "epoch": 0.12, "grad_norm": 0.5177851319313049, "learning_rate": 9.636158500351713e-05, "loss": 3.072, "step": 4240 }, { "epoch": 0.12, "grad_norm": 0.5298792719841003, "learning_rate": 9.635289975104009e-05, "loss": 3.1232, "step": 4245 }, { "epoch": 0.13, "grad_norm": 0.516472578048706, "learning_rate": 9.634420453700413e-05, "loss": 3.1575, "step": 4250 }, { "epoch": 0.13, "grad_norm": 0.6039904952049255, "learning_rate": 9.633549936327789e-05, "loss": 3.0745, "step": 4255 }, { "epoch": 0.13, "grad_norm": 0.6696873307228088, "learning_rate": 9.63267842317322e-05, "loss": 3.1115, "step": 4260 }, { "epoch": 0.13, "grad_norm": 0.6165772080421448, "learning_rate": 9.631805914424e-05, "loss": 3.082, "step": 4265 }, { "epoch": 0.13, "grad_norm": 0.569606602191925, "learning_rate": 9.630932410267637e-05, "loss": 3.0665, "step": 4270 }, { "epoch": 0.13, "grad_norm": 0.5298318266868591, "learning_rate": 9.630057910891852e-05, "loss": 3.0672, "step": 4275 }, { "epoch": 0.13, "grad_norm": 0.5888398885726929, "learning_rate": 9.629182416484582e-05, "loss": 3.1262, "step": 4280 }, { "epoch": 0.13, "grad_norm": 0.5507888793945312, "learning_rate": 9.628305927233975e-05, "loss": 3.173, "step": 4285 }, { "epoch": 0.13, "grad_norm": 0.5697919130325317, "learning_rate": 9.627428443328397e-05, "loss": 3.0368, "step": 4290 }, { "epoch": 0.13, "grad_norm": 0.5659337639808655, "learning_rate": 9.626549964956424e-05, "loss": 3.1766, "step": 4295 }, { "epoch": 0.13, "grad_norm": 0.569694459438324, "learning_rate": 9.625670492306848e-05, "loss": 3.2081, "step": 4300 }, { "epoch": 0.13, "grad_norm": 0.5750781297683716, "learning_rate": 9.624790025568673e-05, "loss": 3.1664, "step": 4305 }, { "epoch": 0.13, "grad_norm": 0.5014953017234802, "learning_rate": 9.623908564931117e-05, "loss": 3.1979, "step": 4310 }, { "epoch": 0.13, "grad_norm": 0.5211942195892334, "learning_rate": 9.623026110583612e-05, "loss": 3.0566, "step": 4315 }, { "epoch": 0.13, "grad_norm": 0.5310171842575073, "learning_rate": 9.622142662715808e-05, "loss": 3.104, "step": 4320 }, { "epoch": 0.13, "grad_norm": 0.5478355884552002, "learning_rate": 9.621258221517556e-05, "loss": 3.057, "step": 4325 }, { "epoch": 0.13, "grad_norm": 0.9798442721366882, "learning_rate": 9.620372787178934e-05, "loss": 3.0448, "step": 4330 }, { "epoch": 0.13, "grad_norm": 0.5133621096611023, "learning_rate": 9.619486359890228e-05, "loss": 3.0456, "step": 4335 }, { "epoch": 0.13, "grad_norm": 0.5344738960266113, "learning_rate": 9.618598939841934e-05, "loss": 3.0378, "step": 4340 }, { "epoch": 0.13, "grad_norm": 0.5612080097198486, "learning_rate": 9.617710527224768e-05, "loss": 3.054, "step": 4345 }, { "epoch": 0.13, "grad_norm": 0.5906364321708679, "learning_rate": 9.616821122229656e-05, "loss": 3.0759, "step": 4350 }, { "epoch": 0.13, "grad_norm": 0.5263745188713074, "learning_rate": 9.615930725047732e-05, "loss": 3.0693, "step": 4355 }, { "epoch": 0.13, "grad_norm": 0.5237680077552795, "learning_rate": 9.615039335870357e-05, "loss": 3.1847, "step": 4360 }, { "epoch": 0.13, "grad_norm": 0.5503563284873962, "learning_rate": 9.614146954889091e-05, "loss": 3.142, "step": 4365 }, { "epoch": 0.13, "grad_norm": 10.55534553527832, "learning_rate": 9.613253582295714e-05, "loss": 3.1331, "step": 4370 }, { "epoch": 0.13, "grad_norm": 0.5358182191848755, "learning_rate": 9.612359218282216e-05, "loss": 3.0146, "step": 4375 }, { "epoch": 0.13, "grad_norm": 0.520017147064209, "learning_rate": 9.611463863040807e-05, "loss": 3.0867, "step": 4380 }, { "epoch": 0.13, "grad_norm": 0.7102450728416443, "learning_rate": 9.610567516763902e-05, "loss": 3.0605, "step": 4385 }, { "epoch": 0.13, "grad_norm": 0.5345319509506226, "learning_rate": 9.60967017964413e-05, "loss": 3.1008, "step": 4390 }, { "epoch": 0.13, "grad_norm": 0.5229172110557556, "learning_rate": 9.608771851874338e-05, "loss": 3.1201, "step": 4395 }, { "epoch": 0.13, "grad_norm": 0.5421702265739441, "learning_rate": 9.607872533647584e-05, "loss": 3.0912, "step": 4400 }, { "epoch": 0.13, "grad_norm": 0.5800228118896484, "learning_rate": 9.606972225157134e-05, "loss": 3.1037, "step": 4405 }, { "epoch": 0.13, "grad_norm": 0.5550212264060974, "learning_rate": 9.606070926596471e-05, "loss": 3.062, "step": 4410 }, { "epoch": 0.13, "grad_norm": 0.5273239016532898, "learning_rate": 9.605168638159294e-05, "loss": 3.0848, "step": 4415 }, { "epoch": 0.13, "grad_norm": 0.5022825002670288, "learning_rate": 9.604265360039507e-05, "loss": 3.0243, "step": 4420 }, { "epoch": 0.13, "grad_norm": 0.5575874447822571, "learning_rate": 9.603361092431234e-05, "loss": 3.1242, "step": 4425 }, { "epoch": 0.13, "grad_norm": 0.5153757333755493, "learning_rate": 9.602455835528804e-05, "loss": 3.0194, "step": 4430 }, { "epoch": 0.13, "grad_norm": 0.5221821069717407, "learning_rate": 9.601549589526767e-05, "loss": 3.0332, "step": 4435 }, { "epoch": 0.13, "grad_norm": 0.5272677540779114, "learning_rate": 9.60064235461988e-05, "loss": 3.1566, "step": 4440 }, { "epoch": 0.13, "grad_norm": 0.5317149758338928, "learning_rate": 9.599734131003114e-05, "loss": 3.0359, "step": 4445 }, { "epoch": 0.13, "grad_norm": 0.5088919997215271, "learning_rate": 9.598824918871654e-05, "loss": 3.1633, "step": 4450 }, { "epoch": 0.13, "grad_norm": 0.531982421875, "learning_rate": 9.597914718420892e-05, "loss": 3.1995, "step": 4455 }, { "epoch": 0.13, "grad_norm": 0.5027827620506287, "learning_rate": 9.597003529846442e-05, "loss": 3.0674, "step": 4460 }, { "epoch": 0.13, "grad_norm": 0.5019488334655762, "learning_rate": 9.59609135334412e-05, "loss": 3.1799, "step": 4465 }, { "epoch": 0.13, "grad_norm": 0.5120329260826111, "learning_rate": 9.595178189109963e-05, "loss": 2.9769, "step": 4470 }, { "epoch": 0.13, "grad_norm": 2.57012677192688, "learning_rate": 9.594264037340212e-05, "loss": 3.0765, "step": 4475 }, { "epoch": 0.13, "grad_norm": 0.5074518322944641, "learning_rate": 9.593348898231326e-05, "loss": 3.0489, "step": 4480 }, { "epoch": 0.13, "grad_norm": 0.5302559733390808, "learning_rate": 9.592432771979978e-05, "loss": 3.1278, "step": 4485 }, { "epoch": 0.13, "grad_norm": 0.4837750494480133, "learning_rate": 9.591515658783046e-05, "loss": 3.0984, "step": 4490 }, { "epoch": 0.13, "grad_norm": 0.5362811088562012, "learning_rate": 9.590597558837626e-05, "loss": 3.1047, "step": 4495 }, { "epoch": 0.13, "grad_norm": 0.5272698998451233, "learning_rate": 9.589678472341023e-05, "loss": 3.1382, "step": 4500 }, { "epoch": 0.13, "grad_norm": 0.5020481944084167, "learning_rate": 9.588758399490755e-05, "loss": 3.1039, "step": 4505 }, { "epoch": 0.13, "grad_norm": 0.5235087871551514, "learning_rate": 9.587837340484553e-05, "loss": 2.948, "step": 4510 }, { "epoch": 0.13, "grad_norm": 0.514712929725647, "learning_rate": 9.586915295520358e-05, "loss": 3.0893, "step": 4515 }, { "epoch": 0.13, "grad_norm": 0.5612702369689941, "learning_rate": 9.585992264796324e-05, "loss": 3.1393, "step": 4520 }, { "epoch": 0.13, "grad_norm": 0.5178195834159851, "learning_rate": 9.585068248510817e-05, "loss": 3.0669, "step": 4525 }, { "epoch": 0.13, "grad_norm": 0.5152338147163391, "learning_rate": 9.584143246862414e-05, "loss": 2.9912, "step": 4530 }, { "epoch": 0.13, "grad_norm": 0.5524663329124451, "learning_rate": 9.583217260049907e-05, "loss": 3.1318, "step": 4535 }, { "epoch": 0.13, "grad_norm": 0.5060458183288574, "learning_rate": 9.582290288272291e-05, "loss": 3.0143, "step": 4540 }, { "epoch": 0.13, "grad_norm": 0.5535380244255066, "learning_rate": 9.581362331728784e-05, "loss": 3.1352, "step": 4545 }, { "epoch": 0.13, "grad_norm": 0.5326555967330933, "learning_rate": 9.580433390618808e-05, "loss": 3.1001, "step": 4550 }, { "epoch": 0.13, "grad_norm": 0.5096615552902222, "learning_rate": 9.579503465142e-05, "loss": 3.0501, "step": 4555 }, { "epoch": 0.13, "grad_norm": 0.5706288814544678, "learning_rate": 9.578572555498207e-05, "loss": 3.0881, "step": 4560 }, { "epoch": 0.13, "grad_norm": 0.5220559239387512, "learning_rate": 9.577640661887488e-05, "loss": 3.0869, "step": 4565 }, { "epoch": 0.13, "grad_norm": 0.5586305260658264, "learning_rate": 9.576707784510113e-05, "loss": 3.0696, "step": 4570 }, { "epoch": 0.13, "grad_norm": 0.5194656252861023, "learning_rate": 9.575773923566563e-05, "loss": 3.135, "step": 4575 }, { "epoch": 0.13, "grad_norm": 0.5799388885498047, "learning_rate": 9.574839079257532e-05, "loss": 3.0004, "step": 4580 }, { "epoch": 0.13, "grad_norm": 0.5133780837059021, "learning_rate": 9.573903251783925e-05, "loss": 3.1584, "step": 4585 }, { "epoch": 0.14, "grad_norm": 0.5471716523170471, "learning_rate": 9.572966441346858e-05, "loss": 3.0835, "step": 4590 }, { "epoch": 0.14, "grad_norm": 0.5336796641349792, "learning_rate": 9.572028648147657e-05, "loss": 3.095, "step": 4595 }, { "epoch": 0.14, "grad_norm": 0.5175730586051941, "learning_rate": 9.571089872387862e-05, "loss": 3.0703, "step": 4600 }, { "epoch": 0.14, "grad_norm": 0.5207660794258118, "learning_rate": 9.57015011426922e-05, "loss": 3.0941, "step": 4605 }, { "epoch": 0.14, "grad_norm": 0.5199393630027771, "learning_rate": 9.569209373993693e-05, "loss": 3.2183, "step": 4610 }, { "epoch": 0.14, "grad_norm": 0.5576893091201782, "learning_rate": 9.568267651763451e-05, "loss": 3.1681, "step": 4615 }, { "epoch": 0.14, "grad_norm": 0.5399788022041321, "learning_rate": 9.56732494778088e-05, "loss": 3.1425, "step": 4620 }, { "epoch": 0.14, "grad_norm": 0.5213262438774109, "learning_rate": 9.56638126224857e-05, "loss": 3.1954, "step": 4625 }, { "epoch": 0.14, "grad_norm": 0.5174756050109863, "learning_rate": 9.565436595369328e-05, "loss": 3.0734, "step": 4630 }, { "epoch": 0.14, "grad_norm": 0.5284809470176697, "learning_rate": 9.564490947346168e-05, "loss": 3.0284, "step": 4635 }, { "epoch": 0.14, "grad_norm": 0.5150734186172485, "learning_rate": 9.563544318382316e-05, "loss": 3.1912, "step": 4640 }, { "epoch": 0.14, "grad_norm": 0.5697650909423828, "learning_rate": 9.56259670868121e-05, "loss": 3.0384, "step": 4645 }, { "epoch": 0.14, "grad_norm": 0.543328046798706, "learning_rate": 9.5616481184465e-05, "loss": 3.0214, "step": 4650 }, { "epoch": 0.14, "grad_norm": 0.5958728194236755, "learning_rate": 9.560698547882041e-05, "loss": 3.0367, "step": 4655 }, { "epoch": 0.14, "grad_norm": 0.4995012581348419, "learning_rate": 9.559747997191903e-05, "loss": 3.007, "step": 4660 }, { "epoch": 0.14, "grad_norm": 0.5143396258354187, "learning_rate": 9.558796466580366e-05, "loss": 3.0162, "step": 4665 }, { "epoch": 0.14, "grad_norm": 0.5007093548774719, "learning_rate": 9.557843956251924e-05, "loss": 3.068, "step": 4670 }, { "epoch": 0.14, "grad_norm": 0.5443606376647949, "learning_rate": 9.556890466411272e-05, "loss": 3.0729, "step": 4675 }, { "epoch": 0.14, "grad_norm": 0.5251206755638123, "learning_rate": 9.555935997263328e-05, "loss": 3.1282, "step": 4680 }, { "epoch": 0.14, "grad_norm": 0.5012190937995911, "learning_rate": 9.554980549013209e-05, "loss": 3.1122, "step": 4685 }, { "epoch": 0.14, "grad_norm": 0.5653667449951172, "learning_rate": 9.554024121866249e-05, "loss": 3.1276, "step": 4690 }, { "epoch": 0.14, "grad_norm": 0.4983196258544922, "learning_rate": 9.55306671602799e-05, "loss": 3.0505, "step": 4695 }, { "epoch": 0.14, "grad_norm": 0.49600112438201904, "learning_rate": 9.552108331704189e-05, "loss": 3.1358, "step": 4700 }, { "epoch": 0.14, "grad_norm": 0.49811333417892456, "learning_rate": 9.551148969100804e-05, "loss": 3.0744, "step": 4705 }, { "epoch": 0.14, "grad_norm": 0.5215647220611572, "learning_rate": 9.550188628424012e-05, "loss": 3.0659, "step": 4710 }, { "epoch": 0.14, "grad_norm": 0.5546525716781616, "learning_rate": 9.549227309880197e-05, "loss": 3.1065, "step": 4715 }, { "epoch": 0.14, "grad_norm": 0.5639951825141907, "learning_rate": 9.548265013675951e-05, "loss": 3.1515, "step": 4720 }, { "epoch": 0.14, "grad_norm": 0.5374469757080078, "learning_rate": 9.54730174001808e-05, "loss": 3.0651, "step": 4725 }, { "epoch": 0.14, "grad_norm": 0.5137919783592224, "learning_rate": 9.546337489113597e-05, "loss": 3.142, "step": 4730 }, { "epoch": 0.14, "grad_norm": 0.5165963768959045, "learning_rate": 9.545372261169725e-05, "loss": 3.1697, "step": 4735 }, { "epoch": 0.14, "grad_norm": 0.5428669452667236, "learning_rate": 9.544406056393903e-05, "loss": 3.0201, "step": 4740 }, { "epoch": 0.14, "grad_norm": 0.5309314131736755, "learning_rate": 9.54343887499377e-05, "loss": 3.0544, "step": 4745 }, { "epoch": 0.14, "grad_norm": 0.5170466303825378, "learning_rate": 9.542470717177182e-05, "loss": 2.9631, "step": 4750 }, { "epoch": 0.14, "grad_norm": 0.5819863677024841, "learning_rate": 9.541501583152204e-05, "loss": 3.1027, "step": 4755 }, { "epoch": 0.14, "grad_norm": 0.5361101031303406, "learning_rate": 9.540531473127106e-05, "loss": 3.038, "step": 4760 }, { "epoch": 0.14, "grad_norm": 0.5083801746368408, "learning_rate": 9.539560387310375e-05, "loss": 3.0989, "step": 4765 }, { "epoch": 0.14, "grad_norm": 0.5344492793083191, "learning_rate": 9.538588325910702e-05, "loss": 3.1359, "step": 4770 }, { "epoch": 0.14, "grad_norm": 0.4992101788520813, "learning_rate": 9.53761528913699e-05, "loss": 3.0081, "step": 4775 }, { "epoch": 0.14, "grad_norm": 0.5070109367370605, "learning_rate": 9.536641277198353e-05, "loss": 3.0443, "step": 4780 }, { "epoch": 0.14, "grad_norm": 0.5071985125541687, "learning_rate": 9.535666290304111e-05, "loss": 3.1087, "step": 4785 }, { "epoch": 0.14, "grad_norm": 0.5037854313850403, "learning_rate": 9.534690328663796e-05, "loss": 2.9498, "step": 4790 }, { "epoch": 0.14, "grad_norm": 0.5272377729415894, "learning_rate": 9.533713392487149e-05, "loss": 3.0317, "step": 4795 }, { "epoch": 0.14, "grad_norm": 0.49802547693252563, "learning_rate": 9.532735481984121e-05, "loss": 3.1561, "step": 4800 }, { "epoch": 0.14, "grad_norm": 0.4864078462123871, "learning_rate": 9.53175659736487e-05, "loss": 3.0902, "step": 4805 }, { "epoch": 0.14, "grad_norm": 0.5195857882499695, "learning_rate": 9.530776738839766e-05, "loss": 3.0474, "step": 4810 }, { "epoch": 0.14, "grad_norm": 0.5044553279876709, "learning_rate": 9.529795906619387e-05, "loss": 2.9877, "step": 4815 }, { "epoch": 0.14, "grad_norm": 0.5671506524085999, "learning_rate": 9.528814100914522e-05, "loss": 3.1365, "step": 4820 }, { "epoch": 0.14, "grad_norm": 0.5479797720909119, "learning_rate": 9.52783132193617e-05, "loss": 3.0146, "step": 4825 }, { "epoch": 0.14, "grad_norm": 0.5190348029136658, "learning_rate": 9.52684756989553e-05, "loss": 3.0364, "step": 4830 }, { "epoch": 0.14, "grad_norm": 0.6265324950218201, "learning_rate": 9.525862845004022e-05, "loss": 3.118, "step": 4835 }, { "epoch": 0.14, "grad_norm": 0.5010820031166077, "learning_rate": 9.52487714747327e-05, "loss": 2.8929, "step": 4840 }, { "epoch": 0.14, "grad_norm": 0.514413595199585, "learning_rate": 9.523890477515105e-05, "loss": 3.0982, "step": 4845 }, { "epoch": 0.14, "grad_norm": 0.7186405658721924, "learning_rate": 9.522902835341573e-05, "loss": 3.1508, "step": 4850 }, { "epoch": 0.14, "grad_norm": 1.1123183965682983, "learning_rate": 9.521914221164922e-05, "loss": 3.0503, "step": 4855 }, { "epoch": 0.14, "grad_norm": 0.47867465019226074, "learning_rate": 9.520924635197612e-05, "loss": 3.0979, "step": 4860 }, { "epoch": 0.14, "grad_norm": 0.49478164315223694, "learning_rate": 9.519934077652312e-05, "loss": 3.055, "step": 4865 }, { "epoch": 0.14, "grad_norm": 0.49822601675987244, "learning_rate": 9.518942548741902e-05, "loss": 3.0725, "step": 4870 }, { "epoch": 0.14, "grad_norm": 0.49988454580307007, "learning_rate": 9.517950048679467e-05, "loss": 3.0935, "step": 4875 }, { "epoch": 0.14, "grad_norm": 0.531676709651947, "learning_rate": 9.516956577678302e-05, "loss": 3.0301, "step": 4880 }, { "epoch": 0.14, "grad_norm": 0.5330474972724915, "learning_rate": 9.51596213595191e-05, "loss": 3.0504, "step": 4885 }, { "epoch": 0.14, "grad_norm": 0.521399736404419, "learning_rate": 9.514966723714004e-05, "loss": 3.0831, "step": 4890 }, { "epoch": 0.14, "grad_norm": 0.5652825236320496, "learning_rate": 9.513970341178506e-05, "loss": 2.9182, "step": 4895 }, { "epoch": 0.14, "grad_norm": 0.5362108945846558, "learning_rate": 9.512972988559545e-05, "loss": 3.0093, "step": 4900 }, { "epoch": 0.14, "grad_norm": 0.5451188683509827, "learning_rate": 9.511974666071458e-05, "loss": 3.0191, "step": 4905 }, { "epoch": 0.14, "grad_norm": 0.9234301447868347, "learning_rate": 9.510975373928792e-05, "loss": 2.9501, "step": 4910 }, { "epoch": 0.14, "grad_norm": 0.5368677377700806, "learning_rate": 9.509975112346303e-05, "loss": 2.9601, "step": 4915 }, { "epoch": 0.14, "grad_norm": 0.5041046738624573, "learning_rate": 9.508973881538952e-05, "loss": 3.0482, "step": 4920 }, { "epoch": 0.14, "grad_norm": 0.5924731492996216, "learning_rate": 9.507971681721911e-05, "loss": 3.0461, "step": 4925 }, { "epoch": 0.15, "grad_norm": 0.5285022258758545, "learning_rate": 9.50696851311056e-05, "loss": 2.9217, "step": 4930 }, { "epoch": 0.15, "grad_norm": 0.4870918095111847, "learning_rate": 9.505964375920489e-05, "loss": 2.9827, "step": 4935 }, { "epoch": 0.15, "grad_norm": 0.5939784646034241, "learning_rate": 9.50495927036749e-05, "loss": 3.07, "step": 4940 }, { "epoch": 0.15, "grad_norm": 0.4943069517612457, "learning_rate": 9.503953196667569e-05, "loss": 3.0191, "step": 4945 }, { "epoch": 0.15, "grad_norm": 0.5104547142982483, "learning_rate": 9.502946155036938e-05, "loss": 3.0239, "step": 4950 }, { "epoch": 0.15, "grad_norm": 0.5070548057556152, "learning_rate": 9.50193814569202e-05, "loss": 3.0437, "step": 4955 }, { "epoch": 0.15, "grad_norm": 0.5032899379730225, "learning_rate": 9.500929168849439e-05, "loss": 3.1372, "step": 4960 }, { "epoch": 0.15, "grad_norm": 0.5015889406204224, "learning_rate": 9.499919224726032e-05, "loss": 3.0354, "step": 4965 }, { "epoch": 0.15, "grad_norm": 0.5241037011146545, "learning_rate": 9.498908313538846e-05, "loss": 3.0369, "step": 4970 }, { "epoch": 0.15, "grad_norm": 0.522500216960907, "learning_rate": 9.49789643550513e-05, "loss": 3.0678, "step": 4975 }, { "epoch": 0.15, "grad_norm": 0.5252301096916199, "learning_rate": 9.496883590842343e-05, "loss": 3.0547, "step": 4980 }, { "epoch": 0.15, "grad_norm": 0.5270678997039795, "learning_rate": 9.495869779768154e-05, "loss": 3.0568, "step": 4985 }, { "epoch": 0.15, "grad_norm": 0.528436541557312, "learning_rate": 9.494855002500438e-05, "loss": 3.0113, "step": 4990 }, { "epoch": 0.15, "grad_norm": 0.5054397583007812, "learning_rate": 9.493839259257278e-05, "loss": 3.102, "step": 4995 }, { "epoch": 0.15, "grad_norm": 0.5004453063011169, "learning_rate": 9.492822550256962e-05, "loss": 3.1075, "step": 5000 }, { "epoch": 0.15, "grad_norm": 0.4878995418548584, "learning_rate": 9.49180487571799e-05, "loss": 3.0219, "step": 5005 }, { "epoch": 0.15, "grad_norm": 0.4876302480697632, "learning_rate": 9.490786235859066e-05, "loss": 3.0332, "step": 5010 }, { "epoch": 0.15, "grad_norm": 0.5370482206344604, "learning_rate": 9.489766630899104e-05, "loss": 3.0488, "step": 5015 }, { "epoch": 0.15, "grad_norm": 0.49171024560928345, "learning_rate": 9.488746061057223e-05, "loss": 3.02, "step": 5020 }, { "epoch": 0.15, "grad_norm": 0.4974074959754944, "learning_rate": 9.487724526552753e-05, "loss": 3.0404, "step": 5025 }, { "epoch": 0.15, "grad_norm": 0.5005912780761719, "learning_rate": 9.486702027605224e-05, "loss": 3.0616, "step": 5030 }, { "epoch": 0.15, "grad_norm": 0.5206892490386963, "learning_rate": 9.485678564434381e-05, "loss": 3.0684, "step": 5035 }, { "epoch": 0.15, "grad_norm": 0.5634109973907471, "learning_rate": 9.484654137260176e-05, "loss": 3.0666, "step": 5040 }, { "epoch": 0.15, "grad_norm": 0.5925890803337097, "learning_rate": 9.483628746302761e-05, "loss": 3.0616, "step": 5045 }, { "epoch": 0.15, "grad_norm": 0.5095767974853516, "learning_rate": 9.482602391782501e-05, "loss": 3.043, "step": 5050 }, { "epoch": 0.15, "grad_norm": 0.5136855840682983, "learning_rate": 9.481575073919968e-05, "loss": 3.0937, "step": 5055 }, { "epoch": 0.15, "grad_norm": 0.5081953406333923, "learning_rate": 9.480546792935939e-05, "loss": 3.1419, "step": 5060 }, { "epoch": 0.15, "grad_norm": 0.5051447153091431, "learning_rate": 9.479517549051398e-05, "loss": 2.9941, "step": 5065 }, { "epoch": 0.15, "grad_norm": 0.5147465467453003, "learning_rate": 9.478487342487536e-05, "loss": 2.9424, "step": 5070 }, { "epoch": 0.15, "grad_norm": 1.177407145500183, "learning_rate": 9.477456173465754e-05, "loss": 3.1209, "step": 5075 }, { "epoch": 0.15, "grad_norm": 0.5322585105895996, "learning_rate": 9.476424042207655e-05, "loss": 2.9618, "step": 5080 }, { "epoch": 0.15, "grad_norm": 0.4936616122722626, "learning_rate": 9.475390948935054e-05, "loss": 3.0777, "step": 5085 }, { "epoch": 0.15, "grad_norm": 0.5346768498420715, "learning_rate": 9.474356893869967e-05, "loss": 3.0421, "step": 5090 }, { "epoch": 0.15, "grad_norm": 0.4830736815929413, "learning_rate": 9.47332187723462e-05, "loss": 3.0257, "step": 5095 }, { "epoch": 0.15, "grad_norm": 0.49701055884361267, "learning_rate": 9.472285899251445e-05, "loss": 3.0261, "step": 5100 }, { "epoch": 0.15, "grad_norm": 0.48945847153663635, "learning_rate": 9.471248960143085e-05, "loss": 3.1045, "step": 5105 }, { "epoch": 0.15, "grad_norm": 0.474223256111145, "learning_rate": 9.47021106013238e-05, "loss": 3.0318, "step": 5110 }, { "epoch": 0.15, "grad_norm": 0.4993312954902649, "learning_rate": 9.469172199442383e-05, "loss": 3.09, "step": 5115 }, { "epoch": 0.15, "grad_norm": 0.5077476501464844, "learning_rate": 9.468132378296355e-05, "loss": 2.9892, "step": 5120 }, { "epoch": 0.15, "grad_norm": 0.4965546131134033, "learning_rate": 9.467091596917758e-05, "loss": 3.0535, "step": 5125 }, { "epoch": 0.15, "grad_norm": 0.5162185430526733, "learning_rate": 9.466049855530265e-05, "loss": 3.0457, "step": 5130 }, { "epoch": 0.15, "grad_norm": 0.5104540586471558, "learning_rate": 9.465007154357752e-05, "loss": 3.0103, "step": 5135 }, { "epoch": 0.15, "grad_norm": 0.5045089721679688, "learning_rate": 9.463963493624304e-05, "loss": 3.0413, "step": 5140 }, { "epoch": 0.15, "grad_norm": 0.4799700975418091, "learning_rate": 9.462918873554209e-05, "loss": 2.984, "step": 5145 }, { "epoch": 0.15, "grad_norm": 0.5253822207450867, "learning_rate": 9.461873294371965e-05, "loss": 3.1016, "step": 5150 }, { "epoch": 0.15, "grad_norm": 0.5181733965873718, "learning_rate": 9.460826756302275e-05, "loss": 2.9799, "step": 5155 }, { "epoch": 0.15, "grad_norm": 0.4755523204803467, "learning_rate": 9.459779259570045e-05, "loss": 3.0934, "step": 5160 }, { "epoch": 0.15, "grad_norm": 0.5031744837760925, "learning_rate": 9.458730804400389e-05, "loss": 3.1306, "step": 5165 }, { "epoch": 0.15, "grad_norm": 0.506716787815094, "learning_rate": 9.457681391018629e-05, "loss": 3.0827, "step": 5170 }, { "epoch": 0.15, "grad_norm": 0.5016564726829529, "learning_rate": 9.45663101965029e-05, "loss": 2.9219, "step": 5175 }, { "epoch": 0.15, "grad_norm": 0.728834331035614, "learning_rate": 9.455579690521106e-05, "loss": 2.9994, "step": 5180 }, { "epoch": 0.15, "grad_norm": 0.5111843347549438, "learning_rate": 9.454527403857013e-05, "loss": 3.0316, "step": 5185 }, { "epoch": 0.15, "grad_norm": 0.49837398529052734, "learning_rate": 9.453474159884155e-05, "loss": 3.0739, "step": 5190 }, { "epoch": 0.15, "grad_norm": 0.4847373068332672, "learning_rate": 9.452419958828883e-05, "loss": 3.0981, "step": 5195 }, { "epoch": 0.15, "grad_norm": 0.4936524033546448, "learning_rate": 9.45136480091775e-05, "loss": 2.8664, "step": 5200 }, { "epoch": 0.15, "grad_norm": 0.49032071232795715, "learning_rate": 9.450308686377516e-05, "loss": 2.959, "step": 5205 }, { "epoch": 0.15, "grad_norm": 0.4707845151424408, "learning_rate": 9.449251615435151e-05, "loss": 3.0518, "step": 5210 }, { "epoch": 0.15, "grad_norm": 0.5165637731552124, "learning_rate": 9.448193588317822e-05, "loss": 3.125, "step": 5215 }, { "epoch": 0.15, "grad_norm": 0.48747512698173523, "learning_rate": 9.447134605252912e-05, "loss": 3.0576, "step": 5220 }, { "epoch": 0.15, "grad_norm": 0.48917874693870544, "learning_rate": 9.446074666468e-05, "loss": 3.0169, "step": 5225 }, { "epoch": 0.15, "grad_norm": 0.5593256950378418, "learning_rate": 9.445013772190874e-05, "loss": 2.9819, "step": 5230 }, { "epoch": 0.15, "grad_norm": 0.5486271977424622, "learning_rate": 9.443951922649529e-05, "loss": 3.0284, "step": 5235 }, { "epoch": 0.15, "grad_norm": 0.5279741287231445, "learning_rate": 9.442889118072162e-05, "loss": 3.087, "step": 5240 }, { "epoch": 0.15, "grad_norm": 0.5427518486976624, "learning_rate": 9.441825358687181e-05, "loss": 2.9782, "step": 5245 }, { "epoch": 0.15, "grad_norm": 0.48713958263397217, "learning_rate": 9.440760644723192e-05, "loss": 2.9578, "step": 5250 }, { "epoch": 0.15, "grad_norm": 0.4994124174118042, "learning_rate": 9.439694976409008e-05, "loss": 2.9965, "step": 5255 }, { "epoch": 0.15, "grad_norm": 0.5118008852005005, "learning_rate": 9.438628353973653e-05, "loss": 3.0438, "step": 5260 }, { "epoch": 0.15, "grad_norm": 0.46790313720703125, "learning_rate": 9.437560777646348e-05, "loss": 3.0036, "step": 5265 }, { "epoch": 0.16, "grad_norm": 0.49292781949043274, "learning_rate": 9.436492247656523e-05, "loss": 3.0133, "step": 5270 }, { "epoch": 0.16, "grad_norm": 0.4969801902770996, "learning_rate": 9.435422764233814e-05, "loss": 3.0423, "step": 5275 }, { "epoch": 0.16, "grad_norm": 0.5201324820518494, "learning_rate": 9.43435232760806e-05, "loss": 3.0226, "step": 5280 }, { "epoch": 0.16, "grad_norm": 0.5249904990196228, "learning_rate": 9.433280938009304e-05, "loss": 3.0765, "step": 5285 }, { "epoch": 0.16, "grad_norm": 0.526862382888794, "learning_rate": 9.432208595667797e-05, "loss": 3.0021, "step": 5290 }, { "epoch": 0.16, "grad_norm": 0.5065636038780212, "learning_rate": 9.431135300813989e-05, "loss": 3.0139, "step": 5295 }, { "epoch": 0.16, "grad_norm": 0.5190463066101074, "learning_rate": 9.430061053678542e-05, "loss": 2.9892, "step": 5300 }, { "epoch": 0.16, "grad_norm": 0.5004535913467407, "learning_rate": 9.42898585449232e-05, "loss": 3.0415, "step": 5305 }, { "epoch": 0.16, "grad_norm": 0.4874188005924225, "learning_rate": 9.427909703486388e-05, "loss": 2.9507, "step": 5310 }, { "epoch": 0.16, "grad_norm": 0.5044699311256409, "learning_rate": 9.426832600892019e-05, "loss": 3.09, "step": 5315 }, { "epoch": 0.16, "grad_norm": 0.5577706098556519, "learning_rate": 9.425754546940691e-05, "loss": 3.0043, "step": 5320 }, { "epoch": 0.16, "grad_norm": 0.5886462926864624, "learning_rate": 9.424675541864081e-05, "loss": 3.1281, "step": 5325 }, { "epoch": 0.16, "grad_norm": 0.5398657917976379, "learning_rate": 9.423595585894082e-05, "loss": 3.1151, "step": 5330 }, { "epoch": 0.16, "grad_norm": 0.5169557929039001, "learning_rate": 9.422514679262778e-05, "loss": 2.9705, "step": 5335 }, { "epoch": 0.16, "grad_norm": 0.48493072390556335, "learning_rate": 9.421432822202465e-05, "loss": 3.0334, "step": 5340 }, { "epoch": 0.16, "grad_norm": 0.5041825175285339, "learning_rate": 9.420350014945643e-05, "loss": 3.0315, "step": 5345 }, { "epoch": 0.16, "grad_norm": 0.4795912802219391, "learning_rate": 9.419266257725013e-05, "loss": 3.0659, "step": 5350 }, { "epoch": 0.16, "grad_norm": 0.47209104895591736, "learning_rate": 9.418181550773483e-05, "loss": 2.9666, "step": 5355 }, { "epoch": 0.16, "grad_norm": 0.48331135511398315, "learning_rate": 9.417095894324163e-05, "loss": 3.1246, "step": 5360 }, { "epoch": 0.16, "grad_norm": 0.5206527709960938, "learning_rate": 9.41600928861037e-05, "loss": 3.0521, "step": 5365 }, { "epoch": 0.16, "grad_norm": 0.5320144295692444, "learning_rate": 9.41492173386562e-05, "loss": 3.0476, "step": 5370 }, { "epoch": 0.16, "grad_norm": 0.9905712604522705, "learning_rate": 9.413833230323639e-05, "loss": 3.0389, "step": 5375 }, { "epoch": 0.16, "grad_norm": 0.5034389495849609, "learning_rate": 9.412743778218351e-05, "loss": 2.9838, "step": 5380 }, { "epoch": 0.16, "grad_norm": 0.5254191160202026, "learning_rate": 9.41165337778389e-05, "loss": 2.9822, "step": 5385 }, { "epoch": 0.16, "grad_norm": 0.527946412563324, "learning_rate": 9.410562029254589e-05, "loss": 2.918, "step": 5390 }, { "epoch": 0.16, "grad_norm": 0.5893475413322449, "learning_rate": 9.409469732864986e-05, "loss": 2.9719, "step": 5395 }, { "epoch": 0.16, "grad_norm": 0.4664258360862732, "learning_rate": 9.408376488849824e-05, "loss": 2.96, "step": 5400 }, { "epoch": 0.16, "grad_norm": 0.5170448422431946, "learning_rate": 9.407282297444048e-05, "loss": 3.0929, "step": 5405 }, { "epoch": 0.16, "grad_norm": 0.5280725955963135, "learning_rate": 9.406187158882808e-05, "loss": 2.9791, "step": 5410 }, { "epoch": 0.16, "grad_norm": 0.5107510089874268, "learning_rate": 9.405091073401458e-05, "loss": 2.9599, "step": 5415 }, { "epoch": 0.16, "grad_norm": 0.5203644633293152, "learning_rate": 9.40399404123555e-05, "loss": 3.0596, "step": 5420 }, { "epoch": 0.16, "grad_norm": 0.4728945195674896, "learning_rate": 9.402896062620848e-05, "loss": 3.0487, "step": 5425 }, { "epoch": 0.16, "grad_norm": 0.5794260501861572, "learning_rate": 9.401797137793313e-05, "loss": 2.9949, "step": 5430 }, { "epoch": 0.16, "grad_norm": 0.5104303359985352, "learning_rate": 9.400697266989112e-05, "loss": 2.9534, "step": 5435 }, { "epoch": 0.16, "grad_norm": 0.5043449401855469, "learning_rate": 9.399596450444617e-05, "loss": 2.9859, "step": 5440 }, { "epoch": 0.16, "grad_norm": 0.528290331363678, "learning_rate": 9.398494688396398e-05, "loss": 3.0687, "step": 5445 }, { "epoch": 0.16, "grad_norm": 0.4701516926288605, "learning_rate": 9.397391981081232e-05, "loss": 3.0169, "step": 5450 }, { "epoch": 0.16, "grad_norm": 0.49247756600379944, "learning_rate": 9.3962883287361e-05, "loss": 2.9877, "step": 5455 }, { "epoch": 0.16, "grad_norm": 0.509986162185669, "learning_rate": 9.395183731598183e-05, "loss": 3.0484, "step": 5460 }, { "epoch": 0.16, "grad_norm": 0.5060095191001892, "learning_rate": 9.394078189904867e-05, "loss": 3.0338, "step": 5465 }, { "epoch": 0.16, "grad_norm": 0.4821600019931793, "learning_rate": 9.392971703893741e-05, "loss": 2.8797, "step": 5470 }, { "epoch": 0.16, "grad_norm": 0.46230044960975647, "learning_rate": 9.391864273802595e-05, "loss": 3.0421, "step": 5475 }, { "epoch": 0.16, "grad_norm": 0.5188688039779663, "learning_rate": 9.390755899869425e-05, "loss": 3.1004, "step": 5480 }, { "epoch": 0.16, "grad_norm": 0.5231472253799438, "learning_rate": 9.38964658233243e-05, "loss": 2.9955, "step": 5485 }, { "epoch": 0.16, "grad_norm": 0.4917570948600769, "learning_rate": 9.388536321430005e-05, "loss": 2.954, "step": 5490 }, { "epoch": 0.16, "grad_norm": 0.49111977219581604, "learning_rate": 9.387425117400757e-05, "loss": 3.0457, "step": 5495 }, { "epoch": 0.16, "grad_norm": 0.507766842842102, "learning_rate": 9.386312970483489e-05, "loss": 2.997, "step": 5500 }, { "epoch": 0.16, "grad_norm": 0.48916691541671753, "learning_rate": 9.385199880917211e-05, "loss": 3.0345, "step": 5505 }, { "epoch": 0.16, "grad_norm": 0.5062963366508484, "learning_rate": 9.384085848941132e-05, "loss": 3.0429, "step": 5510 }, { "epoch": 0.16, "grad_norm": 0.5123204588890076, "learning_rate": 9.382970874794666e-05, "loss": 2.9649, "step": 5515 }, { "epoch": 0.16, "grad_norm": 0.5111730694770813, "learning_rate": 9.381854958717429e-05, "loss": 3.063, "step": 5520 }, { "epoch": 0.16, "grad_norm": 0.49461618065834045, "learning_rate": 9.380738100949238e-05, "loss": 3.0276, "step": 5525 }, { "epoch": 0.16, "grad_norm": 0.4839038848876953, "learning_rate": 9.379620301730114e-05, "loss": 2.9222, "step": 5530 }, { "epoch": 0.16, "grad_norm": 0.49852049350738525, "learning_rate": 9.378501561300282e-05, "loss": 3.0289, "step": 5535 }, { "epoch": 0.16, "grad_norm": 0.4883766770362854, "learning_rate": 9.377381879900162e-05, "loss": 3.0441, "step": 5540 }, { "epoch": 0.16, "grad_norm": 0.49078792333602905, "learning_rate": 9.376261257770384e-05, "loss": 3.0099, "step": 5545 }, { "epoch": 0.16, "grad_norm": 0.4860727787017822, "learning_rate": 9.37513969515178e-05, "loss": 3.0078, "step": 5550 }, { "epoch": 0.16, "grad_norm": 0.498821496963501, "learning_rate": 9.374017192285378e-05, "loss": 3.0745, "step": 5555 }, { "epoch": 0.16, "grad_norm": 0.5146870613098145, "learning_rate": 9.372893749412413e-05, "loss": 2.9548, "step": 5560 }, { "epoch": 0.16, "grad_norm": 0.4697757959365845, "learning_rate": 9.371769366774321e-05, "loss": 2.9432, "step": 5565 }, { "epoch": 0.16, "grad_norm": 0.49438199400901794, "learning_rate": 9.370644044612738e-05, "loss": 3.0148, "step": 5570 }, { "epoch": 0.16, "grad_norm": 0.48534783720970154, "learning_rate": 9.369517783169505e-05, "loss": 3.0268, "step": 5575 }, { "epoch": 0.16, "grad_norm": 0.49150192737579346, "learning_rate": 9.368390582686663e-05, "loss": 2.9731, "step": 5580 }, { "epoch": 0.16, "grad_norm": 0.4905414283275604, "learning_rate": 9.367262443406456e-05, "loss": 2.9564, "step": 5585 }, { "epoch": 0.16, "grad_norm": 0.48296448588371277, "learning_rate": 9.366133365571328e-05, "loss": 3.0936, "step": 5590 }, { "epoch": 0.16, "grad_norm": 0.5082160234451294, "learning_rate": 9.365003349423924e-05, "loss": 3.0886, "step": 5595 }, { "epoch": 0.16, "grad_norm": 0.4993349015712738, "learning_rate": 9.363872395207097e-05, "loss": 2.946, "step": 5600 }, { "epoch": 0.16, "grad_norm": 0.47904837131500244, "learning_rate": 9.362740503163891e-05, "loss": 2.9816, "step": 5605 }, { "epoch": 0.17, "grad_norm": 0.4718692898750305, "learning_rate": 9.361607673537563e-05, "loss": 3.0121, "step": 5610 }, { "epoch": 0.17, "grad_norm": 0.4774525463581085, "learning_rate": 9.36047390657156e-05, "loss": 2.9902, "step": 5615 }, { "epoch": 0.17, "grad_norm": 0.4893141984939575, "learning_rate": 9.359339202509543e-05, "loss": 3.012, "step": 5620 }, { "epoch": 0.17, "grad_norm": 0.49078673124313354, "learning_rate": 9.358203561595363e-05, "loss": 3.0435, "step": 5625 }, { "epoch": 0.17, "grad_norm": 0.5009681582450867, "learning_rate": 9.357066984073079e-05, "loss": 2.983, "step": 5630 }, { "epoch": 0.17, "grad_norm": 0.4553608298301697, "learning_rate": 9.355929470186948e-05, "loss": 3.0874, "step": 5635 }, { "epoch": 0.17, "grad_norm": 0.48020368814468384, "learning_rate": 9.354791020181431e-05, "loss": 3.0907, "step": 5640 }, { "epoch": 0.17, "grad_norm": 0.47585999965667725, "learning_rate": 9.353651634301188e-05, "loss": 3.0403, "step": 5645 }, { "epoch": 0.17, "grad_norm": 0.4846518337726593, "learning_rate": 9.352511312791083e-05, "loss": 2.9288, "step": 5650 }, { "epoch": 0.17, "grad_norm": 0.4731703996658325, "learning_rate": 9.351370055896175e-05, "loss": 2.9059, "step": 5655 }, { "epoch": 0.17, "grad_norm": 0.47612443566322327, "learning_rate": 9.350227863861732e-05, "loss": 2.9393, "step": 5660 }, { "epoch": 0.17, "grad_norm": 0.4715484380722046, "learning_rate": 9.349084736933215e-05, "loss": 3.0919, "step": 5665 }, { "epoch": 0.17, "grad_norm": 0.4899269640445709, "learning_rate": 9.347940675356292e-05, "loss": 3.108, "step": 5670 }, { "epoch": 0.17, "grad_norm": 0.4837105870246887, "learning_rate": 9.346795679376832e-05, "loss": 3.0311, "step": 5675 }, { "epoch": 0.17, "grad_norm": 0.48081597685813904, "learning_rate": 9.345649749240898e-05, "loss": 2.9254, "step": 5680 }, { "epoch": 0.17, "grad_norm": 0.48932644724845886, "learning_rate": 9.344502885194763e-05, "loss": 2.9557, "step": 5685 }, { "epoch": 0.17, "grad_norm": 0.48303529620170593, "learning_rate": 9.343355087484894e-05, "loss": 3.0615, "step": 5690 }, { "epoch": 0.17, "grad_norm": 0.47035014629364014, "learning_rate": 9.342206356357956e-05, "loss": 2.9795, "step": 5695 }, { "epoch": 0.17, "grad_norm": 0.5164473652839661, "learning_rate": 9.341056692060828e-05, "loss": 2.9261, "step": 5700 }, { "epoch": 0.17, "grad_norm": 0.48115551471710205, "learning_rate": 9.339906094840575e-05, "loss": 3.0711, "step": 5705 }, { "epoch": 0.17, "grad_norm": 0.4849070608615875, "learning_rate": 9.33875456494447e-05, "loss": 3.0082, "step": 5710 }, { "epoch": 0.17, "grad_norm": 0.49722084403038025, "learning_rate": 9.337602102619985e-05, "loss": 3.062, "step": 5715 }, { "epoch": 0.17, "grad_norm": 0.4830116629600525, "learning_rate": 9.336448708114789e-05, "loss": 2.9833, "step": 5720 }, { "epoch": 0.17, "grad_norm": 0.4678457975387573, "learning_rate": 9.33529438167676e-05, "loss": 3.0071, "step": 5725 }, { "epoch": 0.17, "grad_norm": 0.464857280254364, "learning_rate": 9.334139123553967e-05, "loss": 2.9641, "step": 5730 }, { "epoch": 0.17, "grad_norm": 0.49367448687553406, "learning_rate": 9.332982933994683e-05, "loss": 2.9553, "step": 5735 }, { "epoch": 0.17, "grad_norm": 0.4856371283531189, "learning_rate": 9.331825813247381e-05, "loss": 2.9678, "step": 5740 }, { "epoch": 0.17, "grad_norm": 0.5203416347503662, "learning_rate": 9.330667761560739e-05, "loss": 3.0057, "step": 5745 }, { "epoch": 0.17, "grad_norm": 0.49711576104164124, "learning_rate": 9.329508779183623e-05, "loss": 2.914, "step": 5750 }, { "epoch": 0.17, "grad_norm": 0.5314367413520813, "learning_rate": 9.328348866365111e-05, "loss": 3.0116, "step": 5755 }, { "epoch": 0.17, "grad_norm": 0.5206249356269836, "learning_rate": 9.327188023354474e-05, "loss": 3.0293, "step": 5760 }, { "epoch": 0.17, "grad_norm": 0.4911401569843292, "learning_rate": 9.326026250401188e-05, "loss": 3.0087, "step": 5765 }, { "epoch": 0.17, "grad_norm": 0.5011433362960815, "learning_rate": 9.324863547754924e-05, "loss": 2.985, "step": 5770 }, { "epoch": 0.17, "grad_norm": 0.4635113477706909, "learning_rate": 9.323699915665555e-05, "loss": 2.9967, "step": 5775 }, { "epoch": 0.17, "grad_norm": 0.4469456076622009, "learning_rate": 9.322535354383153e-05, "loss": 3.0271, "step": 5780 }, { "epoch": 0.17, "grad_norm": 0.4868432581424713, "learning_rate": 9.321369864157992e-05, "loss": 3.091, "step": 5785 }, { "epoch": 0.17, "grad_norm": 0.46113985776901245, "learning_rate": 9.320203445240543e-05, "loss": 3.023, "step": 5790 }, { "epoch": 0.17, "grad_norm": 0.4704936146736145, "learning_rate": 9.319036097881477e-05, "loss": 3.0137, "step": 5795 }, { "epoch": 0.17, "grad_norm": 0.5014192461967468, "learning_rate": 9.317867822331667e-05, "loss": 3.0043, "step": 5800 }, { "epoch": 0.17, "grad_norm": 0.4753495454788208, "learning_rate": 9.31669861884218e-05, "loss": 2.9929, "step": 5805 }, { "epoch": 0.17, "grad_norm": 0.620468020439148, "learning_rate": 9.315528487664289e-05, "loss": 2.9644, "step": 5810 }, { "epoch": 0.17, "grad_norm": 0.4698863923549652, "learning_rate": 9.314357429049462e-05, "loss": 2.9582, "step": 5815 }, { "epoch": 0.17, "grad_norm": 0.4722135365009308, "learning_rate": 9.313185443249367e-05, "loss": 3.0534, "step": 5820 }, { "epoch": 0.17, "grad_norm": 0.48420825600624084, "learning_rate": 9.312012530515875e-05, "loss": 2.9358, "step": 5825 }, { "epoch": 0.17, "grad_norm": 0.47350627183914185, "learning_rate": 9.310838691101048e-05, "loss": 2.989, "step": 5830 }, { "epoch": 0.17, "grad_norm": 0.5510419011116028, "learning_rate": 9.309663925257156e-05, "loss": 3.0177, "step": 5835 }, { "epoch": 0.17, "grad_norm": 0.5616435408592224, "learning_rate": 9.308488233236663e-05, "loss": 3.0394, "step": 5840 }, { "epoch": 0.17, "grad_norm": 0.49273791909217834, "learning_rate": 9.307311615292234e-05, "loss": 3.0263, "step": 5845 }, { "epoch": 0.17, "grad_norm": 0.5135101079940796, "learning_rate": 9.306134071676734e-05, "loss": 3.1071, "step": 5850 }, { "epoch": 0.17, "grad_norm": 0.5018031597137451, "learning_rate": 9.30495560264322e-05, "loss": 2.969, "step": 5855 }, { "epoch": 0.17, "grad_norm": 0.488640159368515, "learning_rate": 9.303776208444958e-05, "loss": 3.0805, "step": 5860 }, { "epoch": 0.17, "grad_norm": 0.46880149841308594, "learning_rate": 9.302595889335405e-05, "loss": 2.9215, "step": 5865 }, { "epoch": 0.17, "grad_norm": 0.48733651638031006, "learning_rate": 9.301414645568223e-05, "loss": 3.0093, "step": 5870 }, { "epoch": 0.17, "grad_norm": 0.48846063017845154, "learning_rate": 9.300232477397266e-05, "loss": 2.9807, "step": 5875 }, { "epoch": 0.17, "grad_norm": 0.4708818793296814, "learning_rate": 9.299049385076592e-05, "loss": 2.9613, "step": 5880 }, { "epoch": 0.17, "grad_norm": 0.4999903440475464, "learning_rate": 9.297865368860456e-05, "loss": 2.8886, "step": 5885 }, { "epoch": 0.17, "grad_norm": 0.4652180075645447, "learning_rate": 9.296680429003311e-05, "loss": 2.9309, "step": 5890 }, { "epoch": 0.17, "grad_norm": 0.4733195900917053, "learning_rate": 9.295494565759808e-05, "loss": 2.9978, "step": 5895 }, { "epoch": 0.17, "grad_norm": 0.5219632983207703, "learning_rate": 9.294307779384796e-05, "loss": 3.0519, "step": 5900 }, { "epoch": 0.17, "grad_norm": 0.4783526360988617, "learning_rate": 9.293120070133327e-05, "loss": 3.0077, "step": 5905 }, { "epoch": 0.17, "grad_norm": 0.48892921209335327, "learning_rate": 9.291931438260646e-05, "loss": 3.0382, "step": 5910 }, { "epoch": 0.17, "grad_norm": 0.5030211806297302, "learning_rate": 9.290741884022198e-05, "loss": 3.0267, "step": 5915 }, { "epoch": 0.17, "grad_norm": 0.4809192717075348, "learning_rate": 9.28955140767363e-05, "loss": 2.9736, "step": 5920 }, { "epoch": 0.17, "grad_norm": 0.4907720983028412, "learning_rate": 9.288360009470776e-05, "loss": 2.9882, "step": 5925 }, { "epoch": 0.17, "grad_norm": 0.5322912931442261, "learning_rate": 9.287167689669683e-05, "loss": 2.9123, "step": 5930 }, { "epoch": 0.17, "grad_norm": 0.46553829312324524, "learning_rate": 9.285974448526585e-05, "loss": 2.9994, "step": 5935 }, { "epoch": 0.17, "grad_norm": 0.4999275803565979, "learning_rate": 9.284780286297918e-05, "loss": 2.9498, "step": 5940 }, { "epoch": 0.17, "grad_norm": 0.47500064969062805, "learning_rate": 9.283585203240317e-05, "loss": 3.0027, "step": 5945 }, { "epoch": 0.18, "grad_norm": 0.46592390537261963, "learning_rate": 9.282389199610613e-05, "loss": 2.9656, "step": 5950 }, { "epoch": 0.18, "grad_norm": 0.45659705996513367, "learning_rate": 9.281192275665836e-05, "loss": 2.9922, "step": 5955 }, { "epoch": 0.18, "grad_norm": 0.4941502511501312, "learning_rate": 9.279994431663212e-05, "loss": 2.9474, "step": 5960 }, { "epoch": 0.18, "grad_norm": 3.2555606365203857, "learning_rate": 9.278795667860167e-05, "loss": 2.9467, "step": 5965 }, { "epoch": 0.18, "grad_norm": 0.4769304692745209, "learning_rate": 9.277595984514325e-05, "loss": 2.9534, "step": 5970 }, { "epoch": 0.18, "grad_norm": 0.48253417015075684, "learning_rate": 9.276395381883502e-05, "loss": 3.0031, "step": 5975 }, { "epoch": 0.18, "grad_norm": 0.49596214294433594, "learning_rate": 9.275193860225722e-05, "loss": 2.969, "step": 5980 }, { "epoch": 0.18, "grad_norm": 0.4897612929344177, "learning_rate": 9.273991419799193e-05, "loss": 2.9912, "step": 5985 }, { "epoch": 0.18, "grad_norm": 0.4993102252483368, "learning_rate": 9.272788060862334e-05, "loss": 3.0563, "step": 5990 }, { "epoch": 0.18, "grad_norm": 0.4987744987010956, "learning_rate": 9.271583783673751e-05, "loss": 2.9863, "step": 5995 }, { "epoch": 0.18, "grad_norm": 0.5047387480735779, "learning_rate": 9.270378588492255e-05, "loss": 2.9747, "step": 6000 }, { "epoch": 0.18, "grad_norm": 0.5020679235458374, "learning_rate": 9.269172475576849e-05, "loss": 2.9918, "step": 6005 }, { "epoch": 0.18, "grad_norm": 0.48928937315940857, "learning_rate": 9.267965445186733e-05, "loss": 2.8946, "step": 6010 }, { "epoch": 0.18, "grad_norm": 0.4627678096294403, "learning_rate": 9.266757497581308e-05, "loss": 3.0515, "step": 6015 }, { "epoch": 0.18, "grad_norm": 0.536028265953064, "learning_rate": 9.265548633020174e-05, "loss": 2.9113, "step": 6020 }, { "epoch": 0.18, "grad_norm": 0.4701045751571655, "learning_rate": 9.264338851763117e-05, "loss": 3.0161, "step": 6025 }, { "epoch": 0.18, "grad_norm": 0.47526925802230835, "learning_rate": 9.263128154070133e-05, "loss": 2.9527, "step": 6030 }, { "epoch": 0.18, "grad_norm": 0.48833751678466797, "learning_rate": 9.261916540201407e-05, "loss": 2.9964, "step": 6035 }, { "epoch": 0.18, "grad_norm": 0.503982424736023, "learning_rate": 9.260704010417325e-05, "loss": 3.025, "step": 6040 }, { "epoch": 0.18, "grad_norm": 0.4825071692466736, "learning_rate": 9.259490564978466e-05, "loss": 2.9068, "step": 6045 }, { "epoch": 0.18, "grad_norm": 0.46499887108802795, "learning_rate": 9.258276204145609e-05, "loss": 2.983, "step": 6050 }, { "epoch": 0.18, "grad_norm": 0.47848033905029297, "learning_rate": 9.257060928179726e-05, "loss": 2.9846, "step": 6055 }, { "epoch": 0.18, "grad_norm": 0.4789679944515228, "learning_rate": 9.255844737341992e-05, "loss": 2.9351, "step": 6060 }, { "epoch": 0.18, "grad_norm": 0.4682319760322571, "learning_rate": 9.254627631893773e-05, "loss": 2.954, "step": 6065 }, { "epoch": 0.18, "grad_norm": 0.48575130105018616, "learning_rate": 9.253409612096635e-05, "loss": 2.9731, "step": 6070 }, { "epoch": 0.18, "grad_norm": 0.504866898059845, "learning_rate": 9.252190678212335e-05, "loss": 2.9398, "step": 6075 }, { "epoch": 0.18, "grad_norm": 0.46055760979652405, "learning_rate": 9.250970830502834e-05, "loss": 2.9809, "step": 6080 }, { "epoch": 0.18, "grad_norm": 0.47376757860183716, "learning_rate": 9.249750069230285e-05, "loss": 3.0626, "step": 6085 }, { "epoch": 0.18, "grad_norm": 0.46104568243026733, "learning_rate": 9.248528394657036e-05, "loss": 2.9923, "step": 6090 }, { "epoch": 0.18, "grad_norm": 0.48255741596221924, "learning_rate": 9.247305807045637e-05, "loss": 2.9725, "step": 6095 }, { "epoch": 0.18, "grad_norm": 0.49375593662261963, "learning_rate": 9.246082306658827e-05, "loss": 3.0005, "step": 6100 }, { "epoch": 0.18, "grad_norm": 0.4716639816761017, "learning_rate": 9.244857893759545e-05, "loss": 2.9948, "step": 6105 }, { "epoch": 0.18, "grad_norm": 0.4614144563674927, "learning_rate": 9.243632568610929e-05, "loss": 2.993, "step": 6110 }, { "epoch": 0.18, "grad_norm": 0.4558739960193634, "learning_rate": 9.242406331476307e-05, "loss": 2.9233, "step": 6115 }, { "epoch": 0.18, "grad_norm": 0.4794633984565735, "learning_rate": 9.241179182619206e-05, "loss": 2.9812, "step": 6120 }, { "epoch": 0.18, "grad_norm": 0.4598528742790222, "learning_rate": 9.239951122303349e-05, "loss": 2.9149, "step": 6125 }, { "epoch": 0.18, "grad_norm": 0.47477206587791443, "learning_rate": 9.238722150792658e-05, "loss": 3.0054, "step": 6130 }, { "epoch": 0.18, "grad_norm": 0.48073726892471313, "learning_rate": 9.237492268351242e-05, "loss": 2.9364, "step": 6135 }, { "epoch": 0.18, "grad_norm": 0.4842113256454468, "learning_rate": 9.236261475243415e-05, "loss": 2.9594, "step": 6140 }, { "epoch": 0.18, "grad_norm": 0.45292335748672485, "learning_rate": 9.23502977173368e-05, "loss": 2.987, "step": 6145 }, { "epoch": 0.18, "grad_norm": 0.4511399567127228, "learning_rate": 9.233797158086742e-05, "loss": 2.9305, "step": 6150 }, { "epoch": 0.18, "grad_norm": 0.4790697991847992, "learning_rate": 9.232563634567495e-05, "loss": 2.9257, "step": 6155 }, { "epoch": 0.18, "grad_norm": 0.4590892791748047, "learning_rate": 9.231329201441034e-05, "loss": 3.0452, "step": 6160 }, { "epoch": 0.18, "grad_norm": 0.4549653232097626, "learning_rate": 9.230093858972646e-05, "loss": 2.9865, "step": 6165 }, { "epoch": 0.18, "grad_norm": 0.49256616830825806, "learning_rate": 9.228857607427816e-05, "loss": 3.0285, "step": 6170 }, { "epoch": 0.18, "grad_norm": 0.4668876528739929, "learning_rate": 9.227620447072221e-05, "loss": 2.9379, "step": 6175 }, { "epoch": 0.18, "grad_norm": 0.4834196865558624, "learning_rate": 9.226382378171738e-05, "loss": 3.0011, "step": 6180 }, { "epoch": 0.18, "grad_norm": 0.4850182831287384, "learning_rate": 9.225143400992435e-05, "loss": 2.9875, "step": 6185 }, { "epoch": 0.18, "grad_norm": 0.4665440618991852, "learning_rate": 9.223903515800576e-05, "loss": 2.9323, "step": 6190 }, { "epoch": 0.18, "grad_norm": 0.4914660155773163, "learning_rate": 9.222662722862622e-05, "loss": 3.0072, "step": 6195 }, { "epoch": 0.18, "grad_norm": 0.48132970929145813, "learning_rate": 9.221421022445228e-05, "loss": 2.9016, "step": 6200 }, { "epoch": 0.18, "grad_norm": 0.53288733959198, "learning_rate": 9.220178414815242e-05, "loss": 2.9918, "step": 6205 }, { "epoch": 0.18, "grad_norm": 0.4755805432796478, "learning_rate": 9.218934900239713e-05, "loss": 3.0104, "step": 6210 }, { "epoch": 0.18, "grad_norm": 0.46284207701683044, "learning_rate": 9.217690478985878e-05, "loss": 3.0227, "step": 6215 }, { "epoch": 0.18, "grad_norm": 0.466816782951355, "learning_rate": 9.216445151321171e-05, "loss": 2.9896, "step": 6220 }, { "epoch": 0.18, "grad_norm": 0.4499899446964264, "learning_rate": 9.215198917513223e-05, "loss": 2.9951, "step": 6225 }, { "epoch": 0.18, "grad_norm": 0.4471033811569214, "learning_rate": 9.213951777829861e-05, "loss": 2.876, "step": 6230 }, { "epoch": 0.18, "grad_norm": 0.46359407901763916, "learning_rate": 9.212703732539098e-05, "loss": 3.0161, "step": 6235 }, { "epoch": 0.18, "grad_norm": 0.5274807214736938, "learning_rate": 9.211454781909153e-05, "loss": 2.9558, "step": 6240 }, { "epoch": 0.18, "grad_norm": 0.5123795866966248, "learning_rate": 9.21020492620843e-05, "loss": 2.9433, "step": 6245 }, { "epoch": 0.18, "grad_norm": 0.4780196249485016, "learning_rate": 9.208954165705534e-05, "loss": 3.0784, "step": 6250 }, { "epoch": 0.18, "grad_norm": 0.49553558230400085, "learning_rate": 9.207702500669262e-05, "loss": 2.9275, "step": 6255 }, { "epoch": 0.18, "grad_norm": 0.4668084979057312, "learning_rate": 9.206449931368604e-05, "loss": 2.925, "step": 6260 }, { "epoch": 0.18, "grad_norm": 0.44680842757225037, "learning_rate": 9.205196458072747e-05, "loss": 3.0023, "step": 6265 }, { "epoch": 0.18, "grad_norm": 0.47806334495544434, "learning_rate": 9.203942081051071e-05, "loss": 2.8947, "step": 6270 }, { "epoch": 0.18, "grad_norm": 0.4853799045085907, "learning_rate": 9.20268680057315e-05, "loss": 2.9048, "step": 6275 }, { "epoch": 0.18, "grad_norm": 0.46591871976852417, "learning_rate": 9.201430616908752e-05, "loss": 3.0146, "step": 6280 }, { "epoch": 0.18, "grad_norm": 0.46370717883110046, "learning_rate": 9.200173530327842e-05, "loss": 2.9721, "step": 6285 }, { "epoch": 0.19, "grad_norm": 0.4716801643371582, "learning_rate": 9.198915541100575e-05, "loss": 2.9744, "step": 6290 }, { "epoch": 0.19, "grad_norm": 0.4748280942440033, "learning_rate": 9.1976566494973e-05, "loss": 2.9343, "step": 6295 }, { "epoch": 0.19, "grad_norm": 0.4797299802303314, "learning_rate": 9.196396855788563e-05, "loss": 3.0031, "step": 6300 }, { "epoch": 0.19, "grad_norm": 0.46222159266471863, "learning_rate": 9.195136160245104e-05, "loss": 2.9354, "step": 6305 }, { "epoch": 0.19, "grad_norm": 0.4493109881877899, "learning_rate": 9.193874563137853e-05, "loss": 3.0634, "step": 6310 }, { "epoch": 0.19, "grad_norm": 0.4668532907962799, "learning_rate": 9.192612064737934e-05, "loss": 2.9755, "step": 6315 }, { "epoch": 0.19, "grad_norm": 0.476399302482605, "learning_rate": 9.19134866531667e-05, "loss": 2.9729, "step": 6320 }, { "epoch": 0.19, "grad_norm": 0.4801667630672455, "learning_rate": 9.190084365145575e-05, "loss": 2.9712, "step": 6325 }, { "epoch": 0.19, "grad_norm": 0.4622453451156616, "learning_rate": 9.188819164496354e-05, "loss": 2.8918, "step": 6330 }, { "epoch": 0.19, "grad_norm": 0.47751423716545105, "learning_rate": 9.187553063640906e-05, "loss": 2.9242, "step": 6335 }, { "epoch": 0.19, "grad_norm": 0.45603206753730774, "learning_rate": 9.186286062851327e-05, "loss": 3.0038, "step": 6340 }, { "epoch": 0.19, "grad_norm": 0.47120070457458496, "learning_rate": 9.185018162399905e-05, "loss": 2.9482, "step": 6345 }, { "epoch": 0.19, "grad_norm": 0.4731959402561188, "learning_rate": 9.183749362559118e-05, "loss": 2.9952, "step": 6350 }, { "epoch": 0.19, "grad_norm": 0.4725649654865265, "learning_rate": 9.182479663601642e-05, "loss": 2.9717, "step": 6355 }, { "epoch": 0.19, "grad_norm": 0.4515538811683655, "learning_rate": 9.181209065800342e-05, "loss": 3.0032, "step": 6360 }, { "epoch": 0.19, "grad_norm": 0.47481176257133484, "learning_rate": 9.17993756942828e-05, "loss": 3.0298, "step": 6365 }, { "epoch": 0.19, "grad_norm": 0.4599917531013489, "learning_rate": 9.17866517475871e-05, "loss": 2.9351, "step": 6370 }, { "epoch": 0.19, "grad_norm": 0.46684667468070984, "learning_rate": 9.177391882065077e-05, "loss": 3.0671, "step": 6375 }, { "epoch": 0.19, "grad_norm": 0.45776793360710144, "learning_rate": 9.176117691621017e-05, "loss": 2.9864, "step": 6380 }, { "epoch": 0.19, "grad_norm": 0.4541158080101013, "learning_rate": 9.17484260370037e-05, "loss": 2.9791, "step": 6385 }, { "epoch": 0.19, "grad_norm": 0.46621742844581604, "learning_rate": 9.173566618577156e-05, "loss": 2.9663, "step": 6390 }, { "epoch": 0.19, "grad_norm": 0.4670698642730713, "learning_rate": 9.172289736525594e-05, "loss": 2.9604, "step": 6395 }, { "epoch": 0.19, "grad_norm": 0.4682334065437317, "learning_rate": 9.171011957820097e-05, "loss": 2.9458, "step": 6400 }, { "epoch": 0.19, "grad_norm": 0.442361980676651, "learning_rate": 9.169733282735265e-05, "loss": 2.9929, "step": 6405 }, { "epoch": 0.19, "grad_norm": 0.4448869526386261, "learning_rate": 9.168453711545895e-05, "loss": 2.9096, "step": 6410 }, { "epoch": 0.19, "grad_norm": 0.4681752026081085, "learning_rate": 9.167173244526977e-05, "loss": 2.9212, "step": 6415 }, { "epoch": 0.19, "grad_norm": 0.48632150888442993, "learning_rate": 9.165891881953692e-05, "loss": 3.0339, "step": 6420 }, { "epoch": 0.19, "grad_norm": 0.4651564955711365, "learning_rate": 9.164609624101414e-05, "loss": 3.047, "step": 6425 }, { "epoch": 0.19, "grad_norm": 0.48560383915901184, "learning_rate": 9.163326471245707e-05, "loss": 3.0109, "step": 6430 }, { "epoch": 0.19, "grad_norm": 0.45293229818344116, "learning_rate": 9.162042423662331e-05, "loss": 3.0392, "step": 6435 }, { "epoch": 0.19, "grad_norm": 0.4713640809059143, "learning_rate": 9.160757481627237e-05, "loss": 2.9841, "step": 6440 }, { "epoch": 0.19, "grad_norm": 0.4639374613761902, "learning_rate": 9.159471645416567e-05, "loss": 2.9326, "step": 6445 }, { "epoch": 0.19, "grad_norm": 0.4834689199924469, "learning_rate": 9.158184915306655e-05, "loss": 2.9497, "step": 6450 }, { "epoch": 0.19, "grad_norm": 0.6186258792877197, "learning_rate": 9.156897291574034e-05, "loss": 3.0028, "step": 6455 }, { "epoch": 0.19, "grad_norm": 0.4426053464412689, "learning_rate": 9.155608774495415e-05, "loss": 2.9018, "step": 6460 }, { "epoch": 0.19, "grad_norm": 0.48873472213745117, "learning_rate": 9.154319364347715e-05, "loss": 2.8931, "step": 6465 }, { "epoch": 0.19, "grad_norm": 0.45704931020736694, "learning_rate": 9.153029061408037e-05, "loss": 3.0683, "step": 6470 }, { "epoch": 0.19, "grad_norm": 0.4610084891319275, "learning_rate": 9.151737865953671e-05, "loss": 2.9146, "step": 6475 }, { "epoch": 0.19, "grad_norm": 0.4579439163208008, "learning_rate": 9.150445778262109e-05, "loss": 2.956, "step": 6480 }, { "epoch": 0.19, "grad_norm": 0.47710561752319336, "learning_rate": 9.149152798611027e-05, "loss": 2.932, "step": 6485 }, { "epoch": 0.19, "grad_norm": 0.46434521675109863, "learning_rate": 9.147858927278296e-05, "loss": 2.9676, "step": 6490 }, { "epoch": 0.19, "grad_norm": 0.5159434080123901, "learning_rate": 9.14656416454198e-05, "loss": 2.9271, "step": 6495 }, { "epoch": 0.19, "grad_norm": 0.4928341209888458, "learning_rate": 9.14526851068033e-05, "loss": 2.9742, "step": 6500 }, { "epoch": 0.19, "grad_norm": 0.47781506180763245, "learning_rate": 9.143971965971789e-05, "loss": 2.9434, "step": 6505 }, { "epoch": 0.19, "grad_norm": 0.4806775152683258, "learning_rate": 9.142674530694999e-05, "loss": 2.9462, "step": 6510 }, { "epoch": 0.19, "grad_norm": 0.45032501220703125, "learning_rate": 9.141376205128784e-05, "loss": 2.912, "step": 6515 }, { "epoch": 0.19, "grad_norm": 0.43274399638175964, "learning_rate": 9.140076989552162e-05, "loss": 2.9541, "step": 6520 }, { "epoch": 0.19, "grad_norm": 0.4643862843513489, "learning_rate": 9.138776884244347e-05, "loss": 2.9833, "step": 6525 }, { "epoch": 0.19, "grad_norm": 0.4594551920890808, "learning_rate": 9.137475889484739e-05, "loss": 2.9653, "step": 6530 }, { "epoch": 0.19, "grad_norm": 0.4554634392261505, "learning_rate": 9.13617400555293e-05, "loss": 2.9153, "step": 6535 }, { "epoch": 0.19, "grad_norm": 0.4643602967262268, "learning_rate": 9.134871232728706e-05, "loss": 3.0003, "step": 6540 }, { "epoch": 0.19, "grad_norm": 0.45453253388404846, "learning_rate": 9.133567571292038e-05, "loss": 2.9479, "step": 6545 }, { "epoch": 0.19, "grad_norm": 0.4724680185317993, "learning_rate": 9.132263021523096e-05, "loss": 2.931, "step": 6550 }, { "epoch": 0.19, "grad_norm": 0.4669559895992279, "learning_rate": 9.130957583702236e-05, "loss": 3.0642, "step": 6555 }, { "epoch": 0.19, "grad_norm": 0.4696744680404663, "learning_rate": 9.129651258110003e-05, "loss": 2.8941, "step": 6560 }, { "epoch": 0.19, "grad_norm": 0.47664403915405273, "learning_rate": 9.128344045027137e-05, "loss": 2.9395, "step": 6565 }, { "epoch": 0.19, "grad_norm": 0.4575192630290985, "learning_rate": 9.127035944734568e-05, "loss": 2.9891, "step": 6570 }, { "epoch": 0.19, "grad_norm": 0.7435876727104187, "learning_rate": 9.125726957513414e-05, "loss": 2.9805, "step": 6575 }, { "epoch": 0.19, "grad_norm": 0.4616797864437103, "learning_rate": 9.124417083644988e-05, "loss": 2.9791, "step": 6580 }, { "epoch": 0.19, "grad_norm": 0.45778268575668335, "learning_rate": 9.12310632341079e-05, "loss": 2.9553, "step": 6585 }, { "epoch": 0.19, "grad_norm": 0.46301910281181335, "learning_rate": 9.121794677092509e-05, "loss": 2.7883, "step": 6590 }, { "epoch": 0.19, "grad_norm": 0.4407591223716736, "learning_rate": 9.120482144972029e-05, "loss": 2.9919, "step": 6595 }, { "epoch": 0.19, "grad_norm": 0.462947279214859, "learning_rate": 9.119168727331424e-05, "loss": 2.9144, "step": 6600 }, { "epoch": 0.19, "grad_norm": 0.47778481245040894, "learning_rate": 9.117854424452952e-05, "loss": 2.9691, "step": 6605 }, { "epoch": 0.19, "grad_norm": 0.5103906393051147, "learning_rate": 9.116539236619071e-05, "loss": 2.9188, "step": 6610 }, { "epoch": 0.19, "grad_norm": 0.4464859366416931, "learning_rate": 9.115223164112418e-05, "loss": 2.9654, "step": 6615 }, { "epoch": 0.19, "grad_norm": 0.47829532623291016, "learning_rate": 9.113906207215832e-05, "loss": 2.9989, "step": 6620 }, { "epoch": 0.19, "grad_norm": 0.4847690165042877, "learning_rate": 9.112588366212333e-05, "loss": 2.8989, "step": 6625 }, { "epoch": 0.2, "grad_norm": 0.47465983033180237, "learning_rate": 9.111269641385134e-05, "loss": 2.9126, "step": 6630 }, { "epoch": 0.2, "grad_norm": 0.45469507575035095, "learning_rate": 9.10995003301764e-05, "loss": 2.95, "step": 6635 }, { "epoch": 0.2, "grad_norm": 0.4525333344936371, "learning_rate": 9.10862954139344e-05, "loss": 2.9382, "step": 6640 }, { "epoch": 0.2, "grad_norm": 0.46923592686653137, "learning_rate": 9.107308166796322e-05, "loss": 3.0043, "step": 6645 }, { "epoch": 0.2, "grad_norm": 4.369565486907959, "learning_rate": 9.105985909510255e-05, "loss": 2.9333, "step": 6650 }, { "epoch": 0.2, "grad_norm": 0.45938149094581604, "learning_rate": 9.104662769819404e-05, "loss": 2.9576, "step": 6655 }, { "epoch": 0.2, "grad_norm": 0.4658982753753662, "learning_rate": 9.103338748008118e-05, "loss": 3.0471, "step": 6660 }, { "epoch": 0.2, "grad_norm": 0.4435548782348633, "learning_rate": 9.102013844360939e-05, "loss": 2.9548, "step": 6665 }, { "epoch": 0.2, "grad_norm": 0.45993366837501526, "learning_rate": 9.100688059162599e-05, "loss": 2.887, "step": 6670 }, { "epoch": 0.2, "grad_norm": 0.4602290689945221, "learning_rate": 9.099361392698018e-05, "loss": 3.0801, "step": 6675 }, { "epoch": 0.2, "grad_norm": 0.4423500597476959, "learning_rate": 9.098033845252307e-05, "loss": 2.8832, "step": 6680 }, { "epoch": 0.2, "grad_norm": 0.48345962166786194, "learning_rate": 9.096705417110762e-05, "loss": 2.9691, "step": 6685 }, { "epoch": 0.2, "grad_norm": 0.47332748770713806, "learning_rate": 9.095376108558875e-05, "loss": 2.907, "step": 6690 }, { "epoch": 0.2, "grad_norm": 0.47301605343818665, "learning_rate": 9.094045919882322e-05, "loss": 2.9259, "step": 6695 }, { "epoch": 0.2, "grad_norm": 0.4567819833755493, "learning_rate": 9.092714851366968e-05, "loss": 2.9471, "step": 6700 }, { "epoch": 0.2, "grad_norm": 0.48256784677505493, "learning_rate": 9.091382903298872e-05, "loss": 2.8682, "step": 6705 }, { "epoch": 0.2, "grad_norm": 0.47909629344940186, "learning_rate": 9.090050075964276e-05, "loss": 2.952, "step": 6710 }, { "epoch": 0.2, "grad_norm": 0.4361881911754608, "learning_rate": 9.088716369649616e-05, "loss": 2.8812, "step": 6715 }, { "epoch": 0.2, "grad_norm": 0.47402554750442505, "learning_rate": 9.087381784641515e-05, "loss": 2.9258, "step": 6720 }, { "epoch": 0.2, "grad_norm": 0.4567302465438843, "learning_rate": 9.086046321226783e-05, "loss": 2.9816, "step": 6725 }, { "epoch": 0.2, "grad_norm": 0.46572384238243103, "learning_rate": 9.08470997969242e-05, "loss": 2.9271, "step": 6730 }, { "epoch": 0.2, "grad_norm": 0.4568386673927307, "learning_rate": 9.083372760325617e-05, "loss": 2.9218, "step": 6735 }, { "epoch": 0.2, "grad_norm": 0.4668714106082916, "learning_rate": 9.08203466341375e-05, "loss": 2.8609, "step": 6740 }, { "epoch": 0.2, "grad_norm": 0.4474034905433655, "learning_rate": 9.080695689244385e-05, "loss": 2.9464, "step": 6745 }, { "epoch": 0.2, "grad_norm": 0.467439204454422, "learning_rate": 9.079355838105279e-05, "loss": 2.999, "step": 6750 }, { "epoch": 0.2, "grad_norm": 0.4955962300300598, "learning_rate": 9.078015110284373e-05, "loss": 2.9196, "step": 6755 }, { "epoch": 0.2, "grad_norm": 0.476922869682312, "learning_rate": 9.076673506069801e-05, "loss": 3.0324, "step": 6760 }, { "epoch": 0.2, "grad_norm": 0.4612782299518585, "learning_rate": 9.07533102574988e-05, "loss": 2.9032, "step": 6765 }, { "epoch": 0.2, "grad_norm": 0.44608959555625916, "learning_rate": 9.07398766961312e-05, "loss": 2.9164, "step": 6770 }, { "epoch": 0.2, "grad_norm": 0.4526824653148651, "learning_rate": 9.072643437948218e-05, "loss": 2.9652, "step": 6775 }, { "epoch": 0.2, "grad_norm": 0.4631563127040863, "learning_rate": 9.07129833104406e-05, "loss": 2.9118, "step": 6780 }, { "epoch": 0.2, "grad_norm": 0.48164522647857666, "learning_rate": 9.069952349189714e-05, "loss": 2.9935, "step": 6785 }, { "epoch": 0.2, "grad_norm": 0.4462948441505432, "learning_rate": 9.068605492674446e-05, "loss": 3.0251, "step": 6790 }, { "epoch": 0.2, "grad_norm": 0.4385778605937958, "learning_rate": 9.0672577617877e-05, "loss": 2.8282, "step": 6795 }, { "epoch": 0.2, "grad_norm": 0.49770697951316833, "learning_rate": 9.065909156819118e-05, "loss": 2.9504, "step": 6800 }, { "epoch": 0.2, "grad_norm": 0.4530959129333496, "learning_rate": 9.06455967805852e-05, "loss": 2.8528, "step": 6805 }, { "epoch": 0.2, "grad_norm": 0.4602462649345398, "learning_rate": 9.063209325795921e-05, "loss": 2.9781, "step": 6810 }, { "epoch": 0.2, "grad_norm": 0.4609087407588959, "learning_rate": 9.06185810032152e-05, "loss": 2.9447, "step": 6815 }, { "epoch": 0.2, "grad_norm": 0.47063148021698, "learning_rate": 9.060506001925707e-05, "loss": 2.9651, "step": 6820 }, { "epoch": 0.2, "grad_norm": 0.4614410400390625, "learning_rate": 9.059153030899051e-05, "loss": 2.8868, "step": 6825 }, { "epoch": 0.2, "grad_norm": 0.4601299464702606, "learning_rate": 9.05779918753232e-05, "loss": 2.9022, "step": 6830 }, { "epoch": 0.2, "grad_norm": 0.4785211980342865, "learning_rate": 9.056444472116465e-05, "loss": 2.912, "step": 6835 }, { "epoch": 0.2, "grad_norm": 0.45824021100997925, "learning_rate": 9.055088884942622e-05, "loss": 2.9382, "step": 6840 }, { "epoch": 0.2, "grad_norm": 0.45250529050827026, "learning_rate": 9.053732426302116e-05, "loss": 2.9403, "step": 6845 }, { "epoch": 0.2, "grad_norm": 0.4856509566307068, "learning_rate": 9.052375096486458e-05, "loss": 2.9773, "step": 6850 }, { "epoch": 0.2, "grad_norm": 0.44042065739631653, "learning_rate": 9.051016895787352e-05, "loss": 2.9548, "step": 6855 }, { "epoch": 0.2, "grad_norm": 0.4589398205280304, "learning_rate": 9.049657824496678e-05, "loss": 3.0052, "step": 6860 }, { "epoch": 0.2, "grad_norm": 0.4901938736438751, "learning_rate": 9.048297882906516e-05, "loss": 3.0542, "step": 6865 }, { "epoch": 0.2, "grad_norm": 0.4781782627105713, "learning_rate": 9.046937071309124e-05, "loss": 2.9852, "step": 6870 }, { "epoch": 0.2, "grad_norm": 0.5225309133529663, "learning_rate": 9.04557538999695e-05, "loss": 2.9186, "step": 6875 }, { "epoch": 0.2, "grad_norm": 0.5181701183319092, "learning_rate": 9.04421283926263e-05, "loss": 3.0155, "step": 6880 }, { "epoch": 0.2, "grad_norm": 0.45996883511543274, "learning_rate": 9.042849419398982e-05, "loss": 2.8887, "step": 6885 }, { "epoch": 0.2, "grad_norm": 0.5226542949676514, "learning_rate": 9.041485130699018e-05, "loss": 2.9089, "step": 6890 }, { "epoch": 0.2, "grad_norm": 0.44956931471824646, "learning_rate": 9.040119973455934e-05, "loss": 3.0262, "step": 6895 }, { "epoch": 0.2, "grad_norm": 0.46628448367118835, "learning_rate": 9.038753947963108e-05, "loss": 2.9173, "step": 6900 }, { "epoch": 0.2, "grad_norm": 0.48515626788139343, "learning_rate": 9.03738705451411e-05, "loss": 2.9453, "step": 6905 }, { "epoch": 0.2, "grad_norm": 0.5062236785888672, "learning_rate": 9.036019293402697e-05, "loss": 2.9195, "step": 6910 }, { "epoch": 0.2, "grad_norm": 0.4511181116104126, "learning_rate": 9.034650664922807e-05, "loss": 2.9715, "step": 6915 }, { "epoch": 0.2, "grad_norm": 0.4419093132019043, "learning_rate": 9.033281169368568e-05, "loss": 2.9531, "step": 6920 }, { "epoch": 0.2, "grad_norm": 0.6523056626319885, "learning_rate": 9.031910807034296e-05, "loss": 2.9984, "step": 6925 }, { "epoch": 0.2, "grad_norm": 0.45996445417404175, "learning_rate": 9.03053957821449e-05, "loss": 2.9792, "step": 6930 }, { "epoch": 0.2, "grad_norm": 0.47275975346565247, "learning_rate": 9.029167483203837e-05, "loss": 2.923, "step": 6935 }, { "epoch": 0.2, "grad_norm": 0.4936447739601135, "learning_rate": 9.027794522297211e-05, "loss": 2.9693, "step": 6940 }, { "epoch": 0.2, "grad_norm": 0.4772507846355438, "learning_rate": 9.026420695789668e-05, "loss": 3.0293, "step": 6945 }, { "epoch": 0.2, "grad_norm": 0.5134108066558838, "learning_rate": 9.025046003976457e-05, "loss": 2.9314, "step": 6950 }, { "epoch": 0.2, "grad_norm": 0.47191184759140015, "learning_rate": 9.023670447153003e-05, "loss": 2.9772, "step": 6955 }, { "epoch": 0.2, "grad_norm": 0.5327199101448059, "learning_rate": 9.022294025614926e-05, "loss": 2.8785, "step": 6960 }, { "epoch": 0.2, "grad_norm": 0.4714873433113098, "learning_rate": 9.02091673965803e-05, "loss": 2.9737, "step": 6965 }, { "epoch": 0.21, "grad_norm": 0.49054715037345886, "learning_rate": 9.0195385895783e-05, "loss": 2.9196, "step": 6970 }, { "epoch": 0.21, "grad_norm": 0.488847017288208, "learning_rate": 9.018159575671914e-05, "loss": 2.9635, "step": 6975 }, { "epoch": 0.21, "grad_norm": 0.5730308890342712, "learning_rate": 9.016779698235227e-05, "loss": 2.9808, "step": 6980 }, { "epoch": 0.21, "grad_norm": 0.44021111726760864, "learning_rate": 9.015398957564787e-05, "loss": 2.9872, "step": 6985 }, { "epoch": 0.21, "grad_norm": 0.4310663938522339, "learning_rate": 9.014017353957324e-05, "loss": 2.835, "step": 6990 }, { "epoch": 0.21, "grad_norm": 0.47780075669288635, "learning_rate": 9.012634887709754e-05, "loss": 2.8727, "step": 6995 }, { "epoch": 0.21, "grad_norm": 0.4927649199962616, "learning_rate": 9.01125155911918e-05, "loss": 2.9305, "step": 7000 }, { "epoch": 0.21, "grad_norm": 0.4666902720928192, "learning_rate": 9.009867368482887e-05, "loss": 2.9484, "step": 7005 }, { "epoch": 0.21, "grad_norm": 0.4701758623123169, "learning_rate": 9.008482316098348e-05, "loss": 2.9398, "step": 7010 }, { "epoch": 0.21, "grad_norm": 0.48070985078811646, "learning_rate": 9.007096402263221e-05, "loss": 3.0277, "step": 7015 }, { "epoch": 0.21, "grad_norm": 0.5022538900375366, "learning_rate": 9.005709627275348e-05, "loss": 2.9404, "step": 7020 }, { "epoch": 0.21, "grad_norm": 0.4779338538646698, "learning_rate": 9.004321991432753e-05, "loss": 2.9569, "step": 7025 }, { "epoch": 0.21, "grad_norm": 0.45550164580345154, "learning_rate": 9.002933495033657e-05, "loss": 2.9857, "step": 7030 }, { "epoch": 0.21, "grad_norm": 0.47266799211502075, "learning_rate": 9.001544138376447e-05, "loss": 2.819, "step": 7035 }, { "epoch": 0.21, "grad_norm": 0.4292304515838623, "learning_rate": 9.000153921759714e-05, "loss": 2.9555, "step": 7040 }, { "epoch": 0.21, "grad_norm": 0.4516564905643463, "learning_rate": 8.998762845482223e-05, "loss": 2.9593, "step": 7045 }, { "epoch": 0.21, "grad_norm": 0.45262211561203003, "learning_rate": 8.997370909842921e-05, "loss": 2.995, "step": 7050 }, { "epoch": 0.21, "grad_norm": 0.43858468532562256, "learning_rate": 8.99597811514095e-05, "loss": 2.9314, "step": 7055 }, { "epoch": 0.21, "grad_norm": 0.4946633577346802, "learning_rate": 8.99458446167563e-05, "loss": 2.9383, "step": 7060 }, { "epoch": 0.21, "grad_norm": 0.48850664496421814, "learning_rate": 8.993189949746468e-05, "loss": 2.9542, "step": 7065 }, { "epoch": 0.21, "grad_norm": 0.44823333621025085, "learning_rate": 8.99179457965315e-05, "loss": 2.9847, "step": 7070 }, { "epoch": 0.21, "grad_norm": 0.48472633957862854, "learning_rate": 8.990398351695555e-05, "loss": 2.9255, "step": 7075 }, { "epoch": 0.21, "grad_norm": 0.4536420702934265, "learning_rate": 8.989001266173738e-05, "loss": 2.8845, "step": 7080 }, { "epoch": 0.21, "grad_norm": 0.4765770435333252, "learning_rate": 8.987603323387948e-05, "loss": 2.9308, "step": 7085 }, { "epoch": 0.21, "grad_norm": 0.45561283826828003, "learning_rate": 8.986204523638606e-05, "loss": 2.93, "step": 7090 }, { "epoch": 0.21, "grad_norm": 0.489464670419693, "learning_rate": 8.984804867226328e-05, "loss": 2.9041, "step": 7095 }, { "epoch": 0.21, "grad_norm": 0.4473949670791626, "learning_rate": 8.98340435445191e-05, "loss": 2.9912, "step": 7100 }, { "epoch": 0.21, "grad_norm": 0.4628908634185791, "learning_rate": 8.982002985616331e-05, "loss": 2.9619, "step": 7105 }, { "epoch": 0.21, "grad_norm": 0.4352634847164154, "learning_rate": 8.980600761020754e-05, "loss": 2.9064, "step": 7110 }, { "epoch": 0.21, "grad_norm": 0.4561644196510315, "learning_rate": 8.979197680966527e-05, "loss": 2.9581, "step": 7115 }, { "epoch": 0.21, "grad_norm": 0.4591280221939087, "learning_rate": 8.977793745755182e-05, "loss": 3.0121, "step": 7120 }, { "epoch": 0.21, "grad_norm": 0.45428478717803955, "learning_rate": 8.976388955688434e-05, "loss": 2.8871, "step": 7125 }, { "epoch": 0.21, "grad_norm": 0.4693281352519989, "learning_rate": 8.974983311068184e-05, "loss": 2.8977, "step": 7130 }, { "epoch": 0.21, "grad_norm": 0.4521077573299408, "learning_rate": 8.973576812196513e-05, "loss": 2.9213, "step": 7135 }, { "epoch": 0.21, "grad_norm": 0.47429290413856506, "learning_rate": 8.972169459375686e-05, "loss": 2.9146, "step": 7140 }, { "epoch": 0.21, "grad_norm": 0.4559466242790222, "learning_rate": 8.970761252908156e-05, "loss": 2.8922, "step": 7145 }, { "epoch": 0.21, "grad_norm": 0.4517318308353424, "learning_rate": 8.969352193096554e-05, "loss": 2.913, "step": 7150 }, { "epoch": 0.21, "grad_norm": 0.46866947412490845, "learning_rate": 8.967942280243699e-05, "loss": 2.8893, "step": 7155 }, { "epoch": 0.21, "grad_norm": 0.4548055827617645, "learning_rate": 8.966531514652588e-05, "loss": 2.9564, "step": 7160 }, { "epoch": 0.21, "grad_norm": 0.45299315452575684, "learning_rate": 8.965119896626406e-05, "loss": 2.9019, "step": 7165 }, { "epoch": 0.21, "grad_norm": 0.45056039094924927, "learning_rate": 8.963707426468519e-05, "loss": 2.8501, "step": 7170 }, { "epoch": 0.21, "grad_norm": 0.6215821504592896, "learning_rate": 8.962294104482478e-05, "loss": 3.0253, "step": 7175 }, { "epoch": 0.21, "grad_norm": 0.502091109752655, "learning_rate": 8.960879930972014e-05, "loss": 2.9073, "step": 7180 }, { "epoch": 0.21, "grad_norm": 0.4572749733924866, "learning_rate": 8.959464906241044e-05, "loss": 2.9305, "step": 7185 }, { "epoch": 0.21, "grad_norm": 0.4659406542778015, "learning_rate": 8.958049030593665e-05, "loss": 2.9494, "step": 7190 }, { "epoch": 0.21, "grad_norm": 0.4434795677661896, "learning_rate": 8.956632304334158e-05, "loss": 2.9321, "step": 7195 }, { "epoch": 0.21, "grad_norm": 0.45167842507362366, "learning_rate": 8.955214727766989e-05, "loss": 2.8702, "step": 7200 }, { "epoch": 0.21, "grad_norm": 0.4828484058380127, "learning_rate": 8.953796301196807e-05, "loss": 2.926, "step": 7205 }, { "epoch": 0.21, "grad_norm": 0.5859330892562866, "learning_rate": 8.952377024928437e-05, "loss": 2.9068, "step": 7210 }, { "epoch": 0.21, "grad_norm": 0.45516887307167053, "learning_rate": 8.950956899266895e-05, "loss": 2.9603, "step": 7215 }, { "epoch": 0.21, "grad_norm": 0.4532371461391449, "learning_rate": 8.949535924517374e-05, "loss": 2.9967, "step": 7220 }, { "epoch": 0.21, "grad_norm": 0.44361749291419983, "learning_rate": 8.948114100985252e-05, "loss": 2.9986, "step": 7225 }, { "epoch": 0.21, "grad_norm": 0.4483746886253357, "learning_rate": 8.946691428976088e-05, "loss": 2.9677, "step": 7230 }, { "epoch": 0.21, "grad_norm": 0.4466780424118042, "learning_rate": 8.945267908795625e-05, "loss": 2.8593, "step": 7235 }, { "epoch": 0.21, "grad_norm": 0.4502350389957428, "learning_rate": 8.943843540749785e-05, "loss": 2.9512, "step": 7240 }, { "epoch": 0.21, "grad_norm": 0.4534597098827362, "learning_rate": 8.942418325144679e-05, "loss": 3.0079, "step": 7245 }, { "epoch": 0.21, "grad_norm": 0.5048699975013733, "learning_rate": 8.940992262286592e-05, "loss": 2.907, "step": 7250 }, { "epoch": 0.21, "grad_norm": 0.4674881398677826, "learning_rate": 8.939565352481996e-05, "loss": 2.9634, "step": 7255 }, { "epoch": 0.21, "grad_norm": 0.44938409328460693, "learning_rate": 8.938137596037544e-05, "loss": 2.98, "step": 7260 }, { "epoch": 0.21, "grad_norm": 0.45789435505867004, "learning_rate": 8.93670899326007e-05, "loss": 2.8968, "step": 7265 }, { "epoch": 0.21, "grad_norm": 0.4755264222621918, "learning_rate": 8.93527954445659e-05, "loss": 2.9832, "step": 7270 }, { "epoch": 0.21, "grad_norm": 0.4380793273448944, "learning_rate": 8.933849249934306e-05, "loss": 2.886, "step": 7275 }, { "epoch": 0.21, "grad_norm": 0.47951391339302063, "learning_rate": 8.932418110000594e-05, "loss": 2.8896, "step": 7280 }, { "epoch": 0.21, "grad_norm": 0.4556163549423218, "learning_rate": 8.930986124963016e-05, "loss": 2.8306, "step": 7285 }, { "epoch": 0.21, "grad_norm": 0.9070758819580078, "learning_rate": 8.92955329512932e-05, "loss": 2.993, "step": 7290 }, { "epoch": 0.21, "grad_norm": 0.4507231116294861, "learning_rate": 8.928119620807426e-05, "loss": 2.9621, "step": 7295 }, { "epoch": 0.21, "grad_norm": 0.4422430992126465, "learning_rate": 8.926685102305443e-05, "loss": 2.9482, "step": 7300 }, { "epoch": 0.21, "grad_norm": 0.46826139092445374, "learning_rate": 8.925249739931659e-05, "loss": 2.9345, "step": 7305 }, { "epoch": 0.22, "grad_norm": 0.4998292028903961, "learning_rate": 8.923813533994543e-05, "loss": 2.9193, "step": 7310 }, { "epoch": 0.22, "grad_norm": 0.43495914340019226, "learning_rate": 8.922376484802745e-05, "loss": 2.8998, "step": 7315 }, { "epoch": 0.22, "grad_norm": 0.4390137493610382, "learning_rate": 8.920938592665098e-05, "loss": 2.9854, "step": 7320 }, { "epoch": 0.22, "grad_norm": 0.4625038802623749, "learning_rate": 8.919499857890613e-05, "loss": 2.8258, "step": 7325 }, { "epoch": 0.22, "grad_norm": 0.472850501537323, "learning_rate": 8.918060280788486e-05, "loss": 3.0413, "step": 7330 }, { "epoch": 0.22, "grad_norm": 0.4573769271373749, "learning_rate": 8.916619861668092e-05, "loss": 2.9211, "step": 7335 }, { "epoch": 0.22, "grad_norm": 0.4446994662284851, "learning_rate": 8.915178600838985e-05, "loss": 2.9966, "step": 7340 }, { "epoch": 0.22, "grad_norm": 0.45436859130859375, "learning_rate": 8.913736498610906e-05, "loss": 2.9162, "step": 7345 }, { "epoch": 0.22, "grad_norm": 0.4972015619277954, "learning_rate": 8.912293555293767e-05, "loss": 2.9851, "step": 7350 }, { "epoch": 0.22, "grad_norm": 0.49918612837791443, "learning_rate": 8.910849771197671e-05, "loss": 2.9121, "step": 7355 }, { "epoch": 0.22, "grad_norm": 0.4337681233882904, "learning_rate": 8.909405146632896e-05, "loss": 2.8175, "step": 7360 }, { "epoch": 0.22, "grad_norm": 0.47636768221855164, "learning_rate": 8.907959681909902e-05, "loss": 3.0178, "step": 7365 }, { "epoch": 0.22, "grad_norm": 0.46312278509140015, "learning_rate": 8.906513377339327e-05, "loss": 2.8965, "step": 7370 }, { "epoch": 0.22, "grad_norm": 0.4651830196380615, "learning_rate": 8.905066233231995e-05, "loss": 2.9181, "step": 7375 }, { "epoch": 0.22, "grad_norm": 0.4596237540245056, "learning_rate": 8.903618249898906e-05, "loss": 2.9807, "step": 7380 }, { "epoch": 0.22, "grad_norm": 0.4380989670753479, "learning_rate": 8.902169427651242e-05, "loss": 2.9644, "step": 7385 }, { "epoch": 0.22, "grad_norm": 0.43792226910591125, "learning_rate": 8.900719766800365e-05, "loss": 2.8803, "step": 7390 }, { "epoch": 0.22, "grad_norm": 0.44277381896972656, "learning_rate": 8.899269267657817e-05, "loss": 2.9286, "step": 7395 }, { "epoch": 0.22, "grad_norm": 0.44835683703422546, "learning_rate": 8.89781793053532e-05, "loss": 2.9526, "step": 7400 }, { "epoch": 0.22, "grad_norm": 0.5993913412094116, "learning_rate": 8.896365755744775e-05, "loss": 2.9354, "step": 7405 }, { "epoch": 0.22, "grad_norm": 0.4593859314918518, "learning_rate": 8.894912743598268e-05, "loss": 2.8702, "step": 7410 }, { "epoch": 0.22, "grad_norm": 0.46167582273483276, "learning_rate": 8.893458894408058e-05, "loss": 2.9438, "step": 7415 }, { "epoch": 0.22, "grad_norm": 0.4928232729434967, "learning_rate": 8.89200420848659e-05, "loss": 2.8969, "step": 7420 }, { "epoch": 0.22, "grad_norm": 0.45526769757270813, "learning_rate": 8.890548686146483e-05, "loss": 3.0441, "step": 7425 }, { "epoch": 0.22, "grad_norm": 0.44477003812789917, "learning_rate": 8.889092327700541e-05, "loss": 2.9901, "step": 7430 }, { "epoch": 0.22, "grad_norm": 0.4587313234806061, "learning_rate": 8.887635133461745e-05, "loss": 2.9544, "step": 7435 }, { "epoch": 0.22, "grad_norm": 0.46520110964775085, "learning_rate": 8.886177103743257e-05, "loss": 2.9031, "step": 7440 }, { "epoch": 0.22, "grad_norm": 0.4482729434967041, "learning_rate": 8.884718238858416e-05, "loss": 2.9382, "step": 7445 }, { "epoch": 0.22, "grad_norm": 0.4511565864086151, "learning_rate": 8.883258539120742e-05, "loss": 2.9461, "step": 7450 }, { "epoch": 0.22, "grad_norm": 0.4647030830383301, "learning_rate": 8.881798004843938e-05, "loss": 2.8062, "step": 7455 }, { "epoch": 0.22, "grad_norm": 0.40865373611450195, "learning_rate": 8.880336636341878e-05, "loss": 2.8114, "step": 7460 }, { "epoch": 0.22, "grad_norm": 0.4731404781341553, "learning_rate": 8.878874433928623e-05, "loss": 2.9027, "step": 7465 }, { "epoch": 0.22, "grad_norm": 0.43039804697036743, "learning_rate": 8.877411397918411e-05, "loss": 2.9697, "step": 7470 }, { "epoch": 0.22, "grad_norm": 0.42162805795669556, "learning_rate": 8.875947528625658e-05, "loss": 2.8924, "step": 7475 }, { "epoch": 0.22, "grad_norm": 0.44448405504226685, "learning_rate": 8.874482826364958e-05, "loss": 2.9663, "step": 7480 }, { "epoch": 0.22, "grad_norm": 0.4422028958797455, "learning_rate": 8.873017291451086e-05, "loss": 2.8292, "step": 7485 }, { "epoch": 0.22, "grad_norm": 0.46044015884399414, "learning_rate": 8.871550924198999e-05, "loss": 2.9903, "step": 7490 }, { "epoch": 0.22, "grad_norm": 0.46944817900657654, "learning_rate": 8.870083724923826e-05, "loss": 2.9469, "step": 7495 }, { "epoch": 0.22, "grad_norm": 0.42327573895454407, "learning_rate": 8.868615693940879e-05, "loss": 2.9654, "step": 7500 }, { "epoch": 0.22, "grad_norm": 0.46775275468826294, "learning_rate": 8.86714683156565e-05, "loss": 2.9292, "step": 7505 }, { "epoch": 0.22, "grad_norm": 0.44510504603385925, "learning_rate": 8.865677138113803e-05, "loss": 2.8743, "step": 7510 }, { "epoch": 0.22, "grad_norm": 0.4568474292755127, "learning_rate": 8.864206613901189e-05, "loss": 2.9417, "step": 7515 }, { "epoch": 0.22, "grad_norm": 0.4678153693675995, "learning_rate": 8.862735259243832e-05, "loss": 2.876, "step": 7520 }, { "epoch": 0.22, "grad_norm": 0.4582410752773285, "learning_rate": 8.861263074457941e-05, "loss": 2.8778, "step": 7525 }, { "epoch": 0.22, "grad_norm": 0.4348117709159851, "learning_rate": 8.859790059859893e-05, "loss": 3.0082, "step": 7530 }, { "epoch": 0.22, "grad_norm": 0.42666441202163696, "learning_rate": 8.858316215766252e-05, "loss": 2.8959, "step": 7535 }, { "epoch": 0.22, "grad_norm": 0.4482584297657013, "learning_rate": 8.856841542493754e-05, "loss": 2.9013, "step": 7540 }, { "epoch": 0.22, "grad_norm": 0.4473045766353607, "learning_rate": 8.855366040359321e-05, "loss": 2.8924, "step": 7545 }, { "epoch": 0.22, "grad_norm": 0.451092004776001, "learning_rate": 8.853889709680045e-05, "loss": 2.8959, "step": 7550 }, { "epoch": 0.22, "grad_norm": 0.46163779497146606, "learning_rate": 8.852412550773202e-05, "loss": 2.9413, "step": 7555 }, { "epoch": 0.22, "grad_norm": 0.45264533162117004, "learning_rate": 8.850934563956244e-05, "loss": 2.9236, "step": 7560 }, { "epoch": 0.22, "grad_norm": 0.4523696005344391, "learning_rate": 8.849455749546797e-05, "loss": 2.9063, "step": 7565 }, { "epoch": 0.22, "grad_norm": 0.5099847316741943, "learning_rate": 8.847976107862673e-05, "loss": 2.9605, "step": 7570 }, { "epoch": 0.22, "grad_norm": 0.46268847584724426, "learning_rate": 8.846495639221854e-05, "loss": 2.9971, "step": 7575 }, { "epoch": 0.22, "grad_norm": 0.4633960723876953, "learning_rate": 8.845014343942505e-05, "loss": 2.9676, "step": 7580 }, { "epoch": 0.22, "grad_norm": 0.46767738461494446, "learning_rate": 8.843532222342965e-05, "loss": 2.9178, "step": 7585 }, { "epoch": 0.22, "grad_norm": 0.4438166916370392, "learning_rate": 8.842049274741753e-05, "loss": 3.0433, "step": 7590 }, { "epoch": 0.22, "grad_norm": 0.4861506521701813, "learning_rate": 8.840565501457565e-05, "loss": 2.8726, "step": 7595 }, { "epoch": 0.22, "grad_norm": 0.430468887090683, "learning_rate": 8.839080902809274e-05, "loss": 2.921, "step": 7600 }, { "epoch": 0.22, "grad_norm": 0.4771009087562561, "learning_rate": 8.83759547911593e-05, "loss": 2.8253, "step": 7605 }, { "epoch": 0.22, "grad_norm": 0.46754589676856995, "learning_rate": 8.83610923069676e-05, "loss": 2.838, "step": 7610 }, { "epoch": 0.22, "grad_norm": 0.45665979385375977, "learning_rate": 8.83462215787117e-05, "loss": 2.9275, "step": 7615 }, { "epoch": 0.22, "grad_norm": 0.5135639309883118, "learning_rate": 8.833134260958745e-05, "loss": 2.9317, "step": 7620 }, { "epoch": 0.22, "grad_norm": 0.4757919907569885, "learning_rate": 8.831645540279238e-05, "loss": 2.8925, "step": 7625 }, { "epoch": 0.22, "grad_norm": 0.5027908682823181, "learning_rate": 8.830155996152591e-05, "loss": 2.7988, "step": 7630 }, { "epoch": 0.22, "grad_norm": 0.4380546808242798, "learning_rate": 8.828665628898918e-05, "loss": 2.7961, "step": 7635 }, { "epoch": 0.22, "grad_norm": 0.4832897186279297, "learning_rate": 8.827174438838503e-05, "loss": 2.9442, "step": 7640 }, { "epoch": 0.22, "grad_norm": 0.5382148027420044, "learning_rate": 8.825682426291819e-05, "loss": 2.9363, "step": 7645 }, { "epoch": 0.23, "grad_norm": 0.4504719376564026, "learning_rate": 8.824189591579505e-05, "loss": 2.7752, "step": 7650 }, { "epoch": 0.23, "grad_norm": 0.4627530574798584, "learning_rate": 8.822695935022387e-05, "loss": 2.864, "step": 7655 }, { "epoch": 0.23, "grad_norm": 0.44624432921409607, "learning_rate": 8.821201456941458e-05, "loss": 2.8754, "step": 7660 }, { "epoch": 0.23, "grad_norm": 0.4356198012828827, "learning_rate": 8.819706157657893e-05, "loss": 2.8587, "step": 7665 }, { "epoch": 0.23, "grad_norm": 0.4348682165145874, "learning_rate": 8.818210037493042e-05, "loss": 2.9985, "step": 7670 }, { "epoch": 0.23, "grad_norm": 0.4344383776187897, "learning_rate": 8.816713096768431e-05, "loss": 2.8192, "step": 7675 }, { "epoch": 0.23, "grad_norm": 0.45120200514793396, "learning_rate": 8.815215335805765e-05, "loss": 2.9429, "step": 7680 }, { "epoch": 0.23, "grad_norm": 0.47065722942352295, "learning_rate": 8.81371675492692e-05, "loss": 2.9253, "step": 7685 }, { "epoch": 0.23, "grad_norm": 0.48139718174934387, "learning_rate": 8.812217354453956e-05, "loss": 2.9643, "step": 7690 }, { "epoch": 0.23, "grad_norm": 0.42573028802871704, "learning_rate": 8.810717134709099e-05, "loss": 2.9439, "step": 7695 }, { "epoch": 0.23, "grad_norm": 0.4795282185077667, "learning_rate": 8.809216096014759e-05, "loss": 2.9586, "step": 7700 }, { "epoch": 0.23, "grad_norm": 0.47545498609542847, "learning_rate": 8.807714238693521e-05, "loss": 2.9636, "step": 7705 }, { "epoch": 0.23, "grad_norm": 0.4640045464038849, "learning_rate": 8.806211563068145e-05, "loss": 2.9223, "step": 7710 }, { "epoch": 0.23, "grad_norm": 0.4597218930721283, "learning_rate": 8.804708069461563e-05, "loss": 2.9094, "step": 7715 }, { "epoch": 0.23, "grad_norm": 0.44956883788108826, "learning_rate": 8.803203758196887e-05, "loss": 2.9144, "step": 7720 }, { "epoch": 0.23, "grad_norm": 0.5148975849151611, "learning_rate": 8.801698629597407e-05, "loss": 2.9533, "step": 7725 }, { "epoch": 0.23, "grad_norm": 0.44650816917419434, "learning_rate": 8.800192683986583e-05, "loss": 2.9102, "step": 7730 }, { "epoch": 0.23, "grad_norm": 0.4833897650241852, "learning_rate": 8.798685921688053e-05, "loss": 2.9038, "step": 7735 }, { "epoch": 0.23, "grad_norm": 0.4616844058036804, "learning_rate": 8.797178343025631e-05, "loss": 2.9411, "step": 7740 }, { "epoch": 0.23, "grad_norm": 0.42749759554862976, "learning_rate": 8.795669948323308e-05, "loss": 2.9496, "step": 7745 }, { "epoch": 0.23, "grad_norm": 0.4483602046966553, "learning_rate": 8.794160737905245e-05, "loss": 2.8863, "step": 7750 }, { "epoch": 0.23, "grad_norm": 0.43929412961006165, "learning_rate": 8.792650712095784e-05, "loss": 2.8399, "step": 7755 }, { "epoch": 0.23, "grad_norm": 0.4711315333843231, "learning_rate": 8.791139871219441e-05, "loss": 2.8687, "step": 7760 }, { "epoch": 0.23, "grad_norm": 0.47226446866989136, "learning_rate": 8.789628215600902e-05, "loss": 2.9427, "step": 7765 }, { "epoch": 0.23, "grad_norm": 0.46355098485946655, "learning_rate": 8.788115745565036e-05, "loss": 2.8801, "step": 7770 }, { "epoch": 0.23, "grad_norm": 0.4372268319129944, "learning_rate": 8.786602461436884e-05, "loss": 2.8462, "step": 7775 }, { "epoch": 0.23, "grad_norm": 0.44118043780326843, "learning_rate": 8.785088363541657e-05, "loss": 2.8791, "step": 7780 }, { "epoch": 0.23, "grad_norm": 0.4596155881881714, "learning_rate": 8.78357345220475e-05, "loss": 2.954, "step": 7785 }, { "epoch": 0.23, "grad_norm": 0.44278618693351746, "learning_rate": 8.782057727751722e-05, "loss": 2.9151, "step": 7790 }, { "epoch": 0.23, "grad_norm": 0.45208224654197693, "learning_rate": 8.78054119050832e-05, "loss": 2.9043, "step": 7795 }, { "epoch": 0.23, "grad_norm": 0.4443177878856659, "learning_rate": 8.779023840800452e-05, "loss": 2.8729, "step": 7800 }, { "epoch": 0.23, "grad_norm": 0.4190120995044708, "learning_rate": 8.777505678954211e-05, "loss": 2.8818, "step": 7805 }, { "epoch": 0.23, "grad_norm": 0.47247540950775146, "learning_rate": 8.775986705295858e-05, "loss": 2.9576, "step": 7810 }, { "epoch": 0.23, "grad_norm": 0.4241117537021637, "learning_rate": 8.774466920151833e-05, "loss": 2.8537, "step": 7815 }, { "epoch": 0.23, "grad_norm": 0.5121718049049377, "learning_rate": 8.772946323848746e-05, "loss": 2.8199, "step": 7820 }, { "epoch": 0.23, "grad_norm": 0.42898470163345337, "learning_rate": 8.771424916713386e-05, "loss": 2.9061, "step": 7825 }, { "epoch": 0.23, "grad_norm": 0.4626065790653229, "learning_rate": 8.769902699072712e-05, "loss": 2.8735, "step": 7830 }, { "epoch": 0.23, "grad_norm": 0.45111188292503357, "learning_rate": 8.768379671253859e-05, "loss": 2.9366, "step": 7835 }, { "epoch": 0.23, "grad_norm": 0.44420769810676575, "learning_rate": 8.76685583358414e-05, "loss": 2.8923, "step": 7840 }, { "epoch": 0.23, "grad_norm": 0.4473263621330261, "learning_rate": 8.765331186391033e-05, "loss": 2.8287, "step": 7845 }, { "epoch": 0.23, "grad_norm": 0.4404882490634918, "learning_rate": 8.763805730002199e-05, "loss": 2.8711, "step": 7850 }, { "epoch": 0.23, "grad_norm": 0.43105772137641907, "learning_rate": 8.762279464745466e-05, "loss": 2.9368, "step": 7855 }, { "epoch": 0.23, "grad_norm": 0.43499937653541565, "learning_rate": 8.760752390948843e-05, "loss": 2.8629, "step": 7860 }, { "epoch": 0.23, "grad_norm": 0.43869948387145996, "learning_rate": 8.759224508940505e-05, "loss": 2.8528, "step": 7865 }, { "epoch": 0.23, "grad_norm": 0.44875940680503845, "learning_rate": 8.757695819048808e-05, "loss": 2.929, "step": 7870 }, { "epoch": 0.23, "grad_norm": 0.4563370943069458, "learning_rate": 8.756166321602273e-05, "loss": 2.8673, "step": 7875 }, { "epoch": 0.23, "grad_norm": 0.42610496282577515, "learning_rate": 8.754636016929604e-05, "loss": 2.8742, "step": 7880 }, { "epoch": 0.23, "grad_norm": 0.44013074040412903, "learning_rate": 8.753104905359674e-05, "loss": 2.8191, "step": 7885 }, { "epoch": 0.23, "grad_norm": 0.4431072771549225, "learning_rate": 8.751572987221526e-05, "loss": 2.8708, "step": 7890 }, { "epoch": 0.23, "grad_norm": 0.4185353219509125, "learning_rate": 8.750040262844384e-05, "loss": 2.9158, "step": 7895 }, { "epoch": 0.23, "grad_norm": 0.457633912563324, "learning_rate": 8.748506732557637e-05, "loss": 2.9388, "step": 7900 }, { "epoch": 0.23, "grad_norm": 0.4381994605064392, "learning_rate": 8.746972396690853e-05, "loss": 2.8775, "step": 7905 }, { "epoch": 0.23, "grad_norm": 0.43807053565979004, "learning_rate": 8.745437255573773e-05, "loss": 2.9173, "step": 7910 }, { "epoch": 0.23, "grad_norm": 0.453977108001709, "learning_rate": 8.743901309536307e-05, "loss": 2.9033, "step": 7915 }, { "epoch": 0.23, "grad_norm": 0.44279560446739197, "learning_rate": 8.742364558908543e-05, "loss": 2.9122, "step": 7920 }, { "epoch": 0.23, "grad_norm": 0.4332951009273529, "learning_rate": 8.740827004020735e-05, "loss": 2.8738, "step": 7925 }, { "epoch": 0.23, "grad_norm": 0.4265264868736267, "learning_rate": 8.739288645203319e-05, "loss": 2.8668, "step": 7930 }, { "epoch": 0.23, "grad_norm": 0.4492313861846924, "learning_rate": 8.737749482786897e-05, "loss": 2.8756, "step": 7935 }, { "epoch": 0.23, "grad_norm": 0.4224439561367035, "learning_rate": 8.736209517102247e-05, "loss": 2.8485, "step": 7940 }, { "epoch": 0.23, "grad_norm": 0.43115100264549255, "learning_rate": 8.734668748480315e-05, "loss": 2.8326, "step": 7945 }, { "epoch": 0.23, "grad_norm": 0.5153506398200989, "learning_rate": 8.733127177252225e-05, "loss": 2.8583, "step": 7950 }, { "epoch": 0.23, "grad_norm": 0.470271497964859, "learning_rate": 8.731584803749271e-05, "loss": 2.822, "step": 7955 }, { "epoch": 0.23, "grad_norm": 0.42905890941619873, "learning_rate": 8.73004162830292e-05, "loss": 2.8028, "step": 7960 }, { "epoch": 0.23, "grad_norm": 0.44920849800109863, "learning_rate": 8.728497651244812e-05, "loss": 2.8725, "step": 7965 }, { "epoch": 0.23, "grad_norm": 0.42933303117752075, "learning_rate": 8.726952872906755e-05, "loss": 2.9124, "step": 7970 }, { "epoch": 0.23, "grad_norm": 0.448479026556015, "learning_rate": 8.725407293620737e-05, "loss": 2.765, "step": 7975 }, { "epoch": 0.23, "grad_norm": 0.5003622174263, "learning_rate": 8.72386091371891e-05, "loss": 2.8414, "step": 7980 }, { "epoch": 0.23, "grad_norm": 0.4555703401565552, "learning_rate": 8.722313733533603e-05, "loss": 2.812, "step": 7985 }, { "epoch": 0.24, "grad_norm": 0.4524518847465515, "learning_rate": 8.720765753397317e-05, "loss": 2.8961, "step": 7990 }, { "epoch": 0.24, "grad_norm": 0.43950599431991577, "learning_rate": 8.719216973642723e-05, "loss": 2.9079, "step": 7995 }, { "epoch": 0.24, "grad_norm": 0.4454759359359741, "learning_rate": 8.717667394602664e-05, "loss": 2.8148, "step": 8000 } ], "logging_steps": 5, "max_steps": 33984, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 1000, "total_flos": 1.987874932339507e+19, "train_batch_size": 16, "trial_name": null, "trial_params": null }