diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,20833 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9999663084127893, + "eval_steps": 500, + "global_step": 14840, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 6.7383174421347e-05, + "grad_norm": 44.709284858028134, + "learning_rate": 6.738544474393531e-10, + "loss": 3.4965, + "step": 1 + }, + { + "epoch": 0.00033691587210673494, + "grad_norm": 44.640535530285916, + "learning_rate": 3.3692722371967655e-09, + "loss": 3.356, + "step": 5 + }, + { + "epoch": 0.0006738317442134699, + "grad_norm": 51.56048341252242, + "learning_rate": 6.738544474393531e-09, + "loss": 3.3383, + "step": 10 + }, + { + "epoch": 0.001010747616320205, + "grad_norm": 53.90723002649116, + "learning_rate": 1.0107816711590296e-08, + "loss": 3.3735, + "step": 15 + }, + { + "epoch": 0.0013476634884269398, + "grad_norm": 46.1101675459699, + "learning_rate": 1.3477088948787062e-08, + "loss": 3.3197, + "step": 20 + }, + { + "epoch": 0.0016845793605336748, + "grad_norm": 43.15160877782767, + "learning_rate": 1.6846361185983825e-08, + "loss": 3.2933, + "step": 25 + }, + { + "epoch": 0.00202149523264041, + "grad_norm": 38.262340082474964, + "learning_rate": 2.021563342318059e-08, + "loss": 3.2625, + "step": 30 + }, + { + "epoch": 0.0023584111047471445, + "grad_norm": 43.90995030755105, + "learning_rate": 2.3584905660377358e-08, + "loss": 3.3362, + "step": 35 + }, + { + "epoch": 0.0026953269768538795, + "grad_norm": 49.80340871131932, + "learning_rate": 2.6954177897574124e-08, + "loss": 3.3537, + "step": 40 + }, + { + "epoch": 0.0030322428489606146, + "grad_norm": 44.81802216376179, + "learning_rate": 3.032345013477089e-08, + "loss": 3.379, + "step": 45 + }, + { + "epoch": 0.0033691587210673496, + "grad_norm": 50.77404389674636, + "learning_rate": 3.369272237196765e-08, + "loss": 3.4307, + "step": 50 + }, + { + "epoch": 0.0037060745931740842, + "grad_norm": 46.42613238025317, + "learning_rate": 3.706199460916442e-08, + "loss": 3.3911, + "step": 55 + }, + { + "epoch": 0.00404299046528082, + "grad_norm": 46.258120740760205, + "learning_rate": 4.043126684636118e-08, + "loss": 3.2756, + "step": 60 + }, + { + "epoch": 0.004379906337387554, + "grad_norm": 43.14076176958262, + "learning_rate": 4.380053908355795e-08, + "loss": 3.355, + "step": 65 + }, + { + "epoch": 0.004716822209494289, + "grad_norm": 46.81420499567808, + "learning_rate": 4.7169811320754715e-08, + "loss": 3.3903, + "step": 70 + }, + { + "epoch": 0.005053738081601024, + "grad_norm": 44.52936544164706, + "learning_rate": 5.053908355795148e-08, + "loss": 3.2939, + "step": 75 + }, + { + "epoch": 0.005390653953707759, + "grad_norm": 34.86781526806953, + "learning_rate": 5.390835579514825e-08, + "loss": 3.198, + "step": 80 + }, + { + "epoch": 0.005727569825814494, + "grad_norm": 29.870304957206507, + "learning_rate": 5.727762803234501e-08, + "loss": 3.0484, + "step": 85 + }, + { + "epoch": 0.006064485697921229, + "grad_norm": 35.58464490708436, + "learning_rate": 6.064690026954177e-08, + "loss": 3.201, + "step": 90 + }, + { + "epoch": 0.006401401570027964, + "grad_norm": 34.646361724277995, + "learning_rate": 6.401617250673854e-08, + "loss": 3.2132, + "step": 95 + }, + { + "epoch": 0.006738317442134699, + "grad_norm": 32.994364624399495, + "learning_rate": 6.73854447439353e-08, + "loss": 3.225, + "step": 100 + }, + { + "epoch": 0.007075233314241434, + "grad_norm": 33.94119605988894, + "learning_rate": 7.075471698113207e-08, + "loss": 3.2271, + "step": 105 + }, + { + "epoch": 0.0074121491863481685, + "grad_norm": 22.362130330257422, + "learning_rate": 7.412398921832884e-08, + "loss": 3.0439, + "step": 110 + }, + { + "epoch": 0.0077490650584549035, + "grad_norm": 20.209448650967722, + "learning_rate": 7.749326145552561e-08, + "loss": 3.015, + "step": 115 + }, + { + "epoch": 0.00808598093056164, + "grad_norm": 19.57320336750467, + "learning_rate": 8.086253369272237e-08, + "loss": 2.9764, + "step": 120 + }, + { + "epoch": 0.008422896802668374, + "grad_norm": 18.543985000117726, + "learning_rate": 8.423180592991913e-08, + "loss": 2.9175, + "step": 125 + }, + { + "epoch": 0.008759812674775108, + "grad_norm": 18.684249606371242, + "learning_rate": 8.76010781671159e-08, + "loss": 3.0318, + "step": 130 + }, + { + "epoch": 0.009096728546881844, + "grad_norm": 16.654807296468853, + "learning_rate": 9.097035040431267e-08, + "loss": 2.9674, + "step": 135 + }, + { + "epoch": 0.009433644418988578, + "grad_norm": 15.119507215593755, + "learning_rate": 9.433962264150943e-08, + "loss": 2.8187, + "step": 140 + }, + { + "epoch": 0.009770560291095314, + "grad_norm": 13.383270486575368, + "learning_rate": 9.770889487870619e-08, + "loss": 2.8611, + "step": 145 + }, + { + "epoch": 0.010107476163202048, + "grad_norm": 11.350525217429713, + "learning_rate": 1.0107816711590296e-07, + "loss": 2.7627, + "step": 150 + }, + { + "epoch": 0.010444392035308784, + "grad_norm": 10.567978119152889, + "learning_rate": 1.0444743935309973e-07, + "loss": 2.7982, + "step": 155 + }, + { + "epoch": 0.010781307907415518, + "grad_norm": 11.623938979272912, + "learning_rate": 1.078167115902965e-07, + "loss": 2.7693, + "step": 160 + }, + { + "epoch": 0.011118223779522254, + "grad_norm": 10.879180679917159, + "learning_rate": 1.1118598382749325e-07, + "loss": 2.7249, + "step": 165 + }, + { + "epoch": 0.011455139651628988, + "grad_norm": 10.08425226855161, + "learning_rate": 1.1455525606469002e-07, + "loss": 2.7114, + "step": 170 + }, + { + "epoch": 0.011792055523735722, + "grad_norm": 10.568577238407425, + "learning_rate": 1.1792452830188679e-07, + "loss": 2.7127, + "step": 175 + }, + { + "epoch": 0.012128971395842458, + "grad_norm": 8.851324037587174, + "learning_rate": 1.2129380053908355e-07, + "loss": 2.7347, + "step": 180 + }, + { + "epoch": 0.012465887267949192, + "grad_norm": 9.055687625401953, + "learning_rate": 1.2466307277628032e-07, + "loss": 2.7112, + "step": 185 + }, + { + "epoch": 0.012802803140055928, + "grad_norm": 9.473681984573032, + "learning_rate": 1.280323450134771e-07, + "loss": 2.6401, + "step": 190 + }, + { + "epoch": 0.013139719012162663, + "grad_norm": 8.258411293131203, + "learning_rate": 1.3140161725067383e-07, + "loss": 2.6933, + "step": 195 + }, + { + "epoch": 0.013476634884269399, + "grad_norm": 8.251773198341327, + "learning_rate": 1.347708894878706e-07, + "loss": 2.6369, + "step": 200 + }, + { + "epoch": 0.013813550756376133, + "grad_norm": 8.076392324410342, + "learning_rate": 1.3814016172506737e-07, + "loss": 2.6968, + "step": 205 + }, + { + "epoch": 0.014150466628482869, + "grad_norm": 8.399380306646764, + "learning_rate": 1.4150943396226414e-07, + "loss": 2.6753, + "step": 210 + }, + { + "epoch": 0.014487382500589603, + "grad_norm": 8.114213020798976, + "learning_rate": 1.448787061994609e-07, + "loss": 2.5847, + "step": 215 + }, + { + "epoch": 0.014824298372696337, + "grad_norm": 9.006493080130776, + "learning_rate": 1.4824797843665768e-07, + "loss": 2.6046, + "step": 220 + }, + { + "epoch": 0.015161214244803073, + "grad_norm": 9.572718178175688, + "learning_rate": 1.5161725067385445e-07, + "loss": 2.577, + "step": 225 + }, + { + "epoch": 0.015498130116909807, + "grad_norm": 8.010214290776423, + "learning_rate": 1.5498652291105122e-07, + "loss": 2.5808, + "step": 230 + }, + { + "epoch": 0.01583504598901654, + "grad_norm": 8.65551262467679, + "learning_rate": 1.58355795148248e-07, + "loss": 2.5189, + "step": 235 + }, + { + "epoch": 0.01617196186112328, + "grad_norm": 7.23054831726289, + "learning_rate": 1.6172506738544473e-07, + "loss": 2.5421, + "step": 240 + }, + { + "epoch": 0.016508877733230013, + "grad_norm": 8.850682737865846, + "learning_rate": 1.650943396226415e-07, + "loss": 2.5049, + "step": 245 + }, + { + "epoch": 0.016845793605336747, + "grad_norm": 8.26897640609425, + "learning_rate": 1.6846361185983827e-07, + "loss": 2.497, + "step": 250 + }, + { + "epoch": 0.01718270947744348, + "grad_norm": 7.774269288496068, + "learning_rate": 1.7183288409703504e-07, + "loss": 2.5256, + "step": 255 + }, + { + "epoch": 0.017519625349550216, + "grad_norm": 7.363973310169826, + "learning_rate": 1.752021563342318e-07, + "loss": 2.4144, + "step": 260 + }, + { + "epoch": 0.017856541221656953, + "grad_norm": 8.517367385880418, + "learning_rate": 1.7857142857142858e-07, + "loss": 2.5122, + "step": 265 + }, + { + "epoch": 0.018193457093763687, + "grad_norm": 7.193035308993747, + "learning_rate": 1.8194070080862535e-07, + "loss": 2.4599, + "step": 270 + }, + { + "epoch": 0.01853037296587042, + "grad_norm": 7.101515399204041, + "learning_rate": 1.853099730458221e-07, + "loss": 2.4951, + "step": 275 + }, + { + "epoch": 0.018867288837977156, + "grad_norm": 6.722538155052685, + "learning_rate": 1.8867924528301886e-07, + "loss": 2.4476, + "step": 280 + }, + { + "epoch": 0.019204204710083893, + "grad_norm": 6.943285735322818, + "learning_rate": 1.920485175202156e-07, + "loss": 2.413, + "step": 285 + }, + { + "epoch": 0.019541120582190628, + "grad_norm": 7.561710369578133, + "learning_rate": 1.9541778975741237e-07, + "loss": 2.4349, + "step": 290 + }, + { + "epoch": 0.019878036454297362, + "grad_norm": 8.469766417330744, + "learning_rate": 1.9878706199460914e-07, + "loss": 2.4515, + "step": 295 + }, + { + "epoch": 0.020214952326404096, + "grad_norm": 7.543991515105139, + "learning_rate": 2.021563342318059e-07, + "loss": 2.4827, + "step": 300 + }, + { + "epoch": 0.02055186819851083, + "grad_norm": 6.931353522398812, + "learning_rate": 2.0552560646900268e-07, + "loss": 2.3415, + "step": 305 + }, + { + "epoch": 0.020888784070617568, + "grad_norm": 8.001618096058628, + "learning_rate": 2.0889487870619945e-07, + "loss": 2.4162, + "step": 310 + }, + { + "epoch": 0.021225699942724302, + "grad_norm": 7.264662723671408, + "learning_rate": 2.1226415094339622e-07, + "loss": 2.4595, + "step": 315 + }, + { + "epoch": 0.021562615814831036, + "grad_norm": 7.210766841546526, + "learning_rate": 2.15633423180593e-07, + "loss": 2.4141, + "step": 320 + }, + { + "epoch": 0.02189953168693777, + "grad_norm": 8.07327269473754, + "learning_rate": 2.1900269541778973e-07, + "loss": 2.4338, + "step": 325 + }, + { + "epoch": 0.022236447559044508, + "grad_norm": 7.20614839275969, + "learning_rate": 2.223719676549865e-07, + "loss": 2.4118, + "step": 330 + }, + { + "epoch": 0.022573363431151242, + "grad_norm": 6.439804675900733, + "learning_rate": 2.2574123989218327e-07, + "loss": 2.3203, + "step": 335 + }, + { + "epoch": 0.022910279303257976, + "grad_norm": 7.145016131156796, + "learning_rate": 2.2911051212938004e-07, + "loss": 2.3943, + "step": 340 + }, + { + "epoch": 0.02324719517536471, + "grad_norm": 6.778656371306502, + "learning_rate": 2.324797843665768e-07, + "loss": 2.4236, + "step": 345 + }, + { + "epoch": 0.023584111047471445, + "grad_norm": 6.722819499030103, + "learning_rate": 2.3584905660377358e-07, + "loss": 2.364, + "step": 350 + }, + { + "epoch": 0.023921026919578182, + "grad_norm": 6.927729717159921, + "learning_rate": 2.392183288409703e-07, + "loss": 2.4121, + "step": 355 + }, + { + "epoch": 0.024257942791684917, + "grad_norm": 7.406480382151978, + "learning_rate": 2.425876010781671e-07, + "loss": 2.379, + "step": 360 + }, + { + "epoch": 0.02459485866379165, + "grad_norm": 7.47620372710672, + "learning_rate": 2.4595687331536387e-07, + "loss": 2.3708, + "step": 365 + }, + { + "epoch": 0.024931774535898385, + "grad_norm": 8.008131038928196, + "learning_rate": 2.4932614555256063e-07, + "loss": 2.3382, + "step": 370 + }, + { + "epoch": 0.025268690408005123, + "grad_norm": 7.200531642388836, + "learning_rate": 2.526954177897574e-07, + "loss": 2.304, + "step": 375 + }, + { + "epoch": 0.025605606280111857, + "grad_norm": 7.44236759253605, + "learning_rate": 2.560646900269542e-07, + "loss": 2.2896, + "step": 380 + }, + { + "epoch": 0.02594252215221859, + "grad_norm": 7.7421899686207984, + "learning_rate": 2.5943396226415094e-07, + "loss": 2.2727, + "step": 385 + }, + { + "epoch": 0.026279438024325325, + "grad_norm": 7.444999671536141, + "learning_rate": 2.6280323450134766e-07, + "loss": 2.3202, + "step": 390 + }, + { + "epoch": 0.02661635389643206, + "grad_norm": 7.695216978308538, + "learning_rate": 2.661725067385445e-07, + "loss": 2.3857, + "step": 395 + }, + { + "epoch": 0.026953269768538797, + "grad_norm": 7.069350012569017, + "learning_rate": 2.695417789757412e-07, + "loss": 2.2707, + "step": 400 + }, + { + "epoch": 0.02729018564064553, + "grad_norm": 7.068751554462215, + "learning_rate": 2.72911051212938e-07, + "loss": 2.2706, + "step": 405 + }, + { + "epoch": 0.027627101512752265, + "grad_norm": 6.929312319633547, + "learning_rate": 2.7628032345013474e-07, + "loss": 2.2942, + "step": 410 + }, + { + "epoch": 0.027964017384859, + "grad_norm": 7.258446906044968, + "learning_rate": 2.7964959568733156e-07, + "loss": 2.2749, + "step": 415 + }, + { + "epoch": 0.028300933256965737, + "grad_norm": 7.212702331192978, + "learning_rate": 2.830188679245283e-07, + "loss": 2.3326, + "step": 420 + }, + { + "epoch": 0.02863784912907247, + "grad_norm": 6.708270480284244, + "learning_rate": 2.863881401617251e-07, + "loss": 2.3198, + "step": 425 + }, + { + "epoch": 0.028974765001179206, + "grad_norm": 6.502809724253282, + "learning_rate": 2.897574123989218e-07, + "loss": 2.2732, + "step": 430 + }, + { + "epoch": 0.02931168087328594, + "grad_norm": 6.628997528648987, + "learning_rate": 2.9312668463611853e-07, + "loss": 2.2332, + "step": 435 + }, + { + "epoch": 0.029648596745392674, + "grad_norm": 7.325695254112328, + "learning_rate": 2.9649595687331536e-07, + "loss": 2.3131, + "step": 440 + }, + { + "epoch": 0.02998551261749941, + "grad_norm": 8.473803247223215, + "learning_rate": 2.9986522911051207e-07, + "loss": 2.2512, + "step": 445 + }, + { + "epoch": 0.030322428489606146, + "grad_norm": 6.921554435986179, + "learning_rate": 3.032345013477089e-07, + "loss": 2.2517, + "step": 450 + }, + { + "epoch": 0.03065934436171288, + "grad_norm": 7.30639146819376, + "learning_rate": 3.066037735849056e-07, + "loss": 2.2854, + "step": 455 + }, + { + "epoch": 0.030996260233819614, + "grad_norm": 7.431388997634885, + "learning_rate": 3.0997304582210244e-07, + "loss": 2.189, + "step": 460 + }, + { + "epoch": 0.03133317610592635, + "grad_norm": 6.792538164315586, + "learning_rate": 3.1334231805929915e-07, + "loss": 2.2807, + "step": 465 + }, + { + "epoch": 0.03167009197803308, + "grad_norm": 7.172802025644924, + "learning_rate": 3.16711590296496e-07, + "loss": 2.2403, + "step": 470 + }, + { + "epoch": 0.03200700785013982, + "grad_norm": 7.396610718898199, + "learning_rate": 3.200808625336927e-07, + "loss": 2.2779, + "step": 475 + }, + { + "epoch": 0.03234392372224656, + "grad_norm": 6.645540214064375, + "learning_rate": 3.2345013477088946e-07, + "loss": 2.275, + "step": 480 + }, + { + "epoch": 0.03268083959435329, + "grad_norm": 6.9885548745948896, + "learning_rate": 3.2681940700808623e-07, + "loss": 2.263, + "step": 485 + }, + { + "epoch": 0.033017755466460026, + "grad_norm": 6.83925180479448, + "learning_rate": 3.30188679245283e-07, + "loss": 2.2113, + "step": 490 + }, + { + "epoch": 0.03335467133856676, + "grad_norm": 6.814370294667984, + "learning_rate": 3.3355795148247977e-07, + "loss": 2.298, + "step": 495 + }, + { + "epoch": 0.033691587210673495, + "grad_norm": 6.750421685926166, + "learning_rate": 3.3692722371967654e-07, + "loss": 2.2178, + "step": 500 + }, + { + "epoch": 0.03402850308278023, + "grad_norm": 7.5770117147083225, + "learning_rate": 3.402964959568733e-07, + "loss": 2.1977, + "step": 505 + }, + { + "epoch": 0.03436541895488696, + "grad_norm": 7.294251237933485, + "learning_rate": 3.436657681940701e-07, + "loss": 2.2226, + "step": 510 + }, + { + "epoch": 0.0347023348269937, + "grad_norm": 6.998637054729564, + "learning_rate": 3.4703504043126685e-07, + "loss": 2.2606, + "step": 515 + }, + { + "epoch": 0.03503925069910043, + "grad_norm": 6.680594853350787, + "learning_rate": 3.504043126684636e-07, + "loss": 2.2916, + "step": 520 + }, + { + "epoch": 0.03537616657120717, + "grad_norm": 7.092023051416907, + "learning_rate": 3.5377358490566033e-07, + "loss": 2.2562, + "step": 525 + }, + { + "epoch": 0.03571308244331391, + "grad_norm": 6.4188706045554085, + "learning_rate": 3.5714285714285716e-07, + "loss": 2.1833, + "step": 530 + }, + { + "epoch": 0.03604999831542064, + "grad_norm": 7.093402002469701, + "learning_rate": 3.605121293800539e-07, + "loss": 2.2543, + "step": 535 + }, + { + "epoch": 0.036386914187527375, + "grad_norm": 7.249732464353174, + "learning_rate": 3.638814016172507e-07, + "loss": 2.2095, + "step": 540 + }, + { + "epoch": 0.03672383005963411, + "grad_norm": 7.0654911164895084, + "learning_rate": 3.672506738544474e-07, + "loss": 2.2404, + "step": 545 + }, + { + "epoch": 0.03706074593174084, + "grad_norm": 6.603314898627233, + "learning_rate": 3.706199460916442e-07, + "loss": 2.2267, + "step": 550 + }, + { + "epoch": 0.03739766180384758, + "grad_norm": 6.781857607869617, + "learning_rate": 3.7398921832884095e-07, + "loss": 2.2001, + "step": 555 + }, + { + "epoch": 0.03773457767595431, + "grad_norm": 6.976038643674048, + "learning_rate": 3.773584905660377e-07, + "loss": 2.2232, + "step": 560 + }, + { + "epoch": 0.038071493548061046, + "grad_norm": 6.403705006233041, + "learning_rate": 3.807277628032345e-07, + "loss": 2.1761, + "step": 565 + }, + { + "epoch": 0.03840840942016779, + "grad_norm": 6.661006329480535, + "learning_rate": 3.840970350404312e-07, + "loss": 2.2408, + "step": 570 + }, + { + "epoch": 0.03874532529227452, + "grad_norm": 6.543237506923167, + "learning_rate": 3.8746630727762803e-07, + "loss": 2.177, + "step": 575 + }, + { + "epoch": 0.039082241164381255, + "grad_norm": 7.382219270622229, + "learning_rate": 3.9083557951482475e-07, + "loss": 2.1816, + "step": 580 + }, + { + "epoch": 0.03941915703648799, + "grad_norm": 6.406997261594138, + "learning_rate": 3.9420485175202157e-07, + "loss": 2.2046, + "step": 585 + }, + { + "epoch": 0.039756072908594724, + "grad_norm": 7.4326501687473465, + "learning_rate": 3.975741239892183e-07, + "loss": 2.2355, + "step": 590 + }, + { + "epoch": 0.04009298878070146, + "grad_norm": 6.902366223177666, + "learning_rate": 4.009433962264151e-07, + "loss": 2.2618, + "step": 595 + }, + { + "epoch": 0.04042990465280819, + "grad_norm": 6.688891128794694, + "learning_rate": 4.043126684636118e-07, + "loss": 2.1724, + "step": 600 + }, + { + "epoch": 0.040766820524914926, + "grad_norm": 6.772799117362383, + "learning_rate": 4.076819407008086e-07, + "loss": 2.2251, + "step": 605 + }, + { + "epoch": 0.04110373639702166, + "grad_norm": 7.2951116139212395, + "learning_rate": 4.1105121293800537e-07, + "loss": 2.2227, + "step": 610 + }, + { + "epoch": 0.0414406522691284, + "grad_norm": 7.314163667705275, + "learning_rate": 4.1442048517520213e-07, + "loss": 2.1938, + "step": 615 + }, + { + "epoch": 0.041777568141235136, + "grad_norm": 7.136287025015161, + "learning_rate": 4.177897574123989e-07, + "loss": 2.2072, + "step": 620 + }, + { + "epoch": 0.04211448401334187, + "grad_norm": 6.6315151337205664, + "learning_rate": 4.211590296495957e-07, + "loss": 2.1836, + "step": 625 + }, + { + "epoch": 0.042451399885448604, + "grad_norm": 6.599360278893471, + "learning_rate": 4.2452830188679244e-07, + "loss": 2.1652, + "step": 630 + }, + { + "epoch": 0.04278831575755534, + "grad_norm": 6.9184973647199595, + "learning_rate": 4.278975741239892e-07, + "loss": 2.2163, + "step": 635 + }, + { + "epoch": 0.04312523162966207, + "grad_norm": 7.832951280915868, + "learning_rate": 4.31266846361186e-07, + "loss": 2.1826, + "step": 640 + }, + { + "epoch": 0.04346214750176881, + "grad_norm": 6.941910380094433, + "learning_rate": 4.3463611859838275e-07, + "loss": 2.1137, + "step": 645 + }, + { + "epoch": 0.04379906337387554, + "grad_norm": 6.563567550854655, + "learning_rate": 4.3800539083557947e-07, + "loss": 2.1699, + "step": 650 + }, + { + "epoch": 0.044135979245982275, + "grad_norm": 6.309193421506049, + "learning_rate": 4.413746630727763e-07, + "loss": 2.2038, + "step": 655 + }, + { + "epoch": 0.044472895118089016, + "grad_norm": 7.1931350816677115, + "learning_rate": 4.44743935309973e-07, + "loss": 2.1559, + "step": 660 + }, + { + "epoch": 0.04480981099019575, + "grad_norm": 7.352209899108727, + "learning_rate": 4.481132075471698e-07, + "loss": 2.2001, + "step": 665 + }, + { + "epoch": 0.045146726862302484, + "grad_norm": 6.731481980763938, + "learning_rate": 4.5148247978436655e-07, + "loss": 2.1885, + "step": 670 + }, + { + "epoch": 0.04548364273440922, + "grad_norm": 6.4333683639934724, + "learning_rate": 4.548517520215633e-07, + "loss": 2.2209, + "step": 675 + }, + { + "epoch": 0.04582055860651595, + "grad_norm": 6.555299933379583, + "learning_rate": 4.582210242587601e-07, + "loss": 2.1986, + "step": 680 + }, + { + "epoch": 0.04615747447862269, + "grad_norm": 6.555677397835128, + "learning_rate": 4.6159029649595686e-07, + "loss": 2.2115, + "step": 685 + }, + { + "epoch": 0.04649439035072942, + "grad_norm": 6.8354360632248135, + "learning_rate": 4.649595687331536e-07, + "loss": 2.1043, + "step": 690 + }, + { + "epoch": 0.046831306222836155, + "grad_norm": 6.728423130936894, + "learning_rate": 4.6832884097035034e-07, + "loss": 2.1579, + "step": 695 + }, + { + "epoch": 0.04716822209494289, + "grad_norm": 6.904472227996776, + "learning_rate": 4.7169811320754717e-07, + "loss": 2.1363, + "step": 700 + }, + { + "epoch": 0.04750513796704963, + "grad_norm": 6.802051471379565, + "learning_rate": 4.750673854447439e-07, + "loss": 2.1569, + "step": 705 + }, + { + "epoch": 0.047842053839156365, + "grad_norm": 6.58219453779912, + "learning_rate": 4.784366576819407e-07, + "loss": 2.2192, + "step": 710 + }, + { + "epoch": 0.0481789697112631, + "grad_norm": 7.071304617996877, + "learning_rate": 4.818059299191375e-07, + "loss": 2.1741, + "step": 715 + }, + { + "epoch": 0.04851588558336983, + "grad_norm": 6.383783135641038, + "learning_rate": 4.851752021563342e-07, + "loss": 2.1671, + "step": 720 + }, + { + "epoch": 0.04885280145547657, + "grad_norm": 7.060000529687919, + "learning_rate": 4.88544474393531e-07, + "loss": 2.1408, + "step": 725 + }, + { + "epoch": 0.0491897173275833, + "grad_norm": 6.9599946143852405, + "learning_rate": 4.919137466307277e-07, + "loss": 2.1848, + "step": 730 + }, + { + "epoch": 0.049526633199690036, + "grad_norm": 7.182506932662675, + "learning_rate": 4.952830188679246e-07, + "loss": 2.155, + "step": 735 + }, + { + "epoch": 0.04986354907179677, + "grad_norm": 7.176366825569252, + "learning_rate": 4.986522911051213e-07, + "loss": 2.0861, + "step": 740 + }, + { + "epoch": 0.050200464943903504, + "grad_norm": 7.2983414546411565, + "learning_rate": 5.020215633423181e-07, + "loss": 2.1813, + "step": 745 + }, + { + "epoch": 0.050537380816010245, + "grad_norm": 6.829539638304298, + "learning_rate": 5.053908355795148e-07, + "loss": 2.1295, + "step": 750 + }, + { + "epoch": 0.05087429668811698, + "grad_norm": 6.222282062474875, + "learning_rate": 5.087601078167115e-07, + "loss": 2.1948, + "step": 755 + }, + { + "epoch": 0.051211212560223714, + "grad_norm": 6.863565766148611, + "learning_rate": 5.121293800539083e-07, + "loss": 2.1797, + "step": 760 + }, + { + "epoch": 0.05154812843233045, + "grad_norm": 6.415782001462684, + "learning_rate": 5.154986522911052e-07, + "loss": 2.1882, + "step": 765 + }, + { + "epoch": 0.05188504430443718, + "grad_norm": 6.77615019110364, + "learning_rate": 5.188679245283019e-07, + "loss": 2.1906, + "step": 770 + }, + { + "epoch": 0.052221960176543916, + "grad_norm": 6.460890212003313, + "learning_rate": 5.222371967654986e-07, + "loss": 2.1472, + "step": 775 + }, + { + "epoch": 0.05255887604865065, + "grad_norm": 6.347677993132583, + "learning_rate": 5.256064690026953e-07, + "loss": 2.1341, + "step": 780 + }, + { + "epoch": 0.052895791920757385, + "grad_norm": 6.751500536981585, + "learning_rate": 5.289757412398921e-07, + "loss": 2.1972, + "step": 785 + }, + { + "epoch": 0.05323270779286412, + "grad_norm": 7.361128453658074, + "learning_rate": 5.32345013477089e-07, + "loss": 2.1438, + "step": 790 + }, + { + "epoch": 0.05356962366497086, + "grad_norm": 6.662796897278697, + "learning_rate": 5.357142857142857e-07, + "loss": 2.1632, + "step": 795 + }, + { + "epoch": 0.053906539537077594, + "grad_norm": 6.094649059417771, + "learning_rate": 5.390835579514824e-07, + "loss": 2.0973, + "step": 800 + }, + { + "epoch": 0.05424345540918433, + "grad_norm": 7.282836195110237, + "learning_rate": 5.424528301886792e-07, + "loss": 2.133, + "step": 805 + }, + { + "epoch": 0.05458037128129106, + "grad_norm": 6.23571863669226, + "learning_rate": 5.45822102425876e-07, + "loss": 2.1828, + "step": 810 + }, + { + "epoch": 0.0549172871533978, + "grad_norm": 7.386022383658537, + "learning_rate": 5.491913746630728e-07, + "loss": 2.0529, + "step": 815 + }, + { + "epoch": 0.05525420302550453, + "grad_norm": 9.35422370123258, + "learning_rate": 5.525606469002695e-07, + "loss": 2.1733, + "step": 820 + }, + { + "epoch": 0.055591118897611265, + "grad_norm": 7.192922966723721, + "learning_rate": 5.559299191374662e-07, + "loss": 2.1616, + "step": 825 + }, + { + "epoch": 0.055928034769718, + "grad_norm": 6.225514128765482, + "learning_rate": 5.592991913746631e-07, + "loss": 2.1473, + "step": 830 + }, + { + "epoch": 0.05626495064182473, + "grad_norm": 6.788375025181168, + "learning_rate": 5.626684636118598e-07, + "loss": 2.1498, + "step": 835 + }, + { + "epoch": 0.056601866513931474, + "grad_norm": 6.722245770758269, + "learning_rate": 5.660377358490566e-07, + "loss": 2.1252, + "step": 840 + }, + { + "epoch": 0.05693878238603821, + "grad_norm": 6.609890705250101, + "learning_rate": 5.694070080862533e-07, + "loss": 2.1435, + "step": 845 + }, + { + "epoch": 0.05727569825814494, + "grad_norm": 6.603484908307562, + "learning_rate": 5.727762803234502e-07, + "loss": 2.1841, + "step": 850 + }, + { + "epoch": 0.05761261413025168, + "grad_norm": 6.378869444678416, + "learning_rate": 5.761455525606469e-07, + "loss": 2.1831, + "step": 855 + }, + { + "epoch": 0.05794953000235841, + "grad_norm": 6.184926864820992, + "learning_rate": 5.795148247978436e-07, + "loss": 2.2116, + "step": 860 + }, + { + "epoch": 0.058286445874465145, + "grad_norm": 6.900192026852389, + "learning_rate": 5.828840970350404e-07, + "loss": 2.0611, + "step": 865 + }, + { + "epoch": 0.05862336174657188, + "grad_norm": 6.848190090768429, + "learning_rate": 5.862533692722371e-07, + "loss": 2.1072, + "step": 870 + }, + { + "epoch": 0.058960277618678614, + "grad_norm": 7.775201016865638, + "learning_rate": 5.89622641509434e-07, + "loss": 2.1282, + "step": 875 + }, + { + "epoch": 0.05929719349078535, + "grad_norm": 7.029087955968144, + "learning_rate": 5.929919137466307e-07, + "loss": 2.1134, + "step": 880 + }, + { + "epoch": 0.05963410936289209, + "grad_norm": 7.451133466059003, + "learning_rate": 5.963611859838274e-07, + "loss": 2.1239, + "step": 885 + }, + { + "epoch": 0.05997102523499882, + "grad_norm": 6.8121792270326065, + "learning_rate": 5.997304582210241e-07, + "loss": 2.1138, + "step": 890 + }, + { + "epoch": 0.06030794110710556, + "grad_norm": 6.705718982799264, + "learning_rate": 6.030997304582211e-07, + "loss": 2.0839, + "step": 895 + }, + { + "epoch": 0.06064485697921229, + "grad_norm": 6.870521558441548, + "learning_rate": 6.064690026954178e-07, + "loss": 2.0964, + "step": 900 + }, + { + "epoch": 0.060981772851319026, + "grad_norm": 7.000218922277794, + "learning_rate": 6.098382749326145e-07, + "loss": 2.0962, + "step": 905 + }, + { + "epoch": 0.06131868872342576, + "grad_norm": 6.953179430165108, + "learning_rate": 6.132075471698112e-07, + "loss": 2.1055, + "step": 910 + }, + { + "epoch": 0.061655604595532494, + "grad_norm": 6.928934208954968, + "learning_rate": 6.16576819407008e-07, + "loss": 2.1092, + "step": 915 + }, + { + "epoch": 0.06199252046763923, + "grad_norm": 6.701423119406395, + "learning_rate": 6.199460916442049e-07, + "loss": 2.0719, + "step": 920 + }, + { + "epoch": 0.06232943633974596, + "grad_norm": 6.0010161747456, + "learning_rate": 6.233153638814016e-07, + "loss": 2.0888, + "step": 925 + }, + { + "epoch": 0.0626663522118527, + "grad_norm": 7.097553267015575, + "learning_rate": 6.266846361185983e-07, + "loss": 2.0802, + "step": 930 + }, + { + "epoch": 0.06300326808395944, + "grad_norm": 6.2986464398316775, + "learning_rate": 6.300539083557951e-07, + "loss": 2.1257, + "step": 935 + }, + { + "epoch": 0.06334018395606617, + "grad_norm": 6.773220981525513, + "learning_rate": 6.33423180592992e-07, + "loss": 2.0675, + "step": 940 + }, + { + "epoch": 0.0636770998281729, + "grad_norm": 6.64075808444217, + "learning_rate": 6.367924528301887e-07, + "loss": 2.0588, + "step": 945 + }, + { + "epoch": 0.06401401570027963, + "grad_norm": 6.704739584304528, + "learning_rate": 6.401617250673854e-07, + "loss": 2.0639, + "step": 950 + }, + { + "epoch": 0.06435093157238637, + "grad_norm": 6.450367347861596, + "learning_rate": 6.435309973045822e-07, + "loss": 2.1685, + "step": 955 + }, + { + "epoch": 0.06468784744449312, + "grad_norm": 6.803757171220706, + "learning_rate": 6.469002695417789e-07, + "loss": 2.1149, + "step": 960 + }, + { + "epoch": 0.06502476331659984, + "grad_norm": 7.0080744698550355, + "learning_rate": 6.502695417789757e-07, + "loss": 2.1161, + "step": 965 + }, + { + "epoch": 0.06536167918870658, + "grad_norm": 7.172981850795997, + "learning_rate": 6.536388140161725e-07, + "loss": 2.0494, + "step": 970 + }, + { + "epoch": 0.06569859506081331, + "grad_norm": 6.87372774725762, + "learning_rate": 6.570080862533693e-07, + "loss": 2.0477, + "step": 975 + }, + { + "epoch": 0.06603551093292005, + "grad_norm": 6.134165071155244, + "learning_rate": 6.60377358490566e-07, + "loss": 2.0151, + "step": 980 + }, + { + "epoch": 0.06637242680502678, + "grad_norm": 6.658460133134156, + "learning_rate": 6.637466307277628e-07, + "loss": 2.0934, + "step": 985 + }, + { + "epoch": 0.06670934267713352, + "grad_norm": 6.29191690274154, + "learning_rate": 6.671159029649595e-07, + "loss": 2.0307, + "step": 990 + }, + { + "epoch": 0.06704625854924025, + "grad_norm": 6.519934752973173, + "learning_rate": 6.704851752021563e-07, + "loss": 2.0922, + "step": 995 + }, + { + "epoch": 0.06738317442134699, + "grad_norm": 6.84965030274981, + "learning_rate": 6.738544474393531e-07, + "loss": 2.1385, + "step": 1000 + }, + { + "epoch": 0.06772009029345373, + "grad_norm": 6.503717367498557, + "learning_rate": 6.772237196765498e-07, + "loss": 2.082, + "step": 1005 + }, + { + "epoch": 0.06805700616556046, + "grad_norm": 7.219950548525877, + "learning_rate": 6.805929919137466e-07, + "loss": 2.0827, + "step": 1010 + }, + { + "epoch": 0.0683939220376672, + "grad_norm": 6.714972141308813, + "learning_rate": 6.839622641509433e-07, + "loss": 2.1236, + "step": 1015 + }, + { + "epoch": 0.06873083790977393, + "grad_norm": 6.1863794506801035, + "learning_rate": 6.873315363881402e-07, + "loss": 2.0694, + "step": 1020 + }, + { + "epoch": 0.06906775378188067, + "grad_norm": 6.519857271262621, + "learning_rate": 6.907008086253369e-07, + "loss": 2.066, + "step": 1025 + }, + { + "epoch": 0.0694046696539874, + "grad_norm": 6.061987355990859, + "learning_rate": 6.940700808625337e-07, + "loss": 2.0226, + "step": 1030 + }, + { + "epoch": 0.06974158552609414, + "grad_norm": 6.906351823694033, + "learning_rate": 6.974393530997304e-07, + "loss": 2.1324, + "step": 1035 + }, + { + "epoch": 0.07007850139820086, + "grad_norm": 5.925534436650087, + "learning_rate": 7.008086253369272e-07, + "loss": 2.1207, + "step": 1040 + }, + { + "epoch": 0.0704154172703076, + "grad_norm": 7.300437774124698, + "learning_rate": 7.04177897574124e-07, + "loss": 2.0932, + "step": 1045 + }, + { + "epoch": 0.07075233314241434, + "grad_norm": 7.208321480205573, + "learning_rate": 7.075471698113207e-07, + "loss": 2.1145, + "step": 1050 + }, + { + "epoch": 0.07108924901452107, + "grad_norm": 6.523539997207994, + "learning_rate": 7.109164420485175e-07, + "loss": 2.1084, + "step": 1055 + }, + { + "epoch": 0.07142616488662781, + "grad_norm": 6.6532730740194905, + "learning_rate": 7.142857142857143e-07, + "loss": 2.0747, + "step": 1060 + }, + { + "epoch": 0.07176308075873454, + "grad_norm": 5.910525317424687, + "learning_rate": 7.17654986522911e-07, + "loss": 2.0412, + "step": 1065 + }, + { + "epoch": 0.07209999663084128, + "grad_norm": 6.671481682607736, + "learning_rate": 7.210242587601077e-07, + "loss": 2.0943, + "step": 1070 + }, + { + "epoch": 0.07243691250294801, + "grad_norm": 7.410527719144032, + "learning_rate": 7.243935309973046e-07, + "loss": 2.089, + "step": 1075 + }, + { + "epoch": 0.07277382837505475, + "grad_norm": 6.998731809026702, + "learning_rate": 7.277628032345014e-07, + "loss": 2.0711, + "step": 1080 + }, + { + "epoch": 0.07311074424716148, + "grad_norm": 6.168070160591785, + "learning_rate": 7.311320754716981e-07, + "loss": 2.1038, + "step": 1085 + }, + { + "epoch": 0.07344766011926822, + "grad_norm": 6.445599092013598, + "learning_rate": 7.345013477088948e-07, + "loss": 2.1231, + "step": 1090 + }, + { + "epoch": 0.07378457599137496, + "grad_norm": 6.847043979859011, + "learning_rate": 7.378706199460915e-07, + "loss": 2.0792, + "step": 1095 + }, + { + "epoch": 0.07412149186348169, + "grad_norm": 6.118594411094878, + "learning_rate": 7.412398921832884e-07, + "loss": 2.0615, + "step": 1100 + }, + { + "epoch": 0.07445840773558843, + "grad_norm": 6.4444397980063535, + "learning_rate": 7.446091644204852e-07, + "loss": 2.0941, + "step": 1105 + }, + { + "epoch": 0.07479532360769515, + "grad_norm": 6.900009739269609, + "learning_rate": 7.479784366576819e-07, + "loss": 2.0874, + "step": 1110 + }, + { + "epoch": 0.0751322394798019, + "grad_norm": 6.715563896490011, + "learning_rate": 7.513477088948786e-07, + "loss": 2.0702, + "step": 1115 + }, + { + "epoch": 0.07546915535190862, + "grad_norm": 6.683671984170299, + "learning_rate": 7.547169811320754e-07, + "loss": 2.0257, + "step": 1120 + }, + { + "epoch": 0.07580607122401536, + "grad_norm": 6.50723958019259, + "learning_rate": 7.580862533692723e-07, + "loss": 2.0737, + "step": 1125 + }, + { + "epoch": 0.07614298709612209, + "grad_norm": 6.061015680440867, + "learning_rate": 7.61455525606469e-07, + "loss": 2.0992, + "step": 1130 + }, + { + "epoch": 0.07647990296822883, + "grad_norm": 6.587328595415548, + "learning_rate": 7.648247978436657e-07, + "loss": 2.0405, + "step": 1135 + }, + { + "epoch": 0.07681681884033557, + "grad_norm": 6.387267569159134, + "learning_rate": 7.681940700808624e-07, + "loss": 2.0936, + "step": 1140 + }, + { + "epoch": 0.0771537347124423, + "grad_norm": 6.5689891488973045, + "learning_rate": 7.715633423180593e-07, + "loss": 2.0483, + "step": 1145 + }, + { + "epoch": 0.07749065058454904, + "grad_norm": 6.98294967939874, + "learning_rate": 7.749326145552561e-07, + "loss": 2.0804, + "step": 1150 + }, + { + "epoch": 0.07782756645665577, + "grad_norm": 6.8502328744808665, + "learning_rate": 7.783018867924528e-07, + "loss": 2.0386, + "step": 1155 + }, + { + "epoch": 0.07816448232876251, + "grad_norm": 6.330263851851983, + "learning_rate": 7.816711590296495e-07, + "loss": 2.0429, + "step": 1160 + }, + { + "epoch": 0.07850139820086924, + "grad_norm": 6.558472938499403, + "learning_rate": 7.850404312668463e-07, + "loss": 2.1344, + "step": 1165 + }, + { + "epoch": 0.07883831407297598, + "grad_norm": 6.793056886526596, + "learning_rate": 7.884097035040431e-07, + "loss": 2.053, + "step": 1170 + }, + { + "epoch": 0.0791752299450827, + "grad_norm": 6.334091419449102, + "learning_rate": 7.917789757412399e-07, + "loss": 2.0741, + "step": 1175 + }, + { + "epoch": 0.07951214581718945, + "grad_norm": 6.995659281890644, + "learning_rate": 7.951482479784366e-07, + "loss": 2.0299, + "step": 1180 + }, + { + "epoch": 0.07984906168929619, + "grad_norm": 6.057490760704781, + "learning_rate": 7.985175202156334e-07, + "loss": 2.0367, + "step": 1185 + }, + { + "epoch": 0.08018597756140292, + "grad_norm": 6.812009588247003, + "learning_rate": 8.018867924528302e-07, + "loss": 2.1046, + "step": 1190 + }, + { + "epoch": 0.08052289343350966, + "grad_norm": 6.880635288270827, + "learning_rate": 8.052560646900269e-07, + "loss": 2.0334, + "step": 1195 + }, + { + "epoch": 0.08085980930561638, + "grad_norm": 6.440406473747084, + "learning_rate": 8.086253369272237e-07, + "loss": 2.0336, + "step": 1200 + }, + { + "epoch": 0.08119672517772313, + "grad_norm": 6.2159531204319896, + "learning_rate": 8.119946091644204e-07, + "loss": 2.0666, + "step": 1205 + }, + { + "epoch": 0.08153364104982985, + "grad_norm": 6.596526872386366, + "learning_rate": 8.153638814016172e-07, + "loss": 2.1116, + "step": 1210 + }, + { + "epoch": 0.0818705569219366, + "grad_norm": 6.215209904994124, + "learning_rate": 8.18733153638814e-07, + "loss": 2.0579, + "step": 1215 + }, + { + "epoch": 0.08220747279404332, + "grad_norm": 6.583210332500679, + "learning_rate": 8.221024258760107e-07, + "loss": 2.014, + "step": 1220 + }, + { + "epoch": 0.08254438866615006, + "grad_norm": 6.212012798535623, + "learning_rate": 8.254716981132074e-07, + "loss": 2.0458, + "step": 1225 + }, + { + "epoch": 0.0828813045382568, + "grad_norm": 6.470292156093085, + "learning_rate": 8.288409703504043e-07, + "loss": 2.0949, + "step": 1230 + }, + { + "epoch": 0.08321822041036353, + "grad_norm": 6.0866198998833685, + "learning_rate": 8.322102425876011e-07, + "loss": 2.0176, + "step": 1235 + }, + { + "epoch": 0.08355513628247027, + "grad_norm": 6.635224587579871, + "learning_rate": 8.355795148247978e-07, + "loss": 2.0963, + "step": 1240 + }, + { + "epoch": 0.083892052154577, + "grad_norm": 5.712838593846866, + "learning_rate": 8.389487870619945e-07, + "loss": 2.0503, + "step": 1245 + }, + { + "epoch": 0.08422896802668374, + "grad_norm": 5.933358327100002, + "learning_rate": 8.423180592991913e-07, + "loss": 2.0016, + "step": 1250 + }, + { + "epoch": 0.08456588389879047, + "grad_norm": 7.143451933361089, + "learning_rate": 8.456873315363881e-07, + "loss": 2.0677, + "step": 1255 + }, + { + "epoch": 0.08490279977089721, + "grad_norm": 6.703706638919056, + "learning_rate": 8.490566037735849e-07, + "loss": 2.01, + "step": 1260 + }, + { + "epoch": 0.08523971564300394, + "grad_norm": 6.0965320200893345, + "learning_rate": 8.524258760107816e-07, + "loss": 1.988, + "step": 1265 + }, + { + "epoch": 0.08557663151511068, + "grad_norm": 7.193765211165279, + "learning_rate": 8.557951482479784e-07, + "loss": 2.1143, + "step": 1270 + }, + { + "epoch": 0.08591354738721742, + "grad_norm": 6.301021380513151, + "learning_rate": 8.591644204851751e-07, + "loss": 2.049, + "step": 1275 + }, + { + "epoch": 0.08625046325932414, + "grad_norm": 6.904570371972653, + "learning_rate": 8.62533692722372e-07, + "loss": 2.0629, + "step": 1280 + }, + { + "epoch": 0.08658737913143089, + "grad_norm": 6.214009445437056, + "learning_rate": 8.659029649595687e-07, + "loss": 2.0645, + "step": 1285 + }, + { + "epoch": 0.08692429500353761, + "grad_norm": 6.6425747142232785, + "learning_rate": 8.692722371967655e-07, + "loss": 2.1192, + "step": 1290 + }, + { + "epoch": 0.08726121087564435, + "grad_norm": 6.715501139786349, + "learning_rate": 8.726415094339622e-07, + "loss": 2.0797, + "step": 1295 + }, + { + "epoch": 0.08759812674775108, + "grad_norm": 7.117189534146049, + "learning_rate": 8.760107816711589e-07, + "loss": 2.1073, + "step": 1300 + }, + { + "epoch": 0.08793504261985782, + "grad_norm": 6.677619622948294, + "learning_rate": 8.793800539083558e-07, + "loss": 2.0255, + "step": 1305 + }, + { + "epoch": 0.08827195849196455, + "grad_norm": 6.784702166999388, + "learning_rate": 8.827493261455526e-07, + "loss": 2.0463, + "step": 1310 + }, + { + "epoch": 0.08860887436407129, + "grad_norm": 6.6534308801544615, + "learning_rate": 8.861185983827493e-07, + "loss": 2.0355, + "step": 1315 + }, + { + "epoch": 0.08894579023617803, + "grad_norm": 6.344151360801522, + "learning_rate": 8.89487870619946e-07, + "loss": 2.0036, + "step": 1320 + }, + { + "epoch": 0.08928270610828476, + "grad_norm": 6.580775236275343, + "learning_rate": 8.928571428571428e-07, + "loss": 2.1212, + "step": 1325 + }, + { + "epoch": 0.0896196219803915, + "grad_norm": 6.968641119999037, + "learning_rate": 8.962264150943396e-07, + "loss": 2.0261, + "step": 1330 + }, + { + "epoch": 0.08995653785249823, + "grad_norm": 7.747892608432803, + "learning_rate": 8.995956873315364e-07, + "loss": 2.0266, + "step": 1335 + }, + { + "epoch": 0.09029345372460497, + "grad_norm": 7.06577595704722, + "learning_rate": 9.029649595687331e-07, + "loss": 2.0086, + "step": 1340 + }, + { + "epoch": 0.0906303695967117, + "grad_norm": 6.2867926464117065, + "learning_rate": 9.063342318059298e-07, + "loss": 2.0108, + "step": 1345 + }, + { + "epoch": 0.09096728546881844, + "grad_norm": 6.481546616940274, + "learning_rate": 9.097035040431266e-07, + "loss": 2.0175, + "step": 1350 + }, + { + "epoch": 0.09130420134092516, + "grad_norm": 6.801814398606904, + "learning_rate": 9.130727762803235e-07, + "loss": 2.0203, + "step": 1355 + }, + { + "epoch": 0.0916411172130319, + "grad_norm": 5.801249089738751, + "learning_rate": 9.164420485175202e-07, + "loss": 2.0812, + "step": 1360 + }, + { + "epoch": 0.09197803308513865, + "grad_norm": 5.945627363410938, + "learning_rate": 9.198113207547169e-07, + "loss": 2.0239, + "step": 1365 + }, + { + "epoch": 0.09231494895724537, + "grad_norm": 5.930218401587245, + "learning_rate": 9.231805929919137e-07, + "loss": 2.0187, + "step": 1370 + }, + { + "epoch": 0.09265186482935212, + "grad_norm": 6.886170229635843, + "learning_rate": 9.265498652291105e-07, + "loss": 2.0579, + "step": 1375 + }, + { + "epoch": 0.09298878070145884, + "grad_norm": 6.505667387657664, + "learning_rate": 9.299191374663073e-07, + "loss": 2.0544, + "step": 1380 + }, + { + "epoch": 0.09332569657356558, + "grad_norm": 6.402314934607021, + "learning_rate": 9.33288409703504e-07, + "loss": 2.0511, + "step": 1385 + }, + { + "epoch": 0.09366261244567231, + "grad_norm": 6.242297912959462, + "learning_rate": 9.366576819407007e-07, + "loss": 2.0305, + "step": 1390 + }, + { + "epoch": 0.09399952831777905, + "grad_norm": 5.557824750137879, + "learning_rate": 9.400269541778976e-07, + "loss": 1.975, + "step": 1395 + }, + { + "epoch": 0.09433644418988578, + "grad_norm": 6.438317330999812, + "learning_rate": 9.433962264150943e-07, + "loss": 2.0444, + "step": 1400 + }, + { + "epoch": 0.09467336006199252, + "grad_norm": 6.829458975236009, + "learning_rate": 9.46765498652291e-07, + "loss": 2.0997, + "step": 1405 + }, + { + "epoch": 0.09501027593409926, + "grad_norm": 6.052680993841202, + "learning_rate": 9.501347708894878e-07, + "loss": 1.9701, + "step": 1410 + }, + { + "epoch": 0.09534719180620599, + "grad_norm": 6.145341438647098, + "learning_rate": 9.535040431266847e-07, + "loss": 2.0437, + "step": 1415 + }, + { + "epoch": 0.09568410767831273, + "grad_norm": 6.386909941250297, + "learning_rate": 9.568733153638813e-07, + "loss": 1.9988, + "step": 1420 + }, + { + "epoch": 0.09602102355041946, + "grad_norm": 7.2117096771751585, + "learning_rate": 9.60242587601078e-07, + "loss": 1.9924, + "step": 1425 + }, + { + "epoch": 0.0963579394225262, + "grad_norm": 7.090905319078084, + "learning_rate": 9.63611859838275e-07, + "loss": 2.0736, + "step": 1430 + }, + { + "epoch": 0.09669485529463293, + "grad_norm": 5.765675979737204, + "learning_rate": 9.669811320754717e-07, + "loss": 2.0255, + "step": 1435 + }, + { + "epoch": 0.09703177116673967, + "grad_norm": 6.302835725851336, + "learning_rate": 9.703504043126684e-07, + "loss": 1.9923, + "step": 1440 + }, + { + "epoch": 0.0973686870388464, + "grad_norm": 6.105386769901037, + "learning_rate": 9.73719676549865e-07, + "loss": 2.0062, + "step": 1445 + }, + { + "epoch": 0.09770560291095313, + "grad_norm": 5.779479200910098, + "learning_rate": 9.77088948787062e-07, + "loss": 1.9981, + "step": 1450 + }, + { + "epoch": 0.09804251878305988, + "grad_norm": 6.478566809649276, + "learning_rate": 9.804582210242587e-07, + "loss": 2.014, + "step": 1455 + }, + { + "epoch": 0.0983794346551666, + "grad_norm": 6.462510705631694, + "learning_rate": 9.838274932614555e-07, + "loss": 2.0597, + "step": 1460 + }, + { + "epoch": 0.09871635052727334, + "grad_norm": 7.517939119617698, + "learning_rate": 9.871967654986522e-07, + "loss": 2.0555, + "step": 1465 + }, + { + "epoch": 0.09905326639938007, + "grad_norm": 6.944192458999394, + "learning_rate": 9.90566037735849e-07, + "loss": 1.9715, + "step": 1470 + }, + { + "epoch": 0.09939018227148681, + "grad_norm": 6.3353086122850275, + "learning_rate": 9.939353099730458e-07, + "loss": 1.9726, + "step": 1475 + }, + { + "epoch": 0.09972709814359354, + "grad_norm": 6.701570960194593, + "learning_rate": 9.973045822102425e-07, + "loss": 2.0269, + "step": 1480 + }, + { + "epoch": 0.10006401401570028, + "grad_norm": 6.056963246639003, + "learning_rate": 9.999999861679377e-07, + "loss": 1.9261, + "step": 1485 + }, + { + "epoch": 0.10040092988780701, + "grad_norm": 6.84014658632624, + "learning_rate": 9.999995020458434e-07, + "loss": 1.9683, + "step": 1490 + }, + { + "epoch": 0.10073784575991375, + "grad_norm": 6.615520534333261, + "learning_rate": 9.99998326321407e-07, + "loss": 2.0027, + "step": 1495 + }, + { + "epoch": 0.10107476163202049, + "grad_norm": 6.417598276308663, + "learning_rate": 9.999964589962556e-07, + "loss": 1.9595, + "step": 1500 + }, + { + "epoch": 0.10141167750412722, + "grad_norm": 6.838049024969315, + "learning_rate": 9.999939000729715e-07, + "loss": 2.0318, + "step": 1505 + }, + { + "epoch": 0.10174859337623396, + "grad_norm": 5.9861108568338315, + "learning_rate": 9.999906495550946e-07, + "loss": 2.0101, + "step": 1510 + }, + { + "epoch": 0.10208550924834069, + "grad_norm": 6.672821704365139, + "learning_rate": 9.999867074471207e-07, + "loss": 2.0185, + "step": 1515 + }, + { + "epoch": 0.10242242512044743, + "grad_norm": 5.864345861096738, + "learning_rate": 9.99982073754503e-07, + "loss": 1.9601, + "step": 1520 + }, + { + "epoch": 0.10275934099255415, + "grad_norm": 5.6384613813717435, + "learning_rate": 9.999767484836502e-07, + "loss": 1.9809, + "step": 1525 + }, + { + "epoch": 0.1030962568646609, + "grad_norm": 6.607737085201073, + "learning_rate": 9.999707316419288e-07, + "loss": 1.9371, + "step": 1530 + }, + { + "epoch": 0.10343317273676762, + "grad_norm": 5.92107920065298, + "learning_rate": 9.99964023237661e-07, + "loss": 1.9583, + "step": 1535 + }, + { + "epoch": 0.10377008860887436, + "grad_norm": 6.797382899990634, + "learning_rate": 9.999566232801261e-07, + "loss": 1.9857, + "step": 1540 + }, + { + "epoch": 0.1041070044809811, + "grad_norm": 6.4291929016349, + "learning_rate": 9.9994853177956e-07, + "loss": 1.9806, + "step": 1545 + }, + { + "epoch": 0.10444392035308783, + "grad_norm": 6.354612439425369, + "learning_rate": 9.999397487471543e-07, + "loss": 2.0837, + "step": 1550 + }, + { + "epoch": 0.10478083622519457, + "grad_norm": 6.203296467397325, + "learning_rate": 9.999302741950582e-07, + "loss": 2.0008, + "step": 1555 + }, + { + "epoch": 0.1051177520973013, + "grad_norm": 6.765947240996871, + "learning_rate": 9.999201081363768e-07, + "loss": 2.018, + "step": 1560 + }, + { + "epoch": 0.10545466796940804, + "grad_norm": 6.364878195328014, + "learning_rate": 9.99909250585172e-07, + "loss": 2.0204, + "step": 1565 + }, + { + "epoch": 0.10579158384151477, + "grad_norm": 6.249427081573669, + "learning_rate": 9.998977015564617e-07, + "loss": 2.0793, + "step": 1570 + }, + { + "epoch": 0.10612849971362151, + "grad_norm": 6.090825980208063, + "learning_rate": 9.998854610662209e-07, + "loss": 2.0201, + "step": 1575 + }, + { + "epoch": 0.10646541558572824, + "grad_norm": 6.387247022298909, + "learning_rate": 9.998725291313805e-07, + "loss": 2.0238, + "step": 1580 + }, + { + "epoch": 0.10680233145783498, + "grad_norm": 10.962328465461098, + "learning_rate": 9.998589057698283e-07, + "loss": 2.0252, + "step": 1585 + }, + { + "epoch": 0.10713924732994172, + "grad_norm": 6.758082756394557, + "learning_rate": 9.99844591000408e-07, + "loss": 1.9437, + "step": 1590 + }, + { + "epoch": 0.10747616320204845, + "grad_norm": 6.409721660335127, + "learning_rate": 9.9982958484292e-07, + "loss": 2.0102, + "step": 1595 + }, + { + "epoch": 0.10781307907415519, + "grad_norm": 6.521098857160827, + "learning_rate": 9.99813887318121e-07, + "loss": 1.9555, + "step": 1600 + }, + { + "epoch": 0.10814999494626192, + "grad_norm": 6.719644060906998, + "learning_rate": 9.997974984477236e-07, + "loss": 2.0169, + "step": 1605 + }, + { + "epoch": 0.10848691081836866, + "grad_norm": 6.2454936784635, + "learning_rate": 9.99780418254397e-07, + "loss": 1.9815, + "step": 1610 + }, + { + "epoch": 0.10882382669047538, + "grad_norm": 6.301452184455107, + "learning_rate": 9.99762646761767e-07, + "loss": 2.0398, + "step": 1615 + }, + { + "epoch": 0.10916074256258212, + "grad_norm": 6.6193329719459495, + "learning_rate": 9.99744183994415e-07, + "loss": 1.9679, + "step": 1620 + }, + { + "epoch": 0.10949765843468885, + "grad_norm": 6.451959168059164, + "learning_rate": 9.997250299778788e-07, + "loss": 2.014, + "step": 1625 + }, + { + "epoch": 0.1098345743067956, + "grad_norm": 6.906957587107302, + "learning_rate": 9.997051847386524e-07, + "loss": 2.024, + "step": 1630 + }, + { + "epoch": 0.11017149017890233, + "grad_norm": 6.132267437341465, + "learning_rate": 9.996846483041858e-07, + "loss": 2.0683, + "step": 1635 + }, + { + "epoch": 0.11050840605100906, + "grad_norm": 6.58193618216666, + "learning_rate": 9.99663420702885e-07, + "loss": 2.0585, + "step": 1640 + }, + { + "epoch": 0.1108453219231158, + "grad_norm": 6.375002932094628, + "learning_rate": 9.996415019641124e-07, + "loss": 1.9758, + "step": 1645 + }, + { + "epoch": 0.11118223779522253, + "grad_norm": 6.196537733272767, + "learning_rate": 9.996188921181861e-07, + "loss": 1.9839, + "step": 1650 + }, + { + "epoch": 0.11151915366732927, + "grad_norm": 6.578190053853053, + "learning_rate": 9.9959559119638e-07, + "loss": 1.966, + "step": 1655 + }, + { + "epoch": 0.111856069539436, + "grad_norm": 6.219530006448483, + "learning_rate": 9.995715992309244e-07, + "loss": 2.0064, + "step": 1660 + }, + { + "epoch": 0.11219298541154274, + "grad_norm": 6.08905725259499, + "learning_rate": 9.995469162550048e-07, + "loss": 1.9459, + "step": 1665 + }, + { + "epoch": 0.11252990128364947, + "grad_norm": 6.718246145997958, + "learning_rate": 9.99521542302763e-07, + "loss": 1.9835, + "step": 1670 + }, + { + "epoch": 0.11286681715575621, + "grad_norm": 6.254668566153578, + "learning_rate": 9.994954774092962e-07, + "loss": 1.9425, + "step": 1675 + }, + { + "epoch": 0.11320373302786295, + "grad_norm": 6.53959891506213, + "learning_rate": 9.994687216106579e-07, + "loss": 1.9768, + "step": 1680 + }, + { + "epoch": 0.11354064889996968, + "grad_norm": 6.0849150556619565, + "learning_rate": 9.994412749438564e-07, + "loss": 1.9348, + "step": 1685 + }, + { + "epoch": 0.11387756477207642, + "grad_norm": 6.256083226566406, + "learning_rate": 9.994131374468565e-07, + "loss": 1.949, + "step": 1690 + }, + { + "epoch": 0.11421448064418314, + "grad_norm": 5.7376740963750885, + "learning_rate": 9.993843091585782e-07, + "loss": 1.9952, + "step": 1695 + }, + { + "epoch": 0.11455139651628989, + "grad_norm": 6.115524205739567, + "learning_rate": 9.993547901188966e-07, + "loss": 1.9722, + "step": 1700 + }, + { + "epoch": 0.11488831238839661, + "grad_norm": 6.245389899651034, + "learning_rate": 9.993245803686426e-07, + "loss": 1.925, + "step": 1705 + }, + { + "epoch": 0.11522522826050335, + "grad_norm": 6.247598728078289, + "learning_rate": 9.992936799496029e-07, + "loss": 1.9901, + "step": 1710 + }, + { + "epoch": 0.11556214413261008, + "grad_norm": 6.325351967024773, + "learning_rate": 9.99262088904519e-07, + "loss": 1.9469, + "step": 1715 + }, + { + "epoch": 0.11589906000471682, + "grad_norm": 6.393709375328121, + "learning_rate": 9.992298072770877e-07, + "loss": 2.0358, + "step": 1720 + }, + { + "epoch": 0.11623597587682356, + "grad_norm": 6.288602341315024, + "learning_rate": 9.991968351119612e-07, + "loss": 1.9845, + "step": 1725 + }, + { + "epoch": 0.11657289174893029, + "grad_norm": 6.754113952908161, + "learning_rate": 9.991631724547467e-07, + "loss": 1.9789, + "step": 1730 + }, + { + "epoch": 0.11690980762103703, + "grad_norm": 6.39680247647427, + "learning_rate": 9.99128819352007e-07, + "loss": 2.0299, + "step": 1735 + }, + { + "epoch": 0.11724672349314376, + "grad_norm": 6.269165973260075, + "learning_rate": 9.99093775851259e-07, + "loss": 2.0239, + "step": 1740 + }, + { + "epoch": 0.1175836393652505, + "grad_norm": 5.989228704848224, + "learning_rate": 9.990580420009755e-07, + "loss": 2.0186, + "step": 1745 + }, + { + "epoch": 0.11792055523735723, + "grad_norm": 6.264711564747278, + "learning_rate": 9.990216178505835e-07, + "loss": 1.9824, + "step": 1750 + }, + { + "epoch": 0.11825747110946397, + "grad_norm": 6.220647194759505, + "learning_rate": 9.989845034504651e-07, + "loss": 1.9392, + "step": 1755 + }, + { + "epoch": 0.1185943869815707, + "grad_norm": 5.876106626219256, + "learning_rate": 9.989466988519572e-07, + "loss": 1.9485, + "step": 1760 + }, + { + "epoch": 0.11893130285367744, + "grad_norm": 6.628947048760207, + "learning_rate": 9.989082041073517e-07, + "loss": 1.9315, + "step": 1765 + }, + { + "epoch": 0.11926821872578418, + "grad_norm": 6.623615345814385, + "learning_rate": 9.988690192698944e-07, + "loss": 1.9931, + "step": 1770 + }, + { + "epoch": 0.1196051345978909, + "grad_norm": 6.281816857100239, + "learning_rate": 9.988291443937857e-07, + "loss": 2.001, + "step": 1775 + }, + { + "epoch": 0.11994205046999765, + "grad_norm": 5.808970810910876, + "learning_rate": 9.987885795341816e-07, + "loss": 1.9608, + "step": 1780 + }, + { + "epoch": 0.12027896634210437, + "grad_norm": 6.0557446750595165, + "learning_rate": 9.987473247471908e-07, + "loss": 1.9572, + "step": 1785 + }, + { + "epoch": 0.12061588221421111, + "grad_norm": 7.277415642252567, + "learning_rate": 9.98705380089878e-07, + "loss": 1.9735, + "step": 1790 + }, + { + "epoch": 0.12095279808631784, + "grad_norm": 6.413165468216076, + "learning_rate": 9.986627456202608e-07, + "loss": 2.0268, + "step": 1795 + }, + { + "epoch": 0.12128971395842458, + "grad_norm": 6.842142941489323, + "learning_rate": 9.986194213973113e-07, + "loss": 1.9498, + "step": 1800 + }, + { + "epoch": 0.12162662983053131, + "grad_norm": 6.284595686241228, + "learning_rate": 9.985754074809562e-07, + "loss": 1.9252, + "step": 1805 + }, + { + "epoch": 0.12196354570263805, + "grad_norm": 5.898293427626579, + "learning_rate": 9.985307039320756e-07, + "loss": 1.952, + "step": 1810 + }, + { + "epoch": 0.12230046157474479, + "grad_norm": 6.039978658719599, + "learning_rate": 9.98485310812504e-07, + "loss": 1.9808, + "step": 1815 + }, + { + "epoch": 0.12263737744685152, + "grad_norm": 5.694946193557863, + "learning_rate": 9.98439228185029e-07, + "loss": 1.9814, + "step": 1820 + }, + { + "epoch": 0.12297429331895826, + "grad_norm": 6.545566967262388, + "learning_rate": 9.983924561133927e-07, + "loss": 1.8983, + "step": 1825 + }, + { + "epoch": 0.12331120919106499, + "grad_norm": 6.728834164161067, + "learning_rate": 9.983449946622906e-07, + "loss": 1.9285, + "step": 1830 + }, + { + "epoch": 0.12364812506317173, + "grad_norm": 5.91994081131544, + "learning_rate": 9.982968438973714e-07, + "loss": 1.9673, + "step": 1835 + }, + { + "epoch": 0.12398504093527846, + "grad_norm": 6.386494266635038, + "learning_rate": 9.982480038852375e-07, + "loss": 2.0266, + "step": 1840 + }, + { + "epoch": 0.1243219568073852, + "grad_norm": 6.441402002868234, + "learning_rate": 9.98198474693445e-07, + "loss": 1.968, + "step": 1845 + }, + { + "epoch": 0.12465887267949192, + "grad_norm": 6.740829824949441, + "learning_rate": 9.981482563905025e-07, + "loss": 2.0041, + "step": 1850 + }, + { + "epoch": 0.12499578855159867, + "grad_norm": 5.710071008321578, + "learning_rate": 9.980973490458728e-07, + "loss": 1.9811, + "step": 1855 + }, + { + "epoch": 0.1253327044237054, + "grad_norm": 6.18536895105298, + "learning_rate": 9.980457527299708e-07, + "loss": 1.9666, + "step": 1860 + }, + { + "epoch": 0.12566962029581213, + "grad_norm": 6.304203854455156, + "learning_rate": 9.979934675141652e-07, + "loss": 1.9504, + "step": 1865 + }, + { + "epoch": 0.12600653616791888, + "grad_norm": 6.359105927612574, + "learning_rate": 9.979404934707771e-07, + "loss": 1.934, + "step": 1870 + }, + { + "epoch": 0.12634345204002562, + "grad_norm": 7.021272358637824, + "learning_rate": 9.978868306730804e-07, + "loss": 1.9559, + "step": 1875 + }, + { + "epoch": 0.12668036791213233, + "grad_norm": 8.721181121099121, + "learning_rate": 9.978324791953018e-07, + "loss": 1.9553, + "step": 1880 + }, + { + "epoch": 0.12701728378423907, + "grad_norm": 5.928281083657215, + "learning_rate": 9.97777439112621e-07, + "loss": 1.9623, + "step": 1885 + }, + { + "epoch": 0.1273541996563458, + "grad_norm": 7.636936649110356, + "learning_rate": 9.977217105011693e-07, + "loss": 1.9405, + "step": 1890 + }, + { + "epoch": 0.12769111552845255, + "grad_norm": 6.378639048271833, + "learning_rate": 9.97665293438031e-07, + "loss": 1.9831, + "step": 1895 + }, + { + "epoch": 0.12802803140055927, + "grad_norm": 6.416618144613928, + "learning_rate": 9.976081880012426e-07, + "loss": 1.919, + "step": 1900 + }, + { + "epoch": 0.128364947272666, + "grad_norm": 6.5865892724690625, + "learning_rate": 9.975503942697925e-07, + "loss": 1.9804, + "step": 1905 + }, + { + "epoch": 0.12870186314477275, + "grad_norm": 6.38034752020973, + "learning_rate": 9.974919123236217e-07, + "loss": 1.9639, + "step": 1910 + }, + { + "epoch": 0.1290387790168795, + "grad_norm": 6.453173099164715, + "learning_rate": 9.974327422436223e-07, + "loss": 1.9276, + "step": 1915 + }, + { + "epoch": 0.12937569488898623, + "grad_norm": 5.566683381219873, + "learning_rate": 9.97372884111639e-07, + "loss": 1.9437, + "step": 1920 + }, + { + "epoch": 0.12971261076109294, + "grad_norm": 6.424615749315436, + "learning_rate": 9.97312338010468e-07, + "loss": 1.996, + "step": 1925 + }, + { + "epoch": 0.13004952663319969, + "grad_norm": 6.189066134613252, + "learning_rate": 9.97251104023857e-07, + "loss": 2.0165, + "step": 1930 + }, + { + "epoch": 0.13038644250530643, + "grad_norm": 6.153302464363939, + "learning_rate": 9.971891822365048e-07, + "loss": 1.963, + "step": 1935 + }, + { + "epoch": 0.13072335837741317, + "grad_norm": 5.898027522950926, + "learning_rate": 9.971265727340627e-07, + "loss": 1.9446, + "step": 1940 + }, + { + "epoch": 0.13106027424951988, + "grad_norm": 6.621778499253866, + "learning_rate": 9.970632756031322e-07, + "loss": 2.0041, + "step": 1945 + }, + { + "epoch": 0.13139719012162662, + "grad_norm": 6.622837129275348, + "learning_rate": 9.969992909312658e-07, + "loss": 1.9858, + "step": 1950 + }, + { + "epoch": 0.13173410599373336, + "grad_norm": 6.052858967214188, + "learning_rate": 9.969346188069684e-07, + "loss": 1.9916, + "step": 1955 + }, + { + "epoch": 0.1320710218658401, + "grad_norm": 6.179560277489928, + "learning_rate": 9.968692593196943e-07, + "loss": 1.8877, + "step": 1960 + }, + { + "epoch": 0.13240793773794685, + "grad_norm": 6.162522901829469, + "learning_rate": 9.968032125598493e-07, + "loss": 2.0524, + "step": 1965 + }, + { + "epoch": 0.13274485361005356, + "grad_norm": 5.902608589754359, + "learning_rate": 9.967364786187894e-07, + "loss": 1.9449, + "step": 1970 + }, + { + "epoch": 0.1330817694821603, + "grad_norm": 6.323335173726953, + "learning_rate": 9.96669057588822e-07, + "loss": 1.9734, + "step": 1975 + }, + { + "epoch": 0.13341868535426704, + "grad_norm": 5.8379906828434285, + "learning_rate": 9.966009495632037e-07, + "loss": 1.9404, + "step": 1980 + }, + { + "epoch": 0.13375560122637378, + "grad_norm": 5.991919976178825, + "learning_rate": 9.965321546361421e-07, + "loss": 1.956, + "step": 1985 + }, + { + "epoch": 0.1340925170984805, + "grad_norm": 5.967331390409548, + "learning_rate": 9.964626729027948e-07, + "loss": 1.953, + "step": 1990 + }, + { + "epoch": 0.13442943297058724, + "grad_norm": 6.281340926030527, + "learning_rate": 9.963925044592695e-07, + "loss": 1.9961, + "step": 1995 + }, + { + "epoch": 0.13476634884269398, + "grad_norm": 6.741635709263175, + "learning_rate": 9.963216494026235e-07, + "loss": 1.912, + "step": 2000 + }, + { + "epoch": 0.13510326471480072, + "grad_norm": 6.08250513108592, + "learning_rate": 9.962501078308636e-07, + "loss": 1.9913, + "step": 2005 + }, + { + "epoch": 0.13544018058690746, + "grad_norm": 6.60765627452018, + "learning_rate": 9.96177879842947e-07, + "loss": 1.9449, + "step": 2010 + }, + { + "epoch": 0.13577709645901417, + "grad_norm": 6.566107141142278, + "learning_rate": 9.961049655387799e-07, + "loss": 1.9879, + "step": 2015 + }, + { + "epoch": 0.13611401233112091, + "grad_norm": 6.335526898300328, + "learning_rate": 9.960313650192175e-07, + "loss": 1.9439, + "step": 2020 + }, + { + "epoch": 0.13645092820322766, + "grad_norm": 6.277578595751896, + "learning_rate": 9.959570783860647e-07, + "loss": 1.9433, + "step": 2025 + }, + { + "epoch": 0.1367878440753344, + "grad_norm": 5.9244821304571085, + "learning_rate": 9.958821057420752e-07, + "loss": 1.9744, + "step": 2030 + }, + { + "epoch": 0.1371247599474411, + "grad_norm": 6.466226232433509, + "learning_rate": 9.958064471909513e-07, + "loss": 1.8964, + "step": 2035 + }, + { + "epoch": 0.13746167581954785, + "grad_norm": 6.203526305769233, + "learning_rate": 9.95730102837345e-07, + "loss": 1.9388, + "step": 2040 + }, + { + "epoch": 0.1377985916916546, + "grad_norm": 5.760768572732212, + "learning_rate": 9.956530727868558e-07, + "loss": 2.0363, + "step": 2045 + }, + { + "epoch": 0.13813550756376133, + "grad_norm": 5.932958764540261, + "learning_rate": 9.955753571460322e-07, + "loss": 1.974, + "step": 2050 + }, + { + "epoch": 0.13847242343586808, + "grad_norm": 6.4230773291754355, + "learning_rate": 9.95496956022371e-07, + "loss": 1.9678, + "step": 2055 + }, + { + "epoch": 0.1388093393079748, + "grad_norm": 6.824106716998237, + "learning_rate": 9.95417869524317e-07, + "loss": 1.9545, + "step": 2060 + }, + { + "epoch": 0.13914625518008153, + "grad_norm": 5.823178409847294, + "learning_rate": 9.953380977612633e-07, + "loss": 1.9823, + "step": 2065 + }, + { + "epoch": 0.13948317105218827, + "grad_norm": 6.257993657815616, + "learning_rate": 9.952576408435505e-07, + "loss": 2.0147, + "step": 2070 + }, + { + "epoch": 0.139820086924295, + "grad_norm": 6.2285691514819055, + "learning_rate": 9.951764988824674e-07, + "loss": 1.9516, + "step": 2075 + }, + { + "epoch": 0.14015700279640173, + "grad_norm": 6.261536455663687, + "learning_rate": 9.950946719902498e-07, + "loss": 1.9726, + "step": 2080 + }, + { + "epoch": 0.14049391866850847, + "grad_norm": 5.744470229374497, + "learning_rate": 9.950121602800813e-07, + "loss": 1.915, + "step": 2085 + }, + { + "epoch": 0.1408308345406152, + "grad_norm": 6.253873692382304, + "learning_rate": 9.949289638660922e-07, + "loss": 1.9089, + "step": 2090 + }, + { + "epoch": 0.14116775041272195, + "grad_norm": 6.137871997869609, + "learning_rate": 9.948450828633608e-07, + "loss": 1.9406, + "step": 2095 + }, + { + "epoch": 0.1415046662848287, + "grad_norm": 6.105512836268031, + "learning_rate": 9.947605173879115e-07, + "loss": 1.9098, + "step": 2100 + }, + { + "epoch": 0.1418415821569354, + "grad_norm": 6.35463151695237, + "learning_rate": 9.94675267556716e-07, + "loss": 1.9155, + "step": 2105 + }, + { + "epoch": 0.14217849802904214, + "grad_norm": 6.31848845137408, + "learning_rate": 9.94589333487692e-07, + "loss": 1.9398, + "step": 2110 + }, + { + "epoch": 0.14251541390114889, + "grad_norm": 6.328687308623832, + "learning_rate": 9.945027152997046e-07, + "loss": 2.0107, + "step": 2115 + }, + { + "epoch": 0.14285232977325563, + "grad_norm": 6.426232569229285, + "learning_rate": 9.944154131125642e-07, + "loss": 1.9887, + "step": 2120 + }, + { + "epoch": 0.14318924564536234, + "grad_norm": 6.532309209899036, + "learning_rate": 9.94327427047028e-07, + "loss": 1.9792, + "step": 2125 + }, + { + "epoch": 0.14352616151746908, + "grad_norm": 7.123050490115521, + "learning_rate": 9.942387572247983e-07, + "loss": 1.9612, + "step": 2130 + }, + { + "epoch": 0.14386307738957582, + "grad_norm": 6.195588758010388, + "learning_rate": 9.941494037685243e-07, + "loss": 1.973, + "step": 2135 + }, + { + "epoch": 0.14419999326168256, + "grad_norm": 5.832825874342457, + "learning_rate": 9.940593668017998e-07, + "loss": 1.947, + "step": 2140 + }, + { + "epoch": 0.1445369091337893, + "grad_norm": 6.986313603080345, + "learning_rate": 9.93968646449165e-07, + "loss": 1.9623, + "step": 2145 + }, + { + "epoch": 0.14487382500589602, + "grad_norm": 6.546972459648574, + "learning_rate": 9.938772428361045e-07, + "loss": 1.9806, + "step": 2150 + }, + { + "epoch": 0.14521074087800276, + "grad_norm": 6.958183204055856, + "learning_rate": 9.937851560890484e-07, + "loss": 1.9149, + "step": 2155 + }, + { + "epoch": 0.1455476567501095, + "grad_norm": 6.132052769974667, + "learning_rate": 9.936923863353717e-07, + "loss": 1.9865, + "step": 2160 + }, + { + "epoch": 0.14588457262221624, + "grad_norm": 6.241571590304282, + "learning_rate": 9.935989337033939e-07, + "loss": 1.9518, + "step": 2165 + }, + { + "epoch": 0.14622148849432295, + "grad_norm": 5.9203250462384505, + "learning_rate": 9.935047983223794e-07, + "loss": 1.8987, + "step": 2170 + }, + { + "epoch": 0.1465584043664297, + "grad_norm": 5.985556544125865, + "learning_rate": 9.934099803225367e-07, + "loss": 1.9217, + "step": 2175 + }, + { + "epoch": 0.14689532023853644, + "grad_norm": 5.730478097660369, + "learning_rate": 9.933144798350188e-07, + "loss": 1.9666, + "step": 2180 + }, + { + "epoch": 0.14723223611064318, + "grad_norm": 6.707764179086582, + "learning_rate": 9.932182969919228e-07, + "loss": 2.0318, + "step": 2185 + }, + { + "epoch": 0.14756915198274992, + "grad_norm": 5.5659888513546685, + "learning_rate": 9.931214319262885e-07, + "loss": 1.9082, + "step": 2190 + }, + { + "epoch": 0.14790606785485663, + "grad_norm": 6.374951323196418, + "learning_rate": 9.930238847721013e-07, + "loss": 1.9317, + "step": 2195 + }, + { + "epoch": 0.14824298372696337, + "grad_norm": 6.434039061973071, + "learning_rate": 9.929256556642884e-07, + "loss": 1.864, + "step": 2200 + }, + { + "epoch": 0.14857989959907011, + "grad_norm": 6.367451884410294, + "learning_rate": 9.92826744738721e-07, + "loss": 1.9982, + "step": 2205 + }, + { + "epoch": 0.14891681547117686, + "grad_norm": 6.6815782297960675, + "learning_rate": 9.927271521322134e-07, + "loss": 1.9709, + "step": 2210 + }, + { + "epoch": 0.14925373134328357, + "grad_norm": 6.516744855920352, + "learning_rate": 9.926268779825224e-07, + "loss": 1.9566, + "step": 2215 + }, + { + "epoch": 0.1495906472153903, + "grad_norm": 5.730312993883115, + "learning_rate": 9.925259224283484e-07, + "loss": 1.9446, + "step": 2220 + }, + { + "epoch": 0.14992756308749705, + "grad_norm": 6.467174318204597, + "learning_rate": 9.924242856093332e-07, + "loss": 1.9286, + "step": 2225 + }, + { + "epoch": 0.1502644789596038, + "grad_norm": 5.428856070091508, + "learning_rate": 9.923219676660614e-07, + "loss": 1.9131, + "step": 2230 + }, + { + "epoch": 0.15060139483171053, + "grad_norm": 5.820529761802369, + "learning_rate": 9.922189687400603e-07, + "loss": 1.9089, + "step": 2235 + }, + { + "epoch": 0.15093831070381725, + "grad_norm": 6.529777214957106, + "learning_rate": 9.921152889737984e-07, + "loss": 1.9469, + "step": 2240 + }, + { + "epoch": 0.151275226575924, + "grad_norm": 6.345390439569736, + "learning_rate": 9.92010928510686e-07, + "loss": 1.9395, + "step": 2245 + }, + { + "epoch": 0.15161214244803073, + "grad_norm": 6.62897291068532, + "learning_rate": 9.919058874950754e-07, + "loss": 1.9595, + "step": 2250 + }, + { + "epoch": 0.15194905832013747, + "grad_norm": 5.504413660555053, + "learning_rate": 9.9180016607226e-07, + "loss": 1.911, + "step": 2255 + }, + { + "epoch": 0.15228597419224418, + "grad_norm": 6.519560511343574, + "learning_rate": 9.916937643884737e-07, + "loss": 1.9846, + "step": 2260 + }, + { + "epoch": 0.15262289006435092, + "grad_norm": 6.975549195640513, + "learning_rate": 9.915866825908927e-07, + "loss": 1.977, + "step": 2265 + }, + { + "epoch": 0.15295980593645767, + "grad_norm": 5.767514386057251, + "learning_rate": 9.914789208276329e-07, + "loss": 2.0107, + "step": 2270 + }, + { + "epoch": 0.1532967218085644, + "grad_norm": 6.74373390316998, + "learning_rate": 9.913704792477511e-07, + "loss": 1.9681, + "step": 2275 + }, + { + "epoch": 0.15363363768067115, + "grad_norm": 5.4136605604834305, + "learning_rate": 9.91261358001244e-07, + "loss": 1.8636, + "step": 2280 + }, + { + "epoch": 0.15397055355277786, + "grad_norm": 6.113357161218679, + "learning_rate": 9.911515572390495e-07, + "loss": 1.9542, + "step": 2285 + }, + { + "epoch": 0.1543074694248846, + "grad_norm": 6.212873246787139, + "learning_rate": 9.91041077113044e-07, + "loss": 1.9415, + "step": 2290 + }, + { + "epoch": 0.15464438529699134, + "grad_norm": 6.310708548707665, + "learning_rate": 9.909299177760445e-07, + "loss": 1.9649, + "step": 2295 + }, + { + "epoch": 0.15498130116909808, + "grad_norm": 6.076947896759005, + "learning_rate": 9.90818079381807e-07, + "loss": 2.0335, + "step": 2300 + }, + { + "epoch": 0.1553182170412048, + "grad_norm": 6.587950508829805, + "learning_rate": 9.907055620850277e-07, + "loss": 1.918, + "step": 2305 + }, + { + "epoch": 0.15565513291331154, + "grad_norm": 6.945749330710282, + "learning_rate": 9.905923660413409e-07, + "loss": 1.8805, + "step": 2310 + }, + { + "epoch": 0.15599204878541828, + "grad_norm": 6.642951255259172, + "learning_rate": 9.904784914073196e-07, + "loss": 1.9538, + "step": 2315 + }, + { + "epoch": 0.15632896465752502, + "grad_norm": 6.302001230717917, + "learning_rate": 9.903639383404765e-07, + "loss": 1.9267, + "step": 2320 + }, + { + "epoch": 0.15666588052963176, + "grad_norm": 6.509732828699096, + "learning_rate": 9.902487069992618e-07, + "loss": 1.8941, + "step": 2325 + }, + { + "epoch": 0.15700279640173848, + "grad_norm": 6.472613626661652, + "learning_rate": 9.901327975430645e-07, + "loss": 1.9944, + "step": 2330 + }, + { + "epoch": 0.15733971227384522, + "grad_norm": 6.249063324051331, + "learning_rate": 9.900162101322106e-07, + "loss": 1.923, + "step": 2335 + }, + { + "epoch": 0.15767662814595196, + "grad_norm": 6.534842584367619, + "learning_rate": 9.898989449279653e-07, + "loss": 1.8986, + "step": 2340 + }, + { + "epoch": 0.1580135440180587, + "grad_norm": 6.129324353524715, + "learning_rate": 9.8978100209253e-07, + "loss": 1.9573, + "step": 2345 + }, + { + "epoch": 0.1583504598901654, + "grad_norm": 5.824833199759791, + "learning_rate": 9.89662381789044e-07, + "loss": 1.9809, + "step": 2350 + }, + { + "epoch": 0.15868737576227215, + "grad_norm": 6.143494469198798, + "learning_rate": 9.89543084181584e-07, + "loss": 1.8954, + "step": 2355 + }, + { + "epoch": 0.1590242916343789, + "grad_norm": 6.651525954696755, + "learning_rate": 9.894231094351628e-07, + "loss": 1.9549, + "step": 2360 + }, + { + "epoch": 0.15936120750648564, + "grad_norm": 5.827719542621931, + "learning_rate": 9.893024577157303e-07, + "loss": 1.9758, + "step": 2365 + }, + { + "epoch": 0.15969812337859238, + "grad_norm": 6.101811273783142, + "learning_rate": 9.891811291901727e-07, + "loss": 1.9159, + "step": 2370 + }, + { + "epoch": 0.1600350392506991, + "grad_norm": 6.166144715136228, + "learning_rate": 9.890591240263124e-07, + "loss": 1.9213, + "step": 2375 + }, + { + "epoch": 0.16037195512280583, + "grad_norm": 6.391645333892786, + "learning_rate": 9.889364423929075e-07, + "loss": 1.9033, + "step": 2380 + }, + { + "epoch": 0.16070887099491257, + "grad_norm": 6.144735650701305, + "learning_rate": 9.888130844596524e-07, + "loss": 1.9093, + "step": 2385 + }, + { + "epoch": 0.1610457868670193, + "grad_norm": 6.010045391891963, + "learning_rate": 9.88689050397176e-07, + "loss": 1.9673, + "step": 2390 + }, + { + "epoch": 0.16138270273912603, + "grad_norm": 6.10557730478434, + "learning_rate": 9.885643403770431e-07, + "loss": 1.8925, + "step": 2395 + }, + { + "epoch": 0.16171961861123277, + "grad_norm": 6.608433974354653, + "learning_rate": 9.884389545717538e-07, + "loss": 1.932, + "step": 2400 + }, + { + "epoch": 0.1620565344833395, + "grad_norm": 6.7409320976993, + "learning_rate": 9.88312893154742e-07, + "loss": 1.8663, + "step": 2405 + }, + { + "epoch": 0.16239345035544625, + "grad_norm": 5.941750402794903, + "learning_rate": 9.881861563003766e-07, + "loss": 2.0028, + "step": 2410 + }, + { + "epoch": 0.162730366227553, + "grad_norm": 6.014502158961904, + "learning_rate": 9.880587441839613e-07, + "loss": 1.9156, + "step": 2415 + }, + { + "epoch": 0.1630672820996597, + "grad_norm": 6.571932040836407, + "learning_rate": 9.87930656981733e-07, + "loss": 1.9331, + "step": 2420 + }, + { + "epoch": 0.16340419797176645, + "grad_norm": 6.066820980535922, + "learning_rate": 9.878018948708625e-07, + "loss": 1.8293, + "step": 2425 + }, + { + "epoch": 0.1637411138438732, + "grad_norm": 6.200211415486074, + "learning_rate": 9.876724580294546e-07, + "loss": 1.9523, + "step": 2430 + }, + { + "epoch": 0.16407802971597993, + "grad_norm": 5.933206529903557, + "learning_rate": 9.875423466365471e-07, + "loss": 1.9296, + "step": 2435 + }, + { + "epoch": 0.16441494558808664, + "grad_norm": 5.820888744313868, + "learning_rate": 9.874115608721107e-07, + "loss": 1.8975, + "step": 2440 + }, + { + "epoch": 0.16475186146019338, + "grad_norm": 6.413197920174298, + "learning_rate": 9.872801009170492e-07, + "loss": 1.9998, + "step": 2445 + }, + { + "epoch": 0.16508877733230012, + "grad_norm": 5.594241105149758, + "learning_rate": 9.871479669531988e-07, + "loss": 1.8642, + "step": 2450 + }, + { + "epoch": 0.16542569320440687, + "grad_norm": 6.90776423556707, + "learning_rate": 9.87015159163328e-07, + "loss": 1.9158, + "step": 2455 + }, + { + "epoch": 0.1657626090765136, + "grad_norm": 6.742200847923345, + "learning_rate": 9.868816777311372e-07, + "loss": 1.9738, + "step": 2460 + }, + { + "epoch": 0.16609952494862032, + "grad_norm": 6.242575863453409, + "learning_rate": 9.867475228412592e-07, + "loss": 1.9085, + "step": 2465 + }, + { + "epoch": 0.16643644082072706, + "grad_norm": 7.077187802421562, + "learning_rate": 9.866126946792572e-07, + "loss": 1.8906, + "step": 2470 + }, + { + "epoch": 0.1667733566928338, + "grad_norm": 6.158021286459406, + "learning_rate": 9.864771934316268e-07, + "loss": 1.913, + "step": 2475 + }, + { + "epoch": 0.16711027256494054, + "grad_norm": 6.122768934226615, + "learning_rate": 9.863410192857938e-07, + "loss": 1.9653, + "step": 2480 + }, + { + "epoch": 0.16744718843704726, + "grad_norm": 5.9040493114933925, + "learning_rate": 9.862041724301154e-07, + "loss": 1.9317, + "step": 2485 + }, + { + "epoch": 0.167784104309154, + "grad_norm": 6.091505622820139, + "learning_rate": 9.860666530538787e-07, + "loss": 1.9795, + "step": 2490 + }, + { + "epoch": 0.16812102018126074, + "grad_norm": 5.55852182181646, + "learning_rate": 9.859284613473017e-07, + "loss": 1.931, + "step": 2495 + }, + { + "epoch": 0.16845793605336748, + "grad_norm": 5.727558368634606, + "learning_rate": 9.857895975015318e-07, + "loss": 1.952, + "step": 2500 + }, + { + "epoch": 0.16879485192547422, + "grad_norm": 6.048697203389933, + "learning_rate": 9.856500617086463e-07, + "loss": 1.9123, + "step": 2505 + }, + { + "epoch": 0.16913176779758093, + "grad_norm": 6.200625030191421, + "learning_rate": 9.85509854161652e-07, + "loss": 1.8941, + "step": 2510 + }, + { + "epoch": 0.16946868366968768, + "grad_norm": 5.954977123661459, + "learning_rate": 9.853689750544849e-07, + "loss": 1.9823, + "step": 2515 + }, + { + "epoch": 0.16980559954179442, + "grad_norm": 6.080978939587595, + "learning_rate": 9.852274245820095e-07, + "loss": 1.928, + "step": 2520 + }, + { + "epoch": 0.17014251541390116, + "grad_norm": 7.661482448972233, + "learning_rate": 9.850852029400198e-07, + "loss": 1.9531, + "step": 2525 + }, + { + "epoch": 0.17047943128600787, + "grad_norm": 6.286433787084899, + "learning_rate": 9.849423103252374e-07, + "loss": 1.9028, + "step": 2530 + }, + { + "epoch": 0.1708163471581146, + "grad_norm": 6.3328085180536835, + "learning_rate": 9.84798746935312e-07, + "loss": 1.9073, + "step": 2535 + }, + { + "epoch": 0.17115326303022135, + "grad_norm": 6.181759007221361, + "learning_rate": 9.846545129688217e-07, + "loss": 1.8744, + "step": 2540 + }, + { + "epoch": 0.1714901789023281, + "grad_norm": 6.234298316262344, + "learning_rate": 9.845096086252716e-07, + "loss": 1.8956, + "step": 2545 + }, + { + "epoch": 0.17182709477443484, + "grad_norm": 5.998336546341908, + "learning_rate": 9.843640341050944e-07, + "loss": 1.9759, + "step": 2550 + }, + { + "epoch": 0.17216401064654155, + "grad_norm": 6.128708725791349, + "learning_rate": 9.842177896096493e-07, + "loss": 1.9879, + "step": 2555 + }, + { + "epoch": 0.1725009265186483, + "grad_norm": 6.013293725561392, + "learning_rate": 9.84070875341223e-07, + "loss": 1.9182, + "step": 2560 + }, + { + "epoch": 0.17283784239075503, + "grad_norm": 5.897020597604511, + "learning_rate": 9.83923291503028e-07, + "loss": 1.9216, + "step": 2565 + }, + { + "epoch": 0.17317475826286177, + "grad_norm": 6.147461118394779, + "learning_rate": 9.837750382992033e-07, + "loss": 1.9187, + "step": 2570 + }, + { + "epoch": 0.17351167413496849, + "grad_norm": 6.668024086437924, + "learning_rate": 9.836261159348135e-07, + "loss": 1.9328, + "step": 2575 + }, + { + "epoch": 0.17384859000707523, + "grad_norm": 6.175494197320494, + "learning_rate": 9.834765246158488e-07, + "loss": 1.9537, + "step": 2580 + }, + { + "epoch": 0.17418550587918197, + "grad_norm": 6.3176648197191065, + "learning_rate": 9.83326264549225e-07, + "loss": 1.9553, + "step": 2585 + }, + { + "epoch": 0.1745224217512887, + "grad_norm": 5.770121391032835, + "learning_rate": 9.83175335942783e-07, + "loss": 1.8756, + "step": 2590 + }, + { + "epoch": 0.17485933762339545, + "grad_norm": 6.12386478367424, + "learning_rate": 9.830237390052876e-07, + "loss": 1.8831, + "step": 2595 + }, + { + "epoch": 0.17519625349550216, + "grad_norm": 6.6115302969443, + "learning_rate": 9.82871473946429e-07, + "loss": 1.9609, + "step": 2600 + }, + { + "epoch": 0.1755331693676089, + "grad_norm": 6.51170744218166, + "learning_rate": 9.82718540976821e-07, + "loss": 1.9285, + "step": 2605 + }, + { + "epoch": 0.17587008523971565, + "grad_norm": 6.505287511470605, + "learning_rate": 9.825649403080015e-07, + "loss": 1.9212, + "step": 2610 + }, + { + "epoch": 0.1762070011118224, + "grad_norm": 6.1451891960370295, + "learning_rate": 9.824106721524317e-07, + "loss": 1.8681, + "step": 2615 + }, + { + "epoch": 0.1765439169839291, + "grad_norm": 6.153611709519882, + "learning_rate": 9.822557367234962e-07, + "loss": 1.9253, + "step": 2620 + }, + { + "epoch": 0.17688083285603584, + "grad_norm": 6.464659492133994, + "learning_rate": 9.82100134235503e-07, + "loss": 1.9525, + "step": 2625 + }, + { + "epoch": 0.17721774872814258, + "grad_norm": 7.049383941068097, + "learning_rate": 9.819438649036823e-07, + "loss": 1.9579, + "step": 2630 + }, + { + "epoch": 0.17755466460024932, + "grad_norm": 6.424061501155142, + "learning_rate": 9.817869289441864e-07, + "loss": 1.8578, + "step": 2635 + }, + { + "epoch": 0.17789158047235606, + "grad_norm": 6.040140871168396, + "learning_rate": 9.816293265740907e-07, + "loss": 1.9384, + "step": 2640 + }, + { + "epoch": 0.17822849634446278, + "grad_norm": 5.889576761087303, + "learning_rate": 9.81471058011391e-07, + "loss": 1.9539, + "step": 2645 + }, + { + "epoch": 0.17856541221656952, + "grad_norm": 6.08103957840839, + "learning_rate": 9.81312123475006e-07, + "loss": 1.933, + "step": 2650 + }, + { + "epoch": 0.17890232808867626, + "grad_norm": 6.525506562314697, + "learning_rate": 9.811525231847746e-07, + "loss": 1.9735, + "step": 2655 + }, + { + "epoch": 0.179239243960783, + "grad_norm": 6.399928080344757, + "learning_rate": 9.809922573614569e-07, + "loss": 1.902, + "step": 2660 + }, + { + "epoch": 0.17957615983288971, + "grad_norm": 6.7137591420055776, + "learning_rate": 9.808313262267337e-07, + "loss": 1.9049, + "step": 2665 + }, + { + "epoch": 0.17991307570499646, + "grad_norm": 6.283847903141634, + "learning_rate": 9.806697300032057e-07, + "loss": 1.9117, + "step": 2670 + }, + { + "epoch": 0.1802499915771032, + "grad_norm": 6.852639330917095, + "learning_rate": 9.805074689143938e-07, + "loss": 1.9271, + "step": 2675 + }, + { + "epoch": 0.18058690744920994, + "grad_norm": 6.646951296296031, + "learning_rate": 9.803445431847388e-07, + "loss": 1.8809, + "step": 2680 + }, + { + "epoch": 0.18092382332131668, + "grad_norm": 5.935782369576825, + "learning_rate": 9.801809530396003e-07, + "loss": 2.0025, + "step": 2685 + }, + { + "epoch": 0.1812607391934234, + "grad_norm": 5.9066632972863475, + "learning_rate": 9.800166987052572e-07, + "loss": 1.8773, + "step": 2690 + }, + { + "epoch": 0.18159765506553013, + "grad_norm": 5.9865063710361905, + "learning_rate": 9.798517804089072e-07, + "loss": 2.0383, + "step": 2695 + }, + { + "epoch": 0.18193457093763687, + "grad_norm": 5.631020483070982, + "learning_rate": 9.796861983786661e-07, + "loss": 1.862, + "step": 2700 + }, + { + "epoch": 0.18227148680974362, + "grad_norm": 5.856519798162219, + "learning_rate": 9.795199528435682e-07, + "loss": 1.9101, + "step": 2705 + }, + { + "epoch": 0.18260840268185033, + "grad_norm": 6.222085140585286, + "learning_rate": 9.793530440335654e-07, + "loss": 1.8835, + "step": 2710 + }, + { + "epoch": 0.18294531855395707, + "grad_norm": 6.359042832937223, + "learning_rate": 9.791854721795264e-07, + "loss": 1.8879, + "step": 2715 + }, + { + "epoch": 0.1832822344260638, + "grad_norm": 6.19407136291253, + "learning_rate": 9.790172375132385e-07, + "loss": 1.9003, + "step": 2720 + }, + { + "epoch": 0.18361915029817055, + "grad_norm": 5.6877990141134305, + "learning_rate": 9.788483402674041e-07, + "loss": 1.9884, + "step": 2725 + }, + { + "epoch": 0.1839560661702773, + "grad_norm": 6.151314409969232, + "learning_rate": 9.786787806756434e-07, + "loss": 1.9299, + "step": 2730 + }, + { + "epoch": 0.184292982042384, + "grad_norm": 6.575318390036048, + "learning_rate": 9.78508558972492e-07, + "loss": 1.934, + "step": 2735 + }, + { + "epoch": 0.18462989791449075, + "grad_norm": 5.919173357209151, + "learning_rate": 9.783376753934015e-07, + "loss": 1.9586, + "step": 2740 + }, + { + "epoch": 0.1849668137865975, + "grad_norm": 6.286929037909757, + "learning_rate": 9.781661301747393e-07, + "loss": 1.8594, + "step": 2745 + }, + { + "epoch": 0.18530372965870423, + "grad_norm": 6.31734888329098, + "learning_rate": 9.779939235537879e-07, + "loss": 1.9917, + "step": 2750 + }, + { + "epoch": 0.18564064553081094, + "grad_norm": 6.564405236880653, + "learning_rate": 9.778210557687443e-07, + "loss": 1.9372, + "step": 2755 + }, + { + "epoch": 0.18597756140291768, + "grad_norm": 6.256750421006695, + "learning_rate": 9.776475270587205e-07, + "loss": 1.9249, + "step": 2760 + }, + { + "epoch": 0.18631447727502443, + "grad_norm": 5.989838119364058, + "learning_rate": 9.774733376637421e-07, + "loss": 1.9155, + "step": 2765 + }, + { + "epoch": 0.18665139314713117, + "grad_norm": 6.446979854135191, + "learning_rate": 9.772984878247493e-07, + "loss": 1.9294, + "step": 2770 + }, + { + "epoch": 0.1869883090192379, + "grad_norm": 6.762973678300427, + "learning_rate": 9.771229777835952e-07, + "loss": 1.8937, + "step": 2775 + }, + { + "epoch": 0.18732522489134462, + "grad_norm": 5.787808739396207, + "learning_rate": 9.769468077830466e-07, + "loss": 1.89, + "step": 2780 + }, + { + "epoch": 0.18766214076345136, + "grad_norm": 6.000820283110658, + "learning_rate": 9.767699780667827e-07, + "loss": 1.9394, + "step": 2785 + }, + { + "epoch": 0.1879990566355581, + "grad_norm": 6.393820981303887, + "learning_rate": 9.765924888793955e-07, + "loss": 1.86, + "step": 2790 + }, + { + "epoch": 0.18833597250766484, + "grad_norm": 5.877371777184333, + "learning_rate": 9.76414340466389e-07, + "loss": 1.9419, + "step": 2795 + }, + { + "epoch": 0.18867288837977156, + "grad_norm": 6.908257437131703, + "learning_rate": 9.762355330741794e-07, + "loss": 1.9607, + "step": 2800 + }, + { + "epoch": 0.1890098042518783, + "grad_norm": 5.869255461740951, + "learning_rate": 9.760560669500941e-07, + "loss": 1.925, + "step": 2805 + }, + { + "epoch": 0.18934672012398504, + "grad_norm": 5.975145370261582, + "learning_rate": 9.758759423423716e-07, + "loss": 1.9067, + "step": 2810 + }, + { + "epoch": 0.18968363599609178, + "grad_norm": 6.276438455622569, + "learning_rate": 9.756951595001617e-07, + "loss": 1.8599, + "step": 2815 + }, + { + "epoch": 0.19002055186819852, + "grad_norm": 6.386035436547608, + "learning_rate": 9.755137186735238e-07, + "loss": 1.9101, + "step": 2820 + }, + { + "epoch": 0.19035746774030524, + "grad_norm": 6.498761340697311, + "learning_rate": 9.753316201134282e-07, + "loss": 1.9375, + "step": 2825 + }, + { + "epoch": 0.19069438361241198, + "grad_norm": 5.868215854791331, + "learning_rate": 9.75148864071755e-07, + "loss": 1.8861, + "step": 2830 + }, + { + "epoch": 0.19103129948451872, + "grad_norm": 6.259936017759626, + "learning_rate": 9.74965450801293e-07, + "loss": 1.9049, + "step": 2835 + }, + { + "epoch": 0.19136821535662546, + "grad_norm": 6.343131958265802, + "learning_rate": 9.747813805557408e-07, + "loss": 1.898, + "step": 2840 + }, + { + "epoch": 0.19170513122873217, + "grad_norm": 6.048544900470273, + "learning_rate": 9.745966535897054e-07, + "loss": 1.8809, + "step": 2845 + }, + { + "epoch": 0.19204204710083891, + "grad_norm": 5.9281878870137, + "learning_rate": 9.744112701587024e-07, + "loss": 1.9059, + "step": 2850 + }, + { + "epoch": 0.19237896297294566, + "grad_norm": 6.21419556733591, + "learning_rate": 9.742252305191551e-07, + "loss": 1.8776, + "step": 2855 + }, + { + "epoch": 0.1927158788450524, + "grad_norm": 6.4341300465391775, + "learning_rate": 9.740385349283946e-07, + "loss": 1.9311, + "step": 2860 + }, + { + "epoch": 0.19305279471715914, + "grad_norm": 6.42152192546941, + "learning_rate": 9.738511836446596e-07, + "loss": 1.9562, + "step": 2865 + }, + { + "epoch": 0.19338971058926585, + "grad_norm": 5.824133775486518, + "learning_rate": 9.736631769270957e-07, + "loss": 1.9235, + "step": 2870 + }, + { + "epoch": 0.1937266264613726, + "grad_norm": 6.247305043191556, + "learning_rate": 9.734745150357544e-07, + "loss": 1.9226, + "step": 2875 + }, + { + "epoch": 0.19406354233347933, + "grad_norm": 6.080141784771128, + "learning_rate": 9.732851982315944e-07, + "loss": 1.9777, + "step": 2880 + }, + { + "epoch": 0.19440045820558607, + "grad_norm": 6.322132356228769, + "learning_rate": 9.730952267764796e-07, + "loss": 1.896, + "step": 2885 + }, + { + "epoch": 0.1947373740776928, + "grad_norm": 6.433030057968718, + "learning_rate": 9.729046009331798e-07, + "loss": 1.877, + "step": 2890 + }, + { + "epoch": 0.19507428994979953, + "grad_norm": 6.10661292779823, + "learning_rate": 9.727133209653696e-07, + "loss": 1.9049, + "step": 2895 + }, + { + "epoch": 0.19541120582190627, + "grad_norm": 5.892069850886527, + "learning_rate": 9.72521387137629e-07, + "loss": 1.9512, + "step": 2900 + }, + { + "epoch": 0.195748121694013, + "grad_norm": 6.26071736541495, + "learning_rate": 9.723287997154419e-07, + "loss": 1.9702, + "step": 2905 + }, + { + "epoch": 0.19608503756611975, + "grad_norm": 5.794806052244091, + "learning_rate": 9.72135558965196e-07, + "loss": 1.921, + "step": 2910 + }, + { + "epoch": 0.19642195343822647, + "grad_norm": 5.809504947809818, + "learning_rate": 9.719416651541837e-07, + "loss": 1.8608, + "step": 2915 + }, + { + "epoch": 0.1967588693103332, + "grad_norm": 5.814187161790213, + "learning_rate": 9.717471185505996e-07, + "loss": 1.8828, + "step": 2920 + }, + { + "epoch": 0.19709578518243995, + "grad_norm": 6.093734929582276, + "learning_rate": 9.715519194235422e-07, + "loss": 1.8619, + "step": 2925 + }, + { + "epoch": 0.1974327010545467, + "grad_norm": 6.3046806166987714, + "learning_rate": 9.713560680430117e-07, + "loss": 1.9335, + "step": 2930 + }, + { + "epoch": 0.1977696169266534, + "grad_norm": 6.427520474069085, + "learning_rate": 9.71159564679911e-07, + "loss": 1.909, + "step": 2935 + }, + { + "epoch": 0.19810653279876014, + "grad_norm": 6.232853782135515, + "learning_rate": 9.709624096060449e-07, + "loss": 1.8993, + "step": 2940 + }, + { + "epoch": 0.19844344867086688, + "grad_norm": 5.853044309136511, + "learning_rate": 9.707646030941192e-07, + "loss": 1.8984, + "step": 2945 + }, + { + "epoch": 0.19878036454297363, + "grad_norm": 6.136548097509919, + "learning_rate": 9.705661454177416e-07, + "loss": 1.9118, + "step": 2950 + }, + { + "epoch": 0.19911728041508037, + "grad_norm": 6.534000467489757, + "learning_rate": 9.703670368514192e-07, + "loss": 1.9262, + "step": 2955 + }, + { + "epoch": 0.19945419628718708, + "grad_norm": 6.290539857874704, + "learning_rate": 9.701672776705609e-07, + "loss": 1.9088, + "step": 2960 + }, + { + "epoch": 0.19979111215929382, + "grad_norm": 6.302727180291852, + "learning_rate": 9.699668681514746e-07, + "loss": 1.9414, + "step": 2965 + }, + { + "epoch": 0.20012802803140056, + "grad_norm": 5.664375469611846, + "learning_rate": 9.697658085713676e-07, + "loss": 1.9472, + "step": 2970 + }, + { + "epoch": 0.2004649439035073, + "grad_norm": 6.866737640209192, + "learning_rate": 9.695640992083471e-07, + "loss": 1.8569, + "step": 2975 + }, + { + "epoch": 0.20080185977561402, + "grad_norm": 6.152666956064643, + "learning_rate": 9.693617403414188e-07, + "loss": 1.9055, + "step": 2980 + }, + { + "epoch": 0.20113877564772076, + "grad_norm": 6.514327490736937, + "learning_rate": 9.691587322504865e-07, + "loss": 1.8868, + "step": 2985 + }, + { + "epoch": 0.2014756915198275, + "grad_norm": 6.682738501012025, + "learning_rate": 9.68955075216352e-07, + "loss": 1.9672, + "step": 2990 + }, + { + "epoch": 0.20181260739193424, + "grad_norm": 6.522921716857152, + "learning_rate": 9.687507695207154e-07, + "loss": 1.9148, + "step": 2995 + }, + { + "epoch": 0.20214952326404098, + "grad_norm": 6.318875468016384, + "learning_rate": 9.685458154461731e-07, + "loss": 1.878, + "step": 3000 + }, + { + "epoch": 0.2024864391361477, + "grad_norm": 5.609414606888322, + "learning_rate": 9.683402132762193e-07, + "loss": 1.937, + "step": 3005 + }, + { + "epoch": 0.20282335500825444, + "grad_norm": 6.2216093413120985, + "learning_rate": 9.68133963295244e-07, + "loss": 1.9736, + "step": 3010 + }, + { + "epoch": 0.20316027088036118, + "grad_norm": 6.189142393589986, + "learning_rate": 9.679270657885334e-07, + "loss": 1.8393, + "step": 3015 + }, + { + "epoch": 0.20349718675246792, + "grad_norm": 6.203512003745445, + "learning_rate": 9.677195210422693e-07, + "loss": 1.9523, + "step": 3020 + }, + { + "epoch": 0.20383410262457463, + "grad_norm": 6.144164019790604, + "learning_rate": 9.675113293435288e-07, + "loss": 1.9024, + "step": 3025 + }, + { + "epoch": 0.20417101849668137, + "grad_norm": 5.712284742254756, + "learning_rate": 9.673024909802841e-07, + "loss": 1.8785, + "step": 3030 + }, + { + "epoch": 0.2045079343687881, + "grad_norm": 5.670735815579872, + "learning_rate": 9.670930062414017e-07, + "loss": 1.8413, + "step": 3035 + }, + { + "epoch": 0.20484485024089485, + "grad_norm": 6.638834689424272, + "learning_rate": 9.66882875416642e-07, + "loss": 1.9462, + "step": 3040 + }, + { + "epoch": 0.2051817661130016, + "grad_norm": 6.2519172937020535, + "learning_rate": 9.666720987966595e-07, + "loss": 1.903, + "step": 3045 + }, + { + "epoch": 0.2055186819851083, + "grad_norm": 6.05215038719685, + "learning_rate": 9.664606766730012e-07, + "loss": 1.9139, + "step": 3050 + }, + { + "epoch": 0.20585559785721505, + "grad_norm": 6.0549888495357465, + "learning_rate": 9.662486093381082e-07, + "loss": 1.905, + "step": 3055 + }, + { + "epoch": 0.2061925137293218, + "grad_norm": 5.955088538325352, + "learning_rate": 9.660358970853126e-07, + "loss": 1.8374, + "step": 3060 + }, + { + "epoch": 0.20652942960142853, + "grad_norm": 5.905908530023023, + "learning_rate": 9.658225402088395e-07, + "loss": 1.8623, + "step": 3065 + }, + { + "epoch": 0.20686634547353525, + "grad_norm": 5.872087794008919, + "learning_rate": 9.656085390038058e-07, + "loss": 1.9301, + "step": 3070 + }, + { + "epoch": 0.207203261345642, + "grad_norm": 6.220493564964145, + "learning_rate": 9.653938937662187e-07, + "loss": 1.9098, + "step": 3075 + }, + { + "epoch": 0.20754017721774873, + "grad_norm": 5.761561040327359, + "learning_rate": 9.651786047929772e-07, + "loss": 1.8888, + "step": 3080 + }, + { + "epoch": 0.20787709308985547, + "grad_norm": 6.119415940843171, + "learning_rate": 9.649626723818702e-07, + "loss": 1.9355, + "step": 3085 + }, + { + "epoch": 0.2082140089619622, + "grad_norm": 6.212140870550058, + "learning_rate": 9.647460968315767e-07, + "loss": 1.9884, + "step": 3090 + }, + { + "epoch": 0.20855092483406892, + "grad_norm": 6.31960790183109, + "learning_rate": 9.645288784416652e-07, + "loss": 1.8941, + "step": 3095 + }, + { + "epoch": 0.20888784070617566, + "grad_norm": 6.198869449733293, + "learning_rate": 9.643110175125935e-07, + "loss": 1.9418, + "step": 3100 + }, + { + "epoch": 0.2092247565782824, + "grad_norm": 5.994224284054857, + "learning_rate": 9.640925143457084e-07, + "loss": 1.8749, + "step": 3105 + }, + { + "epoch": 0.20956167245038915, + "grad_norm": 6.534637990302735, + "learning_rate": 9.638733692432448e-07, + "loss": 1.9156, + "step": 3110 + }, + { + "epoch": 0.20989858832249586, + "grad_norm": 6.350636283397908, + "learning_rate": 9.636535825083252e-07, + "loss": 1.9408, + "step": 3115 + }, + { + "epoch": 0.2102355041946026, + "grad_norm": 5.747429709584084, + "learning_rate": 9.634331544449601e-07, + "loss": 1.9303, + "step": 3120 + }, + { + "epoch": 0.21057242006670934, + "grad_norm": 6.050823517306343, + "learning_rate": 9.632120853580472e-07, + "loss": 1.9323, + "step": 3125 + }, + { + "epoch": 0.21090933593881608, + "grad_norm": 5.676142963813065, + "learning_rate": 9.6299037555337e-07, + "loss": 1.865, + "step": 3130 + }, + { + "epoch": 0.21124625181092282, + "grad_norm": 6.296718830562775, + "learning_rate": 9.627680253375997e-07, + "loss": 1.8542, + "step": 3135 + }, + { + "epoch": 0.21158316768302954, + "grad_norm": 6.182241743324326, + "learning_rate": 9.625450350182918e-07, + "loss": 1.9004, + "step": 3140 + }, + { + "epoch": 0.21192008355513628, + "grad_norm": 5.703077632023605, + "learning_rate": 9.62321404903888e-07, + "loss": 1.8512, + "step": 3145 + }, + { + "epoch": 0.21225699942724302, + "grad_norm": 5.649919083346945, + "learning_rate": 9.620971353037148e-07, + "loss": 1.8742, + "step": 3150 + }, + { + "epoch": 0.21259391529934976, + "grad_norm": 6.162255987071309, + "learning_rate": 9.618722265279835e-07, + "loss": 1.9086, + "step": 3155 + }, + { + "epoch": 0.21293083117145647, + "grad_norm": 6.877459967522398, + "learning_rate": 9.61646678887789e-07, + "loss": 1.9141, + "step": 3160 + }, + { + "epoch": 0.21326774704356322, + "grad_norm": 7.064015072257185, + "learning_rate": 9.614204926951102e-07, + "loss": 1.9449, + "step": 3165 + }, + { + "epoch": 0.21360466291566996, + "grad_norm": 5.8916080827898965, + "learning_rate": 9.611936682628095e-07, + "loss": 1.9092, + "step": 3170 + }, + { + "epoch": 0.2139415787877767, + "grad_norm": 5.964533043114863, + "learning_rate": 9.609662059046315e-07, + "loss": 1.8819, + "step": 3175 + }, + { + "epoch": 0.21427849465988344, + "grad_norm": 5.916309521260202, + "learning_rate": 9.607381059352038e-07, + "loss": 1.8941, + "step": 3180 + }, + { + "epoch": 0.21461541053199015, + "grad_norm": 6.145903770013445, + "learning_rate": 9.605093686700353e-07, + "loss": 1.9525, + "step": 3185 + }, + { + "epoch": 0.2149523264040969, + "grad_norm": 6.26210111100357, + "learning_rate": 9.602799944255172e-07, + "loss": 1.8631, + "step": 3190 + }, + { + "epoch": 0.21528924227620364, + "grad_norm": 6.033360078420039, + "learning_rate": 9.60049983518921e-07, + "loss": 1.9637, + "step": 3195 + }, + { + "epoch": 0.21562615814831038, + "grad_norm": 5.862437650692546, + "learning_rate": 9.598193362683995e-07, + "loss": 1.8619, + "step": 3200 + }, + { + "epoch": 0.2159630740204171, + "grad_norm": 6.846493212362844, + "learning_rate": 9.59588052992985e-07, + "loss": 1.8711, + "step": 3205 + }, + { + "epoch": 0.21629998989252383, + "grad_norm": 6.054162969621891, + "learning_rate": 9.5935613401259e-07, + "loss": 1.9317, + "step": 3210 + }, + { + "epoch": 0.21663690576463057, + "grad_norm": 5.853489681208746, + "learning_rate": 9.591235796480064e-07, + "loss": 1.9221, + "step": 3215 + }, + { + "epoch": 0.2169738216367373, + "grad_norm": 6.026177124643552, + "learning_rate": 9.588903902209048e-07, + "loss": 1.8886, + "step": 3220 + }, + { + "epoch": 0.21731073750884405, + "grad_norm": 6.026706526511821, + "learning_rate": 9.586565660538343e-07, + "loss": 1.875, + "step": 3225 + }, + { + "epoch": 0.21764765338095077, + "grad_norm": 5.931027220154246, + "learning_rate": 9.584221074702217e-07, + "loss": 1.9771, + "step": 3230 + }, + { + "epoch": 0.2179845692530575, + "grad_norm": 6.178018868545221, + "learning_rate": 9.581870147943715e-07, + "loss": 1.893, + "step": 3235 + }, + { + "epoch": 0.21832148512516425, + "grad_norm": 5.983546923887888, + "learning_rate": 9.579512883514656e-07, + "loss": 1.9382, + "step": 3240 + }, + { + "epoch": 0.218658400997271, + "grad_norm": 6.313877194235666, + "learning_rate": 9.577149284675619e-07, + "loss": 1.9233, + "step": 3245 + }, + { + "epoch": 0.2189953168693777, + "grad_norm": 5.647925594777457, + "learning_rate": 9.574779354695951e-07, + "loss": 1.9285, + "step": 3250 + }, + { + "epoch": 0.21933223274148445, + "grad_norm": 6.052584204662514, + "learning_rate": 9.572403096853754e-07, + "loss": 1.8937, + "step": 3255 + }, + { + "epoch": 0.2196691486135912, + "grad_norm": 5.78982516243247, + "learning_rate": 9.570020514435878e-07, + "loss": 1.9158, + "step": 3260 + }, + { + "epoch": 0.22000606448569793, + "grad_norm": 6.5555394076193165, + "learning_rate": 9.567631610737929e-07, + "loss": 1.8403, + "step": 3265 + }, + { + "epoch": 0.22034298035780467, + "grad_norm": 6.244304553460857, + "learning_rate": 9.565236389064255e-07, + "loss": 1.9387, + "step": 3270 + }, + { + "epoch": 0.22067989622991138, + "grad_norm": 6.019749558522282, + "learning_rate": 9.562834852727935e-07, + "loss": 1.9252, + "step": 3275 + }, + { + "epoch": 0.22101681210201812, + "grad_norm": 6.030689411078262, + "learning_rate": 9.560427005050793e-07, + "loss": 1.9028, + "step": 3280 + }, + { + "epoch": 0.22135372797412486, + "grad_norm": 5.829221689829186, + "learning_rate": 9.55801284936338e-07, + "loss": 1.88, + "step": 3285 + }, + { + "epoch": 0.2216906438462316, + "grad_norm": 6.352115907311686, + "learning_rate": 9.555592389004966e-07, + "loss": 1.8563, + "step": 3290 + }, + { + "epoch": 0.22202755971833832, + "grad_norm": 6.346082954813021, + "learning_rate": 9.553165627323548e-07, + "loss": 1.9118, + "step": 3295 + }, + { + "epoch": 0.22236447559044506, + "grad_norm": 6.139627054619066, + "learning_rate": 9.55073256767584e-07, + "loss": 1.8908, + "step": 3300 + }, + { + "epoch": 0.2227013914625518, + "grad_norm": 6.214924215026803, + "learning_rate": 9.548293213427262e-07, + "loss": 1.8842, + "step": 3305 + }, + { + "epoch": 0.22303830733465854, + "grad_norm": 5.962888824637297, + "learning_rate": 9.545847567951944e-07, + "loss": 1.9077, + "step": 3310 + }, + { + "epoch": 0.22337522320676528, + "grad_norm": 6.353914244158636, + "learning_rate": 9.543395634632721e-07, + "loss": 1.9208, + "step": 3315 + }, + { + "epoch": 0.223712139078872, + "grad_norm": 5.778557004198218, + "learning_rate": 9.540937416861117e-07, + "loss": 1.8881, + "step": 3320 + }, + { + "epoch": 0.22404905495097874, + "grad_norm": 7.2073119475486545, + "learning_rate": 9.538472918037356e-07, + "loss": 1.9496, + "step": 3325 + }, + { + "epoch": 0.22438597082308548, + "grad_norm": 5.727299882070468, + "learning_rate": 9.536002141570348e-07, + "loss": 1.946, + "step": 3330 + }, + { + "epoch": 0.22472288669519222, + "grad_norm": 5.994342052634025, + "learning_rate": 9.533525090877688e-07, + "loss": 1.9398, + "step": 3335 + }, + { + "epoch": 0.22505980256729893, + "grad_norm": 6.327096679874867, + "learning_rate": 9.531041769385641e-07, + "loss": 1.8929, + "step": 3340 + }, + { + "epoch": 0.22539671843940567, + "grad_norm": 6.51167490064826, + "learning_rate": 9.528552180529161e-07, + "loss": 2.0104, + "step": 3345 + }, + { + "epoch": 0.22573363431151242, + "grad_norm": 6.516651466812798, + "learning_rate": 9.526056327751856e-07, + "loss": 1.9016, + "step": 3350 + }, + { + "epoch": 0.22607055018361916, + "grad_norm": 6.6127345474256956, + "learning_rate": 9.523554214506006e-07, + "loss": 1.9314, + "step": 3355 + }, + { + "epoch": 0.2264074660557259, + "grad_norm": 6.839303175267801, + "learning_rate": 9.521045844252551e-07, + "loss": 1.9305, + "step": 3360 + }, + { + "epoch": 0.2267443819278326, + "grad_norm": 6.186255355465, + "learning_rate": 9.518531220461084e-07, + "loss": 1.8877, + "step": 3365 + }, + { + "epoch": 0.22708129779993935, + "grad_norm": 5.369845972157766, + "learning_rate": 9.516010346609845e-07, + "loss": 1.8766, + "step": 3370 + }, + { + "epoch": 0.2274182136720461, + "grad_norm": 6.378683661332048, + "learning_rate": 9.513483226185723e-07, + "loss": 1.9803, + "step": 3375 + }, + { + "epoch": 0.22775512954415283, + "grad_norm": 6.484312873315589, + "learning_rate": 9.510949862684248e-07, + "loss": 1.8739, + "step": 3380 + }, + { + "epoch": 0.22809204541625955, + "grad_norm": 6.0136471572220165, + "learning_rate": 9.508410259609583e-07, + "loss": 1.917, + "step": 3385 + }, + { + "epoch": 0.2284289612883663, + "grad_norm": 6.250184322801729, + "learning_rate": 9.505864420474522e-07, + "loss": 1.9138, + "step": 3390 + }, + { + "epoch": 0.22876587716047303, + "grad_norm": 6.3250547018073195, + "learning_rate": 9.503312348800485e-07, + "loss": 1.8632, + "step": 3395 + }, + { + "epoch": 0.22910279303257977, + "grad_norm": 5.740473890542984, + "learning_rate": 9.500754048117514e-07, + "loss": 1.8823, + "step": 3400 + }, + { + "epoch": 0.2294397089046865, + "grad_norm": 5.777022768497205, + "learning_rate": 9.498189521964263e-07, + "loss": 1.8323, + "step": 3405 + }, + { + "epoch": 0.22977662477679323, + "grad_norm": 6.20964502865978, + "learning_rate": 9.495618773888006e-07, + "loss": 1.8181, + "step": 3410 + }, + { + "epoch": 0.23011354064889997, + "grad_norm": 6.134239103081429, + "learning_rate": 9.49304180744461e-07, + "loss": 1.9124, + "step": 3415 + }, + { + "epoch": 0.2304504565210067, + "grad_norm": 5.766924227242812, + "learning_rate": 9.490458626198556e-07, + "loss": 1.8891, + "step": 3420 + }, + { + "epoch": 0.23078737239311345, + "grad_norm": 6.476492599044412, + "learning_rate": 9.487869233722915e-07, + "loss": 1.879, + "step": 3425 + }, + { + "epoch": 0.23112428826522016, + "grad_norm": 6.1492500389967875, + "learning_rate": 9.485273633599348e-07, + "loss": 1.8936, + "step": 3430 + }, + { + "epoch": 0.2314612041373269, + "grad_norm": 6.107479826667597, + "learning_rate": 9.482671829418107e-07, + "loss": 1.8926, + "step": 3435 + }, + { + "epoch": 0.23179812000943364, + "grad_norm": 5.833724685300883, + "learning_rate": 9.480063824778024e-07, + "loss": 1.8725, + "step": 3440 + }, + { + "epoch": 0.23213503588154039, + "grad_norm": 6.169800273715307, + "learning_rate": 9.477449623286505e-07, + "loss": 1.8361, + "step": 3445 + }, + { + "epoch": 0.23247195175364713, + "grad_norm": 5.717025967015645, + "learning_rate": 9.474829228559529e-07, + "loss": 1.9336, + "step": 3450 + }, + { + "epoch": 0.23280886762575384, + "grad_norm": 6.571826219281641, + "learning_rate": 9.472202644221643e-07, + "loss": 1.9702, + "step": 3455 + }, + { + "epoch": 0.23314578349786058, + "grad_norm": 6.267842479579244, + "learning_rate": 9.469569873905955e-07, + "loss": 1.8596, + "step": 3460 + }, + { + "epoch": 0.23348269936996732, + "grad_norm": 5.971915133277161, + "learning_rate": 9.466930921254128e-07, + "loss": 1.8525, + "step": 3465 + }, + { + "epoch": 0.23381961524207406, + "grad_norm": 6.562137662722172, + "learning_rate": 9.464285789916376e-07, + "loss": 1.8986, + "step": 3470 + }, + { + "epoch": 0.23415653111418078, + "grad_norm": 6.385638926464063, + "learning_rate": 9.461634483551464e-07, + "loss": 1.843, + "step": 3475 + }, + { + "epoch": 0.23449344698628752, + "grad_norm": 5.845903723462544, + "learning_rate": 9.458977005826691e-07, + "loss": 1.8584, + "step": 3480 + }, + { + "epoch": 0.23483036285839426, + "grad_norm": 6.2980063600222085, + "learning_rate": 9.456313360417899e-07, + "loss": 1.9289, + "step": 3485 + }, + { + "epoch": 0.235167278730501, + "grad_norm": 5.683558465001443, + "learning_rate": 9.453643551009459e-07, + "loss": 1.8656, + "step": 3490 + }, + { + "epoch": 0.23550419460260774, + "grad_norm": 6.508247982606044, + "learning_rate": 9.450967581294265e-07, + "loss": 1.9393, + "step": 3495 + }, + { + "epoch": 0.23584111047471445, + "grad_norm": 5.700511178061109, + "learning_rate": 9.448285454973737e-07, + "loss": 1.8814, + "step": 3500 + }, + { + "epoch": 0.2361780263468212, + "grad_norm": 6.903117920839508, + "learning_rate": 9.445597175757806e-07, + "loss": 1.8339, + "step": 3505 + }, + { + "epoch": 0.23651494221892794, + "grad_norm": 5.983876339449441, + "learning_rate": 9.442902747364918e-07, + "loss": 1.8379, + "step": 3510 + }, + { + "epoch": 0.23685185809103468, + "grad_norm": 5.918219144963591, + "learning_rate": 9.440202173522022e-07, + "loss": 1.9062, + "step": 3515 + }, + { + "epoch": 0.2371887739631414, + "grad_norm": 6.053975291994282, + "learning_rate": 9.437495457964568e-07, + "loss": 1.8807, + "step": 3520 + }, + { + "epoch": 0.23752568983524813, + "grad_norm": 5.791738879936693, + "learning_rate": 9.434782604436502e-07, + "loss": 1.8712, + "step": 3525 + }, + { + "epoch": 0.23786260570735487, + "grad_norm": 6.018683919389132, + "learning_rate": 9.432063616690258e-07, + "loss": 1.8687, + "step": 3530 + }, + { + "epoch": 0.23819952157946161, + "grad_norm": 6.984256425631296, + "learning_rate": 9.429338498486758e-07, + "loss": 1.842, + "step": 3535 + }, + { + "epoch": 0.23853643745156836, + "grad_norm": 6.014792567520862, + "learning_rate": 9.426607253595402e-07, + "loss": 1.915, + "step": 3540 + }, + { + "epoch": 0.23887335332367507, + "grad_norm": 5.52165584469138, + "learning_rate": 9.423869885794063e-07, + "loss": 1.896, + "step": 3545 + }, + { + "epoch": 0.2392102691957818, + "grad_norm": 6.341307272489993, + "learning_rate": 9.421126398869086e-07, + "loss": 1.9376, + "step": 3550 + }, + { + "epoch": 0.23954718506788855, + "grad_norm": 6.3188324705440015, + "learning_rate": 9.418376796615279e-07, + "loss": 1.8711, + "step": 3555 + }, + { + "epoch": 0.2398841009399953, + "grad_norm": 6.035398818050593, + "learning_rate": 9.415621082835908e-07, + "loss": 1.8746, + "step": 3560 + }, + { + "epoch": 0.240221016812102, + "grad_norm": 5.711076837553516, + "learning_rate": 9.412859261342691e-07, + "loss": 1.8408, + "step": 3565 + }, + { + "epoch": 0.24055793268420875, + "grad_norm": 5.993468075806861, + "learning_rate": 9.410091335955798e-07, + "loss": 1.8672, + "step": 3570 + }, + { + "epoch": 0.2408948485563155, + "grad_norm": 6.360069219854257, + "learning_rate": 9.407317310503841e-07, + "loss": 1.8968, + "step": 3575 + }, + { + "epoch": 0.24123176442842223, + "grad_norm": 6.56717329121729, + "learning_rate": 9.404537188823869e-07, + "loss": 1.9586, + "step": 3580 + }, + { + "epoch": 0.24156868030052897, + "grad_norm": 5.796197742968857, + "learning_rate": 9.40175097476136e-07, + "loss": 1.8776, + "step": 3585 + }, + { + "epoch": 0.24190559617263568, + "grad_norm": 6.672312881743043, + "learning_rate": 9.398958672170225e-07, + "loss": 1.8598, + "step": 3590 + }, + { + "epoch": 0.24224251204474243, + "grad_norm": 6.144503989213299, + "learning_rate": 9.396160284912795e-07, + "loss": 1.9091, + "step": 3595 + }, + { + "epoch": 0.24257942791684917, + "grad_norm": 6.0819379510788245, + "learning_rate": 9.393355816859813e-07, + "loss": 1.8915, + "step": 3600 + }, + { + "epoch": 0.2429163437889559, + "grad_norm": 6.323366435634516, + "learning_rate": 9.390545271890437e-07, + "loss": 1.9043, + "step": 3605 + }, + { + "epoch": 0.24325325966106262, + "grad_norm": 6.4339698201446875, + "learning_rate": 9.387728653892233e-07, + "loss": 1.881, + "step": 3610 + }, + { + "epoch": 0.24359017553316936, + "grad_norm": 6.001463813574768, + "learning_rate": 9.384905966761159e-07, + "loss": 1.9053, + "step": 3615 + }, + { + "epoch": 0.2439270914052761, + "grad_norm": 6.399446616197819, + "learning_rate": 9.382077214401576e-07, + "loss": 1.8755, + "step": 3620 + }, + { + "epoch": 0.24426400727738284, + "grad_norm": 6.66748750013314, + "learning_rate": 9.379242400726232e-07, + "loss": 1.9006, + "step": 3625 + }, + { + "epoch": 0.24460092314948959, + "grad_norm": 6.826259347914525, + "learning_rate": 9.376401529656257e-07, + "loss": 1.9043, + "step": 3630 + }, + { + "epoch": 0.2449378390215963, + "grad_norm": 5.891290217702254, + "learning_rate": 9.373554605121161e-07, + "loss": 1.8244, + "step": 3635 + }, + { + "epoch": 0.24527475489370304, + "grad_norm": 6.181375359799177, + "learning_rate": 9.370701631058828e-07, + "loss": 1.8704, + "step": 3640 + }, + { + "epoch": 0.24561167076580978, + "grad_norm": 6.187612179223154, + "learning_rate": 9.367842611415508e-07, + "loss": 1.9051, + "step": 3645 + }, + { + "epoch": 0.24594858663791652, + "grad_norm": 5.964662147997069, + "learning_rate": 9.364977550145816e-07, + "loss": 1.8619, + "step": 3650 + }, + { + "epoch": 0.24628550251002324, + "grad_norm": 5.965890038499228, + "learning_rate": 9.362106451212721e-07, + "loss": 1.9405, + "step": 3655 + }, + { + "epoch": 0.24662241838212998, + "grad_norm": 6.664341876347663, + "learning_rate": 9.359229318587545e-07, + "loss": 1.9019, + "step": 3660 + }, + { + "epoch": 0.24695933425423672, + "grad_norm": 6.259088317611612, + "learning_rate": 9.356346156249954e-07, + "loss": 1.903, + "step": 3665 + }, + { + "epoch": 0.24729625012634346, + "grad_norm": 6.153712987251788, + "learning_rate": 9.353456968187958e-07, + "loss": 1.9271, + "step": 3670 + }, + { + "epoch": 0.2476331659984502, + "grad_norm": 5.8187434752693115, + "learning_rate": 9.350561758397897e-07, + "loss": 1.899, + "step": 3675 + }, + { + "epoch": 0.2479700818705569, + "grad_norm": 6.9370464485783545, + "learning_rate": 9.347660530884442e-07, + "loss": 1.9089, + "step": 3680 + }, + { + "epoch": 0.24830699774266365, + "grad_norm": 5.993727430508131, + "learning_rate": 9.344753289660592e-07, + "loss": 1.8896, + "step": 3685 + }, + { + "epoch": 0.2486439136147704, + "grad_norm": 6.832030622937063, + "learning_rate": 9.34184003874766e-07, + "loss": 1.9308, + "step": 3690 + }, + { + "epoch": 0.24898082948687714, + "grad_norm": 6.42103736138985, + "learning_rate": 9.338920782175269e-07, + "loss": 1.9103, + "step": 3695 + }, + { + "epoch": 0.24931774535898385, + "grad_norm": 6.567901996041012, + "learning_rate": 9.335995523981355e-07, + "loss": 1.9014, + "step": 3700 + }, + { + "epoch": 0.2496546612310906, + "grad_norm": 6.527187605310147, + "learning_rate": 9.333064268212153e-07, + "loss": 1.8274, + "step": 3705 + }, + { + "epoch": 0.24999157710319733, + "grad_norm": 6.482190165711818, + "learning_rate": 9.330127018922193e-07, + "loss": 1.902, + "step": 3710 + }, + { + "epoch": 0.2503284929753041, + "grad_norm": 5.668326143123714, + "learning_rate": 9.327183780174296e-07, + "loss": 1.8553, + "step": 3715 + }, + { + "epoch": 0.2506654088474108, + "grad_norm": 6.277307710196522, + "learning_rate": 9.324234556039567e-07, + "loss": 1.869, + "step": 3720 + }, + { + "epoch": 0.25100232471951756, + "grad_norm": 6.070880841042985, + "learning_rate": 9.321279350597393e-07, + "loss": 1.8632, + "step": 3725 + }, + { + "epoch": 0.25133924059162427, + "grad_norm": 5.564591682369898, + "learning_rate": 9.31831816793543e-07, + "loss": 1.8442, + "step": 3730 + }, + { + "epoch": 0.251676156463731, + "grad_norm": 5.911710322350655, + "learning_rate": 9.315351012149605e-07, + "loss": 1.8729, + "step": 3735 + }, + { + "epoch": 0.25201307233583775, + "grad_norm": 5.69969301095379, + "learning_rate": 9.312377887344105e-07, + "loss": 1.8273, + "step": 3740 + }, + { + "epoch": 0.25234998820794446, + "grad_norm": 5.758190049455089, + "learning_rate": 9.309398797631374e-07, + "loss": 1.8973, + "step": 3745 + }, + { + "epoch": 0.25268690408005123, + "grad_norm": 5.829441665364131, + "learning_rate": 9.306413747132108e-07, + "loss": 1.836, + "step": 3750 + }, + { + "epoch": 0.25302381995215795, + "grad_norm": 7.494080400403322, + "learning_rate": 9.303422739975246e-07, + "loss": 1.9453, + "step": 3755 + }, + { + "epoch": 0.25336073582426466, + "grad_norm": 6.367300908264392, + "learning_rate": 9.300425780297968e-07, + "loss": 1.8818, + "step": 3760 + }, + { + "epoch": 0.25369765169637143, + "grad_norm": 6.432621170142442, + "learning_rate": 9.297422872245686e-07, + "loss": 1.8613, + "step": 3765 + }, + { + "epoch": 0.25403456756847814, + "grad_norm": 6.108541333471499, + "learning_rate": 9.294414019972043e-07, + "loss": 1.8835, + "step": 3770 + }, + { + "epoch": 0.2543714834405849, + "grad_norm": 6.7536525309650335, + "learning_rate": 9.291399227638898e-07, + "loss": 1.8982, + "step": 3775 + }, + { + "epoch": 0.2547083993126916, + "grad_norm": 5.872721539921024, + "learning_rate": 9.288378499416332e-07, + "loss": 1.8544, + "step": 3780 + }, + { + "epoch": 0.25504531518479834, + "grad_norm": 6.476884234266525, + "learning_rate": 9.285351839482634e-07, + "loss": 1.8855, + "step": 3785 + }, + { + "epoch": 0.2553822310569051, + "grad_norm": 6.189916648502145, + "learning_rate": 9.2823192520243e-07, + "loss": 1.8873, + "step": 3790 + }, + { + "epoch": 0.2557191469290118, + "grad_norm": 5.983897060069876, + "learning_rate": 9.27928074123602e-07, + "loss": 1.8622, + "step": 3795 + }, + { + "epoch": 0.25605606280111853, + "grad_norm": 6.0808815520588615, + "learning_rate": 9.276236311320684e-07, + "loss": 1.8932, + "step": 3800 + }, + { + "epoch": 0.2563929786732253, + "grad_norm": 5.946657227019734, + "learning_rate": 9.273185966489365e-07, + "loss": 1.9228, + "step": 3805 + }, + { + "epoch": 0.256729894545332, + "grad_norm": 5.973295448688016, + "learning_rate": 9.270129710961318e-07, + "loss": 1.8135, + "step": 3810 + }, + { + "epoch": 0.2570668104174388, + "grad_norm": 5.556850084434461, + "learning_rate": 9.267067548963974e-07, + "loss": 1.8453, + "step": 3815 + }, + { + "epoch": 0.2574037262895455, + "grad_norm": 6.380485974028437, + "learning_rate": 9.263999484732934e-07, + "loss": 1.9156, + "step": 3820 + }, + { + "epoch": 0.2577406421616522, + "grad_norm": 6.097594854335576, + "learning_rate": 9.260925522511962e-07, + "loss": 1.8943, + "step": 3825 + }, + { + "epoch": 0.258077558033759, + "grad_norm": 5.871255691492415, + "learning_rate": 9.257845666552984e-07, + "loss": 1.9059, + "step": 3830 + }, + { + "epoch": 0.2584144739058657, + "grad_norm": 6.011724125372125, + "learning_rate": 9.254759921116073e-07, + "loss": 1.8983, + "step": 3835 + }, + { + "epoch": 0.25875138977797246, + "grad_norm": 6.4163291191100615, + "learning_rate": 9.251668290469452e-07, + "loss": 1.8849, + "step": 3840 + }, + { + "epoch": 0.2590883056500792, + "grad_norm": 5.812907082905663, + "learning_rate": 9.248570778889484e-07, + "loss": 1.8675, + "step": 3845 + }, + { + "epoch": 0.2594252215221859, + "grad_norm": 6.324746139412556, + "learning_rate": 9.245467390660664e-07, + "loss": 1.8303, + "step": 3850 + }, + { + "epoch": 0.25976213739429266, + "grad_norm": 6.0773552505761606, + "learning_rate": 9.242358130075618e-07, + "loss": 1.9011, + "step": 3855 + }, + { + "epoch": 0.26009905326639937, + "grad_norm": 5.809493160293788, + "learning_rate": 9.239243001435093e-07, + "loss": 1.8332, + "step": 3860 + }, + { + "epoch": 0.26043596913850614, + "grad_norm": 5.92861603094248, + "learning_rate": 9.236122009047957e-07, + "loss": 1.8137, + "step": 3865 + }, + { + "epoch": 0.26077288501061285, + "grad_norm": 6.5347777348808735, + "learning_rate": 9.232995157231182e-07, + "loss": 1.8711, + "step": 3870 + }, + { + "epoch": 0.26110980088271957, + "grad_norm": 5.818275626030584, + "learning_rate": 9.229862450309851e-07, + "loss": 1.8525, + "step": 3875 + }, + { + "epoch": 0.26144671675482634, + "grad_norm": 6.042244920418748, + "learning_rate": 9.226723892617141e-07, + "loss": 1.8957, + "step": 3880 + }, + { + "epoch": 0.26178363262693305, + "grad_norm": 5.006543607090774, + "learning_rate": 9.223579488494327e-07, + "loss": 1.8471, + "step": 3885 + }, + { + "epoch": 0.26212054849903976, + "grad_norm": 5.959886320776497, + "learning_rate": 9.220429242290763e-07, + "loss": 1.8556, + "step": 3890 + }, + { + "epoch": 0.26245746437114653, + "grad_norm": 6.192797328450058, + "learning_rate": 9.217273158363894e-07, + "loss": 1.8184, + "step": 3895 + }, + { + "epoch": 0.26279438024325324, + "grad_norm": 5.392472106463434, + "learning_rate": 9.214111241079232e-07, + "loss": 1.9066, + "step": 3900 + }, + { + "epoch": 0.26313129611536, + "grad_norm": 5.644651513241768, + "learning_rate": 9.21094349481036e-07, + "loss": 1.9127, + "step": 3905 + }, + { + "epoch": 0.2634682119874667, + "grad_norm": 6.226679672454185, + "learning_rate": 9.207769923938924e-07, + "loss": 1.8451, + "step": 3910 + }, + { + "epoch": 0.26380512785957344, + "grad_norm": 5.695454556727201, + "learning_rate": 9.204590532854627e-07, + "loss": 1.7827, + "step": 3915 + }, + { + "epoch": 0.2641420437316802, + "grad_norm": 5.98442814459207, + "learning_rate": 9.20140532595522e-07, + "loss": 1.9433, + "step": 3920 + }, + { + "epoch": 0.2644789596037869, + "grad_norm": 5.8671791162801155, + "learning_rate": 9.198214307646504e-07, + "loss": 1.8417, + "step": 3925 + }, + { + "epoch": 0.2648158754758937, + "grad_norm": 5.870417252327208, + "learning_rate": 9.195017482342313e-07, + "loss": 1.9082, + "step": 3930 + }, + { + "epoch": 0.2651527913480004, + "grad_norm": 6.4741759723264956, + "learning_rate": 9.191814854464514e-07, + "loss": 1.8705, + "step": 3935 + }, + { + "epoch": 0.2654897072201071, + "grad_norm": 5.862959275378989, + "learning_rate": 9.188606428443002e-07, + "loss": 1.8633, + "step": 3940 + }, + { + "epoch": 0.2658266230922139, + "grad_norm": 5.7751274179245, + "learning_rate": 9.185392208715692e-07, + "loss": 1.7832, + "step": 3945 + }, + { + "epoch": 0.2661635389643206, + "grad_norm": 6.172667926664809, + "learning_rate": 9.182172199728513e-07, + "loss": 1.8827, + "step": 3950 + }, + { + "epoch": 0.26650045483642737, + "grad_norm": 6.481705605622134, + "learning_rate": 9.178946405935398e-07, + "loss": 1.8974, + "step": 3955 + }, + { + "epoch": 0.2668373707085341, + "grad_norm": 6.208374646399624, + "learning_rate": 9.175714831798287e-07, + "loss": 1.8953, + "step": 3960 + }, + { + "epoch": 0.2671742865806408, + "grad_norm": 6.391722651555033, + "learning_rate": 9.172477481787113e-07, + "loss": 1.8014, + "step": 3965 + }, + { + "epoch": 0.26751120245274757, + "grad_norm": 6.037855150211001, + "learning_rate": 9.169234360379796e-07, + "loss": 1.8232, + "step": 3970 + }, + { + "epoch": 0.2678481183248543, + "grad_norm": 6.695825556952006, + "learning_rate": 9.165985472062244e-07, + "loss": 1.8885, + "step": 3975 + }, + { + "epoch": 0.268185034196961, + "grad_norm": 5.964301600300989, + "learning_rate": 9.162730821328337e-07, + "loss": 1.8774, + "step": 3980 + }, + { + "epoch": 0.26852195006906776, + "grad_norm": 6.396623330843612, + "learning_rate": 9.159470412679928e-07, + "loss": 1.9039, + "step": 3985 + }, + { + "epoch": 0.2688588659411745, + "grad_norm": 6.719419412679612, + "learning_rate": 9.156204250626836e-07, + "loss": 1.8773, + "step": 3990 + }, + { + "epoch": 0.26919578181328124, + "grad_norm": 6.879794042484075, + "learning_rate": 9.152932339686833e-07, + "loss": 1.9109, + "step": 3995 + }, + { + "epoch": 0.26953269768538796, + "grad_norm": 5.975921987217448, + "learning_rate": 9.149654684385647e-07, + "loss": 1.8618, + "step": 4000 + }, + { + "epoch": 0.26986961355749467, + "grad_norm": 6.629166390382459, + "learning_rate": 9.146371289256952e-07, + "loss": 1.798, + "step": 4005 + }, + { + "epoch": 0.27020652942960144, + "grad_norm": 5.949398882914333, + "learning_rate": 9.143082158842359e-07, + "loss": 1.9286, + "step": 4010 + }, + { + "epoch": 0.27054344530170815, + "grad_norm": 5.810972748863941, + "learning_rate": 9.139787297691413e-07, + "loss": 1.906, + "step": 4015 + }, + { + "epoch": 0.2708803611738149, + "grad_norm": 6.2120367421089115, + "learning_rate": 9.136486710361586e-07, + "loss": 1.9277, + "step": 4020 + }, + { + "epoch": 0.27121727704592163, + "grad_norm": 5.823041275387848, + "learning_rate": 9.13318040141827e-07, + "loss": 1.8543, + "step": 4025 + }, + { + "epoch": 0.27155419291802835, + "grad_norm": 6.214201315153956, + "learning_rate": 9.129868375434774e-07, + "loss": 1.8462, + "step": 4030 + }, + { + "epoch": 0.2718911087901351, + "grad_norm": 6.719918600313508, + "learning_rate": 9.12655063699231e-07, + "loss": 1.8796, + "step": 4035 + }, + { + "epoch": 0.27222802466224183, + "grad_norm": 6.3796230972963865, + "learning_rate": 9.123227190679994e-07, + "loss": 1.803, + "step": 4040 + }, + { + "epoch": 0.2725649405343486, + "grad_norm": 6.325649258604688, + "learning_rate": 9.119898041094838e-07, + "loss": 1.9, + "step": 4045 + }, + { + "epoch": 0.2729018564064553, + "grad_norm": 6.165527100110748, + "learning_rate": 9.116563192841741e-07, + "loss": 1.8718, + "step": 4050 + }, + { + "epoch": 0.273238772278562, + "grad_norm": 6.505461302130346, + "learning_rate": 9.113222650533486e-07, + "loss": 1.9228, + "step": 4055 + }, + { + "epoch": 0.2735756881506688, + "grad_norm": 6.263801274705381, + "learning_rate": 9.109876418790731e-07, + "loss": 1.8632, + "step": 4060 + }, + { + "epoch": 0.2739126040227755, + "grad_norm": 6.661484279149622, + "learning_rate": 9.106524502242004e-07, + "loss": 1.8669, + "step": 4065 + }, + { + "epoch": 0.2742495198948822, + "grad_norm": 5.738641723072569, + "learning_rate": 9.103166905523699e-07, + "loss": 1.8212, + "step": 4070 + }, + { + "epoch": 0.274586435766989, + "grad_norm": 6.551069538879794, + "learning_rate": 9.099803633280059e-07, + "loss": 1.9684, + "step": 4075 + }, + { + "epoch": 0.2749233516390957, + "grad_norm": 6.532119039589357, + "learning_rate": 9.096434690163184e-07, + "loss": 1.8796, + "step": 4080 + }, + { + "epoch": 0.27526026751120247, + "grad_norm": 5.881963986981689, + "learning_rate": 9.093060080833019e-07, + "loss": 1.8195, + "step": 4085 + }, + { + "epoch": 0.2755971833833092, + "grad_norm": 6.412172720197176, + "learning_rate": 9.089679809957343e-07, + "loss": 1.8372, + "step": 4090 + }, + { + "epoch": 0.2759340992554159, + "grad_norm": 7.471271072846122, + "learning_rate": 9.086293882211768e-07, + "loss": 1.9177, + "step": 4095 + }, + { + "epoch": 0.27627101512752267, + "grad_norm": 6.198988270277827, + "learning_rate": 9.082902302279726e-07, + "loss": 1.8632, + "step": 4100 + }, + { + "epoch": 0.2766079309996294, + "grad_norm": 5.767940298563196, + "learning_rate": 9.079505074852476e-07, + "loss": 1.8877, + "step": 4105 + }, + { + "epoch": 0.27694484687173615, + "grad_norm": 6.230279674826198, + "learning_rate": 9.076102204629082e-07, + "loss": 1.8431, + "step": 4110 + }, + { + "epoch": 0.27728176274384286, + "grad_norm": 5.949567529985799, + "learning_rate": 9.072693696316411e-07, + "loss": 1.8633, + "step": 4115 + }, + { + "epoch": 0.2776186786159496, + "grad_norm": 5.928871922518486, + "learning_rate": 9.069279554629137e-07, + "loss": 1.8668, + "step": 4120 + }, + { + "epoch": 0.27795559448805635, + "grad_norm": 6.081834973234626, + "learning_rate": 9.06585978428972e-07, + "loss": 1.8543, + "step": 4125 + }, + { + "epoch": 0.27829251036016306, + "grad_norm": 6.142536394883528, + "learning_rate": 9.062434390028407e-07, + "loss": 1.8877, + "step": 4130 + }, + { + "epoch": 0.27862942623226983, + "grad_norm": 5.741508220751473, + "learning_rate": 9.059003376583223e-07, + "loss": 1.877, + "step": 4135 + }, + { + "epoch": 0.27896634210437654, + "grad_norm": 5.705602714030965, + "learning_rate": 9.055566748699968e-07, + "loss": 1.9019, + "step": 4140 + }, + { + "epoch": 0.27930325797648325, + "grad_norm": 5.940713375585646, + "learning_rate": 9.052124511132204e-07, + "loss": 1.8499, + "step": 4145 + }, + { + "epoch": 0.27964017384859, + "grad_norm": 5.985593133684585, + "learning_rate": 9.04867666864126e-07, + "loss": 1.8314, + "step": 4150 + }, + { + "epoch": 0.27997708972069674, + "grad_norm": 7.084259902044295, + "learning_rate": 9.045223225996207e-07, + "loss": 1.8601, + "step": 4155 + }, + { + "epoch": 0.28031400559280345, + "grad_norm": 6.259469749054458, + "learning_rate": 9.041764187973871e-07, + "loss": 1.8467, + "step": 4160 + }, + { + "epoch": 0.2806509214649102, + "grad_norm": 6.211124902310236, + "learning_rate": 9.038299559358815e-07, + "loss": 1.8573, + "step": 4165 + }, + { + "epoch": 0.28098783733701693, + "grad_norm": 6.0058531329183875, + "learning_rate": 9.034829344943331e-07, + "loss": 1.8623, + "step": 4170 + }, + { + "epoch": 0.2813247532091237, + "grad_norm": 6.136862011053358, + "learning_rate": 9.031353549527444e-07, + "loss": 1.8878, + "step": 4175 + }, + { + "epoch": 0.2816616690812304, + "grad_norm": 6.167062756427111, + "learning_rate": 9.027872177918894e-07, + "loss": 1.8131, + "step": 4180 + }, + { + "epoch": 0.28199858495333713, + "grad_norm": 5.533797632844438, + "learning_rate": 9.024385234933134e-07, + "loss": 1.9209, + "step": 4185 + }, + { + "epoch": 0.2823355008254439, + "grad_norm": 5.695979000250038, + "learning_rate": 9.020892725393326e-07, + "loss": 1.807, + "step": 4190 + }, + { + "epoch": 0.2826724166975506, + "grad_norm": 6.377546908781681, + "learning_rate": 9.017394654130332e-07, + "loss": 1.8577, + "step": 4195 + }, + { + "epoch": 0.2830093325696574, + "grad_norm": 5.756307597482796, + "learning_rate": 9.013891025982703e-07, + "loss": 1.8913, + "step": 4200 + }, + { + "epoch": 0.2833462484417641, + "grad_norm": 6.166064092582664, + "learning_rate": 9.010381845796677e-07, + "loss": 1.823, + "step": 4205 + }, + { + "epoch": 0.2836831643138708, + "grad_norm": 6.488806989342812, + "learning_rate": 9.006867118426178e-07, + "loss": 1.8323, + "step": 4210 + }, + { + "epoch": 0.2840200801859776, + "grad_norm": 6.27973058379338, + "learning_rate": 9.003346848732793e-07, + "loss": 1.8536, + "step": 4215 + }, + { + "epoch": 0.2843569960580843, + "grad_norm": 5.344380427344846, + "learning_rate": 8.999821041585787e-07, + "loss": 1.8111, + "step": 4220 + }, + { + "epoch": 0.28469391193019106, + "grad_norm": 6.075842058898948, + "learning_rate": 8.996289701862072e-07, + "loss": 1.9303, + "step": 4225 + }, + { + "epoch": 0.28503082780229777, + "grad_norm": 6.265878132390271, + "learning_rate": 8.99275283444622e-07, + "loss": 1.8807, + "step": 4230 + }, + { + "epoch": 0.2853677436744045, + "grad_norm": 5.7219422084040055, + "learning_rate": 8.989210444230449e-07, + "loss": 1.8753, + "step": 4235 + }, + { + "epoch": 0.28570465954651125, + "grad_norm": 6.215050142707896, + "learning_rate": 8.985662536114612e-07, + "loss": 1.884, + "step": 4240 + }, + { + "epoch": 0.28604157541861797, + "grad_norm": 6.110998708733845, + "learning_rate": 8.9821091150062e-07, + "loss": 1.8409, + "step": 4245 + }, + { + "epoch": 0.2863784912907247, + "grad_norm": 6.257750293856616, + "learning_rate": 8.978550185820323e-07, + "loss": 1.8368, + "step": 4250 + }, + { + "epoch": 0.28671540716283145, + "grad_norm": 6.054417802124323, + "learning_rate": 8.974985753479718e-07, + "loss": 1.8733, + "step": 4255 + }, + { + "epoch": 0.28705232303493816, + "grad_norm": 6.043179751280904, + "learning_rate": 8.971415822914726e-07, + "loss": 1.8588, + "step": 4260 + }, + { + "epoch": 0.28738923890704493, + "grad_norm": 6.14176472404905, + "learning_rate": 8.967840399063298e-07, + "loss": 1.8271, + "step": 4265 + }, + { + "epoch": 0.28772615477915164, + "grad_norm": 6.174678032249467, + "learning_rate": 8.964259486870982e-07, + "loss": 1.8321, + "step": 4270 + }, + { + "epoch": 0.28806307065125836, + "grad_norm": 5.8832790823420975, + "learning_rate": 8.960673091290916e-07, + "loss": 1.8336, + "step": 4275 + }, + { + "epoch": 0.2883999865233651, + "grad_norm": 6.488845467073408, + "learning_rate": 8.957081217283825e-07, + "loss": 1.8753, + "step": 4280 + }, + { + "epoch": 0.28873690239547184, + "grad_norm": 5.749007187923349, + "learning_rate": 8.953483869818013e-07, + "loss": 1.8418, + "step": 4285 + }, + { + "epoch": 0.2890738182675786, + "grad_norm": 6.049504947740589, + "learning_rate": 8.949881053869348e-07, + "loss": 1.8525, + "step": 4290 + }, + { + "epoch": 0.2894107341396853, + "grad_norm": 6.396690173802817, + "learning_rate": 8.946272774421271e-07, + "loss": 1.8286, + "step": 4295 + }, + { + "epoch": 0.28974765001179204, + "grad_norm": 6.108405259192714, + "learning_rate": 8.942659036464775e-07, + "loss": 1.859, + "step": 4300 + }, + { + "epoch": 0.2900845658838988, + "grad_norm": 5.98212203777026, + "learning_rate": 8.939039844998403e-07, + "loss": 1.869, + "step": 4305 + }, + { + "epoch": 0.2904214817560055, + "grad_norm": 5.941860278201532, + "learning_rate": 8.935415205028243e-07, + "loss": 1.8408, + "step": 4310 + }, + { + "epoch": 0.2907583976281123, + "grad_norm": 5.72615071195908, + "learning_rate": 8.931785121567921e-07, + "loss": 1.8055, + "step": 4315 + }, + { + "epoch": 0.291095313500219, + "grad_norm": 6.117559850305973, + "learning_rate": 8.928149599638588e-07, + "loss": 1.8843, + "step": 4320 + }, + { + "epoch": 0.2914322293723257, + "grad_norm": 6.450369657828759, + "learning_rate": 8.924508644268921e-07, + "loss": 1.9653, + "step": 4325 + }, + { + "epoch": 0.2917691452444325, + "grad_norm": 5.856381777788301, + "learning_rate": 8.920862260495111e-07, + "loss": 1.8706, + "step": 4330 + }, + { + "epoch": 0.2921060611165392, + "grad_norm": 6.354337305203465, + "learning_rate": 8.917210453360859e-07, + "loss": 1.8893, + "step": 4335 + }, + { + "epoch": 0.2924429769886459, + "grad_norm": 6.906110311477241, + "learning_rate": 8.913553227917365e-07, + "loss": 1.806, + "step": 4340 + }, + { + "epoch": 0.2927798928607527, + "grad_norm": 6.179040612378133, + "learning_rate": 8.909890589223329e-07, + "loss": 1.86, + "step": 4345 + }, + { + "epoch": 0.2931168087328594, + "grad_norm": 5.978516176145858, + "learning_rate": 8.906222542344932e-07, + "loss": 1.8849, + "step": 4350 + }, + { + "epoch": 0.29345372460496616, + "grad_norm": 6.002208675843779, + "learning_rate": 8.902549092355839e-07, + "loss": 1.8798, + "step": 4355 + }, + { + "epoch": 0.2937906404770729, + "grad_norm": 6.2186044385766195, + "learning_rate": 8.898870244337189e-07, + "loss": 1.9232, + "step": 4360 + }, + { + "epoch": 0.2941275563491796, + "grad_norm": 6.153726869453081, + "learning_rate": 8.895186003377586e-07, + "loss": 1.8648, + "step": 4365 + }, + { + "epoch": 0.29446447222128636, + "grad_norm": 6.919968948046176, + "learning_rate": 8.891496374573095e-07, + "loss": 1.9309, + "step": 4370 + }, + { + "epoch": 0.29480138809339307, + "grad_norm": 6.058343737353778, + "learning_rate": 8.887801363027233e-07, + "loss": 1.8841, + "step": 4375 + }, + { + "epoch": 0.29513830396549984, + "grad_norm": 5.937731974461376, + "learning_rate": 8.884100973850962e-07, + "loss": 1.9063, + "step": 4380 + }, + { + "epoch": 0.29547521983760655, + "grad_norm": 5.801901159807343, + "learning_rate": 8.880395212162684e-07, + "loss": 1.8964, + "step": 4385 + }, + { + "epoch": 0.29581213570971326, + "grad_norm": 6.297331913420712, + "learning_rate": 8.87668408308823e-07, + "loss": 1.7967, + "step": 4390 + }, + { + "epoch": 0.29614905158182003, + "grad_norm": 5.897350111858126, + "learning_rate": 8.872967591760856e-07, + "loss": 1.8394, + "step": 4395 + }, + { + "epoch": 0.29648596745392675, + "grad_norm": 5.817511472756013, + "learning_rate": 8.869245743321234e-07, + "loss": 1.8911, + "step": 4400 + }, + { + "epoch": 0.2968228833260335, + "grad_norm": 6.1323690508053215, + "learning_rate": 8.865518542917452e-07, + "loss": 1.782, + "step": 4405 + }, + { + "epoch": 0.29715979919814023, + "grad_norm": 5.936037133542599, + "learning_rate": 8.861785995704991e-07, + "loss": 1.8871, + "step": 4410 + }, + { + "epoch": 0.29749671507024694, + "grad_norm": 5.782678112860589, + "learning_rate": 8.858048106846735e-07, + "loss": 1.8751, + "step": 4415 + }, + { + "epoch": 0.2978336309423537, + "grad_norm": 6.143794616391879, + "learning_rate": 8.854304881512955e-07, + "loss": 1.8205, + "step": 4420 + }, + { + "epoch": 0.2981705468144604, + "grad_norm": 6.267922935330008, + "learning_rate": 8.850556324881302e-07, + "loss": 1.9108, + "step": 4425 + }, + { + "epoch": 0.29850746268656714, + "grad_norm": 5.700982678029616, + "learning_rate": 8.846802442136804e-07, + "loss": 1.8323, + "step": 4430 + }, + { + "epoch": 0.2988443785586739, + "grad_norm": 5.875620333234486, + "learning_rate": 8.843043238471853e-07, + "loss": 1.8779, + "step": 4435 + }, + { + "epoch": 0.2991812944307806, + "grad_norm": 6.514010796547643, + "learning_rate": 8.839278719086201e-07, + "loss": 1.8766, + "step": 4440 + }, + { + "epoch": 0.2995182103028874, + "grad_norm": 6.078146151760929, + "learning_rate": 8.835508889186956e-07, + "loss": 1.8954, + "step": 4445 + }, + { + "epoch": 0.2998551261749941, + "grad_norm": 6.811365534277628, + "learning_rate": 8.83173375398857e-07, + "loss": 1.8825, + "step": 4450 + }, + { + "epoch": 0.3001920420471008, + "grad_norm": 5.954896691271629, + "learning_rate": 8.827953318712831e-07, + "loss": 1.8576, + "step": 4455 + }, + { + "epoch": 0.3005289579192076, + "grad_norm": 5.68557129820593, + "learning_rate": 8.824167588588861e-07, + "loss": 1.8302, + "step": 4460 + }, + { + "epoch": 0.3008658737913143, + "grad_norm": 6.256891159580008, + "learning_rate": 8.820376568853105e-07, + "loss": 1.8263, + "step": 4465 + }, + { + "epoch": 0.30120278966342107, + "grad_norm": 5.930780965618538, + "learning_rate": 8.816580264749325e-07, + "loss": 1.8577, + "step": 4470 + }, + { + "epoch": 0.3015397055355278, + "grad_norm": 5.597980182580963, + "learning_rate": 8.81277868152859e-07, + "loss": 1.874, + "step": 4475 + }, + { + "epoch": 0.3018766214076345, + "grad_norm": 6.170497154326832, + "learning_rate": 8.808971824449274e-07, + "loss": 1.8562, + "step": 4480 + }, + { + "epoch": 0.30221353727974126, + "grad_norm": 5.543906347986211, + "learning_rate": 8.805159698777045e-07, + "loss": 1.8477, + "step": 4485 + }, + { + "epoch": 0.302550453151848, + "grad_norm": 5.950757055714713, + "learning_rate": 8.801342309784858e-07, + "loss": 1.8363, + "step": 4490 + }, + { + "epoch": 0.30288736902395474, + "grad_norm": 5.555554802786965, + "learning_rate": 8.79751966275295e-07, + "loss": 1.7955, + "step": 4495 + }, + { + "epoch": 0.30322428489606146, + "grad_norm": 6.0842683476071695, + "learning_rate": 8.793691762968827e-07, + "loss": 1.838, + "step": 4500 + }, + { + "epoch": 0.30356120076816817, + "grad_norm": 6.3854828636373755, + "learning_rate": 8.789858615727264e-07, + "loss": 1.8526, + "step": 4505 + }, + { + "epoch": 0.30389811664027494, + "grad_norm": 5.849310768859681, + "learning_rate": 8.786020226330295e-07, + "loss": 1.8327, + "step": 4510 + }, + { + "epoch": 0.30423503251238165, + "grad_norm": 5.90106872765922, + "learning_rate": 8.782176600087203e-07, + "loss": 1.7905, + "step": 4515 + }, + { + "epoch": 0.30457194838448837, + "grad_norm": 5.901668679720049, + "learning_rate": 8.778327742314513e-07, + "loss": 1.8567, + "step": 4520 + }, + { + "epoch": 0.30490886425659514, + "grad_norm": 5.8797100239213655, + "learning_rate": 8.77447365833599e-07, + "loss": 1.8321, + "step": 4525 + }, + { + "epoch": 0.30524578012870185, + "grad_norm": 6.2990354182925135, + "learning_rate": 8.770614353482628e-07, + "loss": 1.8714, + "step": 4530 + }, + { + "epoch": 0.3055826960008086, + "grad_norm": 5.873265394004258, + "learning_rate": 8.766749833092638e-07, + "loss": 1.9256, + "step": 4535 + }, + { + "epoch": 0.30591961187291533, + "grad_norm": 5.880033630724263, + "learning_rate": 8.76288010251145e-07, + "loss": 1.8538, + "step": 4540 + }, + { + "epoch": 0.30625652774502204, + "grad_norm": 6.21041901452895, + "learning_rate": 8.759005167091697e-07, + "loss": 1.9027, + "step": 4545 + }, + { + "epoch": 0.3065934436171288, + "grad_norm": 6.363251755485691, + "learning_rate": 8.755125032193214e-07, + "loss": 1.788, + "step": 4550 + }, + { + "epoch": 0.3069303594892355, + "grad_norm": 6.23031482525298, + "learning_rate": 8.751239703183026e-07, + "loss": 1.8438, + "step": 4555 + }, + { + "epoch": 0.3072672753613423, + "grad_norm": 6.133022983823034, + "learning_rate": 8.747349185435348e-07, + "loss": 1.8369, + "step": 4560 + }, + { + "epoch": 0.307604191233449, + "grad_norm": 6.285822907751822, + "learning_rate": 8.743453484331562e-07, + "loss": 1.9024, + "step": 4565 + }, + { + "epoch": 0.3079411071055557, + "grad_norm": 5.835690715657944, + "learning_rate": 8.73955260526023e-07, + "loss": 1.8256, + "step": 4570 + }, + { + "epoch": 0.3082780229776625, + "grad_norm": 5.847371813030593, + "learning_rate": 8.735646553617069e-07, + "loss": 1.9136, + "step": 4575 + }, + { + "epoch": 0.3086149388497692, + "grad_norm": 6.043595537844784, + "learning_rate": 8.731735334804953e-07, + "loss": 1.8674, + "step": 4580 + }, + { + "epoch": 0.308951854721876, + "grad_norm": 5.956889810584438, + "learning_rate": 8.727818954233904e-07, + "loss": 1.8581, + "step": 4585 + }, + { + "epoch": 0.3092887705939827, + "grad_norm": 6.324696749321057, + "learning_rate": 8.723897417321084e-07, + "loss": 1.8447, + "step": 4590 + }, + { + "epoch": 0.3096256864660894, + "grad_norm": 6.276116542740643, + "learning_rate": 8.719970729490788e-07, + "loss": 1.873, + "step": 4595 + }, + { + "epoch": 0.30996260233819617, + "grad_norm": 6.411851214725662, + "learning_rate": 8.716038896174432e-07, + "loss": 1.8802, + "step": 4600 + }, + { + "epoch": 0.3102995182103029, + "grad_norm": 6.138820411276758, + "learning_rate": 8.712101922810551e-07, + "loss": 1.8773, + "step": 4605 + }, + { + "epoch": 0.3106364340824096, + "grad_norm": 5.872851484478736, + "learning_rate": 8.708159814844793e-07, + "loss": 1.8485, + "step": 4610 + }, + { + "epoch": 0.31097334995451636, + "grad_norm": 6.031467898907255, + "learning_rate": 8.704212577729905e-07, + "loss": 1.897, + "step": 4615 + }, + { + "epoch": 0.3113102658266231, + "grad_norm": 5.9956951123885345, + "learning_rate": 8.700260216925728e-07, + "loss": 1.8205, + "step": 4620 + }, + { + "epoch": 0.31164718169872985, + "grad_norm": 6.056845802633199, + "learning_rate": 8.696302737899192e-07, + "loss": 1.8795, + "step": 4625 + }, + { + "epoch": 0.31198409757083656, + "grad_norm": 6.221681582297336, + "learning_rate": 8.692340146124308e-07, + "loss": 1.8269, + "step": 4630 + }, + { + "epoch": 0.3123210134429433, + "grad_norm": 6.553618267443545, + "learning_rate": 8.688372447082153e-07, + "loss": 1.8764, + "step": 4635 + }, + { + "epoch": 0.31265792931505004, + "grad_norm": 6.213539651433188, + "learning_rate": 8.684399646260876e-07, + "loss": 1.8413, + "step": 4640 + }, + { + "epoch": 0.31299484518715676, + "grad_norm": 6.257873473467152, + "learning_rate": 8.680421749155677e-07, + "loss": 1.8354, + "step": 4645 + }, + { + "epoch": 0.3133317610592635, + "grad_norm": 6.354806527386589, + "learning_rate": 8.676438761268808e-07, + "loss": 1.9039, + "step": 4650 + }, + { + "epoch": 0.31366867693137024, + "grad_norm": 6.240255748070445, + "learning_rate": 8.672450688109563e-07, + "loss": 1.9127, + "step": 4655 + }, + { + "epoch": 0.31400559280347695, + "grad_norm": 6.274554512254039, + "learning_rate": 8.668457535194267e-07, + "loss": 1.9452, + "step": 4660 + }, + { + "epoch": 0.3143425086755837, + "grad_norm": 5.680213033367091, + "learning_rate": 8.664459308046274e-07, + "loss": 1.8823, + "step": 4665 + }, + { + "epoch": 0.31467942454769043, + "grad_norm": 6.026744567200448, + "learning_rate": 8.660456012195957e-07, + "loss": 1.8442, + "step": 4670 + }, + { + "epoch": 0.3150163404197972, + "grad_norm": 6.031466170099904, + "learning_rate": 8.656447653180699e-07, + "loss": 1.8205, + "step": 4675 + }, + { + "epoch": 0.3153532562919039, + "grad_norm": 6.34630037939122, + "learning_rate": 8.652434236544886e-07, + "loss": 1.8885, + "step": 4680 + }, + { + "epoch": 0.31569017216401063, + "grad_norm": 6.0108468232784, + "learning_rate": 8.648415767839899e-07, + "loss": 1.8477, + "step": 4685 + }, + { + "epoch": 0.3160270880361174, + "grad_norm": 6.056101862232991, + "learning_rate": 8.644392252624108e-07, + "loss": 1.8009, + "step": 4690 + }, + { + "epoch": 0.3163640039082241, + "grad_norm": 5.549149326299132, + "learning_rate": 8.640363696462869e-07, + "loss": 1.8893, + "step": 4695 + }, + { + "epoch": 0.3167009197803308, + "grad_norm": 5.748085710916815, + "learning_rate": 8.636330104928499e-07, + "loss": 1.8659, + "step": 4700 + }, + { + "epoch": 0.3170378356524376, + "grad_norm": 6.615576133533302, + "learning_rate": 8.632291483600289e-07, + "loss": 1.8309, + "step": 4705 + }, + { + "epoch": 0.3173747515245443, + "grad_norm": 6.1506018149830055, + "learning_rate": 8.628247838064485e-07, + "loss": 1.8303, + "step": 4710 + }, + { + "epoch": 0.3177116673966511, + "grad_norm": 6.332160685216612, + "learning_rate": 8.624199173914279e-07, + "loss": 1.9227, + "step": 4715 + }, + { + "epoch": 0.3180485832687578, + "grad_norm": 6.259268021975856, + "learning_rate": 8.620145496749811e-07, + "loss": 1.8573, + "step": 4720 + }, + { + "epoch": 0.3183854991408645, + "grad_norm": 6.438165923545349, + "learning_rate": 8.616086812178151e-07, + "loss": 1.8553, + "step": 4725 + }, + { + "epoch": 0.31872241501297127, + "grad_norm": 5.938254916301874, + "learning_rate": 8.612023125813296e-07, + "loss": 1.8766, + "step": 4730 + }, + { + "epoch": 0.319059330885078, + "grad_norm": 5.6026652144423, + "learning_rate": 8.607954443276162e-07, + "loss": 1.9063, + "step": 4735 + }, + { + "epoch": 0.31939624675718475, + "grad_norm": 5.860335890704336, + "learning_rate": 8.603880770194574e-07, + "loss": 1.8725, + "step": 4740 + }, + { + "epoch": 0.31973316262929147, + "grad_norm": 5.920443790572611, + "learning_rate": 8.59980211220326e-07, + "loss": 1.8371, + "step": 4745 + }, + { + "epoch": 0.3200700785013982, + "grad_norm": 6.175976373732354, + "learning_rate": 8.595718474943849e-07, + "loss": 1.8342, + "step": 4750 + }, + { + "epoch": 0.32040699437350495, + "grad_norm": 6.303028162479854, + "learning_rate": 8.591629864064851e-07, + "loss": 1.8751, + "step": 4755 + }, + { + "epoch": 0.32074391024561166, + "grad_norm": 6.430617243948466, + "learning_rate": 8.587536285221655e-07, + "loss": 1.9291, + "step": 4760 + }, + { + "epoch": 0.32108082611771843, + "grad_norm": 6.1534051447562055, + "learning_rate": 8.583437744076527e-07, + "loss": 1.8666, + "step": 4765 + }, + { + "epoch": 0.32141774198982515, + "grad_norm": 6.535863631168725, + "learning_rate": 8.579334246298592e-07, + "loss": 1.8779, + "step": 4770 + }, + { + "epoch": 0.32175465786193186, + "grad_norm": 5.71107753763816, + "learning_rate": 8.575225797563834e-07, + "loss": 1.911, + "step": 4775 + }, + { + "epoch": 0.3220915737340386, + "grad_norm": 5.886190655326756, + "learning_rate": 8.571112403555083e-07, + "loss": 1.851, + "step": 4780 + }, + { + "epoch": 0.32242848960614534, + "grad_norm": 6.095741702151825, + "learning_rate": 8.566994069962012e-07, + "loss": 1.8704, + "step": 4785 + }, + { + "epoch": 0.32276540547825205, + "grad_norm": 6.02265285150199, + "learning_rate": 8.562870802481126e-07, + "loss": 1.8763, + "step": 4790 + }, + { + "epoch": 0.3231023213503588, + "grad_norm": 6.5312090704657795, + "learning_rate": 8.55874260681575e-07, + "loss": 1.8484, + "step": 4795 + }, + { + "epoch": 0.32343923722246554, + "grad_norm": 6.285132315448252, + "learning_rate": 8.554609488676032e-07, + "loss": 1.8557, + "step": 4800 + }, + { + "epoch": 0.3237761530945723, + "grad_norm": 6.0685785815626705, + "learning_rate": 8.550471453778925e-07, + "loss": 1.8416, + "step": 4805 + }, + { + "epoch": 0.324113068966679, + "grad_norm": 6.044458918895983, + "learning_rate": 8.546328507848184e-07, + "loss": 1.8412, + "step": 4810 + }, + { + "epoch": 0.32444998483878573, + "grad_norm": 5.8373873569240855, + "learning_rate": 8.542180656614358e-07, + "loss": 1.7964, + "step": 4815 + }, + { + "epoch": 0.3247869007108925, + "grad_norm": 6.1204039997103035, + "learning_rate": 8.538027905814778e-07, + "loss": 1.8239, + "step": 4820 + }, + { + "epoch": 0.3251238165829992, + "grad_norm": 5.906274807482711, + "learning_rate": 8.533870261193556e-07, + "loss": 1.7694, + "step": 4825 + }, + { + "epoch": 0.325460732455106, + "grad_norm": 6.21208348038065, + "learning_rate": 8.529707728501571e-07, + "loss": 1.8148, + "step": 4830 + }, + { + "epoch": 0.3257976483272127, + "grad_norm": 6.014309137015161, + "learning_rate": 8.525540313496462e-07, + "loss": 1.836, + "step": 4835 + }, + { + "epoch": 0.3261345641993194, + "grad_norm": 6.0431816763200805, + "learning_rate": 8.521368021942623e-07, + "loss": 1.9126, + "step": 4840 + }, + { + "epoch": 0.3264714800714262, + "grad_norm": 6.376773643711095, + "learning_rate": 8.517190859611195e-07, + "loss": 1.806, + "step": 4845 + }, + { + "epoch": 0.3268083959435329, + "grad_norm": 6.863001260214309, + "learning_rate": 8.513008832280053e-07, + "loss": 1.9519, + "step": 4850 + }, + { + "epoch": 0.32714531181563966, + "grad_norm": 6.75341977852346, + "learning_rate": 8.508821945733802e-07, + "loss": 1.8704, + "step": 4855 + }, + { + "epoch": 0.3274822276877464, + "grad_norm": 6.397518156192358, + "learning_rate": 8.504630205763768e-07, + "loss": 1.8695, + "step": 4860 + }, + { + "epoch": 0.3278191435598531, + "grad_norm": 6.182422350969933, + "learning_rate": 8.500433618167992e-07, + "loss": 1.8985, + "step": 4865 + }, + { + "epoch": 0.32815605943195986, + "grad_norm": 6.1420508752244265, + "learning_rate": 8.496232188751222e-07, + "loss": 1.8443, + "step": 4870 + }, + { + "epoch": 0.32849297530406657, + "grad_norm": 5.590831241686126, + "learning_rate": 8.492025923324897e-07, + "loss": 1.8332, + "step": 4875 + }, + { + "epoch": 0.3288298911761733, + "grad_norm": 6.260067475048003, + "learning_rate": 8.487814827707152e-07, + "loss": 1.8677, + "step": 4880 + }, + { + "epoch": 0.32916680704828005, + "grad_norm": 6.460645221149728, + "learning_rate": 8.483598907722795e-07, + "loss": 1.8868, + "step": 4885 + }, + { + "epoch": 0.32950372292038677, + "grad_norm": 5.771598644199668, + "learning_rate": 8.479378169203317e-07, + "loss": 1.7935, + "step": 4890 + }, + { + "epoch": 0.32984063879249353, + "grad_norm": 6.205757641829904, + "learning_rate": 8.475152617986869e-07, + "loss": 1.8203, + "step": 4895 + }, + { + "epoch": 0.33017755466460025, + "grad_norm": 6.0547319795697, + "learning_rate": 8.470922259918254e-07, + "loss": 1.815, + "step": 4900 + }, + { + "epoch": 0.33051447053670696, + "grad_norm": 5.81444773035736, + "learning_rate": 8.466687100848935e-07, + "loss": 1.8813, + "step": 4905 + }, + { + "epoch": 0.33085138640881373, + "grad_norm": 6.352411344793242, + "learning_rate": 8.462447146637006e-07, + "loss": 1.8406, + "step": 4910 + }, + { + "epoch": 0.33118830228092044, + "grad_norm": 5.769658940167598, + "learning_rate": 8.458202403147199e-07, + "loss": 1.8585, + "step": 4915 + }, + { + "epoch": 0.3315252181530272, + "grad_norm": 6.334129261483883, + "learning_rate": 8.453952876250867e-07, + "loss": 1.8119, + "step": 4920 + }, + { + "epoch": 0.3318621340251339, + "grad_norm": 6.3371647076546145, + "learning_rate": 8.449698571825984e-07, + "loss": 1.8941, + "step": 4925 + }, + { + "epoch": 0.33219904989724064, + "grad_norm": 6.332880954598852, + "learning_rate": 8.445439495757127e-07, + "loss": 1.8444, + "step": 4930 + }, + { + "epoch": 0.3325359657693474, + "grad_norm": 6.026022147996478, + "learning_rate": 8.44117565393548e-07, + "loss": 1.8098, + "step": 4935 + }, + { + "epoch": 0.3328728816414541, + "grad_norm": 6.068938398940136, + "learning_rate": 8.436907052258808e-07, + "loss": 1.896, + "step": 4940 + }, + { + "epoch": 0.3332097975135609, + "grad_norm": 6.273354421417618, + "learning_rate": 8.432633696631473e-07, + "loss": 1.9326, + "step": 4945 + }, + { + "epoch": 0.3335467133856676, + "grad_norm": 5.900730766734711, + "learning_rate": 8.428355592964405e-07, + "loss": 1.8764, + "step": 4950 + }, + { + "epoch": 0.3338836292577743, + "grad_norm": 5.742195611900634, + "learning_rate": 8.424072747175102e-07, + "loss": 1.8853, + "step": 4955 + }, + { + "epoch": 0.3342205451298811, + "grad_norm": 5.65151284794071, + "learning_rate": 8.419785165187621e-07, + "loss": 1.762, + "step": 4960 + }, + { + "epoch": 0.3345574610019878, + "grad_norm": 6.1205570132491145, + "learning_rate": 8.415492852932573e-07, + "loss": 1.8033, + "step": 4965 + }, + { + "epoch": 0.3348943768740945, + "grad_norm": 6.316447919486459, + "learning_rate": 8.41119581634711e-07, + "loss": 1.9436, + "step": 4970 + }, + { + "epoch": 0.3352312927462013, + "grad_norm": 5.923608163568519, + "learning_rate": 8.406894061374918e-07, + "loss": 1.9223, + "step": 4975 + }, + { + "epoch": 0.335568208618308, + "grad_norm": 6.211087505657584, + "learning_rate": 8.402587593966213e-07, + "loss": 1.8231, + "step": 4980 + }, + { + "epoch": 0.33590512449041476, + "grad_norm": 5.960407764689357, + "learning_rate": 8.398276420077726e-07, + "loss": 1.8805, + "step": 4985 + }, + { + "epoch": 0.3362420403625215, + "grad_norm": 6.353999152530212, + "learning_rate": 8.393960545672698e-07, + "loss": 1.8367, + "step": 4990 + }, + { + "epoch": 0.3365789562346282, + "grad_norm": 6.102007260398823, + "learning_rate": 8.389639976720873e-07, + "loss": 1.8398, + "step": 4995 + }, + { + "epoch": 0.33691587210673496, + "grad_norm": 6.530971846468027, + "learning_rate": 8.385314719198487e-07, + "loss": 1.8589, + "step": 5000 + }, + { + "epoch": 0.3372527879788417, + "grad_norm": 5.855794124939447, + "learning_rate": 8.380984779088264e-07, + "loss": 1.8611, + "step": 5005 + }, + { + "epoch": 0.33758970385094844, + "grad_norm": 6.670146157816599, + "learning_rate": 8.376650162379404e-07, + "loss": 1.8852, + "step": 5010 + }, + { + "epoch": 0.33792661972305515, + "grad_norm": 5.686028080563687, + "learning_rate": 8.372310875067572e-07, + "loss": 1.8913, + "step": 5015 + }, + { + "epoch": 0.33826353559516187, + "grad_norm": 6.097937051428109, + "learning_rate": 8.367966923154899e-07, + "loss": 1.7633, + "step": 5020 + }, + { + "epoch": 0.33860045146726864, + "grad_norm": 5.9742644322023075, + "learning_rate": 8.363618312649967e-07, + "loss": 1.8452, + "step": 5025 + }, + { + "epoch": 0.33893736733937535, + "grad_norm": 6.181641201859419, + "learning_rate": 8.359265049567796e-07, + "loss": 1.8412, + "step": 5030 + }, + { + "epoch": 0.3392742832114821, + "grad_norm": 5.2367742865731985, + "learning_rate": 8.35490713992985e-07, + "loss": 1.861, + "step": 5035 + }, + { + "epoch": 0.33961119908358883, + "grad_norm": 6.292690765399438, + "learning_rate": 8.350544589764015e-07, + "loss": 1.8772, + "step": 5040 + }, + { + "epoch": 0.33994811495569555, + "grad_norm": 6.051654185566775, + "learning_rate": 8.346177405104595e-07, + "loss": 1.8692, + "step": 5045 + }, + { + "epoch": 0.3402850308278023, + "grad_norm": 6.310395089126805, + "learning_rate": 8.341805591992308e-07, + "loss": 1.8857, + "step": 5050 + }, + { + "epoch": 0.34062194669990903, + "grad_norm": 5.8038677244137515, + "learning_rate": 8.337429156474272e-07, + "loss": 1.7971, + "step": 5055 + }, + { + "epoch": 0.34095886257201574, + "grad_norm": 5.620673928318591, + "learning_rate": 8.333048104603999e-07, + "loss": 1.7906, + "step": 5060 + }, + { + "epoch": 0.3412957784441225, + "grad_norm": 5.925283173151023, + "learning_rate": 8.328662442441388e-07, + "loss": 1.7609, + "step": 5065 + }, + { + "epoch": 0.3416326943162292, + "grad_norm": 5.5148787672747135, + "learning_rate": 8.32427217605271e-07, + "loss": 1.8625, + "step": 5070 + }, + { + "epoch": 0.341969610188336, + "grad_norm": 5.889799319547815, + "learning_rate": 8.319877311510612e-07, + "loss": 1.8872, + "step": 5075 + }, + { + "epoch": 0.3423065260604427, + "grad_norm": 5.9730991305417644, + "learning_rate": 8.315477854894095e-07, + "loss": 1.8302, + "step": 5080 + }, + { + "epoch": 0.3426434419325494, + "grad_norm": 6.112512298326469, + "learning_rate": 8.311073812288513e-07, + "loss": 1.8695, + "step": 5085 + }, + { + "epoch": 0.3429803578046562, + "grad_norm": 5.960158481758416, + "learning_rate": 8.306665189785567e-07, + "loss": 1.8504, + "step": 5090 + }, + { + "epoch": 0.3433172736767629, + "grad_norm": 5.762403166494906, + "learning_rate": 8.302251993483289e-07, + "loss": 1.8598, + "step": 5095 + }, + { + "epoch": 0.34365418954886967, + "grad_norm": 6.407712181083279, + "learning_rate": 8.297834229486039e-07, + "loss": 1.8336, + "step": 5100 + }, + { + "epoch": 0.3439911054209764, + "grad_norm": 5.83772252154374, + "learning_rate": 8.293411903904496e-07, + "loss": 1.8033, + "step": 5105 + }, + { + "epoch": 0.3443280212930831, + "grad_norm": 5.833573633058513, + "learning_rate": 8.288985022855645e-07, + "loss": 1.9005, + "step": 5110 + }, + { + "epoch": 0.34466493716518987, + "grad_norm": 6.066836504891038, + "learning_rate": 8.284553592462778e-07, + "loss": 1.8898, + "step": 5115 + }, + { + "epoch": 0.3450018530372966, + "grad_norm": 5.819693757597866, + "learning_rate": 8.280117618855475e-07, + "loss": 1.892, + "step": 5120 + }, + { + "epoch": 0.34533876890940335, + "grad_norm": 6.175735031767354, + "learning_rate": 8.2756771081696e-07, + "loss": 1.8891, + "step": 5125 + }, + { + "epoch": 0.34567568478151006, + "grad_norm": 5.861251644799363, + "learning_rate": 8.271232066547296e-07, + "loss": 1.837, + "step": 5130 + }, + { + "epoch": 0.3460126006536168, + "grad_norm": 5.4275825974852685, + "learning_rate": 8.266782500136971e-07, + "loss": 1.8856, + "step": 5135 + }, + { + "epoch": 0.34634951652572354, + "grad_norm": 6.545104428227057, + "learning_rate": 8.262328415093293e-07, + "loss": 1.8737, + "step": 5140 + }, + { + "epoch": 0.34668643239783026, + "grad_norm": 5.818410117388077, + "learning_rate": 8.257869817577179e-07, + "loss": 1.8721, + "step": 5145 + }, + { + "epoch": 0.34702334826993697, + "grad_norm": 5.848986584909808, + "learning_rate": 8.253406713755786e-07, + "loss": 1.8334, + "step": 5150 + }, + { + "epoch": 0.34736026414204374, + "grad_norm": 6.225066432843631, + "learning_rate": 8.24893910980251e-07, + "loss": 1.7952, + "step": 5155 + }, + { + "epoch": 0.34769718001415045, + "grad_norm": 6.2022839915786525, + "learning_rate": 8.244467011896965e-07, + "loss": 1.7771, + "step": 5160 + }, + { + "epoch": 0.3480340958862572, + "grad_norm": 5.666990913330156, + "learning_rate": 8.239990426224986e-07, + "loss": 1.8682, + "step": 5165 + }, + { + "epoch": 0.34837101175836394, + "grad_norm": 5.725148875884774, + "learning_rate": 8.235509358978611e-07, + "loss": 1.8808, + "step": 5170 + }, + { + "epoch": 0.34870792763047065, + "grad_norm": 6.191906465540604, + "learning_rate": 8.231023816356081e-07, + "loss": 1.808, + "step": 5175 + }, + { + "epoch": 0.3490448435025774, + "grad_norm": 6.300765434889074, + "learning_rate": 8.226533804561826e-07, + "loss": 1.826, + "step": 5180 + }, + { + "epoch": 0.34938175937468413, + "grad_norm": 6.120294191430685, + "learning_rate": 8.222039329806456e-07, + "loss": 1.8224, + "step": 5185 + }, + { + "epoch": 0.3497186752467909, + "grad_norm": 6.035289816758244, + "learning_rate": 8.217540398306757e-07, + "loss": 1.7948, + "step": 5190 + }, + { + "epoch": 0.3500555911188976, + "grad_norm": 6.165541268890447, + "learning_rate": 8.213037016285679e-07, + "loss": 1.8672, + "step": 5195 + }, + { + "epoch": 0.3503925069910043, + "grad_norm": 6.244795566784756, + "learning_rate": 8.208529189972325e-07, + "loss": 1.8504, + "step": 5200 + }, + { + "epoch": 0.3507294228631111, + "grad_norm": 6.429283407205037, + "learning_rate": 8.204016925601951e-07, + "loss": 1.8786, + "step": 5205 + }, + { + "epoch": 0.3510663387352178, + "grad_norm": 6.505410428263308, + "learning_rate": 8.199500229415945e-07, + "loss": 1.8322, + "step": 5210 + }, + { + "epoch": 0.3514032546073246, + "grad_norm": 5.931535804172131, + "learning_rate": 8.19497910766183e-07, + "loss": 1.8532, + "step": 5215 + }, + { + "epoch": 0.3517401704794313, + "grad_norm": 5.8669374305422135, + "learning_rate": 8.19045356659325e-07, + "loss": 1.858, + "step": 5220 + }, + { + "epoch": 0.352077086351538, + "grad_norm": 6.587711874884489, + "learning_rate": 8.185923612469958e-07, + "loss": 1.814, + "step": 5225 + }, + { + "epoch": 0.3524140022236448, + "grad_norm": 6.025476451555223, + "learning_rate": 8.181389251557817e-07, + "loss": 1.834, + "step": 5230 + }, + { + "epoch": 0.3527509180957515, + "grad_norm": 6.144844744356217, + "learning_rate": 8.176850490128782e-07, + "loss": 1.8182, + "step": 5235 + }, + { + "epoch": 0.3530878339678582, + "grad_norm": 5.977324102576289, + "learning_rate": 8.172307334460892e-07, + "loss": 1.9074, + "step": 5240 + }, + { + "epoch": 0.35342474983996497, + "grad_norm": 6.500223474064341, + "learning_rate": 8.167759790838273e-07, + "loss": 1.8393, + "step": 5245 + }, + { + "epoch": 0.3537616657120717, + "grad_norm": 6.276845349036307, + "learning_rate": 8.163207865551111e-07, + "loss": 1.7774, + "step": 5250 + }, + { + "epoch": 0.35409858158417845, + "grad_norm": 6.341988614555882, + "learning_rate": 8.158651564895657e-07, + "loss": 1.791, + "step": 5255 + }, + { + "epoch": 0.35443549745628516, + "grad_norm": 6.143402965283996, + "learning_rate": 8.154090895174215e-07, + "loss": 1.8424, + "step": 5260 + }, + { + "epoch": 0.3547724133283919, + "grad_norm": 6.025218077365625, + "learning_rate": 8.149525862695131e-07, + "loss": 1.8894, + "step": 5265 + }, + { + "epoch": 0.35510932920049865, + "grad_norm": 5.665571111499496, + "learning_rate": 8.144956473772784e-07, + "loss": 1.8342, + "step": 5270 + }, + { + "epoch": 0.35544624507260536, + "grad_norm": 6.059993906864897, + "learning_rate": 8.140382734727581e-07, + "loss": 1.8624, + "step": 5275 + }, + { + "epoch": 0.35578316094471213, + "grad_norm": 5.834225595488122, + "learning_rate": 8.135804651885946e-07, + "loss": 1.8717, + "step": 5280 + }, + { + "epoch": 0.35612007681681884, + "grad_norm": 6.162555810291783, + "learning_rate": 8.131222231580313e-07, + "loss": 1.814, + "step": 5285 + }, + { + "epoch": 0.35645699268892556, + "grad_norm": 6.0975790501152485, + "learning_rate": 8.126635480149107e-07, + "loss": 1.8628, + "step": 5290 + }, + { + "epoch": 0.3567939085610323, + "grad_norm": 6.105019766737801, + "learning_rate": 8.122044403936759e-07, + "loss": 1.8526, + "step": 5295 + }, + { + "epoch": 0.35713082443313904, + "grad_norm": 6.306625818941117, + "learning_rate": 8.117449009293668e-07, + "loss": 1.8393, + "step": 5300 + }, + { + "epoch": 0.3574677403052458, + "grad_norm": 6.135991639401845, + "learning_rate": 8.112849302576212e-07, + "loss": 1.845, + "step": 5305 + }, + { + "epoch": 0.3578046561773525, + "grad_norm": 5.463394231566084, + "learning_rate": 8.108245290146735e-07, + "loss": 1.7506, + "step": 5310 + }, + { + "epoch": 0.35814157204945923, + "grad_norm": 6.291456002634215, + "learning_rate": 8.103636978373534e-07, + "loss": 1.7982, + "step": 5315 + }, + { + "epoch": 0.358478487921566, + "grad_norm": 6.004992246013968, + "learning_rate": 8.099024373630854e-07, + "loss": 1.7842, + "step": 5320 + }, + { + "epoch": 0.3588154037936727, + "grad_norm": 6.023428148718635, + "learning_rate": 8.094407482298877e-07, + "loss": 1.8286, + "step": 5325 + }, + { + "epoch": 0.35915231966577943, + "grad_norm": 6.107016316566931, + "learning_rate": 8.089786310763716e-07, + "loss": 1.8459, + "step": 5330 + }, + { + "epoch": 0.3594892355378862, + "grad_norm": 5.944550432690106, + "learning_rate": 8.085160865417403e-07, + "loss": 1.7915, + "step": 5335 + }, + { + "epoch": 0.3598261514099929, + "grad_norm": 5.8836749674593944, + "learning_rate": 8.080531152657884e-07, + "loss": 1.8045, + "step": 5340 + }, + { + "epoch": 0.3601630672820997, + "grad_norm": 6.529445698249722, + "learning_rate": 8.075897178889002e-07, + "loss": 1.83, + "step": 5345 + }, + { + "epoch": 0.3604999831542064, + "grad_norm": 6.2484120930888, + "learning_rate": 8.071258950520501e-07, + "loss": 1.8925, + "step": 5350 + }, + { + "epoch": 0.3608368990263131, + "grad_norm": 5.739450767655285, + "learning_rate": 8.066616473968005e-07, + "loss": 1.887, + "step": 5355 + }, + { + "epoch": 0.3611738148984199, + "grad_norm": 6.136883926564676, + "learning_rate": 8.061969755653013e-07, + "loss": 1.8956, + "step": 5360 + }, + { + "epoch": 0.3615107307705266, + "grad_norm": 6.099150976787944, + "learning_rate": 8.0573188020029e-07, + "loss": 1.8737, + "step": 5365 + }, + { + "epoch": 0.36184764664263336, + "grad_norm": 5.92471013455527, + "learning_rate": 8.052663619450889e-07, + "loss": 1.7661, + "step": 5370 + }, + { + "epoch": 0.36218456251474007, + "grad_norm": 6.235057387424368, + "learning_rate": 8.048004214436058e-07, + "loss": 1.8556, + "step": 5375 + }, + { + "epoch": 0.3625214783868468, + "grad_norm": 6.17438780027338, + "learning_rate": 8.043340593403325e-07, + "loss": 1.9215, + "step": 5380 + }, + { + "epoch": 0.36285839425895355, + "grad_norm": 6.424536207341481, + "learning_rate": 8.038672762803437e-07, + "loss": 1.8529, + "step": 5385 + }, + { + "epoch": 0.36319531013106027, + "grad_norm": 5.795068157540271, + "learning_rate": 8.034000729092967e-07, + "loss": 1.8689, + "step": 5390 + }, + { + "epoch": 0.36353222600316704, + "grad_norm": 6.7129075662939055, + "learning_rate": 8.029324498734299e-07, + "loss": 1.8011, + "step": 5395 + }, + { + "epoch": 0.36386914187527375, + "grad_norm": 6.016389228550978, + "learning_rate": 8.024644078195625e-07, + "loss": 1.8532, + "step": 5400 + }, + { + "epoch": 0.36420605774738046, + "grad_norm": 6.239532053660758, + "learning_rate": 8.01995947395093e-07, + "loss": 1.8392, + "step": 5405 + }, + { + "epoch": 0.36454297361948723, + "grad_norm": 5.916493777487104, + "learning_rate": 8.015270692479988e-07, + "loss": 1.9247, + "step": 5410 + }, + { + "epoch": 0.36487988949159395, + "grad_norm": 6.612015986424919, + "learning_rate": 8.010577740268347e-07, + "loss": 1.8547, + "step": 5415 + }, + { + "epoch": 0.36521680536370066, + "grad_norm": 6.181898400080323, + "learning_rate": 8.005880623807331e-07, + "loss": 1.764, + "step": 5420 + }, + { + "epoch": 0.3655537212358074, + "grad_norm": 6.493674123294548, + "learning_rate": 8.001179349594016e-07, + "loss": 1.8997, + "step": 5425 + }, + { + "epoch": 0.36589063710791414, + "grad_norm": 5.998232222470107, + "learning_rate": 7.996473924131236e-07, + "loss": 1.909, + "step": 5430 + }, + { + "epoch": 0.3662275529800209, + "grad_norm": 6.2309580428499185, + "learning_rate": 7.991764353927562e-07, + "loss": 1.8379, + "step": 5435 + }, + { + "epoch": 0.3665644688521276, + "grad_norm": 6.1097747554443975, + "learning_rate": 7.987050645497302e-07, + "loss": 1.8493, + "step": 5440 + }, + { + "epoch": 0.36690138472423434, + "grad_norm": 7.217536394961145, + "learning_rate": 7.982332805360486e-07, + "loss": 1.8435, + "step": 5445 + }, + { + "epoch": 0.3672383005963411, + "grad_norm": 6.426087127620301, + "learning_rate": 7.977610840042856e-07, + "loss": 1.8376, + "step": 5450 + }, + { + "epoch": 0.3675752164684478, + "grad_norm": 5.852501492146653, + "learning_rate": 7.972884756075867e-07, + "loss": 1.9335, + "step": 5455 + }, + { + "epoch": 0.3679121323405546, + "grad_norm": 6.68642611327528, + "learning_rate": 7.968154559996665e-07, + "loss": 1.8761, + "step": 5460 + }, + { + "epoch": 0.3682490482126613, + "grad_norm": 5.9075032622037575, + "learning_rate": 7.963420258348086e-07, + "loss": 1.8118, + "step": 5465 + }, + { + "epoch": 0.368585964084768, + "grad_norm": 5.775018051587756, + "learning_rate": 7.958681857678645e-07, + "loss": 1.841, + "step": 5470 + }, + { + "epoch": 0.3689228799568748, + "grad_norm": 5.519542365682442, + "learning_rate": 7.953939364542523e-07, + "loss": 1.7795, + "step": 5475 + }, + { + "epoch": 0.3692597958289815, + "grad_norm": 6.051018669323594, + "learning_rate": 7.949192785499573e-07, + "loss": 1.8035, + "step": 5480 + }, + { + "epoch": 0.36959671170108827, + "grad_norm": 6.173790364293265, + "learning_rate": 7.944442127115285e-07, + "loss": 1.8174, + "step": 5485 + }, + { + "epoch": 0.369933627573195, + "grad_norm": 6.27057349475925, + "learning_rate": 7.939687395960802e-07, + "loss": 1.8228, + "step": 5490 + }, + { + "epoch": 0.3702705434453017, + "grad_norm": 6.403064812356834, + "learning_rate": 7.934928598612895e-07, + "loss": 1.7758, + "step": 5495 + }, + { + "epoch": 0.37060745931740846, + "grad_norm": 5.78875486509689, + "learning_rate": 7.930165741653964e-07, + "loss": 1.8362, + "step": 5500 + }, + { + "epoch": 0.3709443751895152, + "grad_norm": 6.209627219262116, + "learning_rate": 7.925398831672018e-07, + "loss": 1.8188, + "step": 5505 + }, + { + "epoch": 0.3712812910616219, + "grad_norm": 5.910263475062196, + "learning_rate": 7.920627875260679e-07, + "loss": 1.9461, + "step": 5510 + }, + { + "epoch": 0.37161820693372866, + "grad_norm": 5.907600463890629, + "learning_rate": 7.91585287901916e-07, + "loss": 1.8488, + "step": 5515 + }, + { + "epoch": 0.37195512280583537, + "grad_norm": 6.022008103501391, + "learning_rate": 7.911073849552267e-07, + "loss": 1.8528, + "step": 5520 + }, + { + "epoch": 0.37229203867794214, + "grad_norm": 6.297851834336272, + "learning_rate": 7.906290793470382e-07, + "loss": 1.8826, + "step": 5525 + }, + { + "epoch": 0.37262895455004885, + "grad_norm": 6.821824970703982, + "learning_rate": 7.901503717389458e-07, + "loss": 1.8499, + "step": 5530 + }, + { + "epoch": 0.37296587042215557, + "grad_norm": 6.543670341910588, + "learning_rate": 7.896712627931004e-07, + "loss": 1.8973, + "step": 5535 + }, + { + "epoch": 0.37330278629426233, + "grad_norm": 5.839271026858895, + "learning_rate": 7.891917531722087e-07, + "loss": 1.8527, + "step": 5540 + }, + { + "epoch": 0.37363970216636905, + "grad_norm": 5.805332246145514, + "learning_rate": 7.887118435395314e-07, + "loss": 1.8391, + "step": 5545 + }, + { + "epoch": 0.3739766180384758, + "grad_norm": 6.093441102979253, + "learning_rate": 7.882315345588823e-07, + "loss": 1.8382, + "step": 5550 + }, + { + "epoch": 0.37431353391058253, + "grad_norm": 5.9164472483146255, + "learning_rate": 7.877508268946275e-07, + "loss": 1.8068, + "step": 5555 + }, + { + "epoch": 0.37465044978268924, + "grad_norm": 5.783501174118739, + "learning_rate": 7.87269721211685e-07, + "loss": 1.8326, + "step": 5560 + }, + { + "epoch": 0.374987365654796, + "grad_norm": 6.099236664954797, + "learning_rate": 7.86788218175523e-07, + "loss": 1.8907, + "step": 5565 + }, + { + "epoch": 0.3753242815269027, + "grad_norm": 6.015599331347923, + "learning_rate": 7.863063184521595e-07, + "loss": 1.7902, + "step": 5570 + }, + { + "epoch": 0.3756611973990095, + "grad_norm": 6.030584133610423, + "learning_rate": 7.858240227081611e-07, + "loss": 1.7967, + "step": 5575 + }, + { + "epoch": 0.3759981132711162, + "grad_norm": 6.1561594264866875, + "learning_rate": 7.85341331610642e-07, + "loss": 1.8738, + "step": 5580 + }, + { + "epoch": 0.3763350291432229, + "grad_norm": 5.7668037858810735, + "learning_rate": 7.848582458272637e-07, + "loss": 1.8994, + "step": 5585 + }, + { + "epoch": 0.3766719450153297, + "grad_norm": 6.820449431402006, + "learning_rate": 7.843747660262333e-07, + "loss": 1.7353, + "step": 5590 + }, + { + "epoch": 0.3770088608874364, + "grad_norm": 5.731863695815324, + "learning_rate": 7.838908928763028e-07, + "loss": 1.842, + "step": 5595 + }, + { + "epoch": 0.3773457767595431, + "grad_norm": 6.0491543709457805, + "learning_rate": 7.834066270467689e-07, + "loss": 1.8412, + "step": 5600 + }, + { + "epoch": 0.3776826926316499, + "grad_norm": 6.312515900909037, + "learning_rate": 7.829219692074707e-07, + "loss": 1.9166, + "step": 5605 + }, + { + "epoch": 0.3780196085037566, + "grad_norm": 5.629903210882783, + "learning_rate": 7.824369200287899e-07, + "loss": 1.8475, + "step": 5610 + }, + { + "epoch": 0.37835652437586337, + "grad_norm": 5.798215148864628, + "learning_rate": 7.819514801816496e-07, + "loss": 1.8445, + "step": 5615 + }, + { + "epoch": 0.3786934402479701, + "grad_norm": 5.746602784624546, + "learning_rate": 7.814656503375128e-07, + "loss": 1.8142, + "step": 5620 + }, + { + "epoch": 0.3790303561200768, + "grad_norm": 6.212384108148857, + "learning_rate": 7.809794311683828e-07, + "loss": 1.8433, + "step": 5625 + }, + { + "epoch": 0.37936727199218356, + "grad_norm": 5.5219568352553186, + "learning_rate": 7.804928233468006e-07, + "loss": 1.8577, + "step": 5630 + }, + { + "epoch": 0.3797041878642903, + "grad_norm": 5.923816755049023, + "learning_rate": 7.80005827545845e-07, + "loss": 1.8968, + "step": 5635 + }, + { + "epoch": 0.38004110373639705, + "grad_norm": 6.447298966369827, + "learning_rate": 7.795184444391318e-07, + "loss": 1.8263, + "step": 5640 + }, + { + "epoch": 0.38037801960850376, + "grad_norm": 5.700024915629011, + "learning_rate": 7.790306747008119e-07, + "loss": 1.801, + "step": 5645 + }, + { + "epoch": 0.3807149354806105, + "grad_norm": 5.693902051404934, + "learning_rate": 7.785425190055719e-07, + "loss": 1.8346, + "step": 5650 + }, + { + "epoch": 0.38105185135271724, + "grad_norm": 6.623591217723914, + "learning_rate": 7.780539780286312e-07, + "loss": 1.7844, + "step": 5655 + }, + { + "epoch": 0.38138876722482395, + "grad_norm": 5.924554882226798, + "learning_rate": 7.77565052445743e-07, + "loss": 1.8806, + "step": 5660 + }, + { + "epoch": 0.3817256830969307, + "grad_norm": 5.804924090163828, + "learning_rate": 7.770757429331919e-07, + "loss": 1.8837, + "step": 5665 + }, + { + "epoch": 0.38206259896903744, + "grad_norm": 6.307078378352449, + "learning_rate": 7.765860501677939e-07, + "loss": 1.8476, + "step": 5670 + }, + { + "epoch": 0.38239951484114415, + "grad_norm": 5.64288963732036, + "learning_rate": 7.760959748268949e-07, + "loss": 1.7921, + "step": 5675 + }, + { + "epoch": 0.3827364307132509, + "grad_norm": 6.0330255144031755, + "learning_rate": 7.756055175883701e-07, + "loss": 1.8648, + "step": 5680 + }, + { + "epoch": 0.38307334658535763, + "grad_norm": 6.226955158037713, + "learning_rate": 7.751146791306231e-07, + "loss": 1.8733, + "step": 5685 + }, + { + "epoch": 0.38341026245746435, + "grad_norm": 5.831296049474745, + "learning_rate": 7.746234601325843e-07, + "loss": 1.8237, + "step": 5690 + }, + { + "epoch": 0.3837471783295711, + "grad_norm": 6.138631169330866, + "learning_rate": 7.741318612737111e-07, + "loss": 1.8234, + "step": 5695 + }, + { + "epoch": 0.38408409420167783, + "grad_norm": 5.9623003368564955, + "learning_rate": 7.73639883233986e-07, + "loss": 1.7892, + "step": 5700 + }, + { + "epoch": 0.3844210100737846, + "grad_norm": 5.900675764770894, + "learning_rate": 7.731475266939158e-07, + "loss": 1.7886, + "step": 5705 + }, + { + "epoch": 0.3847579259458913, + "grad_norm": 6.598946949032724, + "learning_rate": 7.726547923345313e-07, + "loss": 1.832, + "step": 5710 + }, + { + "epoch": 0.385094841817998, + "grad_norm": 6.274691790815466, + "learning_rate": 7.721616808373855e-07, + "loss": 1.8692, + "step": 5715 + }, + { + "epoch": 0.3854317576901048, + "grad_norm": 6.068967193945496, + "learning_rate": 7.716681928845532e-07, + "loss": 1.8135, + "step": 5720 + }, + { + "epoch": 0.3857686735622115, + "grad_norm": 5.71438663141986, + "learning_rate": 7.711743291586298e-07, + "loss": 1.8432, + "step": 5725 + }, + { + "epoch": 0.3861055894343183, + "grad_norm": 5.908196677526078, + "learning_rate": 7.706800903427309e-07, + "loss": 1.76, + "step": 5730 + }, + { + "epoch": 0.386442505306425, + "grad_norm": 5.980404418034513, + "learning_rate": 7.701854771204905e-07, + "loss": 1.8551, + "step": 5735 + }, + { + "epoch": 0.3867794211785317, + "grad_norm": 5.825977075219841, + "learning_rate": 7.696904901760606e-07, + "loss": 1.8143, + "step": 5740 + }, + { + "epoch": 0.38711633705063847, + "grad_norm": 6.147112371868563, + "learning_rate": 7.691951301941102e-07, + "loss": 1.8429, + "step": 5745 + }, + { + "epoch": 0.3874532529227452, + "grad_norm": 6.679128666452009, + "learning_rate": 7.68699397859824e-07, + "loss": 1.9134, + "step": 5750 + }, + { + "epoch": 0.38779016879485195, + "grad_norm": 5.97814777855936, + "learning_rate": 7.682032938589023e-07, + "loss": 1.8395, + "step": 5755 + }, + { + "epoch": 0.38812708466695867, + "grad_norm": 5.950552045235804, + "learning_rate": 7.677068188775589e-07, + "loss": 1.7672, + "step": 5760 + }, + { + "epoch": 0.3884640005390654, + "grad_norm": 5.814114869159672, + "learning_rate": 7.67209973602521e-07, + "loss": 1.8458, + "step": 5765 + }, + { + "epoch": 0.38880091641117215, + "grad_norm": 6.0582073054991055, + "learning_rate": 7.667127587210282e-07, + "loss": 1.8802, + "step": 5770 + }, + { + "epoch": 0.38913783228327886, + "grad_norm": 6.334265700850607, + "learning_rate": 7.66215174920831e-07, + "loss": 1.8564, + "step": 5775 + }, + { + "epoch": 0.3894747481553856, + "grad_norm": 6.002444518331614, + "learning_rate": 7.657172228901905e-07, + "loss": 1.8211, + "step": 5780 + }, + { + "epoch": 0.38981166402749234, + "grad_norm": 6.1022759973086815, + "learning_rate": 7.652189033178766e-07, + "loss": 1.8229, + "step": 5785 + }, + { + "epoch": 0.39014857989959906, + "grad_norm": 6.413997576789767, + "learning_rate": 7.647202168931683e-07, + "loss": 1.7785, + "step": 5790 + }, + { + "epoch": 0.3904854957717058, + "grad_norm": 6.069802887796607, + "learning_rate": 7.642211643058516e-07, + "loss": 1.8022, + "step": 5795 + }, + { + "epoch": 0.39082241164381254, + "grad_norm": 5.7749180044113, + "learning_rate": 7.637217462462189e-07, + "loss": 1.8342, + "step": 5800 + }, + { + "epoch": 0.39115932751591925, + "grad_norm": 5.841685430277685, + "learning_rate": 7.632219634050685e-07, + "loss": 1.8166, + "step": 5805 + }, + { + "epoch": 0.391496243388026, + "grad_norm": 6.113028925689592, + "learning_rate": 7.62721816473703e-07, + "loss": 1.8851, + "step": 5810 + }, + { + "epoch": 0.39183315926013274, + "grad_norm": 6.574877319300081, + "learning_rate": 7.622213061439287e-07, + "loss": 1.7852, + "step": 5815 + }, + { + "epoch": 0.3921700751322395, + "grad_norm": 6.591103453708863, + "learning_rate": 7.617204331080544e-07, + "loss": 1.8207, + "step": 5820 + }, + { + "epoch": 0.3925069910043462, + "grad_norm": 6.161561709386749, + "learning_rate": 7.612191980588907e-07, + "loss": 1.8254, + "step": 5825 + }, + { + "epoch": 0.39284390687645293, + "grad_norm": 5.913954308903591, + "learning_rate": 7.60717601689749e-07, + "loss": 1.8521, + "step": 5830 + }, + { + "epoch": 0.3931808227485597, + "grad_norm": 6.061050670183792, + "learning_rate": 7.602156446944405e-07, + "loss": 1.8292, + "step": 5835 + }, + { + "epoch": 0.3935177386206664, + "grad_norm": 6.056114290565816, + "learning_rate": 7.597133277672751e-07, + "loss": 1.8017, + "step": 5840 + }, + { + "epoch": 0.3938546544927732, + "grad_norm": 6.659560363196615, + "learning_rate": 7.592106516030607e-07, + "loss": 1.7902, + "step": 5845 + }, + { + "epoch": 0.3941915703648799, + "grad_norm": 6.349531573185438, + "learning_rate": 7.587076168971022e-07, + "loss": 1.7835, + "step": 5850 + }, + { + "epoch": 0.3945284862369866, + "grad_norm": 5.973870975420067, + "learning_rate": 7.582042243451998e-07, + "loss": 1.839, + "step": 5855 + }, + { + "epoch": 0.3948654021090934, + "grad_norm": 6.399608469649776, + "learning_rate": 7.577004746436494e-07, + "loss": 1.8484, + "step": 5860 + }, + { + "epoch": 0.3952023179812001, + "grad_norm": 6.391523749018398, + "learning_rate": 7.571963684892404e-07, + "loss": 1.8203, + "step": 5865 + }, + { + "epoch": 0.3955392338533068, + "grad_norm": 5.55251848998224, + "learning_rate": 7.566919065792558e-07, + "loss": 1.8143, + "step": 5870 + }, + { + "epoch": 0.3958761497254136, + "grad_norm": 6.461778389781404, + "learning_rate": 7.561870896114704e-07, + "loss": 1.7778, + "step": 5875 + }, + { + "epoch": 0.3962130655975203, + "grad_norm": 5.7346585594091115, + "learning_rate": 7.556819182841498e-07, + "loss": 1.7838, + "step": 5880 + }, + { + "epoch": 0.39654998146962706, + "grad_norm": 5.843198641258833, + "learning_rate": 7.551763932960502e-07, + "loss": 1.8738, + "step": 5885 + }, + { + "epoch": 0.39688689734173377, + "grad_norm": 5.993845274934363, + "learning_rate": 7.546705153464168e-07, + "loss": 1.8547, + "step": 5890 + }, + { + "epoch": 0.3972238132138405, + "grad_norm": 5.7984474701034925, + "learning_rate": 7.54164285134983e-07, + "loss": 1.8138, + "step": 5895 + }, + { + "epoch": 0.39756072908594725, + "grad_norm": 6.040293907002973, + "learning_rate": 7.536577033619696e-07, + "loss": 1.8491, + "step": 5900 + }, + { + "epoch": 0.39789764495805396, + "grad_norm": 5.9285219940169664, + "learning_rate": 7.531507707280836e-07, + "loss": 1.816, + "step": 5905 + }, + { + "epoch": 0.39823456083016073, + "grad_norm": 6.001261820811947, + "learning_rate": 7.526434879345171e-07, + "loss": 1.8569, + "step": 5910 + }, + { + "epoch": 0.39857147670226745, + "grad_norm": 6.279600146634121, + "learning_rate": 7.521358556829469e-07, + "loss": 1.8775, + "step": 5915 + }, + { + "epoch": 0.39890839257437416, + "grad_norm": 5.919665080614914, + "learning_rate": 7.51627874675533e-07, + "loss": 1.8932, + "step": 5920 + }, + { + "epoch": 0.39924530844648093, + "grad_norm": 6.237664010789953, + "learning_rate": 7.511195456149177e-07, + "loss": 1.8613, + "step": 5925 + }, + { + "epoch": 0.39958222431858764, + "grad_norm": 5.453939969270052, + "learning_rate": 7.50610869204225e-07, + "loss": 1.8167, + "step": 5930 + }, + { + "epoch": 0.3999191401906944, + "grad_norm": 6.414570793475843, + "learning_rate": 7.50101846147059e-07, + "loss": 1.9238, + "step": 5935 + }, + { + "epoch": 0.4002560560628011, + "grad_norm": 6.214005751571211, + "learning_rate": 7.495924771475037e-07, + "loss": 1.8575, + "step": 5940 + }, + { + "epoch": 0.40059297193490784, + "grad_norm": 6.264117402993965, + "learning_rate": 7.490827629101211e-07, + "loss": 1.8432, + "step": 5945 + }, + { + "epoch": 0.4009298878070146, + "grad_norm": 5.876493353273485, + "learning_rate": 7.485727041399513e-07, + "loss": 1.7964, + "step": 5950 + }, + { + "epoch": 0.4012668036791213, + "grad_norm": 5.563447588486703, + "learning_rate": 7.480623015425105e-07, + "loss": 1.7979, + "step": 5955 + }, + { + "epoch": 0.40160371955122803, + "grad_norm": 5.576251643765876, + "learning_rate": 7.475515558237909e-07, + "loss": 1.8397, + "step": 5960 + }, + { + "epoch": 0.4019406354233348, + "grad_norm": 5.895209504068423, + "learning_rate": 7.470404676902587e-07, + "loss": 1.8415, + "step": 5965 + }, + { + "epoch": 0.4022775512954415, + "grad_norm": 5.977193260631528, + "learning_rate": 7.465290378488544e-07, + "loss": 1.9116, + "step": 5970 + }, + { + "epoch": 0.4026144671675483, + "grad_norm": 6.187315083034187, + "learning_rate": 7.460172670069909e-07, + "loss": 1.8695, + "step": 5975 + }, + { + "epoch": 0.402951383039655, + "grad_norm": 5.699459860019443, + "learning_rate": 7.455051558725524e-07, + "loss": 1.8654, + "step": 5980 + }, + { + "epoch": 0.4032882989117617, + "grad_norm": 5.988742641572804, + "learning_rate": 7.449927051538944e-07, + "loss": 1.7901, + "step": 5985 + }, + { + "epoch": 0.4036252147838685, + "grad_norm": 6.166914520856474, + "learning_rate": 7.444799155598419e-07, + "loss": 1.8478, + "step": 5990 + }, + { + "epoch": 0.4039621306559752, + "grad_norm": 6.290632854081438, + "learning_rate": 7.439667877996884e-07, + "loss": 1.7836, + "step": 5995 + }, + { + "epoch": 0.40429904652808196, + "grad_norm": 5.578242942924098, + "learning_rate": 7.434533225831951e-07, + "loss": 1.8646, + "step": 6000 + }, + { + "epoch": 0.4046359624001887, + "grad_norm": 5.922649460674123, + "learning_rate": 7.429395206205908e-07, + "loss": 1.8159, + "step": 6005 + }, + { + "epoch": 0.4049728782722954, + "grad_norm": 5.973950129607427, + "learning_rate": 7.424253826225689e-07, + "loss": 1.8179, + "step": 6010 + }, + { + "epoch": 0.40530979414440216, + "grad_norm": 6.1535636579164565, + "learning_rate": 7.419109093002887e-07, + "loss": 1.785, + "step": 6015 + }, + { + "epoch": 0.40564671001650887, + "grad_norm": 5.99390268127196, + "learning_rate": 7.413961013653725e-07, + "loss": 1.8474, + "step": 6020 + }, + { + "epoch": 0.40598362588861564, + "grad_norm": 6.095296832831377, + "learning_rate": 7.408809595299057e-07, + "loss": 1.8098, + "step": 6025 + }, + { + "epoch": 0.40632054176072235, + "grad_norm": 5.854910765823886, + "learning_rate": 7.403654845064358e-07, + "loss": 1.8641, + "step": 6030 + }, + { + "epoch": 0.40665745763282907, + "grad_norm": 6.488094157305255, + "learning_rate": 7.398496770079709e-07, + "loss": 1.845, + "step": 6035 + }, + { + "epoch": 0.40699437350493584, + "grad_norm": 5.888696212338066, + "learning_rate": 7.393335377479792e-07, + "loss": 1.8355, + "step": 6040 + }, + { + "epoch": 0.40733128937704255, + "grad_norm": 5.6364358032730415, + "learning_rate": 7.388170674403872e-07, + "loss": 1.7867, + "step": 6045 + }, + { + "epoch": 0.40766820524914926, + "grad_norm": 6.120652995181019, + "learning_rate": 7.383002667995804e-07, + "loss": 1.8407, + "step": 6050 + }, + { + "epoch": 0.40800512112125603, + "grad_norm": 5.9456629334484115, + "learning_rate": 7.377831365404001e-07, + "loss": 1.7944, + "step": 6055 + }, + { + "epoch": 0.40834203699336274, + "grad_norm": 5.926271311098251, + "learning_rate": 7.372656773781442e-07, + "loss": 1.8741, + "step": 6060 + }, + { + "epoch": 0.4086789528654695, + "grad_norm": 5.874696662625113, + "learning_rate": 7.367478900285654e-07, + "loss": 1.8321, + "step": 6065 + }, + { + "epoch": 0.4090158687375762, + "grad_norm": 6.207699178257415, + "learning_rate": 7.362297752078702e-07, + "loss": 1.7815, + "step": 6070 + }, + { + "epoch": 0.40935278460968294, + "grad_norm": 6.0143311305393565, + "learning_rate": 7.357113336327181e-07, + "loss": 1.854, + "step": 6075 + }, + { + "epoch": 0.4096897004817897, + "grad_norm": 6.137764433429077, + "learning_rate": 7.351925660202207e-07, + "loss": 1.8032, + "step": 6080 + }, + { + "epoch": 0.4100266163538964, + "grad_norm": 5.952972611372365, + "learning_rate": 7.346734730879407e-07, + "loss": 1.8577, + "step": 6085 + }, + { + "epoch": 0.4103635322260032, + "grad_norm": 5.93496116315055, + "learning_rate": 7.341540555538902e-07, + "loss": 1.8655, + "step": 6090 + }, + { + "epoch": 0.4107004480981099, + "grad_norm": 6.035523911748048, + "learning_rate": 7.33634314136531e-07, + "loss": 1.8782, + "step": 6095 + }, + { + "epoch": 0.4110373639702166, + "grad_norm": 6.50895901457185, + "learning_rate": 7.331142495547724e-07, + "loss": 1.8557, + "step": 6100 + }, + { + "epoch": 0.4113742798423234, + "grad_norm": 6.3296523573714465, + "learning_rate": 7.325938625279709e-07, + "loss": 1.771, + "step": 6105 + }, + { + "epoch": 0.4117111957144301, + "grad_norm": 6.06063180713123, + "learning_rate": 7.320731537759293e-07, + "loss": 1.9209, + "step": 6110 + }, + { + "epoch": 0.41204811158653687, + "grad_norm": 6.291587004914758, + "learning_rate": 7.315521240188944e-07, + "loss": 1.8921, + "step": 6115 + }, + { + "epoch": 0.4123850274586436, + "grad_norm": 7.144681065930621, + "learning_rate": 7.310307739775585e-07, + "loss": 1.8353, + "step": 6120 + }, + { + "epoch": 0.4127219433307503, + "grad_norm": 6.537598454236774, + "learning_rate": 7.305091043730557e-07, + "loss": 1.8657, + "step": 6125 + }, + { + "epoch": 0.41305885920285706, + "grad_norm": 6.139638408298624, + "learning_rate": 7.299871159269626e-07, + "loss": 1.8291, + "step": 6130 + }, + { + "epoch": 0.4133957750749638, + "grad_norm": 6.052600414436344, + "learning_rate": 7.294648093612968e-07, + "loss": 1.8582, + "step": 6135 + }, + { + "epoch": 0.4137326909470705, + "grad_norm": 5.737200220672097, + "learning_rate": 7.28942185398516e-07, + "loss": 1.8016, + "step": 6140 + }, + { + "epoch": 0.41406960681917726, + "grad_norm": 5.788875630264053, + "learning_rate": 7.284192447615168e-07, + "loss": 1.8144, + "step": 6145 + }, + { + "epoch": 0.414406522691284, + "grad_norm": 5.809695140298498, + "learning_rate": 7.278959881736338e-07, + "loss": 1.8528, + "step": 6150 + }, + { + "epoch": 0.41474343856339074, + "grad_norm": 6.291846673442152, + "learning_rate": 7.273724163586387e-07, + "loss": 1.8146, + "step": 6155 + }, + { + "epoch": 0.41508035443549746, + "grad_norm": 6.592279506836028, + "learning_rate": 7.268485300407394e-07, + "loss": 1.8373, + "step": 6160 + }, + { + "epoch": 0.41541727030760417, + "grad_norm": 5.2404367912538135, + "learning_rate": 7.263243299445783e-07, + "loss": 1.8687, + "step": 6165 + }, + { + "epoch": 0.41575418617971094, + "grad_norm": 6.237826143972818, + "learning_rate": 7.257998167952322e-07, + "loss": 1.7559, + "step": 6170 + }, + { + "epoch": 0.41609110205181765, + "grad_norm": 5.644649333598512, + "learning_rate": 7.25274991318211e-07, + "loss": 1.7745, + "step": 6175 + }, + { + "epoch": 0.4164280179239244, + "grad_norm": 6.221269732535096, + "learning_rate": 7.247498542394566e-07, + "loss": 1.888, + "step": 6180 + }, + { + "epoch": 0.41676493379603113, + "grad_norm": 5.846110634880196, + "learning_rate": 7.242244062853416e-07, + "loss": 1.8297, + "step": 6185 + }, + { + "epoch": 0.41710184966813785, + "grad_norm": 6.0081486154977295, + "learning_rate": 7.236986481826688e-07, + "loss": 1.8962, + "step": 6190 + }, + { + "epoch": 0.4174387655402446, + "grad_norm": 5.714180142339893, + "learning_rate": 7.231725806586699e-07, + "loss": 1.798, + "step": 6195 + }, + { + "epoch": 0.41777568141235133, + "grad_norm": 5.924083302663139, + "learning_rate": 7.22646204441005e-07, + "loss": 1.8671, + "step": 6200 + }, + { + "epoch": 0.4181125972844581, + "grad_norm": 5.884534396453543, + "learning_rate": 7.221195202577606e-07, + "loss": 1.8473, + "step": 6205 + }, + { + "epoch": 0.4184495131565648, + "grad_norm": 6.505335975345712, + "learning_rate": 7.215925288374496e-07, + "loss": 1.845, + "step": 6210 + }, + { + "epoch": 0.4187864290286715, + "grad_norm": 6.361143535314032, + "learning_rate": 7.210652309090098e-07, + "loss": 1.8562, + "step": 6215 + }, + { + "epoch": 0.4191233449007783, + "grad_norm": 6.269558040309333, + "learning_rate": 7.205376272018025e-07, + "loss": 1.8842, + "step": 6220 + }, + { + "epoch": 0.419460260772885, + "grad_norm": 6.033684433696401, + "learning_rate": 7.200097184456128e-07, + "loss": 1.8744, + "step": 6225 + }, + { + "epoch": 0.4197971766449917, + "grad_norm": 5.893842528747161, + "learning_rate": 7.19481505370647e-07, + "loss": 1.9029, + "step": 6230 + }, + { + "epoch": 0.4201340925170985, + "grad_norm": 6.067247445280829, + "learning_rate": 7.189529887075327e-07, + "loss": 1.8535, + "step": 6235 + }, + { + "epoch": 0.4204710083892052, + "grad_norm": 6.019248123337608, + "learning_rate": 7.184241691873174e-07, + "loss": 1.8598, + "step": 6240 + }, + { + "epoch": 0.42080792426131197, + "grad_norm": 5.885826420361127, + "learning_rate": 7.178950475414675e-07, + "loss": 1.8843, + "step": 6245 + }, + { + "epoch": 0.4211448401334187, + "grad_norm": 5.994346100649246, + "learning_rate": 7.173656245018671e-07, + "loss": 1.8454, + "step": 6250 + }, + { + "epoch": 0.4214817560055254, + "grad_norm": 5.683347136227909, + "learning_rate": 7.168359008008177e-07, + "loss": 1.8528, + "step": 6255 + }, + { + "epoch": 0.42181867187763217, + "grad_norm": 6.303528633412533, + "learning_rate": 7.163058771710358e-07, + "loss": 1.8118, + "step": 6260 + }, + { + "epoch": 0.4221555877497389, + "grad_norm": 5.838897189398607, + "learning_rate": 7.157755543456539e-07, + "loss": 1.8554, + "step": 6265 + }, + { + "epoch": 0.42249250362184565, + "grad_norm": 6.102573455498519, + "learning_rate": 7.152449330582173e-07, + "loss": 1.8515, + "step": 6270 + }, + { + "epoch": 0.42282941949395236, + "grad_norm": 6.170469429143662, + "learning_rate": 7.147140140426848e-07, + "loss": 1.8718, + "step": 6275 + }, + { + "epoch": 0.4231663353660591, + "grad_norm": 6.603648144277644, + "learning_rate": 7.141827980334265e-07, + "loss": 1.7583, + "step": 6280 + }, + { + "epoch": 0.42350325123816585, + "grad_norm": 5.710276452721297, + "learning_rate": 7.136512857652239e-07, + "loss": 1.8225, + "step": 6285 + }, + { + "epoch": 0.42384016711027256, + "grad_norm": 6.305744171898564, + "learning_rate": 7.131194779732681e-07, + "loss": 1.8758, + "step": 6290 + }, + { + "epoch": 0.4241770829823793, + "grad_norm": 5.935812543100673, + "learning_rate": 7.125873753931586e-07, + "loss": 1.7446, + "step": 6295 + }, + { + "epoch": 0.42451399885448604, + "grad_norm": 6.4033336349656205, + "learning_rate": 7.120549787609029e-07, + "loss": 1.8007, + "step": 6300 + }, + { + "epoch": 0.42485091472659275, + "grad_norm": 5.988053466649247, + "learning_rate": 7.115222888129156e-07, + "loss": 1.8339, + "step": 6305 + }, + { + "epoch": 0.4251878305986995, + "grad_norm": 5.839719119201011, + "learning_rate": 7.109893062860161e-07, + "loss": 1.7753, + "step": 6310 + }, + { + "epoch": 0.42552474647080624, + "grad_norm": 6.1546266562601355, + "learning_rate": 7.104560319174296e-07, + "loss": 1.9032, + "step": 6315 + }, + { + "epoch": 0.42586166234291295, + "grad_norm": 5.868935206755803, + "learning_rate": 7.099224664447841e-07, + "loss": 1.8336, + "step": 6320 + }, + { + "epoch": 0.4261985782150197, + "grad_norm": 6.897803862049745, + "learning_rate": 7.093886106061106e-07, + "loss": 1.8933, + "step": 6325 + }, + { + "epoch": 0.42653549408712643, + "grad_norm": 6.298122444093038, + "learning_rate": 7.088544651398421e-07, + "loss": 1.8468, + "step": 6330 + }, + { + "epoch": 0.4268724099592332, + "grad_norm": 5.946687882042922, + "learning_rate": 7.083200307848115e-07, + "loss": 1.8509, + "step": 6335 + }, + { + "epoch": 0.4272093258313399, + "grad_norm": 5.9703726574219695, + "learning_rate": 7.077853082802516e-07, + "loss": 1.8723, + "step": 6340 + }, + { + "epoch": 0.42754624170344663, + "grad_norm": 5.940092483411608, + "learning_rate": 7.072502983657939e-07, + "loss": 1.7817, + "step": 6345 + }, + { + "epoch": 0.4278831575755534, + "grad_norm": 5.99693275214663, + "learning_rate": 7.067150017814676e-07, + "loss": 1.8194, + "step": 6350 + }, + { + "epoch": 0.4282200734476601, + "grad_norm": 6.365386492382532, + "learning_rate": 7.061794192676979e-07, + "loss": 1.8713, + "step": 6355 + }, + { + "epoch": 0.4285569893197669, + "grad_norm": 6.137677012183777, + "learning_rate": 7.056435515653058e-07, + "loss": 1.889, + "step": 6360 + }, + { + "epoch": 0.4288939051918736, + "grad_norm": 6.197193970097729, + "learning_rate": 7.051073994155068e-07, + "loss": 1.8624, + "step": 6365 + }, + { + "epoch": 0.4292308210639803, + "grad_norm": 6.262005951949787, + "learning_rate": 7.045709635599098e-07, + "loss": 1.8439, + "step": 6370 + }, + { + "epoch": 0.4295677369360871, + "grad_norm": 6.5893693825632, + "learning_rate": 7.040342447405161e-07, + "loss": 1.7952, + "step": 6375 + }, + { + "epoch": 0.4299046528081938, + "grad_norm": 6.342831308535354, + "learning_rate": 7.034972436997184e-07, + "loss": 1.7968, + "step": 6380 + }, + { + "epoch": 0.4302415686803005, + "grad_norm": 5.704318035103627, + "learning_rate": 7.029599611803e-07, + "loss": 1.8215, + "step": 6385 + }, + { + "epoch": 0.43057848455240727, + "grad_norm": 6.116744457100223, + "learning_rate": 7.024223979254331e-07, + "loss": 1.8791, + "step": 6390 + }, + { + "epoch": 0.430915400424514, + "grad_norm": 6.176754693260745, + "learning_rate": 7.018845546786787e-07, + "loss": 1.8132, + "step": 6395 + }, + { + "epoch": 0.43125231629662075, + "grad_norm": 5.751107796307478, + "learning_rate": 7.013464321839845e-07, + "loss": 1.7256, + "step": 6400 + }, + { + "epoch": 0.43158923216872747, + "grad_norm": 5.652758571110035, + "learning_rate": 7.00808031185685e-07, + "loss": 1.7953, + "step": 6405 + }, + { + "epoch": 0.4319261480408342, + "grad_norm": 6.005003107499712, + "learning_rate": 7.002693524284997e-07, + "loss": 1.8096, + "step": 6410 + }, + { + "epoch": 0.43226306391294095, + "grad_norm": 6.071057162652461, + "learning_rate": 6.997303966575322e-07, + "loss": 1.8587, + "step": 6415 + }, + { + "epoch": 0.43259997978504766, + "grad_norm": 6.363069927240701, + "learning_rate": 6.991911646182696e-07, + "loss": 1.8238, + "step": 6420 + }, + { + "epoch": 0.43293689565715443, + "grad_norm": 7.006943760155347, + "learning_rate": 6.986516570565809e-07, + "loss": 1.943, + "step": 6425 + }, + { + "epoch": 0.43327381152926114, + "grad_norm": 5.627017420401183, + "learning_rate": 6.981118747187163e-07, + "loss": 1.7897, + "step": 6430 + }, + { + "epoch": 0.43361072740136786, + "grad_norm": 5.4444971068326895, + "learning_rate": 6.975718183513056e-07, + "loss": 1.7919, + "step": 6435 + }, + { + "epoch": 0.4339476432734746, + "grad_norm": 6.062332184148201, + "learning_rate": 6.970314887013585e-07, + "loss": 1.7791, + "step": 6440 + }, + { + "epoch": 0.43428455914558134, + "grad_norm": 5.89355423464885, + "learning_rate": 6.964908865162617e-07, + "loss": 1.8423, + "step": 6445 + }, + { + "epoch": 0.4346214750176881, + "grad_norm": 6.6258832394007765, + "learning_rate": 6.959500125437801e-07, + "loss": 1.888, + "step": 6450 + }, + { + "epoch": 0.4349583908897948, + "grad_norm": 6.365315795081244, + "learning_rate": 6.954088675320534e-07, + "loss": 1.7969, + "step": 6455 + }, + { + "epoch": 0.43529530676190153, + "grad_norm": 5.981109456662887, + "learning_rate": 6.948674522295969e-07, + "loss": 1.8629, + "step": 6460 + }, + { + "epoch": 0.4356322226340083, + "grad_norm": 6.009688578755199, + "learning_rate": 6.943257673852993e-07, + "loss": 1.7985, + "step": 6465 + }, + { + "epoch": 0.435969138506115, + "grad_norm": 6.299881918823882, + "learning_rate": 6.937838137484225e-07, + "loss": 1.905, + "step": 6470 + }, + { + "epoch": 0.43630605437822173, + "grad_norm": 6.220203094141343, + "learning_rate": 6.932415920686001e-07, + "loss": 1.7251, + "step": 6475 + }, + { + "epoch": 0.4366429702503285, + "grad_norm": 5.711822479774969, + "learning_rate": 6.926991030958362e-07, + "loss": 1.8137, + "step": 6480 + }, + { + "epoch": 0.4369798861224352, + "grad_norm": 6.410027467003412, + "learning_rate": 6.921563475805051e-07, + "loss": 1.7679, + "step": 6485 + }, + { + "epoch": 0.437316801994542, + "grad_norm": 5.907460533318916, + "learning_rate": 6.916133262733493e-07, + "loss": 1.7704, + "step": 6490 + }, + { + "epoch": 0.4376537178666487, + "grad_norm": 6.55648674383762, + "learning_rate": 6.910700399254793e-07, + "loss": 1.836, + "step": 6495 + }, + { + "epoch": 0.4379906337387554, + "grad_norm": 5.720812174587427, + "learning_rate": 6.905264892883721e-07, + "loss": 1.7987, + "step": 6500 + }, + { + "epoch": 0.4383275496108622, + "grad_norm": 5.349790019313271, + "learning_rate": 6.899826751138701e-07, + "loss": 1.8159, + "step": 6505 + }, + { + "epoch": 0.4386644654829689, + "grad_norm": 6.281596880445255, + "learning_rate": 6.894385981541804e-07, + "loss": 1.8379, + "step": 6510 + }, + { + "epoch": 0.43900138135507566, + "grad_norm": 6.043708122899696, + "learning_rate": 6.888942591618736e-07, + "loss": 1.8159, + "step": 6515 + }, + { + "epoch": 0.4393382972271824, + "grad_norm": 6.024522211212892, + "learning_rate": 6.883496588898827e-07, + "loss": 1.7956, + "step": 6520 + }, + { + "epoch": 0.4396752130992891, + "grad_norm": 6.26133167110833, + "learning_rate": 6.87804798091502e-07, + "loss": 1.8311, + "step": 6525 + }, + { + "epoch": 0.44001212897139586, + "grad_norm": 6.350548816132551, + "learning_rate": 6.872596775203864e-07, + "loss": 1.8546, + "step": 6530 + }, + { + "epoch": 0.44034904484350257, + "grad_norm": 5.911948063541635, + "learning_rate": 6.867142979305498e-07, + "loss": 1.7843, + "step": 6535 + }, + { + "epoch": 0.44068596071560934, + "grad_norm": 5.978493138464683, + "learning_rate": 6.861686600763648e-07, + "loss": 1.8331, + "step": 6540 + }, + { + "epoch": 0.44102287658771605, + "grad_norm": 6.2087819303121154, + "learning_rate": 6.856227647125607e-07, + "loss": 1.8502, + "step": 6545 + }, + { + "epoch": 0.44135979245982276, + "grad_norm": 6.320344130956893, + "learning_rate": 6.850766125942235e-07, + "loss": 1.8667, + "step": 6550 + }, + { + "epoch": 0.44169670833192953, + "grad_norm": 6.142752431697697, + "learning_rate": 6.84530204476794e-07, + "loss": 1.8206, + "step": 6555 + }, + { + "epoch": 0.44203362420403625, + "grad_norm": 6.055474371152951, + "learning_rate": 6.839835411160673e-07, + "loss": 1.8678, + "step": 6560 + }, + { + "epoch": 0.44237054007614296, + "grad_norm": 5.817969085642228, + "learning_rate": 6.834366232681915e-07, + "loss": 1.8149, + "step": 6565 + }, + { + "epoch": 0.44270745594824973, + "grad_norm": 5.76853321731449, + "learning_rate": 6.828894516896664e-07, + "loss": 1.7588, + "step": 6570 + }, + { + "epoch": 0.44304437182035644, + "grad_norm": 6.149317072156233, + "learning_rate": 6.823420271373433e-07, + "loss": 1.8405, + "step": 6575 + }, + { + "epoch": 0.4433812876924632, + "grad_norm": 6.408757055709368, + "learning_rate": 6.817943503684232e-07, + "loss": 1.8757, + "step": 6580 + }, + { + "epoch": 0.4437182035645699, + "grad_norm": 5.7234737809931095, + "learning_rate": 6.812464221404558e-07, + "loss": 1.806, + "step": 6585 + }, + { + "epoch": 0.44405511943667664, + "grad_norm": 5.956521950229173, + "learning_rate": 6.806982432113388e-07, + "loss": 1.835, + "step": 6590 + }, + { + "epoch": 0.4443920353087834, + "grad_norm": 5.847432210874048, + "learning_rate": 6.801498143393168e-07, + "loss": 1.8316, + "step": 6595 + }, + { + "epoch": 0.4447289511808901, + "grad_norm": 6.25690887107104, + "learning_rate": 6.796011362829794e-07, + "loss": 1.8865, + "step": 6600 + }, + { + "epoch": 0.4450658670529969, + "grad_norm": 6.128754946117051, + "learning_rate": 6.790522098012621e-07, + "loss": 1.8238, + "step": 6605 + }, + { + "epoch": 0.4454027829251036, + "grad_norm": 6.087198236864975, + "learning_rate": 6.785030356534428e-07, + "loss": 1.8193, + "step": 6610 + }, + { + "epoch": 0.4457396987972103, + "grad_norm": 6.429468035527822, + "learning_rate": 6.779536145991427e-07, + "loss": 1.8451, + "step": 6615 + }, + { + "epoch": 0.4460766146693171, + "grad_norm": 6.34746638921528, + "learning_rate": 6.774039473983243e-07, + "loss": 1.7525, + "step": 6620 + }, + { + "epoch": 0.4464135305414238, + "grad_norm": 6.139901961547965, + "learning_rate": 6.768540348112906e-07, + "loss": 1.8237, + "step": 6625 + }, + { + "epoch": 0.44675044641353057, + "grad_norm": 5.66015536542511, + "learning_rate": 6.763038775986842e-07, + "loss": 1.8506, + "step": 6630 + }, + { + "epoch": 0.4470873622856373, + "grad_norm": 5.9272363359583204, + "learning_rate": 6.757534765214858e-07, + "loss": 1.8595, + "step": 6635 + }, + { + "epoch": 0.447424278157744, + "grad_norm": 6.161894844697172, + "learning_rate": 6.752028323410134e-07, + "loss": 1.7805, + "step": 6640 + }, + { + "epoch": 0.44776119402985076, + "grad_norm": 5.84449768884819, + "learning_rate": 6.746519458189214e-07, + "loss": 1.8152, + "step": 6645 + }, + { + "epoch": 0.4480981099019575, + "grad_norm": 5.657979812571092, + "learning_rate": 6.741008177171993e-07, + "loss": 1.8209, + "step": 6650 + }, + { + "epoch": 0.4484350257740642, + "grad_norm": 5.944430010944899, + "learning_rate": 6.735494487981711e-07, + "loss": 1.7561, + "step": 6655 + }, + { + "epoch": 0.44877194164617096, + "grad_norm": 5.520891888250897, + "learning_rate": 6.729978398244935e-07, + "loss": 1.7838, + "step": 6660 + }, + { + "epoch": 0.44910885751827767, + "grad_norm": 5.996486165077548, + "learning_rate": 6.724459915591551e-07, + "loss": 1.8489, + "step": 6665 + }, + { + "epoch": 0.44944577339038444, + "grad_norm": 5.733752546274423, + "learning_rate": 6.718939047654763e-07, + "loss": 1.8305, + "step": 6670 + }, + { + "epoch": 0.44978268926249115, + "grad_norm": 5.820810666342823, + "learning_rate": 6.713415802071064e-07, + "loss": 1.857, + "step": 6675 + }, + { + "epoch": 0.45011960513459787, + "grad_norm": 5.650158926744362, + "learning_rate": 6.707890186480244e-07, + "loss": 1.8326, + "step": 6680 + }, + { + "epoch": 0.45045652100670464, + "grad_norm": 5.977907228957739, + "learning_rate": 6.702362208525366e-07, + "loss": 1.7733, + "step": 6685 + }, + { + "epoch": 0.45079343687881135, + "grad_norm": 5.644462707892311, + "learning_rate": 6.696831875852763e-07, + "loss": 1.7398, + "step": 6690 + }, + { + "epoch": 0.4511303527509181, + "grad_norm": 5.5200713986530685, + "learning_rate": 6.691299196112025e-07, + "loss": 1.8783, + "step": 6695 + }, + { + "epoch": 0.45146726862302483, + "grad_norm": 5.829144396241241, + "learning_rate": 6.685764176955991e-07, + "loss": 1.77, + "step": 6700 + }, + { + "epoch": 0.45180418449513154, + "grad_norm": 5.815077874644452, + "learning_rate": 6.680226826040727e-07, + "loss": 1.8806, + "step": 6705 + }, + { + "epoch": 0.4521411003672383, + "grad_norm": 6.016668642678767, + "learning_rate": 6.674687151025535e-07, + "loss": 1.7405, + "step": 6710 + }, + { + "epoch": 0.452478016239345, + "grad_norm": 5.5436173978059315, + "learning_rate": 6.669145159572924e-07, + "loss": 1.8083, + "step": 6715 + }, + { + "epoch": 0.4528149321114518, + "grad_norm": 5.498353466010969, + "learning_rate": 6.663600859348615e-07, + "loss": 1.8109, + "step": 6720 + }, + { + "epoch": 0.4531518479835585, + "grad_norm": 5.857461806397999, + "learning_rate": 6.658054258021513e-07, + "loss": 1.8539, + "step": 6725 + }, + { + "epoch": 0.4534887638556652, + "grad_norm": 5.4613036126443335, + "learning_rate": 6.652505363263712e-07, + "loss": 1.8247, + "step": 6730 + }, + { + "epoch": 0.453825679727772, + "grad_norm": 6.035072147805387, + "learning_rate": 6.646954182750478e-07, + "loss": 1.7817, + "step": 6735 + }, + { + "epoch": 0.4541625955998787, + "grad_norm": 5.595146968290168, + "learning_rate": 6.641400724160234e-07, + "loss": 1.7759, + "step": 6740 + }, + { + "epoch": 0.4544995114719854, + "grad_norm": 5.565775929671658, + "learning_rate": 6.635844995174561e-07, + "loss": 1.8401, + "step": 6745 + }, + { + "epoch": 0.4548364273440922, + "grad_norm": 6.211211135256073, + "learning_rate": 6.630287003478176e-07, + "loss": 1.852, + "step": 6750 + }, + { + "epoch": 0.4551733432161989, + "grad_norm": 5.8516613231945, + "learning_rate": 6.624726756758927e-07, + "loss": 1.7437, + "step": 6755 + }, + { + "epoch": 0.45551025908830567, + "grad_norm": 6.200671363752341, + "learning_rate": 6.619164262707782e-07, + "loss": 1.8437, + "step": 6760 + }, + { + "epoch": 0.4558471749604124, + "grad_norm": 5.932673768489387, + "learning_rate": 6.613599529018815e-07, + "loss": 1.7869, + "step": 6765 + }, + { + "epoch": 0.4561840908325191, + "grad_norm": 6.133480058883383, + "learning_rate": 6.608032563389198e-07, + "loss": 1.8157, + "step": 6770 + }, + { + "epoch": 0.45652100670462586, + "grad_norm": 5.913849561325004, + "learning_rate": 6.602463373519196e-07, + "loss": 1.8171, + "step": 6775 + }, + { + "epoch": 0.4568579225767326, + "grad_norm": 5.5977482506304215, + "learning_rate": 6.596891967112143e-07, + "loss": 1.7908, + "step": 6780 + }, + { + "epoch": 0.45719483844883935, + "grad_norm": 5.493955784529846, + "learning_rate": 6.59131835187444e-07, + "loss": 1.7221, + "step": 6785 + }, + { + "epoch": 0.45753175432094606, + "grad_norm": 6.311459816981047, + "learning_rate": 6.58574253551555e-07, + "loss": 1.8193, + "step": 6790 + }, + { + "epoch": 0.4578686701930528, + "grad_norm": 6.415344094889377, + "learning_rate": 6.580164525747973e-07, + "loss": 1.7941, + "step": 6795 + }, + { + "epoch": 0.45820558606515954, + "grad_norm": 6.105511233378877, + "learning_rate": 6.574584330287247e-07, + "loss": 1.8424, + "step": 6800 + }, + { + "epoch": 0.45854250193726626, + "grad_norm": 5.7943538228072065, + "learning_rate": 6.569001956851932e-07, + "loss": 1.8572, + "step": 6805 + }, + { + "epoch": 0.458879417809373, + "grad_norm": 5.421853273516655, + "learning_rate": 6.563417413163601e-07, + "loss": 1.8344, + "step": 6810 + }, + { + "epoch": 0.45921633368147974, + "grad_norm": 5.637838882491133, + "learning_rate": 6.55783070694683e-07, + "loss": 1.8268, + "step": 6815 + }, + { + "epoch": 0.45955324955358645, + "grad_norm": 5.643321632821095, + "learning_rate": 6.55224184592918e-07, + "loss": 1.8258, + "step": 6820 + }, + { + "epoch": 0.4598901654256932, + "grad_norm": 5.998748203165365, + "learning_rate": 6.546650837841203e-07, + "loss": 1.8257, + "step": 6825 + }, + { + "epoch": 0.46022708129779993, + "grad_norm": 5.965057345432399, + "learning_rate": 6.541057690416414e-07, + "loss": 1.7399, + "step": 6830 + }, + { + "epoch": 0.46056399716990665, + "grad_norm": 6.1720570763462055, + "learning_rate": 6.535462411391284e-07, + "loss": 1.7994, + "step": 6835 + }, + { + "epoch": 0.4609009130420134, + "grad_norm": 5.86341018032716, + "learning_rate": 6.529865008505244e-07, + "loss": 1.799, + "step": 6840 + }, + { + "epoch": 0.46123782891412013, + "grad_norm": 6.746892080799635, + "learning_rate": 6.524265489500651e-07, + "loss": 1.8779, + "step": 6845 + }, + { + "epoch": 0.4615747447862269, + "grad_norm": 5.9098714454502295, + "learning_rate": 6.518663862122794e-07, + "loss": 1.734, + "step": 6850 + }, + { + "epoch": 0.4619116606583336, + "grad_norm": 5.735456527744707, + "learning_rate": 6.513060134119878e-07, + "loss": 1.7316, + "step": 6855 + }, + { + "epoch": 0.4622485765304403, + "grad_norm": 5.890116939620567, + "learning_rate": 6.507454313243015e-07, + "loss": 1.7896, + "step": 6860 + }, + { + "epoch": 0.4625854924025471, + "grad_norm": 5.983532007280158, + "learning_rate": 6.50184640724621e-07, + "loss": 1.8345, + "step": 6865 + }, + { + "epoch": 0.4629224082746538, + "grad_norm": 6.199632944093907, + "learning_rate": 6.496236423886351e-07, + "loss": 1.8626, + "step": 6870 + }, + { + "epoch": 0.4632593241467606, + "grad_norm": 6.237427141315285, + "learning_rate": 6.490624370923201e-07, + "loss": 1.8034, + "step": 6875 + }, + { + "epoch": 0.4635962400188673, + "grad_norm": 5.84748187391112, + "learning_rate": 6.485010256119388e-07, + "loss": 1.7816, + "step": 6880 + }, + { + "epoch": 0.463933155890974, + "grad_norm": 5.64468712531096, + "learning_rate": 6.479394087240389e-07, + "loss": 1.8375, + "step": 6885 + }, + { + "epoch": 0.46427007176308077, + "grad_norm": 6.152420602370605, + "learning_rate": 6.473775872054521e-07, + "loss": 1.8895, + "step": 6890 + }, + { + "epoch": 0.4646069876351875, + "grad_norm": 6.681683712578151, + "learning_rate": 6.468155618332936e-07, + "loss": 1.7815, + "step": 6895 + }, + { + "epoch": 0.46494390350729425, + "grad_norm": 6.095659229972454, + "learning_rate": 6.462533333849599e-07, + "loss": 1.8292, + "step": 6900 + }, + { + "epoch": 0.46528081937940097, + "grad_norm": 5.448666114017835, + "learning_rate": 6.456909026381292e-07, + "loss": 1.7838, + "step": 6905 + }, + { + "epoch": 0.4656177352515077, + "grad_norm": 6.188048616642648, + "learning_rate": 6.451282703707591e-07, + "loss": 1.7777, + "step": 6910 + }, + { + "epoch": 0.46595465112361445, + "grad_norm": 6.326139726045801, + "learning_rate": 6.445654373610854e-07, + "loss": 1.8323, + "step": 6915 + }, + { + "epoch": 0.46629156699572116, + "grad_norm": 6.10356320446364, + "learning_rate": 6.440024043876229e-07, + "loss": 1.8562, + "step": 6920 + }, + { + "epoch": 0.4666284828678279, + "grad_norm": 5.708886402068291, + "learning_rate": 6.434391722291618e-07, + "loss": 1.8621, + "step": 6925 + }, + { + "epoch": 0.46696539873993465, + "grad_norm": 6.11508671033799, + "learning_rate": 6.428757416647683e-07, + "loss": 1.7965, + "step": 6930 + }, + { + "epoch": 0.46730231461204136, + "grad_norm": 6.32816603703681, + "learning_rate": 6.42312113473783e-07, + "loss": 1.8077, + "step": 6935 + }, + { + "epoch": 0.4676392304841481, + "grad_norm": 5.96196567374692, + "learning_rate": 6.417482884358196e-07, + "loss": 1.8217, + "step": 6940 + }, + { + "epoch": 0.46797614635625484, + "grad_norm": 6.021121675554487, + "learning_rate": 6.411842673307648e-07, + "loss": 1.8155, + "step": 6945 + }, + { + "epoch": 0.46831306222836155, + "grad_norm": 6.098892855175968, + "learning_rate": 6.406200509387756e-07, + "loss": 1.7478, + "step": 6950 + }, + { + "epoch": 0.4686499781004683, + "grad_norm": 6.282127245845644, + "learning_rate": 6.400556400402796e-07, + "loss": 1.8861, + "step": 6955 + }, + { + "epoch": 0.46898689397257504, + "grad_norm": 6.018028715278134, + "learning_rate": 6.394910354159736e-07, + "loss": 1.7715, + "step": 6960 + }, + { + "epoch": 0.4693238098446818, + "grad_norm": 5.798697336927142, + "learning_rate": 6.389262378468219e-07, + "loss": 1.8519, + "step": 6965 + }, + { + "epoch": 0.4696607257167885, + "grad_norm": 5.5797942228856705, + "learning_rate": 6.38361248114056e-07, + "loss": 1.7976, + "step": 6970 + }, + { + "epoch": 0.46999764158889523, + "grad_norm": 5.6996680970531015, + "learning_rate": 6.377960669991733e-07, + "loss": 1.9102, + "step": 6975 + }, + { + "epoch": 0.470334557461002, + "grad_norm": 6.235143725814519, + "learning_rate": 6.372306952839353e-07, + "loss": 1.8289, + "step": 6980 + }, + { + "epoch": 0.4706714733331087, + "grad_norm": 5.893001299421735, + "learning_rate": 6.36665133750368e-07, + "loss": 1.8837, + "step": 6985 + }, + { + "epoch": 0.4710083892052155, + "grad_norm": 5.879461530673223, + "learning_rate": 6.360993831807593e-07, + "loss": 1.7474, + "step": 6990 + }, + { + "epoch": 0.4713453050773222, + "grad_norm": 6.002930353556234, + "learning_rate": 6.355334443576589e-07, + "loss": 1.8266, + "step": 6995 + }, + { + "epoch": 0.4716822209494289, + "grad_norm": 5.58752845843444, + "learning_rate": 6.349673180638769e-07, + "loss": 1.7796, + "step": 7000 + }, + { + "epoch": 0.4720191368215357, + "grad_norm": 6.352723281095035, + "learning_rate": 6.344010050824824e-07, + "loss": 1.8354, + "step": 7005 + }, + { + "epoch": 0.4723560526936424, + "grad_norm": 5.846163318879495, + "learning_rate": 6.338345061968032e-07, + "loss": 1.8195, + "step": 7010 + }, + { + "epoch": 0.4726929685657491, + "grad_norm": 5.884896796215211, + "learning_rate": 6.33267822190424e-07, + "loss": 1.7535, + "step": 7015 + }, + { + "epoch": 0.4730298844378559, + "grad_norm": 5.774564823273729, + "learning_rate": 6.327009538471853e-07, + "loss": 1.8374, + "step": 7020 + }, + { + "epoch": 0.4733668003099626, + "grad_norm": 6.0627901365864565, + "learning_rate": 6.321339019511828e-07, + "loss": 1.8027, + "step": 7025 + }, + { + "epoch": 0.47370371618206936, + "grad_norm": 5.788578834075243, + "learning_rate": 6.315666672867664e-07, + "loss": 1.8547, + "step": 7030 + }, + { + "epoch": 0.47404063205417607, + "grad_norm": 6.177124253642273, + "learning_rate": 6.309992506385385e-07, + "loss": 1.7444, + "step": 7035 + }, + { + "epoch": 0.4743775479262828, + "grad_norm": 5.982361591747417, + "learning_rate": 6.304316527913531e-07, + "loss": 1.7825, + "step": 7040 + }, + { + "epoch": 0.47471446379838955, + "grad_norm": 6.061226999093581, + "learning_rate": 6.29863874530315e-07, + "loss": 1.7663, + "step": 7045 + }, + { + "epoch": 0.47505137967049627, + "grad_norm": 6.222315463271125, + "learning_rate": 6.292959166407785e-07, + "loss": 1.8639, + "step": 7050 + }, + { + "epoch": 0.47538829554260303, + "grad_norm": 6.34970457479949, + "learning_rate": 6.287277799083466e-07, + "loss": 1.8037, + "step": 7055 + }, + { + "epoch": 0.47572521141470975, + "grad_norm": 6.03519967937727, + "learning_rate": 6.281594651188693e-07, + "loss": 1.7412, + "step": 7060 + }, + { + "epoch": 0.47606212728681646, + "grad_norm": 5.751601424455161, + "learning_rate": 6.275909730584431e-07, + "loss": 1.7887, + "step": 7065 + }, + { + "epoch": 0.47639904315892323, + "grad_norm": 6.115573647645315, + "learning_rate": 6.270223045134095e-07, + "loss": 1.8296, + "step": 7070 + }, + { + "epoch": 0.47673595903102994, + "grad_norm": 6.074604130360429, + "learning_rate": 6.264534602703546e-07, + "loss": 1.7821, + "step": 7075 + }, + { + "epoch": 0.4770728749031367, + "grad_norm": 6.424918913746082, + "learning_rate": 6.25884441116107e-07, + "loss": 1.865, + "step": 7080 + }, + { + "epoch": 0.4774097907752434, + "grad_norm": 5.845730128232722, + "learning_rate": 6.253152478377375e-07, + "loss": 1.797, + "step": 7085 + }, + { + "epoch": 0.47774670664735014, + "grad_norm": 5.957759182701193, + "learning_rate": 6.247458812225576e-07, + "loss": 1.8609, + "step": 7090 + }, + { + "epoch": 0.4780836225194569, + "grad_norm": 5.921689289922398, + "learning_rate": 6.241763420581188e-07, + "loss": 1.8602, + "step": 7095 + }, + { + "epoch": 0.4784205383915636, + "grad_norm": 5.610549553444218, + "learning_rate": 6.23606631132211e-07, + "loss": 1.8074, + "step": 7100 + }, + { + "epoch": 0.47875745426367033, + "grad_norm": 5.877721261294871, + "learning_rate": 6.23036749232862e-07, + "loss": 1.8379, + "step": 7105 + }, + { + "epoch": 0.4790943701357771, + "grad_norm": 6.284830816460788, + "learning_rate": 6.224666971483355e-07, + "loss": 1.7489, + "step": 7110 + }, + { + "epoch": 0.4794312860078838, + "grad_norm": 6.054222103634587, + "learning_rate": 6.218964756671315e-07, + "loss": 1.783, + "step": 7115 + }, + { + "epoch": 0.4797682018799906, + "grad_norm": 6.127182510154835, + "learning_rate": 6.213260855779834e-07, + "loss": 1.8093, + "step": 7120 + }, + { + "epoch": 0.4801051177520973, + "grad_norm": 5.902080284141082, + "learning_rate": 6.207555276698584e-07, + "loss": 1.8552, + "step": 7125 + }, + { + "epoch": 0.480442033624204, + "grad_norm": 5.945336485857451, + "learning_rate": 6.201848027319556e-07, + "loss": 1.8258, + "step": 7130 + }, + { + "epoch": 0.4807789494963108, + "grad_norm": 5.888205608422822, + "learning_rate": 6.196139115537054e-07, + "loss": 1.7935, + "step": 7135 + }, + { + "epoch": 0.4811158653684175, + "grad_norm": 6.233570018542256, + "learning_rate": 6.190428549247677e-07, + "loss": 1.8432, + "step": 7140 + }, + { + "epoch": 0.48145278124052426, + "grad_norm": 5.755763880495322, + "learning_rate": 6.184716336350316e-07, + "loss": 1.8871, + "step": 7145 + }, + { + "epoch": 0.481789697112631, + "grad_norm": 6.524975298684033, + "learning_rate": 6.179002484746137e-07, + "loss": 1.838, + "step": 7150 + }, + { + "epoch": 0.4821266129847377, + "grad_norm": 5.819047158294165, + "learning_rate": 6.173287002338577e-07, + "loss": 1.8088, + "step": 7155 + }, + { + "epoch": 0.48246352885684446, + "grad_norm": 5.975474876362582, + "learning_rate": 6.167569897033322e-07, + "loss": 1.8079, + "step": 7160 + }, + { + "epoch": 0.4828004447289512, + "grad_norm": 5.71735739384533, + "learning_rate": 6.16185117673831e-07, + "loss": 1.8762, + "step": 7165 + }, + { + "epoch": 0.48313736060105794, + "grad_norm": 6.123157625142532, + "learning_rate": 6.15613084936371e-07, + "loss": 1.8784, + "step": 7170 + }, + { + "epoch": 0.48347427647316465, + "grad_norm": 5.767255419265751, + "learning_rate": 6.150408922821911e-07, + "loss": 1.8505, + "step": 7175 + }, + { + "epoch": 0.48381119234527137, + "grad_norm": 6.395606763447789, + "learning_rate": 6.144685405027518e-07, + "loss": 1.8165, + "step": 7180 + }, + { + "epoch": 0.48414810821737814, + "grad_norm": 5.917857732608529, + "learning_rate": 6.138960303897335e-07, + "loss": 1.8577, + "step": 7185 + }, + { + "epoch": 0.48448502408948485, + "grad_norm": 6.1936529585107465, + "learning_rate": 6.133233627350355e-07, + "loss": 1.7746, + "step": 7190 + }, + { + "epoch": 0.48482193996159156, + "grad_norm": 5.564280808926514, + "learning_rate": 6.127505383307754e-07, + "loss": 1.7958, + "step": 7195 + }, + { + "epoch": 0.48515885583369833, + "grad_norm": 5.747234003023109, + "learning_rate": 6.121775579692873e-07, + "loss": 1.8808, + "step": 7200 + }, + { + "epoch": 0.48549577170580505, + "grad_norm": 6.451438702965074, + "learning_rate": 6.116044224431212e-07, + "loss": 1.8364, + "step": 7205 + }, + { + "epoch": 0.4858326875779118, + "grad_norm": 6.027079769431342, + "learning_rate": 6.110311325450416e-07, + "loss": 1.74, + "step": 7210 + }, + { + "epoch": 0.48616960345001853, + "grad_norm": 6.240032794806677, + "learning_rate": 6.104576890680263e-07, + "loss": 1.8064, + "step": 7215 + }, + { + "epoch": 0.48650651932212524, + "grad_norm": 6.510974143301528, + "learning_rate": 6.098840928052663e-07, + "loss": 1.8691, + "step": 7220 + }, + { + "epoch": 0.486843435194232, + "grad_norm": 6.165787234194551, + "learning_rate": 6.093103445501629e-07, + "loss": 1.7635, + "step": 7225 + }, + { + "epoch": 0.4871803510663387, + "grad_norm": 6.3112427035306595, + "learning_rate": 6.087364450963286e-07, + "loss": 1.8033, + "step": 7230 + }, + { + "epoch": 0.4875172669384455, + "grad_norm": 6.162132319624708, + "learning_rate": 6.081623952375843e-07, + "loss": 1.8591, + "step": 7235 + }, + { + "epoch": 0.4878541828105522, + "grad_norm": 6.168167615327447, + "learning_rate": 6.075881957679593e-07, + "loss": 1.8161, + "step": 7240 + }, + { + "epoch": 0.4881910986826589, + "grad_norm": 6.532602770115309, + "learning_rate": 6.0701384748169e-07, + "loss": 1.8449, + "step": 7245 + }, + { + "epoch": 0.4885280145547657, + "grad_norm": 6.211021003395661, + "learning_rate": 6.064393511732181e-07, + "loss": 1.8488, + "step": 7250 + }, + { + "epoch": 0.4888649304268724, + "grad_norm": 6.206539533335064, + "learning_rate": 6.058647076371906e-07, + "loss": 1.8284, + "step": 7255 + }, + { + "epoch": 0.48920184629897917, + "grad_norm": 6.036174190363687, + "learning_rate": 6.052899176684579e-07, + "loss": 1.7717, + "step": 7260 + }, + { + "epoch": 0.4895387621710859, + "grad_norm": 6.047303397207279, + "learning_rate": 6.047149820620729e-07, + "loss": 1.7952, + "step": 7265 + }, + { + "epoch": 0.4898756780431926, + "grad_norm": 6.07327877713165, + "learning_rate": 6.0413990161329e-07, + "loss": 1.8041, + "step": 7270 + }, + { + "epoch": 0.49021259391529937, + "grad_norm": 5.821598916540012, + "learning_rate": 6.035646771175642e-07, + "loss": 1.8248, + "step": 7275 + }, + { + "epoch": 0.4905495097874061, + "grad_norm": 6.079050090689548, + "learning_rate": 6.029893093705491e-07, + "loss": 1.8119, + "step": 7280 + }, + { + "epoch": 0.4908864256595128, + "grad_norm": 5.628298752995402, + "learning_rate": 6.024137991680973e-07, + "loss": 1.7406, + "step": 7285 + }, + { + "epoch": 0.49122334153161956, + "grad_norm": 5.732384094969437, + "learning_rate": 6.018381473062575e-07, + "loss": 1.8597, + "step": 7290 + }, + { + "epoch": 0.4915602574037263, + "grad_norm": 6.352928731604198, + "learning_rate": 6.012623545812754e-07, + "loss": 1.7484, + "step": 7295 + }, + { + "epoch": 0.49189717327583304, + "grad_norm": 6.042895881529252, + "learning_rate": 6.006864217895906e-07, + "loss": 1.8524, + "step": 7300 + }, + { + "epoch": 0.49223408914793976, + "grad_norm": 5.7435529216099654, + "learning_rate": 6.001103497278369e-07, + "loss": 1.8325, + "step": 7305 + }, + { + "epoch": 0.49257100502004647, + "grad_norm": 6.724766787405004, + "learning_rate": 5.995341391928408e-07, + "loss": 1.8231, + "step": 7310 + }, + { + "epoch": 0.49290792089215324, + "grad_norm": 5.889020383079155, + "learning_rate": 5.9895779098162e-07, + "loss": 1.8216, + "step": 7315 + }, + { + "epoch": 0.49324483676425995, + "grad_norm": 5.854780725466525, + "learning_rate": 5.983813058913829e-07, + "loss": 1.7734, + "step": 7320 + }, + { + "epoch": 0.4935817526363667, + "grad_norm": 5.453517656480012, + "learning_rate": 5.978046847195272e-07, + "loss": 1.8492, + "step": 7325 + }, + { + "epoch": 0.49391866850847344, + "grad_norm": 6.459742174870257, + "learning_rate": 5.97227928263639e-07, + "loss": 1.8713, + "step": 7330 + }, + { + "epoch": 0.49425558438058015, + "grad_norm": 5.857832814151433, + "learning_rate": 5.96651037321491e-07, + "loss": 1.8556, + "step": 7335 + }, + { + "epoch": 0.4945925002526869, + "grad_norm": 5.596533367239473, + "learning_rate": 5.960740126910425e-07, + "loss": 1.7902, + "step": 7340 + }, + { + "epoch": 0.49492941612479363, + "grad_norm": 6.226649318452835, + "learning_rate": 5.954968551704373e-07, + "loss": 1.8362, + "step": 7345 + }, + { + "epoch": 0.4952663319969004, + "grad_norm": 6.110264934214935, + "learning_rate": 5.949195655580032e-07, + "loss": 1.8022, + "step": 7350 + }, + { + "epoch": 0.4956032478690071, + "grad_norm": 6.618903452310688, + "learning_rate": 5.943421446522509e-07, + "loss": 1.8546, + "step": 7355 + }, + { + "epoch": 0.4959401637411138, + "grad_norm": 6.149369987083455, + "learning_rate": 5.93764593251872e-07, + "loss": 1.7979, + "step": 7360 + }, + { + "epoch": 0.4962770796132206, + "grad_norm": 6.8812512541470685, + "learning_rate": 5.931869121557397e-07, + "loss": 1.8307, + "step": 7365 + }, + { + "epoch": 0.4966139954853273, + "grad_norm": 6.040974955829878, + "learning_rate": 5.926091021629055e-07, + "loss": 1.7826, + "step": 7370 + }, + { + "epoch": 0.496950911357434, + "grad_norm": 5.937707008894662, + "learning_rate": 5.920311640726e-07, + "loss": 1.7873, + "step": 7375 + }, + { + "epoch": 0.4972878272295408, + "grad_norm": 6.470436151813095, + "learning_rate": 5.914530986842307e-07, + "loss": 1.8437, + "step": 7380 + }, + { + "epoch": 0.4976247431016475, + "grad_norm": 6.3141277294968186, + "learning_rate": 5.908749067973809e-07, + "loss": 1.8271, + "step": 7385 + }, + { + "epoch": 0.4979616589737543, + "grad_norm": 5.903283938829289, + "learning_rate": 5.902965892118093e-07, + "loss": 1.8597, + "step": 7390 + }, + { + "epoch": 0.498298574845861, + "grad_norm": 5.773979245288634, + "learning_rate": 5.89718146727448e-07, + "loss": 1.8186, + "step": 7395 + }, + { + "epoch": 0.4986354907179677, + "grad_norm": 5.905823787046488, + "learning_rate": 5.891395801444026e-07, + "loss": 1.8116, + "step": 7400 + }, + { + "epoch": 0.49897240659007447, + "grad_norm": 6.486666074459213, + "learning_rate": 5.885608902629496e-07, + "loss": 1.7862, + "step": 7405 + }, + { + "epoch": 0.4993093224621812, + "grad_norm": 6.126712963890818, + "learning_rate": 5.879820778835364e-07, + "loss": 1.8475, + "step": 7410 + }, + { + "epoch": 0.49964623833428795, + "grad_norm": 6.769072318754673, + "learning_rate": 5.874031438067799e-07, + "loss": 1.8015, + "step": 7415 + }, + { + "epoch": 0.49998315420639466, + "grad_norm": 6.260585391210161, + "learning_rate": 5.868240888334652e-07, + "loss": 1.8085, + "step": 7420 + }, + { + "epoch": 0.5003200700785014, + "grad_norm": 6.346729595046292, + "learning_rate": 5.862449137645444e-07, + "loss": 1.8223, + "step": 7425 + }, + { + "epoch": 0.5006569859506081, + "grad_norm": 5.907134243732254, + "learning_rate": 5.856656194011365e-07, + "loss": 1.7711, + "step": 7430 + }, + { + "epoch": 0.5009939018227149, + "grad_norm": 6.063402006959808, + "learning_rate": 5.850862065445243e-07, + "loss": 1.7685, + "step": 7435 + }, + { + "epoch": 0.5013308176948216, + "grad_norm": 5.890743753183375, + "learning_rate": 5.845066759961557e-07, + "loss": 1.8192, + "step": 7440 + }, + { + "epoch": 0.5016677335669283, + "grad_norm": 5.660746063214769, + "learning_rate": 5.839270285576407e-07, + "loss": 1.8722, + "step": 7445 + }, + { + "epoch": 0.5020046494390351, + "grad_norm": 5.733382666328982, + "learning_rate": 5.833472650307509e-07, + "loss": 1.7618, + "step": 7450 + }, + { + "epoch": 0.5023415653111418, + "grad_norm": 5.958776596261261, + "learning_rate": 5.827673862174192e-07, + "loss": 1.873, + "step": 7455 + }, + { + "epoch": 0.5026784811832485, + "grad_norm": 5.773391012038019, + "learning_rate": 5.821873929197371e-07, + "loss": 1.8226, + "step": 7460 + }, + { + "epoch": 0.5030153970553553, + "grad_norm": 6.876209726907723, + "learning_rate": 5.81607285939955e-07, + "loss": 1.7926, + "step": 7465 + }, + { + "epoch": 0.503352312927462, + "grad_norm": 6.83351949530394, + "learning_rate": 5.810270660804805e-07, + "loss": 1.8214, + "step": 7470 + }, + { + "epoch": 0.5036892287995688, + "grad_norm": 5.695338110099444, + "learning_rate": 5.80446734143877e-07, + "loss": 1.7878, + "step": 7475 + }, + { + "epoch": 0.5040261446716755, + "grad_norm": 6.321842315715411, + "learning_rate": 5.798662909328633e-07, + "loss": 1.7732, + "step": 7480 + }, + { + "epoch": 0.5043630605437822, + "grad_norm": 5.8003219342595695, + "learning_rate": 5.792857372503119e-07, + "loss": 1.8207, + "step": 7485 + }, + { + "epoch": 0.5046999764158889, + "grad_norm": 6.206954096469576, + "learning_rate": 5.787050738992481e-07, + "loss": 1.8191, + "step": 7490 + }, + { + "epoch": 0.5050368922879956, + "grad_norm": 6.3174541848095345, + "learning_rate": 5.781243016828492e-07, + "loss": 1.8286, + "step": 7495 + }, + { + "epoch": 0.5053738081601025, + "grad_norm": 5.943268852990081, + "learning_rate": 5.775434214044427e-07, + "loss": 1.8347, + "step": 7500 + }, + { + "epoch": 0.5057107240322092, + "grad_norm": 6.366552386945912, + "learning_rate": 5.769624338675057e-07, + "loss": 1.8694, + "step": 7505 + }, + { + "epoch": 0.5060476399043159, + "grad_norm": 6.09106660535801, + "learning_rate": 5.763813398756637e-07, + "loss": 1.8571, + "step": 7510 + }, + { + "epoch": 0.5063845557764226, + "grad_norm": 5.927983640648523, + "learning_rate": 5.758001402326895e-07, + "loss": 1.8495, + "step": 7515 + }, + { + "epoch": 0.5067214716485293, + "grad_norm": 6.347023539008799, + "learning_rate": 5.752188357425019e-07, + "loss": 1.8598, + "step": 7520 + }, + { + "epoch": 0.5070583875206361, + "grad_norm": 5.7220954371710535, + "learning_rate": 5.746374272091648e-07, + "loss": 1.9033, + "step": 7525 + }, + { + "epoch": 0.5073953033927429, + "grad_norm": 5.716357729963673, + "learning_rate": 5.74055915436886e-07, + "loss": 1.8422, + "step": 7530 + }, + { + "epoch": 0.5077322192648496, + "grad_norm": 5.888138683912625, + "learning_rate": 5.734743012300162e-07, + "loss": 1.7706, + "step": 7535 + }, + { + "epoch": 0.5080691351369563, + "grad_norm": 6.074956239497393, + "learning_rate": 5.728925853930475e-07, + "loss": 1.7751, + "step": 7540 + }, + { + "epoch": 0.508406051009063, + "grad_norm": 6.30224985754695, + "learning_rate": 5.72310768730613e-07, + "loss": 1.8185, + "step": 7545 + }, + { + "epoch": 0.5087429668811698, + "grad_norm": 6.273905491547863, + "learning_rate": 5.717288520474849e-07, + "loss": 1.8798, + "step": 7550 + }, + { + "epoch": 0.5090798827532765, + "grad_norm": 5.877838326642534, + "learning_rate": 5.711468361485739e-07, + "loss": 1.798, + "step": 7555 + }, + { + "epoch": 0.5094167986253832, + "grad_norm": 5.551724826009454, + "learning_rate": 5.70564721838928e-07, + "loss": 1.8133, + "step": 7560 + }, + { + "epoch": 0.50975371449749, + "grad_norm": 6.386798899011169, + "learning_rate": 5.69982509923731e-07, + "loss": 1.8332, + "step": 7565 + }, + { + "epoch": 0.5100906303695967, + "grad_norm": 6.417824627395978, + "learning_rate": 5.694002012083022e-07, + "loss": 1.8929, + "step": 7570 + }, + { + "epoch": 0.5104275462417035, + "grad_norm": 5.731757336532826, + "learning_rate": 5.688177964980946e-07, + "loss": 1.7638, + "step": 7575 + }, + { + "epoch": 0.5107644621138102, + "grad_norm": 5.918225192432386, + "learning_rate": 5.682352965986935e-07, + "loss": 1.7946, + "step": 7580 + }, + { + "epoch": 0.5111013779859169, + "grad_norm": 5.849000801521007, + "learning_rate": 5.676527023158169e-07, + "loss": 1.8941, + "step": 7585 + }, + { + "epoch": 0.5114382938580236, + "grad_norm": 6.564241414658141, + "learning_rate": 5.670700144553122e-07, + "loss": 1.8329, + "step": 7590 + }, + { + "epoch": 0.5117752097301304, + "grad_norm": 6.08295990900783, + "learning_rate": 5.664872338231571e-07, + "loss": 1.8346, + "step": 7595 + }, + { + "epoch": 0.5121121256022371, + "grad_norm": 5.805622916909698, + "learning_rate": 5.659043612254573e-07, + "loss": 1.8188, + "step": 7600 + }, + { + "epoch": 0.5124490414743439, + "grad_norm": 6.197266946989277, + "learning_rate": 5.653213974684455e-07, + "loss": 1.7496, + "step": 7605 + }, + { + "epoch": 0.5127859573464506, + "grad_norm": 5.924833947325567, + "learning_rate": 5.647383433584807e-07, + "loss": 1.7737, + "step": 7610 + }, + { + "epoch": 0.5131228732185573, + "grad_norm": 6.03776739730078, + "learning_rate": 5.641551997020472e-07, + "loss": 1.8332, + "step": 7615 + }, + { + "epoch": 0.513459789090664, + "grad_norm": 6.1153821883699955, + "learning_rate": 5.635719673057524e-07, + "loss": 1.8323, + "step": 7620 + }, + { + "epoch": 0.5137967049627707, + "grad_norm": 5.832380420575743, + "learning_rate": 5.629886469763273e-07, + "loss": 1.7999, + "step": 7625 + }, + { + "epoch": 0.5141336208348776, + "grad_norm": 5.859611583687684, + "learning_rate": 5.624052395206239e-07, + "loss": 1.8414, + "step": 7630 + }, + { + "epoch": 0.5144705367069843, + "grad_norm": 6.172500864509078, + "learning_rate": 5.618217457456151e-07, + "loss": 1.7948, + "step": 7635 + }, + { + "epoch": 0.514807452579091, + "grad_norm": 5.9844119060841825, + "learning_rate": 5.612381664583928e-07, + "loss": 1.8405, + "step": 7640 + }, + { + "epoch": 0.5151443684511977, + "grad_norm": 5.884734313121189, + "learning_rate": 5.606545024661674e-07, + "loss": 1.8402, + "step": 7645 + }, + { + "epoch": 0.5154812843233044, + "grad_norm": 5.697399014618891, + "learning_rate": 5.600707545762667e-07, + "loss": 1.8065, + "step": 7650 + }, + { + "epoch": 0.5158182001954112, + "grad_norm": 5.964664311558702, + "learning_rate": 5.594869235961342e-07, + "loss": 1.849, + "step": 7655 + }, + { + "epoch": 0.516155116067518, + "grad_norm": 5.833081179358123, + "learning_rate": 5.589030103333282e-07, + "loss": 1.8678, + "step": 7660 + }, + { + "epoch": 0.5164920319396247, + "grad_norm": 5.943839246531296, + "learning_rate": 5.583190155955215e-07, + "loss": 1.7551, + "step": 7665 + }, + { + "epoch": 0.5168289478117314, + "grad_norm": 6.184069754194008, + "learning_rate": 5.57734940190499e-07, + "loss": 1.779, + "step": 7670 + }, + { + "epoch": 0.5171658636838381, + "grad_norm": 5.988286442003116, + "learning_rate": 5.571507849261572e-07, + "loss": 1.7786, + "step": 7675 + }, + { + "epoch": 0.5175027795559449, + "grad_norm": 6.104875157714636, + "learning_rate": 5.565665506105035e-07, + "loss": 1.7767, + "step": 7680 + }, + { + "epoch": 0.5178396954280516, + "grad_norm": 5.992583843253686, + "learning_rate": 5.559822380516539e-07, + "loss": 1.8409, + "step": 7685 + }, + { + "epoch": 0.5181766113001584, + "grad_norm": 6.788235374453657, + "learning_rate": 5.553978480578335e-07, + "loss": 1.797, + "step": 7690 + }, + { + "epoch": 0.5185135271722651, + "grad_norm": 5.480013162835297, + "learning_rate": 5.548133814373738e-07, + "loss": 1.8014, + "step": 7695 + }, + { + "epoch": 0.5188504430443718, + "grad_norm": 6.2047091655648074, + "learning_rate": 5.542288389987128e-07, + "loss": 1.8545, + "step": 7700 + }, + { + "epoch": 0.5191873589164786, + "grad_norm": 5.749872893831946, + "learning_rate": 5.536442215503929e-07, + "loss": 1.8309, + "step": 7705 + }, + { + "epoch": 0.5195242747885853, + "grad_norm": 6.013603014933779, + "learning_rate": 5.530595299010606e-07, + "loss": 1.7978, + "step": 7710 + }, + { + "epoch": 0.519861190660692, + "grad_norm": 6.665749598193633, + "learning_rate": 5.524747648594651e-07, + "loss": 1.8907, + "step": 7715 + }, + { + "epoch": 0.5201981065327987, + "grad_norm": 5.806706969722841, + "learning_rate": 5.518899272344568e-07, + "loss": 1.8127, + "step": 7720 + }, + { + "epoch": 0.5205350224049055, + "grad_norm": 5.8100736277669816, + "learning_rate": 5.513050178349866e-07, + "loss": 1.8411, + "step": 7725 + }, + { + "epoch": 0.5208719382770123, + "grad_norm": 6.2881327818412815, + "learning_rate": 5.507200374701048e-07, + "loss": 1.8624, + "step": 7730 + }, + { + "epoch": 0.521208854149119, + "grad_norm": 5.809423166099506, + "learning_rate": 5.501349869489596e-07, + "loss": 1.897, + "step": 7735 + }, + { + "epoch": 0.5215457700212257, + "grad_norm": 6.093243605312457, + "learning_rate": 5.495498670807967e-07, + "loss": 1.812, + "step": 7740 + }, + { + "epoch": 0.5218826858933324, + "grad_norm": 6.223494857280929, + "learning_rate": 5.489646786749574e-07, + "loss": 1.8101, + "step": 7745 + }, + { + "epoch": 0.5222196017654391, + "grad_norm": 6.071573078294997, + "learning_rate": 5.483794225408777e-07, + "loss": 1.8178, + "step": 7750 + }, + { + "epoch": 0.522556517637546, + "grad_norm": 5.693615324631069, + "learning_rate": 5.477940994880877e-07, + "loss": 1.6893, + "step": 7755 + }, + { + "epoch": 0.5228934335096527, + "grad_norm": 5.909887730062095, + "learning_rate": 5.472087103262094e-07, + "loss": 1.8432, + "step": 7760 + }, + { + "epoch": 0.5232303493817594, + "grad_norm": 5.870756025959482, + "learning_rate": 5.46623255864957e-07, + "loss": 1.819, + "step": 7765 + }, + { + "epoch": 0.5235672652538661, + "grad_norm": 5.649860781638532, + "learning_rate": 5.460377369141345e-07, + "loss": 1.8674, + "step": 7770 + }, + { + "epoch": 0.5239041811259728, + "grad_norm": 6.094869124377087, + "learning_rate": 5.454521542836351e-07, + "loss": 1.7859, + "step": 7775 + }, + { + "epoch": 0.5242410969980795, + "grad_norm": 6.4473743410813285, + "learning_rate": 5.448665087834405e-07, + "loss": 1.7703, + "step": 7780 + }, + { + "epoch": 0.5245780128701863, + "grad_norm": 6.560011859707494, + "learning_rate": 5.442808012236192e-07, + "loss": 1.7882, + "step": 7785 + }, + { + "epoch": 0.5249149287422931, + "grad_norm": 5.970137427490027, + "learning_rate": 5.436950324143251e-07, + "loss": 1.8465, + "step": 7790 + }, + { + "epoch": 0.5252518446143998, + "grad_norm": 5.925088761392246, + "learning_rate": 5.431092031657973e-07, + "loss": 1.7958, + "step": 7795 + }, + { + "epoch": 0.5255887604865065, + "grad_norm": 5.802514268517773, + "learning_rate": 5.425233142883585e-07, + "loss": 1.8614, + "step": 7800 + }, + { + "epoch": 0.5259256763586132, + "grad_norm": 5.855082730998635, + "learning_rate": 5.419373665924136e-07, + "loss": 1.7962, + "step": 7805 + }, + { + "epoch": 0.52626259223072, + "grad_norm": 5.429074232700685, + "learning_rate": 5.413513608884491e-07, + "loss": 1.8338, + "step": 7810 + }, + { + "epoch": 0.5265995081028267, + "grad_norm": 5.6742030630773055, + "learning_rate": 5.407652979870315e-07, + "loss": 1.8029, + "step": 7815 + }, + { + "epoch": 0.5269364239749335, + "grad_norm": 5.908502877639631, + "learning_rate": 5.401791786988068e-07, + "loss": 1.8193, + "step": 7820 + }, + { + "epoch": 0.5272733398470402, + "grad_norm": 5.524188749937658, + "learning_rate": 5.395930038344986e-07, + "loss": 1.8287, + "step": 7825 + }, + { + "epoch": 0.5276102557191469, + "grad_norm": 5.966495919240401, + "learning_rate": 5.390067742049073e-07, + "loss": 1.7747, + "step": 7830 + }, + { + "epoch": 0.5279471715912537, + "grad_norm": 6.028530102625983, + "learning_rate": 5.384204906209097e-07, + "loss": 1.7576, + "step": 7835 + }, + { + "epoch": 0.5282840874633604, + "grad_norm": 5.612125576553678, + "learning_rate": 5.378341538934566e-07, + "loss": 1.7362, + "step": 7840 + }, + { + "epoch": 0.5286210033354671, + "grad_norm": 6.444931841113274, + "learning_rate": 5.372477648335725e-07, + "loss": 1.8254, + "step": 7845 + }, + { + "epoch": 0.5289579192075738, + "grad_norm": 6.079946107054406, + "learning_rate": 5.366613242523544e-07, + "loss": 1.799, + "step": 7850 + }, + { + "epoch": 0.5292948350796806, + "grad_norm": 5.9372575273559365, + "learning_rate": 5.360748329609702e-07, + "loss": 1.8219, + "step": 7855 + }, + { + "epoch": 0.5296317509517874, + "grad_norm": 5.73892160169138, + "learning_rate": 5.354882917706586e-07, + "loss": 1.8598, + "step": 7860 + }, + { + "epoch": 0.5299686668238941, + "grad_norm": 6.080503146116818, + "learning_rate": 5.349017014927267e-07, + "loss": 1.8695, + "step": 7865 + }, + { + "epoch": 0.5303055826960008, + "grad_norm": 5.624633270336337, + "learning_rate": 5.343150629385496e-07, + "loss": 1.8171, + "step": 7870 + }, + { + "epoch": 0.5306424985681075, + "grad_norm": 6.59293279441881, + "learning_rate": 5.337283769195696e-07, + "loss": 1.7082, + "step": 7875 + }, + { + "epoch": 0.5309794144402142, + "grad_norm": 6.030473422245009, + "learning_rate": 5.331416442472941e-07, + "loss": 1.7242, + "step": 7880 + }, + { + "epoch": 0.5313163303123211, + "grad_norm": 6.394202607617881, + "learning_rate": 5.325548657332956e-07, + "loss": 1.7612, + "step": 7885 + }, + { + "epoch": 0.5316532461844278, + "grad_norm": 5.85821715132492, + "learning_rate": 5.319680421892095e-07, + "loss": 1.7571, + "step": 7890 + }, + { + "epoch": 0.5319901620565345, + "grad_norm": 6.318447507037375, + "learning_rate": 5.313811744267336e-07, + "loss": 1.8143, + "step": 7895 + }, + { + "epoch": 0.5323270779286412, + "grad_norm": 6.31265305059435, + "learning_rate": 5.30794263257627e-07, + "loss": 1.8149, + "step": 7900 + }, + { + "epoch": 0.5326639938007479, + "grad_norm": 5.780708946571393, + "learning_rate": 5.302073094937089e-07, + "loss": 1.7723, + "step": 7905 + }, + { + "epoch": 0.5330009096728547, + "grad_norm": 5.855290550087316, + "learning_rate": 5.296203139468571e-07, + "loss": 1.8075, + "step": 7910 + }, + { + "epoch": 0.5333378255449615, + "grad_norm": 5.943977721142977, + "learning_rate": 5.290332774290077e-07, + "loss": 1.7952, + "step": 7915 + }, + { + "epoch": 0.5336747414170682, + "grad_norm": 6.036595542010709, + "learning_rate": 5.284462007521528e-07, + "loss": 1.8388, + "step": 7920 + }, + { + "epoch": 0.5340116572891749, + "grad_norm": 5.683505340786709, + "learning_rate": 5.278590847283407e-07, + "loss": 1.7796, + "step": 7925 + }, + { + "epoch": 0.5343485731612816, + "grad_norm": 5.803826687134672, + "learning_rate": 5.27271930169674e-07, + "loss": 1.7972, + "step": 7930 + }, + { + "epoch": 0.5346854890333884, + "grad_norm": 6.286942676524797, + "learning_rate": 5.266847378883079e-07, + "loss": 1.8071, + "step": 7935 + }, + { + "epoch": 0.5350224049054951, + "grad_norm": 6.13011871952327, + "learning_rate": 5.260975086964507e-07, + "loss": 1.828, + "step": 7940 + }, + { + "epoch": 0.5353593207776018, + "grad_norm": 6.465750035776844, + "learning_rate": 5.255102434063612e-07, + "loss": 1.8118, + "step": 7945 + }, + { + "epoch": 0.5356962366497086, + "grad_norm": 6.173065557365732, + "learning_rate": 5.249229428303486e-07, + "loss": 1.8468, + "step": 7950 + }, + { + "epoch": 0.5360331525218153, + "grad_norm": 6.419517709100149, + "learning_rate": 5.243356077807704e-07, + "loss": 1.8322, + "step": 7955 + }, + { + "epoch": 0.536370068393922, + "grad_norm": 5.558598585066808, + "learning_rate": 5.237482390700319e-07, + "loss": 1.8718, + "step": 7960 + }, + { + "epoch": 0.5367069842660288, + "grad_norm": 5.719795797419367, + "learning_rate": 5.231608375105852e-07, + "loss": 1.7335, + "step": 7965 + }, + { + "epoch": 0.5370439001381355, + "grad_norm": 5.539968021502532, + "learning_rate": 5.225734039149277e-07, + "loss": 1.8108, + "step": 7970 + }, + { + "epoch": 0.5373808160102422, + "grad_norm": 5.609804958042693, + "learning_rate": 5.219859390956012e-07, + "loss": 1.7208, + "step": 7975 + }, + { + "epoch": 0.537717731882349, + "grad_norm": 6.023459731644414, + "learning_rate": 5.213984438651904e-07, + "loss": 1.7977, + "step": 7980 + }, + { + "epoch": 0.5380546477544557, + "grad_norm": 6.35720441432873, + "learning_rate": 5.208109190363222e-07, + "loss": 1.8483, + "step": 7985 + }, + { + "epoch": 0.5383915636265625, + "grad_norm": 6.258031001258222, + "learning_rate": 5.202233654216649e-07, + "loss": 1.8099, + "step": 7990 + }, + { + "epoch": 0.5387284794986692, + "grad_norm": 6.416922084645679, + "learning_rate": 5.196357838339259e-07, + "loss": 1.7425, + "step": 7995 + }, + { + "epoch": 0.5390653953707759, + "grad_norm": 5.6567156518400346, + "learning_rate": 5.190481750858516e-07, + "loss": 1.8003, + "step": 8000 + }, + { + "epoch": 0.5394023112428826, + "grad_norm": 6.3126246059511, + "learning_rate": 5.184605399902262e-07, + "loss": 1.8061, + "step": 8005 + }, + { + "epoch": 0.5397392271149893, + "grad_norm": 5.56295908767564, + "learning_rate": 5.178728793598699e-07, + "loss": 1.8358, + "step": 8010 + }, + { + "epoch": 0.5400761429870962, + "grad_norm": 6.186318540703496, + "learning_rate": 5.172851940076387e-07, + "loss": 1.8142, + "step": 8015 + }, + { + "epoch": 0.5404130588592029, + "grad_norm": 6.2506892198636255, + "learning_rate": 5.166974847464223e-07, + "loss": 1.8905, + "step": 8020 + }, + { + "epoch": 0.5407499747313096, + "grad_norm": 6.313235579462093, + "learning_rate": 5.161097523891437e-07, + "loss": 1.8203, + "step": 8025 + }, + { + "epoch": 0.5410868906034163, + "grad_norm": 6.171835646132762, + "learning_rate": 5.15521997748758e-07, + "loss": 1.8299, + "step": 8030 + }, + { + "epoch": 0.541423806475523, + "grad_norm": 5.409317737701678, + "learning_rate": 5.149342216382511e-07, + "loss": 1.7366, + "step": 8035 + }, + { + "epoch": 0.5417607223476298, + "grad_norm": 5.9172269055877615, + "learning_rate": 5.143464248706381e-07, + "loss": 1.7989, + "step": 8040 + }, + { + "epoch": 0.5420976382197366, + "grad_norm": 6.3600919594870335, + "learning_rate": 5.137586082589633e-07, + "loss": 1.8075, + "step": 8045 + }, + { + "epoch": 0.5424345540918433, + "grad_norm": 7.187853592405848, + "learning_rate": 5.131707726162983e-07, + "loss": 1.8324, + "step": 8050 + }, + { + "epoch": 0.54277146996395, + "grad_norm": 6.1090388109161635, + "learning_rate": 5.125829187557406e-07, + "loss": 1.7967, + "step": 8055 + }, + { + "epoch": 0.5431083858360567, + "grad_norm": 5.638278745047237, + "learning_rate": 5.119950474904137e-07, + "loss": 1.8695, + "step": 8060 + }, + { + "epoch": 0.5434453017081635, + "grad_norm": 5.663124594313191, + "learning_rate": 5.114071596334642e-07, + "loss": 1.7512, + "step": 8065 + }, + { + "epoch": 0.5437822175802702, + "grad_norm": 6.2665370581477795, + "learning_rate": 5.108192559980623e-07, + "loss": 1.844, + "step": 8070 + }, + { + "epoch": 0.544119133452377, + "grad_norm": 5.931434130921906, + "learning_rate": 5.102313373974e-07, + "loss": 1.876, + "step": 8075 + }, + { + "epoch": 0.5444560493244837, + "grad_norm": 5.980178228411234, + "learning_rate": 5.096434046446898e-07, + "loss": 1.7887, + "step": 8080 + }, + { + "epoch": 0.5447929651965904, + "grad_norm": 5.3738710489491295, + "learning_rate": 5.090554585531639e-07, + "loss": 1.7753, + "step": 8085 + }, + { + "epoch": 0.5451298810686972, + "grad_norm": 6.216952883756544, + "learning_rate": 5.084674999360729e-07, + "loss": 1.7961, + "step": 8090 + }, + { + "epoch": 0.5454667969408039, + "grad_norm": 6.152275028612343, + "learning_rate": 5.078795296066846e-07, + "loss": 1.8922, + "step": 8095 + }, + { + "epoch": 0.5458037128129106, + "grad_norm": 5.72793787759331, + "learning_rate": 5.072915483782833e-07, + "loss": 1.7049, + "step": 8100 + }, + { + "epoch": 0.5461406286850173, + "grad_norm": 5.9020152342288945, + "learning_rate": 5.067035570641678e-07, + "loss": 1.8247, + "step": 8105 + }, + { + "epoch": 0.546477544557124, + "grad_norm": 6.035798593915844, + "learning_rate": 5.061155564776517e-07, + "loss": 1.807, + "step": 8110 + }, + { + "epoch": 0.5468144604292309, + "grad_norm": 5.672397446189978, + "learning_rate": 5.055275474320609e-07, + "loss": 1.8352, + "step": 8115 + }, + { + "epoch": 0.5471513763013376, + "grad_norm": 6.009849527428413, + "learning_rate": 5.049395307407328e-07, + "loss": 1.8354, + "step": 8120 + }, + { + "epoch": 0.5474882921734443, + "grad_norm": 5.699405940003927, + "learning_rate": 5.04351507217016e-07, + "loss": 1.7494, + "step": 8125 + }, + { + "epoch": 0.547825208045551, + "grad_norm": 6.144976474040944, + "learning_rate": 5.03763477674268e-07, + "loss": 1.7602, + "step": 8130 + }, + { + "epoch": 0.5481621239176577, + "grad_norm": 6.340348103559182, + "learning_rate": 5.031754429258549e-07, + "loss": 1.8694, + "step": 8135 + }, + { + "epoch": 0.5484990397897644, + "grad_norm": 6.0788344731363555, + "learning_rate": 5.025874037851499e-07, + "loss": 1.8078, + "step": 8140 + }, + { + "epoch": 0.5488359556618713, + "grad_norm": 5.593092481508395, + "learning_rate": 5.019993610655322e-07, + "loss": 1.8018, + "step": 8145 + }, + { + "epoch": 0.549172871533978, + "grad_norm": 6.096728640322998, + "learning_rate": 5.014113155803863e-07, + "loss": 1.8155, + "step": 8150 + }, + { + "epoch": 0.5495097874060847, + "grad_norm": 5.870453527758819, + "learning_rate": 5.008232681430999e-07, + "loss": 1.8291, + "step": 8155 + }, + { + "epoch": 0.5498467032781914, + "grad_norm": 5.891573814310423, + "learning_rate": 5.002352195670643e-07, + "loss": 1.6946, + "step": 8160 + }, + { + "epoch": 0.5501836191502981, + "grad_norm": 6.217148650257244, + "learning_rate": 4.996471706656715e-07, + "loss": 1.7996, + "step": 8165 + }, + { + "epoch": 0.5505205350224049, + "grad_norm": 6.003361257252515, + "learning_rate": 4.990591222523142e-07, + "loss": 1.8232, + "step": 8170 + }, + { + "epoch": 0.5508574508945117, + "grad_norm": 5.517791176689642, + "learning_rate": 4.984710751403849e-07, + "loss": 1.7899, + "step": 8175 + }, + { + "epoch": 0.5511943667666184, + "grad_norm": 5.751512833055983, + "learning_rate": 4.978830301432738e-07, + "loss": 1.8628, + "step": 8180 + }, + { + "epoch": 0.5515312826387251, + "grad_norm": 6.65622502830475, + "learning_rate": 4.97294988074368e-07, + "loss": 1.8543, + "step": 8185 + }, + { + "epoch": 0.5518681985108318, + "grad_norm": 5.9681636803859766, + "learning_rate": 4.96706949747051e-07, + "loss": 1.755, + "step": 8190 + }, + { + "epoch": 0.5522051143829386, + "grad_norm": 5.96613622311191, + "learning_rate": 4.961189159747015e-07, + "loss": 1.8139, + "step": 8195 + }, + { + "epoch": 0.5525420302550453, + "grad_norm": 5.939980712328125, + "learning_rate": 4.955308875706905e-07, + "loss": 1.8386, + "step": 8200 + }, + { + "epoch": 0.552878946127152, + "grad_norm": 6.453800027528329, + "learning_rate": 4.94942865348383e-07, + "loss": 1.8015, + "step": 8205 + }, + { + "epoch": 0.5532158619992588, + "grad_norm": 5.962843605569349, + "learning_rate": 4.943548501211351e-07, + "loss": 1.8115, + "step": 8210 + }, + { + "epoch": 0.5535527778713655, + "grad_norm": 5.894927809262481, + "learning_rate": 4.937668427022924e-07, + "loss": 1.7926, + "step": 8215 + }, + { + "epoch": 0.5538896937434723, + "grad_norm": 5.923480477608677, + "learning_rate": 4.931788439051909e-07, + "loss": 1.7601, + "step": 8220 + }, + { + "epoch": 0.554226609615579, + "grad_norm": 6.050904501779012, + "learning_rate": 4.925908545431537e-07, + "loss": 1.7591, + "step": 8225 + }, + { + "epoch": 0.5545635254876857, + "grad_norm": 6.3323162814561815, + "learning_rate": 4.920028754294915e-07, + "loss": 1.7958, + "step": 8230 + }, + { + "epoch": 0.5549004413597924, + "grad_norm": 6.318956138145999, + "learning_rate": 4.914149073775003e-07, + "loss": 1.8323, + "step": 8235 + }, + { + "epoch": 0.5552373572318992, + "grad_norm": 5.959228809440483, + "learning_rate": 4.908269512004613e-07, + "loss": 1.7767, + "step": 8240 + }, + { + "epoch": 0.555574273104006, + "grad_norm": 6.102020489729409, + "learning_rate": 4.902390077116392e-07, + "loss": 1.7939, + "step": 8245 + }, + { + "epoch": 0.5559111889761127, + "grad_norm": 6.357928982728381, + "learning_rate": 4.896510777242805e-07, + "loss": 1.753, + "step": 8250 + }, + { + "epoch": 0.5562481048482194, + "grad_norm": 6.131919994254941, + "learning_rate": 4.890631620516141e-07, + "loss": 1.7993, + "step": 8255 + }, + { + "epoch": 0.5565850207203261, + "grad_norm": 6.255342386981415, + "learning_rate": 4.88475261506848e-07, + "loss": 1.771, + "step": 8260 + }, + { + "epoch": 0.5569219365924328, + "grad_norm": 6.11740983610477, + "learning_rate": 4.878873769031702e-07, + "loss": 1.8473, + "step": 8265 + }, + { + "epoch": 0.5572588524645397, + "grad_norm": 5.82647949361438, + "learning_rate": 4.872995090537459e-07, + "loss": 1.7722, + "step": 8270 + }, + { + "epoch": 0.5575957683366464, + "grad_norm": 6.220989210945246, + "learning_rate": 4.867116587717179e-07, + "loss": 1.8331, + "step": 8275 + }, + { + "epoch": 0.5579326842087531, + "grad_norm": 5.494470821652724, + "learning_rate": 4.861238268702039e-07, + "loss": 1.798, + "step": 8280 + }, + { + "epoch": 0.5582696000808598, + "grad_norm": 6.406158728808393, + "learning_rate": 4.855360141622965e-07, + "loss": 1.8354, + "step": 8285 + }, + { + "epoch": 0.5586065159529665, + "grad_norm": 5.482861394950503, + "learning_rate": 4.849482214610623e-07, + "loss": 1.8003, + "step": 8290 + }, + { + "epoch": 0.5589434318250733, + "grad_norm": 5.867393207721527, + "learning_rate": 4.843604495795392e-07, + "loss": 1.8327, + "step": 8295 + }, + { + "epoch": 0.55928034769718, + "grad_norm": 6.322610690410823, + "learning_rate": 4.83772699330737e-07, + "loss": 1.7719, + "step": 8300 + }, + { + "epoch": 0.5596172635692868, + "grad_norm": 6.042827407314324, + "learning_rate": 4.831849715276355e-07, + "loss": 1.7846, + "step": 8305 + }, + { + "epoch": 0.5599541794413935, + "grad_norm": 6.449090524228536, + "learning_rate": 4.825972669831834e-07, + "loss": 1.7506, + "step": 8310 + }, + { + "epoch": 0.5602910953135002, + "grad_norm": 6.010172412411375, + "learning_rate": 4.82009586510297e-07, + "loss": 1.7813, + "step": 8315 + }, + { + "epoch": 0.5606280111856069, + "grad_norm": 5.862016493948776, + "learning_rate": 4.814219309218594e-07, + "loss": 1.7944, + "step": 8320 + }, + { + "epoch": 0.5609649270577137, + "grad_norm": 5.560071531221416, + "learning_rate": 4.808343010307199e-07, + "loss": 1.8738, + "step": 8325 + }, + { + "epoch": 0.5613018429298204, + "grad_norm": 6.017795379555469, + "learning_rate": 4.802466976496911e-07, + "loss": 1.8071, + "step": 8330 + }, + { + "epoch": 0.5616387588019272, + "grad_norm": 6.220898694052647, + "learning_rate": 4.796591215915498e-07, + "loss": 1.8234, + "step": 8335 + }, + { + "epoch": 0.5619756746740339, + "grad_norm": 6.061974235031019, + "learning_rate": 4.79071573669035e-07, + "loss": 1.8432, + "step": 8340 + }, + { + "epoch": 0.5623125905461406, + "grad_norm": 5.588859747562851, + "learning_rate": 4.784840546948463e-07, + "loss": 1.7834, + "step": 8345 + }, + { + "epoch": 0.5626495064182474, + "grad_norm": 5.981374345991413, + "learning_rate": 4.778965654816435e-07, + "loss": 1.7511, + "step": 8350 + }, + { + "epoch": 0.5629864222903541, + "grad_norm": 6.024335588504932, + "learning_rate": 4.773091068420455e-07, + "loss": 1.8607, + "step": 8355 + }, + { + "epoch": 0.5633233381624608, + "grad_norm": 5.760163267537706, + "learning_rate": 4.767216795886281e-07, + "loss": 1.822, + "step": 8360 + }, + { + "epoch": 0.5636602540345675, + "grad_norm": 6.0879150017865475, + "learning_rate": 4.761342845339246e-07, + "loss": 1.8162, + "step": 8365 + }, + { + "epoch": 0.5639971699066743, + "grad_norm": 6.335467066206414, + "learning_rate": 4.7554692249042345e-07, + "loss": 1.8653, + "step": 8370 + }, + { + "epoch": 0.5643340857787811, + "grad_norm": 6.132419754520845, + "learning_rate": 4.7495959427056754e-07, + "loss": 1.8353, + "step": 8375 + }, + { + "epoch": 0.5646710016508878, + "grad_norm": 5.63466915885368, + "learning_rate": 4.743723006867523e-07, + "loss": 1.7984, + "step": 8380 + }, + { + "epoch": 0.5650079175229945, + "grad_norm": 5.864284044006815, + "learning_rate": 4.737850425513263e-07, + "loss": 1.8202, + "step": 8385 + }, + { + "epoch": 0.5653448333951012, + "grad_norm": 5.707122660502422, + "learning_rate": 4.731978206765884e-07, + "loss": 1.8131, + "step": 8390 + }, + { + "epoch": 0.5656817492672079, + "grad_norm": 6.145138605313598, + "learning_rate": 4.726106358747871e-07, + "loss": 1.8188, + "step": 8395 + }, + { + "epoch": 0.5660186651393148, + "grad_norm": 5.810736149763557, + "learning_rate": 4.720234889581203e-07, + "loss": 1.854, + "step": 8400 + }, + { + "epoch": 0.5663555810114215, + "grad_norm": 6.178151042174748, + "learning_rate": 4.714363807387333e-07, + "loss": 1.8249, + "step": 8405 + }, + { + "epoch": 0.5666924968835282, + "grad_norm": 6.048303318074244, + "learning_rate": 4.708493120287175e-07, + "loss": 1.8319, + "step": 8410 + }, + { + "epoch": 0.5670294127556349, + "grad_norm": 5.506193309921177, + "learning_rate": 4.7026228364010984e-07, + "loss": 1.7959, + "step": 8415 + }, + { + "epoch": 0.5673663286277416, + "grad_norm": 6.236662710850324, + "learning_rate": 4.69675296384892e-07, + "loss": 1.8518, + "step": 8420 + }, + { + "epoch": 0.5677032444998484, + "grad_norm": 5.832229589759707, + "learning_rate": 4.6908835107498775e-07, + "loss": 1.8272, + "step": 8425 + }, + { + "epoch": 0.5680401603719551, + "grad_norm": 5.7248194948890445, + "learning_rate": 4.685014485222637e-07, + "loss": 1.8037, + "step": 8430 + }, + { + "epoch": 0.5683770762440619, + "grad_norm": 5.864436881700974, + "learning_rate": 4.679145895385269e-07, + "loss": 1.8137, + "step": 8435 + }, + { + "epoch": 0.5687139921161686, + "grad_norm": 6.05240631758938, + "learning_rate": 4.673277749355245e-07, + "loss": 1.7116, + "step": 8440 + }, + { + "epoch": 0.5690509079882753, + "grad_norm": 6.322483796907963, + "learning_rate": 4.667410055249417e-07, + "loss": 1.7331, + "step": 8445 + }, + { + "epoch": 0.5693878238603821, + "grad_norm": 6.218327051131996, + "learning_rate": 4.6615428211840154e-07, + "loss": 1.7826, + "step": 8450 + }, + { + "epoch": 0.5697247397324888, + "grad_norm": 5.964493626003666, + "learning_rate": 4.655676055274637e-07, + "loss": 1.8168, + "step": 8455 + }, + { + "epoch": 0.5700616556045955, + "grad_norm": 6.231161584629056, + "learning_rate": 4.6498097656362247e-07, + "loss": 1.8235, + "step": 8460 + }, + { + "epoch": 0.5703985714767023, + "grad_norm": 6.198185695145207, + "learning_rate": 4.643943960383067e-07, + "loss": 1.7135, + "step": 8465 + }, + { + "epoch": 0.570735487348809, + "grad_norm": 6.042628919234112, + "learning_rate": 4.638078647628782e-07, + "loss": 1.9163, + "step": 8470 + }, + { + "epoch": 0.5710724032209158, + "grad_norm": 6.169541691639398, + "learning_rate": 4.632213835486305e-07, + "loss": 1.8465, + "step": 8475 + }, + { + "epoch": 0.5714093190930225, + "grad_norm": 6.080905752465426, + "learning_rate": 4.626349532067879e-07, + "loss": 1.8297, + "step": 8480 + }, + { + "epoch": 0.5717462349651292, + "grad_norm": 5.721896292040939, + "learning_rate": 4.620485745485046e-07, + "loss": 1.7479, + "step": 8485 + }, + { + "epoch": 0.5720831508372359, + "grad_norm": 5.945565469765499, + "learning_rate": 4.6146224838486287e-07, + "loss": 1.8715, + "step": 8490 + }, + { + "epoch": 0.5724200667093426, + "grad_norm": 5.659262127072452, + "learning_rate": 4.6087597552687275e-07, + "loss": 1.8375, + "step": 8495 + }, + { + "epoch": 0.5727569825814494, + "grad_norm": 5.794077318723356, + "learning_rate": 4.602897567854705e-07, + "loss": 1.7981, + "step": 8500 + }, + { + "epoch": 0.5730938984535562, + "grad_norm": 6.099467621941091, + "learning_rate": 4.5970359297151733e-07, + "loss": 1.7755, + "step": 8505 + }, + { + "epoch": 0.5734308143256629, + "grad_norm": 5.984648189742544, + "learning_rate": 4.591174848957986e-07, + "loss": 1.8419, + "step": 8510 + }, + { + "epoch": 0.5737677301977696, + "grad_norm": 6.048661924764337, + "learning_rate": 4.585314333690224e-07, + "loss": 1.816, + "step": 8515 + }, + { + "epoch": 0.5741046460698763, + "grad_norm": 5.522305365781618, + "learning_rate": 4.579454392018192e-07, + "loss": 1.7642, + "step": 8520 + }, + { + "epoch": 0.574441561941983, + "grad_norm": 5.53435047065202, + "learning_rate": 4.5735950320473915e-07, + "loss": 1.7835, + "step": 8525 + }, + { + "epoch": 0.5747784778140899, + "grad_norm": 5.945621879771232, + "learning_rate": 4.5677362618825265e-07, + "loss": 1.7643, + "step": 8530 + }, + { + "epoch": 0.5751153936861966, + "grad_norm": 6.032046377900596, + "learning_rate": 4.5618780896274866e-07, + "loss": 1.7984, + "step": 8535 + }, + { + "epoch": 0.5754523095583033, + "grad_norm": 6.444772493536185, + "learning_rate": 4.556020523385326e-07, + "loss": 1.7324, + "step": 8540 + }, + { + "epoch": 0.57578922543041, + "grad_norm": 5.792224697201741, + "learning_rate": 4.55016357125827e-07, + "loss": 1.7728, + "step": 8545 + }, + { + "epoch": 0.5761261413025167, + "grad_norm": 5.753574407901725, + "learning_rate": 4.5443072413476877e-07, + "loss": 1.7846, + "step": 8550 + }, + { + "epoch": 0.5764630571746235, + "grad_norm": 6.350023820699443, + "learning_rate": 4.5384515417540914e-07, + "loss": 1.7999, + "step": 8555 + }, + { + "epoch": 0.5767999730467303, + "grad_norm": 5.6366643382723876, + "learning_rate": 4.5325964805771187e-07, + "loss": 1.8145, + "step": 8560 + }, + { + "epoch": 0.577136888918837, + "grad_norm": 5.750861217365839, + "learning_rate": 4.526742065915528e-07, + "loss": 1.7306, + "step": 8565 + }, + { + "epoch": 0.5774738047909437, + "grad_norm": 5.4864565976564466, + "learning_rate": 4.520888305867181e-07, + "loss": 1.773, + "step": 8570 + }, + { + "epoch": 0.5778107206630504, + "grad_norm": 6.158976718779292, + "learning_rate": 4.5150352085290315e-07, + "loss": 1.8288, + "step": 8575 + }, + { + "epoch": 0.5781476365351572, + "grad_norm": 5.909685923044137, + "learning_rate": 4.5091827819971207e-07, + "loss": 1.8347, + "step": 8580 + }, + { + "epoch": 0.5784845524072639, + "grad_norm": 6.080997476219573, + "learning_rate": 4.503331034366563e-07, + "loss": 1.8089, + "step": 8585 + }, + { + "epoch": 0.5788214682793706, + "grad_norm": 5.724875421513874, + "learning_rate": 4.4974799737315274e-07, + "loss": 1.7667, + "step": 8590 + }, + { + "epoch": 0.5791583841514774, + "grad_norm": 5.610253640773116, + "learning_rate": 4.491629608185237e-07, + "loss": 1.7924, + "step": 8595 + }, + { + "epoch": 0.5794953000235841, + "grad_norm": 5.610898049083143, + "learning_rate": 4.485779945819956e-07, + "loss": 1.7901, + "step": 8600 + }, + { + "epoch": 0.5798322158956909, + "grad_norm": 5.81753396746651, + "learning_rate": 4.479930994726968e-07, + "loss": 1.7487, + "step": 8605 + }, + { + "epoch": 0.5801691317677976, + "grad_norm": 6.1155229721373745, + "learning_rate": 4.474082762996581e-07, + "loss": 1.8492, + "step": 8610 + }, + { + "epoch": 0.5805060476399043, + "grad_norm": 6.165815910325223, + "learning_rate": 4.468235258718105e-07, + "loss": 1.8476, + "step": 8615 + }, + { + "epoch": 0.580842963512011, + "grad_norm": 5.803183552453552, + "learning_rate": 4.4623884899798397e-07, + "loss": 1.8046, + "step": 8620 + }, + { + "epoch": 0.5811798793841177, + "grad_norm": 6.357851078191232, + "learning_rate": 4.4565424648690743e-07, + "loss": 1.7539, + "step": 8625 + }, + { + "epoch": 0.5815167952562246, + "grad_norm": 6.508465623286144, + "learning_rate": 4.450697191472067e-07, + "loss": 1.7605, + "step": 8630 + }, + { + "epoch": 0.5818537111283313, + "grad_norm": 5.792009077737253, + "learning_rate": 4.4448526778740327e-07, + "loss": 1.7607, + "step": 8635 + }, + { + "epoch": 0.582190627000438, + "grad_norm": 6.377605622379964, + "learning_rate": 4.439008932159138e-07, + "loss": 1.8341, + "step": 8640 + }, + { + "epoch": 0.5825275428725447, + "grad_norm": 5.989708484254575, + "learning_rate": 4.4331659624104876e-07, + "loss": 1.8957, + "step": 8645 + }, + { + "epoch": 0.5828644587446514, + "grad_norm": 6.135183719688721, + "learning_rate": 4.427323776710117e-07, + "loss": 1.8505, + "step": 8650 + }, + { + "epoch": 0.5832013746167583, + "grad_norm": 6.263770305683316, + "learning_rate": 4.4214823831389663e-07, + "loss": 1.8154, + "step": 8655 + }, + { + "epoch": 0.583538290488865, + "grad_norm": 5.800608275245644, + "learning_rate": 4.41564178977689e-07, + "loss": 1.7261, + "step": 8660 + }, + { + "epoch": 0.5838752063609717, + "grad_norm": 5.809205967021153, + "learning_rate": 4.4098020047026343e-07, + "loss": 1.7584, + "step": 8665 + }, + { + "epoch": 0.5842121222330784, + "grad_norm": 5.963419654218539, + "learning_rate": 4.4039630359938194e-07, + "loss": 1.7529, + "step": 8670 + }, + { + "epoch": 0.5845490381051851, + "grad_norm": 5.8426129848767845, + "learning_rate": 4.3981248917269477e-07, + "loss": 1.824, + "step": 8675 + }, + { + "epoch": 0.5848859539772918, + "grad_norm": 5.944106275409081, + "learning_rate": 4.3922875799773735e-07, + "loss": 1.7739, + "step": 8680 + }, + { + "epoch": 0.5852228698493986, + "grad_norm": 5.773197551120905, + "learning_rate": 4.386451108819302e-07, + "loss": 1.7682, + "step": 8685 + }, + { + "epoch": 0.5855597857215054, + "grad_norm": 6.552569491362873, + "learning_rate": 4.380615486325774e-07, + "loss": 1.8046, + "step": 8690 + }, + { + "epoch": 0.5858967015936121, + "grad_norm": 5.456979786224337, + "learning_rate": 4.3747807205686616e-07, + "loss": 1.7698, + "step": 8695 + }, + { + "epoch": 0.5862336174657188, + "grad_norm": 5.11890654536534, + "learning_rate": 4.3689468196186433e-07, + "loss": 1.7561, + "step": 8700 + }, + { + "epoch": 0.5865705333378255, + "grad_norm": 5.716069902489952, + "learning_rate": 4.36311379154521e-07, + "loss": 1.8601, + "step": 8705 + }, + { + "epoch": 0.5869074492099323, + "grad_norm": 6.22330264411751, + "learning_rate": 4.3572816444166406e-07, + "loss": 1.8129, + "step": 8710 + }, + { + "epoch": 0.587244365082039, + "grad_norm": 6.086909858486789, + "learning_rate": 4.351450386299996e-07, + "loss": 1.721, + "step": 8715 + }, + { + "epoch": 0.5875812809541457, + "grad_norm": 6.297426986410737, + "learning_rate": 4.3456200252611075e-07, + "loss": 1.8179, + "step": 8720 + }, + { + "epoch": 0.5879181968262525, + "grad_norm": 5.840418602695064, + "learning_rate": 4.3397905693645653e-07, + "loss": 1.8302, + "step": 8725 + }, + { + "epoch": 0.5882551126983592, + "grad_norm": 6.3296952466437695, + "learning_rate": 4.3339620266737116e-07, + "loss": 1.7801, + "step": 8730 + }, + { + "epoch": 0.588592028570466, + "grad_norm": 5.703161646952239, + "learning_rate": 4.328134405250617e-07, + "loss": 1.7261, + "step": 8735 + }, + { + "epoch": 0.5889289444425727, + "grad_norm": 6.296431483288242, + "learning_rate": 4.322307713156085e-07, + "loss": 1.7399, + "step": 8740 + }, + { + "epoch": 0.5892658603146794, + "grad_norm": 6.319426786873787, + "learning_rate": 4.316481958449634e-07, + "loss": 1.7458, + "step": 8745 + }, + { + "epoch": 0.5896027761867861, + "grad_norm": 5.616103643263215, + "learning_rate": 4.310657149189478e-07, + "loss": 1.8584, + "step": 8750 + }, + { + "epoch": 0.5899396920588929, + "grad_norm": 6.108036015594825, + "learning_rate": 4.3048332934325325e-07, + "loss": 1.8297, + "step": 8755 + }, + { + "epoch": 0.5902766079309997, + "grad_norm": 6.515815846454065, + "learning_rate": 4.2990103992343893e-07, + "loss": 1.771, + "step": 8760 + }, + { + "epoch": 0.5906135238031064, + "grad_norm": 5.921183552077239, + "learning_rate": 4.2931884746493107e-07, + "loss": 1.8703, + "step": 8765 + }, + { + "epoch": 0.5909504396752131, + "grad_norm": 6.096104391048515, + "learning_rate": 4.287367527730216e-07, + "loss": 1.7736, + "step": 8770 + }, + { + "epoch": 0.5912873555473198, + "grad_norm": 6.05992060448168, + "learning_rate": 4.2815475665286766e-07, + "loss": 1.8258, + "step": 8775 + }, + { + "epoch": 0.5916242714194265, + "grad_norm": 6.535494644184692, + "learning_rate": 4.2757285990948993e-07, + "loss": 1.8059, + "step": 8780 + }, + { + "epoch": 0.5919611872915334, + "grad_norm": 6.1860767579422795, + "learning_rate": 4.269910633477711e-07, + "loss": 1.7594, + "step": 8785 + }, + { + "epoch": 0.5922981031636401, + "grad_norm": 5.972961366882138, + "learning_rate": 4.264093677724561e-07, + "loss": 1.8025, + "step": 8790 + }, + { + "epoch": 0.5926350190357468, + "grad_norm": 5.556665434802162, + "learning_rate": 4.2582777398814966e-07, + "loss": 1.7414, + "step": 8795 + }, + { + "epoch": 0.5929719349078535, + "grad_norm": 5.92972532751255, + "learning_rate": 4.252462827993158e-07, + "loss": 1.7877, + "step": 8800 + }, + { + "epoch": 0.5933088507799602, + "grad_norm": 6.056102795509996, + "learning_rate": 4.246648950102765e-07, + "loss": 1.8636, + "step": 8805 + }, + { + "epoch": 0.593645766652067, + "grad_norm": 5.980939951020172, + "learning_rate": 4.240836114252112e-07, + "loss": 1.8036, + "step": 8810 + }, + { + "epoch": 0.5939826825241737, + "grad_norm": 6.003334901011222, + "learning_rate": 4.2350243284815445e-07, + "loss": 1.7733, + "step": 8815 + }, + { + "epoch": 0.5943195983962805, + "grad_norm": 5.955573412552551, + "learning_rate": 4.229213600829963e-07, + "loss": 1.8063, + "step": 8820 + }, + { + "epoch": 0.5946565142683872, + "grad_norm": 5.656293844274432, + "learning_rate": 4.223403939334802e-07, + "loss": 1.8226, + "step": 8825 + }, + { + "epoch": 0.5949934301404939, + "grad_norm": 6.749213331616512, + "learning_rate": 4.217595352032017e-07, + "loss": 1.819, + "step": 8830 + }, + { + "epoch": 0.5953303460126007, + "grad_norm": 6.117182599544254, + "learning_rate": 4.2117878469560834e-07, + "loss": 1.7613, + "step": 8835 + }, + { + "epoch": 0.5956672618847074, + "grad_norm": 5.520177923235137, + "learning_rate": 4.205981432139978e-07, + "loss": 1.768, + "step": 8840 + }, + { + "epoch": 0.5960041777568141, + "grad_norm": 5.711090146795336, + "learning_rate": 4.200176115615169e-07, + "loss": 1.8229, + "step": 8845 + }, + { + "epoch": 0.5963410936289208, + "grad_norm": 6.036353569783169, + "learning_rate": 4.1943719054116027e-07, + "loss": 1.7739, + "step": 8850 + }, + { + "epoch": 0.5966780095010276, + "grad_norm": 6.3115691145794965, + "learning_rate": 4.1885688095577e-07, + "loss": 1.7397, + "step": 8855 + }, + { + "epoch": 0.5970149253731343, + "grad_norm": 5.7168631691824245, + "learning_rate": 4.182766836080339e-07, + "loss": 1.753, + "step": 8860 + }, + { + "epoch": 0.5973518412452411, + "grad_norm": 6.186256876910364, + "learning_rate": 4.176965993004842e-07, + "loss": 1.8123, + "step": 8865 + }, + { + "epoch": 0.5976887571173478, + "grad_norm": 6.505307800637823, + "learning_rate": 4.171166288354971e-07, + "loss": 1.8067, + "step": 8870 + }, + { + "epoch": 0.5980256729894545, + "grad_norm": 5.760914031493992, + "learning_rate": 4.165367730152917e-07, + "loss": 1.8102, + "step": 8875 + }, + { + "epoch": 0.5983625888615612, + "grad_norm": 6.056216167167739, + "learning_rate": 4.1595703264192737e-07, + "loss": 1.8183, + "step": 8880 + }, + { + "epoch": 0.598699504733668, + "grad_norm": 5.852866203850851, + "learning_rate": 4.15377408517305e-07, + "loss": 1.7875, + "step": 8885 + }, + { + "epoch": 0.5990364206057748, + "grad_norm": 6.083621089321414, + "learning_rate": 4.147979014431642e-07, + "loss": 1.7886, + "step": 8890 + }, + { + "epoch": 0.5993733364778815, + "grad_norm": 6.6265834284264695, + "learning_rate": 4.142185122210823e-07, + "loss": 1.7928, + "step": 8895 + }, + { + "epoch": 0.5997102523499882, + "grad_norm": 5.789010083005185, + "learning_rate": 4.136392416524742e-07, + "loss": 1.8099, + "step": 8900 + }, + { + "epoch": 0.6000471682220949, + "grad_norm": 6.131350960422217, + "learning_rate": 4.1306009053859043e-07, + "loss": 1.7875, + "step": 8905 + }, + { + "epoch": 0.6003840840942016, + "grad_norm": 6.264733608982983, + "learning_rate": 4.124810596805166e-07, + "loss": 1.7994, + "step": 8910 + }, + { + "epoch": 0.6007209999663085, + "grad_norm": 5.829929322028521, + "learning_rate": 4.119021498791712e-07, + "loss": 1.8682, + "step": 8915 + }, + { + "epoch": 0.6010579158384152, + "grad_norm": 5.515885939117128, + "learning_rate": 4.113233619353062e-07, + "loss": 1.78, + "step": 8920 + }, + { + "epoch": 0.6013948317105219, + "grad_norm": 6.529160464329195, + "learning_rate": 4.107446966495044e-07, + "loss": 1.8267, + "step": 8925 + }, + { + "epoch": 0.6017317475826286, + "grad_norm": 5.767404449030353, + "learning_rate": 4.101661548221792e-07, + "loss": 1.8162, + "step": 8930 + }, + { + "epoch": 0.6020686634547353, + "grad_norm": 6.2522730607410795, + "learning_rate": 4.0958773725357297e-07, + "loss": 1.7927, + "step": 8935 + }, + { + "epoch": 0.6024055793268421, + "grad_norm": 5.920238162873758, + "learning_rate": 4.0900944474375674e-07, + "loss": 1.8076, + "step": 8940 + }, + { + "epoch": 0.6027424951989488, + "grad_norm": 6.070363136220485, + "learning_rate": 4.084312780926279e-07, + "loss": 1.76, + "step": 8945 + }, + { + "epoch": 0.6030794110710556, + "grad_norm": 6.080971532825581, + "learning_rate": 4.0785323809991006e-07, + "loss": 1.8373, + "step": 8950 + }, + { + "epoch": 0.6034163269431623, + "grad_norm": 6.008686213516527, + "learning_rate": 4.072753255651521e-07, + "loss": 1.7578, + "step": 8955 + }, + { + "epoch": 0.603753242815269, + "grad_norm": 6.373636076172385, + "learning_rate": 4.066975412877255e-07, + "loss": 1.7397, + "step": 8960 + }, + { + "epoch": 0.6040901586873758, + "grad_norm": 7.110282397737251, + "learning_rate": 4.0611988606682544e-07, + "loss": 1.7594, + "step": 8965 + }, + { + "epoch": 0.6044270745594825, + "grad_norm": 6.569745758459806, + "learning_rate": 4.0554236070146785e-07, + "loss": 1.8107, + "step": 8970 + }, + { + "epoch": 0.6047639904315892, + "grad_norm": 6.442153100708266, + "learning_rate": 4.0496496599048963e-07, + "loss": 1.7975, + "step": 8975 + }, + { + "epoch": 0.605100906303696, + "grad_norm": 6.337402589376834, + "learning_rate": 4.0438770273254624e-07, + "loss": 1.7996, + "step": 8980 + }, + { + "epoch": 0.6054378221758027, + "grad_norm": 6.379223050210244, + "learning_rate": 4.038105717261119e-07, + "loss": 1.757, + "step": 8985 + }, + { + "epoch": 0.6057747380479095, + "grad_norm": 5.746548813457159, + "learning_rate": 4.03233573769478e-07, + "loss": 1.8244, + "step": 8990 + }, + { + "epoch": 0.6061116539200162, + "grad_norm": 6.244048991447363, + "learning_rate": 4.026567096607511e-07, + "loss": 1.7605, + "step": 8995 + }, + { + "epoch": 0.6064485697921229, + "grad_norm": 5.893004419934275, + "learning_rate": 4.020799801978535e-07, + "loss": 1.8646, + "step": 9000 + }, + { + "epoch": 0.6067854856642296, + "grad_norm": 6.3581750988564565, + "learning_rate": 4.015033861785208e-07, + "loss": 1.8016, + "step": 9005 + }, + { + "epoch": 0.6071224015363363, + "grad_norm": 6.024642368549191, + "learning_rate": 4.0092692840030126e-07, + "loss": 1.8321, + "step": 9010 + }, + { + "epoch": 0.6074593174084432, + "grad_norm": 5.692437973796029, + "learning_rate": 4.003506076605547e-07, + "loss": 1.7497, + "step": 9015 + }, + { + "epoch": 0.6077962332805499, + "grad_norm": 5.904037102274744, + "learning_rate": 3.997744247564519e-07, + "loss": 1.7012, + "step": 9020 + }, + { + "epoch": 0.6081331491526566, + "grad_norm": 5.770842989006306, + "learning_rate": 3.9919838048497197e-07, + "loss": 1.8748, + "step": 9025 + }, + { + "epoch": 0.6084700650247633, + "grad_norm": 5.717267913626994, + "learning_rate": 3.98622475642903e-07, + "loss": 1.7992, + "step": 9030 + }, + { + "epoch": 0.60880698089687, + "grad_norm": 6.54894021053903, + "learning_rate": 3.980467110268405e-07, + "loss": 1.7435, + "step": 9035 + }, + { + "epoch": 0.6091438967689767, + "grad_norm": 6.259824241149931, + "learning_rate": 3.9747108743318493e-07, + "loss": 1.8146, + "step": 9040 + }, + { + "epoch": 0.6094808126410836, + "grad_norm": 5.981839546695795, + "learning_rate": 3.968956056581428e-07, + "loss": 1.7738, + "step": 9045 + }, + { + "epoch": 0.6098177285131903, + "grad_norm": 6.055280113756847, + "learning_rate": 3.9632026649772366e-07, + "loss": 1.7324, + "step": 9050 + }, + { + "epoch": 0.610154644385297, + "grad_norm": 6.289372477937633, + "learning_rate": 3.9574507074774056e-07, + "loss": 1.7778, + "step": 9055 + }, + { + "epoch": 0.6104915602574037, + "grad_norm": 5.875270432715915, + "learning_rate": 3.951700192038072e-07, + "loss": 1.7313, + "step": 9060 + }, + { + "epoch": 0.6108284761295104, + "grad_norm": 5.598591559503298, + "learning_rate": 3.945951126613387e-07, + "loss": 1.8703, + "step": 9065 + }, + { + "epoch": 0.6111653920016172, + "grad_norm": 6.69707404500823, + "learning_rate": 3.9402035191554937e-07, + "loss": 1.7974, + "step": 9070 + }, + { + "epoch": 0.611502307873724, + "grad_norm": 6.008857949179827, + "learning_rate": 3.934457377614514e-07, + "loss": 1.7459, + "step": 9075 + }, + { + "epoch": 0.6118392237458307, + "grad_norm": 5.617095082901805, + "learning_rate": 3.9287127099385483e-07, + "loss": 1.8318, + "step": 9080 + }, + { + "epoch": 0.6121761396179374, + "grad_norm": 6.535128437503064, + "learning_rate": 3.9229695240736567e-07, + "loss": 1.7986, + "step": 9085 + }, + { + "epoch": 0.6125130554900441, + "grad_norm": 6.152619512707627, + "learning_rate": 3.917227827963846e-07, + "loss": 1.7792, + "step": 9090 + }, + { + "epoch": 0.6128499713621509, + "grad_norm": 6.3313341167472155, + "learning_rate": 3.9114876295510653e-07, + "loss": 1.7732, + "step": 9095 + }, + { + "epoch": 0.6131868872342576, + "grad_norm": 5.895957841166245, + "learning_rate": 3.9057489367751947e-07, + "loss": 1.7714, + "step": 9100 + }, + { + "epoch": 0.6135238031063643, + "grad_norm": 5.798092152940961, + "learning_rate": 3.900011757574024e-07, + "loss": 1.862, + "step": 9105 + }, + { + "epoch": 0.613860718978471, + "grad_norm": 6.545396318492506, + "learning_rate": 3.894276099883258e-07, + "loss": 1.7627, + "step": 9110 + }, + { + "epoch": 0.6141976348505778, + "grad_norm": 6.161928532867873, + "learning_rate": 3.888541971636492e-07, + "loss": 1.7453, + "step": 9115 + }, + { + "epoch": 0.6145345507226846, + "grad_norm": 6.748188653900746, + "learning_rate": 3.8828093807652095e-07, + "loss": 1.8917, + "step": 9120 + }, + { + "epoch": 0.6148714665947913, + "grad_norm": 6.159848242287639, + "learning_rate": 3.8770783351987605e-07, + "loss": 1.7577, + "step": 9125 + }, + { + "epoch": 0.615208382466898, + "grad_norm": 6.534801005504792, + "learning_rate": 3.8713488428643656e-07, + "loss": 1.7931, + "step": 9130 + }, + { + "epoch": 0.6155452983390047, + "grad_norm": 5.739673674024533, + "learning_rate": 3.8656209116870906e-07, + "loss": 1.7721, + "step": 9135 + }, + { + "epoch": 0.6158822142111114, + "grad_norm": 5.7311702430866145, + "learning_rate": 3.859894549589847e-07, + "loss": 1.8119, + "step": 9140 + }, + { + "epoch": 0.6162191300832183, + "grad_norm": 6.415780867852155, + "learning_rate": 3.854169764493371e-07, + "loss": 1.8124, + "step": 9145 + }, + { + "epoch": 0.616556045955325, + "grad_norm": 5.955671102064489, + "learning_rate": 3.848446564316223e-07, + "loss": 1.7789, + "step": 9150 + }, + { + "epoch": 0.6168929618274317, + "grad_norm": 6.461847161048739, + "learning_rate": 3.8427249569747656e-07, + "loss": 1.791, + "step": 9155 + }, + { + "epoch": 0.6172298776995384, + "grad_norm": 6.0663090140076745, + "learning_rate": 3.8370049503831614e-07, + "loss": 1.7913, + "step": 9160 + }, + { + "epoch": 0.6175667935716451, + "grad_norm": 6.03812934458201, + "learning_rate": 3.8312865524533606e-07, + "loss": 1.773, + "step": 9165 + }, + { + "epoch": 0.617903709443752, + "grad_norm": 5.945622065488643, + "learning_rate": 3.825569771095082e-07, + "loss": 1.8358, + "step": 9170 + }, + { + "epoch": 0.6182406253158587, + "grad_norm": 6.041236208673518, + "learning_rate": 3.819854614215814e-07, + "loss": 1.8262, + "step": 9175 + }, + { + "epoch": 0.6185775411879654, + "grad_norm": 6.344915324522393, + "learning_rate": 3.814141089720796e-07, + "loss": 1.832, + "step": 9180 + }, + { + "epoch": 0.6189144570600721, + "grad_norm": 6.343605951445983, + "learning_rate": 3.8084292055130126e-07, + "loss": 1.8455, + "step": 9185 + }, + { + "epoch": 0.6192513729321788, + "grad_norm": 6.0890477862561, + "learning_rate": 3.8027189694931715e-07, + "loss": 1.7903, + "step": 9190 + }, + { + "epoch": 0.6195882888042856, + "grad_norm": 5.747051553245304, + "learning_rate": 3.797010389559708e-07, + "loss": 1.8008, + "step": 9195 + }, + { + "epoch": 0.6199252046763923, + "grad_norm": 6.455198765907777, + "learning_rate": 3.7913034736087677e-07, + "loss": 1.8564, + "step": 9200 + }, + { + "epoch": 0.620262120548499, + "grad_norm": 6.107645595637579, + "learning_rate": 3.785598229534186e-07, + "loss": 1.8038, + "step": 9205 + }, + { + "epoch": 0.6205990364206058, + "grad_norm": 5.6893786974616205, + "learning_rate": 3.7798946652274943e-07, + "loss": 1.7916, + "step": 9210 + }, + { + "epoch": 0.6209359522927125, + "grad_norm": 5.386804515229501, + "learning_rate": 3.7741927885778966e-07, + "loss": 1.7064, + "step": 9215 + }, + { + "epoch": 0.6212728681648192, + "grad_norm": 5.5903465122048805, + "learning_rate": 3.768492607472263e-07, + "loss": 1.8083, + "step": 9220 + }, + { + "epoch": 0.621609784036926, + "grad_norm": 5.597070370599875, + "learning_rate": 3.7627941297951183e-07, + "loss": 1.8165, + "step": 9225 + }, + { + "epoch": 0.6219466999090327, + "grad_norm": 5.70072119730533, + "learning_rate": 3.7570973634286334e-07, + "loss": 1.7952, + "step": 9230 + }, + { + "epoch": 0.6222836157811394, + "grad_norm": 5.808800928790782, + "learning_rate": 3.7514023162526066e-07, + "loss": 1.7582, + "step": 9235 + }, + { + "epoch": 0.6226205316532462, + "grad_norm": 5.88125644092123, + "learning_rate": 3.745708996144463e-07, + "loss": 1.832, + "step": 9240 + }, + { + "epoch": 0.6229574475253529, + "grad_norm": 6.384851274473344, + "learning_rate": 3.740017410979239e-07, + "loss": 1.7643, + "step": 9245 + }, + { + "epoch": 0.6232943633974597, + "grad_norm": 6.047883416599075, + "learning_rate": 3.734327568629569e-07, + "loss": 1.8043, + "step": 9250 + }, + { + "epoch": 0.6236312792695664, + "grad_norm": 5.338272475290313, + "learning_rate": 3.728639476965678e-07, + "loss": 1.7449, + "step": 9255 + }, + { + "epoch": 0.6239681951416731, + "grad_norm": 6.428658869577709, + "learning_rate": 3.7229531438553664e-07, + "loss": 1.7997, + "step": 9260 + }, + { + "epoch": 0.6243051110137798, + "grad_norm": 6.026885048415576, + "learning_rate": 3.7172685771640076e-07, + "loss": 1.8192, + "step": 9265 + }, + { + "epoch": 0.6246420268858865, + "grad_norm": 6.122237637003847, + "learning_rate": 3.7115857847545264e-07, + "loss": 1.7574, + "step": 9270 + }, + { + "epoch": 0.6249789427579934, + "grad_norm": 5.558903228452016, + "learning_rate": 3.7059047744873955e-07, + "loss": 1.8024, + "step": 9275 + }, + { + "epoch": 0.6253158586301001, + "grad_norm": 5.430476107786122, + "learning_rate": 3.700225554220626e-07, + "loss": 1.8056, + "step": 9280 + }, + { + "epoch": 0.6256527745022068, + "grad_norm": 5.82118125541433, + "learning_rate": 3.694548131809747e-07, + "loss": 1.7898, + "step": 9285 + }, + { + "epoch": 0.6259896903743135, + "grad_norm": 5.714807325686535, + "learning_rate": 3.6888725151078024e-07, + "loss": 1.8007, + "step": 9290 + }, + { + "epoch": 0.6263266062464202, + "grad_norm": 5.669878903492906, + "learning_rate": 3.683198711965345e-07, + "loss": 1.746, + "step": 9295 + }, + { + "epoch": 0.626663522118527, + "grad_norm": 5.814699385958482, + "learning_rate": 3.677526730230408e-07, + "loss": 1.7676, + "step": 9300 + }, + { + "epoch": 0.6270004379906338, + "grad_norm": 6.813971359453809, + "learning_rate": 3.671856577748512e-07, + "loss": 1.7549, + "step": 9305 + }, + { + "epoch": 0.6273373538627405, + "grad_norm": 6.810563784485118, + "learning_rate": 3.666188262362648e-07, + "loss": 1.8067, + "step": 9310 + }, + { + "epoch": 0.6276742697348472, + "grad_norm": 6.310703666355027, + "learning_rate": 3.660521791913265e-07, + "loss": 1.8092, + "step": 9315 + }, + { + "epoch": 0.6280111856069539, + "grad_norm": 5.895296847285542, + "learning_rate": 3.654857174238256e-07, + "loss": 1.7903, + "step": 9320 + }, + { + "epoch": 0.6283481014790607, + "grad_norm": 5.588323528807347, + "learning_rate": 3.649194417172957e-07, + "loss": 1.8755, + "step": 9325 + }, + { + "epoch": 0.6286850173511674, + "grad_norm": 5.7839080812815435, + "learning_rate": 3.6435335285501283e-07, + "loss": 1.7056, + "step": 9330 + }, + { + "epoch": 0.6290219332232742, + "grad_norm": 5.5844659216765065, + "learning_rate": 3.6378745161999426e-07, + "loss": 1.8071, + "step": 9335 + }, + { + "epoch": 0.6293588490953809, + "grad_norm": 5.824070138199869, + "learning_rate": 3.632217387949983e-07, + "loss": 1.8042, + "step": 9340 + }, + { + "epoch": 0.6296957649674876, + "grad_norm": 5.711934278426134, + "learning_rate": 3.626562151625223e-07, + "loss": 1.7275, + "step": 9345 + }, + { + "epoch": 0.6300326808395944, + "grad_norm": 5.778338968005331, + "learning_rate": 3.6209088150480173e-07, + "loss": 1.7343, + "step": 9350 + }, + { + "epoch": 0.6303695967117011, + "grad_norm": 6.311857290704742, + "learning_rate": 3.6152573860380964e-07, + "loss": 1.7612, + "step": 9355 + }, + { + "epoch": 0.6307065125838078, + "grad_norm": 5.999783421284286, + "learning_rate": 3.6096078724125544e-07, + "loss": 1.8523, + "step": 9360 + }, + { + "epoch": 0.6310434284559145, + "grad_norm": 6.3198427864797315, + "learning_rate": 3.603960281985828e-07, + "loss": 1.8345, + "step": 9365 + }, + { + "epoch": 0.6313803443280213, + "grad_norm": 6.104616999431709, + "learning_rate": 3.5983146225697007e-07, + "loss": 1.7699, + "step": 9370 + }, + { + "epoch": 0.6317172602001281, + "grad_norm": 6.1785780292512085, + "learning_rate": 3.5926709019732855e-07, + "loss": 1.8611, + "step": 9375 + }, + { + "epoch": 0.6320541760722348, + "grad_norm": 5.938144134115797, + "learning_rate": 3.587029128003006e-07, + "loss": 1.7861, + "step": 9380 + }, + { + "epoch": 0.6323910919443415, + "grad_norm": 6.224929650940109, + "learning_rate": 3.581389308462601e-07, + "loss": 1.7648, + "step": 9385 + }, + { + "epoch": 0.6327280078164482, + "grad_norm": 6.061589086959369, + "learning_rate": 3.5757514511531016e-07, + "loss": 1.819, + "step": 9390 + }, + { + "epoch": 0.6330649236885549, + "grad_norm": 5.951184996035118, + "learning_rate": 3.5701155638728297e-07, + "loss": 1.8019, + "step": 9395 + }, + { + "epoch": 0.6334018395606617, + "grad_norm": 5.682637936259567, + "learning_rate": 3.564481654417374e-07, + "loss": 1.7767, + "step": 9400 + }, + { + "epoch": 0.6337387554327685, + "grad_norm": 5.676942423843257, + "learning_rate": 3.558849730579594e-07, + "loss": 1.8106, + "step": 9405 + }, + { + "epoch": 0.6340756713048752, + "grad_norm": 6.070636918840517, + "learning_rate": 3.553219800149603e-07, + "loss": 1.7765, + "step": 9410 + }, + { + "epoch": 0.6344125871769819, + "grad_norm": 5.891991459452166, + "learning_rate": 3.547591870914752e-07, + "loss": 1.7422, + "step": 9415 + }, + { + "epoch": 0.6347495030490886, + "grad_norm": 6.0285641037028315, + "learning_rate": 3.5419659506596287e-07, + "loss": 1.7854, + "step": 9420 + }, + { + "epoch": 0.6350864189211953, + "grad_norm": 5.939362084214073, + "learning_rate": 3.536342047166039e-07, + "loss": 1.7696, + "step": 9425 + }, + { + "epoch": 0.6354233347933022, + "grad_norm": 5.869634005185154, + "learning_rate": 3.530720168213001e-07, + "loss": 1.7639, + "step": 9430 + }, + { + "epoch": 0.6357602506654089, + "grad_norm": 5.6278059236397855, + "learning_rate": 3.5251003215767305e-07, + "loss": 1.7762, + "step": 9435 + }, + { + "epoch": 0.6360971665375156, + "grad_norm": 6.050098422180009, + "learning_rate": 3.519482515030636e-07, + "loss": 1.7914, + "step": 9440 + }, + { + "epoch": 0.6364340824096223, + "grad_norm": 5.726733009746023, + "learning_rate": 3.5138667563452983e-07, + "loss": 1.8208, + "step": 9445 + }, + { + "epoch": 0.636770998281729, + "grad_norm": 7.0847167131878335, + "learning_rate": 3.5082530532884703e-07, + "loss": 1.7328, + "step": 9450 + }, + { + "epoch": 0.6371079141538358, + "grad_norm": 5.768645926917814, + "learning_rate": 3.5026414136250607e-07, + "loss": 1.7745, + "step": 9455 + }, + { + "epoch": 0.6374448300259425, + "grad_norm": 5.255887609055351, + "learning_rate": 3.497031845117124e-07, + "loss": 1.8214, + "step": 9460 + }, + { + "epoch": 0.6377817458980493, + "grad_norm": 6.164645047794775, + "learning_rate": 3.4914243555238476e-07, + "loss": 1.8279, + "step": 9465 + }, + { + "epoch": 0.638118661770156, + "grad_norm": 5.8015039903705645, + "learning_rate": 3.4858189526015453e-07, + "loss": 1.7587, + "step": 9470 + }, + { + "epoch": 0.6384555776422627, + "grad_norm": 6.166820116860601, + "learning_rate": 3.4802156441036467e-07, + "loss": 1.7864, + "step": 9475 + }, + { + "epoch": 0.6387924935143695, + "grad_norm": 5.652602800993565, + "learning_rate": 3.4746144377806785e-07, + "loss": 1.787, + "step": 9480 + }, + { + "epoch": 0.6391294093864762, + "grad_norm": 6.215195219229495, + "learning_rate": 3.4690153413802653e-07, + "loss": 1.8208, + "step": 9485 + }, + { + "epoch": 0.6394663252585829, + "grad_norm": 6.017628489290521, + "learning_rate": 3.4634183626471125e-07, + "loss": 1.8729, + "step": 9490 + }, + { + "epoch": 0.6398032411306896, + "grad_norm": 6.027983940121836, + "learning_rate": 3.457823509322992e-07, + "loss": 1.842, + "step": 9495 + }, + { + "epoch": 0.6401401570027964, + "grad_norm": 6.276935654397374, + "learning_rate": 3.452230789146741e-07, + "loss": 1.8351, + "step": 9500 + }, + { + "epoch": 0.6404770728749032, + "grad_norm": 6.351214097699412, + "learning_rate": 3.4466402098542435e-07, + "loss": 1.7684, + "step": 9505 + }, + { + "epoch": 0.6408139887470099, + "grad_norm": 6.052932413421503, + "learning_rate": 3.441051779178422e-07, + "loss": 1.7954, + "step": 9510 + }, + { + "epoch": 0.6411509046191166, + "grad_norm": 5.871050354755707, + "learning_rate": 3.4354655048492277e-07, + "loss": 1.7081, + "step": 9515 + }, + { + "epoch": 0.6414878204912233, + "grad_norm": 5.43280461728629, + "learning_rate": 3.429881394593629e-07, + "loss": 1.7498, + "step": 9520 + }, + { + "epoch": 0.64182473636333, + "grad_norm": 6.277219662282607, + "learning_rate": 3.4242994561356043e-07, + "loss": 1.7507, + "step": 9525 + }, + { + "epoch": 0.6421616522354369, + "grad_norm": 5.964958889609506, + "learning_rate": 3.4187196971961185e-07, + "loss": 1.8509, + "step": 9530 + }, + { + "epoch": 0.6424985681075436, + "grad_norm": 6.233012701053734, + "learning_rate": 3.4131421254931326e-07, + "loss": 1.816, + "step": 9535 + }, + { + "epoch": 0.6428354839796503, + "grad_norm": 5.718317294931223, + "learning_rate": 3.4075667487415785e-07, + "loss": 1.8602, + "step": 9540 + }, + { + "epoch": 0.643172399851757, + "grad_norm": 6.076698722005338, + "learning_rate": 3.4019935746533474e-07, + "loss": 1.8048, + "step": 9545 + }, + { + "epoch": 0.6435093157238637, + "grad_norm": 6.344118541757631, + "learning_rate": 3.3964226109372884e-07, + "loss": 1.8794, + "step": 9550 + }, + { + "epoch": 0.6438462315959704, + "grad_norm": 5.824838401908134, + "learning_rate": 3.390853865299195e-07, + "loss": 1.7912, + "step": 9555 + }, + { + "epoch": 0.6441831474680773, + "grad_norm": 6.209879017865747, + "learning_rate": 3.385287345441786e-07, + "loss": 1.8699, + "step": 9560 + }, + { + "epoch": 0.644520063340184, + "grad_norm": 6.082107554907744, + "learning_rate": 3.3797230590647073e-07, + "loss": 1.782, + "step": 9565 + }, + { + "epoch": 0.6448569792122907, + "grad_norm": 6.251059910397102, + "learning_rate": 3.374161013864515e-07, + "loss": 1.7841, + "step": 9570 + }, + { + "epoch": 0.6451938950843974, + "grad_norm": 6.216762679041962, + "learning_rate": 3.368601217534661e-07, + "loss": 1.8344, + "step": 9575 + }, + { + "epoch": 0.6455308109565041, + "grad_norm": 6.1689755886139235, + "learning_rate": 3.3630436777654903e-07, + "loss": 1.8488, + "step": 9580 + }, + { + "epoch": 0.6458677268286109, + "grad_norm": 6.150257163492163, + "learning_rate": 3.357488402244227e-07, + "loss": 1.8074, + "step": 9585 + }, + { + "epoch": 0.6462046427007176, + "grad_norm": 6.405818650009187, + "learning_rate": 3.3519353986549604e-07, + "loss": 1.7121, + "step": 9590 + }, + { + "epoch": 0.6465415585728244, + "grad_norm": 6.4015349458926645, + "learning_rate": 3.346384674678639e-07, + "loss": 1.8637, + "step": 9595 + }, + { + "epoch": 0.6468784744449311, + "grad_norm": 6.022453268076306, + "learning_rate": 3.3408362379930576e-07, + "loss": 1.7349, + "step": 9600 + }, + { + "epoch": 0.6472153903170378, + "grad_norm": 5.707068311915508, + "learning_rate": 3.335290096272849e-07, + "loss": 1.7547, + "step": 9605 + }, + { + "epoch": 0.6475523061891446, + "grad_norm": 6.157901580868346, + "learning_rate": 3.3297462571894673e-07, + "loss": 1.7596, + "step": 9610 + }, + { + "epoch": 0.6478892220612513, + "grad_norm": 5.651431847485237, + "learning_rate": 3.3242047284111857e-07, + "loss": 1.7731, + "step": 9615 + }, + { + "epoch": 0.648226137933358, + "grad_norm": 5.84959347365055, + "learning_rate": 3.3186655176030826e-07, + "loss": 1.8499, + "step": 9620 + }, + { + "epoch": 0.6485630538054648, + "grad_norm": 5.442395829120497, + "learning_rate": 3.3131286324270234e-07, + "loss": 1.6875, + "step": 9625 + }, + { + "epoch": 0.6488999696775715, + "grad_norm": 5.576850818923356, + "learning_rate": 3.3075940805416654e-07, + "loss": 1.7739, + "step": 9630 + }, + { + "epoch": 0.6492368855496783, + "grad_norm": 5.776880524135601, + "learning_rate": 3.3020618696024316e-07, + "loss": 1.8221, + "step": 9635 + }, + { + "epoch": 0.649573801421785, + "grad_norm": 6.1563117132326, + "learning_rate": 3.2965320072615113e-07, + "loss": 1.83, + "step": 9640 + }, + { + "epoch": 0.6499107172938917, + "grad_norm": 6.116669362115531, + "learning_rate": 3.2910045011678424e-07, + "loss": 1.7656, + "step": 9645 + }, + { + "epoch": 0.6502476331659984, + "grad_norm": 5.766920463158223, + "learning_rate": 3.2854793589671046e-07, + "loss": 1.8503, + "step": 9650 + }, + { + "epoch": 0.6505845490381051, + "grad_norm": 5.7263462025342875, + "learning_rate": 3.279956588301712e-07, + "loss": 1.7948, + "step": 9655 + }, + { + "epoch": 0.650921464910212, + "grad_norm": 5.62578071524695, + "learning_rate": 3.274436196810789e-07, + "loss": 1.7502, + "step": 9660 + }, + { + "epoch": 0.6512583807823187, + "grad_norm": 6.186976960735761, + "learning_rate": 3.268918192130178e-07, + "loss": 1.7973, + "step": 9665 + }, + { + "epoch": 0.6515952966544254, + "grad_norm": 6.454846474675306, + "learning_rate": 3.263402581892415e-07, + "loss": 1.7346, + "step": 9670 + }, + { + "epoch": 0.6519322125265321, + "grad_norm": 6.0382204126825885, + "learning_rate": 3.257889373726726e-07, + "loss": 1.843, + "step": 9675 + }, + { + "epoch": 0.6522691283986388, + "grad_norm": 5.732458930505232, + "learning_rate": 3.252378575259013e-07, + "loss": 1.7876, + "step": 9680 + }, + { + "epoch": 0.6526060442707456, + "grad_norm": 5.8367309166096915, + "learning_rate": 3.246870194111849e-07, + "loss": 1.7943, + "step": 9685 + }, + { + "epoch": 0.6529429601428524, + "grad_norm": 6.303598679052027, + "learning_rate": 3.2413642379044557e-07, + "loss": 1.7715, + "step": 9690 + }, + { + "epoch": 0.6532798760149591, + "grad_norm": 6.023790656021271, + "learning_rate": 3.235860714252708e-07, + "loss": 1.8188, + "step": 9695 + }, + { + "epoch": 0.6536167918870658, + "grad_norm": 5.831181231694035, + "learning_rate": 3.2303596307691137e-07, + "loss": 1.7846, + "step": 9700 + }, + { + "epoch": 0.6539537077591725, + "grad_norm": 5.747323594425591, + "learning_rate": 3.2248609950628023e-07, + "loss": 1.8437, + "step": 9705 + }, + { + "epoch": 0.6542906236312793, + "grad_norm": 6.344734553809549, + "learning_rate": 3.219364814739522e-07, + "loss": 1.7766, + "step": 9710 + }, + { + "epoch": 0.654627539503386, + "grad_norm": 6.531726147190786, + "learning_rate": 3.2138710974016226e-07, + "loss": 1.7615, + "step": 9715 + }, + { + "epoch": 0.6549644553754927, + "grad_norm": 5.850481024867147, + "learning_rate": 3.208379850648046e-07, + "loss": 1.8543, + "step": 9720 + }, + { + "epoch": 0.6553013712475995, + "grad_norm": 5.845927633738255, + "learning_rate": 3.202891082074318e-07, + "loss": 1.8064, + "step": 9725 + }, + { + "epoch": 0.6556382871197062, + "grad_norm": 5.677222851765273, + "learning_rate": 3.197404799272537e-07, + "loss": 1.7991, + "step": 9730 + }, + { + "epoch": 0.6559752029918129, + "grad_norm": 6.57853406738191, + "learning_rate": 3.191921009831365e-07, + "loss": 1.7932, + "step": 9735 + }, + { + "epoch": 0.6563121188639197, + "grad_norm": 6.208108133755243, + "learning_rate": 3.1864397213360093e-07, + "loss": 1.8054, + "step": 9740 + }, + { + "epoch": 0.6566490347360264, + "grad_norm": 6.360434354173842, + "learning_rate": 3.180960941368223e-07, + "loss": 1.8019, + "step": 9745 + }, + { + "epoch": 0.6569859506081331, + "grad_norm": 5.962283203943862, + "learning_rate": 3.175484677506288e-07, + "loss": 1.7591, + "step": 9750 + }, + { + "epoch": 0.6573228664802399, + "grad_norm": 6.134085347065576, + "learning_rate": 3.1700109373250063e-07, + "loss": 1.7648, + "step": 9755 + }, + { + "epoch": 0.6576597823523466, + "grad_norm": 6.481907425268963, + "learning_rate": 3.1645397283956843e-07, + "loss": 1.7872, + "step": 9760 + }, + { + "epoch": 0.6579966982244534, + "grad_norm": 5.6435553605771664, + "learning_rate": 3.159071058286138e-07, + "loss": 1.7808, + "step": 9765 + }, + { + "epoch": 0.6583336140965601, + "grad_norm": 5.912519778324625, + "learning_rate": 3.1536049345606586e-07, + "loss": 1.8875, + "step": 9770 + }, + { + "epoch": 0.6586705299686668, + "grad_norm": 5.678335891008834, + "learning_rate": 3.1481413647800247e-07, + "loss": 1.663, + "step": 9775 + }, + { + "epoch": 0.6590074458407735, + "grad_norm": 5.9482407468813, + "learning_rate": 3.14268035650148e-07, + "loss": 1.8166, + "step": 9780 + }, + { + "epoch": 0.6593443617128802, + "grad_norm": 5.722651494191131, + "learning_rate": 3.137221917278723e-07, + "loss": 1.8757, + "step": 9785 + }, + { + "epoch": 0.6596812775849871, + "grad_norm": 6.054054603796881, + "learning_rate": 3.1317660546618986e-07, + "loss": 1.7174, + "step": 9790 + }, + { + "epoch": 0.6600181934570938, + "grad_norm": 6.112296477029174, + "learning_rate": 3.1263127761975917e-07, + "loss": 1.804, + "step": 9795 + }, + { + "epoch": 0.6603551093292005, + "grad_norm": 6.24833577705031, + "learning_rate": 3.12086208942881e-07, + "loss": 1.83, + "step": 9800 + }, + { + "epoch": 0.6606920252013072, + "grad_norm": 5.796380373627383, + "learning_rate": 3.1154140018949736e-07, + "loss": 1.7987, + "step": 9805 + }, + { + "epoch": 0.6610289410734139, + "grad_norm": 6.091806447131728, + "learning_rate": 3.1099685211319116e-07, + "loss": 1.7743, + "step": 9810 + }, + { + "epoch": 0.6613658569455207, + "grad_norm": 6.099517758922633, + "learning_rate": 3.104525654671849e-07, + "loss": 1.7681, + "step": 9815 + }, + { + "epoch": 0.6617027728176275, + "grad_norm": 6.068348509741613, + "learning_rate": 3.099085410043386e-07, + "loss": 1.7287, + "step": 9820 + }, + { + "epoch": 0.6620396886897342, + "grad_norm": 5.435497802729357, + "learning_rate": 3.0936477947715064e-07, + "loss": 1.7398, + "step": 9825 + }, + { + "epoch": 0.6623766045618409, + "grad_norm": 5.978768977959412, + "learning_rate": 3.088212816377552e-07, + "loss": 1.7762, + "step": 9830 + }, + { + "epoch": 0.6627135204339476, + "grad_norm": 5.529402155256499, + "learning_rate": 3.0827804823792157e-07, + "loss": 1.7906, + "step": 9835 + }, + { + "epoch": 0.6630504363060544, + "grad_norm": 5.745634282596181, + "learning_rate": 3.077350800290537e-07, + "loss": 1.8153, + "step": 9840 + }, + { + "epoch": 0.6633873521781611, + "grad_norm": 5.678976914931346, + "learning_rate": 3.071923777621885e-07, + "loss": 1.7453, + "step": 9845 + }, + { + "epoch": 0.6637242680502679, + "grad_norm": 6.0133152567955275, + "learning_rate": 3.066499421879948e-07, + "loss": 1.887, + "step": 9850 + }, + { + "epoch": 0.6640611839223746, + "grad_norm": 6.145888236549586, + "learning_rate": 3.0610777405677286e-07, + "loss": 1.7914, + "step": 9855 + }, + { + "epoch": 0.6643980997944813, + "grad_norm": 5.481352078753557, + "learning_rate": 3.05565874118453e-07, + "loss": 1.7645, + "step": 9860 + }, + { + "epoch": 0.6647350156665881, + "grad_norm": 6.155766828058924, + "learning_rate": 3.050242431225948e-07, + "loss": 1.7972, + "step": 9865 + }, + { + "epoch": 0.6650719315386948, + "grad_norm": 6.096995270749912, + "learning_rate": 3.0448288181838487e-07, + "loss": 1.7869, + "step": 9870 + }, + { + "epoch": 0.6654088474108015, + "grad_norm": 5.923689399349553, + "learning_rate": 3.0394179095463804e-07, + "loss": 1.8212, + "step": 9875 + }, + { + "epoch": 0.6657457632829082, + "grad_norm": 6.183785970604984, + "learning_rate": 3.0340097127979426e-07, + "loss": 1.8298, + "step": 9880 + }, + { + "epoch": 0.666082679155015, + "grad_norm": 6.130929170868858, + "learning_rate": 3.0286042354191844e-07, + "loss": 1.8104, + "step": 9885 + }, + { + "epoch": 0.6664195950271218, + "grad_norm": 5.884804572018233, + "learning_rate": 3.0232014848869955e-07, + "loss": 1.8011, + "step": 9890 + }, + { + "epoch": 0.6667565108992285, + "grad_norm": 6.118852994598744, + "learning_rate": 3.0178014686744966e-07, + "loss": 1.761, + "step": 9895 + }, + { + "epoch": 0.6670934267713352, + "grad_norm": 5.391353549438332, + "learning_rate": 3.0124041942510175e-07, + "loss": 1.8032, + "step": 9900 + }, + { + "epoch": 0.6674303426434419, + "grad_norm": 6.996706242225225, + "learning_rate": 3.007009669082103e-07, + "loss": 1.7749, + "step": 9905 + }, + { + "epoch": 0.6677672585155486, + "grad_norm": 5.769021683253544, + "learning_rate": 3.001617900629496e-07, + "loss": 1.8067, + "step": 9910 + }, + { + "epoch": 0.6681041743876553, + "grad_norm": 6.106192789372139, + "learning_rate": 2.996228896351119e-07, + "loss": 1.8022, + "step": 9915 + }, + { + "epoch": 0.6684410902597622, + "grad_norm": 5.8423030937909335, + "learning_rate": 2.9908426637010773e-07, + "loss": 1.8518, + "step": 9920 + }, + { + "epoch": 0.6687780061318689, + "grad_norm": 5.975794455383276, + "learning_rate": 2.98545921012964e-07, + "loss": 1.7861, + "step": 9925 + }, + { + "epoch": 0.6691149220039756, + "grad_norm": 6.185465462038598, + "learning_rate": 2.9800785430832354e-07, + "loss": 1.751, + "step": 9930 + }, + { + "epoch": 0.6694518378760823, + "grad_norm": 5.959682622502289, + "learning_rate": 2.9747006700044295e-07, + "loss": 1.7959, + "step": 9935 + }, + { + "epoch": 0.669788753748189, + "grad_norm": 5.644052495462014, + "learning_rate": 2.969325598331932e-07, + "loss": 1.7526, + "step": 9940 + }, + { + "epoch": 0.6701256696202958, + "grad_norm": 5.560965520647446, + "learning_rate": 2.9639533355005773e-07, + "loss": 1.7729, + "step": 9945 + }, + { + "epoch": 0.6704625854924026, + "grad_norm": 5.6497755675616865, + "learning_rate": 2.958583888941306e-07, + "loss": 1.7338, + "step": 9950 + }, + { + "epoch": 0.6707995013645093, + "grad_norm": 6.305541997526218, + "learning_rate": 2.9532172660811745e-07, + "loss": 1.8171, + "step": 9955 + }, + { + "epoch": 0.671136417236616, + "grad_norm": 5.908429580273309, + "learning_rate": 2.9478534743433247e-07, + "loss": 1.7819, + "step": 9960 + }, + { + "epoch": 0.6714733331087227, + "grad_norm": 5.222620415906802, + "learning_rate": 2.9424925211469876e-07, + "loss": 1.7729, + "step": 9965 + }, + { + "epoch": 0.6718102489808295, + "grad_norm": 6.572254603044038, + "learning_rate": 2.9371344139074645e-07, + "loss": 1.7642, + "step": 9970 + }, + { + "epoch": 0.6721471648529362, + "grad_norm": 6.327822112736418, + "learning_rate": 2.9317791600361243e-07, + "loss": 1.6583, + "step": 9975 + }, + { + "epoch": 0.672484080725043, + "grad_norm": 5.574113505265418, + "learning_rate": 2.9264267669403833e-07, + "loss": 1.8007, + "step": 9980 + }, + { + "epoch": 0.6728209965971497, + "grad_norm": 5.98463445361315, + "learning_rate": 2.921077242023706e-07, + "loss": 1.7971, + "step": 9985 + }, + { + "epoch": 0.6731579124692564, + "grad_norm": 6.304951408982249, + "learning_rate": 2.9157305926855893e-07, + "loss": 1.7754, + "step": 9990 + }, + { + "epoch": 0.6734948283413632, + "grad_norm": 5.969025928551944, + "learning_rate": 2.910386826321549e-07, + "loss": 1.8124, + "step": 9995 + }, + { + "epoch": 0.6738317442134699, + "grad_norm": 6.2610465949388034, + "learning_rate": 2.905045950323114e-07, + "loss": 1.7963, + "step": 10000 + }, + { + "epoch": 0.6741686600855766, + "grad_norm": 6.168419758973521, + "learning_rate": 2.899707972077817e-07, + "loss": 1.7718, + "step": 10005 + }, + { + "epoch": 0.6745055759576833, + "grad_norm": 6.8413981566739155, + "learning_rate": 2.8943728989691857e-07, + "loss": 1.8168, + "step": 10010 + }, + { + "epoch": 0.6748424918297901, + "grad_norm": 6.184366090083486, + "learning_rate": 2.88904073837672e-07, + "loss": 1.7857, + "step": 10015 + }, + { + "epoch": 0.6751794077018969, + "grad_norm": 5.907308449749499, + "learning_rate": 2.883711497675899e-07, + "loss": 1.7391, + "step": 10020 + }, + { + "epoch": 0.6755163235740036, + "grad_norm": 5.657863323168557, + "learning_rate": 2.878385184238163e-07, + "loss": 1.7462, + "step": 10025 + }, + { + "epoch": 0.6758532394461103, + "grad_norm": 6.0183947518407885, + "learning_rate": 2.8730618054308964e-07, + "loss": 1.7306, + "step": 10030 + }, + { + "epoch": 0.676190155318217, + "grad_norm": 5.743693775609563, + "learning_rate": 2.8677413686174325e-07, + "loss": 1.8358, + "step": 10035 + }, + { + "epoch": 0.6765270711903237, + "grad_norm": 6.764854467811837, + "learning_rate": 2.8624238811570325e-07, + "loss": 1.7915, + "step": 10040 + }, + { + "epoch": 0.6768639870624306, + "grad_norm": 5.519785427555744, + "learning_rate": 2.8571093504048737e-07, + "loss": 1.7907, + "step": 10045 + }, + { + "epoch": 0.6772009029345373, + "grad_norm": 5.738090769801983, + "learning_rate": 2.851797783712049e-07, + "loss": 1.709, + "step": 10050 + }, + { + "epoch": 0.677537818806644, + "grad_norm": 6.1121353044767055, + "learning_rate": 2.8464891884255515e-07, + "loss": 1.8104, + "step": 10055 + }, + { + "epoch": 0.6778747346787507, + "grad_norm": 6.082231792635815, + "learning_rate": 2.8411835718882593e-07, + "loss": 1.8035, + "step": 10060 + }, + { + "epoch": 0.6782116505508574, + "grad_norm": 5.757889132681298, + "learning_rate": 2.835880941438936e-07, + "loss": 1.7574, + "step": 10065 + }, + { + "epoch": 0.6785485664229642, + "grad_norm": 5.8861819403022295, + "learning_rate": 2.8305813044122093e-07, + "loss": 1.7928, + "step": 10070 + }, + { + "epoch": 0.678885482295071, + "grad_norm": 6.509539103297996, + "learning_rate": 2.8252846681385734e-07, + "loss": 1.8665, + "step": 10075 + }, + { + "epoch": 0.6792223981671777, + "grad_norm": 6.3546347574074655, + "learning_rate": 2.8199910399443625e-07, + "loss": 1.8221, + "step": 10080 + }, + { + "epoch": 0.6795593140392844, + "grad_norm": 5.934429157563714, + "learning_rate": 2.8147004271517584e-07, + "loss": 1.7238, + "step": 10085 + }, + { + "epoch": 0.6798962299113911, + "grad_norm": 5.866984996633733, + "learning_rate": 2.8094128370787694e-07, + "loss": 1.8562, + "step": 10090 + }, + { + "epoch": 0.6802331457834978, + "grad_norm": 6.064723412686676, + "learning_rate": 2.8041282770392196e-07, + "loss": 1.7556, + "step": 10095 + }, + { + "epoch": 0.6805700616556046, + "grad_norm": 6.549753460763149, + "learning_rate": 2.7988467543427454e-07, + "loss": 1.722, + "step": 10100 + }, + { + "epoch": 0.6809069775277113, + "grad_norm": 6.655895352398308, + "learning_rate": 2.7935682762947837e-07, + "loss": 1.8285, + "step": 10105 + }, + { + "epoch": 0.6812438933998181, + "grad_norm": 5.802651770334309, + "learning_rate": 2.788292850196553e-07, + "loss": 1.7974, + "step": 10110 + }, + { + "epoch": 0.6815808092719248, + "grad_norm": 5.861414227938647, + "learning_rate": 2.783020483345057e-07, + "loss": 1.737, + "step": 10115 + }, + { + "epoch": 0.6819177251440315, + "grad_norm": 5.698821956793516, + "learning_rate": 2.777751183033067e-07, + "loss": 1.8651, + "step": 10120 + }, + { + "epoch": 0.6822546410161383, + "grad_norm": 5.688820682473993, + "learning_rate": 2.772484956549107e-07, + "loss": 1.7865, + "step": 10125 + }, + { + "epoch": 0.682591556888245, + "grad_norm": 6.358195618306636, + "learning_rate": 2.7672218111774566e-07, + "loss": 1.8499, + "step": 10130 + }, + { + "epoch": 0.6829284727603517, + "grad_norm": 5.6611682532718985, + "learning_rate": 2.7619617541981287e-07, + "loss": 1.7677, + "step": 10135 + }, + { + "epoch": 0.6832653886324584, + "grad_norm": 6.2197481978105325, + "learning_rate": 2.756704792886869e-07, + "loss": 1.7721, + "step": 10140 + }, + { + "epoch": 0.6836023045045652, + "grad_norm": 5.844100559661615, + "learning_rate": 2.7514509345151347e-07, + "loss": 1.7516, + "step": 10145 + }, + { + "epoch": 0.683939220376672, + "grad_norm": 6.384694537823895, + "learning_rate": 2.746200186350097e-07, + "loss": 1.8386, + "step": 10150 + }, + { + "epoch": 0.6842761362487787, + "grad_norm": 5.510665281210932, + "learning_rate": 2.740952555654622e-07, + "loss": 1.8099, + "step": 10155 + }, + { + "epoch": 0.6846130521208854, + "grad_norm": 6.1634268813462985, + "learning_rate": 2.735708049687262e-07, + "loss": 1.7247, + "step": 10160 + }, + { + "epoch": 0.6849499679929921, + "grad_norm": 5.964714807031205, + "learning_rate": 2.730466675702251e-07, + "loss": 1.8141, + "step": 10165 + }, + { + "epoch": 0.6852868838650988, + "grad_norm": 6.170392267124102, + "learning_rate": 2.7252284409494906e-07, + "loss": 1.7619, + "step": 10170 + }, + { + "epoch": 0.6856237997372057, + "grad_norm": 5.509149793923222, + "learning_rate": 2.7199933526745364e-07, + "loss": 1.8118, + "step": 10175 + }, + { + "epoch": 0.6859607156093124, + "grad_norm": 6.54628143404462, + "learning_rate": 2.714761418118596e-07, + "loss": 1.7321, + "step": 10180 + }, + { + "epoch": 0.6862976314814191, + "grad_norm": 7.2477390544693945, + "learning_rate": 2.7095326445185143e-07, + "loss": 1.8118, + "step": 10185 + }, + { + "epoch": 0.6866345473535258, + "grad_norm": 5.822832431612233, + "learning_rate": 2.704307039106759e-07, + "loss": 1.7449, + "step": 10190 + }, + { + "epoch": 0.6869714632256325, + "grad_norm": 5.624150215875334, + "learning_rate": 2.6990846091114205e-07, + "loss": 1.8633, + "step": 10195 + }, + { + "epoch": 0.6873083790977393, + "grad_norm": 5.557146257249171, + "learning_rate": 2.6938653617561967e-07, + "loss": 1.8189, + "step": 10200 + }, + { + "epoch": 0.687645294969846, + "grad_norm": 5.6563267734125935, + "learning_rate": 2.688649304260383e-07, + "loss": 1.7889, + "step": 10205 + }, + { + "epoch": 0.6879822108419528, + "grad_norm": 6.075955336314197, + "learning_rate": 2.683436443838859e-07, + "loss": 1.81, + "step": 10210 + }, + { + "epoch": 0.6883191267140595, + "grad_norm": 6.259216573221612, + "learning_rate": 2.678226787702086e-07, + "loss": 1.8303, + "step": 10215 + }, + { + "epoch": 0.6886560425861662, + "grad_norm": 5.884748810679717, + "learning_rate": 2.673020343056094e-07, + "loss": 1.8741, + "step": 10220 + }, + { + "epoch": 0.688992958458273, + "grad_norm": 6.144015918665497, + "learning_rate": 2.6678171171024657e-07, + "loss": 1.8529, + "step": 10225 + }, + { + "epoch": 0.6893298743303797, + "grad_norm": 5.50532456214314, + "learning_rate": 2.6626171170383373e-07, + "loss": 1.7156, + "step": 10230 + }, + { + "epoch": 0.6896667902024864, + "grad_norm": 5.9031282172064605, + "learning_rate": 2.6574203500563776e-07, + "loss": 1.8002, + "step": 10235 + }, + { + "epoch": 0.6900037060745932, + "grad_norm": 5.789618897247442, + "learning_rate": 2.6522268233447894e-07, + "loss": 1.8153, + "step": 10240 + }, + { + "epoch": 0.6903406219466999, + "grad_norm": 5.9841417775446475, + "learning_rate": 2.6470365440872866e-07, + "loss": 1.7516, + "step": 10245 + }, + { + "epoch": 0.6906775378188067, + "grad_norm": 6.137156396832881, + "learning_rate": 2.641849519463099e-07, + "loss": 1.8213, + "step": 10250 + }, + { + "epoch": 0.6910144536909134, + "grad_norm": 5.63256141938642, + "learning_rate": 2.6366657566469465e-07, + "loss": 1.763, + "step": 10255 + }, + { + "epoch": 0.6913513695630201, + "grad_norm": 5.7892144401063605, + "learning_rate": 2.631485262809043e-07, + "loss": 1.7543, + "step": 10260 + }, + { + "epoch": 0.6916882854351268, + "grad_norm": 5.635590710985062, + "learning_rate": 2.6263080451150797e-07, + "loss": 1.7936, + "step": 10265 + }, + { + "epoch": 0.6920252013072336, + "grad_norm": 6.226656648440483, + "learning_rate": 2.621134110726217e-07, + "loss": 1.8707, + "step": 10270 + }, + { + "epoch": 0.6923621171793403, + "grad_norm": 6.159797742322993, + "learning_rate": 2.6159634667990683e-07, + "loss": 1.8383, + "step": 10275 + }, + { + "epoch": 0.6926990330514471, + "grad_norm": 6.186357246890332, + "learning_rate": 2.610796120485701e-07, + "loss": 1.7995, + "step": 10280 + }, + { + "epoch": 0.6930359489235538, + "grad_norm": 5.607970114608954, + "learning_rate": 2.605632078933623e-07, + "loss": 1.7274, + "step": 10285 + }, + { + "epoch": 0.6933728647956605, + "grad_norm": 6.434772493415341, + "learning_rate": 2.600471349285763e-07, + "loss": 1.8332, + "step": 10290 + }, + { + "epoch": 0.6937097806677672, + "grad_norm": 6.214320065920484, + "learning_rate": 2.5953139386804764e-07, + "loss": 1.79, + "step": 10295 + }, + { + "epoch": 0.6940466965398739, + "grad_norm": 6.399925193314172, + "learning_rate": 2.5901598542515256e-07, + "loss": 1.7514, + "step": 10300 + }, + { + "epoch": 0.6943836124119808, + "grad_norm": 5.732194115105609, + "learning_rate": 2.5850091031280684e-07, + "loss": 1.8028, + "step": 10305 + }, + { + "epoch": 0.6947205282840875, + "grad_norm": 5.820562272225368, + "learning_rate": 2.579861692434658e-07, + "loss": 1.7914, + "step": 10310 + }, + { + "epoch": 0.6950574441561942, + "grad_norm": 5.970701069273288, + "learning_rate": 2.574717629291222e-07, + "loss": 1.7709, + "step": 10315 + }, + { + "epoch": 0.6953943600283009, + "grad_norm": 5.609714862141825, + "learning_rate": 2.5695769208130615e-07, + "loss": 1.7738, + "step": 10320 + }, + { + "epoch": 0.6957312759004076, + "grad_norm": 5.730048947871614, + "learning_rate": 2.564439574110833e-07, + "loss": 1.7921, + "step": 10325 + }, + { + "epoch": 0.6960681917725144, + "grad_norm": 5.885438524238012, + "learning_rate": 2.559305596290547e-07, + "loss": 1.814, + "step": 10330 + }, + { + "epoch": 0.6964051076446212, + "grad_norm": 5.999810851894432, + "learning_rate": 2.554174994453555e-07, + "loss": 1.7399, + "step": 10335 + }, + { + "epoch": 0.6967420235167279, + "grad_norm": 5.971456247977796, + "learning_rate": 2.549047775696532e-07, + "loss": 1.7531, + "step": 10340 + }, + { + "epoch": 0.6970789393888346, + "grad_norm": 6.045304895294886, + "learning_rate": 2.543923947111481e-07, + "loss": 1.7802, + "step": 10345 + }, + { + "epoch": 0.6974158552609413, + "grad_norm": 5.898273349939425, + "learning_rate": 2.538803515785714e-07, + "loss": 1.8081, + "step": 10350 + }, + { + "epoch": 0.6977527711330481, + "grad_norm": 5.716447889359928, + "learning_rate": 2.5336864888018393e-07, + "loss": 1.7536, + "step": 10355 + }, + { + "epoch": 0.6980896870051548, + "grad_norm": 6.1167888708064, + "learning_rate": 2.528572873237761e-07, + "loss": 1.7771, + "step": 10360 + }, + { + "epoch": 0.6984266028772615, + "grad_norm": 5.814373279694408, + "learning_rate": 2.5234626761666647e-07, + "loss": 1.8857, + "step": 10365 + }, + { + "epoch": 0.6987635187493683, + "grad_norm": 5.595119203770916, + "learning_rate": 2.5183559046570036e-07, + "loss": 1.8684, + "step": 10370 + }, + { + "epoch": 0.699100434621475, + "grad_norm": 5.746260139787474, + "learning_rate": 2.513252565772496e-07, + "loss": 1.8153, + "step": 10375 + }, + { + "epoch": 0.6994373504935818, + "grad_norm": 6.070923517455952, + "learning_rate": 2.5081526665721133e-07, + "loss": 1.779, + "step": 10380 + }, + { + "epoch": 0.6997742663656885, + "grad_norm": 5.654721545450662, + "learning_rate": 2.503056214110062e-07, + "loss": 1.7561, + "step": 10385 + }, + { + "epoch": 0.7001111822377952, + "grad_norm": 6.076854962239542, + "learning_rate": 2.497963215435789e-07, + "loss": 1.7356, + "step": 10390 + }, + { + "epoch": 0.7004480981099019, + "grad_norm": 6.267075920486195, + "learning_rate": 2.492873677593964e-07, + "loss": 1.72, + "step": 10395 + }, + { + "epoch": 0.7007850139820087, + "grad_norm": 6.495036916039556, + "learning_rate": 2.4877876076244626e-07, + "loss": 1.8802, + "step": 10400 + }, + { + "epoch": 0.7011219298541155, + "grad_norm": 6.268581169228354, + "learning_rate": 2.482705012562367e-07, + "loss": 1.7963, + "step": 10405 + }, + { + "epoch": 0.7014588457262222, + "grad_norm": 5.710070783985173, + "learning_rate": 2.4776258994379546e-07, + "loss": 1.8051, + "step": 10410 + }, + { + "epoch": 0.7017957615983289, + "grad_norm": 6.111357846006603, + "learning_rate": 2.4725502752766883e-07, + "loss": 1.795, + "step": 10415 + }, + { + "epoch": 0.7021326774704356, + "grad_norm": 5.907713749893674, + "learning_rate": 2.4674781470991967e-07, + "loss": 1.7229, + "step": 10420 + }, + { + "epoch": 0.7024695933425423, + "grad_norm": 6.204994213224908, + "learning_rate": 2.462409521921282e-07, + "loss": 1.761, + "step": 10425 + }, + { + "epoch": 0.7028065092146492, + "grad_norm": 5.823631147973025, + "learning_rate": 2.4573444067538985e-07, + "loss": 1.8227, + "step": 10430 + }, + { + "epoch": 0.7031434250867559, + "grad_norm": 6.33472545976378, + "learning_rate": 2.4522828086031404e-07, + "loss": 1.8701, + "step": 10435 + }, + { + "epoch": 0.7034803409588626, + "grad_norm": 6.345376423431866, + "learning_rate": 2.4472247344702424e-07, + "loss": 1.7783, + "step": 10440 + }, + { + "epoch": 0.7038172568309693, + "grad_norm": 6.2754842195154685, + "learning_rate": 2.442170191351566e-07, + "loss": 1.8856, + "step": 10445 + }, + { + "epoch": 0.704154172703076, + "grad_norm": 6.106369137710223, + "learning_rate": 2.4371191862385816e-07, + "loss": 1.7274, + "step": 10450 + }, + { + "epoch": 0.7044910885751827, + "grad_norm": 6.1777861768742675, + "learning_rate": 2.4320717261178715e-07, + "loss": 1.808, + "step": 10455 + }, + { + "epoch": 0.7048280044472895, + "grad_norm": 6.247165515267102, + "learning_rate": 2.4270278179711163e-07, + "loss": 1.7741, + "step": 10460 + }, + { + "epoch": 0.7051649203193963, + "grad_norm": 6.397283503643184, + "learning_rate": 2.4219874687750754e-07, + "loss": 1.8014, + "step": 10465 + }, + { + "epoch": 0.705501836191503, + "grad_norm": 6.283686491438517, + "learning_rate": 2.4169506855015923e-07, + "loss": 1.8509, + "step": 10470 + }, + { + "epoch": 0.7058387520636097, + "grad_norm": 6.181410447908662, + "learning_rate": 2.4119174751175787e-07, + "loss": 1.7941, + "step": 10475 + }, + { + "epoch": 0.7061756679357164, + "grad_norm": 6.104807693363187, + "learning_rate": 2.406887844584998e-07, + "loss": 1.7714, + "step": 10480 + }, + { + "epoch": 0.7065125838078232, + "grad_norm": 5.877543591957879, + "learning_rate": 2.401861800860868e-07, + "loss": 1.7639, + "step": 10485 + }, + { + "epoch": 0.7068494996799299, + "grad_norm": 5.546451747516014, + "learning_rate": 2.396839350897241e-07, + "loss": 1.8287, + "step": 10490 + }, + { + "epoch": 0.7071864155520367, + "grad_norm": 6.258500520120998, + "learning_rate": 2.391820501641203e-07, + "loss": 1.8497, + "step": 10495 + }, + { + "epoch": 0.7075233314241434, + "grad_norm": 6.476113202277884, + "learning_rate": 2.3868052600348524e-07, + "loss": 1.8027, + "step": 10500 + }, + { + "epoch": 0.7078602472962501, + "grad_norm": 5.551605271516278, + "learning_rate": 2.381793633015305e-07, + "loss": 1.7701, + "step": 10505 + }, + { + "epoch": 0.7081971631683569, + "grad_norm": 6.2164882176078065, + "learning_rate": 2.3767856275146748e-07, + "loss": 1.7781, + "step": 10510 + }, + { + "epoch": 0.7085340790404636, + "grad_norm": 5.782676813508456, + "learning_rate": 2.3717812504600616e-07, + "loss": 1.7576, + "step": 10515 + }, + { + "epoch": 0.7088709949125703, + "grad_norm": 5.835660530958148, + "learning_rate": 2.3667805087735516e-07, + "loss": 1.8002, + "step": 10520 + }, + { + "epoch": 0.709207910784677, + "grad_norm": 5.988522457949844, + "learning_rate": 2.3617834093722033e-07, + "loss": 1.7542, + "step": 10525 + }, + { + "epoch": 0.7095448266567838, + "grad_norm": 6.439152375610268, + "learning_rate": 2.3567899591680317e-07, + "loss": 1.7935, + "step": 10530 + }, + { + "epoch": 0.7098817425288906, + "grad_norm": 5.567500010220148, + "learning_rate": 2.351800165068008e-07, + "loss": 1.7577, + "step": 10535 + }, + { + "epoch": 0.7102186584009973, + "grad_norm": 5.603156857304425, + "learning_rate": 2.346814033974047e-07, + "loss": 1.8004, + "step": 10540 + }, + { + "epoch": 0.710555574273104, + "grad_norm": 6.1572449811841565, + "learning_rate": 2.3418315727829962e-07, + "loss": 1.8318, + "step": 10545 + }, + { + "epoch": 0.7108924901452107, + "grad_norm": 6.078503199343157, + "learning_rate": 2.336852788386623e-07, + "loss": 1.8324, + "step": 10550 + }, + { + "epoch": 0.7112294060173174, + "grad_norm": 6.059107998684295, + "learning_rate": 2.331877687671614e-07, + "loss": 1.7983, + "step": 10555 + }, + { + "epoch": 0.7115663218894243, + "grad_norm": 5.977209253927193, + "learning_rate": 2.3269062775195596e-07, + "loss": 1.7753, + "step": 10560 + }, + { + "epoch": 0.711903237761531, + "grad_norm": 5.965166803624046, + "learning_rate": 2.321938564806944e-07, + "loss": 1.839, + "step": 10565 + }, + { + "epoch": 0.7122401536336377, + "grad_norm": 6.191642482071603, + "learning_rate": 2.3169745564051353e-07, + "loss": 1.7777, + "step": 10570 + }, + { + "epoch": 0.7125770695057444, + "grad_norm": 5.93892798040206, + "learning_rate": 2.3120142591803825e-07, + "loss": 1.8269, + "step": 10575 + }, + { + "epoch": 0.7129139853778511, + "grad_norm": 6.285189080508176, + "learning_rate": 2.307057679993797e-07, + "loss": 1.739, + "step": 10580 + }, + { + "epoch": 0.7132509012499579, + "grad_norm": 5.737737960252139, + "learning_rate": 2.30210482570135e-07, + "loss": 1.8098, + "step": 10585 + }, + { + "epoch": 0.7135878171220646, + "grad_norm": 5.784964645632823, + "learning_rate": 2.2971557031538607e-07, + "loss": 1.7862, + "step": 10590 + }, + { + "epoch": 0.7139247329941714, + "grad_norm": 5.754820146278622, + "learning_rate": 2.2922103191969828e-07, + "loss": 1.8497, + "step": 10595 + }, + { + "epoch": 0.7142616488662781, + "grad_norm": 6.062341451589527, + "learning_rate": 2.2872686806712032e-07, + "loss": 1.758, + "step": 10600 + }, + { + "epoch": 0.7145985647383848, + "grad_norm": 6.037386034436624, + "learning_rate": 2.2823307944118254e-07, + "loss": 1.7276, + "step": 10605 + }, + { + "epoch": 0.7149354806104916, + "grad_norm": 6.117163210581693, + "learning_rate": 2.2773966672489665e-07, + "loss": 1.8482, + "step": 10610 + }, + { + "epoch": 0.7152723964825983, + "grad_norm": 5.934994056880478, + "learning_rate": 2.2724663060075368e-07, + "loss": 1.7635, + "step": 10615 + }, + { + "epoch": 0.715609312354705, + "grad_norm": 5.888424249364107, + "learning_rate": 2.2675397175072437e-07, + "loss": 1.7819, + "step": 10620 + }, + { + "epoch": 0.7159462282268118, + "grad_norm": 5.24051766818146, + "learning_rate": 2.2626169085625762e-07, + "loss": 1.7494, + "step": 10625 + }, + { + "epoch": 0.7162831440989185, + "grad_norm": 5.624907639128319, + "learning_rate": 2.25769788598279e-07, + "loss": 1.7867, + "step": 10630 + }, + { + "epoch": 0.7166200599710252, + "grad_norm": 6.004337111763599, + "learning_rate": 2.2527826565719084e-07, + "loss": 1.7229, + "step": 10635 + }, + { + "epoch": 0.716956975843132, + "grad_norm": 5.898393504503456, + "learning_rate": 2.2478712271287087e-07, + "loss": 1.8171, + "step": 10640 + }, + { + "epoch": 0.7172938917152387, + "grad_norm": 5.941991199667987, + "learning_rate": 2.2429636044467059e-07, + "loss": 1.783, + "step": 10645 + }, + { + "epoch": 0.7176308075873454, + "grad_norm": 5.456828579892596, + "learning_rate": 2.2380597953141573e-07, + "loss": 1.8494, + "step": 10650 + }, + { + "epoch": 0.7179677234594521, + "grad_norm": 5.815674818794205, + "learning_rate": 2.2331598065140396e-07, + "loss": 1.7785, + "step": 10655 + }, + { + "epoch": 0.7183046393315589, + "grad_norm": 5.847892274384262, + "learning_rate": 2.228263644824045e-07, + "loss": 1.8753, + "step": 10660 + }, + { + "epoch": 0.7186415552036657, + "grad_norm": 5.486645013061549, + "learning_rate": 2.2233713170165757e-07, + "loss": 1.7817, + "step": 10665 + }, + { + "epoch": 0.7189784710757724, + "grad_norm": 6.219147030446047, + "learning_rate": 2.2184828298587298e-07, + "loss": 1.8092, + "step": 10670 + }, + { + "epoch": 0.7193153869478791, + "grad_norm": 5.4866490812706195, + "learning_rate": 2.213598190112294e-07, + "loss": 1.7201, + "step": 10675 + }, + { + "epoch": 0.7196523028199858, + "grad_norm": 6.119492342078513, + "learning_rate": 2.2087174045337275e-07, + "loss": 1.7692, + "step": 10680 + }, + { + "epoch": 0.7199892186920925, + "grad_norm": 5.758138381844213, + "learning_rate": 2.2038404798741644e-07, + "loss": 1.7833, + "step": 10685 + }, + { + "epoch": 0.7203261345641994, + "grad_norm": 5.701487797413592, + "learning_rate": 2.1989674228793987e-07, + "loss": 1.7566, + "step": 10690 + }, + { + "epoch": 0.7206630504363061, + "grad_norm": 6.482581377556036, + "learning_rate": 2.1940982402898684e-07, + "loss": 1.8066, + "step": 10695 + }, + { + "epoch": 0.7209999663084128, + "grad_norm": 5.6849295801038355, + "learning_rate": 2.1892329388406582e-07, + "loss": 1.7557, + "step": 10700 + }, + { + "epoch": 0.7213368821805195, + "grad_norm": 5.952259208162528, + "learning_rate": 2.1843715252614847e-07, + "loss": 1.7513, + "step": 10705 + }, + { + "epoch": 0.7216737980526262, + "grad_norm": 6.19002719248764, + "learning_rate": 2.179514006276681e-07, + "loss": 1.801, + "step": 10710 + }, + { + "epoch": 0.722010713924733, + "grad_norm": 5.404105686010048, + "learning_rate": 2.1746603886051978e-07, + "loss": 1.8273, + "step": 10715 + }, + { + "epoch": 0.7223476297968398, + "grad_norm": 5.88089250460163, + "learning_rate": 2.169810678960591e-07, + "loss": 1.768, + "step": 10720 + }, + { + "epoch": 0.7226845456689465, + "grad_norm": 5.449763025065613, + "learning_rate": 2.1649648840510047e-07, + "loss": 1.7353, + "step": 10725 + }, + { + "epoch": 0.7230214615410532, + "grad_norm": 6.096695445268082, + "learning_rate": 2.1601230105791751e-07, + "loss": 1.8199, + "step": 10730 + }, + { + "epoch": 0.7233583774131599, + "grad_norm": 6.397301804989678, + "learning_rate": 2.1552850652424077e-07, + "loss": 1.8495, + "step": 10735 + }, + { + "epoch": 0.7236952932852667, + "grad_norm": 6.029146969394674, + "learning_rate": 2.150451054732581e-07, + "loss": 1.7442, + "step": 10740 + }, + { + "epoch": 0.7240322091573734, + "grad_norm": 6.426485862249062, + "learning_rate": 2.1456209857361246e-07, + "loss": 1.769, + "step": 10745 + }, + { + "epoch": 0.7243691250294801, + "grad_norm": 5.915318242136389, + "learning_rate": 2.1407948649340208e-07, + "loss": 1.7982, + "step": 10750 + }, + { + "epoch": 0.7247060409015869, + "grad_norm": 6.419855312273165, + "learning_rate": 2.1359726990017908e-07, + "loss": 1.7202, + "step": 10755 + }, + { + "epoch": 0.7250429567736936, + "grad_norm": 6.343992364049308, + "learning_rate": 2.13115449460948e-07, + "loss": 1.8065, + "step": 10760 + }, + { + "epoch": 0.7253798726458004, + "grad_norm": 6.240947236635965, + "learning_rate": 2.12634025842166e-07, + "loss": 1.7197, + "step": 10765 + }, + { + "epoch": 0.7257167885179071, + "grad_norm": 6.320174097494002, + "learning_rate": 2.1215299970974132e-07, + "loss": 1.8149, + "step": 10770 + }, + { + "epoch": 0.7260537043900138, + "grad_norm": 5.766045008893195, + "learning_rate": 2.116723717290318e-07, + "loss": 1.8156, + "step": 10775 + }, + { + "epoch": 0.7263906202621205, + "grad_norm": 6.130324245414857, + "learning_rate": 2.111921425648453e-07, + "loss": 1.8454, + "step": 10780 + }, + { + "epoch": 0.7267275361342272, + "grad_norm": 6.192434248424656, + "learning_rate": 2.1071231288143777e-07, + "loss": 1.8161, + "step": 10785 + }, + { + "epoch": 0.7270644520063341, + "grad_norm": 5.846908813819069, + "learning_rate": 2.1023288334251222e-07, + "loss": 1.7336, + "step": 10790 + }, + { + "epoch": 0.7274013678784408, + "grad_norm": 5.245586756603191, + "learning_rate": 2.0975385461121864e-07, + "loss": 1.8321, + "step": 10795 + }, + { + "epoch": 0.7277382837505475, + "grad_norm": 5.920551725229809, + "learning_rate": 2.0927522735015268e-07, + "loss": 1.6794, + "step": 10800 + }, + { + "epoch": 0.7280751996226542, + "grad_norm": 5.618144476884494, + "learning_rate": 2.0879700222135416e-07, + "loss": 1.792, + "step": 10805 + }, + { + "epoch": 0.7284121154947609, + "grad_norm": 6.032158334331574, + "learning_rate": 2.083191798863072e-07, + "loss": 1.783, + "step": 10810 + }, + { + "epoch": 0.7287490313668676, + "grad_norm": 6.065083852382277, + "learning_rate": 2.0784176100593836e-07, + "loss": 1.7731, + "step": 10815 + }, + { + "epoch": 0.7290859472389745, + "grad_norm": 6.262727055878644, + "learning_rate": 2.0736474624061655e-07, + "loss": 1.8491, + "step": 10820 + }, + { + "epoch": 0.7294228631110812, + "grad_norm": 5.725691879050536, + "learning_rate": 2.0688813625015123e-07, + "loss": 1.7585, + "step": 10825 + }, + { + "epoch": 0.7297597789831879, + "grad_norm": 6.127578280833137, + "learning_rate": 2.064119316937923e-07, + "loss": 1.8265, + "step": 10830 + }, + { + "epoch": 0.7300966948552946, + "grad_norm": 5.911039793824732, + "learning_rate": 2.0593613323022907e-07, + "loss": 1.75, + "step": 10835 + }, + { + "epoch": 0.7304336107274013, + "grad_norm": 6.138482118358594, + "learning_rate": 2.054607415175884e-07, + "loss": 1.802, + "step": 10840 + }, + { + "epoch": 0.7307705265995081, + "grad_norm": 6.265596175254885, + "learning_rate": 2.0498575721343525e-07, + "loss": 1.8145, + "step": 10845 + }, + { + "epoch": 0.7311074424716149, + "grad_norm": 6.2040452798334895, + "learning_rate": 2.0451118097477093e-07, + "loss": 1.7859, + "step": 10850 + }, + { + "epoch": 0.7314443583437216, + "grad_norm": 6.550091577089053, + "learning_rate": 2.0403701345803186e-07, + "loss": 1.7903, + "step": 10855 + }, + { + "epoch": 0.7317812742158283, + "grad_norm": 6.022289220773169, + "learning_rate": 2.0356325531908952e-07, + "loss": 1.787, + "step": 10860 + }, + { + "epoch": 0.732118190087935, + "grad_norm": 5.63615689024889, + "learning_rate": 2.0308990721324926e-07, + "loss": 1.7921, + "step": 10865 + }, + { + "epoch": 0.7324551059600418, + "grad_norm": 5.554562145897487, + "learning_rate": 2.0261696979524873e-07, + "loss": 1.7947, + "step": 10870 + }, + { + "epoch": 0.7327920218321485, + "grad_norm": 5.90754768571259, + "learning_rate": 2.0214444371925793e-07, + "loss": 1.6978, + "step": 10875 + }, + { + "epoch": 0.7331289377042552, + "grad_norm": 5.737704638109933, + "learning_rate": 2.0167232963887787e-07, + "loss": 1.7567, + "step": 10880 + }, + { + "epoch": 0.733465853576362, + "grad_norm": 6.505256107780983, + "learning_rate": 2.0120062820713974e-07, + "loss": 1.7953, + "step": 10885 + }, + { + "epoch": 0.7338027694484687, + "grad_norm": 6.0759411860109855, + "learning_rate": 2.0072934007650345e-07, + "loss": 1.7824, + "step": 10890 + }, + { + "epoch": 0.7341396853205755, + "grad_norm": 5.67638008354974, + "learning_rate": 2.0025846589885798e-07, + "loss": 1.8058, + "step": 10895 + }, + { + "epoch": 0.7344766011926822, + "grad_norm": 5.722828442355223, + "learning_rate": 1.99788006325519e-07, + "loss": 1.7025, + "step": 10900 + }, + { + "epoch": 0.7348135170647889, + "grad_norm": 5.832985351284273, + "learning_rate": 1.9931796200722943e-07, + "loss": 1.7731, + "step": 10905 + }, + { + "epoch": 0.7351504329368956, + "grad_norm": 5.781715493966306, + "learning_rate": 1.9884833359415698e-07, + "loss": 1.7884, + "step": 10910 + }, + { + "epoch": 0.7354873488090024, + "grad_norm": 5.648892528599704, + "learning_rate": 1.9837912173589494e-07, + "loss": 1.8245, + "step": 10915 + }, + { + "epoch": 0.7358242646811092, + "grad_norm": 6.213430881285172, + "learning_rate": 1.979103270814596e-07, + "loss": 1.8618, + "step": 10920 + }, + { + "epoch": 0.7361611805532159, + "grad_norm": 5.833744394817774, + "learning_rate": 1.9744195027929072e-07, + "loss": 1.7965, + "step": 10925 + }, + { + "epoch": 0.7364980964253226, + "grad_norm": 6.295540209087126, + "learning_rate": 1.9697399197725023e-07, + "loss": 1.8203, + "step": 10930 + }, + { + "epoch": 0.7368350122974293, + "grad_norm": 5.9347340029367786, + "learning_rate": 1.965064528226204e-07, + "loss": 1.8165, + "step": 10935 + }, + { + "epoch": 0.737171928169536, + "grad_norm": 6.24051717756245, + "learning_rate": 1.9603933346210445e-07, + "loss": 1.8334, + "step": 10940 + }, + { + "epoch": 0.7375088440416429, + "grad_norm": 6.077355157061982, + "learning_rate": 1.9557263454182476e-07, + "loss": 1.7887, + "step": 10945 + }, + { + "epoch": 0.7378457599137496, + "grad_norm": 5.672952748842877, + "learning_rate": 1.9510635670732216e-07, + "loss": 1.7845, + "step": 10950 + }, + { + "epoch": 0.7381826757858563, + "grad_norm": 6.078037272901075, + "learning_rate": 1.946405006035548e-07, + "loss": 1.7856, + "step": 10955 + }, + { + "epoch": 0.738519591657963, + "grad_norm": 5.934637815732029, + "learning_rate": 1.9417506687489772e-07, + "loss": 1.8035, + "step": 10960 + }, + { + "epoch": 0.7388565075300697, + "grad_norm": 5.667637574057072, + "learning_rate": 1.937100561651418e-07, + "loss": 1.7679, + "step": 10965 + }, + { + "epoch": 0.7391934234021765, + "grad_norm": 6.121851679797381, + "learning_rate": 1.9324546911749246e-07, + "loss": 1.7986, + "step": 10970 + }, + { + "epoch": 0.7395303392742832, + "grad_norm": 5.452560236133865, + "learning_rate": 1.9278130637456957e-07, + "loss": 1.7125, + "step": 10975 + }, + { + "epoch": 0.73986725514639, + "grad_norm": 6.117974741639468, + "learning_rate": 1.923175685784056e-07, + "loss": 1.791, + "step": 10980 + }, + { + "epoch": 0.7402041710184967, + "grad_norm": 6.418307670126074, + "learning_rate": 1.9185425637044567e-07, + "loss": 1.7894, + "step": 10985 + }, + { + "epoch": 0.7405410868906034, + "grad_norm": 5.9432415931639415, + "learning_rate": 1.9139137039154584e-07, + "loss": 1.8261, + "step": 10990 + }, + { + "epoch": 0.7408780027627101, + "grad_norm": 5.899504950225264, + "learning_rate": 1.9092891128197308e-07, + "loss": 1.7587, + "step": 10995 + }, + { + "epoch": 0.7412149186348169, + "grad_norm": 6.154385724870445, + "learning_rate": 1.904668796814033e-07, + "loss": 1.8855, + "step": 11000 + }, + { + "epoch": 0.7415518345069236, + "grad_norm": 6.4245473778375795, + "learning_rate": 1.9000527622892154e-07, + "loss": 1.8116, + "step": 11005 + }, + { + "epoch": 0.7418887503790303, + "grad_norm": 6.0428091022921535, + "learning_rate": 1.895441015630206e-07, + "loss": 1.7959, + "step": 11010 + }, + { + "epoch": 0.7422256662511371, + "grad_norm": 5.998009175042043, + "learning_rate": 1.8908335632160011e-07, + "loss": 1.8721, + "step": 11015 + }, + { + "epoch": 0.7425625821232438, + "grad_norm": 5.824437955606014, + "learning_rate": 1.8862304114196542e-07, + "loss": 1.7818, + "step": 11020 + }, + { + "epoch": 0.7428994979953506, + "grad_norm": 6.578699865645307, + "learning_rate": 1.8816315666082744e-07, + "loss": 1.7565, + "step": 11025 + }, + { + "epoch": 0.7432364138674573, + "grad_norm": 6.113247957569191, + "learning_rate": 1.877037035143013e-07, + "loss": 1.8026, + "step": 11030 + }, + { + "epoch": 0.743573329739564, + "grad_norm": 5.882144085105441, + "learning_rate": 1.8724468233790512e-07, + "loss": 1.7853, + "step": 11035 + }, + { + "epoch": 0.7439102456116707, + "grad_norm": 6.557823258170281, + "learning_rate": 1.867860937665599e-07, + "loss": 1.8762, + "step": 11040 + }, + { + "epoch": 0.7442471614837775, + "grad_norm": 5.474147682422214, + "learning_rate": 1.8632793843458827e-07, + "loss": 1.823, + "step": 11045 + }, + { + "epoch": 0.7445840773558843, + "grad_norm": 6.581490827678887, + "learning_rate": 1.8587021697571313e-07, + "loss": 1.7813, + "step": 11050 + }, + { + "epoch": 0.744920993227991, + "grad_norm": 6.447231328360916, + "learning_rate": 1.854129300230578e-07, + "loss": 1.8047, + "step": 11055 + }, + { + "epoch": 0.7452579091000977, + "grad_norm": 6.041430889705571, + "learning_rate": 1.849560782091445e-07, + "loss": 1.7589, + "step": 11060 + }, + { + "epoch": 0.7455948249722044, + "grad_norm": 6.245656556925254, + "learning_rate": 1.8449966216589319e-07, + "loss": 1.8047, + "step": 11065 + }, + { + "epoch": 0.7459317408443111, + "grad_norm": 6.341804719685981, + "learning_rate": 1.8404368252462128e-07, + "loss": 1.753, + "step": 11070 + }, + { + "epoch": 0.746268656716418, + "grad_norm": 6.114522479869793, + "learning_rate": 1.8358813991604262e-07, + "loss": 1.8459, + "step": 11075 + }, + { + "epoch": 0.7466055725885247, + "grad_norm": 5.82791681785764, + "learning_rate": 1.8313303497026673e-07, + "loss": 1.7887, + "step": 11080 + }, + { + "epoch": 0.7469424884606314, + "grad_norm": 6.059091931920949, + "learning_rate": 1.8267836831679718e-07, + "loss": 1.8508, + "step": 11085 + }, + { + "epoch": 0.7472794043327381, + "grad_norm": 6.184374099826963, + "learning_rate": 1.8222414058453183e-07, + "loss": 1.8226, + "step": 11090 + }, + { + "epoch": 0.7476163202048448, + "grad_norm": 5.768389403876578, + "learning_rate": 1.8177035240176136e-07, + "loss": 1.8282, + "step": 11095 + }, + { + "epoch": 0.7479532360769516, + "grad_norm": 6.626367557762699, + "learning_rate": 1.8131700439616803e-07, + "loss": 1.8547, + "step": 11100 + }, + { + "epoch": 0.7482901519490583, + "grad_norm": 5.830974589056405, + "learning_rate": 1.8086409719482576e-07, + "loss": 1.729, + "step": 11105 + }, + { + "epoch": 0.7486270678211651, + "grad_norm": 5.856454265986918, + "learning_rate": 1.8041163142419857e-07, + "loss": 1.8279, + "step": 11110 + }, + { + "epoch": 0.7489639836932718, + "grad_norm": 6.178371847517803, + "learning_rate": 1.7995960771013962e-07, + "loss": 1.7677, + "step": 11115 + }, + { + "epoch": 0.7493008995653785, + "grad_norm": 6.657651397465904, + "learning_rate": 1.7950802667789107e-07, + "loss": 1.8415, + "step": 11120 + }, + { + "epoch": 0.7496378154374853, + "grad_norm": 5.550990585298974, + "learning_rate": 1.7905688895208259e-07, + "loss": 1.7762, + "step": 11125 + }, + { + "epoch": 0.749974731309592, + "grad_norm": 6.099324547733622, + "learning_rate": 1.7860619515673032e-07, + "loss": 1.8136, + "step": 11130 + }, + { + "epoch": 0.7503116471816987, + "grad_norm": 6.650418350028828, + "learning_rate": 1.7815594591523687e-07, + "loss": 1.8521, + "step": 11135 + }, + { + "epoch": 0.7506485630538055, + "grad_norm": 5.902004000771492, + "learning_rate": 1.777061418503898e-07, + "loss": 1.8051, + "step": 11140 + }, + { + "epoch": 0.7509854789259122, + "grad_norm": 5.952130300679766, + "learning_rate": 1.7725678358436053e-07, + "loss": 1.784, + "step": 11145 + }, + { + "epoch": 0.751322394798019, + "grad_norm": 5.775716403203176, + "learning_rate": 1.7680787173870454e-07, + "loss": 1.7143, + "step": 11150 + }, + { + "epoch": 0.7516593106701257, + "grad_norm": 5.483388083311626, + "learning_rate": 1.763594069343589e-07, + "loss": 1.7443, + "step": 11155 + }, + { + "epoch": 0.7519962265422324, + "grad_norm": 6.1697764587867265, + "learning_rate": 1.7591138979164337e-07, + "loss": 1.7364, + "step": 11160 + }, + { + "epoch": 0.7523331424143391, + "grad_norm": 5.761947711806687, + "learning_rate": 1.7546382093025758e-07, + "loss": 1.8046, + "step": 11165 + }, + { + "epoch": 0.7526700582864458, + "grad_norm": 6.194710744116213, + "learning_rate": 1.7501670096928162e-07, + "loss": 1.8163, + "step": 11170 + }, + { + "epoch": 0.7530069741585526, + "grad_norm": 5.488662638622846, + "learning_rate": 1.7457003052717473e-07, + "loss": 1.859, + "step": 11175 + }, + { + "epoch": 0.7533438900306594, + "grad_norm": 6.690003543021317, + "learning_rate": 1.741238102217738e-07, + "loss": 1.7725, + "step": 11180 + }, + { + "epoch": 0.7536808059027661, + "grad_norm": 5.803437120207633, + "learning_rate": 1.736780406702937e-07, + "loss": 1.778, + "step": 11185 + }, + { + "epoch": 0.7540177217748728, + "grad_norm": 6.069859658324239, + "learning_rate": 1.7323272248932564e-07, + "loss": 1.7409, + "step": 11190 + }, + { + "epoch": 0.7543546376469795, + "grad_norm": 5.500334561391225, + "learning_rate": 1.727878562948362e-07, + "loss": 1.769, + "step": 11195 + }, + { + "epoch": 0.7546915535190862, + "grad_norm": 6.708781480901355, + "learning_rate": 1.723434427021671e-07, + "loss": 1.7128, + "step": 11200 + }, + { + "epoch": 0.7550284693911931, + "grad_norm": 5.826615006643447, + "learning_rate": 1.7189948232603412e-07, + "loss": 1.815, + "step": 11205 + }, + { + "epoch": 0.7553653852632998, + "grad_norm": 5.4713789742043275, + "learning_rate": 1.7145597578052557e-07, + "loss": 1.8527, + "step": 11210 + }, + { + "epoch": 0.7557023011354065, + "grad_norm": 5.9722168111861, + "learning_rate": 1.7101292367910259e-07, + "loss": 1.8327, + "step": 11215 + }, + { + "epoch": 0.7560392170075132, + "grad_norm": 5.927637118304886, + "learning_rate": 1.7057032663459768e-07, + "loss": 1.8339, + "step": 11220 + }, + { + "epoch": 0.7563761328796199, + "grad_norm": 6.7802666022955815, + "learning_rate": 1.701281852592134e-07, + "loss": 1.7988, + "step": 11225 + }, + { + "epoch": 0.7567130487517267, + "grad_norm": 5.797191803923179, + "learning_rate": 1.696865001645228e-07, + "loss": 1.7488, + "step": 11230 + }, + { + "epoch": 0.7570499646238334, + "grad_norm": 5.441139282117675, + "learning_rate": 1.6924527196146692e-07, + "loss": 1.8553, + "step": 11235 + }, + { + "epoch": 0.7573868804959402, + "grad_norm": 5.812956263017639, + "learning_rate": 1.6880450126035572e-07, + "loss": 1.7806, + "step": 11240 + }, + { + "epoch": 0.7577237963680469, + "grad_norm": 6.166337691279962, + "learning_rate": 1.683641886708655e-07, + "loss": 1.7556, + "step": 11245 + }, + { + "epoch": 0.7580607122401536, + "grad_norm": 5.87741529416599, + "learning_rate": 1.6792433480203955e-07, + "loss": 1.7676, + "step": 11250 + }, + { + "epoch": 0.7583976281122604, + "grad_norm": 6.040766300149007, + "learning_rate": 1.674849402622865e-07, + "loss": 1.8061, + "step": 11255 + }, + { + "epoch": 0.7587345439843671, + "grad_norm": 5.897769345655216, + "learning_rate": 1.6704600565937927e-07, + "loss": 1.8181, + "step": 11260 + }, + { + "epoch": 0.7590714598564738, + "grad_norm": 6.171163777497132, + "learning_rate": 1.6660753160045498e-07, + "loss": 1.738, + "step": 11265 + }, + { + "epoch": 0.7594083757285806, + "grad_norm": 5.981331595498035, + "learning_rate": 1.6616951869201378e-07, + "loss": 1.7648, + "step": 11270 + }, + { + "epoch": 0.7597452916006873, + "grad_norm": 5.305640332140947, + "learning_rate": 1.6573196753991747e-07, + "loss": 1.8503, + "step": 11275 + }, + { + "epoch": 0.7600822074727941, + "grad_norm": 6.0520841045058456, + "learning_rate": 1.652948787493896e-07, + "loss": 1.797, + "step": 11280 + }, + { + "epoch": 0.7604191233449008, + "grad_norm": 6.148814790534166, + "learning_rate": 1.64858252925014e-07, + "loss": 1.8173, + "step": 11285 + }, + { + "epoch": 0.7607560392170075, + "grad_norm": 5.979828710800014, + "learning_rate": 1.6442209067073442e-07, + "loss": 1.7896, + "step": 11290 + }, + { + "epoch": 0.7610929550891142, + "grad_norm": 6.217740605002487, + "learning_rate": 1.639863925898527e-07, + "loss": 1.812, + "step": 11295 + }, + { + "epoch": 0.761429870961221, + "grad_norm": 6.185363065792439, + "learning_rate": 1.6355115928502934e-07, + "loss": 1.8223, + "step": 11300 + }, + { + "epoch": 0.7617667868333278, + "grad_norm": 6.068122134218195, + "learning_rate": 1.6311639135828176e-07, + "loss": 1.8278, + "step": 11305 + }, + { + "epoch": 0.7621037027054345, + "grad_norm": 6.41089152387682, + "learning_rate": 1.6268208941098344e-07, + "loss": 1.7763, + "step": 11310 + }, + { + "epoch": 0.7624406185775412, + "grad_norm": 5.670824470630969, + "learning_rate": 1.6224825404386326e-07, + "loss": 1.7438, + "step": 11315 + }, + { + "epoch": 0.7627775344496479, + "grad_norm": 5.613534488092596, + "learning_rate": 1.6181488585700541e-07, + "loss": 1.6914, + "step": 11320 + }, + { + "epoch": 0.7631144503217546, + "grad_norm": 6.215093315606554, + "learning_rate": 1.6138198544984692e-07, + "loss": 1.8232, + "step": 11325 + }, + { + "epoch": 0.7634513661938614, + "grad_norm": 5.922779615098384, + "learning_rate": 1.609495534211785e-07, + "loss": 1.8005, + "step": 11330 + }, + { + "epoch": 0.7637882820659682, + "grad_norm": 5.777262125949082, + "learning_rate": 1.6051759036914286e-07, + "loss": 1.7512, + "step": 11335 + }, + { + "epoch": 0.7641251979380749, + "grad_norm": 6.045493587778366, + "learning_rate": 1.6008609689123364e-07, + "loss": 1.8303, + "step": 11340 + }, + { + "epoch": 0.7644621138101816, + "grad_norm": 6.160481197853635, + "learning_rate": 1.596550735842953e-07, + "loss": 1.8254, + "step": 11345 + }, + { + "epoch": 0.7647990296822883, + "grad_norm": 5.995522803991697, + "learning_rate": 1.5922452104452204e-07, + "loss": 1.795, + "step": 11350 + }, + { + "epoch": 0.765135945554395, + "grad_norm": 6.407506795307969, + "learning_rate": 1.5879443986745678e-07, + "loss": 1.8201, + "step": 11355 + }, + { + "epoch": 0.7654728614265018, + "grad_norm": 5.868369233442601, + "learning_rate": 1.583648306479901e-07, + "loss": 1.7845, + "step": 11360 + }, + { + "epoch": 0.7658097772986086, + "grad_norm": 6.300762474878201, + "learning_rate": 1.5793569398036032e-07, + "loss": 1.8119, + "step": 11365 + }, + { + "epoch": 0.7661466931707153, + "grad_norm": 5.732440809710967, + "learning_rate": 1.57507030458152e-07, + "loss": 1.7298, + "step": 11370 + }, + { + "epoch": 0.766483609042822, + "grad_norm": 5.691483180721791, + "learning_rate": 1.5707884067429471e-07, + "loss": 1.8783, + "step": 11375 + }, + { + "epoch": 0.7668205249149287, + "grad_norm": 5.888135695526149, + "learning_rate": 1.566511252210635e-07, + "loss": 1.7552, + "step": 11380 + }, + { + "epoch": 0.7671574407870355, + "grad_norm": 5.475719381961438, + "learning_rate": 1.5622388469007696e-07, + "loss": 1.7779, + "step": 11385 + }, + { + "epoch": 0.7674943566591422, + "grad_norm": 5.736297394223776, + "learning_rate": 1.5579711967229652e-07, + "loss": 1.7512, + "step": 11390 + }, + { + "epoch": 0.7678312725312489, + "grad_norm": 6.108820433367013, + "learning_rate": 1.5537083075802647e-07, + "loss": 1.8151, + "step": 11395 + }, + { + "epoch": 0.7681681884033557, + "grad_norm": 5.645549075230665, + "learning_rate": 1.5494501853691195e-07, + "loss": 1.8157, + "step": 11400 + }, + { + "epoch": 0.7685051042754624, + "grad_norm": 6.416767723799877, + "learning_rate": 1.5451968359793927e-07, + "loss": 1.7385, + "step": 11405 + }, + { + "epoch": 0.7688420201475692, + "grad_norm": 5.795195428364132, + "learning_rate": 1.5409482652943396e-07, + "loss": 1.753, + "step": 11410 + }, + { + "epoch": 0.7691789360196759, + "grad_norm": 6.287596532670337, + "learning_rate": 1.536704479190611e-07, + "loss": 1.7284, + "step": 11415 + }, + { + "epoch": 0.7695158518917826, + "grad_norm": 5.7722465474507425, + "learning_rate": 1.5324654835382384e-07, + "loss": 1.8132, + "step": 11420 + }, + { + "epoch": 0.7698527677638893, + "grad_norm": 5.810495984958396, + "learning_rate": 1.5282312842006238e-07, + "loss": 1.7963, + "step": 11425 + }, + { + "epoch": 0.770189683635996, + "grad_norm": 5.7818877383121015, + "learning_rate": 1.5240018870345388e-07, + "loss": 1.7625, + "step": 11430 + }, + { + "epoch": 0.7705265995081029, + "grad_norm": 6.4616000691306015, + "learning_rate": 1.519777297890113e-07, + "loss": 1.7305, + "step": 11435 + }, + { + "epoch": 0.7708635153802096, + "grad_norm": 6.9500507066394235, + "learning_rate": 1.5155575226108198e-07, + "loss": 1.7646, + "step": 11440 + }, + { + "epoch": 0.7712004312523163, + "grad_norm": 5.90587314198153, + "learning_rate": 1.51134256703348e-07, + "loss": 1.8436, + "step": 11445 + }, + { + "epoch": 0.771537347124423, + "grad_norm": 5.866892874363843, + "learning_rate": 1.5071324369882478e-07, + "loss": 1.7458, + "step": 11450 + }, + { + "epoch": 0.7718742629965297, + "grad_norm": 6.291434417162354, + "learning_rate": 1.5029271382985964e-07, + "loss": 1.7889, + "step": 11455 + }, + { + "epoch": 0.7722111788686365, + "grad_norm": 6.101683917084692, + "learning_rate": 1.498726676781323e-07, + "loss": 1.8393, + "step": 11460 + }, + { + "epoch": 0.7725480947407433, + "grad_norm": 6.448679173286852, + "learning_rate": 1.4945310582465327e-07, + "loss": 1.823, + "step": 11465 + }, + { + "epoch": 0.77288501061285, + "grad_norm": 6.03477234470622, + "learning_rate": 1.4903402884976262e-07, + "loss": 1.8207, + "step": 11470 + }, + { + "epoch": 0.7732219264849567, + "grad_norm": 6.7380284585238135, + "learning_rate": 1.4861543733313065e-07, + "loss": 1.8021, + "step": 11475 + }, + { + "epoch": 0.7735588423570634, + "grad_norm": 6.077995753476859, + "learning_rate": 1.4819733185375531e-07, + "loss": 1.8847, + "step": 11480 + }, + { + "epoch": 0.7738957582291702, + "grad_norm": 5.664292238827305, + "learning_rate": 1.4777971298996288e-07, + "loss": 1.7623, + "step": 11485 + }, + { + "epoch": 0.7742326741012769, + "grad_norm": 5.799868960607819, + "learning_rate": 1.4736258131940605e-07, + "loss": 1.7456, + "step": 11490 + }, + { + "epoch": 0.7745695899733837, + "grad_norm": 6.433673442490591, + "learning_rate": 1.4694593741906403e-07, + "loss": 1.8046, + "step": 11495 + }, + { + "epoch": 0.7749065058454904, + "grad_norm": 6.079136779626077, + "learning_rate": 1.4652978186524135e-07, + "loss": 1.8136, + "step": 11500 + }, + { + "epoch": 0.7752434217175971, + "grad_norm": 5.575230853873149, + "learning_rate": 1.4611411523356653e-07, + "loss": 1.8148, + "step": 11505 + }, + { + "epoch": 0.7755803375897039, + "grad_norm": 5.9996748971291, + "learning_rate": 1.4569893809899242e-07, + "loss": 1.8543, + "step": 11510 + }, + { + "epoch": 0.7759172534618106, + "grad_norm": 5.931034717600519, + "learning_rate": 1.452842510357946e-07, + "loss": 1.756, + "step": 11515 + }, + { + "epoch": 0.7762541693339173, + "grad_norm": 5.768234880310048, + "learning_rate": 1.4487005461757051e-07, + "loss": 1.7537, + "step": 11520 + }, + { + "epoch": 0.776591085206024, + "grad_norm": 5.848819600001102, + "learning_rate": 1.4445634941723927e-07, + "loss": 1.7961, + "step": 11525 + }, + { + "epoch": 0.7769280010781308, + "grad_norm": 5.865945761273097, + "learning_rate": 1.4404313600704054e-07, + "loss": 1.8757, + "step": 11530 + }, + { + "epoch": 0.7772649169502375, + "grad_norm": 6.156370938984119, + "learning_rate": 1.4363041495853334e-07, + "loss": 1.7886, + "step": 11535 + }, + { + "epoch": 0.7776018328223443, + "grad_norm": 6.286006570843523, + "learning_rate": 1.4321818684259607e-07, + "loss": 1.8275, + "step": 11540 + }, + { + "epoch": 0.777938748694451, + "grad_norm": 5.702969599119352, + "learning_rate": 1.4280645222942535e-07, + "loss": 1.786, + "step": 11545 + }, + { + "epoch": 0.7782756645665577, + "grad_norm": 5.670263286040782, + "learning_rate": 1.4239521168853458e-07, + "loss": 1.7547, + "step": 11550 + }, + { + "epoch": 0.7786125804386644, + "grad_norm": 5.877323637618684, + "learning_rate": 1.4198446578875444e-07, + "loss": 1.7754, + "step": 11555 + }, + { + "epoch": 0.7789494963107712, + "grad_norm": 5.950199480950997, + "learning_rate": 1.4157421509823119e-07, + "loss": 1.7511, + "step": 11560 + }, + { + "epoch": 0.779286412182878, + "grad_norm": 5.888421854077837, + "learning_rate": 1.4116446018442608e-07, + "loss": 1.7078, + "step": 11565 + }, + { + "epoch": 0.7796233280549847, + "grad_norm": 6.034680017416559, + "learning_rate": 1.4075520161411425e-07, + "loss": 1.7349, + "step": 11570 + }, + { + "epoch": 0.7799602439270914, + "grad_norm": 6.307386970077885, + "learning_rate": 1.403464399533849e-07, + "loss": 1.7923, + "step": 11575 + }, + { + "epoch": 0.7802971597991981, + "grad_norm": 5.967516151887473, + "learning_rate": 1.3993817576763983e-07, + "loss": 1.8707, + "step": 11580 + }, + { + "epoch": 0.7806340756713048, + "grad_norm": 6.351426294821415, + "learning_rate": 1.3953040962159207e-07, + "loss": 1.7889, + "step": 11585 + }, + { + "epoch": 0.7809709915434117, + "grad_norm": 6.192822266741291, + "learning_rate": 1.3912314207926657e-07, + "loss": 1.77, + "step": 11590 + }, + { + "epoch": 0.7813079074155184, + "grad_norm": 5.8783177680109855, + "learning_rate": 1.3871637370399824e-07, + "loss": 1.7897, + "step": 11595 + }, + { + "epoch": 0.7816448232876251, + "grad_norm": 6.222300669334982, + "learning_rate": 1.3831010505843139e-07, + "loss": 1.7571, + "step": 11600 + }, + { + "epoch": 0.7819817391597318, + "grad_norm": 5.5964193211778825, + "learning_rate": 1.3790433670451927e-07, + "loss": 1.729, + "step": 11605 + }, + { + "epoch": 0.7823186550318385, + "grad_norm": 5.724469476942663, + "learning_rate": 1.374990692035235e-07, + "loss": 1.7733, + "step": 11610 + }, + { + "epoch": 0.7826555709039453, + "grad_norm": 6.081763550531529, + "learning_rate": 1.3709430311601205e-07, + "loss": 1.7551, + "step": 11615 + }, + { + "epoch": 0.782992486776052, + "grad_norm": 5.656334636067317, + "learning_rate": 1.366900390018601e-07, + "loss": 1.689, + "step": 11620 + }, + { + "epoch": 0.7833294026481588, + "grad_norm": 5.899100005447112, + "learning_rate": 1.3628627742024812e-07, + "loss": 1.8493, + "step": 11625 + }, + { + "epoch": 0.7836663185202655, + "grad_norm": 6.416010977631392, + "learning_rate": 1.3588301892966182e-07, + "loss": 1.6631, + "step": 11630 + }, + { + "epoch": 0.7840032343923722, + "grad_norm": 6.2420816157066845, + "learning_rate": 1.3548026408789044e-07, + "loss": 1.8145, + "step": 11635 + }, + { + "epoch": 0.784340150264479, + "grad_norm": 6.394093878050669, + "learning_rate": 1.350780134520272e-07, + "loss": 1.8385, + "step": 11640 + }, + { + "epoch": 0.7846770661365857, + "grad_norm": 5.898670325238131, + "learning_rate": 1.3467626757846733e-07, + "loss": 1.765, + "step": 11645 + }, + { + "epoch": 0.7850139820086924, + "grad_norm": 5.755709771745298, + "learning_rate": 1.342750270229085e-07, + "loss": 1.7411, + "step": 11650 + }, + { + "epoch": 0.7853508978807991, + "grad_norm": 6.361755561484261, + "learning_rate": 1.338742923403487e-07, + "loss": 1.7003, + "step": 11655 + }, + { + "epoch": 0.7856878137529059, + "grad_norm": 6.475887456311801, + "learning_rate": 1.3347406408508694e-07, + "loss": 1.7245, + "step": 11660 + }, + { + "epoch": 0.7860247296250127, + "grad_norm": 5.87715144272717, + "learning_rate": 1.3307434281072106e-07, + "loss": 1.827, + "step": 11665 + }, + { + "epoch": 0.7863616454971194, + "grad_norm": 6.139319666913612, + "learning_rate": 1.326751290701481e-07, + "loss": 1.7936, + "step": 11670 + }, + { + "epoch": 0.7866985613692261, + "grad_norm": 5.760142588635934, + "learning_rate": 1.3227642341556306e-07, + "loss": 1.8124, + "step": 11675 + }, + { + "epoch": 0.7870354772413328, + "grad_norm": 5.731411372422863, + "learning_rate": 1.318782263984577e-07, + "loss": 1.813, + "step": 11680 + }, + { + "epoch": 0.7873723931134395, + "grad_norm": 6.028570522927185, + "learning_rate": 1.314805385696207e-07, + "loss": 1.8055, + "step": 11685 + }, + { + "epoch": 0.7877093089855464, + "grad_norm": 6.4147651771131144, + "learning_rate": 1.3108336047913633e-07, + "loss": 1.8518, + "step": 11690 + }, + { + "epoch": 0.7880462248576531, + "grad_norm": 5.745385502619762, + "learning_rate": 1.3068669267638377e-07, + "loss": 1.766, + "step": 11695 + }, + { + "epoch": 0.7883831407297598, + "grad_norm": 5.753592382263267, + "learning_rate": 1.3029053571003619e-07, + "loss": 1.8171, + "step": 11700 + }, + { + "epoch": 0.7887200566018665, + "grad_norm": 6.094153318282813, + "learning_rate": 1.2989489012806033e-07, + "loss": 1.8266, + "step": 11705 + }, + { + "epoch": 0.7890569724739732, + "grad_norm": 6.626686965131564, + "learning_rate": 1.294997564777157e-07, + "loss": 1.7849, + "step": 11710 + }, + { + "epoch": 0.7893938883460799, + "grad_norm": 6.089503582912961, + "learning_rate": 1.291051353055534e-07, + "loss": 1.8466, + "step": 11715 + }, + { + "epoch": 0.7897308042181868, + "grad_norm": 5.949166550268341, + "learning_rate": 1.28711027157416e-07, + "loss": 1.7636, + "step": 11720 + }, + { + "epoch": 0.7900677200902935, + "grad_norm": 5.777412800846406, + "learning_rate": 1.2831743257843597e-07, + "loss": 1.7805, + "step": 11725 + }, + { + "epoch": 0.7904046359624002, + "grad_norm": 6.030785841040514, + "learning_rate": 1.279243521130361e-07, + "loss": 1.7691, + "step": 11730 + }, + { + "epoch": 0.7907415518345069, + "grad_norm": 6.159706148659863, + "learning_rate": 1.2753178630492733e-07, + "loss": 1.874, + "step": 11735 + }, + { + "epoch": 0.7910784677066136, + "grad_norm": 5.984302184404192, + "learning_rate": 1.271397356971094e-07, + "loss": 1.7516, + "step": 11740 + }, + { + "epoch": 0.7914153835787204, + "grad_norm": 6.29530906949322, + "learning_rate": 1.267482008318687e-07, + "loss": 1.7803, + "step": 11745 + }, + { + "epoch": 0.7917522994508271, + "grad_norm": 5.999071523327007, + "learning_rate": 1.2635718225077884e-07, + "loss": 1.9044, + "step": 11750 + }, + { + "epoch": 0.7920892153229339, + "grad_norm": 6.009971558825039, + "learning_rate": 1.259666804946991e-07, + "loss": 1.7475, + "step": 11755 + }, + { + "epoch": 0.7924261311950406, + "grad_norm": 6.1762343420480725, + "learning_rate": 1.2557669610377397e-07, + "loss": 1.7928, + "step": 11760 + }, + { + "epoch": 0.7927630470671473, + "grad_norm": 6.274370196479858, + "learning_rate": 1.25187229617432e-07, + "loss": 1.8475, + "step": 11765 + }, + { + "epoch": 0.7930999629392541, + "grad_norm": 5.619732178273275, + "learning_rate": 1.247982815743857e-07, + "loss": 1.7979, + "step": 11770 + }, + { + "epoch": 0.7934368788113608, + "grad_norm": 6.37694088286426, + "learning_rate": 1.2440985251263054e-07, + "loss": 1.7106, + "step": 11775 + }, + { + "epoch": 0.7937737946834675, + "grad_norm": 5.846264824270768, + "learning_rate": 1.2402194296944363e-07, + "loss": 1.7965, + "step": 11780 + }, + { + "epoch": 0.7941107105555743, + "grad_norm": 6.051659571401686, + "learning_rate": 1.236345534813839e-07, + "loss": 1.7276, + "step": 11785 + }, + { + "epoch": 0.794447626427681, + "grad_norm": 6.040361548706306, + "learning_rate": 1.2324768458429107e-07, + "loss": 1.7819, + "step": 11790 + }, + { + "epoch": 0.7947845422997878, + "grad_norm": 6.5893551244691, + "learning_rate": 1.228613368132842e-07, + "loss": 1.8357, + "step": 11795 + }, + { + "epoch": 0.7951214581718945, + "grad_norm": 5.556745304140574, + "learning_rate": 1.2247551070276207e-07, + "loss": 1.6579, + "step": 11800 + }, + { + "epoch": 0.7954583740440012, + "grad_norm": 5.90446121886169, + "learning_rate": 1.2209020678640176e-07, + "loss": 1.8439, + "step": 11805 + }, + { + "epoch": 0.7957952899161079, + "grad_norm": 6.108546411118622, + "learning_rate": 1.2170542559715775e-07, + "loss": 1.7384, + "step": 11810 + }, + { + "epoch": 0.7961322057882146, + "grad_norm": 6.183593777221998, + "learning_rate": 1.2132116766726196e-07, + "loss": 1.8085, + "step": 11815 + }, + { + "epoch": 0.7964691216603215, + "grad_norm": 5.981494846943721, + "learning_rate": 1.2093743352822206e-07, + "loss": 1.8076, + "step": 11820 + }, + { + "epoch": 0.7968060375324282, + "grad_norm": 7.444395690512402, + "learning_rate": 1.2055422371082168e-07, + "loss": 1.7539, + "step": 11825 + }, + { + "epoch": 0.7971429534045349, + "grad_norm": 5.876372023884193, + "learning_rate": 1.2017153874511865e-07, + "loss": 1.7757, + "step": 11830 + }, + { + "epoch": 0.7974798692766416, + "grad_norm": 6.370625401977648, + "learning_rate": 1.1978937916044534e-07, + "loss": 1.8109, + "step": 11835 + }, + { + "epoch": 0.7978167851487483, + "grad_norm": 5.883242118263617, + "learning_rate": 1.1940774548540733e-07, + "loss": 1.7298, + "step": 11840 + }, + { + "epoch": 0.7981537010208551, + "grad_norm": 5.767584882456747, + "learning_rate": 1.1902663824788233e-07, + "loss": 1.7702, + "step": 11845 + }, + { + "epoch": 0.7984906168929619, + "grad_norm": 5.756847394591811, + "learning_rate": 1.1864605797502031e-07, + "loss": 1.7993, + "step": 11850 + }, + { + "epoch": 0.7988275327650686, + "grad_norm": 5.412926796228684, + "learning_rate": 1.1826600519324237e-07, + "loss": 1.8253, + "step": 11855 + }, + { + "epoch": 0.7991644486371753, + "grad_norm": 6.040919321374024, + "learning_rate": 1.1788648042823956e-07, + "loss": 1.7589, + "step": 11860 + }, + { + "epoch": 0.799501364509282, + "grad_norm": 6.478672926194837, + "learning_rate": 1.1750748420497298e-07, + "loss": 1.7519, + "step": 11865 + }, + { + "epoch": 0.7998382803813888, + "grad_norm": 6.375820517208503, + "learning_rate": 1.1712901704767253e-07, + "loss": 1.7681, + "step": 11870 + }, + { + "epoch": 0.8001751962534955, + "grad_norm": 5.709068291536156, + "learning_rate": 1.1675107947983615e-07, + "loss": 1.8144, + "step": 11875 + }, + { + "epoch": 0.8005121121256022, + "grad_norm": 6.054350290149295, + "learning_rate": 1.1637367202422943e-07, + "loss": 1.8536, + "step": 11880 + }, + { + "epoch": 0.800849027997709, + "grad_norm": 5.698352041079042, + "learning_rate": 1.159967952028848e-07, + "loss": 1.8335, + "step": 11885 + }, + { + "epoch": 0.8011859438698157, + "grad_norm": 6.031855898334881, + "learning_rate": 1.1562044953710032e-07, + "loss": 1.7183, + "step": 11890 + }, + { + "epoch": 0.8015228597419224, + "grad_norm": 5.625308544157251, + "learning_rate": 1.152446355474398e-07, + "loss": 1.7847, + "step": 11895 + }, + { + "epoch": 0.8018597756140292, + "grad_norm": 6.683973873807723, + "learning_rate": 1.1486935375373124e-07, + "loss": 1.8403, + "step": 11900 + }, + { + "epoch": 0.8021966914861359, + "grad_norm": 6.641861847519315, + "learning_rate": 1.1449460467506689e-07, + "loss": 1.7658, + "step": 11905 + }, + { + "epoch": 0.8025336073582426, + "grad_norm": 5.610412974235367, + "learning_rate": 1.1412038882980174e-07, + "loss": 1.7931, + "step": 11910 + }, + { + "epoch": 0.8028705232303494, + "grad_norm": 5.685183233588973, + "learning_rate": 1.1374670673555348e-07, + "loss": 1.7837, + "step": 11915 + }, + { + "epoch": 0.8032074391024561, + "grad_norm": 5.611660132961658, + "learning_rate": 1.1337355890920169e-07, + "loss": 1.8168, + "step": 11920 + }, + { + "epoch": 0.8035443549745629, + "grad_norm": 5.403506032815681, + "learning_rate": 1.130009458668863e-07, + "loss": 1.7992, + "step": 11925 + }, + { + "epoch": 0.8038812708466696, + "grad_norm": 5.79197296037922, + "learning_rate": 1.1262886812400813e-07, + "loss": 1.7595, + "step": 11930 + }, + { + "epoch": 0.8042181867187763, + "grad_norm": 5.8569410896518574, + "learning_rate": 1.1225732619522754e-07, + "loss": 1.7474, + "step": 11935 + }, + { + "epoch": 0.804555102590883, + "grad_norm": 6.025795706365198, + "learning_rate": 1.118863205944633e-07, + "loss": 1.7532, + "step": 11940 + }, + { + "epoch": 0.8048920184629897, + "grad_norm": 6.168101917389683, + "learning_rate": 1.1151585183489266e-07, + "loss": 1.8026, + "step": 11945 + }, + { + "epoch": 0.8052289343350966, + "grad_norm": 5.827955134646608, + "learning_rate": 1.1114592042895044e-07, + "loss": 1.775, + "step": 11950 + }, + { + "epoch": 0.8055658502072033, + "grad_norm": 6.076578864373635, + "learning_rate": 1.1077652688832772e-07, + "loss": 1.7875, + "step": 11955 + }, + { + "epoch": 0.80590276607931, + "grad_norm": 5.7168850287245245, + "learning_rate": 1.1040767172397209e-07, + "loss": 1.7375, + "step": 11960 + }, + { + "epoch": 0.8062396819514167, + "grad_norm": 6.282872282911842, + "learning_rate": 1.1003935544608612e-07, + "loss": 1.777, + "step": 11965 + }, + { + "epoch": 0.8065765978235234, + "grad_norm": 5.887798609993405, + "learning_rate": 1.0967157856412739e-07, + "loss": 1.7622, + "step": 11970 + }, + { + "epoch": 0.8069135136956302, + "grad_norm": 6.199457852026214, + "learning_rate": 1.093043415868069e-07, + "loss": 1.8875, + "step": 11975 + }, + { + "epoch": 0.807250429567737, + "grad_norm": 5.745866008462109, + "learning_rate": 1.0893764502208891e-07, + "loss": 1.8143, + "step": 11980 + }, + { + "epoch": 0.8075873454398437, + "grad_norm": 6.110266261481142, + "learning_rate": 1.0857148937719063e-07, + "loss": 1.7584, + "step": 11985 + }, + { + "epoch": 0.8079242613119504, + "grad_norm": 5.969686237894243, + "learning_rate": 1.0820587515858054e-07, + "loss": 1.8105, + "step": 11990 + }, + { + "epoch": 0.8082611771840571, + "grad_norm": 5.759859891411606, + "learning_rate": 1.078408028719785e-07, + "loss": 1.7588, + "step": 11995 + }, + { + "epoch": 0.8085980930561639, + "grad_norm": 5.579575443228889, + "learning_rate": 1.0747627302235491e-07, + "loss": 1.8207, + "step": 12000 + }, + { + "epoch": 0.8089350089282706, + "grad_norm": 6.213088918697938, + "learning_rate": 1.0711228611392936e-07, + "loss": 1.8158, + "step": 12005 + }, + { + "epoch": 0.8092719248003774, + "grad_norm": 6.19662993284419, + "learning_rate": 1.0674884265017086e-07, + "loss": 1.7965, + "step": 12010 + }, + { + "epoch": 0.8096088406724841, + "grad_norm": 5.992976800633545, + "learning_rate": 1.0638594313379678e-07, + "loss": 1.7008, + "step": 12015 + }, + { + "epoch": 0.8099457565445908, + "grad_norm": 6.142320900455945, + "learning_rate": 1.060235880667717e-07, + "loss": 1.839, + "step": 12020 + }, + { + "epoch": 0.8102826724166976, + "grad_norm": 5.560359240436788, + "learning_rate": 1.056617779503074e-07, + "loss": 1.7725, + "step": 12025 + }, + { + "epoch": 0.8106195882888043, + "grad_norm": 6.437317541471631, + "learning_rate": 1.053005132848619e-07, + "loss": 1.7889, + "step": 12030 + }, + { + "epoch": 0.810956504160911, + "grad_norm": 5.943008760322463, + "learning_rate": 1.0493979457013874e-07, + "loss": 1.8495, + "step": 12035 + }, + { + "epoch": 0.8112934200330177, + "grad_norm": 5.612578556134183, + "learning_rate": 1.0457962230508599e-07, + "loss": 1.768, + "step": 12040 + }, + { + "epoch": 0.8116303359051245, + "grad_norm": 6.331371365186378, + "learning_rate": 1.042199969878963e-07, + "loss": 1.7489, + "step": 12045 + }, + { + "epoch": 0.8119672517772313, + "grad_norm": 5.594775080040045, + "learning_rate": 1.0386091911600564e-07, + "loss": 1.7299, + "step": 12050 + }, + { + "epoch": 0.812304167649338, + "grad_norm": 5.993230773092134, + "learning_rate": 1.0350238918609244e-07, + "loss": 1.7439, + "step": 12055 + }, + { + "epoch": 0.8126410835214447, + "grad_norm": 5.79719115106528, + "learning_rate": 1.0314440769407784e-07, + "loss": 1.7425, + "step": 12060 + }, + { + "epoch": 0.8129779993935514, + "grad_norm": 6.163239487153185, + "learning_rate": 1.0278697513512375e-07, + "loss": 1.6992, + "step": 12065 + }, + { + "epoch": 0.8133149152656581, + "grad_norm": 6.019555217322019, + "learning_rate": 1.0243009200363295e-07, + "loss": 1.7035, + "step": 12070 + }, + { + "epoch": 0.8136518311377648, + "grad_norm": 6.411126780956076, + "learning_rate": 1.0207375879324853e-07, + "loss": 1.7633, + "step": 12075 + }, + { + "epoch": 0.8139887470098717, + "grad_norm": 6.560863671091432, + "learning_rate": 1.0171797599685284e-07, + "loss": 1.8065, + "step": 12080 + }, + { + "epoch": 0.8143256628819784, + "grad_norm": 5.61227341277097, + "learning_rate": 1.0136274410656653e-07, + "loss": 1.8068, + "step": 12085 + }, + { + "epoch": 0.8146625787540851, + "grad_norm": 5.998206753083885, + "learning_rate": 1.010080636137487e-07, + "loss": 1.773, + "step": 12090 + }, + { + "epoch": 0.8149994946261918, + "grad_norm": 6.074166764653066, + "learning_rate": 1.0065393500899549e-07, + "loss": 1.695, + "step": 12095 + }, + { + "epoch": 0.8153364104982985, + "grad_norm": 5.886255551150027, + "learning_rate": 1.0030035878213988e-07, + "loss": 1.7873, + "step": 12100 + }, + { + "epoch": 0.8156733263704053, + "grad_norm": 5.998981616215432, + "learning_rate": 9.994733542225037e-08, + "loss": 1.7402, + "step": 12105 + }, + { + "epoch": 0.8160102422425121, + "grad_norm": 5.68317572627976, + "learning_rate": 9.959486541763118e-08, + "loss": 1.7603, + "step": 12110 + }, + { + "epoch": 0.8163471581146188, + "grad_norm": 5.900621893113269, + "learning_rate": 9.924294925582105e-08, + "loss": 1.8141, + "step": 12115 + }, + { + "epoch": 0.8166840739867255, + "grad_norm": 6.333895946186688, + "learning_rate": 9.889158742359233e-08, + "loss": 1.7808, + "step": 12120 + }, + { + "epoch": 0.8170209898588322, + "grad_norm": 5.947606154639251, + "learning_rate": 9.854078040695096e-08, + "loss": 1.8038, + "step": 12125 + }, + { + "epoch": 0.817357905730939, + "grad_norm": 5.642752987160135, + "learning_rate": 9.819052869113543e-08, + "loss": 1.7817, + "step": 12130 + }, + { + "epoch": 0.8176948216030457, + "grad_norm": 6.10066295676423, + "learning_rate": 9.784083276061578e-08, + "loss": 1.773, + "step": 12135 + }, + { + "epoch": 0.8180317374751525, + "grad_norm": 6.390361061803448, + "learning_rate": 9.749169309909382e-08, + "loss": 1.7252, + "step": 12140 + }, + { + "epoch": 0.8183686533472592, + "grad_norm": 5.605942954144895, + "learning_rate": 9.714311018950139e-08, + "loss": 1.7809, + "step": 12145 + }, + { + "epoch": 0.8187055692193659, + "grad_norm": 5.633651526457674, + "learning_rate": 9.67950845140007e-08, + "loss": 1.8077, + "step": 12150 + }, + { + "epoch": 0.8190424850914727, + "grad_norm": 5.803438387797127, + "learning_rate": 9.644761655398276e-08, + "loss": 1.7152, + "step": 12155 + }, + { + "epoch": 0.8193794009635794, + "grad_norm": 6.908491498949583, + "learning_rate": 9.61007067900675e-08, + "loss": 1.835, + "step": 12160 + }, + { + "epoch": 0.8197163168356861, + "grad_norm": 5.867281972222165, + "learning_rate": 9.575435570210266e-08, + "loss": 1.8139, + "step": 12165 + }, + { + "epoch": 0.8200532327077928, + "grad_norm": 5.7296379548628105, + "learning_rate": 9.5408563769163e-08, + "loss": 1.7873, + "step": 12170 + }, + { + "epoch": 0.8203901485798996, + "grad_norm": 5.1827663356538896, + "learning_rate": 9.506333146955009e-08, + "loss": 1.7037, + "step": 12175 + }, + { + "epoch": 0.8207270644520064, + "grad_norm": 5.3980116057512335, + "learning_rate": 9.471865928079148e-08, + "loss": 1.7945, + "step": 12180 + }, + { + "epoch": 0.8210639803241131, + "grad_norm": 5.4658793637910925, + "learning_rate": 9.437454767963954e-08, + "loss": 1.7733, + "step": 12185 + }, + { + "epoch": 0.8214008961962198, + "grad_norm": 6.1213965153921075, + "learning_rate": 9.403099714207174e-08, + "loss": 1.7969, + "step": 12190 + }, + { + "epoch": 0.8217378120683265, + "grad_norm": 5.838123846027778, + "learning_rate": 9.368800814328931e-08, + "loss": 1.8021, + "step": 12195 + }, + { + "epoch": 0.8220747279404332, + "grad_norm": 6.339463633874931, + "learning_rate": 9.334558115771646e-08, + "loss": 1.789, + "step": 12200 + }, + { + "epoch": 0.8224116438125401, + "grad_norm": 5.5623582051309155, + "learning_rate": 9.300371665900048e-08, + "loss": 1.7557, + "step": 12205 + }, + { + "epoch": 0.8227485596846468, + "grad_norm": 6.361797379331405, + "learning_rate": 9.266241512001044e-08, + "loss": 1.7047, + "step": 12210 + }, + { + "epoch": 0.8230854755567535, + "grad_norm": 5.952969064534977, + "learning_rate": 9.23216770128365e-08, + "loss": 1.7918, + "step": 12215 + }, + { + "epoch": 0.8234223914288602, + "grad_norm": 6.3011377196826155, + "learning_rate": 9.19815028087898e-08, + "loss": 1.7846, + "step": 12220 + }, + { + "epoch": 0.8237593073009669, + "grad_norm": 5.5847611704130316, + "learning_rate": 9.164189297840147e-08, + "loss": 1.8166, + "step": 12225 + }, + { + "epoch": 0.8240962231730737, + "grad_norm": 6.510883218408153, + "learning_rate": 9.130284799142179e-08, + "loss": 1.7582, + "step": 12230 + }, + { + "epoch": 0.8244331390451805, + "grad_norm": 6.09019683484805, + "learning_rate": 9.09643683168197e-08, + "loss": 1.775, + "step": 12235 + }, + { + "epoch": 0.8247700549172872, + "grad_norm": 5.852527126802162, + "learning_rate": 9.062645442278244e-08, + "loss": 1.8143, + "step": 12240 + }, + { + "epoch": 0.8251069707893939, + "grad_norm": 5.8185776473103505, + "learning_rate": 9.028910677671469e-08, + "loss": 1.86, + "step": 12245 + }, + { + "epoch": 0.8254438866615006, + "grad_norm": 6.0410974739836405, + "learning_rate": 8.995232584523754e-08, + "loss": 1.7393, + "step": 12250 + }, + { + "epoch": 0.8257808025336073, + "grad_norm": 6.076512425747793, + "learning_rate": 8.961611209418851e-08, + "loss": 1.783, + "step": 12255 + }, + { + "epoch": 0.8261177184057141, + "grad_norm": 5.948019238377317, + "learning_rate": 8.928046598862065e-08, + "loss": 1.7837, + "step": 12260 + }, + { + "epoch": 0.8264546342778208, + "grad_norm": 6.210822208377314, + "learning_rate": 8.894538799280138e-08, + "loss": 1.8413, + "step": 12265 + }, + { + "epoch": 0.8267915501499276, + "grad_norm": 5.430421973136374, + "learning_rate": 8.861087857021282e-08, + "loss": 1.7717, + "step": 12270 + }, + { + "epoch": 0.8271284660220343, + "grad_norm": 6.478405775826868, + "learning_rate": 8.827693818355048e-08, + "loss": 1.6905, + "step": 12275 + }, + { + "epoch": 0.827465381894141, + "grad_norm": 6.100376083406547, + "learning_rate": 8.794356729472252e-08, + "loss": 1.7629, + "step": 12280 + }, + { + "epoch": 0.8278022977662478, + "grad_norm": 5.678122613542731, + "learning_rate": 8.76107663648497e-08, + "loss": 1.7382, + "step": 12285 + }, + { + "epoch": 0.8281392136383545, + "grad_norm": 5.648135451717112, + "learning_rate": 8.727853585426436e-08, + "loss": 1.8082, + "step": 12290 + }, + { + "epoch": 0.8284761295104612, + "grad_norm": 5.782391731975333, + "learning_rate": 8.694687622250963e-08, + "loss": 1.7921, + "step": 12295 + }, + { + "epoch": 0.828813045382568, + "grad_norm": 5.854072035771789, + "learning_rate": 8.661578792833907e-08, + "loss": 1.8038, + "step": 12300 + }, + { + "epoch": 0.8291499612546747, + "grad_norm": 5.731142272858889, + "learning_rate": 8.628527142971632e-08, + "loss": 1.6797, + "step": 12305 + }, + { + "epoch": 0.8294868771267815, + "grad_norm": 5.81642283052816, + "learning_rate": 8.595532718381338e-08, + "loss": 1.7891, + "step": 12310 + }, + { + "epoch": 0.8298237929988882, + "grad_norm": 6.010154250940546, + "learning_rate": 8.562595564701153e-08, + "loss": 1.8121, + "step": 12315 + }, + { + "epoch": 0.8301607088709949, + "grad_norm": 6.179374881780338, + "learning_rate": 8.529715727489912e-08, + "loss": 1.8339, + "step": 12320 + }, + { + "epoch": 0.8304976247431016, + "grad_norm": 5.734377900742119, + "learning_rate": 8.496893252227238e-08, + "loss": 1.7897, + "step": 12325 + }, + { + "epoch": 0.8308345406152083, + "grad_norm": 5.466608782130143, + "learning_rate": 8.464128184313346e-08, + "loss": 1.7676, + "step": 12330 + }, + { + "epoch": 0.8311714564873152, + "grad_norm": 6.156550917567134, + "learning_rate": 8.431420569069093e-08, + "loss": 1.7143, + "step": 12335 + }, + { + "epoch": 0.8315083723594219, + "grad_norm": 6.212082720502393, + "learning_rate": 8.398770451735865e-08, + "loss": 1.7742, + "step": 12340 + }, + { + "epoch": 0.8318452882315286, + "grad_norm": 6.203682522334206, + "learning_rate": 8.366177877475473e-08, + "loss": 1.7803, + "step": 12345 + }, + { + "epoch": 0.8321822041036353, + "grad_norm": 5.773188612460619, + "learning_rate": 8.333642891370174e-08, + "loss": 1.8809, + "step": 12350 + }, + { + "epoch": 0.832519119975742, + "grad_norm": 5.572837441862478, + "learning_rate": 8.301165538422577e-08, + "loss": 1.8031, + "step": 12355 + }, + { + "epoch": 0.8328560358478488, + "grad_norm": 6.1136780454024136, + "learning_rate": 8.268745863555521e-08, + "loss": 1.7058, + "step": 12360 + }, + { + "epoch": 0.8331929517199556, + "grad_norm": 6.072274649537594, + "learning_rate": 8.236383911612116e-08, + "loss": 1.7625, + "step": 12365 + }, + { + "epoch": 0.8335298675920623, + "grad_norm": 5.642550548273279, + "learning_rate": 8.204079727355611e-08, + "loss": 1.7976, + "step": 12370 + }, + { + "epoch": 0.833866783464169, + "grad_norm": 6.656470966965112, + "learning_rate": 8.171833355469354e-08, + "loss": 1.8501, + "step": 12375 + }, + { + "epoch": 0.8342036993362757, + "grad_norm": 6.377445803343679, + "learning_rate": 8.139644840556703e-08, + "loss": 1.7944, + "step": 12380 + }, + { + "epoch": 0.8345406152083825, + "grad_norm": 6.003788291285669, + "learning_rate": 8.107514227141032e-08, + "loss": 1.7504, + "step": 12385 + }, + { + "epoch": 0.8348775310804892, + "grad_norm": 5.763130376288233, + "learning_rate": 8.075441559665569e-08, + "loss": 1.8019, + "step": 12390 + }, + { + "epoch": 0.835214446952596, + "grad_norm": 5.365364742414105, + "learning_rate": 8.04342688249346e-08, + "loss": 1.8251, + "step": 12395 + }, + { + "epoch": 0.8355513628247027, + "grad_norm": 5.943722593508419, + "learning_rate": 8.011470239907558e-08, + "loss": 1.7841, + "step": 12400 + }, + { + "epoch": 0.8358882786968094, + "grad_norm": 6.280412446558371, + "learning_rate": 7.979571676110525e-08, + "loss": 1.7724, + "step": 12405 + }, + { + "epoch": 0.8362251945689162, + "grad_norm": 6.41052769353216, + "learning_rate": 7.947731235224614e-08, + "loss": 1.8516, + "step": 12410 + }, + { + "epoch": 0.8365621104410229, + "grad_norm": 5.8087270438920156, + "learning_rate": 7.915948961291729e-08, + "loss": 1.7758, + "step": 12415 + }, + { + "epoch": 0.8368990263131296, + "grad_norm": 5.948851731569554, + "learning_rate": 7.884224898273322e-08, + "loss": 1.7559, + "step": 12420 + }, + { + "epoch": 0.8372359421852363, + "grad_norm": 6.26399600567096, + "learning_rate": 7.852559090050276e-08, + "loss": 1.7629, + "step": 12425 + }, + { + "epoch": 0.837572858057343, + "grad_norm": 6.214317608558152, + "learning_rate": 7.820951580422952e-08, + "loss": 1.7334, + "step": 12430 + }, + { + "epoch": 0.8379097739294498, + "grad_norm": 5.705699682881419, + "learning_rate": 7.789402413111041e-08, + "loss": 1.7162, + "step": 12435 + }, + { + "epoch": 0.8382466898015566, + "grad_norm": 5.854199053151604, + "learning_rate": 7.757911631753556e-08, + "loss": 1.7659, + "step": 12440 + }, + { + "epoch": 0.8385836056736633, + "grad_norm": 5.782064879077919, + "learning_rate": 7.72647927990871e-08, + "loss": 1.8207, + "step": 12445 + }, + { + "epoch": 0.83892052154577, + "grad_norm": 5.730477635589501, + "learning_rate": 7.695105401053942e-08, + "loss": 1.8184, + "step": 12450 + }, + { + "epoch": 0.8392574374178767, + "grad_norm": 6.009518784132229, + "learning_rate": 7.663790038585794e-08, + "loss": 1.8255, + "step": 12455 + }, + { + "epoch": 0.8395943532899834, + "grad_norm": 5.80924413423034, + "learning_rate": 7.63253323581985e-08, + "loss": 1.774, + "step": 12460 + }, + { + "epoch": 0.8399312691620903, + "grad_norm": 6.029972989497406, + "learning_rate": 7.601335035990714e-08, + "loss": 1.8408, + "step": 12465 + }, + { + "epoch": 0.840268185034197, + "grad_norm": 6.012125031968545, + "learning_rate": 7.57019548225194e-08, + "loss": 1.7452, + "step": 12470 + }, + { + "epoch": 0.8406051009063037, + "grad_norm": 6.3070354195548255, + "learning_rate": 7.539114617675941e-08, + "loss": 1.8147, + "step": 12475 + }, + { + "epoch": 0.8409420167784104, + "grad_norm": 5.467419693731685, + "learning_rate": 7.508092485253936e-08, + "loss": 1.7499, + "step": 12480 + }, + { + "epoch": 0.8412789326505171, + "grad_norm": 6.055406576016901, + "learning_rate": 7.477129127895954e-08, + "loss": 1.8029, + "step": 12485 + }, + { + "epoch": 0.8416158485226239, + "grad_norm": 5.736119441083093, + "learning_rate": 7.446224588430678e-08, + "loss": 1.7793, + "step": 12490 + }, + { + "epoch": 0.8419527643947307, + "grad_norm": 5.502943142670245, + "learning_rate": 7.415378909605457e-08, + "loss": 1.8578, + "step": 12495 + }, + { + "epoch": 0.8422896802668374, + "grad_norm": 6.003707840545835, + "learning_rate": 7.384592134086231e-08, + "loss": 1.7574, + "step": 12500 + }, + { + "epoch": 0.8426265961389441, + "grad_norm": 6.0955461129701005, + "learning_rate": 7.353864304457463e-08, + "loss": 1.8106, + "step": 12505 + }, + { + "epoch": 0.8429635120110508, + "grad_norm": 6.613263301440216, + "learning_rate": 7.323195463222054e-08, + "loss": 1.7904, + "step": 12510 + }, + { + "epoch": 0.8433004278831576, + "grad_norm": 5.851838960250193, + "learning_rate": 7.292585652801331e-08, + "loss": 1.7135, + "step": 12515 + }, + { + "epoch": 0.8436373437552643, + "grad_norm": 5.870921208338694, + "learning_rate": 7.262034915534993e-08, + "loss": 1.7985, + "step": 12520 + }, + { + "epoch": 0.843974259627371, + "grad_norm": 6.341944056999602, + "learning_rate": 7.231543293680969e-08, + "loss": 1.7518, + "step": 12525 + }, + { + "epoch": 0.8443111754994778, + "grad_norm": 6.7081791221760065, + "learning_rate": 7.20111082941548e-08, + "loss": 1.8246, + "step": 12530 + }, + { + "epoch": 0.8446480913715845, + "grad_norm": 5.923306053836086, + "learning_rate": 7.170737564832902e-08, + "loss": 1.8312, + "step": 12535 + }, + { + "epoch": 0.8449850072436913, + "grad_norm": 6.095365797264817, + "learning_rate": 7.14042354194569e-08, + "loss": 1.8126, + "step": 12540 + }, + { + "epoch": 0.845321923115798, + "grad_norm": 5.684604484852161, + "learning_rate": 7.110168802684408e-08, + "loss": 1.7082, + "step": 12545 + }, + { + "epoch": 0.8456588389879047, + "grad_norm": 5.725222684944491, + "learning_rate": 7.079973388897592e-08, + "loss": 1.8111, + "step": 12550 + }, + { + "epoch": 0.8459957548600114, + "grad_norm": 5.693320014755603, + "learning_rate": 7.049837342351706e-08, + "loss": 1.7548, + "step": 12555 + }, + { + "epoch": 0.8463326707321182, + "grad_norm": 6.398363624106793, + "learning_rate": 7.019760704731131e-08, + "loss": 1.8213, + "step": 12560 + }, + { + "epoch": 0.846669586604225, + "grad_norm": 5.860647280733894, + "learning_rate": 6.989743517638053e-08, + "loss": 1.7709, + "step": 12565 + }, + { + "epoch": 0.8470065024763317, + "grad_norm": 6.148625041417426, + "learning_rate": 6.959785822592402e-08, + "loss": 1.826, + "step": 12570 + }, + { + "epoch": 0.8473434183484384, + "grad_norm": 6.417824137742353, + "learning_rate": 6.929887661031864e-08, + "loss": 1.7246, + "step": 12575 + }, + { + "epoch": 0.8476803342205451, + "grad_norm": 5.64895725010932, + "learning_rate": 6.900049074311753e-08, + "loss": 1.7656, + "step": 12580 + }, + { + "epoch": 0.8480172500926518, + "grad_norm": 5.972290040076362, + "learning_rate": 6.870270103705e-08, + "loss": 1.7489, + "step": 12585 + }, + { + "epoch": 0.8483541659647587, + "grad_norm": 5.887133857519855, + "learning_rate": 6.840550790402027e-08, + "loss": 1.7962, + "step": 12590 + }, + { + "epoch": 0.8486910818368654, + "grad_norm": 5.782128457023704, + "learning_rate": 6.810891175510792e-08, + "loss": 1.8129, + "step": 12595 + }, + { + "epoch": 0.8490279977089721, + "grad_norm": 5.874892138790784, + "learning_rate": 6.781291300056647e-08, + "loss": 1.7947, + "step": 12600 + }, + { + "epoch": 0.8493649135810788, + "grad_norm": 5.812368931391869, + "learning_rate": 6.751751204982309e-08, + "loss": 1.6731, + "step": 12605 + }, + { + "epoch": 0.8497018294531855, + "grad_norm": 6.020082431564708, + "learning_rate": 6.722270931147827e-08, + "loss": 1.8078, + "step": 12610 + }, + { + "epoch": 0.8500387453252922, + "grad_norm": 5.960115495572332, + "learning_rate": 6.692850519330506e-08, + "loss": 1.7458, + "step": 12615 + }, + { + "epoch": 0.850375661197399, + "grad_norm": 6.307539795503096, + "learning_rate": 6.66349001022481e-08, + "loss": 1.7037, + "step": 12620 + }, + { + "epoch": 0.8507125770695058, + "grad_norm": 5.545091703305936, + "learning_rate": 6.634189444442389e-08, + "loss": 1.8488, + "step": 12625 + }, + { + "epoch": 0.8510494929416125, + "grad_norm": 6.254283824668138, + "learning_rate": 6.604948862511977e-08, + "loss": 1.8558, + "step": 12630 + }, + { + "epoch": 0.8513864088137192, + "grad_norm": 6.109361798563485, + "learning_rate": 6.575768304879292e-08, + "loss": 1.8072, + "step": 12635 + }, + { + "epoch": 0.8517233246858259, + "grad_norm": 5.779243983837955, + "learning_rate": 6.546647811907091e-08, + "loss": 1.6969, + "step": 12640 + }, + { + "epoch": 0.8520602405579327, + "grad_norm": 5.6901474533716385, + "learning_rate": 6.517587423874988e-08, + "loss": 1.8094, + "step": 12645 + }, + { + "epoch": 0.8523971564300394, + "grad_norm": 5.452805325410532, + "learning_rate": 6.48858718097951e-08, + "loss": 1.695, + "step": 12650 + }, + { + "epoch": 0.8527340723021462, + "grad_norm": 5.545529646061517, + "learning_rate": 6.459647123333956e-08, + "loss": 1.6861, + "step": 12655 + }, + { + "epoch": 0.8530709881742529, + "grad_norm": 6.1029815867327, + "learning_rate": 6.430767290968387e-08, + "loss": 1.7565, + "step": 12660 + }, + { + "epoch": 0.8534079040463596, + "grad_norm": 5.876283483843472, + "learning_rate": 6.401947723829576e-08, + "loss": 1.7649, + "step": 12665 + }, + { + "epoch": 0.8537448199184664, + "grad_norm": 5.860811960055388, + "learning_rate": 6.373188461780904e-08, + "loss": 1.7763, + "step": 12670 + }, + { + "epoch": 0.8540817357905731, + "grad_norm": 5.71660515150186, + "learning_rate": 6.344489544602371e-08, + "loss": 1.8172, + "step": 12675 + }, + { + "epoch": 0.8544186516626798, + "grad_norm": 6.166026903880695, + "learning_rate": 6.315851011990498e-08, + "loss": 1.8399, + "step": 12680 + }, + { + "epoch": 0.8547555675347865, + "grad_norm": 5.601035725888883, + "learning_rate": 6.28727290355826e-08, + "loss": 1.7438, + "step": 12685 + }, + { + "epoch": 0.8550924834068933, + "grad_norm": 6.425970981523913, + "learning_rate": 6.258755258835075e-08, + "loss": 1.7938, + "step": 12690 + }, + { + "epoch": 0.8554293992790001, + "grad_norm": 5.882378548711108, + "learning_rate": 6.230298117266736e-08, + "loss": 1.7494, + "step": 12695 + }, + { + "epoch": 0.8557663151511068, + "grad_norm": 6.343902811827818, + "learning_rate": 6.201901518215313e-08, + "loss": 1.764, + "step": 12700 + }, + { + "epoch": 0.8561032310232135, + "grad_norm": 6.253978212706272, + "learning_rate": 6.173565500959165e-08, + "loss": 1.8211, + "step": 12705 + }, + { + "epoch": 0.8564401468953202, + "grad_norm": 6.608042387592941, + "learning_rate": 6.14529010469284e-08, + "loss": 1.819, + "step": 12710 + }, + { + "epoch": 0.8567770627674269, + "grad_norm": 5.752172995272666, + "learning_rate": 6.117075368527053e-08, + "loss": 1.6747, + "step": 12715 + }, + { + "epoch": 0.8571139786395338, + "grad_norm": 5.840794832771322, + "learning_rate": 6.088921331488566e-08, + "loss": 1.7578, + "step": 12720 + }, + { + "epoch": 0.8574508945116405, + "grad_norm": 5.849034173026425, + "learning_rate": 6.060828032520249e-08, + "loss": 1.7759, + "step": 12725 + }, + { + "epoch": 0.8577878103837472, + "grad_norm": 6.213661331265872, + "learning_rate": 6.032795510480904e-08, + "loss": 1.7317, + "step": 12730 + }, + { + "epoch": 0.8581247262558539, + "grad_norm": 5.6898733401141515, + "learning_rate": 6.004823804145276e-08, + "loss": 1.8348, + "step": 12735 + }, + { + "epoch": 0.8584616421279606, + "grad_norm": 5.859861873519631, + "learning_rate": 5.976912952204016e-08, + "loss": 1.797, + "step": 12740 + }, + { + "epoch": 0.8587985580000674, + "grad_norm": 5.76133259829159, + "learning_rate": 5.9490629932635815e-08, + "loss": 1.8532, + "step": 12745 + }, + { + "epoch": 0.8591354738721741, + "grad_norm": 6.129986293259829, + "learning_rate": 5.921273965846191e-08, + "loss": 1.7627, + "step": 12750 + }, + { + "epoch": 0.8594723897442809, + "grad_norm": 5.945438078193672, + "learning_rate": 5.893545908389807e-08, + "loss": 1.6993, + "step": 12755 + }, + { + "epoch": 0.8598093056163876, + "grad_norm": 5.893702356532198, + "learning_rate": 5.865878859248058e-08, + "loss": 1.7023, + "step": 12760 + }, + { + "epoch": 0.8601462214884943, + "grad_norm": 6.134347621385907, + "learning_rate": 5.838272856690146e-08, + "loss": 1.7135, + "step": 12765 + }, + { + "epoch": 0.860483137360601, + "grad_norm": 5.8408265930376775, + "learning_rate": 5.810727938900878e-08, + "loss": 1.7885, + "step": 12770 + }, + { + "epoch": 0.8608200532327078, + "grad_norm": 5.967188418245232, + "learning_rate": 5.7832441439805536e-08, + "loss": 1.8186, + "step": 12775 + }, + { + "epoch": 0.8611569691048145, + "grad_norm": 5.883049229755388, + "learning_rate": 5.755821509944925e-08, + "loss": 1.7955, + "step": 12780 + }, + { + "epoch": 0.8614938849769213, + "grad_norm": 6.2544522511675895, + "learning_rate": 5.728460074725133e-08, + "loss": 1.7627, + "step": 12785 + }, + { + "epoch": 0.861830800849028, + "grad_norm": 5.786365914538278, + "learning_rate": 5.701159876167688e-08, + "loss": 1.7497, + "step": 12790 + }, + { + "epoch": 0.8621677167211347, + "grad_norm": 6.071667040551235, + "learning_rate": 5.673920952034406e-08, + "loss": 1.7676, + "step": 12795 + }, + { + "epoch": 0.8625046325932415, + "grad_norm": 5.83884850067031, + "learning_rate": 5.646743340002302e-08, + "loss": 1.7544, + "step": 12800 + }, + { + "epoch": 0.8628415484653482, + "grad_norm": 6.399413975233235, + "learning_rate": 5.619627077663636e-08, + "loss": 1.8448, + "step": 12805 + }, + { + "epoch": 0.8631784643374549, + "grad_norm": 6.006783852696938, + "learning_rate": 5.5925722025257746e-08, + "loss": 1.8516, + "step": 12810 + }, + { + "epoch": 0.8635153802095616, + "grad_norm": 6.048382439323067, + "learning_rate": 5.5655787520111966e-08, + "loss": 1.8342, + "step": 12815 + }, + { + "epoch": 0.8638522960816684, + "grad_norm": 5.9506691951333135, + "learning_rate": 5.538646763457389e-08, + "loss": 1.7532, + "step": 12820 + }, + { + "epoch": 0.8641892119537752, + "grad_norm": 5.7101654823015116, + "learning_rate": 5.511776274116864e-08, + "loss": 1.7317, + "step": 12825 + }, + { + "epoch": 0.8645261278258819, + "grad_norm": 6.3845287996072475, + "learning_rate": 5.484967321157019e-08, + "loss": 1.7862, + "step": 12830 + }, + { + "epoch": 0.8648630436979886, + "grad_norm": 6.234434470893632, + "learning_rate": 5.4582199416601746e-08, + "loss": 1.7275, + "step": 12835 + }, + { + "epoch": 0.8651999595700953, + "grad_norm": 6.22322611604707, + "learning_rate": 5.43153417262347e-08, + "loss": 1.8236, + "step": 12840 + }, + { + "epoch": 0.865536875442202, + "grad_norm": 6.468448015402212, + "learning_rate": 5.404910050958833e-08, + "loss": 1.7972, + "step": 12845 + }, + { + "epoch": 0.8658737913143089, + "grad_norm": 5.814429402895073, + "learning_rate": 5.378347613492884e-08, + "loss": 1.7756, + "step": 12850 + }, + { + "epoch": 0.8662107071864156, + "grad_norm": 5.7511147012293, + "learning_rate": 5.351846896966966e-08, + "loss": 1.8017, + "step": 12855 + }, + { + "epoch": 0.8665476230585223, + "grad_norm": 6.609525748717229, + "learning_rate": 5.32540793803703e-08, + "loss": 1.7982, + "step": 12860 + }, + { + "epoch": 0.866884538930629, + "grad_norm": 5.593059673213316, + "learning_rate": 5.299030773273594e-08, + "loss": 1.7657, + "step": 12865 + }, + { + "epoch": 0.8672214548027357, + "grad_norm": 6.277557730074275, + "learning_rate": 5.272715439161718e-08, + "loss": 1.7661, + "step": 12870 + }, + { + "epoch": 0.8675583706748425, + "grad_norm": 6.0356327137348575, + "learning_rate": 5.246461972100941e-08, + "loss": 1.7348, + "step": 12875 + }, + { + "epoch": 0.8678952865469493, + "grad_norm": 5.819516781146134, + "learning_rate": 5.220270408405197e-08, + "loss": 1.7576, + "step": 12880 + }, + { + "epoch": 0.868232202419056, + "grad_norm": 5.901222351752382, + "learning_rate": 5.194140784302836e-08, + "loss": 1.7543, + "step": 12885 + }, + { + "epoch": 0.8685691182911627, + "grad_norm": 5.102802335028869, + "learning_rate": 5.168073135936496e-08, + "loss": 1.7689, + "step": 12890 + }, + { + "epoch": 0.8689060341632694, + "grad_norm": 6.347703972438555, + "learning_rate": 5.1420674993631285e-08, + "loss": 1.7232, + "step": 12895 + }, + { + "epoch": 0.8692429500353762, + "grad_norm": 5.80051625292983, + "learning_rate": 5.116123910553854e-08, + "loss": 1.8073, + "step": 12900 + }, + { + "epoch": 0.8695798659074829, + "grad_norm": 5.540327273203947, + "learning_rate": 5.0902424053940406e-08, + "loss": 1.8433, + "step": 12905 + }, + { + "epoch": 0.8699167817795896, + "grad_norm": 5.9566276471635335, + "learning_rate": 5.064423019683106e-08, + "loss": 1.7564, + "step": 12910 + }, + { + "epoch": 0.8702536976516964, + "grad_norm": 6.138415912931555, + "learning_rate": 5.0386657891346e-08, + "loss": 1.7506, + "step": 12915 + }, + { + "epoch": 0.8705906135238031, + "grad_norm": 5.255444258029073, + "learning_rate": 5.012970749376083e-08, + "loss": 1.7524, + "step": 12920 + }, + { + "epoch": 0.8709275293959099, + "grad_norm": 6.279316610951135, + "learning_rate": 4.987337935949087e-08, + "loss": 1.8029, + "step": 12925 + }, + { + "epoch": 0.8712644452680166, + "grad_norm": 5.837133118530122, + "learning_rate": 4.961767384309068e-08, + "loss": 1.8114, + "step": 12930 + }, + { + "epoch": 0.8716013611401233, + "grad_norm": 5.4440077222971555, + "learning_rate": 4.936259129825376e-08, + "loss": 1.7704, + "step": 12935 + }, + { + "epoch": 0.87193827701223, + "grad_norm": 5.8098683047516495, + "learning_rate": 4.9108132077811836e-08, + "loss": 1.7495, + "step": 12940 + }, + { + "epoch": 0.8722751928843367, + "grad_norm": 6.079084932181276, + "learning_rate": 4.885429653373435e-08, + "loss": 1.7877, + "step": 12945 + }, + { + "epoch": 0.8726121087564435, + "grad_norm": 6.484226794140705, + "learning_rate": 4.860108501712823e-08, + "loss": 1.796, + "step": 12950 + }, + { + "epoch": 0.8729490246285503, + "grad_norm": 5.944286752397985, + "learning_rate": 4.834849787823725e-08, + "loss": 1.7539, + "step": 12955 + }, + { + "epoch": 0.873285940500657, + "grad_norm": 6.719801982192994, + "learning_rate": 4.809653546644132e-08, + "loss": 1.8585, + "step": 12960 + }, + { + "epoch": 0.8736228563727637, + "grad_norm": 6.021944012578693, + "learning_rate": 4.7845198130256395e-08, + "loss": 1.8295, + "step": 12965 + }, + { + "epoch": 0.8739597722448704, + "grad_norm": 5.682973842187247, + "learning_rate": 4.759448621733403e-08, + "loss": 1.753, + "step": 12970 + }, + { + "epoch": 0.8742966881169771, + "grad_norm": 6.027055507715319, + "learning_rate": 4.7344400074460276e-08, + "loss": 1.7675, + "step": 12975 + }, + { + "epoch": 0.874633603989084, + "grad_norm": 6.019246359137967, + "learning_rate": 4.709494004755571e-08, + "loss": 1.7944, + "step": 12980 + }, + { + "epoch": 0.8749705198611907, + "grad_norm": 6.205925692533681, + "learning_rate": 4.684610648167503e-08, + "loss": 1.7383, + "step": 12985 + }, + { + "epoch": 0.8753074357332974, + "grad_norm": 6.541486904892061, + "learning_rate": 4.659789972100647e-08, + "loss": 1.7923, + "step": 12990 + }, + { + "epoch": 0.8756443516054041, + "grad_norm": 5.7486065087911085, + "learning_rate": 4.635032010887097e-08, + "loss": 1.8291, + "step": 12995 + }, + { + "epoch": 0.8759812674775108, + "grad_norm": 5.637034264598444, + "learning_rate": 4.610336798772213e-08, + "loss": 1.7808, + "step": 13000 + }, + { + "epoch": 0.8763181833496176, + "grad_norm": 5.740217106810944, + "learning_rate": 4.5857043699145834e-08, + "loss": 1.8048, + "step": 13005 + }, + { + "epoch": 0.8766550992217244, + "grad_norm": 6.194258312455204, + "learning_rate": 4.5611347583859095e-08, + "loss": 1.7798, + "step": 13010 + }, + { + "epoch": 0.8769920150938311, + "grad_norm": 5.900531799707303, + "learning_rate": 4.536627998171033e-08, + "loss": 1.8157, + "step": 13015 + }, + { + "epoch": 0.8773289309659378, + "grad_norm": 6.470340001417238, + "learning_rate": 4.512184123167867e-08, + "loss": 1.74, + "step": 13020 + }, + { + "epoch": 0.8776658468380445, + "grad_norm": 5.817971283498176, + "learning_rate": 4.487803167187304e-08, + "loss": 1.7794, + "step": 13025 + }, + { + "epoch": 0.8780027627101513, + "grad_norm": 5.790877494976903, + "learning_rate": 4.463485163953246e-08, + "loss": 1.8452, + "step": 13030 + }, + { + "epoch": 0.878339678582258, + "grad_norm": 5.960756473041569, + "learning_rate": 4.4392301471025074e-08, + "loss": 1.8384, + "step": 13035 + }, + { + "epoch": 0.8786765944543647, + "grad_norm": 6.030953838356603, + "learning_rate": 4.415038150184758e-08, + "loss": 1.7726, + "step": 13040 + }, + { + "epoch": 0.8790135103264715, + "grad_norm": 6.256975255497656, + "learning_rate": 4.3909092066625245e-08, + "loss": 1.7452, + "step": 13045 + }, + { + "epoch": 0.8793504261985782, + "grad_norm": 5.843379556994611, + "learning_rate": 4.366843349911109e-08, + "loss": 1.7354, + "step": 13050 + }, + { + "epoch": 0.879687342070685, + "grad_norm": 6.33305600966883, + "learning_rate": 4.342840613218546e-08, + "loss": 1.7203, + "step": 13055 + }, + { + "epoch": 0.8800242579427917, + "grad_norm": 6.100154791697077, + "learning_rate": 4.318901029785571e-08, + "loss": 1.8078, + "step": 13060 + }, + { + "epoch": 0.8803611738148984, + "grad_norm": 5.77833918849904, + "learning_rate": 4.2950246327255523e-08, + "loss": 1.7375, + "step": 13065 + }, + { + "epoch": 0.8806980896870051, + "grad_norm": 5.645289655518625, + "learning_rate": 4.271211455064483e-08, + "loss": 1.8231, + "step": 13070 + }, + { + "epoch": 0.8810350055591119, + "grad_norm": 6.012684369012068, + "learning_rate": 4.2474615297408754e-08, + "loss": 1.7604, + "step": 13075 + }, + { + "epoch": 0.8813719214312187, + "grad_norm": 5.441699463373052, + "learning_rate": 4.223774889605775e-08, + "loss": 1.7773, + "step": 13080 + }, + { + "epoch": 0.8817088373033254, + "grad_norm": 6.0339587220276325, + "learning_rate": 4.200151567422699e-08, + "loss": 1.8166, + "step": 13085 + }, + { + "epoch": 0.8820457531754321, + "grad_norm": 6.067671922831784, + "learning_rate": 4.176591595867557e-08, + "loss": 1.8074, + "step": 13090 + }, + { + "epoch": 0.8823826690475388, + "grad_norm": 5.630193597022125, + "learning_rate": 4.153095007528645e-08, + "loss": 1.7895, + "step": 13095 + }, + { + "epoch": 0.8827195849196455, + "grad_norm": 5.897156624279815, + "learning_rate": 4.1296618349066e-08, + "loss": 1.723, + "step": 13100 + }, + { + "epoch": 0.8830565007917524, + "grad_norm": 5.59628907202942, + "learning_rate": 4.106292110414311e-08, + "loss": 1.8166, + "step": 13105 + }, + { + "epoch": 0.8833934166638591, + "grad_norm": 5.629179966186753, + "learning_rate": 4.082985866376926e-08, + "loss": 1.8097, + "step": 13110 + }, + { + "epoch": 0.8837303325359658, + "grad_norm": 6.399112941808455, + "learning_rate": 4.05974313503179e-08, + "loss": 1.8253, + "step": 13115 + }, + { + "epoch": 0.8840672484080725, + "grad_norm": 5.787318489151215, + "learning_rate": 4.036563948528393e-08, + "loss": 1.7877, + "step": 13120 + }, + { + "epoch": 0.8844041642801792, + "grad_norm": 5.519712124099288, + "learning_rate": 4.01344833892831e-08, + "loss": 1.8243, + "step": 13125 + }, + { + "epoch": 0.8847410801522859, + "grad_norm": 6.393037581588508, + "learning_rate": 3.990396338205204e-08, + "loss": 1.7897, + "step": 13130 + }, + { + "epoch": 0.8850779960243927, + "grad_norm": 6.216475201686147, + "learning_rate": 3.967407978244747e-08, + "loss": 1.6919, + "step": 13135 + }, + { + "epoch": 0.8854149118964995, + "grad_norm": 6.0911141234125, + "learning_rate": 3.944483290844575e-08, + "loss": 1.7612, + "step": 13140 + }, + { + "epoch": 0.8857518277686062, + "grad_norm": 6.14238918360982, + "learning_rate": 3.9216223077142394e-08, + "loss": 1.8431, + "step": 13145 + }, + { + "epoch": 0.8860887436407129, + "grad_norm": 5.873505883296761, + "learning_rate": 3.8988250604752135e-08, + "loss": 1.7477, + "step": 13150 + }, + { + "epoch": 0.8864256595128196, + "grad_norm": 6.636804154561093, + "learning_rate": 3.876091580660762e-08, + "loss": 1.8773, + "step": 13155 + }, + { + "epoch": 0.8867625753849264, + "grad_norm": 5.990492412539213, + "learning_rate": 3.853421899715992e-08, + "loss": 1.7506, + "step": 13160 + }, + { + "epoch": 0.8870994912570331, + "grad_norm": 5.796005866130032, + "learning_rate": 3.8308160489977424e-08, + "loss": 1.8137, + "step": 13165 + }, + { + "epoch": 0.8874364071291398, + "grad_norm": 6.212100150861587, + "learning_rate": 3.808274059774552e-08, + "loss": 1.8317, + "step": 13170 + }, + { + "epoch": 0.8877733230012466, + "grad_norm": 6.143632521898199, + "learning_rate": 3.785795963226646e-08, + "loss": 1.7555, + "step": 13175 + }, + { + "epoch": 0.8881102388733533, + "grad_norm": 5.9097879315428, + "learning_rate": 3.7633817904458574e-08, + "loss": 1.8154, + "step": 13180 + }, + { + "epoch": 0.8884471547454601, + "grad_norm": 6.222570248163253, + "learning_rate": 3.741031572435615e-08, + "loss": 1.7622, + "step": 13185 + }, + { + "epoch": 0.8887840706175668, + "grad_norm": 5.99694342465949, + "learning_rate": 3.718745340110868e-08, + "loss": 1.7687, + "step": 13190 + }, + { + "epoch": 0.8891209864896735, + "grad_norm": 6.024623734589238, + "learning_rate": 3.6965231242980624e-08, + "loss": 1.7684, + "step": 13195 + }, + { + "epoch": 0.8894579023617802, + "grad_norm": 6.0660869612201544, + "learning_rate": 3.6743649557351265e-08, + "loss": 1.8178, + "step": 13200 + }, + { + "epoch": 0.889794818233887, + "grad_norm": 5.584094991029224, + "learning_rate": 3.652270865071344e-08, + "loss": 1.7802, + "step": 13205 + }, + { + "epoch": 0.8901317341059938, + "grad_norm": 5.19943663551813, + "learning_rate": 3.630240882867408e-08, + "loss": 1.714, + "step": 13210 + }, + { + "epoch": 0.8904686499781005, + "grad_norm": 5.960605902380365, + "learning_rate": 3.608275039595332e-08, + "loss": 1.8067, + "step": 13215 + }, + { + "epoch": 0.8908055658502072, + "grad_norm": 6.364002982000837, + "learning_rate": 3.5863733656383844e-08, + "loss": 1.7506, + "step": 13220 + }, + { + "epoch": 0.8911424817223139, + "grad_norm": 5.831937518355378, + "learning_rate": 3.564535891291115e-08, + "loss": 1.7529, + "step": 13225 + }, + { + "epoch": 0.8914793975944206, + "grad_norm": 5.90539818137521, + "learning_rate": 3.542762646759234e-08, + "loss": 1.7931, + "step": 13230 + }, + { + "epoch": 0.8918163134665275, + "grad_norm": 6.244337889693646, + "learning_rate": 3.521053662159629e-08, + "loss": 1.7838, + "step": 13235 + }, + { + "epoch": 0.8921532293386342, + "grad_norm": 5.636325790259478, + "learning_rate": 3.499408967520295e-08, + "loss": 1.7723, + "step": 13240 + }, + { + "epoch": 0.8924901452107409, + "grad_norm": 5.909428853004949, + "learning_rate": 3.477828592780319e-08, + "loss": 1.7905, + "step": 13245 + }, + { + "epoch": 0.8928270610828476, + "grad_norm": 5.743071012947205, + "learning_rate": 3.456312567789793e-08, + "loss": 1.8337, + "step": 13250 + }, + { + "epoch": 0.8931639769549543, + "grad_norm": 5.470554904588339, + "learning_rate": 3.4348609223098125e-08, + "loss": 1.7911, + "step": 13255 + }, + { + "epoch": 0.8935008928270611, + "grad_norm": 5.665375115697222, + "learning_rate": 3.41347368601243e-08, + "loss": 1.7307, + "step": 13260 + }, + { + "epoch": 0.8938378086991678, + "grad_norm": 6.481341411904355, + "learning_rate": 3.39215088848061e-08, + "loss": 1.8396, + "step": 13265 + }, + { + "epoch": 0.8941747245712746, + "grad_norm": 6.225439326082594, + "learning_rate": 3.370892559208155e-08, + "loss": 1.7568, + "step": 13270 + }, + { + "epoch": 0.8945116404433813, + "grad_norm": 5.906481797833057, + "learning_rate": 3.34969872759972e-08, + "loss": 1.8076, + "step": 13275 + }, + { + "epoch": 0.894848556315488, + "grad_norm": 5.856494710175204, + "learning_rate": 3.328569422970762e-08, + "loss": 1.8648, + "step": 13280 + }, + { + "epoch": 0.8951854721875948, + "grad_norm": 5.753105226825638, + "learning_rate": 3.307504674547429e-08, + "loss": 1.7815, + "step": 13285 + }, + { + "epoch": 0.8955223880597015, + "grad_norm": 6.095595815839566, + "learning_rate": 3.286504511466631e-08, + "loss": 1.8328, + "step": 13290 + }, + { + "epoch": 0.8958593039318082, + "grad_norm": 6.288610606607138, + "learning_rate": 3.265568962775927e-08, + "loss": 1.7913, + "step": 13295 + }, + { + "epoch": 0.896196219803915, + "grad_norm": 6.5022942580461125, + "learning_rate": 3.2446980574334706e-08, + "loss": 1.7869, + "step": 13300 + }, + { + "epoch": 0.8965331356760217, + "grad_norm": 6.6487512915161275, + "learning_rate": 3.2238918243080505e-08, + "loss": 1.7442, + "step": 13305 + }, + { + "epoch": 0.8968700515481284, + "grad_norm": 6.732808022858596, + "learning_rate": 3.203150292178952e-08, + "loss": 1.8084, + "step": 13310 + }, + { + "epoch": 0.8972069674202352, + "grad_norm": 5.897197143896169, + "learning_rate": 3.182473489736004e-08, + "loss": 1.7462, + "step": 13315 + }, + { + "epoch": 0.8975438832923419, + "grad_norm": 6.842935750718153, + "learning_rate": 3.161861445579478e-08, + "loss": 1.8202, + "step": 13320 + }, + { + "epoch": 0.8978807991644486, + "grad_norm": 6.089128345755971, + "learning_rate": 3.1413141882200736e-08, + "loss": 1.7682, + "step": 13325 + }, + { + "epoch": 0.8982177150365553, + "grad_norm": 5.780557396715805, + "learning_rate": 3.120831746078895e-08, + "loss": 1.7929, + "step": 13330 + }, + { + "epoch": 0.898554630908662, + "grad_norm": 5.928980468404957, + "learning_rate": 3.100414147487368e-08, + "loss": 1.8481, + "step": 13335 + }, + { + "epoch": 0.8988915467807689, + "grad_norm": 6.0861515824792844, + "learning_rate": 3.0800614206872413e-08, + "loss": 1.7645, + "step": 13340 + }, + { + "epoch": 0.8992284626528756, + "grad_norm": 6.467715458691118, + "learning_rate": 3.059773593830539e-08, + "loss": 1.7795, + "step": 13345 + }, + { + "epoch": 0.8995653785249823, + "grad_norm": 5.804373053320775, + "learning_rate": 3.039550694979492e-08, + "loss": 1.7391, + "step": 13350 + }, + { + "epoch": 0.899902294397089, + "grad_norm": 5.747917639408956, + "learning_rate": 3.019392752106548e-08, + "loss": 1.7619, + "step": 13355 + }, + { + "epoch": 0.9002392102691957, + "grad_norm": 5.824464537511226, + "learning_rate": 2.9992997930942954e-08, + "loss": 1.7164, + "step": 13360 + }, + { + "epoch": 0.9005761261413026, + "grad_norm": 6.003811614386364, + "learning_rate": 2.979271845735426e-08, + "loss": 1.6998, + "step": 13365 + }, + { + "epoch": 0.9009130420134093, + "grad_norm": 5.8609792524761914, + "learning_rate": 2.9593089377327242e-08, + "loss": 1.7337, + "step": 13370 + }, + { + "epoch": 0.901249957885516, + "grad_norm": 5.948332344827435, + "learning_rate": 2.9394110966990184e-08, + "loss": 1.7973, + "step": 13375 + }, + { + "epoch": 0.9015868737576227, + "grad_norm": 5.945282741761599, + "learning_rate": 2.9195783501570982e-08, + "loss": 1.8228, + "step": 13380 + }, + { + "epoch": 0.9019237896297294, + "grad_norm": 6.262748243809702, + "learning_rate": 2.8998107255397643e-08, + "loss": 1.7952, + "step": 13385 + }, + { + "epoch": 0.9022607055018362, + "grad_norm": 5.640388167355125, + "learning_rate": 2.880108250189689e-08, + "loss": 1.7446, + "step": 13390 + }, + { + "epoch": 0.902597621373943, + "grad_norm": 5.800344956552423, + "learning_rate": 2.860470951359478e-08, + "loss": 1.7547, + "step": 13395 + }, + { + "epoch": 0.9029345372460497, + "grad_norm": 6.291107803853359, + "learning_rate": 2.8408988562115488e-08, + "loss": 1.7873, + "step": 13400 + }, + { + "epoch": 0.9032714531181564, + "grad_norm": 6.022821073549721, + "learning_rate": 2.8213919918181393e-08, + "loss": 1.8376, + "step": 13405 + }, + { + "epoch": 0.9036083689902631, + "grad_norm": 5.900130211549773, + "learning_rate": 2.8019503851612837e-08, + "loss": 1.7953, + "step": 13410 + }, + { + "epoch": 0.9039452848623699, + "grad_norm": 5.901590236995989, + "learning_rate": 2.782574063132703e-08, + "loss": 1.8029, + "step": 13415 + }, + { + "epoch": 0.9042822007344766, + "grad_norm": 5.913123808511502, + "learning_rate": 2.7632630525338597e-08, + "loss": 1.7911, + "step": 13420 + }, + { + "epoch": 0.9046191166065833, + "grad_norm": 6.486077531057897, + "learning_rate": 2.7440173800758583e-08, + "loss": 1.8126, + "step": 13425 + }, + { + "epoch": 0.90495603247869, + "grad_norm": 5.924773230014183, + "learning_rate": 2.7248370723794268e-08, + "loss": 1.7774, + "step": 13430 + }, + { + "epoch": 0.9052929483507968, + "grad_norm": 6.107738959168259, + "learning_rate": 2.7057221559748822e-08, + "loss": 1.7895, + "step": 13435 + }, + { + "epoch": 0.9056298642229036, + "grad_norm": 5.945232419314283, + "learning_rate": 2.6866726573021025e-08, + "loss": 1.7537, + "step": 13440 + }, + { + "epoch": 0.9059667800950103, + "grad_norm": 6.242713673098733, + "learning_rate": 2.667688602710455e-08, + "loss": 1.709, + "step": 13445 + }, + { + "epoch": 0.906303695967117, + "grad_norm": 5.810035940437741, + "learning_rate": 2.648770018458807e-08, + "loss": 1.7855, + "step": 13450 + }, + { + "epoch": 0.9066406118392237, + "grad_norm": 5.876560332140629, + "learning_rate": 2.6299169307154535e-08, + "loss": 1.7288, + "step": 13455 + }, + { + "epoch": 0.9069775277113304, + "grad_norm": 6.026710876222, + "learning_rate": 2.611129365558118e-08, + "loss": 1.7491, + "step": 13460 + }, + { + "epoch": 0.9073144435834373, + "grad_norm": 5.90332142357701, + "learning_rate": 2.592407348973852e-08, + "loss": 1.7657, + "step": 13465 + }, + { + "epoch": 0.907651359455544, + "grad_norm": 5.567054827768613, + "learning_rate": 2.573750906859079e-08, + "loss": 1.7703, + "step": 13470 + }, + { + "epoch": 0.9079882753276507, + "grad_norm": 6.275015331234111, + "learning_rate": 2.5551600650194906e-08, + "loss": 1.7642, + "step": 13475 + }, + { + "epoch": 0.9083251911997574, + "grad_norm": 5.80583328955002, + "learning_rate": 2.536634849170055e-08, + "loss": 1.7379, + "step": 13480 + }, + { + "epoch": 0.9086621070718641, + "grad_norm": 6.1557870962000045, + "learning_rate": 2.5181752849349593e-08, + "loss": 1.8367, + "step": 13485 + }, + { + "epoch": 0.9089990229439708, + "grad_norm": 6.037242378415563, + "learning_rate": 2.4997813978476e-08, + "loss": 1.7544, + "step": 13490 + }, + { + "epoch": 0.9093359388160777, + "grad_norm": 5.672781989201412, + "learning_rate": 2.481453213350493e-08, + "loss": 1.7149, + "step": 13495 + }, + { + "epoch": 0.9096728546881844, + "grad_norm": 5.80097338376941, + "learning_rate": 2.463190756795308e-08, + "loss": 1.8256, + "step": 13500 + }, + { + "epoch": 0.9100097705602911, + "grad_norm": 5.341807531960438, + "learning_rate": 2.4449940534427836e-08, + "loss": 1.7936, + "step": 13505 + }, + { + "epoch": 0.9103466864323978, + "grad_norm": 5.820785045693387, + "learning_rate": 2.4268631284627027e-08, + "loss": 1.7686, + "step": 13510 + }, + { + "epoch": 0.9106836023045045, + "grad_norm": 5.816002660909503, + "learning_rate": 2.408798006933882e-08, + "loss": 1.7303, + "step": 13515 + }, + { + "epoch": 0.9110205181766113, + "grad_norm": 6.275964406450363, + "learning_rate": 2.3907987138440945e-08, + "loss": 1.8489, + "step": 13520 + }, + { + "epoch": 0.911357434048718, + "grad_norm": 6.08682021587138, + "learning_rate": 2.3728652740900856e-08, + "loss": 1.8246, + "step": 13525 + }, + { + "epoch": 0.9116943499208248, + "grad_norm": 6.302297868580049, + "learning_rate": 2.3549977124774857e-08, + "loss": 1.8028, + "step": 13530 + }, + { + "epoch": 0.9120312657929315, + "grad_norm": 6.206494648513327, + "learning_rate": 2.337196053720819e-08, + "loss": 1.8341, + "step": 13535 + }, + { + "epoch": 0.9123681816650382, + "grad_norm": 5.733867734946892, + "learning_rate": 2.319460322443456e-08, + "loss": 1.8267, + "step": 13540 + }, + { + "epoch": 0.912705097537145, + "grad_norm": 6.233397928155692, + "learning_rate": 2.301790543177551e-08, + "loss": 1.8284, + "step": 13545 + }, + { + "epoch": 0.9130420134092517, + "grad_norm": 5.902961201033665, + "learning_rate": 2.284186740364069e-08, + "loss": 1.8105, + "step": 13550 + }, + { + "epoch": 0.9133789292813584, + "grad_norm": 5.83122816008699, + "learning_rate": 2.266648938352672e-08, + "loss": 1.765, + "step": 13555 + }, + { + "epoch": 0.9137158451534652, + "grad_norm": 5.692387788783081, + "learning_rate": 2.249177161401783e-08, + "loss": 1.7427, + "step": 13560 + }, + { + "epoch": 0.9140527610255719, + "grad_norm": 6.083052684924437, + "learning_rate": 2.2317714336784422e-08, + "loss": 1.78, + "step": 13565 + }, + { + "epoch": 0.9143896768976787, + "grad_norm": 6.01194897181711, + "learning_rate": 2.21443177925838e-08, + "loss": 1.7904, + "step": 13570 + }, + { + "epoch": 0.9147265927697854, + "grad_norm": 6.2538897195592735, + "learning_rate": 2.1971582221258944e-08, + "loss": 1.6636, + "step": 13575 + }, + { + "epoch": 0.9150635086418921, + "grad_norm": 5.338562271328569, + "learning_rate": 2.1799507861738788e-08, + "loss": 1.7815, + "step": 13580 + }, + { + "epoch": 0.9154004245139988, + "grad_norm": 6.02697364861854, + "learning_rate": 2.1628094952037713e-08, + "loss": 1.7755, + "step": 13585 + }, + { + "epoch": 0.9157373403861055, + "grad_norm": 5.6174748611517025, + "learning_rate": 2.1457343729255062e-08, + "loss": 1.7606, + "step": 13590 + }, + { + "epoch": 0.9160742562582124, + "grad_norm": 5.953965768567391, + "learning_rate": 2.128725442957491e-08, + "loss": 1.7886, + "step": 13595 + }, + { + "epoch": 0.9164111721303191, + "grad_norm": 5.733695202042228, + "learning_rate": 2.111782728826583e-08, + "loss": 1.7728, + "step": 13600 + }, + { + "epoch": 0.9167480880024258, + "grad_norm": 6.073387061708105, + "learning_rate": 2.0949062539680486e-08, + "loss": 1.8331, + "step": 13605 + }, + { + "epoch": 0.9170850038745325, + "grad_norm": 5.96599535385796, + "learning_rate": 2.07809604172553e-08, + "loss": 1.7551, + "step": 13610 + }, + { + "epoch": 0.9174219197466392, + "grad_norm": 6.058916705027718, + "learning_rate": 2.0613521153510115e-08, + "loss": 1.7635, + "step": 13615 + }, + { + "epoch": 0.917758835618746, + "grad_norm": 6.206503726449759, + "learning_rate": 2.0446744980048002e-08, + "loss": 1.777, + "step": 13620 + }, + { + "epoch": 0.9180957514908528, + "grad_norm": 6.121730585820273, + "learning_rate": 2.0280632127554708e-08, + "loss": 1.8304, + "step": 13625 + }, + { + "epoch": 0.9184326673629595, + "grad_norm": 5.909262400028377, + "learning_rate": 2.011518282579855e-08, + "loss": 1.7962, + "step": 13630 + }, + { + "epoch": 0.9187695832350662, + "grad_norm": 5.951133427267503, + "learning_rate": 1.9950397303630075e-08, + "loss": 1.8755, + "step": 13635 + }, + { + "epoch": 0.9191064991071729, + "grad_norm": 5.749682108699682, + "learning_rate": 1.9786275788981565e-08, + "loss": 1.7573, + "step": 13640 + }, + { + "epoch": 0.9194434149792797, + "grad_norm": 5.932217476759413, + "learning_rate": 1.9622818508866823e-08, + "loss": 1.7832, + "step": 13645 + }, + { + "epoch": 0.9197803308513864, + "grad_norm": 6.105403050825645, + "learning_rate": 1.9460025689381043e-08, + "loss": 1.8365, + "step": 13650 + }, + { + "epoch": 0.9201172467234932, + "grad_norm": 5.352964036275184, + "learning_rate": 1.9297897555700216e-08, + "loss": 1.7852, + "step": 13655 + }, + { + "epoch": 0.9204541625955999, + "grad_norm": 5.644291930764029, + "learning_rate": 1.9136434332080898e-08, + "loss": 1.7718, + "step": 13660 + }, + { + "epoch": 0.9207910784677066, + "grad_norm": 5.807485679937865, + "learning_rate": 1.8975636241860048e-08, + "loss": 1.7835, + "step": 13665 + }, + { + "epoch": 0.9211279943398133, + "grad_norm": 5.9129764742967765, + "learning_rate": 1.8815503507454644e-08, + "loss": 1.7171, + "step": 13670 + }, + { + "epoch": 0.9214649102119201, + "grad_norm": 5.975444115877748, + "learning_rate": 1.8656036350361117e-08, + "loss": 1.751, + "step": 13675 + }, + { + "epoch": 0.9218018260840268, + "grad_norm": 6.144902579922595, + "learning_rate": 1.8497234991155463e-08, + "loss": 1.7783, + "step": 13680 + }, + { + "epoch": 0.9221387419561335, + "grad_norm": 5.666472776654721, + "learning_rate": 1.8339099649492762e-08, + "loss": 1.7597, + "step": 13685 + }, + { + "epoch": 0.9224756578282403, + "grad_norm": 5.9809396720752, + "learning_rate": 1.8181630544106653e-08, + "loss": 1.7526, + "step": 13690 + }, + { + "epoch": 0.922812573700347, + "grad_norm": 6.83292618583234, + "learning_rate": 1.8024827892809346e-08, + "loss": 1.8244, + "step": 13695 + }, + { + "epoch": 0.9231494895724538, + "grad_norm": 5.86478171310939, + "learning_rate": 1.7868691912491352e-08, + "loss": 1.7762, + "step": 13700 + }, + { + "epoch": 0.9234864054445605, + "grad_norm": 6.790313884006856, + "learning_rate": 1.77132228191208e-08, + "loss": 1.8513, + "step": 13705 + }, + { + "epoch": 0.9238233213166672, + "grad_norm": 5.9937155458488345, + "learning_rate": 1.7558420827743505e-08, + "loss": 1.7219, + "step": 13710 + }, + { + "epoch": 0.9241602371887739, + "grad_norm": 5.797828927156594, + "learning_rate": 1.7404286152482573e-08, + "loss": 1.8354, + "step": 13715 + }, + { + "epoch": 0.9244971530608806, + "grad_norm": 6.102906661047269, + "learning_rate": 1.725081900653791e-08, + "loss": 1.7955, + "step": 13720 + }, + { + "epoch": 0.9248340689329875, + "grad_norm": 6.010033747492002, + "learning_rate": 1.7098019602186376e-08, + "loss": 1.7936, + "step": 13725 + }, + { + "epoch": 0.9251709848050942, + "grad_norm": 5.893664038194936, + "learning_rate": 1.6945888150780797e-08, + "loss": 1.7985, + "step": 13730 + }, + { + "epoch": 0.9255079006772009, + "grad_norm": 5.822012915972412, + "learning_rate": 1.6794424862750568e-08, + "loss": 1.7945, + "step": 13735 + }, + { + "epoch": 0.9258448165493076, + "grad_norm": 6.3394192953564, + "learning_rate": 1.664362994760038e-08, + "loss": 1.7824, + "step": 13740 + }, + { + "epoch": 0.9261817324214143, + "grad_norm": 6.035176725514034, + "learning_rate": 1.649350361391083e-08, + "loss": 1.7793, + "step": 13745 + }, + { + "epoch": 0.9265186482935212, + "grad_norm": 6.654659837949139, + "learning_rate": 1.6344046069337646e-08, + "loss": 1.7988, + "step": 13750 + }, + { + "epoch": 0.9268555641656279, + "grad_norm": 5.585511061286529, + "learning_rate": 1.6195257520611182e-08, + "loss": 1.8061, + "step": 13755 + }, + { + "epoch": 0.9271924800377346, + "grad_norm": 5.778820047963369, + "learning_rate": 1.604713817353681e-08, + "loss": 1.7819, + "step": 13760 + }, + { + "epoch": 0.9275293959098413, + "grad_norm": 6.078945827605927, + "learning_rate": 1.5899688232994147e-08, + "loss": 1.7825, + "step": 13765 + }, + { + "epoch": 0.927866311781948, + "grad_norm": 5.951451870166644, + "learning_rate": 1.5752907902936707e-08, + "loss": 1.839, + "step": 13770 + }, + { + "epoch": 0.9282032276540548, + "grad_norm": 5.916252930427277, + "learning_rate": 1.560679738639198e-08, + "loss": 1.7983, + "step": 13775 + }, + { + "epoch": 0.9285401435261615, + "grad_norm": 5.771537516382258, + "learning_rate": 1.5461356885461075e-08, + "loss": 1.8014, + "step": 13780 + }, + { + "epoch": 0.9288770593982683, + "grad_norm": 6.107394671320099, + "learning_rate": 1.5316586601317905e-08, + "loss": 1.7635, + "step": 13785 + }, + { + "epoch": 0.929213975270375, + "grad_norm": 5.785638235806199, + "learning_rate": 1.5172486734209788e-08, + "loss": 1.7735, + "step": 13790 + }, + { + "epoch": 0.9295508911424817, + "grad_norm": 5.730850538387305, + "learning_rate": 1.502905748345651e-08, + "loss": 1.8355, + "step": 13795 + }, + { + "epoch": 0.9298878070145885, + "grad_norm": 5.875893406145162, + "learning_rate": 1.4886299047450257e-08, + "loss": 1.8188, + "step": 13800 + }, + { + "epoch": 0.9302247228866952, + "grad_norm": 5.813097498313215, + "learning_rate": 1.4744211623655356e-08, + "loss": 1.7418, + "step": 13805 + }, + { + "epoch": 0.9305616387588019, + "grad_norm": 5.935994794147698, + "learning_rate": 1.4602795408607982e-08, + "loss": 1.852, + "step": 13810 + }, + { + "epoch": 0.9308985546309086, + "grad_norm": 5.807523266406351, + "learning_rate": 1.4462050597915942e-08, + "loss": 1.7984, + "step": 13815 + }, + { + "epoch": 0.9312354705030154, + "grad_norm": 5.730729459958924, + "learning_rate": 1.4321977386258289e-08, + "loss": 1.7641, + "step": 13820 + }, + { + "epoch": 0.9315723863751222, + "grad_norm": 5.986047538097552, + "learning_rate": 1.4182575967385092e-08, + "loss": 1.791, + "step": 13825 + }, + { + "epoch": 0.9319093022472289, + "grad_norm": 6.129565707883838, + "learning_rate": 1.4043846534117331e-08, + "loss": 1.8389, + "step": 13830 + }, + { + "epoch": 0.9322462181193356, + "grad_norm": 5.760689255089029, + "learning_rate": 1.3905789278346347e-08, + "loss": 1.8328, + "step": 13835 + }, + { + "epoch": 0.9325831339914423, + "grad_norm": 6.0668336057579255, + "learning_rate": 1.3768404391033717e-08, + "loss": 1.735, + "step": 13840 + }, + { + "epoch": 0.932920049863549, + "grad_norm": 5.885732912533342, + "learning_rate": 1.3631692062211209e-08, + "loss": 1.7884, + "step": 13845 + }, + { + "epoch": 0.9332569657356558, + "grad_norm": 5.388443134954862, + "learning_rate": 1.3495652480979947e-08, + "loss": 1.7078, + "step": 13850 + }, + { + "epoch": 0.9335938816077626, + "grad_norm": 6.045883802907099, + "learning_rate": 1.3360285835510854e-08, + "loss": 1.7818, + "step": 13855 + }, + { + "epoch": 0.9339307974798693, + "grad_norm": 5.517887293622092, + "learning_rate": 1.322559231304382e-08, + "loss": 1.7703, + "step": 13860 + }, + { + "epoch": 0.934267713351976, + "grad_norm": 5.541469046961143, + "learning_rate": 1.3091572099887816e-08, + "loss": 1.7328, + "step": 13865 + }, + { + "epoch": 0.9346046292240827, + "grad_norm": 6.56871395739785, + "learning_rate": 1.2958225381420329e-08, + "loss": 1.8307, + "step": 13870 + }, + { + "epoch": 0.9349415450961894, + "grad_norm": 5.667419893283266, + "learning_rate": 1.282555234208732e-08, + "loss": 1.7326, + "step": 13875 + }, + { + "epoch": 0.9352784609682963, + "grad_norm": 5.838203900467372, + "learning_rate": 1.2693553165403104e-08, + "loss": 1.8029, + "step": 13880 + }, + { + "epoch": 0.935615376840403, + "grad_norm": 5.711248775025883, + "learning_rate": 1.2562228033949628e-08, + "loss": 1.7949, + "step": 13885 + }, + { + "epoch": 0.9359522927125097, + "grad_norm": 5.676609731870054, + "learning_rate": 1.243157712937659e-08, + "loss": 1.7805, + "step": 13890 + }, + { + "epoch": 0.9362892085846164, + "grad_norm": 5.8302983836456725, + "learning_rate": 1.230160063240121e-08, + "loss": 1.7759, + "step": 13895 + }, + { + "epoch": 0.9366261244567231, + "grad_norm": 6.030113265479034, + "learning_rate": 1.2172298722807617e-08, + "loss": 1.7496, + "step": 13900 + }, + { + "epoch": 0.9369630403288299, + "grad_norm": 6.57072889368483, + "learning_rate": 1.204367157944708e-08, + "loss": 1.7831, + "step": 13905 + }, + { + "epoch": 0.9372999562009366, + "grad_norm": 6.038296639335411, + "learning_rate": 1.19157193802375e-08, + "loss": 1.747, + "step": 13910 + }, + { + "epoch": 0.9376368720730434, + "grad_norm": 5.814924414623662, + "learning_rate": 1.1788442302163026e-08, + "loss": 1.7092, + "step": 13915 + }, + { + "epoch": 0.9379737879451501, + "grad_norm": 5.262932683835106, + "learning_rate": 1.1661840521274168e-08, + "loss": 1.7298, + "step": 13920 + }, + { + "epoch": 0.9383107038172568, + "grad_norm": 5.546576533364832, + "learning_rate": 1.1535914212687237e-08, + "loss": 1.7256, + "step": 13925 + }, + { + "epoch": 0.9386476196893636, + "grad_norm": 6.17537281360325, + "learning_rate": 1.1410663550584287e-08, + "loss": 1.8097, + "step": 13930 + }, + { + "epoch": 0.9389845355614703, + "grad_norm": 5.773326142297128, + "learning_rate": 1.1286088708212793e-08, + "loss": 1.7578, + "step": 13935 + }, + { + "epoch": 0.939321451433577, + "grad_norm": 5.893824855852281, + "learning_rate": 1.1162189857885362e-08, + "loss": 1.6737, + "step": 13940 + }, + { + "epoch": 0.9396583673056838, + "grad_norm": 6.293068011289075, + "learning_rate": 1.1038967170979741e-08, + "loss": 1.8323, + "step": 13945 + }, + { + "epoch": 0.9399952831777905, + "grad_norm": 6.3178382632779675, + "learning_rate": 1.0916420817938254e-08, + "loss": 1.7639, + "step": 13950 + }, + { + "epoch": 0.9403321990498973, + "grad_norm": 6.366970292404799, + "learning_rate": 1.0794550968267701e-08, + "loss": 1.7629, + "step": 13955 + }, + { + "epoch": 0.940669114922004, + "grad_norm": 5.228237120022797, + "learning_rate": 1.0673357790539294e-08, + "loss": 1.7724, + "step": 13960 + }, + { + "epoch": 0.9410060307941107, + "grad_norm": 5.925892786800767, + "learning_rate": 1.0552841452388105e-08, + "loss": 1.7607, + "step": 13965 + }, + { + "epoch": 0.9413429466662174, + "grad_norm": 6.2915894761092614, + "learning_rate": 1.0433002120513123e-08, + "loss": 1.8291, + "step": 13970 + }, + { + "epoch": 0.9416798625383241, + "grad_norm": 5.612014170575601, + "learning_rate": 1.0313839960676751e-08, + "loss": 1.7242, + "step": 13975 + }, + { + "epoch": 0.942016778410431, + "grad_norm": 5.38062147344516, + "learning_rate": 1.019535513770492e-08, + "loss": 1.7061, + "step": 13980 + }, + { + "epoch": 0.9423536942825377, + "grad_norm": 6.023856824550423, + "learning_rate": 1.0077547815486476e-08, + "loss": 1.8322, + "step": 13985 + }, + { + "epoch": 0.9426906101546444, + "grad_norm": 5.912398539147373, + "learning_rate": 9.960418156973238e-09, + "loss": 1.8146, + "step": 13990 + }, + { + "epoch": 0.9430275260267511, + "grad_norm": 6.1113631057581586, + "learning_rate": 9.843966324179609e-09, + "loss": 1.7683, + "step": 13995 + }, + { + "epoch": 0.9433644418988578, + "grad_norm": 5.615958243889113, + "learning_rate": 9.728192478182573e-09, + "loss": 1.7788, + "step": 14000 + }, + { + "epoch": 0.9437013577709646, + "grad_norm": 5.854145953223302, + "learning_rate": 9.613096779121089e-09, + "loss": 1.7846, + "step": 14005 + }, + { + "epoch": 0.9440382736430714, + "grad_norm": 6.076063145121326, + "learning_rate": 9.498679386196417e-09, + "loss": 1.6587, + "step": 14010 + }, + { + "epoch": 0.9443751895151781, + "grad_norm": 6.399734349425361, + "learning_rate": 9.384940457671186e-09, + "loss": 1.7681, + "step": 14015 + }, + { + "epoch": 0.9447121053872848, + "grad_norm": 6.338424938973493, + "learning_rate": 9.271880150869882e-09, + "loss": 1.829, + "step": 14020 + }, + { + "epoch": 0.9450490212593915, + "grad_norm": 5.985103027928022, + "learning_rate": 9.15949862217824e-09, + "loss": 1.7963, + "step": 14025 + }, + { + "epoch": 0.9453859371314982, + "grad_norm": 6.383395928238325, + "learning_rate": 9.04779602704292e-09, + "loss": 1.8399, + "step": 14030 + }, + { + "epoch": 0.945722853003605, + "grad_norm": 5.7326489327199655, + "learning_rate": 8.936772519971769e-09, + "loss": 1.7752, + "step": 14035 + }, + { + "epoch": 0.9460597688757117, + "grad_norm": 5.816190646797327, + "learning_rate": 8.826428254533169e-09, + "loss": 1.8245, + "step": 14040 + }, + { + "epoch": 0.9463966847478185, + "grad_norm": 5.440371053867433, + "learning_rate": 8.716763383355862e-09, + "loss": 1.7861, + "step": 14045 + }, + { + "epoch": 0.9467336006199252, + "grad_norm": 6.137934997478103, + "learning_rate": 8.607778058129122e-09, + "loss": 1.8045, + "step": 14050 + }, + { + "epoch": 0.9470705164920319, + "grad_norm": 5.8353013118261945, + "learning_rate": 8.499472429601972e-09, + "loss": 1.7325, + "step": 14055 + }, + { + "epoch": 0.9474074323641387, + "grad_norm": 5.964783791471046, + "learning_rate": 8.391846647583468e-09, + "loss": 1.8082, + "step": 14060 + }, + { + "epoch": 0.9477443482362454, + "grad_norm": 6.710260912678535, + "learning_rate": 8.284900860942246e-09, + "loss": 1.8287, + "step": 14065 + }, + { + "epoch": 0.9480812641083521, + "grad_norm": 6.257280332373422, + "learning_rate": 8.178635217606367e-09, + "loss": 1.7582, + "step": 14070 + }, + { + "epoch": 0.9484181799804589, + "grad_norm": 5.687185998282569, + "learning_rate": 8.073049864563142e-09, + "loss": 1.7209, + "step": 14075 + }, + { + "epoch": 0.9487550958525656, + "grad_norm": 5.285333365404428, + "learning_rate": 7.968144947858801e-09, + "loss": 1.7414, + "step": 14080 + }, + { + "epoch": 0.9490920117246724, + "grad_norm": 6.183760283797594, + "learning_rate": 7.863920612598496e-09, + "loss": 1.789, + "step": 14085 + }, + { + "epoch": 0.9494289275967791, + "grad_norm": 5.718203089285701, + "learning_rate": 7.760377002945961e-09, + "loss": 1.7936, + "step": 14090 + }, + { + "epoch": 0.9497658434688858, + "grad_norm": 5.577317551701951, + "learning_rate": 7.657514262123354e-09, + "loss": 1.7411, + "step": 14095 + }, + { + "epoch": 0.9501027593409925, + "grad_norm": 7.000453905423672, + "learning_rate": 7.55533253241103e-09, + "loss": 1.7849, + "step": 14100 + }, + { + "epoch": 0.9504396752130992, + "grad_norm": 5.744188185276807, + "learning_rate": 7.453831955147428e-09, + "loss": 1.7766, + "step": 14105 + }, + { + "epoch": 0.9507765910852061, + "grad_norm": 5.311037237991293, + "learning_rate": 7.353012670728631e-09, + "loss": 1.7438, + "step": 14110 + }, + { + "epoch": 0.9511135069573128, + "grad_norm": 5.705160326224774, + "learning_rate": 7.252874818608645e-09, + "loss": 1.7423, + "step": 14115 + }, + { + "epoch": 0.9514504228294195, + "grad_norm": 5.941266511069281, + "learning_rate": 7.153418537298617e-09, + "loss": 1.8092, + "step": 14120 + }, + { + "epoch": 0.9517873387015262, + "grad_norm": 6.028018288023328, + "learning_rate": 7.0546439643671685e-09, + "loss": 1.7414, + "step": 14125 + }, + { + "epoch": 0.9521242545736329, + "grad_norm": 6.0984894459924615, + "learning_rate": 6.9565512364398445e-09, + "loss": 1.799, + "step": 14130 + }, + { + "epoch": 0.9524611704457397, + "grad_norm": 5.810943364613546, + "learning_rate": 6.859140489199167e-09, + "loss": 1.7548, + "step": 14135 + }, + { + "epoch": 0.9527980863178465, + "grad_norm": 6.05019358569247, + "learning_rate": 6.762411857384187e-09, + "loss": 1.8126, + "step": 14140 + }, + { + "epoch": 0.9531350021899532, + "grad_norm": 6.012558005711744, + "learning_rate": 6.666365474790492e-09, + "loss": 1.7978, + "step": 14145 + }, + { + "epoch": 0.9534719180620599, + "grad_norm": 6.316428272394365, + "learning_rate": 6.571001474270144e-09, + "loss": 1.7144, + "step": 14150 + }, + { + "epoch": 0.9538088339341666, + "grad_norm": 5.825366175142781, + "learning_rate": 6.4763199877311825e-09, + "loss": 1.7862, + "step": 14155 + }, + { + "epoch": 0.9541457498062734, + "grad_norm": 5.678194546571653, + "learning_rate": 6.382321146137571e-09, + "loss": 1.848, + "step": 14160 + }, + { + "epoch": 0.9544826656783801, + "grad_norm": 6.0270794934572285, + "learning_rate": 6.28900507950908e-09, + "loss": 1.8092, + "step": 14165 + }, + { + "epoch": 0.9548195815504869, + "grad_norm": 6.029126967150501, + "learning_rate": 6.196371916921073e-09, + "loss": 1.7877, + "step": 14170 + }, + { + "epoch": 0.9551564974225936, + "grad_norm": 6.231081190815106, + "learning_rate": 6.1044217865043325e-09, + "loss": 1.7842, + "step": 14175 + }, + { + "epoch": 0.9554934132947003, + "grad_norm": 5.951369955420382, + "learning_rate": 6.013154815444732e-09, + "loss": 1.8343, + "step": 14180 + }, + { + "epoch": 0.9558303291668071, + "grad_norm": 5.698513479571637, + "learning_rate": 5.922571129983456e-09, + "loss": 1.8377, + "step": 14185 + }, + { + "epoch": 0.9561672450389138, + "grad_norm": 6.336560120824533, + "learning_rate": 5.832670855416277e-09, + "loss": 1.8043, + "step": 14190 + }, + { + "epoch": 0.9565041609110205, + "grad_norm": 5.567693621557861, + "learning_rate": 5.7434541160938375e-09, + "loss": 1.7857, + "step": 14195 + }, + { + "epoch": 0.9568410767831272, + "grad_norm": 5.602719462601326, + "learning_rate": 5.6549210354212565e-09, + "loss": 1.8139, + "step": 14200 + }, + { + "epoch": 0.957177992655234, + "grad_norm": 6.252668788603584, + "learning_rate": 5.567071735858131e-09, + "loss": 1.7325, + "step": 14205 + }, + { + "epoch": 0.9575149085273407, + "grad_norm": 6.186365782591999, + "learning_rate": 5.4799063389179834e-09, + "loss": 1.7848, + "step": 14210 + }, + { + "epoch": 0.9578518243994475, + "grad_norm": 6.0170910772312345, + "learning_rate": 5.393424965168702e-09, + "loss": 1.8098, + "step": 14215 + }, + { + "epoch": 0.9581887402715542, + "grad_norm": 6.416217500285821, + "learning_rate": 5.307627734231657e-09, + "loss": 1.8432, + "step": 14220 + }, + { + "epoch": 0.9585256561436609, + "grad_norm": 6.079095377891874, + "learning_rate": 5.222514764782193e-09, + "loss": 1.8535, + "step": 14225 + }, + { + "epoch": 0.9588625720157676, + "grad_norm": 5.691306381084669, + "learning_rate": 5.138086174549083e-09, + "loss": 1.8069, + "step": 14230 + }, + { + "epoch": 0.9591994878878743, + "grad_norm": 6.315757537930107, + "learning_rate": 5.054342080314522e-09, + "loss": 1.7752, + "step": 14235 + }, + { + "epoch": 0.9595364037599812, + "grad_norm": 5.6811246579307255, + "learning_rate": 4.97128259791374e-09, + "loss": 1.7077, + "step": 14240 + }, + { + "epoch": 0.9598733196320879, + "grad_norm": 5.875555272618595, + "learning_rate": 4.888907842235113e-09, + "loss": 1.7176, + "step": 14245 + }, + { + "epoch": 0.9602102355041946, + "grad_norm": 5.940205015447577, + "learning_rate": 4.807217927220053e-09, + "loss": 1.7786, + "step": 14250 + }, + { + "epoch": 0.9605471513763013, + "grad_norm": 6.595986946123306, + "learning_rate": 4.726212965862342e-09, + "loss": 1.7806, + "step": 14255 + }, + { + "epoch": 0.960884067248408, + "grad_norm": 6.100938914389518, + "learning_rate": 4.645893070208684e-09, + "loss": 1.7441, + "step": 14260 + }, + { + "epoch": 0.9612209831205148, + "grad_norm": 6.225212166717535, + "learning_rate": 4.566258351357988e-09, + "loss": 1.7242, + "step": 14265 + }, + { + "epoch": 0.9615578989926216, + "grad_norm": 6.038611752498317, + "learning_rate": 4.48730891946153e-09, + "loss": 1.7141, + "step": 14270 + }, + { + "epoch": 0.9618948148647283, + "grad_norm": 5.687997722577268, + "learning_rate": 4.409044883722568e-09, + "loss": 1.7426, + "step": 14275 + }, + { + "epoch": 0.962231730736835, + "grad_norm": 5.65260149981692, + "learning_rate": 4.331466352396396e-09, + "loss": 1.8174, + "step": 14280 + }, + { + "epoch": 0.9625686466089417, + "grad_norm": 5.542702072962062, + "learning_rate": 4.2545734327902315e-09, + "loss": 1.7413, + "step": 14285 + }, + { + "epoch": 0.9629055624810485, + "grad_norm": 6.429780119232116, + "learning_rate": 4.178366231262665e-09, + "loss": 1.7694, + "step": 14290 + }, + { + "epoch": 0.9632424783531552, + "grad_norm": 5.910097048238468, + "learning_rate": 4.102844853224041e-09, + "loss": 1.7907, + "step": 14295 + }, + { + "epoch": 0.963579394225262, + "grad_norm": 6.045190253764579, + "learning_rate": 4.028009403135968e-09, + "loss": 1.774, + "step": 14300 + }, + { + "epoch": 0.9639163100973687, + "grad_norm": 6.098928200355735, + "learning_rate": 3.95385998451131e-09, + "loss": 1.8191, + "step": 14305 + }, + { + "epoch": 0.9642532259694754, + "grad_norm": 5.455149987735236, + "learning_rate": 3.880396699913968e-09, + "loss": 1.7874, + "step": 14310 + }, + { + "epoch": 0.9645901418415822, + "grad_norm": 6.359343013532195, + "learning_rate": 3.807619650958827e-09, + "loss": 1.7665, + "step": 14315 + }, + { + "epoch": 0.9649270577136889, + "grad_norm": 6.159889529402087, + "learning_rate": 3.7355289383115276e-09, + "loss": 1.8139, + "step": 14320 + }, + { + "epoch": 0.9652639735857956, + "grad_norm": 5.9449316060723785, + "learning_rate": 3.664124661688417e-09, + "loss": 1.8311, + "step": 14325 + }, + { + "epoch": 0.9656008894579023, + "grad_norm": 6.470751557779755, + "learning_rate": 3.5934069198562677e-09, + "loss": 1.7104, + "step": 14330 + }, + { + "epoch": 0.9659378053300091, + "grad_norm": 5.800731201565026, + "learning_rate": 3.5233758106322787e-09, + "loss": 1.7851, + "step": 14335 + }, + { + "epoch": 0.9662747212021159, + "grad_norm": 6.006742828646833, + "learning_rate": 3.4540314308839635e-09, + "loss": 1.7698, + "step": 14340 + }, + { + "epoch": 0.9666116370742226, + "grad_norm": 5.905075320021472, + "learning_rate": 3.385373876528874e-09, + "loss": 1.8125, + "step": 14345 + }, + { + "epoch": 0.9669485529463293, + "grad_norm": 6.577938888492687, + "learning_rate": 3.3174032425345444e-09, + "loss": 1.7513, + "step": 14350 + }, + { + "epoch": 0.967285468818436, + "grad_norm": 5.903574805231385, + "learning_rate": 3.250119622918379e-09, + "loss": 1.7226, + "step": 14355 + }, + { + "epoch": 0.9676223846905427, + "grad_norm": 6.078005184292544, + "learning_rate": 3.1835231107474323e-09, + "loss": 1.7905, + "step": 14360 + }, + { + "epoch": 0.9679593005626496, + "grad_norm": 6.042836567494886, + "learning_rate": 3.1176137981385185e-09, + "loss": 1.7973, + "step": 14365 + }, + { + "epoch": 0.9682962164347563, + "grad_norm": 5.65602428557149, + "learning_rate": 3.0523917762576568e-09, + "loss": 1.7457, + "step": 14370 + }, + { + "epoch": 0.968633132306863, + "grad_norm": 5.773086987424218, + "learning_rate": 2.9878571353204595e-09, + "loss": 1.8666, + "step": 14375 + }, + { + "epoch": 0.9689700481789697, + "grad_norm": 6.0373431050475315, + "learning_rate": 2.924009964591578e-09, + "loss": 1.7397, + "step": 14380 + }, + { + "epoch": 0.9693069640510764, + "grad_norm": 6.094144908034533, + "learning_rate": 2.8608503523848136e-09, + "loss": 1.8231, + "step": 14385 + }, + { + "epoch": 0.9696438799231831, + "grad_norm": 6.33990389312829, + "learning_rate": 2.7983783860629496e-09, + "loss": 1.699, + "step": 14390 + }, + { + "epoch": 0.96998079579529, + "grad_norm": 5.7053703145794605, + "learning_rate": 2.7365941520375303e-09, + "loss": 1.7927, + "step": 14395 + }, + { + "epoch": 0.9703177116673967, + "grad_norm": 5.636532734952386, + "learning_rate": 2.6754977357689724e-09, + "loss": 1.7631, + "step": 14400 + }, + { + "epoch": 0.9706546275395034, + "grad_norm": 5.892415294536311, + "learning_rate": 2.6150892217660647e-09, + "loss": 1.7085, + "step": 14405 + }, + { + "epoch": 0.9709915434116101, + "grad_norm": 6.6153894309409615, + "learning_rate": 2.5553686935864126e-09, + "loss": 1.8487, + "step": 14410 + }, + { + "epoch": 0.9713284592837168, + "grad_norm": 6.00592621079739, + "learning_rate": 2.496336233835661e-09, + "loss": 1.7502, + "step": 14415 + }, + { + "epoch": 0.9716653751558236, + "grad_norm": 5.883380002615908, + "learning_rate": 2.437991924167937e-09, + "loss": 1.834, + "step": 14420 + }, + { + "epoch": 0.9720022910279303, + "grad_norm": 5.457301773132191, + "learning_rate": 2.380335845285464e-09, + "loss": 1.7772, + "step": 14425 + }, + { + "epoch": 0.9723392069000371, + "grad_norm": 6.009693677479863, + "learning_rate": 2.323368076938448e-09, + "loss": 1.7901, + "step": 14430 + }, + { + "epoch": 0.9726761227721438, + "grad_norm": 5.926199674258243, + "learning_rate": 2.2670886979250235e-09, + "loss": 1.7362, + "step": 14435 + }, + { + "epoch": 0.9730130386442505, + "grad_norm": 6.6228262552651955, + "learning_rate": 2.211497786091143e-09, + "loss": 1.7798, + "step": 14440 + }, + { + "epoch": 0.9733499545163573, + "grad_norm": 5.525416957164009, + "learning_rate": 2.1565954183306313e-09, + "loss": 1.7603, + "step": 14445 + }, + { + "epoch": 0.973686870388464, + "grad_norm": 6.128785233954746, + "learning_rate": 2.1023816705846853e-09, + "loss": 1.7899, + "step": 14450 + }, + { + "epoch": 0.9740237862605707, + "grad_norm": 5.597950039158548, + "learning_rate": 2.048856617842043e-09, + "loss": 1.6975, + "step": 14455 + }, + { + "epoch": 0.9743607021326774, + "grad_norm": 6.07135782851224, + "learning_rate": 1.9960203341389813e-09, + "loss": 1.7247, + "step": 14460 + }, + { + "epoch": 0.9746976180047842, + "grad_norm": 5.877722240882944, + "learning_rate": 1.943872892558929e-09, + "loss": 1.8057, + "step": 14465 + }, + { + "epoch": 0.975034533876891, + "grad_norm": 5.852579274221657, + "learning_rate": 1.8924143652325196e-09, + "loss": 1.7745, + "step": 14470 + }, + { + "epoch": 0.9753714497489977, + "grad_norm": 6.176901784449129, + "learning_rate": 1.8416448233374848e-09, + "loss": 1.7142, + "step": 14475 + }, + { + "epoch": 0.9757083656211044, + "grad_norm": 6.208387127082091, + "learning_rate": 1.79156433709865e-09, + "loss": 1.8073, + "step": 14480 + }, + { + "epoch": 0.9760452814932111, + "grad_norm": 5.5727389870896715, + "learning_rate": 1.742172975787548e-09, + "loss": 1.748, + "step": 14485 + }, + { + "epoch": 0.9763821973653178, + "grad_norm": 5.9251930429951445, + "learning_rate": 1.6934708077226411e-09, + "loss": 1.7577, + "step": 14490 + }, + { + "epoch": 0.9767191132374247, + "grad_norm": 5.954705180523381, + "learning_rate": 1.6454579002690982e-09, + "loss": 1.7851, + "step": 14495 + }, + { + "epoch": 0.9770560291095314, + "grad_norm": 6.19174715813602, + "learning_rate": 1.5981343198386288e-09, + "loss": 1.7829, + "step": 14500 + }, + { + "epoch": 0.9773929449816381, + "grad_norm": 5.893212756172858, + "learning_rate": 1.5515001318895382e-09, + "loss": 1.7271, + "step": 14505 + }, + { + "epoch": 0.9777298608537448, + "grad_norm": 5.568240553890549, + "learning_rate": 1.5055554009264505e-09, + "loss": 1.8086, + "step": 14510 + }, + { + "epoch": 0.9780667767258515, + "grad_norm": 5.900260642183186, + "learning_rate": 1.4603001905004187e-09, + "loss": 1.7413, + "step": 14515 + }, + { + "epoch": 0.9784036925979583, + "grad_norm": 5.940174718298427, + "learning_rate": 1.4157345632087592e-09, + "loss": 1.6663, + "step": 14520 + }, + { + "epoch": 0.978740608470065, + "grad_norm": 5.696923193549704, + "learning_rate": 1.37185858069494e-09, + "loss": 1.7419, + "step": 14525 + }, + { + "epoch": 0.9790775243421718, + "grad_norm": 5.732780541053396, + "learning_rate": 1.328672303648415e-09, + "loss": 1.7824, + "step": 14530 + }, + { + "epoch": 0.9794144402142785, + "grad_norm": 6.147023994098516, + "learning_rate": 1.2861757918046778e-09, + "loss": 1.7452, + "step": 14535 + }, + { + "epoch": 0.9797513560863852, + "grad_norm": 6.103750948785661, + "learning_rate": 1.2443691039452642e-09, + "loss": 1.7967, + "step": 14540 + }, + { + "epoch": 0.980088271958492, + "grad_norm": 6.290929925062352, + "learning_rate": 1.203252297897417e-09, + "loss": 1.7872, + "step": 14545 + }, + { + "epoch": 0.9804251878305987, + "grad_norm": 5.631296340413579, + "learning_rate": 1.1628254305340869e-09, + "loss": 1.7577, + "step": 14550 + }, + { + "epoch": 0.9807621037027054, + "grad_norm": 6.151781649826895, + "learning_rate": 1.123088557773988e-09, + "loss": 1.7795, + "step": 14555 + }, + { + "epoch": 0.9810990195748122, + "grad_norm": 5.878601681301532, + "learning_rate": 1.0840417345814312e-09, + "loss": 1.7934, + "step": 14560 + }, + { + "epoch": 0.9814359354469189, + "grad_norm": 5.661532017486243, + "learning_rate": 1.0456850149662134e-09, + "loss": 1.7786, + "step": 14565 + }, + { + "epoch": 0.9817728513190256, + "grad_norm": 5.498959828209034, + "learning_rate": 1.0080184519835056e-09, + "loss": 1.7411, + "step": 14570 + }, + { + "epoch": 0.9821097671911324, + "grad_norm": 5.9308419780727695, + "learning_rate": 9.71042097734076e-10, + "loss": 1.6922, + "step": 14575 + }, + { + "epoch": 0.9824466830632391, + "grad_norm": 6.224672314052866, + "learning_rate": 9.347560033637347e-10, + "loss": 1.7872, + "step": 14580 + }, + { + "epoch": 0.9827835989353458, + "grad_norm": 6.456012846690479, + "learning_rate": 8.991602190636105e-10, + "loss": 1.7946, + "step": 14585 + }, + { + "epoch": 0.9831205148074526, + "grad_norm": 5.885385615000342, + "learning_rate": 8.642547940700961e-10, + "loss": 1.7394, + "step": 14590 + }, + { + "epoch": 0.9834574306795593, + "grad_norm": 6.523936838219438, + "learning_rate": 8.300397766644595e-10, + "loss": 1.7704, + "step": 14595 + }, + { + "epoch": 0.9837943465516661, + "grad_norm": 5.787459818959744, + "learning_rate": 7.965152141732878e-10, + "loss": 1.7182, + "step": 14600 + }, + { + "epoch": 0.9841312624237728, + "grad_norm": 6.66665537166297, + "learning_rate": 7.636811529678211e-10, + "loss": 1.9044, + "step": 14605 + }, + { + "epoch": 0.9844681782958795, + "grad_norm": 6.253212410487395, + "learning_rate": 7.315376384643968e-10, + "loss": 1.7786, + "step": 14610 + }, + { + "epoch": 0.9848050941679862, + "grad_norm": 6.212902508637188, + "learning_rate": 7.000847151240608e-10, + "loss": 1.8143, + "step": 14615 + }, + { + "epoch": 0.9851420100400929, + "grad_norm": 5.743679508987371, + "learning_rate": 6.693224264527897e-10, + "loss": 1.7831, + "step": 14620 + }, + { + "epoch": 0.9854789259121998, + "grad_norm": 5.626780106553643, + "learning_rate": 6.392508150011023e-10, + "loss": 1.7868, + "step": 14625 + }, + { + "epoch": 0.9858158417843065, + "grad_norm": 5.914322606080147, + "learning_rate": 6.098699223641701e-10, + "loss": 1.7559, + "step": 14630 + }, + { + "epoch": 0.9861527576564132, + "grad_norm": 5.280356384405962, + "learning_rate": 5.811797891819847e-10, + "loss": 1.7467, + "step": 14635 + }, + { + "epoch": 0.9864896735285199, + "grad_norm": 6.098016592220372, + "learning_rate": 5.531804551387464e-10, + "loss": 1.7997, + "step": 14640 + }, + { + "epoch": 0.9868265894006266, + "grad_norm": 5.913509913806897, + "learning_rate": 5.258719589634198e-10, + "loss": 1.6823, + "step": 14645 + }, + { + "epoch": 0.9871635052727334, + "grad_norm": 6.600818579713024, + "learning_rate": 4.992543384291781e-10, + "loss": 1.8086, + "step": 14650 + }, + { + "epoch": 0.9875004211448402, + "grad_norm": 6.463420633584194, + "learning_rate": 4.733276303537925e-10, + "loss": 1.8205, + "step": 14655 + }, + { + "epoch": 0.9878373370169469, + "grad_norm": 5.856138268133559, + "learning_rate": 4.480918705991321e-10, + "loss": 1.7879, + "step": 14660 + }, + { + "epoch": 0.9881742528890536, + "grad_norm": 6.322129837266916, + "learning_rate": 4.235470940715524e-10, + "loss": 1.8006, + "step": 14665 + }, + { + "epoch": 0.9885111687611603, + "grad_norm": 6.141656233222583, + "learning_rate": 3.99693334721507e-10, + "loss": 1.7482, + "step": 14670 + }, + { + "epoch": 0.9888480846332671, + "grad_norm": 6.526943362952002, + "learning_rate": 3.765306255436029e-10, + "loss": 1.8075, + "step": 14675 + }, + { + "epoch": 0.9891850005053738, + "grad_norm": 5.793751726786318, + "learning_rate": 3.540589985766562e-10, + "loss": 1.7946, + "step": 14680 + }, + { + "epoch": 0.9895219163774805, + "grad_norm": 5.973903896846691, + "learning_rate": 3.322784849036364e-10, + "loss": 1.7779, + "step": 14685 + }, + { + "epoch": 0.9898588322495873, + "grad_norm": 6.014872978513893, + "learning_rate": 3.11189114651389e-10, + "loss": 1.6545, + "step": 14690 + }, + { + "epoch": 0.990195748121694, + "grad_norm": 6.368229704231421, + "learning_rate": 2.9079091699091287e-10, + "loss": 1.7579, + "step": 14695 + }, + { + "epoch": 0.9905326639938008, + "grad_norm": 6.193988622544747, + "learning_rate": 2.710839201370829e-10, + "loss": 1.8368, + "step": 14700 + }, + { + "epoch": 0.9908695798659075, + "grad_norm": 5.758847930865848, + "learning_rate": 2.5206815134881655e-10, + "loss": 1.7821, + "step": 14705 + }, + { + "epoch": 0.9912064957380142, + "grad_norm": 5.69684814939484, + "learning_rate": 2.337436369287404e-10, + "loss": 1.7896, + "step": 14710 + }, + { + "epoch": 0.9915434116101209, + "grad_norm": 5.638836518174001, + "learning_rate": 2.1611040222346833e-10, + "loss": 1.7734, + "step": 14715 + }, + { + "epoch": 0.9918803274822277, + "grad_norm": 5.899552088739335, + "learning_rate": 1.9916847162343432e-10, + "loss": 1.7912, + "step": 14720 + }, + { + "epoch": 0.9922172433543345, + "grad_norm": 5.829892034986946, + "learning_rate": 1.829178685627264e-10, + "loss": 1.7899, + "step": 14725 + }, + { + "epoch": 0.9925541592264412, + "grad_norm": 5.795901536554887, + "learning_rate": 1.6735861551936402e-10, + "loss": 1.728, + "step": 14730 + }, + { + "epoch": 0.9928910750985479, + "grad_norm": 5.766864718819964, + "learning_rate": 1.5249073401502055e-10, + "loss": 1.7406, + "step": 14735 + }, + { + "epoch": 0.9932279909706546, + "grad_norm": 6.08955650931636, + "learning_rate": 1.3831424461496766e-10, + "loss": 1.7832, + "step": 14740 + }, + { + "epoch": 0.9935649068427613, + "grad_norm": 6.069229823009566, + "learning_rate": 1.2482916692824197e-10, + "loss": 1.7347, + "step": 14745 + }, + { + "epoch": 0.993901822714868, + "grad_norm": 5.865594094745515, + "learning_rate": 1.1203551960742297e-10, + "loss": 1.7604, + "step": 14750 + }, + { + "epoch": 0.9942387385869749, + "grad_norm": 6.41217202434423, + "learning_rate": 9.993332034891056e-11, + "loss": 1.7317, + "step": 14755 + }, + { + "epoch": 0.9945756544590816, + "grad_norm": 5.818435701385782, + "learning_rate": 8.852258589236994e-11, + "loss": 1.7556, + "step": 14760 + }, + { + "epoch": 0.9949125703311883, + "grad_norm": 6.258726795030299, + "learning_rate": 7.780333202134226e-11, + "loss": 1.7599, + "step": 14765 + }, + { + "epoch": 0.995249486203295, + "grad_norm": 5.7871168282439545, + "learning_rate": 6.777557356263397e-11, + "loss": 1.7978, + "step": 14770 + }, + { + "epoch": 0.9955864020754017, + "grad_norm": 5.995386180636011, + "learning_rate": 5.843932438681643e-11, + "loss": 1.8203, + "step": 14775 + }, + { + "epoch": 0.9959233179475085, + "grad_norm": 5.821396562583734, + "learning_rate": 4.97945974077818e-11, + "loss": 1.8173, + "step": 14780 + }, + { + "epoch": 0.9962602338196153, + "grad_norm": 6.117647963442301, + "learning_rate": 4.1841404582965143e-11, + "loss": 1.7567, + "step": 14785 + }, + { + "epoch": 0.996597149691722, + "grad_norm": 5.655528276296895, + "learning_rate": 3.457975691334436e-11, + "loss": 1.681, + "step": 14790 + }, + { + "epoch": 0.9969340655638287, + "grad_norm": 5.646705806265958, + "learning_rate": 2.800966444316266e-11, + "loss": 1.7546, + "step": 14795 + }, + { + "epoch": 0.9972709814359354, + "grad_norm": 5.905296090122313, + "learning_rate": 2.213113626026164e-11, + "loss": 1.8007, + "step": 14800 + }, + { + "epoch": 0.9976078973080422, + "grad_norm": 5.922158957749158, + "learning_rate": 1.6944180495914728e-11, + "loss": 1.704, + "step": 14805 + }, + { + "epoch": 0.9979448131801489, + "grad_norm": 6.204962127147208, + "learning_rate": 1.2448804324660667e-11, + "loss": 1.8127, + "step": 14810 + }, + { + "epoch": 0.9982817290522557, + "grad_norm": 6.224102373975581, + "learning_rate": 8.645013964581061e-12, + "loss": 1.8241, + "step": 14815 + }, + { + "epoch": 0.9986186449243624, + "grad_norm": 5.952709226062345, + "learning_rate": 5.532814677133846e-12, + "loss": 1.7962, + "step": 14820 + }, + { + "epoch": 0.9989555607964691, + "grad_norm": 6.246353867410941, + "learning_rate": 3.112210767042267e-12, + "loss": 1.804, + "step": 14825 + }, + { + "epoch": 0.9992924766685759, + "grad_norm": 6.163743778465804, + "learning_rate": 1.383205582516922e-12, + "loss": 1.6971, + "step": 14830 + }, + { + "epoch": 0.9996293925406826, + "grad_norm": 5.689291557459708, + "learning_rate": 3.4580151520025024e-13, + "loss": 1.8645, + "step": 14835 + }, + { + "epoch": 0.9999663084127893, + "grad_norm": 6.244802182387895, + "learning_rate": 0.0, + "loss": 1.8053, + "step": 14840 + }, + { + "epoch": 0.9999663084127893, + "eval_loss": NaN, + "eval_runtime": 77.4202, + "eval_samples_per_second": 61.973, + "eval_steps_per_second": 7.75, + "step": 14840 + }, + { + "epoch": 0.9999663084127893, + "step": 14840, + "total_flos": 1.2604318649008783e+18, + "train_loss": 1.874135019435394, + "train_runtime": 22896.7966, + "train_samples_per_second": 20.741, + "train_steps_per_second": 0.648 + } + ], + "logging_steps": 5, + "max_steps": 14840, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 1.2604318649008783e+18, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}