{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 6870, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.002183406113537118, "grad_norm": 15.530065061351076, "learning_rate": 2.9112081513828244e-07, "loss": 3.5144, "step": 5 }, { "epoch": 0.004366812227074236, "grad_norm": 14.048089091944986, "learning_rate": 6.550218340611353e-07, "loss": 3.3991, "step": 10 }, { "epoch": 0.006550218340611353, "grad_norm": 10.38275753075565, "learning_rate": 1.0189228529839884e-06, "loss": 3.4492, "step": 15 }, { "epoch": 0.008733624454148471, "grad_norm": 9.374594765056237, "learning_rate": 1.3828238719068415e-06, "loss": 3.3321, "step": 20 }, { "epoch": 0.010917030567685589, "grad_norm": 6.58125566549559, "learning_rate": 1.7467248908296944e-06, "loss": 3.0597, "step": 25 }, { "epoch": 0.013100436681222707, "grad_norm": 7.80412598343367, "learning_rate": 2.1106259097525473e-06, "loss": 2.9508, "step": 30 }, { "epoch": 0.015283842794759825, "grad_norm": 7.524557299709741, "learning_rate": 2.4745269286754004e-06, "loss": 2.612, "step": 35 }, { "epoch": 0.017467248908296942, "grad_norm": 6.2330781402437365, "learning_rate": 2.8384279475982535e-06, "loss": 2.3435, "step": 40 }, { "epoch": 0.019650655021834062, "grad_norm": 5.3971059460100825, "learning_rate": 3.2023289665211066e-06, "loss": 1.9812, "step": 45 }, { "epoch": 0.021834061135371178, "grad_norm": 5.173910417939837, "learning_rate": 3.5662299854439593e-06, "loss": 1.6557, "step": 50 }, { "epoch": 0.024017467248908297, "grad_norm": 4.343854825431366, "learning_rate": 3.930131004366813e-06, "loss": 1.4334, "step": 55 }, { "epoch": 0.026200873362445413, "grad_norm": 3.318343002698737, "learning_rate": 4.2940320232896655e-06, "loss": 1.2603, "step": 60 }, { "epoch": 0.028384279475982533, "grad_norm": 3.3649234822213425, "learning_rate": 4.657933042212519e-06, "loss": 1.128, "step": 65 }, { "epoch": 0.03056768558951965, "grad_norm": 4.651497105373944, "learning_rate": 5.021834061135372e-06, "loss": 1.0685, "step": 70 }, { "epoch": 0.03275109170305677, "grad_norm": 2.1122195307286127, "learning_rate": 5.385735080058224e-06, "loss": 0.9865, "step": 75 }, { "epoch": 0.034934497816593885, "grad_norm": 3.264835592266701, "learning_rate": 5.749636098981077e-06, "loss": 0.8563, "step": 80 }, { "epoch": 0.03711790393013101, "grad_norm": 4.854625880384626, "learning_rate": 6.11353711790393e-06, "loss": 0.813, "step": 85 }, { "epoch": 0.039301310043668124, "grad_norm": 2.2913931555435223, "learning_rate": 6.477438136826783e-06, "loss": 0.8189, "step": 90 }, { "epoch": 0.04148471615720524, "grad_norm": 1.4161513553023397, "learning_rate": 6.841339155749636e-06, "loss": 0.7341, "step": 95 }, { "epoch": 0.043668122270742356, "grad_norm": 1.2517988600799463, "learning_rate": 7.205240174672489e-06, "loss": 0.7264, "step": 100 }, { "epoch": 0.04585152838427948, "grad_norm": 4.403128247316299, "learning_rate": 7.569141193595343e-06, "loss": 0.7105, "step": 105 }, { "epoch": 0.048034934497816595, "grad_norm": 2.6065584518468996, "learning_rate": 7.933042212518195e-06, "loss": 0.6883, "step": 110 }, { "epoch": 0.05021834061135371, "grad_norm": 1.2548695388960658, "learning_rate": 8.296943231441049e-06, "loss": 0.6444, "step": 115 }, { "epoch": 0.05240174672489083, "grad_norm": 1.2420608105158033, "learning_rate": 8.660844250363902e-06, "loss": 0.6228, "step": 120 }, { "epoch": 0.05458515283842795, "grad_norm": 1.6213174255897973, "learning_rate": 9.024745269286754e-06, "loss": 0.6091, "step": 125 }, { "epoch": 0.056768558951965066, "grad_norm": 1.1139361754005903, "learning_rate": 9.388646288209607e-06, "loss": 0.6049, "step": 130 }, { "epoch": 0.05895196506550218, "grad_norm": 2.019221687744237, "learning_rate": 9.75254730713246e-06, "loss": 0.6334, "step": 135 }, { "epoch": 0.0611353711790393, "grad_norm": 1.4614530692241856, "learning_rate": 1.0116448326055314e-05, "loss": 0.5484, "step": 140 }, { "epoch": 0.06331877729257641, "grad_norm": 1.7546701255165822, "learning_rate": 1.0480349344978165e-05, "loss": 0.5777, "step": 145 }, { "epoch": 0.06550218340611354, "grad_norm": 1.5028564001758502, "learning_rate": 1.084425036390102e-05, "loss": 0.5427, "step": 150 }, { "epoch": 0.06768558951965066, "grad_norm": 0.8737189391813837, "learning_rate": 1.1208151382823872e-05, "loss": 0.5501, "step": 155 }, { "epoch": 0.06986899563318777, "grad_norm": 0.7913219748536747, "learning_rate": 1.1572052401746727e-05, "loss": 0.5404, "step": 160 }, { "epoch": 0.07205240174672489, "grad_norm": 1.2931917942966242, "learning_rate": 1.1935953420669578e-05, "loss": 0.5317, "step": 165 }, { "epoch": 0.07423580786026202, "grad_norm": 1.7815089914348687, "learning_rate": 1.2299854439592432e-05, "loss": 0.5391, "step": 170 }, { "epoch": 0.07641921397379912, "grad_norm": 1.588354964609756, "learning_rate": 1.2663755458515283e-05, "loss": 0.5163, "step": 175 }, { "epoch": 0.07860262008733625, "grad_norm": 0.7324777658190085, "learning_rate": 1.3027656477438139e-05, "loss": 0.5206, "step": 180 }, { "epoch": 0.08078602620087336, "grad_norm": 0.9320526641008475, "learning_rate": 1.339155749636099e-05, "loss": 0.51, "step": 185 }, { "epoch": 0.08296943231441048, "grad_norm": 1.0473996131197105, "learning_rate": 1.3755458515283843e-05, "loss": 0.4995, "step": 190 }, { "epoch": 0.0851528384279476, "grad_norm": 1.1005598187154526, "learning_rate": 1.4119359534206697e-05, "loss": 0.4992, "step": 195 }, { "epoch": 0.08733624454148471, "grad_norm": 1.3835272450195761, "learning_rate": 1.448326055312955e-05, "loss": 0.5157, "step": 200 }, { "epoch": 0.08951965065502183, "grad_norm": 0.8363321954594061, "learning_rate": 1.4847161572052404e-05, "loss": 0.4884, "step": 205 }, { "epoch": 0.09170305676855896, "grad_norm": 0.6751320611923934, "learning_rate": 1.5211062590975255e-05, "loss": 0.4837, "step": 210 }, { "epoch": 0.09388646288209607, "grad_norm": 0.6597124878200359, "learning_rate": 1.557496360989811e-05, "loss": 0.4614, "step": 215 }, { "epoch": 0.09606986899563319, "grad_norm": 1.0002616519579557, "learning_rate": 1.5938864628820962e-05, "loss": 0.5178, "step": 220 }, { "epoch": 0.0982532751091703, "grad_norm": 2.094018394615991, "learning_rate": 1.6302765647743815e-05, "loss": 0.4785, "step": 225 }, { "epoch": 0.10043668122270742, "grad_norm": 0.8762721338958381, "learning_rate": 1.6666666666666667e-05, "loss": 0.4867, "step": 230 }, { "epoch": 0.10262008733624454, "grad_norm": 1.3388349700796345, "learning_rate": 1.703056768558952e-05, "loss": 0.4781, "step": 235 }, { "epoch": 0.10480349344978165, "grad_norm": 1.0507336135636423, "learning_rate": 1.7394468704512373e-05, "loss": 0.4732, "step": 240 }, { "epoch": 0.10698689956331878, "grad_norm": 0.7938940927456775, "learning_rate": 1.7758369723435225e-05, "loss": 0.4651, "step": 245 }, { "epoch": 0.1091703056768559, "grad_norm": 1.0437122295269432, "learning_rate": 1.812227074235808e-05, "loss": 0.4585, "step": 250 }, { "epoch": 0.11135371179039301, "grad_norm": 0.67350165285708, "learning_rate": 1.848617176128093e-05, "loss": 0.4609, "step": 255 }, { "epoch": 0.11353711790393013, "grad_norm": 0.5808251314496068, "learning_rate": 1.8850072780203784e-05, "loss": 0.4549, "step": 260 }, { "epoch": 0.11572052401746726, "grad_norm": 1.8396131242666172, "learning_rate": 1.921397379912664e-05, "loss": 0.4783, "step": 265 }, { "epoch": 0.11790393013100436, "grad_norm": 0.9680443797016299, "learning_rate": 1.9577874818049492e-05, "loss": 0.455, "step": 270 }, { "epoch": 0.12008733624454149, "grad_norm": 0.7014587496843943, "learning_rate": 1.9941775836972345e-05, "loss": 0.4479, "step": 275 }, { "epoch": 0.1222707423580786, "grad_norm": 0.9218263081064745, "learning_rate": 2.0305676855895198e-05, "loss": 0.456, "step": 280 }, { "epoch": 0.12445414847161572, "grad_norm": 0.5840696636129238, "learning_rate": 2.066957787481805e-05, "loss": 0.4318, "step": 285 }, { "epoch": 0.12663755458515283, "grad_norm": 0.6901911501394837, "learning_rate": 2.1033478893740903e-05, "loss": 0.4494, "step": 290 }, { "epoch": 0.12882096069868995, "grad_norm": 0.5256416173639158, "learning_rate": 2.1397379912663756e-05, "loss": 0.4604, "step": 295 }, { "epoch": 0.13100436681222707, "grad_norm": 0.8643348227959773, "learning_rate": 2.176128093158661e-05, "loss": 0.4529, "step": 300 }, { "epoch": 0.1331877729257642, "grad_norm": 0.845191908182343, "learning_rate": 2.2125181950509464e-05, "loss": 0.4462, "step": 305 }, { "epoch": 0.13537117903930132, "grad_norm": 0.6911667544775351, "learning_rate": 2.2489082969432314e-05, "loss": 0.4353, "step": 310 }, { "epoch": 0.13755458515283842, "grad_norm": 0.6842730240845932, "learning_rate": 2.2852983988355166e-05, "loss": 0.4425, "step": 315 }, { "epoch": 0.13973799126637554, "grad_norm": 0.6095243486936803, "learning_rate": 2.3216885007278022e-05, "loss": 0.4375, "step": 320 }, { "epoch": 0.14192139737991266, "grad_norm": 0.9176516550209551, "learning_rate": 2.3580786026200875e-05, "loss": 0.4539, "step": 325 }, { "epoch": 0.14410480349344978, "grad_norm": 0.8430335294570548, "learning_rate": 2.3944687045123728e-05, "loss": 0.4347, "step": 330 }, { "epoch": 0.1462882096069869, "grad_norm": 0.7088826431924614, "learning_rate": 2.430858806404658e-05, "loss": 0.4149, "step": 335 }, { "epoch": 0.14847161572052403, "grad_norm": 0.6180099894880557, "learning_rate": 2.4672489082969433e-05, "loss": 0.422, "step": 340 }, { "epoch": 0.15065502183406113, "grad_norm": 0.6817744275567881, "learning_rate": 2.503639010189229e-05, "loss": 0.4418, "step": 345 }, { "epoch": 0.15283842794759825, "grad_norm": 0.6940745586404697, "learning_rate": 2.540029112081514e-05, "loss": 0.4404, "step": 350 }, { "epoch": 0.15502183406113537, "grad_norm": 0.5754528111989647, "learning_rate": 2.576419213973799e-05, "loss": 0.4263, "step": 355 }, { "epoch": 0.1572052401746725, "grad_norm": 1.0591998260395188, "learning_rate": 2.6128093158660844e-05, "loss": 0.4299, "step": 360 }, { "epoch": 0.15938864628820962, "grad_norm": 0.7123628186533127, "learning_rate": 2.64919941775837e-05, "loss": 0.4196, "step": 365 }, { "epoch": 0.1615720524017467, "grad_norm": 0.6457647967870623, "learning_rate": 2.6855895196506553e-05, "loss": 0.4174, "step": 370 }, { "epoch": 0.16375545851528384, "grad_norm": 0.9935438036526899, "learning_rate": 2.7219796215429405e-05, "loss": 0.3993, "step": 375 }, { "epoch": 0.16593886462882096, "grad_norm": 0.883212399644436, "learning_rate": 2.7583697234352258e-05, "loss": 0.4345, "step": 380 }, { "epoch": 0.16812227074235808, "grad_norm": 0.948205577828456, "learning_rate": 2.794759825327511e-05, "loss": 0.4441, "step": 385 }, { "epoch": 0.1703056768558952, "grad_norm": 0.8223542412825284, "learning_rate": 2.831149927219796e-05, "loss": 0.4029, "step": 390 }, { "epoch": 0.17248908296943233, "grad_norm": 0.7182654852607071, "learning_rate": 2.867540029112082e-05, "loss": 0.4342, "step": 395 }, { "epoch": 0.17467248908296942, "grad_norm": 0.6862331015738732, "learning_rate": 2.9039301310043672e-05, "loss": 0.4044, "step": 400 }, { "epoch": 0.17685589519650655, "grad_norm": 0.9719904263631477, "learning_rate": 2.940320232896652e-05, "loss": 0.4114, "step": 405 }, { "epoch": 0.17903930131004367, "grad_norm": 0.8428735007204966, "learning_rate": 2.9767103347889374e-05, "loss": 0.4124, "step": 410 }, { "epoch": 0.1812227074235808, "grad_norm": 0.8900932432816406, "learning_rate": 3.0131004366812227e-05, "loss": 0.4096, "step": 415 }, { "epoch": 0.18340611353711792, "grad_norm": 0.658732643364676, "learning_rate": 3.0494905385735083e-05, "loss": 0.4019, "step": 420 }, { "epoch": 0.185589519650655, "grad_norm": 0.5717380212732919, "learning_rate": 3.0858806404657935e-05, "loss": 0.4125, "step": 425 }, { "epoch": 0.18777292576419213, "grad_norm": 0.7561360040926198, "learning_rate": 3.122270742358079e-05, "loss": 0.3979, "step": 430 }, { "epoch": 0.18995633187772926, "grad_norm": 0.7154647354779691, "learning_rate": 3.158660844250364e-05, "loss": 0.411, "step": 435 }, { "epoch": 0.19213973799126638, "grad_norm": 1.106580393251376, "learning_rate": 3.1950509461426493e-05, "loss": 0.4001, "step": 440 }, { "epoch": 0.1943231441048035, "grad_norm": 0.8277158276869705, "learning_rate": 3.2314410480349346e-05, "loss": 0.3982, "step": 445 }, { "epoch": 0.1965065502183406, "grad_norm": 0.6695298962156193, "learning_rate": 3.26783114992722e-05, "loss": 0.4129, "step": 450 }, { "epoch": 0.19868995633187772, "grad_norm": 0.8926693328948101, "learning_rate": 3.304221251819505e-05, "loss": 0.4173, "step": 455 }, { "epoch": 0.20087336244541484, "grad_norm": 0.9134165249644083, "learning_rate": 3.3406113537117904e-05, "loss": 0.3947, "step": 460 }, { "epoch": 0.20305676855895197, "grad_norm": 0.6780077341796262, "learning_rate": 3.377001455604076e-05, "loss": 0.4066, "step": 465 }, { "epoch": 0.2052401746724891, "grad_norm": 0.8359819073368681, "learning_rate": 3.413391557496361e-05, "loss": 0.4054, "step": 470 }, { "epoch": 0.2074235807860262, "grad_norm": 0.49801067958580947, "learning_rate": 3.449781659388647e-05, "loss": 0.3923, "step": 475 }, { "epoch": 0.2096069868995633, "grad_norm": 0.7389933814419388, "learning_rate": 3.486171761280932e-05, "loss": 0.4068, "step": 480 }, { "epoch": 0.21179039301310043, "grad_norm": 0.4893296535673842, "learning_rate": 3.5225618631732174e-05, "loss": 0.4002, "step": 485 }, { "epoch": 0.21397379912663755, "grad_norm": 0.5093678162190949, "learning_rate": 3.558951965065502e-05, "loss": 0.4068, "step": 490 }, { "epoch": 0.21615720524017468, "grad_norm": 0.9016225376861969, "learning_rate": 3.595342066957787e-05, "loss": 0.4081, "step": 495 }, { "epoch": 0.2183406113537118, "grad_norm": 0.5844526927479975, "learning_rate": 3.6317321688500726e-05, "loss": 0.4053, "step": 500 }, { "epoch": 0.2205240174672489, "grad_norm": 0.6328384104413857, "learning_rate": 3.6681222707423585e-05, "loss": 0.391, "step": 505 }, { "epoch": 0.22270742358078602, "grad_norm": 0.6295475420727077, "learning_rate": 3.704512372634644e-05, "loss": 0.4061, "step": 510 }, { "epoch": 0.22489082969432314, "grad_norm": 0.6065260581614148, "learning_rate": 3.740902474526929e-05, "loss": 0.4107, "step": 515 }, { "epoch": 0.22707423580786026, "grad_norm": 1.004727626897767, "learning_rate": 3.777292576419214e-05, "loss": 0.3998, "step": 520 }, { "epoch": 0.2292576419213974, "grad_norm": 1.1129157537413497, "learning_rate": 3.8136826783114996e-05, "loss": 0.4005, "step": 525 }, { "epoch": 0.2314410480349345, "grad_norm": 0.9960094773975019, "learning_rate": 3.850072780203784e-05, "loss": 0.4098, "step": 530 }, { "epoch": 0.2336244541484716, "grad_norm": 0.5262503731188081, "learning_rate": 3.88646288209607e-05, "loss": 0.3965, "step": 535 }, { "epoch": 0.23580786026200873, "grad_norm": 0.6294913749088836, "learning_rate": 3.9228529839883554e-05, "loss": 0.3962, "step": 540 }, { "epoch": 0.23799126637554585, "grad_norm": 0.6672651864348986, "learning_rate": 3.9592430858806406e-05, "loss": 0.3976, "step": 545 }, { "epoch": 0.24017467248908297, "grad_norm": 0.6736374930092932, "learning_rate": 3.995633187772926e-05, "loss": 0.3932, "step": 550 }, { "epoch": 0.2423580786026201, "grad_norm": 0.9025467677457151, "learning_rate": 4.032023289665211e-05, "loss": 0.4163, "step": 555 }, { "epoch": 0.2445414847161572, "grad_norm": 0.6615595441983655, "learning_rate": 4.068413391557497e-05, "loss": 0.3885, "step": 560 }, { "epoch": 0.24672489082969432, "grad_norm": 0.747795749818688, "learning_rate": 4.104803493449782e-05, "loss": 0.3979, "step": 565 }, { "epoch": 0.24890829694323144, "grad_norm": 0.5395748717140287, "learning_rate": 4.141193595342067e-05, "loss": 0.4052, "step": 570 }, { "epoch": 0.25109170305676853, "grad_norm": 0.5834684407168078, "learning_rate": 4.177583697234352e-05, "loss": 0.3931, "step": 575 }, { "epoch": 0.25327510917030566, "grad_norm": 0.6149399366377413, "learning_rate": 4.2139737991266375e-05, "loss": 0.4199, "step": 580 }, { "epoch": 0.2554585152838428, "grad_norm": 0.6439187925066173, "learning_rate": 4.250363901018923e-05, "loss": 0.3817, "step": 585 }, { "epoch": 0.2576419213973799, "grad_norm": 0.7335269690355521, "learning_rate": 4.286754002911209e-05, "loss": 0.4146, "step": 590 }, { "epoch": 0.259825327510917, "grad_norm": 0.5948480037959886, "learning_rate": 4.323144104803494e-05, "loss": 0.3894, "step": 595 }, { "epoch": 0.26200873362445415, "grad_norm": 0.7265886088550407, "learning_rate": 4.359534206695779e-05, "loss": 0.4099, "step": 600 }, { "epoch": 0.26419213973799127, "grad_norm": 0.8146189489413153, "learning_rate": 4.395924308588064e-05, "loss": 0.3937, "step": 605 }, { "epoch": 0.2663755458515284, "grad_norm": 0.7479919188399701, "learning_rate": 4.432314410480349e-05, "loss": 0.4041, "step": 610 }, { "epoch": 0.2685589519650655, "grad_norm": 0.6451793400170305, "learning_rate": 4.468704512372635e-05, "loss": 0.3923, "step": 615 }, { "epoch": 0.27074235807860264, "grad_norm": 0.6731070339864949, "learning_rate": 4.50509461426492e-05, "loss": 0.3933, "step": 620 }, { "epoch": 0.27292576419213976, "grad_norm": 0.7633021827753019, "learning_rate": 4.5414847161572056e-05, "loss": 0.3887, "step": 625 }, { "epoch": 0.27510917030567683, "grad_norm": 0.5318275158910994, "learning_rate": 4.577874818049491e-05, "loss": 0.3767, "step": 630 }, { "epoch": 0.27729257641921395, "grad_norm": 0.6053347759447276, "learning_rate": 4.614264919941776e-05, "loss": 0.3996, "step": 635 }, { "epoch": 0.2794759825327511, "grad_norm": 0.8239318519157607, "learning_rate": 4.6506550218340614e-05, "loss": 0.3966, "step": 640 }, { "epoch": 0.2816593886462882, "grad_norm": 0.6239015579696312, "learning_rate": 4.687045123726347e-05, "loss": 0.4006, "step": 645 }, { "epoch": 0.2838427947598253, "grad_norm": 0.5211608695178815, "learning_rate": 4.723435225618632e-05, "loss": 0.3923, "step": 650 }, { "epoch": 0.28602620087336245, "grad_norm": 0.6758177772158337, "learning_rate": 4.759825327510917e-05, "loss": 0.402, "step": 655 }, { "epoch": 0.28820960698689957, "grad_norm": 0.7053141851010356, "learning_rate": 4.7962154294032025e-05, "loss": 0.4099, "step": 660 }, { "epoch": 0.2903930131004367, "grad_norm": 0.6398424235305709, "learning_rate": 4.832605531295488e-05, "loss": 0.3886, "step": 665 }, { "epoch": 0.2925764192139738, "grad_norm": 0.6225872999277295, "learning_rate": 4.868995633187774e-05, "loss": 0.3706, "step": 670 }, { "epoch": 0.29475982532751094, "grad_norm": 0.5559572514117758, "learning_rate": 4.905385735080058e-05, "loss": 0.4012, "step": 675 }, { "epoch": 0.29694323144104806, "grad_norm": 0.5432885290294175, "learning_rate": 4.9417758369723435e-05, "loss": 0.3842, "step": 680 }, { "epoch": 0.29912663755458513, "grad_norm": 0.6340294335348677, "learning_rate": 4.978165938864629e-05, "loss": 0.3855, "step": 685 }, { "epoch": 0.30131004366812225, "grad_norm": 0.7186273990805551, "learning_rate": 4.999998709163553e-05, "loss": 0.3705, "step": 690 }, { "epoch": 0.3034934497816594, "grad_norm": 0.6162542024063481, "learning_rate": 4.999984187268831e-05, "loss": 0.3828, "step": 695 }, { "epoch": 0.3056768558951965, "grad_norm": 0.588350309953633, "learning_rate": 4.9999535300278666e-05, "loss": 0.3763, "step": 700 }, { "epoch": 0.3078602620087336, "grad_norm": 0.5721216258057977, "learning_rate": 4.999906737638528e-05, "loss": 0.3928, "step": 705 }, { "epoch": 0.31004366812227074, "grad_norm": 1.0000585963765807, "learning_rate": 4.9998438104028226e-05, "loss": 0.3926, "step": 710 }, { "epoch": 0.31222707423580787, "grad_norm": 2.2978661693892355, "learning_rate": 4.999764748726891e-05, "loss": 0.385, "step": 715 }, { "epoch": 0.314410480349345, "grad_norm": 1.8144056170150589, "learning_rate": 4.9996695531210135e-05, "loss": 0.4175, "step": 720 }, { "epoch": 0.3165938864628821, "grad_norm": 0.7618077273846784, "learning_rate": 4.9995582241996e-05, "loss": 0.4076, "step": 725 }, { "epoch": 0.31877729257641924, "grad_norm": 0.8570756785316501, "learning_rate": 4.9994307626811865e-05, "loss": 0.3912, "step": 730 }, { "epoch": 0.32096069868995636, "grad_norm": 0.6575118539772534, "learning_rate": 4.999287169388433e-05, "loss": 0.3913, "step": 735 }, { "epoch": 0.3231441048034934, "grad_norm": 0.7341799315850096, "learning_rate": 4.9991274452481164e-05, "loss": 0.4078, "step": 740 }, { "epoch": 0.32532751091703055, "grad_norm": 0.6776721860026673, "learning_rate": 4.9989515912911234e-05, "loss": 0.3957, "step": 745 }, { "epoch": 0.32751091703056767, "grad_norm": 0.6298219471990844, "learning_rate": 4.998759608652449e-05, "loss": 0.4049, "step": 750 }, { "epoch": 0.3296943231441048, "grad_norm": 0.5126548037982753, "learning_rate": 4.998551498571182e-05, "loss": 0.3929, "step": 755 }, { "epoch": 0.3318777292576419, "grad_norm": 0.6814430013566892, "learning_rate": 4.9983272623905045e-05, "loss": 0.3823, "step": 760 }, { "epoch": 0.33406113537117904, "grad_norm": 0.5643156159588145, "learning_rate": 4.9980869015576736e-05, "loss": 0.3761, "step": 765 }, { "epoch": 0.33624454148471616, "grad_norm": 0.4499045190804366, "learning_rate": 4.997830417624024e-05, "loss": 0.3867, "step": 770 }, { "epoch": 0.3384279475982533, "grad_norm": 0.7955611854097291, "learning_rate": 4.9975578122449485e-05, "loss": 0.3953, "step": 775 }, { "epoch": 0.3406113537117904, "grad_norm": 0.5150140528502942, "learning_rate": 4.99726908717989e-05, "loss": 0.3769, "step": 780 }, { "epoch": 0.34279475982532753, "grad_norm": 0.5896604319461464, "learning_rate": 4.9969642442923325e-05, "loss": 0.3776, "step": 785 }, { "epoch": 0.34497816593886466, "grad_norm": 0.7333319728063213, "learning_rate": 4.996643285549787e-05, "loss": 0.3912, "step": 790 }, { "epoch": 0.3471615720524017, "grad_norm": 0.975012641228752, "learning_rate": 4.996306213023778e-05, "loss": 0.374, "step": 795 }, { "epoch": 0.34934497816593885, "grad_norm": 0.6363914305855141, "learning_rate": 4.995953028889833e-05, "loss": 0.3908, "step": 800 }, { "epoch": 0.35152838427947597, "grad_norm": 0.7256711049149246, "learning_rate": 4.995583735427465e-05, "loss": 0.3804, "step": 805 }, { "epoch": 0.3537117903930131, "grad_norm": 0.5201509210204703, "learning_rate": 4.9951983350201604e-05, "loss": 0.3787, "step": 810 }, { "epoch": 0.3558951965065502, "grad_norm": 0.5937818267570265, "learning_rate": 4.9947968301553625e-05, "loss": 0.3681, "step": 815 }, { "epoch": 0.35807860262008734, "grad_norm": 0.5805445788983403, "learning_rate": 4.994379223424456e-05, "loss": 0.3632, "step": 820 }, { "epoch": 0.36026200873362446, "grad_norm": 0.4804762583689749, "learning_rate": 4.99394551752275e-05, "loss": 0.3729, "step": 825 }, { "epoch": 0.3624454148471616, "grad_norm": 0.4780935144472406, "learning_rate": 4.9934957152494585e-05, "loss": 0.366, "step": 830 }, { "epoch": 0.3646288209606987, "grad_norm": 0.46427385625879564, "learning_rate": 4.993029819507687e-05, "loss": 0.3641, "step": 835 }, { "epoch": 0.36681222707423583, "grad_norm": 0.5283252072301436, "learning_rate": 4.9925478333044114e-05, "loss": 0.3797, "step": 840 }, { "epoch": 0.36899563318777295, "grad_norm": 0.5593669993363974, "learning_rate": 4.992049759750455e-05, "loss": 0.3776, "step": 845 }, { "epoch": 0.37117903930131, "grad_norm": 0.6282087870379117, "learning_rate": 4.9915356020604754e-05, "loss": 0.3723, "step": 850 }, { "epoch": 0.37336244541484714, "grad_norm": 0.4876112040234335, "learning_rate": 4.9910053635529376e-05, "loss": 0.3918, "step": 855 }, { "epoch": 0.37554585152838427, "grad_norm": 0.4642413234429665, "learning_rate": 4.9904590476500955e-05, "loss": 0.3809, "step": 860 }, { "epoch": 0.3777292576419214, "grad_norm": 0.44822878442026826, "learning_rate": 4.989896657877972e-05, "loss": 0.37, "step": 865 }, { "epoch": 0.3799126637554585, "grad_norm": 0.4846478506665109, "learning_rate": 4.98931819786633e-05, "loss": 0.3742, "step": 870 }, { "epoch": 0.38209606986899564, "grad_norm": 0.4166354564636158, "learning_rate": 4.988723671348654e-05, "loss": 0.3658, "step": 875 }, { "epoch": 0.38427947598253276, "grad_norm": 0.49223225972882195, "learning_rate": 4.9881130821621255e-05, "loss": 0.3914, "step": 880 }, { "epoch": 0.3864628820960699, "grad_norm": 0.526884770677872, "learning_rate": 4.987486434247596e-05, "loss": 0.3754, "step": 885 }, { "epoch": 0.388646288209607, "grad_norm": 0.5232997418172833, "learning_rate": 4.9868437316495636e-05, "loss": 0.3724, "step": 890 }, { "epoch": 0.39082969432314413, "grad_norm": 0.48376676230007215, "learning_rate": 4.9861849785161465e-05, "loss": 0.3673, "step": 895 }, { "epoch": 0.3930131004366812, "grad_norm": 0.5294590310273921, "learning_rate": 4.9855101790990556e-05, "loss": 0.3727, "step": 900 }, { "epoch": 0.3951965065502183, "grad_norm": 0.47896380765481217, "learning_rate": 4.9848193377535665e-05, "loss": 0.3841, "step": 905 }, { "epoch": 0.39737991266375544, "grad_norm": 0.607838261659882, "learning_rate": 4.9841124589384944e-05, "loss": 0.3723, "step": 910 }, { "epoch": 0.39956331877729256, "grad_norm": 0.5096948643379262, "learning_rate": 4.983389547216161e-05, "loss": 0.3707, "step": 915 }, { "epoch": 0.4017467248908297, "grad_norm": 0.38304781823406214, "learning_rate": 4.9826506072523684e-05, "loss": 0.3545, "step": 920 }, { "epoch": 0.4039301310043668, "grad_norm": 0.587506061988026, "learning_rate": 4.981895643816367e-05, "loss": 0.3718, "step": 925 }, { "epoch": 0.40611353711790393, "grad_norm": 0.5967719221364212, "learning_rate": 4.9811246617808275e-05, "loss": 0.3684, "step": 930 }, { "epoch": 0.40829694323144106, "grad_norm": 0.6935599142313973, "learning_rate": 4.980337666121806e-05, "loss": 0.3789, "step": 935 }, { "epoch": 0.4104803493449782, "grad_norm": 0.5766117112149044, "learning_rate": 4.979534661918712e-05, "loss": 0.3542, "step": 940 }, { "epoch": 0.4126637554585153, "grad_norm": 0.43795584226862594, "learning_rate": 4.978715654354281e-05, "loss": 0.3595, "step": 945 }, { "epoch": 0.4148471615720524, "grad_norm": 0.4430509318673314, "learning_rate": 4.9778806487145326e-05, "loss": 0.3666, "step": 950 }, { "epoch": 0.4170305676855895, "grad_norm": 0.4481773430888534, "learning_rate": 4.977029650388744e-05, "loss": 0.3654, "step": 955 }, { "epoch": 0.4192139737991266, "grad_norm": 0.4657407424462558, "learning_rate": 4.976162664869411e-05, "loss": 0.3542, "step": 960 }, { "epoch": 0.42139737991266374, "grad_norm": 0.5287646283190662, "learning_rate": 4.975279697752213e-05, "loss": 0.3594, "step": 965 }, { "epoch": 0.42358078602620086, "grad_norm": 0.4887455966923865, "learning_rate": 4.974380754735979e-05, "loss": 0.3669, "step": 970 }, { "epoch": 0.425764192139738, "grad_norm": 0.43872446154490324, "learning_rate": 4.9734658416226476e-05, "loss": 0.3616, "step": 975 }, { "epoch": 0.4279475982532751, "grad_norm": 0.6424735520245978, "learning_rate": 4.972534964317233e-05, "loss": 0.369, "step": 980 }, { "epoch": 0.43013100436681223, "grad_norm": 0.48958723865804354, "learning_rate": 4.9715881288277835e-05, "loss": 0.3594, "step": 985 }, { "epoch": 0.43231441048034935, "grad_norm": 0.4479761879052664, "learning_rate": 4.9706253412653446e-05, "loss": 0.3549, "step": 990 }, { "epoch": 0.4344978165938865, "grad_norm": 0.5435346773895244, "learning_rate": 4.9696466078439216e-05, "loss": 0.3671, "step": 995 }, { "epoch": 0.4366812227074236, "grad_norm": 0.5308581082448025, "learning_rate": 4.968651934880434e-05, "loss": 0.365, "step": 1000 }, { "epoch": 0.4388646288209607, "grad_norm": 0.39896740398055186, "learning_rate": 4.96764132879468e-05, "loss": 0.3598, "step": 1005 }, { "epoch": 0.4410480349344978, "grad_norm": 0.5067575654052805, "learning_rate": 4.9666147961092924e-05, "loss": 0.3659, "step": 1010 }, { "epoch": 0.4432314410480349, "grad_norm": 0.5592601406097657, "learning_rate": 4.9655723434496984e-05, "loss": 0.3719, "step": 1015 }, { "epoch": 0.44541484716157204, "grad_norm": 0.4413849291952618, "learning_rate": 4.9645139775440716e-05, "loss": 0.3652, "step": 1020 }, { "epoch": 0.44759825327510916, "grad_norm": 0.3910244442044385, "learning_rate": 4.9634397052232983e-05, "loss": 0.3691, "step": 1025 }, { "epoch": 0.4497816593886463, "grad_norm": 0.46071657374615793, "learning_rate": 4.9623495334209226e-05, "loss": 0.3624, "step": 1030 }, { "epoch": 0.4519650655021834, "grad_norm": 0.4226092000790114, "learning_rate": 4.96124346917311e-05, "loss": 0.3676, "step": 1035 }, { "epoch": 0.45414847161572053, "grad_norm": 0.44419459860035787, "learning_rate": 4.960121519618597e-05, "loss": 0.332, "step": 1040 }, { "epoch": 0.45633187772925765, "grad_norm": 0.3577954443167031, "learning_rate": 4.9589836919986475e-05, "loss": 0.3512, "step": 1045 }, { "epoch": 0.4585152838427948, "grad_norm": 0.5820450329129828, "learning_rate": 4.957829993657005e-05, "loss": 0.3501, "step": 1050 }, { "epoch": 0.4606986899563319, "grad_norm": 0.5150191150181325, "learning_rate": 4.956660432039846e-05, "loss": 0.3597, "step": 1055 }, { "epoch": 0.462882096069869, "grad_norm": 0.4319371225817441, "learning_rate": 4.955475014695729e-05, "loss": 0.3676, "step": 1060 }, { "epoch": 0.4650655021834061, "grad_norm": 0.5533636569254564, "learning_rate": 4.954273749275552e-05, "loss": 0.3727, "step": 1065 }, { "epoch": 0.4672489082969432, "grad_norm": 0.4958192610268456, "learning_rate": 4.9530566435324963e-05, "loss": 0.3488, "step": 1070 }, { "epoch": 0.46943231441048033, "grad_norm": 0.5108202025863451, "learning_rate": 4.9518237053219816e-05, "loss": 0.3709, "step": 1075 }, { "epoch": 0.47161572052401746, "grad_norm": 0.45372615060906546, "learning_rate": 4.950574942601611e-05, "loss": 0.3475, "step": 1080 }, { "epoch": 0.4737991266375546, "grad_norm": 0.43181104589299185, "learning_rate": 4.9493103634311245e-05, "loss": 0.3527, "step": 1085 }, { "epoch": 0.4759825327510917, "grad_norm": 0.49622919369152685, "learning_rate": 4.948029975972341e-05, "loss": 0.34, "step": 1090 }, { "epoch": 0.4781659388646288, "grad_norm": 0.5352955894735107, "learning_rate": 4.9467337884891127e-05, "loss": 0.3578, "step": 1095 }, { "epoch": 0.48034934497816595, "grad_norm": 0.48767956685977704, "learning_rate": 4.945421809347266e-05, "loss": 0.3437, "step": 1100 }, { "epoch": 0.48253275109170307, "grad_norm": 0.5172947775347728, "learning_rate": 4.944094047014547e-05, "loss": 0.3582, "step": 1105 }, { "epoch": 0.4847161572052402, "grad_norm": 0.4633995942526625, "learning_rate": 4.942750510060574e-05, "loss": 0.3369, "step": 1110 }, { "epoch": 0.4868995633187773, "grad_norm": 0.5384683843923247, "learning_rate": 4.9413912071567756e-05, "loss": 0.3683, "step": 1115 }, { "epoch": 0.4890829694323144, "grad_norm": 0.428948742370993, "learning_rate": 4.940016147076336e-05, "loss": 0.3425, "step": 1120 }, { "epoch": 0.4912663755458515, "grad_norm": 0.5160388037360805, "learning_rate": 4.938625338694139e-05, "loss": 0.3609, "step": 1125 }, { "epoch": 0.49344978165938863, "grad_norm": 0.45537160535635335, "learning_rate": 4.937218790986712e-05, "loss": 0.3532, "step": 1130 }, { "epoch": 0.49563318777292575, "grad_norm": 0.4996736665030406, "learning_rate": 4.935796513032165e-05, "loss": 0.3434, "step": 1135 }, { "epoch": 0.4978165938864629, "grad_norm": 0.5155894229667807, "learning_rate": 4.934358514010137e-05, "loss": 0.3569, "step": 1140 }, { "epoch": 0.5, "grad_norm": 0.38793250116768685, "learning_rate": 4.93290480320173e-05, "loss": 0.3545, "step": 1145 }, { "epoch": 0.5021834061135371, "grad_norm": 0.36522790796535515, "learning_rate": 4.931435389989455e-05, "loss": 0.3482, "step": 1150 }, { "epoch": 0.5043668122270742, "grad_norm": 0.46762083247492375, "learning_rate": 4.929950283857167e-05, "loss": 0.3704, "step": 1155 }, { "epoch": 0.5065502183406113, "grad_norm": 0.45795606465334227, "learning_rate": 4.9284494943900085e-05, "loss": 0.3646, "step": 1160 }, { "epoch": 0.5087336244541485, "grad_norm": 0.42995508768886065, "learning_rate": 4.926933031274344e-05, "loss": 0.3626, "step": 1165 }, { "epoch": 0.5109170305676856, "grad_norm": 0.4224390944477537, "learning_rate": 4.925400904297698e-05, "loss": 0.347, "step": 1170 }, { "epoch": 0.5131004366812227, "grad_norm": 0.5046566006479356, "learning_rate": 4.923853123348693e-05, "loss": 0.3678, "step": 1175 }, { "epoch": 0.5152838427947598, "grad_norm": 0.5310348130320324, "learning_rate": 4.922289698416983e-05, "loss": 0.3568, "step": 1180 }, { "epoch": 0.517467248908297, "grad_norm": 0.44019956240689734, "learning_rate": 4.920710639593196e-05, "loss": 0.3555, "step": 1185 }, { "epoch": 0.519650655021834, "grad_norm": 0.36638621076470246, "learning_rate": 4.919115957068859e-05, "loss": 0.3529, "step": 1190 }, { "epoch": 0.5218340611353712, "grad_norm": 0.4899606255024857, "learning_rate": 4.917505661136339e-05, "loss": 0.3491, "step": 1195 }, { "epoch": 0.5240174672489083, "grad_norm": 0.501601743259068, "learning_rate": 4.915879762188774e-05, "loss": 0.3615, "step": 1200 }, { "epoch": 0.5262008733624454, "grad_norm": 0.6025124136368822, "learning_rate": 4.9142382707200096e-05, "loss": 0.3553, "step": 1205 }, { "epoch": 0.5283842794759825, "grad_norm": 0.445329859368048, "learning_rate": 4.912581197324524e-05, "loss": 0.3633, "step": 1210 }, { "epoch": 0.5305676855895196, "grad_norm": 0.6418024421311272, "learning_rate": 4.9109085526973674e-05, "loss": 0.3476, "step": 1215 }, { "epoch": 0.5327510917030568, "grad_norm": 0.4430605711065509, "learning_rate": 4.909220347634087e-05, "loss": 0.359, "step": 1220 }, { "epoch": 0.5349344978165939, "grad_norm": 0.4354421223792483, "learning_rate": 4.907516593030662e-05, "loss": 0.3515, "step": 1225 }, { "epoch": 0.537117903930131, "grad_norm": 0.47281124382089545, "learning_rate": 4.905797299883429e-05, "loss": 0.3629, "step": 1230 }, { "epoch": 0.5393013100436681, "grad_norm": 0.4468031895316623, "learning_rate": 4.904062479289015e-05, "loss": 0.3484, "step": 1235 }, { "epoch": 0.5414847161572053, "grad_norm": 0.4659670701752814, "learning_rate": 4.902312142444264e-05, "loss": 0.3526, "step": 1240 }, { "epoch": 0.5436681222707423, "grad_norm": 0.371869503352742, "learning_rate": 4.900546300646162e-05, "loss": 0.3366, "step": 1245 }, { "epoch": 0.5458515283842795, "grad_norm": 0.42423288318858376, "learning_rate": 4.89876496529177e-05, "loss": 0.3543, "step": 1250 }, { "epoch": 0.5480349344978166, "grad_norm": 0.4609253407888565, "learning_rate": 4.896968147878145e-05, "loss": 0.3565, "step": 1255 }, { "epoch": 0.5502183406113537, "grad_norm": 0.3237267708873633, "learning_rate": 4.895155860002271e-05, "loss": 0.3377, "step": 1260 }, { "epoch": 0.5524017467248908, "grad_norm": 0.4546639718401833, "learning_rate": 4.893328113360975e-05, "loss": 0.3535, "step": 1265 }, { "epoch": 0.5545851528384279, "grad_norm": 0.43148670011250856, "learning_rate": 4.891484919750865e-05, "loss": 0.344, "step": 1270 }, { "epoch": 0.5567685589519651, "grad_norm": 0.3991105241595635, "learning_rate": 4.889626291068241e-05, "loss": 0.339, "step": 1275 }, { "epoch": 0.5589519650655022, "grad_norm": 0.48987054857422774, "learning_rate": 4.887752239309027e-05, "loss": 0.3391, "step": 1280 }, { "epoch": 0.5611353711790393, "grad_norm": 0.4609741215989259, "learning_rate": 4.885862776568689e-05, "loss": 0.3576, "step": 1285 }, { "epoch": 0.5633187772925764, "grad_norm": 0.50704867138633, "learning_rate": 4.883957915042158e-05, "loss": 0.3496, "step": 1290 }, { "epoch": 0.5655021834061136, "grad_norm": 0.48159077762660746, "learning_rate": 4.882037667023752e-05, "loss": 0.3361, "step": 1295 }, { "epoch": 0.5676855895196506, "grad_norm": 0.7062981902269905, "learning_rate": 4.880102044907096e-05, "loss": 0.3677, "step": 1300 }, { "epoch": 0.5698689956331878, "grad_norm": 0.4888521580220186, "learning_rate": 4.878151061185042e-05, "loss": 0.3294, "step": 1305 }, { "epoch": 0.5720524017467249, "grad_norm": 0.4540470014267252, "learning_rate": 4.876184728449589e-05, "loss": 0.3488, "step": 1310 }, { "epoch": 0.574235807860262, "grad_norm": 0.3930373040922702, "learning_rate": 4.874203059391802e-05, "loss": 0.3364, "step": 1315 }, { "epoch": 0.5764192139737991, "grad_norm": 0.4451157942220829, "learning_rate": 4.8722060668017267e-05, "loss": 0.3562, "step": 1320 }, { "epoch": 0.5786026200873362, "grad_norm": 0.3942536599077881, "learning_rate": 4.870193763568312e-05, "loss": 0.3619, "step": 1325 }, { "epoch": 0.5807860262008734, "grad_norm": 0.48861179907906854, "learning_rate": 4.8681661626793246e-05, "loss": 0.3468, "step": 1330 }, { "epoch": 0.5829694323144105, "grad_norm": 0.42138612451614066, "learning_rate": 4.866123277221264e-05, "loss": 0.347, "step": 1335 }, { "epoch": 0.5851528384279476, "grad_norm": 0.4074290406551753, "learning_rate": 4.864065120379278e-05, "loss": 0.3573, "step": 1340 }, { "epoch": 0.5873362445414847, "grad_norm": 0.3987198118209572, "learning_rate": 4.861991705437081e-05, "loss": 0.3598, "step": 1345 }, { "epoch": 0.5895196506550219, "grad_norm": 0.46460108590367194, "learning_rate": 4.859903045776866e-05, "loss": 0.3524, "step": 1350 }, { "epoch": 0.5917030567685589, "grad_norm": 0.4803498793618286, "learning_rate": 4.857799154879215e-05, "loss": 0.3605, "step": 1355 }, { "epoch": 0.5938864628820961, "grad_norm": 0.5901063012029109, "learning_rate": 4.8556800463230165e-05, "loss": 0.3437, "step": 1360 }, { "epoch": 0.5960698689956332, "grad_norm": 0.4641267585126181, "learning_rate": 4.8535457337853794e-05, "loss": 0.3285, "step": 1365 }, { "epoch": 0.5982532751091703, "grad_norm": 0.3647796263943895, "learning_rate": 4.8513962310415396e-05, "loss": 0.3463, "step": 1370 }, { "epoch": 0.6004366812227074, "grad_norm": 0.3648828726715139, "learning_rate": 4.849231551964771e-05, "loss": 0.3535, "step": 1375 }, { "epoch": 0.6026200873362445, "grad_norm": 0.40542343011346316, "learning_rate": 4.847051710526302e-05, "loss": 0.3437, "step": 1380 }, { "epoch": 0.6048034934497817, "grad_norm": 0.350582105718404, "learning_rate": 4.8448567207952205e-05, "loss": 0.3549, "step": 1385 }, { "epoch": 0.6069868995633187, "grad_norm": 0.39947990507599496, "learning_rate": 4.8426465969383826e-05, "loss": 0.3243, "step": 1390 }, { "epoch": 0.6091703056768559, "grad_norm": 0.4129694646452404, "learning_rate": 4.840421353220326e-05, "loss": 0.3251, "step": 1395 }, { "epoch": 0.611353711790393, "grad_norm": 0.3820909710575323, "learning_rate": 4.83818100400317e-05, "loss": 0.334, "step": 1400 }, { "epoch": 0.6135371179039302, "grad_norm": 0.4069293886007247, "learning_rate": 4.835925563746532e-05, "loss": 0.3551, "step": 1405 }, { "epoch": 0.6157205240174672, "grad_norm": 0.4582430996524103, "learning_rate": 4.833655047007427e-05, "loss": 0.3504, "step": 1410 }, { "epoch": 0.6179039301310044, "grad_norm": 0.3828069417511488, "learning_rate": 4.831369468440177e-05, "loss": 0.3358, "step": 1415 }, { "epoch": 0.6200873362445415, "grad_norm": 0.379141828052868, "learning_rate": 4.8290688427963175e-05, "loss": 0.3315, "step": 1420 }, { "epoch": 0.6222707423580786, "grad_norm": 0.4059377645684439, "learning_rate": 4.8267531849244974e-05, "loss": 0.3403, "step": 1425 }, { "epoch": 0.6244541484716157, "grad_norm": 0.3781489511039065, "learning_rate": 4.8244225097703885e-05, "loss": 0.351, "step": 1430 }, { "epoch": 0.6266375545851528, "grad_norm": 0.36811899180808383, "learning_rate": 4.8220768323765866e-05, "loss": 0.345, "step": 1435 }, { "epoch": 0.62882096069869, "grad_norm": 0.48695298818921573, "learning_rate": 4.819716167882513e-05, "loss": 0.343, "step": 1440 }, { "epoch": 0.631004366812227, "grad_norm": 0.41424246411412347, "learning_rate": 4.8173405315243216e-05, "loss": 0.3381, "step": 1445 }, { "epoch": 0.6331877729257642, "grad_norm": 0.5982905072375967, "learning_rate": 4.814949938634793e-05, "loss": 0.3334, "step": 1450 }, { "epoch": 0.6353711790393013, "grad_norm": 0.41432212938369534, "learning_rate": 4.8125444046432434e-05, "loss": 0.3387, "step": 1455 }, { "epoch": 0.6375545851528385, "grad_norm": 0.3970797596419897, "learning_rate": 4.810123945075421e-05, "loss": 0.3485, "step": 1460 }, { "epoch": 0.6397379912663755, "grad_norm": 0.36649659825973896, "learning_rate": 4.807688575553406e-05, "loss": 0.3381, "step": 1465 }, { "epoch": 0.6419213973799127, "grad_norm": 0.4151615051412959, "learning_rate": 4.80523831179551e-05, "loss": 0.3316, "step": 1470 }, { "epoch": 0.6441048034934498, "grad_norm": 0.41296409849005844, "learning_rate": 4.8027731696161735e-05, "loss": 0.3378, "step": 1475 }, { "epoch": 0.6462882096069869, "grad_norm": 0.43599613227274564, "learning_rate": 4.8002931649258674e-05, "loss": 0.3343, "step": 1480 }, { "epoch": 0.648471615720524, "grad_norm": 0.5041455481985673, "learning_rate": 4.7977983137309855e-05, "loss": 0.3337, "step": 1485 }, { "epoch": 0.6506550218340611, "grad_norm": 0.41603606102503005, "learning_rate": 4.795288632133747e-05, "loss": 0.3328, "step": 1490 }, { "epoch": 0.6528384279475983, "grad_norm": 0.4091756291458897, "learning_rate": 4.792764136332084e-05, "loss": 0.3558, "step": 1495 }, { "epoch": 0.6550218340611353, "grad_norm": 0.3535456736539121, "learning_rate": 4.790224842619548e-05, "loss": 0.3305, "step": 1500 }, { "epoch": 0.6572052401746725, "grad_norm": 0.4204136805374324, "learning_rate": 4.787670767385194e-05, "loss": 0.3399, "step": 1505 }, { "epoch": 0.6593886462882096, "grad_norm": 0.4138703097114033, "learning_rate": 4.785101927113482e-05, "loss": 0.3423, "step": 1510 }, { "epoch": 0.6615720524017468, "grad_norm": 0.42867964938598574, "learning_rate": 4.782518338384168e-05, "loss": 0.3474, "step": 1515 }, { "epoch": 0.6637554585152838, "grad_norm": 0.3655864787630958, "learning_rate": 4.779920017872197e-05, "loss": 0.3426, "step": 1520 }, { "epoch": 0.665938864628821, "grad_norm": 0.31312668856989573, "learning_rate": 4.7773069823475945e-05, "loss": 0.3317, "step": 1525 }, { "epoch": 0.6681222707423581, "grad_norm": 0.3720008834724695, "learning_rate": 4.774679248675362e-05, "loss": 0.3468, "step": 1530 }, { "epoch": 0.6703056768558951, "grad_norm": 0.3500906347775602, "learning_rate": 4.772036833815363e-05, "loss": 0.3305, "step": 1535 }, { "epoch": 0.6724890829694323, "grad_norm": 0.4324016765021572, "learning_rate": 4.7693797548222166e-05, "loss": 0.3226, "step": 1540 }, { "epoch": 0.6746724890829694, "grad_norm": 0.40696438387223544, "learning_rate": 4.766708028845188e-05, "loss": 0.3325, "step": 1545 }, { "epoch": 0.6768558951965066, "grad_norm": 0.4155361411655341, "learning_rate": 4.764021673128075e-05, "loss": 0.3482, "step": 1550 }, { "epoch": 0.6790393013100436, "grad_norm": 0.4890226077260248, "learning_rate": 4.761320705009099e-05, "loss": 0.3457, "step": 1555 }, { "epoch": 0.6812227074235808, "grad_norm": 0.42425030996602786, "learning_rate": 4.758605141920792e-05, "loss": 0.3512, "step": 1560 }, { "epoch": 0.6834061135371179, "grad_norm": 0.3594877731084044, "learning_rate": 4.755875001389887e-05, "loss": 0.3434, "step": 1565 }, { "epoch": 0.6855895196506551, "grad_norm": 0.3979846319153574, "learning_rate": 4.753130301037199e-05, "loss": 0.3337, "step": 1570 }, { "epoch": 0.6877729257641921, "grad_norm": 0.402834791445396, "learning_rate": 4.7503710585775174e-05, "loss": 0.3387, "step": 1575 }, { "epoch": 0.6899563318777293, "grad_norm": 0.34220820957249626, "learning_rate": 4.747597291819487e-05, "loss": 0.3439, "step": 1580 }, { "epoch": 0.6921397379912664, "grad_norm": 0.3831695794480096, "learning_rate": 4.744809018665495e-05, "loss": 0.3373, "step": 1585 }, { "epoch": 0.6943231441048034, "grad_norm": 0.40957271083541796, "learning_rate": 4.742006257111558e-05, "loss": 0.3358, "step": 1590 }, { "epoch": 0.6965065502183406, "grad_norm": 0.3989291887534643, "learning_rate": 4.7391890252472016e-05, "loss": 0.3624, "step": 1595 }, { "epoch": 0.6986899563318777, "grad_norm": 0.47006740071344755, "learning_rate": 4.736357341255344e-05, "loss": 0.3427, "step": 1600 }, { "epoch": 0.7008733624454149, "grad_norm": 0.5495070509701956, "learning_rate": 4.7335112234121824e-05, "loss": 0.3332, "step": 1605 }, { "epoch": 0.7030567685589519, "grad_norm": 0.37491891454833864, "learning_rate": 4.730650690087072e-05, "loss": 0.3406, "step": 1610 }, { "epoch": 0.7052401746724891, "grad_norm": 0.43036507029194554, "learning_rate": 4.727775759742408e-05, "loss": 0.3431, "step": 1615 }, { "epoch": 0.7074235807860262, "grad_norm": 0.3820782925001961, "learning_rate": 4.724886450933505e-05, "loss": 0.3436, "step": 1620 }, { "epoch": 0.7096069868995634, "grad_norm": 0.4860255736060773, "learning_rate": 4.721982782308481e-05, "loss": 0.3523, "step": 1625 }, { "epoch": 0.7117903930131004, "grad_norm": 0.3695141770304634, "learning_rate": 4.719064772608134e-05, "loss": 0.3462, "step": 1630 }, { "epoch": 0.7139737991266376, "grad_norm": 0.4003960638271693, "learning_rate": 4.7161324406658225e-05, "loss": 0.3286, "step": 1635 }, { "epoch": 0.7161572052401747, "grad_norm": 0.3530404563540438, "learning_rate": 4.713185805407342e-05, "loss": 0.3236, "step": 1640 }, { "epoch": 0.7183406113537117, "grad_norm": 0.584075905055993, "learning_rate": 4.7102248858508044e-05, "loss": 0.3544, "step": 1645 }, { "epoch": 0.7205240174672489, "grad_norm": 0.46166165713972934, "learning_rate": 4.7072497011065156e-05, "loss": 0.3298, "step": 1650 }, { "epoch": 0.722707423580786, "grad_norm": 0.3355056455618887, "learning_rate": 4.704260270376853e-05, "loss": 0.3264, "step": 1655 }, { "epoch": 0.7248908296943232, "grad_norm": 0.39311481965484996, "learning_rate": 4.7012566129561365e-05, "loss": 0.3381, "step": 1660 }, { "epoch": 0.7270742358078602, "grad_norm": 0.38148657969468514, "learning_rate": 4.6982387482305104e-05, "loss": 0.3322, "step": 1665 }, { "epoch": 0.7292576419213974, "grad_norm": 0.44427131750278887, "learning_rate": 4.6952066956778155e-05, "loss": 0.3268, "step": 1670 }, { "epoch": 0.7314410480349345, "grad_norm": 0.3857743094672692, "learning_rate": 4.692160474867463e-05, "loss": 0.3324, "step": 1675 }, { "epoch": 0.7336244541484717, "grad_norm": 0.3899786180476667, "learning_rate": 4.6891001054603056e-05, "loss": 0.3374, "step": 1680 }, { "epoch": 0.7358078602620087, "grad_norm": 0.3696166334234957, "learning_rate": 4.6860256072085196e-05, "loss": 0.3453, "step": 1685 }, { "epoch": 0.7379912663755459, "grad_norm": 0.30167847925632657, "learning_rate": 4.6829369999554664e-05, "loss": 0.3297, "step": 1690 }, { "epoch": 0.740174672489083, "grad_norm": 0.37202185004193333, "learning_rate": 4.6798343036355704e-05, "loss": 0.3058, "step": 1695 }, { "epoch": 0.74235807860262, "grad_norm": 0.3965117699988383, "learning_rate": 4.6767175382741914e-05, "loss": 0.3367, "step": 1700 }, { "epoch": 0.7445414847161572, "grad_norm": 0.37265664937861914, "learning_rate": 4.673586723987491e-05, "loss": 0.3504, "step": 1705 }, { "epoch": 0.7467248908296943, "grad_norm": 0.42824465747750634, "learning_rate": 4.670441880982307e-05, "loss": 0.3323, "step": 1710 }, { "epoch": 0.7489082969432315, "grad_norm": 0.3708258588256483, "learning_rate": 4.667283029556019e-05, "loss": 0.3288, "step": 1715 }, { "epoch": 0.7510917030567685, "grad_norm": 0.4271790614273471, "learning_rate": 4.664110190096421e-05, "loss": 0.3436, "step": 1720 }, { "epoch": 0.7532751091703057, "grad_norm": 0.37662914132162023, "learning_rate": 4.660923383081588e-05, "loss": 0.3401, "step": 1725 }, { "epoch": 0.7554585152838428, "grad_norm": 0.3574814773268414, "learning_rate": 4.657722629079744e-05, "loss": 0.3359, "step": 1730 }, { "epoch": 0.75764192139738, "grad_norm": 0.44811194231414175, "learning_rate": 4.6545079487491285e-05, "loss": 0.3466, "step": 1735 }, { "epoch": 0.759825327510917, "grad_norm": 0.5433785036597635, "learning_rate": 4.6512793628378657e-05, "loss": 0.3348, "step": 1740 }, { "epoch": 0.7620087336244541, "grad_norm": 0.424131707206492, "learning_rate": 4.648036892183828e-05, "loss": 0.3381, "step": 1745 }, { "epoch": 0.7641921397379913, "grad_norm": 0.3485719470067596, "learning_rate": 4.644780557714502e-05, "loss": 0.3338, "step": 1750 }, { "epoch": 0.7663755458515283, "grad_norm": 0.3595654663623888, "learning_rate": 4.641510380446854e-05, "loss": 0.3238, "step": 1755 }, { "epoch": 0.7685589519650655, "grad_norm": 0.4262032500390272, "learning_rate": 4.638226381487195e-05, "loss": 0.3373, "step": 1760 }, { "epoch": 0.7707423580786026, "grad_norm": 0.44078101888220983, "learning_rate": 4.634928582031043e-05, "loss": 0.3449, "step": 1765 }, { "epoch": 0.7729257641921398, "grad_norm": 0.4054216447358104, "learning_rate": 4.631617003362988e-05, "loss": 0.3337, "step": 1770 }, { "epoch": 0.7751091703056768, "grad_norm": 0.3848597633812537, "learning_rate": 4.628291666856551e-05, "loss": 0.3211, "step": 1775 }, { "epoch": 0.777292576419214, "grad_norm": 0.3408957154658452, "learning_rate": 4.6249525939740514e-05, "loss": 0.3306, "step": 1780 }, { "epoch": 0.7794759825327511, "grad_norm": 0.4050082743023021, "learning_rate": 4.6215998062664636e-05, "loss": 0.336, "step": 1785 }, { "epoch": 0.7816593886462883, "grad_norm": 0.4303256630650177, "learning_rate": 4.61823332537328e-05, "loss": 0.3271, "step": 1790 }, { "epoch": 0.7838427947598253, "grad_norm": 0.3879573361229342, "learning_rate": 4.6148531730223735e-05, "loss": 0.3309, "step": 1795 }, { "epoch": 0.7860262008733624, "grad_norm": 0.3340993854538669, "learning_rate": 4.611459371029853e-05, "loss": 0.3273, "step": 1800 }, { "epoch": 0.7882096069868996, "grad_norm": 0.41220645591308713, "learning_rate": 4.6080519412999244e-05, "loss": 0.3256, "step": 1805 }, { "epoch": 0.7903930131004366, "grad_norm": 0.4607924969239619, "learning_rate": 4.604630905824751e-05, "loss": 0.3344, "step": 1810 }, { "epoch": 0.7925764192139738, "grad_norm": 0.3986158317545471, "learning_rate": 4.601196286684309e-05, "loss": 0.3301, "step": 1815 }, { "epoch": 0.7947598253275109, "grad_norm": 0.31961695415950014, "learning_rate": 4.597748106046247e-05, "loss": 0.3305, "step": 1820 }, { "epoch": 0.7969432314410481, "grad_norm": 0.390531096769233, "learning_rate": 4.5942863861657394e-05, "loss": 0.3359, "step": 1825 }, { "epoch": 0.7991266375545851, "grad_norm": 0.3431703506154723, "learning_rate": 4.590811149385348e-05, "loss": 0.3422, "step": 1830 }, { "epoch": 0.8013100436681223, "grad_norm": 0.41192305785453975, "learning_rate": 4.5873224181348736e-05, "loss": 0.3247, "step": 1835 }, { "epoch": 0.8034934497816594, "grad_norm": 0.2924629037593409, "learning_rate": 4.5838202149312145e-05, "loss": 0.3304, "step": 1840 }, { "epoch": 0.8056768558951966, "grad_norm": 0.5634336476609664, "learning_rate": 4.580304562378218e-05, "loss": 0.3303, "step": 1845 }, { "epoch": 0.8078602620087336, "grad_norm": 0.3544990534025258, "learning_rate": 4.576775483166536e-05, "loss": 0.3234, "step": 1850 }, { "epoch": 0.8100436681222707, "grad_norm": 0.3459858077037921, "learning_rate": 4.5732330000734775e-05, "loss": 0.3319, "step": 1855 }, { "epoch": 0.8122270742358079, "grad_norm": 0.3524884912271132, "learning_rate": 4.569677135962864e-05, "loss": 0.3278, "step": 1860 }, { "epoch": 0.8144104803493449, "grad_norm": 0.42168404868370957, "learning_rate": 4.5661079137848817e-05, "loss": 0.314, "step": 1865 }, { "epoch": 0.8165938864628821, "grad_norm": 0.4177554810132261, "learning_rate": 4.562525356575928e-05, "loss": 0.319, "step": 1870 }, { "epoch": 0.8187772925764192, "grad_norm": 0.3406360499086555, "learning_rate": 4.558929487458472e-05, "loss": 0.3307, "step": 1875 }, { "epoch": 0.8209606986899564, "grad_norm": 0.5953338048353011, "learning_rate": 4.5553203296408966e-05, "loss": 0.3363, "step": 1880 }, { "epoch": 0.8231441048034934, "grad_norm": 0.3755495334670722, "learning_rate": 4.551697906417353e-05, "loss": 0.3228, "step": 1885 }, { "epoch": 0.8253275109170306, "grad_norm": 0.5244192452428544, "learning_rate": 4.5480622411676105e-05, "loss": 0.3371, "step": 1890 }, { "epoch": 0.8275109170305677, "grad_norm": 0.42718167146451547, "learning_rate": 4.5444133573569056e-05, "loss": 0.3394, "step": 1895 }, { "epoch": 0.8296943231441049, "grad_norm": 0.5786026070553736, "learning_rate": 4.540751278535788e-05, "loss": 0.3413, "step": 1900 }, { "epoch": 0.8318777292576419, "grad_norm": 0.3994255104166588, "learning_rate": 4.5370760283399714e-05, "loss": 0.3237, "step": 1905 }, { "epoch": 0.834061135371179, "grad_norm": 0.465218260660747, "learning_rate": 4.533387630490179e-05, "loss": 0.3258, "step": 1910 }, { "epoch": 0.8362445414847162, "grad_norm": 0.3503020390555209, "learning_rate": 4.529686108791993e-05, "loss": 0.3365, "step": 1915 }, { "epoch": 0.8384279475982532, "grad_norm": 0.2937649633191906, "learning_rate": 4.525971487135697e-05, "loss": 0.3261, "step": 1920 }, { "epoch": 0.8406113537117904, "grad_norm": 0.3640608075443947, "learning_rate": 4.522243789496125e-05, "loss": 0.3314, "step": 1925 }, { "epoch": 0.8427947598253275, "grad_norm": 0.46642738830210834, "learning_rate": 4.518503039932508e-05, "loss": 0.3294, "step": 1930 }, { "epoch": 0.8449781659388647, "grad_norm": 0.49375163248568255, "learning_rate": 4.5147492625883133e-05, "loss": 0.3346, "step": 1935 }, { "epoch": 0.8471615720524017, "grad_norm": 0.43745545950003145, "learning_rate": 4.510982481691093e-05, "loss": 0.3356, "step": 1940 }, { "epoch": 0.8493449781659389, "grad_norm": 0.3849745896242002, "learning_rate": 4.5072027215523256e-05, "loss": 0.328, "step": 1945 }, { "epoch": 0.851528384279476, "grad_norm": 0.323723856108267, "learning_rate": 4.503410006567262e-05, "loss": 0.3137, "step": 1950 }, { "epoch": 0.8537117903930131, "grad_norm": 0.3497043951204244, "learning_rate": 4.499604361214764e-05, "loss": 0.3367, "step": 1955 }, { "epoch": 0.8558951965065502, "grad_norm": 0.3279206934648897, "learning_rate": 4.495785810057151e-05, "loss": 0.3305, "step": 1960 }, { "epoch": 0.8580786026200873, "grad_norm": 0.35534435710755663, "learning_rate": 4.4919543777400354e-05, "loss": 0.3148, "step": 1965 }, { "epoch": 0.8602620087336245, "grad_norm": 0.35737396766118557, "learning_rate": 4.488110088992168e-05, "loss": 0.3365, "step": 1970 }, { "epoch": 0.8624454148471615, "grad_norm": 0.3811344665888169, "learning_rate": 4.4842529686252774e-05, "loss": 0.3412, "step": 1975 }, { "epoch": 0.8646288209606987, "grad_norm": 0.3586035760646476, "learning_rate": 4.4803830415339124e-05, "loss": 0.3319, "step": 1980 }, { "epoch": 0.8668122270742358, "grad_norm": 0.35951201381552045, "learning_rate": 4.476500332695275e-05, "loss": 0.3123, "step": 1985 }, { "epoch": 0.868995633187773, "grad_norm": 0.4398889292154198, "learning_rate": 4.472604867169063e-05, "loss": 0.3258, "step": 1990 }, { "epoch": 0.87117903930131, "grad_norm": 0.45684784821543106, "learning_rate": 4.468696670097312e-05, "loss": 0.3321, "step": 1995 }, { "epoch": 0.8733624454148472, "grad_norm": 0.4624735089982277, "learning_rate": 4.464775766704225e-05, "loss": 0.339, "step": 2000 }, { "epoch": 0.8755458515283843, "grad_norm": 0.39373793988338507, "learning_rate": 4.4608421822960165e-05, "loss": 0.3232, "step": 2005 }, { "epoch": 0.8777292576419214, "grad_norm": 0.4793265990786189, "learning_rate": 4.456895942260745e-05, "loss": 0.3115, "step": 2010 }, { "epoch": 0.8799126637554585, "grad_norm": 0.4596917524946721, "learning_rate": 4.4529370720681516e-05, "loss": 0.3327, "step": 2015 }, { "epoch": 0.8820960698689956, "grad_norm": 0.41630714877910346, "learning_rate": 4.4489655972694946e-05, "loss": 0.3234, "step": 2020 }, { "epoch": 0.8842794759825328, "grad_norm": 0.433659367959683, "learning_rate": 4.444981543497384e-05, "loss": 0.3268, "step": 2025 }, { "epoch": 0.8864628820960698, "grad_norm": 0.4911763094493765, "learning_rate": 4.440984936465618e-05, "loss": 0.3283, "step": 2030 }, { "epoch": 0.888646288209607, "grad_norm": 0.43509708304067723, "learning_rate": 4.436975801969013e-05, "loss": 0.3166, "step": 2035 }, { "epoch": 0.8908296943231441, "grad_norm": 0.32181145619100665, "learning_rate": 4.4329541658832454e-05, "loss": 0.3275, "step": 2040 }, { "epoch": 0.8930131004366813, "grad_norm": 0.5173193806980154, "learning_rate": 4.428920054164673e-05, "loss": 0.3367, "step": 2045 }, { "epoch": 0.8951965065502183, "grad_norm": 0.31983525834019055, "learning_rate": 4.424873492850177e-05, "loss": 0.3332, "step": 2050 }, { "epoch": 0.8973799126637555, "grad_norm": 0.3070335963036919, "learning_rate": 4.42081450805699e-05, "loss": 0.3416, "step": 2055 }, { "epoch": 0.8995633187772926, "grad_norm": 0.33240317525320323, "learning_rate": 4.4167431259825274e-05, "loss": 0.3214, "step": 2060 }, { "epoch": 0.9017467248908297, "grad_norm": 0.6261250657003081, "learning_rate": 4.41265937290422e-05, "loss": 0.3403, "step": 2065 }, { "epoch": 0.9039301310043668, "grad_norm": 0.6469352618052645, "learning_rate": 4.408563275179341e-05, "loss": 0.3161, "step": 2070 }, { "epoch": 0.9061135371179039, "grad_norm": 0.5128181692159853, "learning_rate": 4.404454859244841e-05, "loss": 0.3513, "step": 2075 }, { "epoch": 0.9082969432314411, "grad_norm": 0.4054137512768172, "learning_rate": 4.400334151617171e-05, "loss": 0.3551, "step": 2080 }, { "epoch": 0.9104803493449781, "grad_norm": 0.5160259798496531, "learning_rate": 4.39620117889212e-05, "loss": 0.3202, "step": 2085 }, { "epoch": 0.9126637554585153, "grad_norm": 1.180298141482782, "learning_rate": 4.3920559677446335e-05, "loss": 0.3264, "step": 2090 }, { "epoch": 0.9148471615720524, "grad_norm": 0.4807710172298594, "learning_rate": 4.387898544928648e-05, "loss": 0.3382, "step": 2095 }, { "epoch": 0.9170305676855895, "grad_norm": 0.5325040390840762, "learning_rate": 4.383728937276915e-05, "loss": 0.3408, "step": 2100 }, { "epoch": 0.9192139737991266, "grad_norm": 0.4423809835776037, "learning_rate": 4.37954717170083e-05, "loss": 0.3336, "step": 2105 }, { "epoch": 0.9213973799126638, "grad_norm": 0.5498410601529987, "learning_rate": 4.375353275190259e-05, "loss": 0.3309, "step": 2110 }, { "epoch": 0.9235807860262009, "grad_norm": 0.37076013082197373, "learning_rate": 4.371147274813362e-05, "loss": 0.3414, "step": 2115 }, { "epoch": 0.925764192139738, "grad_norm": 0.403897556358072, "learning_rate": 4.3669291977164185e-05, "loss": 0.3308, "step": 2120 }, { "epoch": 0.9279475982532751, "grad_norm": 0.4050123744090438, "learning_rate": 4.362699071123655e-05, "loss": 0.3383, "step": 2125 }, { "epoch": 0.9301310043668122, "grad_norm": 0.4499616242086888, "learning_rate": 4.358456922337068e-05, "loss": 0.3247, "step": 2130 }, { "epoch": 0.9323144104803494, "grad_norm": 0.41076931445914805, "learning_rate": 4.354202778736245e-05, "loss": 0.3325, "step": 2135 }, { "epoch": 0.9344978165938864, "grad_norm": 0.6474266802788862, "learning_rate": 4.3499366677781927e-05, "loss": 0.3384, "step": 2140 }, { "epoch": 0.9366812227074236, "grad_norm": 0.43958950888079396, "learning_rate": 4.3456586169971555e-05, "loss": 0.3155, "step": 2145 }, { "epoch": 0.9388646288209607, "grad_norm": 0.430386825850428, "learning_rate": 4.341368654004442e-05, "loss": 0.3407, "step": 2150 }, { "epoch": 0.9410480349344978, "grad_norm": 0.4400594923102947, "learning_rate": 4.3370668064882406e-05, "loss": 0.3227, "step": 2155 }, { "epoch": 0.9432314410480349, "grad_norm": 0.49619465693365633, "learning_rate": 4.332753102213447e-05, "loss": 0.3488, "step": 2160 }, { "epoch": 0.9454148471615721, "grad_norm": 0.2953653819485504, "learning_rate": 4.328427569021481e-05, "loss": 0.3237, "step": 2165 }, { "epoch": 0.9475982532751092, "grad_norm": 0.3277990570496694, "learning_rate": 4.3240902348301134e-05, "loss": 0.3137, "step": 2170 }, { "epoch": 0.9497816593886463, "grad_norm": 0.31994132171967327, "learning_rate": 4.319741127633273e-05, "loss": 0.3263, "step": 2175 }, { "epoch": 0.9519650655021834, "grad_norm": 0.3159818487824089, "learning_rate": 4.315380275500879e-05, "loss": 0.3342, "step": 2180 }, { "epoch": 0.9541484716157205, "grad_norm": 0.37710542197615304, "learning_rate": 4.311007706578654e-05, "loss": 0.3202, "step": 2185 }, { "epoch": 0.9563318777292577, "grad_norm": 0.39548613497208596, "learning_rate": 4.306623449087939e-05, "loss": 0.3182, "step": 2190 }, { "epoch": 0.9585152838427947, "grad_norm": 0.3737799555086228, "learning_rate": 4.3022275313255215e-05, "loss": 0.3202, "step": 2195 }, { "epoch": 0.9606986899563319, "grad_norm": 0.3302761729103403, "learning_rate": 4.2978199816634404e-05, "loss": 0.3134, "step": 2200 }, { "epoch": 0.962882096069869, "grad_norm": 0.3169568046490519, "learning_rate": 4.293400828548812e-05, "loss": 0.3281, "step": 2205 }, { "epoch": 0.9650655021834061, "grad_norm": 0.3469828187440111, "learning_rate": 4.2889701005036444e-05, "loss": 0.3257, "step": 2210 }, { "epoch": 0.9672489082969432, "grad_norm": 0.41804346589863545, "learning_rate": 4.28452782612465e-05, "loss": 0.3395, "step": 2215 }, { "epoch": 0.9694323144104804, "grad_norm": 0.33364679303892797, "learning_rate": 4.2800740340830636e-05, "loss": 0.3127, "step": 2220 }, { "epoch": 0.9716157205240175, "grad_norm": 0.3192103795670632, "learning_rate": 4.275608753124459e-05, "loss": 0.3139, "step": 2225 }, { "epoch": 0.9737991266375546, "grad_norm": 0.43164296461188517, "learning_rate": 4.271132012068559e-05, "loss": 0.3144, "step": 2230 }, { "epoch": 0.9759825327510917, "grad_norm": 0.36786016009963335, "learning_rate": 4.266643839809054e-05, "loss": 0.3217, "step": 2235 }, { "epoch": 0.9781659388646288, "grad_norm": 0.40515456026599855, "learning_rate": 4.262144265313412e-05, "loss": 0.3263, "step": 2240 }, { "epoch": 0.980349344978166, "grad_norm": 0.40316039408536825, "learning_rate": 4.2576333176226944e-05, "loss": 0.3174, "step": 2245 }, { "epoch": 0.982532751091703, "grad_norm": 0.4036377506232831, "learning_rate": 4.253111025851366e-05, "loss": 0.3324, "step": 2250 }, { "epoch": 0.9847161572052402, "grad_norm": 0.38600938504920373, "learning_rate": 4.2485774191871084e-05, "loss": 0.3148, "step": 2255 }, { "epoch": 0.9868995633187773, "grad_norm": 0.3724866092235095, "learning_rate": 4.244032526890633e-05, "loss": 0.32, "step": 2260 }, { "epoch": 0.9890829694323144, "grad_norm": 0.45130824685960164, "learning_rate": 4.239476378295489e-05, "loss": 0.3208, "step": 2265 }, { "epoch": 0.9912663755458515, "grad_norm": 0.3614299023632802, "learning_rate": 4.2349090028078766e-05, "loss": 0.3108, "step": 2270 }, { "epoch": 0.9934497816593887, "grad_norm": 0.3529974003854878, "learning_rate": 4.230330429906456e-05, "loss": 0.3114, "step": 2275 }, { "epoch": 0.9956331877729258, "grad_norm": 0.3358112816928416, "learning_rate": 4.225740689142158e-05, "loss": 0.3168, "step": 2280 }, { "epoch": 0.9978165938864629, "grad_norm": 0.3907341561097172, "learning_rate": 4.221139810137992e-05, "loss": 0.3174, "step": 2285 }, { "epoch": 1.0, "grad_norm": 0.37377145244747273, "learning_rate": 4.216527822588857e-05, "loss": 0.3168, "step": 2290 }, { "epoch": 1.002183406113537, "grad_norm": 0.40852537721961585, "learning_rate": 4.2119047562613476e-05, "loss": 0.2941, "step": 2295 }, { "epoch": 1.0043668122270741, "grad_norm": 0.39878675388447493, "learning_rate": 4.207270640993561e-05, "loss": 0.2787, "step": 2300 }, { "epoch": 1.0065502183406114, "grad_norm": 0.346225297182408, "learning_rate": 4.202625506694911e-05, "loss": 0.2686, "step": 2305 }, { "epoch": 1.0087336244541485, "grad_norm": 0.34208833101355196, "learning_rate": 4.197969383345925e-05, "loss": 0.2809, "step": 2310 }, { "epoch": 1.0109170305676856, "grad_norm": 0.4002016670275176, "learning_rate": 4.19330230099806e-05, "loss": 0.2979, "step": 2315 }, { "epoch": 1.0131004366812226, "grad_norm": 0.4630743803988168, "learning_rate": 4.188624289773501e-05, "loss": 0.2967, "step": 2320 }, { "epoch": 1.01528384279476, "grad_norm": 0.3568696275266243, "learning_rate": 4.1839353798649717e-05, "loss": 0.2758, "step": 2325 }, { "epoch": 1.017467248908297, "grad_norm": 0.3119249400022314, "learning_rate": 4.1792356015355375e-05, "loss": 0.2766, "step": 2330 }, { "epoch": 1.019650655021834, "grad_norm": 0.3696696936235827, "learning_rate": 4.17452498511841e-05, "loss": 0.286, "step": 2335 }, { "epoch": 1.0218340611353711, "grad_norm": 0.3084313639844132, "learning_rate": 4.169803561016754e-05, "loss": 0.2856, "step": 2340 }, { "epoch": 1.0240174672489082, "grad_norm": 0.3909495635210935, "learning_rate": 4.1650713597034845e-05, "loss": 0.287, "step": 2345 }, { "epoch": 1.0262008733624455, "grad_norm": 0.44840025564529157, "learning_rate": 4.16032841172108e-05, "loss": 0.2927, "step": 2350 }, { "epoch": 1.0283842794759825, "grad_norm": 0.41340997676020685, "learning_rate": 4.155574747681375e-05, "loss": 0.2858, "step": 2355 }, { "epoch": 1.0305676855895196, "grad_norm": 0.3994213300123044, "learning_rate": 4.1508103982653715e-05, "loss": 0.2991, "step": 2360 }, { "epoch": 1.0327510917030567, "grad_norm": 0.33838520158365676, "learning_rate": 4.146035394223034e-05, "loss": 0.2776, "step": 2365 }, { "epoch": 1.034934497816594, "grad_norm": 0.3368322156288848, "learning_rate": 4.1412497663730954e-05, "loss": 0.2748, "step": 2370 }, { "epoch": 1.037117903930131, "grad_norm": 0.3141352669129108, "learning_rate": 4.1364535456028555e-05, "loss": 0.2841, "step": 2375 }, { "epoch": 1.039301310043668, "grad_norm": 0.2818967412459555, "learning_rate": 4.1316467628679834e-05, "loss": 0.2619, "step": 2380 }, { "epoch": 1.0414847161572052, "grad_norm": 0.28565982519466604, "learning_rate": 4.126829449192316e-05, "loss": 0.2857, "step": 2385 }, { "epoch": 1.0436681222707425, "grad_norm": 0.3210535970776655, "learning_rate": 4.1220016356676615e-05, "loss": 0.2787, "step": 2390 }, { "epoch": 1.0458515283842795, "grad_norm": 0.3919954733516357, "learning_rate": 4.117163353453593e-05, "loss": 0.2824, "step": 2395 }, { "epoch": 1.0480349344978166, "grad_norm": 0.2781986797396297, "learning_rate": 4.112314633777251e-05, "loss": 0.2831, "step": 2400 }, { "epoch": 1.0502183406113537, "grad_norm": 0.3117143962813532, "learning_rate": 4.107455507933144e-05, "loss": 0.2741, "step": 2405 }, { "epoch": 1.0524017467248907, "grad_norm": 0.29749854620448457, "learning_rate": 4.102586007282938e-05, "loss": 0.2784, "step": 2410 }, { "epoch": 1.054585152838428, "grad_norm": 0.3011787067490516, "learning_rate": 4.0977061632552656e-05, "loss": 0.2822, "step": 2415 }, { "epoch": 1.056768558951965, "grad_norm": 0.27733319238652004, "learning_rate": 4.092816007345515e-05, "loss": 0.2918, "step": 2420 }, { "epoch": 1.0589519650655022, "grad_norm": 0.3178714706734469, "learning_rate": 4.087915571115629e-05, "loss": 0.2657, "step": 2425 }, { "epoch": 1.0611353711790392, "grad_norm": 0.33147612300712626, "learning_rate": 4.083004886193902e-05, "loss": 0.2821, "step": 2430 }, { "epoch": 1.0633187772925765, "grad_norm": 0.32015236720081114, "learning_rate": 4.0780839842747754e-05, "loss": 0.2876, "step": 2435 }, { "epoch": 1.0655021834061136, "grad_norm": 0.2637393126437815, "learning_rate": 4.073152897118632e-05, "loss": 0.2902, "step": 2440 }, { "epoch": 1.0676855895196506, "grad_norm": 0.2971292010190096, "learning_rate": 4.0682116565515935e-05, "loss": 0.294, "step": 2445 }, { "epoch": 1.0698689956331877, "grad_norm": 0.32151195135466043, "learning_rate": 4.063260294465311e-05, "loss": 0.2754, "step": 2450 }, { "epoch": 1.072052401746725, "grad_norm": 0.29344847997392354, "learning_rate": 4.058298842816765e-05, "loss": 0.2756, "step": 2455 }, { "epoch": 1.074235807860262, "grad_norm": 0.38141464593456376, "learning_rate": 4.0533273336280516e-05, "loss": 0.2749, "step": 2460 }, { "epoch": 1.0764192139737991, "grad_norm": 0.3519071752481233, "learning_rate": 4.048345798986186e-05, "loss": 0.2937, "step": 2465 }, { "epoch": 1.0786026200873362, "grad_norm": 0.27974724944596935, "learning_rate": 4.043354271042884e-05, "loss": 0.2772, "step": 2470 }, { "epoch": 1.0807860262008733, "grad_norm": 0.31883200998020006, "learning_rate": 4.038352782014362e-05, "loss": 0.2783, "step": 2475 }, { "epoch": 1.0829694323144106, "grad_norm": 0.3107436150415206, "learning_rate": 4.033341364181128e-05, "loss": 0.2845, "step": 2480 }, { "epoch": 1.0851528384279476, "grad_norm": 0.3631114446623769, "learning_rate": 4.028320049887771e-05, "loss": 0.2815, "step": 2485 }, { "epoch": 1.0873362445414847, "grad_norm": 0.29440410304916975, "learning_rate": 4.023288871542752e-05, "loss": 0.2767, "step": 2490 }, { "epoch": 1.0895196506550218, "grad_norm": 0.3013728586687223, "learning_rate": 4.0182478616182015e-05, "loss": 0.2802, "step": 2495 }, { "epoch": 1.091703056768559, "grad_norm": 0.30015696374357026, "learning_rate": 4.013197052649699e-05, "loss": 0.2885, "step": 2500 }, { "epoch": 1.0938864628820961, "grad_norm": 0.334054468972433, "learning_rate": 4.0081364772360725e-05, "loss": 0.2849, "step": 2505 }, { "epoch": 1.0960698689956332, "grad_norm": 0.29580484838617715, "learning_rate": 4.003066168039183e-05, "loss": 0.2748, "step": 2510 }, { "epoch": 1.0982532751091703, "grad_norm": 0.3454428161947495, "learning_rate": 3.9979861577837155e-05, "loss": 0.2783, "step": 2515 }, { "epoch": 1.1004366812227073, "grad_norm": 0.3354653145039301, "learning_rate": 3.9928964792569655e-05, "loss": 0.2811, "step": 2520 }, { "epoch": 1.1026200873362446, "grad_norm": 0.3257148621799395, "learning_rate": 3.987797165308633e-05, "loss": 0.2805, "step": 2525 }, { "epoch": 1.1048034934497817, "grad_norm": 0.3093405188202137, "learning_rate": 3.982688248850604e-05, "loss": 0.2929, "step": 2530 }, { "epoch": 1.1069868995633187, "grad_norm": 0.42822993107757057, "learning_rate": 3.977569762856741e-05, "loss": 0.2913, "step": 2535 }, { "epoch": 1.1091703056768558, "grad_norm": 0.40323824571841865, "learning_rate": 3.9724417403626705e-05, "loss": 0.2834, "step": 2540 }, { "epoch": 1.111353711790393, "grad_norm": 0.30813240614187476, "learning_rate": 3.9673042144655697e-05, "loss": 0.2927, "step": 2545 }, { "epoch": 1.1135371179039302, "grad_norm": 0.2807299601361123, "learning_rate": 3.9621572183239515e-05, "loss": 0.2778, "step": 2550 }, { "epoch": 1.1157205240174672, "grad_norm": 0.34539465437923245, "learning_rate": 3.957000785157452e-05, "loss": 0.3014, "step": 2555 }, { "epoch": 1.1179039301310043, "grad_norm": 0.36001154980798994, "learning_rate": 3.951834948246616e-05, "loss": 0.268, "step": 2560 }, { "epoch": 1.1200873362445414, "grad_norm": 0.3659504200165266, "learning_rate": 3.946659740932681e-05, "loss": 0.2999, "step": 2565 }, { "epoch": 1.1222707423580787, "grad_norm": 0.310758158636425, "learning_rate": 3.9414751966173634e-05, "loss": 0.2788, "step": 2570 }, { "epoch": 1.1244541484716157, "grad_norm": 0.2658948188578196, "learning_rate": 3.93628134876264e-05, "loss": 0.2764, "step": 2575 }, { "epoch": 1.1266375545851528, "grad_norm": 0.3038259778170334, "learning_rate": 3.9310782308905395e-05, "loss": 0.2809, "step": 2580 }, { "epoch": 1.1288209606986899, "grad_norm": 0.3098003775343629, "learning_rate": 3.925865876582915e-05, "loss": 0.2668, "step": 2585 }, { "epoch": 1.1310043668122272, "grad_norm": 0.29478313929574995, "learning_rate": 3.920644319481237e-05, "loss": 0.2762, "step": 2590 }, { "epoch": 1.1331877729257642, "grad_norm": 0.2715517904632468, "learning_rate": 3.915413593286371e-05, "loss": 0.2828, "step": 2595 }, { "epoch": 1.1353711790393013, "grad_norm": 0.2546601742913032, "learning_rate": 3.910173731758362e-05, "loss": 0.2752, "step": 2600 }, { "epoch": 1.1375545851528384, "grad_norm": 0.3020454024382658, "learning_rate": 3.904924768716216e-05, "loss": 0.2764, "step": 2605 }, { "epoch": 1.1397379912663754, "grad_norm": 0.3404890082601122, "learning_rate": 3.89966673803768e-05, "loss": 0.2746, "step": 2610 }, { "epoch": 1.1419213973799127, "grad_norm": 0.2918728991626841, "learning_rate": 3.894399673659031e-05, "loss": 0.2911, "step": 2615 }, { "epoch": 1.1441048034934498, "grad_norm": 0.32777823728250605, "learning_rate": 3.889123609574843e-05, "loss": 0.2687, "step": 2620 }, { "epoch": 1.1462882096069869, "grad_norm": 0.3417220224406648, "learning_rate": 3.883838579837781e-05, "loss": 0.29, "step": 2625 }, { "epoch": 1.1484716157205241, "grad_norm": 0.3206470787242044, "learning_rate": 3.8785446185583755e-05, "loss": 0.2852, "step": 2630 }, { "epoch": 1.1506550218340612, "grad_norm": 0.2896917429374818, "learning_rate": 3.873241759904801e-05, "loss": 0.2844, "step": 2635 }, { "epoch": 1.1528384279475983, "grad_norm": 0.319012549688518, "learning_rate": 3.8679300381026585e-05, "loss": 0.2926, "step": 2640 }, { "epoch": 1.1550218340611353, "grad_norm": 0.27491269743019364, "learning_rate": 3.862609487434752e-05, "loss": 0.2701, "step": 2645 }, { "epoch": 1.1572052401746724, "grad_norm": 0.30736358656559387, "learning_rate": 3.857280142240871e-05, "loss": 0.2813, "step": 2650 }, { "epoch": 1.1593886462882097, "grad_norm": 0.27847789277867213, "learning_rate": 3.851942036917565e-05, "loss": 0.2989, "step": 2655 }, { "epoch": 1.1615720524017468, "grad_norm": 0.3178443314858301, "learning_rate": 3.846595205917922e-05, "loss": 0.2831, "step": 2660 }, { "epoch": 1.1637554585152838, "grad_norm": 0.27407489870985063, "learning_rate": 3.8412396837513484e-05, "loss": 0.2838, "step": 2665 }, { "epoch": 1.165938864628821, "grad_norm": 0.31079163261067777, "learning_rate": 3.835875504983344e-05, "loss": 0.2896, "step": 2670 }, { "epoch": 1.1681222707423582, "grad_norm": 0.330388527773976, "learning_rate": 3.830502704235282e-05, "loss": 0.281, "step": 2675 }, { "epoch": 1.1703056768558953, "grad_norm": 0.28816900411883056, "learning_rate": 3.825121316184181e-05, "loss": 0.2883, "step": 2680 }, { "epoch": 1.1724890829694323, "grad_norm": 0.3328938276296137, "learning_rate": 3.8197313755624855e-05, "loss": 0.266, "step": 2685 }, { "epoch": 1.1746724890829694, "grad_norm": 0.3824721867696784, "learning_rate": 3.8143329171578366e-05, "loss": 0.2792, "step": 2690 }, { "epoch": 1.1768558951965065, "grad_norm": 0.33578112498653234, "learning_rate": 3.808925975812855e-05, "loss": 0.2994, "step": 2695 }, { "epoch": 1.1790393013100438, "grad_norm": 0.35312074931486553, "learning_rate": 3.803510586424908e-05, "loss": 0.2909, "step": 2700 }, { "epoch": 1.1812227074235808, "grad_norm": 0.3016960624575352, "learning_rate": 3.798086783945892e-05, "loss": 0.2829, "step": 2705 }, { "epoch": 1.1834061135371179, "grad_norm": 0.355359254971825, "learning_rate": 3.792654603381998e-05, "loss": 0.2843, "step": 2710 }, { "epoch": 1.185589519650655, "grad_norm": 0.3700674234618425, "learning_rate": 3.7872140797934956e-05, "loss": 0.2719, "step": 2715 }, { "epoch": 1.1877729257641922, "grad_norm": 0.44056476547643436, "learning_rate": 3.781765248294499e-05, "loss": 0.2759, "step": 2720 }, { "epoch": 1.1899563318777293, "grad_norm": 0.29079420170197845, "learning_rate": 3.7763081440527435e-05, "loss": 0.2857, "step": 2725 }, { "epoch": 1.1921397379912664, "grad_norm": 0.312707282955329, "learning_rate": 3.7708428022893596e-05, "loss": 0.2863, "step": 2730 }, { "epoch": 1.1943231441048034, "grad_norm": 0.2684008615845062, "learning_rate": 3.7653692582786416e-05, "loss": 0.272, "step": 2735 }, { "epoch": 1.1965065502183405, "grad_norm": 0.35782293675245397, "learning_rate": 3.7598875473478246e-05, "loss": 0.2715, "step": 2740 }, { "epoch": 1.1986899563318778, "grad_norm": 0.30301767808558216, "learning_rate": 3.7543977048768534e-05, "loss": 0.271, "step": 2745 }, { "epoch": 1.2008733624454149, "grad_norm": 0.3151321773465957, "learning_rate": 3.748899766298157e-05, "loss": 0.2845, "step": 2750 }, { "epoch": 1.203056768558952, "grad_norm": 0.2925859858510235, "learning_rate": 3.743393767096416e-05, "loss": 0.2755, "step": 2755 }, { "epoch": 1.205240174672489, "grad_norm": 0.2859805443710802, "learning_rate": 3.737879742808337e-05, "loss": 0.2793, "step": 2760 }, { "epoch": 1.2074235807860263, "grad_norm": 0.25461387164265176, "learning_rate": 3.732357729022421e-05, "loss": 0.2877, "step": 2765 }, { "epoch": 1.2096069868995634, "grad_norm": 0.24813193774999823, "learning_rate": 3.7268277613787356e-05, "loss": 0.2869, "step": 2770 }, { "epoch": 1.2117903930131004, "grad_norm": 0.3276586080798546, "learning_rate": 3.721289875568683e-05, "loss": 0.2853, "step": 2775 }, { "epoch": 1.2139737991266375, "grad_norm": 0.3351580255362412, "learning_rate": 3.715744107334773e-05, "loss": 0.2699, "step": 2780 }, { "epoch": 1.2161572052401746, "grad_norm": 0.30882004844208494, "learning_rate": 3.7101904924703865e-05, "loss": 0.2823, "step": 2785 }, { "epoch": 1.2183406113537119, "grad_norm": 0.37760317246667774, "learning_rate": 3.704629066819549e-05, "loss": 0.2771, "step": 2790 }, { "epoch": 1.220524017467249, "grad_norm": 0.32894843491571224, "learning_rate": 3.6990598662767014e-05, "loss": 0.2699, "step": 2795 }, { "epoch": 1.222707423580786, "grad_norm": 0.3476225470830814, "learning_rate": 3.693482926786461e-05, "loss": 0.2726, "step": 2800 }, { "epoch": 1.224890829694323, "grad_norm": 0.3118303300626587, "learning_rate": 3.687898284343396e-05, "loss": 0.2891, "step": 2805 }, { "epoch": 1.2270742358078603, "grad_norm": 0.3281515137713933, "learning_rate": 3.68230597499179e-05, "loss": 0.2805, "step": 2810 }, { "epoch": 1.2292576419213974, "grad_norm": 0.281506457906696, "learning_rate": 3.676706034825409e-05, "loss": 0.3019, "step": 2815 }, { "epoch": 1.2314410480349345, "grad_norm": 0.2918564930941491, "learning_rate": 3.6710984999872735e-05, "loss": 0.2748, "step": 2820 }, { "epoch": 1.2336244541484715, "grad_norm": 0.301158978005547, "learning_rate": 3.665483406669418e-05, "loss": 0.2904, "step": 2825 }, { "epoch": 1.2358078602620086, "grad_norm": 0.29117932725379103, "learning_rate": 3.6598607911126614e-05, "loss": 0.2856, "step": 2830 }, { "epoch": 1.237991266375546, "grad_norm": 0.2736978249597251, "learning_rate": 3.654230689606372e-05, "loss": 0.2659, "step": 2835 }, { "epoch": 1.240174672489083, "grad_norm": 0.2595195793059447, "learning_rate": 3.648593138488235e-05, "loss": 0.278, "step": 2840 }, { "epoch": 1.24235807860262, "grad_norm": 0.31138011895796025, "learning_rate": 3.6429481741440176e-05, "loss": 0.2804, "step": 2845 }, { "epoch": 1.244541484716157, "grad_norm": 0.28491002649715325, "learning_rate": 3.63729583300733e-05, "loss": 0.278, "step": 2850 }, { "epoch": 1.2467248908296944, "grad_norm": 0.28329885485362183, "learning_rate": 3.6316361515593966e-05, "loss": 0.2874, "step": 2855 }, { "epoch": 1.2489082969432315, "grad_norm": 0.36368480663196673, "learning_rate": 3.625969166328815e-05, "loss": 0.276, "step": 2860 }, { "epoch": 1.2510917030567685, "grad_norm": 0.2700235560170439, "learning_rate": 3.620294913891326e-05, "loss": 0.2884, "step": 2865 }, { "epoch": 1.2532751091703056, "grad_norm": 0.25418847496227326, "learning_rate": 3.614613430869571e-05, "loss": 0.2859, "step": 2870 }, { "epoch": 1.2554585152838427, "grad_norm": 0.28230477688636335, "learning_rate": 3.608924753932862e-05, "loss": 0.2884, "step": 2875 }, { "epoch": 1.25764192139738, "grad_norm": 0.31064202314287603, "learning_rate": 3.6032289197969374e-05, "loss": 0.2858, "step": 2880 }, { "epoch": 1.259825327510917, "grad_norm": 0.272199049103352, "learning_rate": 3.5975259652237333e-05, "loss": 0.287, "step": 2885 }, { "epoch": 1.262008733624454, "grad_norm": 0.23256991109619604, "learning_rate": 3.5918159270211424e-05, "loss": 0.2734, "step": 2890 }, { "epoch": 1.2641921397379914, "grad_norm": 0.2926902120837456, "learning_rate": 3.586098842042773e-05, "loss": 0.2774, "step": 2895 }, { "epoch": 1.2663755458515285, "grad_norm": 0.36624779554626996, "learning_rate": 3.580374747187719e-05, "loss": 0.2775, "step": 2900 }, { "epoch": 1.2685589519650655, "grad_norm": 0.29331268838279223, "learning_rate": 3.574643679400313e-05, "loss": 0.2747, "step": 2905 }, { "epoch": 1.2707423580786026, "grad_norm": 0.27621174524527253, "learning_rate": 3.568905675669895e-05, "loss": 0.2882, "step": 2910 }, { "epoch": 1.2729257641921397, "grad_norm": 0.2771703632616136, "learning_rate": 3.5631607730305694e-05, "loss": 0.2664, "step": 2915 }, { "epoch": 1.2751091703056767, "grad_norm": 0.33588301272251375, "learning_rate": 3.5574090085609696e-05, "loss": 0.2798, "step": 2920 }, { "epoch": 1.277292576419214, "grad_norm": 0.35009121663132425, "learning_rate": 3.551650419384013e-05, "loss": 0.2833, "step": 2925 }, { "epoch": 1.279475982532751, "grad_norm": 0.2687807965233062, "learning_rate": 3.545885042666669e-05, "loss": 0.2738, "step": 2930 }, { "epoch": 1.2816593886462881, "grad_norm": 0.36331931036035, "learning_rate": 3.5401129156197105e-05, "loss": 0.2833, "step": 2935 }, { "epoch": 1.2838427947598254, "grad_norm": 0.3828328306666377, "learning_rate": 3.534334075497482e-05, "loss": 0.2821, "step": 2940 }, { "epoch": 1.2860262008733625, "grad_norm": 0.30473362904993256, "learning_rate": 3.5285485595976534e-05, "loss": 0.2821, "step": 2945 }, { "epoch": 1.2882096069868996, "grad_norm": 0.3499861198156561, "learning_rate": 3.5227564052609816e-05, "loss": 0.2724, "step": 2950 }, { "epoch": 1.2903930131004366, "grad_norm": 0.2938833486098817, "learning_rate": 3.5169576498710696e-05, "loss": 0.278, "step": 2955 }, { "epoch": 1.2925764192139737, "grad_norm": 0.28084189095150464, "learning_rate": 3.511152330854125e-05, "loss": 0.2679, "step": 2960 }, { "epoch": 1.294759825327511, "grad_norm": 0.3199421686229126, "learning_rate": 3.505340485678717e-05, "loss": 0.2819, "step": 2965 }, { "epoch": 1.296943231441048, "grad_norm": 0.3149994117491474, "learning_rate": 3.499522151855537e-05, "loss": 0.2975, "step": 2970 }, { "epoch": 1.2991266375545851, "grad_norm": 0.2899693019311649, "learning_rate": 3.493697366937154e-05, "loss": 0.2706, "step": 2975 }, { "epoch": 1.3013100436681222, "grad_norm": 0.33913499909626604, "learning_rate": 3.487866168517775e-05, "loss": 0.294, "step": 2980 }, { "epoch": 1.3034934497816595, "grad_norm": 0.3769147657930795, "learning_rate": 3.4820285942330003e-05, "loss": 0.2818, "step": 2985 }, { "epoch": 1.3056768558951966, "grad_norm": 0.32801615796861155, "learning_rate": 3.476184681759581e-05, "loss": 0.28, "step": 2990 }, { "epoch": 1.3078602620087336, "grad_norm": 0.27357707367933753, "learning_rate": 3.4703344688151776e-05, "loss": 0.2783, "step": 2995 }, { "epoch": 1.3100436681222707, "grad_norm": 0.37622537438391995, "learning_rate": 3.46447799315811e-05, "loss": 0.2819, "step": 3000 }, { "epoch": 1.3122270742358078, "grad_norm": 0.3550543812495112, "learning_rate": 3.458615292587125e-05, "loss": 0.2792, "step": 3005 }, { "epoch": 1.314410480349345, "grad_norm": 0.2816496834529316, "learning_rate": 3.452746404941143e-05, "loss": 0.2859, "step": 3010 }, { "epoch": 1.3165938864628821, "grad_norm": 0.3066581463920424, "learning_rate": 3.446871368099017e-05, "loss": 0.2746, "step": 3015 }, { "epoch": 1.3187772925764192, "grad_norm": 0.2708116400301654, "learning_rate": 3.440990219979289e-05, "loss": 0.2863, "step": 3020 }, { "epoch": 1.3209606986899565, "grad_norm": 0.2497661453451677, "learning_rate": 3.435102998539942e-05, "loss": 0.2702, "step": 3025 }, { "epoch": 1.3231441048034935, "grad_norm": 0.2651767333352664, "learning_rate": 3.4292097417781595e-05, "loss": 0.2695, "step": 3030 }, { "epoch": 1.3253275109170306, "grad_norm": 0.2695245871370125, "learning_rate": 3.4233104877300775e-05, "loss": 0.267, "step": 3035 }, { "epoch": 1.3275109170305677, "grad_norm": 0.27077504903552757, "learning_rate": 3.41740527447054e-05, "loss": 0.2662, "step": 3040 }, { "epoch": 1.3296943231441047, "grad_norm": 0.2519617614699693, "learning_rate": 3.411494140112852e-05, "loss": 0.2716, "step": 3045 }, { "epoch": 1.3318777292576418, "grad_norm": 0.2809728902320245, "learning_rate": 3.4055771228085347e-05, "loss": 0.2879, "step": 3050 }, { "epoch": 1.334061135371179, "grad_norm": 0.26727497601257166, "learning_rate": 3.399654260747078e-05, "loss": 0.2725, "step": 3055 }, { "epoch": 1.3362445414847162, "grad_norm": 0.34198746633106286, "learning_rate": 3.3937255921556974e-05, "loss": 0.2877, "step": 3060 }, { "epoch": 1.3384279475982532, "grad_norm": 0.2984040980470414, "learning_rate": 3.3877911552990826e-05, "loss": 0.2793, "step": 3065 }, { "epoch": 1.3406113537117905, "grad_norm": 0.2865962833130215, "learning_rate": 3.3818509884791514e-05, "loss": 0.273, "step": 3070 }, { "epoch": 1.3427947598253276, "grad_norm": 0.2988181873287166, "learning_rate": 3.375905130034808e-05, "loss": 0.275, "step": 3075 }, { "epoch": 1.3449781659388647, "grad_norm": 0.29108551078775924, "learning_rate": 3.369953618341688e-05, "loss": 0.2874, "step": 3080 }, { "epoch": 1.3471615720524017, "grad_norm": 0.29395564574021693, "learning_rate": 3.3639964918119135e-05, "loss": 0.2756, "step": 3085 }, { "epoch": 1.3493449781659388, "grad_norm": 0.2834037344232317, "learning_rate": 3.3580337888938515e-05, "loss": 0.2795, "step": 3090 }, { "epoch": 1.3515283842794759, "grad_norm": 0.2963729779205327, "learning_rate": 3.3520655480718515e-05, "loss": 0.2821, "step": 3095 }, { "epoch": 1.3537117903930131, "grad_norm": 0.2719250597184558, "learning_rate": 3.346091807866013e-05, "loss": 0.2799, "step": 3100 }, { "epoch": 1.3558951965065502, "grad_norm": 0.345480183079055, "learning_rate": 3.3401126068319244e-05, "loss": 0.2753, "step": 3105 }, { "epoch": 1.3580786026200873, "grad_norm": 0.2731322103388952, "learning_rate": 3.3341279835604224e-05, "loss": 0.2806, "step": 3110 }, { "epoch": 1.3602620087336246, "grad_norm": 0.3377724432178437, "learning_rate": 3.328137976677339e-05, "loss": 0.2952, "step": 3115 }, { "epoch": 1.3624454148471616, "grad_norm": 0.3490702531757625, "learning_rate": 3.322142624843253e-05, "loss": 0.2883, "step": 3120 }, { "epoch": 1.3646288209606987, "grad_norm": 0.23733767943987039, "learning_rate": 3.3161419667532384e-05, "loss": 0.2811, "step": 3125 }, { "epoch": 1.3668122270742358, "grad_norm": 0.29640032968995195, "learning_rate": 3.3101360411366205e-05, "loss": 0.2714, "step": 3130 }, { "epoch": 1.3689956331877728, "grad_norm": 0.38335436641971393, "learning_rate": 3.304124886756719e-05, "loss": 0.2773, "step": 3135 }, { "epoch": 1.37117903930131, "grad_norm": 0.2614713678332863, "learning_rate": 3.2981085424106036e-05, "loss": 0.2716, "step": 3140 }, { "epoch": 1.3733624454148472, "grad_norm": 0.271556464815005, "learning_rate": 3.2920870469288375e-05, "loss": 0.2929, "step": 3145 }, { "epoch": 1.3755458515283843, "grad_norm": 0.2984513438432199, "learning_rate": 3.2860604391752344e-05, "loss": 0.2767, "step": 3150 }, { "epoch": 1.3777292576419213, "grad_norm": 0.35226423449647926, "learning_rate": 3.2800287580466e-05, "loss": 0.2803, "step": 3155 }, { "epoch": 1.3799126637554586, "grad_norm": 0.2762721435202866, "learning_rate": 3.273992042472487e-05, "loss": 0.2635, "step": 3160 }, { "epoch": 1.3820960698689957, "grad_norm": 0.2849399596588635, "learning_rate": 3.2679503314149395e-05, "loss": 0.2828, "step": 3165 }, { "epoch": 1.3842794759825328, "grad_norm": 0.3240823060545034, "learning_rate": 3.261903663868243e-05, "loss": 0.2772, "step": 3170 }, { "epoch": 1.3864628820960698, "grad_norm": 0.29529588632218107, "learning_rate": 3.2558520788586764e-05, "loss": 0.2709, "step": 3175 }, { "epoch": 1.388646288209607, "grad_norm": 0.2880598231126244, "learning_rate": 3.249795615444253e-05, "loss": 0.2649, "step": 3180 }, { "epoch": 1.3908296943231442, "grad_norm": 0.34085771304801965, "learning_rate": 3.243734312714475e-05, "loss": 0.2767, "step": 3185 }, { "epoch": 1.3930131004366813, "grad_norm": 0.24864813820746293, "learning_rate": 3.2376682097900754e-05, "loss": 0.2793, "step": 3190 }, { "epoch": 1.3951965065502183, "grad_norm": 0.3059620331889871, "learning_rate": 3.2315973458227714e-05, "loss": 0.2875, "step": 3195 }, { "epoch": 1.3973799126637554, "grad_norm": 0.24526372956599277, "learning_rate": 3.225521759995007e-05, "loss": 0.2821, "step": 3200 }, { "epoch": 1.3995633187772927, "grad_norm": 0.2972629377196201, "learning_rate": 3.219441491519703e-05, "loss": 0.2742, "step": 3205 }, { "epoch": 1.4017467248908297, "grad_norm": 0.30368222569014375, "learning_rate": 3.2133565796400046e-05, "loss": 0.2794, "step": 3210 }, { "epoch": 1.4039301310043668, "grad_norm": 0.27087580419910157, "learning_rate": 3.2072670636290195e-05, "loss": 0.2686, "step": 3215 }, { "epoch": 1.4061135371179039, "grad_norm": 0.3097415372011331, "learning_rate": 3.20117298278958e-05, "loss": 0.2744, "step": 3220 }, { "epoch": 1.408296943231441, "grad_norm": 0.265200535542659, "learning_rate": 3.195074376453975e-05, "loss": 0.2904, "step": 3225 }, { "epoch": 1.4104803493449782, "grad_norm": 0.26447053395871756, "learning_rate": 3.188971283983703e-05, "loss": 0.2633, "step": 3230 }, { "epoch": 1.4126637554585153, "grad_norm": 0.25630242114154234, "learning_rate": 3.1828637447692187e-05, "loss": 0.2755, "step": 3235 }, { "epoch": 1.4148471615720524, "grad_norm": 0.26664274806120797, "learning_rate": 3.176751798229673e-05, "loss": 0.2802, "step": 3240 }, { "epoch": 1.4170305676855894, "grad_norm": 0.3278108102940346, "learning_rate": 3.1706354838126687e-05, "loss": 0.2585, "step": 3245 }, { "epoch": 1.4192139737991267, "grad_norm": 0.24680968950816637, "learning_rate": 3.1645148409939915e-05, "loss": 0.2758, "step": 3250 }, { "epoch": 1.4213973799126638, "grad_norm": 0.2843693481761553, "learning_rate": 3.158389909277371e-05, "loss": 0.2857, "step": 3255 }, { "epoch": 1.4235807860262009, "grad_norm": 0.27926774884591915, "learning_rate": 3.152260728194213e-05, "loss": 0.2827, "step": 3260 }, { "epoch": 1.425764192139738, "grad_norm": 0.33114453429263857, "learning_rate": 3.1461273373033516e-05, "loss": 0.2791, "step": 3265 }, { "epoch": 1.427947598253275, "grad_norm": 0.2523111291701471, "learning_rate": 3.1399897761907914e-05, "loss": 0.2785, "step": 3270 }, { "epoch": 1.4301310043668123, "grad_norm": 0.2669982957777147, "learning_rate": 3.133848084469454e-05, "loss": 0.2723, "step": 3275 }, { "epoch": 1.4323144104803494, "grad_norm": 0.2551505338781039, "learning_rate": 3.1277023017789164e-05, "loss": 0.2701, "step": 3280 }, { "epoch": 1.4344978165938864, "grad_norm": 0.2729328232441411, "learning_rate": 3.121552467785164e-05, "loss": 0.2756, "step": 3285 }, { "epoch": 1.4366812227074237, "grad_norm": 0.31413945020452155, "learning_rate": 3.1153986221803284e-05, "loss": 0.2634, "step": 3290 }, { "epoch": 1.4388646288209608, "grad_norm": 0.27459622465012024, "learning_rate": 3.109240804682431e-05, "loss": 0.2749, "step": 3295 }, { "epoch": 1.4410480349344978, "grad_norm": 0.3428676752152493, "learning_rate": 3.103079055035131e-05, "loss": 0.2737, "step": 3300 }, { "epoch": 1.443231441048035, "grad_norm": 0.28826863252341717, "learning_rate": 3.096913413007467e-05, "loss": 0.2716, "step": 3305 }, { "epoch": 1.445414847161572, "grad_norm": 0.3003661345160089, "learning_rate": 3.090743918393596e-05, "loss": 0.2711, "step": 3310 }, { "epoch": 1.447598253275109, "grad_norm": 0.2726421109267964, "learning_rate": 3.0845706110125435e-05, "loss": 0.2755, "step": 3315 }, { "epoch": 1.4497816593886463, "grad_norm": 0.2633565745117208, "learning_rate": 3.078393530707942e-05, "loss": 0.2708, "step": 3320 }, { "epoch": 1.4519650655021834, "grad_norm": 0.3590173494625973, "learning_rate": 3.072212717347776e-05, "loss": 0.2802, "step": 3325 }, { "epoch": 1.4541484716157205, "grad_norm": 0.2841165489479175, "learning_rate": 3.0660282108241235e-05, "loss": 0.2793, "step": 3330 }, { "epoch": 1.4563318777292578, "grad_norm": 0.27787135238996175, "learning_rate": 3.059840051052898e-05, "loss": 0.2915, "step": 3335 }, { "epoch": 1.4585152838427948, "grad_norm": 0.29883394182548156, "learning_rate": 3.053648277973592e-05, "loss": 0.2667, "step": 3340 }, { "epoch": 1.460698689956332, "grad_norm": 0.2897989278866621, "learning_rate": 3.04745293154902e-05, "loss": 0.2863, "step": 3345 }, { "epoch": 1.462882096069869, "grad_norm": 0.2413426041774491, "learning_rate": 3.0412540517650582e-05, "loss": 0.2686, "step": 3350 }, { "epoch": 1.465065502183406, "grad_norm": 0.3092426474858127, "learning_rate": 3.03505167863039e-05, "loss": 0.2861, "step": 3355 }, { "epoch": 1.467248908296943, "grad_norm": 0.37271590450228287, "learning_rate": 3.0288458521762414e-05, "loss": 0.2804, "step": 3360 }, { "epoch": 1.4694323144104804, "grad_norm": 0.32074172110600374, "learning_rate": 3.02263661245613e-05, "loss": 0.2673, "step": 3365 }, { "epoch": 1.4716157205240175, "grad_norm": 0.26878027230807305, "learning_rate": 3.016423999545603e-05, "loss": 0.2805, "step": 3370 }, { "epoch": 1.4737991266375545, "grad_norm": 0.2611222636406977, "learning_rate": 3.0102080535419775e-05, "loss": 0.2852, "step": 3375 }, { "epoch": 1.4759825327510918, "grad_norm": 0.32711140245362424, "learning_rate": 3.0039888145640845e-05, "loss": 0.2768, "step": 3380 }, { "epoch": 1.4781659388646289, "grad_norm": 0.2933942315221079, "learning_rate": 2.9977663227520065e-05, "loss": 0.2733, "step": 3385 }, { "epoch": 1.480349344978166, "grad_norm": 0.3239911145447236, "learning_rate": 2.991540618266822e-05, "loss": 0.2689, "step": 3390 }, { "epoch": 1.482532751091703, "grad_norm": 0.31053885049245833, "learning_rate": 2.9853117412903443e-05, "loss": 0.259, "step": 3395 }, { "epoch": 1.48471615720524, "grad_norm": 0.29558451827845905, "learning_rate": 2.979079732024862e-05, "loss": 0.2676, "step": 3400 }, { "epoch": 1.4868995633187774, "grad_norm": 0.2919836637875955, "learning_rate": 2.9728446306928813e-05, "loss": 0.274, "step": 3405 }, { "epoch": 1.4890829694323144, "grad_norm": 0.2914986515005281, "learning_rate": 2.966606477536863e-05, "loss": 0.265, "step": 3410 }, { "epoch": 1.4912663755458515, "grad_norm": 0.24516343045152553, "learning_rate": 2.960365312818967e-05, "loss": 0.2745, "step": 3415 }, { "epoch": 1.4934497816593886, "grad_norm": 0.31298311253304867, "learning_rate": 2.9541211768207883e-05, "loss": 0.2718, "step": 3420 }, { "epoch": 1.4956331877729259, "grad_norm": 0.3182623875469872, "learning_rate": 2.9478741098431027e-05, "loss": 0.2747, "step": 3425 }, { "epoch": 1.497816593886463, "grad_norm": 0.31623459423577244, "learning_rate": 2.9416241522055987e-05, "loss": 0.3007, "step": 3430 }, { "epoch": 1.5, "grad_norm": 0.3659591897091837, "learning_rate": 2.9353713442466246e-05, "loss": 0.2797, "step": 3435 }, { "epoch": 1.502183406113537, "grad_norm": 0.29892370752426095, "learning_rate": 2.929115726322924e-05, "loss": 0.2711, "step": 3440 }, { "epoch": 1.5043668122270741, "grad_norm": 0.36468217436567685, "learning_rate": 2.9228573388093765e-05, "loss": 0.2781, "step": 3445 }, { "epoch": 1.5065502183406112, "grad_norm": 0.36523975525457747, "learning_rate": 2.916596222098738e-05, "loss": 0.2782, "step": 3450 }, { "epoch": 1.5087336244541485, "grad_norm": 0.24799351938184713, "learning_rate": 2.9103324166013777e-05, "loss": 0.2723, "step": 3455 }, { "epoch": 1.5109170305676856, "grad_norm": 0.3395645662027844, "learning_rate": 2.9040659627450195e-05, "loss": 0.2734, "step": 3460 }, { "epoch": 1.5131004366812228, "grad_norm": 0.2312365794117943, "learning_rate": 2.897796900974481e-05, "loss": 0.2695, "step": 3465 }, { "epoch": 1.51528384279476, "grad_norm": 0.2733485343314042, "learning_rate": 2.8915252717514108e-05, "loss": 0.2753, "step": 3470 }, { "epoch": 1.517467248908297, "grad_norm": 0.30853988506735613, "learning_rate": 2.8852511155540286e-05, "loss": 0.2805, "step": 3475 }, { "epoch": 1.519650655021834, "grad_norm": 0.298316972177523, "learning_rate": 2.8789744728768632e-05, "loss": 0.2674, "step": 3480 }, { "epoch": 1.5218340611353711, "grad_norm": 0.33489826338103257, "learning_rate": 2.8726953842304917e-05, "loss": 0.2781, "step": 3485 }, { "epoch": 1.5240174672489082, "grad_norm": 0.29180816733941406, "learning_rate": 2.8664138901412784e-05, "loss": 0.2828, "step": 3490 }, { "epoch": 1.5262008733624453, "grad_norm": 0.2652928534046041, "learning_rate": 2.8601300311511135e-05, "loss": 0.2704, "step": 3495 }, { "epoch": 1.5283842794759825, "grad_norm": 0.2478372696723595, "learning_rate": 2.8538438478171493e-05, "loss": 0.2709, "step": 3500 }, { "epoch": 1.5305676855895196, "grad_norm": 0.29590356196390866, "learning_rate": 2.8475553807115392e-05, "loss": 0.2645, "step": 3505 }, { "epoch": 1.532751091703057, "grad_norm": 0.2404106523342164, "learning_rate": 2.8412646704211787e-05, "loss": 0.2716, "step": 3510 }, { "epoch": 1.534934497816594, "grad_norm": 0.2913911540734402, "learning_rate": 2.8349717575474395e-05, "loss": 0.2812, "step": 3515 }, { "epoch": 1.537117903930131, "grad_norm": 0.27619563545275305, "learning_rate": 2.8286766827059103e-05, "loss": 0.2852, "step": 3520 }, { "epoch": 1.539301310043668, "grad_norm": 0.25296146919214546, "learning_rate": 2.8223794865261326e-05, "loss": 0.2706, "step": 3525 }, { "epoch": 1.5414847161572052, "grad_norm": 0.4120208095199771, "learning_rate": 2.8160802096513396e-05, "loss": 0.2655, "step": 3530 }, { "epoch": 1.5436681222707422, "grad_norm": 0.2797662392218346, "learning_rate": 2.809778892738194e-05, "loss": 0.2799, "step": 3535 }, { "epoch": 1.5458515283842795, "grad_norm": 0.30491897219029096, "learning_rate": 2.8034755764565252e-05, "loss": 0.2703, "step": 3540 }, { "epoch": 1.5480349344978166, "grad_norm": 0.2830893746174212, "learning_rate": 2.797170301489066e-05, "loss": 0.2716, "step": 3545 }, { "epoch": 1.5502183406113537, "grad_norm": 0.23485142267745251, "learning_rate": 2.7908631085311933e-05, "loss": 0.2617, "step": 3550 }, { "epoch": 1.552401746724891, "grad_norm": 0.3021327029538882, "learning_rate": 2.78455403829066e-05, "loss": 0.2638, "step": 3555 }, { "epoch": 1.554585152838428, "grad_norm": 0.2621866662981608, "learning_rate": 2.7782431314873377e-05, "loss": 0.2678, "step": 3560 }, { "epoch": 1.556768558951965, "grad_norm": 0.2747099055332518, "learning_rate": 2.7719304288529503e-05, "loss": 0.2714, "step": 3565 }, { "epoch": 1.5589519650655022, "grad_norm": 0.27602610129495325, "learning_rate": 2.7656159711308127e-05, "loss": 0.2754, "step": 3570 }, { "epoch": 1.5611353711790392, "grad_norm": 0.2563630850581704, "learning_rate": 2.7592997990755665e-05, "loss": 0.2674, "step": 3575 }, { "epoch": 1.5633187772925763, "grad_norm": 0.25276036483666414, "learning_rate": 2.75298195345292e-05, "loss": 0.2689, "step": 3580 }, { "epoch": 1.5655021834061136, "grad_norm": 0.2568576656221897, "learning_rate": 2.7466624750393804e-05, "loss": 0.2624, "step": 3585 }, { "epoch": 1.5676855895196506, "grad_norm": 0.2540380161679821, "learning_rate": 2.740341404621996e-05, "loss": 0.2704, "step": 3590 }, { "epoch": 1.569868995633188, "grad_norm": 0.2691098927474828, "learning_rate": 2.7340187829980884e-05, "loss": 0.2596, "step": 3595 }, { "epoch": 1.572052401746725, "grad_norm": 0.27708252166503045, "learning_rate": 2.7276946509749905e-05, "loss": 0.2753, "step": 3600 }, { "epoch": 1.574235807860262, "grad_norm": 0.2828335584788292, "learning_rate": 2.7213690493697852e-05, "loss": 0.2813, "step": 3605 }, { "epoch": 1.5764192139737991, "grad_norm": 0.29157807173338984, "learning_rate": 2.7150420190090397e-05, "loss": 0.2654, "step": 3610 }, { "epoch": 1.5786026200873362, "grad_norm": 0.2621619308124079, "learning_rate": 2.7087136007285417e-05, "loss": 0.269, "step": 3615 }, { "epoch": 1.5807860262008733, "grad_norm": 0.27055849925403364, "learning_rate": 2.7023838353730386e-05, "loss": 0.2603, "step": 3620 }, { "epoch": 1.5829694323144103, "grad_norm": 0.28812667281668847, "learning_rate": 2.6960527637959704e-05, "loss": 0.2615, "step": 3625 }, { "epoch": 1.5851528384279476, "grad_norm": 0.3648305241785186, "learning_rate": 2.689720426859208e-05, "loss": 0.2745, "step": 3630 }, { "epoch": 1.5873362445414847, "grad_norm": 0.3579531717893759, "learning_rate": 2.68338686543279e-05, "loss": 0.2754, "step": 3635 }, { "epoch": 1.589519650655022, "grad_norm": 0.30496840176534284, "learning_rate": 2.6770521203946582e-05, "loss": 0.2687, "step": 3640 }, { "epoch": 1.591703056768559, "grad_norm": 0.3316407097346181, "learning_rate": 2.670716232630392e-05, "loss": 0.2708, "step": 3645 }, { "epoch": 1.5938864628820961, "grad_norm": 0.30884262120013556, "learning_rate": 2.6643792430329473e-05, "loss": 0.2691, "step": 3650 }, { "epoch": 1.5960698689956332, "grad_norm": 0.24575164955804643, "learning_rate": 2.6580411925023916e-05, "loss": 0.2895, "step": 3655 }, { "epoch": 1.5982532751091703, "grad_norm": 0.2403162634268798, "learning_rate": 2.6517021219456402e-05, "loss": 0.2768, "step": 3660 }, { "epoch": 1.6004366812227073, "grad_norm": 0.7286289540628894, "learning_rate": 2.6453620722761896e-05, "loss": 0.2772, "step": 3665 }, { "epoch": 1.6026200873362444, "grad_norm": 0.2698469031360812, "learning_rate": 2.6390210844138593e-05, "loss": 0.2772, "step": 3670 }, { "epoch": 1.6048034934497817, "grad_norm": 0.29594772177235845, "learning_rate": 2.6326791992845202e-05, "loss": 0.2746, "step": 3675 }, { "epoch": 1.6069868995633187, "grad_norm": 0.3432286485907405, "learning_rate": 2.626336457819837e-05, "loss": 0.2627, "step": 3680 }, { "epoch": 1.609170305676856, "grad_norm": 0.47848404072485456, "learning_rate": 2.619992900957e-05, "loss": 0.2896, "step": 3685 }, { "epoch": 1.611353711790393, "grad_norm": 0.400345398790263, "learning_rate": 2.613648569638464e-05, "loss": 0.2832, "step": 3690 }, { "epoch": 1.6135371179039302, "grad_norm": 0.26168308925039074, "learning_rate": 2.6073035048116796e-05, "loss": 0.2693, "step": 3695 }, { "epoch": 1.6157205240174672, "grad_norm": 0.33192980592740756, "learning_rate": 2.6009577474288333e-05, "loss": 0.2759, "step": 3700 }, { "epoch": 1.6179039301310043, "grad_norm": 0.2787526637545283, "learning_rate": 2.5946113384465813e-05, "loss": 0.2642, "step": 3705 }, { "epoch": 1.6200873362445414, "grad_norm": 0.27338870781942926, "learning_rate": 2.5882643188257856e-05, "loss": 0.2706, "step": 3710 }, { "epoch": 1.6222707423580784, "grad_norm": 0.2435797974873236, "learning_rate": 2.5819167295312485e-05, "loss": 0.2724, "step": 3715 }, { "epoch": 1.6244541484716157, "grad_norm": 0.29781182108189724, "learning_rate": 2.57556861153145e-05, "loss": 0.2707, "step": 3720 }, { "epoch": 1.6266375545851528, "grad_norm": 0.32478484905425753, "learning_rate": 2.569220005798281e-05, "loss": 0.2725, "step": 3725 }, { "epoch": 1.62882096069869, "grad_norm": 0.2669336390561608, "learning_rate": 2.5628709533067824e-05, "loss": 0.2713, "step": 3730 }, { "epoch": 1.6310043668122272, "grad_norm": 0.2573882639282421, "learning_rate": 2.556521495034877e-05, "loss": 0.2606, "step": 3735 }, { "epoch": 1.6331877729257642, "grad_norm": 0.2705880029160448, "learning_rate": 2.550171671963108e-05, "loss": 0.2575, "step": 3740 }, { "epoch": 1.6353711790393013, "grad_norm": 0.27796104371371233, "learning_rate": 2.543821525074371e-05, "loss": 0.2699, "step": 3745 }, { "epoch": 1.6375545851528384, "grad_norm": 0.2968772131918911, "learning_rate": 2.537471095353653e-05, "loss": 0.2604, "step": 3750 }, { "epoch": 1.6397379912663754, "grad_norm": 0.30473359891717344, "learning_rate": 2.5311204237877657e-05, "loss": 0.2715, "step": 3755 }, { "epoch": 1.6419213973799127, "grad_norm": 0.29986937512458306, "learning_rate": 2.524769551365083e-05, "loss": 0.2569, "step": 3760 }, { "epoch": 1.6441048034934498, "grad_norm": 0.29756570977438257, "learning_rate": 2.5184185190752752e-05, "loss": 0.2805, "step": 3765 }, { "epoch": 1.6462882096069869, "grad_norm": 0.24764879064376663, "learning_rate": 2.512067367909041e-05, "loss": 0.2723, "step": 3770 }, { "epoch": 1.6484716157205241, "grad_norm": 0.2536796536666786, "learning_rate": 2.505716138857851e-05, "loss": 0.2692, "step": 3775 }, { "epoch": 1.6506550218340612, "grad_norm": 0.27985489044371364, "learning_rate": 2.499364872913676e-05, "loss": 0.2736, "step": 3780 }, { "epoch": 1.6528384279475983, "grad_norm": 0.24159737014671576, "learning_rate": 2.4930136110687246e-05, "loss": 0.2711, "step": 3785 }, { "epoch": 1.6550218340611353, "grad_norm": 0.28234624412572884, "learning_rate": 2.486662394315181e-05, "loss": 0.2865, "step": 3790 }, { "epoch": 1.6572052401746724, "grad_norm": 0.2491348077927823, "learning_rate": 2.4803112636449355e-05, "loss": 0.2731, "step": 3795 }, { "epoch": 1.6593886462882095, "grad_norm": 0.31780593847271627, "learning_rate": 2.473960260049325e-05, "loss": 0.2789, "step": 3800 }, { "epoch": 1.6615720524017468, "grad_norm": 0.2685872565371744, "learning_rate": 2.4676094245188658e-05, "loss": 0.2933, "step": 3805 }, { "epoch": 1.6637554585152838, "grad_norm": 0.263863501194976, "learning_rate": 2.461258798042989e-05, "loss": 0.2685, "step": 3810 }, { "epoch": 1.6659388646288211, "grad_norm": 0.21766584856793741, "learning_rate": 2.454908421609777e-05, "loss": 0.2552, "step": 3815 }, { "epoch": 1.6681222707423582, "grad_norm": 0.27428707449471423, "learning_rate": 2.4485583362056974e-05, "loss": 0.2786, "step": 3820 }, { "epoch": 1.6703056768558953, "grad_norm": 0.2835969591758734, "learning_rate": 2.4422085828153403e-05, "loss": 0.2662, "step": 3825 }, { "epoch": 1.6724890829694323, "grad_norm": 0.29526168010306597, "learning_rate": 2.435859202421153e-05, "loss": 0.269, "step": 3830 }, { "epoch": 1.6746724890829694, "grad_norm": 0.30337514206743876, "learning_rate": 2.429510236003175e-05, "loss": 0.2597, "step": 3835 }, { "epoch": 1.6768558951965065, "grad_norm": 0.2584878592869018, "learning_rate": 2.4231617245387745e-05, "loss": 0.2709, "step": 3840 }, { "epoch": 1.6790393013100435, "grad_norm": 0.27852560485634303, "learning_rate": 2.416813709002382e-05, "loss": 0.2712, "step": 3845 }, { "epoch": 1.6812227074235808, "grad_norm": 0.31941354537015787, "learning_rate": 2.4104662303652288e-05, "loss": 0.286, "step": 3850 }, { "epoch": 1.6834061135371179, "grad_norm": 0.2639262509863191, "learning_rate": 2.4041193295950788e-05, "loss": 0.2679, "step": 3855 }, { "epoch": 1.6855895196506552, "grad_norm": 0.27648769301544285, "learning_rate": 2.39777304765597e-05, "loss": 0.2764, "step": 3860 }, { "epoch": 1.6877729257641922, "grad_norm": 0.2557231698124579, "learning_rate": 2.391427425507943e-05, "loss": 0.2773, "step": 3865 }, { "epoch": 1.6899563318777293, "grad_norm": 0.2835965272930869, "learning_rate": 2.385082504106781e-05, "loss": 0.2655, "step": 3870 }, { "epoch": 1.6921397379912664, "grad_norm": 0.30219688777437337, "learning_rate": 2.378738324403744e-05, "loss": 0.2649, "step": 3875 }, { "epoch": 1.6943231441048034, "grad_norm": 0.286618784377373, "learning_rate": 2.3723949273453073e-05, "loss": 0.2668, "step": 3880 }, { "epoch": 1.6965065502183405, "grad_norm": 0.31422639752909204, "learning_rate": 2.3660523538728904e-05, "loss": 0.2671, "step": 3885 }, { "epoch": 1.6986899563318776, "grad_norm": 0.24899445840743908, "learning_rate": 2.3597106449226015e-05, "loss": 0.2603, "step": 3890 }, { "epoch": 1.7008733624454149, "grad_norm": 0.28778299708675603, "learning_rate": 2.3533698414249668e-05, "loss": 0.2532, "step": 3895 }, { "epoch": 1.703056768558952, "grad_norm": 0.3010236993612902, "learning_rate": 2.3470299843046693e-05, "loss": 0.2607, "step": 3900 }, { "epoch": 1.7052401746724892, "grad_norm": 0.2505565783956947, "learning_rate": 2.340691114480284e-05, "loss": 0.2675, "step": 3905 }, { "epoch": 1.7074235807860263, "grad_norm": 0.2534199791327587, "learning_rate": 2.334353272864013e-05, "loss": 0.2727, "step": 3910 }, { "epoch": 1.7096069868995634, "grad_norm": 0.2759471253942165, "learning_rate": 2.3280165003614227e-05, "loss": 0.2663, "step": 3915 }, { "epoch": 1.7117903930131004, "grad_norm": 0.2958147406172001, "learning_rate": 2.3216808378711788e-05, "loss": 0.2705, "step": 3920 }, { "epoch": 1.7139737991266375, "grad_norm": 0.26156474466311214, "learning_rate": 2.3153463262847835e-05, "loss": 0.242, "step": 3925 }, { "epoch": 1.7161572052401746, "grad_norm": 0.29968058749209264, "learning_rate": 2.3090130064863113e-05, "loss": 0.2681, "step": 3930 }, { "epoch": 1.7183406113537116, "grad_norm": 0.24513602810052307, "learning_rate": 2.3026809193521423e-05, "loss": 0.2746, "step": 3935 }, { "epoch": 1.720524017467249, "grad_norm": 0.2624738237172384, "learning_rate": 2.296350105750703e-05, "loss": 0.2682, "step": 3940 }, { "epoch": 1.722707423580786, "grad_norm": 0.21915397316253904, "learning_rate": 2.2900206065421997e-05, "loss": 0.26, "step": 3945 }, { "epoch": 1.7248908296943233, "grad_norm": 0.248859562817194, "learning_rate": 2.2836924625783554e-05, "loss": 0.2776, "step": 3950 }, { "epoch": 1.7270742358078603, "grad_norm": 0.2941037210948601, "learning_rate": 2.2773657147021464e-05, "loss": 0.281, "step": 3955 }, { "epoch": 1.7292576419213974, "grad_norm": 0.22756003553556517, "learning_rate": 2.271040403747537e-05, "loss": 0.2597, "step": 3960 }, { "epoch": 1.7314410480349345, "grad_norm": 0.274949859317834, "learning_rate": 2.2647165705392186e-05, "loss": 0.2697, "step": 3965 }, { "epoch": 1.7336244541484715, "grad_norm": 0.2797174304909422, "learning_rate": 2.2583942558923456e-05, "loss": 0.2716, "step": 3970 }, { "epoch": 1.7358078602620086, "grad_norm": 0.26704187553881054, "learning_rate": 2.2520735006122688e-05, "loss": 0.274, "step": 3975 }, { "epoch": 1.737991266375546, "grad_norm": 0.284507084979283, "learning_rate": 2.2457543454942786e-05, "loss": 0.2587, "step": 3980 }, { "epoch": 1.740174672489083, "grad_norm": 0.2699117094732979, "learning_rate": 2.2394368313233332e-05, "loss": 0.2619, "step": 3985 }, { "epoch": 1.74235807860262, "grad_norm": 0.25250488525353865, "learning_rate": 2.2331209988738024e-05, "loss": 0.2602, "step": 3990 }, { "epoch": 1.7445414847161573, "grad_norm": 0.24206728711747397, "learning_rate": 2.226806888909202e-05, "loss": 0.262, "step": 3995 }, { "epoch": 1.7467248908296944, "grad_norm": 0.2907832129799141, "learning_rate": 2.2204945421819292e-05, "loss": 0.2636, "step": 4000 }, { "epoch": 1.7489082969432315, "grad_norm": 0.29699912162411024, "learning_rate": 2.2141839994330027e-05, "loss": 0.2747, "step": 4005 }, { "epoch": 1.7510917030567685, "grad_norm": 0.2505983156960274, "learning_rate": 2.207875301391796e-05, "loss": 0.2706, "step": 4010 }, { "epoch": 1.7532751091703056, "grad_norm": 0.20887238006814582, "learning_rate": 2.2015684887757776e-05, "loss": 0.2544, "step": 4015 }, { "epoch": 1.7554585152838427, "grad_norm": 0.2350828605757233, "learning_rate": 2.195263602290248e-05, "loss": 0.2759, "step": 4020 }, { "epoch": 1.75764192139738, "grad_norm": 0.27919466794671643, "learning_rate": 2.188960682628074e-05, "loss": 0.2468, "step": 4025 }, { "epoch": 1.759825327510917, "grad_norm": 0.24755062875053999, "learning_rate": 2.1826597704694306e-05, "loss": 0.2699, "step": 4030 }, { "epoch": 1.762008733624454, "grad_norm": 0.28930040508547633, "learning_rate": 2.176360906481533e-05, "loss": 0.27, "step": 4035 }, { "epoch": 1.7641921397379914, "grad_norm": 0.2463760850445127, "learning_rate": 2.1700641313183807e-05, "loss": 0.2604, "step": 4040 }, { "epoch": 1.7663755458515285, "grad_norm": 0.24878427896848646, "learning_rate": 2.163769485620489e-05, "loss": 0.2647, "step": 4045 }, { "epoch": 1.7685589519650655, "grad_norm": 0.3491349213382724, "learning_rate": 2.1574770100146295e-05, "loss": 0.2717, "step": 4050 }, { "epoch": 1.7707423580786026, "grad_norm": 0.26013086945480984, "learning_rate": 2.1511867451135693e-05, "loss": 0.2599, "step": 4055 }, { "epoch": 1.7729257641921397, "grad_norm": 0.2900875329162927, "learning_rate": 2.1448987315158048e-05, "loss": 0.2686, "step": 4060 }, { "epoch": 1.7751091703056767, "grad_norm": 0.2653600164327816, "learning_rate": 2.1386130098053036e-05, "loss": 0.2627, "step": 4065 }, { "epoch": 1.777292576419214, "grad_norm": 0.27303311100886735, "learning_rate": 2.1323296205512418e-05, "loss": 0.2708, "step": 4070 }, { "epoch": 1.779475982532751, "grad_norm": 0.29210046654985766, "learning_rate": 2.1260486043077388e-05, "loss": 0.2581, "step": 4075 }, { "epoch": 1.7816593886462884, "grad_norm": 0.284149374163666, "learning_rate": 2.1197700016136014e-05, "loss": 0.266, "step": 4080 }, { "epoch": 1.7838427947598254, "grad_norm": 0.2409781119667348, "learning_rate": 2.113493852992056e-05, "loss": 0.2571, "step": 4085 }, { "epoch": 1.7860262008733625, "grad_norm": 0.20843436142983807, "learning_rate": 2.1072201989504913e-05, "loss": 0.2733, "step": 4090 }, { "epoch": 1.7882096069868996, "grad_norm": 0.23230387555929172, "learning_rate": 2.1009490799801965e-05, "loss": 0.269, "step": 4095 }, { "epoch": 1.7903930131004366, "grad_norm": 0.2694983772096665, "learning_rate": 2.0946805365560966e-05, "loss": 0.2745, "step": 4100 }, { "epoch": 1.7925764192139737, "grad_norm": 0.24747448881142187, "learning_rate": 2.0884146091364956e-05, "loss": 0.2659, "step": 4105 }, { "epoch": 1.7947598253275108, "grad_norm": 0.22323270338570358, "learning_rate": 2.0821513381628132e-05, "loss": 0.2645, "step": 4110 }, { "epoch": 1.796943231441048, "grad_norm": 0.2580592437636737, "learning_rate": 2.0758907640593228e-05, "loss": 0.2563, "step": 4115 }, { "epoch": 1.7991266375545851, "grad_norm": 0.2411641382308118, "learning_rate": 2.0696329272328932e-05, "loss": 0.2721, "step": 4120 }, { "epoch": 1.8013100436681224, "grad_norm": 0.25806558253779854, "learning_rate": 2.063377868072724e-05, "loss": 0.2616, "step": 4125 }, { "epoch": 1.8034934497816595, "grad_norm": 0.2455964016028092, "learning_rate": 2.0571256269500893e-05, "loss": 0.271, "step": 4130 }, { "epoch": 1.8056768558951966, "grad_norm": 0.25469457274681495, "learning_rate": 2.0508762442180745e-05, "loss": 0.2683, "step": 4135 }, { "epoch": 1.8078602620087336, "grad_norm": 0.2734524817312364, "learning_rate": 2.0446297602113167e-05, "loss": 0.2782, "step": 4140 }, { "epoch": 1.8100436681222707, "grad_norm": 0.2711480537954455, "learning_rate": 2.0383862152457432e-05, "loss": 0.2636, "step": 4145 }, { "epoch": 1.8122270742358078, "grad_norm": 0.23192737600539587, "learning_rate": 2.0321456496183134e-05, "loss": 0.276, "step": 4150 }, { "epoch": 1.8144104803493448, "grad_norm": 0.2610658805286943, "learning_rate": 2.0259081036067563e-05, "loss": 0.2706, "step": 4155 }, { "epoch": 1.8165938864628821, "grad_norm": 0.2670079724190112, "learning_rate": 2.0196736174693126e-05, "loss": 0.2529, "step": 4160 }, { "epoch": 1.8187772925764192, "grad_norm": 0.25454455186632396, "learning_rate": 2.013442231444474e-05, "loss": 0.2649, "step": 4165 }, { "epoch": 1.8209606986899565, "grad_norm": 0.2777703715553109, "learning_rate": 2.007213985750724e-05, "loss": 0.265, "step": 4170 }, { "epoch": 1.8231441048034935, "grad_norm": 0.2654974710451674, "learning_rate": 2.000988920586275e-05, "loss": 0.2653, "step": 4175 }, { "epoch": 1.8253275109170306, "grad_norm": 0.28984168869856813, "learning_rate": 1.994767076128816e-05, "loss": 0.2572, "step": 4180 }, { "epoch": 1.8275109170305677, "grad_norm": 0.2620326885283614, "learning_rate": 1.988548492535246e-05, "loss": 0.2754, "step": 4185 }, { "epoch": 1.8296943231441047, "grad_norm": 0.24235633152361688, "learning_rate": 1.9823332099414185e-05, "loss": 0.2717, "step": 4190 }, { "epoch": 1.8318777292576418, "grad_norm": 0.24991695967334396, "learning_rate": 1.9761212684618824e-05, "loss": 0.25, "step": 4195 }, { "epoch": 1.8340611353711789, "grad_norm": 0.21335483219532675, "learning_rate": 1.9699127081896214e-05, "loss": 0.2513, "step": 4200 }, { "epoch": 1.8362445414847162, "grad_norm": 0.2515339708285586, "learning_rate": 1.963707569195797e-05, "loss": 0.2524, "step": 4205 }, { "epoch": 1.8384279475982532, "grad_norm": 0.23802145795043225, "learning_rate": 1.9575058915294885e-05, "loss": 0.2652, "step": 4210 }, { "epoch": 1.8406113537117905, "grad_norm": 0.30056095101017133, "learning_rate": 1.9513077152174363e-05, "loss": 0.2776, "step": 4215 }, { "epoch": 1.8427947598253276, "grad_norm": 0.2915453674361558, "learning_rate": 1.9451130802637822e-05, "loss": 0.2588, "step": 4220 }, { "epoch": 1.8449781659388647, "grad_norm": 0.30777863799002625, "learning_rate": 1.9389220266498102e-05, "loss": 0.2701, "step": 4225 }, { "epoch": 1.8471615720524017, "grad_norm": 0.2597777533142635, "learning_rate": 1.9327345943336904e-05, "loss": 0.2645, "step": 4230 }, { "epoch": 1.8493449781659388, "grad_norm": 0.3080444941319507, "learning_rate": 1.9265508232502208e-05, "loss": 0.2749, "step": 4235 }, { "epoch": 1.8515283842794759, "grad_norm": 0.2831226100931665, "learning_rate": 1.9203707533105687e-05, "loss": 0.2598, "step": 4240 }, { "epoch": 1.8537117903930131, "grad_norm": 0.27124968567558605, "learning_rate": 1.914194424402015e-05, "loss": 0.271, "step": 4245 }, { "epoch": 1.8558951965065502, "grad_norm": 0.24173431823915756, "learning_rate": 1.9080218763876917e-05, "loss": 0.2512, "step": 4250 }, { "epoch": 1.8580786026200873, "grad_norm": 0.27847599015903607, "learning_rate": 1.901853149106332e-05, "loss": 0.263, "step": 4255 }, { "epoch": 1.8602620087336246, "grad_norm": 0.348403165189276, "learning_rate": 1.8956882823720075e-05, "loss": 0.2657, "step": 4260 }, { "epoch": 1.8624454148471616, "grad_norm": 0.24233816768582403, "learning_rate": 1.889527315973874e-05, "loss": 0.26, "step": 4265 }, { "epoch": 1.8646288209606987, "grad_norm": 0.23840655638016398, "learning_rate": 1.8833702896759132e-05, "loss": 0.2636, "step": 4270 }, { "epoch": 1.8668122270742358, "grad_norm": 0.2549222906252635, "learning_rate": 1.8772172432166767e-05, "loss": 0.2744, "step": 4275 }, { "epoch": 1.8689956331877728, "grad_norm": 0.25475120516780736, "learning_rate": 1.87106821630903e-05, "loss": 0.2748, "step": 4280 }, { "epoch": 1.87117903930131, "grad_norm": 0.24088260430162353, "learning_rate": 1.8649232486398955e-05, "loss": 0.2638, "step": 4285 }, { "epoch": 1.8733624454148472, "grad_norm": 0.2556026511315039, "learning_rate": 1.8587823798699958e-05, "loss": 0.2601, "step": 4290 }, { "epoch": 1.8755458515283843, "grad_norm": 0.2700718784491458, "learning_rate": 1.8526456496335997e-05, "loss": 0.2525, "step": 4295 }, { "epoch": 1.8777292576419216, "grad_norm": 0.27667715242466545, "learning_rate": 1.846513097538265e-05, "loss": 0.2547, "step": 4300 }, { "epoch": 1.8799126637554586, "grad_norm": 0.2488594588357838, "learning_rate": 1.8403847631645815e-05, "loss": 0.2747, "step": 4305 }, { "epoch": 1.8820960698689957, "grad_norm": 0.2486103000672603, "learning_rate": 1.834260686065919e-05, "loss": 0.2755, "step": 4310 }, { "epoch": 1.8842794759825328, "grad_norm": 0.2673003900981902, "learning_rate": 1.8281409057681685e-05, "loss": 0.2556, "step": 4315 }, { "epoch": 1.8864628820960698, "grad_norm": 0.27816423985109073, "learning_rate": 1.8220254617694904e-05, "loss": 0.2725, "step": 4320 }, { "epoch": 1.888646288209607, "grad_norm": 0.21891089387179588, "learning_rate": 1.815914393540057e-05, "loss": 0.2672, "step": 4325 }, { "epoch": 1.890829694323144, "grad_norm": 0.3027560989030244, "learning_rate": 1.8098077405217973e-05, "loss": 0.2681, "step": 4330 }, { "epoch": 1.8930131004366813, "grad_norm": 0.2956885379322635, "learning_rate": 1.8037055421281453e-05, "loss": 0.2754, "step": 4335 }, { "epoch": 1.8951965065502183, "grad_norm": 0.27344213459308064, "learning_rate": 1.7976078377437843e-05, "loss": 0.254, "step": 4340 }, { "epoch": 1.8973799126637556, "grad_norm": 0.25908804251433426, "learning_rate": 1.7915146667243912e-05, "loss": 0.2709, "step": 4345 }, { "epoch": 1.8995633187772927, "grad_norm": 0.27387774174550766, "learning_rate": 1.785426068396385e-05, "loss": 0.2724, "step": 4350 }, { "epoch": 1.9017467248908297, "grad_norm": 0.2349626027532333, "learning_rate": 1.7793420820566698e-05, "loss": 0.2685, "step": 4355 }, { "epoch": 1.9039301310043668, "grad_norm": 0.27519209684053136, "learning_rate": 1.7732627469723867e-05, "loss": 0.2665, "step": 4360 }, { "epoch": 1.9061135371179039, "grad_norm": 0.2237579472013528, "learning_rate": 1.7671881023806526e-05, "loss": 0.2629, "step": 4365 }, { "epoch": 1.908296943231441, "grad_norm": 0.24463587078424295, "learning_rate": 1.7611181874883133e-05, "loss": 0.2637, "step": 4370 }, { "epoch": 1.910480349344978, "grad_norm": 0.30256331204318965, "learning_rate": 1.755053041471689e-05, "loss": 0.2626, "step": 4375 }, { "epoch": 1.9126637554585153, "grad_norm": 0.23702938555275288, "learning_rate": 1.748992703476318e-05, "loss": 0.2621, "step": 4380 }, { "epoch": 1.9148471615720524, "grad_norm": 0.29244916287768213, "learning_rate": 1.7429372126167105e-05, "loss": 0.2492, "step": 4385 }, { "epoch": 1.9170305676855897, "grad_norm": 0.2588931757550786, "learning_rate": 1.7368866079760887e-05, "loss": 0.2615, "step": 4390 }, { "epoch": 1.9192139737991267, "grad_norm": 0.25196463301087313, "learning_rate": 1.7308409286061407e-05, "loss": 0.2686, "step": 4395 }, { "epoch": 1.9213973799126638, "grad_norm": 0.2158443773081268, "learning_rate": 1.7248002135267645e-05, "loss": 0.2666, "step": 4400 }, { "epoch": 1.9235807860262009, "grad_norm": 0.287220788904416, "learning_rate": 1.7187645017258196e-05, "loss": 0.2735, "step": 4405 }, { "epoch": 1.925764192139738, "grad_norm": 0.23107125712315182, "learning_rate": 1.712733832158872e-05, "loss": 0.2613, "step": 4410 }, { "epoch": 1.927947598253275, "grad_norm": 0.24853969589485442, "learning_rate": 1.7067082437489436e-05, "loss": 0.2653, "step": 4415 }, { "epoch": 1.930131004366812, "grad_norm": 0.2767451923179264, "learning_rate": 1.7006877753862632e-05, "loss": 0.262, "step": 4420 }, { "epoch": 1.9323144104803494, "grad_norm": 0.22927033548053927, "learning_rate": 1.6946724659280133e-05, "loss": 0.2542, "step": 4425 }, { "epoch": 1.9344978165938864, "grad_norm": 0.23110650040140782, "learning_rate": 1.6886623541980796e-05, "loss": 0.2757, "step": 4430 }, { "epoch": 1.9366812227074237, "grad_norm": 0.23731032145161848, "learning_rate": 1.682657478986801e-05, "loss": 0.2645, "step": 4435 }, { "epoch": 1.9388646288209608, "grad_norm": 0.25552976383740716, "learning_rate": 1.6766578790507192e-05, "loss": 0.2595, "step": 4440 }, { "epoch": 1.9410480349344978, "grad_norm": 0.2155945620763327, "learning_rate": 1.6706635931123277e-05, "loss": 0.2596, "step": 4445 }, { "epoch": 1.943231441048035, "grad_norm": 0.24247256828652752, "learning_rate": 1.664674659859822e-05, "loss": 0.2493, "step": 4450 }, { "epoch": 1.945414847161572, "grad_norm": 0.2516593577449982, "learning_rate": 1.658691117946853e-05, "loss": 0.2761, "step": 4455 }, { "epoch": 1.947598253275109, "grad_norm": 0.21991257872575132, "learning_rate": 1.6527130059922717e-05, "loss": 0.2632, "step": 4460 }, { "epoch": 1.9497816593886463, "grad_norm": 0.261259312013636, "learning_rate": 1.646740362579886e-05, "loss": 0.2615, "step": 4465 }, { "epoch": 1.9519650655021834, "grad_norm": 0.2318715733570351, "learning_rate": 1.6407732262582067e-05, "loss": 0.2654, "step": 4470 }, { "epoch": 1.9541484716157205, "grad_norm": 0.22989633286543176, "learning_rate": 1.634811635540201e-05, "loss": 0.2593, "step": 4475 }, { "epoch": 1.9563318777292578, "grad_norm": 0.24017270832452148, "learning_rate": 1.6288556289030452e-05, "loss": 0.2676, "step": 4480 }, { "epoch": 1.9585152838427948, "grad_norm": 0.25142686822052845, "learning_rate": 1.6229052447878743e-05, "loss": 0.2541, "step": 4485 }, { "epoch": 1.960698689956332, "grad_norm": 0.22732754650526263, "learning_rate": 1.6169605215995345e-05, "loss": 0.259, "step": 4490 }, { "epoch": 1.962882096069869, "grad_norm": 0.24853011047008652, "learning_rate": 1.6110214977063345e-05, "loss": 0.2559, "step": 4495 }, { "epoch": 1.965065502183406, "grad_norm": 0.2557640948437553, "learning_rate": 1.6050882114398e-05, "loss": 0.2508, "step": 4500 }, { "epoch": 1.967248908296943, "grad_norm": 0.24324444806418238, "learning_rate": 1.5991607010944242e-05, "loss": 0.2607, "step": 4505 }, { "epoch": 1.9694323144104804, "grad_norm": 0.23773093829535516, "learning_rate": 1.5932390049274227e-05, "loss": 0.2644, "step": 4510 }, { "epoch": 1.9716157205240175, "grad_norm": 0.2532308876036313, "learning_rate": 1.587323161158484e-05, "loss": 0.272, "step": 4515 }, { "epoch": 1.9737991266375547, "grad_norm": 0.2617214304004118, "learning_rate": 1.5814132079695232e-05, "loss": 0.2649, "step": 4520 }, { "epoch": 1.9759825327510918, "grad_norm": 0.27129287136393265, "learning_rate": 1.57550918350444e-05, "loss": 0.2591, "step": 4525 }, { "epoch": 1.9781659388646289, "grad_norm": 0.24468851478815726, "learning_rate": 1.5696111258688657e-05, "loss": 0.256, "step": 4530 }, { "epoch": 1.980349344978166, "grad_norm": 0.23891980205993738, "learning_rate": 1.563719073129923e-05, "loss": 0.2655, "step": 4535 }, { "epoch": 1.982532751091703, "grad_norm": 0.21564016827597185, "learning_rate": 1.557833063315976e-05, "loss": 0.2593, "step": 4540 }, { "epoch": 1.98471615720524, "grad_norm": 0.2744337627938704, "learning_rate": 1.5519531344163873e-05, "loss": 0.2593, "step": 4545 }, { "epoch": 1.9868995633187772, "grad_norm": 0.239539464353039, "learning_rate": 1.5460793243812732e-05, "loss": 0.2645, "step": 4550 }, { "epoch": 1.9890829694323144, "grad_norm": 0.2652718460025595, "learning_rate": 1.5402116711212566e-05, "loss": 0.2712, "step": 4555 }, { "epoch": 1.9912663755458515, "grad_norm": 0.23601629641684965, "learning_rate": 1.5343502125072236e-05, "loss": 0.2542, "step": 4560 }, { "epoch": 1.9934497816593888, "grad_norm": 0.23315305328846694, "learning_rate": 1.5284949863700805e-05, "loss": 0.2663, "step": 4565 }, { "epoch": 1.9956331877729259, "grad_norm": 0.23782090171706022, "learning_rate": 1.5226460305005044e-05, "loss": 0.2622, "step": 4570 }, { "epoch": 1.997816593886463, "grad_norm": 0.22388379965681177, "learning_rate": 1.5168033826487072e-05, "loss": 0.2692, "step": 4575 }, { "epoch": 2.0, "grad_norm": 0.27446739099024114, "learning_rate": 1.5109670805241846e-05, "loss": 0.2572, "step": 4580 }, { "epoch": 2.002183406113537, "grad_norm": 0.29167629982584226, "learning_rate": 1.505137161795478e-05, "loss": 0.2211, "step": 4585 }, { "epoch": 2.004366812227074, "grad_norm": 0.2877855784905206, "learning_rate": 1.4993136640899286e-05, "loss": 0.2147, "step": 4590 }, { "epoch": 2.006550218340611, "grad_norm": 0.2589731295275126, "learning_rate": 1.4934966249934334e-05, "loss": 0.2052, "step": 4595 }, { "epoch": 2.0087336244541483, "grad_norm": 0.24025382992800187, "learning_rate": 1.487686082050207e-05, "loss": 0.2162, "step": 4600 }, { "epoch": 2.010917030567686, "grad_norm": 0.22612669576702268, "learning_rate": 1.4818820727625348e-05, "loss": 0.2071, "step": 4605 }, { "epoch": 2.013100436681223, "grad_norm": 0.2212365851012379, "learning_rate": 1.4760846345905344e-05, "loss": 0.2139, "step": 4610 }, { "epoch": 2.01528384279476, "grad_norm": 0.20584034227093886, "learning_rate": 1.4702938049519105e-05, "loss": 0.2168, "step": 4615 }, { "epoch": 2.017467248908297, "grad_norm": 0.21936675852848891, "learning_rate": 1.4645096212217167e-05, "loss": 0.2073, "step": 4620 }, { "epoch": 2.019650655021834, "grad_norm": 0.26100976524835695, "learning_rate": 1.4587321207321115e-05, "loss": 0.2233, "step": 4625 }, { "epoch": 2.021834061135371, "grad_norm": 0.2056350483707369, "learning_rate": 1.4529613407721192e-05, "loss": 0.2016, "step": 4630 }, { "epoch": 2.024017467248908, "grad_norm": 0.21939396929192034, "learning_rate": 1.4471973185873888e-05, "loss": 0.2092, "step": 4635 }, { "epoch": 2.0262008733624453, "grad_norm": 0.20467895973459885, "learning_rate": 1.4414400913799535e-05, "loss": 0.2026, "step": 4640 }, { "epoch": 2.0283842794759823, "grad_norm": 0.20303046431819063, "learning_rate": 1.4356896963079885e-05, "loss": 0.2034, "step": 4645 }, { "epoch": 2.03056768558952, "grad_norm": 0.2147257371596202, "learning_rate": 1.4299461704855759e-05, "loss": 0.2038, "step": 4650 }, { "epoch": 2.032751091703057, "grad_norm": 0.2709689872203016, "learning_rate": 1.424209550982461e-05, "loss": 0.2151, "step": 4655 }, { "epoch": 2.034934497816594, "grad_norm": 0.22708728340724726, "learning_rate": 1.4184798748238148e-05, "loss": 0.2176, "step": 4660 }, { "epoch": 2.037117903930131, "grad_norm": 0.25203147090946515, "learning_rate": 1.4127571789899946e-05, "loss": 0.2077, "step": 4665 }, { "epoch": 2.039301310043668, "grad_norm": 0.23983946431243502, "learning_rate": 1.4070415004163056e-05, "loss": 0.2163, "step": 4670 }, { "epoch": 2.041484716157205, "grad_norm": 0.24658657899765385, "learning_rate": 1.4013328759927625e-05, "loss": 0.2139, "step": 4675 }, { "epoch": 2.0436681222707422, "grad_norm": 0.22543241349761747, "learning_rate": 1.39563134256385e-05, "loss": 0.2019, "step": 4680 }, { "epoch": 2.0458515283842793, "grad_norm": 0.22236910097875634, "learning_rate": 1.3899369369282888e-05, "loss": 0.2183, "step": 4685 }, { "epoch": 2.0480349344978164, "grad_norm": 0.20505749283945437, "learning_rate": 1.3842496958387938e-05, "loss": 0.2188, "step": 4690 }, { "epoch": 2.050218340611354, "grad_norm": 0.2025603121212584, "learning_rate": 1.3785696560018374e-05, "loss": 0.2038, "step": 4695 }, { "epoch": 2.052401746724891, "grad_norm": 0.2263051334719587, "learning_rate": 1.372896854077416e-05, "loss": 0.206, "step": 4700 }, { "epoch": 2.054585152838428, "grad_norm": 0.2321594137504455, "learning_rate": 1.3672313266788106e-05, "loss": 0.2133, "step": 4705 }, { "epoch": 2.056768558951965, "grad_norm": 0.2122041960683085, "learning_rate": 1.3615731103723507e-05, "loss": 0.2126, "step": 4710 }, { "epoch": 2.058951965065502, "grad_norm": 0.24859114020141224, "learning_rate": 1.3559222416771793e-05, "loss": 0.2061, "step": 4715 }, { "epoch": 2.061135371179039, "grad_norm": 0.21853834679540102, "learning_rate": 1.350278757065016e-05, "loss": 0.2109, "step": 4720 }, { "epoch": 2.0633187772925763, "grad_norm": 0.2074388039569142, "learning_rate": 1.3446426929599229e-05, "loss": 0.2137, "step": 4725 }, { "epoch": 2.0655021834061134, "grad_norm": 0.21863003094886263, "learning_rate": 1.3390140857380678e-05, "loss": 0.1967, "step": 4730 }, { "epoch": 2.067685589519651, "grad_norm": 0.25260685298764257, "learning_rate": 1.3333929717274919e-05, "loss": 0.2255, "step": 4735 }, { "epoch": 2.069868995633188, "grad_norm": 0.23960033740421333, "learning_rate": 1.3277793872078732e-05, "loss": 0.2195, "step": 4740 }, { "epoch": 2.072052401746725, "grad_norm": 0.23761525700313252, "learning_rate": 1.3221733684102905e-05, "loss": 0.2128, "step": 4745 }, { "epoch": 2.074235807860262, "grad_norm": 0.21987521365657883, "learning_rate": 1.3165749515169967e-05, "loss": 0.2115, "step": 4750 }, { "epoch": 2.076419213973799, "grad_norm": 0.223938388935573, "learning_rate": 1.310984172661177e-05, "loss": 0.2181, "step": 4755 }, { "epoch": 2.078602620087336, "grad_norm": 0.2521445985500392, "learning_rate": 1.305401067926722e-05, "loss": 0.2143, "step": 4760 }, { "epoch": 2.0807860262008733, "grad_norm": 0.23287630698929376, "learning_rate": 1.2998256733479897e-05, "loss": 0.2002, "step": 4765 }, { "epoch": 2.0829694323144103, "grad_norm": 0.2600254334217533, "learning_rate": 1.294258024909577e-05, "loss": 0.2041, "step": 4770 }, { "epoch": 2.0851528384279474, "grad_norm": 0.2926480056967554, "learning_rate": 1.288698158546085e-05, "loss": 0.2206, "step": 4775 }, { "epoch": 2.087336244541485, "grad_norm": 0.21634943547113875, "learning_rate": 1.2831461101418884e-05, "loss": 0.2166, "step": 4780 }, { "epoch": 2.089519650655022, "grad_norm": 0.2398092797244653, "learning_rate": 1.2776019155309035e-05, "loss": 0.2228, "step": 4785 }, { "epoch": 2.091703056768559, "grad_norm": 0.22191669799449495, "learning_rate": 1.2720656104963569e-05, "loss": 0.2146, "step": 4790 }, { "epoch": 2.093886462882096, "grad_norm": 0.22249356654838032, "learning_rate": 1.2665372307705525e-05, "loss": 0.2173, "step": 4795 }, { "epoch": 2.096069868995633, "grad_norm": 0.2338757269866602, "learning_rate": 1.2610168120346455e-05, "loss": 0.2203, "step": 4800 }, { "epoch": 2.0982532751091703, "grad_norm": 0.20569835239307332, "learning_rate": 1.2555043899184077e-05, "loss": 0.2169, "step": 4805 }, { "epoch": 2.1004366812227073, "grad_norm": 0.18840672821070764, "learning_rate": 1.2500000000000006e-05, "loss": 0.2131, "step": 4810 }, { "epoch": 2.1026200873362444, "grad_norm": 0.229982583270256, "learning_rate": 1.2445036778057434e-05, "loss": 0.2155, "step": 4815 }, { "epoch": 2.1048034934497815, "grad_norm": 0.21678928738468423, "learning_rate": 1.2390154588098854e-05, "loss": 0.2151, "step": 4820 }, { "epoch": 2.106986899563319, "grad_norm": 0.20815591769350697, "learning_rate": 1.233535378434376e-05, "loss": 0.2075, "step": 4825 }, { "epoch": 2.109170305676856, "grad_norm": 0.22153362078682773, "learning_rate": 1.2280634720486364e-05, "loss": 0.2136, "step": 4830 }, { "epoch": 2.111353711790393, "grad_norm": 0.21312931677811298, "learning_rate": 1.2225997749693318e-05, "loss": 0.21, "step": 4835 }, { "epoch": 2.11353711790393, "grad_norm": 0.23940821453611505, "learning_rate": 1.2171443224601443e-05, "loss": 0.2141, "step": 4840 }, { "epoch": 2.1157205240174672, "grad_norm": 0.2242248154195571, "learning_rate": 1.2116971497315402e-05, "loss": 0.208, "step": 4845 }, { "epoch": 2.1179039301310043, "grad_norm": 0.21221180992392646, "learning_rate": 1.2062582919405502e-05, "loss": 0.2093, "step": 4850 }, { "epoch": 2.1200873362445414, "grad_norm": 0.22445864486046593, "learning_rate": 1.200827784190537e-05, "loss": 0.2146, "step": 4855 }, { "epoch": 2.1222707423580784, "grad_norm": 0.2301722383710042, "learning_rate": 1.1954056615309728e-05, "loss": 0.2084, "step": 4860 }, { "epoch": 2.1244541484716155, "grad_norm": 0.2135558798950853, "learning_rate": 1.1899919589572092e-05, "loss": 0.2126, "step": 4865 }, { "epoch": 2.126637554585153, "grad_norm": 0.2073308539610393, "learning_rate": 1.1845867114102532e-05, "loss": 0.2063, "step": 4870 }, { "epoch": 2.12882096069869, "grad_norm": 0.2049312257048255, "learning_rate": 1.1791899537765416e-05, "loss": 0.2224, "step": 4875 }, { "epoch": 2.131004366812227, "grad_norm": 0.24321581226405323, "learning_rate": 1.173801720887717e-05, "loss": 0.2218, "step": 4880 }, { "epoch": 2.1331877729257642, "grad_norm": 0.21578319625044598, "learning_rate": 1.1684220475203986e-05, "loss": 0.199, "step": 4885 }, { "epoch": 2.1353711790393013, "grad_norm": 0.2162898287189938, "learning_rate": 1.1630509683959643e-05, "loss": 0.2121, "step": 4890 }, { "epoch": 2.1375545851528384, "grad_norm": 0.24512267530291906, "learning_rate": 1.1576885181803215e-05, "loss": 0.2171, "step": 4895 }, { "epoch": 2.1397379912663754, "grad_norm": 0.19666073074843837, "learning_rate": 1.1523347314836858e-05, "loss": 0.2135, "step": 4900 }, { "epoch": 2.1419213973799125, "grad_norm": 0.21309169492333768, "learning_rate": 1.1469896428603563e-05, "loss": 0.22, "step": 4905 }, { "epoch": 2.14410480349345, "grad_norm": 0.2097865812131667, "learning_rate": 1.1416532868084937e-05, "loss": 0.2022, "step": 4910 }, { "epoch": 2.146288209606987, "grad_norm": 0.22140108936049668, "learning_rate": 1.1363256977698968e-05, "loss": 0.2157, "step": 4915 }, { "epoch": 2.148471615720524, "grad_norm": 0.20612571861461668, "learning_rate": 1.1310069101297807e-05, "loss": 0.212, "step": 4920 }, { "epoch": 2.150655021834061, "grad_norm": 0.2172078322760656, "learning_rate": 1.1256969582165544e-05, "loss": 0.21, "step": 4925 }, { "epoch": 2.1528384279475983, "grad_norm": 0.22655837718087724, "learning_rate": 1.1203958763016007e-05, "loss": 0.1969, "step": 4930 }, { "epoch": 2.1550218340611353, "grad_norm": 0.23562847936093284, "learning_rate": 1.1151036985990509e-05, "loss": 0.209, "step": 4935 }, { "epoch": 2.1572052401746724, "grad_norm": 0.21438898804486692, "learning_rate": 1.1098204592655703e-05, "loss": 0.2144, "step": 4940 }, { "epoch": 2.1593886462882095, "grad_norm": 0.22341501985465956, "learning_rate": 1.1045461924001325e-05, "loss": 0.2095, "step": 4945 }, { "epoch": 2.1615720524017465, "grad_norm": 0.2154153191432171, "learning_rate": 1.0992809320438016e-05, "loss": 0.2071, "step": 4950 }, { "epoch": 2.1637554585152836, "grad_norm": 0.19531686390702627, "learning_rate": 1.0940247121795123e-05, "loss": 0.2095, "step": 4955 }, { "epoch": 2.165938864628821, "grad_norm": 0.22132766807603957, "learning_rate": 1.08877756673185e-05, "loss": 0.2174, "step": 4960 }, { "epoch": 2.168122270742358, "grad_norm": 0.22517823657634206, "learning_rate": 1.0835395295668327e-05, "loss": 0.2179, "step": 4965 }, { "epoch": 2.1703056768558953, "grad_norm": 0.21142342412266563, "learning_rate": 1.0783106344916913e-05, "loss": 0.2051, "step": 4970 }, { "epoch": 2.1724890829694323, "grad_norm": 0.20941158976258883, "learning_rate": 1.073090915254652e-05, "loss": 0.2221, "step": 4975 }, { "epoch": 2.1746724890829694, "grad_norm": 0.21006000563282354, "learning_rate": 1.06788040554472e-05, "loss": 0.2003, "step": 4980 }, { "epoch": 2.1768558951965065, "grad_norm": 0.19371851293041617, "learning_rate": 1.0626791389914573e-05, "loss": 0.2029, "step": 4985 }, { "epoch": 2.1790393013100435, "grad_norm": 0.2202294387310257, "learning_rate": 1.0574871491647723e-05, "loss": 0.2141, "step": 4990 }, { "epoch": 2.1812227074235806, "grad_norm": 0.22059840659575228, "learning_rate": 1.0523044695746975e-05, "loss": 0.2121, "step": 4995 }, { "epoch": 2.183406113537118, "grad_norm": 0.2066703461639902, "learning_rate": 1.0471311336711772e-05, "loss": 0.2123, "step": 5000 }, { "epoch": 2.185589519650655, "grad_norm": 0.20919147411357605, "learning_rate": 1.0419671748438487e-05, "loss": 0.2166, "step": 5005 }, { "epoch": 2.1877729257641922, "grad_norm": 0.1934804236700962, "learning_rate": 1.0368126264218282e-05, "loss": 0.2109, "step": 5010 }, { "epoch": 2.1899563318777293, "grad_norm": 0.2089470879409506, "learning_rate": 1.0316675216734952e-05, "loss": 0.2057, "step": 5015 }, { "epoch": 2.1921397379912664, "grad_norm": 0.19986998684970525, "learning_rate": 1.026531893806279e-05, "loss": 0.2225, "step": 5020 }, { "epoch": 2.1943231441048034, "grad_norm": 0.22134729099448455, "learning_rate": 1.0214057759664424e-05, "loss": 0.2161, "step": 5025 }, { "epoch": 2.1965065502183405, "grad_norm": 0.20179325283452076, "learning_rate": 1.0162892012388703e-05, "loss": 0.223, "step": 5030 }, { "epoch": 2.1986899563318776, "grad_norm": 0.20653676340802546, "learning_rate": 1.0111822026468515e-05, "loss": 0.2127, "step": 5035 }, { "epoch": 2.2008733624454146, "grad_norm": 0.20118347192898964, "learning_rate": 1.0060848131518716e-05, "loss": 0.2132, "step": 5040 }, { "epoch": 2.203056768558952, "grad_norm": 0.2127487431663625, "learning_rate": 1.0009970656533965e-05, "loss": 0.2138, "step": 5045 }, { "epoch": 2.2052401746724892, "grad_norm": 0.18873587826004806, "learning_rate": 9.95918992988661e-06, "loss": 0.2159, "step": 5050 }, { "epoch": 2.2074235807860263, "grad_norm": 0.21267677980729546, "learning_rate": 9.908506279324564e-06, "loss": 0.2094, "step": 5055 }, { "epoch": 2.2096069868995634, "grad_norm": 0.20619431733016733, "learning_rate": 9.857920031969198e-06, "loss": 0.2066, "step": 5060 }, { "epoch": 2.2117903930131004, "grad_norm": 0.2223339638565872, "learning_rate": 9.807431514313225e-06, "loss": 0.2315, "step": 5065 }, { "epoch": 2.2139737991266375, "grad_norm": 0.2351717184649456, "learning_rate": 9.757041052218587e-06, "loss": 0.2192, "step": 5070 }, { "epoch": 2.2161572052401746, "grad_norm": 0.18769468217068455, "learning_rate": 9.70674897091437e-06, "loss": 0.21, "step": 5075 }, { "epoch": 2.2183406113537116, "grad_norm": 0.24549048829311135, "learning_rate": 9.656555594994687e-06, "loss": 0.2025, "step": 5080 }, { "epoch": 2.2205240174672487, "grad_norm": 0.20415216492897895, "learning_rate": 9.606461248416573e-06, "loss": 0.2016, "step": 5085 }, { "epoch": 2.222707423580786, "grad_norm": 0.2005430180173854, "learning_rate": 9.556466254497931e-06, "loss": 0.2098, "step": 5090 }, { "epoch": 2.2248908296943233, "grad_norm": 0.22132570053896475, "learning_rate": 9.506570935915419e-06, "loss": 0.2087, "step": 5095 }, { "epoch": 2.2270742358078603, "grad_norm": 0.19989818333839046, "learning_rate": 9.456775614702364e-06, "loss": 0.2079, "step": 5100 }, { "epoch": 2.2292576419213974, "grad_norm": 0.2473690640544336, "learning_rate": 9.407080612246697e-06, "loss": 0.2107, "step": 5105 }, { "epoch": 2.2314410480349345, "grad_norm": 0.21359326359845074, "learning_rate": 9.357486249288891e-06, "loss": 0.2254, "step": 5110 }, { "epoch": 2.2336244541484715, "grad_norm": 0.2355252231386045, "learning_rate": 9.307992845919844e-06, "loss": 0.2032, "step": 5115 }, { "epoch": 2.2358078602620086, "grad_norm": 0.1990769526625258, "learning_rate": 9.258600721578872e-06, "loss": 0.2076, "step": 5120 }, { "epoch": 2.2379912663755457, "grad_norm": 0.2244729672302675, "learning_rate": 9.209310195051582e-06, "loss": 0.2186, "step": 5125 }, { "epoch": 2.2401746724890828, "grad_norm": 0.19571847441525664, "learning_rate": 9.160121584467878e-06, "loss": 0.1942, "step": 5130 }, { "epoch": 2.2423580786026203, "grad_norm": 0.2174940143142323, "learning_rate": 9.111035207299878e-06, "loss": 0.2155, "step": 5135 }, { "epoch": 2.2445414847161573, "grad_norm": 0.20301405504923226, "learning_rate": 9.062051380359857e-06, "loss": 0.2155, "step": 5140 }, { "epoch": 2.2467248908296944, "grad_norm": 0.2381569949311249, "learning_rate": 9.01317041979822e-06, "loss": 0.2191, "step": 5145 }, { "epoch": 2.2489082969432315, "grad_norm": 0.20675760421426403, "learning_rate": 8.964392641101455e-06, "loss": 0.198, "step": 5150 }, { "epoch": 2.2510917030567685, "grad_norm": 0.18783552296846562, "learning_rate": 8.915718359090086e-06, "loss": 0.2049, "step": 5155 }, { "epoch": 2.2532751091703056, "grad_norm": 0.20565919831917942, "learning_rate": 8.867147887916663e-06, "loss": 0.2227, "step": 5160 }, { "epoch": 2.2554585152838427, "grad_norm": 0.205583376077084, "learning_rate": 8.818681541063714e-06, "loss": 0.2118, "step": 5165 }, { "epoch": 2.2576419213973797, "grad_norm": 0.18747729101469687, "learning_rate": 8.770319631341745e-06, "loss": 0.2014, "step": 5170 }, { "epoch": 2.2598253275109172, "grad_norm": 0.1903688682819849, "learning_rate": 8.722062470887169e-06, "loss": 0.1973, "step": 5175 }, { "epoch": 2.2620087336244543, "grad_norm": 0.20544378450917616, "learning_rate": 8.673910371160368e-06, "loss": 0.2132, "step": 5180 }, { "epoch": 2.2641921397379914, "grad_norm": 0.22300030961664347, "learning_rate": 8.625863642943623e-06, "loss": 0.2102, "step": 5185 }, { "epoch": 2.2663755458515285, "grad_norm": 0.24719313551585811, "learning_rate": 8.577922596339142e-06, "loss": 0.2091, "step": 5190 }, { "epoch": 2.2685589519650655, "grad_norm": 0.19395953090113746, "learning_rate": 8.530087540767029e-06, "loss": 0.2119, "step": 5195 }, { "epoch": 2.2707423580786026, "grad_norm": 0.2057558795183291, "learning_rate": 8.482358784963312e-06, "loss": 0.2088, "step": 5200 }, { "epoch": 2.2729257641921397, "grad_norm": 0.194298675532194, "learning_rate": 8.43473663697794e-06, "loss": 0.1949, "step": 5205 }, { "epoch": 2.2751091703056767, "grad_norm": 0.2059192410189242, "learning_rate": 8.387221404172796e-06, "loss": 0.2036, "step": 5210 }, { "epoch": 2.277292576419214, "grad_norm": 0.20744243911102733, "learning_rate": 8.339813393219714e-06, "loss": 0.2146, "step": 5215 }, { "epoch": 2.279475982532751, "grad_norm": 0.2043265849228992, "learning_rate": 8.292512910098508e-06, "loss": 0.2206, "step": 5220 }, { "epoch": 2.2816593886462884, "grad_norm": 0.2100815969871642, "learning_rate": 8.245320260094955e-06, "loss": 0.2199, "step": 5225 }, { "epoch": 2.2838427947598254, "grad_norm": 0.20976956534920874, "learning_rate": 8.198235747798893e-06, "loss": 0.2168, "step": 5230 }, { "epoch": 2.2860262008733625, "grad_norm": 0.2050232275697135, "learning_rate": 8.151259677102211e-06, "loss": 0.216, "step": 5235 }, { "epoch": 2.2882096069868996, "grad_norm": 4.122598477737714, "learning_rate": 8.104392351196885e-06, "loss": 0.2198, "step": 5240 }, { "epoch": 2.2903930131004366, "grad_norm": 0.2237909328728828, "learning_rate": 8.057634072573048e-06, "loss": 0.2173, "step": 5245 }, { "epoch": 2.2925764192139737, "grad_norm": 0.21176821690953787, "learning_rate": 8.010985143017008e-06, "loss": 0.2157, "step": 5250 }, { "epoch": 2.2947598253275108, "grad_norm": 0.8545824488764707, "learning_rate": 7.964445863609329e-06, "loss": 0.215, "step": 5255 }, { "epoch": 2.2969432314410483, "grad_norm": 0.23797387294881264, "learning_rate": 7.91801653472286e-06, "loss": 0.1993, "step": 5260 }, { "epoch": 2.2991266375545854, "grad_norm": 0.209126744021443, "learning_rate": 7.871697456020819e-06, "loss": 0.2285, "step": 5265 }, { "epoch": 2.3013100436681224, "grad_norm": 0.2036921534501079, "learning_rate": 7.825488926454851e-06, "loss": 0.2137, "step": 5270 }, { "epoch": 2.3034934497816595, "grad_norm": 0.19891864557452632, "learning_rate": 7.77939124426308e-06, "loss": 0.2094, "step": 5275 }, { "epoch": 2.3056768558951966, "grad_norm": 0.21772742093397335, "learning_rate": 7.733404706968215e-06, "loss": 0.2097, "step": 5280 }, { "epoch": 2.3078602620087336, "grad_norm": 0.1920838796289188, "learning_rate": 7.687529611375616e-06, "loss": 0.2161, "step": 5285 }, { "epoch": 2.3100436681222707, "grad_norm": 0.20114220947132505, "learning_rate": 7.64176625357138e-06, "loss": 0.2041, "step": 5290 }, { "epoch": 2.3122270742358078, "grad_norm": 0.20058172964283594, "learning_rate": 7.596114928920422e-06, "loss": 0.2124, "step": 5295 }, { "epoch": 2.314410480349345, "grad_norm": 0.20110764788896288, "learning_rate": 7.55057593206458e-06, "loss": 0.2018, "step": 5300 }, { "epoch": 2.316593886462882, "grad_norm": 0.2033865906846183, "learning_rate": 7.505149556920699e-06, "loss": 0.2139, "step": 5305 }, { "epoch": 2.3187772925764194, "grad_norm": 0.21910248265310586, "learning_rate": 7.459836096678757e-06, "loss": 0.2153, "step": 5310 }, { "epoch": 2.3209606986899565, "grad_norm": 0.2026773384631961, "learning_rate": 7.41463584379995e-06, "loss": 0.215, "step": 5315 }, { "epoch": 2.3231441048034935, "grad_norm": 0.2127225287015194, "learning_rate": 7.369549090014821e-06, "loss": 0.212, "step": 5320 }, { "epoch": 2.3253275109170306, "grad_norm": 0.20869494683552092, "learning_rate": 7.32457612632135e-06, "loss": 0.216, "step": 5325 }, { "epoch": 2.3275109170305677, "grad_norm": 0.19166630407586727, "learning_rate": 7.279717242983117e-06, "loss": 0.2166, "step": 5330 }, { "epoch": 2.3296943231441047, "grad_norm": 0.2004851412037729, "learning_rate": 7.234972729527401e-06, "loss": 0.207, "step": 5335 }, { "epoch": 2.331877729257642, "grad_norm": 0.18554259866541592, "learning_rate": 7.190342874743311e-06, "loss": 0.201, "step": 5340 }, { "epoch": 2.334061135371179, "grad_norm": 0.21209107877243255, "learning_rate": 7.145827966679936e-06, "loss": 0.1968, "step": 5345 }, { "epoch": 2.3362445414847164, "grad_norm": 0.20153718074405388, "learning_rate": 7.101428292644477e-06, "loss": 0.2176, "step": 5350 }, { "epoch": 2.3384279475982535, "grad_norm": 0.18992678255749726, "learning_rate": 7.057144139200378e-06, "loss": 0.2013, "step": 5355 }, { "epoch": 2.3406113537117905, "grad_norm": 0.1830289392115265, "learning_rate": 7.0129757921655355e-06, "loss": 0.1979, "step": 5360 }, { "epoch": 2.3427947598253276, "grad_norm": 0.19740674215886508, "learning_rate": 6.968923536610356e-06, "loss": 0.2078, "step": 5365 }, { "epoch": 2.3449781659388647, "grad_norm": 0.2121990875989476, "learning_rate": 6.924987656856002e-06, "loss": 0.1936, "step": 5370 }, { "epoch": 2.3471615720524017, "grad_norm": 0.22075994643117536, "learning_rate": 6.881168436472524e-06, "loss": 0.2114, "step": 5375 }, { "epoch": 2.349344978165939, "grad_norm": 0.20805940844514723, "learning_rate": 6.837466158277025e-06, "loss": 0.1941, "step": 5380 }, { "epoch": 2.351528384279476, "grad_norm": 0.18880443913500422, "learning_rate": 6.7938811043318475e-06, "loss": 0.2151, "step": 5385 }, { "epoch": 2.353711790393013, "grad_norm": 0.23050444783405694, "learning_rate": 6.750413555942742e-06, "loss": 0.2066, "step": 5390 }, { "epoch": 2.35589519650655, "grad_norm": 0.19659728337288793, "learning_rate": 6.707063793657065e-06, "loss": 0.2168, "step": 5395 }, { "epoch": 2.3580786026200875, "grad_norm": 0.2044183886617955, "learning_rate": 6.663832097261952e-06, "loss": 0.2193, "step": 5400 }, { "epoch": 2.3602620087336246, "grad_norm": 0.19473776562391912, "learning_rate": 6.620718745782525e-06, "loss": 0.2031, "step": 5405 }, { "epoch": 2.3624454148471616, "grad_norm": 0.19433516742290474, "learning_rate": 6.577724017480094e-06, "loss": 0.2147, "step": 5410 }, { "epoch": 2.3646288209606987, "grad_norm": 0.2002897275632141, "learning_rate": 6.534848189850329e-06, "loss": 0.1992, "step": 5415 }, { "epoch": 2.3668122270742358, "grad_norm": 0.1981020777647114, "learning_rate": 6.492091539621517e-06, "loss": 0.2029, "step": 5420 }, { "epoch": 2.368995633187773, "grad_norm": 0.18263004359615023, "learning_rate": 6.449454342752748e-06, "loss": 0.2018, "step": 5425 }, { "epoch": 2.37117903930131, "grad_norm": 0.1878131474845262, "learning_rate": 6.406936874432134e-06, "loss": 0.2037, "step": 5430 }, { "epoch": 2.373362445414847, "grad_norm": 0.204062897390976, "learning_rate": 6.364539409075043e-06, "loss": 0.2017, "step": 5435 }, { "epoch": 2.3755458515283845, "grad_norm": 0.19374315646095067, "learning_rate": 6.322262220322314e-06, "loss": 0.2093, "step": 5440 }, { "epoch": 2.3777292576419216, "grad_norm": 0.20250412793089098, "learning_rate": 6.280105581038509e-06, "loss": 0.2031, "step": 5445 }, { "epoch": 2.3799126637554586, "grad_norm": 0.18639353569932318, "learning_rate": 6.238069763310136e-06, "loss": 0.2082, "step": 5450 }, { "epoch": 2.3820960698689957, "grad_norm": 0.19128091651488877, "learning_rate": 6.1961550384438984e-06, "loss": 0.2066, "step": 5455 }, { "epoch": 2.3842794759825328, "grad_norm": 0.19608351533598506, "learning_rate": 6.154361676964956e-06, "loss": 0.2165, "step": 5460 }, { "epoch": 2.38646288209607, "grad_norm": 0.18656964338831433, "learning_rate": 6.112689948615144e-06, "loss": 0.2019, "step": 5465 }, { "epoch": 2.388646288209607, "grad_norm": 0.18488662819515192, "learning_rate": 6.071140122351276e-06, "loss": 0.2018, "step": 5470 }, { "epoch": 2.390829694323144, "grad_norm": 0.19839674782985758, "learning_rate": 6.029712466343382e-06, "loss": 0.2057, "step": 5475 }, { "epoch": 2.393013100436681, "grad_norm": 0.1935316894500808, "learning_rate": 5.988407247972982e-06, "loss": 0.2101, "step": 5480 }, { "epoch": 2.395196506550218, "grad_norm": 0.19014181610486888, "learning_rate": 5.947224733831364e-06, "loss": 0.214, "step": 5485 }, { "epoch": 2.3973799126637556, "grad_norm": 0.2047537532894175, "learning_rate": 5.906165189717855e-06, "loss": 0.2069, "step": 5490 }, { "epoch": 2.3995633187772927, "grad_norm": 0.20244514937691652, "learning_rate": 5.8652288806381215e-06, "loss": 0.2093, "step": 5495 }, { "epoch": 2.4017467248908297, "grad_norm": 0.2029682194929424, "learning_rate": 5.824416070802438e-06, "loss": 0.1989, "step": 5500 }, { "epoch": 2.403930131004367, "grad_norm": 0.19862209705141703, "learning_rate": 5.7837270236239995e-06, "loss": 0.1988, "step": 5505 }, { "epoch": 2.406113537117904, "grad_norm": 0.19816634806985825, "learning_rate": 5.74316200171722e-06, "loss": 0.2042, "step": 5510 }, { "epoch": 2.408296943231441, "grad_norm": 0.20129123615306388, "learning_rate": 5.70272126689601e-06, "loss": 0.2036, "step": 5515 }, { "epoch": 2.410480349344978, "grad_norm": 0.18987017520409014, "learning_rate": 5.662405080172134e-06, "loss": 0.2078, "step": 5520 }, { "epoch": 2.4126637554585155, "grad_norm": 0.19052375793050977, "learning_rate": 5.622213701753487e-06, "loss": 0.2231, "step": 5525 }, { "epoch": 2.4148471615720526, "grad_norm": 0.19323234507341908, "learning_rate": 5.582147391042433e-06, "loss": 0.2092, "step": 5530 }, { "epoch": 2.4170305676855897, "grad_norm": 0.1924708259179898, "learning_rate": 5.542206406634129e-06, "loss": 0.2078, "step": 5535 }, { "epoch": 2.4192139737991267, "grad_norm": 0.207006675227268, "learning_rate": 5.5023910063148445e-06, "loss": 0.2109, "step": 5540 }, { "epoch": 2.421397379912664, "grad_norm": 0.2002952114074735, "learning_rate": 5.46270144706032e-06, "loss": 0.198, "step": 5545 }, { "epoch": 2.423580786026201, "grad_norm": 0.2091856914324174, "learning_rate": 5.423137985034085e-06, "loss": 0.2138, "step": 5550 }, { "epoch": 2.425764192139738, "grad_norm": 0.22941186423542465, "learning_rate": 5.3837008755858145e-06, "loss": 0.2013, "step": 5555 }, { "epoch": 2.427947598253275, "grad_norm": 0.1974516927188841, "learning_rate": 5.3443903732496966e-06, "loss": 0.2069, "step": 5560 }, { "epoch": 2.430131004366812, "grad_norm": 0.1861177274579643, "learning_rate": 5.305206731742748e-06, "loss": 0.2086, "step": 5565 }, { "epoch": 2.432314410480349, "grad_norm": 0.196399769855382, "learning_rate": 5.266150203963224e-06, "loss": 0.2155, "step": 5570 }, { "epoch": 2.4344978165938866, "grad_norm": 0.20583463409268135, "learning_rate": 5.227221041988956e-06, "loss": 0.2055, "step": 5575 }, { "epoch": 2.4366812227074237, "grad_norm": 0.19145217163247738, "learning_rate": 5.188419497075739e-06, "loss": 0.2045, "step": 5580 }, { "epoch": 2.438864628820961, "grad_norm": 0.2300625133365478, "learning_rate": 5.1497458196556995e-06, "loss": 0.2192, "step": 5585 }, { "epoch": 2.441048034934498, "grad_norm": 0.19252331533699785, "learning_rate": 5.111200259335688e-06, "loss": 0.2089, "step": 5590 }, { "epoch": 2.443231441048035, "grad_norm": 0.20014832340012859, "learning_rate": 5.072783064895664e-06, "loss": 0.2243, "step": 5595 }, { "epoch": 2.445414847161572, "grad_norm": 0.19507458711265246, "learning_rate": 5.034494484287086e-06, "loss": 0.2195, "step": 5600 }, { "epoch": 2.447598253275109, "grad_norm": 0.18618941255923363, "learning_rate": 4.996334764631322e-06, "loss": 0.2034, "step": 5605 }, { "epoch": 2.449781659388646, "grad_norm": 0.2223628562088218, "learning_rate": 4.958304152218046e-06, "loss": 0.2076, "step": 5610 }, { "epoch": 2.4519650655021836, "grad_norm": 0.22749716704810904, "learning_rate": 4.920402892503648e-06, "loss": 0.2142, "step": 5615 }, { "epoch": 2.4541484716157207, "grad_norm": 0.20326433502448837, "learning_rate": 4.882631230109655e-06, "loss": 0.206, "step": 5620 }, { "epoch": 2.4563318777292578, "grad_norm": 0.20525401555309347, "learning_rate": 4.844989408821149e-06, "loss": 0.2042, "step": 5625 }, { "epoch": 2.458515283842795, "grad_norm": 0.20210008215232908, "learning_rate": 4.807477671585195e-06, "loss": 0.1999, "step": 5630 }, { "epoch": 2.460698689956332, "grad_norm": 0.21710829320874975, "learning_rate": 4.770096260509274e-06, "loss": 0.203, "step": 5635 }, { "epoch": 2.462882096069869, "grad_norm": 0.18303629687950113, "learning_rate": 4.732845416859713e-06, "loss": 0.2134, "step": 5640 }, { "epoch": 2.465065502183406, "grad_norm": 0.19284778474305014, "learning_rate": 4.695725381060137e-06, "loss": 0.2062, "step": 5645 }, { "epoch": 2.467248908296943, "grad_norm": 0.18825083490369524, "learning_rate": 4.658736392689922e-06, "loss": 0.2214, "step": 5650 }, { "epoch": 2.46943231441048, "grad_norm": 0.20310843234802933, "learning_rate": 4.6218786904826175e-06, "loss": 0.2132, "step": 5655 }, { "epoch": 2.4716157205240172, "grad_norm": 0.21879459935317394, "learning_rate": 4.58515251232445e-06, "loss": 0.2193, "step": 5660 }, { "epoch": 2.4737991266375547, "grad_norm": 0.2064948920674535, "learning_rate": 4.548558095252758e-06, "loss": 0.206, "step": 5665 }, { "epoch": 2.475982532751092, "grad_norm": 0.20819617169964372, "learning_rate": 4.512095675454472e-06, "loss": 0.2116, "step": 5670 }, { "epoch": 2.478165938864629, "grad_norm": 0.19013108098717665, "learning_rate": 4.475765488264588e-06, "loss": 0.2031, "step": 5675 }, { "epoch": 2.480349344978166, "grad_norm": 0.19875013946254505, "learning_rate": 4.439567768164651e-06, "loss": 0.2017, "step": 5680 }, { "epoch": 2.482532751091703, "grad_norm": 0.235432313345855, "learning_rate": 4.4035027487812325e-06, "loss": 0.2119, "step": 5685 }, { "epoch": 2.48471615720524, "grad_norm": 0.19883691503756268, "learning_rate": 4.367570662884438e-06, "loss": 0.1997, "step": 5690 }, { "epoch": 2.486899563318777, "grad_norm": 0.19289587740272107, "learning_rate": 4.331771742386395e-06, "loss": 0.2096, "step": 5695 }, { "epoch": 2.489082969432314, "grad_norm": 0.18568302603914705, "learning_rate": 4.296106218339757e-06, "loss": 0.2056, "step": 5700 }, { "epoch": 2.4912663755458517, "grad_norm": 0.1886391942932925, "learning_rate": 4.260574320936203e-06, "loss": 0.2022, "step": 5705 }, { "epoch": 2.493449781659389, "grad_norm": 0.19736454250903648, "learning_rate": 4.225176279504975e-06, "loss": 0.2139, "step": 5710 }, { "epoch": 2.495633187772926, "grad_norm": 0.20848042828630423, "learning_rate": 4.1899123225113825e-06, "loss": 0.2033, "step": 5715 }, { "epoch": 2.497816593886463, "grad_norm": 0.18156063895028207, "learning_rate": 4.154782677555324e-06, "loss": 0.2055, "step": 5720 }, { "epoch": 2.5, "grad_norm": 0.1915665995869679, "learning_rate": 4.119787571369829e-06, "loss": 0.2056, "step": 5725 }, { "epoch": 2.502183406113537, "grad_norm": 0.20789030655261928, "learning_rate": 4.084927229819591e-06, "loss": 0.1969, "step": 5730 }, { "epoch": 2.504366812227074, "grad_norm": 0.19567843626163597, "learning_rate": 4.050201877899501e-06, "loss": 0.1967, "step": 5735 }, { "epoch": 2.506550218340611, "grad_norm": 0.19524615840080548, "learning_rate": 4.0156117397332074e-06, "loss": 0.2017, "step": 5740 }, { "epoch": 2.5087336244541483, "grad_norm": 0.1851828139613455, "learning_rate": 3.9811570385716705e-06, "loss": 0.2187, "step": 5745 }, { "epoch": 2.5109170305676853, "grad_norm": 0.18768129563315683, "learning_rate": 3.9468379967917025e-06, "loss": 0.2139, "step": 5750 }, { "epoch": 2.513100436681223, "grad_norm": 0.1908316174021218, "learning_rate": 3.912654835894564e-06, "loss": 0.2085, "step": 5755 }, { "epoch": 2.51528384279476, "grad_norm": 0.18887997682497717, "learning_rate": 3.878607776504489e-06, "loss": 0.2089, "step": 5760 }, { "epoch": 2.517467248908297, "grad_norm": 0.18931507612361673, "learning_rate": 3.844697038367314e-06, "loss": 0.2118, "step": 5765 }, { "epoch": 2.519650655021834, "grad_norm": 0.1955389549996342, "learning_rate": 3.8109228403490268e-06, "loss": 0.2041, "step": 5770 }, { "epoch": 2.521834061135371, "grad_norm": 0.1985311672510908, "learning_rate": 3.7772854004343494e-06, "loss": 0.2117, "step": 5775 }, { "epoch": 2.524017467248908, "grad_norm": 0.18193407723396662, "learning_rate": 3.7437849357253555e-06, "loss": 0.2051, "step": 5780 }, { "epoch": 2.5262008733624453, "grad_norm": 0.19363438642248365, "learning_rate": 3.7104216624400502e-06, "loss": 0.2112, "step": 5785 }, { "epoch": 2.5283842794759828, "grad_norm": 0.18728937794992326, "learning_rate": 3.677195795910984e-06, "loss": 0.2052, "step": 5790 }, { "epoch": 2.53056768558952, "grad_norm": 0.17522592702667172, "learning_rate": 3.6441075505838586e-06, "loss": 0.2108, "step": 5795 }, { "epoch": 2.532751091703057, "grad_norm": 0.19669895879396115, "learning_rate": 3.6111571400161387e-06, "loss": 0.2107, "step": 5800 }, { "epoch": 2.534934497816594, "grad_norm": 0.2086966810335374, "learning_rate": 3.578344776875689e-06, "loss": 0.2025, "step": 5805 }, { "epoch": 2.537117903930131, "grad_norm": 0.18427750014405722, "learning_rate": 3.5456706729393774e-06, "loss": 0.2089, "step": 5810 }, { "epoch": 2.539301310043668, "grad_norm": 0.19759644279021854, "learning_rate": 3.5131350390917333e-06, "loss": 0.2136, "step": 5815 }, { "epoch": 2.541484716157205, "grad_norm": 0.18832000886509567, "learning_rate": 3.4807380853235704e-06, "loss": 0.2206, "step": 5820 }, { "epoch": 2.5436681222707422, "grad_norm": 0.1939256790542837, "learning_rate": 3.4484800207306405e-06, "loss": 0.2104, "step": 5825 }, { "epoch": 2.5458515283842793, "grad_norm": 0.1981092726688764, "learning_rate": 3.4163610535122738e-06, "loss": 0.2082, "step": 5830 }, { "epoch": 2.5480349344978164, "grad_norm": 0.20236118691003865, "learning_rate": 3.384381390970048e-06, "loss": 0.2046, "step": 5835 }, { "epoch": 2.5502183406113534, "grad_norm": 0.19004473471111177, "learning_rate": 3.3525412395064375e-06, "loss": 0.2023, "step": 5840 }, { "epoch": 2.552401746724891, "grad_norm": 0.20444360368222714, "learning_rate": 3.32084080462349e-06, "loss": 0.2027, "step": 5845 }, { "epoch": 2.554585152838428, "grad_norm": 0.2175490290108786, "learning_rate": 3.2892802909214955e-06, "loss": 0.2096, "step": 5850 }, { "epoch": 2.556768558951965, "grad_norm": 0.19247531436865034, "learning_rate": 3.2578599020976707e-06, "loss": 0.1951, "step": 5855 }, { "epoch": 2.558951965065502, "grad_norm": 0.1874278832882864, "learning_rate": 3.2265798409448376e-06, "loss": 0.211, "step": 5860 }, { "epoch": 2.561135371179039, "grad_norm": 0.19407368076943168, "learning_rate": 3.1954403093501227e-06, "loss": 0.2179, "step": 5865 }, { "epoch": 2.5633187772925763, "grad_norm": 0.176804419787979, "learning_rate": 3.1644415082936417e-06, "loss": 0.2017, "step": 5870 }, { "epoch": 2.565502183406114, "grad_norm": 0.1936425739896731, "learning_rate": 3.1335836378472233e-06, "loss": 0.1963, "step": 5875 }, { "epoch": 2.567685589519651, "grad_norm": 0.18623478504104157, "learning_rate": 3.10286689717309e-06, "loss": 0.2076, "step": 5880 }, { "epoch": 2.569868995633188, "grad_norm": 0.18973768721599718, "learning_rate": 3.0722914845225963e-06, "loss": 0.211, "step": 5885 }, { "epoch": 2.572052401746725, "grad_norm": 0.1887388707833972, "learning_rate": 3.04185759723494e-06, "loss": 0.1983, "step": 5890 }, { "epoch": 2.574235807860262, "grad_norm": 0.18187528811137918, "learning_rate": 3.0115654317358884e-06, "loss": 0.2119, "step": 5895 }, { "epoch": 2.576419213973799, "grad_norm": 0.19057590033301217, "learning_rate": 2.981415183536504e-06, "loss": 0.212, "step": 5900 }, { "epoch": 2.578602620087336, "grad_norm": 0.18728042390500685, "learning_rate": 2.9514070472318965e-06, "loss": 0.1956, "step": 5905 }, { "epoch": 2.5807860262008733, "grad_norm": 0.18967646290907864, "learning_rate": 2.9215412164999594e-06, "loss": 0.2006, "step": 5910 }, { "epoch": 2.5829694323144103, "grad_norm": 0.19120785023883197, "learning_rate": 2.891817884100123e-06, "loss": 0.192, "step": 5915 }, { "epoch": 2.5851528384279474, "grad_norm": 0.18671520762510882, "learning_rate": 2.862237241872101e-06, "loss": 0.2215, "step": 5920 }, { "epoch": 2.5873362445414845, "grad_norm": 0.1891615762513396, "learning_rate": 2.832799480734663e-06, "loss": 0.2244, "step": 5925 }, { "epoch": 2.589519650655022, "grad_norm": 0.1807780047483408, "learning_rate": 2.8035047906844e-06, "loss": 0.2053, "step": 5930 }, { "epoch": 2.591703056768559, "grad_norm": 0.18284608120565451, "learning_rate": 2.774353360794493e-06, "loss": 0.2047, "step": 5935 }, { "epoch": 2.593886462882096, "grad_norm": 0.18110685745874158, "learning_rate": 2.7453453792134986e-06, "loss": 0.213, "step": 5940 }, { "epoch": 2.596069868995633, "grad_norm": 0.18936818198939037, "learning_rate": 2.7164810331641366e-06, "loss": 0.2038, "step": 5945 }, { "epoch": 2.5982532751091703, "grad_norm": 0.19029850620608788, "learning_rate": 2.687760508942064e-06, "loss": 0.2116, "step": 5950 }, { "epoch": 2.6004366812227073, "grad_norm": 0.19115405588229892, "learning_rate": 2.659183991914696e-06, "loss": 0.2145, "step": 5955 }, { "epoch": 2.6026200873362444, "grad_norm": 0.1945513238775359, "learning_rate": 2.6307516665200012e-06, "loss": 0.208, "step": 5960 }, { "epoch": 2.604803493449782, "grad_norm": 0.19282829361579937, "learning_rate": 2.602463716265302e-06, "loss": 0.2179, "step": 5965 }, { "epoch": 2.606986899563319, "grad_norm": 0.2002570604911309, "learning_rate": 2.5743203237261065e-06, "loss": 0.2095, "step": 5970 }, { "epoch": 2.609170305676856, "grad_norm": 0.18334332369689957, "learning_rate": 2.5463216705449155e-06, "loss": 0.2185, "step": 5975 }, { "epoch": 2.611353711790393, "grad_norm": 0.181736048243745, "learning_rate": 2.5184679374300553e-06, "loss": 0.2085, "step": 5980 }, { "epoch": 2.61353711790393, "grad_norm": 0.17385940038429637, "learning_rate": 2.4907593041545163e-06, "loss": 0.2167, "step": 5985 }, { "epoch": 2.6157205240174672, "grad_norm": 0.17922911028497665, "learning_rate": 2.4631959495547833e-06, "loss": 0.1995, "step": 5990 }, { "epoch": 2.6179039301310043, "grad_norm": 0.17779619304763405, "learning_rate": 2.4357780515297e-06, "loss": 0.1941, "step": 5995 }, { "epoch": 2.6200873362445414, "grad_norm": 0.19092236134949003, "learning_rate": 2.408505787039281e-06, "loss": 0.2023, "step": 6000 }, { "epoch": 2.6222707423580784, "grad_norm": 0.21537785339457424, "learning_rate": 2.381379332103623e-06, "loss": 0.2039, "step": 6005 }, { "epoch": 2.6244541484716155, "grad_norm": 0.19076636501936398, "learning_rate": 2.3543988618017237e-06, "loss": 0.2007, "step": 6010 }, { "epoch": 2.6266375545851526, "grad_norm": 0.18091299115157525, "learning_rate": 2.327564550270381e-06, "loss": 0.2003, "step": 6015 }, { "epoch": 2.62882096069869, "grad_norm": 0.17652429068296252, "learning_rate": 2.3008765707030532e-06, "loss": 0.1987, "step": 6020 }, { "epoch": 2.631004366812227, "grad_norm": 0.19337565491737743, "learning_rate": 2.2743350953487424e-06, "loss": 0.2161, "step": 6025 }, { "epoch": 2.6331877729257642, "grad_norm": 0.19521279815086381, "learning_rate": 2.2479402955108945e-06, "loss": 0.1994, "step": 6030 }, { "epoch": 2.6353711790393013, "grad_norm": 0.18448803162094932, "learning_rate": 2.2216923415462802e-06, "loss": 0.2071, "step": 6035 }, { "epoch": 2.6375545851528384, "grad_norm": 0.19952616634096593, "learning_rate": 2.1955914028639e-06, "loss": 0.1997, "step": 6040 }, { "epoch": 2.6397379912663754, "grad_norm": 0.17820128872734733, "learning_rate": 2.1696376479238947e-06, "loss": 0.2095, "step": 6045 }, { "epoch": 2.641921397379913, "grad_norm": 0.18942931129920965, "learning_rate": 2.1438312442364467e-06, "loss": 0.2065, "step": 6050 }, { "epoch": 2.64410480349345, "grad_norm": 0.1880722865584005, "learning_rate": 2.1181723583607177e-06, "loss": 0.2011, "step": 6055 }, { "epoch": 2.646288209606987, "grad_norm": 0.19525082185206008, "learning_rate": 2.092661155903758e-06, "loss": 0.2075, "step": 6060 }, { "epoch": 2.648471615720524, "grad_norm": 0.17350084934940418, "learning_rate": 2.067297801519441e-06, "loss": 0.1998, "step": 6065 }, { "epoch": 2.650655021834061, "grad_norm": 0.18323734039467143, "learning_rate": 2.0420824589074073e-06, "loss": 0.2066, "step": 6070 }, { "epoch": 2.6528384279475983, "grad_norm": 0.19042032301324274, "learning_rate": 2.0170152908119964e-06, "loss": 0.2078, "step": 6075 }, { "epoch": 2.6550218340611353, "grad_norm": 0.1822344566119162, "learning_rate": 1.9920964590212088e-06, "loss": 0.2026, "step": 6080 }, { "epoch": 2.6572052401746724, "grad_norm": 0.18350664419747995, "learning_rate": 1.967326124365651e-06, "loss": 0.2046, "step": 6085 }, { "epoch": 2.6593886462882095, "grad_norm": 0.17900525181219676, "learning_rate": 1.9427044467175054e-06, "loss": 0.2053, "step": 6090 }, { "epoch": 2.6615720524017465, "grad_norm": 0.18662259458308622, "learning_rate": 1.9182315849894996e-06, "loss": 0.1969, "step": 6095 }, { "epoch": 2.6637554585152836, "grad_norm": 0.18210726045913153, "learning_rate": 1.8939076971338592e-06, "loss": 0.2133, "step": 6100 }, { "epoch": 2.665938864628821, "grad_norm": 0.18936892016307003, "learning_rate": 1.8697329401413333e-06, "loss": 0.2036, "step": 6105 }, { "epoch": 2.668122270742358, "grad_norm": 0.17873699633744589, "learning_rate": 1.8457074700401284e-06, "loss": 0.2015, "step": 6110 }, { "epoch": 2.6703056768558953, "grad_norm": 0.19259954300544802, "learning_rate": 1.821831441894939e-06, "loss": 0.2168, "step": 6115 }, { "epoch": 2.6724890829694323, "grad_norm": 0.1765084388420357, "learning_rate": 1.7981050098059364e-06, "loss": 0.2051, "step": 6120 }, { "epoch": 2.6746724890829694, "grad_norm": 0.18953853048678723, "learning_rate": 1.7745283269077622e-06, "loss": 0.1945, "step": 6125 }, { "epoch": 2.6768558951965065, "grad_norm": 0.17824046665997959, "learning_rate": 1.751101545368558e-06, "loss": 0.1987, "step": 6130 }, { "epoch": 2.6790393013100435, "grad_norm": 0.1839474436743472, "learning_rate": 1.7278248163889787e-06, "loss": 0.2297, "step": 6135 }, { "epoch": 2.681222707423581, "grad_norm": 0.198016592733994, "learning_rate": 1.7046982902011954e-06, "loss": 0.2004, "step": 6140 }, { "epoch": 2.683406113537118, "grad_norm": 0.1754478072579575, "learning_rate": 1.6817221160679603e-06, "loss": 0.2165, "step": 6145 }, { "epoch": 2.685589519650655, "grad_norm": 0.1739282991837851, "learning_rate": 1.658896442281624e-06, "loss": 0.2139, "step": 6150 }, { "epoch": 2.6877729257641922, "grad_norm": 0.1752625270758878, "learning_rate": 1.636221416163175e-06, "loss": 0.1918, "step": 6155 }, { "epoch": 2.6899563318777293, "grad_norm": 0.1779820497152234, "learning_rate": 1.6136971840612996e-06, "loss": 0.1899, "step": 6160 }, { "epoch": 2.6921397379912664, "grad_norm": 0.18605348108430886, "learning_rate": 1.5913238913514317e-06, "loss": 0.2034, "step": 6165 }, { "epoch": 2.6943231441048034, "grad_norm": 0.18327039260922082, "learning_rate": 1.569101682434812e-06, "loss": 0.2151, "step": 6170 }, { "epoch": 2.6965065502183405, "grad_norm": 0.18420649704706576, "learning_rate": 1.5470307007375617e-06, "loss": 0.1999, "step": 6175 }, { "epoch": 2.6986899563318776, "grad_norm": 0.18496496761599715, "learning_rate": 1.5251110887097547e-06, "loss": 0.2015, "step": 6180 }, { "epoch": 2.7008733624454146, "grad_norm": 0.18359525195308157, "learning_rate": 1.5033429878245047e-06, "loss": 0.1964, "step": 6185 }, { "epoch": 2.7030567685589517, "grad_norm": 0.18877940096775106, "learning_rate": 1.4817265385770273e-06, "loss": 0.219, "step": 6190 }, { "epoch": 2.7052401746724892, "grad_norm": 0.17210679893859862, "learning_rate": 1.460261880483768e-06, "loss": 0.2087, "step": 6195 }, { "epoch": 2.7074235807860263, "grad_norm": 0.18497422573663475, "learning_rate": 1.4389491520814785e-06, "loss": 0.2073, "step": 6200 }, { "epoch": 2.7096069868995634, "grad_norm": 0.18280651214761798, "learning_rate": 1.417788490926328e-06, "loss": 0.2049, "step": 6205 }, { "epoch": 2.7117903930131004, "grad_norm": 0.21697399944996257, "learning_rate": 1.3967800335930203e-06, "loss": 0.1966, "step": 6210 }, { "epoch": 2.7139737991266375, "grad_norm": 0.17484224090531814, "learning_rate": 1.3759239156739046e-06, "loss": 0.2004, "step": 6215 }, { "epoch": 2.7161572052401746, "grad_norm": 0.17922630540901408, "learning_rate": 1.355220271778107e-06, "loss": 0.2171, "step": 6220 }, { "epoch": 2.7183406113537116, "grad_norm": 0.1788064963960434, "learning_rate": 1.3346692355306583e-06, "loss": 0.21, "step": 6225 }, { "epoch": 2.720524017467249, "grad_norm": 0.17658228011851998, "learning_rate": 1.3142709395716325e-06, "loss": 0.2041, "step": 6230 }, { "epoch": 2.722707423580786, "grad_norm": 0.19747059238001105, "learning_rate": 1.294025515555289e-06, "loss": 0.2141, "step": 6235 }, { "epoch": 2.7248908296943233, "grad_norm": 0.19234938030154514, "learning_rate": 1.2739330941492211e-06, "loss": 0.2194, "step": 6240 }, { "epoch": 2.7270742358078603, "grad_norm": 0.18116820313254028, "learning_rate": 1.253993805033521e-06, "loss": 0.2026, "step": 6245 }, { "epoch": 2.7292576419213974, "grad_norm": 0.18091742916208972, "learning_rate": 1.2342077768999372e-06, "loss": 0.2074, "step": 6250 }, { "epoch": 2.7314410480349345, "grad_norm": 0.18173772524843324, "learning_rate": 1.2145751374510462e-06, "loss": 0.1982, "step": 6255 }, { "epoch": 2.7336244541484715, "grad_norm": 0.18111352269398215, "learning_rate": 1.1950960133994189e-06, "loss": 0.2113, "step": 6260 }, { "epoch": 2.7358078602620086, "grad_norm": 0.17797289847881081, "learning_rate": 1.175770530466816e-06, "loss": 0.1985, "step": 6265 }, { "epoch": 2.7379912663755457, "grad_norm": 0.18092908617426118, "learning_rate": 1.156598813383372e-06, "loss": 0.2228, "step": 6270 }, { "epoch": 2.7401746724890828, "grad_norm": 0.18444520409610834, "learning_rate": 1.1375809858867853e-06, "loss": 0.2114, "step": 6275 }, { "epoch": 2.74235807860262, "grad_norm": 0.1725831612805252, "learning_rate": 1.1187171707215228e-06, "loss": 0.2016, "step": 6280 }, { "epoch": 2.7445414847161573, "grad_norm": 0.1822891732232196, "learning_rate": 1.1000074896380386e-06, "loss": 0.2046, "step": 6285 }, { "epoch": 2.7467248908296944, "grad_norm": 0.1837802202840671, "learning_rate": 1.0814520633919577e-06, "loss": 0.2063, "step": 6290 }, { "epoch": 2.7489082969432315, "grad_norm": 0.19854848233423214, "learning_rate": 1.063051011743335e-06, "loss": 0.2026, "step": 6295 }, { "epoch": 2.7510917030567685, "grad_norm": 0.1690617798747192, "learning_rate": 1.0448044534558582e-06, "loss": 0.2042, "step": 6300 }, { "epoch": 2.7532751091703056, "grad_norm": 0.18680449525384113, "learning_rate": 1.026712506296082e-06, "loss": 0.2198, "step": 6305 }, { "epoch": 2.7554585152838427, "grad_norm": 0.1690572517735115, "learning_rate": 1.0087752870326817e-06, "loss": 0.2128, "step": 6310 }, { "epoch": 2.75764192139738, "grad_norm": 0.1741781541277763, "learning_rate": 9.909929114356842e-07, "loss": 0.2026, "step": 6315 }, { "epoch": 2.7598253275109172, "grad_norm": 0.17546914742624392, "learning_rate": 9.733654942757353e-07, "loss": 0.1994, "step": 6320 }, { "epoch": 2.7620087336244543, "grad_norm": 0.17660338305720158, "learning_rate": 9.558931493233392e-07, "loss": 0.204, "step": 6325 }, { "epoch": 2.7641921397379914, "grad_norm": 0.1769127719144767, "learning_rate": 9.385759893481477e-07, "loss": 0.1996, "step": 6330 }, { "epoch": 2.7663755458515285, "grad_norm": 0.17837387586213685, "learning_rate": 9.214141261182196e-07, "loss": 0.2156, "step": 6335 }, { "epoch": 2.7685589519650655, "grad_norm": 0.18449098828135574, "learning_rate": 9.044076703992904e-07, "loss": 0.208, "step": 6340 }, { "epoch": 2.7707423580786026, "grad_norm": 0.18352552163108934, "learning_rate": 8.875567319540784e-07, "loss": 0.2074, "step": 6345 }, { "epoch": 2.7729257641921397, "grad_norm": 0.18503191612346148, "learning_rate": 8.708614195415604e-07, "loss": 0.2, "step": 6350 }, { "epoch": 2.7751091703056767, "grad_norm": 0.1937080570900333, "learning_rate": 8.543218409162778e-07, "loss": 0.2082, "step": 6355 }, { "epoch": 2.777292576419214, "grad_norm": 0.18510996106818742, "learning_rate": 8.379381028276401e-07, "loss": 0.2056, "step": 6360 }, { "epoch": 2.779475982532751, "grad_norm": 0.1777159695537935, "learning_rate": 8.217103110192276e-07, "loss": 0.2061, "step": 6365 }, { "epoch": 2.7816593886462884, "grad_norm": 0.1752542433085556, "learning_rate": 8.056385702281177e-07, "loss": 0.2007, "step": 6370 }, { "epoch": 2.7838427947598254, "grad_norm": 0.1789778389216568, "learning_rate": 7.89722984184213e-07, "loss": 0.2046, "step": 6375 }, { "epoch": 2.7860262008733625, "grad_norm": 0.17685450533969785, "learning_rate": 7.739636556095525e-07, "loss": 0.2065, "step": 6380 }, { "epoch": 2.7882096069868996, "grad_norm": 0.186062026864878, "learning_rate": 7.583606862176712e-07, "loss": 0.2032, "step": 6385 }, { "epoch": 2.7903930131004366, "grad_norm": 0.16820621198730876, "learning_rate": 7.429141767129333e-07, "loss": 0.2037, "step": 6390 }, { "epoch": 2.7925764192139737, "grad_norm": 0.18272612053201714, "learning_rate": 7.276242267898775e-07, "loss": 0.2077, "step": 6395 }, { "epoch": 2.7947598253275108, "grad_norm": 0.1818746307663275, "learning_rate": 7.12490935132587e-07, "loss": 0.2126, "step": 6400 }, { "epoch": 2.7969432314410483, "grad_norm": 0.1755002936425013, "learning_rate": 6.97514399414037e-07, "loss": 0.2032, "step": 6405 }, { "epoch": 2.7991266375545854, "grad_norm": 0.18013946700206887, "learning_rate": 6.826947162954733e-07, "loss": 0.2066, "step": 6410 }, { "epoch": 2.8013100436681224, "grad_norm": 0.17849038284790902, "learning_rate": 6.68031981425793e-07, "loss": 0.2055, "step": 6415 }, { "epoch": 2.8034934497816595, "grad_norm": 0.21383468189806515, "learning_rate": 6.535262894409116e-07, "loss": 0.2171, "step": 6420 }, { "epoch": 2.8056768558951966, "grad_norm": 0.1707908400592636, "learning_rate": 6.391777339631727e-07, "loss": 0.1968, "step": 6425 }, { "epoch": 2.8078602620087336, "grad_norm": 0.1928726259141998, "learning_rate": 6.249864076007167e-07, "loss": 0.2078, "step": 6430 }, { "epoch": 2.8100436681222707, "grad_norm": 0.18080723535203885, "learning_rate": 6.109524019469154e-07, "loss": 0.2052, "step": 6435 }, { "epoch": 2.8122270742358078, "grad_norm": 0.1785838290629133, "learning_rate": 5.970758075797556e-07, "loss": 0.1978, "step": 6440 }, { "epoch": 2.814410480349345, "grad_norm": 0.17615554784359172, "learning_rate": 5.833567140612673e-07, "loss": 0.205, "step": 6445 }, { "epoch": 2.816593886462882, "grad_norm": 0.18081617254199714, "learning_rate": 5.69795209936938e-07, "loss": 0.2085, "step": 6450 }, { "epoch": 2.818777292576419, "grad_norm": 0.19233898776215452, "learning_rate": 5.563913827351469e-07, "loss": 0.2094, "step": 6455 }, { "epoch": 2.8209606986899565, "grad_norm": 0.18538569170027117, "learning_rate": 5.431453189666036e-07, "loss": 0.2133, "step": 6460 }, { "epoch": 2.8231441048034935, "grad_norm": 0.16627030997203804, "learning_rate": 5.30057104123774e-07, "loss": 0.1946, "step": 6465 }, { "epoch": 2.8253275109170306, "grad_norm": 0.18546465605942317, "learning_rate": 5.171268226803416e-07, "loss": 0.2014, "step": 6470 }, { "epoch": 2.8275109170305677, "grad_norm": 0.18279009388047487, "learning_rate": 5.043545580906695e-07, "loss": 0.2228, "step": 6475 }, { "epoch": 2.8296943231441047, "grad_norm": 0.17398692192324666, "learning_rate": 4.917403927892389e-07, "loss": 0.208, "step": 6480 }, { "epoch": 2.831877729257642, "grad_norm": 0.17416683242727254, "learning_rate": 4.792844081901309e-07, "loss": 0.1994, "step": 6485 }, { "epoch": 2.834061135371179, "grad_norm": 0.177823166130051, "learning_rate": 4.6698668468651006e-07, "loss": 0.1939, "step": 6490 }, { "epoch": 2.8362445414847164, "grad_norm": 0.19673421756206907, "learning_rate": 4.548473016500854e-07, "loss": 0.2161, "step": 6495 }, { "epoch": 2.8384279475982535, "grad_norm": 0.1811334337251195, "learning_rate": 4.4286633743061413e-07, "loss": 0.2082, "step": 6500 }, { "epoch": 2.8406113537117905, "grad_norm": 0.17564798960003072, "learning_rate": 4.310438693553853e-07, "loss": 0.2066, "step": 6505 }, { "epoch": 2.8427947598253276, "grad_norm": 0.1794074378421751, "learning_rate": 4.193799737287285e-07, "loss": 0.2093, "step": 6510 }, { "epoch": 2.8449781659388647, "grad_norm": 0.18373063476787263, "learning_rate": 4.0787472583151964e-07, "loss": 0.2235, "step": 6515 }, { "epoch": 2.8471615720524017, "grad_norm": 0.17915751006104885, "learning_rate": 3.965281999206899e-07, "loss": 0.1956, "step": 6520 }, { "epoch": 2.849344978165939, "grad_norm": 0.17324382598114016, "learning_rate": 3.853404692287538e-07, "loss": 0.2074, "step": 6525 }, { "epoch": 2.851528384279476, "grad_norm": 0.18692578826449793, "learning_rate": 3.743116059633317e-07, "loss": 0.2174, "step": 6530 }, { "epoch": 2.853711790393013, "grad_norm": 0.18933592310694983, "learning_rate": 3.6344168130668097e-07, "loss": 0.2058, "step": 6535 }, { "epoch": 2.85589519650655, "grad_norm": 0.18309703282903245, "learning_rate": 3.527307654152462e-07, "loss": 0.203, "step": 6540 }, { "epoch": 2.858078602620087, "grad_norm": 0.17336689816177772, "learning_rate": 3.421789274192011e-07, "loss": 0.189, "step": 6545 }, { "epoch": 2.8602620087336246, "grad_norm": 0.17273324095781836, "learning_rate": 3.3178623542199916e-07, "loss": 0.2009, "step": 6550 }, { "epoch": 2.8624454148471616, "grad_norm": 0.18054845843338757, "learning_rate": 3.2155275649994045e-07, "loss": 0.2125, "step": 6555 }, { "epoch": 2.8646288209606987, "grad_norm": 0.17486255516315685, "learning_rate": 3.1147855670172755e-07, "loss": 0.2014, "step": 6560 }, { "epoch": 2.8668122270742358, "grad_norm": 0.16994065735909195, "learning_rate": 3.015637010480576e-07, "loss": 0.1991, "step": 6565 }, { "epoch": 2.868995633187773, "grad_norm": 0.17823602403047226, "learning_rate": 2.918082535311839e-07, "loss": 0.2107, "step": 6570 }, { "epoch": 2.87117903930131, "grad_norm": 0.18534656992358386, "learning_rate": 2.822122771145158e-07, "loss": 0.1988, "step": 6575 }, { "epoch": 2.8733624454148474, "grad_norm": 0.17359168867052116, "learning_rate": 2.727758337322056e-07, "loss": 0.2006, "step": 6580 }, { "epoch": 2.8755458515283845, "grad_norm": 0.1818021640887144, "learning_rate": 2.6349898428875143e-07, "loss": 0.2107, "step": 6585 }, { "epoch": 2.8777292576419216, "grad_norm": 0.187022315664445, "learning_rate": 2.543817886586003e-07, "loss": 0.2109, "step": 6590 }, { "epoch": 2.8799126637554586, "grad_norm": 0.18065592152437351, "learning_rate": 2.454243056857708e-07, "loss": 0.2155, "step": 6595 }, { "epoch": 2.8820960698689957, "grad_norm": 0.1803124984333343, "learning_rate": 2.366265931834588e-07, "loss": 0.2005, "step": 6600 }, { "epoch": 2.8842794759825328, "grad_norm": 0.17978461103858198, "learning_rate": 2.2798870793368787e-07, "loss": 0.2135, "step": 6605 }, { "epoch": 2.88646288209607, "grad_norm": 0.1838336256596766, "learning_rate": 2.1951070568691222e-07, "loss": 0.2131, "step": 6610 }, { "epoch": 2.888646288209607, "grad_norm": 0.18102300117813305, "learning_rate": 2.1119264116168104e-07, "loss": 0.2064, "step": 6615 }, { "epoch": 2.890829694323144, "grad_norm": 0.1743140768844891, "learning_rate": 2.030345680442719e-07, "loss": 0.1962, "step": 6620 }, { "epoch": 2.893013100436681, "grad_norm": 0.18304778762459922, "learning_rate": 1.950365389883524e-07, "loss": 0.2145, "step": 6625 }, { "epoch": 2.895196506550218, "grad_norm": 0.17457199470182785, "learning_rate": 1.8719860561463853e-07, "loss": 0.1986, "step": 6630 }, { "epoch": 2.8973799126637556, "grad_norm": 0.17803797345436936, "learning_rate": 1.7952081851055346e-07, "loss": 0.2087, "step": 6635 }, { "epoch": 2.8995633187772927, "grad_norm": 0.19249639309638913, "learning_rate": 1.720032272299138e-07, "loss": 0.208, "step": 6640 }, { "epoch": 2.9017467248908297, "grad_norm": 0.1931146601533778, "learning_rate": 1.6464588029260485e-07, "loss": 0.2159, "step": 6645 }, { "epoch": 2.903930131004367, "grad_norm": 0.17050733886371794, "learning_rate": 1.5744882518425875e-07, "loss": 0.2079, "step": 6650 }, { "epoch": 2.906113537117904, "grad_norm": 0.1821050184025676, "learning_rate": 1.5041210835596287e-07, "loss": 0.2113, "step": 6655 }, { "epoch": 2.908296943231441, "grad_norm": 0.1767629872679953, "learning_rate": 1.4353577522394635e-07, "loss": 0.2043, "step": 6660 }, { "epoch": 2.910480349344978, "grad_norm": 0.1770612573000096, "learning_rate": 1.3681987016929963e-07, "loss": 0.2086, "step": 6665 }, { "epoch": 2.9126637554585155, "grad_norm": 0.17785323864755895, "learning_rate": 1.3026443653767762e-07, "loss": 0.2004, "step": 6670 }, { "epoch": 2.9148471615720526, "grad_norm": 0.17642382590921754, "learning_rate": 1.2386951663901913e-07, "loss": 0.206, "step": 6675 }, { "epoch": 2.9170305676855897, "grad_norm": 0.17596251154223802, "learning_rate": 1.1763515174729179e-07, "loss": 0.1986, "step": 6680 }, { "epoch": 2.9192139737991267, "grad_norm": 0.1727279406074698, "learning_rate": 1.115613821002004e-07, "loss": 0.2038, "step": 6685 }, { "epoch": 2.921397379912664, "grad_norm": 0.1761644833661898, "learning_rate": 1.0564824689894282e-07, "loss": 0.2117, "step": 6690 }, { "epoch": 2.923580786026201, "grad_norm": 0.1772235899771806, "learning_rate": 9.989578430795732e-08, "loss": 0.2014, "step": 6695 }, { "epoch": 2.925764192139738, "grad_norm": 0.17942614591659065, "learning_rate": 9.430403145466448e-08, "loss": 0.2109, "step": 6700 }, { "epoch": 2.927947598253275, "grad_norm": 0.17253943064578878, "learning_rate": 8.887302442924517e-08, "loss": 0.2003, "step": 6705 }, { "epoch": 2.930131004366812, "grad_norm": 0.17852850650326985, "learning_rate": 8.360279828438789e-08, "loss": 0.2094, "step": 6710 }, { "epoch": 2.932314410480349, "grad_norm": 0.17307368947339777, "learning_rate": 7.849338703508069e-08, "loss": 0.2064, "step": 6715 }, { "epoch": 2.934497816593886, "grad_norm": 0.1793388736193528, "learning_rate": 7.3544823658378e-08, "loss": 0.1927, "step": 6720 }, { "epoch": 2.9366812227074237, "grad_norm": 0.1803659744144063, "learning_rate": 6.875714009319523e-08, "loss": 0.2076, "step": 6725 }, { "epoch": 2.938864628820961, "grad_norm": 0.17865387702225788, "learning_rate": 6.413036724010058e-08, "loss": 0.2003, "step": 6730 }, { "epoch": 2.941048034934498, "grad_norm": 0.17741136523028206, "learning_rate": 5.966453496111524e-08, "loss": 0.2144, "step": 6735 }, { "epoch": 2.943231441048035, "grad_norm": 0.17550555860513464, "learning_rate": 5.5359672079524637e-08, "loss": 0.2153, "step": 6740 }, { "epoch": 2.945414847161572, "grad_norm": 0.17696250741970354, "learning_rate": 5.121580637968137e-08, "loss": 0.2054, "step": 6745 }, { "epoch": 2.947598253275109, "grad_norm": 0.17454155560080192, "learning_rate": 4.7232964606841454e-08, "loss": 0.1971, "step": 6750 }, { "epoch": 2.9497816593886466, "grad_norm": 0.17698232159770602, "learning_rate": 4.341117246697834e-08, "loss": 0.2085, "step": 6755 }, { "epoch": 2.9519650655021836, "grad_norm": 0.17354434091915077, "learning_rate": 3.9750454626624745e-08, "loss": 0.2137, "step": 6760 }, { "epoch": 2.9541484716157207, "grad_norm": 0.17097280016052646, "learning_rate": 3.6250834712708846e-08, "loss": 0.1945, "step": 6765 }, { "epoch": 2.9563318777292578, "grad_norm": 0.17551121068500722, "learning_rate": 3.2912335312407226e-08, "loss": 0.1865, "step": 6770 }, { "epoch": 2.958515283842795, "grad_norm": 0.17583508061228037, "learning_rate": 2.9734977972989407e-08, "loss": 0.2107, "step": 6775 }, { "epoch": 2.960698689956332, "grad_norm": 0.17683109325722057, "learning_rate": 2.6718783201692964e-08, "loss": 0.2127, "step": 6780 }, { "epoch": 2.962882096069869, "grad_norm": 0.18803887975352115, "learning_rate": 2.3863770465576418e-08, "loss": 0.2156, "step": 6785 }, { "epoch": 2.965065502183406, "grad_norm": 0.1771870084804161, "learning_rate": 2.116995819140821e-08, "loss": 0.2182, "step": 6790 }, { "epoch": 2.967248908296943, "grad_norm": 0.18459436255752454, "learning_rate": 1.8637363765536264e-08, "loss": 0.2009, "step": 6795 }, { "epoch": 2.96943231441048, "grad_norm": 0.17264295233596325, "learning_rate": 1.6266003533774166e-08, "loss": 0.201, "step": 6800 }, { "epoch": 2.9716157205240172, "grad_norm": 0.18069190701339236, "learning_rate": 1.405589280130959e-08, "loss": 0.214, "step": 6805 }, { "epoch": 2.9737991266375547, "grad_norm": 0.17111956645749835, "learning_rate": 1.2007045832590492e-08, "loss": 0.1958, "step": 6810 }, { "epoch": 2.975982532751092, "grad_norm": 0.17292900306530656, "learning_rate": 1.0119475851247395e-08, "loss": 0.1952, "step": 6815 }, { "epoch": 2.978165938864629, "grad_norm": 0.1806661006510654, "learning_rate": 8.393195039993473e-09, "loss": 0.207, "step": 6820 }, { "epoch": 2.980349344978166, "grad_norm": 0.1805502156590632, "learning_rate": 6.82821454055238e-09, "loss": 0.2077, "step": 6825 }, { "epoch": 2.982532751091703, "grad_norm": 0.17317279258008123, "learning_rate": 5.42454445359164e-09, "loss": 0.2, "step": 6830 }, { "epoch": 2.98471615720524, "grad_norm": 0.17476834735543392, "learning_rate": 4.1821938386477075e-09, "loss": 0.2028, "step": 6835 }, { "epoch": 2.986899563318777, "grad_norm": 0.18884495387086833, "learning_rate": 3.1011707140732317e-09, "loss": 0.2119, "step": 6840 }, { "epoch": 2.9890829694323147, "grad_norm": 0.17552656950302298, "learning_rate": 2.1814820569898698e-09, "loss": 0.2146, "step": 6845 }, { "epoch": 2.9912663755458517, "grad_norm": 0.17245278444580078, "learning_rate": 1.4231338032300034e-09, "loss": 0.213, "step": 6850 }, { "epoch": 2.993449781659389, "grad_norm": 0.17656984635938758, "learning_rate": 8.261308473089813e-10, "loss": 0.2045, "step": 6855 }, { "epoch": 2.995633187772926, "grad_norm": 0.18334478736300916, "learning_rate": 3.9047704239181247e-10, "loss": 0.2067, "step": 6860 }, { "epoch": 2.997816593886463, "grad_norm": 0.1826422374600176, "learning_rate": 1.1617520026818706e-10, "loss": 0.1943, "step": 6865 }, { "epoch": 3.0, "grad_norm": 0.1888467433612867, "learning_rate": 3.2270913274956573e-12, "loss": 0.1971, "step": 6870 }, { "epoch": 3.0, "step": 6870, "total_flos": 1.440651216420864e+16, "train_loss": 0.30719185047274594, "train_runtime": 200812.4774, "train_samples_per_second": 1.094, "train_steps_per_second": 0.034 } ], "logging_steps": 5, "max_steps": 6870, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 10000000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1.440651216420864e+16, "train_batch_size": 1, "trial_name": null, "trial_params": null }