{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 5.659309564233164, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.005659309564233163, "grad_norm": 14.506192207336426, "learning_rate": 1.8e-06, "loss": 1.1688, "step": 10 }, { "epoch": 0.011318619128466326, "grad_norm": 8.839621543884277, "learning_rate": 3.8e-06, "loss": 0.9318, "step": 20 }, { "epoch": 0.01697792869269949, "grad_norm": 3.10554838180542, "learning_rate": 5.8e-06, "loss": 0.4084, "step": 30 }, { "epoch": 0.022637238256932653, "grad_norm": 4.209934234619141, "learning_rate": 7.8e-06, "loss": 0.289, "step": 40 }, { "epoch": 0.028296547821165818, "grad_norm": 3.8156583309173584, "learning_rate": 9.800000000000001e-06, "loss": 0.2076, "step": 50 }, { "epoch": 0.03395585738539898, "grad_norm": 1.7666521072387695, "learning_rate": 1.18e-05, "loss": 0.1527, "step": 60 }, { "epoch": 0.039615166949632144, "grad_norm": 2.237152099609375, "learning_rate": 1.3800000000000002e-05, "loss": 0.1597, "step": 70 }, { "epoch": 0.045274476513865305, "grad_norm": 2.67939829826355, "learning_rate": 1.58e-05, "loss": 0.1255, "step": 80 }, { "epoch": 0.050933786078098474, "grad_norm": 1.1937289237976074, "learning_rate": 1.78e-05, "loss": 0.119, "step": 90 }, { "epoch": 0.056593095642331635, "grad_norm": 1.8054336309432983, "learning_rate": 1.9800000000000004e-05, "loss": 0.1256, "step": 100 }, { "epoch": 0.0622524052065648, "grad_norm": 1.6763436794281006, "learning_rate": 2.18e-05, "loss": 0.1201, "step": 110 }, { "epoch": 0.06791171477079797, "grad_norm": 2.8526389598846436, "learning_rate": 2.38e-05, "loss": 0.1375, "step": 120 }, { "epoch": 0.07357102433503113, "grad_norm": 3.1760990619659424, "learning_rate": 2.58e-05, "loss": 0.1141, "step": 130 }, { "epoch": 0.07923033389926429, "grad_norm": 2.534616231918335, "learning_rate": 2.7800000000000005e-05, "loss": 0.1632, "step": 140 }, { "epoch": 0.08488964346349745, "grad_norm": 1.8274834156036377, "learning_rate": 2.98e-05, "loss": 0.1228, "step": 150 }, { "epoch": 0.09054895302773061, "grad_norm": 1.4464740753173828, "learning_rate": 3.18e-05, "loss": 0.1255, "step": 160 }, { "epoch": 0.09620826259196379, "grad_norm": 2.235652208328247, "learning_rate": 3.38e-05, "loss": 0.1048, "step": 170 }, { "epoch": 0.10186757215619695, "grad_norm": 2.356511116027832, "learning_rate": 3.58e-05, "loss": 0.1135, "step": 180 }, { "epoch": 0.10752688172043011, "grad_norm": 3.055460214614868, "learning_rate": 3.7800000000000004e-05, "loss": 0.1326, "step": 190 }, { "epoch": 0.11318619128466327, "grad_norm": 1.6213253736495972, "learning_rate": 3.9800000000000005e-05, "loss": 0.1094, "step": 200 }, { "epoch": 0.11884550084889643, "grad_norm": 3.7455012798309326, "learning_rate": 4.18e-05, "loss": 0.1197, "step": 210 }, { "epoch": 0.1245048104131296, "grad_norm": 1.7306926250457764, "learning_rate": 4.38e-05, "loss": 0.1331, "step": 220 }, { "epoch": 0.13016411997736277, "grad_norm": 1.8728036880493164, "learning_rate": 4.58e-05, "loss": 0.102, "step": 230 }, { "epoch": 0.13582342954159593, "grad_norm": 1.3134273290634155, "learning_rate": 4.78e-05, "loss": 0.1013, "step": 240 }, { "epoch": 0.1414827391058291, "grad_norm": 1.5936145782470703, "learning_rate": 4.9800000000000004e-05, "loss": 0.0903, "step": 250 }, { "epoch": 0.14714204867006225, "grad_norm": 1.9866862297058105, "learning_rate": 5.1800000000000005e-05, "loss": 0.0755, "step": 260 }, { "epoch": 0.15280135823429541, "grad_norm": 1.147523045539856, "learning_rate": 5.380000000000001e-05, "loss": 0.0921, "step": 270 }, { "epoch": 0.15846066779852858, "grad_norm": 0.8689554333686829, "learning_rate": 5.580000000000001e-05, "loss": 0.0637, "step": 280 }, { "epoch": 0.16411997736276174, "grad_norm": 2.5970981121063232, "learning_rate": 5.7799999999999995e-05, "loss": 0.1011, "step": 290 }, { "epoch": 0.1697792869269949, "grad_norm": 1.6930752992630005, "learning_rate": 5.9800000000000003e-05, "loss": 0.1222, "step": 300 }, { "epoch": 0.17543859649122806, "grad_norm": 1.0483814477920532, "learning_rate": 6.18e-05, "loss": 0.1165, "step": 310 }, { "epoch": 0.18109790605546122, "grad_norm": 0.8814326524734497, "learning_rate": 6.38e-05, "loss": 0.0794, "step": 320 }, { "epoch": 0.1867572156196944, "grad_norm": 0.8466112017631531, "learning_rate": 6.58e-05, "loss": 0.1113, "step": 330 }, { "epoch": 0.19241652518392757, "grad_norm": 1.5122456550598145, "learning_rate": 6.780000000000001e-05, "loss": 0.1026, "step": 340 }, { "epoch": 0.19807583474816073, "grad_norm": 2.2806904315948486, "learning_rate": 6.98e-05, "loss": 0.1015, "step": 350 }, { "epoch": 0.2037351443123939, "grad_norm": 1.844056487083435, "learning_rate": 7.18e-05, "loss": 0.072, "step": 360 }, { "epoch": 0.20939445387662706, "grad_norm": 1.286731481552124, "learning_rate": 7.38e-05, "loss": 0.0841, "step": 370 }, { "epoch": 0.21505376344086022, "grad_norm": 1.5619851350784302, "learning_rate": 7.58e-05, "loss": 0.0916, "step": 380 }, { "epoch": 0.22071307300509338, "grad_norm": 1.0341761112213135, "learning_rate": 7.780000000000001e-05, "loss": 0.0915, "step": 390 }, { "epoch": 0.22637238256932654, "grad_norm": 1.0657052993774414, "learning_rate": 7.98e-05, "loss": 0.1015, "step": 400 }, { "epoch": 0.2320316921335597, "grad_norm": 1.3276731967926025, "learning_rate": 8.18e-05, "loss": 0.0945, "step": 410 }, { "epoch": 0.23769100169779286, "grad_norm": 1.2397652864456177, "learning_rate": 8.38e-05, "loss": 0.0741, "step": 420 }, { "epoch": 0.24335031126202603, "grad_norm": 1.7711883783340454, "learning_rate": 8.58e-05, "loss": 0.0996, "step": 430 }, { "epoch": 0.2490096208262592, "grad_norm": 1.4832528829574585, "learning_rate": 8.78e-05, "loss": 0.1061, "step": 440 }, { "epoch": 0.2546689303904924, "grad_norm": 1.2473294734954834, "learning_rate": 8.98e-05, "loss": 0.0826, "step": 450 }, { "epoch": 0.26032823995472554, "grad_norm": 1.1587681770324707, "learning_rate": 9.180000000000001e-05, "loss": 0.0941, "step": 460 }, { "epoch": 0.2659875495189587, "grad_norm": 1.194387674331665, "learning_rate": 9.38e-05, "loss": 0.0918, "step": 470 }, { "epoch": 0.27164685908319186, "grad_norm": 1.5423600673675537, "learning_rate": 9.58e-05, "loss": 0.0927, "step": 480 }, { "epoch": 0.277306168647425, "grad_norm": 0.9861327409744263, "learning_rate": 9.78e-05, "loss": 0.1115, "step": 490 }, { "epoch": 0.2829654782116582, "grad_norm": 1.40666663646698, "learning_rate": 9.98e-05, "loss": 0.0966, "step": 500 }, { "epoch": 0.28862478777589134, "grad_norm": 1.3362163305282593, "learning_rate": 9.9999778549206e-05, "loss": 0.1125, "step": 510 }, { "epoch": 0.2942840973401245, "grad_norm": 1.251626968383789, "learning_rate": 9.999901304280685e-05, "loss": 0.0908, "step": 520 }, { "epoch": 0.29994340690435767, "grad_norm": 1.2477635145187378, "learning_rate": 9.999770075521164e-05, "loss": 0.0834, "step": 530 }, { "epoch": 0.30560271646859083, "grad_norm": 0.7544317245483398, "learning_rate": 9.99958417007713e-05, "loss": 0.0707, "step": 540 }, { "epoch": 0.311262026032824, "grad_norm": 1.800075888633728, "learning_rate": 9.999343589981615e-05, "loss": 0.1028, "step": 550 }, { "epoch": 0.31692133559705715, "grad_norm": 1.1727581024169922, "learning_rate": 9.999048337865568e-05, "loss": 0.1114, "step": 560 }, { "epoch": 0.3225806451612903, "grad_norm": 1.3092683553695679, "learning_rate": 9.998698416957815e-05, "loss": 0.0754, "step": 570 }, { "epoch": 0.3282399547255235, "grad_norm": 1.2468684911727905, "learning_rate": 9.998293831085037e-05, "loss": 0.0945, "step": 580 }, { "epoch": 0.33389926428975664, "grad_norm": 1.1072725057601929, "learning_rate": 9.997834584671719e-05, "loss": 0.0827, "step": 590 }, { "epoch": 0.3395585738539898, "grad_norm": 1.7513043880462646, "learning_rate": 9.997320682740107e-05, "loss": 0.1094, "step": 600 }, { "epoch": 0.34521788341822296, "grad_norm": 1.4385285377502441, "learning_rate": 9.996752130910149e-05, "loss": 0.1009, "step": 610 }, { "epoch": 0.3508771929824561, "grad_norm": 0.9453766942024231, "learning_rate": 9.99612893539944e-05, "loss": 0.1067, "step": 620 }, { "epoch": 0.3565365025466893, "grad_norm": 1.1654232740402222, "learning_rate": 9.995451103023144e-05, "loss": 0.0853, "step": 630 }, { "epoch": 0.36219581211092244, "grad_norm": 1.4993220567703247, "learning_rate": 9.994718641193928e-05, "loss": 0.108, "step": 640 }, { "epoch": 0.3678551216751556, "grad_norm": 1.2022749185562134, "learning_rate": 9.993931557921874e-05, "loss": 0.0941, "step": 650 }, { "epoch": 0.3735144312393888, "grad_norm": 1.6157125234603882, "learning_rate": 9.993089861814402e-05, "loss": 0.1124, "step": 660 }, { "epoch": 0.379173740803622, "grad_norm": 0.8949654698371887, "learning_rate": 9.992193562076166e-05, "loss": 0.0776, "step": 670 }, { "epoch": 0.38483305036785515, "grad_norm": 1.2268325090408325, "learning_rate": 9.991242668508954e-05, "loss": 0.0925, "step": 680 }, { "epoch": 0.3904923599320883, "grad_norm": 1.0547950267791748, "learning_rate": 9.990237191511587e-05, "loss": 0.0786, "step": 690 }, { "epoch": 0.39615166949632147, "grad_norm": 1.033515214920044, "learning_rate": 9.989177142079802e-05, "loss": 0.085, "step": 700 }, { "epoch": 0.40181097906055463, "grad_norm": 0.7889791131019592, "learning_rate": 9.988062531806126e-05, "loss": 0.0634, "step": 710 }, { "epoch": 0.4074702886247878, "grad_norm": 1.0996527671813965, "learning_rate": 9.986893372879762e-05, "loss": 0.1017, "step": 720 }, { "epoch": 0.41312959818902095, "grad_norm": 1.2688416242599487, "learning_rate": 9.985669678086443e-05, "loss": 0.0874, "step": 730 }, { "epoch": 0.4187889077532541, "grad_norm": 1.4887137413024902, "learning_rate": 9.984391460808298e-05, "loss": 0.0614, "step": 740 }, { "epoch": 0.4244482173174873, "grad_norm": 1.156822681427002, "learning_rate": 9.983058735023709e-05, "loss": 0.083, "step": 750 }, { "epoch": 0.43010752688172044, "grad_norm": 0.937840461730957, "learning_rate": 9.98167151530715e-05, "loss": 0.1015, "step": 760 }, { "epoch": 0.4357668364459536, "grad_norm": 1.1325842142105103, "learning_rate": 9.980229816829034e-05, "loss": 0.077, "step": 770 }, { "epoch": 0.44142614601018676, "grad_norm": 1.1565141677856445, "learning_rate": 9.978733655355544e-05, "loss": 0.0907, "step": 780 }, { "epoch": 0.4470854555744199, "grad_norm": 0.8821191191673279, "learning_rate": 9.977183047248464e-05, "loss": 0.0761, "step": 790 }, { "epoch": 0.4527447651386531, "grad_norm": 1.1629949808120728, "learning_rate": 9.975578009464992e-05, "loss": 0.0747, "step": 800 }, { "epoch": 0.45840407470288624, "grad_norm": 0.719484806060791, "learning_rate": 9.97391855955757e-05, "loss": 0.0823, "step": 810 }, { "epoch": 0.4640633842671194, "grad_norm": 1.6020772457122803, "learning_rate": 9.972204715673669e-05, "loss": 0.0937, "step": 820 }, { "epoch": 0.46972269383135257, "grad_norm": 0.9074944257736206, "learning_rate": 9.970436496555617e-05, "loss": 0.1057, "step": 830 }, { "epoch": 0.47538200339558573, "grad_norm": 0.8391576409339905, "learning_rate": 9.968613921540373e-05, "loss": 0.0726, "step": 840 }, { "epoch": 0.4810413129598189, "grad_norm": 1.3409184217453003, "learning_rate": 9.966737010559326e-05, "loss": 0.0928, "step": 850 }, { "epoch": 0.48670062252405205, "grad_norm": 0.6904551982879639, "learning_rate": 9.964805784138072e-05, "loss": 0.0677, "step": 860 }, { "epoch": 0.4923599320882852, "grad_norm": 1.3853803873062134, "learning_rate": 9.962820263396195e-05, "loss": 0.0811, "step": 870 }, { "epoch": 0.4980192416525184, "grad_norm": 1.2435743808746338, "learning_rate": 9.960780470047033e-05, "loss": 0.0846, "step": 880 }, { "epoch": 0.5036785512167515, "grad_norm": 0.7919995784759521, "learning_rate": 9.958686426397437e-05, "loss": 0.0672, "step": 890 }, { "epoch": 0.5093378607809848, "grad_norm": 1.1187522411346436, "learning_rate": 9.956538155347534e-05, "loss": 0.0602, "step": 900 }, { "epoch": 0.5149971703452179, "grad_norm": 1.1229113340377808, "learning_rate": 9.95433568039047e-05, "loss": 0.0688, "step": 910 }, { "epoch": 0.5206564799094511, "grad_norm": 1.0347950458526611, "learning_rate": 9.952079025612162e-05, "loss": 0.0591, "step": 920 }, { "epoch": 0.5263157894736842, "grad_norm": 1.2555299997329712, "learning_rate": 9.949768215691022e-05, "loss": 0.0733, "step": 930 }, { "epoch": 0.5319750990379174, "grad_norm": 0.9478985667228699, "learning_rate": 9.9474032758977e-05, "loss": 0.0681, "step": 940 }, { "epoch": 0.5376344086021505, "grad_norm": 0.9187138676643372, "learning_rate": 9.944984232094794e-05, "loss": 0.066, "step": 950 }, { "epoch": 0.5432937181663837, "grad_norm": 0.7259000539779663, "learning_rate": 9.942511110736584e-05, "loss": 0.0796, "step": 960 }, { "epoch": 0.5489530277306168, "grad_norm": 0.6589801907539368, "learning_rate": 9.939983938868726e-05, "loss": 0.0728, "step": 970 }, { "epoch": 0.55461233729485, "grad_norm": 1.0876269340515137, "learning_rate": 9.93740274412797e-05, "loss": 0.0637, "step": 980 }, { "epoch": 0.5602716468590832, "grad_norm": 0.8798189759254456, "learning_rate": 9.934767554741846e-05, "loss": 0.0596, "step": 990 }, { "epoch": 0.5659309564233164, "grad_norm": 0.92113196849823, "learning_rate": 9.932078399528361e-05, "loss": 0.0877, "step": 1000 }, { "epoch": 0.5715902659875495, "grad_norm": 0.9167202711105347, "learning_rate": 9.929335307895689e-05, "loss": 0.0652, "step": 1010 }, { "epoch": 0.5772495755517827, "grad_norm": 1.027732014656067, "learning_rate": 9.926538309841839e-05, "loss": 0.0778, "step": 1020 }, { "epoch": 0.5829088851160158, "grad_norm": 0.6727028489112854, "learning_rate": 9.923687435954334e-05, "loss": 0.0525, "step": 1030 }, { "epoch": 0.588568194680249, "grad_norm": 1.1351747512817383, "learning_rate": 9.920782717409873e-05, "loss": 0.0598, "step": 1040 }, { "epoch": 0.5942275042444821, "grad_norm": 1.283753752708435, "learning_rate": 9.917824185973994e-05, "loss": 0.0719, "step": 1050 }, { "epoch": 0.5998868138087153, "grad_norm": 0.9144543409347534, "learning_rate": 9.914811874000723e-05, "loss": 0.0625, "step": 1060 }, { "epoch": 0.6055461233729486, "grad_norm": 0.7145938277244568, "learning_rate": 9.911745814432218e-05, "loss": 0.0595, "step": 1070 }, { "epoch": 0.6112054329371817, "grad_norm": 0.742757260799408, "learning_rate": 9.90862604079842e-05, "loss": 0.0681, "step": 1080 }, { "epoch": 0.6168647425014149, "grad_norm": 0.6105904579162598, "learning_rate": 9.90545258721667e-05, "loss": 0.0638, "step": 1090 }, { "epoch": 0.622524052065648, "grad_norm": 0.6754553914070129, "learning_rate": 9.90222548839135e-05, "loss": 0.0605, "step": 1100 }, { "epoch": 0.6281833616298812, "grad_norm": 0.9949795603752136, "learning_rate": 9.898944779613495e-05, "loss": 0.101, "step": 1110 }, { "epoch": 0.6338426711941143, "grad_norm": 0.4132203459739685, "learning_rate": 9.89561049676041e-05, "loss": 0.0607, "step": 1120 }, { "epoch": 0.6395019807583475, "grad_norm": 1.0556095838546753, "learning_rate": 9.89222267629528e-05, "loss": 0.0657, "step": 1130 }, { "epoch": 0.6451612903225806, "grad_norm": 0.6254858374595642, "learning_rate": 9.888781355266763e-05, "loss": 0.0659, "step": 1140 }, { "epoch": 0.6508205998868138, "grad_norm": 0.676054060459137, "learning_rate": 9.885286571308598e-05, "loss": 0.0764, "step": 1150 }, { "epoch": 0.656479909451047, "grad_norm": 0.6152122020721436, "learning_rate": 9.881738362639182e-05, "loss": 0.0581, "step": 1160 }, { "epoch": 0.6621392190152802, "grad_norm": 0.9762745499610901, "learning_rate": 9.878136768061154e-05, "loss": 0.0695, "step": 1170 }, { "epoch": 0.6677985285795133, "grad_norm": 0.7954534292221069, "learning_rate": 9.874481826960979e-05, "loss": 0.0556, "step": 1180 }, { "epoch": 0.6734578381437465, "grad_norm": 0.7913191914558411, "learning_rate": 9.870773579308503e-05, "loss": 0.0713, "step": 1190 }, { "epoch": 0.6791171477079796, "grad_norm": 1.2579820156097412, "learning_rate": 9.867012065656533e-05, "loss": 0.0762, "step": 1200 }, { "epoch": 0.6847764572722128, "grad_norm": 1.192491888999939, "learning_rate": 9.863197327140376e-05, "loss": 0.0905, "step": 1210 }, { "epoch": 0.6904357668364459, "grad_norm": 1.1196470260620117, "learning_rate": 9.859329405477403e-05, "loss": 0.0679, "step": 1220 }, { "epoch": 0.6960950764006791, "grad_norm": 0.9146842956542969, "learning_rate": 9.855408342966585e-05, "loss": 0.0698, "step": 1230 }, { "epoch": 0.7017543859649122, "grad_norm": 0.5354324579238892, "learning_rate": 9.851434182488033e-05, "loss": 0.0731, "step": 1240 }, { "epoch": 0.7074136955291455, "grad_norm": 0.7416520714759827, "learning_rate": 9.84740696750253e-05, "loss": 0.0809, "step": 1250 }, { "epoch": 0.7130730050933786, "grad_norm": 0.7727601528167725, "learning_rate": 9.843326742051055e-05, "loss": 0.0729, "step": 1260 }, { "epoch": 0.7187323146576118, "grad_norm": 0.7373871207237244, "learning_rate": 9.839193550754297e-05, "loss": 0.0395, "step": 1270 }, { "epoch": 0.7243916242218449, "grad_norm": 0.9921310544013977, "learning_rate": 9.835007438812177e-05, "loss": 0.062, "step": 1280 }, { "epoch": 0.7300509337860781, "grad_norm": 0.685374915599823, "learning_rate": 9.830768452003341e-05, "loss": 0.0563, "step": 1290 }, { "epoch": 0.7357102433503112, "grad_norm": 0.997697651386261, "learning_rate": 9.826476636684671e-05, "loss": 0.0616, "step": 1300 }, { "epoch": 0.7413695529145444, "grad_norm": 1.0013808012008667, "learning_rate": 9.822132039790773e-05, "loss": 0.0662, "step": 1310 }, { "epoch": 0.7470288624787776, "grad_norm": 0.8838229179382324, "learning_rate": 9.817734708833461e-05, "loss": 0.0642, "step": 1320 }, { "epoch": 0.7526881720430108, "grad_norm": 0.9542595744132996, "learning_rate": 9.813284691901243e-05, "loss": 0.0553, "step": 1330 }, { "epoch": 0.758347481607244, "grad_norm": 0.5350849628448486, "learning_rate": 9.808782037658792e-05, "loss": 0.0431, "step": 1340 }, { "epoch": 0.7640067911714771, "grad_norm": 0.8132240772247314, "learning_rate": 9.804226795346411e-05, "loss": 0.0717, "step": 1350 }, { "epoch": 0.7696661007357103, "grad_norm": 0.9164685010910034, "learning_rate": 9.799619014779503e-05, "loss": 0.0692, "step": 1360 }, { "epoch": 0.7753254102999434, "grad_norm": 0.9166828393936157, "learning_rate": 9.794958746348013e-05, "loss": 0.0525, "step": 1370 }, { "epoch": 0.7809847198641766, "grad_norm": 0.5645913481712341, "learning_rate": 9.790246041015896e-05, "loss": 0.0488, "step": 1380 }, { "epoch": 0.7866440294284097, "grad_norm": 0.8486982583999634, "learning_rate": 9.785480950320538e-05, "loss": 0.0623, "step": 1390 }, { "epoch": 0.7923033389926429, "grad_norm": 0.9856374859809875, "learning_rate": 9.78066352637221e-05, "loss": 0.0427, "step": 1400 }, { "epoch": 0.797962648556876, "grad_norm": 0.9763184189796448, "learning_rate": 9.775793821853488e-05, "loss": 0.0662, "step": 1410 }, { "epoch": 0.8036219581211093, "grad_norm": 0.5652852654457092, "learning_rate": 9.77087189001868e-05, "loss": 0.0624, "step": 1420 }, { "epoch": 0.8092812676853424, "grad_norm": 0.8615242838859558, "learning_rate": 9.765897784693243e-05, "loss": 0.0575, "step": 1430 }, { "epoch": 0.8149405772495756, "grad_norm": 1.01888108253479, "learning_rate": 9.760871560273197e-05, "loss": 0.0723, "step": 1440 }, { "epoch": 0.8205998868138087, "grad_norm": 0.571245551109314, "learning_rate": 9.755793271724526e-05, "loss": 0.0723, "step": 1450 }, { "epoch": 0.8262591963780419, "grad_norm": 0.5587224960327148, "learning_rate": 9.750662974582584e-05, "loss": 0.0567, "step": 1460 }, { "epoch": 0.831918505942275, "grad_norm": 0.983090877532959, "learning_rate": 9.745480724951473e-05, "loss": 0.0614, "step": 1470 }, { "epoch": 0.8375778155065082, "grad_norm": 0.7602547407150269, "learning_rate": 9.740246579503447e-05, "loss": 0.0452, "step": 1480 }, { "epoch": 0.8432371250707413, "grad_norm": 0.7132800817489624, "learning_rate": 9.734960595478284e-05, "loss": 0.0544, "step": 1490 }, { "epoch": 0.8488964346349746, "grad_norm": 0.4686054289340973, "learning_rate": 9.729622830682657e-05, "loss": 0.0565, "step": 1500 }, { "epoch": 0.8545557441992077, "grad_norm": 1.0880781412124634, "learning_rate": 9.724233343489504e-05, "loss": 0.0585, "step": 1510 }, { "epoch": 0.8602150537634409, "grad_norm": 1.143423318862915, "learning_rate": 9.718792192837396e-05, "loss": 0.0611, "step": 1520 }, { "epoch": 0.865874363327674, "grad_norm": 0.7778660655021667, "learning_rate": 9.713299438229886e-05, "loss": 0.0582, "step": 1530 }, { "epoch": 0.8715336728919072, "grad_norm": 0.733134925365448, "learning_rate": 9.707755139734855e-05, "loss": 0.0616, "step": 1540 }, { "epoch": 0.8771929824561403, "grad_norm": 1.2035789489746094, "learning_rate": 9.702159357983866e-05, "loss": 0.0646, "step": 1550 }, { "epoch": 0.8828522920203735, "grad_norm": 1.4871290922164917, "learning_rate": 9.696512154171492e-05, "loss": 0.0747, "step": 1560 }, { "epoch": 0.8885116015846066, "grad_norm": 1.6757930517196655, "learning_rate": 9.690813590054645e-05, "loss": 0.0632, "step": 1570 }, { "epoch": 0.8941709111488398, "grad_norm": 0.7295994758605957, "learning_rate": 9.685063727951914e-05, "loss": 0.0537, "step": 1580 }, { "epoch": 0.8998302207130731, "grad_norm": 0.8169795870780945, "learning_rate": 9.679262630742865e-05, "loss": 0.0697, "step": 1590 }, { "epoch": 0.9054895302773062, "grad_norm": 0.8239006400108337, "learning_rate": 9.673410361867373e-05, "loss": 0.0714, "step": 1600 }, { "epoch": 0.9111488398415394, "grad_norm": 0.7182111144065857, "learning_rate": 9.667506985324909e-05, "loss": 0.0434, "step": 1610 }, { "epoch": 0.9168081494057725, "grad_norm": 0.7491622567176819, "learning_rate": 9.661552565673855e-05, "loss": 0.0551, "step": 1620 }, { "epoch": 0.9224674589700057, "grad_norm": 0.7354273796081543, "learning_rate": 9.655547168030789e-05, "loss": 0.1136, "step": 1630 }, { "epoch": 0.9281267685342388, "grad_norm": 0.7655503153800964, "learning_rate": 9.649490858069777e-05, "loss": 0.0772, "step": 1640 }, { "epoch": 0.933786078098472, "grad_norm": 1.1188385486602783, "learning_rate": 9.643383702021658e-05, "loss": 0.0753, "step": 1650 }, { "epoch": 0.9394453876627051, "grad_norm": 0.9626965522766113, "learning_rate": 9.637225766673307e-05, "loss": 0.0622, "step": 1660 }, { "epoch": 0.9451046972269384, "grad_norm": 0.7576481699943542, "learning_rate": 9.631017119366922e-05, "loss": 0.0546, "step": 1670 }, { "epoch": 0.9507640067911715, "grad_norm": 0.6366353034973145, "learning_rate": 9.624757827999273e-05, "loss": 0.0536, "step": 1680 }, { "epoch": 0.9564233163554047, "grad_norm": 0.6421340703964233, "learning_rate": 9.618447961020971e-05, "loss": 0.0568, "step": 1690 }, { "epoch": 0.9620826259196378, "grad_norm": 0.6514986753463745, "learning_rate": 9.612087587435707e-05, "loss": 0.0519, "step": 1700 }, { "epoch": 0.967741935483871, "grad_norm": 1.2975069284439087, "learning_rate": 9.605676776799508e-05, "loss": 0.065, "step": 1710 }, { "epoch": 0.9734012450481041, "grad_norm": 0.791313648223877, "learning_rate": 9.599215599219973e-05, "loss": 0.058, "step": 1720 }, { "epoch": 0.9790605546123373, "grad_norm": 0.9186238050460815, "learning_rate": 9.592704125355505e-05, "loss": 0.0575, "step": 1730 }, { "epoch": 0.9847198641765704, "grad_norm": 0.6246191263198853, "learning_rate": 9.586142426414538e-05, "loss": 0.0476, "step": 1740 }, { "epoch": 0.9903791737408036, "grad_norm": 0.6215393543243408, "learning_rate": 9.57953057415476e-05, "loss": 0.0735, "step": 1750 }, { "epoch": 0.9960384833050367, "grad_norm": 0.6971435546875, "learning_rate": 9.572868640882328e-05, "loss": 0.0547, "step": 1760 }, { "epoch": 1.0016977928692699, "grad_norm": 0.7553586959838867, "learning_rate": 9.56615669945108e-05, "loss": 0.0457, "step": 1770 }, { "epoch": 1.007357102433503, "grad_norm": 0.7966594099998474, "learning_rate": 9.55939482326173e-05, "loss": 0.0611, "step": 1780 }, { "epoch": 1.0130164119977363, "grad_norm": 0.9479690194129944, "learning_rate": 9.552583086261069e-05, "loss": 0.0422, "step": 1790 }, { "epoch": 1.0186757215619695, "grad_norm": 0.8228363990783691, "learning_rate": 9.545721562941168e-05, "loss": 0.0734, "step": 1800 }, { "epoch": 1.0243350311262025, "grad_norm": 0.9321158528327942, "learning_rate": 9.538810328338543e-05, "loss": 0.0504, "step": 1810 }, { "epoch": 1.0299943406904357, "grad_norm": 0.6621798276901245, "learning_rate": 9.531849458033349e-05, "loss": 0.0732, "step": 1820 }, { "epoch": 1.035653650254669, "grad_norm": 0.7835595011711121, "learning_rate": 9.524839028148547e-05, "loss": 0.0595, "step": 1830 }, { "epoch": 1.0413129598189022, "grad_norm": 0.6544864773750305, "learning_rate": 9.517779115349077e-05, "loss": 0.0849, "step": 1840 }, { "epoch": 1.0469722693831351, "grad_norm": 1.661457896232605, "learning_rate": 9.510669796841014e-05, "loss": 0.0986, "step": 1850 }, { "epoch": 1.0526315789473684, "grad_norm": 0.8697350025177002, "learning_rate": 9.503511150370727e-05, "loss": 0.0744, "step": 1860 }, { "epoch": 1.0582908885116016, "grad_norm": 0.8535569906234741, "learning_rate": 9.496303254224024e-05, "loss": 0.0646, "step": 1870 }, { "epoch": 1.0639501980758348, "grad_norm": 0.5321581959724426, "learning_rate": 9.489046187225306e-05, "loss": 0.0447, "step": 1880 }, { "epoch": 1.069609507640068, "grad_norm": 0.6567709445953369, "learning_rate": 9.481740028736692e-05, "loss": 0.0466, "step": 1890 }, { "epoch": 1.075268817204301, "grad_norm": 0.6248086094856262, "learning_rate": 9.474384858657164e-05, "loss": 0.0517, "step": 1900 }, { "epoch": 1.0809281267685342, "grad_norm": 0.7937766313552856, "learning_rate": 9.466980757421679e-05, "loss": 0.0683, "step": 1910 }, { "epoch": 1.0865874363327674, "grad_norm": 1.025101900100708, "learning_rate": 9.459527806000305e-05, "loss": 0.0595, "step": 1920 }, { "epoch": 1.0922467458970007, "grad_norm": 0.7608640789985657, "learning_rate": 9.452026085897325e-05, "loss": 0.0428, "step": 1930 }, { "epoch": 1.0979060554612337, "grad_norm": 0.49536848068237305, "learning_rate": 9.444475679150348e-05, "loss": 0.0453, "step": 1940 }, { "epoch": 1.1035653650254669, "grad_norm": 0.609743595123291, "learning_rate": 9.436876668329411e-05, "loss": 0.0376, "step": 1950 }, { "epoch": 1.1092246745897, "grad_norm": 0.5043293833732605, "learning_rate": 9.429229136536079e-05, "loss": 0.0579, "step": 1960 }, { "epoch": 1.1148839841539333, "grad_norm": 0.5044483542442322, "learning_rate": 9.421533167402534e-05, "loss": 0.045, "step": 1970 }, { "epoch": 1.1205432937181663, "grad_norm": 0.5287529826164246, "learning_rate": 9.413788845090666e-05, "loss": 0.047, "step": 1980 }, { "epoch": 1.1262026032823995, "grad_norm": 0.5549773573875427, "learning_rate": 9.405996254291136e-05, "loss": 0.0429, "step": 1990 }, { "epoch": 1.1318619128466327, "grad_norm": 0.4656256139278412, "learning_rate": 9.398155480222474e-05, "loss": 0.0604, "step": 2000 }, { "epoch": 1.137521222410866, "grad_norm": 0.6236544847488403, "learning_rate": 9.390266608630128e-05, "loss": 0.0763, "step": 2010 }, { "epoch": 1.143180531975099, "grad_norm": 0.4793378710746765, "learning_rate": 9.38232972578553e-05, "loss": 0.054, "step": 2020 }, { "epoch": 1.1488398415393322, "grad_norm": 0.6340910792350769, "learning_rate": 9.374344918485164e-05, "loss": 0.0424, "step": 2030 }, { "epoch": 1.1544991511035654, "grad_norm": 0.6548799276351929, "learning_rate": 9.366312274049602e-05, "loss": 0.0628, "step": 2040 }, { "epoch": 1.1601584606677986, "grad_norm": 0.646457850933075, "learning_rate": 9.358231880322554e-05, "loss": 0.0623, "step": 2050 }, { "epoch": 1.1658177702320316, "grad_norm": 0.7396799921989441, "learning_rate": 9.350103825669916e-05, "loss": 0.0516, "step": 2060 }, { "epoch": 1.1714770797962648, "grad_norm": 0.5293351411819458, "learning_rate": 9.341928198978787e-05, "loss": 0.0464, "step": 2070 }, { "epoch": 1.177136389360498, "grad_norm": 0.8705312609672546, "learning_rate": 9.333705089656512e-05, "loss": 0.0511, "step": 2080 }, { "epoch": 1.1827956989247312, "grad_norm": 0.7432714104652405, "learning_rate": 9.325434587629698e-05, "loss": 0.0481, "step": 2090 }, { "epoch": 1.1884550084889645, "grad_norm": 0.7466390132904053, "learning_rate": 9.31711678334323e-05, "loss": 0.0553, "step": 2100 }, { "epoch": 1.1941143180531975, "grad_norm": 0.9503251314163208, "learning_rate": 9.308751767759282e-05, "loss": 0.0679, "step": 2110 }, { "epoch": 1.1997736276174307, "grad_norm": 0.770002007484436, "learning_rate": 9.300339632356325e-05, "loss": 0.049, "step": 2120 }, { "epoch": 1.2054329371816639, "grad_norm": 1.1000735759735107, "learning_rate": 9.291880469128124e-05, "loss": 0.0561, "step": 2130 }, { "epoch": 1.2110922467458969, "grad_norm": 0.695549488067627, "learning_rate": 9.283374370582732e-05, "loss": 0.0378, "step": 2140 }, { "epoch": 1.21675155631013, "grad_norm": 1.4372471570968628, "learning_rate": 9.274821429741482e-05, "loss": 0.062, "step": 2150 }, { "epoch": 1.2224108658743633, "grad_norm": 0.7232362031936646, "learning_rate": 9.266221740137961e-05, "loss": 0.0404, "step": 2160 }, { "epoch": 1.2280701754385965, "grad_norm": 0.6811502575874329, "learning_rate": 9.257575395817001e-05, "loss": 0.0485, "step": 2170 }, { "epoch": 1.2337294850028298, "grad_norm": 1.0631557703018188, "learning_rate": 9.248882491333637e-05, "loss": 0.056, "step": 2180 }, { "epoch": 1.2393887945670627, "grad_norm": 0.5628013014793396, "learning_rate": 9.240143121752076e-05, "loss": 0.0463, "step": 2190 }, { "epoch": 1.245048104131296, "grad_norm": 0.42298072576522827, "learning_rate": 9.23135738264467e-05, "loss": 0.0403, "step": 2200 }, { "epoch": 1.2507074136955292, "grad_norm": 0.7429174184799194, "learning_rate": 9.222525370090849e-05, "loss": 0.0484, "step": 2210 }, { "epoch": 1.2563667232597622, "grad_norm": 0.6854789853096008, "learning_rate": 9.213647180676088e-05, "loss": 0.0431, "step": 2220 }, { "epoch": 1.2620260328239954, "grad_norm": 0.680793285369873, "learning_rate": 9.204722911490846e-05, "loss": 0.0527, "step": 2230 }, { "epoch": 1.2676853423882286, "grad_norm": 0.6899171471595764, "learning_rate": 9.1957526601295e-05, "loss": 0.0448, "step": 2240 }, { "epoch": 1.2733446519524618, "grad_norm": 0.6361444592475891, "learning_rate": 9.186736524689281e-05, "loss": 0.0461, "step": 2250 }, { "epoch": 1.279003961516695, "grad_norm": 0.7933545112609863, "learning_rate": 9.177674603769204e-05, "loss": 0.0461, "step": 2260 }, { "epoch": 1.284663271080928, "grad_norm": 0.8648480772972107, "learning_rate": 9.168566996468983e-05, "loss": 0.043, "step": 2270 }, { "epoch": 1.2903225806451613, "grad_norm": 0.685413658618927, "learning_rate": 9.159413802387951e-05, "loss": 0.0542, "step": 2280 }, { "epoch": 1.2959818902093945, "grad_norm": 0.8425791263580322, "learning_rate": 9.150215121623974e-05, "loss": 0.0691, "step": 2290 }, { "epoch": 1.3016411997736277, "grad_norm": 0.5436529517173767, "learning_rate": 9.140971054772349e-05, "loss": 0.05, "step": 2300 }, { "epoch": 1.307300509337861, "grad_norm": 0.9092062711715698, "learning_rate": 9.131681702924713e-05, "loss": 0.0503, "step": 2310 }, { "epoch": 1.312959818902094, "grad_norm": 0.602950394153595, "learning_rate": 9.122347167667926e-05, "loss": 0.042, "step": 2320 }, { "epoch": 1.3186191284663271, "grad_norm": 0.7941600680351257, "learning_rate": 9.112967551082973e-05, "loss": 0.0499, "step": 2330 }, { "epoch": 1.3242784380305603, "grad_norm": 0.786613941192627, "learning_rate": 9.103542955743835e-05, "loss": 0.0487, "step": 2340 }, { "epoch": 1.3299377475947933, "grad_norm": 0.6765789985656738, "learning_rate": 9.094073484716381e-05, "loss": 0.0533, "step": 2350 }, { "epoch": 1.3355970571590265, "grad_norm": 0.7006556987762451, "learning_rate": 9.084559241557226e-05, "loss": 0.0528, "step": 2360 }, { "epoch": 1.3412563667232598, "grad_norm": 0.7602994441986084, "learning_rate": 9.075000330312608e-05, "loss": 0.0528, "step": 2370 }, { "epoch": 1.346915676287493, "grad_norm": 1.3614463806152344, "learning_rate": 9.065396855517253e-05, "loss": 0.0578, "step": 2380 }, { "epoch": 1.3525749858517262, "grad_norm": 0.7470871210098267, "learning_rate": 9.055748922193219e-05, "loss": 0.0562, "step": 2390 }, { "epoch": 1.3582342954159592, "grad_norm": 0.7833662629127502, "learning_rate": 9.046056635848761e-05, "loss": 0.0514, "step": 2400 }, { "epoch": 1.3638936049801924, "grad_norm": 0.6521514058113098, "learning_rate": 9.036320102477169e-05, "loss": 0.0432, "step": 2410 }, { "epoch": 1.3695529145444256, "grad_norm": 0.6200571060180664, "learning_rate": 9.02653942855561e-05, "loss": 0.0732, "step": 2420 }, { "epoch": 1.3752122241086586, "grad_norm": 0.7283011078834534, "learning_rate": 9.016714721043971e-05, "loss": 0.0452, "step": 2430 }, { "epoch": 1.3808715336728918, "grad_norm": 0.7621099948883057, "learning_rate": 9.006846087383675e-05, "loss": 0.0515, "step": 2440 }, { "epoch": 1.386530843237125, "grad_norm": 0.7514267563819885, "learning_rate": 8.996933635496523e-05, "loss": 0.0394, "step": 2450 }, { "epoch": 1.3921901528013583, "grad_norm": 0.7734144926071167, "learning_rate": 8.986977473783498e-05, "loss": 0.0486, "step": 2460 }, { "epoch": 1.3978494623655915, "grad_norm": 0.6210748553276062, "learning_rate": 8.97697771112359e-05, "loss": 0.039, "step": 2470 }, { "epoch": 1.4035087719298245, "grad_norm": 0.5797284245491028, "learning_rate": 8.966934456872602e-05, "loss": 0.0456, "step": 2480 }, { "epoch": 1.4091680814940577, "grad_norm": 0.626136064529419, "learning_rate": 8.95684782086195e-05, "loss": 0.0541, "step": 2490 }, { "epoch": 1.414827391058291, "grad_norm": 0.2991216778755188, "learning_rate": 8.946717913397476e-05, "loss": 0.036, "step": 2500 }, { "epoch": 1.4204867006225241, "grad_norm": 0.5612489581108093, "learning_rate": 8.93654484525822e-05, "loss": 0.0338, "step": 2510 }, { "epoch": 1.4261460101867574, "grad_norm": 0.4333272874355316, "learning_rate": 8.926328727695226e-05, "loss": 0.0444, "step": 2520 }, { "epoch": 1.4318053197509903, "grad_norm": 0.5372508764266968, "learning_rate": 8.916069672430319e-05, "loss": 0.0377, "step": 2530 }, { "epoch": 1.4374646293152236, "grad_norm": 0.5010121464729309, "learning_rate": 8.905767791654884e-05, "loss": 0.0353, "step": 2540 }, { "epoch": 1.4431239388794568, "grad_norm": 0.5633225440979004, "learning_rate": 8.895423198028638e-05, "loss": 0.0481, "step": 2550 }, { "epoch": 1.4487832484436898, "grad_norm": 0.6071559190750122, "learning_rate": 8.885036004678402e-05, "loss": 0.049, "step": 2560 }, { "epoch": 1.454442558007923, "grad_norm": 0.8190767765045166, "learning_rate": 8.874606325196857e-05, "loss": 0.0511, "step": 2570 }, { "epoch": 1.4601018675721562, "grad_norm": 0.6717620491981506, "learning_rate": 8.864134273641304e-05, "loss": 0.0446, "step": 2580 }, { "epoch": 1.4657611771363894, "grad_norm": 0.9050039052963257, "learning_rate": 8.853619964532427e-05, "loss": 0.0454, "step": 2590 }, { "epoch": 1.4714204867006226, "grad_norm": 0.6766312122344971, "learning_rate": 8.843063512853019e-05, "loss": 0.0462, "step": 2600 }, { "epoch": 1.4770797962648556, "grad_norm": 0.4999978244304657, "learning_rate": 8.832465034046749e-05, "loss": 0.0533, "step": 2610 }, { "epoch": 1.4827391058290889, "grad_norm": 0.4319976568222046, "learning_rate": 8.821824644016882e-05, "loss": 0.0442, "step": 2620 }, { "epoch": 1.488398415393322, "grad_norm": 0.7079813480377197, "learning_rate": 8.811142459125019e-05, "loss": 0.0556, "step": 2630 }, { "epoch": 1.494057724957555, "grad_norm": 0.9638972282409668, "learning_rate": 8.800418596189822e-05, "loss": 0.0527, "step": 2640 }, { "epoch": 1.4997170345217883, "grad_norm": 0.7943388223648071, "learning_rate": 8.789653172485737e-05, "loss": 0.0585, "step": 2650 }, { "epoch": 1.5053763440860215, "grad_norm": 0.629218339920044, "learning_rate": 8.778846305741715e-05, "loss": 0.0497, "step": 2660 }, { "epoch": 1.5110356536502547, "grad_norm": 0.5880003571510315, "learning_rate": 8.767998114139918e-05, "loss": 0.0573, "step": 2670 }, { "epoch": 1.516694963214488, "grad_norm": 0.6889175176620483, "learning_rate": 8.757108716314429e-05, "loss": 0.0681, "step": 2680 }, { "epoch": 1.522354272778721, "grad_norm": 0.3247281312942505, "learning_rate": 8.746178231349962e-05, "loss": 0.0431, "step": 2690 }, { "epoch": 1.5280135823429541, "grad_norm": 0.6539881229400635, "learning_rate": 8.735206778780549e-05, "loss": 0.0343, "step": 2700 }, { "epoch": 1.5336728919071874, "grad_norm": 0.6488775610923767, "learning_rate": 8.724194478588234e-05, "loss": 0.047, "step": 2710 }, { "epoch": 1.5393322014714204, "grad_norm": 0.8400557041168213, "learning_rate": 8.713141451201772e-05, "loss": 0.0754, "step": 2720 }, { "epoch": 1.5449915110356538, "grad_norm": 0.5844991207122803, "learning_rate": 8.702047817495295e-05, "loss": 0.0409, "step": 2730 }, { "epoch": 1.5506508205998868, "grad_norm": 1.1271346807479858, "learning_rate": 8.69091369878701e-05, "loss": 0.0522, "step": 2740 }, { "epoch": 1.55631013016412, "grad_norm": 0.8452271223068237, "learning_rate": 8.679739216837849e-05, "loss": 0.0551, "step": 2750 }, { "epoch": 1.5619694397283532, "grad_norm": 0.6306042075157166, "learning_rate": 8.66852449385016e-05, "loss": 0.0467, "step": 2760 }, { "epoch": 1.5676287492925862, "grad_norm": 0.616023600101471, "learning_rate": 8.657269652466356e-05, "loss": 0.0477, "step": 2770 }, { "epoch": 1.5732880588568194, "grad_norm": 0.9345172047615051, "learning_rate": 8.645974815767577e-05, "loss": 0.0502, "step": 2780 }, { "epoch": 1.5789473684210527, "grad_norm": 0.723246157169342, "learning_rate": 8.634640107272351e-05, "loss": 0.0545, "step": 2790 }, { "epoch": 1.5846066779852856, "grad_norm": 0.49784278869628906, "learning_rate": 8.623265650935234e-05, "loss": 0.0518, "step": 2800 }, { "epoch": 1.590265987549519, "grad_norm": 0.5574387907981873, "learning_rate": 8.611851571145456e-05, "loss": 0.0627, "step": 2810 }, { "epoch": 1.595925297113752, "grad_norm": 0.6356534957885742, "learning_rate": 8.600397992725566e-05, "loss": 0.0616, "step": 2820 }, { "epoch": 1.6015846066779853, "grad_norm": 0.6085881590843201, "learning_rate": 8.588905040930061e-05, "loss": 0.0573, "step": 2830 }, { "epoch": 1.6072439162422185, "grad_norm": 0.49020498991012573, "learning_rate": 8.577372841444022e-05, "loss": 0.0544, "step": 2840 }, { "epoch": 1.6129032258064515, "grad_norm": 0.8643253445625305, "learning_rate": 8.565801520381736e-05, "loss": 0.0648, "step": 2850 }, { "epoch": 1.6185625353706847, "grad_norm": 0.626186728477478, "learning_rate": 8.554191204285313e-05, "loss": 0.0479, "step": 2860 }, { "epoch": 1.624221844934918, "grad_norm": 0.4494522511959076, "learning_rate": 8.542542020123315e-05, "loss": 0.042, "step": 2870 }, { "epoch": 1.629881154499151, "grad_norm": 0.766018807888031, "learning_rate": 8.530854095289347e-05, "loss": 0.0475, "step": 2880 }, { "epoch": 1.6355404640633844, "grad_norm": 0.5061770081520081, "learning_rate": 8.519127557600688e-05, "loss": 0.052, "step": 2890 }, { "epoch": 1.6411997736276174, "grad_norm": 0.7420628666877747, "learning_rate": 8.507362535296871e-05, "loss": 0.0575, "step": 2900 }, { "epoch": 1.6468590831918506, "grad_norm": 0.5609028339385986, "learning_rate": 8.495559157038299e-05, "loss": 0.045, "step": 2910 }, { "epoch": 1.6525183927560838, "grad_norm": 0.3088698387145996, "learning_rate": 8.483717551904823e-05, "loss": 0.0527, "step": 2920 }, { "epoch": 1.6581777023203168, "grad_norm": 0.43243131041526794, "learning_rate": 8.47183784939434e-05, "loss": 0.0353, "step": 2930 }, { "epoch": 1.6638370118845502, "grad_norm": 0.8786044716835022, "learning_rate": 8.459920179421374e-05, "loss": 0.0464, "step": 2940 }, { "epoch": 1.6694963214487832, "grad_norm": 0.441781610250473, "learning_rate": 8.447964672315656e-05, "loss": 0.0365, "step": 2950 }, { "epoch": 1.6751556310130165, "grad_norm": 0.5830533504486084, "learning_rate": 8.435971458820692e-05, "loss": 0.039, "step": 2960 }, { "epoch": 1.6808149405772497, "grad_norm": 0.4785899817943573, "learning_rate": 8.423940670092345e-05, "loss": 0.0406, "step": 2970 }, { "epoch": 1.6864742501414827, "grad_norm": 0.9931580424308777, "learning_rate": 8.411872437697394e-05, "loss": 0.0501, "step": 2980 }, { "epoch": 1.6921335597057159, "grad_norm": 0.41342973709106445, "learning_rate": 8.399766893612096e-05, "loss": 0.041, "step": 2990 }, { "epoch": 1.697792869269949, "grad_norm": 0.47967806458473206, "learning_rate": 8.38762417022074e-05, "loss": 0.0538, "step": 3000 }, { "epoch": 1.703452178834182, "grad_norm": 0.76365065574646, "learning_rate": 8.375444400314204e-05, "loss": 0.0553, "step": 3010 }, { "epoch": 1.7091114883984155, "grad_norm": 0.6275987029075623, "learning_rate": 8.3632277170885e-05, "loss": 0.0313, "step": 3020 }, { "epoch": 1.7147707979626485, "grad_norm": 0.8967359662055969, "learning_rate": 8.350974254143318e-05, "loss": 0.0444, "step": 3030 }, { "epoch": 1.7204301075268817, "grad_norm": 0.43494531512260437, "learning_rate": 8.338684145480566e-05, "loss": 0.0392, "step": 3040 }, { "epoch": 1.726089417091115, "grad_norm": 0.6516427397727966, "learning_rate": 8.326357525502904e-05, "loss": 0.0452, "step": 3050 }, { "epoch": 1.731748726655348, "grad_norm": 0.5176613926887512, "learning_rate": 8.313994529012273e-05, "loss": 0.0395, "step": 3060 }, { "epoch": 1.7374080362195812, "grad_norm": 0.5854260325431824, "learning_rate": 8.301595291208422e-05, "loss": 0.0338, "step": 3070 }, { "epoch": 1.7430673457838144, "grad_norm": 0.9170899987220764, "learning_rate": 8.289159947687427e-05, "loss": 0.0522, "step": 3080 }, { "epoch": 1.7487266553480474, "grad_norm": 0.5007301568984985, "learning_rate": 8.276688634440216e-05, "loss": 0.0458, "step": 3090 }, { "epoch": 1.7543859649122808, "grad_norm": 0.5097112059593201, "learning_rate": 8.26418148785107e-05, "loss": 0.0467, "step": 3100 }, { "epoch": 1.7600452744765138, "grad_norm": 0.5052613019943237, "learning_rate": 8.251638644696141e-05, "loss": 0.0349, "step": 3110 }, { "epoch": 1.765704584040747, "grad_norm": 0.35661929845809937, "learning_rate": 8.23906024214195e-05, "loss": 0.0303, "step": 3120 }, { "epoch": 1.7713638936049803, "grad_norm": 0.23859961330890656, "learning_rate": 8.226446417743897e-05, "loss": 0.0343, "step": 3130 }, { "epoch": 1.7770232031692133, "grad_norm": 0.5440965294837952, "learning_rate": 8.213797309444742e-05, "loss": 0.0464, "step": 3140 }, { "epoch": 1.7826825127334465, "grad_norm": 0.7961117625236511, "learning_rate": 8.201113055573105e-05, "loss": 0.0406, "step": 3150 }, { "epoch": 1.7883418222976797, "grad_norm": 0.508758544921875, "learning_rate": 8.188393794841958e-05, "loss": 0.0442, "step": 3160 }, { "epoch": 1.794001131861913, "grad_norm": 0.5264890193939209, "learning_rate": 8.175639666347094e-05, "loss": 0.0336, "step": 3170 }, { "epoch": 1.7996604414261461, "grad_norm": 0.7343180179595947, "learning_rate": 8.162850809565623e-05, "loss": 0.0552, "step": 3180 }, { "epoch": 1.8053197509903791, "grad_norm": 0.6147553324699402, "learning_rate": 8.150027364354431e-05, "loss": 0.0489, "step": 3190 }, { "epoch": 1.8109790605546123, "grad_norm": 0.6342501640319824, "learning_rate": 8.137169470948662e-05, "loss": 0.0514, "step": 3200 }, { "epoch": 1.8166383701188455, "grad_norm": 0.6614866256713867, "learning_rate": 8.124277269960179e-05, "loss": 0.0385, "step": 3210 }, { "epoch": 1.8222976796830785, "grad_norm": 0.5475081205368042, "learning_rate": 8.111350902376023e-05, "loss": 0.0471, "step": 3220 }, { "epoch": 1.827956989247312, "grad_norm": 0.37897610664367676, "learning_rate": 8.098390509556883e-05, "loss": 0.0465, "step": 3230 }, { "epoch": 1.833616298811545, "grad_norm": 0.5038880705833435, "learning_rate": 8.085396233235536e-05, "loss": 0.0451, "step": 3240 }, { "epoch": 1.8392756083757782, "grad_norm": 0.6834940910339355, "learning_rate": 8.072368215515306e-05, "loss": 0.0601, "step": 3250 }, { "epoch": 1.8449349179400114, "grad_norm": 0.7319331169128418, "learning_rate": 8.059306598868506e-05, "loss": 0.033, "step": 3260 }, { "epoch": 1.8505942275042444, "grad_norm": 0.8380953669548035, "learning_rate": 8.046211526134888e-05, "loss": 0.0464, "step": 3270 }, { "epoch": 1.8562535370684776, "grad_norm": 0.4649491310119629, "learning_rate": 8.033083140520065e-05, "loss": 0.0506, "step": 3280 }, { "epoch": 1.8619128466327108, "grad_norm": 0.4398675858974457, "learning_rate": 8.019921585593962e-05, "loss": 0.0372, "step": 3290 }, { "epoch": 1.8675721561969438, "grad_norm": 0.6413957476615906, "learning_rate": 8.006727005289232e-05, "loss": 0.0414, "step": 3300 }, { "epoch": 1.8732314657611773, "grad_norm": 0.4823025166988373, "learning_rate": 7.993499543899692e-05, "loss": 0.0424, "step": 3310 }, { "epoch": 1.8788907753254103, "grad_norm": 0.47241145372390747, "learning_rate": 7.980239346078742e-05, "loss": 0.0644, "step": 3320 }, { "epoch": 1.8845500848896435, "grad_norm": 1.091975212097168, "learning_rate": 7.966946556837778e-05, "loss": 0.0576, "step": 3330 }, { "epoch": 1.8902093944538767, "grad_norm": 0.5624361038208008, "learning_rate": 7.953621321544616e-05, "loss": 0.0348, "step": 3340 }, { "epoch": 1.8958687040181097, "grad_norm": 0.5308264493942261, "learning_rate": 7.940263785921896e-05, "loss": 0.0547, "step": 3350 }, { "epoch": 1.901528013582343, "grad_norm": 0.5924705862998962, "learning_rate": 7.926874096045482e-05, "loss": 0.0311, "step": 3360 }, { "epoch": 1.9071873231465761, "grad_norm": 0.671329140663147, "learning_rate": 7.913452398342881e-05, "loss": 0.0432, "step": 3370 }, { "epoch": 1.9128466327108091, "grad_norm": 0.42409855127334595, "learning_rate": 7.89999883959163e-05, "loss": 0.0627, "step": 3380 }, { "epoch": 1.9185059422750426, "grad_norm": 0.8025627732276917, "learning_rate": 7.886513566917687e-05, "loss": 0.0433, "step": 3390 }, { "epoch": 1.9241652518392756, "grad_norm": 0.8151332139968872, "learning_rate": 7.872996727793838e-05, "loss": 0.0464, "step": 3400 }, { "epoch": 1.9298245614035088, "grad_norm": 0.4845048487186432, "learning_rate": 7.859448470038069e-05, "loss": 0.0454, "step": 3410 }, { "epoch": 1.935483870967742, "grad_norm": 0.37321311235427856, "learning_rate": 7.845868941811956e-05, "loss": 0.034, "step": 3420 }, { "epoch": 1.941143180531975, "grad_norm": 0.7626938223838806, "learning_rate": 7.832258291619043e-05, "loss": 0.038, "step": 3430 }, { "epoch": 1.9468024900962084, "grad_norm": 0.7977511882781982, "learning_rate": 7.81861666830322e-05, "loss": 0.0457, "step": 3440 }, { "epoch": 1.9524617996604414, "grad_norm": 0.807888388633728, "learning_rate": 7.804944221047097e-05, "loss": 0.0386, "step": 3450 }, { "epoch": 1.9581211092246746, "grad_norm": 0.6336291432380676, "learning_rate": 7.791241099370364e-05, "loss": 0.0399, "step": 3460 }, { "epoch": 1.9637804187889079, "grad_norm": 0.8386486768722534, "learning_rate": 7.777507453128163e-05, "loss": 0.0588, "step": 3470 }, { "epoch": 1.9694397283531409, "grad_norm": 0.736029326915741, "learning_rate": 7.763743432509451e-05, "loss": 0.0427, "step": 3480 }, { "epoch": 1.975099037917374, "grad_norm": 0.6496866941452026, "learning_rate": 7.749949188035353e-05, "loss": 0.0469, "step": 3490 }, { "epoch": 1.9807583474816073, "grad_norm": 0.4239360988140106, "learning_rate": 7.736124870557516e-05, "loss": 0.0332, "step": 3500 }, { "epoch": 1.9864176570458403, "grad_norm": 0.7490051984786987, "learning_rate": 7.722270631256459e-05, "loss": 0.0424, "step": 3510 }, { "epoch": 1.9920769666100737, "grad_norm": 0.8489423394203186, "learning_rate": 7.708386621639925e-05, "loss": 0.0427, "step": 3520 }, { "epoch": 1.9977362761743067, "grad_norm": 0.4871460795402527, "learning_rate": 7.694472993541219e-05, "loss": 0.0413, "step": 3530 }, { "epoch": 2.0033955857385397, "grad_norm": 0.5171214938163757, "learning_rate": 7.680529899117547e-05, "loss": 0.0497, "step": 3540 }, { "epoch": 2.009054895302773, "grad_norm": 0.49935320019721985, "learning_rate": 7.666557490848358e-05, "loss": 0.0428, "step": 3550 }, { "epoch": 2.014714204867006, "grad_norm": 0.6899267435073853, "learning_rate": 7.65255592153367e-05, "loss": 0.0393, "step": 3560 }, { "epoch": 2.0203735144312396, "grad_norm": 0.6978103518486023, "learning_rate": 7.638525344292402e-05, "loss": 0.039, "step": 3570 }, { "epoch": 2.0260328239954726, "grad_norm": 0.4505501091480255, "learning_rate": 7.624465912560697e-05, "loss": 0.0379, "step": 3580 }, { "epoch": 2.0316921335597056, "grad_norm": 0.6588614583015442, "learning_rate": 7.610377780090249e-05, "loss": 0.0468, "step": 3590 }, { "epoch": 2.037351443123939, "grad_norm": 0.47646525502204895, "learning_rate": 7.596261100946618e-05, "loss": 0.0314, "step": 3600 }, { "epoch": 2.043010752688172, "grad_norm": 0.5849595665931702, "learning_rate": 7.582116029507542e-05, "loss": 0.0423, "step": 3610 }, { "epoch": 2.048670062252405, "grad_norm": 0.7265652418136597, "learning_rate": 7.56794272046126e-05, "loss": 0.0428, "step": 3620 }, { "epoch": 2.0543293718166384, "grad_norm": 0.6475759148597717, "learning_rate": 7.55374132880481e-05, "loss": 0.035, "step": 3630 }, { "epoch": 2.0599886813808714, "grad_norm": 1.025071382522583, "learning_rate": 7.539512009842333e-05, "loss": 0.0364, "step": 3640 }, { "epoch": 2.065647990945105, "grad_norm": 0.7076042294502258, "learning_rate": 7.525254919183382e-05, "loss": 0.0292, "step": 3650 }, { "epoch": 2.071307300509338, "grad_norm": 0.5352051258087158, "learning_rate": 7.510970212741215e-05, "loss": 0.0381, "step": 3660 }, { "epoch": 2.076966610073571, "grad_norm": 0.6327780485153198, "learning_rate": 7.496658046731096e-05, "loss": 0.0351, "step": 3670 }, { "epoch": 2.0826259196378043, "grad_norm": 0.4013051986694336, "learning_rate": 7.482318577668578e-05, "loss": 0.0329, "step": 3680 }, { "epoch": 2.0882852292020373, "grad_norm": 0.3340098261833191, "learning_rate": 7.467951962367796e-05, "loss": 0.0374, "step": 3690 }, { "epoch": 2.0939445387662703, "grad_norm": 0.5839859843254089, "learning_rate": 7.453558357939755e-05, "loss": 0.0418, "step": 3700 }, { "epoch": 2.0996038483305037, "grad_norm": 1.155350923538208, "learning_rate": 7.439137921790606e-05, "loss": 0.0321, "step": 3710 }, { "epoch": 2.1052631578947367, "grad_norm": 0.4015667140483856, "learning_rate": 7.42469081161993e-05, "loss": 0.0283, "step": 3720 }, { "epoch": 2.11092246745897, "grad_norm": 0.424580454826355, "learning_rate": 7.410217185419006e-05, "loss": 0.0344, "step": 3730 }, { "epoch": 2.116581777023203, "grad_norm": 0.37337833642959595, "learning_rate": 7.395717201469095e-05, "loss": 0.0271, "step": 3740 }, { "epoch": 2.122241086587436, "grad_norm": 0.41936591267585754, "learning_rate": 7.381191018339696e-05, "loss": 0.0354, "step": 3750 }, { "epoch": 2.1279003961516696, "grad_norm": 0.621281087398529, "learning_rate": 7.36663879488682e-05, "loss": 0.0377, "step": 3760 }, { "epoch": 2.1335597057159026, "grad_norm": 0.5020797848701477, "learning_rate": 7.352060690251254e-05, "loss": 0.0533, "step": 3770 }, { "epoch": 2.139219015280136, "grad_norm": 0.8350841403007507, "learning_rate": 7.337456863856811e-05, "loss": 0.0557, "step": 3780 }, { "epoch": 2.144878324844369, "grad_norm": 0.8346255421638489, "learning_rate": 7.3228274754086e-05, "loss": 0.047, "step": 3790 }, { "epoch": 2.150537634408602, "grad_norm": 0.881129264831543, "learning_rate": 7.308172684891267e-05, "loss": 0.0726, "step": 3800 }, { "epoch": 2.1561969439728355, "grad_norm": 0.6044751405715942, "learning_rate": 7.293492652567255e-05, "loss": 0.0345, "step": 3810 }, { "epoch": 2.1618562535370685, "grad_norm": 0.5880212187767029, "learning_rate": 7.278787538975043e-05, "loss": 0.0498, "step": 3820 }, { "epoch": 2.1675155631013014, "grad_norm": 0.8480750322341919, "learning_rate": 7.2640575049274e-05, "loss": 0.049, "step": 3830 }, { "epoch": 2.173174872665535, "grad_norm": 0.6445808410644531, "learning_rate": 7.249302711509616e-05, "loss": 0.0352, "step": 3840 }, { "epoch": 2.178834182229768, "grad_norm": 0.5758914351463318, "learning_rate": 7.23452332007775e-05, "loss": 0.0426, "step": 3850 }, { "epoch": 2.1844934917940013, "grad_norm": 0.6860023736953735, "learning_rate": 7.219719492256858e-05, "loss": 0.04, "step": 3860 }, { "epoch": 2.1901528013582343, "grad_norm": 0.46748650074005127, "learning_rate": 7.20489138993923e-05, "loss": 0.0379, "step": 3870 }, { "epoch": 2.1958121109224673, "grad_norm": 0.7767139077186584, "learning_rate": 7.190039175282614e-05, "loss": 0.0309, "step": 3880 }, { "epoch": 2.2014714204867007, "grad_norm": 0.5371547937393188, "learning_rate": 7.175163010708455e-05, "loss": 0.0415, "step": 3890 }, { "epoch": 2.2071307300509337, "grad_norm": 0.7050542831420898, "learning_rate": 7.1602630589001e-05, "loss": 0.0384, "step": 3900 }, { "epoch": 2.212790039615167, "grad_norm": 0.966188907623291, "learning_rate": 7.14533948280104e-05, "loss": 0.0434, "step": 3910 }, { "epoch": 2.2184493491794, "grad_norm": 0.794049859046936, "learning_rate": 7.130392445613109e-05, "loss": 0.039, "step": 3920 }, { "epoch": 2.224108658743633, "grad_norm": 0.5228994488716125, "learning_rate": 7.115422110794711e-05, "loss": 0.0487, "step": 3930 }, { "epoch": 2.2297679683078666, "grad_norm": 0.6201066374778748, "learning_rate": 7.100428642059033e-05, "loss": 0.0588, "step": 3940 }, { "epoch": 2.2354272778720996, "grad_norm": 0.4783194661140442, "learning_rate": 7.08541220337224e-05, "loss": 0.0476, "step": 3950 }, { "epoch": 2.2410865874363326, "grad_norm": 0.5686235427856445, "learning_rate": 7.070372958951706e-05, "loss": 0.0491, "step": 3960 }, { "epoch": 2.246745897000566, "grad_norm": 0.38256895542144775, "learning_rate": 7.055311073264194e-05, "loss": 0.0291, "step": 3970 }, { "epoch": 2.252405206564799, "grad_norm": 0.5158995985984802, "learning_rate": 7.040226711024077e-05, "loss": 0.0356, "step": 3980 }, { "epoch": 2.258064516129032, "grad_norm": 0.5597074627876282, "learning_rate": 7.02512003719152e-05, "loss": 0.0312, "step": 3990 }, { "epoch": 2.2637238256932655, "grad_norm": 0.5796484351158142, "learning_rate": 7.00999121697069e-05, "loss": 0.0403, "step": 4000 }, { "epoch": 2.2693831352574985, "grad_norm": 0.43213656544685364, "learning_rate": 6.99484041580794e-05, "loss": 0.0372, "step": 4010 }, { "epoch": 2.275042444821732, "grad_norm": 0.8799070715904236, "learning_rate": 6.979667799390004e-05, "loss": 0.0287, "step": 4020 }, { "epoch": 2.280701754385965, "grad_norm": 0.5011523962020874, "learning_rate": 6.964473533642185e-05, "loss": 0.0357, "step": 4030 }, { "epoch": 2.286361063950198, "grad_norm": 0.4229520261287689, "learning_rate": 6.949257784726539e-05, "loss": 0.0246, "step": 4040 }, { "epoch": 2.2920203735144313, "grad_norm": 0.8594684600830078, "learning_rate": 6.934020719040056e-05, "loss": 0.0415, "step": 4050 }, { "epoch": 2.2976796830786643, "grad_norm": 0.5908093452453613, "learning_rate": 6.918762503212848e-05, "loss": 0.0328, "step": 4060 }, { "epoch": 2.3033389926428978, "grad_norm": 0.239020437002182, "learning_rate": 6.903483304106319e-05, "loss": 0.0266, "step": 4070 }, { "epoch": 2.3089983022071308, "grad_norm": 0.7022054195404053, "learning_rate": 6.888183288811341e-05, "loss": 0.0464, "step": 4080 }, { "epoch": 2.3146576117713638, "grad_norm": 0.8178216218948364, "learning_rate": 6.87286262464643e-05, "loss": 0.0446, "step": 4090 }, { "epoch": 2.320316921335597, "grad_norm": 0.5563772320747375, "learning_rate": 6.857521479155915e-05, "loss": 0.0273, "step": 4100 }, { "epoch": 2.32597623089983, "grad_norm": 1.1647361516952515, "learning_rate": 6.842160020108104e-05, "loss": 0.0343, "step": 4110 }, { "epoch": 2.331635540464063, "grad_norm": 0.5406816005706787, "learning_rate": 6.826778415493455e-05, "loss": 0.0335, "step": 4120 }, { "epoch": 2.3372948500282966, "grad_norm": 0.5563187003135681, "learning_rate": 6.811376833522729e-05, "loss": 0.0402, "step": 4130 }, { "epoch": 2.3429541595925296, "grad_norm": 0.5906765460968018, "learning_rate": 6.795955442625159e-05, "loss": 0.0461, "step": 4140 }, { "epoch": 2.348613469156763, "grad_norm": 0.4335680603981018, "learning_rate": 6.780514411446608e-05, "loss": 0.0369, "step": 4150 }, { "epoch": 2.354272778720996, "grad_norm": 0.38988831639289856, "learning_rate": 6.765053908847716e-05, "loss": 0.0363, "step": 4160 }, { "epoch": 2.359932088285229, "grad_norm": 0.39675620198249817, "learning_rate": 6.749574103902064e-05, "loss": 0.0384, "step": 4170 }, { "epoch": 2.3655913978494625, "grad_norm": 0.4085799753665924, "learning_rate": 6.734075165894317e-05, "loss": 0.0324, "step": 4180 }, { "epoch": 2.3712507074136955, "grad_norm": 0.4118804931640625, "learning_rate": 6.71855726431838e-05, "loss": 0.0302, "step": 4190 }, { "epoch": 2.376910016977929, "grad_norm": 0.4113968312740326, "learning_rate": 6.703020568875538e-05, "loss": 0.0238, "step": 4200 }, { "epoch": 2.382569326542162, "grad_norm": 0.4689813554286957, "learning_rate": 6.687465249472603e-05, "loss": 0.0374, "step": 4210 }, { "epoch": 2.388228636106395, "grad_norm": 0.6826220750808716, "learning_rate": 6.671891476220055e-05, "loss": 0.0393, "step": 4220 }, { "epoch": 2.3938879456706283, "grad_norm": 0.9193120002746582, "learning_rate": 6.656299419430183e-05, "loss": 0.0447, "step": 4230 }, { "epoch": 2.3995472552348613, "grad_norm": 0.62978595495224, "learning_rate": 6.640689249615223e-05, "loss": 0.0501, "step": 4240 }, { "epoch": 2.4052065647990943, "grad_norm": 0.6345353126525879, "learning_rate": 6.625061137485491e-05, "loss": 0.0448, "step": 4250 }, { "epoch": 2.4108658743633278, "grad_norm": 0.8327631950378418, "learning_rate": 6.609415253947517e-05, "loss": 0.0466, "step": 4260 }, { "epoch": 2.4165251839275608, "grad_norm": 0.4961281716823578, "learning_rate": 6.593751770102178e-05, "loss": 0.0396, "step": 4270 }, { "epoch": 2.4221844934917938, "grad_norm": 0.47766831517219543, "learning_rate": 6.578070857242823e-05, "loss": 0.0425, "step": 4280 }, { "epoch": 2.427843803056027, "grad_norm": 0.5118998885154724, "learning_rate": 6.562372686853402e-05, "loss": 0.038, "step": 4290 }, { "epoch": 2.43350311262026, "grad_norm": 0.5439344048500061, "learning_rate": 6.546657430606593e-05, "loss": 0.0358, "step": 4300 }, { "epoch": 2.4391624221844936, "grad_norm": 0.8164300918579102, "learning_rate": 6.530925260361918e-05, "loss": 0.0281, "step": 4310 }, { "epoch": 2.4448217317487266, "grad_norm": 0.6689225435256958, "learning_rate": 6.515176348163871e-05, "loss": 0.0587, "step": 4320 }, { "epoch": 2.4504810413129596, "grad_norm": 0.4825805425643921, "learning_rate": 6.499410866240032e-05, "loss": 0.0347, "step": 4330 }, { "epoch": 2.456140350877193, "grad_norm": 0.5034956336021423, "learning_rate": 6.48362898699919e-05, "loss": 0.0274, "step": 4340 }, { "epoch": 2.461799660441426, "grad_norm": 0.45087069272994995, "learning_rate": 6.467830883029443e-05, "loss": 0.0291, "step": 4350 }, { "epoch": 2.4674589700056595, "grad_norm": 0.7518239617347717, "learning_rate": 6.452016727096326e-05, "loss": 0.0311, "step": 4360 }, { "epoch": 2.4731182795698925, "grad_norm": 0.3112383782863617, "learning_rate": 6.436186692140916e-05, "loss": 0.0286, "step": 4370 }, { "epoch": 2.4787775891341255, "grad_norm": 0.6246465444564819, "learning_rate": 6.420340951277938e-05, "loss": 0.0332, "step": 4380 }, { "epoch": 2.484436898698359, "grad_norm": 0.4958966076374054, "learning_rate": 6.404479677793874e-05, "loss": 0.0274, "step": 4390 }, { "epoch": 2.490096208262592, "grad_norm": 0.46587511897087097, "learning_rate": 6.388603045145075e-05, "loss": 0.0361, "step": 4400 }, { "epoch": 2.495755517826825, "grad_norm": 0.7619835138320923, "learning_rate": 6.372711226955843e-05, "loss": 0.0284, "step": 4410 }, { "epoch": 2.5014148273910584, "grad_norm": 0.3014388084411621, "learning_rate": 6.356804397016564e-05, "loss": 0.0408, "step": 4420 }, { "epoch": 2.5070741369552914, "grad_norm": 0.5010896921157837, "learning_rate": 6.340882729281779e-05, "loss": 0.0377, "step": 4430 }, { "epoch": 2.5127334465195243, "grad_norm": 0.47420522570610046, "learning_rate": 6.324946397868294e-05, "loss": 0.0406, "step": 4440 }, { "epoch": 2.518392756083758, "grad_norm": 0.7290257811546326, "learning_rate": 6.308995577053276e-05, "loss": 0.0452, "step": 4450 }, { "epoch": 2.524052065647991, "grad_norm": 0.5837309956550598, "learning_rate": 6.293030441272347e-05, "loss": 0.0405, "step": 4460 }, { "epoch": 2.5297113752122242, "grad_norm": 0.6096610426902771, "learning_rate": 6.277051165117677e-05, "loss": 0.0339, "step": 4470 }, { "epoch": 2.535370684776457, "grad_norm": 0.673401951789856, "learning_rate": 6.261057923336064e-05, "loss": 0.0339, "step": 4480 }, { "epoch": 2.5410299943406907, "grad_norm": 0.726558268070221, "learning_rate": 6.245050890827042e-05, "loss": 0.0326, "step": 4490 }, { "epoch": 2.5466893039049237, "grad_norm": 0.8387537002563477, "learning_rate": 6.229030242640952e-05, "loss": 0.0391, "step": 4500 }, { "epoch": 2.5523486134691566, "grad_norm": 0.904774010181427, "learning_rate": 6.212996153977037e-05, "loss": 0.0531, "step": 4510 }, { "epoch": 2.55800792303339, "grad_norm": 0.5178611874580383, "learning_rate": 6.196948800181523e-05, "loss": 0.0339, "step": 4520 }, { "epoch": 2.563667232597623, "grad_norm": 0.5299800038337708, "learning_rate": 6.180888356745695e-05, "loss": 0.0364, "step": 4530 }, { "epoch": 2.569326542161856, "grad_norm": 0.515885591506958, "learning_rate": 6.164814999303995e-05, "loss": 0.0358, "step": 4540 }, { "epoch": 2.5749858517260895, "grad_norm": 0.5696628093719482, "learning_rate": 6.148728903632081e-05, "loss": 0.0407, "step": 4550 }, { "epoch": 2.5806451612903225, "grad_norm": 0.45487692952156067, "learning_rate": 6.132630245644921e-05, "loss": 0.0294, "step": 4560 }, { "epoch": 2.5863044708545555, "grad_norm": 0.5714346170425415, "learning_rate": 6.116519201394857e-05, "loss": 0.0322, "step": 4570 }, { "epoch": 2.591963780418789, "grad_norm": 0.3328671455383301, "learning_rate": 6.10039594706969e-05, "loss": 0.0379, "step": 4580 }, { "epoch": 2.597623089983022, "grad_norm": 0.5089686512947083, "learning_rate": 6.084260658990744e-05, "loss": 0.0393, "step": 4590 }, { "epoch": 2.6032823995472554, "grad_norm": 0.4880335330963135, "learning_rate": 6.068113513610943e-05, "loss": 0.0346, "step": 4600 }, { "epoch": 2.6089417091114884, "grad_norm": 0.3896096348762512, "learning_rate": 6.0519546875128876e-05, "loss": 0.0328, "step": 4610 }, { "epoch": 2.614601018675722, "grad_norm": 0.8328500390052795, "learning_rate": 6.035784357406906e-05, "loss": 0.0486, "step": 4620 }, { "epoch": 2.620260328239955, "grad_norm": 0.8080193400382996, "learning_rate": 6.01960270012914e-05, "loss": 0.0339, "step": 4630 }, { "epoch": 2.625919637804188, "grad_norm": 0.5415770411491394, "learning_rate": 6.003409892639599e-05, "loss": 0.0391, "step": 4640 }, { "epoch": 2.6315789473684212, "grad_norm": 0.6859717965126038, "learning_rate": 5.9872061120202336e-05, "loss": 0.031, "step": 4650 }, { "epoch": 2.6372382569326542, "grad_norm": 0.6185169816017151, "learning_rate": 5.9709915354729914e-05, "loss": 0.0375, "step": 4660 }, { "epoch": 2.6428975664968872, "grad_norm": 0.5620154142379761, "learning_rate": 5.9547663403178824e-05, "loss": 0.0308, "step": 4670 }, { "epoch": 2.6485568760611207, "grad_norm": 0.3828488886356354, "learning_rate": 5.9385307039910445e-05, "loss": 0.0374, "step": 4680 }, { "epoch": 2.6542161856253537, "grad_norm": 0.5923011898994446, "learning_rate": 5.922284804042792e-05, "loss": 0.0322, "step": 4690 }, { "epoch": 2.6598754951895867, "grad_norm": 0.4397177994251251, "learning_rate": 5.906028818135687e-05, "loss": 0.0413, "step": 4700 }, { "epoch": 2.66553480475382, "grad_norm": 0.4128766357898712, "learning_rate": 5.889762924042585e-05, "loss": 0.0279, "step": 4710 }, { "epoch": 2.671194114318053, "grad_norm": 0.47550642490386963, "learning_rate": 5.873487299644699e-05, "loss": 0.0363, "step": 4720 }, { "epoch": 2.676853423882286, "grad_norm": 0.6374404430389404, "learning_rate": 5.857202122929649e-05, "loss": 0.0366, "step": 4730 }, { "epoch": 2.6825127334465195, "grad_norm": 0.3073052763938904, "learning_rate": 5.840907571989518e-05, "loss": 0.025, "step": 4740 }, { "epoch": 2.688172043010753, "grad_norm": 0.5139605402946472, "learning_rate": 5.824603825018904e-05, "loss": 0.0363, "step": 4750 }, { "epoch": 2.693831352574986, "grad_norm": 0.6438497304916382, "learning_rate": 5.808291060312975e-05, "loss": 0.0336, "step": 4760 }, { "epoch": 2.699490662139219, "grad_norm": 0.8812306523323059, "learning_rate": 5.7919694562655083e-05, "loss": 0.0401, "step": 4770 }, { "epoch": 2.7051499717034524, "grad_norm": 0.35547712445259094, "learning_rate": 5.775639191366954e-05, "loss": 0.0545, "step": 4780 }, { "epoch": 2.7108092812676854, "grad_norm": 0.3681725561618805, "learning_rate": 5.75930044420247e-05, "loss": 0.033, "step": 4790 }, { "epoch": 2.7164685908319184, "grad_norm": 0.3613463044166565, "learning_rate": 5.74295339344998e-05, "loss": 0.0247, "step": 4800 }, { "epoch": 2.722127900396152, "grad_norm": 0.37430834770202637, "learning_rate": 5.726598217878211e-05, "loss": 0.0276, "step": 4810 }, { "epoch": 2.727787209960385, "grad_norm": 0.763028621673584, "learning_rate": 5.71023509634474e-05, "loss": 0.0489, "step": 4820 }, { "epoch": 2.733446519524618, "grad_norm": 0.9758390784263611, "learning_rate": 5.693864207794049e-05, "loss": 0.0353, "step": 4830 }, { "epoch": 2.7391058290888513, "grad_norm": 0.6927124261856079, "learning_rate": 5.677485731255545e-05, "loss": 0.0402, "step": 4840 }, { "epoch": 2.7447651386530842, "grad_norm": 0.4335505962371826, "learning_rate": 5.6610998458416296e-05, "loss": 0.0323, "step": 4850 }, { "epoch": 2.7504244482173172, "grad_norm": 0.642966091632843, "learning_rate": 5.644706730745716e-05, "loss": 0.0328, "step": 4860 }, { "epoch": 2.7560837577815507, "grad_norm": 0.4464760422706604, "learning_rate": 5.628306565240287e-05, "loss": 0.0293, "step": 4870 }, { "epoch": 2.7617430673457837, "grad_norm": 0.4965338110923767, "learning_rate": 5.611899528674923e-05, "loss": 0.0284, "step": 4880 }, { "epoch": 2.767402376910017, "grad_norm": 0.7036542892456055, "learning_rate": 5.595485800474349e-05, "loss": 0.0504, "step": 4890 }, { "epoch": 2.77306168647425, "grad_norm": 0.29114800691604614, "learning_rate": 5.579065560136467e-05, "loss": 0.0285, "step": 4900 }, { "epoch": 2.7787209960384835, "grad_norm": 0.6166044473648071, "learning_rate": 5.562638987230392e-05, "loss": 0.026, "step": 4910 }, { "epoch": 2.7843803056027165, "grad_norm": 0.637798011302948, "learning_rate": 5.546206261394498e-05, "loss": 0.0361, "step": 4920 }, { "epoch": 2.7900396151669495, "grad_norm": 0.3737649917602539, "learning_rate": 5.529767562334437e-05, "loss": 0.0255, "step": 4930 }, { "epoch": 2.795698924731183, "grad_norm": 0.5660009980201721, "learning_rate": 5.5133230698211926e-05, "loss": 0.0311, "step": 4940 }, { "epoch": 2.801358234295416, "grad_norm": 0.40252718329429626, "learning_rate": 5.496872963689096e-05, "loss": 0.0305, "step": 4950 }, { "epoch": 2.807017543859649, "grad_norm": 0.6519100666046143, "learning_rate": 5.4804174238338756e-05, "loss": 0.0279, "step": 4960 }, { "epoch": 2.8126768534238824, "grad_norm": 0.3294301927089691, "learning_rate": 5.463956630210678e-05, "loss": 0.0375, "step": 4970 }, { "epoch": 2.8183361629881154, "grad_norm": 0.3593049645423889, "learning_rate": 5.4474907628321046e-05, "loss": 0.0304, "step": 4980 }, { "epoch": 2.8239954725523484, "grad_norm": 0.3873436152935028, "learning_rate": 5.431020001766244e-05, "loss": 0.031, "step": 4990 }, { "epoch": 2.829654782116582, "grad_norm": 0.7387383580207825, "learning_rate": 5.4145445271346986e-05, "loss": 0.0448, "step": 5000 }, { "epoch": 2.835314091680815, "grad_norm": 0.5776609182357788, "learning_rate": 5.398064519110622e-05, "loss": 0.0412, "step": 5010 }, { "epoch": 2.8409734012450483, "grad_norm": 0.34580954909324646, "learning_rate": 5.3815801579167394e-05, "loss": 0.0363, "step": 5020 }, { "epoch": 2.8466327108092813, "grad_norm": 0.3362334072589874, "learning_rate": 5.365091623823382e-05, "loss": 0.0291, "step": 5030 }, { "epoch": 2.8522920203735147, "grad_norm": 0.320499986410141, "learning_rate": 5.348599097146521e-05, "loss": 0.0213, "step": 5040 }, { "epoch": 2.8579513299377477, "grad_norm": 0.5263609290122986, "learning_rate": 5.3321027582457836e-05, "loss": 0.0308, "step": 5050 }, { "epoch": 2.8636106395019807, "grad_norm": 0.34841442108154297, "learning_rate": 5.315602787522491e-05, "loss": 0.027, "step": 5060 }, { "epoch": 2.869269949066214, "grad_norm": 0.47967901825904846, "learning_rate": 5.299099365417678e-05, "loss": 0.0405, "step": 5070 }, { "epoch": 2.874929258630447, "grad_norm": 0.8032472729682922, "learning_rate": 5.2825926724101236e-05, "loss": 0.0363, "step": 5080 }, { "epoch": 2.88058856819468, "grad_norm": 0.5181282758712769, "learning_rate": 5.26608288901438e-05, "loss": 0.0348, "step": 5090 }, { "epoch": 2.8862478777589136, "grad_norm": 0.4246527850627899, "learning_rate": 5.24957019577879e-05, "loss": 0.0375, "step": 5100 }, { "epoch": 2.8919071873231466, "grad_norm": 0.9782278537750244, "learning_rate": 5.2330547732835266e-05, "loss": 0.041, "step": 5110 }, { "epoch": 2.8975664968873796, "grad_norm": 0.8772940039634705, "learning_rate": 5.2165368021385996e-05, "loss": 0.0389, "step": 5120 }, { "epoch": 2.903225806451613, "grad_norm": 0.5722554326057434, "learning_rate": 5.200016462981897e-05, "loss": 0.0335, "step": 5130 }, { "epoch": 2.908885116015846, "grad_norm": 0.5590901970863342, "learning_rate": 5.1834939364772015e-05, "loss": 0.0343, "step": 5140 }, { "epoch": 2.914544425580079, "grad_norm": 0.4556905925273895, "learning_rate": 5.166969403312214e-05, "loss": 0.0328, "step": 5150 }, { "epoch": 2.9202037351443124, "grad_norm": 0.7099003791809082, "learning_rate": 5.1504430441965844e-05, "loss": 0.0374, "step": 5160 }, { "epoch": 2.9258630447085454, "grad_norm": 0.518718957901001, "learning_rate": 5.133915039859923e-05, "loss": 0.0264, "step": 5170 }, { "epoch": 2.931522354272779, "grad_norm": 0.32491767406463623, "learning_rate": 5.1173855710498444e-05, "loss": 0.0258, "step": 5180 }, { "epoch": 2.937181663837012, "grad_norm": 0.5351004004478455, "learning_rate": 5.100854818529967e-05, "loss": 0.0417, "step": 5190 }, { "epoch": 2.9428409734012453, "grad_norm": 0.4286375343799591, "learning_rate": 5.084322963077951e-05, "loss": 0.0273, "step": 5200 }, { "epoch": 2.9485002829654783, "grad_norm": 0.6190813779830933, "learning_rate": 5.067790185483522e-05, "loss": 0.0329, "step": 5210 }, { "epoch": 2.9541595925297113, "grad_norm": 0.6987380981445312, "learning_rate": 5.0512566665464844e-05, "loss": 0.0241, "step": 5220 }, { "epoch": 2.9598189020939447, "grad_norm": 0.3838080167770386, "learning_rate": 5.034722587074755e-05, "loss": 0.0282, "step": 5230 }, { "epoch": 2.9654782116581777, "grad_norm": 0.6405492424964905, "learning_rate": 5.018188127882375e-05, "loss": 0.0351, "step": 5240 }, { "epoch": 2.9711375212224107, "grad_norm": 0.5722482800483704, "learning_rate": 5.0016534697875417e-05, "loss": 0.0315, "step": 5250 }, { "epoch": 2.976796830786644, "grad_norm": 0.5997118949890137, "learning_rate": 4.9851187936106294e-05, "loss": 0.0244, "step": 5260 }, { "epoch": 2.982456140350877, "grad_norm": 0.3658483028411865, "learning_rate": 4.968584280172206e-05, "loss": 0.0302, "step": 5270 }, { "epoch": 2.98811544991511, "grad_norm": 0.43296298384666443, "learning_rate": 4.95205011029106e-05, "loss": 0.0369, "step": 5280 }, { "epoch": 2.9937747594793436, "grad_norm": 0.26728492975234985, "learning_rate": 4.935516464782227e-05, "loss": 0.0343, "step": 5290 }, { "epoch": 2.9994340690435766, "grad_norm": 0.6191973090171814, "learning_rate": 4.918983524455003e-05, "loss": 0.025, "step": 5300 }, { "epoch": 3.00509337860781, "grad_norm": 0.4005588889122009, "learning_rate": 4.9024514701109766e-05, "loss": 0.02, "step": 5310 }, { "epoch": 3.010752688172043, "grad_norm": 0.4203624725341797, "learning_rate": 4.885920482542043e-05, "loss": 0.0254, "step": 5320 }, { "epoch": 3.016411997736276, "grad_norm": 0.41406959295272827, "learning_rate": 4.869390742528438e-05, "loss": 0.0389, "step": 5330 }, { "epoch": 3.0220713073005094, "grad_norm": 0.3125941753387451, "learning_rate": 4.852862430836744e-05, "loss": 0.0284, "step": 5340 }, { "epoch": 3.0277306168647424, "grad_norm": 0.6467152237892151, "learning_rate": 4.836335728217933e-05, "loss": 0.0294, "step": 5350 }, { "epoch": 3.033389926428976, "grad_norm": 0.46983492374420166, "learning_rate": 4.819810815405379e-05, "loss": 0.0271, "step": 5360 }, { "epoch": 3.039049235993209, "grad_norm": 0.31741878390312195, "learning_rate": 4.803287873112877e-05, "loss": 0.0267, "step": 5370 }, { "epoch": 3.044708545557442, "grad_norm": 0.28420332074165344, "learning_rate": 4.786767082032681e-05, "loss": 0.0265, "step": 5380 }, { "epoch": 3.0503678551216753, "grad_norm": 0.5284949541091919, "learning_rate": 4.77024862283351e-05, "loss": 0.0208, "step": 5390 }, { "epoch": 3.0560271646859083, "grad_norm": 0.5990930795669556, "learning_rate": 4.753732676158593e-05, "loss": 0.0288, "step": 5400 }, { "epoch": 3.0616864742501413, "grad_norm": 0.4833022654056549, "learning_rate": 4.737219422623672e-05, "loss": 0.0245, "step": 5410 }, { "epoch": 3.0673457838143747, "grad_norm": 0.4014180600643158, "learning_rate": 4.720709042815044e-05, "loss": 0.0263, "step": 5420 }, { "epoch": 3.0730050933786077, "grad_norm": 0.2624795138835907, "learning_rate": 4.704201717287578e-05, "loss": 0.0203, "step": 5430 }, { "epoch": 3.078664402942841, "grad_norm": 0.5820720791816711, "learning_rate": 4.6876976265627404e-05, "loss": 0.0372, "step": 5440 }, { "epoch": 3.084323712507074, "grad_norm": 0.4845047891139984, "learning_rate": 4.671196951126626e-05, "loss": 0.0205, "step": 5450 }, { "epoch": 3.089983022071307, "grad_norm": 0.3671550750732422, "learning_rate": 4.654699871427971e-05, "loss": 0.0252, "step": 5460 }, { "epoch": 3.0956423316355406, "grad_norm": 0.645203173160553, "learning_rate": 4.6382065678762034e-05, "loss": 0.0296, "step": 5470 }, { "epoch": 3.1013016411997736, "grad_norm": 0.2925666272640228, "learning_rate": 4.6217172208394424e-05, "loss": 0.0213, "step": 5480 }, { "epoch": 3.1069609507640066, "grad_norm": 0.40954533219337463, "learning_rate": 4.605232010642549e-05, "loss": 0.0284, "step": 5490 }, { "epoch": 3.11262026032824, "grad_norm": 0.377370685338974, "learning_rate": 4.588751117565142e-05, "loss": 0.0331, "step": 5500 }, { "epoch": 3.118279569892473, "grad_norm": 0.4786716103553772, "learning_rate": 4.5722747218396214e-05, "loss": 0.033, "step": 5510 }, { "epoch": 3.1239388794567065, "grad_norm": 0.2982279360294342, "learning_rate": 4.5558030036492194e-05, "loss": 0.0236, "step": 5520 }, { "epoch": 3.1295981890209394, "grad_norm": 0.3954036235809326, "learning_rate": 4.539336143125999e-05, "loss": 0.026, "step": 5530 }, { "epoch": 3.1352574985851724, "grad_norm": 0.46236851811408997, "learning_rate": 4.522874320348916e-05, "loss": 0.0202, "step": 5540 }, { "epoch": 3.140916808149406, "grad_norm": 0.5871313214302063, "learning_rate": 4.506417715341821e-05, "loss": 0.029, "step": 5550 }, { "epoch": 3.146576117713639, "grad_norm": 0.3890996277332306, "learning_rate": 4.489966508071511e-05, "loss": 0.0233, "step": 5560 }, { "epoch": 3.152235427277872, "grad_norm": 0.6300520300865173, "learning_rate": 4.4735208784457575e-05, "loss": 0.0327, "step": 5570 }, { "epoch": 3.1578947368421053, "grad_norm": 0.3230230510234833, "learning_rate": 4.457081006311325e-05, "loss": 0.0234, "step": 5580 }, { "epoch": 3.1635540464063383, "grad_norm": 0.5044695734977722, "learning_rate": 4.440647071452027e-05, "loss": 0.0263, "step": 5590 }, { "epoch": 3.1692133559705717, "grad_norm": 0.42241913080215454, "learning_rate": 4.424219253586737e-05, "loss": 0.0238, "step": 5600 }, { "epoch": 3.1748726655348047, "grad_norm": 0.506126880645752, "learning_rate": 4.407797732367443e-05, "loss": 0.0278, "step": 5610 }, { "epoch": 3.1805319750990377, "grad_norm": 0.3469611406326294, "learning_rate": 4.391382687377268e-05, "loss": 0.0186, "step": 5620 }, { "epoch": 3.186191284663271, "grad_norm": 0.3190121054649353, "learning_rate": 4.374974298128512e-05, "loss": 0.0229, "step": 5630 }, { "epoch": 3.191850594227504, "grad_norm": 0.224530428647995, "learning_rate": 4.358572744060699e-05, "loss": 0.0218, "step": 5640 }, { "epoch": 3.1975099037917376, "grad_norm": 0.583054780960083, "learning_rate": 4.342178204538588e-05, "loss": 0.0362, "step": 5650 }, { "epoch": 3.2031692133559706, "grad_norm": 0.5004111528396606, "learning_rate": 4.325790858850241e-05, "loss": 0.0223, "step": 5660 }, { "epoch": 3.2088285229202036, "grad_norm": 0.4681668281555176, "learning_rate": 4.309410886205043e-05, "loss": 0.0257, "step": 5670 }, { "epoch": 3.214487832484437, "grad_norm": 0.25892138481140137, "learning_rate": 4.293038465731752e-05, "loss": 0.0309, "step": 5680 }, { "epoch": 3.22014714204867, "grad_norm": 0.659767210483551, "learning_rate": 4.276673776476533e-05, "loss": 0.0349, "step": 5690 }, { "epoch": 3.225806451612903, "grad_norm": 0.5110019445419312, "learning_rate": 4.260316997401007e-05, "loss": 0.032, "step": 5700 }, { "epoch": 3.2314657611771365, "grad_norm": 0.8555366396903992, "learning_rate": 4.243968307380293e-05, "loss": 0.0266, "step": 5710 }, { "epoch": 3.2371250707413695, "grad_norm": 0.37470516562461853, "learning_rate": 4.22762788520104e-05, "loss": 0.0279, "step": 5720 }, { "epoch": 3.242784380305603, "grad_norm": 0.7342115044593811, "learning_rate": 4.211295909559491e-05, "loss": 0.0273, "step": 5730 }, { "epoch": 3.248443689869836, "grad_norm": 0.6888209581375122, "learning_rate": 4.194972559059511e-05, "loss": 0.0249, "step": 5740 }, { "epoch": 3.254102999434069, "grad_norm": 0.5219807624816895, "learning_rate": 4.178658012210651e-05, "loss": 0.0232, "step": 5750 }, { "epoch": 3.2597623089983023, "grad_norm": 0.8390067219734192, "learning_rate": 4.162352447426177e-05, "loss": 0.0256, "step": 5760 }, { "epoch": 3.2654216185625353, "grad_norm": 0.6247023940086365, "learning_rate": 4.146056043021135e-05, "loss": 0.0281, "step": 5770 }, { "epoch": 3.2710809281267688, "grad_norm": 0.4523935616016388, "learning_rate": 4.1297689772103944e-05, "loss": 0.0286, "step": 5780 }, { "epoch": 3.2767402376910018, "grad_norm": 0.5269297361373901, "learning_rate": 4.113491428106694e-05, "loss": 0.0268, "step": 5790 }, { "epoch": 3.2823995472552348, "grad_norm": 0.48714950680732727, "learning_rate": 4.0972235737187055e-05, "loss": 0.0321, "step": 5800 }, { "epoch": 3.288058856819468, "grad_norm": 0.5200141072273254, "learning_rate": 4.080965591949076e-05, "loss": 0.0192, "step": 5810 }, { "epoch": 3.293718166383701, "grad_norm": 0.5562568306922913, "learning_rate": 4.0647176605924924e-05, "loss": 0.0223, "step": 5820 }, { "epoch": 3.299377475947934, "grad_norm": 0.5858548879623413, "learning_rate": 4.0484799573337255e-05, "loss": 0.0297, "step": 5830 }, { "epoch": 3.3050367855121676, "grad_norm": 0.33869001269340515, "learning_rate": 4.032252659745699e-05, "loss": 0.0251, "step": 5840 }, { "epoch": 3.3106960950764006, "grad_norm": 0.45930442214012146, "learning_rate": 4.016035945287539e-05, "loss": 0.0187, "step": 5850 }, { "epoch": 3.3163554046406336, "grad_norm": 0.5437840819358826, "learning_rate": 3.999829991302635e-05, "loss": 0.0303, "step": 5860 }, { "epoch": 3.322014714204867, "grad_norm": 0.5779958367347717, "learning_rate": 3.983634975016707e-05, "loss": 0.0327, "step": 5870 }, { "epoch": 3.3276740237691, "grad_norm": 0.8774251937866211, "learning_rate": 3.967451073535854e-05, "loss": 0.0276, "step": 5880 }, { "epoch": 3.3333333333333335, "grad_norm": 0.6216500401496887, "learning_rate": 3.951278463844633e-05, "loss": 0.033, "step": 5890 }, { "epoch": 3.3389926428975665, "grad_norm": 0.5506871342658997, "learning_rate": 3.935117322804111e-05, "loss": 0.0454, "step": 5900 }, { "epoch": 3.3446519524617995, "grad_norm": 0.7217140793800354, "learning_rate": 3.918967827149938e-05, "loss": 0.0332, "step": 5910 }, { "epoch": 3.350311262026033, "grad_norm": 0.6366479992866516, "learning_rate": 3.9028301534904094e-05, "loss": 0.026, "step": 5920 }, { "epoch": 3.355970571590266, "grad_norm": 0.4815925657749176, "learning_rate": 3.88670447830454e-05, "loss": 0.0232, "step": 5930 }, { "epoch": 3.3616298811544993, "grad_norm": 0.9379783272743225, "learning_rate": 3.870590977940132e-05, "loss": 0.0277, "step": 5940 }, { "epoch": 3.3672891907187323, "grad_norm": 0.4577707052230835, "learning_rate": 3.8544898286118404e-05, "loss": 0.0233, "step": 5950 }, { "epoch": 3.3729485002829653, "grad_norm": 0.36198124289512634, "learning_rate": 3.838401206399257e-05, "loss": 0.0228, "step": 5960 }, { "epoch": 3.3786078098471988, "grad_norm": 0.3477252423763275, "learning_rate": 3.822325287244975e-05, "loss": 0.0193, "step": 5970 }, { "epoch": 3.3842671194114318, "grad_norm": 0.5713914036750793, "learning_rate": 3.8062622469526725e-05, "loss": 0.0276, "step": 5980 }, { "epoch": 3.3899264289756648, "grad_norm": 0.6365956664085388, "learning_rate": 3.790212261185183e-05, "loss": 0.0328, "step": 5990 }, { "epoch": 3.395585738539898, "grad_norm": 0.7436591386795044, "learning_rate": 3.7741755054625794e-05, "loss": 0.0343, "step": 6000 }, { "epoch": 3.401245048104131, "grad_norm": 1.1299272775650024, "learning_rate": 3.758152155160255e-05, "loss": 0.0435, "step": 6010 }, { "epoch": 3.4069043576683646, "grad_norm": 0.3610575497150421, "learning_rate": 3.742142385506999e-05, "loss": 0.023, "step": 6020 }, { "epoch": 3.4125636672325976, "grad_norm": 0.5312737226486206, "learning_rate": 3.72614637158309e-05, "loss": 0.0263, "step": 6030 }, { "epoch": 3.4182229767968306, "grad_norm": 0.5494548678398132, "learning_rate": 3.710164288318371e-05, "loss": 0.032, "step": 6040 }, { "epoch": 3.423882286361064, "grad_norm": 0.4840650260448456, "learning_rate": 3.694196310490345e-05, "loss": 0.0264, "step": 6050 }, { "epoch": 3.429541595925297, "grad_norm": 0.5452207922935486, "learning_rate": 3.678242612722259e-05, "loss": 0.0278, "step": 6060 }, { "epoch": 3.4352009054895305, "grad_norm": 0.3490133285522461, "learning_rate": 3.6623033694811953e-05, "loss": 0.0211, "step": 6070 }, { "epoch": 3.4408602150537635, "grad_norm": 0.39896267652511597, "learning_rate": 3.6463787550761665e-05, "loss": 0.0354, "step": 6080 }, { "epoch": 3.4465195246179965, "grad_norm": 0.5362284779548645, "learning_rate": 3.630468943656202e-05, "loss": 0.0293, "step": 6090 }, { "epoch": 3.45217883418223, "grad_norm": 0.4318121075630188, "learning_rate": 3.6145741092084523e-05, "loss": 0.0317, "step": 6100 }, { "epoch": 3.457838143746463, "grad_norm": 0.5870428681373596, "learning_rate": 3.598694425556278e-05, "loss": 0.0269, "step": 6110 }, { "epoch": 3.463497453310696, "grad_norm": 0.5336974263191223, "learning_rate": 3.58283006635736e-05, "loss": 0.0257, "step": 6120 }, { "epoch": 3.4691567628749294, "grad_norm": 0.406678169965744, "learning_rate": 3.566981205101781e-05, "loss": 0.0183, "step": 6130 }, { "epoch": 3.4748160724391624, "grad_norm": 0.634650707244873, "learning_rate": 3.5511480151101556e-05, "loss": 0.0249, "step": 6140 }, { "epoch": 3.4804753820033953, "grad_norm": 0.27335265278816223, "learning_rate": 3.5353306695317104e-05, "loss": 0.031, "step": 6150 }, { "epoch": 3.486134691567629, "grad_norm": 0.3521565794944763, "learning_rate": 3.519529341342402e-05, "loss": 0.0414, "step": 6160 }, { "epoch": 3.491794001131862, "grad_norm": 0.3947822153568268, "learning_rate": 3.503744203343026e-05, "loss": 0.0339, "step": 6170 }, { "epoch": 3.497453310696095, "grad_norm": 0.4416903257369995, "learning_rate": 3.487975428157318e-05, "loss": 0.027, "step": 6180 }, { "epoch": 3.503112620260328, "grad_norm": 0.4432899057865143, "learning_rate": 3.472223188230083e-05, "loss": 0.0288, "step": 6190 }, { "epoch": 3.5087719298245617, "grad_norm": 0.45889851450920105, "learning_rate": 3.4564876558252866e-05, "loss": 0.0241, "step": 6200 }, { "epoch": 3.5144312393887946, "grad_norm": 0.4277569353580475, "learning_rate": 3.440769003024195e-05, "loss": 0.0254, "step": 6210 }, { "epoch": 3.5200905489530276, "grad_norm": 0.47739022970199585, "learning_rate": 3.425067401723477e-05, "loss": 0.0268, "step": 6220 }, { "epoch": 3.525749858517261, "grad_norm": 0.789215087890625, "learning_rate": 3.409383023633325e-05, "loss": 0.0303, "step": 6230 }, { "epoch": 3.531409168081494, "grad_norm": 0.2821437418460846, "learning_rate": 3.3937160402755894e-05, "loss": 0.0258, "step": 6240 }, { "epoch": 3.537068477645727, "grad_norm": 0.4800223410129547, "learning_rate": 3.378066622981885e-05, "loss": 0.0338, "step": 6250 }, { "epoch": 3.5427277872099605, "grad_norm": 0.3567802608013153, "learning_rate": 3.362434942891738e-05, "loss": 0.0243, "step": 6260 }, { "epoch": 3.5483870967741935, "grad_norm": 0.4538363814353943, "learning_rate": 3.346821170950693e-05, "loss": 0.0249, "step": 6270 }, { "epoch": 3.5540464063384265, "grad_norm": 0.36534032225608826, "learning_rate": 3.3312254779084585e-05, "loss": 0.028, "step": 6280 }, { "epoch": 3.55970571590266, "grad_norm": 0.4637555778026581, "learning_rate": 3.315648034317039e-05, "loss": 0.0309, "step": 6290 }, { "epoch": 3.565365025466893, "grad_norm": 0.31388649344444275, "learning_rate": 3.3000890105288564e-05, "loss": 0.0321, "step": 6300 }, { "epoch": 3.571024335031126, "grad_norm": 0.5784611105918884, "learning_rate": 3.284548576694908e-05, "loss": 0.0261, "step": 6310 }, { "epoch": 3.5766836445953594, "grad_norm": 0.4061717092990875, "learning_rate": 3.2690269027628815e-05, "loss": 0.0224, "step": 6320 }, { "epoch": 3.582342954159593, "grad_norm": 0.34512901306152344, "learning_rate": 3.253524158475324e-05, "loss": 0.0274, "step": 6330 }, { "epoch": 3.588002263723826, "grad_norm": 0.2854452133178711, "learning_rate": 3.238040513367757e-05, "loss": 0.031, "step": 6340 }, { "epoch": 3.593661573288059, "grad_norm": 0.4145554304122925, "learning_rate": 3.222576136766843e-05, "loss": 0.0248, "step": 6350 }, { "epoch": 3.5993208828522922, "grad_norm": 0.4759776294231415, "learning_rate": 3.2071311977885324e-05, "loss": 0.0341, "step": 6360 }, { "epoch": 3.6049801924165252, "grad_norm": 0.39346179366111755, "learning_rate": 3.191705865336197e-05, "loss": 0.0162, "step": 6370 }, { "epoch": 3.6106395019807582, "grad_norm": 0.25690701603889465, "learning_rate": 3.1763003080988075e-05, "loss": 0.0231, "step": 6380 }, { "epoch": 3.6162988115449917, "grad_norm": 0.5031017065048218, "learning_rate": 3.160914694549063e-05, "loss": 0.033, "step": 6390 }, { "epoch": 3.6219581211092247, "grad_norm": 0.4231465458869934, "learning_rate": 3.145549192941573e-05, "loss": 0.0217, "step": 6400 }, { "epoch": 3.6276174306734577, "grad_norm": 0.5258224606513977, "learning_rate": 3.130203971310999e-05, "loss": 0.0222, "step": 6410 }, { "epoch": 3.633276740237691, "grad_norm": 0.3988799750804901, "learning_rate": 3.114879197470225e-05, "loss": 0.0264, "step": 6420 }, { "epoch": 3.638936049801924, "grad_norm": 0.33015578985214233, "learning_rate": 3.0995750390085285e-05, "loss": 0.0226, "step": 6430 }, { "epoch": 3.644595359366157, "grad_norm": 0.30468347668647766, "learning_rate": 3.084291663289728e-05, "loss": 0.0223, "step": 6440 }, { "epoch": 3.6502546689303905, "grad_norm": 0.31607723236083984, "learning_rate": 3.069029237450375e-05, "loss": 0.0143, "step": 6450 }, { "epoch": 3.6559139784946235, "grad_norm": 0.31934604048728943, "learning_rate": 3.053787928397911e-05, "loss": 0.0251, "step": 6460 }, { "epoch": 3.661573288058857, "grad_norm": 0.34353768825531006, "learning_rate": 3.0385679028088526e-05, "loss": 0.0223, "step": 6470 }, { "epoch": 3.66723259762309, "grad_norm": 0.4591050446033478, "learning_rate": 3.023369327126959e-05, "loss": 0.0171, "step": 6480 }, { "epoch": 3.6728919071873234, "grad_norm": 0.4569452106952667, "learning_rate": 3.0081923675614198e-05, "loss": 0.0262, "step": 6490 }, { "epoch": 3.6785512167515564, "grad_norm": 0.3188132643699646, "learning_rate": 2.993037190085034e-05, "loss": 0.0209, "step": 6500 }, { "epoch": 3.6842105263157894, "grad_norm": 0.3605244755744934, "learning_rate": 2.977903960432392e-05, "loss": 0.0196, "step": 6510 }, { "epoch": 3.689869835880023, "grad_norm": 0.39840972423553467, "learning_rate": 2.9627928440980722e-05, "loss": 0.0207, "step": 6520 }, { "epoch": 3.695529145444256, "grad_norm": 0.32871973514556885, "learning_rate": 2.9477040063348183e-05, "loss": 0.0223, "step": 6530 }, { "epoch": 3.701188455008489, "grad_norm": 0.6761415004730225, "learning_rate": 2.9326376121517456e-05, "loss": 0.0324, "step": 6540 }, { "epoch": 3.7068477645727222, "grad_norm": 0.3579584062099457, "learning_rate": 2.9175938263125236e-05, "loss": 0.0177, "step": 6550 }, { "epoch": 3.7125070741369552, "grad_norm": 0.8756101131439209, "learning_rate": 2.9025728133335873e-05, "loss": 0.0329, "step": 6560 }, { "epoch": 3.7181663837011882, "grad_norm": 0.5549876093864441, "learning_rate": 2.8875747374823288e-05, "loss": 0.0207, "step": 6570 }, { "epoch": 3.7238256932654217, "grad_norm": 0.3724808394908905, "learning_rate": 2.872599762775298e-05, "loss": 0.0341, "step": 6580 }, { "epoch": 3.7294850028296547, "grad_norm": 0.6460519433021545, "learning_rate": 2.857648052976425e-05, "loss": 0.0311, "step": 6590 }, { "epoch": 3.735144312393888, "grad_norm": 0.48827290534973145, "learning_rate": 2.8427197715952047e-05, "loss": 0.0302, "step": 6600 }, { "epoch": 3.740803621958121, "grad_norm": 0.33132031559944153, "learning_rate": 2.8278150818849393e-05, "loss": 0.016, "step": 6610 }, { "epoch": 3.7464629315223545, "grad_norm": 0.3171054720878601, "learning_rate": 2.812934146840922e-05, "loss": 0.0175, "step": 6620 }, { "epoch": 3.7521222410865875, "grad_norm": 0.4657711684703827, "learning_rate": 2.7980771291986764e-05, "loss": 0.0199, "step": 6630 }, { "epoch": 3.7577815506508205, "grad_norm": 0.288437157869339, "learning_rate": 2.783244191432167e-05, "loss": 0.0194, "step": 6640 }, { "epoch": 3.763440860215054, "grad_norm": 0.44233739376068115, "learning_rate": 2.768435495752022e-05, "loss": 0.024, "step": 6650 }, { "epoch": 3.769100169779287, "grad_norm": 0.32472509145736694, "learning_rate": 2.753651204103771e-05, "loss": 0.0171, "step": 6660 }, { "epoch": 3.77475947934352, "grad_norm": 0.3336578607559204, "learning_rate": 2.7388914781660523e-05, "loss": 0.0194, "step": 6670 }, { "epoch": 3.7804187889077534, "grad_norm": 0.6103697419166565, "learning_rate": 2.7241564793488693e-05, "loss": 0.0206, "step": 6680 }, { "epoch": 3.7860780984719864, "grad_norm": 0.3058294653892517, "learning_rate": 2.7094463687918037e-05, "loss": 0.0187, "step": 6690 }, { "epoch": 3.7917374080362194, "grad_norm": 0.3847391903400421, "learning_rate": 2.694761307362268e-05, "loss": 0.0269, "step": 6700 }, { "epoch": 3.797396717600453, "grad_norm": 0.4608091115951538, "learning_rate": 2.6801014556537467e-05, "loss": 0.024, "step": 6710 }, { "epoch": 3.803056027164686, "grad_norm": 0.23236976563930511, "learning_rate": 2.6654669739840243e-05, "loss": 0.0289, "step": 6720 }, { "epoch": 3.808715336728919, "grad_norm": 0.3241671919822693, "learning_rate": 2.650858022393451e-05, "loss": 0.0226, "step": 6730 }, { "epoch": 3.8143746462931523, "grad_norm": 0.2845827043056488, "learning_rate": 2.6362747606431747e-05, "loss": 0.0358, "step": 6740 }, { "epoch": 3.8200339558573853, "grad_norm": 0.8334740400314331, "learning_rate": 2.6217173482134172e-05, "loss": 0.0259, "step": 6750 }, { "epoch": 3.8256932654216187, "grad_norm": 0.3732583522796631, "learning_rate": 2.6071859443017044e-05, "loss": 0.0216, "step": 6760 }, { "epoch": 3.8313525749858517, "grad_norm": 0.6674580574035645, "learning_rate": 2.5926807078211414e-05, "loss": 0.025, "step": 6770 }, { "epoch": 3.837011884550085, "grad_norm": 0.4493323564529419, "learning_rate": 2.5782017973986728e-05, "loss": 0.0212, "step": 6780 }, { "epoch": 3.842671194114318, "grad_norm": 0.40532025694847107, "learning_rate": 2.5637493713733374e-05, "loss": 0.0287, "step": 6790 }, { "epoch": 3.848330503678551, "grad_norm": 0.5855029225349426, "learning_rate": 2.549323587794559e-05, "loss": 0.0211, "step": 6800 }, { "epoch": 3.8539898132427846, "grad_norm": 0.642335832118988, "learning_rate": 2.5349246044203895e-05, "loss": 0.02, "step": 6810 }, { "epoch": 3.8596491228070176, "grad_norm": 0.24802717566490173, "learning_rate": 2.520552578715808e-05, "loss": 0.0181, "step": 6820 }, { "epoch": 3.8653084323712505, "grad_norm": 0.4547887146472931, "learning_rate": 2.506207667850981e-05, "loss": 0.0179, "step": 6830 }, { "epoch": 3.870967741935484, "grad_norm": 0.39535802602767944, "learning_rate": 2.4918900286995555e-05, "loss": 0.0347, "step": 6840 }, { "epoch": 3.876627051499717, "grad_norm": 0.6071020364761353, "learning_rate": 2.4775998178369458e-05, "loss": 0.0314, "step": 6850 }, { "epoch": 3.88228636106395, "grad_norm": 0.44565150141716003, "learning_rate": 2.4633371915386017e-05, "loss": 0.022, "step": 6860 }, { "epoch": 3.8879456706281834, "grad_norm": 0.6935356259346008, "learning_rate": 2.4491023057783235e-05, "loss": 0.0303, "step": 6870 }, { "epoch": 3.8936049801924164, "grad_norm": 0.406706303358078, "learning_rate": 2.4348953162265375e-05, "loss": 0.0307, "step": 6880 }, { "epoch": 3.89926428975665, "grad_norm": 0.33041876554489136, "learning_rate": 2.420716378248607e-05, "loss": 0.023, "step": 6890 }, { "epoch": 3.904923599320883, "grad_norm": 0.4668940007686615, "learning_rate": 2.4065656469031266e-05, "loss": 0.0328, "step": 6900 }, { "epoch": 3.9105829088851163, "grad_norm": 0.4430356025695801, "learning_rate": 2.3924432769402268e-05, "loss": 0.0231, "step": 6910 }, { "epoch": 3.9162422184493493, "grad_norm": 0.9641812443733215, "learning_rate": 2.3783494227998844e-05, "loss": 0.0229, "step": 6920 }, { "epoch": 3.9219015280135823, "grad_norm": 0.3597238063812256, "learning_rate": 2.3642842386102264e-05, "loss": 0.0234, "step": 6930 }, { "epoch": 3.9275608375778157, "grad_norm": 0.3482387661933899, "learning_rate": 2.3502478781858567e-05, "loss": 0.0261, "step": 6940 }, { "epoch": 3.9332201471420487, "grad_norm": 0.348222017288208, "learning_rate": 2.3362404950261628e-05, "loss": 0.0198, "step": 6950 }, { "epoch": 3.9388794567062817, "grad_norm": 0.4402678608894348, "learning_rate": 2.3222622423136458e-05, "loss": 0.0179, "step": 6960 }, { "epoch": 3.944538766270515, "grad_norm": 0.3544287085533142, "learning_rate": 2.3083132729122332e-05, "loss": 0.0204, "step": 6970 }, { "epoch": 3.950198075834748, "grad_norm": 0.7409390211105347, "learning_rate": 2.294393739365621e-05, "loss": 0.0246, "step": 6980 }, { "epoch": 3.955857385398981, "grad_norm": 0.6544904708862305, "learning_rate": 2.2805037938956e-05, "loss": 0.0347, "step": 6990 }, { "epoch": 3.9615166949632146, "grad_norm": 0.5040102601051331, "learning_rate": 2.266643588400386e-05, "loss": 0.02, "step": 7000 }, { "epoch": 3.9671760045274476, "grad_norm": 0.2518904209136963, "learning_rate": 2.252813274452969e-05, "loss": 0.0354, "step": 7010 }, { "epoch": 3.9728353140916806, "grad_norm": 0.4520137310028076, "learning_rate": 2.2390130032994427e-05, "loss": 0.0186, "step": 7020 }, { "epoch": 3.978494623655914, "grad_norm": 0.276895672082901, "learning_rate": 2.2252429258573633e-05, "loss": 0.0236, "step": 7030 }, { "epoch": 3.984153933220147, "grad_norm": 0.27124443650245667, "learning_rate": 2.2115031927140904e-05, "loss": 0.0199, "step": 7040 }, { "epoch": 3.9898132427843804, "grad_norm": 0.3213484287261963, "learning_rate": 2.1977939541251463e-05, "loss": 0.0176, "step": 7050 }, { "epoch": 3.9954725523486134, "grad_norm": 0.5860376954078674, "learning_rate": 2.1841153600125684e-05, "loss": 0.0247, "step": 7060 }, { "epoch": 4.001131861912847, "grad_norm": 0.2233385443687439, "learning_rate": 2.170467559963267e-05, "loss": 0.0227, "step": 7070 }, { "epoch": 4.006791171477079, "grad_norm": 0.3871348798274994, "learning_rate": 2.1568507032273982e-05, "loss": 0.028, "step": 7080 }, { "epoch": 4.012450481041313, "grad_norm": 0.6563515663146973, "learning_rate": 2.1432649387167264e-05, "loss": 0.0277, "step": 7090 }, { "epoch": 4.018109790605546, "grad_norm": 0.545616090297699, "learning_rate": 2.1297104150029973e-05, "loss": 0.0202, "step": 7100 }, { "epoch": 4.023769100169779, "grad_norm": 0.2558697462081909, "learning_rate": 2.116187280316307e-05, "loss": 0.0228, "step": 7110 }, { "epoch": 4.029428409734012, "grad_norm": 0.31855130195617676, "learning_rate": 2.1026956825434908e-05, "loss": 0.0203, "step": 7120 }, { "epoch": 4.035087719298246, "grad_norm": 0.23440046608448029, "learning_rate": 2.0892357692265017e-05, "loss": 0.0154, "step": 7130 }, { "epoch": 4.040747028862479, "grad_norm": 0.33446216583251953, "learning_rate": 2.0758076875607947e-05, "loss": 0.0322, "step": 7140 }, { "epoch": 4.046406338426712, "grad_norm": 0.2720486521720886, "learning_rate": 2.0624115843937207e-05, "loss": 0.032, "step": 7150 }, { "epoch": 4.052065647990945, "grad_norm": 0.4796851575374603, "learning_rate": 2.0490476062229157e-05, "loss": 0.0177, "step": 7160 }, { "epoch": 4.057724957555179, "grad_norm": 0.3703482151031494, "learning_rate": 2.035715899194704e-05, "loss": 0.0232, "step": 7170 }, { "epoch": 4.063384267119411, "grad_norm": 0.5428142547607422, "learning_rate": 2.022416609102499e-05, "loss": 0.0206, "step": 7180 }, { "epoch": 4.069043576683645, "grad_norm": 0.4209301769733429, "learning_rate": 2.009149881385205e-05, "loss": 0.0294, "step": 7190 }, { "epoch": 4.074702886247878, "grad_norm": 0.21641351282596588, "learning_rate": 1.995915861125634e-05, "loss": 0.0141, "step": 7200 }, { "epoch": 4.080362195812111, "grad_norm": 0.5476838946342468, "learning_rate": 1.9827146930489065e-05, "loss": 0.0164, "step": 7210 }, { "epoch": 4.086021505376344, "grad_norm": 0.4869919717311859, "learning_rate": 1.9695465215208848e-05, "loss": 0.0238, "step": 7220 }, { "epoch": 4.0916808149405774, "grad_norm": 0.164511039853096, "learning_rate": 1.9564114905465813e-05, "loss": 0.0179, "step": 7230 }, { "epoch": 4.09734012450481, "grad_norm": 0.3579815924167633, "learning_rate": 1.9433097437685936e-05, "loss": 0.0236, "step": 7240 }, { "epoch": 4.102999434069043, "grad_norm": 0.46660980582237244, "learning_rate": 1.930241424465521e-05, "loss": 0.0237, "step": 7250 }, { "epoch": 4.108658743633277, "grad_norm": 0.4709692597389221, "learning_rate": 1.9172066755504115e-05, "loss": 0.0229, "step": 7260 }, { "epoch": 4.11431805319751, "grad_norm": 0.21272122859954834, "learning_rate": 1.9042056395691914e-05, "loss": 0.0152, "step": 7270 }, { "epoch": 4.119977362761743, "grad_norm": 0.3281286060810089, "learning_rate": 1.8912384586991066e-05, "loss": 0.013, "step": 7280 }, { "epoch": 4.125636672325976, "grad_norm": 0.43052250146865845, "learning_rate": 1.8783052747471717e-05, "loss": 0.0164, "step": 7290 }, { "epoch": 4.13129598189021, "grad_norm": 0.37970268726348877, "learning_rate": 1.865406229148611e-05, "loss": 0.0232, "step": 7300 }, { "epoch": 4.136955291454442, "grad_norm": 0.46434861421585083, "learning_rate": 1.8525414629653233e-05, "loss": 0.0248, "step": 7310 }, { "epoch": 4.142614601018676, "grad_norm": 0.28842875361442566, "learning_rate": 1.8397111168843255e-05, "loss": 0.0107, "step": 7320 }, { "epoch": 4.148273910582909, "grad_norm": 0.36098814010620117, "learning_rate": 1.8269153312162323e-05, "loss": 0.0188, "step": 7330 }, { "epoch": 4.153933220147142, "grad_norm": 0.35521721839904785, "learning_rate": 1.8141542458937054e-05, "loss": 0.018, "step": 7340 }, { "epoch": 4.159592529711375, "grad_norm": 0.36936667561531067, "learning_rate": 1.8014280004699268e-05, "loss": 0.015, "step": 7350 }, { "epoch": 4.165251839275609, "grad_norm": 0.2968006134033203, "learning_rate": 1.788736734117078e-05, "loss": 0.0161, "step": 7360 }, { "epoch": 4.170911148839841, "grad_norm": 0.6336629390716553, "learning_rate": 1.7760805856248152e-05, "loss": 0.0235, "step": 7370 }, { "epoch": 4.176570458404075, "grad_norm": 0.4255976378917694, "learning_rate": 1.7634596933987518e-05, "loss": 0.017, "step": 7380 }, { "epoch": 4.182229767968308, "grad_norm": 0.3603611886501312, "learning_rate": 1.7508741954589404e-05, "loss": 0.023, "step": 7390 }, { "epoch": 4.187889077532541, "grad_norm": 0.49528074264526367, "learning_rate": 1.7383242294383717e-05, "loss": 0.0178, "step": 7400 }, { "epoch": 4.193548387096774, "grad_norm": 0.28436949849128723, "learning_rate": 1.7258099325814632e-05, "loss": 0.0181, "step": 7410 }, { "epoch": 4.1992076966610075, "grad_norm": 0.497397243976593, "learning_rate": 1.7133314417425594e-05, "loss": 0.0202, "step": 7420 }, { "epoch": 4.204867006225241, "grad_norm": 0.48627689480781555, "learning_rate": 1.7008888933844408e-05, "loss": 0.0291, "step": 7430 }, { "epoch": 4.2105263157894735, "grad_norm": 0.2106613963842392, "learning_rate": 1.6884824235768172e-05, "loss": 0.0136, "step": 7440 }, { "epoch": 4.216185625353707, "grad_norm": 0.24168846011161804, "learning_rate": 1.6761121679948592e-05, "loss": 0.0249, "step": 7450 }, { "epoch": 4.22184493491794, "grad_norm": 0.46466273069381714, "learning_rate": 1.663778261917695e-05, "loss": 0.0195, "step": 7460 }, { "epoch": 4.227504244482173, "grad_norm": 0.6745490431785583, "learning_rate": 1.651480840226952e-05, "loss": 0.0219, "step": 7470 }, { "epoch": 4.233163554046406, "grad_norm": 0.21338589489459991, "learning_rate": 1.639220037405258e-05, "loss": 0.0214, "step": 7480 }, { "epoch": 4.23882286361064, "grad_norm": 0.39897558093070984, "learning_rate": 1.6269959875347906e-05, "loss": 0.0239, "step": 7490 }, { "epoch": 4.244482173174872, "grad_norm": 0.4335300326347351, "learning_rate": 1.614808824295802e-05, "loss": 0.0185, "step": 7500 }, { "epoch": 4.250141482739106, "grad_norm": 0.3278849422931671, "learning_rate": 1.602658680965152e-05, "loss": 0.0162, "step": 7510 }, { "epoch": 4.255800792303339, "grad_norm": 0.2869303524494171, "learning_rate": 1.5905456904148686e-05, "loss": 0.0155, "step": 7520 }, { "epoch": 4.261460101867572, "grad_norm": 0.43340399861335754, "learning_rate": 1.57846998511067e-05, "loss": 0.0146, "step": 7530 }, { "epoch": 4.267119411431805, "grad_norm": 0.4681690037250519, "learning_rate": 1.566431697110538e-05, "loss": 0.0151, "step": 7540 }, { "epoch": 4.272778720996039, "grad_norm": 0.39153608679771423, "learning_rate": 1.554430958063259e-05, "loss": 0.0222, "step": 7550 }, { "epoch": 4.278438030560272, "grad_norm": 0.2339085340499878, "learning_rate": 1.5424678992069912e-05, "loss": 0.0152, "step": 7560 }, { "epoch": 4.284097340124505, "grad_norm": 0.30296626687049866, "learning_rate": 1.5305426513678362e-05, "loss": 0.0287, "step": 7570 }, { "epoch": 4.289756649688738, "grad_norm": 0.40544393658638, "learning_rate": 1.518655344958388e-05, "loss": 0.0166, "step": 7580 }, { "epoch": 4.2954159592529715, "grad_norm": 0.38694241642951965, "learning_rate": 1.5068061099763275e-05, "loss": 0.0278, "step": 7590 }, { "epoch": 4.301075268817204, "grad_norm": 0.27713772654533386, "learning_rate": 1.494995076002988e-05, "loss": 0.0256, "step": 7600 }, { "epoch": 4.3067345783814375, "grad_norm": 0.5354537963867188, "learning_rate": 1.4832223722019456e-05, "loss": 0.0268, "step": 7610 }, { "epoch": 4.312393887945671, "grad_norm": 0.27956539392471313, "learning_rate": 1.4714881273176035e-05, "loss": 0.0207, "step": 7620 }, { "epoch": 4.3180531975099035, "grad_norm": 0.3939666450023651, "learning_rate": 1.4597924696737835e-05, "loss": 0.0186, "step": 7630 }, { "epoch": 4.323712507074137, "grad_norm": 0.3555355668067932, "learning_rate": 1.4481355271723252e-05, "loss": 0.0126, "step": 7640 }, { "epoch": 4.32937181663837, "grad_norm": 0.45406582951545715, "learning_rate": 1.4365174272916809e-05, "loss": 0.0212, "step": 7650 }, { "epoch": 4.335031126202603, "grad_norm": 0.41438305377960205, "learning_rate": 1.4249382970855319e-05, "loss": 0.0231, "step": 7660 }, { "epoch": 4.340690435766836, "grad_norm": 0.1785048395395279, "learning_rate": 1.4133982631813903e-05, "loss": 0.0157, "step": 7670 }, { "epoch": 4.34634974533107, "grad_norm": 0.37624722719192505, "learning_rate": 1.4018974517792194e-05, "loss": 0.021, "step": 7680 }, { "epoch": 4.352009054895303, "grad_norm": 0.3121233582496643, "learning_rate": 1.390435988650048e-05, "loss": 0.0203, "step": 7690 }, { "epoch": 4.357668364459536, "grad_norm": 0.4113762676715851, "learning_rate": 1.3790139991346006e-05, "loss": 0.0126, "step": 7700 }, { "epoch": 4.363327674023769, "grad_norm": 0.4757934510707855, "learning_rate": 1.367631608141926e-05, "loss": 0.0165, "step": 7710 }, { "epoch": 4.368986983588003, "grad_norm": 0.4884026348590851, "learning_rate": 1.3562889401480278e-05, "loss": 0.028, "step": 7720 }, { "epoch": 4.374646293152235, "grad_norm": 0.44030284881591797, "learning_rate": 1.3449861191945074e-05, "loss": 0.0168, "step": 7730 }, { "epoch": 4.380305602716469, "grad_norm": 0.2994650602340698, "learning_rate": 1.3337232688872009e-05, "loss": 0.0135, "step": 7740 }, { "epoch": 4.385964912280702, "grad_norm": 0.2755189538002014, "learning_rate": 1.3225005123948364e-05, "loss": 0.0191, "step": 7750 }, { "epoch": 4.391624221844935, "grad_norm": 0.434190958738327, "learning_rate": 1.311317972447681e-05, "loss": 0.0177, "step": 7760 }, { "epoch": 4.397283531409168, "grad_norm": 0.4349735677242279, "learning_rate": 1.3001757713361996e-05, "loss": 0.0201, "step": 7770 }, { "epoch": 4.4029428409734015, "grad_norm": 0.30483144521713257, "learning_rate": 1.2890740309097204e-05, "loss": 0.0233, "step": 7780 }, { "epoch": 4.408602150537634, "grad_norm": 0.2641529440879822, "learning_rate": 1.2780128725750944e-05, "loss": 0.0168, "step": 7790 }, { "epoch": 4.4142614601018675, "grad_norm": 0.28446123003959656, "learning_rate": 1.266992417295379e-05, "loss": 0.0137, "step": 7800 }, { "epoch": 4.419920769666101, "grad_norm": 0.43393540382385254, "learning_rate": 1.2560127855885073e-05, "loss": 0.0202, "step": 7810 }, { "epoch": 4.425580079230334, "grad_norm": 0.39954882860183716, "learning_rate": 1.2450740975259745e-05, "loss": 0.0166, "step": 7820 }, { "epoch": 4.431239388794567, "grad_norm": 0.4561370313167572, "learning_rate": 1.234176472731517e-05, "loss": 0.0174, "step": 7830 }, { "epoch": 4.4368986983588, "grad_norm": 0.2922416031360626, "learning_rate": 1.2233200303798158e-05, "loss": 0.0119, "step": 7840 }, { "epoch": 4.442558007923034, "grad_norm": 0.3477330207824707, "learning_rate": 1.2125048891951846e-05, "loss": 0.0217, "step": 7850 }, { "epoch": 4.448217317487266, "grad_norm": 0.381676584482193, "learning_rate": 1.2017311674502745e-05, "loss": 0.0209, "step": 7860 }, { "epoch": 4.4538766270515, "grad_norm": 0.5662440657615662, "learning_rate": 1.1909989829647822e-05, "loss": 0.0275, "step": 7870 }, { "epoch": 4.459535936615733, "grad_norm": 0.22791215777397156, "learning_rate": 1.1803084531041553e-05, "loss": 0.0197, "step": 7880 }, { "epoch": 4.465195246179966, "grad_norm": 0.36128297448158264, "learning_rate": 1.1696596947783162e-05, "loss": 0.0152, "step": 7890 }, { "epoch": 4.470854555744199, "grad_norm": 0.5986441373825073, "learning_rate": 1.1590528244403803e-05, "loss": 0.0228, "step": 7900 }, { "epoch": 4.476513865308433, "grad_norm": 0.37303632497787476, "learning_rate": 1.148487958085382e-05, "loss": 0.0278, "step": 7910 }, { "epoch": 4.482173174872665, "grad_norm": 0.5574079751968384, "learning_rate": 1.1379652112490086e-05, "loss": 0.0296, "step": 7920 }, { "epoch": 4.487832484436899, "grad_norm": 0.4958466589450836, "learning_rate": 1.1274846990063315e-05, "loss": 0.022, "step": 7930 }, { "epoch": 4.493491794001132, "grad_norm": 0.6211679577827454, "learning_rate": 1.117046535970554e-05, "loss": 0.0199, "step": 7940 }, { "epoch": 4.499151103565365, "grad_norm": 0.2561228573322296, "learning_rate": 1.106650836291755e-05, "loss": 0.0204, "step": 7950 }, { "epoch": 4.504810413129598, "grad_norm": 0.4060359001159668, "learning_rate": 1.0962977136556418e-05, "loss": 0.016, "step": 7960 }, { "epoch": 4.5104697226938315, "grad_norm": 0.47181347012519836, "learning_rate": 1.0859872812823024e-05, "loss": 0.0199, "step": 7970 }, { "epoch": 4.516129032258064, "grad_norm": 0.22250010073184967, "learning_rate": 1.0757196519249747e-05, "loss": 0.0174, "step": 7980 }, { "epoch": 4.5217883418222975, "grad_norm": 0.3580704629421234, "learning_rate": 1.0654949378688077e-05, "loss": 0.0125, "step": 7990 }, { "epoch": 4.527447651386531, "grad_norm": 0.21652798354625702, "learning_rate": 1.0553132509296376e-05, "loss": 0.0194, "step": 8000 }, { "epoch": 4.533106960950764, "grad_norm": 0.3229580223560333, "learning_rate": 1.0451747024527613e-05, "loss": 0.0155, "step": 8010 }, { "epoch": 4.538766270514997, "grad_norm": 0.31920507550239563, "learning_rate": 1.0350794033117189e-05, "loss": 0.0187, "step": 8020 }, { "epoch": 4.54442558007923, "grad_norm": 0.17051509022712708, "learning_rate": 1.0250274639070856e-05, "loss": 0.0147, "step": 8030 }, { "epoch": 4.550084889643464, "grad_norm": 0.2453758269548416, "learning_rate": 1.0150189941652599e-05, "loss": 0.0127, "step": 8040 }, { "epoch": 4.555744199207696, "grad_norm": 0.27052152156829834, "learning_rate": 1.0050541035372635e-05, "loss": 0.0159, "step": 8050 }, { "epoch": 4.56140350877193, "grad_norm": 0.2720597982406616, "learning_rate": 9.951329009975458e-06, "loss": 0.018, "step": 8060 }, { "epoch": 4.567062818336163, "grad_norm": 0.20707684755325317, "learning_rate": 9.852554950427845e-06, "loss": 0.0143, "step": 8070 }, { "epoch": 4.572722127900396, "grad_norm": 0.37861260771751404, "learning_rate": 9.754219936907105e-06, "loss": 0.018, "step": 8080 }, { "epoch": 4.578381437464629, "grad_norm": 0.26799649000167847, "learning_rate": 9.656325044789194e-06, "loss": 0.0273, "step": 8090 }, { "epoch": 4.584040747028863, "grad_norm": 0.2633807361125946, "learning_rate": 9.55887134463697e-06, "loss": 0.0293, "step": 8100 }, { "epoch": 4.589700056593095, "grad_norm": 0.13268502056598663, "learning_rate": 9.461859902188475e-06, "loss": 0.0191, "step": 8110 }, { "epoch": 4.595359366157329, "grad_norm": 0.32166966795921326, "learning_rate": 9.365291778345303e-06, "loss": 0.0165, "step": 8120 }, { "epoch": 4.601018675721562, "grad_norm": 0.42073673009872437, "learning_rate": 9.269168029160991e-06, "loss": 0.0215, "step": 8130 }, { "epoch": 4.6066779852857955, "grad_norm": 0.2821817398071289, "learning_rate": 9.173489705829447e-06, "loss": 0.0296, "step": 8140 }, { "epoch": 4.612337294850028, "grad_norm": 0.42356571555137634, "learning_rate": 9.078257854673516e-06, "loss": 0.014, "step": 8150 }, { "epoch": 4.6179966044142615, "grad_norm": 0.313291072845459, "learning_rate": 8.983473517133429e-06, "loss": 0.0217, "step": 8160 }, { "epoch": 4.623655913978495, "grad_norm": 0.6416903734207153, "learning_rate": 8.889137729755537e-06, "loss": 0.0185, "step": 8170 }, { "epoch": 4.6293152235427275, "grad_norm": 0.26457569003105164, "learning_rate": 8.79525152418087e-06, "loss": 0.0193, "step": 8180 }, { "epoch": 4.634974533106961, "grad_norm": 0.1700776219367981, "learning_rate": 8.701815927133961e-06, "loss": 0.0153, "step": 8190 }, { "epoch": 4.640633842671194, "grad_norm": 0.1664831042289734, "learning_rate": 8.608831960411534e-06, "loss": 0.0151, "step": 8200 }, { "epoch": 4.646293152235427, "grad_norm": 0.13171158730983734, "learning_rate": 8.516300640871321e-06, "loss": 0.0118, "step": 8210 }, { "epoch": 4.65195246179966, "grad_norm": 0.17852510511875153, "learning_rate": 8.424222980421038e-06, "loss": 0.0153, "step": 8220 }, { "epoch": 4.657611771363894, "grad_norm": 0.3008864223957062, "learning_rate": 8.332599986007184e-06, "loss": 0.0133, "step": 8230 }, { "epoch": 4.663271080928126, "grad_norm": 0.3318842053413391, "learning_rate": 8.241432659604203e-06, "loss": 0.0175, "step": 8240 }, { "epoch": 4.66893039049236, "grad_norm": 0.16649554669857025, "learning_rate": 8.150721998203331e-06, "loss": 0.0133, "step": 8250 }, { "epoch": 4.674589700056593, "grad_norm": 0.23691275715827942, "learning_rate": 8.06046899380184e-06, "loss": 0.0087, "step": 8260 }, { "epoch": 4.680249009620827, "grad_norm": 0.2552962303161621, "learning_rate": 7.970674633392133e-06, "loss": 0.0129, "step": 8270 }, { "epoch": 4.685908319185059, "grad_norm": 0.2878318428993225, "learning_rate": 7.881339898950924e-06, "loss": 0.0166, "step": 8280 }, { "epoch": 4.691567628749293, "grad_norm": 0.21222403645515442, "learning_rate": 7.792465767428597e-06, "loss": 0.0207, "step": 8290 }, { "epoch": 4.697226938313526, "grad_norm": 0.3449961245059967, "learning_rate": 7.704053210738376e-06, "loss": 0.0188, "step": 8300 }, { "epoch": 4.702886247877759, "grad_norm": 0.12952852249145508, "learning_rate": 7.6161031957458494e-06, "loss": 0.0116, "step": 8310 }, { "epoch": 4.708545557441992, "grad_norm": 0.17324061691761017, "learning_rate": 7.5286166842582605e-06, "loss": 0.0144, "step": 8320 }, { "epoch": 4.7142048670062255, "grad_norm": 0.7605016231536865, "learning_rate": 7.4415946330140814e-06, "loss": 0.019, "step": 8330 }, { "epoch": 4.719864176570458, "grad_norm": 0.23774351179599762, "learning_rate": 7.3550379936725644e-06, "loss": 0.0102, "step": 8340 }, { "epoch": 4.7255234861346915, "grad_norm": 0.5098357796669006, "learning_rate": 7.2689477128032035e-06, "loss": 0.0187, "step": 8350 }, { "epoch": 4.731182795698925, "grad_norm": 0.48201507329940796, "learning_rate": 7.183324731875551e-06, "loss": 0.018, "step": 8360 }, { "epoch": 4.7368421052631575, "grad_norm": 0.3019934594631195, "learning_rate": 7.098169987248782e-06, "loss": 0.0113, "step": 8370 }, { "epoch": 4.742501414827391, "grad_norm": 0.495006263256073, "learning_rate": 7.013484410161553e-06, "loss": 0.0218, "step": 8380 }, { "epoch": 4.748160724391624, "grad_norm": 0.33595865964889526, "learning_rate": 6.92926892672176e-06, "loss": 0.0187, "step": 8390 }, { "epoch": 4.753820033955858, "grad_norm": 0.44678542017936707, "learning_rate": 6.845524457896446e-06, "loss": 0.0146, "step": 8400 }, { "epoch": 4.75947934352009, "grad_norm": 0.15218102931976318, "learning_rate": 6.7622519195017165e-06, "loss": 0.0187, "step": 8410 }, { "epoch": 4.765138653084324, "grad_norm": 0.22666513919830322, "learning_rate": 6.679452222192684e-06, "loss": 0.0135, "step": 8420 }, { "epoch": 4.770797962648556, "grad_norm": 0.3472532331943512, "learning_rate": 6.597126271453579e-06, "loss": 0.015, "step": 8430 }, { "epoch": 4.77645727221279, "grad_norm": 0.4426356852054596, "learning_rate": 6.51527496758782e-06, "loss": 0.0174, "step": 8440 }, { "epoch": 4.782116581777023, "grad_norm": 0.177797332406044, "learning_rate": 6.433899205708155e-06, "loss": 0.0205, "step": 8450 }, { "epoch": 4.787775891341257, "grad_norm": 0.4254482388496399, "learning_rate": 6.352999875726856e-06, "loss": 0.0147, "step": 8460 }, { "epoch": 4.793435200905489, "grad_norm": 0.3482140898704529, "learning_rate": 6.272577862346052e-06, "loss": 0.0144, "step": 8470 }, { "epoch": 4.799094510469723, "grad_norm": 0.36761054396629333, "learning_rate": 6.192634045047996e-06, "loss": 0.0127, "step": 8480 }, { "epoch": 4.804753820033956, "grad_norm": 0.17115193605422974, "learning_rate": 6.113169298085458e-06, "loss": 0.0169, "step": 8490 }, { "epoch": 4.810413129598189, "grad_norm": 0.3664368987083435, "learning_rate": 6.034184490472195e-06, "loss": 0.0199, "step": 8500 }, { "epoch": 4.816072439162422, "grad_norm": 0.25585654377937317, "learning_rate": 5.955680485973386e-06, "loss": 0.0218, "step": 8510 }, { "epoch": 4.8217317487266556, "grad_norm": 0.1754721999168396, "learning_rate": 5.877658143096265e-06, "loss": 0.0108, "step": 8520 }, { "epoch": 4.827391058290889, "grad_norm": 0.5733800530433655, "learning_rate": 5.800118315080661e-06, "loss": 0.0144, "step": 8530 }, { "epoch": 4.8330503678551215, "grad_norm": 0.23090040683746338, "learning_rate": 5.723061849889716e-06, "loss": 0.0163, "step": 8540 }, { "epoch": 4.838709677419355, "grad_norm": 0.179311603307724, "learning_rate": 5.646489590200604e-06, "loss": 0.0226, "step": 8550 }, { "epoch": 4.8443689869835875, "grad_norm": 0.3688429892063141, "learning_rate": 5.570402373395256e-06, "loss": 0.0187, "step": 8560 }, { "epoch": 4.850028296547821, "grad_norm": 0.24394282698631287, "learning_rate": 5.494801031551305e-06, "loss": 0.0364, "step": 8570 }, { "epoch": 4.855687606112054, "grad_norm": 0.3167862296104431, "learning_rate": 5.41968639143291e-06, "loss": 0.0197, "step": 8580 }, { "epoch": 4.861346915676288, "grad_norm": 0.5820156335830688, "learning_rate": 5.345059274481751e-06, "loss": 0.025, "step": 8590 }, { "epoch": 4.86700622524052, "grad_norm": 0.226038858294487, "learning_rate": 5.270920496808002e-06, "loss": 0.0093, "step": 8600 }, { "epoch": 4.872665534804754, "grad_norm": 0.12436351925134659, "learning_rate": 5.1972708691814695e-06, "loss": 0.0096, "step": 8610 }, { "epoch": 4.878324844368987, "grad_norm": 0.23296402394771576, "learning_rate": 5.124111197022674e-06, "loss": 0.0182, "step": 8620 }, { "epoch": 4.88398415393322, "grad_norm": 0.4382390081882477, "learning_rate": 5.051442280394081e-06, "loss": 0.0236, "step": 8630 }, { "epoch": 4.889643463497453, "grad_norm": 0.21248957514762878, "learning_rate": 4.979264913991322e-06, "loss": 0.013, "step": 8640 }, { "epoch": 4.895302773061687, "grad_norm": 0.37292659282684326, "learning_rate": 4.907579887134489e-06, "loss": 0.0181, "step": 8650 }, { "epoch": 4.900962082625919, "grad_norm": 0.14314575493335724, "learning_rate": 4.836387983759572e-06, "loss": 0.0185, "step": 8660 }, { "epoch": 4.906621392190153, "grad_norm": 0.25116947293281555, "learning_rate": 4.765689982409816e-06, "loss": 0.0107, "step": 8670 }, { "epoch": 4.912280701754386, "grad_norm": 0.24996954202651978, "learning_rate": 4.695486656227233e-06, "loss": 0.0128, "step": 8680 }, { "epoch": 4.917940011318619, "grad_norm": 0.19975966215133667, "learning_rate": 4.625778772944156e-06, "loss": 0.0111, "step": 8690 }, { "epoch": 4.923599320882852, "grad_norm": 0.32479751110076904, "learning_rate": 4.556567094874825e-06, "loss": 0.0158, "step": 8700 }, { "epoch": 4.929258630447086, "grad_norm": 0.14419406652450562, "learning_rate": 4.487852378907059e-06, "loss": 0.0128, "step": 8710 }, { "epoch": 4.934917940011319, "grad_norm": 0.31757551431655884, "learning_rate": 4.419635376493986e-06, "loss": 0.0224, "step": 8720 }, { "epoch": 4.9405772495755516, "grad_norm": 0.38625890016555786, "learning_rate": 4.351916833645825e-06, "loss": 0.0171, "step": 8730 }, { "epoch": 4.946236559139785, "grad_norm": 0.28305289149284363, "learning_rate": 4.284697490921691e-06, "loss": 0.0105, "step": 8740 }, { "epoch": 4.951895868704018, "grad_norm": 0.5939130187034607, "learning_rate": 4.2179780834215585e-06, "loss": 0.0206, "step": 8750 }, { "epoch": 4.957555178268251, "grad_norm": 0.45923322439193726, "learning_rate": 4.151759340778178e-06, "loss": 0.0139, "step": 8760 }, { "epoch": 4.963214487832484, "grad_norm": 0.17273946106433868, "learning_rate": 4.086041987149109e-06, "loss": 0.017, "step": 8770 }, { "epoch": 4.968873797396718, "grad_norm": 0.32904428243637085, "learning_rate": 4.020826741208811e-06, "loss": 0.0125, "step": 8780 }, { "epoch": 4.97453310696095, "grad_norm": 0.5931819081306458, "learning_rate": 3.956114316140746e-06, "loss": 0.0224, "step": 8790 }, { "epoch": 4.980192416525184, "grad_norm": 0.5407785773277283, "learning_rate": 3.891905419629643e-06, "loss": 0.0147, "step": 8800 }, { "epoch": 4.985851726089417, "grad_norm": 0.19974103569984436, "learning_rate": 3.8282007538536946e-06, "loss": 0.0141, "step": 8810 }, { "epoch": 4.99151103565365, "grad_norm": 0.33167314529418945, "learning_rate": 3.7650010154769265e-06, "loss": 0.014, "step": 8820 }, { "epoch": 4.997170345217883, "grad_norm": 0.21541465818881989, "learning_rate": 3.7023068956415608e-06, "loss": 0.0158, "step": 8830 }, { "epoch": 5.002829654782117, "grad_norm": 0.15989074110984802, "learning_rate": 3.6401190799604303e-06, "loss": 0.0123, "step": 8840 }, { "epoch": 5.00848896434635, "grad_norm": 0.38332706689834595, "learning_rate": 3.578438248509536e-06, "loss": 0.0101, "step": 8850 }, { "epoch": 5.014148273910583, "grad_norm": 0.25430190563201904, "learning_rate": 3.5172650758205583e-06, "loss": 0.0146, "step": 8860 }, { "epoch": 5.019807583474816, "grad_norm": 0.3547205626964569, "learning_rate": 3.45660023087353e-06, "loss": 0.0122, "step": 8870 }, { "epoch": 5.02546689303905, "grad_norm": 0.21467237174510956, "learning_rate": 3.3964443770894528e-06, "loss": 0.0194, "step": 8880 }, { "epoch": 5.031126202603282, "grad_norm": 0.2861272096633911, "learning_rate": 3.3367981723231245e-06, "loss": 0.0182, "step": 8890 }, { "epoch": 5.036785512167516, "grad_norm": 0.40209969878196716, "learning_rate": 3.2776622688558746e-06, "loss": 0.0146, "step": 8900 }, { "epoch": 5.042444821731749, "grad_norm": 0.38302555680274963, "learning_rate": 3.2190373133884677e-06, "loss": 0.0146, "step": 8910 }, { "epoch": 5.048104131295982, "grad_norm": 0.34767237305641174, "learning_rate": 3.1609239470340446e-06, "loss": 0.016, "step": 8920 }, { "epoch": 5.053763440860215, "grad_norm": 0.187392920255661, "learning_rate": 3.1033228053110373e-06, "loss": 0.0129, "step": 8930 }, { "epoch": 5.0594227504244484, "grad_norm": 0.3076670169830322, "learning_rate": 3.0462345181363314e-06, "loss": 0.0142, "step": 8940 }, { "epoch": 5.065082059988681, "grad_norm": 0.24371568858623505, "learning_rate": 2.9896597098182654e-06, "loss": 0.0177, "step": 8950 }, { "epoch": 5.070741369552914, "grad_norm": 0.20141568779945374, "learning_rate": 2.933598999049891e-06, "loss": 0.0088, "step": 8960 }, { "epoch": 5.076400679117148, "grad_norm": 0.18076495826244354, "learning_rate": 2.8780529989021697e-06, "loss": 0.01, "step": 8970 }, { "epoch": 5.082059988681381, "grad_norm": 0.16600769758224487, "learning_rate": 2.823022316817242e-06, "loss": 0.0113, "step": 8980 }, { "epoch": 5.087719298245614, "grad_norm": 0.26359665393829346, "learning_rate": 2.7685075546018456e-06, "loss": 0.0168, "step": 8990 }, { "epoch": 5.093378607809847, "grad_norm": 0.32241716980934143, "learning_rate": 2.7145093084206598e-06, "loss": 0.032, "step": 9000 }, { "epoch": 5.099037917374081, "grad_norm": 0.6367690563201904, "learning_rate": 2.661028168789892e-06, "loss": 0.0136, "step": 9010 }, { "epoch": 5.104697226938313, "grad_norm": 0.3206840753555298, "learning_rate": 2.6080647205706855e-06, "loss": 0.0093, "step": 9020 }, { "epoch": 5.110356536502547, "grad_norm": 0.3884199559688568, "learning_rate": 2.555619542962834e-06, "loss": 0.0195, "step": 9030 }, { "epoch": 5.11601584606678, "grad_norm": 0.20583972334861755, "learning_rate": 2.503693209498409e-06, "loss": 0.0154, "step": 9040 }, { "epoch": 5.121675155631013, "grad_norm": 0.3200293481349945, "learning_rate": 2.452286288035449e-06, "loss": 0.0163, "step": 9050 }, { "epoch": 5.127334465195246, "grad_norm": 0.33171623945236206, "learning_rate": 2.4013993407518363e-06, "loss": 0.0187, "step": 9060 }, { "epoch": 5.13299377475948, "grad_norm": 0.19206714630126953, "learning_rate": 2.351032924139063e-06, "loss": 0.0107, "step": 9070 }, { "epoch": 5.138653084323712, "grad_norm": 0.2751140296459198, "learning_rate": 2.30118758899619e-06, "loss": 0.0099, "step": 9080 }, { "epoch": 5.144312393887946, "grad_norm": 0.24944934248924255, "learning_rate": 2.2518638804238157e-06, "loss": 0.0249, "step": 9090 }, { "epoch": 5.149971703452179, "grad_norm": 0.5298340916633606, "learning_rate": 2.203062337818118e-06, "loss": 0.0208, "step": 9100 }, { "epoch": 5.155631013016412, "grad_norm": 0.20584413409233093, "learning_rate": 2.1547834948649483e-06, "loss": 0.0105, "step": 9110 }, { "epoch": 5.161290322580645, "grad_norm": 0.2957272529602051, "learning_rate": 2.1070278795340017e-06, "loss": 0.017, "step": 9120 }, { "epoch": 5.1669496321448785, "grad_norm": 0.17460356652736664, "learning_rate": 2.059796014073029e-06, "loss": 0.0233, "step": 9130 }, { "epoch": 5.172608941709112, "grad_norm": 0.3751950263977051, "learning_rate": 2.01308841500214e-06, "loss": 0.01, "step": 9140 }, { "epoch": 5.1782682512733444, "grad_norm": 0.16106566786766052, "learning_rate": 1.9669055931081704e-06, "loss": 0.0163, "step": 9150 }, { "epoch": 5.183927560837578, "grad_norm": 0.3512350022792816, "learning_rate": 1.9212480534390507e-06, "loss": 0.0115, "step": 9160 }, { "epoch": 5.189586870401811, "grad_norm": 0.09471239149570465, "learning_rate": 1.8761162952983246e-06, "loss": 0.0129, "step": 9170 }, { "epoch": 5.195246179966044, "grad_norm": 0.3261120915412903, "learning_rate": 1.8315108122396618e-06, "loss": 0.0141, "step": 9180 }, { "epoch": 5.200905489530277, "grad_norm": 0.14130228757858276, "learning_rate": 1.787432092061475e-06, "loss": 0.0105, "step": 9190 }, { "epoch": 5.206564799094511, "grad_norm": 0.33845049142837524, "learning_rate": 1.743880616801602e-06, "loss": 0.014, "step": 9200 }, { "epoch": 5.212224108658743, "grad_norm": 0.5717782974243164, "learning_rate": 1.7008568627319865e-06, "loss": 0.0163, "step": 9210 }, { "epoch": 5.217883418222977, "grad_norm": 0.43369486927986145, "learning_rate": 1.6583613003535226e-06, "loss": 0.0229, "step": 9220 }, { "epoch": 5.22354272778721, "grad_norm": 0.24910669028759003, "learning_rate": 1.6163943943908522e-06, "loss": 0.0152, "step": 9230 }, { "epoch": 5.229202037351443, "grad_norm": 0.25324907898902893, "learning_rate": 1.5749566037873476e-06, "loss": 0.0125, "step": 9240 }, { "epoch": 5.234861346915676, "grad_norm": 0.1568821221590042, "learning_rate": 1.5340483817000428e-06, "loss": 0.0242, "step": 9250 }, { "epoch": 5.24052065647991, "grad_norm": 0.31048205494880676, "learning_rate": 1.4936701754947101e-06, "loss": 0.0159, "step": 9260 }, { "epoch": 5.246179966044142, "grad_norm": 0.5406427979469299, "learning_rate": 1.4538224267409361e-06, "loss": 0.0124, "step": 9270 }, { "epoch": 5.251839275608376, "grad_norm": 0.4390304386615753, "learning_rate": 1.414505571207314e-06, "loss": 0.0129, "step": 9280 }, { "epoch": 5.257498585172609, "grad_norm": 0.19887875020503998, "learning_rate": 1.3757200388566816e-06, "loss": 0.02, "step": 9290 }, { "epoch": 5.2631578947368425, "grad_norm": 0.1655147671699524, "learning_rate": 1.3374662538414074e-06, "loss": 0.0154, "step": 9300 }, { "epoch": 5.268817204301075, "grad_norm": 0.2915429472923279, "learning_rate": 1.2997446344987617e-06, "loss": 0.0239, "step": 9310 }, { "epoch": 5.2744765138653085, "grad_norm": 0.384115993976593, "learning_rate": 1.262555593346315e-06, "loss": 0.0182, "step": 9320 }, { "epoch": 5.280135823429542, "grad_norm": 0.19533221423625946, "learning_rate": 1.2258995370774685e-06, "loss": 0.0271, "step": 9330 }, { "epoch": 5.2857951329937745, "grad_norm": 0.36194220185279846, "learning_rate": 1.1897768665569798e-06, "loss": 0.0177, "step": 9340 }, { "epoch": 5.291454442558008, "grad_norm": 0.2928744852542877, "learning_rate": 1.1541879768165954e-06, "loss": 0.0097, "step": 9350 }, { "epoch": 5.297113752122241, "grad_norm": 0.41840288043022156, "learning_rate": 1.1191332570507085e-06, "loss": 0.0134, "step": 9360 }, { "epoch": 5.302773061686474, "grad_norm": 0.5431681871414185, "learning_rate": 1.0846130906121132e-06, "loss": 0.0227, "step": 9370 }, { "epoch": 5.308432371250707, "grad_norm": 0.27649563550949097, "learning_rate": 1.0506278550078131e-06, "loss": 0.0139, "step": 9380 }, { "epoch": 5.314091680814941, "grad_norm": 0.27125483751296997, "learning_rate": 1.0171779218949185e-06, "loss": 0.0144, "step": 9390 }, { "epoch": 5.319750990379173, "grad_norm": 0.27272504568099976, "learning_rate": 9.842636570765174e-07, "loss": 0.0173, "step": 9400 }, { "epoch": 5.325410299943407, "grad_norm": 0.3312033414840698, "learning_rate": 9.518854204977612e-07, "loss": 0.0273, "step": 9410 }, { "epoch": 5.33106960950764, "grad_norm": 0.29762130975723267, "learning_rate": 9.200435662418349e-07, "loss": 0.0111, "step": 9420 }, { "epoch": 5.336728919071874, "grad_norm": 0.2990401089191437, "learning_rate": 8.887384425261658e-07, "loss": 0.0159, "step": 9430 }, { "epoch": 5.342388228636106, "grad_norm": 0.4782230257987976, "learning_rate": 8.579703916985648e-07, "loss": 0.0118, "step": 9440 }, { "epoch": 5.34804753820034, "grad_norm": 0.21451210975646973, "learning_rate": 8.277397502335194e-07, "loss": 0.0146, "step": 9450 }, { "epoch": 5.353706847764573, "grad_norm": 0.3441447913646698, "learning_rate": 7.980468487284675e-07, "loss": 0.0082, "step": 9460 }, { "epoch": 5.359366157328806, "grad_norm": 0.35575538873672485, "learning_rate": 7.688920119002297e-07, "loss": 0.0168, "step": 9470 }, { "epoch": 5.365025466893039, "grad_norm": 0.27196231484413147, "learning_rate": 7.402755585814269e-07, "loss": 0.0205, "step": 9480 }, { "epoch": 5.3706847764572725, "grad_norm": 0.23656480014324188, "learning_rate": 7.121978017170073e-07, "loss": 0.0081, "step": 9490 }, { "epoch": 5.376344086021505, "grad_norm": 0.10927730053663254, "learning_rate": 6.846590483608306e-07, "loss": 0.0115, "step": 9500 }, { "epoch": 5.3820033955857385, "grad_norm": 0.43316495418548584, "learning_rate": 6.576595996722834e-07, "loss": 0.0136, "step": 9510 }, { "epoch": 5.387662705149972, "grad_norm": 0.24791935086250305, "learning_rate": 6.311997509130141e-07, "loss": 0.0077, "step": 9520 }, { "epoch": 5.3933220147142045, "grad_norm": 0.17370861768722534, "learning_rate": 6.052797914436803e-07, "loss": 0.0218, "step": 9530 }, { "epoch": 5.398981324278438, "grad_norm": 0.561249852180481, "learning_rate": 5.799000047208181e-07, "loss": 0.0149, "step": 9540 }, { "epoch": 5.404640633842671, "grad_norm": 0.1811702996492386, "learning_rate": 5.550606682937054e-07, "loss": 0.025, "step": 9550 }, { "epoch": 5.410299943406905, "grad_norm": 0.4653280973434448, "learning_rate": 5.307620538013481e-07, "loss": 0.0163, "step": 9560 }, { "epoch": 5.415959252971137, "grad_norm": 0.30958282947540283, "learning_rate": 5.070044269694874e-07, "loss": 0.0155, "step": 9570 }, { "epoch": 5.421618562535371, "grad_norm": 0.3639081120491028, "learning_rate": 4.837880476077417e-07, "loss": 0.0216, "step": 9580 }, { "epoch": 5.427277872099604, "grad_norm": 0.16603532433509827, "learning_rate": 4.6111316960670835e-07, "loss": 0.0133, "step": 9590 }, { "epoch": 5.432937181663837, "grad_norm": 0.2216721922159195, "learning_rate": 4.389800409352218e-07, "loss": 0.016, "step": 9600 }, { "epoch": 5.43859649122807, "grad_norm": 0.27220794558525085, "learning_rate": 4.173889036376277e-07, "loss": 0.0154, "step": 9610 }, { "epoch": 5.444255800792304, "grad_norm": 0.191203311085701, "learning_rate": 3.963399938311463e-07, "loss": 0.0144, "step": 9620 }, { "epoch": 5.449915110356536, "grad_norm": 0.17640888690948486, "learning_rate": 3.7583354170328545e-07, "loss": 0.0179, "step": 9630 }, { "epoch": 5.45557441992077, "grad_norm": 0.09340283274650574, "learning_rate": 3.558697715093207e-07, "loss": 0.012, "step": 9640 }, { "epoch": 5.461233729485003, "grad_norm": 0.1469435691833496, "learning_rate": 3.3644890156983576e-07, "loss": 0.0173, "step": 9650 }, { "epoch": 5.466893039049236, "grad_norm": 0.24418553709983826, "learning_rate": 3.175711442683638e-07, "loss": 0.015, "step": 9660 }, { "epoch": 5.472552348613469, "grad_norm": 0.19976800680160522, "learning_rate": 2.9923670604902197e-07, "loss": 0.0095, "step": 9670 }, { "epoch": 5.4782116581777025, "grad_norm": 0.18360495567321777, "learning_rate": 2.814457874143028e-07, "loss": 0.0151, "step": 9680 }, { "epoch": 5.483870967741936, "grad_norm": 0.44020748138427734, "learning_rate": 2.641985829228366e-07, "loss": 0.012, "step": 9690 }, { "epoch": 5.4895302773061685, "grad_norm": 0.21414652466773987, "learning_rate": 2.474952811872877e-07, "loss": 0.0093, "step": 9700 }, { "epoch": 5.495189586870402, "grad_norm": 0.29714521765708923, "learning_rate": 2.3133606487228397e-07, "loss": 0.0157, "step": 9710 }, { "epoch": 5.5008488964346345, "grad_norm": 0.11837577819824219, "learning_rate": 2.157211106924295e-07, "loss": 0.0073, "step": 9720 }, { "epoch": 5.506508205998868, "grad_norm": 0.1690998524427414, "learning_rate": 2.006505894103672e-07, "loss": 0.0169, "step": 9730 }, { "epoch": 5.512167515563101, "grad_norm": 0.14515431225299835, "learning_rate": 1.8612466583489696e-07, "loss": 0.0161, "step": 9740 }, { "epoch": 5.517826825127335, "grad_norm": 0.4104876220226288, "learning_rate": 1.7214349881918834e-07, "loss": 0.0221, "step": 9750 }, { "epoch": 5.523486134691567, "grad_norm": 0.14933814108371735, "learning_rate": 1.5870724125904845e-07, "loss": 0.0093, "step": 9760 }, { "epoch": 5.529145444255801, "grad_norm": 0.3486264646053314, "learning_rate": 1.4581604009124006e-07, "loss": 0.0169, "step": 9770 }, { "epoch": 5.534804753820034, "grad_norm": 0.4453631341457367, "learning_rate": 1.334700362918717e-07, "loss": 0.0184, "step": 9780 }, { "epoch": 5.540464063384267, "grad_norm": 0.22598163783550262, "learning_rate": 1.2166936487486015e-07, "loss": 0.0131, "step": 9790 }, { "epoch": 5.5461233729485, "grad_norm": 0.21766339242458344, "learning_rate": 1.1041415489045914e-07, "loss": 0.0124, "step": 9800 }, { "epoch": 5.551782682512734, "grad_norm": 0.26409411430358887, "learning_rate": 9.970452942384412e-08, "loss": 0.0109, "step": 9810 }, { "epoch": 5.557441992076967, "grad_norm": 0.3744887113571167, "learning_rate": 8.954060559375754e-08, "loss": 0.0242, "step": 9820 }, { "epoch": 5.5631013016412, "grad_norm": 0.23755554854869843, "learning_rate": 7.99224945512489e-08, "loss": 0.0148, "step": 9830 }, { "epoch": 5.568760611205433, "grad_norm": 0.6062955260276794, "learning_rate": 7.085030147843675e-08, "loss": 0.02, "step": 9840 }, { "epoch": 5.574419920769666, "grad_norm": 0.18671253323554993, "learning_rate": 6.232412558736523e-08, "loss": 0.0189, "step": 9850 }, { "epoch": 5.580079230333899, "grad_norm": 0.18308652937412262, "learning_rate": 5.434406011893822e-08, "loss": 0.0132, "step": 9860 }, { "epoch": 5.5857385398981325, "grad_norm": 0.21284526586532593, "learning_rate": 4.6910192341864664e-08, "loss": 0.0094, "step": 9870 }, { "epoch": 5.591397849462366, "grad_norm": 0.3942609131336212, "learning_rate": 4.0022603551737035e-08, "loss": 0.0128, "step": 9880 }, { "epoch": 5.5970571590265985, "grad_norm": 0.14613506197929382, "learning_rate": 3.3681369070120985e-08, "loss": 0.0106, "step": 9890 }, { "epoch": 5.602716468590832, "grad_norm": 0.22569288313388824, "learning_rate": 2.7886558243744866e-08, "loss": 0.0161, "step": 9900 }, { "epoch": 5.608375778155065, "grad_norm": 0.2921515107154846, "learning_rate": 2.2638234443722596e-08, "loss": 0.0161, "step": 9910 }, { "epoch": 5.614035087719298, "grad_norm": 0.3103187084197998, "learning_rate": 1.7936455064887504e-08, "loss": 0.0131, "step": 9920 }, { "epoch": 5.619694397283531, "grad_norm": 0.20665791630744934, "learning_rate": 1.378127152514841e-08, "loss": 0.0182, "step": 9930 }, { "epoch": 5.625353706847765, "grad_norm": 0.2384548783302307, "learning_rate": 1.0172729264917857e-08, "loss": 0.0114, "step": 9940 }, { "epoch": 5.631013016411997, "grad_norm": 0.2406599372625351, "learning_rate": 7.1108677466458215e-09, "loss": 0.0138, "step": 9950 }, { "epoch": 5.636672325976231, "grad_norm": 0.13295477628707886, "learning_rate": 4.595720454353414e-09, "loss": 0.0105, "step": 9960 }, { "epoch": 5.642331635540464, "grad_norm": 0.2226768136024475, "learning_rate": 2.627314893294264e-09, "loss": 0.0097, "step": 9970 }, { "epoch": 5.647990945104697, "grad_norm": 0.6699599027633667, "learning_rate": 1.2056725896270048e-09, "loss": 0.023, "step": 9980 }, { "epoch": 5.65365025466893, "grad_norm": 0.2477726936340332, "learning_rate": 3.308090902098826e-10, "loss": 0.0181, "step": 9990 }, { "epoch": 5.659309564233164, "grad_norm": 0.2732313871383667, "learning_rate": 2.7339624120159555e-12, "loss": 0.014, "step": 10000 }, { "epoch": 5.659309564233164, "step": 10000, "total_flos": 0.0, "train_loss": 0.04171483232602477, "train_runtime": 7285.8266, "train_samples_per_second": 10.98, "train_steps_per_second": 1.373 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 6, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }