{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 2.609462261199951, "learning_rate": 1.8e-06, "loss": 1.1485, "step": 10 }, { "grad_norm": 2.7307465076446533, "learning_rate": 3.8e-06, "loss": 1.1216, "step": 20 }, { "grad_norm": 0.778983473777771, "learning_rate": 5.8e-06, "loss": 1.0928, "step": 30 }, { "grad_norm": 0.47953301668167114, "learning_rate": 7.8e-06, "loss": 1.0813, "step": 40 }, { "grad_norm": 0.6872252821922302, "learning_rate": 9.800000000000001e-06, "loss": 1.0678, "step": 50 }, { "grad_norm": 0.43077972531318665, "learning_rate": 1.18e-05, "loss": 1.0591, "step": 60 }, { "grad_norm": 0.7190169095993042, "learning_rate": 1.3800000000000002e-05, "loss": 1.068, "step": 70 }, { "grad_norm": 0.7817668914794922, "learning_rate": 1.58e-05, "loss": 1.0537, "step": 80 }, { "grad_norm": 0.3384284973144531, "learning_rate": 1.78e-05, "loss": 1.0556, "step": 90 }, { "grad_norm": 0.5667229294776917, "learning_rate": 1.9800000000000004e-05, "loss": 1.0337, "step": 100 }, { "grad_norm": 0.6704097986221313, "learning_rate": 2.18e-05, "loss": 1.0298, "step": 110 }, { "grad_norm": 0.6423372030258179, "learning_rate": 2.38e-05, "loss": 1.029, "step": 120 }, { "grad_norm": 0.412604421377182, "learning_rate": 2.58e-05, "loss": 1.0323, "step": 130 }, { "grad_norm": 0.6234939098358154, "learning_rate": 2.7800000000000005e-05, "loss": 1.0164, "step": 140 }, { "grad_norm": 1.8988213539123535, "learning_rate": 2.98e-05, "loss": 0.9607, "step": 150 }, { "grad_norm": 1.5028096437454224, "learning_rate": 3.18e-05, "loss": 0.912, "step": 160 }, { "grad_norm": 2.726884365081787, "learning_rate": 3.38e-05, "loss": 0.7625, "step": 170 }, { "grad_norm": 2.3534679412841797, "learning_rate": 3.58e-05, "loss": 0.6689, "step": 180 }, { "grad_norm": 1.6922287940979004, "learning_rate": 3.7800000000000004e-05, "loss": 0.517, "step": 190 }, { "grad_norm": 3.0068304538726807, "learning_rate": 3.9800000000000005e-05, "loss": 0.4073, "step": 200 }, { "grad_norm": 1.8579225540161133, "learning_rate": 4.18e-05, "loss": 0.3039, "step": 210 }, { "grad_norm": 2.4592034816741943, "learning_rate": 4.38e-05, "loss": 0.2385, "step": 220 }, { "grad_norm": 1.5148944854736328, "learning_rate": 4.58e-05, "loss": 0.1944, "step": 230 }, { "grad_norm": 1.4749828577041626, "learning_rate": 4.78e-05, "loss": 0.1727, "step": 240 }, { "grad_norm": 1.717528223991394, "learning_rate": 4.9800000000000004e-05, "loss": 0.1581, "step": 250 }, { "grad_norm": 1.616878867149353, "learning_rate": 5.1800000000000005e-05, "loss": 0.1484, "step": 260 }, { "grad_norm": 1.7140049934387207, "learning_rate": 5.380000000000001e-05, "loss": 0.16, "step": 270 }, { "grad_norm": 1.0581860542297363, "learning_rate": 5.580000000000001e-05, "loss": 0.1421, "step": 280 }, { "grad_norm": 1.33902907371521, "learning_rate": 5.7799999999999995e-05, "loss": 0.1492, "step": 290 }, { "grad_norm": 1.3512868881225586, "learning_rate": 5.9800000000000003e-05, "loss": 0.1352, "step": 300 }, { "grad_norm": 0.9916190505027771, "learning_rate": 6.18e-05, "loss": 0.1428, "step": 310 }, { "grad_norm": 1.178504228591919, "learning_rate": 6.38e-05, "loss": 0.1259, "step": 320 }, { "grad_norm": 1.197808861732483, "learning_rate": 6.58e-05, "loss": 0.1201, "step": 330 }, { "grad_norm": 1.2348016500473022, "learning_rate": 6.780000000000001e-05, "loss": 0.118, "step": 340 }, { "grad_norm": 0.6368618607521057, "learning_rate": 6.98e-05, "loss": 0.102, "step": 350 }, { "grad_norm": 1.1706050634384155, "learning_rate": 7.18e-05, "loss": 0.1097, "step": 360 }, { "grad_norm": 1.0973981618881226, "learning_rate": 7.38e-05, "loss": 0.1201, "step": 370 }, { "grad_norm": 0.8612812757492065, "learning_rate": 7.58e-05, "loss": 0.1078, "step": 380 }, { "grad_norm": 0.9290992617607117, "learning_rate": 7.780000000000001e-05, "loss": 0.1082, "step": 390 }, { "grad_norm": 0.8851697444915771, "learning_rate": 7.98e-05, "loss": 0.1033, "step": 400 }, { "grad_norm": 0.5730687975883484, "learning_rate": 8.18e-05, "loss": 0.1086, "step": 410 }, { "grad_norm": 0.6562680006027222, "learning_rate": 8.38e-05, "loss": 0.1036, "step": 420 }, { "grad_norm": 0.9072203636169434, "learning_rate": 8.58e-05, "loss": 0.1017, "step": 430 }, { "grad_norm": 0.898476779460907, "learning_rate": 8.78e-05, "loss": 0.1018, "step": 440 }, { "grad_norm": 0.6041813492774963, "learning_rate": 8.98e-05, "loss": 0.0925, "step": 450 }, { "grad_norm": 0.6205312013626099, "learning_rate": 9.180000000000001e-05, "loss": 0.0939, "step": 460 }, { "grad_norm": 0.7066349387168884, "learning_rate": 9.38e-05, "loss": 0.1096, "step": 470 }, { "grad_norm": 0.5705448985099792, "learning_rate": 9.58e-05, "loss": 0.1075, "step": 480 }, { "grad_norm": 0.8207343220710754, "learning_rate": 9.78e-05, "loss": 0.0997, "step": 490 }, { "grad_norm": 0.5730529427528381, "learning_rate": 9.98e-05, "loss": 0.0931, "step": 500 }, { "grad_norm": 0.7002720236778259, "learning_rate": 9.9999778549206e-05, "loss": 0.1027, "step": 510 }, { "grad_norm": 1.052080750465393, "learning_rate": 9.999901304280685e-05, "loss": 0.0902, "step": 520 }, { "grad_norm": 0.5516476631164551, "learning_rate": 9.999770075521164e-05, "loss": 0.0954, "step": 530 }, { "grad_norm": 0.7560146450996399, "learning_rate": 9.99958417007713e-05, "loss": 0.093, "step": 540 }, { "grad_norm": 0.6026809215545654, "learning_rate": 9.999343589981615e-05, "loss": 0.0862, "step": 550 }, { "grad_norm": 0.4387367367744446, "learning_rate": 9.999048337865568e-05, "loss": 0.0815, "step": 560 }, { "grad_norm": 0.5430311560630798, "learning_rate": 9.998698416957815e-05, "loss": 0.0823, "step": 570 }, { "grad_norm": 0.5647290945053101, "learning_rate": 9.998293831085037e-05, "loss": 0.0841, "step": 580 }, { "grad_norm": 0.6483649015426636, "learning_rate": 9.997834584671719e-05, "loss": 0.0845, "step": 590 }, { "grad_norm": 0.4731362462043762, "learning_rate": 9.997320682740107e-05, "loss": 0.0852, "step": 600 }, { "grad_norm": 0.6121919751167297, "learning_rate": 9.996752130910149e-05, "loss": 0.086, "step": 610 }, { "grad_norm": 0.5331981778144836, "learning_rate": 9.99612893539944e-05, "loss": 0.077, "step": 620 }, { "grad_norm": 0.4118482172489166, "learning_rate": 9.995451103023144e-05, "loss": 0.0687, "step": 630 }, { "grad_norm": 0.6658375263214111, "learning_rate": 9.994718641193928e-05, "loss": 0.0669, "step": 640 }, { "grad_norm": 0.41122791171073914, "learning_rate": 9.993931557921874e-05, "loss": 0.0717, "step": 650 }, { "grad_norm": 0.6048871874809265, "learning_rate": 9.993089861814402e-05, "loss": 0.0676, "step": 660 }, { "grad_norm": 0.44824308156967163, "learning_rate": 9.992193562076166e-05, "loss": 0.0803, "step": 670 }, { "grad_norm": 0.6873279213905334, "learning_rate": 9.991242668508954e-05, "loss": 0.0795, "step": 680 }, { "grad_norm": 0.6303627490997314, "learning_rate": 9.990237191511587e-05, "loss": 0.0775, "step": 690 }, { "grad_norm": 0.4801505506038666, "learning_rate": 9.989177142079802e-05, "loss": 0.076, "step": 700 }, { "grad_norm": 0.4206171929836273, "learning_rate": 9.988062531806126e-05, "loss": 0.0884, "step": 710 }, { "grad_norm": 0.5291579365730286, "learning_rate": 9.986893372879762e-05, "loss": 0.0695, "step": 720 }, { "grad_norm": 0.7211747169494629, "learning_rate": 9.985669678086443e-05, "loss": 0.0774, "step": 730 }, { "grad_norm": 0.5550020337104797, "learning_rate": 9.984391460808298e-05, "loss": 0.0607, "step": 740 }, { "grad_norm": 0.620403528213501, "learning_rate": 9.983058735023709e-05, "loss": 0.0694, "step": 750 }, { "grad_norm": 0.5498299598693848, "learning_rate": 9.98167151530715e-05, "loss": 0.061, "step": 760 }, { "grad_norm": 0.5352259278297424, "learning_rate": 9.980229816829034e-05, "loss": 0.0681, "step": 770 }, { "grad_norm": 0.3503713309764862, "learning_rate": 9.978733655355544e-05, "loss": 0.0671, "step": 780 }, { "grad_norm": 0.4149607717990875, "learning_rate": 9.977183047248464e-05, "loss": 0.0632, "step": 790 }, { "grad_norm": 0.45517802238464355, "learning_rate": 9.975578009464992e-05, "loss": 0.069, "step": 800 }, { "grad_norm": 0.2692798972129822, "learning_rate": 9.97391855955757e-05, "loss": 0.0749, "step": 810 }, { "grad_norm": 0.3512372672557831, "learning_rate": 9.972204715673669e-05, "loss": 0.0671, "step": 820 }, { "grad_norm": 0.41873735189437866, "learning_rate": 9.970436496555617e-05, "loss": 0.0696, "step": 830 }, { "grad_norm": 0.3570493757724762, "learning_rate": 9.968613921540373e-05, "loss": 0.0626, "step": 840 }, { "grad_norm": 0.44994211196899414, "learning_rate": 9.966737010559326e-05, "loss": 0.0717, "step": 850 }, { "grad_norm": 0.40255212783813477, "learning_rate": 9.964805784138072e-05, "loss": 0.0651, "step": 860 }, { "grad_norm": 0.5730339884757996, "learning_rate": 9.962820263396195e-05, "loss": 0.0703, "step": 870 }, { "grad_norm": 0.36409586668014526, "learning_rate": 9.960780470047033e-05, "loss": 0.0636, "step": 880 }, { "grad_norm": 0.3740072548389435, "learning_rate": 9.958686426397437e-05, "loss": 0.0615, "step": 890 }, { "grad_norm": 0.5327548384666443, "learning_rate": 9.956538155347534e-05, "loss": 0.0699, "step": 900 }, { "grad_norm": 0.45331472158432007, "learning_rate": 9.95433568039047e-05, "loss": 0.0584, "step": 910 }, { "grad_norm": 0.3615391254425049, "learning_rate": 9.952079025612162e-05, "loss": 0.0644, "step": 920 }, { "grad_norm": 0.44736239314079285, "learning_rate": 9.949768215691022e-05, "loss": 0.0556, "step": 930 }, { "grad_norm": 0.4774295687675476, "learning_rate": 9.9474032758977e-05, "loss": 0.055, "step": 940 }, { "grad_norm": 0.5144764184951782, "learning_rate": 9.944984232094794e-05, "loss": 0.0706, "step": 950 }, { "grad_norm": 0.4668825566768646, "learning_rate": 9.942511110736584e-05, "loss": 0.053, "step": 960 }, { "grad_norm": 0.487498015165329, "learning_rate": 9.939983938868726e-05, "loss": 0.0575, "step": 970 }, { "grad_norm": 0.4372357428073883, "learning_rate": 9.93740274412797e-05, "loss": 0.0611, "step": 980 }, { "grad_norm": 0.46888917684555054, "learning_rate": 9.934767554741846e-05, "loss": 0.0645, "step": 990 }, { "grad_norm": 0.3818019926548004, "learning_rate": 9.932078399528361e-05, "loss": 0.0588, "step": 1000 }, { "grad_norm": 0.417123019695282, "learning_rate": 9.929335307895689e-05, "loss": 0.0577, "step": 1010 }, { "grad_norm": 0.36975085735321045, "learning_rate": 9.926538309841839e-05, "loss": 0.0547, "step": 1020 }, { "grad_norm": 0.4730314612388611, "learning_rate": 9.923687435954334e-05, "loss": 0.0605, "step": 1030 }, { "grad_norm": 0.47617942094802856, "learning_rate": 9.920782717409873e-05, "loss": 0.0589, "step": 1040 }, { "grad_norm": 0.35871610045433044, "learning_rate": 9.917824185973994e-05, "loss": 0.0508, "step": 1050 }, { "grad_norm": 0.4433595538139343, "learning_rate": 9.914811874000723e-05, "loss": 0.0681, "step": 1060 }, { "grad_norm": 0.41465121507644653, "learning_rate": 9.911745814432218e-05, "loss": 0.0589, "step": 1070 }, { "grad_norm": 0.4832235276699066, "learning_rate": 9.90862604079842e-05, "loss": 0.0664, "step": 1080 }, { "grad_norm": 0.5261781811714172, "learning_rate": 9.90545258721667e-05, "loss": 0.051, "step": 1090 }, { "grad_norm": 0.5406185388565063, "learning_rate": 9.90222548839135e-05, "loss": 0.0593, "step": 1100 }, { "grad_norm": 0.45040756464004517, "learning_rate": 9.898944779613495e-05, "loss": 0.0605, "step": 1110 }, { "grad_norm": 0.30359071493148804, "learning_rate": 9.89561049676041e-05, "loss": 0.0541, "step": 1120 }, { "grad_norm": 0.35110658407211304, "learning_rate": 9.89222267629528e-05, "loss": 0.06, "step": 1130 }, { "grad_norm": 0.4363695979118347, "learning_rate": 9.888781355266763e-05, "loss": 0.0599, "step": 1140 }, { "grad_norm": 0.22849243879318237, "learning_rate": 9.885286571308598e-05, "loss": 0.0569, "step": 1150 }, { "grad_norm": 0.36995771527290344, "learning_rate": 9.881738362639182e-05, "loss": 0.0529, "step": 1160 }, { "grad_norm": 0.35861286520957947, "learning_rate": 9.878136768061154e-05, "loss": 0.0562, "step": 1170 }, { "grad_norm": 0.4145205020904541, "learning_rate": 9.874481826960979e-05, "loss": 0.0541, "step": 1180 }, { "grad_norm": 0.30212199687957764, "learning_rate": 9.870773579308503e-05, "loss": 0.055, "step": 1190 }, { "grad_norm": 0.2626578211784363, "learning_rate": 9.867012065656533e-05, "loss": 0.0617, "step": 1200 }, { "grad_norm": 0.3705542981624603, "learning_rate": 9.863197327140376e-05, "loss": 0.0563, "step": 1210 }, { "grad_norm": 0.3843325674533844, "learning_rate": 9.859329405477403e-05, "loss": 0.0507, "step": 1220 }, { "grad_norm": 0.320382684469223, "learning_rate": 9.855408342966585e-05, "loss": 0.0565, "step": 1230 }, { "grad_norm": 0.2753898799419403, "learning_rate": 9.851434182488033e-05, "loss": 0.051, "step": 1240 }, { "grad_norm": 0.37669724225997925, "learning_rate": 9.84740696750253e-05, "loss": 0.0596, "step": 1250 }, { "grad_norm": 0.39258167147636414, "learning_rate": 9.843326742051055e-05, "loss": 0.0529, "step": 1260 }, { "grad_norm": 0.3084080219268799, "learning_rate": 9.839193550754297e-05, "loss": 0.0562, "step": 1270 }, { "grad_norm": 0.5414196848869324, "learning_rate": 9.835007438812177e-05, "loss": 0.0556, "step": 1280 }, { "grad_norm": 0.3001902997493744, "learning_rate": 9.830768452003341e-05, "loss": 0.0586, "step": 1290 }, { "grad_norm": 0.35467296838760376, "learning_rate": 9.826476636684671e-05, "loss": 0.056, "step": 1300 }, { "grad_norm": 0.37087416648864746, "learning_rate": 9.822132039790773e-05, "loss": 0.0569, "step": 1310 }, { "grad_norm": 0.4359745979309082, "learning_rate": 9.817734708833461e-05, "loss": 0.0596, "step": 1320 }, { "grad_norm": 0.32818081974983215, "learning_rate": 9.813284691901243e-05, "loss": 0.0489, "step": 1330 }, { "grad_norm": 0.336981862783432, "learning_rate": 9.808782037658792e-05, "loss": 0.0551, "step": 1340 }, { "grad_norm": 0.289148211479187, "learning_rate": 9.804226795346411e-05, "loss": 0.0505, "step": 1350 }, { "grad_norm": 0.49363142251968384, "learning_rate": 9.799619014779503e-05, "loss": 0.0549, "step": 1360 }, { "grad_norm": 0.399759978055954, "learning_rate": 9.794958746348013e-05, "loss": 0.0447, "step": 1370 }, { "grad_norm": 0.3471735417842865, "learning_rate": 9.790246041015896e-05, "loss": 0.0463, "step": 1380 }, { "grad_norm": 0.41345447301864624, "learning_rate": 9.785480950320538e-05, "loss": 0.0456, "step": 1390 }, { "grad_norm": 0.40723294019699097, "learning_rate": 9.78066352637221e-05, "loss": 0.0551, "step": 1400 }, { "grad_norm": 0.32937395572662354, "learning_rate": 9.775793821853488e-05, "loss": 0.0436, "step": 1410 }, { "grad_norm": 0.43093496561050415, "learning_rate": 9.77087189001868e-05, "loss": 0.053, "step": 1420 }, { "grad_norm": 0.28276047110557556, "learning_rate": 9.765897784693243e-05, "loss": 0.0479, "step": 1430 }, { "grad_norm": 0.25817418098449707, "learning_rate": 9.760871560273197e-05, "loss": 0.0488, "step": 1440 }, { "grad_norm": 0.3129417300224304, "learning_rate": 9.755793271724526e-05, "loss": 0.056, "step": 1450 }, { "grad_norm": 0.31618309020996094, "learning_rate": 9.750662974582584e-05, "loss": 0.0493, "step": 1460 }, { "grad_norm": 0.3680717349052429, "learning_rate": 9.745480724951473e-05, "loss": 0.0521, "step": 1470 }, { "grad_norm": 0.24478930234909058, "learning_rate": 9.740246579503447e-05, "loss": 0.0552, "step": 1480 }, { "grad_norm": 0.4973839819431305, "learning_rate": 9.734960595478284e-05, "loss": 0.0511, "step": 1490 }, { "grad_norm": 0.3382282853126526, "learning_rate": 9.729622830682657e-05, "loss": 0.0562, "step": 1500 }, { "grad_norm": 0.34821027517318726, "learning_rate": 9.724233343489504e-05, "loss": 0.0522, "step": 1510 }, { "grad_norm": 0.43229493498802185, "learning_rate": 9.718792192837396e-05, "loss": 0.0544, "step": 1520 }, { "grad_norm": 0.3701668977737427, "learning_rate": 9.713299438229886e-05, "loss": 0.0535, "step": 1530 }, { "grad_norm": 0.36101725697517395, "learning_rate": 9.707755139734855e-05, "loss": 0.0559, "step": 1540 }, { "grad_norm": 0.3739871680736542, "learning_rate": 9.702159357983866e-05, "loss": 0.0525, "step": 1550 }, { "grad_norm": 0.3881339132785797, "learning_rate": 9.696512154171492e-05, "loss": 0.0532, "step": 1560 }, { "grad_norm": 0.3372054994106293, "learning_rate": 9.690813590054645e-05, "loss": 0.0463, "step": 1570 }, { "grad_norm": 0.34329137206077576, "learning_rate": 9.685063727951914e-05, "loss": 0.0472, "step": 1580 }, { "grad_norm": 0.3130674362182617, "learning_rate": 9.679262630742865e-05, "loss": 0.0462, "step": 1590 }, { "grad_norm": 0.3507116138935089, "learning_rate": 9.673410361867373e-05, "loss": 0.0455, "step": 1600 }, { "grad_norm": 0.4377332031726837, "learning_rate": 9.667506985324909e-05, "loss": 0.0417, "step": 1610 }, { "grad_norm": 0.27128973603248596, "learning_rate": 9.661552565673855e-05, "loss": 0.0387, "step": 1620 }, { "grad_norm": 0.25216543674468994, "learning_rate": 9.655547168030789e-05, "loss": 0.043, "step": 1630 }, { "grad_norm": 0.4172189235687256, "learning_rate": 9.649490858069777e-05, "loss": 0.0455, "step": 1640 }, { "grad_norm": 0.32755693793296814, "learning_rate": 9.643383702021658e-05, "loss": 0.051, "step": 1650 }, { "grad_norm": 0.4242418706417084, "learning_rate": 9.637225766673307e-05, "loss": 0.0405, "step": 1660 }, { "grad_norm": 0.2995989918708801, "learning_rate": 9.631017119366922e-05, "loss": 0.0467, "step": 1670 }, { "grad_norm": 0.4804268777370453, "learning_rate": 9.624757827999273e-05, "loss": 0.0474, "step": 1680 }, { "grad_norm": 0.30098140239715576, "learning_rate": 9.618447961020971e-05, "loss": 0.049, "step": 1690 }, { "grad_norm": 0.4148315489292145, "learning_rate": 9.612087587435707e-05, "loss": 0.0475, "step": 1700 }, { "grad_norm": 0.34601667523384094, "learning_rate": 9.605676776799508e-05, "loss": 0.045, "step": 1710 }, { "grad_norm": 0.3090530335903168, "learning_rate": 9.599215599219973e-05, "loss": 0.0431, "step": 1720 }, { "grad_norm": 0.3857743740081787, "learning_rate": 9.592704125355505e-05, "loss": 0.0445, "step": 1730 }, { "grad_norm": 0.2537960410118103, "learning_rate": 9.586142426414538e-05, "loss": 0.04, "step": 1740 }, { "grad_norm": 0.24324947595596313, "learning_rate": 9.57953057415476e-05, "loss": 0.0447, "step": 1750 }, { "grad_norm": 0.2522665560245514, "learning_rate": 9.572868640882328e-05, "loss": 0.0484, "step": 1760 }, { "grad_norm": 0.3916299045085907, "learning_rate": 9.56615669945108e-05, "loss": 0.042, "step": 1770 }, { "grad_norm": 0.28435930609703064, "learning_rate": 9.55939482326173e-05, "loss": 0.046, "step": 1780 }, { "grad_norm": 0.2893117666244507, "learning_rate": 9.552583086261069e-05, "loss": 0.0509, "step": 1790 }, { "grad_norm": 0.28712761402130127, "learning_rate": 9.545721562941168e-05, "loss": 0.0455, "step": 1800 }, { "grad_norm": 0.42221757769584656, "learning_rate": 9.538810328338543e-05, "loss": 0.0413, "step": 1810 }, { "grad_norm": 0.2665775716304779, "learning_rate": 9.531849458033349e-05, "loss": 0.0404, "step": 1820 }, { "grad_norm": 0.2509636878967285, "learning_rate": 9.524839028148547e-05, "loss": 0.0395, "step": 1830 }, { "grad_norm": 0.39951369166374207, "learning_rate": 9.517779115349077e-05, "loss": 0.0395, "step": 1840 }, { "grad_norm": 0.27439939975738525, "learning_rate": 9.510669796841014e-05, "loss": 0.0427, "step": 1850 }, { "grad_norm": 0.39046451449394226, "learning_rate": 9.503511150370727e-05, "loss": 0.0468, "step": 1860 }, { "grad_norm": 0.3335532248020172, "learning_rate": 9.496303254224024e-05, "loss": 0.0468, "step": 1870 }, { "grad_norm": 0.42565518617630005, "learning_rate": 9.489046187225306e-05, "loss": 0.0391, "step": 1880 }, { "grad_norm": 0.2588476240634918, "learning_rate": 9.481740028736692e-05, "loss": 0.0355, "step": 1890 }, { "grad_norm": 0.35043540596961975, "learning_rate": 9.474384858657164e-05, "loss": 0.0519, "step": 1900 }, { "grad_norm": 0.3746950924396515, "learning_rate": 9.466980757421679e-05, "loss": 0.0429, "step": 1910 }, { "grad_norm": 0.3585033714771271, "learning_rate": 9.459527806000305e-05, "loss": 0.0437, "step": 1920 }, { "grad_norm": 0.47203877568244934, "learning_rate": 9.452026085897325e-05, "loss": 0.0444, "step": 1930 }, { "grad_norm": 0.36170506477355957, "learning_rate": 9.444475679150348e-05, "loss": 0.0432, "step": 1940 }, { "grad_norm": 0.3162188231945038, "learning_rate": 9.436876668329411e-05, "loss": 0.0411, "step": 1950 }, { "grad_norm": 0.3193727135658264, "learning_rate": 9.429229136536079e-05, "loss": 0.0365, "step": 1960 }, { "grad_norm": 0.31013405323028564, "learning_rate": 9.421533167402534e-05, "loss": 0.0456, "step": 1970 }, { "grad_norm": 0.3845147490501404, "learning_rate": 9.413788845090666e-05, "loss": 0.0531, "step": 1980 }, { "grad_norm": 0.41048377752304077, "learning_rate": 9.405996254291136e-05, "loss": 0.0417, "step": 1990 }, { "grad_norm": 0.3203062117099762, "learning_rate": 9.398155480222474e-05, "loss": 0.038, "step": 2000 }, { "grad_norm": 0.2670988142490387, "learning_rate": 9.390266608630128e-05, "loss": 0.0396, "step": 2010 }, { "grad_norm": 0.32089856266975403, "learning_rate": 9.38232972578553e-05, "loss": 0.0435, "step": 2020 }, { "grad_norm": 0.22675646841526031, "learning_rate": 9.374344918485164e-05, "loss": 0.0443, "step": 2030 }, { "grad_norm": 0.307250052690506, "learning_rate": 9.366312274049602e-05, "loss": 0.0456, "step": 2040 }, { "grad_norm": 0.24730075895786285, "learning_rate": 9.358231880322554e-05, "loss": 0.0374, "step": 2050 }, { "grad_norm": 0.1980336457490921, "learning_rate": 9.350103825669916e-05, "loss": 0.0351, "step": 2060 }, { "grad_norm": 0.3448292315006256, "learning_rate": 9.341928198978787e-05, "loss": 0.0402, "step": 2070 }, { "grad_norm": 0.21278829872608185, "learning_rate": 9.333705089656512e-05, "loss": 0.0348, "step": 2080 }, { "grad_norm": 0.22461259365081787, "learning_rate": 9.325434587629698e-05, "loss": 0.0367, "step": 2090 }, { "grad_norm": 0.29972174763679504, "learning_rate": 9.31711678334323e-05, "loss": 0.0359, "step": 2100 }, { "grad_norm": 0.2837313711643219, "learning_rate": 9.308751767759282e-05, "loss": 0.0384, "step": 2110 }, { "grad_norm": 0.37176111340522766, "learning_rate": 9.300339632356325e-05, "loss": 0.0337, "step": 2120 }, { "grad_norm": 0.3224641680717468, "learning_rate": 9.291880469128124e-05, "loss": 0.0435, "step": 2130 }, { "grad_norm": 0.3874353766441345, "learning_rate": 9.283374370582732e-05, "loss": 0.0383, "step": 2140 }, { "grad_norm": 0.2514594495296478, "learning_rate": 9.274821429741482e-05, "loss": 0.0479, "step": 2150 }, { "grad_norm": 0.29283609986305237, "learning_rate": 9.266221740137961e-05, "loss": 0.0398, "step": 2160 }, { "grad_norm": 0.22666414082050323, "learning_rate": 9.257575395817001e-05, "loss": 0.037, "step": 2170 }, { "grad_norm": 0.38336607813835144, "learning_rate": 9.248882491333637e-05, "loss": 0.0401, "step": 2180 }, { "grad_norm": 0.3453926742076874, "learning_rate": 9.240143121752076e-05, "loss": 0.0373, "step": 2190 }, { "grad_norm": 0.18881116807460785, "learning_rate": 9.23135738264467e-05, "loss": 0.041, "step": 2200 }, { "grad_norm": 0.2322729229927063, "learning_rate": 9.222525370090849e-05, "loss": 0.0377, "step": 2210 }, { "grad_norm": 0.3147348165512085, "learning_rate": 9.213647180676088e-05, "loss": 0.0358, "step": 2220 }, { "grad_norm": 0.39017215371131897, "learning_rate": 9.204722911490846e-05, "loss": 0.0395, "step": 2230 }, { "grad_norm": 0.2994228005409241, "learning_rate": 9.1957526601295e-05, "loss": 0.0351, "step": 2240 }, { "grad_norm": 0.22871950268745422, "learning_rate": 9.186736524689281e-05, "loss": 0.0431, "step": 2250 }, { "grad_norm": 0.3205336034297943, "learning_rate": 9.177674603769204e-05, "loss": 0.0394, "step": 2260 }, { "grad_norm": 0.32509955763816833, "learning_rate": 9.168566996468983e-05, "loss": 0.0406, "step": 2270 }, { "grad_norm": 0.3418531119823456, "learning_rate": 9.159413802387951e-05, "loss": 0.0457, "step": 2280 }, { "grad_norm": 0.29726359248161316, "learning_rate": 9.150215121623974e-05, "loss": 0.043, "step": 2290 }, { "grad_norm": 0.21875575184822083, "learning_rate": 9.140971054772349e-05, "loss": 0.0351, "step": 2300 }, { "grad_norm": 0.3160592317581177, "learning_rate": 9.131681702924713e-05, "loss": 0.0435, "step": 2310 }, { "grad_norm": 0.35614874958992004, "learning_rate": 9.122347167667926e-05, "loss": 0.0337, "step": 2320 }, { "grad_norm": 0.3039318919181824, "learning_rate": 9.112967551082973e-05, "loss": 0.0487, "step": 2330 }, { "grad_norm": 0.2893160879611969, "learning_rate": 9.103542955743835e-05, "loss": 0.0351, "step": 2340 }, { "grad_norm": 0.298188716173172, "learning_rate": 9.094073484716381e-05, "loss": 0.0408, "step": 2350 }, { "grad_norm": 0.271395742893219, "learning_rate": 9.084559241557226e-05, "loss": 0.0337, "step": 2360 }, { "grad_norm": 0.28065118193626404, "learning_rate": 9.075000330312608e-05, "loss": 0.0337, "step": 2370 }, { "grad_norm": 0.3361847400665283, "learning_rate": 9.065396855517253e-05, "loss": 0.0344, "step": 2380 }, { "grad_norm": 0.2705598771572113, "learning_rate": 9.055748922193219e-05, "loss": 0.0387, "step": 2390 }, { "grad_norm": 0.34754690527915955, "learning_rate": 9.046056635848761e-05, "loss": 0.0386, "step": 2400 }, { "grad_norm": 0.3356654644012451, "learning_rate": 9.036320102477169e-05, "loss": 0.0329, "step": 2410 }, { "grad_norm": 0.1732822209596634, "learning_rate": 9.02653942855561e-05, "loss": 0.0341, "step": 2420 }, { "grad_norm": 0.2813863158226013, "learning_rate": 9.016714721043971e-05, "loss": 0.0344, "step": 2430 }, { "grad_norm": 0.2501205801963806, "learning_rate": 9.006846087383675e-05, "loss": 0.0355, "step": 2440 }, { "grad_norm": 0.21508678793907166, "learning_rate": 8.996933635496523e-05, "loss": 0.0333, "step": 2450 }, { "grad_norm": 0.2784106135368347, "learning_rate": 8.986977473783498e-05, "loss": 0.0404, "step": 2460 }, { "grad_norm": 0.2749677002429962, "learning_rate": 8.97697771112359e-05, "loss": 0.0347, "step": 2470 }, { "grad_norm": 0.31252336502075195, "learning_rate": 8.966934456872602e-05, "loss": 0.0357, "step": 2480 }, { "grad_norm": 0.2891818881034851, "learning_rate": 8.95684782086195e-05, "loss": 0.0301, "step": 2490 }, { "grad_norm": 0.2712235748767853, "learning_rate": 8.946717913397476e-05, "loss": 0.0383, "step": 2500 }, { "grad_norm": 0.2813902199268341, "learning_rate": 8.93654484525822e-05, "loss": 0.0371, "step": 2510 }, { "grad_norm": 0.3089042603969574, "learning_rate": 8.926328727695226e-05, "loss": 0.0342, "step": 2520 }, { "grad_norm": 0.291174978017807, "learning_rate": 8.916069672430319e-05, "loss": 0.0308, "step": 2530 }, { "grad_norm": 0.36839067935943604, "learning_rate": 8.905767791654884e-05, "loss": 0.0396, "step": 2540 }, { "grad_norm": 0.44516414403915405, "learning_rate": 8.895423198028638e-05, "loss": 0.0477, "step": 2550 }, { "grad_norm": 0.2058655172586441, "learning_rate": 8.885036004678402e-05, "loss": 0.0392, "step": 2560 }, { "grad_norm": 0.19877126812934875, "learning_rate": 8.874606325196857e-05, "loss": 0.0339, "step": 2570 }, { "grad_norm": 0.286281019449234, "learning_rate": 8.864134273641304e-05, "loss": 0.0372, "step": 2580 }, { "grad_norm": 0.19290633499622345, "learning_rate": 8.853619964532427e-05, "loss": 0.0388, "step": 2590 }, { "grad_norm": 0.38533124327659607, "learning_rate": 8.843063512853019e-05, "loss": 0.0345, "step": 2600 }, { "grad_norm": 0.24691641330718994, "learning_rate": 8.832465034046749e-05, "loss": 0.034, "step": 2610 }, { "grad_norm": 0.20866183936595917, "learning_rate": 8.821824644016882e-05, "loss": 0.0343, "step": 2620 }, { "grad_norm": 0.3668222427368164, "learning_rate": 8.811142459125019e-05, "loss": 0.0342, "step": 2630 }, { "grad_norm": 0.27398979663848877, "learning_rate": 8.800418596189822e-05, "loss": 0.0298, "step": 2640 }, { "grad_norm": 0.2508249878883362, "learning_rate": 8.789653172485737e-05, "loss": 0.0353, "step": 2650 }, { "grad_norm": 0.27359575033187866, "learning_rate": 8.778846305741715e-05, "loss": 0.0479, "step": 2660 }, { "grad_norm": 0.25260868668556213, "learning_rate": 8.767998114139918e-05, "loss": 0.0374, "step": 2670 }, { "grad_norm": 0.2010546624660492, "learning_rate": 8.757108716314429e-05, "loss": 0.0399, "step": 2680 }, { "grad_norm": 0.31015515327453613, "learning_rate": 8.746178231349962e-05, "loss": 0.0322, "step": 2690 }, { "grad_norm": 0.30862241983413696, "learning_rate": 8.735206778780549e-05, "loss": 0.0371, "step": 2700 }, { "grad_norm": 0.32803478837013245, "learning_rate": 8.724194478588234e-05, "loss": 0.0387, "step": 2710 }, { "grad_norm": 0.2550460398197174, "learning_rate": 8.713141451201772e-05, "loss": 0.0364, "step": 2720 }, { "grad_norm": 0.3427490293979645, "learning_rate": 8.702047817495295e-05, "loss": 0.0419, "step": 2730 }, { "grad_norm": 0.4121761918067932, "learning_rate": 8.69091369878701e-05, "loss": 0.0384, "step": 2740 }, { "grad_norm": 0.2661398649215698, "learning_rate": 8.679739216837849e-05, "loss": 0.0322, "step": 2750 }, { "grad_norm": 0.27513355016708374, "learning_rate": 8.66852449385016e-05, "loss": 0.0336, "step": 2760 }, { "grad_norm": 0.2936078608036041, "learning_rate": 8.657269652466356e-05, "loss": 0.0334, "step": 2770 }, { "grad_norm": 0.2643412947654724, "learning_rate": 8.645974815767577e-05, "loss": 0.0339, "step": 2780 }, { "grad_norm": 0.28178632259368896, "learning_rate": 8.634640107272351e-05, "loss": 0.0314, "step": 2790 }, { "grad_norm": 0.3185223937034607, "learning_rate": 8.623265650935234e-05, "loss": 0.0327, "step": 2800 }, { "grad_norm": 0.21827954053878784, "learning_rate": 8.611851571145456e-05, "loss": 0.0351, "step": 2810 }, { "grad_norm": 0.24170783162117004, "learning_rate": 8.600397992725566e-05, "loss": 0.0291, "step": 2820 }, { "grad_norm": 0.277829110622406, "learning_rate": 8.588905040930061e-05, "loss": 0.0372, "step": 2830 }, { "grad_norm": 0.27800947427749634, "learning_rate": 8.577372841444022e-05, "loss": 0.0328, "step": 2840 }, { "grad_norm": 0.3463057279586792, "learning_rate": 8.565801520381736e-05, "loss": 0.0327, "step": 2850 }, { "grad_norm": 0.27984192967414856, "learning_rate": 8.554191204285313e-05, "loss": 0.033, "step": 2860 }, { "grad_norm": 0.274709016084671, "learning_rate": 8.542542020123315e-05, "loss": 0.0339, "step": 2870 }, { "grad_norm": 0.23455101251602173, "learning_rate": 8.530854095289347e-05, "loss": 0.0306, "step": 2880 }, { "grad_norm": 0.23256126046180725, "learning_rate": 8.519127557600688e-05, "loss": 0.0325, "step": 2890 }, { "grad_norm": 0.20642027258872986, "learning_rate": 8.507362535296871e-05, "loss": 0.0352, "step": 2900 }, { "grad_norm": 0.2514182925224304, "learning_rate": 8.495559157038299e-05, "loss": 0.0466, "step": 2910 }, { "grad_norm": 0.21313071250915527, "learning_rate": 8.483717551904823e-05, "loss": 0.0317, "step": 2920 }, { "grad_norm": 0.3188178837299347, "learning_rate": 8.47183784939434e-05, "loss": 0.0318, "step": 2930 }, { "grad_norm": 0.26879018545150757, "learning_rate": 8.459920179421374e-05, "loss": 0.0314, "step": 2940 }, { "grad_norm": 0.29836633801460266, "learning_rate": 8.447964672315656e-05, "loss": 0.0355, "step": 2950 }, { "grad_norm": 0.24870367348194122, "learning_rate": 8.435971458820692e-05, "loss": 0.038, "step": 2960 }, { "grad_norm": 0.2699428200721741, "learning_rate": 8.423940670092345e-05, "loss": 0.0369, "step": 2970 }, { "grad_norm": 0.2697128653526306, "learning_rate": 8.411872437697394e-05, "loss": 0.0327, "step": 2980 }, { "grad_norm": 0.2513008415699005, "learning_rate": 8.399766893612096e-05, "loss": 0.0343, "step": 2990 }, { "grad_norm": 0.2980031669139862, "learning_rate": 8.38762417022074e-05, "loss": 0.0397, "step": 3000 }, { "grad_norm": 0.2906007170677185, "learning_rate": 8.375444400314204e-05, "loss": 0.0365, "step": 3010 }, { "grad_norm": 0.29300904273986816, "learning_rate": 8.3632277170885e-05, "loss": 0.0386, "step": 3020 }, { "grad_norm": 0.27136802673339844, "learning_rate": 8.350974254143318e-05, "loss": 0.0311, "step": 3030 }, { "grad_norm": 0.28083622455596924, "learning_rate": 8.338684145480566e-05, "loss": 0.0294, "step": 3040 }, { "grad_norm": 0.26537853479385376, "learning_rate": 8.326357525502904e-05, "loss": 0.0328, "step": 3050 }, { "grad_norm": 0.32124215364456177, "learning_rate": 8.313994529012273e-05, "loss": 0.0358, "step": 3060 }, { "grad_norm": 0.32729002833366394, "learning_rate": 8.301595291208422e-05, "loss": 0.0372, "step": 3070 }, { "grad_norm": 0.2976420223712921, "learning_rate": 8.289159947687427e-05, "loss": 0.0355, "step": 3080 }, { "grad_norm": 0.2107667773962021, "learning_rate": 8.276688634440216e-05, "loss": 0.0341, "step": 3090 }, { "grad_norm": 0.27912428975105286, "learning_rate": 8.26418148785107e-05, "loss": 0.036, "step": 3100 }, { "grad_norm": 0.21707573533058167, "learning_rate": 8.251638644696141e-05, "loss": 0.0373, "step": 3110 }, { "grad_norm": 0.19843187928199768, "learning_rate": 8.23906024214195e-05, "loss": 0.0327, "step": 3120 }, { "grad_norm": 0.20462940633296967, "learning_rate": 8.226446417743897e-05, "loss": 0.0324, "step": 3130 }, { "grad_norm": 0.17077213525772095, "learning_rate": 8.213797309444742e-05, "loss": 0.0343, "step": 3140 }, { "grad_norm": 0.22978489100933075, "learning_rate": 8.201113055573105e-05, "loss": 0.0326, "step": 3150 }, { "grad_norm": 0.37337610125541687, "learning_rate": 8.188393794841958e-05, "loss": 0.0295, "step": 3160 }, { "grad_norm": 0.3887908160686493, "learning_rate": 8.175639666347094e-05, "loss": 0.0345, "step": 3170 }, { "grad_norm": 0.3370738923549652, "learning_rate": 8.162850809565623e-05, "loss": 0.0317, "step": 3180 }, { "grad_norm": 0.24584807455539703, "learning_rate": 8.150027364354431e-05, "loss": 0.0313, "step": 3190 }, { "grad_norm": 0.3442938029766083, "learning_rate": 8.137169470948662e-05, "loss": 0.0416, "step": 3200 }, { "grad_norm": 0.2810775637626648, "learning_rate": 8.124277269960179e-05, "loss": 0.0363, "step": 3210 }, { "grad_norm": 0.19405275583267212, "learning_rate": 8.111350902376023e-05, "loss": 0.0305, "step": 3220 }, { "grad_norm": 0.3256344497203827, "learning_rate": 8.098390509556883e-05, "loss": 0.0304, "step": 3230 }, { "grad_norm": 0.264670729637146, "learning_rate": 8.085396233235536e-05, "loss": 0.026, "step": 3240 }, { "grad_norm": 0.2893134653568268, "learning_rate": 8.072368215515306e-05, "loss": 0.0266, "step": 3250 }, { "grad_norm": 0.2728324234485626, "learning_rate": 8.059306598868506e-05, "loss": 0.0392, "step": 3260 }, { "grad_norm": 0.3051053583621979, "learning_rate": 8.046211526134888e-05, "loss": 0.0346, "step": 3270 }, { "grad_norm": 0.2062731683254242, "learning_rate": 8.033083140520065e-05, "loss": 0.0285, "step": 3280 }, { "grad_norm": 0.2715912163257599, "learning_rate": 8.019921585593962e-05, "loss": 0.032, "step": 3290 }, { "grad_norm": 0.3166407346725464, "learning_rate": 8.006727005289232e-05, "loss": 0.0282, "step": 3300 }, { "grad_norm": 0.2292829304933548, "learning_rate": 7.993499543899692e-05, "loss": 0.0292, "step": 3310 }, { "grad_norm": 0.33925873041152954, "learning_rate": 7.980239346078742e-05, "loss": 0.0287, "step": 3320 }, { "grad_norm": 0.28408676385879517, "learning_rate": 7.966946556837778e-05, "loss": 0.0321, "step": 3330 }, { "grad_norm": 0.16675904393196106, "learning_rate": 7.953621321544616e-05, "loss": 0.0297, "step": 3340 }, { "grad_norm": 0.26386404037475586, "learning_rate": 7.940263785921896e-05, "loss": 0.0352, "step": 3350 }, { "grad_norm": 0.18476559221744537, "learning_rate": 7.926874096045482e-05, "loss": 0.0293, "step": 3360 }, { "grad_norm": 0.4043673872947693, "learning_rate": 7.913452398342881e-05, "loss": 0.0427, "step": 3370 }, { "grad_norm": 0.3320227265357971, "learning_rate": 7.89999883959163e-05, "loss": 0.0358, "step": 3380 }, { "grad_norm": 0.2546858787536621, "learning_rate": 7.886513566917687e-05, "loss": 0.0348, "step": 3390 }, { "grad_norm": 0.31003737449645996, "learning_rate": 7.872996727793838e-05, "loss": 0.0377, "step": 3400 }, { "grad_norm": 0.28634345531463623, "learning_rate": 7.859448470038069e-05, "loss": 0.0303, "step": 3410 }, { "grad_norm": 0.17955291271209717, "learning_rate": 7.845868941811956e-05, "loss": 0.0327, "step": 3420 }, { "grad_norm": 0.2258872240781784, "learning_rate": 7.832258291619043e-05, "loss": 0.0295, "step": 3430 }, { "grad_norm": 0.2839079201221466, "learning_rate": 7.81861666830322e-05, "loss": 0.0296, "step": 3440 }, { "grad_norm": 0.18423597514629364, "learning_rate": 7.804944221047097e-05, "loss": 0.0258, "step": 3450 }, { "grad_norm": 0.32803013920783997, "learning_rate": 7.791241099370364e-05, "loss": 0.0351, "step": 3460 }, { "grad_norm": 0.15780730545520782, "learning_rate": 7.777507453128163e-05, "loss": 0.0277, "step": 3470 }, { "grad_norm": 0.20855095982551575, "learning_rate": 7.763743432509451e-05, "loss": 0.0299, "step": 3480 }, { "grad_norm": 0.2275158166885376, "learning_rate": 7.749949188035353e-05, "loss": 0.0294, "step": 3490 }, { "grad_norm": 0.2686370313167572, "learning_rate": 7.736124870557516e-05, "loss": 0.038, "step": 3500 }, { "grad_norm": 0.34551769495010376, "learning_rate": 7.722270631256459e-05, "loss": 0.0336, "step": 3510 }, { "grad_norm": 0.23830966651439667, "learning_rate": 7.708386621639925e-05, "loss": 0.0311, "step": 3520 }, { "grad_norm": 0.2267194390296936, "learning_rate": 7.694472993541219e-05, "loss": 0.0335, "step": 3530 }, { "grad_norm": 0.20236381888389587, "learning_rate": 7.680529899117547e-05, "loss": 0.0345, "step": 3540 }, { "grad_norm": 0.194863423705101, "learning_rate": 7.666557490848358e-05, "loss": 0.0323, "step": 3550 }, { "grad_norm": 0.20245641469955444, "learning_rate": 7.65255592153367e-05, "loss": 0.0308, "step": 3560 }, { "grad_norm": 0.25559133291244507, "learning_rate": 7.638525344292402e-05, "loss": 0.0321, "step": 3570 }, { "grad_norm": 0.2867175042629242, "learning_rate": 7.624465912560697e-05, "loss": 0.0346, "step": 3580 }, { "grad_norm": 0.16951186954975128, "learning_rate": 7.610377780090249e-05, "loss": 0.0289, "step": 3590 }, { "grad_norm": 0.1567351371049881, "learning_rate": 7.596261100946618e-05, "loss": 0.0322, "step": 3600 }, { "grad_norm": 0.21078485250473022, "learning_rate": 7.582116029507542e-05, "loss": 0.027, "step": 3610 }, { "grad_norm": 0.3169976472854614, "learning_rate": 7.56794272046126e-05, "loss": 0.0291, "step": 3620 }, { "grad_norm": 0.21792450547218323, "learning_rate": 7.55374132880481e-05, "loss": 0.0339, "step": 3630 }, { "grad_norm": 0.2316005676984787, "learning_rate": 7.539512009842333e-05, "loss": 0.0316, "step": 3640 }, { "grad_norm": 0.22255703806877136, "learning_rate": 7.525254919183382e-05, "loss": 0.034, "step": 3650 }, { "grad_norm": 0.1907738596200943, "learning_rate": 7.510970212741215e-05, "loss": 0.026, "step": 3660 }, { "grad_norm": 0.18570856750011444, "learning_rate": 7.496658046731096e-05, "loss": 0.0302, "step": 3670 }, { "grad_norm": 0.17549049854278564, "learning_rate": 7.482318577668578e-05, "loss": 0.033, "step": 3680 }, { "grad_norm": 0.25804534554481506, "learning_rate": 7.467951962367796e-05, "loss": 0.0343, "step": 3690 }, { "grad_norm": 0.2328014224767685, "learning_rate": 7.453558357939755e-05, "loss": 0.0289, "step": 3700 }, { "grad_norm": 0.2675175070762634, "learning_rate": 7.439137921790606e-05, "loss": 0.0349, "step": 3710 }, { "grad_norm": 0.26529091596603394, "learning_rate": 7.42469081161993e-05, "loss": 0.031, "step": 3720 }, { "grad_norm": 0.2291182279586792, "learning_rate": 7.410217185419006e-05, "loss": 0.0238, "step": 3730 }, { "grad_norm": 0.266123503446579, "learning_rate": 7.395717201469095e-05, "loss": 0.0301, "step": 3740 }, { "grad_norm": 0.2607523500919342, "learning_rate": 7.381191018339696e-05, "loss": 0.0291, "step": 3750 }, { "grad_norm": 0.2305401712656021, "learning_rate": 7.36663879488682e-05, "loss": 0.0272, "step": 3760 }, { "grad_norm": 0.258056104183197, "learning_rate": 7.352060690251254e-05, "loss": 0.0341, "step": 3770 }, { "grad_norm": 0.1916736215353012, "learning_rate": 7.337456863856811e-05, "loss": 0.0286, "step": 3780 }, { "grad_norm": 0.19997639954090118, "learning_rate": 7.3228274754086e-05, "loss": 0.028, "step": 3790 }, { "grad_norm": 0.2288093864917755, "learning_rate": 7.308172684891267e-05, "loss": 0.0302, "step": 3800 }, { "grad_norm": 0.1596165895462036, "learning_rate": 7.293492652567255e-05, "loss": 0.0244, "step": 3810 }, { "grad_norm": 0.18738606572151184, "learning_rate": 7.278787538975043e-05, "loss": 0.0319, "step": 3820 }, { "grad_norm": 0.27061331272125244, "learning_rate": 7.2640575049274e-05, "loss": 0.0324, "step": 3830 }, { "grad_norm": 0.19814740121364594, "learning_rate": 7.249302711509616e-05, "loss": 0.0271, "step": 3840 }, { "grad_norm": 0.24830563366413116, "learning_rate": 7.23452332007775e-05, "loss": 0.0263, "step": 3850 }, { "grad_norm": 0.2686096727848053, "learning_rate": 7.219719492256858e-05, "loss": 0.031, "step": 3860 }, { "grad_norm": 0.2932262718677521, "learning_rate": 7.20489138993923e-05, "loss": 0.0276, "step": 3870 }, { "grad_norm": 0.22418922185897827, "learning_rate": 7.190039175282614e-05, "loss": 0.0289, "step": 3880 }, { "grad_norm": 0.2663893401622772, "learning_rate": 7.175163010708455e-05, "loss": 0.0287, "step": 3890 }, { "grad_norm": 0.18200421333312988, "learning_rate": 7.1602630589001e-05, "loss": 0.0289, "step": 3900 }, { "grad_norm": 0.24957586824893951, "learning_rate": 7.14533948280104e-05, "loss": 0.0278, "step": 3910 }, { "grad_norm": 0.20494110882282257, "learning_rate": 7.130392445613109e-05, "loss": 0.0288, "step": 3920 }, { "grad_norm": 0.33091986179351807, "learning_rate": 7.115422110794711e-05, "loss": 0.0339, "step": 3930 }, { "grad_norm": 0.2313767671585083, "learning_rate": 7.100428642059033e-05, "loss": 0.0338, "step": 3940 }, { "grad_norm": 0.25328120589256287, "learning_rate": 7.08541220337224e-05, "loss": 0.0302, "step": 3950 }, { "grad_norm": 0.3268498182296753, "learning_rate": 7.070372958951706e-05, "loss": 0.0288, "step": 3960 }, { "grad_norm": 0.23085163533687592, "learning_rate": 7.055311073264194e-05, "loss": 0.0313, "step": 3970 }, { "grad_norm": 0.15610571205615997, "learning_rate": 7.040226711024077e-05, "loss": 0.025, "step": 3980 }, { "grad_norm": 0.35929346084594727, "learning_rate": 7.02512003719152e-05, "loss": 0.0268, "step": 3990 }, { "grad_norm": 0.16080673038959503, "learning_rate": 7.00999121697069e-05, "loss": 0.0296, "step": 4000 }, { "grad_norm": 0.2296619713306427, "learning_rate": 6.99484041580794e-05, "loss": 0.0282, "step": 4010 }, { "grad_norm": 0.18837708234786987, "learning_rate": 6.979667799390004e-05, "loss": 0.0265, "step": 4020 }, { "grad_norm": 0.22696955502033234, "learning_rate": 6.964473533642185e-05, "loss": 0.0326, "step": 4030 }, { "grad_norm": 0.13756400346755981, "learning_rate": 6.949257784726539e-05, "loss": 0.0285, "step": 4040 }, { "grad_norm": 0.22082467377185822, "learning_rate": 6.934020719040056e-05, "loss": 0.0249, "step": 4050 }, { "grad_norm": 0.20093458890914917, "learning_rate": 6.918762503212848e-05, "loss": 0.0306, "step": 4060 }, { "grad_norm": 0.20564433932304382, "learning_rate": 6.903483304106319e-05, "loss": 0.0301, "step": 4070 }, { "grad_norm": 0.2906251847743988, "learning_rate": 6.888183288811341e-05, "loss": 0.0243, "step": 4080 }, { "grad_norm": 0.21277932822704315, "learning_rate": 6.87286262464643e-05, "loss": 0.0269, "step": 4090 }, { "grad_norm": 0.26809293031692505, "learning_rate": 6.857521479155915e-05, "loss": 0.0309, "step": 4100 }, { "grad_norm": 0.274167001247406, "learning_rate": 6.842160020108104e-05, "loss": 0.0293, "step": 4110 }, { "grad_norm": 0.3058222234249115, "learning_rate": 6.826778415493455e-05, "loss": 0.0315, "step": 4120 }, { "grad_norm": 0.36119914054870605, "learning_rate": 6.811376833522729e-05, "loss": 0.0307, "step": 4130 }, { "grad_norm": 0.31132474541664124, "learning_rate": 6.795955442625159e-05, "loss": 0.0317, "step": 4140 }, { "grad_norm": 0.23346687853336334, "learning_rate": 6.780514411446608e-05, "loss": 0.0274, "step": 4150 }, { "grad_norm": 0.26099225878715515, "learning_rate": 6.765053908847716e-05, "loss": 0.0249, "step": 4160 }, { "grad_norm": 0.25587785243988037, "learning_rate": 6.749574103902064e-05, "loss": 0.0315, "step": 4170 }, { "grad_norm": 0.3356243073940277, "learning_rate": 6.734075165894317e-05, "loss": 0.034, "step": 4180 }, { "grad_norm": 0.2186242640018463, "learning_rate": 6.71855726431838e-05, "loss": 0.0311, "step": 4190 }, { "grad_norm": 0.24243560433387756, "learning_rate": 6.703020568875538e-05, "loss": 0.0256, "step": 4200 }, { "grad_norm": 0.14695316553115845, "learning_rate": 6.687465249472603e-05, "loss": 0.0244, "step": 4210 }, { "grad_norm": 0.20663321018218994, "learning_rate": 6.671891476220055e-05, "loss": 0.025, "step": 4220 }, { "grad_norm": 0.2125050574541092, "learning_rate": 6.656299419430183e-05, "loss": 0.0241, "step": 4230 }, { "grad_norm": 0.22032848000526428, "learning_rate": 6.640689249615223e-05, "loss": 0.0306, "step": 4240 }, { "grad_norm": 0.2904788851737976, "learning_rate": 6.625061137485491e-05, "loss": 0.0355, "step": 4250 }, { "grad_norm": 0.1588737666606903, "learning_rate": 6.609415253947517e-05, "loss": 0.0316, "step": 4260 }, { "grad_norm": 0.2341064214706421, "learning_rate": 6.593751770102178e-05, "loss": 0.0284, "step": 4270 }, { "grad_norm": 0.27394556999206543, "learning_rate": 6.578070857242823e-05, "loss": 0.0284, "step": 4280 }, { "grad_norm": 0.24494680762290955, "learning_rate": 6.562372686853402e-05, "loss": 0.0264, "step": 4290 }, { "grad_norm": 0.24473708868026733, "learning_rate": 6.546657430606593e-05, "loss": 0.0279, "step": 4300 }, { "grad_norm": 0.3806935250759125, "learning_rate": 6.530925260361918e-05, "loss": 0.0244, "step": 4310 }, { "grad_norm": 0.248376727104187, "learning_rate": 6.515176348163871e-05, "loss": 0.0276, "step": 4320 }, { "grad_norm": 0.1574171632528305, "learning_rate": 6.499410866240032e-05, "loss": 0.0239, "step": 4330 }, { "grad_norm": 0.2769834101200104, "learning_rate": 6.48362898699919e-05, "loss": 0.0266, "step": 4340 }, { "grad_norm": 0.31519681215286255, "learning_rate": 6.467830883029443e-05, "loss": 0.0264, "step": 4350 }, { "grad_norm": 0.27628466486930847, "learning_rate": 6.452016727096326e-05, "loss": 0.0246, "step": 4360 }, { "grad_norm": 0.21230652928352356, "learning_rate": 6.436186692140916e-05, "loss": 0.0247, "step": 4370 }, { "grad_norm": 0.18257810175418854, "learning_rate": 6.420340951277938e-05, "loss": 0.0277, "step": 4380 }, { "grad_norm": 0.2835541367530823, "learning_rate": 6.404479677793874e-05, "loss": 0.0303, "step": 4390 }, { "grad_norm": 0.22322863340377808, "learning_rate": 6.388603045145075e-05, "loss": 0.036, "step": 4400 }, { "grad_norm": 0.20386968553066254, "learning_rate": 6.372711226955843e-05, "loss": 0.0293, "step": 4410 }, { "grad_norm": 0.20988278090953827, "learning_rate": 6.356804397016564e-05, "loss": 0.0294, "step": 4420 }, { "grad_norm": 0.20170122385025024, "learning_rate": 6.340882729281779e-05, "loss": 0.0278, "step": 4430 }, { "grad_norm": 0.23380529880523682, "learning_rate": 6.324946397868294e-05, "loss": 0.0298, "step": 4440 }, { "grad_norm": 0.1665024310350418, "learning_rate": 6.308995577053276e-05, "loss": 0.0234, "step": 4450 }, { "grad_norm": 0.22018113732337952, "learning_rate": 6.293030441272347e-05, "loss": 0.0258, "step": 4460 }, { "grad_norm": 0.20327134430408478, "learning_rate": 6.277051165117677e-05, "loss": 0.0247, "step": 4470 }, { "grad_norm": 0.18797895312309265, "learning_rate": 6.261057923336064e-05, "loss": 0.0256, "step": 4480 }, { "grad_norm": 0.25495514273643494, "learning_rate": 6.245050890827042e-05, "loss": 0.0269, "step": 4490 }, { "grad_norm": 0.2462357133626938, "learning_rate": 6.229030242640952e-05, "loss": 0.0253, "step": 4500 }, { "grad_norm": 0.2605406641960144, "learning_rate": 6.212996153977037e-05, "loss": 0.0346, "step": 4510 }, { "grad_norm": 0.2589111328125, "learning_rate": 6.196948800181523e-05, "loss": 0.0271, "step": 4520 }, { "grad_norm": 0.1928110122680664, "learning_rate": 6.180888356745695e-05, "loss": 0.0253, "step": 4530 }, { "grad_norm": 0.27965453267097473, "learning_rate": 6.164814999303995e-05, "loss": 0.0252, "step": 4540 }, { "grad_norm": 0.20948413014411926, "learning_rate": 6.148728903632081e-05, "loss": 0.0219, "step": 4550 }, { "grad_norm": 0.22863753139972687, "learning_rate": 6.132630245644921e-05, "loss": 0.0224, "step": 4560 }, { "grad_norm": 0.19739468395709991, "learning_rate": 6.116519201394857e-05, "loss": 0.026, "step": 4570 }, { "grad_norm": 0.24227376282215118, "learning_rate": 6.10039594706969e-05, "loss": 0.026, "step": 4580 }, { "grad_norm": 0.1937863528728485, "learning_rate": 6.084260658990744e-05, "loss": 0.0251, "step": 4590 }, { "grad_norm": 0.20023277401924133, "learning_rate": 6.068113513610943e-05, "loss": 0.0251, "step": 4600 }, { "grad_norm": 0.25035014748573303, "learning_rate": 6.0519546875128876e-05, "loss": 0.0272, "step": 4610 }, { "grad_norm": 0.20950275659561157, "learning_rate": 6.035784357406906e-05, "loss": 0.0323, "step": 4620 }, { "grad_norm": 0.1854996532201767, "learning_rate": 6.01960270012914e-05, "loss": 0.0302, "step": 4630 }, { "grad_norm": 0.31139492988586426, "learning_rate": 6.003409892639599e-05, "loss": 0.0279, "step": 4640 }, { "grad_norm": 0.23125475645065308, "learning_rate": 5.9872061120202336e-05, "loss": 0.0262, "step": 4650 }, { "grad_norm": 0.2486438900232315, "learning_rate": 5.9709915354729914e-05, "loss": 0.0334, "step": 4660 }, { "grad_norm": 0.21768805384635925, "learning_rate": 5.9547663403178824e-05, "loss": 0.0263, "step": 4670 }, { "grad_norm": 0.19530339539051056, "learning_rate": 5.9385307039910445e-05, "loss": 0.0189, "step": 4680 }, { "grad_norm": 0.22342173755168915, "learning_rate": 5.922284804042792e-05, "loss": 0.0298, "step": 4690 }, { "grad_norm": 0.212743878364563, "learning_rate": 5.906028818135687e-05, "loss": 0.0316, "step": 4700 }, { "grad_norm": 0.21438507735729218, "learning_rate": 5.889762924042585e-05, "loss": 0.0234, "step": 4710 }, { "grad_norm": 0.19083158671855927, "learning_rate": 5.873487299644699e-05, "loss": 0.028, "step": 4720 }, { "grad_norm": 0.2479652315378189, "learning_rate": 5.857202122929649e-05, "loss": 0.0259, "step": 4730 }, { "grad_norm": 0.2593628168106079, "learning_rate": 5.840907571989518e-05, "loss": 0.0218, "step": 4740 }, { "grad_norm": 0.20351749658584595, "learning_rate": 5.824603825018904e-05, "loss": 0.0279, "step": 4750 }, { "grad_norm": 0.14537928998470306, "learning_rate": 5.808291060312975e-05, "loss": 0.0241, "step": 4760 }, { "grad_norm": 0.2960677742958069, "learning_rate": 5.7919694562655083e-05, "loss": 0.0233, "step": 4770 }, { "grad_norm": 0.19677580893039703, "learning_rate": 5.775639191366954e-05, "loss": 0.0236, "step": 4780 }, { "grad_norm": 0.2516234815120697, "learning_rate": 5.75930044420247e-05, "loss": 0.0233, "step": 4790 }, { "grad_norm": 0.249427929520607, "learning_rate": 5.74295339344998e-05, "loss": 0.0251, "step": 4800 }, { "grad_norm": 0.22791625559329987, "learning_rate": 5.726598217878211e-05, "loss": 0.0205, "step": 4810 }, { "grad_norm": 0.2155335247516632, "learning_rate": 5.71023509634474e-05, "loss": 0.0203, "step": 4820 }, { "grad_norm": 0.1742229014635086, "learning_rate": 5.693864207794049e-05, "loss": 0.0268, "step": 4830 }, { "grad_norm": 0.22918163239955902, "learning_rate": 5.677485731255545e-05, "loss": 0.0372, "step": 4840 }, { "grad_norm": 0.21200183033943176, "learning_rate": 5.6610998458416296e-05, "loss": 0.0252, "step": 4850 }, { "grad_norm": 0.2557915151119232, "learning_rate": 5.644706730745716e-05, "loss": 0.0256, "step": 4860 }, { "grad_norm": 0.15262071788311005, "learning_rate": 5.628306565240287e-05, "loss": 0.0236, "step": 4870 }, { "grad_norm": 0.16425473988056183, "learning_rate": 5.611899528674923e-05, "loss": 0.0288, "step": 4880 }, { "grad_norm": 0.2221406251192093, "learning_rate": 5.595485800474349e-05, "loss": 0.0263, "step": 4890 }, { "grad_norm": 0.23902589082717896, "learning_rate": 5.579065560136467e-05, "loss": 0.0274, "step": 4900 }, { "grad_norm": 0.20001167058944702, "learning_rate": 5.562638987230392e-05, "loss": 0.0251, "step": 4910 }, { "grad_norm": 0.28417783975601196, "learning_rate": 5.546206261394498e-05, "loss": 0.0225, "step": 4920 }, { "grad_norm": 0.22098001837730408, "learning_rate": 5.529767562334437e-05, "loss": 0.028, "step": 4930 }, { "grad_norm": 0.22791694104671478, "learning_rate": 5.5133230698211926e-05, "loss": 0.023, "step": 4940 }, { "grad_norm": 0.1528070718050003, "learning_rate": 5.496872963689096e-05, "loss": 0.0262, "step": 4950 }, { "grad_norm": 0.21025723218917847, "learning_rate": 5.4804174238338756e-05, "loss": 0.0217, "step": 4960 }, { "grad_norm": 0.1836385279893875, "learning_rate": 5.463956630210678e-05, "loss": 0.0291, "step": 4970 }, { "grad_norm": 0.22137781977653503, "learning_rate": 5.4474907628321046e-05, "loss": 0.0299, "step": 4980 }, { "grad_norm": 0.24128715693950653, "learning_rate": 5.431020001766244e-05, "loss": 0.0239, "step": 4990 }, { "grad_norm": 0.20182283222675323, "learning_rate": 5.4145445271346986e-05, "loss": 0.0245, "step": 5000 }, { "grad_norm": 0.13724316656589508, "learning_rate": 5.398064519110622e-05, "loss": 0.0232, "step": 5010 }, { "grad_norm": 0.26050880551338196, "learning_rate": 5.3815801579167394e-05, "loss": 0.0251, "step": 5020 }, { "grad_norm": 0.2751350700855255, "learning_rate": 5.365091623823382e-05, "loss": 0.027, "step": 5030 }, { "grad_norm": 0.14769554138183594, "learning_rate": 5.348599097146521e-05, "loss": 0.0232, "step": 5040 }, { "grad_norm": 0.25758570432662964, "learning_rate": 5.3321027582457836e-05, "loss": 0.0235, "step": 5050 }, { "grad_norm": 0.2353949397802353, "learning_rate": 5.315602787522491e-05, "loss": 0.0257, "step": 5060 }, { "grad_norm": 0.200799360871315, "learning_rate": 5.299099365417678e-05, "loss": 0.0198, "step": 5070 }, { "grad_norm": 0.1694364845752716, "learning_rate": 5.2825926724101236e-05, "loss": 0.025, "step": 5080 }, { "grad_norm": 0.19024622440338135, "learning_rate": 5.26608288901438e-05, "loss": 0.0231, "step": 5090 }, { "grad_norm": 0.24238838255405426, "learning_rate": 5.24957019577879e-05, "loss": 0.0271, "step": 5100 }, { "grad_norm": 0.25330886244773865, "learning_rate": 5.2330547732835266e-05, "loss": 0.0228, "step": 5110 }, { "grad_norm": 0.21432088315486908, "learning_rate": 5.2165368021385996e-05, "loss": 0.0339, "step": 5120 }, { "grad_norm": 0.2222827821969986, "learning_rate": 5.200016462981897e-05, "loss": 0.0217, "step": 5130 }, { "grad_norm": 0.2205737978219986, "learning_rate": 5.1834939364772015e-05, "loss": 0.0218, "step": 5140 }, { "grad_norm": 0.23958593606948853, "learning_rate": 5.166969403312214e-05, "loss": 0.0224, "step": 5150 }, { "grad_norm": 0.26182669401168823, "learning_rate": 5.1504430441965844e-05, "loss": 0.0273, "step": 5160 }, { "grad_norm": 0.19815179705619812, "learning_rate": 5.133915039859923e-05, "loss": 0.0187, "step": 5170 }, { "grad_norm": 0.22778578102588654, "learning_rate": 5.1173855710498444e-05, "loss": 0.0221, "step": 5180 }, { "grad_norm": 0.22530879080295563, "learning_rate": 5.100854818529967e-05, "loss": 0.0244, "step": 5190 }, { "grad_norm": 0.1883741021156311, "learning_rate": 5.084322963077951e-05, "loss": 0.0282, "step": 5200 }, { "grad_norm": 0.18606232106685638, "learning_rate": 5.067790185483522e-05, "loss": 0.023, "step": 5210 }, { "grad_norm": 0.17160527408123016, "learning_rate": 5.0512566665464844e-05, "loss": 0.0233, "step": 5220 }, { "grad_norm": 0.23687072098255157, "learning_rate": 5.034722587074755e-05, "loss": 0.0258, "step": 5230 }, { "grad_norm": 0.18745245039463043, "learning_rate": 5.018188127882375e-05, "loss": 0.0184, "step": 5240 }, { "grad_norm": 0.16375330090522766, "learning_rate": 5.0016534697875417e-05, "loss": 0.0205, "step": 5250 }, { "grad_norm": 0.17337261140346527, "learning_rate": 4.9851187936106294e-05, "loss": 0.0233, "step": 5260 }, { "grad_norm": 0.13573043048381805, "learning_rate": 4.968584280172206e-05, "loss": 0.0237, "step": 5270 }, { "grad_norm": 0.12221360951662064, "learning_rate": 4.95205011029106e-05, "loss": 0.0227, "step": 5280 }, { "grad_norm": 0.17494088411331177, "learning_rate": 4.935516464782227e-05, "loss": 0.0231, "step": 5290 }, { "grad_norm": 0.2583000659942627, "learning_rate": 4.918983524455003e-05, "loss": 0.0236, "step": 5300 }, { "grad_norm": 0.25226151943206787, "learning_rate": 4.9024514701109766e-05, "loss": 0.0214, "step": 5310 }, { "grad_norm": 0.2820110321044922, "learning_rate": 4.885920482542043e-05, "loss": 0.0251, "step": 5320 }, { "grad_norm": 0.19450557231903076, "learning_rate": 4.869390742528438e-05, "loss": 0.0205, "step": 5330 }, { "grad_norm": 0.22665174305438995, "learning_rate": 4.852862430836744e-05, "loss": 0.0245, "step": 5340 }, { "grad_norm": 0.21274635195732117, "learning_rate": 4.836335728217933e-05, "loss": 0.0219, "step": 5350 }, { "grad_norm": 0.22021105885505676, "learning_rate": 4.819810815405379e-05, "loss": 0.0229, "step": 5360 }, { "grad_norm": 0.22876478731632233, "learning_rate": 4.803287873112877e-05, "loss": 0.027, "step": 5370 }, { "grad_norm": 0.25273939967155457, "learning_rate": 4.786767082032681e-05, "loss": 0.0216, "step": 5380 }, { "grad_norm": 0.18715810775756836, "learning_rate": 4.77024862283351e-05, "loss": 0.0205, "step": 5390 }, { "grad_norm": 0.1712145209312439, "learning_rate": 4.753732676158593e-05, "loss": 0.0225, "step": 5400 }, { "grad_norm": 0.2069764882326126, "learning_rate": 4.737219422623672e-05, "loss": 0.0213, "step": 5410 }, { "grad_norm": 0.22927632927894592, "learning_rate": 4.720709042815044e-05, "loss": 0.0199, "step": 5420 }, { "grad_norm": 0.19533266127109528, "learning_rate": 4.704201717287578e-05, "loss": 0.0242, "step": 5430 }, { "grad_norm": 0.29584628343582153, "learning_rate": 4.6876976265627404e-05, "loss": 0.0222, "step": 5440 }, { "grad_norm": 0.2191377580165863, "learning_rate": 4.671196951126626e-05, "loss": 0.0263, "step": 5450 }, { "grad_norm": 0.16778653860092163, "learning_rate": 4.654699871427971e-05, "loss": 0.0213, "step": 5460 }, { "grad_norm": 0.1397499144077301, "learning_rate": 4.6382065678762034e-05, "loss": 0.0236, "step": 5470 }, { "grad_norm": 0.16911858320236206, "learning_rate": 4.6217172208394424e-05, "loss": 0.0227, "step": 5480 }, { "grad_norm": 0.19770485162734985, "learning_rate": 4.605232010642549e-05, "loss": 0.0176, "step": 5490 }, { "grad_norm": 0.1593470573425293, "learning_rate": 4.588751117565142e-05, "loss": 0.0197, "step": 5500 }, { "grad_norm": 0.1721043437719345, "learning_rate": 4.5722747218396214e-05, "loss": 0.0232, "step": 5510 }, { "grad_norm": 0.16751620173454285, "learning_rate": 4.5558030036492194e-05, "loss": 0.0175, "step": 5520 }, { "grad_norm": 0.21106623113155365, "learning_rate": 4.539336143125999e-05, "loss": 0.0251, "step": 5530 }, { "grad_norm": 0.22368739545345306, "learning_rate": 4.522874320348916e-05, "loss": 0.0211, "step": 5540 }, { "grad_norm": 0.12836293876171112, "learning_rate": 4.506417715341821e-05, "loss": 0.0176, "step": 5550 }, { "grad_norm": 0.1723211109638214, "learning_rate": 4.489966508071511e-05, "loss": 0.0206, "step": 5560 }, { "grad_norm": 0.2214130312204361, "learning_rate": 4.4735208784457575e-05, "loss": 0.0186, "step": 5570 }, { "grad_norm": 0.19197236001491547, "learning_rate": 4.457081006311325e-05, "loss": 0.0228, "step": 5580 }, { "grad_norm": 0.14758263528347015, "learning_rate": 4.440647071452027e-05, "loss": 0.0206, "step": 5590 }, { "grad_norm": 0.20461399853229523, "learning_rate": 4.424219253586737e-05, "loss": 0.0245, "step": 5600 }, { "grad_norm": 0.18232451379299164, "learning_rate": 4.407797732367443e-05, "loss": 0.0258, "step": 5610 }, { "grad_norm": 0.20427417755126953, "learning_rate": 4.391382687377268e-05, "loss": 0.0214, "step": 5620 }, { "grad_norm": 0.2036174088716507, "learning_rate": 4.374974298128512e-05, "loss": 0.0186, "step": 5630 }, { "grad_norm": 0.24022623896598816, "learning_rate": 4.358572744060699e-05, "loss": 0.0262, "step": 5640 }, { "grad_norm": 0.23279570043087006, "learning_rate": 4.342178204538588e-05, "loss": 0.0159, "step": 5650 }, { "grad_norm": 0.18340960144996643, "learning_rate": 4.325790858850241e-05, "loss": 0.0199, "step": 5660 }, { "grad_norm": 0.16987384855747223, "learning_rate": 4.309410886205043e-05, "loss": 0.0223, "step": 5670 }, { "grad_norm": 0.18575528264045715, "learning_rate": 4.293038465731752e-05, "loss": 0.019, "step": 5680 }, { "grad_norm": 0.23115025460720062, "learning_rate": 4.276673776476533e-05, "loss": 0.0232, "step": 5690 }, { "grad_norm": 0.13905857503414154, "learning_rate": 4.260316997401007e-05, "loss": 0.0173, "step": 5700 }, { "grad_norm": 0.24076686799526215, "learning_rate": 4.243968307380293e-05, "loss": 0.0237, "step": 5710 }, { "grad_norm": 0.22415952384471893, "learning_rate": 4.22762788520104e-05, "loss": 0.0232, "step": 5720 }, { "grad_norm": 0.17500349879264832, "learning_rate": 4.211295909559491e-05, "loss": 0.0242, "step": 5730 }, { "grad_norm": 0.17348487675189972, "learning_rate": 4.194972559059511e-05, "loss": 0.0183, "step": 5740 }, { "grad_norm": 0.1685681790113449, "learning_rate": 4.178658012210651e-05, "loss": 0.0206, "step": 5750 }, { "grad_norm": 0.19496312737464905, "learning_rate": 4.162352447426177e-05, "loss": 0.0195, "step": 5760 }, { "grad_norm": 0.18542470037937164, "learning_rate": 4.146056043021135e-05, "loss": 0.022, "step": 5770 }, { "grad_norm": 0.22575882077217102, "learning_rate": 4.1297689772103944e-05, "loss": 0.023, "step": 5780 }, { "grad_norm": 0.13409879803657532, "learning_rate": 4.113491428106694e-05, "loss": 0.02, "step": 5790 }, { "grad_norm": 0.1499057412147522, "learning_rate": 4.0972235737187055e-05, "loss": 0.0202, "step": 5800 }, { "grad_norm": 0.22806742787361145, "learning_rate": 4.080965591949076e-05, "loss": 0.0188, "step": 5810 }, { "grad_norm": 0.2341807335615158, "learning_rate": 4.0647176605924924e-05, "loss": 0.0184, "step": 5820 }, { "grad_norm": 0.33441999554634094, "learning_rate": 4.0484799573337255e-05, "loss": 0.0243, "step": 5830 }, { "grad_norm": 0.1586579531431198, "learning_rate": 4.032252659745699e-05, "loss": 0.0216, "step": 5840 }, { "grad_norm": 0.1978447288274765, "learning_rate": 4.016035945287539e-05, "loss": 0.0279, "step": 5850 }, { "grad_norm": 0.20895929634571075, "learning_rate": 3.999829991302635e-05, "loss": 0.0227, "step": 5860 }, { "grad_norm": 0.21069805324077606, "learning_rate": 3.983634975016707e-05, "loss": 0.0186, "step": 5870 }, { "grad_norm": 0.18085743486881256, "learning_rate": 3.967451073535854e-05, "loss": 0.0294, "step": 5880 }, { "grad_norm": 0.16747790575027466, "learning_rate": 3.951278463844633e-05, "loss": 0.0194, "step": 5890 }, { "grad_norm": 0.2697499096393585, "learning_rate": 3.935117322804111e-05, "loss": 0.0236, "step": 5900 }, { "grad_norm": 0.1996539831161499, "learning_rate": 3.918967827149938e-05, "loss": 0.0224, "step": 5910 }, { "grad_norm": 0.2189987301826477, "learning_rate": 3.9028301534904094e-05, "loss": 0.0221, "step": 5920 }, { "grad_norm": 0.19814705848693848, "learning_rate": 3.88670447830454e-05, "loss": 0.0188, "step": 5930 }, { "grad_norm": 0.14466208219528198, "learning_rate": 3.870590977940132e-05, "loss": 0.0195, "step": 5940 }, { "grad_norm": 0.1508033275604248, "learning_rate": 3.8544898286118404e-05, "loss": 0.0199, "step": 5950 }, { "grad_norm": 0.25172799825668335, "learning_rate": 3.838401206399257e-05, "loss": 0.0178, "step": 5960 }, { "grad_norm": 0.2094907909631729, "learning_rate": 3.822325287244975e-05, "loss": 0.0174, "step": 5970 }, { "grad_norm": 0.21287955343723297, "learning_rate": 3.8062622469526725e-05, "loss": 0.0192, "step": 5980 }, { "grad_norm": 0.2298693209886551, "learning_rate": 3.790212261185183e-05, "loss": 0.0182, "step": 5990 }, { "grad_norm": 0.2104862481355667, "learning_rate": 3.7741755054625794e-05, "loss": 0.0207, "step": 6000 }, { "grad_norm": 0.26982924342155457, "learning_rate": 3.758152155160255e-05, "loss": 0.0176, "step": 6010 }, { "grad_norm": 0.18295921385288239, "learning_rate": 3.742142385506999e-05, "loss": 0.0203, "step": 6020 }, { "grad_norm": 0.16817566752433777, "learning_rate": 3.72614637158309e-05, "loss": 0.015, "step": 6030 }, { "grad_norm": 0.16283245384693146, "learning_rate": 3.710164288318371e-05, "loss": 0.0205, "step": 6040 }, { "grad_norm": 0.17177154123783112, "learning_rate": 3.694196310490345e-05, "loss": 0.0154, "step": 6050 }, { "grad_norm": 0.2619665265083313, "learning_rate": 3.678242612722259e-05, "loss": 0.0207, "step": 6060 }, { "grad_norm": 0.2173738330602646, "learning_rate": 3.6623033694811953e-05, "loss": 0.0159, "step": 6070 }, { "grad_norm": 0.14981365203857422, "learning_rate": 3.6463787550761665e-05, "loss": 0.0212, "step": 6080 }, { "grad_norm": 0.15248903632164001, "learning_rate": 3.630468943656202e-05, "loss": 0.0187, "step": 6090 }, { "grad_norm": 0.22192512452602386, "learning_rate": 3.6145741092084523e-05, "loss": 0.0167, "step": 6100 }, { "grad_norm": 0.18760988116264343, "learning_rate": 3.598694425556278e-05, "loss": 0.0176, "step": 6110 }, { "grad_norm": 0.18520580232143402, "learning_rate": 3.58283006635736e-05, "loss": 0.0223, "step": 6120 }, { "grad_norm": 0.15981748700141907, "learning_rate": 3.566981205101781e-05, "loss": 0.0212, "step": 6130 }, { "grad_norm": 0.16330958902835846, "learning_rate": 3.5511480151101556e-05, "loss": 0.0154, "step": 6140 }, { "grad_norm": 0.17073288559913635, "learning_rate": 3.5353306695317104e-05, "loss": 0.0166, "step": 6150 }, { "grad_norm": 0.17583197355270386, "learning_rate": 3.519529341342402e-05, "loss": 0.0232, "step": 6160 }, { "grad_norm": 0.17077790200710297, "learning_rate": 3.503744203343026e-05, "loss": 0.0199, "step": 6170 }, { "grad_norm": 0.20763163268566132, "learning_rate": 3.487975428157318e-05, "loss": 0.0218, "step": 6180 }, { "grad_norm": 0.20854730904102325, "learning_rate": 3.472223188230083e-05, "loss": 0.0222, "step": 6190 }, { "grad_norm": 0.19341234862804413, "learning_rate": 3.4564876558252866e-05, "loss": 0.0214, "step": 6200 }, { "grad_norm": 0.18530413508415222, "learning_rate": 3.440769003024195e-05, "loss": 0.0204, "step": 6210 }, { "grad_norm": 0.15070335566997528, "learning_rate": 3.425067401723477e-05, "loss": 0.0176, "step": 6220 }, { "grad_norm": 0.17049677670001984, "learning_rate": 3.409383023633325e-05, "loss": 0.0185, "step": 6230 }, { "grad_norm": 0.206292524933815, "learning_rate": 3.3937160402755894e-05, "loss": 0.0173, "step": 6240 }, { "grad_norm": 0.2206525355577469, "learning_rate": 3.378066622981885e-05, "loss": 0.0194, "step": 6250 }, { "grad_norm": 0.23360085487365723, "learning_rate": 3.362434942891738e-05, "loss": 0.0219, "step": 6260 }, { "grad_norm": 0.16536471247673035, "learning_rate": 3.346821170950693e-05, "loss": 0.0175, "step": 6270 }, { "grad_norm": 0.13412506878376007, "learning_rate": 3.3312254779084585e-05, "loss": 0.0201, "step": 6280 }, { "grad_norm": 0.12446487694978714, "learning_rate": 3.315648034317039e-05, "loss": 0.0217, "step": 6290 }, { "grad_norm": 0.2169237732887268, "learning_rate": 3.3000890105288564e-05, "loss": 0.0177, "step": 6300 }, { "grad_norm": 0.15866364538669586, "learning_rate": 3.284548576694908e-05, "loss": 0.0191, "step": 6310 }, { "grad_norm": 0.14268960058689117, "learning_rate": 3.2690269027628815e-05, "loss": 0.0157, "step": 6320 }, { "grad_norm": 0.23870660364627838, "learning_rate": 3.253524158475324e-05, "loss": 0.0163, "step": 6330 }, { "grad_norm": 0.1867639720439911, "learning_rate": 3.238040513367757e-05, "loss": 0.0209, "step": 6340 }, { "grad_norm": 0.19012349843978882, "learning_rate": 3.222576136766843e-05, "loss": 0.0143, "step": 6350 }, { "grad_norm": 0.187286838889122, "learning_rate": 3.2071311977885324e-05, "loss": 0.0178, "step": 6360 }, { "grad_norm": 0.2175855189561844, "learning_rate": 3.191705865336197e-05, "loss": 0.0221, "step": 6370 }, { "grad_norm": 0.17796802520751953, "learning_rate": 3.1763003080988075e-05, "loss": 0.0165, "step": 6380 }, { "grad_norm": 0.16998334228992462, "learning_rate": 3.160914694549063e-05, "loss": 0.0192, "step": 6390 }, { "grad_norm": 0.17466141283512115, "learning_rate": 3.145549192941573e-05, "loss": 0.018, "step": 6400 }, { "grad_norm": 0.21003904938697815, "learning_rate": 3.130203971310999e-05, "loss": 0.0167, "step": 6410 }, { "grad_norm": 0.1871408224105835, "learning_rate": 3.114879197470225e-05, "loss": 0.0197, "step": 6420 }, { "grad_norm": 0.14272694289684296, "learning_rate": 3.0995750390085285e-05, "loss": 0.0162, "step": 6430 }, { "grad_norm": 0.1589004248380661, "learning_rate": 3.084291663289728e-05, "loss": 0.0186, "step": 6440 }, { "grad_norm": 0.1706986278295517, "learning_rate": 3.069029237450375e-05, "loss": 0.0145, "step": 6450 }, { "grad_norm": 0.2153434157371521, "learning_rate": 3.053787928397911e-05, "loss": 0.0164, "step": 6460 }, { "grad_norm": 0.33653825521469116, "learning_rate": 3.0385679028088526e-05, "loss": 0.0219, "step": 6470 }, { "grad_norm": 0.22163210809230804, "learning_rate": 3.023369327126959e-05, "loss": 0.0189, "step": 6480 }, { "grad_norm": 0.187764972448349, "learning_rate": 3.0081923675614198e-05, "loss": 0.0152, "step": 6490 }, { "grad_norm": 0.2183144986629486, "learning_rate": 2.993037190085034e-05, "loss": 0.0158, "step": 6500 }, { "grad_norm": 0.16626107692718506, "learning_rate": 2.977903960432392e-05, "loss": 0.0189, "step": 6510 }, { "grad_norm": 0.1924884021282196, "learning_rate": 2.9627928440980722e-05, "loss": 0.023, "step": 6520 }, { "grad_norm": 0.20345288515090942, "learning_rate": 2.9477040063348183e-05, "loss": 0.0172, "step": 6530 }, { "grad_norm": 0.11487789452075958, "learning_rate": 2.9326376121517456e-05, "loss": 0.016, "step": 6540 }, { "grad_norm": 0.11667237430810928, "learning_rate": 2.9175938263125236e-05, "loss": 0.0161, "step": 6550 }, { "grad_norm": 0.13387639820575714, "learning_rate": 2.9025728133335873e-05, "loss": 0.0178, "step": 6560 }, { "grad_norm": 0.17790280282497406, "learning_rate": 2.8875747374823288e-05, "loss": 0.026, "step": 6570 }, { "grad_norm": 0.19654490053653717, "learning_rate": 2.872599762775298e-05, "loss": 0.0246, "step": 6580 }, { "grad_norm": 0.18908821046352386, "learning_rate": 2.857648052976425e-05, "loss": 0.0172, "step": 6590 }, { "grad_norm": 0.11827342212200165, "learning_rate": 2.8427197715952047e-05, "loss": 0.0173, "step": 6600 }, { "grad_norm": 0.20089268684387207, "learning_rate": 2.8278150818849393e-05, "loss": 0.0175, "step": 6610 }, { "grad_norm": 0.17044885456562042, "learning_rate": 2.812934146840922e-05, "loss": 0.0159, "step": 6620 }, { "grad_norm": 0.2770296037197113, "learning_rate": 2.7980771291986764e-05, "loss": 0.0222, "step": 6630 }, { "grad_norm": 0.17394866049289703, "learning_rate": 2.783244191432167e-05, "loss": 0.0178, "step": 6640 }, { "grad_norm": 0.2376834601163864, "learning_rate": 2.768435495752022e-05, "loss": 0.0167, "step": 6650 }, { "grad_norm": 0.18354684114456177, "learning_rate": 2.753651204103771e-05, "loss": 0.0165, "step": 6660 }, { "grad_norm": 0.17763622105121613, "learning_rate": 2.7388914781660523e-05, "loss": 0.0172, "step": 6670 }, { "grad_norm": 0.15997667610645294, "learning_rate": 2.7241564793488693e-05, "loss": 0.0193, "step": 6680 }, { "grad_norm": 0.21509884297847748, "learning_rate": 2.7094463687918037e-05, "loss": 0.0173, "step": 6690 }, { "grad_norm": 0.2478136569261551, "learning_rate": 2.694761307362268e-05, "loss": 0.0152, "step": 6700 }, { "grad_norm": 0.19748015701770782, "learning_rate": 2.6801014556537467e-05, "loss": 0.0156, "step": 6710 }, { "grad_norm": 0.2463788241147995, "learning_rate": 2.6654669739840243e-05, "loss": 0.0164, "step": 6720 }, { "grad_norm": 0.260615736246109, "learning_rate": 2.650858022393451e-05, "loss": 0.0188, "step": 6730 }, { "grad_norm": 0.13328254222869873, "learning_rate": 2.6362747606431747e-05, "loss": 0.0203, "step": 6740 }, { "grad_norm": 0.23007121682167053, "learning_rate": 2.6217173482134172e-05, "loss": 0.017, "step": 6750 }, { "grad_norm": 0.19547820091247559, "learning_rate": 2.6071859443017044e-05, "loss": 0.0149, "step": 6760 }, { "grad_norm": 0.17939288914203644, "learning_rate": 2.5926807078211414e-05, "loss": 0.0144, "step": 6770 }, { "grad_norm": 0.15607008337974548, "learning_rate": 2.5782017973986728e-05, "loss": 0.0155, "step": 6780 }, { "grad_norm": 0.14892666041851044, "learning_rate": 2.5637493713733374e-05, "loss": 0.0159, "step": 6790 }, { "grad_norm": 0.1887890249490738, "learning_rate": 2.549323587794559e-05, "loss": 0.024, "step": 6800 }, { "grad_norm": 0.20272813737392426, "learning_rate": 2.5349246044203895e-05, "loss": 0.0142, "step": 6810 }, { "grad_norm": 0.20968875288963318, "learning_rate": 2.520552578715808e-05, "loss": 0.0152, "step": 6820 }, { "grad_norm": 0.15231072902679443, "learning_rate": 2.506207667850981e-05, "loss": 0.0139, "step": 6830 }, { "grad_norm": 0.23315376043319702, "learning_rate": 2.4918900286995555e-05, "loss": 0.0155, "step": 6840 }, { "grad_norm": 0.11334145069122314, "learning_rate": 2.4775998178369458e-05, "loss": 0.0142, "step": 6850 }, { "grad_norm": 0.2368345409631729, "learning_rate": 2.4633371915386017e-05, "loss": 0.021, "step": 6860 }, { "grad_norm": 0.1593906730413437, "learning_rate": 2.4491023057783235e-05, "loss": 0.0196, "step": 6870 }, { "grad_norm": 0.24972380697727203, "learning_rate": 2.4348953162265375e-05, "loss": 0.0233, "step": 6880 }, { "grad_norm": 0.12066525965929031, "learning_rate": 2.420716378248607e-05, "loss": 0.0197, "step": 6890 }, { "grad_norm": 0.12942016124725342, "learning_rate": 2.4065656469031266e-05, "loss": 0.0167, "step": 6900 }, { "grad_norm": 0.19439370930194855, "learning_rate": 2.3924432769402268e-05, "loss": 0.0174, "step": 6910 }, { "grad_norm": 0.21388225257396698, "learning_rate": 2.3783494227998844e-05, "loss": 0.0227, "step": 6920 }, { "grad_norm": 0.17400234937667847, "learning_rate": 2.3642842386102264e-05, "loss": 0.0151, "step": 6930 }, { "grad_norm": 0.2021106779575348, "learning_rate": 2.3502478781858567e-05, "loss": 0.0135, "step": 6940 }, { "grad_norm": 0.18549764156341553, "learning_rate": 2.3362404950261628e-05, "loss": 0.0169, "step": 6950 }, { "grad_norm": 0.16581761837005615, "learning_rate": 2.3222622423136458e-05, "loss": 0.0157, "step": 6960 }, { "grad_norm": 0.2570318877696991, "learning_rate": 2.3083132729122332e-05, "loss": 0.0157, "step": 6970 }, { "grad_norm": 0.14701613783836365, "learning_rate": 2.294393739365621e-05, "loss": 0.0158, "step": 6980 }, { "grad_norm": 0.16766297817230225, "learning_rate": 2.2805037938956e-05, "loss": 0.0188, "step": 6990 }, { "grad_norm": 0.17437700927257538, "learning_rate": 2.266643588400386e-05, "loss": 0.0145, "step": 7000 }, { "grad_norm": 0.1380719691514969, "learning_rate": 2.252813274452969e-05, "loss": 0.0161, "step": 7010 }, { "grad_norm": 0.14155253767967224, "learning_rate": 2.2390130032994427e-05, "loss": 0.0116, "step": 7020 }, { "grad_norm": 0.1739863008260727, "learning_rate": 2.2252429258573633e-05, "loss": 0.0159, "step": 7030 }, { "grad_norm": 0.11838425695896149, "learning_rate": 2.2115031927140904e-05, "loss": 0.018, "step": 7040 }, { "grad_norm": 0.18357719480991364, "learning_rate": 2.1977939541251463e-05, "loss": 0.0209, "step": 7050 }, { "grad_norm": 0.1741139143705368, "learning_rate": 2.1841153600125684e-05, "loss": 0.0126, "step": 7060 }, { "grad_norm": 0.15315450727939606, "learning_rate": 2.170467559963267e-05, "loss": 0.017, "step": 7070 }, { "grad_norm": 0.11113177984952927, "learning_rate": 2.1568507032273982e-05, "loss": 0.0132, "step": 7080 }, { "grad_norm": 0.11522191762924194, "learning_rate": 2.1432649387167264e-05, "loss": 0.0143, "step": 7090 }, { "grad_norm": 0.15075652301311493, "learning_rate": 2.1297104150029973e-05, "loss": 0.0147, "step": 7100 }, { "grad_norm": 0.09744875878095627, "learning_rate": 2.116187280316307e-05, "loss": 0.0157, "step": 7110 }, { "grad_norm": 0.1728011816740036, "learning_rate": 2.1026956825434908e-05, "loss": 0.0176, "step": 7120 }, { "grad_norm": 0.1996811032295227, "learning_rate": 2.0892357692265017e-05, "loss": 0.015, "step": 7130 }, { "grad_norm": 0.1624578833580017, "learning_rate": 2.0758076875607947e-05, "loss": 0.0194, "step": 7140 }, { "grad_norm": 0.16990268230438232, "learning_rate": 2.0624115843937207e-05, "loss": 0.0169, "step": 7150 }, { "grad_norm": 0.18371562659740448, "learning_rate": 2.0490476062229157e-05, "loss": 0.0193, "step": 7160 }, { "grad_norm": 0.13989897072315216, "learning_rate": 2.035715899194704e-05, "loss": 0.0159, "step": 7170 }, { "grad_norm": 0.11404819041490555, "learning_rate": 2.022416609102499e-05, "loss": 0.0142, "step": 7180 }, { "grad_norm": 0.1869916319847107, "learning_rate": 2.009149881385205e-05, "loss": 0.0217, "step": 7190 }, { "grad_norm": 0.16575314104557037, "learning_rate": 1.995915861125634e-05, "loss": 0.0158, "step": 7200 }, { "grad_norm": 0.19422870874404907, "learning_rate": 1.9827146930489065e-05, "loss": 0.0134, "step": 7210 }, { "grad_norm": 0.2006867378950119, "learning_rate": 1.9695465215208848e-05, "loss": 0.0212, "step": 7220 }, { "grad_norm": 0.14603765308856964, "learning_rate": 1.9564114905465813e-05, "loss": 0.0142, "step": 7230 }, { "grad_norm": 0.116471067070961, "learning_rate": 1.9433097437685936e-05, "loss": 0.0156, "step": 7240 }, { "grad_norm": 0.16552303731441498, "learning_rate": 1.930241424465521e-05, "loss": 0.0132, "step": 7250 }, { "grad_norm": 0.1597685068845749, "learning_rate": 1.9172066755504115e-05, "loss": 0.0164, "step": 7260 }, { "grad_norm": 0.2048540860414505, "learning_rate": 1.9042056395691914e-05, "loss": 0.0167, "step": 7270 }, { "grad_norm": 0.10383892804384232, "learning_rate": 1.8912384586991066e-05, "loss": 0.013, "step": 7280 }, { "grad_norm": 0.15818072855472565, "learning_rate": 1.8783052747471717e-05, "loss": 0.0133, "step": 7290 }, { "grad_norm": 0.09440428018569946, "learning_rate": 1.865406229148611e-05, "loss": 0.0157, "step": 7300 }, { "grad_norm": 0.12263920903205872, "learning_rate": 1.8525414629653233e-05, "loss": 0.0174, "step": 7310 }, { "grad_norm": 0.178569033741951, "learning_rate": 1.8397111168843255e-05, "loss": 0.0146, "step": 7320 }, { "grad_norm": 0.21510741114616394, "learning_rate": 1.8269153312162323e-05, "loss": 0.0176, "step": 7330 }, { "grad_norm": 0.17760683596134186, "learning_rate": 1.8141542458937054e-05, "loss": 0.0141, "step": 7340 }, { "grad_norm": 0.11459831893444061, "learning_rate": 1.8014280004699268e-05, "loss": 0.0135, "step": 7350 }, { "grad_norm": 0.25474002957344055, "learning_rate": 1.788736734117078e-05, "loss": 0.0128, "step": 7360 }, { "grad_norm": 0.20476779341697693, "learning_rate": 1.7760805856248152e-05, "loss": 0.0145, "step": 7370 }, { "grad_norm": 0.14919638633728027, "learning_rate": 1.7634596933987518e-05, "loss": 0.0159, "step": 7380 }, { "grad_norm": 0.10745817422866821, "learning_rate": 1.7508741954589404e-05, "loss": 0.0167, "step": 7390 }, { "grad_norm": 0.12611889839172363, "learning_rate": 1.7383242294383717e-05, "loss": 0.0145, "step": 7400 }, { "grad_norm": 0.21467800438404083, "learning_rate": 1.7258099325814632e-05, "loss": 0.0118, "step": 7410 }, { "grad_norm": 0.1465645730495453, "learning_rate": 1.7133314417425594e-05, "loss": 0.0161, "step": 7420 }, { "grad_norm": 0.1166311576962471, "learning_rate": 1.7008888933844408e-05, "loss": 0.0131, "step": 7430 }, { "grad_norm": 0.14885519444942474, "learning_rate": 1.6884824235768172e-05, "loss": 0.0165, "step": 7440 }, { "grad_norm": 0.1303434520959854, "learning_rate": 1.6761121679948592e-05, "loss": 0.0129, "step": 7450 }, { "grad_norm": 0.12084059417247772, "learning_rate": 1.663778261917695e-05, "loss": 0.0166, "step": 7460 }, { "grad_norm": 0.16177496314048767, "learning_rate": 1.651480840226952e-05, "loss": 0.0147, "step": 7470 }, { "grad_norm": 0.2134389579296112, "learning_rate": 1.639220037405258e-05, "loss": 0.0162, "step": 7480 }, { "grad_norm": 0.1597629189491272, "learning_rate": 1.6269959875347906e-05, "loss": 0.0151, "step": 7490 }, { "grad_norm": 0.16108790040016174, "learning_rate": 1.614808824295802e-05, "loss": 0.0132, "step": 7500 }, { "grad_norm": 0.2026890218257904, "learning_rate": 1.602658680965152e-05, "loss": 0.012, "step": 7510 }, { "grad_norm": 0.13074465095996857, "learning_rate": 1.5905456904148686e-05, "loss": 0.0176, "step": 7520 }, { "grad_norm": 0.1662781983613968, "learning_rate": 1.57846998511067e-05, "loss": 0.0168, "step": 7530 }, { "grad_norm": 0.14472974836826324, "learning_rate": 1.566431697110538e-05, "loss": 0.0126, "step": 7540 }, { "grad_norm": 0.14783832430839539, "learning_rate": 1.554430958063259e-05, "loss": 0.0127, "step": 7550 }, { "grad_norm": 0.2390289604663849, "learning_rate": 1.5424678992069912e-05, "loss": 0.0143, "step": 7560 }, { "grad_norm": 0.23600034415721893, "learning_rate": 1.5305426513678362e-05, "loss": 0.0147, "step": 7570 }, { "grad_norm": 0.17943447828292847, "learning_rate": 1.518655344958388e-05, "loss": 0.018, "step": 7580 }, { "grad_norm": 0.233170285820961, "learning_rate": 1.5068061099763275e-05, "loss": 0.0159, "step": 7590 }, { "grad_norm": 0.17150527238845825, "learning_rate": 1.494995076002988e-05, "loss": 0.0162, "step": 7600 }, { "grad_norm": 0.24240827560424805, "learning_rate": 1.4832223722019456e-05, "loss": 0.0114, "step": 7610 }, { "grad_norm": 0.10364258289337158, "learning_rate": 1.4714881273176035e-05, "loss": 0.013, "step": 7620 }, { "grad_norm": 0.10322117060422897, "learning_rate": 1.4597924696737835e-05, "loss": 0.0141, "step": 7630 }, { "grad_norm": 0.13579167425632477, "learning_rate": 1.4481355271723252e-05, "loss": 0.0126, "step": 7640 }, { "grad_norm": 0.16095535457134247, "learning_rate": 1.4365174272916809e-05, "loss": 0.019, "step": 7650 }, { "grad_norm": 0.18823829293251038, "learning_rate": 1.4249382970855319e-05, "loss": 0.0139, "step": 7660 }, { "grad_norm": 0.18771116435527802, "learning_rate": 1.4133982631813903e-05, "loss": 0.0164, "step": 7670 }, { "grad_norm": 0.13662363588809967, "learning_rate": 1.4018974517792194e-05, "loss": 0.0118, "step": 7680 }, { "grad_norm": 0.11648961156606674, "learning_rate": 1.390435988650048e-05, "loss": 0.0133, "step": 7690 }, { "grad_norm": 0.18967899680137634, "learning_rate": 1.3790139991346006e-05, "loss": 0.014, "step": 7700 }, { "grad_norm": 0.2753754258155823, "learning_rate": 1.367631608141926e-05, "loss": 0.0149, "step": 7710 }, { "grad_norm": 0.13524499535560608, "learning_rate": 1.3562889401480278e-05, "loss": 0.0126, "step": 7720 }, { "grad_norm": 0.0912032425403595, "learning_rate": 1.3449861191945074e-05, "loss": 0.0174, "step": 7730 }, { "grad_norm": 0.1640395224094391, "learning_rate": 1.3337232688872009e-05, "loss": 0.0139, "step": 7740 }, { "grad_norm": 0.12938587367534637, "learning_rate": 1.3225005123948364e-05, "loss": 0.0143, "step": 7750 }, { "grad_norm": 0.20532652735710144, "learning_rate": 1.311317972447681e-05, "loss": 0.0134, "step": 7760 }, { "grad_norm": 0.15617311000823975, "learning_rate": 1.3001757713361996e-05, "loss": 0.0104, "step": 7770 }, { "grad_norm": 0.22980402410030365, "learning_rate": 1.2890740309097204e-05, "loss": 0.0157, "step": 7780 }, { "grad_norm": 0.1644316464662552, "learning_rate": 1.2780128725750944e-05, "loss": 0.0114, "step": 7790 }, { "grad_norm": 0.1280447244644165, "learning_rate": 1.266992417295379e-05, "loss": 0.0148, "step": 7800 }, { "grad_norm": 0.20253905653953552, "learning_rate": 1.2560127855885073e-05, "loss": 0.0155, "step": 7810 }, { "grad_norm": 0.18069352209568024, "learning_rate": 1.2450740975259745e-05, "loss": 0.0181, "step": 7820 }, { "grad_norm": 0.1316608041524887, "learning_rate": 1.234176472731517e-05, "loss": 0.0215, "step": 7830 }, { "grad_norm": 0.21290437877178192, "learning_rate": 1.2233200303798158e-05, "loss": 0.0143, "step": 7840 }, { "grad_norm": 0.12037481367588043, "learning_rate": 1.2125048891951846e-05, "loss": 0.0129, "step": 7850 }, { "grad_norm": 0.10183481872081757, "learning_rate": 1.2017311674502745e-05, "loss": 0.0116, "step": 7860 }, { "grad_norm": 0.13539209961891174, "learning_rate": 1.1909989829647822e-05, "loss": 0.0129, "step": 7870 }, { "grad_norm": 0.1954246610403061, "learning_rate": 1.1803084531041553e-05, "loss": 0.0139, "step": 7880 }, { "grad_norm": 0.20504838228225708, "learning_rate": 1.1696596947783162e-05, "loss": 0.0135, "step": 7890 }, { "grad_norm": 0.1564154475927353, "learning_rate": 1.1590528244403803e-05, "loss": 0.0132, "step": 7900 }, { "grad_norm": 0.2089209407567978, "learning_rate": 1.148487958085382e-05, "loss": 0.0188, "step": 7910 }, { "grad_norm": 0.11926914751529694, "learning_rate": 1.1379652112490086e-05, "loss": 0.0125, "step": 7920 }, { "grad_norm": 0.2154795378446579, "learning_rate": 1.1274846990063315e-05, "loss": 0.0154, "step": 7930 }, { "grad_norm": 0.11410186439752579, "learning_rate": 1.117046535970554e-05, "loss": 0.013, "step": 7940 }, { "grad_norm": 0.13840824365615845, "learning_rate": 1.106650836291755e-05, "loss": 0.0188, "step": 7950 }, { "grad_norm": 0.1824643313884735, "learning_rate": 1.0962977136556418e-05, "loss": 0.0191, "step": 7960 }, { "grad_norm": 0.17118573188781738, "learning_rate": 1.0859872812823024e-05, "loss": 0.015, "step": 7970 }, { "grad_norm": 0.305040568113327, "learning_rate": 1.0757196519249747e-05, "loss": 0.0209, "step": 7980 }, { "grad_norm": 0.0827808678150177, "learning_rate": 1.0654949378688077e-05, "loss": 0.0141, "step": 7990 }, { "grad_norm": 0.11013500392436981, "learning_rate": 1.0553132509296376e-05, "loss": 0.0143, "step": 8000 }, { "grad_norm": 0.1727442443370819, "learning_rate": 1.0451747024527613e-05, "loss": 0.0142, "step": 8010 }, { "grad_norm": 0.08842010796070099, "learning_rate": 1.0350794033117189e-05, "loss": 0.0143, "step": 8020 }, { "grad_norm": 0.1054481714963913, "learning_rate": 1.0250274639070856e-05, "loss": 0.0163, "step": 8030 }, { "grad_norm": 0.12533679604530334, "learning_rate": 1.0150189941652599e-05, "loss": 0.0119, "step": 8040 }, { "grad_norm": 0.17964580655097961, "learning_rate": 1.0050541035372635e-05, "loss": 0.0118, "step": 8050 }, { "grad_norm": 0.14818887412548065, "learning_rate": 9.951329009975458e-06, "loss": 0.0121, "step": 8060 }, { "grad_norm": 0.14204365015029907, "learning_rate": 9.852554950427845e-06, "loss": 0.0127, "step": 8070 }, { "grad_norm": 0.13218338787555695, "learning_rate": 9.754219936907105e-06, "loss": 0.0132, "step": 8080 }, { "grad_norm": 0.11850284785032272, "learning_rate": 9.656325044789194e-06, "loss": 0.0108, "step": 8090 }, { "grad_norm": 0.17401188611984253, "learning_rate": 9.55887134463697e-06, "loss": 0.0107, "step": 8100 }, { "grad_norm": 0.14256110787391663, "learning_rate": 9.461859902188475e-06, "loss": 0.0141, "step": 8110 }, { "grad_norm": 0.1347450166940689, "learning_rate": 9.365291778345303e-06, "loss": 0.013, "step": 8120 }, { "grad_norm": 0.07804623991250992, "learning_rate": 9.269168029160991e-06, "loss": 0.0093, "step": 8130 }, { "grad_norm": 0.20732715725898743, "learning_rate": 9.173489705829447e-06, "loss": 0.014, "step": 8140 }, { "grad_norm": 0.1815032660961151, "learning_rate": 9.078257854673516e-06, "loss": 0.0138, "step": 8150 }, { "grad_norm": 0.09171808511018753, "learning_rate": 8.983473517133429e-06, "loss": 0.0121, "step": 8160 }, { "grad_norm": 0.15266616642475128, "learning_rate": 8.889137729755537e-06, "loss": 0.0117, "step": 8170 }, { "grad_norm": 0.13094878196716309, "learning_rate": 8.79525152418087e-06, "loss": 0.0144, "step": 8180 }, { "grad_norm": 0.1509494036436081, "learning_rate": 8.701815927133961e-06, "loss": 0.0128, "step": 8190 }, { "grad_norm": 0.1392393708229065, "learning_rate": 8.608831960411534e-06, "loss": 0.0121, "step": 8200 }, { "grad_norm": 0.13514764606952667, "learning_rate": 8.516300640871321e-06, "loss": 0.0121, "step": 8210 }, { "grad_norm": 0.10357410460710526, "learning_rate": 8.424222980421038e-06, "loss": 0.0144, "step": 8220 }, { "grad_norm": 0.1921055167913437, "learning_rate": 8.332599986007184e-06, "loss": 0.0136, "step": 8230 }, { "grad_norm": 0.11742070317268372, "learning_rate": 8.241432659604203e-06, "loss": 0.0104, "step": 8240 }, { "grad_norm": 0.15381766855716705, "learning_rate": 8.150721998203331e-06, "loss": 0.0098, "step": 8250 }, { "grad_norm": 0.12187135219573975, "learning_rate": 8.06046899380184e-06, "loss": 0.013, "step": 8260 }, { "grad_norm": 0.24860753118991852, "learning_rate": 7.970674633392133e-06, "loss": 0.0176, "step": 8270 }, { "grad_norm": 0.14835579693317413, "learning_rate": 7.881339898950924e-06, "loss": 0.0126, "step": 8280 }, { "grad_norm": 0.1429685652256012, "learning_rate": 7.792465767428597e-06, "loss": 0.0124, "step": 8290 }, { "grad_norm": 0.13771310448646545, "learning_rate": 7.704053210738376e-06, "loss": 0.0095, "step": 8300 }, { "grad_norm": 0.16650184988975525, "learning_rate": 7.6161031957458494e-06, "loss": 0.0167, "step": 8310 }, { "grad_norm": 0.1301259845495224, "learning_rate": 7.5286166842582605e-06, "loss": 0.0139, "step": 8320 }, { "grad_norm": 0.15172867476940155, "learning_rate": 7.4415946330140814e-06, "loss": 0.0116, "step": 8330 }, { "grad_norm": 0.12610171735286713, "learning_rate": 7.3550379936725644e-06, "loss": 0.0116, "step": 8340 }, { "grad_norm": 0.20198939740657806, "learning_rate": 7.2689477128032035e-06, "loss": 0.0133, "step": 8350 }, { "grad_norm": 0.12228169292211533, "learning_rate": 7.183324731875551e-06, "loss": 0.0113, "step": 8360 }, { "grad_norm": 0.22058534622192383, "learning_rate": 7.098169987248782e-06, "loss": 0.0096, "step": 8370 }, { "grad_norm": 0.08793634921312332, "learning_rate": 7.013484410161553e-06, "loss": 0.0135, "step": 8380 }, { "grad_norm": 0.1392078697681427, "learning_rate": 6.92926892672176e-06, "loss": 0.0161, "step": 8390 }, { "grad_norm": 0.09317264705896378, "learning_rate": 6.845524457896446e-06, "loss": 0.011, "step": 8400 }, { "grad_norm": 0.11677222698926926, "learning_rate": 6.7622519195017165e-06, "loss": 0.0119, "step": 8410 }, { "grad_norm": 0.09971068054437637, "learning_rate": 6.679452222192684e-06, "loss": 0.0116, "step": 8420 }, { "grad_norm": 0.15154249966144562, "learning_rate": 6.597126271453579e-06, "loss": 0.0108, "step": 8430 }, { "grad_norm": 0.14583270251750946, "learning_rate": 6.51527496758782e-06, "loss": 0.011, "step": 8440 }, { "grad_norm": 0.10512930899858475, "learning_rate": 6.433899205708155e-06, "loss": 0.0108, "step": 8450 }, { "grad_norm": 0.11432874202728271, "learning_rate": 6.352999875726856e-06, "loss": 0.009, "step": 8460 }, { "grad_norm": 0.21080876886844635, "learning_rate": 6.272577862346052e-06, "loss": 0.0106, "step": 8470 }, { "grad_norm": 0.19748060405254364, "learning_rate": 6.192634045047996e-06, "loss": 0.0195, "step": 8480 }, { "grad_norm": 0.16188432276248932, "learning_rate": 6.113169298085458e-06, "loss": 0.014, "step": 8490 }, { "grad_norm": 0.16545750200748444, "learning_rate": 6.034184490472195e-06, "loss": 0.0108, "step": 8500 }, { "grad_norm": 0.16926530003547668, "learning_rate": 5.955680485973386e-06, "loss": 0.0117, "step": 8510 }, { "grad_norm": 0.16979670524597168, "learning_rate": 5.877658143096265e-06, "loss": 0.0127, "step": 8520 }, { "grad_norm": 0.14450082182884216, "learning_rate": 5.800118315080661e-06, "loss": 0.0106, "step": 8530 }, { "grad_norm": 0.1763642281293869, "learning_rate": 5.723061849889716e-06, "loss": 0.0101, "step": 8540 }, { "grad_norm": 0.16128256916999817, "learning_rate": 5.646489590200604e-06, "loss": 0.0139, "step": 8550 }, { "grad_norm": 0.1535434126853943, "learning_rate": 5.570402373395256e-06, "loss": 0.0107, "step": 8560 }, { "grad_norm": 0.1365794837474823, "learning_rate": 5.494801031551305e-06, "loss": 0.0141, "step": 8570 }, { "grad_norm": 0.13815127313137054, "learning_rate": 5.41968639143291e-06, "loss": 0.0109, "step": 8580 }, { "grad_norm": 0.16882140934467316, "learning_rate": 5.345059274481751e-06, "loss": 0.0116, "step": 8590 }, { "grad_norm": 0.13646328449249268, "learning_rate": 5.270920496808002e-06, "loss": 0.0164, "step": 8600 }, { "grad_norm": 0.18481159210205078, "learning_rate": 5.1972708691814695e-06, "loss": 0.0139, "step": 8610 }, { "grad_norm": 0.10143931955099106, "learning_rate": 5.124111197022674e-06, "loss": 0.0143, "step": 8620 }, { "grad_norm": 0.1860588639974594, "learning_rate": 5.051442280394081e-06, "loss": 0.0114, "step": 8630 }, { "grad_norm": 0.14322184026241302, "learning_rate": 4.979264913991322e-06, "loss": 0.0114, "step": 8640 }, { "grad_norm": 0.14105692505836487, "learning_rate": 4.907579887134489e-06, "loss": 0.0149, "step": 8650 }, { "grad_norm": 0.17398186028003693, "learning_rate": 4.836387983759572e-06, "loss": 0.0117, "step": 8660 }, { "grad_norm": 0.15849202871322632, "learning_rate": 4.765689982409816e-06, "loss": 0.0121, "step": 8670 }, { "grad_norm": 0.1484318971633911, "learning_rate": 4.695486656227233e-06, "loss": 0.0165, "step": 8680 }, { "grad_norm": 0.11097422987222672, "learning_rate": 4.625778772944156e-06, "loss": 0.0112, "step": 8690 }, { "grad_norm": 0.100319042801857, "learning_rate": 4.556567094874825e-06, "loss": 0.0104, "step": 8700 }, { "grad_norm": 0.12142813205718994, "learning_rate": 4.487852378907059e-06, "loss": 0.0098, "step": 8710 }, { "grad_norm": 0.1289249211549759, "learning_rate": 4.419635376493986e-06, "loss": 0.0107, "step": 8720 }, { "grad_norm": 0.1285937875509262, "learning_rate": 4.351916833645825e-06, "loss": 0.0142, "step": 8730 }, { "grad_norm": 0.21277806162834167, "learning_rate": 4.284697490921691e-06, "loss": 0.0122, "step": 8740 }, { "grad_norm": 0.18576200306415558, "learning_rate": 4.2179780834215585e-06, "loss": 0.0125, "step": 8750 }, { "grad_norm": 0.21574409306049347, "learning_rate": 4.151759340778178e-06, "loss": 0.0151, "step": 8760 }, { "grad_norm": 0.1464664340019226, "learning_rate": 4.086041987149109e-06, "loss": 0.0117, "step": 8770 }, { "grad_norm": 0.1805085688829422, "learning_rate": 4.020826741208811e-06, "loss": 0.0123, "step": 8780 }, { "grad_norm": 0.197200745344162, "learning_rate": 3.956114316140746e-06, "loss": 0.0148, "step": 8790 }, { "grad_norm": 0.16501112282276154, "learning_rate": 3.891905419629643e-06, "loss": 0.0123, "step": 8800 }, { "grad_norm": 0.10883435606956482, "learning_rate": 3.8282007538536946e-06, "loss": 0.0163, "step": 8810 }, { "grad_norm": 0.2673315703868866, "learning_rate": 3.7650010154769265e-06, "loss": 0.0139, "step": 8820 }, { "grad_norm": 0.1797867864370346, "learning_rate": 3.7023068956415608e-06, "loss": 0.0123, "step": 8830 }, { "grad_norm": 0.17199642956256866, "learning_rate": 3.6401190799604303e-06, "loss": 0.0091, "step": 8840 }, { "grad_norm": 0.13386261463165283, "learning_rate": 3.578438248509536e-06, "loss": 0.012, "step": 8850 }, { "grad_norm": 0.13790687918663025, "learning_rate": 3.5172650758205583e-06, "loss": 0.0134, "step": 8860 }, { "grad_norm": 0.09710917621850967, "learning_rate": 3.45660023087353e-06, "loss": 0.0095, "step": 8870 }, { "grad_norm": 0.1823161095380783, "learning_rate": 3.3964443770894528e-06, "loss": 0.0121, "step": 8880 }, { "grad_norm": 0.10127899795770645, "learning_rate": 3.3367981723231245e-06, "loss": 0.0096, "step": 8890 }, { "grad_norm": 0.1882726103067398, "learning_rate": 3.2776622688558746e-06, "loss": 0.014, "step": 8900 }, { "grad_norm": 0.1851186901330948, "learning_rate": 3.2190373133884677e-06, "loss": 0.0094, "step": 8910 }, { "grad_norm": 0.1470097452402115, "learning_rate": 3.1609239470340446e-06, "loss": 0.0105, "step": 8920 }, { "grad_norm": 0.08990567922592163, "learning_rate": 3.1033228053110373e-06, "loss": 0.009, "step": 8930 }, { "grad_norm": 0.11339270323514938, "learning_rate": 3.0462345181363314e-06, "loss": 0.0104, "step": 8940 }, { "grad_norm": 0.10732753574848175, "learning_rate": 2.9896597098182654e-06, "loss": 0.0135, "step": 8950 }, { "grad_norm": 0.15574195981025696, "learning_rate": 2.933598999049891e-06, "loss": 0.0129, "step": 8960 }, { "grad_norm": 0.087925024330616, "learning_rate": 2.8780529989021697e-06, "loss": 0.0096, "step": 8970 }, { "grad_norm": 0.1065230667591095, "learning_rate": 2.823022316817242e-06, "loss": 0.0144, "step": 8980 }, { "grad_norm": 0.09873256832361221, "learning_rate": 2.7685075546018456e-06, "loss": 0.0122, "step": 8990 }, { "grad_norm": 0.127954363822937, "learning_rate": 2.7145093084206598e-06, "loss": 0.0172, "step": 9000 }, { "grad_norm": 0.12901662290096283, "learning_rate": 2.661028168789892e-06, "loss": 0.0115, "step": 9010 }, { "grad_norm": 0.16218611598014832, "learning_rate": 2.6080647205706855e-06, "loss": 0.0094, "step": 9020 }, { "grad_norm": 0.0849599838256836, "learning_rate": 2.555619542962834e-06, "loss": 0.0098, "step": 9030 }, { "grad_norm": 0.12140654772520065, "learning_rate": 2.503693209498409e-06, "loss": 0.0099, "step": 9040 }, { "grad_norm": 0.13725794851779938, "learning_rate": 2.452286288035449e-06, "loss": 0.0122, "step": 9050 }, { "grad_norm": 0.09011051058769226, "learning_rate": 2.4013993407518363e-06, "loss": 0.0127, "step": 9060 }, { "grad_norm": 0.13556823134422302, "learning_rate": 2.351032924139063e-06, "loss": 0.0114, "step": 9070 }, { "grad_norm": 0.13893501460552216, "learning_rate": 2.30118758899619e-06, "loss": 0.0119, "step": 9080 }, { "grad_norm": 0.07889826595783234, "learning_rate": 2.2518638804238157e-06, "loss": 0.0102, "step": 9090 }, { "grad_norm": 0.09053882211446762, "learning_rate": 2.203062337818118e-06, "loss": 0.0114, "step": 9100 }, { "grad_norm": 0.1259828358888626, "learning_rate": 2.1547834948649483e-06, "loss": 0.013, "step": 9110 }, { "grad_norm": 0.15469038486480713, "learning_rate": 2.1070278795340017e-06, "loss": 0.0104, "step": 9120 }, { "grad_norm": 0.10178510844707489, "learning_rate": 2.059796014073029e-06, "loss": 0.0073, "step": 9130 }, { "grad_norm": 0.17118637263774872, "learning_rate": 2.01308841500214e-06, "loss": 0.0115, "step": 9140 }, { "grad_norm": 0.17713071405887604, "learning_rate": 1.9669055931081704e-06, "loss": 0.0162, "step": 9150 }, { "grad_norm": 0.11722289025783539, "learning_rate": 1.9212480534390507e-06, "loss": 0.0105, "step": 9160 }, { "grad_norm": 0.17232359945774078, "learning_rate": 1.8761162952983246e-06, "loss": 0.0132, "step": 9170 }, { "grad_norm": 0.15275292098522186, "learning_rate": 1.8315108122396618e-06, "loss": 0.0117, "step": 9180 }, { "grad_norm": 0.08959384262561798, "learning_rate": 1.787432092061475e-06, "loss": 0.0095, "step": 9190 }, { "grad_norm": 0.0914398580789566, "learning_rate": 1.743880616801602e-06, "loss": 0.0128, "step": 9200 }, { "grad_norm": 0.14314721524715424, "learning_rate": 1.7008568627319865e-06, "loss": 0.0112, "step": 9210 }, { "grad_norm": 0.14288179576396942, "learning_rate": 1.6583613003535226e-06, "loss": 0.0146, "step": 9220 }, { "grad_norm": 0.14377747476100922, "learning_rate": 1.6163943943908522e-06, "loss": 0.0105, "step": 9230 }, { "grad_norm": 0.15320050716400146, "learning_rate": 1.5749566037873476e-06, "loss": 0.0116, "step": 9240 }, { "grad_norm": 0.1702478677034378, "learning_rate": 1.5340483817000428e-06, "loss": 0.013, "step": 9250 }, { "grad_norm": 0.10916151851415634, "learning_rate": 1.4936701754947101e-06, "loss": 0.011, "step": 9260 }, { "grad_norm": 0.09552150219678879, "learning_rate": 1.4538224267409361e-06, "loss": 0.0116, "step": 9270 }, { "grad_norm": 0.08339200913906097, "learning_rate": 1.414505571207314e-06, "loss": 0.0085, "step": 9280 }, { "grad_norm": 0.11374591290950775, "learning_rate": 1.3757200388566816e-06, "loss": 0.0105, "step": 9290 }, { "grad_norm": 0.25294533371925354, "learning_rate": 1.3374662538414074e-06, "loss": 0.0124, "step": 9300 }, { "grad_norm": 0.11946651339530945, "learning_rate": 1.2997446344987617e-06, "loss": 0.0106, "step": 9310 }, { "grad_norm": 0.11976730823516846, "learning_rate": 1.262555593346315e-06, "loss": 0.0116, "step": 9320 }, { "grad_norm": 0.23559392988681793, "learning_rate": 1.2258995370774685e-06, "loss": 0.0123, "step": 9330 }, { "grad_norm": 0.12007373571395874, "learning_rate": 1.1897768665569798e-06, "loss": 0.0122, "step": 9340 }, { "grad_norm": 0.09195871651172638, "learning_rate": 1.1541879768165954e-06, "loss": 0.0089, "step": 9350 }, { "grad_norm": 0.13674812018871307, "learning_rate": 1.1191332570507085e-06, "loss": 0.0126, "step": 9360 }, { "grad_norm": 0.11076989769935608, "learning_rate": 1.0846130906121132e-06, "loss": 0.0163, "step": 9370 }, { "grad_norm": 0.15767857432365417, "learning_rate": 1.0506278550078131e-06, "loss": 0.0137, "step": 9380 }, { "grad_norm": 0.1805792897939682, "learning_rate": 1.0171779218949185e-06, "loss": 0.0116, "step": 9390 }, { "grad_norm": 0.11813852190971375, "learning_rate": 9.842636570765174e-07, "loss": 0.0085, "step": 9400 }, { "grad_norm": 0.1467324048280716, "learning_rate": 9.518854204977612e-07, "loss": 0.0098, "step": 9410 }, { "grad_norm": 0.10415437072515488, "learning_rate": 9.200435662418349e-07, "loss": 0.011, "step": 9420 }, { "grad_norm": 0.13425996899604797, "learning_rate": 8.887384425261658e-07, "loss": 0.0111, "step": 9430 }, { "grad_norm": 0.13137972354888916, "learning_rate": 8.579703916985648e-07, "loss": 0.0141, "step": 9440 }, { "grad_norm": 0.08584709465503693, "learning_rate": 8.277397502335194e-07, "loss": 0.0104, "step": 9450 }, { "grad_norm": 0.187907874584198, "learning_rate": 7.980468487284675e-07, "loss": 0.0138, "step": 9460 }, { "grad_norm": 0.10054634511470795, "learning_rate": 7.688920119002297e-07, "loss": 0.0114, "step": 9470 }, { "grad_norm": 0.09902559220790863, "learning_rate": 7.402755585814269e-07, "loss": 0.0114, "step": 9480 }, { "grad_norm": 0.15654601156711578, "learning_rate": 7.121978017170073e-07, "loss": 0.0139, "step": 9490 }, { "grad_norm": 0.1847333014011383, "learning_rate": 6.846590483608306e-07, "loss": 0.0133, "step": 9500 }, { "grad_norm": 0.1458956003189087, "learning_rate": 6.576595996722834e-07, "loss": 0.0114, "step": 9510 }, { "grad_norm": 0.18196991086006165, "learning_rate": 6.311997509130141e-07, "loss": 0.0143, "step": 9520 }, { "grad_norm": 0.11081382632255554, "learning_rate": 6.052797914436803e-07, "loss": 0.0126, "step": 9530 }, { "grad_norm": 0.11541574448347092, "learning_rate": 5.799000047208181e-07, "loss": 0.0152, "step": 9540 }, { "grad_norm": 0.17209942638874054, "learning_rate": 5.550606682937054e-07, "loss": 0.0131, "step": 9550 }, { "grad_norm": 0.12133768200874329, "learning_rate": 5.307620538013481e-07, "loss": 0.0145, "step": 9560 }, { "grad_norm": 0.1505117565393448, "learning_rate": 5.070044269694874e-07, "loss": 0.0114, "step": 9570 }, { "grad_norm": 0.14388573169708252, "learning_rate": 4.837880476077417e-07, "loss": 0.0119, "step": 9580 }, { "grad_norm": 0.06610798835754395, "learning_rate": 4.6111316960670835e-07, "loss": 0.0087, "step": 9590 }, { "grad_norm": 0.22115370631217957, "learning_rate": 4.389800409352218e-07, "loss": 0.0134, "step": 9600 }, { "grad_norm": 0.18885287642478943, "learning_rate": 4.173889036376277e-07, "loss": 0.0147, "step": 9610 }, { "grad_norm": 0.13984602689743042, "learning_rate": 3.963399938311463e-07, "loss": 0.0133, "step": 9620 }, { "grad_norm": 0.11732786893844604, "learning_rate": 3.7583354170328545e-07, "loss": 0.0131, "step": 9630 }, { "grad_norm": 0.09076925367116928, "learning_rate": 3.558697715093207e-07, "loss": 0.0088, "step": 9640 }, { "grad_norm": 0.26557376980781555, "learning_rate": 3.3644890156983576e-07, "loss": 0.0134, "step": 9650 }, { "grad_norm": 0.11162034422159195, "learning_rate": 3.175711442683638e-07, "loss": 0.012, "step": 9660 }, { "grad_norm": 0.11327049136161804, "learning_rate": 2.9923670604902197e-07, "loss": 0.0111, "step": 9670 }, { "grad_norm": 0.16933433711528778, "learning_rate": 2.814457874143028e-07, "loss": 0.0127, "step": 9680 }, { "grad_norm": 0.11748065054416656, "learning_rate": 2.641985829228366e-07, "loss": 0.0094, "step": 9690 }, { "grad_norm": 0.12200001627206802, "learning_rate": 2.474952811872877e-07, "loss": 0.0102, "step": 9700 }, { "grad_norm": 0.17700079083442688, "learning_rate": 2.3133606487228397e-07, "loss": 0.0146, "step": 9710 }, { "grad_norm": 0.11939152330160141, "learning_rate": 2.157211106924295e-07, "loss": 0.0134, "step": 9720 }, { "grad_norm": 0.1380860060453415, "learning_rate": 2.006505894103672e-07, "loss": 0.013, "step": 9730 }, { "grad_norm": 0.11468132585287094, "learning_rate": 1.8612466583489696e-07, "loss": 0.0104, "step": 9740 }, { "grad_norm": 0.11605285108089447, "learning_rate": 1.7214349881918834e-07, "loss": 0.0106, "step": 9750 }, { "grad_norm": 0.11929124593734741, "learning_rate": 1.5870724125904845e-07, "loss": 0.0095, "step": 9760 }, { "grad_norm": 0.11584172397851944, "learning_rate": 1.4581604009124006e-07, "loss": 0.011, "step": 9770 }, { "grad_norm": 0.1461670994758606, "learning_rate": 1.334700362918717e-07, "loss": 0.0119, "step": 9780 }, { "grad_norm": 0.10434766113758087, "learning_rate": 1.2166936487486015e-07, "loss": 0.0136, "step": 9790 }, { "grad_norm": 0.12522736191749573, "learning_rate": 1.1041415489045914e-07, "loss": 0.009, "step": 9800 }, { "grad_norm": 0.12443745881319046, "learning_rate": 9.970452942384412e-08, "loss": 0.0116, "step": 9810 }, { "grad_norm": 0.14427554607391357, "learning_rate": 8.954060559375754e-08, "loss": 0.0084, "step": 9820 }, { "grad_norm": 0.10228047519922256, "learning_rate": 7.99224945512489e-08, "loss": 0.0164, "step": 9830 }, { "grad_norm": 0.11378460377454758, "learning_rate": 7.085030147843675e-08, "loss": 0.0128, "step": 9840 }, { "grad_norm": 0.1570298671722412, "learning_rate": 6.232412558736523e-08, "loss": 0.0129, "step": 9850 }, { "grad_norm": 0.1602962166070938, "learning_rate": 5.434406011893822e-08, "loss": 0.0121, "step": 9860 }, { "grad_norm": 0.1493944674730301, "learning_rate": 4.6910192341864664e-08, "loss": 0.0104, "step": 9870 }, { "grad_norm": 0.1174878180027008, "learning_rate": 4.0022603551737035e-08, "loss": 0.0131, "step": 9880 }, { "grad_norm": 0.14381143450737, "learning_rate": 3.3681369070120985e-08, "loss": 0.0154, "step": 9890 }, { "grad_norm": 0.13277053833007812, "learning_rate": 2.7886558243744866e-08, "loss": 0.0114, "step": 9900 }, { "grad_norm": 0.15353800356388092, "learning_rate": 2.2638234443722596e-08, "loss": 0.0133, "step": 9910 }, { "grad_norm": 0.1338016390800476, "learning_rate": 1.7936455064887504e-08, "loss": 0.0135, "step": 9920 }, { "grad_norm": 0.08933908492326736, "learning_rate": 1.378127152514841e-08, "loss": 0.0079, "step": 9930 }, { "grad_norm": 0.09656242281198502, "learning_rate": 1.0172729264917857e-08, "loss": 0.0146, "step": 9940 }, { "grad_norm": 0.2050967961549759, "learning_rate": 7.1108677466458215e-09, "loss": 0.012, "step": 9950 }, { "grad_norm": 0.12416277080774307, "learning_rate": 4.595720454353414e-09, "loss": 0.012, "step": 9960 }, { "grad_norm": 0.2629949152469635, "learning_rate": 2.627314893294264e-09, "loss": 0.0118, "step": 9970 }, { "grad_norm": 0.1364159882068634, "learning_rate": 1.2056725896270048e-09, "loss": 0.0111, "step": 9980 }, { "grad_norm": 0.14430800080299377, "learning_rate": 3.308090902098826e-10, "loss": 0.0135, "step": 9990 }, { "grad_norm": 0.14944520592689514, "learning_rate": 2.7339624120159555e-12, "loss": 0.0147, "step": 10000 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 5000, "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": 32, "trial_name": null, "trial_params": null }