{ "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": 2.5159618854522705, "learning_rate": 1.8e-06, "loss": 1.1483, "step": 10 }, { "grad_norm": 1.229854702949524, "learning_rate": 3.8e-06, "loss": 1.1304, "step": 20 }, { "grad_norm": 0.5788406133651733, "learning_rate": 5.8e-06, "loss": 1.0995, "step": 30 }, { "grad_norm": 0.5237457752227783, "learning_rate": 7.8e-06, "loss": 1.1111, "step": 40 }, { "grad_norm": 0.6178592443466187, "learning_rate": 9.800000000000001e-06, "loss": 1.087, "step": 50 }, { "grad_norm": 0.5519043803215027, "learning_rate": 1.18e-05, "loss": 1.0591, "step": 60 }, { "grad_norm": 0.4879607856273651, "learning_rate": 1.3800000000000002e-05, "loss": 1.0801, "step": 70 }, { "grad_norm": 0.44365695118904114, "learning_rate": 1.58e-05, "loss": 1.0609, "step": 80 }, { "grad_norm": 0.45549723505973816, "learning_rate": 1.78e-05, "loss": 1.0609, "step": 90 }, { "grad_norm": 0.5034562349319458, "learning_rate": 1.9800000000000004e-05, "loss": 1.0492, "step": 100 }, { "grad_norm": 0.5842446088790894, "learning_rate": 2.18e-05, "loss": 1.0367, "step": 110 }, { "grad_norm": 0.5683846473693848, "learning_rate": 2.38e-05, "loss": 1.0385, "step": 120 }, { "grad_norm": 0.4842475950717926, "learning_rate": 2.58e-05, "loss": 1.0403, "step": 130 }, { "grad_norm": 0.5306119322776794, "learning_rate": 2.7800000000000005e-05, "loss": 1.0189, "step": 140 }, { "grad_norm": 1.9963287115097046, "learning_rate": 2.98e-05, "loss": 0.9516, "step": 150 }, { "grad_norm": 1.812093734741211, "learning_rate": 3.18e-05, "loss": 0.9092, "step": 160 }, { "grad_norm": 1.4375126361846924, "learning_rate": 3.38e-05, "loss": 0.834, "step": 170 }, { "grad_norm": 2.010869026184082, "learning_rate": 3.58e-05, "loss": 0.7375, "step": 180 }, { "grad_norm": 2.413586139678955, "learning_rate": 3.7800000000000004e-05, "loss": 0.5582, "step": 190 }, { "grad_norm": 1.9026707410812378, "learning_rate": 3.9800000000000005e-05, "loss": 0.4421, "step": 200 }, { "grad_norm": 2.250880718231201, "learning_rate": 4.18e-05, "loss": 0.3566, "step": 210 }, { "grad_norm": 2.1882007122039795, "learning_rate": 4.38e-05, "loss": 0.281, "step": 220 }, { "grad_norm": 2.1763956546783447, "learning_rate": 4.58e-05, "loss": 0.2188, "step": 230 }, { "grad_norm": 1.598118782043457, "learning_rate": 4.78e-05, "loss": 0.1885, "step": 240 }, { "grad_norm": 1.7943470478057861, "learning_rate": 4.9800000000000004e-05, "loss": 0.1683, "step": 250 }, { "grad_norm": 1.1090458631515503, "learning_rate": 5.1800000000000005e-05, "loss": 0.1526, "step": 260 }, { "grad_norm": 1.7006478309631348, "learning_rate": 5.380000000000001e-05, "loss": 0.167, "step": 270 }, { "grad_norm": 1.1285202503204346, "learning_rate": 5.580000000000001e-05, "loss": 0.1539, "step": 280 }, { "grad_norm": 1.2011115550994873, "learning_rate": 5.7799999999999995e-05, "loss": 0.1673, "step": 290 }, { "grad_norm": 1.6653646230697632, "learning_rate": 5.9800000000000003e-05, "loss": 0.1324, "step": 300 }, { "grad_norm": 1.439849615097046, "learning_rate": 6.18e-05, "loss": 0.1499, "step": 310 }, { "grad_norm": 1.2204383611679077, "learning_rate": 6.38e-05, "loss": 0.1346, "step": 320 }, { "grad_norm": 1.333120346069336, "learning_rate": 6.58e-05, "loss": 0.126, "step": 330 }, { "grad_norm": 1.0728588104248047, "learning_rate": 6.780000000000001e-05, "loss": 0.1182, "step": 340 }, { "grad_norm": 1.0763732194900513, "learning_rate": 6.98e-05, "loss": 0.1145, "step": 350 }, { "grad_norm": 1.1664179563522339, "learning_rate": 7.18e-05, "loss": 0.1114, "step": 360 }, { "grad_norm": 0.9064451456069946, "learning_rate": 7.38e-05, "loss": 0.1198, "step": 370 }, { "grad_norm": 0.978127658367157, "learning_rate": 7.58e-05, "loss": 0.1186, "step": 380 }, { "grad_norm": 0.6878655552864075, "learning_rate": 7.780000000000001e-05, "loss": 0.117, "step": 390 }, { "grad_norm": 0.7018166780471802, "learning_rate": 7.98e-05, "loss": 0.1135, "step": 400 }, { "grad_norm": 0.8144962787628174, "learning_rate": 8.18e-05, "loss": 0.1127, "step": 410 }, { "grad_norm": 0.9641146063804626, "learning_rate": 8.38e-05, "loss": 0.1085, "step": 420 }, { "grad_norm": 0.8674544095993042, "learning_rate": 8.58e-05, "loss": 0.1109, "step": 430 }, { "grad_norm": 1.1523188352584839, "learning_rate": 8.78e-05, "loss": 0.1074, "step": 440 }, { "grad_norm": 0.5703819990158081, "learning_rate": 8.98e-05, "loss": 0.1094, "step": 450 }, { "grad_norm": 0.7837931513786316, "learning_rate": 9.180000000000001e-05, "loss": 0.1038, "step": 460 }, { "grad_norm": 0.8676462173461914, "learning_rate": 9.38e-05, "loss": 0.1111, "step": 470 }, { "grad_norm": 0.6141104698181152, "learning_rate": 9.58e-05, "loss": 0.1058, "step": 480 }, { "grad_norm": 0.920661211013794, "learning_rate": 9.78e-05, "loss": 0.0972, "step": 490 }, { "grad_norm": 0.444709837436676, "learning_rate": 9.98e-05, "loss": 0.0974, "step": 500 }, { "grad_norm": 0.7144810557365417, "learning_rate": 9.9999778549206e-05, "loss": 0.1051, "step": 510 }, { "grad_norm": 0.6596009135246277, "learning_rate": 9.999901304280685e-05, "loss": 0.0912, "step": 520 }, { "grad_norm": 0.46732577681541443, "learning_rate": 9.999770075521164e-05, "loss": 0.1004, "step": 530 }, { "grad_norm": 0.7965353727340698, "learning_rate": 9.99958417007713e-05, "loss": 0.095, "step": 540 }, { "grad_norm": 0.6141448020935059, "learning_rate": 9.999343589981615e-05, "loss": 0.0873, "step": 550 }, { "grad_norm": 0.5732608437538147, "learning_rate": 9.999048337865568e-05, "loss": 0.0825, "step": 560 }, { "grad_norm": 0.7064613103866577, "learning_rate": 9.998698416957815e-05, "loss": 0.0857, "step": 570 }, { "grad_norm": 0.6934483647346497, "learning_rate": 9.998293831085037e-05, "loss": 0.0839, "step": 580 }, { "grad_norm": 0.6228412985801697, "learning_rate": 9.997834584671719e-05, "loss": 0.0928, "step": 590 }, { "grad_norm": 0.45418229699134827, "learning_rate": 9.997320682740107e-05, "loss": 0.0878, "step": 600 }, { "grad_norm": 0.5538354516029358, "learning_rate": 9.996752130910149e-05, "loss": 0.1002, "step": 610 }, { "grad_norm": 0.6151629686355591, "learning_rate": 9.99612893539944e-05, "loss": 0.0835, "step": 620 }, { "grad_norm": 0.4220874309539795, "learning_rate": 9.995451103023144e-05, "loss": 0.077, "step": 630 }, { "grad_norm": 0.5480417013168335, "learning_rate": 9.994718641193928e-05, "loss": 0.0751, "step": 640 }, { "grad_norm": 0.550042450428009, "learning_rate": 9.993931557921874e-05, "loss": 0.0799, "step": 650 }, { "grad_norm": 0.5008789300918579, "learning_rate": 9.993089861814402e-05, "loss": 0.0804, "step": 660 }, { "grad_norm": 0.5455918908119202, "learning_rate": 9.992193562076166e-05, "loss": 0.0909, "step": 670 }, { "grad_norm": 0.4892705976963043, "learning_rate": 9.991242668508954e-05, "loss": 0.0828, "step": 680 }, { "grad_norm": 0.5675092935562134, "learning_rate": 9.990237191511587e-05, "loss": 0.0799, "step": 690 }, { "grad_norm": 0.47399237751960754, "learning_rate": 9.989177142079802e-05, "loss": 0.0817, "step": 700 }, { "grad_norm": 0.5346032977104187, "learning_rate": 9.988062531806126e-05, "loss": 0.089, "step": 710 }, { "grad_norm": 0.40706565976142883, "learning_rate": 9.986893372879762e-05, "loss": 0.0725, "step": 720 }, { "grad_norm": 0.5299361348152161, "learning_rate": 9.985669678086443e-05, "loss": 0.0771, "step": 730 }, { "grad_norm": 0.3075653314590454, "learning_rate": 9.984391460808298e-05, "loss": 0.064, "step": 740 }, { "grad_norm": 0.6659806966781616, "learning_rate": 9.983058735023709e-05, "loss": 0.0767, "step": 750 }, { "grad_norm": 0.483489066362381, "learning_rate": 9.98167151530715e-05, "loss": 0.0678, "step": 760 }, { "grad_norm": 0.6130730509757996, "learning_rate": 9.980229816829034e-05, "loss": 0.0694, "step": 770 }, { "grad_norm": 0.5382611155509949, "learning_rate": 9.978733655355544e-05, "loss": 0.0735, "step": 780 }, { "grad_norm": 0.400630384683609, "learning_rate": 9.977183047248464e-05, "loss": 0.0694, "step": 790 }, { "grad_norm": 0.7054753303527832, "learning_rate": 9.975578009464992e-05, "loss": 0.0779, "step": 800 }, { "grad_norm": 0.4207785725593567, "learning_rate": 9.97391855955757e-05, "loss": 0.0801, "step": 810 }, { "grad_norm": 0.42646199464797974, "learning_rate": 9.972204715673669e-05, "loss": 0.0733, "step": 820 }, { "grad_norm": 0.4028584659099579, "learning_rate": 9.970436496555617e-05, "loss": 0.0683, "step": 830 }, { "grad_norm": 0.7651212811470032, "learning_rate": 9.968613921540373e-05, "loss": 0.0726, "step": 840 }, { "grad_norm": 0.43106693029403687, "learning_rate": 9.966737010559326e-05, "loss": 0.08, "step": 850 }, { "grad_norm": 0.3023202121257782, "learning_rate": 9.964805784138072e-05, "loss": 0.0713, "step": 860 }, { "grad_norm": 0.39201438426971436, "learning_rate": 9.962820263396195e-05, "loss": 0.0695, "step": 870 }, { "grad_norm": 0.45070594549179077, "learning_rate": 9.960780470047033e-05, "loss": 0.0657, "step": 880 }, { "grad_norm": 0.41378459334373474, "learning_rate": 9.958686426397437e-05, "loss": 0.0671, "step": 890 }, { "grad_norm": 0.36704525351524353, "learning_rate": 9.956538155347534e-05, "loss": 0.074, "step": 900 }, { "grad_norm": 0.40086475014686584, "learning_rate": 9.95433568039047e-05, "loss": 0.0603, "step": 910 }, { "grad_norm": 0.41832470893859863, "learning_rate": 9.952079025612162e-05, "loss": 0.074, "step": 920 }, { "grad_norm": 0.39717358350753784, "learning_rate": 9.949768215691022e-05, "loss": 0.0641, "step": 930 }, { "grad_norm": 0.5014980435371399, "learning_rate": 9.9474032758977e-05, "loss": 0.058, "step": 940 }, { "grad_norm": 0.4757887125015259, "learning_rate": 9.944984232094794e-05, "loss": 0.0759, "step": 950 }, { "grad_norm": 0.3897148668766022, "learning_rate": 9.942511110736584e-05, "loss": 0.053, "step": 960 }, { "grad_norm": 0.27132555842399597, "learning_rate": 9.939983938868726e-05, "loss": 0.0562, "step": 970 }, { "grad_norm": 0.4776926636695862, "learning_rate": 9.93740274412797e-05, "loss": 0.0632, "step": 980 }, { "grad_norm": 0.4840758442878723, "learning_rate": 9.934767554741846e-05, "loss": 0.0681, "step": 990 }, { "grad_norm": 0.43904194235801697, "learning_rate": 9.932078399528361e-05, "loss": 0.0614, "step": 1000 }, { "grad_norm": 0.426669716835022, "learning_rate": 9.929335307895689e-05, "loss": 0.0594, "step": 1010 }, { "grad_norm": 0.3193824291229248, "learning_rate": 9.926538309841839e-05, "loss": 0.0606, "step": 1020 }, { "grad_norm": 0.5331795811653137, "learning_rate": 9.923687435954334e-05, "loss": 0.0634, "step": 1030 }, { "grad_norm": 0.38970547914505005, "learning_rate": 9.920782717409873e-05, "loss": 0.0647, "step": 1040 }, { "grad_norm": 0.5359234809875488, "learning_rate": 9.917824185973994e-05, "loss": 0.0568, "step": 1050 }, { "grad_norm": 0.3400614559650421, "learning_rate": 9.914811874000723e-05, "loss": 0.0716, "step": 1060 }, { "grad_norm": 0.414949506521225, "learning_rate": 9.911745814432218e-05, "loss": 0.0617, "step": 1070 }, { "grad_norm": 0.36002618074417114, "learning_rate": 9.90862604079842e-05, "loss": 0.066, "step": 1080 }, { "grad_norm": 0.4223342835903168, "learning_rate": 9.90545258721667e-05, "loss": 0.0511, "step": 1090 }, { "grad_norm": 0.3083857595920563, "learning_rate": 9.90222548839135e-05, "loss": 0.0587, "step": 1100 }, { "grad_norm": 0.3328525125980377, "learning_rate": 9.898944779613495e-05, "loss": 0.0594, "step": 1110 }, { "grad_norm": 0.3414061963558197, "learning_rate": 9.89561049676041e-05, "loss": 0.0598, "step": 1120 }, { "grad_norm": 0.2961043417453766, "learning_rate": 9.89222267629528e-05, "loss": 0.0685, "step": 1130 }, { "grad_norm": 0.3253377676010132, "learning_rate": 9.888781355266763e-05, "loss": 0.0614, "step": 1140 }, { "grad_norm": 0.3054792881011963, "learning_rate": 9.885286571308598e-05, "loss": 0.0576, "step": 1150 }, { "grad_norm": 0.35794559121131897, "learning_rate": 9.881738362639182e-05, "loss": 0.059, "step": 1160 }, { "grad_norm": 0.5221659541130066, "learning_rate": 9.878136768061154e-05, "loss": 0.0565, "step": 1170 }, { "grad_norm": 0.3117104470729828, "learning_rate": 9.874481826960979e-05, "loss": 0.0528, "step": 1180 }, { "grad_norm": 0.38916710019111633, "learning_rate": 9.870773579308503e-05, "loss": 0.0546, "step": 1190 }, { "grad_norm": 0.4094749391078949, "learning_rate": 9.867012065656533e-05, "loss": 0.0588, "step": 1200 }, { "grad_norm": 0.2930634915828705, "learning_rate": 9.863197327140376e-05, "loss": 0.057, "step": 1210 }, { "grad_norm": 0.2838492691516876, "learning_rate": 9.859329405477403e-05, "loss": 0.0505, "step": 1220 }, { "grad_norm": 0.425595223903656, "learning_rate": 9.855408342966585e-05, "loss": 0.0615, "step": 1230 }, { "grad_norm": 0.4051397144794464, "learning_rate": 9.851434182488033e-05, "loss": 0.0555, "step": 1240 }, { "grad_norm": 0.29593825340270996, "learning_rate": 9.84740696750253e-05, "loss": 0.0593, "step": 1250 }, { "grad_norm": 0.32431620359420776, "learning_rate": 9.843326742051055e-05, "loss": 0.0546, "step": 1260 }, { "grad_norm": 0.34110841155052185, "learning_rate": 9.839193550754297e-05, "loss": 0.0629, "step": 1270 }, { "grad_norm": 0.3189084827899933, "learning_rate": 9.835007438812177e-05, "loss": 0.0589, "step": 1280 }, { "grad_norm": 0.306033194065094, "learning_rate": 9.830768452003341e-05, "loss": 0.061, "step": 1290 }, { "grad_norm": 0.3669886291027069, "learning_rate": 9.826476636684671e-05, "loss": 0.0547, "step": 1300 }, { "grad_norm": 0.42673438787460327, "learning_rate": 9.822132039790773e-05, "loss": 0.0544, "step": 1310 }, { "grad_norm": 0.5216823816299438, "learning_rate": 9.817734708833461e-05, "loss": 0.0585, "step": 1320 }, { "grad_norm": 0.464231938123703, "learning_rate": 9.813284691901243e-05, "loss": 0.0579, "step": 1330 }, { "grad_norm": 0.39660927653312683, "learning_rate": 9.808782037658792e-05, "loss": 0.0584, "step": 1340 }, { "grad_norm": 0.33324843645095825, "learning_rate": 9.804226795346411e-05, "loss": 0.0493, "step": 1350 }, { "grad_norm": 0.34605899453163147, "learning_rate": 9.799619014779503e-05, "loss": 0.0545, "step": 1360 }, { "grad_norm": 0.2854861617088318, "learning_rate": 9.794958746348013e-05, "loss": 0.0441, "step": 1370 }, { "grad_norm": 0.37449151277542114, "learning_rate": 9.790246041015896e-05, "loss": 0.0477, "step": 1380 }, { "grad_norm": 0.31299933791160583, "learning_rate": 9.785480950320538e-05, "loss": 0.0493, "step": 1390 }, { "grad_norm": 0.3539699912071228, "learning_rate": 9.78066352637221e-05, "loss": 0.0543, "step": 1400 }, { "grad_norm": 0.42367759346961975, "learning_rate": 9.775793821853488e-05, "loss": 0.0445, "step": 1410 }, { "grad_norm": 0.3910120129585266, "learning_rate": 9.77087189001868e-05, "loss": 0.0532, "step": 1420 }, { "grad_norm": 0.293381929397583, "learning_rate": 9.765897784693243e-05, "loss": 0.0513, "step": 1430 }, { "grad_norm": 0.37230536341667175, "learning_rate": 9.760871560273197e-05, "loss": 0.0514, "step": 1440 }, { "grad_norm": 0.3689688742160797, "learning_rate": 9.755793271724526e-05, "loss": 0.0512, "step": 1450 }, { "grad_norm": 0.3543822169303894, "learning_rate": 9.750662974582584e-05, "loss": 0.0537, "step": 1460 }, { "grad_norm": 0.34433937072753906, "learning_rate": 9.745480724951473e-05, "loss": 0.0522, "step": 1470 }, { "grad_norm": 0.33119767904281616, "learning_rate": 9.740246579503447e-05, "loss": 0.0549, "step": 1480 }, { "grad_norm": 0.31917229294776917, "learning_rate": 9.734960595478284e-05, "loss": 0.0477, "step": 1490 }, { "grad_norm": 0.3343929052352905, "learning_rate": 9.729622830682657e-05, "loss": 0.0556, "step": 1500 }, { "grad_norm": 0.33983084559440613, "learning_rate": 9.724233343489504e-05, "loss": 0.0507, "step": 1510 }, { "grad_norm": 0.4354788661003113, "learning_rate": 9.718792192837396e-05, "loss": 0.0552, "step": 1520 }, { "grad_norm": 0.3280980885028839, "learning_rate": 9.713299438229886e-05, "loss": 0.0569, "step": 1530 }, { "grad_norm": 0.4725235402584076, "learning_rate": 9.707755139734855e-05, "loss": 0.0589, "step": 1540 }, { "grad_norm": 0.29983919858932495, "learning_rate": 9.702159357983866e-05, "loss": 0.0526, "step": 1550 }, { "grad_norm": 0.2426820993423462, "learning_rate": 9.696512154171492e-05, "loss": 0.0493, "step": 1560 }, { "grad_norm": 0.2913431227207184, "learning_rate": 9.690813590054645e-05, "loss": 0.0476, "step": 1570 }, { "grad_norm": 0.3784482181072235, "learning_rate": 9.685063727951914e-05, "loss": 0.0482, "step": 1580 }, { "grad_norm": 0.24481727182865143, "learning_rate": 9.679262630742865e-05, "loss": 0.0493, "step": 1590 }, { "grad_norm": 0.3738441467285156, "learning_rate": 9.673410361867373e-05, "loss": 0.05, "step": 1600 }, { "grad_norm": 0.22309787571430206, "learning_rate": 9.667506985324909e-05, "loss": 0.0421, "step": 1610 }, { "grad_norm": 0.3819243013858795, "learning_rate": 9.661552565673855e-05, "loss": 0.0395, "step": 1620 }, { "grad_norm": 0.3560880422592163, "learning_rate": 9.655547168030789e-05, "loss": 0.0459, "step": 1630 }, { "grad_norm": 0.4970574975013733, "learning_rate": 9.649490858069777e-05, "loss": 0.0499, "step": 1640 }, { "grad_norm": 0.292142391204834, "learning_rate": 9.643383702021658e-05, "loss": 0.056, "step": 1650 }, { "grad_norm": 0.40684160590171814, "learning_rate": 9.637225766673307e-05, "loss": 0.0404, "step": 1660 }, { "grad_norm": 0.4257897436618805, "learning_rate": 9.631017119366922e-05, "loss": 0.0502, "step": 1670 }, { "grad_norm": 0.35576918721199036, "learning_rate": 9.624757827999273e-05, "loss": 0.0501, "step": 1680 }, { "grad_norm": 0.3620152771472931, "learning_rate": 9.618447961020971e-05, "loss": 0.0517, "step": 1690 }, { "grad_norm": 0.35370972752571106, "learning_rate": 9.612087587435707e-05, "loss": 0.0477, "step": 1700 }, { "grad_norm": 0.4586949646472931, "learning_rate": 9.605676776799508e-05, "loss": 0.044, "step": 1710 }, { "grad_norm": 0.3300144672393799, "learning_rate": 9.599215599219973e-05, "loss": 0.0487, "step": 1720 }, { "grad_norm": 0.49373117089271545, "learning_rate": 9.592704125355505e-05, "loss": 0.0473, "step": 1730 }, { "grad_norm": 0.2639038562774658, "learning_rate": 9.586142426414538e-05, "loss": 0.0424, "step": 1740 }, { "grad_norm": 0.3513612747192383, "learning_rate": 9.57953057415476e-05, "loss": 0.0483, "step": 1750 }, { "grad_norm": 0.31420576572418213, "learning_rate": 9.572868640882328e-05, "loss": 0.0498, "step": 1760 }, { "grad_norm": 0.41801580786705017, "learning_rate": 9.56615669945108e-05, "loss": 0.0505, "step": 1770 }, { "grad_norm": 0.2436353713274002, "learning_rate": 9.55939482326173e-05, "loss": 0.0491, "step": 1780 }, { "grad_norm": 0.24478358030319214, "learning_rate": 9.552583086261069e-05, "loss": 0.0525, "step": 1790 }, { "grad_norm": 0.31968528032302856, "learning_rate": 9.545721562941168e-05, "loss": 0.0442, "step": 1800 }, { "grad_norm": 0.35239654779434204, "learning_rate": 9.538810328338543e-05, "loss": 0.0442, "step": 1810 }, { "grad_norm": 0.27907559275627136, "learning_rate": 9.531849458033349e-05, "loss": 0.0437, "step": 1820 }, { "grad_norm": 0.2652444839477539, "learning_rate": 9.524839028148547e-05, "loss": 0.0423, "step": 1830 }, { "grad_norm": 0.3920736610889435, "learning_rate": 9.517779115349077e-05, "loss": 0.0447, "step": 1840 }, { "grad_norm": 0.2938689589500427, "learning_rate": 9.510669796841014e-05, "loss": 0.0399, "step": 1850 }, { "grad_norm": 0.3294620215892792, "learning_rate": 9.503511150370727e-05, "loss": 0.0479, "step": 1860 }, { "grad_norm": 0.35458725690841675, "learning_rate": 9.496303254224024e-05, "loss": 0.0468, "step": 1870 }, { "grad_norm": 0.3131880462169647, "learning_rate": 9.489046187225306e-05, "loss": 0.0425, "step": 1880 }, { "grad_norm": 0.2784838378429413, "learning_rate": 9.481740028736692e-05, "loss": 0.0424, "step": 1890 }, { "grad_norm": 0.32683059573173523, "learning_rate": 9.474384858657164e-05, "loss": 0.0503, "step": 1900 }, { "grad_norm": 0.28824374079704285, "learning_rate": 9.466980757421679e-05, "loss": 0.0465, "step": 1910 }, { "grad_norm": 0.2978738248348236, "learning_rate": 9.459527806000305e-05, "loss": 0.0463, "step": 1920 }, { "grad_norm": 0.3237500786781311, "learning_rate": 9.452026085897325e-05, "loss": 0.0476, "step": 1930 }, { "grad_norm": 0.31228598952293396, "learning_rate": 9.444475679150348e-05, "loss": 0.0463, "step": 1940 }, { "grad_norm": 0.40382838249206543, "learning_rate": 9.436876668329411e-05, "loss": 0.0463, "step": 1950 }, { "grad_norm": 0.30576321482658386, "learning_rate": 9.429229136536079e-05, "loss": 0.0374, "step": 1960 }, { "grad_norm": 0.24859489500522614, "learning_rate": 9.421533167402534e-05, "loss": 0.0493, "step": 1970 }, { "grad_norm": 0.3831899166107178, "learning_rate": 9.413788845090666e-05, "loss": 0.0521, "step": 1980 }, { "grad_norm": 0.39726337790489197, "learning_rate": 9.405996254291136e-05, "loss": 0.0434, "step": 1990 }, { "grad_norm": 0.3918340802192688, "learning_rate": 9.398155480222474e-05, "loss": 0.0429, "step": 2000 }, { "grad_norm": 0.327368825674057, "learning_rate": 9.390266608630128e-05, "loss": 0.0408, "step": 2010 }, { "grad_norm": 0.3698867857456207, "learning_rate": 9.38232972578553e-05, "loss": 0.0444, "step": 2020 }, { "grad_norm": 0.4605638086795807, "learning_rate": 9.374344918485164e-05, "loss": 0.0457, "step": 2030 }, { "grad_norm": 0.2379828691482544, "learning_rate": 9.366312274049602e-05, "loss": 0.0491, "step": 2040 }, { "grad_norm": 0.2519533336162567, "learning_rate": 9.358231880322554e-05, "loss": 0.0412, "step": 2050 }, { "grad_norm": 0.20280233025550842, "learning_rate": 9.350103825669916e-05, "loss": 0.0374, "step": 2060 }, { "grad_norm": 0.42747077345848083, "learning_rate": 9.341928198978787e-05, "loss": 0.038, "step": 2070 }, { "grad_norm": 0.3578861653804779, "learning_rate": 9.333705089656512e-05, "loss": 0.0377, "step": 2080 }, { "grad_norm": 0.40609949827194214, "learning_rate": 9.325434587629698e-05, "loss": 0.0415, "step": 2090 }, { "grad_norm": 0.2940372824668884, "learning_rate": 9.31711678334323e-05, "loss": 0.0387, "step": 2100 }, { "grad_norm": 0.24989618360996246, "learning_rate": 9.308751767759282e-05, "loss": 0.0404, "step": 2110 }, { "grad_norm": 0.3179076910018921, "learning_rate": 9.300339632356325e-05, "loss": 0.0388, "step": 2120 }, { "grad_norm": 0.3278786838054657, "learning_rate": 9.291880469128124e-05, "loss": 0.043, "step": 2130 }, { "grad_norm": 0.40317773818969727, "learning_rate": 9.283374370582732e-05, "loss": 0.044, "step": 2140 }, { "grad_norm": 0.2676919996738434, "learning_rate": 9.274821429741482e-05, "loss": 0.0495, "step": 2150 }, { "grad_norm": 0.30166029930114746, "learning_rate": 9.266221740137961e-05, "loss": 0.0398, "step": 2160 }, { "grad_norm": 0.2604588270187378, "learning_rate": 9.257575395817001e-05, "loss": 0.0389, "step": 2170 }, { "grad_norm": 0.3449733555316925, "learning_rate": 9.248882491333637e-05, "loss": 0.0435, "step": 2180 }, { "grad_norm": 0.33755919337272644, "learning_rate": 9.240143121752076e-05, "loss": 0.0396, "step": 2190 }, { "grad_norm": 0.2523297071456909, "learning_rate": 9.23135738264467e-05, "loss": 0.0408, "step": 2200 }, { "grad_norm": 0.2879725992679596, "learning_rate": 9.222525370090849e-05, "loss": 0.0367, "step": 2210 }, { "grad_norm": 0.31816625595092773, "learning_rate": 9.213647180676088e-05, "loss": 0.0388, "step": 2220 }, { "grad_norm": 0.3764864206314087, "learning_rate": 9.204722911490846e-05, "loss": 0.0454, "step": 2230 }, { "grad_norm": 0.27936139702796936, "learning_rate": 9.1957526601295e-05, "loss": 0.0347, "step": 2240 }, { "grad_norm": 0.46504807472229004, "learning_rate": 9.186736524689281e-05, "loss": 0.0445, "step": 2250 }, { "grad_norm": 0.33609825372695923, "learning_rate": 9.177674603769204e-05, "loss": 0.0423, "step": 2260 }, { "grad_norm": 0.3260129392147064, "learning_rate": 9.168566996468983e-05, "loss": 0.0396, "step": 2270 }, { "grad_norm": 0.3423762321472168, "learning_rate": 9.159413802387951e-05, "loss": 0.0446, "step": 2280 }, { "grad_norm": 0.2547859847545624, "learning_rate": 9.150215121623974e-05, "loss": 0.0397, "step": 2290 }, { "grad_norm": 0.3185064196586609, "learning_rate": 9.140971054772349e-05, "loss": 0.0399, "step": 2300 }, { "grad_norm": 0.27693429589271545, "learning_rate": 9.131681702924713e-05, "loss": 0.0458, "step": 2310 }, { "grad_norm": 0.2701265513896942, "learning_rate": 9.122347167667926e-05, "loss": 0.0356, "step": 2320 }, { "grad_norm": 0.31347665190696716, "learning_rate": 9.112967551082973e-05, "loss": 0.0475, "step": 2330 }, { "grad_norm": 0.26438936591148376, "learning_rate": 9.103542955743835e-05, "loss": 0.0368, "step": 2340 }, { "grad_norm": 0.44931694865226746, "learning_rate": 9.094073484716381e-05, "loss": 0.0423, "step": 2350 }, { "grad_norm": 0.27881142497062683, "learning_rate": 9.084559241557226e-05, "loss": 0.0388, "step": 2360 }, { "grad_norm": 0.3351340889930725, "learning_rate": 9.075000330312608e-05, "loss": 0.0381, "step": 2370 }, { "grad_norm": 0.2496800720691681, "learning_rate": 9.065396855517253e-05, "loss": 0.031, "step": 2380 }, { "grad_norm": 0.41475576162338257, "learning_rate": 9.055748922193219e-05, "loss": 0.0407, "step": 2390 }, { "grad_norm": 0.265460729598999, "learning_rate": 9.046056635848761e-05, "loss": 0.0433, "step": 2400 }, { "grad_norm": 0.2581157982349396, "learning_rate": 9.036320102477169e-05, "loss": 0.0367, "step": 2410 }, { "grad_norm": 0.24478290975093842, "learning_rate": 9.02653942855561e-05, "loss": 0.0361, "step": 2420 }, { "grad_norm": 0.26577359437942505, "learning_rate": 9.016714721043971e-05, "loss": 0.0341, "step": 2430 }, { "grad_norm": 0.28633931279182434, "learning_rate": 9.006846087383675e-05, "loss": 0.0464, "step": 2440 }, { "grad_norm": 0.3152526915073395, "learning_rate": 8.996933635496523e-05, "loss": 0.0371, "step": 2450 }, { "grad_norm": 0.29289478063583374, "learning_rate": 8.986977473783498e-05, "loss": 0.0442, "step": 2460 }, { "grad_norm": 0.3587043583393097, "learning_rate": 8.97697771112359e-05, "loss": 0.0373, "step": 2470 }, { "grad_norm": 0.24506734311580658, "learning_rate": 8.966934456872602e-05, "loss": 0.037, "step": 2480 }, { "grad_norm": 0.2384117692708969, "learning_rate": 8.95684782086195e-05, "loss": 0.0321, "step": 2490 }, { "grad_norm": 0.314713716506958, "learning_rate": 8.946717913397476e-05, "loss": 0.0396, "step": 2500 }, { "grad_norm": 0.19089624285697937, "learning_rate": 8.93654484525822e-05, "loss": 0.0367, "step": 2510 }, { "grad_norm": 0.30495622754096985, "learning_rate": 8.926328727695226e-05, "loss": 0.0369, "step": 2520 }, { "grad_norm": 0.26448914408683777, "learning_rate": 8.916069672430319e-05, "loss": 0.031, "step": 2530 }, { "grad_norm": 0.26216375827789307, "learning_rate": 8.905767791654884e-05, "loss": 0.0367, "step": 2540 }, { "grad_norm": 0.38782116770744324, "learning_rate": 8.895423198028638e-05, "loss": 0.0437, "step": 2550 }, { "grad_norm": 0.3100297749042511, "learning_rate": 8.885036004678402e-05, "loss": 0.0387, "step": 2560 }, { "grad_norm": 0.2529044449329376, "learning_rate": 8.874606325196857e-05, "loss": 0.0368, "step": 2570 }, { "grad_norm": 0.21800687909126282, "learning_rate": 8.864134273641304e-05, "loss": 0.0398, "step": 2580 }, { "grad_norm": 0.2019965797662735, "learning_rate": 8.853619964532427e-05, "loss": 0.0399, "step": 2590 }, { "grad_norm": 0.3156558573246002, "learning_rate": 8.843063512853019e-05, "loss": 0.0373, "step": 2600 }, { "grad_norm": 0.293620228767395, "learning_rate": 8.832465034046749e-05, "loss": 0.0376, "step": 2610 }, { "grad_norm": 0.2764570713043213, "learning_rate": 8.821824644016882e-05, "loss": 0.0367, "step": 2620 }, { "grad_norm": 0.1768033504486084, "learning_rate": 8.811142459125019e-05, "loss": 0.0344, "step": 2630 }, { "grad_norm": 0.2290610671043396, "learning_rate": 8.800418596189822e-05, "loss": 0.0298, "step": 2640 }, { "grad_norm": 0.3569720685482025, "learning_rate": 8.789653172485737e-05, "loss": 0.0403, "step": 2650 }, { "grad_norm": 0.2659302353858948, "learning_rate": 8.778846305741715e-05, "loss": 0.0462, "step": 2660 }, { "grad_norm": 0.3115200996398926, "learning_rate": 8.767998114139918e-05, "loss": 0.0381, "step": 2670 }, { "grad_norm": 0.2642381191253662, "learning_rate": 8.757108716314429e-05, "loss": 0.0413, "step": 2680 }, { "grad_norm": 0.2170819342136383, "learning_rate": 8.746178231349962e-05, "loss": 0.0352, "step": 2690 }, { "grad_norm": 0.2556818425655365, "learning_rate": 8.735206778780549e-05, "loss": 0.0451, "step": 2700 }, { "grad_norm": 0.3087403178215027, "learning_rate": 8.724194478588234e-05, "loss": 0.0403, "step": 2710 }, { "grad_norm": 0.229642853140831, "learning_rate": 8.713141451201772e-05, "loss": 0.0358, "step": 2720 }, { "grad_norm": 0.30421188473701477, "learning_rate": 8.702047817495295e-05, "loss": 0.0381, "step": 2730 }, { "grad_norm": 0.22514581680297852, "learning_rate": 8.69091369878701e-05, "loss": 0.0336, "step": 2740 }, { "grad_norm": 0.2712033689022064, "learning_rate": 8.679739216837849e-05, "loss": 0.0287, "step": 2750 }, { "grad_norm": 0.18467025458812714, "learning_rate": 8.66852449385016e-05, "loss": 0.0336, "step": 2760 }, { "grad_norm": 0.33869704604148865, "learning_rate": 8.657269652466356e-05, "loss": 0.0366, "step": 2770 }, { "grad_norm": 0.24359700083732605, "learning_rate": 8.645974815767577e-05, "loss": 0.0327, "step": 2780 }, { "grad_norm": 0.21421508491039276, "learning_rate": 8.634640107272351e-05, "loss": 0.0275, "step": 2790 }, { "grad_norm": 0.22281470894813538, "learning_rate": 8.623265650935234e-05, "loss": 0.0365, "step": 2800 }, { "grad_norm": 0.3095913827419281, "learning_rate": 8.611851571145456e-05, "loss": 0.0326, "step": 2810 }, { "grad_norm": 0.3226255178451538, "learning_rate": 8.600397992725566e-05, "loss": 0.0318, "step": 2820 }, { "grad_norm": 0.2663918137550354, "learning_rate": 8.588905040930061e-05, "loss": 0.0337, "step": 2830 }, { "grad_norm": 0.21763144433498383, "learning_rate": 8.577372841444022e-05, "loss": 0.0341, "step": 2840 }, { "grad_norm": 0.35835328698158264, "learning_rate": 8.565801520381736e-05, "loss": 0.0323, "step": 2850 }, { "grad_norm": 0.27201056480407715, "learning_rate": 8.554191204285313e-05, "loss": 0.0336, "step": 2860 }, { "grad_norm": 0.2764015793800354, "learning_rate": 8.542542020123315e-05, "loss": 0.0302, "step": 2870 }, { "grad_norm": 0.2368517816066742, "learning_rate": 8.530854095289347e-05, "loss": 0.0286, "step": 2880 }, { "grad_norm": 0.32102054357528687, "learning_rate": 8.519127557600688e-05, "loss": 0.0307, "step": 2890 }, { "grad_norm": 0.19790856540203094, "learning_rate": 8.507362535296871e-05, "loss": 0.0339, "step": 2900 }, { "grad_norm": 0.357888400554657, "learning_rate": 8.495559157038299e-05, "loss": 0.0457, "step": 2910 }, { "grad_norm": 0.286602258682251, "learning_rate": 8.483717551904823e-05, "loss": 0.0341, "step": 2920 }, { "grad_norm": 0.26493215560913086, "learning_rate": 8.47183784939434e-05, "loss": 0.0344, "step": 2930 }, { "grad_norm": 0.18114492297172546, "learning_rate": 8.459920179421374e-05, "loss": 0.0331, "step": 2940 }, { "grad_norm": 0.2885872423648834, "learning_rate": 8.447964672315656e-05, "loss": 0.0327, "step": 2950 }, { "grad_norm": 0.22596657276153564, "learning_rate": 8.435971458820692e-05, "loss": 0.0378, "step": 2960 }, { "grad_norm": 0.36850401759147644, "learning_rate": 8.423940670092345e-05, "loss": 0.0386, "step": 2970 }, { "grad_norm": 0.29241660237312317, "learning_rate": 8.411872437697394e-05, "loss": 0.0329, "step": 2980 }, { "grad_norm": 0.32035908102989197, "learning_rate": 8.399766893612096e-05, "loss": 0.0319, "step": 2990 }, { "grad_norm": 0.3289054036140442, "learning_rate": 8.38762417022074e-05, "loss": 0.0365, "step": 3000 }, { "grad_norm": 0.26399320363998413, "learning_rate": 8.375444400314204e-05, "loss": 0.0354, "step": 3010 }, { "grad_norm": 0.22729003429412842, "learning_rate": 8.3632277170885e-05, "loss": 0.0403, "step": 3020 }, { "grad_norm": 0.22307905554771423, "learning_rate": 8.350974254143318e-05, "loss": 0.0335, "step": 3030 }, { "grad_norm": 0.3093739151954651, "learning_rate": 8.338684145480566e-05, "loss": 0.0303, "step": 3040 }, { "grad_norm": 0.30450448393821716, "learning_rate": 8.326357525502904e-05, "loss": 0.0387, "step": 3050 }, { "grad_norm": 0.24762652814388275, "learning_rate": 8.313994529012273e-05, "loss": 0.0367, "step": 3060 }, { "grad_norm": 0.20990091562271118, "learning_rate": 8.301595291208422e-05, "loss": 0.0377, "step": 3070 }, { "grad_norm": 0.2918902337551117, "learning_rate": 8.289159947687427e-05, "loss": 0.0331, "step": 3080 }, { "grad_norm": 0.2855488955974579, "learning_rate": 8.276688634440216e-05, "loss": 0.0342, "step": 3090 }, { "grad_norm": 0.254739373922348, "learning_rate": 8.26418148785107e-05, "loss": 0.0359, "step": 3100 }, { "grad_norm": 0.22317546606063843, "learning_rate": 8.251638644696141e-05, "loss": 0.0373, "step": 3110 }, { "grad_norm": 0.18980999290943146, "learning_rate": 8.23906024214195e-05, "loss": 0.0318, "step": 3120 }, { "grad_norm": 0.32512664794921875, "learning_rate": 8.226446417743897e-05, "loss": 0.0367, "step": 3130 }, { "grad_norm": 0.2628178298473358, "learning_rate": 8.213797309444742e-05, "loss": 0.0391, "step": 3140 }, { "grad_norm": 0.2750248610973358, "learning_rate": 8.201113055573105e-05, "loss": 0.0379, "step": 3150 }, { "grad_norm": 0.33209916949272156, "learning_rate": 8.188393794841958e-05, "loss": 0.0333, "step": 3160 }, { "grad_norm": 0.2867058217525482, "learning_rate": 8.175639666347094e-05, "loss": 0.0352, "step": 3170 }, { "grad_norm": 0.27029573917388916, "learning_rate": 8.162850809565623e-05, "loss": 0.0339, "step": 3180 }, { "grad_norm": 0.2590360939502716, "learning_rate": 8.150027364354431e-05, "loss": 0.0319, "step": 3190 }, { "grad_norm": 0.2756377160549164, "learning_rate": 8.137169470948662e-05, "loss": 0.0415, "step": 3200 }, { "grad_norm": 0.2657574713230133, "learning_rate": 8.124277269960179e-05, "loss": 0.0357, "step": 3210 }, { "grad_norm": 0.2988937497138977, "learning_rate": 8.111350902376023e-05, "loss": 0.0312, "step": 3220 }, { "grad_norm": 0.20664525032043457, "learning_rate": 8.098390509556883e-05, "loss": 0.0312, "step": 3230 }, { "grad_norm": 0.38841408491134644, "learning_rate": 8.085396233235536e-05, "loss": 0.0288, "step": 3240 }, { "grad_norm": 0.22538471221923828, "learning_rate": 8.072368215515306e-05, "loss": 0.0273, "step": 3250 }, { "grad_norm": 0.1948246955871582, "learning_rate": 8.059306598868506e-05, "loss": 0.0404, "step": 3260 }, { "grad_norm": 0.2921832203865051, "learning_rate": 8.046211526134888e-05, "loss": 0.0331, "step": 3270 }, { "grad_norm": 0.2974282205104828, "learning_rate": 8.033083140520065e-05, "loss": 0.03, "step": 3280 }, { "grad_norm": 0.23619142174720764, "learning_rate": 8.019921585593962e-05, "loss": 0.0316, "step": 3290 }, { "grad_norm": 0.33081120252609253, "learning_rate": 8.006727005289232e-05, "loss": 0.0311, "step": 3300 }, { "grad_norm": 0.22083830833435059, "learning_rate": 7.993499543899692e-05, "loss": 0.0298, "step": 3310 }, { "grad_norm": 0.190445676445961, "learning_rate": 7.980239346078742e-05, "loss": 0.0272, "step": 3320 }, { "grad_norm": 0.2742348611354828, "learning_rate": 7.966946556837778e-05, "loss": 0.0299, "step": 3330 }, { "grad_norm": 0.27944785356521606, "learning_rate": 7.953621321544616e-05, "loss": 0.0311, "step": 3340 }, { "grad_norm": 0.22460418939590454, "learning_rate": 7.940263785921896e-05, "loss": 0.0391, "step": 3350 }, { "grad_norm": 0.22456350922584534, "learning_rate": 7.926874096045482e-05, "loss": 0.0302, "step": 3360 }, { "grad_norm": 0.25177833437919617, "learning_rate": 7.913452398342881e-05, "loss": 0.0412, "step": 3370 }, { "grad_norm": 0.21285733580589294, "learning_rate": 7.89999883959163e-05, "loss": 0.0315, "step": 3380 }, { "grad_norm": 0.33911028504371643, "learning_rate": 7.886513566917687e-05, "loss": 0.0352, "step": 3390 }, { "grad_norm": 0.2258351743221283, "learning_rate": 7.872996727793838e-05, "loss": 0.0347, "step": 3400 }, { "grad_norm": 0.17767034471035004, "learning_rate": 7.859448470038069e-05, "loss": 0.0285, "step": 3410 }, { "grad_norm": 0.2087840884923935, "learning_rate": 7.845868941811956e-05, "loss": 0.03, "step": 3420 }, { "grad_norm": 0.25187647342681885, "learning_rate": 7.832258291619043e-05, "loss": 0.0284, "step": 3430 }, { "grad_norm": 0.32244664430618286, "learning_rate": 7.81861666830322e-05, "loss": 0.0322, "step": 3440 }, { "grad_norm": 0.21062393486499786, "learning_rate": 7.804944221047097e-05, "loss": 0.0281, "step": 3450 }, { "grad_norm": 0.3759423494338989, "learning_rate": 7.791241099370364e-05, "loss": 0.0355, "step": 3460 }, { "grad_norm": 0.2141740918159485, "learning_rate": 7.777507453128163e-05, "loss": 0.0272, "step": 3470 }, { "grad_norm": 0.23246589303016663, "learning_rate": 7.763743432509451e-05, "loss": 0.0285, "step": 3480 }, { "grad_norm": 0.23324766755104065, "learning_rate": 7.749949188035353e-05, "loss": 0.0299, "step": 3490 }, { "grad_norm": 0.2150503247976303, "learning_rate": 7.736124870557516e-05, "loss": 0.0382, "step": 3500 }, { "grad_norm": 0.2522960901260376, "learning_rate": 7.722270631256459e-05, "loss": 0.031, "step": 3510 }, { "grad_norm": 0.19911709427833557, "learning_rate": 7.708386621639925e-05, "loss": 0.0285, "step": 3520 }, { "grad_norm": 0.18453550338745117, "learning_rate": 7.694472993541219e-05, "loss": 0.0275, "step": 3530 }, { "grad_norm": 0.21897374093532562, "learning_rate": 7.680529899117547e-05, "loss": 0.0355, "step": 3540 }, { "grad_norm": 0.2338583618402481, "learning_rate": 7.666557490848358e-05, "loss": 0.0296, "step": 3550 }, { "grad_norm": 0.26813045144081116, "learning_rate": 7.65255592153367e-05, "loss": 0.0316, "step": 3560 }, { "grad_norm": 0.22822201251983643, "learning_rate": 7.638525344292402e-05, "loss": 0.035, "step": 3570 }, { "grad_norm": 0.17838403582572937, "learning_rate": 7.624465912560697e-05, "loss": 0.0319, "step": 3580 }, { "grad_norm": 0.3971710205078125, "learning_rate": 7.610377780090249e-05, "loss": 0.0288, "step": 3590 }, { "grad_norm": 0.2704290747642517, "learning_rate": 7.596261100946618e-05, "loss": 0.0351, "step": 3600 }, { "grad_norm": 0.2077568769454956, "learning_rate": 7.582116029507542e-05, "loss": 0.0265, "step": 3610 }, { "grad_norm": 0.22542044520378113, "learning_rate": 7.56794272046126e-05, "loss": 0.0299, "step": 3620 }, { "grad_norm": 0.27705544233322144, "learning_rate": 7.55374132880481e-05, "loss": 0.0334, "step": 3630 }, { "grad_norm": 0.2664991021156311, "learning_rate": 7.539512009842333e-05, "loss": 0.0345, "step": 3640 }, { "grad_norm": 0.2080715000629425, "learning_rate": 7.525254919183382e-05, "loss": 0.0338, "step": 3650 }, { "grad_norm": 0.1543269008398056, "learning_rate": 7.510970212741215e-05, "loss": 0.0232, "step": 3660 }, { "grad_norm": 0.31717175245285034, "learning_rate": 7.496658046731096e-05, "loss": 0.0285, "step": 3670 }, { "grad_norm": 0.3038731813430786, "learning_rate": 7.482318577668578e-05, "loss": 0.0332, "step": 3680 }, { "grad_norm": 0.37994128465652466, "learning_rate": 7.467951962367796e-05, "loss": 0.0374, "step": 3690 }, { "grad_norm": 0.31009748578071594, "learning_rate": 7.453558357939755e-05, "loss": 0.0307, "step": 3700 }, { "grad_norm": 0.17653970420360565, "learning_rate": 7.439137921790606e-05, "loss": 0.0331, "step": 3710 }, { "grad_norm": 0.3015185296535492, "learning_rate": 7.42469081161993e-05, "loss": 0.0321, "step": 3720 }, { "grad_norm": 0.21379120647907257, "learning_rate": 7.410217185419006e-05, "loss": 0.0267, "step": 3730 }, { "grad_norm": 0.2472841739654541, "learning_rate": 7.395717201469095e-05, "loss": 0.0358, "step": 3740 }, { "grad_norm": 0.3473665118217468, "learning_rate": 7.381191018339696e-05, "loss": 0.0326, "step": 3750 }, { "grad_norm": 0.2506972849369049, "learning_rate": 7.36663879488682e-05, "loss": 0.03, "step": 3760 }, { "grad_norm": 0.2238173931837082, "learning_rate": 7.352060690251254e-05, "loss": 0.036, "step": 3770 }, { "grad_norm": 0.23032395541667938, "learning_rate": 7.337456863856811e-05, "loss": 0.0325, "step": 3780 }, { "grad_norm": 0.26016494631767273, "learning_rate": 7.3228274754086e-05, "loss": 0.0277, "step": 3790 }, { "grad_norm": 0.24500225484371185, "learning_rate": 7.308172684891267e-05, "loss": 0.0335, "step": 3800 }, { "grad_norm": 0.19754865765571594, "learning_rate": 7.293492652567255e-05, "loss": 0.0257, "step": 3810 }, { "grad_norm": 0.15811826288700104, "learning_rate": 7.278787538975043e-05, "loss": 0.031, "step": 3820 }, { "grad_norm": 0.19066336750984192, "learning_rate": 7.2640575049274e-05, "loss": 0.0328, "step": 3830 }, { "grad_norm": 0.2515644133090973, "learning_rate": 7.249302711509616e-05, "loss": 0.0295, "step": 3840 }, { "grad_norm": 0.3684943914413452, "learning_rate": 7.23452332007775e-05, "loss": 0.0288, "step": 3850 }, { "grad_norm": 0.15881241858005524, "learning_rate": 7.219719492256858e-05, "loss": 0.0308, "step": 3860 }, { "grad_norm": 0.2603793740272522, "learning_rate": 7.20489138993923e-05, "loss": 0.0286, "step": 3870 }, { "grad_norm": 0.20173440873622894, "learning_rate": 7.190039175282614e-05, "loss": 0.028, "step": 3880 }, { "grad_norm": 0.2176423966884613, "learning_rate": 7.175163010708455e-05, "loss": 0.0263, "step": 3890 }, { "grad_norm": 0.20771844685077667, "learning_rate": 7.1602630589001e-05, "loss": 0.0291, "step": 3900 }, { "grad_norm": 0.25005754828453064, "learning_rate": 7.14533948280104e-05, "loss": 0.0304, "step": 3910 }, { "grad_norm": 0.22443783283233643, "learning_rate": 7.130392445613109e-05, "loss": 0.0278, "step": 3920 }, { "grad_norm": 0.24558424949645996, "learning_rate": 7.115422110794711e-05, "loss": 0.0286, "step": 3930 }, { "grad_norm": 0.23003438115119934, "learning_rate": 7.100428642059033e-05, "loss": 0.0341, "step": 3940 }, { "grad_norm": 0.21803398430347443, "learning_rate": 7.08541220337224e-05, "loss": 0.0331, "step": 3950 }, { "grad_norm": 0.38766953349113464, "learning_rate": 7.070372958951706e-05, "loss": 0.0313, "step": 3960 }, { "grad_norm": 0.324675589799881, "learning_rate": 7.055311073264194e-05, "loss": 0.0375, "step": 3970 }, { "grad_norm": 0.3639219105243683, "learning_rate": 7.040226711024077e-05, "loss": 0.0303, "step": 3980 }, { "grad_norm": 0.21190956234931946, "learning_rate": 7.02512003719152e-05, "loss": 0.0274, "step": 3990 }, { "grad_norm": 0.31770941615104675, "learning_rate": 7.00999121697069e-05, "loss": 0.0333, "step": 4000 }, { "grad_norm": 0.23885898292064667, "learning_rate": 6.99484041580794e-05, "loss": 0.0265, "step": 4010 }, { "grad_norm": 0.17833659052848816, "learning_rate": 6.979667799390004e-05, "loss": 0.0257, "step": 4020 }, { "grad_norm": 0.28732091188430786, "learning_rate": 6.964473533642185e-05, "loss": 0.0293, "step": 4030 }, { "grad_norm": 0.1983463317155838, "learning_rate": 6.949257784726539e-05, "loss": 0.0279, "step": 4040 }, { "grad_norm": 0.31734007596969604, "learning_rate": 6.934020719040056e-05, "loss": 0.0287, "step": 4050 }, { "grad_norm": 0.2545399069786072, "learning_rate": 6.918762503212848e-05, "loss": 0.0328, "step": 4060 }, { "grad_norm": 0.20350800454616547, "learning_rate": 6.903483304106319e-05, "loss": 0.0299, "step": 4070 }, { "grad_norm": 0.2234880030155182, "learning_rate": 6.888183288811341e-05, "loss": 0.025, "step": 4080 }, { "grad_norm": 0.1783427596092224, "learning_rate": 6.87286262464643e-05, "loss": 0.0284, "step": 4090 }, { "grad_norm": 0.27758923172950745, "learning_rate": 6.857521479155915e-05, "loss": 0.0319, "step": 4100 }, { "grad_norm": 0.3199763596057892, "learning_rate": 6.842160020108104e-05, "loss": 0.0275, "step": 4110 }, { "grad_norm": 0.27340933680534363, "learning_rate": 6.826778415493455e-05, "loss": 0.0304, "step": 4120 }, { "grad_norm": 0.36284250020980835, "learning_rate": 6.811376833522729e-05, "loss": 0.03, "step": 4130 }, { "grad_norm": 0.24351419508457184, "learning_rate": 6.795955442625159e-05, "loss": 0.0311, "step": 4140 }, { "grad_norm": 0.23593714833259583, "learning_rate": 6.780514411446608e-05, "loss": 0.0279, "step": 4150 }, { "grad_norm": 0.190011665225029, "learning_rate": 6.765053908847716e-05, "loss": 0.0243, "step": 4160 }, { "grad_norm": 0.23605172336101532, "learning_rate": 6.749574103902064e-05, "loss": 0.0294, "step": 4170 }, { "grad_norm": 0.2562548518180847, "learning_rate": 6.734075165894317e-05, "loss": 0.033, "step": 4180 }, { "grad_norm": 0.2442295253276825, "learning_rate": 6.71855726431838e-05, "loss": 0.0275, "step": 4190 }, { "grad_norm": 0.1766994446516037, "learning_rate": 6.703020568875538e-05, "loss": 0.0238, "step": 4200 }, { "grad_norm": 0.20202086865901947, "learning_rate": 6.687465249472603e-05, "loss": 0.023, "step": 4210 }, { "grad_norm": 0.2444705367088318, "learning_rate": 6.671891476220055e-05, "loss": 0.0235, "step": 4220 }, { "grad_norm": 0.1772482842206955, "learning_rate": 6.656299419430183e-05, "loss": 0.0245, "step": 4230 }, { "grad_norm": 0.1747276932001114, "learning_rate": 6.640689249615223e-05, "loss": 0.0285, "step": 4240 }, { "grad_norm": 0.2806522846221924, "learning_rate": 6.625061137485491e-05, "loss": 0.0301, "step": 4250 }, { "grad_norm": 0.28968319296836853, "learning_rate": 6.609415253947517e-05, "loss": 0.0317, "step": 4260 }, { "grad_norm": 0.19500422477722168, "learning_rate": 6.593751770102178e-05, "loss": 0.0277, "step": 4270 }, { "grad_norm": 0.26789239048957825, "learning_rate": 6.578070857242823e-05, "loss": 0.0269, "step": 4280 }, { "grad_norm": 0.2419128715991974, "learning_rate": 6.562372686853402e-05, "loss": 0.0248, "step": 4290 }, { "grad_norm": 0.26919662952423096, "learning_rate": 6.546657430606593e-05, "loss": 0.0293, "step": 4300 }, { "grad_norm": 0.33603113889694214, "learning_rate": 6.530925260361918e-05, "loss": 0.0268, "step": 4310 }, { "grad_norm": 0.2292005568742752, "learning_rate": 6.515176348163871e-05, "loss": 0.0294, "step": 4320 }, { "grad_norm": 0.17253372073173523, "learning_rate": 6.499410866240032e-05, "loss": 0.0239, "step": 4330 }, { "grad_norm": 0.19797582924365997, "learning_rate": 6.48362898699919e-05, "loss": 0.0259, "step": 4340 }, { "grad_norm": 0.28790074586868286, "learning_rate": 6.467830883029443e-05, "loss": 0.0245, "step": 4350 }, { "grad_norm": 0.20172841846942902, "learning_rate": 6.452016727096326e-05, "loss": 0.0237, "step": 4360 }, { "grad_norm": 0.23226380348205566, "learning_rate": 6.436186692140916e-05, "loss": 0.0243, "step": 4370 }, { "grad_norm": 0.2298421710729599, "learning_rate": 6.420340951277938e-05, "loss": 0.025, "step": 4380 }, { "grad_norm": 0.20976707339286804, "learning_rate": 6.404479677793874e-05, "loss": 0.0291, "step": 4390 }, { "grad_norm": 0.23290212452411652, "learning_rate": 6.388603045145075e-05, "loss": 0.031, "step": 4400 }, { "grad_norm": 0.20056304335594177, "learning_rate": 6.372711226955843e-05, "loss": 0.0268, "step": 4410 }, { "grad_norm": 0.1931198090314865, "learning_rate": 6.356804397016564e-05, "loss": 0.0314, "step": 4420 }, { "grad_norm": 0.18088491261005402, "learning_rate": 6.340882729281779e-05, "loss": 0.0275, "step": 4430 }, { "grad_norm": 0.2132810801267624, "learning_rate": 6.324946397868294e-05, "loss": 0.0329, "step": 4440 }, { "grad_norm": 0.17969046533107758, "learning_rate": 6.308995577053276e-05, "loss": 0.0248, "step": 4450 }, { "grad_norm": 0.19553053379058838, "learning_rate": 6.293030441272347e-05, "loss": 0.0261, "step": 4460 }, { "grad_norm": 0.2618257701396942, "learning_rate": 6.277051165117677e-05, "loss": 0.025, "step": 4470 }, { "grad_norm": 0.14612287282943726, "learning_rate": 6.261057923336064e-05, "loss": 0.0248, "step": 4480 }, { "grad_norm": 0.3589058220386505, "learning_rate": 6.245050890827042e-05, "loss": 0.0275, "step": 4490 }, { "grad_norm": 0.43023210763931274, "learning_rate": 6.229030242640952e-05, "loss": 0.029, "step": 4500 }, { "grad_norm": 0.20643489062786102, "learning_rate": 6.212996153977037e-05, "loss": 0.0336, "step": 4510 }, { "grad_norm": 0.2154926359653473, "learning_rate": 6.196948800181523e-05, "loss": 0.0251, "step": 4520 }, { "grad_norm": 0.1552824229001999, "learning_rate": 6.180888356745695e-05, "loss": 0.0286, "step": 4530 }, { "grad_norm": 0.20980259776115417, "learning_rate": 6.164814999303995e-05, "loss": 0.0227, "step": 4540 }, { "grad_norm": 0.26989027857780457, "learning_rate": 6.148728903632081e-05, "loss": 0.0238, "step": 4550 }, { "grad_norm": 0.253307044506073, "learning_rate": 6.132630245644921e-05, "loss": 0.0253, "step": 4560 }, { "grad_norm": 0.29325348138809204, "learning_rate": 6.116519201394857e-05, "loss": 0.0269, "step": 4570 }, { "grad_norm": 0.24940632283687592, "learning_rate": 6.10039594706969e-05, "loss": 0.0251, "step": 4580 }, { "grad_norm": 0.16976597905158997, "learning_rate": 6.084260658990744e-05, "loss": 0.0244, "step": 4590 }, { "grad_norm": 0.22269468009471893, "learning_rate": 6.068113513610943e-05, "loss": 0.0248, "step": 4600 }, { "grad_norm": 0.21847006678581238, "learning_rate": 6.0519546875128876e-05, "loss": 0.0248, "step": 4610 }, { "grad_norm": 0.1659715622663498, "learning_rate": 6.035784357406906e-05, "loss": 0.0285, "step": 4620 }, { "grad_norm": 0.27241092920303345, "learning_rate": 6.01960270012914e-05, "loss": 0.0261, "step": 4630 }, { "grad_norm": 0.2532859146595001, "learning_rate": 6.003409892639599e-05, "loss": 0.0263, "step": 4640 }, { "grad_norm": 0.24068626761436462, "learning_rate": 5.9872061120202336e-05, "loss": 0.0259, "step": 4650 }, { "grad_norm": 0.21822616457939148, "learning_rate": 5.9709915354729914e-05, "loss": 0.033, "step": 4660 }, { "grad_norm": 0.17084018886089325, "learning_rate": 5.9547663403178824e-05, "loss": 0.026, "step": 4670 }, { "grad_norm": 0.15443822741508484, "learning_rate": 5.9385307039910445e-05, "loss": 0.0187, "step": 4680 }, { "grad_norm": 0.18807466328144073, "learning_rate": 5.922284804042792e-05, "loss": 0.0259, "step": 4690 }, { "grad_norm": 0.2835083603858948, "learning_rate": 5.906028818135687e-05, "loss": 0.0297, "step": 4700 }, { "grad_norm": 0.2203025072813034, "learning_rate": 5.889762924042585e-05, "loss": 0.0229, "step": 4710 }, { "grad_norm": 0.16474661231040955, "learning_rate": 5.873487299644699e-05, "loss": 0.0259, "step": 4720 }, { "grad_norm": 0.18733949959278107, "learning_rate": 5.857202122929649e-05, "loss": 0.0259, "step": 4730 }, { "grad_norm": 0.21967047452926636, "learning_rate": 5.840907571989518e-05, "loss": 0.0217, "step": 4740 }, { "grad_norm": 0.23394455015659332, "learning_rate": 5.824603825018904e-05, "loss": 0.0277, "step": 4750 }, { "grad_norm": 0.20997564494609833, "learning_rate": 5.808291060312975e-05, "loss": 0.0226, "step": 4760 }, { "grad_norm": 0.1558031439781189, "learning_rate": 5.7919694562655083e-05, "loss": 0.0244, "step": 4770 }, { "grad_norm": 0.19557452201843262, "learning_rate": 5.775639191366954e-05, "loss": 0.0261, "step": 4780 }, { "grad_norm": 0.208068385720253, "learning_rate": 5.75930044420247e-05, "loss": 0.0259, "step": 4790 }, { "grad_norm": 0.17421892285346985, "learning_rate": 5.74295339344998e-05, "loss": 0.0275, "step": 4800 }, { "grad_norm": 0.1661977767944336, "learning_rate": 5.726598217878211e-05, "loss": 0.0228, "step": 4810 }, { "grad_norm": 0.23800338804721832, "learning_rate": 5.71023509634474e-05, "loss": 0.0214, "step": 4820 }, { "grad_norm": 0.1766684502363205, "learning_rate": 5.693864207794049e-05, "loss": 0.0274, "step": 4830 }, { "grad_norm": 0.1398058533668518, "learning_rate": 5.677485731255545e-05, "loss": 0.0339, "step": 4840 }, { "grad_norm": 0.22359806299209595, "learning_rate": 5.6610998458416296e-05, "loss": 0.0245, "step": 4850 }, { "grad_norm": 0.18239201605319977, "learning_rate": 5.644706730745716e-05, "loss": 0.0268, "step": 4860 }, { "grad_norm": 0.18287858366966248, "learning_rate": 5.628306565240287e-05, "loss": 0.0227, "step": 4870 }, { "grad_norm": 0.19487044215202332, "learning_rate": 5.611899528674923e-05, "loss": 0.028, "step": 4880 }, { "grad_norm": 0.23102310299873352, "learning_rate": 5.595485800474349e-05, "loss": 0.0238, "step": 4890 }, { "grad_norm": 0.25905781984329224, "learning_rate": 5.579065560136467e-05, "loss": 0.027, "step": 4900 }, { "grad_norm": 0.2631303369998932, "learning_rate": 5.562638987230392e-05, "loss": 0.0246, "step": 4910 }, { "grad_norm": 0.2523619532585144, "learning_rate": 5.546206261394498e-05, "loss": 0.0207, "step": 4920 }, { "grad_norm": 0.23928526043891907, "learning_rate": 5.529767562334437e-05, "loss": 0.025, "step": 4930 }, { "grad_norm": 0.26391521096229553, "learning_rate": 5.5133230698211926e-05, "loss": 0.0212, "step": 4940 }, { "grad_norm": 0.22025996446609497, "learning_rate": 5.496872963689096e-05, "loss": 0.0247, "step": 4950 }, { "grad_norm": 0.17973317205905914, "learning_rate": 5.4804174238338756e-05, "loss": 0.0209, "step": 4960 }, { "grad_norm": 0.20982912182807922, "learning_rate": 5.463956630210678e-05, "loss": 0.031, "step": 4970 }, { "grad_norm": 0.2510877847671509, "learning_rate": 5.4474907628321046e-05, "loss": 0.0299, "step": 4980 }, { "grad_norm": 0.21110424399375916, "learning_rate": 5.431020001766244e-05, "loss": 0.0226, "step": 4990 }, { "grad_norm": 0.17870676517486572, "learning_rate": 5.4145445271346986e-05, "loss": 0.024, "step": 5000 }, { "grad_norm": 0.21111856400966644, "learning_rate": 5.398064519110622e-05, "loss": 0.0243, "step": 5010 }, { "grad_norm": 0.298432320356369, "learning_rate": 5.3815801579167394e-05, "loss": 0.0262, "step": 5020 }, { "grad_norm": 0.17039087414741516, "learning_rate": 5.365091623823382e-05, "loss": 0.0247, "step": 5030 }, { "grad_norm": 0.16314338147640228, "learning_rate": 5.348599097146521e-05, "loss": 0.0236, "step": 5040 }, { "grad_norm": 0.23423925042152405, "learning_rate": 5.3321027582457836e-05, "loss": 0.0244, "step": 5050 }, { "grad_norm": 0.25296518206596375, "learning_rate": 5.315602787522491e-05, "loss": 0.026, "step": 5060 }, { "grad_norm": 0.2403746396303177, "learning_rate": 5.299099365417678e-05, "loss": 0.0205, "step": 5070 }, { "grad_norm": 0.23002105951309204, "learning_rate": 5.2825926724101236e-05, "loss": 0.0251, "step": 5080 }, { "grad_norm": 0.21524211764335632, "learning_rate": 5.26608288901438e-05, "loss": 0.0268, "step": 5090 }, { "grad_norm": 0.2229151427745819, "learning_rate": 5.24957019577879e-05, "loss": 0.0285, "step": 5100 }, { "grad_norm": 0.22927090525627136, "learning_rate": 5.2330547732835266e-05, "loss": 0.0239, "step": 5110 }, { "grad_norm": 0.22728806734085083, "learning_rate": 5.2165368021385996e-05, "loss": 0.0329, "step": 5120 }, { "grad_norm": 0.366022527217865, "learning_rate": 5.200016462981897e-05, "loss": 0.0223, "step": 5130 }, { "grad_norm": 0.21690018475055695, "learning_rate": 5.1834939364772015e-05, "loss": 0.0226, "step": 5140 }, { "grad_norm": 0.2649517357349396, "learning_rate": 5.166969403312214e-05, "loss": 0.0222, "step": 5150 }, { "grad_norm": 0.2824750542640686, "learning_rate": 5.1504430441965844e-05, "loss": 0.0258, "step": 5160 }, { "grad_norm": 0.17414304614067078, "learning_rate": 5.133915039859923e-05, "loss": 0.0174, "step": 5170 }, { "grad_norm": 0.1713828593492508, "learning_rate": 5.1173855710498444e-05, "loss": 0.0222, "step": 5180 }, { "grad_norm": 0.1995028853416443, "learning_rate": 5.100854818529967e-05, "loss": 0.0233, "step": 5190 }, { "grad_norm": 0.2799699902534485, "learning_rate": 5.084322963077951e-05, "loss": 0.0298, "step": 5200 }, { "grad_norm": 0.2097620815038681, "learning_rate": 5.067790185483522e-05, "loss": 0.024, "step": 5210 }, { "grad_norm": 0.2677308917045593, "learning_rate": 5.0512566665464844e-05, "loss": 0.0239, "step": 5220 }, { "grad_norm": 0.2056334763765335, "learning_rate": 5.034722587074755e-05, "loss": 0.024, "step": 5230 }, { "grad_norm": 0.15316806733608246, "learning_rate": 5.018188127882375e-05, "loss": 0.0209, "step": 5240 }, { "grad_norm": 0.234547957777977, "learning_rate": 5.0016534697875417e-05, "loss": 0.0185, "step": 5250 }, { "grad_norm": 0.1341109573841095, "learning_rate": 4.9851187936106294e-05, "loss": 0.0248, "step": 5260 }, { "grad_norm": 0.22220894694328308, "learning_rate": 4.968584280172206e-05, "loss": 0.0238, "step": 5270 }, { "grad_norm": 0.24161013960838318, "learning_rate": 4.95205011029106e-05, "loss": 0.0266, "step": 5280 }, { "grad_norm": 0.281843900680542, "learning_rate": 4.935516464782227e-05, "loss": 0.0256, "step": 5290 }, { "grad_norm": 0.2089860737323761, "learning_rate": 4.918983524455003e-05, "loss": 0.0233, "step": 5300 }, { "grad_norm": 0.2313580960035324, "learning_rate": 4.9024514701109766e-05, "loss": 0.0223, "step": 5310 }, { "grad_norm": 0.24522802233695984, "learning_rate": 4.885920482542043e-05, "loss": 0.024, "step": 5320 }, { "grad_norm": 0.2235432267189026, "learning_rate": 4.869390742528438e-05, "loss": 0.0205, "step": 5330 }, { "grad_norm": 0.1975218802690506, "learning_rate": 4.852862430836744e-05, "loss": 0.0238, "step": 5340 }, { "grad_norm": 0.20714668929576874, "learning_rate": 4.836335728217933e-05, "loss": 0.0215, "step": 5350 }, { "grad_norm": 0.1792767196893692, "learning_rate": 4.819810815405379e-05, "loss": 0.0247, "step": 5360 }, { "grad_norm": 0.2726396918296814, "learning_rate": 4.803287873112877e-05, "loss": 0.0274, "step": 5370 }, { "grad_norm": 0.3014656603336334, "learning_rate": 4.786767082032681e-05, "loss": 0.0239, "step": 5380 }, { "grad_norm": 0.14074298739433289, "learning_rate": 4.77024862283351e-05, "loss": 0.0202, "step": 5390 }, { "grad_norm": 0.16625763475894928, "learning_rate": 4.753732676158593e-05, "loss": 0.0241, "step": 5400 }, { "grad_norm": 0.20003578066825867, "learning_rate": 4.737219422623672e-05, "loss": 0.0228, "step": 5410 }, { "grad_norm": 0.19405880570411682, "learning_rate": 4.720709042815044e-05, "loss": 0.0179, "step": 5420 }, { "grad_norm": 0.30463993549346924, "learning_rate": 4.704201717287578e-05, "loss": 0.0241, "step": 5430 }, { "grad_norm": 0.2033459097146988, "learning_rate": 4.6876976265627404e-05, "loss": 0.0225, "step": 5440 }, { "grad_norm": 0.3117787837982178, "learning_rate": 4.671196951126626e-05, "loss": 0.0277, "step": 5450 }, { "grad_norm": 0.1801569014787674, "learning_rate": 4.654699871427971e-05, "loss": 0.0213, "step": 5460 }, { "grad_norm": 0.1678101271390915, "learning_rate": 4.6382065678762034e-05, "loss": 0.022, "step": 5470 }, { "grad_norm": 0.1992630660533905, "learning_rate": 4.6217172208394424e-05, "loss": 0.0214, "step": 5480 }, { "grad_norm": 0.25973883271217346, "learning_rate": 4.605232010642549e-05, "loss": 0.0181, "step": 5490 }, { "grad_norm": 0.16325701773166656, "learning_rate": 4.588751117565142e-05, "loss": 0.0196, "step": 5500 }, { "grad_norm": 0.2112061083316803, "learning_rate": 4.5722747218396214e-05, "loss": 0.0203, "step": 5510 }, { "grad_norm": 0.12029478698968887, "learning_rate": 4.5558030036492194e-05, "loss": 0.0173, "step": 5520 }, { "grad_norm": 0.17809319496154785, "learning_rate": 4.539336143125999e-05, "loss": 0.0269, "step": 5530 }, { "grad_norm": 0.14287729561328888, "learning_rate": 4.522874320348916e-05, "loss": 0.0199, "step": 5540 }, { "grad_norm": 0.1885291337966919, "learning_rate": 4.506417715341821e-05, "loss": 0.0166, "step": 5550 }, { "grad_norm": 0.1869751214981079, "learning_rate": 4.489966508071511e-05, "loss": 0.0222, "step": 5560 }, { "grad_norm": 0.2115202099084854, "learning_rate": 4.4735208784457575e-05, "loss": 0.0185, "step": 5570 }, { "grad_norm": 0.23135606944561005, "learning_rate": 4.457081006311325e-05, "loss": 0.0226, "step": 5580 }, { "grad_norm": 0.20014838874340057, "learning_rate": 4.440647071452027e-05, "loss": 0.0204, "step": 5590 }, { "grad_norm": 0.2833021283149719, "learning_rate": 4.424219253586737e-05, "loss": 0.0249, "step": 5600 }, { "grad_norm": 0.1479504406452179, "learning_rate": 4.407797732367443e-05, "loss": 0.0249, "step": 5610 }, { "grad_norm": 0.23114106059074402, "learning_rate": 4.391382687377268e-05, "loss": 0.0214, "step": 5620 }, { "grad_norm": 0.17118139564990997, "learning_rate": 4.374974298128512e-05, "loss": 0.0187, "step": 5630 }, { "grad_norm": 0.21469072997570038, "learning_rate": 4.358572744060699e-05, "loss": 0.0246, "step": 5640 }, { "grad_norm": 0.2397519201040268, "learning_rate": 4.342178204538588e-05, "loss": 0.0162, "step": 5650 }, { "grad_norm": 0.2292308211326599, "learning_rate": 4.325790858850241e-05, "loss": 0.0213, "step": 5660 }, { "grad_norm": 0.19618074595928192, "learning_rate": 4.309410886205043e-05, "loss": 0.0203, "step": 5670 }, { "grad_norm": 0.16086728870868683, "learning_rate": 4.293038465731752e-05, "loss": 0.0186, "step": 5680 }, { "grad_norm": 0.1940697282552719, "learning_rate": 4.276673776476533e-05, "loss": 0.0213, "step": 5690 }, { "grad_norm": 0.13102318346500397, "learning_rate": 4.260316997401007e-05, "loss": 0.0173, "step": 5700 }, { "grad_norm": 0.2293800562620163, "learning_rate": 4.243968307380293e-05, "loss": 0.0231, "step": 5710 }, { "grad_norm": 0.3290698528289795, "learning_rate": 4.22762788520104e-05, "loss": 0.0245, "step": 5720 }, { "grad_norm": 0.19899433851242065, "learning_rate": 4.211295909559491e-05, "loss": 0.0263, "step": 5730 }, { "grad_norm": 0.18115076422691345, "learning_rate": 4.194972559059511e-05, "loss": 0.0192, "step": 5740 }, { "grad_norm": 0.17197217047214508, "learning_rate": 4.178658012210651e-05, "loss": 0.0186, "step": 5750 }, { "grad_norm": 0.1555451601743698, "learning_rate": 4.162352447426177e-05, "loss": 0.0198, "step": 5760 }, { "grad_norm": 0.30404624342918396, "learning_rate": 4.146056043021135e-05, "loss": 0.02, "step": 5770 }, { "grad_norm": 0.1828131377696991, "learning_rate": 4.1297689772103944e-05, "loss": 0.0222, "step": 5780 }, { "grad_norm": 0.1611815243959427, "learning_rate": 4.113491428106694e-05, "loss": 0.0215, "step": 5790 }, { "grad_norm": 0.16920843720436096, "learning_rate": 4.0972235737187055e-05, "loss": 0.0199, "step": 5800 }, { "grad_norm": 0.20992609858512878, "learning_rate": 4.080965591949076e-05, "loss": 0.0171, "step": 5810 }, { "grad_norm": 0.17002899944782257, "learning_rate": 4.0647176605924924e-05, "loss": 0.0196, "step": 5820 }, { "grad_norm": 0.23627449572086334, "learning_rate": 4.0484799573337255e-05, "loss": 0.0229, "step": 5830 }, { "grad_norm": 0.18887124955654144, "learning_rate": 4.032252659745699e-05, "loss": 0.02, "step": 5840 }, { "grad_norm": 0.21620744466781616, "learning_rate": 4.016035945287539e-05, "loss": 0.0257, "step": 5850 }, { "grad_norm": 0.17924706637859344, "learning_rate": 3.999829991302635e-05, "loss": 0.0195, "step": 5860 }, { "grad_norm": 0.2103392481803894, "learning_rate": 3.983634975016707e-05, "loss": 0.0192, "step": 5870 }, { "grad_norm": 0.24414408206939697, "learning_rate": 3.967451073535854e-05, "loss": 0.0284, "step": 5880 }, { "grad_norm": 0.20437967777252197, "learning_rate": 3.951278463844633e-05, "loss": 0.0173, "step": 5890 }, { "grad_norm": 0.15973792970180511, "learning_rate": 3.935117322804111e-05, "loss": 0.0204, "step": 5900 }, { "grad_norm": 0.22679762542247772, "learning_rate": 3.918967827149938e-05, "loss": 0.0224, "step": 5910 }, { "grad_norm": 0.17461122572422028, "learning_rate": 3.9028301534904094e-05, "loss": 0.0198, "step": 5920 }, { "grad_norm": 0.2074856013059616, "learning_rate": 3.88670447830454e-05, "loss": 0.017, "step": 5930 }, { "grad_norm": 0.22945290803909302, "learning_rate": 3.870590977940132e-05, "loss": 0.0199, "step": 5940 }, { "grad_norm": 0.20542582869529724, "learning_rate": 3.8544898286118404e-05, "loss": 0.0208, "step": 5950 }, { "grad_norm": 0.1813216507434845, "learning_rate": 3.838401206399257e-05, "loss": 0.0181, "step": 5960 }, { "grad_norm": 0.17595194280147552, "learning_rate": 3.822325287244975e-05, "loss": 0.0159, "step": 5970 }, { "grad_norm": 0.2058447152376175, "learning_rate": 3.8062622469526725e-05, "loss": 0.0214, "step": 5980 }, { "grad_norm": 0.18866387009620667, "learning_rate": 3.790212261185183e-05, "loss": 0.0173, "step": 5990 }, { "grad_norm": 0.20874284207820892, "learning_rate": 3.7741755054625794e-05, "loss": 0.0205, "step": 6000 }, { "grad_norm": 0.24301007390022278, "learning_rate": 3.758152155160255e-05, "loss": 0.0172, "step": 6010 }, { "grad_norm": 0.1861373335123062, "learning_rate": 3.742142385506999e-05, "loss": 0.0184, "step": 6020 }, { "grad_norm": 0.16388267278671265, "learning_rate": 3.72614637158309e-05, "loss": 0.0161, "step": 6030 }, { "grad_norm": 0.12648065388202667, "learning_rate": 3.710164288318371e-05, "loss": 0.0221, "step": 6040 }, { "grad_norm": 0.14910662174224854, "learning_rate": 3.694196310490345e-05, "loss": 0.0146, "step": 6050 }, { "grad_norm": 0.2183721363544464, "learning_rate": 3.678242612722259e-05, "loss": 0.0196, "step": 6060 }, { "grad_norm": 0.18195432424545288, "learning_rate": 3.6623033694811953e-05, "loss": 0.0171, "step": 6070 }, { "grad_norm": 0.19163085520267487, "learning_rate": 3.6463787550761665e-05, "loss": 0.023, "step": 6080 }, { "grad_norm": 0.23740535974502563, "learning_rate": 3.630468943656202e-05, "loss": 0.0183, "step": 6090 }, { "grad_norm": 0.2018851935863495, "learning_rate": 3.6145741092084523e-05, "loss": 0.0163, "step": 6100 }, { "grad_norm": 0.10151472687721252, "learning_rate": 3.598694425556278e-05, "loss": 0.0173, "step": 6110 }, { "grad_norm": 0.20645231008529663, "learning_rate": 3.58283006635736e-05, "loss": 0.0208, "step": 6120 }, { "grad_norm": 0.1677972376346588, "learning_rate": 3.566981205101781e-05, "loss": 0.0191, "step": 6130 }, { "grad_norm": 0.1409769058227539, "learning_rate": 3.5511480151101556e-05, "loss": 0.0169, "step": 6140 }, { "grad_norm": 0.12926335632801056, "learning_rate": 3.5353306695317104e-05, "loss": 0.0155, "step": 6150 }, { "grad_norm": 0.19563069939613342, "learning_rate": 3.519529341342402e-05, "loss": 0.0217, "step": 6160 }, { "grad_norm": 0.1810111105442047, "learning_rate": 3.503744203343026e-05, "loss": 0.0189, "step": 6170 }, { "grad_norm": 0.17030303180217743, "learning_rate": 3.487975428157318e-05, "loss": 0.0177, "step": 6180 }, { "grad_norm": 0.2339194416999817, "learning_rate": 3.472223188230083e-05, "loss": 0.023, "step": 6190 }, { "grad_norm": 0.18104776740074158, "learning_rate": 3.4564876558252866e-05, "loss": 0.0187, "step": 6200 }, { "grad_norm": 0.19230380654335022, "learning_rate": 3.440769003024195e-05, "loss": 0.0194, "step": 6210 }, { "grad_norm": 0.2133266180753708, "learning_rate": 3.425067401723477e-05, "loss": 0.0176, "step": 6220 }, { "grad_norm": 0.16459503769874573, "learning_rate": 3.409383023633325e-05, "loss": 0.0179, "step": 6230 }, { "grad_norm": 0.18029530346393585, "learning_rate": 3.3937160402755894e-05, "loss": 0.0173, "step": 6240 }, { "grad_norm": 0.16047437489032745, "learning_rate": 3.378066622981885e-05, "loss": 0.0207, "step": 6250 }, { "grad_norm": 0.2770056128501892, "learning_rate": 3.362434942891738e-05, "loss": 0.0216, "step": 6260 }, { "grad_norm": 0.15078945457935333, "learning_rate": 3.346821170950693e-05, "loss": 0.0181, "step": 6270 }, { "grad_norm": 0.14760951697826385, "learning_rate": 3.3312254779084585e-05, "loss": 0.0199, "step": 6280 }, { "grad_norm": 0.13010315597057343, "learning_rate": 3.315648034317039e-05, "loss": 0.0222, "step": 6290 }, { "grad_norm": 0.12723341584205627, "learning_rate": 3.3000890105288564e-05, "loss": 0.0156, "step": 6300 }, { "grad_norm": 0.23955503106117249, "learning_rate": 3.284548576694908e-05, "loss": 0.0195, "step": 6310 }, { "grad_norm": 0.10950811207294464, "learning_rate": 3.2690269027628815e-05, "loss": 0.0148, "step": 6320 }, { "grad_norm": 0.16154851019382477, "learning_rate": 3.253524158475324e-05, "loss": 0.0151, "step": 6330 }, { "grad_norm": 0.15673905611038208, "learning_rate": 3.238040513367757e-05, "loss": 0.0185, "step": 6340 }, { "grad_norm": 0.1478872001171112, "learning_rate": 3.222576136766843e-05, "loss": 0.0142, "step": 6350 }, { "grad_norm": 0.18054033815860748, "learning_rate": 3.2071311977885324e-05, "loss": 0.0171, "step": 6360 }, { "grad_norm": 0.16510236263275146, "learning_rate": 3.191705865336197e-05, "loss": 0.0201, "step": 6370 }, { "grad_norm": 0.21213585138320923, "learning_rate": 3.1763003080988075e-05, "loss": 0.0173, "step": 6380 }, { "grad_norm": 0.22388684749603271, "learning_rate": 3.160914694549063e-05, "loss": 0.0186, "step": 6390 }, { "grad_norm": 0.15494391322135925, "learning_rate": 3.145549192941573e-05, "loss": 0.0188, "step": 6400 }, { "grad_norm": 0.17399117350578308, "learning_rate": 3.130203971310999e-05, "loss": 0.0176, "step": 6410 }, { "grad_norm": 0.15635186433792114, "learning_rate": 3.114879197470225e-05, "loss": 0.0185, "step": 6420 }, { "grad_norm": 0.172160804271698, "learning_rate": 3.0995750390085285e-05, "loss": 0.0161, "step": 6430 }, { "grad_norm": 0.1847938448190689, "learning_rate": 3.084291663289728e-05, "loss": 0.0178, "step": 6440 }, { "grad_norm": 0.2734600007534027, "learning_rate": 3.069029237450375e-05, "loss": 0.0155, "step": 6450 }, { "grad_norm": 0.17941828072071075, "learning_rate": 3.053787928397911e-05, "loss": 0.0168, "step": 6460 }, { "grad_norm": 0.19831398129463196, "learning_rate": 3.0385679028088526e-05, "loss": 0.0194, "step": 6470 }, { "grad_norm": 0.17675912380218506, "learning_rate": 3.023369327126959e-05, "loss": 0.0158, "step": 6480 }, { "grad_norm": 0.17891480028629303, "learning_rate": 3.0081923675614198e-05, "loss": 0.0149, "step": 6490 }, { "grad_norm": 0.1752539575099945, "learning_rate": 2.993037190085034e-05, "loss": 0.0157, "step": 6500 }, { "grad_norm": 0.19107332825660706, "learning_rate": 2.977903960432392e-05, "loss": 0.0198, "step": 6510 }, { "grad_norm": 0.30129343271255493, "learning_rate": 2.9627928440980722e-05, "loss": 0.023, "step": 6520 }, { "grad_norm": 0.15403415262699127, "learning_rate": 2.9477040063348183e-05, "loss": 0.0167, "step": 6530 }, { "grad_norm": 0.15855643153190613, "learning_rate": 2.9326376121517456e-05, "loss": 0.0164, "step": 6540 }, { "grad_norm": 0.1749446988105774, "learning_rate": 2.9175938263125236e-05, "loss": 0.0169, "step": 6550 }, { "grad_norm": 0.1152254194021225, "learning_rate": 2.9025728133335873e-05, "loss": 0.018, "step": 6560 }, { "grad_norm": 0.1264336258172989, "learning_rate": 2.8875747374823288e-05, "loss": 0.0216, "step": 6570 }, { "grad_norm": 0.19339090585708618, "learning_rate": 2.872599762775298e-05, "loss": 0.0246, "step": 6580 }, { "grad_norm": 0.21167321503162384, "learning_rate": 2.857648052976425e-05, "loss": 0.0169, "step": 6590 }, { "grad_norm": 0.21779845654964447, "learning_rate": 2.8427197715952047e-05, "loss": 0.0176, "step": 6600 }, { "grad_norm": 0.18816585838794708, "learning_rate": 2.8278150818849393e-05, "loss": 0.0163, "step": 6610 }, { "grad_norm": 0.14975960552692413, "learning_rate": 2.812934146840922e-05, "loss": 0.0142, "step": 6620 }, { "grad_norm": 0.2150605171918869, "learning_rate": 2.7980771291986764e-05, "loss": 0.0207, "step": 6630 }, { "grad_norm": 0.2699699401855469, "learning_rate": 2.783244191432167e-05, "loss": 0.0178, "step": 6640 }, { "grad_norm": 0.20075759291648865, "learning_rate": 2.768435495752022e-05, "loss": 0.0161, "step": 6650 }, { "grad_norm": 0.1815977245569229, "learning_rate": 2.753651204103771e-05, "loss": 0.0165, "step": 6660 }, { "grad_norm": 0.19166889786720276, "learning_rate": 2.7388914781660523e-05, "loss": 0.016, "step": 6670 }, { "grad_norm": 0.16243691742420197, "learning_rate": 2.7241564793488693e-05, "loss": 0.0182, "step": 6680 }, { "grad_norm": 0.22660264372825623, "learning_rate": 2.7094463687918037e-05, "loss": 0.0167, "step": 6690 }, { "grad_norm": 0.23117853701114655, "learning_rate": 2.694761307362268e-05, "loss": 0.0162, "step": 6700 }, { "grad_norm": 0.1355389654636383, "learning_rate": 2.6801014556537467e-05, "loss": 0.0157, "step": 6710 }, { "grad_norm": 0.1914055347442627, "learning_rate": 2.6654669739840243e-05, "loss": 0.0144, "step": 6720 }, { "grad_norm": 0.20606271922588348, "learning_rate": 2.650858022393451e-05, "loss": 0.0184, "step": 6730 }, { "grad_norm": 0.1971818506717682, "learning_rate": 2.6362747606431747e-05, "loss": 0.0195, "step": 6740 }, { "grad_norm": 0.11258693039417267, "learning_rate": 2.6217173482134172e-05, "loss": 0.0153, "step": 6750 }, { "grad_norm": 0.12520867586135864, "learning_rate": 2.6071859443017044e-05, "loss": 0.0158, "step": 6760 }, { "grad_norm": 0.19072768092155457, "learning_rate": 2.5926807078211414e-05, "loss": 0.0143, "step": 6770 }, { "grad_norm": 0.2020331174135208, "learning_rate": 2.5782017973986728e-05, "loss": 0.0142, "step": 6780 }, { "grad_norm": 0.1382087916135788, "learning_rate": 2.5637493713733374e-05, "loss": 0.0171, "step": 6790 }, { "grad_norm": 0.20873375236988068, "learning_rate": 2.549323587794559e-05, "loss": 0.0238, "step": 6800 }, { "grad_norm": 0.1455889344215393, "learning_rate": 2.5349246044203895e-05, "loss": 0.0126, "step": 6810 }, { "grad_norm": 0.16566771268844604, "learning_rate": 2.520552578715808e-05, "loss": 0.0166, "step": 6820 }, { "grad_norm": 0.20062410831451416, "learning_rate": 2.506207667850981e-05, "loss": 0.013, "step": 6830 }, { "grad_norm": 0.23433108627796173, "learning_rate": 2.4918900286995555e-05, "loss": 0.0159, "step": 6840 }, { "grad_norm": 0.10260957479476929, "learning_rate": 2.4775998178369458e-05, "loss": 0.0126, "step": 6850 }, { "grad_norm": 0.13390065729618073, "learning_rate": 2.4633371915386017e-05, "loss": 0.0196, "step": 6860 }, { "grad_norm": 0.14119313657283783, "learning_rate": 2.4491023057783235e-05, "loss": 0.0188, "step": 6870 }, { "grad_norm": 0.16310876607894897, "learning_rate": 2.4348953162265375e-05, "loss": 0.021, "step": 6880 }, { "grad_norm": 0.14795207977294922, "learning_rate": 2.420716378248607e-05, "loss": 0.0174, "step": 6890 }, { "grad_norm": 0.19894002377986908, "learning_rate": 2.4065656469031266e-05, "loss": 0.0151, "step": 6900 }, { "grad_norm": 0.17174188792705536, "learning_rate": 2.3924432769402268e-05, "loss": 0.0167, "step": 6910 }, { "grad_norm": 0.2161995768547058, "learning_rate": 2.3783494227998844e-05, "loss": 0.023, "step": 6920 }, { "grad_norm": 0.14266380667686462, "learning_rate": 2.3642842386102264e-05, "loss": 0.0139, "step": 6930 }, { "grad_norm": 0.09098489582538605, "learning_rate": 2.3502478781858567e-05, "loss": 0.0127, "step": 6940 }, { "grad_norm": 0.1875651776790619, "learning_rate": 2.3362404950261628e-05, "loss": 0.0183, "step": 6950 }, { "grad_norm": 0.18232321739196777, "learning_rate": 2.3222622423136458e-05, "loss": 0.0146, "step": 6960 }, { "grad_norm": 0.1709059625864029, "learning_rate": 2.3083132729122332e-05, "loss": 0.0157, "step": 6970 }, { "grad_norm": 0.16723498702049255, "learning_rate": 2.294393739365621e-05, "loss": 0.015, "step": 6980 }, { "grad_norm": 0.12622317671775818, "learning_rate": 2.2805037938956e-05, "loss": 0.0169, "step": 6990 }, { "grad_norm": 0.1713910549879074, "learning_rate": 2.266643588400386e-05, "loss": 0.013, "step": 7000 }, { "grad_norm": 0.15605196356773376, "learning_rate": 2.252813274452969e-05, "loss": 0.0164, "step": 7010 }, { "grad_norm": 0.12056981027126312, "learning_rate": 2.2390130032994427e-05, "loss": 0.0129, "step": 7020 }, { "grad_norm": 0.17300763726234436, "learning_rate": 2.2252429258573633e-05, "loss": 0.015, "step": 7030 }, { "grad_norm": 0.16801564395427704, "learning_rate": 2.2115031927140904e-05, "loss": 0.0183, "step": 7040 }, { "grad_norm": 0.17644909024238586, "learning_rate": 2.1977939541251463e-05, "loss": 0.0213, "step": 7050 }, { "grad_norm": 0.1529550552368164, "learning_rate": 2.1841153600125684e-05, "loss": 0.0127, "step": 7060 }, { "grad_norm": 0.12212110310792923, "learning_rate": 2.170467559963267e-05, "loss": 0.0168, "step": 7070 }, { "grad_norm": 0.15709425508975983, "learning_rate": 2.1568507032273982e-05, "loss": 0.014, "step": 7080 }, { "grad_norm": 0.14139342308044434, "learning_rate": 2.1432649387167264e-05, "loss": 0.0135, "step": 7090 }, { "grad_norm": 0.11711273342370987, "learning_rate": 2.1297104150029973e-05, "loss": 0.0144, "step": 7100 }, { "grad_norm": 0.2023629993200302, "learning_rate": 2.116187280316307e-05, "loss": 0.0163, "step": 7110 }, { "grad_norm": 0.20753523707389832, "learning_rate": 2.1026956825434908e-05, "loss": 0.0163, "step": 7120 }, { "grad_norm": 0.15429365634918213, "learning_rate": 2.0892357692265017e-05, "loss": 0.0151, "step": 7130 }, { "grad_norm": 0.19455178081989288, "learning_rate": 2.0758076875607947e-05, "loss": 0.016, "step": 7140 }, { "grad_norm": 0.11026325076818466, "learning_rate": 2.0624115843937207e-05, "loss": 0.0174, "step": 7150 }, { "grad_norm": 0.14076511561870575, "learning_rate": 2.0490476062229157e-05, "loss": 0.0196, "step": 7160 }, { "grad_norm": 0.14516936242580414, "learning_rate": 2.035715899194704e-05, "loss": 0.0147, "step": 7170 }, { "grad_norm": 0.15679407119750977, "learning_rate": 2.022416609102499e-05, "loss": 0.0152, "step": 7180 }, { "grad_norm": 0.16737958788871765, "learning_rate": 2.009149881385205e-05, "loss": 0.0202, "step": 7190 }, { "grad_norm": 0.13998618721961975, "learning_rate": 1.995915861125634e-05, "loss": 0.0149, "step": 7200 }, { "grad_norm": 0.11836197227239609, "learning_rate": 1.9827146930489065e-05, "loss": 0.0123, "step": 7210 }, { "grad_norm": 0.1566004753112793, "learning_rate": 1.9695465215208848e-05, "loss": 0.0204, "step": 7220 }, { "grad_norm": 0.1454019844532013, "learning_rate": 1.9564114905465813e-05, "loss": 0.013, "step": 7230 }, { "grad_norm": 0.12373020499944687, "learning_rate": 1.9433097437685936e-05, "loss": 0.016, "step": 7240 }, { "grad_norm": 0.20374807715415955, "learning_rate": 1.930241424465521e-05, "loss": 0.0138, "step": 7250 }, { "grad_norm": 0.22306931018829346, "learning_rate": 1.9172066755504115e-05, "loss": 0.0152, "step": 7260 }, { "grad_norm": 0.201716348528862, "learning_rate": 1.9042056395691914e-05, "loss": 0.0155, "step": 7270 }, { "grad_norm": 0.10736740380525589, "learning_rate": 1.8912384586991066e-05, "loss": 0.0125, "step": 7280 }, { "grad_norm": 0.22348569333553314, "learning_rate": 1.8783052747471717e-05, "loss": 0.0137, "step": 7290 }, { "grad_norm": 0.143004909157753, "learning_rate": 1.865406229148611e-05, "loss": 0.017, "step": 7300 }, { "grad_norm": 0.12815743684768677, "learning_rate": 1.8525414629653233e-05, "loss": 0.0193, "step": 7310 }, { "grad_norm": 0.17485801875591278, "learning_rate": 1.8397111168843255e-05, "loss": 0.0138, "step": 7320 }, { "grad_norm": 0.22644290328025818, "learning_rate": 1.8269153312162323e-05, "loss": 0.0181, "step": 7330 }, { "grad_norm": 0.1598215252161026, "learning_rate": 1.8141542458937054e-05, "loss": 0.0137, "step": 7340 }, { "grad_norm": 0.13555458188056946, "learning_rate": 1.8014280004699268e-05, "loss": 0.0131, "step": 7350 }, { "grad_norm": 0.1912456601858139, "learning_rate": 1.788736734117078e-05, "loss": 0.0125, "step": 7360 }, { "grad_norm": 0.13225045800209045, "learning_rate": 1.7760805856248152e-05, "loss": 0.0139, "step": 7370 }, { "grad_norm": 0.14121341705322266, "learning_rate": 1.7634596933987518e-05, "loss": 0.0149, "step": 7380 }, { "grad_norm": 0.12116909772157669, "learning_rate": 1.7508741954589404e-05, "loss": 0.0164, "step": 7390 }, { "grad_norm": 0.09469825029373169, "learning_rate": 1.7383242294383717e-05, "loss": 0.0143, "step": 7400 }, { "grad_norm": 0.14066258072853088, "learning_rate": 1.7258099325814632e-05, "loss": 0.0111, "step": 7410 }, { "grad_norm": 0.1475728154182434, "learning_rate": 1.7133314417425594e-05, "loss": 0.0157, "step": 7420 }, { "grad_norm": 0.17055334150791168, "learning_rate": 1.7008888933844408e-05, "loss": 0.0141, "step": 7430 }, { "grad_norm": 0.15132835507392883, "learning_rate": 1.6884824235768172e-05, "loss": 0.015, "step": 7440 }, { "grad_norm": 0.11495436728000641, "learning_rate": 1.6761121679948592e-05, "loss": 0.0124, "step": 7450 }, { "grad_norm": 0.1160864308476448, "learning_rate": 1.663778261917695e-05, "loss": 0.016, "step": 7460 }, { "grad_norm": 0.196089968085289, "learning_rate": 1.651480840226952e-05, "loss": 0.0133, "step": 7470 }, { "grad_norm": 0.15958842635154724, "learning_rate": 1.639220037405258e-05, "loss": 0.0138, "step": 7480 }, { "grad_norm": 0.19039584696292877, "learning_rate": 1.6269959875347906e-05, "loss": 0.014, "step": 7490 }, { "grad_norm": 0.1763318032026291, "learning_rate": 1.614808824295802e-05, "loss": 0.0145, "step": 7500 }, { "grad_norm": 0.16834403574466705, "learning_rate": 1.602658680965152e-05, "loss": 0.0112, "step": 7510 }, { "grad_norm": 0.12094952166080475, "learning_rate": 1.5905456904148686e-05, "loss": 0.0171, "step": 7520 }, { "grad_norm": 0.252076119184494, "learning_rate": 1.57846998511067e-05, "loss": 0.0173, "step": 7530 }, { "grad_norm": 0.13076864182949066, "learning_rate": 1.566431697110538e-05, "loss": 0.0122, "step": 7540 }, { "grad_norm": 0.20389549434185028, "learning_rate": 1.554430958063259e-05, "loss": 0.0128, "step": 7550 }, { "grad_norm": 0.1678963154554367, "learning_rate": 1.5424678992069912e-05, "loss": 0.0118, "step": 7560 }, { "grad_norm": 0.19656077027320862, "learning_rate": 1.5305426513678362e-05, "loss": 0.0149, "step": 7570 }, { "grad_norm": 0.20102985203266144, "learning_rate": 1.518655344958388e-05, "loss": 0.0163, "step": 7580 }, { "grad_norm": 0.16944259405136108, "learning_rate": 1.5068061099763275e-05, "loss": 0.0142, "step": 7590 }, { "grad_norm": 0.13637343049049377, "learning_rate": 1.494995076002988e-05, "loss": 0.0157, "step": 7600 }, { "grad_norm": 0.1748809814453125, "learning_rate": 1.4832223722019456e-05, "loss": 0.0115, "step": 7610 }, { "grad_norm": 0.09317122399806976, "learning_rate": 1.4714881273176035e-05, "loss": 0.011, "step": 7620 }, { "grad_norm": 0.16769948601722717, "learning_rate": 1.4597924696737835e-05, "loss": 0.0155, "step": 7630 }, { "grad_norm": 0.1249789223074913, "learning_rate": 1.4481355271723252e-05, "loss": 0.0131, "step": 7640 }, { "grad_norm": 0.13603630661964417, "learning_rate": 1.4365174272916809e-05, "loss": 0.0167, "step": 7650 }, { "grad_norm": 0.1430763155221939, "learning_rate": 1.4249382970855319e-05, "loss": 0.0134, "step": 7660 }, { "grad_norm": 0.1413087695837021, "learning_rate": 1.4133982631813903e-05, "loss": 0.0148, "step": 7670 }, { "grad_norm": 0.1687745600938797, "learning_rate": 1.4018974517792194e-05, "loss": 0.0119, "step": 7680 }, { "grad_norm": 0.13151030242443085, "learning_rate": 1.390435988650048e-05, "loss": 0.0138, "step": 7690 }, { "grad_norm": 0.13335809111595154, "learning_rate": 1.3790139991346006e-05, "loss": 0.0133, "step": 7700 }, { "grad_norm": 0.2066105157136917, "learning_rate": 1.367631608141926e-05, "loss": 0.0135, "step": 7710 }, { "grad_norm": 0.15266694128513336, "learning_rate": 1.3562889401480278e-05, "loss": 0.0112, "step": 7720 }, { "grad_norm": 0.12902896106243134, "learning_rate": 1.3449861191945074e-05, "loss": 0.0145, "step": 7730 }, { "grad_norm": 0.1378546506166458, "learning_rate": 1.3337232688872009e-05, "loss": 0.0124, "step": 7740 }, { "grad_norm": 0.14824151992797852, "learning_rate": 1.3225005123948364e-05, "loss": 0.0126, "step": 7750 }, { "grad_norm": 0.1550481766462326, "learning_rate": 1.311317972447681e-05, "loss": 0.0122, "step": 7760 }, { "grad_norm": 0.15472941100597382, "learning_rate": 1.3001757713361996e-05, "loss": 0.0088, "step": 7770 }, { "grad_norm": 0.22371751070022583, "learning_rate": 1.2890740309097204e-05, "loss": 0.0143, "step": 7780 }, { "grad_norm": 0.16064859926700592, "learning_rate": 1.2780128725750944e-05, "loss": 0.0117, "step": 7790 }, { "grad_norm": 0.15371255576610565, "learning_rate": 1.266992417295379e-05, "loss": 0.0131, "step": 7800 }, { "grad_norm": 0.16012048721313477, "learning_rate": 1.2560127855885073e-05, "loss": 0.0141, "step": 7810 }, { "grad_norm": 0.2112874835729599, "learning_rate": 1.2450740975259745e-05, "loss": 0.0182, "step": 7820 }, { "grad_norm": 0.1497241109609604, "learning_rate": 1.234176472731517e-05, "loss": 0.0187, "step": 7830 }, { "grad_norm": 0.20241333544254303, "learning_rate": 1.2233200303798158e-05, "loss": 0.0151, "step": 7840 }, { "grad_norm": 0.137581929564476, "learning_rate": 1.2125048891951846e-05, "loss": 0.0115, "step": 7850 }, { "grad_norm": 0.10353267937898636, "learning_rate": 1.2017311674502745e-05, "loss": 0.0103, "step": 7860 }, { "grad_norm": 0.20903342962265015, "learning_rate": 1.1909989829647822e-05, "loss": 0.0131, "step": 7870 }, { "grad_norm": 0.1234491616487503, "learning_rate": 1.1803084531041553e-05, "loss": 0.0132, "step": 7880 }, { "grad_norm": 0.1918737143278122, "learning_rate": 1.1696596947783162e-05, "loss": 0.0128, "step": 7890 }, { "grad_norm": 0.16622322797775269, "learning_rate": 1.1590528244403803e-05, "loss": 0.0132, "step": 7900 }, { "grad_norm": 0.2549535632133484, "learning_rate": 1.148487958085382e-05, "loss": 0.0174, "step": 7910 }, { "grad_norm": 0.12393693625926971, "learning_rate": 1.1379652112490086e-05, "loss": 0.0126, "step": 7920 }, { "grad_norm": 0.24038994312286377, "learning_rate": 1.1274846990063315e-05, "loss": 0.0132, "step": 7930 }, { "grad_norm": 0.17276456952095032, "learning_rate": 1.117046535970554e-05, "loss": 0.0127, "step": 7940 }, { "grad_norm": 0.12973998486995697, "learning_rate": 1.106650836291755e-05, "loss": 0.0187, "step": 7950 }, { "grad_norm": 0.2425190508365631, "learning_rate": 1.0962977136556418e-05, "loss": 0.021, "step": 7960 }, { "grad_norm": 0.22836294770240784, "learning_rate": 1.0859872812823024e-05, "loss": 0.0139, "step": 7970 }, { "grad_norm": 0.2183670848608017, "learning_rate": 1.0757196519249747e-05, "loss": 0.0213, "step": 7980 }, { "grad_norm": 0.09989073127508163, "learning_rate": 1.0654949378688077e-05, "loss": 0.0137, "step": 7990 }, { "grad_norm": 0.10476111620664597, "learning_rate": 1.0553132509296376e-05, "loss": 0.0127, "step": 8000 }, { "grad_norm": 0.15584993362426758, "learning_rate": 1.0451747024527613e-05, "loss": 0.0156, "step": 8010 }, { "grad_norm": 0.07291298359632492, "learning_rate": 1.0350794033117189e-05, "loss": 0.0129, "step": 8020 }, { "grad_norm": 0.17351505160331726, "learning_rate": 1.0250274639070856e-05, "loss": 0.0141, "step": 8030 }, { "grad_norm": 0.14552554488182068, "learning_rate": 1.0150189941652599e-05, "loss": 0.011, "step": 8040 }, { "grad_norm": 0.15430666506290436, "learning_rate": 1.0050541035372635e-05, "loss": 0.0102, "step": 8050 }, { "grad_norm": 0.14589077234268188, "learning_rate": 9.951329009975458e-06, "loss": 0.0137, "step": 8060 }, { "grad_norm": 0.11308623105287552, "learning_rate": 9.852554950427845e-06, "loss": 0.0117, "step": 8070 }, { "grad_norm": 0.21355678141117096, "learning_rate": 9.754219936907105e-06, "loss": 0.0107, "step": 8080 }, { "grad_norm": 0.18101254105567932, "learning_rate": 9.656325044789194e-06, "loss": 0.0111, "step": 8090 }, { "grad_norm": 0.15565314888954163, "learning_rate": 9.55887134463697e-06, "loss": 0.0117, "step": 8100 }, { "grad_norm": 0.12406432628631592, "learning_rate": 9.461859902188475e-06, "loss": 0.0119, "step": 8110 }, { "grad_norm": 0.12049220502376556, "learning_rate": 9.365291778345303e-06, "loss": 0.0129, "step": 8120 }, { "grad_norm": 0.09245448559522629, "learning_rate": 9.269168029160991e-06, "loss": 0.0099, "step": 8130 }, { "grad_norm": 0.18314129114151, "learning_rate": 9.173489705829447e-06, "loss": 0.0144, "step": 8140 }, { "grad_norm": 0.14671066403388977, "learning_rate": 9.078257854673516e-06, "loss": 0.0127, "step": 8150 }, { "grad_norm": 0.09395632147789001, "learning_rate": 8.983473517133429e-06, "loss": 0.0113, "step": 8160 }, { "grad_norm": 0.1617269366979599, "learning_rate": 8.889137729755537e-06, "loss": 0.0115, "step": 8170 }, { "grad_norm": 0.11325367540121078, "learning_rate": 8.79525152418087e-06, "loss": 0.0147, "step": 8180 }, { "grad_norm": 0.09093060344457626, "learning_rate": 8.701815927133961e-06, "loss": 0.0127, "step": 8190 }, { "grad_norm": 0.13928936421871185, "learning_rate": 8.608831960411534e-06, "loss": 0.0136, "step": 8200 }, { "grad_norm": 0.11112497001886368, "learning_rate": 8.516300640871321e-06, "loss": 0.0118, "step": 8210 }, { "grad_norm": 0.23459313809871674, "learning_rate": 8.424222980421038e-06, "loss": 0.0131, "step": 8220 }, { "grad_norm": 0.18964238464832306, "learning_rate": 8.332599986007184e-06, "loss": 0.0132, "step": 8230 }, { "grad_norm": 0.07244930416345596, "learning_rate": 8.241432659604203e-06, "loss": 0.0099, "step": 8240 }, { "grad_norm": 0.16244952380657196, "learning_rate": 8.150721998203331e-06, "loss": 0.0107, "step": 8250 }, { "grad_norm": 0.1344965100288391, "learning_rate": 8.06046899380184e-06, "loss": 0.0121, "step": 8260 }, { "grad_norm": 0.2448025345802307, "learning_rate": 7.970674633392133e-06, "loss": 0.018, "step": 8270 }, { "grad_norm": 0.14939352869987488, "learning_rate": 7.881339898950924e-06, "loss": 0.0116, "step": 8280 }, { "grad_norm": 0.12850721180438995, "learning_rate": 7.792465767428597e-06, "loss": 0.0119, "step": 8290 }, { "grad_norm": 0.08892274647951126, "learning_rate": 7.704053210738376e-06, "loss": 0.0092, "step": 8300 }, { "grad_norm": 0.19393090903759003, "learning_rate": 7.6161031957458494e-06, "loss": 0.0157, "step": 8310 }, { "grad_norm": 0.14795632660388947, "learning_rate": 7.5286166842582605e-06, "loss": 0.0126, "step": 8320 }, { "grad_norm": 0.114802286028862, "learning_rate": 7.4415946330140814e-06, "loss": 0.0113, "step": 8330 }, { "grad_norm": 0.1105973869562149, "learning_rate": 7.3550379936725644e-06, "loss": 0.0111, "step": 8340 }, { "grad_norm": 0.0866873487830162, "learning_rate": 7.2689477128032035e-06, "loss": 0.0127, "step": 8350 }, { "grad_norm": 0.12462054193019867, "learning_rate": 7.183324731875551e-06, "loss": 0.0109, "step": 8360 }, { "grad_norm": 0.2077144831418991, "learning_rate": 7.098169987248782e-06, "loss": 0.0089, "step": 8370 }, { "grad_norm": 0.13467471301555634, "learning_rate": 7.013484410161553e-06, "loss": 0.0132, "step": 8380 }, { "grad_norm": 0.15024136006832123, "learning_rate": 6.92926892672176e-06, "loss": 0.0138, "step": 8390 }, { "grad_norm": 0.115578792989254, "learning_rate": 6.845524457896446e-06, "loss": 0.0096, "step": 8400 }, { "grad_norm": 0.11852545291185379, "learning_rate": 6.7622519195017165e-06, "loss": 0.0108, "step": 8410 }, { "grad_norm": 0.07927063852548599, "learning_rate": 6.679452222192684e-06, "loss": 0.0109, "step": 8420 }, { "grad_norm": 0.09844058752059937, "learning_rate": 6.597126271453579e-06, "loss": 0.011, "step": 8430 }, { "grad_norm": 0.15749429166316986, "learning_rate": 6.51527496758782e-06, "loss": 0.0101, "step": 8440 }, { "grad_norm": 0.12616148591041565, "learning_rate": 6.433899205708155e-06, "loss": 0.0106, "step": 8450 }, { "grad_norm": 0.1520530730485916, "learning_rate": 6.352999875726856e-06, "loss": 0.0094, "step": 8460 }, { "grad_norm": 0.14901386201381683, "learning_rate": 6.272577862346052e-06, "loss": 0.0097, "step": 8470 }, { "grad_norm": 0.17188186943531036, "learning_rate": 6.192634045047996e-06, "loss": 0.0179, "step": 8480 }, { "grad_norm": 0.14584150910377502, "learning_rate": 6.113169298085458e-06, "loss": 0.0136, "step": 8490 }, { "grad_norm": 0.2611314654350281, "learning_rate": 6.034184490472195e-06, "loss": 0.011, "step": 8500 }, { "grad_norm": 0.14182092249393463, "learning_rate": 5.955680485973386e-06, "loss": 0.0107, "step": 8510 }, { "grad_norm": 0.16524113714694977, "learning_rate": 5.877658143096265e-06, "loss": 0.0108, "step": 8520 }, { "grad_norm": 0.11848839372396469, "learning_rate": 5.800118315080661e-06, "loss": 0.0093, "step": 8530 }, { "grad_norm": 0.15183329582214355, "learning_rate": 5.723061849889716e-06, "loss": 0.0095, "step": 8540 }, { "grad_norm": 0.1282571703195572, "learning_rate": 5.646489590200604e-06, "loss": 0.0124, "step": 8550 }, { "grad_norm": 0.12772516906261444, "learning_rate": 5.570402373395256e-06, "loss": 0.0104, "step": 8560 }, { "grad_norm": 0.10616809129714966, "learning_rate": 5.494801031551305e-06, "loss": 0.0152, "step": 8570 }, { "grad_norm": 0.17324693500995636, "learning_rate": 5.41968639143291e-06, "loss": 0.01, "step": 8580 }, { "grad_norm": 0.1224069818854332, "learning_rate": 5.345059274481751e-06, "loss": 0.0101, "step": 8590 }, { "grad_norm": 0.14806406199932098, "learning_rate": 5.270920496808002e-06, "loss": 0.0141, "step": 8600 }, { "grad_norm": 0.21798059344291687, "learning_rate": 5.1972708691814695e-06, "loss": 0.0129, "step": 8610 }, { "grad_norm": 0.12504763901233673, "learning_rate": 5.124111197022674e-06, "loss": 0.0126, "step": 8620 }, { "grad_norm": 0.2139354795217514, "learning_rate": 5.051442280394081e-06, "loss": 0.0106, "step": 8630 }, { "grad_norm": 0.13048024475574493, "learning_rate": 4.979264913991322e-06, "loss": 0.0099, "step": 8640 }, { "grad_norm": 0.16292615234851837, "learning_rate": 4.907579887134489e-06, "loss": 0.014, "step": 8650 }, { "grad_norm": 0.19152607023715973, "learning_rate": 4.836387983759572e-06, "loss": 0.0106, "step": 8660 }, { "grad_norm": 0.1169474795460701, "learning_rate": 4.765689982409816e-06, "loss": 0.01, "step": 8670 }, { "grad_norm": 0.14360105991363525, "learning_rate": 4.695486656227233e-06, "loss": 0.0149, "step": 8680 }, { "grad_norm": 0.1019250825047493, "learning_rate": 4.625778772944156e-06, "loss": 0.0103, "step": 8690 }, { "grad_norm": 0.09302256256341934, "learning_rate": 4.556567094874825e-06, "loss": 0.0111, "step": 8700 }, { "grad_norm": 0.11409218609333038, "learning_rate": 4.487852378907059e-06, "loss": 0.0084, "step": 8710 }, { "grad_norm": 0.1498062163591385, "learning_rate": 4.419635376493986e-06, "loss": 0.0101, "step": 8720 }, { "grad_norm": 0.11267298460006714, "learning_rate": 4.351916833645825e-06, "loss": 0.0132, "step": 8730 }, { "grad_norm": 0.1822999268770218, "learning_rate": 4.284697490921691e-06, "loss": 0.0125, "step": 8740 }, { "grad_norm": 0.210024893283844, "learning_rate": 4.2179780834215585e-06, "loss": 0.0124, "step": 8750 }, { "grad_norm": 0.1692306250333786, "learning_rate": 4.151759340778178e-06, "loss": 0.0127, "step": 8760 }, { "grad_norm": 0.16332361102104187, "learning_rate": 4.086041987149109e-06, "loss": 0.0096, "step": 8770 }, { "grad_norm": 0.16440030932426453, "learning_rate": 4.020826741208811e-06, "loss": 0.0125, "step": 8780 }, { "grad_norm": 0.1810166984796524, "learning_rate": 3.956114316140746e-06, "loss": 0.0138, "step": 8790 }, { "grad_norm": 0.14973098039627075, "learning_rate": 3.891905419629643e-06, "loss": 0.0124, "step": 8800 }, { "grad_norm": 0.11333898454904556, "learning_rate": 3.8282007538536946e-06, "loss": 0.0147, "step": 8810 }, { "grad_norm": 0.2362566441297531, "learning_rate": 3.7650010154769265e-06, "loss": 0.012, "step": 8820 }, { "grad_norm": 0.18750977516174316, "learning_rate": 3.7023068956415608e-06, "loss": 0.0122, "step": 8830 }, { "grad_norm": 0.14243942499160767, "learning_rate": 3.6401190799604303e-06, "loss": 0.0091, "step": 8840 }, { "grad_norm": 0.13702280819416046, "learning_rate": 3.578438248509536e-06, "loss": 0.0112, "step": 8850 }, { "grad_norm": 0.1270366609096527, "learning_rate": 3.5172650758205583e-06, "loss": 0.0126, "step": 8860 }, { "grad_norm": 0.12612245976924896, "learning_rate": 3.45660023087353e-06, "loss": 0.009, "step": 8870 }, { "grad_norm": 0.10695841163396835, "learning_rate": 3.3964443770894528e-06, "loss": 0.0121, "step": 8880 }, { "grad_norm": 0.1714233160018921, "learning_rate": 3.3367981723231245e-06, "loss": 0.0103, "step": 8890 }, { "grad_norm": 0.22548113763332367, "learning_rate": 3.2776622688558746e-06, "loss": 0.0145, "step": 8900 }, { "grad_norm": 0.16675962507724762, "learning_rate": 3.2190373133884677e-06, "loss": 0.0087, "step": 8910 }, { "grad_norm": 0.12807071208953857, "learning_rate": 3.1609239470340446e-06, "loss": 0.0102, "step": 8920 }, { "grad_norm": 0.09657885134220123, "learning_rate": 3.1033228053110373e-06, "loss": 0.0085, "step": 8930 }, { "grad_norm": 0.09678257256746292, "learning_rate": 3.0462345181363314e-06, "loss": 0.0093, "step": 8940 }, { "grad_norm": 0.14196868240833282, "learning_rate": 2.9896597098182654e-06, "loss": 0.0131, "step": 8950 }, { "grad_norm": 0.17089851200580597, "learning_rate": 2.933598999049891e-06, "loss": 0.0121, "step": 8960 }, { "grad_norm": 0.11916597187519073, "learning_rate": 2.8780529989021697e-06, "loss": 0.0082, "step": 8970 }, { "grad_norm": 0.09864859282970428, "learning_rate": 2.823022316817242e-06, "loss": 0.0136, "step": 8980 }, { "grad_norm": 0.1239866390824318, "learning_rate": 2.7685075546018456e-06, "loss": 0.0121, "step": 8990 }, { "grad_norm": 0.13456425070762634, "learning_rate": 2.7145093084206598e-06, "loss": 0.0161, "step": 9000 }, { "grad_norm": 0.12325014919042587, "learning_rate": 2.661028168789892e-06, "loss": 0.0102, "step": 9010 }, { "grad_norm": 0.15166881680488586, "learning_rate": 2.6080647205706855e-06, "loss": 0.0093, "step": 9020 }, { "grad_norm": 0.07158375531435013, "learning_rate": 2.555619542962834e-06, "loss": 0.0091, "step": 9030 }, { "grad_norm": 0.1377096027135849, "learning_rate": 2.503693209498409e-06, "loss": 0.0098, "step": 9040 }, { "grad_norm": 0.08473791182041168, "learning_rate": 2.452286288035449e-06, "loss": 0.0095, "step": 9050 }, { "grad_norm": 0.05557389184832573, "learning_rate": 2.4013993407518363e-06, "loss": 0.011, "step": 9060 }, { "grad_norm": 0.14798206090927124, "learning_rate": 2.351032924139063e-06, "loss": 0.0103, "step": 9070 }, { "grad_norm": 0.12694066762924194, "learning_rate": 2.30118758899619e-06, "loss": 0.0103, "step": 9080 }, { "grad_norm": 0.06677090376615524, "learning_rate": 2.2518638804238157e-06, "loss": 0.0104, "step": 9090 }, { "grad_norm": 0.08540564030408859, "learning_rate": 2.203062337818118e-06, "loss": 0.011, "step": 9100 }, { "grad_norm": 0.1247575506567955, "learning_rate": 2.1547834948649483e-06, "loss": 0.0122, "step": 9110 }, { "grad_norm": 0.15721207857131958, "learning_rate": 2.1070278795340017e-06, "loss": 0.0111, "step": 9120 }, { "grad_norm": 0.10585247725248337, "learning_rate": 2.059796014073029e-06, "loss": 0.0077, "step": 9130 }, { "grad_norm": 0.15764175355434418, "learning_rate": 2.01308841500214e-06, "loss": 0.0119, "step": 9140 }, { "grad_norm": 0.21145889163017273, "learning_rate": 1.9669055931081704e-06, "loss": 0.0154, "step": 9150 }, { "grad_norm": 0.09447361528873444, "learning_rate": 1.9212480534390507e-06, "loss": 0.0117, "step": 9160 }, { "grad_norm": 0.21924106776714325, "learning_rate": 1.8761162952983246e-06, "loss": 0.0145, "step": 9170 }, { "grad_norm": 0.2662036120891571, "learning_rate": 1.8315108122396618e-06, "loss": 0.0127, "step": 9180 }, { "grad_norm": 0.06394612789154053, "learning_rate": 1.787432092061475e-06, "loss": 0.0086, "step": 9190 }, { "grad_norm": 0.052848123013973236, "learning_rate": 1.743880616801602e-06, "loss": 0.013, "step": 9200 }, { "grad_norm": 0.1914844959974289, "learning_rate": 1.7008568627319865e-06, "loss": 0.0103, "step": 9210 }, { "grad_norm": 0.20355594158172607, "learning_rate": 1.6583613003535226e-06, "loss": 0.0145, "step": 9220 }, { "grad_norm": 0.1600801944732666, "learning_rate": 1.6163943943908522e-06, "loss": 0.0102, "step": 9230 }, { "grad_norm": 0.1370668262243271, "learning_rate": 1.5749566037873476e-06, "loss": 0.0103, "step": 9240 }, { "grad_norm": 0.19000869989395142, "learning_rate": 1.5340483817000428e-06, "loss": 0.0129, "step": 9250 }, { "grad_norm": 0.1023942232131958, "learning_rate": 1.4936701754947101e-06, "loss": 0.0089, "step": 9260 }, { "grad_norm": 0.09788645058870316, "learning_rate": 1.4538224267409361e-06, "loss": 0.0111, "step": 9270 }, { "grad_norm": 0.10348238050937653, "learning_rate": 1.414505571207314e-06, "loss": 0.0086, "step": 9280 }, { "grad_norm": 0.1035490408539772, "learning_rate": 1.3757200388566816e-06, "loss": 0.0108, "step": 9290 }, { "grad_norm": 0.13091067969799042, "learning_rate": 1.3374662538414074e-06, "loss": 0.0124, "step": 9300 }, { "grad_norm": 0.1418973058462143, "learning_rate": 1.2997446344987617e-06, "loss": 0.0115, "step": 9310 }, { "grad_norm": 0.17107412219047546, "learning_rate": 1.262555593346315e-06, "loss": 0.0112, "step": 9320 }, { "grad_norm": 0.23288598656654358, "learning_rate": 1.2258995370774685e-06, "loss": 0.0124, "step": 9330 }, { "grad_norm": 0.12423767149448395, "learning_rate": 1.1897768665569798e-06, "loss": 0.0115, "step": 9340 }, { "grad_norm": 0.13479694724082947, "learning_rate": 1.1541879768165954e-06, "loss": 0.0092, "step": 9350 }, { "grad_norm": 0.17695064842700958, "learning_rate": 1.1191332570507085e-06, "loss": 0.0122, "step": 9360 }, { "grad_norm": 0.0883454829454422, "learning_rate": 1.0846130906121132e-06, "loss": 0.0153, "step": 9370 }, { "grad_norm": 0.10093779861927032, "learning_rate": 1.0506278550078131e-06, "loss": 0.0133, "step": 9380 }, { "grad_norm": 0.15328559279441833, "learning_rate": 1.0171779218949185e-06, "loss": 0.0129, "step": 9390 }, { "grad_norm": 0.1118769571185112, "learning_rate": 9.842636570765174e-07, "loss": 0.0083, "step": 9400 }, { "grad_norm": 0.11771470308303833, "learning_rate": 9.518854204977612e-07, "loss": 0.0094, "step": 9410 }, { "grad_norm": 0.16304083168506622, "learning_rate": 9.200435662418349e-07, "loss": 0.011, "step": 9420 }, { "grad_norm": 0.1582677662372589, "learning_rate": 8.887384425261658e-07, "loss": 0.0095, "step": 9430 }, { "grad_norm": 0.1486797332763672, "learning_rate": 8.579703916985648e-07, "loss": 0.0144, "step": 9440 }, { "grad_norm": 0.04845510795712471, "learning_rate": 8.277397502335194e-07, "loss": 0.0097, "step": 9450 }, { "grad_norm": 0.15196345746517181, "learning_rate": 7.980468487284675e-07, "loss": 0.0121, "step": 9460 }, { "grad_norm": 0.08732646703720093, "learning_rate": 7.688920119002297e-07, "loss": 0.0128, "step": 9470 }, { "grad_norm": 0.09132163971662521, "learning_rate": 7.402755585814269e-07, "loss": 0.0093, "step": 9480 }, { "grad_norm": 0.14294582605361938, "learning_rate": 7.121978017170073e-07, "loss": 0.0122, "step": 9490 }, { "grad_norm": 0.2266436368227005, "learning_rate": 6.846590483608306e-07, "loss": 0.0143, "step": 9500 }, { "grad_norm": 0.16442354023456573, "learning_rate": 6.576595996722834e-07, "loss": 0.0096, "step": 9510 }, { "grad_norm": 0.21678096055984497, "learning_rate": 6.311997509130141e-07, "loss": 0.0151, "step": 9520 }, { "grad_norm": 0.14133839309215546, "learning_rate": 6.052797914436803e-07, "loss": 0.0106, "step": 9530 }, { "grad_norm": 0.10911430418491364, "learning_rate": 5.799000047208181e-07, "loss": 0.0157, "step": 9540 }, { "grad_norm": 0.24235397577285767, "learning_rate": 5.550606682937054e-07, "loss": 0.0137, "step": 9550 }, { "grad_norm": 0.14801129698753357, "learning_rate": 5.307620538013481e-07, "loss": 0.0117, "step": 9560 }, { "grad_norm": 0.18895979225635529, "learning_rate": 5.070044269694874e-07, "loss": 0.0121, "step": 9570 }, { "grad_norm": 0.14139409363269806, "learning_rate": 4.837880476077417e-07, "loss": 0.0128, "step": 9580 }, { "grad_norm": 0.08082684874534607, "learning_rate": 4.6111316960670835e-07, "loss": 0.008, "step": 9590 }, { "grad_norm": 0.14877291023731232, "learning_rate": 4.389800409352218e-07, "loss": 0.0097, "step": 9600 }, { "grad_norm": 0.21980130672454834, "learning_rate": 4.173889036376277e-07, "loss": 0.0145, "step": 9610 }, { "grad_norm": 0.1517840325832367, "learning_rate": 3.963399938311463e-07, "loss": 0.0121, "step": 9620 }, { "grad_norm": 0.15636564791202545, "learning_rate": 3.7583354170328545e-07, "loss": 0.0126, "step": 9630 }, { "grad_norm": 0.07844378799200058, "learning_rate": 3.558697715093207e-07, "loss": 0.0077, "step": 9640 }, { "grad_norm": 0.19841763377189636, "learning_rate": 3.3644890156983576e-07, "loss": 0.0126, "step": 9650 }, { "grad_norm": 0.11830093711614609, "learning_rate": 3.175711442683638e-07, "loss": 0.0097, "step": 9660 }, { "grad_norm": 0.10069432109594345, "learning_rate": 2.9923670604902197e-07, "loss": 0.0112, "step": 9670 }, { "grad_norm": 0.18374690413475037, "learning_rate": 2.814457874143028e-07, "loss": 0.0114, "step": 9680 }, { "grad_norm": 0.09517468512058258, "learning_rate": 2.641985829228366e-07, "loss": 0.0095, "step": 9690 }, { "grad_norm": 0.10329031199216843, "learning_rate": 2.474952811872877e-07, "loss": 0.01, "step": 9700 }, { "grad_norm": 0.14360502362251282, "learning_rate": 2.3133606487228397e-07, "loss": 0.0131, "step": 9710 }, { "grad_norm": 0.1737045794725418, "learning_rate": 2.157211106924295e-07, "loss": 0.0123, "step": 9720 }, { "grad_norm": 0.165406733751297, "learning_rate": 2.006505894103672e-07, "loss": 0.0119, "step": 9730 }, { "grad_norm": 0.09217849373817444, "learning_rate": 1.8612466583489696e-07, "loss": 0.0106, "step": 9740 }, { "grad_norm": 0.1349356770515442, "learning_rate": 1.7214349881918834e-07, "loss": 0.0112, "step": 9750 }, { "grad_norm": 0.1287662237882614, "learning_rate": 1.5870724125904845e-07, "loss": 0.0087, "step": 9760 }, { "grad_norm": 0.09644419699907303, "learning_rate": 1.4581604009124006e-07, "loss": 0.0108, "step": 9770 }, { "grad_norm": 0.155265673995018, "learning_rate": 1.334700362918717e-07, "loss": 0.0112, "step": 9780 }, { "grad_norm": 0.1518457531929016, "learning_rate": 1.2166936487486015e-07, "loss": 0.0138, "step": 9790 }, { "grad_norm": 0.10346577316522598, "learning_rate": 1.1041415489045914e-07, "loss": 0.0075, "step": 9800 }, { "grad_norm": 0.14470964670181274, "learning_rate": 9.970452942384412e-08, "loss": 0.0109, "step": 9810 }, { "grad_norm": 0.13003452122211456, "learning_rate": 8.954060559375754e-08, "loss": 0.0084, "step": 9820 }, { "grad_norm": 0.10539904981851578, "learning_rate": 7.99224945512489e-08, "loss": 0.0128, "step": 9830 }, { "grad_norm": 0.09676070511341095, "learning_rate": 7.085030147843675e-08, "loss": 0.013, "step": 9840 }, { "grad_norm": 0.15782088041305542, "learning_rate": 6.232412558736523e-08, "loss": 0.0133, "step": 9850 }, { "grad_norm": 0.1917269378900528, "learning_rate": 5.434406011893822e-08, "loss": 0.0137, "step": 9860 }, { "grad_norm": 0.15100739896297455, "learning_rate": 4.6910192341864664e-08, "loss": 0.0096, "step": 9870 }, { "grad_norm": 0.15185561776161194, "learning_rate": 4.0022603551737035e-08, "loss": 0.0123, "step": 9880 }, { "grad_norm": 0.19557790458202362, "learning_rate": 3.3681369070120985e-08, "loss": 0.0148, "step": 9890 }, { "grad_norm": 0.11138531565666199, "learning_rate": 2.7886558243744866e-08, "loss": 0.0124, "step": 9900 }, { "grad_norm": 0.1685207486152649, "learning_rate": 2.2638234443722596e-08, "loss": 0.013, "step": 9910 }, { "grad_norm": 0.2191893458366394, "learning_rate": 1.7936455064887504e-08, "loss": 0.0136, "step": 9920 }, { "grad_norm": 0.11110971868038177, "learning_rate": 1.378127152514841e-08, "loss": 0.0074, "step": 9930 }, { "grad_norm": 0.12800949811935425, "learning_rate": 1.0172729264917857e-08, "loss": 0.0139, "step": 9940 }, { "grad_norm": 0.18108631670475006, "learning_rate": 7.1108677466458215e-09, "loss": 0.0128, "step": 9950 }, { "grad_norm": 0.13491609692573547, "learning_rate": 4.595720454353414e-09, "loss": 0.0122, "step": 9960 }, { "grad_norm": 0.1606258898973465, "learning_rate": 2.627314893294264e-09, "loss": 0.0109, "step": 9970 }, { "grad_norm": 0.11117072403430939, "learning_rate": 1.2056725896270048e-09, "loss": 0.0118, "step": 9980 }, { "grad_norm": 0.1367674618959427, "learning_rate": 3.308090902098826e-10, "loss": 0.0118, "step": 9990 }, { "grad_norm": 0.13445894420146942, "learning_rate": 2.7339624120159555e-12, "loss": 0.0144, "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 }