{ "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": 2.4938712120056152, "learning_rate": 1.8e-06, "loss": 1.149, "step": 10 }, { "grad_norm": 1.4163585901260376, "learning_rate": 3.8e-06, "loss": 1.1247, "step": 20 }, { "grad_norm": 0.5394372940063477, "learning_rate": 5.8e-06, "loss": 1.103, "step": 30 }, { "grad_norm": 0.9632188081741333, "learning_rate": 7.8e-06, "loss": 1.1119, "step": 40 }, { "grad_norm": 0.5876862406730652, "learning_rate": 9.800000000000001e-06, "loss": 1.0779, "step": 50 }, { "grad_norm": 0.48418310284614563, "learning_rate": 1.18e-05, "loss": 1.0618, "step": 60 }, { "grad_norm": 0.7189286351203918, "learning_rate": 1.3800000000000002e-05, "loss": 1.0718, "step": 70 }, { "grad_norm": 0.36253997683525085, "learning_rate": 1.58e-05, "loss": 1.0688, "step": 80 }, { "grad_norm": 0.403954416513443, "learning_rate": 1.78e-05, "loss": 1.0615, "step": 90 }, { "grad_norm": 0.5885539054870605, "learning_rate": 1.9800000000000004e-05, "loss": 1.0402, "step": 100 }, { "grad_norm": 0.6073641180992126, "learning_rate": 2.18e-05, "loss": 1.0305, "step": 110 }, { "grad_norm": 0.6402820348739624, "learning_rate": 2.38e-05, "loss": 1.0306, "step": 120 }, { "grad_norm": 0.48018601536750793, "learning_rate": 2.58e-05, "loss": 1.0276, "step": 130 }, { "grad_norm": 0.7982405424118042, "learning_rate": 2.7800000000000005e-05, "loss": 1.011, "step": 140 }, { "grad_norm": 1.7149966955184937, "learning_rate": 2.98e-05, "loss": 0.9281, "step": 150 }, { "grad_norm": 1.5796239376068115, "learning_rate": 3.18e-05, "loss": 0.8816, "step": 160 }, { "grad_norm": 1.9514416456222534, "learning_rate": 3.38e-05, "loss": 0.7236, "step": 170 }, { "grad_norm": 2.136714220046997, "learning_rate": 3.58e-05, "loss": 0.5758, "step": 180 }, { "grad_norm": 2.2087130546569824, "learning_rate": 3.7800000000000004e-05, "loss": 0.3993, "step": 190 }, { "grad_norm": 2.7446579933166504, "learning_rate": 3.9800000000000005e-05, "loss": 0.3111, "step": 200 }, { "grad_norm": 2.1047074794769287, "learning_rate": 4.18e-05, "loss": 0.2648, "step": 210 }, { "grad_norm": 2.1818909645080566, "learning_rate": 4.38e-05, "loss": 0.2288, "step": 220 }, { "grad_norm": 1.5412745475769043, "learning_rate": 4.58e-05, "loss": 0.2029, "step": 230 }, { "grad_norm": 1.5120090246200562, "learning_rate": 4.78e-05, "loss": 0.1678, "step": 240 }, { "grad_norm": 1.2502282857894897, "learning_rate": 4.9800000000000004e-05, "loss": 0.1532, "step": 250 }, { "grad_norm": 1.2709970474243164, "learning_rate": 5.1800000000000005e-05, "loss": 0.1447, "step": 260 }, { "grad_norm": 1.2660635709762573, "learning_rate": 5.380000000000001e-05, "loss": 0.1659, "step": 270 }, { "grad_norm": 0.9219812154769897, "learning_rate": 5.580000000000001e-05, "loss": 0.1404, "step": 280 }, { "grad_norm": 1.1851774454116821, "learning_rate": 5.7799999999999995e-05, "loss": 0.1534, "step": 290 }, { "grad_norm": 1.2766985893249512, "learning_rate": 5.9800000000000003e-05, "loss": 0.1359, "step": 300 }, { "grad_norm": 1.3988244533538818, "learning_rate": 6.18e-05, "loss": 0.1505, "step": 310 }, { "grad_norm": 0.8213768601417542, "learning_rate": 6.38e-05, "loss": 0.1179, "step": 320 }, { "grad_norm": 1.0496826171875, "learning_rate": 6.58e-05, "loss": 0.1187, "step": 330 }, { "grad_norm": 0.9927864074707031, "learning_rate": 6.780000000000001e-05, "loss": 0.1105, "step": 340 }, { "grad_norm": 0.821165144443512, "learning_rate": 6.98e-05, "loss": 0.1026, "step": 350 }, { "grad_norm": 1.1925801038742065, "learning_rate": 7.18e-05, "loss": 0.1208, "step": 360 }, { "grad_norm": 0.8459774851799011, "learning_rate": 7.38e-05, "loss": 0.1313, "step": 370 }, { "grad_norm": 1.1768630743026733, "learning_rate": 7.58e-05, "loss": 0.1115, "step": 380 }, { "grad_norm": 0.6162320375442505, "learning_rate": 7.780000000000001e-05, "loss": 0.1196, "step": 390 }, { "grad_norm": 0.7858737111091614, "learning_rate": 7.98e-05, "loss": 0.1063, "step": 400 }, { "grad_norm": 0.5725531578063965, "learning_rate": 8.18e-05, "loss": 0.1099, "step": 410 }, { "grad_norm": 0.8981689810752869, "learning_rate": 8.38e-05, "loss": 0.1083, "step": 420 }, { "grad_norm": 0.797149658203125, "learning_rate": 8.58e-05, "loss": 0.1086, "step": 430 }, { "grad_norm": 1.0583909749984741, "learning_rate": 8.78e-05, "loss": 0.1075, "step": 440 }, { "grad_norm": 0.6379928588867188, "learning_rate": 8.98e-05, "loss": 0.105, "step": 450 }, { "grad_norm": 0.8783487677574158, "learning_rate": 9.180000000000001e-05, "loss": 0.1051, "step": 460 }, { "grad_norm": 0.6223819255828857, "learning_rate": 9.38e-05, "loss": 0.1133, "step": 470 }, { "grad_norm": 0.6023280024528503, "learning_rate": 9.58e-05, "loss": 0.1106, "step": 480 }, { "grad_norm": 1.0111503601074219, "learning_rate": 9.78e-05, "loss": 0.0996, "step": 490 }, { "grad_norm": 0.7443103790283203, "learning_rate": 9.98e-05, "loss": 0.1029, "step": 500 }, { "grad_norm": 0.8028576374053955, "learning_rate": 9.9999778549206e-05, "loss": 0.1058, "step": 510 }, { "grad_norm": 0.754014253616333, "learning_rate": 9.999901304280685e-05, "loss": 0.0938, "step": 520 }, { "grad_norm": 0.6088759303092957, "learning_rate": 9.999770075521164e-05, "loss": 0.1001, "step": 530 }, { "grad_norm": 0.6760469675064087, "learning_rate": 9.99958417007713e-05, "loss": 0.091, "step": 540 }, { "grad_norm": 0.45930418372154236, "learning_rate": 9.999343589981615e-05, "loss": 0.0931, "step": 550 }, { "grad_norm": 0.4519936740398407, "learning_rate": 9.999048337865568e-05, "loss": 0.0844, "step": 560 }, { "grad_norm": 0.7299916744232178, "learning_rate": 9.998698416957815e-05, "loss": 0.0817, "step": 570 }, { "grad_norm": 0.6385707855224609, "learning_rate": 9.998293831085037e-05, "loss": 0.0864, "step": 580 }, { "grad_norm": 0.6345222592353821, "learning_rate": 9.997834584671719e-05, "loss": 0.0883, "step": 590 }, { "grad_norm": 0.554981529712677, "learning_rate": 9.997320682740107e-05, "loss": 0.0876, "step": 600 }, { "grad_norm": 0.6528695821762085, "learning_rate": 9.996752130910149e-05, "loss": 0.098, "step": 610 }, { "grad_norm": 0.7737570405006409, "learning_rate": 9.99612893539944e-05, "loss": 0.0878, "step": 620 }, { "grad_norm": 0.4518269896507263, "learning_rate": 9.995451103023144e-05, "loss": 0.0744, "step": 630 }, { "grad_norm": 0.3672517240047455, "learning_rate": 9.994718641193928e-05, "loss": 0.0755, "step": 640 }, { "grad_norm": 0.45259639620780945, "learning_rate": 9.993931557921874e-05, "loss": 0.0786, "step": 650 }, { "grad_norm": 0.571256697177887, "learning_rate": 9.993089861814402e-05, "loss": 0.0791, "step": 660 }, { "grad_norm": 0.7500366568565369, "learning_rate": 9.992193562076166e-05, "loss": 0.0935, "step": 670 }, { "grad_norm": 0.5490401387214661, "learning_rate": 9.991242668508954e-05, "loss": 0.089, "step": 680 }, { "grad_norm": 0.589253306388855, "learning_rate": 9.990237191511587e-05, "loss": 0.0763, "step": 690 }, { "grad_norm": 0.4584391117095947, "learning_rate": 9.989177142079802e-05, "loss": 0.0802, "step": 700 }, { "grad_norm": 0.522132933139801, "learning_rate": 9.988062531806126e-05, "loss": 0.0869, "step": 710 }, { "grad_norm": 0.46739551424980164, "learning_rate": 9.986893372879762e-05, "loss": 0.0725, "step": 720 }, { "grad_norm": 0.46067553758621216, "learning_rate": 9.985669678086443e-05, "loss": 0.0747, "step": 730 }, { "grad_norm": 0.6948791742324829, "learning_rate": 9.984391460808298e-05, "loss": 0.0612, "step": 740 }, { "grad_norm": 0.44356176257133484, "learning_rate": 9.983058735023709e-05, "loss": 0.0728, "step": 750 }, { "grad_norm": 0.4867616891860962, "learning_rate": 9.98167151530715e-05, "loss": 0.0643, "step": 760 }, { "grad_norm": 0.5087238550186157, "learning_rate": 9.980229816829034e-05, "loss": 0.0727, "step": 770 }, { "grad_norm": 0.3429226875305176, "learning_rate": 9.978733655355544e-05, "loss": 0.0673, "step": 780 }, { "grad_norm": 0.3515297472476959, "learning_rate": 9.977183047248464e-05, "loss": 0.0648, "step": 790 }, { "grad_norm": 0.5170158743858337, "learning_rate": 9.975578009464992e-05, "loss": 0.073, "step": 800 }, { "grad_norm": 0.32225289940834045, "learning_rate": 9.97391855955757e-05, "loss": 0.0757, "step": 810 }, { "grad_norm": 0.5070093274116516, "learning_rate": 9.972204715673669e-05, "loss": 0.0713, "step": 820 }, { "grad_norm": 0.4836777150630951, "learning_rate": 9.970436496555617e-05, "loss": 0.0704, "step": 830 }, { "grad_norm": 0.5138593912124634, "learning_rate": 9.968613921540373e-05, "loss": 0.0654, "step": 840 }, { "grad_norm": 0.3828106224536896, "learning_rate": 9.966737010559326e-05, "loss": 0.0812, "step": 850 }, { "grad_norm": 0.425285667181015, "learning_rate": 9.964805784138072e-05, "loss": 0.068, "step": 860 }, { "grad_norm": 0.4094374179840088, "learning_rate": 9.962820263396195e-05, "loss": 0.0738, "step": 870 }, { "grad_norm": 0.48152482509613037, "learning_rate": 9.960780470047033e-05, "loss": 0.0692, "step": 880 }, { "grad_norm": 0.39264413714408875, "learning_rate": 9.958686426397437e-05, "loss": 0.0595, "step": 890 }, { "grad_norm": 0.3702973425388336, "learning_rate": 9.956538155347534e-05, "loss": 0.0728, "step": 900 }, { "grad_norm": 0.3418523669242859, "learning_rate": 9.95433568039047e-05, "loss": 0.0588, "step": 910 }, { "grad_norm": 0.41219237446784973, "learning_rate": 9.952079025612162e-05, "loss": 0.0693, "step": 920 }, { "grad_norm": 0.48037222027778625, "learning_rate": 9.949768215691022e-05, "loss": 0.0604, "step": 930 }, { "grad_norm": 0.4843182861804962, "learning_rate": 9.9474032758977e-05, "loss": 0.0597, "step": 940 }, { "grad_norm": 0.45763182640075684, "learning_rate": 9.944984232094794e-05, "loss": 0.0726, "step": 950 }, { "grad_norm": 0.46674877405166626, "learning_rate": 9.942511110736584e-05, "loss": 0.0573, "step": 960 }, { "grad_norm": 0.3402610421180725, "learning_rate": 9.939983938868726e-05, "loss": 0.0578, "step": 970 }, { "grad_norm": 0.514228880405426, "learning_rate": 9.93740274412797e-05, "loss": 0.0633, "step": 980 }, { "grad_norm": 0.5390498042106628, "learning_rate": 9.934767554741846e-05, "loss": 0.0689, "step": 990 }, { "grad_norm": 0.4683254361152649, "learning_rate": 9.932078399528361e-05, "loss": 0.0626, "step": 1000 }, { "grad_norm": 0.4931696951389313, "learning_rate": 9.929335307895689e-05, "loss": 0.0572, "step": 1010 }, { "grad_norm": 0.41609814763069153, "learning_rate": 9.926538309841839e-05, "loss": 0.0625, "step": 1020 }, { "grad_norm": 0.485098659992218, "learning_rate": 9.923687435954334e-05, "loss": 0.0663, "step": 1030 }, { "grad_norm": 0.43627384305000305, "learning_rate": 9.920782717409873e-05, "loss": 0.0656, "step": 1040 }, { "grad_norm": 0.6335458159446716, "learning_rate": 9.917824185973994e-05, "loss": 0.0622, "step": 1050 }, { "grad_norm": 0.4830600917339325, "learning_rate": 9.914811874000723e-05, "loss": 0.0725, "step": 1060 }, { "grad_norm": 0.3551991581916809, "learning_rate": 9.911745814432218e-05, "loss": 0.0649, "step": 1070 }, { "grad_norm": 0.6572110652923584, "learning_rate": 9.90862604079842e-05, "loss": 0.0667, "step": 1080 }, { "grad_norm": 0.31550419330596924, "learning_rate": 9.90545258721667e-05, "loss": 0.0483, "step": 1090 }, { "grad_norm": 0.424724280834198, "learning_rate": 9.90222548839135e-05, "loss": 0.0577, "step": 1100 }, { "grad_norm": 0.411609411239624, "learning_rate": 9.898944779613495e-05, "loss": 0.0582, "step": 1110 }, { "grad_norm": 0.403708279132843, "learning_rate": 9.89561049676041e-05, "loss": 0.0604, "step": 1120 }, { "grad_norm": 0.524554431438446, "learning_rate": 9.89222267629528e-05, "loss": 0.0625, "step": 1130 }, { "grad_norm": 0.4202245771884918, "learning_rate": 9.888781355266763e-05, "loss": 0.0613, "step": 1140 }, { "grad_norm": 0.25448477268218994, "learning_rate": 9.885286571308598e-05, "loss": 0.0579, "step": 1150 }, { "grad_norm": 0.3334948420524597, "learning_rate": 9.881738362639182e-05, "loss": 0.0561, "step": 1160 }, { "grad_norm": 0.3942888379096985, "learning_rate": 9.878136768061154e-05, "loss": 0.0561, "step": 1170 }, { "grad_norm": 0.4101827144622803, "learning_rate": 9.874481826960979e-05, "loss": 0.0538, "step": 1180 }, { "grad_norm": 0.49180328845977783, "learning_rate": 9.870773579308503e-05, "loss": 0.0551, "step": 1190 }, { "grad_norm": 0.44954514503479004, "learning_rate": 9.867012065656533e-05, "loss": 0.062, "step": 1200 }, { "grad_norm": 0.420514851808548, "learning_rate": 9.863197327140376e-05, "loss": 0.0578, "step": 1210 }, { "grad_norm": 0.486957848072052, "learning_rate": 9.859329405477403e-05, "loss": 0.0527, "step": 1220 }, { "grad_norm": 0.4154336750507355, "learning_rate": 9.855408342966585e-05, "loss": 0.0598, "step": 1230 }, { "grad_norm": 0.3420454263687134, "learning_rate": 9.851434182488033e-05, "loss": 0.0539, "step": 1240 }, { "grad_norm": 0.3399484157562256, "learning_rate": 9.84740696750253e-05, "loss": 0.0613, "step": 1250 }, { "grad_norm": 0.33110183477401733, "learning_rate": 9.843326742051055e-05, "loss": 0.0537, "step": 1260 }, { "grad_norm": 0.3873280882835388, "learning_rate": 9.839193550754297e-05, "loss": 0.0633, "step": 1270 }, { "grad_norm": 0.4030843675136566, "learning_rate": 9.835007438812177e-05, "loss": 0.0567, "step": 1280 }, { "grad_norm": 0.384939581155777, "learning_rate": 9.830768452003341e-05, "loss": 0.0637, "step": 1290 }, { "grad_norm": 0.36267516016960144, "learning_rate": 9.826476636684671e-05, "loss": 0.0556, "step": 1300 }, { "grad_norm": 0.4910261631011963, "learning_rate": 9.822132039790773e-05, "loss": 0.0597, "step": 1310 }, { "grad_norm": 0.40302979946136475, "learning_rate": 9.817734708833461e-05, "loss": 0.0658, "step": 1320 }, { "grad_norm": 0.3806966543197632, "learning_rate": 9.813284691901243e-05, "loss": 0.0547, "step": 1330 }, { "grad_norm": 0.5657820701599121, "learning_rate": 9.808782037658792e-05, "loss": 0.0582, "step": 1340 }, { "grad_norm": 0.4249174892902374, "learning_rate": 9.804226795346411e-05, "loss": 0.0525, "step": 1350 }, { "grad_norm": 0.3852368891239166, "learning_rate": 9.799619014779503e-05, "loss": 0.054, "step": 1360 }, { "grad_norm": 0.30641093850135803, "learning_rate": 9.794958746348013e-05, "loss": 0.0451, "step": 1370 }, { "grad_norm": 0.3716799020767212, "learning_rate": 9.790246041015896e-05, "loss": 0.0505, "step": 1380 }, { "grad_norm": 0.4790414571762085, "learning_rate": 9.785480950320538e-05, "loss": 0.0497, "step": 1390 }, { "grad_norm": 0.37867796421051025, "learning_rate": 9.78066352637221e-05, "loss": 0.0544, "step": 1400 }, { "grad_norm": 0.5180774927139282, "learning_rate": 9.775793821853488e-05, "loss": 0.0485, "step": 1410 }, { "grad_norm": 0.361497163772583, "learning_rate": 9.77087189001868e-05, "loss": 0.0548, "step": 1420 }, { "grad_norm": 0.44058680534362793, "learning_rate": 9.765897784693243e-05, "loss": 0.0504, "step": 1430 }, { "grad_norm": 0.377067506313324, "learning_rate": 9.760871560273197e-05, "loss": 0.0531, "step": 1440 }, { "grad_norm": 0.3615536689758301, "learning_rate": 9.755793271724526e-05, "loss": 0.0559, "step": 1450 }, { "grad_norm": 0.24905550479888916, "learning_rate": 9.750662974582584e-05, "loss": 0.0534, "step": 1460 }, { "grad_norm": 0.3822507858276367, "learning_rate": 9.745480724951473e-05, "loss": 0.0524, "step": 1470 }, { "grad_norm": 0.40116965770721436, "learning_rate": 9.740246579503447e-05, "loss": 0.0588, "step": 1480 }, { "grad_norm": 0.4540383219718933, "learning_rate": 9.734960595478284e-05, "loss": 0.0491, "step": 1490 }, { "grad_norm": 0.35509923100471497, "learning_rate": 9.729622830682657e-05, "loss": 0.0595, "step": 1500 }, { "grad_norm": 0.3072110116481781, "learning_rate": 9.724233343489504e-05, "loss": 0.0498, "step": 1510 }, { "grad_norm": 0.4348505437374115, "learning_rate": 9.718792192837396e-05, "loss": 0.0584, "step": 1520 }, { "grad_norm": 0.4681381285190582, "learning_rate": 9.713299438229886e-05, "loss": 0.0565, "step": 1530 }, { "grad_norm": 0.43342989683151245, "learning_rate": 9.707755139734855e-05, "loss": 0.0612, "step": 1540 }, { "grad_norm": 0.4543539881706238, "learning_rate": 9.702159357983866e-05, "loss": 0.0537, "step": 1550 }, { "grad_norm": 0.3873662054538727, "learning_rate": 9.696512154171492e-05, "loss": 0.0547, "step": 1560 }, { "grad_norm": 0.33500927686691284, "learning_rate": 9.690813590054645e-05, "loss": 0.0493, "step": 1570 }, { "grad_norm": 0.5049640536308289, "learning_rate": 9.685063727951914e-05, "loss": 0.0478, "step": 1580 }, { "grad_norm": 0.4916016459465027, "learning_rate": 9.679262630742865e-05, "loss": 0.051, "step": 1590 }, { "grad_norm": 0.407440721988678, "learning_rate": 9.673410361867373e-05, "loss": 0.0504, "step": 1600 }, { "grad_norm": 0.30330148339271545, "learning_rate": 9.667506985324909e-05, "loss": 0.0438, "step": 1610 }, { "grad_norm": 0.2702285349369049, "learning_rate": 9.661552565673855e-05, "loss": 0.0408, "step": 1620 }, { "grad_norm": 0.3102802038192749, "learning_rate": 9.655547168030789e-05, "loss": 0.0453, "step": 1630 }, { "grad_norm": 0.41987520456314087, "learning_rate": 9.649490858069777e-05, "loss": 0.0486, "step": 1640 }, { "grad_norm": 0.4578944146633148, "learning_rate": 9.643383702021658e-05, "loss": 0.0546, "step": 1650 }, { "grad_norm": 0.2824367582798004, "learning_rate": 9.637225766673307e-05, "loss": 0.041, "step": 1660 }, { "grad_norm": 0.41557398438453674, "learning_rate": 9.631017119366922e-05, "loss": 0.0507, "step": 1670 }, { "grad_norm": 0.4481380581855774, "learning_rate": 9.624757827999273e-05, "loss": 0.0523, "step": 1680 }, { "grad_norm": 0.2056555151939392, "learning_rate": 9.618447961020971e-05, "loss": 0.0532, "step": 1690 }, { "grad_norm": 0.36813294887542725, "learning_rate": 9.612087587435707e-05, "loss": 0.0483, "step": 1700 }, { "grad_norm": 0.30924245715141296, "learning_rate": 9.605676776799508e-05, "loss": 0.0474, "step": 1710 }, { "grad_norm": 0.3423722982406616, "learning_rate": 9.599215599219973e-05, "loss": 0.0477, "step": 1720 }, { "grad_norm": 0.3364288806915283, "learning_rate": 9.592704125355505e-05, "loss": 0.0471, "step": 1730 }, { "grad_norm": 0.3145952522754669, "learning_rate": 9.586142426414538e-05, "loss": 0.0462, "step": 1740 }, { "grad_norm": 0.42746543884277344, "learning_rate": 9.57953057415476e-05, "loss": 0.0478, "step": 1750 }, { "grad_norm": 0.39381465315818787, "learning_rate": 9.572868640882328e-05, "loss": 0.0554, "step": 1760 }, { "grad_norm": 0.29161179065704346, "learning_rate": 9.56615669945108e-05, "loss": 0.0484, "step": 1770 }, { "grad_norm": 0.32075440883636475, "learning_rate": 9.55939482326173e-05, "loss": 0.0455, "step": 1780 }, { "grad_norm": 0.39323264360427856, "learning_rate": 9.552583086261069e-05, "loss": 0.0528, "step": 1790 }, { "grad_norm": 0.3446557819843292, "learning_rate": 9.545721562941168e-05, "loss": 0.0455, "step": 1800 }, { "grad_norm": 0.35655298829078674, "learning_rate": 9.538810328338543e-05, "loss": 0.0435, "step": 1810 }, { "grad_norm": 0.28898605704307556, "learning_rate": 9.531849458033349e-05, "loss": 0.0423, "step": 1820 }, { "grad_norm": 0.3521408438682556, "learning_rate": 9.524839028148547e-05, "loss": 0.0421, "step": 1830 }, { "grad_norm": 0.34939441084861755, "learning_rate": 9.517779115349077e-05, "loss": 0.0447, "step": 1840 }, { "grad_norm": 0.2448209971189499, "learning_rate": 9.510669796841014e-05, "loss": 0.0437, "step": 1850 }, { "grad_norm": 0.3223055601119995, "learning_rate": 9.503511150370727e-05, "loss": 0.0484, "step": 1860 }, { "grad_norm": 0.29390156269073486, "learning_rate": 9.496303254224024e-05, "loss": 0.0468, "step": 1870 }, { "grad_norm": 0.30412185192108154, "learning_rate": 9.489046187225306e-05, "loss": 0.0385, "step": 1880 }, { "grad_norm": 0.36074450612068176, "learning_rate": 9.481740028736692e-05, "loss": 0.0389, "step": 1890 }, { "grad_norm": 0.33187755942344666, "learning_rate": 9.474384858657164e-05, "loss": 0.0507, "step": 1900 }, { "grad_norm": 0.32691606879234314, "learning_rate": 9.466980757421679e-05, "loss": 0.0517, "step": 1910 }, { "grad_norm": 0.5117631554603577, "learning_rate": 9.459527806000305e-05, "loss": 0.0475, "step": 1920 }, { "grad_norm": 0.41090837121009827, "learning_rate": 9.452026085897325e-05, "loss": 0.0494, "step": 1930 }, { "grad_norm": 0.3536020517349243, "learning_rate": 9.444475679150348e-05, "loss": 0.0455, "step": 1940 }, { "grad_norm": 0.37706005573272705, "learning_rate": 9.436876668329411e-05, "loss": 0.0453, "step": 1950 }, { "grad_norm": 0.2635892927646637, "learning_rate": 9.429229136536079e-05, "loss": 0.0367, "step": 1960 }, { "grad_norm": 0.3455093204975128, "learning_rate": 9.421533167402534e-05, "loss": 0.0493, "step": 1970 }, { "grad_norm": 0.33207792043685913, "learning_rate": 9.413788845090666e-05, "loss": 0.0531, "step": 1980 }, { "grad_norm": 0.36056169867515564, "learning_rate": 9.405996254291136e-05, "loss": 0.0421, "step": 1990 }, { "grad_norm": 0.3565305471420288, "learning_rate": 9.398155480222474e-05, "loss": 0.0407, "step": 2000 }, { "grad_norm": 0.31847113370895386, "learning_rate": 9.390266608630128e-05, "loss": 0.0417, "step": 2010 }, { "grad_norm": 0.37380245327949524, "learning_rate": 9.38232972578553e-05, "loss": 0.0452, "step": 2020 }, { "grad_norm": 0.29371991753578186, "learning_rate": 9.374344918485164e-05, "loss": 0.0503, "step": 2030 }, { "grad_norm": 0.27426791191101074, "learning_rate": 9.366312274049602e-05, "loss": 0.052, "step": 2040 }, { "grad_norm": 0.2717961370944977, "learning_rate": 9.358231880322554e-05, "loss": 0.0434, "step": 2050 }, { "grad_norm": 0.3620748221874237, "learning_rate": 9.350103825669916e-05, "loss": 0.0435, "step": 2060 }, { "grad_norm": 0.4252539575099945, "learning_rate": 9.341928198978787e-05, "loss": 0.0406, "step": 2070 }, { "grad_norm": 0.3430294394493103, "learning_rate": 9.333705089656512e-05, "loss": 0.0403, "step": 2080 }, { "grad_norm": 0.42273756861686707, "learning_rate": 9.325434587629698e-05, "loss": 0.0426, "step": 2090 }, { "grad_norm": 0.358388751745224, "learning_rate": 9.31711678334323e-05, "loss": 0.0422, "step": 2100 }, { "grad_norm": 0.3798494040966034, "learning_rate": 9.308751767759282e-05, "loss": 0.0433, "step": 2110 }, { "grad_norm": 0.23689408600330353, "learning_rate": 9.300339632356325e-05, "loss": 0.0363, "step": 2120 }, { "grad_norm": 0.314285546541214, "learning_rate": 9.291880469128124e-05, "loss": 0.0448, "step": 2130 }, { "grad_norm": 0.3283476233482361, "learning_rate": 9.283374370582732e-05, "loss": 0.0456, "step": 2140 }, { "grad_norm": 0.3064684271812439, "learning_rate": 9.274821429741482e-05, "loss": 0.0487, "step": 2150 }, { "grad_norm": 0.2701602876186371, "learning_rate": 9.266221740137961e-05, "loss": 0.0438, "step": 2160 }, { "grad_norm": 0.22183829545974731, "learning_rate": 9.257575395817001e-05, "loss": 0.039, "step": 2170 }, { "grad_norm": 0.3263572156429291, "learning_rate": 9.248882491333637e-05, "loss": 0.0415, "step": 2180 }, { "grad_norm": 0.3883651793003082, "learning_rate": 9.240143121752076e-05, "loss": 0.0381, "step": 2190 }, { "grad_norm": 0.22430305182933807, "learning_rate": 9.23135738264467e-05, "loss": 0.0421, "step": 2200 }, { "grad_norm": 0.26308751106262207, "learning_rate": 9.222525370090849e-05, "loss": 0.0372, "step": 2210 }, { "grad_norm": 0.3491934537887573, "learning_rate": 9.213647180676088e-05, "loss": 0.037, "step": 2220 }, { "grad_norm": 0.399991512298584, "learning_rate": 9.204722911490846e-05, "loss": 0.0439, "step": 2230 }, { "grad_norm": 0.24338486790657043, "learning_rate": 9.1957526601295e-05, "loss": 0.0391, "step": 2240 }, { "grad_norm": 0.27984070777893066, "learning_rate": 9.186736524689281e-05, "loss": 0.0429, "step": 2250 }, { "grad_norm": 0.3628758192062378, "learning_rate": 9.177674603769204e-05, "loss": 0.045, "step": 2260 }, { "grad_norm": 0.2437887042760849, "learning_rate": 9.168566996468983e-05, "loss": 0.0367, "step": 2270 }, { "grad_norm": 0.26792553067207336, "learning_rate": 9.159413802387951e-05, "loss": 0.0445, "step": 2280 }, { "grad_norm": 0.2326800376176834, "learning_rate": 9.150215121623974e-05, "loss": 0.0413, "step": 2290 }, { "grad_norm": 0.226214200258255, "learning_rate": 9.140971054772349e-05, "loss": 0.0335, "step": 2300 }, { "grad_norm": 0.29084980487823486, "learning_rate": 9.131681702924713e-05, "loss": 0.0444, "step": 2310 }, { "grad_norm": 0.29175591468811035, "learning_rate": 9.122347167667926e-05, "loss": 0.0333, "step": 2320 }, { "grad_norm": 0.2684953510761261, "learning_rate": 9.112967551082973e-05, "loss": 0.0448, "step": 2330 }, { "grad_norm": 0.3047690689563751, "learning_rate": 9.103542955743835e-05, "loss": 0.0372, "step": 2340 }, { "grad_norm": 0.3676929175853729, "learning_rate": 9.094073484716381e-05, "loss": 0.0437, "step": 2350 }, { "grad_norm": 0.3885331451892853, "learning_rate": 9.084559241557226e-05, "loss": 0.0383, "step": 2360 }, { "grad_norm": 0.13384123146533966, "learning_rate": 9.075000330312608e-05, "loss": 0.0367, "step": 2370 }, { "grad_norm": 0.2833608090877533, "learning_rate": 9.065396855517253e-05, "loss": 0.0338, "step": 2380 }, { "grad_norm": 0.28043001890182495, "learning_rate": 9.055748922193219e-05, "loss": 0.041, "step": 2390 }, { "grad_norm": 0.20895858108997345, "learning_rate": 9.046056635848761e-05, "loss": 0.0399, "step": 2400 }, { "grad_norm": 0.2684876620769501, "learning_rate": 9.036320102477169e-05, "loss": 0.0347, "step": 2410 }, { "grad_norm": 0.3110980689525604, "learning_rate": 9.02653942855561e-05, "loss": 0.0374, "step": 2420 }, { "grad_norm": 0.36423078179359436, "learning_rate": 9.016714721043971e-05, "loss": 0.0364, "step": 2430 }, { "grad_norm": 0.24248777329921722, "learning_rate": 9.006846087383675e-05, "loss": 0.0426, "step": 2440 }, { "grad_norm": 0.22214096784591675, "learning_rate": 8.996933635496523e-05, "loss": 0.0361, "step": 2450 }, { "grad_norm": 0.2939525246620178, "learning_rate": 8.986977473783498e-05, "loss": 0.0415, "step": 2460 }, { "grad_norm": 0.2388240396976471, "learning_rate": 8.97697771112359e-05, "loss": 0.036, "step": 2470 }, { "grad_norm": 0.3107961118221283, "learning_rate": 8.966934456872602e-05, "loss": 0.035, "step": 2480 }, { "grad_norm": 0.20910818874835968, "learning_rate": 8.95684782086195e-05, "loss": 0.0323, "step": 2490 }, { "grad_norm": 0.42585983872413635, "learning_rate": 8.946717913397476e-05, "loss": 0.0413, "step": 2500 }, { "grad_norm": 0.3215937316417694, "learning_rate": 8.93654484525822e-05, "loss": 0.0436, "step": 2510 }, { "grad_norm": 0.2785462737083435, "learning_rate": 8.926328727695226e-05, "loss": 0.0374, "step": 2520 }, { "grad_norm": 0.4290221035480499, "learning_rate": 8.916069672430319e-05, "loss": 0.0335, "step": 2530 }, { "grad_norm": 0.27457764744758606, "learning_rate": 8.905767791654884e-05, "loss": 0.0388, "step": 2540 }, { "grad_norm": 0.375864177942276, "learning_rate": 8.895423198028638e-05, "loss": 0.0427, "step": 2550 }, { "grad_norm": 0.30524152517318726, "learning_rate": 8.885036004678402e-05, "loss": 0.0388, "step": 2560 }, { "grad_norm": 0.25175073742866516, "learning_rate": 8.874606325196857e-05, "loss": 0.037, "step": 2570 }, { "grad_norm": 0.3221883177757263, "learning_rate": 8.864134273641304e-05, "loss": 0.0393, "step": 2580 }, { "grad_norm": 0.22268454730510712, "learning_rate": 8.853619964532427e-05, "loss": 0.0406, "step": 2590 }, { "grad_norm": 0.30943942070007324, "learning_rate": 8.843063512853019e-05, "loss": 0.0342, "step": 2600 }, { "grad_norm": 0.34148910641670227, "learning_rate": 8.832465034046749e-05, "loss": 0.0339, "step": 2610 }, { "grad_norm": 0.368591845035553, "learning_rate": 8.821824644016882e-05, "loss": 0.0365, "step": 2620 }, { "grad_norm": 0.24631820619106293, "learning_rate": 8.811142459125019e-05, "loss": 0.0335, "step": 2630 }, { "grad_norm": 0.2928312420845032, "learning_rate": 8.800418596189822e-05, "loss": 0.0323, "step": 2640 }, { "grad_norm": 0.3286888301372528, "learning_rate": 8.789653172485737e-05, "loss": 0.0405, "step": 2650 }, { "grad_norm": 0.32013002038002014, "learning_rate": 8.778846305741715e-05, "loss": 0.0493, "step": 2660 }, { "grad_norm": 0.31326818466186523, "learning_rate": 8.767998114139918e-05, "loss": 0.0386, "step": 2670 }, { "grad_norm": 0.3746146559715271, "learning_rate": 8.757108716314429e-05, "loss": 0.0456, "step": 2680 }, { "grad_norm": 0.2563594877719879, "learning_rate": 8.746178231349962e-05, "loss": 0.037, "step": 2690 }, { "grad_norm": 0.2953929603099823, "learning_rate": 8.735206778780549e-05, "loss": 0.0407, "step": 2700 }, { "grad_norm": 0.31035730242729187, "learning_rate": 8.724194478588234e-05, "loss": 0.0396, "step": 2710 }, { "grad_norm": 0.23785659670829773, "learning_rate": 8.713141451201772e-05, "loss": 0.0379, "step": 2720 }, { "grad_norm": 0.2431260496377945, "learning_rate": 8.702047817495295e-05, "loss": 0.0361, "step": 2730 }, { "grad_norm": 0.26889485120773315, "learning_rate": 8.69091369878701e-05, "loss": 0.0352, "step": 2740 }, { "grad_norm": 0.25768664479255676, "learning_rate": 8.679739216837849e-05, "loss": 0.0309, "step": 2750 }, { "grad_norm": 0.21640291810035706, "learning_rate": 8.66852449385016e-05, "loss": 0.0305, "step": 2760 }, { "grad_norm": 0.239796444773674, "learning_rate": 8.657269652466356e-05, "loss": 0.0329, "step": 2770 }, { "grad_norm": 0.1949763000011444, "learning_rate": 8.645974815767577e-05, "loss": 0.0339, "step": 2780 }, { "grad_norm": 0.22950531542301178, "learning_rate": 8.634640107272351e-05, "loss": 0.0318, "step": 2790 }, { "grad_norm": 0.4042598605155945, "learning_rate": 8.623265650935234e-05, "loss": 0.0375, "step": 2800 }, { "grad_norm": 0.26357343792915344, "learning_rate": 8.611851571145456e-05, "loss": 0.0381, "step": 2810 }, { "grad_norm": 0.28535282611846924, "learning_rate": 8.600397992725566e-05, "loss": 0.0329, "step": 2820 }, { "grad_norm": 0.26484018564224243, "learning_rate": 8.588905040930061e-05, "loss": 0.0348, "step": 2830 }, { "grad_norm": 0.42581886053085327, "learning_rate": 8.577372841444022e-05, "loss": 0.0352, "step": 2840 }, { "grad_norm": 0.3328506648540497, "learning_rate": 8.565801520381736e-05, "loss": 0.0371, "step": 2850 }, { "grad_norm": 0.2353639304637909, "learning_rate": 8.554191204285313e-05, "loss": 0.0329, "step": 2860 }, { "grad_norm": 0.24978923797607422, "learning_rate": 8.542542020123315e-05, "loss": 0.0329, "step": 2870 }, { "grad_norm": 0.3346234858036041, "learning_rate": 8.530854095289347e-05, "loss": 0.0336, "step": 2880 }, { "grad_norm": 0.17244651913642883, "learning_rate": 8.519127557600688e-05, "loss": 0.0355, "step": 2890 }, { "grad_norm": 0.2736976146697998, "learning_rate": 8.507362535296871e-05, "loss": 0.0362, "step": 2900 }, { "grad_norm": 0.25759074091911316, "learning_rate": 8.495559157038299e-05, "loss": 0.0453, "step": 2910 }, { "grad_norm": 0.20118193328380585, "learning_rate": 8.483717551904823e-05, "loss": 0.0349, "step": 2920 }, { "grad_norm": 0.26836928725242615, "learning_rate": 8.47183784939434e-05, "loss": 0.0314, "step": 2930 }, { "grad_norm": 0.21301132440567017, "learning_rate": 8.459920179421374e-05, "loss": 0.0339, "step": 2940 }, { "grad_norm": 0.20741698145866394, "learning_rate": 8.447964672315656e-05, "loss": 0.031, "step": 2950 }, { "grad_norm": 0.2821630835533142, "learning_rate": 8.435971458820692e-05, "loss": 0.0393, "step": 2960 }, { "grad_norm": 0.26507481932640076, "learning_rate": 8.423940670092345e-05, "loss": 0.0388, "step": 2970 }, { "grad_norm": 0.23084531724452972, "learning_rate": 8.411872437697394e-05, "loss": 0.0331, "step": 2980 }, { "grad_norm": 0.2546499967575073, "learning_rate": 8.399766893612096e-05, "loss": 0.032, "step": 2990 }, { "grad_norm": 0.3551200330257416, "learning_rate": 8.38762417022074e-05, "loss": 0.0404, "step": 3000 }, { "grad_norm": 0.31873995065689087, "learning_rate": 8.375444400314204e-05, "loss": 0.0344, "step": 3010 }, { "grad_norm": 0.40460678935050964, "learning_rate": 8.3632277170885e-05, "loss": 0.0416, "step": 3020 }, { "grad_norm": 0.31619250774383545, "learning_rate": 8.350974254143318e-05, "loss": 0.0334, "step": 3030 }, { "grad_norm": 0.24041129648685455, "learning_rate": 8.338684145480566e-05, "loss": 0.0301, "step": 3040 }, { "grad_norm": 0.2873077988624573, "learning_rate": 8.326357525502904e-05, "loss": 0.0356, "step": 3050 }, { "grad_norm": 0.3374482989311218, "learning_rate": 8.313994529012273e-05, "loss": 0.0361, "step": 3060 }, { "grad_norm": 0.26342105865478516, "learning_rate": 8.301595291208422e-05, "loss": 0.0384, "step": 3070 }, { "grad_norm": 0.24637039005756378, "learning_rate": 8.289159947687427e-05, "loss": 0.0359, "step": 3080 }, { "grad_norm": 0.21619956195354462, "learning_rate": 8.276688634440216e-05, "loss": 0.0354, "step": 3090 }, { "grad_norm": 0.3332933783531189, "learning_rate": 8.26418148785107e-05, "loss": 0.0349, "step": 3100 }, { "grad_norm": 0.421780526638031, "learning_rate": 8.251638644696141e-05, "loss": 0.0417, "step": 3110 }, { "grad_norm": 0.20134341716766357, "learning_rate": 8.23906024214195e-05, "loss": 0.033, "step": 3120 }, { "grad_norm": 0.21278683841228485, "learning_rate": 8.226446417743897e-05, "loss": 0.0373, "step": 3130 }, { "grad_norm": 0.24749384820461273, "learning_rate": 8.213797309444742e-05, "loss": 0.0343, "step": 3140 }, { "grad_norm": 0.22110119462013245, "learning_rate": 8.201113055573105e-05, "loss": 0.0331, "step": 3150 }, { "grad_norm": 0.3886474668979645, "learning_rate": 8.188393794841958e-05, "loss": 0.0331, "step": 3160 }, { "grad_norm": 0.2292635440826416, "learning_rate": 8.175639666347094e-05, "loss": 0.0363, "step": 3170 }, { "grad_norm": 0.25323307514190674, "learning_rate": 8.162850809565623e-05, "loss": 0.0336, "step": 3180 }, { "grad_norm": 0.21227145195007324, "learning_rate": 8.150027364354431e-05, "loss": 0.0323, "step": 3190 }, { "grad_norm": 0.2609926462173462, "learning_rate": 8.137169470948662e-05, "loss": 0.0356, "step": 3200 }, { "grad_norm": 0.2936052978038788, "learning_rate": 8.124277269960179e-05, "loss": 0.0322, "step": 3210 }, { "grad_norm": 0.20430290699005127, "learning_rate": 8.111350902376023e-05, "loss": 0.0276, "step": 3220 }, { "grad_norm": 0.24877238273620605, "learning_rate": 8.098390509556883e-05, "loss": 0.0319, "step": 3230 }, { "grad_norm": 0.2633545994758606, "learning_rate": 8.085396233235536e-05, "loss": 0.0276, "step": 3240 }, { "grad_norm": 0.263155996799469, "learning_rate": 8.072368215515306e-05, "loss": 0.0275, "step": 3250 }, { "grad_norm": 0.2651858627796173, "learning_rate": 8.059306598868506e-05, "loss": 0.0415, "step": 3260 }, { "grad_norm": 0.2887791395187378, "learning_rate": 8.046211526134888e-05, "loss": 0.0312, "step": 3270 }, { "grad_norm": 0.2017965018749237, "learning_rate": 8.033083140520065e-05, "loss": 0.0279, "step": 3280 }, { "grad_norm": 0.18208840489387512, "learning_rate": 8.019921585593962e-05, "loss": 0.0312, "step": 3290 }, { "grad_norm": 0.3216135501861572, "learning_rate": 8.006727005289232e-05, "loss": 0.0276, "step": 3300 }, { "grad_norm": 0.3061944246292114, "learning_rate": 7.993499543899692e-05, "loss": 0.0296, "step": 3310 }, { "grad_norm": 0.2093459665775299, "learning_rate": 7.980239346078742e-05, "loss": 0.0267, "step": 3320 }, { "grad_norm": 0.2369282990694046, "learning_rate": 7.966946556837778e-05, "loss": 0.0307, "step": 3330 }, { "grad_norm": 0.18099355697631836, "learning_rate": 7.953621321544616e-05, "loss": 0.0279, "step": 3340 }, { "grad_norm": 0.3071058988571167, "learning_rate": 7.940263785921896e-05, "loss": 0.0361, "step": 3350 }, { "grad_norm": 0.21619005501270294, "learning_rate": 7.926874096045482e-05, "loss": 0.0285, "step": 3360 }, { "grad_norm": 0.34027624130249023, "learning_rate": 7.913452398342881e-05, "loss": 0.04, "step": 3370 }, { "grad_norm": 0.18769048154354095, "learning_rate": 7.89999883959163e-05, "loss": 0.0335, "step": 3380 }, { "grad_norm": 0.3169165253639221, "learning_rate": 7.886513566917687e-05, "loss": 0.0366, "step": 3390 }, { "grad_norm": 0.27813392877578735, "learning_rate": 7.872996727793838e-05, "loss": 0.0371, "step": 3400 }, { "grad_norm": 0.22508178651332855, "learning_rate": 7.859448470038069e-05, "loss": 0.0311, "step": 3410 }, { "grad_norm": 0.26989710330963135, "learning_rate": 7.845868941811956e-05, "loss": 0.0357, "step": 3420 }, { "grad_norm": 0.2365150898694992, "learning_rate": 7.832258291619043e-05, "loss": 0.03, "step": 3430 }, { "grad_norm": 0.23985950648784637, "learning_rate": 7.81861666830322e-05, "loss": 0.0335, "step": 3440 }, { "grad_norm": 0.19496580958366394, "learning_rate": 7.804944221047097e-05, "loss": 0.0294, "step": 3450 }, { "grad_norm": 0.3390446901321411, "learning_rate": 7.791241099370364e-05, "loss": 0.0379, "step": 3460 }, { "grad_norm": 0.19493456184864044, "learning_rate": 7.777507453128163e-05, "loss": 0.0301, "step": 3470 }, { "grad_norm": 0.2823702096939087, "learning_rate": 7.763743432509451e-05, "loss": 0.0326, "step": 3480 }, { "grad_norm": 0.3537236154079437, "learning_rate": 7.749949188035353e-05, "loss": 0.0321, "step": 3490 }, { "grad_norm": 0.3415128290653229, "learning_rate": 7.736124870557516e-05, "loss": 0.0433, "step": 3500 }, { "grad_norm": 0.2136930227279663, "learning_rate": 7.722270631256459e-05, "loss": 0.0339, "step": 3510 }, { "grad_norm": 0.20249636471271515, "learning_rate": 7.708386621639925e-05, "loss": 0.0292, "step": 3520 }, { "grad_norm": 0.2890123724937439, "learning_rate": 7.694472993541219e-05, "loss": 0.0336, "step": 3530 }, { "grad_norm": 0.26712900400161743, "learning_rate": 7.680529899117547e-05, "loss": 0.0369, "step": 3540 }, { "grad_norm": 0.25592684745788574, "learning_rate": 7.666557490848358e-05, "loss": 0.0318, "step": 3550 }, { "grad_norm": 0.23626044392585754, "learning_rate": 7.65255592153367e-05, "loss": 0.032, "step": 3560 }, { "grad_norm": 0.28349146246910095, "learning_rate": 7.638525344292402e-05, "loss": 0.0364, "step": 3570 }, { "grad_norm": 0.2167208045721054, "learning_rate": 7.624465912560697e-05, "loss": 0.0311, "step": 3580 }, { "grad_norm": 0.3048813045024872, "learning_rate": 7.610377780090249e-05, "loss": 0.0303, "step": 3590 }, { "grad_norm": 0.18003325164318085, "learning_rate": 7.596261100946618e-05, "loss": 0.0294, "step": 3600 }, { "grad_norm": 0.204031839966774, "learning_rate": 7.582116029507542e-05, "loss": 0.0265, "step": 3610 }, { "grad_norm": 0.31757301092147827, "learning_rate": 7.56794272046126e-05, "loss": 0.0303, "step": 3620 }, { "grad_norm": 0.2157532274723053, "learning_rate": 7.55374132880481e-05, "loss": 0.0328, "step": 3630 }, { "grad_norm": 0.23537680506706238, "learning_rate": 7.539512009842333e-05, "loss": 0.0332, "step": 3640 }, { "grad_norm": 0.33641180396080017, "learning_rate": 7.525254919183382e-05, "loss": 0.0344, "step": 3650 }, { "grad_norm": 0.21259333193302155, "learning_rate": 7.510970212741215e-05, "loss": 0.0238, "step": 3660 }, { "grad_norm": 0.21127231419086456, "learning_rate": 7.496658046731096e-05, "loss": 0.0301, "step": 3670 }, { "grad_norm": 0.24336831271648407, "learning_rate": 7.482318577668578e-05, "loss": 0.0313, "step": 3680 }, { "grad_norm": 0.2727046310901642, "learning_rate": 7.467951962367796e-05, "loss": 0.0345, "step": 3690 }, { "grad_norm": 0.2592078745365143, "learning_rate": 7.453558357939755e-05, "loss": 0.0281, "step": 3700 }, { "grad_norm": 0.17212733626365662, "learning_rate": 7.439137921790606e-05, "loss": 0.0348, "step": 3710 }, { "grad_norm": 0.2228129357099533, "learning_rate": 7.42469081161993e-05, "loss": 0.0321, "step": 3720 }, { "grad_norm": 0.298423171043396, "learning_rate": 7.410217185419006e-05, "loss": 0.0259, "step": 3730 }, { "grad_norm": 0.35815808176994324, "learning_rate": 7.395717201469095e-05, "loss": 0.037, "step": 3740 }, { "grad_norm": 0.3189627528190613, "learning_rate": 7.381191018339696e-05, "loss": 0.0322, "step": 3750 }, { "grad_norm": 0.1740846186876297, "learning_rate": 7.36663879488682e-05, "loss": 0.0308, "step": 3760 }, { "grad_norm": 0.2678832411766052, "learning_rate": 7.352060690251254e-05, "loss": 0.0342, "step": 3770 }, { "grad_norm": 0.23539084196090698, "learning_rate": 7.337456863856811e-05, "loss": 0.0282, "step": 3780 }, { "grad_norm": 0.18756552040576935, "learning_rate": 7.3228274754086e-05, "loss": 0.0274, "step": 3790 }, { "grad_norm": 0.3606250286102295, "learning_rate": 7.308172684891267e-05, "loss": 0.0349, "step": 3800 }, { "grad_norm": 0.26585501432418823, "learning_rate": 7.293492652567255e-05, "loss": 0.0249, "step": 3810 }, { "grad_norm": 0.17047196626663208, "learning_rate": 7.278787538975043e-05, "loss": 0.0346, "step": 3820 }, { "grad_norm": 0.28121426701545715, "learning_rate": 7.2640575049274e-05, "loss": 0.0341, "step": 3830 }, { "grad_norm": 0.22555971145629883, "learning_rate": 7.249302711509616e-05, "loss": 0.0291, "step": 3840 }, { "grad_norm": 0.24895784258842468, "learning_rate": 7.23452332007775e-05, "loss": 0.028, "step": 3850 }, { "grad_norm": 0.22032113373279572, "learning_rate": 7.219719492256858e-05, "loss": 0.0303, "step": 3860 }, { "grad_norm": 0.2677270770072937, "learning_rate": 7.20489138993923e-05, "loss": 0.0315, "step": 3870 }, { "grad_norm": 0.20324906706809998, "learning_rate": 7.190039175282614e-05, "loss": 0.0288, "step": 3880 }, { "grad_norm": 0.20269349217414856, "learning_rate": 7.175163010708455e-05, "loss": 0.0253, "step": 3890 }, { "grad_norm": 0.184677392244339, "learning_rate": 7.1602630589001e-05, "loss": 0.0302, "step": 3900 }, { "grad_norm": 0.3061683475971222, "learning_rate": 7.14533948280104e-05, "loss": 0.0317, "step": 3910 }, { "grad_norm": 0.24308982491493225, "learning_rate": 7.130392445613109e-05, "loss": 0.031, "step": 3920 }, { "grad_norm": 0.24938377737998962, "learning_rate": 7.115422110794711e-05, "loss": 0.0315, "step": 3930 }, { "grad_norm": 0.25510725378990173, "learning_rate": 7.100428642059033e-05, "loss": 0.0317, "step": 3940 }, { "grad_norm": 0.3143344819545746, "learning_rate": 7.08541220337224e-05, "loss": 0.0342, "step": 3950 }, { "grad_norm": 0.2725895345211029, "learning_rate": 7.070372958951706e-05, "loss": 0.0287, "step": 3960 }, { "grad_norm": 0.24504853785037994, "learning_rate": 7.055311073264194e-05, "loss": 0.0313, "step": 3970 }, { "grad_norm": 0.3629746735095978, "learning_rate": 7.040226711024077e-05, "loss": 0.0308, "step": 3980 }, { "grad_norm": 0.245266854763031, "learning_rate": 7.02512003719152e-05, "loss": 0.0286, "step": 3990 }, { "grad_norm": 0.17016740143299103, "learning_rate": 7.00999121697069e-05, "loss": 0.033, "step": 4000 }, { "grad_norm": 0.31733542680740356, "learning_rate": 6.99484041580794e-05, "loss": 0.0289, "step": 4010 }, { "grad_norm": 0.2647671401500702, "learning_rate": 6.979667799390004e-05, "loss": 0.0286, "step": 4020 }, { "grad_norm": 0.2501753866672516, "learning_rate": 6.964473533642185e-05, "loss": 0.0344, "step": 4030 }, { "grad_norm": 0.16188398003578186, "learning_rate": 6.949257784726539e-05, "loss": 0.0321, "step": 4040 }, { "grad_norm": 0.17279411852359772, "learning_rate": 6.934020719040056e-05, "loss": 0.0269, "step": 4050 }, { "grad_norm": 0.2892407178878784, "learning_rate": 6.918762503212848e-05, "loss": 0.0345, "step": 4060 }, { "grad_norm": 0.327754408121109, "learning_rate": 6.903483304106319e-05, "loss": 0.0303, "step": 4070 }, { "grad_norm": 0.23895633220672607, "learning_rate": 6.888183288811341e-05, "loss": 0.0263, "step": 4080 }, { "grad_norm": 0.2391633242368698, "learning_rate": 6.87286262464643e-05, "loss": 0.0278, "step": 4090 }, { "grad_norm": 0.2875942587852478, "learning_rate": 6.857521479155915e-05, "loss": 0.0318, "step": 4100 }, { "grad_norm": 0.23478403687477112, "learning_rate": 6.842160020108104e-05, "loss": 0.0303, "step": 4110 }, { "grad_norm": 0.2621610462665558, "learning_rate": 6.826778415493455e-05, "loss": 0.0306, "step": 4120 }, { "grad_norm": 0.2485862374305725, "learning_rate": 6.811376833522729e-05, "loss": 0.0278, "step": 4130 }, { "grad_norm": 0.31626594066619873, "learning_rate": 6.795955442625159e-05, "loss": 0.0293, "step": 4140 }, { "grad_norm": 0.25329089164733887, "learning_rate": 6.780514411446608e-05, "loss": 0.0253, "step": 4150 }, { "grad_norm": 0.23457878828048706, "learning_rate": 6.765053908847716e-05, "loss": 0.0241, "step": 4160 }, { "grad_norm": 0.20784923434257507, "learning_rate": 6.749574103902064e-05, "loss": 0.031, "step": 4170 }, { "grad_norm": 0.2782142162322998, "learning_rate": 6.734075165894317e-05, "loss": 0.0351, "step": 4180 }, { "grad_norm": 0.20590895414352417, "learning_rate": 6.71855726431838e-05, "loss": 0.0304, "step": 4190 }, { "grad_norm": 0.2027372270822525, "learning_rate": 6.703020568875538e-05, "loss": 0.0256, "step": 4200 }, { "grad_norm": 0.21629966795444489, "learning_rate": 6.687465249472603e-05, "loss": 0.0224, "step": 4210 }, { "grad_norm": 0.21328003704547882, "learning_rate": 6.671891476220055e-05, "loss": 0.0246, "step": 4220 }, { "grad_norm": 0.19752737879753113, "learning_rate": 6.656299419430183e-05, "loss": 0.024, "step": 4230 }, { "grad_norm": 0.23028455674648285, "learning_rate": 6.640689249615223e-05, "loss": 0.0279, "step": 4240 }, { "grad_norm": 0.3835226893424988, "learning_rate": 6.625061137485491e-05, "loss": 0.033, "step": 4250 }, { "grad_norm": 0.11433912068605423, "learning_rate": 6.609415253947517e-05, "loss": 0.0316, "step": 4260 }, { "grad_norm": 0.23865659534931183, "learning_rate": 6.593751770102178e-05, "loss": 0.0268, "step": 4270 }, { "grad_norm": 0.27540498971939087, "learning_rate": 6.578070857242823e-05, "loss": 0.0249, "step": 4280 }, { "grad_norm": 0.19155386090278625, "learning_rate": 6.562372686853402e-05, "loss": 0.0258, "step": 4290 }, { "grad_norm": 0.24003063142299652, "learning_rate": 6.546657430606593e-05, "loss": 0.0281, "step": 4300 }, { "grad_norm": 0.2698249816894531, "learning_rate": 6.530925260361918e-05, "loss": 0.0238, "step": 4310 }, { "grad_norm": 0.223853200674057, "learning_rate": 6.515176348163871e-05, "loss": 0.0269, "step": 4320 }, { "grad_norm": 0.2344471663236618, "learning_rate": 6.499410866240032e-05, "loss": 0.0257, "step": 4330 }, { "grad_norm": 0.24914981424808502, "learning_rate": 6.48362898699919e-05, "loss": 0.0244, "step": 4340 }, { "grad_norm": 0.3356197476387024, "learning_rate": 6.467830883029443e-05, "loss": 0.026, "step": 4350 }, { "grad_norm": 0.2091979831457138, "learning_rate": 6.452016727096326e-05, "loss": 0.024, "step": 4360 }, { "grad_norm": 0.23164969682693481, "learning_rate": 6.436186692140916e-05, "loss": 0.0258, "step": 4370 }, { "grad_norm": 0.1934380978345871, "learning_rate": 6.420340951277938e-05, "loss": 0.0236, "step": 4380 }, { "grad_norm": 0.1940460503101349, "learning_rate": 6.404479677793874e-05, "loss": 0.0284, "step": 4390 }, { "grad_norm": 0.3145342171192169, "learning_rate": 6.388603045145075e-05, "loss": 0.0339, "step": 4400 }, { "grad_norm": 0.2623482644557953, "learning_rate": 6.372711226955843e-05, "loss": 0.0283, "step": 4410 }, { "grad_norm": 0.23927292227745056, "learning_rate": 6.356804397016564e-05, "loss": 0.0315, "step": 4420 }, { "grad_norm": 0.1701492965221405, "learning_rate": 6.340882729281779e-05, "loss": 0.0281, "step": 4430 }, { "grad_norm": 0.24823914468288422, "learning_rate": 6.324946397868294e-05, "loss": 0.0329, "step": 4440 }, { "grad_norm": 0.1654198318719864, "learning_rate": 6.308995577053276e-05, "loss": 0.0227, "step": 4450 }, { "grad_norm": 0.19930078089237213, "learning_rate": 6.293030441272347e-05, "loss": 0.0258, "step": 4460 }, { "grad_norm": 0.1882615089416504, "learning_rate": 6.277051165117677e-05, "loss": 0.0257, "step": 4470 }, { "grad_norm": 0.23363041877746582, "learning_rate": 6.261057923336064e-05, "loss": 0.026, "step": 4480 }, { "grad_norm": 0.2629242539405823, "learning_rate": 6.245050890827042e-05, "loss": 0.0261, "step": 4490 }, { "grad_norm": 0.19916780292987823, "learning_rate": 6.229030242640952e-05, "loss": 0.0271, "step": 4500 }, { "grad_norm": 0.2795523703098297, "learning_rate": 6.212996153977037e-05, "loss": 0.0327, "step": 4510 }, { "grad_norm": 0.1577005237340927, "learning_rate": 6.196948800181523e-05, "loss": 0.0241, "step": 4520 }, { "grad_norm": 0.21297843754291534, "learning_rate": 6.180888356745695e-05, "loss": 0.0271, "step": 4530 }, { "grad_norm": 0.22162441909313202, "learning_rate": 6.164814999303995e-05, "loss": 0.0254, "step": 4540 }, { "grad_norm": 0.2922469973564148, "learning_rate": 6.148728903632081e-05, "loss": 0.0225, "step": 4550 }, { "grad_norm": 0.1920434683561325, "learning_rate": 6.132630245644921e-05, "loss": 0.0249, "step": 4560 }, { "grad_norm": 0.2500826120376587, "learning_rate": 6.116519201394857e-05, "loss": 0.0273, "step": 4570 }, { "grad_norm": 0.24933739006519318, "learning_rate": 6.10039594706969e-05, "loss": 0.0255, "step": 4580 }, { "grad_norm": 0.17798496782779694, "learning_rate": 6.084260658990744e-05, "loss": 0.0228, "step": 4590 }, { "grad_norm": 0.2469681054353714, "learning_rate": 6.068113513610943e-05, "loss": 0.0255, "step": 4600 }, { "grad_norm": 0.21480140089988708, "learning_rate": 6.0519546875128876e-05, "loss": 0.0286, "step": 4610 }, { "grad_norm": 0.1382124274969101, "learning_rate": 6.035784357406906e-05, "loss": 0.0282, "step": 4620 }, { "grad_norm": 0.2649359703063965, "learning_rate": 6.01960270012914e-05, "loss": 0.0277, "step": 4630 }, { "grad_norm": 0.18258926272392273, "learning_rate": 6.003409892639599e-05, "loss": 0.0277, "step": 4640 }, { "grad_norm": 0.26712682843208313, "learning_rate": 5.9872061120202336e-05, "loss": 0.0284, "step": 4650 }, { "grad_norm": 0.321340411901474, "learning_rate": 5.9709915354729914e-05, "loss": 0.0356, "step": 4660 }, { "grad_norm": 0.26174652576446533, "learning_rate": 5.9547663403178824e-05, "loss": 0.0272, "step": 4670 }, { "grad_norm": 0.18546070158481598, "learning_rate": 5.9385307039910445e-05, "loss": 0.0185, "step": 4680 }, { "grad_norm": 0.18037548661231995, "learning_rate": 5.922284804042792e-05, "loss": 0.0301, "step": 4690 }, { "grad_norm": 0.259575217962265, "learning_rate": 5.906028818135687e-05, "loss": 0.0317, "step": 4700 }, { "grad_norm": 0.13854414224624634, "learning_rate": 5.889762924042585e-05, "loss": 0.023, "step": 4710 }, { "grad_norm": 0.1512906700372696, "learning_rate": 5.873487299644699e-05, "loss": 0.0273, "step": 4720 }, { "grad_norm": 0.2657434344291687, "learning_rate": 5.857202122929649e-05, "loss": 0.0258, "step": 4730 }, { "grad_norm": 0.23746590316295624, "learning_rate": 5.840907571989518e-05, "loss": 0.0245, "step": 4740 }, { "grad_norm": 0.2601969242095947, "learning_rate": 5.824603825018904e-05, "loss": 0.026, "step": 4750 }, { "grad_norm": 0.1772381216287613, "learning_rate": 5.808291060312975e-05, "loss": 0.0247, "step": 4760 }, { "grad_norm": 0.21261322498321533, "learning_rate": 5.7919694562655083e-05, "loss": 0.0274, "step": 4770 }, { "grad_norm": 0.1394602358341217, "learning_rate": 5.775639191366954e-05, "loss": 0.0259, "step": 4780 }, { "grad_norm": 0.2280806005001068, "learning_rate": 5.75930044420247e-05, "loss": 0.0231, "step": 4790 }, { "grad_norm": 0.22793671488761902, "learning_rate": 5.74295339344998e-05, "loss": 0.0259, "step": 4800 }, { "grad_norm": 0.16661079227924347, "learning_rate": 5.726598217878211e-05, "loss": 0.0221, "step": 4810 }, { "grad_norm": 0.21578985452651978, "learning_rate": 5.71023509634474e-05, "loss": 0.0216, "step": 4820 }, { "grad_norm": 0.21352419257164001, "learning_rate": 5.693864207794049e-05, "loss": 0.0288, "step": 4830 }, { "grad_norm": 0.2209060788154602, "learning_rate": 5.677485731255545e-05, "loss": 0.0336, "step": 4840 }, { "grad_norm": 0.2761784493923187, "learning_rate": 5.6610998458416296e-05, "loss": 0.0244, "step": 4850 }, { "grad_norm": 0.2526509165763855, "learning_rate": 5.644706730745716e-05, "loss": 0.0267, "step": 4860 }, { "grad_norm": 0.17073377966880798, "learning_rate": 5.628306565240287e-05, "loss": 0.0232, "step": 4870 }, { "grad_norm": 0.21943002939224243, "learning_rate": 5.611899528674923e-05, "loss": 0.0285, "step": 4880 }, { "grad_norm": 0.27038800716400146, "learning_rate": 5.595485800474349e-05, "loss": 0.0265, "step": 4890 }, { "grad_norm": 0.19568046927452087, "learning_rate": 5.579065560136467e-05, "loss": 0.0267, "step": 4900 }, { "grad_norm": 0.2279411405324936, "learning_rate": 5.562638987230392e-05, "loss": 0.0263, "step": 4910 }, { "grad_norm": 0.3438696265220642, "learning_rate": 5.546206261394498e-05, "loss": 0.0233, "step": 4920 }, { "grad_norm": 0.25858235359191895, "learning_rate": 5.529767562334437e-05, "loss": 0.0292, "step": 4930 }, { "grad_norm": 0.27226218581199646, "learning_rate": 5.5133230698211926e-05, "loss": 0.0222, "step": 4940 }, { "grad_norm": 0.21062999963760376, "learning_rate": 5.496872963689096e-05, "loss": 0.0251, "step": 4950 }, { "grad_norm": 0.2091674655675888, "learning_rate": 5.4804174238338756e-05, "loss": 0.0223, "step": 4960 }, { "grad_norm": 0.2473215013742447, "learning_rate": 5.463956630210678e-05, "loss": 0.029, "step": 4970 }, { "grad_norm": 0.1932680755853653, "learning_rate": 5.4474907628321046e-05, "loss": 0.031, "step": 4980 }, { "grad_norm": 0.17771746218204498, "learning_rate": 5.431020001766244e-05, "loss": 0.0245, "step": 4990 }, { "grad_norm": 0.2277432233095169, "learning_rate": 5.4145445271346986e-05, "loss": 0.0241, "step": 5000 }, { "grad_norm": 0.14626404643058777, "learning_rate": 5.398064519110622e-05, "loss": 0.0242, "step": 5010 }, { "grad_norm": 0.19513532519340515, "learning_rate": 5.3815801579167394e-05, "loss": 0.0242, "step": 5020 }, { "grad_norm": 0.18844881653785706, "learning_rate": 5.365091623823382e-05, "loss": 0.0257, "step": 5030 }, { "grad_norm": 0.24951714277267456, "learning_rate": 5.348599097146521e-05, "loss": 0.0244, "step": 5040 }, { "grad_norm": 0.24955077469348907, "learning_rate": 5.3321027582457836e-05, "loss": 0.0245, "step": 5050 }, { "grad_norm": 0.24488501250743866, "learning_rate": 5.315602787522491e-05, "loss": 0.0249, "step": 5060 }, { "grad_norm": 0.22254887223243713, "learning_rate": 5.299099365417678e-05, "loss": 0.0184, "step": 5070 }, { "grad_norm": 0.2150186151266098, "learning_rate": 5.2825926724101236e-05, "loss": 0.025, "step": 5080 }, { "grad_norm": 0.2329549938440323, "learning_rate": 5.26608288901438e-05, "loss": 0.0248, "step": 5090 }, { "grad_norm": 0.25964900851249695, "learning_rate": 5.24957019577879e-05, "loss": 0.0281, "step": 5100 }, { "grad_norm": 0.21950477361679077, "learning_rate": 5.2330547732835266e-05, "loss": 0.0226, "step": 5110 }, { "grad_norm": 0.26472753286361694, "learning_rate": 5.2165368021385996e-05, "loss": 0.0324, "step": 5120 }, { "grad_norm": 0.19229494035243988, "learning_rate": 5.200016462981897e-05, "loss": 0.0222, "step": 5130 }, { "grad_norm": 0.17324385046958923, "learning_rate": 5.1834939364772015e-05, "loss": 0.0225, "step": 5140 }, { "grad_norm": 0.2529517114162445, "learning_rate": 5.166969403312214e-05, "loss": 0.0237, "step": 5150 }, { "grad_norm": 0.2019347995519638, "learning_rate": 5.1504430441965844e-05, "loss": 0.0246, "step": 5160 }, { "grad_norm": 0.15839067101478577, "learning_rate": 5.133915039859923e-05, "loss": 0.0184, "step": 5170 }, { "grad_norm": 0.15650928020477295, "learning_rate": 5.1173855710498444e-05, "loss": 0.0222, "step": 5180 }, { "grad_norm": 0.27009451389312744, "learning_rate": 5.100854818529967e-05, "loss": 0.0252, "step": 5190 }, { "grad_norm": 0.26238104701042175, "learning_rate": 5.084322963077951e-05, "loss": 0.0286, "step": 5200 }, { "grad_norm": 0.27602532505989075, "learning_rate": 5.067790185483522e-05, "loss": 0.0252, "step": 5210 }, { "grad_norm": 0.1367495208978653, "learning_rate": 5.0512566665464844e-05, "loss": 0.0243, "step": 5220 }, { "grad_norm": 0.21731014549732208, "learning_rate": 5.034722587074755e-05, "loss": 0.0261, "step": 5230 }, { "grad_norm": 0.1448562741279602, "learning_rate": 5.018188127882375e-05, "loss": 0.0203, "step": 5240 }, { "grad_norm": 0.21827775239944458, "learning_rate": 5.0016534697875417e-05, "loss": 0.0193, "step": 5250 }, { "grad_norm": 0.26346448063850403, "learning_rate": 4.9851187936106294e-05, "loss": 0.0272, "step": 5260 }, { "grad_norm": 0.1559494584798813, "learning_rate": 4.968584280172206e-05, "loss": 0.0251, "step": 5270 }, { "grad_norm": 0.18296435475349426, "learning_rate": 4.95205011029106e-05, "loss": 0.025, "step": 5280 }, { "grad_norm": 0.18214404582977295, "learning_rate": 4.935516464782227e-05, "loss": 0.0261, "step": 5290 }, { "grad_norm": 0.22872810065746307, "learning_rate": 4.918983524455003e-05, "loss": 0.0244, "step": 5300 }, { "grad_norm": 0.24100956320762634, "learning_rate": 4.9024514701109766e-05, "loss": 0.0232, "step": 5310 }, { "grad_norm": 0.3035481870174408, "learning_rate": 4.885920482542043e-05, "loss": 0.0242, "step": 5320 }, { "grad_norm": 0.15349286794662476, "learning_rate": 4.869390742528438e-05, "loss": 0.0198, "step": 5330 }, { "grad_norm": 0.23986560106277466, "learning_rate": 4.852862430836744e-05, "loss": 0.0262, "step": 5340 }, { "grad_norm": 0.24374152719974518, "learning_rate": 4.836335728217933e-05, "loss": 0.0223, "step": 5350 }, { "grad_norm": 0.2568149268627167, "learning_rate": 4.819810815405379e-05, "loss": 0.0242, "step": 5360 }, { "grad_norm": 0.2882293462753296, "learning_rate": 4.803287873112877e-05, "loss": 0.0267, "step": 5370 }, { "grad_norm": 0.22537153959274292, "learning_rate": 4.786767082032681e-05, "loss": 0.021, "step": 5380 }, { "grad_norm": 0.1875973343849182, "learning_rate": 4.77024862283351e-05, "loss": 0.0214, "step": 5390 }, { "grad_norm": 0.12758386135101318, "learning_rate": 4.753732676158593e-05, "loss": 0.0211, "step": 5400 }, { "grad_norm": 0.15391485393047333, "learning_rate": 4.737219422623672e-05, "loss": 0.0227, "step": 5410 }, { "grad_norm": 0.2379147857427597, "learning_rate": 4.720709042815044e-05, "loss": 0.0203, "step": 5420 }, { "grad_norm": 0.27782538533210754, "learning_rate": 4.704201717287578e-05, "loss": 0.0235, "step": 5430 }, { "grad_norm": 0.24118074774742126, "learning_rate": 4.6876976265627404e-05, "loss": 0.0236, "step": 5440 }, { "grad_norm": 0.26524657011032104, "learning_rate": 4.671196951126626e-05, "loss": 0.0279, "step": 5450 }, { "grad_norm": 0.19693705439567566, "learning_rate": 4.654699871427971e-05, "loss": 0.0209, "step": 5460 }, { "grad_norm": 0.1477838009595871, "learning_rate": 4.6382065678762034e-05, "loss": 0.0216, "step": 5470 }, { "grad_norm": 0.16909027099609375, "learning_rate": 4.6217172208394424e-05, "loss": 0.0204, "step": 5480 }, { "grad_norm": 0.22319193184375763, "learning_rate": 4.605232010642549e-05, "loss": 0.019, "step": 5490 }, { "grad_norm": 0.19781042635440826, "learning_rate": 4.588751117565142e-05, "loss": 0.0193, "step": 5500 }, { "grad_norm": 0.18684546649456024, "learning_rate": 4.5722747218396214e-05, "loss": 0.0244, "step": 5510 }, { "grad_norm": 0.1417679339647293, "learning_rate": 4.5558030036492194e-05, "loss": 0.0177, "step": 5520 }, { "grad_norm": 0.25321561098098755, "learning_rate": 4.539336143125999e-05, "loss": 0.0274, "step": 5530 }, { "grad_norm": 0.18102231621742249, "learning_rate": 4.522874320348916e-05, "loss": 0.0196, "step": 5540 }, { "grad_norm": 0.19369767606258392, "learning_rate": 4.506417715341821e-05, "loss": 0.0183, "step": 5550 }, { "grad_norm": 0.15654641389846802, "learning_rate": 4.489966508071511e-05, "loss": 0.0201, "step": 5560 }, { "grad_norm": 0.18179334700107574, "learning_rate": 4.4735208784457575e-05, "loss": 0.0181, "step": 5570 }, { "grad_norm": 0.20368140935897827, "learning_rate": 4.457081006311325e-05, "loss": 0.0249, "step": 5580 }, { "grad_norm": 0.15744532644748688, "learning_rate": 4.440647071452027e-05, "loss": 0.0223, "step": 5590 }, { "grad_norm": 0.27984732389450073, "learning_rate": 4.424219253586737e-05, "loss": 0.0246, "step": 5600 }, { "grad_norm": 0.22008205950260162, "learning_rate": 4.407797732367443e-05, "loss": 0.0267, "step": 5610 }, { "grad_norm": 0.25910887122154236, "learning_rate": 4.391382687377268e-05, "loss": 0.0255, "step": 5620 }, { "grad_norm": 0.20075449347496033, "learning_rate": 4.374974298128512e-05, "loss": 0.0202, "step": 5630 }, { "grad_norm": 0.20055940747261047, "learning_rate": 4.358572744060699e-05, "loss": 0.0264, "step": 5640 }, { "grad_norm": 0.22007609903812408, "learning_rate": 4.342178204538588e-05, "loss": 0.0174, "step": 5650 }, { "grad_norm": 0.14645157754421234, "learning_rate": 4.325790858850241e-05, "loss": 0.0222, "step": 5660 }, { "grad_norm": 0.2203606516122818, "learning_rate": 4.309410886205043e-05, "loss": 0.0242, "step": 5670 }, { "grad_norm": 0.20661097764968872, "learning_rate": 4.293038465731752e-05, "loss": 0.0205, "step": 5680 }, { "grad_norm": 0.23800870776176453, "learning_rate": 4.276673776476533e-05, "loss": 0.0236, "step": 5690 }, { "grad_norm": 0.1541551649570465, "learning_rate": 4.260316997401007e-05, "loss": 0.0161, "step": 5700 }, { "grad_norm": 0.20120784640312195, "learning_rate": 4.243968307380293e-05, "loss": 0.0235, "step": 5710 }, { "grad_norm": 0.28917548060417175, "learning_rate": 4.22762788520104e-05, "loss": 0.0267, "step": 5720 }, { "grad_norm": 0.1613108068704605, "learning_rate": 4.211295909559491e-05, "loss": 0.0289, "step": 5730 }, { "grad_norm": 0.17306523025035858, "learning_rate": 4.194972559059511e-05, "loss": 0.0203, "step": 5740 }, { "grad_norm": 0.29139244556427, "learning_rate": 4.178658012210651e-05, "loss": 0.0212, "step": 5750 }, { "grad_norm": 0.16063804924488068, "learning_rate": 4.162352447426177e-05, "loss": 0.0233, "step": 5760 }, { "grad_norm": 0.2160862386226654, "learning_rate": 4.146056043021135e-05, "loss": 0.0217, "step": 5770 }, { "grad_norm": 0.16396670043468475, "learning_rate": 4.1297689772103944e-05, "loss": 0.0237, "step": 5780 }, { "grad_norm": 0.18755421042442322, "learning_rate": 4.113491428106694e-05, "loss": 0.0195, "step": 5790 }, { "grad_norm": 0.15634946525096893, "learning_rate": 4.0972235737187055e-05, "loss": 0.0215, "step": 5800 }, { "grad_norm": 0.13430272042751312, "learning_rate": 4.080965591949076e-05, "loss": 0.018, "step": 5810 }, { "grad_norm": 0.20626617968082428, "learning_rate": 4.0647176605924924e-05, "loss": 0.0198, "step": 5820 }, { "grad_norm": 0.37165969610214233, "learning_rate": 4.0484799573337255e-05, "loss": 0.0247, "step": 5830 }, { "grad_norm": 0.17410384118556976, "learning_rate": 4.032252659745699e-05, "loss": 0.0207, "step": 5840 }, { "grad_norm": 0.2360416054725647, "learning_rate": 4.016035945287539e-05, "loss": 0.0268, "step": 5850 }, { "grad_norm": 0.21655277907848358, "learning_rate": 3.999829991302635e-05, "loss": 0.0204, "step": 5860 }, { "grad_norm": 0.2546917498111725, "learning_rate": 3.983634975016707e-05, "loss": 0.0204, "step": 5870 }, { "grad_norm": 0.20379948616027832, "learning_rate": 3.967451073535854e-05, "loss": 0.0287, "step": 5880 }, { "grad_norm": 0.17221467196941376, "learning_rate": 3.951278463844633e-05, "loss": 0.0201, "step": 5890 }, { "grad_norm": 0.15472696721553802, "learning_rate": 3.935117322804111e-05, "loss": 0.0231, "step": 5900 }, { "grad_norm": 0.18871457874774933, "learning_rate": 3.918967827149938e-05, "loss": 0.0208, "step": 5910 }, { "grad_norm": 0.19256338477134705, "learning_rate": 3.9028301534904094e-05, "loss": 0.0213, "step": 5920 }, { "grad_norm": 0.24404627084732056, "learning_rate": 3.88670447830454e-05, "loss": 0.0195, "step": 5930 }, { "grad_norm": 0.18381774425506592, "learning_rate": 3.870590977940132e-05, "loss": 0.0218, "step": 5940 }, { "grad_norm": 0.1988138109445572, "learning_rate": 3.8544898286118404e-05, "loss": 0.019, "step": 5950 }, { "grad_norm": 0.23373179137706757, "learning_rate": 3.838401206399257e-05, "loss": 0.0175, "step": 5960 }, { "grad_norm": 0.20487549901008606, "learning_rate": 3.822325287244975e-05, "loss": 0.0187, "step": 5970 }, { "grad_norm": 0.20276184380054474, "learning_rate": 3.8062622469526725e-05, "loss": 0.0207, "step": 5980 }, { "grad_norm": 0.1880190074443817, "learning_rate": 3.790212261185183e-05, "loss": 0.0193, "step": 5990 }, { "grad_norm": 0.21077567338943481, "learning_rate": 3.7741755054625794e-05, "loss": 0.0195, "step": 6000 }, { "grad_norm": 0.2877955734729767, "learning_rate": 3.758152155160255e-05, "loss": 0.0178, "step": 6010 }, { "grad_norm": 0.20962396264076233, "learning_rate": 3.742142385506999e-05, "loss": 0.0192, "step": 6020 }, { "grad_norm": 0.14927442371845245, "learning_rate": 3.72614637158309e-05, "loss": 0.0183, "step": 6030 }, { "grad_norm": 0.22256891429424286, "learning_rate": 3.710164288318371e-05, "loss": 0.0233, "step": 6040 }, { "grad_norm": 0.1716308444738388, "learning_rate": 3.694196310490345e-05, "loss": 0.0158, "step": 6050 }, { "grad_norm": 0.15721526741981506, "learning_rate": 3.678242612722259e-05, "loss": 0.0197, "step": 6060 }, { "grad_norm": 0.19593234360218048, "learning_rate": 3.6623033694811953e-05, "loss": 0.0171, "step": 6070 }, { "grad_norm": 0.16975939273834229, "learning_rate": 3.6463787550761665e-05, "loss": 0.0219, "step": 6080 }, { "grad_norm": 0.18472842872142792, "learning_rate": 3.630468943656202e-05, "loss": 0.0178, "step": 6090 }, { "grad_norm": 0.2210434228181839, "learning_rate": 3.6145741092084523e-05, "loss": 0.016, "step": 6100 }, { "grad_norm": 0.13858677446842194, "learning_rate": 3.598694425556278e-05, "loss": 0.0156, "step": 6110 }, { "grad_norm": 0.20435364544391632, "learning_rate": 3.58283006635736e-05, "loss": 0.0212, "step": 6120 }, { "grad_norm": 0.19628238677978516, "learning_rate": 3.566981205101781e-05, "loss": 0.019, "step": 6130 }, { "grad_norm": 0.1378137320280075, "learning_rate": 3.5511480151101556e-05, "loss": 0.0154, "step": 6140 }, { "grad_norm": 0.208942711353302, "learning_rate": 3.5353306695317104e-05, "loss": 0.016, "step": 6150 }, { "grad_norm": 0.2685665786266327, "learning_rate": 3.519529341342402e-05, "loss": 0.0212, "step": 6160 }, { "grad_norm": 0.20916691422462463, "learning_rate": 3.503744203343026e-05, "loss": 0.021, "step": 6170 }, { "grad_norm": 0.1822635382413864, "learning_rate": 3.487975428157318e-05, "loss": 0.0207, "step": 6180 }, { "grad_norm": 0.22418911755084991, "learning_rate": 3.472223188230083e-05, "loss": 0.0231, "step": 6190 }, { "grad_norm": 0.2651776373386383, "learning_rate": 3.4564876558252866e-05, "loss": 0.0184, "step": 6200 }, { "grad_norm": 0.18448644876480103, "learning_rate": 3.440769003024195e-05, "loss": 0.0182, "step": 6210 }, { "grad_norm": 0.14920015633106232, "learning_rate": 3.425067401723477e-05, "loss": 0.0178, "step": 6220 }, { "grad_norm": 0.140127494931221, "learning_rate": 3.409383023633325e-05, "loss": 0.0183, "step": 6230 }, { "grad_norm": 0.17689046263694763, "learning_rate": 3.3937160402755894e-05, "loss": 0.0185, "step": 6240 }, { "grad_norm": 0.158067524433136, "learning_rate": 3.378066622981885e-05, "loss": 0.0201, "step": 6250 }, { "grad_norm": 0.25078022480010986, "learning_rate": 3.362434942891738e-05, "loss": 0.0231, "step": 6260 }, { "grad_norm": 0.1817896068096161, "learning_rate": 3.346821170950693e-05, "loss": 0.019, "step": 6270 }, { "grad_norm": 0.1831316351890564, "learning_rate": 3.3312254779084585e-05, "loss": 0.0216, "step": 6280 }, { "grad_norm": 0.19478438794612885, "learning_rate": 3.315648034317039e-05, "loss": 0.0218, "step": 6290 }, { "grad_norm": 0.1527704894542694, "learning_rate": 3.3000890105288564e-05, "loss": 0.0163, "step": 6300 }, { "grad_norm": 0.16880416870117188, "learning_rate": 3.284548576694908e-05, "loss": 0.019, "step": 6310 }, { "grad_norm": 0.13074038922786713, "learning_rate": 3.2690269027628815e-05, "loss": 0.0178, "step": 6320 }, { "grad_norm": 0.13471795618534088, "learning_rate": 3.253524158475324e-05, "loss": 0.018, "step": 6330 }, { "grad_norm": 0.184020534157753, "learning_rate": 3.238040513367757e-05, "loss": 0.0191, "step": 6340 }, { "grad_norm": 0.1522371768951416, "learning_rate": 3.222576136766843e-05, "loss": 0.0145, "step": 6350 }, { "grad_norm": 0.11753250658512115, "learning_rate": 3.2071311977885324e-05, "loss": 0.0181, "step": 6360 }, { "grad_norm": 0.19269528985023499, "learning_rate": 3.191705865336197e-05, "loss": 0.0219, "step": 6370 }, { "grad_norm": 0.17957369983196259, "learning_rate": 3.1763003080988075e-05, "loss": 0.0176, "step": 6380 }, { "grad_norm": 0.19532904028892517, "learning_rate": 3.160914694549063e-05, "loss": 0.0197, "step": 6390 }, { "grad_norm": 0.17244452238082886, "learning_rate": 3.145549192941573e-05, "loss": 0.0179, "step": 6400 }, { "grad_norm": 0.2408980429172516, "learning_rate": 3.130203971310999e-05, "loss": 0.02, "step": 6410 }, { "grad_norm": 0.219632089138031, "learning_rate": 3.114879197470225e-05, "loss": 0.02, "step": 6420 }, { "grad_norm": 0.15346218645572662, "learning_rate": 3.0995750390085285e-05, "loss": 0.0151, "step": 6430 }, { "grad_norm": 0.1592087745666504, "learning_rate": 3.084291663289728e-05, "loss": 0.0191, "step": 6440 }, { "grad_norm": 0.16185326874256134, "learning_rate": 3.069029237450375e-05, "loss": 0.0146, "step": 6450 }, { "grad_norm": 0.16734297573566437, "learning_rate": 3.053787928397911e-05, "loss": 0.0156, "step": 6460 }, { "grad_norm": 0.21206697821617126, "learning_rate": 3.0385679028088526e-05, "loss": 0.0204, "step": 6470 }, { "grad_norm": 0.2032633125782013, "learning_rate": 3.023369327126959e-05, "loss": 0.017, "step": 6480 }, { "grad_norm": 0.15359030663967133, "learning_rate": 3.0081923675614198e-05, "loss": 0.0164, "step": 6490 }, { "grad_norm": 0.20097380876541138, "learning_rate": 2.993037190085034e-05, "loss": 0.0156, "step": 6500 }, { "grad_norm": 0.18181000649929047, "learning_rate": 2.977903960432392e-05, "loss": 0.0174, "step": 6510 }, { "grad_norm": 0.1900976002216339, "learning_rate": 2.9627928440980722e-05, "loss": 0.0235, "step": 6520 }, { "grad_norm": 0.17095741629600525, "learning_rate": 2.9477040063348183e-05, "loss": 0.0197, "step": 6530 }, { "grad_norm": 0.25206059217453003, "learning_rate": 2.9326376121517456e-05, "loss": 0.0187, "step": 6540 }, { "grad_norm": 0.1710679829120636, "learning_rate": 2.9175938263125236e-05, "loss": 0.0182, "step": 6550 }, { "grad_norm": 0.10189634561538696, "learning_rate": 2.9025728133335873e-05, "loss": 0.0178, "step": 6560 }, { "grad_norm": 0.19464664161205292, "learning_rate": 2.8875747374823288e-05, "loss": 0.0223, "step": 6570 }, { "grad_norm": 0.22288554906845093, "learning_rate": 2.872599762775298e-05, "loss": 0.0256, "step": 6580 }, { "grad_norm": 0.22217154502868652, "learning_rate": 2.857648052976425e-05, "loss": 0.0177, "step": 6590 }, { "grad_norm": 0.1568957269191742, "learning_rate": 2.8427197715952047e-05, "loss": 0.0174, "step": 6600 }, { "grad_norm": 0.16788138449192047, "learning_rate": 2.8278150818849393e-05, "loss": 0.0179, "step": 6610 }, { "grad_norm": 0.18023543059825897, "learning_rate": 2.812934146840922e-05, "loss": 0.0172, "step": 6620 }, { "grad_norm": 0.23732759058475494, "learning_rate": 2.7980771291986764e-05, "loss": 0.0217, "step": 6630 }, { "grad_norm": 0.2055542767047882, "learning_rate": 2.783244191432167e-05, "loss": 0.0191, "step": 6640 }, { "grad_norm": 0.16190384328365326, "learning_rate": 2.768435495752022e-05, "loss": 0.0159, "step": 6650 }, { "grad_norm": 0.20430581271648407, "learning_rate": 2.753651204103771e-05, "loss": 0.0172, "step": 6660 }, { "grad_norm": 0.14682504534721375, "learning_rate": 2.7388914781660523e-05, "loss": 0.0162, "step": 6670 }, { "grad_norm": 0.15921834111213684, "learning_rate": 2.7241564793488693e-05, "loss": 0.0187, "step": 6680 }, { "grad_norm": 0.22177007794380188, "learning_rate": 2.7094463687918037e-05, "loss": 0.0168, "step": 6690 }, { "grad_norm": 0.24490861594676971, "learning_rate": 2.694761307362268e-05, "loss": 0.0156, "step": 6700 }, { "grad_norm": 0.19955235719680786, "learning_rate": 2.6801014556537467e-05, "loss": 0.0161, "step": 6710 }, { "grad_norm": 0.24833202362060547, "learning_rate": 2.6654669739840243e-05, "loss": 0.0174, "step": 6720 }, { "grad_norm": 0.23644858598709106, "learning_rate": 2.650858022393451e-05, "loss": 0.0195, "step": 6730 }, { "grad_norm": 0.1827281415462494, "learning_rate": 2.6362747606431747e-05, "loss": 0.02, "step": 6740 }, { "grad_norm": 0.14655549824237823, "learning_rate": 2.6217173482134172e-05, "loss": 0.0151, "step": 6750 }, { "grad_norm": 0.1415199637413025, "learning_rate": 2.6071859443017044e-05, "loss": 0.0163, "step": 6760 }, { "grad_norm": 0.23147845268249512, "learning_rate": 2.5926807078211414e-05, "loss": 0.014, "step": 6770 }, { "grad_norm": 0.18513616919517517, "learning_rate": 2.5782017973986728e-05, "loss": 0.0155, "step": 6780 }, { "grad_norm": 0.13637782633304596, "learning_rate": 2.5637493713733374e-05, "loss": 0.0155, "step": 6790 }, { "grad_norm": 0.1584867686033249, "learning_rate": 2.549323587794559e-05, "loss": 0.0229, "step": 6800 }, { "grad_norm": 0.13858868181705475, "learning_rate": 2.5349246044203895e-05, "loss": 0.0131, "step": 6810 }, { "grad_norm": 0.17394240200519562, "learning_rate": 2.520552578715808e-05, "loss": 0.0162, "step": 6820 }, { "grad_norm": 0.1673695296049118, "learning_rate": 2.506207667850981e-05, "loss": 0.0131, "step": 6830 }, { "grad_norm": 0.2173657864332199, "learning_rate": 2.4918900286995555e-05, "loss": 0.0157, "step": 6840 }, { "grad_norm": 0.13986285030841827, "learning_rate": 2.4775998178369458e-05, "loss": 0.0123, "step": 6850 }, { "grad_norm": 0.18820303678512573, "learning_rate": 2.4633371915386017e-05, "loss": 0.019, "step": 6860 }, { "grad_norm": 0.15979260206222534, "learning_rate": 2.4491023057783235e-05, "loss": 0.0176, "step": 6870 }, { "grad_norm": 0.29124119877815247, "learning_rate": 2.4348953162265375e-05, "loss": 0.0235, "step": 6880 }, { "grad_norm": 0.14994558691978455, "learning_rate": 2.420716378248607e-05, "loss": 0.0173, "step": 6890 }, { "grad_norm": 0.19411176443099976, "learning_rate": 2.4065656469031266e-05, "loss": 0.0159, "step": 6900 }, { "grad_norm": 0.18173612654209137, "learning_rate": 2.3924432769402268e-05, "loss": 0.0185, "step": 6910 }, { "grad_norm": 0.2811588943004608, "learning_rate": 2.3783494227998844e-05, "loss": 0.023, "step": 6920 }, { "grad_norm": 0.2056000530719757, "learning_rate": 2.3642842386102264e-05, "loss": 0.0132, "step": 6930 }, { "grad_norm": 0.1833900511264801, "learning_rate": 2.3502478781858567e-05, "loss": 0.0135, "step": 6940 }, { "grad_norm": 0.22436781227588654, "learning_rate": 2.3362404950261628e-05, "loss": 0.0188, "step": 6950 }, { "grad_norm": 0.22013536095619202, "learning_rate": 2.3222622423136458e-05, "loss": 0.0173, "step": 6960 }, { "grad_norm": 0.13023504614830017, "learning_rate": 2.3083132729122332e-05, "loss": 0.0159, "step": 6970 }, { "grad_norm": 0.1646192967891693, "learning_rate": 2.294393739365621e-05, "loss": 0.0152, "step": 6980 }, { "grad_norm": 0.16340629756450653, "learning_rate": 2.2805037938956e-05, "loss": 0.0198, "step": 6990 }, { "grad_norm": 0.24428610503673553, "learning_rate": 2.266643588400386e-05, "loss": 0.0141, "step": 7000 }, { "grad_norm": 0.15410342812538147, "learning_rate": 2.252813274452969e-05, "loss": 0.0167, "step": 7010 }, { "grad_norm": 0.1458609402179718, "learning_rate": 2.2390130032994427e-05, "loss": 0.013, "step": 7020 }, { "grad_norm": 0.11972752958536148, "learning_rate": 2.2252429258573633e-05, "loss": 0.0155, "step": 7030 }, { "grad_norm": 0.09356407821178436, "learning_rate": 2.2115031927140904e-05, "loss": 0.0196, "step": 7040 }, { "grad_norm": 0.1808558851480484, "learning_rate": 2.1977939541251463e-05, "loss": 0.0218, "step": 7050 }, { "grad_norm": 0.14423266053199768, "learning_rate": 2.1841153600125684e-05, "loss": 0.0145, "step": 7060 }, { "grad_norm": 0.19404086470603943, "learning_rate": 2.170467559963267e-05, "loss": 0.0173, "step": 7070 }, { "grad_norm": 0.19107075035572052, "learning_rate": 2.1568507032273982e-05, "loss": 0.0134, "step": 7080 }, { "grad_norm": 0.11125738173723221, "learning_rate": 2.1432649387167264e-05, "loss": 0.0148, "step": 7090 }, { "grad_norm": 0.11647651344537735, "learning_rate": 2.1297104150029973e-05, "loss": 0.0163, "step": 7100 }, { "grad_norm": 0.15378396213054657, "learning_rate": 2.116187280316307e-05, "loss": 0.0162, "step": 7110 }, { "grad_norm": 0.17042005062103271, "learning_rate": 2.1026956825434908e-05, "loss": 0.0171, "step": 7120 }, { "grad_norm": 0.16227971017360687, "learning_rate": 2.0892357692265017e-05, "loss": 0.0135, "step": 7130 }, { "grad_norm": 0.1611928641796112, "learning_rate": 2.0758076875607947e-05, "loss": 0.0169, "step": 7140 }, { "grad_norm": 0.14948996901512146, "learning_rate": 2.0624115843937207e-05, "loss": 0.0176, "step": 7150 }, { "grad_norm": 0.13957589864730835, "learning_rate": 2.0490476062229157e-05, "loss": 0.0217, "step": 7160 }, { "grad_norm": 0.15609119832515717, "learning_rate": 2.035715899194704e-05, "loss": 0.0146, "step": 7170 }, { "grad_norm": 0.17707312107086182, "learning_rate": 2.022416609102499e-05, "loss": 0.0148, "step": 7180 }, { "grad_norm": 0.15267689526081085, "learning_rate": 2.009149881385205e-05, "loss": 0.0207, "step": 7190 }, { "grad_norm": 0.1135406345129013, "learning_rate": 1.995915861125634e-05, "loss": 0.0135, "step": 7200 }, { "grad_norm": 0.12519434094429016, "learning_rate": 1.9827146930489065e-05, "loss": 0.0118, "step": 7210 }, { "grad_norm": 0.24776341021060944, "learning_rate": 1.9695465215208848e-05, "loss": 0.0195, "step": 7220 }, { "grad_norm": 0.19259370863437653, "learning_rate": 1.9564114905465813e-05, "loss": 0.0134, "step": 7230 }, { "grad_norm": 0.08698273450136185, "learning_rate": 1.9433097437685936e-05, "loss": 0.0158, "step": 7240 }, { "grad_norm": 0.20530691742897034, "learning_rate": 1.930241424465521e-05, "loss": 0.0135, "step": 7250 }, { "grad_norm": 0.21093280613422394, "learning_rate": 1.9172066755504115e-05, "loss": 0.0151, "step": 7260 }, { "grad_norm": 0.16378900408744812, "learning_rate": 1.9042056395691914e-05, "loss": 0.0164, "step": 7270 }, { "grad_norm": 0.1526462435722351, "learning_rate": 1.8912384586991066e-05, "loss": 0.0139, "step": 7280 }, { "grad_norm": 0.1873576045036316, "learning_rate": 1.8783052747471717e-05, "loss": 0.0155, "step": 7290 }, { "grad_norm": 0.18176430463790894, "learning_rate": 1.865406229148611e-05, "loss": 0.0163, "step": 7300 }, { "grad_norm": 0.10951041430234909, "learning_rate": 1.8525414629653233e-05, "loss": 0.0175, "step": 7310 }, { "grad_norm": 0.17176972329616547, "learning_rate": 1.8397111168843255e-05, "loss": 0.0153, "step": 7320 }, { "grad_norm": 0.17705020308494568, "learning_rate": 1.8269153312162323e-05, "loss": 0.0182, "step": 7330 }, { "grad_norm": 0.1507093906402588, "learning_rate": 1.8141542458937054e-05, "loss": 0.0157, "step": 7340 }, { "grad_norm": 0.11911935359239578, "learning_rate": 1.8014280004699268e-05, "loss": 0.0133, "step": 7350 }, { "grad_norm": 0.2908618748188019, "learning_rate": 1.788736734117078e-05, "loss": 0.0154, "step": 7360 }, { "grad_norm": 0.13536831736564636, "learning_rate": 1.7760805856248152e-05, "loss": 0.0148, "step": 7370 }, { "grad_norm": 0.13650037348270416, "learning_rate": 1.7634596933987518e-05, "loss": 0.0157, "step": 7380 }, { "grad_norm": 0.11142794042825699, "learning_rate": 1.7508741954589404e-05, "loss": 0.0171, "step": 7390 }, { "grad_norm": 0.09929151833057404, "learning_rate": 1.7383242294383717e-05, "loss": 0.0155, "step": 7400 }, { "grad_norm": 0.147218719124794, "learning_rate": 1.7258099325814632e-05, "loss": 0.0121, "step": 7410 }, { "grad_norm": 0.11491416394710541, "learning_rate": 1.7133314417425594e-05, "loss": 0.0166, "step": 7420 }, { "grad_norm": 0.15126653015613556, "learning_rate": 1.7008888933844408e-05, "loss": 0.0141, "step": 7430 }, { "grad_norm": 0.15106645226478577, "learning_rate": 1.6884824235768172e-05, "loss": 0.0163, "step": 7440 }, { "grad_norm": 0.13374722003936768, "learning_rate": 1.6761121679948592e-05, "loss": 0.0125, "step": 7450 }, { "grad_norm": 0.1123928427696228, "learning_rate": 1.663778261917695e-05, "loss": 0.0178, "step": 7460 }, { "grad_norm": 0.21889421343803406, "learning_rate": 1.651480840226952e-05, "loss": 0.015, "step": 7470 }, { "grad_norm": 0.1740415096282959, "learning_rate": 1.639220037405258e-05, "loss": 0.0165, "step": 7480 }, { "grad_norm": 0.19887235760688782, "learning_rate": 1.6269959875347906e-05, "loss": 0.0156, "step": 7490 }, { "grad_norm": 0.1645178198814392, "learning_rate": 1.614808824295802e-05, "loss": 0.0147, "step": 7500 }, { "grad_norm": 0.20531408488750458, "learning_rate": 1.602658680965152e-05, "loss": 0.0124, "step": 7510 }, { "grad_norm": 0.13613347709178925, "learning_rate": 1.5905456904148686e-05, "loss": 0.0183, "step": 7520 }, { "grad_norm": 0.15945537388324738, "learning_rate": 1.57846998511067e-05, "loss": 0.0178, "step": 7530 }, { "grad_norm": 0.10745138674974442, "learning_rate": 1.566431697110538e-05, "loss": 0.0128, "step": 7540 }, { "grad_norm": 0.19295713305473328, "learning_rate": 1.554430958063259e-05, "loss": 0.0128, "step": 7550 }, { "grad_norm": 0.17681176960468292, "learning_rate": 1.5424678992069912e-05, "loss": 0.0129, "step": 7560 }, { "grad_norm": 0.24248047173023224, "learning_rate": 1.5305426513678362e-05, "loss": 0.0153, "step": 7570 }, { "grad_norm": 0.1605902463197708, "learning_rate": 1.518655344958388e-05, "loss": 0.0163, "step": 7580 }, { "grad_norm": 0.16095761954784393, "learning_rate": 1.5068061099763275e-05, "loss": 0.015, "step": 7590 }, { "grad_norm": 0.14404016733169556, "learning_rate": 1.494995076002988e-05, "loss": 0.016, "step": 7600 }, { "grad_norm": 0.17697139084339142, "learning_rate": 1.4832223722019456e-05, "loss": 0.0122, "step": 7610 }, { "grad_norm": 0.16327296197414398, "learning_rate": 1.4714881273176035e-05, "loss": 0.0142, "step": 7620 }, { "grad_norm": 0.1537860929965973, "learning_rate": 1.4597924696737835e-05, "loss": 0.0145, "step": 7630 }, { "grad_norm": 0.1758957952260971, "learning_rate": 1.4481355271723252e-05, "loss": 0.0144, "step": 7640 }, { "grad_norm": 0.11952029168605804, "learning_rate": 1.4365174272916809e-05, "loss": 0.0167, "step": 7650 }, { "grad_norm": 0.2211906611919403, "learning_rate": 1.4249382970855319e-05, "loss": 0.0156, "step": 7660 }, { "grad_norm": 0.14674614369869232, "learning_rate": 1.4133982631813903e-05, "loss": 0.016, "step": 7670 }, { "grad_norm": 0.1289716511964798, "learning_rate": 1.4018974517792194e-05, "loss": 0.0133, "step": 7680 }, { "grad_norm": 0.17677290737628937, "learning_rate": 1.390435988650048e-05, "loss": 0.0143, "step": 7690 }, { "grad_norm": 0.13738666474819183, "learning_rate": 1.3790139991346006e-05, "loss": 0.0141, "step": 7700 }, { "grad_norm": 0.25329065322875977, "learning_rate": 1.367631608141926e-05, "loss": 0.0154, "step": 7710 }, { "grad_norm": 0.13695746660232544, "learning_rate": 1.3562889401480278e-05, "loss": 0.0128, "step": 7720 }, { "grad_norm": 0.07748368382453918, "learning_rate": 1.3449861191945074e-05, "loss": 0.0151, "step": 7730 }, { "grad_norm": 0.1905820667743683, "learning_rate": 1.3337232688872009e-05, "loss": 0.0125, "step": 7740 }, { "grad_norm": 0.15634295344352722, "learning_rate": 1.3225005123948364e-05, "loss": 0.0144, "step": 7750 }, { "grad_norm": 0.14439263939857483, "learning_rate": 1.311317972447681e-05, "loss": 0.0135, "step": 7760 }, { "grad_norm": 0.14225336909294128, "learning_rate": 1.3001757713361996e-05, "loss": 0.0104, "step": 7770 }, { "grad_norm": 0.15916700661182404, "learning_rate": 1.2890740309097204e-05, "loss": 0.0132, "step": 7780 }, { "grad_norm": 0.13703936338424683, "learning_rate": 1.2780128725750944e-05, "loss": 0.0111, "step": 7790 }, { "grad_norm": 0.12058528512716293, "learning_rate": 1.266992417295379e-05, "loss": 0.0137, "step": 7800 }, { "grad_norm": 0.17314349114894867, "learning_rate": 1.2560127855885073e-05, "loss": 0.0137, "step": 7810 }, { "grad_norm": 0.24393682181835175, "learning_rate": 1.2450740975259745e-05, "loss": 0.017, "step": 7820 }, { "grad_norm": 0.14594079554080963, "learning_rate": 1.234176472731517e-05, "loss": 0.0209, "step": 7830 }, { "grad_norm": 0.1699279248714447, "learning_rate": 1.2233200303798158e-05, "loss": 0.0135, "step": 7840 }, { "grad_norm": 0.15933039784431458, "learning_rate": 1.2125048891951846e-05, "loss": 0.0134, "step": 7850 }, { "grad_norm": 0.11323697865009308, "learning_rate": 1.2017311674502745e-05, "loss": 0.0118, "step": 7860 }, { "grad_norm": 0.14450101554393768, "learning_rate": 1.1909989829647822e-05, "loss": 0.0129, "step": 7870 }, { "grad_norm": 0.12027398496866226, "learning_rate": 1.1803084531041553e-05, "loss": 0.0135, "step": 7880 }, { "grad_norm": 0.17379429936408997, "learning_rate": 1.1696596947783162e-05, "loss": 0.014, "step": 7890 }, { "grad_norm": 0.16011232137680054, "learning_rate": 1.1590528244403803e-05, "loss": 0.0126, "step": 7900 }, { "grad_norm": 0.24031603336334229, "learning_rate": 1.148487958085382e-05, "loss": 0.0194, "step": 7910 }, { "grad_norm": 0.11084669083356857, "learning_rate": 1.1379652112490086e-05, "loss": 0.0122, "step": 7920 }, { "grad_norm": 0.21068330109119415, "learning_rate": 1.1274846990063315e-05, "loss": 0.0137, "step": 7930 }, { "grad_norm": 0.14663046598434448, "learning_rate": 1.117046535970554e-05, "loss": 0.0129, "step": 7940 }, { "grad_norm": 0.12927627563476562, "learning_rate": 1.106650836291755e-05, "loss": 0.0195, "step": 7950 }, { "grad_norm": 0.1668599396944046, "learning_rate": 1.0962977136556418e-05, "loss": 0.0208, "step": 7960 }, { "grad_norm": 0.14867907762527466, "learning_rate": 1.0859872812823024e-05, "loss": 0.0159, "step": 7970 }, { "grad_norm": 0.1827944964170456, "learning_rate": 1.0757196519249747e-05, "loss": 0.0204, "step": 7980 }, { "grad_norm": 0.19044557213783264, "learning_rate": 1.0654949378688077e-05, "loss": 0.013, "step": 7990 }, { "grad_norm": 0.12824290990829468, "learning_rate": 1.0553132509296376e-05, "loss": 0.013, "step": 8000 }, { "grad_norm": 0.24974271655082703, "learning_rate": 1.0451747024527613e-05, "loss": 0.0148, "step": 8010 }, { "grad_norm": 0.09178805351257324, "learning_rate": 1.0350794033117189e-05, "loss": 0.0135, "step": 8020 }, { "grad_norm": 0.08583728224039078, "learning_rate": 1.0250274639070856e-05, "loss": 0.0144, "step": 8030 }, { "grad_norm": 0.15035970509052277, "learning_rate": 1.0150189941652599e-05, "loss": 0.0116, "step": 8040 }, { "grad_norm": 0.18453428149223328, "learning_rate": 1.0050541035372635e-05, "loss": 0.0108, "step": 8050 }, { "grad_norm": 0.17961032688617706, "learning_rate": 9.951329009975458e-06, "loss": 0.0137, "step": 8060 }, { "grad_norm": 0.18854476511478424, "learning_rate": 9.852554950427845e-06, "loss": 0.0138, "step": 8070 }, { "grad_norm": 0.08875742554664612, "learning_rate": 9.754219936907105e-06, "loss": 0.0131, "step": 8080 }, { "grad_norm": 0.09984441101551056, "learning_rate": 9.656325044789194e-06, "loss": 0.0111, "step": 8090 }, { "grad_norm": 0.15604174137115479, "learning_rate": 9.55887134463697e-06, "loss": 0.0125, "step": 8100 }, { "grad_norm": 0.12224118411540985, "learning_rate": 9.461859902188475e-06, "loss": 0.0133, "step": 8110 }, { "grad_norm": 0.15887723863124847, "learning_rate": 9.365291778345303e-06, "loss": 0.0148, "step": 8120 }, { "grad_norm": 0.14053001999855042, "learning_rate": 9.269168029160991e-06, "loss": 0.0101, "step": 8130 }, { "grad_norm": 0.18022020161151886, "learning_rate": 9.173489705829447e-06, "loss": 0.0152, "step": 8140 }, { "grad_norm": 0.19975541532039642, "learning_rate": 9.078257854673516e-06, "loss": 0.0152, "step": 8150 }, { "grad_norm": 0.09093631058931351, "learning_rate": 8.983473517133429e-06, "loss": 0.0119, "step": 8160 }, { "grad_norm": 0.17694291472434998, "learning_rate": 8.889137729755537e-06, "loss": 0.0123, "step": 8170 }, { "grad_norm": 0.13523834943771362, "learning_rate": 8.79525152418087e-06, "loss": 0.0157, "step": 8180 }, { "grad_norm": 0.10322197526693344, "learning_rate": 8.701815927133961e-06, "loss": 0.0128, "step": 8190 }, { "grad_norm": 0.08569823950529099, "learning_rate": 8.608831960411534e-06, "loss": 0.0126, "step": 8200 }, { "grad_norm": 0.1410985291004181, "learning_rate": 8.516300640871321e-06, "loss": 0.0128, "step": 8210 }, { "grad_norm": 0.1109914481639862, "learning_rate": 8.424222980421038e-06, "loss": 0.0149, "step": 8220 }, { "grad_norm": 0.17577391862869263, "learning_rate": 8.332599986007184e-06, "loss": 0.0129, "step": 8230 }, { "grad_norm": 0.08530381321907043, "learning_rate": 8.241432659604203e-06, "loss": 0.012, "step": 8240 }, { "grad_norm": 0.12259893864393234, "learning_rate": 8.150721998203331e-06, "loss": 0.0109, "step": 8250 }, { "grad_norm": 0.13014091551303864, "learning_rate": 8.06046899380184e-06, "loss": 0.0133, "step": 8260 }, { "grad_norm": 0.21877069771289825, "learning_rate": 7.970674633392133e-06, "loss": 0.0205, "step": 8270 }, { "grad_norm": 0.14529435336589813, "learning_rate": 7.881339898950924e-06, "loss": 0.0116, "step": 8280 }, { "grad_norm": 0.14572905004024506, "learning_rate": 7.792465767428597e-06, "loss": 0.0127, "step": 8290 }, { "grad_norm": 0.0887068584561348, "learning_rate": 7.704053210738376e-06, "loss": 0.0092, "step": 8300 }, { "grad_norm": 0.15583845973014832, "learning_rate": 7.6161031957458494e-06, "loss": 0.0166, "step": 8310 }, { "grad_norm": 0.19063858687877655, "learning_rate": 7.5286166842582605e-06, "loss": 0.015, "step": 8320 }, { "grad_norm": 0.13634401559829712, "learning_rate": 7.4415946330140814e-06, "loss": 0.0135, "step": 8330 }, { "grad_norm": 0.11061529070138931, "learning_rate": 7.3550379936725644e-06, "loss": 0.0112, "step": 8340 }, { "grad_norm": 0.09183473885059357, "learning_rate": 7.2689477128032035e-06, "loss": 0.0134, "step": 8350 }, { "grad_norm": 0.17421060800552368, "learning_rate": 7.183324731875551e-06, "loss": 0.0123, "step": 8360 }, { "grad_norm": 0.09720266610383987, "learning_rate": 7.098169987248782e-06, "loss": 0.0098, "step": 8370 }, { "grad_norm": 0.12905731797218323, "learning_rate": 7.013484410161553e-06, "loss": 0.0133, "step": 8380 }, { "grad_norm": 0.15862785279750824, "learning_rate": 6.92926892672176e-06, "loss": 0.016, "step": 8390 }, { "grad_norm": 0.11174117028713226, "learning_rate": 6.845524457896446e-06, "loss": 0.0108, "step": 8400 }, { "grad_norm": 0.14331871271133423, "learning_rate": 6.7622519195017165e-06, "loss": 0.0117, "step": 8410 }, { "grad_norm": 0.10665357112884521, "learning_rate": 6.679452222192684e-06, "loss": 0.0109, "step": 8420 }, { "grad_norm": 0.1437055617570877, "learning_rate": 6.597126271453579e-06, "loss": 0.0116, "step": 8430 }, { "grad_norm": 0.16377826035022736, "learning_rate": 6.51527496758782e-06, "loss": 0.011, "step": 8440 }, { "grad_norm": 0.13747504353523254, "learning_rate": 6.433899205708155e-06, "loss": 0.0105, "step": 8450 }, { "grad_norm": 0.08898147940635681, "learning_rate": 6.352999875726856e-06, "loss": 0.0089, "step": 8460 }, { "grad_norm": 0.1428779661655426, "learning_rate": 6.272577862346052e-06, "loss": 0.0097, "step": 8470 }, { "grad_norm": 0.1656123846769333, "learning_rate": 6.192634045047996e-06, "loss": 0.0178, "step": 8480 }, { "grad_norm": 0.1553337574005127, "learning_rate": 6.113169298085458e-06, "loss": 0.0136, "step": 8490 }, { "grad_norm": 0.1642589420080185, "learning_rate": 6.034184490472195e-06, "loss": 0.0107, "step": 8500 }, { "grad_norm": 0.17997749149799347, "learning_rate": 5.955680485973386e-06, "loss": 0.0131, "step": 8510 }, { "grad_norm": 0.16968750953674316, "learning_rate": 5.877658143096265e-06, "loss": 0.012, "step": 8520 }, { "grad_norm": 0.1405196189880371, "learning_rate": 5.800118315080661e-06, "loss": 0.0096, "step": 8530 }, { "grad_norm": 0.19422386586666107, "learning_rate": 5.723061849889716e-06, "loss": 0.0101, "step": 8540 }, { "grad_norm": 0.13218717277050018, "learning_rate": 5.646489590200604e-06, "loss": 0.013, "step": 8550 }, { "grad_norm": 0.11157453805208206, "learning_rate": 5.570402373395256e-06, "loss": 0.0103, "step": 8560 }, { "grad_norm": 0.14696305990219116, "learning_rate": 5.494801031551305e-06, "loss": 0.0161, "step": 8570 }, { "grad_norm": 0.13735026121139526, "learning_rate": 5.41968639143291e-06, "loss": 0.0114, "step": 8580 }, { "grad_norm": 0.21167024970054626, "learning_rate": 5.345059274481751e-06, "loss": 0.0111, "step": 8590 }, { "grad_norm": 0.14754518866539001, "learning_rate": 5.270920496808002e-06, "loss": 0.0134, "step": 8600 }, { "grad_norm": 0.23152942955493927, "learning_rate": 5.1972708691814695e-06, "loss": 0.0134, "step": 8610 }, { "grad_norm": 0.12972496449947357, "learning_rate": 5.124111197022674e-06, "loss": 0.0142, "step": 8620 }, { "grad_norm": 0.18226200342178345, "learning_rate": 5.051442280394081e-06, "loss": 0.0101, "step": 8630 }, { "grad_norm": 0.14536815881729126, "learning_rate": 4.979264913991322e-06, "loss": 0.0117, "step": 8640 }, { "grad_norm": 0.16113154590129852, "learning_rate": 4.907579887134489e-06, "loss": 0.0155, "step": 8650 }, { "grad_norm": 0.19396112859249115, "learning_rate": 4.836387983759572e-06, "loss": 0.0114, "step": 8660 }, { "grad_norm": 0.17245811223983765, "learning_rate": 4.765689982409816e-06, "loss": 0.0114, "step": 8670 }, { "grad_norm": 0.13400056958198547, "learning_rate": 4.695486656227233e-06, "loss": 0.0147, "step": 8680 }, { "grad_norm": 0.09909213334321976, "learning_rate": 4.625778772944156e-06, "loss": 0.0114, "step": 8690 }, { "grad_norm": 0.12754808366298676, "learning_rate": 4.556567094874825e-06, "loss": 0.0129, "step": 8700 }, { "grad_norm": 0.11776209622621536, "learning_rate": 4.487852378907059e-06, "loss": 0.0095, "step": 8710 }, { "grad_norm": 0.1655227690935135, "learning_rate": 4.419635376493986e-06, "loss": 0.0122, "step": 8720 }, { "grad_norm": 0.12749098241329193, "learning_rate": 4.351916833645825e-06, "loss": 0.0142, "step": 8730 }, { "grad_norm": 0.1500367969274521, "learning_rate": 4.284697490921691e-06, "loss": 0.0119, "step": 8740 }, { "grad_norm": 0.14680761098861694, "learning_rate": 4.2179780834215585e-06, "loss": 0.0131, "step": 8750 }, { "grad_norm": 0.21032112836837769, "learning_rate": 4.151759340778178e-06, "loss": 0.0137, "step": 8760 }, { "grad_norm": 0.13834387063980103, "learning_rate": 4.086041987149109e-06, "loss": 0.0116, "step": 8770 }, { "grad_norm": 0.22972474992275238, "learning_rate": 4.020826741208811e-06, "loss": 0.0129, "step": 8780 }, { "grad_norm": 0.1458142101764679, "learning_rate": 3.956114316140746e-06, "loss": 0.0163, "step": 8790 }, { "grad_norm": 0.15836627781391144, "learning_rate": 3.891905419629643e-06, "loss": 0.0104, "step": 8800 }, { "grad_norm": 0.11640172451734543, "learning_rate": 3.8282007538536946e-06, "loss": 0.0151, "step": 8810 }, { "grad_norm": 0.1954166740179062, "learning_rate": 3.7650010154769265e-06, "loss": 0.0133, "step": 8820 }, { "grad_norm": 0.17176075279712677, "learning_rate": 3.7023068956415608e-06, "loss": 0.0114, "step": 8830 }, { "grad_norm": 0.079417884349823, "learning_rate": 3.6401190799604303e-06, "loss": 0.0089, "step": 8840 }, { "grad_norm": 0.15849106013774872, "learning_rate": 3.578438248509536e-06, "loss": 0.0122, "step": 8850 }, { "grad_norm": 0.09843376278877258, "learning_rate": 3.5172650758205583e-06, "loss": 0.0142, "step": 8860 }, { "grad_norm": 0.1148449257016182, "learning_rate": 3.45660023087353e-06, "loss": 0.0099, "step": 8870 }, { "grad_norm": 0.14963871240615845, "learning_rate": 3.3964443770894528e-06, "loss": 0.0131, "step": 8880 }, { "grad_norm": 0.17750944197177887, "learning_rate": 3.3367981723231245e-06, "loss": 0.0103, "step": 8890 }, { "grad_norm": 0.21314233541488647, "learning_rate": 3.2776622688558746e-06, "loss": 0.0151, "step": 8900 }, { "grad_norm": 0.17068736255168915, "learning_rate": 3.2190373133884677e-06, "loss": 0.0086, "step": 8910 }, { "grad_norm": 0.16469429433345795, "learning_rate": 3.1609239470340446e-06, "loss": 0.0101, "step": 8920 }, { "grad_norm": 0.11852634698152542, "learning_rate": 3.1033228053110373e-06, "loss": 0.0094, "step": 8930 }, { "grad_norm": 0.10194798558950424, "learning_rate": 3.0462345181363314e-06, "loss": 0.0096, "step": 8940 }, { "grad_norm": 0.14574280381202698, "learning_rate": 2.9896597098182654e-06, "loss": 0.0145, "step": 8950 }, { "grad_norm": 0.21890714764595032, "learning_rate": 2.933598999049891e-06, "loss": 0.015, "step": 8960 }, { "grad_norm": 0.11626438796520233, "learning_rate": 2.8780529989021697e-06, "loss": 0.0098, "step": 8970 }, { "grad_norm": 0.11380519717931747, "learning_rate": 2.823022316817242e-06, "loss": 0.0134, "step": 8980 }, { "grad_norm": 0.14622558653354645, "learning_rate": 2.7685075546018456e-06, "loss": 0.0118, "step": 8990 }, { "grad_norm": 0.13948456943035126, "learning_rate": 2.7145093084206598e-06, "loss": 0.0153, "step": 9000 }, { "grad_norm": 0.08545006811618805, "learning_rate": 2.661028168789892e-06, "loss": 0.0116, "step": 9010 }, { "grad_norm": 0.15577736496925354, "learning_rate": 2.6080647205706855e-06, "loss": 0.0096, "step": 9020 }, { "grad_norm": 0.07484491914510727, "learning_rate": 2.555619542962834e-06, "loss": 0.0101, "step": 9030 }, { "grad_norm": 0.08874287456274033, "learning_rate": 2.503693209498409e-06, "loss": 0.0095, "step": 9040 }, { "grad_norm": 0.10317665338516235, "learning_rate": 2.452286288035449e-06, "loss": 0.0122, "step": 9050 }, { "grad_norm": 0.06863200664520264, "learning_rate": 2.4013993407518363e-06, "loss": 0.0117, "step": 9060 }, { "grad_norm": 0.15778382122516632, "learning_rate": 2.351032924139063e-06, "loss": 0.0094, "step": 9070 }, { "grad_norm": 0.11813550442457199, "learning_rate": 2.30118758899619e-06, "loss": 0.011, "step": 9080 }, { "grad_norm": 0.07327229529619217, "learning_rate": 2.2518638804238157e-06, "loss": 0.0103, "step": 9090 }, { "grad_norm": 0.09998597949743271, "learning_rate": 2.203062337818118e-06, "loss": 0.0116, "step": 9100 }, { "grad_norm": 0.17128537595272064, "learning_rate": 2.1547834948649483e-06, "loss": 0.0135, "step": 9110 }, { "grad_norm": 0.1948433369398117, "learning_rate": 2.1070278795340017e-06, "loss": 0.0108, "step": 9120 }, { "grad_norm": 0.10191141068935394, "learning_rate": 2.059796014073029e-06, "loss": 0.0086, "step": 9130 }, { "grad_norm": 0.1465614140033722, "learning_rate": 2.01308841500214e-06, "loss": 0.0126, "step": 9140 }, { "grad_norm": 0.1827053725719452, "learning_rate": 1.9669055931081704e-06, "loss": 0.0156, "step": 9150 }, { "grad_norm": 0.12135829776525497, "learning_rate": 1.9212480534390507e-06, "loss": 0.0122, "step": 9160 }, { "grad_norm": 0.2075703889131546, "learning_rate": 1.8761162952983246e-06, "loss": 0.0132, "step": 9170 }, { "grad_norm": 0.19804269075393677, "learning_rate": 1.8315108122396618e-06, "loss": 0.0122, "step": 9180 }, { "grad_norm": 0.0665370523929596, "learning_rate": 1.787432092061475e-06, "loss": 0.0091, "step": 9190 }, { "grad_norm": 0.08151435852050781, "learning_rate": 1.743880616801602e-06, "loss": 0.0138, "step": 9200 }, { "grad_norm": 0.19633319973945618, "learning_rate": 1.7008568627319865e-06, "loss": 0.0102, "step": 9210 }, { "grad_norm": 0.17719054222106934, "learning_rate": 1.6583613003535226e-06, "loss": 0.0147, "step": 9220 }, { "grad_norm": 0.17035585641860962, "learning_rate": 1.6163943943908522e-06, "loss": 0.0107, "step": 9230 }, { "grad_norm": 0.16717898845672607, "learning_rate": 1.5749566037873476e-06, "loss": 0.011, "step": 9240 }, { "grad_norm": 0.19066427648067474, "learning_rate": 1.5340483817000428e-06, "loss": 0.0127, "step": 9250 }, { "grad_norm": 0.15096762776374817, "learning_rate": 1.4936701754947101e-06, "loss": 0.0119, "step": 9260 }, { "grad_norm": 0.12303285300731659, "learning_rate": 1.4538224267409361e-06, "loss": 0.0119, "step": 9270 }, { "grad_norm": 0.13047857582569122, "learning_rate": 1.414505571207314e-06, "loss": 0.009, "step": 9280 }, { "grad_norm": 0.10579918324947357, "learning_rate": 1.3757200388566816e-06, "loss": 0.0116, "step": 9290 }, { "grad_norm": 0.19899429380893707, "learning_rate": 1.3374662538414074e-06, "loss": 0.0134, "step": 9300 }, { "grad_norm": 0.1685827076435089, "learning_rate": 1.2997446344987617e-06, "loss": 0.0118, "step": 9310 }, { "grad_norm": 0.12367964535951614, "learning_rate": 1.262555593346315e-06, "loss": 0.0121, "step": 9320 }, { "grad_norm": 0.17311914265155792, "learning_rate": 1.2258995370774685e-06, "loss": 0.0126, "step": 9330 }, { "grad_norm": 0.12149365246295929, "learning_rate": 1.1897768665569798e-06, "loss": 0.0125, "step": 9340 }, { "grad_norm": 0.10873806476593018, "learning_rate": 1.1541879768165954e-06, "loss": 0.0095, "step": 9350 }, { "grad_norm": 0.1434645652770996, "learning_rate": 1.1191332570507085e-06, "loss": 0.0125, "step": 9360 }, { "grad_norm": 0.09381873160600662, "learning_rate": 1.0846130906121132e-06, "loss": 0.0151, "step": 9370 }, { "grad_norm": 0.14483855664730072, "learning_rate": 1.0506278550078131e-06, "loss": 0.0147, "step": 9380 }, { "grad_norm": 0.17843471467494965, "learning_rate": 1.0171779218949185e-06, "loss": 0.0141, "step": 9390 }, { "grad_norm": 0.12885360419750214, "learning_rate": 9.842636570765174e-07, "loss": 0.0084, "step": 9400 }, { "grad_norm": 0.1116366982460022, "learning_rate": 9.518854204977612e-07, "loss": 0.0097, "step": 9410 }, { "grad_norm": 0.15187424421310425, "learning_rate": 9.200435662418349e-07, "loss": 0.0133, "step": 9420 }, { "grad_norm": 0.17909541726112366, "learning_rate": 8.887384425261658e-07, "loss": 0.0102, "step": 9430 }, { "grad_norm": 0.19550544023513794, "learning_rate": 8.579703916985648e-07, "loss": 0.0149, "step": 9440 }, { "grad_norm": 0.04908235743641853, "learning_rate": 8.277397502335194e-07, "loss": 0.0103, "step": 9450 }, { "grad_norm": 0.1798756867647171, "learning_rate": 7.980468487284675e-07, "loss": 0.0133, "step": 9460 }, { "grad_norm": 0.11874180287122726, "learning_rate": 7.688920119002297e-07, "loss": 0.012, "step": 9470 }, { "grad_norm": 0.11021319776773453, "learning_rate": 7.402755585814269e-07, "loss": 0.0107, "step": 9480 }, { "grad_norm": 0.14386406540870667, "learning_rate": 7.121978017170073e-07, "loss": 0.0123, "step": 9490 }, { "grad_norm": 0.20570838451385498, "learning_rate": 6.846590483608306e-07, "loss": 0.0139, "step": 9500 }, { "grad_norm": 0.16686032712459564, "learning_rate": 6.576595996722834e-07, "loss": 0.0112, "step": 9510 }, { "grad_norm": 0.18355204164981842, "learning_rate": 6.311997509130141e-07, "loss": 0.0146, "step": 9520 }, { "grad_norm": 0.22738204896450043, "learning_rate": 6.052797914436803e-07, "loss": 0.011, "step": 9530 }, { "grad_norm": 0.10810808837413788, "learning_rate": 5.799000047208181e-07, "loss": 0.0151, "step": 9540 }, { "grad_norm": 0.18656066060066223, "learning_rate": 5.550606682937054e-07, "loss": 0.0154, "step": 9550 }, { "grad_norm": 0.1370597928762436, "learning_rate": 5.307620538013481e-07, "loss": 0.014, "step": 9560 }, { "grad_norm": 0.1516852080821991, "learning_rate": 5.070044269694874e-07, "loss": 0.0117, "step": 9570 }, { "grad_norm": 0.18828630447387695, "learning_rate": 4.837880476077417e-07, "loss": 0.0139, "step": 9580 }, { "grad_norm": 0.06233203411102295, "learning_rate": 4.6111316960670835e-07, "loss": 0.0093, "step": 9590 }, { "grad_norm": 0.1417885571718216, "learning_rate": 4.389800409352218e-07, "loss": 0.0112, "step": 9600 }, { "grad_norm": 0.25167733430862427, "learning_rate": 4.173889036376277e-07, "loss": 0.0156, "step": 9610 }, { "grad_norm": 0.15105876326560974, "learning_rate": 3.963399938311463e-07, "loss": 0.0122, "step": 9620 }, { "grad_norm": 0.15377460420131683, "learning_rate": 3.7583354170328545e-07, "loss": 0.0143, "step": 9630 }, { "grad_norm": 0.069511279463768, "learning_rate": 3.558697715093207e-07, "loss": 0.0089, "step": 9640 }, { "grad_norm": 0.18190111219882965, "learning_rate": 3.3644890156983576e-07, "loss": 0.0119, "step": 9650 }, { "grad_norm": 0.1350547820329666, "learning_rate": 3.175711442683638e-07, "loss": 0.0123, "step": 9660 }, { "grad_norm": 0.13941970467567444, "learning_rate": 2.9923670604902197e-07, "loss": 0.0119, "step": 9670 }, { "grad_norm": 0.19192935526371002, "learning_rate": 2.814457874143028e-07, "loss": 0.0134, "step": 9680 }, { "grad_norm": 0.13127397000789642, "learning_rate": 2.641985829228366e-07, "loss": 0.0102, "step": 9690 }, { "grad_norm": 0.09755296260118484, "learning_rate": 2.474952811872877e-07, "loss": 0.0103, "step": 9700 }, { "grad_norm": 0.09053207188844681, "learning_rate": 2.3133606487228397e-07, "loss": 0.0121, "step": 9710 }, { "grad_norm": 0.12107087671756744, "learning_rate": 2.157211106924295e-07, "loss": 0.0131, "step": 9720 }, { "grad_norm": 0.2682044506072998, "learning_rate": 2.006505894103672e-07, "loss": 0.0125, "step": 9730 }, { "grad_norm": 0.06316626071929932, "learning_rate": 1.8612466583489696e-07, "loss": 0.0101, "step": 9740 }, { "grad_norm": 0.09964451193809509, "learning_rate": 1.7214349881918834e-07, "loss": 0.0103, "step": 9750 }, { "grad_norm": 0.11377908289432526, "learning_rate": 1.5870724125904845e-07, "loss": 0.011, "step": 9760 }, { "grad_norm": 0.06940179318189621, "learning_rate": 1.4581604009124006e-07, "loss": 0.0101, "step": 9770 }, { "grad_norm": 0.14716176688671112, "learning_rate": 1.334700362918717e-07, "loss": 0.0111, "step": 9780 }, { "grad_norm": 0.13093708455562592, "learning_rate": 1.2166936487486015e-07, "loss": 0.0129, "step": 9790 }, { "grad_norm": 0.09536731243133545, "learning_rate": 1.1041415489045914e-07, "loss": 0.008, "step": 9800 }, { "grad_norm": 0.14768612384796143, "learning_rate": 9.970452942384412e-08, "loss": 0.0126, "step": 9810 }, { "grad_norm": 0.17343644797801971, "learning_rate": 8.954060559375754e-08, "loss": 0.009, "step": 9820 }, { "grad_norm": 0.14313387870788574, "learning_rate": 7.99224945512489e-08, "loss": 0.0151, "step": 9830 }, { "grad_norm": 0.11853434890508652, "learning_rate": 7.085030147843675e-08, "loss": 0.014, "step": 9840 }, { "grad_norm": 0.1830025464296341, "learning_rate": 6.232412558736523e-08, "loss": 0.0144, "step": 9850 }, { "grad_norm": 0.14608250558376312, "learning_rate": 5.434406011893822e-08, "loss": 0.0132, "step": 9860 }, { "grad_norm": 0.24710986018180847, "learning_rate": 4.6910192341864664e-08, "loss": 0.011, "step": 9870 }, { "grad_norm": 0.13796855509281158, "learning_rate": 4.0022603551737035e-08, "loss": 0.0124, "step": 9880 }, { "grad_norm": 0.15818729996681213, "learning_rate": 3.3681369070120985e-08, "loss": 0.0164, "step": 9890 }, { "grad_norm": 0.11452880501747131, "learning_rate": 2.7886558243744866e-08, "loss": 0.0119, "step": 9900 }, { "grad_norm": 0.16273202002048492, "learning_rate": 2.2638234443722596e-08, "loss": 0.0128, "step": 9910 }, { "grad_norm": 0.1674075424671173, "learning_rate": 1.7936455064887504e-08, "loss": 0.0144, "step": 9920 }, { "grad_norm": 0.11069000512361526, "learning_rate": 1.378127152514841e-08, "loss": 0.0079, "step": 9930 }, { "grad_norm": 0.12962037324905396, "learning_rate": 1.0172729264917857e-08, "loss": 0.0144, "step": 9940 }, { "grad_norm": 0.20579205453395844, "learning_rate": 7.1108677466458215e-09, "loss": 0.0147, "step": 9950 }, { "grad_norm": 0.14993710815906525, "learning_rate": 4.595720454353414e-09, "loss": 0.0118, "step": 9960 }, { "grad_norm": 0.15225958824157715, "learning_rate": 2.627314893294264e-09, "loss": 0.0115, "step": 9970 }, { "grad_norm": 0.10546813160181046, "learning_rate": 1.2056725896270048e-09, "loss": 0.011, "step": 9980 }, { "grad_norm": 0.12684182822704315, "learning_rate": 3.308090902098826e-10, "loss": 0.0123, "step": 9990 }, { "grad_norm": 0.17403271794319153, "learning_rate": 2.7339624120159555e-12, "loss": 0.0128, "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 }