{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.6, "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": 2.6956980228424072, "learning_rate": 1.8e-06, "loss": 1.1661, "step": 10 }, { "grad_norm": 1.908838152885437, "learning_rate": 3.8e-06, "loss": 1.1289, "step": 20 }, { "grad_norm": 0.8673072457313538, "learning_rate": 5.8e-06, "loss": 1.0677, "step": 30 }, { "grad_norm": 0.8871041536331177, "learning_rate": 7.8e-06, "loss": 1.0497, "step": 40 }, { "grad_norm": 0.7078301310539246, "learning_rate": 9.800000000000001e-06, "loss": 1.0613, "step": 50 }, { "grad_norm": 0.6471117734909058, "learning_rate": 1.18e-05, "loss": 1.0667, "step": 60 }, { "grad_norm": 0.71360844373703, "learning_rate": 1.3800000000000002e-05, "loss": 1.0635, "step": 70 }, { "grad_norm": 0.8140817880630493, "learning_rate": 1.58e-05, "loss": 1.0675, "step": 80 }, { "grad_norm": 0.5220566987991333, "learning_rate": 1.78e-05, "loss": 1.0462, "step": 90 }, { "grad_norm": 0.40389445424079895, "learning_rate": 1.9800000000000004e-05, "loss": 1.0536, "step": 100 }, { "grad_norm": 0.40283504128456116, "learning_rate": 2.18e-05, "loss": 1.0486, "step": 110 }, { "grad_norm": 0.5318396687507629, "learning_rate": 2.38e-05, "loss": 1.0336, "step": 120 }, { "grad_norm": 0.3760662078857422, "learning_rate": 2.58e-05, "loss": 1.0421, "step": 130 }, { "grad_norm": 0.6521379947662354, "learning_rate": 2.7800000000000005e-05, "loss": 1.0257, "step": 140 }, { "grad_norm": 1.4326046705245972, "learning_rate": 2.98e-05, "loss": 0.972, "step": 150 }, { "grad_norm": 1.5177102088928223, "learning_rate": 3.18e-05, "loss": 0.9129, "step": 160 }, { "grad_norm": 1.0445795059204102, "learning_rate": 3.38e-05, "loss": 0.8824, "step": 170 }, { "grad_norm": 1.8929375410079956, "learning_rate": 3.58e-05, "loss": 0.8328, "step": 180 }, { "grad_norm": 1.9234741926193237, "learning_rate": 3.7800000000000004e-05, "loss": 0.7208, "step": 190 }, { "grad_norm": 2.0339930057525635, "learning_rate": 3.9800000000000005e-05, "loss": 0.6035, "step": 200 }, { "grad_norm": 2.2198052406311035, "learning_rate": 4.18e-05, "loss": 0.4436, "step": 210 }, { "grad_norm": 2.5448694229125977, "learning_rate": 4.38e-05, "loss": 0.2995, "step": 220 }, { "grad_norm": 2.438976526260376, "learning_rate": 4.58e-05, "loss": 0.2238, "step": 230 }, { "grad_norm": 1.8254433870315552, "learning_rate": 4.78e-05, "loss": 0.1835, "step": 240 }, { "grad_norm": 1.5078606605529785, "learning_rate": 4.9800000000000004e-05, "loss": 0.1575, "step": 250 }, { "grad_norm": 1.18130362033844, "learning_rate": 5.1800000000000005e-05, "loss": 0.1487, "step": 260 }, { "grad_norm": 1.734771728515625, "learning_rate": 5.380000000000001e-05, "loss": 0.1567, "step": 270 }, { "grad_norm": 1.3391666412353516, "learning_rate": 5.580000000000001e-05, "loss": 0.1318, "step": 280 }, { "grad_norm": 1.360201120376587, "learning_rate": 5.7799999999999995e-05, "loss": 0.14, "step": 290 }, { "grad_norm": 1.8701971769332886, "learning_rate": 5.9800000000000003e-05, "loss": 0.126, "step": 300 }, { "grad_norm": 1.2580232620239258, "learning_rate": 6.18e-05, "loss": 0.1337, "step": 310 }, { "grad_norm": 1.0569756031036377, "learning_rate": 6.38e-05, "loss": 0.113, "step": 320 }, { "grad_norm": 1.0326085090637207, "learning_rate": 6.58e-05, "loss": 0.1001, "step": 330 }, { "grad_norm": 1.2274115085601807, "learning_rate": 6.780000000000001e-05, "loss": 0.1051, "step": 340 }, { "grad_norm": 0.8806998133659363, "learning_rate": 6.98e-05, "loss": 0.0988, "step": 350 }, { "grad_norm": 0.9884889721870422, "learning_rate": 7.18e-05, "loss": 0.1092, "step": 360 }, { "grad_norm": 0.9650731086730957, "learning_rate": 7.38e-05, "loss": 0.1167, "step": 370 }, { "grad_norm": 1.2655612230300903, "learning_rate": 7.58e-05, "loss": 0.0968, "step": 380 }, { "grad_norm": 0.755662202835083, "learning_rate": 7.780000000000001e-05, "loss": 0.106, "step": 390 }, { "grad_norm": 0.7026094198226929, "learning_rate": 7.98e-05, "loss": 0.1001, "step": 400 }, { "grad_norm": 0.6749414801597595, "learning_rate": 8.18e-05, "loss": 0.101, "step": 410 }, { "grad_norm": 0.7273757457733154, "learning_rate": 8.38e-05, "loss": 0.0938, "step": 420 }, { "grad_norm": 0.9588629603385925, "learning_rate": 8.58e-05, "loss": 0.1004, "step": 430 }, { "grad_norm": 0.759544849395752, "learning_rate": 8.78e-05, "loss": 0.0917, "step": 440 }, { "grad_norm": 0.6082161664962769, "learning_rate": 8.98e-05, "loss": 0.1035, "step": 450 }, { "grad_norm": 0.75276118516922, "learning_rate": 9.180000000000001e-05, "loss": 0.1089, "step": 460 }, { "grad_norm": 0.7040882706642151, "learning_rate": 9.38e-05, "loss": 0.0997, "step": 470 }, { "grad_norm": 0.6963564157485962, "learning_rate": 9.58e-05, "loss": 0.1013, "step": 480 }, { "grad_norm": 0.7889728546142578, "learning_rate": 9.78e-05, "loss": 0.0888, "step": 490 }, { "grad_norm": 0.6861390471458435, "learning_rate": 9.98e-05, "loss": 0.0993, "step": 500 }, { "grad_norm": 0.683660089969635, "learning_rate": 9.9999778549206e-05, "loss": 0.0975, "step": 510 }, { "grad_norm": 0.7197879552841187, "learning_rate": 9.999901304280685e-05, "loss": 0.0819, "step": 520 }, { "grad_norm": 0.5837870240211487, "learning_rate": 9.999770075521164e-05, "loss": 0.0863, "step": 530 }, { "grad_norm": 0.5479613542556763, "learning_rate": 9.99958417007713e-05, "loss": 0.0786, "step": 540 }, { "grad_norm": 0.5500237941741943, "learning_rate": 9.999343589981615e-05, "loss": 0.0898, "step": 550 }, { "grad_norm": 0.4174486994743347, "learning_rate": 9.999048337865568e-05, "loss": 0.0786, "step": 560 }, { "grad_norm": 0.7582618594169617, "learning_rate": 9.998698416957815e-05, "loss": 0.0746, "step": 570 }, { "grad_norm": 0.38146519660949707, "learning_rate": 9.998293831085037e-05, "loss": 0.0769, "step": 580 }, { "grad_norm": 0.7051944136619568, "learning_rate": 9.997834584671719e-05, "loss": 0.0762, "step": 590 }, { "grad_norm": 0.5390310883522034, "learning_rate": 9.997320682740107e-05, "loss": 0.0778, "step": 600 }, { "grad_norm": 0.4269278049468994, "learning_rate": 9.996752130910149e-05, "loss": 0.0844, "step": 610 }, { "grad_norm": 0.48361289501190186, "learning_rate": 9.99612893539944e-05, "loss": 0.0715, "step": 620 }, { "grad_norm": 0.5045555830001831, "learning_rate": 9.995451103023144e-05, "loss": 0.0642, "step": 630 }, { "grad_norm": 0.5716514587402344, "learning_rate": 9.994718641193928e-05, "loss": 0.067, "step": 640 }, { "grad_norm": 0.6318100690841675, "learning_rate": 9.993931557921874e-05, "loss": 0.066, "step": 650 }, { "grad_norm": 0.7169504165649414, "learning_rate": 9.993089861814402e-05, "loss": 0.0723, "step": 660 }, { "grad_norm": 0.6121976971626282, "learning_rate": 9.992193562076166e-05, "loss": 0.0824, "step": 670 }, { "grad_norm": 0.5035982728004456, "learning_rate": 9.991242668508954e-05, "loss": 0.0759, "step": 680 }, { "grad_norm": 0.5040209293365479, "learning_rate": 9.990237191511587e-05, "loss": 0.0657, "step": 690 }, { "grad_norm": 0.5007683634757996, "learning_rate": 9.989177142079802e-05, "loss": 0.0709, "step": 700 }, { "grad_norm": 0.45832380652427673, "learning_rate": 9.988062531806126e-05, "loss": 0.0778, "step": 710 }, { "grad_norm": 0.48925894498825073, "learning_rate": 9.986893372879762e-05, "loss": 0.0688, "step": 720 }, { "grad_norm": 0.3998444676399231, "learning_rate": 9.985669678086443e-05, "loss": 0.0673, "step": 730 }, { "grad_norm": 0.44561105966567993, "learning_rate": 9.984391460808298e-05, "loss": 0.0601, "step": 740 }, { "grad_norm": 0.4195992052555084, "learning_rate": 9.983058735023709e-05, "loss": 0.0618, "step": 750 }, { "grad_norm": 0.490553081035614, "learning_rate": 9.98167151530715e-05, "loss": 0.0613, "step": 760 }, { "grad_norm": 0.5012132525444031, "learning_rate": 9.980229816829034e-05, "loss": 0.0598, "step": 770 }, { "grad_norm": 0.5100612640380859, "learning_rate": 9.978733655355544e-05, "loss": 0.066, "step": 780 }, { "grad_norm": 0.47101524472236633, "learning_rate": 9.977183047248464e-05, "loss": 0.0683, "step": 790 }, { "grad_norm": 0.5815221071243286, "learning_rate": 9.975578009464992e-05, "loss": 0.0642, "step": 800 }, { "grad_norm": 0.33351296186447144, "learning_rate": 9.97391855955757e-05, "loss": 0.0627, "step": 810 }, { "grad_norm": 0.4773010015487671, "learning_rate": 9.972204715673669e-05, "loss": 0.0597, "step": 820 }, { "grad_norm": 0.6065933108329773, "learning_rate": 9.970436496555617e-05, "loss": 0.0656, "step": 830 }, { "grad_norm": 0.4566664397716522, "learning_rate": 9.968613921540373e-05, "loss": 0.0629, "step": 840 }, { "grad_norm": 0.36297500133514404, "learning_rate": 9.966737010559326e-05, "loss": 0.0682, "step": 850 }, { "grad_norm": 0.4747133255004883, "learning_rate": 9.964805784138072e-05, "loss": 0.0644, "step": 860 }, { "grad_norm": 0.3558894395828247, "learning_rate": 9.962820263396195e-05, "loss": 0.0562, "step": 870 }, { "grad_norm": 0.4547067880630493, "learning_rate": 9.960780470047033e-05, "loss": 0.0543, "step": 880 }, { "grad_norm": 0.36591267585754395, "learning_rate": 9.958686426397437e-05, "loss": 0.053, "step": 890 }, { "grad_norm": 0.5208208560943604, "learning_rate": 9.956538155347534e-05, "loss": 0.0638, "step": 900 }, { "grad_norm": 0.5751354098320007, "learning_rate": 9.95433568039047e-05, "loss": 0.0626, "step": 910 }, { "grad_norm": 0.33825773000717163, "learning_rate": 9.952079025612162e-05, "loss": 0.0617, "step": 920 }, { "grad_norm": 0.4860726296901703, "learning_rate": 9.949768215691022e-05, "loss": 0.058, "step": 930 }, { "grad_norm": 0.5611844062805176, "learning_rate": 9.9474032758977e-05, "loss": 0.0587, "step": 940 }, { "grad_norm": 0.5879578590393066, "learning_rate": 9.944984232094794e-05, "loss": 0.0608, "step": 950 }, { "grad_norm": 0.486156165599823, "learning_rate": 9.942511110736584e-05, "loss": 0.0544, "step": 960 }, { "grad_norm": 0.4980018436908722, "learning_rate": 9.939983938868726e-05, "loss": 0.0641, "step": 970 }, { "grad_norm": 0.43133726716041565, "learning_rate": 9.93740274412797e-05, "loss": 0.063, "step": 980 }, { "grad_norm": 0.31840652227401733, "learning_rate": 9.934767554741846e-05, "loss": 0.0597, "step": 990 }, { "grad_norm": 0.46251097321510315, "learning_rate": 9.932078399528361e-05, "loss": 0.057, "step": 1000 }, { "grad_norm": 0.4224526584148407, "learning_rate": 9.929335307895689e-05, "loss": 0.0526, "step": 1010 }, { "grad_norm": 0.4052707850933075, "learning_rate": 9.926538309841839e-05, "loss": 0.0563, "step": 1020 }, { "grad_norm": 0.5739420056343079, "learning_rate": 9.923687435954334e-05, "loss": 0.0574, "step": 1030 }, { "grad_norm": 0.4307857155799866, "learning_rate": 9.920782717409873e-05, "loss": 0.0579, "step": 1040 }, { "grad_norm": 0.407103568315506, "learning_rate": 9.917824185973994e-05, "loss": 0.05, "step": 1050 }, { "grad_norm": 0.4271583557128906, "learning_rate": 9.914811874000723e-05, "loss": 0.0595, "step": 1060 }, { "grad_norm": 0.3994213044643402, "learning_rate": 9.911745814432218e-05, "loss": 0.0566, "step": 1070 }, { "grad_norm": 0.32001420855522156, "learning_rate": 9.90862604079842e-05, "loss": 0.0555, "step": 1080 }, { "grad_norm": 0.3987141251564026, "learning_rate": 9.90545258721667e-05, "loss": 0.0432, "step": 1090 }, { "grad_norm": 0.5398167967796326, "learning_rate": 9.90222548839135e-05, "loss": 0.0533, "step": 1100 }, { "grad_norm": 0.325234979391098, "learning_rate": 9.898944779613495e-05, "loss": 0.0561, "step": 1110 }, { "grad_norm": 0.4157761335372925, "learning_rate": 9.89561049676041e-05, "loss": 0.0515, "step": 1120 }, { "grad_norm": 0.4650302231311798, "learning_rate": 9.89222267629528e-05, "loss": 0.0617, "step": 1130 }, { "grad_norm": 0.4272788166999817, "learning_rate": 9.888781355266763e-05, "loss": 0.0509, "step": 1140 }, { "grad_norm": 0.3848171830177307, "learning_rate": 9.885286571308598e-05, "loss": 0.0535, "step": 1150 }, { "grad_norm": 0.5255935788154602, "learning_rate": 9.881738362639182e-05, "loss": 0.0573, "step": 1160 }, { "grad_norm": 0.5066208839416504, "learning_rate": 9.878136768061154e-05, "loss": 0.051, "step": 1170 }, { "grad_norm": 0.2994983196258545, "learning_rate": 9.874481826960979e-05, "loss": 0.0554, "step": 1180 }, { "grad_norm": 0.3597926199436188, "learning_rate": 9.870773579308503e-05, "loss": 0.0535, "step": 1190 }, { "grad_norm": 0.6144452095031738, "learning_rate": 9.867012065656533e-05, "loss": 0.0581, "step": 1200 }, { "grad_norm": 0.3906228840351105, "learning_rate": 9.863197327140376e-05, "loss": 0.0528, "step": 1210 }, { "grad_norm": 0.43274977803230286, "learning_rate": 9.859329405477403e-05, "loss": 0.057, "step": 1220 }, { "grad_norm": 0.32804086804389954, "learning_rate": 9.855408342966585e-05, "loss": 0.063, "step": 1230 }, { "grad_norm": 0.45003542304039, "learning_rate": 9.851434182488033e-05, "loss": 0.0578, "step": 1240 }, { "grad_norm": 0.394601047039032, "learning_rate": 9.84740696750253e-05, "loss": 0.0596, "step": 1250 }, { "grad_norm": 0.40644606947898865, "learning_rate": 9.843326742051055e-05, "loss": 0.0507, "step": 1260 }, { "grad_norm": 0.383413165807724, "learning_rate": 9.839193550754297e-05, "loss": 0.0582, "step": 1270 }, { "grad_norm": 0.4818800687789917, "learning_rate": 9.835007438812177e-05, "loss": 0.0553, "step": 1280 }, { "grad_norm": 0.5241859555244446, "learning_rate": 9.830768452003341e-05, "loss": 0.0626, "step": 1290 }, { "grad_norm": 0.3347008526325226, "learning_rate": 9.826476636684671e-05, "loss": 0.0562, "step": 1300 }, { "grad_norm": 0.4377828538417816, "learning_rate": 9.822132039790773e-05, "loss": 0.0525, "step": 1310 }, { "grad_norm": 0.41226622462272644, "learning_rate": 9.817734708833461e-05, "loss": 0.0497, "step": 1320 }, { "grad_norm": 0.35821035504341125, "learning_rate": 9.813284691901243e-05, "loss": 0.0495, "step": 1330 }, { "grad_norm": 0.3222714364528656, "learning_rate": 9.808782037658792e-05, "loss": 0.0504, "step": 1340 }, { "grad_norm": 0.3628571927547455, "learning_rate": 9.804226795346411e-05, "loss": 0.0418, "step": 1350 }, { "grad_norm": 0.5117912292480469, "learning_rate": 9.799619014779503e-05, "loss": 0.0497, "step": 1360 }, { "grad_norm": 0.25170013308525085, "learning_rate": 9.794958746348013e-05, "loss": 0.0495, "step": 1370 }, { "grad_norm": 0.33992668986320496, "learning_rate": 9.790246041015896e-05, "loss": 0.0508, "step": 1380 }, { "grad_norm": 0.2997431755065918, "learning_rate": 9.785480950320538e-05, "loss": 0.0449, "step": 1390 }, { "grad_norm": 0.5067160725593567, "learning_rate": 9.78066352637221e-05, "loss": 0.0469, "step": 1400 }, { "grad_norm": 0.30482926964759827, "learning_rate": 9.775793821853488e-05, "loss": 0.0453, "step": 1410 }, { "grad_norm": 0.3082956075668335, "learning_rate": 9.77087189001868e-05, "loss": 0.048, "step": 1420 }, { "grad_norm": 0.4568119943141937, "learning_rate": 9.765897784693243e-05, "loss": 0.0502, "step": 1430 }, { "grad_norm": 0.28716057538986206, "learning_rate": 9.760871560273197e-05, "loss": 0.0549, "step": 1440 }, { "grad_norm": 0.35360240936279297, "learning_rate": 9.755793271724526e-05, "loss": 0.0595, "step": 1450 }, { "grad_norm": 0.2708582580089569, "learning_rate": 9.750662974582584e-05, "loss": 0.049, "step": 1460 }, { "grad_norm": 0.4497932493686676, "learning_rate": 9.745480724951473e-05, "loss": 0.0519, "step": 1470 }, { "grad_norm": 0.40248093008995056, "learning_rate": 9.740246579503447e-05, "loss": 0.0545, "step": 1480 }, { "grad_norm": 0.326463520526886, "learning_rate": 9.734960595478284e-05, "loss": 0.0518, "step": 1490 }, { "grad_norm": 0.46301496028900146, "learning_rate": 9.729622830682657e-05, "loss": 0.0542, "step": 1500 }, { "grad_norm": 0.39056462049484253, "learning_rate": 9.724233343489504e-05, "loss": 0.0497, "step": 1510 }, { "grad_norm": 0.30494236946105957, "learning_rate": 9.718792192837396e-05, "loss": 0.0465, "step": 1520 }, { "grad_norm": 0.2788922190666199, "learning_rate": 9.713299438229886e-05, "loss": 0.0452, "step": 1530 }, { "grad_norm": 0.4535343050956726, "learning_rate": 9.707755139734855e-05, "loss": 0.0509, "step": 1540 }, { "grad_norm": 0.410661906003952, "learning_rate": 9.702159357983866e-05, "loss": 0.0573, "step": 1550 }, { "grad_norm": 0.32712242007255554, "learning_rate": 9.696512154171492e-05, "loss": 0.0485, "step": 1560 }, { "grad_norm": 0.2284832000732422, "learning_rate": 9.690813590054645e-05, "loss": 0.0423, "step": 1570 }, { "grad_norm": 0.37343117594718933, "learning_rate": 9.685063727951914e-05, "loss": 0.0482, "step": 1580 }, { "grad_norm": 0.3524080216884613, "learning_rate": 9.679262630742865e-05, "loss": 0.0548, "step": 1590 }, { "grad_norm": 0.49708884954452515, "learning_rate": 9.673410361867373e-05, "loss": 0.0481, "step": 1600 }, { "grad_norm": 0.3279217779636383, "learning_rate": 9.667506985324909e-05, "loss": 0.0494, "step": 1610 }, { "grad_norm": 0.2762748897075653, "learning_rate": 9.661552565673855e-05, "loss": 0.0471, "step": 1620 }, { "grad_norm": 0.4197017252445221, "learning_rate": 9.655547168030789e-05, "loss": 0.0482, "step": 1630 }, { "grad_norm": 0.38950929045677185, "learning_rate": 9.649490858069777e-05, "loss": 0.0485, "step": 1640 }, { "grad_norm": 0.37052738666534424, "learning_rate": 9.643383702021658e-05, "loss": 0.0483, "step": 1650 }, { "grad_norm": 0.3302462100982666, "learning_rate": 9.637225766673307e-05, "loss": 0.0444, "step": 1660 }, { "grad_norm": 0.35473397374153137, "learning_rate": 9.631017119366922e-05, "loss": 0.0543, "step": 1670 }, { "grad_norm": 0.23816119134426117, "learning_rate": 9.624757827999273e-05, "loss": 0.0546, "step": 1680 }, { "grad_norm": 0.4294321537017822, "learning_rate": 9.618447961020971e-05, "loss": 0.0521, "step": 1690 }, { "grad_norm": 0.45564281940460205, "learning_rate": 9.612087587435707e-05, "loss": 0.0501, "step": 1700 }, { "grad_norm": 0.33394744992256165, "learning_rate": 9.605676776799508e-05, "loss": 0.0467, "step": 1710 }, { "grad_norm": 0.31159186363220215, "learning_rate": 9.599215599219973e-05, "loss": 0.0482, "step": 1720 }, { "grad_norm": 0.4700748026371002, "learning_rate": 9.592704125355505e-05, "loss": 0.0527, "step": 1730 }, { "grad_norm": 0.33702802658081055, "learning_rate": 9.586142426414538e-05, "loss": 0.0441, "step": 1740 }, { "grad_norm": 0.3642789423465729, "learning_rate": 9.57953057415476e-05, "loss": 0.0474, "step": 1750 }, { "grad_norm": 0.2561584711074829, "learning_rate": 9.572868640882328e-05, "loss": 0.0463, "step": 1760 }, { "grad_norm": 0.2665463089942932, "learning_rate": 9.56615669945108e-05, "loss": 0.0495, "step": 1770 }, { "grad_norm": 0.2568252682685852, "learning_rate": 9.55939482326173e-05, "loss": 0.0448, "step": 1780 }, { "grad_norm": 0.3730902373790741, "learning_rate": 9.552583086261069e-05, "loss": 0.0521, "step": 1790 }, { "grad_norm": 0.4158942997455597, "learning_rate": 9.545721562941168e-05, "loss": 0.045, "step": 1800 }, { "grad_norm": 0.31290382146835327, "learning_rate": 9.538810328338543e-05, "loss": 0.0428, "step": 1810 }, { "grad_norm": 0.34673401713371277, "learning_rate": 9.531849458033349e-05, "loss": 0.0492, "step": 1820 }, { "grad_norm": 0.32160207629203796, "learning_rate": 9.524839028148547e-05, "loss": 0.0484, "step": 1830 }, { "grad_norm": 0.3650602102279663, "learning_rate": 9.517779115349077e-05, "loss": 0.0524, "step": 1840 }, { "grad_norm": 0.35613611340522766, "learning_rate": 9.510669796841014e-05, "loss": 0.0514, "step": 1850 }, { "grad_norm": 0.24479940533638, "learning_rate": 9.503511150370727e-05, "loss": 0.0491, "step": 1860 }, { "grad_norm": 0.3608837425708771, "learning_rate": 9.496303254224024e-05, "loss": 0.0454, "step": 1870 }, { "grad_norm": 0.3112425208091736, "learning_rate": 9.489046187225306e-05, "loss": 0.0455, "step": 1880 }, { "grad_norm": 0.331745982170105, "learning_rate": 9.481740028736692e-05, "loss": 0.0463, "step": 1890 }, { "grad_norm": 0.27865085005760193, "learning_rate": 9.474384858657164e-05, "loss": 0.0538, "step": 1900 }, { "grad_norm": 0.26767468452453613, "learning_rate": 9.466980757421679e-05, "loss": 0.0477, "step": 1910 }, { "grad_norm": 0.25685328245162964, "learning_rate": 9.459527806000305e-05, "loss": 0.0471, "step": 1920 }, { "grad_norm": 0.47773152589797974, "learning_rate": 9.452026085897325e-05, "loss": 0.0501, "step": 1930 }, { "grad_norm": 0.27584904432296753, "learning_rate": 9.444475679150348e-05, "loss": 0.0442, "step": 1940 }, { "grad_norm": 0.4923325479030609, "learning_rate": 9.436876668329411e-05, "loss": 0.0492, "step": 1950 }, { "grad_norm": 0.32955610752105713, "learning_rate": 9.429229136536079e-05, "loss": 0.0449, "step": 1960 }, { "grad_norm": 0.1964392513036728, "learning_rate": 9.421533167402534e-05, "loss": 0.0448, "step": 1970 }, { "grad_norm": 0.37957942485809326, "learning_rate": 9.413788845090666e-05, "loss": 0.0556, "step": 1980 }, { "grad_norm": 0.3430396318435669, "learning_rate": 9.405996254291136e-05, "loss": 0.0462, "step": 1990 }, { "grad_norm": 0.23329870402812958, "learning_rate": 9.398155480222474e-05, "loss": 0.0469, "step": 2000 }, { "grad_norm": 0.24158401787281036, "learning_rate": 9.390266608630128e-05, "loss": 0.0421, "step": 2010 }, { "grad_norm": 0.3294133245944977, "learning_rate": 9.38232972578553e-05, "loss": 0.0442, "step": 2020 }, { "grad_norm": 0.24810637533664703, "learning_rate": 9.374344918485164e-05, "loss": 0.0486, "step": 2030 }, { "grad_norm": 0.28352993726730347, "learning_rate": 9.366312274049602e-05, "loss": 0.0497, "step": 2040 }, { "grad_norm": 0.3007359206676483, "learning_rate": 9.358231880322554e-05, "loss": 0.0418, "step": 2050 }, { "grad_norm": 0.255808025598526, "learning_rate": 9.350103825669916e-05, "loss": 0.0373, "step": 2060 }, { "grad_norm": 0.30319085717201233, "learning_rate": 9.341928198978787e-05, "loss": 0.0442, "step": 2070 }, { "grad_norm": 0.27748292684555054, "learning_rate": 9.333705089656512e-05, "loss": 0.0365, "step": 2080 }, { "grad_norm": 0.20545822381973267, "learning_rate": 9.325434587629698e-05, "loss": 0.0403, "step": 2090 }, { "grad_norm": 0.3240455389022827, "learning_rate": 9.31711678334323e-05, "loss": 0.0369, "step": 2100 }, { "grad_norm": 0.296354204416275, "learning_rate": 9.308751767759282e-05, "loss": 0.0402, "step": 2110 }, { "grad_norm": 0.2712908685207367, "learning_rate": 9.300339632356325e-05, "loss": 0.0407, "step": 2120 }, { "grad_norm": 0.30893176794052124, "learning_rate": 9.291880469128124e-05, "loss": 0.0412, "step": 2130 }, { "grad_norm": 0.35958197712898254, "learning_rate": 9.283374370582732e-05, "loss": 0.0408, "step": 2140 }, { "grad_norm": 0.2797650098800659, "learning_rate": 9.274821429741482e-05, "loss": 0.045, "step": 2150 }, { "grad_norm": 0.5062182545661926, "learning_rate": 9.266221740137961e-05, "loss": 0.0398, "step": 2160 }, { "grad_norm": 0.3966743052005768, "learning_rate": 9.257575395817001e-05, "loss": 0.039, "step": 2170 }, { "grad_norm": 0.33677244186401367, "learning_rate": 9.248882491333637e-05, "loss": 0.0462, "step": 2180 }, { "grad_norm": 0.32986190915107727, "learning_rate": 9.240143121752076e-05, "loss": 0.0445, "step": 2190 }, { "grad_norm": 0.24580863118171692, "learning_rate": 9.23135738264467e-05, "loss": 0.0421, "step": 2200 }, { "grad_norm": 0.30002644658088684, "learning_rate": 9.222525370090849e-05, "loss": 0.0419, "step": 2210 }, { "grad_norm": 0.3718232214450836, "learning_rate": 9.213647180676088e-05, "loss": 0.0342, "step": 2220 }, { "grad_norm": 0.3451251685619354, "learning_rate": 9.204722911490846e-05, "loss": 0.041, "step": 2230 }, { "grad_norm": 0.245164155960083, "learning_rate": 9.1957526601295e-05, "loss": 0.0387, "step": 2240 }, { "grad_norm": 0.29164502024650574, "learning_rate": 9.186736524689281e-05, "loss": 0.047, "step": 2250 }, { "grad_norm": 0.2925788462162018, "learning_rate": 9.177674603769204e-05, "loss": 0.0485, "step": 2260 }, { "grad_norm": 0.36472469568252563, "learning_rate": 9.168566996468983e-05, "loss": 0.0393, "step": 2270 }, { "grad_norm": 0.3461534380912781, "learning_rate": 9.159413802387951e-05, "loss": 0.0453, "step": 2280 }, { "grad_norm": 0.3834392726421356, "learning_rate": 9.150215121623974e-05, "loss": 0.0381, "step": 2290 }, { "grad_norm": 0.4060312807559967, "learning_rate": 9.140971054772349e-05, "loss": 0.0388, "step": 2300 }, { "grad_norm": 0.39975062012672424, "learning_rate": 9.131681702924713e-05, "loss": 0.0449, "step": 2310 }, { "grad_norm": 0.3716065585613251, "learning_rate": 9.122347167667926e-05, "loss": 0.0363, "step": 2320 }, { "grad_norm": 0.29488712549209595, "learning_rate": 9.112967551082973e-05, "loss": 0.0472, "step": 2330 }, { "grad_norm": 0.32794609665870667, "learning_rate": 9.103542955743835e-05, "loss": 0.0407, "step": 2340 }, { "grad_norm": 0.3162729740142822, "learning_rate": 9.094073484716381e-05, "loss": 0.0499, "step": 2350 }, { "grad_norm": 0.3124503195285797, "learning_rate": 9.084559241557226e-05, "loss": 0.0368, "step": 2360 }, { "grad_norm": 0.3742011487483978, "learning_rate": 9.075000330312608e-05, "loss": 0.0406, "step": 2370 }, { "grad_norm": 0.3292102515697479, "learning_rate": 9.065396855517253e-05, "loss": 0.0402, "step": 2380 }, { "grad_norm": 0.3746188282966614, "learning_rate": 9.055748922193219e-05, "loss": 0.0437, "step": 2390 }, { "grad_norm": 0.3140893578529358, "learning_rate": 9.046056635848761e-05, "loss": 0.046, "step": 2400 }, { "grad_norm": 0.4070205092430115, "learning_rate": 9.036320102477169e-05, "loss": 0.045, "step": 2410 }, { "grad_norm": 0.33836811780929565, "learning_rate": 9.02653942855561e-05, "loss": 0.0384, "step": 2420 }, { "grad_norm": 0.28187307715415955, "learning_rate": 9.016714721043971e-05, "loss": 0.0354, "step": 2430 }, { "grad_norm": 0.21064117550849915, "learning_rate": 9.006846087383675e-05, "loss": 0.0362, "step": 2440 }, { "grad_norm": 0.20335541665554047, "learning_rate": 8.996933635496523e-05, "loss": 0.0339, "step": 2450 }, { "grad_norm": 0.2884037494659424, "learning_rate": 8.986977473783498e-05, "loss": 0.037, "step": 2460 }, { "grad_norm": 0.3026350736618042, "learning_rate": 8.97697771112359e-05, "loss": 0.0347, "step": 2470 }, { "grad_norm": 0.2787318527698517, "learning_rate": 8.966934456872602e-05, "loss": 0.0373, "step": 2480 }, { "grad_norm": 0.21673502027988434, "learning_rate": 8.95684782086195e-05, "loss": 0.0366, "step": 2490 }, { "grad_norm": 0.6515851020812988, "learning_rate": 8.946717913397476e-05, "loss": 0.0417, "step": 2500 }, { "grad_norm": 0.24351195991039276, "learning_rate": 8.93654484525822e-05, "loss": 0.0401, "step": 2510 }, { "grad_norm": 0.24626921117305756, "learning_rate": 8.926328727695226e-05, "loss": 0.0383, "step": 2520 }, { "grad_norm": 0.35909584164619446, "learning_rate": 8.916069672430319e-05, "loss": 0.0365, "step": 2530 }, { "grad_norm": 0.2594301402568817, "learning_rate": 8.905767791654884e-05, "loss": 0.037, "step": 2540 }, { "grad_norm": 0.3230316936969757, "learning_rate": 8.895423198028638e-05, "loss": 0.0398, "step": 2550 }, { "grad_norm": 0.31597062945365906, "learning_rate": 8.885036004678402e-05, "loss": 0.0379, "step": 2560 }, { "grad_norm": 0.32177606225013733, "learning_rate": 8.874606325196857e-05, "loss": 0.0394, "step": 2570 }, { "grad_norm": 0.25245916843414307, "learning_rate": 8.864134273641304e-05, "loss": 0.0434, "step": 2580 }, { "grad_norm": 0.1703503280878067, "learning_rate": 8.853619964532427e-05, "loss": 0.036, "step": 2590 }, { "grad_norm": 0.2980024218559265, "learning_rate": 8.843063512853019e-05, "loss": 0.0354, "step": 2600 }, { "grad_norm": 0.2179039567708969, "learning_rate": 8.832465034046749e-05, "loss": 0.0348, "step": 2610 }, { "grad_norm": 0.3365156948566437, "learning_rate": 8.821824644016882e-05, "loss": 0.0363, "step": 2620 }, { "grad_norm": 0.3875049650669098, "learning_rate": 8.811142459125019e-05, "loss": 0.0361, "step": 2630 }, { "grad_norm": 0.2622855007648468, "learning_rate": 8.800418596189822e-05, "loss": 0.0329, "step": 2640 }, { "grad_norm": 0.45283034443855286, "learning_rate": 8.789653172485737e-05, "loss": 0.0374, "step": 2650 }, { "grad_norm": 0.600604236125946, "learning_rate": 8.778846305741715e-05, "loss": 0.0553, "step": 2660 }, { "grad_norm": 0.202572301030159, "learning_rate": 8.767998114139918e-05, "loss": 0.0432, "step": 2670 }, { "grad_norm": 0.4384573698043823, "learning_rate": 8.757108716314429e-05, "loss": 0.0451, "step": 2680 }, { "grad_norm": 0.4603714346885681, "learning_rate": 8.746178231349962e-05, "loss": 0.0438, "step": 2690 }, { "grad_norm": 0.3617015480995178, "learning_rate": 8.735206778780549e-05, "loss": 0.0426, "step": 2700 }, { "grad_norm": 0.2864076495170593, "learning_rate": 8.724194478588234e-05, "loss": 0.0405, "step": 2710 }, { "grad_norm": 0.24608391523361206, "learning_rate": 8.713141451201772e-05, "loss": 0.036, "step": 2720 }, { "grad_norm": 0.269539475440979, "learning_rate": 8.702047817495295e-05, "loss": 0.0384, "step": 2730 }, { "grad_norm": 0.2684914767742157, "learning_rate": 8.69091369878701e-05, "loss": 0.0361, "step": 2740 }, { "grad_norm": 0.2894139587879181, "learning_rate": 8.679739216837849e-05, "loss": 0.03, "step": 2750 }, { "grad_norm": 0.22977973520755768, "learning_rate": 8.66852449385016e-05, "loss": 0.04, "step": 2760 }, { "grad_norm": 0.21344155073165894, "learning_rate": 8.657269652466356e-05, "loss": 0.0365, "step": 2770 }, { "grad_norm": 0.28664329648017883, "learning_rate": 8.645974815767577e-05, "loss": 0.0359, "step": 2780 }, { "grad_norm": 0.2940286099910736, "learning_rate": 8.634640107272351e-05, "loss": 0.0376, "step": 2790 }, { "grad_norm": 0.7882370948791504, "learning_rate": 8.623265650935234e-05, "loss": 0.0406, "step": 2800 }, { "grad_norm": 0.2541744112968445, "learning_rate": 8.611851571145456e-05, "loss": 0.0384, "step": 2810 }, { "grad_norm": 0.30241698026657104, "learning_rate": 8.600397992725566e-05, "loss": 0.0373, "step": 2820 }, { "grad_norm": 0.2790631055831909, "learning_rate": 8.588905040930061e-05, "loss": 0.0363, "step": 2830 }, { "grad_norm": 0.2962036430835724, "learning_rate": 8.577372841444022e-05, "loss": 0.0409, "step": 2840 }, { "grad_norm": 0.28642210364341736, "learning_rate": 8.565801520381736e-05, "loss": 0.0405, "step": 2850 }, { "grad_norm": 0.36998873949050903, "learning_rate": 8.554191204285313e-05, "loss": 0.0385, "step": 2860 }, { "grad_norm": 0.265214204788208, "learning_rate": 8.542542020123315e-05, "loss": 0.0376, "step": 2870 }, { "grad_norm": 0.2859579622745514, "learning_rate": 8.530854095289347e-05, "loss": 0.0367, "step": 2880 }, { "grad_norm": 0.30964896082878113, "learning_rate": 8.519127557600688e-05, "loss": 0.0349, "step": 2890 }, { "grad_norm": 0.29006829857826233, "learning_rate": 8.507362535296871e-05, "loss": 0.0392, "step": 2900 }, { "grad_norm": 0.3030093014240265, "learning_rate": 8.495559157038299e-05, "loss": 0.0449, "step": 2910 }, { "grad_norm": 0.2128904163837433, "learning_rate": 8.483717551904823e-05, "loss": 0.033, "step": 2920 }, { "grad_norm": 0.3433743119239807, "learning_rate": 8.47183784939434e-05, "loss": 0.0378, "step": 2930 }, { "grad_norm": 0.27203765511512756, "learning_rate": 8.459920179421374e-05, "loss": 0.0401, "step": 2940 }, { "grad_norm": 0.4420972764492035, "learning_rate": 8.447964672315656e-05, "loss": 0.0381, "step": 2950 }, { "grad_norm": 0.3673306703567505, "learning_rate": 8.435971458820692e-05, "loss": 0.0457, "step": 2960 }, { "grad_norm": 0.25317999720573425, "learning_rate": 8.423940670092345e-05, "loss": 0.0457, "step": 2970 }, { "grad_norm": 0.3223629593849182, "learning_rate": 8.411872437697394e-05, "loss": 0.0382, "step": 2980 }, { "grad_norm": 0.2370004504919052, "learning_rate": 8.399766893612096e-05, "loss": 0.0343, "step": 2990 }, { "grad_norm": 0.29405468702316284, "learning_rate": 8.38762417022074e-05, "loss": 0.0392, "step": 3000 }, { "grad_norm": 0.4289790391921997, "learning_rate": 8.375444400314204e-05, "loss": 0.037, "step": 3010 }, { "grad_norm": 0.31298673152923584, "learning_rate": 8.3632277170885e-05, "loss": 0.0454, "step": 3020 }, { "grad_norm": 0.3962829113006592, "learning_rate": 8.350974254143318e-05, "loss": 0.0362, "step": 3030 }, { "grad_norm": 0.2606299817562103, "learning_rate": 8.338684145480566e-05, "loss": 0.0342, "step": 3040 }, { "grad_norm": 0.2971605658531189, "learning_rate": 8.326357525502904e-05, "loss": 0.0384, "step": 3050 }, { "grad_norm": 0.23222973942756653, "learning_rate": 8.313994529012273e-05, "loss": 0.0354, "step": 3060 }, { "grad_norm": 0.23696400225162506, "learning_rate": 8.301595291208422e-05, "loss": 0.0377, "step": 3070 }, { "grad_norm": 0.2377704530954361, "learning_rate": 8.289159947687427e-05, "loss": 0.0358, "step": 3080 }, { "grad_norm": 0.2908557653427124, "learning_rate": 8.276688634440216e-05, "loss": 0.0339, "step": 3090 }, { "grad_norm": 0.2976522147655487, "learning_rate": 8.26418148785107e-05, "loss": 0.0373, "step": 3100 }, { "grad_norm": 0.24224083125591278, "learning_rate": 8.251638644696141e-05, "loss": 0.0395, "step": 3110 }, { "grad_norm": 0.2737254798412323, "learning_rate": 8.23906024214195e-05, "loss": 0.0355, "step": 3120 }, { "grad_norm": 0.1535027027130127, "learning_rate": 8.226446417743897e-05, "loss": 0.0406, "step": 3130 }, { "grad_norm": 0.17368154227733612, "learning_rate": 8.213797309444742e-05, "loss": 0.0352, "step": 3140 }, { "grad_norm": 0.41940975189208984, "learning_rate": 8.201113055573105e-05, "loss": 0.0376, "step": 3150 }, { "grad_norm": 0.44374775886535645, "learning_rate": 8.188393794841958e-05, "loss": 0.0413, "step": 3160 }, { "grad_norm": 0.2870585024356842, "learning_rate": 8.175639666347094e-05, "loss": 0.0376, "step": 3170 }, { "grad_norm": 0.15355916321277618, "learning_rate": 8.162850809565623e-05, "loss": 0.0317, "step": 3180 }, { "grad_norm": 0.3080750107765198, "learning_rate": 8.150027364354431e-05, "loss": 0.0348, "step": 3190 }, { "grad_norm": 0.21497546136379242, "learning_rate": 8.137169470948662e-05, "loss": 0.0399, "step": 3200 }, { "grad_norm": 0.24774391949176788, "learning_rate": 8.124277269960179e-05, "loss": 0.0402, "step": 3210 }, { "grad_norm": 0.20315490663051605, "learning_rate": 8.111350902376023e-05, "loss": 0.0319, "step": 3220 }, { "grad_norm": 0.23530274629592896, "learning_rate": 8.098390509556883e-05, "loss": 0.031, "step": 3230 }, { "grad_norm": 0.34867796301841736, "learning_rate": 8.085396233235536e-05, "loss": 0.0291, "step": 3240 }, { "grad_norm": 0.25846192240715027, "learning_rate": 8.072368215515306e-05, "loss": 0.0309, "step": 3250 }, { "grad_norm": 0.37157535552978516, "learning_rate": 8.059306598868506e-05, "loss": 0.0393, "step": 3260 }, { "grad_norm": 0.21389614045619965, "learning_rate": 8.046211526134888e-05, "loss": 0.0344, "step": 3270 }, { "grad_norm": 0.1898447722196579, "learning_rate": 8.033083140520065e-05, "loss": 0.034, "step": 3280 }, { "grad_norm": 0.23761814832687378, "learning_rate": 8.019921585593962e-05, "loss": 0.0351, "step": 3290 }, { "grad_norm": 0.3429633677005768, "learning_rate": 8.006727005289232e-05, "loss": 0.0299, "step": 3300 }, { "grad_norm": 0.18289905786514282, "learning_rate": 7.993499543899692e-05, "loss": 0.0294, "step": 3310 }, { "grad_norm": 0.2843772768974304, "learning_rate": 7.980239346078742e-05, "loss": 0.0327, "step": 3320 }, { "grad_norm": 0.42163923382759094, "learning_rate": 7.966946556837778e-05, "loss": 0.0377, "step": 3330 }, { "grad_norm": 0.25373753905296326, "learning_rate": 7.953621321544616e-05, "loss": 0.0336, "step": 3340 }, { "grad_norm": 0.22870062291622162, "learning_rate": 7.940263785921896e-05, "loss": 0.0391, "step": 3350 }, { "grad_norm": 0.22196009755134583, "learning_rate": 7.926874096045482e-05, "loss": 0.0312, "step": 3360 }, { "grad_norm": 0.31909167766571045, "learning_rate": 7.913452398342881e-05, "loss": 0.0494, "step": 3370 }, { "grad_norm": 0.2545294463634491, "learning_rate": 7.89999883959163e-05, "loss": 0.0342, "step": 3380 }, { "grad_norm": 0.30069422721862793, "learning_rate": 7.886513566917687e-05, "loss": 0.0346, "step": 3390 }, { "grad_norm": 0.3574341833591461, "learning_rate": 7.872996727793838e-05, "loss": 0.0394, "step": 3400 }, { "grad_norm": 0.28775718808174133, "learning_rate": 7.859448470038069e-05, "loss": 0.0346, "step": 3410 }, { "grad_norm": 0.23815351724624634, "learning_rate": 7.845868941811956e-05, "loss": 0.0327, "step": 3420 }, { "grad_norm": 0.28751444816589355, "learning_rate": 7.832258291619043e-05, "loss": 0.0324, "step": 3430 }, { "grad_norm": 0.3596091866493225, "learning_rate": 7.81861666830322e-05, "loss": 0.0341, "step": 3440 }, { "grad_norm": 0.33906108140945435, "learning_rate": 7.804944221047097e-05, "loss": 0.0327, "step": 3450 }, { "grad_norm": 0.2798650562763214, "learning_rate": 7.791241099370364e-05, "loss": 0.0381, "step": 3460 }, { "grad_norm": 0.17338016629219055, "learning_rate": 7.777507453128163e-05, "loss": 0.0286, "step": 3470 }, { "grad_norm": 0.23408222198486328, "learning_rate": 7.763743432509451e-05, "loss": 0.0321, "step": 3480 }, { "grad_norm": 0.26579993963241577, "learning_rate": 7.749949188035353e-05, "loss": 0.0316, "step": 3490 }, { "grad_norm": 0.3647434413433075, "learning_rate": 7.736124870557516e-05, "loss": 0.0419, "step": 3500 }, { "grad_norm": 0.21104422211647034, "learning_rate": 7.722270631256459e-05, "loss": 0.0365, "step": 3510 }, { "grad_norm": 0.2116510272026062, "learning_rate": 7.708386621639925e-05, "loss": 0.0301, "step": 3520 }, { "grad_norm": 0.20376160740852356, "learning_rate": 7.694472993541219e-05, "loss": 0.0355, "step": 3530 }, { "grad_norm": 0.28653788566589355, "learning_rate": 7.680529899117547e-05, "loss": 0.0384, "step": 3540 }, { "grad_norm": 0.21193797886371613, "learning_rate": 7.666557490848358e-05, "loss": 0.0308, "step": 3550 }, { "grad_norm": 0.20635604858398438, "learning_rate": 7.65255592153367e-05, "loss": 0.0318, "step": 3560 }, { "grad_norm": 0.32040151953697205, "learning_rate": 7.638525344292402e-05, "loss": 0.0375, "step": 3570 }, { "grad_norm": 0.2371794581413269, "learning_rate": 7.624465912560697e-05, "loss": 0.0376, "step": 3580 }, { "grad_norm": 0.19234181940555573, "learning_rate": 7.610377780090249e-05, "loss": 0.0312, "step": 3590 }, { "grad_norm": 0.22629988193511963, "learning_rate": 7.596261100946618e-05, "loss": 0.0338, "step": 3600 }, { "grad_norm": 0.2880770266056061, "learning_rate": 7.582116029507542e-05, "loss": 0.03, "step": 3610 }, { "grad_norm": 0.2479832023382187, "learning_rate": 7.56794272046126e-05, "loss": 0.0324, "step": 3620 }, { "grad_norm": 0.24249543249607086, "learning_rate": 7.55374132880481e-05, "loss": 0.0333, "step": 3630 }, { "grad_norm": 0.22627884149551392, "learning_rate": 7.539512009842333e-05, "loss": 0.0351, "step": 3640 }, { "grad_norm": 0.2360038310289383, "learning_rate": 7.525254919183382e-05, "loss": 0.0333, "step": 3650 }, { "grad_norm": 0.3291195034980774, "learning_rate": 7.510970212741215e-05, "loss": 0.0299, "step": 3660 }, { "grad_norm": 0.38330790400505066, "learning_rate": 7.496658046731096e-05, "loss": 0.0318, "step": 3670 }, { "grad_norm": 0.25102487206459045, "learning_rate": 7.482318577668578e-05, "loss": 0.0318, "step": 3680 }, { "grad_norm": 0.24644972383975983, "learning_rate": 7.467951962367796e-05, "loss": 0.0344, "step": 3690 }, { "grad_norm": 0.2787819802761078, "learning_rate": 7.453558357939755e-05, "loss": 0.0339, "step": 3700 }, { "grad_norm": 0.31794267892837524, "learning_rate": 7.439137921790606e-05, "loss": 0.0344, "step": 3710 }, { "grad_norm": 0.20875516533851624, "learning_rate": 7.42469081161993e-05, "loss": 0.0375, "step": 3720 }, { "grad_norm": 0.2492925226688385, "learning_rate": 7.410217185419006e-05, "loss": 0.0334, "step": 3730 }, { "grad_norm": 0.2691625654697418, "learning_rate": 7.395717201469095e-05, "loss": 0.0376, "step": 3740 }, { "grad_norm": 0.3722934126853943, "learning_rate": 7.381191018339696e-05, "loss": 0.0352, "step": 3750 }, { "grad_norm": 0.21331264078617096, "learning_rate": 7.36663879488682e-05, "loss": 0.0344, "step": 3760 }, { "grad_norm": 0.32906973361968994, "learning_rate": 7.352060690251254e-05, "loss": 0.0376, "step": 3770 }, { "grad_norm": 0.3062564432621002, "learning_rate": 7.337456863856811e-05, "loss": 0.0342, "step": 3780 }, { "grad_norm": 0.3081783652305603, "learning_rate": 7.3228274754086e-05, "loss": 0.0365, "step": 3790 }, { "grad_norm": 0.21354940533638, "learning_rate": 7.308172684891267e-05, "loss": 0.0372, "step": 3800 }, { "grad_norm": 0.2509887218475342, "learning_rate": 7.293492652567255e-05, "loss": 0.0366, "step": 3810 }, { "grad_norm": 0.2332848310470581, "learning_rate": 7.278787538975043e-05, "loss": 0.0383, "step": 3820 }, { "grad_norm": 0.27271968126296997, "learning_rate": 7.2640575049274e-05, "loss": 0.0342, "step": 3830 }, { "grad_norm": 0.26460155844688416, "learning_rate": 7.249302711509616e-05, "loss": 0.0306, "step": 3840 }, { "grad_norm": 0.3441675901412964, "learning_rate": 7.23452332007775e-05, "loss": 0.0294, "step": 3850 }, { "grad_norm": 0.26211312413215637, "learning_rate": 7.219719492256858e-05, "loss": 0.032, "step": 3860 }, { "grad_norm": 0.30294981598854065, "learning_rate": 7.20489138993923e-05, "loss": 0.0327, "step": 3870 }, { "grad_norm": 0.2654699981212616, "learning_rate": 7.190039175282614e-05, "loss": 0.0311, "step": 3880 }, { "grad_norm": 0.18821047246456146, "learning_rate": 7.175163010708455e-05, "loss": 0.029, "step": 3890 }, { "grad_norm": 0.14636529982089996, "learning_rate": 7.1602630589001e-05, "loss": 0.032, "step": 3900 }, { "grad_norm": 0.2175510674715042, "learning_rate": 7.14533948280104e-05, "loss": 0.0339, "step": 3910 }, { "grad_norm": 0.19719205796718597, "learning_rate": 7.130392445613109e-05, "loss": 0.0293, "step": 3920 }, { "grad_norm": 0.28057318925857544, "learning_rate": 7.115422110794711e-05, "loss": 0.0325, "step": 3930 }, { "grad_norm": 0.26272523403167725, "learning_rate": 7.100428642059033e-05, "loss": 0.0338, "step": 3940 }, { "grad_norm": 0.24825690686702728, "learning_rate": 7.08541220337224e-05, "loss": 0.0334, "step": 3950 }, { "grad_norm": 0.21773886680603027, "learning_rate": 7.070372958951706e-05, "loss": 0.0277, "step": 3960 }, { "grad_norm": 0.23684903979301453, "learning_rate": 7.055311073264194e-05, "loss": 0.0367, "step": 3970 }, { "grad_norm": 0.3153875172138214, "learning_rate": 7.040226711024077e-05, "loss": 0.034, "step": 3980 }, { "grad_norm": 0.15356598794460297, "learning_rate": 7.02512003719152e-05, "loss": 0.0319, "step": 3990 }, { "grad_norm": 0.2538532614707947, "learning_rate": 7.00999121697069e-05, "loss": 0.0347, "step": 4000 }, { "grad_norm": 0.17953459918498993, "learning_rate": 6.99484041580794e-05, "loss": 0.0297, "step": 4010 }, { "grad_norm": 0.22616443037986755, "learning_rate": 6.979667799390004e-05, "loss": 0.0255, "step": 4020 }, { "grad_norm": 0.24502547085285187, "learning_rate": 6.964473533642185e-05, "loss": 0.0302, "step": 4030 }, { "grad_norm": 0.2551388144493103, "learning_rate": 6.949257784726539e-05, "loss": 0.0323, "step": 4040 }, { "grad_norm": 0.23204657435417175, "learning_rate": 6.934020719040056e-05, "loss": 0.0274, "step": 4050 }, { "grad_norm": 0.3056252896785736, "learning_rate": 6.918762503212848e-05, "loss": 0.0351, "step": 4060 }, { "grad_norm": 0.24859943985939026, "learning_rate": 6.903483304106319e-05, "loss": 0.0303, "step": 4070 }, { "grad_norm": 0.32507219910621643, "learning_rate": 6.888183288811341e-05, "loss": 0.0258, "step": 4080 }, { "grad_norm": 0.18772199749946594, "learning_rate": 6.87286262464643e-05, "loss": 0.0304, "step": 4090 }, { "grad_norm": 0.22234755754470825, "learning_rate": 6.857521479155915e-05, "loss": 0.0305, "step": 4100 }, { "grad_norm": 0.2716682553291321, "learning_rate": 6.842160020108104e-05, "loss": 0.0304, "step": 4110 }, { "grad_norm": 0.3107684850692749, "learning_rate": 6.826778415493455e-05, "loss": 0.0307, "step": 4120 }, { "grad_norm": 0.24197353422641754, "learning_rate": 6.811376833522729e-05, "loss": 0.0266, "step": 4130 }, { "grad_norm": 0.28156036138534546, "learning_rate": 6.795955442625159e-05, "loss": 0.0355, "step": 4140 }, { "grad_norm": 0.3198966681957245, "learning_rate": 6.780514411446608e-05, "loss": 0.032, "step": 4150 }, { "grad_norm": 0.2845383584499359, "learning_rate": 6.765053908847716e-05, "loss": 0.0283, "step": 4160 }, { "grad_norm": 0.22354809939861298, "learning_rate": 6.749574103902064e-05, "loss": 0.0317, "step": 4170 }, { "grad_norm": 0.18603889644145966, "learning_rate": 6.734075165894317e-05, "loss": 0.0372, "step": 4180 }, { "grad_norm": 0.26540815830230713, "learning_rate": 6.71855726431838e-05, "loss": 0.0384, "step": 4190 }, { "grad_norm": 0.2067757099866867, "learning_rate": 6.703020568875538e-05, "loss": 0.0283, "step": 4200 }, { "grad_norm": 0.26586946845054626, "learning_rate": 6.687465249472603e-05, "loss": 0.0257, "step": 4210 }, { "grad_norm": 0.30017343163490295, "learning_rate": 6.671891476220055e-05, "loss": 0.0284, "step": 4220 }, { "grad_norm": 0.2723040282726288, "learning_rate": 6.656299419430183e-05, "loss": 0.0261, "step": 4230 }, { "grad_norm": 0.15793976187705994, "learning_rate": 6.640689249615223e-05, "loss": 0.0311, "step": 4240 }, { "grad_norm": 0.21047021448612213, "learning_rate": 6.625061137485491e-05, "loss": 0.0287, "step": 4250 }, { "grad_norm": 0.2234886735677719, "learning_rate": 6.609415253947517e-05, "loss": 0.033, "step": 4260 }, { "grad_norm": 0.23535245656967163, "learning_rate": 6.593751770102178e-05, "loss": 0.0292, "step": 4270 }, { "grad_norm": 0.2390592396259308, "learning_rate": 6.578070857242823e-05, "loss": 0.0293, "step": 4280 }, { "grad_norm": 0.2743031084537506, "learning_rate": 6.562372686853402e-05, "loss": 0.032, "step": 4290 }, { "grad_norm": 0.2415972501039505, "learning_rate": 6.546657430606593e-05, "loss": 0.0321, "step": 4300 }, { "grad_norm": 0.16081291437149048, "learning_rate": 6.530925260361918e-05, "loss": 0.027, "step": 4310 }, { "grad_norm": 0.274310439825058, "learning_rate": 6.515176348163871e-05, "loss": 0.0308, "step": 4320 }, { "grad_norm": 0.2042841911315918, "learning_rate": 6.499410866240032e-05, "loss": 0.0299, "step": 4330 }, { "grad_norm": 0.26616671681404114, "learning_rate": 6.48362898699919e-05, "loss": 0.0317, "step": 4340 }, { "grad_norm": 0.22311647236347198, "learning_rate": 6.467830883029443e-05, "loss": 0.0291, "step": 4350 }, { "grad_norm": 0.3545197546482086, "learning_rate": 6.452016727096326e-05, "loss": 0.0317, "step": 4360 }, { "grad_norm": 0.2211519330739975, "learning_rate": 6.436186692140916e-05, "loss": 0.0284, "step": 4370 }, { "grad_norm": 0.19756735861301422, "learning_rate": 6.420340951277938e-05, "loss": 0.0281, "step": 4380 }, { "grad_norm": 0.29953208565711975, "learning_rate": 6.404479677793874e-05, "loss": 0.0328, "step": 4390 }, { "grad_norm": 0.22879450023174286, "learning_rate": 6.388603045145075e-05, "loss": 0.0366, "step": 4400 }, { "grad_norm": 0.2704363465309143, "learning_rate": 6.372711226955843e-05, "loss": 0.0315, "step": 4410 }, { "grad_norm": 0.30063214898109436, "learning_rate": 6.356804397016564e-05, "loss": 0.0307, "step": 4420 }, { "grad_norm": 0.1842574030160904, "learning_rate": 6.340882729281779e-05, "loss": 0.03, "step": 4430 }, { "grad_norm": 0.22554689645767212, "learning_rate": 6.324946397868294e-05, "loss": 0.0334, "step": 4440 }, { "grad_norm": 0.2885400950908661, "learning_rate": 6.308995577053276e-05, "loss": 0.0262, "step": 4450 }, { "grad_norm": 0.18989217281341553, "learning_rate": 6.293030441272347e-05, "loss": 0.0298, "step": 4460 }, { "grad_norm": 0.24906837940216064, "learning_rate": 6.277051165117677e-05, "loss": 0.0255, "step": 4470 }, { "grad_norm": 0.25435349345207214, "learning_rate": 6.261057923336064e-05, "loss": 0.0298, "step": 4480 }, { "grad_norm": 0.3191067576408386, "learning_rate": 6.245050890827042e-05, "loss": 0.0334, "step": 4490 }, { "grad_norm": 0.19303469359874725, "learning_rate": 6.229030242640952e-05, "loss": 0.0293, "step": 4500 }, { "grad_norm": 0.15651026368141174, "learning_rate": 6.212996153977037e-05, "loss": 0.0301, "step": 4510 }, { "grad_norm": 0.14190660417079926, "learning_rate": 6.196948800181523e-05, "loss": 0.0257, "step": 4520 }, { "grad_norm": 0.18370695412158966, "learning_rate": 6.180888356745695e-05, "loss": 0.0271, "step": 4530 }, { "grad_norm": 0.25455665588378906, "learning_rate": 6.164814999303995e-05, "loss": 0.0278, "step": 4540 }, { "grad_norm": 0.2657298445701599, "learning_rate": 6.148728903632081e-05, "loss": 0.0242, "step": 4550 }, { "grad_norm": 0.2646745443344116, "learning_rate": 6.132630245644921e-05, "loss": 0.0239, "step": 4560 }, { "grad_norm": 0.23278462886810303, "learning_rate": 6.116519201394857e-05, "loss": 0.0341, "step": 4570 }, { "grad_norm": 0.16402935981750488, "learning_rate": 6.10039594706969e-05, "loss": 0.028, "step": 4580 }, { "grad_norm": 0.1654949188232422, "learning_rate": 6.084260658990744e-05, "loss": 0.0291, "step": 4590 }, { "grad_norm": 0.26383063197135925, "learning_rate": 6.068113513610943e-05, "loss": 0.0282, "step": 4600 }, { "grad_norm": 0.1793452650308609, "learning_rate": 6.0519546875128876e-05, "loss": 0.0259, "step": 4610 }, { "grad_norm": 0.1650751233100891, "learning_rate": 6.035784357406906e-05, "loss": 0.0274, "step": 4620 }, { "grad_norm": 0.1727399230003357, "learning_rate": 6.01960270012914e-05, "loss": 0.0277, "step": 4630 }, { "grad_norm": 0.27257946133613586, "learning_rate": 6.003409892639599e-05, "loss": 0.0262, "step": 4640 }, { "grad_norm": 0.38527849316596985, "learning_rate": 5.9872061120202336e-05, "loss": 0.0308, "step": 4650 }, { "grad_norm": 0.26593899726867676, "learning_rate": 5.9709915354729914e-05, "loss": 0.0427, "step": 4660 }, { "grad_norm": 0.2331356704235077, "learning_rate": 5.9547663403178824e-05, "loss": 0.0378, "step": 4670 }, { "grad_norm": 0.18385440111160278, "learning_rate": 5.9385307039910445e-05, "loss": 0.0261, "step": 4680 }, { "grad_norm": 0.210539311170578, "learning_rate": 5.922284804042792e-05, "loss": 0.0341, "step": 4690 }, { "grad_norm": 0.32465529441833496, "learning_rate": 5.906028818135687e-05, "loss": 0.0313, "step": 4700 }, { "grad_norm": 0.19285862147808075, "learning_rate": 5.889762924042585e-05, "loss": 0.029, "step": 4710 }, { "grad_norm": 0.17221784591674805, "learning_rate": 5.873487299644699e-05, "loss": 0.0268, "step": 4720 }, { "grad_norm": 0.20987318456172943, "learning_rate": 5.857202122929649e-05, "loss": 0.0287, "step": 4730 }, { "grad_norm": 0.22214433550834656, "learning_rate": 5.840907571989518e-05, "loss": 0.0245, "step": 4740 }, { "grad_norm": 0.3116675019264221, "learning_rate": 5.824603825018904e-05, "loss": 0.0289, "step": 4750 }, { "grad_norm": 0.13698796927928925, "learning_rate": 5.808291060312975e-05, "loss": 0.0258, "step": 4760 }, { "grad_norm": 0.25833991169929504, "learning_rate": 5.7919694562655083e-05, "loss": 0.0258, "step": 4770 }, { "grad_norm": 0.23839454352855682, "learning_rate": 5.775639191366954e-05, "loss": 0.031, "step": 4780 }, { "grad_norm": 0.23593635857105255, "learning_rate": 5.75930044420247e-05, "loss": 0.0245, "step": 4790 }, { "grad_norm": 0.22100837528705597, "learning_rate": 5.74295339344998e-05, "loss": 0.0305, "step": 4800 }, { "grad_norm": 0.21723754703998566, "learning_rate": 5.726598217878211e-05, "loss": 0.0269, "step": 4810 }, { "grad_norm": 0.20138925313949585, "learning_rate": 5.71023509634474e-05, "loss": 0.0251, "step": 4820 }, { "grad_norm": 0.19356589019298553, "learning_rate": 5.693864207794049e-05, "loss": 0.0282, "step": 4830 }, { "grad_norm": 0.17575696110725403, "learning_rate": 5.677485731255545e-05, "loss": 0.0344, "step": 4840 }, { "grad_norm": 0.2050069123506546, "learning_rate": 5.6610998458416296e-05, "loss": 0.0265, "step": 4850 }, { "grad_norm": 0.22531214356422424, "learning_rate": 5.644706730745716e-05, "loss": 0.026, "step": 4860 }, { "grad_norm": 0.18739935755729675, "learning_rate": 5.628306565240287e-05, "loss": 0.0254, "step": 4870 }, { "grad_norm": 0.2051858752965927, "learning_rate": 5.611899528674923e-05, "loss": 0.0276, "step": 4880 }, { "grad_norm": 0.30509981513023376, "learning_rate": 5.595485800474349e-05, "loss": 0.0254, "step": 4890 }, { "grad_norm": 0.2612735629081726, "learning_rate": 5.579065560136467e-05, "loss": 0.0293, "step": 4900 }, { "grad_norm": 0.27787521481513977, "learning_rate": 5.562638987230392e-05, "loss": 0.0286, "step": 4910 }, { "grad_norm": 0.17541448771953583, "learning_rate": 5.546206261394498e-05, "loss": 0.0233, "step": 4920 }, { "grad_norm": 0.19974945485591888, "learning_rate": 5.529767562334437e-05, "loss": 0.0267, "step": 4930 }, { "grad_norm": 0.22405220568180084, "learning_rate": 5.5133230698211926e-05, "loss": 0.0263, "step": 4940 }, { "grad_norm": 0.17551493644714355, "learning_rate": 5.496872963689096e-05, "loss": 0.0298, "step": 4950 }, { "grad_norm": 0.179105743765831, "learning_rate": 5.4804174238338756e-05, "loss": 0.0229, "step": 4960 }, { "grad_norm": 0.18640023469924927, "learning_rate": 5.463956630210678e-05, "loss": 0.0312, "step": 4970 }, { "grad_norm": 0.18185877799987793, "learning_rate": 5.4474907628321046e-05, "loss": 0.0316, "step": 4980 }, { "grad_norm": 0.233022540807724, "learning_rate": 5.431020001766244e-05, "loss": 0.026, "step": 4990 }, { "grad_norm": 0.19441919028759003, "learning_rate": 5.4145445271346986e-05, "loss": 0.0292, "step": 5000 }, { "grad_norm": 0.15909437835216522, "learning_rate": 5.398064519110622e-05, "loss": 0.027, "step": 5010 }, { "grad_norm": 0.29164817929267883, "learning_rate": 5.3815801579167394e-05, "loss": 0.0258, "step": 5020 }, { "grad_norm": 0.25241750478744507, "learning_rate": 5.365091623823382e-05, "loss": 0.0263, "step": 5030 }, { "grad_norm": 0.2550692558288574, "learning_rate": 5.348599097146521e-05, "loss": 0.0258, "step": 5040 }, { "grad_norm": 0.26224735379219055, "learning_rate": 5.3321027582457836e-05, "loss": 0.0248, "step": 5050 }, { "grad_norm": 0.20547939836978912, "learning_rate": 5.315602787522491e-05, "loss": 0.0265, "step": 5060 }, { "grad_norm": 0.2606329619884491, "learning_rate": 5.299099365417678e-05, "loss": 0.0201, "step": 5070 }, { "grad_norm": 0.19851054251194, "learning_rate": 5.2825926724101236e-05, "loss": 0.0265, "step": 5080 }, { "grad_norm": 0.15555155277252197, "learning_rate": 5.26608288901438e-05, "loss": 0.0242, "step": 5090 }, { "grad_norm": 0.23422931134700775, "learning_rate": 5.24957019577879e-05, "loss": 0.0287, "step": 5100 }, { "grad_norm": 0.17410996556282043, "learning_rate": 5.2330547732835266e-05, "loss": 0.0263, "step": 5110 }, { "grad_norm": 0.16141533851623535, "learning_rate": 5.2165368021385996e-05, "loss": 0.0306, "step": 5120 }, { "grad_norm": 0.2299593985080719, "learning_rate": 5.200016462981897e-05, "loss": 0.0215, "step": 5130 }, { "grad_norm": 0.1776454746723175, "learning_rate": 5.1834939364772015e-05, "loss": 0.0243, "step": 5140 }, { "grad_norm": 0.23985803127288818, "learning_rate": 5.166969403312214e-05, "loss": 0.0262, "step": 5150 }, { "grad_norm": 0.1949138343334198, "learning_rate": 5.1504430441965844e-05, "loss": 0.0292, "step": 5160 }, { "grad_norm": 0.22921544313430786, "learning_rate": 5.133915039859923e-05, "loss": 0.0245, "step": 5170 }, { "grad_norm": 0.23496897518634796, "learning_rate": 5.1173855710498444e-05, "loss": 0.0245, "step": 5180 }, { "grad_norm": 0.22369922697544098, "learning_rate": 5.100854818529967e-05, "loss": 0.0307, "step": 5190 }, { "grad_norm": 0.3506659269332886, "learning_rate": 5.084322963077951e-05, "loss": 0.03, "step": 5200 }, { "grad_norm": 0.1961509734392166, "learning_rate": 5.067790185483522e-05, "loss": 0.0266, "step": 5210 }, { "grad_norm": 0.16531285643577576, "learning_rate": 5.0512566665464844e-05, "loss": 0.0247, "step": 5220 }, { "grad_norm": 0.19502408802509308, "learning_rate": 5.034722587074755e-05, "loss": 0.0247, "step": 5230 }, { "grad_norm": 0.12796403467655182, "learning_rate": 5.018188127882375e-05, "loss": 0.0225, "step": 5240 }, { "grad_norm": 0.1667669415473938, "learning_rate": 5.0016534697875417e-05, "loss": 0.0238, "step": 5250 }, { "grad_norm": 0.17065398395061493, "learning_rate": 4.9851187936106294e-05, "loss": 0.0244, "step": 5260 }, { "grad_norm": 0.23487922549247742, "learning_rate": 4.968584280172206e-05, "loss": 0.0302, "step": 5270 }, { "grad_norm": 0.2468186467885971, "learning_rate": 4.95205011029106e-05, "loss": 0.0264, "step": 5280 }, { "grad_norm": 0.27279043197631836, "learning_rate": 4.935516464782227e-05, "loss": 0.0245, "step": 5290 }, { "grad_norm": 0.1981157511472702, "learning_rate": 4.918983524455003e-05, "loss": 0.0273, "step": 5300 }, { "grad_norm": 0.17381031811237335, "learning_rate": 4.9024514701109766e-05, "loss": 0.0283, "step": 5310 }, { "grad_norm": 0.1580193042755127, "learning_rate": 4.885920482542043e-05, "loss": 0.0264, "step": 5320 }, { "grad_norm": 0.2992728650569916, "learning_rate": 4.869390742528438e-05, "loss": 0.0252, "step": 5330 }, { "grad_norm": 0.23415395617485046, "learning_rate": 4.852862430836744e-05, "loss": 0.0284, "step": 5340 }, { "grad_norm": 0.22398489713668823, "learning_rate": 4.836335728217933e-05, "loss": 0.0304, "step": 5350 }, { "grad_norm": 0.27576330304145813, "learning_rate": 4.819810815405379e-05, "loss": 0.0266, "step": 5360 }, { "grad_norm": 0.18810507655143738, "learning_rate": 4.803287873112877e-05, "loss": 0.0277, "step": 5370 }, { "grad_norm": 0.2624048888683319, "learning_rate": 4.786767082032681e-05, "loss": 0.0264, "step": 5380 }, { "grad_norm": 0.23779258131980896, "learning_rate": 4.77024862283351e-05, "loss": 0.0274, "step": 5390 }, { "grad_norm": 0.29530954360961914, "learning_rate": 4.753732676158593e-05, "loss": 0.024, "step": 5400 }, { "grad_norm": 0.22838789224624634, "learning_rate": 4.737219422623672e-05, "loss": 0.0261, "step": 5410 }, { "grad_norm": 0.16590280830860138, "learning_rate": 4.720709042815044e-05, "loss": 0.0235, "step": 5420 }, { "grad_norm": 0.2111019492149353, "learning_rate": 4.704201717287578e-05, "loss": 0.0229, "step": 5430 }, { "grad_norm": 0.2716585397720337, "learning_rate": 4.6876976265627404e-05, "loss": 0.0263, "step": 5440 }, { "grad_norm": 0.23845458030700684, "learning_rate": 4.671196951126626e-05, "loss": 0.0277, "step": 5450 }, { "grad_norm": 0.23560231924057007, "learning_rate": 4.654699871427971e-05, "loss": 0.0266, "step": 5460 }, { "grad_norm": 0.16281796991825104, "learning_rate": 4.6382065678762034e-05, "loss": 0.0225, "step": 5470 }, { "grad_norm": 0.2318909913301468, "learning_rate": 4.6217172208394424e-05, "loss": 0.028, "step": 5480 }, { "grad_norm": 0.28669074177742004, "learning_rate": 4.605232010642549e-05, "loss": 0.0224, "step": 5490 }, { "grad_norm": 0.24503110349178314, "learning_rate": 4.588751117565142e-05, "loss": 0.022, "step": 5500 }, { "grad_norm": 0.19491009414196014, "learning_rate": 4.5722747218396214e-05, "loss": 0.0269, "step": 5510 }, { "grad_norm": 0.12442632019519806, "learning_rate": 4.5558030036492194e-05, "loss": 0.0203, "step": 5520 }, { "grad_norm": 0.20821109414100647, "learning_rate": 4.539336143125999e-05, "loss": 0.0311, "step": 5530 }, { "grad_norm": 0.18892186880111694, "learning_rate": 4.522874320348916e-05, "loss": 0.02, "step": 5540 }, { "grad_norm": 0.2050052136182785, "learning_rate": 4.506417715341821e-05, "loss": 0.0232, "step": 5550 }, { "grad_norm": 0.18678373098373413, "learning_rate": 4.489966508071511e-05, "loss": 0.0198, "step": 5560 }, { "grad_norm": 0.23457950353622437, "learning_rate": 4.4735208784457575e-05, "loss": 0.0264, "step": 5570 }, { "grad_norm": 0.166547954082489, "learning_rate": 4.457081006311325e-05, "loss": 0.0209, "step": 5580 }, { "grad_norm": 0.1793099343776703, "learning_rate": 4.440647071452027e-05, "loss": 0.0235, "step": 5590 }, { "grad_norm": 0.23061560094356537, "learning_rate": 4.424219253586737e-05, "loss": 0.0227, "step": 5600 }, { "grad_norm": 0.21990767121315002, "learning_rate": 4.407797732367443e-05, "loss": 0.0303, "step": 5610 }, { "grad_norm": 0.2818678021430969, "learning_rate": 4.391382687377268e-05, "loss": 0.0229, "step": 5620 }, { "grad_norm": 0.321310818195343, "learning_rate": 4.374974298128512e-05, "loss": 0.0214, "step": 5630 }, { "grad_norm": 0.24316588044166565, "learning_rate": 4.358572744060699e-05, "loss": 0.0297, "step": 5640 }, { "grad_norm": 0.2073911875486374, "learning_rate": 4.342178204538588e-05, "loss": 0.0257, "step": 5650 }, { "grad_norm": 0.195980042219162, "learning_rate": 4.325790858850241e-05, "loss": 0.0248, "step": 5660 }, { "grad_norm": 0.16604967415332794, "learning_rate": 4.309410886205043e-05, "loss": 0.0239, "step": 5670 }, { "grad_norm": 0.1955181211233139, "learning_rate": 4.293038465731752e-05, "loss": 0.0254, "step": 5680 }, { "grad_norm": 0.18106386065483093, "learning_rate": 4.276673776476533e-05, "loss": 0.0308, "step": 5690 }, { "grad_norm": 0.08949582278728485, "learning_rate": 4.260316997401007e-05, "loss": 0.0217, "step": 5700 }, { "grad_norm": 0.28356561064720154, "learning_rate": 4.243968307380293e-05, "loss": 0.0246, "step": 5710 }, { "grad_norm": 0.1785511076450348, "learning_rate": 4.22762788520104e-05, "loss": 0.0283, "step": 5720 }, { "grad_norm": 0.253016859292984, "learning_rate": 4.211295909559491e-05, "loss": 0.0286, "step": 5730 }, { "grad_norm": 0.2753322124481201, "learning_rate": 4.194972559059511e-05, "loss": 0.0238, "step": 5740 }, { "grad_norm": 0.16788065433502197, "learning_rate": 4.178658012210651e-05, "loss": 0.0235, "step": 5750 }, { "grad_norm": 0.1692957729101181, "learning_rate": 4.162352447426177e-05, "loss": 0.0195, "step": 5760 }, { "grad_norm": 0.17816145718097687, "learning_rate": 4.146056043021135e-05, "loss": 0.0205, "step": 5770 }, { "grad_norm": 0.18158046901226044, "learning_rate": 4.1297689772103944e-05, "loss": 0.0244, "step": 5780 }, { "grad_norm": 0.242145374417305, "learning_rate": 4.113491428106694e-05, "loss": 0.0259, "step": 5790 }, { "grad_norm": 0.21062354743480682, "learning_rate": 4.0972235737187055e-05, "loss": 0.0245, "step": 5800 }, { "grad_norm": 0.18233442306518555, "learning_rate": 4.080965591949076e-05, "loss": 0.0249, "step": 5810 }, { "grad_norm": 0.2482534497976303, "learning_rate": 4.0647176605924924e-05, "loss": 0.0251, "step": 5820 }, { "grad_norm": 0.24134398996829987, "learning_rate": 4.0484799573337255e-05, "loss": 0.0304, "step": 5830 }, { "grad_norm": 0.2638164758682251, "learning_rate": 4.032252659745699e-05, "loss": 0.0262, "step": 5840 }, { "grad_norm": 0.23320716619491577, "learning_rate": 4.016035945287539e-05, "loss": 0.0264, "step": 5850 }, { "grad_norm": 0.19648794829845428, "learning_rate": 3.999829991302635e-05, "loss": 0.0216, "step": 5860 }, { "grad_norm": 0.19325271248817444, "learning_rate": 3.983634975016707e-05, "loss": 0.024, "step": 5870 }, { "grad_norm": 0.1748272329568863, "learning_rate": 3.967451073535854e-05, "loss": 0.0297, "step": 5880 }, { "grad_norm": 0.269959419965744, "learning_rate": 3.951278463844633e-05, "loss": 0.0255, "step": 5890 }, { "grad_norm": 0.29978659749031067, "learning_rate": 3.935117322804111e-05, "loss": 0.0282, "step": 5900 }, { "grad_norm": 0.19957485795021057, "learning_rate": 3.918967827149938e-05, "loss": 0.0245, "step": 5910 }, { "grad_norm": 0.20154502987861633, "learning_rate": 3.9028301534904094e-05, "loss": 0.0227, "step": 5920 }, { "grad_norm": 0.1569104790687561, "learning_rate": 3.88670447830454e-05, "loss": 0.0221, "step": 5930 }, { "grad_norm": 0.18918085098266602, "learning_rate": 3.870590977940132e-05, "loss": 0.0246, "step": 5940 }, { "grad_norm": 0.17519505321979523, "learning_rate": 3.8544898286118404e-05, "loss": 0.0198, "step": 5950 }, { "grad_norm": 0.1880214661359787, "learning_rate": 3.838401206399257e-05, "loss": 0.0237, "step": 5960 }, { "grad_norm": 0.8893473148345947, "learning_rate": 3.822325287244975e-05, "loss": 0.0218, "step": 5970 }, { "grad_norm": 0.2133779078722, "learning_rate": 3.8062622469526725e-05, "loss": 0.0231, "step": 5980 }, { "grad_norm": 0.26420339941978455, "learning_rate": 3.790212261185183e-05, "loss": 0.0226, "step": 5990 }, { "grad_norm": 0.22284920513629913, "learning_rate": 3.7741755054625794e-05, "loss": 0.0258, "step": 6000 }, { "grad_norm": 0.20438985526561737, "learning_rate": 3.758152155160255e-05, "loss": 0.0223, "step": 6010 }, { "grad_norm": 0.158140167593956, "learning_rate": 3.742142385506999e-05, "loss": 0.0304, "step": 6020 }, { "grad_norm": 0.2087687999010086, "learning_rate": 3.72614637158309e-05, "loss": 0.0222, "step": 6030 }, { "grad_norm": 0.16037289798259735, "learning_rate": 3.710164288318371e-05, "loss": 0.0238, "step": 6040 }, { "grad_norm": 0.17881503701210022, "learning_rate": 3.694196310490345e-05, "loss": 0.0196, "step": 6050 }, { "grad_norm": 0.17834429442882538, "learning_rate": 3.678242612722259e-05, "loss": 0.0194, "step": 6060 }, { "grad_norm": 0.217429518699646, "learning_rate": 3.6623033694811953e-05, "loss": 0.0203, "step": 6070 }, { "grad_norm": 0.18372195959091187, "learning_rate": 3.6463787550761665e-05, "loss": 0.0268, "step": 6080 }, { "grad_norm": 0.3576352596282959, "learning_rate": 3.630468943656202e-05, "loss": 0.0257, "step": 6090 }, { "grad_norm": 0.21325308084487915, "learning_rate": 3.6145741092084523e-05, "loss": 0.0235, "step": 6100 }, { "grad_norm": 0.15054462850093842, "learning_rate": 3.598694425556278e-05, "loss": 0.0203, "step": 6110 }, { "grad_norm": 0.17576907575130463, "learning_rate": 3.58283006635736e-05, "loss": 0.0228, "step": 6120 }, { "grad_norm": 0.12895582616329193, "learning_rate": 3.566981205101781e-05, "loss": 0.0223, "step": 6130 }, { "grad_norm": 0.13337048888206482, "learning_rate": 3.5511480151101556e-05, "loss": 0.0197, "step": 6140 }, { "grad_norm": 0.19493821263313293, "learning_rate": 3.5353306695317104e-05, "loss": 0.0188, "step": 6150 }, { "grad_norm": 0.15973666310310364, "learning_rate": 3.519529341342402e-05, "loss": 0.0208, "step": 6160 }, { "grad_norm": 0.194875106215477, "learning_rate": 3.503744203343026e-05, "loss": 0.0206, "step": 6170 }, { "grad_norm": 0.15493808686733246, "learning_rate": 3.487975428157318e-05, "loss": 0.0221, "step": 6180 }, { "grad_norm": 0.2392806112766266, "learning_rate": 3.472223188230083e-05, "loss": 0.0246, "step": 6190 }, { "grad_norm": 0.17829817533493042, "learning_rate": 3.4564876558252866e-05, "loss": 0.0219, "step": 6200 }, { "grad_norm": 0.133379727602005, "learning_rate": 3.440769003024195e-05, "loss": 0.021, "step": 6210 }, { "grad_norm": 0.17275074124336243, "learning_rate": 3.425067401723477e-05, "loss": 0.0196, "step": 6220 }, { "grad_norm": 0.17307311296463013, "learning_rate": 3.409383023633325e-05, "loss": 0.0229, "step": 6230 }, { "grad_norm": 0.29553088545799255, "learning_rate": 3.3937160402755894e-05, "loss": 0.025, "step": 6240 }, { "grad_norm": 0.14471402764320374, "learning_rate": 3.378066622981885e-05, "loss": 0.0224, "step": 6250 }, { "grad_norm": 0.14043006300926208, "learning_rate": 3.362434942891738e-05, "loss": 0.0268, "step": 6260 }, { "grad_norm": 0.22659337520599365, "learning_rate": 3.346821170950693e-05, "loss": 0.0266, "step": 6270 }, { "grad_norm": 0.10525943338871002, "learning_rate": 3.3312254779084585e-05, "loss": 0.0201, "step": 6280 }, { "grad_norm": 0.17259424924850464, "learning_rate": 3.315648034317039e-05, "loss": 0.0242, "step": 6290 }, { "grad_norm": 0.23512648046016693, "learning_rate": 3.3000890105288564e-05, "loss": 0.0232, "step": 6300 }, { "grad_norm": 0.3587051331996918, "learning_rate": 3.284548576694908e-05, "loss": 0.0222, "step": 6310 }, { "grad_norm": 0.2243478149175644, "learning_rate": 3.2690269027628815e-05, "loss": 0.0196, "step": 6320 }, { "grad_norm": 0.1820303052663803, "learning_rate": 3.253524158475324e-05, "loss": 0.0209, "step": 6330 }, { "grad_norm": 0.13899880647659302, "learning_rate": 3.238040513367757e-05, "loss": 0.0208, "step": 6340 }, { "grad_norm": 0.19786477088928223, "learning_rate": 3.222576136766843e-05, "loss": 0.0197, "step": 6350 }, { "grad_norm": 0.2029048204421997, "learning_rate": 3.2071311977885324e-05, "loss": 0.0211, "step": 6360 }, { "grad_norm": 0.14796499907970428, "learning_rate": 3.191705865336197e-05, "loss": 0.0219, "step": 6370 }, { "grad_norm": 0.24770914018154144, "learning_rate": 3.1763003080988075e-05, "loss": 0.0194, "step": 6380 }, { "grad_norm": 0.20868290960788727, "learning_rate": 3.160914694549063e-05, "loss": 0.0215, "step": 6390 }, { "grad_norm": 0.21524359285831451, "learning_rate": 3.145549192941573e-05, "loss": 0.0224, "step": 6400 }, { "grad_norm": 0.18463124334812164, "learning_rate": 3.130203971310999e-05, "loss": 0.0233, "step": 6410 }, { "grad_norm": 0.24082276225090027, "learning_rate": 3.114879197470225e-05, "loss": 0.0234, "step": 6420 }, { "grad_norm": 0.17373405396938324, "learning_rate": 3.0995750390085285e-05, "loss": 0.0179, "step": 6430 }, { "grad_norm": 0.12774558365345, "learning_rate": 3.084291663289728e-05, "loss": 0.0214, "step": 6440 }, { "grad_norm": 0.17318451404571533, "learning_rate": 3.069029237450375e-05, "loss": 0.0166, "step": 6450 }, { "grad_norm": 0.19220612943172455, "learning_rate": 3.053787928397911e-05, "loss": 0.018, "step": 6460 }, { "grad_norm": 0.21162372827529907, "learning_rate": 3.0385679028088526e-05, "loss": 0.0226, "step": 6470 }, { "grad_norm": 0.17719335854053497, "learning_rate": 3.023369327126959e-05, "loss": 0.0173, "step": 6480 }, { "grad_norm": 0.17820969223976135, "learning_rate": 3.0081923675614198e-05, "loss": 0.0174, "step": 6490 }, { "grad_norm": 0.17941966652870178, "learning_rate": 2.993037190085034e-05, "loss": 0.0191, "step": 6500 }, { "grad_norm": 0.17816117405891418, "learning_rate": 2.977903960432392e-05, "loss": 0.0217, "step": 6510 }, { "grad_norm": 0.3016155958175659, "learning_rate": 2.9627928440980722e-05, "loss": 0.0253, "step": 6520 }, { "grad_norm": 0.1797763556241989, "learning_rate": 2.9477040063348183e-05, "loss": 0.0204, "step": 6530 }, { "grad_norm": 0.1276247799396515, "learning_rate": 2.9326376121517456e-05, "loss": 0.0212, "step": 6540 }, { "grad_norm": 0.1688711941242218, "learning_rate": 2.9175938263125236e-05, "loss": 0.0201, "step": 6550 }, { "grad_norm": 0.19279010593891144, "learning_rate": 2.9025728133335873e-05, "loss": 0.0201, "step": 6560 }, { "grad_norm": 0.21296808123588562, "learning_rate": 2.8875747374823288e-05, "loss": 0.0249, "step": 6570 }, { "grad_norm": 0.19720396399497986, "learning_rate": 2.872599762775298e-05, "loss": 0.0258, "step": 6580 }, { "grad_norm": 0.16150468587875366, "learning_rate": 2.857648052976425e-05, "loss": 0.0211, "step": 6590 }, { "grad_norm": 0.24893704056739807, "learning_rate": 2.8427197715952047e-05, "loss": 0.0205, "step": 6600 }, { "grad_norm": 0.16543982923030853, "learning_rate": 2.8278150818849393e-05, "loss": 0.0193, "step": 6610 }, { "grad_norm": 0.18192972242832184, "learning_rate": 2.812934146840922e-05, "loss": 0.0184, "step": 6620 }, { "grad_norm": 0.2210838496685028, "learning_rate": 2.7980771291986764e-05, "loss": 0.0212, "step": 6630 }, { "grad_norm": 0.21865949034690857, "learning_rate": 2.783244191432167e-05, "loss": 0.019, "step": 6640 }, { "grad_norm": 0.194326713681221, "learning_rate": 2.768435495752022e-05, "loss": 0.0202, "step": 6650 }, { "grad_norm": 0.18440577387809753, "learning_rate": 2.753651204103771e-05, "loss": 0.0206, "step": 6660 }, { "grad_norm": 0.21295639872550964, "learning_rate": 2.7388914781660523e-05, "loss": 0.0214, "step": 6670 }, { "grad_norm": 0.16129761934280396, "learning_rate": 2.7241564793488693e-05, "loss": 0.0212, "step": 6680 }, { "grad_norm": 0.11878236383199692, "learning_rate": 2.7094463687918037e-05, "loss": 0.0201, "step": 6690 }, { "grad_norm": 0.22634334862232208, "learning_rate": 2.694761307362268e-05, "loss": 0.0194, "step": 6700 }, { "grad_norm": 0.1719747632741928, "learning_rate": 2.6801014556537467e-05, "loss": 0.0196, "step": 6710 }, { "grad_norm": 0.16810236871242523, "learning_rate": 2.6654669739840243e-05, "loss": 0.0191, "step": 6720 }, { "grad_norm": 0.23939171433448792, "learning_rate": 2.650858022393451e-05, "loss": 0.0222, "step": 6730 }, { "grad_norm": 0.16029053926467896, "learning_rate": 2.6362747606431747e-05, "loss": 0.0193, "step": 6740 }, { "grad_norm": 0.139109268784523, "learning_rate": 2.6217173482134172e-05, "loss": 0.0185, "step": 6750 }, { "grad_norm": 0.19581027328968048, "learning_rate": 2.6071859443017044e-05, "loss": 0.0184, "step": 6760 }, { "grad_norm": 0.1664178967475891, "learning_rate": 2.5926807078211414e-05, "loss": 0.0172, "step": 6770 }, { "grad_norm": 0.15122197568416595, "learning_rate": 2.5782017973986728e-05, "loss": 0.0162, "step": 6780 }, { "grad_norm": 0.2806341350078583, "learning_rate": 2.5637493713733374e-05, "loss": 0.018, "step": 6790 }, { "grad_norm": 0.23505569994449615, "learning_rate": 2.549323587794559e-05, "loss": 0.0247, "step": 6800 }, { "grad_norm": 0.14424313604831696, "learning_rate": 2.5349246044203895e-05, "loss": 0.0165, "step": 6810 }, { "grad_norm": 0.17154046893119812, "learning_rate": 2.520552578715808e-05, "loss": 0.0174, "step": 6820 }, { "grad_norm": 0.2045929729938507, "learning_rate": 2.506207667850981e-05, "loss": 0.0176, "step": 6830 }, { "grad_norm": 0.179784893989563, "learning_rate": 2.4918900286995555e-05, "loss": 0.0192, "step": 6840 }, { "grad_norm": 0.1777564436197281, "learning_rate": 2.4775998178369458e-05, "loss": 0.0182, "step": 6850 }, { "grad_norm": 0.2086399495601654, "learning_rate": 2.4633371915386017e-05, "loss": 0.0203, "step": 6860 }, { "grad_norm": 0.18741224706172943, "learning_rate": 2.4491023057783235e-05, "loss": 0.0224, "step": 6870 }, { "grad_norm": 0.25924691557884216, "learning_rate": 2.4348953162265375e-05, "loss": 0.0249, "step": 6880 }, { "grad_norm": 0.1669986993074417, "learning_rate": 2.420716378248607e-05, "loss": 0.0199, "step": 6890 }, { "grad_norm": 0.19221723079681396, "learning_rate": 2.4065656469031266e-05, "loss": 0.019, "step": 6900 }, { "grad_norm": 0.21825043857097626, "learning_rate": 2.3924432769402268e-05, "loss": 0.0187, "step": 6910 }, { "grad_norm": 0.20272719860076904, "learning_rate": 2.3783494227998844e-05, "loss": 0.0267, "step": 6920 }, { "grad_norm": 0.12414345145225525, "learning_rate": 2.3642842386102264e-05, "loss": 0.02, "step": 6930 }, { "grad_norm": 0.1465044468641281, "learning_rate": 2.3502478781858567e-05, "loss": 0.0186, "step": 6940 }, { "grad_norm": 0.17612896859645844, "learning_rate": 2.3362404950261628e-05, "loss": 0.0208, "step": 6950 }, { "grad_norm": 0.19352567195892334, "learning_rate": 2.3222622423136458e-05, "loss": 0.0192, "step": 6960 }, { "grad_norm": 0.15316592156887054, "learning_rate": 2.3083132729122332e-05, "loss": 0.0188, "step": 6970 }, { "grad_norm": 0.16287732124328613, "learning_rate": 2.294393739365621e-05, "loss": 0.0181, "step": 6980 }, { "grad_norm": 0.17146830260753632, "learning_rate": 2.2805037938956e-05, "loss": 0.0184, "step": 6990 }, { "grad_norm": 0.19281062483787537, "learning_rate": 2.266643588400386e-05, "loss": 0.0187, "step": 7000 }, { "grad_norm": 0.20195621252059937, "learning_rate": 2.252813274452969e-05, "loss": 0.0192, "step": 7010 }, { "grad_norm": 0.10406572371721268, "learning_rate": 2.2390130032994427e-05, "loss": 0.0142, "step": 7020 }, { "grad_norm": 0.11163413524627686, "learning_rate": 2.2252429258573633e-05, "loss": 0.0156, "step": 7030 }, { "grad_norm": 0.19183745980262756, "learning_rate": 2.2115031927140904e-05, "loss": 0.0201, "step": 7040 }, { "grad_norm": 0.1770429015159607, "learning_rate": 2.1977939541251463e-05, "loss": 0.0228, "step": 7050 }, { "grad_norm": 0.14570020139217377, "learning_rate": 2.1841153600125684e-05, "loss": 0.016, "step": 7060 }, { "grad_norm": 0.2018880993127823, "learning_rate": 2.170467559963267e-05, "loss": 0.0197, "step": 7070 }, { "grad_norm": 0.12946443259716034, "learning_rate": 2.1568507032273982e-05, "loss": 0.0168, "step": 7080 }, { "grad_norm": 0.16385304927825928, "learning_rate": 2.1432649387167264e-05, "loss": 0.0168, "step": 7090 }, { "grad_norm": 0.13708727061748505, "learning_rate": 2.1297104150029973e-05, "loss": 0.0166, "step": 7100 }, { "grad_norm": 0.2472526580095291, "learning_rate": 2.116187280316307e-05, "loss": 0.018, "step": 7110 }, { "grad_norm": 0.22644086182117462, "learning_rate": 2.1026956825434908e-05, "loss": 0.0204, "step": 7120 }, { "grad_norm": 0.17362110316753387, "learning_rate": 2.0892357692265017e-05, "loss": 0.0171, "step": 7130 }, { "grad_norm": 0.2007298320531845, "learning_rate": 2.0758076875607947e-05, "loss": 0.0188, "step": 7140 }, { "grad_norm": 0.1114509105682373, "learning_rate": 2.0624115843937207e-05, "loss": 0.019, "step": 7150 }, { "grad_norm": 0.11876010149717331, "learning_rate": 2.0490476062229157e-05, "loss": 0.022, "step": 7160 }, { "grad_norm": 0.2686578035354614, "learning_rate": 2.035715899194704e-05, "loss": 0.0179, "step": 7170 }, { "grad_norm": 0.12457459419965744, "learning_rate": 2.022416609102499e-05, "loss": 0.0154, "step": 7180 }, { "grad_norm": 0.10102657228708267, "learning_rate": 2.009149881385205e-05, "loss": 0.0226, "step": 7190 }, { "grad_norm": 0.14772836863994598, "learning_rate": 1.995915861125634e-05, "loss": 0.0178, "step": 7200 }, { "grad_norm": 0.12683475017547607, "learning_rate": 1.9827146930489065e-05, "loss": 0.0143, "step": 7210 }, { "grad_norm": 0.2532680630683899, "learning_rate": 1.9695465215208848e-05, "loss": 0.0212, "step": 7220 }, { "grad_norm": 0.11957696825265884, "learning_rate": 1.9564114905465813e-05, "loss": 0.0164, "step": 7230 }, { "grad_norm": 0.12867221236228943, "learning_rate": 1.9433097437685936e-05, "loss": 0.0168, "step": 7240 }, { "grad_norm": 0.18440400063991547, "learning_rate": 1.930241424465521e-05, "loss": 0.0164, "step": 7250 }, { "grad_norm": 0.18408997356891632, "learning_rate": 1.9172066755504115e-05, "loss": 0.0175, "step": 7260 }, { "grad_norm": 0.15457284450531006, "learning_rate": 1.9042056395691914e-05, "loss": 0.0211, "step": 7270 }, { "grad_norm": 0.11811310052871704, "learning_rate": 1.8912384586991066e-05, "loss": 0.0145, "step": 7280 }, { "grad_norm": 0.12857763469219208, "learning_rate": 1.8783052747471717e-05, "loss": 0.0154, "step": 7290 }, { "grad_norm": 0.19257484376430511, "learning_rate": 1.865406229148611e-05, "loss": 0.0192, "step": 7300 }, { "grad_norm": 0.13709598779678345, "learning_rate": 1.8525414629653233e-05, "loss": 0.0185, "step": 7310 }, { "grad_norm": 0.18893171846866608, "learning_rate": 1.8397111168843255e-05, "loss": 0.0162, "step": 7320 }, { "grad_norm": 0.17083989083766937, "learning_rate": 1.8269153312162323e-05, "loss": 0.0186, "step": 7330 }, { "grad_norm": 0.13018789887428284, "learning_rate": 1.8141542458937054e-05, "loss": 0.0165, "step": 7340 }, { "grad_norm": 0.1095215454697609, "learning_rate": 1.8014280004699268e-05, "loss": 0.0173, "step": 7350 }, { "grad_norm": 0.1645497828722, "learning_rate": 1.788736734117078e-05, "loss": 0.0143, "step": 7360 }, { "grad_norm": 0.12877090275287628, "learning_rate": 1.7760805856248152e-05, "loss": 0.0178, "step": 7370 }, { "grad_norm": 0.15835006535053253, "learning_rate": 1.7634596933987518e-05, "loss": 0.0199, "step": 7380 }, { "grad_norm": 0.15329298377037048, "learning_rate": 1.7508741954589404e-05, "loss": 0.0176, "step": 7390 }, { "grad_norm": 0.13807955384254456, "learning_rate": 1.7383242294383717e-05, "loss": 0.0159, "step": 7400 }, { "grad_norm": 0.1560157984495163, "learning_rate": 1.7258099325814632e-05, "loss": 0.0142, "step": 7410 }, { "grad_norm": 0.16062749922275543, "learning_rate": 1.7133314417425594e-05, "loss": 0.0178, "step": 7420 }, { "grad_norm": 0.13994382321834564, "learning_rate": 1.7008888933844408e-05, "loss": 0.0168, "step": 7430 }, { "grad_norm": 0.12487193197011948, "learning_rate": 1.6884824235768172e-05, "loss": 0.0201, "step": 7440 }, { "grad_norm": 0.1565808802843094, "learning_rate": 1.6761121679948592e-05, "loss": 0.0159, "step": 7450 }, { "grad_norm": 0.16437600553035736, "learning_rate": 1.663778261917695e-05, "loss": 0.0172, "step": 7460 }, { "grad_norm": 0.17170003056526184, "learning_rate": 1.651480840226952e-05, "loss": 0.0162, "step": 7470 }, { "grad_norm": 0.12132951617240906, "learning_rate": 1.639220037405258e-05, "loss": 0.0192, "step": 7480 }, { "grad_norm": 0.14489589631557465, "learning_rate": 1.6269959875347906e-05, "loss": 0.0177, "step": 7490 }, { "grad_norm": 0.17209839820861816, "learning_rate": 1.614808824295802e-05, "loss": 0.0162, "step": 7500 }, { "grad_norm": 0.14911475777626038, "learning_rate": 1.602658680965152e-05, "loss": 0.0143, "step": 7510 }, { "grad_norm": 0.13209745287895203, "learning_rate": 1.5905456904148686e-05, "loss": 0.0188, "step": 7520 }, { "grad_norm": 0.15718616545200348, "learning_rate": 1.57846998511067e-05, "loss": 0.0199, "step": 7530 }, { "grad_norm": 0.1673203855752945, "learning_rate": 1.566431697110538e-05, "loss": 0.0159, "step": 7540 }, { "grad_norm": 0.15472687780857086, "learning_rate": 1.554430958063259e-05, "loss": 0.0167, "step": 7550 }, { "grad_norm": 0.17345958948135376, "learning_rate": 1.5424678992069912e-05, "loss": 0.0145, "step": 7560 }, { "grad_norm": 0.28122711181640625, "learning_rate": 1.5305426513678362e-05, "loss": 0.0169, "step": 7570 }, { "grad_norm": 0.18371731042861938, "learning_rate": 1.518655344958388e-05, "loss": 0.0188, "step": 7580 }, { "grad_norm": 0.17419502139091492, "learning_rate": 1.5068061099763275e-05, "loss": 0.0171, "step": 7590 }, { "grad_norm": 0.09389843046665192, "learning_rate": 1.494995076002988e-05, "loss": 0.0161, "step": 7600 }, { "grad_norm": 0.15366578102111816, "learning_rate": 1.4832223722019456e-05, "loss": 0.0136, "step": 7610 }, { "grad_norm": 0.15734989941120148, "learning_rate": 1.4714881273176035e-05, "loss": 0.0175, "step": 7620 }, { "grad_norm": 0.11974746733903885, "learning_rate": 1.4597924696737835e-05, "loss": 0.0156, "step": 7630 }, { "grad_norm": 0.13332171738147736, "learning_rate": 1.4481355271723252e-05, "loss": 0.0147, "step": 7640 }, { "grad_norm": 0.17544510960578918, "learning_rate": 1.4365174272916809e-05, "loss": 0.0169, "step": 7650 }, { "grad_norm": 0.15159939229488373, "learning_rate": 1.4249382970855319e-05, "loss": 0.0171, "step": 7660 }, { "grad_norm": 0.179666668176651, "learning_rate": 1.4133982631813903e-05, "loss": 0.0168, "step": 7670 }, { "grad_norm": 0.17890037596225739, "learning_rate": 1.4018974517792194e-05, "loss": 0.0154, "step": 7680 }, { "grad_norm": 0.14395786821842194, "learning_rate": 1.390435988650048e-05, "loss": 0.0167, "step": 7690 }, { "grad_norm": 0.1503734141588211, "learning_rate": 1.3790139991346006e-05, "loss": 0.0159, "step": 7700 }, { "grad_norm": 0.16791732609272003, "learning_rate": 1.367631608141926e-05, "loss": 0.02, "step": 7710 }, { "grad_norm": 0.1589469164609909, "learning_rate": 1.3562889401480278e-05, "loss": 0.0161, "step": 7720 }, { "grad_norm": 0.19974438846111298, "learning_rate": 1.3449861191945074e-05, "loss": 0.0161, "step": 7730 }, { "grad_norm": 0.1391679048538208, "learning_rate": 1.3337232688872009e-05, "loss": 0.0145, "step": 7740 }, { "grad_norm": 0.1625400185585022, "learning_rate": 1.3225005123948364e-05, "loss": 0.0157, "step": 7750 }, { "grad_norm": 0.15353775024414062, "learning_rate": 1.311317972447681e-05, "loss": 0.0138, "step": 7760 }, { "grad_norm": 0.14045576751232147, "learning_rate": 1.3001757713361996e-05, "loss": 0.0126, "step": 7770 }, { "grad_norm": 0.1758357137441635, "learning_rate": 1.2890740309097204e-05, "loss": 0.0187, "step": 7780 }, { "grad_norm": 0.22632494568824768, "learning_rate": 1.2780128725750944e-05, "loss": 0.0163, "step": 7790 }, { "grad_norm": 0.13876816630363464, "learning_rate": 1.266992417295379e-05, "loss": 0.0173, "step": 7800 }, { "grad_norm": 0.1254066526889801, "learning_rate": 1.2560127855885073e-05, "loss": 0.0176, "step": 7810 }, { "grad_norm": 0.18325023353099823, "learning_rate": 1.2450740975259745e-05, "loss": 0.0186, "step": 7820 }, { "grad_norm": 0.1624547243118286, "learning_rate": 1.234176472731517e-05, "loss": 0.0216, "step": 7830 }, { "grad_norm": 0.1877821981906891, "learning_rate": 1.2233200303798158e-05, "loss": 0.0168, "step": 7840 }, { "grad_norm": 0.13350386917591095, "learning_rate": 1.2125048891951846e-05, "loss": 0.0151, "step": 7850 }, { "grad_norm": 0.14304585754871368, "learning_rate": 1.2017311674502745e-05, "loss": 0.0154, "step": 7860 }, { "grad_norm": 0.17110075056552887, "learning_rate": 1.1909989829647822e-05, "loss": 0.0154, "step": 7870 }, { "grad_norm": 0.14714130759239197, "learning_rate": 1.1803084531041553e-05, "loss": 0.0164, "step": 7880 }, { "grad_norm": 0.17469313740730286, "learning_rate": 1.1696596947783162e-05, "loss": 0.0178, "step": 7890 }, { "grad_norm": 0.18000158667564392, "learning_rate": 1.1590528244403803e-05, "loss": 0.0175, "step": 7900 }, { "grad_norm": 0.1818334460258484, "learning_rate": 1.148487958085382e-05, "loss": 0.0164, "step": 7910 }, { "grad_norm": 0.14082416892051697, "learning_rate": 1.1379652112490086e-05, "loss": 0.0158, "step": 7920 }, { "grad_norm": 0.250521183013916, "learning_rate": 1.1274846990063315e-05, "loss": 0.017, "step": 7930 }, { "grad_norm": 0.22370409965515137, "learning_rate": 1.117046535970554e-05, "loss": 0.0172, "step": 7940 }, { "grad_norm": 0.17268650233745575, "learning_rate": 1.106650836291755e-05, "loss": 0.0208, "step": 7950 }, { "grad_norm": 0.19990529119968414, "learning_rate": 1.0962977136556418e-05, "loss": 0.0231, "step": 7960 }, { "grad_norm": 0.20242197811603546, "learning_rate": 1.0859872812823024e-05, "loss": 0.0182, "step": 7970 }, { "grad_norm": 0.292499303817749, "learning_rate": 1.0757196519249747e-05, "loss": 0.0245, "step": 7980 }, { "grad_norm": 0.1894037276506424, "learning_rate": 1.0654949378688077e-05, "loss": 0.0174, "step": 7990 }, { "grad_norm": 0.1939881145954132, "learning_rate": 1.0553132509296376e-05, "loss": 0.0179, "step": 8000 }, { "grad_norm": 0.11809485405683517, "learning_rate": 1.0451747024527613e-05, "loss": 0.019, "step": 8010 }, { "grad_norm": 0.14520397782325745, "learning_rate": 1.0350794033117189e-05, "loss": 0.0165, "step": 8020 }, { "grad_norm": 0.12728576362133026, "learning_rate": 1.0250274639070856e-05, "loss": 0.0192, "step": 8030 }, { "grad_norm": 0.1828681230545044, "learning_rate": 1.0150189941652599e-05, "loss": 0.0148, "step": 8040 }, { "grad_norm": 0.22145496308803558, "learning_rate": 1.0050541035372635e-05, "loss": 0.0163, "step": 8050 }, { "grad_norm": 0.09645770490169525, "learning_rate": 9.951329009975458e-06, "loss": 0.0165, "step": 8060 }, { "grad_norm": 0.1380506455898285, "learning_rate": 9.852554950427845e-06, "loss": 0.0139, "step": 8070 }, { "grad_norm": 0.7945582270622253, "learning_rate": 9.754219936907105e-06, "loss": 0.0171, "step": 8080 }, { "grad_norm": 0.11642926186323166, "learning_rate": 9.656325044789194e-06, "loss": 0.0147, "step": 8090 }, { "grad_norm": 0.1195152997970581, "learning_rate": 9.55887134463697e-06, "loss": 0.0159, "step": 8100 }, { "grad_norm": 0.1575433313846588, "learning_rate": 9.461859902188475e-06, "loss": 0.0185, "step": 8110 }, { "grad_norm": 0.1328561007976532, "learning_rate": 9.365291778345303e-06, "loss": 0.0182, "step": 8120 }, { "grad_norm": 0.13686002790927887, "learning_rate": 9.269168029160991e-06, "loss": 0.0145, "step": 8130 }, { "grad_norm": 0.2300335317850113, "learning_rate": 9.173489705829447e-06, "loss": 0.0165, "step": 8140 }, { "grad_norm": 0.16561192274093628, "learning_rate": 9.078257854673516e-06, "loss": 0.0182, "step": 8150 }, { "grad_norm": 0.09173670411109924, "learning_rate": 8.983473517133429e-06, "loss": 0.015, "step": 8160 }, { "grad_norm": 0.1712142825126648, "learning_rate": 8.889137729755537e-06, "loss": 0.0178, "step": 8170 }, { "grad_norm": 0.1319228559732437, "learning_rate": 8.79525152418087e-06, "loss": 0.02, "step": 8180 }, { "grad_norm": 0.12378199398517609, "learning_rate": 8.701815927133961e-06, "loss": 0.0177, "step": 8190 }, { "grad_norm": 0.19220711290836334, "learning_rate": 8.608831960411534e-06, "loss": 0.0167, "step": 8200 }, { "grad_norm": 0.152066171169281, "learning_rate": 8.516300640871321e-06, "loss": 0.0141, "step": 8210 }, { "grad_norm": 0.15020455420017242, "learning_rate": 8.424222980421038e-06, "loss": 0.0158, "step": 8220 }, { "grad_norm": 0.2237347513437271, "learning_rate": 8.332599986007184e-06, "loss": 0.0165, "step": 8230 }, { "grad_norm": 0.09469515085220337, "learning_rate": 8.241432659604203e-06, "loss": 0.014, "step": 8240 }, { "grad_norm": 0.10305121541023254, "learning_rate": 8.150721998203331e-06, "loss": 0.0145, "step": 8250 }, { "grad_norm": 0.2260267734527588, "learning_rate": 8.06046899380184e-06, "loss": 0.0133, "step": 8260 }, { "grad_norm": 0.24169157445430756, "learning_rate": 7.970674633392133e-06, "loss": 0.0204, "step": 8270 }, { "grad_norm": 0.11277104914188385, "learning_rate": 7.881339898950924e-06, "loss": 0.0132, "step": 8280 }, { "grad_norm": 0.16091711819171906, "learning_rate": 7.792465767428597e-06, "loss": 0.0128, "step": 8290 }, { "grad_norm": 0.11624167859554291, "learning_rate": 7.704053210738376e-06, "loss": 0.0135, "step": 8300 }, { "grad_norm": 0.22575534880161285, "learning_rate": 7.6161031957458494e-06, "loss": 0.0207, "step": 8310 }, { "grad_norm": 0.16817863285541534, "learning_rate": 7.5286166842582605e-06, "loss": 0.019, "step": 8320 }, { "grad_norm": 0.21053095161914825, "learning_rate": 7.4415946330140814e-06, "loss": 0.0169, "step": 8330 }, { "grad_norm": 0.14203409850597382, "learning_rate": 7.3550379936725644e-06, "loss": 0.0141, "step": 8340 }, { "grad_norm": 0.13740839064121246, "learning_rate": 7.2689477128032035e-06, "loss": 0.0145, "step": 8350 }, { "grad_norm": 0.14529792964458466, "learning_rate": 7.183324731875551e-06, "loss": 0.0161, "step": 8360 }, { "grad_norm": 0.1367732137441635, "learning_rate": 7.098169987248782e-06, "loss": 0.0125, "step": 8370 }, { "grad_norm": 0.10992296040058136, "learning_rate": 7.013484410161553e-06, "loss": 0.0152, "step": 8380 }, { "grad_norm": 0.15979985892772675, "learning_rate": 6.92926892672176e-06, "loss": 0.0144, "step": 8390 }, { "grad_norm": 0.12830288708209991, "learning_rate": 6.845524457896446e-06, "loss": 0.0117, "step": 8400 }, { "grad_norm": 0.13029831647872925, "learning_rate": 6.7622519195017165e-06, "loss": 0.0174, "step": 8410 }, { "grad_norm": 0.15612126886844635, "learning_rate": 6.679452222192684e-06, "loss": 0.0135, "step": 8420 }, { "grad_norm": 0.16025367379188538, "learning_rate": 6.597126271453579e-06, "loss": 0.0143, "step": 8430 }, { "grad_norm": 0.14055493474006653, "learning_rate": 6.51527496758782e-06, "loss": 0.0139, "step": 8440 }, { "grad_norm": 0.12680508196353912, "learning_rate": 6.433899205708155e-06, "loss": 0.0134, "step": 8450 }, { "grad_norm": 0.15970245003700256, "learning_rate": 6.352999875726856e-06, "loss": 0.014, "step": 8460 }, { "grad_norm": 0.11459923535585403, "learning_rate": 6.272577862346052e-06, "loss": 0.0144, "step": 8470 }, { "grad_norm": 0.19914846122264862, "learning_rate": 6.192634045047996e-06, "loss": 0.0212, "step": 8480 }, { "grad_norm": 0.17813663184642792, "learning_rate": 6.113169298085458e-06, "loss": 0.0163, "step": 8490 }, { "grad_norm": 0.19133634865283966, "learning_rate": 6.034184490472195e-06, "loss": 0.0159, "step": 8500 }, { "grad_norm": 0.14250624179840088, "learning_rate": 5.955680485973386e-06, "loss": 0.0141, "step": 8510 }, { "grad_norm": 0.1711205393075943, "learning_rate": 5.877658143096265e-06, "loss": 0.0165, "step": 8520 }, { "grad_norm": 0.08294077962636948, "learning_rate": 5.800118315080661e-06, "loss": 0.0129, "step": 8530 }, { "grad_norm": 0.16110099852085114, "learning_rate": 5.723061849889716e-06, "loss": 0.0148, "step": 8540 }, { "grad_norm": 0.290877103805542, "learning_rate": 5.646489590200604e-06, "loss": 0.0167, "step": 8550 }, { "grad_norm": 0.07924945652484894, "learning_rate": 5.570402373395256e-06, "loss": 0.0113, "step": 8560 }, { "grad_norm": 0.1851269006729126, "learning_rate": 5.494801031551305e-06, "loss": 0.0167, "step": 8570 }, { "grad_norm": 0.09695053845643997, "learning_rate": 5.41968639143291e-06, "loss": 0.0149, "step": 8580 }, { "grad_norm": 0.13390277326107025, "learning_rate": 5.345059274481751e-06, "loss": 0.0147, "step": 8590 }, { "grad_norm": 0.15931259095668793, "learning_rate": 5.270920496808002e-06, "loss": 0.0159, "step": 8600 }, { "grad_norm": 0.1652296781539917, "learning_rate": 5.1972708691814695e-06, "loss": 0.0159, "step": 8610 }, { "grad_norm": 0.13737909495830536, "learning_rate": 5.124111197022674e-06, "loss": 0.0146, "step": 8620 }, { "grad_norm": 0.12145192176103592, "learning_rate": 5.051442280394081e-06, "loss": 0.0126, "step": 8630 }, { "grad_norm": 0.15984362363815308, "learning_rate": 4.979264913991322e-06, "loss": 0.013, "step": 8640 }, { "grad_norm": 0.14123108983039856, "learning_rate": 4.907579887134489e-06, "loss": 0.014, "step": 8650 }, { "grad_norm": 0.15598590672016144, "learning_rate": 4.836387983759572e-06, "loss": 0.0148, "step": 8660 }, { "grad_norm": 0.10904663801193237, "learning_rate": 4.765689982409816e-06, "loss": 0.0138, "step": 8670 }, { "grad_norm": 0.1333862543106079, "learning_rate": 4.695486656227233e-06, "loss": 0.0179, "step": 8680 }, { "grad_norm": 0.08314937353134155, "learning_rate": 4.625778772944156e-06, "loss": 0.0139, "step": 8690 }, { "grad_norm": 0.09639056771993637, "learning_rate": 4.556567094874825e-06, "loss": 0.0141, "step": 8700 }, { "grad_norm": 0.11278994381427765, "learning_rate": 4.487852378907059e-06, "loss": 0.0115, "step": 8710 }, { "grad_norm": 0.11118289828300476, "learning_rate": 4.419635376493986e-06, "loss": 0.0133, "step": 8720 }, { "grad_norm": 0.1554441899061203, "learning_rate": 4.351916833645825e-06, "loss": 0.0145, "step": 8730 }, { "grad_norm": 0.16225364804267883, "learning_rate": 4.284697490921691e-06, "loss": 0.016, "step": 8740 }, { "grad_norm": 0.15683670341968536, "learning_rate": 4.2179780834215585e-06, "loss": 0.0159, "step": 8750 }, { "grad_norm": 0.1803421825170517, "learning_rate": 4.151759340778178e-06, "loss": 0.0164, "step": 8760 }, { "grad_norm": 0.13370923697948456, "learning_rate": 4.086041987149109e-06, "loss": 0.0137, "step": 8770 }, { "grad_norm": 0.1665254682302475, "learning_rate": 4.020826741208811e-06, "loss": 0.017, "step": 8780 }, { "grad_norm": 0.1833808571100235, "learning_rate": 3.956114316140746e-06, "loss": 0.0154, "step": 8790 }, { "grad_norm": 0.12698817253112793, "learning_rate": 3.891905419629643e-06, "loss": 0.0129, "step": 8800 }, { "grad_norm": 0.1011931523680687, "learning_rate": 3.8282007538536946e-06, "loss": 0.0187, "step": 8810 }, { "grad_norm": 0.22995996475219727, "learning_rate": 3.7650010154769265e-06, "loss": 0.0156, "step": 8820 }, { "grad_norm": 0.12448207288980484, "learning_rate": 3.7023068956415608e-06, "loss": 0.0141, "step": 8830 }, { "grad_norm": 0.12978610396385193, "learning_rate": 3.6401190799604303e-06, "loss": 0.0127, "step": 8840 }, { "grad_norm": 0.20795798301696777, "learning_rate": 3.578438248509536e-06, "loss": 0.0167, "step": 8850 }, { "grad_norm": 0.16256587207317352, "learning_rate": 3.5172650758205583e-06, "loss": 0.0163, "step": 8860 }, { "grad_norm": 0.12686973810195923, "learning_rate": 3.45660023087353e-06, "loss": 0.014, "step": 8870 }, { "grad_norm": 0.10828518867492676, "learning_rate": 3.3964443770894528e-06, "loss": 0.0134, "step": 8880 }, { "grad_norm": 0.10737987607717514, "learning_rate": 3.3367981723231245e-06, "loss": 0.0118, "step": 8890 }, { "grad_norm": 0.3040982186794281, "learning_rate": 3.2776622688558746e-06, "loss": 0.0166, "step": 8900 }, { "grad_norm": 0.16585928201675415, "learning_rate": 3.2190373133884677e-06, "loss": 0.0135, "step": 8910 }, { "grad_norm": 0.1141418069601059, "learning_rate": 3.1609239470340446e-06, "loss": 0.0122, "step": 8920 }, { "grad_norm": 0.08903242647647858, "learning_rate": 3.1033228053110373e-06, "loss": 0.0116, "step": 8930 }, { "grad_norm": 0.12849244475364685, "learning_rate": 3.0462345181363314e-06, "loss": 0.0128, "step": 8940 }, { "grad_norm": 0.13789059221744537, "learning_rate": 2.9896597098182654e-06, "loss": 0.0156, "step": 8950 }, { "grad_norm": 0.20985892415046692, "learning_rate": 2.933598999049891e-06, "loss": 0.0141, "step": 8960 }, { "grad_norm": 0.12013689428567886, "learning_rate": 2.8780529989021697e-06, "loss": 0.012, "step": 8970 }, { "grad_norm": 0.09641115367412567, "learning_rate": 2.823022316817242e-06, "loss": 0.0173, "step": 8980 }, { "grad_norm": 0.12721452116966248, "learning_rate": 2.7685075546018456e-06, "loss": 0.0153, "step": 8990 }, { "grad_norm": 0.1013830378651619, "learning_rate": 2.7145093084206598e-06, "loss": 0.0142, "step": 9000 }, { "grad_norm": 0.1277765929698944, "learning_rate": 2.661028168789892e-06, "loss": 0.0124, "step": 9010 }, { "grad_norm": 0.21034492552280426, "learning_rate": 2.6080647205706855e-06, "loss": 0.0137, "step": 9020 }, { "grad_norm": 0.13637621700763702, "learning_rate": 2.555619542962834e-06, "loss": 0.0145, "step": 9030 }, { "grad_norm": 0.17367185652256012, "learning_rate": 2.503693209498409e-06, "loss": 0.0138, "step": 9040 }, { "grad_norm": 0.09734096378087997, "learning_rate": 2.452286288035449e-06, "loss": 0.0135, "step": 9050 }, { "grad_norm": 0.12714159488677979, "learning_rate": 2.4013993407518363e-06, "loss": 0.0135, "step": 9060 }, { "grad_norm": 0.19327189028263092, "learning_rate": 2.351032924139063e-06, "loss": 0.013, "step": 9070 }, { "grad_norm": 0.12542635202407837, "learning_rate": 2.30118758899619e-06, "loss": 0.0129, "step": 9080 }, { "grad_norm": 0.09104075282812119, "learning_rate": 2.2518638804238157e-06, "loss": 0.0151, "step": 9090 }, { "grad_norm": 0.142464280128479, "learning_rate": 2.203062337818118e-06, "loss": 0.0148, "step": 9100 }, { "grad_norm": 0.12345139682292938, "learning_rate": 2.1547834948649483e-06, "loss": 0.0139, "step": 9110 }, { "grad_norm": 0.19940927624702454, "learning_rate": 2.1070278795340017e-06, "loss": 0.014, "step": 9120 }, { "grad_norm": 0.14735978841781616, "learning_rate": 2.059796014073029e-06, "loss": 0.0136, "step": 9130 }, { "grad_norm": 0.15923340618610382, "learning_rate": 2.01308841500214e-06, "loss": 0.0131, "step": 9140 }, { "grad_norm": 0.15401338040828705, "learning_rate": 1.9669055931081704e-06, "loss": 0.0151, "step": 9150 }, { "grad_norm": 0.10889153182506561, "learning_rate": 1.9212480534390507e-06, "loss": 0.0154, "step": 9160 }, { "grad_norm": 0.15817081928253174, "learning_rate": 1.8761162952983246e-06, "loss": 0.0161, "step": 9170 }, { "grad_norm": 0.15868796408176422, "learning_rate": 1.8315108122396618e-06, "loss": 0.0146, "step": 9180 }, { "grad_norm": 0.10401670634746552, "learning_rate": 1.787432092061475e-06, "loss": 0.0131, "step": 9190 }, { "grad_norm": 0.173081636428833, "learning_rate": 1.743880616801602e-06, "loss": 0.0142, "step": 9200 }, { "grad_norm": 0.17190033197402954, "learning_rate": 1.7008568627319865e-06, "loss": 0.0169, "step": 9210 }, { "grad_norm": 0.22819815576076508, "learning_rate": 1.6583613003535226e-06, "loss": 0.0173, "step": 9220 }, { "grad_norm": 0.11464705318212509, "learning_rate": 1.6163943943908522e-06, "loss": 0.0143, "step": 9230 }, { "grad_norm": 0.12012580782175064, "learning_rate": 1.5749566037873476e-06, "loss": 0.0112, "step": 9240 }, { "grad_norm": 0.12822668254375458, "learning_rate": 1.5340483817000428e-06, "loss": 0.0148, "step": 9250 }, { "grad_norm": 0.08638016134500504, "learning_rate": 1.4936701754947101e-06, "loss": 0.0119, "step": 9260 }, { "grad_norm": 0.12327328324317932, "learning_rate": 1.4538224267409361e-06, "loss": 0.0171, "step": 9270 }, { "grad_norm": 0.11788056045770645, "learning_rate": 1.414505571207314e-06, "loss": 0.0123, "step": 9280 }, { "grad_norm": 0.13440822064876556, "learning_rate": 1.3757200388566816e-06, "loss": 0.0129, "step": 9290 }, { "grad_norm": 0.147916778922081, "learning_rate": 1.3374662538414074e-06, "loss": 0.0155, "step": 9300 }, { "grad_norm": 0.40608835220336914, "learning_rate": 1.2997446344987617e-06, "loss": 0.0151, "step": 9310 }, { "grad_norm": 0.21104620397090912, "learning_rate": 1.262555593346315e-06, "loss": 0.013, "step": 9320 }, { "grad_norm": 0.1760464608669281, "learning_rate": 1.2258995370774685e-06, "loss": 0.0145, "step": 9330 }, { "grad_norm": 0.11093448847532272, "learning_rate": 1.1897768665569798e-06, "loss": 0.0142, "step": 9340 }, { "grad_norm": 0.13234373927116394, "learning_rate": 1.1541879768165954e-06, "loss": 0.0128, "step": 9350 }, { "grad_norm": 0.1820833534002304, "learning_rate": 1.1191332570507085e-06, "loss": 0.0158, "step": 9360 }, { "grad_norm": 0.11815668642520905, "learning_rate": 1.0846130906121132e-06, "loss": 0.0163, "step": 9370 }, { "grad_norm": 0.09018229693174362, "learning_rate": 1.0506278550078131e-06, "loss": 0.016, "step": 9380 }, { "grad_norm": 0.11231695860624313, "learning_rate": 1.0171779218949185e-06, "loss": 0.0121, "step": 9390 }, { "grad_norm": 0.1269022822380066, "learning_rate": 9.842636570765174e-07, "loss": 0.0121, "step": 9400 }, { "grad_norm": 0.19012972712516785, "learning_rate": 9.518854204977612e-07, "loss": 0.0122, "step": 9410 }, { "grad_norm": 0.17924317717552185, "learning_rate": 9.200435662418349e-07, "loss": 0.0143, "step": 9420 }, { "grad_norm": 0.24052727222442627, "learning_rate": 8.887384425261658e-07, "loss": 0.0134, "step": 9430 }, { "grad_norm": 0.17198549211025238, "learning_rate": 8.579703916985648e-07, "loss": 0.0159, "step": 9440 }, { "grad_norm": 0.10455486178398132, "learning_rate": 8.277397502335194e-07, "loss": 0.0133, "step": 9450 }, { "grad_norm": 0.14557500183582306, "learning_rate": 7.980468487284675e-07, "loss": 0.0134, "step": 9460 }, { "grad_norm": 0.1947624683380127, "learning_rate": 7.688920119002297e-07, "loss": 0.0151, "step": 9470 }, { "grad_norm": 0.17279574275016785, "learning_rate": 7.402755585814269e-07, "loss": 0.0132, "step": 9480 }, { "grad_norm": 0.1667862981557846, "learning_rate": 7.121978017170073e-07, "loss": 0.0152, "step": 9490 }, { "grad_norm": 0.1782369166612625, "learning_rate": 6.846590483608306e-07, "loss": 0.0171, "step": 9500 }, { "grad_norm": 0.11378748714923859, "learning_rate": 6.576595996722834e-07, "loss": 0.0148, "step": 9510 }, { "grad_norm": 0.1494322419166565, "learning_rate": 6.311997509130141e-07, "loss": 0.0178, "step": 9520 }, { "grad_norm": 0.15421955287456512, "learning_rate": 6.052797914436803e-07, "loss": 0.0156, "step": 9530 }, { "grad_norm": 0.18674427270889282, "learning_rate": 5.799000047208181e-07, "loss": 0.0171, "step": 9540 }, { "grad_norm": 0.1291435807943344, "learning_rate": 5.550606682937054e-07, "loss": 0.0156, "step": 9550 }, { "grad_norm": 0.1153743639588356, "learning_rate": 5.307620538013481e-07, "loss": 0.0149, "step": 9560 }, { "grad_norm": 0.14372242987155914, "learning_rate": 5.070044269694874e-07, "loss": 0.015, "step": 9570 }, { "grad_norm": 0.12678027153015137, "learning_rate": 4.837880476077417e-07, "loss": 0.016, "step": 9580 }, { "grad_norm": 0.1187908723950386, "learning_rate": 4.6111316960670835e-07, "loss": 0.0122, "step": 9590 }, { "grad_norm": 0.13331301510334015, "learning_rate": 4.389800409352218e-07, "loss": 0.0149, "step": 9600 }, { "grad_norm": 0.2063882052898407, "learning_rate": 4.173889036376277e-07, "loss": 0.0185, "step": 9610 }, { "grad_norm": 0.16114045679569244, "learning_rate": 3.963399938311463e-07, "loss": 0.0208, "step": 9620 }, { "grad_norm": 0.22279466688632965, "learning_rate": 3.7583354170328545e-07, "loss": 0.0186, "step": 9630 }, { "grad_norm": 0.10700543969869614, "learning_rate": 3.558697715093207e-07, "loss": 0.0128, "step": 9640 }, { "grad_norm": 0.2722342610359192, "learning_rate": 3.3644890156983576e-07, "loss": 0.0158, "step": 9650 }, { "grad_norm": 0.11648640036582947, "learning_rate": 3.175711442683638e-07, "loss": 0.0145, "step": 9660 }, { "grad_norm": 0.13997343182563782, "learning_rate": 2.9923670604902197e-07, "loss": 0.0152, "step": 9670 }, { "grad_norm": 0.11420878767967224, "learning_rate": 2.814457874143028e-07, "loss": 0.0123, "step": 9680 }, { "grad_norm": 0.14629878103733063, "learning_rate": 2.641985829228366e-07, "loss": 0.0113, "step": 9690 }, { "grad_norm": 0.13049235939979553, "learning_rate": 2.474952811872877e-07, "loss": 0.0135, "step": 9700 }, { "grad_norm": 0.18770433962345123, "learning_rate": 2.3133606487228397e-07, "loss": 0.0174, "step": 9710 }, { "grad_norm": 0.15080510079860687, "learning_rate": 2.157211106924295e-07, "loss": 0.0159, "step": 9720 }, { "grad_norm": 0.19432778656482697, "learning_rate": 2.006505894103672e-07, "loss": 0.016, "step": 9730 }, { "grad_norm": 0.06694447994232178, "learning_rate": 1.8612466583489696e-07, "loss": 0.0135, "step": 9740 }, { "grad_norm": 0.13827185332775116, "learning_rate": 1.7214349881918834e-07, "loss": 0.0134, "step": 9750 }, { "grad_norm": 0.16339640319347382, "learning_rate": 1.5870724125904845e-07, "loss": 0.0112, "step": 9760 }, { "grad_norm": 0.18306228518486023, "learning_rate": 1.4581604009124006e-07, "loss": 0.0145, "step": 9770 }, { "grad_norm": 0.1019846722483635, "learning_rate": 1.334700362918717e-07, "loss": 0.012, "step": 9780 }, { "grad_norm": 0.1288602650165558, "learning_rate": 1.2166936487486015e-07, "loss": 0.0162, "step": 9790 }, { "grad_norm": 0.14368805289268494, "learning_rate": 1.1041415489045914e-07, "loss": 0.0115, "step": 9800 }, { "grad_norm": 0.2050599455833435, "learning_rate": 9.970452942384412e-08, "loss": 0.0159, "step": 9810 }, { "grad_norm": 0.19492647051811218, "learning_rate": 8.954060559375754e-08, "loss": 0.0111, "step": 9820 }, { "grad_norm": 0.1194487139582634, "learning_rate": 7.99224945512489e-08, "loss": 0.0177, "step": 9830 }, { "grad_norm": 0.1959109753370285, "learning_rate": 7.085030147843675e-08, "loss": 0.0165, "step": 9840 }, { "grad_norm": 0.1559261828660965, "learning_rate": 6.232412558736523e-08, "loss": 0.0165, "step": 9850 }, { "grad_norm": 0.22403892874717712, "learning_rate": 5.434406011893822e-08, "loss": 0.016, "step": 9860 }, { "grad_norm": 0.20102401077747345, "learning_rate": 4.6910192341864664e-08, "loss": 0.0139, "step": 9870 }, { "grad_norm": 0.17690938711166382, "learning_rate": 4.0022603551737035e-08, "loss": 0.0149, "step": 9880 }, { "grad_norm": 0.13541299104690552, "learning_rate": 3.3681369070120985e-08, "loss": 0.0157, "step": 9890 }, { "grad_norm": 0.12647123634815216, "learning_rate": 2.7886558243744866e-08, "loss": 0.0145, "step": 9900 }, { "grad_norm": 0.1283811777830124, "learning_rate": 2.2638234443722596e-08, "loss": 0.0166, "step": 9910 }, { "grad_norm": 0.16622725129127502, "learning_rate": 1.7936455064887504e-08, "loss": 0.0186, "step": 9920 }, { "grad_norm": 0.16733846068382263, "learning_rate": 1.378127152514841e-08, "loss": 0.0112, "step": 9930 }, { "grad_norm": 0.16455897688865662, "learning_rate": 1.0172729264917857e-08, "loss": 0.0151, "step": 9940 }, { "grad_norm": 0.2071903944015503, "learning_rate": 7.1108677466458215e-09, "loss": 0.0137, "step": 9950 }, { "grad_norm": 0.17285338044166565, "learning_rate": 4.595720454353414e-09, "loss": 0.0124, "step": 9960 }, { "grad_norm": 0.2841334640979767, "learning_rate": 2.627314893294264e-09, "loss": 0.0145, "step": 9970 }, { "grad_norm": 0.12008201330900192, "learning_rate": 1.2056725896270048e-09, "loss": 0.0115, "step": 9980 }, { "grad_norm": 0.13479606807231903, "learning_rate": 3.308090902098826e-10, "loss": 0.014, "step": 9990 }, { "grad_norm": 0.1615888625383377, "learning_rate": 2.7339624120159555e-12, "loss": 0.0163, "step": 10000 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 4000, "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 }