{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.6884768009185791, "learning_rate": 1.8e-06, "loss": 1.4744, "step": 10 }, { "grad_norm": 0.5435450077056885, "learning_rate": 3.8e-06, "loss": 1.461, "step": 20 }, { "grad_norm": 0.9047264456748962, "learning_rate": 5.8e-06, "loss": 1.3836, "step": 30 }, { "grad_norm": 0.6196959614753723, "learning_rate": 7.8e-06, "loss": 1.2982, "step": 40 }, { "grad_norm": 0.3882925808429718, "learning_rate": 9.800000000000001e-06, "loss": 1.2132, "step": 50 }, { "grad_norm": 0.655003547668457, "learning_rate": 1.18e-05, "loss": 1.1696, "step": 60 }, { "grad_norm": 0.7774730324745178, "learning_rate": 1.3800000000000002e-05, "loss": 1.1367, "step": 70 }, { "grad_norm": 0.60676509141922, "learning_rate": 1.58e-05, "loss": 1.1113, "step": 80 }, { "grad_norm": 0.44719481468200684, "learning_rate": 1.78e-05, "loss": 1.0964, "step": 90 }, { "grad_norm": 0.7986728549003601, "learning_rate": 1.9800000000000004e-05, "loss": 1.0696, "step": 100 }, { "grad_norm": 1.0873409509658813, "learning_rate": 2.18e-05, "loss": 1.0126, "step": 110 }, { "grad_norm": 0.7478275895118713, "learning_rate": 2.38e-05, "loss": 0.913, "step": 120 }, { "grad_norm": 0.8428552150726318, "learning_rate": 2.58e-05, "loss": 0.784, "step": 130 }, { "grad_norm": 1.3388457298278809, "learning_rate": 2.7800000000000005e-05, "loss": 0.6813, "step": 140 }, { "grad_norm": 1.5308880805969238, "learning_rate": 2.98e-05, "loss": 0.5887, "step": 150 }, { "grad_norm": 1.438978672027588, "learning_rate": 3.18e-05, "loss": 0.506, "step": 160 }, { "grad_norm": 1.6501655578613281, "learning_rate": 3.38e-05, "loss": 0.4359, "step": 170 }, { "grad_norm": 1.5194191932678223, "learning_rate": 3.58e-05, "loss": 0.3799, "step": 180 }, { "grad_norm": 1.3467830419540405, "learning_rate": 3.7800000000000004e-05, "loss": 0.3253, "step": 190 }, { "grad_norm": 1.5457572937011719, "learning_rate": 3.9800000000000005e-05, "loss": 0.2776, "step": 200 }, { "grad_norm": 1.8355919122695923, "learning_rate": 4.18e-05, "loss": 0.2358, "step": 210 }, { "grad_norm": 1.1384572982788086, "learning_rate": 4.38e-05, "loss": 0.2177, "step": 220 }, { "grad_norm": 1.7481200695037842, "learning_rate": 4.58e-05, "loss": 0.202, "step": 230 }, { "grad_norm": 1.2863574028015137, "learning_rate": 4.78e-05, "loss": 0.1757, "step": 240 }, { "grad_norm": 1.3261154890060425, "learning_rate": 4.9800000000000004e-05, "loss": 0.1653, "step": 250 }, { "grad_norm": 1.6747429370880127, "learning_rate": 5.1800000000000005e-05, "loss": 0.1567, "step": 260 }, { "grad_norm": 1.1158037185668945, "learning_rate": 5.380000000000001e-05, "loss": 0.1611, "step": 270 }, { "grad_norm": 1.659161925315857, "learning_rate": 5.580000000000001e-05, "loss": 0.1471, "step": 280 }, { "grad_norm": 1.3028548955917358, "learning_rate": 5.7799999999999995e-05, "loss": 0.153, "step": 290 }, { "grad_norm": 1.321920394897461, "learning_rate": 5.9800000000000003e-05, "loss": 0.143, "step": 300 }, { "grad_norm": 1.2718724012374878, "learning_rate": 6.18e-05, "loss": 0.1352, "step": 310 }, { "grad_norm": 1.378176212310791, "learning_rate": 6.38e-05, "loss": 0.1371, "step": 320 }, { "grad_norm": 1.248529314994812, "learning_rate": 6.58e-05, "loss": 0.1353, "step": 330 }, { "grad_norm": 1.1998169422149658, "learning_rate": 6.780000000000001e-05, "loss": 0.1263, "step": 340 }, { "grad_norm": 1.041830062866211, "learning_rate": 6.98e-05, "loss": 0.1225, "step": 350 }, { "grad_norm": 1.228946328163147, "learning_rate": 7.18e-05, "loss": 0.126, "step": 360 }, { "grad_norm": 1.210856556892395, "learning_rate": 7.38e-05, "loss": 0.132, "step": 370 }, { "grad_norm": 1.276041865348816, "learning_rate": 7.58e-05, "loss": 0.1288, "step": 380 }, { "grad_norm": 1.296594500541687, "learning_rate": 7.780000000000001e-05, "loss": 0.1248, "step": 390 }, { "grad_norm": 1.1049566268920898, "learning_rate": 7.98e-05, "loss": 0.1184, "step": 400 }, { "grad_norm": 1.2943700551986694, "learning_rate": 8.18e-05, "loss": 0.1125, "step": 410 }, { "grad_norm": 1.0258386135101318, "learning_rate": 8.38e-05, "loss": 0.1208, "step": 420 }, { "grad_norm": 1.1422423124313354, "learning_rate": 8.58e-05, "loss": 0.1185, "step": 430 }, { "grad_norm": 1.2890939712524414, "learning_rate": 8.78e-05, "loss": 0.1165, "step": 440 }, { "grad_norm": 1.0085821151733398, "learning_rate": 8.98e-05, "loss": 0.1153, "step": 450 }, { "grad_norm": 1.0268089771270752, "learning_rate": 9.180000000000001e-05, "loss": 0.1099, "step": 460 }, { "grad_norm": 0.9510951042175293, "learning_rate": 9.38e-05, "loss": 0.1111, "step": 470 }, { "grad_norm": 1.0179111957550049, "learning_rate": 9.58e-05, "loss": 0.1122, "step": 480 }, { "grad_norm": 1.1109613180160522, "learning_rate": 9.78e-05, "loss": 0.1097, "step": 490 }, { "grad_norm": 1.0083664655685425, "learning_rate": 9.98e-05, "loss": 0.1115, "step": 500 }, { "grad_norm": 0.9720937609672546, "learning_rate": 9.9999778549206e-05, "loss": 0.1075, "step": 510 }, { "grad_norm": 0.9505049586296082, "learning_rate": 9.999901304280685e-05, "loss": 0.101, "step": 520 }, { "grad_norm": 0.7659870982170105, "learning_rate": 9.999770075521164e-05, "loss": 0.0984, "step": 530 }, { "grad_norm": 0.7466753125190735, "learning_rate": 9.99958417007713e-05, "loss": 0.1043, "step": 540 }, { "grad_norm": 0.9827821850776672, "learning_rate": 9.999343589981615e-05, "loss": 0.0982, "step": 550 }, { "grad_norm": 0.9333674311637878, "learning_rate": 9.999048337865568e-05, "loss": 0.0914, "step": 560 }, { "grad_norm": 1.0197436809539795, "learning_rate": 9.998698416957815e-05, "loss": 0.0967, "step": 570 }, { "grad_norm": 0.8936194777488708, "learning_rate": 9.998293831085037e-05, "loss": 0.0915, "step": 580 }, { "grad_norm": 0.9632366895675659, "learning_rate": 9.997834584671719e-05, "loss": 0.095, "step": 590 }, { "grad_norm": 0.885512113571167, "learning_rate": 9.997320682740107e-05, "loss": 0.0932, "step": 600 }, { "grad_norm": 0.9485294222831726, "learning_rate": 9.996752130910149e-05, "loss": 0.0894, "step": 610 }, { "grad_norm": 0.8002691864967346, "learning_rate": 9.99612893539944e-05, "loss": 0.0882, "step": 620 }, { "grad_norm": 0.804033637046814, "learning_rate": 9.995451103023144e-05, "loss": 0.0869, "step": 630 }, { "grad_norm": 0.8636960387229919, "learning_rate": 9.994718641193928e-05, "loss": 0.0882, "step": 640 }, { "grad_norm": 0.9110060930252075, "learning_rate": 9.993931557921874e-05, "loss": 0.0864, "step": 650 }, { "grad_norm": 1.0052390098571777, "learning_rate": 9.993089861814402e-05, "loss": 0.0815, "step": 660 }, { "grad_norm": 0.8719620704650879, "learning_rate": 9.992193562076166e-05, "loss": 0.0869, "step": 670 }, { "grad_norm": 0.8186468482017517, "learning_rate": 9.991242668508954e-05, "loss": 0.082, "step": 680 }, { "grad_norm": 0.809939444065094, "learning_rate": 9.990237191511587e-05, "loss": 0.0778, "step": 690 }, { "grad_norm": 0.8026238083839417, "learning_rate": 9.989177142079802e-05, "loss": 0.0791, "step": 700 }, { "grad_norm": 0.8680374622344971, "learning_rate": 9.988062531806126e-05, "loss": 0.0795, "step": 710 }, { "grad_norm": 0.9117174744606018, "learning_rate": 9.986893372879762e-05, "loss": 0.0776, "step": 720 }, { "grad_norm": 0.6612201929092407, "learning_rate": 9.985669678086443e-05, "loss": 0.0802, "step": 730 }, { "grad_norm": 0.9553166031837463, "learning_rate": 9.984391460808298e-05, "loss": 0.0788, "step": 740 }, { "grad_norm": 0.8051835894584656, "learning_rate": 9.983058735023709e-05, "loss": 0.0769, "step": 750 }, { "grad_norm": 0.7538738250732422, "learning_rate": 9.98167151530715e-05, "loss": 0.0766, "step": 760 }, { "grad_norm": 0.7742618322372437, "learning_rate": 9.980229816829034e-05, "loss": 0.0774, "step": 770 }, { "grad_norm": 0.873926043510437, "learning_rate": 9.978733655355544e-05, "loss": 0.0771, "step": 780 }, { "grad_norm": 0.6845922470092773, "learning_rate": 9.977183047248464e-05, "loss": 0.075, "step": 790 }, { "grad_norm": 0.6262147426605225, "learning_rate": 9.975578009464992e-05, "loss": 0.0792, "step": 800 }, { "grad_norm": 0.6965222358703613, "learning_rate": 9.97391855955757e-05, "loss": 0.0741, "step": 810 }, { "grad_norm": 0.6920332908630371, "learning_rate": 9.972204715673669e-05, "loss": 0.0736, "step": 820 }, { "grad_norm": 0.7292290925979614, "learning_rate": 9.970436496555617e-05, "loss": 0.0678, "step": 830 }, { "grad_norm": 0.7118574380874634, "learning_rate": 9.968613921540373e-05, "loss": 0.069, "step": 840 }, { "grad_norm": 0.7643498778343201, "learning_rate": 9.966737010559326e-05, "loss": 0.0751, "step": 850 }, { "grad_norm": 0.6831203699111938, "learning_rate": 9.964805784138072e-05, "loss": 0.0685, "step": 860 }, { "grad_norm": 0.6466369032859802, "learning_rate": 9.962820263396195e-05, "loss": 0.0681, "step": 870 }, { "grad_norm": 0.7740229964256287, "learning_rate": 9.960780470047033e-05, "loss": 0.0707, "step": 880 }, { "grad_norm": 0.6474797129631042, "learning_rate": 9.958686426397437e-05, "loss": 0.0694, "step": 890 }, { "grad_norm": 0.6896506547927856, "learning_rate": 9.956538155347534e-05, "loss": 0.0677, "step": 900 }, { "grad_norm": 0.7501163482666016, "learning_rate": 9.95433568039047e-05, "loss": 0.0669, "step": 910 }, { "grad_norm": 0.7086248993873596, "learning_rate": 9.952079025612162e-05, "loss": 0.0656, "step": 920 }, { "grad_norm": 0.7878040671348572, "learning_rate": 9.949768215691022e-05, "loss": 0.068, "step": 930 }, { "grad_norm": 0.5746245980262756, "learning_rate": 9.9474032758977e-05, "loss": 0.0656, "step": 940 }, { "grad_norm": 0.6804682612419128, "learning_rate": 9.944984232094794e-05, "loss": 0.0672, "step": 950 }, { "grad_norm": 0.8384517431259155, "learning_rate": 9.942511110736584e-05, "loss": 0.0634, "step": 960 }, { "grad_norm": 0.9291036128997803, "learning_rate": 9.939983938868726e-05, "loss": 0.0655, "step": 970 }, { "grad_norm": 0.7508569955825806, "learning_rate": 9.93740274412797e-05, "loss": 0.0636, "step": 980 }, { "grad_norm": 0.7587980031967163, "learning_rate": 9.934767554741846e-05, "loss": 0.0644, "step": 990 }, { "grad_norm": 0.7202167510986328, "learning_rate": 9.932078399528361e-05, "loss": 0.0641, "step": 1000 }, { "grad_norm": 0.6816259622573853, "learning_rate": 9.929335307895689e-05, "loss": 0.0635, "step": 1010 }, { "grad_norm": 0.6750280857086182, "learning_rate": 9.926538309841839e-05, "loss": 0.0616, "step": 1020 }, { "grad_norm": 0.6307911276817322, "learning_rate": 9.923687435954334e-05, "loss": 0.0643, "step": 1030 }, { "grad_norm": 0.6810320019721985, "learning_rate": 9.920782717409873e-05, "loss": 0.0603, "step": 1040 }, { "grad_norm": 0.6102986335754395, "learning_rate": 9.917824185973994e-05, "loss": 0.0614, "step": 1050 }, { "grad_norm": 0.5722846984863281, "learning_rate": 9.914811874000723e-05, "loss": 0.0566, "step": 1060 }, { "grad_norm": 0.6585894823074341, "learning_rate": 9.911745814432218e-05, "loss": 0.0586, "step": 1070 }, { "grad_norm": 0.45641443133354187, "learning_rate": 9.90862604079842e-05, "loss": 0.0559, "step": 1080 }, { "grad_norm": 0.5736437439918518, "learning_rate": 9.90545258721667e-05, "loss": 0.0541, "step": 1090 }, { "grad_norm": 0.643835186958313, "learning_rate": 9.90222548839135e-05, "loss": 0.0594, "step": 1100 }, { "grad_norm": 0.6161701679229736, "learning_rate": 9.898944779613495e-05, "loss": 0.0597, "step": 1110 }, { "grad_norm": 0.48352810740470886, "learning_rate": 9.89561049676041e-05, "loss": 0.0556, "step": 1120 }, { "grad_norm": 0.6591546535491943, "learning_rate": 9.89222267629528e-05, "loss": 0.0575, "step": 1130 }, { "grad_norm": 0.6489901542663574, "learning_rate": 9.888781355266763e-05, "loss": 0.0534, "step": 1140 }, { "grad_norm": 0.563843309879303, "learning_rate": 9.885286571308598e-05, "loss": 0.0548, "step": 1150 }, { "grad_norm": 0.6797248125076294, "learning_rate": 9.881738362639182e-05, "loss": 0.0565, "step": 1160 }, { "grad_norm": 0.6767581701278687, "learning_rate": 9.878136768061154e-05, "loss": 0.0574, "step": 1170 }, { "grad_norm": 0.623855710029602, "learning_rate": 9.874481826960979e-05, "loss": 0.0579, "step": 1180 }, { "grad_norm": 0.6848495006561279, "learning_rate": 9.870773579308503e-05, "loss": 0.0546, "step": 1190 }, { "grad_norm": 0.7150700092315674, "learning_rate": 9.867012065656533e-05, "loss": 0.0536, "step": 1200 }, { "grad_norm": 0.6190496683120728, "learning_rate": 9.863197327140376e-05, "loss": 0.0577, "step": 1210 }, { "grad_norm": 0.6939300298690796, "learning_rate": 9.859329405477403e-05, "loss": 0.0572, "step": 1220 }, { "grad_norm": 0.6264883875846863, "learning_rate": 9.855408342966585e-05, "loss": 0.053, "step": 1230 }, { "grad_norm": 0.6426411271095276, "learning_rate": 9.851434182488033e-05, "loss": 0.053, "step": 1240 }, { "grad_norm": 0.7734183669090271, "learning_rate": 9.84740696750253e-05, "loss": 0.0571, "step": 1250 }, { "grad_norm": 0.604672908782959, "learning_rate": 9.843326742051055e-05, "loss": 0.0535, "step": 1260 }, { "grad_norm": 0.8048153519630432, "learning_rate": 9.839193550754297e-05, "loss": 0.0515, "step": 1270 }, { "grad_norm": 0.6076591610908508, "learning_rate": 9.835007438812177e-05, "loss": 0.0563, "step": 1280 }, { "grad_norm": 0.7367923855781555, "learning_rate": 9.830768452003341e-05, "loss": 0.0515, "step": 1290 }, { "grad_norm": 0.6216933131217957, "learning_rate": 9.826476636684671e-05, "loss": 0.0501, "step": 1300 }, { "grad_norm": 0.5908995866775513, "learning_rate": 9.822132039790773e-05, "loss": 0.0495, "step": 1310 }, { "grad_norm": 0.5623757839202881, "learning_rate": 9.817734708833461e-05, "loss": 0.0495, "step": 1320 }, { "grad_norm": 0.6854605078697205, "learning_rate": 9.813284691901243e-05, "loss": 0.0513, "step": 1330 }, { "grad_norm": 0.5624321699142456, "learning_rate": 9.808782037658792e-05, "loss": 0.0516, "step": 1340 }, { "grad_norm": 0.5714311003684998, "learning_rate": 9.804226795346411e-05, "loss": 0.0508, "step": 1350 }, { "grad_norm": 0.5043617486953735, "learning_rate": 9.799619014779503e-05, "loss": 0.0484, "step": 1360 }, { "grad_norm": 0.6112223863601685, "learning_rate": 9.794958746348013e-05, "loss": 0.0514, "step": 1370 }, { "grad_norm": 0.7058053612709045, "learning_rate": 9.790246041015896e-05, "loss": 0.0474, "step": 1380 }, { "grad_norm": 0.6250683069229126, "learning_rate": 9.785480950320538e-05, "loss": 0.0496, "step": 1390 }, { "grad_norm": 0.4950709342956543, "learning_rate": 9.78066352637221e-05, "loss": 0.0483, "step": 1400 }, { "grad_norm": 0.5634369850158691, "learning_rate": 9.775793821853488e-05, "loss": 0.0505, "step": 1410 }, { "grad_norm": 0.5947253704071045, "learning_rate": 9.77087189001868e-05, "loss": 0.047, "step": 1420 }, { "grad_norm": 0.5646452903747559, "learning_rate": 9.765897784693243e-05, "loss": 0.051, "step": 1430 }, { "grad_norm": 0.5483761429786682, "learning_rate": 9.760871560273197e-05, "loss": 0.0479, "step": 1440 }, { "grad_norm": 0.5943145751953125, "learning_rate": 9.755793271724526e-05, "loss": 0.0492, "step": 1450 }, { "grad_norm": 0.5690364837646484, "learning_rate": 9.750662974582584e-05, "loss": 0.048, "step": 1460 }, { "grad_norm": 0.6954290270805359, "learning_rate": 9.745480724951473e-05, "loss": 0.0517, "step": 1470 }, { "grad_norm": 0.5464627146720886, "learning_rate": 9.740246579503447e-05, "loss": 0.0482, "step": 1480 }, { "grad_norm": 0.5543454885482788, "learning_rate": 9.734960595478284e-05, "loss": 0.0448, "step": 1490 }, { "grad_norm": 0.7237008810043335, "learning_rate": 9.729622830682657e-05, "loss": 0.0453, "step": 1500 }, { "grad_norm": 0.6331709623336792, "learning_rate": 9.724233343489504e-05, "loss": 0.0459, "step": 1510 }, { "grad_norm": 0.496570348739624, "learning_rate": 9.718792192837396e-05, "loss": 0.0456, "step": 1520 }, { "grad_norm": 0.46912509202957153, "learning_rate": 9.713299438229886e-05, "loss": 0.0453, "step": 1530 }, { "grad_norm": 0.6254875063896179, "learning_rate": 9.707755139734855e-05, "loss": 0.0448, "step": 1540 }, { "grad_norm": 0.518423318862915, "learning_rate": 9.702159357983866e-05, "loss": 0.0482, "step": 1550 }, { "grad_norm": 0.5152536630630493, "learning_rate": 9.696512154171492e-05, "loss": 0.0437, "step": 1560 }, { "grad_norm": 0.5807472467422485, "learning_rate": 9.690813590054645e-05, "loss": 0.0454, "step": 1570 }, { "grad_norm": 0.6077199578285217, "learning_rate": 9.685063727951914e-05, "loss": 0.0429, "step": 1580 }, { "grad_norm": 0.4908650815486908, "learning_rate": 9.679262630742865e-05, "loss": 0.0414, "step": 1590 }, { "grad_norm": 0.5801288485527039, "learning_rate": 9.673410361867373e-05, "loss": 0.0432, "step": 1600 }, { "grad_norm": 0.6001010537147522, "learning_rate": 9.667506985324909e-05, "loss": 0.0441, "step": 1610 }, { "grad_norm": 0.6290978789329529, "learning_rate": 9.661552565673855e-05, "loss": 0.0411, "step": 1620 }, { "grad_norm": 0.5781362652778625, "learning_rate": 9.655547168030789e-05, "loss": 0.044, "step": 1630 }, { "grad_norm": 0.6214280724525452, "learning_rate": 9.649490858069777e-05, "loss": 0.0433, "step": 1640 }, { "grad_norm": 0.5994082689285278, "learning_rate": 9.643383702021658e-05, "loss": 0.0452, "step": 1650 }, { "grad_norm": 0.527083694934845, "learning_rate": 9.637225766673307e-05, "loss": 0.0435, "step": 1660 }, { "grad_norm": 0.6873682141304016, "learning_rate": 9.631017119366922e-05, "loss": 0.0427, "step": 1670 }, { "grad_norm": 0.5397573113441467, "learning_rate": 9.624757827999273e-05, "loss": 0.0429, "step": 1680 }, { "grad_norm": 0.5558175444602966, "learning_rate": 9.618447961020971e-05, "loss": 0.0437, "step": 1690 }, { "grad_norm": 0.6952885985374451, "learning_rate": 9.612087587435707e-05, "loss": 0.0452, "step": 1700 }, { "grad_norm": 0.5797290205955505, "learning_rate": 9.605676776799508e-05, "loss": 0.0401, "step": 1710 }, { "grad_norm": 0.7723774313926697, "learning_rate": 9.599215599219973e-05, "loss": 0.0457, "step": 1720 }, { "grad_norm": 0.6276054382324219, "learning_rate": 9.592704125355505e-05, "loss": 0.045, "step": 1730 }, { "grad_norm": 0.669849157333374, "learning_rate": 9.586142426414538e-05, "loss": 0.042, "step": 1740 }, { "grad_norm": 0.4703032970428467, "learning_rate": 9.57953057415476e-05, "loss": 0.0416, "step": 1750 }, { "grad_norm": 0.4303815960884094, "learning_rate": 9.572868640882328e-05, "loss": 0.041, "step": 1760 }, { "grad_norm": 0.5113366842269897, "learning_rate": 9.56615669945108e-05, "loss": 0.0417, "step": 1770 }, { "grad_norm": 0.42997127771377563, "learning_rate": 9.55939482326173e-05, "loss": 0.0411, "step": 1780 }, { "grad_norm": 0.6484348773956299, "learning_rate": 9.552583086261069e-05, "loss": 0.0418, "step": 1790 }, { "grad_norm": 0.5123947858810425, "learning_rate": 9.545721562941168e-05, "loss": 0.0411, "step": 1800 }, { "grad_norm": 0.5017980337142944, "learning_rate": 9.538810328338543e-05, "loss": 0.0395, "step": 1810 }, { "grad_norm": 0.5054149031639099, "learning_rate": 9.531849458033349e-05, "loss": 0.0406, "step": 1820 }, { "grad_norm": 0.5568170547485352, "learning_rate": 9.524839028148547e-05, "loss": 0.041, "step": 1830 }, { "grad_norm": 0.6279913783073425, "learning_rate": 9.517779115349077e-05, "loss": 0.041, "step": 1840 }, { "grad_norm": 0.6407419443130493, "learning_rate": 9.510669796841014e-05, "loss": 0.0411, "step": 1850 }, { "grad_norm": 0.655894935131073, "learning_rate": 9.503511150370727e-05, "loss": 0.0406, "step": 1860 }, { "grad_norm": 0.5322553515434265, "learning_rate": 9.496303254224024e-05, "loss": 0.0403, "step": 1870 }, { "grad_norm": 0.5445989370346069, "learning_rate": 9.489046187225306e-05, "loss": 0.0401, "step": 1880 }, { "grad_norm": 0.6330538392066956, "learning_rate": 9.481740028736692e-05, "loss": 0.0385, "step": 1890 }, { "grad_norm": 0.529361367225647, "learning_rate": 9.474384858657164e-05, "loss": 0.0416, "step": 1900 }, { "grad_norm": 0.5612173676490784, "learning_rate": 9.466980757421679e-05, "loss": 0.0373, "step": 1910 }, { "grad_norm": 0.46560075879096985, "learning_rate": 9.459527806000305e-05, "loss": 0.0389, "step": 1920 }, { "grad_norm": 0.3968232274055481, "learning_rate": 9.452026085897325e-05, "loss": 0.0394, "step": 1930 }, { "grad_norm": 0.6718124747276306, "learning_rate": 9.444475679150348e-05, "loss": 0.0395, "step": 1940 }, { "grad_norm": 0.6392882466316223, "learning_rate": 9.436876668329411e-05, "loss": 0.0383, "step": 1950 }, { "grad_norm": 0.5263699889183044, "learning_rate": 9.429229136536079e-05, "loss": 0.0445, "step": 1960 }, { "grad_norm": 0.5392703413963318, "learning_rate": 9.421533167402534e-05, "loss": 0.0384, "step": 1970 }, { "grad_norm": 0.357778400182724, "learning_rate": 9.413788845090666e-05, "loss": 0.0391, "step": 1980 }, { "grad_norm": 0.6305422782897949, "learning_rate": 9.405996254291136e-05, "loss": 0.0423, "step": 1990 }, { "grad_norm": 0.5278794765472412, "learning_rate": 9.398155480222474e-05, "loss": 0.0406, "step": 2000 }, { "grad_norm": 0.4771844148635864, "learning_rate": 9.390266608630128e-05, "loss": 0.0382, "step": 2010 }, { "grad_norm": 0.5370548367500305, "learning_rate": 9.38232972578553e-05, "loss": 0.0366, "step": 2020 }, { "grad_norm": 0.44940972328186035, "learning_rate": 9.374344918485164e-05, "loss": 0.0351, "step": 2030 }, { "grad_norm": 0.44597116112709045, "learning_rate": 9.366312274049602e-05, "loss": 0.0351, "step": 2040 }, { "grad_norm": 0.5676587820053101, "learning_rate": 9.358231880322554e-05, "loss": 0.0364, "step": 2050 }, { "grad_norm": 0.6230515837669373, "learning_rate": 9.350103825669916e-05, "loss": 0.0369, "step": 2060 }, { "grad_norm": 0.38392457365989685, "learning_rate": 9.341928198978787e-05, "loss": 0.0377, "step": 2070 }, { "grad_norm": 0.536754846572876, "learning_rate": 9.333705089656512e-05, "loss": 0.0377, "step": 2080 }, { "grad_norm": 0.6331470012664795, "learning_rate": 9.325434587629698e-05, "loss": 0.0367, "step": 2090 }, { "grad_norm": 0.717631459236145, "learning_rate": 9.31711678334323e-05, "loss": 0.035, "step": 2100 }, { "grad_norm": 0.5035995841026306, "learning_rate": 9.308751767759282e-05, "loss": 0.0358, "step": 2110 }, { "grad_norm": 0.43534913659095764, "learning_rate": 9.300339632356325e-05, "loss": 0.0365, "step": 2120 }, { "grad_norm": 0.48784056305885315, "learning_rate": 9.291880469128124e-05, "loss": 0.0369, "step": 2130 }, { "grad_norm": 0.5468835234642029, "learning_rate": 9.283374370582732e-05, "loss": 0.0349, "step": 2140 }, { "grad_norm": 0.6240045428276062, "learning_rate": 9.274821429741482e-05, "loss": 0.0366, "step": 2150 }, { "grad_norm": 0.5389606952667236, "learning_rate": 9.266221740137961e-05, "loss": 0.0351, "step": 2160 }, { "grad_norm": 0.7081879377365112, "learning_rate": 9.257575395817001e-05, "loss": 0.037, "step": 2170 }, { "grad_norm": 0.650422215461731, "learning_rate": 9.248882491333637e-05, "loss": 0.0342, "step": 2180 }, { "grad_norm": 0.5459349155426025, "learning_rate": 9.240143121752076e-05, "loss": 0.0364, "step": 2190 }, { "grad_norm": 0.4988988935947418, "learning_rate": 9.23135738264467e-05, "loss": 0.0372, "step": 2200 }, { "grad_norm": 0.5022599101066589, "learning_rate": 9.222525370090849e-05, "loss": 0.0367, "step": 2210 }, { "grad_norm": 0.48840683698654175, "learning_rate": 9.213647180676088e-05, "loss": 0.0365, "step": 2220 }, { "grad_norm": 0.5591930150985718, "learning_rate": 9.204722911490846e-05, "loss": 0.0339, "step": 2230 }, { "grad_norm": 0.4881974756717682, "learning_rate": 9.1957526601295e-05, "loss": 0.0337, "step": 2240 }, { "grad_norm": 0.46559590101242065, "learning_rate": 9.186736524689281e-05, "loss": 0.0338, "step": 2250 }, { "grad_norm": 0.510321855545044, "learning_rate": 9.177674603769204e-05, "loss": 0.0316, "step": 2260 }, { "grad_norm": 0.5647909641265869, "learning_rate": 9.168566996468983e-05, "loss": 0.0338, "step": 2270 }, { "grad_norm": 0.44700849056243896, "learning_rate": 9.159413802387951e-05, "loss": 0.0352, "step": 2280 }, { "grad_norm": 0.5236690640449524, "learning_rate": 9.150215121623974e-05, "loss": 0.0335, "step": 2290 }, { "grad_norm": 0.637067973613739, "learning_rate": 9.140971054772349e-05, "loss": 0.0361, "step": 2300 }, { "grad_norm": 0.40049421787261963, "learning_rate": 9.131681702924713e-05, "loss": 0.0353, "step": 2310 }, { "grad_norm": 0.4820675253868103, "learning_rate": 9.122347167667926e-05, "loss": 0.0338, "step": 2320 }, { "grad_norm": 0.5388680100440979, "learning_rate": 9.112967551082973e-05, "loss": 0.0337, "step": 2330 }, { "grad_norm": 0.5734873414039612, "learning_rate": 9.103542955743835e-05, "loss": 0.0344, "step": 2340 }, { "grad_norm": 0.4512770473957062, "learning_rate": 9.094073484716381e-05, "loss": 0.0352, "step": 2350 }, { "grad_norm": 0.5278335809707642, "learning_rate": 9.084559241557226e-05, "loss": 0.0351, "step": 2360 }, { "grad_norm": 0.4799845218658447, "learning_rate": 9.075000330312608e-05, "loss": 0.034, "step": 2370 }, { "grad_norm": 0.48106276988983154, "learning_rate": 9.065396855517253e-05, "loss": 0.036, "step": 2380 }, { "grad_norm": 0.4671836495399475, "learning_rate": 9.055748922193219e-05, "loss": 0.036, "step": 2390 }, { "grad_norm": 0.48815926909446716, "learning_rate": 9.046056635848761e-05, "loss": 0.0335, "step": 2400 }, { "grad_norm": 0.46270161867141724, "learning_rate": 9.036320102477169e-05, "loss": 0.0357, "step": 2410 }, { "grad_norm": 0.4139176905155182, "learning_rate": 9.02653942855561e-05, "loss": 0.0337, "step": 2420 }, { "grad_norm": 0.4781225919723511, "learning_rate": 9.016714721043971e-05, "loss": 0.0338, "step": 2430 }, { "grad_norm": 0.5322824120521545, "learning_rate": 9.006846087383675e-05, "loss": 0.034, "step": 2440 }, { "grad_norm": 0.41263505816459656, "learning_rate": 8.996933635496523e-05, "loss": 0.0321, "step": 2450 }, { "grad_norm": 0.43716228008270264, "learning_rate": 8.986977473783498e-05, "loss": 0.0336, "step": 2460 }, { "grad_norm": 0.4647403359413147, "learning_rate": 8.97697771112359e-05, "loss": 0.0328, "step": 2470 }, { "grad_norm": 0.5513673424720764, "learning_rate": 8.966934456872602e-05, "loss": 0.0349, "step": 2480 }, { "grad_norm": 0.4306408762931824, "learning_rate": 8.95684782086195e-05, "loss": 0.0337, "step": 2490 }, { "grad_norm": 0.43048956990242004, "learning_rate": 8.946717913397476e-05, "loss": 0.0322, "step": 2500 }, { "grad_norm": 0.3641330599784851, "learning_rate": 8.93654484525822e-05, "loss": 0.0337, "step": 2510 }, { "grad_norm": 0.5518286228179932, "learning_rate": 8.926328727695226e-05, "loss": 0.0323, "step": 2520 }, { "grad_norm": 0.48069480061531067, "learning_rate": 8.916069672430319e-05, "loss": 0.0317, "step": 2530 }, { "grad_norm": 0.5597518086433411, "learning_rate": 8.905767791654884e-05, "loss": 0.0306, "step": 2540 }, { "grad_norm": 0.39868655800819397, "learning_rate": 8.895423198028638e-05, "loss": 0.0321, "step": 2550 }, { "grad_norm": 0.5725629925727844, "learning_rate": 8.885036004678402e-05, "loss": 0.032, "step": 2560 }, { "grad_norm": 0.4501100182533264, "learning_rate": 8.874606325196857e-05, "loss": 0.0339, "step": 2570 }, { "grad_norm": 0.47274723649024963, "learning_rate": 8.864134273641304e-05, "loss": 0.0327, "step": 2580 }, { "grad_norm": 0.48845234513282776, "learning_rate": 8.853619964532427e-05, "loss": 0.0282, "step": 2590 }, { "grad_norm": 0.560067355632782, "learning_rate": 8.843063512853019e-05, "loss": 0.032, "step": 2600 }, { "grad_norm": 0.46626779437065125, "learning_rate": 8.832465034046749e-05, "loss": 0.0313, "step": 2610 }, { "grad_norm": 0.55399090051651, "learning_rate": 8.821824644016882e-05, "loss": 0.0306, "step": 2620 }, { "grad_norm": 0.5233787894248962, "learning_rate": 8.811142459125019e-05, "loss": 0.031, "step": 2630 }, { "grad_norm": 0.49803784489631653, "learning_rate": 8.800418596189822e-05, "loss": 0.0298, "step": 2640 }, { "grad_norm": 0.5696395039558411, "learning_rate": 8.789653172485737e-05, "loss": 0.0308, "step": 2650 }, { "grad_norm": 0.5385692715644836, "learning_rate": 8.778846305741715e-05, "loss": 0.0298, "step": 2660 }, { "grad_norm": 0.5355715155601501, "learning_rate": 8.767998114139918e-05, "loss": 0.0304, "step": 2670 }, { "grad_norm": 0.45811524987220764, "learning_rate": 8.757108716314429e-05, "loss": 0.029, "step": 2680 }, { "grad_norm": 0.6332003474235535, "learning_rate": 8.746178231349962e-05, "loss": 0.0314, "step": 2690 }, { "grad_norm": 0.47556614875793457, "learning_rate": 8.735206778780549e-05, "loss": 0.0302, "step": 2700 }, { "grad_norm": 0.49704328179359436, "learning_rate": 8.724194478588234e-05, "loss": 0.0316, "step": 2710 }, { "grad_norm": 0.43984368443489075, "learning_rate": 8.713141451201772e-05, "loss": 0.0286, "step": 2720 }, { "grad_norm": 0.5174685120582581, "learning_rate": 8.702047817495295e-05, "loss": 0.0292, "step": 2730 }, { "grad_norm": 0.6484087705612183, "learning_rate": 8.69091369878701e-05, "loss": 0.0307, "step": 2740 }, { "grad_norm": 0.4972013533115387, "learning_rate": 8.679739216837849e-05, "loss": 0.0326, "step": 2750 }, { "grad_norm": 0.4111463129520416, "learning_rate": 8.66852449385016e-05, "loss": 0.0285, "step": 2760 }, { "grad_norm": 0.5287098288536072, "learning_rate": 8.657269652466356e-05, "loss": 0.0307, "step": 2770 }, { "grad_norm": 0.45857205986976624, "learning_rate": 8.645974815767577e-05, "loss": 0.0322, "step": 2780 }, { "grad_norm": 0.4227924644947052, "learning_rate": 8.634640107272351e-05, "loss": 0.0331, "step": 2790 }, { "grad_norm": 0.42045554518699646, "learning_rate": 8.623265650935234e-05, "loss": 0.0324, "step": 2800 }, { "grad_norm": 0.3867752254009247, "learning_rate": 8.611851571145456e-05, "loss": 0.0324, "step": 2810 }, { "grad_norm": 0.45348119735717773, "learning_rate": 8.600397992725566e-05, "loss": 0.0298, "step": 2820 }, { "grad_norm": 0.5112945437431335, "learning_rate": 8.588905040930061e-05, "loss": 0.0296, "step": 2830 }, { "grad_norm": 0.3713921904563904, "learning_rate": 8.577372841444022e-05, "loss": 0.0299, "step": 2840 }, { "grad_norm": 0.47615671157836914, "learning_rate": 8.565801520381736e-05, "loss": 0.0279, "step": 2850 }, { "grad_norm": 0.5480076670646667, "learning_rate": 8.554191204285313e-05, "loss": 0.0299, "step": 2860 }, { "grad_norm": 0.3847907483577728, "learning_rate": 8.542542020123315e-05, "loss": 0.0323, "step": 2870 }, { "grad_norm": 0.4974842965602875, "learning_rate": 8.530854095289347e-05, "loss": 0.0292, "step": 2880 }, { "grad_norm": 0.38240358233451843, "learning_rate": 8.519127557600688e-05, "loss": 0.0299, "step": 2890 }, { "grad_norm": 0.42244869470596313, "learning_rate": 8.507362535296871e-05, "loss": 0.0281, "step": 2900 }, { "grad_norm": 0.515114426612854, "learning_rate": 8.495559157038299e-05, "loss": 0.0293, "step": 2910 }, { "grad_norm": 0.4594149589538574, "learning_rate": 8.483717551904823e-05, "loss": 0.0276, "step": 2920 }, { "grad_norm": 0.38395774364471436, "learning_rate": 8.47183784939434e-05, "loss": 0.029, "step": 2930 }, { "grad_norm": 0.3728574216365814, "learning_rate": 8.459920179421374e-05, "loss": 0.0284, "step": 2940 }, { "grad_norm": 0.509965181350708, "learning_rate": 8.447964672315656e-05, "loss": 0.0273, "step": 2950 }, { "grad_norm": 0.3596603572368622, "learning_rate": 8.435971458820692e-05, "loss": 0.0292, "step": 2960 }, { "grad_norm": 0.5644559264183044, "learning_rate": 8.423940670092345e-05, "loss": 0.0302, "step": 2970 }, { "grad_norm": 0.4705330431461334, "learning_rate": 8.411872437697394e-05, "loss": 0.029, "step": 2980 }, { "grad_norm": 0.536916196346283, "learning_rate": 8.399766893612096e-05, "loss": 0.0271, "step": 2990 }, { "grad_norm": 0.713305652141571, "learning_rate": 8.38762417022074e-05, "loss": 0.0283, "step": 3000 }, { "grad_norm": 0.5698578953742981, "learning_rate": 8.375444400314204e-05, "loss": 0.0299, "step": 3010 }, { "grad_norm": 0.5044066905975342, "learning_rate": 8.3632277170885e-05, "loss": 0.0291, "step": 3020 }, { "grad_norm": 0.4217812716960907, "learning_rate": 8.350974254143318e-05, "loss": 0.0283, "step": 3030 }, { "grad_norm": 0.35898059606552124, "learning_rate": 8.338684145480566e-05, "loss": 0.0295, "step": 3040 }, { "grad_norm": 0.34342560172080994, "learning_rate": 8.326357525502904e-05, "loss": 0.028, "step": 3050 }, { "grad_norm": 0.47930920124053955, "learning_rate": 8.313994529012273e-05, "loss": 0.0294, "step": 3060 }, { "grad_norm": 0.35611844062805176, "learning_rate": 8.301595291208422e-05, "loss": 0.0284, "step": 3070 }, { "grad_norm": 0.5228999853134155, "learning_rate": 8.289159947687427e-05, "loss": 0.0269, "step": 3080 }, { "grad_norm": 0.501822829246521, "learning_rate": 8.276688634440216e-05, "loss": 0.0278, "step": 3090 }, { "grad_norm": 0.4105527400970459, "learning_rate": 8.26418148785107e-05, "loss": 0.0293, "step": 3100 }, { "grad_norm": 0.33322545886039734, "learning_rate": 8.251638644696141e-05, "loss": 0.028, "step": 3110 }, { "grad_norm": 0.4939309358596802, "learning_rate": 8.23906024214195e-05, "loss": 0.0267, "step": 3120 }, { "grad_norm": 0.4345974922180176, "learning_rate": 8.226446417743897e-05, "loss": 0.0258, "step": 3130 }, { "grad_norm": 0.41716814041137695, "learning_rate": 8.213797309444742e-05, "loss": 0.0246, "step": 3140 }, { "grad_norm": 0.4974120855331421, "learning_rate": 8.201113055573105e-05, "loss": 0.0268, "step": 3150 }, { "grad_norm": 0.5228216648101807, "learning_rate": 8.188393794841958e-05, "loss": 0.0266, "step": 3160 }, { "grad_norm": 0.47184276580810547, "learning_rate": 8.175639666347094e-05, "loss": 0.0275, "step": 3170 }, { "grad_norm": 0.5031725764274597, "learning_rate": 8.162850809565623e-05, "loss": 0.0283, "step": 3180 }, { "grad_norm": 0.42637404799461365, "learning_rate": 8.150027364354431e-05, "loss": 0.026, "step": 3190 }, { "grad_norm": 0.4197865426540375, "learning_rate": 8.137169470948662e-05, "loss": 0.0288, "step": 3200 }, { "grad_norm": 0.42720651626586914, "learning_rate": 8.124277269960179e-05, "loss": 0.0288, "step": 3210 }, { "grad_norm": 0.45464473962783813, "learning_rate": 8.111350902376023e-05, "loss": 0.0262, "step": 3220 }, { "grad_norm": 0.3929649591445923, "learning_rate": 8.098390509556883e-05, "loss": 0.0257, "step": 3230 }, { "grad_norm": 0.3582702875137329, "learning_rate": 8.085396233235536e-05, "loss": 0.025, "step": 3240 }, { "grad_norm": 0.4618740379810333, "learning_rate": 8.072368215515306e-05, "loss": 0.0276, "step": 3250 }, { "grad_norm": 0.5441854000091553, "learning_rate": 8.059306598868506e-05, "loss": 0.0275, "step": 3260 }, { "grad_norm": 0.42596638202667236, "learning_rate": 8.046211526134888e-05, "loss": 0.0262, "step": 3270 }, { "grad_norm": 0.40560075640678406, "learning_rate": 8.033083140520065e-05, "loss": 0.028, "step": 3280 }, { "grad_norm": 0.3976554274559021, "learning_rate": 8.019921585593962e-05, "loss": 0.0256, "step": 3290 }, { "grad_norm": 0.3523779809474945, "learning_rate": 8.006727005289232e-05, "loss": 0.0263, "step": 3300 }, { "grad_norm": 0.4528290033340454, "learning_rate": 7.993499543899692e-05, "loss": 0.0274, "step": 3310 }, { "grad_norm": 0.4202211797237396, "learning_rate": 7.980239346078742e-05, "loss": 0.0287, "step": 3320 }, { "grad_norm": 0.47111716866493225, "learning_rate": 7.966946556837778e-05, "loss": 0.0267, "step": 3330 }, { "grad_norm": 0.3857215344905853, "learning_rate": 7.953621321544616e-05, "loss": 0.0263, "step": 3340 }, { "grad_norm": 0.5243465304374695, "learning_rate": 7.940263785921896e-05, "loss": 0.0252, "step": 3350 }, { "grad_norm": 0.39726972579956055, "learning_rate": 7.926874096045482e-05, "loss": 0.0279, "step": 3360 }, { "grad_norm": 0.45503106713294983, "learning_rate": 7.913452398342881e-05, "loss": 0.0258, "step": 3370 }, { "grad_norm": 0.45970919728279114, "learning_rate": 7.89999883959163e-05, "loss": 0.0298, "step": 3380 }, { "grad_norm": 0.5347476601600647, "learning_rate": 7.886513566917687e-05, "loss": 0.0282, "step": 3390 }, { "grad_norm": 0.4340955317020416, "learning_rate": 7.872996727793838e-05, "loss": 0.0265, "step": 3400 }, { "grad_norm": 0.5048775672912598, "learning_rate": 7.859448470038069e-05, "loss": 0.0272, "step": 3410 }, { "grad_norm": 0.44894444942474365, "learning_rate": 7.845868941811956e-05, "loss": 0.0261, "step": 3420 }, { "grad_norm": 0.4693596065044403, "learning_rate": 7.832258291619043e-05, "loss": 0.0255, "step": 3430 }, { "grad_norm": 0.39503955841064453, "learning_rate": 7.81861666830322e-05, "loss": 0.0261, "step": 3440 }, { "grad_norm": 0.5348406434059143, "learning_rate": 7.804944221047097e-05, "loss": 0.0261, "step": 3450 }, { "grad_norm": 0.582147479057312, "learning_rate": 7.791241099370364e-05, "loss": 0.0267, "step": 3460 }, { "grad_norm": 0.36838778853416443, "learning_rate": 7.777507453128163e-05, "loss": 0.0248, "step": 3470 }, { "grad_norm": 0.36497950553894043, "learning_rate": 7.763743432509451e-05, "loss": 0.0257, "step": 3480 }, { "grad_norm": 0.39281174540519714, "learning_rate": 7.749949188035353e-05, "loss": 0.0254, "step": 3490 }, { "grad_norm": 0.41327956318855286, "learning_rate": 7.736124870557516e-05, "loss": 0.0239, "step": 3500 }, { "grad_norm": 0.3645923435688019, "learning_rate": 7.722270631256459e-05, "loss": 0.0256, "step": 3510 }, { "grad_norm": 0.3965131342411041, "learning_rate": 7.708386621639925e-05, "loss": 0.0241, "step": 3520 }, { "grad_norm": 0.40945279598236084, "learning_rate": 7.694472993541219e-05, "loss": 0.024, "step": 3530 }, { "grad_norm": 0.35620400309562683, "learning_rate": 7.680529899117547e-05, "loss": 0.0259, "step": 3540 }, { "grad_norm": 0.4255146086215973, "learning_rate": 7.666557490848358e-05, "loss": 0.025, "step": 3550 }, { "grad_norm": 0.3932982385158539, "learning_rate": 7.65255592153367e-05, "loss": 0.0258, "step": 3560 }, { "grad_norm": 0.42475467920303345, "learning_rate": 7.638525344292402e-05, "loss": 0.0264, "step": 3570 }, { "grad_norm": 0.4374462366104126, "learning_rate": 7.624465912560697e-05, "loss": 0.024, "step": 3580 }, { "grad_norm": 0.2781524658203125, "learning_rate": 7.610377780090249e-05, "loss": 0.0269, "step": 3590 }, { "grad_norm": 0.3377465605735779, "learning_rate": 7.596261100946618e-05, "loss": 0.0254, "step": 3600 }, { "grad_norm": 0.35418862104415894, "learning_rate": 7.582116029507542e-05, "loss": 0.0232, "step": 3610 }, { "grad_norm": 0.3646920621395111, "learning_rate": 7.56794272046126e-05, "loss": 0.0276, "step": 3620 }, { "grad_norm": 0.3237231969833374, "learning_rate": 7.55374132880481e-05, "loss": 0.0234, "step": 3630 }, { "grad_norm": 0.31907379627227783, "learning_rate": 7.539512009842333e-05, "loss": 0.0263, "step": 3640 }, { "grad_norm": 0.3929639756679535, "learning_rate": 7.525254919183382e-05, "loss": 0.0251, "step": 3650 }, { "grad_norm": 0.4671182632446289, "learning_rate": 7.510970212741215e-05, "loss": 0.0258, "step": 3660 }, { "grad_norm": 0.36579927802085876, "learning_rate": 7.496658046731096e-05, "loss": 0.023, "step": 3670 }, { "grad_norm": 0.4392993152141571, "learning_rate": 7.482318577668578e-05, "loss": 0.0268, "step": 3680 }, { "grad_norm": 0.5119951963424683, "learning_rate": 7.467951962367796e-05, "loss": 0.0265, "step": 3690 }, { "grad_norm": 0.4307856261730194, "learning_rate": 7.453558357939755e-05, "loss": 0.0255, "step": 3700 }, { "grad_norm": 0.3914544880390167, "learning_rate": 7.439137921790606e-05, "loss": 0.024, "step": 3710 }, { "grad_norm": 0.4817541837692261, "learning_rate": 7.42469081161993e-05, "loss": 0.024, "step": 3720 }, { "grad_norm": 0.4258405268192291, "learning_rate": 7.410217185419006e-05, "loss": 0.0262, "step": 3730 }, { "grad_norm": 0.4522794187068939, "learning_rate": 7.395717201469095e-05, "loss": 0.0234, "step": 3740 }, { "grad_norm": 0.45494088530540466, "learning_rate": 7.381191018339696e-05, "loss": 0.0235, "step": 3750 }, { "grad_norm": 0.3842923939228058, "learning_rate": 7.36663879488682e-05, "loss": 0.0235, "step": 3760 }, { "grad_norm": 0.40246981382369995, "learning_rate": 7.352060690251254e-05, "loss": 0.0227, "step": 3770 }, { "grad_norm": 0.4384191036224365, "learning_rate": 7.337456863856811e-05, "loss": 0.0256, "step": 3780 }, { "grad_norm": 0.3425196409225464, "learning_rate": 7.3228274754086e-05, "loss": 0.0223, "step": 3790 }, { "grad_norm": 0.4337756633758545, "learning_rate": 7.308172684891267e-05, "loss": 0.0243, "step": 3800 }, { "grad_norm": 0.4420734941959381, "learning_rate": 7.293492652567255e-05, "loss": 0.0233, "step": 3810 }, { "grad_norm": 0.5090165734291077, "learning_rate": 7.278787538975043e-05, "loss": 0.0247, "step": 3820 }, { "grad_norm": 0.3871191143989563, "learning_rate": 7.2640575049274e-05, "loss": 0.0258, "step": 3830 }, { "grad_norm": 0.3648150861263275, "learning_rate": 7.249302711509616e-05, "loss": 0.0242, "step": 3840 }, { "grad_norm": 0.30189186334609985, "learning_rate": 7.23452332007775e-05, "loss": 0.0248, "step": 3850 }, { "grad_norm": 0.4441606104373932, "learning_rate": 7.219719492256858e-05, "loss": 0.0234, "step": 3860 }, { "grad_norm": 0.3776841163635254, "learning_rate": 7.20489138993923e-05, "loss": 0.0231, "step": 3870 }, { "grad_norm": 0.360429972410202, "learning_rate": 7.190039175282614e-05, "loss": 0.0243, "step": 3880 }, { "grad_norm": 0.36907291412353516, "learning_rate": 7.175163010708455e-05, "loss": 0.0254, "step": 3890 }, { "grad_norm": 0.33200153708457947, "learning_rate": 7.1602630589001e-05, "loss": 0.0227, "step": 3900 }, { "grad_norm": 0.3350712060928345, "learning_rate": 7.14533948280104e-05, "loss": 0.021, "step": 3910 }, { "grad_norm": 0.40215158462524414, "learning_rate": 7.130392445613109e-05, "loss": 0.0239, "step": 3920 }, { "grad_norm": 0.3822092115879059, "learning_rate": 7.115422110794711e-05, "loss": 0.0232, "step": 3930 }, { "grad_norm": 0.510712206363678, "learning_rate": 7.100428642059033e-05, "loss": 0.0228, "step": 3940 }, { "grad_norm": 0.5172283053398132, "learning_rate": 7.08541220337224e-05, "loss": 0.0259, "step": 3950 }, { "grad_norm": 0.2803525924682617, "learning_rate": 7.070372958951706e-05, "loss": 0.0236, "step": 3960 }, { "grad_norm": 0.3997623920440674, "learning_rate": 7.055311073264194e-05, "loss": 0.0244, "step": 3970 }, { "grad_norm": 0.427029550075531, "learning_rate": 7.040226711024077e-05, "loss": 0.0242, "step": 3980 }, { "grad_norm": 0.34151220321655273, "learning_rate": 7.02512003719152e-05, "loss": 0.0253, "step": 3990 }, { "grad_norm": 0.42390501499176025, "learning_rate": 7.00999121697069e-05, "loss": 0.0215, "step": 4000 }, { "grad_norm": 0.4360036551952362, "learning_rate": 6.99484041580794e-05, "loss": 0.0209, "step": 4010 }, { "grad_norm": 0.41873621940612793, "learning_rate": 6.979667799390004e-05, "loss": 0.0231, "step": 4020 }, { "grad_norm": 0.41821345686912537, "learning_rate": 6.964473533642185e-05, "loss": 0.023, "step": 4030 }, { "grad_norm": 0.43036431074142456, "learning_rate": 6.949257784726539e-05, "loss": 0.0228, "step": 4040 }, { "grad_norm": 0.3201462924480438, "learning_rate": 6.934020719040056e-05, "loss": 0.0237, "step": 4050 }, { "grad_norm": 0.48180466890335083, "learning_rate": 6.918762503212848e-05, "loss": 0.0223, "step": 4060 }, { "grad_norm": 0.3788181245326996, "learning_rate": 6.903483304106319e-05, "loss": 0.022, "step": 4070 }, { "grad_norm": 0.3761153221130371, "learning_rate": 6.888183288811341e-05, "loss": 0.0228, "step": 4080 }, { "grad_norm": 0.3631659150123596, "learning_rate": 6.87286262464643e-05, "loss": 0.0227, "step": 4090 }, { "grad_norm": 0.3802398443222046, "learning_rate": 6.857521479155915e-05, "loss": 0.023, "step": 4100 }, { "grad_norm": 0.34178775548934937, "learning_rate": 6.842160020108104e-05, "loss": 0.0207, "step": 4110 }, { "grad_norm": 0.3590947389602661, "learning_rate": 6.826778415493455e-05, "loss": 0.0215, "step": 4120 }, { "grad_norm": 0.3021015226840973, "learning_rate": 6.811376833522729e-05, "loss": 0.0222, "step": 4130 }, { "grad_norm": 0.4180317521095276, "learning_rate": 6.795955442625159e-05, "loss": 0.0212, "step": 4140 }, { "grad_norm": 0.33779770135879517, "learning_rate": 6.780514411446608e-05, "loss": 0.022, "step": 4150 }, { "grad_norm": 0.3947369456291199, "learning_rate": 6.765053908847716e-05, "loss": 0.0222, "step": 4160 }, { "grad_norm": 0.3775780200958252, "learning_rate": 6.749574103902064e-05, "loss": 0.0216, "step": 4170 }, { "grad_norm": 0.3824998140335083, "learning_rate": 6.734075165894317e-05, "loss": 0.0219, "step": 4180 }, { "grad_norm": 0.3695768713951111, "learning_rate": 6.71855726431838e-05, "loss": 0.0191, "step": 4190 }, { "grad_norm": 0.39735138416290283, "learning_rate": 6.703020568875538e-05, "loss": 0.0219, "step": 4200 }, { "grad_norm": 0.35551631450653076, "learning_rate": 6.687465249472603e-05, "loss": 0.0223, "step": 4210 }, { "grad_norm": 0.40660127997398376, "learning_rate": 6.671891476220055e-05, "loss": 0.022, "step": 4220 }, { "grad_norm": 0.4183954894542694, "learning_rate": 6.656299419430183e-05, "loss": 0.025, "step": 4230 }, { "grad_norm": 0.3806694746017456, "learning_rate": 6.640689249615223e-05, "loss": 0.0219, "step": 4240 }, { "grad_norm": 0.4369344413280487, "learning_rate": 6.625061137485491e-05, "loss": 0.0235, "step": 4250 }, { "grad_norm": 0.3795107305049896, "learning_rate": 6.609415253947517e-05, "loss": 0.0217, "step": 4260 }, { "grad_norm": 0.42946112155914307, "learning_rate": 6.593751770102178e-05, "loss": 0.023, "step": 4270 }, { "grad_norm": 0.45787689089775085, "learning_rate": 6.578070857242823e-05, "loss": 0.0221, "step": 4280 }, { "grad_norm": 0.3975658118724823, "learning_rate": 6.562372686853402e-05, "loss": 0.0242, "step": 4290 }, { "grad_norm": 0.3673941195011139, "learning_rate": 6.546657430606593e-05, "loss": 0.0206, "step": 4300 }, { "grad_norm": 0.3379276394844055, "learning_rate": 6.530925260361918e-05, "loss": 0.0214, "step": 4310 }, { "grad_norm": 0.4009743928909302, "learning_rate": 6.515176348163871e-05, "loss": 0.0226, "step": 4320 }, { "grad_norm": 0.37054339051246643, "learning_rate": 6.499410866240032e-05, "loss": 0.022, "step": 4330 }, { "grad_norm": 0.3670901656150818, "learning_rate": 6.48362898699919e-05, "loss": 0.0231, "step": 4340 }, { "grad_norm": 0.33396339416503906, "learning_rate": 6.467830883029443e-05, "loss": 0.0185, "step": 4350 }, { "grad_norm": 0.4144164025783539, "learning_rate": 6.452016727096326e-05, "loss": 0.0217, "step": 4360 }, { "grad_norm": 0.46327540278434753, "learning_rate": 6.436186692140916e-05, "loss": 0.0219, "step": 4370 }, { "grad_norm": 0.42627495527267456, "learning_rate": 6.420340951277938e-05, "loss": 0.0204, "step": 4380 }, { "grad_norm": 0.46319931745529175, "learning_rate": 6.404479677793874e-05, "loss": 0.0237, "step": 4390 }, { "grad_norm": 0.48768728971481323, "learning_rate": 6.388603045145075e-05, "loss": 0.0201, "step": 4400 }, { "grad_norm": 0.3541359007358551, "learning_rate": 6.372711226955843e-05, "loss": 0.0208, "step": 4410 }, { "grad_norm": 0.365697979927063, "learning_rate": 6.356804397016564e-05, "loss": 0.0225, "step": 4420 }, { "grad_norm": 0.47155770659446716, "learning_rate": 6.340882729281779e-05, "loss": 0.0223, "step": 4430 }, { "grad_norm": 0.3902604579925537, "learning_rate": 6.324946397868294e-05, "loss": 0.0217, "step": 4440 }, { "grad_norm": 0.4499368369579315, "learning_rate": 6.308995577053276e-05, "loss": 0.0231, "step": 4450 }, { "grad_norm": 0.458696573972702, "learning_rate": 6.293030441272347e-05, "loss": 0.023, "step": 4460 }, { "grad_norm": 0.42503297328948975, "learning_rate": 6.277051165117677e-05, "loss": 0.0243, "step": 4470 }, { "grad_norm": 0.44061288237571716, "learning_rate": 6.261057923336064e-05, "loss": 0.0218, "step": 4480 }, { "grad_norm": 0.34737083315849304, "learning_rate": 6.245050890827042e-05, "loss": 0.0222, "step": 4490 }, { "grad_norm": 0.34495919942855835, "learning_rate": 6.229030242640952e-05, "loss": 0.0231, "step": 4500 }, { "grad_norm": 0.28552642464637756, "learning_rate": 6.212996153977037e-05, "loss": 0.0233, "step": 4510 }, { "grad_norm": 0.2908640205860138, "learning_rate": 6.196948800181523e-05, "loss": 0.022, "step": 4520 }, { "grad_norm": 0.4566967189311981, "learning_rate": 6.180888356745695e-05, "loss": 0.0211, "step": 4530 }, { "grad_norm": 0.38995006680488586, "learning_rate": 6.164814999303995e-05, "loss": 0.0207, "step": 4540 }, { "grad_norm": 0.36497780680656433, "learning_rate": 6.148728903632081e-05, "loss": 0.0205, "step": 4550 }, { "grad_norm": 0.3991524875164032, "learning_rate": 6.132630245644921e-05, "loss": 0.02, "step": 4560 }, { "grad_norm": 0.3814873695373535, "learning_rate": 6.116519201394857e-05, "loss": 0.0193, "step": 4570 }, { "grad_norm": 0.32147884368896484, "learning_rate": 6.10039594706969e-05, "loss": 0.0205, "step": 4580 }, { "grad_norm": 0.3776841461658478, "learning_rate": 6.084260658990744e-05, "loss": 0.0196, "step": 4590 }, { "grad_norm": 0.32612505555152893, "learning_rate": 6.068113513610943e-05, "loss": 0.0206, "step": 4600 }, { "grad_norm": 0.362383097410202, "learning_rate": 6.0519546875128876e-05, "loss": 0.0211, "step": 4610 }, { "grad_norm": 0.32785895466804504, "learning_rate": 6.035784357406906e-05, "loss": 0.0201, "step": 4620 }, { "grad_norm": 0.3206351697444916, "learning_rate": 6.01960270012914e-05, "loss": 0.021, "step": 4630 }, { "grad_norm": 0.39958441257476807, "learning_rate": 6.003409892639599e-05, "loss": 0.0221, "step": 4640 }, { "grad_norm": 0.3972451984882355, "learning_rate": 5.9872061120202336e-05, "loss": 0.021, "step": 4650 }, { "grad_norm": 0.34831133484840393, "learning_rate": 5.9709915354729914e-05, "loss": 0.0191, "step": 4660 }, { "grad_norm": 0.34851184487342834, "learning_rate": 5.9547663403178824e-05, "loss": 0.0214, "step": 4670 }, { "grad_norm": 0.3693978488445282, "learning_rate": 5.9385307039910445e-05, "loss": 0.0205, "step": 4680 }, { "grad_norm": 0.3223094344139099, "learning_rate": 5.922284804042792e-05, "loss": 0.0188, "step": 4690 }, { "grad_norm": 0.3133412301540375, "learning_rate": 5.906028818135687e-05, "loss": 0.0201, "step": 4700 }, { "grad_norm": 0.2581075131893158, "learning_rate": 5.889762924042585e-05, "loss": 0.0194, "step": 4710 }, { "grad_norm": 0.3379593789577484, "learning_rate": 5.873487299644699e-05, "loss": 0.019, "step": 4720 }, { "grad_norm": 0.38564348220825195, "learning_rate": 5.857202122929649e-05, "loss": 0.0203, "step": 4730 }, { "grad_norm": 0.4237644374370575, "learning_rate": 5.840907571989518e-05, "loss": 0.0204, "step": 4740 }, { "grad_norm": 0.39771631360054016, "learning_rate": 5.824603825018904e-05, "loss": 0.0181, "step": 4750 }, { "grad_norm": 0.43610477447509766, "learning_rate": 5.808291060312975e-05, "loss": 0.0209, "step": 4760 }, { "grad_norm": 0.28062424063682556, "learning_rate": 5.7919694562655083e-05, "loss": 0.0198, "step": 4770 }, { "grad_norm": 0.2931273877620697, "learning_rate": 5.775639191366954e-05, "loss": 0.0188, "step": 4780 }, { "grad_norm": 0.35215064883232117, "learning_rate": 5.75930044420247e-05, "loss": 0.0207, "step": 4790 }, { "grad_norm": 0.3879399001598358, "learning_rate": 5.74295339344998e-05, "loss": 0.0199, "step": 4800 }, { "grad_norm": 0.37896856665611267, "learning_rate": 5.726598217878211e-05, "loss": 0.0192, "step": 4810 }, { "grad_norm": 0.3535725474357605, "learning_rate": 5.71023509634474e-05, "loss": 0.0201, "step": 4820 }, { "grad_norm": 0.4016187787055969, "learning_rate": 5.693864207794049e-05, "loss": 0.0181, "step": 4830 }, { "grad_norm": 0.303587943315506, "learning_rate": 5.677485731255545e-05, "loss": 0.0194, "step": 4840 }, { "grad_norm": 0.41879087686538696, "learning_rate": 5.6610998458416296e-05, "loss": 0.0203, "step": 4850 }, { "grad_norm": 0.29371756315231323, "learning_rate": 5.644706730745716e-05, "loss": 0.0199, "step": 4860 }, { "grad_norm": 0.4103497266769409, "learning_rate": 5.628306565240287e-05, "loss": 0.0201, "step": 4870 }, { "grad_norm": 0.37845170497894287, "learning_rate": 5.611899528674923e-05, "loss": 0.0193, "step": 4880 }, { "grad_norm": 0.3922991156578064, "learning_rate": 5.595485800474349e-05, "loss": 0.0202, "step": 4890 }, { "grad_norm": 0.26364538073539734, "learning_rate": 5.579065560136467e-05, "loss": 0.0195, "step": 4900 }, { "grad_norm": 0.34728729724884033, "learning_rate": 5.562638987230392e-05, "loss": 0.017, "step": 4910 }, { "grad_norm": 0.3298253118991852, "learning_rate": 5.546206261394498e-05, "loss": 0.0184, "step": 4920 }, { "grad_norm": 0.32188817858695984, "learning_rate": 5.529767562334437e-05, "loss": 0.02, "step": 4930 }, { "grad_norm": 0.272776335477829, "learning_rate": 5.5133230698211926e-05, "loss": 0.0192, "step": 4940 }, { "grad_norm": 0.41054901480674744, "learning_rate": 5.496872963689096e-05, "loss": 0.02, "step": 4950 }, { "grad_norm": 0.3224942982196808, "learning_rate": 5.4804174238338756e-05, "loss": 0.0205, "step": 4960 }, { "grad_norm": 0.3163520097732544, "learning_rate": 5.463956630210678e-05, "loss": 0.0202, "step": 4970 }, { "grad_norm": 0.37926673889160156, "learning_rate": 5.4474907628321046e-05, "loss": 0.0189, "step": 4980 }, { "grad_norm": 0.2396094799041748, "learning_rate": 5.431020001766244e-05, "loss": 0.0179, "step": 4990 }, { "grad_norm": 0.3139626681804657, "learning_rate": 5.4145445271346986e-05, "loss": 0.018, "step": 5000 }, { "grad_norm": 0.3266680836677551, "learning_rate": 5.398064519110622e-05, "loss": 0.0176, "step": 5010 }, { "grad_norm": 0.3144661486148834, "learning_rate": 5.3815801579167394e-05, "loss": 0.0179, "step": 5020 }, { "grad_norm": 0.23234283924102783, "learning_rate": 5.365091623823382e-05, "loss": 0.0173, "step": 5030 }, { "grad_norm": 0.36290574073791504, "learning_rate": 5.348599097146521e-05, "loss": 0.0194, "step": 5040 }, { "grad_norm": 0.31773337721824646, "learning_rate": 5.3321027582457836e-05, "loss": 0.0203, "step": 5050 }, { "grad_norm": 0.3832032084465027, "learning_rate": 5.315602787522491e-05, "loss": 0.0182, "step": 5060 }, { "grad_norm": 0.37287789583206177, "learning_rate": 5.299099365417678e-05, "loss": 0.0175, "step": 5070 }, { "grad_norm": 0.41710200905799866, "learning_rate": 5.2825926724101236e-05, "loss": 0.017, "step": 5080 }, { "grad_norm": 0.34887146949768066, "learning_rate": 5.26608288901438e-05, "loss": 0.0199, "step": 5090 }, { "grad_norm": 0.3933245539665222, "learning_rate": 5.24957019577879e-05, "loss": 0.017, "step": 5100 }, { "grad_norm": 0.43839412927627563, "learning_rate": 5.2330547732835266e-05, "loss": 0.0187, "step": 5110 }, { "grad_norm": 0.27346116304397583, "learning_rate": 5.2165368021385996e-05, "loss": 0.0172, "step": 5120 }, { "grad_norm": 0.347281813621521, "learning_rate": 5.200016462981897e-05, "loss": 0.0175, "step": 5130 }, { "grad_norm": 0.3731026351451874, "learning_rate": 5.1834939364772015e-05, "loss": 0.0188, "step": 5140 }, { "grad_norm": 0.32215848565101624, "learning_rate": 5.166969403312214e-05, "loss": 0.019, "step": 5150 }, { "grad_norm": 0.32148465514183044, "learning_rate": 5.1504430441965844e-05, "loss": 0.0165, "step": 5160 }, { "grad_norm": 0.33479586243629456, "learning_rate": 5.133915039859923e-05, "loss": 0.0184, "step": 5170 }, { "grad_norm": 0.3004407286643982, "learning_rate": 5.1173855710498444e-05, "loss": 0.0174, "step": 5180 }, { "grad_norm": 0.3273955285549164, "learning_rate": 5.100854818529967e-05, "loss": 0.0168, "step": 5190 }, { "grad_norm": 0.3490319550037384, "learning_rate": 5.084322963077951e-05, "loss": 0.0195, "step": 5200 }, { "grad_norm": 0.2791002094745636, "learning_rate": 5.067790185483522e-05, "loss": 0.0177, "step": 5210 }, { "grad_norm": 0.3798407018184662, "learning_rate": 5.0512566665464844e-05, "loss": 0.0159, "step": 5220 }, { "grad_norm": 0.3125234842300415, "learning_rate": 5.034722587074755e-05, "loss": 0.0173, "step": 5230 }, { "grad_norm": 0.27127504348754883, "learning_rate": 5.018188127882375e-05, "loss": 0.0167, "step": 5240 }, { "grad_norm": 0.2915577292442322, "learning_rate": 5.0016534697875417e-05, "loss": 0.0196, "step": 5250 }, { "grad_norm": 0.3667474091053009, "learning_rate": 4.9851187936106294e-05, "loss": 0.016, "step": 5260 }, { "grad_norm": 0.3036137521266937, "learning_rate": 4.968584280172206e-05, "loss": 0.0192, "step": 5270 }, { "grad_norm": 0.3458300828933716, "learning_rate": 4.95205011029106e-05, "loss": 0.0185, "step": 5280 }, { "grad_norm": 0.2940741181373596, "learning_rate": 4.935516464782227e-05, "loss": 0.0175, "step": 5290 }, { "grad_norm": 0.3159944415092468, "learning_rate": 4.918983524455003e-05, "loss": 0.0169, "step": 5300 }, { "grad_norm": 0.3169877231121063, "learning_rate": 4.9024514701109766e-05, "loss": 0.0175, "step": 5310 }, { "grad_norm": 0.31846627593040466, "learning_rate": 4.885920482542043e-05, "loss": 0.0199, "step": 5320 }, { "grad_norm": 0.36222904920578003, "learning_rate": 4.869390742528438e-05, "loss": 0.0171, "step": 5330 }, { "grad_norm": 0.31586378812789917, "learning_rate": 4.852862430836744e-05, "loss": 0.0175, "step": 5340 }, { "grad_norm": 0.3819626271724701, "learning_rate": 4.836335728217933e-05, "loss": 0.0182, "step": 5350 }, { "grad_norm": 0.34948858618736267, "learning_rate": 4.819810815405379e-05, "loss": 0.0179, "step": 5360 }, { "grad_norm": 0.28954964876174927, "learning_rate": 4.803287873112877e-05, "loss": 0.0186, "step": 5370 }, { "grad_norm": 0.29904448986053467, "learning_rate": 4.786767082032681e-05, "loss": 0.0178, "step": 5380 }, { "grad_norm": 0.28503966331481934, "learning_rate": 4.77024862283351e-05, "loss": 0.0156, "step": 5390 }, { "grad_norm": 0.27454105019569397, "learning_rate": 4.753732676158593e-05, "loss": 0.0174, "step": 5400 }, { "grad_norm": 0.21128273010253906, "learning_rate": 4.737219422623672e-05, "loss": 0.0172, "step": 5410 }, { "grad_norm": 0.38476717472076416, "learning_rate": 4.720709042815044e-05, "loss": 0.016, "step": 5420 }, { "grad_norm": 0.33306175470352173, "learning_rate": 4.704201717287578e-05, "loss": 0.0173, "step": 5430 }, { "grad_norm": 0.28761449456214905, "learning_rate": 4.6876976265627404e-05, "loss": 0.0169, "step": 5440 }, { "grad_norm": 0.306238055229187, "learning_rate": 4.671196951126626e-05, "loss": 0.0161, "step": 5450 }, { "grad_norm": 0.3490322232246399, "learning_rate": 4.654699871427971e-05, "loss": 0.0166, "step": 5460 }, { "grad_norm": 0.4287564754486084, "learning_rate": 4.6382065678762034e-05, "loss": 0.0173, "step": 5470 }, { "grad_norm": 0.32195451855659485, "learning_rate": 4.6217172208394424e-05, "loss": 0.0156, "step": 5480 }, { "grad_norm": 0.24442841112613678, "learning_rate": 4.605232010642549e-05, "loss": 0.0174, "step": 5490 }, { "grad_norm": 0.2617293894290924, "learning_rate": 4.588751117565142e-05, "loss": 0.0167, "step": 5500 }, { "grad_norm": 0.41320738196372986, "learning_rate": 4.5722747218396214e-05, "loss": 0.0171, "step": 5510 }, { "grad_norm": 0.4450415372848511, "learning_rate": 4.5558030036492194e-05, "loss": 0.0166, "step": 5520 }, { "grad_norm": 0.29307112097740173, "learning_rate": 4.539336143125999e-05, "loss": 0.0183, "step": 5530 }, { "grad_norm": 0.40844473242759705, "learning_rate": 4.522874320348916e-05, "loss": 0.0163, "step": 5540 }, { "grad_norm": 0.2830883860588074, "learning_rate": 4.506417715341821e-05, "loss": 0.0165, "step": 5550 }, { "grad_norm": 0.29526257514953613, "learning_rate": 4.489966508071511e-05, "loss": 0.0168, "step": 5560 }, { "grad_norm": 0.43474912643432617, "learning_rate": 4.4735208784457575e-05, "loss": 0.0167, "step": 5570 }, { "grad_norm": 0.2804695963859558, "learning_rate": 4.457081006311325e-05, "loss": 0.0159, "step": 5580 }, { "grad_norm": 0.31013211607933044, "learning_rate": 4.440647071452027e-05, "loss": 0.0164, "step": 5590 }, { "grad_norm": 0.26033687591552734, "learning_rate": 4.424219253586737e-05, "loss": 0.0183, "step": 5600 }, { "grad_norm": 0.2758661210536957, "learning_rate": 4.407797732367443e-05, "loss": 0.0159, "step": 5610 }, { "grad_norm": 0.27171555161476135, "learning_rate": 4.391382687377268e-05, "loss": 0.0186, "step": 5620 }, { "grad_norm": 0.24751637876033783, "learning_rate": 4.374974298128512e-05, "loss": 0.0153, "step": 5630 }, { "grad_norm": 0.2765371799468994, "learning_rate": 4.358572744060699e-05, "loss": 0.015, "step": 5640 }, { "grad_norm": 0.2294846922159195, "learning_rate": 4.342178204538588e-05, "loss": 0.0151, "step": 5650 }, { "grad_norm": 0.26203376054763794, "learning_rate": 4.325790858850241e-05, "loss": 0.0154, "step": 5660 }, { "grad_norm": 0.3525017201900482, "learning_rate": 4.309410886205043e-05, "loss": 0.0164, "step": 5670 }, { "grad_norm": 0.2888972759246826, "learning_rate": 4.293038465731752e-05, "loss": 0.0171, "step": 5680 }, { "grad_norm": 0.34767410159111023, "learning_rate": 4.276673776476533e-05, "loss": 0.0174, "step": 5690 }, { "grad_norm": 0.3591817319393158, "learning_rate": 4.260316997401007e-05, "loss": 0.0169, "step": 5700 }, { "grad_norm": 0.3320533037185669, "learning_rate": 4.243968307380293e-05, "loss": 0.0167, "step": 5710 }, { "grad_norm": 0.26119765639305115, "learning_rate": 4.22762788520104e-05, "loss": 0.015, "step": 5720 }, { "grad_norm": 0.26788055896759033, "learning_rate": 4.211295909559491e-05, "loss": 0.0166, "step": 5730 }, { "grad_norm": 0.235127791762352, "learning_rate": 4.194972559059511e-05, "loss": 0.0149, "step": 5740 }, { "grad_norm": 0.269861102104187, "learning_rate": 4.178658012210651e-05, "loss": 0.016, "step": 5750 }, { "grad_norm": 0.3097539246082306, "learning_rate": 4.162352447426177e-05, "loss": 0.0168, "step": 5760 }, { "grad_norm": 0.33859768509864807, "learning_rate": 4.146056043021135e-05, "loss": 0.0184, "step": 5770 }, { "grad_norm": 0.34026962518692017, "learning_rate": 4.1297689772103944e-05, "loss": 0.0171, "step": 5780 }, { "grad_norm": 0.36599791049957275, "learning_rate": 4.113491428106694e-05, "loss": 0.0165, "step": 5790 }, { "grad_norm": 0.33167728781700134, "learning_rate": 4.0972235737187055e-05, "loss": 0.0161, "step": 5800 }, { "grad_norm": 0.2664805054664612, "learning_rate": 4.080965591949076e-05, "loss": 0.0186, "step": 5810 }, { "grad_norm": 0.33038291335105896, "learning_rate": 4.0647176605924924e-05, "loss": 0.0183, "step": 5820 }, { "grad_norm": 0.32955172657966614, "learning_rate": 4.0484799573337255e-05, "loss": 0.0159, "step": 5830 }, { "grad_norm": 0.30267274379730225, "learning_rate": 4.032252659745699e-05, "loss": 0.0155, "step": 5840 }, { "grad_norm": 0.3740771412849426, "learning_rate": 4.016035945287539e-05, "loss": 0.0171, "step": 5850 }, { "grad_norm": 0.3034801781177521, "learning_rate": 3.999829991302635e-05, "loss": 0.0154, "step": 5860 }, { "grad_norm": 0.29304537177085876, "learning_rate": 3.983634975016707e-05, "loss": 0.0163, "step": 5870 }, { "grad_norm": 0.24175730347633362, "learning_rate": 3.967451073535854e-05, "loss": 0.016, "step": 5880 }, { "grad_norm": 0.31063467264175415, "learning_rate": 3.951278463844633e-05, "loss": 0.0144, "step": 5890 }, { "grad_norm": 0.23879215121269226, "learning_rate": 3.935117322804111e-05, "loss": 0.0169, "step": 5900 }, { "grad_norm": 0.2856917083263397, "learning_rate": 3.918967827149938e-05, "loss": 0.0163, "step": 5910 }, { "grad_norm": 0.34703442454338074, "learning_rate": 3.9028301534904094e-05, "loss": 0.0164, "step": 5920 }, { "grad_norm": 0.294552743434906, "learning_rate": 3.88670447830454e-05, "loss": 0.0151, "step": 5930 }, { "grad_norm": 0.31474730372428894, "learning_rate": 3.870590977940132e-05, "loss": 0.0171, "step": 5940 }, { "grad_norm": 0.27652087807655334, "learning_rate": 3.8544898286118404e-05, "loss": 0.0155, "step": 5950 }, { "grad_norm": 0.2814768850803375, "learning_rate": 3.838401206399257e-05, "loss": 0.0156, "step": 5960 }, { "grad_norm": 0.32818129658699036, "learning_rate": 3.822325287244975e-05, "loss": 0.0193, "step": 5970 }, { "grad_norm": 0.35848745703697205, "learning_rate": 3.8062622469526725e-05, "loss": 0.0164, "step": 5980 }, { "grad_norm": 0.5209701657295227, "learning_rate": 3.790212261185183e-05, "loss": 0.0162, "step": 5990 }, { "grad_norm": 0.35661599040031433, "learning_rate": 3.7741755054625794e-05, "loss": 0.0163, "step": 6000 }, { "grad_norm": 0.24497860670089722, "learning_rate": 3.758152155160255e-05, "loss": 0.0162, "step": 6010 }, { "grad_norm": 0.2971506416797638, "learning_rate": 3.742142385506999e-05, "loss": 0.0144, "step": 6020 }, { "grad_norm": 0.3219272792339325, "learning_rate": 3.72614637158309e-05, "loss": 0.0154, "step": 6030 }, { "grad_norm": 0.40087923407554626, "learning_rate": 3.710164288318371e-05, "loss": 0.0145, "step": 6040 }, { "grad_norm": 0.33105772733688354, "learning_rate": 3.694196310490345e-05, "loss": 0.0152, "step": 6050 }, { "grad_norm": 0.2773398756980896, "learning_rate": 3.678242612722259e-05, "loss": 0.0155, "step": 6060 }, { "grad_norm": 0.2388211339712143, "learning_rate": 3.6623033694811953e-05, "loss": 0.0149, "step": 6070 }, { "grad_norm": 0.20104274153709412, "learning_rate": 3.6463787550761665e-05, "loss": 0.0148, "step": 6080 }, { "grad_norm": 0.3465093970298767, "learning_rate": 3.630468943656202e-05, "loss": 0.0153, "step": 6090 }, { "grad_norm": 0.36019083857536316, "learning_rate": 3.6145741092084523e-05, "loss": 0.0156, "step": 6100 }, { "grad_norm": 0.28671571612358093, "learning_rate": 3.598694425556278e-05, "loss": 0.0157, "step": 6110 }, { "grad_norm": 0.3652366101741791, "learning_rate": 3.58283006635736e-05, "loss": 0.0159, "step": 6120 }, { "grad_norm": 0.27596062421798706, "learning_rate": 3.566981205101781e-05, "loss": 0.0137, "step": 6130 }, { "grad_norm": 0.22577039897441864, "learning_rate": 3.5511480151101556e-05, "loss": 0.0153, "step": 6140 }, { "grad_norm": 0.2608244717121124, "learning_rate": 3.5353306695317104e-05, "loss": 0.0148, "step": 6150 }, { "grad_norm": 0.22657227516174316, "learning_rate": 3.519529341342402e-05, "loss": 0.015, "step": 6160 }, { "grad_norm": 0.22019316256046295, "learning_rate": 3.503744203343026e-05, "loss": 0.0145, "step": 6170 }, { "grad_norm": 0.3199502229690552, "learning_rate": 3.487975428157318e-05, "loss": 0.0156, "step": 6180 }, { "grad_norm": 0.2646932899951935, "learning_rate": 3.472223188230083e-05, "loss": 0.0143, "step": 6190 }, { "grad_norm": 0.3268091380596161, "learning_rate": 3.4564876558252866e-05, "loss": 0.0145, "step": 6200 }, { "grad_norm": 0.33983325958251953, "learning_rate": 3.440769003024195e-05, "loss": 0.0141, "step": 6210 }, { "grad_norm": 0.33610665798187256, "learning_rate": 3.425067401723477e-05, "loss": 0.0159, "step": 6220 }, { "grad_norm": 0.21606391668319702, "learning_rate": 3.409383023633325e-05, "loss": 0.0136, "step": 6230 }, { "grad_norm": 0.3107696771621704, "learning_rate": 3.3937160402755894e-05, "loss": 0.0151, "step": 6240 }, { "grad_norm": 0.2496861219406128, "learning_rate": 3.378066622981885e-05, "loss": 0.0155, "step": 6250 }, { "grad_norm": 0.2572075128555298, "learning_rate": 3.362434942891738e-05, "loss": 0.0143, "step": 6260 }, { "grad_norm": 0.2679691016674042, "learning_rate": 3.346821170950693e-05, "loss": 0.0136, "step": 6270 }, { "grad_norm": 0.31219303607940674, "learning_rate": 3.3312254779084585e-05, "loss": 0.0154, "step": 6280 }, { "grad_norm": 0.24941672384738922, "learning_rate": 3.315648034317039e-05, "loss": 0.0161, "step": 6290 }, { "grad_norm": 0.23058953881263733, "learning_rate": 3.3000890105288564e-05, "loss": 0.0133, "step": 6300 }, { "grad_norm": 0.2573181688785553, "learning_rate": 3.284548576694908e-05, "loss": 0.0127, "step": 6310 }, { "grad_norm": 0.22785234451293945, "learning_rate": 3.2690269027628815e-05, "loss": 0.0148, "step": 6320 }, { "grad_norm": 0.23109222948551178, "learning_rate": 3.253524158475324e-05, "loss": 0.0136, "step": 6330 }, { "grad_norm": 0.26725316047668457, "learning_rate": 3.238040513367757e-05, "loss": 0.0137, "step": 6340 }, { "grad_norm": 0.23967809975147247, "learning_rate": 3.222576136766843e-05, "loss": 0.015, "step": 6350 }, { "grad_norm": 0.25730836391448975, "learning_rate": 3.2071311977885324e-05, "loss": 0.0157, "step": 6360 }, { "grad_norm": 0.20204809308052063, "learning_rate": 3.191705865336197e-05, "loss": 0.0148, "step": 6370 }, { "grad_norm": 0.239023819565773, "learning_rate": 3.1763003080988075e-05, "loss": 0.0158, "step": 6380 }, { "grad_norm": 0.2214677482843399, "learning_rate": 3.160914694549063e-05, "loss": 0.0148, "step": 6390 }, { "grad_norm": 0.2470880150794983, "learning_rate": 3.145549192941573e-05, "loss": 0.0154, "step": 6400 }, { "grad_norm": 0.3535122573375702, "learning_rate": 3.130203971310999e-05, "loss": 0.0145, "step": 6410 }, { "grad_norm": 0.2641960382461548, "learning_rate": 3.114879197470225e-05, "loss": 0.0141, "step": 6420 }, { "grad_norm": 0.33345454931259155, "learning_rate": 3.0995750390085285e-05, "loss": 0.0151, "step": 6430 }, { "grad_norm": 0.2601906955242157, "learning_rate": 3.084291663289728e-05, "loss": 0.0139, "step": 6440 }, { "grad_norm": 0.29147717356681824, "learning_rate": 3.069029237450375e-05, "loss": 0.0136, "step": 6450 }, { "grad_norm": 0.2313554435968399, "learning_rate": 3.053787928397911e-05, "loss": 0.0137, "step": 6460 }, { "grad_norm": 0.2394963949918747, "learning_rate": 3.0385679028088526e-05, "loss": 0.0143, "step": 6470 }, { "grad_norm": 0.2498079389333725, "learning_rate": 3.023369327126959e-05, "loss": 0.0137, "step": 6480 }, { "grad_norm": 0.2616010904312134, "learning_rate": 3.0081923675614198e-05, "loss": 0.0132, "step": 6490 }, { "grad_norm": 0.21279749274253845, "learning_rate": 2.993037190085034e-05, "loss": 0.0148, "step": 6500 }, { "grad_norm": 0.26042068004608154, "learning_rate": 2.977903960432392e-05, "loss": 0.014, "step": 6510 }, { "grad_norm": 0.2625766098499298, "learning_rate": 2.9627928440980722e-05, "loss": 0.0131, "step": 6520 }, { "grad_norm": 0.22035294771194458, "learning_rate": 2.9477040063348183e-05, "loss": 0.015, "step": 6530 }, { "grad_norm": 0.27604538202285767, "learning_rate": 2.9326376121517456e-05, "loss": 0.013, "step": 6540 }, { "grad_norm": 0.26744845509529114, "learning_rate": 2.9175938263125236e-05, "loss": 0.0142, "step": 6550 }, { "grad_norm": 0.19801142811775208, "learning_rate": 2.9025728133335873e-05, "loss": 0.0142, "step": 6560 }, { "grad_norm": 0.33632710576057434, "learning_rate": 2.8875747374823288e-05, "loss": 0.0136, "step": 6570 }, { "grad_norm": 0.32133644819259644, "learning_rate": 2.872599762775298e-05, "loss": 0.0128, "step": 6580 }, { "grad_norm": 0.2646152377128601, "learning_rate": 2.857648052976425e-05, "loss": 0.0138, "step": 6590 }, { "grad_norm": 0.22432151436805725, "learning_rate": 2.8427197715952047e-05, "loss": 0.013, "step": 6600 }, { "grad_norm": 0.2753683030605316, "learning_rate": 2.8278150818849393e-05, "loss": 0.0133, "step": 6610 }, { "grad_norm": 0.23748110234737396, "learning_rate": 2.812934146840922e-05, "loss": 0.013, "step": 6620 }, { "grad_norm": 0.23313763737678528, "learning_rate": 2.7980771291986764e-05, "loss": 0.0144, "step": 6630 }, { "grad_norm": 0.24659840762615204, "learning_rate": 2.783244191432167e-05, "loss": 0.0138, "step": 6640 }, { "grad_norm": 0.24917156994342804, "learning_rate": 2.768435495752022e-05, "loss": 0.0149, "step": 6650 }, { "grad_norm": 0.2637504041194916, "learning_rate": 2.753651204103771e-05, "loss": 0.015, "step": 6660 }, { "grad_norm": 0.21695435047149658, "learning_rate": 2.7388914781660523e-05, "loss": 0.0133, "step": 6670 }, { "grad_norm": 0.23899517953395844, "learning_rate": 2.7241564793488693e-05, "loss": 0.0142, "step": 6680 }, { "grad_norm": 0.23502898216247559, "learning_rate": 2.7094463687918037e-05, "loss": 0.0126, "step": 6690 }, { "grad_norm": 0.20827923715114594, "learning_rate": 2.694761307362268e-05, "loss": 0.0137, "step": 6700 }, { "grad_norm": 0.24863681197166443, "learning_rate": 2.6801014556537467e-05, "loss": 0.0144, "step": 6710 }, { "grad_norm": 0.30441343784332275, "learning_rate": 2.6654669739840243e-05, "loss": 0.015, "step": 6720 }, { "grad_norm": 0.2653089165687561, "learning_rate": 2.650858022393451e-05, "loss": 0.013, "step": 6730 }, { "grad_norm": 0.26582351326942444, "learning_rate": 2.6362747606431747e-05, "loss": 0.0143, "step": 6740 }, { "grad_norm": 0.2640269994735718, "learning_rate": 2.6217173482134172e-05, "loss": 0.0138, "step": 6750 }, { "grad_norm": 0.2718997001647949, "learning_rate": 2.6071859443017044e-05, "loss": 0.013, "step": 6760 }, { "grad_norm": 0.23644915223121643, "learning_rate": 2.5926807078211414e-05, "loss": 0.0133, "step": 6770 }, { "grad_norm": 0.24741601943969727, "learning_rate": 2.5782017973986728e-05, "loss": 0.0145, "step": 6780 }, { "grad_norm": 0.3438808023929596, "learning_rate": 2.5637493713733374e-05, "loss": 0.0139, "step": 6790 }, { "grad_norm": 0.24648793041706085, "learning_rate": 2.549323587794559e-05, "loss": 0.0145, "step": 6800 }, { "grad_norm": 0.2806674838066101, "learning_rate": 2.5349246044203895e-05, "loss": 0.0124, "step": 6810 }, { "grad_norm": 0.2633695900440216, "learning_rate": 2.520552578715808e-05, "loss": 0.0131, "step": 6820 }, { "grad_norm": 0.27935320138931274, "learning_rate": 2.506207667850981e-05, "loss": 0.0137, "step": 6830 }, { "grad_norm": 0.2817527651786804, "learning_rate": 2.4918900286995555e-05, "loss": 0.0141, "step": 6840 }, { "grad_norm": 0.23134270310401917, "learning_rate": 2.4775998178369458e-05, "loss": 0.0139, "step": 6850 }, { "grad_norm": 0.34125643968582153, "learning_rate": 2.4633371915386017e-05, "loss": 0.0131, "step": 6860 }, { "grad_norm": 0.2661719024181366, "learning_rate": 2.4491023057783235e-05, "loss": 0.0141, "step": 6870 }, { "grad_norm": 0.2387714684009552, "learning_rate": 2.4348953162265375e-05, "loss": 0.0152, "step": 6880 }, { "grad_norm": 0.21910713613033295, "learning_rate": 2.420716378248607e-05, "loss": 0.0114, "step": 6890 }, { "grad_norm": 0.24722081422805786, "learning_rate": 2.4065656469031266e-05, "loss": 0.012, "step": 6900 }, { "grad_norm": 0.2326001226902008, "learning_rate": 2.3924432769402268e-05, "loss": 0.0121, "step": 6910 }, { "grad_norm": 0.18272905051708221, "learning_rate": 2.3783494227998844e-05, "loss": 0.0142, "step": 6920 }, { "grad_norm": 0.25469258427619934, "learning_rate": 2.3642842386102264e-05, "loss": 0.0116, "step": 6930 }, { "grad_norm": 0.19339416921138763, "learning_rate": 2.3502478781858567e-05, "loss": 0.0143, "step": 6940 }, { "grad_norm": 0.22971296310424805, "learning_rate": 2.3362404950261628e-05, "loss": 0.0118, "step": 6950 }, { "grad_norm": 0.19808506965637207, "learning_rate": 2.3222622423136458e-05, "loss": 0.0133, "step": 6960 }, { "grad_norm": 0.23411384224891663, "learning_rate": 2.3083132729122332e-05, "loss": 0.0114, "step": 6970 }, { "grad_norm": 0.2048424482345581, "learning_rate": 2.294393739365621e-05, "loss": 0.0131, "step": 6980 }, { "grad_norm": 0.22214441001415253, "learning_rate": 2.2805037938956e-05, "loss": 0.0136, "step": 6990 }, { "grad_norm": 0.16321013867855072, "learning_rate": 2.266643588400386e-05, "loss": 0.0114, "step": 7000 }, { "grad_norm": 0.22840356826782227, "learning_rate": 2.252813274452969e-05, "loss": 0.0135, "step": 7010 }, { "grad_norm": 0.17697709798812866, "learning_rate": 2.2390130032994427e-05, "loss": 0.0112, "step": 7020 }, { "grad_norm": 0.2366112321615219, "learning_rate": 2.2252429258573633e-05, "loss": 0.0132, "step": 7030 }, { "grad_norm": 0.18608340620994568, "learning_rate": 2.2115031927140904e-05, "loss": 0.0109, "step": 7040 }, { "grad_norm": 0.22012066841125488, "learning_rate": 2.1977939541251463e-05, "loss": 0.012, "step": 7050 }, { "grad_norm": 0.23690955340862274, "learning_rate": 2.1841153600125684e-05, "loss": 0.0125, "step": 7060 }, { "grad_norm": 0.24370461702346802, "learning_rate": 2.170467559963267e-05, "loss": 0.0122, "step": 7070 }, { "grad_norm": 0.2025620937347412, "learning_rate": 2.1568507032273982e-05, "loss": 0.0118, "step": 7080 }, { "grad_norm": 0.1817142367362976, "learning_rate": 2.1432649387167264e-05, "loss": 0.0127, "step": 7090 }, { "grad_norm": 0.2520084083080292, "learning_rate": 2.1297104150029973e-05, "loss": 0.0127, "step": 7100 }, { "grad_norm": 0.19251765310764313, "learning_rate": 2.116187280316307e-05, "loss": 0.0131, "step": 7110 }, { "grad_norm": 0.24097765982151031, "learning_rate": 2.1026956825434908e-05, "loss": 0.0128, "step": 7120 }, { "grad_norm": 0.2637593448162079, "learning_rate": 2.0892357692265017e-05, "loss": 0.0125, "step": 7130 }, { "grad_norm": 0.22581739723682404, "learning_rate": 2.0758076875607947e-05, "loss": 0.0118, "step": 7140 }, { "grad_norm": 0.1722562164068222, "learning_rate": 2.0624115843937207e-05, "loss": 0.0116, "step": 7150 }, { "grad_norm": 0.15994882583618164, "learning_rate": 2.0490476062229157e-05, "loss": 0.0132, "step": 7160 }, { "grad_norm": 0.16781404614448547, "learning_rate": 2.035715899194704e-05, "loss": 0.0118, "step": 7170 }, { "grad_norm": 0.20209604501724243, "learning_rate": 2.022416609102499e-05, "loss": 0.0111, "step": 7180 }, { "grad_norm": 0.21168842911720276, "learning_rate": 2.009149881385205e-05, "loss": 0.0116, "step": 7190 }, { "grad_norm": 0.16990382969379425, "learning_rate": 1.995915861125634e-05, "loss": 0.0122, "step": 7200 }, { "grad_norm": 0.1730867177248001, "learning_rate": 1.9827146930489065e-05, "loss": 0.0101, "step": 7210 }, { "grad_norm": 0.2187972515821457, "learning_rate": 1.9695465215208848e-05, "loss": 0.0123, "step": 7220 }, { "grad_norm": 0.21004584431648254, "learning_rate": 1.9564114905465813e-05, "loss": 0.012, "step": 7230 }, { "grad_norm": 0.17119227349758148, "learning_rate": 1.9433097437685936e-05, "loss": 0.011, "step": 7240 }, { "grad_norm": 0.14043289422988892, "learning_rate": 1.930241424465521e-05, "loss": 0.0118, "step": 7250 }, { "grad_norm": 0.20318934321403503, "learning_rate": 1.9172066755504115e-05, "loss": 0.012, "step": 7260 }, { "grad_norm": 0.16159075498580933, "learning_rate": 1.9042056395691914e-05, "loss": 0.0111, "step": 7270 }, { "grad_norm": 0.21174514293670654, "learning_rate": 1.8912384586991066e-05, "loss": 0.0135, "step": 7280 }, { "grad_norm": 0.23087285459041595, "learning_rate": 1.8783052747471717e-05, "loss": 0.011, "step": 7290 }, { "grad_norm": 0.314473420381546, "learning_rate": 1.865406229148611e-05, "loss": 0.0121, "step": 7300 }, { "grad_norm": 0.22004060447216034, "learning_rate": 1.8525414629653233e-05, "loss": 0.0126, "step": 7310 }, { "grad_norm": 0.16314730048179626, "learning_rate": 1.8397111168843255e-05, "loss": 0.0117, "step": 7320 }, { "grad_norm": 0.23096422851085663, "learning_rate": 1.8269153312162323e-05, "loss": 0.0115, "step": 7330 }, { "grad_norm": 0.19100229442119598, "learning_rate": 1.8141542458937054e-05, "loss": 0.012, "step": 7340 }, { "grad_norm": 0.18090316653251648, "learning_rate": 1.8014280004699268e-05, "loss": 0.0118, "step": 7350 }, { "grad_norm": 0.18095086514949799, "learning_rate": 1.788736734117078e-05, "loss": 0.0113, "step": 7360 }, { "grad_norm": 0.1991056650876999, "learning_rate": 1.7760805856248152e-05, "loss": 0.0118, "step": 7370 }, { "grad_norm": 0.24725079536437988, "learning_rate": 1.7634596933987518e-05, "loss": 0.0117, "step": 7380 }, { "grad_norm": 0.1726284921169281, "learning_rate": 1.7508741954589404e-05, "loss": 0.0111, "step": 7390 }, { "grad_norm": 0.20586198568344116, "learning_rate": 1.7383242294383717e-05, "loss": 0.0121, "step": 7400 }, { "grad_norm": 0.1874207854270935, "learning_rate": 1.7258099325814632e-05, "loss": 0.0111, "step": 7410 }, { "grad_norm": 0.18513351678848267, "learning_rate": 1.7133314417425594e-05, "loss": 0.0121, "step": 7420 }, { "grad_norm": 0.19456328451633453, "learning_rate": 1.7008888933844408e-05, "loss": 0.0105, "step": 7430 }, { "grad_norm": 0.2279425859451294, "learning_rate": 1.6884824235768172e-05, "loss": 0.012, "step": 7440 }, { "grad_norm": 0.22582899034023285, "learning_rate": 1.6761121679948592e-05, "loss": 0.0121, "step": 7450 }, { "grad_norm": 0.23311367630958557, "learning_rate": 1.663778261917695e-05, "loss": 0.0122, "step": 7460 }, { "grad_norm": 0.17170944809913635, "learning_rate": 1.651480840226952e-05, "loss": 0.0126, "step": 7470 }, { "grad_norm": 0.24613086879253387, "learning_rate": 1.639220037405258e-05, "loss": 0.0103, "step": 7480 }, { "grad_norm": 0.15609629452228546, "learning_rate": 1.6269959875347906e-05, "loss": 0.0116, "step": 7490 }, { "grad_norm": 0.18583336472511292, "learning_rate": 1.614808824295802e-05, "loss": 0.0112, "step": 7500 }, { "grad_norm": 0.16003058850765228, "learning_rate": 1.602658680965152e-05, "loss": 0.0126, "step": 7510 }, { "grad_norm": 0.19989557564258575, "learning_rate": 1.5905456904148686e-05, "loss": 0.0101, "step": 7520 }, { "grad_norm": 0.1978452056646347, "learning_rate": 1.57846998511067e-05, "loss": 0.0121, "step": 7530 }, { "grad_norm": 0.2141692191362381, "learning_rate": 1.566431697110538e-05, "loss": 0.0135, "step": 7540 }, { "grad_norm": 0.1725344955921173, "learning_rate": 1.554430958063259e-05, "loss": 0.0133, "step": 7550 }, { "grad_norm": 0.2010943591594696, "learning_rate": 1.5424678992069912e-05, "loss": 0.0121, "step": 7560 }, { "grad_norm": 0.24755266308784485, "learning_rate": 1.5305426513678362e-05, "loss": 0.0111, "step": 7570 }, { "grad_norm": 0.23008891940116882, "learning_rate": 1.518655344958388e-05, "loss": 0.0103, "step": 7580 }, { "grad_norm": 0.18001632392406464, "learning_rate": 1.5068061099763275e-05, "loss": 0.0111, "step": 7590 }, { "grad_norm": 0.17482762038707733, "learning_rate": 1.494995076002988e-05, "loss": 0.0121, "step": 7600 }, { "grad_norm": 0.18560713529586792, "learning_rate": 1.4832223722019456e-05, "loss": 0.011, "step": 7610 }, { "grad_norm": 0.21285954117774963, "learning_rate": 1.4714881273176035e-05, "loss": 0.0119, "step": 7620 }, { "grad_norm": 0.2147040069103241, "learning_rate": 1.4597924696737835e-05, "loss": 0.0113, "step": 7630 }, { "grad_norm": 0.19684305787086487, "learning_rate": 1.4481355271723252e-05, "loss": 0.0128, "step": 7640 }, { "grad_norm": 0.19444303214550018, "learning_rate": 1.4365174272916809e-05, "loss": 0.0128, "step": 7650 }, { "grad_norm": 0.18290269374847412, "learning_rate": 1.4249382970855319e-05, "loss": 0.0097, "step": 7660 }, { "grad_norm": 0.1847521960735321, "learning_rate": 1.4133982631813903e-05, "loss": 0.0107, "step": 7670 }, { "grad_norm": 0.16351696848869324, "learning_rate": 1.4018974517792194e-05, "loss": 0.0116, "step": 7680 }, { "grad_norm": 0.28729915618896484, "learning_rate": 1.390435988650048e-05, "loss": 0.0119, "step": 7690 }, { "grad_norm": 0.21513687074184418, "learning_rate": 1.3790139991346006e-05, "loss": 0.0121, "step": 7700 }, { "grad_norm": 0.2532550096511841, "learning_rate": 1.367631608141926e-05, "loss": 0.0115, "step": 7710 }, { "grad_norm": 0.1791357547044754, "learning_rate": 1.3562889401480278e-05, "loss": 0.0113, "step": 7720 }, { "grad_norm": 0.17590001225471497, "learning_rate": 1.3449861191945074e-05, "loss": 0.0102, "step": 7730 }, { "grad_norm": 0.1789025366306305, "learning_rate": 1.3337232688872009e-05, "loss": 0.011, "step": 7740 }, { "grad_norm": 0.1950041800737381, "learning_rate": 1.3225005123948364e-05, "loss": 0.0125, "step": 7750 }, { "grad_norm": 0.1787525713443756, "learning_rate": 1.311317972447681e-05, "loss": 0.0115, "step": 7760 }, { "grad_norm": 0.20550596714019775, "learning_rate": 1.3001757713361996e-05, "loss": 0.011, "step": 7770 }, { "grad_norm": 0.19651085138320923, "learning_rate": 1.2890740309097204e-05, "loss": 0.0117, "step": 7780 }, { "grad_norm": 0.2014620155096054, "learning_rate": 1.2780128725750944e-05, "loss": 0.0122, "step": 7790 }, { "grad_norm": 0.25769302248954773, "learning_rate": 1.266992417295379e-05, "loss": 0.0099, "step": 7800 }, { "grad_norm": 0.18175539374351501, "learning_rate": 1.2560127855885073e-05, "loss": 0.01, "step": 7810 }, { "grad_norm": 0.20194993913173676, "learning_rate": 1.2450740975259745e-05, "loss": 0.0099, "step": 7820 }, { "grad_norm": 0.2262062430381775, "learning_rate": 1.234176472731517e-05, "loss": 0.011, "step": 7830 }, { "grad_norm": 0.21755559742450714, "learning_rate": 1.2233200303798158e-05, "loss": 0.01, "step": 7840 }, { "grad_norm": 0.1692352294921875, "learning_rate": 1.2125048891951846e-05, "loss": 0.0099, "step": 7850 }, { "grad_norm": 0.16326846182346344, "learning_rate": 1.2017311674502745e-05, "loss": 0.011, "step": 7860 }, { "grad_norm": 0.2140989452600479, "learning_rate": 1.1909989829647822e-05, "loss": 0.0108, "step": 7870 }, { "grad_norm": 0.20244278013706207, "learning_rate": 1.1803084531041553e-05, "loss": 0.0115, "step": 7880 }, { "grad_norm": 0.26726818084716797, "learning_rate": 1.1696596947783162e-05, "loss": 0.011, "step": 7890 }, { "grad_norm": 0.20713664591312408, "learning_rate": 1.1590528244403803e-05, "loss": 0.0123, "step": 7900 }, { "grad_norm": 0.20319293439388275, "learning_rate": 1.148487958085382e-05, "loss": 0.0106, "step": 7910 }, { "grad_norm": 0.16766761243343353, "learning_rate": 1.1379652112490086e-05, "loss": 0.0111, "step": 7920 }, { "grad_norm": 0.16615155339241028, "learning_rate": 1.1274846990063315e-05, "loss": 0.0126, "step": 7930 }, { "grad_norm": 0.21019206941127777, "learning_rate": 1.117046535970554e-05, "loss": 0.0105, "step": 7940 }, { "grad_norm": 0.17922504246234894, "learning_rate": 1.106650836291755e-05, "loss": 0.0114, "step": 7950 }, { "grad_norm": 0.19980747997760773, "learning_rate": 1.0962977136556418e-05, "loss": 0.0103, "step": 7960 }, { "grad_norm": 0.1286524385213852, "learning_rate": 1.0859872812823024e-05, "loss": 0.0104, "step": 7970 }, { "grad_norm": 0.22386574745178223, "learning_rate": 1.0757196519249747e-05, "loss": 0.0104, "step": 7980 }, { "grad_norm": 0.21618032455444336, "learning_rate": 1.0654949378688077e-05, "loss": 0.0113, "step": 7990 }, { "grad_norm": 0.23711183667182922, "learning_rate": 1.0553132509296376e-05, "loss": 0.0104, "step": 8000 }, { "grad_norm": 0.2142760157585144, "learning_rate": 1.0451747024527613e-05, "loss": 0.0102, "step": 8010 }, { "grad_norm": 0.17201751470565796, "learning_rate": 1.0350794033117189e-05, "loss": 0.01, "step": 8020 }, { "grad_norm": 0.17084836959838867, "learning_rate": 1.0250274639070856e-05, "loss": 0.0108, "step": 8030 }, { "grad_norm": 0.17862458527088165, "learning_rate": 1.0150189941652599e-05, "loss": 0.0106, "step": 8040 }, { "grad_norm": 0.22199933230876923, "learning_rate": 1.0050541035372635e-05, "loss": 0.0106, "step": 8050 }, { "grad_norm": 0.16829898953437805, "learning_rate": 9.951329009975458e-06, "loss": 0.0102, "step": 8060 }, { "grad_norm": 0.13257305324077606, "learning_rate": 9.852554950427845e-06, "loss": 0.0105, "step": 8070 }, { "grad_norm": 0.17554251849651337, "learning_rate": 9.754219936907105e-06, "loss": 0.0105, "step": 8080 }, { "grad_norm": 0.19697682559490204, "learning_rate": 9.656325044789194e-06, "loss": 0.0102, "step": 8090 }, { "grad_norm": 0.17778280377388, "learning_rate": 9.55887134463697e-06, "loss": 0.0115, "step": 8100 }, { "grad_norm": 0.21413202583789825, "learning_rate": 9.461859902188475e-06, "loss": 0.011, "step": 8110 }, { "grad_norm": 0.1668155938386917, "learning_rate": 9.365291778345303e-06, "loss": 0.0095, "step": 8120 }, { "grad_norm": 0.16499124467372894, "learning_rate": 9.269168029160991e-06, "loss": 0.0111, "step": 8130 }, { "grad_norm": 0.16271968185901642, "learning_rate": 9.173489705829447e-06, "loss": 0.0113, "step": 8140 }, { "grad_norm": 0.20346985757350922, "learning_rate": 9.078257854673516e-06, "loss": 0.0131, "step": 8150 }, { "grad_norm": 0.19347979128360748, "learning_rate": 8.983473517133429e-06, "loss": 0.01, "step": 8160 }, { "grad_norm": 0.18587663769721985, "learning_rate": 8.889137729755537e-06, "loss": 0.0107, "step": 8170 }, { "grad_norm": 0.1593473255634308, "learning_rate": 8.79525152418087e-06, "loss": 0.01, "step": 8180 }, { "grad_norm": 0.1668710559606552, "learning_rate": 8.701815927133961e-06, "loss": 0.0093, "step": 8190 }, { "grad_norm": 0.19755819439888, "learning_rate": 8.608831960411534e-06, "loss": 0.0111, "step": 8200 }, { "grad_norm": 0.19110320508480072, "learning_rate": 8.516300640871321e-06, "loss": 0.0096, "step": 8210 }, { "grad_norm": 0.16933202743530273, "learning_rate": 8.424222980421038e-06, "loss": 0.0105, "step": 8220 }, { "grad_norm": 0.1578066498041153, "learning_rate": 8.332599986007184e-06, "loss": 0.0096, "step": 8230 }, { "grad_norm": 0.1904328167438507, "learning_rate": 8.241432659604203e-06, "loss": 0.0104, "step": 8240 }, { "grad_norm": 0.15530110895633698, "learning_rate": 8.150721998203331e-06, "loss": 0.0099, "step": 8250 }, { "grad_norm": 0.16483114659786224, "learning_rate": 8.06046899380184e-06, "loss": 0.0097, "step": 8260 }, { "grad_norm": 0.14825624227523804, "learning_rate": 7.970674633392133e-06, "loss": 0.0096, "step": 8270 }, { "grad_norm": 0.17839965224266052, "learning_rate": 7.881339898950924e-06, "loss": 0.0096, "step": 8280 }, { "grad_norm": 0.17693009972572327, "learning_rate": 7.792465767428597e-06, "loss": 0.0101, "step": 8290 }, { "grad_norm": 0.16681832075119019, "learning_rate": 7.704053210738376e-06, "loss": 0.0118, "step": 8300 }, { "grad_norm": 0.17237776517868042, "learning_rate": 7.6161031957458494e-06, "loss": 0.0092, "step": 8310 }, { "grad_norm": 0.21229785680770874, "learning_rate": 7.5286166842582605e-06, "loss": 0.0107, "step": 8320 }, { "grad_norm": 0.2030494511127472, "learning_rate": 7.4415946330140814e-06, "loss": 0.0098, "step": 8330 }, { "grad_norm": 0.1648033857345581, "learning_rate": 7.3550379936725644e-06, "loss": 0.0107, "step": 8340 }, { "grad_norm": 0.15091152489185333, "learning_rate": 7.2689477128032035e-06, "loss": 0.0114, "step": 8350 }, { "grad_norm": 0.2017657458782196, "learning_rate": 7.183324731875551e-06, "loss": 0.0101, "step": 8360 }, { "grad_norm": 0.16651184856891632, "learning_rate": 7.098169987248782e-06, "loss": 0.0093, "step": 8370 }, { "grad_norm": 0.15628720819950104, "learning_rate": 7.013484410161553e-06, "loss": 0.0107, "step": 8380 }, { "grad_norm": 0.13926927745342255, "learning_rate": 6.92926892672176e-06, "loss": 0.0096, "step": 8390 }, { "grad_norm": 0.13018466532230377, "learning_rate": 6.845524457896446e-06, "loss": 0.0103, "step": 8400 }, { "grad_norm": 0.15522430837154388, "learning_rate": 6.7622519195017165e-06, "loss": 0.0113, "step": 8410 }, { "grad_norm": 0.1361219733953476, "learning_rate": 6.679452222192684e-06, "loss": 0.01, "step": 8420 }, { "grad_norm": 0.17699995636940002, "learning_rate": 6.597126271453579e-06, "loss": 0.0098, "step": 8430 }, { "grad_norm": 0.1929134875535965, "learning_rate": 6.51527496758782e-06, "loss": 0.0097, "step": 8440 }, { "grad_norm": 0.18266186118125916, "learning_rate": 6.433899205708155e-06, "loss": 0.0106, "step": 8450 }, { "grad_norm": 0.1484767347574234, "learning_rate": 6.352999875726856e-06, "loss": 0.0093, "step": 8460 }, { "grad_norm": 0.18770654499530792, "learning_rate": 6.272577862346052e-06, "loss": 0.0088, "step": 8470 }, { "grad_norm": 0.14020554721355438, "learning_rate": 6.192634045047996e-06, "loss": 0.0095, "step": 8480 }, { "grad_norm": 0.14608430862426758, "learning_rate": 6.113169298085458e-06, "loss": 0.009, "step": 8490 }, { "grad_norm": 0.16182179749011993, "learning_rate": 6.034184490472195e-06, "loss": 0.0107, "step": 8500 }, { "grad_norm": 0.14036253094673157, "learning_rate": 5.955680485973386e-06, "loss": 0.0102, "step": 8510 }, { "grad_norm": 0.11773405969142914, "learning_rate": 5.877658143096265e-06, "loss": 0.0093, "step": 8520 }, { "grad_norm": 0.17549166083335876, "learning_rate": 5.800118315080661e-06, "loss": 0.0097, "step": 8530 }, { "grad_norm": 0.20212304592132568, "learning_rate": 5.723061849889716e-06, "loss": 0.0095, "step": 8540 }, { "grad_norm": 0.20131273567676544, "learning_rate": 5.646489590200604e-06, "loss": 0.0097, "step": 8550 }, { "grad_norm": 0.15936528146266937, "learning_rate": 5.570402373395256e-06, "loss": 0.0098, "step": 8560 }, { "grad_norm": 0.18682752549648285, "learning_rate": 5.494801031551305e-06, "loss": 0.0101, "step": 8570 }, { "grad_norm": 0.16506889462471008, "learning_rate": 5.41968639143291e-06, "loss": 0.01, "step": 8580 }, { "grad_norm": 0.13726378977298737, "learning_rate": 5.345059274481751e-06, "loss": 0.01, "step": 8590 }, { "grad_norm": 0.15286356210708618, "learning_rate": 5.270920496808002e-06, "loss": 0.0092, "step": 8600 }, { "grad_norm": 0.17122718691825867, "learning_rate": 5.1972708691814695e-06, "loss": 0.011, "step": 8610 }, { "grad_norm": 0.14659050107002258, "learning_rate": 5.124111197022674e-06, "loss": 0.0094, "step": 8620 }, { "grad_norm": 0.1481146663427353, "learning_rate": 5.051442280394081e-06, "loss": 0.0098, "step": 8630 }, { "grad_norm": 0.15521688759326935, "learning_rate": 4.979264913991322e-06, "loss": 0.0089, "step": 8640 }, { "grad_norm": 0.1377221941947937, "learning_rate": 4.907579887134489e-06, "loss": 0.0097, "step": 8650 }, { "grad_norm": 0.12472935765981674, "learning_rate": 4.836387983759572e-06, "loss": 0.0102, "step": 8660 }, { "grad_norm": 0.1450311541557312, "learning_rate": 4.765689982409816e-06, "loss": 0.0091, "step": 8670 }, { "grad_norm": 0.1343529373407364, "learning_rate": 4.695486656227233e-06, "loss": 0.0093, "step": 8680 }, { "grad_norm": 0.13973450660705566, "learning_rate": 4.625778772944156e-06, "loss": 0.0099, "step": 8690 }, { "grad_norm": 0.14927607774734497, "learning_rate": 4.556567094874825e-06, "loss": 0.0116, "step": 8700 }, { "grad_norm": 0.12080688029527664, "learning_rate": 4.487852378907059e-06, "loss": 0.0087, "step": 8710 }, { "grad_norm": 0.10517770051956177, "learning_rate": 4.419635376493986e-06, "loss": 0.0083, "step": 8720 }, { "grad_norm": 0.17983846366405487, "learning_rate": 4.351916833645825e-06, "loss": 0.0095, "step": 8730 }, { "grad_norm": 0.13668473064899445, "learning_rate": 4.284697490921691e-06, "loss": 0.0097, "step": 8740 }, { "grad_norm": 0.1187635287642479, "learning_rate": 4.2179780834215585e-06, "loss": 0.0095, "step": 8750 }, { "grad_norm": 0.15661875903606415, "learning_rate": 4.151759340778178e-06, "loss": 0.0095, "step": 8760 }, { "grad_norm": 0.16183887422084808, "learning_rate": 4.086041987149109e-06, "loss": 0.0106, "step": 8770 }, { "grad_norm": 0.11340538412332535, "learning_rate": 4.020826741208811e-06, "loss": 0.0102, "step": 8780 }, { "grad_norm": 0.18345151841640472, "learning_rate": 3.956114316140746e-06, "loss": 0.0106, "step": 8790 }, { "grad_norm": 0.15115945041179657, "learning_rate": 3.891905419629643e-06, "loss": 0.01, "step": 8800 }, { "grad_norm": 0.13596104085445404, "learning_rate": 3.8282007538536946e-06, "loss": 0.0082, "step": 8810 }, { "grad_norm": 0.14496482908725739, "learning_rate": 3.7650010154769265e-06, "loss": 0.011, "step": 8820 }, { "grad_norm": 0.17712140083312988, "learning_rate": 3.7023068956415608e-06, "loss": 0.0104, "step": 8830 }, { "grad_norm": 0.1323084533214569, "learning_rate": 3.6401190799604303e-06, "loss": 0.0086, "step": 8840 }, { "grad_norm": 0.1280461996793747, "learning_rate": 3.578438248509536e-06, "loss": 0.0103, "step": 8850 }, { "grad_norm": 0.13297851383686066, "learning_rate": 3.5172650758205583e-06, "loss": 0.0084, "step": 8860 }, { "grad_norm": 0.09793499112129211, "learning_rate": 3.45660023087353e-06, "loss": 0.0088, "step": 8870 }, { "grad_norm": 0.13547968864440918, "learning_rate": 3.3964443770894528e-06, "loss": 0.0083, "step": 8880 }, { "grad_norm": 0.11533573269844055, "learning_rate": 3.3367981723231245e-06, "loss": 0.0089, "step": 8890 }, { "grad_norm": 0.09506341069936752, "learning_rate": 3.2776622688558746e-06, "loss": 0.0093, "step": 8900 }, { "grad_norm": 0.15880908071994781, "learning_rate": 3.2190373133884677e-06, "loss": 0.01, "step": 8910 }, { "grad_norm": 0.13512764871120453, "learning_rate": 3.1609239470340446e-06, "loss": 0.0093, "step": 8920 }, { "grad_norm": 0.1267048567533493, "learning_rate": 3.1033228053110373e-06, "loss": 0.0094, "step": 8930 }, { "grad_norm": 0.09866093844175339, "learning_rate": 3.0462345181363314e-06, "loss": 0.0091, "step": 8940 }, { "grad_norm": 0.11662023514509201, "learning_rate": 2.9896597098182654e-06, "loss": 0.0087, "step": 8950 }, { "grad_norm": 0.1487470418214798, "learning_rate": 2.933598999049891e-06, "loss": 0.0092, "step": 8960 }, { "grad_norm": 0.09962930530309677, "learning_rate": 2.8780529989021697e-06, "loss": 0.0091, "step": 8970 }, { "grad_norm": 0.10952994972467422, "learning_rate": 2.823022316817242e-06, "loss": 0.0086, "step": 8980 }, { "grad_norm": 0.11238310486078262, "learning_rate": 2.7685075546018456e-06, "loss": 0.0089, "step": 8990 }, { "grad_norm": 0.11257796734571457, "learning_rate": 2.7145093084206598e-06, "loss": 0.0087, "step": 9000 }, { "grad_norm": 0.11522279679775238, "learning_rate": 2.661028168789892e-06, "loss": 0.009, "step": 9010 }, { "grad_norm": 0.11765235662460327, "learning_rate": 2.6080647205706855e-06, "loss": 0.0095, "step": 9020 }, { "grad_norm": 0.09889334440231323, "learning_rate": 2.555619542962834e-06, "loss": 0.0087, "step": 9030 }, { "grad_norm": 0.12180913239717484, "learning_rate": 2.503693209498409e-06, "loss": 0.0095, "step": 9040 }, { "grad_norm": 0.09369821846485138, "learning_rate": 2.452286288035449e-06, "loss": 0.0089, "step": 9050 }, { "grad_norm": 0.11615947633981705, "learning_rate": 2.4013993407518363e-06, "loss": 0.009, "step": 9060 }, { "grad_norm": 0.16930943727493286, "learning_rate": 2.351032924139063e-06, "loss": 0.01, "step": 9070 }, { "grad_norm": 0.10949096828699112, "learning_rate": 2.30118758899619e-06, "loss": 0.0081, "step": 9080 }, { "grad_norm": 0.09550078958272934, "learning_rate": 2.2518638804238157e-06, "loss": 0.0106, "step": 9090 }, { "grad_norm": 0.09553427249193192, "learning_rate": 2.203062337818118e-06, "loss": 0.0084, "step": 9100 }, { "grad_norm": 0.11716356873512268, "learning_rate": 2.1547834948649483e-06, "loss": 0.0096, "step": 9110 }, { "grad_norm": 0.10805660486221313, "learning_rate": 2.1070278795340017e-06, "loss": 0.0091, "step": 9120 }, { "grad_norm": 0.09806843847036362, "learning_rate": 2.059796014073029e-06, "loss": 0.0089, "step": 9130 }, { "grad_norm": 0.13436640799045563, "learning_rate": 2.01308841500214e-06, "loss": 0.0094, "step": 9140 }, { "grad_norm": 0.09462229162454605, "learning_rate": 1.9669055931081704e-06, "loss": 0.0091, "step": 9150 }, { "grad_norm": 0.09547465294599533, "learning_rate": 1.9212480534390507e-06, "loss": 0.0091, "step": 9160 }, { "grad_norm": 0.11512733995914459, "learning_rate": 1.8761162952983246e-06, "loss": 0.0105, "step": 9170 }, { "grad_norm": 0.13469257950782776, "learning_rate": 1.8315108122396618e-06, "loss": 0.009, "step": 9180 }, { "grad_norm": 0.11139988154172897, "learning_rate": 1.787432092061475e-06, "loss": 0.01, "step": 9190 }, { "grad_norm": 0.10881217569112778, "learning_rate": 1.743880616801602e-06, "loss": 0.0093, "step": 9200 }, { "grad_norm": 0.12722499668598175, "learning_rate": 1.7008568627319865e-06, "loss": 0.0106, "step": 9210 }, { "grad_norm": 0.0840773954987526, "learning_rate": 1.6583613003535226e-06, "loss": 0.0094, "step": 9220 }, { "grad_norm": 0.11305548995733261, "learning_rate": 1.6163943943908522e-06, "loss": 0.0096, "step": 9230 }, { "grad_norm": 0.10006562620401382, "learning_rate": 1.5749566037873476e-06, "loss": 0.0087, "step": 9240 }, { "grad_norm": 0.08213262259960175, "learning_rate": 1.5340483817000428e-06, "loss": 0.0092, "step": 9250 }, { "grad_norm": 0.09439334273338318, "learning_rate": 1.4936701754947101e-06, "loss": 0.0103, "step": 9260 }, { "grad_norm": 0.13202977180480957, "learning_rate": 1.4538224267409361e-06, "loss": 0.0097, "step": 9270 }, { "grad_norm": 0.09729668498039246, "learning_rate": 1.414505571207314e-06, "loss": 0.0087, "step": 9280 }, { "grad_norm": 0.09637151658535004, "learning_rate": 1.3757200388566816e-06, "loss": 0.0093, "step": 9290 }, { "grad_norm": 0.12193208932876587, "learning_rate": 1.3374662538414074e-06, "loss": 0.01, "step": 9300 }, { "grad_norm": 0.10521063953638077, "learning_rate": 1.2997446344987617e-06, "loss": 0.0085, "step": 9310 }, { "grad_norm": 0.10748616605997086, "learning_rate": 1.262555593346315e-06, "loss": 0.0089, "step": 9320 }, { "grad_norm": 0.09921427816152573, "learning_rate": 1.2258995370774685e-06, "loss": 0.0079, "step": 9330 }, { "grad_norm": 0.0961531400680542, "learning_rate": 1.1897768665569798e-06, "loss": 0.0084, "step": 9340 }, { "grad_norm": 0.11345779150724411, "learning_rate": 1.1541879768165954e-06, "loss": 0.0094, "step": 9350 }, { "grad_norm": 0.09206047654151917, "learning_rate": 1.1191332570507085e-06, "loss": 0.0084, "step": 9360 }, { "grad_norm": 0.08921259641647339, "learning_rate": 1.0846130906121132e-06, "loss": 0.0093, "step": 9370 }, { "grad_norm": 0.08341587334871292, "learning_rate": 1.0506278550078131e-06, "loss": 0.0089, "step": 9380 }, { "grad_norm": 0.07334251701831818, "learning_rate": 1.0171779218949185e-06, "loss": 0.0091, "step": 9390 }, { "grad_norm": 0.09756631404161453, "learning_rate": 9.842636570765174e-07, "loss": 0.0098, "step": 9400 }, { "grad_norm": 0.09916376322507858, "learning_rate": 9.518854204977612e-07, "loss": 0.0093, "step": 9410 }, { "grad_norm": 0.07355308532714844, "learning_rate": 9.200435662418349e-07, "loss": 0.009, "step": 9420 }, { "grad_norm": 0.08397398144006729, "learning_rate": 8.887384425261658e-07, "loss": 0.0096, "step": 9430 }, { "grad_norm": 0.08130810409784317, "learning_rate": 8.579703916985648e-07, "loss": 0.0085, "step": 9440 }, { "grad_norm": 0.08272337913513184, "learning_rate": 8.277397502335194e-07, "loss": 0.0082, "step": 9450 }, { "grad_norm": 0.09818243980407715, "learning_rate": 7.980468487284675e-07, "loss": 0.0095, "step": 9460 }, { "grad_norm": 0.06783220171928406, "learning_rate": 7.688920119002297e-07, "loss": 0.008, "step": 9470 }, { "grad_norm": 0.11458965390920639, "learning_rate": 7.402755585814269e-07, "loss": 0.0091, "step": 9480 }, { "grad_norm": 0.08334201574325562, "learning_rate": 7.121978017170073e-07, "loss": 0.0102, "step": 9490 }, { "grad_norm": 0.11537005007266998, "learning_rate": 6.846590483608306e-07, "loss": 0.0093, "step": 9500 }, { "grad_norm": 0.08983808755874634, "learning_rate": 6.576595996722834e-07, "loss": 0.0078, "step": 9510 }, { "grad_norm": 0.10231666266918182, "learning_rate": 6.311997509130141e-07, "loss": 0.0104, "step": 9520 }, { "grad_norm": 0.09165944159030914, "learning_rate": 6.052797914436803e-07, "loss": 0.0104, "step": 9530 }, { "grad_norm": 0.07615209370851517, "learning_rate": 5.799000047208181e-07, "loss": 0.0096, "step": 9540 }, { "grad_norm": 0.11403731256723404, "learning_rate": 5.550606682937054e-07, "loss": 0.0096, "step": 9550 }, { "grad_norm": 0.10134383291006088, "learning_rate": 5.307620538013481e-07, "loss": 0.0105, "step": 9560 }, { "grad_norm": 0.11685805022716522, "learning_rate": 5.070044269694874e-07, "loss": 0.0103, "step": 9570 }, { "grad_norm": 0.09262718260288239, "learning_rate": 4.837880476077417e-07, "loss": 0.0093, "step": 9580 }, { "grad_norm": 0.08297484368085861, "learning_rate": 4.6111316960670835e-07, "loss": 0.0085, "step": 9590 }, { "grad_norm": 0.07831259816884995, "learning_rate": 4.389800409352218e-07, "loss": 0.0101, "step": 9600 }, { "grad_norm": 0.10337094217538834, "learning_rate": 4.173889036376277e-07, "loss": 0.009, "step": 9610 }, { "grad_norm": 0.1181417927145958, "learning_rate": 3.963399938311463e-07, "loss": 0.0093, "step": 9620 }, { "grad_norm": 0.07525420933961868, "learning_rate": 3.7583354170328545e-07, "loss": 0.0102, "step": 9630 }, { "grad_norm": 0.09781687706708908, "learning_rate": 3.558697715093207e-07, "loss": 0.0091, "step": 9640 }, { "grad_norm": 0.07795120775699615, "learning_rate": 3.3644890156983576e-07, "loss": 0.009, "step": 9650 }, { "grad_norm": 0.1063016951084137, "learning_rate": 3.175711442683638e-07, "loss": 0.0102, "step": 9660 }, { "grad_norm": 0.12205851823091507, "learning_rate": 2.9923670604902197e-07, "loss": 0.0084, "step": 9670 }, { "grad_norm": 0.0832120031118393, "learning_rate": 2.814457874143028e-07, "loss": 0.0083, "step": 9680 }, { "grad_norm": 0.08002273738384247, "learning_rate": 2.641985829228366e-07, "loss": 0.0104, "step": 9690 }, { "grad_norm": 0.07582097500562668, "learning_rate": 2.474952811872877e-07, "loss": 0.0101, "step": 9700 }, { "grad_norm": 0.08389899134635925, "learning_rate": 2.3133606487228397e-07, "loss": 0.0096, "step": 9710 }, { "grad_norm": 0.09777085483074188, "learning_rate": 2.157211106924295e-07, "loss": 0.0097, "step": 9720 }, { "grad_norm": 0.08385897427797318, "learning_rate": 2.006505894103672e-07, "loss": 0.0076, "step": 9730 }, { "grad_norm": 0.07970065623521805, "learning_rate": 1.8612466583489696e-07, "loss": 0.008, "step": 9740 }, { "grad_norm": 0.07993104308843613, "learning_rate": 1.7214349881918834e-07, "loss": 0.0079, "step": 9750 }, { "grad_norm": 0.10009945929050446, "learning_rate": 1.5870724125904845e-07, "loss": 0.0096, "step": 9760 }, { "grad_norm": 0.07894735783338547, "learning_rate": 1.4581604009124006e-07, "loss": 0.0086, "step": 9770 }, { "grad_norm": 0.1091090515255928, "learning_rate": 1.334700362918717e-07, "loss": 0.0095, "step": 9780 }, { "grad_norm": 0.08564490079879761, "learning_rate": 1.2166936487486015e-07, "loss": 0.0092, "step": 9790 }, { "grad_norm": 0.0926039069890976, "learning_rate": 1.1041415489045914e-07, "loss": 0.0084, "step": 9800 }, { "grad_norm": 0.09205314517021179, "learning_rate": 9.970452942384412e-08, "loss": 0.0086, "step": 9810 }, { "grad_norm": 0.08366527408361435, "learning_rate": 8.954060559375754e-08, "loss": 0.0088, "step": 9820 }, { "grad_norm": 0.11565835028886795, "learning_rate": 7.99224945512489e-08, "loss": 0.0104, "step": 9830 }, { "grad_norm": 0.0974796712398529, "learning_rate": 7.085030147843675e-08, "loss": 0.0094, "step": 9840 }, { "grad_norm": 0.07913940399885178, "learning_rate": 6.232412558736523e-08, "loss": 0.0082, "step": 9850 }, { "grad_norm": 0.08574803173542023, "learning_rate": 5.434406011893822e-08, "loss": 0.008, "step": 9860 }, { "grad_norm": 0.11111103743314743, "learning_rate": 4.6910192341864664e-08, "loss": 0.0092, "step": 9870 }, { "grad_norm": 0.07517047226428986, "learning_rate": 4.0022603551737035e-08, "loss": 0.0104, "step": 9880 }, { "grad_norm": 0.0997793972492218, "learning_rate": 3.3681369070120985e-08, "loss": 0.0085, "step": 9890 }, { "grad_norm": 0.08471164107322693, "learning_rate": 2.7886558243744866e-08, "loss": 0.0099, "step": 9900 }, { "grad_norm": 0.08656098693609238, "learning_rate": 2.2638234443722596e-08, "loss": 0.0087, "step": 9910 }, { "grad_norm": 0.08476077765226364, "learning_rate": 1.7936455064887504e-08, "loss": 0.0091, "step": 9920 }, { "grad_norm": 0.07970105856657028, "learning_rate": 1.378127152514841e-08, "loss": 0.0088, "step": 9930 }, { "grad_norm": 0.09936222434043884, "learning_rate": 1.0172729264917857e-08, "loss": 0.0089, "step": 9940 }, { "grad_norm": 0.08048803359270096, "learning_rate": 7.1108677466458215e-09, "loss": 0.0083, "step": 9950 }, { "grad_norm": 0.08394958823919296, "learning_rate": 4.595720454353414e-09, "loss": 0.0098, "step": 9960 }, { "grad_norm": 0.07573619484901428, "learning_rate": 2.627314893294264e-09, "loss": 0.01, "step": 9970 }, { "grad_norm": 0.10208114981651306, "learning_rate": 1.2056725896270048e-09, "loss": 0.0102, "step": 9980 }, { "grad_norm": 0.09716355055570602, "learning_rate": 3.308090902098826e-10, "loss": 0.0085, "step": 9990 }, { "grad_norm": 0.09812839329242706, "learning_rate": 2.7339624120159555e-12, "loss": 0.0086, "step": 10000 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 128, "trial_name": null, "trial_params": null }