{ "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": 0.5780179500579834, "learning_rate": 1.8e-06, "loss": 1.1855, "step": 10 }, { "grad_norm": 0.5869970917701721, "learning_rate": 3.8e-06, "loss": 1.1858, "step": 20 }, { "grad_norm": 1.0656380653381348, "learning_rate": 5.8e-06, "loss": 1.1769, "step": 30 }, { "grad_norm": 0.44934725761413574, "learning_rate": 7.8e-06, "loss": 1.1748, "step": 40 }, { "grad_norm": 0.45334547758102417, "learning_rate": 9.800000000000001e-06, "loss": 1.1822, "step": 50 }, { "grad_norm": 0.34318044781684875, "learning_rate": 1.18e-05, "loss": 1.1725, "step": 60 }, { "grad_norm": 0.5219594240188599, "learning_rate": 1.3800000000000002e-05, "loss": 1.1683, "step": 70 }, { "grad_norm": 0.4110032021999359, "learning_rate": 1.58e-05, "loss": 1.1808, "step": 80 }, { "grad_norm": 0.4219168424606323, "learning_rate": 1.78e-05, "loss": 1.1499, "step": 90 }, { "grad_norm": 1.6549280881881714, "learning_rate": 1.9800000000000004e-05, "loss": 1.1529, "step": 100 }, { "grad_norm": 1.5253068208694458, "learning_rate": 2.18e-05, "loss": 1.1358, "step": 110 }, { "grad_norm": 2.034780263900757, "learning_rate": 2.38e-05, "loss": 1.0965, "step": 120 }, { "grad_norm": 0.8334375619888306, "learning_rate": 2.58e-05, "loss": 1.0875, "step": 130 }, { "grad_norm": 1.663697361946106, "learning_rate": 2.7800000000000005e-05, "loss": 1.0284, "step": 140 }, { "grad_norm": 1.1360939741134644, "learning_rate": 2.98e-05, "loss": 1.0245, "step": 150 }, { "grad_norm": 1.5169981718063354, "learning_rate": 3.18e-05, "loss": 0.9499, "step": 160 }, { "grad_norm": 1.689755916595459, "learning_rate": 3.38e-05, "loss": 0.8959, "step": 170 }, { "grad_norm": 2.594494104385376, "learning_rate": 3.58e-05, "loss": 0.8787, "step": 180 }, { "grad_norm": 1.6937029361724854, "learning_rate": 3.7800000000000004e-05, "loss": 0.7971, "step": 190 }, { "grad_norm": 1.634263038635254, "learning_rate": 3.9800000000000005e-05, "loss": 0.6838, "step": 200 }, { "grad_norm": 2.959101676940918, "learning_rate": 4.18e-05, "loss": 0.5431, "step": 210 }, { "grad_norm": 2.6135573387145996, "learning_rate": 4.38e-05, "loss": 0.4867, "step": 220 }, { "grad_norm": 2.34859299659729, "learning_rate": 4.58e-05, "loss": 0.4265, "step": 230 }, { "grad_norm": 1.6821973323822021, "learning_rate": 4.78e-05, "loss": 0.3832, "step": 240 }, { "grad_norm": 1.7778671979904175, "learning_rate": 4.9800000000000004e-05, "loss": 0.2804, "step": 250 }, { "grad_norm": 1.9755197763442993, "learning_rate": 5.1800000000000005e-05, "loss": 0.2236, "step": 260 }, { "grad_norm": 2.127776861190796, "learning_rate": 5.380000000000001e-05, "loss": 0.2425, "step": 270 }, { "grad_norm": 1.266705870628357, "learning_rate": 5.580000000000001e-05, "loss": 0.2324, "step": 280 }, { "grad_norm": 4.997979164123535, "learning_rate": 5.7799999999999995e-05, "loss": 0.2664, "step": 290 }, { "grad_norm": 1.3988759517669678, "learning_rate": 5.9800000000000003e-05, "loss": 0.2932, "step": 300 }, { "grad_norm": 0.9591935873031616, "learning_rate": 6.18e-05, "loss": 0.2601, "step": 310 }, { "grad_norm": 1.8011797666549683, "learning_rate": 6.38e-05, "loss": 0.203, "step": 320 }, { "grad_norm": 1.4436049461364746, "learning_rate": 6.58e-05, "loss": 0.2411, "step": 330 }, { "grad_norm": 1.3680981397628784, "learning_rate": 6.780000000000001e-05, "loss": 0.209, "step": 340 }, { "grad_norm": 1.7653027772903442, "learning_rate": 6.98e-05, "loss": 0.2005, "step": 350 }, { "grad_norm": 1.3353748321533203, "learning_rate": 7.18e-05, "loss": 0.2145, "step": 360 }, { "grad_norm": 1.5616081953048706, "learning_rate": 7.38e-05, "loss": 0.2313, "step": 370 }, { "grad_norm": 1.6857037544250488, "learning_rate": 7.58e-05, "loss": 0.2029, "step": 380 }, { "grad_norm": 1.550398349761963, "learning_rate": 7.780000000000001e-05, "loss": 0.227, "step": 390 }, { "grad_norm": 1.26262366771698, "learning_rate": 7.98e-05, "loss": 0.2122, "step": 400 }, { "grad_norm": 1.5547126531600952, "learning_rate": 8.18e-05, "loss": 0.2226, "step": 410 }, { "grad_norm": 1.4188966751098633, "learning_rate": 8.38e-05, "loss": 0.2305, "step": 420 }, { "grad_norm": 0.9161266088485718, "learning_rate": 8.58e-05, "loss": 0.2309, "step": 430 }, { "grad_norm": 0.9617606401443481, "learning_rate": 8.78e-05, "loss": 0.21, "step": 440 }, { "grad_norm": 1.0878945589065552, "learning_rate": 8.98e-05, "loss": 0.1761, "step": 450 }, { "grad_norm": 1.3132082223892212, "learning_rate": 9.180000000000001e-05, "loss": 0.2062, "step": 460 }, { "grad_norm": 1.0124036073684692, "learning_rate": 9.38e-05, "loss": 0.2071, "step": 470 }, { "grad_norm": 1.1576775312423706, "learning_rate": 9.58e-05, "loss": 0.2144, "step": 480 }, { "grad_norm": 1.8741527795791626, "learning_rate": 9.78e-05, "loss": 0.2289, "step": 490 }, { "grad_norm": 0.8291168212890625, "learning_rate": 9.98e-05, "loss": 0.1679, "step": 500 }, { "grad_norm": 1.041609525680542, "learning_rate": 9.9999778549206e-05, "loss": 0.2196, "step": 510 }, { "grad_norm": 1.681339979171753, "learning_rate": 9.999901304280685e-05, "loss": 0.2272, "step": 520 }, { "grad_norm": 1.3579230308532715, "learning_rate": 9.999770075521164e-05, "loss": 0.2095, "step": 530 }, { "grad_norm": 1.0099557638168335, "learning_rate": 9.99958417007713e-05, "loss": 0.2054, "step": 540 }, { "grad_norm": 0.7613399028778076, "learning_rate": 9.999343589981615e-05, "loss": 0.1976, "step": 550 }, { "grad_norm": 1.1024012565612793, "learning_rate": 9.999048337865568e-05, "loss": 0.2007, "step": 560 }, { "grad_norm": 1.194237470626831, "learning_rate": 9.998698416957815e-05, "loss": 0.1582, "step": 570 }, { "grad_norm": 1.5290417671203613, "learning_rate": 9.998293831085037e-05, "loss": 0.1763, "step": 580 }, { "grad_norm": 0.6369971036911011, "learning_rate": 9.997834584671719e-05, "loss": 0.1628, "step": 590 }, { "grad_norm": 0.7296329736709595, "learning_rate": 9.997320682740107e-05, "loss": 0.1707, "step": 600 }, { "grad_norm": 1.011475920677185, "learning_rate": 9.996752130910149e-05, "loss": 0.1731, "step": 610 }, { "grad_norm": 1.7991652488708496, "learning_rate": 9.99612893539944e-05, "loss": 0.1821, "step": 620 }, { "grad_norm": 0.9146873950958252, "learning_rate": 9.995451103023144e-05, "loss": 0.1987, "step": 630 }, { "grad_norm": 0.8445677757263184, "learning_rate": 9.994718641193928e-05, "loss": 0.1813, "step": 640 }, { "grad_norm": 0.8563655018806458, "learning_rate": 9.993931557921874e-05, "loss": 0.1533, "step": 650 }, { "grad_norm": 0.9622541666030884, "learning_rate": 9.993089861814402e-05, "loss": 0.1805, "step": 660 }, { "grad_norm": 1.6675599813461304, "learning_rate": 9.992193562076166e-05, "loss": 0.1483, "step": 670 }, { "grad_norm": 0.7461128234863281, "learning_rate": 9.991242668508954e-05, "loss": 0.1558, "step": 680 }, { "grad_norm": 1.3172097206115723, "learning_rate": 9.990237191511587e-05, "loss": 0.1593, "step": 690 }, { "grad_norm": 0.9307062029838562, "learning_rate": 9.989177142079802e-05, "loss": 0.1664, "step": 700 }, { "grad_norm": 0.789570152759552, "learning_rate": 9.988062531806126e-05, "loss": 0.1525, "step": 710 }, { "grad_norm": 1.0092016458511353, "learning_rate": 9.986893372879762e-05, "loss": 0.1609, "step": 720 }, { "grad_norm": 1.1027886867523193, "learning_rate": 9.985669678086443e-05, "loss": 0.1694, "step": 730 }, { "grad_norm": 0.5743829607963562, "learning_rate": 9.984391460808298e-05, "loss": 0.1204, "step": 740 }, { "grad_norm": 0.6262876987457275, "learning_rate": 9.983058735023709e-05, "loss": 0.1308, "step": 750 }, { "grad_norm": 0.5650350451469421, "learning_rate": 9.98167151530715e-05, "loss": 0.1464, "step": 760 }, { "grad_norm": 0.7405779361724854, "learning_rate": 9.980229816829034e-05, "loss": 0.182, "step": 770 }, { "grad_norm": 1.3788007497787476, "learning_rate": 9.978733655355544e-05, "loss": 0.192, "step": 780 }, { "grad_norm": 0.6555464863777161, "learning_rate": 9.977183047248464e-05, "loss": 0.1363, "step": 790 }, { "grad_norm": 0.7306974530220032, "learning_rate": 9.975578009464992e-05, "loss": 0.1427, "step": 800 }, { "grad_norm": 0.8651738166809082, "learning_rate": 9.97391855955757e-05, "loss": 0.1592, "step": 810 }, { "grad_norm": 0.7441361546516418, "learning_rate": 9.972204715673669e-05, "loss": 0.1588, "step": 820 }, { "grad_norm": 0.42081770300865173, "learning_rate": 9.970436496555617e-05, "loss": 0.1731, "step": 830 }, { "grad_norm": 0.9355372786521912, "learning_rate": 9.968613921540373e-05, "loss": 0.1569, "step": 840 }, { "grad_norm": 0.8102606534957886, "learning_rate": 9.966737010559326e-05, "loss": 0.1852, "step": 850 }, { "grad_norm": 0.6917514204978943, "learning_rate": 9.964805784138072e-05, "loss": 0.1408, "step": 860 }, { "grad_norm": 0.6040403842926025, "learning_rate": 9.962820263396195e-05, "loss": 0.1587, "step": 870 }, { "grad_norm": 1.0503604412078857, "learning_rate": 9.960780470047033e-05, "loss": 0.1477, "step": 880 }, { "grad_norm": 0.8523648977279663, "learning_rate": 9.958686426397437e-05, "loss": 0.1484, "step": 890 }, { "grad_norm": 0.5573827028274536, "learning_rate": 9.956538155347534e-05, "loss": 0.135, "step": 900 }, { "grad_norm": 0.6877114176750183, "learning_rate": 9.95433568039047e-05, "loss": 0.1462, "step": 910 }, { "grad_norm": 0.8519322276115417, "learning_rate": 9.952079025612162e-05, "loss": 0.1386, "step": 920 }, { "grad_norm": 0.7193958163261414, "learning_rate": 9.949768215691022e-05, "loss": 0.1749, "step": 930 }, { "grad_norm": 0.9248827695846558, "learning_rate": 9.9474032758977e-05, "loss": 0.1522, "step": 940 }, { "grad_norm": 1.2417088747024536, "learning_rate": 9.944984232094794e-05, "loss": 0.1416, "step": 950 }, { "grad_norm": 0.5647571086883545, "learning_rate": 9.942511110736584e-05, "loss": 0.1347, "step": 960 }, { "grad_norm": 0.6503048539161682, "learning_rate": 9.939983938868726e-05, "loss": 0.1604, "step": 970 }, { "grad_norm": 0.9128164052963257, "learning_rate": 9.93740274412797e-05, "loss": 0.1432, "step": 980 }, { "grad_norm": 0.6832290887832642, "learning_rate": 9.934767554741846e-05, "loss": 0.152, "step": 990 }, { "grad_norm": 1.1181143522262573, "learning_rate": 9.932078399528361e-05, "loss": 0.1237, "step": 1000 }, { "grad_norm": 0.6594109535217285, "learning_rate": 9.929335307895689e-05, "loss": 0.1358, "step": 1010 }, { "grad_norm": 0.9479378461837769, "learning_rate": 9.926538309841839e-05, "loss": 0.1571, "step": 1020 }, { "grad_norm": 0.7370449900627136, "learning_rate": 9.923687435954334e-05, "loss": 0.1414, "step": 1030 }, { "grad_norm": 0.8689610362052917, "learning_rate": 9.920782717409873e-05, "loss": 0.1453, "step": 1040 }, { "grad_norm": 1.7723618745803833, "learning_rate": 9.917824185973994e-05, "loss": 0.1567, "step": 1050 }, { "grad_norm": 0.5684078931808472, "learning_rate": 9.914811874000723e-05, "loss": 0.1416, "step": 1060 }, { "grad_norm": 0.760726809501648, "learning_rate": 9.911745814432218e-05, "loss": 0.1489, "step": 1070 }, { "grad_norm": 0.9661853909492493, "learning_rate": 9.90862604079842e-05, "loss": 0.1642, "step": 1080 }, { "grad_norm": 0.5236778259277344, "learning_rate": 9.90545258721667e-05, "loss": 0.1406, "step": 1090 }, { "grad_norm": 0.7970632910728455, "learning_rate": 9.90222548839135e-05, "loss": 0.1259, "step": 1100 }, { "grad_norm": 0.9028500318527222, "learning_rate": 9.898944779613495e-05, "loss": 0.1454, "step": 1110 }, { "grad_norm": 1.312590479850769, "learning_rate": 9.89561049676041e-05, "loss": 0.1191, "step": 1120 }, { "grad_norm": 0.6038645505905151, "learning_rate": 9.89222267629528e-05, "loss": 0.1472, "step": 1130 }, { "grad_norm": 0.730760395526886, "learning_rate": 9.888781355266763e-05, "loss": 0.1869, "step": 1140 }, { "grad_norm": 0.4676041901111603, "learning_rate": 9.885286571308598e-05, "loss": 0.1533, "step": 1150 }, { "grad_norm": 0.5219357013702393, "learning_rate": 9.881738362639182e-05, "loss": 0.1482, "step": 1160 }, { "grad_norm": 0.6638937592506409, "learning_rate": 9.878136768061154e-05, "loss": 0.135, "step": 1170 }, { "grad_norm": 0.7467520833015442, "learning_rate": 9.874481826960979e-05, "loss": 0.1122, "step": 1180 }, { "grad_norm": 0.7374361157417297, "learning_rate": 9.870773579308503e-05, "loss": 0.1422, "step": 1190 }, { "grad_norm": 0.7214168906211853, "learning_rate": 9.867012065656533e-05, "loss": 0.1476, "step": 1200 }, { "grad_norm": 0.7666581273078918, "learning_rate": 9.863197327140376e-05, "loss": 0.1276, "step": 1210 }, { "grad_norm": 0.8063265681266785, "learning_rate": 9.859329405477403e-05, "loss": 0.145, "step": 1220 }, { "grad_norm": 0.6530053615570068, "learning_rate": 9.855408342966585e-05, "loss": 0.1205, "step": 1230 }, { "grad_norm": 1.3437426090240479, "learning_rate": 9.851434182488033e-05, "loss": 0.143, "step": 1240 }, { "grad_norm": 0.7637791633605957, "learning_rate": 9.84740696750253e-05, "loss": 0.1241, "step": 1250 }, { "grad_norm": 1.3913887739181519, "learning_rate": 9.843326742051055e-05, "loss": 0.139, "step": 1260 }, { "grad_norm": 0.6103953719139099, "learning_rate": 9.839193550754297e-05, "loss": 0.1125, "step": 1270 }, { "grad_norm": 0.5727524161338806, "learning_rate": 9.835007438812177e-05, "loss": 0.1415, "step": 1280 }, { "grad_norm": 0.7588467001914978, "learning_rate": 9.830768452003341e-05, "loss": 0.1336, "step": 1290 }, { "grad_norm": 0.9322163462638855, "learning_rate": 9.826476636684671e-05, "loss": 0.1422, "step": 1300 }, { "grad_norm": 0.5724161863327026, "learning_rate": 9.822132039790773e-05, "loss": 0.1471, "step": 1310 }, { "grad_norm": 0.6532781720161438, "learning_rate": 9.817734708833461e-05, "loss": 0.1431, "step": 1320 }, { "grad_norm": 0.7093478441238403, "learning_rate": 9.813284691901243e-05, "loss": 0.1213, "step": 1330 }, { "grad_norm": 0.4390380382537842, "learning_rate": 9.808782037658792e-05, "loss": 0.1406, "step": 1340 }, { "grad_norm": 1.0854969024658203, "learning_rate": 9.804226795346411e-05, "loss": 0.1619, "step": 1350 }, { "grad_norm": 0.7211141586303711, "learning_rate": 9.799619014779503e-05, "loss": 0.1271, "step": 1360 }, { "grad_norm": 0.6482756733894348, "learning_rate": 9.794958746348013e-05, "loss": 0.141, "step": 1370 }, { "grad_norm": 0.660712480545044, "learning_rate": 9.790246041015896e-05, "loss": 0.116, "step": 1380 }, { "grad_norm": 0.5669071078300476, "learning_rate": 9.785480950320538e-05, "loss": 0.1351, "step": 1390 }, { "grad_norm": 0.5253428816795349, "learning_rate": 9.78066352637221e-05, "loss": 0.1241, "step": 1400 }, { "grad_norm": 0.6068577766418457, "learning_rate": 9.775793821853488e-05, "loss": 0.1473, "step": 1410 }, { "grad_norm": 0.682200014591217, "learning_rate": 9.77087189001868e-05, "loss": 0.1223, "step": 1420 }, { "grad_norm": 0.6875131130218506, "learning_rate": 9.765897784693243e-05, "loss": 0.1154, "step": 1430 }, { "grad_norm": 0.9316884279251099, "learning_rate": 9.760871560273197e-05, "loss": 0.1398, "step": 1440 }, { "grad_norm": 1.0771321058273315, "learning_rate": 9.755793271724526e-05, "loss": 0.1518, "step": 1450 }, { "grad_norm": 0.4644845128059387, "learning_rate": 9.750662974582584e-05, "loss": 0.122, "step": 1460 }, { "grad_norm": 1.1517488956451416, "learning_rate": 9.745480724951473e-05, "loss": 0.1371, "step": 1470 }, { "grad_norm": 0.8959156274795532, "learning_rate": 9.740246579503447e-05, "loss": 0.1385, "step": 1480 }, { "grad_norm": 0.6989287734031677, "learning_rate": 9.734960595478284e-05, "loss": 0.1362, "step": 1490 }, { "grad_norm": 0.9027889370918274, "learning_rate": 9.729622830682657e-05, "loss": 0.1545, "step": 1500 }, { "grad_norm": 0.5657837986946106, "learning_rate": 9.724233343489504e-05, "loss": 0.1156, "step": 1510 }, { "grad_norm": 1.2955174446105957, "learning_rate": 9.718792192837396e-05, "loss": 0.1321, "step": 1520 }, { "grad_norm": 0.7190059423446655, "learning_rate": 9.713299438229886e-05, "loss": 0.1346, "step": 1530 }, { "grad_norm": 0.5783393383026123, "learning_rate": 9.707755139734855e-05, "loss": 0.1427, "step": 1540 }, { "grad_norm": 0.70136958360672, "learning_rate": 9.702159357983866e-05, "loss": 0.1342, "step": 1550 }, { "grad_norm": 0.6306099891662598, "learning_rate": 9.696512154171492e-05, "loss": 0.142, "step": 1560 }, { "grad_norm": 0.5954473614692688, "learning_rate": 9.690813590054645e-05, "loss": 0.1154, "step": 1570 }, { "grad_norm": 0.9227553606033325, "learning_rate": 9.685063727951914e-05, "loss": 0.1365, "step": 1580 }, { "grad_norm": 0.5813431739807129, "learning_rate": 9.679262630742865e-05, "loss": 0.1344, "step": 1590 }, { "grad_norm": 0.7154408693313599, "learning_rate": 9.673410361867373e-05, "loss": 0.1155, "step": 1600 }, { "grad_norm": 0.8946419954299927, "learning_rate": 9.667506985324909e-05, "loss": 0.1049, "step": 1610 }, { "grad_norm": 0.5063388347625732, "learning_rate": 9.661552565673855e-05, "loss": 0.1198, "step": 1620 }, { "grad_norm": 0.9255726933479309, "learning_rate": 9.655547168030789e-05, "loss": 0.1555, "step": 1630 }, { "grad_norm": 1.0362030267715454, "learning_rate": 9.649490858069777e-05, "loss": 0.1402, "step": 1640 }, { "grad_norm": 0.7516462802886963, "learning_rate": 9.643383702021658e-05, "loss": 0.1605, "step": 1650 }, { "grad_norm": 1.0271644592285156, "learning_rate": 9.637225766673307e-05, "loss": 0.1609, "step": 1660 }, { "grad_norm": 0.7086920142173767, "learning_rate": 9.631017119366922e-05, "loss": 0.1352, "step": 1670 }, { "grad_norm": 1.0367600917816162, "learning_rate": 9.624757827999273e-05, "loss": 0.138, "step": 1680 }, { "grad_norm": 0.6368134021759033, "learning_rate": 9.618447961020971e-05, "loss": 0.1375, "step": 1690 }, { "grad_norm": 1.1094588041305542, "learning_rate": 9.612087587435707e-05, "loss": 0.1426, "step": 1700 }, { "grad_norm": 0.6850295066833496, "learning_rate": 9.605676776799508e-05, "loss": 0.1408, "step": 1710 }, { "grad_norm": 0.5318161249160767, "learning_rate": 9.599215599219973e-05, "loss": 0.1307, "step": 1720 }, { "grad_norm": 1.091141939163208, "learning_rate": 9.592704125355505e-05, "loss": 0.1401, "step": 1730 }, { "grad_norm": 1.046175241470337, "learning_rate": 9.586142426414538e-05, "loss": 0.1357, "step": 1740 }, { "grad_norm": 1.1542227268218994, "learning_rate": 9.57953057415476e-05, "loss": 0.1364, "step": 1750 }, { "grad_norm": 1.0420176982879639, "learning_rate": 9.572868640882328e-05, "loss": 0.1345, "step": 1760 }, { "grad_norm": 1.3089466094970703, "learning_rate": 9.56615669945108e-05, "loss": 0.1465, "step": 1770 }, { "grad_norm": 0.8598980903625488, "learning_rate": 9.55939482326173e-05, "loss": 0.1667, "step": 1780 }, { "grad_norm": 0.7444891333580017, "learning_rate": 9.552583086261069e-05, "loss": 0.1351, "step": 1790 }, { "grad_norm": 0.5432271957397461, "learning_rate": 9.545721562941168e-05, "loss": 0.119, "step": 1800 }, { "grad_norm": 0.5851179361343384, "learning_rate": 9.538810328338543e-05, "loss": 0.1147, "step": 1810 }, { "grad_norm": 0.7954565286636353, "learning_rate": 9.531849458033349e-05, "loss": 0.13, "step": 1820 }, { "grad_norm": 2.3662149906158447, "learning_rate": 9.524839028148547e-05, "loss": 0.1643, "step": 1830 }, { "grad_norm": 0.672012448310852, "learning_rate": 9.517779115349077e-05, "loss": 0.1355, "step": 1840 }, { "grad_norm": 0.8759785890579224, "learning_rate": 9.510669796841014e-05, "loss": 0.1627, "step": 1850 }, { "grad_norm": 0.640409529209137, "learning_rate": 9.503511150370727e-05, "loss": 0.1527, "step": 1860 }, { "grad_norm": 0.7285005450248718, "learning_rate": 9.496303254224024e-05, "loss": 0.107, "step": 1870 }, { "grad_norm": 0.44621723890304565, "learning_rate": 9.489046187225306e-05, "loss": 0.099, "step": 1880 }, { "grad_norm": 0.6594659090042114, "learning_rate": 9.481740028736692e-05, "loss": 0.1279, "step": 1890 }, { "grad_norm": 0.47400325536727905, "learning_rate": 9.474384858657164e-05, "loss": 0.1212, "step": 1900 }, { "grad_norm": 0.6146695613861084, "learning_rate": 9.466980757421679e-05, "loss": 0.1367, "step": 1910 }, { "grad_norm": 0.5293046832084656, "learning_rate": 9.459527806000305e-05, "loss": 0.1162, "step": 1920 }, { "grad_norm": 0.8982950448989868, "learning_rate": 9.452026085897325e-05, "loss": 0.123, "step": 1930 }, { "grad_norm": 0.5693746209144592, "learning_rate": 9.444475679150348e-05, "loss": 0.1253, "step": 1940 }, { "grad_norm": 0.726219117641449, "learning_rate": 9.436876668329411e-05, "loss": 0.1267, "step": 1950 }, { "grad_norm": 0.9716258645057678, "learning_rate": 9.429229136536079e-05, "loss": 0.1815, "step": 1960 }, { "grad_norm": 0.9436078071594238, "learning_rate": 9.421533167402534e-05, "loss": 0.1264, "step": 1970 }, { "grad_norm": 0.6088870167732239, "learning_rate": 9.413788845090666e-05, "loss": 0.1354, "step": 1980 }, { "grad_norm": 0.6033660769462585, "learning_rate": 9.405996254291136e-05, "loss": 0.1247, "step": 1990 }, { "grad_norm": 1.3232632875442505, "learning_rate": 9.398155480222474e-05, "loss": 0.1287, "step": 2000 }, { "grad_norm": 0.5999051332473755, "learning_rate": 9.390266608630128e-05, "loss": 0.1542, "step": 2010 }, { "grad_norm": 0.7801201939582825, "learning_rate": 9.38232972578553e-05, "loss": 0.1285, "step": 2020 }, { "grad_norm": 0.540688157081604, "learning_rate": 9.374344918485164e-05, "loss": 0.1252, "step": 2030 }, { "grad_norm": 0.7680627107620239, "learning_rate": 9.366312274049602e-05, "loss": 0.1381, "step": 2040 }, { "grad_norm": 0.664294958114624, "learning_rate": 9.358231880322554e-05, "loss": 0.1435, "step": 2050 }, { "grad_norm": 0.7024267315864563, "learning_rate": 9.350103825669916e-05, "loss": 0.1304, "step": 2060 }, { "grad_norm": 0.49538540840148926, "learning_rate": 9.341928198978787e-05, "loss": 0.1095, "step": 2070 }, { "grad_norm": 0.6042807698249817, "learning_rate": 9.333705089656512e-05, "loss": 0.1169, "step": 2080 }, { "grad_norm": 0.554349958896637, "learning_rate": 9.325434587629698e-05, "loss": 0.114, "step": 2090 }, { "grad_norm": 0.40326714515686035, "learning_rate": 9.31711678334323e-05, "loss": 0.0935, "step": 2100 }, { "grad_norm": 0.7731055617332458, "learning_rate": 9.308751767759282e-05, "loss": 0.1301, "step": 2110 }, { "grad_norm": 0.8644314408302307, "learning_rate": 9.300339632356325e-05, "loss": 0.1301, "step": 2120 }, { "grad_norm": 1.236546277999878, "learning_rate": 9.291880469128124e-05, "loss": 0.1337, "step": 2130 }, { "grad_norm": 0.4555363357067108, "learning_rate": 9.283374370582732e-05, "loss": 0.106, "step": 2140 }, { "grad_norm": 0.5409621000289917, "learning_rate": 9.274821429741482e-05, "loss": 0.1362, "step": 2150 }, { "grad_norm": 0.8813226819038391, "learning_rate": 9.266221740137961e-05, "loss": 0.1159, "step": 2160 }, { "grad_norm": 0.7029048800468445, "learning_rate": 9.257575395817001e-05, "loss": 0.1438, "step": 2170 }, { "grad_norm": 0.7038227319717407, "learning_rate": 9.248882491333637e-05, "loss": 0.1279, "step": 2180 }, { "grad_norm": 0.7189838886260986, "learning_rate": 9.240143121752076e-05, "loss": 0.1135, "step": 2190 }, { "grad_norm": 0.9633091688156128, "learning_rate": 9.23135738264467e-05, "loss": 0.1305, "step": 2200 }, { "grad_norm": 0.47805964946746826, "learning_rate": 9.222525370090849e-05, "loss": 0.1328, "step": 2210 }, { "grad_norm": 0.605929970741272, "learning_rate": 9.213647180676088e-05, "loss": 0.1132, "step": 2220 }, { "grad_norm": 0.4386877715587616, "learning_rate": 9.204722911490846e-05, "loss": 0.1223, "step": 2230 }, { "grad_norm": 0.7663604617118835, "learning_rate": 9.1957526601295e-05, "loss": 0.1174, "step": 2240 }, { "grad_norm": 0.4937477111816406, "learning_rate": 9.186736524689281e-05, "loss": 0.1187, "step": 2250 }, { "grad_norm": 0.4460795521736145, "learning_rate": 9.177674603769204e-05, "loss": 0.0929, "step": 2260 }, { "grad_norm": 0.6728976964950562, "learning_rate": 9.168566996468983e-05, "loss": 0.102, "step": 2270 }, { "grad_norm": 0.3893265724182129, "learning_rate": 9.159413802387951e-05, "loss": 0.0901, "step": 2280 }, { "grad_norm": 0.5107865929603577, "learning_rate": 9.150215121623974e-05, "loss": 0.1126, "step": 2290 }, { "grad_norm": 0.38916125893592834, "learning_rate": 9.140971054772349e-05, "loss": 0.092, "step": 2300 }, { "grad_norm": 0.5199502110481262, "learning_rate": 9.131681702924713e-05, "loss": 0.1082, "step": 2310 }, { "grad_norm": 0.6223076581954956, "learning_rate": 9.122347167667926e-05, "loss": 0.1002, "step": 2320 }, { "grad_norm": 0.42556697130203247, "learning_rate": 9.112967551082973e-05, "loss": 0.1296, "step": 2330 }, { "grad_norm": 0.4137495458126068, "learning_rate": 9.103542955743835e-05, "loss": 0.1056, "step": 2340 }, { "grad_norm": 0.4414235055446625, "learning_rate": 9.094073484716381e-05, "loss": 0.1101, "step": 2350 }, { "grad_norm": 0.5673224329948425, "learning_rate": 9.084559241557226e-05, "loss": 0.1144, "step": 2360 }, { "grad_norm": 0.43088409304618835, "learning_rate": 9.075000330312608e-05, "loss": 0.1247, "step": 2370 }, { "grad_norm": 0.3968179225921631, "learning_rate": 9.065396855517253e-05, "loss": 0.1211, "step": 2380 }, { "grad_norm": 0.4962259531021118, "learning_rate": 9.055748922193219e-05, "loss": 0.1209, "step": 2390 }, { "grad_norm": 0.5517762899398804, "learning_rate": 9.046056635848761e-05, "loss": 0.1144, "step": 2400 }, { "grad_norm": 0.5910106897354126, "learning_rate": 9.036320102477169e-05, "loss": 0.1126, "step": 2410 }, { "grad_norm": 0.7645018696784973, "learning_rate": 9.02653942855561e-05, "loss": 0.1516, "step": 2420 }, { "grad_norm": 0.5800725817680359, "learning_rate": 9.016714721043971e-05, "loss": 0.1203, "step": 2430 }, { "grad_norm": 0.685245931148529, "learning_rate": 9.006846087383675e-05, "loss": 0.1362, "step": 2440 }, { "grad_norm": 0.5141698122024536, "learning_rate": 8.996933635496523e-05, "loss": 0.1118, "step": 2450 }, { "grad_norm": 1.060425043106079, "learning_rate": 8.986977473783498e-05, "loss": 0.1293, "step": 2460 }, { "grad_norm": 0.5873779654502869, "learning_rate": 8.97697771112359e-05, "loss": 0.1246, "step": 2470 }, { "grad_norm": 0.5342249274253845, "learning_rate": 8.966934456872602e-05, "loss": 0.1203, "step": 2480 }, { "grad_norm": 0.9915500283241272, "learning_rate": 8.95684782086195e-05, "loss": 0.1184, "step": 2490 }, { "grad_norm": 0.9389453530311584, "learning_rate": 8.946717913397476e-05, "loss": 0.1041, "step": 2500 }, { "grad_norm": 0.47131210565567017, "learning_rate": 8.93654484525822e-05, "loss": 0.1029, "step": 2510 }, { "grad_norm": 0.5389177203178406, "learning_rate": 8.926328727695226e-05, "loss": 0.1091, "step": 2520 }, { "grad_norm": 0.41263818740844727, "learning_rate": 8.916069672430319e-05, "loss": 0.104, "step": 2530 }, { "grad_norm": 0.5178645253181458, "learning_rate": 8.905767791654884e-05, "loss": 0.1117, "step": 2540 }, { "grad_norm": 0.4925541877746582, "learning_rate": 8.895423198028638e-05, "loss": 0.1163, "step": 2550 }, { "grad_norm": 0.4278854727745056, "learning_rate": 8.885036004678402e-05, "loss": 0.1071, "step": 2560 }, { "grad_norm": 0.7558642029762268, "learning_rate": 8.874606325196857e-05, "loss": 0.1267, "step": 2570 }, { "grad_norm": 0.4618389904499054, "learning_rate": 8.864134273641304e-05, "loss": 0.0953, "step": 2580 }, { "grad_norm": 0.4521310031414032, "learning_rate": 8.853619964532427e-05, "loss": 0.1099, "step": 2590 }, { "grad_norm": 0.6213418245315552, "learning_rate": 8.843063512853019e-05, "loss": 0.1546, "step": 2600 }, { "grad_norm": 0.7035230398178101, "learning_rate": 8.832465034046749e-05, "loss": 0.1252, "step": 2610 }, { "grad_norm": 0.8496027588844299, "learning_rate": 8.821824644016882e-05, "loss": 0.1027, "step": 2620 }, { "grad_norm": 0.5012723207473755, "learning_rate": 8.811142459125019e-05, "loss": 0.1297, "step": 2630 }, { "grad_norm": 0.5326700210571289, "learning_rate": 8.800418596189822e-05, "loss": 0.138, "step": 2640 }, { "grad_norm": 0.583854079246521, "learning_rate": 8.789653172485737e-05, "loss": 0.1283, "step": 2650 }, { "grad_norm": 0.47870713472366333, "learning_rate": 8.778846305741715e-05, "loss": 0.1247, "step": 2660 }, { "grad_norm": 0.6043845415115356, "learning_rate": 8.767998114139918e-05, "loss": 0.1376, "step": 2670 }, { "grad_norm": 0.5682188868522644, "learning_rate": 8.757108716314429e-05, "loss": 0.1256, "step": 2680 }, { "grad_norm": 0.5210828185081482, "learning_rate": 8.746178231349962e-05, "loss": 0.1133, "step": 2690 }, { "grad_norm": 0.7691870331764221, "learning_rate": 8.735206778780549e-05, "loss": 0.1051, "step": 2700 }, { "grad_norm": 0.513430118560791, "learning_rate": 8.724194478588234e-05, "loss": 0.1125, "step": 2710 }, { "grad_norm": 0.5964224934577942, "learning_rate": 8.713141451201772e-05, "loss": 0.1271, "step": 2720 }, { "grad_norm": 0.45631977915763855, "learning_rate": 8.702047817495295e-05, "loss": 0.1183, "step": 2730 }, { "grad_norm": 1.344876766204834, "learning_rate": 8.69091369878701e-05, "loss": 0.1164, "step": 2740 }, { "grad_norm": 0.9987202286720276, "learning_rate": 8.679739216837849e-05, "loss": 0.1291, "step": 2750 }, { "grad_norm": 0.4542425572872162, "learning_rate": 8.66852449385016e-05, "loss": 0.1191, "step": 2760 }, { "grad_norm": 0.6505128741264343, "learning_rate": 8.657269652466356e-05, "loss": 0.1073, "step": 2770 }, { "grad_norm": 0.6039180755615234, "learning_rate": 8.645974815767577e-05, "loss": 0.1118, "step": 2780 }, { "grad_norm": 0.8731012344360352, "learning_rate": 8.634640107272351e-05, "loss": 0.1115, "step": 2790 }, { "grad_norm": 0.7360416054725647, "learning_rate": 8.623265650935234e-05, "loss": 0.1092, "step": 2800 }, { "grad_norm": 0.6806288361549377, "learning_rate": 8.611851571145456e-05, "loss": 0.1344, "step": 2810 }, { "grad_norm": 0.9552347660064697, "learning_rate": 8.600397992725566e-05, "loss": 0.1039, "step": 2820 }, { "grad_norm": 0.47237882018089294, "learning_rate": 8.588905040930061e-05, "loss": 0.0971, "step": 2830 }, { "grad_norm": 1.428812861442566, "learning_rate": 8.577372841444022e-05, "loss": 0.1354, "step": 2840 }, { "grad_norm": 0.45926663279533386, "learning_rate": 8.565801520381736e-05, "loss": 0.11, "step": 2850 }, { "grad_norm": 0.6788615584373474, "learning_rate": 8.554191204285313e-05, "loss": 0.1384, "step": 2860 }, { "grad_norm": 0.5688839554786682, "learning_rate": 8.542542020123315e-05, "loss": 0.0979, "step": 2870 }, { "grad_norm": 0.739556074142456, "learning_rate": 8.530854095289347e-05, "loss": 0.1119, "step": 2880 }, { "grad_norm": 0.429339736700058, "learning_rate": 8.519127557600688e-05, "loss": 0.1105, "step": 2890 }, { "grad_norm": 0.650242269039154, "learning_rate": 8.507362535296871e-05, "loss": 0.1113, "step": 2900 }, { "grad_norm": 0.6444766521453857, "learning_rate": 8.495559157038299e-05, "loss": 0.0949, "step": 2910 }, { "grad_norm": 0.3491423726081848, "learning_rate": 8.483717551904823e-05, "loss": 0.1182, "step": 2920 }, { "grad_norm": 0.9333589673042297, "learning_rate": 8.47183784939434e-05, "loss": 0.0908, "step": 2930 }, { "grad_norm": 0.584991991519928, "learning_rate": 8.459920179421374e-05, "loss": 0.0946, "step": 2940 }, { "grad_norm": 0.42924031615257263, "learning_rate": 8.447964672315656e-05, "loss": 0.0999, "step": 2950 }, { "grad_norm": 0.6555279493331909, "learning_rate": 8.435971458820692e-05, "loss": 0.1124, "step": 2960 }, { "grad_norm": 0.6194357872009277, "learning_rate": 8.423940670092345e-05, "loss": 0.0989, "step": 2970 }, { "grad_norm": 0.5603435635566711, "learning_rate": 8.411872437697394e-05, "loss": 0.0977, "step": 2980 }, { "grad_norm": 0.5386211276054382, "learning_rate": 8.399766893612096e-05, "loss": 0.1145, "step": 2990 }, { "grad_norm": 0.6309317946434021, "learning_rate": 8.38762417022074e-05, "loss": 0.0998, "step": 3000 }, { "grad_norm": 0.8125643134117126, "learning_rate": 8.375444400314204e-05, "loss": 0.0858, "step": 3010 }, { "grad_norm": 0.7054994702339172, "learning_rate": 8.3632277170885e-05, "loss": 0.0989, "step": 3020 }, { "grad_norm": 0.3920750617980957, "learning_rate": 8.350974254143318e-05, "loss": 0.0913, "step": 3030 }, { "grad_norm": 0.5084161758422852, "learning_rate": 8.338684145480566e-05, "loss": 0.122, "step": 3040 }, { "grad_norm": 0.5673415064811707, "learning_rate": 8.326357525502904e-05, "loss": 0.1165, "step": 3050 }, { "grad_norm": 0.801711916923523, "learning_rate": 8.313994529012273e-05, "loss": 0.1132, "step": 3060 }, { "grad_norm": 0.5055217146873474, "learning_rate": 8.301595291208422e-05, "loss": 0.0898, "step": 3070 }, { "grad_norm": 0.4956274628639221, "learning_rate": 8.289159947687427e-05, "loss": 0.1266, "step": 3080 }, { "grad_norm": 1.0802046060562134, "learning_rate": 8.276688634440216e-05, "loss": 0.1357, "step": 3090 }, { "grad_norm": 0.8836338520050049, "learning_rate": 8.26418148785107e-05, "loss": 0.106, "step": 3100 }, { "grad_norm": 0.6314801573753357, "learning_rate": 8.251638644696141e-05, "loss": 0.1058, "step": 3110 }, { "grad_norm": 0.4846455752849579, "learning_rate": 8.23906024214195e-05, "loss": 0.0953, "step": 3120 }, { "grad_norm": 0.7425668835639954, "learning_rate": 8.226446417743897e-05, "loss": 0.0943, "step": 3130 }, { "grad_norm": 0.6256777048110962, "learning_rate": 8.213797309444742e-05, "loss": 0.1106, "step": 3140 }, { "grad_norm": 0.48116856813430786, "learning_rate": 8.201113055573105e-05, "loss": 0.1013, "step": 3150 }, { "grad_norm": 1.3814300298690796, "learning_rate": 8.188393794841958e-05, "loss": 0.108, "step": 3160 }, { "grad_norm": 0.5032051205635071, "learning_rate": 8.175639666347094e-05, "loss": 0.0954, "step": 3170 }, { "grad_norm": 0.740406334400177, "learning_rate": 8.162850809565623e-05, "loss": 0.1092, "step": 3180 }, { "grad_norm": 0.4717388153076172, "learning_rate": 8.150027364354431e-05, "loss": 0.1079, "step": 3190 }, { "grad_norm": 0.7672988176345825, "learning_rate": 8.137169470948662e-05, "loss": 0.1057, "step": 3200 }, { "grad_norm": 0.5005557537078857, "learning_rate": 8.124277269960179e-05, "loss": 0.1021, "step": 3210 }, { "grad_norm": 0.6313621401786804, "learning_rate": 8.111350902376023e-05, "loss": 0.1186, "step": 3220 }, { "grad_norm": 0.3950771987438202, "learning_rate": 8.098390509556883e-05, "loss": 0.1242, "step": 3230 }, { "grad_norm": 0.633834183216095, "learning_rate": 8.085396233235536e-05, "loss": 0.1001, "step": 3240 }, { "grad_norm": 0.40161705017089844, "learning_rate": 8.072368215515306e-05, "loss": 0.1164, "step": 3250 }, { "grad_norm": 0.5127862095832825, "learning_rate": 8.059306598868506e-05, "loss": 0.1143, "step": 3260 }, { "grad_norm": 0.43583595752716064, "learning_rate": 8.046211526134888e-05, "loss": 0.1158, "step": 3270 }, { "grad_norm": 0.5329251289367676, "learning_rate": 8.033083140520065e-05, "loss": 0.0939, "step": 3280 }, { "grad_norm": 0.41451963782310486, "learning_rate": 8.019921585593962e-05, "loss": 0.1048, "step": 3290 }, { "grad_norm": 0.37818682193756104, "learning_rate": 8.006727005289232e-05, "loss": 0.084, "step": 3300 }, { "grad_norm": 0.5068961977958679, "learning_rate": 7.993499543899692e-05, "loss": 0.1011, "step": 3310 }, { "grad_norm": 0.7691237330436707, "learning_rate": 7.980239346078742e-05, "loss": 0.1146, "step": 3320 }, { "grad_norm": 0.5319196581840515, "learning_rate": 7.966946556837778e-05, "loss": 0.0835, "step": 3330 }, { "grad_norm": 0.876521646976471, "learning_rate": 7.953621321544616e-05, "loss": 0.0922, "step": 3340 }, { "grad_norm": 0.5534335970878601, "learning_rate": 7.940263785921896e-05, "loss": 0.0969, "step": 3350 }, { "grad_norm": 0.5090944170951843, "learning_rate": 7.926874096045482e-05, "loss": 0.0836, "step": 3360 }, { "grad_norm": 0.9975054264068604, "learning_rate": 7.913452398342881e-05, "loss": 0.1034, "step": 3370 }, { "grad_norm": 0.5329188108444214, "learning_rate": 7.89999883959163e-05, "loss": 0.1293, "step": 3380 }, { "grad_norm": 0.4581211805343628, "learning_rate": 7.886513566917687e-05, "loss": 0.089, "step": 3390 }, { "grad_norm": 0.49225670099258423, "learning_rate": 7.872996727793838e-05, "loss": 0.098, "step": 3400 }, { "grad_norm": 0.6229566335678101, "learning_rate": 7.859448470038069e-05, "loss": 0.1033, "step": 3410 }, { "grad_norm": 1.0110636949539185, "learning_rate": 7.845868941811956e-05, "loss": 0.1112, "step": 3420 }, { "grad_norm": 0.8129494786262512, "learning_rate": 7.832258291619043e-05, "loss": 0.1177, "step": 3430 }, { "grad_norm": 0.40623798966407776, "learning_rate": 7.81861666830322e-05, "loss": 0.1161, "step": 3440 }, { "grad_norm": 0.3818419873714447, "learning_rate": 7.804944221047097e-05, "loss": 0.0956, "step": 3450 }, { "grad_norm": 0.4899725615978241, "learning_rate": 7.791241099370364e-05, "loss": 0.081, "step": 3460 }, { "grad_norm": 0.5021271109580994, "learning_rate": 7.777507453128163e-05, "loss": 0.0999, "step": 3470 }, { "grad_norm": 0.5628650784492493, "learning_rate": 7.763743432509451e-05, "loss": 0.0981, "step": 3480 }, { "grad_norm": 0.2970348000526428, "learning_rate": 7.749949188035353e-05, "loss": 0.0872, "step": 3490 }, { "grad_norm": 3.434886932373047, "learning_rate": 7.736124870557516e-05, "loss": 0.0832, "step": 3500 }, { "grad_norm": 0.4831876754760742, "learning_rate": 7.722270631256459e-05, "loss": 0.0937, "step": 3510 }, { "grad_norm": 0.4069261848926544, "learning_rate": 7.708386621639925e-05, "loss": 0.0899, "step": 3520 }, { "grad_norm": 0.6318170428276062, "learning_rate": 7.694472993541219e-05, "loss": 0.0991, "step": 3530 }, { "grad_norm": 2.8973448276519775, "learning_rate": 7.680529899117547e-05, "loss": 0.1073, "step": 3540 }, { "grad_norm": 0.7681806683540344, "learning_rate": 7.666557490848358e-05, "loss": 0.0991, "step": 3550 }, { "grad_norm": 0.5337274074554443, "learning_rate": 7.65255592153367e-05, "loss": 0.0904, "step": 3560 }, { "grad_norm": 0.5897412300109863, "learning_rate": 7.638525344292402e-05, "loss": 0.1019, "step": 3570 }, { "grad_norm": 0.4395844340324402, "learning_rate": 7.624465912560697e-05, "loss": 0.1092, "step": 3580 }, { "grad_norm": 0.5541544556617737, "learning_rate": 7.610377780090249e-05, "loss": 0.1105, "step": 3590 }, { "grad_norm": 0.6564070582389832, "learning_rate": 7.596261100946618e-05, "loss": 0.1141, "step": 3600 }, { "grad_norm": 0.6231480836868286, "learning_rate": 7.582116029507542e-05, "loss": 0.0985, "step": 3610 }, { "grad_norm": 0.6433252096176147, "learning_rate": 7.56794272046126e-05, "loss": 0.0941, "step": 3620 }, { "grad_norm": 0.3914188742637634, "learning_rate": 7.55374132880481e-05, "loss": 0.0902, "step": 3630 }, { "grad_norm": 0.5077407956123352, "learning_rate": 7.539512009842333e-05, "loss": 0.0989, "step": 3640 }, { "grad_norm": 0.6180402636528015, "learning_rate": 7.525254919183382e-05, "loss": 0.1018, "step": 3650 }, { "grad_norm": 0.438075989484787, "learning_rate": 7.510970212741215e-05, "loss": 0.0926, "step": 3660 }, { "grad_norm": 1.4384263753890991, "learning_rate": 7.496658046731096e-05, "loss": 0.1012, "step": 3670 }, { "grad_norm": 0.493194043636322, "learning_rate": 7.482318577668578e-05, "loss": 0.0913, "step": 3680 }, { "grad_norm": 0.3817712664604187, "learning_rate": 7.467951962367796e-05, "loss": 0.0982, "step": 3690 }, { "grad_norm": 0.5547468066215515, "learning_rate": 7.453558357939755e-05, "loss": 0.1001, "step": 3700 }, { "grad_norm": 0.48424816131591797, "learning_rate": 7.439137921790606e-05, "loss": 0.0971, "step": 3710 }, { "grad_norm": 0.39593541622161865, "learning_rate": 7.42469081161993e-05, "loss": 0.1085, "step": 3720 }, { "grad_norm": 0.4937167763710022, "learning_rate": 7.410217185419006e-05, "loss": 0.0983, "step": 3730 }, { "grad_norm": 0.509681224822998, "learning_rate": 7.395717201469095e-05, "loss": 0.0987, "step": 3740 }, { "grad_norm": 0.4673025608062744, "learning_rate": 7.381191018339696e-05, "loss": 0.095, "step": 3750 }, { "grad_norm": 0.5049256086349487, "learning_rate": 7.36663879488682e-05, "loss": 0.0878, "step": 3760 }, { "grad_norm": 0.6004349589347839, "learning_rate": 7.352060690251254e-05, "loss": 0.104, "step": 3770 }, { "grad_norm": 0.5094652771949768, "learning_rate": 7.337456863856811e-05, "loss": 0.0965, "step": 3780 }, { "grad_norm": 0.4386717677116394, "learning_rate": 7.3228274754086e-05, "loss": 0.0973, "step": 3790 }, { "grad_norm": 0.6509039998054504, "learning_rate": 7.308172684891267e-05, "loss": 0.131, "step": 3800 }, { "grad_norm": 0.3697807490825653, "learning_rate": 7.293492652567255e-05, "loss": 0.0776, "step": 3810 }, { "grad_norm": 0.6300870180130005, "learning_rate": 7.278787538975043e-05, "loss": 0.1202, "step": 3820 }, { "grad_norm": 0.568928599357605, "learning_rate": 7.2640575049274e-05, "loss": 0.0769, "step": 3830 }, { "grad_norm": 0.5327763557434082, "learning_rate": 7.249302711509616e-05, "loss": 0.0824, "step": 3840 }, { "grad_norm": 0.5911914706230164, "learning_rate": 7.23452332007775e-05, "loss": 0.1154, "step": 3850 }, { "grad_norm": 0.46360528469085693, "learning_rate": 7.219719492256858e-05, "loss": 0.0812, "step": 3860 }, { "grad_norm": 0.42489010095596313, "learning_rate": 7.20489138993923e-05, "loss": 0.0874, "step": 3870 }, { "grad_norm": 0.5154345631599426, "learning_rate": 7.190039175282614e-05, "loss": 0.068, "step": 3880 }, { "grad_norm": 0.4092108905315399, "learning_rate": 7.175163010708455e-05, "loss": 0.0843, "step": 3890 }, { "grad_norm": 0.4314195513725281, "learning_rate": 7.1602630589001e-05, "loss": 0.0922, "step": 3900 }, { "grad_norm": 0.5822885632514954, "learning_rate": 7.14533948280104e-05, "loss": 0.0837, "step": 3910 }, { "grad_norm": 0.6443356275558472, "learning_rate": 7.130392445613109e-05, "loss": 0.1021, "step": 3920 }, { "grad_norm": 0.43977099657058716, "learning_rate": 7.115422110794711e-05, "loss": 0.0835, "step": 3930 }, { "grad_norm": 0.4812179505825043, "learning_rate": 7.100428642059033e-05, "loss": 0.0791, "step": 3940 }, { "grad_norm": 0.4314152002334595, "learning_rate": 7.08541220337224e-05, "loss": 0.0811, "step": 3950 }, { "grad_norm": 0.4761147201061249, "learning_rate": 7.070372958951706e-05, "loss": 0.0922, "step": 3960 }, { "grad_norm": 0.5310417413711548, "learning_rate": 7.055311073264194e-05, "loss": 0.0667, "step": 3970 }, { "grad_norm": 0.44499823451042175, "learning_rate": 7.040226711024077e-05, "loss": 0.1004, "step": 3980 }, { "grad_norm": 0.44270697236061096, "learning_rate": 7.02512003719152e-05, "loss": 0.0946, "step": 3990 }, { "grad_norm": 0.5509179830551147, "learning_rate": 7.00999121697069e-05, "loss": 0.0879, "step": 4000 }, { "grad_norm": 0.3742004334926605, "learning_rate": 6.99484041580794e-05, "loss": 0.0879, "step": 4010 }, { "grad_norm": 1.3894963264465332, "learning_rate": 6.979667799390004e-05, "loss": 0.1073, "step": 4020 }, { "grad_norm": 0.45007896423339844, "learning_rate": 6.964473533642185e-05, "loss": 0.0877, "step": 4030 }, { "grad_norm": 0.754334568977356, "learning_rate": 6.949257784726539e-05, "loss": 0.0874, "step": 4040 }, { "grad_norm": 0.630352258682251, "learning_rate": 6.934020719040056e-05, "loss": 0.0895, "step": 4050 }, { "grad_norm": 0.6873381733894348, "learning_rate": 6.918762503212848e-05, "loss": 0.0881, "step": 4060 }, { "grad_norm": 0.43182840943336487, "learning_rate": 6.903483304106319e-05, "loss": 0.0865, "step": 4070 }, { "grad_norm": 0.37536755204200745, "learning_rate": 6.888183288811341e-05, "loss": 0.0978, "step": 4080 }, { "grad_norm": 0.41249892115592957, "learning_rate": 6.87286262464643e-05, "loss": 0.082, "step": 4090 }, { "grad_norm": 0.526856541633606, "learning_rate": 6.857521479155915e-05, "loss": 0.0936, "step": 4100 }, { "grad_norm": 0.5684453248977661, "learning_rate": 6.842160020108104e-05, "loss": 0.1129, "step": 4110 }, { "grad_norm": 0.5500288605690002, "learning_rate": 6.826778415493455e-05, "loss": 0.1013, "step": 4120 }, { "grad_norm": 0.46216273307800293, "learning_rate": 6.811376833522729e-05, "loss": 0.106, "step": 4130 }, { "grad_norm": 0.4234270453453064, "learning_rate": 6.795955442625159e-05, "loss": 0.1051, "step": 4140 }, { "grad_norm": 0.6794913411140442, "learning_rate": 6.780514411446608e-05, "loss": 0.1224, "step": 4150 }, { "grad_norm": 0.47639283537864685, "learning_rate": 6.765053908847716e-05, "loss": 0.0965, "step": 4160 }, { "grad_norm": 0.5768946409225464, "learning_rate": 6.749574103902064e-05, "loss": 0.0895, "step": 4170 }, { "grad_norm": 0.6823899745941162, "learning_rate": 6.734075165894317e-05, "loss": 0.0885, "step": 4180 }, { "grad_norm": 0.4899390935897827, "learning_rate": 6.71855726431838e-05, "loss": 0.0884, "step": 4190 }, { "grad_norm": 0.5871185064315796, "learning_rate": 6.703020568875538e-05, "loss": 0.0919, "step": 4200 }, { "grad_norm": 0.8028411269187927, "learning_rate": 6.687465249472603e-05, "loss": 0.102, "step": 4210 }, { "grad_norm": 0.31998544931411743, "learning_rate": 6.671891476220055e-05, "loss": 0.1111, "step": 4220 }, { "grad_norm": 0.5346073508262634, "learning_rate": 6.656299419430183e-05, "loss": 0.109, "step": 4230 }, { "grad_norm": 0.42429614067077637, "learning_rate": 6.640689249615223e-05, "loss": 0.0966, "step": 4240 }, { "grad_norm": 0.445103257894516, "learning_rate": 6.625061137485491e-05, "loss": 0.0948, "step": 4250 }, { "grad_norm": 0.46631312370300293, "learning_rate": 6.609415253947517e-05, "loss": 0.0963, "step": 4260 }, { "grad_norm": 0.34065407514572144, "learning_rate": 6.593751770102178e-05, "loss": 0.0998, "step": 4270 }, { "grad_norm": 0.5899305939674377, "learning_rate": 6.578070857242823e-05, "loss": 0.0914, "step": 4280 }, { "grad_norm": 0.39978891611099243, "learning_rate": 6.562372686853402e-05, "loss": 0.106, "step": 4290 }, { "grad_norm": 0.4549619257450104, "learning_rate": 6.546657430606593e-05, "loss": 0.0839, "step": 4300 }, { "grad_norm": 0.5061110258102417, "learning_rate": 6.530925260361918e-05, "loss": 0.1121, "step": 4310 }, { "grad_norm": 0.5547327995300293, "learning_rate": 6.515176348163871e-05, "loss": 0.116, "step": 4320 }, { "grad_norm": 0.33788561820983887, "learning_rate": 6.499410866240032e-05, "loss": 0.0805, "step": 4330 }, { "grad_norm": 0.5540539622306824, "learning_rate": 6.48362898699919e-05, "loss": 0.0905, "step": 4340 }, { "grad_norm": 0.575571596622467, "learning_rate": 6.467830883029443e-05, "loss": 0.0952, "step": 4350 }, { "grad_norm": 0.5708086490631104, "learning_rate": 6.452016727096326e-05, "loss": 0.0838, "step": 4360 }, { "grad_norm": 0.40521448850631714, "learning_rate": 6.436186692140916e-05, "loss": 0.0858, "step": 4370 }, { "grad_norm": 0.4029323160648346, "learning_rate": 6.420340951277938e-05, "loss": 0.087, "step": 4380 }, { "grad_norm": 0.5552173256874084, "learning_rate": 6.404479677793874e-05, "loss": 0.0704, "step": 4390 }, { "grad_norm": 0.41451993584632874, "learning_rate": 6.388603045145075e-05, "loss": 0.0942, "step": 4400 }, { "grad_norm": 0.3800111413002014, "learning_rate": 6.372711226955843e-05, "loss": 0.0914, "step": 4410 }, { "grad_norm": 0.2892295718193054, "learning_rate": 6.356804397016564e-05, "loss": 0.0929, "step": 4420 }, { "grad_norm": 0.7497897744178772, "learning_rate": 6.340882729281779e-05, "loss": 0.0884, "step": 4430 }, { "grad_norm": 0.3840358257293701, "learning_rate": 6.324946397868294e-05, "loss": 0.092, "step": 4440 }, { "grad_norm": 0.4108794331550598, "learning_rate": 6.308995577053276e-05, "loss": 0.0864, "step": 4450 }, { "grad_norm": 0.38527926802635193, "learning_rate": 6.293030441272347e-05, "loss": 0.0766, "step": 4460 }, { "grad_norm": 0.5243363976478577, "learning_rate": 6.277051165117677e-05, "loss": 0.0835, "step": 4470 }, { "grad_norm": 0.3804589807987213, "learning_rate": 6.261057923336064e-05, "loss": 0.0778, "step": 4480 }, { "grad_norm": 0.5868896842002869, "learning_rate": 6.245050890827042e-05, "loss": 0.093, "step": 4490 }, { "grad_norm": 0.5329678058624268, "learning_rate": 6.229030242640952e-05, "loss": 0.0959, "step": 4500 }, { "grad_norm": 0.6053011417388916, "learning_rate": 6.212996153977037e-05, "loss": 0.1069, "step": 4510 }, { "grad_norm": 0.27183061838150024, "learning_rate": 6.196948800181523e-05, "loss": 0.0951, "step": 4520 }, { "grad_norm": 0.5163459777832031, "learning_rate": 6.180888356745695e-05, "loss": 0.0904, "step": 4530 }, { "grad_norm": 0.5805942416191101, "learning_rate": 6.164814999303995e-05, "loss": 0.1018, "step": 4540 }, { "grad_norm": 0.4547424018383026, "learning_rate": 6.148728903632081e-05, "loss": 0.0986, "step": 4550 }, { "grad_norm": 0.6734144687652588, "learning_rate": 6.132630245644921e-05, "loss": 0.0941, "step": 4560 }, { "grad_norm": 0.45316675305366516, "learning_rate": 6.116519201394857e-05, "loss": 0.0973, "step": 4570 }, { "grad_norm": 0.5132927298545837, "learning_rate": 6.10039594706969e-05, "loss": 0.0924, "step": 4580 }, { "grad_norm": 0.688420295715332, "learning_rate": 6.084260658990744e-05, "loss": 0.0956, "step": 4590 }, { "grad_norm": 0.7064704298973083, "learning_rate": 6.068113513610943e-05, "loss": 0.1009, "step": 4600 }, { "grad_norm": 0.6397907733917236, "learning_rate": 6.0519546875128876e-05, "loss": 0.0932, "step": 4610 }, { "grad_norm": 0.4156625270843506, "learning_rate": 6.035784357406906e-05, "loss": 0.0975, "step": 4620 }, { "grad_norm": 0.48149043321609497, "learning_rate": 6.01960270012914e-05, "loss": 0.1078, "step": 4630 }, { "grad_norm": 0.39828813076019287, "learning_rate": 6.003409892639599e-05, "loss": 0.0924, "step": 4640 }, { "grad_norm": 0.4838493764400482, "learning_rate": 5.9872061120202336e-05, "loss": 0.0843, "step": 4650 }, { "grad_norm": 0.34178808331489563, "learning_rate": 5.9709915354729914e-05, "loss": 0.0859, "step": 4660 }, { "grad_norm": 0.7623035907745361, "learning_rate": 5.9547663403178824e-05, "loss": 0.1045, "step": 4670 }, { "grad_norm": 0.6738616228103638, "learning_rate": 5.9385307039910445e-05, "loss": 0.0792, "step": 4680 }, { "grad_norm": 0.4055043160915375, "learning_rate": 5.922284804042792e-05, "loss": 0.1196, "step": 4690 }, { "grad_norm": 0.39592602849006653, "learning_rate": 5.906028818135687e-05, "loss": 0.0919, "step": 4700 }, { "grad_norm": 0.5241408944129944, "learning_rate": 5.889762924042585e-05, "loss": 0.0992, "step": 4710 }, { "grad_norm": 0.7048568725585938, "learning_rate": 5.873487299644699e-05, "loss": 0.0972, "step": 4720 }, { "grad_norm": 0.4049266278743744, "learning_rate": 5.857202122929649e-05, "loss": 0.1022, "step": 4730 }, { "grad_norm": 0.451406329870224, "learning_rate": 5.840907571989518e-05, "loss": 0.0958, "step": 4740 }, { "grad_norm": 0.3363693654537201, "learning_rate": 5.824603825018904e-05, "loss": 0.0883, "step": 4750 }, { "grad_norm": 0.43708670139312744, "learning_rate": 5.808291060312975e-05, "loss": 0.0996, "step": 4760 }, { "grad_norm": 0.5052967071533203, "learning_rate": 5.7919694562655083e-05, "loss": 0.0937, "step": 4770 }, { "grad_norm": 0.6338623762130737, "learning_rate": 5.775639191366954e-05, "loss": 0.0975, "step": 4780 }, { "grad_norm": 0.5449447631835938, "learning_rate": 5.75930044420247e-05, "loss": 0.0974, "step": 4790 }, { "grad_norm": 0.5309737920761108, "learning_rate": 5.74295339344998e-05, "loss": 0.083, "step": 4800 }, { "grad_norm": 1.2652665376663208, "learning_rate": 5.726598217878211e-05, "loss": 0.1004, "step": 4810 }, { "grad_norm": 0.5122917294502258, "learning_rate": 5.71023509634474e-05, "loss": 0.0993, "step": 4820 }, { "grad_norm": 0.4122774004936218, "learning_rate": 5.693864207794049e-05, "loss": 0.0938, "step": 4830 }, { "grad_norm": 0.28626084327697754, "learning_rate": 5.677485731255545e-05, "loss": 0.0835, "step": 4840 }, { "grad_norm": 0.345565527677536, "learning_rate": 5.6610998458416296e-05, "loss": 0.0737, "step": 4850 }, { "grad_norm": 0.5571615099906921, "learning_rate": 5.644706730745716e-05, "loss": 0.0848, "step": 4860 }, { "grad_norm": 0.34856364130973816, "learning_rate": 5.628306565240287e-05, "loss": 0.0775, "step": 4870 }, { "grad_norm": 0.45140308141708374, "learning_rate": 5.611899528674923e-05, "loss": 0.095, "step": 4880 }, { "grad_norm": 0.47641852498054504, "learning_rate": 5.595485800474349e-05, "loss": 0.0963, "step": 4890 }, { "grad_norm": 0.8866299390792847, "learning_rate": 5.579065560136467e-05, "loss": 0.0809, "step": 4900 }, { "grad_norm": 0.7897559404373169, "learning_rate": 5.562638987230392e-05, "loss": 0.0864, "step": 4910 }, { "grad_norm": 0.4387589395046234, "learning_rate": 5.546206261394498e-05, "loss": 0.0926, "step": 4920 }, { "grad_norm": 0.8544060587882996, "learning_rate": 5.529767562334437e-05, "loss": 0.0825, "step": 4930 }, { "grad_norm": 0.5818145871162415, "learning_rate": 5.5133230698211926e-05, "loss": 0.0797, "step": 4940 }, { "grad_norm": 0.5924447774887085, "learning_rate": 5.496872963689096e-05, "loss": 0.086, "step": 4950 }, { "grad_norm": 0.681376039981842, "learning_rate": 5.4804174238338756e-05, "loss": 0.0869, "step": 4960 }, { "grad_norm": 0.5320497155189514, "learning_rate": 5.463956630210678e-05, "loss": 0.0965, "step": 4970 }, { "grad_norm": 0.41286584734916687, "learning_rate": 5.4474907628321046e-05, "loss": 0.0921, "step": 4980 }, { "grad_norm": 0.47107991576194763, "learning_rate": 5.431020001766244e-05, "loss": 0.0889, "step": 4990 }, { "grad_norm": 0.39125165343284607, "learning_rate": 5.4145445271346986e-05, "loss": 0.1056, "step": 5000 }, { "grad_norm": 0.6862531900405884, "learning_rate": 5.398064519110622e-05, "loss": 0.094, "step": 5010 }, { "grad_norm": 0.350613534450531, "learning_rate": 5.3815801579167394e-05, "loss": 0.0782, "step": 5020 }, { "grad_norm": 0.3431878685951233, "learning_rate": 5.365091623823382e-05, "loss": 0.0878, "step": 5030 }, { "grad_norm": 0.44780752062797546, "learning_rate": 5.348599097146521e-05, "loss": 0.0756, "step": 5040 }, { "grad_norm": 0.39826542139053345, "learning_rate": 5.3321027582457836e-05, "loss": 0.0815, "step": 5050 }, { "grad_norm": 0.49251148104667664, "learning_rate": 5.315602787522491e-05, "loss": 0.0694, "step": 5060 }, { "grad_norm": 0.6309869289398193, "learning_rate": 5.299099365417678e-05, "loss": 0.0925, "step": 5070 }, { "grad_norm": 0.5354300737380981, "learning_rate": 5.2825926724101236e-05, "loss": 0.0872, "step": 5080 }, { "grad_norm": 0.5926438570022583, "learning_rate": 5.26608288901438e-05, "loss": 0.0685, "step": 5090 }, { "grad_norm": 0.7714500427246094, "learning_rate": 5.24957019577879e-05, "loss": 0.0873, "step": 5100 }, { "grad_norm": 0.6278719902038574, "learning_rate": 5.2330547732835266e-05, "loss": 0.0893, "step": 5110 }, { "grad_norm": 0.6609734892845154, "learning_rate": 5.2165368021385996e-05, "loss": 0.1001, "step": 5120 }, { "grad_norm": 0.4641166627407074, "learning_rate": 5.200016462981897e-05, "loss": 0.0911, "step": 5130 }, { "grad_norm": 0.6038467288017273, "learning_rate": 5.1834939364772015e-05, "loss": 0.098, "step": 5140 }, { "grad_norm": 0.41543564200401306, "learning_rate": 5.166969403312214e-05, "loss": 0.0825, "step": 5150 }, { "grad_norm": 0.574292778968811, "learning_rate": 5.1504430441965844e-05, "loss": 0.1121, "step": 5160 }, { "grad_norm": 0.7103983759880066, "learning_rate": 5.133915039859923e-05, "loss": 0.0876, "step": 5170 }, { "grad_norm": 0.34789013862609863, "learning_rate": 5.1173855710498444e-05, "loss": 0.0729, "step": 5180 }, { "grad_norm": 0.4640834331512451, "learning_rate": 5.100854818529967e-05, "loss": 0.1018, "step": 5190 }, { "grad_norm": 0.29679492115974426, "learning_rate": 5.084322963077951e-05, "loss": 0.0697, "step": 5200 }, { "grad_norm": 0.4756428599357605, "learning_rate": 5.067790185483522e-05, "loss": 0.0883, "step": 5210 }, { "grad_norm": 0.37873443961143494, "learning_rate": 5.0512566665464844e-05, "loss": 0.0932, "step": 5220 }, { "grad_norm": 0.3493492305278778, "learning_rate": 5.034722587074755e-05, "loss": 0.0718, "step": 5230 }, { "grad_norm": 0.8235279321670532, "learning_rate": 5.018188127882375e-05, "loss": 0.0762, "step": 5240 }, { "grad_norm": 0.4805556535720825, "learning_rate": 5.0016534697875417e-05, "loss": 0.1056, "step": 5250 }, { "grad_norm": 0.5494657158851624, "learning_rate": 4.9851187936106294e-05, "loss": 0.0828, "step": 5260 }, { "grad_norm": 0.4475446045398712, "learning_rate": 4.968584280172206e-05, "loss": 0.0952, "step": 5270 }, { "grad_norm": 0.4109076261520386, "learning_rate": 4.95205011029106e-05, "loss": 0.0908, "step": 5280 }, { "grad_norm": 0.451121062040329, "learning_rate": 4.935516464782227e-05, "loss": 0.0807, "step": 5290 }, { "grad_norm": 0.34987643361091614, "learning_rate": 4.918983524455003e-05, "loss": 0.0728, "step": 5300 }, { "grad_norm": 0.5376846194267273, "learning_rate": 4.9024514701109766e-05, "loss": 0.0842, "step": 5310 }, { "grad_norm": 0.3152400255203247, "learning_rate": 4.885920482542043e-05, "loss": 0.0833, "step": 5320 }, { "grad_norm": 0.39066582918167114, "learning_rate": 4.869390742528438e-05, "loss": 0.0912, "step": 5330 }, { "grad_norm": 0.45404544472694397, "learning_rate": 4.852862430836744e-05, "loss": 0.0772, "step": 5340 }, { "grad_norm": 0.5292527079582214, "learning_rate": 4.836335728217933e-05, "loss": 0.0733, "step": 5350 }, { "grad_norm": 0.4296908378601074, "learning_rate": 4.819810815405379e-05, "loss": 0.0781, "step": 5360 }, { "grad_norm": 0.6098202466964722, "learning_rate": 4.803287873112877e-05, "loss": 0.0871, "step": 5370 }, { "grad_norm": 0.41479721665382385, "learning_rate": 4.786767082032681e-05, "loss": 0.0696, "step": 5380 }, { "grad_norm": 0.4190218448638916, "learning_rate": 4.77024862283351e-05, "loss": 0.0779, "step": 5390 }, { "grad_norm": 0.3043394088745117, "learning_rate": 4.753732676158593e-05, "loss": 0.0835, "step": 5400 }, { "grad_norm": 0.8789849877357483, "learning_rate": 4.737219422623672e-05, "loss": 0.0763, "step": 5410 }, { "grad_norm": 0.7131383419036865, "learning_rate": 4.720709042815044e-05, "loss": 0.0811, "step": 5420 }, { "grad_norm": 0.5399758219718933, "learning_rate": 4.704201717287578e-05, "loss": 0.0787, "step": 5430 }, { "grad_norm": 0.5642313361167908, "learning_rate": 4.6876976265627404e-05, "loss": 0.0983, "step": 5440 }, { "grad_norm": 0.5702924132347107, "learning_rate": 4.671196951126626e-05, "loss": 0.0734, "step": 5450 }, { "grad_norm": 0.4508060812950134, "learning_rate": 4.654699871427971e-05, "loss": 0.0841, "step": 5460 }, { "grad_norm": 0.3975818455219269, "learning_rate": 4.6382065678762034e-05, "loss": 0.0805, "step": 5470 }, { "grad_norm": 0.7214745879173279, "learning_rate": 4.6217172208394424e-05, "loss": 0.0825, "step": 5480 }, { "grad_norm": 0.3697737157344818, "learning_rate": 4.605232010642549e-05, "loss": 0.0768, "step": 5490 }, { "grad_norm": 0.360867440700531, "learning_rate": 4.588751117565142e-05, "loss": 0.0811, "step": 5500 }, { "grad_norm": 0.39117133617401123, "learning_rate": 4.5722747218396214e-05, "loss": 0.0832, "step": 5510 }, { "grad_norm": 0.3379577696323395, "learning_rate": 4.5558030036492194e-05, "loss": 0.063, "step": 5520 }, { "grad_norm": 0.49066630005836487, "learning_rate": 4.539336143125999e-05, "loss": 0.0854, "step": 5530 }, { "grad_norm": 0.3705296218395233, "learning_rate": 4.522874320348916e-05, "loss": 0.0687, "step": 5540 }, { "grad_norm": 0.6204256415367126, "learning_rate": 4.506417715341821e-05, "loss": 0.0968, "step": 5550 }, { "grad_norm": 0.5469505190849304, "learning_rate": 4.489966508071511e-05, "loss": 0.0831, "step": 5560 }, { "grad_norm": 0.3545830249786377, "learning_rate": 4.4735208784457575e-05, "loss": 0.0824, "step": 5570 }, { "grad_norm": 0.44615694880485535, "learning_rate": 4.457081006311325e-05, "loss": 0.0911, "step": 5580 }, { "grad_norm": 0.4059657156467438, "learning_rate": 4.440647071452027e-05, "loss": 0.0773, "step": 5590 }, { "grad_norm": 0.37141457200050354, "learning_rate": 4.424219253586737e-05, "loss": 0.0833, "step": 5600 }, { "grad_norm": 0.5122745037078857, "learning_rate": 4.407797732367443e-05, "loss": 0.085, "step": 5610 }, { "grad_norm": 0.3963432013988495, "learning_rate": 4.391382687377268e-05, "loss": 0.0696, "step": 5620 }, { "grad_norm": 0.3497571647167206, "learning_rate": 4.374974298128512e-05, "loss": 0.0685, "step": 5630 }, { "grad_norm": 0.3579045832157135, "learning_rate": 4.358572744060699e-05, "loss": 0.0732, "step": 5640 }, { "grad_norm": 0.4091202914714813, "learning_rate": 4.342178204538588e-05, "loss": 0.0945, "step": 5650 }, { "grad_norm": 0.3646412789821625, "learning_rate": 4.325790858850241e-05, "loss": 0.0728, "step": 5660 }, { "grad_norm": 0.4868878424167633, "learning_rate": 4.309410886205043e-05, "loss": 0.0903, "step": 5670 }, { "grad_norm": 0.3419983386993408, "learning_rate": 4.293038465731752e-05, "loss": 0.0918, "step": 5680 }, { "grad_norm": 0.57663494348526, "learning_rate": 4.276673776476533e-05, "loss": 0.0782, "step": 5690 }, { "grad_norm": 0.32256126403808594, "learning_rate": 4.260316997401007e-05, "loss": 0.0772, "step": 5700 }, { "grad_norm": 0.4896562695503235, "learning_rate": 4.243968307380293e-05, "loss": 0.0844, "step": 5710 }, { "grad_norm": 0.44715356826782227, "learning_rate": 4.22762788520104e-05, "loss": 0.0853, "step": 5720 }, { "grad_norm": 0.5560845136642456, "learning_rate": 4.211295909559491e-05, "loss": 0.0714, "step": 5730 }, { "grad_norm": 0.41671445965766907, "learning_rate": 4.194972559059511e-05, "loss": 0.0952, "step": 5740 }, { "grad_norm": 0.3581355810165405, "learning_rate": 4.178658012210651e-05, "loss": 0.0799, "step": 5750 }, { "grad_norm": 0.6145849823951721, "learning_rate": 4.162352447426177e-05, "loss": 0.0778, "step": 5760 }, { "grad_norm": 0.46420586109161377, "learning_rate": 4.146056043021135e-05, "loss": 0.0848, "step": 5770 }, { "grad_norm": 0.6919949054718018, "learning_rate": 4.1297689772103944e-05, "loss": 0.1014, "step": 5780 }, { "grad_norm": 0.5602481365203857, "learning_rate": 4.113491428106694e-05, "loss": 0.0716, "step": 5790 }, { "grad_norm": 0.525973379611969, "learning_rate": 4.0972235737187055e-05, "loss": 0.0958, "step": 5800 }, { "grad_norm": 0.6218112707138062, "learning_rate": 4.080965591949076e-05, "loss": 0.0852, "step": 5810 }, { "grad_norm": 0.4099525809288025, "learning_rate": 4.0647176605924924e-05, "loss": 0.0856, "step": 5820 }, { "grad_norm": 0.36325082182884216, "learning_rate": 4.0484799573337255e-05, "loss": 0.0864, "step": 5830 }, { "grad_norm": 0.4957577884197235, "learning_rate": 4.032252659745699e-05, "loss": 0.0748, "step": 5840 }, { "grad_norm": 0.5453797578811646, "learning_rate": 4.016035945287539e-05, "loss": 0.0764, "step": 5850 }, { "grad_norm": 0.4988965690135956, "learning_rate": 3.999829991302635e-05, "loss": 0.0823, "step": 5860 }, { "grad_norm": 0.2878781855106354, "learning_rate": 3.983634975016707e-05, "loss": 0.0732, "step": 5870 }, { "grad_norm": 0.3241180181503296, "learning_rate": 3.967451073535854e-05, "loss": 0.0823, "step": 5880 }, { "grad_norm": 0.4814620614051819, "learning_rate": 3.951278463844633e-05, "loss": 0.0958, "step": 5890 }, { "grad_norm": 0.497734010219574, "learning_rate": 3.935117322804111e-05, "loss": 0.1198, "step": 5900 }, { "grad_norm": 0.4364153742790222, "learning_rate": 3.918967827149938e-05, "loss": 0.0956, "step": 5910 }, { "grad_norm": 0.3721391558647156, "learning_rate": 3.9028301534904094e-05, "loss": 0.0769, "step": 5920 }, { "grad_norm": 0.3159007132053375, "learning_rate": 3.88670447830454e-05, "loss": 0.0578, "step": 5930 }, { "grad_norm": 0.5735059380531311, "learning_rate": 3.870590977940132e-05, "loss": 0.0835, "step": 5940 }, { "grad_norm": 0.7289217114448547, "learning_rate": 3.8544898286118404e-05, "loss": 0.0902, "step": 5950 }, { "grad_norm": 0.3191511034965515, "learning_rate": 3.838401206399257e-05, "loss": 0.0677, "step": 5960 }, { "grad_norm": 0.5466225743293762, "learning_rate": 3.822325287244975e-05, "loss": 0.0969, "step": 5970 }, { "grad_norm": 0.4763900637626648, "learning_rate": 3.8062622469526725e-05, "loss": 0.0954, "step": 5980 }, { "grad_norm": 0.512734055519104, "learning_rate": 3.790212261185183e-05, "loss": 0.0914, "step": 5990 }, { "grad_norm": 0.38841450214385986, "learning_rate": 3.7741755054625794e-05, "loss": 0.1079, "step": 6000 }, { "grad_norm": 0.3968331515789032, "learning_rate": 3.758152155160255e-05, "loss": 0.0993, "step": 6010 }, { "grad_norm": 0.546953558921814, "learning_rate": 3.742142385506999e-05, "loss": 0.0746, "step": 6020 }, { "grad_norm": 0.4161210358142853, "learning_rate": 3.72614637158309e-05, "loss": 0.0822, "step": 6030 }, { "grad_norm": 0.504655659198761, "learning_rate": 3.710164288318371e-05, "loss": 0.0776, "step": 6040 }, { "grad_norm": 0.5824733972549438, "learning_rate": 3.694196310490345e-05, "loss": 0.0812, "step": 6050 }, { "grad_norm": 0.3908507525920868, "learning_rate": 3.678242612722259e-05, "loss": 0.0662, "step": 6060 }, { "grad_norm": 0.3547900915145874, "learning_rate": 3.6623033694811953e-05, "loss": 0.0785, "step": 6070 }, { "grad_norm": 0.38695043325424194, "learning_rate": 3.6463787550761665e-05, "loss": 0.0909, "step": 6080 }, { "grad_norm": 0.3548903167247772, "learning_rate": 3.630468943656202e-05, "loss": 0.0934, "step": 6090 }, { "grad_norm": 0.45635277032852173, "learning_rate": 3.6145741092084523e-05, "loss": 0.0847, "step": 6100 }, { "grad_norm": 0.422884464263916, "learning_rate": 3.598694425556278e-05, "loss": 0.0728, "step": 6110 }, { "grad_norm": 0.4719564616680145, "learning_rate": 3.58283006635736e-05, "loss": 0.0828, "step": 6120 }, { "grad_norm": 0.4873234033584595, "learning_rate": 3.566981205101781e-05, "loss": 0.0619, "step": 6130 }, { "grad_norm": 0.3228538930416107, "learning_rate": 3.5511480151101556e-05, "loss": 0.0817, "step": 6140 }, { "grad_norm": 0.40462708473205566, "learning_rate": 3.5353306695317104e-05, "loss": 0.0958, "step": 6150 }, { "grad_norm": 0.445445716381073, "learning_rate": 3.519529341342402e-05, "loss": 0.0978, "step": 6160 }, { "grad_norm": 0.4996490478515625, "learning_rate": 3.503744203343026e-05, "loss": 0.078, "step": 6170 }, { "grad_norm": 0.3199898898601532, "learning_rate": 3.487975428157318e-05, "loss": 0.0652, "step": 6180 }, { "grad_norm": 0.26956820487976074, "learning_rate": 3.472223188230083e-05, "loss": 0.0882, "step": 6190 }, { "grad_norm": 0.5824195146560669, "learning_rate": 3.4564876558252866e-05, "loss": 0.078, "step": 6200 }, { "grad_norm": 0.46209609508514404, "learning_rate": 3.440769003024195e-05, "loss": 0.0867, "step": 6210 }, { "grad_norm": 0.37930580973625183, "learning_rate": 3.425067401723477e-05, "loss": 0.0891, "step": 6220 }, { "grad_norm": 0.2800642251968384, "learning_rate": 3.409383023633325e-05, "loss": 0.0757, "step": 6230 }, { "grad_norm": 0.5696620345115662, "learning_rate": 3.3937160402755894e-05, "loss": 0.0773, "step": 6240 }, { "grad_norm": 0.4432618021965027, "learning_rate": 3.378066622981885e-05, "loss": 0.074, "step": 6250 }, { "grad_norm": 0.433745414018631, "learning_rate": 3.362434942891738e-05, "loss": 0.0761, "step": 6260 }, { "grad_norm": 0.5789763331413269, "learning_rate": 3.346821170950693e-05, "loss": 0.0785, "step": 6270 }, { "grad_norm": 0.3254248797893524, "learning_rate": 3.3312254779084585e-05, "loss": 0.0796, "step": 6280 }, { "grad_norm": 0.44405749440193176, "learning_rate": 3.315648034317039e-05, "loss": 0.0875, "step": 6290 }, { "grad_norm": 0.4471359848976135, "learning_rate": 3.3000890105288564e-05, "loss": 0.0698, "step": 6300 }, { "grad_norm": 0.32520610094070435, "learning_rate": 3.284548576694908e-05, "loss": 0.0671, "step": 6310 }, { "grad_norm": 0.4187530279159546, "learning_rate": 3.2690269027628815e-05, "loss": 0.069, "step": 6320 }, { "grad_norm": 0.3989262580871582, "learning_rate": 3.253524158475324e-05, "loss": 0.068, "step": 6330 }, { "grad_norm": 0.5916138887405396, "learning_rate": 3.238040513367757e-05, "loss": 0.0882, "step": 6340 }, { "grad_norm": 0.4049164950847626, "learning_rate": 3.222576136766843e-05, "loss": 0.0708, "step": 6350 }, { "grad_norm": 0.27160242199897766, "learning_rate": 3.2071311977885324e-05, "loss": 0.0817, "step": 6360 }, { "grad_norm": 0.44213786721229553, "learning_rate": 3.191705865336197e-05, "loss": 0.07, "step": 6370 }, { "grad_norm": 0.36541691422462463, "learning_rate": 3.1763003080988075e-05, "loss": 0.0635, "step": 6380 }, { "grad_norm": 0.41078269481658936, "learning_rate": 3.160914694549063e-05, "loss": 0.0778, "step": 6390 }, { "grad_norm": 0.5698886513710022, "learning_rate": 3.145549192941573e-05, "loss": 0.0858, "step": 6400 }, { "grad_norm": 0.29118743538856506, "learning_rate": 3.130203971310999e-05, "loss": 0.0723, "step": 6410 }, { "grad_norm": 0.2796599566936493, "learning_rate": 3.114879197470225e-05, "loss": 0.0755, "step": 6420 }, { "grad_norm": 0.3855741024017334, "learning_rate": 3.0995750390085285e-05, "loss": 0.064, "step": 6430 }, { "grad_norm": 0.3842144310474396, "learning_rate": 3.084291663289728e-05, "loss": 0.0755, "step": 6440 }, { "grad_norm": 0.4576573669910431, "learning_rate": 3.069029237450375e-05, "loss": 0.0838, "step": 6450 }, { "grad_norm": 0.22724193334579468, "learning_rate": 3.053787928397911e-05, "loss": 0.0781, "step": 6460 }, { "grad_norm": 0.4728739857673645, "learning_rate": 3.0385679028088526e-05, "loss": 0.0732, "step": 6470 }, { "grad_norm": 0.4177311062812805, "learning_rate": 3.023369327126959e-05, "loss": 0.0595, "step": 6480 }, { "grad_norm": 0.4850885570049286, "learning_rate": 3.0081923675614198e-05, "loss": 0.0819, "step": 6490 }, { "grad_norm": 0.5045800805091858, "learning_rate": 2.993037190085034e-05, "loss": 0.0811, "step": 6500 }, { "grad_norm": 0.2982923686504364, "learning_rate": 2.977903960432392e-05, "loss": 0.0755, "step": 6510 }, { "grad_norm": 0.4610765874385834, "learning_rate": 2.9627928440980722e-05, "loss": 0.0688, "step": 6520 }, { "grad_norm": 0.5172238349914551, "learning_rate": 2.9477040063348183e-05, "loss": 0.0754, "step": 6530 }, { "grad_norm": 0.3337225914001465, "learning_rate": 2.9326376121517456e-05, "loss": 0.0939, "step": 6540 }, { "grad_norm": 0.5745890736579895, "learning_rate": 2.9175938263125236e-05, "loss": 0.0749, "step": 6550 }, { "grad_norm": 0.3162401616573334, "learning_rate": 2.9025728133335873e-05, "loss": 0.0714, "step": 6560 }, { "grad_norm": 0.35456663370132446, "learning_rate": 2.8875747374823288e-05, "loss": 0.0809, "step": 6570 }, { "grad_norm": 0.4011017978191376, "learning_rate": 2.872599762775298e-05, "loss": 0.0695, "step": 6580 }, { "grad_norm": 0.44703322649002075, "learning_rate": 2.857648052976425e-05, "loss": 0.0797, "step": 6590 }, { "grad_norm": 0.4158497452735901, "learning_rate": 2.8427197715952047e-05, "loss": 0.0957, "step": 6600 }, { "grad_norm": 0.4107862710952759, "learning_rate": 2.8278150818849393e-05, "loss": 0.0741, "step": 6610 }, { "grad_norm": 1.4406017065048218, "learning_rate": 2.812934146840922e-05, "loss": 0.0854, "step": 6620 }, { "grad_norm": 0.46371498703956604, "learning_rate": 2.7980771291986764e-05, "loss": 0.0723, "step": 6630 }, { "grad_norm": 0.7529252767562866, "learning_rate": 2.783244191432167e-05, "loss": 0.0628, "step": 6640 }, { "grad_norm": 0.3839459717273712, "learning_rate": 2.768435495752022e-05, "loss": 0.092, "step": 6650 }, { "grad_norm": 0.3733179271221161, "learning_rate": 2.753651204103771e-05, "loss": 0.0619, "step": 6660 }, { "grad_norm": 0.3301064968109131, "learning_rate": 2.7388914781660523e-05, "loss": 0.0683, "step": 6670 }, { "grad_norm": 0.42661985754966736, "learning_rate": 2.7241564793488693e-05, "loss": 0.0932, "step": 6680 }, { "grad_norm": 0.3856300711631775, "learning_rate": 2.7094463687918037e-05, "loss": 0.0887, "step": 6690 }, { "grad_norm": 0.3584440350532532, "learning_rate": 2.694761307362268e-05, "loss": 0.0718, "step": 6700 }, { "grad_norm": 0.41910549998283386, "learning_rate": 2.6801014556537467e-05, "loss": 0.0843, "step": 6710 }, { "grad_norm": 0.3737851679325104, "learning_rate": 2.6654669739840243e-05, "loss": 0.0853, "step": 6720 }, { "grad_norm": 0.5173537135124207, "learning_rate": 2.650858022393451e-05, "loss": 0.076, "step": 6730 }, { "grad_norm": 0.35249125957489014, "learning_rate": 2.6362747606431747e-05, "loss": 0.1061, "step": 6740 }, { "grad_norm": 0.4537646472454071, "learning_rate": 2.6217173482134172e-05, "loss": 0.0767, "step": 6750 }, { "grad_norm": 1.344586968421936, "learning_rate": 2.6071859443017044e-05, "loss": 0.0713, "step": 6760 }, { "grad_norm": 0.37521737813949585, "learning_rate": 2.5926807078211414e-05, "loss": 0.0727, "step": 6770 }, { "grad_norm": 0.3582833409309387, "learning_rate": 2.5782017973986728e-05, "loss": 0.0818, "step": 6780 }, { "grad_norm": 0.5952498316764832, "learning_rate": 2.5637493713733374e-05, "loss": 0.0766, "step": 6790 }, { "grad_norm": 0.40153881907463074, "learning_rate": 2.549323587794559e-05, "loss": 0.0822, "step": 6800 }, { "grad_norm": 0.4498429298400879, "learning_rate": 2.5349246044203895e-05, "loss": 0.0784, "step": 6810 }, { "grad_norm": 0.4070703685283661, "learning_rate": 2.520552578715808e-05, "loss": 0.0717, "step": 6820 }, { "grad_norm": 0.5056887269020081, "learning_rate": 2.506207667850981e-05, "loss": 0.0781, "step": 6830 }, { "grad_norm": 0.36663922667503357, "learning_rate": 2.4918900286995555e-05, "loss": 0.0772, "step": 6840 }, { "grad_norm": 0.31992608308792114, "learning_rate": 2.4775998178369458e-05, "loss": 0.0761, "step": 6850 }, { "grad_norm": 0.38594529032707214, "learning_rate": 2.4633371915386017e-05, "loss": 0.0544, "step": 6860 }, { "grad_norm": 0.37852048873901367, "learning_rate": 2.4491023057783235e-05, "loss": 0.0761, "step": 6870 }, { "grad_norm": 0.3775743842124939, "learning_rate": 2.4348953162265375e-05, "loss": 0.0817, "step": 6880 }, { "grad_norm": 0.4313925802707672, "learning_rate": 2.420716378248607e-05, "loss": 0.0783, "step": 6890 }, { "grad_norm": 0.4117332696914673, "learning_rate": 2.4065656469031266e-05, "loss": 0.077, "step": 6900 }, { "grad_norm": 0.319731205701828, "learning_rate": 2.3924432769402268e-05, "loss": 0.0777, "step": 6910 }, { "grad_norm": 0.4458310008049011, "learning_rate": 2.3783494227998844e-05, "loss": 0.0702, "step": 6920 }, { "grad_norm": 0.25544869899749756, "learning_rate": 2.3642842386102264e-05, "loss": 0.0712, "step": 6930 }, { "grad_norm": 0.3450974225997925, "learning_rate": 2.3502478781858567e-05, "loss": 0.0705, "step": 6940 }, { "grad_norm": 0.48896971344947815, "learning_rate": 2.3362404950261628e-05, "loss": 0.0605, "step": 6950 }, { "grad_norm": 0.6590002179145813, "learning_rate": 2.3222622423136458e-05, "loss": 0.0684, "step": 6960 }, { "grad_norm": 0.4526193141937256, "learning_rate": 2.3083132729122332e-05, "loss": 0.0821, "step": 6970 }, { "grad_norm": 0.4742291569709778, "learning_rate": 2.294393739365621e-05, "loss": 0.0775, "step": 6980 }, { "grad_norm": 0.6493641138076782, "learning_rate": 2.2805037938956e-05, "loss": 0.0656, "step": 6990 }, { "grad_norm": 0.5979931950569153, "learning_rate": 2.266643588400386e-05, "loss": 0.079, "step": 7000 }, { "grad_norm": 0.32816651463508606, "learning_rate": 2.252813274452969e-05, "loss": 0.09, "step": 7010 }, { "grad_norm": 0.42768433690071106, "learning_rate": 2.2390130032994427e-05, "loss": 0.0645, "step": 7020 }, { "grad_norm": 0.38517889380455017, "learning_rate": 2.2252429258573633e-05, "loss": 0.0755, "step": 7030 }, { "grad_norm": 0.6428894400596619, "learning_rate": 2.2115031927140904e-05, "loss": 0.0705, "step": 7040 }, { "grad_norm": 0.4560980498790741, "learning_rate": 2.1977939541251463e-05, "loss": 0.0666, "step": 7050 }, { "grad_norm": 0.3638777434825897, "learning_rate": 2.1841153600125684e-05, "loss": 0.0659, "step": 7060 }, { "grad_norm": 0.3320036232471466, "learning_rate": 2.170467559963267e-05, "loss": 0.0653, "step": 7070 }, { "grad_norm": 0.5185064077377319, "learning_rate": 2.1568507032273982e-05, "loss": 0.094, "step": 7080 }, { "grad_norm": 0.4568357765674591, "learning_rate": 2.1432649387167264e-05, "loss": 0.0749, "step": 7090 }, { "grad_norm": 0.5098255276679993, "learning_rate": 2.1297104150029973e-05, "loss": 0.0753, "step": 7100 }, { "grad_norm": 0.4780910313129425, "learning_rate": 2.116187280316307e-05, "loss": 0.0823, "step": 7110 }, { "grad_norm": 0.356631875038147, "learning_rate": 2.1026956825434908e-05, "loss": 0.068, "step": 7120 }, { "grad_norm": 1.1051487922668457, "learning_rate": 2.0892357692265017e-05, "loss": 0.0703, "step": 7130 }, { "grad_norm": 0.4998587965965271, "learning_rate": 2.0758076875607947e-05, "loss": 0.0858, "step": 7140 }, { "grad_norm": 0.3493131995201111, "learning_rate": 2.0624115843937207e-05, "loss": 0.0868, "step": 7150 }, { "grad_norm": 0.631033182144165, "learning_rate": 2.0490476062229157e-05, "loss": 0.0782, "step": 7160 }, { "grad_norm": 0.35530272126197815, "learning_rate": 2.035715899194704e-05, "loss": 0.0735, "step": 7170 }, { "grad_norm": 0.40505343675613403, "learning_rate": 2.022416609102499e-05, "loss": 0.061, "step": 7180 }, { "grad_norm": 0.632790744304657, "learning_rate": 2.009149881385205e-05, "loss": 0.0804, "step": 7190 }, { "grad_norm": 0.3478289544582367, "learning_rate": 1.995915861125634e-05, "loss": 0.0644, "step": 7200 }, { "grad_norm": 0.3802562952041626, "learning_rate": 1.9827146930489065e-05, "loss": 0.0623, "step": 7210 }, { "grad_norm": 0.34591779112815857, "learning_rate": 1.9695465215208848e-05, "loss": 0.0757, "step": 7220 }, { "grad_norm": 0.3096807301044464, "learning_rate": 1.9564114905465813e-05, "loss": 0.0594, "step": 7230 }, { "grad_norm": 0.49927088618278503, "learning_rate": 1.9433097437685936e-05, "loss": 0.0726, "step": 7240 }, { "grad_norm": 0.27938950061798096, "learning_rate": 1.930241424465521e-05, "loss": 0.0743, "step": 7250 }, { "grad_norm": 0.40096601843833923, "learning_rate": 1.9172066755504115e-05, "loss": 0.0672, "step": 7260 }, { "grad_norm": 0.5010851621627808, "learning_rate": 1.9042056395691914e-05, "loss": 0.0605, "step": 7270 }, { "grad_norm": 0.5956852436065674, "learning_rate": 1.8912384586991066e-05, "loss": 0.067, "step": 7280 }, { "grad_norm": 0.3866286277770996, "learning_rate": 1.8783052747471717e-05, "loss": 0.0701, "step": 7290 }, { "grad_norm": 0.42603155970573425, "learning_rate": 1.865406229148611e-05, "loss": 0.0706, "step": 7300 }, { "grad_norm": 0.3312182128429413, "learning_rate": 1.8525414629653233e-05, "loss": 0.069, "step": 7310 }, { "grad_norm": 0.3676312267780304, "learning_rate": 1.8397111168843255e-05, "loss": 0.0538, "step": 7320 }, { "grad_norm": 0.422480970621109, "learning_rate": 1.8269153312162323e-05, "loss": 0.0696, "step": 7330 }, { "grad_norm": 0.44964325428009033, "learning_rate": 1.8141542458937054e-05, "loss": 0.061, "step": 7340 }, { "grad_norm": 0.3502974510192871, "learning_rate": 1.8014280004699268e-05, "loss": 0.0637, "step": 7350 }, { "grad_norm": 0.4156774580478668, "learning_rate": 1.788736734117078e-05, "loss": 0.0757, "step": 7360 }, { "grad_norm": 0.4451335668563843, "learning_rate": 1.7760805856248152e-05, "loss": 0.068, "step": 7370 }, { "grad_norm": 0.6710352897644043, "learning_rate": 1.7634596933987518e-05, "loss": 0.0802, "step": 7380 }, { "grad_norm": 0.5433879494667053, "learning_rate": 1.7508741954589404e-05, "loss": 0.0715, "step": 7390 }, { "grad_norm": 0.4147975444793701, "learning_rate": 1.7383242294383717e-05, "loss": 0.0833, "step": 7400 }, { "grad_norm": 0.3811854124069214, "learning_rate": 1.7258099325814632e-05, "loss": 0.0737, "step": 7410 }, { "grad_norm": 0.3429500162601471, "learning_rate": 1.7133314417425594e-05, "loss": 0.0778, "step": 7420 }, { "grad_norm": 0.3635982275009155, "learning_rate": 1.7008888933844408e-05, "loss": 0.082, "step": 7430 }, { "grad_norm": 0.4766620099544525, "learning_rate": 1.6884824235768172e-05, "loss": 0.0712, "step": 7440 }, { "grad_norm": 0.30849799513816833, "learning_rate": 1.6761121679948592e-05, "loss": 0.0638, "step": 7450 }, { "grad_norm": 0.3875696659088135, "learning_rate": 1.663778261917695e-05, "loss": 0.0689, "step": 7460 }, { "grad_norm": 0.36167389154434204, "learning_rate": 1.651480840226952e-05, "loss": 0.0824, "step": 7470 }, { "grad_norm": 0.454254150390625, "learning_rate": 1.639220037405258e-05, "loss": 0.0823, "step": 7480 }, { "grad_norm": 0.43230605125427246, "learning_rate": 1.6269959875347906e-05, "loss": 0.0751, "step": 7490 }, { "grad_norm": 0.3847936689853668, "learning_rate": 1.614808824295802e-05, "loss": 0.0667, "step": 7500 }, { "grad_norm": 0.31421366333961487, "learning_rate": 1.602658680965152e-05, "loss": 0.0612, "step": 7510 }, { "grad_norm": 0.329708993434906, "learning_rate": 1.5905456904148686e-05, "loss": 0.0656, "step": 7520 }, { "grad_norm": 0.25004416704177856, "learning_rate": 1.57846998511067e-05, "loss": 0.0711, "step": 7530 }, { "grad_norm": 0.3635454475879669, "learning_rate": 1.566431697110538e-05, "loss": 0.0588, "step": 7540 }, { "grad_norm": 0.32505130767822266, "learning_rate": 1.554430958063259e-05, "loss": 0.0748, "step": 7550 }, { "grad_norm": 0.291025847196579, "learning_rate": 1.5424678992069912e-05, "loss": 0.0746, "step": 7560 }, { "grad_norm": 2.5982906818389893, "learning_rate": 1.5305426513678362e-05, "loss": 0.0683, "step": 7570 }, { "grad_norm": 0.36373472213745117, "learning_rate": 1.518655344958388e-05, "loss": 0.0777, "step": 7580 }, { "grad_norm": 0.3067559599876404, "learning_rate": 1.5068061099763275e-05, "loss": 0.0734, "step": 7590 }, { "grad_norm": 0.402044415473938, "learning_rate": 1.494995076002988e-05, "loss": 0.0758, "step": 7600 }, { "grad_norm": 0.49911919236183167, "learning_rate": 1.4832223722019456e-05, "loss": 0.0723, "step": 7610 }, { "grad_norm": 1.145955204963684, "learning_rate": 1.4714881273176035e-05, "loss": 0.0755, "step": 7620 }, { "grad_norm": 0.3083437979221344, "learning_rate": 1.4597924696737835e-05, "loss": 0.0759, "step": 7630 }, { "grad_norm": 0.4502699673175812, "learning_rate": 1.4481355271723252e-05, "loss": 0.0758, "step": 7640 }, { "grad_norm": 0.41489067673683167, "learning_rate": 1.4365174272916809e-05, "loss": 0.0645, "step": 7650 }, { "grad_norm": 0.4562056064605713, "learning_rate": 1.4249382970855319e-05, "loss": 0.068, "step": 7660 }, { "grad_norm": 0.490709513425827, "learning_rate": 1.4133982631813903e-05, "loss": 0.0783, "step": 7670 }, { "grad_norm": 0.2677809000015259, "learning_rate": 1.4018974517792194e-05, "loss": 0.0665, "step": 7680 }, { "grad_norm": 0.4502719044685364, "learning_rate": 1.390435988650048e-05, "loss": 0.0693, "step": 7690 }, { "grad_norm": 0.4056742489337921, "learning_rate": 1.3790139991346006e-05, "loss": 0.0613, "step": 7700 }, { "grad_norm": 0.36056429147720337, "learning_rate": 1.367631608141926e-05, "loss": 0.0772, "step": 7710 }, { "grad_norm": 0.3969501852989197, "learning_rate": 1.3562889401480278e-05, "loss": 0.0651, "step": 7720 }, { "grad_norm": 0.3061199188232422, "learning_rate": 1.3449861191945074e-05, "loss": 0.0645, "step": 7730 }, { "grad_norm": 0.3171863555908203, "learning_rate": 1.3337232688872009e-05, "loss": 0.0669, "step": 7740 }, { "grad_norm": 0.2687610685825348, "learning_rate": 1.3225005123948364e-05, "loss": 0.0581, "step": 7750 }, { "grad_norm": 0.3050493896007538, "learning_rate": 1.311317972447681e-05, "loss": 0.0631, "step": 7760 }, { "grad_norm": 0.3260588049888611, "learning_rate": 1.3001757713361996e-05, "loss": 0.0638, "step": 7770 }, { "grad_norm": 0.41647979617118835, "learning_rate": 1.2890740309097204e-05, "loss": 0.0691, "step": 7780 }, { "grad_norm": 0.3611864447593689, "learning_rate": 1.2780128725750944e-05, "loss": 0.0639, "step": 7790 }, { "grad_norm": 0.23494739830493927, "learning_rate": 1.266992417295379e-05, "loss": 0.0688, "step": 7800 }, { "grad_norm": 0.3532504439353943, "learning_rate": 1.2560127855885073e-05, "loss": 0.0584, "step": 7810 }, { "grad_norm": 0.3809933662414551, "learning_rate": 1.2450740975259745e-05, "loss": 0.053, "step": 7820 }, { "grad_norm": 0.3667725920677185, "learning_rate": 1.234176472731517e-05, "loss": 0.06, "step": 7830 }, { "grad_norm": 0.3875812590122223, "learning_rate": 1.2233200303798158e-05, "loss": 0.061, "step": 7840 }, { "grad_norm": 0.2897678315639496, "learning_rate": 1.2125048891951846e-05, "loss": 0.0625, "step": 7850 }, { "grad_norm": 0.3473932445049286, "learning_rate": 1.2017311674502745e-05, "loss": 0.0692, "step": 7860 }, { "grad_norm": 0.2771158814430237, "learning_rate": 1.1909989829647822e-05, "loss": 0.0837, "step": 7870 }, { "grad_norm": 0.42535659670829773, "learning_rate": 1.1803084531041553e-05, "loss": 0.0639, "step": 7880 }, { "grad_norm": 0.2160670906305313, "learning_rate": 1.1696596947783162e-05, "loss": 0.0684, "step": 7890 }, { "grad_norm": 0.40612348914146423, "learning_rate": 1.1590528244403803e-05, "loss": 0.0655, "step": 7900 }, { "grad_norm": 0.40680959820747375, "learning_rate": 1.148487958085382e-05, "loss": 0.0712, "step": 7910 }, { "grad_norm": 0.31693732738494873, "learning_rate": 1.1379652112490086e-05, "loss": 0.0869, "step": 7920 }, { "grad_norm": 0.6027331352233887, "learning_rate": 1.1274846990063315e-05, "loss": 0.0782, "step": 7930 }, { "grad_norm": 0.37115952372550964, "learning_rate": 1.117046535970554e-05, "loss": 0.0719, "step": 7940 }, { "grad_norm": 0.39793339371681213, "learning_rate": 1.106650836291755e-05, "loss": 0.0651, "step": 7950 }, { "grad_norm": 0.3654366135597229, "learning_rate": 1.0962977136556418e-05, "loss": 0.0639, "step": 7960 }, { "grad_norm": 0.5946422815322876, "learning_rate": 1.0859872812823024e-05, "loss": 0.0631, "step": 7970 }, { "grad_norm": 0.31593021750450134, "learning_rate": 1.0757196519249747e-05, "loss": 0.0638, "step": 7980 }, { "grad_norm": 0.456306517124176, "learning_rate": 1.0654949378688077e-05, "loss": 0.0698, "step": 7990 }, { "grad_norm": 0.5309772491455078, "learning_rate": 1.0553132509296376e-05, "loss": 0.0683, "step": 8000 }, { "grad_norm": 0.3638041019439697, "learning_rate": 1.0451747024527613e-05, "loss": 0.0723, "step": 8010 }, { "grad_norm": 0.3474266529083252, "learning_rate": 1.0350794033117189e-05, "loss": 0.0712, "step": 8020 }, { "grad_norm": 0.3768191635608673, "learning_rate": 1.0250274639070856e-05, "loss": 0.0657, "step": 8030 }, { "grad_norm": 0.27658402919769287, "learning_rate": 1.0150189941652599e-05, "loss": 0.0591, "step": 8040 }, { "grad_norm": 0.31934595108032227, "learning_rate": 1.0050541035372635e-05, "loss": 0.0663, "step": 8050 }, { "grad_norm": 0.322737455368042, "learning_rate": 9.951329009975458e-06, "loss": 0.0565, "step": 8060 }, { "grad_norm": 0.476762592792511, "learning_rate": 9.852554950427845e-06, "loss": 0.059, "step": 8070 }, { "grad_norm": 0.6329071521759033, "learning_rate": 9.754219936907105e-06, "loss": 0.0854, "step": 8080 }, { "grad_norm": 0.29337766766548157, "learning_rate": 9.656325044789194e-06, "loss": 0.0587, "step": 8090 }, { "grad_norm": 0.2130904346704483, "learning_rate": 9.55887134463697e-06, "loss": 0.071, "step": 8100 }, { "grad_norm": 0.2876536548137665, "learning_rate": 9.461859902188475e-06, "loss": 0.0755, "step": 8110 }, { "grad_norm": 0.2752240002155304, "learning_rate": 9.365291778345303e-06, "loss": 0.0658, "step": 8120 }, { "grad_norm": 0.44440704584121704, "learning_rate": 9.269168029160991e-06, "loss": 0.077, "step": 8130 }, { "grad_norm": 0.28074294328689575, "learning_rate": 9.173489705829447e-06, "loss": 0.0733, "step": 8140 }, { "grad_norm": 0.33253300189971924, "learning_rate": 9.078257854673516e-06, "loss": 0.071, "step": 8150 }, { "grad_norm": 0.5165941715240479, "learning_rate": 8.983473517133429e-06, "loss": 0.0712, "step": 8160 }, { "grad_norm": 0.4846840798854828, "learning_rate": 8.889137729755537e-06, "loss": 0.0628, "step": 8170 }, { "grad_norm": 0.3533477485179901, "learning_rate": 8.79525152418087e-06, "loss": 0.0649, "step": 8180 }, { "grad_norm": 0.32317987084388733, "learning_rate": 8.701815927133961e-06, "loss": 0.0692, "step": 8190 }, { "grad_norm": 0.32873451709747314, "learning_rate": 8.608831960411534e-06, "loss": 0.0586, "step": 8200 }, { "grad_norm": 0.3284474313259125, "learning_rate": 8.516300640871321e-06, "loss": 0.0534, "step": 8210 }, { "grad_norm": 0.4055301249027252, "learning_rate": 8.424222980421038e-06, "loss": 0.065, "step": 8220 }, { "grad_norm": 0.4032268524169922, "learning_rate": 8.332599986007184e-06, "loss": 0.0553, "step": 8230 }, { "grad_norm": 0.3618662357330322, "learning_rate": 8.241432659604203e-06, "loss": 0.0656, "step": 8240 }, { "grad_norm": 0.2021513432264328, "learning_rate": 8.150721998203331e-06, "loss": 0.0544, "step": 8250 }, { "grad_norm": 0.5143668055534363, "learning_rate": 8.06046899380184e-06, "loss": 0.064, "step": 8260 }, { "grad_norm": 0.39325425028800964, "learning_rate": 7.970674633392133e-06, "loss": 0.0617, "step": 8270 }, { "grad_norm": 0.3878862261772156, "learning_rate": 7.881339898950924e-06, "loss": 0.0674, "step": 8280 }, { "grad_norm": 0.6087312698364258, "learning_rate": 7.792465767428597e-06, "loss": 0.0706, "step": 8290 }, { "grad_norm": 0.3374445140361786, "learning_rate": 7.704053210738376e-06, "loss": 0.0713, "step": 8300 }, { "grad_norm": 0.3850906789302826, "learning_rate": 7.6161031957458494e-06, "loss": 0.0532, "step": 8310 }, { "grad_norm": 0.31918320059776306, "learning_rate": 7.5286166842582605e-06, "loss": 0.0531, "step": 8320 }, { "grad_norm": 0.6599282026290894, "learning_rate": 7.4415946330140814e-06, "loss": 0.0631, "step": 8330 }, { "grad_norm": 0.2786236107349396, "learning_rate": 7.3550379936725644e-06, "loss": 0.0568, "step": 8340 }, { "grad_norm": 0.34030529856681824, "learning_rate": 7.2689477128032035e-06, "loss": 0.0851, "step": 8350 }, { "grad_norm": 0.4037209749221802, "learning_rate": 7.183324731875551e-06, "loss": 0.061, "step": 8360 }, { "grad_norm": 0.3887912929058075, "learning_rate": 7.098169987248782e-06, "loss": 0.064, "step": 8370 }, { "grad_norm": 0.6399078965187073, "learning_rate": 7.013484410161553e-06, "loss": 0.0651, "step": 8380 }, { "grad_norm": 0.431021511554718, "learning_rate": 6.92926892672176e-06, "loss": 0.0514, "step": 8390 }, { "grad_norm": 0.3372241258621216, "learning_rate": 6.845524457896446e-06, "loss": 0.0556, "step": 8400 }, { "grad_norm": 0.5195845365524292, "learning_rate": 6.7622519195017165e-06, "loss": 0.0647, "step": 8410 }, { "grad_norm": 0.2356649935245514, "learning_rate": 6.679452222192684e-06, "loss": 0.06, "step": 8420 }, { "grad_norm": 0.3620017170906067, "learning_rate": 6.597126271453579e-06, "loss": 0.0536, "step": 8430 }, { "grad_norm": 0.24648191034793854, "learning_rate": 6.51527496758782e-06, "loss": 0.0575, "step": 8440 }, { "grad_norm": 0.39949268102645874, "learning_rate": 6.433899205708155e-06, "loss": 0.0575, "step": 8450 }, { "grad_norm": 0.4155108332633972, "learning_rate": 6.352999875726856e-06, "loss": 0.0569, "step": 8460 }, { "grad_norm": 0.42015165090560913, "learning_rate": 6.272577862346052e-06, "loss": 0.0488, "step": 8470 }, { "grad_norm": 0.24066294729709625, "learning_rate": 6.192634045047996e-06, "loss": 0.0545, "step": 8480 }, { "grad_norm": 0.40263667702674866, "learning_rate": 6.113169298085458e-06, "loss": 0.0644, "step": 8490 }, { "grad_norm": 0.5174115300178528, "learning_rate": 6.034184490472195e-06, "loss": 0.0654, "step": 8500 }, { "grad_norm": 0.27164226770401, "learning_rate": 5.955680485973386e-06, "loss": 0.0583, "step": 8510 }, { "grad_norm": 0.4941122829914093, "learning_rate": 5.877658143096265e-06, "loss": 0.042, "step": 8520 }, { "grad_norm": 0.2729173004627228, "learning_rate": 5.800118315080661e-06, "loss": 0.0637, "step": 8530 }, { "grad_norm": 0.29822373390197754, "learning_rate": 5.723061849889716e-06, "loss": 0.0545, "step": 8540 }, { "grad_norm": 0.24252235889434814, "learning_rate": 5.646489590200604e-06, "loss": 0.0578, "step": 8550 }, { "grad_norm": 0.253322958946228, "learning_rate": 5.570402373395256e-06, "loss": 0.0588, "step": 8560 }, { "grad_norm": 0.37817808985710144, "learning_rate": 5.494801031551305e-06, "loss": 0.0855, "step": 8570 }, { "grad_norm": 0.3456450402736664, "learning_rate": 5.41968639143291e-06, "loss": 0.0615, "step": 8580 }, { "grad_norm": 0.2630425989627838, "learning_rate": 5.345059274481751e-06, "loss": 0.0659, "step": 8590 }, { "grad_norm": 0.27888405323028564, "learning_rate": 5.270920496808002e-06, "loss": 0.058, "step": 8600 }, { "grad_norm": 0.4681285321712494, "learning_rate": 5.1972708691814695e-06, "loss": 0.0611, "step": 8610 }, { "grad_norm": 0.3960498869419098, "learning_rate": 5.124111197022674e-06, "loss": 0.0578, "step": 8620 }, { "grad_norm": 0.7145316004753113, "learning_rate": 5.051442280394081e-06, "loss": 0.0598, "step": 8630 }, { "grad_norm": 0.31400689482688904, "learning_rate": 4.979264913991322e-06, "loss": 0.0653, "step": 8640 }, { "grad_norm": 0.2672719657421112, "learning_rate": 4.907579887134489e-06, "loss": 0.0803, "step": 8650 }, { "grad_norm": 0.3060336112976074, "learning_rate": 4.836387983759572e-06, "loss": 0.0517, "step": 8660 }, { "grad_norm": 0.3286443054676056, "learning_rate": 4.765689982409816e-06, "loss": 0.0539, "step": 8670 }, { "grad_norm": 0.33614036440849304, "learning_rate": 4.695486656227233e-06, "loss": 0.0565, "step": 8680 }, { "grad_norm": 0.21835413575172424, "learning_rate": 4.625778772944156e-06, "loss": 0.0566, "step": 8690 }, { "grad_norm": 0.6201481223106384, "learning_rate": 4.556567094874825e-06, "loss": 0.0614, "step": 8700 }, { "grad_norm": 0.38892146944999695, "learning_rate": 4.487852378907059e-06, "loss": 0.0826, "step": 8710 }, { "grad_norm": 0.6803016662597656, "learning_rate": 4.419635376493986e-06, "loss": 0.0809, "step": 8720 }, { "grad_norm": 0.4124367833137512, "learning_rate": 4.351916833645825e-06, "loss": 0.0534, "step": 8730 }, { "grad_norm": 0.3587772846221924, "learning_rate": 4.284697490921691e-06, "loss": 0.0543, "step": 8740 }, { "grad_norm": 0.276333749294281, "learning_rate": 4.2179780834215585e-06, "loss": 0.0644, "step": 8750 }, { "grad_norm": 0.3985384702682495, "learning_rate": 4.151759340778178e-06, "loss": 0.065, "step": 8760 }, { "grad_norm": 0.42064395546913147, "learning_rate": 4.086041987149109e-06, "loss": 0.0759, "step": 8770 }, { "grad_norm": 0.336568146944046, "learning_rate": 4.020826741208811e-06, "loss": 0.0701, "step": 8780 }, { "grad_norm": 0.34179866313934326, "learning_rate": 3.956114316140746e-06, "loss": 0.067, "step": 8790 }, { "grad_norm": 0.28545862436294556, "learning_rate": 3.891905419629643e-06, "loss": 0.0533, "step": 8800 }, { "grad_norm": 0.454970121383667, "learning_rate": 3.8282007538536946e-06, "loss": 0.0557, "step": 8810 }, { "grad_norm": 0.3128204345703125, "learning_rate": 3.7650010154769265e-06, "loss": 0.057, "step": 8820 }, { "grad_norm": 0.728003978729248, "learning_rate": 3.7023068956415608e-06, "loss": 0.0535, "step": 8830 }, { "grad_norm": 0.3874084949493408, "learning_rate": 3.6401190799604303e-06, "loss": 0.0588, "step": 8840 }, { "grad_norm": 0.29409363865852356, "learning_rate": 3.578438248509536e-06, "loss": 0.053, "step": 8850 }, { "grad_norm": 0.5673768520355225, "learning_rate": 3.5172650758205583e-06, "loss": 0.0677, "step": 8860 }, { "grad_norm": 0.5056778788566589, "learning_rate": 3.45660023087353e-06, "loss": 0.0666, "step": 8870 }, { "grad_norm": 0.35132861137390137, "learning_rate": 3.3964443770894528e-06, "loss": 0.0745, "step": 8880 }, { "grad_norm": 0.35812243819236755, "learning_rate": 3.3367981723231245e-06, "loss": 0.0724, "step": 8890 }, { "grad_norm": 0.25863274931907654, "learning_rate": 3.2776622688558746e-06, "loss": 0.0654, "step": 8900 }, { "grad_norm": 0.2230181097984314, "learning_rate": 3.2190373133884677e-06, "loss": 0.0518, "step": 8910 }, { "grad_norm": 0.5070387721061707, "learning_rate": 3.1609239470340446e-06, "loss": 0.0684, "step": 8920 }, { "grad_norm": 0.38593173027038574, "learning_rate": 3.1033228053110373e-06, "loss": 0.0633, "step": 8930 }, { "grad_norm": 0.3043648302555084, "learning_rate": 3.0462345181363314e-06, "loss": 0.0635, "step": 8940 }, { "grad_norm": 0.31729623675346375, "learning_rate": 2.9896597098182654e-06, "loss": 0.0619, "step": 8950 }, { "grad_norm": 0.7212941646575928, "learning_rate": 2.933598999049891e-06, "loss": 0.0528, "step": 8960 }, { "grad_norm": 0.30422481894493103, "learning_rate": 2.8780529989021697e-06, "loss": 0.0546, "step": 8970 }, { "grad_norm": 0.37249085307121277, "learning_rate": 2.823022316817242e-06, "loss": 0.067, "step": 8980 }, { "grad_norm": 0.5135778784751892, "learning_rate": 2.7685075546018456e-06, "loss": 0.085, "step": 8990 }, { "grad_norm": 0.39063557982444763, "learning_rate": 2.7145093084206598e-06, "loss": 0.07, "step": 9000 }, { "grad_norm": 0.35842862725257874, "learning_rate": 2.661028168789892e-06, "loss": 0.0634, "step": 9010 }, { "grad_norm": 0.4435299336910248, "learning_rate": 2.6080647205706855e-06, "loss": 0.0652, "step": 9020 }, { "grad_norm": 0.46711552143096924, "learning_rate": 2.555619542962834e-06, "loss": 0.0641, "step": 9030 }, { "grad_norm": 0.3262532949447632, "learning_rate": 2.503693209498409e-06, "loss": 0.0711, "step": 9040 }, { "grad_norm": 0.29012560844421387, "learning_rate": 2.452286288035449e-06, "loss": 0.0577, "step": 9050 }, { "grad_norm": 0.3589058816432953, "learning_rate": 2.4013993407518363e-06, "loss": 0.0782, "step": 9060 }, { "grad_norm": 0.32655617594718933, "learning_rate": 2.351032924139063e-06, "loss": 0.0461, "step": 9070 }, { "grad_norm": 0.4130423367023468, "learning_rate": 2.30118758899619e-06, "loss": 0.0654, "step": 9080 }, { "grad_norm": 0.20899498462677002, "learning_rate": 2.2518638804238157e-06, "loss": 0.066, "step": 9090 }, { "grad_norm": 0.5182575583457947, "learning_rate": 2.203062337818118e-06, "loss": 0.0629, "step": 9100 }, { "grad_norm": 0.3034808337688446, "learning_rate": 2.1547834948649483e-06, "loss": 0.0602, "step": 9110 }, { "grad_norm": 0.37853437662124634, "learning_rate": 2.1070278795340017e-06, "loss": 0.062, "step": 9120 }, { "grad_norm": 0.3598875105381012, "learning_rate": 2.059796014073029e-06, "loss": 0.061, "step": 9130 }, { "grad_norm": 0.22988058626651764, "learning_rate": 2.01308841500214e-06, "loss": 0.0512, "step": 9140 }, { "grad_norm": 0.32057061791419983, "learning_rate": 1.9669055931081704e-06, "loss": 0.063, "step": 9150 }, { "grad_norm": 0.3072979152202606, "learning_rate": 1.9212480534390507e-06, "loss": 0.0535, "step": 9160 }, { "grad_norm": 0.3584454357624054, "learning_rate": 1.8761162952983246e-06, "loss": 0.0521, "step": 9170 }, { "grad_norm": 0.5642172694206238, "learning_rate": 1.8315108122396618e-06, "loss": 0.058, "step": 9180 }, { "grad_norm": 0.41956979036331177, "learning_rate": 1.787432092061475e-06, "loss": 0.0679, "step": 9190 }, { "grad_norm": 0.3429662585258484, "learning_rate": 1.743880616801602e-06, "loss": 0.0715, "step": 9200 }, { "grad_norm": 0.34560486674308777, "learning_rate": 1.7008568627319865e-06, "loss": 0.0711, "step": 9210 }, { "grad_norm": 0.32336634397506714, "learning_rate": 1.6583613003535226e-06, "loss": 0.0708, "step": 9220 }, { "grad_norm": 0.3006320297718048, "learning_rate": 1.6163943943908522e-06, "loss": 0.0654, "step": 9230 }, { "grad_norm": 1.8685743808746338, "learning_rate": 1.5749566037873476e-06, "loss": 0.0691, "step": 9240 }, { "grad_norm": 0.49167078733444214, "learning_rate": 1.5340483817000428e-06, "loss": 0.0599, "step": 9250 }, { "grad_norm": 0.31972095370292664, "learning_rate": 1.4936701754947101e-06, "loss": 0.0528, "step": 9260 }, { "grad_norm": 0.4235546290874481, "learning_rate": 1.4538224267409361e-06, "loss": 0.0597, "step": 9270 }, { "grad_norm": 0.4883168339729309, "learning_rate": 1.414505571207314e-06, "loss": 0.0555, "step": 9280 }, { "grad_norm": 0.47365331649780273, "learning_rate": 1.3757200388566816e-06, "loss": 0.0716, "step": 9290 }, { "grad_norm": 0.4217505156993866, "learning_rate": 1.3374662538414074e-06, "loss": 0.0725, "step": 9300 }, { "grad_norm": 0.34823206067085266, "learning_rate": 1.2997446344987617e-06, "loss": 0.0711, "step": 9310 }, { "grad_norm": 0.3186560273170471, "learning_rate": 1.262555593346315e-06, "loss": 0.0757, "step": 9320 }, { "grad_norm": 0.353429913520813, "learning_rate": 1.2258995370774685e-06, "loss": 0.0619, "step": 9330 }, { "grad_norm": 0.2643960416316986, "learning_rate": 1.1897768665569798e-06, "loss": 0.0665, "step": 9340 }, { "grad_norm": 0.48919573426246643, "learning_rate": 1.1541879768165954e-06, "loss": 0.0617, "step": 9350 }, { "grad_norm": 0.31620702147483826, "learning_rate": 1.1191332570507085e-06, "loss": 0.0564, "step": 9360 }, { "grad_norm": 0.4109647274017334, "learning_rate": 1.0846130906121132e-06, "loss": 0.0695, "step": 9370 }, { "grad_norm": 0.4819319546222687, "learning_rate": 1.0506278550078131e-06, "loss": 0.0681, "step": 9380 }, { "grad_norm": 0.4952082931995392, "learning_rate": 1.0171779218949185e-06, "loss": 0.0705, "step": 9390 }, { "grad_norm": 0.323355108499527, "learning_rate": 9.842636570765174e-07, "loss": 0.0781, "step": 9400 }, { "grad_norm": 0.3570466637611389, "learning_rate": 9.518854204977612e-07, "loss": 0.0718, "step": 9410 }, { "grad_norm": 0.6221121549606323, "learning_rate": 9.200435662418349e-07, "loss": 0.0661, "step": 9420 }, { "grad_norm": 0.2938351333141327, "learning_rate": 8.887384425261658e-07, "loss": 0.0554, "step": 9430 }, { "grad_norm": 0.4560299515724182, "learning_rate": 8.579703916985648e-07, "loss": 0.061, "step": 9440 }, { "grad_norm": 0.24459919333457947, "learning_rate": 8.277397502335194e-07, "loss": 0.0455, "step": 9450 }, { "grad_norm": 0.35994449257850647, "learning_rate": 7.980468487284675e-07, "loss": 0.053, "step": 9460 }, { "grad_norm": 0.5594532489776611, "learning_rate": 7.688920119002297e-07, "loss": 0.0627, "step": 9470 }, { "grad_norm": 0.4399980306625366, "learning_rate": 7.402755585814269e-07, "loss": 0.0581, "step": 9480 }, { "grad_norm": 0.5443531274795532, "learning_rate": 7.121978017170073e-07, "loss": 0.0539, "step": 9490 }, { "grad_norm": 0.28078052401542664, "learning_rate": 6.846590483608306e-07, "loss": 0.0623, "step": 9500 }, { "grad_norm": 0.3159867525100708, "learning_rate": 6.576595996722834e-07, "loss": 0.0603, "step": 9510 }, { "grad_norm": 0.2947486340999603, "learning_rate": 6.311997509130141e-07, "loss": 0.0582, "step": 9520 }, { "grad_norm": 0.6113506555557251, "learning_rate": 6.052797914436803e-07, "loss": 0.0662, "step": 9530 }, { "grad_norm": 0.5673233866691589, "learning_rate": 5.799000047208181e-07, "loss": 0.0729, "step": 9540 }, { "grad_norm": 0.3114430606365204, "learning_rate": 5.550606682937054e-07, "loss": 0.0588, "step": 9550 }, { "grad_norm": 0.3426528573036194, "learning_rate": 5.307620538013481e-07, "loss": 0.0573, "step": 9560 }, { "grad_norm": 0.2445455640554428, "learning_rate": 5.070044269694874e-07, "loss": 0.0634, "step": 9570 }, { "grad_norm": 0.3405572772026062, "learning_rate": 4.837880476077417e-07, "loss": 0.062, "step": 9580 }, { "grad_norm": 0.2529284358024597, "learning_rate": 4.6111316960670835e-07, "loss": 0.0587, "step": 9590 }, { "grad_norm": 0.2960663139820099, "learning_rate": 4.389800409352218e-07, "loss": 0.0584, "step": 9600 }, { "grad_norm": 0.5044896006584167, "learning_rate": 4.173889036376277e-07, "loss": 0.0632, "step": 9610 }, { "grad_norm": 0.3009117841720581, "learning_rate": 3.963399938311463e-07, "loss": 0.0549, "step": 9620 }, { "grad_norm": 0.6443825960159302, "learning_rate": 3.7583354170328545e-07, "loss": 0.0572, "step": 9630 }, { "grad_norm": 0.3018619418144226, "learning_rate": 3.558697715093207e-07, "loss": 0.0661, "step": 9640 }, { "grad_norm": 0.3821899890899658, "learning_rate": 3.3644890156983576e-07, "loss": 0.0666, "step": 9650 }, { "grad_norm": 0.2947966158390045, "learning_rate": 3.175711442683638e-07, "loss": 0.0633, "step": 9660 }, { "grad_norm": 0.23648110032081604, "learning_rate": 2.9923670604902197e-07, "loss": 0.0634, "step": 9670 }, { "grad_norm": 0.38341328501701355, "learning_rate": 2.814457874143028e-07, "loss": 0.0535, "step": 9680 }, { "grad_norm": 0.4490429759025574, "learning_rate": 2.641985829228366e-07, "loss": 0.0548, "step": 9690 }, { "grad_norm": 0.34937039017677307, "learning_rate": 2.474952811872877e-07, "loss": 0.051, "step": 9700 }, { "grad_norm": 0.3383079171180725, "learning_rate": 2.3133606487228397e-07, "loss": 0.0532, "step": 9710 }, { "grad_norm": 0.3796030580997467, "learning_rate": 2.157211106924295e-07, "loss": 0.0582, "step": 9720 }, { "grad_norm": 0.3596123158931732, "learning_rate": 2.006505894103672e-07, "loss": 0.0698, "step": 9730 }, { "grad_norm": 0.4103030264377594, "learning_rate": 1.8612466583489696e-07, "loss": 0.0505, "step": 9740 }, { "grad_norm": 0.4245615303516388, "learning_rate": 1.7214349881918834e-07, "loss": 0.0569, "step": 9750 }, { "grad_norm": 0.35942643880844116, "learning_rate": 1.5870724125904845e-07, "loss": 0.0678, "step": 9760 }, { "grad_norm": 0.35717880725860596, "learning_rate": 1.4581604009124006e-07, "loss": 0.0655, "step": 9770 }, { "grad_norm": 0.3359903395175934, "learning_rate": 1.334700362918717e-07, "loss": 0.0588, "step": 9780 }, { "grad_norm": 0.32164308428764343, "learning_rate": 1.2166936487486015e-07, "loss": 0.0568, "step": 9790 }, { "grad_norm": 0.26678669452667236, "learning_rate": 1.1041415489045914e-07, "loss": 0.0547, "step": 9800 }, { "grad_norm": 0.31802651286125183, "learning_rate": 9.970452942384412e-08, "loss": 0.0615, "step": 9810 }, { "grad_norm": 0.3192044496536255, "learning_rate": 8.954060559375754e-08, "loss": 0.0635, "step": 9820 }, { "grad_norm": 0.32221218943595886, "learning_rate": 7.99224945512489e-08, "loss": 0.0634, "step": 9830 }, { "grad_norm": 0.26733824610710144, "learning_rate": 7.085030147843675e-08, "loss": 0.067, "step": 9840 }, { "grad_norm": 0.20583797991275787, "learning_rate": 6.232412558736523e-08, "loss": 0.0703, "step": 9850 }, { "grad_norm": 0.4932723343372345, "learning_rate": 5.434406011893822e-08, "loss": 0.0509, "step": 9860 }, { "grad_norm": 0.5286715030670166, "learning_rate": 4.6910192341864664e-08, "loss": 0.0647, "step": 9870 }, { "grad_norm": 0.3384794294834137, "learning_rate": 4.0022603551737035e-08, "loss": 0.0699, "step": 9880 }, { "grad_norm": 0.6526644825935364, "learning_rate": 3.3681369070120985e-08, "loss": 0.062, "step": 9890 }, { "grad_norm": 0.29839667677879333, "learning_rate": 2.7886558243744866e-08, "loss": 0.0582, "step": 9900 }, { "grad_norm": 0.39546728134155273, "learning_rate": 2.2638234443722596e-08, "loss": 0.0566, "step": 9910 }, { "grad_norm": 0.2450965791940689, "learning_rate": 1.7936455064887504e-08, "loss": 0.0652, "step": 9920 }, { "grad_norm": 0.3420693874359131, "learning_rate": 1.378127152514841e-08, "loss": 0.0536, "step": 9930 }, { "grad_norm": 0.44208481907844543, "learning_rate": 1.0172729264917857e-08, "loss": 0.062, "step": 9940 }, { "grad_norm": 0.388689786195755, "learning_rate": 7.1108677466458215e-09, "loss": 0.0586, "step": 9950 }, { "grad_norm": 0.31959807872772217, "learning_rate": 4.595720454353414e-09, "loss": 0.0503, "step": 9960 }, { "grad_norm": 0.2958472669124603, "learning_rate": 2.627314893294264e-09, "loss": 0.0725, "step": 9970 }, { "grad_norm": 0.9079338908195496, "learning_rate": 1.2056725896270048e-09, "loss": 0.0682, "step": 9980 }, { "grad_norm": 0.33804085850715637, "learning_rate": 3.308090902098826e-10, "loss": 0.0652, "step": 9990 }, { "grad_norm": 0.5289033651351929, "learning_rate": 2.7339624120159555e-12, "loss": 0.0694, "step": 10000 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 10000, "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": 8, "trial_name": null, "trial_params": null }