{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.6666666666666666, "eval_steps": 500, "global_step": 20000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.1528514623641968, "learning_rate": 6.000000000000001e-07, "loss": 1.1443, "step": 10 }, { "grad_norm": 1.385394811630249, "learning_rate": 1.2666666666666667e-06, "loss": 1.1511, "step": 20 }, { "grad_norm": 1.0653809309005737, "learning_rate": 1.9333333333333336e-06, "loss": 1.148, "step": 30 }, { "grad_norm": 0.6685559749603271, "learning_rate": 2.6e-06, "loss": 1.1349, "step": 40 }, { "grad_norm": 0.42683279514312744, "learning_rate": 3.2666666666666666e-06, "loss": 1.127, "step": 50 }, { "grad_norm": 0.4504030644893646, "learning_rate": 3.9333333333333335e-06, "loss": 1.1169, "step": 60 }, { "grad_norm": 0.4557018578052521, "learning_rate": 4.6e-06, "loss": 1.1179, "step": 70 }, { "grad_norm": 0.34378084540367126, "learning_rate": 5.266666666666667e-06, "loss": 1.1123, "step": 80 }, { "grad_norm": 0.3076566159725189, "learning_rate": 5.933333333333334e-06, "loss": 1.0955, "step": 90 }, { "grad_norm": 0.3332269489765167, "learning_rate": 6.6e-06, "loss": 1.1103, "step": 100 }, { "grad_norm": 0.602932333946228, "learning_rate": 7.266666666666668e-06, "loss": 1.1139, "step": 110 }, { "grad_norm": 0.3543250560760498, "learning_rate": 7.933333333333334e-06, "loss": 1.1131, "step": 120 }, { "grad_norm": 0.38544756174087524, "learning_rate": 8.599999999999999e-06, "loss": 1.1097, "step": 130 }, { "grad_norm": 0.4270870089530945, "learning_rate": 9.266666666666667e-06, "loss": 1.0912, "step": 140 }, { "grad_norm": 0.556666374206543, "learning_rate": 9.933333333333334e-06, "loss": 1.0899, "step": 150 }, { "grad_norm": 0.6506553292274475, "learning_rate": 1.06e-05, "loss": 1.0743, "step": 160 }, { "grad_norm": 0.5043882727622986, "learning_rate": 1.1266666666666667e-05, "loss": 1.0617, "step": 170 }, { "grad_norm": 0.5745680928230286, "learning_rate": 1.1933333333333333e-05, "loss": 1.057, "step": 180 }, { "grad_norm": 0.49374717473983765, "learning_rate": 1.2600000000000001e-05, "loss": 1.0634, "step": 190 }, { "grad_norm": 0.5588945746421814, "learning_rate": 1.3266666666666666e-05, "loss": 1.0467, "step": 200 }, { "grad_norm": 0.44214844703674316, "learning_rate": 1.3933333333333334e-05, "loss": 1.0685, "step": 210 }, { "grad_norm": 0.4926779568195343, "learning_rate": 1.4599999999999999e-05, "loss": 1.0597, "step": 220 }, { "grad_norm": 0.5118926167488098, "learning_rate": 1.5266666666666667e-05, "loss": 1.066, "step": 230 }, { "grad_norm": 0.5368437767028809, "learning_rate": 1.5933333333333332e-05, "loss": 1.0524, "step": 240 }, { "grad_norm": 0.5136311650276184, "learning_rate": 1.66e-05, "loss": 1.0541, "step": 250 }, { "grad_norm": 0.6221996545791626, "learning_rate": 1.726666666666667e-05, "loss": 1.0549, "step": 260 }, { "grad_norm": 0.46564388275146484, "learning_rate": 1.7933333333333337e-05, "loss": 1.0434, "step": 270 }, { "grad_norm": 0.5669482946395874, "learning_rate": 1.86e-05, "loss": 1.0439, "step": 280 }, { "grad_norm": 0.46580296754837036, "learning_rate": 1.926666666666667e-05, "loss": 1.0518, "step": 290 }, { "grad_norm": 0.5167627930641174, "learning_rate": 1.9933333333333334e-05, "loss": 1.0378, "step": 300 }, { "grad_norm": 0.579369306564331, "learning_rate": 2.06e-05, "loss": 1.0348, "step": 310 }, { "grad_norm": 0.7498450875282288, "learning_rate": 2.1266666666666667e-05, "loss": 1.0297, "step": 320 }, { "grad_norm": 0.5921820998191833, "learning_rate": 2.1933333333333332e-05, "loss": 1.0286, "step": 330 }, { "grad_norm": 0.6954401135444641, "learning_rate": 2.26e-05, "loss": 1.0179, "step": 340 }, { "grad_norm": 0.6771758794784546, "learning_rate": 2.326666666666667e-05, "loss": 1.0232, "step": 350 }, { "grad_norm": 0.6090942621231079, "learning_rate": 2.3933333333333337e-05, "loss": 1.0254, "step": 360 }, { "grad_norm": 0.490033894777298, "learning_rate": 2.46e-05, "loss": 1.0149, "step": 370 }, { "grad_norm": 0.47561225295066833, "learning_rate": 2.5266666666666666e-05, "loss": 0.9988, "step": 380 }, { "grad_norm": 0.7549184560775757, "learning_rate": 2.5933333333333338e-05, "loss": 0.9752, "step": 390 }, { "grad_norm": 0.739684522151947, "learning_rate": 2.6600000000000003e-05, "loss": 0.9683, "step": 400 }, { "grad_norm": 0.835106611251831, "learning_rate": 2.7266666666666668e-05, "loss": 0.9426, "step": 410 }, { "grad_norm": 1.1115773916244507, "learning_rate": 2.7933333333333332e-05, "loss": 0.9312, "step": 420 }, { "grad_norm": 1.1162455081939697, "learning_rate": 2.86e-05, "loss": 0.9072, "step": 430 }, { "grad_norm": 0.8143486976623535, "learning_rate": 2.926666666666667e-05, "loss": 0.8631, "step": 440 }, { "grad_norm": 0.9558935761451721, "learning_rate": 2.9933333333333337e-05, "loss": 0.817, "step": 450 }, { "grad_norm": 1.1171176433563232, "learning_rate": 3.06e-05, "loss": 0.788, "step": 460 }, { "grad_norm": 1.0326383113861084, "learning_rate": 3.126666666666666e-05, "loss": 0.7486, "step": 470 }, { "grad_norm": 1.1813828945159912, "learning_rate": 3.1933333333333335e-05, "loss": 0.711, "step": 480 }, { "grad_norm": 1.3364958763122559, "learning_rate": 3.26e-05, "loss": 0.6752, "step": 490 }, { "grad_norm": 1.3715628385543823, "learning_rate": 3.326666666666667e-05, "loss": 0.6272, "step": 500 }, { "grad_norm": 1.099719762802124, "learning_rate": 3.3933333333333336e-05, "loss": 0.6105, "step": 510 }, { "grad_norm": 1.1769758462905884, "learning_rate": 3.46e-05, "loss": 0.5706, "step": 520 }, { "grad_norm": 1.3166848421096802, "learning_rate": 3.526666666666667e-05, "loss": 0.5501, "step": 530 }, { "grad_norm": 1.022141695022583, "learning_rate": 3.593333333333334e-05, "loss": 0.4726, "step": 540 }, { "grad_norm": 1.4008371829986572, "learning_rate": 3.66e-05, "loss": 0.4758, "step": 550 }, { "grad_norm": 1.2278764247894287, "learning_rate": 3.726666666666667e-05, "loss": 0.4412, "step": 560 }, { "grad_norm": 1.1129118204116821, "learning_rate": 3.793333333333334e-05, "loss": 0.3892, "step": 570 }, { "grad_norm": 1.149166226387024, "learning_rate": 3.86e-05, "loss": 0.3616, "step": 580 }, { "grad_norm": 1.2555584907531738, "learning_rate": 3.926666666666667e-05, "loss": 0.3479, "step": 590 }, { "grad_norm": 1.2826015949249268, "learning_rate": 3.993333333333333e-05, "loss": 0.3312, "step": 600 }, { "grad_norm": 1.2750180959701538, "learning_rate": 4.0600000000000004e-05, "loss": 0.3153, "step": 610 }, { "grad_norm": 1.2580409049987793, "learning_rate": 4.126666666666667e-05, "loss": 0.292, "step": 620 }, { "grad_norm": 1.2718510627746582, "learning_rate": 4.1933333333333334e-05, "loss": 0.289, "step": 630 }, { "grad_norm": 1.4107351303100586, "learning_rate": 4.26e-05, "loss": 0.2713, "step": 640 }, { "grad_norm": 1.2467225790023804, "learning_rate": 4.3266666666666664e-05, "loss": 0.2561, "step": 650 }, { "grad_norm": 1.1022698879241943, "learning_rate": 4.3933333333333335e-05, "loss": 0.2627, "step": 660 }, { "grad_norm": 0.9434885382652283, "learning_rate": 4.46e-05, "loss": 0.2112, "step": 670 }, { "grad_norm": 1.442144751548767, "learning_rate": 4.526666666666667e-05, "loss": 0.2357, "step": 680 }, { "grad_norm": 1.2791364192962646, "learning_rate": 4.5933333333333336e-05, "loss": 0.1974, "step": 690 }, { "grad_norm": 1.2810497283935547, "learning_rate": 4.660000000000001e-05, "loss": 0.203, "step": 700 }, { "grad_norm": 1.2554171085357666, "learning_rate": 4.726666666666667e-05, "loss": 0.1764, "step": 710 }, { "grad_norm": 1.2241530418395996, "learning_rate": 4.793333333333334e-05, "loss": 0.1745, "step": 720 }, { "grad_norm": 1.2599396705627441, "learning_rate": 4.86e-05, "loss": 0.1712, "step": 730 }, { "grad_norm": 1.3620973825454712, "learning_rate": 4.926666666666667e-05, "loss": 0.1396, "step": 740 }, { "grad_norm": 1.0858829021453857, "learning_rate": 4.993333333333334e-05, "loss": 0.1618, "step": 750 }, { "grad_norm": 0.8151256442070007, "learning_rate": 5.0600000000000003e-05, "loss": 0.1556, "step": 760 }, { "grad_norm": 0.9104342460632324, "learning_rate": 5.1266666666666675e-05, "loss": 0.17, "step": 770 }, { "grad_norm": 1.4416277408599854, "learning_rate": 5.193333333333333e-05, "loss": 0.1979, "step": 780 }, { "grad_norm": 0.9919344782829285, "learning_rate": 5.2600000000000005e-05, "loss": 0.1475, "step": 790 }, { "grad_norm": 0.888515055179596, "learning_rate": 5.326666666666666e-05, "loss": 0.1463, "step": 800 }, { "grad_norm": 1.0972806215286255, "learning_rate": 5.3933333333333334e-05, "loss": 0.1406, "step": 810 }, { "grad_norm": 1.0611846446990967, "learning_rate": 5.4600000000000006e-05, "loss": 0.1433, "step": 820 }, { "grad_norm": 0.9166650772094727, "learning_rate": 5.5266666666666664e-05, "loss": 0.1657, "step": 830 }, { "grad_norm": 0.9015916585922241, "learning_rate": 5.5933333333333335e-05, "loss": 0.1553, "step": 840 }, { "grad_norm": 1.1537153720855713, "learning_rate": 5.66e-05, "loss": 0.1537, "step": 850 }, { "grad_norm": 0.997890293598175, "learning_rate": 5.726666666666667e-05, "loss": 0.1445, "step": 860 }, { "grad_norm": 1.060499906539917, "learning_rate": 5.7933333333333337e-05, "loss": 0.1633, "step": 870 }, { "grad_norm": 0.7399134635925293, "learning_rate": 5.86e-05, "loss": 0.1371, "step": 880 }, { "grad_norm": 0.735348105430603, "learning_rate": 5.926666666666667e-05, "loss": 0.1439, "step": 890 }, { "grad_norm": 0.7918251752853394, "learning_rate": 5.9933333333333345e-05, "loss": 0.1318, "step": 900 }, { "grad_norm": 0.8497225046157837, "learning_rate": 6.06e-05, "loss": 0.1386, "step": 910 }, { "grad_norm": 0.7786096334457397, "learning_rate": 6.126666666666667e-05, "loss": 0.1235, "step": 920 }, { "grad_norm": 1.0084772109985352, "learning_rate": 6.193333333333333e-05, "loss": 0.1377, "step": 930 }, { "grad_norm": 0.8829622268676758, "learning_rate": 6.26e-05, "loss": 0.1454, "step": 940 }, { "grad_norm": 0.6258594989776611, "learning_rate": 6.326666666666667e-05, "loss": 0.1219, "step": 950 }, { "grad_norm": 0.9196205139160156, "learning_rate": 6.393333333333333e-05, "loss": 0.1354, "step": 960 }, { "grad_norm": 0.9631812572479248, "learning_rate": 6.460000000000001e-05, "loss": 0.161, "step": 970 }, { "grad_norm": 0.7544456720352173, "learning_rate": 6.526666666666666e-05, "loss": 0.1209, "step": 980 }, { "grad_norm": 0.9052571654319763, "learning_rate": 6.593333333333334e-05, "loss": 0.1177, "step": 990 }, { "grad_norm": 0.7887353301048279, "learning_rate": 6.66e-05, "loss": 0.1339, "step": 1000 }, { "grad_norm": 0.7791980504989624, "learning_rate": 6.726666666666667e-05, "loss": 0.1312, "step": 1010 }, { "grad_norm": 0.7570158243179321, "learning_rate": 6.793333333333334e-05, "loss": 0.139, "step": 1020 }, { "grad_norm": 0.7318595051765442, "learning_rate": 6.860000000000001e-05, "loss": 0.1172, "step": 1030 }, { "grad_norm": 0.6829156279563904, "learning_rate": 6.926666666666667e-05, "loss": 0.1221, "step": 1040 }, { "grad_norm": 0.8809564709663391, "learning_rate": 6.993333333333334e-05, "loss": 0.1266, "step": 1050 }, { "grad_norm": 0.6686496734619141, "learning_rate": 7.06e-05, "loss": 0.128, "step": 1060 }, { "grad_norm": 0.8521612882614136, "learning_rate": 7.126666666666667e-05, "loss": 0.1436, "step": 1070 }, { "grad_norm": 0.8797381520271301, "learning_rate": 7.193333333333334e-05, "loss": 0.1435, "step": 1080 }, { "grad_norm": 0.7127763032913208, "learning_rate": 7.26e-05, "loss": 0.1246, "step": 1090 }, { "grad_norm": 1.1113970279693604, "learning_rate": 7.326666666666667e-05, "loss": 0.1128, "step": 1100 }, { "grad_norm": 0.8404185175895691, "learning_rate": 7.393333333333333e-05, "loss": 0.1636, "step": 1110 }, { "grad_norm": 0.6565252542495728, "learning_rate": 7.46e-05, "loss": 0.1115, "step": 1120 }, { "grad_norm": 0.6725383400917053, "learning_rate": 7.526666666666668e-05, "loss": 0.1289, "step": 1130 }, { "grad_norm": 0.8664162158966064, "learning_rate": 7.593333333333334e-05, "loss": 0.1629, "step": 1140 }, { "grad_norm": 0.6621503233909607, "learning_rate": 7.66e-05, "loss": 0.1528, "step": 1150 }, { "grad_norm": 0.7894150614738464, "learning_rate": 7.726666666666667e-05, "loss": 0.1307, "step": 1160 }, { "grad_norm": 0.5448333621025085, "learning_rate": 7.793333333333333e-05, "loss": 0.1285, "step": 1170 }, { "grad_norm": 0.6217865347862244, "learning_rate": 7.860000000000001e-05, "loss": 0.1039, "step": 1180 }, { "grad_norm": 0.6798178553581238, "learning_rate": 7.926666666666666e-05, "loss": 0.1212, "step": 1190 }, { "grad_norm": 0.9362872838973999, "learning_rate": 7.993333333333334e-05, "loss": 0.1292, "step": 1200 }, { "grad_norm": 1.1560794115066528, "learning_rate": 8.060000000000001e-05, "loss": 0.148, "step": 1210 }, { "grad_norm": 0.5845258831977844, "learning_rate": 8.126666666666667e-05, "loss": 0.1235, "step": 1220 }, { "grad_norm": 0.5799212455749512, "learning_rate": 8.193333333333334e-05, "loss": 0.1216, "step": 1230 }, { "grad_norm": 0.5902547240257263, "learning_rate": 8.26e-05, "loss": 0.1358, "step": 1240 }, { "grad_norm": 0.7473225593566895, "learning_rate": 8.326666666666667e-05, "loss": 0.1169, "step": 1250 }, { "grad_norm": 0.6877524852752686, "learning_rate": 8.393333333333335e-05, "loss": 0.1195, "step": 1260 }, { "grad_norm": 0.5932507514953613, "learning_rate": 8.46e-05, "loss": 0.0941, "step": 1270 }, { "grad_norm": 0.47514939308166504, "learning_rate": 8.526666666666667e-05, "loss": 0.115, "step": 1280 }, { "grad_norm": 0.5834448933601379, "learning_rate": 8.593333333333333e-05, "loss": 0.1083, "step": 1290 }, { "grad_norm": 0.5778251886367798, "learning_rate": 8.66e-05, "loss": 0.1101, "step": 1300 }, { "grad_norm": 0.5453944206237793, "learning_rate": 8.726666666666667e-05, "loss": 0.1087, "step": 1310 }, { "grad_norm": 0.5924835205078125, "learning_rate": 8.793333333333333e-05, "loss": 0.1249, "step": 1320 }, { "grad_norm": 0.6033546924591064, "learning_rate": 8.86e-05, "loss": 0.1028, "step": 1330 }, { "grad_norm": 0.5413753390312195, "learning_rate": 8.926666666666668e-05, "loss": 0.1084, "step": 1340 }, { "grad_norm": 0.6905558705329895, "learning_rate": 8.993333333333334e-05, "loss": 0.135, "step": 1350 }, { "grad_norm": 0.8058986067771912, "learning_rate": 9.06e-05, "loss": 0.1243, "step": 1360 }, { "grad_norm": 0.7169292569160461, "learning_rate": 9.126666666666667e-05, "loss": 0.1105, "step": 1370 }, { "grad_norm": 0.6363813877105713, "learning_rate": 9.193333333333334e-05, "loss": 0.0983, "step": 1380 }, { "grad_norm": 0.691824734210968, "learning_rate": 9.260000000000001e-05, "loss": 0.1182, "step": 1390 }, { "grad_norm": 0.6535894274711609, "learning_rate": 9.326666666666667e-05, "loss": 0.1048, "step": 1400 }, { "grad_norm": 0.5367677211761475, "learning_rate": 9.393333333333334e-05, "loss": 0.1116, "step": 1410 }, { "grad_norm": 0.573167085647583, "learning_rate": 9.46e-05, "loss": 0.1117, "step": 1420 }, { "grad_norm": 0.6312726736068726, "learning_rate": 9.526666666666667e-05, "loss": 0.0997, "step": 1430 }, { "grad_norm": 0.6530237197875977, "learning_rate": 9.593333333333334e-05, "loss": 0.1089, "step": 1440 }, { "grad_norm": 0.5789769887924194, "learning_rate": 9.66e-05, "loss": 0.1294, "step": 1450 }, { "grad_norm": 0.7314766645431519, "learning_rate": 9.726666666666667e-05, "loss": 0.1232, "step": 1460 }, { "grad_norm": 0.5134965181350708, "learning_rate": 9.793333333333333e-05, "loss": 0.1265, "step": 1470 }, { "grad_norm": 0.4784684479236603, "learning_rate": 9.86e-05, "loss": 0.1056, "step": 1480 }, { "grad_norm": 0.4422905147075653, "learning_rate": 9.926666666666668e-05, "loss": 0.1055, "step": 1490 }, { "grad_norm": 0.6670015454292297, "learning_rate": 9.993333333333334e-05, "loss": 0.1139, "step": 1500 }, { "grad_norm": 0.5424293279647827, "learning_rate": 9.999997539434007e-05, "loss": 0.1004, "step": 1510 }, { "grad_norm": 0.6592455506324768, "learning_rate": 9.999989033776898e-05, "loss": 0.1192, "step": 1520 }, { "grad_norm": 0.5724893808364868, "learning_rate": 9.999974452661641e-05, "loss": 0.1182, "step": 1530 }, { "grad_norm": 0.5243816375732422, "learning_rate": 9.999953796105959e-05, "loss": 0.1032, "step": 1540 }, { "grad_norm": 0.47374847531318665, "learning_rate": 9.999927064134949e-05, "loss": 0.1, "step": 1550 }, { "grad_norm": 0.5260648131370544, "learning_rate": 9.999894256781095e-05, "loss": 0.1223, "step": 1560 }, { "grad_norm": 0.5934615731239319, "learning_rate": 9.99985537408426e-05, "loss": 0.0896, "step": 1570 }, { "grad_norm": 0.5248888731002808, "learning_rate": 9.999810416091688e-05, "loss": 0.1018, "step": 1580 }, { "grad_norm": 0.7135077118873596, "learning_rate": 9.99975938285801e-05, "loss": 0.1175, "step": 1590 }, { "grad_norm": 0.5935297608375549, "learning_rate": 9.999702274445236e-05, "loss": 0.1038, "step": 1600 }, { "grad_norm": 0.44233593344688416, "learning_rate": 9.999639090922756e-05, "loss": 0.0898, "step": 1610 }, { "grad_norm": 0.5554015636444092, "learning_rate": 9.999569832367346e-05, "loss": 0.1001, "step": 1620 }, { "grad_norm": 0.44629737734794617, "learning_rate": 9.999494498863162e-05, "loss": 0.1327, "step": 1630 }, { "grad_norm": 0.6388291716575623, "learning_rate": 9.99941309050174e-05, "loss": 0.1089, "step": 1640 }, { "grad_norm": 0.6431708931922913, "learning_rate": 9.999325607381999e-05, "loss": 0.1151, "step": 1650 }, { "grad_norm": 0.4875712990760803, "learning_rate": 9.999232049610238e-05, "loss": 0.1102, "step": 1660 }, { "grad_norm": 0.5300871133804321, "learning_rate": 9.999132417300142e-05, "loss": 0.092, "step": 1670 }, { "grad_norm": 0.6332004070281982, "learning_rate": 9.99902671057277e-05, "loss": 0.1006, "step": 1680 }, { "grad_norm": 0.5419436097145081, "learning_rate": 9.998914929556569e-05, "loss": 0.0966, "step": 1690 }, { "grad_norm": 0.5577907562255859, "learning_rate": 9.998797074387361e-05, "loss": 0.1046, "step": 1700 }, { "grad_norm": 0.6427026987075806, "learning_rate": 9.99867314520835e-05, "loss": 0.1184, "step": 1710 }, { "grad_norm": 0.47363460063934326, "learning_rate": 9.998543142170126e-05, "loss": 0.1048, "step": 1720 }, { "grad_norm": 0.4323226809501648, "learning_rate": 9.99840706543065e-05, "loss": 0.1018, "step": 1730 }, { "grad_norm": 0.5850794315338135, "learning_rate": 9.998264915155274e-05, "loss": 0.1074, "step": 1740 }, { "grad_norm": 0.6395797729492188, "learning_rate": 9.998116691516718e-05, "loss": 0.1114, "step": 1750 }, { "grad_norm": 0.7132892608642578, "learning_rate": 9.997962394695091e-05, "loss": 0.1038, "step": 1760 }, { "grad_norm": 0.5341310501098633, "learning_rate": 9.997802024877875e-05, "loss": 0.0893, "step": 1770 }, { "grad_norm": 0.4844887852668762, "learning_rate": 9.99763558225994e-05, "loss": 0.1247, "step": 1780 }, { "grad_norm": 0.5799452066421509, "learning_rate": 9.997463067043526e-05, "loss": 0.104, "step": 1790 }, { "grad_norm": 0.48172006011009216, "learning_rate": 9.997284479438253e-05, "loss": 0.1054, "step": 1800 }, { "grad_norm": 0.48875951766967773, "learning_rate": 9.997099819661127e-05, "loss": 0.1005, "step": 1810 }, { "grad_norm": 0.6384109258651733, "learning_rate": 9.996909087936524e-05, "loss": 0.1126, "step": 1820 }, { "grad_norm": 0.7000404596328735, "learning_rate": 9.996712284496201e-05, "loss": 0.1145, "step": 1830 }, { "grad_norm": 0.4848990738391876, "learning_rate": 9.996509409579293e-05, "loss": 0.1119, "step": 1840 }, { "grad_norm": 0.5118829607963562, "learning_rate": 9.996300463432312e-05, "loss": 0.1266, "step": 1850 }, { "grad_norm": 0.4704790711402893, "learning_rate": 9.996085446309148e-05, "loss": 0.1196, "step": 1860 }, { "grad_norm": 0.4748091697692871, "learning_rate": 9.995864358471066e-05, "loss": 0.1071, "step": 1870 }, { "grad_norm": 0.5323951840400696, "learning_rate": 9.99563720018671e-05, "loss": 0.0888, "step": 1880 }, { "grad_norm": 0.4245382249355316, "learning_rate": 9.995403971732098e-05, "loss": 0.0936, "step": 1890 }, { "grad_norm": 0.49264609813690186, "learning_rate": 9.995164673390625e-05, "loss": 0.0866, "step": 1900 }, { "grad_norm": 0.5234168171882629, "learning_rate": 9.994919305453059e-05, "loss": 0.1146, "step": 1910 }, { "grad_norm": 0.6327356100082397, "learning_rate": 9.994667868217548e-05, "loss": 0.0959, "step": 1920 }, { "grad_norm": 0.39227941632270813, "learning_rate": 9.99441036198961e-05, "loss": 0.0933, "step": 1930 }, { "grad_norm": 0.4861612915992737, "learning_rate": 9.99414678708214e-05, "loss": 0.0993, "step": 1940 }, { "grad_norm": 0.5046985149383545, "learning_rate": 9.993877143815407e-05, "loss": 0.0873, "step": 1950 }, { "grad_norm": 0.5337247252464294, "learning_rate": 9.993601432517053e-05, "loss": 0.1151, "step": 1960 }, { "grad_norm": 0.4153483808040619, "learning_rate": 9.993319653522091e-05, "loss": 0.082, "step": 1970 }, { "grad_norm": 0.5904155373573303, "learning_rate": 9.993031807172911e-05, "loss": 0.1141, "step": 1980 }, { "grad_norm": 0.5496517419815063, "learning_rate": 9.992737893819273e-05, "loss": 0.0905, "step": 1990 }, { "grad_norm": 0.44806545972824097, "learning_rate": 9.992437913818312e-05, "loss": 0.1036, "step": 2000 }, { "grad_norm": 0.5436198711395264, "learning_rate": 9.992131867534526e-05, "loss": 0.1231, "step": 2010 }, { "grad_norm": 0.6306231021881104, "learning_rate": 9.991819755339796e-05, "loss": 0.1025, "step": 2020 }, { "grad_norm": 0.5338940024375916, "learning_rate": 9.991501577613365e-05, "loss": 0.0786, "step": 2030 }, { "grad_norm": 0.5160660743713379, "learning_rate": 9.99117733474185e-05, "loss": 0.0995, "step": 2040 }, { "grad_norm": 0.5081049203872681, "learning_rate": 9.990847027119234e-05, "loss": 0.098, "step": 2050 }, { "grad_norm": 0.41104191541671753, "learning_rate": 9.990510655146877e-05, "loss": 0.0924, "step": 2060 }, { "grad_norm": 0.4931928217411041, "learning_rate": 9.990168219233496e-05, "loss": 0.0876, "step": 2070 }, { "grad_norm": 0.5019201636314392, "learning_rate": 9.989819719795188e-05, "loss": 0.0996, "step": 2080 }, { "grad_norm": 0.5325645804405212, "learning_rate": 9.989465157255412e-05, "loss": 0.0987, "step": 2090 }, { "grad_norm": 0.600153923034668, "learning_rate": 9.989104532044994e-05, "loss": 0.1007, "step": 2100 }, { "grad_norm": 0.45202505588531494, "learning_rate": 9.988737844602128e-05, "loss": 0.1143, "step": 2110 }, { "grad_norm": 0.4158287048339844, "learning_rate": 9.988365095372372e-05, "loss": 0.0846, "step": 2120 }, { "grad_norm": 0.46842241287231445, "learning_rate": 9.987986284808654e-05, "loss": 0.093, "step": 2130 }, { "grad_norm": 0.44180116057395935, "learning_rate": 9.987601413371264e-05, "loss": 0.0804, "step": 2140 }, { "grad_norm": 0.4573896527290344, "learning_rate": 9.987210481527855e-05, "loss": 0.11, "step": 2150 }, { "grad_norm": 0.3827414810657501, "learning_rate": 9.98681348975345e-05, "loss": 0.0931, "step": 2160 }, { "grad_norm": 0.3857544958591461, "learning_rate": 9.986410438530427e-05, "loss": 0.0871, "step": 2170 }, { "grad_norm": 0.4049851894378662, "learning_rate": 9.986001328348534e-05, "loss": 0.1026, "step": 2180 }, { "grad_norm": 0.572483479976654, "learning_rate": 9.985586159704878e-05, "loss": 0.0909, "step": 2190 }, { "grad_norm": 0.4478301703929901, "learning_rate": 9.985164933103929e-05, "loss": 0.0867, "step": 2200 }, { "grad_norm": 0.3501303791999817, "learning_rate": 9.984737649057513e-05, "loss": 0.0868, "step": 2210 }, { "grad_norm": 0.5308849811553955, "learning_rate": 9.984304308084827e-05, "loss": 0.0905, "step": 2220 }, { "grad_norm": 0.36159029603004456, "learning_rate": 9.983864910712416e-05, "loss": 0.0834, "step": 2230 }, { "grad_norm": 0.3906756043434143, "learning_rate": 9.98341945747419e-05, "loss": 0.0813, "step": 2240 }, { "grad_norm": 0.45696622133255005, "learning_rate": 9.98296794891142e-05, "loss": 0.1036, "step": 2250 }, { "grad_norm": 0.38583239912986755, "learning_rate": 9.982510385572725e-05, "loss": 0.0822, "step": 2260 }, { "grad_norm": 0.3984242379665375, "learning_rate": 9.982046768014094e-05, "loss": 0.0785, "step": 2270 }, { "grad_norm": 0.3725263774394989, "learning_rate": 9.981577096798863e-05, "loss": 0.0785, "step": 2280 }, { "grad_norm": 0.6681470274925232, "learning_rate": 9.981101372497727e-05, "loss": 0.1172, "step": 2290 }, { "grad_norm": 0.4247549772262573, "learning_rate": 9.980619595688737e-05, "loss": 0.0792, "step": 2300 }, { "grad_norm": 0.379131942987442, "learning_rate": 9.980131766957295e-05, "loss": 0.0863, "step": 2310 }, { "grad_norm": 0.43463680148124695, "learning_rate": 9.979637886896163e-05, "loss": 0.0822, "step": 2320 }, { "grad_norm": 0.3485284745693207, "learning_rate": 9.979137956105447e-05, "loss": 0.1017, "step": 2330 }, { "grad_norm": 0.3427072763442993, "learning_rate": 9.978631975192613e-05, "loss": 0.0884, "step": 2340 }, { "grad_norm": 0.40517082810401917, "learning_rate": 9.978119944772475e-05, "loss": 0.0926, "step": 2350 }, { "grad_norm": 0.3981311619281769, "learning_rate": 9.977601865467197e-05, "loss": 0.0885, "step": 2360 }, { "grad_norm": 0.402889609336853, "learning_rate": 9.977077737906297e-05, "loss": 0.0881, "step": 2370 }, { "grad_norm": 0.39083513617515564, "learning_rate": 9.976547562726636e-05, "loss": 0.0927, "step": 2380 }, { "grad_norm": 0.3705734312534332, "learning_rate": 9.976011340572429e-05, "loss": 0.0928, "step": 2390 }, { "grad_norm": 0.4495719075202942, "learning_rate": 9.975469072095237e-05, "loss": 0.0893, "step": 2400 }, { "grad_norm": 0.3956231474876404, "learning_rate": 9.974920757953965e-05, "loss": 0.0895, "step": 2410 }, { "grad_norm": 0.46903565526008606, "learning_rate": 9.97436639881487e-05, "loss": 0.1279, "step": 2420 }, { "grad_norm": 0.4369224011898041, "learning_rate": 9.973805995351545e-05, "loss": 0.1002, "step": 2430 }, { "grad_norm": 0.5357306599617004, "learning_rate": 9.973239548244939e-05, "loss": 0.1008, "step": 2440 }, { "grad_norm": 0.4203925132751465, "learning_rate": 9.972667058183333e-05, "loss": 0.0837, "step": 2450 }, { "grad_norm": 0.4359923005104065, "learning_rate": 9.972088525862362e-05, "loss": 0.0813, "step": 2460 }, { "grad_norm": 0.5051701664924622, "learning_rate": 9.971503951984995e-05, "loss": 0.0818, "step": 2470 }, { "grad_norm": 0.4398789405822754, "learning_rate": 9.970913337261543e-05, "loss": 0.0968, "step": 2480 }, { "grad_norm": 0.40990710258483887, "learning_rate": 9.97031668240966e-05, "loss": 0.0852, "step": 2490 }, { "grad_norm": 0.3925633132457733, "learning_rate": 9.969713988154339e-05, "loss": 0.0718, "step": 2500 }, { "grad_norm": 0.3791011869907379, "learning_rate": 9.969105255227906e-05, "loss": 0.0729, "step": 2510 }, { "grad_norm": 0.4883902072906494, "learning_rate": 9.968490484370035e-05, "loss": 0.0807, "step": 2520 }, { "grad_norm": 0.38710153102874756, "learning_rate": 9.967869676327726e-05, "loss": 0.0712, "step": 2530 }, { "grad_norm": 0.4061811864376068, "learning_rate": 9.96724283185532e-05, "loss": 0.0837, "step": 2540 }, { "grad_norm": 0.366252064704895, "learning_rate": 9.966609951714494e-05, "loss": 0.0942, "step": 2550 }, { "grad_norm": 0.3941636383533478, "learning_rate": 9.965971036674255e-05, "loss": 0.0915, "step": 2560 }, { "grad_norm": 0.4702654480934143, "learning_rate": 9.965326087510947e-05, "loss": 0.0881, "step": 2570 }, { "grad_norm": 0.336631178855896, "learning_rate": 9.964675105008244e-05, "loss": 0.0842, "step": 2580 }, { "grad_norm": 0.3802169859409332, "learning_rate": 9.964018089957147e-05, "loss": 0.081, "step": 2590 }, { "grad_norm": 0.3639869689941406, "learning_rate": 9.963355043155996e-05, "loss": 0.0889, "step": 2600 }, { "grad_norm": 0.44370871782302856, "learning_rate": 9.962685965410455e-05, "loss": 0.0856, "step": 2610 }, { "grad_norm": 0.4446844160556793, "learning_rate": 9.962010857533514e-05, "loss": 0.076, "step": 2620 }, { "grad_norm": 0.5931194424629211, "learning_rate": 9.961329720345493e-05, "loss": 0.0801, "step": 2630 }, { "grad_norm": 0.42624562978744507, "learning_rate": 9.96064255467404e-05, "loss": 0.0947, "step": 2640 }, { "grad_norm": 0.31102001667022705, "learning_rate": 9.959949361354126e-05, "loss": 0.0945, "step": 2650 }, { "grad_norm": 0.49821650981903076, "learning_rate": 9.959250141228045e-05, "loss": 0.0947, "step": 2660 }, { "grad_norm": 0.43530580401420593, "learning_rate": 9.958544895145414e-05, "loss": 0.1008, "step": 2670 }, { "grad_norm": 0.3874141573905945, "learning_rate": 9.957833623963177e-05, "loss": 0.108, "step": 2680 }, { "grad_norm": 0.3330201506614685, "learning_rate": 9.957116328545593e-05, "loss": 0.0795, "step": 2690 }, { "grad_norm": 0.37822309136390686, "learning_rate": 9.956393009764244e-05, "loss": 0.0703, "step": 2700 }, { "grad_norm": 0.3816111385822296, "learning_rate": 9.955663668498032e-05, "loss": 0.0797, "step": 2710 }, { "grad_norm": 0.45314934849739075, "learning_rate": 9.954928305633173e-05, "loss": 0.1025, "step": 2720 }, { "grad_norm": 0.46508336067199707, "learning_rate": 9.954186922063204e-05, "loss": 0.0843, "step": 2730 }, { "grad_norm": 0.32668551802635193, "learning_rate": 9.953439518688974e-05, "loss": 0.0799, "step": 2740 }, { "grad_norm": 0.3508513867855072, "learning_rate": 9.952686096418652e-05, "loss": 0.0877, "step": 2750 }, { "grad_norm": 0.4518144428730011, "learning_rate": 9.951926656167714e-05, "loss": 0.0799, "step": 2760 }, { "grad_norm": 0.36115148663520813, "learning_rate": 9.951161198858953e-05, "loss": 0.084, "step": 2770 }, { "grad_norm": 0.3468472957611084, "learning_rate": 9.950389725422471e-05, "loss": 0.0931, "step": 2780 }, { "grad_norm": 0.4942661225795746, "learning_rate": 9.949612236795682e-05, "loss": 0.0876, "step": 2790 }, { "grad_norm": 0.39886510372161865, "learning_rate": 9.948828733923305e-05, "loss": 0.0781, "step": 2800 }, { "grad_norm": 0.4604346752166748, "learning_rate": 9.948039217757374e-05, "loss": 0.1043, "step": 2810 }, { "grad_norm": 0.43620482087135315, "learning_rate": 9.947243689257225e-05, "loss": 0.0882, "step": 2820 }, { "grad_norm": 0.3865968585014343, "learning_rate": 9.946442149389497e-05, "loss": 0.0823, "step": 2830 }, { "grad_norm": 0.4155484437942505, "learning_rate": 9.945634599128139e-05, "loss": 0.0976, "step": 2840 }, { "grad_norm": 0.4456450045108795, "learning_rate": 9.944821039454402e-05, "loss": 0.1003, "step": 2850 }, { "grad_norm": 0.32964199781417847, "learning_rate": 9.944001471356835e-05, "loss": 0.0807, "step": 2860 }, { "grad_norm": 0.4506993889808655, "learning_rate": 9.94317589583129e-05, "loss": 0.0784, "step": 2870 }, { "grad_norm": 0.3558453917503357, "learning_rate": 9.942344313880922e-05, "loss": 0.0728, "step": 2880 }, { "grad_norm": 0.39189061522483826, "learning_rate": 9.941506726516179e-05, "loss": 0.0835, "step": 2890 }, { "grad_norm": 0.3118578791618347, "learning_rate": 9.94066313475481e-05, "loss": 0.0906, "step": 2900 }, { "grad_norm": 0.5767695307731628, "learning_rate": 9.939813539621857e-05, "loss": 0.0862, "step": 2910 }, { "grad_norm": 0.4470449388027191, "learning_rate": 9.93895794214966e-05, "loss": 0.0862, "step": 2920 }, { "grad_norm": 0.4411753714084625, "learning_rate": 9.938096343377852e-05, "loss": 0.07, "step": 2930 }, { "grad_norm": 0.377742737531662, "learning_rate": 9.937228744353353e-05, "loss": 0.0721, "step": 2940 }, { "grad_norm": 0.4582287669181824, "learning_rate": 9.936355146130379e-05, "loss": 0.0743, "step": 2950 }, { "grad_norm": 0.35823315382003784, "learning_rate": 9.935475549770435e-05, "loss": 0.0891, "step": 2960 }, { "grad_norm": 0.3655441403388977, "learning_rate": 9.934589956342315e-05, "loss": 0.0851, "step": 2970 }, { "grad_norm": 0.46333637833595276, "learning_rate": 9.933698366922093e-05, "loss": 0.086, "step": 2980 }, { "grad_norm": 0.4037494659423828, "learning_rate": 9.93280078259314e-05, "loss": 0.0986, "step": 2990 }, { "grad_norm": 0.42796680331230164, "learning_rate": 9.931897204446104e-05, "loss": 0.0974, "step": 3000 }, { "grad_norm": 0.39216646552085876, "learning_rate": 9.930987633578915e-05, "loss": 0.0776, "step": 3010 }, { "grad_norm": 0.40791165828704834, "learning_rate": 9.93007207109679e-05, "loss": 0.0799, "step": 3020 }, { "grad_norm": 0.29888173937797546, "learning_rate": 9.929150518112224e-05, "loss": 0.0777, "step": 3030 }, { "grad_norm": 0.36772024631500244, "learning_rate": 9.928222975744991e-05, "loss": 0.0852, "step": 3040 }, { "grad_norm": 0.40149396657943726, "learning_rate": 9.92728944512214e-05, "loss": 0.0837, "step": 3050 }, { "grad_norm": 0.4948907196521759, "learning_rate": 9.926349927378001e-05, "loss": 0.0869, "step": 3060 }, { "grad_norm": 0.2891085147857666, "learning_rate": 9.925404423654174e-05, "loss": 0.0701, "step": 3070 }, { "grad_norm": 0.4487878382205963, "learning_rate": 9.924452935099537e-05, "loss": 0.0814, "step": 3080 }, { "grad_norm": 0.5200735330581665, "learning_rate": 9.92349546287024e-05, "loss": 0.1063, "step": 3090 }, { "grad_norm": 0.4704391658306122, "learning_rate": 9.9225320081297e-05, "loss": 0.0879, "step": 3100 }, { "grad_norm": 0.3586313724517822, "learning_rate": 9.921562572048606e-05, "loss": 0.0762, "step": 3110 }, { "grad_norm": 0.41285327076911926, "learning_rate": 9.920587155804913e-05, "loss": 0.072, "step": 3120 }, { "grad_norm": 0.3336651921272278, "learning_rate": 9.919605760583845e-05, "loss": 0.0726, "step": 3130 }, { "grad_norm": 0.4118559956550598, "learning_rate": 9.91861838757789e-05, "loss": 0.0783, "step": 3140 }, { "grad_norm": 0.30703088641166687, "learning_rate": 9.917625037986798e-05, "loss": 0.0729, "step": 3150 }, { "grad_norm": 0.521151602268219, "learning_rate": 9.916625713017583e-05, "loss": 0.0797, "step": 3160 }, { "grad_norm": 0.31585174798965454, "learning_rate": 9.915620413884519e-05, "loss": 0.0809, "step": 3170 }, { "grad_norm": 0.3435116112232208, "learning_rate": 9.914609141809139e-05, "loss": 0.0962, "step": 3180 }, { "grad_norm": 0.2976171374320984, "learning_rate": 9.913591898020235e-05, "loss": 0.0924, "step": 3190 }, { "grad_norm": 0.37509864568710327, "learning_rate": 9.912568683753853e-05, "loss": 0.0808, "step": 3200 }, { "grad_norm": 0.35193321108818054, "learning_rate": 9.911539500253295e-05, "loss": 0.0724, "step": 3210 }, { "grad_norm": 0.5125724673271179, "learning_rate": 9.910504348769118e-05, "loss": 0.1067, "step": 3220 }, { "grad_norm": 0.33397117257118225, "learning_rate": 9.909463230559127e-05, "loss": 0.0923, "step": 3230 }, { "grad_norm": 0.42479535937309265, "learning_rate": 9.908416146888376e-05, "loss": 0.0755, "step": 3240 }, { "grad_norm": 0.41101396083831787, "learning_rate": 9.907363099029175e-05, "loss": 0.0875, "step": 3250 }, { "grad_norm": 0.43507862091064453, "learning_rate": 9.906304088261073e-05, "loss": 0.088, "step": 3260 }, { "grad_norm": 0.37717124819755554, "learning_rate": 9.905239115870872e-05, "loss": 0.0799, "step": 3270 }, { "grad_norm": 0.3911370038986206, "learning_rate": 9.90416818315261e-05, "loss": 0.0692, "step": 3280 }, { "grad_norm": 0.45878443121910095, "learning_rate": 9.903091291407573e-05, "loss": 0.0771, "step": 3290 }, { "grad_norm": 0.33283111453056335, "learning_rate": 9.902008441944286e-05, "loss": 0.0764, "step": 3300 }, { "grad_norm": 0.3886740207672119, "learning_rate": 9.900919636078512e-05, "loss": 0.0858, "step": 3310 }, { "grad_norm": 0.3212319314479828, "learning_rate": 9.899824875133255e-05, "loss": 0.1037, "step": 3320 }, { "grad_norm": 0.33279192447662354, "learning_rate": 9.898724160438749e-05, "loss": 0.0784, "step": 3330 }, { "grad_norm": 0.41042816638946533, "learning_rate": 9.89761749333247e-05, "loss": 0.0793, "step": 3340 }, { "grad_norm": 0.4451558291912079, "learning_rate": 9.896504875159122e-05, "loss": 0.078, "step": 3350 }, { "grad_norm": 0.41818737983703613, "learning_rate": 9.89538630727064e-05, "loss": 0.0751, "step": 3360 }, { "grad_norm": 0.2754257321357727, "learning_rate": 9.894261791026189e-05, "loss": 0.0784, "step": 3370 }, { "grad_norm": 0.3439501225948334, "learning_rate": 9.893131327792165e-05, "loss": 0.1231, "step": 3380 }, { "grad_norm": 0.4869624972343445, "learning_rate": 9.891994918942182e-05, "loss": 0.0726, "step": 3390 }, { "grad_norm": 0.3778139054775238, "learning_rate": 9.890852565857092e-05, "loss": 0.0793, "step": 3400 }, { "grad_norm": 0.30616796016693115, "learning_rate": 9.889704269924954e-05, "loss": 0.0784, "step": 3410 }, { "grad_norm": 0.4344424307346344, "learning_rate": 9.888550032541059e-05, "loss": 0.0667, "step": 3420 }, { "grad_norm": 0.3332769274711609, "learning_rate": 9.887389855107916e-05, "loss": 0.0789, "step": 3430 }, { "grad_norm": 0.3972417414188385, "learning_rate": 9.886223739035248e-05, "loss": 0.0854, "step": 3440 }, { "grad_norm": 0.34755605459213257, "learning_rate": 9.885051685739997e-05, "loss": 0.0721, "step": 3450 }, { "grad_norm": 0.3642781376838684, "learning_rate": 9.883873696646316e-05, "loss": 0.069, "step": 3460 }, { "grad_norm": 0.36320212483406067, "learning_rate": 9.882689773185575e-05, "loss": 0.0816, "step": 3470 }, { "grad_norm": 0.35150855779647827, "learning_rate": 9.881499916796353e-05, "loss": 0.0824, "step": 3480 }, { "grad_norm": 0.3586837649345398, "learning_rate": 9.880304128924434e-05, "loss": 0.0822, "step": 3490 }, { "grad_norm": 0.3634629249572754, "learning_rate": 9.879102411022817e-05, "loss": 0.067, "step": 3500 }, { "grad_norm": 0.3698098957538605, "learning_rate": 9.877894764551703e-05, "loss": 0.0715, "step": 3510 }, { "grad_norm": 0.2860109508037567, "learning_rate": 9.876681190978494e-05, "loss": 0.0703, "step": 3520 }, { "grad_norm": 0.36054855585098267, "learning_rate": 9.875461691777797e-05, "loss": 0.0754, "step": 3530 }, { "grad_norm": 0.3345522880554199, "learning_rate": 9.874236268431417e-05, "loss": 0.0788, "step": 3540 }, { "grad_norm": 0.3613918721675873, "learning_rate": 9.873004922428361e-05, "loss": 0.0828, "step": 3550 }, { "grad_norm": 0.3282865583896637, "learning_rate": 9.871767655264829e-05, "loss": 0.0649, "step": 3560 }, { "grad_norm": 0.39156654477119446, "learning_rate": 9.87052446844422e-05, "loss": 0.0916, "step": 3570 }, { "grad_norm": 0.34197404980659485, "learning_rate": 9.869275363477122e-05, "loss": 0.0934, "step": 3580 }, { "grad_norm": 0.40387439727783203, "learning_rate": 9.868020341881312e-05, "loss": 0.0934, "step": 3590 }, { "grad_norm": 0.42230018973350525, "learning_rate": 9.866759405181765e-05, "loss": 0.0796, "step": 3600 }, { "grad_norm": 0.3891996443271637, "learning_rate": 9.865492554910633e-05, "loss": 0.0792, "step": 3610 }, { "grad_norm": 0.32976168394088745, "learning_rate": 9.864219792607262e-05, "loss": 0.075, "step": 3620 }, { "grad_norm": 0.38950809836387634, "learning_rate": 9.862941119818177e-05, "loss": 0.0633, "step": 3630 }, { "grad_norm": 0.4041704535484314, "learning_rate": 9.861656538097086e-05, "loss": 0.0731, "step": 3640 }, { "grad_norm": 0.36189284920692444, "learning_rate": 9.860366049004877e-05, "loss": 0.0815, "step": 3650 }, { "grad_norm": 0.39249521493911743, "learning_rate": 9.859069654109615e-05, "loss": 0.0745, "step": 3660 }, { "grad_norm": 0.31761810183525085, "learning_rate": 9.857767354986545e-05, "loss": 0.0789, "step": 3670 }, { "grad_norm": 0.40047577023506165, "learning_rate": 9.856459153218078e-05, "loss": 0.0769, "step": 3680 }, { "grad_norm": 0.35106852650642395, "learning_rate": 9.855145050393808e-05, "loss": 0.069, "step": 3690 }, { "grad_norm": 0.2970884442329407, "learning_rate": 9.85382504811049e-05, "loss": 0.068, "step": 3700 }, { "grad_norm": 0.43520399928092957, "learning_rate": 9.852499147972054e-05, "loss": 0.0779, "step": 3710 }, { "grad_norm": 0.42946574091911316, "learning_rate": 9.851167351589592e-05, "loss": 0.0774, "step": 3720 }, { "grad_norm": 0.3652973771095276, "learning_rate": 9.849829660581363e-05, "loss": 0.0622, "step": 3730 }, { "grad_norm": 0.33622822165489197, "learning_rate": 9.848486076572787e-05, "loss": 0.0656, "step": 3740 }, { "grad_norm": 0.2991586923599243, "learning_rate": 9.847136601196446e-05, "loss": 0.0796, "step": 3750 }, { "grad_norm": 0.41241970658302307, "learning_rate": 9.845781236092078e-05, "loss": 0.0602, "step": 3760 }, { "grad_norm": 0.44666680693626404, "learning_rate": 9.844419982906583e-05, "loss": 0.0792, "step": 3770 }, { "grad_norm": 0.3048223555088043, "learning_rate": 9.843052843294008e-05, "loss": 0.0745, "step": 3780 }, { "grad_norm": 0.3770085573196411, "learning_rate": 9.841679818915559e-05, "loss": 0.0669, "step": 3790 }, { "grad_norm": 0.347969114780426, "learning_rate": 9.840300911439591e-05, "loss": 0.1006, "step": 3800 }, { "grad_norm": 0.2679331302642822, "learning_rate": 9.838916122541603e-05, "loss": 0.0652, "step": 3810 }, { "grad_norm": 0.3670705258846283, "learning_rate": 9.837525453904246e-05, "loss": 0.0764, "step": 3820 }, { "grad_norm": 0.339759886264801, "learning_rate": 9.836128907217314e-05, "loss": 0.0635, "step": 3830 }, { "grad_norm": 0.3293312191963196, "learning_rate": 9.834726484177743e-05, "loss": 0.0679, "step": 3840 }, { "grad_norm": 0.37540292739868164, "learning_rate": 9.833318186489609e-05, "loss": 0.0823, "step": 3850 }, { "grad_norm": 0.37442871928215027, "learning_rate": 9.831904015864126e-05, "loss": 0.0637, "step": 3860 }, { "grad_norm": 0.35796207189559937, "learning_rate": 9.830483974019645e-05, "loss": 0.065, "step": 3870 }, { "grad_norm": 0.42522764205932617, "learning_rate": 9.82905806268165e-05, "loss": 0.0647, "step": 3880 }, { "grad_norm": 0.3776293694972992, "learning_rate": 9.82762628358276e-05, "loss": 0.0706, "step": 3890 }, { "grad_norm": 0.3231024444103241, "learning_rate": 9.826188638462718e-05, "loss": 0.0754, "step": 3900 }, { "grad_norm": 0.34337443113327026, "learning_rate": 9.824745129068402e-05, "loss": 0.0663, "step": 3910 }, { "grad_norm": 0.34903934597969055, "learning_rate": 9.82329575715381e-05, "loss": 0.0776, "step": 3920 }, { "grad_norm": 0.32079896330833435, "learning_rate": 9.821840524480066e-05, "loss": 0.0678, "step": 3930 }, { "grad_norm": 0.35653141140937805, "learning_rate": 9.820379432815414e-05, "loss": 0.0853, "step": 3940 }, { "grad_norm": 0.29968544840812683, "learning_rate": 9.81891248393522e-05, "loss": 0.0746, "step": 3950 }, { "grad_norm": 0.4232248067855835, "learning_rate": 9.817439679621963e-05, "loss": 0.0843, "step": 3960 }, { "grad_norm": 0.42368650436401367, "learning_rate": 9.815961021665243e-05, "loss": 0.0633, "step": 3970 }, { "grad_norm": 0.4561800956726074, "learning_rate": 9.814476511861763e-05, "loss": 0.076, "step": 3980 }, { "grad_norm": 0.4436664283275604, "learning_rate": 9.812986152015348e-05, "loss": 0.0777, "step": 3990 }, { "grad_norm": 0.37550675868988037, "learning_rate": 9.811489943936922e-05, "loss": 0.0798, "step": 4000 }, { "grad_norm": 0.3713410496711731, "learning_rate": 9.809987889444522e-05, "loss": 0.0736, "step": 4010 }, { "grad_norm": 0.3269137144088745, "learning_rate": 9.808479990363282e-05, "loss": 0.0706, "step": 4020 }, { "grad_norm": 0.37039443850517273, "learning_rate": 9.806966248525445e-05, "loss": 0.0768, "step": 4030 }, { "grad_norm": 0.33045583963394165, "learning_rate": 9.805446665770348e-05, "loss": 0.0555, "step": 4040 }, { "grad_norm": 0.32815346121788025, "learning_rate": 9.803921243944429e-05, "loss": 0.0635, "step": 4050 }, { "grad_norm": 0.33596983551979065, "learning_rate": 9.802389984901218e-05, "loss": 0.0652, "step": 4060 }, { "grad_norm": 0.3996434509754181, "learning_rate": 9.80085289050134e-05, "loss": 0.0621, "step": 4070 }, { "grad_norm": 0.30065611004829407, "learning_rate": 9.799309962612508e-05, "loss": 0.0773, "step": 4080 }, { "grad_norm": 0.29930317401885986, "learning_rate": 9.797761203109527e-05, "loss": 0.0711, "step": 4090 }, { "grad_norm": 0.37649670243263245, "learning_rate": 9.796206613874283e-05, "loss": 0.0609, "step": 4100 }, { "grad_norm": 0.3771814703941345, "learning_rate": 9.794646196795754e-05, "loss": 0.0705, "step": 4110 }, { "grad_norm": 0.4576139748096466, "learning_rate": 9.793079953769987e-05, "loss": 0.0765, "step": 4120 }, { "grad_norm": 0.3788709044456482, "learning_rate": 9.79150788670012e-05, "loss": 0.0687, "step": 4130 }, { "grad_norm": 0.4399116635322571, "learning_rate": 9.78992999749636e-05, "loss": 0.0691, "step": 4140 }, { "grad_norm": 0.34906673431396484, "learning_rate": 9.788346288075994e-05, "loss": 0.0864, "step": 4150 }, { "grad_norm": 0.3049456477165222, "learning_rate": 9.786756760363373e-05, "loss": 0.0667, "step": 4160 }, { "grad_norm": 0.30384546518325806, "learning_rate": 9.78516141628993e-05, "loss": 0.062, "step": 4170 }, { "grad_norm": 0.39459142088890076, "learning_rate": 9.783560257794154e-05, "loss": 0.0676, "step": 4180 }, { "grad_norm": 0.3868788480758667, "learning_rate": 9.781953286821603e-05, "loss": 0.0672, "step": 4190 }, { "grad_norm": 0.3191371262073517, "learning_rate": 9.780340505324901e-05, "loss": 0.0704, "step": 4200 }, { "grad_norm": 0.2825709283351898, "learning_rate": 9.778721915263727e-05, "loss": 0.0673, "step": 4210 }, { "grad_norm": 0.264251172542572, "learning_rate": 9.777097518604824e-05, "loss": 0.0791, "step": 4220 }, { "grad_norm": 0.358866423368454, "learning_rate": 9.775467317321984e-05, "loss": 0.0831, "step": 4230 }, { "grad_norm": 0.3779652714729309, "learning_rate": 9.773831313396055e-05, "loss": 0.078, "step": 4240 }, { "grad_norm": 0.3109897971153259, "learning_rate": 9.77218950881494e-05, "loss": 0.0736, "step": 4250 }, { "grad_norm": 0.4136536121368408, "learning_rate": 9.770541905573583e-05, "loss": 0.0776, "step": 4260 }, { "grad_norm": 0.3865390121936798, "learning_rate": 9.768888505673976e-05, "loss": 0.0753, "step": 4270 }, { "grad_norm": 0.3999148905277252, "learning_rate": 9.767229311125162e-05, "loss": 0.0628, "step": 4280 }, { "grad_norm": 0.3760717213153839, "learning_rate": 9.765564323943211e-05, "loss": 0.0714, "step": 4290 }, { "grad_norm": 0.3280054032802582, "learning_rate": 9.763893546151244e-05, "loss": 0.0608, "step": 4300 }, { "grad_norm": 0.3601842224597931, "learning_rate": 9.762216979779412e-05, "loss": 0.07, "step": 4310 }, { "grad_norm": 0.39024171233177185, "learning_rate": 9.760534626864902e-05, "loss": 0.0977, "step": 4320 }, { "grad_norm": 0.375029593706131, "learning_rate": 9.758846489451931e-05, "loss": 0.0647, "step": 4330 }, { "grad_norm": 0.31465184688568115, "learning_rate": 9.757152569591748e-05, "loss": 0.0556, "step": 4340 }, { "grad_norm": 0.33993780612945557, "learning_rate": 9.75545286934262e-05, "loss": 0.0514, "step": 4350 }, { "grad_norm": 0.4763353168964386, "learning_rate": 9.753747390769847e-05, "loss": 0.0604, "step": 4360 }, { "grad_norm": 0.2842327654361725, "learning_rate": 9.752036135945744e-05, "loss": 0.0715, "step": 4370 }, { "grad_norm": 0.25574740767478943, "learning_rate": 9.750319106949649e-05, "loss": 0.0658, "step": 4380 }, { "grad_norm": 0.3205084502696991, "learning_rate": 9.748596305867913e-05, "loss": 0.0553, "step": 4390 }, { "grad_norm": 0.278474897146225, "learning_rate": 9.746867734793903e-05, "loss": 0.0797, "step": 4400 }, { "grad_norm": 0.31896743178367615, "learning_rate": 9.745133395827993e-05, "loss": 0.0681, "step": 4410 }, { "grad_norm": 0.3283815383911133, "learning_rate": 9.743393291077572e-05, "loss": 0.0793, "step": 4420 }, { "grad_norm": 0.3668324053287506, "learning_rate": 9.741647422657028e-05, "loss": 0.0698, "step": 4430 }, { "grad_norm": 0.3674434721469879, "learning_rate": 9.739895792687758e-05, "loss": 0.0686, "step": 4440 }, { "grad_norm": 0.2978595793247223, "learning_rate": 9.738138403298157e-05, "loss": 0.0754, "step": 4450 }, { "grad_norm": 0.38436946272850037, "learning_rate": 9.736375256623619e-05, "loss": 0.0662, "step": 4460 }, { "grad_norm": 0.43455782532691956, "learning_rate": 9.734606354806533e-05, "loss": 0.0665, "step": 4470 }, { "grad_norm": 0.2865617871284485, "learning_rate": 9.73283169999628e-05, "loss": 0.0757, "step": 4480 }, { "grad_norm": 0.2843056917190552, "learning_rate": 9.731051294349238e-05, "loss": 0.0658, "step": 4490 }, { "grad_norm": 0.2967698872089386, "learning_rate": 9.729265140028762e-05, "loss": 0.0665, "step": 4500 }, { "grad_norm": 0.3098256289958954, "learning_rate": 9.727473239205201e-05, "loss": 0.0968, "step": 4510 }, { "grad_norm": 0.3413183391094208, "learning_rate": 9.725675594055883e-05, "loss": 0.0723, "step": 4520 }, { "grad_norm": 0.32757699489593506, "learning_rate": 9.723872206765116e-05, "loss": 0.0714, "step": 4530 }, { "grad_norm": 0.35564085841178894, "learning_rate": 9.722063079524185e-05, "loss": 0.085, "step": 4540 }, { "grad_norm": 0.28030404448509216, "learning_rate": 9.720248214531351e-05, "loss": 0.078, "step": 4550 }, { "grad_norm": 0.2936975955963135, "learning_rate": 9.718427613991848e-05, "loss": 0.0638, "step": 4560 }, { "grad_norm": 0.3403068780899048, "learning_rate": 9.716601280117873e-05, "loss": 0.075, "step": 4570 }, { "grad_norm": 0.3703251779079437, "learning_rate": 9.714769215128596e-05, "loss": 0.0661, "step": 4580 }, { "grad_norm": 0.2512607276439667, "learning_rate": 9.712931421250152e-05, "loss": 0.0631, "step": 4590 }, { "grad_norm": 0.30206209421157837, "learning_rate": 9.711087900715627e-05, "loss": 0.0631, "step": 4600 }, { "grad_norm": 0.5077812075614929, "learning_rate": 9.709238655765078e-05, "loss": 0.0701, "step": 4610 }, { "grad_norm": 0.28740963339805603, "learning_rate": 9.707383688645511e-05, "loss": 0.0597, "step": 4620 }, { "grad_norm": 0.4405771493911743, "learning_rate": 9.705523001610883e-05, "loss": 0.0736, "step": 4630 }, { "grad_norm": 0.30801647901535034, "learning_rate": 9.703656596922107e-05, "loss": 0.0695, "step": 4640 }, { "grad_norm": 0.37961673736572266, "learning_rate": 9.70178447684704e-05, "loss": 0.0607, "step": 4650 }, { "grad_norm": 0.33177971839904785, "learning_rate": 9.699906643660483e-05, "loss": 0.0699, "step": 4660 }, { "grad_norm": 0.36621785163879395, "learning_rate": 9.698023099644185e-05, "loss": 0.0649, "step": 4670 }, { "grad_norm": 0.31450119614601135, "learning_rate": 9.696133847086823e-05, "loss": 0.069, "step": 4680 }, { "grad_norm": 0.3675762414932251, "learning_rate": 9.694238888284022e-05, "loss": 0.0801, "step": 4690 }, { "grad_norm": 0.3316011428833008, "learning_rate": 9.692338225538333e-05, "loss": 0.0722, "step": 4700 }, { "grad_norm": 0.35293522477149963, "learning_rate": 9.690431861159241e-05, "loss": 0.0678, "step": 4710 }, { "grad_norm": 0.37698665261268616, "learning_rate": 9.688519797463161e-05, "loss": 0.0626, "step": 4720 }, { "grad_norm": 0.27976125478744507, "learning_rate": 9.686602036773426e-05, "loss": 0.0829, "step": 4730 }, { "grad_norm": 0.41071072220802307, "learning_rate": 9.684678581420302e-05, "loss": 0.0671, "step": 4740 }, { "grad_norm": 0.30273082852363586, "learning_rate": 9.682749433740962e-05, "loss": 0.056, "step": 4750 }, { "grad_norm": 0.3606887757778168, "learning_rate": 9.680814596079507e-05, "loss": 0.0707, "step": 4760 }, { "grad_norm": 0.42008915543556213, "learning_rate": 9.678874070786945e-05, "loss": 0.0742, "step": 4770 }, { "grad_norm": 0.4121362864971161, "learning_rate": 9.676927860221199e-05, "loss": 0.0834, "step": 4780 }, { "grad_norm": 0.41364142298698425, "learning_rate": 9.674975966747097e-05, "loss": 0.0594, "step": 4790 }, { "grad_norm": 0.3277043104171753, "learning_rate": 9.673018392736374e-05, "loss": 0.0672, "step": 4800 }, { "grad_norm": 0.38060125708580017, "learning_rate": 9.671055140567667e-05, "loss": 0.0711, "step": 4810 }, { "grad_norm": 0.3424491286277771, "learning_rate": 9.669086212626511e-05, "loss": 0.074, "step": 4820 }, { "grad_norm": 0.3166457712650299, "learning_rate": 9.667111611305341e-05, "loss": 0.0677, "step": 4830 }, { "grad_norm": 0.37099090218544006, "learning_rate": 9.665131339003486e-05, "loss": 0.0681, "step": 4840 }, { "grad_norm": 0.32075679302215576, "learning_rate": 9.663145398127158e-05, "loss": 0.0669, "step": 4850 }, { "grad_norm": 0.2723931074142456, "learning_rate": 9.661153791089467e-05, "loss": 0.0599, "step": 4860 }, { "grad_norm": 0.3884235918521881, "learning_rate": 9.659156520310402e-05, "loss": 0.0681, "step": 4870 }, { "grad_norm": 0.36726728081703186, "learning_rate": 9.657153588216835e-05, "loss": 0.0621, "step": 4880 }, { "grad_norm": 0.3652724623680115, "learning_rate": 9.655144997242516e-05, "loss": 0.0848, "step": 4890 }, { "grad_norm": 0.3042035698890686, "learning_rate": 9.653130749828075e-05, "loss": 0.0568, "step": 4900 }, { "grad_norm": 0.4046174883842468, "learning_rate": 9.65111084842101e-05, "loss": 0.0652, "step": 4910 }, { "grad_norm": 0.28950849175453186, "learning_rate": 9.649085295475695e-05, "loss": 0.0699, "step": 4920 }, { "grad_norm": 0.36150306463241577, "learning_rate": 9.647054093453365e-05, "loss": 0.0739, "step": 4930 }, { "grad_norm": 0.3459465801715851, "learning_rate": 9.645017244822123e-05, "loss": 0.0649, "step": 4940 }, { "grad_norm": 0.31890907883644104, "learning_rate": 9.642974752056931e-05, "loss": 0.0695, "step": 4950 }, { "grad_norm": 0.31441301107406616, "learning_rate": 9.640926617639613e-05, "loss": 0.0597, "step": 4960 }, { "grad_norm": 0.2928606867790222, "learning_rate": 9.638872844058843e-05, "loss": 0.0748, "step": 4970 }, { "grad_norm": 0.32576534152030945, "learning_rate": 9.63681343381015e-05, "loss": 0.0691, "step": 4980 }, { "grad_norm": 0.35987862944602966, "learning_rate": 9.634748389395914e-05, "loss": 0.0684, "step": 4990 }, { "grad_norm": 0.4038085341453552, "learning_rate": 9.632677713325353e-05, "loss": 0.0809, "step": 5000 }, { "grad_norm": 0.36913588643074036, "learning_rate": 9.63060140811454e-05, "loss": 0.0755, "step": 5010 }, { "grad_norm": 0.4843292236328125, "learning_rate": 9.628519476286379e-05, "loss": 0.0735, "step": 5020 }, { "grad_norm": 0.2809702754020691, "learning_rate": 9.626431920370612e-05, "loss": 0.0582, "step": 5030 }, { "grad_norm": 0.33649909496307373, "learning_rate": 9.624338742903819e-05, "loss": 0.0618, "step": 5040 }, { "grad_norm": 0.3287201523780823, "learning_rate": 9.622239946429406e-05, "loss": 0.075, "step": 5050 }, { "grad_norm": 0.33899977803230286, "learning_rate": 9.620135533497609e-05, "loss": 0.0469, "step": 5060 }, { "grad_norm": 0.28731414675712585, "learning_rate": 9.61802550666549e-05, "loss": 0.0857, "step": 5070 }, { "grad_norm": 0.3478144407272339, "learning_rate": 9.615909868496928e-05, "loss": 0.0696, "step": 5080 }, { "grad_norm": 0.3435647785663605, "learning_rate": 9.613788621562622e-05, "loss": 0.0609, "step": 5090 }, { "grad_norm": 0.22869904339313507, "learning_rate": 9.611661768440093e-05, "loss": 0.0591, "step": 5100 }, { "grad_norm": 0.30646634101867676, "learning_rate": 9.609529311713661e-05, "loss": 0.0694, "step": 5110 }, { "grad_norm": 0.34552913904190063, "learning_rate": 9.607391253974466e-05, "loss": 0.0812, "step": 5120 }, { "grad_norm": 0.2576040029525757, "learning_rate": 9.605247597820448e-05, "loss": 0.0812, "step": 5130 }, { "grad_norm": 0.3567649722099304, "learning_rate": 9.603098345856354e-05, "loss": 0.0655, "step": 5140 }, { "grad_norm": 0.32950764894485474, "learning_rate": 9.600943500693725e-05, "loss": 0.0582, "step": 5150 }, { "grad_norm": 0.28899064660072327, "learning_rate": 9.598783064950902e-05, "loss": 0.0795, "step": 5160 }, { "grad_norm": 0.28894123435020447, "learning_rate": 9.596617041253018e-05, "loss": 0.0664, "step": 5170 }, { "grad_norm": 0.35477086901664734, "learning_rate": 9.594445432231996e-05, "loss": 0.0545, "step": 5180 }, { "grad_norm": 0.32432466745376587, "learning_rate": 9.592268240526547e-05, "loss": 0.0841, "step": 5190 }, { "grad_norm": 0.2890274226665497, "learning_rate": 9.590085468782162e-05, "loss": 0.0522, "step": 5200 }, { "grad_norm": 0.3483738601207733, "learning_rate": 9.587897119651116e-05, "loss": 0.0617, "step": 5210 }, { "grad_norm": 0.44481441378593445, "learning_rate": 9.585703195792459e-05, "loss": 0.0615, "step": 5220 }, { "grad_norm": 0.33214178681373596, "learning_rate": 9.583503699872016e-05, "loss": 0.0656, "step": 5230 }, { "grad_norm": 0.2405916452407837, "learning_rate": 9.581298634562381e-05, "loss": 0.0686, "step": 5240 }, { "grad_norm": 0.2829635441303253, "learning_rate": 9.579088002542917e-05, "loss": 0.0833, "step": 5250 }, { "grad_norm": 0.37046289443969727, "learning_rate": 9.57687180649975e-05, "loss": 0.0604, "step": 5260 }, { "grad_norm": 0.32492294907569885, "learning_rate": 9.574650049125768e-05, "loss": 0.0654, "step": 5270 }, { "grad_norm": 0.32311946153640747, "learning_rate": 9.572422733120614e-05, "loss": 0.0767, "step": 5280 }, { "grad_norm": 0.33394747972488403, "learning_rate": 9.570189861190689e-05, "loss": 0.0688, "step": 5290 }, { "grad_norm": 0.2944694459438324, "learning_rate": 9.56795143604914e-05, "loss": 0.0558, "step": 5300 }, { "grad_norm": 0.339233934879303, "learning_rate": 9.565707460415869e-05, "loss": 0.0547, "step": 5310 }, { "grad_norm": 0.3358080983161926, "learning_rate": 9.563457937017515e-05, "loss": 0.0629, "step": 5320 }, { "grad_norm": 0.37991535663604736, "learning_rate": 9.56120286858746e-05, "loss": 0.0774, "step": 5330 }, { "grad_norm": 0.32406139373779297, "learning_rate": 9.558942257865829e-05, "loss": 0.0673, "step": 5340 }, { "grad_norm": 0.280856192111969, "learning_rate": 9.556676107599472e-05, "loss": 0.0595, "step": 5350 }, { "grad_norm": 0.33333641290664673, "learning_rate": 9.554404420541978e-05, "loss": 0.0647, "step": 5360 }, { "grad_norm": 0.27470484375953674, "learning_rate": 9.55212719945366e-05, "loss": 0.0564, "step": 5370 }, { "grad_norm": 0.3289695382118225, "learning_rate": 9.549844447101559e-05, "loss": 0.0581, "step": 5380 }, { "grad_norm": 0.30457091331481934, "learning_rate": 9.547556166259433e-05, "loss": 0.0616, "step": 5390 }, { "grad_norm": 0.3400229513645172, "learning_rate": 9.545262359707756e-05, "loss": 0.0694, "step": 5400 }, { "grad_norm": 0.3404929041862488, "learning_rate": 9.542963030233724e-05, "loss": 0.058, "step": 5410 }, { "grad_norm": 0.29423561692237854, "learning_rate": 9.540658180631237e-05, "loss": 0.0613, "step": 5420 }, { "grad_norm": 0.3646376430988312, "learning_rate": 9.538347813700904e-05, "loss": 0.0548, "step": 5430 }, { "grad_norm": 0.3316885828971863, "learning_rate": 9.536031932250036e-05, "loss": 0.0792, "step": 5440 }, { "grad_norm": 0.26589637994766235, "learning_rate": 9.533710539092653e-05, "loss": 0.0444, "step": 5450 }, { "grad_norm": 0.2743433713912964, "learning_rate": 9.531383637049464e-05, "loss": 0.0583, "step": 5460 }, { "grad_norm": 0.2999128997325897, "learning_rate": 9.529051228947875e-05, "loss": 0.0561, "step": 5470 }, { "grad_norm": 0.31379467248916626, "learning_rate": 9.52671331762198e-05, "loss": 0.057, "step": 5480 }, { "grad_norm": 0.2803536653518677, "learning_rate": 9.524369905912565e-05, "loss": 0.0587, "step": 5490 }, { "grad_norm": 0.3245692849159241, "learning_rate": 9.522020996667092e-05, "loss": 0.0634, "step": 5500 }, { "grad_norm": 0.2900449335575104, "learning_rate": 9.519666592739709e-05, "loss": 0.0672, "step": 5510 }, { "grad_norm": 0.26660019159317017, "learning_rate": 9.517306696991241e-05, "loss": 0.059, "step": 5520 }, { "grad_norm": 0.28617027401924133, "learning_rate": 9.51494131228918e-05, "loss": 0.0597, "step": 5530 }, { "grad_norm": 0.29317325353622437, "learning_rate": 9.512570441507695e-05, "loss": 0.0547, "step": 5540 }, { "grad_norm": 0.2587851583957672, "learning_rate": 9.510194087527615e-05, "loss": 0.0579, "step": 5550 }, { "grad_norm": 0.33443427085876465, "learning_rate": 9.507812253236435e-05, "loss": 0.0589, "step": 5560 }, { "grad_norm": 0.2698667347431183, "learning_rate": 9.505424941528309e-05, "loss": 0.0594, "step": 5570 }, { "grad_norm": 0.3618665337562561, "learning_rate": 9.503032155304046e-05, "loss": 0.0804, "step": 5580 }, { "grad_norm": 0.26600879430770874, "learning_rate": 9.500633897471106e-05, "loss": 0.0565, "step": 5590 }, { "grad_norm": 0.3710070252418518, "learning_rate": 9.498230170943596e-05, "loss": 0.0693, "step": 5600 }, { "grad_norm": 0.3222614824771881, "learning_rate": 9.495820978642275e-05, "loss": 0.0577, "step": 5610 }, { "grad_norm": 0.274471253156662, "learning_rate": 9.493406323494535e-05, "loss": 0.052, "step": 5620 }, { "grad_norm": 0.31425178050994873, "learning_rate": 9.490986208434413e-05, "loss": 0.0539, "step": 5630 }, { "grad_norm": 0.3415668308734894, "learning_rate": 9.488560636402577e-05, "loss": 0.058, "step": 5640 }, { "grad_norm": 0.34010058641433716, "learning_rate": 9.486129610346321e-05, "loss": 0.0691, "step": 5650 }, { "grad_norm": 0.29943808913230896, "learning_rate": 9.483693133219576e-05, "loss": 0.057, "step": 5660 }, { "grad_norm": 0.30773842334747314, "learning_rate": 9.481251207982888e-05, "loss": 0.059, "step": 5670 }, { "grad_norm": 0.33717119693756104, "learning_rate": 9.47880383760343e-05, "loss": 0.0689, "step": 5680 }, { "grad_norm": 0.34518131613731384, "learning_rate": 9.476351025054983e-05, "loss": 0.0647, "step": 5690 }, { "grad_norm": 0.33139568567276, "learning_rate": 9.473892773317952e-05, "loss": 0.0611, "step": 5700 }, { "grad_norm": 0.24814656376838684, "learning_rate": 9.471429085379338e-05, "loss": 0.066, "step": 5710 }, { "grad_norm": 0.32591357827186584, "learning_rate": 9.468959964232757e-05, "loss": 0.0545, "step": 5720 }, { "grad_norm": 0.36171090602874756, "learning_rate": 9.466485412878425e-05, "loss": 0.0594, "step": 5730 }, { "grad_norm": 0.25330281257629395, "learning_rate": 9.464005434323154e-05, "loss": 0.0702, "step": 5740 }, { "grad_norm": 0.26813772320747375, "learning_rate": 9.461520031580352e-05, "loss": 0.0562, "step": 5750 }, { "grad_norm": 0.28434452414512634, "learning_rate": 9.459029207670019e-05, "loss": 0.0635, "step": 5760 }, { "grad_norm": 0.2507632374763489, "learning_rate": 9.456532965618737e-05, "loss": 0.0572, "step": 5770 }, { "grad_norm": 0.26459982991218567, "learning_rate": 9.454031308459681e-05, "loss": 0.0715, "step": 5780 }, { "grad_norm": 0.3322916328907013, "learning_rate": 9.451524239232595e-05, "loss": 0.0559, "step": 5790 }, { "grad_norm": 0.27088460326194763, "learning_rate": 9.449011760983809e-05, "loss": 0.0721, "step": 5800 }, { "grad_norm": 0.3380279839038849, "learning_rate": 9.446493876766218e-05, "loss": 0.0593, "step": 5810 }, { "grad_norm": 0.30934378504753113, "learning_rate": 9.44397058963929e-05, "loss": 0.06, "step": 5820 }, { "grad_norm": 0.3194241523742676, "learning_rate": 9.441441902669056e-05, "loss": 0.0681, "step": 5830 }, { "grad_norm": 0.21575912833213806, "learning_rate": 9.43890781892811e-05, "loss": 0.0592, "step": 5840 }, { "grad_norm": 0.24440687894821167, "learning_rate": 9.436368341495603e-05, "loss": 0.0478, "step": 5850 }, { "grad_norm": 0.2604001760482788, "learning_rate": 9.43382347345724e-05, "loss": 0.0707, "step": 5860 }, { "grad_norm": 0.3628251254558563, "learning_rate": 9.431273217905273e-05, "loss": 0.0653, "step": 5870 }, { "grad_norm": 0.27745911478996277, "learning_rate": 9.428717577938504e-05, "loss": 0.0684, "step": 5880 }, { "grad_norm": 0.30416756868362427, "learning_rate": 9.426156556662276e-05, "loss": 0.0643, "step": 5890 }, { "grad_norm": 0.27996647357940674, "learning_rate": 9.423590157188474e-05, "loss": 0.0817, "step": 5900 }, { "grad_norm": 0.3198047876358032, "learning_rate": 9.421018382635513e-05, "loss": 0.0716, "step": 5910 }, { "grad_norm": 0.36215877532958984, "learning_rate": 9.418441236128343e-05, "loss": 0.0621, "step": 5920 }, { "grad_norm": 0.29409343004226685, "learning_rate": 9.41585872079844e-05, "loss": 0.0549, "step": 5930 }, { "grad_norm": 0.32419055700302124, "learning_rate": 9.413270839783802e-05, "loss": 0.0632, "step": 5940 }, { "grad_norm": 0.32091644406318665, "learning_rate": 9.41067759622895e-05, "loss": 0.0567, "step": 5950 }, { "grad_norm": 0.2937156856060028, "learning_rate": 9.408078993284917e-05, "loss": 0.0535, "step": 5960 }, { "grad_norm": 0.3482649326324463, "learning_rate": 9.405475034109254e-05, "loss": 0.0638, "step": 5970 }, { "grad_norm": 0.2738930583000183, "learning_rate": 9.402865721866015e-05, "loss": 0.0635, "step": 5980 }, { "grad_norm": 0.3629935681819916, "learning_rate": 9.400251059725762e-05, "loss": 0.0763, "step": 5990 }, { "grad_norm": 0.4087563157081604, "learning_rate": 9.397631050865554e-05, "loss": 0.0866, "step": 6000 }, { "grad_norm": 0.30971622467041016, "learning_rate": 9.395005698468949e-05, "loss": 0.0671, "step": 6010 }, { "grad_norm": 0.24732452630996704, "learning_rate": 9.392375005726e-05, "loss": 0.0603, "step": 6020 }, { "grad_norm": 0.3049730658531189, "learning_rate": 9.389738975833242e-05, "loss": 0.0604, "step": 6030 }, { "grad_norm": 0.33987388014793396, "learning_rate": 9.387097611993707e-05, "loss": 0.0631, "step": 6040 }, { "grad_norm": 0.35611629486083984, "learning_rate": 9.384450917416894e-05, "loss": 0.0752, "step": 6050 }, { "grad_norm": 0.2588232755661011, "learning_rate": 9.381798895318792e-05, "loss": 0.0586, "step": 6060 }, { "grad_norm": 0.30615851283073425, "learning_rate": 9.379141548921853e-05, "loss": 0.0623, "step": 6070 }, { "grad_norm": 0.34001755714416504, "learning_rate": 9.376478881455009e-05, "loss": 0.0689, "step": 6080 }, { "grad_norm": 0.34220191836357117, "learning_rate": 9.373810896153647e-05, "loss": 0.0684, "step": 6090 }, { "grad_norm": 0.3081852197647095, "learning_rate": 9.371137596259623e-05, "loss": 0.0777, "step": 6100 }, { "grad_norm": 0.30905458331108093, "learning_rate": 9.368458985021248e-05, "loss": 0.0554, "step": 6110 }, { "grad_norm": 0.21939627826213837, "learning_rate": 9.365775065693287e-05, "loss": 0.0638, "step": 6120 }, { "grad_norm": 0.24481797218322754, "learning_rate": 9.363085841536957e-05, "loss": 0.0516, "step": 6130 }, { "grad_norm": 0.3698243498802185, "learning_rate": 9.360391315819917e-05, "loss": 0.0583, "step": 6140 }, { "grad_norm": 0.22680851817131042, "learning_rate": 9.357691491816269e-05, "loss": 0.0789, "step": 6150 }, { "grad_norm": 0.3375963866710663, "learning_rate": 9.354986372806557e-05, "loss": 0.0881, "step": 6160 }, { "grad_norm": 0.2559952437877655, "learning_rate": 9.352275962077752e-05, "loss": 0.0665, "step": 6170 }, { "grad_norm": 0.3269023895263672, "learning_rate": 9.349560262923262e-05, "loss": 0.0555, "step": 6180 }, { "grad_norm": 0.33286795020103455, "learning_rate": 9.346839278642913e-05, "loss": 0.0689, "step": 6190 }, { "grad_norm": 0.3095006048679352, "learning_rate": 9.344113012542963e-05, "loss": 0.0551, "step": 6200 }, { "grad_norm": 0.3369254171848297, "learning_rate": 9.341381467936079e-05, "loss": 0.0636, "step": 6210 }, { "grad_norm": 0.34415602684020996, "learning_rate": 9.338644648141346e-05, "loss": 0.0536, "step": 6220 }, { "grad_norm": 0.2738407850265503, "learning_rate": 9.335902556484257e-05, "loss": 0.0577, "step": 6230 }, { "grad_norm": 0.29314205050468445, "learning_rate": 9.333155196296712e-05, "loss": 0.0635, "step": 6240 }, { "grad_norm": 0.376844197511673, "learning_rate": 9.330402570917016e-05, "loss": 0.0725, "step": 6250 }, { "grad_norm": 0.26317062973976135, "learning_rate": 9.327644683689865e-05, "loss": 0.0529, "step": 6260 }, { "grad_norm": 0.28763848543167114, "learning_rate": 9.324881537966354e-05, "loss": 0.0572, "step": 6270 }, { "grad_norm": 0.3053903579711914, "learning_rate": 9.322113137103964e-05, "loss": 0.0556, "step": 6280 }, { "grad_norm": 0.27420687675476074, "learning_rate": 9.319339484466564e-05, "loss": 0.0553, "step": 6290 }, { "grad_norm": 0.25371405482292175, "learning_rate": 9.316560583424403e-05, "loss": 0.0564, "step": 6300 }, { "grad_norm": 0.283191055059433, "learning_rate": 9.313776437354108e-05, "loss": 0.0526, "step": 6310 }, { "grad_norm": 0.2978106141090393, "learning_rate": 9.31098704963868e-05, "loss": 0.0563, "step": 6320 }, { "grad_norm": 0.2851193845272064, "learning_rate": 9.308192423667485e-05, "loss": 0.0531, "step": 6330 }, { "grad_norm": 0.325709730386734, "learning_rate": 9.305392562836262e-05, "loss": 0.0649, "step": 6340 }, { "grad_norm": 0.3214518427848816, "learning_rate": 9.3025874705471e-05, "loss": 0.0573, "step": 6350 }, { "grad_norm": 0.3116633892059326, "learning_rate": 9.299777150208457e-05, "loss": 0.061, "step": 6360 }, { "grad_norm": 0.2861455976963043, "learning_rate": 9.296961605235133e-05, "loss": 0.0556, "step": 6370 }, { "grad_norm": 0.310829222202301, "learning_rate": 9.29414083904828e-05, "loss": 0.0549, "step": 6380 }, { "grad_norm": 0.32811516523361206, "learning_rate": 9.291314855075397e-05, "loss": 0.0646, "step": 6390 }, { "grad_norm": 0.2620982229709625, "learning_rate": 9.288483656750322e-05, "loss": 0.0615, "step": 6400 }, { "grad_norm": 0.24432693421840668, "learning_rate": 9.285647247513225e-05, "loss": 0.0429, "step": 6410 }, { "grad_norm": 0.23530341684818268, "learning_rate": 9.282805630810614e-05, "loss": 0.0583, "step": 6420 }, { "grad_norm": 0.2995782494544983, "learning_rate": 9.279958810095317e-05, "loss": 0.0565, "step": 6430 }, { "grad_norm": 0.2374049872159958, "learning_rate": 9.277106788826494e-05, "loss": 0.057, "step": 6440 }, { "grad_norm": 0.3054257035255432, "learning_rate": 9.274249570469618e-05, "loss": 0.052, "step": 6450 }, { "grad_norm": 0.27932992577552795, "learning_rate": 9.271387158496476e-05, "loss": 0.0492, "step": 6460 }, { "grad_norm": 0.2024971842765808, "learning_rate": 9.268519556385173e-05, "loss": 0.0491, "step": 6470 }, { "grad_norm": 0.24359631538391113, "learning_rate": 9.265646767620113e-05, "loss": 0.0488, "step": 6480 }, { "grad_norm": 0.32073676586151123, "learning_rate": 9.262768795692006e-05, "loss": 0.0668, "step": 6490 }, { "grad_norm": 0.31750747561454773, "learning_rate": 9.259885644097861e-05, "loss": 0.0471, "step": 6500 }, { "grad_norm": 0.25658541917800903, "learning_rate": 9.256997316340976e-05, "loss": 0.0449, "step": 6510 }, { "grad_norm": 0.21716344356536865, "learning_rate": 9.254103815930943e-05, "loss": 0.0555, "step": 6520 }, { "grad_norm": 0.31931352615356445, "learning_rate": 9.251205146383636e-05, "loss": 0.054, "step": 6530 }, { "grad_norm": 0.2929620146751404, "learning_rate": 9.248301311221216e-05, "loss": 0.0606, "step": 6540 }, { "grad_norm": 0.39821046590805054, "learning_rate": 9.245392313972115e-05, "loss": 0.0515, "step": 6550 }, { "grad_norm": 0.2783571183681488, "learning_rate": 9.242478158171037e-05, "loss": 0.0558, "step": 6560 }, { "grad_norm": 0.2597090005874634, "learning_rate": 9.239558847358957e-05, "loss": 0.0597, "step": 6570 }, { "grad_norm": 0.2624075710773468, "learning_rate": 9.236634385083114e-05, "loss": 0.045, "step": 6580 }, { "grad_norm": 0.2923348844051361, "learning_rate": 9.233704774897005e-05, "loss": 0.0739, "step": 6590 }, { "grad_norm": 0.31554779410362244, "learning_rate": 9.230770020360382e-05, "loss": 0.0711, "step": 6600 }, { "grad_norm": 0.31558459997177124, "learning_rate": 9.227830125039247e-05, "loss": 0.0505, "step": 6610 }, { "grad_norm": 0.2715090215206146, "learning_rate": 9.224885092505853e-05, "loss": 0.0522, "step": 6620 }, { "grad_norm": 0.2902199923992157, "learning_rate": 9.221934926338691e-05, "loss": 0.0475, "step": 6630 }, { "grad_norm": 0.2692619562149048, "learning_rate": 9.21897963012249e-05, "loss": 0.0469, "step": 6640 }, { "grad_norm": 0.29514452815055847, "learning_rate": 9.216019207448217e-05, "loss": 0.0633, "step": 6650 }, { "grad_norm": 0.3086773157119751, "learning_rate": 9.21305366191306e-05, "loss": 0.05, "step": 6660 }, { "grad_norm": 0.33134493231773376, "learning_rate": 9.21008299712044e-05, "loss": 0.0452, "step": 6670 }, { "grad_norm": 0.2896968722343445, "learning_rate": 9.207107216679993e-05, "loss": 0.0586, "step": 6680 }, { "grad_norm": 0.4009014666080475, "learning_rate": 9.204126324207574e-05, "loss": 0.0597, "step": 6690 }, { "grad_norm": 0.28062134981155396, "learning_rate": 9.201140323325248e-05, "loss": 0.0577, "step": 6700 }, { "grad_norm": 0.24070149660110474, "learning_rate": 9.198149217661287e-05, "loss": 0.0581, "step": 6710 }, { "grad_norm": 0.3009938895702362, "learning_rate": 9.195153010850167e-05, "loss": 0.066, "step": 6720 }, { "grad_norm": 0.262368381023407, "learning_rate": 9.192151706532562e-05, "loss": 0.0547, "step": 6730 }, { "grad_norm": 0.27898305654525757, "learning_rate": 9.189145308355337e-05, "loss": 0.0825, "step": 6740 }, { "grad_norm": 0.26537737250328064, "learning_rate": 9.186133819971555e-05, "loss": 0.0501, "step": 6750 }, { "grad_norm": 0.20543497800827026, "learning_rate": 9.183117245040455e-05, "loss": 0.0551, "step": 6760 }, { "grad_norm": 0.31179550290107727, "learning_rate": 9.180095587227458e-05, "loss": 0.0486, "step": 6770 }, { "grad_norm": 0.35982269048690796, "learning_rate": 9.177068850204167e-05, "loss": 0.0603, "step": 6780 }, { "grad_norm": 0.29707014560699463, "learning_rate": 9.174037037648351e-05, "loss": 0.0498, "step": 6790 }, { "grad_norm": 0.2911478281021118, "learning_rate": 9.171000153243947e-05, "loss": 0.0656, "step": 6800 }, { "grad_norm": 0.3358289301395416, "learning_rate": 9.167958200681057e-05, "loss": 0.0452, "step": 6810 }, { "grad_norm": 0.3521217405796051, "learning_rate": 9.164911183655943e-05, "loss": 0.0615, "step": 6820 }, { "grad_norm": 0.3227016031742096, "learning_rate": 9.161859105871013e-05, "loss": 0.0574, "step": 6830 }, { "grad_norm": 0.3175215721130371, "learning_rate": 9.158801971034832e-05, "loss": 0.0579, "step": 6840 }, { "grad_norm": 0.28898903727531433, "learning_rate": 9.155739782862107e-05, "loss": 0.0518, "step": 6850 }, { "grad_norm": 0.2803167402744293, "learning_rate": 9.152672545073687e-05, "loss": 0.0468, "step": 6860 }, { "grad_norm": 0.3059990704059601, "learning_rate": 9.149600261396551e-05, "loss": 0.0743, "step": 6870 }, { "grad_norm": 0.2578364312648773, "learning_rate": 9.146522935563817e-05, "loss": 0.0644, "step": 6880 }, { "grad_norm": 0.3416660726070404, "learning_rate": 9.143440571314723e-05, "loss": 0.0551, "step": 6890 }, { "grad_norm": 0.32495394349098206, "learning_rate": 9.140353172394636e-05, "loss": 0.0629, "step": 6900 }, { "grad_norm": 0.22883176803588867, "learning_rate": 9.137260742555032e-05, "loss": 0.0446, "step": 6910 }, { "grad_norm": 0.28669771552085876, "learning_rate": 9.13416328555351e-05, "loss": 0.0628, "step": 6920 }, { "grad_norm": 0.2978442311286926, "learning_rate": 9.131060805153769e-05, "loss": 0.0585, "step": 6930 }, { "grad_norm": 0.2976245880126953, "learning_rate": 9.127953305125618e-05, "loss": 0.0458, "step": 6940 }, { "grad_norm": 0.2992052435874939, "learning_rate": 9.124840789244958e-05, "loss": 0.0483, "step": 6950 }, { "grad_norm": 0.31257694959640503, "learning_rate": 9.121723261293794e-05, "loss": 0.0538, "step": 6960 }, { "grad_norm": 0.24986298382282257, "learning_rate": 9.118600725060214e-05, "loss": 0.0543, "step": 6970 }, { "grad_norm": 0.26425743103027344, "learning_rate": 9.115473184338392e-05, "loss": 0.0607, "step": 6980 }, { "grad_norm": 0.31712400913238525, "learning_rate": 9.11234064292859e-05, "loss": 0.0506, "step": 6990 }, { "grad_norm": 0.31658363342285156, "learning_rate": 9.109203104637137e-05, "loss": 0.0592, "step": 7000 }, { "grad_norm": 0.29683810472488403, "learning_rate": 9.10606057327644e-05, "loss": 0.0654, "step": 7010 }, { "grad_norm": 0.3693622052669525, "learning_rate": 9.10291305266497e-05, "loss": 0.0537, "step": 7020 }, { "grad_norm": 0.33074280619621277, "learning_rate": 9.099760546627261e-05, "loss": 0.0754, "step": 7030 }, { "grad_norm": 0.29221969842910767, "learning_rate": 9.096603058993907e-05, "loss": 0.0543, "step": 7040 }, { "grad_norm": 0.3026493787765503, "learning_rate": 9.093440593601552e-05, "loss": 0.0478, "step": 7050 }, { "grad_norm": 0.241587296128273, "learning_rate": 9.090273154292889e-05, "loss": 0.0509, "step": 7060 }, { "grad_norm": 0.24857619404792786, "learning_rate": 9.087100744916656e-05, "loss": 0.0564, "step": 7070 }, { "grad_norm": 0.34108248353004456, "learning_rate": 9.08392336932763e-05, "loss": 0.0714, "step": 7080 }, { "grad_norm": 0.3444642424583435, "learning_rate": 9.080741031386619e-05, "loss": 0.0587, "step": 7090 }, { "grad_norm": 0.28803908824920654, "learning_rate": 9.077553734960468e-05, "loss": 0.0515, "step": 7100 }, { "grad_norm": 0.28228259086608887, "learning_rate": 9.074361483922041e-05, "loss": 0.065, "step": 7110 }, { "grad_norm": 0.29784703254699707, "learning_rate": 9.071164282150224e-05, "loss": 0.0475, "step": 7120 }, { "grad_norm": 0.23805853724479675, "learning_rate": 9.067962133529919e-05, "loss": 0.0498, "step": 7130 }, { "grad_norm": 0.267630010843277, "learning_rate": 9.064755041952036e-05, "loss": 0.0695, "step": 7140 }, { "grad_norm": 0.322786808013916, "learning_rate": 9.061543011313497e-05, "loss": 0.064, "step": 7150 }, { "grad_norm": 0.40472379326820374, "learning_rate": 9.05832604551722e-05, "loss": 0.064, "step": 7160 }, { "grad_norm": 0.2969406843185425, "learning_rate": 9.055104148472124e-05, "loss": 0.066, "step": 7170 }, { "grad_norm": 0.30160200595855713, "learning_rate": 9.051877324093114e-05, "loss": 0.0539, "step": 7180 }, { "grad_norm": 0.23984181880950928, "learning_rate": 9.04864557630109e-05, "loss": 0.0632, "step": 7190 }, { "grad_norm": 0.277883917093277, "learning_rate": 9.045408909022928e-05, "loss": 0.0524, "step": 7200 }, { "grad_norm": 0.2720932960510254, "learning_rate": 9.042167326191484e-05, "loss": 0.0517, "step": 7210 }, { "grad_norm": 0.3136080205440521, "learning_rate": 9.038920831745587e-05, "loss": 0.0553, "step": 7220 }, { "grad_norm": 0.22797536849975586, "learning_rate": 9.035669429630036e-05, "loss": 0.0483, "step": 7230 }, { "grad_norm": 0.3152064383029938, "learning_rate": 9.032413123795588e-05, "loss": 0.05, "step": 7240 }, { "grad_norm": 0.24178220331668854, "learning_rate": 9.029151918198962e-05, "loss": 0.0536, "step": 7250 }, { "grad_norm": 0.24881818890571594, "learning_rate": 9.025885816802833e-05, "loss": 0.0471, "step": 7260 }, { "grad_norm": 0.3024463951587677, "learning_rate": 9.02261482357582e-05, "loss": 0.0457, "step": 7270 }, { "grad_norm": 0.25216028094291687, "learning_rate": 9.019338942492485e-05, "loss": 0.049, "step": 7280 }, { "grad_norm": 0.22370260953903198, "learning_rate": 9.01605817753334e-05, "loss": 0.0518, "step": 7290 }, { "grad_norm": 0.3325144350528717, "learning_rate": 9.012772532684818e-05, "loss": 0.0518, "step": 7300 }, { "grad_norm": 0.30344507098197937, "learning_rate": 9.00948201193929e-05, "loss": 0.0546, "step": 7310 }, { "grad_norm": 0.2602061629295349, "learning_rate": 9.006186619295047e-05, "loss": 0.0481, "step": 7320 }, { "grad_norm": 0.28807756304740906, "learning_rate": 9.002886358756305e-05, "loss": 0.0503, "step": 7330 }, { "grad_norm": 0.3184116780757904, "learning_rate": 8.999581234333188e-05, "loss": 0.0534, "step": 7340 }, { "grad_norm": 0.2343742400407791, "learning_rate": 8.996271250041734e-05, "loss": 0.0424, "step": 7350 }, { "grad_norm": 0.26885560154914856, "learning_rate": 8.99295640990389e-05, "loss": 0.0485, "step": 7360 }, { "grad_norm": 0.27294883131980896, "learning_rate": 8.989636717947494e-05, "loss": 0.0534, "step": 7370 }, { "grad_norm": 0.2733928859233856, "learning_rate": 8.986312178206291e-05, "loss": 0.0473, "step": 7380 }, { "grad_norm": 0.3421197235584259, "learning_rate": 8.982982794719904e-05, "loss": 0.0536, "step": 7390 }, { "grad_norm": 0.24818827211856842, "learning_rate": 8.979648571533852e-05, "loss": 0.0488, "step": 7400 }, { "grad_norm": 0.33971843123435974, "learning_rate": 8.976309512699528e-05, "loss": 0.0514, "step": 7410 }, { "grad_norm": 0.2864566147327423, "learning_rate": 8.972965622274205e-05, "loss": 0.0663, "step": 7420 }, { "grad_norm": 0.2659139633178711, "learning_rate": 8.969616904321026e-05, "loss": 0.0615, "step": 7430 }, { "grad_norm": 0.2871437966823578, "learning_rate": 8.966263362908998e-05, "loss": 0.0553, "step": 7440 }, { "grad_norm": 0.231905996799469, "learning_rate": 8.962905002112989e-05, "loss": 0.0505, "step": 7450 }, { "grad_norm": 0.2859322130680084, "learning_rate": 8.959541826013724e-05, "loss": 0.0568, "step": 7460 }, { "grad_norm": 0.27701035141944885, "learning_rate": 8.956173838697781e-05, "loss": 0.0513, "step": 7470 }, { "grad_norm": 0.34368452429771423, "learning_rate": 8.952801044257581e-05, "loss": 0.0606, "step": 7480 }, { "grad_norm": 0.2816283702850342, "learning_rate": 8.949423446791387e-05, "loss": 0.0554, "step": 7490 }, { "grad_norm": 0.24058087170124054, "learning_rate": 8.9460410504033e-05, "loss": 0.044, "step": 7500 }, { "grad_norm": 0.30176934599876404, "learning_rate": 8.942653859203248e-05, "loss": 0.0437, "step": 7510 }, { "grad_norm": 0.22638750076293945, "learning_rate": 8.93926187730699e-05, "loss": 0.045, "step": 7520 }, { "grad_norm": 0.23853576183319092, "learning_rate": 8.935865108836102e-05, "loss": 0.0411, "step": 7530 }, { "grad_norm": 0.21600733697414398, "learning_rate": 8.932463557917981e-05, "loss": 0.0505, "step": 7540 }, { "grad_norm": 0.35305488109588623, "learning_rate": 8.929057228685829e-05, "loss": 0.0519, "step": 7550 }, { "grad_norm": 0.2669561803340912, "learning_rate": 8.925646125278658e-05, "loss": 0.0508, "step": 7560 }, { "grad_norm": 0.22806251049041748, "learning_rate": 8.92223025184128e-05, "loss": 0.0609, "step": 7570 }, { "grad_norm": 0.3022954761981964, "learning_rate": 8.918809612524303e-05, "loss": 0.0653, "step": 7580 }, { "grad_norm": 0.24308346211910248, "learning_rate": 8.91538421148413e-05, "loss": 0.0577, "step": 7590 }, { "grad_norm": 0.27458855509757996, "learning_rate": 8.91195405288294e-05, "loss": 0.0705, "step": 7600 }, { "grad_norm": 0.34469375014305115, "learning_rate": 8.908519140888705e-05, "loss": 0.0562, "step": 7610 }, { "grad_norm": 0.28360602259635925, "learning_rate": 8.905079479675164e-05, "loss": 0.0518, "step": 7620 }, { "grad_norm": 0.25513869524002075, "learning_rate": 8.901635073421831e-05, "loss": 0.0535, "step": 7630 }, { "grad_norm": 0.2796258330345154, "learning_rate": 8.898185926313983e-05, "loss": 0.0519, "step": 7640 }, { "grad_norm": 0.25643911957740784, "learning_rate": 8.894732042542659e-05, "loss": 0.0492, "step": 7650 }, { "grad_norm": 0.250540554523468, "learning_rate": 8.891273426304654e-05, "loss": 0.0479, "step": 7660 }, { "grad_norm": 0.25919821858406067, "learning_rate": 8.887810081802515e-05, "loss": 0.0521, "step": 7670 }, { "grad_norm": 0.24039512872695923, "learning_rate": 8.884342013244529e-05, "loss": 0.0595, "step": 7680 }, { "grad_norm": 0.22423216700553894, "learning_rate": 8.880869224844725e-05, "loss": 0.0505, "step": 7690 }, { "grad_norm": 0.25921905040740967, "learning_rate": 8.877391720822875e-05, "loss": 0.0414, "step": 7700 }, { "grad_norm": 0.2976139783859253, "learning_rate": 8.873909505404468e-05, "loss": 0.0593, "step": 7710 }, { "grad_norm": 0.2532261908054352, "learning_rate": 8.870422582820726e-05, "loss": 0.0545, "step": 7720 }, { "grad_norm": 0.2918912172317505, "learning_rate": 8.866930957308588e-05, "loss": 0.0543, "step": 7730 }, { "grad_norm": 0.2974960505962372, "learning_rate": 8.86343463311071e-05, "loss": 0.053, "step": 7740 }, { "grad_norm": 0.2276851385831833, "learning_rate": 8.859933614475452e-05, "loss": 0.0445, "step": 7750 }, { "grad_norm": 0.3020394742488861, "learning_rate": 8.85642790565689e-05, "loss": 0.0553, "step": 7760 }, { "grad_norm": 0.30414915084838867, "learning_rate": 8.852917510914782e-05, "loss": 0.0605, "step": 7770 }, { "grad_norm": 0.2669791579246521, "learning_rate": 8.84940243451459e-05, "loss": 0.0563, "step": 7780 }, { "grad_norm": 0.3816332221031189, "learning_rate": 8.845882680727469e-05, "loss": 0.0514, "step": 7790 }, { "grad_norm": 0.3032526969909668, "learning_rate": 8.842358253830246e-05, "loss": 0.053, "step": 7800 }, { "grad_norm": 0.22727279365062714, "learning_rate": 8.838829158105434e-05, "loss": 0.0514, "step": 7810 }, { "grad_norm": 0.26629239320755005, "learning_rate": 8.835295397841217e-05, "loss": 0.041, "step": 7820 }, { "grad_norm": 0.27909791469573975, "learning_rate": 8.831756977331447e-05, "loss": 0.0445, "step": 7830 }, { "grad_norm": 0.26475268602371216, "learning_rate": 8.828213900875638e-05, "loss": 0.0433, "step": 7840 }, { "grad_norm": 0.2663150727748871, "learning_rate": 8.824666172778964e-05, "loss": 0.0465, "step": 7850 }, { "grad_norm": 0.2263398915529251, "learning_rate": 8.821113797352245e-05, "loss": 0.0514, "step": 7860 }, { "grad_norm": 0.24593374133110046, "learning_rate": 8.817556778911956e-05, "loss": 0.0612, "step": 7870 }, { "grad_norm": 0.28834420442581177, "learning_rate": 8.81399512178021e-05, "loss": 0.0496, "step": 7880 }, { "grad_norm": 0.2974545359611511, "learning_rate": 8.810428830284752e-05, "loss": 0.0469, "step": 7890 }, { "grad_norm": 0.2690908908843994, "learning_rate": 8.806857908758967e-05, "loss": 0.0608, "step": 7900 }, { "grad_norm": 0.2526589632034302, "learning_rate": 8.80328236154186e-05, "loss": 0.054, "step": 7910 }, { "grad_norm": 0.25868749618530273, "learning_rate": 8.799702192978056e-05, "loss": 0.072, "step": 7920 }, { "grad_norm": 0.26528292894363403, "learning_rate": 8.7961174074178e-05, "loss": 0.046, "step": 7930 }, { "grad_norm": 0.2738575041294098, "learning_rate": 8.792528009216942e-05, "loss": 0.0498, "step": 7940 }, { "grad_norm": 0.2932920455932617, "learning_rate": 8.788934002736943e-05, "loss": 0.05, "step": 7950 }, { "grad_norm": 0.48074695467948914, "learning_rate": 8.785335392344857e-05, "loss": 0.0519, "step": 7960 }, { "grad_norm": 0.24512407183647156, "learning_rate": 8.781732182413336e-05, "loss": 0.0476, "step": 7970 }, { "grad_norm": 0.2894071936607361, "learning_rate": 8.778124377320618e-05, "loss": 0.0525, "step": 7980 }, { "grad_norm": 0.24978435039520264, "learning_rate": 8.774511981450529e-05, "loss": 0.0443, "step": 7990 }, { "grad_norm": 0.29027390480041504, "learning_rate": 8.770894999192467e-05, "loss": 0.0433, "step": 8000 }, { "grad_norm": 0.23820069432258606, "learning_rate": 8.767273434941412e-05, "loss": 0.043, "step": 8010 }, { "grad_norm": 0.2417587786912918, "learning_rate": 8.7636472930979e-05, "loss": 0.0469, "step": 8020 }, { "grad_norm": 0.2664194703102112, "learning_rate": 8.76001657806804e-05, "loss": 0.0419, "step": 8030 }, { "grad_norm": 0.25502514839172363, "learning_rate": 8.756381294263491e-05, "loss": 0.0445, "step": 8040 }, { "grad_norm": 0.2552069127559662, "learning_rate": 8.752741446101464e-05, "loss": 0.0487, "step": 8050 }, { "grad_norm": 0.2212906926870346, "learning_rate": 8.749097038004722e-05, "loss": 0.0377, "step": 8060 }, { "grad_norm": 0.21560050547122955, "learning_rate": 8.745448074401561e-05, "loss": 0.0439, "step": 8070 }, { "grad_norm": 0.32789865136146545, "learning_rate": 8.741794559725818e-05, "loss": 0.0596, "step": 8080 }, { "grad_norm": 0.2761930525302887, "learning_rate": 8.738136498416856e-05, "loss": 0.0508, "step": 8090 }, { "grad_norm": 0.26075479388237, "learning_rate": 8.734473894919564e-05, "loss": 0.0534, "step": 8100 }, { "grad_norm": 0.2957226037979126, "learning_rate": 8.730806753684353e-05, "loss": 0.055, "step": 8110 }, { "grad_norm": 0.299996018409729, "learning_rate": 8.727135079167144e-05, "loss": 0.0464, "step": 8120 }, { "grad_norm": 0.3013080060482025, "learning_rate": 8.723458875829368e-05, "loss": 0.0613, "step": 8130 }, { "grad_norm": 0.4580949544906616, "learning_rate": 8.719778148137958e-05, "loss": 0.0522, "step": 8140 }, { "grad_norm": 0.20117861032485962, "learning_rate": 8.716092900565346e-05, "loss": 0.0461, "step": 8150 }, { "grad_norm": 0.35306739807128906, "learning_rate": 8.712403137589456e-05, "loss": 0.0591, "step": 8160 }, { "grad_norm": 0.3968307077884674, "learning_rate": 8.708708863693697e-05, "loss": 0.0547, "step": 8170 }, { "grad_norm": 0.2264263778924942, "learning_rate": 8.70501008336696e-05, "loss": 0.0465, "step": 8180 }, { "grad_norm": 0.252871036529541, "learning_rate": 8.701306801103611e-05, "loss": 0.0534, "step": 8190 }, { "grad_norm": 0.26898086071014404, "learning_rate": 8.697599021403489e-05, "loss": 0.0476, "step": 8200 }, { "grad_norm": 0.24235332012176514, "learning_rate": 8.693886748771896e-05, "loss": 0.0405, "step": 8210 }, { "grad_norm": 0.251179039478302, "learning_rate": 8.690169987719593e-05, "loss": 0.0469, "step": 8220 }, { "grad_norm": 0.20761802792549133, "learning_rate": 8.686448742762791e-05, "loss": 0.0482, "step": 8230 }, { "grad_norm": 0.2517324388027191, "learning_rate": 8.682723018423161e-05, "loss": 0.0487, "step": 8240 }, { "grad_norm": 0.28511324524879456, "learning_rate": 8.678992819227804e-05, "loss": 0.0472, "step": 8250 }, { "grad_norm": 0.19987966120243073, "learning_rate": 8.675258149709265e-05, "loss": 0.0411, "step": 8260 }, { "grad_norm": 0.3056575059890747, "learning_rate": 8.67151901440552e-05, "loss": 0.0413, "step": 8270 }, { "grad_norm": 0.2385588139295578, "learning_rate": 8.667775417859971e-05, "loss": 0.0534, "step": 8280 }, { "grad_norm": 0.2711584270000458, "learning_rate": 8.664027364621441e-05, "loss": 0.0369, "step": 8290 }, { "grad_norm": 0.25731542706489563, "learning_rate": 8.660274859244168e-05, "loss": 0.0553, "step": 8300 }, { "grad_norm": 0.2701854407787323, "learning_rate": 8.656517906287799e-05, "loss": 0.0481, "step": 8310 }, { "grad_norm": 0.27506452798843384, "learning_rate": 8.652756510317387e-05, "loss": 0.036, "step": 8320 }, { "grad_norm": 0.21722854673862457, "learning_rate": 8.648990675903382e-05, "loss": 0.05, "step": 8330 }, { "grad_norm": 0.21630197763442993, "learning_rate": 8.64522040762163e-05, "loss": 0.0386, "step": 8340 }, { "grad_norm": 0.22009441256523132, "learning_rate": 8.64144571005336e-05, "loss": 0.0515, "step": 8350 }, { "grad_norm": 0.2339830845594406, "learning_rate": 8.637666587785184e-05, "loss": 0.0431, "step": 8360 }, { "grad_norm": 0.23702920973300934, "learning_rate": 8.633883045409096e-05, "loss": 0.0429, "step": 8370 }, { "grad_norm": 0.27942943572998047, "learning_rate": 8.630095087522457e-05, "loss": 0.0523, "step": 8380 }, { "grad_norm": 0.1936587542295456, "learning_rate": 8.62630271872799e-05, "loss": 0.0386, "step": 8390 }, { "grad_norm": 0.2626267075538635, "learning_rate": 8.62250594363378e-05, "loss": 0.0419, "step": 8400 }, { "grad_norm": 0.24887345731258392, "learning_rate": 8.61870476685327e-05, "loss": 0.0511, "step": 8410 }, { "grad_norm": 0.2754501700401306, "learning_rate": 8.614899193005248e-05, "loss": 0.0414, "step": 8420 }, { "grad_norm": 0.2810872793197632, "learning_rate": 8.611089226713842e-05, "loss": 0.0446, "step": 8430 }, { "grad_norm": 0.23976877331733704, "learning_rate": 8.607274872608521e-05, "loss": 0.0466, "step": 8440 }, { "grad_norm": 0.29919353127479553, "learning_rate": 8.603456135324088e-05, "loss": 0.0462, "step": 8450 }, { "grad_norm": 0.29875996708869934, "learning_rate": 8.599633019500665e-05, "loss": 0.0445, "step": 8460 }, { "grad_norm": 0.267062783241272, "learning_rate": 8.595805529783702e-05, "loss": 0.0466, "step": 8470 }, { "grad_norm": 0.23112919926643372, "learning_rate": 8.591973670823958e-05, "loss": 0.0426, "step": 8480 }, { "grad_norm": 0.34670862555503845, "learning_rate": 8.588137447277503e-05, "loss": 0.0485, "step": 8490 }, { "grad_norm": 0.2950081527233124, "learning_rate": 8.584296863805709e-05, "loss": 0.0595, "step": 8500 }, { "grad_norm": 0.28798621892929077, "learning_rate": 8.580451925075248e-05, "loss": 0.0456, "step": 8510 }, { "grad_norm": 0.29371100664138794, "learning_rate": 8.576602635758085e-05, "loss": 0.0363, "step": 8520 }, { "grad_norm": 0.27705618739128113, "learning_rate": 8.572749000531469e-05, "loss": 0.0488, "step": 8530 }, { "grad_norm": 0.31781938672065735, "learning_rate": 8.568891024077926e-05, "loss": 0.0486, "step": 8540 }, { "grad_norm": 0.2287498116493225, "learning_rate": 8.565028711085265e-05, "loss": 0.0494, "step": 8550 }, { "grad_norm": 0.30090707540512085, "learning_rate": 8.561162066246561e-05, "loss": 0.0514, "step": 8560 }, { "grad_norm": 0.34750980138778687, "learning_rate": 8.557291094260152e-05, "loss": 0.0799, "step": 8570 }, { "grad_norm": 0.2207563817501068, "learning_rate": 8.553415799829629e-05, "loss": 0.0586, "step": 8580 }, { "grad_norm": 0.20844745635986328, "learning_rate": 8.549536187663847e-05, "loss": 0.0518, "step": 8590 }, { "grad_norm": 0.21608540415763855, "learning_rate": 8.545652262476897e-05, "loss": 0.0371, "step": 8600 }, { "grad_norm": 0.2636551260948181, "learning_rate": 8.541764028988115e-05, "loss": 0.0389, "step": 8610 }, { "grad_norm": 0.2458539456129074, "learning_rate": 8.537871491922071e-05, "loss": 0.0506, "step": 8620 }, { "grad_norm": 0.26190024614334106, "learning_rate": 8.533974656008566e-05, "loss": 0.04, "step": 8630 }, { "grad_norm": 0.28759729862213135, "learning_rate": 8.530073525982621e-05, "loss": 0.0485, "step": 8640 }, { "grad_norm": 0.2911844253540039, "learning_rate": 8.526168106584476e-05, "loss": 0.0516, "step": 8650 }, { "grad_norm": 0.24320611357688904, "learning_rate": 8.522258402559586e-05, "loss": 0.0416, "step": 8660 }, { "grad_norm": 0.27092230319976807, "learning_rate": 8.518344418658613e-05, "loss": 0.0405, "step": 8670 }, { "grad_norm": 0.27483871579170227, "learning_rate": 8.51442615963741e-05, "loss": 0.0318, "step": 8680 }, { "grad_norm": 0.22191360592842102, "learning_rate": 8.510503630257034e-05, "loss": 0.0443, "step": 8690 }, { "grad_norm": 0.2575194537639618, "learning_rate": 8.50657683528373e-05, "loss": 0.0429, "step": 8700 }, { "grad_norm": 0.24161295592784882, "learning_rate": 8.502645779488922e-05, "loss": 0.052, "step": 8710 }, { "grad_norm": 0.3259291350841522, "learning_rate": 8.498710467649213e-05, "loss": 0.0617, "step": 8720 }, { "grad_norm": 0.21492913365364075, "learning_rate": 8.494770904546381e-05, "loss": 0.0397, "step": 8730 }, { "grad_norm": 0.22493666410446167, "learning_rate": 8.490827094967363e-05, "loss": 0.0382, "step": 8740 }, { "grad_norm": 0.27042335271835327, "learning_rate": 8.486879043704263e-05, "loss": 0.0511, "step": 8750 }, { "grad_norm": 0.2624862790107727, "learning_rate": 8.482926755554332e-05, "loss": 0.0395, "step": 8760 }, { "grad_norm": 0.25154152512550354, "learning_rate": 8.478970235319975e-05, "loss": 0.0465, "step": 8770 }, { "grad_norm": 0.28452378511428833, "learning_rate": 8.475009487808737e-05, "loss": 0.0501, "step": 8780 }, { "grad_norm": 0.2794736623764038, "learning_rate": 8.471044517833298e-05, "loss": 0.0422, "step": 8790 }, { "grad_norm": 0.23355667293071747, "learning_rate": 8.467075330211474e-05, "loss": 0.034, "step": 8800 }, { "grad_norm": 0.2942545413970947, "learning_rate": 8.463101929766198e-05, "loss": 0.0463, "step": 8810 }, { "grad_norm": 0.2104250192642212, "learning_rate": 8.459124321325528e-05, "loss": 0.0416, "step": 8820 }, { "grad_norm": 0.2468242198228836, "learning_rate": 8.455142509722634e-05, "loss": 0.0438, "step": 8830 }, { "grad_norm": 0.2591128349304199, "learning_rate": 8.451156499795792e-05, "loss": 0.0433, "step": 8840 }, { "grad_norm": 0.2196366935968399, "learning_rate": 8.447166296388381e-05, "loss": 0.0375, "step": 8850 }, { "grad_norm": 0.2734740376472473, "learning_rate": 8.443171904348872e-05, "loss": 0.0366, "step": 8860 }, { "grad_norm": 0.31208929419517517, "learning_rate": 8.43917332853083e-05, "loss": 0.0455, "step": 8870 }, { "grad_norm": 0.25306111574172974, "learning_rate": 8.435170573792901e-05, "loss": 0.0445, "step": 8880 }, { "grad_norm": 0.27458199858665466, "learning_rate": 8.431163644998808e-05, "loss": 0.0452, "step": 8890 }, { "grad_norm": 0.2181319147348404, "learning_rate": 8.42715254701735e-05, "loss": 0.0443, "step": 8900 }, { "grad_norm": 0.22568073868751526, "learning_rate": 8.42313728472239e-05, "loss": 0.0424, "step": 8910 }, { "grad_norm": 0.2852737605571747, "learning_rate": 8.419117862992844e-05, "loss": 0.0593, "step": 8920 }, { "grad_norm": 0.29081934690475464, "learning_rate": 8.415094286712694e-05, "loss": 0.0435, "step": 8930 }, { "grad_norm": 0.2440153807401657, "learning_rate": 8.411066560770965e-05, "loss": 0.0474, "step": 8940 }, { "grad_norm": 0.19539517164230347, "learning_rate": 8.407034690061721e-05, "loss": 0.0396, "step": 8950 }, { "grad_norm": 0.2322612851858139, "learning_rate": 8.402998679484067e-05, "loss": 0.0361, "step": 8960 }, { "grad_norm": 0.21792368590831757, "learning_rate": 8.398958533942135e-05, "loss": 0.0341, "step": 8970 }, { "grad_norm": 0.28550052642822266, "learning_rate": 8.394914258345083e-05, "loss": 0.0464, "step": 8980 }, { "grad_norm": 0.2693115174770355, "learning_rate": 8.390865857607089e-05, "loss": 0.0705, "step": 8990 }, { "grad_norm": 0.3633045554161072, "learning_rate": 8.386813336647339e-05, "loss": 0.0493, "step": 9000 }, { "grad_norm": 0.22543634474277496, "learning_rate": 8.38275670039003e-05, "loss": 0.0364, "step": 9010 }, { "grad_norm": 0.26402097940444946, "learning_rate": 8.378695953764356e-05, "loss": 0.044, "step": 9020 }, { "grad_norm": 0.21990586817264557, "learning_rate": 8.374631101704509e-05, "loss": 0.0521, "step": 9030 }, { "grad_norm": 0.26469701528549194, "learning_rate": 8.370562149149665e-05, "loss": 0.0509, "step": 9040 }, { "grad_norm": 0.23545585572719574, "learning_rate": 8.366489101043989e-05, "loss": 0.0407, "step": 9050 }, { "grad_norm": 0.23479630053043365, "learning_rate": 8.362411962336614e-05, "loss": 0.0518, "step": 9060 }, { "grad_norm": 0.20438818633556366, "learning_rate": 8.358330737981651e-05, "loss": 0.0371, "step": 9070 }, { "grad_norm": 0.1889590620994568, "learning_rate": 8.35424543293817e-05, "loss": 0.0378, "step": 9080 }, { "grad_norm": 0.3126935362815857, "learning_rate": 8.350156052170206e-05, "loss": 0.0533, "step": 9090 }, { "grad_norm": 0.2519415616989136, "learning_rate": 8.346062600646738e-05, "loss": 0.0547, "step": 9100 }, { "grad_norm": 0.35271504521369934, "learning_rate": 8.341965083341697e-05, "loss": 0.0464, "step": 9110 }, { "grad_norm": 0.2967512309551239, "learning_rate": 8.337863505233953e-05, "loss": 0.0447, "step": 9120 }, { "grad_norm": 0.25248798727989197, "learning_rate": 8.33375787130731e-05, "loss": 0.0506, "step": 9130 }, { "grad_norm": 0.25679951906204224, "learning_rate": 8.3296481865505e-05, "loss": 0.036, "step": 9140 }, { "grad_norm": 0.2641292214393616, "learning_rate": 8.325534455957181e-05, "loss": 0.0532, "step": 9150 }, { "grad_norm": 0.22401762008666992, "learning_rate": 8.321416684525917e-05, "loss": 0.0331, "step": 9160 }, { "grad_norm": 0.2568923532962799, "learning_rate": 8.317294877260193e-05, "loss": 0.0429, "step": 9170 }, { "grad_norm": 0.2670583426952362, "learning_rate": 8.313169039168394e-05, "loss": 0.0431, "step": 9180 }, { "grad_norm": 0.2686108946800232, "learning_rate": 8.3090391752638e-05, "loss": 0.0341, "step": 9190 }, { "grad_norm": 0.28109800815582275, "learning_rate": 8.304905290564586e-05, "loss": 0.0457, "step": 9200 }, { "grad_norm": 0.21770186722278595, "learning_rate": 8.300767390093813e-05, "loss": 0.056, "step": 9210 }, { "grad_norm": 0.18378987908363342, "learning_rate": 8.296625478879417e-05, "loss": 0.0515, "step": 9220 }, { "grad_norm": 0.2301870882511139, "learning_rate": 8.292479561954214e-05, "loss": 0.0443, "step": 9230 }, { "grad_norm": 0.30590444803237915, "learning_rate": 8.288329644355883e-05, "loss": 0.0499, "step": 9240 }, { "grad_norm": 0.18814727663993835, "learning_rate": 8.284175731126965e-05, "loss": 0.0342, "step": 9250 }, { "grad_norm": 0.23684948682785034, "learning_rate": 8.280017827314854e-05, "loss": 0.0392, "step": 9260 }, { "grad_norm": 0.2521929442882538, "learning_rate": 8.2758559379718e-05, "loss": 0.0419, "step": 9270 }, { "grad_norm": 0.2673289477825165, "learning_rate": 8.271690068154887e-05, "loss": 0.048, "step": 9280 }, { "grad_norm": 0.33278775215148926, "learning_rate": 8.26752022292604e-05, "loss": 0.0534, "step": 9290 }, { "grad_norm": 0.24429400265216827, "learning_rate": 8.263346407352018e-05, "loss": 0.0508, "step": 9300 }, { "grad_norm": 0.38916319608688354, "learning_rate": 8.259168626504395e-05, "loss": 0.0432, "step": 9310 }, { "grad_norm": 0.22466930747032166, "learning_rate": 8.254986885459572e-05, "loss": 0.0595, "step": 9320 }, { "grad_norm": 0.19832603633403778, "learning_rate": 8.250801189298757e-05, "loss": 0.0414, "step": 9330 }, { "grad_norm": 0.26373961567878723, "learning_rate": 8.246611543107967e-05, "loss": 0.0483, "step": 9340 }, { "grad_norm": 0.20057500898838043, "learning_rate": 8.242417951978018e-05, "loss": 0.0412, "step": 9350 }, { "grad_norm": 0.2535959482192993, "learning_rate": 8.238220421004517e-05, "loss": 0.0378, "step": 9360 }, { "grad_norm": 0.19792145490646362, "learning_rate": 8.234018955287859e-05, "loss": 0.0521, "step": 9370 }, { "grad_norm": 0.16096755862236023, "learning_rate": 8.229813559933224e-05, "loss": 0.0405, "step": 9380 }, { "grad_norm": 0.21708257496356964, "learning_rate": 8.22560424005056e-05, "loss": 0.0407, "step": 9390 }, { "grad_norm": 0.25643229484558105, "learning_rate": 8.22139100075459e-05, "loss": 0.049, "step": 9400 }, { "grad_norm": 0.23652513325214386, "learning_rate": 8.217173847164799e-05, "loss": 0.0481, "step": 9410 }, { "grad_norm": 0.29153189063072205, "learning_rate": 8.212952784405422e-05, "loss": 0.0419, "step": 9420 }, { "grad_norm": 0.3846096992492676, "learning_rate": 8.208727817605451e-05, "loss": 0.0393, "step": 9430 }, { "grad_norm": 0.28406786918640137, "learning_rate": 8.204498951898618e-05, "loss": 0.0391, "step": 9440 }, { "grad_norm": 0.3028539717197418, "learning_rate": 8.200266192423395e-05, "loss": 0.0378, "step": 9450 }, { "grad_norm": 0.16957193613052368, "learning_rate": 8.196029544322981e-05, "loss": 0.0349, "step": 9460 }, { "grad_norm": 0.27391359210014343, "learning_rate": 8.191789012745305e-05, "loss": 0.0583, "step": 9470 }, { "grad_norm": 0.16656799614429474, "learning_rate": 8.187544602843014e-05, "loss": 0.0369, "step": 9480 }, { "grad_norm": 0.20624010264873505, "learning_rate": 8.183296319773466e-05, "loss": 0.0401, "step": 9490 }, { "grad_norm": 0.21962164342403412, "learning_rate": 8.179044168698721e-05, "loss": 0.0337, "step": 9500 }, { "grad_norm": 0.1763055920600891, "learning_rate": 8.174788154785548e-05, "loss": 0.034, "step": 9510 }, { "grad_norm": 0.24307899177074432, "learning_rate": 8.170528283205403e-05, "loss": 0.0444, "step": 9520 }, { "grad_norm": 0.2650991678237915, "learning_rate": 8.166264559134435e-05, "loss": 0.0432, "step": 9530 }, { "grad_norm": 0.3472405970096588, "learning_rate": 8.161996987753467e-05, "loss": 0.0484, "step": 9540 }, { "grad_norm": 0.16898582875728607, "learning_rate": 8.157725574248e-05, "loss": 0.0384, "step": 9550 }, { "grad_norm": 0.27579790353775024, "learning_rate": 8.153450323808209e-05, "loss": 0.0494, "step": 9560 }, { "grad_norm": 0.2798709571361542, "learning_rate": 8.149171241628923e-05, "loss": 0.0588, "step": 9570 }, { "grad_norm": 0.300478458404541, "learning_rate": 8.144888332909631e-05, "loss": 0.047, "step": 9580 }, { "grad_norm": 0.21779417991638184, "learning_rate": 8.14060160285447e-05, "loss": 0.0375, "step": 9590 }, { "grad_norm": 0.33049634099006653, "learning_rate": 8.136311056672223e-05, "loss": 0.0363, "step": 9600 }, { "grad_norm": 0.2596324682235718, "learning_rate": 8.132016699576308e-05, "loss": 0.0384, "step": 9610 }, { "grad_norm": 0.2808871865272522, "learning_rate": 8.127718536784771e-05, "loss": 0.0431, "step": 9620 }, { "grad_norm": 0.2244134396314621, "learning_rate": 8.12341657352029e-05, "loss": 0.0351, "step": 9630 }, { "grad_norm": 0.3169289827346802, "learning_rate": 8.119110815010151e-05, "loss": 0.0354, "step": 9640 }, { "grad_norm": 0.39019590616226196, "learning_rate": 8.114801266486261e-05, "loss": 0.039, "step": 9650 }, { "grad_norm": 0.23711971938610077, "learning_rate": 8.110487933185122e-05, "loss": 0.0406, "step": 9660 }, { "grad_norm": 0.2399277687072754, "learning_rate": 8.106170820347847e-05, "loss": 0.049, "step": 9670 }, { "grad_norm": 0.21163716912269592, "learning_rate": 8.101849933220132e-05, "loss": 0.0348, "step": 9680 }, { "grad_norm": 0.15759004652500153, "learning_rate": 8.097525277052264e-05, "loss": 0.0414, "step": 9690 }, { "grad_norm": 0.24726495146751404, "learning_rate": 8.093196857099106e-05, "loss": 0.036, "step": 9700 }, { "grad_norm": 0.19588416814804077, "learning_rate": 8.088864678620096e-05, "loss": 0.0363, "step": 9710 }, { "grad_norm": 0.2239535003900528, "learning_rate": 8.084528746879243e-05, "loss": 0.0472, "step": 9720 }, { "grad_norm": 0.2357938140630722, "learning_rate": 8.080189067145108e-05, "loss": 0.0539, "step": 9730 }, { "grad_norm": 0.2593938112258911, "learning_rate": 8.075845644690813e-05, "loss": 0.0384, "step": 9740 }, { "grad_norm": 0.21965909004211426, "learning_rate": 8.07149848479403e-05, "loss": 0.0374, "step": 9750 }, { "grad_norm": 0.19287362694740295, "learning_rate": 8.067147592736962e-05, "loss": 0.0439, "step": 9760 }, { "grad_norm": 0.21761581301689148, "learning_rate": 8.062792973806357e-05, "loss": 0.0382, "step": 9770 }, { "grad_norm": 0.3522203266620636, "learning_rate": 8.058434633293485e-05, "loss": 0.0418, "step": 9780 }, { "grad_norm": 0.20352689921855927, "learning_rate": 8.054072576494143e-05, "loss": 0.0423, "step": 9790 }, { "grad_norm": 0.32283347845077515, "learning_rate": 8.04970680870864e-05, "loss": 0.034, "step": 9800 }, { "grad_norm": 0.2942480146884918, "learning_rate": 8.045337335241793e-05, "loss": 0.0386, "step": 9810 }, { "grad_norm": 0.31796011328697205, "learning_rate": 8.040964161402932e-05, "loss": 0.0467, "step": 9820 }, { "grad_norm": 0.2257106602191925, "learning_rate": 8.036587292505869e-05, "loss": 0.043, "step": 9830 }, { "grad_norm": 0.20455415546894073, "learning_rate": 8.032206733868912e-05, "loss": 0.0504, "step": 9840 }, { "grad_norm": 0.22516673803329468, "learning_rate": 8.027822490814858e-05, "loss": 0.0438, "step": 9850 }, { "grad_norm": 0.2701977491378784, "learning_rate": 8.023434568670972e-05, "loss": 0.0348, "step": 9860 }, { "grad_norm": 0.19640228152275085, "learning_rate": 8.019042972768992e-05, "loss": 0.0361, "step": 9870 }, { "grad_norm": 0.24725179374217987, "learning_rate": 8.014647708445124e-05, "loss": 0.0482, "step": 9880 }, { "grad_norm": 0.24354586005210876, "learning_rate": 8.010248781040027e-05, "loss": 0.0315, "step": 9890 }, { "grad_norm": 0.1855171024799347, "learning_rate": 8.005846195898814e-05, "loss": 0.043, "step": 9900 }, { "grad_norm": 0.2570749819278717, "learning_rate": 8.00143995837104e-05, "loss": 0.0417, "step": 9910 }, { "grad_norm": 0.2468637377023697, "learning_rate": 7.997030073810698e-05, "loss": 0.0476, "step": 9920 }, { "grad_norm": 0.22067344188690186, "learning_rate": 7.992616547576217e-05, "loss": 0.0319, "step": 9930 }, { "grad_norm": 0.19970707595348358, "learning_rate": 7.988199385030445e-05, "loss": 0.0373, "step": 9940 }, { "grad_norm": 0.2628572881221771, "learning_rate": 7.98377859154065e-05, "loss": 0.0316, "step": 9950 }, { "grad_norm": 0.208957701921463, "learning_rate": 7.979354172478516e-05, "loss": 0.0337, "step": 9960 }, { "grad_norm": 0.2231970876455307, "learning_rate": 7.974926133220127e-05, "loss": 0.0439, "step": 9970 }, { "grad_norm": 0.2116335928440094, "learning_rate": 7.970494479145967e-05, "loss": 0.04, "step": 9980 }, { "grad_norm": 0.2541116178035736, "learning_rate": 7.966059215640919e-05, "loss": 0.0417, "step": 9990 }, { "grad_norm": 0.2910885810852051, "learning_rate": 7.961620348094241e-05, "loss": 0.0544, "step": 10000 }, { "grad_norm": 0.21513716876506805, "learning_rate": 7.957177881899577e-05, "loss": 0.0447, "step": 10010 }, { "grad_norm": 0.21461637318134308, "learning_rate": 7.952731822454944e-05, "loss": 0.0459, "step": 10020 }, { "grad_norm": 0.23180069029331207, "learning_rate": 7.948282175162722e-05, "loss": 0.0367, "step": 10030 }, { "grad_norm": 0.27447235584259033, "learning_rate": 7.943828945429652e-05, "loss": 0.0509, "step": 10040 }, { "grad_norm": 0.2721457779407501, "learning_rate": 7.939372138666827e-05, "loss": 0.0417, "step": 10050 }, { "grad_norm": 0.24187959730625153, "learning_rate": 7.934911760289692e-05, "loss": 0.0502, "step": 10060 }, { "grad_norm": 0.24366912245750427, "learning_rate": 7.930447815718022e-05, "loss": 0.0319, "step": 10070 }, { "grad_norm": 0.23423758149147034, "learning_rate": 7.925980310375934e-05, "loss": 0.0432, "step": 10080 }, { "grad_norm": 0.22259816527366638, "learning_rate": 7.921509249691865e-05, "loss": 0.0382, "step": 10090 }, { "grad_norm": 0.27035683393478394, "learning_rate": 7.917034639098579e-05, "loss": 0.0414, "step": 10100 }, { "grad_norm": 0.14527547359466553, "learning_rate": 7.912556484033146e-05, "loss": 0.0309, "step": 10110 }, { "grad_norm": 0.25964194536209106, "learning_rate": 7.908074789936952e-05, "loss": 0.0343, "step": 10120 }, { "grad_norm": 0.22951532900333405, "learning_rate": 7.903589562255673e-05, "loss": 0.0487, "step": 10130 }, { "grad_norm": 0.283926397562027, "learning_rate": 7.899100806439285e-05, "loss": 0.0356, "step": 10140 }, { "grad_norm": 0.22075264155864716, "learning_rate": 7.894608527942049e-05, "loss": 0.0463, "step": 10150 }, { "grad_norm": 0.23929134011268616, "learning_rate": 7.89011273222251e-05, "loss": 0.0457, "step": 10160 }, { "grad_norm": 0.22705943882465363, "learning_rate": 7.885613424743482e-05, "loss": 0.0393, "step": 10170 }, { "grad_norm": 0.20997200906276703, "learning_rate": 7.881110610972044e-05, "loss": 0.0491, "step": 10180 }, { "grad_norm": 0.2516878843307495, "learning_rate": 7.876604296379544e-05, "loss": 0.0521, "step": 10190 }, { "grad_norm": 0.17547576129436493, "learning_rate": 7.872094486441579e-05, "loss": 0.0469, "step": 10200 }, { "grad_norm": 0.19798053801059723, "learning_rate": 7.867581186637991e-05, "loss": 0.045, "step": 10210 }, { "grad_norm": 0.18334461748600006, "learning_rate": 7.863064402452866e-05, "loss": 0.0333, "step": 10220 }, { "grad_norm": 0.22966639697551727, "learning_rate": 7.858544139374524e-05, "loss": 0.0425, "step": 10230 }, { "grad_norm": 0.19323423504829407, "learning_rate": 7.854020402895508e-05, "loss": 0.0418, "step": 10240 }, { "grad_norm": 0.19293759763240814, "learning_rate": 7.849493198512587e-05, "loss": 0.0391, "step": 10250 }, { "grad_norm": 0.23573896288871765, "learning_rate": 7.84496253172674e-05, "loss": 0.0379, "step": 10260 }, { "grad_norm": 0.23406943678855896, "learning_rate": 7.840428408043155e-05, "loss": 0.0446, "step": 10270 }, { "grad_norm": 0.25668907165527344, "learning_rate": 7.835890832971218e-05, "loss": 0.0459, "step": 10280 }, { "grad_norm": 0.24066990613937378, "learning_rate": 7.831349812024512e-05, "loss": 0.0472, "step": 10290 }, { "grad_norm": 0.2807666063308716, "learning_rate": 7.826805350720807e-05, "loss": 0.0406, "step": 10300 }, { "grad_norm": 0.2792661488056183, "learning_rate": 7.822257454582049e-05, "loss": 0.0379, "step": 10310 }, { "grad_norm": 0.2074809968471527, "learning_rate": 7.817706129134363e-05, "loss": 0.0454, "step": 10320 }, { "grad_norm": 0.2765476107597351, "learning_rate": 7.813151379908036e-05, "loss": 0.04, "step": 10330 }, { "grad_norm": 0.21375752985477448, "learning_rate": 7.808593212437523e-05, "loss": 0.0444, "step": 10340 }, { "grad_norm": 0.26423248648643494, "learning_rate": 7.804031632261421e-05, "loss": 0.0515, "step": 10350 }, { "grad_norm": 0.2407953441143036, "learning_rate": 7.799466644922484e-05, "loss": 0.0413, "step": 10360 }, { "grad_norm": 0.21558886766433716, "learning_rate": 7.794898255967602e-05, "loss": 0.0347, "step": 10370 }, { "grad_norm": 0.20214705169200897, "learning_rate": 7.790326470947795e-05, "loss": 0.0419, "step": 10380 }, { "grad_norm": 0.2293347418308258, "learning_rate": 7.785751295418217e-05, "loss": 0.0381, "step": 10390 }, { "grad_norm": 0.2281937152147293, "learning_rate": 7.781172734938136e-05, "loss": 0.0423, "step": 10400 }, { "grad_norm": 0.24225731194019318, "learning_rate": 7.776590795070933e-05, "loss": 0.0432, "step": 10410 }, { "grad_norm": 0.2024509608745575, "learning_rate": 7.772005481384099e-05, "loss": 0.0458, "step": 10420 }, { "grad_norm": 0.2956198751926422, "learning_rate": 7.767416799449223e-05, "loss": 0.0324, "step": 10430 }, { "grad_norm": 0.2598063349723816, "learning_rate": 7.762824754841984e-05, "loss": 0.0355, "step": 10440 }, { "grad_norm": 0.25561872124671936, "learning_rate": 7.758229353142152e-05, "loss": 0.047, "step": 10450 }, { "grad_norm": 0.22605006396770477, "learning_rate": 7.753630599933572e-05, "loss": 0.0395, "step": 10460 }, { "grad_norm": 0.2262575924396515, "learning_rate": 7.749028500804161e-05, "loss": 0.0348, "step": 10470 }, { "grad_norm": 0.2061920017004013, "learning_rate": 7.744423061345906e-05, "loss": 0.0343, "step": 10480 }, { "grad_norm": 0.18750597536563873, "learning_rate": 7.73981428715485e-05, "loss": 0.0334, "step": 10490 }, { "grad_norm": 0.23163531720638275, "learning_rate": 7.735202183831085e-05, "loss": 0.0341, "step": 10500 }, { "grad_norm": 0.20083948969841003, "learning_rate": 7.730586756978757e-05, "loss": 0.0471, "step": 10510 }, { "grad_norm": 0.22803887724876404, "learning_rate": 7.72596801220604e-05, "loss": 0.0319, "step": 10520 }, { "grad_norm": 0.15561164915561676, "learning_rate": 7.721345955125147e-05, "loss": 0.0437, "step": 10530 }, { "grad_norm": 0.22287897765636444, "learning_rate": 7.71672059135231e-05, "loss": 0.0326, "step": 10540 }, { "grad_norm": 0.1741456836462021, "learning_rate": 7.712091926507787e-05, "loss": 0.0283, "step": 10550 }, { "grad_norm": 0.21366147696971893, "learning_rate": 7.70745996621584e-05, "loss": 0.0322, "step": 10560 }, { "grad_norm": 0.22503922879695892, "learning_rate": 7.702824716104735e-05, "loss": 0.0304, "step": 10570 }, { "grad_norm": 0.20203983783721924, "learning_rate": 7.698186181806743e-05, "loss": 0.0423, "step": 10580 }, { "grad_norm": 0.21008794009685516, "learning_rate": 7.693544368958116e-05, "loss": 0.0386, "step": 10590 }, { "grad_norm": 0.20781303942203522, "learning_rate": 7.688899283199096e-05, "loss": 0.0514, "step": 10600 }, { "grad_norm": 0.1906471699476242, "learning_rate": 7.684250930173901e-05, "loss": 0.0621, "step": 10610 }, { "grad_norm": 0.28375154733657837, "learning_rate": 7.679599315530718e-05, "loss": 0.0581, "step": 10620 }, { "grad_norm": 0.15197792649269104, "learning_rate": 7.674944444921695e-05, "loss": 0.0326, "step": 10630 }, { "grad_norm": 0.17161054909229279, "learning_rate": 7.670286324002944e-05, "loss": 0.0403, "step": 10640 }, { "grad_norm": 0.14621169865131378, "learning_rate": 7.665624958434514e-05, "loss": 0.0477, "step": 10650 }, { "grad_norm": 0.2014542669057846, "learning_rate": 7.66096035388041e-05, "loss": 0.0445, "step": 10660 }, { "grad_norm": 0.24865767359733582, "learning_rate": 7.656292516008563e-05, "loss": 0.044, "step": 10670 }, { "grad_norm": 0.18066716194152832, "learning_rate": 7.651621450490837e-05, "loss": 0.0304, "step": 10680 }, { "grad_norm": 0.2286992073059082, "learning_rate": 7.646947163003017e-05, "loss": 0.031, "step": 10690 }, { "grad_norm": 0.23579055070877075, "learning_rate": 7.642269659224803e-05, "loss": 0.0603, "step": 10700 }, { "grad_norm": 0.22920671105384827, "learning_rate": 7.637588944839803e-05, "loss": 0.0491, "step": 10710 }, { "grad_norm": 0.14447563886642456, "learning_rate": 7.632905025535528e-05, "loss": 0.0359, "step": 10720 }, { "grad_norm": 0.21368440985679626, "learning_rate": 7.628217907003378e-05, "loss": 0.0361, "step": 10730 }, { "grad_norm": 0.19534438848495483, "learning_rate": 7.623527594938649e-05, "loss": 0.0316, "step": 10740 }, { "grad_norm": 0.22025732696056366, "learning_rate": 7.618834095040509e-05, "loss": 0.0399, "step": 10750 }, { "grad_norm": 0.3441934287548065, "learning_rate": 7.614137413012e-05, "loss": 0.0446, "step": 10760 }, { "grad_norm": 0.28608062863349915, "learning_rate": 7.609437554560042e-05, "loss": 0.0434, "step": 10770 }, { "grad_norm": 0.22394655644893646, "learning_rate": 7.604734525395398e-05, "loss": 0.0453, "step": 10780 }, { "grad_norm": 0.20351824164390564, "learning_rate": 7.600028331232699e-05, "loss": 0.0358, "step": 10790 }, { "grad_norm": 0.2478836625814438, "learning_rate": 7.595318977790408e-05, "loss": 0.0468, "step": 10800 }, { "grad_norm": 0.23694613575935364, "learning_rate": 7.590606470790836e-05, "loss": 0.0333, "step": 10810 }, { "grad_norm": 0.24223344027996063, "learning_rate": 7.585890815960125e-05, "loss": 0.0404, "step": 10820 }, { "grad_norm": 0.1631132960319519, "learning_rate": 7.581172019028238e-05, "loss": 0.0392, "step": 10830 }, { "grad_norm": 0.22024089097976685, "learning_rate": 7.576450085728959e-05, "loss": 0.0479, "step": 10840 }, { "grad_norm": 0.2285163402557373, "learning_rate": 7.571725021799885e-05, "loss": 0.0389, "step": 10850 }, { "grad_norm": 0.26019567251205444, "learning_rate": 7.566996832982409e-05, "loss": 0.0301, "step": 10860 }, { "grad_norm": 0.18268737196922302, "learning_rate": 7.56226552502173e-05, "loss": 0.0362, "step": 10870 }, { "grad_norm": 0.2807478904724121, "learning_rate": 7.557531103666833e-05, "loss": 0.0405, "step": 10880 }, { "grad_norm": 0.17552682757377625, "learning_rate": 7.552793574670485e-05, "loss": 0.0384, "step": 10890 }, { "grad_norm": 0.2608664631843567, "learning_rate": 7.54805294378923e-05, "loss": 0.0449, "step": 10900 }, { "grad_norm": 0.22736406326293945, "learning_rate": 7.543309216783384e-05, "loss": 0.0356, "step": 10910 }, { "grad_norm": 0.256570965051651, "learning_rate": 7.53856239941702e-05, "loss": 0.0383, "step": 10920 }, { "grad_norm": 0.23745015263557434, "learning_rate": 7.533812497457972e-05, "loss": 0.0482, "step": 10930 }, { "grad_norm": 0.24123995006084442, "learning_rate": 7.529059516677814e-05, "loss": 0.0448, "step": 10940 }, { "grad_norm": 0.18594416975975037, "learning_rate": 7.524303462851872e-05, "loss": 0.0323, "step": 10950 }, { "grad_norm": 0.21350662410259247, "learning_rate": 7.519544341759192e-05, "loss": 0.03, "step": 10960 }, { "grad_norm": 0.1693103015422821, "learning_rate": 7.514782159182562e-05, "loss": 0.0321, "step": 10970 }, { "grad_norm": 0.18776792287826538, "learning_rate": 7.510016920908481e-05, "loss": 0.033, "step": 10980 }, { "grad_norm": 0.2281103879213333, "learning_rate": 7.505248632727158e-05, "loss": 0.0325, "step": 10990 }, { "grad_norm": 0.21889925003051758, "learning_rate": 7.50047730043252e-05, "loss": 0.0362, "step": 11000 }, { "grad_norm": 0.25138047337532043, "learning_rate": 7.495702929822183e-05, "loss": 0.032, "step": 11010 }, { "grad_norm": 0.22701409459114075, "learning_rate": 7.490925526697455e-05, "loss": 0.0368, "step": 11020 }, { "grad_norm": 0.298861026763916, "learning_rate": 7.486145096863334e-05, "loss": 0.038, "step": 11030 }, { "grad_norm": 0.2517353892326355, "learning_rate": 7.481361646128491e-05, "loss": 0.0352, "step": 11040 }, { "grad_norm": 0.18295548856258392, "learning_rate": 7.476575180305271e-05, "loss": 0.0308, "step": 11050 }, { "grad_norm": 0.1780187040567398, "learning_rate": 7.471785705209682e-05, "loss": 0.046, "step": 11060 }, { "grad_norm": 0.2478165477514267, "learning_rate": 7.466993226661387e-05, "loss": 0.0304, "step": 11070 }, { "grad_norm": 0.22152885794639587, "learning_rate": 7.462197750483699e-05, "loss": 0.0385, "step": 11080 }, { "grad_norm": 0.327821284532547, "learning_rate": 7.457399282503574e-05, "loss": 0.0382, "step": 11090 }, { "grad_norm": 0.2564924359321594, "learning_rate": 7.452597828551604e-05, "loss": 0.0447, "step": 11100 }, { "grad_norm": 0.25822699069976807, "learning_rate": 7.447793394462006e-05, "loss": 0.0418, "step": 11110 }, { "grad_norm": 0.18866516649723053, "learning_rate": 7.442985986072624e-05, "loss": 0.0374, "step": 11120 }, { "grad_norm": 0.17779546976089478, "learning_rate": 7.438175609224908e-05, "loss": 0.0395, "step": 11130 }, { "grad_norm": 0.2204521894454956, "learning_rate": 7.433362269763924e-05, "loss": 0.0338, "step": 11140 }, { "grad_norm": 0.160470649600029, "learning_rate": 7.428545973538329e-05, "loss": 0.0246, "step": 11150 }, { "grad_norm": 0.2592422366142273, "learning_rate": 7.42372672640038e-05, "loss": 0.0381, "step": 11160 }, { "grad_norm": 0.17459969222545624, "learning_rate": 7.418904534205917e-05, "loss": 0.0339, "step": 11170 }, { "grad_norm": 0.25367820262908936, "learning_rate": 7.414079402814356e-05, "loss": 0.0377, "step": 11180 }, { "grad_norm": 0.266734778881073, "learning_rate": 7.409251338088687e-05, "loss": 0.0435, "step": 11190 }, { "grad_norm": 0.1828782707452774, "learning_rate": 7.404420345895467e-05, "loss": 0.0404, "step": 11200 }, { "grad_norm": 0.25150710344314575, "learning_rate": 7.399586432104804e-05, "loss": 0.0337, "step": 11210 }, { "grad_norm": 0.2177266925573349, "learning_rate": 7.394749602590358e-05, "loss": 0.0355, "step": 11220 }, { "grad_norm": 0.2668229043483734, "learning_rate": 7.389909863229336e-05, "loss": 0.0483, "step": 11230 }, { "grad_norm": 0.15293827652931213, "learning_rate": 7.385067219902477e-05, "loss": 0.0353, "step": 11240 }, { "grad_norm": 0.2464955598115921, "learning_rate": 7.380221678494049e-05, "loss": 0.027, "step": 11250 }, { "grad_norm": 0.24195197224617004, "learning_rate": 7.37537324489184e-05, "loss": 0.0287, "step": 11260 }, { "grad_norm": 0.2145175337791443, "learning_rate": 7.370521924987155e-05, "loss": 0.0311, "step": 11270 }, { "grad_norm": 0.2280535101890564, "learning_rate": 7.365667724674806e-05, "loss": 0.0468, "step": 11280 }, { "grad_norm": 0.2039894163608551, "learning_rate": 7.360810649853105e-05, "loss": 0.0387, "step": 11290 }, { "grad_norm": 0.2793155610561371, "learning_rate": 7.355950706423853e-05, "loss": 0.036, "step": 11300 }, { "grad_norm": 0.22918646037578583, "learning_rate": 7.351087900292342e-05, "loss": 0.0326, "step": 11310 }, { "grad_norm": 0.18763580918312073, "learning_rate": 7.346222237367339e-05, "loss": 0.0359, "step": 11320 }, { "grad_norm": 0.20314472913742065, "learning_rate": 7.341353723561081e-05, "loss": 0.0371, "step": 11330 }, { "grad_norm": 0.22601304948329926, "learning_rate": 7.336482364789277e-05, "loss": 0.0297, "step": 11340 }, { "grad_norm": 0.1856689751148224, "learning_rate": 7.331608166971082e-05, "loss": 0.052, "step": 11350 }, { "grad_norm": 0.24075965583324432, "learning_rate": 7.326731136029108e-05, "loss": 0.0324, "step": 11360 }, { "grad_norm": 0.15768487751483917, "learning_rate": 7.321851277889408e-05, "loss": 0.0287, "step": 11370 }, { "grad_norm": 0.15501593053340912, "learning_rate": 7.316968598481469e-05, "loss": 0.0349, "step": 11380 }, { "grad_norm": 0.2004115730524063, "learning_rate": 7.312083103738206e-05, "loss": 0.0437, "step": 11390 }, { "grad_norm": 0.1856250762939453, "learning_rate": 7.307194799595958e-05, "loss": 0.0427, "step": 11400 }, { "grad_norm": 0.3031299412250519, "learning_rate": 7.302303691994473e-05, "loss": 0.0351, "step": 11410 }, { "grad_norm": 0.21357467770576477, "learning_rate": 7.29740978687691e-05, "loss": 0.05, "step": 11420 }, { "grad_norm": 0.18904723227024078, "learning_rate": 7.292513090189825e-05, "loss": 0.0331, "step": 11430 }, { "grad_norm": 0.17944231629371643, "learning_rate": 7.287613607883163e-05, "loss": 0.0338, "step": 11440 }, { "grad_norm": 0.21458730101585388, "learning_rate": 7.282711345910263e-05, "loss": 0.039, "step": 11450 }, { "grad_norm": 0.27236565947532654, "learning_rate": 7.27780631022783e-05, "loss": 0.0451, "step": 11460 }, { "grad_norm": 0.24009928107261658, "learning_rate": 7.272898506795948e-05, "loss": 0.0326, "step": 11470 }, { "grad_norm": 0.2749861776828766, "learning_rate": 7.267987941578058e-05, "loss": 0.0324, "step": 11480 }, { "grad_norm": 0.2728468179702759, "learning_rate": 7.263074620540962e-05, "loss": 0.041, "step": 11490 }, { "grad_norm": 0.18730558454990387, "learning_rate": 7.25815854965481e-05, "loss": 0.0364, "step": 11500 }, { "grad_norm": 0.21756869554519653, "learning_rate": 7.253239734893089e-05, "loss": 0.0281, "step": 11510 }, { "grad_norm": 0.21789298951625824, "learning_rate": 7.248318182232623e-05, "loss": 0.0332, "step": 11520 }, { "grad_norm": 0.2051064521074295, "learning_rate": 7.243393897653565e-05, "loss": 0.0274, "step": 11530 }, { "grad_norm": 0.18722832202911377, "learning_rate": 7.23846688713938e-05, "loss": 0.0354, "step": 11540 }, { "grad_norm": 0.20284681022167206, "learning_rate": 7.233537156676854e-05, "loss": 0.0401, "step": 11550 }, { "grad_norm": 0.24660086631774902, "learning_rate": 7.228604712256076e-05, "loss": 0.0363, "step": 11560 }, { "grad_norm": 0.25477492809295654, "learning_rate": 7.223669559870426e-05, "loss": 0.0294, "step": 11570 }, { "grad_norm": 0.29753777384757996, "learning_rate": 7.218731705516585e-05, "loss": 0.0436, "step": 11580 }, { "grad_norm": 0.22822757065296173, "learning_rate": 7.21379115519451e-05, "loss": 0.0366, "step": 11590 }, { "grad_norm": 0.18386752903461456, "learning_rate": 7.20884791490743e-05, "loss": 0.0296, "step": 11600 }, { "grad_norm": 0.1793716996908188, "learning_rate": 7.203901990661857e-05, "loss": 0.0491, "step": 11610 }, { "grad_norm": 0.28362375497817993, "learning_rate": 7.198953388467549e-05, "loss": 0.0516, "step": 11620 }, { "grad_norm": 0.25571319460868835, "learning_rate": 7.194002114337526e-05, "loss": 0.0585, "step": 11630 }, { "grad_norm": 0.20558087527751923, "learning_rate": 7.189048174288054e-05, "loss": 0.0341, "step": 11640 }, { "grad_norm": 0.25998634099960327, "learning_rate": 7.184091574338636e-05, "loss": 0.0403, "step": 11650 }, { "grad_norm": 0.1997186839580536, "learning_rate": 7.179132320512009e-05, "loss": 0.0413, "step": 11660 }, { "grad_norm": 0.18591244518756866, "learning_rate": 7.174170418834134e-05, "loss": 0.0255, "step": 11670 }, { "grad_norm": 0.22342169284820557, "learning_rate": 7.169205875334189e-05, "loss": 0.043, "step": 11680 }, { "grad_norm": 0.21058760583400726, "learning_rate": 7.164238696044562e-05, "loss": 0.0349, "step": 11690 }, { "grad_norm": 0.20036543905735016, "learning_rate": 7.159268887000846e-05, "loss": 0.0351, "step": 11700 }, { "grad_norm": 0.2219921499490738, "learning_rate": 7.154296454241827e-05, "loss": 0.0316, "step": 11710 }, { "grad_norm": 0.26134175062179565, "learning_rate": 7.149321403809479e-05, "loss": 0.0366, "step": 11720 }, { "grad_norm": 0.21305516362190247, "learning_rate": 7.144343741748956e-05, "loss": 0.0305, "step": 11730 }, { "grad_norm": 0.17928600311279297, "learning_rate": 7.139363474108589e-05, "loss": 0.0343, "step": 11740 }, { "grad_norm": 0.2807321548461914, "learning_rate": 7.134380606939874e-05, "loss": 0.0442, "step": 11750 }, { "grad_norm": 0.17526468634605408, "learning_rate": 7.12939514629746e-05, "loss": 0.0308, "step": 11760 }, { "grad_norm": 0.21607904136180878, "learning_rate": 7.124407098239155e-05, "loss": 0.0334, "step": 11770 }, { "grad_norm": 0.24536313116550446, "learning_rate": 7.119416468825908e-05, "loss": 0.0292, "step": 11780 }, { "grad_norm": 0.21912555396556854, "learning_rate": 7.114423264121804e-05, "loss": 0.0324, "step": 11790 }, { "grad_norm": 0.23140616714954376, "learning_rate": 7.109427490194056e-05, "loss": 0.0339, "step": 11800 }, { "grad_norm": 0.20326338708400726, "learning_rate": 7.104429153113001e-05, "loss": 0.0459, "step": 11810 }, { "grad_norm": 0.17200860381126404, "learning_rate": 7.099428258952092e-05, "loss": 0.035, "step": 11820 }, { "grad_norm": 0.22446954250335693, "learning_rate": 7.094424813787883e-05, "loss": 0.0481, "step": 11830 }, { "grad_norm": 0.20550380647182465, "learning_rate": 7.089418823700034e-05, "loss": 0.0322, "step": 11840 }, { "grad_norm": 0.1756659746170044, "learning_rate": 7.084410294771298e-05, "loss": 0.0338, "step": 11850 }, { "grad_norm": 0.18976259231567383, "learning_rate": 7.079399233087504e-05, "loss": 0.0386, "step": 11860 }, { "grad_norm": 0.18804162740707397, "learning_rate": 7.074385644737568e-05, "loss": 0.0296, "step": 11870 }, { "grad_norm": 0.20449590682983398, "learning_rate": 7.069369535813473e-05, "loss": 0.0487, "step": 11880 }, { "grad_norm": 0.14807897806167603, "learning_rate": 7.06435091241026e-05, "loss": 0.0279, "step": 11890 }, { "grad_norm": 0.26914849877357483, "learning_rate": 7.059329780626034e-05, "loss": 0.0325, "step": 11900 }, { "grad_norm": 0.2273307889699936, "learning_rate": 7.054306146561944e-05, "loss": 0.0393, "step": 11910 }, { "grad_norm": 0.19894933700561523, "learning_rate": 7.049280016322178e-05, "loss": 0.0338, "step": 11920 }, { "grad_norm": 0.1787034124135971, "learning_rate": 7.044251396013957e-05, "loss": 0.0367, "step": 11930 }, { "grad_norm": 0.2262076735496521, "learning_rate": 7.039220291747528e-05, "loss": 0.0343, "step": 11940 }, { "grad_norm": 0.19826503098011017, "learning_rate": 7.034186709636159e-05, "loss": 0.0258, "step": 11950 }, { "grad_norm": 0.2578589916229248, "learning_rate": 7.029150655796129e-05, "loss": 0.0304, "step": 11960 }, { "grad_norm": 0.2281111776828766, "learning_rate": 7.024112136346712e-05, "loss": 0.035, "step": 11970 }, { "grad_norm": 0.28937196731567383, "learning_rate": 7.01907115741019e-05, "loss": 0.0387, "step": 11980 }, { "grad_norm": 0.1993597000837326, "learning_rate": 7.014027725111825e-05, "loss": 0.0353, "step": 11990 }, { "grad_norm": 0.23850713670253754, "learning_rate": 7.008981845579864e-05, "loss": 0.043, "step": 12000 }, { "grad_norm": 0.18557646870613098, "learning_rate": 7.003933524945528e-05, "loss": 0.0387, "step": 12010 }, { "grad_norm": 0.16059325635433197, "learning_rate": 6.998882769342998e-05, "loss": 0.0295, "step": 12020 }, { "grad_norm": 0.19098864495754242, "learning_rate": 6.993829584909423e-05, "loss": 0.0306, "step": 12030 }, { "grad_norm": 0.19327397644519806, "learning_rate": 6.988773977784895e-05, "loss": 0.0361, "step": 12040 }, { "grad_norm": 0.19993968307971954, "learning_rate": 6.983715954112454e-05, "loss": 0.0373, "step": 12050 }, { "grad_norm": 0.22963137924671173, "learning_rate": 6.978655520038079e-05, "loss": 0.0355, "step": 12060 }, { "grad_norm": 0.23809117078781128, "learning_rate": 6.97359268171067e-05, "loss": 0.0541, "step": 12070 }, { "grad_norm": 0.24691952764987946, "learning_rate": 6.968527445282056e-05, "loss": 0.033, "step": 12080 }, { "grad_norm": 0.18024635314941406, "learning_rate": 6.963459816906977e-05, "loss": 0.0275, "step": 12090 }, { "grad_norm": 0.2038385570049286, "learning_rate": 6.958389802743077e-05, "loss": 0.0397, "step": 12100 }, { "grad_norm": 0.17748123407363892, "learning_rate": 6.953317408950903e-05, "loss": 0.0305, "step": 12110 }, { "grad_norm": 0.2357974797487259, "learning_rate": 6.94824264169389e-05, "loss": 0.0265, "step": 12120 }, { "grad_norm": 0.23855142295360565, "learning_rate": 6.94316550713836e-05, "loss": 0.0344, "step": 12130 }, { "grad_norm": 0.17587412893772125, "learning_rate": 6.938086011453513e-05, "loss": 0.0252, "step": 12140 }, { "grad_norm": 0.2223414033651352, "learning_rate": 6.93300416081141e-05, "loss": 0.0292, "step": 12150 }, { "grad_norm": 0.20358902215957642, "learning_rate": 6.927919961386984e-05, "loss": 0.026, "step": 12160 }, { "grad_norm": 0.15331725776195526, "learning_rate": 6.922833419358013e-05, "loss": 0.0307, "step": 12170 }, { "grad_norm": 0.20770806074142456, "learning_rate": 6.917744540905124e-05, "loss": 0.0451, "step": 12180 }, { "grad_norm": 0.2627474069595337, "learning_rate": 6.912653332211787e-05, "loss": 0.0346, "step": 12190 }, { "grad_norm": 0.2068651020526886, "learning_rate": 6.9075597994643e-05, "loss": 0.0422, "step": 12200 }, { "grad_norm": 0.207931786775589, "learning_rate": 6.902463948851787e-05, "loss": 0.0542, "step": 12210 }, { "grad_norm": 0.18511824309825897, "learning_rate": 6.897365786566184e-05, "loss": 0.033, "step": 12220 }, { "grad_norm": 0.24139876663684845, "learning_rate": 6.892265318802242e-05, "loss": 0.0379, "step": 12230 }, { "grad_norm": 0.20225629210472107, "learning_rate": 6.887162551757507e-05, "loss": 0.0288, "step": 12240 }, { "grad_norm": 0.2574456036090851, "learning_rate": 6.882057491632326e-05, "loss": 0.0465, "step": 12250 }, { "grad_norm": 0.2313331961631775, "learning_rate": 6.876950144629824e-05, "loss": 0.0398, "step": 12260 }, { "grad_norm": 0.24418342113494873, "learning_rate": 6.87184051695591e-05, "loss": 0.0387, "step": 12270 }, { "grad_norm": 0.20967651903629303, "learning_rate": 6.866728614819268e-05, "loss": 0.0307, "step": 12280 }, { "grad_norm": 0.2302282303571701, "learning_rate": 6.861614444431337e-05, "loss": 0.0409, "step": 12290 }, { "grad_norm": 0.21022088825702667, "learning_rate": 6.856498012006318e-05, "loss": 0.0268, "step": 12300 }, { "grad_norm": 0.20633940398693085, "learning_rate": 6.851379323761157e-05, "loss": 0.0317, "step": 12310 }, { "grad_norm": 0.2107572704553604, "learning_rate": 6.846258385915545e-05, "loss": 0.0369, "step": 12320 }, { "grad_norm": 0.1409638524055481, "learning_rate": 6.841135204691902e-05, "loss": 0.0283, "step": 12330 }, { "grad_norm": 0.20625650882720947, "learning_rate": 6.836009786315377e-05, "loss": 0.0373, "step": 12340 }, { "grad_norm": 0.18149420619010925, "learning_rate": 6.830882137013839e-05, "loss": 0.0311, "step": 12350 }, { "grad_norm": 0.19056148827075958, "learning_rate": 6.825752263017863e-05, "loss": 0.0354, "step": 12360 }, { "grad_norm": 0.2885838747024536, "learning_rate": 6.820620170560731e-05, "loss": 0.0294, "step": 12370 }, { "grad_norm": 0.21567915380001068, "learning_rate": 6.815485865878418e-05, "loss": 0.031, "step": 12380 }, { "grad_norm": 0.21836021542549133, "learning_rate": 6.810349355209588e-05, "loss": 0.0347, "step": 12390 }, { "grad_norm": 0.1669064164161682, "learning_rate": 6.805210644795588e-05, "loss": 0.0406, "step": 12400 }, { "grad_norm": 0.2267451286315918, "learning_rate": 6.800069740880433e-05, "loss": 0.0411, "step": 12410 }, { "grad_norm": 0.25267183780670166, "learning_rate": 6.794926649710807e-05, "loss": 0.0275, "step": 12420 }, { "grad_norm": 0.2376917004585266, "learning_rate": 6.789781377536052e-05, "loss": 0.0462, "step": 12430 }, { "grad_norm": 0.21754929423332214, "learning_rate": 6.784633930608158e-05, "loss": 0.0424, "step": 12440 }, { "grad_norm": 0.23343171179294586, "learning_rate": 6.779484315181759e-05, "loss": 0.0335, "step": 12450 }, { "grad_norm": 0.17943526804447174, "learning_rate": 6.774332537514122e-05, "loss": 0.0409, "step": 12460 }, { "grad_norm": 0.17528562247753143, "learning_rate": 6.769178603865143e-05, "loss": 0.0323, "step": 12470 }, { "grad_norm": 0.1302635669708252, "learning_rate": 6.764022520497337e-05, "loss": 0.0276, "step": 12480 }, { "grad_norm": 0.22071188688278198, "learning_rate": 6.758864293675833e-05, "loss": 0.0295, "step": 12490 }, { "grad_norm": 0.19273672997951508, "learning_rate": 6.753703929668362e-05, "loss": 0.0373, "step": 12500 }, { "grad_norm": 0.2646324038505554, "learning_rate": 6.748541434745253e-05, "loss": 0.0466, "step": 12510 }, { "grad_norm": 0.2976017892360687, "learning_rate": 6.743376815179424e-05, "loss": 0.0387, "step": 12520 }, { "grad_norm": 0.25198742747306824, "learning_rate": 6.738210077246376e-05, "loss": 0.0332, "step": 12530 }, { "grad_norm": 0.22069761157035828, "learning_rate": 6.733041227224181e-05, "loss": 0.0427, "step": 12540 }, { "grad_norm": 0.26411744952201843, "learning_rate": 6.72787027139348e-05, "loss": 0.0386, "step": 12550 }, { "grad_norm": 0.20788641273975372, "learning_rate": 6.72269721603747e-05, "loss": 0.0307, "step": 12560 }, { "grad_norm": 0.2759261727333069, "learning_rate": 6.717522067441904e-05, "loss": 0.0393, "step": 12570 }, { "grad_norm": 0.20782579481601715, "learning_rate": 6.712344831895074e-05, "loss": 0.0317, "step": 12580 }, { "grad_norm": 0.2841245234012604, "learning_rate": 6.707165515687811e-05, "loss": 0.0424, "step": 12590 }, { "grad_norm": 0.22693940997123718, "learning_rate": 6.70198412511347e-05, "loss": 0.0367, "step": 12600 }, { "grad_norm": 0.2937001585960388, "learning_rate": 6.69680066646793e-05, "loss": 0.0307, "step": 12610 }, { "grad_norm": 0.15974771976470947, "learning_rate": 6.691615146049583e-05, "loss": 0.0296, "step": 12620 }, { "grad_norm": 0.21191959083080292, "learning_rate": 6.686427570159324e-05, "loss": 0.0278, "step": 12630 }, { "grad_norm": 0.26813802123069763, "learning_rate": 6.681237945100548e-05, "loss": 0.0467, "step": 12640 }, { "grad_norm": 0.18900974094867706, "learning_rate": 6.676046277179139e-05, "loss": 0.0307, "step": 12650 }, { "grad_norm": 0.23223617672920227, "learning_rate": 6.670852572703462e-05, "loss": 0.0472, "step": 12660 }, { "grad_norm": 0.1825520545244217, "learning_rate": 6.665656837984359e-05, "loss": 0.0347, "step": 12670 }, { "grad_norm": 0.22293280065059662, "learning_rate": 6.660459079335135e-05, "loss": 0.0446, "step": 12680 }, { "grad_norm": 0.20572049915790558, "learning_rate": 6.655259303071558e-05, "loss": 0.0385, "step": 12690 }, { "grad_norm": 0.19207362830638885, "learning_rate": 6.650057515511848e-05, "loss": 0.0349, "step": 12700 }, { "grad_norm": 0.23845228552818298, "learning_rate": 6.644853722976667e-05, "loss": 0.0296, "step": 12710 }, { "grad_norm": 0.2009299397468567, "learning_rate": 6.639647931789113e-05, "loss": 0.0356, "step": 12720 }, { "grad_norm": 0.1822335422039032, "learning_rate": 6.634440148274713e-05, "loss": 0.0299, "step": 12730 }, { "grad_norm": 0.1785789430141449, "learning_rate": 6.629230378761415e-05, "loss": 0.0365, "step": 12740 }, { "grad_norm": 0.18480674922466278, "learning_rate": 6.624018629579581e-05, "loss": 0.0278, "step": 12750 }, { "grad_norm": 0.22947920858860016, "learning_rate": 6.618804907061976e-05, "loss": 0.0459, "step": 12760 }, { "grad_norm": 0.2577440142631531, "learning_rate": 6.613589217543766e-05, "loss": 0.0507, "step": 12770 }, { "grad_norm": 0.2271593064069748, "learning_rate": 6.608371567362504e-05, "loss": 0.0373, "step": 12780 }, { "grad_norm": 0.20656532049179077, "learning_rate": 6.60315196285813e-05, "loss": 0.0391, "step": 12790 }, { "grad_norm": 0.22931186854839325, "learning_rate": 6.597930410372954e-05, "loss": 0.0353, "step": 12800 }, { "grad_norm": 0.21692313253879547, "learning_rate": 6.592706916251653e-05, "loss": 0.038, "step": 12810 }, { "grad_norm": 0.16582746803760529, "learning_rate": 6.587481486841267e-05, "loss": 0.0404, "step": 12820 }, { "grad_norm": 0.29444846510887146, "learning_rate": 6.582254128491184e-05, "loss": 0.0373, "step": 12830 }, { "grad_norm": 0.17845116555690765, "learning_rate": 6.577024847553139e-05, "loss": 0.0314, "step": 12840 }, { "grad_norm": 0.2097546011209488, "learning_rate": 6.571793650381201e-05, "loss": 0.034, "step": 12850 }, { "grad_norm": 0.15086252987384796, "learning_rate": 6.566560543331766e-05, "loss": 0.039, "step": 12860 }, { "grad_norm": 0.20577900111675262, "learning_rate": 6.561325532763554e-05, "loss": 0.0297, "step": 12870 }, { "grad_norm": 0.19441434741020203, "learning_rate": 6.556088625037597e-05, "loss": 0.0268, "step": 12880 }, { "grad_norm": 0.195923313498497, "learning_rate": 6.55084982651723e-05, "loss": 0.0305, "step": 12890 }, { "grad_norm": 0.19240650534629822, "learning_rate": 6.54560914356809e-05, "loss": 0.0456, "step": 12900 }, { "grad_norm": 0.1837531328201294, "learning_rate": 6.540366582558097e-05, "loss": 0.0282, "step": 12910 }, { "grad_norm": 0.2300800383090973, "learning_rate": 6.53512214985746e-05, "loss": 0.0254, "step": 12920 }, { "grad_norm": 0.17217852175235748, "learning_rate": 6.529875851838658e-05, "loss": 0.0322, "step": 12930 }, { "grad_norm": 0.16264893114566803, "learning_rate": 6.52462769487644e-05, "loss": 0.0318, "step": 12940 }, { "grad_norm": 0.1393808275461197, "learning_rate": 6.519377685347807e-05, "loss": 0.0259, "step": 12950 }, { "grad_norm": 0.17456215620040894, "learning_rate": 6.514125829632021e-05, "loss": 0.0265, "step": 12960 }, { "grad_norm": 0.2666062116622925, "learning_rate": 6.508872134110577e-05, "loss": 0.0301, "step": 12970 }, { "grad_norm": 0.22711321711540222, "learning_rate": 6.503616605167213e-05, "loss": 0.0263, "step": 12980 }, { "grad_norm": 0.1959354132413864, "learning_rate": 6.498359249187893e-05, "loss": 0.0302, "step": 12990 }, { "grad_norm": 0.23334607481956482, "learning_rate": 6.493100072560799e-05, "loss": 0.0307, "step": 13000 }, { "grad_norm": 0.18357446789741516, "learning_rate": 6.487839081676327e-05, "loss": 0.0509, "step": 13010 }, { "grad_norm": 0.1780903935432434, "learning_rate": 6.482576282927076e-05, "loss": 0.0435, "step": 13020 }, { "grad_norm": 0.2285001426935196, "learning_rate": 6.477311682707845e-05, "loss": 0.0447, "step": 13030 }, { "grad_norm": 0.21082235872745514, "learning_rate": 6.472045287415616e-05, "loss": 0.0293, "step": 13040 }, { "grad_norm": 0.2524701952934265, "learning_rate": 6.466777103449558e-05, "loss": 0.0279, "step": 13050 }, { "grad_norm": 0.2176787555217743, "learning_rate": 6.461507137211012e-05, "loss": 0.0307, "step": 13060 }, { "grad_norm": 0.2490358203649521, "learning_rate": 6.456235395103483e-05, "loss": 0.0434, "step": 13070 }, { "grad_norm": 0.2356301248073578, "learning_rate": 6.450961883532634e-05, "loss": 0.0341, "step": 13080 }, { "grad_norm": 0.16627436876296997, "learning_rate": 6.445686608906283e-05, "loss": 0.0336, "step": 13090 }, { "grad_norm": 0.230295792222023, "learning_rate": 6.44040957763438e-05, "loss": 0.0349, "step": 13100 }, { "grad_norm": 0.16852883994579315, "learning_rate": 6.435130796129018e-05, "loss": 0.0267, "step": 13110 }, { "grad_norm": 0.18314160406589508, "learning_rate": 6.429850270804416e-05, "loss": 0.0245, "step": 13120 }, { "grad_norm": 0.20287933945655823, "learning_rate": 6.424568008076909e-05, "loss": 0.0439, "step": 13130 }, { "grad_norm": 0.23461277782917023, "learning_rate": 6.419284014364943e-05, "loss": 0.0374, "step": 13140 }, { "grad_norm": 0.18870609998703003, "learning_rate": 6.41399829608907e-05, "loss": 0.0341, "step": 13150 }, { "grad_norm": 0.21097031235694885, "learning_rate": 6.408710859671938e-05, "loss": 0.0275, "step": 13160 }, { "grad_norm": 0.2141493260860443, "learning_rate": 6.403421711538278e-05, "loss": 0.028, "step": 13170 }, { "grad_norm": 0.1577443778514862, "learning_rate": 6.398130858114903e-05, "loss": 0.0307, "step": 13180 }, { "grad_norm": 0.20675460994243622, "learning_rate": 6.392838305830701e-05, "loss": 0.0282, "step": 13190 }, { "grad_norm": 0.23158922791481018, "learning_rate": 6.387544061116621e-05, "loss": 0.0327, "step": 13200 }, { "grad_norm": 0.16248519718647003, "learning_rate": 6.382248130405671e-05, "loss": 0.02, "step": 13210 }, { "grad_norm": 0.14729534089565277, "learning_rate": 6.376950520132905e-05, "loss": 0.0247, "step": 13220 }, { "grad_norm": 0.24640007317066193, "learning_rate": 6.371651236735418e-05, "loss": 0.0309, "step": 13230 }, { "grad_norm": 0.20975331962108612, "learning_rate": 6.366350286652341e-05, "loss": 0.0441, "step": 13240 }, { "grad_norm": 0.24000178277492523, "learning_rate": 6.361047676324827e-05, "loss": 0.0273, "step": 13250 }, { "grad_norm": 0.21374762058258057, "learning_rate": 6.355743412196047e-05, "loss": 0.0349, "step": 13260 }, { "grad_norm": 0.17652471363544464, "learning_rate": 6.350437500711184e-05, "loss": 0.0317, "step": 13270 }, { "grad_norm": 0.23415538668632507, "learning_rate": 6.345129948317418e-05, "loss": 0.0297, "step": 13280 }, { "grad_norm": 0.19566041231155396, "learning_rate": 6.33982076146393e-05, "loss": 0.0261, "step": 13290 }, { "grad_norm": 0.2317422777414322, "learning_rate": 6.334509946601879e-05, "loss": 0.0283, "step": 13300 }, { "grad_norm": 0.143731951713562, "learning_rate": 6.329197510184406e-05, "loss": 0.0357, "step": 13310 }, { "grad_norm": 0.2541100084781647, "learning_rate": 6.323883458666624e-05, "loss": 0.043, "step": 13320 }, { "grad_norm": 0.2262498140335083, "learning_rate": 6.318567798505605e-05, "loss": 0.0318, "step": 13330 }, { "grad_norm": 0.20021890103816986, "learning_rate": 6.313250536160378e-05, "loss": 0.0316, "step": 13340 }, { "grad_norm": 0.18915250897407532, "learning_rate": 6.307931678091917e-05, "loss": 0.0267, "step": 13350 }, { "grad_norm": 0.1565304696559906, "learning_rate": 6.302611230763138e-05, "loss": 0.0342, "step": 13360 }, { "grad_norm": 0.18347346782684326, "learning_rate": 6.297289200638887e-05, "loss": 0.0355, "step": 13370 }, { "grad_norm": 0.2785630226135254, "learning_rate": 6.29196559418593e-05, "loss": 0.0314, "step": 13380 }, { "grad_norm": 0.2014230191707611, "learning_rate": 6.286640417872951e-05, "loss": 0.0387, "step": 13390 }, { "grad_norm": 0.20572972297668457, "learning_rate": 6.281313678170542e-05, "loss": 0.0413, "step": 13400 }, { "grad_norm": 0.20329295098781586, "learning_rate": 6.275985381551194e-05, "loss": 0.0351, "step": 13410 }, { "grad_norm": 0.1239226907491684, "learning_rate": 6.270655534489292e-05, "loss": 0.0346, "step": 13420 }, { "grad_norm": 0.27133694291114807, "learning_rate": 6.265324143461098e-05, "loss": 0.0304, "step": 13430 }, { "grad_norm": 0.1723014861345291, "learning_rate": 6.259991214944758e-05, "loss": 0.0301, "step": 13440 }, { "grad_norm": 0.33376774191856384, "learning_rate": 6.254656755420283e-05, "loss": 0.0489, "step": 13450 }, { "grad_norm": 0.2029126137495041, "learning_rate": 6.249320771369545e-05, "loss": 0.0412, "step": 13460 }, { "grad_norm": 0.19274817407131195, "learning_rate": 6.243983269276263e-05, "loss": 0.0471, "step": 13470 }, { "grad_norm": 0.16773293912410736, "learning_rate": 6.238644255626012e-05, "loss": 0.0407, "step": 13480 }, { "grad_norm": 0.1813092678785324, "learning_rate": 6.233303736906193e-05, "loss": 0.0445, "step": 13490 }, { "grad_norm": 0.22510384023189545, "learning_rate": 6.227961719606039e-05, "loss": 0.0259, "step": 13500 }, { "grad_norm": 0.19456100463867188, "learning_rate": 6.222618210216608e-05, "loss": 0.0433, "step": 13510 }, { "grad_norm": 0.16481295228004456, "learning_rate": 6.217273215230767e-05, "loss": 0.0444, "step": 13520 }, { "grad_norm": 0.20593784749507904, "learning_rate": 6.211926741143188e-05, "loss": 0.0383, "step": 13530 }, { "grad_norm": 0.16661639511585236, "learning_rate": 6.20657879445034e-05, "loss": 0.0369, "step": 13540 }, { "grad_norm": 0.215183287858963, "learning_rate": 6.201229381650485e-05, "loss": 0.0272, "step": 13550 }, { "grad_norm": 0.22547286748886108, "learning_rate": 6.195878509243661e-05, "loss": 0.0387, "step": 13560 }, { "grad_norm": 0.18910841643810272, "learning_rate": 6.190526183731685e-05, "loss": 0.0278, "step": 13570 }, { "grad_norm": 0.247131809592247, "learning_rate": 6.185172411618138e-05, "loss": 0.034, "step": 13580 }, { "grad_norm": 0.15028734505176544, "learning_rate": 6.179817199408355e-05, "loss": 0.0428, "step": 13590 }, { "grad_norm": 0.23681946098804474, "learning_rate": 6.174460553609426e-05, "loss": 0.056, "step": 13600 }, { "grad_norm": 0.1996772140264511, "learning_rate": 6.16910248073018e-05, "loss": 0.0294, "step": 13610 }, { "grad_norm": 0.18251745402812958, "learning_rate": 6.16374298728118e-05, "loss": 0.0446, "step": 13620 }, { "grad_norm": 0.21300379931926727, "learning_rate": 6.158382079774716e-05, "loss": 0.0282, "step": 13630 }, { "grad_norm": 0.1994907408952713, "learning_rate": 6.153019764724799e-05, "loss": 0.0326, "step": 13640 }, { "grad_norm": 0.2106035351753235, "learning_rate": 6.147656048647144e-05, "loss": 0.0408, "step": 13650 }, { "grad_norm": 0.30005118250846863, "learning_rate": 6.142290938059173e-05, "loss": 0.041, "step": 13660 }, { "grad_norm": 0.2009482979774475, "learning_rate": 6.136924439480001e-05, "loss": 0.0307, "step": 13670 }, { "grad_norm": 0.2016477882862091, "learning_rate": 6.13155655943043e-05, "loss": 0.0283, "step": 13680 }, { "grad_norm": 0.26654449105262756, "learning_rate": 6.126187304432941e-05, "loss": 0.0386, "step": 13690 }, { "grad_norm": 0.18891321122646332, "learning_rate": 6.120816681011684e-05, "loss": 0.0338, "step": 13700 }, { "grad_norm": 0.22431381046772003, "learning_rate": 6.115444695692474e-05, "loss": 0.0305, "step": 13710 }, { "grad_norm": 0.18426962196826935, "learning_rate": 6.110071355002779e-05, "loss": 0.026, "step": 13720 }, { "grad_norm": 0.2011214792728424, "learning_rate": 6.104696665471714e-05, "loss": 0.025, "step": 13730 }, { "grad_norm": 0.14333146810531616, "learning_rate": 6.099320633630036e-05, "loss": 0.0238, "step": 13740 }, { "grad_norm": 0.14756740629673004, "learning_rate": 6.093943266010128e-05, "loss": 0.0317, "step": 13750 }, { "grad_norm": 0.27212756872177124, "learning_rate": 6.088564569146e-05, "loss": 0.0514, "step": 13760 }, { "grad_norm": 0.166418194770813, "learning_rate": 6.083184549573274e-05, "loss": 0.0258, "step": 13770 }, { "grad_norm": 0.16172556579113007, "learning_rate": 6.077803213829184e-05, "loss": 0.028, "step": 13780 }, { "grad_norm": 0.21507908403873444, "learning_rate": 6.0724205684525594e-05, "loss": 0.0313, "step": 13790 }, { "grad_norm": 0.1925247311592102, "learning_rate": 6.0670366199838215e-05, "loss": 0.0268, "step": 13800 }, { "grad_norm": 0.19990235567092896, "learning_rate": 6.061651374964974e-05, "loss": 0.0374, "step": 13810 }, { "grad_norm": 0.19695794582366943, "learning_rate": 6.0562648399396004e-05, "loss": 0.034, "step": 13820 }, { "grad_norm": 0.19952715933322906, "learning_rate": 6.050877021452845e-05, "loss": 0.0362, "step": 13830 }, { "grad_norm": 0.15110434591770172, "learning_rate": 6.04548792605142e-05, "loss": 0.0376, "step": 13840 }, { "grad_norm": 0.23436006903648376, "learning_rate": 6.04009756028358e-05, "loss": 0.0264, "step": 13850 }, { "grad_norm": 0.13880528509616852, "learning_rate": 6.0347059306991294e-05, "loss": 0.0311, "step": 13860 }, { "grad_norm": 0.12060677260160446, "learning_rate": 6.029313043849407e-05, "loss": 0.0367, "step": 13870 }, { "grad_norm": 0.14109617471694946, "learning_rate": 6.023918906287279e-05, "loss": 0.0224, "step": 13880 }, { "grad_norm": 0.23832552134990692, "learning_rate": 6.01852352456713e-05, "loss": 0.0406, "step": 13890 }, { "grad_norm": 0.13085034489631653, "learning_rate": 6.013126905244858e-05, "loss": 0.0238, "step": 13900 }, { "grad_norm": 0.2091747224330902, "learning_rate": 6.007729054877864e-05, "loss": 0.0313, "step": 13910 }, { "grad_norm": 0.19043736159801483, "learning_rate": 6.0023299800250466e-05, "loss": 0.0394, "step": 13920 }, { "grad_norm": 0.20711570978164673, "learning_rate": 5.9969296872467894e-05, "loss": 0.0355, "step": 13930 }, { "grad_norm": 0.1581268310546875, "learning_rate": 5.9915281831049585e-05, "loss": 0.0301, "step": 13940 }, { "grad_norm": 0.1576516181230545, "learning_rate": 5.98612547416289e-05, "loss": 0.0287, "step": 13950 }, { "grad_norm": 0.2333119660615921, "learning_rate": 5.9807215669853855e-05, "loss": 0.0348, "step": 13960 }, { "grad_norm": 0.2993478775024414, "learning_rate": 5.9753164681387e-05, "loss": 0.0536, "step": 13970 }, { "grad_norm": 0.2733190953731537, "learning_rate": 5.969910184190539e-05, "loss": 0.039, "step": 13980 }, { "grad_norm": 0.16756166517734528, "learning_rate": 5.9645027217100477e-05, "loss": 0.0359, "step": 13990 }, { "grad_norm": 0.2577950358390808, "learning_rate": 5.959094087267804e-05, "loss": 0.0341, "step": 14000 }, { "grad_norm": 0.19272616505622864, "learning_rate": 5.9536842874358065e-05, "loss": 0.027, "step": 14010 }, { "grad_norm": 0.20314079523086548, "learning_rate": 5.9482733287874734e-05, "loss": 0.0449, "step": 14020 }, { "grad_norm": 0.15251007676124573, "learning_rate": 5.942861217897631e-05, "loss": 0.0341, "step": 14030 }, { "grad_norm": 0.17837302386760712, "learning_rate": 5.9374479613425e-05, "loss": 0.0265, "step": 14040 }, { "grad_norm": 0.1343550831079483, "learning_rate": 5.932033565699704e-05, "loss": 0.0243, "step": 14050 }, { "grad_norm": 0.24749648571014404, "learning_rate": 5.926618037548237e-05, "loss": 0.0342, "step": 14060 }, { "grad_norm": 0.17731140553951263, "learning_rate": 5.9212013834684824e-05, "loss": 0.0338, "step": 14070 }, { "grad_norm": 0.18263112008571625, "learning_rate": 5.9157836100421824e-05, "loss": 0.0319, "step": 14080 }, { "grad_norm": 0.1422201693058014, "learning_rate": 5.910364723852444e-05, "loss": 0.0286, "step": 14090 }, { "grad_norm": 0.18987135589122772, "learning_rate": 5.904944731483723e-05, "loss": 0.0382, "step": 14100 }, { "grad_norm": 0.20541095733642578, "learning_rate": 5.899523639521825e-05, "loss": 0.0344, "step": 14110 }, { "grad_norm": 0.1406869888305664, "learning_rate": 5.894101454553883e-05, "loss": 0.0271, "step": 14120 }, { "grad_norm": 0.19461899995803833, "learning_rate": 5.888678183168368e-05, "loss": 0.0389, "step": 14130 }, { "grad_norm": 0.18807439506053925, "learning_rate": 5.883253831955061e-05, "loss": 0.0348, "step": 14140 }, { "grad_norm": 0.18073450028896332, "learning_rate": 5.8778284075050625e-05, "loss": 0.0268, "step": 14150 }, { "grad_norm": 0.22673805058002472, "learning_rate": 5.872401916410777e-05, "loss": 0.0517, "step": 14160 }, { "grad_norm": 0.27724531292915344, "learning_rate": 5.8669743652659014e-05, "loss": 0.0339, "step": 14170 }, { "grad_norm": 0.23960255086421967, "learning_rate": 5.86154576066542e-05, "loss": 0.0308, "step": 14180 }, { "grad_norm": 0.24701827764511108, "learning_rate": 5.856116109205602e-05, "loss": 0.0396, "step": 14190 }, { "grad_norm": 0.1822662651538849, "learning_rate": 5.850685417483983e-05, "loss": 0.0338, "step": 14200 }, { "grad_norm": 0.16172094643115997, "learning_rate": 5.8452536920993684e-05, "loss": 0.0316, "step": 14210 }, { "grad_norm": 0.1632997989654541, "learning_rate": 5.839820939651817e-05, "loss": 0.0399, "step": 14220 }, { "grad_norm": 0.184631809592247, "learning_rate": 5.834387166742632e-05, "loss": 0.0289, "step": 14230 }, { "grad_norm": 0.22746214270591736, "learning_rate": 5.828952379974364e-05, "loss": 0.0323, "step": 14240 }, { "grad_norm": 0.22836022078990936, "learning_rate": 5.8235165859507864e-05, "loss": 0.0392, "step": 14250 }, { "grad_norm": 0.16332457959651947, "learning_rate": 5.818079791276907e-05, "loss": 0.0287, "step": 14260 }, { "grad_norm": 0.1986784040927887, "learning_rate": 5.8126420025589415e-05, "loss": 0.0454, "step": 14270 }, { "grad_norm": 0.2056928128004074, "learning_rate": 5.807203226404313e-05, "loss": 0.0331, "step": 14280 }, { "grad_norm": 0.2104800045490265, "learning_rate": 5.801763469421652e-05, "loss": 0.0241, "step": 14290 }, { "grad_norm": 0.14998765289783478, "learning_rate": 5.796322738220774e-05, "loss": 0.0321, "step": 14300 }, { "grad_norm": 0.19466973841190338, "learning_rate": 5.79088103941268e-05, "loss": 0.0279, "step": 14310 }, { "grad_norm": 0.20867998898029327, "learning_rate": 5.785438379609549e-05, "loss": 0.046, "step": 14320 }, { "grad_norm": 0.1769104152917862, "learning_rate": 5.779994765424724e-05, "loss": 0.0271, "step": 14330 }, { "grad_norm": 0.13131432235240936, "learning_rate": 5.7745502034727114e-05, "loss": 0.0306, "step": 14340 }, { "grad_norm": 0.1973579376935959, "learning_rate": 5.769104700369164e-05, "loss": 0.0318, "step": 14350 }, { "grad_norm": 0.17501404881477356, "learning_rate": 5.7636582627308854e-05, "loss": 0.0278, "step": 14360 }, { "grad_norm": 0.1878010630607605, "learning_rate": 5.7582108971758095e-05, "loss": 0.0314, "step": 14370 }, { "grad_norm": 0.20265860855579376, "learning_rate": 5.752762610322999e-05, "loss": 0.0302, "step": 14380 }, { "grad_norm": 0.2528924345970154, "learning_rate": 5.747313408792636e-05, "loss": 0.0437, "step": 14390 }, { "grad_norm": 0.1707998812198639, "learning_rate": 5.741863299206014e-05, "loss": 0.0327, "step": 14400 }, { "grad_norm": 0.18738983571529388, "learning_rate": 5.73641228818553e-05, "loss": 0.0284, "step": 14410 }, { "grad_norm": 0.13990353047847748, "learning_rate": 5.730960382354676e-05, "loss": 0.029, "step": 14420 }, { "grad_norm": 0.12964816391468048, "learning_rate": 5.725507588338035e-05, "loss": 0.0288, "step": 14430 }, { "grad_norm": 0.18815720081329346, "learning_rate": 5.7200539127612604e-05, "loss": 0.0247, "step": 14440 }, { "grad_norm": 0.1902053952217102, "learning_rate": 5.714599362251087e-05, "loss": 0.0304, "step": 14450 }, { "grad_norm": 0.1474921703338623, "learning_rate": 5.709143943435307e-05, "loss": 0.0313, "step": 14460 }, { "grad_norm": 0.18455132842063904, "learning_rate": 5.7036876629427646e-05, "loss": 0.0361, "step": 14470 }, { "grad_norm": 0.18754181265830994, "learning_rate": 5.698230527403361e-05, "loss": 0.0268, "step": 14480 }, { "grad_norm": 0.1865972876548767, "learning_rate": 5.6927725434480274e-05, "loss": 0.0277, "step": 14490 }, { "grad_norm": 0.16402579843997955, "learning_rate": 5.687313717708728e-05, "loss": 0.0421, "step": 14500 }, { "grad_norm": 0.22958630323410034, "learning_rate": 5.681854056818453e-05, "loss": 0.0332, "step": 14510 }, { "grad_norm": 0.16093195974826813, "learning_rate": 5.676393567411205e-05, "loss": 0.0237, "step": 14520 }, { "grad_norm": 0.2527540922164917, "learning_rate": 5.670932256121991e-05, "loss": 0.0483, "step": 14530 }, { "grad_norm": 0.16379407048225403, "learning_rate": 5.6654701295868216e-05, "loss": 0.0242, "step": 14540 }, { "grad_norm": 0.2278575599193573, "learning_rate": 5.660007194442697e-05, "loss": 0.0253, "step": 14550 }, { "grad_norm": 0.14975406229496002, "learning_rate": 5.6545434573275946e-05, "loss": 0.0279, "step": 14560 }, { "grad_norm": 0.21807970106601715, "learning_rate": 5.649078924880472e-05, "loss": 0.0238, "step": 14570 }, { "grad_norm": 0.21484090387821198, "learning_rate": 5.643613603741252e-05, "loss": 0.0347, "step": 14580 }, { "grad_norm": 0.1796574741601944, "learning_rate": 5.6381475005508154e-05, "loss": 0.0453, "step": 14590 }, { "grad_norm": 0.17536048591136932, "learning_rate": 5.632680621950992e-05, "loss": 0.0313, "step": 14600 }, { "grad_norm": 0.15919505059719086, "learning_rate": 5.627212974584555e-05, "loss": 0.025, "step": 14610 }, { "grad_norm": 0.14604268968105316, "learning_rate": 5.62174456509521e-05, "loss": 0.025, "step": 14620 }, { "grad_norm": 0.1436149924993515, "learning_rate": 5.616275400127594e-05, "loss": 0.0291, "step": 14630 }, { "grad_norm": 0.18078698217868805, "learning_rate": 5.610805486327254e-05, "loss": 0.022, "step": 14640 }, { "grad_norm": 0.20685580372810364, "learning_rate": 5.6053348303406536e-05, "loss": 0.0322, "step": 14650 }, { "grad_norm": 0.20441849529743195, "learning_rate": 5.599863438815156e-05, "loss": 0.0273, "step": 14660 }, { "grad_norm": 0.19948896765708923, "learning_rate": 5.594391318399017e-05, "loss": 0.0381, "step": 14670 }, { "grad_norm": 0.24139241874217987, "learning_rate": 5.588918475741377e-05, "loss": 0.0284, "step": 14680 }, { "grad_norm": 0.18850520253181458, "learning_rate": 5.5834449174922586e-05, "loss": 0.0323, "step": 14690 }, { "grad_norm": 0.14516226947307587, "learning_rate": 5.57797065030255e-05, "loss": 0.0343, "step": 14700 }, { "grad_norm": 0.23033872246742249, "learning_rate": 5.5724956808240016e-05, "loss": 0.0367, "step": 14710 }, { "grad_norm": 0.1910519301891327, "learning_rate": 5.567020015709219e-05, "loss": 0.0309, "step": 14720 }, { "grad_norm": 0.12635844945907593, "learning_rate": 5.561543661611649e-05, "loss": 0.0304, "step": 14730 }, { "grad_norm": 0.17694571614265442, "learning_rate": 5.556066625185583e-05, "loss": 0.0341, "step": 14740 }, { "grad_norm": 0.13062593340873718, "learning_rate": 5.550588913086131e-05, "loss": 0.0315, "step": 14750 }, { "grad_norm": 0.21010665595531464, "learning_rate": 5.545110531969234e-05, "loss": 0.0337, "step": 14760 }, { "grad_norm": 0.1744592785835266, "learning_rate": 5.539631488491641e-05, "loss": 0.031, "step": 14770 }, { "grad_norm": 0.2283664494752884, "learning_rate": 5.534151789310904e-05, "loss": 0.0302, "step": 14780 }, { "grad_norm": 0.18571536242961884, "learning_rate": 5.528671441085376e-05, "loss": 0.023, "step": 14790 }, { "grad_norm": 0.22915449738502502, "learning_rate": 5.523190450474197e-05, "loss": 0.0317, "step": 14800 }, { "grad_norm": 0.16852588951587677, "learning_rate": 5.5177088241372865e-05, "loss": 0.022, "step": 14810 }, { "grad_norm": 0.16375407576560974, "learning_rate": 5.512226568735338e-05, "loss": 0.0374, "step": 14820 }, { "grad_norm": 0.18927864730358124, "learning_rate": 5.506743690929809e-05, "loss": 0.0394, "step": 14830 }, { "grad_norm": 0.18168985843658447, "learning_rate": 5.501260197382912e-05, "loss": 0.0435, "step": 14840 }, { "grad_norm": 0.16342012584209442, "learning_rate": 5.49577609475761e-05, "loss": 0.0463, "step": 14850 }, { "grad_norm": 0.19524148106575012, "learning_rate": 5.490291389717603e-05, "loss": 0.0266, "step": 14860 }, { "grad_norm": 0.1744721233844757, "learning_rate": 5.484806088927329e-05, "loss": 0.0258, "step": 14870 }, { "grad_norm": 0.24693726003170013, "learning_rate": 5.479320199051942e-05, "loss": 0.0317, "step": 14880 }, { "grad_norm": 0.18724927306175232, "learning_rate": 5.473833726757314e-05, "loss": 0.0271, "step": 14890 }, { "grad_norm": 0.2372477799654007, "learning_rate": 5.4683466787100304e-05, "loss": 0.0262, "step": 14900 }, { "grad_norm": 0.21471421420574188, "learning_rate": 5.462859061577369e-05, "loss": 0.0309, "step": 14910 }, { "grad_norm": 0.15774790942668915, "learning_rate": 5.4573708820273026e-05, "loss": 0.0248, "step": 14920 }, { "grad_norm": 0.2296208143234253, "learning_rate": 5.451882146728488e-05, "loss": 0.0257, "step": 14930 }, { "grad_norm": 0.13449518382549286, "learning_rate": 5.446392862350255e-05, "loss": 0.0377, "step": 14940 }, { "grad_norm": 0.2033645212650299, "learning_rate": 5.440903035562603e-05, "loss": 0.0406, "step": 14950 }, { "grad_norm": 0.2234881967306137, "learning_rate": 5.435412673036188e-05, "loss": 0.0294, "step": 14960 }, { "grad_norm": 0.27564162015914917, "learning_rate": 5.429921781442318e-05, "loss": 0.0302, "step": 14970 }, { "grad_norm": 0.17778237164020538, "learning_rate": 5.424430367452945e-05, "loss": 0.026, "step": 14980 }, { "grad_norm": 0.19313441216945648, "learning_rate": 5.418938437740655e-05, "loss": 0.0295, "step": 14990 }, { "grad_norm": 0.24673429131507874, "learning_rate": 5.413445998978658e-05, "loss": 0.0382, "step": 15000 }, { "grad_norm": 0.1659199595451355, "learning_rate": 5.407953057840789e-05, "loss": 0.0236, "step": 15010 }, { "grad_norm": 0.20957469940185547, "learning_rate": 5.4024596210014853e-05, "loss": 0.0277, "step": 15020 }, { "grad_norm": 0.19345799088478088, "learning_rate": 5.3969656951357935e-05, "loss": 0.0265, "step": 15030 }, { "grad_norm": 0.1927626132965088, "learning_rate": 5.3914712869193504e-05, "loss": 0.0402, "step": 15040 }, { "grad_norm": 0.1690162569284439, "learning_rate": 5.385976403028381e-05, "loss": 0.0441, "step": 15050 }, { "grad_norm": 0.21517270803451538, "learning_rate": 5.3804810501396875e-05, "loss": 0.041, "step": 15060 }, { "grad_norm": 0.1479446440935135, "learning_rate": 5.374985234930639e-05, "loss": 0.0284, "step": 15070 }, { "grad_norm": 0.168551966547966, "learning_rate": 5.3694889640791724e-05, "loss": 0.029, "step": 15080 }, { "grad_norm": 0.19270923733711243, "learning_rate": 5.3639922442637736e-05, "loss": 0.0352, "step": 15090 }, { "grad_norm": 0.15864215791225433, "learning_rate": 5.358495082163476e-05, "loss": 0.0289, "step": 15100 }, { "grad_norm": 0.13892872631549835, "learning_rate": 5.35299748445785e-05, "loss": 0.0295, "step": 15110 }, { "grad_norm": 0.19725961983203888, "learning_rate": 5.3474994578269944e-05, "loss": 0.0232, "step": 15120 }, { "grad_norm": 0.1501840502023697, "learning_rate": 5.34200100895153e-05, "loss": 0.0265, "step": 15130 }, { "grad_norm": 0.1944458782672882, "learning_rate": 5.3365021445125916e-05, "loss": 0.018, "step": 15140 }, { "grad_norm": 0.2049548178911209, "learning_rate": 5.331002871191817e-05, "loss": 0.0291, "step": 15150 }, { "grad_norm": 0.18238000571727753, "learning_rate": 5.3255031956713443e-05, "loss": 0.0297, "step": 15160 }, { "grad_norm": 0.2033488154411316, "learning_rate": 5.320003124633795e-05, "loss": 0.0346, "step": 15170 }, { "grad_norm": 0.15310707688331604, "learning_rate": 5.314502664762274e-05, "loss": 0.0306, "step": 15180 }, { "grad_norm": 0.20466585457324982, "learning_rate": 5.3090018227403604e-05, "loss": 0.0241, "step": 15190 }, { "grad_norm": 0.1683662235736847, "learning_rate": 5.3035006052520955e-05, "loss": 0.0328, "step": 15200 }, { "grad_norm": 0.1209966167807579, "learning_rate": 5.297999018981976e-05, "loss": 0.029, "step": 15210 }, { "grad_norm": 0.23305171728134155, "learning_rate": 5.2924970706149505e-05, "loss": 0.0321, "step": 15220 }, { "grad_norm": 0.18811765313148499, "learning_rate": 5.2869947668364014e-05, "loss": 0.0223, "step": 15230 }, { "grad_norm": 0.17748038470745087, "learning_rate": 5.2814921143321504e-05, "loss": 0.0246, "step": 15240 }, { "grad_norm": 0.23303374648094177, "learning_rate": 5.275989119788436e-05, "loss": 0.0447, "step": 15250 }, { "grad_norm": 0.22319155931472778, "learning_rate": 5.270485789891919e-05, "loss": 0.034, "step": 15260 }, { "grad_norm": 0.15018324553966522, "learning_rate": 5.264982131329661e-05, "loss": 0.0307, "step": 15270 }, { "grad_norm": 0.16031254827976227, "learning_rate": 5.259478150789128e-05, "loss": 0.0263, "step": 15280 }, { "grad_norm": 0.17276938259601593, "learning_rate": 5.253973854958173e-05, "loss": 0.0417, "step": 15290 }, { "grad_norm": 0.20104090869426727, "learning_rate": 5.2484692505250376e-05, "loss": 0.033, "step": 15300 }, { "grad_norm": 0.11932551860809326, "learning_rate": 5.242964344178333e-05, "loss": 0.0358, "step": 15310 }, { "grad_norm": 0.19675248861312866, "learning_rate": 5.237459142607041e-05, "loss": 0.0231, "step": 15320 }, { "grad_norm": 0.18975374102592468, "learning_rate": 5.231953652500498e-05, "loss": 0.0247, "step": 15330 }, { "grad_norm": 0.18572260439395905, "learning_rate": 5.2264478805483976e-05, "loss": 0.0302, "step": 15340 }, { "grad_norm": 0.17348280549049377, "learning_rate": 5.220941833440768e-05, "loss": 0.0353, "step": 15350 }, { "grad_norm": 0.1869698017835617, "learning_rate": 5.215435517867977e-05, "loss": 0.0228, "step": 15360 }, { "grad_norm": 0.17397625744342804, "learning_rate": 5.209928940520719e-05, "loss": 0.0285, "step": 15370 }, { "grad_norm": 0.15967972576618195, "learning_rate": 5.204422108090004e-05, "loss": 0.0271, "step": 15380 }, { "grad_norm": 0.19761362671852112, "learning_rate": 5.19891502726715e-05, "loss": 0.0274, "step": 15390 }, { "grad_norm": 0.18547473847866058, "learning_rate": 5.193407704743782e-05, "loss": 0.0309, "step": 15400 }, { "grad_norm": 0.18234094977378845, "learning_rate": 5.187900147211815e-05, "loss": 0.0233, "step": 15410 }, { "grad_norm": 0.26438456773757935, "learning_rate": 5.182392361363453e-05, "loss": 0.0312, "step": 15420 }, { "grad_norm": 0.17008134722709656, "learning_rate": 5.176884353891172e-05, "loss": 0.0468, "step": 15430 }, { "grad_norm": 0.16362731158733368, "learning_rate": 5.171376131487722e-05, "loss": 0.0252, "step": 15440 }, { "grad_norm": 0.1849975883960724, "learning_rate": 5.165867700846113e-05, "loss": 0.0333, "step": 15450 }, { "grad_norm": 0.16272678971290588, "learning_rate": 5.160359068659606e-05, "loss": 0.0273, "step": 15460 }, { "grad_norm": 0.1985752135515213, "learning_rate": 5.154850241621712e-05, "loss": 0.0237, "step": 15470 }, { "grad_norm": 0.18709860742092133, "learning_rate": 5.149341226426172e-05, "loss": 0.0321, "step": 15480 }, { "grad_norm": 0.21969223022460938, "learning_rate": 5.14383202976696e-05, "loss": 0.0353, "step": 15490 }, { "grad_norm": 0.19120453298091888, "learning_rate": 5.1383226583382684e-05, "loss": 0.0287, "step": 15500 }, { "grad_norm": 0.17317569255828857, "learning_rate": 5.132813118834504e-05, "loss": 0.0338, "step": 15510 }, { "grad_norm": 0.15216568112373352, "learning_rate": 5.1273034179502777e-05, "loss": 0.021, "step": 15520 }, { "grad_norm": 0.17817220091819763, "learning_rate": 5.1217935623803945e-05, "loss": 0.0238, "step": 15530 }, { "grad_norm": 0.18722756206989288, "learning_rate": 5.1162835588198476e-05, "loss": 0.0277, "step": 15540 }, { "grad_norm": 0.20363815128803253, "learning_rate": 5.110773413963813e-05, "loss": 0.0258, "step": 15550 }, { "grad_norm": 0.14147475361824036, "learning_rate": 5.105263134507636e-05, "loss": 0.031, "step": 15560 }, { "grad_norm": 0.2621809244155884, "learning_rate": 5.099752727146824e-05, "loss": 0.0383, "step": 15570 }, { "grad_norm": 0.19362476468086243, "learning_rate": 5.094242198577042e-05, "loss": 0.0389, "step": 15580 }, { "grad_norm": 0.12420185655355453, "learning_rate": 5.088731555494102e-05, "loss": 0.0259, "step": 15590 }, { "grad_norm": 0.15179501473903656, "learning_rate": 5.0832208045939556e-05, "loss": 0.0242, "step": 15600 }, { "grad_norm": 0.15309332311153412, "learning_rate": 5.0777099525726844e-05, "loss": 0.0317, "step": 15610 }, { "grad_norm": 0.19083943963050842, "learning_rate": 5.072199006126494e-05, "loss": 0.0225, "step": 15620 }, { "grad_norm": 0.2122243493795395, "learning_rate": 5.0666879719517026e-05, "loss": 0.0268, "step": 15630 }, { "grad_norm": 0.17942848801612854, "learning_rate": 5.061176856744737e-05, "loss": 0.0283, "step": 15640 }, { "grad_norm": 0.16964101791381836, "learning_rate": 5.0556656672021205e-05, "loss": 0.0304, "step": 15650 }, { "grad_norm": 0.27288371324539185, "learning_rate": 5.050154410020472e-05, "loss": 0.0352, "step": 15660 }, { "grad_norm": 0.1564224809408188, "learning_rate": 5.044643091896485e-05, "loss": 0.0254, "step": 15670 }, { "grad_norm": 0.12880170345306396, "learning_rate": 5.039131719526932e-05, "loss": 0.0225, "step": 15680 }, { "grad_norm": 0.1853754073381424, "learning_rate": 5.03362029960865e-05, "loss": 0.029, "step": 15690 }, { "grad_norm": 0.21202430129051208, "learning_rate": 5.028108838838533e-05, "loss": 0.0309, "step": 15700 }, { "grad_norm": 0.1619756668806076, "learning_rate": 5.0225973439135276e-05, "loss": 0.0322, "step": 15710 }, { "grad_norm": 0.186361625790596, "learning_rate": 5.017085821530617e-05, "loss": 0.0249, "step": 15720 }, { "grad_norm": 0.224025160074234, "learning_rate": 5.011574278386822e-05, "loss": 0.0409, "step": 15730 }, { "grad_norm": 0.16389955580234528, "learning_rate": 5.006062721179189e-05, "loss": 0.0259, "step": 15740 }, { "grad_norm": 0.1542915254831314, "learning_rate": 5.000551156604777e-05, "loss": 0.0344, "step": 15750 }, { "grad_norm": 0.16380098462104797, "learning_rate": 4.9950395913606594e-05, "loss": 0.0293, "step": 15760 }, { "grad_norm": 0.2256404459476471, "learning_rate": 4.989528032143903e-05, "loss": 0.0316, "step": 15770 }, { "grad_norm": 0.16901950538158417, "learning_rate": 4.984016485651578e-05, "loss": 0.029, "step": 15780 }, { "grad_norm": 0.22757162153720856, "learning_rate": 4.9785049585807274e-05, "loss": 0.0383, "step": 15790 }, { "grad_norm": 0.19711369276046753, "learning_rate": 4.972993457628381e-05, "loss": 0.0332, "step": 15800 }, { "grad_norm": 0.16102544963359833, "learning_rate": 4.9674819894915306e-05, "loss": 0.0276, "step": 15810 }, { "grad_norm": 0.18006177246570587, "learning_rate": 4.9619705608671265e-05, "loss": 0.0281, "step": 15820 }, { "grad_norm": 0.17261824011802673, "learning_rate": 4.956459178452079e-05, "loss": 0.0277, "step": 15830 }, { "grad_norm": 0.1722591072320938, "learning_rate": 4.950947848943235e-05, "loss": 0.0265, "step": 15840 }, { "grad_norm": 0.20743298530578613, "learning_rate": 4.9454365790373805e-05, "loss": 0.0277, "step": 15850 }, { "grad_norm": 0.159715473651886, "learning_rate": 4.939925375431226e-05, "loss": 0.0341, "step": 15860 }, { "grad_norm": 0.20955756306648254, "learning_rate": 4.9344142448214046e-05, "loss": 0.0372, "step": 15870 }, { "grad_norm": 0.15517686307430267, "learning_rate": 4.928903193904461e-05, "loss": 0.0355, "step": 15880 }, { "grad_norm": 0.16315455734729767, "learning_rate": 4.923392229376842e-05, "loss": 0.0292, "step": 15890 }, { "grad_norm": 0.1228923574090004, "learning_rate": 4.9178813579348865e-05, "loss": 0.0201, "step": 15900 }, { "grad_norm": 0.15196537971496582, "learning_rate": 4.9123705862748253e-05, "loss": 0.0255, "step": 15910 }, { "grad_norm": 0.1537640392780304, "learning_rate": 4.9068599210927627e-05, "loss": 0.0323, "step": 15920 }, { "grad_norm": 0.21058905124664307, "learning_rate": 4.901349369084681e-05, "loss": 0.0355, "step": 15930 }, { "grad_norm": 0.12025801837444305, "learning_rate": 4.8958389369464156e-05, "loss": 0.0264, "step": 15940 }, { "grad_norm": 0.20650848746299744, "learning_rate": 4.890328631373666e-05, "loss": 0.0207, "step": 15950 }, { "grad_norm": 0.17561031877994537, "learning_rate": 4.88481845906197e-05, "loss": 0.0238, "step": 15960 }, { "grad_norm": 0.17418698966503143, "learning_rate": 4.879308426706707e-05, "loss": 0.0361, "step": 15970 }, { "grad_norm": 0.2112257480621338, "learning_rate": 4.873798541003084e-05, "loss": 0.0236, "step": 15980 }, { "grad_norm": 0.15911011397838593, "learning_rate": 4.868288808646136e-05, "loss": 0.038, "step": 15990 }, { "grad_norm": 0.22617973387241364, "learning_rate": 4.862779236330705e-05, "loss": 0.0293, "step": 16000 }, { "grad_norm": 0.15487836301326752, "learning_rate": 4.8572698307514395e-05, "loss": 0.0292, "step": 16010 }, { "grad_norm": 0.24215149879455566, "learning_rate": 4.85176059860279e-05, "loss": 0.0278, "step": 16020 }, { "grad_norm": 0.20623484253883362, "learning_rate": 4.846251546578989e-05, "loss": 0.0345, "step": 16030 }, { "grad_norm": 0.18627691268920898, "learning_rate": 4.840742681374058e-05, "loss": 0.0308, "step": 16040 }, { "grad_norm": 0.19475041329860687, "learning_rate": 4.835234009681787e-05, "loss": 0.027, "step": 16050 }, { "grad_norm": 0.17404621839523315, "learning_rate": 4.8297255381957296e-05, "loss": 0.0252, "step": 16060 }, { "grad_norm": 0.15633514523506165, "learning_rate": 4.824217273609198e-05, "loss": 0.0281, "step": 16070 }, { "grad_norm": 0.16332794725894928, "learning_rate": 4.8187092226152555e-05, "loss": 0.0295, "step": 16080 }, { "grad_norm": 0.20800045132637024, "learning_rate": 4.8132013919067016e-05, "loss": 0.0311, "step": 16090 }, { "grad_norm": 0.17396536469459534, "learning_rate": 4.807693788176071e-05, "loss": 0.0316, "step": 16100 }, { "grad_norm": 0.16399413347244263, "learning_rate": 4.8021864181156214e-05, "loss": 0.0294, "step": 16110 }, { "grad_norm": 0.24819256365299225, "learning_rate": 4.796679288417326e-05, "loss": 0.0394, "step": 16120 }, { "grad_norm": 0.21177524328231812, "learning_rate": 4.791172405772866e-05, "loss": 0.0318, "step": 16130 }, { "grad_norm": 0.17257872223854065, "learning_rate": 4.785665776873625e-05, "loss": 0.0218, "step": 16140 }, { "grad_norm": 0.14141876995563507, "learning_rate": 4.7801594084106763e-05, "loss": 0.0354, "step": 16150 }, { "grad_norm": 0.15414516627788544, "learning_rate": 4.774653307074775e-05, "loss": 0.0325, "step": 16160 }, { "grad_norm": 0.1581634134054184, "learning_rate": 4.769147479556355e-05, "loss": 0.026, "step": 16170 }, { "grad_norm": 0.1281193047761917, "learning_rate": 4.763641932545515e-05, "loss": 0.0189, "step": 16180 }, { "grad_norm": 0.23443517088890076, "learning_rate": 4.758136672732013e-05, "loss": 0.0282, "step": 16190 }, { "grad_norm": 0.2073829621076584, "learning_rate": 4.752631706805261e-05, "loss": 0.0255, "step": 16200 }, { "grad_norm": 0.13341279327869415, "learning_rate": 4.7471270414543114e-05, "loss": 0.037, "step": 16210 }, { "grad_norm": 0.1463790386915207, "learning_rate": 4.741622683367849e-05, "loss": 0.0228, "step": 16220 }, { "grad_norm": 0.1431562751531601, "learning_rate": 4.736118639234191e-05, "loss": 0.0407, "step": 16230 }, { "grad_norm": 0.2260499894618988, "learning_rate": 4.730614915741267e-05, "loss": 0.0326, "step": 16240 }, { "grad_norm": 0.1604788601398468, "learning_rate": 4.7251115195766234e-05, "loss": 0.0267, "step": 16250 }, { "grad_norm": 0.18636147677898407, "learning_rate": 4.719608457427404e-05, "loss": 0.0296, "step": 16260 }, { "grad_norm": 0.1761908382177353, "learning_rate": 4.714105735980347e-05, "loss": 0.0257, "step": 16270 }, { "grad_norm": 0.15951822698116302, "learning_rate": 4.70860336192178e-05, "loss": 0.0321, "step": 16280 }, { "grad_norm": 0.14094442129135132, "learning_rate": 4.7031013419376034e-05, "loss": 0.0277, "step": 16290 }, { "grad_norm": 0.1281169354915619, "learning_rate": 4.6975996827132926e-05, "loss": 0.0194, "step": 16300 }, { "grad_norm": 0.20577271282672882, "learning_rate": 4.692098390933883e-05, "loss": 0.0315, "step": 16310 }, { "grad_norm": 0.19472183287143707, "learning_rate": 4.6865974732839615e-05, "loss": 0.0232, "step": 16320 }, { "grad_norm": 0.1630740910768509, "learning_rate": 4.6810969364476615e-05, "loss": 0.036, "step": 16330 }, { "grad_norm": 0.16056525707244873, "learning_rate": 4.675596787108653e-05, "loss": 0.0257, "step": 16340 }, { "grad_norm": 0.18710577487945557, "learning_rate": 4.670097031950139e-05, "loss": 0.0189, "step": 16350 }, { "grad_norm": 0.16045401990413666, "learning_rate": 4.664597677654838e-05, "loss": 0.0307, "step": 16360 }, { "grad_norm": 0.1509973555803299, "learning_rate": 4.659098730904986e-05, "loss": 0.033, "step": 16370 }, { "grad_norm": 0.18295647203922272, "learning_rate": 4.6536001983823204e-05, "loss": 0.0252, "step": 16380 }, { "grad_norm": 0.17948667705059052, "learning_rate": 4.648102086768078e-05, "loss": 0.0333, "step": 16390 }, { "grad_norm": 0.159071147441864, "learning_rate": 4.6426044027429786e-05, "loss": 0.0287, "step": 16400 }, { "grad_norm": 0.18523085117340088, "learning_rate": 4.637107152987234e-05, "loss": 0.0297, "step": 16410 }, { "grad_norm": 0.23403644561767578, "learning_rate": 4.631610344180515e-05, "loss": 0.0279, "step": 16420 }, { "grad_norm": 0.20961448550224304, "learning_rate": 4.626113983001965e-05, "loss": 0.0293, "step": 16430 }, { "grad_norm": 0.14151686429977417, "learning_rate": 4.620618076130182e-05, "loss": 0.0259, "step": 16440 }, { "grad_norm": 0.1762949675321579, "learning_rate": 4.6151226302432074e-05, "loss": 0.0235, "step": 16450 }, { "grad_norm": 0.1636640727519989, "learning_rate": 4.6096276520185304e-05, "loss": 0.0281, "step": 16460 }, { "grad_norm": 0.2867969572544098, "learning_rate": 4.604133148133066e-05, "loss": 0.0252, "step": 16470 }, { "grad_norm": 0.20009568333625793, "learning_rate": 4.598639125263155e-05, "loss": 0.0325, "step": 16480 }, { "grad_norm": 0.1700575351715088, "learning_rate": 4.593145590084553e-05, "loss": 0.0309, "step": 16490 }, { "grad_norm": 0.2472802996635437, "learning_rate": 4.58765254927242e-05, "loss": 0.0247, "step": 16500 }, { "grad_norm": 0.15075381100177765, "learning_rate": 4.582160009501323e-05, "loss": 0.0239, "step": 16510 }, { "grad_norm": 0.1888672411441803, "learning_rate": 4.5766679774452143e-05, "loss": 0.0357, "step": 16520 }, { "grad_norm": 0.22058872878551483, "learning_rate": 4.571176459777431e-05, "loss": 0.0311, "step": 16530 }, { "grad_norm": 0.18016763031482697, "learning_rate": 4.5656854631706846e-05, "loss": 0.0242, "step": 16540 }, { "grad_norm": 0.3112098276615143, "learning_rate": 4.560194994297054e-05, "loss": 0.0273, "step": 16550 }, { "grad_norm": 0.23655253648757935, "learning_rate": 4.554705059827974e-05, "loss": 0.0375, "step": 16560 }, { "grad_norm": 0.10532896220684052, "learning_rate": 4.5492156664342365e-05, "loss": 0.0268, "step": 16570 }, { "grad_norm": 0.1650882512331009, "learning_rate": 4.5437268207859695e-05, "loss": 0.0349, "step": 16580 }, { "grad_norm": 0.1384940892457962, "learning_rate": 4.538238529552641e-05, "loss": 0.0207, "step": 16590 }, { "grad_norm": 0.1891254037618637, "learning_rate": 4.5327507994030394e-05, "loss": 0.0213, "step": 16600 }, { "grad_norm": 0.13388070464134216, "learning_rate": 4.527263637005273e-05, "loss": 0.0184, "step": 16610 }, { "grad_norm": 0.13386081159114838, "learning_rate": 4.521777049026766e-05, "loss": 0.0188, "step": 16620 }, { "grad_norm": 0.18159665167331696, "learning_rate": 4.5162910421342386e-05, "loss": 0.0219, "step": 16630 }, { "grad_norm": 0.2732070982456207, "learning_rate": 4.510805622993706e-05, "loss": 0.0324, "step": 16640 }, { "grad_norm": 0.13001444935798645, "learning_rate": 4.505320798270467e-05, "loss": 0.0325, "step": 16650 }, { "grad_norm": 0.1268036812543869, "learning_rate": 4.499836574629105e-05, "loss": 0.0282, "step": 16660 }, { "grad_norm": 0.15025870501995087, "learning_rate": 4.494352958733465e-05, "loss": 0.0402, "step": 16670 }, { "grad_norm": 0.2308923751115799, "learning_rate": 4.488869957246663e-05, "loss": 0.0238, "step": 16680 }, { "grad_norm": 0.21403959393501282, "learning_rate": 4.4833875768310575e-05, "loss": 0.0273, "step": 16690 }, { "grad_norm": 0.17247822880744934, "learning_rate": 4.47790582414826e-05, "loss": 0.0245, "step": 16700 }, { "grad_norm": 0.13079269230365753, "learning_rate": 4.472424705859114e-05, "loss": 0.0291, "step": 16710 }, { "grad_norm": 0.19283173978328705, "learning_rate": 4.466944228623701e-05, "loss": 0.0407, "step": 16720 }, { "grad_norm": 0.21732202172279358, "learning_rate": 4.461464399101312e-05, "loss": 0.0357, "step": 16730 }, { "grad_norm": 0.15358823537826538, "learning_rate": 4.45598522395046e-05, "loss": 0.0254, "step": 16740 }, { "grad_norm": 0.1670280396938324, "learning_rate": 4.4505067098288577e-05, "loss": 0.0348, "step": 16750 }, { "grad_norm": 0.1907881200313568, "learning_rate": 4.445028863393417e-05, "loss": 0.022, "step": 16760 }, { "grad_norm": 0.14847977459430695, "learning_rate": 4.439551691300236e-05, "loss": 0.0239, "step": 16770 }, { "grad_norm": 0.2179231196641922, "learning_rate": 4.434075200204599e-05, "loss": 0.0279, "step": 16780 }, { "grad_norm": 0.14085212349891663, "learning_rate": 4.428599396760956e-05, "loss": 0.0222, "step": 16790 }, { "grad_norm": 0.15063989162445068, "learning_rate": 4.423124287622926e-05, "loss": 0.0263, "step": 16800 }, { "grad_norm": 0.1632469892501831, "learning_rate": 4.417649879443282e-05, "loss": 0.0222, "step": 16810 }, { "grad_norm": 0.13959118723869324, "learning_rate": 4.4121761788739444e-05, "loss": 0.0235, "step": 16820 }, { "grad_norm": 0.17851382493972778, "learning_rate": 4.4067031925659805e-05, "loss": 0.0261, "step": 16830 }, { "grad_norm": 0.22051312029361725, "learning_rate": 4.4012309271695816e-05, "loss": 0.0205, "step": 16840 }, { "grad_norm": 0.17327040433883667, "learning_rate": 4.395759389334067e-05, "loss": 0.0249, "step": 16850 }, { "grad_norm": 0.1499096304178238, "learning_rate": 4.3902885857078685e-05, "loss": 0.0312, "step": 16860 }, { "grad_norm": 0.16273732483386993, "learning_rate": 4.384818522938531e-05, "loss": 0.0199, "step": 16870 }, { "grad_norm": 0.15806205570697784, "learning_rate": 4.379349207672695e-05, "loss": 0.0209, "step": 16880 }, { "grad_norm": 0.19559495151042938, "learning_rate": 4.373880646556098e-05, "loss": 0.0293, "step": 16890 }, { "grad_norm": 0.20053917169570923, "learning_rate": 4.3684128462335533e-05, "loss": 0.0203, "step": 16900 }, { "grad_norm": 0.17291860282421112, "learning_rate": 4.362945813348955e-05, "loss": 0.0228, "step": 16910 }, { "grad_norm": 0.1621454507112503, "learning_rate": 4.357479554545262e-05, "loss": 0.0267, "step": 16920 }, { "grad_norm": 0.21902555227279663, "learning_rate": 4.352014076464499e-05, "loss": 0.0306, "step": 16930 }, { "grad_norm": 0.17015263438224792, "learning_rate": 4.346549385747734e-05, "loss": 0.025, "step": 16940 }, { "grad_norm": 0.17932765185832977, "learning_rate": 4.34108548903508e-05, "loss": 0.0385, "step": 16950 }, { "grad_norm": 0.19885432720184326, "learning_rate": 4.335622392965689e-05, "loss": 0.0255, "step": 16960 }, { "grad_norm": 0.16841638088226318, "learning_rate": 4.330160104177738e-05, "loss": 0.0283, "step": 16970 }, { "grad_norm": 0.1437029242515564, "learning_rate": 4.32469862930842e-05, "loss": 0.027, "step": 16980 }, { "grad_norm": 0.22905980050563812, "learning_rate": 4.319237974993946e-05, "loss": 0.033, "step": 16990 }, { "grad_norm": 0.27510014176368713, "learning_rate": 4.3137781478695236e-05, "loss": 0.0314, "step": 17000 }, { "grad_norm": 0.19176854193210602, "learning_rate": 4.308319154569358e-05, "loss": 0.038, "step": 17010 }, { "grad_norm": 0.16606642305850983, "learning_rate": 4.302861001726641e-05, "loss": 0.0394, "step": 17020 }, { "grad_norm": 0.13132554292678833, "learning_rate": 4.2974036959735415e-05, "loss": 0.0195, "step": 17030 }, { "grad_norm": 0.16896776854991913, "learning_rate": 4.2919472439412034e-05, "loss": 0.0209, "step": 17040 }, { "grad_norm": 0.14945730566978455, "learning_rate": 4.286491652259729e-05, "loss": 0.0263, "step": 17050 }, { "grad_norm": 0.2339295595884323, "learning_rate": 4.281036927558178e-05, "loss": 0.0368, "step": 17060 }, { "grad_norm": 0.1317826807498932, "learning_rate": 4.275583076464552e-05, "loss": 0.0214, "step": 17070 }, { "grad_norm": 0.18599583208560944, "learning_rate": 4.270130105605794e-05, "loss": 0.0232, "step": 17080 }, { "grad_norm": 0.14964690804481506, "learning_rate": 4.264678021607782e-05, "loss": 0.0276, "step": 17090 }, { "grad_norm": 0.1664372682571411, "learning_rate": 4.25922683109531e-05, "loss": 0.0219, "step": 17100 }, { "grad_norm": 0.21258482336997986, "learning_rate": 4.25377654069209e-05, "loss": 0.0291, "step": 17110 }, { "grad_norm": 0.1834336221218109, "learning_rate": 4.248327157020737e-05, "loss": 0.0242, "step": 17120 }, { "grad_norm": 0.20884910225868225, "learning_rate": 4.242878686702763e-05, "loss": 0.0232, "step": 17130 }, { "grad_norm": 0.14685863256454468, "learning_rate": 4.23743113635858e-05, "loss": 0.0203, "step": 17140 }, { "grad_norm": 0.18898893892765045, "learning_rate": 4.2319845126074705e-05, "loss": 0.0309, "step": 17150 }, { "grad_norm": 0.21508552134037018, "learning_rate": 4.226538822067598e-05, "loss": 0.0288, "step": 17160 }, { "grad_norm": 0.19032879173755646, "learning_rate": 4.2210940713559896e-05, "loss": 0.0339, "step": 17170 }, { "grad_norm": 0.19031527638435364, "learning_rate": 4.21565026708853e-05, "loss": 0.0235, "step": 17180 }, { "grad_norm": 0.18016071617603302, "learning_rate": 4.2102074158799535e-05, "loss": 0.0213, "step": 17190 }, { "grad_norm": 0.16415634751319885, "learning_rate": 4.204765524343841e-05, "loss": 0.0302, "step": 17200 }, { "grad_norm": 0.16044238209724426, "learning_rate": 4.1993245990926036e-05, "loss": 0.0218, "step": 17210 }, { "grad_norm": 0.15630461275577545, "learning_rate": 4.193884646737474e-05, "loss": 0.0259, "step": 17220 }, { "grad_norm": 0.16304215788841248, "learning_rate": 4.1884456738885124e-05, "loss": 0.0203, "step": 17230 }, { "grad_norm": 0.16586464643478394, "learning_rate": 4.1830076871545795e-05, "loss": 0.023, "step": 17240 }, { "grad_norm": 0.3129137456417084, "learning_rate": 4.177570693143346e-05, "loss": 0.0404, "step": 17250 }, { "grad_norm": 0.14831073582172394, "learning_rate": 4.172134698461271e-05, "loss": 0.022, "step": 17260 }, { "grad_norm": 0.17091797292232513, "learning_rate": 4.166699709713599e-05, "loss": 0.0213, "step": 17270 }, { "grad_norm": 0.1409478783607483, "learning_rate": 4.1612657335043555e-05, "loss": 0.0187, "step": 17280 }, { "grad_norm": 0.15984086692333221, "learning_rate": 4.155832776436331e-05, "loss": 0.0293, "step": 17290 }, { "grad_norm": 0.17393511533737183, "learning_rate": 4.1504008451110845e-05, "loss": 0.0234, "step": 17300 }, { "grad_norm": 0.1451442539691925, "learning_rate": 4.1449699461289225e-05, "loss": 0.0228, "step": 17310 }, { "grad_norm": 0.13682995736598969, "learning_rate": 4.139540086088901e-05, "loss": 0.0169, "step": 17320 }, { "grad_norm": 0.12546406686306, "learning_rate": 4.1341112715888106e-05, "loss": 0.0232, "step": 17330 }, { "grad_norm": 0.16703002154827118, "learning_rate": 4.128683509225172e-05, "loss": 0.027, "step": 17340 }, { "grad_norm": 0.1445690542459488, "learning_rate": 4.123256805593231e-05, "loss": 0.0213, "step": 17350 }, { "grad_norm": 0.13827911019325256, "learning_rate": 4.117831167286943e-05, "loss": 0.0297, "step": 17360 }, { "grad_norm": 0.21285881102085114, "learning_rate": 4.1124066008989685e-05, "loss": 0.0274, "step": 17370 }, { "grad_norm": 0.13088181614875793, "learning_rate": 4.106983113020669e-05, "loss": 0.0213, "step": 17380 }, { "grad_norm": 0.15594898164272308, "learning_rate": 4.101560710242094e-05, "loss": 0.0213, "step": 17390 }, { "grad_norm": 0.19192615151405334, "learning_rate": 4.096139399151971e-05, "loss": 0.0292, "step": 17400 }, { "grad_norm": 0.13994117081165314, "learning_rate": 4.090719186337709e-05, "loss": 0.0165, "step": 17410 }, { "grad_norm": 0.19912520051002502, "learning_rate": 4.0853000783853746e-05, "loss": 0.0284, "step": 17420 }, { "grad_norm": 0.12665323913097382, "learning_rate": 4.0798820818796956e-05, "loss": 0.022, "step": 17430 }, { "grad_norm": 0.21452359855175018, "learning_rate": 4.074465203404048e-05, "loss": 0.0301, "step": 17440 }, { "grad_norm": 0.18347513675689697, "learning_rate": 4.0690494495404495e-05, "loss": 0.0264, "step": 17450 }, { "grad_norm": 0.117930106818676, "learning_rate": 4.0636348268695535e-05, "loss": 0.0228, "step": 17460 }, { "grad_norm": 0.12115605175495148, "learning_rate": 4.058221341970637e-05, "loss": 0.021, "step": 17470 }, { "grad_norm": 0.19469726085662842, "learning_rate": 4.0528090014215945e-05, "loss": 0.0247, "step": 17480 }, { "grad_norm": 0.1699804812669754, "learning_rate": 4.047397811798929e-05, "loss": 0.0231, "step": 17490 }, { "grad_norm": 0.13194504380226135, "learning_rate": 4.041987779677745e-05, "loss": 0.0213, "step": 17500 }, { "grad_norm": 0.22040104866027832, "learning_rate": 4.036578911631745e-05, "loss": 0.0278, "step": 17510 }, { "grad_norm": 0.14719656109809875, "learning_rate": 4.031171214233211e-05, "loss": 0.0269, "step": 17520 }, { "grad_norm": 0.14870822429656982, "learning_rate": 4.025764694053008e-05, "loss": 0.0249, "step": 17530 }, { "grad_norm": 0.16434639692306519, "learning_rate": 4.020359357660566e-05, "loss": 0.0333, "step": 17540 }, { "grad_norm": 0.16411030292510986, "learning_rate": 4.014955211623875e-05, "loss": 0.0216, "step": 17550 }, { "grad_norm": 0.16052471101284027, "learning_rate": 4.0095522625094895e-05, "loss": 0.0386, "step": 17560 }, { "grad_norm": 0.17466910183429718, "learning_rate": 4.004150516882497e-05, "loss": 0.0326, "step": 17570 }, { "grad_norm": 0.20175005495548248, "learning_rate": 3.99874998130653e-05, "loss": 0.0289, "step": 17580 }, { "grad_norm": 0.18267810344696045, "learning_rate": 3.9933506623437455e-05, "loss": 0.0341, "step": 17590 }, { "grad_norm": 0.15119655430316925, "learning_rate": 3.9879525665548276e-05, "loss": 0.0402, "step": 17600 }, { "grad_norm": 0.17913520336151123, "learning_rate": 3.9825557004989704e-05, "loss": 0.0323, "step": 17610 }, { "grad_norm": 0.1616043746471405, "learning_rate": 3.977160070733877e-05, "loss": 0.0263, "step": 17620 }, { "grad_norm": 0.16602373123168945, "learning_rate": 3.971765683815746e-05, "loss": 0.0216, "step": 17630 }, { "grad_norm": 0.15362051129341125, "learning_rate": 3.966372546299265e-05, "loss": 0.0286, "step": 17640 }, { "grad_norm": 0.23251408338546753, "learning_rate": 3.960980664737604e-05, "loss": 0.0233, "step": 17650 }, { "grad_norm": 0.20657192170619965, "learning_rate": 3.955590045682408e-05, "loss": 0.0382, "step": 17660 }, { "grad_norm": 0.15085983276367188, "learning_rate": 3.950200695683788e-05, "loss": 0.0296, "step": 17670 }, { "grad_norm": 0.25893867015838623, "learning_rate": 3.944812621290314e-05, "loss": 0.0329, "step": 17680 }, { "grad_norm": 0.13914451003074646, "learning_rate": 3.939425829049002e-05, "loss": 0.027, "step": 17690 }, { "grad_norm": 0.14555813372135162, "learning_rate": 3.9340403255053135e-05, "loss": 0.028, "step": 17700 }, { "grad_norm": 0.2418401688337326, "learning_rate": 3.9286561172031407e-05, "loss": 0.0312, "step": 17710 }, { "grad_norm": 0.14400777220726013, "learning_rate": 3.923273210684809e-05, "loss": 0.0264, "step": 17720 }, { "grad_norm": 0.13172689080238342, "learning_rate": 3.9178916124910555e-05, "loss": 0.0234, "step": 17730 }, { "grad_norm": 0.19687415659427643, "learning_rate": 3.9125113291610274e-05, "loss": 0.0353, "step": 17740 }, { "grad_norm": 0.20358501374721527, "learning_rate": 3.907132367232278e-05, "loss": 0.0415, "step": 17750 }, { "grad_norm": 0.17561247944831848, "learning_rate": 3.901754733240752e-05, "loss": 0.0329, "step": 17760 }, { "grad_norm": 0.2149197906255722, "learning_rate": 3.896378433720785e-05, "loss": 0.0211, "step": 17770 }, { "grad_norm": 0.1562180519104004, "learning_rate": 3.891003475205086e-05, "loss": 0.0339, "step": 17780 }, { "grad_norm": 0.14612264931201935, "learning_rate": 3.885629864224736e-05, "loss": 0.0157, "step": 17790 }, { "grad_norm": 0.16199229657649994, "learning_rate": 3.880257607309178e-05, "loss": 0.019, "step": 17800 }, { "grad_norm": 0.1995002031326294, "learning_rate": 3.874886710986213e-05, "loss": 0.0322, "step": 17810 }, { "grad_norm": 0.1794530600309372, "learning_rate": 3.8695171817819834e-05, "loss": 0.0274, "step": 17820 }, { "grad_norm": 0.155242457985878, "learning_rate": 3.8641490262209774e-05, "loss": 0.0243, "step": 17830 }, { "grad_norm": 0.1399874985218048, "learning_rate": 3.8587822508260084e-05, "loss": 0.0212, "step": 17840 }, { "grad_norm": 0.23274807631969452, "learning_rate": 3.853416862118214e-05, "loss": 0.0237, "step": 17850 }, { "grad_norm": 0.12246403843164444, "learning_rate": 3.848052866617049e-05, "loss": 0.0273, "step": 17860 }, { "grad_norm": 0.16975194215774536, "learning_rate": 3.84269027084027e-05, "loss": 0.0248, "step": 17870 }, { "grad_norm": 0.18328401446342468, "learning_rate": 3.83732908130394e-05, "loss": 0.0206, "step": 17880 }, { "grad_norm": 0.1475679576396942, "learning_rate": 3.831969304522409e-05, "loss": 0.0267, "step": 17890 }, { "grad_norm": 0.19079506397247314, "learning_rate": 3.826610947008312e-05, "loss": 0.0202, "step": 17900 }, { "grad_norm": 0.22402462363243103, "learning_rate": 3.821254015272559e-05, "loss": 0.0286, "step": 17910 }, { "grad_norm": 0.11997716128826141, "learning_rate": 3.8158985158243214e-05, "loss": 0.0211, "step": 17920 }, { "grad_norm": 0.1305195540189743, "learning_rate": 3.810544455171044e-05, "loss": 0.0262, "step": 17930 }, { "grad_norm": 0.17184890806674957, "learning_rate": 3.8051918398184114e-05, "loss": 0.0228, "step": 17940 }, { "grad_norm": 0.14496375620365143, "learning_rate": 3.799840676270356e-05, "loss": 0.0307, "step": 17950 }, { "grad_norm": 0.1810416579246521, "learning_rate": 3.794490971029048e-05, "loss": 0.0269, "step": 17960 }, { "grad_norm": 0.1426318734884262, "learning_rate": 3.789142730594881e-05, "loss": 0.02, "step": 17970 }, { "grad_norm": 0.13621002435684204, "learning_rate": 3.783795961466471e-05, "loss": 0.0365, "step": 17980 }, { "grad_norm": 0.2016277313232422, "learning_rate": 3.778450670140651e-05, "loss": 0.0185, "step": 17990 }, { "grad_norm": 0.16420386731624603, "learning_rate": 3.773106863112451e-05, "loss": 0.0291, "step": 18000 }, { "grad_norm": 0.1628560721874237, "learning_rate": 3.7677645468751e-05, "loss": 0.0372, "step": 18010 }, { "grad_norm": 0.1833002269268036, "learning_rate": 3.762423727920018e-05, "loss": 0.0299, "step": 18020 }, { "grad_norm": 0.11595755815505981, "learning_rate": 3.7570844127367995e-05, "loss": 0.0327, "step": 18030 }, { "grad_norm": 0.1749364733695984, "learning_rate": 3.751746607813221e-05, "loss": 0.0303, "step": 18040 }, { "grad_norm": 0.17241103947162628, "learning_rate": 3.746410319635217e-05, "loss": 0.0284, "step": 18050 }, { "grad_norm": 0.17131824791431427, "learning_rate": 3.7410755546868806e-05, "loss": 0.0264, "step": 18060 }, { "grad_norm": 0.24329139292240143, "learning_rate": 3.735742319450454e-05, "loss": 0.0244, "step": 18070 }, { "grad_norm": 0.18366271257400513, "learning_rate": 3.7304106204063184e-05, "loss": 0.0226, "step": 18080 }, { "grad_norm": 0.15827353298664093, "learning_rate": 3.725080464032996e-05, "loss": 0.0241, "step": 18090 }, { "grad_norm": 0.14490318298339844, "learning_rate": 3.719751856807125e-05, "loss": 0.0346, "step": 18100 }, { "grad_norm": 0.18667484819889069, "learning_rate": 3.71442480520347e-05, "loss": 0.0215, "step": 18110 }, { "grad_norm": 0.1924593448638916, "learning_rate": 3.709099315694897e-05, "loss": 0.0261, "step": 18120 }, { "grad_norm": 0.21613511443138123, "learning_rate": 3.703775394752378e-05, "loss": 0.0283, "step": 18130 }, { "grad_norm": 0.11856229603290558, "learning_rate": 3.698453048844983e-05, "loss": 0.0183, "step": 18140 }, { "grad_norm": 0.16210848093032837, "learning_rate": 3.693132284439861e-05, "loss": 0.0228, "step": 18150 }, { "grad_norm": 0.1695358157157898, "learning_rate": 3.6878131080022414e-05, "loss": 0.0249, "step": 18160 }, { "grad_norm": 0.18427854776382446, "learning_rate": 3.682495525995429e-05, "loss": 0.0251, "step": 18170 }, { "grad_norm": 0.1782171130180359, "learning_rate": 3.6771795448807846e-05, "loss": 0.023, "step": 18180 }, { "grad_norm": 0.08838263899087906, "learning_rate": 3.671865171117724e-05, "loss": 0.027, "step": 18190 }, { "grad_norm": 0.12233658134937286, "learning_rate": 3.666552411163718e-05, "loss": 0.0183, "step": 18200 }, { "grad_norm": 0.18663670122623444, "learning_rate": 3.661241271474269e-05, "loss": 0.0238, "step": 18210 }, { "grad_norm": 0.14684736728668213, "learning_rate": 3.6559317585029116e-05, "loss": 0.0179, "step": 18220 }, { "grad_norm": 0.14432795345783234, "learning_rate": 3.650623878701204e-05, "loss": 0.0223, "step": 18230 }, { "grad_norm": 0.18092234432697296, "learning_rate": 3.645317638518721e-05, "loss": 0.0261, "step": 18240 }, { "grad_norm": 0.20206302404403687, "learning_rate": 3.640013044403046e-05, "loss": 0.0297, "step": 18250 }, { "grad_norm": 0.2654450237751007, "learning_rate": 3.634710102799761e-05, "loss": 0.0229, "step": 18260 }, { "grad_norm": 0.151868999004364, "learning_rate": 3.6294088201524394e-05, "loss": 0.028, "step": 18270 }, { "grad_norm": 0.1873083859682083, "learning_rate": 3.62410920290264e-05, "loss": 0.0182, "step": 18280 }, { "grad_norm": 0.17980080842971802, "learning_rate": 3.6188112574898954e-05, "loss": 0.026, "step": 18290 }, { "grad_norm": 0.18606188893318176, "learning_rate": 3.6135149903517115e-05, "loss": 0.0245, "step": 18300 }, { "grad_norm": 0.10925909131765366, "learning_rate": 3.608220407923552e-05, "loss": 0.0248, "step": 18310 }, { "grad_norm": 0.15704317390918732, "learning_rate": 3.602927516638833e-05, "loss": 0.0269, "step": 18320 }, { "grad_norm": 0.1189785748720169, "learning_rate": 3.5976363229289165e-05, "loss": 0.0217, "step": 18330 }, { "grad_norm": 0.18034513294696808, "learning_rate": 3.5923468332231e-05, "loss": 0.0259, "step": 18340 }, { "grad_norm": 0.13018353283405304, "learning_rate": 3.587059053948616e-05, "loss": 0.0224, "step": 18350 }, { "grad_norm": 0.1722467839717865, "learning_rate": 3.5817729915306134e-05, "loss": 0.0218, "step": 18360 }, { "grad_norm": 0.14306074380874634, "learning_rate": 3.5764886523921566e-05, "loss": 0.0194, "step": 18370 }, { "grad_norm": 0.1502135992050171, "learning_rate": 3.571206042954214e-05, "loss": 0.0225, "step": 18380 }, { "grad_norm": 0.10373854637145996, "learning_rate": 3.5659251696356566e-05, "loss": 0.0193, "step": 18390 }, { "grad_norm": 0.1957133412361145, "learning_rate": 3.56064603885324e-05, "loss": 0.0341, "step": 18400 }, { "grad_norm": 0.1799539178609848, "learning_rate": 3.5553686570216115e-05, "loss": 0.0291, "step": 18410 }, { "grad_norm": 0.14451481401920319, "learning_rate": 3.550093030553284e-05, "loss": 0.0292, "step": 18420 }, { "grad_norm": 0.15833961963653564, "learning_rate": 3.544819165858642e-05, "loss": 0.0212, "step": 18430 }, { "grad_norm": 0.167887881398201, "learning_rate": 3.539547069345926e-05, "loss": 0.0188, "step": 18440 }, { "grad_norm": 0.14814037084579468, "learning_rate": 3.5342767474212344e-05, "loss": 0.0294, "step": 18450 }, { "grad_norm": 0.23323796689510345, "learning_rate": 3.529008206488502e-05, "loss": 0.0256, "step": 18460 }, { "grad_norm": 0.223677858710289, "learning_rate": 3.5237414529495055e-05, "loss": 0.0313, "step": 18470 }, { "grad_norm": 0.11891154944896698, "learning_rate": 3.5184764932038454e-05, "loss": 0.0248, "step": 18480 }, { "grad_norm": 0.20489263534545898, "learning_rate": 3.513213333648945e-05, "loss": 0.0173, "step": 18490 }, { "grad_norm": 0.1560223549604416, "learning_rate": 3.5079519806800374e-05, "loss": 0.0315, "step": 18500 }, { "grad_norm": 0.15365861356258392, "learning_rate": 3.502692440690165e-05, "loss": 0.0241, "step": 18510 }, { "grad_norm": 0.21275393664836884, "learning_rate": 3.497434720070165e-05, "loss": 0.022, "step": 18520 }, { "grad_norm": 0.17309117317199707, "learning_rate": 3.4921788252086616e-05, "loss": 0.029, "step": 18530 }, { "grad_norm": 0.1138497143983841, "learning_rate": 3.486924762492065e-05, "loss": 0.0185, "step": 18540 }, { "grad_norm": 0.11253515630960464, "learning_rate": 3.4816725383045534e-05, "loss": 0.0235, "step": 18550 }, { "grad_norm": 0.11933323740959167, "learning_rate": 3.476422159028079e-05, "loss": 0.0385, "step": 18560 }, { "grad_norm": 0.1655617207288742, "learning_rate": 3.471173631042345e-05, "loss": 0.0228, "step": 18570 }, { "grad_norm": 0.1685613989830017, "learning_rate": 3.465926960724808e-05, "loss": 0.0203, "step": 18580 }, { "grad_norm": 0.1398773342370987, "learning_rate": 3.460682154450666e-05, "loss": 0.0212, "step": 18590 }, { "grad_norm": 0.22249695658683777, "learning_rate": 3.4554392185928564e-05, "loss": 0.0381, "step": 18600 }, { "grad_norm": 0.16222411394119263, "learning_rate": 3.450198159522037e-05, "loss": 0.0211, "step": 18610 }, { "grad_norm": 0.12296365201473236, "learning_rate": 3.444958983606592e-05, "loss": 0.0341, "step": 18620 }, { "grad_norm": 0.23759505152702332, "learning_rate": 3.439721697212612e-05, "loss": 0.0527, "step": 18630 }, { "grad_norm": 0.17970189452171326, "learning_rate": 3.4344863067038954e-05, "loss": 0.025, "step": 18640 }, { "grad_norm": 0.158095583319664, "learning_rate": 3.429252818441935e-05, "loss": 0.0214, "step": 18650 }, { "grad_norm": 0.18040098249912262, "learning_rate": 3.4240212387859094e-05, "loss": 0.0269, "step": 18660 }, { "grad_norm": 0.1593809425830841, "learning_rate": 3.418791574092686e-05, "loss": 0.0229, "step": 18670 }, { "grad_norm": 0.1647575944662094, "learning_rate": 3.413563830716796e-05, "loss": 0.0254, "step": 18680 }, { "grad_norm": 0.12763381004333496, "learning_rate": 3.408338015010445e-05, "loss": 0.0217, "step": 18690 }, { "grad_norm": 0.14211711287498474, "learning_rate": 3.40311413332349e-05, "loss": 0.0192, "step": 18700 }, { "grad_norm": 0.13089966773986816, "learning_rate": 3.3978921920034365e-05, "loss": 0.0207, "step": 18710 }, { "grad_norm": 0.11268056184053421, "learning_rate": 3.392672197395441e-05, "loss": 0.0147, "step": 18720 }, { "grad_norm": 0.1731192022562027, "learning_rate": 3.387454155842287e-05, "loss": 0.0502, "step": 18730 }, { "grad_norm": 0.14089438319206238, "learning_rate": 3.382238073684386e-05, "loss": 0.0174, "step": 18740 }, { "grad_norm": 0.18928129971027374, "learning_rate": 3.377023957259771e-05, "loss": 0.0166, "step": 18750 }, { "grad_norm": 0.19431087374687195, "learning_rate": 3.3718118129040835e-05, "loss": 0.0221, "step": 18760 }, { "grad_norm": 0.12010464072227478, "learning_rate": 3.3666016469505724e-05, "loss": 0.0375, "step": 18770 }, { "grad_norm": 0.22619199752807617, "learning_rate": 3.361393465730079e-05, "loss": 0.0334, "step": 18780 }, { "grad_norm": 0.1857004463672638, "learning_rate": 3.3561872755710366e-05, "loss": 0.0259, "step": 18790 }, { "grad_norm": 0.18373975157737732, "learning_rate": 3.350983082799456e-05, "loss": 0.02, "step": 18800 }, { "grad_norm": 0.1058323085308075, "learning_rate": 3.34578089373892e-05, "loss": 0.0251, "step": 18810 }, { "grad_norm": 0.17902441322803497, "learning_rate": 3.340580714710581e-05, "loss": 0.0274, "step": 18820 }, { "grad_norm": 0.11939459294080734, "learning_rate": 3.3353825520331466e-05, "loss": 0.021, "step": 18830 }, { "grad_norm": 0.15240921080112457, "learning_rate": 3.330186412022876e-05, "loss": 0.0227, "step": 18840 }, { "grad_norm": 0.1627514660358429, "learning_rate": 3.324992300993568e-05, "loss": 0.0233, "step": 18850 }, { "grad_norm": 0.16863752901554108, "learning_rate": 3.319800225256556e-05, "loss": 0.0287, "step": 18860 }, { "grad_norm": 0.1545286774635315, "learning_rate": 3.314610191120702e-05, "loss": 0.0261, "step": 18870 }, { "grad_norm": 0.12654347717761993, "learning_rate": 3.30942220489239e-05, "loss": 0.0213, "step": 18880 }, { "grad_norm": 0.16069471836090088, "learning_rate": 3.3042362728755086e-05, "loss": 0.0197, "step": 18890 }, { "grad_norm": 0.1308940351009369, "learning_rate": 3.299052401371456e-05, "loss": 0.0395, "step": 18900 }, { "grad_norm": 0.14129002392292023, "learning_rate": 3.293870596679125e-05, "loss": 0.0164, "step": 18910 }, { "grad_norm": 0.1646890491247177, "learning_rate": 3.288690865094895e-05, "loss": 0.0211, "step": 18920 }, { "grad_norm": 0.13675245642662048, "learning_rate": 3.283513212912632e-05, "loss": 0.0278, "step": 18930 }, { "grad_norm": 0.14048780500888824, "learning_rate": 3.2783376464236684e-05, "loss": 0.0186, "step": 18940 }, { "grad_norm": 0.11824901401996613, "learning_rate": 3.273164171916806e-05, "loss": 0.0163, "step": 18950 }, { "grad_norm": 0.135278582572937, "learning_rate": 3.267992795678306e-05, "loss": 0.0263, "step": 18960 }, { "grad_norm": 0.17301444709300995, "learning_rate": 3.2628235239918744e-05, "loss": 0.0255, "step": 18970 }, { "grad_norm": 0.16236703097820282, "learning_rate": 3.2576563631386695e-05, "loss": 0.024, "step": 18980 }, { "grad_norm": 0.15360984206199646, "learning_rate": 3.252491319397275e-05, "loss": 0.0204, "step": 18990 }, { "grad_norm": 0.17739365994930267, "learning_rate": 3.247328399043706e-05, "loss": 0.0294, "step": 19000 }, { "grad_norm": 0.137454092502594, "learning_rate": 3.242167608351399e-05, "loss": 0.0267, "step": 19010 }, { "grad_norm": 0.17715775966644287, "learning_rate": 3.2370089535911986e-05, "loss": 0.025, "step": 19020 }, { "grad_norm": 0.1424257755279541, "learning_rate": 3.23185244103136e-05, "loss": 0.0181, "step": 19030 }, { "grad_norm": 0.13088086247444153, "learning_rate": 3.22669807693753e-05, "loss": 0.0297, "step": 19040 }, { "grad_norm": 0.1950674057006836, "learning_rate": 3.2215458675727495e-05, "loss": 0.0265, "step": 19050 }, { "grad_norm": 0.19069045782089233, "learning_rate": 3.216395819197438e-05, "loss": 0.0232, "step": 19060 }, { "grad_norm": 0.11797426640987396, "learning_rate": 3.211247938069387e-05, "loss": 0.0304, "step": 19070 }, { "grad_norm": 0.18977919220924377, "learning_rate": 3.206102230443759e-05, "loss": 0.0222, "step": 19080 }, { "grad_norm": 0.14537763595581055, "learning_rate": 3.2009587025730764e-05, "loss": 0.0317, "step": 19090 }, { "grad_norm": 0.15562433004379272, "learning_rate": 3.195817360707207e-05, "loss": 0.0345, "step": 19100 }, { "grad_norm": 0.11960604041814804, "learning_rate": 3.19067821109337e-05, "loss": 0.0232, "step": 19110 }, { "grad_norm": 0.1511765420436859, "learning_rate": 3.185541259976114e-05, "loss": 0.0176, "step": 19120 }, { "grad_norm": 0.14871153235435486, "learning_rate": 3.180406513597316e-05, "loss": 0.0325, "step": 19130 }, { "grad_norm": 0.1625516563653946, "learning_rate": 3.1752739781961835e-05, "loss": 0.0189, "step": 19140 }, { "grad_norm": 0.11996357142925262, "learning_rate": 3.170143660009228e-05, "loss": 0.0121, "step": 19150 }, { "grad_norm": 0.15694525837898254, "learning_rate": 3.16501556527027e-05, "loss": 0.0259, "step": 19160 }, { "grad_norm": 0.1491648554801941, "learning_rate": 3.1598897002104265e-05, "loss": 0.0186, "step": 19170 }, { "grad_norm": 0.11707095056772232, "learning_rate": 3.154766071058108e-05, "loss": 0.0227, "step": 19180 }, { "grad_norm": 0.13991980254650116, "learning_rate": 3.149644684039008e-05, "loss": 0.0368, "step": 19190 }, { "grad_norm": 0.15939636528491974, "learning_rate": 3.144525545376095e-05, "loss": 0.0229, "step": 19200 }, { "grad_norm": 0.15413333475589752, "learning_rate": 3.139408661289603e-05, "loss": 0.0191, "step": 19210 }, { "grad_norm": 0.21792426705360413, "learning_rate": 3.134294037997032e-05, "loss": 0.0193, "step": 19220 }, { "grad_norm": 0.16204224526882172, "learning_rate": 3.129181681713127e-05, "loss": 0.0246, "step": 19230 }, { "grad_norm": 0.1580965518951416, "learning_rate": 3.1240715986498855e-05, "loss": 0.0233, "step": 19240 }, { "grad_norm": 0.13536816835403442, "learning_rate": 3.1189637950165394e-05, "loss": 0.0219, "step": 19250 }, { "grad_norm": 0.1415414810180664, "learning_rate": 3.1138582770195544e-05, "loss": 0.0174, "step": 19260 }, { "grad_norm": 0.10943274199962616, "learning_rate": 3.108755050862615e-05, "loss": 0.0158, "step": 19270 }, { "grad_norm": 0.17461733520030975, "learning_rate": 3.1036541227466204e-05, "loss": 0.0229, "step": 19280 }, { "grad_norm": 0.22713017463684082, "learning_rate": 3.0985554988696786e-05, "loss": 0.0347, "step": 19290 }, { "grad_norm": 0.13479343056678772, "learning_rate": 3.093459185427102e-05, "loss": 0.0254, "step": 19300 }, { "grad_norm": 0.17659170925617218, "learning_rate": 3.0883651886113905e-05, "loss": 0.0205, "step": 19310 }, { "grad_norm": 0.13995130360126495, "learning_rate": 3.0832735146122294e-05, "loss": 0.0304, "step": 19320 }, { "grad_norm": 0.13722623884677887, "learning_rate": 3.078184169616485e-05, "loss": 0.0401, "step": 19330 }, { "grad_norm": 0.1958882212638855, "learning_rate": 3.073097159808187e-05, "loss": 0.0423, "step": 19340 }, { "grad_norm": 0.12051232904195786, "learning_rate": 3.068012491368537e-05, "loss": 0.0318, "step": 19350 }, { "grad_norm": 0.11143656075000763, "learning_rate": 3.062930170475885e-05, "loss": 0.016, "step": 19360 }, { "grad_norm": 0.118914395570755, "learning_rate": 3.057850203305729e-05, "loss": 0.0212, "step": 19370 }, { "grad_norm": 0.12956653535366058, "learning_rate": 3.052772596030708e-05, "loss": 0.0286, "step": 19380 }, { "grad_norm": 0.12789255380630493, "learning_rate": 3.0476973548205943e-05, "loss": 0.0269, "step": 19390 }, { "grad_norm": 0.12757675349712372, "learning_rate": 3.0426244858422847e-05, "loss": 0.0213, "step": 19400 }, { "grad_norm": 0.13550017774105072, "learning_rate": 3.0375539952597943e-05, "loss": 0.0193, "step": 19410 }, { "grad_norm": 0.19020958244800568, "learning_rate": 3.0324858892342468e-05, "loss": 0.02, "step": 19420 }, { "grad_norm": 0.2496158331632614, "learning_rate": 3.0274201739238672e-05, "loss": 0.0378, "step": 19430 }, { "grad_norm": 0.11468277126550674, "learning_rate": 3.0223568554839786e-05, "loss": 0.018, "step": 19440 }, { "grad_norm": 0.14433610439300537, "learning_rate": 3.0172959400669886e-05, "loss": 0.0196, "step": 19450 }, { "grad_norm": 0.1439269334077835, "learning_rate": 3.0122374338223902e-05, "loss": 0.021, "step": 19460 }, { "grad_norm": 0.12016113847494125, "learning_rate": 3.0071813428967427e-05, "loss": 0.0306, "step": 19470 }, { "grad_norm": 0.14949537813663483, "learning_rate": 3.0021276734336746e-05, "loss": 0.0178, "step": 19480 }, { "grad_norm": 0.1447705328464508, "learning_rate": 2.997076431573871e-05, "loss": 0.0343, "step": 19490 }, { "grad_norm": 0.14605894684791565, "learning_rate": 2.9920276234550636e-05, "loss": 0.0162, "step": 19500 }, { "grad_norm": 0.1835075318813324, "learning_rate": 2.986981255212035e-05, "loss": 0.0363, "step": 19510 }, { "grad_norm": 0.14121422171592712, "learning_rate": 2.9819373329765977e-05, "loss": 0.0165, "step": 19520 }, { "grad_norm": 0.122889444231987, "learning_rate": 2.9768958628775902e-05, "loss": 0.0185, "step": 19530 }, { "grad_norm": 0.13280485570430756, "learning_rate": 2.9718568510408762e-05, "loss": 0.0192, "step": 19540 }, { "grad_norm": 0.14454151690006256, "learning_rate": 2.9668203035893272e-05, "loss": 0.0229, "step": 19550 }, { "grad_norm": 0.10855939984321594, "learning_rate": 2.9617862266428287e-05, "loss": 0.03, "step": 19560 }, { "grad_norm": 0.1231502965092659, "learning_rate": 2.9567546263182556e-05, "loss": 0.0236, "step": 19570 }, { "grad_norm": 0.17368441820144653, "learning_rate": 2.951725508729476e-05, "loss": 0.025, "step": 19580 }, { "grad_norm": 0.1703839898109436, "learning_rate": 2.946698879987344e-05, "loss": 0.0221, "step": 19590 }, { "grad_norm": 0.17846380174160004, "learning_rate": 2.9416747461996853e-05, "loss": 0.0267, "step": 19600 }, { "grad_norm": 0.12962299585342407, "learning_rate": 2.9366531134712973e-05, "loss": 0.0242, "step": 19610 }, { "grad_norm": 0.1317129284143448, "learning_rate": 2.9316339879039367e-05, "loss": 0.0234, "step": 19620 }, { "grad_norm": 0.09920349717140198, "learning_rate": 2.9266173755963167e-05, "loss": 0.0241, "step": 19630 }, { "grad_norm": 0.14105162024497986, "learning_rate": 2.9216032826440926e-05, "loss": 0.0185, "step": 19640 }, { "grad_norm": 0.13886553049087524, "learning_rate": 2.9165917151398592e-05, "loss": 0.0243, "step": 19650 }, { "grad_norm": 0.15728013217449188, "learning_rate": 2.9115826791731428e-05, "loss": 0.0199, "step": 19660 }, { "grad_norm": 0.134812593460083, "learning_rate": 2.906576180830398e-05, "loss": 0.0183, "step": 19670 }, { "grad_norm": 0.17238885164260864, "learning_rate": 2.9015722261949917e-05, "loss": 0.0184, "step": 19680 }, { "grad_norm": 0.11582757532596588, "learning_rate": 2.8965708213471986e-05, "loss": 0.0281, "step": 19690 }, { "grad_norm": 0.17534799873828888, "learning_rate": 2.8915719723641975e-05, "loss": 0.0221, "step": 19700 }, { "grad_norm": 0.1359989196062088, "learning_rate": 2.8865756853200604e-05, "loss": 0.0286, "step": 19710 }, { "grad_norm": 0.20485445857048035, "learning_rate": 2.8815819662857503e-05, "loss": 0.0251, "step": 19720 }, { "grad_norm": 0.15821757912635803, "learning_rate": 2.876590821329105e-05, "loss": 0.0193, "step": 19730 }, { "grad_norm": 0.13033734261989594, "learning_rate": 2.871602256514836e-05, "loss": 0.0217, "step": 19740 }, { "grad_norm": 0.1674693375825882, "learning_rate": 2.8666162779045203e-05, "loss": 0.0149, "step": 19750 }, { "grad_norm": 0.14524652063846588, "learning_rate": 2.8616328915565904e-05, "loss": 0.0307, "step": 19760 }, { "grad_norm": 0.16099661588668823, "learning_rate": 2.856652103526334e-05, "loss": 0.0207, "step": 19770 }, { "grad_norm": 0.12125895172357559, "learning_rate": 2.8516739198658755e-05, "loss": 0.0262, "step": 19780 }, { "grad_norm": 0.15415190160274506, "learning_rate": 2.846698346624177e-05, "loss": 0.0275, "step": 19790 }, { "grad_norm": 0.1308627873659134, "learning_rate": 2.841725389847032e-05, "loss": 0.0204, "step": 19800 }, { "grad_norm": 0.17017002403736115, "learning_rate": 2.8367550555770506e-05, "loss": 0.0218, "step": 19810 }, { "grad_norm": 0.11846572160720825, "learning_rate": 2.831787349853655e-05, "loss": 0.0243, "step": 19820 }, { "grad_norm": 0.11906851828098297, "learning_rate": 2.8268222787130806e-05, "loss": 0.0139, "step": 19830 }, { "grad_norm": 0.12394123524427414, "learning_rate": 2.821859848188355e-05, "loss": 0.0198, "step": 19840 }, { "grad_norm": 0.2048010677099228, "learning_rate": 2.816900064309299e-05, "loss": 0.0371, "step": 19850 }, { "grad_norm": 0.1835079789161682, "learning_rate": 2.811942933102517e-05, "loss": 0.0304, "step": 19860 }, { "grad_norm": 0.17645493149757385, "learning_rate": 2.806988460591391e-05, "loss": 0.017, "step": 19870 }, { "grad_norm": 0.17804789543151855, "learning_rate": 2.802036652796074e-05, "loss": 0.0297, "step": 19880 }, { "grad_norm": 0.12346853315830231, "learning_rate": 2.797087515733478e-05, "loss": 0.0243, "step": 19890 }, { "grad_norm": 0.20005930960178375, "learning_rate": 2.7921410554172723e-05, "loss": 0.0414, "step": 19900 }, { "grad_norm": 0.12366753071546555, "learning_rate": 2.787197277857871e-05, "loss": 0.0228, "step": 19910 }, { "grad_norm": 0.1443571299314499, "learning_rate": 2.782256189062429e-05, "loss": 0.0213, "step": 19920 }, { "grad_norm": 0.1801394522190094, "learning_rate": 2.777317795034839e-05, "loss": 0.0221, "step": 19930 }, { "grad_norm": 0.15387219190597534, "learning_rate": 2.7723821017757112e-05, "loss": 0.0221, "step": 19940 }, { "grad_norm": 0.12056508660316467, "learning_rate": 2.7674491152823822e-05, "loss": 0.0203, "step": 19950 }, { "grad_norm": 0.12686704099178314, "learning_rate": 2.7625188415488944e-05, "loss": 0.0232, "step": 19960 }, { "grad_norm": 0.1879570186138153, "learning_rate": 2.7575912865659924e-05, "loss": 0.0239, "step": 19970 }, { "grad_norm": 0.1918022334575653, "learning_rate": 2.7526664563211245e-05, "loss": 0.024, "step": 19980 }, { "grad_norm": 0.13535237312316895, "learning_rate": 2.7477443567984224e-05, "loss": 0.0233, "step": 19990 }, { "grad_norm": 0.15216320753097534, "learning_rate": 2.7428249939787e-05, "loss": 0.0249, "step": 20000 } ], "logging_steps": 10, "max_steps": 30000, "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": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }