{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 6000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.3615550994873047, "learning_rate": 3e-06, "loss": 1.4557, "step": 10 }, { "grad_norm": 0.8587501049041748, "learning_rate": 6.333333333333334e-06, "loss": 1.424, "step": 20 }, { "grad_norm": 0.3398872911930084, "learning_rate": 9.666666666666667e-06, "loss": 1.3389, "step": 30 }, { "grad_norm": 0.44489237666130066, "learning_rate": 1.3000000000000001e-05, "loss": 1.2756, "step": 40 }, { "grad_norm": 0.49807339906692505, "learning_rate": 1.6333333333333335e-05, "loss": 1.1748, "step": 50 }, { "grad_norm": 0.33505111932754517, "learning_rate": 1.9666666666666666e-05, "loss": 1.11, "step": 60 }, { "grad_norm": 0.37047985196113586, "learning_rate": 2.3000000000000003e-05, "loss": 1.0889, "step": 70 }, { "grad_norm": 0.29498252272605896, "learning_rate": 2.633333333333333e-05, "loss": 1.0758, "step": 80 }, { "grad_norm": 0.3145069181919098, "learning_rate": 2.9666666666666672e-05, "loss": 1.0602, "step": 90 }, { "grad_norm": 0.3874819278717041, "learning_rate": 3.3e-05, "loss": 1.0518, "step": 100 }, { "grad_norm": 0.41450005769729614, "learning_rate": 3.633333333333333e-05, "loss": 1.0352, "step": 110 }, { "grad_norm": 0.5897858738899231, "learning_rate": 3.966666666666667e-05, "loss": 1.0111, "step": 120 }, { "grad_norm": 0.595295786857605, "learning_rate": 4.3e-05, "loss": 0.9768, "step": 130 }, { "grad_norm": 0.4137485921382904, "learning_rate": 4.633333333333333e-05, "loss": 0.9478, "step": 140 }, { "grad_norm": 0.6388778686523438, "learning_rate": 4.966666666666667e-05, "loss": 0.9013, "step": 150 }, { "grad_norm": 0.7980178594589233, "learning_rate": 5.300000000000001e-05, "loss": 0.8418, "step": 160 }, { "grad_norm": 0.6096994280815125, "learning_rate": 5.633333333333334e-05, "loss": 0.7827, "step": 170 }, { "grad_norm": 0.7750433087348938, "learning_rate": 5.966666666666667e-05, "loss": 0.736, "step": 180 }, { "grad_norm": 0.7485987544059753, "learning_rate": 6.3e-05, "loss": 0.6688, "step": 190 }, { "grad_norm": 0.9166963696479797, "learning_rate": 6.633333333333334e-05, "loss": 0.6113, "step": 200 }, { "grad_norm": 1.1108864545822144, "learning_rate": 6.966666666666668e-05, "loss": 0.5597, "step": 210 }, { "grad_norm": 1.0937623977661133, "learning_rate": 7.3e-05, "loss": 0.5089, "step": 220 }, { "grad_norm": 1.240064263343811, "learning_rate": 7.633333333333334e-05, "loss": 0.4596, "step": 230 }, { "grad_norm": 1.176957368850708, "learning_rate": 7.966666666666666e-05, "loss": 0.4209, "step": 240 }, { "grad_norm": 0.8346590399742126, "learning_rate": 8.3e-05, "loss": 0.3922, "step": 250 }, { "grad_norm": 1.5463416576385498, "learning_rate": 8.633333333333334e-05, "loss": 0.3626, "step": 260 }, { "grad_norm": 0.7868688106536865, "learning_rate": 8.966666666666666e-05, "loss": 0.3346, "step": 270 }, { "grad_norm": 1.1313313245773315, "learning_rate": 9.300000000000001e-05, "loss": 0.306, "step": 280 }, { "grad_norm": 1.101473093032837, "learning_rate": 9.633333333333335e-05, "loss": 0.2881, "step": 290 }, { "grad_norm": 1.817094326019287, "learning_rate": 9.966666666666667e-05, "loss": 0.2582, "step": 300 }, { "grad_norm": 1.2253403663635254, "learning_rate": 9.999938485971279e-05, "loss": 0.2284, "step": 310 }, { "grad_norm": 0.8121590614318848, "learning_rate": 9.999725846827562e-05, "loss": 0.2163, "step": 320 }, { "grad_norm": 1.1257294416427612, "learning_rate": 9.999361329594254e-05, "loss": 0.206, "step": 330 }, { "grad_norm": 0.8087614178657532, "learning_rate": 9.998844945344405e-05, "loss": 0.191, "step": 340 }, { "grad_norm": 0.7762327790260315, "learning_rate": 9.99817670976436e-05, "loss": 0.1737, "step": 350 }, { "grad_norm": 1.1138006448745728, "learning_rate": 9.997356643153303e-05, "loss": 0.1529, "step": 360 }, { "grad_norm": 0.8183638453483582, "learning_rate": 9.996384770422629e-05, "loss": 0.1377, "step": 370 }, { "grad_norm": 1.0174615383148193, "learning_rate": 9.995261121095194e-05, "loss": 0.1274, "step": 380 }, { "grad_norm": 0.9137309193611145, "learning_rate": 9.993985729304408e-05, "loss": 0.1228, "step": 390 }, { "grad_norm": 0.9494693875312805, "learning_rate": 9.992558633793212e-05, "loss": 0.1141, "step": 400 }, { "grad_norm": 0.8476019501686096, "learning_rate": 9.990979877912891e-05, "loss": 0.1086, "step": 410 }, { "grad_norm": 0.9416108727455139, "learning_rate": 9.989249509621759e-05, "loss": 0.1122, "step": 420 }, { "grad_norm": 0.8482936024665833, "learning_rate": 9.987367581483705e-05, "loss": 0.1098, "step": 430 }, { "grad_norm": 0.9496634602546692, "learning_rate": 9.985334150666592e-05, "loss": 0.105, "step": 440 }, { "grad_norm": 1.0568591356277466, "learning_rate": 9.983149278940526e-05, "loss": 0.0989, "step": 450 }, { "grad_norm": 1.0233659744262695, "learning_rate": 9.980813032675974e-05, "loss": 0.0874, "step": 460 }, { "grad_norm": 0.8151543140411377, "learning_rate": 9.978325482841753e-05, "loss": 0.0809, "step": 470 }, { "grad_norm": 0.8303743600845337, "learning_rate": 9.975686705002867e-05, "loss": 0.0779, "step": 480 }, { "grad_norm": 0.6682484745979309, "learning_rate": 9.972896779318219e-05, "loss": 0.0726, "step": 490 }, { "grad_norm": 0.8442310690879822, "learning_rate": 9.969955790538175e-05, "loss": 0.0764, "step": 500 }, { "grad_norm": 0.7364897131919861, "learning_rate": 9.966863828001982e-05, "loss": 0.0721, "step": 510 }, { "grad_norm": 0.8933179378509521, "learning_rate": 9.963620985635065e-05, "loss": 0.069, "step": 520 }, { "grad_norm": 0.714433491230011, "learning_rate": 9.960227361946164e-05, "loss": 0.0666, "step": 530 }, { "grad_norm": 0.8506194949150085, "learning_rate": 9.95668306002435e-05, "loss": 0.0678, "step": 540 }, { "grad_norm": 0.7329393625259399, "learning_rate": 9.952988187535886e-05, "loss": 0.0657, "step": 550 }, { "grad_norm": 0.7777057886123657, "learning_rate": 9.949142856720961e-05, "loss": 0.0635, "step": 560 }, { "grad_norm": 0.8185117244720459, "learning_rate": 9.945147184390278e-05, "loss": 0.0697, "step": 570 }, { "grad_norm": 0.7965772151947021, "learning_rate": 9.941001291921512e-05, "loss": 0.0666, "step": 580 }, { "grad_norm": 0.6764131188392639, "learning_rate": 9.936705305255612e-05, "loss": 0.0661, "step": 590 }, { "grad_norm": 0.692888081073761, "learning_rate": 9.932259354892984e-05, "loss": 0.0642, "step": 600 }, { "grad_norm": 0.7117837071418762, "learning_rate": 9.927663575889521e-05, "loss": 0.0616, "step": 610 }, { "grad_norm": 0.774642825126648, "learning_rate": 9.922918107852504e-05, "loss": 0.0631, "step": 620 }, { "grad_norm": 0.8402009606361389, "learning_rate": 9.918023094936363e-05, "loss": 0.0613, "step": 630 }, { "grad_norm": 0.8676945567131042, "learning_rate": 9.912978685838294e-05, "loss": 0.062, "step": 640 }, { "grad_norm": 0.6354483962059021, "learning_rate": 9.90778503379374e-05, "loss": 0.0595, "step": 650 }, { "grad_norm": 0.7966945767402649, "learning_rate": 9.902442296571743e-05, "loss": 0.0573, "step": 660 }, { "grad_norm": 0.8080548644065857, "learning_rate": 9.896950636470147e-05, "loss": 0.0644, "step": 670 }, { "grad_norm": 0.8381810784339905, "learning_rate": 9.891310220310666e-05, "loss": 0.0599, "step": 680 }, { "grad_norm": 0.6608224511146545, "learning_rate": 9.885521219433823e-05, "loss": 0.0557, "step": 690 }, { "grad_norm": 0.6613548994064331, "learning_rate": 9.879583809693738e-05, "loss": 0.0563, "step": 700 }, { "grad_norm": 0.8921074867248535, "learning_rate": 9.873498171452789e-05, "loss": 0.056, "step": 710 }, { "grad_norm": 0.732887327671051, "learning_rate": 9.867264489576135e-05, "loss": 0.0567, "step": 720 }, { "grad_norm": 0.8288407921791077, "learning_rate": 9.860882953426099e-05, "loss": 0.0578, "step": 730 }, { "grad_norm": 0.7719908952713013, "learning_rate": 9.854353756856412e-05, "loss": 0.0567, "step": 740 }, { "grad_norm": 0.7751674056053162, "learning_rate": 9.847677098206332e-05, "loss": 0.0555, "step": 750 }, { "grad_norm": 0.5783160924911499, "learning_rate": 9.840853180294608e-05, "loss": 0.0545, "step": 760 }, { "grad_norm": 0.7684217095375061, "learning_rate": 9.833882210413332e-05, "loss": 0.0568, "step": 770 }, { "grad_norm": 0.7103134393692017, "learning_rate": 9.826764400321633e-05, "loss": 0.0557, "step": 780 }, { "grad_norm": 0.7729565501213074, "learning_rate": 9.819499966239243e-05, "loss": 0.0559, "step": 790 }, { "grad_norm": 0.7751089334487915, "learning_rate": 9.812089128839938e-05, "loss": 0.0579, "step": 800 }, { "grad_norm": 0.6948894262313843, "learning_rate": 9.804532113244828e-05, "loss": 0.0526, "step": 810 }, { "grad_norm": 0.6186414957046509, "learning_rate": 9.796829149015517e-05, "loss": 0.0536, "step": 820 }, { "grad_norm": 0.7275042533874512, "learning_rate": 9.788980470147132e-05, "loss": 0.0517, "step": 830 }, { "grad_norm": 0.7078156471252441, "learning_rate": 9.780986315061218e-05, "loss": 0.0532, "step": 840 }, { "grad_norm": 0.676052987575531, "learning_rate": 9.772846926598491e-05, "loss": 0.0579, "step": 850 }, { "grad_norm": 0.8566974401473999, "learning_rate": 9.76456255201146e-05, "loss": 0.0543, "step": 860 }, { "grad_norm": 0.6115126609802246, "learning_rate": 9.756133442956923e-05, "loss": 0.0523, "step": 870 }, { "grad_norm": 0.6624767780303955, "learning_rate": 9.747559855488313e-05, "loss": 0.0526, "step": 880 }, { "grad_norm": 0.6324094533920288, "learning_rate": 9.73884205004793e-05, "loss": 0.0532, "step": 890 }, { "grad_norm": 0.7360411286354065, "learning_rate": 9.729980291459019e-05, "loss": 0.0519, "step": 900 }, { "grad_norm": 0.6209548115730286, "learning_rate": 9.720974848917735e-05, "loss": 0.05, "step": 910 }, { "grad_norm": 0.7278928160667419, "learning_rate": 9.711825995984957e-05, "loss": 0.0506, "step": 920 }, { "grad_norm": 0.7817022800445557, "learning_rate": 9.702534010577991e-05, "loss": 0.0515, "step": 930 }, { "grad_norm": 0.6537401080131531, "learning_rate": 9.693099174962103e-05, "loss": 0.0511, "step": 940 }, { "grad_norm": 0.6222761273384094, "learning_rate": 9.683521775741977e-05, "loss": 0.0508, "step": 950 }, { "grad_norm": 0.6970085501670837, "learning_rate": 9.673802103852979e-05, "loss": 0.0489, "step": 960 }, { "grad_norm": 0.5753750205039978, "learning_rate": 9.663940454552342e-05, "loss": 0.0498, "step": 970 }, { "grad_norm": 0.6854166388511658, "learning_rate": 9.65393712741018e-05, "loss": 0.0476, "step": 980 }, { "grad_norm": 0.7109822034835815, "learning_rate": 9.6437924263004e-05, "loss": 0.0493, "step": 990 }, { "grad_norm": 0.6544412970542908, "learning_rate": 9.63350665939146e-05, "loss": 0.0482, "step": 1000 }, { "grad_norm": 0.8334299921989441, "learning_rate": 9.623080139137023e-05, "loss": 0.0475, "step": 1010 }, { "grad_norm": 0.5907924175262451, "learning_rate": 9.612513182266447e-05, "loss": 0.0493, "step": 1020 }, { "grad_norm": 0.6483442783355713, "learning_rate": 9.601806109775179e-05, "loss": 0.0502, "step": 1030 }, { "grad_norm": 0.6074472665786743, "learning_rate": 9.590959246914995e-05, "loss": 0.0474, "step": 1040 }, { "grad_norm": 0.6958264112472534, "learning_rate": 9.579972923184122e-05, "loss": 0.0488, "step": 1050 }, { "grad_norm": 0.5658279061317444, "learning_rate": 9.568847472317232e-05, "loss": 0.0454, "step": 1060 }, { "grad_norm": 0.6736823916435242, "learning_rate": 9.557583232275303e-05, "loss": 0.0474, "step": 1070 }, { "grad_norm": 0.6027989983558655, "learning_rate": 9.546180545235344e-05, "loss": 0.0445, "step": 1080 }, { "grad_norm": 0.591834545135498, "learning_rate": 9.534639757580013e-05, "loss": 0.044, "step": 1090 }, { "grad_norm": 0.5577622056007385, "learning_rate": 9.522961219887092e-05, "loss": 0.0459, "step": 1100 }, { "grad_norm": 0.5711413621902466, "learning_rate": 9.511145286918828e-05, "loss": 0.0478, "step": 1110 }, { "grad_norm": 0.6266237497329712, "learning_rate": 9.499192317611167e-05, "loss": 0.0443, "step": 1120 }, { "grad_norm": 0.6088590621948242, "learning_rate": 9.487102675062851e-05, "loss": 0.0469, "step": 1130 }, { "grad_norm": 0.5997032523155212, "learning_rate": 9.474876726524374e-05, "loss": 0.0434, "step": 1140 }, { "grad_norm": 0.6687015891075134, "learning_rate": 9.462514843386845e-05, "loss": 0.0448, "step": 1150 }, { "grad_norm": 0.5682821869850159, "learning_rate": 9.450017401170689e-05, "loss": 0.0465, "step": 1160 }, { "grad_norm": 0.7369022369384766, "learning_rate": 9.437384779514256e-05, "loss": 0.0461, "step": 1170 }, { "grad_norm": 0.5957316756248474, "learning_rate": 9.424617362162271e-05, "loss": 0.0469, "step": 1180 }, { "grad_norm": 0.628410279750824, "learning_rate": 9.411715536954196e-05, "loss": 0.0441, "step": 1190 }, { "grad_norm": 0.6107819080352783, "learning_rate": 9.39867969581243e-05, "loss": 0.0443, "step": 1200 }, { "grad_norm": 0.6843348741531372, "learning_rate": 9.385510234730415e-05, "loss": 0.0447, "step": 1210 }, { "grad_norm": 0.7711231708526611, "learning_rate": 9.372207553760603e-05, "loss": 0.0436, "step": 1220 }, { "grad_norm": 0.5622444748878479, "learning_rate": 9.358772057002312e-05, "loss": 0.0426, "step": 1230 }, { "grad_norm": 0.6071102619171143, "learning_rate": 9.345204152589428e-05, "loss": 0.0428, "step": 1240 }, { "grad_norm": 0.57755446434021, "learning_rate": 9.331504252678037e-05, "loss": 0.0458, "step": 1250 }, { "grad_norm": 0.5819398760795593, "learning_rate": 9.317672773433876e-05, "loss": 0.0439, "step": 1260 }, { "grad_norm": 0.5703418254852295, "learning_rate": 9.30371013501972e-05, "loss": 0.044, "step": 1270 }, { "grad_norm": 0.6551567912101746, "learning_rate": 9.289616761582587e-05, "loss": 0.046, "step": 1280 }, { "grad_norm": 0.7065009474754333, "learning_rate": 9.275393081240882e-05, "loss": 0.0425, "step": 1290 }, { "grad_norm": 0.6561137437820435, "learning_rate": 9.261039526071374e-05, "loss": 0.0452, "step": 1300 }, { "grad_norm": 0.5200664401054382, "learning_rate": 9.246556532096078e-05, "loss": 0.0429, "step": 1310 }, { "grad_norm": 0.6974889039993286, "learning_rate": 9.231944539269009e-05, "loss": 0.0429, "step": 1320 }, { "grad_norm": 0.6593557000160217, "learning_rate": 9.217203991462815e-05, "loss": 0.0436, "step": 1330 }, { "grad_norm": 0.6332272291183472, "learning_rate": 9.202335336455296e-05, "loss": 0.0421, "step": 1340 }, { "grad_norm": 0.511134147644043, "learning_rate": 9.187339025915802e-05, "loss": 0.0417, "step": 1350 }, { "grad_norm": 0.5553621649742126, "learning_rate": 9.17221551539151e-05, "loss": 0.0399, "step": 1360 }, { "grad_norm": 0.5665884017944336, "learning_rate": 9.156965264293586e-05, "loss": 0.0414, "step": 1370 }, { "grad_norm": 0.5823286175727844, "learning_rate": 9.141588735883232e-05, "loss": 0.0402, "step": 1380 }, { "grad_norm": 0.6147422194480896, "learning_rate": 9.126086397257612e-05, "loss": 0.0419, "step": 1390 }, { "grad_norm": 0.619178056716919, "learning_rate": 9.110458719335659e-05, "loss": 0.042, "step": 1400 }, { "grad_norm": 0.5730240345001221, "learning_rate": 9.094706176843777e-05, "loss": 0.0437, "step": 1410 }, { "grad_norm": 0.5200291275978088, "learning_rate": 9.078829248301417e-05, "loss": 0.039, "step": 1420 }, { "grad_norm": 0.6252971291542053, "learning_rate": 9.062828416006539e-05, "loss": 0.0426, "step": 1430 }, { "grad_norm": 0.7060213685035706, "learning_rate": 9.046704166020961e-05, "loss": 0.0407, "step": 1440 }, { "grad_norm": 0.5895113945007324, "learning_rate": 9.030456988155596e-05, "loss": 0.0408, "step": 1450 }, { "grad_norm": 0.4463379979133606, "learning_rate": 9.014087375955573e-05, "loss": 0.0402, "step": 1460 }, { "grad_norm": 0.6615645885467529, "learning_rate": 8.997595826685243e-05, "loss": 0.0446, "step": 1470 }, { "grad_norm": 0.4830588102340698, "learning_rate": 8.980982841313074e-05, "loss": 0.0409, "step": 1480 }, { "grad_norm": 0.47620710730552673, "learning_rate": 8.964248924496435e-05, "loss": 0.0384, "step": 1490 }, { "grad_norm": 0.5986263155937195, "learning_rate": 8.947394584566258e-05, "loss": 0.0389, "step": 1500 }, { "grad_norm": 0.4985571503639221, "learning_rate": 8.930420333511606e-05, "loss": 0.0394, "step": 1510 }, { "grad_norm": 0.6862359046936035, "learning_rate": 8.913326686964117e-05, "loss": 0.0393, "step": 1520 }, { "grad_norm": 0.6026191711425781, "learning_rate": 8.89611416418234e-05, "loss": 0.0385, "step": 1530 }, { "grad_norm": 0.6502306461334229, "learning_rate": 8.878783288035957e-05, "loss": 0.0386, "step": 1540 }, { "grad_norm": 0.6114603877067566, "learning_rate": 8.86133458498991e-05, "loss": 0.0406, "step": 1550 }, { "grad_norm": 0.6185985207557678, "learning_rate": 8.843768585088393e-05, "loss": 0.0385, "step": 1560 }, { "grad_norm": 0.5343775749206543, "learning_rate": 8.82608582193877e-05, "loss": 0.0404, "step": 1570 }, { "grad_norm": 0.5147342085838318, "learning_rate": 8.80828683269535e-05, "loss": 0.0386, "step": 1580 }, { "grad_norm": 0.578025758266449, "learning_rate": 8.790372158043074e-05, "loss": 0.0378, "step": 1590 }, { "grad_norm": 0.5646693706512451, "learning_rate": 8.772342342181095e-05, "loss": 0.039, "step": 1600 }, { "grad_norm": 0.6474507451057434, "learning_rate": 8.75419793280624e-05, "loss": 0.0376, "step": 1610 }, { "grad_norm": 0.5903915762901306, "learning_rate": 8.735939481096378e-05, "loss": 0.0383, "step": 1620 }, { "grad_norm": 0.5238552689552307, "learning_rate": 8.717567541693673e-05, "loss": 0.0403, "step": 1630 }, { "grad_norm": 0.7640721201896667, "learning_rate": 8.699082672687734e-05, "loss": 0.0391, "step": 1640 }, { "grad_norm": 0.41057243943214417, "learning_rate": 8.680485435598673e-05, "loss": 0.0421, "step": 1650 }, { "grad_norm": 0.49183857440948486, "learning_rate": 8.661776395360029e-05, "loss": 0.0388, "step": 1660 }, { "grad_norm": 0.6991832256317139, "learning_rate": 8.642956120301626e-05, "loss": 0.0372, "step": 1670 }, { "grad_norm": 0.6492579579353333, "learning_rate": 8.624025182132292e-05, "loss": 0.038, "step": 1680 }, { "grad_norm": 0.43906593322753906, "learning_rate": 8.604984155922506e-05, "loss": 0.0379, "step": 1690 }, { "grad_norm": 0.5015841126441956, "learning_rate": 8.585833620086918e-05, "loss": 0.0371, "step": 1700 }, { "grad_norm": 0.5101574659347534, "learning_rate": 8.566574156366784e-05, "loss": 0.0351, "step": 1710 }, { "grad_norm": 0.9203280210494995, "learning_rate": 8.547206349812298e-05, "loss": 0.0395, "step": 1720 }, { "grad_norm": 0.5471207499504089, "learning_rate": 8.527730788764805e-05, "loss": 0.0399, "step": 1730 }, { "grad_norm": 0.5866566896438599, "learning_rate": 8.508148064838948e-05, "loss": 0.0366, "step": 1740 }, { "grad_norm": 0.6344946026802063, "learning_rate": 8.488458772904684e-05, "loss": 0.0369, "step": 1750 }, { "grad_norm": 0.46788814663887024, "learning_rate": 8.468663511069217e-05, "loss": 0.0363, "step": 1760 }, { "grad_norm": 0.46914973855018616, "learning_rate": 8.448762880658825e-05, "loss": 0.0374, "step": 1770 }, { "grad_norm": 0.6406647562980652, "learning_rate": 8.428757486200603e-05, "loss": 0.0366, "step": 1780 }, { "grad_norm": 0.4758830964565277, "learning_rate": 8.40864793540409e-05, "loss": 0.0395, "step": 1790 }, { "grad_norm": 0.5328498482704163, "learning_rate": 8.388434839142813e-05, "loss": 0.0377, "step": 1800 }, { "grad_norm": 0.47568827867507935, "learning_rate": 8.368118811435726e-05, "loss": 0.0353, "step": 1810 }, { "grad_norm": 0.5234317779541016, "learning_rate": 8.347700469428564e-05, "loss": 0.0348, "step": 1820 }, { "grad_norm": 0.5909441709518433, "learning_rate": 8.327180433375091e-05, "loss": 0.0365, "step": 1830 }, { "grad_norm": 0.5138852000236511, "learning_rate": 8.306559326618259e-05, "loss": 0.0353, "step": 1840 }, { "grad_norm": 0.5205146074295044, "learning_rate": 8.285837775571276e-05, "loss": 0.0353, "step": 1850 }, { "grad_norm": 0.5465312004089355, "learning_rate": 8.265016409698573e-05, "loss": 0.0368, "step": 1860 }, { "grad_norm": 0.5244470238685608, "learning_rate": 8.244095861496686e-05, "loss": 0.0366, "step": 1870 }, { "grad_norm": 0.4607563018798828, "learning_rate": 8.223076766475035e-05, "loss": 0.0358, "step": 1880 }, { "grad_norm": 0.566045880317688, "learning_rate": 8.201959763136633e-05, "loss": 0.0342, "step": 1890 }, { "grad_norm": 0.5815367698669434, "learning_rate": 8.180745492958674e-05, "loss": 0.0359, "step": 1900 }, { "grad_norm": 0.5056940913200378, "learning_rate": 8.159434600373061e-05, "loss": 0.0338, "step": 1910 }, { "grad_norm": 0.6393750905990601, "learning_rate": 8.138027732746818e-05, "loss": 0.0359, "step": 1920 }, { "grad_norm": 0.3954344689846039, "learning_rate": 8.116525540362434e-05, "loss": 0.0348, "step": 1930 }, { "grad_norm": 0.5240972638130188, "learning_rate": 8.094928676398101e-05, "loss": 0.0351, "step": 1940 }, { "grad_norm": 0.5751314759254456, "learning_rate": 8.073237796907882e-05, "loss": 0.0344, "step": 1950 }, { "grad_norm": 0.638588547706604, "learning_rate": 8.051453560801772e-05, "loss": 0.0393, "step": 1960 }, { "grad_norm": 0.5971081852912903, "learning_rate": 8.029576629825687e-05, "loss": 0.0341, "step": 1970 }, { "grad_norm": 0.538398027420044, "learning_rate": 8.007607668541362e-05, "loss": 0.0352, "step": 1980 }, { "grad_norm": 0.516894519329071, "learning_rate": 7.985547344306161e-05, "loss": 0.0385, "step": 1990 }, { "grad_norm": 0.5202428698539734, "learning_rate": 7.963396327252812e-05, "loss": 0.0373, "step": 2000 }, { "grad_norm": 0.6087396144866943, "learning_rate": 7.941155290269038e-05, "loss": 0.0352, "step": 2010 }, { "grad_norm": 0.5531020760536194, "learning_rate": 7.918824908977123e-05, "loss": 0.0333, "step": 2020 }, { "grad_norm": 0.4421527087688446, "learning_rate": 7.896405861713394e-05, "loss": 0.0336, "step": 2030 }, { "grad_norm": 0.5156572461128235, "learning_rate": 7.873898829507606e-05, "loss": 0.0336, "step": 2040 }, { "grad_norm": 0.4285777807235718, "learning_rate": 7.851304496062254e-05, "loss": 0.0337, "step": 2050 }, { "grad_norm": 0.4402537941932678, "learning_rate": 7.828623547731818e-05, "loss": 0.0338, "step": 2060 }, { "grad_norm": 0.5571301579475403, "learning_rate": 7.80585667350189e-05, "loss": 0.0347, "step": 2070 }, { "grad_norm": 0.4796355366706848, "learning_rate": 7.783004564968263e-05, "loss": 0.0338, "step": 2080 }, { "grad_norm": 0.4996072053909302, "learning_rate": 7.760067916315921e-05, "loss": 0.033, "step": 2090 }, { "grad_norm": 0.5444074273109436, "learning_rate": 7.737047424297941e-05, "loss": 0.0332, "step": 2100 }, { "grad_norm": 0.5246517658233643, "learning_rate": 7.713943788214337e-05, "loss": 0.0328, "step": 2110 }, { "grad_norm": 0.45199844241142273, "learning_rate": 7.690757709890812e-05, "loss": 0.0329, "step": 2120 }, { "grad_norm": 0.5004163980484009, "learning_rate": 7.66748989365744e-05, "loss": 0.0328, "step": 2130 }, { "grad_norm": 0.44086408615112305, "learning_rate": 7.644141046327271e-05, "loss": 0.0311, "step": 2140 }, { "grad_norm": 0.47598645091056824, "learning_rate": 7.620711877174866e-05, "loss": 0.0326, "step": 2150 }, { "grad_norm": 0.49366524815559387, "learning_rate": 7.597203097914732e-05, "loss": 0.034, "step": 2160 }, { "grad_norm": 0.6028842329978943, "learning_rate": 7.573615422679726e-05, "loss": 0.0349, "step": 2170 }, { "grad_norm": 0.44428279995918274, "learning_rate": 7.549949567999345e-05, "loss": 0.0312, "step": 2180 }, { "grad_norm": 0.5870749950408936, "learning_rate": 7.526206252777968e-05, "loss": 0.0345, "step": 2190 }, { "grad_norm": 0.6251340508460999, "learning_rate": 7.50238619827301e-05, "loss": 0.0337, "step": 2200 }, { "grad_norm": 0.5077648162841797, "learning_rate": 7.478490128073022e-05, "loss": 0.0336, "step": 2210 }, { "grad_norm": 0.5786702036857605, "learning_rate": 7.454518768075704e-05, "loss": 0.0331, "step": 2220 }, { "grad_norm": 0.4865873456001282, "learning_rate": 7.430472846465856e-05, "loss": 0.032, "step": 2230 }, { "grad_norm": 0.5625883340835571, "learning_rate": 7.406353093693253e-05, "loss": 0.0326, "step": 2240 }, { "grad_norm": 0.5131894946098328, "learning_rate": 7.382160242450469e-05, "loss": 0.033, "step": 2250 }, { "grad_norm": 0.4549960196018219, "learning_rate": 7.357895027650598e-05, "loss": 0.0311, "step": 2260 }, { "grad_norm": 0.46319058537483215, "learning_rate": 7.333558186404958e-05, "loss": 0.0319, "step": 2270 }, { "grad_norm": 0.5134724378585815, "learning_rate": 7.309150458000668e-05, "loss": 0.033, "step": 2280 }, { "grad_norm": 0.5446333289146423, "learning_rate": 7.284672583878219e-05, "loss": 0.0325, "step": 2290 }, { "grad_norm": 0.5468949675559998, "learning_rate": 7.260125307608929e-05, "loss": 0.0337, "step": 2300 }, { "grad_norm": 0.4193164110183716, "learning_rate": 7.235509374872373e-05, "loss": 0.033, "step": 2310 }, { "grad_norm": 0.4370434582233429, "learning_rate": 7.210825533433719e-05, "loss": 0.0318, "step": 2320 }, { "grad_norm": 0.44524067640304565, "learning_rate": 7.186074533121013e-05, "loss": 0.0328, "step": 2330 }, { "grad_norm": 0.45491671562194824, "learning_rate": 7.161257125802413e-05, "loss": 0.0321, "step": 2340 }, { "grad_norm": 0.5783659219741821, "learning_rate": 7.136374065363334e-05, "loss": 0.034, "step": 2350 }, { "grad_norm": 0.44561490416526794, "learning_rate": 7.11142610768356e-05, "loss": 0.0323, "step": 2360 }, { "grad_norm": 0.4409330189228058, "learning_rate": 7.086414010614276e-05, "loss": 0.0313, "step": 2370 }, { "grad_norm": 0.5489862561225891, "learning_rate": 7.061338533955043e-05, "loss": 0.0337, "step": 2380 }, { "grad_norm": 0.6935354471206665, "learning_rate": 7.036200439430725e-05, "loss": 0.0338, "step": 2390 }, { "grad_norm": 0.47750264406204224, "learning_rate": 7.01100049066835e-05, "loss": 0.0324, "step": 2400 }, { "grad_norm": 0.4267221689224243, "learning_rate": 6.985739453173903e-05, "loss": 0.0336, "step": 2410 }, { "grad_norm": 0.5106669068336487, "learning_rate": 6.960418094309085e-05, "loss": 0.0323, "step": 2420 }, { "grad_norm": 0.3738442361354828, "learning_rate": 6.93503718326799e-05, "loss": 0.0315, "step": 2430 }, { "grad_norm": 0.5010490417480469, "learning_rate": 6.909597491053751e-05, "loss": 0.031, "step": 2440 }, { "grad_norm": 0.6164571642875671, "learning_rate": 6.884099790455113e-05, "loss": 0.0312, "step": 2450 }, { "grad_norm": 0.6351388096809387, "learning_rate": 6.858544856022952e-05, "loss": 0.033, "step": 2460 }, { "grad_norm": 0.46630266308784485, "learning_rate": 6.83293346404676e-05, "loss": 0.0333, "step": 2470 }, { "grad_norm": 0.44635745882987976, "learning_rate": 6.80726639253105e-05, "loss": 0.0348, "step": 2480 }, { "grad_norm": 0.5177208185195923, "learning_rate": 6.781544421171732e-05, "loss": 0.0327, "step": 2490 }, { "grad_norm": 0.5045998096466064, "learning_rate": 6.755768331332424e-05, "loss": 0.0331, "step": 2500 }, { "grad_norm": 0.42711469531059265, "learning_rate": 6.729938906020713e-05, "loss": 0.0325, "step": 2510 }, { "grad_norm": 0.38780301809310913, "learning_rate": 6.704056929864376e-05, "loss": 0.0306, "step": 2520 }, { "grad_norm": 0.44992300868034363, "learning_rate": 6.67812318908754e-05, "loss": 0.0298, "step": 2530 }, { "grad_norm": 0.3885252773761749, "learning_rate": 6.6521384714868e-05, "loss": 0.0291, "step": 2540 }, { "grad_norm": 0.5003411769866943, "learning_rate": 6.626103566407295e-05, "loss": 0.0298, "step": 2550 }, { "grad_norm": 0.4585190415382385, "learning_rate": 6.600019264718713e-05, "loss": 0.0304, "step": 2560 }, { "grad_norm": 0.5360947847366333, "learning_rate": 6.573886358791285e-05, "loss": 0.0321, "step": 2570 }, { "grad_norm": 0.4568477272987366, "learning_rate": 6.547705642471703e-05, "loss": 0.0311, "step": 2580 }, { "grad_norm": 0.34080684185028076, "learning_rate": 6.521477911059008e-05, "loss": 0.0279, "step": 2590 }, { "grad_norm": 0.4268318712711334, "learning_rate": 6.495203961280434e-05, "loss": 0.0302, "step": 2600 }, { "grad_norm": 0.4563651382923126, "learning_rate": 6.468884591267204e-05, "loss": 0.0302, "step": 2610 }, { "grad_norm": 0.564641535282135, "learning_rate": 6.44252060053028e-05, "loss": 0.0298, "step": 2620 }, { "grad_norm": 0.5848179459571838, "learning_rate": 6.416112789936086e-05, "loss": 0.0311, "step": 2630 }, { "grad_norm": 0.4817412495613098, "learning_rate": 6.389661961682173e-05, "loss": 0.0299, "step": 2640 }, { "grad_norm": 0.4469004273414612, "learning_rate": 6.363168919272846e-05, "loss": 0.0293, "step": 2650 }, { "grad_norm": 0.46473902463912964, "learning_rate": 6.336634467494768e-05, "loss": 0.0288, "step": 2660 }, { "grad_norm": 0.5113839507102966, "learning_rate": 6.310059412392505e-05, "loss": 0.031, "step": 2670 }, { "grad_norm": 0.46477293968200684, "learning_rate": 6.283444561244042e-05, "loss": 0.029, "step": 2680 }, { "grad_norm": 0.4217671751976013, "learning_rate": 6.256790722536251e-05, "loss": 0.0288, "step": 2690 }, { "grad_norm": 0.47681549191474915, "learning_rate": 6.230098705940354e-05, "loss": 0.0291, "step": 2700 }, { "grad_norm": 0.4174683392047882, "learning_rate": 6.203369322287306e-05, "loss": 0.0289, "step": 2710 }, { "grad_norm": 0.3565898537635803, "learning_rate": 6.17660338354317e-05, "loss": 0.028, "step": 2720 }, { "grad_norm": 0.42243969440460205, "learning_rate": 6.149801702784456e-05, "loss": 0.0278, "step": 2730 }, { "grad_norm": 0.4470120668411255, "learning_rate": 6.122965094173424e-05, "loss": 0.029, "step": 2740 }, { "grad_norm": 0.47333022952079773, "learning_rate": 6.0960943729333374e-05, "loss": 0.0307, "step": 2750 }, { "grad_norm": 0.4755380153656006, "learning_rate": 6.069190355323717e-05, "loss": 0.0282, "step": 2760 }, { "grad_norm": 0.44144678115844727, "learning_rate": 6.042253858615532e-05, "loss": 0.0286, "step": 2770 }, { "grad_norm": 0.4741808772087097, "learning_rate": 6.015285701066382e-05, "loss": 0.0306, "step": 2780 }, { "grad_norm": 0.456865519285202, "learning_rate": 5.988286701895631e-05, "loss": 0.0324, "step": 2790 }, { "grad_norm": 0.4278053641319275, "learning_rate": 5.961257681259535e-05, "loss": 0.0311, "step": 2800 }, { "grad_norm": 0.4751148819923401, "learning_rate": 5.934199460226317e-05, "loss": 0.03, "step": 2810 }, { "grad_norm": 0.4468934237957001, "learning_rate": 5.9071128607512285e-05, "loss": 0.029, "step": 2820 }, { "grad_norm": 0.31696584820747375, "learning_rate": 5.8799987056515804e-05, "loss": 0.0288, "step": 2830 }, { "grad_norm": 0.315243124961853, "learning_rate": 5.8528578185817514e-05, "loss": 0.0296, "step": 2840 }, { "grad_norm": 0.40967297554016113, "learning_rate": 5.825691024008162e-05, "loss": 0.0285, "step": 2850 }, { "grad_norm": 0.41214829683303833, "learning_rate": 5.798499147184233e-05, "loss": 0.0285, "step": 2860 }, { "grad_norm": 0.44682854413986206, "learning_rate": 5.771283014125317e-05, "loss": 0.0309, "step": 2870 }, { "grad_norm": 0.5166850686073303, "learning_rate": 5.7440434515836064e-05, "loss": 0.0286, "step": 2880 }, { "grad_norm": 0.46400532126426697, "learning_rate": 5.7167812870230094e-05, "loss": 0.0287, "step": 2890 }, { "grad_norm": 0.4673866033554077, "learning_rate": 5.689497348594035e-05, "loss": 0.0285, "step": 2900 }, { "grad_norm": 0.4213738441467285, "learning_rate": 5.662192465108613e-05, "loss": 0.0294, "step": 2910 }, { "grad_norm": 0.5419370532035828, "learning_rate": 5.634867466014932e-05, "loss": 0.0277, "step": 2920 }, { "grad_norm": 0.451986700296402, "learning_rate": 5.607523181372234e-05, "loss": 0.0293, "step": 2930 }, { "grad_norm": 0.48955249786376953, "learning_rate": 5.5801604418256117e-05, "loss": 0.0283, "step": 2940 }, { "grad_norm": 0.4409627318382263, "learning_rate": 5.552780078580756e-05, "loss": 0.0267, "step": 2950 }, { "grad_norm": 0.4321872591972351, "learning_rate": 5.525382923378728e-05, "loss": 0.0279, "step": 2960 }, { "grad_norm": 0.4211738407611847, "learning_rate": 5.49796980847068e-05, "loss": 0.0303, "step": 2970 }, { "grad_norm": 0.4374380111694336, "learning_rate": 5.470541566592573e-05, "loss": 0.0294, "step": 2980 }, { "grad_norm": 0.37320834398269653, "learning_rate": 5.443099030939887e-05, "loss": 0.0277, "step": 2990 }, { "grad_norm": 0.38404184579849243, "learning_rate": 5.415643035142309e-05, "loss": 0.0272, "step": 3000 }, { "grad_norm": 0.3510778546333313, "learning_rate": 5.3881744132384104e-05, "loss": 0.0282, "step": 3010 }, { "grad_norm": 0.4736660420894623, "learning_rate": 5.360693999650303e-05, "loss": 0.0285, "step": 3020 }, { "grad_norm": 0.43348023295402527, "learning_rate": 5.3332026291583016e-05, "loss": 0.0281, "step": 3030 }, { "grad_norm": 0.4993012845516205, "learning_rate": 5.305701136875566e-05, "loss": 0.0292, "step": 3040 }, { "grad_norm": 0.45350855588912964, "learning_rate": 5.278190358222721e-05, "loss": 0.0282, "step": 3050 }, { "grad_norm": 0.3909547030925751, "learning_rate": 5.25067112890249e-05, "loss": 0.0287, "step": 3060 }, { "grad_norm": 0.32262375950813293, "learning_rate": 5.2231442848743064e-05, "loss": 0.0282, "step": 3070 }, { "grad_norm": 0.3546123504638672, "learning_rate": 5.1956106623289145e-05, "loss": 0.0265, "step": 3080 }, { "grad_norm": 0.36697161197662354, "learning_rate": 5.168071097662972e-05, "loss": 0.0261, "step": 3090 }, { "grad_norm": 0.6001777052879333, "learning_rate": 5.1405264274536445e-05, "loss": 0.0279, "step": 3100 }, { "grad_norm": 0.46627113223075867, "learning_rate": 5.112977488433188e-05, "loss": 0.0281, "step": 3110 }, { "grad_norm": 0.3822088837623596, "learning_rate": 5.085425117463533e-05, "loss": 0.0275, "step": 3120 }, { "grad_norm": 0.40099310874938965, "learning_rate": 5.057870151510864e-05, "loss": 0.0268, "step": 3130 }, { "grad_norm": 0.4191964268684387, "learning_rate": 5.030313427620197e-05, "loss": 0.0266, "step": 3140 }, { "grad_norm": 0.41972440481185913, "learning_rate": 5.0027557828899426e-05, "loss": 0.0271, "step": 3150 }, { "grad_norm": 0.39679303765296936, "learning_rate": 4.975198054446492e-05, "loss": 0.0265, "step": 3160 }, { "grad_norm": 0.4636761546134949, "learning_rate": 4.947641079418773e-05, "loss": 0.0267, "step": 3170 }, { "grad_norm": 0.44737759232521057, "learning_rate": 4.920085694912828e-05, "loss": 0.0284, "step": 3180 }, { "grad_norm": 0.38707834482192993, "learning_rate": 4.892532737986387e-05, "loss": 0.0265, "step": 3190 }, { "grad_norm": 0.4388893246650696, "learning_rate": 4.864983045623434e-05, "loss": 0.0293, "step": 3200 }, { "grad_norm": 0.4819057881832123, "learning_rate": 4.837437454708784e-05, "loss": 0.0282, "step": 3210 }, { "grad_norm": 0.3984449803829193, "learning_rate": 4.809896802002662e-05, "loss": 0.0261, "step": 3220 }, { "grad_norm": 0.42898425459861755, "learning_rate": 4.7823619241152854e-05, "loss": 0.026, "step": 3230 }, { "grad_norm": 0.337990939617157, "learning_rate": 4.754833657481445e-05, "loss": 0.025, "step": 3240 }, { "grad_norm": 0.4331616461277008, "learning_rate": 4.7273128383351015e-05, "loss": 0.0278, "step": 3250 }, { "grad_norm": 0.5213151574134827, "learning_rate": 4.699800302683981e-05, "loss": 0.0276, "step": 3260 }, { "grad_norm": 0.3841381371021271, "learning_rate": 4.6722968862841806e-05, "loss": 0.0253, "step": 3270 }, { "grad_norm": 0.45123445987701416, "learning_rate": 4.6448034246147754e-05, "loss": 0.0269, "step": 3280 }, { "grad_norm": 0.4308590292930603, "learning_rate": 4.6173207528524476e-05, "loss": 0.0257, "step": 3290 }, { "grad_norm": 0.3546655774116516, "learning_rate": 4.58984970584611e-05, "loss": 0.0264, "step": 3300 }, { "grad_norm": 0.42120057344436646, "learning_rate": 4.562391118091544e-05, "loss": 0.0277, "step": 3310 }, { "grad_norm": 0.4904358685016632, "learning_rate": 4.534945823706056e-05, "loss": 0.0287, "step": 3320 }, { "grad_norm": 0.3876945376396179, "learning_rate": 4.507514656403137e-05, "loss": 0.0271, "step": 3330 }, { "grad_norm": 0.3710775673389435, "learning_rate": 4.480098449467132e-05, "loss": 0.0262, "step": 3340 }, { "grad_norm": 0.34494373202323914, "learning_rate": 4.452698035727929e-05, "loss": 0.0256, "step": 3350 }, { "grad_norm": 0.3994169235229492, "learning_rate": 4.425314247535668e-05, "loss": 0.0273, "step": 3360 }, { "grad_norm": 0.3998394310474396, "learning_rate": 4.3979479167354477e-05, "loss": 0.0266, "step": 3370 }, { "grad_norm": 0.41353845596313477, "learning_rate": 4.370599874642055e-05, "loss": 0.0292, "step": 3380 }, { "grad_norm": 0.4640069901943207, "learning_rate": 4.3432709520147205e-05, "loss": 0.0275, "step": 3390 }, { "grad_norm": 0.4176912307739258, "learning_rate": 4.315961979031875e-05, "loss": 0.0265, "step": 3400 }, { "grad_norm": 0.3660881817340851, "learning_rate": 4.2886737852659325e-05, "loss": 0.0266, "step": 3410 }, { "grad_norm": 0.4644610583782196, "learning_rate": 4.261407199658093e-05, "loss": 0.0263, "step": 3420 }, { "grad_norm": 0.4425753951072693, "learning_rate": 4.234163050493158e-05, "loss": 0.0249, "step": 3430 }, { "grad_norm": 0.4039178192615509, "learning_rate": 4.2069421653743706e-05, "loss": 0.0264, "step": 3440 }, { "grad_norm": 0.36323806643486023, "learning_rate": 4.179745371198276e-05, "loss": 0.0254, "step": 3450 }, { "grad_norm": 0.39530399441719055, "learning_rate": 4.1525734941296026e-05, "loss": 0.0266, "step": 3460 }, { "grad_norm": 0.34070858359336853, "learning_rate": 4.125427359576162e-05, "loss": 0.0255, "step": 3470 }, { "grad_norm": 0.4473513066768646, "learning_rate": 4.0983077921637815e-05, "loss": 0.0257, "step": 3480 }, { "grad_norm": 0.3686147630214691, "learning_rate": 4.07121561571125e-05, "loss": 0.0258, "step": 3490 }, { "grad_norm": 0.3667031526565552, "learning_rate": 4.044151653205292e-05, "loss": 0.0237, "step": 3500 }, { "grad_norm": 0.3009558916091919, "learning_rate": 4.0171167267755696e-05, "loss": 0.0247, "step": 3510 }, { "grad_norm": 0.4037996530532837, "learning_rate": 3.9901116576697083e-05, "loss": 0.0244, "step": 3520 }, { "grad_norm": 0.36101481318473816, "learning_rate": 3.963137266228349e-05, "loss": 0.0239, "step": 3530 }, { "grad_norm": 0.41857394576072693, "learning_rate": 3.93619437186023e-05, "loss": 0.0266, "step": 3540 }, { "grad_norm": 0.414919912815094, "learning_rate": 3.9092837930172884e-05, "loss": 0.0253, "step": 3550 }, { "grad_norm": 0.30529114603996277, "learning_rate": 3.8824063471698105e-05, "loss": 0.0252, "step": 3560 }, { "grad_norm": 0.3524399995803833, "learning_rate": 3.855562850781589e-05, "loss": 0.0251, "step": 3570 }, { "grad_norm": 0.3394084572792053, "learning_rate": 3.828754119285123e-05, "loss": 0.0248, "step": 3580 }, { "grad_norm": 0.3748822510242462, "learning_rate": 3.801980967056851e-05, "loss": 0.0259, "step": 3590 }, { "grad_norm": 0.3017618954181671, "learning_rate": 3.77524420739241e-05, "loss": 0.0256, "step": 3600 }, { "grad_norm": 0.404254287481308, "learning_rate": 3.748544652481927e-05, "loss": 0.0241, "step": 3610 }, { "grad_norm": 0.38028210401535034, "learning_rate": 3.721883113385353e-05, "loss": 0.0273, "step": 3620 }, { "grad_norm": 0.30036142468452454, "learning_rate": 3.695260400007819e-05, "loss": 0.0232, "step": 3630 }, { "grad_norm": 0.35517075657844543, "learning_rate": 3.6686773210750385e-05, "loss": 0.0253, "step": 3640 }, { "grad_norm": 0.35242655873298645, "learning_rate": 3.642134684108737e-05, "loss": 0.0246, "step": 3650 }, { "grad_norm": 0.40823134779930115, "learning_rate": 3.615633295402123e-05, "loss": 0.0257, "step": 3660 }, { "grad_norm": 0.344542533159256, "learning_rate": 3.5891739599953945e-05, "loss": 0.0232, "step": 3670 }, { "grad_norm": 0.422847718000412, "learning_rate": 3.5627574816512846e-05, "loss": 0.0271, "step": 3680 }, { "grad_norm": 0.42910683155059814, "learning_rate": 3.536384662830648e-05, "loss": 0.0266, "step": 3690 }, { "grad_norm": 0.37712565064430237, "learning_rate": 3.5100563046680764e-05, "loss": 0.0245, "step": 3700 }, { "grad_norm": 0.3819398581981659, "learning_rate": 3.483773206947572e-05, "loss": 0.0233, "step": 3710 }, { "grad_norm": 0.4083232581615448, "learning_rate": 3.457536168078247e-05, "loss": 0.0235, "step": 3720 }, { "grad_norm": 0.48376643657684326, "learning_rate": 3.431345985070067e-05, "loss": 0.025, "step": 3730 }, { "grad_norm": 0.3357716500759125, "learning_rate": 3.40520345350965e-05, "loss": 0.0235, "step": 3740 }, { "grad_norm": 0.29756292700767517, "learning_rate": 3.379109367536089e-05, "loss": 0.0234, "step": 3750 }, { "grad_norm": 0.3365173637866974, "learning_rate": 3.3530645198168295e-05, "loss": 0.024, "step": 3760 }, { "grad_norm": 0.3592134118080139, "learning_rate": 3.327069701523595e-05, "loss": 0.0233, "step": 3770 }, { "grad_norm": 0.31844982504844666, "learning_rate": 3.301125702308353e-05, "loss": 0.025, "step": 3780 }, { "grad_norm": 0.30566251277923584, "learning_rate": 3.275233310279321e-05, "loss": 0.0232, "step": 3790 }, { "grad_norm": 0.35913974046707153, "learning_rate": 3.249393311977037e-05, "loss": 0.024, "step": 3800 }, { "grad_norm": 0.29653260111808777, "learning_rate": 3.223606492350451e-05, "loss": 0.0233, "step": 3810 }, { "grad_norm": 0.39305341243743896, "learning_rate": 3.197873634733096e-05, "loss": 0.0243, "step": 3820 }, { "grad_norm": 0.35840174555778503, "learning_rate": 3.172195520819285e-05, "loss": 0.0248, "step": 3830 }, { "grad_norm": 0.5032244324684143, "learning_rate": 3.146572930640362e-05, "loss": 0.0235, "step": 3840 }, { "grad_norm": 0.4091183543205261, "learning_rate": 3.121006642541014e-05, "loss": 0.0248, "step": 3850 }, { "grad_norm": 0.35277608036994934, "learning_rate": 3.095497433155626e-05, "loss": 0.0237, "step": 3860 }, { "grad_norm": 0.3423890769481659, "learning_rate": 3.070046077384682e-05, "loss": 0.0234, "step": 3870 }, { "grad_norm": 0.3125498294830322, "learning_rate": 3.0446533483712304e-05, "loss": 0.0229, "step": 3880 }, { "grad_norm": 0.32279565930366516, "learning_rate": 3.0193200174774038e-05, "loss": 0.0258, "step": 3890 }, { "grad_norm": 0.33911263942718506, "learning_rate": 2.994046854260974e-05, "loss": 0.0218, "step": 3900 }, { "grad_norm": 0.3048379123210907, "learning_rate": 2.9688346264519866e-05, "loss": 0.0219, "step": 3910 }, { "grad_norm": 0.31707772612571716, "learning_rate": 2.943684099929436e-05, "loss": 0.0238, "step": 3920 }, { "grad_norm": 0.35387957096099854, "learning_rate": 2.918596038697995e-05, "loss": 0.0237, "step": 3930 }, { "grad_norm": 0.34473472833633423, "learning_rate": 2.8935712048648112e-05, "loss": 0.0224, "step": 3940 }, { "grad_norm": 0.31038379669189453, "learning_rate": 2.8686103586163626e-05, "loss": 0.0234, "step": 3950 }, { "grad_norm": 0.3322765827178955, "learning_rate": 2.843714258195346e-05, "loss": 0.023, "step": 3960 }, { "grad_norm": 0.3639306426048279, "learning_rate": 2.8188836598776662e-05, "loss": 0.0238, "step": 3970 }, { "grad_norm": 0.3464415371417999, "learning_rate": 2.7941193179494484e-05, "loss": 0.0241, "step": 3980 }, { "grad_norm": 0.36975568532943726, "learning_rate": 2.7694219846841262e-05, "loss": 0.0251, "step": 3990 }, { "grad_norm": 0.29754698276519775, "learning_rate": 2.7447924103195976e-05, "loss": 0.022, "step": 4000 }, { "grad_norm": 0.27461138367652893, "learning_rate": 2.7202313430354253e-05, "loss": 0.0212, "step": 4010 }, { "grad_norm": 0.3545227348804474, "learning_rate": 2.695739528930111e-05, "loss": 0.0224, "step": 4020 }, { "grad_norm": 0.34755411744117737, "learning_rate": 2.67131771199844e-05, "loss": 0.0224, "step": 4030 }, { "grad_norm": 0.3036070764064789, "learning_rate": 2.6469666341088677e-05, "loss": 0.023, "step": 4040 }, { "grad_norm": 0.2977795898914337, "learning_rate": 2.6226870349809885e-05, "loss": 0.0233, "step": 4050 }, { "grad_norm": 0.3287654221057892, "learning_rate": 2.5984796521630737e-05, "loss": 0.0222, "step": 4060 }, { "grad_norm": 0.22366271913051605, "learning_rate": 2.574345221009653e-05, "loss": 0.022, "step": 4070 }, { "grad_norm": 0.34360748529434204, "learning_rate": 2.5502844746591804e-05, "loss": 0.0232, "step": 4080 }, { "grad_norm": 0.3500519096851349, "learning_rate": 2.526298144011775e-05, "loss": 0.0223, "step": 4090 }, { "grad_norm": 0.35333576798439026, "learning_rate": 2.5023869577070013e-05, "loss": 0.0227, "step": 4100 }, { "grad_norm": 0.3594508767127991, "learning_rate": 2.478551642101743e-05, "loss": 0.0216, "step": 4110 }, { "grad_norm": 0.3093178868293762, "learning_rate": 2.4547929212481435e-05, "loss": 0.0217, "step": 4120 }, { "grad_norm": 0.35326287150382996, "learning_rate": 2.4311115168716013e-05, "loss": 0.0224, "step": 4130 }, { "grad_norm": 0.28599607944488525, "learning_rate": 2.4075081483488494e-05, "loss": 0.0215, "step": 4140 }, { "grad_norm": 0.2827727496623993, "learning_rate": 2.3839835326861104e-05, "loss": 0.0218, "step": 4150 }, { "grad_norm": 0.27437281608581543, "learning_rate": 2.3605383844972966e-05, "loss": 0.0218, "step": 4160 }, { "grad_norm": 0.2667216658592224, "learning_rate": 2.3371734159823284e-05, "loss": 0.0215, "step": 4170 }, { "grad_norm": 0.27455681562423706, "learning_rate": 2.3138893369054766e-05, "loss": 0.0229, "step": 4180 }, { "grad_norm": 0.308096706867218, "learning_rate": 2.2906868545738102e-05, "loss": 0.0205, "step": 4190 }, { "grad_norm": 0.27138468623161316, "learning_rate": 2.2675666738157186e-05, "loss": 0.0218, "step": 4200 }, { "grad_norm": 0.25461405515670776, "learning_rate": 2.2445294969594844e-05, "loss": 0.0222, "step": 4210 }, { "grad_norm": 0.30527886748313904, "learning_rate": 2.22157602381196e-05, "loss": 0.0215, "step": 4220 }, { "grad_norm": 0.3930235505104065, "learning_rate": 2.1987069516373098e-05, "loss": 0.0235, "step": 4230 }, { "grad_norm": 0.3291780948638916, "learning_rate": 2.1759229751358217e-05, "loss": 0.0215, "step": 4240 }, { "grad_norm": 0.27991819381713867, "learning_rate": 2.1532247864228084e-05, "loss": 0.0229, "step": 4250 }, { "grad_norm": 0.2986624538898468, "learning_rate": 2.1306130750075865e-05, "loss": 0.0212, "step": 4260 }, { "grad_norm": 0.283047616481781, "learning_rate": 2.1080885277725236e-05, "loss": 0.0217, "step": 4270 }, { "grad_norm": 0.2986035645008087, "learning_rate": 2.085651828952175e-05, "loss": 0.0217, "step": 4280 }, { "grad_norm": 0.27685680985450745, "learning_rate": 2.063303660112506e-05, "loss": 0.0234, "step": 4290 }, { "grad_norm": 0.350088894367218, "learning_rate": 2.0410447001301753e-05, "loss": 0.0207, "step": 4300 }, { "grad_norm": 0.3324001431465149, "learning_rate": 2.0188756251719203e-05, "loss": 0.0213, "step": 4310 }, { "grad_norm": 0.3323828876018524, "learning_rate": 1.9967971086740195e-05, "loss": 0.0225, "step": 4320 }, { "grad_norm": 0.3833501636981964, "learning_rate": 1.974809821321827e-05, "loss": 0.0217, "step": 4330 }, { "grad_norm": 0.3206261396408081, "learning_rate": 1.9529144310294023e-05, "loss": 0.0221, "step": 4340 }, { "grad_norm": 0.2521507740020752, "learning_rate": 1.9311116029192278e-05, "loss": 0.0198, "step": 4350 }, { "grad_norm": 0.26638540625572205, "learning_rate": 1.909401999301993e-05, "loss": 0.0213, "step": 4360 }, { "grad_norm": 0.2637273073196411, "learning_rate": 1.887786279656482e-05, "loss": 0.0203, "step": 4370 }, { "grad_norm": 0.2609107196331024, "learning_rate": 1.8662651006095387e-05, "loss": 0.0191, "step": 4380 }, { "grad_norm": 0.31063973903656006, "learning_rate": 1.8448391159161204e-05, "loss": 0.0226, "step": 4390 }, { "grad_norm": 0.26557642221450806, "learning_rate": 1.8235089764394408e-05, "loss": 0.02, "step": 4400 }, { "grad_norm": 0.2592714726924896, "learning_rate": 1.8022753301311935e-05, "loss": 0.0205, "step": 4410 }, { "grad_norm": 0.3369557857513428, "learning_rate": 1.7811388220118707e-05, "loss": 0.0215, "step": 4420 }, { "grad_norm": 0.40116578340530396, "learning_rate": 1.7601000941511757e-05, "loss": 0.0214, "step": 4430 }, { "grad_norm": 0.33265015482902527, "learning_rate": 1.7391597856485083e-05, "loss": 0.0212, "step": 4440 }, { "grad_norm": 0.33101966977119446, "learning_rate": 1.7183185326135543e-05, "loss": 0.0227, "step": 4450 }, { "grad_norm": 0.36064624786376953, "learning_rate": 1.6975769681469705e-05, "loss": 0.0209, "step": 4460 }, { "grad_norm": 0.2728773057460785, "learning_rate": 1.676935722321139e-05, "loss": 0.0222, "step": 4470 }, { "grad_norm": 0.33906540274620056, "learning_rate": 1.6563954221610355e-05, "loss": 0.0212, "step": 4480 }, { "grad_norm": 0.3477151393890381, "learning_rate": 1.6359566916251845e-05, "loss": 0.0214, "step": 4490 }, { "grad_norm": 0.2769169807434082, "learning_rate": 1.615620151586697e-05, "loss": 0.0224, "step": 4500 }, { "grad_norm": 0.31241461634635925, "learning_rate": 1.5953864198144135e-05, "loss": 0.0225, "step": 4510 }, { "grad_norm": 0.27812764048576355, "learning_rate": 1.5752561109541447e-05, "loss": 0.0217, "step": 4520 }, { "grad_norm": 0.2607854902744293, "learning_rate": 1.5552298365099882e-05, "loss": 0.02, "step": 4530 }, { "grad_norm": 0.35663941502571106, "learning_rate": 1.5353082048257596e-05, "loss": 0.0201, "step": 4540 }, { "grad_norm": 0.26552900671958923, "learning_rate": 1.5154918210665148e-05, "loss": 0.0202, "step": 4550 }, { "grad_norm": 0.2690965235233307, "learning_rate": 1.4957812872001614e-05, "loss": 0.0199, "step": 4560 }, { "grad_norm": 0.2682316303253174, "learning_rate": 1.4761772019791748e-05, "loss": 0.0196, "step": 4570 }, { "grad_norm": 0.20987866818904877, "learning_rate": 1.4566801609224096e-05, "loss": 0.0201, "step": 4580 }, { "grad_norm": 0.2587372958660126, "learning_rate": 1.4372907562970079e-05, "loss": 0.0198, "step": 4590 }, { "grad_norm": 0.2135162204504013, "learning_rate": 1.4180095771004154e-05, "loss": 0.0206, "step": 4600 }, { "grad_norm": 0.308379590511322, "learning_rate": 1.3988372090424773e-05, "loss": 0.0207, "step": 4610 }, { "grad_norm": 0.24547438323497772, "learning_rate": 1.3797742345276521e-05, "loss": 0.0196, "step": 4620 }, { "grad_norm": 0.26744821667671204, "learning_rate": 1.3608212326373249e-05, "loss": 0.0214, "step": 4630 }, { "grad_norm": 0.2596748471260071, "learning_rate": 1.3419787791122062e-05, "loss": 0.0223, "step": 4640 }, { "grad_norm": 0.30168330669403076, "learning_rate": 1.323247446334847e-05, "loss": 0.0202, "step": 4650 }, { "grad_norm": 0.28446125984191895, "learning_rate": 1.3046278033122577e-05, "loss": 0.019, "step": 4660 }, { "grad_norm": 0.2737242877483368, "learning_rate": 1.286120415658611e-05, "loss": 0.0216, "step": 4670 }, { "grad_norm": 0.22798103094100952, "learning_rate": 1.2677258455780683e-05, "loss": 0.0202, "step": 4680 }, { "grad_norm": 0.23216412961483002, "learning_rate": 1.2494446518477022e-05, "loss": 0.0189, "step": 4690 }, { "grad_norm": 0.2562443017959595, "learning_rate": 1.2312773898005175e-05, "loss": 0.0197, "step": 4700 }, { "grad_norm": 0.309608519077301, "learning_rate": 1.2132246113085822e-05, "loss": 0.0189, "step": 4710 }, { "grad_norm": 0.2388632744550705, "learning_rate": 1.1952868647662696e-05, "loss": 0.019, "step": 4720 }, { "grad_norm": 0.28642135858535767, "learning_rate": 1.1774646950735913e-05, "loss": 0.0194, "step": 4730 }, { "grad_norm": 0.23370078206062317, "learning_rate": 1.1597586436196473e-05, "loss": 0.0185, "step": 4740 }, { "grad_norm": 0.1894742250442505, "learning_rate": 1.1421692482661856e-05, "loss": 0.0175, "step": 4750 }, { "grad_norm": 0.18605804443359375, "learning_rate": 1.124697043331256e-05, "loss": 0.0192, "step": 4760 }, { "grad_norm": 0.18748939037322998, "learning_rate": 1.107342559572977e-05, "loss": 0.0183, "step": 4770 }, { "grad_norm": 0.18629539012908936, "learning_rate": 1.090106324173426e-05, "loss": 0.0182, "step": 4780 }, { "grad_norm": 0.22638608515262604, "learning_rate": 1.0729888607226113e-05, "loss": 0.0195, "step": 4790 }, { "grad_norm": 0.2516600787639618, "learning_rate": 1.0559906892025745e-05, "loss": 0.0191, "step": 4800 }, { "grad_norm": 0.20124773681163788, "learning_rate": 1.0391123259715906e-05, "loss": 0.0183, "step": 4810 }, { "grad_norm": 0.22951306402683258, "learning_rate": 1.0223542837484839e-05, "loss": 0.019, "step": 4820 }, { "grad_norm": 0.21238106489181519, "learning_rate": 1.0057170715970559e-05, "loss": 0.0178, "step": 4830 }, { "grad_norm": 0.21776369214057922, "learning_rate": 9.892011949106172e-06, "loss": 0.0192, "step": 4840 }, { "grad_norm": 0.19844797253608704, "learning_rate": 9.728071553966339e-06, "loss": 0.0181, "step": 4850 }, { "grad_norm": 0.2200976014137268, "learning_rate": 9.56535451061496e-06, "loss": 0.019, "step": 4860 }, { "grad_norm": 0.2437240481376648, "learning_rate": 9.403865761953779e-06, "loss": 0.0194, "step": 4870 }, { "grad_norm": 0.209197998046875, "learning_rate": 9.243610213572285e-06, "loss": 0.0194, "step": 4880 }, { "grad_norm": 0.20026618242263794, "learning_rate": 9.084592733598735e-06, "loss": 0.0206, "step": 4890 }, { "grad_norm": 0.24568377435207367, "learning_rate": 8.92681815255219e-06, "loss": 0.0189, "step": 4900 }, { "grad_norm": 0.2347182184457779, "learning_rate": 8.770291263195819e-06, "loss": 0.017, "step": 4910 }, { "grad_norm": 0.19473488628864288, "learning_rate": 8.615016820391342e-06, "loss": 0.0183, "step": 4920 }, { "grad_norm": 0.23291820287704468, "learning_rate": 8.460999540954517e-06, "loss": 0.0197, "step": 4930 }, { "grad_norm": 0.18584541976451874, "learning_rate": 8.308244103511909e-06, "loss": 0.0196, "step": 4940 }, { "grad_norm": 0.22126774489879608, "learning_rate": 8.156755148358764e-06, "loss": 0.0193, "step": 4950 }, { "grad_norm": 0.19748686254024506, "learning_rate": 8.00653727731801e-06, "loss": 0.0198, "step": 4960 }, { "grad_norm": 0.172799214720726, "learning_rate": 7.857595053600513e-06, "loss": 0.0191, "step": 4970 }, { "grad_norm": 0.20784571766853333, "learning_rate": 7.709933001666431e-06, "loss": 0.018, "step": 4980 }, { "grad_norm": 0.22004172205924988, "learning_rate": 7.56355560708778e-06, "loss": 0.0179, "step": 4990 }, { "grad_norm": 0.19802697002887726, "learning_rate": 7.418467316412158e-06, "loss": 0.018, "step": 5000 }, { "grad_norm": 0.25336119532585144, "learning_rate": 7.2746725370277435e-06, "loss": 0.0178, "step": 5010 }, { "grad_norm": 0.21458624303340912, "learning_rate": 7.132175637029293e-06, "loss": 0.0179, "step": 5020 }, { "grad_norm": 0.2250562608242035, "learning_rate": 6.9909809450855345e-06, "loss": 0.0172, "step": 5030 }, { "grad_norm": 0.22691819071769714, "learning_rate": 6.851092750307686e-06, "loss": 0.0188, "step": 5040 }, { "grad_norm": 0.17341260612010956, "learning_rate": 6.712515302119077e-06, "loss": 0.0198, "step": 5050 }, { "grad_norm": 0.1902797669172287, "learning_rate": 6.575252810126143e-06, "loss": 0.0179, "step": 5060 }, { "grad_norm": 0.17719143629074097, "learning_rate": 6.439309443990532e-06, "loss": 0.0171, "step": 5070 }, { "grad_norm": 0.16415466368198395, "learning_rate": 6.304689333302416e-06, "loss": 0.0164, "step": 5080 }, { "grad_norm": 0.19889427721500397, "learning_rate": 6.171396567455051e-06, "loss": 0.0189, "step": 5090 }, { "grad_norm": 0.1713903248310089, "learning_rate": 6.039435195520604e-06, "loss": 0.0169, "step": 5100 }, { "grad_norm": 0.15443091094493866, "learning_rate": 5.908809226127054e-06, "loss": 0.0178, "step": 5110 }, { "grad_norm": 0.20828746259212494, "learning_rate": 5.779522627336537e-06, "loss": 0.0176, "step": 5120 }, { "grad_norm": 0.17979548871517181, "learning_rate": 5.651579326524709e-06, "loss": 0.0175, "step": 5130 }, { "grad_norm": 0.22293756902217865, "learning_rate": 5.524983210261481e-06, "loss": 0.0179, "step": 5140 }, { "grad_norm": 0.20830951631069183, "learning_rate": 5.399738124192988e-06, "loss": 0.0182, "step": 5150 }, { "grad_norm": 0.17396672070026398, "learning_rate": 5.2758478729247164e-06, "loss": 0.0167, "step": 5160 }, { "grad_norm": 0.17180149257183075, "learning_rate": 5.153316219905946e-06, "loss": 0.0181, "step": 5170 }, { "grad_norm": 0.18279190361499786, "learning_rate": 5.032146887315448e-06, "loss": 0.0171, "step": 5180 }, { "grad_norm": 0.18074536323547363, "learning_rate": 4.91234355594839e-06, "loss": 0.0166, "step": 5190 }, { "grad_norm": 0.15373946726322174, "learning_rate": 4.7939098651045235e-06, "loss": 0.0188, "step": 5200 }, { "grad_norm": 0.17076745629310608, "learning_rate": 4.67684941247768e-06, "loss": 0.017, "step": 5210 }, { "grad_norm": 0.19420327246189117, "learning_rate": 4.5611657540464036e-06, "loss": 0.0163, "step": 5220 }, { "grad_norm": 0.1708861142396927, "learning_rate": 4.446862403965984e-06, "loss": 0.0175, "step": 5230 }, { "grad_norm": 0.13070233166217804, "learning_rate": 4.333942834461702e-06, "loss": 0.0174, "step": 5240 }, { "grad_norm": 0.16838034987449646, "learning_rate": 4.222410475723326e-06, "loss": 0.02, "step": 5250 }, { "grad_norm": 0.16201151907444, "learning_rate": 4.112268715800943e-06, "loss": 0.0164, "step": 5260 }, { "grad_norm": 0.1617550253868103, "learning_rate": 4.003520900502028e-06, "loss": 0.0185, "step": 5270 }, { "grad_norm": 0.1492706686258316, "learning_rate": 3.8961703332898e-06, "loss": 0.017, "step": 5280 }, { "grad_norm": 0.1480911672115326, "learning_rate": 3.790220275182854e-06, "loss": 0.0172, "step": 5290 }, { "grad_norm": 0.15708868205547333, "learning_rate": 3.685673944656176e-06, "loss": 0.0169, "step": 5300 }, { "grad_norm": 0.1614643633365631, "learning_rate": 3.582534517543268e-06, "loss": 0.0173, "step": 5310 }, { "grad_norm": 0.15201203525066376, "learning_rate": 3.4808051269397512e-06, "loss": 0.0188, "step": 5320 }, { "grad_norm": 0.14969263970851898, "learning_rate": 3.380488863108183e-06, "loss": 0.017, "step": 5330 }, { "grad_norm": 0.13973307609558105, "learning_rate": 3.2815887733841365e-06, "loss": 0.0171, "step": 5340 }, { "grad_norm": 0.15208420157432556, "learning_rate": 3.1841078620836683e-06, "loss": 0.0171, "step": 5350 }, { "grad_norm": 0.17642419040203094, "learning_rate": 3.0880490904120874e-06, "loss": 0.0173, "step": 5360 }, { "grad_norm": 0.1703454852104187, "learning_rate": 2.9934153763739205e-06, "loss": 0.0177, "step": 5370 }, { "grad_norm": 0.14440421760082245, "learning_rate": 2.9002095946843277e-06, "loss": 0.0172, "step": 5380 }, { "grad_norm": 0.13597236573696136, "learning_rate": 2.8084345766817676e-06, "loss": 0.0158, "step": 5390 }, { "grad_norm": 0.1392473429441452, "learning_rate": 2.718093110241976e-06, "loss": 0.0175, "step": 5400 }, { "grad_norm": 0.16542956233024597, "learning_rate": 2.6291879396933004e-06, "loss": 0.0177, "step": 5410 }, { "grad_norm": 0.13195013999938965, "learning_rate": 2.541721765733318e-06, "loss": 0.0158, "step": 5420 }, { "grad_norm": 0.15574543178081512, "learning_rate": 2.455697245346783e-06, "loss": 0.017, "step": 5430 }, { "grad_norm": 0.14603905379772186, "learning_rate": 2.371116991724953e-06, "loss": 0.0174, "step": 5440 }, { "grad_norm": 0.17199069261550903, "learning_rate": 2.2879835741861586e-06, "loss": 0.0167, "step": 5450 }, { "grad_norm": 0.17186951637268066, "learning_rate": 2.206299518097804e-06, "loss": 0.0171, "step": 5460 }, { "grad_norm": 0.1496783345937729, "learning_rate": 2.1260673047996227e-06, "loss": 0.0166, "step": 5470 }, { "grad_norm": 0.13315020501613617, "learning_rate": 2.047289371528299e-06, "loss": 0.0161, "step": 5480 }, { "grad_norm": 0.12859411537647247, "learning_rate": 1.96996811134344e-06, "loss": 0.0174, "step": 5490 }, { "grad_norm": 0.12755151093006134, "learning_rate": 1.8941058730549132e-06, "loss": 0.0174, "step": 5500 }, { "grad_norm": 0.14982177317142487, "learning_rate": 1.8197049611514194e-06, "loss": 0.0167, "step": 5510 }, { "grad_norm": 0.16311031579971313, "learning_rate": 1.7467676357305561e-06, "loss": 0.016, "step": 5520 }, { "grad_norm": 0.1535111963748932, "learning_rate": 1.6752961124301415e-06, "loss": 0.0178, "step": 5530 }, { "grad_norm": 0.1346113234758377, "learning_rate": 1.6052925623609049e-06, "loss": 0.0163, "step": 5540 }, { "grad_norm": 0.131708562374115, "learning_rate": 1.5367591120405256e-06, "loss": 0.0164, "step": 5550 }, { "grad_norm": 0.12153219431638718, "learning_rate": 1.4696978433290653e-06, "loss": 0.017, "step": 5560 }, { "grad_norm": 0.1188512071967125, "learning_rate": 1.4041107933656928e-06, "loss": 0.0164, "step": 5570 }, { "grad_norm": 0.13979417085647583, "learning_rate": 1.339999954506821e-06, "loss": 0.0159, "step": 5580 }, { "grad_norm": 0.09836265444755554, "learning_rate": 1.2773672742655784e-06, "loss": 0.0165, "step": 5590 }, { "grad_norm": 0.15529687702655792, "learning_rate": 1.2162146552526399e-06, "loss": 0.0182, "step": 5600 }, { "grad_norm": 0.12691031396389008, "learning_rate": 1.1565439551184664e-06, "loss": 0.0165, "step": 5610 }, { "grad_norm": 0.1308346390724182, "learning_rate": 1.0983569864968346e-06, "loss": 0.0184, "step": 5620 }, { "grad_norm": 0.12242095917463303, "learning_rate": 1.0416555169497688e-06, "loss": 0.0161, "step": 5630 }, { "grad_norm": 0.11082255840301514, "learning_rate": 9.864412689139123e-07, "loss": 0.0154, "step": 5640 }, { "grad_norm": 0.09780777990818024, "learning_rate": 9.327159196481138e-07, "loss": 0.0158, "step": 5650 }, { "grad_norm": 0.10955962538719177, "learning_rate": 8.804811011825398e-07, "loss": 0.016, "step": 5660 }, { "grad_norm": 0.14597293734550476, "learning_rate": 8.297384002690866e-07, "loss": 0.0168, "step": 5670 }, { "grad_norm": 0.13006150722503662, "learning_rate": 7.804893583331696e-07, "loss": 0.0175, "step": 5680 }, { "grad_norm": 0.12242487072944641, "learning_rate": 7.32735471426893e-07, "loss": 0.0177, "step": 5690 }, { "grad_norm": 0.13306714594364166, "learning_rate": 6.864781901836259e-07, "loss": 0.0173, "step": 5700 }, { "grad_norm": 0.11279428750276566, "learning_rate": 6.417189197739093e-07, "loss": 0.0168, "step": 5710 }, { "grad_norm": 0.1149117648601532, "learning_rate": 5.984590198627849e-07, "loss": 0.0157, "step": 5720 }, { "grad_norm": 0.10999634861946106, "learning_rate": 5.566998045685112e-07, "loss": 0.0172, "step": 5730 }, { "grad_norm": 0.1290031224489212, "learning_rate": 5.164425424226016e-07, "loss": 0.0156, "step": 5740 }, { "grad_norm": 0.11297288537025452, "learning_rate": 4.776884563313266e-07, "loss": 0.0163, "step": 5750 }, { "grad_norm": 0.13406500220298767, "learning_rate": 4.404387235385443e-07, "loss": 0.0169, "step": 5760 }, { "grad_norm": 0.11824876815080643, "learning_rate": 4.0469447558995065e-07, "loss": 0.0188, "step": 5770 }, { "grad_norm": 0.12567949295043945, "learning_rate": 3.7045679829870175e-07, "loss": 0.0175, "step": 5780 }, { "grad_norm": 0.09721609950065613, "learning_rate": 3.377267317124233e-07, "loss": 0.0175, "step": 5790 }, { "grad_norm": 0.11334440857172012, "learning_rate": 3.0650527008162513e-07, "loss": 0.017, "step": 5800 }, { "grad_norm": 0.13209737837314606, "learning_rate": 2.767933618295082e-07, "loss": 0.0197, "step": 5810 }, { "grad_norm": 0.14339503645896912, "learning_rate": 2.485919095231326e-07, "loss": 0.0187, "step": 5820 }, { "grad_norm": 0.10382021963596344, "learning_rate": 2.219017698460002e-07, "loss": 0.0165, "step": 5830 }, { "grad_norm": 0.10608837008476257, "learning_rate": 1.9672375357206452e-07, "loss": 0.016, "step": 5840 }, { "grad_norm": 0.12135794758796692, "learning_rate": 1.73058625541056e-07, "loss": 0.0174, "step": 5850 }, { "grad_norm": 0.1002826914191246, "learning_rate": 1.5090710463527836e-07, "loss": 0.0166, "step": 5860 }, { "grad_norm": 0.11978262662887573, "learning_rate": 1.3026986375776485e-07, "loss": 0.0173, "step": 5870 }, { "grad_norm": 0.12693767249584198, "learning_rate": 1.1114752981183917e-07, "loss": 0.0166, "step": 5880 }, { "grad_norm": 0.12360755354166031, "learning_rate": 9.354068368204739e-08, "loss": 0.0154, "step": 5890 }, { "grad_norm": 0.09292574971914291, "learning_rate": 7.744986021656076e-08, "loss": 0.017, "step": 5900 }, { "grad_norm": 0.15183936059474945, "learning_rate": 6.287554821087783e-08, "loss": 0.0177, "step": 5910 }, { "grad_norm": 0.10597635060548782, "learning_rate": 4.981819039300284e-08, "loss": 0.0171, "step": 5920 }, { "grad_norm": 0.10936059802770615, "learning_rate": 3.827818341000655e-08, "loss": 0.0162, "step": 5930 }, { "grad_norm": 0.1095290556550026, "learning_rate": 2.8255877815946963e-08, "loss": 0.0182, "step": 5940 }, { "grad_norm": 0.09919020533561707, "learning_rate": 1.9751578061244504e-08, "loss": 0.0167, "step": 5950 }, { "grad_norm": 0.1258183717727661, "learning_rate": 1.2765542483417214e-08, "loss": 0.0172, "step": 5960 }, { "grad_norm": 0.13377799093723297, "learning_rate": 7.2979832992592365e-09, "loss": 0.0195, "step": 5970 }, { "grad_norm": 0.13954906165599823, "learning_rate": 3.349066598362649e-09, "loss": 0.0176, "step": 5980 }, { "grad_norm": 0.10413049906492233, "learning_rate": 9.189123380826114e-10, "loss": 0.0167, "step": 5990 }, { "grad_norm": 0.11175262182950974, "learning_rate": 7.594339912486703e-12, "loss": 0.0174, "step": 6000 } ], "logging_steps": 10, "max_steps": 6000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 6000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 128, "trial_name": null, "trial_params": null }