{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 2336, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0004280821917808219, "grad_norm": 163.2266845703125, "learning_rate": 9.995719178082192e-06, "loss": 1.7422, "step": 1 }, { "epoch": 0.0008561643835616438, "grad_norm": 118.90894317626953, "learning_rate": 9.991438356164384e-06, "loss": 1.2969, "step": 2 }, { "epoch": 0.0012842465753424657, "grad_norm": 117.6906967163086, "learning_rate": 9.987157534246576e-06, "loss": 1.1309, "step": 3 }, { "epoch": 0.0017123287671232876, "grad_norm": 50.19585418701172, "learning_rate": 9.982876712328769e-06, "loss": 0.5508, "step": 4 }, { "epoch": 0.0021404109589041095, "grad_norm": 80.31368255615234, "learning_rate": 9.978595890410959e-06, "loss": 0.4639, "step": 5 }, { "epoch": 0.0025684931506849314, "grad_norm": 50.50063705444336, "learning_rate": 9.97431506849315e-06, "loss": 0.4922, "step": 6 }, { "epoch": 0.0029965753424657533, "grad_norm": 65.76918029785156, "learning_rate": 9.970034246575342e-06, "loss": 0.8359, "step": 7 }, { "epoch": 0.003424657534246575, "grad_norm": 48.21067428588867, "learning_rate": 9.965753424657536e-06, "loss": 0.3955, "step": 8 }, { "epoch": 0.003852739726027397, "grad_norm": 81.70868682861328, "learning_rate": 9.961472602739727e-06, "loss": 1.4395, "step": 9 }, { "epoch": 0.004280821917808219, "grad_norm": 51.5203742980957, "learning_rate": 9.957191780821919e-06, "loss": 0.7041, "step": 10 }, { "epoch": 0.004708904109589041, "grad_norm": 31.164913177490234, "learning_rate": 9.95291095890411e-06, "loss": 0.3242, "step": 11 }, { "epoch": 0.005136986301369863, "grad_norm": 26.639284133911133, "learning_rate": 9.948630136986302e-06, "loss": 0.365, "step": 12 }, { "epoch": 0.005565068493150685, "grad_norm": 58.62739944458008, "learning_rate": 9.944349315068494e-06, "loss": 0.8047, "step": 13 }, { "epoch": 0.0059931506849315065, "grad_norm": 72.5135498046875, "learning_rate": 9.940068493150685e-06, "loss": 0.8105, "step": 14 }, { "epoch": 0.006421232876712328, "grad_norm": 66.16786193847656, "learning_rate": 9.935787671232877e-06, "loss": 1.1426, "step": 15 }, { "epoch": 0.00684931506849315, "grad_norm": 49.01142883300781, "learning_rate": 9.931506849315069e-06, "loss": 0.7441, "step": 16 }, { "epoch": 0.007277397260273972, "grad_norm": 56.357269287109375, "learning_rate": 9.92722602739726e-06, "loss": 0.7988, "step": 17 }, { "epoch": 0.007705479452054794, "grad_norm": 30.090015411376953, "learning_rate": 9.922945205479454e-06, "loss": 0.251, "step": 18 }, { "epoch": 0.008133561643835616, "grad_norm": 43.15403366088867, "learning_rate": 9.918664383561645e-06, "loss": 0.8496, "step": 19 }, { "epoch": 0.008561643835616438, "grad_norm": 26.33038902282715, "learning_rate": 9.914383561643837e-06, "loss": 0.5645, "step": 20 }, { "epoch": 0.00898972602739726, "grad_norm": 29.77689552307129, "learning_rate": 9.910102739726028e-06, "loss": 0.3594, "step": 21 }, { "epoch": 0.009417808219178082, "grad_norm": 48.89308166503906, "learning_rate": 9.90582191780822e-06, "loss": 0.5039, "step": 22 }, { "epoch": 0.009845890410958904, "grad_norm": 42.678977966308594, "learning_rate": 9.901541095890412e-06, "loss": 0.4424, "step": 23 }, { "epoch": 0.010273972602739725, "grad_norm": 77.9081802368164, "learning_rate": 9.897260273972603e-06, "loss": 1.1953, "step": 24 }, { "epoch": 0.010702054794520547, "grad_norm": 49.19930648803711, "learning_rate": 9.892979452054795e-06, "loss": 0.2832, "step": 25 }, { "epoch": 0.01113013698630137, "grad_norm": 77.62024688720703, "learning_rate": 9.888698630136987e-06, "loss": 1.1602, "step": 26 }, { "epoch": 0.011558219178082191, "grad_norm": 47.39247131347656, "learning_rate": 9.884417808219178e-06, "loss": 0.5938, "step": 27 }, { "epoch": 0.011986301369863013, "grad_norm": 33.812538146972656, "learning_rate": 9.880136986301372e-06, "loss": 0.5889, "step": 28 }, { "epoch": 0.012414383561643835, "grad_norm": 27.215848922729492, "learning_rate": 9.875856164383563e-06, "loss": 0.4067, "step": 29 }, { "epoch": 0.012842465753424657, "grad_norm": 37.977394104003906, "learning_rate": 9.871575342465755e-06, "loss": 0.5703, "step": 30 }, { "epoch": 0.013270547945205479, "grad_norm": 21.77713394165039, "learning_rate": 9.867294520547946e-06, "loss": 0.3599, "step": 31 }, { "epoch": 0.0136986301369863, "grad_norm": 15.376811027526855, "learning_rate": 9.863013698630138e-06, "loss": 0.2498, "step": 32 }, { "epoch": 0.014126712328767123, "grad_norm": 10.837747573852539, "learning_rate": 9.85873287671233e-06, "loss": 0.0626, "step": 33 }, { "epoch": 0.014554794520547944, "grad_norm": 44.55924987792969, "learning_rate": 9.854452054794521e-06, "loss": 0.5439, "step": 34 }, { "epoch": 0.014982876712328766, "grad_norm": 34.450809478759766, "learning_rate": 9.850171232876713e-06, "loss": 0.1338, "step": 35 }, { "epoch": 0.015410958904109588, "grad_norm": 42.857093811035156, "learning_rate": 9.845890410958905e-06, "loss": 0.4434, "step": 36 }, { "epoch": 0.015839041095890412, "grad_norm": 41.09054946899414, "learning_rate": 9.841609589041096e-06, "loss": 0.6562, "step": 37 }, { "epoch": 0.016267123287671232, "grad_norm": 43.41743087768555, "learning_rate": 9.837328767123288e-06, "loss": 0.4844, "step": 38 }, { "epoch": 0.016695205479452056, "grad_norm": 76.30646514892578, "learning_rate": 9.833047945205481e-06, "loss": 0.519, "step": 39 }, { "epoch": 0.017123287671232876, "grad_norm": 66.53297424316406, "learning_rate": 9.828767123287673e-06, "loss": 0.6021, "step": 40 }, { "epoch": 0.0175513698630137, "grad_norm": 35.35666275024414, "learning_rate": 9.824486301369864e-06, "loss": 0.3784, "step": 41 }, { "epoch": 0.01797945205479452, "grad_norm": 66.11115264892578, "learning_rate": 9.820205479452054e-06, "loss": 0.8262, "step": 42 }, { "epoch": 0.018407534246575343, "grad_norm": 12.854829788208008, "learning_rate": 9.815924657534248e-06, "loss": 0.0774, "step": 43 }, { "epoch": 0.018835616438356163, "grad_norm": 26.50552749633789, "learning_rate": 9.81164383561644e-06, "loss": 0.313, "step": 44 }, { "epoch": 0.019263698630136987, "grad_norm": 59.839263916015625, "learning_rate": 9.807363013698631e-06, "loss": 0.5889, "step": 45 }, { "epoch": 0.019691780821917807, "grad_norm": 41.4076042175293, "learning_rate": 9.803082191780823e-06, "loss": 0.4512, "step": 46 }, { "epoch": 0.02011986301369863, "grad_norm": 52.38404083251953, "learning_rate": 9.798801369863014e-06, "loss": 0.4341, "step": 47 }, { "epoch": 0.02054794520547945, "grad_norm": 94.31884002685547, "learning_rate": 9.794520547945206e-06, "loss": 1.3066, "step": 48 }, { "epoch": 0.020976027397260275, "grad_norm": 45.07887268066406, "learning_rate": 9.7902397260274e-06, "loss": 0.2788, "step": 49 }, { "epoch": 0.021404109589041095, "grad_norm": 35.69428634643555, "learning_rate": 9.78595890410959e-06, "loss": 0.2617, "step": 50 }, { "epoch": 0.02183219178082192, "grad_norm": 49.2519645690918, "learning_rate": 9.78167808219178e-06, "loss": 0.2021, "step": 51 }, { "epoch": 0.02226027397260274, "grad_norm": 34.78032684326172, "learning_rate": 9.777397260273972e-06, "loss": 0.1367, "step": 52 }, { "epoch": 0.022688356164383562, "grad_norm": 180.70758056640625, "learning_rate": 9.773116438356166e-06, "loss": 2.0557, "step": 53 }, { "epoch": 0.023116438356164382, "grad_norm": 44.840415954589844, "learning_rate": 9.768835616438357e-06, "loss": 0.2007, "step": 54 }, { "epoch": 0.023544520547945206, "grad_norm": 103.3006820678711, "learning_rate": 9.764554794520549e-06, "loss": 1.1592, "step": 55 }, { "epoch": 0.023972602739726026, "grad_norm": 123.79376220703125, "learning_rate": 9.76027397260274e-06, "loss": 0.9277, "step": 56 }, { "epoch": 0.02440068493150685, "grad_norm": 94.13998413085938, "learning_rate": 9.755993150684932e-06, "loss": 1.0566, "step": 57 }, { "epoch": 0.02482876712328767, "grad_norm": 61.686767578125, "learning_rate": 9.751712328767124e-06, "loss": 0.5391, "step": 58 }, { "epoch": 0.025256849315068493, "grad_norm": 95.09843444824219, "learning_rate": 9.747431506849317e-06, "loss": 1.2383, "step": 59 }, { "epoch": 0.025684931506849314, "grad_norm": 49.338706970214844, "learning_rate": 9.743150684931507e-06, "loss": 0.585, "step": 60 }, { "epoch": 0.026113013698630137, "grad_norm": 68.59893035888672, "learning_rate": 9.738869863013699e-06, "loss": 0.6484, "step": 61 }, { "epoch": 0.026541095890410957, "grad_norm": 50.83188247680664, "learning_rate": 9.73458904109589e-06, "loss": 0.4683, "step": 62 }, { "epoch": 0.02696917808219178, "grad_norm": 70.9660873413086, "learning_rate": 9.730308219178084e-06, "loss": 0.8398, "step": 63 }, { "epoch": 0.0273972602739726, "grad_norm": 22.691551208496094, "learning_rate": 9.726027397260275e-06, "loss": 0.2114, "step": 64 }, { "epoch": 0.027825342465753425, "grad_norm": 57.38175582885742, "learning_rate": 9.721746575342467e-06, "loss": 0.6953, "step": 65 }, { "epoch": 0.028253424657534245, "grad_norm": 24.31278419494629, "learning_rate": 9.717465753424659e-06, "loss": 0.0945, "step": 66 }, { "epoch": 0.02868150684931507, "grad_norm": 58.020423889160156, "learning_rate": 9.71318493150685e-06, "loss": 0.9297, "step": 67 }, { "epoch": 0.02910958904109589, "grad_norm": 51.30492401123047, "learning_rate": 9.708904109589042e-06, "loss": 0.4922, "step": 68 }, { "epoch": 0.029537671232876712, "grad_norm": 115.82005310058594, "learning_rate": 9.704623287671233e-06, "loss": 0.6094, "step": 69 }, { "epoch": 0.029965753424657533, "grad_norm": 46.8453369140625, "learning_rate": 9.700342465753425e-06, "loss": 0.2905, "step": 70 }, { "epoch": 0.030393835616438356, "grad_norm": 56.09636688232422, "learning_rate": 9.696061643835617e-06, "loss": 0.2026, "step": 71 }, { "epoch": 0.030821917808219176, "grad_norm": 42.543582916259766, "learning_rate": 9.691780821917808e-06, "loss": 0.3647, "step": 72 }, { "epoch": 0.03125, "grad_norm": 65.88575744628906, "learning_rate": 9.6875e-06, "loss": 0.8838, "step": 73 }, { "epoch": 0.031678082191780824, "grad_norm": 43.704833984375, "learning_rate": 9.683219178082193e-06, "loss": 0.3853, "step": 74 }, { "epoch": 0.03210616438356165, "grad_norm": 29.904958724975586, "learning_rate": 9.678938356164385e-06, "loss": 0.2568, "step": 75 }, { "epoch": 0.032534246575342464, "grad_norm": 33.839019775390625, "learning_rate": 9.674657534246577e-06, "loss": 0.1411, "step": 76 }, { "epoch": 0.03296232876712329, "grad_norm": 42.87808609008789, "learning_rate": 9.670376712328768e-06, "loss": 0.3745, "step": 77 }, { "epoch": 0.03339041095890411, "grad_norm": 14.992167472839355, "learning_rate": 9.66609589041096e-06, "loss": 0.0513, "step": 78 }, { "epoch": 0.033818493150684935, "grad_norm": 43.532371520996094, "learning_rate": 9.661815068493151e-06, "loss": 0.5078, "step": 79 }, { "epoch": 0.03424657534246575, "grad_norm": 36.9251708984375, "learning_rate": 9.657534246575343e-06, "loss": 0.3452, "step": 80 }, { "epoch": 0.034674657534246575, "grad_norm": 28.610567092895508, "learning_rate": 9.653253424657535e-06, "loss": 0.282, "step": 81 }, { "epoch": 0.0351027397260274, "grad_norm": 47.53772735595703, "learning_rate": 9.648972602739726e-06, "loss": 0.487, "step": 82 }, { "epoch": 0.03553082191780822, "grad_norm": 45.757686614990234, "learning_rate": 9.644691780821918e-06, "loss": 0.4766, "step": 83 }, { "epoch": 0.03595890410958904, "grad_norm": 58.78816604614258, "learning_rate": 9.640410958904111e-06, "loss": 0.9688, "step": 84 }, { "epoch": 0.03638698630136986, "grad_norm": 111.34010314941406, "learning_rate": 9.636130136986303e-06, "loss": 1.7471, "step": 85 }, { "epoch": 0.036815068493150686, "grad_norm": 42.7132682800293, "learning_rate": 9.631849315068495e-06, "loss": 0.7178, "step": 86 }, { "epoch": 0.03724315068493151, "grad_norm": 88.56800842285156, "learning_rate": 9.627568493150685e-06, "loss": 0.582, "step": 87 }, { "epoch": 0.03767123287671233, "grad_norm": 48.32605743408203, "learning_rate": 9.623287671232878e-06, "loss": 0.3745, "step": 88 }, { "epoch": 0.03809931506849315, "grad_norm": 80.7742919921875, "learning_rate": 9.61900684931507e-06, "loss": 0.6471, "step": 89 }, { "epoch": 0.038527397260273974, "grad_norm": 32.47557830810547, "learning_rate": 9.614726027397261e-06, "loss": 0.3096, "step": 90 }, { "epoch": 0.0389554794520548, "grad_norm": 53.78463363647461, "learning_rate": 9.610445205479453e-06, "loss": 0.5657, "step": 91 }, { "epoch": 0.039383561643835614, "grad_norm": 11.662224769592285, "learning_rate": 9.606164383561644e-06, "loss": 0.0432, "step": 92 }, { "epoch": 0.03981164383561644, "grad_norm": 54.706214904785156, "learning_rate": 9.601883561643836e-06, "loss": 0.5068, "step": 93 }, { "epoch": 0.04023972602739726, "grad_norm": 36.1409912109375, "learning_rate": 9.59760273972603e-06, "loss": 0.5074, "step": 94 }, { "epoch": 0.040667808219178085, "grad_norm": 29.356271743774414, "learning_rate": 9.593321917808221e-06, "loss": 0.4761, "step": 95 }, { "epoch": 0.0410958904109589, "grad_norm": 73.32432556152344, "learning_rate": 9.589041095890411e-06, "loss": 0.8477, "step": 96 }, { "epoch": 0.041523972602739725, "grad_norm": 75.54895782470703, "learning_rate": 9.584760273972603e-06, "loss": 1.1465, "step": 97 }, { "epoch": 0.04195205479452055, "grad_norm": 30.29691505432129, "learning_rate": 9.580479452054796e-06, "loss": 0.1902, "step": 98 }, { "epoch": 0.04238013698630137, "grad_norm": 25.872453689575195, "learning_rate": 9.576198630136987e-06, "loss": 0.305, "step": 99 }, { "epoch": 0.04280821917808219, "grad_norm": 16.537263870239258, "learning_rate": 9.571917808219179e-06, "loss": 0.1003, "step": 100 }, { "epoch": 0.04323630136986301, "grad_norm": 27.05809783935547, "learning_rate": 9.56763698630137e-06, "loss": 0.3013, "step": 101 }, { "epoch": 0.04366438356164384, "grad_norm": 42.295204162597656, "learning_rate": 9.563356164383562e-06, "loss": 0.5579, "step": 102 }, { "epoch": 0.04409246575342466, "grad_norm": 46.98442840576172, "learning_rate": 9.559075342465754e-06, "loss": 0.499, "step": 103 }, { "epoch": 0.04452054794520548, "grad_norm": 40.21479034423828, "learning_rate": 9.554794520547946e-06, "loss": 0.3672, "step": 104 }, { "epoch": 0.0449486301369863, "grad_norm": 27.482547760009766, "learning_rate": 9.550513698630137e-06, "loss": 0.312, "step": 105 }, { "epoch": 0.045376712328767124, "grad_norm": 42.102352142333984, "learning_rate": 9.546232876712329e-06, "loss": 0.4111, "step": 106 }, { "epoch": 0.04580479452054795, "grad_norm": 39.58338928222656, "learning_rate": 9.54195205479452e-06, "loss": 0.335, "step": 107 }, { "epoch": 0.046232876712328765, "grad_norm": 28.716392517089844, "learning_rate": 9.537671232876712e-06, "loss": 0.1155, "step": 108 }, { "epoch": 0.04666095890410959, "grad_norm": 51.15207290649414, "learning_rate": 9.533390410958905e-06, "loss": 0.1726, "step": 109 }, { "epoch": 0.04708904109589041, "grad_norm": 54.00579833984375, "learning_rate": 9.529109589041097e-06, "loss": 0.6201, "step": 110 }, { "epoch": 0.047517123287671235, "grad_norm": 55.49365997314453, "learning_rate": 9.524828767123289e-06, "loss": 0.3022, "step": 111 }, { "epoch": 0.04794520547945205, "grad_norm": 50.05632019042969, "learning_rate": 9.52054794520548e-06, "loss": 0.333, "step": 112 }, { "epoch": 0.048373287671232876, "grad_norm": 117.1716537475586, "learning_rate": 9.516267123287672e-06, "loss": 1.2188, "step": 113 }, { "epoch": 0.0488013698630137, "grad_norm": 47.18971252441406, "learning_rate": 9.511986301369864e-06, "loss": 0.4805, "step": 114 }, { "epoch": 0.04922945205479452, "grad_norm": 29.63117790222168, "learning_rate": 9.507705479452055e-06, "loss": 0.3142, "step": 115 }, { "epoch": 0.04965753424657534, "grad_norm": 32.12873840332031, "learning_rate": 9.503424657534247e-06, "loss": 0.2753, "step": 116 }, { "epoch": 0.05008561643835616, "grad_norm": 84.57488250732422, "learning_rate": 9.499143835616439e-06, "loss": 0.666, "step": 117 }, { "epoch": 0.05051369863013699, "grad_norm": 19.031274795532227, "learning_rate": 9.49486301369863e-06, "loss": 0.1104, "step": 118 }, { "epoch": 0.05094178082191781, "grad_norm": 92.2868881225586, "learning_rate": 9.490582191780823e-06, "loss": 1.2754, "step": 119 }, { "epoch": 0.05136986301369863, "grad_norm": 16.880685806274414, "learning_rate": 9.486301369863015e-06, "loss": 0.1113, "step": 120 }, { "epoch": 0.05179794520547945, "grad_norm": 35.29542541503906, "learning_rate": 9.482020547945207e-06, "loss": 0.183, "step": 121 }, { "epoch": 0.052226027397260275, "grad_norm": 12.117992401123047, "learning_rate": 9.477739726027398e-06, "loss": 0.0419, "step": 122 }, { "epoch": 0.0526541095890411, "grad_norm": 13.47909164428711, "learning_rate": 9.47345890410959e-06, "loss": 0.0979, "step": 123 }, { "epoch": 0.053082191780821915, "grad_norm": 55.554847717285156, "learning_rate": 9.469178082191782e-06, "loss": 0.333, "step": 124 }, { "epoch": 0.05351027397260274, "grad_norm": 84.60369873046875, "learning_rate": 9.464897260273973e-06, "loss": 1.5195, "step": 125 }, { "epoch": 0.05393835616438356, "grad_norm": 30.786800384521484, "learning_rate": 9.460616438356165e-06, "loss": 0.1035, "step": 126 }, { "epoch": 0.054366438356164386, "grad_norm": 94.16078186035156, "learning_rate": 9.456335616438357e-06, "loss": 0.9502, "step": 127 }, { "epoch": 0.0547945205479452, "grad_norm": 63.53156661987305, "learning_rate": 9.452054794520548e-06, "loss": 0.3003, "step": 128 }, { "epoch": 0.055222602739726026, "grad_norm": 35.526771545410156, "learning_rate": 9.447773972602741e-06, "loss": 0.0728, "step": 129 }, { "epoch": 0.05565068493150685, "grad_norm": 78.25074005126953, "learning_rate": 9.443493150684933e-06, "loss": 0.3237, "step": 130 }, { "epoch": 0.05607876712328767, "grad_norm": 88.91143798828125, "learning_rate": 9.439212328767125e-06, "loss": 1.1025, "step": 131 }, { "epoch": 0.05650684931506849, "grad_norm": 23.621562957763672, "learning_rate": 9.434931506849316e-06, "loss": 0.0715, "step": 132 }, { "epoch": 0.056934931506849314, "grad_norm": 66.71867370605469, "learning_rate": 9.430650684931508e-06, "loss": 0.7058, "step": 133 }, { "epoch": 0.05736301369863014, "grad_norm": 82.85579681396484, "learning_rate": 9.4263698630137e-06, "loss": 0.6074, "step": 134 }, { "epoch": 0.05779109589041096, "grad_norm": 14.319716453552246, "learning_rate": 9.422089041095891e-06, "loss": 0.0315, "step": 135 }, { "epoch": 0.05821917808219178, "grad_norm": 48.554527282714844, "learning_rate": 9.417808219178083e-06, "loss": 0.2295, "step": 136 }, { "epoch": 0.0586472602739726, "grad_norm": 214.54042053222656, "learning_rate": 9.413527397260275e-06, "loss": 3.3359, "step": 137 }, { "epoch": 0.059075342465753425, "grad_norm": 38.910667419433594, "learning_rate": 9.409246575342466e-06, "loss": 0.3572, "step": 138 }, { "epoch": 0.05950342465753425, "grad_norm": 75.3293685913086, "learning_rate": 9.404965753424658e-06, "loss": 0.908, "step": 139 }, { "epoch": 0.059931506849315065, "grad_norm": 47.39945602416992, "learning_rate": 9.400684931506851e-06, "loss": 0.5874, "step": 140 }, { "epoch": 0.06035958904109589, "grad_norm": 66.14482879638672, "learning_rate": 9.396404109589043e-06, "loss": 0.6099, "step": 141 }, { "epoch": 0.06078767123287671, "grad_norm": 49.586753845214844, "learning_rate": 9.392123287671233e-06, "loss": 0.4082, "step": 142 }, { "epoch": 0.061215753424657536, "grad_norm": 39.09817886352539, "learning_rate": 9.387842465753424e-06, "loss": 0.1222, "step": 143 }, { "epoch": 0.06164383561643835, "grad_norm": 31.635862350463867, "learning_rate": 9.383561643835618e-06, "loss": 0.2689, "step": 144 }, { "epoch": 0.062071917808219176, "grad_norm": 39.722808837890625, "learning_rate": 9.37928082191781e-06, "loss": 0.6655, "step": 145 }, { "epoch": 0.0625, "grad_norm": 28.22904396057129, "learning_rate": 9.375000000000001e-06, "loss": 0.0806, "step": 146 }, { "epoch": 0.06292808219178082, "grad_norm": 65.94910430908203, "learning_rate": 9.370719178082193e-06, "loss": 0.7681, "step": 147 }, { "epoch": 0.06335616438356165, "grad_norm": 46.708412170410156, "learning_rate": 9.366438356164384e-06, "loss": 0.7578, "step": 148 }, { "epoch": 0.06378424657534247, "grad_norm": 41.5900764465332, "learning_rate": 9.362157534246576e-06, "loss": 0.5845, "step": 149 }, { "epoch": 0.0642123287671233, "grad_norm": 36.07903289794922, "learning_rate": 9.357876712328769e-06, "loss": 0.4512, "step": 150 }, { "epoch": 0.0646404109589041, "grad_norm": 44.07269287109375, "learning_rate": 9.353595890410959e-06, "loss": 0.1777, "step": 151 }, { "epoch": 0.06506849315068493, "grad_norm": 26.03644371032715, "learning_rate": 9.34931506849315e-06, "loss": 0.0771, "step": 152 }, { "epoch": 0.06549657534246575, "grad_norm": 21.55603790283203, "learning_rate": 9.345034246575342e-06, "loss": 0.1191, "step": 153 }, { "epoch": 0.06592465753424658, "grad_norm": 17.84248161315918, "learning_rate": 9.340753424657536e-06, "loss": 0.0387, "step": 154 }, { "epoch": 0.0663527397260274, "grad_norm": 7.136550426483154, "learning_rate": 9.336472602739727e-06, "loss": 0.0396, "step": 155 }, { "epoch": 0.06678082191780822, "grad_norm": 44.68436813354492, "learning_rate": 9.332191780821919e-06, "loss": 0.5281, "step": 156 }, { "epoch": 0.06720890410958905, "grad_norm": 55.19449234008789, "learning_rate": 9.32791095890411e-06, "loss": 0.4492, "step": 157 }, { "epoch": 0.06763698630136987, "grad_norm": 84.2194595336914, "learning_rate": 9.323630136986302e-06, "loss": 1.0137, "step": 158 }, { "epoch": 0.06806506849315068, "grad_norm": 53.6732292175293, "learning_rate": 9.319349315068494e-06, "loss": 0.563, "step": 159 }, { "epoch": 0.0684931506849315, "grad_norm": 64.67556762695312, "learning_rate": 9.315068493150685e-06, "loss": 0.6777, "step": 160 }, { "epoch": 0.06892123287671233, "grad_norm": 75.4085922241211, "learning_rate": 9.310787671232877e-06, "loss": 1.207, "step": 161 }, { "epoch": 0.06934931506849315, "grad_norm": 37.94438934326172, "learning_rate": 9.306506849315069e-06, "loss": 0.2366, "step": 162 }, { "epoch": 0.06977739726027397, "grad_norm": 38.889774322509766, "learning_rate": 9.30222602739726e-06, "loss": 0.3499, "step": 163 }, { "epoch": 0.0702054794520548, "grad_norm": 35.34381866455078, "learning_rate": 9.297945205479454e-06, "loss": 0.4927, "step": 164 }, { "epoch": 0.07063356164383562, "grad_norm": 49.14551544189453, "learning_rate": 9.293664383561645e-06, "loss": 0.5483, "step": 165 }, { "epoch": 0.07106164383561644, "grad_norm": 33.285499572753906, "learning_rate": 9.289383561643837e-06, "loss": 0.1514, "step": 166 }, { "epoch": 0.07148972602739725, "grad_norm": 44.87728500366211, "learning_rate": 9.285102739726028e-06, "loss": 0.335, "step": 167 }, { "epoch": 0.07191780821917808, "grad_norm": 105.69696807861328, "learning_rate": 9.28082191780822e-06, "loss": 0.9668, "step": 168 }, { "epoch": 0.0723458904109589, "grad_norm": 48.49049758911133, "learning_rate": 9.276541095890412e-06, "loss": 0.3359, "step": 169 }, { "epoch": 0.07277397260273973, "grad_norm": 60.151302337646484, "learning_rate": 9.272260273972603e-06, "loss": 0.2583, "step": 170 }, { "epoch": 0.07320205479452055, "grad_norm": 48.52293395996094, "learning_rate": 9.267979452054795e-06, "loss": 0.6616, "step": 171 }, { "epoch": 0.07363013698630137, "grad_norm": 24.146413803100586, "learning_rate": 9.263698630136987e-06, "loss": 0.2061, "step": 172 }, { "epoch": 0.0740582191780822, "grad_norm": 18.99797821044922, "learning_rate": 9.259417808219178e-06, "loss": 0.1609, "step": 173 }, { "epoch": 0.07448630136986302, "grad_norm": 31.66946792602539, "learning_rate": 9.25513698630137e-06, "loss": 0.2869, "step": 174 }, { "epoch": 0.07491438356164383, "grad_norm": 21.496623992919922, "learning_rate": 9.250856164383563e-06, "loss": 0.15, "step": 175 }, { "epoch": 0.07534246575342465, "grad_norm": 28.544666290283203, "learning_rate": 9.246575342465755e-06, "loss": 0.291, "step": 176 }, { "epoch": 0.07577054794520548, "grad_norm": 38.538814544677734, "learning_rate": 9.242294520547946e-06, "loss": 0.439, "step": 177 }, { "epoch": 0.0761986301369863, "grad_norm": 24.289535522460938, "learning_rate": 9.238013698630136e-06, "loss": 0.1097, "step": 178 }, { "epoch": 0.07662671232876712, "grad_norm": 35.688507080078125, "learning_rate": 9.23373287671233e-06, "loss": 0.4103, "step": 179 }, { "epoch": 0.07705479452054795, "grad_norm": 38.46387481689453, "learning_rate": 9.229452054794521e-06, "loss": 0.2798, "step": 180 }, { "epoch": 0.07748287671232877, "grad_norm": 50.39382553100586, "learning_rate": 9.225171232876713e-06, "loss": 0.5273, "step": 181 }, { "epoch": 0.0779109589041096, "grad_norm": 80.03727722167969, "learning_rate": 9.220890410958905e-06, "loss": 0.7324, "step": 182 }, { "epoch": 0.0783390410958904, "grad_norm": 39.46415328979492, "learning_rate": 9.216609589041096e-06, "loss": 0.342, "step": 183 }, { "epoch": 0.07876712328767123, "grad_norm": 11.882387161254883, "learning_rate": 9.212328767123288e-06, "loss": 0.0403, "step": 184 }, { "epoch": 0.07919520547945205, "grad_norm": 35.587646484375, "learning_rate": 9.208047945205481e-06, "loss": 0.6038, "step": 185 }, { "epoch": 0.07962328767123288, "grad_norm": 63.19414138793945, "learning_rate": 9.203767123287673e-06, "loss": 0.5957, "step": 186 }, { "epoch": 0.0800513698630137, "grad_norm": 16.239625930786133, "learning_rate": 9.199486301369864e-06, "loss": 0.0386, "step": 187 }, { "epoch": 0.08047945205479452, "grad_norm": 91.97357177734375, "learning_rate": 9.195205479452054e-06, "loss": 0.8369, "step": 188 }, { "epoch": 0.08090753424657535, "grad_norm": 23.62206268310547, "learning_rate": 9.190924657534248e-06, "loss": 0.1511, "step": 189 }, { "epoch": 0.08133561643835617, "grad_norm": 143.5996551513672, "learning_rate": 9.18664383561644e-06, "loss": 1.6445, "step": 190 }, { "epoch": 0.08176369863013698, "grad_norm": 37.86130905151367, "learning_rate": 9.182363013698631e-06, "loss": 0.1907, "step": 191 }, { "epoch": 0.0821917808219178, "grad_norm": 62.148250579833984, "learning_rate": 9.178082191780823e-06, "loss": 0.3618, "step": 192 }, { "epoch": 0.08261986301369863, "grad_norm": 51.59618377685547, "learning_rate": 9.173801369863014e-06, "loss": 0.6575, "step": 193 }, { "epoch": 0.08304794520547945, "grad_norm": 96.1744155883789, "learning_rate": 9.169520547945206e-06, "loss": 0.7754, "step": 194 }, { "epoch": 0.08347602739726027, "grad_norm": 38.116214752197266, "learning_rate": 9.1652397260274e-06, "loss": 0.3682, "step": 195 }, { "epoch": 0.0839041095890411, "grad_norm": 8.462751388549805, "learning_rate": 9.160958904109591e-06, "loss": 0.0662, "step": 196 }, { "epoch": 0.08433219178082192, "grad_norm": 56.697025299072266, "learning_rate": 9.15667808219178e-06, "loss": 0.3604, "step": 197 }, { "epoch": 0.08476027397260275, "grad_norm": 52.983314514160156, "learning_rate": 9.152397260273972e-06, "loss": 0.186, "step": 198 }, { "epoch": 0.08518835616438356, "grad_norm": 56.10969543457031, "learning_rate": 9.148116438356166e-06, "loss": 0.252, "step": 199 }, { "epoch": 0.08561643835616438, "grad_norm": 10.476598739624023, "learning_rate": 9.143835616438357e-06, "loss": 0.0155, "step": 200 }, { "epoch": 0.0860445205479452, "grad_norm": 95.80377197265625, "learning_rate": 9.139554794520549e-06, "loss": 1.0732, "step": 201 }, { "epoch": 0.08647260273972603, "grad_norm": 57.76522445678711, "learning_rate": 9.13527397260274e-06, "loss": 0.1965, "step": 202 }, { "epoch": 0.08690068493150685, "grad_norm": 65.23712921142578, "learning_rate": 9.130993150684932e-06, "loss": 0.8074, "step": 203 }, { "epoch": 0.08732876712328767, "grad_norm": 67.85092163085938, "learning_rate": 9.126712328767124e-06, "loss": 0.6807, "step": 204 }, { "epoch": 0.0877568493150685, "grad_norm": 116.7297134399414, "learning_rate": 9.122431506849316e-06, "loss": 1.9629, "step": 205 }, { "epoch": 0.08818493150684932, "grad_norm": 17.460460662841797, "learning_rate": 9.118150684931507e-06, "loss": 0.0918, "step": 206 }, { "epoch": 0.08861301369863013, "grad_norm": 22.521827697753906, "learning_rate": 9.113869863013699e-06, "loss": 0.1381, "step": 207 }, { "epoch": 0.08904109589041095, "grad_norm": 61.357662200927734, "learning_rate": 9.10958904109589e-06, "loss": 0.4126, "step": 208 }, { "epoch": 0.08946917808219178, "grad_norm": 40.329261779785156, "learning_rate": 9.105308219178082e-06, "loss": 0.2578, "step": 209 }, { "epoch": 0.0898972602739726, "grad_norm": 36.668785095214844, "learning_rate": 9.101027397260275e-06, "loss": 0.0823, "step": 210 }, { "epoch": 0.09032534246575342, "grad_norm": 120.11133575439453, "learning_rate": 9.096746575342467e-06, "loss": 0.7204, "step": 211 }, { "epoch": 0.09075342465753425, "grad_norm": 15.322932243347168, "learning_rate": 9.092465753424659e-06, "loss": 0.1869, "step": 212 }, { "epoch": 0.09118150684931507, "grad_norm": 16.17852783203125, "learning_rate": 9.08818493150685e-06, "loss": 0.0837, "step": 213 }, { "epoch": 0.0916095890410959, "grad_norm": 109.42291259765625, "learning_rate": 9.083904109589042e-06, "loss": 0.9033, "step": 214 }, { "epoch": 0.0920376712328767, "grad_norm": 75.98442840576172, "learning_rate": 9.079623287671234e-06, "loss": 0.915, "step": 215 }, { "epoch": 0.09246575342465753, "grad_norm": 108.29961395263672, "learning_rate": 9.075342465753425e-06, "loss": 1.1074, "step": 216 }, { "epoch": 0.09289383561643835, "grad_norm": 17.57246971130371, "learning_rate": 9.071061643835617e-06, "loss": 0.0948, "step": 217 }, { "epoch": 0.09332191780821918, "grad_norm": 120.14564514160156, "learning_rate": 9.066780821917808e-06, "loss": 0.9766, "step": 218 }, { "epoch": 0.09375, "grad_norm": 34.483882904052734, "learning_rate": 9.0625e-06, "loss": 0.1245, "step": 219 }, { "epoch": 0.09417808219178082, "grad_norm": 39.514644622802734, "learning_rate": 9.058219178082193e-06, "loss": 0.5201, "step": 220 }, { "epoch": 0.09460616438356165, "grad_norm": 27.897550582885742, "learning_rate": 9.053938356164385e-06, "loss": 0.1829, "step": 221 }, { "epoch": 0.09503424657534247, "grad_norm": 70.43628692626953, "learning_rate": 9.049657534246577e-06, "loss": 0.7012, "step": 222 }, { "epoch": 0.0954623287671233, "grad_norm": 19.070241928100586, "learning_rate": 9.045376712328768e-06, "loss": 0.0787, "step": 223 }, { "epoch": 0.0958904109589041, "grad_norm": 64.4140625, "learning_rate": 9.04109589041096e-06, "loss": 0.5088, "step": 224 }, { "epoch": 0.09631849315068493, "grad_norm": 81.9620361328125, "learning_rate": 9.036815068493152e-06, "loss": 0.7358, "step": 225 }, { "epoch": 0.09674657534246575, "grad_norm": 60.969825744628906, "learning_rate": 9.032534246575343e-06, "loss": 0.4883, "step": 226 }, { "epoch": 0.09717465753424658, "grad_norm": 37.3078498840332, "learning_rate": 9.028253424657535e-06, "loss": 0.3036, "step": 227 }, { "epoch": 0.0976027397260274, "grad_norm": 11.907888412475586, "learning_rate": 9.023972602739726e-06, "loss": 0.054, "step": 228 }, { "epoch": 0.09803082191780822, "grad_norm": 23.08271026611328, "learning_rate": 9.019691780821918e-06, "loss": 0.1863, "step": 229 }, { "epoch": 0.09845890410958905, "grad_norm": 109.9488525390625, "learning_rate": 9.015410958904111e-06, "loss": 1.377, "step": 230 }, { "epoch": 0.09888698630136987, "grad_norm": 53.925804138183594, "learning_rate": 9.011130136986303e-06, "loss": 0.3386, "step": 231 }, { "epoch": 0.09931506849315068, "grad_norm": 7.760956287384033, "learning_rate": 9.006849315068495e-06, "loss": 0.0185, "step": 232 }, { "epoch": 0.0997431506849315, "grad_norm": 36.14971160888672, "learning_rate": 9.002568493150685e-06, "loss": 0.4839, "step": 233 }, { "epoch": 0.10017123287671233, "grad_norm": 10.891535758972168, "learning_rate": 8.998287671232878e-06, "loss": 0.0249, "step": 234 }, { "epoch": 0.10059931506849315, "grad_norm": 41.08586502075195, "learning_rate": 8.99400684931507e-06, "loss": 0.1882, "step": 235 }, { "epoch": 0.10102739726027397, "grad_norm": 37.54701614379883, "learning_rate": 8.989726027397261e-06, "loss": 0.2107, "step": 236 }, { "epoch": 0.1014554794520548, "grad_norm": 53.09024429321289, "learning_rate": 8.985445205479453e-06, "loss": 0.6573, "step": 237 }, { "epoch": 0.10188356164383562, "grad_norm": 32.37310028076172, "learning_rate": 8.981164383561644e-06, "loss": 0.0946, "step": 238 }, { "epoch": 0.10231164383561644, "grad_norm": 4.636551856994629, "learning_rate": 8.976883561643836e-06, "loss": 0.0192, "step": 239 }, { "epoch": 0.10273972602739725, "grad_norm": 38.00944900512695, "learning_rate": 8.972602739726028e-06, "loss": 0.2426, "step": 240 }, { "epoch": 0.10316780821917808, "grad_norm": 48.147239685058594, "learning_rate": 8.968321917808221e-06, "loss": 0.2014, "step": 241 }, { "epoch": 0.1035958904109589, "grad_norm": 9.609210014343262, "learning_rate": 8.964041095890411e-06, "loss": 0.0601, "step": 242 }, { "epoch": 0.10402397260273973, "grad_norm": 34.37236022949219, "learning_rate": 8.959760273972603e-06, "loss": 0.2742, "step": 243 }, { "epoch": 0.10445205479452055, "grad_norm": 4.758100509643555, "learning_rate": 8.955479452054794e-06, "loss": 0.0184, "step": 244 }, { "epoch": 0.10488013698630137, "grad_norm": 61.11134719848633, "learning_rate": 8.951198630136988e-06, "loss": 0.3456, "step": 245 }, { "epoch": 0.1053082191780822, "grad_norm": 67.84508514404297, "learning_rate": 8.94691780821918e-06, "loss": 0.4766, "step": 246 }, { "epoch": 0.10573630136986302, "grad_norm": 37.642887115478516, "learning_rate": 8.94263698630137e-06, "loss": 0.166, "step": 247 }, { "epoch": 0.10616438356164383, "grad_norm": 62.398292541503906, "learning_rate": 8.938356164383562e-06, "loss": 0.1968, "step": 248 }, { "epoch": 0.10659246575342465, "grad_norm": 70.62897491455078, "learning_rate": 8.934075342465754e-06, "loss": 0.5869, "step": 249 }, { "epoch": 0.10702054794520548, "grad_norm": 54.89447784423828, "learning_rate": 8.929794520547946e-06, "loss": 0.1831, "step": 250 }, { "epoch": 0.1074486301369863, "grad_norm": 38.33755111694336, "learning_rate": 8.925513698630137e-06, "loss": 0.1384, "step": 251 }, { "epoch": 0.10787671232876712, "grad_norm": 53.85675811767578, "learning_rate": 8.921232876712329e-06, "loss": 0.1135, "step": 252 }, { "epoch": 0.10830479452054795, "grad_norm": 32.73790740966797, "learning_rate": 8.91695205479452e-06, "loss": 0.1434, "step": 253 }, { "epoch": 0.10873287671232877, "grad_norm": 117.09986114501953, "learning_rate": 8.912671232876712e-06, "loss": 0.9849, "step": 254 }, { "epoch": 0.1091609589041096, "grad_norm": 68.32984161376953, "learning_rate": 8.908390410958906e-06, "loss": 0.2601, "step": 255 }, { "epoch": 0.1095890410958904, "grad_norm": 47.095462799072266, "learning_rate": 8.904109589041097e-06, "loss": 0.1807, "step": 256 }, { "epoch": 0.11001712328767123, "grad_norm": 32.43281936645508, "learning_rate": 8.899828767123289e-06, "loss": 0.0684, "step": 257 }, { "epoch": 0.11044520547945205, "grad_norm": 32.04557800292969, "learning_rate": 8.89554794520548e-06, "loss": 0.1162, "step": 258 }, { "epoch": 0.11087328767123288, "grad_norm": 26.665889739990234, "learning_rate": 8.891267123287672e-06, "loss": 0.041, "step": 259 }, { "epoch": 0.1113013698630137, "grad_norm": 48.212730407714844, "learning_rate": 8.886986301369864e-06, "loss": 0.4425, "step": 260 }, { "epoch": 0.11172945205479452, "grad_norm": 64.91281127929688, "learning_rate": 8.882705479452055e-06, "loss": 0.4561, "step": 261 }, { "epoch": 0.11215753424657535, "grad_norm": 7.845885753631592, "learning_rate": 8.878424657534247e-06, "loss": 0.0074, "step": 262 }, { "epoch": 0.11258561643835617, "grad_norm": 64.28694915771484, "learning_rate": 8.874143835616439e-06, "loss": 0.4678, "step": 263 }, { "epoch": 0.11301369863013698, "grad_norm": 8.632591247558594, "learning_rate": 8.86986301369863e-06, "loss": 0.0159, "step": 264 }, { "epoch": 0.1134417808219178, "grad_norm": 57.08514404296875, "learning_rate": 8.865582191780824e-06, "loss": 0.5645, "step": 265 }, { "epoch": 0.11386986301369863, "grad_norm": 38.54975509643555, "learning_rate": 8.861301369863015e-06, "loss": 0.332, "step": 266 }, { "epoch": 0.11429794520547945, "grad_norm": 52.48422622680664, "learning_rate": 8.857020547945207e-06, "loss": 0.2217, "step": 267 }, { "epoch": 0.11472602739726027, "grad_norm": 45.460113525390625, "learning_rate": 8.852739726027398e-06, "loss": 0.1436, "step": 268 }, { "epoch": 0.1151541095890411, "grad_norm": 72.46959686279297, "learning_rate": 8.84845890410959e-06, "loss": 0.3394, "step": 269 }, { "epoch": 0.11558219178082192, "grad_norm": 20.02398109436035, "learning_rate": 8.844178082191782e-06, "loss": 0.0464, "step": 270 }, { "epoch": 0.11601027397260275, "grad_norm": 109.50933837890625, "learning_rate": 8.839897260273973e-06, "loss": 0.6108, "step": 271 }, { "epoch": 0.11643835616438356, "grad_norm": 77.8636703491211, "learning_rate": 8.835616438356165e-06, "loss": 0.552, "step": 272 }, { "epoch": 0.11686643835616438, "grad_norm": 66.06000518798828, "learning_rate": 8.831335616438357e-06, "loss": 0.3643, "step": 273 }, { "epoch": 0.1172945205479452, "grad_norm": 54.06517791748047, "learning_rate": 8.827054794520548e-06, "loss": 0.1506, "step": 274 }, { "epoch": 0.11772260273972603, "grad_norm": 141.74339294433594, "learning_rate": 8.82277397260274e-06, "loss": 1.3018, "step": 275 }, { "epoch": 0.11815068493150685, "grad_norm": 52.25667190551758, "learning_rate": 8.818493150684933e-06, "loss": 0.3232, "step": 276 }, { "epoch": 0.11857876712328767, "grad_norm": 30.076414108276367, "learning_rate": 8.814212328767125e-06, "loss": 0.1031, "step": 277 }, { "epoch": 0.1190068493150685, "grad_norm": 37.0654296875, "learning_rate": 8.809931506849316e-06, "loss": 0.139, "step": 278 }, { "epoch": 0.11943493150684932, "grad_norm": 71.22837829589844, "learning_rate": 8.805650684931506e-06, "loss": 0.502, "step": 279 }, { "epoch": 0.11986301369863013, "grad_norm": 58.90175247192383, "learning_rate": 8.8013698630137e-06, "loss": 0.5029, "step": 280 }, { "epoch": 0.12029109589041095, "grad_norm": 48.439701080322266, "learning_rate": 8.797089041095891e-06, "loss": 0.1665, "step": 281 }, { "epoch": 0.12071917808219178, "grad_norm": 11.688260078430176, "learning_rate": 8.792808219178083e-06, "loss": 0.0427, "step": 282 }, { "epoch": 0.1211472602739726, "grad_norm": 4.981200218200684, "learning_rate": 8.788527397260275e-06, "loss": 0.0216, "step": 283 }, { "epoch": 0.12157534246575342, "grad_norm": 5.941740036010742, "learning_rate": 8.784246575342466e-06, "loss": 0.016, "step": 284 }, { "epoch": 0.12200342465753425, "grad_norm": 64.41452026367188, "learning_rate": 8.779965753424658e-06, "loss": 0.1775, "step": 285 }, { "epoch": 0.12243150684931507, "grad_norm": 46.653106689453125, "learning_rate": 8.775684931506851e-06, "loss": 0.3066, "step": 286 }, { "epoch": 0.1228595890410959, "grad_norm": 70.7468490600586, "learning_rate": 8.771404109589043e-06, "loss": 0.6025, "step": 287 }, { "epoch": 0.1232876712328767, "grad_norm": 15.12166976928711, "learning_rate": 8.767123287671233e-06, "loss": 0.0605, "step": 288 }, { "epoch": 0.12371575342465753, "grad_norm": 64.68742370605469, "learning_rate": 8.762842465753424e-06, "loss": 0.4492, "step": 289 }, { "epoch": 0.12414383561643835, "grad_norm": 154.65289306640625, "learning_rate": 8.758561643835618e-06, "loss": 0.5527, "step": 290 }, { "epoch": 0.12457191780821918, "grad_norm": 22.422409057617188, "learning_rate": 8.75428082191781e-06, "loss": 0.0825, "step": 291 }, { "epoch": 0.125, "grad_norm": 101.2022933959961, "learning_rate": 8.750000000000001e-06, "loss": 0.5625, "step": 292 }, { "epoch": 0.1254280821917808, "grad_norm": 104.07276916503906, "learning_rate": 8.745719178082193e-06, "loss": 0.791, "step": 293 }, { "epoch": 0.12585616438356165, "grad_norm": 128.5898895263672, "learning_rate": 8.741438356164384e-06, "loss": 0.8477, "step": 294 }, { "epoch": 0.12628424657534246, "grad_norm": 33.713836669921875, "learning_rate": 8.737157534246576e-06, "loss": 0.1279, "step": 295 }, { "epoch": 0.1267123287671233, "grad_norm": 8.506799697875977, "learning_rate": 8.732876712328769e-06, "loss": 0.0144, "step": 296 }, { "epoch": 0.1271404109589041, "grad_norm": 32.11957931518555, "learning_rate": 8.728595890410959e-06, "loss": 0.0525, "step": 297 }, { "epoch": 0.12756849315068494, "grad_norm": 135.8148651123047, "learning_rate": 8.72431506849315e-06, "loss": 0.8989, "step": 298 }, { "epoch": 0.12799657534246575, "grad_norm": 88.31269073486328, "learning_rate": 8.720034246575342e-06, "loss": 0.541, "step": 299 }, { "epoch": 0.1284246575342466, "grad_norm": 50.81577682495117, "learning_rate": 8.715753424657536e-06, "loss": 0.1554, "step": 300 }, { "epoch": 0.1288527397260274, "grad_norm": 36.67142868041992, "learning_rate": 8.711472602739727e-06, "loss": 0.0936, "step": 301 }, { "epoch": 0.1292808219178082, "grad_norm": 90.8173599243164, "learning_rate": 8.707191780821919e-06, "loss": 0.9378, "step": 302 }, { "epoch": 0.12970890410958905, "grad_norm": 74.02561950683594, "learning_rate": 8.70291095890411e-06, "loss": 0.3287, "step": 303 }, { "epoch": 0.13013698630136986, "grad_norm": 161.94371032714844, "learning_rate": 8.698630136986302e-06, "loss": 0.8264, "step": 304 }, { "epoch": 0.1305650684931507, "grad_norm": 14.539388656616211, "learning_rate": 8.694349315068494e-06, "loss": 0.0395, "step": 305 }, { "epoch": 0.1309931506849315, "grad_norm": 46.877479553222656, "learning_rate": 8.690068493150685e-06, "loss": 0.1481, "step": 306 }, { "epoch": 0.13142123287671234, "grad_norm": 91.92957305908203, "learning_rate": 8.685787671232877e-06, "loss": 0.8167, "step": 307 }, { "epoch": 0.13184931506849315, "grad_norm": 147.99049377441406, "learning_rate": 8.681506849315069e-06, "loss": 1.3047, "step": 308 }, { "epoch": 0.13227739726027396, "grad_norm": 31.805007934570312, "learning_rate": 8.67722602739726e-06, "loss": 0.142, "step": 309 }, { "epoch": 0.1327054794520548, "grad_norm": 23.34716033935547, "learning_rate": 8.672945205479452e-06, "loss": 0.0766, "step": 310 }, { "epoch": 0.1331335616438356, "grad_norm": 15.696638107299805, "learning_rate": 8.668664383561645e-06, "loss": 0.115, "step": 311 }, { "epoch": 0.13356164383561644, "grad_norm": 9.273674011230469, "learning_rate": 8.664383561643837e-06, "loss": 0.0193, "step": 312 }, { "epoch": 0.13398972602739725, "grad_norm": 26.187429428100586, "learning_rate": 8.660102739726029e-06, "loss": 0.0957, "step": 313 }, { "epoch": 0.1344178082191781, "grad_norm": 19.599716186523438, "learning_rate": 8.65582191780822e-06, "loss": 0.0423, "step": 314 }, { "epoch": 0.1348458904109589, "grad_norm": 43.581050872802734, "learning_rate": 8.651541095890412e-06, "loss": 0.1069, "step": 315 }, { "epoch": 0.13527397260273974, "grad_norm": 281.81787109375, "learning_rate": 8.647260273972603e-06, "loss": 1.501, "step": 316 }, { "epoch": 0.13570205479452055, "grad_norm": 15.016854286193848, "learning_rate": 8.642979452054795e-06, "loss": 0.0771, "step": 317 }, { "epoch": 0.13613013698630136, "grad_norm": 136.5507049560547, "learning_rate": 8.638698630136987e-06, "loss": 0.3523, "step": 318 }, { "epoch": 0.1365582191780822, "grad_norm": 61.46116256713867, "learning_rate": 8.634417808219178e-06, "loss": 0.3057, "step": 319 }, { "epoch": 0.136986301369863, "grad_norm": 5.453299045562744, "learning_rate": 8.63013698630137e-06, "loss": 0.0139, "step": 320 }, { "epoch": 0.13741438356164384, "grad_norm": 16.06694793701172, "learning_rate": 8.625856164383563e-06, "loss": 0.0686, "step": 321 }, { "epoch": 0.13784246575342465, "grad_norm": 166.1921844482422, "learning_rate": 8.621575342465755e-06, "loss": 0.9229, "step": 322 }, { "epoch": 0.1382705479452055, "grad_norm": 45.807918548583984, "learning_rate": 8.617294520547947e-06, "loss": 0.4268, "step": 323 }, { "epoch": 0.1386986301369863, "grad_norm": 67.76055145263672, "learning_rate": 8.613013698630136e-06, "loss": 0.2993, "step": 324 }, { "epoch": 0.1391267123287671, "grad_norm": 68.28023529052734, "learning_rate": 8.60873287671233e-06, "loss": 0.2686, "step": 325 }, { "epoch": 0.13955479452054795, "grad_norm": 16.03226089477539, "learning_rate": 8.604452054794521e-06, "loss": 0.0365, "step": 326 }, { "epoch": 0.13998287671232876, "grad_norm": 12.287507057189941, "learning_rate": 8.600171232876713e-06, "loss": 0.0195, "step": 327 }, { "epoch": 0.1404109589041096, "grad_norm": 35.907752990722656, "learning_rate": 8.595890410958905e-06, "loss": 0.1327, "step": 328 }, { "epoch": 0.1408390410958904, "grad_norm": 48.862178802490234, "learning_rate": 8.591609589041096e-06, "loss": 0.2104, "step": 329 }, { "epoch": 0.14126712328767124, "grad_norm": 69.06820678710938, "learning_rate": 8.587328767123288e-06, "loss": 0.2886, "step": 330 }, { "epoch": 0.14169520547945205, "grad_norm": 117.0577392578125, "learning_rate": 8.583047945205481e-06, "loss": 0.4004, "step": 331 }, { "epoch": 0.1421232876712329, "grad_norm": 69.75260162353516, "learning_rate": 8.578767123287673e-06, "loss": 0.369, "step": 332 }, { "epoch": 0.1425513698630137, "grad_norm": 74.98434448242188, "learning_rate": 8.574486301369865e-06, "loss": 0.5264, "step": 333 }, { "epoch": 0.1429794520547945, "grad_norm": 126.62271118164062, "learning_rate": 8.570205479452054e-06, "loss": 0.2847, "step": 334 }, { "epoch": 0.14340753424657535, "grad_norm": 39.073307037353516, "learning_rate": 8.565924657534246e-06, "loss": 0.2692, "step": 335 }, { "epoch": 0.14383561643835616, "grad_norm": 32.434967041015625, "learning_rate": 8.56164383561644e-06, "loss": 0.0771, "step": 336 }, { "epoch": 0.144263698630137, "grad_norm": 122.9669418334961, "learning_rate": 8.557363013698631e-06, "loss": 0.4845, "step": 337 }, { "epoch": 0.1446917808219178, "grad_norm": 117.14293670654297, "learning_rate": 8.553082191780823e-06, "loss": 0.8438, "step": 338 }, { "epoch": 0.14511986301369864, "grad_norm": 15.454106330871582, "learning_rate": 8.548801369863014e-06, "loss": 0.045, "step": 339 }, { "epoch": 0.14554794520547945, "grad_norm": 31.3294677734375, "learning_rate": 8.544520547945206e-06, "loss": 0.0616, "step": 340 }, { "epoch": 0.14597602739726026, "grad_norm": 105.02311706542969, "learning_rate": 8.540239726027398e-06, "loss": 0.1552, "step": 341 }, { "epoch": 0.1464041095890411, "grad_norm": 23.53959083557129, "learning_rate": 8.535958904109591e-06, "loss": 0.0627, "step": 342 }, { "epoch": 0.1468321917808219, "grad_norm": 27.087482452392578, "learning_rate": 8.531678082191781e-06, "loss": 0.1152, "step": 343 }, { "epoch": 0.14726027397260275, "grad_norm": 7.306534767150879, "learning_rate": 8.527397260273972e-06, "loss": 0.02, "step": 344 }, { "epoch": 0.14768835616438356, "grad_norm": 27.709033966064453, "learning_rate": 8.523116438356164e-06, "loss": 0.0451, "step": 345 }, { "epoch": 0.1481164383561644, "grad_norm": 70.24214935302734, "learning_rate": 8.518835616438357e-06, "loss": 0.5278, "step": 346 }, { "epoch": 0.1485445205479452, "grad_norm": 69.5459976196289, "learning_rate": 8.514554794520549e-06, "loss": 0.249, "step": 347 }, { "epoch": 0.14897260273972604, "grad_norm": 14.626334190368652, "learning_rate": 8.51027397260274e-06, "loss": 0.0392, "step": 348 }, { "epoch": 0.14940068493150685, "grad_norm": 23.599042892456055, "learning_rate": 8.505993150684932e-06, "loss": 0.0928, "step": 349 }, { "epoch": 0.14982876712328766, "grad_norm": 54.4442138671875, "learning_rate": 8.501712328767124e-06, "loss": 0.1936, "step": 350 }, { "epoch": 0.1502568493150685, "grad_norm": 15.525128364562988, "learning_rate": 8.497431506849316e-06, "loss": 0.0187, "step": 351 }, { "epoch": 0.1506849315068493, "grad_norm": 52.476654052734375, "learning_rate": 8.493150684931507e-06, "loss": 0.3615, "step": 352 }, { "epoch": 0.15111301369863014, "grad_norm": 22.190654754638672, "learning_rate": 8.488869863013699e-06, "loss": 0.1316, "step": 353 }, { "epoch": 0.15154109589041095, "grad_norm": 30.005868911743164, "learning_rate": 8.48458904109589e-06, "loss": 0.0518, "step": 354 }, { "epoch": 0.1519691780821918, "grad_norm": 85.0967025756836, "learning_rate": 8.480308219178082e-06, "loss": 0.4414, "step": 355 }, { "epoch": 0.1523972602739726, "grad_norm": 152.321044921875, "learning_rate": 8.476027397260275e-06, "loss": 2.2856, "step": 356 }, { "epoch": 0.1528253424657534, "grad_norm": 32.371856689453125, "learning_rate": 8.471746575342467e-06, "loss": 0.1064, "step": 357 }, { "epoch": 0.15325342465753425, "grad_norm": 26.163942337036133, "learning_rate": 8.467465753424659e-06, "loss": 0.2437, "step": 358 }, { "epoch": 0.15368150684931506, "grad_norm": 28.543643951416016, "learning_rate": 8.46318493150685e-06, "loss": 0.0925, "step": 359 }, { "epoch": 0.1541095890410959, "grad_norm": 17.703163146972656, "learning_rate": 8.458904109589042e-06, "loss": 0.0477, "step": 360 }, { "epoch": 0.1545376712328767, "grad_norm": 35.669437408447266, "learning_rate": 8.454623287671234e-06, "loss": 0.0703, "step": 361 }, { "epoch": 0.15496575342465754, "grad_norm": 95.41531372070312, "learning_rate": 8.450342465753425e-06, "loss": 0.8867, "step": 362 }, { "epoch": 0.15539383561643835, "grad_norm": 47.916690826416016, "learning_rate": 8.446061643835617e-06, "loss": 0.1835, "step": 363 }, { "epoch": 0.1558219178082192, "grad_norm": 13.686189651489258, "learning_rate": 8.441780821917808e-06, "loss": 0.0412, "step": 364 }, { "epoch": 0.15625, "grad_norm": 29.718189239501953, "learning_rate": 8.4375e-06, "loss": 0.1434, "step": 365 }, { "epoch": 0.1566780821917808, "grad_norm": 30.06578254699707, "learning_rate": 8.433219178082193e-06, "loss": 0.0598, "step": 366 }, { "epoch": 0.15710616438356165, "grad_norm": 52.38542556762695, "learning_rate": 8.428938356164385e-06, "loss": 0.1583, "step": 367 }, { "epoch": 0.15753424657534246, "grad_norm": 18.387348175048828, "learning_rate": 8.424657534246577e-06, "loss": 0.0325, "step": 368 }, { "epoch": 0.1579623287671233, "grad_norm": 72.9970703125, "learning_rate": 8.420376712328768e-06, "loss": 0.4585, "step": 369 }, { "epoch": 0.1583904109589041, "grad_norm": 95.2360610961914, "learning_rate": 8.416095890410958e-06, "loss": 0.4034, "step": 370 }, { "epoch": 0.15881849315068494, "grad_norm": 89.58851623535156, "learning_rate": 8.411815068493152e-06, "loss": 0.8083, "step": 371 }, { "epoch": 0.15924657534246575, "grad_norm": 37.05683898925781, "learning_rate": 8.407534246575343e-06, "loss": 0.1254, "step": 372 }, { "epoch": 0.1596746575342466, "grad_norm": 19.151887893676758, "learning_rate": 8.403253424657535e-06, "loss": 0.0333, "step": 373 }, { "epoch": 0.1601027397260274, "grad_norm": 53.85642623901367, "learning_rate": 8.398972602739726e-06, "loss": 0.2534, "step": 374 }, { "epoch": 0.1605308219178082, "grad_norm": 69.13060760498047, "learning_rate": 8.394691780821918e-06, "loss": 0.5388, "step": 375 }, { "epoch": 0.16095890410958905, "grad_norm": 123.2313232421875, "learning_rate": 8.39041095890411e-06, "loss": 0.5753, "step": 376 }, { "epoch": 0.16138698630136986, "grad_norm": 53.2192497253418, "learning_rate": 8.386130136986303e-06, "loss": 0.2144, "step": 377 }, { "epoch": 0.1618150684931507, "grad_norm": 54.52056121826172, "learning_rate": 8.381849315068495e-06, "loss": 0.332, "step": 378 }, { "epoch": 0.1622431506849315, "grad_norm": 18.94780921936035, "learning_rate": 8.377568493150685e-06, "loss": 0.0275, "step": 379 }, { "epoch": 0.16267123287671234, "grad_norm": 64.26736450195312, "learning_rate": 8.373287671232876e-06, "loss": 0.1584, "step": 380 }, { "epoch": 0.16309931506849315, "grad_norm": 73.84407043457031, "learning_rate": 8.36900684931507e-06, "loss": 0.3676, "step": 381 }, { "epoch": 0.16352739726027396, "grad_norm": 33.125770568847656, "learning_rate": 8.364726027397261e-06, "loss": 0.0897, "step": 382 }, { "epoch": 0.1639554794520548, "grad_norm": 23.03815460205078, "learning_rate": 8.360445205479453e-06, "loss": 0.0657, "step": 383 }, { "epoch": 0.1643835616438356, "grad_norm": 117.94865417480469, "learning_rate": 8.356164383561644e-06, "loss": 0.4825, "step": 384 }, { "epoch": 0.16481164383561644, "grad_norm": 217.17152404785156, "learning_rate": 8.351883561643836e-06, "loss": 1.8047, "step": 385 }, { "epoch": 0.16523972602739725, "grad_norm": 51.64556884765625, "learning_rate": 8.347602739726028e-06, "loss": 0.3049, "step": 386 }, { "epoch": 0.1656678082191781, "grad_norm": 19.442771911621094, "learning_rate": 8.343321917808221e-06, "loss": 0.0443, "step": 387 }, { "epoch": 0.1660958904109589, "grad_norm": 12.147891998291016, "learning_rate": 8.339041095890411e-06, "loss": 0.0129, "step": 388 }, { "epoch": 0.16652397260273974, "grad_norm": 135.79783630371094, "learning_rate": 8.334760273972603e-06, "loss": 0.8618, "step": 389 }, { "epoch": 0.16695205479452055, "grad_norm": 49.41326904296875, "learning_rate": 8.330479452054794e-06, "loss": 0.2112, "step": 390 }, { "epoch": 0.16738013698630136, "grad_norm": 7.714687824249268, "learning_rate": 8.326198630136988e-06, "loss": 0.0192, "step": 391 }, { "epoch": 0.1678082191780822, "grad_norm": 34.35576248168945, "learning_rate": 8.32191780821918e-06, "loss": 0.2192, "step": 392 }, { "epoch": 0.168236301369863, "grad_norm": 23.709880828857422, "learning_rate": 8.31763698630137e-06, "loss": 0.0732, "step": 393 }, { "epoch": 0.16866438356164384, "grad_norm": 11.744807243347168, "learning_rate": 8.313356164383562e-06, "loss": 0.049, "step": 394 }, { "epoch": 0.16909246575342465, "grad_norm": 18.15147590637207, "learning_rate": 8.309075342465754e-06, "loss": 0.0299, "step": 395 }, { "epoch": 0.1695205479452055, "grad_norm": 149.55770874023438, "learning_rate": 8.304794520547946e-06, "loss": 0.8519, "step": 396 }, { "epoch": 0.1699486301369863, "grad_norm": 28.994413375854492, "learning_rate": 8.300513698630137e-06, "loss": 0.0508, "step": 397 }, { "epoch": 0.1703767123287671, "grad_norm": 32.71704864501953, "learning_rate": 8.296232876712329e-06, "loss": 0.1824, "step": 398 }, { "epoch": 0.17080479452054795, "grad_norm": 102.74718475341797, "learning_rate": 8.29195205479452e-06, "loss": 0.4092, "step": 399 }, { "epoch": 0.17123287671232876, "grad_norm": 36.74433517456055, "learning_rate": 8.287671232876712e-06, "loss": 0.185, "step": 400 }, { "epoch": 0.1716609589041096, "grad_norm": 97.72563934326172, "learning_rate": 8.283390410958904e-06, "loss": 0.2309, "step": 401 }, { "epoch": 0.1720890410958904, "grad_norm": 107.80477142333984, "learning_rate": 8.279109589041097e-06, "loss": 1.3736, "step": 402 }, { "epoch": 0.17251712328767124, "grad_norm": 144.7368927001953, "learning_rate": 8.274828767123289e-06, "loss": 2.168, "step": 403 }, { "epoch": 0.17294520547945205, "grad_norm": 6.275152683258057, "learning_rate": 8.27054794520548e-06, "loss": 0.0154, "step": 404 }, { "epoch": 0.1733732876712329, "grad_norm": 40.719539642333984, "learning_rate": 8.266267123287672e-06, "loss": 0.1774, "step": 405 }, { "epoch": 0.1738013698630137, "grad_norm": 8.962625503540039, "learning_rate": 8.261986301369864e-06, "loss": 0.0352, "step": 406 }, { "epoch": 0.1742294520547945, "grad_norm": 61.032081604003906, "learning_rate": 8.257705479452055e-06, "loss": 0.2205, "step": 407 }, { "epoch": 0.17465753424657535, "grad_norm": 22.021076202392578, "learning_rate": 8.253424657534247e-06, "loss": 0.1046, "step": 408 }, { "epoch": 0.17508561643835616, "grad_norm": 50.16923904418945, "learning_rate": 8.249143835616439e-06, "loss": 0.2157, "step": 409 }, { "epoch": 0.175513698630137, "grad_norm": 60.680152893066406, "learning_rate": 8.24486301369863e-06, "loss": 0.2749, "step": 410 }, { "epoch": 0.1759417808219178, "grad_norm": 14.29369068145752, "learning_rate": 8.240582191780822e-06, "loss": 0.044, "step": 411 }, { "epoch": 0.17636986301369864, "grad_norm": 11.453596115112305, "learning_rate": 8.236301369863015e-06, "loss": 0.0219, "step": 412 }, { "epoch": 0.17679794520547945, "grad_norm": 3.6018426418304443, "learning_rate": 8.232020547945207e-06, "loss": 0.0093, "step": 413 }, { "epoch": 0.17722602739726026, "grad_norm": 24.375032424926758, "learning_rate": 8.227739726027398e-06, "loss": 0.0404, "step": 414 }, { "epoch": 0.1776541095890411, "grad_norm": 56.732452392578125, "learning_rate": 8.22345890410959e-06, "loss": 0.166, "step": 415 }, { "epoch": 0.1780821917808219, "grad_norm": 165.12954711914062, "learning_rate": 8.219178082191782e-06, "loss": 0.5665, "step": 416 }, { "epoch": 0.17851027397260275, "grad_norm": 9.569149017333984, "learning_rate": 8.214897260273973e-06, "loss": 0.033, "step": 417 }, { "epoch": 0.17893835616438356, "grad_norm": 72.97818756103516, "learning_rate": 8.210616438356165e-06, "loss": 0.1615, "step": 418 }, { "epoch": 0.1793664383561644, "grad_norm": 69.98970794677734, "learning_rate": 8.206335616438357e-06, "loss": 0.3115, "step": 419 }, { "epoch": 0.1797945205479452, "grad_norm": 84.51908111572266, "learning_rate": 8.202054794520548e-06, "loss": 0.2405, "step": 420 }, { "epoch": 0.18022260273972604, "grad_norm": 136.4776153564453, "learning_rate": 8.19777397260274e-06, "loss": 0.728, "step": 421 }, { "epoch": 0.18065068493150685, "grad_norm": 67.04364776611328, "learning_rate": 8.193493150684933e-06, "loss": 0.2666, "step": 422 }, { "epoch": 0.18107876712328766, "grad_norm": 127.31564331054688, "learning_rate": 8.189212328767125e-06, "loss": 0.5212, "step": 423 }, { "epoch": 0.1815068493150685, "grad_norm": 11.789568901062012, "learning_rate": 8.184931506849316e-06, "loss": 0.0336, "step": 424 }, { "epoch": 0.1819349315068493, "grad_norm": 103.89586639404297, "learning_rate": 8.180650684931506e-06, "loss": 1.0869, "step": 425 }, { "epoch": 0.18236301369863014, "grad_norm": 18.406875610351562, "learning_rate": 8.1763698630137e-06, "loss": 0.0239, "step": 426 }, { "epoch": 0.18279109589041095, "grad_norm": 123.96774291992188, "learning_rate": 8.172089041095891e-06, "loss": 0.6501, "step": 427 }, { "epoch": 0.1832191780821918, "grad_norm": 8.668130874633789, "learning_rate": 8.167808219178083e-06, "loss": 0.0227, "step": 428 }, { "epoch": 0.1836472602739726, "grad_norm": 23.353988647460938, "learning_rate": 8.163527397260275e-06, "loss": 0.0712, "step": 429 }, { "epoch": 0.1840753424657534, "grad_norm": 81.748291015625, "learning_rate": 8.159246575342466e-06, "loss": 0.4292, "step": 430 }, { "epoch": 0.18450342465753425, "grad_norm": 65.16221618652344, "learning_rate": 8.154965753424658e-06, "loss": 0.0958, "step": 431 }, { "epoch": 0.18493150684931506, "grad_norm": 295.2866516113281, "learning_rate": 8.150684931506851e-06, "loss": 0.5054, "step": 432 }, { "epoch": 0.1853595890410959, "grad_norm": 62.82515335083008, "learning_rate": 8.146404109589043e-06, "loss": 0.299, "step": 433 }, { "epoch": 0.1857876712328767, "grad_norm": 4.258349895477295, "learning_rate": 8.142123287671233e-06, "loss": 0.011, "step": 434 }, { "epoch": 0.18621575342465754, "grad_norm": 60.002967834472656, "learning_rate": 8.137842465753424e-06, "loss": 0.4827, "step": 435 }, { "epoch": 0.18664383561643835, "grad_norm": 144.71421813964844, "learning_rate": 8.133561643835616e-06, "loss": 0.9691, "step": 436 }, { "epoch": 0.1870719178082192, "grad_norm": 7.554676532745361, "learning_rate": 8.12928082191781e-06, "loss": 0.019, "step": 437 }, { "epoch": 0.1875, "grad_norm": 26.081762313842773, "learning_rate": 8.125000000000001e-06, "loss": 0.1851, "step": 438 }, { "epoch": 0.1879280821917808, "grad_norm": 32.47056579589844, "learning_rate": 8.120719178082193e-06, "loss": 0.0507, "step": 439 }, { "epoch": 0.18835616438356165, "grad_norm": 31.24920654296875, "learning_rate": 8.116438356164384e-06, "loss": 0.058, "step": 440 }, { "epoch": 0.18878424657534246, "grad_norm": 68.88462829589844, "learning_rate": 8.112157534246576e-06, "loss": 0.1941, "step": 441 }, { "epoch": 0.1892123287671233, "grad_norm": 200.69619750976562, "learning_rate": 8.107876712328768e-06, "loss": 0.8787, "step": 442 }, { "epoch": 0.1896404109589041, "grad_norm": 39.96286392211914, "learning_rate": 8.103595890410959e-06, "loss": 0.1562, "step": 443 }, { "epoch": 0.19006849315068494, "grad_norm": 71.73566436767578, "learning_rate": 8.09931506849315e-06, "loss": 0.292, "step": 444 }, { "epoch": 0.19049657534246575, "grad_norm": 26.57863426208496, "learning_rate": 8.095034246575342e-06, "loss": 0.0815, "step": 445 }, { "epoch": 0.1909246575342466, "grad_norm": 71.0126724243164, "learning_rate": 8.090753424657534e-06, "loss": 0.1779, "step": 446 }, { "epoch": 0.1913527397260274, "grad_norm": 73.09654235839844, "learning_rate": 8.086472602739727e-06, "loss": 0.3105, "step": 447 }, { "epoch": 0.1917808219178082, "grad_norm": 4.8552021980285645, "learning_rate": 8.082191780821919e-06, "loss": 0.0135, "step": 448 }, { "epoch": 0.19220890410958905, "grad_norm": 18.21192169189453, "learning_rate": 8.07791095890411e-06, "loss": 0.0299, "step": 449 }, { "epoch": 0.19263698630136986, "grad_norm": 44.2227668762207, "learning_rate": 8.073630136986302e-06, "loss": 0.2471, "step": 450 }, { "epoch": 0.1930650684931507, "grad_norm": 59.5881462097168, "learning_rate": 8.069349315068494e-06, "loss": 0.1214, "step": 451 }, { "epoch": 0.1934931506849315, "grad_norm": 57.74563217163086, "learning_rate": 8.065068493150686e-06, "loss": 0.5046, "step": 452 }, { "epoch": 0.19392123287671234, "grad_norm": 81.63369750976562, "learning_rate": 8.060787671232877e-06, "loss": 0.2528, "step": 453 }, { "epoch": 0.19434931506849315, "grad_norm": 23.080135345458984, "learning_rate": 8.056506849315069e-06, "loss": 0.0921, "step": 454 }, { "epoch": 0.19477739726027396, "grad_norm": 21.80877113342285, "learning_rate": 8.05222602739726e-06, "loss": 0.0503, "step": 455 }, { "epoch": 0.1952054794520548, "grad_norm": 65.44474792480469, "learning_rate": 8.047945205479452e-06, "loss": 0.3682, "step": 456 }, { "epoch": 0.1956335616438356, "grad_norm": 291.75982666015625, "learning_rate": 8.043664383561645e-06, "loss": 0.7959, "step": 457 }, { "epoch": 0.19606164383561644, "grad_norm": 27.08806610107422, "learning_rate": 8.039383561643837e-06, "loss": 0.0638, "step": 458 }, { "epoch": 0.19648972602739725, "grad_norm": 44.05672836303711, "learning_rate": 8.035102739726029e-06, "loss": 0.2368, "step": 459 }, { "epoch": 0.1969178082191781, "grad_norm": 111.96686553955078, "learning_rate": 8.03082191780822e-06, "loss": 0.7705, "step": 460 }, { "epoch": 0.1973458904109589, "grad_norm": 43.08165740966797, "learning_rate": 8.026541095890412e-06, "loss": 0.1881, "step": 461 }, { "epoch": 0.19777397260273974, "grad_norm": 8.988324165344238, "learning_rate": 8.022260273972603e-06, "loss": 0.0291, "step": 462 }, { "epoch": 0.19820205479452055, "grad_norm": 17.92043113708496, "learning_rate": 8.017979452054795e-06, "loss": 0.0409, "step": 463 }, { "epoch": 0.19863013698630136, "grad_norm": 191.11143493652344, "learning_rate": 8.013698630136987e-06, "loss": 1.0013, "step": 464 }, { "epoch": 0.1990582191780822, "grad_norm": 123.35166931152344, "learning_rate": 8.009417808219178e-06, "loss": 0.5732, "step": 465 }, { "epoch": 0.199486301369863, "grad_norm": 132.7969207763672, "learning_rate": 8.00513698630137e-06, "loss": 2.6979, "step": 466 }, { "epoch": 0.19991438356164384, "grad_norm": 50.9564323425293, "learning_rate": 8.000856164383562e-06, "loss": 0.1727, "step": 467 }, { "epoch": 0.20034246575342465, "grad_norm": 86.68875885009766, "learning_rate": 7.996575342465755e-06, "loss": 0.4635, "step": 468 }, { "epoch": 0.2007705479452055, "grad_norm": 96.85285186767578, "learning_rate": 7.992294520547947e-06, "loss": 0.5205, "step": 469 }, { "epoch": 0.2011986301369863, "grad_norm": 39.35941696166992, "learning_rate": 7.988013698630137e-06, "loss": 0.0823, "step": 470 }, { "epoch": 0.2016267123287671, "grad_norm": 61.601375579833984, "learning_rate": 7.983732876712328e-06, "loss": 0.3396, "step": 471 }, { "epoch": 0.20205479452054795, "grad_norm": 44.64535140991211, "learning_rate": 7.979452054794521e-06, "loss": 0.2207, "step": 472 }, { "epoch": 0.20248287671232876, "grad_norm": 14.725532531738281, "learning_rate": 7.975171232876713e-06, "loss": 0.0275, "step": 473 }, { "epoch": 0.2029109589041096, "grad_norm": 70.2264404296875, "learning_rate": 7.970890410958905e-06, "loss": 0.6777, "step": 474 }, { "epoch": 0.2033390410958904, "grad_norm": 41.509300231933594, "learning_rate": 7.966609589041096e-06, "loss": 0.2705, "step": 475 }, { "epoch": 0.20376712328767124, "grad_norm": 22.661699295043945, "learning_rate": 7.962328767123288e-06, "loss": 0.0526, "step": 476 }, { "epoch": 0.20419520547945205, "grad_norm": 78.91494750976562, "learning_rate": 7.95804794520548e-06, "loss": 0.5059, "step": 477 }, { "epoch": 0.2046232876712329, "grad_norm": 37.66532516479492, "learning_rate": 7.953767123287673e-06, "loss": 0.1389, "step": 478 }, { "epoch": 0.2050513698630137, "grad_norm": 56.2984504699707, "learning_rate": 7.949486301369865e-06, "loss": 0.1818, "step": 479 }, { "epoch": 0.2054794520547945, "grad_norm": 7.991452217102051, "learning_rate": 7.945205479452055e-06, "loss": 0.0261, "step": 480 }, { "epoch": 0.20590753424657535, "grad_norm": 108.90692138671875, "learning_rate": 7.940924657534246e-06, "loss": 0.5928, "step": 481 }, { "epoch": 0.20633561643835616, "grad_norm": 82.06063079833984, "learning_rate": 7.93664383561644e-06, "loss": 0.447, "step": 482 }, { "epoch": 0.206763698630137, "grad_norm": 22.254690170288086, "learning_rate": 7.932363013698631e-06, "loss": 0.0707, "step": 483 }, { "epoch": 0.2071917808219178, "grad_norm": 44.442745208740234, "learning_rate": 7.928082191780823e-06, "loss": 0.124, "step": 484 }, { "epoch": 0.20761986301369864, "grad_norm": 23.431697845458984, "learning_rate": 7.923801369863014e-06, "loss": 0.1113, "step": 485 }, { "epoch": 0.20804794520547945, "grad_norm": 30.050079345703125, "learning_rate": 7.919520547945206e-06, "loss": 0.158, "step": 486 }, { "epoch": 0.20847602739726026, "grad_norm": 29.796293258666992, "learning_rate": 7.915239726027398e-06, "loss": 0.0697, "step": 487 }, { "epoch": 0.2089041095890411, "grad_norm": 168.51315307617188, "learning_rate": 7.910958904109591e-06, "loss": 0.8149, "step": 488 }, { "epoch": 0.2093321917808219, "grad_norm": 75.45913696289062, "learning_rate": 7.906678082191781e-06, "loss": 0.2637, "step": 489 }, { "epoch": 0.20976027397260275, "grad_norm": 43.08189392089844, "learning_rate": 7.902397260273973e-06, "loss": 0.2964, "step": 490 }, { "epoch": 0.21018835616438356, "grad_norm": 34.841773986816406, "learning_rate": 7.898116438356164e-06, "loss": 0.0824, "step": 491 }, { "epoch": 0.2106164383561644, "grad_norm": 107.3261489868164, "learning_rate": 7.893835616438357e-06, "loss": 0.5146, "step": 492 }, { "epoch": 0.2110445205479452, "grad_norm": 22.32187843322754, "learning_rate": 7.889554794520549e-06, "loss": 0.057, "step": 493 }, { "epoch": 0.21147260273972604, "grad_norm": 57.04218673706055, "learning_rate": 7.88527397260274e-06, "loss": 0.6281, "step": 494 }, { "epoch": 0.21190068493150685, "grad_norm": 6.401570796966553, "learning_rate": 7.880993150684932e-06, "loss": 0.0052, "step": 495 }, { "epoch": 0.21232876712328766, "grad_norm": 68.44953918457031, "learning_rate": 7.876712328767124e-06, "loss": 0.5293, "step": 496 }, { "epoch": 0.2127568493150685, "grad_norm": 26.146785736083984, "learning_rate": 7.872431506849316e-06, "loss": 0.072, "step": 497 }, { "epoch": 0.2131849315068493, "grad_norm": 46.900047302246094, "learning_rate": 7.868150684931507e-06, "loss": 0.1309, "step": 498 }, { "epoch": 0.21361301369863014, "grad_norm": 131.1702880859375, "learning_rate": 7.863869863013699e-06, "loss": 0.4961, "step": 499 }, { "epoch": 0.21404109589041095, "grad_norm": 16.822818756103516, "learning_rate": 7.85958904109589e-06, "loss": 0.0552, "step": 500 }, { "epoch": 0.2144691780821918, "grad_norm": 26.82673454284668, "learning_rate": 7.855308219178082e-06, "loss": 0.09, "step": 501 }, { "epoch": 0.2148972602739726, "grad_norm": 16.971660614013672, "learning_rate": 7.851027397260274e-06, "loss": 0.0307, "step": 502 }, { "epoch": 0.2153253424657534, "grad_norm": 2.7754628658294678, "learning_rate": 7.846746575342467e-06, "loss": 0.0058, "step": 503 }, { "epoch": 0.21575342465753425, "grad_norm": 67.32703399658203, "learning_rate": 7.842465753424659e-06, "loss": 0.1811, "step": 504 }, { "epoch": 0.21618150684931506, "grad_norm": 78.63072204589844, "learning_rate": 7.83818493150685e-06, "loss": 0.2158, "step": 505 }, { "epoch": 0.2166095890410959, "grad_norm": 23.777822494506836, "learning_rate": 7.833904109589042e-06, "loss": 0.0664, "step": 506 }, { "epoch": 0.2170376712328767, "grad_norm": 85.40617370605469, "learning_rate": 7.829623287671234e-06, "loss": 0.324, "step": 507 }, { "epoch": 0.21746575342465754, "grad_norm": 58.77161407470703, "learning_rate": 7.825342465753425e-06, "loss": 0.1624, "step": 508 }, { "epoch": 0.21789383561643835, "grad_norm": 43.78720474243164, "learning_rate": 7.821061643835617e-06, "loss": 0.1247, "step": 509 }, { "epoch": 0.2183219178082192, "grad_norm": 27.038026809692383, "learning_rate": 7.816780821917809e-06, "loss": 0.1211, "step": 510 }, { "epoch": 0.21875, "grad_norm": 117.07380676269531, "learning_rate": 7.8125e-06, "loss": 0.9597, "step": 511 }, { "epoch": 0.2191780821917808, "grad_norm": 29.46851348876953, "learning_rate": 7.808219178082192e-06, "loss": 0.0518, "step": 512 }, { "epoch": 0.21960616438356165, "grad_norm": 270.5569152832031, "learning_rate": 7.803938356164385e-06, "loss": 1.4758, "step": 513 }, { "epoch": 0.22003424657534246, "grad_norm": 22.611888885498047, "learning_rate": 7.799657534246577e-06, "loss": 0.0431, "step": 514 }, { "epoch": 0.2204623287671233, "grad_norm": 12.542022705078125, "learning_rate": 7.795376712328768e-06, "loss": 0.0147, "step": 515 }, { "epoch": 0.2208904109589041, "grad_norm": 68.031005859375, "learning_rate": 7.791095890410958e-06, "loss": 0.2042, "step": 516 }, { "epoch": 0.22131849315068494, "grad_norm": 106.87434387207031, "learning_rate": 7.786815068493152e-06, "loss": 0.4688, "step": 517 }, { "epoch": 0.22174657534246575, "grad_norm": 1.9663132429122925, "learning_rate": 7.782534246575343e-06, "loss": 0.002, "step": 518 }, { "epoch": 0.2221746575342466, "grad_norm": 17.649272918701172, "learning_rate": 7.778253424657535e-06, "loss": 0.0377, "step": 519 }, { "epoch": 0.2226027397260274, "grad_norm": 26.762678146362305, "learning_rate": 7.773972602739727e-06, "loss": 0.0464, "step": 520 }, { "epoch": 0.2230308219178082, "grad_norm": 61.747135162353516, "learning_rate": 7.769691780821918e-06, "loss": 0.1821, "step": 521 }, { "epoch": 0.22345890410958905, "grad_norm": 100.49110412597656, "learning_rate": 7.76541095890411e-06, "loss": 0.329, "step": 522 }, { "epoch": 0.22388698630136986, "grad_norm": 11.70064640045166, "learning_rate": 7.761130136986303e-06, "loss": 0.0149, "step": 523 }, { "epoch": 0.2243150684931507, "grad_norm": 7.090034008026123, "learning_rate": 7.756849315068495e-06, "loss": 0.0252, "step": 524 }, { "epoch": 0.2247431506849315, "grad_norm": 89.45475006103516, "learning_rate": 7.752568493150685e-06, "loss": 0.7598, "step": 525 }, { "epoch": 0.22517123287671234, "grad_norm": 98.85481262207031, "learning_rate": 7.748287671232876e-06, "loss": 0.272, "step": 526 }, { "epoch": 0.22559931506849315, "grad_norm": 133.96432495117188, "learning_rate": 7.74400684931507e-06, "loss": 0.3225, "step": 527 }, { "epoch": 0.22602739726027396, "grad_norm": 53.594242095947266, "learning_rate": 7.739726027397261e-06, "loss": 0.1082, "step": 528 }, { "epoch": 0.2264554794520548, "grad_norm": 9.469794273376465, "learning_rate": 7.735445205479453e-06, "loss": 0.0079, "step": 529 }, { "epoch": 0.2268835616438356, "grad_norm": 8.829156875610352, "learning_rate": 7.731164383561645e-06, "loss": 0.0173, "step": 530 }, { "epoch": 0.22731164383561644, "grad_norm": 33.17487335205078, "learning_rate": 7.726883561643836e-06, "loss": 0.0636, "step": 531 }, { "epoch": 0.22773972602739725, "grad_norm": 14.252382278442383, "learning_rate": 7.722602739726028e-06, "loss": 0.0274, "step": 532 }, { "epoch": 0.2281678082191781, "grad_norm": 30.04306411743164, "learning_rate": 7.718321917808221e-06, "loss": 0.1294, "step": 533 }, { "epoch": 0.2285958904109589, "grad_norm": 2.298100709915161, "learning_rate": 7.714041095890411e-06, "loss": 0.0013, "step": 534 }, { "epoch": 0.22902397260273974, "grad_norm": 22.12137222290039, "learning_rate": 7.709760273972603e-06, "loss": 0.0947, "step": 535 }, { "epoch": 0.22945205479452055, "grad_norm": 98.56476593017578, "learning_rate": 7.705479452054794e-06, "loss": 0.541, "step": 536 }, { "epoch": 0.22988013698630136, "grad_norm": 66.25592041015625, "learning_rate": 7.701198630136986e-06, "loss": 0.6309, "step": 537 }, { "epoch": 0.2303082191780822, "grad_norm": 74.68619537353516, "learning_rate": 7.69691780821918e-06, "loss": 0.1833, "step": 538 }, { "epoch": 0.230736301369863, "grad_norm": 49.24323654174805, "learning_rate": 7.692636986301371e-06, "loss": 0.1067, "step": 539 }, { "epoch": 0.23116438356164384, "grad_norm": 90.04692840576172, "learning_rate": 7.688356164383563e-06, "loss": 0.21, "step": 540 }, { "epoch": 0.23159246575342465, "grad_norm": 22.979005813598633, "learning_rate": 7.684075342465754e-06, "loss": 0.036, "step": 541 }, { "epoch": 0.2320205479452055, "grad_norm": 96.52161407470703, "learning_rate": 7.679794520547946e-06, "loss": 0.2627, "step": 542 }, { "epoch": 0.2324486301369863, "grad_norm": 61.27184295654297, "learning_rate": 7.675513698630137e-06, "loss": 0.3613, "step": 543 }, { "epoch": 0.2328767123287671, "grad_norm": 56.99726486206055, "learning_rate": 7.671232876712329e-06, "loss": 0.187, "step": 544 }, { "epoch": 0.23330479452054795, "grad_norm": 16.743865966796875, "learning_rate": 7.66695205479452e-06, "loss": 0.035, "step": 545 }, { "epoch": 0.23373287671232876, "grad_norm": 74.56373596191406, "learning_rate": 7.662671232876712e-06, "loss": 0.4227, "step": 546 }, { "epoch": 0.2341609589041096, "grad_norm": 68.41123962402344, "learning_rate": 7.658390410958904e-06, "loss": 0.3865, "step": 547 }, { "epoch": 0.2345890410958904, "grad_norm": 51.32826614379883, "learning_rate": 7.654109589041097e-06, "loss": 0.6667, "step": 548 }, { "epoch": 0.23501712328767124, "grad_norm": 89.78392028808594, "learning_rate": 7.649828767123289e-06, "loss": 0.3252, "step": 549 }, { "epoch": 0.23544520547945205, "grad_norm": 153.05174255371094, "learning_rate": 7.64554794520548e-06, "loss": 0.5256, "step": 550 }, { "epoch": 0.2358732876712329, "grad_norm": 110.25298309326172, "learning_rate": 7.641267123287672e-06, "loss": 0.6084, "step": 551 }, { "epoch": 0.2363013698630137, "grad_norm": 23.095800399780273, "learning_rate": 7.636986301369864e-06, "loss": 0.0331, "step": 552 }, { "epoch": 0.2367294520547945, "grad_norm": 22.217445373535156, "learning_rate": 7.632705479452055e-06, "loss": 0.0607, "step": 553 }, { "epoch": 0.23715753424657535, "grad_norm": 192.19688415527344, "learning_rate": 7.628424657534247e-06, "loss": 0.6672, "step": 554 }, { "epoch": 0.23758561643835616, "grad_norm": 80.33145141601562, "learning_rate": 7.624143835616439e-06, "loss": 0.4561, "step": 555 }, { "epoch": 0.238013698630137, "grad_norm": 17.056669235229492, "learning_rate": 7.61986301369863e-06, "loss": 0.0187, "step": 556 }, { "epoch": 0.2384417808219178, "grad_norm": 92.52870178222656, "learning_rate": 7.615582191780823e-06, "loss": 0.3694, "step": 557 }, { "epoch": 0.23886986301369864, "grad_norm": 34.541908264160156, "learning_rate": 7.611301369863014e-06, "loss": 0.1918, "step": 558 }, { "epoch": 0.23929794520547945, "grad_norm": 73.72164916992188, "learning_rate": 7.607020547945206e-06, "loss": 0.4377, "step": 559 }, { "epoch": 0.23972602739726026, "grad_norm": 21.241411209106445, "learning_rate": 7.6027397260273985e-06, "loss": 0.0581, "step": 560 }, { "epoch": 0.2401541095890411, "grad_norm": 5.856642723083496, "learning_rate": 7.59845890410959e-06, "loss": 0.0186, "step": 561 }, { "epoch": 0.2405821917808219, "grad_norm": 20.41284942626953, "learning_rate": 7.594178082191781e-06, "loss": 0.0464, "step": 562 }, { "epoch": 0.24101027397260275, "grad_norm": 75.43649291992188, "learning_rate": 7.5898972602739726e-06, "loss": 0.2036, "step": 563 }, { "epoch": 0.24143835616438356, "grad_norm": 73.34547424316406, "learning_rate": 7.585616438356165e-06, "loss": 0.1382, "step": 564 }, { "epoch": 0.2418664383561644, "grad_norm": 97.1091537475586, "learning_rate": 7.581335616438357e-06, "loss": 0.1597, "step": 565 }, { "epoch": 0.2422945205479452, "grad_norm": 11.451400756835938, "learning_rate": 7.577054794520548e-06, "loss": 0.0212, "step": 566 }, { "epoch": 0.24272260273972604, "grad_norm": 8.382686614990234, "learning_rate": 7.572773972602741e-06, "loss": 0.0198, "step": 567 }, { "epoch": 0.24315068493150685, "grad_norm": 16.28538703918457, "learning_rate": 7.568493150684932e-06, "loss": 0.0171, "step": 568 }, { "epoch": 0.24357876712328766, "grad_norm": 29.36876678466797, "learning_rate": 7.564212328767124e-06, "loss": 0.0944, "step": 569 }, { "epoch": 0.2440068493150685, "grad_norm": 2.247208833694458, "learning_rate": 7.5599315068493165e-06, "loss": 0.0085, "step": 570 }, { "epoch": 0.2444349315068493, "grad_norm": 33.92767333984375, "learning_rate": 7.555650684931507e-06, "loss": 0.0775, "step": 571 }, { "epoch": 0.24486301369863014, "grad_norm": 42.226051330566406, "learning_rate": 7.551369863013699e-06, "loss": 0.1113, "step": 572 }, { "epoch": 0.24529109589041095, "grad_norm": 36.8174934387207, "learning_rate": 7.5470890410958906e-06, "loss": 0.1409, "step": 573 }, { "epoch": 0.2457191780821918, "grad_norm": 29.062646865844727, "learning_rate": 7.542808219178083e-06, "loss": 0.045, "step": 574 }, { "epoch": 0.2461472602739726, "grad_norm": 31.840742111206055, "learning_rate": 7.538527397260275e-06, "loss": 0.0625, "step": 575 }, { "epoch": 0.2465753424657534, "grad_norm": 104.6715087890625, "learning_rate": 7.534246575342466e-06, "loss": 0.2224, "step": 576 }, { "epoch": 0.24700342465753425, "grad_norm": 70.09893798828125, "learning_rate": 7.529965753424659e-06, "loss": 0.3589, "step": 577 }, { "epoch": 0.24743150684931506, "grad_norm": 118.00634765625, "learning_rate": 7.52568493150685e-06, "loss": 0.1735, "step": 578 }, { "epoch": 0.2478595890410959, "grad_norm": 95.87020111083984, "learning_rate": 7.521404109589042e-06, "loss": 0.1334, "step": 579 }, { "epoch": 0.2482876712328767, "grad_norm": 97.64168548583984, "learning_rate": 7.517123287671233e-06, "loss": 0.2988, "step": 580 }, { "epoch": 0.24871575342465754, "grad_norm": 27.673921585083008, "learning_rate": 7.512842465753425e-06, "loss": 0.0275, "step": 581 }, { "epoch": 0.24914383561643835, "grad_norm": 7.856906890869141, "learning_rate": 7.508561643835617e-06, "loss": 0.0075, "step": 582 }, { "epoch": 0.2495719178082192, "grad_norm": 57.55317687988281, "learning_rate": 7.5042808219178086e-06, "loss": 0.206, "step": 583 }, { "epoch": 0.25, "grad_norm": 4.763654708862305, "learning_rate": 7.500000000000001e-06, "loss": 0.0112, "step": 584 }, { "epoch": 0.2504280821917808, "grad_norm": 95.7210693359375, "learning_rate": 7.495719178082193e-06, "loss": 0.2783, "step": 585 }, { "epoch": 0.2508561643835616, "grad_norm": 2.9004921913146973, "learning_rate": 7.491438356164384e-06, "loss": 0.0021, "step": 586 }, { "epoch": 0.2512842465753425, "grad_norm": 56.52210998535156, "learning_rate": 7.487157534246576e-06, "loss": 0.2039, "step": 587 }, { "epoch": 0.2517123287671233, "grad_norm": 7.09292459487915, "learning_rate": 7.482876712328768e-06, "loss": 0.0237, "step": 588 }, { "epoch": 0.2521404109589041, "grad_norm": 14.720183372497559, "learning_rate": 7.478595890410959e-06, "loss": 0.0353, "step": 589 }, { "epoch": 0.2525684931506849, "grad_norm": 20.860889434814453, "learning_rate": 7.474315068493151e-06, "loss": 0.0401, "step": 590 }, { "epoch": 0.2529965753424658, "grad_norm": 334.3153076171875, "learning_rate": 7.4700342465753424e-06, "loss": 0.7425, "step": 591 }, { "epoch": 0.2534246575342466, "grad_norm": 90.02655029296875, "learning_rate": 7.465753424657535e-06, "loss": 0.2339, "step": 592 }, { "epoch": 0.2538527397260274, "grad_norm": 49.38987731933594, "learning_rate": 7.4614726027397266e-06, "loss": 0.1111, "step": 593 }, { "epoch": 0.2542808219178082, "grad_norm": 318.6791687011719, "learning_rate": 7.457191780821918e-06, "loss": 1.2622, "step": 594 }, { "epoch": 0.254708904109589, "grad_norm": 9.896432876586914, "learning_rate": 7.452910958904111e-06, "loss": 0.0081, "step": 595 }, { "epoch": 0.2551369863013699, "grad_norm": 6.4284796714782715, "learning_rate": 7.448630136986302e-06, "loss": 0.0064, "step": 596 }, { "epoch": 0.2555650684931507, "grad_norm": 11.007637023925781, "learning_rate": 7.444349315068494e-06, "loss": 0.0086, "step": 597 }, { "epoch": 0.2559931506849315, "grad_norm": 182.78367614746094, "learning_rate": 7.440068493150685e-06, "loss": 0.9126, "step": 598 }, { "epoch": 0.2564212328767123, "grad_norm": 49.29271697998047, "learning_rate": 7.435787671232877e-06, "loss": 0.2307, "step": 599 }, { "epoch": 0.2568493150684932, "grad_norm": 100.42945861816406, "learning_rate": 7.431506849315069e-06, "loss": 0.2881, "step": 600 }, { "epoch": 0.257277397260274, "grad_norm": 45.95989227294922, "learning_rate": 7.4272260273972604e-06, "loss": 0.1907, "step": 601 }, { "epoch": 0.2577054794520548, "grad_norm": 94.66627502441406, "learning_rate": 7.422945205479453e-06, "loss": 0.7324, "step": 602 }, { "epoch": 0.2581335616438356, "grad_norm": 95.62594604492188, "learning_rate": 7.4186643835616446e-06, "loss": 0.5456, "step": 603 }, { "epoch": 0.2585616438356164, "grad_norm": 32.52836227416992, "learning_rate": 7.414383561643836e-06, "loss": 0.0642, "step": 604 }, { "epoch": 0.2589897260273973, "grad_norm": 262.517578125, "learning_rate": 7.410102739726029e-06, "loss": 0.4297, "step": 605 }, { "epoch": 0.2594178082191781, "grad_norm": 36.34096908569336, "learning_rate": 7.40582191780822e-06, "loss": 0.0908, "step": 606 }, { "epoch": 0.2598458904109589, "grad_norm": 5.151368618011475, "learning_rate": 7.401541095890411e-06, "loss": 0.0124, "step": 607 }, { "epoch": 0.2602739726027397, "grad_norm": 41.76686477661133, "learning_rate": 7.397260273972603e-06, "loss": 0.1053, "step": 608 }, { "epoch": 0.2607020547945205, "grad_norm": 74.2865982055664, "learning_rate": 7.392979452054795e-06, "loss": 0.6653, "step": 609 }, { "epoch": 0.2611301369863014, "grad_norm": 34.553993225097656, "learning_rate": 7.388698630136987e-06, "loss": 0.21, "step": 610 }, { "epoch": 0.2615582191780822, "grad_norm": 236.61729431152344, "learning_rate": 7.3844178082191784e-06, "loss": 0.6637, "step": 611 }, { "epoch": 0.261986301369863, "grad_norm": 55.93560028076172, "learning_rate": 7.380136986301371e-06, "loss": 0.1228, "step": 612 }, { "epoch": 0.2624143835616438, "grad_norm": 108.90987396240234, "learning_rate": 7.3758561643835626e-06, "loss": 0.2217, "step": 613 }, { "epoch": 0.2628424657534247, "grad_norm": 51.13361358642578, "learning_rate": 7.371575342465754e-06, "loss": 0.2389, "step": 614 }, { "epoch": 0.2632705479452055, "grad_norm": 19.939138412475586, "learning_rate": 7.367294520547947e-06, "loss": 0.0284, "step": 615 }, { "epoch": 0.2636986301369863, "grad_norm": 14.23261547088623, "learning_rate": 7.3630136986301374e-06, "loss": 0.0226, "step": 616 }, { "epoch": 0.2641267123287671, "grad_norm": 66.28099822998047, "learning_rate": 7.358732876712329e-06, "loss": 0.2708, "step": 617 }, { "epoch": 0.2645547945205479, "grad_norm": 8.92492961883545, "learning_rate": 7.354452054794521e-06, "loss": 0.0104, "step": 618 }, { "epoch": 0.2649828767123288, "grad_norm": 6.167044639587402, "learning_rate": 7.350171232876712e-06, "loss": 0.0155, "step": 619 }, { "epoch": 0.2654109589041096, "grad_norm": 24.597652435302734, "learning_rate": 7.345890410958905e-06, "loss": 0.0627, "step": 620 }, { "epoch": 0.2658390410958904, "grad_norm": 15.109396934509277, "learning_rate": 7.3416095890410964e-06, "loss": 0.0376, "step": 621 }, { "epoch": 0.2662671232876712, "grad_norm": 20.913543701171875, "learning_rate": 7.337328767123288e-06, "loss": 0.0832, "step": 622 }, { "epoch": 0.2666952054794521, "grad_norm": 42.59825134277344, "learning_rate": 7.3330479452054806e-06, "loss": 0.1003, "step": 623 }, { "epoch": 0.2671232876712329, "grad_norm": 54.10437774658203, "learning_rate": 7.328767123287672e-06, "loss": 0.1199, "step": 624 }, { "epoch": 0.2675513698630137, "grad_norm": 28.593109130859375, "learning_rate": 7.324486301369864e-06, "loss": 0.0388, "step": 625 }, { "epoch": 0.2679794520547945, "grad_norm": 47.63380432128906, "learning_rate": 7.320205479452055e-06, "loss": 0.167, "step": 626 }, { "epoch": 0.2684075342465753, "grad_norm": 101.06744384765625, "learning_rate": 7.315924657534247e-06, "loss": 0.7295, "step": 627 }, { "epoch": 0.2688356164383562, "grad_norm": 3.1795494556427, "learning_rate": 7.311643835616439e-06, "loss": 0.0028, "step": 628 }, { "epoch": 0.269263698630137, "grad_norm": 41.142974853515625, "learning_rate": 7.30736301369863e-06, "loss": 0.0424, "step": 629 }, { "epoch": 0.2696917808219178, "grad_norm": 132.26124572753906, "learning_rate": 7.303082191780823e-06, "loss": 0.4273, "step": 630 }, { "epoch": 0.2701198630136986, "grad_norm": 12.163527488708496, "learning_rate": 7.2988013698630144e-06, "loss": 0.0116, "step": 631 }, { "epoch": 0.2705479452054795, "grad_norm": 11.390554428100586, "learning_rate": 7.294520547945206e-06, "loss": 0.0105, "step": 632 }, { "epoch": 0.2709760273972603, "grad_norm": 71.94181060791016, "learning_rate": 7.2902397260273985e-06, "loss": 0.0894, "step": 633 }, { "epoch": 0.2714041095890411, "grad_norm": 8.997446060180664, "learning_rate": 7.28595890410959e-06, "loss": 0.0164, "step": 634 }, { "epoch": 0.2718321917808219, "grad_norm": 2.068051815032959, "learning_rate": 7.281678082191781e-06, "loss": 0.0062, "step": 635 }, { "epoch": 0.2722602739726027, "grad_norm": 49.66343688964844, "learning_rate": 7.277397260273973e-06, "loss": 0.1014, "step": 636 }, { "epoch": 0.2726883561643836, "grad_norm": 73.87043762207031, "learning_rate": 7.273116438356165e-06, "loss": 0.3019, "step": 637 }, { "epoch": 0.2731164383561644, "grad_norm": 71.8984375, "learning_rate": 7.268835616438357e-06, "loss": 0.6669, "step": 638 }, { "epoch": 0.2735445205479452, "grad_norm": 5.218684196472168, "learning_rate": 7.264554794520548e-06, "loss": 0.0094, "step": 639 }, { "epoch": 0.273972602739726, "grad_norm": 7.96542501449585, "learning_rate": 7.260273972602741e-06, "loss": 0.0214, "step": 640 }, { "epoch": 0.2744006849315068, "grad_norm": 47.47689437866211, "learning_rate": 7.2559931506849324e-06, "loss": 0.0776, "step": 641 }, { "epoch": 0.2748287671232877, "grad_norm": 66.14462280273438, "learning_rate": 7.251712328767124e-06, "loss": 0.2258, "step": 642 }, { "epoch": 0.2752568493150685, "grad_norm": 31.02815818786621, "learning_rate": 7.2474315068493165e-06, "loss": 0.0275, "step": 643 }, { "epoch": 0.2756849315068493, "grad_norm": 8.517438888549805, "learning_rate": 7.243150684931507e-06, "loss": 0.0134, "step": 644 }, { "epoch": 0.2761130136986301, "grad_norm": 21.09378433227539, "learning_rate": 7.238869863013699e-06, "loss": 0.0312, "step": 645 }, { "epoch": 0.276541095890411, "grad_norm": 17.719432830810547, "learning_rate": 7.234589041095891e-06, "loss": 0.023, "step": 646 }, { "epoch": 0.2769691780821918, "grad_norm": 105.49552154541016, "learning_rate": 7.230308219178083e-06, "loss": 0.2104, "step": 647 }, { "epoch": 0.2773972602739726, "grad_norm": 7.747145652770996, "learning_rate": 7.226027397260275e-06, "loss": 0.0077, "step": 648 }, { "epoch": 0.2778253424657534, "grad_norm": 11.375240325927734, "learning_rate": 7.221746575342466e-06, "loss": 0.0171, "step": 649 }, { "epoch": 0.2782534246575342, "grad_norm": 41.927738189697266, "learning_rate": 7.217465753424659e-06, "loss": 0.2375, "step": 650 }, { "epoch": 0.2786815068493151, "grad_norm": 20.683643341064453, "learning_rate": 7.2131849315068504e-06, "loss": 0.021, "step": 651 }, { "epoch": 0.2791095890410959, "grad_norm": 55.011531829833984, "learning_rate": 7.208904109589042e-06, "loss": 0.0858, "step": 652 }, { "epoch": 0.2795376712328767, "grad_norm": 26.65351676940918, "learning_rate": 7.204623287671233e-06, "loss": 0.0455, "step": 653 }, { "epoch": 0.2799657534246575, "grad_norm": 83.19271850585938, "learning_rate": 7.2003424657534245e-06, "loss": 0.2292, "step": 654 }, { "epoch": 0.2803938356164384, "grad_norm": 13.489514350891113, "learning_rate": 7.196061643835617e-06, "loss": 0.0203, "step": 655 }, { "epoch": 0.2808219178082192, "grad_norm": 27.34222412109375, "learning_rate": 7.191780821917809e-06, "loss": 0.0538, "step": 656 }, { "epoch": 0.28125, "grad_norm": 86.69216918945312, "learning_rate": 7.1875e-06, "loss": 0.1419, "step": 657 }, { "epoch": 0.2816780821917808, "grad_norm": 79.14155578613281, "learning_rate": 7.183219178082193e-06, "loss": 0.1389, "step": 658 }, { "epoch": 0.2821061643835616, "grad_norm": 92.4407958984375, "learning_rate": 7.178938356164384e-06, "loss": 0.2664, "step": 659 }, { "epoch": 0.2825342465753425, "grad_norm": 17.849285125732422, "learning_rate": 7.174657534246576e-06, "loss": 0.0235, "step": 660 }, { "epoch": 0.2829623287671233, "grad_norm": 10.503944396972656, "learning_rate": 7.1703767123287684e-06, "loss": 0.0094, "step": 661 }, { "epoch": 0.2833904109589041, "grad_norm": 6.342771053314209, "learning_rate": 7.166095890410959e-06, "loss": 0.0214, "step": 662 }, { "epoch": 0.2838184931506849, "grad_norm": 25.34906768798828, "learning_rate": 7.161815068493151e-06, "loss": 0.082, "step": 663 }, { "epoch": 0.2842465753424658, "grad_norm": 36.787376403808594, "learning_rate": 7.1575342465753425e-06, "loss": 0.0373, "step": 664 }, { "epoch": 0.2846746575342466, "grad_norm": 47.88157272338867, "learning_rate": 7.153253424657535e-06, "loss": 0.157, "step": 665 }, { "epoch": 0.2851027397260274, "grad_norm": 31.276655197143555, "learning_rate": 7.148972602739727e-06, "loss": 0.0544, "step": 666 }, { "epoch": 0.2855308219178082, "grad_norm": 134.88014221191406, "learning_rate": 7.144691780821918e-06, "loss": 0.3058, "step": 667 }, { "epoch": 0.285958904109589, "grad_norm": 62.04694747924805, "learning_rate": 7.140410958904111e-06, "loss": 0.1371, "step": 668 }, { "epoch": 0.2863869863013699, "grad_norm": 2.598020553588867, "learning_rate": 7.136130136986302e-06, "loss": 0.0025, "step": 669 }, { "epoch": 0.2868150684931507, "grad_norm": 32.35898208618164, "learning_rate": 7.131849315068494e-06, "loss": 0.0457, "step": 670 }, { "epoch": 0.2872431506849315, "grad_norm": 3.9286282062530518, "learning_rate": 7.127568493150685e-06, "loss": 0.0045, "step": 671 }, { "epoch": 0.2876712328767123, "grad_norm": 293.2462463378906, "learning_rate": 7.123287671232877e-06, "loss": 0.4492, "step": 672 }, { "epoch": 0.2880993150684932, "grad_norm": 46.22662353515625, "learning_rate": 7.119006849315069e-06, "loss": 0.0533, "step": 673 }, { "epoch": 0.288527397260274, "grad_norm": 22.07549476623535, "learning_rate": 7.1147260273972605e-06, "loss": 0.03, "step": 674 }, { "epoch": 0.2889554794520548, "grad_norm": 17.995840072631836, "learning_rate": 7.110445205479453e-06, "loss": 0.0187, "step": 675 }, { "epoch": 0.2893835616438356, "grad_norm": 28.85215187072754, "learning_rate": 7.106164383561645e-06, "loss": 0.0422, "step": 676 }, { "epoch": 0.2898116438356164, "grad_norm": 30.075183868408203, "learning_rate": 7.101883561643836e-06, "loss": 0.0557, "step": 677 }, { "epoch": 0.2902397260273973, "grad_norm": 19.16193389892578, "learning_rate": 7.097602739726029e-06, "loss": 0.0449, "step": 678 }, { "epoch": 0.2906678082191781, "grad_norm": 13.669227600097656, "learning_rate": 7.09332191780822e-06, "loss": 0.0195, "step": 679 }, { "epoch": 0.2910958904109589, "grad_norm": 13.037485122680664, "learning_rate": 7.089041095890411e-06, "loss": 0.0192, "step": 680 }, { "epoch": 0.2915239726027397, "grad_norm": 18.063499450683594, "learning_rate": 7.084760273972603e-06, "loss": 0.023, "step": 681 }, { "epoch": 0.2919520547945205, "grad_norm": 3.417897939682007, "learning_rate": 7.080479452054795e-06, "loss": 0.0108, "step": 682 }, { "epoch": 0.2923801369863014, "grad_norm": 115.08209228515625, "learning_rate": 7.076198630136987e-06, "loss": 0.827, "step": 683 }, { "epoch": 0.2928082191780822, "grad_norm": 42.29871368408203, "learning_rate": 7.0719178082191785e-06, "loss": 0.1132, "step": 684 }, { "epoch": 0.293236301369863, "grad_norm": 4.0236101150512695, "learning_rate": 7.06763698630137e-06, "loss": 0.0018, "step": 685 }, { "epoch": 0.2936643835616438, "grad_norm": 76.28496551513672, "learning_rate": 7.063356164383563e-06, "loss": 0.9036, "step": 686 }, { "epoch": 0.2940924657534247, "grad_norm": 82.69190216064453, "learning_rate": 7.059075342465754e-06, "loss": 0.3111, "step": 687 }, { "epoch": 0.2945205479452055, "grad_norm": 3.840296506881714, "learning_rate": 7.054794520547946e-06, "loss": 0.0019, "step": 688 }, { "epoch": 0.2949486301369863, "grad_norm": 123.97752380371094, "learning_rate": 7.050513698630137e-06, "loss": 0.1168, "step": 689 }, { "epoch": 0.2953767123287671, "grad_norm": 3.9040679931640625, "learning_rate": 7.046232876712329e-06, "loss": 0.0081, "step": 690 }, { "epoch": 0.2958047945205479, "grad_norm": 36.315547943115234, "learning_rate": 7.041952054794521e-06, "loss": 0.1548, "step": 691 }, { "epoch": 0.2962328767123288, "grad_norm": 32.0535774230957, "learning_rate": 7.037671232876712e-06, "loss": 0.0541, "step": 692 }, { "epoch": 0.2966609589041096, "grad_norm": 10.910841941833496, "learning_rate": 7.033390410958905e-06, "loss": 0.0115, "step": 693 }, { "epoch": 0.2970890410958904, "grad_norm": 28.188838958740234, "learning_rate": 7.0291095890410965e-06, "loss": 0.0499, "step": 694 }, { "epoch": 0.2975171232876712, "grad_norm": 291.2646179199219, "learning_rate": 7.024828767123288e-06, "loss": 0.5909, "step": 695 }, { "epoch": 0.2979452054794521, "grad_norm": 14.459863662719727, "learning_rate": 7.020547945205481e-06, "loss": 0.0136, "step": 696 }, { "epoch": 0.2983732876712329, "grad_norm": 107.52633666992188, "learning_rate": 7.016267123287672e-06, "loss": 0.5579, "step": 697 }, { "epoch": 0.2988013698630137, "grad_norm": 8.023235321044922, "learning_rate": 7.011986301369864e-06, "loss": 0.0049, "step": 698 }, { "epoch": 0.2992294520547945, "grad_norm": 11.51945972442627, "learning_rate": 7.007705479452055e-06, "loss": 0.0155, "step": 699 }, { "epoch": 0.2996575342465753, "grad_norm": 10.948917388916016, "learning_rate": 7.003424657534247e-06, "loss": 0.0145, "step": 700 }, { "epoch": 0.3000856164383562, "grad_norm": 80.97380828857422, "learning_rate": 6.999143835616439e-06, "loss": 0.0395, "step": 701 }, { "epoch": 0.300513698630137, "grad_norm": 8.497552871704102, "learning_rate": 6.99486301369863e-06, "loss": 0.0062, "step": 702 }, { "epoch": 0.3009417808219178, "grad_norm": 78.86396789550781, "learning_rate": 6.990582191780823e-06, "loss": 0.135, "step": 703 }, { "epoch": 0.3013698630136986, "grad_norm": 565.851318359375, "learning_rate": 6.9863013698630145e-06, "loss": 1.9149, "step": 704 }, { "epoch": 0.3017979452054795, "grad_norm": 2.973067283630371, "learning_rate": 6.982020547945206e-06, "loss": 0.0019, "step": 705 }, { "epoch": 0.3022260273972603, "grad_norm": 126.46358489990234, "learning_rate": 6.9777397260273986e-06, "loss": 0.3523, "step": 706 }, { "epoch": 0.3026541095890411, "grad_norm": 8.117659568786621, "learning_rate": 6.97345890410959e-06, "loss": 0.0114, "step": 707 }, { "epoch": 0.3030821917808219, "grad_norm": 46.60285949707031, "learning_rate": 6.969178082191781e-06, "loss": 0.0526, "step": 708 }, { "epoch": 0.3035102739726027, "grad_norm": 28.88027572631836, "learning_rate": 6.964897260273973e-06, "loss": 0.03, "step": 709 }, { "epoch": 0.3039383561643836, "grad_norm": 81.90469360351562, "learning_rate": 6.960616438356165e-06, "loss": 0.5602, "step": 710 }, { "epoch": 0.3043664383561644, "grad_norm": 12.723942756652832, "learning_rate": 6.956335616438357e-06, "loss": 0.0143, "step": 711 }, { "epoch": 0.3047945205479452, "grad_norm": 20.726442337036133, "learning_rate": 6.952054794520548e-06, "loss": 0.0334, "step": 712 }, { "epoch": 0.305222602739726, "grad_norm": 24.761749267578125, "learning_rate": 6.947773972602741e-06, "loss": 0.0431, "step": 713 }, { "epoch": 0.3056506849315068, "grad_norm": 61.85054016113281, "learning_rate": 6.9434931506849325e-06, "loss": 0.1456, "step": 714 }, { "epoch": 0.3060787671232877, "grad_norm": 13.25638198852539, "learning_rate": 6.939212328767124e-06, "loss": 0.0144, "step": 715 }, { "epoch": 0.3065068493150685, "grad_norm": 11.779502868652344, "learning_rate": 6.9349315068493166e-06, "loss": 0.0198, "step": 716 }, { "epoch": 0.3069349315068493, "grad_norm": 10.708169937133789, "learning_rate": 6.930650684931507e-06, "loss": 0.0172, "step": 717 }, { "epoch": 0.3073630136986301, "grad_norm": 16.624950408935547, "learning_rate": 6.926369863013699e-06, "loss": 0.0227, "step": 718 }, { "epoch": 0.307791095890411, "grad_norm": 5.336156368255615, "learning_rate": 6.922089041095891e-06, "loss": 0.009, "step": 719 }, { "epoch": 0.3082191780821918, "grad_norm": 19.109262466430664, "learning_rate": 6.917808219178082e-06, "loss": 0.0473, "step": 720 }, { "epoch": 0.3086472602739726, "grad_norm": 4.689062118530273, "learning_rate": 6.913527397260275e-06, "loss": 0.0026, "step": 721 }, { "epoch": 0.3090753424657534, "grad_norm": 153.7244110107422, "learning_rate": 6.909246575342466e-06, "loss": 0.3762, "step": 722 }, { "epoch": 0.3095034246575342, "grad_norm": 45.71989822387695, "learning_rate": 6.904965753424658e-06, "loss": 0.1113, "step": 723 }, { "epoch": 0.3099315068493151, "grad_norm": 217.67835998535156, "learning_rate": 6.9006849315068505e-06, "loss": 0.749, "step": 724 }, { "epoch": 0.3103595890410959, "grad_norm": 180.55604553222656, "learning_rate": 6.896404109589042e-06, "loss": 0.207, "step": 725 }, { "epoch": 0.3107876712328767, "grad_norm": 60.27791976928711, "learning_rate": 6.892123287671233e-06, "loss": 0.0967, "step": 726 }, { "epoch": 0.3112157534246575, "grad_norm": 47.213829040527344, "learning_rate": 6.8878424657534245e-06, "loss": 0.1113, "step": 727 }, { "epoch": 0.3116438356164384, "grad_norm": 23.9641170501709, "learning_rate": 6.883561643835617e-06, "loss": 0.0247, "step": 728 }, { "epoch": 0.3120719178082192, "grad_norm": 12.333625793457031, "learning_rate": 6.879280821917809e-06, "loss": 0.0114, "step": 729 }, { "epoch": 0.3125, "grad_norm": 13.724752426147461, "learning_rate": 6.875e-06, "loss": 0.0146, "step": 730 }, { "epoch": 0.3129280821917808, "grad_norm": 12.22458553314209, "learning_rate": 6.870719178082193e-06, "loss": 0.016, "step": 731 }, { "epoch": 0.3133561643835616, "grad_norm": 8.08575439453125, "learning_rate": 6.866438356164384e-06, "loss": 0.0067, "step": 732 }, { "epoch": 0.3137842465753425, "grad_norm": 14.754509925842285, "learning_rate": 6.862157534246576e-06, "loss": 0.0165, "step": 733 }, { "epoch": 0.3142123287671233, "grad_norm": 25.3981990814209, "learning_rate": 6.8578767123287685e-06, "loss": 0.0376, "step": 734 }, { "epoch": 0.3146404109589041, "grad_norm": 130.7689208984375, "learning_rate": 6.853595890410959e-06, "loss": 0.2422, "step": 735 }, { "epoch": 0.3150684931506849, "grad_norm": 66.85039520263672, "learning_rate": 6.849315068493151e-06, "loss": 0.4971, "step": 736 }, { "epoch": 0.3154965753424658, "grad_norm": 15.581055641174316, "learning_rate": 6.8450342465753425e-06, "loss": 0.0169, "step": 737 }, { "epoch": 0.3159246575342466, "grad_norm": 10.500638008117676, "learning_rate": 6.840753424657535e-06, "loss": 0.0162, "step": 738 }, { "epoch": 0.3163527397260274, "grad_norm": 18.11604881286621, "learning_rate": 6.836472602739727e-06, "loss": 0.0215, "step": 739 }, { "epoch": 0.3167808219178082, "grad_norm": 16.76956558227539, "learning_rate": 6.832191780821918e-06, "loss": 0.0397, "step": 740 }, { "epoch": 0.317208904109589, "grad_norm": 43.52001953125, "learning_rate": 6.827910958904111e-06, "loss": 0.043, "step": 741 }, { "epoch": 0.3176369863013699, "grad_norm": 6.29957389831543, "learning_rate": 6.823630136986302e-06, "loss": 0.005, "step": 742 }, { "epoch": 0.3180650684931507, "grad_norm": 78.44217681884766, "learning_rate": 6.819349315068494e-06, "loss": 0.22, "step": 743 }, { "epoch": 0.3184931506849315, "grad_norm": 2.262739896774292, "learning_rate": 6.815068493150685e-06, "loss": 0.0059, "step": 744 }, { "epoch": 0.3189212328767123, "grad_norm": 7.1645402908325195, "learning_rate": 6.810787671232877e-06, "loss": 0.0094, "step": 745 }, { "epoch": 0.3193493150684932, "grad_norm": 127.20807647705078, "learning_rate": 6.806506849315069e-06, "loss": 0.0886, "step": 746 }, { "epoch": 0.319777397260274, "grad_norm": 21.663501739501953, "learning_rate": 6.8022260273972605e-06, "loss": 0.0203, "step": 747 }, { "epoch": 0.3202054794520548, "grad_norm": 6.34868860244751, "learning_rate": 6.797945205479453e-06, "loss": 0.0083, "step": 748 }, { "epoch": 0.3206335616438356, "grad_norm": 8.686149597167969, "learning_rate": 6.793664383561645e-06, "loss": 0.006, "step": 749 }, { "epoch": 0.3210616438356164, "grad_norm": 27.286664962768555, "learning_rate": 6.789383561643836e-06, "loss": 0.0313, "step": 750 }, { "epoch": 0.3214897260273973, "grad_norm": 58.57149887084961, "learning_rate": 6.785102739726028e-06, "loss": 0.3638, "step": 751 }, { "epoch": 0.3219178082191781, "grad_norm": 27.98064422607422, "learning_rate": 6.78082191780822e-06, "loss": 0.0228, "step": 752 }, { "epoch": 0.3223458904109589, "grad_norm": 504.890625, "learning_rate": 6.776541095890411e-06, "loss": 1.7298, "step": 753 }, { "epoch": 0.3227739726027397, "grad_norm": 1.934202790260315, "learning_rate": 6.772260273972603e-06, "loss": 0.0026, "step": 754 }, { "epoch": 0.3232020547945205, "grad_norm": 17.227943420410156, "learning_rate": 6.767979452054794e-06, "loss": 0.019, "step": 755 }, { "epoch": 0.3236301369863014, "grad_norm": 23.620874404907227, "learning_rate": 6.763698630136987e-06, "loss": 0.0358, "step": 756 }, { "epoch": 0.3240582191780822, "grad_norm": 8.004253387451172, "learning_rate": 6.7594178082191785e-06, "loss": 0.012, "step": 757 }, { "epoch": 0.324486301369863, "grad_norm": 87.92827606201172, "learning_rate": 6.75513698630137e-06, "loss": 0.3543, "step": 758 }, { "epoch": 0.3249143835616438, "grad_norm": 16.671125411987305, "learning_rate": 6.750856164383563e-06, "loss": 0.0179, "step": 759 }, { "epoch": 0.3253424657534247, "grad_norm": 26.35955238342285, "learning_rate": 6.746575342465754e-06, "loss": 0.0471, "step": 760 }, { "epoch": 0.3257705479452055, "grad_norm": 10.599063873291016, "learning_rate": 6.742294520547946e-06, "loss": 0.0085, "step": 761 }, { "epoch": 0.3261986301369863, "grad_norm": 454.0329284667969, "learning_rate": 6.738013698630137e-06, "loss": 0.7407, "step": 762 }, { "epoch": 0.3266267123287671, "grad_norm": 11.343154907226562, "learning_rate": 6.733732876712329e-06, "loss": 0.0122, "step": 763 }, { "epoch": 0.3270547945205479, "grad_norm": 16.3758487701416, "learning_rate": 6.729452054794521e-06, "loss": 0.0224, "step": 764 }, { "epoch": 0.3274828767123288, "grad_norm": 20.14232635498047, "learning_rate": 6.725171232876712e-06, "loss": 0.0292, "step": 765 }, { "epoch": 0.3279109589041096, "grad_norm": 208.4271697998047, "learning_rate": 6.720890410958905e-06, "loss": 0.7407, "step": 766 }, { "epoch": 0.3283390410958904, "grad_norm": 11.07890796661377, "learning_rate": 6.7166095890410965e-06, "loss": 0.0094, "step": 767 }, { "epoch": 0.3287671232876712, "grad_norm": 855.7062377929688, "learning_rate": 6.712328767123288e-06, "loss": 1.8857, "step": 768 }, { "epoch": 0.3291952054794521, "grad_norm": 44.36665725708008, "learning_rate": 6.708047945205481e-06, "loss": 0.0713, "step": 769 }, { "epoch": 0.3296232876712329, "grad_norm": 13.021831512451172, "learning_rate": 6.703767123287672e-06, "loss": 0.0154, "step": 770 }, { "epoch": 0.3300513698630137, "grad_norm": 80.6351318359375, "learning_rate": 6.699486301369864e-06, "loss": 0.2122, "step": 771 }, { "epoch": 0.3304794520547945, "grad_norm": 56.189090728759766, "learning_rate": 6.695205479452055e-06, "loss": 0.5319, "step": 772 }, { "epoch": 0.3309075342465753, "grad_norm": 5.4538164138793945, "learning_rate": 6.690924657534247e-06, "loss": 0.0087, "step": 773 }, { "epoch": 0.3313356164383562, "grad_norm": 14.344651222229004, "learning_rate": 6.686643835616439e-06, "loss": 0.0244, "step": 774 }, { "epoch": 0.331763698630137, "grad_norm": 61.52043151855469, "learning_rate": 6.68236301369863e-06, "loss": 0.1411, "step": 775 }, { "epoch": 0.3321917808219178, "grad_norm": 86.20520782470703, "learning_rate": 6.678082191780823e-06, "loss": 0.0879, "step": 776 }, { "epoch": 0.3326198630136986, "grad_norm": 57.60829162597656, "learning_rate": 6.6738013698630145e-06, "loss": 0.1816, "step": 777 }, { "epoch": 0.3330479452054795, "grad_norm": 57.83295440673828, "learning_rate": 6.669520547945206e-06, "loss": 0.2549, "step": 778 }, { "epoch": 0.3334760273972603, "grad_norm": 37.63810729980469, "learning_rate": 6.665239726027399e-06, "loss": 0.0751, "step": 779 }, { "epoch": 0.3339041095890411, "grad_norm": 15.989691734313965, "learning_rate": 6.66095890410959e-06, "loss": 0.021, "step": 780 }, { "epoch": 0.3343321917808219, "grad_norm": 21.690929412841797, "learning_rate": 6.656678082191781e-06, "loss": 0.0386, "step": 781 }, { "epoch": 0.3347602739726027, "grad_norm": 5.409888744354248, "learning_rate": 6.652397260273973e-06, "loss": 0.0109, "step": 782 }, { "epoch": 0.3351883561643836, "grad_norm": 28.66766357421875, "learning_rate": 6.648116438356165e-06, "loss": 0.0651, "step": 783 }, { "epoch": 0.3356164383561644, "grad_norm": 25.652944564819336, "learning_rate": 6.643835616438357e-06, "loss": 0.0251, "step": 784 }, { "epoch": 0.3360445205479452, "grad_norm": 27.543224334716797, "learning_rate": 6.639554794520548e-06, "loss": 0.0262, "step": 785 }, { "epoch": 0.336472602739726, "grad_norm": 31.819868087768555, "learning_rate": 6.63527397260274e-06, "loss": 0.0665, "step": 786 }, { "epoch": 0.3369006849315068, "grad_norm": 19.81488037109375, "learning_rate": 6.6309931506849325e-06, "loss": 0.0289, "step": 787 }, { "epoch": 0.3373287671232877, "grad_norm": 24.618408203125, "learning_rate": 6.626712328767124e-06, "loss": 0.0332, "step": 788 }, { "epoch": 0.3377568493150685, "grad_norm": 121.25804138183594, "learning_rate": 6.622431506849316e-06, "loss": 0.3293, "step": 789 }, { "epoch": 0.3381849315068493, "grad_norm": 6.387507915496826, "learning_rate": 6.6181506849315065e-06, "loss": 0.0067, "step": 790 }, { "epoch": 0.3386130136986301, "grad_norm": 87.32767486572266, "learning_rate": 6.613869863013699e-06, "loss": 0.3959, "step": 791 }, { "epoch": 0.339041095890411, "grad_norm": 3.0123238563537598, "learning_rate": 6.609589041095891e-06, "loss": 0.0045, "step": 792 }, { "epoch": 0.3394691780821918, "grad_norm": 4.5845947265625, "learning_rate": 6.605308219178082e-06, "loss": 0.0119, "step": 793 }, { "epoch": 0.3398972602739726, "grad_norm": 14.00794506072998, "learning_rate": 6.601027397260275e-06, "loss": 0.0153, "step": 794 }, { "epoch": 0.3403253424657534, "grad_norm": 186.1710968017578, "learning_rate": 6.596746575342466e-06, "loss": 0.4659, "step": 795 }, { "epoch": 0.3407534246575342, "grad_norm": 8.461259841918945, "learning_rate": 6.592465753424658e-06, "loss": 0.0069, "step": 796 }, { "epoch": 0.3411815068493151, "grad_norm": 72.13870239257812, "learning_rate": 6.5881849315068505e-06, "loss": 0.162, "step": 797 }, { "epoch": 0.3416095890410959, "grad_norm": 10.147340774536133, "learning_rate": 6.583904109589042e-06, "loss": 0.0142, "step": 798 }, { "epoch": 0.3420376712328767, "grad_norm": 179.70053100585938, "learning_rate": 6.579623287671233e-06, "loss": 1.0081, "step": 799 }, { "epoch": 0.3424657534246575, "grad_norm": 17.37003517150879, "learning_rate": 6.5753424657534245e-06, "loss": 0.0505, "step": 800 }, { "epoch": 0.3428938356164384, "grad_norm": 33.14570236206055, "learning_rate": 6.571061643835617e-06, "loss": 0.0557, "step": 801 }, { "epoch": 0.3433219178082192, "grad_norm": 15.514074325561523, "learning_rate": 6.566780821917809e-06, "loss": 0.0209, "step": 802 }, { "epoch": 0.34375, "grad_norm": 75.05601501464844, "learning_rate": 6.5625e-06, "loss": 0.7754, "step": 803 }, { "epoch": 0.3441780821917808, "grad_norm": 25.535127639770508, "learning_rate": 6.558219178082193e-06, "loss": 0.0406, "step": 804 }, { "epoch": 0.3446061643835616, "grad_norm": 441.56689453125, "learning_rate": 6.553938356164384e-06, "loss": 0.3735, "step": 805 }, { "epoch": 0.3450342465753425, "grad_norm": 22.877582550048828, "learning_rate": 6.549657534246576e-06, "loss": 0.0176, "step": 806 }, { "epoch": 0.3454623287671233, "grad_norm": 2.8419065475463867, "learning_rate": 6.5453767123287685e-06, "loss": 0.0037, "step": 807 }, { "epoch": 0.3458904109589041, "grad_norm": 76.78167724609375, "learning_rate": 6.541095890410959e-06, "loss": 0.2032, "step": 808 }, { "epoch": 0.3463184931506849, "grad_norm": 7.273828506469727, "learning_rate": 6.536815068493151e-06, "loss": 0.0151, "step": 809 }, { "epoch": 0.3467465753424658, "grad_norm": 37.0320930480957, "learning_rate": 6.5325342465753425e-06, "loss": 0.0396, "step": 810 }, { "epoch": 0.3471746575342466, "grad_norm": 8.880044937133789, "learning_rate": 6.528253424657535e-06, "loss": 0.0109, "step": 811 }, { "epoch": 0.3476027397260274, "grad_norm": 26.836044311523438, "learning_rate": 6.523972602739727e-06, "loss": 0.0497, "step": 812 }, { "epoch": 0.3480308219178082, "grad_norm": 15.911478996276855, "learning_rate": 6.519691780821918e-06, "loss": 0.0182, "step": 813 }, { "epoch": 0.348458904109589, "grad_norm": 7.474634170532227, "learning_rate": 6.515410958904111e-06, "loss": 0.0198, "step": 814 }, { "epoch": 0.3488869863013699, "grad_norm": 5.0682783126831055, "learning_rate": 6.511130136986302e-06, "loss": 0.0264, "step": 815 }, { "epoch": 0.3493150684931507, "grad_norm": 0.7878289818763733, "learning_rate": 6.506849315068494e-06, "loss": 0.0019, "step": 816 }, { "epoch": 0.3497431506849315, "grad_norm": 8.055185317993164, "learning_rate": 6.502568493150685e-06, "loss": 0.0083, "step": 817 }, { "epoch": 0.3501712328767123, "grad_norm": 11.566200256347656, "learning_rate": 6.498287671232877e-06, "loss": 0.0156, "step": 818 }, { "epoch": 0.3505993150684932, "grad_norm": 3.710402488708496, "learning_rate": 6.494006849315069e-06, "loss": 0.0027, "step": 819 }, { "epoch": 0.351027397260274, "grad_norm": 7.8317084312438965, "learning_rate": 6.4897260273972605e-06, "loss": 0.0051, "step": 820 }, { "epoch": 0.3514554794520548, "grad_norm": 6.033999919891357, "learning_rate": 6.485445205479452e-06, "loss": 0.011, "step": 821 }, { "epoch": 0.3518835616438356, "grad_norm": 1.830694317817688, "learning_rate": 6.481164383561645e-06, "loss": 0.0025, "step": 822 }, { "epoch": 0.3523116438356164, "grad_norm": 5.064305305480957, "learning_rate": 6.476883561643836e-06, "loss": 0.007, "step": 823 }, { "epoch": 0.3527397260273973, "grad_norm": 6.223678112030029, "learning_rate": 6.472602739726028e-06, "loss": 0.0075, "step": 824 }, { "epoch": 0.3531678082191781, "grad_norm": 72.25824737548828, "learning_rate": 6.46832191780822e-06, "loss": 0.3499, "step": 825 }, { "epoch": 0.3535958904109589, "grad_norm": 26.88986587524414, "learning_rate": 6.464041095890411e-06, "loss": 0.0502, "step": 826 }, { "epoch": 0.3540239726027397, "grad_norm": 22.039337158203125, "learning_rate": 6.459760273972603e-06, "loss": 0.0348, "step": 827 }, { "epoch": 0.3544520547945205, "grad_norm": 66.24119567871094, "learning_rate": 6.455479452054794e-06, "loss": 0.1115, "step": 828 }, { "epoch": 0.3548801369863014, "grad_norm": 16.62362289428711, "learning_rate": 6.451198630136987e-06, "loss": 0.0196, "step": 829 }, { "epoch": 0.3553082191780822, "grad_norm": 149.7106170654297, "learning_rate": 6.4469178082191785e-06, "loss": 0.7102, "step": 830 }, { "epoch": 0.355736301369863, "grad_norm": 285.2542724609375, "learning_rate": 6.44263698630137e-06, "loss": 2.1409, "step": 831 }, { "epoch": 0.3561643835616438, "grad_norm": 5.94433069229126, "learning_rate": 6.438356164383563e-06, "loss": 0.0037, "step": 832 }, { "epoch": 0.3565924657534247, "grad_norm": 130.2583465576172, "learning_rate": 6.434075342465754e-06, "loss": 0.7834, "step": 833 }, { "epoch": 0.3570205479452055, "grad_norm": 111.3176498413086, "learning_rate": 6.429794520547946e-06, "loss": 0.3678, "step": 834 }, { "epoch": 0.3574486301369863, "grad_norm": 9.774210929870605, "learning_rate": 6.425513698630137e-06, "loss": 0.0068, "step": 835 }, { "epoch": 0.3578767123287671, "grad_norm": 5.70313024520874, "learning_rate": 6.421232876712329e-06, "loss": 0.0065, "step": 836 }, { "epoch": 0.3583047945205479, "grad_norm": 1.3021713495254517, "learning_rate": 6.416952054794521e-06, "loss": 0.0044, "step": 837 }, { "epoch": 0.3587328767123288, "grad_norm": 66.08086395263672, "learning_rate": 6.412671232876712e-06, "loss": 0.2146, "step": 838 }, { "epoch": 0.3591609589041096, "grad_norm": 4.013594627380371, "learning_rate": 6.408390410958905e-06, "loss": 0.0026, "step": 839 }, { "epoch": 0.3595890410958904, "grad_norm": 3.3040263652801514, "learning_rate": 6.4041095890410965e-06, "loss": 0.0031, "step": 840 }, { "epoch": 0.3600171232876712, "grad_norm": 103.75096130371094, "learning_rate": 6.399828767123288e-06, "loss": 0.146, "step": 841 }, { "epoch": 0.3604452054794521, "grad_norm": 72.23934173583984, "learning_rate": 6.395547945205481e-06, "loss": 0.0536, "step": 842 }, { "epoch": 0.3608732876712329, "grad_norm": 152.94384765625, "learning_rate": 6.391267123287672e-06, "loss": 0.2349, "step": 843 }, { "epoch": 0.3613013698630137, "grad_norm": 20.31263542175293, "learning_rate": 6.386986301369864e-06, "loss": 0.0241, "step": 844 }, { "epoch": 0.3617294520547945, "grad_norm": 21.1018009185791, "learning_rate": 6.382705479452055e-06, "loss": 0.0129, "step": 845 }, { "epoch": 0.3621575342465753, "grad_norm": 16.643999099731445, "learning_rate": 6.378424657534247e-06, "loss": 0.0189, "step": 846 }, { "epoch": 0.3625856164383562, "grad_norm": 5.676992893218994, "learning_rate": 6.374143835616439e-06, "loss": 0.0025, "step": 847 }, { "epoch": 0.363013698630137, "grad_norm": 1.1164036989212036, "learning_rate": 6.36986301369863e-06, "loss": 0.0009, "step": 848 }, { "epoch": 0.3634417808219178, "grad_norm": 7.258193492889404, "learning_rate": 6.365582191780823e-06, "loss": 0.0037, "step": 849 }, { "epoch": 0.3638698630136986, "grad_norm": 9.21819019317627, "learning_rate": 6.3613013698630145e-06, "loss": 0.0231, "step": 850 }, { "epoch": 0.3642979452054795, "grad_norm": 20.49305534362793, "learning_rate": 6.357020547945206e-06, "loss": 0.027, "step": 851 }, { "epoch": 0.3647260273972603, "grad_norm": 1.7870659828186035, "learning_rate": 6.352739726027398e-06, "loss": 0.0016, "step": 852 }, { "epoch": 0.3651541095890411, "grad_norm": 7.725332736968994, "learning_rate": 6.34845890410959e-06, "loss": 0.0052, "step": 853 }, { "epoch": 0.3655821917808219, "grad_norm": 62.8889045715332, "learning_rate": 6.344178082191781e-06, "loss": 0.0691, "step": 854 }, { "epoch": 0.3660102739726027, "grad_norm": 100.45838165283203, "learning_rate": 6.339897260273973e-06, "loss": 0.1436, "step": 855 }, { "epoch": 0.3664383561643836, "grad_norm": 12.106488227844238, "learning_rate": 6.335616438356164e-06, "loss": 0.0142, "step": 856 }, { "epoch": 0.3668664383561644, "grad_norm": 19.49967384338379, "learning_rate": 6.331335616438357e-06, "loss": 0.0244, "step": 857 }, { "epoch": 0.3672945205479452, "grad_norm": 60.52241134643555, "learning_rate": 6.327054794520548e-06, "loss": 0.5321, "step": 858 }, { "epoch": 0.367722602739726, "grad_norm": 93.20745086669922, "learning_rate": 6.32277397260274e-06, "loss": 0.7114, "step": 859 }, { "epoch": 0.3681506849315068, "grad_norm": 151.56056213378906, "learning_rate": 6.3184931506849325e-06, "loss": 0.1941, "step": 860 }, { "epoch": 0.3685787671232877, "grad_norm": 12.088443756103516, "learning_rate": 6.314212328767124e-06, "loss": 0.0114, "step": 861 }, { "epoch": 0.3690068493150685, "grad_norm": 9.825630187988281, "learning_rate": 6.309931506849316e-06, "loss": 0.0071, "step": 862 }, { "epoch": 0.3694349315068493, "grad_norm": 15.508722305297852, "learning_rate": 6.3056506849315066e-06, "loss": 0.0136, "step": 863 }, { "epoch": 0.3698630136986301, "grad_norm": 122.10585021972656, "learning_rate": 6.301369863013699e-06, "loss": 0.3805, "step": 864 }, { "epoch": 0.370291095890411, "grad_norm": 192.96530151367188, "learning_rate": 6.297089041095891e-06, "loss": 0.4185, "step": 865 }, { "epoch": 0.3707191780821918, "grad_norm": 21.94679832458496, "learning_rate": 6.292808219178082e-06, "loss": 0.0336, "step": 866 }, { "epoch": 0.3711472602739726, "grad_norm": 11.41549301147461, "learning_rate": 6.288527397260275e-06, "loss": 0.0092, "step": 867 }, { "epoch": 0.3715753424657534, "grad_norm": 8.148255348205566, "learning_rate": 6.284246575342466e-06, "loss": 0.0191, "step": 868 }, { "epoch": 0.3720034246575342, "grad_norm": 10.335281372070312, "learning_rate": 6.279965753424658e-06, "loss": 0.0091, "step": 869 }, { "epoch": 0.3724315068493151, "grad_norm": 181.57301330566406, "learning_rate": 6.2756849315068505e-06, "loss": 0.7219, "step": 870 }, { "epoch": 0.3728595890410959, "grad_norm": 2.4079675674438477, "learning_rate": 6.271404109589042e-06, "loss": 0.0042, "step": 871 }, { "epoch": 0.3732876712328767, "grad_norm": 7.557215213775635, "learning_rate": 6.267123287671233e-06, "loss": 0.0052, "step": 872 }, { "epoch": 0.3737157534246575, "grad_norm": 11.720671653747559, "learning_rate": 6.2628424657534246e-06, "loss": 0.012, "step": 873 }, { "epoch": 0.3741438356164384, "grad_norm": 79.31566619873047, "learning_rate": 6.258561643835617e-06, "loss": 0.4631, "step": 874 }, { "epoch": 0.3745719178082192, "grad_norm": 14.094743728637695, "learning_rate": 6.254280821917809e-06, "loss": 0.0148, "step": 875 }, { "epoch": 0.375, "grad_norm": 12.296183586120605, "learning_rate": 6.25e-06, "loss": 0.0194, "step": 876 }, { "epoch": 0.3754280821917808, "grad_norm": 13.464929580688477, "learning_rate": 6.245719178082193e-06, "loss": 0.0145, "step": 877 }, { "epoch": 0.3758561643835616, "grad_norm": 62.615478515625, "learning_rate": 6.241438356164384e-06, "loss": 0.0769, "step": 878 }, { "epoch": 0.3762842465753425, "grad_norm": 4.890023231506348, "learning_rate": 6.237157534246576e-06, "loss": 0.0044, "step": 879 }, { "epoch": 0.3767123287671233, "grad_norm": 386.76483154296875, "learning_rate": 6.2328767123287685e-06, "loss": 1.1992, "step": 880 }, { "epoch": 0.3771404109589041, "grad_norm": 196.6068572998047, "learning_rate": 6.228595890410959e-06, "loss": 0.3713, "step": 881 }, { "epoch": 0.3775684931506849, "grad_norm": 22.358598709106445, "learning_rate": 6.224315068493151e-06, "loss": 0.0269, "step": 882 }, { "epoch": 0.3779965753424658, "grad_norm": 32.61568069458008, "learning_rate": 6.2200342465753426e-06, "loss": 0.1155, "step": 883 }, { "epoch": 0.3784246575342466, "grad_norm": 58.500362396240234, "learning_rate": 6.215753424657535e-06, "loss": 0.0719, "step": 884 }, { "epoch": 0.3788527397260274, "grad_norm": 9.146284103393555, "learning_rate": 6.211472602739727e-06, "loss": 0.0074, "step": 885 }, { "epoch": 0.3792808219178082, "grad_norm": 36.86732482910156, "learning_rate": 6.207191780821918e-06, "loss": 0.0372, "step": 886 }, { "epoch": 0.379708904109589, "grad_norm": 2.518616199493408, "learning_rate": 6.20291095890411e-06, "loss": 0.0022, "step": 887 }, { "epoch": 0.3801369863013699, "grad_norm": 91.62859344482422, "learning_rate": 6.198630136986302e-06, "loss": 0.0561, "step": 888 }, { "epoch": 0.3805650684931507, "grad_norm": 137.18487548828125, "learning_rate": 6.194349315068494e-06, "loss": 0.5324, "step": 889 }, { "epoch": 0.3809931506849315, "grad_norm": 100.78773498535156, "learning_rate": 6.190068493150685e-06, "loss": 0.214, "step": 890 }, { "epoch": 0.3814212328767123, "grad_norm": 103.11360931396484, "learning_rate": 6.1857876712328764e-06, "loss": 0.0992, "step": 891 }, { "epoch": 0.3818493150684932, "grad_norm": 5.435788154602051, "learning_rate": 6.181506849315069e-06, "loss": 0.0078, "step": 892 }, { "epoch": 0.382277397260274, "grad_norm": 547.0624389648438, "learning_rate": 6.1772260273972606e-06, "loss": 1.9218, "step": 893 }, { "epoch": 0.3827054794520548, "grad_norm": 6.447920799255371, "learning_rate": 6.172945205479452e-06, "loss": 0.0045, "step": 894 }, { "epoch": 0.3831335616438356, "grad_norm": 2.0357208251953125, "learning_rate": 6.168664383561645e-06, "loss": 0.0024, "step": 895 }, { "epoch": 0.3835616438356164, "grad_norm": 58.316654205322266, "learning_rate": 6.164383561643836e-06, "loss": 0.3872, "step": 896 }, { "epoch": 0.3839897260273973, "grad_norm": 123.9785385131836, "learning_rate": 6.160102739726028e-06, "loss": 0.7482, "step": 897 }, { "epoch": 0.3844178082191781, "grad_norm": 41.15146255493164, "learning_rate": 6.15582191780822e-06, "loss": 0.2041, "step": 898 }, { "epoch": 0.3848458904109589, "grad_norm": 132.924560546875, "learning_rate": 6.151541095890411e-06, "loss": 0.0637, "step": 899 }, { "epoch": 0.3852739726027397, "grad_norm": 57.75654220581055, "learning_rate": 6.147260273972603e-06, "loss": 0.0895, "step": 900 }, { "epoch": 0.3857020547945205, "grad_norm": 182.3575897216797, "learning_rate": 6.1429794520547944e-06, "loss": 0.2964, "step": 901 }, { "epoch": 0.3861301369863014, "grad_norm": 20.49941062927246, "learning_rate": 6.138698630136987e-06, "loss": 0.0287, "step": 902 }, { "epoch": 0.3865582191780822, "grad_norm": 19.082088470458984, "learning_rate": 6.1344178082191786e-06, "loss": 0.0242, "step": 903 }, { "epoch": 0.386986301369863, "grad_norm": 65.98912811279297, "learning_rate": 6.13013698630137e-06, "loss": 0.0812, "step": 904 }, { "epoch": 0.3874143835616438, "grad_norm": 23.577014923095703, "learning_rate": 6.125856164383563e-06, "loss": 0.0375, "step": 905 }, { "epoch": 0.3878424657534247, "grad_norm": 41.05651092529297, "learning_rate": 6.121575342465754e-06, "loss": 0.0924, "step": 906 }, { "epoch": 0.3882705479452055, "grad_norm": 2.0852835178375244, "learning_rate": 6.117294520547946e-06, "loss": 0.002, "step": 907 }, { "epoch": 0.3886986301369863, "grad_norm": 11.117595672607422, "learning_rate": 6.113013698630137e-06, "loss": 0.0131, "step": 908 }, { "epoch": 0.3891267123287671, "grad_norm": 19.924936294555664, "learning_rate": 6.108732876712329e-06, "loss": 0.0286, "step": 909 }, { "epoch": 0.3895547945205479, "grad_norm": 7.271633625030518, "learning_rate": 6.104452054794521e-06, "loss": 0.0195, "step": 910 }, { "epoch": 0.3899828767123288, "grad_norm": 5.640427589416504, "learning_rate": 6.1001712328767124e-06, "loss": 0.0029, "step": 911 }, { "epoch": 0.3904109589041096, "grad_norm": 12.804917335510254, "learning_rate": 6.095890410958905e-06, "loss": 0.0114, "step": 912 }, { "epoch": 0.3908390410958904, "grad_norm": 11.020535469055176, "learning_rate": 6.0916095890410966e-06, "loss": 0.0174, "step": 913 }, { "epoch": 0.3912671232876712, "grad_norm": 3.9280107021331787, "learning_rate": 6.087328767123288e-06, "loss": 0.0136, "step": 914 }, { "epoch": 0.3916952054794521, "grad_norm": 16.8043270111084, "learning_rate": 6.083047945205481e-06, "loss": 0.0195, "step": 915 }, { "epoch": 0.3921232876712329, "grad_norm": 21.98556900024414, "learning_rate": 6.078767123287672e-06, "loss": 0.0256, "step": 916 }, { "epoch": 0.3925513698630137, "grad_norm": 58.59423065185547, "learning_rate": 6.074486301369864e-06, "loss": 0.3357, "step": 917 }, { "epoch": 0.3929794520547945, "grad_norm": 14.943151473999023, "learning_rate": 6.070205479452055e-06, "loss": 0.0464, "step": 918 }, { "epoch": 0.3934075342465753, "grad_norm": 101.20857238769531, "learning_rate": 6.065924657534246e-06, "loss": 0.1646, "step": 919 }, { "epoch": 0.3938356164383562, "grad_norm": 104.01217651367188, "learning_rate": 6.061643835616439e-06, "loss": 0.5366, "step": 920 }, { "epoch": 0.394263698630137, "grad_norm": 14.744192123413086, "learning_rate": 6.0573630136986304e-06, "loss": 0.0153, "step": 921 }, { "epoch": 0.3946917808219178, "grad_norm": 14.16304874420166, "learning_rate": 6.053082191780822e-06, "loss": 0.015, "step": 922 }, { "epoch": 0.3951198630136986, "grad_norm": 6.476023197174072, "learning_rate": 6.0488013698630145e-06, "loss": 0.0064, "step": 923 }, { "epoch": 0.3955479452054795, "grad_norm": 13.571688652038574, "learning_rate": 6.044520547945206e-06, "loss": 0.0137, "step": 924 }, { "epoch": 0.3959760273972603, "grad_norm": 83.42941284179688, "learning_rate": 6.040239726027398e-06, "loss": 0.0881, "step": 925 }, { "epoch": 0.3964041095890411, "grad_norm": 13.660935401916504, "learning_rate": 6.03595890410959e-06, "loss": 0.0148, "step": 926 }, { "epoch": 0.3968321917808219, "grad_norm": 12.997106552124023, "learning_rate": 6.031678082191781e-06, "loss": 0.0183, "step": 927 }, { "epoch": 0.3972602739726027, "grad_norm": 67.87445068359375, "learning_rate": 6.027397260273973e-06, "loss": 0.3519, "step": 928 }, { "epoch": 0.3976883561643836, "grad_norm": 33.10548400878906, "learning_rate": 6.023116438356164e-06, "loss": 0.1334, "step": 929 }, { "epoch": 0.3981164383561644, "grad_norm": 20.274568557739258, "learning_rate": 6.018835616438357e-06, "loss": 0.0283, "step": 930 }, { "epoch": 0.3985445205479452, "grad_norm": 22.506881713867188, "learning_rate": 6.0145547945205484e-06, "loss": 0.0325, "step": 931 }, { "epoch": 0.398972602739726, "grad_norm": 46.482948303222656, "learning_rate": 6.01027397260274e-06, "loss": 0.0624, "step": 932 }, { "epoch": 0.3994006849315068, "grad_norm": 35.14398956298828, "learning_rate": 6.0059931506849325e-06, "loss": 0.0812, "step": 933 }, { "epoch": 0.3998287671232877, "grad_norm": 15.867663383483887, "learning_rate": 6.001712328767124e-06, "loss": 0.0183, "step": 934 }, { "epoch": 0.4002568493150685, "grad_norm": 25.753124237060547, "learning_rate": 5.997431506849316e-06, "loss": 0.0436, "step": 935 }, { "epoch": 0.4006849315068493, "grad_norm": 15.811467170715332, "learning_rate": 5.993150684931507e-06, "loss": 0.0263, "step": 936 }, { "epoch": 0.4011130136986301, "grad_norm": 55.820892333984375, "learning_rate": 5.988869863013699e-06, "loss": 0.5389, "step": 937 }, { "epoch": 0.401541095890411, "grad_norm": 6.706796169281006, "learning_rate": 5.984589041095891e-06, "loss": 0.0093, "step": 938 }, { "epoch": 0.4019691780821918, "grad_norm": 23.71198272705078, "learning_rate": 5.980308219178082e-06, "loss": 0.0489, "step": 939 }, { "epoch": 0.4023972602739726, "grad_norm": 87.54012298583984, "learning_rate": 5.976027397260275e-06, "loss": 0.5646, "step": 940 }, { "epoch": 0.4028253424657534, "grad_norm": 26.53270149230957, "learning_rate": 5.9717465753424664e-06, "loss": 0.048, "step": 941 }, { "epoch": 0.4032534246575342, "grad_norm": 9.968974113464355, "learning_rate": 5.967465753424658e-06, "loss": 0.0232, "step": 942 }, { "epoch": 0.4036815068493151, "grad_norm": 5.335632801055908, "learning_rate": 5.9631849315068505e-06, "loss": 0.0124, "step": 943 }, { "epoch": 0.4041095890410959, "grad_norm": 6.9691901206970215, "learning_rate": 5.958904109589042e-06, "loss": 0.0075, "step": 944 }, { "epoch": 0.4045376712328767, "grad_norm": 391.5332336425781, "learning_rate": 5.954623287671233e-06, "loss": 1.0069, "step": 945 }, { "epoch": 0.4049657534246575, "grad_norm": 1.6628025770187378, "learning_rate": 5.950342465753425e-06, "loss": 0.003, "step": 946 }, { "epoch": 0.4053938356164384, "grad_norm": 41.638893127441406, "learning_rate": 5.946061643835617e-06, "loss": 0.5002, "step": 947 }, { "epoch": 0.4058219178082192, "grad_norm": 60.69717025756836, "learning_rate": 5.941780821917809e-06, "loss": 0.4755, "step": 948 }, { "epoch": 0.40625, "grad_norm": 16.42765998840332, "learning_rate": 5.9375e-06, "loss": 0.0198, "step": 949 }, { "epoch": 0.4066780821917808, "grad_norm": 66.93634796142578, "learning_rate": 5.933219178082193e-06, "loss": 0.2036, "step": 950 }, { "epoch": 0.4071061643835616, "grad_norm": 14.892927169799805, "learning_rate": 5.9289383561643844e-06, "loss": 0.0201, "step": 951 }, { "epoch": 0.4075342465753425, "grad_norm": 9.147442817687988, "learning_rate": 5.924657534246576e-06, "loss": 0.0232, "step": 952 }, { "epoch": 0.4079623287671233, "grad_norm": 8.011756896972656, "learning_rate": 5.920376712328768e-06, "loss": 0.0082, "step": 953 }, { "epoch": 0.4083904109589041, "grad_norm": 6.318150043487549, "learning_rate": 5.9160958904109585e-06, "loss": 0.0076, "step": 954 }, { "epoch": 0.4088184931506849, "grad_norm": 7.383048057556152, "learning_rate": 5.911815068493151e-06, "loss": 0.0196, "step": 955 }, { "epoch": 0.4092465753424658, "grad_norm": 4.141871452331543, "learning_rate": 5.907534246575343e-06, "loss": 0.0039, "step": 956 }, { "epoch": 0.4096746575342466, "grad_norm": 1.911180853843689, "learning_rate": 5.903253424657534e-06, "loss": 0.0023, "step": 957 }, { "epoch": 0.4101027397260274, "grad_norm": 2.452007293701172, "learning_rate": 5.898972602739727e-06, "loss": 0.002, "step": 958 }, { "epoch": 0.4105308219178082, "grad_norm": 6.452823162078857, "learning_rate": 5.894691780821918e-06, "loss": 0.0074, "step": 959 }, { "epoch": 0.410958904109589, "grad_norm": 1.1057308912277222, "learning_rate": 5.89041095890411e-06, "loss": 0.0033, "step": 960 }, { "epoch": 0.4113869863013699, "grad_norm": 121.41741943359375, "learning_rate": 5.8861301369863024e-06, "loss": 0.1245, "step": 961 }, { "epoch": 0.4118150684931507, "grad_norm": 26.434307098388672, "learning_rate": 5.881849315068494e-06, "loss": 0.0493, "step": 962 }, { "epoch": 0.4122431506849315, "grad_norm": 481.4786682128906, "learning_rate": 5.877568493150685e-06, "loss": 1.6669, "step": 963 }, { "epoch": 0.4126712328767123, "grad_norm": 33.2419319152832, "learning_rate": 5.8732876712328765e-06, "loss": 0.0574, "step": 964 }, { "epoch": 0.4130993150684932, "grad_norm": 225.33880615234375, "learning_rate": 5.869006849315069e-06, "loss": 0.6272, "step": 965 }, { "epoch": 0.413527397260274, "grad_norm": 107.65780639648438, "learning_rate": 5.864726027397261e-06, "loss": 0.6911, "step": 966 }, { "epoch": 0.4139554794520548, "grad_norm": 4.322863578796387, "learning_rate": 5.860445205479452e-06, "loss": 0.0051, "step": 967 }, { "epoch": 0.4143835616438356, "grad_norm": 5.4713215827941895, "learning_rate": 5.856164383561645e-06, "loss": 0.0092, "step": 968 }, { "epoch": 0.4148116438356164, "grad_norm": 3.8091259002685547, "learning_rate": 5.851883561643836e-06, "loss": 0.0017, "step": 969 }, { "epoch": 0.4152397260273973, "grad_norm": 7.451267242431641, "learning_rate": 5.847602739726028e-06, "loss": 0.0138, "step": 970 }, { "epoch": 0.4156678082191781, "grad_norm": 32.93617630004883, "learning_rate": 5.8433219178082204e-06, "loss": 0.0403, "step": 971 }, { "epoch": 0.4160958904109589, "grad_norm": 49.510276794433594, "learning_rate": 5.839041095890411e-06, "loss": 0.4891, "step": 972 }, { "epoch": 0.4165239726027397, "grad_norm": 40.13460922241211, "learning_rate": 5.834760273972603e-06, "loss": 0.0444, "step": 973 }, { "epoch": 0.4169520547945205, "grad_norm": 8.622764587402344, "learning_rate": 5.8304794520547945e-06, "loss": 0.0066, "step": 974 }, { "epoch": 0.4173801369863014, "grad_norm": 5.671130657196045, "learning_rate": 5.826198630136987e-06, "loss": 0.0051, "step": 975 }, { "epoch": 0.4178082191780822, "grad_norm": 2.098041296005249, "learning_rate": 5.821917808219179e-06, "loss": 0.0039, "step": 976 }, { "epoch": 0.418236301369863, "grad_norm": 13.897394180297852, "learning_rate": 5.81763698630137e-06, "loss": 0.0138, "step": 977 }, { "epoch": 0.4186643835616438, "grad_norm": 25.134187698364258, "learning_rate": 5.813356164383563e-06, "loss": 0.0077, "step": 978 }, { "epoch": 0.4190924657534247, "grad_norm": 24.463863372802734, "learning_rate": 5.809075342465754e-06, "loss": 0.0191, "step": 979 }, { "epoch": 0.4195205479452055, "grad_norm": 113.18438720703125, "learning_rate": 5.804794520547946e-06, "loss": 0.2802, "step": 980 }, { "epoch": 0.4199486301369863, "grad_norm": 7.934897422790527, "learning_rate": 5.800513698630137e-06, "loss": 0.0069, "step": 981 }, { "epoch": 0.4203767123287671, "grad_norm": 10.816283226013184, "learning_rate": 5.796232876712329e-06, "loss": 0.0187, "step": 982 }, { "epoch": 0.4208047945205479, "grad_norm": 347.7380065917969, "learning_rate": 5.791952054794521e-06, "loss": 0.5634, "step": 983 }, { "epoch": 0.4212328767123288, "grad_norm": 2.2849245071411133, "learning_rate": 5.7876712328767125e-06, "loss": 0.0029, "step": 984 }, { "epoch": 0.4216609589041096, "grad_norm": 130.4491424560547, "learning_rate": 5.783390410958904e-06, "loss": 0.4645, "step": 985 }, { "epoch": 0.4220890410958904, "grad_norm": 25.555770874023438, "learning_rate": 5.779109589041097e-06, "loss": 0.0261, "step": 986 }, { "epoch": 0.4225171232876712, "grad_norm": 27.832059860229492, "learning_rate": 5.774828767123288e-06, "loss": 0.0302, "step": 987 }, { "epoch": 0.4229452054794521, "grad_norm": 104.06024932861328, "learning_rate": 5.77054794520548e-06, "loss": 0.3544, "step": 988 }, { "epoch": 0.4233732876712329, "grad_norm": 32.38069152832031, "learning_rate": 5.766267123287672e-06, "loss": 0.0691, "step": 989 }, { "epoch": 0.4238013698630137, "grad_norm": 24.300683975219727, "learning_rate": 5.761986301369864e-06, "loss": 0.0372, "step": 990 }, { "epoch": 0.4242294520547945, "grad_norm": 63.390647888183594, "learning_rate": 5.757705479452055e-06, "loss": 0.1353, "step": 991 }, { "epoch": 0.4246575342465753, "grad_norm": 10.295267105102539, "learning_rate": 5.753424657534246e-06, "loss": 0.02, "step": 992 }, { "epoch": 0.4250856164383562, "grad_norm": 3.5813405513763428, "learning_rate": 5.749143835616439e-06, "loss": 0.014, "step": 993 }, { "epoch": 0.425513698630137, "grad_norm": 33.01933288574219, "learning_rate": 5.7448630136986305e-06, "loss": 0.0459, "step": 994 }, { "epoch": 0.4259417808219178, "grad_norm": 5.795770645141602, "learning_rate": 5.740582191780822e-06, "loss": 0.0058, "step": 995 }, { "epoch": 0.4263698630136986, "grad_norm": 16.074430465698242, "learning_rate": 5.736301369863015e-06, "loss": 0.0181, "step": 996 }, { "epoch": 0.4267979452054795, "grad_norm": 50.64266586303711, "learning_rate": 5.732020547945206e-06, "loss": 0.0782, "step": 997 }, { "epoch": 0.4272260273972603, "grad_norm": 3.7091071605682373, "learning_rate": 5.727739726027398e-06, "loss": 0.0028, "step": 998 }, { "epoch": 0.4276541095890411, "grad_norm": 5.4585280418396, "learning_rate": 5.72345890410959e-06, "loss": 0.0141, "step": 999 }, { "epoch": 0.4280821917808219, "grad_norm": 44.929630279541016, "learning_rate": 5.719178082191781e-06, "loss": 0.0831, "step": 1000 }, { "epoch": 0.4285102739726027, "grad_norm": 30.443269729614258, "learning_rate": 5.714897260273973e-06, "loss": 0.0565, "step": 1001 }, { "epoch": 0.4289383561643836, "grad_norm": 247.00213623046875, "learning_rate": 5.710616438356164e-06, "loss": 0.6668, "step": 1002 }, { "epoch": 0.4293664383561644, "grad_norm": 12.726300239562988, "learning_rate": 5.706335616438357e-06, "loss": 0.0172, "step": 1003 }, { "epoch": 0.4297945205479452, "grad_norm": 14.520042419433594, "learning_rate": 5.7020547945205485e-06, "loss": 0.0242, "step": 1004 }, { "epoch": 0.430222602739726, "grad_norm": 11.606200218200684, "learning_rate": 5.69777397260274e-06, "loss": 0.013, "step": 1005 }, { "epoch": 0.4306506849315068, "grad_norm": 94.65902709960938, "learning_rate": 5.6934931506849326e-06, "loss": 0.2504, "step": 1006 }, { "epoch": 0.4310787671232877, "grad_norm": 51.457435607910156, "learning_rate": 5.689212328767124e-06, "loss": 0.2499, "step": 1007 }, { "epoch": 0.4315068493150685, "grad_norm": 16.64958381652832, "learning_rate": 5.684931506849316e-06, "loss": 0.019, "step": 1008 }, { "epoch": 0.4319349315068493, "grad_norm": 11.332723617553711, "learning_rate": 5.680650684931507e-06, "loss": 0.0154, "step": 1009 }, { "epoch": 0.4323630136986301, "grad_norm": 14.569085121154785, "learning_rate": 5.676369863013699e-06, "loss": 0.0156, "step": 1010 }, { "epoch": 0.432791095890411, "grad_norm": 2.6400537490844727, "learning_rate": 5.672089041095891e-06, "loss": 0.0037, "step": 1011 }, { "epoch": 0.4332191780821918, "grad_norm": 91.3015365600586, "learning_rate": 5.667808219178082e-06, "loss": 0.0903, "step": 1012 }, { "epoch": 0.4336472602739726, "grad_norm": 3.3731813430786133, "learning_rate": 5.663527397260275e-06, "loss": 0.0032, "step": 1013 }, { "epoch": 0.4340753424657534, "grad_norm": 5.062733173370361, "learning_rate": 5.6592465753424665e-06, "loss": 0.0026, "step": 1014 }, { "epoch": 0.4345034246575342, "grad_norm": 119.99620056152344, "learning_rate": 5.654965753424658e-06, "loss": 0.3801, "step": 1015 }, { "epoch": 0.4349315068493151, "grad_norm": 12.00499153137207, "learning_rate": 5.6506849315068506e-06, "loss": 0.0097, "step": 1016 }, { "epoch": 0.4353595890410959, "grad_norm": 22.19403076171875, "learning_rate": 5.646404109589042e-06, "loss": 0.0323, "step": 1017 }, { "epoch": 0.4357876712328767, "grad_norm": 2.039998769760132, "learning_rate": 5.642123287671233e-06, "loss": 0.0011, "step": 1018 }, { "epoch": 0.4362157534246575, "grad_norm": 65.08935546875, "learning_rate": 5.637842465753425e-06, "loss": 0.0784, "step": 1019 }, { "epoch": 0.4366438356164384, "grad_norm": 74.95030212402344, "learning_rate": 5.633561643835616e-06, "loss": 0.3467, "step": 1020 }, { "epoch": 0.4370719178082192, "grad_norm": 6.750339984893799, "learning_rate": 5.629280821917809e-06, "loss": 0.0078, "step": 1021 }, { "epoch": 0.4375, "grad_norm": 3.5614428520202637, "learning_rate": 5.625e-06, "loss": 0.0064, "step": 1022 }, { "epoch": 0.4379280821917808, "grad_norm": 69.19539642333984, "learning_rate": 5.620719178082192e-06, "loss": 0.3293, "step": 1023 }, { "epoch": 0.4383561643835616, "grad_norm": 87.89957427978516, "learning_rate": 5.6164383561643845e-06, "loss": 0.131, "step": 1024 }, { "epoch": 0.4387842465753425, "grad_norm": 1.638093113899231, "learning_rate": 5.612157534246576e-06, "loss": 0.0027, "step": 1025 }, { "epoch": 0.4392123287671233, "grad_norm": 134.06517028808594, "learning_rate": 5.607876712328768e-06, "loss": 0.1852, "step": 1026 }, { "epoch": 0.4396404109589041, "grad_norm": 256.1405334472656, "learning_rate": 5.6035958904109585e-06, "loss": 0.1877, "step": 1027 }, { "epoch": 0.4400684931506849, "grad_norm": 133.2291717529297, "learning_rate": 5.599315068493151e-06, "loss": 0.4312, "step": 1028 }, { "epoch": 0.4404965753424658, "grad_norm": 69.69784545898438, "learning_rate": 5.595034246575343e-06, "loss": 0.1255, "step": 1029 }, { "epoch": 0.4409246575342466, "grad_norm": 12.753862380981445, "learning_rate": 5.590753424657534e-06, "loss": 0.0189, "step": 1030 }, { "epoch": 0.4413527397260274, "grad_norm": 130.07652282714844, "learning_rate": 5.586472602739727e-06, "loss": 0.1098, "step": 1031 }, { "epoch": 0.4417808219178082, "grad_norm": 90.9600830078125, "learning_rate": 5.582191780821918e-06, "loss": 0.1334, "step": 1032 }, { "epoch": 0.442208904109589, "grad_norm": 52.45614242553711, "learning_rate": 5.57791095890411e-06, "loss": 0.4741, "step": 1033 }, { "epoch": 0.4426369863013699, "grad_norm": 49.997833251953125, "learning_rate": 5.5736301369863025e-06, "loss": 0.085, "step": 1034 }, { "epoch": 0.4430650684931507, "grad_norm": 10.235312461853027, "learning_rate": 5.569349315068494e-06, "loss": 0.0086, "step": 1035 }, { "epoch": 0.4434931506849315, "grad_norm": 3.4895875453948975, "learning_rate": 5.565068493150685e-06, "loss": 0.0034, "step": 1036 }, { "epoch": 0.4439212328767123, "grad_norm": 35.31819152832031, "learning_rate": 5.5607876712328765e-06, "loss": 0.0464, "step": 1037 }, { "epoch": 0.4443493150684932, "grad_norm": 8.523208618164062, "learning_rate": 5.556506849315069e-06, "loss": 0.0124, "step": 1038 }, { "epoch": 0.444777397260274, "grad_norm": 1.8957252502441406, "learning_rate": 5.552226027397261e-06, "loss": 0.0049, "step": 1039 }, { "epoch": 0.4452054794520548, "grad_norm": 3.5509464740753174, "learning_rate": 5.547945205479452e-06, "loss": 0.0027, "step": 1040 }, { "epoch": 0.4456335616438356, "grad_norm": 11.53720760345459, "learning_rate": 5.543664383561645e-06, "loss": 0.0113, "step": 1041 }, { "epoch": 0.4460616438356164, "grad_norm": 75.25436401367188, "learning_rate": 5.539383561643836e-06, "loss": 0.7256, "step": 1042 }, { "epoch": 0.4464897260273973, "grad_norm": 6.704931735992432, "learning_rate": 5.535102739726028e-06, "loss": 0.0042, "step": 1043 }, { "epoch": 0.4469178082191781, "grad_norm": 2.722788095474243, "learning_rate": 5.5308219178082205e-06, "loss": 0.0022, "step": 1044 }, { "epoch": 0.4473458904109589, "grad_norm": 1.9630862474441528, "learning_rate": 5.526541095890411e-06, "loss": 0.0037, "step": 1045 }, { "epoch": 0.4477739726027397, "grad_norm": 2.7755000591278076, "learning_rate": 5.522260273972603e-06, "loss": 0.0077, "step": 1046 }, { "epoch": 0.4482020547945205, "grad_norm": 6.125967979431152, "learning_rate": 5.5179794520547945e-06, "loss": 0.0083, "step": 1047 }, { "epoch": 0.4486301369863014, "grad_norm": 2.91430401802063, "learning_rate": 5.513698630136987e-06, "loss": 0.0043, "step": 1048 }, { "epoch": 0.4490582191780822, "grad_norm": 3.851900339126587, "learning_rate": 5.509417808219179e-06, "loss": 0.0025, "step": 1049 }, { "epoch": 0.449486301369863, "grad_norm": 3.933361530303955, "learning_rate": 5.50513698630137e-06, "loss": 0.0021, "step": 1050 }, { "epoch": 0.4499143835616438, "grad_norm": 7.407456874847412, "learning_rate": 5.500856164383562e-06, "loss": 0.0048, "step": 1051 }, { "epoch": 0.4503424657534247, "grad_norm": 108.24439239501953, "learning_rate": 5.496575342465754e-06, "loss": 0.067, "step": 1052 }, { "epoch": 0.4507705479452055, "grad_norm": 592.7552490234375, "learning_rate": 5.492294520547946e-06, "loss": 2.1179, "step": 1053 }, { "epoch": 0.4511986301369863, "grad_norm": 38.160369873046875, "learning_rate": 5.488013698630137e-06, "loss": 0.0561, "step": 1054 }, { "epoch": 0.4516267123287671, "grad_norm": 7.388355731964111, "learning_rate": 5.483732876712328e-06, "loss": 0.0046, "step": 1055 }, { "epoch": 0.4520547945205479, "grad_norm": 56.4323616027832, "learning_rate": 5.479452054794521e-06, "loss": 0.2282, "step": 1056 }, { "epoch": 0.4524828767123288, "grad_norm": 9.330334663391113, "learning_rate": 5.4751712328767125e-06, "loss": 0.0062, "step": 1057 }, { "epoch": 0.4529109589041096, "grad_norm": 14.37112808227539, "learning_rate": 5.470890410958904e-06, "loss": 0.0221, "step": 1058 }, { "epoch": 0.4533390410958904, "grad_norm": 12.330192565917969, "learning_rate": 5.466609589041097e-06, "loss": 0.0126, "step": 1059 }, { "epoch": 0.4537671232876712, "grad_norm": 5.691786766052246, "learning_rate": 5.462328767123288e-06, "loss": 0.0038, "step": 1060 }, { "epoch": 0.4541952054794521, "grad_norm": 197.26828002929688, "learning_rate": 5.45804794520548e-06, "loss": 0.1471, "step": 1061 }, { "epoch": 0.4546232876712329, "grad_norm": 7.102381229400635, "learning_rate": 5.453767123287672e-06, "loss": 0.004, "step": 1062 }, { "epoch": 0.4550513698630137, "grad_norm": 9.380279541015625, "learning_rate": 5.449486301369864e-06, "loss": 0.0072, "step": 1063 }, { "epoch": 0.4554794520547945, "grad_norm": 2.107924699783325, "learning_rate": 5.445205479452055e-06, "loss": 0.0016, "step": 1064 }, { "epoch": 0.4559075342465753, "grad_norm": 0.7310783267021179, "learning_rate": 5.440924657534246e-06, "loss": 0.0014, "step": 1065 }, { "epoch": 0.4563356164383562, "grad_norm": 99.61133575439453, "learning_rate": 5.436643835616439e-06, "loss": 0.0589, "step": 1066 }, { "epoch": 0.456763698630137, "grad_norm": 57.76700210571289, "learning_rate": 5.4323630136986305e-06, "loss": 0.5709, "step": 1067 }, { "epoch": 0.4571917808219178, "grad_norm": 76.9726333618164, "learning_rate": 5.428082191780822e-06, "loss": 0.5312, "step": 1068 }, { "epoch": 0.4576198630136986, "grad_norm": 5.113916873931885, "learning_rate": 5.423801369863015e-06, "loss": 0.0038, "step": 1069 }, { "epoch": 0.4580479452054795, "grad_norm": 33.631290435791016, "learning_rate": 5.419520547945206e-06, "loss": 0.025, "step": 1070 }, { "epoch": 0.4584760273972603, "grad_norm": 5.303081035614014, "learning_rate": 5.415239726027398e-06, "loss": 0.0032, "step": 1071 }, { "epoch": 0.4589041095890411, "grad_norm": 2.867372512817383, "learning_rate": 5.41095890410959e-06, "loss": 0.0032, "step": 1072 }, { "epoch": 0.4593321917808219, "grad_norm": 8.176352500915527, "learning_rate": 5.406678082191781e-06, "loss": 0.02, "step": 1073 }, { "epoch": 0.4597602739726027, "grad_norm": 5.386654376983643, "learning_rate": 5.402397260273973e-06, "loss": 0.004, "step": 1074 }, { "epoch": 0.4601883561643836, "grad_norm": 34.52302932739258, "learning_rate": 5.398116438356164e-06, "loss": 0.0504, "step": 1075 }, { "epoch": 0.4606164383561644, "grad_norm": 7.721161842346191, "learning_rate": 5.393835616438357e-06, "loss": 0.0048, "step": 1076 }, { "epoch": 0.4610445205479452, "grad_norm": 4.925795078277588, "learning_rate": 5.3895547945205485e-06, "loss": 0.0071, "step": 1077 }, { "epoch": 0.461472602739726, "grad_norm": 282.0859375, "learning_rate": 5.38527397260274e-06, "loss": 0.3559, "step": 1078 }, { "epoch": 0.4619006849315068, "grad_norm": 8.077362060546875, "learning_rate": 5.380993150684933e-06, "loss": 0.0062, "step": 1079 }, { "epoch": 0.4623287671232877, "grad_norm": 2.211637020111084, "learning_rate": 5.376712328767124e-06, "loss": 0.0032, "step": 1080 }, { "epoch": 0.4627568493150685, "grad_norm": 4.208451747894287, "learning_rate": 5.372431506849316e-06, "loss": 0.0047, "step": 1081 }, { "epoch": 0.4631849315068493, "grad_norm": 1.9387317895889282, "learning_rate": 5.368150684931507e-06, "loss": 0.0012, "step": 1082 }, { "epoch": 0.4636130136986301, "grad_norm": 8.034186363220215, "learning_rate": 5.363869863013699e-06, "loss": 0.0048, "step": 1083 }, { "epoch": 0.464041095890411, "grad_norm": 1.5299983024597168, "learning_rate": 5.359589041095891e-06, "loss": 0.003, "step": 1084 }, { "epoch": 0.4644691780821918, "grad_norm": 2.7010157108306885, "learning_rate": 5.355308219178082e-06, "loss": 0.0017, "step": 1085 }, { "epoch": 0.4648972602739726, "grad_norm": 0.7659898996353149, "learning_rate": 5.351027397260274e-06, "loss": 0.0013, "step": 1086 }, { "epoch": 0.4653253424657534, "grad_norm": 84.79753875732422, "learning_rate": 5.3467465753424665e-06, "loss": 0.2979, "step": 1087 }, { "epoch": 0.4657534246575342, "grad_norm": 8.105289459228516, "learning_rate": 5.342465753424658e-06, "loss": 0.0045, "step": 1088 }, { "epoch": 0.4661815068493151, "grad_norm": 119.7294692993164, "learning_rate": 5.33818493150685e-06, "loss": 0.2155, "step": 1089 }, { "epoch": 0.4666095890410959, "grad_norm": 6.688464164733887, "learning_rate": 5.333904109589042e-06, "loss": 0.0036, "step": 1090 }, { "epoch": 0.4670376712328767, "grad_norm": 1.2618751525878906, "learning_rate": 5.329623287671233e-06, "loss": 0.002, "step": 1091 }, { "epoch": 0.4674657534246575, "grad_norm": 4.874697208404541, "learning_rate": 5.325342465753425e-06, "loss": 0.002, "step": 1092 }, { "epoch": 0.4678938356164384, "grad_norm": 6.419432163238525, "learning_rate": 5.321061643835616e-06, "loss": 0.005, "step": 1093 }, { "epoch": 0.4683219178082192, "grad_norm": 37.74488830566406, "learning_rate": 5.316780821917809e-06, "loss": 0.0139, "step": 1094 }, { "epoch": 0.46875, "grad_norm": 15.277942657470703, "learning_rate": 5.3125e-06, "loss": 0.0161, "step": 1095 }, { "epoch": 0.4691780821917808, "grad_norm": 0.6786008477210999, "learning_rate": 5.308219178082192e-06, "loss": 0.0003, "step": 1096 }, { "epoch": 0.4696061643835616, "grad_norm": 1.817502498626709, "learning_rate": 5.3039383561643845e-06, "loss": 0.0008, "step": 1097 }, { "epoch": 0.4700342465753425, "grad_norm": 5.423282146453857, "learning_rate": 5.299657534246576e-06, "loss": 0.0054, "step": 1098 }, { "epoch": 0.4704623287671233, "grad_norm": 1.61270010471344, "learning_rate": 5.295376712328768e-06, "loss": 0.0024, "step": 1099 }, { "epoch": 0.4708904109589041, "grad_norm": 4.167280197143555, "learning_rate": 5.2910958904109585e-06, "loss": 0.0016, "step": 1100 }, { "epoch": 0.4713184931506849, "grad_norm": 18.08887481689453, "learning_rate": 5.286815068493151e-06, "loss": 0.0184, "step": 1101 }, { "epoch": 0.4717465753424658, "grad_norm": 104.22652435302734, "learning_rate": 5.282534246575343e-06, "loss": 0.1563, "step": 1102 }, { "epoch": 0.4721746575342466, "grad_norm": 1.2675684690475464, "learning_rate": 5.278253424657534e-06, "loss": 0.0008, "step": 1103 }, { "epoch": 0.4726027397260274, "grad_norm": 12.911229133605957, "learning_rate": 5.273972602739727e-06, "loss": 0.011, "step": 1104 }, { "epoch": 0.4730308219178082, "grad_norm": 15.148597717285156, "learning_rate": 5.269691780821918e-06, "loss": 0.0388, "step": 1105 }, { "epoch": 0.473458904109589, "grad_norm": 73.49935150146484, "learning_rate": 5.26541095890411e-06, "loss": 0.5354, "step": 1106 }, { "epoch": 0.4738869863013699, "grad_norm": 6.196122169494629, "learning_rate": 5.2611301369863025e-06, "loss": 0.0039, "step": 1107 }, { "epoch": 0.4743150684931507, "grad_norm": 144.6229248046875, "learning_rate": 5.256849315068494e-06, "loss": 0.6136, "step": 1108 }, { "epoch": 0.4747431506849315, "grad_norm": 59.92033767700195, "learning_rate": 5.252568493150685e-06, "loss": 0.4651, "step": 1109 }, { "epoch": 0.4751712328767123, "grad_norm": 76.04388427734375, "learning_rate": 5.2482876712328765e-06, "loss": 0.707, "step": 1110 }, { "epoch": 0.4755993150684932, "grad_norm": 10.956904411315918, "learning_rate": 5.244006849315069e-06, "loss": 0.008, "step": 1111 }, { "epoch": 0.476027397260274, "grad_norm": 110.64356231689453, "learning_rate": 5.239726027397261e-06, "loss": 1.2344, "step": 1112 }, { "epoch": 0.4764554794520548, "grad_norm": 2.2644340991973877, "learning_rate": 5.235445205479452e-06, "loss": 0.0049, "step": 1113 }, { "epoch": 0.4768835616438356, "grad_norm": 8.147318840026855, "learning_rate": 5.231164383561645e-06, "loss": 0.0052, "step": 1114 }, { "epoch": 0.4773116438356164, "grad_norm": 10.69848918914795, "learning_rate": 5.226883561643836e-06, "loss": 0.0108, "step": 1115 }, { "epoch": 0.4777397260273973, "grad_norm": 7.460585117340088, "learning_rate": 5.222602739726028e-06, "loss": 0.013, "step": 1116 }, { "epoch": 0.4781678082191781, "grad_norm": 65.72416687011719, "learning_rate": 5.2183219178082205e-06, "loss": 0.5966, "step": 1117 }, { "epoch": 0.4785958904109589, "grad_norm": 2.073444128036499, "learning_rate": 5.214041095890411e-06, "loss": 0.0016, "step": 1118 }, { "epoch": 0.4790239726027397, "grad_norm": 55.670166015625, "learning_rate": 5.209760273972603e-06, "loss": 0.595, "step": 1119 }, { "epoch": 0.4794520547945205, "grad_norm": 11.451766967773438, "learning_rate": 5.2054794520547945e-06, "loss": 0.011, "step": 1120 }, { "epoch": 0.4798801369863014, "grad_norm": 59.376686096191406, "learning_rate": 5.201198630136986e-06, "loss": 0.3404, "step": 1121 }, { "epoch": 0.4803082191780822, "grad_norm": 4.653728008270264, "learning_rate": 5.196917808219179e-06, "loss": 0.0051, "step": 1122 }, { "epoch": 0.480736301369863, "grad_norm": 1.979022741317749, "learning_rate": 5.19263698630137e-06, "loss": 0.0033, "step": 1123 }, { "epoch": 0.4811643835616438, "grad_norm": 2.2930779457092285, "learning_rate": 5.188356164383562e-06, "loss": 0.002, "step": 1124 }, { "epoch": 0.4815924657534247, "grad_norm": 9.91403865814209, "learning_rate": 5.184075342465754e-06, "loss": 0.0096, "step": 1125 }, { "epoch": 0.4820205479452055, "grad_norm": 2.7068562507629395, "learning_rate": 5.179794520547946e-06, "loss": 0.0047, "step": 1126 }, { "epoch": 0.4824486301369863, "grad_norm": 9.261037826538086, "learning_rate": 5.175513698630137e-06, "loss": 0.0079, "step": 1127 }, { "epoch": 0.4828767123287671, "grad_norm": 4.100689888000488, "learning_rate": 5.171232876712328e-06, "loss": 0.0073, "step": 1128 }, { "epoch": 0.4833047945205479, "grad_norm": 8.450872421264648, "learning_rate": 5.166952054794521e-06, "loss": 0.0059, "step": 1129 }, { "epoch": 0.4837328767123288, "grad_norm": 1.3383041620254517, "learning_rate": 5.1626712328767125e-06, "loss": 0.0031, "step": 1130 }, { "epoch": 0.4841609589041096, "grad_norm": 12.379931449890137, "learning_rate": 5.158390410958904e-06, "loss": 0.0129, "step": 1131 }, { "epoch": 0.4845890410958904, "grad_norm": 2.050520658493042, "learning_rate": 5.154109589041097e-06, "loss": 0.0016, "step": 1132 }, { "epoch": 0.4850171232876712, "grad_norm": 76.57479095458984, "learning_rate": 5.149828767123288e-06, "loss": 0.2637, "step": 1133 }, { "epoch": 0.4854452054794521, "grad_norm": 61.557861328125, "learning_rate": 5.14554794520548e-06, "loss": 0.1533, "step": 1134 }, { "epoch": 0.4858732876712329, "grad_norm": 5.484839916229248, "learning_rate": 5.141267123287672e-06, "loss": 0.0034, "step": 1135 }, { "epoch": 0.4863013698630137, "grad_norm": 0.5172860026359558, "learning_rate": 5.136986301369864e-06, "loss": 0.0006, "step": 1136 }, { "epoch": 0.4867294520547945, "grad_norm": 2.6900784969329834, "learning_rate": 5.132705479452055e-06, "loss": 0.0037, "step": 1137 }, { "epoch": 0.4871575342465753, "grad_norm": 2.101417303085327, "learning_rate": 5.128424657534246e-06, "loss": 0.0051, "step": 1138 }, { "epoch": 0.4875856164383562, "grad_norm": 0.8406479358673096, "learning_rate": 5.124143835616439e-06, "loss": 0.0018, "step": 1139 }, { "epoch": 0.488013698630137, "grad_norm": 53.457767486572266, "learning_rate": 5.1198630136986305e-06, "loss": 0.0385, "step": 1140 }, { "epoch": 0.4884417808219178, "grad_norm": 6.119464874267578, "learning_rate": 5.115582191780822e-06, "loss": 0.0044, "step": 1141 }, { "epoch": 0.4888698630136986, "grad_norm": 4.63554573059082, "learning_rate": 5.111301369863015e-06, "loss": 0.0043, "step": 1142 }, { "epoch": 0.4892979452054795, "grad_norm": 11.250064849853516, "learning_rate": 5.107020547945206e-06, "loss": 0.0122, "step": 1143 }, { "epoch": 0.4897260273972603, "grad_norm": 7.577165126800537, "learning_rate": 5.102739726027398e-06, "loss": 0.0088, "step": 1144 }, { "epoch": 0.4901541095890411, "grad_norm": 107.66679382324219, "learning_rate": 5.09845890410959e-06, "loss": 0.0655, "step": 1145 }, { "epoch": 0.4905821917808219, "grad_norm": 459.14508056640625, "learning_rate": 5.094178082191781e-06, "loss": 0.9114, "step": 1146 }, { "epoch": 0.4910102739726027, "grad_norm": 10.55251407623291, "learning_rate": 5.089897260273973e-06, "loss": 0.0088, "step": 1147 }, { "epoch": 0.4914383561643836, "grad_norm": 7.23561954498291, "learning_rate": 5.085616438356164e-06, "loss": 0.0161, "step": 1148 }, { "epoch": 0.4918664383561644, "grad_norm": 69.8739013671875, "learning_rate": 5.081335616438357e-06, "loss": 0.4613, "step": 1149 }, { "epoch": 0.4922945205479452, "grad_norm": 13.904727935791016, "learning_rate": 5.0770547945205485e-06, "loss": 0.0163, "step": 1150 }, { "epoch": 0.492722602739726, "grad_norm": 4.524500846862793, "learning_rate": 5.07277397260274e-06, "loss": 0.0023, "step": 1151 }, { "epoch": 0.4931506849315068, "grad_norm": 71.3212890625, "learning_rate": 5.068493150684932e-06, "loss": 0.7202, "step": 1152 }, { "epoch": 0.4935787671232877, "grad_norm": 3.704495668411255, "learning_rate": 5.064212328767124e-06, "loss": 0.0027, "step": 1153 }, { "epoch": 0.4940068493150685, "grad_norm": 14.180179595947266, "learning_rate": 5.059931506849316e-06, "loss": 0.0099, "step": 1154 }, { "epoch": 0.4944349315068493, "grad_norm": 57.565799713134766, "learning_rate": 5.055650684931507e-06, "loss": 0.0359, "step": 1155 }, { "epoch": 0.4948630136986301, "grad_norm": 8.604300498962402, "learning_rate": 5.051369863013698e-06, "loss": 0.0082, "step": 1156 }, { "epoch": 0.495291095890411, "grad_norm": 12.653788566589355, "learning_rate": 5.047089041095891e-06, "loss": 0.0162, "step": 1157 }, { "epoch": 0.4957191780821918, "grad_norm": 38.62876510620117, "learning_rate": 5.042808219178082e-06, "loss": 0.0317, "step": 1158 }, { "epoch": 0.4961472602739726, "grad_norm": 74.88848114013672, "learning_rate": 5.038527397260274e-06, "loss": 0.2153, "step": 1159 }, { "epoch": 0.4965753424657534, "grad_norm": 34.85064697265625, "learning_rate": 5.0342465753424665e-06, "loss": 0.0242, "step": 1160 }, { "epoch": 0.4970034246575342, "grad_norm": 20.438159942626953, "learning_rate": 5.029965753424658e-06, "loss": 0.0297, "step": 1161 }, { "epoch": 0.4974315068493151, "grad_norm": 16.130592346191406, "learning_rate": 5.02568493150685e-06, "loss": 0.0186, "step": 1162 }, { "epoch": 0.4978595890410959, "grad_norm": 15.19290542602539, "learning_rate": 5.021404109589042e-06, "loss": 0.0159, "step": 1163 }, { "epoch": 0.4982876712328767, "grad_norm": 9.5408296585083, "learning_rate": 5.017123287671233e-06, "loss": 0.0094, "step": 1164 }, { "epoch": 0.4987157534246575, "grad_norm": 11.414407730102539, "learning_rate": 5.012842465753425e-06, "loss": 0.0095, "step": 1165 }, { "epoch": 0.4991438356164384, "grad_norm": 11.906081199645996, "learning_rate": 5.008561643835616e-06, "loss": 0.0098, "step": 1166 }, { "epoch": 0.4995719178082192, "grad_norm": 8.756658554077148, "learning_rate": 5.004280821917809e-06, "loss": 0.0089, "step": 1167 }, { "epoch": 0.5, "grad_norm": 66.50926971435547, "learning_rate": 5e-06, "loss": 0.7325, "step": 1168 }, { "epoch": 0.5004280821917808, "grad_norm": 14.19948673248291, "learning_rate": 4.995719178082192e-06, "loss": 0.0148, "step": 1169 }, { "epoch": 0.5008561643835616, "grad_norm": 13.950363159179688, "learning_rate": 4.9914383561643845e-06, "loss": 0.0229, "step": 1170 }, { "epoch": 0.5012842465753424, "grad_norm": 55.29323959350586, "learning_rate": 4.987157534246575e-06, "loss": 0.0984, "step": 1171 }, { "epoch": 0.5017123287671232, "grad_norm": 2.1963953971862793, "learning_rate": 4.982876712328768e-06, "loss": 0.0015, "step": 1172 }, { "epoch": 0.5021404109589042, "grad_norm": 6.738938331604004, "learning_rate": 4.978595890410959e-06, "loss": 0.0042, "step": 1173 }, { "epoch": 0.502568493150685, "grad_norm": 8.572482109069824, "learning_rate": 4.974315068493151e-06, "loss": 0.0077, "step": 1174 }, { "epoch": 0.5029965753424658, "grad_norm": 20.42005729675293, "learning_rate": 4.970034246575343e-06, "loss": 0.0344, "step": 1175 }, { "epoch": 0.5034246575342466, "grad_norm": 89.95914459228516, "learning_rate": 4.965753424657534e-06, "loss": 0.4775, "step": 1176 }, { "epoch": 0.5038527397260274, "grad_norm": 23.952512741088867, "learning_rate": 4.961472602739727e-06, "loss": 0.0157, "step": 1177 }, { "epoch": 0.5042808219178082, "grad_norm": 12.89033317565918, "learning_rate": 4.957191780821918e-06, "loss": 0.013, "step": 1178 }, { "epoch": 0.504708904109589, "grad_norm": 19.480037689208984, "learning_rate": 4.95291095890411e-06, "loss": 0.0288, "step": 1179 }, { "epoch": 0.5051369863013698, "grad_norm": 18.2451114654541, "learning_rate": 4.948630136986302e-06, "loss": 0.0226, "step": 1180 }, { "epoch": 0.5055650684931506, "grad_norm": 11.92221450805664, "learning_rate": 4.944349315068493e-06, "loss": 0.0194, "step": 1181 }, { "epoch": 0.5059931506849316, "grad_norm": 12.277036666870117, "learning_rate": 4.940068493150686e-06, "loss": 0.0106, "step": 1182 }, { "epoch": 0.5064212328767124, "grad_norm": 8.934948921203613, "learning_rate": 4.935787671232877e-06, "loss": 0.0082, "step": 1183 }, { "epoch": 0.5068493150684932, "grad_norm": 0.9689731001853943, "learning_rate": 4.931506849315069e-06, "loss": 0.0026, "step": 1184 }, { "epoch": 0.507277397260274, "grad_norm": 3.336622714996338, "learning_rate": 4.927226027397261e-06, "loss": 0.0026, "step": 1185 }, { "epoch": 0.5077054794520548, "grad_norm": 14.436705589294434, "learning_rate": 4.922945205479452e-06, "loss": 0.0165, "step": 1186 }, { "epoch": 0.5081335616438356, "grad_norm": 63.632755279541016, "learning_rate": 4.918664383561644e-06, "loss": 0.045, "step": 1187 }, { "epoch": 0.5085616438356164, "grad_norm": 10.411566734313965, "learning_rate": 4.914383561643836e-06, "loss": 0.0095, "step": 1188 }, { "epoch": 0.5089897260273972, "grad_norm": 10.311022758483887, "learning_rate": 4.910102739726027e-06, "loss": 0.0079, "step": 1189 }, { "epoch": 0.509417808219178, "grad_norm": 28.62019920349121, "learning_rate": 4.90582191780822e-06, "loss": 0.0458, "step": 1190 }, { "epoch": 0.509845890410959, "grad_norm": 35.711036682128906, "learning_rate": 4.901541095890411e-06, "loss": 0.0451, "step": 1191 }, { "epoch": 0.5102739726027398, "grad_norm": 104.48796844482422, "learning_rate": 4.897260273972603e-06, "loss": 0.3269, "step": 1192 }, { "epoch": 0.5107020547945206, "grad_norm": 100.91781616210938, "learning_rate": 4.892979452054795e-06, "loss": 0.22, "step": 1193 }, { "epoch": 0.5111301369863014, "grad_norm": 85.9109115600586, "learning_rate": 4.888698630136986e-06, "loss": 0.4619, "step": 1194 }, { "epoch": 0.5115582191780822, "grad_norm": 17.553979873657227, "learning_rate": 4.884417808219179e-06, "loss": 0.0228, "step": 1195 }, { "epoch": 0.511986301369863, "grad_norm": 7.938053131103516, "learning_rate": 4.88013698630137e-06, "loss": 0.0094, "step": 1196 }, { "epoch": 0.5124143835616438, "grad_norm": 8.572015762329102, "learning_rate": 4.875856164383562e-06, "loss": 0.0069, "step": 1197 }, { "epoch": 0.5128424657534246, "grad_norm": 10.718796730041504, "learning_rate": 4.8715753424657536e-06, "loss": 0.0092, "step": 1198 }, { "epoch": 0.5132705479452054, "grad_norm": 0.7646522521972656, "learning_rate": 4.867294520547945e-06, "loss": 0.002, "step": 1199 }, { "epoch": 0.5136986301369864, "grad_norm": 3.676908016204834, "learning_rate": 4.863013698630138e-06, "loss": 0.0061, "step": 1200 }, { "epoch": 0.5141267123287672, "grad_norm": 10.822122573852539, "learning_rate": 4.858732876712329e-06, "loss": 0.0145, "step": 1201 }, { "epoch": 0.514554794520548, "grad_norm": 8.054058074951172, "learning_rate": 4.854452054794521e-06, "loss": 0.0092, "step": 1202 }, { "epoch": 0.5149828767123288, "grad_norm": 7.673189163208008, "learning_rate": 4.8501712328767126e-06, "loss": 0.0065, "step": 1203 }, { "epoch": 0.5154109589041096, "grad_norm": 7.5489420890808105, "learning_rate": 4.845890410958904e-06, "loss": 0.005, "step": 1204 }, { "epoch": 0.5158390410958904, "grad_norm": 8.321257591247559, "learning_rate": 4.841609589041097e-06, "loss": 0.0056, "step": 1205 }, { "epoch": 0.5162671232876712, "grad_norm": 48.92940902709961, "learning_rate": 4.837328767123288e-06, "loss": 0.073, "step": 1206 }, { "epoch": 0.516695205479452, "grad_norm": 61.435752868652344, "learning_rate": 4.83304794520548e-06, "loss": 0.2836, "step": 1207 }, { "epoch": 0.5171232876712328, "grad_norm": 25.4943790435791, "learning_rate": 4.8287671232876716e-06, "loss": 0.0447, "step": 1208 }, { "epoch": 0.5175513698630136, "grad_norm": 13.400500297546387, "learning_rate": 4.824486301369863e-06, "loss": 0.0147, "step": 1209 }, { "epoch": 0.5179794520547946, "grad_norm": 6.7678093910217285, "learning_rate": 4.820205479452056e-06, "loss": 0.0058, "step": 1210 }, { "epoch": 0.5184075342465754, "grad_norm": 10.802887916564941, "learning_rate": 4.815924657534247e-06, "loss": 0.0085, "step": 1211 }, { "epoch": 0.5188356164383562, "grad_norm": 17.120424270629883, "learning_rate": 4.811643835616439e-06, "loss": 0.0208, "step": 1212 }, { "epoch": 0.519263698630137, "grad_norm": 61.196990966796875, "learning_rate": 4.8073630136986306e-06, "loss": 0.4257, "step": 1213 }, { "epoch": 0.5196917808219178, "grad_norm": 2.3595547676086426, "learning_rate": 4.803082191780822e-06, "loss": 0.0034, "step": 1214 }, { "epoch": 0.5201198630136986, "grad_norm": 5.0290446281433105, "learning_rate": 4.798801369863015e-06, "loss": 0.0048, "step": 1215 }, { "epoch": 0.5205479452054794, "grad_norm": 7.802610874176025, "learning_rate": 4.7945205479452054e-06, "loss": 0.005, "step": 1216 }, { "epoch": 0.5209760273972602, "grad_norm": 8.581430435180664, "learning_rate": 4.790239726027398e-06, "loss": 0.0059, "step": 1217 }, { "epoch": 0.521404109589041, "grad_norm": 103.98832702636719, "learning_rate": 4.7859589041095896e-06, "loss": 0.1897, "step": 1218 }, { "epoch": 0.521832191780822, "grad_norm": 49.695167541503906, "learning_rate": 4.781678082191781e-06, "loss": 0.1548, "step": 1219 }, { "epoch": 0.5222602739726028, "grad_norm": 6.994738578796387, "learning_rate": 4.777397260273973e-06, "loss": 0.0141, "step": 1220 }, { "epoch": 0.5226883561643836, "grad_norm": 12.28221321105957, "learning_rate": 4.7731164383561644e-06, "loss": 0.0141, "step": 1221 }, { "epoch": 0.5231164383561644, "grad_norm": 10.515260696411133, "learning_rate": 4.768835616438356e-06, "loss": 0.0089, "step": 1222 }, { "epoch": 0.5235445205479452, "grad_norm": 18.12851905822754, "learning_rate": 4.7645547945205485e-06, "loss": 0.0204, "step": 1223 }, { "epoch": 0.523972602739726, "grad_norm": 9.36546516418457, "learning_rate": 4.76027397260274e-06, "loss": 0.0115, "step": 1224 }, { "epoch": 0.5244006849315068, "grad_norm": 10.878954887390137, "learning_rate": 4.755993150684932e-06, "loss": 0.0105, "step": 1225 }, { "epoch": 0.5248287671232876, "grad_norm": 17.029361724853516, "learning_rate": 4.7517123287671234e-06, "loss": 0.02, "step": 1226 }, { "epoch": 0.5252568493150684, "grad_norm": 6.7078046798706055, "learning_rate": 4.747431506849315e-06, "loss": 0.0119, "step": 1227 }, { "epoch": 0.5256849315068494, "grad_norm": 3.281134605407715, "learning_rate": 4.7431506849315075e-06, "loss": 0.0032, "step": 1228 }, { "epoch": 0.5261130136986302, "grad_norm": 3.9276623725891113, "learning_rate": 4.738869863013699e-06, "loss": 0.0067, "step": 1229 }, { "epoch": 0.526541095890411, "grad_norm": 10.867420196533203, "learning_rate": 4.734589041095891e-06, "loss": 0.0048, "step": 1230 }, { "epoch": 0.5269691780821918, "grad_norm": 2.0156662464141846, "learning_rate": 4.7303082191780824e-06, "loss": 0.0022, "step": 1231 }, { "epoch": 0.5273972602739726, "grad_norm": 654.9430541992188, "learning_rate": 4.726027397260274e-06, "loss": 2.8773, "step": 1232 }, { "epoch": 0.5278253424657534, "grad_norm": 12.064836502075195, "learning_rate": 4.7217465753424665e-06, "loss": 0.0105, "step": 1233 }, { "epoch": 0.5282534246575342, "grad_norm": 6.685173511505127, "learning_rate": 4.717465753424658e-06, "loss": 0.0041, "step": 1234 }, { "epoch": 0.528681506849315, "grad_norm": 12.987231254577637, "learning_rate": 4.71318493150685e-06, "loss": 0.0114, "step": 1235 }, { "epoch": 0.5291095890410958, "grad_norm": 10.298890113830566, "learning_rate": 4.7089041095890414e-06, "loss": 0.0089, "step": 1236 }, { "epoch": 0.5295376712328768, "grad_norm": 23.350555419921875, "learning_rate": 4.704623287671233e-06, "loss": 0.0277, "step": 1237 }, { "epoch": 0.5299657534246576, "grad_norm": 16.90178108215332, "learning_rate": 4.7003424657534255e-06, "loss": 0.0099, "step": 1238 }, { "epoch": 0.5303938356164384, "grad_norm": 30.189075469970703, "learning_rate": 4.696061643835616e-06, "loss": 0.0507, "step": 1239 }, { "epoch": 0.5308219178082192, "grad_norm": 3.834439277648926, "learning_rate": 4.691780821917809e-06, "loss": 0.0045, "step": 1240 }, { "epoch": 0.53125, "grad_norm": 3.724006414413452, "learning_rate": 4.6875000000000004e-06, "loss": 0.0044, "step": 1241 }, { "epoch": 0.5316780821917808, "grad_norm": 176.75755310058594, "learning_rate": 4.683219178082192e-06, "loss": 0.1244, "step": 1242 }, { "epoch": 0.5321061643835616, "grad_norm": 5.83447265625, "learning_rate": 4.6789383561643845e-06, "loss": 0.0037, "step": 1243 }, { "epoch": 0.5325342465753424, "grad_norm": 11.795159339904785, "learning_rate": 4.674657534246575e-06, "loss": 0.0154, "step": 1244 }, { "epoch": 0.5329623287671232, "grad_norm": 7.021984577178955, "learning_rate": 4.670376712328768e-06, "loss": 0.0039, "step": 1245 }, { "epoch": 0.5333904109589042, "grad_norm": 1.0993725061416626, "learning_rate": 4.6660958904109594e-06, "loss": 0.0009, "step": 1246 }, { "epoch": 0.533818493150685, "grad_norm": 138.73333740234375, "learning_rate": 4.661815068493151e-06, "loss": 0.2923, "step": 1247 }, { "epoch": 0.5342465753424658, "grad_norm": 1.6468721628189087, "learning_rate": 4.657534246575343e-06, "loss": 0.001, "step": 1248 }, { "epoch": 0.5346746575342466, "grad_norm": 6.719359874725342, "learning_rate": 4.653253424657534e-06, "loss": 0.0032, "step": 1249 }, { "epoch": 0.5351027397260274, "grad_norm": 4.239284515380859, "learning_rate": 4.648972602739727e-06, "loss": 0.0019, "step": 1250 }, { "epoch": 0.5355308219178082, "grad_norm": 5.000033855438232, "learning_rate": 4.6446917808219184e-06, "loss": 0.0034, "step": 1251 }, { "epoch": 0.535958904109589, "grad_norm": 5.306197643280029, "learning_rate": 4.64041095890411e-06, "loss": 0.0039, "step": 1252 }, { "epoch": 0.5363869863013698, "grad_norm": 1.9134445190429688, "learning_rate": 4.636130136986302e-06, "loss": 0.0046, "step": 1253 }, { "epoch": 0.5368150684931506, "grad_norm": 39.46144104003906, "learning_rate": 4.631849315068493e-06, "loss": 0.0467, "step": 1254 }, { "epoch": 0.5372431506849316, "grad_norm": 7.83928108215332, "learning_rate": 4.627568493150685e-06, "loss": 0.0071, "step": 1255 }, { "epoch": 0.5376712328767124, "grad_norm": 26.302696228027344, "learning_rate": 4.6232876712328774e-06, "loss": 0.026, "step": 1256 }, { "epoch": 0.5380993150684932, "grad_norm": 16.11188316345215, "learning_rate": 4.619006849315068e-06, "loss": 0.0186, "step": 1257 }, { "epoch": 0.538527397260274, "grad_norm": 6.765133857727051, "learning_rate": 4.614726027397261e-06, "loss": 0.0051, "step": 1258 }, { "epoch": 0.5389554794520548, "grad_norm": 6.332149505615234, "learning_rate": 4.610445205479452e-06, "loss": 0.0057, "step": 1259 }, { "epoch": 0.5393835616438356, "grad_norm": 4.269348621368408, "learning_rate": 4.606164383561644e-06, "loss": 0.0023, "step": 1260 }, { "epoch": 0.5398116438356164, "grad_norm": 38.76060485839844, "learning_rate": 4.6018835616438364e-06, "loss": 0.0278, "step": 1261 }, { "epoch": 0.5402397260273972, "grad_norm": 2.9721896648406982, "learning_rate": 4.597602739726027e-06, "loss": 0.0074, "step": 1262 }, { "epoch": 0.540667808219178, "grad_norm": 111.07145690917969, "learning_rate": 4.59332191780822e-06, "loss": 0.459, "step": 1263 }, { "epoch": 0.541095890410959, "grad_norm": 59.67496871948242, "learning_rate": 4.589041095890411e-06, "loss": 0.0528, "step": 1264 }, { "epoch": 0.5415239726027398, "grad_norm": 11.926633834838867, "learning_rate": 4.584760273972603e-06, "loss": 0.0102, "step": 1265 }, { "epoch": 0.5419520547945206, "grad_norm": 12.641128540039062, "learning_rate": 4.5804794520547954e-06, "loss": 0.0117, "step": 1266 }, { "epoch": 0.5423801369863014, "grad_norm": 79.57364654541016, "learning_rate": 4.576198630136986e-06, "loss": 0.4559, "step": 1267 }, { "epoch": 0.5428082191780822, "grad_norm": 11.291837692260742, "learning_rate": 4.571917808219179e-06, "loss": 0.0095, "step": 1268 }, { "epoch": 0.543236301369863, "grad_norm": 10.375357627868652, "learning_rate": 4.56763698630137e-06, "loss": 0.0109, "step": 1269 }, { "epoch": 0.5436643835616438, "grad_norm": 9.463214874267578, "learning_rate": 4.563356164383562e-06, "loss": 0.0067, "step": 1270 }, { "epoch": 0.5440924657534246, "grad_norm": 11.554205894470215, "learning_rate": 4.559075342465754e-06, "loss": 0.0064, "step": 1271 }, { "epoch": 0.5445205479452054, "grad_norm": 0.3323765993118286, "learning_rate": 4.554794520547945e-06, "loss": 0.0009, "step": 1272 }, { "epoch": 0.5449486301369864, "grad_norm": 4.67681360244751, "learning_rate": 4.550513698630138e-06, "loss": 0.0022, "step": 1273 }, { "epoch": 0.5453767123287672, "grad_norm": 5.274454116821289, "learning_rate": 4.546232876712329e-06, "loss": 0.0039, "step": 1274 }, { "epoch": 0.545804794520548, "grad_norm": 1.8316044807434082, "learning_rate": 4.541952054794521e-06, "loss": 0.0019, "step": 1275 }, { "epoch": 0.5462328767123288, "grad_norm": 155.4766387939453, "learning_rate": 4.537671232876713e-06, "loss": 0.1199, "step": 1276 }, { "epoch": 0.5466609589041096, "grad_norm": 2.9463577270507812, "learning_rate": 4.533390410958904e-06, "loss": 0.0016, "step": 1277 }, { "epoch": 0.5470890410958904, "grad_norm": 371.9568176269531, "learning_rate": 4.529109589041097e-06, "loss": 0.8065, "step": 1278 }, { "epoch": 0.5475171232876712, "grad_norm": 2.3611817359924316, "learning_rate": 4.524828767123288e-06, "loss": 0.0023, "step": 1279 }, { "epoch": 0.547945205479452, "grad_norm": 2.187673568725586, "learning_rate": 4.52054794520548e-06, "loss": 0.002, "step": 1280 }, { "epoch": 0.5483732876712328, "grad_norm": 5.614506721496582, "learning_rate": 4.516267123287672e-06, "loss": 0.0027, "step": 1281 }, { "epoch": 0.5488013698630136, "grad_norm": 30.809206008911133, "learning_rate": 4.511986301369863e-06, "loss": 0.0376, "step": 1282 }, { "epoch": 0.5492294520547946, "grad_norm": 62.94560623168945, "learning_rate": 4.507705479452056e-06, "loss": 0.0513, "step": 1283 }, { "epoch": 0.5496575342465754, "grad_norm": 62.186065673828125, "learning_rate": 4.503424657534247e-06, "loss": 0.4273, "step": 1284 }, { "epoch": 0.5500856164383562, "grad_norm": 5.988548755645752, "learning_rate": 4.499143835616439e-06, "loss": 0.0087, "step": 1285 }, { "epoch": 0.550513698630137, "grad_norm": 3.6339237689971924, "learning_rate": 4.494863013698631e-06, "loss": 0.0043, "step": 1286 }, { "epoch": 0.5509417808219178, "grad_norm": 10.668782234191895, "learning_rate": 4.490582191780822e-06, "loss": 0.008, "step": 1287 }, { "epoch": 0.5513698630136986, "grad_norm": 8.648480415344238, "learning_rate": 4.486301369863014e-06, "loss": 0.0091, "step": 1288 }, { "epoch": 0.5517979452054794, "grad_norm": 8.088958740234375, "learning_rate": 4.4820205479452055e-06, "loss": 0.0054, "step": 1289 }, { "epoch": 0.5522260273972602, "grad_norm": 0.5095022320747375, "learning_rate": 4.477739726027397e-06, "loss": 0.0007, "step": 1290 }, { "epoch": 0.552654109589041, "grad_norm": 171.5633544921875, "learning_rate": 4.47345890410959e-06, "loss": 0.1875, "step": 1291 }, { "epoch": 0.553082191780822, "grad_norm": 1.2134604454040527, "learning_rate": 4.469178082191781e-06, "loss": 0.0031, "step": 1292 }, { "epoch": 0.5535102739726028, "grad_norm": 16.14531898498535, "learning_rate": 4.464897260273973e-06, "loss": 0.0108, "step": 1293 }, { "epoch": 0.5539383561643836, "grad_norm": 60.991512298583984, "learning_rate": 4.4606164383561645e-06, "loss": 0.0675, "step": 1294 }, { "epoch": 0.5543664383561644, "grad_norm": 203.21063232421875, "learning_rate": 4.456335616438356e-06, "loss": 0.2817, "step": 1295 }, { "epoch": 0.5547945205479452, "grad_norm": 6.695775508880615, "learning_rate": 4.4520547945205486e-06, "loss": 0.0049, "step": 1296 }, { "epoch": 0.555222602739726, "grad_norm": 52.098472595214844, "learning_rate": 4.44777397260274e-06, "loss": 0.2093, "step": 1297 }, { "epoch": 0.5556506849315068, "grad_norm": 6.677694320678711, "learning_rate": 4.443493150684932e-06, "loss": 0.0066, "step": 1298 }, { "epoch": 0.5560787671232876, "grad_norm": 1.1115559339523315, "learning_rate": 4.4392123287671235e-06, "loss": 0.0019, "step": 1299 }, { "epoch": 0.5565068493150684, "grad_norm": 112.1546630859375, "learning_rate": 4.434931506849315e-06, "loss": 0.2114, "step": 1300 }, { "epoch": 0.5569349315068494, "grad_norm": 2.7021117210388184, "learning_rate": 4.4306506849315076e-06, "loss": 0.0009, "step": 1301 }, { "epoch": 0.5573630136986302, "grad_norm": 3.204270362854004, "learning_rate": 4.426369863013699e-06, "loss": 0.0032, "step": 1302 }, { "epoch": 0.557791095890411, "grad_norm": 5.7866315841674805, "learning_rate": 4.422089041095891e-06, "loss": 0.007, "step": 1303 }, { "epoch": 0.5582191780821918, "grad_norm": 4.766517162322998, "learning_rate": 4.4178082191780825e-06, "loss": 0.0035, "step": 1304 }, { "epoch": 0.5586472602739726, "grad_norm": 2.4729862213134766, "learning_rate": 4.413527397260274e-06, "loss": 0.0029, "step": 1305 }, { "epoch": 0.5590753424657534, "grad_norm": 8.471940994262695, "learning_rate": 4.4092465753424666e-06, "loss": 0.0129, "step": 1306 }, { "epoch": 0.5595034246575342, "grad_norm": 6.214473724365234, "learning_rate": 4.404965753424658e-06, "loss": 0.0037, "step": 1307 }, { "epoch": 0.559931506849315, "grad_norm": 1.439353585243225, "learning_rate": 4.40068493150685e-06, "loss": 0.0025, "step": 1308 }, { "epoch": 0.5603595890410958, "grad_norm": 42.32218551635742, "learning_rate": 4.3964041095890415e-06, "loss": 0.0452, "step": 1309 }, { "epoch": 0.5607876712328768, "grad_norm": 4.737757682800293, "learning_rate": 4.392123287671233e-06, "loss": 0.0038, "step": 1310 }, { "epoch": 0.5612157534246576, "grad_norm": 1.030159592628479, "learning_rate": 4.3878424657534256e-06, "loss": 0.0012, "step": 1311 }, { "epoch": 0.5616438356164384, "grad_norm": 6.447063446044922, "learning_rate": 4.383561643835616e-06, "loss": 0.0043, "step": 1312 }, { "epoch": 0.5620719178082192, "grad_norm": 119.8042984008789, "learning_rate": 4.379280821917809e-06, "loss": 0.0732, "step": 1313 }, { "epoch": 0.5625, "grad_norm": 7.388429641723633, "learning_rate": 4.3750000000000005e-06, "loss": 0.0044, "step": 1314 }, { "epoch": 0.5629280821917808, "grad_norm": 3.267864942550659, "learning_rate": 4.370719178082192e-06, "loss": 0.0015, "step": 1315 }, { "epoch": 0.5633561643835616, "grad_norm": 39.35624313354492, "learning_rate": 4.3664383561643846e-06, "loss": 0.0234, "step": 1316 }, { "epoch": 0.5637842465753424, "grad_norm": 4.906305313110352, "learning_rate": 4.362157534246575e-06, "loss": 0.0024, "step": 1317 }, { "epoch": 0.5642123287671232, "grad_norm": 2.7436280250549316, "learning_rate": 4.357876712328768e-06, "loss": 0.0036, "step": 1318 }, { "epoch": 0.5646404109589042, "grad_norm": 86.44053649902344, "learning_rate": 4.3535958904109595e-06, "loss": 0.1351, "step": 1319 }, { "epoch": 0.565068493150685, "grad_norm": 10.85738754272461, "learning_rate": 4.349315068493151e-06, "loss": 0.0034, "step": 1320 }, { "epoch": 0.5654965753424658, "grad_norm": 34.56877136230469, "learning_rate": 4.345034246575343e-06, "loss": 0.0175, "step": 1321 }, { "epoch": 0.5659246575342466, "grad_norm": 3.4462716579437256, "learning_rate": 4.340753424657534e-06, "loss": 0.0059, "step": 1322 }, { "epoch": 0.5663527397260274, "grad_norm": 2.0911543369293213, "learning_rate": 4.336472602739726e-06, "loss": 0.001, "step": 1323 }, { "epoch": 0.5667808219178082, "grad_norm": 65.26193237304688, "learning_rate": 4.3321917808219185e-06, "loss": 0.0695, "step": 1324 }, { "epoch": 0.567208904109589, "grad_norm": 15.116117477416992, "learning_rate": 4.32791095890411e-06, "loss": 0.0162, "step": 1325 }, { "epoch": 0.5676369863013698, "grad_norm": 5.227285861968994, "learning_rate": 4.323630136986302e-06, "loss": 0.0042, "step": 1326 }, { "epoch": 0.5680650684931506, "grad_norm": 13.618903160095215, "learning_rate": 4.319349315068493e-06, "loss": 0.0058, "step": 1327 }, { "epoch": 0.5684931506849316, "grad_norm": 2.7537317276000977, "learning_rate": 4.315068493150685e-06, "loss": 0.0031, "step": 1328 }, { "epoch": 0.5689212328767124, "grad_norm": 48.94231033325195, "learning_rate": 4.3107876712328775e-06, "loss": 0.4732, "step": 1329 }, { "epoch": 0.5693493150684932, "grad_norm": 28.86013412475586, "learning_rate": 4.306506849315068e-06, "loss": 0.0472, "step": 1330 }, { "epoch": 0.569777397260274, "grad_norm": 9.386947631835938, "learning_rate": 4.302226027397261e-06, "loss": 0.0062, "step": 1331 }, { "epoch": 0.5702054794520548, "grad_norm": 111.19974517822266, "learning_rate": 4.297945205479452e-06, "loss": 0.4078, "step": 1332 }, { "epoch": 0.5706335616438356, "grad_norm": 1.120774745941162, "learning_rate": 4.293664383561644e-06, "loss": 0.003, "step": 1333 }, { "epoch": 0.5710616438356164, "grad_norm": 67.14391326904297, "learning_rate": 4.2893835616438365e-06, "loss": 0.0652, "step": 1334 }, { "epoch": 0.5714897260273972, "grad_norm": 4.1797709465026855, "learning_rate": 4.285102739726027e-06, "loss": 0.0018, "step": 1335 }, { "epoch": 0.571917808219178, "grad_norm": 2.2525241374969482, "learning_rate": 4.28082191780822e-06, "loss": 0.0044, "step": 1336 }, { "epoch": 0.572345890410959, "grad_norm": 1.5364301204681396, "learning_rate": 4.276541095890411e-06, "loss": 0.0014, "step": 1337 }, { "epoch": 0.5727739726027398, "grad_norm": 4.4892096519470215, "learning_rate": 4.272260273972603e-06, "loss": 0.0022, "step": 1338 }, { "epoch": 0.5732020547945206, "grad_norm": 7.955974578857422, "learning_rate": 4.2679794520547955e-06, "loss": 0.0052, "step": 1339 }, { "epoch": 0.5736301369863014, "grad_norm": 63.068695068359375, "learning_rate": 4.263698630136986e-06, "loss": 0.6995, "step": 1340 }, { "epoch": 0.5740582191780822, "grad_norm": 12.122851371765137, "learning_rate": 4.259417808219179e-06, "loss": 0.0096, "step": 1341 }, { "epoch": 0.574486301369863, "grad_norm": 8.563458442687988, "learning_rate": 4.25513698630137e-06, "loss": 0.0074, "step": 1342 }, { "epoch": 0.5749143835616438, "grad_norm": 55.30105209350586, "learning_rate": 4.250856164383562e-06, "loss": 0.4539, "step": 1343 }, { "epoch": 0.5753424657534246, "grad_norm": 0.9474047422409058, "learning_rate": 4.246575342465754e-06, "loss": 0.0017, "step": 1344 }, { "epoch": 0.5757705479452054, "grad_norm": 5.611225128173828, "learning_rate": 4.242294520547945e-06, "loss": 0.0034, "step": 1345 }, { "epoch": 0.5761986301369864, "grad_norm": 6.45542049407959, "learning_rate": 4.238013698630138e-06, "loss": 0.0035, "step": 1346 }, { "epoch": 0.5766267123287672, "grad_norm": 6.021223545074463, "learning_rate": 4.233732876712329e-06, "loss": 0.0049, "step": 1347 }, { "epoch": 0.577054794520548, "grad_norm": 15.790694236755371, "learning_rate": 4.229452054794521e-06, "loss": 0.0086, "step": 1348 }, { "epoch": 0.5774828767123288, "grad_norm": 17.275354385375977, "learning_rate": 4.225171232876713e-06, "loss": 0.0184, "step": 1349 }, { "epoch": 0.5779109589041096, "grad_norm": 14.398591041564941, "learning_rate": 4.220890410958904e-06, "loss": 0.0149, "step": 1350 }, { "epoch": 0.5783390410958904, "grad_norm": 121.00745391845703, "learning_rate": 4.216609589041097e-06, "loss": 0.0653, "step": 1351 }, { "epoch": 0.5787671232876712, "grad_norm": 10.502429008483887, "learning_rate": 4.212328767123288e-06, "loss": 0.0095, "step": 1352 }, { "epoch": 0.579195205479452, "grad_norm": 72.72311401367188, "learning_rate": 4.208047945205479e-06, "loss": 0.6172, "step": 1353 }, { "epoch": 0.5796232876712328, "grad_norm": 17.594356536865234, "learning_rate": 4.203767123287672e-06, "loss": 0.0206, "step": 1354 }, { "epoch": 0.5800513698630136, "grad_norm": 1.492584228515625, "learning_rate": 4.199486301369863e-06, "loss": 0.0007, "step": 1355 }, { "epoch": 0.5804794520547946, "grad_norm": 1.2296305894851685, "learning_rate": 4.195205479452055e-06, "loss": 0.002, "step": 1356 }, { "epoch": 0.5809075342465754, "grad_norm": 7.447392463684082, "learning_rate": 4.190924657534247e-06, "loss": 0.0042, "step": 1357 }, { "epoch": 0.5813356164383562, "grad_norm": 20.35088348388672, "learning_rate": 4.186643835616438e-06, "loss": 0.0248, "step": 1358 }, { "epoch": 0.581763698630137, "grad_norm": 6.785518169403076, "learning_rate": 4.182363013698631e-06, "loss": 0.0048, "step": 1359 }, { "epoch": 0.5821917808219178, "grad_norm": 8.12873649597168, "learning_rate": 4.178082191780822e-06, "loss": 0.0061, "step": 1360 }, { "epoch": 0.5826198630136986, "grad_norm": 2.2282629013061523, "learning_rate": 4.173801369863014e-06, "loss": 0.0008, "step": 1361 }, { "epoch": 0.5830479452054794, "grad_norm": 4.9799065589904785, "learning_rate": 4.1695205479452055e-06, "loss": 0.0031, "step": 1362 }, { "epoch": 0.5834760273972602, "grad_norm": 10.95131778717041, "learning_rate": 4.165239726027397e-06, "loss": 0.0079, "step": 1363 }, { "epoch": 0.583904109589041, "grad_norm": 33.217227935791016, "learning_rate": 4.16095890410959e-06, "loss": 0.0275, "step": 1364 }, { "epoch": 0.584332191780822, "grad_norm": 1.7718178033828735, "learning_rate": 4.156678082191781e-06, "loss": 0.0033, "step": 1365 }, { "epoch": 0.5847602739726028, "grad_norm": 5.8240766525268555, "learning_rate": 4.152397260273973e-06, "loss": 0.0038, "step": 1366 }, { "epoch": 0.5851883561643836, "grad_norm": 81.0331039428711, "learning_rate": 4.1481164383561645e-06, "loss": 0.1011, "step": 1367 }, { "epoch": 0.5856164383561644, "grad_norm": 1.4410375356674194, "learning_rate": 4.143835616438356e-06, "loss": 0.003, "step": 1368 }, { "epoch": 0.5860445205479452, "grad_norm": 5.092958450317383, "learning_rate": 4.139554794520549e-06, "loss": 0.0024, "step": 1369 }, { "epoch": 0.586472602739726, "grad_norm": 2.3007493019104004, "learning_rate": 4.13527397260274e-06, "loss": 0.0014, "step": 1370 }, { "epoch": 0.5869006849315068, "grad_norm": 3.8061540126800537, "learning_rate": 4.130993150684932e-06, "loss": 0.0025, "step": 1371 }, { "epoch": 0.5873287671232876, "grad_norm": 7.091177940368652, "learning_rate": 4.1267123287671235e-06, "loss": 0.0148, "step": 1372 }, { "epoch": 0.5877568493150684, "grad_norm": 5.648877143859863, "learning_rate": 4.122431506849315e-06, "loss": 0.0038, "step": 1373 }, { "epoch": 0.5881849315068494, "grad_norm": 2.2066330909729004, "learning_rate": 4.118150684931508e-06, "loss": 0.0013, "step": 1374 }, { "epoch": 0.5886130136986302, "grad_norm": 4.347330570220947, "learning_rate": 4.113869863013699e-06, "loss": 0.002, "step": 1375 }, { "epoch": 0.589041095890411, "grad_norm": 3.5065274238586426, "learning_rate": 4.109589041095891e-06, "loss": 0.0067, "step": 1376 }, { "epoch": 0.5894691780821918, "grad_norm": 130.03111267089844, "learning_rate": 4.1053082191780825e-06, "loss": 0.4311, "step": 1377 }, { "epoch": 0.5898972602739726, "grad_norm": 121.37728118896484, "learning_rate": 4.101027397260274e-06, "loss": 0.5584, "step": 1378 }, { "epoch": 0.5903253424657534, "grad_norm": 23.14792251586914, "learning_rate": 4.096746575342467e-06, "loss": 0.0397, "step": 1379 }, { "epoch": 0.5907534246575342, "grad_norm": 17.762248992919922, "learning_rate": 4.092465753424658e-06, "loss": 0.0172, "step": 1380 }, { "epoch": 0.591181506849315, "grad_norm": 2.031428098678589, "learning_rate": 4.08818493150685e-06, "loss": 0.0053, "step": 1381 }, { "epoch": 0.5916095890410958, "grad_norm": 82.16519165039062, "learning_rate": 4.0839041095890415e-06, "loss": 0.1115, "step": 1382 }, { "epoch": 0.5920376712328768, "grad_norm": 62.81968688964844, "learning_rate": 4.079623287671233e-06, "loss": 0.4742, "step": 1383 }, { "epoch": 0.5924657534246576, "grad_norm": 7.363317489624023, "learning_rate": 4.075342465753426e-06, "loss": 0.0063, "step": 1384 }, { "epoch": 0.5928938356164384, "grad_norm": 5.410426616668701, "learning_rate": 4.071061643835616e-06, "loss": 0.0055, "step": 1385 }, { "epoch": 0.5933219178082192, "grad_norm": 3.1887741088867188, "learning_rate": 4.066780821917808e-06, "loss": 0.0016, "step": 1386 }, { "epoch": 0.59375, "grad_norm": 7.374100685119629, "learning_rate": 4.0625000000000005e-06, "loss": 0.0042, "step": 1387 }, { "epoch": 0.5941780821917808, "grad_norm": 7.138071537017822, "learning_rate": 4.058219178082192e-06, "loss": 0.0041, "step": 1388 }, { "epoch": 0.5946061643835616, "grad_norm": 16.059730529785156, "learning_rate": 4.053938356164384e-06, "loss": 0.0184, "step": 1389 }, { "epoch": 0.5950342465753424, "grad_norm": 24.38505744934082, "learning_rate": 4.049657534246575e-06, "loss": 0.0231, "step": 1390 }, { "epoch": 0.5954623287671232, "grad_norm": 9.57630729675293, "learning_rate": 4.045376712328767e-06, "loss": 0.0069, "step": 1391 }, { "epoch": 0.5958904109589042, "grad_norm": 92.97953796386719, "learning_rate": 4.0410958904109595e-06, "loss": 0.7056, "step": 1392 }, { "epoch": 0.596318493150685, "grad_norm": 50.18244171142578, "learning_rate": 4.036815068493151e-06, "loss": 0.0616, "step": 1393 }, { "epoch": 0.5967465753424658, "grad_norm": 2.2438783645629883, "learning_rate": 4.032534246575343e-06, "loss": 0.005, "step": 1394 }, { "epoch": 0.5971746575342466, "grad_norm": 6.604089736938477, "learning_rate": 4.028253424657534e-06, "loss": 0.0039, "step": 1395 }, { "epoch": 0.5976027397260274, "grad_norm": 0.6127303242683411, "learning_rate": 4.023972602739726e-06, "loss": 0.0009, "step": 1396 }, { "epoch": 0.5980308219178082, "grad_norm": 3.7500455379486084, "learning_rate": 4.0196917808219185e-06, "loss": 0.0036, "step": 1397 }, { "epoch": 0.598458904109589, "grad_norm": 11.188542366027832, "learning_rate": 4.01541095890411e-06, "loss": 0.0096, "step": 1398 }, { "epoch": 0.5988869863013698, "grad_norm": 1.372507095336914, "learning_rate": 4.011130136986302e-06, "loss": 0.0023, "step": 1399 }, { "epoch": 0.5993150684931506, "grad_norm": 9.313090324401855, "learning_rate": 4.006849315068493e-06, "loss": 0.0102, "step": 1400 }, { "epoch": 0.5997431506849316, "grad_norm": 10.831459045410156, "learning_rate": 4.002568493150685e-06, "loss": 0.0109, "step": 1401 }, { "epoch": 0.6001712328767124, "grad_norm": 1.0780478715896606, "learning_rate": 3.9982876712328775e-06, "loss": 0.0013, "step": 1402 }, { "epoch": 0.6005993150684932, "grad_norm": 5.768706798553467, "learning_rate": 3.994006849315068e-06, "loss": 0.0169, "step": 1403 }, { "epoch": 0.601027397260274, "grad_norm": 6.806074619293213, "learning_rate": 3.989726027397261e-06, "loss": 0.004, "step": 1404 }, { "epoch": 0.6014554794520548, "grad_norm": 1.4164862632751465, "learning_rate": 3.985445205479452e-06, "loss": 0.0013, "step": 1405 }, { "epoch": 0.6018835616438356, "grad_norm": 2.985398769378662, "learning_rate": 3.981164383561644e-06, "loss": 0.002, "step": 1406 }, { "epoch": 0.6023116438356164, "grad_norm": 13.057939529418945, "learning_rate": 3.9768835616438365e-06, "loss": 0.009, "step": 1407 }, { "epoch": 0.6027397260273972, "grad_norm": 0.3426067531108856, "learning_rate": 3.972602739726027e-06, "loss": 0.0012, "step": 1408 }, { "epoch": 0.603167808219178, "grad_norm": 0.6996367573738098, "learning_rate": 3.96832191780822e-06, "loss": 0.0008, "step": 1409 }, { "epoch": 0.603595890410959, "grad_norm": 11.053820610046387, "learning_rate": 3.964041095890411e-06, "loss": 0.0118, "step": 1410 }, { "epoch": 0.6040239726027398, "grad_norm": 1.4751886129379272, "learning_rate": 3.959760273972603e-06, "loss": 0.0007, "step": 1411 }, { "epoch": 0.6044520547945206, "grad_norm": 11.62452507019043, "learning_rate": 3.9554794520547955e-06, "loss": 0.0115, "step": 1412 }, { "epoch": 0.6048801369863014, "grad_norm": 2.903002977371216, "learning_rate": 3.951198630136986e-06, "loss": 0.0025, "step": 1413 }, { "epoch": 0.6053082191780822, "grad_norm": 2.0436832904815674, "learning_rate": 3.946917808219179e-06, "loss": 0.0009, "step": 1414 }, { "epoch": 0.605736301369863, "grad_norm": 167.77008056640625, "learning_rate": 3.94263698630137e-06, "loss": 0.145, "step": 1415 }, { "epoch": 0.6061643835616438, "grad_norm": 10.575725555419922, "learning_rate": 3.938356164383562e-06, "loss": 0.0079, "step": 1416 }, { "epoch": 0.6065924657534246, "grad_norm": 4.829672336578369, "learning_rate": 3.934075342465754e-06, "loss": 0.002, "step": 1417 }, { "epoch": 0.6070205479452054, "grad_norm": 7.858760356903076, "learning_rate": 3.929794520547945e-06, "loss": 0.005, "step": 1418 }, { "epoch": 0.6074486301369864, "grad_norm": 1.3848947286605835, "learning_rate": 3.925513698630137e-06, "loss": 0.0026, "step": 1419 }, { "epoch": 0.6078767123287672, "grad_norm": 33.11077117919922, "learning_rate": 3.921232876712329e-06, "loss": 0.0127, "step": 1420 }, { "epoch": 0.608304794520548, "grad_norm": 7.512510299682617, "learning_rate": 3.916952054794521e-06, "loss": 0.0046, "step": 1421 }, { "epoch": 0.6087328767123288, "grad_norm": 340.4276123046875, "learning_rate": 3.912671232876713e-06, "loss": 0.6642, "step": 1422 }, { "epoch": 0.6091609589041096, "grad_norm": 25.19481658935547, "learning_rate": 3.908390410958904e-06, "loss": 0.0149, "step": 1423 }, { "epoch": 0.6095890410958904, "grad_norm": 176.07200622558594, "learning_rate": 3.904109589041096e-06, "loss": 0.1383, "step": 1424 }, { "epoch": 0.6100171232876712, "grad_norm": 124.2127914428711, "learning_rate": 3.899828767123288e-06, "loss": 0.1117, "step": 1425 }, { "epoch": 0.610445205479452, "grad_norm": 0.8382692933082581, "learning_rate": 3.895547945205479e-06, "loss": 0.0009, "step": 1426 }, { "epoch": 0.6108732876712328, "grad_norm": 1.420507550239563, "learning_rate": 3.891267123287672e-06, "loss": 0.001, "step": 1427 }, { "epoch": 0.6113013698630136, "grad_norm": 57.628108978271484, "learning_rate": 3.886986301369863e-06, "loss": 0.0405, "step": 1428 }, { "epoch": 0.6117294520547946, "grad_norm": 92.24188995361328, "learning_rate": 3.882705479452055e-06, "loss": 0.2292, "step": 1429 }, { "epoch": 0.6121575342465754, "grad_norm": 2.5576236248016357, "learning_rate": 3.878424657534247e-06, "loss": 0.0017, "step": 1430 }, { "epoch": 0.6125856164383562, "grad_norm": 5.143060684204102, "learning_rate": 3.874143835616438e-06, "loss": 0.0021, "step": 1431 }, { "epoch": 0.613013698630137, "grad_norm": 1.6757464408874512, "learning_rate": 3.869863013698631e-06, "loss": 0.0016, "step": 1432 }, { "epoch": 0.6134417808219178, "grad_norm": 4.580164909362793, "learning_rate": 3.865582191780822e-06, "loss": 0.0019, "step": 1433 }, { "epoch": 0.6138698630136986, "grad_norm": 0.9253978133201599, "learning_rate": 3.861301369863014e-06, "loss": 0.0004, "step": 1434 }, { "epoch": 0.6142979452054794, "grad_norm": 6.991001129150391, "learning_rate": 3.8570205479452055e-06, "loss": 0.0039, "step": 1435 }, { "epoch": 0.6147260273972602, "grad_norm": 63.37910842895508, "learning_rate": 3.852739726027397e-06, "loss": 0.818, "step": 1436 }, { "epoch": 0.615154109589041, "grad_norm": 0.9218167662620544, "learning_rate": 3.84845890410959e-06, "loss": 0.0008, "step": 1437 }, { "epoch": 0.615582191780822, "grad_norm": 4.461684703826904, "learning_rate": 3.844178082191781e-06, "loss": 0.0065, "step": 1438 }, { "epoch": 0.6160102739726028, "grad_norm": 122.59984588623047, "learning_rate": 3.839897260273973e-06, "loss": 0.4253, "step": 1439 }, { "epoch": 0.6164383561643836, "grad_norm": 0.2312237173318863, "learning_rate": 3.8356164383561645e-06, "loss": 0.0007, "step": 1440 }, { "epoch": 0.6168664383561644, "grad_norm": 206.74484252929688, "learning_rate": 3.831335616438356e-06, "loss": 0.2356, "step": 1441 }, { "epoch": 0.6172945205479452, "grad_norm": 8.116215705871582, "learning_rate": 3.827054794520549e-06, "loss": 0.0065, "step": 1442 }, { "epoch": 0.617722602739726, "grad_norm": 70.40342712402344, "learning_rate": 3.82277397260274e-06, "loss": 0.4737, "step": 1443 }, { "epoch": 0.6181506849315068, "grad_norm": 13.247394561767578, "learning_rate": 3.818493150684932e-06, "loss": 0.0123, "step": 1444 }, { "epoch": 0.6185787671232876, "grad_norm": 5.628781795501709, "learning_rate": 3.8142123287671235e-06, "loss": 0.0024, "step": 1445 }, { "epoch": 0.6190068493150684, "grad_norm": 32.239662170410156, "learning_rate": 3.809931506849315e-06, "loss": 0.034, "step": 1446 }, { "epoch": 0.6194349315068494, "grad_norm": 42.2091064453125, "learning_rate": 3.805650684931507e-06, "loss": 0.0656, "step": 1447 }, { "epoch": 0.6198630136986302, "grad_norm": 48.30079650878906, "learning_rate": 3.8013698630136993e-06, "loss": 0.0143, "step": 1448 }, { "epoch": 0.620291095890411, "grad_norm": 7.752137184143066, "learning_rate": 3.7970890410958905e-06, "loss": 0.0097, "step": 1449 }, { "epoch": 0.6207191780821918, "grad_norm": 4.709440231323242, "learning_rate": 3.7928082191780825e-06, "loss": 0.0022, "step": 1450 }, { "epoch": 0.6211472602739726, "grad_norm": 4.107389450073242, "learning_rate": 3.788527397260274e-06, "loss": 0.0024, "step": 1451 }, { "epoch": 0.6215753424657534, "grad_norm": 25.196767807006836, "learning_rate": 3.784246575342466e-06, "loss": 0.0118, "step": 1452 }, { "epoch": 0.6220034246575342, "grad_norm": 2.666196823120117, "learning_rate": 3.7799657534246583e-06, "loss": 0.0012, "step": 1453 }, { "epoch": 0.622431506849315, "grad_norm": 5.42006254196167, "learning_rate": 3.7756849315068495e-06, "loss": 0.0024, "step": 1454 }, { "epoch": 0.6228595890410958, "grad_norm": 30.45682716369629, "learning_rate": 3.7714041095890415e-06, "loss": 0.0358, "step": 1455 }, { "epoch": 0.6232876712328768, "grad_norm": 103.69232177734375, "learning_rate": 3.767123287671233e-06, "loss": 0.1003, "step": 1456 }, { "epoch": 0.6237157534246576, "grad_norm": 0.9902276396751404, "learning_rate": 3.762842465753425e-06, "loss": 0.0011, "step": 1457 }, { "epoch": 0.6241438356164384, "grad_norm": 8.185440063476562, "learning_rate": 3.7585616438356164e-06, "loss": 0.0058, "step": 1458 }, { "epoch": 0.6245719178082192, "grad_norm": 3.914591073989868, "learning_rate": 3.7542808219178085e-06, "loss": 0.0021, "step": 1459 }, { "epoch": 0.625, "grad_norm": 1.579645037651062, "learning_rate": 3.7500000000000005e-06, "loss": 0.0023, "step": 1460 }, { "epoch": 0.6254280821917808, "grad_norm": 3.3727524280548096, "learning_rate": 3.745719178082192e-06, "loss": 0.0016, "step": 1461 }, { "epoch": 0.6258561643835616, "grad_norm": 86.2633056640625, "learning_rate": 3.741438356164384e-06, "loss": 0.1625, "step": 1462 }, { "epoch": 0.6262842465753424, "grad_norm": 2.287116289138794, "learning_rate": 3.7371575342465754e-06, "loss": 0.0012, "step": 1463 }, { "epoch": 0.6267123287671232, "grad_norm": 84.17948913574219, "learning_rate": 3.7328767123287675e-06, "loss": 0.0671, "step": 1464 }, { "epoch": 0.6271404109589042, "grad_norm": 97.2966537475586, "learning_rate": 3.728595890410959e-06, "loss": 0.4906, "step": 1465 }, { "epoch": 0.627568493150685, "grad_norm": 3.4898667335510254, "learning_rate": 3.724315068493151e-06, "loss": 0.0032, "step": 1466 }, { "epoch": 0.6279965753424658, "grad_norm": 15.297725677490234, "learning_rate": 3.7200342465753424e-06, "loss": 0.0182, "step": 1467 }, { "epoch": 0.6284246575342466, "grad_norm": 2.224381446838379, "learning_rate": 3.7157534246575344e-06, "loss": 0.004, "step": 1468 }, { "epoch": 0.6288527397260274, "grad_norm": 55.43446731567383, "learning_rate": 3.7114726027397265e-06, "loss": 0.0671, "step": 1469 }, { "epoch": 0.6292808219178082, "grad_norm": 29.2762393951416, "learning_rate": 3.707191780821918e-06, "loss": 0.0272, "step": 1470 }, { "epoch": 0.629708904109589, "grad_norm": 1.3637113571166992, "learning_rate": 3.70291095890411e-06, "loss": 0.0017, "step": 1471 }, { "epoch": 0.6301369863013698, "grad_norm": 5.948508262634277, "learning_rate": 3.6986301369863014e-06, "loss": 0.0034, "step": 1472 }, { "epoch": 0.6305650684931506, "grad_norm": 32.135799407958984, "learning_rate": 3.6943493150684934e-06, "loss": 0.033, "step": 1473 }, { "epoch": 0.6309931506849316, "grad_norm": 2.265585422515869, "learning_rate": 3.6900684931506855e-06, "loss": 0.0033, "step": 1474 }, { "epoch": 0.6314212328767124, "grad_norm": 8.708675384521484, "learning_rate": 3.685787671232877e-06, "loss": 0.0079, "step": 1475 }, { "epoch": 0.6318493150684932, "grad_norm": 225.29019165039062, "learning_rate": 3.6815068493150687e-06, "loss": 0.1887, "step": 1476 }, { "epoch": 0.632277397260274, "grad_norm": 3.3073229789733887, "learning_rate": 3.6772260273972604e-06, "loss": 0.0018, "step": 1477 }, { "epoch": 0.6327054794520548, "grad_norm": 5.2108659744262695, "learning_rate": 3.6729452054794524e-06, "loss": 0.0018, "step": 1478 }, { "epoch": 0.6331335616438356, "grad_norm": 11.371235847473145, "learning_rate": 3.668664383561644e-06, "loss": 0.0104, "step": 1479 }, { "epoch": 0.6335616438356164, "grad_norm": 4.906203269958496, "learning_rate": 3.664383561643836e-06, "loss": 0.0026, "step": 1480 }, { "epoch": 0.6339897260273972, "grad_norm": 4.929328918457031, "learning_rate": 3.6601027397260273e-06, "loss": 0.0032, "step": 1481 }, { "epoch": 0.634417808219178, "grad_norm": 9.273417472839355, "learning_rate": 3.6558219178082194e-06, "loss": 0.0061, "step": 1482 }, { "epoch": 0.634845890410959, "grad_norm": 16.293359756469727, "learning_rate": 3.6515410958904114e-06, "loss": 0.0151, "step": 1483 }, { "epoch": 0.6352739726027398, "grad_norm": 44.26121520996094, "learning_rate": 3.647260273972603e-06, "loss": 0.089, "step": 1484 }, { "epoch": 0.6357020547945206, "grad_norm": 4.846875190734863, "learning_rate": 3.642979452054795e-06, "loss": 0.0032, "step": 1485 }, { "epoch": 0.6361301369863014, "grad_norm": 0.8606888651847839, "learning_rate": 3.6386986301369863e-06, "loss": 0.0008, "step": 1486 }, { "epoch": 0.6365582191780822, "grad_norm": 8.249634742736816, "learning_rate": 3.6344178082191783e-06, "loss": 0.0106, "step": 1487 }, { "epoch": 0.636986301369863, "grad_norm": 1.3124595880508423, "learning_rate": 3.6301369863013704e-06, "loss": 0.0025, "step": 1488 }, { "epoch": 0.6374143835616438, "grad_norm": 0.3741726577281952, "learning_rate": 3.625856164383562e-06, "loss": 0.0008, "step": 1489 }, { "epoch": 0.6378424657534246, "grad_norm": 82.41120147705078, "learning_rate": 3.6215753424657537e-06, "loss": 0.1663, "step": 1490 }, { "epoch": 0.6382705479452054, "grad_norm": 213.1150360107422, "learning_rate": 3.6172945205479453e-06, "loss": 0.5928, "step": 1491 }, { "epoch": 0.6386986301369864, "grad_norm": 33.115882873535156, "learning_rate": 3.6130136986301373e-06, "loss": 0.0408, "step": 1492 }, { "epoch": 0.6391267123287672, "grad_norm": 19.094566345214844, "learning_rate": 3.6087328767123294e-06, "loss": 0.0155, "step": 1493 }, { "epoch": 0.639554794520548, "grad_norm": 117.77619934082031, "learning_rate": 3.604452054794521e-06, "loss": 0.1533, "step": 1494 }, { "epoch": 0.6399828767123288, "grad_norm": 4.553089141845703, "learning_rate": 3.6001712328767122e-06, "loss": 0.0033, "step": 1495 }, { "epoch": 0.6404109589041096, "grad_norm": 1.7492064237594604, "learning_rate": 3.5958904109589043e-06, "loss": 0.0008, "step": 1496 }, { "epoch": 0.6408390410958904, "grad_norm": 79.98994445800781, "learning_rate": 3.5916095890410963e-06, "loss": 0.0623, "step": 1497 }, { "epoch": 0.6412671232876712, "grad_norm": 2.3220348358154297, "learning_rate": 3.587328767123288e-06, "loss": 0.0013, "step": 1498 }, { "epoch": 0.641695205479452, "grad_norm": 1.0320122241973877, "learning_rate": 3.5830479452054796e-06, "loss": 0.0012, "step": 1499 }, { "epoch": 0.6421232876712328, "grad_norm": 6.743963241577148, "learning_rate": 3.5787671232876712e-06, "loss": 0.0048, "step": 1500 }, { "epoch": 0.6425513698630136, "grad_norm": 10.868728637695312, "learning_rate": 3.5744863013698633e-06, "loss": 0.0089, "step": 1501 }, { "epoch": 0.6429794520547946, "grad_norm": 1.9495466947555542, "learning_rate": 3.5702054794520553e-06, "loss": 0.0021, "step": 1502 }, { "epoch": 0.6434075342465754, "grad_norm": 6.411136150360107, "learning_rate": 3.565924657534247e-06, "loss": 0.004, "step": 1503 }, { "epoch": 0.6438356164383562, "grad_norm": 7.954117298126221, "learning_rate": 3.5616438356164386e-06, "loss": 0.0051, "step": 1504 }, { "epoch": 0.644263698630137, "grad_norm": 5.296102523803711, "learning_rate": 3.5573630136986302e-06, "loss": 0.0042, "step": 1505 }, { "epoch": 0.6446917808219178, "grad_norm": 7.186568260192871, "learning_rate": 3.5530821917808223e-06, "loss": 0.0043, "step": 1506 }, { "epoch": 0.6451198630136986, "grad_norm": 1.574050784111023, "learning_rate": 3.5488013698630143e-06, "loss": 0.0021, "step": 1507 }, { "epoch": 0.6455479452054794, "grad_norm": 7.080282211303711, "learning_rate": 3.5445205479452056e-06, "loss": 0.0039, "step": 1508 }, { "epoch": 0.6459760273972602, "grad_norm": 3.806734800338745, "learning_rate": 3.5402397260273976e-06, "loss": 0.003, "step": 1509 }, { "epoch": 0.646404109589041, "grad_norm": 7.583524227142334, "learning_rate": 3.5359589041095892e-06, "loss": 0.0034, "step": 1510 }, { "epoch": 0.646832191780822, "grad_norm": 19.204914093017578, "learning_rate": 3.5316780821917813e-06, "loss": 0.0089, "step": 1511 }, { "epoch": 0.6472602739726028, "grad_norm": 6.462817668914795, "learning_rate": 3.527397260273973e-06, "loss": 0.0036, "step": 1512 }, { "epoch": 0.6476883561643836, "grad_norm": 25.578567504882812, "learning_rate": 3.5231164383561646e-06, "loss": 0.01, "step": 1513 }, { "epoch": 0.6481164383561644, "grad_norm": 2.3506271839141846, "learning_rate": 3.518835616438356e-06, "loss": 0.0031, "step": 1514 }, { "epoch": 0.6485445205479452, "grad_norm": 176.80567932128906, "learning_rate": 3.5145547945205482e-06, "loss": 0.3172, "step": 1515 }, { "epoch": 0.648972602739726, "grad_norm": 62.4288330078125, "learning_rate": 3.5102739726027403e-06, "loss": 0.619, "step": 1516 }, { "epoch": 0.6494006849315068, "grad_norm": 2.850057363510132, "learning_rate": 3.505993150684932e-06, "loss": 0.0008, "step": 1517 }, { "epoch": 0.6498287671232876, "grad_norm": 3.133211374282837, "learning_rate": 3.5017123287671235e-06, "loss": 0.0016, "step": 1518 }, { "epoch": 0.6502568493150684, "grad_norm": 6.875095367431641, "learning_rate": 3.497431506849315e-06, "loss": 0.0065, "step": 1519 }, { "epoch": 0.6506849315068494, "grad_norm": 27.46495819091797, "learning_rate": 3.4931506849315072e-06, "loss": 0.0197, "step": 1520 }, { "epoch": 0.6511130136986302, "grad_norm": 4.435451030731201, "learning_rate": 3.4888698630136993e-06, "loss": 0.0031, "step": 1521 }, { "epoch": 0.651541095890411, "grad_norm": 4.3725361824035645, "learning_rate": 3.4845890410958905e-06, "loss": 0.0017, "step": 1522 }, { "epoch": 0.6519691780821918, "grad_norm": 10.563131332397461, "learning_rate": 3.4803082191780825e-06, "loss": 0.0064, "step": 1523 }, { "epoch": 0.6523972602739726, "grad_norm": 4.029549598693848, "learning_rate": 3.476027397260274e-06, "loss": 0.002, "step": 1524 }, { "epoch": 0.6528253424657534, "grad_norm": 0.5828776955604553, "learning_rate": 3.4717465753424662e-06, "loss": 0.0005, "step": 1525 }, { "epoch": 0.6532534246575342, "grad_norm": 4.144784450531006, "learning_rate": 3.4674657534246583e-06, "loss": 0.0042, "step": 1526 }, { "epoch": 0.653681506849315, "grad_norm": 3.2127463817596436, "learning_rate": 3.4631849315068495e-06, "loss": 0.0015, "step": 1527 }, { "epoch": 0.6541095890410958, "grad_norm": 1.0160362720489502, "learning_rate": 3.458904109589041e-06, "loss": 0.0008, "step": 1528 }, { "epoch": 0.6545376712328768, "grad_norm": 7.308888912200928, "learning_rate": 3.454623287671233e-06, "loss": 0.0041, "step": 1529 }, { "epoch": 0.6549657534246576, "grad_norm": 11.140643119812012, "learning_rate": 3.4503424657534252e-06, "loss": 0.0049, "step": 1530 }, { "epoch": 0.6553938356164384, "grad_norm": 1.364044427871704, "learning_rate": 3.4460616438356164e-06, "loss": 0.0018, "step": 1531 }, { "epoch": 0.6558219178082192, "grad_norm": 5.661051273345947, "learning_rate": 3.4417808219178085e-06, "loss": 0.0035, "step": 1532 }, { "epoch": 0.65625, "grad_norm": 2.0294559001922607, "learning_rate": 3.4375e-06, "loss": 0.0008, "step": 1533 }, { "epoch": 0.6566780821917808, "grad_norm": 3.766801118850708, "learning_rate": 3.433219178082192e-06, "loss": 0.0019, "step": 1534 }, { "epoch": 0.6571061643835616, "grad_norm": 1.559596300125122, "learning_rate": 3.4289383561643842e-06, "loss": 0.0019, "step": 1535 }, { "epoch": 0.6575342465753424, "grad_norm": 25.868913650512695, "learning_rate": 3.4246575342465754e-06, "loss": 0.0176, "step": 1536 }, { "epoch": 0.6579623287671232, "grad_norm": 4.886362075805664, "learning_rate": 3.4203767123287675e-06, "loss": 0.0022, "step": 1537 }, { "epoch": 0.6583904109589042, "grad_norm": 1.2223601341247559, "learning_rate": 3.416095890410959e-06, "loss": 0.0007, "step": 1538 }, { "epoch": 0.658818493150685, "grad_norm": 197.3466339111328, "learning_rate": 3.411815068493151e-06, "loss": 0.3608, "step": 1539 }, { "epoch": 0.6592465753424658, "grad_norm": 5.495297431945801, "learning_rate": 3.4075342465753424e-06, "loss": 0.0024, "step": 1540 }, { "epoch": 0.6596746575342466, "grad_norm": 5.9631524085998535, "learning_rate": 3.4032534246575344e-06, "loss": 0.0029, "step": 1541 }, { "epoch": 0.6601027397260274, "grad_norm": 1.5551867485046387, "learning_rate": 3.3989726027397265e-06, "loss": 0.0018, "step": 1542 }, { "epoch": 0.6605308219178082, "grad_norm": 52.53830337524414, "learning_rate": 3.394691780821918e-06, "loss": 0.5439, "step": 1543 }, { "epoch": 0.660958904109589, "grad_norm": 1.1564918756484985, "learning_rate": 3.39041095890411e-06, "loss": 0.0015, "step": 1544 }, { "epoch": 0.6613869863013698, "grad_norm": 1.7196059226989746, "learning_rate": 3.3861301369863014e-06, "loss": 0.0008, "step": 1545 }, { "epoch": 0.6618150684931506, "grad_norm": 5.593273162841797, "learning_rate": 3.3818493150684934e-06, "loss": 0.0033, "step": 1546 }, { "epoch": 0.6622431506849316, "grad_norm": 3.1256916522979736, "learning_rate": 3.377568493150685e-06, "loss": 0.0009, "step": 1547 }, { "epoch": 0.6626712328767124, "grad_norm": 4.145781993865967, "learning_rate": 3.373287671232877e-06, "loss": 0.0026, "step": 1548 }, { "epoch": 0.6630993150684932, "grad_norm": 2.788656234741211, "learning_rate": 3.3690068493150683e-06, "loss": 0.0015, "step": 1549 }, { "epoch": 0.663527397260274, "grad_norm": 0.764890193939209, "learning_rate": 3.3647260273972604e-06, "loss": 0.0004, "step": 1550 }, { "epoch": 0.6639554794520548, "grad_norm": 0.3082369565963745, "learning_rate": 3.3604452054794524e-06, "loss": 0.0005, "step": 1551 }, { "epoch": 0.6643835616438356, "grad_norm": 17.246828079223633, "learning_rate": 3.356164383561644e-06, "loss": 0.0183, "step": 1552 }, { "epoch": 0.6648116438356164, "grad_norm": 4.747049808502197, "learning_rate": 3.351883561643836e-06, "loss": 0.0037, "step": 1553 }, { "epoch": 0.6652397260273972, "grad_norm": 2.6844513416290283, "learning_rate": 3.3476027397260273e-06, "loss": 0.001, "step": 1554 }, { "epoch": 0.665667808219178, "grad_norm": 52.50829315185547, "learning_rate": 3.3433219178082194e-06, "loss": 0.0322, "step": 1555 }, { "epoch": 0.666095890410959, "grad_norm": 2.903136968612671, "learning_rate": 3.3390410958904114e-06, "loss": 0.002, "step": 1556 }, { "epoch": 0.6665239726027398, "grad_norm": 12.775906562805176, "learning_rate": 3.334760273972603e-06, "loss": 0.0079, "step": 1557 }, { "epoch": 0.6669520547945206, "grad_norm": 19.215044021606445, "learning_rate": 3.330479452054795e-06, "loss": 0.0198, "step": 1558 }, { "epoch": 0.6673801369863014, "grad_norm": 0.17277565598487854, "learning_rate": 3.3261986301369863e-06, "loss": 0.0006, "step": 1559 }, { "epoch": 0.6678082191780822, "grad_norm": 10.719157218933105, "learning_rate": 3.3219178082191784e-06, "loss": 0.0041, "step": 1560 }, { "epoch": 0.668236301369863, "grad_norm": 2.619201421737671, "learning_rate": 3.31763698630137e-06, "loss": 0.0016, "step": 1561 }, { "epoch": 0.6686643835616438, "grad_norm": 81.49394226074219, "learning_rate": 3.313356164383562e-06, "loss": 0.0373, "step": 1562 }, { "epoch": 0.6690924657534246, "grad_norm": 4.3130784034729, "learning_rate": 3.3090753424657533e-06, "loss": 0.0049, "step": 1563 }, { "epoch": 0.6695205479452054, "grad_norm": 5.828455924987793, "learning_rate": 3.3047945205479453e-06, "loss": 0.0033, "step": 1564 }, { "epoch": 0.6699486301369864, "grad_norm": 1.2845667600631714, "learning_rate": 3.3005136986301374e-06, "loss": 0.0012, "step": 1565 }, { "epoch": 0.6703767123287672, "grad_norm": 3.5588693618774414, "learning_rate": 3.296232876712329e-06, "loss": 0.0009, "step": 1566 }, { "epoch": 0.670804794520548, "grad_norm": 2.6474616527557373, "learning_rate": 3.291952054794521e-06, "loss": 0.0009, "step": 1567 }, { "epoch": 0.6712328767123288, "grad_norm": 8.840327262878418, "learning_rate": 3.2876712328767123e-06, "loss": 0.0052, "step": 1568 }, { "epoch": 0.6716609589041096, "grad_norm": 6.170435905456543, "learning_rate": 3.2833904109589043e-06, "loss": 0.007, "step": 1569 }, { "epoch": 0.6720890410958904, "grad_norm": 8.549419403076172, "learning_rate": 3.2791095890410964e-06, "loss": 0.008, "step": 1570 }, { "epoch": 0.6725171232876712, "grad_norm": 4.0339226722717285, "learning_rate": 3.274828767123288e-06, "loss": 0.0032, "step": 1571 }, { "epoch": 0.672945205479452, "grad_norm": 4.795053482055664, "learning_rate": 3.2705479452054796e-06, "loss": 0.002, "step": 1572 }, { "epoch": 0.6733732876712328, "grad_norm": 15.693790435791016, "learning_rate": 3.2662671232876713e-06, "loss": 0.0373, "step": 1573 }, { "epoch": 0.6738013698630136, "grad_norm": 4.3212151527404785, "learning_rate": 3.2619863013698633e-06, "loss": 0.0016, "step": 1574 }, { "epoch": 0.6742294520547946, "grad_norm": 25.472753524780273, "learning_rate": 3.2577054794520554e-06, "loss": 0.0248, "step": 1575 }, { "epoch": 0.6746575342465754, "grad_norm": 0.5839943289756775, "learning_rate": 3.253424657534247e-06, "loss": 0.0009, "step": 1576 }, { "epoch": 0.6750856164383562, "grad_norm": 56.60502624511719, "learning_rate": 3.2491438356164386e-06, "loss": 0.039, "step": 1577 }, { "epoch": 0.675513698630137, "grad_norm": 9.264687538146973, "learning_rate": 3.2448630136986303e-06, "loss": 0.0076, "step": 1578 }, { "epoch": 0.6759417808219178, "grad_norm": 56.452980041503906, "learning_rate": 3.2405821917808223e-06, "loss": 0.049, "step": 1579 }, { "epoch": 0.6763698630136986, "grad_norm": 18.34209632873535, "learning_rate": 3.236301369863014e-06, "loss": 0.0239, "step": 1580 }, { "epoch": 0.6767979452054794, "grad_norm": 7.600927829742432, "learning_rate": 3.2320205479452056e-06, "loss": 0.0049, "step": 1581 }, { "epoch": 0.6772260273972602, "grad_norm": 6.149591445922852, "learning_rate": 3.227739726027397e-06, "loss": 0.0034, "step": 1582 }, { "epoch": 0.677654109589041, "grad_norm": 5.238715648651123, "learning_rate": 3.2234589041095893e-06, "loss": 0.002, "step": 1583 }, { "epoch": 0.678082191780822, "grad_norm": 5.473127841949463, "learning_rate": 3.2191780821917813e-06, "loss": 0.0022, "step": 1584 }, { "epoch": 0.6785102739726028, "grad_norm": 2.882988929748535, "learning_rate": 3.214897260273973e-06, "loss": 0.0033, "step": 1585 }, { "epoch": 0.6789383561643836, "grad_norm": 1.3625125885009766, "learning_rate": 3.2106164383561646e-06, "loss": 0.0004, "step": 1586 }, { "epoch": 0.6793664383561644, "grad_norm": 1.986010193824768, "learning_rate": 3.206335616438356e-06, "loss": 0.0021, "step": 1587 }, { "epoch": 0.6797945205479452, "grad_norm": 338.2096252441406, "learning_rate": 3.2020547945205483e-06, "loss": 0.0949, "step": 1588 }, { "epoch": 0.680222602739726, "grad_norm": 12.366413116455078, "learning_rate": 3.1977739726027403e-06, "loss": 0.0104, "step": 1589 }, { "epoch": 0.6806506849315068, "grad_norm": 15.218355178833008, "learning_rate": 3.193493150684932e-06, "loss": 0.0135, "step": 1590 }, { "epoch": 0.6810787671232876, "grad_norm": 14.947218894958496, "learning_rate": 3.1892123287671236e-06, "loss": 0.018, "step": 1591 }, { "epoch": 0.6815068493150684, "grad_norm": 16.04259490966797, "learning_rate": 3.184931506849315e-06, "loss": 0.0184, "step": 1592 }, { "epoch": 0.6819349315068494, "grad_norm": 6.878523826599121, "learning_rate": 3.1806506849315073e-06, "loss": 0.004, "step": 1593 }, { "epoch": 0.6823630136986302, "grad_norm": 7.495099067687988, "learning_rate": 3.176369863013699e-06, "loss": 0.0042, "step": 1594 }, { "epoch": 0.682791095890411, "grad_norm": 9.372208595275879, "learning_rate": 3.1720890410958905e-06, "loss": 0.0081, "step": 1595 }, { "epoch": 0.6832191780821918, "grad_norm": 3.321232795715332, "learning_rate": 3.167808219178082e-06, "loss": 0.0011, "step": 1596 }, { "epoch": 0.6836472602739726, "grad_norm": 1.0083523988723755, "learning_rate": 3.163527397260274e-06, "loss": 0.0005, "step": 1597 }, { "epoch": 0.6840753424657534, "grad_norm": 1.2982144355773926, "learning_rate": 3.1592465753424663e-06, "loss": 0.0011, "step": 1598 }, { "epoch": 0.6845034246575342, "grad_norm": 11.464179992675781, "learning_rate": 3.154965753424658e-06, "loss": 0.0058, "step": 1599 }, { "epoch": 0.684931506849315, "grad_norm": 132.45025634765625, "learning_rate": 3.1506849315068495e-06, "loss": 0.126, "step": 1600 }, { "epoch": 0.6853595890410958, "grad_norm": 176.4072723388672, "learning_rate": 3.146404109589041e-06, "loss": 0.0547, "step": 1601 }, { "epoch": 0.6857876712328768, "grad_norm": 15.363570213317871, "learning_rate": 3.142123287671233e-06, "loss": 0.0167, "step": 1602 }, { "epoch": 0.6862157534246576, "grad_norm": 14.597602844238281, "learning_rate": 3.1378424657534253e-06, "loss": 0.0156, "step": 1603 }, { "epoch": 0.6866438356164384, "grad_norm": 11.271079063415527, "learning_rate": 3.1335616438356165e-06, "loss": 0.0099, "step": 1604 }, { "epoch": 0.6870719178082192, "grad_norm": 43.16969299316406, "learning_rate": 3.1292808219178085e-06, "loss": 0.0259, "step": 1605 }, { "epoch": 0.6875, "grad_norm": 98.5142822265625, "learning_rate": 3.125e-06, "loss": 0.5608, "step": 1606 }, { "epoch": 0.6879280821917808, "grad_norm": 9.178690910339355, "learning_rate": 3.120719178082192e-06, "loss": 0.0068, "step": 1607 }, { "epoch": 0.6883561643835616, "grad_norm": 6.666224002838135, "learning_rate": 3.1164383561643843e-06, "loss": 0.0038, "step": 1608 }, { "epoch": 0.6887842465753424, "grad_norm": 6.398871898651123, "learning_rate": 3.1121575342465755e-06, "loss": 0.0038, "step": 1609 }, { "epoch": 0.6892123287671232, "grad_norm": 7.254252910614014, "learning_rate": 3.1078767123287675e-06, "loss": 0.008, "step": 1610 }, { "epoch": 0.6896404109589042, "grad_norm": 84.06938934326172, "learning_rate": 3.103595890410959e-06, "loss": 0.0385, "step": 1611 }, { "epoch": 0.690068493150685, "grad_norm": 9.468402862548828, "learning_rate": 3.099315068493151e-06, "loss": 0.0065, "step": 1612 }, { "epoch": 0.6904965753424658, "grad_norm": 9.934799194335938, "learning_rate": 3.0950342465753424e-06, "loss": 0.0072, "step": 1613 }, { "epoch": 0.6909246575342466, "grad_norm": 9.065140724182129, "learning_rate": 3.0907534246575345e-06, "loss": 0.0059, "step": 1614 }, { "epoch": 0.6913527397260274, "grad_norm": 2.240502119064331, "learning_rate": 3.086472602739726e-06, "loss": 0.0009, "step": 1615 }, { "epoch": 0.6917808219178082, "grad_norm": 10.559172630310059, "learning_rate": 3.082191780821918e-06, "loss": 0.0097, "step": 1616 }, { "epoch": 0.692208904109589, "grad_norm": 2.0675525665283203, "learning_rate": 3.07791095890411e-06, "loss": 0.001, "step": 1617 }, { "epoch": 0.6926369863013698, "grad_norm": 1.496484637260437, "learning_rate": 3.0736301369863014e-06, "loss": 0.0021, "step": 1618 }, { "epoch": 0.6930650684931506, "grad_norm": 14.082022666931152, "learning_rate": 3.0693493150684935e-06, "loss": 0.0083, "step": 1619 }, { "epoch": 0.6934931506849316, "grad_norm": 0.9320625066757202, "learning_rate": 3.065068493150685e-06, "loss": 0.0005, "step": 1620 }, { "epoch": 0.6939212328767124, "grad_norm": 12.455010414123535, "learning_rate": 3.060787671232877e-06, "loss": 0.0145, "step": 1621 }, { "epoch": 0.6943493150684932, "grad_norm": 5.520180702209473, "learning_rate": 3.0565068493150684e-06, "loss": 0.0045, "step": 1622 }, { "epoch": 0.694777397260274, "grad_norm": 70.74891662597656, "learning_rate": 3.0522260273972604e-06, "loss": 0.1473, "step": 1623 }, { "epoch": 0.6952054794520548, "grad_norm": 5.151222229003906, "learning_rate": 3.0479452054794525e-06, "loss": 0.0031, "step": 1624 }, { "epoch": 0.6956335616438356, "grad_norm": 8.5337553024292, "learning_rate": 3.043664383561644e-06, "loss": 0.0073, "step": 1625 }, { "epoch": 0.6960616438356164, "grad_norm": 1.449599027633667, "learning_rate": 3.039383561643836e-06, "loss": 0.0007, "step": 1626 }, { "epoch": 0.6964897260273972, "grad_norm": 2.3516628742218018, "learning_rate": 3.0351027397260274e-06, "loss": 0.0009, "step": 1627 }, { "epoch": 0.696917808219178, "grad_norm": 143.76602172851562, "learning_rate": 3.0308219178082194e-06, "loss": 0.34, "step": 1628 }, { "epoch": 0.697345890410959, "grad_norm": 4.010818958282471, "learning_rate": 3.026541095890411e-06, "loss": 0.0028, "step": 1629 }, { "epoch": 0.6977739726027398, "grad_norm": 4.487422943115234, "learning_rate": 3.022260273972603e-06, "loss": 0.0015, "step": 1630 }, { "epoch": 0.6982020547945206, "grad_norm": 9.818235397338867, "learning_rate": 3.017979452054795e-06, "loss": 0.0062, "step": 1631 }, { "epoch": 0.6986301369863014, "grad_norm": 0.9299815893173218, "learning_rate": 3.0136986301369864e-06, "loss": 0.0004, "step": 1632 }, { "epoch": 0.6990582191780822, "grad_norm": 4.448871612548828, "learning_rate": 3.0094178082191784e-06, "loss": 0.0037, "step": 1633 }, { "epoch": 0.699486301369863, "grad_norm": 4.714613437652588, "learning_rate": 3.00513698630137e-06, "loss": 0.0032, "step": 1634 }, { "epoch": 0.6999143835616438, "grad_norm": 5.753317832946777, "learning_rate": 3.000856164383562e-06, "loss": 0.003, "step": 1635 }, { "epoch": 0.7003424657534246, "grad_norm": 2.8652102947235107, "learning_rate": 2.9965753424657533e-06, "loss": 0.0055, "step": 1636 }, { "epoch": 0.7007705479452054, "grad_norm": 1.8467093706130981, "learning_rate": 2.9922945205479453e-06, "loss": 0.0006, "step": 1637 }, { "epoch": 0.7011986301369864, "grad_norm": 2.0887656211853027, "learning_rate": 2.9880136986301374e-06, "loss": 0.001, "step": 1638 }, { "epoch": 0.7016267123287672, "grad_norm": 7.080411434173584, "learning_rate": 2.983732876712329e-06, "loss": 0.0055, "step": 1639 }, { "epoch": 0.702054794520548, "grad_norm": 4.441183567047119, "learning_rate": 2.979452054794521e-06, "loss": 0.0019, "step": 1640 }, { "epoch": 0.7024828767123288, "grad_norm": 8.39742660522461, "learning_rate": 2.9751712328767123e-06, "loss": 0.0059, "step": 1641 }, { "epoch": 0.7029109589041096, "grad_norm": 10.298524856567383, "learning_rate": 2.9708904109589043e-06, "loss": 0.0086, "step": 1642 }, { "epoch": 0.7033390410958904, "grad_norm": 53.1323127746582, "learning_rate": 2.9666095890410964e-06, "loss": 0.0587, "step": 1643 }, { "epoch": 0.7037671232876712, "grad_norm": 13.39601993560791, "learning_rate": 2.962328767123288e-06, "loss": 0.0152, "step": 1644 }, { "epoch": 0.704195205479452, "grad_norm": 5.144074440002441, "learning_rate": 2.9580479452054792e-06, "loss": 0.0027, "step": 1645 }, { "epoch": 0.7046232876712328, "grad_norm": 2.237705707550049, "learning_rate": 2.9537671232876713e-06, "loss": 0.0015, "step": 1646 }, { "epoch": 0.7050513698630136, "grad_norm": 1.2122917175292969, "learning_rate": 2.9494863013698633e-06, "loss": 0.0008, "step": 1647 }, { "epoch": 0.7054794520547946, "grad_norm": 82.07071685791016, "learning_rate": 2.945205479452055e-06, "loss": 0.193, "step": 1648 }, { "epoch": 0.7059075342465754, "grad_norm": 4.0096893310546875, "learning_rate": 2.940924657534247e-06, "loss": 0.0026, "step": 1649 }, { "epoch": 0.7063356164383562, "grad_norm": 1.5330088138580322, "learning_rate": 2.9366438356164382e-06, "loss": 0.0005, "step": 1650 }, { "epoch": 0.706763698630137, "grad_norm": 2.6511662006378174, "learning_rate": 2.9323630136986303e-06, "loss": 0.0009, "step": 1651 }, { "epoch": 0.7071917808219178, "grad_norm": 1.2386120557785034, "learning_rate": 2.9280821917808223e-06, "loss": 0.0009, "step": 1652 }, { "epoch": 0.7076198630136986, "grad_norm": 35.590579986572266, "learning_rate": 2.923801369863014e-06, "loss": 0.0261, "step": 1653 }, { "epoch": 0.7080479452054794, "grad_norm": 180.27830505371094, "learning_rate": 2.9195205479452056e-06, "loss": 0.0625, "step": 1654 }, { "epoch": 0.7084760273972602, "grad_norm": 104.61904907226562, "learning_rate": 2.9152397260273972e-06, "loss": 0.8986, "step": 1655 }, { "epoch": 0.708904109589041, "grad_norm": 2.5711395740509033, "learning_rate": 2.9109589041095893e-06, "loss": 0.0007, "step": 1656 }, { "epoch": 0.709332191780822, "grad_norm": 2.6367461681365967, "learning_rate": 2.9066780821917813e-06, "loss": 0.0022, "step": 1657 }, { "epoch": 0.7097602739726028, "grad_norm": 8.023321151733398, "learning_rate": 2.902397260273973e-06, "loss": 0.0052, "step": 1658 }, { "epoch": 0.7101883561643836, "grad_norm": 7.646404266357422, "learning_rate": 2.8981164383561646e-06, "loss": 0.0044, "step": 1659 }, { "epoch": 0.7106164383561644, "grad_norm": 160.6707305908203, "learning_rate": 2.8938356164383562e-06, "loss": 0.1236, "step": 1660 }, { "epoch": 0.7110445205479452, "grad_norm": 1.7002663612365723, "learning_rate": 2.8895547945205483e-06, "loss": 0.0005, "step": 1661 }, { "epoch": 0.711472602739726, "grad_norm": 3.484675407409668, "learning_rate": 2.88527397260274e-06, "loss": 0.0015, "step": 1662 }, { "epoch": 0.7119006849315068, "grad_norm": 88.17328643798828, "learning_rate": 2.880993150684932e-06, "loss": 0.6281, "step": 1663 }, { "epoch": 0.7123287671232876, "grad_norm": 199.0227813720703, "learning_rate": 2.876712328767123e-06, "loss": 0.3224, "step": 1664 }, { "epoch": 0.7127568493150684, "grad_norm": 0.6613712906837463, "learning_rate": 2.8724315068493152e-06, "loss": 0.0004, "step": 1665 }, { "epoch": 0.7131849315068494, "grad_norm": 3.463466167449951, "learning_rate": 2.8681506849315073e-06, "loss": 0.0013, "step": 1666 }, { "epoch": 0.7136130136986302, "grad_norm": 5.259455680847168, "learning_rate": 2.863869863013699e-06, "loss": 0.0033, "step": 1667 }, { "epoch": 0.714041095890411, "grad_norm": 8.865537643432617, "learning_rate": 2.8595890410958905e-06, "loss": 0.007, "step": 1668 }, { "epoch": 0.7144691780821918, "grad_norm": 6.747955322265625, "learning_rate": 2.855308219178082e-06, "loss": 0.0037, "step": 1669 }, { "epoch": 0.7148972602739726, "grad_norm": 23.76809310913086, "learning_rate": 2.8510273972602742e-06, "loss": 0.0109, "step": 1670 }, { "epoch": 0.7153253424657534, "grad_norm": 0.6451877355575562, "learning_rate": 2.8467465753424663e-06, "loss": 0.0005, "step": 1671 }, { "epoch": 0.7157534246575342, "grad_norm": 1.9164334535598755, "learning_rate": 2.842465753424658e-06, "loss": 0.0007, "step": 1672 }, { "epoch": 0.716181506849315, "grad_norm": 6.329676628112793, "learning_rate": 2.8381849315068495e-06, "loss": 0.0033, "step": 1673 }, { "epoch": 0.7166095890410958, "grad_norm": 8.550711631774902, "learning_rate": 2.833904109589041e-06, "loss": 0.0055, "step": 1674 }, { "epoch": 0.7170376712328768, "grad_norm": 4.48649263381958, "learning_rate": 2.8296232876712332e-06, "loss": 0.0018, "step": 1675 }, { "epoch": 0.7174657534246576, "grad_norm": 4.23390007019043, "learning_rate": 2.8253424657534253e-06, "loss": 0.0016, "step": 1676 }, { "epoch": 0.7178938356164384, "grad_norm": 0.48566845059394836, "learning_rate": 2.8210616438356165e-06, "loss": 0.0004, "step": 1677 }, { "epoch": 0.7183219178082192, "grad_norm": 3.36252760887146, "learning_rate": 2.816780821917808e-06, "loss": 0.0038, "step": 1678 }, { "epoch": 0.71875, "grad_norm": 9.620567321777344, "learning_rate": 2.8125e-06, "loss": 0.0073, "step": 1679 }, { "epoch": 0.7191780821917808, "grad_norm": 2.4459731578826904, "learning_rate": 2.8082191780821922e-06, "loss": 0.0019, "step": 1680 }, { "epoch": 0.7196061643835616, "grad_norm": 5.555056571960449, "learning_rate": 2.803938356164384e-06, "loss": 0.0054, "step": 1681 }, { "epoch": 0.7200342465753424, "grad_norm": 2.347841501235962, "learning_rate": 2.7996575342465755e-06, "loss": 0.0021, "step": 1682 }, { "epoch": 0.7204623287671232, "grad_norm": 1.6606602668762207, "learning_rate": 2.795376712328767e-06, "loss": 0.0005, "step": 1683 }, { "epoch": 0.7208904109589042, "grad_norm": 3.1642956733703613, "learning_rate": 2.791095890410959e-06, "loss": 0.0012, "step": 1684 }, { "epoch": 0.721318493150685, "grad_norm": 8.396029472351074, "learning_rate": 2.7868150684931512e-06, "loss": 0.0148, "step": 1685 }, { "epoch": 0.7217465753424658, "grad_norm": 5.485311031341553, "learning_rate": 2.7825342465753424e-06, "loss": 0.0031, "step": 1686 }, { "epoch": 0.7221746575342466, "grad_norm": 17.260194778442383, "learning_rate": 2.7782534246575345e-06, "loss": 0.0108, "step": 1687 }, { "epoch": 0.7226027397260274, "grad_norm": 3.0003623962402344, "learning_rate": 2.773972602739726e-06, "loss": 0.0048, "step": 1688 }, { "epoch": 0.7230308219178082, "grad_norm": 5.597887992858887, "learning_rate": 2.769691780821918e-06, "loss": 0.0025, "step": 1689 }, { "epoch": 0.723458904109589, "grad_norm": 7.411900043487549, "learning_rate": 2.7654109589041102e-06, "loss": 0.0052, "step": 1690 }, { "epoch": 0.7238869863013698, "grad_norm": 4.9059648513793945, "learning_rate": 2.7611301369863014e-06, "loss": 0.0021, "step": 1691 }, { "epoch": 0.7243150684931506, "grad_norm": 7.791223049163818, "learning_rate": 2.7568493150684935e-06, "loss": 0.0045, "step": 1692 }, { "epoch": 0.7247431506849316, "grad_norm": 5.259424686431885, "learning_rate": 2.752568493150685e-06, "loss": 0.003, "step": 1693 }, { "epoch": 0.7251712328767124, "grad_norm": 0.8124938011169434, "learning_rate": 2.748287671232877e-06, "loss": 0.001, "step": 1694 }, { "epoch": 0.7255993150684932, "grad_norm": 1.7296770811080933, "learning_rate": 2.7440068493150684e-06, "loss": 0.0008, "step": 1695 }, { "epoch": 0.726027397260274, "grad_norm": 2.7262420654296875, "learning_rate": 2.7397260273972604e-06, "loss": 0.0009, "step": 1696 }, { "epoch": 0.7264554794520548, "grad_norm": 6.579883575439453, "learning_rate": 2.735445205479452e-06, "loss": 0.0037, "step": 1697 }, { "epoch": 0.7268835616438356, "grad_norm": 2.726407289505005, "learning_rate": 2.731164383561644e-06, "loss": 0.0031, "step": 1698 }, { "epoch": 0.7273116438356164, "grad_norm": 3.5342764854431152, "learning_rate": 2.726883561643836e-06, "loss": 0.0012, "step": 1699 }, { "epoch": 0.7277397260273972, "grad_norm": 1.9516152143478394, "learning_rate": 2.7226027397260274e-06, "loss": 0.0005, "step": 1700 }, { "epoch": 0.728167808219178, "grad_norm": 1.912180781364441, "learning_rate": 2.7183219178082194e-06, "loss": 0.0022, "step": 1701 }, { "epoch": 0.728595890410959, "grad_norm": 4.588295936584473, "learning_rate": 2.714041095890411e-06, "loss": 0.0045, "step": 1702 }, { "epoch": 0.7290239726027398, "grad_norm": 22.36650848388672, "learning_rate": 2.709760273972603e-06, "loss": 0.0091, "step": 1703 }, { "epoch": 0.7294520547945206, "grad_norm": 1.2226773500442505, "learning_rate": 2.705479452054795e-06, "loss": 0.0011, "step": 1704 }, { "epoch": 0.7298801369863014, "grad_norm": 7.702093601226807, "learning_rate": 2.7011986301369864e-06, "loss": 0.0043, "step": 1705 }, { "epoch": 0.7303082191780822, "grad_norm": 2.660301446914673, "learning_rate": 2.6969178082191784e-06, "loss": 0.0051, "step": 1706 }, { "epoch": 0.730736301369863, "grad_norm": 6.549902439117432, "learning_rate": 2.69263698630137e-06, "loss": 0.0034, "step": 1707 }, { "epoch": 0.7311643835616438, "grad_norm": 4.062997817993164, "learning_rate": 2.688356164383562e-06, "loss": 0.0017, "step": 1708 }, { "epoch": 0.7315924657534246, "grad_norm": 5.627403259277344, "learning_rate": 2.6840753424657533e-06, "loss": 0.0031, "step": 1709 }, { "epoch": 0.7320205479452054, "grad_norm": 1.0681564807891846, "learning_rate": 2.6797945205479454e-06, "loss": 0.0019, "step": 1710 }, { "epoch": 0.7324486301369864, "grad_norm": 0.9374256730079651, "learning_rate": 2.675513698630137e-06, "loss": 0.0007, "step": 1711 }, { "epoch": 0.7328767123287672, "grad_norm": 2.7956769466400146, "learning_rate": 2.671232876712329e-06, "loss": 0.0045, "step": 1712 }, { "epoch": 0.733304794520548, "grad_norm": 6.726290702819824, "learning_rate": 2.666952054794521e-06, "loss": 0.0034, "step": 1713 }, { "epoch": 0.7337328767123288, "grad_norm": 1.3944538831710815, "learning_rate": 2.6626712328767123e-06, "loss": 0.0005, "step": 1714 }, { "epoch": 0.7341609589041096, "grad_norm": 58.86363983154297, "learning_rate": 2.6583904109589044e-06, "loss": 0.481, "step": 1715 }, { "epoch": 0.7345890410958904, "grad_norm": 59.62095642089844, "learning_rate": 2.654109589041096e-06, "loss": 0.051, "step": 1716 }, { "epoch": 0.7350171232876712, "grad_norm": 56.33080291748047, "learning_rate": 2.649828767123288e-06, "loss": 0.0432, "step": 1717 }, { "epoch": 0.735445205479452, "grad_norm": 14.889920234680176, "learning_rate": 2.6455479452054793e-06, "loss": 0.0104, "step": 1718 }, { "epoch": 0.7358732876712328, "grad_norm": 10.357416152954102, "learning_rate": 2.6412671232876713e-06, "loss": 0.008, "step": 1719 }, { "epoch": 0.7363013698630136, "grad_norm": 5.919829845428467, "learning_rate": 2.6369863013698634e-06, "loss": 0.0034, "step": 1720 }, { "epoch": 0.7367294520547946, "grad_norm": 34.16865921020508, "learning_rate": 2.632705479452055e-06, "loss": 0.0343, "step": 1721 }, { "epoch": 0.7371575342465754, "grad_norm": 5.061959743499756, "learning_rate": 2.628424657534247e-06, "loss": 0.002, "step": 1722 }, { "epoch": 0.7375856164383562, "grad_norm": 4.334521293640137, "learning_rate": 2.6241438356164383e-06, "loss": 0.0022, "step": 1723 }, { "epoch": 0.738013698630137, "grad_norm": 4.589346408843994, "learning_rate": 2.6198630136986303e-06, "loss": 0.0021, "step": 1724 }, { "epoch": 0.7384417808219178, "grad_norm": 2.0270659923553467, "learning_rate": 2.6155821917808224e-06, "loss": 0.0024, "step": 1725 }, { "epoch": 0.7388698630136986, "grad_norm": 149.47914123535156, "learning_rate": 2.611301369863014e-06, "loss": 0.1209, "step": 1726 }, { "epoch": 0.7392979452054794, "grad_norm": 9.978551864624023, "learning_rate": 2.6070205479452056e-06, "loss": 0.0071, "step": 1727 }, { "epoch": 0.7397260273972602, "grad_norm": 6.076265811920166, "learning_rate": 2.6027397260273973e-06, "loss": 0.0065, "step": 1728 }, { "epoch": 0.740154109589041, "grad_norm": 1.6027851104736328, "learning_rate": 2.5984589041095893e-06, "loss": 0.0022, "step": 1729 }, { "epoch": 0.740582191780822, "grad_norm": 92.83515167236328, "learning_rate": 2.594178082191781e-06, "loss": 0.3594, "step": 1730 }, { "epoch": 0.7410102739726028, "grad_norm": 9.407625198364258, "learning_rate": 2.589897260273973e-06, "loss": 0.0066, "step": 1731 }, { "epoch": 0.7414383561643836, "grad_norm": 40.04216766357422, "learning_rate": 2.585616438356164e-06, "loss": 0.0368, "step": 1732 }, { "epoch": 0.7418664383561644, "grad_norm": 7.513312816619873, "learning_rate": 2.5813356164383563e-06, "loss": 0.004, "step": 1733 }, { "epoch": 0.7422945205479452, "grad_norm": 3.9087157249450684, "learning_rate": 2.5770547945205483e-06, "loss": 0.0015, "step": 1734 }, { "epoch": 0.742722602739726, "grad_norm": 4.500131130218506, "learning_rate": 2.57277397260274e-06, "loss": 0.0017, "step": 1735 }, { "epoch": 0.7431506849315068, "grad_norm": 1.677527904510498, "learning_rate": 2.568493150684932e-06, "loss": 0.0021, "step": 1736 }, { "epoch": 0.7435787671232876, "grad_norm": 5.9777140617370605, "learning_rate": 2.564212328767123e-06, "loss": 0.0041, "step": 1737 }, { "epoch": 0.7440068493150684, "grad_norm": 2.5368635654449463, "learning_rate": 2.5599315068493153e-06, "loss": 0.0007, "step": 1738 }, { "epoch": 0.7444349315068494, "grad_norm": 2.009899854660034, "learning_rate": 2.5556506849315073e-06, "loss": 0.0023, "step": 1739 }, { "epoch": 0.7448630136986302, "grad_norm": 0.878929615020752, "learning_rate": 2.551369863013699e-06, "loss": 0.0007, "step": 1740 }, { "epoch": 0.745291095890411, "grad_norm": 124.3022232055664, "learning_rate": 2.5470890410958906e-06, "loss": 0.2593, "step": 1741 }, { "epoch": 0.7457191780821918, "grad_norm": 4.143433094024658, "learning_rate": 2.542808219178082e-06, "loss": 0.0015, "step": 1742 }, { "epoch": 0.7461472602739726, "grad_norm": 4.964057922363281, "learning_rate": 2.5385273972602743e-06, "loss": 0.0037, "step": 1743 }, { "epoch": 0.7465753424657534, "grad_norm": 2.308248996734619, "learning_rate": 2.534246575342466e-06, "loss": 0.0014, "step": 1744 }, { "epoch": 0.7470034246575342, "grad_norm": 2.744887113571167, "learning_rate": 2.529965753424658e-06, "loss": 0.0009, "step": 1745 }, { "epoch": 0.747431506849315, "grad_norm": 4.791858673095703, "learning_rate": 2.525684931506849e-06, "loss": 0.003, "step": 1746 }, { "epoch": 0.7478595890410958, "grad_norm": 0.7155559062957764, "learning_rate": 2.521404109589041e-06, "loss": 0.0006, "step": 1747 }, { "epoch": 0.7482876712328768, "grad_norm": 8.799507141113281, "learning_rate": 2.5171232876712333e-06, "loss": 0.0057, "step": 1748 }, { "epoch": 0.7487157534246576, "grad_norm": 20.310829162597656, "learning_rate": 2.512842465753425e-06, "loss": 0.01, "step": 1749 }, { "epoch": 0.7491438356164384, "grad_norm": 6.482624053955078, "learning_rate": 2.5085616438356165e-06, "loss": 0.0034, "step": 1750 }, { "epoch": 0.7495719178082192, "grad_norm": 5.174587249755859, "learning_rate": 2.504280821917808e-06, "loss": 0.0019, "step": 1751 }, { "epoch": 0.75, "grad_norm": 2.5896923542022705, "learning_rate": 2.5e-06, "loss": 0.0009, "step": 1752 }, { "epoch": 0.7504280821917808, "grad_norm": 0.717301070690155, "learning_rate": 2.4957191780821923e-06, "loss": 0.0006, "step": 1753 }, { "epoch": 0.7508561643835616, "grad_norm": 41.71918487548828, "learning_rate": 2.491438356164384e-06, "loss": 0.0298, "step": 1754 }, { "epoch": 0.7512842465753424, "grad_norm": 58.27811813354492, "learning_rate": 2.4871575342465755e-06, "loss": 0.7237, "step": 1755 }, { "epoch": 0.7517123287671232, "grad_norm": 4.752049922943115, "learning_rate": 2.482876712328767e-06, "loss": 0.0031, "step": 1756 }, { "epoch": 0.7521404109589042, "grad_norm": 1.564629077911377, "learning_rate": 2.478595890410959e-06, "loss": 0.0005, "step": 1757 }, { "epoch": 0.752568493150685, "grad_norm": 5.421573638916016, "learning_rate": 2.474315068493151e-06, "loss": 0.0031, "step": 1758 }, { "epoch": 0.7529965753424658, "grad_norm": 3.687087297439575, "learning_rate": 2.470034246575343e-06, "loss": 0.0013, "step": 1759 }, { "epoch": 0.7534246575342466, "grad_norm": 3.4277830123901367, "learning_rate": 2.4657534246575345e-06, "loss": 0.0038, "step": 1760 }, { "epoch": 0.7538527397260274, "grad_norm": 156.96900939941406, "learning_rate": 2.461472602739726e-06, "loss": 0.1647, "step": 1761 }, { "epoch": 0.7542808219178082, "grad_norm": 2.0456349849700928, "learning_rate": 2.457191780821918e-06, "loss": 0.0014, "step": 1762 }, { "epoch": 0.754708904109589, "grad_norm": 0.5045717358589172, "learning_rate": 2.45291095890411e-06, "loss": 0.0006, "step": 1763 }, { "epoch": 0.7551369863013698, "grad_norm": 4.159392356872559, "learning_rate": 2.4486301369863015e-06, "loss": 0.0022, "step": 1764 }, { "epoch": 0.7555650684931506, "grad_norm": 16.560396194458008, "learning_rate": 2.444349315068493e-06, "loss": 0.0136, "step": 1765 }, { "epoch": 0.7559931506849316, "grad_norm": 474.90185546875, "learning_rate": 2.440068493150685e-06, "loss": 1.1879, "step": 1766 }, { "epoch": 0.7564212328767124, "grad_norm": 10.655787467956543, "learning_rate": 2.4357876712328768e-06, "loss": 0.0172, "step": 1767 }, { "epoch": 0.7568493150684932, "grad_norm": 5.5037407875061035, "learning_rate": 2.431506849315069e-06, "loss": 0.0049, "step": 1768 }, { "epoch": 0.757277397260274, "grad_norm": 5.3989973068237305, "learning_rate": 2.4272260273972605e-06, "loss": 0.0024, "step": 1769 }, { "epoch": 0.7577054794520548, "grad_norm": 87.94680786132812, "learning_rate": 2.422945205479452e-06, "loss": 0.5982, "step": 1770 }, { "epoch": 0.7581335616438356, "grad_norm": 57.705787658691406, "learning_rate": 2.418664383561644e-06, "loss": 0.5807, "step": 1771 }, { "epoch": 0.7585616438356164, "grad_norm": 4.955946445465088, "learning_rate": 2.4143835616438358e-06, "loss": 0.0022, "step": 1772 }, { "epoch": 0.7589897260273972, "grad_norm": 7.130711555480957, "learning_rate": 2.410102739726028e-06, "loss": 0.0065, "step": 1773 }, { "epoch": 0.759417808219178, "grad_norm": 2.1886653900146484, "learning_rate": 2.4058219178082195e-06, "loss": 0.0008, "step": 1774 }, { "epoch": 0.759845890410959, "grad_norm": 0.6742653846740723, "learning_rate": 2.401541095890411e-06, "loss": 0.0011, "step": 1775 }, { "epoch": 0.7602739726027398, "grad_norm": 3.7320444583892822, "learning_rate": 2.3972602739726027e-06, "loss": 0.0034, "step": 1776 }, { "epoch": 0.7607020547945206, "grad_norm": 6.7068071365356445, "learning_rate": 2.3929794520547948e-06, "loss": 0.0048, "step": 1777 }, { "epoch": 0.7611301369863014, "grad_norm": 2.7274997234344482, "learning_rate": 2.3886986301369864e-06, "loss": 0.0024, "step": 1778 }, { "epoch": 0.7615582191780822, "grad_norm": 0.6394506692886353, "learning_rate": 2.384417808219178e-06, "loss": 0.001, "step": 1779 }, { "epoch": 0.761986301369863, "grad_norm": 1.7245248556137085, "learning_rate": 2.38013698630137e-06, "loss": 0.0013, "step": 1780 }, { "epoch": 0.7624143835616438, "grad_norm": 6.798206329345703, "learning_rate": 2.3758561643835617e-06, "loss": 0.0036, "step": 1781 }, { "epoch": 0.7628424657534246, "grad_norm": 3.942227840423584, "learning_rate": 2.3715753424657538e-06, "loss": 0.0024, "step": 1782 }, { "epoch": 0.7632705479452054, "grad_norm": 166.9664764404297, "learning_rate": 2.3672945205479454e-06, "loss": 0.3518, "step": 1783 }, { "epoch": 0.7636986301369864, "grad_norm": 4.7455363273620605, "learning_rate": 2.363013698630137e-06, "loss": 0.0018, "step": 1784 }, { "epoch": 0.7641267123287672, "grad_norm": 3.056523561477661, "learning_rate": 2.358732876712329e-06, "loss": 0.0018, "step": 1785 }, { "epoch": 0.764554794520548, "grad_norm": 0.5848721265792847, "learning_rate": 2.3544520547945207e-06, "loss": 0.0007, "step": 1786 }, { "epoch": 0.7649828767123288, "grad_norm": 57.065956115722656, "learning_rate": 2.3501712328767128e-06, "loss": 0.5923, "step": 1787 }, { "epoch": 0.7654109589041096, "grad_norm": 5.5825371742248535, "learning_rate": 2.3458904109589044e-06, "loss": 0.0023, "step": 1788 }, { "epoch": 0.7658390410958904, "grad_norm": 4.67862606048584, "learning_rate": 2.341609589041096e-06, "loss": 0.0029, "step": 1789 }, { "epoch": 0.7662671232876712, "grad_norm": 2.6217315196990967, "learning_rate": 2.3373287671232877e-06, "loss": 0.0017, "step": 1790 }, { "epoch": 0.766695205479452, "grad_norm": 1.9339593648910522, "learning_rate": 2.3330479452054797e-06, "loss": 0.0008, "step": 1791 }, { "epoch": 0.7671232876712328, "grad_norm": 3.2226319313049316, "learning_rate": 2.3287671232876713e-06, "loss": 0.0015, "step": 1792 }, { "epoch": 0.7675513698630136, "grad_norm": 258.5383605957031, "learning_rate": 2.3244863013698634e-06, "loss": 0.2796, "step": 1793 }, { "epoch": 0.7679794520547946, "grad_norm": 3.7849924564361572, "learning_rate": 2.320205479452055e-06, "loss": 0.0032, "step": 1794 }, { "epoch": 0.7684075342465754, "grad_norm": 6.783998489379883, "learning_rate": 2.3159246575342467e-06, "loss": 0.0039, "step": 1795 }, { "epoch": 0.7688356164383562, "grad_norm": 3.619011163711548, "learning_rate": 2.3116438356164387e-06, "loss": 0.0016, "step": 1796 }, { "epoch": 0.769263698630137, "grad_norm": 3.570925712585449, "learning_rate": 2.3073630136986303e-06, "loss": 0.0028, "step": 1797 }, { "epoch": 0.7696917808219178, "grad_norm": 4.224192142486572, "learning_rate": 2.303082191780822e-06, "loss": 0.0089, "step": 1798 }, { "epoch": 0.7701198630136986, "grad_norm": 2.4116737842559814, "learning_rate": 2.2988013698630136e-06, "loss": 0.0028, "step": 1799 }, { "epoch": 0.7705479452054794, "grad_norm": 124.53852081298828, "learning_rate": 2.2945205479452057e-06, "loss": 0.2152, "step": 1800 }, { "epoch": 0.7709760273972602, "grad_norm": 2.73903489112854, "learning_rate": 2.2902397260273977e-06, "loss": 0.0017, "step": 1801 }, { "epoch": 0.771404109589041, "grad_norm": 1.4771884679794312, "learning_rate": 2.2859589041095893e-06, "loss": 0.0006, "step": 1802 }, { "epoch": 0.771832191780822, "grad_norm": 1.5881068706512451, "learning_rate": 2.281678082191781e-06, "loss": 0.001, "step": 1803 }, { "epoch": 0.7722602739726028, "grad_norm": 3.2447099685668945, "learning_rate": 2.2773972602739726e-06, "loss": 0.0036, "step": 1804 }, { "epoch": 0.7726883561643836, "grad_norm": 70.79975128173828, "learning_rate": 2.2731164383561647e-06, "loss": 0.5384, "step": 1805 }, { "epoch": 0.7731164383561644, "grad_norm": 1.407033085823059, "learning_rate": 2.2688356164383563e-06, "loss": 0.0011, "step": 1806 }, { "epoch": 0.7735445205479452, "grad_norm": 5.285018444061279, "learning_rate": 2.2645547945205483e-06, "loss": 0.0027, "step": 1807 }, { "epoch": 0.773972602739726, "grad_norm": 3.906313180923462, "learning_rate": 2.26027397260274e-06, "loss": 0.0023, "step": 1808 }, { "epoch": 0.7744006849315068, "grad_norm": 0.439988374710083, "learning_rate": 2.2559931506849316e-06, "loss": 0.0008, "step": 1809 }, { "epoch": 0.7748287671232876, "grad_norm": 2.462493896484375, "learning_rate": 2.2517123287671237e-06, "loss": 0.0024, "step": 1810 }, { "epoch": 0.7752568493150684, "grad_norm": 4.293789863586426, "learning_rate": 2.2474315068493153e-06, "loss": 0.0019, "step": 1811 }, { "epoch": 0.7756849315068494, "grad_norm": 119.20621490478516, "learning_rate": 2.243150684931507e-06, "loss": 0.1702, "step": 1812 }, { "epoch": 0.7761130136986302, "grad_norm": 0.7937411069869995, "learning_rate": 2.2388698630136986e-06, "loss": 0.0007, "step": 1813 }, { "epoch": 0.776541095890411, "grad_norm": 2.262157917022705, "learning_rate": 2.2345890410958906e-06, "loss": 0.0007, "step": 1814 }, { "epoch": 0.7769691780821918, "grad_norm": 5.747010231018066, "learning_rate": 2.2303082191780822e-06, "loss": 0.0032, "step": 1815 }, { "epoch": 0.7773972602739726, "grad_norm": 5.210381507873535, "learning_rate": 2.2260273972602743e-06, "loss": 0.0045, "step": 1816 }, { "epoch": 0.7778253424657534, "grad_norm": 8.51820182800293, "learning_rate": 2.221746575342466e-06, "loss": 0.005, "step": 1817 }, { "epoch": 0.7782534246575342, "grad_norm": 4.7299041748046875, "learning_rate": 2.2174657534246575e-06, "loss": 0.002, "step": 1818 }, { "epoch": 0.778681506849315, "grad_norm": 114.94373321533203, "learning_rate": 2.2131849315068496e-06, "loss": 0.4995, "step": 1819 }, { "epoch": 0.7791095890410958, "grad_norm": 1.3082828521728516, "learning_rate": 2.2089041095890412e-06, "loss": 0.0005, "step": 1820 }, { "epoch": 0.7795376712328768, "grad_norm": 6.529698371887207, "learning_rate": 2.2046232876712333e-06, "loss": 0.0045, "step": 1821 }, { "epoch": 0.7799657534246576, "grad_norm": 0.6540308594703674, "learning_rate": 2.200342465753425e-06, "loss": 0.0006, "step": 1822 }, { "epoch": 0.7803938356164384, "grad_norm": 2.2330188751220703, "learning_rate": 2.1960616438356165e-06, "loss": 0.0009, "step": 1823 }, { "epoch": 0.7808219178082192, "grad_norm": 5.597017765045166, "learning_rate": 2.191780821917808e-06, "loss": 0.003, "step": 1824 }, { "epoch": 0.78125, "grad_norm": 1.6804215908050537, "learning_rate": 2.1875000000000002e-06, "loss": 0.0016, "step": 1825 }, { "epoch": 0.7816780821917808, "grad_norm": 3.888646364212036, "learning_rate": 2.1832191780821923e-06, "loss": 0.0037, "step": 1826 }, { "epoch": 0.7821061643835616, "grad_norm": 37.31660461425781, "learning_rate": 2.178938356164384e-06, "loss": 0.0482, "step": 1827 }, { "epoch": 0.7825342465753424, "grad_norm": 5.07855224609375, "learning_rate": 2.1746575342465755e-06, "loss": 0.0029, "step": 1828 }, { "epoch": 0.7829623287671232, "grad_norm": 6.412681579589844, "learning_rate": 2.170376712328767e-06, "loss": 0.0049, "step": 1829 }, { "epoch": 0.7833904109589042, "grad_norm": 0.39119061827659607, "learning_rate": 2.1660958904109592e-06, "loss": 0.0003, "step": 1830 }, { "epoch": 0.783818493150685, "grad_norm": 2.632390022277832, "learning_rate": 2.161815068493151e-06, "loss": 0.0008, "step": 1831 }, { "epoch": 0.7842465753424658, "grad_norm": 1.8024394512176514, "learning_rate": 2.1575342465753425e-06, "loss": 0.001, "step": 1832 }, { "epoch": 0.7846746575342466, "grad_norm": 59.881343841552734, "learning_rate": 2.153253424657534e-06, "loss": 0.6662, "step": 1833 }, { "epoch": 0.7851027397260274, "grad_norm": 57.92162322998047, "learning_rate": 2.148972602739726e-06, "loss": 0.6254, "step": 1834 }, { "epoch": 0.7855308219178082, "grad_norm": 8.216519355773926, "learning_rate": 2.1446917808219182e-06, "loss": 0.005, "step": 1835 }, { "epoch": 0.785958904109589, "grad_norm": 25.54343605041504, "learning_rate": 2.14041095890411e-06, "loss": 0.0158, "step": 1836 }, { "epoch": 0.7863869863013698, "grad_norm": 1.2362462282180786, "learning_rate": 2.1361301369863015e-06, "loss": 0.0005, "step": 1837 }, { "epoch": 0.7868150684931506, "grad_norm": 3.45320463180542, "learning_rate": 2.131849315068493e-06, "loss": 0.0012, "step": 1838 }, { "epoch": 0.7872431506849316, "grad_norm": 2.664525032043457, "learning_rate": 2.127568493150685e-06, "loss": 0.0008, "step": 1839 }, { "epoch": 0.7876712328767124, "grad_norm": 1.7475284337997437, "learning_rate": 2.123287671232877e-06, "loss": 0.0012, "step": 1840 }, { "epoch": 0.7880993150684932, "grad_norm": 0.3151736557483673, "learning_rate": 2.119006849315069e-06, "loss": 0.0006, "step": 1841 }, { "epoch": 0.788527397260274, "grad_norm": 4.835506439208984, "learning_rate": 2.1147260273972605e-06, "loss": 0.0024, "step": 1842 }, { "epoch": 0.7889554794520548, "grad_norm": 11.5018949508667, "learning_rate": 2.110445205479452e-06, "loss": 0.0074, "step": 1843 }, { "epoch": 0.7893835616438356, "grad_norm": 7.25272274017334, "learning_rate": 2.106164383561644e-06, "loss": 0.0041, "step": 1844 }, { "epoch": 0.7898116438356164, "grad_norm": 4.006270408630371, "learning_rate": 2.101883561643836e-06, "loss": 0.0023, "step": 1845 }, { "epoch": 0.7902397260273972, "grad_norm": 4.403939247131348, "learning_rate": 2.0976027397260274e-06, "loss": 0.0033, "step": 1846 }, { "epoch": 0.790667808219178, "grad_norm": 2.9222700595855713, "learning_rate": 2.093321917808219e-06, "loss": 0.0015, "step": 1847 }, { "epoch": 0.791095890410959, "grad_norm": 1.1199190616607666, "learning_rate": 2.089041095890411e-06, "loss": 0.0014, "step": 1848 }, { "epoch": 0.7915239726027398, "grad_norm": 1.832234263420105, "learning_rate": 2.0847602739726027e-06, "loss": 0.0006, "step": 1849 }, { "epoch": 0.7919520547945206, "grad_norm": 1.1324692964553833, "learning_rate": 2.080479452054795e-06, "loss": 0.002, "step": 1850 }, { "epoch": 0.7923801369863014, "grad_norm": 1.2470128536224365, "learning_rate": 2.0761986301369864e-06, "loss": 0.0008, "step": 1851 }, { "epoch": 0.7928082191780822, "grad_norm": 67.83740234375, "learning_rate": 2.071917808219178e-06, "loss": 0.4506, "step": 1852 }, { "epoch": 0.793236301369863, "grad_norm": 5.629336357116699, "learning_rate": 2.06763698630137e-06, "loss": 0.0059, "step": 1853 }, { "epoch": 0.7936643835616438, "grad_norm": 2.4384357929229736, "learning_rate": 2.0633561643835617e-06, "loss": 0.0013, "step": 1854 }, { "epoch": 0.7940924657534246, "grad_norm": 14.207345962524414, "learning_rate": 2.059075342465754e-06, "loss": 0.0115, "step": 1855 }, { "epoch": 0.7945205479452054, "grad_norm": 0.4984908699989319, "learning_rate": 2.0547945205479454e-06, "loss": 0.0004, "step": 1856 }, { "epoch": 0.7949486301369864, "grad_norm": 1.13307785987854, "learning_rate": 2.050513698630137e-06, "loss": 0.0005, "step": 1857 }, { "epoch": 0.7953767123287672, "grad_norm": 1.403398036956787, "learning_rate": 2.046232876712329e-06, "loss": 0.0007, "step": 1858 }, { "epoch": 0.795804794520548, "grad_norm": 1.5276222229003906, "learning_rate": 2.0419520547945207e-06, "loss": 0.0016, "step": 1859 }, { "epoch": 0.7962328767123288, "grad_norm": 3.5548596382141113, "learning_rate": 2.037671232876713e-06, "loss": 0.0011, "step": 1860 }, { "epoch": 0.7966609589041096, "grad_norm": 7.326754093170166, "learning_rate": 2.033390410958904e-06, "loss": 0.0074, "step": 1861 }, { "epoch": 0.7970890410958904, "grad_norm": 8.07601261138916, "learning_rate": 2.029109589041096e-06, "loss": 0.0072, "step": 1862 }, { "epoch": 0.7975171232876712, "grad_norm": 1.8106070756912231, "learning_rate": 2.0248287671232877e-06, "loss": 0.0059, "step": 1863 }, { "epoch": 0.797945205479452, "grad_norm": 4.600616455078125, "learning_rate": 2.0205479452054797e-06, "loss": 0.0018, "step": 1864 }, { "epoch": 0.7983732876712328, "grad_norm": 93.08236694335938, "learning_rate": 2.0162671232876714e-06, "loss": 0.5259, "step": 1865 }, { "epoch": 0.7988013698630136, "grad_norm": 169.8060302734375, "learning_rate": 2.011986301369863e-06, "loss": 0.5316, "step": 1866 }, { "epoch": 0.7992294520547946, "grad_norm": 1.8019589185714722, "learning_rate": 2.007705479452055e-06, "loss": 0.002, "step": 1867 }, { "epoch": 0.7996575342465754, "grad_norm": 15.802233695983887, "learning_rate": 2.0034246575342467e-06, "loss": 0.0126, "step": 1868 }, { "epoch": 0.8000856164383562, "grad_norm": 1.712005376815796, "learning_rate": 1.9991438356164387e-06, "loss": 0.0007, "step": 1869 }, { "epoch": 0.800513698630137, "grad_norm": 5.527495861053467, "learning_rate": 1.9948630136986304e-06, "loss": 0.0032, "step": 1870 }, { "epoch": 0.8009417808219178, "grad_norm": 468.1472473144531, "learning_rate": 1.990582191780822e-06, "loss": 0.3242, "step": 1871 }, { "epoch": 0.8013698630136986, "grad_norm": 0.9060855507850647, "learning_rate": 1.9863013698630136e-06, "loss": 0.001, "step": 1872 }, { "epoch": 0.8017979452054794, "grad_norm": 58.026126861572266, "learning_rate": 1.9820205479452057e-06, "loss": 0.5197, "step": 1873 }, { "epoch": 0.8022260273972602, "grad_norm": 0.21686245501041412, "learning_rate": 1.9777397260273977e-06, "loss": 0.0003, "step": 1874 }, { "epoch": 0.802654109589041, "grad_norm": 9.2562255859375, "learning_rate": 1.9734589041095894e-06, "loss": 0.0043, "step": 1875 }, { "epoch": 0.803082191780822, "grad_norm": 4.690082550048828, "learning_rate": 1.969178082191781e-06, "loss": 0.0018, "step": 1876 }, { "epoch": 0.8035102739726028, "grad_norm": 56.147865295410156, "learning_rate": 1.9648972602739726e-06, "loss": 0.5861, "step": 1877 }, { "epoch": 0.8039383561643836, "grad_norm": 124.49285125732422, "learning_rate": 1.9606164383561647e-06, "loss": 0.153, "step": 1878 }, { "epoch": 0.8043664383561644, "grad_norm": 6.530893325805664, "learning_rate": 1.9563356164383563e-06, "loss": 0.0039, "step": 1879 }, { "epoch": 0.8047945205479452, "grad_norm": 2.740473508834839, "learning_rate": 1.952054794520548e-06, "loss": 0.0016, "step": 1880 }, { "epoch": 0.805222602739726, "grad_norm": 73.35096740722656, "learning_rate": 1.9477739726027396e-06, "loss": 0.0488, "step": 1881 }, { "epoch": 0.8056506849315068, "grad_norm": 3.96303129196167, "learning_rate": 1.9434931506849316e-06, "loss": 0.0039, "step": 1882 }, { "epoch": 0.8060787671232876, "grad_norm": 35.65889358520508, "learning_rate": 1.9392123287671237e-06, "loss": 0.0215, "step": 1883 }, { "epoch": 0.8065068493150684, "grad_norm": 2.3047878742218018, "learning_rate": 1.9349315068493153e-06, "loss": 0.0019, "step": 1884 }, { "epoch": 0.8069349315068494, "grad_norm": 59.431312561035156, "learning_rate": 1.930650684931507e-06, "loss": 0.0349, "step": 1885 }, { "epoch": 0.8073630136986302, "grad_norm": 2.230163335800171, "learning_rate": 1.9263698630136986e-06, "loss": 0.0012, "step": 1886 }, { "epoch": 0.807791095890411, "grad_norm": 86.41316223144531, "learning_rate": 1.9220890410958906e-06, "loss": 0.644, "step": 1887 }, { "epoch": 0.8082191780821918, "grad_norm": 67.39104461669922, "learning_rate": 1.9178082191780823e-06, "loss": 0.3501, "step": 1888 }, { "epoch": 0.8086472602739726, "grad_norm": 3.2903497219085693, "learning_rate": 1.9135273972602743e-06, "loss": 0.0016, "step": 1889 }, { "epoch": 0.8090753424657534, "grad_norm": 4.647758960723877, "learning_rate": 1.909246575342466e-06, "loss": 0.003, "step": 1890 }, { "epoch": 0.8095034246575342, "grad_norm": 1.649864912033081, "learning_rate": 1.9049657534246576e-06, "loss": 0.0008, "step": 1891 }, { "epoch": 0.809931506849315, "grad_norm": 3.0110764503479004, "learning_rate": 1.9006849315068496e-06, "loss": 0.0015, "step": 1892 }, { "epoch": 0.8103595890410958, "grad_norm": 9.162644386291504, "learning_rate": 1.8964041095890413e-06, "loss": 0.0044, "step": 1893 }, { "epoch": 0.8107876712328768, "grad_norm": 3.090468406677246, "learning_rate": 1.892123287671233e-06, "loss": 0.0015, "step": 1894 }, { "epoch": 0.8112157534246576, "grad_norm": 151.02084350585938, "learning_rate": 1.8878424657534247e-06, "loss": 0.2776, "step": 1895 }, { "epoch": 0.8116438356164384, "grad_norm": 1.4924538135528564, "learning_rate": 1.8835616438356166e-06, "loss": 0.0022, "step": 1896 }, { "epoch": 0.8120719178082192, "grad_norm": 3.600553512573242, "learning_rate": 1.8792808219178082e-06, "loss": 0.0073, "step": 1897 }, { "epoch": 0.8125, "grad_norm": 23.90523338317871, "learning_rate": 1.8750000000000003e-06, "loss": 0.0091, "step": 1898 }, { "epoch": 0.8129280821917808, "grad_norm": 1.2864543199539185, "learning_rate": 1.870719178082192e-06, "loss": 0.002, "step": 1899 }, { "epoch": 0.8133561643835616, "grad_norm": 0.9033134579658508, "learning_rate": 1.8664383561643837e-06, "loss": 0.0014, "step": 1900 }, { "epoch": 0.8137842465753424, "grad_norm": 4.737240791320801, "learning_rate": 1.8621575342465756e-06, "loss": 0.0032, "step": 1901 }, { "epoch": 0.8142123287671232, "grad_norm": 2.0795350074768066, "learning_rate": 1.8578767123287672e-06, "loss": 0.001, "step": 1902 }, { "epoch": 0.8146404109589042, "grad_norm": 3.7669601440429688, "learning_rate": 1.853595890410959e-06, "loss": 0.0013, "step": 1903 }, { "epoch": 0.815068493150685, "grad_norm": 2.7164621353149414, "learning_rate": 1.8493150684931507e-06, "loss": 0.001, "step": 1904 }, { "epoch": 0.8154965753424658, "grad_norm": 14.071316719055176, "learning_rate": 1.8450342465753427e-06, "loss": 0.0088, "step": 1905 }, { "epoch": 0.8159246575342466, "grad_norm": 3.2982513904571533, "learning_rate": 1.8407534246575344e-06, "loss": 0.0013, "step": 1906 }, { "epoch": 0.8163527397260274, "grad_norm": 13.500249862670898, "learning_rate": 1.8364726027397262e-06, "loss": 0.0075, "step": 1907 }, { "epoch": 0.8167808219178082, "grad_norm": 2.4409143924713135, "learning_rate": 1.832191780821918e-06, "loss": 0.0032, "step": 1908 }, { "epoch": 0.817208904109589, "grad_norm": 8.26755142211914, "learning_rate": 1.8279109589041097e-06, "loss": 0.0052, "step": 1909 }, { "epoch": 0.8176369863013698, "grad_norm": 1.5819567441940308, "learning_rate": 1.8236301369863015e-06, "loss": 0.0008, "step": 1910 }, { "epoch": 0.8180650684931506, "grad_norm": 3.8009421825408936, "learning_rate": 1.8193493150684931e-06, "loss": 0.0019, "step": 1911 }, { "epoch": 0.8184931506849316, "grad_norm": 3.9734132289886475, "learning_rate": 1.8150684931506852e-06, "loss": 0.0028, "step": 1912 }, { "epoch": 0.8189212328767124, "grad_norm": 1.9975414276123047, "learning_rate": 1.8107876712328768e-06, "loss": 0.0013, "step": 1913 }, { "epoch": 0.8193493150684932, "grad_norm": 11.313298225402832, "learning_rate": 1.8065068493150687e-06, "loss": 0.0077, "step": 1914 }, { "epoch": 0.819777397260274, "grad_norm": 3.7634470462799072, "learning_rate": 1.8022260273972605e-06, "loss": 0.0015, "step": 1915 }, { "epoch": 0.8202054794520548, "grad_norm": 6.937271595001221, "learning_rate": 1.7979452054794521e-06, "loss": 0.004, "step": 1916 }, { "epoch": 0.8206335616438356, "grad_norm": 0.9257087111473083, "learning_rate": 1.793664383561644e-06, "loss": 0.0014, "step": 1917 }, { "epoch": 0.8210616438356164, "grad_norm": 0.18984808027744293, "learning_rate": 1.7893835616438356e-06, "loss": 0.0005, "step": 1918 }, { "epoch": 0.8214897260273972, "grad_norm": 2.7968316078186035, "learning_rate": 1.7851027397260277e-06, "loss": 0.0039, "step": 1919 }, { "epoch": 0.821917808219178, "grad_norm": 2.404803991317749, "learning_rate": 1.7808219178082193e-06, "loss": 0.001, "step": 1920 }, { "epoch": 0.822345890410959, "grad_norm": 5.654704570770264, "learning_rate": 1.7765410958904111e-06, "loss": 0.0024, "step": 1921 }, { "epoch": 0.8227739726027398, "grad_norm": 182.4481964111328, "learning_rate": 1.7722602739726028e-06, "loss": 0.0991, "step": 1922 }, { "epoch": 0.8232020547945206, "grad_norm": 31.7225284576416, "learning_rate": 1.7679794520547946e-06, "loss": 0.0677, "step": 1923 }, { "epoch": 0.8236301369863014, "grad_norm": 5.045909881591797, "learning_rate": 1.7636986301369865e-06, "loss": 0.0031, "step": 1924 }, { "epoch": 0.8240582191780822, "grad_norm": 10.991689682006836, "learning_rate": 1.759417808219178e-06, "loss": 0.0112, "step": 1925 }, { "epoch": 0.824486301369863, "grad_norm": 1.5810576677322388, "learning_rate": 1.7551369863013701e-06, "loss": 0.0015, "step": 1926 }, { "epoch": 0.8249143835616438, "grad_norm": 4.2286505699157715, "learning_rate": 1.7508561643835618e-06, "loss": 0.0019, "step": 1927 }, { "epoch": 0.8253424657534246, "grad_norm": 155.42947387695312, "learning_rate": 1.7465753424657536e-06, "loss": 0.387, "step": 1928 }, { "epoch": 0.8257705479452054, "grad_norm": 134.19593811035156, "learning_rate": 1.7422945205479452e-06, "loss": 0.9863, "step": 1929 }, { "epoch": 0.8261986301369864, "grad_norm": 0.5284451842308044, "learning_rate": 1.738013698630137e-06, "loss": 0.0006, "step": 1930 }, { "epoch": 0.8266267123287672, "grad_norm": 0.991439938545227, "learning_rate": 1.7337328767123291e-06, "loss": 0.0014, "step": 1931 }, { "epoch": 0.827054794520548, "grad_norm": 1.8861268758773804, "learning_rate": 1.7294520547945206e-06, "loss": 0.0028, "step": 1932 }, { "epoch": 0.8274828767123288, "grad_norm": 1.3123420476913452, "learning_rate": 1.7251712328767126e-06, "loss": 0.0005, "step": 1933 }, { "epoch": 0.8279109589041096, "grad_norm": 54.32320785522461, "learning_rate": 1.7208904109589042e-06, "loss": 0.5005, "step": 1934 }, { "epoch": 0.8283390410958904, "grad_norm": 103.36758422851562, "learning_rate": 1.716609589041096e-06, "loss": 0.4813, "step": 1935 }, { "epoch": 0.8287671232876712, "grad_norm": 58.999019622802734, "learning_rate": 1.7123287671232877e-06, "loss": 0.548, "step": 1936 }, { "epoch": 0.829195205479452, "grad_norm": 6.888538837432861, "learning_rate": 1.7080479452054796e-06, "loss": 0.0036, "step": 1937 }, { "epoch": 0.8296232876712328, "grad_norm": 5.507605075836182, "learning_rate": 1.7037671232876712e-06, "loss": 0.0032, "step": 1938 }, { "epoch": 0.8300513698630136, "grad_norm": 4.184703826904297, "learning_rate": 1.6994863013698632e-06, "loss": 0.0024, "step": 1939 }, { "epoch": 0.8304794520547946, "grad_norm": 1.4979307651519775, "learning_rate": 1.695205479452055e-06, "loss": 0.0015, "step": 1940 }, { "epoch": 0.8309075342465754, "grad_norm": 6.165715217590332, "learning_rate": 1.6909246575342467e-06, "loss": 0.0047, "step": 1941 }, { "epoch": 0.8313356164383562, "grad_norm": 0.5748972296714783, "learning_rate": 1.6866438356164386e-06, "loss": 0.0008, "step": 1942 }, { "epoch": 0.831763698630137, "grad_norm": 2.1094889640808105, "learning_rate": 1.6823630136986302e-06, "loss": 0.0008, "step": 1943 }, { "epoch": 0.8321917808219178, "grad_norm": 2.353755474090576, "learning_rate": 1.678082191780822e-06, "loss": 0.0009, "step": 1944 }, { "epoch": 0.8326198630136986, "grad_norm": 0.9339281916618347, "learning_rate": 1.6738013698630137e-06, "loss": 0.0007, "step": 1945 }, { "epoch": 0.8330479452054794, "grad_norm": 5.518191814422607, "learning_rate": 1.6695205479452057e-06, "loss": 0.0024, "step": 1946 }, { "epoch": 0.8334760273972602, "grad_norm": 2.2156710624694824, "learning_rate": 1.6652397260273976e-06, "loss": 0.0009, "step": 1947 }, { "epoch": 0.833904109589041, "grad_norm": 0.4846402704715729, "learning_rate": 1.6609589041095892e-06, "loss": 0.0006, "step": 1948 }, { "epoch": 0.834332191780822, "grad_norm": 54.72189712524414, "learning_rate": 1.656678082191781e-06, "loss": 0.4016, "step": 1949 }, { "epoch": 0.8347602739726028, "grad_norm": 2.464877128601074, "learning_rate": 1.6523972602739727e-06, "loss": 0.0009, "step": 1950 }, { "epoch": 0.8351883561643836, "grad_norm": 87.31864166259766, "learning_rate": 1.6481164383561645e-06, "loss": 0.1199, "step": 1951 }, { "epoch": 0.8356164383561644, "grad_norm": 7.16926383972168, "learning_rate": 1.6438356164383561e-06, "loss": 0.0057, "step": 1952 }, { "epoch": 0.8360445205479452, "grad_norm": 5.954171180725098, "learning_rate": 1.6395547945205482e-06, "loss": 0.0027, "step": 1953 }, { "epoch": 0.836472602739726, "grad_norm": 1.5759366750717163, "learning_rate": 1.6352739726027398e-06, "loss": 0.0011, "step": 1954 }, { "epoch": 0.8369006849315068, "grad_norm": 4.073934555053711, "learning_rate": 1.6309931506849317e-06, "loss": 0.002, "step": 1955 }, { "epoch": 0.8373287671232876, "grad_norm": 5.139629364013672, "learning_rate": 1.6267123287671235e-06, "loss": 0.0024, "step": 1956 }, { "epoch": 0.8377568493150684, "grad_norm": 175.98361206054688, "learning_rate": 1.6224315068493151e-06, "loss": 0.3069, "step": 1957 }, { "epoch": 0.8381849315068494, "grad_norm": 123.54338073730469, "learning_rate": 1.618150684931507e-06, "loss": 0.0802, "step": 1958 }, { "epoch": 0.8386130136986302, "grad_norm": 8.355463981628418, "learning_rate": 1.6138698630136986e-06, "loss": 0.0064, "step": 1959 }, { "epoch": 0.839041095890411, "grad_norm": 2.706456422805786, "learning_rate": 1.6095890410958907e-06, "loss": 0.0013, "step": 1960 }, { "epoch": 0.8394691780821918, "grad_norm": 199.66110229492188, "learning_rate": 1.6053082191780823e-06, "loss": 0.5393, "step": 1961 }, { "epoch": 0.8398972602739726, "grad_norm": 10.197649955749512, "learning_rate": 1.6010273972602741e-06, "loss": 0.0079, "step": 1962 }, { "epoch": 0.8403253424657534, "grad_norm": 7.84173583984375, "learning_rate": 1.596746575342466e-06, "loss": 0.0062, "step": 1963 }, { "epoch": 0.8407534246575342, "grad_norm": 31.94739532470703, "learning_rate": 1.5924657534246576e-06, "loss": 0.0105, "step": 1964 }, { "epoch": 0.841181506849315, "grad_norm": 7.568800926208496, "learning_rate": 1.5881849315068494e-06, "loss": 0.0044, "step": 1965 }, { "epoch": 0.8416095890410958, "grad_norm": 1.1610288619995117, "learning_rate": 1.583904109589041e-06, "loss": 0.0005, "step": 1966 }, { "epoch": 0.8420376712328768, "grad_norm": 8.150439262390137, "learning_rate": 1.5796232876712331e-06, "loss": 0.0065, "step": 1967 }, { "epoch": 0.8424657534246576, "grad_norm": 1.2331970930099487, "learning_rate": 1.5753424657534248e-06, "loss": 0.0008, "step": 1968 }, { "epoch": 0.8428938356164384, "grad_norm": 0.7660293579101562, "learning_rate": 1.5710616438356166e-06, "loss": 0.0004, "step": 1969 }, { "epoch": 0.8433219178082192, "grad_norm": 3.4094760417938232, "learning_rate": 1.5667808219178082e-06, "loss": 0.0016, "step": 1970 }, { "epoch": 0.84375, "grad_norm": 2.468914747238159, "learning_rate": 1.5625e-06, "loss": 0.0013, "step": 1971 }, { "epoch": 0.8441780821917808, "grad_norm": 3.7197611331939697, "learning_rate": 1.5582191780821921e-06, "loss": 0.0014, "step": 1972 }, { "epoch": 0.8446061643835616, "grad_norm": 3.7056772708892822, "learning_rate": 1.5539383561643838e-06, "loss": 0.0025, "step": 1973 }, { "epoch": 0.8450342465753424, "grad_norm": 4.4918084144592285, "learning_rate": 1.5496575342465756e-06, "loss": 0.0017, "step": 1974 }, { "epoch": 0.8454623287671232, "grad_norm": 5.482022762298584, "learning_rate": 1.5453767123287672e-06, "loss": 0.0028, "step": 1975 }, { "epoch": 0.8458904109589042, "grad_norm": 1.8710386753082275, "learning_rate": 1.541095890410959e-06, "loss": 0.0039, "step": 1976 }, { "epoch": 0.846318493150685, "grad_norm": 60.82484436035156, "learning_rate": 1.5368150684931507e-06, "loss": 0.6801, "step": 1977 }, { "epoch": 0.8467465753424658, "grad_norm": 1.8804965019226074, "learning_rate": 1.5325342465753425e-06, "loss": 0.0014, "step": 1978 }, { "epoch": 0.8471746575342466, "grad_norm": 4.324956893920898, "learning_rate": 1.5282534246575342e-06, "loss": 0.0043, "step": 1979 }, { "epoch": 0.8476027397260274, "grad_norm": 2.7642526626586914, "learning_rate": 1.5239726027397262e-06, "loss": 0.0023, "step": 1980 }, { "epoch": 0.8480308219178082, "grad_norm": 0.510888397693634, "learning_rate": 1.519691780821918e-06, "loss": 0.0005, "step": 1981 }, { "epoch": 0.848458904109589, "grad_norm": 4.244278907775879, "learning_rate": 1.5154109589041097e-06, "loss": 0.0034, "step": 1982 }, { "epoch": 0.8488869863013698, "grad_norm": 149.32481384277344, "learning_rate": 1.5111301369863015e-06, "loss": 0.1637, "step": 1983 }, { "epoch": 0.8493150684931506, "grad_norm": 2.451603412628174, "learning_rate": 1.5068493150684932e-06, "loss": 0.0013, "step": 1984 }, { "epoch": 0.8497431506849316, "grad_norm": 3.966921329498291, "learning_rate": 1.502568493150685e-06, "loss": 0.0019, "step": 1985 }, { "epoch": 0.8501712328767124, "grad_norm": 56.846168518066406, "learning_rate": 1.4982876712328766e-06, "loss": 0.7042, "step": 1986 }, { "epoch": 0.8505993150684932, "grad_norm": 56.30250549316406, "learning_rate": 1.4940068493150687e-06, "loss": 0.6804, "step": 1987 }, { "epoch": 0.851027397260274, "grad_norm": 2.571352005004883, "learning_rate": 1.4897260273972605e-06, "loss": 0.0011, "step": 1988 }, { "epoch": 0.8514554794520548, "grad_norm": 59.01203536987305, "learning_rate": 1.4854452054794522e-06, "loss": 0.5944, "step": 1989 }, { "epoch": 0.8518835616438356, "grad_norm": 3.255209445953369, "learning_rate": 1.481164383561644e-06, "loss": 0.0013, "step": 1990 }, { "epoch": 0.8523116438356164, "grad_norm": 4.1617817878723145, "learning_rate": 1.4768835616438356e-06, "loss": 0.0018, "step": 1991 }, { "epoch": 0.8527397260273972, "grad_norm": 3.669482707977295, "learning_rate": 1.4726027397260275e-06, "loss": 0.0015, "step": 1992 }, { "epoch": 0.853167808219178, "grad_norm": 199.4161834716797, "learning_rate": 1.4683219178082191e-06, "loss": 0.3424, "step": 1993 }, { "epoch": 0.853595890410959, "grad_norm": 1.4519258737564087, "learning_rate": 1.4640410958904112e-06, "loss": 0.0008, "step": 1994 }, { "epoch": 0.8540239726027398, "grad_norm": 2.990743398666382, "learning_rate": 1.4597602739726028e-06, "loss": 0.0012, "step": 1995 }, { "epoch": 0.8544520547945206, "grad_norm": 5.457079887390137, "learning_rate": 1.4554794520547946e-06, "loss": 0.0028, "step": 1996 }, { "epoch": 0.8548801369863014, "grad_norm": 1.0955448150634766, "learning_rate": 1.4511986301369865e-06, "loss": 0.0007, "step": 1997 }, { "epoch": 0.8553082191780822, "grad_norm": 0.5084691047668457, "learning_rate": 1.4469178082191781e-06, "loss": 0.0007, "step": 1998 }, { "epoch": 0.855736301369863, "grad_norm": 3.332608699798584, "learning_rate": 1.44263698630137e-06, "loss": 0.0015, "step": 1999 }, { "epoch": 0.8561643835616438, "grad_norm": 1.3960236310958862, "learning_rate": 1.4383561643835616e-06, "loss": 0.0005, "step": 2000 }, { "epoch": 0.8565924657534246, "grad_norm": 0.843875527381897, "learning_rate": 1.4340753424657536e-06, "loss": 0.0004, "step": 2001 }, { "epoch": 0.8570205479452054, "grad_norm": 0.47111305594444275, "learning_rate": 1.4297945205479453e-06, "loss": 0.0004, "step": 2002 }, { "epoch": 0.8574486301369864, "grad_norm": 1.5025198459625244, "learning_rate": 1.4255136986301371e-06, "loss": 0.0006, "step": 2003 }, { "epoch": 0.8578767123287672, "grad_norm": 0.671853244304657, "learning_rate": 1.421232876712329e-06, "loss": 0.001, "step": 2004 }, { "epoch": 0.858304794520548, "grad_norm": 2.55116605758667, "learning_rate": 1.4169520547945206e-06, "loss": 0.0008, "step": 2005 }, { "epoch": 0.8587328767123288, "grad_norm": 2.921344518661499, "learning_rate": 1.4126712328767126e-06, "loss": 0.0011, "step": 2006 }, { "epoch": 0.8591609589041096, "grad_norm": 124.09028625488281, "learning_rate": 1.408390410958904e-06, "loss": 0.2941, "step": 2007 }, { "epoch": 0.8595890410958904, "grad_norm": 1.3383313417434692, "learning_rate": 1.4041095890410961e-06, "loss": 0.0008, "step": 2008 }, { "epoch": 0.8600171232876712, "grad_norm": 3.6996521949768066, "learning_rate": 1.3998287671232877e-06, "loss": 0.0016, "step": 2009 }, { "epoch": 0.860445205479452, "grad_norm": 2.9628865718841553, "learning_rate": 1.3955479452054796e-06, "loss": 0.0014, "step": 2010 }, { "epoch": 0.8608732876712328, "grad_norm": 4.604301452636719, "learning_rate": 1.3912671232876712e-06, "loss": 0.0031, "step": 2011 }, { "epoch": 0.8613013698630136, "grad_norm": 1.4125421047210693, "learning_rate": 1.386986301369863e-06, "loss": 0.0011, "step": 2012 }, { "epoch": 0.8617294520547946, "grad_norm": 1.3975892066955566, "learning_rate": 1.3827054794520551e-06, "loss": 0.0015, "step": 2013 }, { "epoch": 0.8621575342465754, "grad_norm": 56.271385192871094, "learning_rate": 1.3784246575342467e-06, "loss": 0.4318, "step": 2014 }, { "epoch": 0.8625856164383562, "grad_norm": 3.626190423965454, "learning_rate": 1.3741438356164386e-06, "loss": 0.0019, "step": 2015 }, { "epoch": 0.863013698630137, "grad_norm": 0.9755136966705322, "learning_rate": 1.3698630136986302e-06, "loss": 0.0003, "step": 2016 }, { "epoch": 0.8634417808219178, "grad_norm": 2.4193813800811768, "learning_rate": 1.365582191780822e-06, "loss": 0.0012, "step": 2017 }, { "epoch": 0.8638698630136986, "grad_norm": 3.8916666507720947, "learning_rate": 1.3613013698630137e-06, "loss": 0.0015, "step": 2018 }, { "epoch": 0.8642979452054794, "grad_norm": 149.45785522460938, "learning_rate": 1.3570205479452055e-06, "loss": 0.2201, "step": 2019 }, { "epoch": 0.8647260273972602, "grad_norm": 5.402375221252441, "learning_rate": 1.3527397260273976e-06, "loss": 0.0043, "step": 2020 }, { "epoch": 0.865154109589041, "grad_norm": 4.682033061981201, "learning_rate": 1.3484589041095892e-06, "loss": 0.0043, "step": 2021 }, { "epoch": 0.865582191780822, "grad_norm": 21.606409072875977, "learning_rate": 1.344178082191781e-06, "loss": 0.0157, "step": 2022 }, { "epoch": 0.8660102739726028, "grad_norm": 46.892826080322266, "learning_rate": 1.3398972602739727e-06, "loss": 0.0292, "step": 2023 }, { "epoch": 0.8664383561643836, "grad_norm": 2.2376158237457275, "learning_rate": 1.3356164383561645e-06, "loss": 0.0006, "step": 2024 }, { "epoch": 0.8668664383561644, "grad_norm": 3.981309175491333, "learning_rate": 1.3313356164383562e-06, "loss": 0.0016, "step": 2025 }, { "epoch": 0.8672945205479452, "grad_norm": 6.0301947593688965, "learning_rate": 1.327054794520548e-06, "loss": 0.0031, "step": 2026 }, { "epoch": 0.867722602739726, "grad_norm": 52.6871223449707, "learning_rate": 1.3227739726027396e-06, "loss": 0.6095, "step": 2027 }, { "epoch": 0.8681506849315068, "grad_norm": 2.2579288482666016, "learning_rate": 1.3184931506849317e-06, "loss": 0.0009, "step": 2028 }, { "epoch": 0.8685787671232876, "grad_norm": 1.7433677911758423, "learning_rate": 1.3142123287671235e-06, "loss": 0.0029, "step": 2029 }, { "epoch": 0.8690068493150684, "grad_norm": 157.83827209472656, "learning_rate": 1.3099315068493152e-06, "loss": 0.2604, "step": 2030 }, { "epoch": 0.8694349315068494, "grad_norm": 0.7379125356674194, "learning_rate": 1.305650684931507e-06, "loss": 0.0005, "step": 2031 }, { "epoch": 0.8698630136986302, "grad_norm": 3.072850465774536, "learning_rate": 1.3013698630136986e-06, "loss": 0.001, "step": 2032 }, { "epoch": 0.870291095890411, "grad_norm": 0.18861767649650574, "learning_rate": 1.2970890410958905e-06, "loss": 0.0007, "step": 2033 }, { "epoch": 0.8707191780821918, "grad_norm": 7.1078033447265625, "learning_rate": 1.292808219178082e-06, "loss": 0.0071, "step": 2034 }, { "epoch": 0.8711472602739726, "grad_norm": 9.716798782348633, "learning_rate": 1.2885273972602742e-06, "loss": 0.0068, "step": 2035 }, { "epoch": 0.8715753424657534, "grad_norm": 1.1444214582443237, "learning_rate": 1.284246575342466e-06, "loss": 0.0005, "step": 2036 }, { "epoch": 0.8720034246575342, "grad_norm": 4.907404899597168, "learning_rate": 1.2799657534246576e-06, "loss": 0.0029, "step": 2037 }, { "epoch": 0.872431506849315, "grad_norm": 2.768296718597412, "learning_rate": 1.2756849315068495e-06, "loss": 0.0014, "step": 2038 }, { "epoch": 0.8728595890410958, "grad_norm": 2.1504311561584473, "learning_rate": 1.271404109589041e-06, "loss": 0.0009, "step": 2039 }, { "epoch": 0.8732876712328768, "grad_norm": 1.0654916763305664, "learning_rate": 1.267123287671233e-06, "loss": 0.0007, "step": 2040 }, { "epoch": 0.8737157534246576, "grad_norm": 10.716083526611328, "learning_rate": 1.2628424657534246e-06, "loss": 0.005, "step": 2041 }, { "epoch": 0.8741438356164384, "grad_norm": 71.5223617553711, "learning_rate": 1.2585616438356166e-06, "loss": 0.6423, "step": 2042 }, { "epoch": 0.8745719178082192, "grad_norm": 3.6259469985961914, "learning_rate": 1.2542808219178083e-06, "loss": 0.0015, "step": 2043 }, { "epoch": 0.875, "grad_norm": 4.318339824676514, "learning_rate": 1.25e-06, "loss": 0.0019, "step": 2044 }, { "epoch": 0.8754280821917808, "grad_norm": 4.178323745727539, "learning_rate": 1.245719178082192e-06, "loss": 0.0016, "step": 2045 }, { "epoch": 0.8758561643835616, "grad_norm": 15.39336109161377, "learning_rate": 1.2414383561643836e-06, "loss": 0.008, "step": 2046 }, { "epoch": 0.8762842465753424, "grad_norm": 11.642847061157227, "learning_rate": 1.2371575342465754e-06, "loss": 0.0111, "step": 2047 }, { "epoch": 0.8767123287671232, "grad_norm": 7.588431358337402, "learning_rate": 1.2328767123287673e-06, "loss": 0.0066, "step": 2048 }, { "epoch": 0.8771404109589042, "grad_norm": 2.099104642868042, "learning_rate": 1.228595890410959e-06, "loss": 0.0014, "step": 2049 }, { "epoch": 0.877568493150685, "grad_norm": 3.5699424743652344, "learning_rate": 1.2243150684931507e-06, "loss": 0.0021, "step": 2050 }, { "epoch": 0.8779965753424658, "grad_norm": 2.7273201942443848, "learning_rate": 1.2200342465753426e-06, "loss": 0.001, "step": 2051 }, { "epoch": 0.8784246575342466, "grad_norm": 2.0622575283050537, "learning_rate": 1.2157534246575344e-06, "loss": 0.0023, "step": 2052 }, { "epoch": 0.8788527397260274, "grad_norm": 1.3167810440063477, "learning_rate": 1.211472602739726e-06, "loss": 0.002, "step": 2053 }, { "epoch": 0.8792808219178082, "grad_norm": 4.148390769958496, "learning_rate": 1.2071917808219179e-06, "loss": 0.0018, "step": 2054 }, { "epoch": 0.879708904109589, "grad_norm": 55.13698196411133, "learning_rate": 1.2029109589041097e-06, "loss": 0.0263, "step": 2055 }, { "epoch": 0.8801369863013698, "grad_norm": 4.334939956665039, "learning_rate": 1.1986301369863014e-06, "loss": 0.0022, "step": 2056 }, { "epoch": 0.8805650684931506, "grad_norm": 26.684728622436523, "learning_rate": 1.1943493150684932e-06, "loss": 0.0191, "step": 2057 }, { "epoch": 0.8809931506849316, "grad_norm": 2.434631586074829, "learning_rate": 1.190068493150685e-06, "loss": 0.0008, "step": 2058 }, { "epoch": 0.8814212328767124, "grad_norm": 7.7087931632995605, "learning_rate": 1.1857876712328769e-06, "loss": 0.0069, "step": 2059 }, { "epoch": 0.8818493150684932, "grad_norm": 4.189228534698486, "learning_rate": 1.1815068493150685e-06, "loss": 0.0039, "step": 2060 }, { "epoch": 0.882277397260274, "grad_norm": 2.996400833129883, "learning_rate": 1.1772260273972604e-06, "loss": 0.0017, "step": 2061 }, { "epoch": 0.8827054794520548, "grad_norm": 1.1859803199768066, "learning_rate": 1.1729452054794522e-06, "loss": 0.0009, "step": 2062 }, { "epoch": 0.8831335616438356, "grad_norm": 1.758534550666809, "learning_rate": 1.1686643835616438e-06, "loss": 0.0016, "step": 2063 }, { "epoch": 0.8835616438356164, "grad_norm": 4.854681968688965, "learning_rate": 1.1643835616438357e-06, "loss": 0.0038, "step": 2064 }, { "epoch": 0.8839897260273972, "grad_norm": 1.3547465801239014, "learning_rate": 1.1601027397260275e-06, "loss": 0.0007, "step": 2065 }, { "epoch": 0.884417808219178, "grad_norm": 49.752540588378906, "learning_rate": 1.1558219178082194e-06, "loss": 0.0272, "step": 2066 }, { "epoch": 0.884845890410959, "grad_norm": 2.116443157196045, "learning_rate": 1.151541095890411e-06, "loss": 0.0015, "step": 2067 }, { "epoch": 0.8852739726027398, "grad_norm": 18.7691707611084, "learning_rate": 1.1472602739726028e-06, "loss": 0.0132, "step": 2068 }, { "epoch": 0.8857020547945206, "grad_norm": 2.0166029930114746, "learning_rate": 1.1429794520547947e-06, "loss": 0.0007, "step": 2069 }, { "epoch": 0.8861301369863014, "grad_norm": 4.277080059051514, "learning_rate": 1.1386986301369863e-06, "loss": 0.0022, "step": 2070 }, { "epoch": 0.8865582191780822, "grad_norm": 3.2369654178619385, "learning_rate": 1.1344178082191781e-06, "loss": 0.0012, "step": 2071 }, { "epoch": 0.886986301369863, "grad_norm": 6.578203201293945, "learning_rate": 1.13013698630137e-06, "loss": 0.0038, "step": 2072 }, { "epoch": 0.8874143835616438, "grad_norm": 2.1225099563598633, "learning_rate": 1.1258561643835618e-06, "loss": 0.0022, "step": 2073 }, { "epoch": 0.8878424657534246, "grad_norm": 5.976751327514648, "learning_rate": 1.1215753424657535e-06, "loss": 0.0036, "step": 2074 }, { "epoch": 0.8882705479452054, "grad_norm": 4.461650371551514, "learning_rate": 1.1172945205479453e-06, "loss": 0.0021, "step": 2075 }, { "epoch": 0.8886986301369864, "grad_norm": 160.5789794921875, "learning_rate": 1.1130136986301371e-06, "loss": 0.2539, "step": 2076 }, { "epoch": 0.8891267123287672, "grad_norm": 3.704806089401245, "learning_rate": 1.1087328767123288e-06, "loss": 0.0013, "step": 2077 }, { "epoch": 0.889554794520548, "grad_norm": 3.7217648029327393, "learning_rate": 1.1044520547945206e-06, "loss": 0.0012, "step": 2078 }, { "epoch": 0.8899828767123288, "grad_norm": 1.2007619142532349, "learning_rate": 1.1001712328767125e-06, "loss": 0.0008, "step": 2079 }, { "epoch": 0.8904109589041096, "grad_norm": 0.6495642066001892, "learning_rate": 1.095890410958904e-06, "loss": 0.0005, "step": 2080 }, { "epoch": 0.8908390410958904, "grad_norm": 2.5095865726470947, "learning_rate": 1.0916095890410961e-06, "loss": 0.0009, "step": 2081 }, { "epoch": 0.8912671232876712, "grad_norm": 3.760390520095825, "learning_rate": 1.0873287671232878e-06, "loss": 0.0019, "step": 2082 }, { "epoch": 0.891695205479452, "grad_norm": 2.380357503890991, "learning_rate": 1.0830479452054796e-06, "loss": 0.0015, "step": 2083 }, { "epoch": 0.8921232876712328, "grad_norm": 18.497936248779297, "learning_rate": 1.0787671232876712e-06, "loss": 0.0222, "step": 2084 }, { "epoch": 0.8925513698630136, "grad_norm": 0.20861221849918365, "learning_rate": 1.074486301369863e-06, "loss": 0.0005, "step": 2085 }, { "epoch": 0.8929794520547946, "grad_norm": 2.0229876041412354, "learning_rate": 1.070205479452055e-06, "loss": 0.0011, "step": 2086 }, { "epoch": 0.8934075342465754, "grad_norm": 2.7299346923828125, "learning_rate": 1.0659246575342466e-06, "loss": 0.0009, "step": 2087 }, { "epoch": 0.8938356164383562, "grad_norm": 6.8298020362854, "learning_rate": 1.0616438356164384e-06, "loss": 0.0037, "step": 2088 }, { "epoch": 0.894263698630137, "grad_norm": 2.245934247970581, "learning_rate": 1.0573630136986302e-06, "loss": 0.002, "step": 2089 }, { "epoch": 0.8946917808219178, "grad_norm": 5.956130504608154, "learning_rate": 1.053082191780822e-06, "loss": 0.003, "step": 2090 }, { "epoch": 0.8951198630136986, "grad_norm": 52.28565216064453, "learning_rate": 1.0488013698630137e-06, "loss": 0.5784, "step": 2091 }, { "epoch": 0.8955479452054794, "grad_norm": 4.013579368591309, "learning_rate": 1.0445205479452056e-06, "loss": 0.0014, "step": 2092 }, { "epoch": 0.8959760273972602, "grad_norm": 173.17239379882812, "learning_rate": 1.0402397260273974e-06, "loss": 0.6254, "step": 2093 }, { "epoch": 0.896404109589041, "grad_norm": 1.8486756086349487, "learning_rate": 1.035958904109589e-06, "loss": 0.0006, "step": 2094 }, { "epoch": 0.896832191780822, "grad_norm": 1.5321124792099, "learning_rate": 1.0316780821917809e-06, "loss": 0.0016, "step": 2095 }, { "epoch": 0.8972602739726028, "grad_norm": 6.74153470993042, "learning_rate": 1.0273972602739727e-06, "loss": 0.0037, "step": 2096 }, { "epoch": 0.8976883561643836, "grad_norm": 6.139674663543701, "learning_rate": 1.0231164383561646e-06, "loss": 0.003, "step": 2097 }, { "epoch": 0.8981164383561644, "grad_norm": 3.3353652954101562, "learning_rate": 1.0188356164383564e-06, "loss": 0.0024, "step": 2098 }, { "epoch": 0.8985445205479452, "grad_norm": 2.2239694595336914, "learning_rate": 1.014554794520548e-06, "loss": 0.0021, "step": 2099 }, { "epoch": 0.898972602739726, "grad_norm": 1.0682430267333984, "learning_rate": 1.0102739726027399e-06, "loss": 0.0009, "step": 2100 }, { "epoch": 0.8994006849315068, "grad_norm": 3.796753406524658, "learning_rate": 1.0059931506849315e-06, "loss": 0.0019, "step": 2101 }, { "epoch": 0.8998287671232876, "grad_norm": 68.56819152832031, "learning_rate": 1.0017123287671233e-06, "loss": 0.5706, "step": 2102 }, { "epoch": 0.9002568493150684, "grad_norm": 1.6789535284042358, "learning_rate": 9.974315068493152e-07, "loss": 0.001, "step": 2103 }, { "epoch": 0.9006849315068494, "grad_norm": 5.1463189125061035, "learning_rate": 9.931506849315068e-07, "loss": 0.0021, "step": 2104 }, { "epoch": 0.9011130136986302, "grad_norm": 1.6061432361602783, "learning_rate": 9.888698630136989e-07, "loss": 0.0005, "step": 2105 }, { "epoch": 0.901541095890411, "grad_norm": 4.261180877685547, "learning_rate": 9.845890410958905e-07, "loss": 0.002, "step": 2106 }, { "epoch": 0.9019691780821918, "grad_norm": 6.1412458419799805, "learning_rate": 9.803082191780823e-07, "loss": 0.0045, "step": 2107 }, { "epoch": 0.9023972602739726, "grad_norm": 1.8362950086593628, "learning_rate": 9.76027397260274e-07, "loss": 0.0008, "step": 2108 }, { "epoch": 0.9028253424657534, "grad_norm": 1.3357676267623901, "learning_rate": 9.717465753424658e-07, "loss": 0.0007, "step": 2109 }, { "epoch": 0.9032534246575342, "grad_norm": 4.518088340759277, "learning_rate": 9.674657534246577e-07, "loss": 0.0035, "step": 2110 }, { "epoch": 0.903681506849315, "grad_norm": 0.33314403891563416, "learning_rate": 9.631849315068493e-07, "loss": 0.0005, "step": 2111 }, { "epoch": 0.9041095890410958, "grad_norm": 0.23400847613811493, "learning_rate": 9.589041095890411e-07, "loss": 0.0004, "step": 2112 }, { "epoch": 0.9045376712328768, "grad_norm": 1.2830311059951782, "learning_rate": 9.54623287671233e-07, "loss": 0.001, "step": 2113 }, { "epoch": 0.9049657534246576, "grad_norm": 4.625979423522949, "learning_rate": 9.503424657534248e-07, "loss": 0.0013, "step": 2114 }, { "epoch": 0.9053938356164384, "grad_norm": 0.7540076971054077, "learning_rate": 9.460616438356166e-07, "loss": 0.0005, "step": 2115 }, { "epoch": 0.9058219178082192, "grad_norm": 1.7743487358093262, "learning_rate": 9.417808219178083e-07, "loss": 0.0006, "step": 2116 }, { "epoch": 0.90625, "grad_norm": 1.436619758605957, "learning_rate": 9.375000000000001e-07, "loss": 0.0007, "step": 2117 }, { "epoch": 0.9066780821917808, "grad_norm": 171.29905700683594, "learning_rate": 9.332191780821919e-07, "loss": 0.2775, "step": 2118 }, { "epoch": 0.9071061643835616, "grad_norm": 2.8127684593200684, "learning_rate": 9.289383561643836e-07, "loss": 0.0022, "step": 2119 }, { "epoch": 0.9075342465753424, "grad_norm": 10.977072715759277, "learning_rate": 9.246575342465753e-07, "loss": 0.0098, "step": 2120 }, { "epoch": 0.9079623287671232, "grad_norm": 1.251521110534668, "learning_rate": 9.203767123287672e-07, "loss": 0.0004, "step": 2121 }, { "epoch": 0.9083904109589042, "grad_norm": 3.8164424896240234, "learning_rate": 9.16095890410959e-07, "loss": 0.0019, "step": 2122 }, { "epoch": 0.908818493150685, "grad_norm": 6.239194393157959, "learning_rate": 9.118150684931508e-07, "loss": 0.0029, "step": 2123 }, { "epoch": 0.9092465753424658, "grad_norm": 1.9325757026672363, "learning_rate": 9.075342465753426e-07, "loss": 0.0013, "step": 2124 }, { "epoch": 0.9096746575342466, "grad_norm": 3.3966634273529053, "learning_rate": 9.032534246575343e-07, "loss": 0.0011, "step": 2125 }, { "epoch": 0.9101027397260274, "grad_norm": 39.9615478515625, "learning_rate": 8.989726027397261e-07, "loss": 0.013, "step": 2126 }, { "epoch": 0.9105308219178082, "grad_norm": 2.6320738792419434, "learning_rate": 8.946917808219178e-07, "loss": 0.0047, "step": 2127 }, { "epoch": 0.910958904109589, "grad_norm": 1.2247581481933594, "learning_rate": 8.904109589041097e-07, "loss": 0.0006, "step": 2128 }, { "epoch": 0.9113869863013698, "grad_norm": 0.21572835743427277, "learning_rate": 8.861301369863014e-07, "loss": 0.0004, "step": 2129 }, { "epoch": 0.9118150684931506, "grad_norm": 1.1117615699768066, "learning_rate": 8.818493150684932e-07, "loss": 0.0006, "step": 2130 }, { "epoch": 0.9122431506849316, "grad_norm": 10.621416091918945, "learning_rate": 8.775684931506851e-07, "loss": 0.0178, "step": 2131 }, { "epoch": 0.9126712328767124, "grad_norm": 3.5150797367095947, "learning_rate": 8.732876712328768e-07, "loss": 0.0017, "step": 2132 }, { "epoch": 0.9130993150684932, "grad_norm": 46.861759185791016, "learning_rate": 8.690068493150685e-07, "loss": 0.0339, "step": 2133 }, { "epoch": 0.913527397260274, "grad_norm": 2.690537929534912, "learning_rate": 8.647260273972603e-07, "loss": 0.0009, "step": 2134 }, { "epoch": 0.9139554794520548, "grad_norm": 4.235171318054199, "learning_rate": 8.604452054794521e-07, "loss": 0.0019, "step": 2135 }, { "epoch": 0.9143835616438356, "grad_norm": 7.5841965675354, "learning_rate": 8.561643835616439e-07, "loss": 0.0049, "step": 2136 }, { "epoch": 0.9148116438356164, "grad_norm": 3.8783857822418213, "learning_rate": 8.518835616438356e-07, "loss": 0.0027, "step": 2137 }, { "epoch": 0.9152397260273972, "grad_norm": 5.373304843902588, "learning_rate": 8.476027397260275e-07, "loss": 0.0028, "step": 2138 }, { "epoch": 0.915667808219178, "grad_norm": 0.6711202263832092, "learning_rate": 8.433219178082193e-07, "loss": 0.0006, "step": 2139 }, { "epoch": 0.916095890410959, "grad_norm": 3.2580244541168213, "learning_rate": 8.39041095890411e-07, "loss": 0.0032, "step": 2140 }, { "epoch": 0.9165239726027398, "grad_norm": 0.9540608525276184, "learning_rate": 8.347602739726029e-07, "loss": 0.0003, "step": 2141 }, { "epoch": 0.9169520547945206, "grad_norm": 1.7682204246520996, "learning_rate": 8.304794520547946e-07, "loss": 0.0006, "step": 2142 }, { "epoch": 0.9173801369863014, "grad_norm": 236.947265625, "learning_rate": 8.261986301369863e-07, "loss": 0.2029, "step": 2143 }, { "epoch": 0.9178082191780822, "grad_norm": 1.1837639808654785, "learning_rate": 8.219178082191781e-07, "loss": 0.0008, "step": 2144 }, { "epoch": 0.918236301369863, "grad_norm": 2.628296375274658, "learning_rate": 8.176369863013699e-07, "loss": 0.0008, "step": 2145 }, { "epoch": 0.9186643835616438, "grad_norm": 0.5763463377952576, "learning_rate": 8.133561643835618e-07, "loss": 0.0004, "step": 2146 }, { "epoch": 0.9190924657534246, "grad_norm": 1.7685126066207886, "learning_rate": 8.090753424657535e-07, "loss": 0.0005, "step": 2147 }, { "epoch": 0.9195205479452054, "grad_norm": 10.738037109375, "learning_rate": 8.047945205479453e-07, "loss": 0.005, "step": 2148 }, { "epoch": 0.9199486301369864, "grad_norm": 1.3255845308303833, "learning_rate": 8.005136986301371e-07, "loss": 0.0008, "step": 2149 }, { "epoch": 0.9203767123287672, "grad_norm": 2.0541303157806396, "learning_rate": 7.962328767123288e-07, "loss": 0.0006, "step": 2150 }, { "epoch": 0.920804794520548, "grad_norm": 0.8713811039924622, "learning_rate": 7.919520547945205e-07, "loss": 0.0007, "step": 2151 }, { "epoch": 0.9212328767123288, "grad_norm": 101.23865509033203, "learning_rate": 7.876712328767124e-07, "loss": 0.372, "step": 2152 }, { "epoch": 0.9216609589041096, "grad_norm": 1.3019075393676758, "learning_rate": 7.833904109589041e-07, "loss": 0.0008, "step": 2153 }, { "epoch": 0.9220890410958904, "grad_norm": 2.518451690673828, "learning_rate": 7.791095890410961e-07, "loss": 0.0008, "step": 2154 }, { "epoch": 0.9225171232876712, "grad_norm": 1.203059434890747, "learning_rate": 7.748287671232878e-07, "loss": 0.0005, "step": 2155 }, { "epoch": 0.922945205479452, "grad_norm": 5.986830711364746, "learning_rate": 7.705479452054795e-07, "loss": 0.006, "step": 2156 }, { "epoch": 0.9233732876712328, "grad_norm": 9.612332344055176, "learning_rate": 7.662671232876713e-07, "loss": 0.0052, "step": 2157 }, { "epoch": 0.9238013698630136, "grad_norm": 6.612191677093506, "learning_rate": 7.619863013698631e-07, "loss": 0.0032, "step": 2158 }, { "epoch": 0.9242294520547946, "grad_norm": 1.5820657014846802, "learning_rate": 7.577054794520549e-07, "loss": 0.0005, "step": 2159 }, { "epoch": 0.9246575342465754, "grad_norm": 2.6876883506774902, "learning_rate": 7.534246575342466e-07, "loss": 0.0009, "step": 2160 }, { "epoch": 0.9250856164383562, "grad_norm": 4.259371280670166, "learning_rate": 7.491438356164383e-07, "loss": 0.0016, "step": 2161 }, { "epoch": 0.925513698630137, "grad_norm": 2.759366273880005, "learning_rate": 7.448630136986303e-07, "loss": 0.0008, "step": 2162 }, { "epoch": 0.9259417808219178, "grad_norm": 4.674362659454346, "learning_rate": 7.40582191780822e-07, "loss": 0.0024, "step": 2163 }, { "epoch": 0.9263698630136986, "grad_norm": 4.183143138885498, "learning_rate": 7.363013698630137e-07, "loss": 0.0015, "step": 2164 }, { "epoch": 0.9267979452054794, "grad_norm": 1.854106068611145, "learning_rate": 7.320205479452056e-07, "loss": 0.0018, "step": 2165 }, { "epoch": 0.9272260273972602, "grad_norm": 0.5782306790351868, "learning_rate": 7.277397260273973e-07, "loss": 0.0003, "step": 2166 }, { "epoch": 0.927654109589041, "grad_norm": 3.3071281909942627, "learning_rate": 7.234589041095891e-07, "loss": 0.0014, "step": 2167 }, { "epoch": 0.928082191780822, "grad_norm": 5.577425479888916, "learning_rate": 7.191780821917808e-07, "loss": 0.0026, "step": 2168 }, { "epoch": 0.9285102739726028, "grad_norm": 0.5914297103881836, "learning_rate": 7.148972602739726e-07, "loss": 0.0008, "step": 2169 }, { "epoch": 0.9289383561643836, "grad_norm": 3.071762800216675, "learning_rate": 7.106164383561645e-07, "loss": 0.0009, "step": 2170 }, { "epoch": 0.9293664383561644, "grad_norm": 1.2038681507110596, "learning_rate": 7.063356164383563e-07, "loss": 0.0008, "step": 2171 }, { "epoch": 0.9297945205479452, "grad_norm": 0.6348571181297302, "learning_rate": 7.020547945205481e-07, "loss": 0.0004, "step": 2172 }, { "epoch": 0.930222602739726, "grad_norm": 0.9480258226394653, "learning_rate": 6.977739726027398e-07, "loss": 0.001, "step": 2173 }, { "epoch": 0.9306506849315068, "grad_norm": 60.0487060546875, "learning_rate": 6.934931506849315e-07, "loss": 0.6799, "step": 2174 }, { "epoch": 0.9310787671232876, "grad_norm": 0.6891542673110962, "learning_rate": 6.892123287671234e-07, "loss": 0.0004, "step": 2175 }, { "epoch": 0.9315068493150684, "grad_norm": 10.98681354522705, "learning_rate": 6.849315068493151e-07, "loss": 0.0093, "step": 2176 }, { "epoch": 0.9319349315068494, "grad_norm": 2.563288688659668, "learning_rate": 6.806506849315068e-07, "loss": 0.0008, "step": 2177 }, { "epoch": 0.9323630136986302, "grad_norm": 1.0155168771743774, "learning_rate": 6.763698630136988e-07, "loss": 0.0015, "step": 2178 }, { "epoch": 0.932791095890411, "grad_norm": 3.3522400856018066, "learning_rate": 6.720890410958905e-07, "loss": 0.001, "step": 2179 }, { "epoch": 0.9332191780821918, "grad_norm": 8.761086463928223, "learning_rate": 6.678082191780823e-07, "loss": 0.0069, "step": 2180 }, { "epoch": 0.9336472602739726, "grad_norm": 0.7912534475326538, "learning_rate": 6.63527397260274e-07, "loss": 0.001, "step": 2181 }, { "epoch": 0.9340753424657534, "grad_norm": 3.87899112701416, "learning_rate": 6.592465753424658e-07, "loss": 0.0014, "step": 2182 }, { "epoch": 0.9345034246575342, "grad_norm": 2.7563817501068115, "learning_rate": 6.549657534246576e-07, "loss": 0.0008, "step": 2183 }, { "epoch": 0.934931506849315, "grad_norm": 7.158295631408691, "learning_rate": 6.506849315068493e-07, "loss": 0.0042, "step": 2184 }, { "epoch": 0.9353595890410958, "grad_norm": 0.9054219126701355, "learning_rate": 6.46404109589041e-07, "loss": 0.0005, "step": 2185 }, { "epoch": 0.9357876712328768, "grad_norm": 3.250142812728882, "learning_rate": 6.42123287671233e-07, "loss": 0.0031, "step": 2186 }, { "epoch": 0.9362157534246576, "grad_norm": 1.438173532485962, "learning_rate": 6.378424657534247e-07, "loss": 0.0008, "step": 2187 }, { "epoch": 0.9366438356164384, "grad_norm": 55.10334396362305, "learning_rate": 6.335616438356165e-07, "loss": 0.6138, "step": 2188 }, { "epoch": 0.9370719178082192, "grad_norm": 7.535523414611816, "learning_rate": 6.292808219178083e-07, "loss": 0.0043, "step": 2189 }, { "epoch": 0.9375, "grad_norm": 1.079433798789978, "learning_rate": 6.25e-07, "loss": 0.0006, "step": 2190 }, { "epoch": 0.9379280821917808, "grad_norm": 4.944845199584961, "learning_rate": 6.207191780821918e-07, "loss": 0.0051, "step": 2191 }, { "epoch": 0.9383561643835616, "grad_norm": 3.7726073265075684, "learning_rate": 6.164383561643836e-07, "loss": 0.001, "step": 2192 }, { "epoch": 0.9387842465753424, "grad_norm": 0.8420928120613098, "learning_rate": 6.121575342465754e-07, "loss": 0.0008, "step": 2193 }, { "epoch": 0.9392123287671232, "grad_norm": 2.3209431171417236, "learning_rate": 6.078767123287672e-07, "loss": 0.0008, "step": 2194 }, { "epoch": 0.9396404109589042, "grad_norm": 0.5438706278800964, "learning_rate": 6.035958904109589e-07, "loss": 0.0004, "step": 2195 }, { "epoch": 0.940068493150685, "grad_norm": 22.072486877441406, "learning_rate": 5.993150684931507e-07, "loss": 0.014, "step": 2196 }, { "epoch": 0.9404965753424658, "grad_norm": 1.1208808422088623, "learning_rate": 5.950342465753425e-07, "loss": 0.0004, "step": 2197 }, { "epoch": 0.9409246575342466, "grad_norm": 3.6858668327331543, "learning_rate": 5.907534246575343e-07, "loss": 0.0018, "step": 2198 }, { "epoch": 0.9413527397260274, "grad_norm": 146.7224884033203, "learning_rate": 5.864726027397261e-07, "loss": 0.115, "step": 2199 }, { "epoch": 0.9417808219178082, "grad_norm": 0.844282329082489, "learning_rate": 5.821917808219178e-07, "loss": 0.0006, "step": 2200 }, { "epoch": 0.942208904109589, "grad_norm": 1.0182464122772217, "learning_rate": 5.779109589041097e-07, "loss": 0.0004, "step": 2201 }, { "epoch": 0.9426369863013698, "grad_norm": 4.7519001960754395, "learning_rate": 5.736301369863014e-07, "loss": 0.0026, "step": 2202 }, { "epoch": 0.9430650684931506, "grad_norm": 3.961663246154785, "learning_rate": 5.693493150684932e-07, "loss": 0.0019, "step": 2203 }, { "epoch": 0.9434931506849316, "grad_norm": 6.793483257293701, "learning_rate": 5.65068493150685e-07, "loss": 0.0036, "step": 2204 }, { "epoch": 0.9439212328767124, "grad_norm": 4.63498067855835, "learning_rate": 5.607876712328767e-07, "loss": 0.0018, "step": 2205 }, { "epoch": 0.9443493150684932, "grad_norm": 3.362103223800659, "learning_rate": 5.565068493150686e-07, "loss": 0.0022, "step": 2206 }, { "epoch": 0.944777397260274, "grad_norm": 1.5044094324111938, "learning_rate": 5.522260273972603e-07, "loss": 0.0007, "step": 2207 }, { "epoch": 0.9452054794520548, "grad_norm": 4.100698471069336, "learning_rate": 5.47945205479452e-07, "loss": 0.0017, "step": 2208 }, { "epoch": 0.9456335616438356, "grad_norm": 55.78254699707031, "learning_rate": 5.436643835616439e-07, "loss": 0.5785, "step": 2209 }, { "epoch": 0.9460616438356164, "grad_norm": 1.108717918395996, "learning_rate": 5.393835616438356e-07, "loss": 0.0004, "step": 2210 }, { "epoch": 0.9464897260273972, "grad_norm": 7.246195316314697, "learning_rate": 5.351027397260275e-07, "loss": 0.0043, "step": 2211 }, { "epoch": 0.946917808219178, "grad_norm": 59.134117126464844, "learning_rate": 5.308219178082192e-07, "loss": 0.6449, "step": 2212 }, { "epoch": 0.947345890410959, "grad_norm": 1.33339524269104, "learning_rate": 5.26541095890411e-07, "loss": 0.0004, "step": 2213 }, { "epoch": 0.9477739726027398, "grad_norm": 0.27476435899734497, "learning_rate": 5.222602739726028e-07, "loss": 0.0004, "step": 2214 }, { "epoch": 0.9482020547945206, "grad_norm": 0.43454962968826294, "learning_rate": 5.179794520547945e-07, "loss": 0.0005, "step": 2215 }, { "epoch": 0.9486301369863014, "grad_norm": 3.919314384460449, "learning_rate": 5.136986301369864e-07, "loss": 0.0025, "step": 2216 }, { "epoch": 0.9490582191780822, "grad_norm": 8.685298919677734, "learning_rate": 5.094178082191782e-07, "loss": 0.0058, "step": 2217 }, { "epoch": 0.949486301369863, "grad_norm": 1.003472089767456, "learning_rate": 5.051369863013699e-07, "loss": 0.0009, "step": 2218 }, { "epoch": 0.9499143835616438, "grad_norm": 5.846353530883789, "learning_rate": 5.008561643835617e-07, "loss": 0.0033, "step": 2219 }, { "epoch": 0.9503424657534246, "grad_norm": 1.119820237159729, "learning_rate": 4.965753424657534e-07, "loss": 0.0003, "step": 2220 }, { "epoch": 0.9507705479452054, "grad_norm": 3.146143674850464, "learning_rate": 4.922945205479453e-07, "loss": 0.0014, "step": 2221 }, { "epoch": 0.9511986301369864, "grad_norm": 62.02906799316406, "learning_rate": 4.88013698630137e-07, "loss": 0.5645, "step": 2222 }, { "epoch": 0.9516267123287672, "grad_norm": 6.412882328033447, "learning_rate": 4.837328767123288e-07, "loss": 0.0031, "step": 2223 }, { "epoch": 0.952054794520548, "grad_norm": 3.3252434730529785, "learning_rate": 4.794520547945206e-07, "loss": 0.0016, "step": 2224 }, { "epoch": 0.9524828767123288, "grad_norm": 2.6210439205169678, "learning_rate": 4.751712328767124e-07, "loss": 0.001, "step": 2225 }, { "epoch": 0.9529109589041096, "grad_norm": 64.07918548583984, "learning_rate": 4.7089041095890414e-07, "loss": 0.0276, "step": 2226 }, { "epoch": 0.9533390410958904, "grad_norm": 0.9277715682983398, "learning_rate": 4.6660958904109593e-07, "loss": 0.0004, "step": 2227 }, { "epoch": 0.9537671232876712, "grad_norm": 7.523828506469727, "learning_rate": 4.6232876712328767e-07, "loss": 0.0077, "step": 2228 }, { "epoch": 0.954195205479452, "grad_norm": 1.6990840435028076, "learning_rate": 4.580479452054795e-07, "loss": 0.0012, "step": 2229 }, { "epoch": 0.9546232876712328, "grad_norm": 2.3483026027679443, "learning_rate": 4.537671232876713e-07, "loss": 0.0018, "step": 2230 }, { "epoch": 0.9550513698630136, "grad_norm": 8.03554630279541, "learning_rate": 4.4948630136986304e-07, "loss": 0.0057, "step": 2231 }, { "epoch": 0.9554794520547946, "grad_norm": 3.5570013523101807, "learning_rate": 4.452054794520548e-07, "loss": 0.0019, "step": 2232 }, { "epoch": 0.9559075342465754, "grad_norm": 2.3923583030700684, "learning_rate": 4.409246575342466e-07, "loss": 0.0013, "step": 2233 }, { "epoch": 0.9563356164383562, "grad_norm": 9.04458236694336, "learning_rate": 4.366438356164384e-07, "loss": 0.0041, "step": 2234 }, { "epoch": 0.956763698630137, "grad_norm": 3.603900909423828, "learning_rate": 4.3236301369863014e-07, "loss": 0.0013, "step": 2235 }, { "epoch": 0.9571917808219178, "grad_norm": 4.102055549621582, "learning_rate": 4.2808219178082193e-07, "loss": 0.0019, "step": 2236 }, { "epoch": 0.9576198630136986, "grad_norm": 4.244271278381348, "learning_rate": 4.2380136986301377e-07, "loss": 0.0033, "step": 2237 }, { "epoch": 0.9580479452054794, "grad_norm": 0.957492470741272, "learning_rate": 4.195205479452055e-07, "loss": 0.0003, "step": 2238 }, { "epoch": 0.9584760273972602, "grad_norm": 4.181631088256836, "learning_rate": 4.152397260273973e-07, "loss": 0.006, "step": 2239 }, { "epoch": 0.958904109589041, "grad_norm": 0.2941834628582001, "learning_rate": 4.1095890410958903e-07, "loss": 0.0003, "step": 2240 }, { "epoch": 0.959332191780822, "grad_norm": 1.1354862451553345, "learning_rate": 4.066780821917809e-07, "loss": 0.0004, "step": 2241 }, { "epoch": 0.9597602739726028, "grad_norm": 3.422067165374756, "learning_rate": 4.0239726027397266e-07, "loss": 0.0013, "step": 2242 }, { "epoch": 0.9601883561643836, "grad_norm": 11.468897819519043, "learning_rate": 3.981164383561644e-07, "loss": 0.0086, "step": 2243 }, { "epoch": 0.9606164383561644, "grad_norm": 1.8269472122192383, "learning_rate": 3.938356164383562e-07, "loss": 0.0006, "step": 2244 }, { "epoch": 0.9610445205479452, "grad_norm": 1.5990266799926758, "learning_rate": 3.8955479452054803e-07, "loss": 0.0005, "step": 2245 }, { "epoch": 0.961472602739726, "grad_norm": 0.2970937490463257, "learning_rate": 3.8527397260273977e-07, "loss": 0.0003, "step": 2246 }, { "epoch": 0.9619006849315068, "grad_norm": 3.9892938137054443, "learning_rate": 3.8099315068493156e-07, "loss": 0.0056, "step": 2247 }, { "epoch": 0.9623287671232876, "grad_norm": 0.20469678938388824, "learning_rate": 3.767123287671233e-07, "loss": 0.0003, "step": 2248 }, { "epoch": 0.9627568493150684, "grad_norm": 1.298866629600525, "learning_rate": 3.7243150684931514e-07, "loss": 0.0003, "step": 2249 }, { "epoch": 0.9631849315068494, "grad_norm": 0.7789490818977356, "learning_rate": 3.6815068493150687e-07, "loss": 0.0012, "step": 2250 }, { "epoch": 0.9636130136986302, "grad_norm": 130.36819458007812, "learning_rate": 3.6386986301369866e-07, "loss": 0.0984, "step": 2251 }, { "epoch": 0.964041095890411, "grad_norm": 3.3625242710113525, "learning_rate": 3.595890410958904e-07, "loss": 0.0017, "step": 2252 }, { "epoch": 0.9644691780821918, "grad_norm": 0.2082357257604599, "learning_rate": 3.5530821917808224e-07, "loss": 0.0004, "step": 2253 }, { "epoch": 0.9648972602739726, "grad_norm": 0.8227162957191467, "learning_rate": 3.5102739726027403e-07, "loss": 0.0005, "step": 2254 }, { "epoch": 0.9653253424657534, "grad_norm": 2.9783785343170166, "learning_rate": 3.4674657534246577e-07, "loss": 0.0009, "step": 2255 }, { "epoch": 0.9657534246575342, "grad_norm": 0.6048452258110046, "learning_rate": 3.4246575342465755e-07, "loss": 0.0005, "step": 2256 }, { "epoch": 0.966181506849315, "grad_norm": 9.97964859008789, "learning_rate": 3.381849315068494e-07, "loss": 0.0071, "step": 2257 }, { "epoch": 0.9666095890410958, "grad_norm": 4.847930908203125, "learning_rate": 3.3390410958904113e-07, "loss": 0.0021, "step": 2258 }, { "epoch": 0.9670376712328768, "grad_norm": 14.008380889892578, "learning_rate": 3.296232876712329e-07, "loss": 0.0134, "step": 2259 }, { "epoch": 0.9674657534246576, "grad_norm": 2.425384521484375, "learning_rate": 3.2534246575342466e-07, "loss": 0.0017, "step": 2260 }, { "epoch": 0.9678938356164384, "grad_norm": 89.8054428100586, "learning_rate": 3.210616438356165e-07, "loss": 0.4965, "step": 2261 }, { "epoch": 0.9683219178082192, "grad_norm": 5.169912815093994, "learning_rate": 3.1678082191780824e-07, "loss": 0.0023, "step": 2262 }, { "epoch": 0.96875, "grad_norm": 1.178281307220459, "learning_rate": 3.125e-07, "loss": 0.002, "step": 2263 }, { "epoch": 0.9691780821917808, "grad_norm": 43.918521881103516, "learning_rate": 3.082191780821918e-07, "loss": 0.0177, "step": 2264 }, { "epoch": 0.9696061643835616, "grad_norm": 0.9219381213188171, "learning_rate": 3.039383561643836e-07, "loss": 0.0003, "step": 2265 }, { "epoch": 0.9700342465753424, "grad_norm": 0.992885947227478, "learning_rate": 2.9965753424657534e-07, "loss": 0.0006, "step": 2266 }, { "epoch": 0.9704623287671232, "grad_norm": 67.21822357177734, "learning_rate": 2.9537671232876713e-07, "loss": 0.5164, "step": 2267 }, { "epoch": 0.9708904109589042, "grad_norm": 2.992672920227051, "learning_rate": 2.910958904109589e-07, "loss": 0.0015, "step": 2268 }, { "epoch": 0.971318493150685, "grad_norm": 2.119882106781006, "learning_rate": 2.868150684931507e-07, "loss": 0.0017, "step": 2269 }, { "epoch": 0.9717465753424658, "grad_norm": 0.6352967619895935, "learning_rate": 2.825342465753425e-07, "loss": 0.0005, "step": 2270 }, { "epoch": 0.9721746575342466, "grad_norm": 6.107473373413086, "learning_rate": 2.782534246575343e-07, "loss": 0.0043, "step": 2271 }, { "epoch": 0.9726027397260274, "grad_norm": 3.6329355239868164, "learning_rate": 2.73972602739726e-07, "loss": 0.0038, "step": 2272 }, { "epoch": 0.9730308219178082, "grad_norm": 0.30696433782577515, "learning_rate": 2.696917808219178e-07, "loss": 0.0004, "step": 2273 }, { "epoch": 0.973458904109589, "grad_norm": 7.601875305175781, "learning_rate": 2.654109589041096e-07, "loss": 0.0082, "step": 2274 }, { "epoch": 0.9738869863013698, "grad_norm": 2.8004698753356934, "learning_rate": 2.611301369863014e-07, "loss": 0.001, "step": 2275 }, { "epoch": 0.9743150684931506, "grad_norm": 4.48164176940918, "learning_rate": 2.568493150684932e-07, "loss": 0.0016, "step": 2276 }, { "epoch": 0.9747431506849316, "grad_norm": 1.5099890232086182, "learning_rate": 2.5256849315068497e-07, "loss": 0.0015, "step": 2277 }, { "epoch": 0.9751712328767124, "grad_norm": 69.79464721679688, "learning_rate": 2.482876712328767e-07, "loss": 0.6654, "step": 2278 }, { "epoch": 0.9755993150684932, "grad_norm": 1.4544225931167603, "learning_rate": 2.440068493150685e-07, "loss": 0.0009, "step": 2279 }, { "epoch": 0.976027397260274, "grad_norm": 2.370265245437622, "learning_rate": 2.397260273972603e-07, "loss": 0.0023, "step": 2280 }, { "epoch": 0.9764554794520548, "grad_norm": 15.790185928344727, "learning_rate": 2.3544520547945207e-07, "loss": 0.0114, "step": 2281 }, { "epoch": 0.9768835616438356, "grad_norm": 93.4534683227539, "learning_rate": 2.3116438356164383e-07, "loss": 0.4702, "step": 2282 }, { "epoch": 0.9773116438356164, "grad_norm": 1.7658931016921997, "learning_rate": 2.2688356164383565e-07, "loss": 0.0017, "step": 2283 }, { "epoch": 0.9777397260273972, "grad_norm": 1.6760776042938232, "learning_rate": 2.226027397260274e-07, "loss": 0.0005, "step": 2284 }, { "epoch": 0.978167808219178, "grad_norm": 196.32626342773438, "learning_rate": 2.183219178082192e-07, "loss": 0.2546, "step": 2285 }, { "epoch": 0.978595890410959, "grad_norm": 0.46252232789993286, "learning_rate": 2.1404109589041096e-07, "loss": 0.0007, "step": 2286 }, { "epoch": 0.9790239726027398, "grad_norm": 1.1673928499221802, "learning_rate": 2.0976027397260275e-07, "loss": 0.0006, "step": 2287 }, { "epoch": 0.9794520547945206, "grad_norm": 0.6844284534454346, "learning_rate": 2.0547945205479452e-07, "loss": 0.0004, "step": 2288 }, { "epoch": 0.9798801369863014, "grad_norm": 0.6158546209335327, "learning_rate": 2.0119863013698633e-07, "loss": 0.0007, "step": 2289 }, { "epoch": 0.9803082191780822, "grad_norm": 4.5730085372924805, "learning_rate": 1.969178082191781e-07, "loss": 0.0023, "step": 2290 }, { "epoch": 0.980736301369863, "grad_norm": 11.496292114257812, "learning_rate": 1.9263698630136988e-07, "loss": 0.0067, "step": 2291 }, { "epoch": 0.9811643835616438, "grad_norm": 0.6309894323348999, "learning_rate": 1.8835616438356165e-07, "loss": 0.0005, "step": 2292 }, { "epoch": 0.9815924657534246, "grad_norm": 2.358217716217041, "learning_rate": 1.8407534246575344e-07, "loss": 0.0012, "step": 2293 }, { "epoch": 0.9820205479452054, "grad_norm": 0.7463250756263733, "learning_rate": 1.797945205479452e-07, "loss": 0.0003, "step": 2294 }, { "epoch": 0.9824486301369864, "grad_norm": 6.301076412200928, "learning_rate": 1.7551369863013701e-07, "loss": 0.0033, "step": 2295 }, { "epoch": 0.9828767123287672, "grad_norm": 2.804046392440796, "learning_rate": 1.7123287671232878e-07, "loss": 0.0009, "step": 2296 }, { "epoch": 0.983304794520548, "grad_norm": 1.2526475191116333, "learning_rate": 1.6695205479452057e-07, "loss": 0.0005, "step": 2297 }, { "epoch": 0.9837328767123288, "grad_norm": 5.170557498931885, "learning_rate": 1.6267123287671233e-07, "loss": 0.0042, "step": 2298 }, { "epoch": 0.9841609589041096, "grad_norm": 13.225316047668457, "learning_rate": 1.5839041095890412e-07, "loss": 0.0096, "step": 2299 }, { "epoch": 0.9845890410958904, "grad_norm": 0.5702086687088013, "learning_rate": 1.541095890410959e-07, "loss": 0.0004, "step": 2300 }, { "epoch": 0.9850171232876712, "grad_norm": 0.7707123756408691, "learning_rate": 1.4982876712328767e-07, "loss": 0.0011, "step": 2301 }, { "epoch": 0.985445205479452, "grad_norm": 3.0253913402557373, "learning_rate": 1.4554794520547946e-07, "loss": 0.0009, "step": 2302 }, { "epoch": 0.9858732876712328, "grad_norm": 0.32685351371765137, "learning_rate": 1.4126712328767125e-07, "loss": 0.0005, "step": 2303 }, { "epoch": 0.9863013698630136, "grad_norm": 1.468679428100586, "learning_rate": 1.36986301369863e-07, "loss": 0.0004, "step": 2304 }, { "epoch": 0.9867294520547946, "grad_norm": 0.34716132283210754, "learning_rate": 1.327054794520548e-07, "loss": 0.0004, "step": 2305 }, { "epoch": 0.9871575342465754, "grad_norm": 1.3313649892807007, "learning_rate": 1.284246575342466e-07, "loss": 0.0011, "step": 2306 }, { "epoch": 0.9875856164383562, "grad_norm": 2.062138795852661, "learning_rate": 1.2414383561643835e-07, "loss": 0.0011, "step": 2307 }, { "epoch": 0.988013698630137, "grad_norm": 19.259302139282227, "learning_rate": 1.1986301369863014e-07, "loss": 0.0086, "step": 2308 }, { "epoch": 0.9884417808219178, "grad_norm": 6.000552654266357, "learning_rate": 1.1558219178082192e-07, "loss": 0.0029, "step": 2309 }, { "epoch": 0.9888698630136986, "grad_norm": 2.814868688583374, "learning_rate": 1.113013698630137e-07, "loss": 0.0009, "step": 2310 }, { "epoch": 0.9892979452054794, "grad_norm": 1.5825250148773193, "learning_rate": 1.0702054794520548e-07, "loss": 0.0005, "step": 2311 }, { "epoch": 0.9897260273972602, "grad_norm": 3.651094436645508, "learning_rate": 1.0273972602739726e-07, "loss": 0.0013, "step": 2312 }, { "epoch": 0.990154109589041, "grad_norm": 1.9178614616394043, "learning_rate": 9.845890410958905e-08, "loss": 0.0014, "step": 2313 }, { "epoch": 0.990582191780822, "grad_norm": 1.7152886390686035, "learning_rate": 9.417808219178082e-08, "loss": 0.0009, "step": 2314 }, { "epoch": 0.9910102739726028, "grad_norm": 3.395320177078247, "learning_rate": 8.98972602739726e-08, "loss": 0.005, "step": 2315 }, { "epoch": 0.9914383561643836, "grad_norm": 0.5748260021209717, "learning_rate": 8.561643835616439e-08, "loss": 0.0003, "step": 2316 }, { "epoch": 0.9918664383561644, "grad_norm": 1.9731606245040894, "learning_rate": 8.133561643835616e-08, "loss": 0.0021, "step": 2317 }, { "epoch": 0.9922945205479452, "grad_norm": 2.424976348876953, "learning_rate": 7.705479452054795e-08, "loss": 0.0006, "step": 2318 }, { "epoch": 0.992722602739726, "grad_norm": 3.3053479194641113, "learning_rate": 7.277397260273973e-08, "loss": 0.0014, "step": 2319 }, { "epoch": 0.9931506849315068, "grad_norm": 4.19493293762207, "learning_rate": 6.84931506849315e-08, "loss": 0.0026, "step": 2320 }, { "epoch": 0.9935787671232876, "grad_norm": 1.398173451423645, "learning_rate": 6.42123287671233e-08, "loss": 0.0015, "step": 2321 }, { "epoch": 0.9940068493150684, "grad_norm": 0.9288395643234253, "learning_rate": 5.993150684931507e-08, "loss": 0.0003, "step": 2322 }, { "epoch": 0.9944349315068494, "grad_norm": 3.4160728454589844, "learning_rate": 5.565068493150685e-08, "loss": 0.0019, "step": 2323 }, { "epoch": 0.9948630136986302, "grad_norm": 2.460414171218872, "learning_rate": 5.136986301369863e-08, "loss": 0.0008, "step": 2324 }, { "epoch": 0.995291095890411, "grad_norm": 1.3318833112716675, "learning_rate": 4.708904109589041e-08, "loss": 0.0003, "step": 2325 }, { "epoch": 0.9957191780821918, "grad_norm": 2.249501943588257, "learning_rate": 4.2808219178082194e-08, "loss": 0.0007, "step": 2326 }, { "epoch": 0.9961472602739726, "grad_norm": 3.5273773670196533, "learning_rate": 3.852739726027398e-08, "loss": 0.0013, "step": 2327 }, { "epoch": 0.9965753424657534, "grad_norm": 2.1203482151031494, "learning_rate": 3.424657534246575e-08, "loss": 0.0007, "step": 2328 }, { "epoch": 0.9970034246575342, "grad_norm": 261.9087829589844, "learning_rate": 2.9965753424657535e-08, "loss": 0.2892, "step": 2329 }, { "epoch": 0.997431506849315, "grad_norm": 0.42973774671554565, "learning_rate": 2.5684931506849315e-08, "loss": 0.0005, "step": 2330 }, { "epoch": 0.9978595890410958, "grad_norm": 60.9056510925293, "learning_rate": 2.1404109589041097e-08, "loss": 0.6368, "step": 2331 }, { "epoch": 0.9982876712328768, "grad_norm": 1.3128621578216553, "learning_rate": 1.7123287671232876e-08, "loss": 0.0017, "step": 2332 }, { "epoch": 0.9987157534246576, "grad_norm": 0.05756198242306709, "learning_rate": 1.2842465753424657e-08, "loss": 0.0002, "step": 2333 }, { "epoch": 0.9991438356164384, "grad_norm": 1.4713302850723267, "learning_rate": 8.561643835616438e-09, "loss": 0.0006, "step": 2334 }, { "epoch": 0.9995719178082192, "grad_norm": 68.62449645996094, "learning_rate": 4.280821917808219e-09, "loss": 0.0272, "step": 2335 }, { "epoch": 1.0, "grad_norm": 59.60505294799805, "learning_rate": 0.0, "loss": 0.6637, "step": 2336 }, { "epoch": 1.0, "eval_accuracy": 0.776049766718507, "eval_loss": 1.1974841356277466, "eval_runtime": 93.017, "eval_samples_per_second": 6.977, "eval_steps_per_second": 6.977, "step": 2336 } ], "logging_steps": 1, "max_steps": 2336, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 500, "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": 1, "trial_name": null, "trial_params": null }