{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.9998725952350618, "eval_steps": 500, "global_step": 3924, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00025480952987641737, "grad_norm": 218.569091796875, "learning_rate": 9.99745158002039e-06, "loss": 1.7422, "step": 1 }, { "epoch": 0.0005096190597528347, "grad_norm": 129.1477508544922, "learning_rate": 9.994903160040775e-06, "loss": 1.2891, "step": 2 }, { "epoch": 0.0007644285896292522, "grad_norm": 86.88077545166016, "learning_rate": 9.992354740061163e-06, "loss": 0.7168, "step": 3 }, { "epoch": 0.0010192381195056695, "grad_norm": 115.41887664794922, "learning_rate": 9.98980632008155e-06, "loss": 0.8242, "step": 4 }, { "epoch": 0.0012740476493820869, "grad_norm": 120.88542938232422, "learning_rate": 9.987257900101937e-06, "loss": 0.8535, "step": 5 }, { "epoch": 0.0015288571792585043, "grad_norm": 95.70926666259766, "learning_rate": 9.984709480122326e-06, "loss": 0.6094, "step": 6 }, { "epoch": 0.0017836667091349217, "grad_norm": 53.63985824584961, "learning_rate": 9.982161060142713e-06, "loss": 0.397, "step": 7 }, { "epoch": 0.002038476239011339, "grad_norm": 78.90757751464844, "learning_rate": 9.9796126401631e-06, "loss": 0.582, "step": 8 }, { "epoch": 0.0022932857688877564, "grad_norm": 43.16275405883789, "learning_rate": 9.977064220183487e-06, "loss": 0.5654, "step": 9 }, { "epoch": 0.0025480952987641738, "grad_norm": 50.69451904296875, "learning_rate": 9.974515800203874e-06, "loss": 0.6016, "step": 10 }, { "epoch": 0.002802904828640591, "grad_norm": 73.22306823730469, "learning_rate": 9.971967380224262e-06, "loss": 0.5918, "step": 11 }, { "epoch": 0.0030577143585170086, "grad_norm": 55.550899505615234, "learning_rate": 9.96941896024465e-06, "loss": 0.5635, "step": 12 }, { "epoch": 0.003312523888393426, "grad_norm": 89.92225646972656, "learning_rate": 9.966870540265036e-06, "loss": 0.7422, "step": 13 }, { "epoch": 0.0035673334182698435, "grad_norm": 49.9693717956543, "learning_rate": 9.964322120285423e-06, "loss": 0.3818, "step": 14 }, { "epoch": 0.003822142948146261, "grad_norm": 88.36708068847656, "learning_rate": 9.961773700305812e-06, "loss": 0.8906, "step": 15 }, { "epoch": 0.004076952478022678, "grad_norm": 69.35698699951172, "learning_rate": 9.959225280326199e-06, "loss": 0.5254, "step": 16 }, { "epoch": 0.004331762007899096, "grad_norm": 61.6408576965332, "learning_rate": 9.956676860346586e-06, "loss": 0.4736, "step": 17 }, { "epoch": 0.004586571537775513, "grad_norm": 108.65873718261719, "learning_rate": 9.954128440366973e-06, "loss": 1.0293, "step": 18 }, { "epoch": 0.004841381067651931, "grad_norm": 64.16191101074219, "learning_rate": 9.95158002038736e-06, "loss": 0.5498, "step": 19 }, { "epoch": 0.0050961905975283476, "grad_norm": 108.84109497070312, "learning_rate": 9.949031600407748e-06, "loss": 0.6367, "step": 20 }, { "epoch": 0.0053510001274047645, "grad_norm": 83.16543579101562, "learning_rate": 9.946483180428135e-06, "loss": 0.5957, "step": 21 }, { "epoch": 0.005605809657281182, "grad_norm": 94.14505004882812, "learning_rate": 9.943934760448522e-06, "loss": 0.4375, "step": 22 }, { "epoch": 0.005860619187157599, "grad_norm": 191.8399658203125, "learning_rate": 9.941386340468911e-06, "loss": 1.3281, "step": 23 }, { "epoch": 0.006115428717034017, "grad_norm": 123.62604522705078, "learning_rate": 9.938837920489298e-06, "loss": 1.0645, "step": 24 }, { "epoch": 0.006370238246910434, "grad_norm": 183.05865478515625, "learning_rate": 9.936289500509685e-06, "loss": 1.5645, "step": 25 }, { "epoch": 0.006625047776786852, "grad_norm": 136.49156188964844, "learning_rate": 9.933741080530073e-06, "loss": 1.1611, "step": 26 }, { "epoch": 0.006879857306663269, "grad_norm": 53.94658660888672, "learning_rate": 9.931192660550459e-06, "loss": 0.293, "step": 27 }, { "epoch": 0.007134666836539687, "grad_norm": 71.97644805908203, "learning_rate": 9.928644240570847e-06, "loss": 0.7275, "step": 28 }, { "epoch": 0.007389476366416104, "grad_norm": 68.1817855834961, "learning_rate": 9.926095820591234e-06, "loss": 0.6113, "step": 29 }, { "epoch": 0.007644285896292522, "grad_norm": 17.173877716064453, "learning_rate": 9.923547400611621e-06, "loss": 0.1052, "step": 30 }, { "epoch": 0.007899095426168939, "grad_norm": 110.92193603515625, "learning_rate": 9.92099898063201e-06, "loss": 1.0439, "step": 31 }, { "epoch": 0.008153904956045356, "grad_norm": 84.88875579833984, "learning_rate": 9.918450560652397e-06, "loss": 0.5586, "step": 32 }, { "epoch": 0.008408714485921773, "grad_norm": 36.61272430419922, "learning_rate": 9.915902140672784e-06, "loss": 0.2305, "step": 33 }, { "epoch": 0.008663524015798191, "grad_norm": 120.29364013671875, "learning_rate": 9.91335372069317e-06, "loss": 1.3945, "step": 34 }, { "epoch": 0.008918333545674608, "grad_norm": 99.75139617919922, "learning_rate": 9.910805300713558e-06, "loss": 1.0073, "step": 35 }, { "epoch": 0.009173143075551025, "grad_norm": 67.17069244384766, "learning_rate": 9.908256880733946e-06, "loss": 0.5459, "step": 36 }, { "epoch": 0.009427952605427442, "grad_norm": 92.02169036865234, "learning_rate": 9.905708460754333e-06, "loss": 0.8896, "step": 37 }, { "epoch": 0.009682762135303861, "grad_norm": 43.90317916870117, "learning_rate": 9.90316004077472e-06, "loss": 0.3867, "step": 38 }, { "epoch": 0.009937571665180278, "grad_norm": 77.52678680419922, "learning_rate": 9.900611620795107e-06, "loss": 0.8115, "step": 39 }, { "epoch": 0.010192381195056695, "grad_norm": 87.4370346069336, "learning_rate": 9.898063200815496e-06, "loss": 0.8418, "step": 40 }, { "epoch": 0.010447190724933112, "grad_norm": 45.098426818847656, "learning_rate": 9.895514780835883e-06, "loss": 0.3496, "step": 41 }, { "epoch": 0.010702000254809529, "grad_norm": 43.89577102661133, "learning_rate": 9.89296636085627e-06, "loss": 0.3418, "step": 42 }, { "epoch": 0.010956809784685948, "grad_norm": 54.5704345703125, "learning_rate": 9.890417940876658e-06, "loss": 0.4434, "step": 43 }, { "epoch": 0.011211619314562365, "grad_norm": 58.51212692260742, "learning_rate": 9.887869520897044e-06, "loss": 0.6045, "step": 44 }, { "epoch": 0.011466428844438782, "grad_norm": 167.21343994140625, "learning_rate": 9.885321100917432e-06, "loss": 1.502, "step": 45 }, { "epoch": 0.011721238374315199, "grad_norm": 110.8761215209961, "learning_rate": 9.88277268093782e-06, "loss": 0.9268, "step": 46 }, { "epoch": 0.011976047904191617, "grad_norm": 109.35682678222656, "learning_rate": 9.880224260958206e-06, "loss": 1.4199, "step": 47 }, { "epoch": 0.012230857434068034, "grad_norm": 55.90480422973633, "learning_rate": 9.877675840978595e-06, "loss": 0.583, "step": 48 }, { "epoch": 0.012485666963944451, "grad_norm": 121.46697998046875, "learning_rate": 9.875127420998982e-06, "loss": 1.1426, "step": 49 }, { "epoch": 0.012740476493820868, "grad_norm": 68.06258392333984, "learning_rate": 9.872579001019369e-06, "loss": 0.7344, "step": 50 }, { "epoch": 0.012995286023697285, "grad_norm": 23.94645881652832, "learning_rate": 9.870030581039756e-06, "loss": 0.1372, "step": 51 }, { "epoch": 0.013250095553573704, "grad_norm": 119.4050064086914, "learning_rate": 9.867482161060143e-06, "loss": 1.293, "step": 52 }, { "epoch": 0.013504905083450121, "grad_norm": 41.196815490722656, "learning_rate": 9.864933741080531e-06, "loss": 0.2842, "step": 53 }, { "epoch": 0.013759714613326538, "grad_norm": 57.59366226196289, "learning_rate": 9.862385321100918e-06, "loss": 0.3584, "step": 54 }, { "epoch": 0.014014524143202955, "grad_norm": 79.16130828857422, "learning_rate": 9.859836901121305e-06, "loss": 0.5332, "step": 55 }, { "epoch": 0.014269333673079374, "grad_norm": 80.99279022216797, "learning_rate": 9.857288481141692e-06, "loss": 0.8623, "step": 56 }, { "epoch": 0.01452414320295579, "grad_norm": 55.88768005371094, "learning_rate": 9.85474006116208e-06, "loss": 0.3657, "step": 57 }, { "epoch": 0.014778952732832208, "grad_norm": 121.56302642822266, "learning_rate": 9.852191641182468e-06, "loss": 1.1265, "step": 58 }, { "epoch": 0.015033762262708625, "grad_norm": 109.28952026367188, "learning_rate": 9.849643221202855e-06, "loss": 1.0938, "step": 59 }, { "epoch": 0.015288571792585044, "grad_norm": 103.24484252929688, "learning_rate": 9.847094801223243e-06, "loss": 1.1543, "step": 60 }, { "epoch": 0.01554338132246146, "grad_norm": 111.74369812011719, "learning_rate": 9.844546381243629e-06, "loss": 0.9927, "step": 61 }, { "epoch": 0.015798190852337878, "grad_norm": 35.825260162353516, "learning_rate": 9.841997961264017e-06, "loss": 0.3198, "step": 62 }, { "epoch": 0.016053000382214296, "grad_norm": 59.730064392089844, "learning_rate": 9.839449541284404e-06, "loss": 0.3599, "step": 63 }, { "epoch": 0.01630780991209071, "grad_norm": 49.016937255859375, "learning_rate": 9.836901121304791e-06, "loss": 0.4268, "step": 64 }, { "epoch": 0.01656261944196713, "grad_norm": 58.81710433959961, "learning_rate": 9.83435270132518e-06, "loss": 0.5205, "step": 65 }, { "epoch": 0.016817428971843545, "grad_norm": 57.783660888671875, "learning_rate": 9.831804281345567e-06, "loss": 0.4961, "step": 66 }, { "epoch": 0.017072238501719964, "grad_norm": 37.23876190185547, "learning_rate": 9.829255861365954e-06, "loss": 0.3857, "step": 67 }, { "epoch": 0.017327048031596383, "grad_norm": 78.01972961425781, "learning_rate": 9.826707441386342e-06, "loss": 0.8008, "step": 68 }, { "epoch": 0.017581857561472798, "grad_norm": 59.551578521728516, "learning_rate": 9.824159021406728e-06, "loss": 0.6758, "step": 69 }, { "epoch": 0.017836667091349217, "grad_norm": 40.85459899902344, "learning_rate": 9.821610601427116e-06, "loss": 0.4326, "step": 70 }, { "epoch": 0.018091476621225632, "grad_norm": 53.46759033203125, "learning_rate": 9.819062181447503e-06, "loss": 0.3596, "step": 71 }, { "epoch": 0.01834628615110205, "grad_norm": 118.14017486572266, "learning_rate": 9.81651376146789e-06, "loss": 1.3262, "step": 72 }, { "epoch": 0.01860109568097847, "grad_norm": 75.58549499511719, "learning_rate": 9.813965341488279e-06, "loss": 0.7549, "step": 73 }, { "epoch": 0.018855905210854885, "grad_norm": 33.628170013427734, "learning_rate": 9.811416921508666e-06, "loss": 0.1709, "step": 74 }, { "epoch": 0.019110714740731304, "grad_norm": 27.694244384765625, "learning_rate": 9.808868501529053e-06, "loss": 0.2397, "step": 75 }, { "epoch": 0.019365524270607722, "grad_norm": 45.57538986206055, "learning_rate": 9.80632008154944e-06, "loss": 0.6104, "step": 76 }, { "epoch": 0.019620333800484138, "grad_norm": 78.93385314941406, "learning_rate": 9.803771661569828e-06, "loss": 0.6113, "step": 77 }, { "epoch": 0.019875143330360556, "grad_norm": 58.1033821105957, "learning_rate": 9.801223241590215e-06, "loss": 0.8271, "step": 78 }, { "epoch": 0.02012995286023697, "grad_norm": 47.3413200378418, "learning_rate": 9.798674821610602e-06, "loss": 0.5063, "step": 79 }, { "epoch": 0.02038476239011339, "grad_norm": 128.58657836914062, "learning_rate": 9.79612640163099e-06, "loss": 0.9453, "step": 80 }, { "epoch": 0.02063957191998981, "grad_norm": 66.90831756591797, "learning_rate": 9.793577981651376e-06, "loss": 0.585, "step": 81 }, { "epoch": 0.020894381449866224, "grad_norm": 95.4129409790039, "learning_rate": 9.791029561671765e-06, "loss": 1.1875, "step": 82 }, { "epoch": 0.021149190979742643, "grad_norm": 35.35508728027344, "learning_rate": 9.788481141692152e-06, "loss": 0.2031, "step": 83 }, { "epoch": 0.021404000509619058, "grad_norm": 145.17066955566406, "learning_rate": 9.785932721712539e-06, "loss": 2.2344, "step": 84 }, { "epoch": 0.021658810039495477, "grad_norm": 74.26435852050781, "learning_rate": 9.783384301732927e-06, "loss": 0.6157, "step": 85 }, { "epoch": 0.021913619569371896, "grad_norm": 65.70437622070312, "learning_rate": 9.780835881753313e-06, "loss": 0.6006, "step": 86 }, { "epoch": 0.02216842909924831, "grad_norm": 71.99554443359375, "learning_rate": 9.778287461773701e-06, "loss": 0.7285, "step": 87 }, { "epoch": 0.02242323862912473, "grad_norm": 63.99489974975586, "learning_rate": 9.775739041794088e-06, "loss": 0.5898, "step": 88 }, { "epoch": 0.02267804815900115, "grad_norm": 106.69953918457031, "learning_rate": 9.773190621814475e-06, "loss": 1.4849, "step": 89 }, { "epoch": 0.022932857688877564, "grad_norm": 50.570762634277344, "learning_rate": 9.770642201834864e-06, "loss": 0.6328, "step": 90 }, { "epoch": 0.023187667218753982, "grad_norm": 42.83375549316406, "learning_rate": 9.76809378185525e-06, "loss": 0.5078, "step": 91 }, { "epoch": 0.023442476748630398, "grad_norm": 126.10667419433594, "learning_rate": 9.765545361875638e-06, "loss": 2.0312, "step": 92 }, { "epoch": 0.023697286278506816, "grad_norm": 50.28273010253906, "learning_rate": 9.762996941896026e-06, "loss": 0.5488, "step": 93 }, { "epoch": 0.023952095808383235, "grad_norm": 68.12547302246094, "learning_rate": 9.760448521916412e-06, "loss": 0.8506, "step": 94 }, { "epoch": 0.02420690533825965, "grad_norm": 72.77415466308594, "learning_rate": 9.7579001019368e-06, "loss": 1.1387, "step": 95 }, { "epoch": 0.02446171486813607, "grad_norm": 35.81525421142578, "learning_rate": 9.755351681957187e-06, "loss": 0.4004, "step": 96 }, { "epoch": 0.024716524398012484, "grad_norm": 46.31241226196289, "learning_rate": 9.752803261977574e-06, "loss": 0.4463, "step": 97 }, { "epoch": 0.024971333927888903, "grad_norm": 44.04067611694336, "learning_rate": 9.750254841997963e-06, "loss": 0.7773, "step": 98 }, { "epoch": 0.02522614345776532, "grad_norm": 37.98672866821289, "learning_rate": 9.74770642201835e-06, "loss": 0.498, "step": 99 }, { "epoch": 0.025480952987641737, "grad_norm": 48.79045867919922, "learning_rate": 9.745158002038737e-06, "loss": 0.4883, "step": 100 }, { "epoch": 0.025735762517518156, "grad_norm": 38.394229888916016, "learning_rate": 9.742609582059124e-06, "loss": 0.5703, "step": 101 }, { "epoch": 0.02599057204739457, "grad_norm": 94.14883422851562, "learning_rate": 9.740061162079512e-06, "loss": 1.832, "step": 102 }, { "epoch": 0.02624538157727099, "grad_norm": 95.95703887939453, "learning_rate": 9.7375127420999e-06, "loss": 0.9609, "step": 103 }, { "epoch": 0.02650019110714741, "grad_norm": 63.93125915527344, "learning_rate": 9.734964322120286e-06, "loss": 1.1035, "step": 104 }, { "epoch": 0.026755000637023824, "grad_norm": 71.4206314086914, "learning_rate": 9.732415902140673e-06, "loss": 1.1406, "step": 105 }, { "epoch": 0.027009810166900242, "grad_norm": 46.97043228149414, "learning_rate": 9.72986748216106e-06, "loss": 0.5205, "step": 106 }, { "epoch": 0.02726461969677666, "grad_norm": 39.0087890625, "learning_rate": 9.727319062181449e-06, "loss": 0.5312, "step": 107 }, { "epoch": 0.027519429226653076, "grad_norm": 48.1061897277832, "learning_rate": 9.724770642201836e-06, "loss": 0.6074, "step": 108 }, { "epoch": 0.027774238756529495, "grad_norm": 48.12303161621094, "learning_rate": 9.722222222222223e-06, "loss": 0.5625, "step": 109 }, { "epoch": 0.02802904828640591, "grad_norm": 39.040157318115234, "learning_rate": 9.719673802242611e-06, "loss": 0.4946, "step": 110 }, { "epoch": 0.02828385781628233, "grad_norm": 44.025272369384766, "learning_rate": 9.717125382262997e-06, "loss": 0.6035, "step": 111 }, { "epoch": 0.028538667346158748, "grad_norm": 40.44293212890625, "learning_rate": 9.714576962283385e-06, "loss": 0.4707, "step": 112 }, { "epoch": 0.028793476876035163, "grad_norm": 52.926002502441406, "learning_rate": 9.712028542303772e-06, "loss": 0.6221, "step": 113 }, { "epoch": 0.02904828640591158, "grad_norm": 30.92253303527832, "learning_rate": 9.709480122324159e-06, "loss": 0.2705, "step": 114 }, { "epoch": 0.029303095935787997, "grad_norm": 53.222511291503906, "learning_rate": 9.706931702344548e-06, "loss": 0.8535, "step": 115 }, { "epoch": 0.029557905465664416, "grad_norm": 50.605525970458984, "learning_rate": 9.704383282364935e-06, "loss": 0.5115, "step": 116 }, { "epoch": 0.029812714995540834, "grad_norm": 57.34376525878906, "learning_rate": 9.701834862385322e-06, "loss": 0.4795, "step": 117 }, { "epoch": 0.03006752452541725, "grad_norm": 45.4526481628418, "learning_rate": 9.699286442405709e-06, "loss": 0.4517, "step": 118 }, { "epoch": 0.03032233405529367, "grad_norm": 76.78860473632812, "learning_rate": 9.696738022426097e-06, "loss": 1.2793, "step": 119 }, { "epoch": 0.030577143585170087, "grad_norm": 53.16313934326172, "learning_rate": 9.694189602446484e-06, "loss": 0.3926, "step": 120 }, { "epoch": 0.030831953115046502, "grad_norm": 59.53965759277344, "learning_rate": 9.691641182466871e-06, "loss": 0.7793, "step": 121 }, { "epoch": 0.03108676264492292, "grad_norm": 49.699546813964844, "learning_rate": 9.689092762487258e-06, "loss": 0.4819, "step": 122 }, { "epoch": 0.03134157217479934, "grad_norm": 44.4760627746582, "learning_rate": 9.686544342507645e-06, "loss": 0.4121, "step": 123 }, { "epoch": 0.031596381704675755, "grad_norm": 71.62140655517578, "learning_rate": 9.683995922528034e-06, "loss": 0.877, "step": 124 }, { "epoch": 0.03185119123455217, "grad_norm": 67.57754516601562, "learning_rate": 9.68144750254842e-06, "loss": 0.3452, "step": 125 }, { "epoch": 0.03210600076442859, "grad_norm": 53.43225860595703, "learning_rate": 9.678899082568808e-06, "loss": 0.3198, "step": 126 }, { "epoch": 0.03236081029430501, "grad_norm": 37.07188415527344, "learning_rate": 9.676350662589196e-06, "loss": 0.3301, "step": 127 }, { "epoch": 0.03261561982418142, "grad_norm": 68.73414611816406, "learning_rate": 9.673802242609582e-06, "loss": 0.666, "step": 128 }, { "epoch": 0.032870429354057845, "grad_norm": 86.94526672363281, "learning_rate": 9.67125382262997e-06, "loss": 0.7783, "step": 129 }, { "epoch": 0.03312523888393426, "grad_norm": 78.11392974853516, "learning_rate": 9.668705402650357e-06, "loss": 0.6201, "step": 130 }, { "epoch": 0.033380048413810676, "grad_norm": 46.56452941894531, "learning_rate": 9.666156982670744e-06, "loss": 0.2793, "step": 131 }, { "epoch": 0.03363485794368709, "grad_norm": 90.6522445678711, "learning_rate": 9.663608562691133e-06, "loss": 0.9141, "step": 132 }, { "epoch": 0.03388966747356351, "grad_norm": 64.54054260253906, "learning_rate": 9.66106014271152e-06, "loss": 0.5332, "step": 133 }, { "epoch": 0.03414447700343993, "grad_norm": 78.5429916381836, "learning_rate": 9.658511722731907e-06, "loss": 0.6768, "step": 134 }, { "epoch": 0.034399286533316344, "grad_norm": 43.193302154541016, "learning_rate": 9.655963302752295e-06, "loss": 0.3857, "step": 135 }, { "epoch": 0.034654096063192766, "grad_norm": 147.53749084472656, "learning_rate": 9.653414882772682e-06, "loss": 1.5034, "step": 136 }, { "epoch": 0.03490890559306918, "grad_norm": 68.73218536376953, "learning_rate": 9.65086646279307e-06, "loss": 0.2637, "step": 137 }, { "epoch": 0.035163715122945596, "grad_norm": 91.99873352050781, "learning_rate": 9.648318042813456e-06, "loss": 0.6914, "step": 138 }, { "epoch": 0.03541852465282202, "grad_norm": 45.22923278808594, "learning_rate": 9.645769622833843e-06, "loss": 0.2769, "step": 139 }, { "epoch": 0.035673334182698434, "grad_norm": 81.6264419555664, "learning_rate": 9.643221202854232e-06, "loss": 1.0229, "step": 140 }, { "epoch": 0.03592814371257485, "grad_norm": 122.22702026367188, "learning_rate": 9.640672782874619e-06, "loss": 0.5544, "step": 141 }, { "epoch": 0.036182953242451264, "grad_norm": 82.97919464111328, "learning_rate": 9.638124362895006e-06, "loss": 0.5898, "step": 142 }, { "epoch": 0.036437762772327686, "grad_norm": 82.19255828857422, "learning_rate": 9.635575942915393e-06, "loss": 0.5742, "step": 143 }, { "epoch": 0.0366925723022041, "grad_norm": 83.8096694946289, "learning_rate": 9.633027522935781e-06, "loss": 0.6187, "step": 144 }, { "epoch": 0.03694738183208052, "grad_norm": 121.03484344482422, "learning_rate": 9.630479102956168e-06, "loss": 0.7961, "step": 145 }, { "epoch": 0.03720219136195694, "grad_norm": 80.63111877441406, "learning_rate": 9.627930682976555e-06, "loss": 0.8154, "step": 146 }, { "epoch": 0.037457000891833354, "grad_norm": 65.17581939697266, "learning_rate": 9.625382262996942e-06, "loss": 0.5098, "step": 147 }, { "epoch": 0.03771181042170977, "grad_norm": 40.070350646972656, "learning_rate": 9.622833843017329e-06, "loss": 0.2778, "step": 148 }, { "epoch": 0.03796661995158619, "grad_norm": 20.47524642944336, "learning_rate": 9.620285423037718e-06, "loss": 0.0751, "step": 149 }, { "epoch": 0.03822142948146261, "grad_norm": 49.788536071777344, "learning_rate": 9.617737003058105e-06, "loss": 0.3403, "step": 150 }, { "epoch": 0.03847623901133902, "grad_norm": 119.06466674804688, "learning_rate": 9.615188583078492e-06, "loss": 0.6162, "step": 151 }, { "epoch": 0.038731048541215445, "grad_norm": 50.26600646972656, "learning_rate": 9.61264016309888e-06, "loss": 0.3516, "step": 152 }, { "epoch": 0.03898585807109186, "grad_norm": 83.78335571289062, "learning_rate": 9.610091743119267e-06, "loss": 0.4785, "step": 153 }, { "epoch": 0.039240667600968275, "grad_norm": 96.27387237548828, "learning_rate": 9.607543323139654e-06, "loss": 0.6113, "step": 154 }, { "epoch": 0.03949547713084469, "grad_norm": 108.59729766845703, "learning_rate": 9.604994903160043e-06, "loss": 1.4551, "step": 155 }, { "epoch": 0.03975028666072111, "grad_norm": 48.23210525512695, "learning_rate": 9.602446483180428e-06, "loss": 0.2307, "step": 156 }, { "epoch": 0.04000509619059753, "grad_norm": 102.53917694091797, "learning_rate": 9.599898063200817e-06, "loss": 0.3662, "step": 157 }, { "epoch": 0.04025990572047394, "grad_norm": 27.39449119567871, "learning_rate": 9.597349643221204e-06, "loss": 0.059, "step": 158 }, { "epoch": 0.040514715250350365, "grad_norm": 89.27335357666016, "learning_rate": 9.59480122324159e-06, "loss": 0.5918, "step": 159 }, { "epoch": 0.04076952478022678, "grad_norm": 91.40609741210938, "learning_rate": 9.59225280326198e-06, "loss": 0.4838, "step": 160 }, { "epoch": 0.041024334310103196, "grad_norm": 115.16098022460938, "learning_rate": 9.589704383282366e-06, "loss": 1.0645, "step": 161 }, { "epoch": 0.04127914383997962, "grad_norm": 46.53141784667969, "learning_rate": 9.587155963302753e-06, "loss": 0.2388, "step": 162 }, { "epoch": 0.04153395336985603, "grad_norm": 154.14964294433594, "learning_rate": 9.58460754332314e-06, "loss": 2.0156, "step": 163 }, { "epoch": 0.04178876289973245, "grad_norm": 90.96394348144531, "learning_rate": 9.582059123343527e-06, "loss": 0.8613, "step": 164 }, { "epoch": 0.04204357242960887, "grad_norm": 24.45330810546875, "learning_rate": 9.579510703363916e-06, "loss": 0.0919, "step": 165 }, { "epoch": 0.042298381959485286, "grad_norm": 32.14163589477539, "learning_rate": 9.576962283384303e-06, "loss": 0.2002, "step": 166 }, { "epoch": 0.0425531914893617, "grad_norm": 43.481101989746094, "learning_rate": 9.57441386340469e-06, "loss": 0.1218, "step": 167 }, { "epoch": 0.042808001019238116, "grad_norm": 57.750999450683594, "learning_rate": 9.571865443425077e-06, "loss": 0.3452, "step": 168 }, { "epoch": 0.04306281054911454, "grad_norm": 80.59860229492188, "learning_rate": 9.569317023445465e-06, "loss": 0.7041, "step": 169 }, { "epoch": 0.043317620078990954, "grad_norm": 149.20265197753906, "learning_rate": 9.566768603465852e-06, "loss": 1.6338, "step": 170 }, { "epoch": 0.04357242960886737, "grad_norm": 32.791175842285156, "learning_rate": 9.56422018348624e-06, "loss": 0.1924, "step": 171 }, { "epoch": 0.04382723913874379, "grad_norm": 42.31593704223633, "learning_rate": 9.561671763506628e-06, "loss": 0.1895, "step": 172 }, { "epoch": 0.044082048668620206, "grad_norm": 172.00379943847656, "learning_rate": 9.559123343527013e-06, "loss": 1.4551, "step": 173 }, { "epoch": 0.04433685819849662, "grad_norm": 54.94070053100586, "learning_rate": 9.556574923547402e-06, "loss": 0.2737, "step": 174 }, { "epoch": 0.044591667728373044, "grad_norm": 76.01490020751953, "learning_rate": 9.554026503567789e-06, "loss": 0.4902, "step": 175 }, { "epoch": 0.04484647725824946, "grad_norm": 174.40432739257812, "learning_rate": 9.551478083588176e-06, "loss": 1.9192, "step": 176 }, { "epoch": 0.045101286788125874, "grad_norm": 92.52979278564453, "learning_rate": 9.548929663608564e-06, "loss": 0.6532, "step": 177 }, { "epoch": 0.0453560963180023, "grad_norm": 180.22305297851562, "learning_rate": 9.546381243628951e-06, "loss": 1.0215, "step": 178 }, { "epoch": 0.04561090584787871, "grad_norm": 129.09922790527344, "learning_rate": 9.543832823649338e-06, "loss": 0.751, "step": 179 }, { "epoch": 0.04586571537775513, "grad_norm": 21.067270278930664, "learning_rate": 9.541284403669727e-06, "loss": 0.0986, "step": 180 }, { "epoch": 0.04612052490763154, "grad_norm": 91.2066650390625, "learning_rate": 9.538735983690112e-06, "loss": 0.7227, "step": 181 }, { "epoch": 0.046375334437507965, "grad_norm": 127.53115844726562, "learning_rate": 9.5361875637105e-06, "loss": 1.0312, "step": 182 }, { "epoch": 0.04663014396738438, "grad_norm": 52.19419860839844, "learning_rate": 9.533639143730888e-06, "loss": 0.3462, "step": 183 }, { "epoch": 0.046884953497260795, "grad_norm": 23.747705459594727, "learning_rate": 9.531090723751275e-06, "loss": 0.1035, "step": 184 }, { "epoch": 0.04713976302713722, "grad_norm": 43.406288146972656, "learning_rate": 9.528542303771663e-06, "loss": 0.2993, "step": 185 }, { "epoch": 0.04739457255701363, "grad_norm": 73.7568588256836, "learning_rate": 9.52599388379205e-06, "loss": 0.7832, "step": 186 }, { "epoch": 0.04764938208689005, "grad_norm": 113.39163970947266, "learning_rate": 9.523445463812437e-06, "loss": 0.8081, "step": 187 }, { "epoch": 0.04790419161676647, "grad_norm": 43.71094512939453, "learning_rate": 9.520897043832824e-06, "loss": 0.3262, "step": 188 }, { "epoch": 0.048159001146642885, "grad_norm": 16.169147491455078, "learning_rate": 9.518348623853211e-06, "loss": 0.074, "step": 189 }, { "epoch": 0.0484138106765193, "grad_norm": 95.02597045898438, "learning_rate": 9.515800203873598e-06, "loss": 0.325, "step": 190 }, { "epoch": 0.048668620206395716, "grad_norm": 26.56520652770996, "learning_rate": 9.513251783893987e-06, "loss": 0.0732, "step": 191 }, { "epoch": 0.04892342973627214, "grad_norm": 50.07223892211914, "learning_rate": 9.510703363914374e-06, "loss": 0.1323, "step": 192 }, { "epoch": 0.04917823926614855, "grad_norm": 219.73020935058594, "learning_rate": 9.50815494393476e-06, "loss": 1.639, "step": 193 }, { "epoch": 0.04943304879602497, "grad_norm": 55.342811584472656, "learning_rate": 9.50560652395515e-06, "loss": 0.1544, "step": 194 }, { "epoch": 0.04968785832590139, "grad_norm": 58.9760627746582, "learning_rate": 9.503058103975536e-06, "loss": 0.1416, "step": 195 }, { "epoch": 0.049942667855777806, "grad_norm": 67.9120101928711, "learning_rate": 9.500509683995923e-06, "loss": 0.1797, "step": 196 }, { "epoch": 0.05019747738565422, "grad_norm": 136.793701171875, "learning_rate": 9.497961264016312e-06, "loss": 0.6755, "step": 197 }, { "epoch": 0.05045228691553064, "grad_norm": 33.05602264404297, "learning_rate": 9.495412844036697e-06, "loss": 0.1199, "step": 198 }, { "epoch": 0.05070709644540706, "grad_norm": 99.4974365234375, "learning_rate": 9.492864424057086e-06, "loss": 0.4309, "step": 199 }, { "epoch": 0.050961905975283474, "grad_norm": 193.66893005371094, "learning_rate": 9.490316004077473e-06, "loss": 1.4194, "step": 200 }, { "epoch": 0.051216715505159896, "grad_norm": 189.33763122558594, "learning_rate": 9.48776758409786e-06, "loss": 1.1902, "step": 201 }, { "epoch": 0.05147152503503631, "grad_norm": 247.53817749023438, "learning_rate": 9.485219164118248e-06, "loss": 1.0016, "step": 202 }, { "epoch": 0.051726334564912727, "grad_norm": 90.2113265991211, "learning_rate": 9.482670744138635e-06, "loss": 0.2542, "step": 203 }, { "epoch": 0.05198114409478914, "grad_norm": 16.512563705444336, "learning_rate": 9.480122324159022e-06, "loss": 0.0591, "step": 204 }, { "epoch": 0.052235953624665564, "grad_norm": 49.5167121887207, "learning_rate": 9.477573904179409e-06, "loss": 0.199, "step": 205 }, { "epoch": 0.05249076315454198, "grad_norm": 51.188228607177734, "learning_rate": 9.475025484199796e-06, "loss": 0.114, "step": 206 }, { "epoch": 0.052745572684418394, "grad_norm": 32.93622589111328, "learning_rate": 9.472477064220185e-06, "loss": 0.1207, "step": 207 }, { "epoch": 0.05300038221429482, "grad_norm": 359.47149658203125, "learning_rate": 9.469928644240572e-06, "loss": 3.2734, "step": 208 }, { "epoch": 0.05325519174417123, "grad_norm": 312.2518615722656, "learning_rate": 9.467380224260959e-06, "loss": 1.0703, "step": 209 }, { "epoch": 0.05351000127404765, "grad_norm": 37.07282257080078, "learning_rate": 9.464831804281346e-06, "loss": 0.0797, "step": 210 }, { "epoch": 0.05376481080392407, "grad_norm": 211.88494873046875, "learning_rate": 9.462283384301734e-06, "loss": 0.6992, "step": 211 }, { "epoch": 0.054019620333800485, "grad_norm": 202.12030029296875, "learning_rate": 9.459734964322121e-06, "loss": 1.0508, "step": 212 }, { "epoch": 0.0542744298636769, "grad_norm": 212.82247924804688, "learning_rate": 9.457186544342508e-06, "loss": 1.7188, "step": 213 }, { "epoch": 0.05452923939355332, "grad_norm": 43.56071090698242, "learning_rate": 9.454638124362897e-06, "loss": 0.092, "step": 214 }, { "epoch": 0.05478404892342974, "grad_norm": 85.43267059326172, "learning_rate": 9.452089704383282e-06, "loss": 0.2646, "step": 215 }, { "epoch": 0.05503885845330615, "grad_norm": 35.799468994140625, "learning_rate": 9.44954128440367e-06, "loss": 0.0968, "step": 216 }, { "epoch": 0.05529366798318257, "grad_norm": 161.7420196533203, "learning_rate": 9.446992864424058e-06, "loss": 1.1367, "step": 217 }, { "epoch": 0.05554847751305899, "grad_norm": 216.98245239257812, "learning_rate": 9.444444444444445e-06, "loss": 0.7979, "step": 218 }, { "epoch": 0.055803287042935405, "grad_norm": 120.71834564208984, "learning_rate": 9.441896024464833e-06, "loss": 0.5664, "step": 219 }, { "epoch": 0.05605809657281182, "grad_norm": 27.95607566833496, "learning_rate": 9.43934760448522e-06, "loss": 0.064, "step": 220 }, { "epoch": 0.05631290610268824, "grad_norm": 95.45901489257812, "learning_rate": 9.436799184505607e-06, "loss": 0.2168, "step": 221 }, { "epoch": 0.05656771563256466, "grad_norm": 147.98501586914062, "learning_rate": 9.434250764525996e-06, "loss": 0.481, "step": 222 }, { "epoch": 0.05682252516244107, "grad_norm": 30.702754974365234, "learning_rate": 9.431702344546381e-06, "loss": 0.0845, "step": 223 }, { "epoch": 0.057077334692317495, "grad_norm": 7.381120681762695, "learning_rate": 9.42915392456677e-06, "loss": 0.0198, "step": 224 }, { "epoch": 0.05733214422219391, "grad_norm": 122.435302734375, "learning_rate": 9.426605504587157e-06, "loss": 0.575, "step": 225 }, { "epoch": 0.057586953752070326, "grad_norm": 22.403907775878906, "learning_rate": 9.424057084607544e-06, "loss": 0.0697, "step": 226 }, { "epoch": 0.05784176328194675, "grad_norm": 56.44075393676758, "learning_rate": 9.421508664627932e-06, "loss": 0.1838, "step": 227 }, { "epoch": 0.05809657281182316, "grad_norm": 41.41353988647461, "learning_rate": 9.41896024464832e-06, "loss": 0.1699, "step": 228 }, { "epoch": 0.05835138234169958, "grad_norm": 149.31405639648438, "learning_rate": 9.416411824668706e-06, "loss": 0.6416, "step": 229 }, { "epoch": 0.058606191871575994, "grad_norm": 335.9905700683594, "learning_rate": 9.413863404689093e-06, "loss": 2.4219, "step": 230 }, { "epoch": 0.058861001401452416, "grad_norm": 98.74788665771484, "learning_rate": 9.411314984709482e-06, "loss": 0.2881, "step": 231 }, { "epoch": 0.05911581093132883, "grad_norm": 44.28268051147461, "learning_rate": 9.408766564729869e-06, "loss": 0.1116, "step": 232 }, { "epoch": 0.05937062046120525, "grad_norm": 23.923595428466797, "learning_rate": 9.406218144750256e-06, "loss": 0.0948, "step": 233 }, { "epoch": 0.05962542999108167, "grad_norm": 44.63673782348633, "learning_rate": 9.403669724770643e-06, "loss": 0.2264, "step": 234 }, { "epoch": 0.059880239520958084, "grad_norm": 104.87225341796875, "learning_rate": 9.40112130479103e-06, "loss": 0.6498, "step": 235 }, { "epoch": 0.0601350490508345, "grad_norm": 276.3097229003906, "learning_rate": 9.398572884811418e-06, "loss": 0.8125, "step": 236 }, { "epoch": 0.06038985858071092, "grad_norm": 36.04986572265625, "learning_rate": 9.396024464831805e-06, "loss": 0.0614, "step": 237 }, { "epoch": 0.06064466811058734, "grad_norm": 144.49658203125, "learning_rate": 9.393476044852192e-06, "loss": 0.9243, "step": 238 }, { "epoch": 0.06089947764046375, "grad_norm": 18.446163177490234, "learning_rate": 9.39092762487258e-06, "loss": 0.0526, "step": 239 }, { "epoch": 0.061154287170340174, "grad_norm": 73.78413391113281, "learning_rate": 9.388379204892966e-06, "loss": 0.4658, "step": 240 }, { "epoch": 0.06140909670021659, "grad_norm": 16.34041976928711, "learning_rate": 9.385830784913355e-06, "loss": 0.0255, "step": 241 }, { "epoch": 0.061663906230093005, "grad_norm": 222.79359436035156, "learning_rate": 9.383282364933742e-06, "loss": 1.3262, "step": 242 }, { "epoch": 0.06191871575996942, "grad_norm": 149.55638122558594, "learning_rate": 9.380733944954129e-06, "loss": 0.2988, "step": 243 }, { "epoch": 0.06217352528984584, "grad_norm": 277.6204833984375, "learning_rate": 9.378185524974517e-06, "loss": 0.9102, "step": 244 }, { "epoch": 0.06242833481972226, "grad_norm": 36.71376037597656, "learning_rate": 9.375637104994904e-06, "loss": 0.1365, "step": 245 }, { "epoch": 0.06268314434959868, "grad_norm": 109.22211456298828, "learning_rate": 9.373088685015291e-06, "loss": 0.4362, "step": 246 }, { "epoch": 0.0629379538794751, "grad_norm": 239.53807067871094, "learning_rate": 9.37054026503568e-06, "loss": 1.7959, "step": 247 }, { "epoch": 0.06319276340935151, "grad_norm": 140.5045623779297, "learning_rate": 9.367991845056067e-06, "loss": 0.6631, "step": 248 }, { "epoch": 0.06344757293922793, "grad_norm": 172.60467529296875, "learning_rate": 9.365443425076454e-06, "loss": 0.8887, "step": 249 }, { "epoch": 0.06370238246910434, "grad_norm": 158.14743041992188, "learning_rate": 9.36289500509684e-06, "loss": 1.2197, "step": 250 }, { "epoch": 0.06395719199898076, "grad_norm": 162.47970581054688, "learning_rate": 9.360346585117228e-06, "loss": 0.4642, "step": 251 }, { "epoch": 0.06421200152885718, "grad_norm": 206.1090545654297, "learning_rate": 9.357798165137616e-06, "loss": 2.6309, "step": 252 }, { "epoch": 0.0644668110587336, "grad_norm": 337.3321838378906, "learning_rate": 9.355249745158003e-06, "loss": 1.8027, "step": 253 }, { "epoch": 0.06472162058861002, "grad_norm": 41.692630767822266, "learning_rate": 9.35270132517839e-06, "loss": 0.1558, "step": 254 }, { "epoch": 0.06497643011848643, "grad_norm": 171.75767517089844, "learning_rate": 9.350152905198777e-06, "loss": 1.9302, "step": 255 }, { "epoch": 0.06523123964836285, "grad_norm": 163.48558044433594, "learning_rate": 9.347604485219166e-06, "loss": 1.0244, "step": 256 }, { "epoch": 0.06548604917823926, "grad_norm": 52.2914924621582, "learning_rate": 9.345056065239553e-06, "loss": 0.1604, "step": 257 }, { "epoch": 0.06574085870811569, "grad_norm": 22.39143180847168, "learning_rate": 9.34250764525994e-06, "loss": 0.0393, "step": 258 }, { "epoch": 0.0659956682379921, "grad_norm": 226.38031005859375, "learning_rate": 9.339959225280327e-06, "loss": 1.334, "step": 259 }, { "epoch": 0.06625047776786852, "grad_norm": 18.66448402404785, "learning_rate": 9.337410805300714e-06, "loss": 0.0604, "step": 260 }, { "epoch": 0.06650528729774494, "grad_norm": 17.126380920410156, "learning_rate": 9.334862385321102e-06, "loss": 0.0205, "step": 261 }, { "epoch": 0.06676009682762135, "grad_norm": 36.977176666259766, "learning_rate": 9.332313965341489e-06, "loss": 0.1129, "step": 262 }, { "epoch": 0.06701490635749777, "grad_norm": 123.706787109375, "learning_rate": 9.329765545361876e-06, "loss": 0.7726, "step": 263 }, { "epoch": 0.06726971588737418, "grad_norm": 107.77877044677734, "learning_rate": 9.327217125382265e-06, "loss": 0.2598, "step": 264 }, { "epoch": 0.06752452541725061, "grad_norm": 316.3159484863281, "learning_rate": 9.32466870540265e-06, "loss": 1.2327, "step": 265 }, { "epoch": 0.06777933494712703, "grad_norm": 54.271331787109375, "learning_rate": 9.322120285423039e-06, "loss": 0.2234, "step": 266 }, { "epoch": 0.06803414447700344, "grad_norm": 30.470794677734375, "learning_rate": 9.319571865443426e-06, "loss": 0.0768, "step": 267 }, { "epoch": 0.06828895400687986, "grad_norm": 95.85240173339844, "learning_rate": 9.317023445463813e-06, "loss": 0.627, "step": 268 }, { "epoch": 0.06854376353675627, "grad_norm": 60.286956787109375, "learning_rate": 9.314475025484201e-06, "loss": 0.1976, "step": 269 }, { "epoch": 0.06879857306663269, "grad_norm": 31.29051399230957, "learning_rate": 9.311926605504588e-06, "loss": 0.0829, "step": 270 }, { "epoch": 0.06905338259650912, "grad_norm": 102.92997741699219, "learning_rate": 9.309378185524975e-06, "loss": 0.4053, "step": 271 }, { "epoch": 0.06930819212638553, "grad_norm": 60.99541091918945, "learning_rate": 9.306829765545362e-06, "loss": 0.1536, "step": 272 }, { "epoch": 0.06956300165626195, "grad_norm": 32.87839126586914, "learning_rate": 9.30428134556575e-06, "loss": 0.0728, "step": 273 }, { "epoch": 0.06981781118613836, "grad_norm": 123.33274841308594, "learning_rate": 9.301732925586138e-06, "loss": 0.5645, "step": 274 }, { "epoch": 0.07007262071601478, "grad_norm": 93.30522918701172, "learning_rate": 9.299184505606525e-06, "loss": 0.2998, "step": 275 }, { "epoch": 0.07032743024589119, "grad_norm": 221.28787231445312, "learning_rate": 9.296636085626912e-06, "loss": 1.1975, "step": 276 }, { "epoch": 0.07058223977576761, "grad_norm": 98.23433685302734, "learning_rate": 9.294087665647298e-06, "loss": 0.1985, "step": 277 }, { "epoch": 0.07083704930564404, "grad_norm": 312.95269775390625, "learning_rate": 9.291539245667687e-06, "loss": 1.5562, "step": 278 }, { "epoch": 0.07109185883552045, "grad_norm": 347.3350524902344, "learning_rate": 9.288990825688074e-06, "loss": 1.0977, "step": 279 }, { "epoch": 0.07134666836539687, "grad_norm": 66.75719451904297, "learning_rate": 9.286442405708461e-06, "loss": 0.1246, "step": 280 }, { "epoch": 0.07160147789527328, "grad_norm": 51.94187545776367, "learning_rate": 9.28389398572885e-06, "loss": 0.0555, "step": 281 }, { "epoch": 0.0718562874251497, "grad_norm": 160.7375030517578, "learning_rate": 9.281345565749235e-06, "loss": 0.6597, "step": 282 }, { "epoch": 0.07211109695502611, "grad_norm": 57.98982238769531, "learning_rate": 9.278797145769624e-06, "loss": 0.1765, "step": 283 }, { "epoch": 0.07236590648490253, "grad_norm": 294.85382080078125, "learning_rate": 9.27624872579001e-06, "loss": 1.6016, "step": 284 }, { "epoch": 0.07262071601477896, "grad_norm": 30.357391357421875, "learning_rate": 9.273700305810397e-06, "loss": 0.027, "step": 285 }, { "epoch": 0.07287552554465537, "grad_norm": 136.214599609375, "learning_rate": 9.271151885830786e-06, "loss": 0.4365, "step": 286 }, { "epoch": 0.07313033507453179, "grad_norm": 100.21566009521484, "learning_rate": 9.268603465851173e-06, "loss": 0.2799, "step": 287 }, { "epoch": 0.0733851446044082, "grad_norm": 28.234405517578125, "learning_rate": 9.26605504587156e-06, "loss": 0.0753, "step": 288 }, { "epoch": 0.07363995413428462, "grad_norm": 21.46152114868164, "learning_rate": 9.263506625891949e-06, "loss": 0.0305, "step": 289 }, { "epoch": 0.07389476366416103, "grad_norm": 51.17667770385742, "learning_rate": 9.260958205912336e-06, "loss": 0.1117, "step": 290 }, { "epoch": 0.07414957319403746, "grad_norm": 34.0002555847168, "learning_rate": 9.258409785932723e-06, "loss": 0.0691, "step": 291 }, { "epoch": 0.07440438272391388, "grad_norm": 212.44985961914062, "learning_rate": 9.25586136595311e-06, "loss": 0.5986, "step": 292 }, { "epoch": 0.0746591922537903, "grad_norm": 43.983367919921875, "learning_rate": 9.253312945973496e-06, "loss": 0.1464, "step": 293 }, { "epoch": 0.07491400178366671, "grad_norm": 12.064727783203125, "learning_rate": 9.250764525993885e-06, "loss": 0.0178, "step": 294 }, { "epoch": 0.07516881131354312, "grad_norm": 14.837090492248535, "learning_rate": 9.248216106014272e-06, "loss": 0.025, "step": 295 }, { "epoch": 0.07542362084341954, "grad_norm": 59.155887603759766, "learning_rate": 9.245667686034659e-06, "loss": 0.1115, "step": 296 }, { "epoch": 0.07567843037329595, "grad_norm": 437.534423828125, "learning_rate": 9.243119266055046e-06, "loss": 1.8698, "step": 297 }, { "epoch": 0.07593323990317238, "grad_norm": 443.5690612792969, "learning_rate": 9.240570846075435e-06, "loss": 5.4219, "step": 298 }, { "epoch": 0.0761880494330488, "grad_norm": 326.8785095214844, "learning_rate": 9.238022426095822e-06, "loss": 2.1965, "step": 299 }, { "epoch": 0.07644285896292521, "grad_norm": 552.1416625976562, "learning_rate": 9.235474006116209e-06, "loss": 1.4038, "step": 300 }, { "epoch": 0.07669766849280163, "grad_norm": 25.277873992919922, "learning_rate": 9.232925586136595e-06, "loss": 0.0206, "step": 301 }, { "epoch": 0.07695247802267804, "grad_norm": 113.59070587158203, "learning_rate": 9.230377166156982e-06, "loss": 0.228, "step": 302 }, { "epoch": 0.07720728755255446, "grad_norm": 35.250572204589844, "learning_rate": 9.227828746177371e-06, "loss": 0.0613, "step": 303 }, { "epoch": 0.07746209708243089, "grad_norm": 469.51129150390625, "learning_rate": 9.225280326197758e-06, "loss": 0.5611, "step": 304 }, { "epoch": 0.0777169066123073, "grad_norm": 15.395145416259766, "learning_rate": 9.222731906218145e-06, "loss": 0.0166, "step": 305 }, { "epoch": 0.07797171614218372, "grad_norm": 74.85309600830078, "learning_rate": 9.220183486238534e-06, "loss": 0.2, "step": 306 }, { "epoch": 0.07822652567206013, "grad_norm": 207.7628173828125, "learning_rate": 9.21763506625892e-06, "loss": 0.6133, "step": 307 }, { "epoch": 0.07848133520193655, "grad_norm": 282.39794921875, "learning_rate": 9.215086646279308e-06, "loss": 0.5869, "step": 308 }, { "epoch": 0.07873614473181297, "grad_norm": 133.83595275878906, "learning_rate": 9.212538226299696e-06, "loss": 0.2723, "step": 309 }, { "epoch": 0.07899095426168938, "grad_norm": 171.92379760742188, "learning_rate": 9.209989806320081e-06, "loss": 1.1079, "step": 310 }, { "epoch": 0.07924576379156581, "grad_norm": 82.58956909179688, "learning_rate": 9.20744138634047e-06, "loss": 0.1338, "step": 311 }, { "epoch": 0.07950057332144222, "grad_norm": 183.9136962890625, "learning_rate": 9.204892966360857e-06, "loss": 0.6738, "step": 312 }, { "epoch": 0.07975538285131864, "grad_norm": 81.39661407470703, "learning_rate": 9.202344546381244e-06, "loss": 0.1177, "step": 313 }, { "epoch": 0.08001019238119506, "grad_norm": 355.1209411621094, "learning_rate": 9.199796126401633e-06, "loss": 1.0957, "step": 314 }, { "epoch": 0.08026500191107147, "grad_norm": 232.779052734375, "learning_rate": 9.19724770642202e-06, "loss": 1.4729, "step": 315 }, { "epoch": 0.08051981144094789, "grad_norm": 86.49019622802734, "learning_rate": 9.194699286442407e-06, "loss": 0.2676, "step": 316 }, { "epoch": 0.08077462097082432, "grad_norm": 100.93199920654297, "learning_rate": 9.192150866462794e-06, "loss": 0.7358, "step": 317 }, { "epoch": 0.08102943050070073, "grad_norm": 131.91976928710938, "learning_rate": 9.18960244648318e-06, "loss": 1.0708, "step": 318 }, { "epoch": 0.08128424003057715, "grad_norm": 28.828283309936523, "learning_rate": 9.187054026503569e-06, "loss": 0.0595, "step": 319 }, { "epoch": 0.08153904956045356, "grad_norm": 25.785743713378906, "learning_rate": 9.184505606523956e-06, "loss": 0.0583, "step": 320 }, { "epoch": 0.08179385909032998, "grad_norm": 34.221527099609375, "learning_rate": 9.181957186544343e-06, "loss": 0.0574, "step": 321 }, { "epoch": 0.08204866862020639, "grad_norm": 17.57350730895996, "learning_rate": 9.17940876656473e-06, "loss": 0.0386, "step": 322 }, { "epoch": 0.0823034781500828, "grad_norm": 130.1326446533203, "learning_rate": 9.176860346585119e-06, "loss": 0.3003, "step": 323 }, { "epoch": 0.08255828767995924, "grad_norm": 62.6984748840332, "learning_rate": 9.174311926605506e-06, "loss": 0.2024, "step": 324 }, { "epoch": 0.08281309720983565, "grad_norm": 31.103540420532227, "learning_rate": 9.171763506625893e-06, "loss": 0.079, "step": 325 }, { "epoch": 0.08306790673971207, "grad_norm": 183.28062438964844, "learning_rate": 9.169215086646281e-06, "loss": 0.7935, "step": 326 }, { "epoch": 0.08332271626958848, "grad_norm": 119.90725708007812, "learning_rate": 9.166666666666666e-06, "loss": 0.5273, "step": 327 }, { "epoch": 0.0835775257994649, "grad_norm": 136.00485229492188, "learning_rate": 9.164118246687055e-06, "loss": 0.316, "step": 328 }, { "epoch": 0.08383233532934131, "grad_norm": 37.78673553466797, "learning_rate": 9.161569826707442e-06, "loss": 0.0786, "step": 329 }, { "epoch": 0.08408714485921774, "grad_norm": 83.40485382080078, "learning_rate": 9.159021406727829e-06, "loss": 0.2568, "step": 330 }, { "epoch": 0.08434195438909416, "grad_norm": 11.905128479003906, "learning_rate": 9.156472986748218e-06, "loss": 0.0206, "step": 331 }, { "epoch": 0.08459676391897057, "grad_norm": 12.747295379638672, "learning_rate": 9.153924566768605e-06, "loss": 0.0306, "step": 332 }, { "epoch": 0.08485157344884699, "grad_norm": 239.58973693847656, "learning_rate": 9.151376146788992e-06, "loss": 0.6639, "step": 333 }, { "epoch": 0.0851063829787234, "grad_norm": 220.24609375, "learning_rate": 9.14882772680938e-06, "loss": 0.8594, "step": 334 }, { "epoch": 0.08536119250859982, "grad_norm": 28.42400550842285, "learning_rate": 9.146279306829765e-06, "loss": 0.0574, "step": 335 }, { "epoch": 0.08561600203847623, "grad_norm": 9.111586570739746, "learning_rate": 9.143730886850154e-06, "loss": 0.0265, "step": 336 }, { "epoch": 0.08587081156835266, "grad_norm": 94.72270202636719, "learning_rate": 9.141182466870541e-06, "loss": 0.2195, "step": 337 }, { "epoch": 0.08612562109822908, "grad_norm": 122.28697204589844, "learning_rate": 9.138634046890928e-06, "loss": 0.1699, "step": 338 }, { "epoch": 0.08638043062810549, "grad_norm": 67.74645233154297, "learning_rate": 9.136085626911317e-06, "loss": 0.0676, "step": 339 }, { "epoch": 0.08663524015798191, "grad_norm": 62.78877258300781, "learning_rate": 9.133537206931704e-06, "loss": 0.5469, "step": 340 }, { "epoch": 0.08689004968785832, "grad_norm": 48.57039260864258, "learning_rate": 9.13098878695209e-06, "loss": 0.1348, "step": 341 }, { "epoch": 0.08714485921773474, "grad_norm": 66.81463623046875, "learning_rate": 9.128440366972477e-06, "loss": 0.2339, "step": 342 }, { "epoch": 0.08739966874761117, "grad_norm": 56.30792236328125, "learning_rate": 9.125891946992864e-06, "loss": 0.1525, "step": 343 }, { "epoch": 0.08765447827748758, "grad_norm": 425.0268249511719, "learning_rate": 9.123343527013251e-06, "loss": 1.4297, "step": 344 }, { "epoch": 0.087909287807364, "grad_norm": 57.313533782958984, "learning_rate": 9.12079510703364e-06, "loss": 0.1024, "step": 345 }, { "epoch": 0.08816409733724041, "grad_norm": 87.43323516845703, "learning_rate": 9.118246687054027e-06, "loss": 0.3564, "step": 346 }, { "epoch": 0.08841890686711683, "grad_norm": 51.800594329833984, "learning_rate": 9.115698267074414e-06, "loss": 0.0974, "step": 347 }, { "epoch": 0.08867371639699324, "grad_norm": 89.66438293457031, "learning_rate": 9.113149847094803e-06, "loss": 0.1414, "step": 348 }, { "epoch": 0.08892852592686966, "grad_norm": 49.58383560180664, "learning_rate": 9.11060142711519e-06, "loss": 0.1104, "step": 349 }, { "epoch": 0.08918333545674609, "grad_norm": 154.91024780273438, "learning_rate": 9.108053007135577e-06, "loss": 0.4082, "step": 350 }, { "epoch": 0.0894381449866225, "grad_norm": 153.825927734375, "learning_rate": 9.105504587155965e-06, "loss": 1.2725, "step": 351 }, { "epoch": 0.08969295451649892, "grad_norm": 33.371734619140625, "learning_rate": 9.10295616717635e-06, "loss": 0.0538, "step": 352 }, { "epoch": 0.08994776404637533, "grad_norm": 22.483362197875977, "learning_rate": 9.100407747196739e-06, "loss": 0.0415, "step": 353 }, { "epoch": 0.09020257357625175, "grad_norm": 329.1225891113281, "learning_rate": 9.097859327217126e-06, "loss": 0.8398, "step": 354 }, { "epoch": 0.09045738310612816, "grad_norm": 6.473784923553467, "learning_rate": 9.095310907237513e-06, "loss": 0.0125, "step": 355 }, { "epoch": 0.0907121926360046, "grad_norm": 60.81599426269531, "learning_rate": 9.092762487257902e-06, "loss": 0.115, "step": 356 }, { "epoch": 0.09096700216588101, "grad_norm": 20.473493576049805, "learning_rate": 9.090214067278289e-06, "loss": 0.0571, "step": 357 }, { "epoch": 0.09122181169575742, "grad_norm": 28.45832061767578, "learning_rate": 9.087665647298676e-06, "loss": 0.0414, "step": 358 }, { "epoch": 0.09147662122563384, "grad_norm": 157.40853881835938, "learning_rate": 9.085117227319062e-06, "loss": 0.7202, "step": 359 }, { "epoch": 0.09173143075551025, "grad_norm": 250.04537963867188, "learning_rate": 9.08256880733945e-06, "loss": 1.2341, "step": 360 }, { "epoch": 0.09198624028538667, "grad_norm": 284.0129089355469, "learning_rate": 9.080020387359838e-06, "loss": 3.6211, "step": 361 }, { "epoch": 0.09224104981526308, "grad_norm": 97.96273803710938, "learning_rate": 9.077471967380225e-06, "loss": 0.3057, "step": 362 }, { "epoch": 0.09249585934513951, "grad_norm": 102.12102508544922, "learning_rate": 9.074923547400612e-06, "loss": 0.266, "step": 363 }, { "epoch": 0.09275066887501593, "grad_norm": 206.81858825683594, "learning_rate": 9.072375127420999e-06, "loss": 0.2295, "step": 364 }, { "epoch": 0.09300547840489234, "grad_norm": 7.732639789581299, "learning_rate": 9.069826707441388e-06, "loss": 0.017, "step": 365 }, { "epoch": 0.09326028793476876, "grad_norm": 133.55545043945312, "learning_rate": 9.067278287461775e-06, "loss": 0.2849, "step": 366 }, { "epoch": 0.09351509746464517, "grad_norm": 80.06969451904297, "learning_rate": 9.064729867482161e-06, "loss": 0.4131, "step": 367 }, { "epoch": 0.09376990699452159, "grad_norm": 37.314029693603516, "learning_rate": 9.06218144750255e-06, "loss": 0.0874, "step": 368 }, { "epoch": 0.09402471652439802, "grad_norm": 377.7703857421875, "learning_rate": 9.059633027522935e-06, "loss": 0.8439, "step": 369 }, { "epoch": 0.09427952605427443, "grad_norm": 79.1003189086914, "learning_rate": 9.057084607543324e-06, "loss": 0.1245, "step": 370 }, { "epoch": 0.09453433558415085, "grad_norm": 19.655742645263672, "learning_rate": 9.054536187563711e-06, "loss": 0.0449, "step": 371 }, { "epoch": 0.09478914511402727, "grad_norm": 227.97918701171875, "learning_rate": 9.051987767584098e-06, "loss": 1.4347, "step": 372 }, { "epoch": 0.09504395464390368, "grad_norm": 31.99831771850586, "learning_rate": 9.049439347604487e-06, "loss": 0.0801, "step": 373 }, { "epoch": 0.0952987641737801, "grad_norm": 4.896319389343262, "learning_rate": 9.046890927624874e-06, "loss": 0.0041, "step": 374 }, { "epoch": 0.09555357370365651, "grad_norm": 160.3213348388672, "learning_rate": 9.04434250764526e-06, "loss": 0.7246, "step": 375 }, { "epoch": 0.09580838323353294, "grad_norm": 43.0701789855957, "learning_rate": 9.041794087665649e-06, "loss": 0.1387, "step": 376 }, { "epoch": 0.09606319276340936, "grad_norm": 36.43310546875, "learning_rate": 9.039245667686034e-06, "loss": 0.0621, "step": 377 }, { "epoch": 0.09631800229328577, "grad_norm": 32.22747802734375, "learning_rate": 9.036697247706423e-06, "loss": 0.075, "step": 378 }, { "epoch": 0.09657281182316219, "grad_norm": 83.71952819824219, "learning_rate": 9.03414882772681e-06, "loss": 0.7339, "step": 379 }, { "epoch": 0.0968276213530386, "grad_norm": 480.2369384765625, "learning_rate": 9.031600407747197e-06, "loss": 2.1562, "step": 380 }, { "epoch": 0.09708243088291502, "grad_norm": 10.179464340209961, "learning_rate": 9.029051987767586e-06, "loss": 0.0235, "step": 381 }, { "epoch": 0.09733724041279143, "grad_norm": 61.2548828125, "learning_rate": 9.026503567787973e-06, "loss": 0.0878, "step": 382 }, { "epoch": 0.09759204994266786, "grad_norm": 379.81072998046875, "learning_rate": 9.02395514780836e-06, "loss": 1.1023, "step": 383 }, { "epoch": 0.09784685947254428, "grad_norm": 367.8057861328125, "learning_rate": 9.021406727828746e-06, "loss": 0.8828, "step": 384 }, { "epoch": 0.09810166900242069, "grad_norm": 15.864148139953613, "learning_rate": 9.018858307849135e-06, "loss": 0.0121, "step": 385 }, { "epoch": 0.0983564785322971, "grad_norm": 81.1125259399414, "learning_rate": 9.016309887869522e-06, "loss": 0.1137, "step": 386 }, { "epoch": 0.09861128806217352, "grad_norm": 28.933372497558594, "learning_rate": 9.013761467889909e-06, "loss": 0.049, "step": 387 }, { "epoch": 0.09886609759204994, "grad_norm": 212.48104858398438, "learning_rate": 9.011213047910296e-06, "loss": 1.0225, "step": 388 }, { "epoch": 0.09912090712192637, "grad_norm": 49.86363220214844, "learning_rate": 9.008664627930683e-06, "loss": 0.106, "step": 389 }, { "epoch": 0.09937571665180278, "grad_norm": 85.92784118652344, "learning_rate": 9.006116207951072e-06, "loss": 0.1449, "step": 390 }, { "epoch": 0.0996305261816792, "grad_norm": 360.6685485839844, "learning_rate": 9.003567787971458e-06, "loss": 0.2815, "step": 391 }, { "epoch": 0.09988533571155561, "grad_norm": 180.4518585205078, "learning_rate": 9.001019367991845e-06, "loss": 0.752, "step": 392 }, { "epoch": 0.10014014524143203, "grad_norm": 3.5057826042175293, "learning_rate": 8.998470948012234e-06, "loss": 0.0049, "step": 393 }, { "epoch": 0.10039495477130844, "grad_norm": 487.2030334472656, "learning_rate": 8.99592252803262e-06, "loss": 1.5312, "step": 394 }, { "epoch": 0.10064976430118486, "grad_norm": 603.1558837890625, "learning_rate": 8.993374108053008e-06, "loss": 0.6201, "step": 395 }, { "epoch": 0.10090457383106129, "grad_norm": 37.78219223022461, "learning_rate": 8.990825688073395e-06, "loss": 0.0745, "step": 396 }, { "epoch": 0.1011593833609377, "grad_norm": 292.55413818359375, "learning_rate": 8.988277268093782e-06, "loss": 0.6655, "step": 397 }, { "epoch": 0.10141419289081412, "grad_norm": 6.80036735534668, "learning_rate": 8.98572884811417e-06, "loss": 0.006, "step": 398 }, { "epoch": 0.10166900242069053, "grad_norm": 16.878122329711914, "learning_rate": 8.983180428134558e-06, "loss": 0.0158, "step": 399 }, { "epoch": 0.10192381195056695, "grad_norm": 164.44036865234375, "learning_rate": 8.980632008154944e-06, "loss": 1.6973, "step": 400 }, { "epoch": 0.10217862148044336, "grad_norm": 36.3424186706543, "learning_rate": 8.978083588175333e-06, "loss": 0.0687, "step": 401 }, { "epoch": 0.10243343101031979, "grad_norm": 41.721500396728516, "learning_rate": 8.97553516819572e-06, "loss": 0.0643, "step": 402 }, { "epoch": 0.10268824054019621, "grad_norm": 71.86236572265625, "learning_rate": 8.972986748216107e-06, "loss": 0.1397, "step": 403 }, { "epoch": 0.10294305007007262, "grad_norm": 14.406981468200684, "learning_rate": 8.970438328236494e-06, "loss": 0.0294, "step": 404 }, { "epoch": 0.10319785959994904, "grad_norm": 963.9365234375, "learning_rate": 8.967889908256881e-06, "loss": 1.6487, "step": 405 }, { "epoch": 0.10345266912982545, "grad_norm": 8.305174827575684, "learning_rate": 8.96534148827727e-06, "loss": 0.007, "step": 406 }, { "epoch": 0.10370747865970187, "grad_norm": 12.301192283630371, "learning_rate": 8.962793068297657e-06, "loss": 0.0453, "step": 407 }, { "epoch": 0.10396228818957828, "grad_norm": 219.79122924804688, "learning_rate": 8.960244648318043e-06, "loss": 0.5395, "step": 408 }, { "epoch": 0.10421709771945471, "grad_norm": 31.62107276916504, "learning_rate": 8.95769622833843e-06, "loss": 0.0535, "step": 409 }, { "epoch": 0.10447190724933113, "grad_norm": 275.3426818847656, "learning_rate": 8.955147808358819e-06, "loss": 1.1875, "step": 410 }, { "epoch": 0.10472671677920754, "grad_norm": 30.74949073791504, "learning_rate": 8.952599388379206e-06, "loss": 0.0547, "step": 411 }, { "epoch": 0.10498152630908396, "grad_norm": 33.42961883544922, "learning_rate": 8.950050968399593e-06, "loss": 0.0822, "step": 412 }, { "epoch": 0.10523633583896037, "grad_norm": 26.316104888916016, "learning_rate": 8.94750254841998e-06, "loss": 0.0429, "step": 413 }, { "epoch": 0.10549114536883679, "grad_norm": 287.6810302734375, "learning_rate": 8.944954128440367e-06, "loss": 0.7198, "step": 414 }, { "epoch": 0.10574595489871322, "grad_norm": 50.971832275390625, "learning_rate": 8.942405708460756e-06, "loss": 0.1795, "step": 415 }, { "epoch": 0.10600076442858963, "grad_norm": 335.1401672363281, "learning_rate": 8.939857288481142e-06, "loss": 2.094, "step": 416 }, { "epoch": 0.10625557395846605, "grad_norm": 31.346248626708984, "learning_rate": 8.93730886850153e-06, "loss": 0.0568, "step": 417 }, { "epoch": 0.10651038348834246, "grad_norm": 82.54093170166016, "learning_rate": 8.934760448521918e-06, "loss": 0.25, "step": 418 }, { "epoch": 0.10676519301821888, "grad_norm": 326.3962707519531, "learning_rate": 8.932212028542305e-06, "loss": 0.5752, "step": 419 }, { "epoch": 0.1070200025480953, "grad_norm": 34.33600997924805, "learning_rate": 8.929663608562692e-06, "loss": 0.1082, "step": 420 }, { "epoch": 0.10727481207797171, "grad_norm": 67.55184173583984, "learning_rate": 8.927115188583079e-06, "loss": 0.1792, "step": 421 }, { "epoch": 0.10752962160784814, "grad_norm": 115.46858215332031, "learning_rate": 8.924566768603466e-06, "loss": 0.4021, "step": 422 }, { "epoch": 0.10778443113772455, "grad_norm": 116.77753448486328, "learning_rate": 8.922018348623855e-06, "loss": 0.5414, "step": 423 }, { "epoch": 0.10803924066760097, "grad_norm": 29.826536178588867, "learning_rate": 8.919469928644241e-06, "loss": 0.0703, "step": 424 }, { "epoch": 0.10829405019747738, "grad_norm": 674.0891723632812, "learning_rate": 8.916921508664628e-06, "loss": 0.5423, "step": 425 }, { "epoch": 0.1085488597273538, "grad_norm": 81.72020721435547, "learning_rate": 8.914373088685015e-06, "loss": 0.3333, "step": 426 }, { "epoch": 0.10880366925723021, "grad_norm": 103.58757781982422, "learning_rate": 8.911824668705404e-06, "loss": 0.4667, "step": 427 }, { "epoch": 0.10905847878710664, "grad_norm": 247.8717041015625, "learning_rate": 8.909276248725791e-06, "loss": 0.1907, "step": 428 }, { "epoch": 0.10931328831698306, "grad_norm": 10.038141250610352, "learning_rate": 8.906727828746178e-06, "loss": 0.0318, "step": 429 }, { "epoch": 0.10956809784685947, "grad_norm": 5.034185886383057, "learning_rate": 8.904179408766565e-06, "loss": 0.0157, "step": 430 }, { "epoch": 0.10982290737673589, "grad_norm": 399.8277893066406, "learning_rate": 8.901630988786952e-06, "loss": 1.6914, "step": 431 }, { "epoch": 0.1100777169066123, "grad_norm": 67.76985168457031, "learning_rate": 8.89908256880734e-06, "loss": 0.1604, "step": 432 }, { "epoch": 0.11033252643648872, "grad_norm": 17.830453872680664, "learning_rate": 8.896534148827727e-06, "loss": 0.0236, "step": 433 }, { "epoch": 0.11058733596636514, "grad_norm": 10.489675521850586, "learning_rate": 8.893985728848114e-06, "loss": 0.0167, "step": 434 }, { "epoch": 0.11084214549624156, "grad_norm": 3.460163116455078, "learning_rate": 8.891437308868503e-06, "loss": 0.0081, "step": 435 }, { "epoch": 0.11109695502611798, "grad_norm": 70.01533508300781, "learning_rate": 8.888888888888888e-06, "loss": 0.2075, "step": 436 }, { "epoch": 0.1113517645559944, "grad_norm": 67.42399597167969, "learning_rate": 8.886340468909277e-06, "loss": 0.0825, "step": 437 }, { "epoch": 0.11160657408587081, "grad_norm": 28.540929794311523, "learning_rate": 8.883792048929664e-06, "loss": 0.0439, "step": 438 }, { "epoch": 0.11186138361574723, "grad_norm": 515.7363891601562, "learning_rate": 8.881243628950051e-06, "loss": 3.0013, "step": 439 }, { "epoch": 0.11211619314562364, "grad_norm": 348.6291809082031, "learning_rate": 8.87869520897044e-06, "loss": 1.4462, "step": 440 }, { "epoch": 0.11237100267550007, "grad_norm": 74.01558685302734, "learning_rate": 8.876146788990826e-06, "loss": 0.1692, "step": 441 }, { "epoch": 0.11262581220537649, "grad_norm": 20.544443130493164, "learning_rate": 8.873598369011213e-06, "loss": 0.0306, "step": 442 }, { "epoch": 0.1128806217352529, "grad_norm": 876.9181518554688, "learning_rate": 8.871049949031602e-06, "loss": 0.4661, "step": 443 }, { "epoch": 0.11313543126512932, "grad_norm": 11.187724113464355, "learning_rate": 8.868501529051989e-06, "loss": 0.0111, "step": 444 }, { "epoch": 0.11339024079500573, "grad_norm": 5.6719865798950195, "learning_rate": 8.865953109072376e-06, "loss": 0.0028, "step": 445 }, { "epoch": 0.11364505032488215, "grad_norm": 10.441142082214355, "learning_rate": 8.863404689092763e-06, "loss": 0.0098, "step": 446 }, { "epoch": 0.11389985985475856, "grad_norm": 8.888696670532227, "learning_rate": 8.86085626911315e-06, "loss": 0.0118, "step": 447 }, { "epoch": 0.11415466938463499, "grad_norm": 45.65511703491211, "learning_rate": 8.858307849133539e-06, "loss": 0.0542, "step": 448 }, { "epoch": 0.1144094789145114, "grad_norm": 15.516458511352539, "learning_rate": 8.855759429153925e-06, "loss": 0.0123, "step": 449 }, { "epoch": 0.11466428844438782, "grad_norm": 9.798154830932617, "learning_rate": 8.853211009174312e-06, "loss": 0.0264, "step": 450 }, { "epoch": 0.11491909797426424, "grad_norm": 97.52962493896484, "learning_rate": 8.8506625891947e-06, "loss": 0.0917, "step": 451 }, { "epoch": 0.11517390750414065, "grad_norm": 361.8843078613281, "learning_rate": 8.848114169215088e-06, "loss": 3.1079, "step": 452 }, { "epoch": 0.11542871703401707, "grad_norm": 37.37089538574219, "learning_rate": 8.845565749235475e-06, "loss": 0.1671, "step": 453 }, { "epoch": 0.1156835265638935, "grad_norm": 12.460262298583984, "learning_rate": 8.843017329255862e-06, "loss": 0.0231, "step": 454 }, { "epoch": 0.11593833609376991, "grad_norm": 240.99464416503906, "learning_rate": 8.840468909276249e-06, "loss": 0.0922, "step": 455 }, { "epoch": 0.11619314562364633, "grad_norm": 31.687517166137695, "learning_rate": 8.837920489296636e-06, "loss": 0.0481, "step": 456 }, { "epoch": 0.11644795515352274, "grad_norm": 2.685251474380493, "learning_rate": 8.835372069317024e-06, "loss": 0.0031, "step": 457 }, { "epoch": 0.11670276468339916, "grad_norm": 25.355867385864258, "learning_rate": 8.832823649337411e-06, "loss": 0.0215, "step": 458 }, { "epoch": 0.11695757421327557, "grad_norm": 39.05842971801758, "learning_rate": 8.830275229357798e-06, "loss": 0.0623, "step": 459 }, { "epoch": 0.11721238374315199, "grad_norm": 25.282665252685547, "learning_rate": 8.827726809378187e-06, "loss": 0.0249, "step": 460 }, { "epoch": 0.11746719327302842, "grad_norm": 19.603761672973633, "learning_rate": 8.825178389398574e-06, "loss": 0.0286, "step": 461 }, { "epoch": 0.11772200280290483, "grad_norm": 55.164833068847656, "learning_rate": 8.822629969418961e-06, "loss": 0.1105, "step": 462 }, { "epoch": 0.11797681233278125, "grad_norm": 29.501110076904297, "learning_rate": 8.82008154943935e-06, "loss": 0.0452, "step": 463 }, { "epoch": 0.11823162186265766, "grad_norm": 372.2466125488281, "learning_rate": 8.817533129459735e-06, "loss": 0.5371, "step": 464 }, { "epoch": 0.11848643139253408, "grad_norm": 699.8214111328125, "learning_rate": 8.814984709480123e-06, "loss": 0.9578, "step": 465 }, { "epoch": 0.1187412409224105, "grad_norm": 226.0756072998047, "learning_rate": 8.81243628950051e-06, "loss": 0.5471, "step": 466 }, { "epoch": 0.11899605045228692, "grad_norm": 21.93132972717285, "learning_rate": 8.809887869520897e-06, "loss": 0.0225, "step": 467 }, { "epoch": 0.11925085998216334, "grad_norm": 25.768108367919922, "learning_rate": 8.807339449541286e-06, "loss": 0.0285, "step": 468 }, { "epoch": 0.11950566951203975, "grad_norm": 8.374308586120605, "learning_rate": 8.804791029561673e-06, "loss": 0.0105, "step": 469 }, { "epoch": 0.11976047904191617, "grad_norm": 189.27609252929688, "learning_rate": 8.80224260958206e-06, "loss": 0.6383, "step": 470 }, { "epoch": 0.12001528857179258, "grad_norm": 74.82435607910156, "learning_rate": 8.799694189602447e-06, "loss": 0.1685, "step": 471 }, { "epoch": 0.120270098101669, "grad_norm": 11.743697166442871, "learning_rate": 8.797145769622834e-06, "loss": 0.0118, "step": 472 }, { "epoch": 0.12052490763154541, "grad_norm": 18.14217758178711, "learning_rate": 8.794597349643222e-06, "loss": 0.0253, "step": 473 }, { "epoch": 0.12077971716142184, "grad_norm": 119.08099365234375, "learning_rate": 8.79204892966361e-06, "loss": 0.4934, "step": 474 }, { "epoch": 0.12103452669129826, "grad_norm": 15.350459098815918, "learning_rate": 8.789500509683996e-06, "loss": 0.0146, "step": 475 }, { "epoch": 0.12128933622117467, "grad_norm": 274.27398681640625, "learning_rate": 8.786952089704383e-06, "loss": 3.4133, "step": 476 }, { "epoch": 0.12154414575105109, "grad_norm": 23.45751953125, "learning_rate": 8.784403669724772e-06, "loss": 0.034, "step": 477 }, { "epoch": 0.1217989552809275, "grad_norm": 169.9596710205078, "learning_rate": 8.781855249745159e-06, "loss": 0.2003, "step": 478 }, { "epoch": 0.12205376481080392, "grad_norm": 205.7943878173828, "learning_rate": 8.779306829765546e-06, "loss": 0.9619, "step": 479 }, { "epoch": 0.12230857434068035, "grad_norm": 3.06111478805542, "learning_rate": 8.776758409785935e-06, "loss": 0.0075, "step": 480 }, { "epoch": 0.12256338387055676, "grad_norm": 30.160781860351562, "learning_rate": 8.77420998980632e-06, "loss": 0.0747, "step": 481 }, { "epoch": 0.12281819340043318, "grad_norm": 78.47171783447266, "learning_rate": 8.771661569826708e-06, "loss": 0.2002, "step": 482 }, { "epoch": 0.1230730029303096, "grad_norm": 30.985435485839844, "learning_rate": 8.769113149847095e-06, "loss": 0.0615, "step": 483 }, { "epoch": 0.12332781246018601, "grad_norm": 36.71837615966797, "learning_rate": 8.766564729867482e-06, "loss": 0.0611, "step": 484 }, { "epoch": 0.12358262199006242, "grad_norm": 29.462932586669922, "learning_rate": 8.764016309887871e-06, "loss": 0.0494, "step": 485 }, { "epoch": 0.12383743151993884, "grad_norm": 90.40989685058594, "learning_rate": 8.761467889908258e-06, "loss": 0.6465, "step": 486 }, { "epoch": 0.12409224104981527, "grad_norm": 84.54421997070312, "learning_rate": 8.758919469928645e-06, "loss": 0.2305, "step": 487 }, { "epoch": 0.12434705057969168, "grad_norm": 16.363229751586914, "learning_rate": 8.756371049949034e-06, "loss": 0.0256, "step": 488 }, { "epoch": 0.1246018601095681, "grad_norm": 23.500694274902344, "learning_rate": 8.753822629969419e-06, "loss": 0.0316, "step": 489 }, { "epoch": 0.12485666963944451, "grad_norm": 94.21842193603516, "learning_rate": 8.751274209989807e-06, "loss": 0.324, "step": 490 }, { "epoch": 0.12511147916932094, "grad_norm": 27.209686279296875, "learning_rate": 8.748725790010194e-06, "loss": 0.0444, "step": 491 }, { "epoch": 0.12536628869919736, "grad_norm": 12.645621299743652, "learning_rate": 8.746177370030581e-06, "loss": 0.0305, "step": 492 }, { "epoch": 0.12562109822907377, "grad_norm": 17.36095428466797, "learning_rate": 8.743628950050968e-06, "loss": 0.0238, "step": 493 }, { "epoch": 0.1258759077589502, "grad_norm": 4.2269978523254395, "learning_rate": 8.741080530071357e-06, "loss": 0.0086, "step": 494 }, { "epoch": 0.1261307172888266, "grad_norm": 31.060813903808594, "learning_rate": 8.738532110091744e-06, "loss": 0.0345, "step": 495 }, { "epoch": 0.12638552681870302, "grad_norm": 24.781206130981445, "learning_rate": 8.735983690112131e-06, "loss": 0.0327, "step": 496 }, { "epoch": 0.12664033634857944, "grad_norm": 21.878314971923828, "learning_rate": 8.73343527013252e-06, "loss": 0.0264, "step": 497 }, { "epoch": 0.12689514587845585, "grad_norm": 25.292024612426758, "learning_rate": 8.730886850152905e-06, "loss": 0.0296, "step": 498 }, { "epoch": 0.12714995540833227, "grad_norm": 4.126364707946777, "learning_rate": 8.728338430173293e-06, "loss": 0.0053, "step": 499 }, { "epoch": 0.12740476493820868, "grad_norm": 73.13129425048828, "learning_rate": 8.72579001019368e-06, "loss": 0.1467, "step": 500 }, { "epoch": 0.1276595744680851, "grad_norm": 99.20863342285156, "learning_rate": 8.723241590214067e-06, "loss": 0.1534, "step": 501 }, { "epoch": 0.1279143839979615, "grad_norm": 28.300031661987305, "learning_rate": 8.720693170234456e-06, "loss": 0.0477, "step": 502 }, { "epoch": 0.12816919352783795, "grad_norm": 170.3609619140625, "learning_rate": 8.718144750254843e-06, "loss": 0.3476, "step": 503 }, { "epoch": 0.12842400305771437, "grad_norm": 30.597957611083984, "learning_rate": 8.71559633027523e-06, "loss": 0.0491, "step": 504 }, { "epoch": 0.12867881258759079, "grad_norm": 309.4017333984375, "learning_rate": 8.713047910295619e-06, "loss": 0.478, "step": 505 }, { "epoch": 0.1289336221174672, "grad_norm": 16.36326789855957, "learning_rate": 8.710499490316004e-06, "loss": 0.0433, "step": 506 }, { "epoch": 0.12918843164734362, "grad_norm": 15.463802337646484, "learning_rate": 8.707951070336392e-06, "loss": 0.0175, "step": 507 }, { "epoch": 0.12944324117722003, "grad_norm": 6.033719062805176, "learning_rate": 8.70540265035678e-06, "loss": 0.0056, "step": 508 }, { "epoch": 0.12969805070709645, "grad_norm": 16.513174057006836, "learning_rate": 8.702854230377166e-06, "loss": 0.0322, "step": 509 }, { "epoch": 0.12995286023697286, "grad_norm": 92.56959533691406, "learning_rate": 8.700305810397555e-06, "loss": 0.3613, "step": 510 }, { "epoch": 0.13020766976684928, "grad_norm": 84.94668579101562, "learning_rate": 8.697757390417942e-06, "loss": 0.0657, "step": 511 }, { "epoch": 0.1304624792967257, "grad_norm": 165.81846618652344, "learning_rate": 8.695208970438329e-06, "loss": 0.7058, "step": 512 }, { "epoch": 0.1307172888266021, "grad_norm": 28.631532669067383, "learning_rate": 8.692660550458716e-06, "loss": 0.0383, "step": 513 }, { "epoch": 0.13097209835647852, "grad_norm": 33.72076416015625, "learning_rate": 8.690112130479103e-06, "loss": 0.0537, "step": 514 }, { "epoch": 0.13122690788635494, "grad_norm": 71.7044677734375, "learning_rate": 8.687563710499491e-06, "loss": 0.1287, "step": 515 }, { "epoch": 0.13148171741623138, "grad_norm": 46.705047607421875, "learning_rate": 8.685015290519878e-06, "loss": 0.0446, "step": 516 }, { "epoch": 0.1317365269461078, "grad_norm": 2.453301191329956, "learning_rate": 8.682466870540265e-06, "loss": 0.0049, "step": 517 }, { "epoch": 0.1319913364759842, "grad_norm": 386.61944580078125, "learning_rate": 8.679918450560652e-06, "loss": 2.026, "step": 518 }, { "epoch": 0.13224614600586063, "grad_norm": 139.7738800048828, "learning_rate": 8.677370030581041e-06, "loss": 0.395, "step": 519 }, { "epoch": 0.13250095553573704, "grad_norm": 29.409221649169922, "learning_rate": 8.674821610601428e-06, "loss": 0.048, "step": 520 }, { "epoch": 0.13275576506561346, "grad_norm": 22.540849685668945, "learning_rate": 8.672273190621815e-06, "loss": 0.0291, "step": 521 }, { "epoch": 0.13301057459548987, "grad_norm": 24.00151824951172, "learning_rate": 8.669724770642203e-06, "loss": 0.0338, "step": 522 }, { "epoch": 0.1332653841253663, "grad_norm": 18.187442779541016, "learning_rate": 8.667176350662589e-06, "loss": 0.0213, "step": 523 }, { "epoch": 0.1335201936552427, "grad_norm": 7.6271820068359375, "learning_rate": 8.664627930682977e-06, "loss": 0.0064, "step": 524 }, { "epoch": 0.13377500318511912, "grad_norm": 33.70282745361328, "learning_rate": 8.662079510703364e-06, "loss": 0.0439, "step": 525 }, { "epoch": 0.13402981271499553, "grad_norm": 62.14427947998047, "learning_rate": 8.659531090723751e-06, "loss": 0.3197, "step": 526 }, { "epoch": 0.13428462224487195, "grad_norm": 29.642961502075195, "learning_rate": 8.65698267074414e-06, "loss": 0.0516, "step": 527 }, { "epoch": 0.13453943177474836, "grad_norm": 65.708740234375, "learning_rate": 8.654434250764527e-06, "loss": 0.146, "step": 528 }, { "epoch": 0.1347942413046248, "grad_norm": 29.779056549072266, "learning_rate": 8.651885830784914e-06, "loss": 0.0598, "step": 529 }, { "epoch": 0.13504905083450122, "grad_norm": 31.765953063964844, "learning_rate": 8.649337410805302e-06, "loss": 0.0574, "step": 530 }, { "epoch": 0.13530386036437764, "grad_norm": 1.1838963031768799, "learning_rate": 8.646788990825688e-06, "loss": 0.0025, "step": 531 }, { "epoch": 0.13555866989425405, "grad_norm": 107.97193145751953, "learning_rate": 8.644240570846076e-06, "loss": 0.362, "step": 532 }, { "epoch": 0.13581347942413047, "grad_norm": 3.066854476928711, "learning_rate": 8.641692150866463e-06, "loss": 0.0061, "step": 533 }, { "epoch": 0.13606828895400688, "grad_norm": 39.946598052978516, "learning_rate": 8.63914373088685e-06, "loss": 0.0825, "step": 534 }, { "epoch": 0.1363230984838833, "grad_norm": 41.54171371459961, "learning_rate": 8.636595310907239e-06, "loss": 0.0908, "step": 535 }, { "epoch": 0.1365779080137597, "grad_norm": 27.314579010009766, "learning_rate": 8.634046890927626e-06, "loss": 0.037, "step": 536 }, { "epoch": 0.13683271754363613, "grad_norm": 102.2757339477539, "learning_rate": 8.631498470948013e-06, "loss": 0.2031, "step": 537 }, { "epoch": 0.13708752707351254, "grad_norm": 8.11296272277832, "learning_rate": 8.6289500509684e-06, "loss": 0.006, "step": 538 }, { "epoch": 0.13734233660338896, "grad_norm": 106.66088104248047, "learning_rate": 8.626401630988788e-06, "loss": 0.0928, "step": 539 }, { "epoch": 0.13759714613326537, "grad_norm": 22.237829208374023, "learning_rate": 8.623853211009175e-06, "loss": 0.0344, "step": 540 }, { "epoch": 0.1378519556631418, "grad_norm": 9.76301097869873, "learning_rate": 8.621304791029562e-06, "loss": 0.0066, "step": 541 }, { "epoch": 0.13810676519301823, "grad_norm": 22.57248878479004, "learning_rate": 8.61875637104995e-06, "loss": 0.0304, "step": 542 }, { "epoch": 0.13836157472289465, "grad_norm": 73.22361755371094, "learning_rate": 8.616207951070336e-06, "loss": 0.0914, "step": 543 }, { "epoch": 0.13861638425277106, "grad_norm": 32.02495193481445, "learning_rate": 8.613659531090725e-06, "loss": 0.0381, "step": 544 }, { "epoch": 0.13887119378264748, "grad_norm": 13.949383735656738, "learning_rate": 8.611111111111112e-06, "loss": 0.0138, "step": 545 }, { "epoch": 0.1391260033125239, "grad_norm": 21.57050323486328, "learning_rate": 8.608562691131499e-06, "loss": 0.0394, "step": 546 }, { "epoch": 0.1393808128424003, "grad_norm": 4.794432163238525, "learning_rate": 8.606014271151887e-06, "loss": 0.0078, "step": 547 }, { "epoch": 0.13963562237227672, "grad_norm": 4.165365695953369, "learning_rate": 8.603465851172273e-06, "loss": 0.0021, "step": 548 }, { "epoch": 0.13989043190215314, "grad_norm": 20.04294776916504, "learning_rate": 8.600917431192661e-06, "loss": 0.0406, "step": 549 }, { "epoch": 0.14014524143202955, "grad_norm": 9.379389762878418, "learning_rate": 8.598369011213048e-06, "loss": 0.0057, "step": 550 }, { "epoch": 0.14040005096190597, "grad_norm": 128.825439453125, "learning_rate": 8.595820591233435e-06, "loss": 0.9618, "step": 551 }, { "epoch": 0.14065486049178239, "grad_norm": 116.0452651977539, "learning_rate": 8.593272171253824e-06, "loss": 0.2168, "step": 552 }, { "epoch": 0.1409096700216588, "grad_norm": 62.865299224853516, "learning_rate": 8.590723751274211e-06, "loss": 0.0361, "step": 553 }, { "epoch": 0.14116447955153522, "grad_norm": 3.876119613647461, "learning_rate": 8.588175331294598e-06, "loss": 0.0027, "step": 554 }, { "epoch": 0.14141928908141163, "grad_norm": 87.6649398803711, "learning_rate": 8.585626911314986e-06, "loss": 0.1467, "step": 555 }, { "epoch": 0.14167409861128807, "grad_norm": 4.8600335121154785, "learning_rate": 8.583078491335373e-06, "loss": 0.0098, "step": 556 }, { "epoch": 0.1419289081411645, "grad_norm": 1.3271070718765259, "learning_rate": 8.58053007135576e-06, "loss": 0.002, "step": 557 }, { "epoch": 0.1421837176710409, "grad_norm": 38.83177185058594, "learning_rate": 8.577981651376147e-06, "loss": 0.0746, "step": 558 }, { "epoch": 0.14243852720091732, "grad_norm": 11.678545951843262, "learning_rate": 8.575433231396534e-06, "loss": 0.0086, "step": 559 }, { "epoch": 0.14269333673079374, "grad_norm": 9.289490699768066, "learning_rate": 8.572884811416923e-06, "loss": 0.0052, "step": 560 }, { "epoch": 0.14294814626067015, "grad_norm": 12.72591495513916, "learning_rate": 8.57033639143731e-06, "loss": 0.0105, "step": 561 }, { "epoch": 0.14320295579054657, "grad_norm": 70.24102020263672, "learning_rate": 8.567787971457697e-06, "loss": 0.4242, "step": 562 }, { "epoch": 0.14345776532042298, "grad_norm": 14.474027633666992, "learning_rate": 8.565239551478084e-06, "loss": 0.0129, "step": 563 }, { "epoch": 0.1437125748502994, "grad_norm": 5.702688694000244, "learning_rate": 8.562691131498472e-06, "loss": 0.0033, "step": 564 }, { "epoch": 0.1439673843801758, "grad_norm": 95.22308349609375, "learning_rate": 8.56014271151886e-06, "loss": 0.264, "step": 565 }, { "epoch": 0.14422219391005223, "grad_norm": 75.52281188964844, "learning_rate": 8.557594291539246e-06, "loss": 0.0695, "step": 566 }, { "epoch": 0.14447700343992864, "grad_norm": 17.631776809692383, "learning_rate": 8.555045871559633e-06, "loss": 0.0204, "step": 567 }, { "epoch": 0.14473181296980506, "grad_norm": 312.894287109375, "learning_rate": 8.55249745158002e-06, "loss": 0.1224, "step": 568 }, { "epoch": 0.1449866224996815, "grad_norm": 2.809800386428833, "learning_rate": 8.549949031600409e-06, "loss": 0.0038, "step": 569 }, { "epoch": 0.14524143202955792, "grad_norm": 12.962081909179688, "learning_rate": 8.547400611620796e-06, "loss": 0.006, "step": 570 }, { "epoch": 0.14549624155943433, "grad_norm": 8.897120475769043, "learning_rate": 8.544852191641183e-06, "loss": 0.0126, "step": 571 }, { "epoch": 0.14575105108931075, "grad_norm": 97.79730987548828, "learning_rate": 8.542303771661571e-06, "loss": 0.377, "step": 572 }, { "epoch": 0.14600586061918716, "grad_norm": 13.457335472106934, "learning_rate": 8.539755351681958e-06, "loss": 0.012, "step": 573 }, { "epoch": 0.14626067014906358, "grad_norm": 283.26617431640625, "learning_rate": 8.537206931702345e-06, "loss": 0.4277, "step": 574 }, { "epoch": 0.14651547967894, "grad_norm": 10.631940841674805, "learning_rate": 8.534658511722732e-06, "loss": 0.0124, "step": 575 }, { "epoch": 0.1467702892088164, "grad_norm": 3.655259609222412, "learning_rate": 8.53211009174312e-06, "loss": 0.0063, "step": 576 }, { "epoch": 0.14702509873869282, "grad_norm": 7.5835795402526855, "learning_rate": 8.529561671763508e-06, "loss": 0.0077, "step": 577 }, { "epoch": 0.14727990826856924, "grad_norm": 3.833108901977539, "learning_rate": 8.527013251783895e-06, "loss": 0.0026, "step": 578 }, { "epoch": 0.14753471779844565, "grad_norm": 2.393767833709717, "learning_rate": 8.524464831804282e-06, "loss": 0.0021, "step": 579 }, { "epoch": 0.14778952732832207, "grad_norm": 1797.753173828125, "learning_rate": 8.521916411824669e-06, "loss": 0.9237, "step": 580 }, { "epoch": 0.14804433685819848, "grad_norm": 3.9528253078460693, "learning_rate": 8.519367991845057e-06, "loss": 0.0135, "step": 581 }, { "epoch": 0.14829914638807493, "grad_norm": 143.27078247070312, "learning_rate": 8.516819571865444e-06, "loss": 0.0466, "step": 582 }, { "epoch": 0.14855395591795134, "grad_norm": 16.966739654541016, "learning_rate": 8.514271151885831e-06, "loss": 0.0179, "step": 583 }, { "epoch": 0.14880876544782776, "grad_norm": 18.038198471069336, "learning_rate": 8.511722731906218e-06, "loss": 0.0245, "step": 584 }, { "epoch": 0.14906357497770417, "grad_norm": 68.3419418334961, "learning_rate": 8.509174311926605e-06, "loss": 0.0908, "step": 585 }, { "epoch": 0.1493183845075806, "grad_norm": 6.604894161224365, "learning_rate": 8.506625891946994e-06, "loss": 0.0042, "step": 586 }, { "epoch": 0.149573194037457, "grad_norm": 8.759491920471191, "learning_rate": 8.50407747196738e-06, "loss": 0.0083, "step": 587 }, { "epoch": 0.14982800356733342, "grad_norm": 22.55419921875, "learning_rate": 8.501529051987768e-06, "loss": 0.0296, "step": 588 }, { "epoch": 0.15008281309720983, "grad_norm": 28.120960235595703, "learning_rate": 8.498980632008156e-06, "loss": 0.0447, "step": 589 }, { "epoch": 0.15033762262708625, "grad_norm": 72.25299835205078, "learning_rate": 8.496432212028543e-06, "loss": 0.0973, "step": 590 }, { "epoch": 0.15059243215696266, "grad_norm": 114.5904769897461, "learning_rate": 8.49388379204893e-06, "loss": 0.089, "step": 591 }, { "epoch": 0.15084724168683908, "grad_norm": 18.136537551879883, "learning_rate": 8.491335372069319e-06, "loss": 0.0356, "step": 592 }, { "epoch": 0.1511020512167155, "grad_norm": 7.583436489105225, "learning_rate": 8.488786952089704e-06, "loss": 0.0042, "step": 593 }, { "epoch": 0.1513568607465919, "grad_norm": 92.7400894165039, "learning_rate": 8.486238532110093e-06, "loss": 0.0406, "step": 594 }, { "epoch": 0.15161167027646835, "grad_norm": 222.20558166503906, "learning_rate": 8.48369011213048e-06, "loss": 0.4065, "step": 595 }, { "epoch": 0.15186647980634477, "grad_norm": 146.93165588378906, "learning_rate": 8.481141692150867e-06, "loss": 0.0771, "step": 596 }, { "epoch": 0.15212128933622118, "grad_norm": 22.145092010498047, "learning_rate": 8.478593272171255e-06, "loss": 0.0297, "step": 597 }, { "epoch": 0.1523760988660976, "grad_norm": 103.70513153076172, "learning_rate": 8.476044852191642e-06, "loss": 0.0782, "step": 598 }, { "epoch": 0.152630908395974, "grad_norm": 137.55499267578125, "learning_rate": 8.47349643221203e-06, "loss": 0.1898, "step": 599 }, { "epoch": 0.15288571792585043, "grad_norm": 29.127765655517578, "learning_rate": 8.470948012232416e-06, "loss": 0.0502, "step": 600 }, { "epoch": 0.15314052745572684, "grad_norm": 19.205307006835938, "learning_rate": 8.468399592252803e-06, "loss": 0.025, "step": 601 }, { "epoch": 0.15339533698560326, "grad_norm": 11.708564758300781, "learning_rate": 8.465851172273192e-06, "loss": 0.0159, "step": 602 }, { "epoch": 0.15365014651547967, "grad_norm": 13.64586353302002, "learning_rate": 8.463302752293579e-06, "loss": 0.025, "step": 603 }, { "epoch": 0.1539049560453561, "grad_norm": 6.630014896392822, "learning_rate": 8.460754332313966e-06, "loss": 0.0088, "step": 604 }, { "epoch": 0.1541597655752325, "grad_norm": 46.29561233520508, "learning_rate": 8.458205912334353e-06, "loss": 0.0941, "step": 605 }, { "epoch": 0.15441457510510892, "grad_norm": 21.089475631713867, "learning_rate": 8.455657492354741e-06, "loss": 0.028, "step": 606 }, { "epoch": 0.15466938463498534, "grad_norm": 30.900400161743164, "learning_rate": 8.453109072375128e-06, "loss": 0.055, "step": 607 }, { "epoch": 0.15492419416486178, "grad_norm": 329.75885009765625, "learning_rate": 8.450560652395515e-06, "loss": 0.3733, "step": 608 }, { "epoch": 0.1551790036947382, "grad_norm": 17.52680778503418, "learning_rate": 8.448012232415902e-06, "loss": 0.0203, "step": 609 }, { "epoch": 0.1554338132246146, "grad_norm": 28.002962112426758, "learning_rate": 8.44546381243629e-06, "loss": 0.0334, "step": 610 }, { "epoch": 0.15568862275449102, "grad_norm": 3.475905656814575, "learning_rate": 8.442915392456678e-06, "loss": 0.003, "step": 611 }, { "epoch": 0.15594343228436744, "grad_norm": 9.859709739685059, "learning_rate": 8.440366972477065e-06, "loss": 0.0097, "step": 612 }, { "epoch": 0.15619824181424385, "grad_norm": 5.394769668579102, "learning_rate": 8.437818552497452e-06, "loss": 0.0065, "step": 613 }, { "epoch": 0.15645305134412027, "grad_norm": 24.128131866455078, "learning_rate": 8.43527013251784e-06, "loss": 0.0266, "step": 614 }, { "epoch": 0.15670786087399668, "grad_norm": 70.71337127685547, "learning_rate": 8.432721712538227e-06, "loss": 0.0449, "step": 615 }, { "epoch": 0.1569626704038731, "grad_norm": 30.41301155090332, "learning_rate": 8.430173292558614e-06, "loss": 0.0558, "step": 616 }, { "epoch": 0.15721747993374952, "grad_norm": 20.49118995666504, "learning_rate": 8.427624872579003e-06, "loss": 0.0258, "step": 617 }, { "epoch": 0.15747228946362593, "grad_norm": 23.030391693115234, "learning_rate": 8.425076452599388e-06, "loss": 0.031, "step": 618 }, { "epoch": 0.15772709899350235, "grad_norm": 52.30411911010742, "learning_rate": 8.422528032619777e-06, "loss": 0.0197, "step": 619 }, { "epoch": 0.15798190852337876, "grad_norm": 304.415283203125, "learning_rate": 8.419979612640164e-06, "loss": 3.094, "step": 620 }, { "epoch": 0.1582367180532552, "grad_norm": 2.6185243129730225, "learning_rate": 8.41743119266055e-06, "loss": 0.0057, "step": 621 }, { "epoch": 0.15849152758313162, "grad_norm": 17.818408966064453, "learning_rate": 8.41488277268094e-06, "loss": 0.0196, "step": 622 }, { "epoch": 0.15874633711300803, "grad_norm": 3.9391210079193115, "learning_rate": 8.412334352701326e-06, "loss": 0.0026, "step": 623 }, { "epoch": 0.15900114664288445, "grad_norm": 7.706480503082275, "learning_rate": 8.409785932721713e-06, "loss": 0.0059, "step": 624 }, { "epoch": 0.15925595617276087, "grad_norm": 4.161251544952393, "learning_rate": 8.4072375127421e-06, "loss": 0.0044, "step": 625 }, { "epoch": 0.15951076570263728, "grad_norm": 55.80327606201172, "learning_rate": 8.404689092762487e-06, "loss": 0.3439, "step": 626 }, { "epoch": 0.1597655752325137, "grad_norm": 9.804763793945312, "learning_rate": 8.402140672782876e-06, "loss": 0.0137, "step": 627 }, { "epoch": 0.1600203847623901, "grad_norm": 4.637109279632568, "learning_rate": 8.399592252803263e-06, "loss": 0.0032, "step": 628 }, { "epoch": 0.16027519429226653, "grad_norm": 7.809397220611572, "learning_rate": 8.39704383282365e-06, "loss": 0.007, "step": 629 }, { "epoch": 0.16053000382214294, "grad_norm": 65.1230239868164, "learning_rate": 8.394495412844037e-06, "loss": 0.5163, "step": 630 }, { "epoch": 0.16078481335201936, "grad_norm": 9.367552757263184, "learning_rate": 8.391946992864425e-06, "loss": 0.0063, "step": 631 }, { "epoch": 0.16103962288189577, "grad_norm": 214.69607543945312, "learning_rate": 8.389398572884812e-06, "loss": 0.1503, "step": 632 }, { "epoch": 0.1612944324117722, "grad_norm": 95.27933502197266, "learning_rate": 8.3868501529052e-06, "loss": 0.0903, "step": 633 }, { "epoch": 0.16154924194164863, "grad_norm": 79.68666076660156, "learning_rate": 8.384301732925588e-06, "loss": 0.6071, "step": 634 }, { "epoch": 0.16180405147152505, "grad_norm": 10.480917930603027, "learning_rate": 8.381753312945973e-06, "loss": 0.0079, "step": 635 }, { "epoch": 0.16205886100140146, "grad_norm": 1.261254072189331, "learning_rate": 8.379204892966362e-06, "loss": 0.0014, "step": 636 }, { "epoch": 0.16231367053127788, "grad_norm": 247.7661895751953, "learning_rate": 8.376656472986749e-06, "loss": 0.2547, "step": 637 }, { "epoch": 0.1625684800611543, "grad_norm": 4.608572006225586, "learning_rate": 8.374108053007136e-06, "loss": 0.0047, "step": 638 }, { "epoch": 0.1628232895910307, "grad_norm": 2.257286787033081, "learning_rate": 8.371559633027524e-06, "loss": 0.0025, "step": 639 }, { "epoch": 0.16307809912090712, "grad_norm": 5.184882640838623, "learning_rate": 8.369011213047911e-06, "loss": 0.0038, "step": 640 }, { "epoch": 0.16333290865078354, "grad_norm": 10.772210121154785, "learning_rate": 8.366462793068298e-06, "loss": 0.0082, "step": 641 }, { "epoch": 0.16358771818065995, "grad_norm": 19.935495376586914, "learning_rate": 8.363914373088685e-06, "loss": 0.0275, "step": 642 }, { "epoch": 0.16384252771053637, "grad_norm": 26.0557918548584, "learning_rate": 8.361365953109072e-06, "loss": 0.0509, "step": 643 }, { "epoch": 0.16409733724041278, "grad_norm": 0.5837207436561584, "learning_rate": 8.35881753312946e-06, "loss": 0.0017, "step": 644 }, { "epoch": 0.1643521467702892, "grad_norm": 41.02202606201172, "learning_rate": 8.356269113149848e-06, "loss": 0.0342, "step": 645 }, { "epoch": 0.1646069563001656, "grad_norm": 1.1491689682006836, "learning_rate": 8.353720693170235e-06, "loss": 0.0024, "step": 646 }, { "epoch": 0.16486176583004206, "grad_norm": 2350.91162109375, "learning_rate": 8.351172273190622e-06, "loss": 0.9882, "step": 647 }, { "epoch": 0.16511657535991847, "grad_norm": 18.017988204956055, "learning_rate": 8.34862385321101e-06, "loss": 0.0204, "step": 648 }, { "epoch": 0.1653713848897949, "grad_norm": 12.318140029907227, "learning_rate": 8.346075433231397e-06, "loss": 0.0129, "step": 649 }, { "epoch": 0.1656261944196713, "grad_norm": 8.462002754211426, "learning_rate": 8.343527013251784e-06, "loss": 0.0051, "step": 650 }, { "epoch": 0.16588100394954772, "grad_norm": 2.1949338912963867, "learning_rate": 8.340978593272173e-06, "loss": 0.0018, "step": 651 }, { "epoch": 0.16613581347942413, "grad_norm": 13.564543724060059, "learning_rate": 8.338430173292558e-06, "loss": 0.0129, "step": 652 }, { "epoch": 0.16639062300930055, "grad_norm": 8.67857837677002, "learning_rate": 8.335881753312947e-06, "loss": 0.0051, "step": 653 }, { "epoch": 0.16664543253917696, "grad_norm": 442.46337890625, "learning_rate": 8.333333333333334e-06, "loss": 3.1172, "step": 654 }, { "epoch": 0.16690024206905338, "grad_norm": 542.9337768554688, "learning_rate": 8.33078491335372e-06, "loss": 1.8564, "step": 655 }, { "epoch": 0.1671550515989298, "grad_norm": 6.099351406097412, "learning_rate": 8.32823649337411e-06, "loss": 0.0093, "step": 656 }, { "epoch": 0.1674098611288062, "grad_norm": 8.745729446411133, "learning_rate": 8.325688073394496e-06, "loss": 0.0051, "step": 657 }, { "epoch": 0.16766467065868262, "grad_norm": 11.187594413757324, "learning_rate": 8.323139653414883e-06, "loss": 0.0078, "step": 658 }, { "epoch": 0.16791948018855904, "grad_norm": 4.602405071258545, "learning_rate": 8.320591233435272e-06, "loss": 0.0034, "step": 659 }, { "epoch": 0.16817428971843548, "grad_norm": 787.7642822265625, "learning_rate": 8.318042813455657e-06, "loss": 0.3943, "step": 660 }, { "epoch": 0.1684290992483119, "grad_norm": 17.031156539916992, "learning_rate": 8.315494393476046e-06, "loss": 0.0109, "step": 661 }, { "epoch": 0.1686839087781883, "grad_norm": 83.1441421508789, "learning_rate": 8.312945973496433e-06, "loss": 0.2218, "step": 662 }, { "epoch": 0.16893871830806473, "grad_norm": 23.66316795349121, "learning_rate": 8.31039755351682e-06, "loss": 0.0328, "step": 663 }, { "epoch": 0.16919352783794114, "grad_norm": 15.793343544006348, "learning_rate": 8.307849133537208e-06, "loss": 0.0164, "step": 664 }, { "epoch": 0.16944833736781756, "grad_norm": 14.819869995117188, "learning_rate": 8.305300713557595e-06, "loss": 0.0139, "step": 665 }, { "epoch": 0.16970314689769397, "grad_norm": 10.905662536621094, "learning_rate": 8.302752293577982e-06, "loss": 0.0084, "step": 666 }, { "epoch": 0.1699579564275704, "grad_norm": 2.22351336479187, "learning_rate": 8.30020387359837e-06, "loss": 0.0007, "step": 667 }, { "epoch": 0.1702127659574468, "grad_norm": 3.1164894104003906, "learning_rate": 8.297655453618758e-06, "loss": 0.0016, "step": 668 }, { "epoch": 0.17046757548732322, "grad_norm": 4.391525745391846, "learning_rate": 8.295107033639145e-06, "loss": 0.0025, "step": 669 }, { "epoch": 0.17072238501719963, "grad_norm": 8.582864761352539, "learning_rate": 8.292558613659532e-06, "loss": 0.0055, "step": 670 }, { "epoch": 0.17097719454707605, "grad_norm": 21.17778968811035, "learning_rate": 8.290010193679919e-06, "loss": 0.0308, "step": 671 }, { "epoch": 0.17123200407695247, "grad_norm": 33.625057220458984, "learning_rate": 8.287461773700306e-06, "loss": 0.0348, "step": 672 }, { "epoch": 0.1714868136068289, "grad_norm": 4.713667392730713, "learning_rate": 8.284913353720694e-06, "loss": 0.0037, "step": 673 }, { "epoch": 0.17174162313670532, "grad_norm": 12.505748748779297, "learning_rate": 8.282364933741081e-06, "loss": 0.0099, "step": 674 }, { "epoch": 0.17199643266658174, "grad_norm": 15.416276931762695, "learning_rate": 8.279816513761468e-06, "loss": 0.0146, "step": 675 }, { "epoch": 0.17225124219645815, "grad_norm": 25.314815521240234, "learning_rate": 8.277268093781857e-06, "loss": 0.0136, "step": 676 }, { "epoch": 0.17250605172633457, "grad_norm": 18.105253219604492, "learning_rate": 8.274719673802242e-06, "loss": 0.0176, "step": 677 }, { "epoch": 0.17276086125621098, "grad_norm": 13.182333946228027, "learning_rate": 8.27217125382263e-06, "loss": 0.0212, "step": 678 }, { "epoch": 0.1730156707860874, "grad_norm": 8.471870422363281, "learning_rate": 8.269622833843018e-06, "loss": 0.007, "step": 679 }, { "epoch": 0.17327048031596382, "grad_norm": 42.56953430175781, "learning_rate": 8.267074413863405e-06, "loss": 0.0345, "step": 680 }, { "epoch": 0.17352528984584023, "grad_norm": 160.38076782226562, "learning_rate": 8.264525993883793e-06, "loss": 3.4714, "step": 681 }, { "epoch": 0.17378009937571665, "grad_norm": 9.037978172302246, "learning_rate": 8.26197757390418e-06, "loss": 0.0166, "step": 682 }, { "epoch": 0.17403490890559306, "grad_norm": 192.0311737060547, "learning_rate": 8.259429153924567e-06, "loss": 0.79, "step": 683 }, { "epoch": 0.17428971843546948, "grad_norm": 158.95333862304688, "learning_rate": 8.256880733944956e-06, "loss": 0.6683, "step": 684 }, { "epoch": 0.1745445279653459, "grad_norm": 4.719536781311035, "learning_rate": 8.254332313965341e-06, "loss": 0.0039, "step": 685 }, { "epoch": 0.17479933749522233, "grad_norm": 568.3900756835938, "learning_rate": 8.25178389398573e-06, "loss": 0.3932, "step": 686 }, { "epoch": 0.17505414702509875, "grad_norm": 4.543330192565918, "learning_rate": 8.249235474006117e-06, "loss": 0.0071, "step": 687 }, { "epoch": 0.17530895655497516, "grad_norm": 2.6091468334198, "learning_rate": 8.246687054026504e-06, "loss": 0.0064, "step": 688 }, { "epoch": 0.17556376608485158, "grad_norm": 5.517385482788086, "learning_rate": 8.244138634046892e-06, "loss": 0.003, "step": 689 }, { "epoch": 0.175818575614728, "grad_norm": 107.25810241699219, "learning_rate": 8.24159021406728e-06, "loss": 0.2546, "step": 690 }, { "epoch": 0.1760733851446044, "grad_norm": 13.176841735839844, "learning_rate": 8.239041794087666e-06, "loss": 0.0147, "step": 691 }, { "epoch": 0.17632819467448083, "grad_norm": 2.8126046657562256, "learning_rate": 8.236493374108053e-06, "loss": 0.0088, "step": 692 }, { "epoch": 0.17658300420435724, "grad_norm": 11.951362609863281, "learning_rate": 8.233944954128442e-06, "loss": 0.0123, "step": 693 }, { "epoch": 0.17683781373423366, "grad_norm": 5.160869121551514, "learning_rate": 8.231396534148829e-06, "loss": 0.0041, "step": 694 }, { "epoch": 0.17709262326411007, "grad_norm": 5.930375576019287, "learning_rate": 8.228848114169216e-06, "loss": 0.0071, "step": 695 }, { "epoch": 0.1773474327939865, "grad_norm": 6.686774730682373, "learning_rate": 8.226299694189603e-06, "loss": 0.0045, "step": 696 }, { "epoch": 0.1776022423238629, "grad_norm": 919.2767944335938, "learning_rate": 8.22375127420999e-06, "loss": 0.2872, "step": 697 }, { "epoch": 0.17785705185373932, "grad_norm": 11.843929290771484, "learning_rate": 8.221202854230378e-06, "loss": 0.0114, "step": 698 }, { "epoch": 0.17811186138361576, "grad_norm": 3.170184850692749, "learning_rate": 8.218654434250765e-06, "loss": 0.0024, "step": 699 }, { "epoch": 0.17836667091349218, "grad_norm": 9.458320617675781, "learning_rate": 8.216106014271152e-06, "loss": 0.0167, "step": 700 }, { "epoch": 0.1786214804433686, "grad_norm": 1.5410457849502563, "learning_rate": 8.21355759429154e-06, "loss": 0.0015, "step": 701 }, { "epoch": 0.178876289973245, "grad_norm": 448.32257080078125, "learning_rate": 8.211009174311926e-06, "loss": 2.8105, "step": 702 }, { "epoch": 0.17913109950312142, "grad_norm": 23.701892852783203, "learning_rate": 8.208460754332315e-06, "loss": 0.0756, "step": 703 }, { "epoch": 0.17938590903299784, "grad_norm": 200.4190216064453, "learning_rate": 8.205912334352702e-06, "loss": 3.0016, "step": 704 }, { "epoch": 0.17964071856287425, "grad_norm": 26.422119140625, "learning_rate": 8.203363914373089e-06, "loss": 0.0202, "step": 705 }, { "epoch": 0.17989552809275067, "grad_norm": 8.571760177612305, "learning_rate": 8.200815494393477e-06, "loss": 0.0053, "step": 706 }, { "epoch": 0.18015033762262708, "grad_norm": 4.7402238845825195, "learning_rate": 8.198267074413864e-06, "loss": 0.0036, "step": 707 }, { "epoch": 0.1804051471525035, "grad_norm": 5.5966668128967285, "learning_rate": 8.195718654434251e-06, "loss": 0.0039, "step": 708 }, { "epoch": 0.1806599566823799, "grad_norm": 275.29461669921875, "learning_rate": 8.19317023445464e-06, "loss": 0.1491, "step": 709 }, { "epoch": 0.18091476621225633, "grad_norm": 4.091842174530029, "learning_rate": 8.190621814475027e-06, "loss": 0.0058, "step": 710 }, { "epoch": 0.18116957574213274, "grad_norm": 2.1052355766296387, "learning_rate": 8.188073394495414e-06, "loss": 0.004, "step": 711 }, { "epoch": 0.1814243852720092, "grad_norm": 1.5857720375061035, "learning_rate": 8.1855249745158e-06, "loss": 0.0039, "step": 712 }, { "epoch": 0.1816791948018856, "grad_norm": 10.00837230682373, "learning_rate": 8.182976554536188e-06, "loss": 0.0107, "step": 713 }, { "epoch": 0.18193400433176202, "grad_norm": 8.12230110168457, "learning_rate": 8.180428134556576e-06, "loss": 0.008, "step": 714 }, { "epoch": 0.18218881386163843, "grad_norm": 482.1853942871094, "learning_rate": 8.177879714576963e-06, "loss": 1.3916, "step": 715 }, { "epoch": 0.18244362339151485, "grad_norm": 13.029902458190918, "learning_rate": 8.17533129459735e-06, "loss": 0.0153, "step": 716 }, { "epoch": 0.18269843292139126, "grad_norm": 12.420220375061035, "learning_rate": 8.172782874617737e-06, "loss": 0.0184, "step": 717 }, { "epoch": 0.18295324245126768, "grad_norm": 9.766034126281738, "learning_rate": 8.170234454638126e-06, "loss": 0.0076, "step": 718 }, { "epoch": 0.1832080519811441, "grad_norm": 4.201150417327881, "learning_rate": 8.167686034658511e-06, "loss": 0.0017, "step": 719 }, { "epoch": 0.1834628615110205, "grad_norm": 11.937125205993652, "learning_rate": 8.1651376146789e-06, "loss": 0.0152, "step": 720 }, { "epoch": 0.18371767104089692, "grad_norm": 5.041351795196533, "learning_rate": 8.162589194699287e-06, "loss": 0.0031, "step": 721 }, { "epoch": 0.18397248057077334, "grad_norm": 1.476393699645996, "learning_rate": 8.160040774719674e-06, "loss": 0.0033, "step": 722 }, { "epoch": 0.18422729010064975, "grad_norm": 0.8453362584114075, "learning_rate": 8.157492354740062e-06, "loss": 0.0014, "step": 723 }, { "epoch": 0.18448209963052617, "grad_norm": 5.35353422164917, "learning_rate": 8.15494393476045e-06, "loss": 0.0035, "step": 724 }, { "epoch": 0.1847369091604026, "grad_norm": 6.762155532836914, "learning_rate": 8.152395514780836e-06, "loss": 0.0053, "step": 725 }, { "epoch": 0.18499171869027903, "grad_norm": 172.69961547851562, "learning_rate": 8.149847094801225e-06, "loss": 0.1633, "step": 726 }, { "epoch": 0.18524652822015544, "grad_norm": 228.3174285888672, "learning_rate": 8.147298674821612e-06, "loss": 0.5698, "step": 727 }, { "epoch": 0.18550133775003186, "grad_norm": 344.9529724121094, "learning_rate": 8.144750254841999e-06, "loss": 3.5372, "step": 728 }, { "epoch": 0.18575614727990827, "grad_norm": 95.5208969116211, "learning_rate": 8.142201834862386e-06, "loss": 0.1949, "step": 729 }, { "epoch": 0.1860109568097847, "grad_norm": 13.727733612060547, "learning_rate": 8.139653414882773e-06, "loss": 0.012, "step": 730 }, { "epoch": 0.1862657663396611, "grad_norm": 24.458459854125977, "learning_rate": 8.137104994903161e-06, "loss": 0.0184, "step": 731 }, { "epoch": 0.18652057586953752, "grad_norm": 16.154903411865234, "learning_rate": 8.134556574923548e-06, "loss": 0.0101, "step": 732 }, { "epoch": 0.18677538539941393, "grad_norm": 87.26277923583984, "learning_rate": 8.132008154943935e-06, "loss": 0.7774, "step": 733 }, { "epoch": 0.18703019492929035, "grad_norm": 123.02075958251953, "learning_rate": 8.129459734964322e-06, "loss": 0.1137, "step": 734 }, { "epoch": 0.18728500445916677, "grad_norm": 15.207338333129883, "learning_rate": 8.12691131498471e-06, "loss": 0.015, "step": 735 }, { "epoch": 0.18753981398904318, "grad_norm": 12.185080528259277, "learning_rate": 8.124362895005098e-06, "loss": 0.011, "step": 736 }, { "epoch": 0.1877946235189196, "grad_norm": 149.4249725341797, "learning_rate": 8.121814475025485e-06, "loss": 0.4315, "step": 737 }, { "epoch": 0.18804943304879604, "grad_norm": 6.199905872344971, "learning_rate": 8.119266055045872e-06, "loss": 0.0116, "step": 738 }, { "epoch": 0.18830424257867245, "grad_norm": 6.252039432525635, "learning_rate": 8.116717635066259e-06, "loss": 0.0039, "step": 739 }, { "epoch": 0.18855905210854887, "grad_norm": 3.601240634918213, "learning_rate": 8.114169215086647e-06, "loss": 0.002, "step": 740 }, { "epoch": 0.18881386163842528, "grad_norm": 3.0869665145874023, "learning_rate": 8.111620795107034e-06, "loss": 0.0013, "step": 741 }, { "epoch": 0.1890686711683017, "grad_norm": 3.771259307861328, "learning_rate": 8.109072375127421e-06, "loss": 0.0021, "step": 742 }, { "epoch": 0.18932348069817811, "grad_norm": 207.44976806640625, "learning_rate": 8.10652395514781e-06, "loss": 0.9807, "step": 743 }, { "epoch": 0.18957829022805453, "grad_norm": 7.4394989013671875, "learning_rate": 8.103975535168197e-06, "loss": 0.0041, "step": 744 }, { "epoch": 0.18983309975793095, "grad_norm": 3.522735357284546, "learning_rate": 8.101427115188584e-06, "loss": 0.0031, "step": 745 }, { "epoch": 0.19008790928780736, "grad_norm": 240.0874786376953, "learning_rate": 8.098878695208972e-06, "loss": 0.2152, "step": 746 }, { "epoch": 0.19034271881768378, "grad_norm": 258.3412780761719, "learning_rate": 8.096330275229358e-06, "loss": 2.5063, "step": 747 }, { "epoch": 0.1905975283475602, "grad_norm": 27.07572364807129, "learning_rate": 8.093781855249746e-06, "loss": 0.046, "step": 748 }, { "epoch": 0.1908523378774366, "grad_norm": 123.6343994140625, "learning_rate": 8.091233435270133e-06, "loss": 3.8954, "step": 749 }, { "epoch": 0.19110714740731302, "grad_norm": 26.699323654174805, "learning_rate": 8.08868501529052e-06, "loss": 0.0419, "step": 750 }, { "epoch": 0.19136195693718946, "grad_norm": 113.14676666259766, "learning_rate": 8.086136595310909e-06, "loss": 0.5013, "step": 751 }, { "epoch": 0.19161676646706588, "grad_norm": 12.461063385009766, "learning_rate": 8.083588175331296e-06, "loss": 0.0142, "step": 752 }, { "epoch": 0.1918715759969423, "grad_norm": 8.081064224243164, "learning_rate": 8.081039755351683e-06, "loss": 0.0123, "step": 753 }, { "epoch": 0.1921263855268187, "grad_norm": 691.4606323242188, "learning_rate": 8.07849133537207e-06, "loss": 1.3288, "step": 754 }, { "epoch": 0.19238119505669513, "grad_norm": 8.771824836730957, "learning_rate": 8.075942915392457e-06, "loss": 0.013, "step": 755 }, { "epoch": 0.19263600458657154, "grad_norm": 14.059209823608398, "learning_rate": 8.073394495412845e-06, "loss": 0.0172, "step": 756 }, { "epoch": 0.19289081411644796, "grad_norm": 6.797913074493408, "learning_rate": 8.070846075433232e-06, "loss": 0.0118, "step": 757 }, { "epoch": 0.19314562364632437, "grad_norm": 4.570469856262207, "learning_rate": 8.068297655453619e-06, "loss": 0.0045, "step": 758 }, { "epoch": 0.1934004331762008, "grad_norm": 1.1206953525543213, "learning_rate": 8.065749235474006e-06, "loss": 0.0053, "step": 759 }, { "epoch": 0.1936552427060772, "grad_norm": 2.416242837905884, "learning_rate": 8.063200815494395e-06, "loss": 0.006, "step": 760 }, { "epoch": 0.19391005223595362, "grad_norm": 10.277372360229492, "learning_rate": 8.060652395514782e-06, "loss": 0.0102, "step": 761 }, { "epoch": 0.19416486176583003, "grad_norm": 6.629180431365967, "learning_rate": 8.058103975535169e-06, "loss": 0.007, "step": 762 }, { "epoch": 0.19441967129570645, "grad_norm": 11.52656364440918, "learning_rate": 8.055555555555557e-06, "loss": 0.0104, "step": 763 }, { "epoch": 0.19467448082558286, "grad_norm": 51.111392974853516, "learning_rate": 8.053007135575943e-06, "loss": 0.4075, "step": 764 }, { "epoch": 0.1949292903554593, "grad_norm": 87.60601806640625, "learning_rate": 8.050458715596331e-06, "loss": 0.4088, "step": 765 }, { "epoch": 0.19518409988533572, "grad_norm": 74.85270690917969, "learning_rate": 8.047910295616718e-06, "loss": 0.057, "step": 766 }, { "epoch": 0.19543890941521214, "grad_norm": 4.844683647155762, "learning_rate": 8.045361875637105e-06, "loss": 0.0049, "step": 767 }, { "epoch": 0.19569371894508855, "grad_norm": 6.800317764282227, "learning_rate": 8.042813455657494e-06, "loss": 0.0045, "step": 768 }, { "epoch": 0.19594852847496497, "grad_norm": 6.884078025817871, "learning_rate": 8.04026503567788e-06, "loss": 0.0106, "step": 769 }, { "epoch": 0.19620333800484138, "grad_norm": 1.1684939861297607, "learning_rate": 8.037716615698268e-06, "loss": 0.002, "step": 770 }, { "epoch": 0.1964581475347178, "grad_norm": 2.382887840270996, "learning_rate": 8.035168195718656e-06, "loss": 0.0026, "step": 771 }, { "epoch": 0.1967129570645942, "grad_norm": 5.3212175369262695, "learning_rate": 8.032619775739042e-06, "loss": 0.0082, "step": 772 }, { "epoch": 0.19696776659447063, "grad_norm": 736.4041137695312, "learning_rate": 8.03007135575943e-06, "loss": 0.5723, "step": 773 }, { "epoch": 0.19722257612434704, "grad_norm": 71.96548461914062, "learning_rate": 8.027522935779817e-06, "loss": 0.6891, "step": 774 }, { "epoch": 0.19747738565422346, "grad_norm": 21.422536849975586, "learning_rate": 8.024974515800204e-06, "loss": 0.0306, "step": 775 }, { "epoch": 0.19773219518409987, "grad_norm": 24.483240127563477, "learning_rate": 8.022426095820593e-06, "loss": 0.0361, "step": 776 }, { "epoch": 0.1979870047139763, "grad_norm": 10.323836326599121, "learning_rate": 8.01987767584098e-06, "loss": 0.008, "step": 777 }, { "epoch": 0.19824181424385273, "grad_norm": 11.626747131347656, "learning_rate": 8.017329255861367e-06, "loss": 0.014, "step": 778 }, { "epoch": 0.19849662377372915, "grad_norm": 1910.42529296875, "learning_rate": 8.014780835881754e-06, "loss": 2.4078, "step": 779 }, { "epoch": 0.19875143330360556, "grad_norm": 12.097496032714844, "learning_rate": 8.01223241590214e-06, "loss": 0.0097, "step": 780 }, { "epoch": 0.19900624283348198, "grad_norm": 1.4545053243637085, "learning_rate": 8.00968399592253e-06, "loss": 0.002, "step": 781 }, { "epoch": 0.1992610523633584, "grad_norm": 9.746658325195312, "learning_rate": 8.007135575942916e-06, "loss": 0.0067, "step": 782 }, { "epoch": 0.1995158618932348, "grad_norm": 5.473706245422363, "learning_rate": 8.004587155963303e-06, "loss": 0.0032, "step": 783 }, { "epoch": 0.19977067142311122, "grad_norm": 41.58401870727539, "learning_rate": 8.00203873598369e-06, "loss": 0.2602, "step": 784 }, { "epoch": 0.20002548095298764, "grad_norm": 67.98558807373047, "learning_rate": 7.999490316004079e-06, "loss": 0.3803, "step": 785 }, { "epoch": 0.20028029048286405, "grad_norm": 79.42501831054688, "learning_rate": 7.996941896024466e-06, "loss": 0.0814, "step": 786 }, { "epoch": 0.20053510001274047, "grad_norm": 21.074947357177734, "learning_rate": 7.994393476044853e-06, "loss": 0.0319, "step": 787 }, { "epoch": 0.20078990954261688, "grad_norm": 16.414112091064453, "learning_rate": 7.991845056065241e-06, "loss": 0.0182, "step": 788 }, { "epoch": 0.2010447190724933, "grad_norm": 43.5382194519043, "learning_rate": 7.989296636085627e-06, "loss": 0.0399, "step": 789 }, { "epoch": 0.20129952860236972, "grad_norm": 13.620686531066895, "learning_rate": 7.986748216106015e-06, "loss": 0.0123, "step": 790 }, { "epoch": 0.20155433813224616, "grad_norm": 0.907649576663971, "learning_rate": 7.984199796126402e-06, "loss": 0.0019, "step": 791 }, { "epoch": 0.20180914766212257, "grad_norm": 2.888389825820923, "learning_rate": 7.981651376146789e-06, "loss": 0.0023, "step": 792 }, { "epoch": 0.202063957191999, "grad_norm": 1.738276481628418, "learning_rate": 7.979102956167178e-06, "loss": 0.0035, "step": 793 }, { "epoch": 0.2023187667218754, "grad_norm": 200.20700073242188, "learning_rate": 7.976554536187565e-06, "loss": 0.2802, "step": 794 }, { "epoch": 0.20257357625175182, "grad_norm": 2.550565719604492, "learning_rate": 7.974006116207952e-06, "loss": 0.0016, "step": 795 }, { "epoch": 0.20282838578162823, "grad_norm": 138.15821838378906, "learning_rate": 7.971457696228339e-06, "loss": 0.234, "step": 796 }, { "epoch": 0.20308319531150465, "grad_norm": 8.389463424682617, "learning_rate": 7.968909276248726e-06, "loss": 0.0092, "step": 797 }, { "epoch": 0.20333800484138106, "grad_norm": 2.280897617340088, "learning_rate": 7.966360856269114e-06, "loss": 0.0031, "step": 798 }, { "epoch": 0.20359281437125748, "grad_norm": 3.4397807121276855, "learning_rate": 7.963812436289501e-06, "loss": 0.0046, "step": 799 }, { "epoch": 0.2038476239011339, "grad_norm": 7.0501580238342285, "learning_rate": 7.961264016309888e-06, "loss": 0.0056, "step": 800 }, { "epoch": 0.2041024334310103, "grad_norm": 14.371070861816406, "learning_rate": 7.958715596330275e-06, "loss": 0.0143, "step": 801 }, { "epoch": 0.20435724296088673, "grad_norm": 5.279518127441406, "learning_rate": 7.956167176350664e-06, "loss": 0.0061, "step": 802 }, { "epoch": 0.20461205249076314, "grad_norm": 17.255367279052734, "learning_rate": 7.95361875637105e-06, "loss": 0.0146, "step": 803 }, { "epoch": 0.20486686202063958, "grad_norm": 8.05230712890625, "learning_rate": 7.951070336391438e-06, "loss": 0.0085, "step": 804 }, { "epoch": 0.205121671550516, "grad_norm": 87.41287994384766, "learning_rate": 7.948521916411826e-06, "loss": 0.7463, "step": 805 }, { "epoch": 0.20537648108039241, "grad_norm": 12.32080078125, "learning_rate": 7.945973496432212e-06, "loss": 0.0151, "step": 806 }, { "epoch": 0.20563129061026883, "grad_norm": 11.955742835998535, "learning_rate": 7.9434250764526e-06, "loss": 0.0117, "step": 807 }, { "epoch": 0.20588610014014525, "grad_norm": 7.503651142120361, "learning_rate": 7.940876656472987e-06, "loss": 0.0051, "step": 808 }, { "epoch": 0.20614090967002166, "grad_norm": 5.031035423278809, "learning_rate": 7.938328236493374e-06, "loss": 0.0028, "step": 809 }, { "epoch": 0.20639571919989808, "grad_norm": 8.738835334777832, "learning_rate": 7.935779816513763e-06, "loss": 0.0067, "step": 810 }, { "epoch": 0.2066505287297745, "grad_norm": 7.2342047691345215, "learning_rate": 7.93323139653415e-06, "loss": 0.004, "step": 811 }, { "epoch": 0.2069053382596509, "grad_norm": 17.276758193969727, "learning_rate": 7.930682976554537e-06, "loss": 0.0223, "step": 812 }, { "epoch": 0.20716014778952732, "grad_norm": 6.109868049621582, "learning_rate": 7.928134556574925e-06, "loss": 0.0028, "step": 813 }, { "epoch": 0.20741495731940374, "grad_norm": 1024.965576171875, "learning_rate": 7.92558613659531e-06, "loss": 2.3133, "step": 814 }, { "epoch": 0.20766976684928015, "grad_norm": 1.4937922954559326, "learning_rate": 7.9230377166157e-06, "loss": 0.001, "step": 815 }, { "epoch": 0.20792457637915657, "grad_norm": 4.803149223327637, "learning_rate": 7.920489296636086e-06, "loss": 0.0054, "step": 816 }, { "epoch": 0.208179385909033, "grad_norm": 352.9211120605469, "learning_rate": 7.917940876656473e-06, "loss": 0.341, "step": 817 }, { "epoch": 0.20843419543890943, "grad_norm": 6.533541679382324, "learning_rate": 7.915392456676862e-06, "loss": 0.0049, "step": 818 }, { "epoch": 0.20868900496878584, "grad_norm": 13.473882675170898, "learning_rate": 7.912844036697249e-06, "loss": 0.0162, "step": 819 }, { "epoch": 0.20894381449866226, "grad_norm": 4.410040378570557, "learning_rate": 7.910295616717636e-06, "loss": 0.0078, "step": 820 }, { "epoch": 0.20919862402853867, "grad_norm": 7.174475193023682, "learning_rate": 7.907747196738023e-06, "loss": 0.0036, "step": 821 }, { "epoch": 0.2094534335584151, "grad_norm": 7.0319132804870605, "learning_rate": 7.905198776758411e-06, "loss": 0.0032, "step": 822 }, { "epoch": 0.2097082430882915, "grad_norm": 4.358388900756836, "learning_rate": 7.902650356778798e-06, "loss": 0.0043, "step": 823 }, { "epoch": 0.20996305261816792, "grad_norm": 156.07241821289062, "learning_rate": 7.900101936799185e-06, "loss": 0.3997, "step": 824 }, { "epoch": 0.21021786214804433, "grad_norm": 24.62611198425293, "learning_rate": 7.897553516819572e-06, "loss": 0.0419, "step": 825 }, { "epoch": 0.21047267167792075, "grad_norm": 55.484161376953125, "learning_rate": 7.895005096839959e-06, "loss": 0.3351, "step": 826 }, { "epoch": 0.21072748120779716, "grad_norm": 43.82518005371094, "learning_rate": 7.892456676860348e-06, "loss": 0.0441, "step": 827 }, { "epoch": 0.21098229073767358, "grad_norm": 28.1529598236084, "learning_rate": 7.889908256880735e-06, "loss": 0.0499, "step": 828 }, { "epoch": 0.21123710026755, "grad_norm": 32.94591522216797, "learning_rate": 7.887359836901122e-06, "loss": 0.0688, "step": 829 }, { "epoch": 0.21149190979742644, "grad_norm": 21.583419799804688, "learning_rate": 7.88481141692151e-06, "loss": 0.0308, "step": 830 }, { "epoch": 0.21174671932730285, "grad_norm": 12.505573272705078, "learning_rate": 7.882262996941895e-06, "loss": 0.0193, "step": 831 }, { "epoch": 0.21200152885717927, "grad_norm": 167.01406860351562, "learning_rate": 7.879714576962284e-06, "loss": 0.2289, "step": 832 }, { "epoch": 0.21225633838705568, "grad_norm": 10.525616645812988, "learning_rate": 7.877166156982671e-06, "loss": 0.0093, "step": 833 }, { "epoch": 0.2125111479169321, "grad_norm": 5.61439323425293, "learning_rate": 7.874617737003058e-06, "loss": 0.004, "step": 834 }, { "epoch": 0.2127659574468085, "grad_norm": 6.062806129455566, "learning_rate": 7.872069317023447e-06, "loss": 0.003, "step": 835 }, { "epoch": 0.21302076697668493, "grad_norm": 9.927224159240723, "learning_rate": 7.869520897043834e-06, "loss": 0.0074, "step": 836 }, { "epoch": 0.21327557650656134, "grad_norm": 10.591646194458008, "learning_rate": 7.86697247706422e-06, "loss": 0.0079, "step": 837 }, { "epoch": 0.21353038603643776, "grad_norm": 10.34538459777832, "learning_rate": 7.86442405708461e-06, "loss": 0.0082, "step": 838 }, { "epoch": 0.21378519556631417, "grad_norm": 466.6644287109375, "learning_rate": 7.861875637104996e-06, "loss": 0.6139, "step": 839 }, { "epoch": 0.2140400050961906, "grad_norm": 0.927978515625, "learning_rate": 7.859327217125383e-06, "loss": 0.0026, "step": 840 }, { "epoch": 0.214294814626067, "grad_norm": 6.513106822967529, "learning_rate": 7.85677879714577e-06, "loss": 0.0069, "step": 841 }, { "epoch": 0.21454962415594342, "grad_norm": 67.68296813964844, "learning_rate": 7.854230377166157e-06, "loss": 0.31, "step": 842 }, { "epoch": 0.21480443368581986, "grad_norm": 61.1998291015625, "learning_rate": 7.851681957186546e-06, "loss": 0.0563, "step": 843 }, { "epoch": 0.21505924321569628, "grad_norm": 70.8736572265625, "learning_rate": 7.849133537206933e-06, "loss": 0.5183, "step": 844 }, { "epoch": 0.2153140527455727, "grad_norm": 64.7689437866211, "learning_rate": 7.84658511722732e-06, "loss": 0.029, "step": 845 }, { "epoch": 0.2155688622754491, "grad_norm": 5.319849491119385, "learning_rate": 7.844036697247707e-06, "loss": 0.0028, "step": 846 }, { "epoch": 0.21582367180532552, "grad_norm": 3.066965103149414, "learning_rate": 7.841488277268095e-06, "loss": 0.0025, "step": 847 }, { "epoch": 0.21607848133520194, "grad_norm": 12.662148475646973, "learning_rate": 7.838939857288482e-06, "loss": 0.0081, "step": 848 }, { "epoch": 0.21633329086507835, "grad_norm": 212.3879852294922, "learning_rate": 7.836391437308869e-06, "loss": 0.598, "step": 849 }, { "epoch": 0.21658810039495477, "grad_norm": 4.502976894378662, "learning_rate": 7.833843017329256e-06, "loss": 0.002, "step": 850 }, { "epoch": 0.21684290992483118, "grad_norm": 1.274155855178833, "learning_rate": 7.831294597349643e-06, "loss": 0.0022, "step": 851 }, { "epoch": 0.2170977194547076, "grad_norm": 0.6867152452468872, "learning_rate": 7.828746177370032e-06, "loss": 0.001, "step": 852 }, { "epoch": 0.21735252898458401, "grad_norm": 4.200469970703125, "learning_rate": 7.826197757390419e-06, "loss": 0.0028, "step": 853 }, { "epoch": 0.21760733851446043, "grad_norm": 1.7612526416778564, "learning_rate": 7.823649337410806e-06, "loss": 0.0027, "step": 854 }, { "epoch": 0.21786214804433685, "grad_norm": 8.225656509399414, "learning_rate": 7.821100917431194e-06, "loss": 0.0055, "step": 855 }, { "epoch": 0.2181169575742133, "grad_norm": 7.8409647941589355, "learning_rate": 7.81855249745158e-06, "loss": 0.0045, "step": 856 }, { "epoch": 0.2183717671040897, "grad_norm": 140.7010040283203, "learning_rate": 7.816004077471968e-06, "loss": 0.442, "step": 857 }, { "epoch": 0.21862657663396612, "grad_norm": 207.52159118652344, "learning_rate": 7.813455657492357e-06, "loss": 3.2502, "step": 858 }, { "epoch": 0.21888138616384253, "grad_norm": 12.458690643310547, "learning_rate": 7.810907237512742e-06, "loss": 0.0113, "step": 859 }, { "epoch": 0.21913619569371895, "grad_norm": 11.633651733398438, "learning_rate": 7.80835881753313e-06, "loss": 0.0096, "step": 860 }, { "epoch": 0.21939100522359536, "grad_norm": 14.775179862976074, "learning_rate": 7.805810397553518e-06, "loss": 0.015, "step": 861 }, { "epoch": 0.21964581475347178, "grad_norm": 6.201473236083984, "learning_rate": 7.803261977573905e-06, "loss": 0.0043, "step": 862 }, { "epoch": 0.2199006242833482, "grad_norm": 6.513566970825195, "learning_rate": 7.800713557594293e-06, "loss": 0.0039, "step": 863 }, { "epoch": 0.2201554338132246, "grad_norm": 83.46454620361328, "learning_rate": 7.79816513761468e-06, "loss": 0.563, "step": 864 }, { "epoch": 0.22041024334310103, "grad_norm": 3.3830487728118896, "learning_rate": 7.795616717635067e-06, "loss": 0.0079, "step": 865 }, { "epoch": 0.22066505287297744, "grad_norm": 0.7122774124145508, "learning_rate": 7.793068297655454e-06, "loss": 0.0017, "step": 866 }, { "epoch": 0.22091986240285386, "grad_norm": 50.44882583618164, "learning_rate": 7.790519877675841e-06, "loss": 0.0417, "step": 867 }, { "epoch": 0.22117467193273027, "grad_norm": 7.045048713684082, "learning_rate": 7.787971457696228e-06, "loss": 0.0045, "step": 868 }, { "epoch": 0.22142948146260671, "grad_norm": 8.233758926391602, "learning_rate": 7.785423037716617e-06, "loss": 0.0087, "step": 869 }, { "epoch": 0.22168429099248313, "grad_norm": 8.370360374450684, "learning_rate": 7.782874617737004e-06, "loss": 0.0063, "step": 870 }, { "epoch": 0.22193910052235954, "grad_norm": 1.875454306602478, "learning_rate": 7.78032619775739e-06, "loss": 0.0015, "step": 871 }, { "epoch": 0.22219391005223596, "grad_norm": 5.443411350250244, "learning_rate": 7.77777777777778e-06, "loss": 0.0058, "step": 872 }, { "epoch": 0.22244871958211238, "grad_norm": 1.528052806854248, "learning_rate": 7.775229357798164e-06, "loss": 0.0023, "step": 873 }, { "epoch": 0.2227035291119888, "grad_norm": 336.9598388671875, "learning_rate": 7.772680937818553e-06, "loss": 0.3628, "step": 874 }, { "epoch": 0.2229583386418652, "grad_norm": 11.13251781463623, "learning_rate": 7.77013251783894e-06, "loss": 0.01, "step": 875 }, { "epoch": 0.22321314817174162, "grad_norm": 18.117284774780273, "learning_rate": 7.767584097859327e-06, "loss": 0.0223, "step": 876 }, { "epoch": 0.22346795770161804, "grad_norm": 16.41920280456543, "learning_rate": 7.765035677879716e-06, "loss": 0.0204, "step": 877 }, { "epoch": 0.22372276723149445, "grad_norm": 14.44711685180664, "learning_rate": 7.762487257900103e-06, "loss": 0.0144, "step": 878 }, { "epoch": 0.22397757676137087, "grad_norm": 1.2297385931015015, "learning_rate": 7.75993883792049e-06, "loss": 0.001, "step": 879 }, { "epoch": 0.22423238629124728, "grad_norm": 2.525031089782715, "learning_rate": 7.757390417940878e-06, "loss": 0.0045, "step": 880 }, { "epoch": 0.2244871958211237, "grad_norm": 0.7896365523338318, "learning_rate": 7.754841997961265e-06, "loss": 0.0014, "step": 881 }, { "epoch": 0.22474200535100014, "grad_norm": 5.889585018157959, "learning_rate": 7.752293577981652e-06, "loss": 0.0041, "step": 882 }, { "epoch": 0.22499681488087656, "grad_norm": 5.4996795654296875, "learning_rate": 7.749745158002039e-06, "loss": 0.0025, "step": 883 }, { "epoch": 0.22525162441075297, "grad_norm": 94.62612915039062, "learning_rate": 7.747196738022426e-06, "loss": 0.0756, "step": 884 }, { "epoch": 0.22550643394062939, "grad_norm": 3.6954033374786377, "learning_rate": 7.744648318042815e-06, "loss": 0.0021, "step": 885 }, { "epoch": 0.2257612434705058, "grad_norm": 386.6903381347656, "learning_rate": 7.742099898063202e-06, "loss": 0.1841, "step": 886 }, { "epoch": 0.22601605300038222, "grad_norm": 97.84149932861328, "learning_rate": 7.739551478083589e-06, "loss": 1.2656, "step": 887 }, { "epoch": 0.22627086253025863, "grad_norm": 10.041919708251953, "learning_rate": 7.737003058103975e-06, "loss": 0.011, "step": 888 }, { "epoch": 0.22652567206013505, "grad_norm": 18.38896369934082, "learning_rate": 7.734454638124364e-06, "loss": 0.0269, "step": 889 }, { "epoch": 0.22678048159001146, "grad_norm": 341.3282470703125, "learning_rate": 7.731906218144751e-06, "loss": 2.7565, "step": 890 }, { "epoch": 0.22703529111988788, "grad_norm": 387.1972351074219, "learning_rate": 7.729357798165138e-06, "loss": 0.2863, "step": 891 }, { "epoch": 0.2272901006497643, "grad_norm": 44.79634094238281, "learning_rate": 7.726809378185525e-06, "loss": 0.3307, "step": 892 }, { "epoch": 0.2275449101796407, "grad_norm": 3.3438644409179688, "learning_rate": 7.724260958205912e-06, "loss": 0.0037, "step": 893 }, { "epoch": 0.22779971970951712, "grad_norm": 12.60893440246582, "learning_rate": 7.7217125382263e-06, "loss": 0.0116, "step": 894 }, { "epoch": 0.22805452923939357, "grad_norm": 13.756491661071777, "learning_rate": 7.719164118246688e-06, "loss": 0.0137, "step": 895 }, { "epoch": 0.22830933876926998, "grad_norm": 83.3353042602539, "learning_rate": 7.716615698267074e-06, "loss": 0.1151, "step": 896 }, { "epoch": 0.2285641482991464, "grad_norm": 2.7178268432617188, "learning_rate": 7.714067278287463e-06, "loss": 0.0027, "step": 897 }, { "epoch": 0.2288189578290228, "grad_norm": 5.558854579925537, "learning_rate": 7.71151885830785e-06, "loss": 0.005, "step": 898 }, { "epoch": 0.22907376735889923, "grad_norm": 1.2551816701889038, "learning_rate": 7.708970438328237e-06, "loss": 0.0028, "step": 899 }, { "epoch": 0.22932857688877564, "grad_norm": 2.5748391151428223, "learning_rate": 7.706422018348626e-06, "loss": 0.0042, "step": 900 }, { "epoch": 0.22958338641865206, "grad_norm": 10.017240524291992, "learning_rate": 7.703873598369011e-06, "loss": 0.008, "step": 901 }, { "epoch": 0.22983819594852847, "grad_norm": 37.88123321533203, "learning_rate": 7.7013251783894e-06, "loss": 0.0757, "step": 902 }, { "epoch": 0.2300930054784049, "grad_norm": 3.317456007003784, "learning_rate": 7.698776758409787e-06, "loss": 0.0025, "step": 903 }, { "epoch": 0.2303478150082813, "grad_norm": 25.264501571655273, "learning_rate": 7.696228338430174e-06, "loss": 0.0403, "step": 904 }, { "epoch": 0.23060262453815772, "grad_norm": 6.7867045402526855, "learning_rate": 7.693679918450562e-06, "loss": 0.0039, "step": 905 }, { "epoch": 0.23085743406803413, "grad_norm": 13.396056175231934, "learning_rate": 7.691131498470949e-06, "loss": 0.0127, "step": 906 }, { "epoch": 0.23111224359791055, "grad_norm": 6.317763328552246, "learning_rate": 7.688583078491336e-06, "loss": 0.0083, "step": 907 }, { "epoch": 0.231367053127787, "grad_norm": 12.72378921508789, "learning_rate": 7.686034658511723e-06, "loss": 0.0111, "step": 908 }, { "epoch": 0.2316218626576634, "grad_norm": 256.6712951660156, "learning_rate": 7.68348623853211e-06, "loss": 0.3659, "step": 909 }, { "epoch": 0.23187667218753982, "grad_norm": 4.448498249053955, "learning_rate": 7.680937818552499e-06, "loss": 0.0035, "step": 910 }, { "epoch": 0.23213148171741624, "grad_norm": 1.3387995958328247, "learning_rate": 7.678389398572886e-06, "loss": 0.0029, "step": 911 }, { "epoch": 0.23238629124729265, "grad_norm": 1.6688767671585083, "learning_rate": 7.675840978593273e-06, "loss": 0.0011, "step": 912 }, { "epoch": 0.23264110077716907, "grad_norm": 1.3107398748397827, "learning_rate": 7.67329255861366e-06, "loss": 0.0019, "step": 913 }, { "epoch": 0.23289591030704548, "grad_norm": 2.031846523284912, "learning_rate": 7.670744138634048e-06, "loss": 0.0009, "step": 914 }, { "epoch": 0.2331507198369219, "grad_norm": 2.573822259902954, "learning_rate": 7.668195718654435e-06, "loss": 0.0013, "step": 915 }, { "epoch": 0.23340552936679831, "grad_norm": 8.096732139587402, "learning_rate": 7.665647298674822e-06, "loss": 0.0048, "step": 916 }, { "epoch": 0.23366033889667473, "grad_norm": 137.30413818359375, "learning_rate": 7.66309887869521e-06, "loss": 0.775, "step": 917 }, { "epoch": 0.23391514842655114, "grad_norm": 8.532777786254883, "learning_rate": 7.660550458715596e-06, "loss": 0.0055, "step": 918 }, { "epoch": 0.23416995795642756, "grad_norm": 19.57465171813965, "learning_rate": 7.658002038735985e-06, "loss": 0.0291, "step": 919 }, { "epoch": 0.23442476748630398, "grad_norm": 14.862135887145996, "learning_rate": 7.655453618756372e-06, "loss": 0.0153, "step": 920 }, { "epoch": 0.23467957701618042, "grad_norm": 8.469969749450684, "learning_rate": 7.652905198776758e-06, "loss": 0.0203, "step": 921 }, { "epoch": 0.23493438654605683, "grad_norm": 13.249506950378418, "learning_rate": 7.650356778797147e-06, "loss": 0.0127, "step": 922 }, { "epoch": 0.23518919607593325, "grad_norm": 6.139504909515381, "learning_rate": 7.647808358817534e-06, "loss": 0.003, "step": 923 }, { "epoch": 0.23544400560580966, "grad_norm": 1.8604609966278076, "learning_rate": 7.645259938837921e-06, "loss": 0.0016, "step": 924 }, { "epoch": 0.23569881513568608, "grad_norm": 1.7501715421676636, "learning_rate": 7.64271151885831e-06, "loss": 0.0025, "step": 925 }, { "epoch": 0.2359536246655625, "grad_norm": 3.0566492080688477, "learning_rate": 7.640163098878695e-06, "loss": 0.0012, "step": 926 }, { "epoch": 0.2362084341954389, "grad_norm": 12.355313301086426, "learning_rate": 7.637614678899084e-06, "loss": 0.0117, "step": 927 }, { "epoch": 0.23646324372531533, "grad_norm": 8.740922927856445, "learning_rate": 7.63506625891947e-06, "loss": 0.0053, "step": 928 }, { "epoch": 0.23671805325519174, "grad_norm": 55.2923698425293, "learning_rate": 7.632517838939857e-06, "loss": 0.563, "step": 929 }, { "epoch": 0.23697286278506816, "grad_norm": 11.18720817565918, "learning_rate": 7.629969418960246e-06, "loss": 0.0117, "step": 930 }, { "epoch": 0.23722767231494457, "grad_norm": 142.33424377441406, "learning_rate": 7.627420998980632e-06, "loss": 0.3417, "step": 931 }, { "epoch": 0.237482481844821, "grad_norm": 35.336326599121094, "learning_rate": 7.62487257900102e-06, "loss": 0.0147, "step": 932 }, { "epoch": 0.2377372913746974, "grad_norm": 10.15295124053955, "learning_rate": 7.622324159021407e-06, "loss": 0.0047, "step": 933 }, { "epoch": 0.23799210090457384, "grad_norm": 11.798460960388184, "learning_rate": 7.619775739041795e-06, "loss": 0.0114, "step": 934 }, { "epoch": 0.23824691043445026, "grad_norm": 14.82275676727295, "learning_rate": 7.617227319062183e-06, "loss": 0.0157, "step": 935 }, { "epoch": 0.23850171996432667, "grad_norm": 13.297581672668457, "learning_rate": 7.6146788990825695e-06, "loss": 0.0082, "step": 936 }, { "epoch": 0.2387565294942031, "grad_norm": 2.5305354595184326, "learning_rate": 7.612130479102957e-06, "loss": 0.0014, "step": 937 }, { "epoch": 0.2390113390240795, "grad_norm": 5.2593092918396, "learning_rate": 7.6095820591233434e-06, "loss": 0.0085, "step": 938 }, { "epoch": 0.23926614855395592, "grad_norm": 0.26241737604141235, "learning_rate": 7.607033639143731e-06, "loss": 0.0008, "step": 939 }, { "epoch": 0.23952095808383234, "grad_norm": 4.504225730895996, "learning_rate": 7.604485219164119e-06, "loss": 0.0023, "step": 940 }, { "epoch": 0.23977576761370875, "grad_norm": 1.2650299072265625, "learning_rate": 7.601936799184506e-06, "loss": 0.0033, "step": 941 }, { "epoch": 0.24003057714358517, "grad_norm": 2.0197315216064453, "learning_rate": 7.599388379204894e-06, "loss": 0.0007, "step": 942 }, { "epoch": 0.24028538667346158, "grad_norm": 4.709583759307861, "learning_rate": 7.596839959225281e-06, "loss": 0.0046, "step": 943 }, { "epoch": 0.240540196203338, "grad_norm": 7.379934310913086, "learning_rate": 7.5942915392456685e-06, "loss": 0.0033, "step": 944 }, { "epoch": 0.2407950057332144, "grad_norm": 77.70902252197266, "learning_rate": 7.5917431192660555e-06, "loss": 0.0334, "step": 945 }, { "epoch": 0.24104981526309083, "grad_norm": 12.164741516113281, "learning_rate": 7.589194699286443e-06, "loss": 0.0122, "step": 946 }, { "epoch": 0.24130462479296727, "grad_norm": 147.49557495117188, "learning_rate": 7.586646279306831e-06, "loss": 3.2503, "step": 947 }, { "epoch": 0.24155943432284369, "grad_norm": 7.852569103240967, "learning_rate": 7.584097859327217e-06, "loss": 0.0049, "step": 948 }, { "epoch": 0.2418142438527201, "grad_norm": 2.2729146480560303, "learning_rate": 7.581549439347605e-06, "loss": 0.0032, "step": 949 }, { "epoch": 0.24206905338259652, "grad_norm": 27.67197608947754, "learning_rate": 7.579001019367992e-06, "loss": 0.0287, "step": 950 }, { "epoch": 0.24232386291247293, "grad_norm": 2.4236226081848145, "learning_rate": 7.57645259938838e-06, "loss": 0.0032, "step": 951 }, { "epoch": 0.24257867244234935, "grad_norm": 3.729179859161377, "learning_rate": 7.5739041794087676e-06, "loss": 0.0017, "step": 952 }, { "epoch": 0.24283348197222576, "grad_norm": 3.8558833599090576, "learning_rate": 7.5713557594291545e-06, "loss": 0.0049, "step": 953 }, { "epoch": 0.24308829150210218, "grad_norm": 1.8311649560928345, "learning_rate": 7.568807339449542e-06, "loss": 0.0012, "step": 954 }, { "epoch": 0.2433431010319786, "grad_norm": 9.815407752990723, "learning_rate": 7.566258919469928e-06, "loss": 0.0074, "step": 955 }, { "epoch": 0.243597910561855, "grad_norm": 5.448746681213379, "learning_rate": 7.563710499490316e-06, "loss": 0.0065, "step": 956 }, { "epoch": 0.24385272009173142, "grad_norm": 10.364989280700684, "learning_rate": 7.561162079510704e-06, "loss": 0.0085, "step": 957 }, { "epoch": 0.24410752962160784, "grad_norm": 3.8324239253997803, "learning_rate": 7.558613659531091e-06, "loss": 0.0022, "step": 958 }, { "epoch": 0.24436233915148425, "grad_norm": 1.637624740600586, "learning_rate": 7.556065239551479e-06, "loss": 0.0045, "step": 959 }, { "epoch": 0.2446171486813607, "grad_norm": 2.528881072998047, "learning_rate": 7.553516819571866e-06, "loss": 0.0031, "step": 960 }, { "epoch": 0.2448719582112371, "grad_norm": 8.625188827514648, "learning_rate": 7.5509683995922535e-06, "loss": 0.0059, "step": 961 }, { "epoch": 0.24512676774111353, "grad_norm": 18.53608512878418, "learning_rate": 7.548419979612641e-06, "loss": 0.0108, "step": 962 }, { "epoch": 0.24538157727098994, "grad_norm": 5.777303218841553, "learning_rate": 7.545871559633028e-06, "loss": 0.0028, "step": 963 }, { "epoch": 0.24563638680086636, "grad_norm": 12.737194061279297, "learning_rate": 7.543323139653416e-06, "loss": 0.0115, "step": 964 }, { "epoch": 0.24589119633074277, "grad_norm": 4.058441638946533, "learning_rate": 7.540774719673802e-06, "loss": 0.0017, "step": 965 }, { "epoch": 0.2461460058606192, "grad_norm": 2.1592040061950684, "learning_rate": 7.53822629969419e-06, "loss": 0.004, "step": 966 }, { "epoch": 0.2464008153904956, "grad_norm": 3.8423545360565186, "learning_rate": 7.535677879714578e-06, "loss": 0.0022, "step": 967 }, { "epoch": 0.24665562492037202, "grad_norm": 0.4578978419303894, "learning_rate": 7.533129459734965e-06, "loss": 0.0011, "step": 968 }, { "epoch": 0.24691043445024843, "grad_norm": 0.32270023226737976, "learning_rate": 7.5305810397553525e-06, "loss": 0.0006, "step": 969 }, { "epoch": 0.24716524398012485, "grad_norm": 0.503840446472168, "learning_rate": 7.5280326197757395e-06, "loss": 0.001, "step": 970 }, { "epoch": 0.24742005351000126, "grad_norm": 9.9978666305542, "learning_rate": 7.525484199796127e-06, "loss": 0.01, "step": 971 }, { "epoch": 0.24767486303987768, "grad_norm": 72.37625885009766, "learning_rate": 7.522935779816515e-06, "loss": 0.5629, "step": 972 }, { "epoch": 0.24792967256975412, "grad_norm": 9.441390037536621, "learning_rate": 7.520387359836901e-06, "loss": 0.0064, "step": 973 }, { "epoch": 0.24818448209963054, "grad_norm": 2.4929840564727783, "learning_rate": 7.517838939857289e-06, "loss": 0.0096, "step": 974 }, { "epoch": 0.24843929162950695, "grad_norm": 35.53825378417969, "learning_rate": 7.515290519877676e-06, "loss": 0.0759, "step": 975 }, { "epoch": 0.24869410115938337, "grad_norm": 0.8715357184410095, "learning_rate": 7.512742099898064e-06, "loss": 0.0004, "step": 976 }, { "epoch": 0.24894891068925978, "grad_norm": 0.5543079972267151, "learning_rate": 7.5101936799184515e-06, "loss": 0.0006, "step": 977 }, { "epoch": 0.2492037202191362, "grad_norm": 8.509117126464844, "learning_rate": 7.5076452599388385e-06, "loss": 0.0055, "step": 978 }, { "epoch": 0.24945852974901261, "grad_norm": 1459.9503173828125, "learning_rate": 7.505096839959226e-06, "loss": 1.7412, "step": 979 }, { "epoch": 0.24971333927888903, "grad_norm": 10.803936958312988, "learning_rate": 7.502548419979612e-06, "loss": 0.0089, "step": 980 }, { "epoch": 0.24996814880876544, "grad_norm": 6.786251068115234, "learning_rate": 7.500000000000001e-06, "loss": 0.0039, "step": 981 }, { "epoch": 0.2502229583386419, "grad_norm": 13.539137840270996, "learning_rate": 7.497451580020389e-06, "loss": 0.0073, "step": 982 }, { "epoch": 0.2504777678685183, "grad_norm": 0.3618035912513733, "learning_rate": 7.494903160040775e-06, "loss": 0.0006, "step": 983 }, { "epoch": 0.2507325773983947, "grad_norm": 55.11130142211914, "learning_rate": 7.492354740061163e-06, "loss": 0.6168, "step": 984 }, { "epoch": 0.2509873869282711, "grad_norm": 175.7881317138672, "learning_rate": 7.48980632008155e-06, "loss": 0.1488, "step": 985 }, { "epoch": 0.25124219645814755, "grad_norm": 17.057329177856445, "learning_rate": 7.4872579001019375e-06, "loss": 0.0207, "step": 986 }, { "epoch": 0.25149700598802394, "grad_norm": 17.80198097229004, "learning_rate": 7.484709480122325e-06, "loss": 0.0251, "step": 987 }, { "epoch": 0.2517518155179004, "grad_norm": 20.30576515197754, "learning_rate": 7.482161060142712e-06, "loss": 0.0279, "step": 988 }, { "epoch": 0.25200662504777677, "grad_norm": 16.685121536254883, "learning_rate": 7.4796126401631e-06, "loss": 0.019, "step": 989 }, { "epoch": 0.2522614345776532, "grad_norm": 8.197447776794434, "learning_rate": 7.477064220183486e-06, "loss": 0.005, "step": 990 }, { "epoch": 0.2525162441075296, "grad_norm": 9.934812545776367, "learning_rate": 7.474515800203874e-06, "loss": 0.0058, "step": 991 }, { "epoch": 0.25277105363740604, "grad_norm": 2.0829079151153564, "learning_rate": 7.471967380224262e-06, "loss": 0.0021, "step": 992 }, { "epoch": 0.2530258631672825, "grad_norm": 6.4389872550964355, "learning_rate": 7.469418960244649e-06, "loss": 0.0105, "step": 993 }, { "epoch": 0.25328067269715887, "grad_norm": 2.595555543899536, "learning_rate": 7.4668705402650365e-06, "loss": 0.0016, "step": 994 }, { "epoch": 0.2535354822270353, "grad_norm": 20.645938873291016, "learning_rate": 7.4643221202854235e-06, "loss": 0.0282, "step": 995 }, { "epoch": 0.2537902917569117, "grad_norm": 6.833658218383789, "learning_rate": 7.461773700305811e-06, "loss": 0.007, "step": 996 }, { "epoch": 0.25404510128678814, "grad_norm": 17.642507553100586, "learning_rate": 7.459225280326199e-06, "loss": 0.0224, "step": 997 }, { "epoch": 0.25429991081666453, "grad_norm": 11.203266143798828, "learning_rate": 7.456676860346585e-06, "loss": 0.0121, "step": 998 }, { "epoch": 0.254554720346541, "grad_norm": 8.960795402526855, "learning_rate": 7.454128440366973e-06, "loss": 0.0059, "step": 999 }, { "epoch": 0.25480952987641736, "grad_norm": 15.366107940673828, "learning_rate": 7.45158002038736e-06, "loss": 0.0164, "step": 1000 }, { "epoch": 0.2550643394062938, "grad_norm": 1.4594751596450806, "learning_rate": 7.449031600407748e-06, "loss": 0.003, "step": 1001 }, { "epoch": 0.2553191489361702, "grad_norm": 9.13967227935791, "learning_rate": 7.4464831804281355e-06, "loss": 0.0054, "step": 1002 }, { "epoch": 0.25557395846604664, "grad_norm": 32.55609130859375, "learning_rate": 7.4439347604485225e-06, "loss": 0.0212, "step": 1003 }, { "epoch": 0.255828767995923, "grad_norm": 13.773650169372559, "learning_rate": 7.44138634046891e-06, "loss": 0.0135, "step": 1004 }, { "epoch": 0.25608357752579947, "grad_norm": 54.41773986816406, "learning_rate": 7.438837920489297e-06, "loss": 0.0435, "step": 1005 }, { "epoch": 0.2563383870556759, "grad_norm": 8.054533958435059, "learning_rate": 7.436289500509685e-06, "loss": 0.0049, "step": 1006 }, { "epoch": 0.2565931965855523, "grad_norm": 11.271697998046875, "learning_rate": 7.433741080530073e-06, "loss": 0.009, "step": 1007 }, { "epoch": 0.25684800611542874, "grad_norm": 7.738370895385742, "learning_rate": 7.431192660550459e-06, "loss": 0.0051, "step": 1008 }, { "epoch": 0.2571028156453051, "grad_norm": 0.8792528510093689, "learning_rate": 7.428644240570847e-06, "loss": 0.0018, "step": 1009 }, { "epoch": 0.25735762517518157, "grad_norm": 853.517822265625, "learning_rate": 7.426095820591234e-06, "loss": 0.6732, "step": 1010 }, { "epoch": 0.25761243470505796, "grad_norm": 6.855268955230713, "learning_rate": 7.4235474006116215e-06, "loss": 0.0043, "step": 1011 }, { "epoch": 0.2578672442349344, "grad_norm": 7.690743923187256, "learning_rate": 7.420998980632009e-06, "loss": 0.0049, "step": 1012 }, { "epoch": 0.2581220537648108, "grad_norm": 77.64906311035156, "learning_rate": 7.418450560652396e-06, "loss": 0.8917, "step": 1013 }, { "epoch": 0.25837686329468723, "grad_norm": 9.186442375183105, "learning_rate": 7.415902140672784e-06, "loss": 0.0103, "step": 1014 }, { "epoch": 0.2586316728245636, "grad_norm": 0.9954903721809387, "learning_rate": 7.41335372069317e-06, "loss": 0.0016, "step": 1015 }, { "epoch": 0.25888648235444006, "grad_norm": 3.177839756011963, "learning_rate": 7.410805300713558e-06, "loss": 0.0032, "step": 1016 }, { "epoch": 0.25914129188431645, "grad_norm": 1.3144434690475464, "learning_rate": 7.408256880733946e-06, "loss": 0.0037, "step": 1017 }, { "epoch": 0.2593961014141929, "grad_norm": 7.9781694412231445, "learning_rate": 7.405708460754333e-06, "loss": 0.0101, "step": 1018 }, { "epoch": 0.25965091094406934, "grad_norm": 3.091196298599243, "learning_rate": 7.4031600407747205e-06, "loss": 0.0014, "step": 1019 }, { "epoch": 0.2599057204739457, "grad_norm": 1.0096731185913086, "learning_rate": 7.4006116207951074e-06, "loss": 0.001, "step": 1020 }, { "epoch": 0.26016053000382217, "grad_norm": 1.5030808448791504, "learning_rate": 7.398063200815495e-06, "loss": 0.0011, "step": 1021 }, { "epoch": 0.26041533953369855, "grad_norm": 5.071967601776123, "learning_rate": 7.395514780835882e-06, "loss": 0.0031, "step": 1022 }, { "epoch": 0.260670149063575, "grad_norm": 45.84017562866211, "learning_rate": 7.39296636085627e-06, "loss": 0.4212, "step": 1023 }, { "epoch": 0.2609249585934514, "grad_norm": 2.2220351696014404, "learning_rate": 7.390417940876658e-06, "loss": 0.0014, "step": 1024 }, { "epoch": 0.2611797681233278, "grad_norm": 2.9073612689971924, "learning_rate": 7.387869520897044e-06, "loss": 0.003, "step": 1025 }, { "epoch": 0.2614345776532042, "grad_norm": 5.375236988067627, "learning_rate": 7.385321100917432e-06, "loss": 0.0035, "step": 1026 }, { "epoch": 0.26168938718308066, "grad_norm": 1.6232194900512695, "learning_rate": 7.382772680937819e-06, "loss": 0.002, "step": 1027 }, { "epoch": 0.26194419671295704, "grad_norm": 3.24564266204834, "learning_rate": 7.3802242609582064e-06, "loss": 0.0025, "step": 1028 }, { "epoch": 0.2621990062428335, "grad_norm": 5.290421485900879, "learning_rate": 7.377675840978594e-06, "loss": 0.0029, "step": 1029 }, { "epoch": 0.2624538157727099, "grad_norm": 857.0712280273438, "learning_rate": 7.375127420998981e-06, "loss": 0.5914, "step": 1030 }, { "epoch": 0.2627086253025863, "grad_norm": 1.7532522678375244, "learning_rate": 7.372579001019369e-06, "loss": 0.001, "step": 1031 }, { "epoch": 0.26296343483246276, "grad_norm": 7.173964977264404, "learning_rate": 7.370030581039755e-06, "loss": 0.0078, "step": 1032 }, { "epoch": 0.26321824436233915, "grad_norm": 6.1166181564331055, "learning_rate": 7.367482161060143e-06, "loss": 0.0032, "step": 1033 }, { "epoch": 0.2634730538922156, "grad_norm": 3.38157057762146, "learning_rate": 7.364933741080531e-06, "loss": 0.0015, "step": 1034 }, { "epoch": 0.263727863422092, "grad_norm": 23.82469940185547, "learning_rate": 7.362385321100918e-06, "loss": 0.0177, "step": 1035 }, { "epoch": 0.2639826729519684, "grad_norm": 1.767769455909729, "learning_rate": 7.3598369011213054e-06, "loss": 0.001, "step": 1036 }, { "epoch": 0.2642374824818448, "grad_norm": 4.912933349609375, "learning_rate": 7.357288481141692e-06, "loss": 0.0025, "step": 1037 }, { "epoch": 0.26449229201172125, "grad_norm": 3.7366995811462402, "learning_rate": 7.35474006116208e-06, "loss": 0.002, "step": 1038 }, { "epoch": 0.26474710154159764, "grad_norm": 177.26316833496094, "learning_rate": 7.352191641182468e-06, "loss": 0.2526, "step": 1039 }, { "epoch": 0.2650019110714741, "grad_norm": 131.920166015625, "learning_rate": 7.349643221202855e-06, "loss": 0.5134, "step": 1040 }, { "epoch": 0.26525672060135047, "grad_norm": 1181.184814453125, "learning_rate": 7.347094801223243e-06, "loss": 1.2601, "step": 1041 }, { "epoch": 0.2655115301312269, "grad_norm": 12.167306900024414, "learning_rate": 7.344546381243629e-06, "loss": 0.0106, "step": 1042 }, { "epoch": 0.2657663396611033, "grad_norm": 19.509143829345703, "learning_rate": 7.341997961264017e-06, "loss": 0.0257, "step": 1043 }, { "epoch": 0.26602114919097974, "grad_norm": 18.57435417175293, "learning_rate": 7.3394495412844045e-06, "loss": 0.0237, "step": 1044 }, { "epoch": 0.2662759587208562, "grad_norm": 4.228957653045654, "learning_rate": 7.336901121304791e-06, "loss": 0.0067, "step": 1045 }, { "epoch": 0.2665307682507326, "grad_norm": 3.1919689178466797, "learning_rate": 7.334352701325179e-06, "loss": 0.0031, "step": 1046 }, { "epoch": 0.266785577780609, "grad_norm": 6.276793479919434, "learning_rate": 7.331804281345566e-06, "loss": 0.0081, "step": 1047 }, { "epoch": 0.2670403873104854, "grad_norm": 185.91551208496094, "learning_rate": 7.329255861365954e-06, "loss": 3.7055, "step": 1048 }, { "epoch": 0.26729519684036185, "grad_norm": 9.110211372375488, "learning_rate": 7.326707441386342e-06, "loss": 0.0065, "step": 1049 }, { "epoch": 0.26755000637023824, "grad_norm": 10.299043655395508, "learning_rate": 7.324159021406728e-06, "loss": 0.0079, "step": 1050 }, { "epoch": 0.2678048159001147, "grad_norm": 36.952064514160156, "learning_rate": 7.321610601427116e-06, "loss": 0.0233, "step": 1051 }, { "epoch": 0.26805962542999107, "grad_norm": 9.072159767150879, "learning_rate": 7.319062181447503e-06, "loss": 0.0081, "step": 1052 }, { "epoch": 0.2683144349598675, "grad_norm": 70.0463638305664, "learning_rate": 7.31651376146789e-06, "loss": 0.0769, "step": 1053 }, { "epoch": 0.2685692444897439, "grad_norm": 98.56433868408203, "learning_rate": 7.313965341488278e-06, "loss": 3.645, "step": 1054 }, { "epoch": 0.26882405401962034, "grad_norm": 1.9257160425186157, "learning_rate": 7.311416921508665e-06, "loss": 0.0027, "step": 1055 }, { "epoch": 0.2690788635494967, "grad_norm": 3.8213553428649902, "learning_rate": 7.308868501529053e-06, "loss": 0.0025, "step": 1056 }, { "epoch": 0.26933367307937317, "grad_norm": 4.761744022369385, "learning_rate": 7.30632008154944e-06, "loss": 0.0046, "step": 1057 }, { "epoch": 0.2695884826092496, "grad_norm": 11.854000091552734, "learning_rate": 7.303771661569828e-06, "loss": 0.0125, "step": 1058 }, { "epoch": 0.269843292139126, "grad_norm": 0.8905635476112366, "learning_rate": 7.3012232415902155e-06, "loss": 0.0024, "step": 1059 }, { "epoch": 0.27009810166900244, "grad_norm": 2.7750048637390137, "learning_rate": 7.298674821610602e-06, "loss": 0.0019, "step": 1060 }, { "epoch": 0.27035291119887883, "grad_norm": 5.6296186447143555, "learning_rate": 7.2961264016309894e-06, "loss": 0.0065, "step": 1061 }, { "epoch": 0.2706077207287553, "grad_norm": 211.83546447753906, "learning_rate": 7.293577981651376e-06, "loss": 0.6854, "step": 1062 }, { "epoch": 0.27086253025863166, "grad_norm": 1.7805100679397583, "learning_rate": 7.291029561671764e-06, "loss": 0.0032, "step": 1063 }, { "epoch": 0.2711173397885081, "grad_norm": 6.184434413909912, "learning_rate": 7.288481141692152e-06, "loss": 0.0106, "step": 1064 }, { "epoch": 0.2713721493183845, "grad_norm": 1.7441006898880005, "learning_rate": 7.285932721712539e-06, "loss": 0.0023, "step": 1065 }, { "epoch": 0.27162695884826094, "grad_norm": 9.939352989196777, "learning_rate": 7.283384301732927e-06, "loss": 0.0091, "step": 1066 }, { "epoch": 0.2718817683781373, "grad_norm": 6.932032108306885, "learning_rate": 7.280835881753313e-06, "loss": 0.0046, "step": 1067 }, { "epoch": 0.27213657790801377, "grad_norm": 6.498997688293457, "learning_rate": 7.278287461773701e-06, "loss": 0.0046, "step": 1068 }, { "epoch": 0.27239138743789015, "grad_norm": 4.079516887664795, "learning_rate": 7.2757390417940884e-06, "loss": 0.0023, "step": 1069 }, { "epoch": 0.2726461969677666, "grad_norm": 4.690783500671387, "learning_rate": 7.273190621814475e-06, "loss": 0.003, "step": 1070 }, { "epoch": 0.27290100649764304, "grad_norm": 4.19354772567749, "learning_rate": 7.270642201834863e-06, "loss": 0.0026, "step": 1071 }, { "epoch": 0.2731558160275194, "grad_norm": 3.904369592666626, "learning_rate": 7.26809378185525e-06, "loss": 0.0023, "step": 1072 }, { "epoch": 0.27341062555739587, "grad_norm": 286.58837890625, "learning_rate": 7.265545361875638e-06, "loss": 0.0909, "step": 1073 }, { "epoch": 0.27366543508727226, "grad_norm": 9.985176086425781, "learning_rate": 7.262996941896026e-06, "loss": 0.0075, "step": 1074 }, { "epoch": 0.2739202446171487, "grad_norm": 9.345090866088867, "learning_rate": 7.260448521916412e-06, "loss": 0.007, "step": 1075 }, { "epoch": 0.2741750541470251, "grad_norm": 3.769547939300537, "learning_rate": 7.2579001019368e-06, "loss": 0.0055, "step": 1076 }, { "epoch": 0.27442986367690153, "grad_norm": 6.246638298034668, "learning_rate": 7.255351681957187e-06, "loss": 0.0044, "step": 1077 }, { "epoch": 0.2746846732067779, "grad_norm": 107.40975189208984, "learning_rate": 7.252803261977574e-06, "loss": 0.0959, "step": 1078 }, { "epoch": 0.27493948273665436, "grad_norm": 0.8831741213798523, "learning_rate": 7.250254841997962e-06, "loss": 0.0016, "step": 1079 }, { "epoch": 0.27519429226653075, "grad_norm": 10.967939376831055, "learning_rate": 7.247706422018349e-06, "loss": 0.0087, "step": 1080 }, { "epoch": 0.2754491017964072, "grad_norm": 12.086177825927734, "learning_rate": 7.245158002038737e-06, "loss": 0.0107, "step": 1081 }, { "epoch": 0.2757039113262836, "grad_norm": 21.87987518310547, "learning_rate": 7.242609582059124e-06, "loss": 0.0347, "step": 1082 }, { "epoch": 0.27595872085616, "grad_norm": 7.2161102294921875, "learning_rate": 7.240061162079512e-06, "loss": 0.004, "step": 1083 }, { "epoch": 0.27621353038603647, "grad_norm": 2.23244571685791, "learning_rate": 7.2375127420998995e-06, "loss": 0.0025, "step": 1084 }, { "epoch": 0.27646833991591285, "grad_norm": 2.5368218421936035, "learning_rate": 7.234964322120286e-06, "loss": 0.0021, "step": 1085 }, { "epoch": 0.2767231494457893, "grad_norm": 1.0255666971206665, "learning_rate": 7.232415902140673e-06, "loss": 0.0008, "step": 1086 }, { "epoch": 0.2769779589756657, "grad_norm": 7.523916244506836, "learning_rate": 7.22986748216106e-06, "loss": 0.0043, "step": 1087 }, { "epoch": 0.2772327685055421, "grad_norm": 94.79924011230469, "learning_rate": 7.227319062181448e-06, "loss": 0.6645, "step": 1088 }, { "epoch": 0.2774875780354185, "grad_norm": 1.4241063594818115, "learning_rate": 7.224770642201836e-06, "loss": 0.001, "step": 1089 }, { "epoch": 0.27774238756529496, "grad_norm": 1.0525959730148315, "learning_rate": 7.222222222222223e-06, "loss": 0.0004, "step": 1090 }, { "epoch": 0.27799719709517134, "grad_norm": 10.219755172729492, "learning_rate": 7.219673802242611e-06, "loss": 0.0079, "step": 1091 }, { "epoch": 0.2782520066250478, "grad_norm": 9.189131736755371, "learning_rate": 7.217125382262997e-06, "loss": 0.0062, "step": 1092 }, { "epoch": 0.2785068161549242, "grad_norm": 69.84357452392578, "learning_rate": 7.214576962283385e-06, "loss": 0.7126, "step": 1093 }, { "epoch": 0.2787616256848006, "grad_norm": 62.83428192138672, "learning_rate": 7.212028542303772e-06, "loss": 0.0361, "step": 1094 }, { "epoch": 0.279016435214677, "grad_norm": 5.571718692779541, "learning_rate": 7.209480122324159e-06, "loss": 0.0026, "step": 1095 }, { "epoch": 0.27927124474455345, "grad_norm": 180.99893188476562, "learning_rate": 7.206931702344547e-06, "loss": 0.4185, "step": 1096 }, { "epoch": 0.27952605427442984, "grad_norm": 37.3760986328125, "learning_rate": 7.204383282364934e-06, "loss": 0.0267, "step": 1097 }, { "epoch": 0.2797808638043063, "grad_norm": 4.92976188659668, "learning_rate": 7.201834862385322e-06, "loss": 0.002, "step": 1098 }, { "epoch": 0.2800356733341827, "grad_norm": 6.43076229095459, "learning_rate": 7.199286442405709e-06, "loss": 0.0033, "step": 1099 }, { "epoch": 0.2802904828640591, "grad_norm": 0.6753835082054138, "learning_rate": 7.196738022426097e-06, "loss": 0.0004, "step": 1100 }, { "epoch": 0.28054529239393555, "grad_norm": 4.34655237197876, "learning_rate": 7.1941896024464845e-06, "loss": 0.0052, "step": 1101 }, { "epoch": 0.28080010192381194, "grad_norm": 2.4949228763580322, "learning_rate": 7.1916411824668706e-06, "loss": 0.0045, "step": 1102 }, { "epoch": 0.2810549114536884, "grad_norm": 12.004493713378906, "learning_rate": 7.189092762487258e-06, "loss": 0.0103, "step": 1103 }, { "epoch": 0.28130972098356477, "grad_norm": 508.8042907714844, "learning_rate": 7.186544342507645e-06, "loss": 3.3166, "step": 1104 }, { "epoch": 0.2815645305134412, "grad_norm": 17.542526245117188, "learning_rate": 7.183995922528033e-06, "loss": 0.0214, "step": 1105 }, { "epoch": 0.2818193400433176, "grad_norm": 7.469714641571045, "learning_rate": 7.181447502548421e-06, "loss": 0.0092, "step": 1106 }, { "epoch": 0.28207414957319404, "grad_norm": 7.862711429595947, "learning_rate": 7.178899082568808e-06, "loss": 0.0046, "step": 1107 }, { "epoch": 0.28232895910307043, "grad_norm": 7.714081764221191, "learning_rate": 7.176350662589196e-06, "loss": 0.0044, "step": 1108 }, { "epoch": 0.2825837686329469, "grad_norm": 2.375978469848633, "learning_rate": 7.173802242609582e-06, "loss": 0.0008, "step": 1109 }, { "epoch": 0.28283857816282326, "grad_norm": 4.552841663360596, "learning_rate": 7.17125382262997e-06, "loss": 0.0019, "step": 1110 }, { "epoch": 0.2830933876926997, "grad_norm": 8.534311294555664, "learning_rate": 7.168705402650357e-06, "loss": 0.0054, "step": 1111 }, { "epoch": 0.28334819722257615, "grad_norm": 0.7981840968132019, "learning_rate": 7.166156982670744e-06, "loss": 0.0007, "step": 1112 }, { "epoch": 0.28360300675245254, "grad_norm": 3.400076150894165, "learning_rate": 7.163608562691132e-06, "loss": 0.0016, "step": 1113 }, { "epoch": 0.283857816282329, "grad_norm": 2.5448877811431885, "learning_rate": 7.161060142711519e-06, "loss": 0.0015, "step": 1114 }, { "epoch": 0.28411262581220537, "grad_norm": 58.992958068847656, "learning_rate": 7.158511722731907e-06, "loss": 0.657, "step": 1115 }, { "epoch": 0.2843674353420818, "grad_norm": 1.2672390937805176, "learning_rate": 7.155963302752295e-06, "loss": 0.0006, "step": 1116 }, { "epoch": 0.2846222448719582, "grad_norm": 59.46980285644531, "learning_rate": 7.153414882772682e-06, "loss": 0.59, "step": 1117 }, { "epoch": 0.28487705440183464, "grad_norm": 2.030886173248291, "learning_rate": 7.1508664627930694e-06, "loss": 0.0014, "step": 1118 }, { "epoch": 0.285131863931711, "grad_norm": 2.025970697402954, "learning_rate": 7.1483180428134555e-06, "loss": 0.0019, "step": 1119 }, { "epoch": 0.28538667346158747, "grad_norm": 578.2003784179688, "learning_rate": 7.145769622833843e-06, "loss": 0.2795, "step": 1120 }, { "epoch": 0.28564148299146386, "grad_norm": 328.8240661621094, "learning_rate": 7.143221202854231e-06, "loss": 0.7464, "step": 1121 }, { "epoch": 0.2858962925213403, "grad_norm": 6.520992755889893, "learning_rate": 7.140672782874618e-06, "loss": 0.0036, "step": 1122 }, { "epoch": 0.2861511020512167, "grad_norm": 3.0316903591156006, "learning_rate": 7.138124362895006e-06, "loss": 0.0015, "step": 1123 }, { "epoch": 0.28640591158109313, "grad_norm": 3.375819683074951, "learning_rate": 7.135575942915393e-06, "loss": 0.0031, "step": 1124 }, { "epoch": 0.2866607211109696, "grad_norm": 1.6158901453018188, "learning_rate": 7.133027522935781e-06, "loss": 0.0007, "step": 1125 }, { "epoch": 0.28691553064084596, "grad_norm": 18.030553817749023, "learning_rate": 7.1304791029561684e-06, "loss": 0.0078, "step": 1126 }, { "epoch": 0.2871703401707224, "grad_norm": 6.241525173187256, "learning_rate": 7.1279306829765545e-06, "loss": 0.0034, "step": 1127 }, { "epoch": 0.2874251497005988, "grad_norm": 3.589994192123413, "learning_rate": 7.125382262996942e-06, "loss": 0.0018, "step": 1128 }, { "epoch": 0.28767995923047524, "grad_norm": 2.383620500564575, "learning_rate": 7.122833843017329e-06, "loss": 0.0009, "step": 1129 }, { "epoch": 0.2879347687603516, "grad_norm": 2.669050455093384, "learning_rate": 7.120285423037717e-06, "loss": 0.0012, "step": 1130 }, { "epoch": 0.28818957829022807, "grad_norm": 7.762860298156738, "learning_rate": 7.117737003058105e-06, "loss": 0.0083, "step": 1131 }, { "epoch": 0.28844438782010445, "grad_norm": 2.8118646144866943, "learning_rate": 7.115188583078492e-06, "loss": 0.0025, "step": 1132 }, { "epoch": 0.2886991973499809, "grad_norm": 1.9972026348114014, "learning_rate": 7.11264016309888e-06, "loss": 0.0031, "step": 1133 }, { "epoch": 0.2889540068798573, "grad_norm": 3.2298309803009033, "learning_rate": 7.110091743119267e-06, "loss": 0.0014, "step": 1134 }, { "epoch": 0.2892088164097337, "grad_norm": 137.86570739746094, "learning_rate": 7.107543323139654e-06, "loss": 0.391, "step": 1135 }, { "epoch": 0.2894636259396101, "grad_norm": 7.984918117523193, "learning_rate": 7.104994903160042e-06, "loss": 0.006, "step": 1136 }, { "epoch": 0.28971843546948656, "grad_norm": 2.095956802368164, "learning_rate": 7.102446483180428e-06, "loss": 0.001, "step": 1137 }, { "epoch": 0.289973244999363, "grad_norm": 5.867272853851318, "learning_rate": 7.099898063200816e-06, "loss": 0.0035, "step": 1138 }, { "epoch": 0.2902280545292394, "grad_norm": 0.6440432667732239, "learning_rate": 7.097349643221203e-06, "loss": 0.0005, "step": 1139 }, { "epoch": 0.29048286405911583, "grad_norm": 1.298549771308899, "learning_rate": 7.094801223241591e-06, "loss": 0.0017, "step": 1140 }, { "epoch": 0.2907376735889922, "grad_norm": 10.497100830078125, "learning_rate": 7.092252803261979e-06, "loss": 0.0085, "step": 1141 }, { "epoch": 0.29099248311886866, "grad_norm": 7.525472164154053, "learning_rate": 7.089704383282366e-06, "loss": 0.0045, "step": 1142 }, { "epoch": 0.29124729264874505, "grad_norm": 10.913243293762207, "learning_rate": 7.087155963302753e-06, "loss": 0.0085, "step": 1143 }, { "epoch": 0.2915021021786215, "grad_norm": 7.507528305053711, "learning_rate": 7.0846075433231395e-06, "loss": 0.0043, "step": 1144 }, { "epoch": 0.2917569117084979, "grad_norm": 5.201216697692871, "learning_rate": 7.082059123343527e-06, "loss": 0.0035, "step": 1145 }, { "epoch": 0.2920117212383743, "grad_norm": 1.810828447341919, "learning_rate": 7.079510703363915e-06, "loss": 0.0006, "step": 1146 }, { "epoch": 0.2922665307682507, "grad_norm": 96.31836700439453, "learning_rate": 7.076962283384302e-06, "loss": 3.9224, "step": 1147 }, { "epoch": 0.29252134029812715, "grad_norm": 7.016152858734131, "learning_rate": 7.07441386340469e-06, "loss": 0.004, "step": 1148 }, { "epoch": 0.29277614982800354, "grad_norm": 8.901239395141602, "learning_rate": 7.071865443425077e-06, "loss": 0.0071, "step": 1149 }, { "epoch": 0.29303095935788, "grad_norm": 124.38504791259766, "learning_rate": 7.069317023445465e-06, "loss": 0.53, "step": 1150 }, { "epoch": 0.2932857688877564, "grad_norm": 5.439929008483887, "learning_rate": 7.066768603465852e-06, "loss": 0.0034, "step": 1151 }, { "epoch": 0.2935405784176328, "grad_norm": 361.53582763671875, "learning_rate": 7.0642201834862385e-06, "loss": 0.5315, "step": 1152 }, { "epoch": 0.29379538794750926, "grad_norm": 0.9768308401107788, "learning_rate": 7.061671763506627e-06, "loss": 0.0012, "step": 1153 }, { "epoch": 0.29405019747738564, "grad_norm": 1.4300659894943237, "learning_rate": 7.059123343527013e-06, "loss": 0.0008, "step": 1154 }, { "epoch": 0.2943050070072621, "grad_norm": 154.76731872558594, "learning_rate": 7.056574923547401e-06, "loss": 0.5122, "step": 1155 }, { "epoch": 0.2945598165371385, "grad_norm": 2.1863934993743896, "learning_rate": 7.054026503567789e-06, "loss": 0.0024, "step": 1156 }, { "epoch": 0.2948146260670149, "grad_norm": 2.2769899368286133, "learning_rate": 7.051478083588176e-06, "loss": 0.0014, "step": 1157 }, { "epoch": 0.2950694355968913, "grad_norm": 4.478065490722656, "learning_rate": 7.048929663608564e-06, "loss": 0.0026, "step": 1158 }, { "epoch": 0.29532424512676775, "grad_norm": 1.0653502941131592, "learning_rate": 7.046381243628951e-06, "loss": 0.0013, "step": 1159 }, { "epoch": 0.29557905465664414, "grad_norm": 3.3568339347839355, "learning_rate": 7.043832823649338e-06, "loss": 0.0028, "step": 1160 }, { "epoch": 0.2958338641865206, "grad_norm": 2.3835268020629883, "learning_rate": 7.041284403669726e-06, "loss": 0.0014, "step": 1161 }, { "epoch": 0.29608867371639697, "grad_norm": 1.4275922775268555, "learning_rate": 7.038735983690112e-06, "loss": 0.0011, "step": 1162 }, { "epoch": 0.2963434832462734, "grad_norm": 5.16569709777832, "learning_rate": 7.0361875637105e-06, "loss": 0.0022, "step": 1163 }, { "epoch": 0.29659829277614985, "grad_norm": 32.20552062988281, "learning_rate": 7.033639143730887e-06, "loss": 0.0383, "step": 1164 }, { "epoch": 0.29685310230602624, "grad_norm": 4.547721862792969, "learning_rate": 7.031090723751275e-06, "loss": 0.0036, "step": 1165 }, { "epoch": 0.2971079118359027, "grad_norm": 6.611216068267822, "learning_rate": 7.028542303771663e-06, "loss": 0.0064, "step": 1166 }, { "epoch": 0.29736272136577907, "grad_norm": 7.474546432495117, "learning_rate": 7.02599388379205e-06, "loss": 0.0028, "step": 1167 }, { "epoch": 0.2976175308956555, "grad_norm": 1.137627363204956, "learning_rate": 7.023445463812437e-06, "loss": 0.0005, "step": 1168 }, { "epoch": 0.2978723404255319, "grad_norm": 5.947595596313477, "learning_rate": 7.0208970438328235e-06, "loss": 0.0035, "step": 1169 }, { "epoch": 0.29812714995540834, "grad_norm": 26.644168853759766, "learning_rate": 7.018348623853211e-06, "loss": 0.0195, "step": 1170 }, { "epoch": 0.29838195948528473, "grad_norm": 2.8528075218200684, "learning_rate": 7.015800203873598e-06, "loss": 0.0011, "step": 1171 }, { "epoch": 0.2986367690151612, "grad_norm": 5.212165355682373, "learning_rate": 7.013251783893986e-06, "loss": 0.0023, "step": 1172 }, { "epoch": 0.29889157854503756, "grad_norm": 5.20858097076416, "learning_rate": 7.010703363914374e-06, "loss": 0.0036, "step": 1173 }, { "epoch": 0.299146388074914, "grad_norm": 38.70458984375, "learning_rate": 7.008154943934761e-06, "loss": 0.0204, "step": 1174 }, { "epoch": 0.2994011976047904, "grad_norm": 3.433619260787964, "learning_rate": 7.005606523955149e-06, "loss": 0.0014, "step": 1175 }, { "epoch": 0.29965600713466684, "grad_norm": 0.36577895283699036, "learning_rate": 7.0030581039755356e-06, "loss": 0.001, "step": 1176 }, { "epoch": 0.2999108166645433, "grad_norm": 1.3443719148635864, "learning_rate": 7.000509683995923e-06, "loss": 0.0006, "step": 1177 }, { "epoch": 0.30016562619441967, "grad_norm": 2.104318380355835, "learning_rate": 6.997961264016311e-06, "loss": 0.001, "step": 1178 }, { "epoch": 0.3004204357242961, "grad_norm": 7.331689834594727, "learning_rate": 6.995412844036697e-06, "loss": 0.0042, "step": 1179 }, { "epoch": 0.3006752452541725, "grad_norm": 10.739809036254883, "learning_rate": 6.992864424057085e-06, "loss": 0.0067, "step": 1180 }, { "epoch": 0.30093005478404894, "grad_norm": 1.923504114151001, "learning_rate": 6.990316004077472e-06, "loss": 0.001, "step": 1181 }, { "epoch": 0.3011848643139253, "grad_norm": 6.208507061004639, "learning_rate": 6.98776758409786e-06, "loss": 0.0049, "step": 1182 }, { "epoch": 0.30143967384380177, "grad_norm": 6.484943389892578, "learning_rate": 6.985219164118248e-06, "loss": 0.0077, "step": 1183 }, { "epoch": 0.30169448337367816, "grad_norm": 6.394095420837402, "learning_rate": 6.9826707441386346e-06, "loss": 0.0038, "step": 1184 }, { "epoch": 0.3019492929035546, "grad_norm": 11.954566955566406, "learning_rate": 6.980122324159022e-06, "loss": 0.01, "step": 1185 }, { "epoch": 0.302204102433431, "grad_norm": 0.7947127819061279, "learning_rate": 6.9775739041794085e-06, "loss": 0.0012, "step": 1186 }, { "epoch": 0.30245891196330743, "grad_norm": 5.4980082511901855, "learning_rate": 6.975025484199796e-06, "loss": 0.0024, "step": 1187 }, { "epoch": 0.3027137214931838, "grad_norm": 3.992075204849243, "learning_rate": 6.972477064220184e-06, "loss": 0.0014, "step": 1188 }, { "epoch": 0.30296853102306026, "grad_norm": 5.817104816436768, "learning_rate": 6.969928644240571e-06, "loss": 0.0027, "step": 1189 }, { "epoch": 0.3032233405529367, "grad_norm": 168.0584716796875, "learning_rate": 6.967380224260959e-06, "loss": 0.2413, "step": 1190 }, { "epoch": 0.3034781500828131, "grad_norm": 5.7129926681518555, "learning_rate": 6.964831804281346e-06, "loss": 0.0026, "step": 1191 }, { "epoch": 0.30373295961268953, "grad_norm": 85.12960815429688, "learning_rate": 6.9622833843017336e-06, "loss": 0.47, "step": 1192 }, { "epoch": 0.3039877691425659, "grad_norm": 11.105542182922363, "learning_rate": 6.959734964322121e-06, "loss": 0.0085, "step": 1193 }, { "epoch": 0.30424257867244237, "grad_norm": 17.91176986694336, "learning_rate": 6.957186544342508e-06, "loss": 0.0226, "step": 1194 }, { "epoch": 0.30449738820231875, "grad_norm": 8.830971717834473, "learning_rate": 6.954638124362896e-06, "loss": 0.007, "step": 1195 }, { "epoch": 0.3047521977321952, "grad_norm": 7.08978796005249, "learning_rate": 6.952089704383282e-06, "loss": 0.0036, "step": 1196 }, { "epoch": 0.3050070072620716, "grad_norm": 1.8349940776824951, "learning_rate": 6.94954128440367e-06, "loss": 0.0034, "step": 1197 }, { "epoch": 0.305261816791948, "grad_norm": 4.004942417144775, "learning_rate": 6.946992864424058e-06, "loss": 0.0044, "step": 1198 }, { "epoch": 0.3055166263218244, "grad_norm": 1.15510094165802, "learning_rate": 6.944444444444445e-06, "loss": 0.0004, "step": 1199 }, { "epoch": 0.30577143585170086, "grad_norm": 2.20208477973938, "learning_rate": 6.941896024464833e-06, "loss": 0.0007, "step": 1200 }, { "epoch": 0.30602624538157724, "grad_norm": 4.003388404846191, "learning_rate": 6.9393476044852195e-06, "loss": 0.0014, "step": 1201 }, { "epoch": 0.3062810549114537, "grad_norm": 2.1341841220855713, "learning_rate": 6.936799184505607e-06, "loss": 0.0008, "step": 1202 }, { "epoch": 0.30653586444133013, "grad_norm": 1.5670182704925537, "learning_rate": 6.934250764525995e-06, "loss": 0.0008, "step": 1203 }, { "epoch": 0.3067906739712065, "grad_norm": 9.724214553833008, "learning_rate": 6.931702344546381e-06, "loss": 0.0103, "step": 1204 }, { "epoch": 0.30704548350108296, "grad_norm": 10.010737419128418, "learning_rate": 6.929153924566769e-06, "loss": 0.0084, "step": 1205 }, { "epoch": 0.30730029303095935, "grad_norm": 1.142236351966858, "learning_rate": 6.926605504587156e-06, "loss": 0.0005, "step": 1206 }, { "epoch": 0.3075551025608358, "grad_norm": 12.97903060913086, "learning_rate": 6.924057084607544e-06, "loss": 0.0116, "step": 1207 }, { "epoch": 0.3078099120907122, "grad_norm": 3.312836170196533, "learning_rate": 6.921508664627932e-06, "loss": 0.0056, "step": 1208 }, { "epoch": 0.3080647216205886, "grad_norm": 0.6309067010879517, "learning_rate": 6.9189602446483185e-06, "loss": 0.0004, "step": 1209 }, { "epoch": 0.308319531150465, "grad_norm": 0.39451077580451965, "learning_rate": 6.916411824668706e-06, "loss": 0.0007, "step": 1210 }, { "epoch": 0.30857434068034145, "grad_norm": 13.661370277404785, "learning_rate": 6.913863404689093e-06, "loss": 0.0147, "step": 1211 }, { "epoch": 0.30882915021021784, "grad_norm": 4.648937702178955, "learning_rate": 6.911314984709481e-06, "loss": 0.0018, "step": 1212 }, { "epoch": 0.3090839597400943, "grad_norm": 4.504700660705566, "learning_rate": 6.908766564729869e-06, "loss": 0.0017, "step": 1213 }, { "epoch": 0.30933876926997067, "grad_norm": 3.413278102874756, "learning_rate": 6.906218144750255e-06, "loss": 0.0014, "step": 1214 }, { "epoch": 0.3095935787998471, "grad_norm": 0.3992582857608795, "learning_rate": 6.903669724770643e-06, "loss": 0.0003, "step": 1215 }, { "epoch": 0.30984838832972356, "grad_norm": 8.307916641235352, "learning_rate": 6.90112130479103e-06, "loss": 0.0049, "step": 1216 }, { "epoch": 0.31010319785959994, "grad_norm": 14.395406723022461, "learning_rate": 6.8985728848114175e-06, "loss": 0.0149, "step": 1217 }, { "epoch": 0.3103580073894764, "grad_norm": 10.306292533874512, "learning_rate": 6.896024464831805e-06, "loss": 0.0077, "step": 1218 }, { "epoch": 0.3106128169193528, "grad_norm": 4.945379257202148, "learning_rate": 6.893476044852192e-06, "loss": 0.0025, "step": 1219 }, { "epoch": 0.3108676264492292, "grad_norm": 1.9216362237930298, "learning_rate": 6.89092762487258e-06, "loss": 0.0005, "step": 1220 }, { "epoch": 0.3111224359791056, "grad_norm": 4.233333110809326, "learning_rate": 6.888379204892966e-06, "loss": 0.0058, "step": 1221 }, { "epoch": 0.31137724550898205, "grad_norm": 6.409354209899902, "learning_rate": 6.885830784913354e-06, "loss": 0.0045, "step": 1222 }, { "epoch": 0.31163205503885844, "grad_norm": 8.914352416992188, "learning_rate": 6.883282364933742e-06, "loss": 0.006, "step": 1223 }, { "epoch": 0.3118868645687349, "grad_norm": 12.951192855834961, "learning_rate": 6.880733944954129e-06, "loss": 0.0099, "step": 1224 }, { "epoch": 0.31214167409861127, "grad_norm": 14.573563575744629, "learning_rate": 6.8781855249745166e-06, "loss": 0.0147, "step": 1225 }, { "epoch": 0.3123964836284877, "grad_norm": 5.8799638748168945, "learning_rate": 6.8756371049949035e-06, "loss": 0.0042, "step": 1226 }, { "epoch": 0.3126512931583641, "grad_norm": 1.2741020917892456, "learning_rate": 6.873088685015291e-06, "loss": 0.0014, "step": 1227 }, { "epoch": 0.31290610268824054, "grad_norm": 2.1011388301849365, "learning_rate": 6.870540265035679e-06, "loss": 0.0011, "step": 1228 }, { "epoch": 0.313160912218117, "grad_norm": 0.17449143528938293, "learning_rate": 6.867991845056066e-06, "loss": 0.0004, "step": 1229 }, { "epoch": 0.31341572174799337, "grad_norm": 14.097684860229492, "learning_rate": 6.865443425076454e-06, "loss": 0.0154, "step": 1230 }, { "epoch": 0.3136705312778698, "grad_norm": 6.698642253875732, "learning_rate": 6.86289500509684e-06, "loss": 0.0032, "step": 1231 }, { "epoch": 0.3139253408077462, "grad_norm": 11.76547622680664, "learning_rate": 6.860346585117228e-06, "loss": 0.0097, "step": 1232 }, { "epoch": 0.31418015033762264, "grad_norm": 43.35481643676758, "learning_rate": 6.8577981651376156e-06, "loss": 0.0227, "step": 1233 }, { "epoch": 0.31443495986749903, "grad_norm": 6.4721150398254395, "learning_rate": 6.8552497451580025e-06, "loss": 0.0034, "step": 1234 }, { "epoch": 0.3146897693973755, "grad_norm": 5.889163017272949, "learning_rate": 6.85270132517839e-06, "loss": 0.0026, "step": 1235 }, { "epoch": 0.31494457892725186, "grad_norm": 8.758374214172363, "learning_rate": 6.850152905198777e-06, "loss": 0.0064, "step": 1236 }, { "epoch": 0.3151993884571283, "grad_norm": 5.815447807312012, "learning_rate": 6.847604485219165e-06, "loss": 0.0034, "step": 1237 }, { "epoch": 0.3154541979870047, "grad_norm": 120.14398193359375, "learning_rate": 6.845056065239553e-06, "loss": 0.047, "step": 1238 }, { "epoch": 0.31570900751688113, "grad_norm": 11.502256393432617, "learning_rate": 6.842507645259939e-06, "loss": 0.0092, "step": 1239 }, { "epoch": 0.3159638170467575, "grad_norm": 18.661029815673828, "learning_rate": 6.839959225280327e-06, "loss": 0.024, "step": 1240 }, { "epoch": 0.31621862657663397, "grad_norm": 21.413724899291992, "learning_rate": 6.837410805300714e-06, "loss": 0.0309, "step": 1241 }, { "epoch": 0.3164734361065104, "grad_norm": 19.835039138793945, "learning_rate": 6.8348623853211015e-06, "loss": 0.027, "step": 1242 }, { "epoch": 0.3167282456363868, "grad_norm": 19.0180606842041, "learning_rate": 6.832313965341489e-06, "loss": 0.0248, "step": 1243 }, { "epoch": 0.31698305516626324, "grad_norm": 13.499749183654785, "learning_rate": 6.829765545361876e-06, "loss": 0.0123, "step": 1244 }, { "epoch": 0.3172378646961396, "grad_norm": 4.93202543258667, "learning_rate": 6.827217125382264e-06, "loss": 0.0044, "step": 1245 }, { "epoch": 0.31749267422601607, "grad_norm": 0.8197073936462402, "learning_rate": 6.82466870540265e-06, "loss": 0.0009, "step": 1246 }, { "epoch": 0.31774748375589246, "grad_norm": 2.7741665840148926, "learning_rate": 6.822120285423038e-06, "loss": 0.0014, "step": 1247 }, { "epoch": 0.3180022932857689, "grad_norm": 5.62083625793457, "learning_rate": 6.819571865443425e-06, "loss": 0.0034, "step": 1248 }, { "epoch": 0.3182571028156453, "grad_norm": 9.277347564697266, "learning_rate": 6.817023445463813e-06, "loss": 0.0084, "step": 1249 }, { "epoch": 0.31851191234552173, "grad_norm": 13.212928771972656, "learning_rate": 6.8144750254842005e-06, "loss": 0.0141, "step": 1250 }, { "epoch": 0.3187667218753981, "grad_norm": 7.179544448852539, "learning_rate": 6.8119266055045875e-06, "loss": 0.0038, "step": 1251 }, { "epoch": 0.31902153140527456, "grad_norm": 3.1385128498077393, "learning_rate": 6.809378185524975e-06, "loss": 0.001, "step": 1252 }, { "epoch": 0.31927634093515095, "grad_norm": 4.188614845275879, "learning_rate": 6.806829765545362e-06, "loss": 0.0016, "step": 1253 }, { "epoch": 0.3195311504650274, "grad_norm": 95.81249237060547, "learning_rate": 6.80428134556575e-06, "loss": 0.0943, "step": 1254 }, { "epoch": 0.31978595999490383, "grad_norm": 1.7140514850616455, "learning_rate": 6.801732925586138e-06, "loss": 0.0005, "step": 1255 }, { "epoch": 0.3200407695247802, "grad_norm": 3.327768564224243, "learning_rate": 6.799184505606524e-06, "loss": 0.0033, "step": 1256 }, { "epoch": 0.32029557905465667, "grad_norm": 0.6673057079315186, "learning_rate": 6.796636085626912e-06, "loss": 0.0006, "step": 1257 }, { "epoch": 0.32055038858453305, "grad_norm": 5.080882549285889, "learning_rate": 6.794087665647299e-06, "loss": 0.0021, "step": 1258 }, { "epoch": 0.3208051981144095, "grad_norm": 1.5604559183120728, "learning_rate": 6.7915392456676865e-06, "loss": 0.002, "step": 1259 }, { "epoch": 0.3210600076442859, "grad_norm": 7.360530376434326, "learning_rate": 6.788990825688074e-06, "loss": 0.004, "step": 1260 }, { "epoch": 0.3213148171741623, "grad_norm": 4.25223445892334, "learning_rate": 6.786442405708461e-06, "loss": 0.0015, "step": 1261 }, { "epoch": 0.3215696267040387, "grad_norm": 0.9515922665596008, "learning_rate": 6.783893985728849e-06, "loss": 0.0004, "step": 1262 }, { "epoch": 0.32182443623391516, "grad_norm": 3.601093053817749, "learning_rate": 6.781345565749235e-06, "loss": 0.0028, "step": 1263 }, { "epoch": 0.32207924576379154, "grad_norm": 1.4529156684875488, "learning_rate": 6.778797145769623e-06, "loss": 0.0008, "step": 1264 }, { "epoch": 0.322334055293668, "grad_norm": 5.582183837890625, "learning_rate": 6.776248725790011e-06, "loss": 0.0025, "step": 1265 }, { "epoch": 0.3225888648235444, "grad_norm": 64.9097900390625, "learning_rate": 6.773700305810398e-06, "loss": 0.6503, "step": 1266 }, { "epoch": 0.3228436743534208, "grad_norm": 10.444249153137207, "learning_rate": 6.7711518858307855e-06, "loss": 0.0087, "step": 1267 }, { "epoch": 0.32309848388329726, "grad_norm": 0.9449276328086853, "learning_rate": 6.7686034658511725e-06, "loss": 0.0003, "step": 1268 }, { "epoch": 0.32335329341317365, "grad_norm": 0.41693875193595886, "learning_rate": 6.76605504587156e-06, "loss": 0.0003, "step": 1269 }, { "epoch": 0.3236081029430501, "grad_norm": 0.4287775754928589, "learning_rate": 6.763506625891948e-06, "loss": 0.0008, "step": 1270 }, { "epoch": 0.3238629124729265, "grad_norm": 4.072312831878662, "learning_rate": 6.760958205912335e-06, "loss": 0.0031, "step": 1271 }, { "epoch": 0.3241177220028029, "grad_norm": 23.2770938873291, "learning_rate": 6.758409785932723e-06, "loss": 0.0513, "step": 1272 }, { "epoch": 0.3243725315326793, "grad_norm": 114.1397476196289, "learning_rate": 6.755861365953109e-06, "loss": 0.5823, "step": 1273 }, { "epoch": 0.32462734106255575, "grad_norm": 1.7250241041183472, "learning_rate": 6.753312945973497e-06, "loss": 0.0005, "step": 1274 }, { "epoch": 0.32488215059243214, "grad_norm": 11.052684783935547, "learning_rate": 6.7507645259938845e-06, "loss": 0.0104, "step": 1275 }, { "epoch": 0.3251369601223086, "grad_norm": 3.510303258895874, "learning_rate": 6.7482161060142715e-06, "loss": 0.0011, "step": 1276 }, { "epoch": 0.32539176965218497, "grad_norm": 10.632192611694336, "learning_rate": 6.745667686034659e-06, "loss": 0.0034, "step": 1277 }, { "epoch": 0.3256465791820614, "grad_norm": 0.9783118963241577, "learning_rate": 6.743119266055046e-06, "loss": 0.0006, "step": 1278 }, { "epoch": 0.3259013887119378, "grad_norm": 7.224955081939697, "learning_rate": 6.740570846075434e-06, "loss": 0.005, "step": 1279 }, { "epoch": 0.32615619824181424, "grad_norm": 1.1408355236053467, "learning_rate": 6.738022426095822e-06, "loss": 0.0009, "step": 1280 }, { "epoch": 0.3264110077716907, "grad_norm": 3.6244094371795654, "learning_rate": 6.735474006116208e-06, "loss": 0.0045, "step": 1281 }, { "epoch": 0.3266658173015671, "grad_norm": 0.5275117754936218, "learning_rate": 6.732925586136596e-06, "loss": 0.0009, "step": 1282 }, { "epoch": 0.3269206268314435, "grad_norm": 66.08354949951172, "learning_rate": 6.730377166156983e-06, "loss": 0.1048, "step": 1283 }, { "epoch": 0.3271754363613199, "grad_norm": 2.345583200454712, "learning_rate": 6.7278287461773705e-06, "loss": 0.0021, "step": 1284 }, { "epoch": 0.32743024589119635, "grad_norm": 3.916280508041382, "learning_rate": 6.725280326197758e-06, "loss": 0.0015, "step": 1285 }, { "epoch": 0.32768505542107274, "grad_norm": 9.411942481994629, "learning_rate": 6.722731906218145e-06, "loss": 0.0065, "step": 1286 }, { "epoch": 0.3279398649509492, "grad_norm": 6.922333717346191, "learning_rate": 6.720183486238533e-06, "loss": 0.004, "step": 1287 }, { "epoch": 0.32819467448082557, "grad_norm": 0.9435524940490723, "learning_rate": 6.71763506625892e-06, "loss": 0.0008, "step": 1288 }, { "epoch": 0.328449484010702, "grad_norm": 1.443350911140442, "learning_rate": 6.715086646279308e-06, "loss": 0.0033, "step": 1289 }, { "epoch": 0.3287042935405784, "grad_norm": 3.1084585189819336, "learning_rate": 6.7125382262996956e-06, "loss": 0.001, "step": 1290 }, { "epoch": 0.32895910307045484, "grad_norm": 10.5692777633667, "learning_rate": 6.709989806320082e-06, "loss": 0.0113, "step": 1291 }, { "epoch": 0.3292139126003312, "grad_norm": 2.6000590324401855, "learning_rate": 6.7074413863404695e-06, "loss": 0.0012, "step": 1292 }, { "epoch": 0.32946872213020767, "grad_norm": 0.9738119840621948, "learning_rate": 6.7048929663608564e-06, "loss": 0.0009, "step": 1293 }, { "epoch": 0.3297235316600841, "grad_norm": 3.983177661895752, "learning_rate": 6.702344546381244e-06, "loss": 0.0013, "step": 1294 }, { "epoch": 0.3299783411899605, "grad_norm": 3.260266065597534, "learning_rate": 6.699796126401632e-06, "loss": 0.0016, "step": 1295 }, { "epoch": 0.33023315071983694, "grad_norm": 3.335101842880249, "learning_rate": 6.697247706422019e-06, "loss": 0.0018, "step": 1296 }, { "epoch": 0.33048796024971333, "grad_norm": 7.07432746887207, "learning_rate": 6.694699286442407e-06, "loss": 0.0036, "step": 1297 }, { "epoch": 0.3307427697795898, "grad_norm": 1.4544627666473389, "learning_rate": 6.692150866462793e-06, "loss": 0.0007, "step": 1298 }, { "epoch": 0.33099757930946616, "grad_norm": 6.748089790344238, "learning_rate": 6.689602446483181e-06, "loss": 0.0041, "step": 1299 }, { "epoch": 0.3312523888393426, "grad_norm": 2.087160110473633, "learning_rate": 6.6870540265035685e-06, "loss": 0.0005, "step": 1300 }, { "epoch": 0.331507198369219, "grad_norm": 101.36763000488281, "learning_rate": 6.6845056065239554e-06, "loss": 4.4067, "step": 1301 }, { "epoch": 0.33176200789909543, "grad_norm": 2.3523287773132324, "learning_rate": 6.681957186544343e-06, "loss": 0.0007, "step": 1302 }, { "epoch": 0.3320168174289718, "grad_norm": 2.6990439891815186, "learning_rate": 6.67940876656473e-06, "loss": 0.0013, "step": 1303 }, { "epoch": 0.33227162695884827, "grad_norm": 7.463613986968994, "learning_rate": 6.676860346585118e-06, "loss": 0.0053, "step": 1304 }, { "epoch": 0.33252643648872465, "grad_norm": 4.400005340576172, "learning_rate": 6.674311926605506e-06, "loss": 0.0033, "step": 1305 }, { "epoch": 0.3327812460186011, "grad_norm": 3.225013494491577, "learning_rate": 6.671763506625893e-06, "loss": 0.0011, "step": 1306 }, { "epoch": 0.33303605554847754, "grad_norm": 2.822521924972534, "learning_rate": 6.6692150866462805e-06, "loss": 0.001, "step": 1307 }, { "epoch": 0.3332908650783539, "grad_norm": 2.707946300506592, "learning_rate": 6.666666666666667e-06, "loss": 0.0026, "step": 1308 }, { "epoch": 0.33354567460823037, "grad_norm": 7.965229511260986, "learning_rate": 6.6641182466870544e-06, "loss": 0.0049, "step": 1309 }, { "epoch": 0.33380048413810676, "grad_norm": 204.60665893554688, "learning_rate": 6.661569826707442e-06, "loss": 0.0743, "step": 1310 }, { "epoch": 0.3340552936679832, "grad_norm": 3.675734758377075, "learning_rate": 6.659021406727829e-06, "loss": 0.0017, "step": 1311 }, { "epoch": 0.3343101031978596, "grad_norm": 1.1427645683288574, "learning_rate": 6.656472986748217e-06, "loss": 0.0008, "step": 1312 }, { "epoch": 0.33456491272773603, "grad_norm": 2.167402505874634, "learning_rate": 6.653924566768604e-06, "loss": 0.0011, "step": 1313 }, { "epoch": 0.3348197222576124, "grad_norm": 4.63723087310791, "learning_rate": 6.651376146788992e-06, "loss": 0.0022, "step": 1314 }, { "epoch": 0.33507453178748886, "grad_norm": 5.385533332824707, "learning_rate": 6.6488277268093796e-06, "loss": 0.0027, "step": 1315 }, { "epoch": 0.33532934131736525, "grad_norm": 6.796685218811035, "learning_rate": 6.646279306829766e-06, "loss": 0.0038, "step": 1316 }, { "epoch": 0.3355841508472417, "grad_norm": 8.604466438293457, "learning_rate": 6.6437308868501535e-06, "loss": 0.0056, "step": 1317 }, { "epoch": 0.3358389603771181, "grad_norm": 2.6935300827026367, "learning_rate": 6.64118246687054e-06, "loss": 0.0027, "step": 1318 }, { "epoch": 0.3360937699069945, "grad_norm": 4.155214786529541, "learning_rate": 6.638634046890928e-06, "loss": 0.0018, "step": 1319 }, { "epoch": 0.33634857943687096, "grad_norm": 3.8676891326904297, "learning_rate": 6.636085626911316e-06, "loss": 0.0018, "step": 1320 }, { "epoch": 0.33660338896674735, "grad_norm": 11.178906440734863, "learning_rate": 6.633537206931703e-06, "loss": 0.0094, "step": 1321 }, { "epoch": 0.3368581984966238, "grad_norm": 6.856326103210449, "learning_rate": 6.630988786952091e-06, "loss": 0.0042, "step": 1322 }, { "epoch": 0.3371130080265002, "grad_norm": 4.469374179840088, "learning_rate": 6.628440366972477e-06, "loss": 0.0038, "step": 1323 }, { "epoch": 0.3373678175563766, "grad_norm": 6.737509250640869, "learning_rate": 6.625891946992865e-06, "loss": 0.004, "step": 1324 }, { "epoch": 0.337622627086253, "grad_norm": 96.98795318603516, "learning_rate": 6.623343527013252e-06, "loss": 3.5033, "step": 1325 }, { "epoch": 0.33787743661612946, "grad_norm": 2.0819497108459473, "learning_rate": 6.620795107033639e-06, "loss": 0.001, "step": 1326 }, { "epoch": 0.33813224614600584, "grad_norm": 3.025082588195801, "learning_rate": 6.618246687054027e-06, "loss": 0.0013, "step": 1327 }, { "epoch": 0.3383870556758823, "grad_norm": 145.63189697265625, "learning_rate": 6.615698267074414e-06, "loss": 1.6596, "step": 1328 }, { "epoch": 0.3386418652057587, "grad_norm": 7.372653007507324, "learning_rate": 6.613149847094802e-06, "loss": 0.0058, "step": 1329 }, { "epoch": 0.3388966747356351, "grad_norm": 1.5263407230377197, "learning_rate": 6.610601427115189e-06, "loss": 0.0017, "step": 1330 }, { "epoch": 0.3391514842655115, "grad_norm": 6.7438435554504395, "learning_rate": 6.608053007135577e-06, "loss": 0.0041, "step": 1331 }, { "epoch": 0.33940629379538795, "grad_norm": 7.118472576141357, "learning_rate": 6.6055045871559645e-06, "loss": 0.0051, "step": 1332 }, { "epoch": 0.3396611033252644, "grad_norm": 1.1746944189071655, "learning_rate": 6.602956167176351e-06, "loss": 0.0014, "step": 1333 }, { "epoch": 0.3399159128551408, "grad_norm": 1.459389328956604, "learning_rate": 6.6004077471967384e-06, "loss": 0.002, "step": 1334 }, { "epoch": 0.3401707223850172, "grad_norm": 6.366577625274658, "learning_rate": 6.597859327217125e-06, "loss": 0.005, "step": 1335 }, { "epoch": 0.3404255319148936, "grad_norm": 11.80704402923584, "learning_rate": 6.595310907237513e-06, "loss": 0.0115, "step": 1336 }, { "epoch": 0.34068034144477005, "grad_norm": 5.115415096282959, "learning_rate": 6.592762487257901e-06, "loss": 0.0032, "step": 1337 }, { "epoch": 0.34093515097464644, "grad_norm": 5.613772392272949, "learning_rate": 6.590214067278288e-06, "loss": 0.0046, "step": 1338 }, { "epoch": 0.3411899605045229, "grad_norm": 5.622154235839844, "learning_rate": 6.587665647298676e-06, "loss": 0.0039, "step": 1339 }, { "epoch": 0.34144477003439927, "grad_norm": 2.9062678813934326, "learning_rate": 6.585117227319062e-06, "loss": 0.0018, "step": 1340 }, { "epoch": 0.3416995795642757, "grad_norm": 0.5034934878349304, "learning_rate": 6.58256880733945e-06, "loss": 0.001, "step": 1341 }, { "epoch": 0.3419543890941521, "grad_norm": 0.4245041310787201, "learning_rate": 6.5800203873598374e-06, "loss": 0.0016, "step": 1342 }, { "epoch": 0.34220919862402854, "grad_norm": 4.736161708831787, "learning_rate": 6.577471967380224e-06, "loss": 0.0027, "step": 1343 }, { "epoch": 0.34246400815390493, "grad_norm": 0.943156361579895, "learning_rate": 6.574923547400612e-06, "loss": 0.0024, "step": 1344 }, { "epoch": 0.3427188176837814, "grad_norm": 1.8880647420883179, "learning_rate": 6.572375127420999e-06, "loss": 0.002, "step": 1345 }, { "epoch": 0.3429736272136578, "grad_norm": 0.2624977231025696, "learning_rate": 6.569826707441387e-06, "loss": 0.0007, "step": 1346 }, { "epoch": 0.3432284367435342, "grad_norm": 3.123929023742676, "learning_rate": 6.567278287461775e-06, "loss": 0.0017, "step": 1347 }, { "epoch": 0.34348324627341065, "grad_norm": 0.9520541429519653, "learning_rate": 6.564729867482162e-06, "loss": 0.0008, "step": 1348 }, { "epoch": 0.34373805580328703, "grad_norm": 175.1728515625, "learning_rate": 6.5621814475025495e-06, "loss": 0.2341, "step": 1349 }, { "epoch": 0.3439928653331635, "grad_norm": 2.614830255508423, "learning_rate": 6.559633027522936e-06, "loss": 0.0019, "step": 1350 }, { "epoch": 0.34424767486303987, "grad_norm": 1.9816035032272339, "learning_rate": 6.557084607543323e-06, "loss": 0.0008, "step": 1351 }, { "epoch": 0.3445024843929163, "grad_norm": 4.059197425842285, "learning_rate": 6.554536187563711e-06, "loss": 0.0018, "step": 1352 }, { "epoch": 0.3447572939227927, "grad_norm": 0.4751805067062378, "learning_rate": 6.551987767584098e-06, "loss": 0.0011, "step": 1353 }, { "epoch": 0.34501210345266914, "grad_norm": 2.9791321754455566, "learning_rate": 6.549439347604486e-06, "loss": 0.0038, "step": 1354 }, { "epoch": 0.3452669129825455, "grad_norm": 1.7992115020751953, "learning_rate": 6.546890927624873e-06, "loss": 0.0006, "step": 1355 }, { "epoch": 0.34552172251242197, "grad_norm": 64.29216003417969, "learning_rate": 6.544342507645261e-06, "loss": 0.0132, "step": 1356 }, { "epoch": 0.34577653204229836, "grad_norm": 1.183692455291748, "learning_rate": 6.5417940876656485e-06, "loss": 0.0009, "step": 1357 }, { "epoch": 0.3460313415721748, "grad_norm": 1.3201334476470947, "learning_rate": 6.539245667686035e-06, "loss": 0.0003, "step": 1358 }, { "epoch": 0.34628615110205124, "grad_norm": 0.5777949094772339, "learning_rate": 6.536697247706422e-06, "loss": 0.0004, "step": 1359 }, { "epoch": 0.34654096063192763, "grad_norm": 9.058926582336426, "learning_rate": 6.534148827726809e-06, "loss": 0.0059, "step": 1360 }, { "epoch": 0.3467957701618041, "grad_norm": 3.4707329273223877, "learning_rate": 6.531600407747197e-06, "loss": 0.0012, "step": 1361 }, { "epoch": 0.34705057969168046, "grad_norm": 2.4490363597869873, "learning_rate": 6.529051987767585e-06, "loss": 0.0014, "step": 1362 }, { "epoch": 0.3473053892215569, "grad_norm": 4.8990654945373535, "learning_rate": 6.526503567787972e-06, "loss": 0.002, "step": 1363 }, { "epoch": 0.3475601987514333, "grad_norm": 5.266764163970947, "learning_rate": 6.52395514780836e-06, "loss": 0.0036, "step": 1364 }, { "epoch": 0.34781500828130973, "grad_norm": 0.03600892052054405, "learning_rate": 6.521406727828747e-06, "loss": 0.0002, "step": 1365 }, { "epoch": 0.3480698178111861, "grad_norm": 3.8224265575408936, "learning_rate": 6.5188583078491345e-06, "loss": 0.0044, "step": 1366 }, { "epoch": 0.34832462734106256, "grad_norm": 0.3815406262874603, "learning_rate": 6.516309887869522e-06, "loss": 0.0002, "step": 1367 }, { "epoch": 0.34857943687093895, "grad_norm": 95.94962310791016, "learning_rate": 6.513761467889908e-06, "loss": 0.1545, "step": 1368 }, { "epoch": 0.3488342464008154, "grad_norm": 5.124578475952148, "learning_rate": 6.511213047910296e-06, "loss": 0.0035, "step": 1369 }, { "epoch": 0.3490890559306918, "grad_norm": 70.0281753540039, "learning_rate": 6.508664627930683e-06, "loss": 0.7814, "step": 1370 }, { "epoch": 0.3493438654605682, "grad_norm": 7.2557291984558105, "learning_rate": 6.506116207951071e-06, "loss": 0.0039, "step": 1371 }, { "epoch": 0.34959867499044467, "grad_norm": 2.632417917251587, "learning_rate": 6.503567787971459e-06, "loss": 0.0006, "step": 1372 }, { "epoch": 0.34985348452032106, "grad_norm": 2.7485315799713135, "learning_rate": 6.501019367991846e-06, "loss": 0.0026, "step": 1373 }, { "epoch": 0.3501082940501975, "grad_norm": 1.1426347494125366, "learning_rate": 6.4984709480122335e-06, "loss": 0.0004, "step": 1374 }, { "epoch": 0.3503631035800739, "grad_norm": 2.5017478466033936, "learning_rate": 6.4959225280326196e-06, "loss": 0.0012, "step": 1375 }, { "epoch": 0.35061791310995033, "grad_norm": 113.52030181884766, "learning_rate": 6.493374108053007e-06, "loss": 0.1076, "step": 1376 }, { "epoch": 0.3508727226398267, "grad_norm": 8.256390571594238, "learning_rate": 6.490825688073395e-06, "loss": 0.0051, "step": 1377 }, { "epoch": 0.35112753216970316, "grad_norm": 118.80034637451172, "learning_rate": 6.488277268093782e-06, "loss": 0.7716, "step": 1378 }, { "epoch": 0.35138234169957955, "grad_norm": 2.130624532699585, "learning_rate": 6.48572884811417e-06, "loss": 0.0014, "step": 1379 }, { "epoch": 0.351637151229456, "grad_norm": 2.498082160949707, "learning_rate": 6.483180428134557e-06, "loss": 0.0022, "step": 1380 }, { "epoch": 0.3518919607593324, "grad_norm": 129.10888671875, "learning_rate": 6.480632008154945e-06, "loss": 0.0563, "step": 1381 }, { "epoch": 0.3521467702892088, "grad_norm": 3.694733142852783, "learning_rate": 6.4780835881753325e-06, "loss": 0.0016, "step": 1382 }, { "epoch": 0.3524015798190852, "grad_norm": 8.444147109985352, "learning_rate": 6.4755351681957194e-06, "loss": 0.0057, "step": 1383 }, { "epoch": 0.35265638934896165, "grad_norm": 7.424720287322998, "learning_rate": 6.472986748216107e-06, "loss": 0.0043, "step": 1384 }, { "epoch": 0.3529111988788381, "grad_norm": 7.569698810577393, "learning_rate": 6.470438328236493e-06, "loss": 0.0049, "step": 1385 }, { "epoch": 0.3531660084087145, "grad_norm": 2.9234344959259033, "learning_rate": 6.467889908256881e-06, "loss": 0.0026, "step": 1386 }, { "epoch": 0.3534208179385909, "grad_norm": 3.576078414916992, "learning_rate": 6.465341488277269e-06, "loss": 0.0012, "step": 1387 }, { "epoch": 0.3536756274684673, "grad_norm": 2.917107582092285, "learning_rate": 6.462793068297656e-06, "loss": 0.0009, "step": 1388 }, { "epoch": 0.35393043699834376, "grad_norm": 5.8224358558654785, "learning_rate": 6.460244648318044e-06, "loss": 0.0026, "step": 1389 }, { "epoch": 0.35418524652822014, "grad_norm": 0.9422265887260437, "learning_rate": 6.457696228338431e-06, "loss": 0.0003, "step": 1390 }, { "epoch": 0.3544400560580966, "grad_norm": 315.40509033203125, "learning_rate": 6.4551478083588184e-06, "loss": 3.6418, "step": 1391 }, { "epoch": 0.354694865587973, "grad_norm": 5.769959926605225, "learning_rate": 6.452599388379206e-06, "loss": 0.0038, "step": 1392 }, { "epoch": 0.3549496751178494, "grad_norm": 0.6822844743728638, "learning_rate": 6.450050968399592e-06, "loss": 0.0009, "step": 1393 }, { "epoch": 0.3552044846477258, "grad_norm": 0.5925134420394897, "learning_rate": 6.44750254841998e-06, "loss": 0.0012, "step": 1394 }, { "epoch": 0.35545929417760225, "grad_norm": 1310.8211669921875, "learning_rate": 6.444954128440367e-06, "loss": 2.8911, "step": 1395 }, { "epoch": 0.35571410370747863, "grad_norm": 2.952317237854004, "learning_rate": 6.442405708460755e-06, "loss": 0.0016, "step": 1396 }, { "epoch": 0.3559689132373551, "grad_norm": 0.5608370304107666, "learning_rate": 6.439857288481142e-06, "loss": 0.0007, "step": 1397 }, { "epoch": 0.3562237227672315, "grad_norm": 0.3334255814552307, "learning_rate": 6.43730886850153e-06, "loss": 0.0009, "step": 1398 }, { "epoch": 0.3564785322971079, "grad_norm": 1.8779187202453613, "learning_rate": 6.4347604485219174e-06, "loss": 0.0005, "step": 1399 }, { "epoch": 0.35673334182698435, "grad_norm": 1.6891659498214722, "learning_rate": 6.432212028542304e-06, "loss": 0.0004, "step": 1400 }, { "epoch": 0.35698815135686074, "grad_norm": 2.549051284790039, "learning_rate": 6.429663608562692e-06, "loss": 0.0012, "step": 1401 }, { "epoch": 0.3572429608867372, "grad_norm": 0.6433826684951782, "learning_rate": 6.427115188583078e-06, "loss": 0.0013, "step": 1402 }, { "epoch": 0.35749777041661357, "grad_norm": 3.5840437412261963, "learning_rate": 6.424566768603466e-06, "loss": 0.0024, "step": 1403 }, { "epoch": 0.35775257994649, "grad_norm": 1.1411893367767334, "learning_rate": 6.422018348623854e-06, "loss": 0.0005, "step": 1404 }, { "epoch": 0.3580073894763664, "grad_norm": 2.067009210586548, "learning_rate": 6.419469928644241e-06, "loss": 0.0006, "step": 1405 }, { "epoch": 0.35826219900624284, "grad_norm": 0.3306000530719757, "learning_rate": 6.416921508664629e-06, "loss": 0.0005, "step": 1406 }, { "epoch": 0.35851700853611923, "grad_norm": 0.3826887607574463, "learning_rate": 6.414373088685016e-06, "loss": 0.0002, "step": 1407 }, { "epoch": 0.3587718180659957, "grad_norm": 3.3978707790374756, "learning_rate": 6.411824668705403e-06, "loss": 0.0018, "step": 1408 }, { "epoch": 0.35902662759587206, "grad_norm": 6.081480979919434, "learning_rate": 6.409276248725791e-06, "loss": 0.0029, "step": 1409 }, { "epoch": 0.3592814371257485, "grad_norm": 2.3695597648620605, "learning_rate": 6.406727828746177e-06, "loss": 0.0013, "step": 1410 }, { "epoch": 0.35953624665562495, "grad_norm": 7.061707496643066, "learning_rate": 6.404179408766565e-06, "loss": 0.0056, "step": 1411 }, { "epoch": 0.35979105618550133, "grad_norm": 8.943286895751953, "learning_rate": 6.401630988786952e-06, "loss": 0.0051, "step": 1412 }, { "epoch": 0.3600458657153778, "grad_norm": 106.78902435302734, "learning_rate": 6.39908256880734e-06, "loss": 4.0003, "step": 1413 }, { "epoch": 0.36030067524525417, "grad_norm": 2.826712131500244, "learning_rate": 6.396534148827728e-06, "loss": 0.0017, "step": 1414 }, { "epoch": 0.3605554847751306, "grad_norm": 2.7189247608184814, "learning_rate": 6.393985728848115e-06, "loss": 0.0008, "step": 1415 }, { "epoch": 0.360810294305007, "grad_norm": 154.3860626220703, "learning_rate": 6.391437308868502e-06, "loss": 0.0584, "step": 1416 }, { "epoch": 0.36106510383488344, "grad_norm": 0.5450733304023743, "learning_rate": 6.3888888888888885e-06, "loss": 0.0008, "step": 1417 }, { "epoch": 0.3613199133647598, "grad_norm": 5.486621379852295, "learning_rate": 6.386340468909276e-06, "loss": 0.0032, "step": 1418 }, { "epoch": 0.36157472289463627, "grad_norm": 3.388237953186035, "learning_rate": 6.383792048929664e-06, "loss": 0.0014, "step": 1419 }, { "epoch": 0.36182953242451266, "grad_norm": 47.337589263916016, "learning_rate": 6.381243628950051e-06, "loss": 0.0285, "step": 1420 }, { "epoch": 0.3620843419543891, "grad_norm": 1.9034467935562134, "learning_rate": 6.378695208970439e-06, "loss": 0.001, "step": 1421 }, { "epoch": 0.3623391514842655, "grad_norm": 0.5274309515953064, "learning_rate": 6.376146788990826e-06, "loss": 0.0008, "step": 1422 }, { "epoch": 0.36259396101414193, "grad_norm": 2.1533565521240234, "learning_rate": 6.373598369011214e-06, "loss": 0.0009, "step": 1423 }, { "epoch": 0.3628487705440184, "grad_norm": 3.3705098628997803, "learning_rate": 6.371049949031601e-06, "loss": 0.0016, "step": 1424 }, { "epoch": 0.36310358007389476, "grad_norm": 4.074306011199951, "learning_rate": 6.368501529051988e-06, "loss": 0.0024, "step": 1425 }, { "epoch": 0.3633583896037712, "grad_norm": 2.630434036254883, "learning_rate": 6.365953109072376e-06, "loss": 0.0012, "step": 1426 }, { "epoch": 0.3636131991336476, "grad_norm": 0.4903945326805115, "learning_rate": 6.363404689092762e-06, "loss": 0.0005, "step": 1427 }, { "epoch": 0.36386800866352403, "grad_norm": 1.4420385360717773, "learning_rate": 6.36085626911315e-06, "loss": 0.0012, "step": 1428 }, { "epoch": 0.3641228181934004, "grad_norm": 0.39493754506111145, "learning_rate": 6.358307849133538e-06, "loss": 0.0009, "step": 1429 }, { "epoch": 0.36437762772327686, "grad_norm": 20.484872817993164, "learning_rate": 6.355759429153925e-06, "loss": 0.0103, "step": 1430 }, { "epoch": 0.36463243725315325, "grad_norm": 49.28974914550781, "learning_rate": 6.353211009174313e-06, "loss": 0.4999, "step": 1431 }, { "epoch": 0.3648872467830297, "grad_norm": 4.563051223754883, "learning_rate": 6.3506625891947e-06, "loss": 0.0024, "step": 1432 }, { "epoch": 0.3651420563129061, "grad_norm": 6.480624675750732, "learning_rate": 6.348114169215087e-06, "loss": 0.0035, "step": 1433 }, { "epoch": 0.3653968658427825, "grad_norm": 1.9946320056915283, "learning_rate": 6.345565749235475e-06, "loss": 0.0009, "step": 1434 }, { "epoch": 0.3656516753726589, "grad_norm": 6.772976398468018, "learning_rate": 6.343017329255861e-06, "loss": 0.0067, "step": 1435 }, { "epoch": 0.36590648490253536, "grad_norm": 203.11126708984375, "learning_rate": 6.340468909276249e-06, "loss": 0.1332, "step": 1436 }, { "epoch": 0.3661612944324118, "grad_norm": 0.580550491809845, "learning_rate": 6.337920489296636e-06, "loss": 0.0006, "step": 1437 }, { "epoch": 0.3664161039622882, "grad_norm": 1.1428468227386475, "learning_rate": 6.335372069317024e-06, "loss": 0.0005, "step": 1438 }, { "epoch": 0.36667091349216463, "grad_norm": 1.446468710899353, "learning_rate": 6.332823649337412e-06, "loss": 0.0011, "step": 1439 }, { "epoch": 0.366925723022041, "grad_norm": 5.635172367095947, "learning_rate": 6.330275229357799e-06, "loss": 0.0031, "step": 1440 }, { "epoch": 0.36718053255191746, "grad_norm": 0.6593403816223145, "learning_rate": 6.327726809378186e-06, "loss": 0.0005, "step": 1441 }, { "epoch": 0.36743534208179385, "grad_norm": 7.689243316650391, "learning_rate": 6.325178389398573e-06, "loss": 0.0089, "step": 1442 }, { "epoch": 0.3676901516116703, "grad_norm": 3.3157641887664795, "learning_rate": 6.322629969418961e-06, "loss": 0.0014, "step": 1443 }, { "epoch": 0.3679449611415467, "grad_norm": 0.7979831695556641, "learning_rate": 6.320081549439349e-06, "loss": 0.0008, "step": 1444 }, { "epoch": 0.3681997706714231, "grad_norm": 5.837172031402588, "learning_rate": 6.317533129459735e-06, "loss": 0.003, "step": 1445 }, { "epoch": 0.3684545802012995, "grad_norm": 0.72694993019104, "learning_rate": 6.314984709480123e-06, "loss": 0.0008, "step": 1446 }, { "epoch": 0.36870938973117595, "grad_norm": 5.834043502807617, "learning_rate": 6.31243628950051e-06, "loss": 0.0032, "step": 1447 }, { "epoch": 0.36896419926105234, "grad_norm": 2.4200026988983154, "learning_rate": 6.309887869520898e-06, "loss": 0.0042, "step": 1448 }, { "epoch": 0.3692190087909288, "grad_norm": 2.5321481227874756, "learning_rate": 6.307339449541285e-06, "loss": 0.0032, "step": 1449 }, { "epoch": 0.3694738183208052, "grad_norm": 2.3165411949157715, "learning_rate": 6.304791029561672e-06, "loss": 0.0034, "step": 1450 }, { "epoch": 0.3697286278506816, "grad_norm": 9.704877853393555, "learning_rate": 6.30224260958206e-06, "loss": 0.007, "step": 1451 }, { "epoch": 0.36998343738055806, "grad_norm": 6.208980560302734, "learning_rate": 6.299694189602446e-06, "loss": 0.0033, "step": 1452 }, { "epoch": 0.37023824691043444, "grad_norm": 4.443699359893799, "learning_rate": 6.297145769622834e-06, "loss": 0.0023, "step": 1453 }, { "epoch": 0.3704930564403109, "grad_norm": 5.294854640960693, "learning_rate": 6.294597349643222e-06, "loss": 0.0023, "step": 1454 }, { "epoch": 0.3707478659701873, "grad_norm": 76.1446762084961, "learning_rate": 6.292048929663609e-06, "loss": 0.4129, "step": 1455 }, { "epoch": 0.3710026755000637, "grad_norm": 3.9307403564453125, "learning_rate": 6.289500509683997e-06, "loss": 0.0015, "step": 1456 }, { "epoch": 0.3712574850299401, "grad_norm": 1.4054863452911377, "learning_rate": 6.2869520897043836e-06, "loss": 0.0005, "step": 1457 }, { "epoch": 0.37151229455981655, "grad_norm": 6.522034645080566, "learning_rate": 6.284403669724771e-06, "loss": 0.004, "step": 1458 }, { "epoch": 0.37176710408969293, "grad_norm": 3.2216415405273438, "learning_rate": 6.281855249745159e-06, "loss": 0.0015, "step": 1459 }, { "epoch": 0.3720219136195694, "grad_norm": 4.507644176483154, "learning_rate": 6.279306829765546e-06, "loss": 0.0015, "step": 1460 }, { "epoch": 0.37227672314944577, "grad_norm": 2.736933708190918, "learning_rate": 6.276758409785934e-06, "loss": 0.0009, "step": 1461 }, { "epoch": 0.3725315326793222, "grad_norm": 0.5517087578773499, "learning_rate": 6.27420998980632e-06, "loss": 0.002, "step": 1462 }, { "epoch": 0.37278634220919865, "grad_norm": 57.965301513671875, "learning_rate": 6.271661569826708e-06, "loss": 0.6809, "step": 1463 }, { "epoch": 0.37304115173907504, "grad_norm": 6.339090347290039, "learning_rate": 6.269113149847096e-06, "loss": 0.0039, "step": 1464 }, { "epoch": 0.3732959612689515, "grad_norm": 331.8721618652344, "learning_rate": 6.2665647298674826e-06, "loss": 0.2778, "step": 1465 }, { "epoch": 0.37355077079882787, "grad_norm": 9.494669914245605, "learning_rate": 6.26401630988787e-06, "loss": 0.0073, "step": 1466 }, { "epoch": 0.3738055803287043, "grad_norm": 8.44284725189209, "learning_rate": 6.261467889908257e-06, "loss": 0.0055, "step": 1467 }, { "epoch": 0.3740603898585807, "grad_norm": 5.271597862243652, "learning_rate": 6.258919469928645e-06, "loss": 0.0034, "step": 1468 }, { "epoch": 0.37431519938845714, "grad_norm": 3.7557594776153564, "learning_rate": 6.256371049949033e-06, "loss": 0.0021, "step": 1469 }, { "epoch": 0.37457000891833353, "grad_norm": 188.57125854492188, "learning_rate": 6.253822629969419e-06, "loss": 0.1774, "step": 1470 }, { "epoch": 0.37482481844821, "grad_norm": 5.525568962097168, "learning_rate": 6.251274209989807e-06, "loss": 0.003, "step": 1471 }, { "epoch": 0.37507962797808636, "grad_norm": 11.939030647277832, "learning_rate": 6.248725790010194e-06, "loss": 0.0104, "step": 1472 }, { "epoch": 0.3753344375079628, "grad_norm": 9.852555274963379, "learning_rate": 6.2461773700305816e-06, "loss": 0.0075, "step": 1473 }, { "epoch": 0.3755892470378392, "grad_norm": 11.611922264099121, "learning_rate": 6.2436289500509685e-06, "loss": 0.0099, "step": 1474 }, { "epoch": 0.37584405656771563, "grad_norm": 2.955352306365967, "learning_rate": 6.241080530071356e-06, "loss": 0.0014, "step": 1475 }, { "epoch": 0.3760988660975921, "grad_norm": 5.100165367126465, "learning_rate": 6.238532110091744e-06, "loss": 0.0023, "step": 1476 }, { "epoch": 0.37635367562746846, "grad_norm": 3.786611795425415, "learning_rate": 6.235983690112131e-06, "loss": 0.0025, "step": 1477 }, { "epoch": 0.3766084851573449, "grad_norm": 2.356764554977417, "learning_rate": 6.233435270132519e-06, "loss": 0.0013, "step": 1478 }, { "epoch": 0.3768632946872213, "grad_norm": 279.2536926269531, "learning_rate": 6.230886850152905e-06, "loss": 0.1697, "step": 1479 }, { "epoch": 0.37711810421709774, "grad_norm": 8.042028427124023, "learning_rate": 6.228338430173293e-06, "loss": 0.0049, "step": 1480 }, { "epoch": 0.3773729137469741, "grad_norm": 12.800419807434082, "learning_rate": 6.225790010193681e-06, "loss": 0.0122, "step": 1481 }, { "epoch": 0.37762772327685057, "grad_norm": 10.365827560424805, "learning_rate": 6.2232415902140675e-06, "loss": 0.0084, "step": 1482 }, { "epoch": 0.37788253280672696, "grad_norm": 6.2335381507873535, "learning_rate": 6.220693170234455e-06, "loss": 0.0033, "step": 1483 }, { "epoch": 0.3781373423366034, "grad_norm": 5.1683783531188965, "learning_rate": 6.218144750254842e-06, "loss": 0.0025, "step": 1484 }, { "epoch": 0.3783921518664798, "grad_norm": 4.630924224853516, "learning_rate": 6.21559633027523e-06, "loss": 0.0024, "step": 1485 }, { "epoch": 0.37864696139635623, "grad_norm": 0.5069142580032349, "learning_rate": 6.213047910295618e-06, "loss": 0.0006, "step": 1486 }, { "epoch": 0.3789017709262326, "grad_norm": 6.288352966308594, "learning_rate": 6.210499490316004e-06, "loss": 0.0033, "step": 1487 }, { "epoch": 0.37915658045610906, "grad_norm": 282.84429931640625, "learning_rate": 6.207951070336392e-06, "loss": 0.4433, "step": 1488 }, { "epoch": 0.3794113899859855, "grad_norm": 9.350077629089355, "learning_rate": 6.205402650356779e-06, "loss": 0.0065, "step": 1489 }, { "epoch": 0.3796661995158619, "grad_norm": 10.535582542419434, "learning_rate": 6.2028542303771665e-06, "loss": 0.0098, "step": 1490 }, { "epoch": 0.37992100904573833, "grad_norm": 8.739837646484375, "learning_rate": 6.200305810397554e-06, "loss": 0.0057, "step": 1491 }, { "epoch": 0.3801758185756147, "grad_norm": 352.43463134765625, "learning_rate": 6.197757390417941e-06, "loss": 0.8036, "step": 1492 }, { "epoch": 0.38043062810549116, "grad_norm": 9.284416198730469, "learning_rate": 6.195208970438329e-06, "loss": 0.0067, "step": 1493 }, { "epoch": 0.38068543763536755, "grad_norm": 7.966891288757324, "learning_rate": 6.192660550458715e-06, "loss": 0.0051, "step": 1494 }, { "epoch": 0.380940247165244, "grad_norm": 2.641827344894409, "learning_rate": 6.190112130479103e-06, "loss": 0.0014, "step": 1495 }, { "epoch": 0.3811950566951204, "grad_norm": 2.633613348007202, "learning_rate": 6.187563710499492e-06, "loss": 0.0009, "step": 1496 }, { "epoch": 0.3814498662249968, "grad_norm": 5.352120399475098, "learning_rate": 6.185015290519878e-06, "loss": 0.0027, "step": 1497 }, { "epoch": 0.3817046757548732, "grad_norm": 7.233755588531494, "learning_rate": 6.1824668705402656e-06, "loss": 0.0048, "step": 1498 }, { "epoch": 0.38195948528474966, "grad_norm": 5.617506980895996, "learning_rate": 6.1799184505606525e-06, "loss": 0.0025, "step": 1499 }, { "epoch": 0.38221429481462604, "grad_norm": 3.5764000415802, "learning_rate": 6.17737003058104e-06, "loss": 0.0019, "step": 1500 }, { "epoch": 0.3824691043445025, "grad_norm": 3.848297595977783, "learning_rate": 6.174821610601428e-06, "loss": 0.0013, "step": 1501 }, { "epoch": 0.38272391387437893, "grad_norm": 3.970505714416504, "learning_rate": 6.172273190621815e-06, "loss": 0.0022, "step": 1502 }, { "epoch": 0.3829787234042553, "grad_norm": 5.108886241912842, "learning_rate": 6.169724770642203e-06, "loss": 0.0029, "step": 1503 }, { "epoch": 0.38323353293413176, "grad_norm": 4.039179801940918, "learning_rate": 6.167176350662589e-06, "loss": 0.0025, "step": 1504 }, { "epoch": 0.38348834246400815, "grad_norm": 7.1075239181518555, "learning_rate": 6.164627930682977e-06, "loss": 0.0039, "step": 1505 }, { "epoch": 0.3837431519938846, "grad_norm": 46.554039001464844, "learning_rate": 6.1620795107033646e-06, "loss": 0.5347, "step": 1506 }, { "epoch": 0.383997961523761, "grad_norm": 10.657620429992676, "learning_rate": 6.1595310907237515e-06, "loss": 0.0084, "step": 1507 }, { "epoch": 0.3842527710536374, "grad_norm": 8.945984840393066, "learning_rate": 6.156982670744139e-06, "loss": 0.0059, "step": 1508 }, { "epoch": 0.3845075805835138, "grad_norm": 5.71747350692749, "learning_rate": 6.154434250764526e-06, "loss": 0.0027, "step": 1509 }, { "epoch": 0.38476239011339025, "grad_norm": 3.9018070697784424, "learning_rate": 6.151885830784914e-06, "loss": 0.0018, "step": 1510 }, { "epoch": 0.38501719964326664, "grad_norm": 1.1895880699157715, "learning_rate": 6.149337410805302e-06, "loss": 0.0012, "step": 1511 }, { "epoch": 0.3852720091731431, "grad_norm": 1.6822720766067505, "learning_rate": 6.146788990825688e-06, "loss": 0.0006, "step": 1512 }, { "epoch": 0.38552681870301947, "grad_norm": 3.013282537460327, "learning_rate": 6.144240570846076e-06, "loss": 0.001, "step": 1513 }, { "epoch": 0.3857816282328959, "grad_norm": 3.983815908432007, "learning_rate": 6.141692150866463e-06, "loss": 0.0023, "step": 1514 }, { "epoch": 0.38603643776277236, "grad_norm": 6.54759407043457, "learning_rate": 6.1391437308868505e-06, "loss": 0.0035, "step": 1515 }, { "epoch": 0.38629124729264874, "grad_norm": 4.830348968505859, "learning_rate": 6.136595310907238e-06, "loss": 0.0026, "step": 1516 }, { "epoch": 0.3865460568225252, "grad_norm": 9.02996826171875, "learning_rate": 6.134046890927625e-06, "loss": 0.0213, "step": 1517 }, { "epoch": 0.3868008663524016, "grad_norm": 1.5277031660079956, "learning_rate": 6.131498470948013e-06, "loss": 0.0024, "step": 1518 }, { "epoch": 0.387055675882278, "grad_norm": 1.5929316282272339, "learning_rate": 6.1289500509684e-06, "loss": 0.0005, "step": 1519 }, { "epoch": 0.3873104854121544, "grad_norm": 5.366246223449707, "learning_rate": 6.126401630988788e-06, "loss": 0.0024, "step": 1520 }, { "epoch": 0.38756529494203085, "grad_norm": 8.068830490112305, "learning_rate": 6.123853211009176e-06, "loss": 0.0048, "step": 1521 }, { "epoch": 0.38782010447190723, "grad_norm": 1.4970139265060425, "learning_rate": 6.121304791029562e-06, "loss": 0.0006, "step": 1522 }, { "epoch": 0.3880749140017837, "grad_norm": 1.1275941133499146, "learning_rate": 6.1187563710499495e-06, "loss": 0.0004, "step": 1523 }, { "epoch": 0.38832972353166006, "grad_norm": 0.43051832914352417, "learning_rate": 6.1162079510703365e-06, "loss": 0.0005, "step": 1524 }, { "epoch": 0.3885845330615365, "grad_norm": 1.1404398679733276, "learning_rate": 6.113659531090724e-06, "loss": 0.0004, "step": 1525 }, { "epoch": 0.3888393425914129, "grad_norm": 3.3642525672912598, "learning_rate": 6.111111111111112e-06, "loss": 0.0012, "step": 1526 }, { "epoch": 0.38909415212128934, "grad_norm": 0.22504213452339172, "learning_rate": 6.108562691131499e-06, "loss": 0.0004, "step": 1527 }, { "epoch": 0.3893489616511657, "grad_norm": 2.1543710231781006, "learning_rate": 6.106014271151887e-06, "loss": 0.0008, "step": 1528 }, { "epoch": 0.38960377118104217, "grad_norm": 1.4942249059677124, "learning_rate": 6.103465851172273e-06, "loss": 0.0004, "step": 1529 }, { "epoch": 0.3898585807109186, "grad_norm": 89.51766204833984, "learning_rate": 6.100917431192661e-06, "loss": 0.0366, "step": 1530 }, { "epoch": 0.390113390240795, "grad_norm": 0.19306407868862152, "learning_rate": 6.0983690112130485e-06, "loss": 0.0003, "step": 1531 }, { "epoch": 0.39036819977067144, "grad_norm": 6.96235990524292, "learning_rate": 6.0958205912334355e-06, "loss": 0.005, "step": 1532 }, { "epoch": 0.39062300930054783, "grad_norm": 43.011085510253906, "learning_rate": 6.093272171253823e-06, "loss": 0.0475, "step": 1533 }, { "epoch": 0.3908778188304243, "grad_norm": 6.903972148895264, "learning_rate": 6.09072375127421e-06, "loss": 0.0038, "step": 1534 }, { "epoch": 0.39113262836030066, "grad_norm": 6.92374849319458, "learning_rate": 6.088175331294598e-06, "loss": 0.0038, "step": 1535 }, { "epoch": 0.3913874378901771, "grad_norm": 5.36497163772583, "learning_rate": 6.085626911314986e-06, "loss": 0.0024, "step": 1536 }, { "epoch": 0.3916422474200535, "grad_norm": 2.9879469871520996, "learning_rate": 6.083078491335373e-06, "loss": 0.0009, "step": 1537 }, { "epoch": 0.39189705694992993, "grad_norm": 2.060433864593506, "learning_rate": 6.080530071355761e-06, "loss": 0.0005, "step": 1538 }, { "epoch": 0.3921518664798063, "grad_norm": 0.8229150176048279, "learning_rate": 6.077981651376147e-06, "loss": 0.0005, "step": 1539 }, { "epoch": 0.39240667600968276, "grad_norm": 1.0453945398330688, "learning_rate": 6.0754332313965345e-06, "loss": 0.0014, "step": 1540 }, { "epoch": 0.39266148553955915, "grad_norm": 5.975008964538574, "learning_rate": 6.072884811416922e-06, "loss": 0.0027, "step": 1541 }, { "epoch": 0.3929162950694356, "grad_norm": 5.43073844909668, "learning_rate": 6.070336391437309e-06, "loss": 0.0024, "step": 1542 }, { "epoch": 0.39317110459931204, "grad_norm": 57.455318450927734, "learning_rate": 6.067787971457697e-06, "loss": 0.5158, "step": 1543 }, { "epoch": 0.3934259141291884, "grad_norm": 1.1057268381118774, "learning_rate": 6.065239551478084e-06, "loss": 0.0013, "step": 1544 }, { "epoch": 0.39368072365906487, "grad_norm": 0.5224020481109619, "learning_rate": 6.062691131498472e-06, "loss": 0.0015, "step": 1545 }, { "epoch": 0.39393553318894126, "grad_norm": 2.4821224212646484, "learning_rate": 6.06014271151886e-06, "loss": 0.0015, "step": 1546 }, { "epoch": 0.3941903427188177, "grad_norm": 5.4434614181518555, "learning_rate": 6.057594291539246e-06, "loss": 0.0023, "step": 1547 }, { "epoch": 0.3944451522486941, "grad_norm": 3.838088274002075, "learning_rate": 6.0550458715596335e-06, "loss": 0.0026, "step": 1548 }, { "epoch": 0.39469996177857053, "grad_norm": 562.3795776367188, "learning_rate": 6.0524974515800205e-06, "loss": 0.1555, "step": 1549 }, { "epoch": 0.3949547713084469, "grad_norm": 1.7236045598983765, "learning_rate": 6.049949031600408e-06, "loss": 0.0011, "step": 1550 }, { "epoch": 0.39520958083832336, "grad_norm": 2.993422269821167, "learning_rate": 6.047400611620795e-06, "loss": 0.0009, "step": 1551 }, { "epoch": 0.39546439036819975, "grad_norm": 0.8216015696525574, "learning_rate": 6.044852191641183e-06, "loss": 0.0007, "step": 1552 }, { "epoch": 0.3957191998980762, "grad_norm": 4.915148735046387, "learning_rate": 6.042303771661571e-06, "loss": 0.0041, "step": 1553 }, { "epoch": 0.3959740094279526, "grad_norm": 2.9007062911987305, "learning_rate": 6.039755351681958e-06, "loss": 0.0009, "step": 1554 }, { "epoch": 0.396228818957829, "grad_norm": 0.46577954292297363, "learning_rate": 6.0372069317023456e-06, "loss": 0.0006, "step": 1555 }, { "epoch": 0.39648362848770546, "grad_norm": 3.389167070388794, "learning_rate": 6.034658511722732e-06, "loss": 0.0014, "step": 1556 }, { "epoch": 0.39673843801758185, "grad_norm": 224.58116149902344, "learning_rate": 6.0321100917431195e-06, "loss": 0.0519, "step": 1557 }, { "epoch": 0.3969932475474583, "grad_norm": 321.6665954589844, "learning_rate": 6.029561671763507e-06, "loss": 0.2912, "step": 1558 }, { "epoch": 0.3972480570773347, "grad_norm": 4.404673099517822, "learning_rate": 6.027013251783894e-06, "loss": 0.0027, "step": 1559 }, { "epoch": 0.3975028666072111, "grad_norm": 2.984898805618286, "learning_rate": 6.024464831804282e-06, "loss": 0.001, "step": 1560 }, { "epoch": 0.3977576761370875, "grad_norm": 20.21275520324707, "learning_rate": 6.021916411824669e-06, "loss": 0.0042, "step": 1561 }, { "epoch": 0.39801248566696396, "grad_norm": 3.394542694091797, "learning_rate": 6.019367991845057e-06, "loss": 0.0019, "step": 1562 }, { "epoch": 0.39826729519684034, "grad_norm": 1.6079838275909424, "learning_rate": 6.0168195718654446e-06, "loss": 0.0008, "step": 1563 }, { "epoch": 0.3985221047267168, "grad_norm": 4.973844051361084, "learning_rate": 6.014271151885831e-06, "loss": 0.002, "step": 1564 }, { "epoch": 0.3987769142565932, "grad_norm": 4.057135581970215, "learning_rate": 6.0117227319062185e-06, "loss": 0.0014, "step": 1565 }, { "epoch": 0.3990317237864696, "grad_norm": 9.903620719909668, "learning_rate": 6.0091743119266054e-06, "loss": 0.0085, "step": 1566 }, { "epoch": 0.399286533316346, "grad_norm": 2.9299538135528564, "learning_rate": 6.006625891946993e-06, "loss": 0.0011, "step": 1567 }, { "epoch": 0.39954134284622245, "grad_norm": 8.518317222595215, "learning_rate": 6.004077471967381e-06, "loss": 0.002, "step": 1568 }, { "epoch": 0.3997961523760989, "grad_norm": 2.2437455654144287, "learning_rate": 6.001529051987768e-06, "loss": 0.0006, "step": 1569 }, { "epoch": 0.4000509619059753, "grad_norm": 850.445556640625, "learning_rate": 5.998980632008156e-06, "loss": 1.5328, "step": 1570 }, { "epoch": 0.4003057714358517, "grad_norm": 48.72916793823242, "learning_rate": 5.996432212028543e-06, "loss": 0.0223, "step": 1571 }, { "epoch": 0.4005605809657281, "grad_norm": 0.319902241230011, "learning_rate": 5.9938837920489305e-06, "loss": 0.0003, "step": 1572 }, { "epoch": 0.40081539049560455, "grad_norm": 1.6748673915863037, "learning_rate": 5.991335372069318e-06, "loss": 0.0015, "step": 1573 }, { "epoch": 0.40107020002548094, "grad_norm": 1043.97705078125, "learning_rate": 5.9887869520897044e-06, "loss": 2.2192, "step": 1574 }, { "epoch": 0.4013250095553574, "grad_norm": 0.8223576545715332, "learning_rate": 5.986238532110092e-06, "loss": 0.0007, "step": 1575 }, { "epoch": 0.40157981908523377, "grad_norm": 0.9472077488899231, "learning_rate": 5.983690112130479e-06, "loss": 0.0022, "step": 1576 }, { "epoch": 0.4018346286151102, "grad_norm": 2.428051471710205, "learning_rate": 5.981141692150867e-06, "loss": 0.0026, "step": 1577 }, { "epoch": 0.4020894381449866, "grad_norm": 3.015967845916748, "learning_rate": 5.978593272171255e-06, "loss": 0.0023, "step": 1578 }, { "epoch": 0.40234424767486304, "grad_norm": 3.742976665496826, "learning_rate": 5.976044852191642e-06, "loss": 0.0011, "step": 1579 }, { "epoch": 0.40259905720473943, "grad_norm": 1.6357455253601074, "learning_rate": 5.9734964322120295e-06, "loss": 0.0011, "step": 1580 }, { "epoch": 0.4028538667346159, "grad_norm": 2.626582622528076, "learning_rate": 5.970948012232416e-06, "loss": 0.0007, "step": 1581 }, { "epoch": 0.4031086762644923, "grad_norm": 2.9174370765686035, "learning_rate": 5.9683995922528034e-06, "loss": 0.0009, "step": 1582 }, { "epoch": 0.4033634857943687, "grad_norm": 3.373260498046875, "learning_rate": 5.965851172273191e-06, "loss": 0.001, "step": 1583 }, { "epoch": 0.40361829532424515, "grad_norm": 1.3197414875030518, "learning_rate": 5.963302752293578e-06, "loss": 0.0004, "step": 1584 }, { "epoch": 0.40387310485412153, "grad_norm": 1.9422587156295776, "learning_rate": 5.960754332313966e-06, "loss": 0.0009, "step": 1585 }, { "epoch": 0.404127914383998, "grad_norm": 9.742303848266602, "learning_rate": 5.958205912334353e-06, "loss": 0.0026, "step": 1586 }, { "epoch": 0.40438272391387436, "grad_norm": 6.592748641967773, "learning_rate": 5.955657492354741e-06, "loss": 0.0031, "step": 1587 }, { "epoch": 0.4046375334437508, "grad_norm": 6.932510852813721, "learning_rate": 5.9531090723751285e-06, "loss": 0.0034, "step": 1588 }, { "epoch": 0.4048923429736272, "grad_norm": 194.2269287109375, "learning_rate": 5.950560652395515e-06, "loss": 0.6218, "step": 1589 }, { "epoch": 0.40514715250350364, "grad_norm": 12.15743350982666, "learning_rate": 5.9480122324159025e-06, "loss": 0.0043, "step": 1590 }, { "epoch": 0.40540196203338, "grad_norm": 2.298551321029663, "learning_rate": 5.945463812436289e-06, "loss": 0.0014, "step": 1591 }, { "epoch": 0.40565677156325647, "grad_norm": 3.9154117107391357, "learning_rate": 5.942915392456677e-06, "loss": 0.0013, "step": 1592 }, { "epoch": 0.40591158109313286, "grad_norm": 0.2658975124359131, "learning_rate": 5.940366972477065e-06, "loss": 0.0004, "step": 1593 }, { "epoch": 0.4061663906230093, "grad_norm": 1.140323281288147, "learning_rate": 5.937818552497452e-06, "loss": 0.0004, "step": 1594 }, { "epoch": 0.40642120015288574, "grad_norm": 3.8882083892822266, "learning_rate": 5.93527013251784e-06, "loss": 0.0014, "step": 1595 }, { "epoch": 0.40667600968276213, "grad_norm": 4.456520080566406, "learning_rate": 5.932721712538227e-06, "loss": 0.0016, "step": 1596 }, { "epoch": 0.4069308192126386, "grad_norm": 6.559423446655273, "learning_rate": 5.9301732925586145e-06, "loss": 0.0032, "step": 1597 }, { "epoch": 0.40718562874251496, "grad_norm": 1.6971416473388672, "learning_rate": 5.927624872579002e-06, "loss": 0.0004, "step": 1598 }, { "epoch": 0.4074404382723914, "grad_norm": 2.626420021057129, "learning_rate": 5.925076452599388e-06, "loss": 0.0009, "step": 1599 }, { "epoch": 0.4076952478022678, "grad_norm": 0.7463902831077576, "learning_rate": 5.922528032619776e-06, "loss": 0.0002, "step": 1600 }, { "epoch": 0.40795005733214423, "grad_norm": 3.5488929748535156, "learning_rate": 5.919979612640163e-06, "loss": 0.0014, "step": 1601 }, { "epoch": 0.4082048668620206, "grad_norm": 0.9367160201072693, "learning_rate": 5.917431192660551e-06, "loss": 0.0002, "step": 1602 }, { "epoch": 0.40845967639189706, "grad_norm": 6.991890907287598, "learning_rate": 5.914882772680939e-06, "loss": 0.0057, "step": 1603 }, { "epoch": 0.40871448592177345, "grad_norm": 2.8258309364318848, "learning_rate": 5.912334352701326e-06, "loss": 0.0011, "step": 1604 }, { "epoch": 0.4089692954516499, "grad_norm": 4.113875865936279, "learning_rate": 5.9097859327217135e-06, "loss": 0.0015, "step": 1605 }, { "epoch": 0.4092241049815263, "grad_norm": 2.655686140060425, "learning_rate": 5.9072375127421e-06, "loss": 0.0011, "step": 1606 }, { "epoch": 0.4094789145114027, "grad_norm": 1.7039399147033691, "learning_rate": 5.904689092762487e-06, "loss": 0.0009, "step": 1607 }, { "epoch": 0.40973372404127917, "grad_norm": 3.7667171955108643, "learning_rate": 5.902140672782875e-06, "loss": 0.0019, "step": 1608 }, { "epoch": 0.40998853357115556, "grad_norm": 62.540470123291016, "learning_rate": 5.899592252803262e-06, "loss": 0.6572, "step": 1609 }, { "epoch": 0.410243343101032, "grad_norm": 1.3685004711151123, "learning_rate": 5.89704383282365e-06, "loss": 0.0013, "step": 1610 }, { "epoch": 0.4104981526309084, "grad_norm": 0.7706711888313293, "learning_rate": 5.894495412844037e-06, "loss": 0.0003, "step": 1611 }, { "epoch": 0.41075296216078483, "grad_norm": 2.8133442401885986, "learning_rate": 5.891946992864425e-06, "loss": 0.0007, "step": 1612 }, { "epoch": 0.4110077716906612, "grad_norm": 2.4828526973724365, "learning_rate": 5.8893985728848125e-06, "loss": 0.0011, "step": 1613 }, { "epoch": 0.41126258122053766, "grad_norm": 3.2332239151000977, "learning_rate": 5.8868501529051995e-06, "loss": 0.0022, "step": 1614 }, { "epoch": 0.41151739075041405, "grad_norm": 3.5658700466156006, "learning_rate": 5.884301732925587e-06, "loss": 0.0011, "step": 1615 }, { "epoch": 0.4117722002802905, "grad_norm": 0.3731498420238495, "learning_rate": 5.881753312945973e-06, "loss": 0.0002, "step": 1616 }, { "epoch": 0.4120270098101669, "grad_norm": 1.1273419857025146, "learning_rate": 5.879204892966361e-06, "loss": 0.0002, "step": 1617 }, { "epoch": 0.4122818193400433, "grad_norm": 2.281574249267578, "learning_rate": 5.876656472986749e-06, "loss": 0.0008, "step": 1618 }, { "epoch": 0.4125366288699197, "grad_norm": 2.812150001525879, "learning_rate": 5.874108053007136e-06, "loss": 0.0007, "step": 1619 }, { "epoch": 0.41279143839979615, "grad_norm": 0.9307936429977417, "learning_rate": 5.871559633027524e-06, "loss": 0.0003, "step": 1620 }, { "epoch": 0.4130462479296726, "grad_norm": 1.1686687469482422, "learning_rate": 5.869011213047911e-06, "loss": 0.0009, "step": 1621 }, { "epoch": 0.413301057459549, "grad_norm": 1.5054919719696045, "learning_rate": 5.8664627930682985e-06, "loss": 0.0004, "step": 1622 }, { "epoch": 0.4135558669894254, "grad_norm": 0.7940207123756409, "learning_rate": 5.863914373088685e-06, "loss": 0.0005, "step": 1623 }, { "epoch": 0.4138106765193018, "grad_norm": 1.504180908203125, "learning_rate": 5.861365953109072e-06, "loss": 0.0003, "step": 1624 }, { "epoch": 0.41406548604917826, "grad_norm": 3.18554949760437, "learning_rate": 5.85881753312946e-06, "loss": 0.0009, "step": 1625 }, { "epoch": 0.41432029557905464, "grad_norm": 2.4228086471557617, "learning_rate": 5.856269113149847e-06, "loss": 0.0006, "step": 1626 }, { "epoch": 0.4145751051089311, "grad_norm": 0.9356958866119385, "learning_rate": 5.853720693170235e-06, "loss": 0.0002, "step": 1627 }, { "epoch": 0.4148299146388075, "grad_norm": 0.7593979239463806, "learning_rate": 5.851172273190622e-06, "loss": 0.0004, "step": 1628 }, { "epoch": 0.4150847241686839, "grad_norm": 2.7891182899475098, "learning_rate": 5.84862385321101e-06, "loss": 0.0008, "step": 1629 }, { "epoch": 0.4153395336985603, "grad_norm": 1.977418065071106, "learning_rate": 5.8460754332313975e-06, "loss": 0.0009, "step": 1630 }, { "epoch": 0.41559434322843675, "grad_norm": 0.6707476377487183, "learning_rate": 5.8435270132517844e-06, "loss": 0.0007, "step": 1631 }, { "epoch": 0.41584915275831313, "grad_norm": 1.8709304332733154, "learning_rate": 5.840978593272172e-06, "loss": 0.0004, "step": 1632 }, { "epoch": 0.4161039622881896, "grad_norm": 4.117403030395508, "learning_rate": 5.838430173292558e-06, "loss": 0.0014, "step": 1633 }, { "epoch": 0.416358771818066, "grad_norm": 5.972495079040527, "learning_rate": 5.835881753312946e-06, "loss": 0.0027, "step": 1634 }, { "epoch": 0.4166135813479424, "grad_norm": 0.46174487471580505, "learning_rate": 5.833333333333334e-06, "loss": 0.0003, "step": 1635 }, { "epoch": 0.41686839087781885, "grad_norm": 1.710745930671692, "learning_rate": 5.830784913353721e-06, "loss": 0.0005, "step": 1636 }, { "epoch": 0.41712320040769524, "grad_norm": 0.45959383249282837, "learning_rate": 5.828236493374109e-06, "loss": 0.0006, "step": 1637 }, { "epoch": 0.4173780099375717, "grad_norm": 0.43235912919044495, "learning_rate": 5.825688073394496e-06, "loss": 0.0007, "step": 1638 }, { "epoch": 0.41763281946744807, "grad_norm": 3.9414010047912598, "learning_rate": 5.8231396534148835e-06, "loss": 0.0012, "step": 1639 }, { "epoch": 0.4178876289973245, "grad_norm": 1.6928209066390991, "learning_rate": 5.820591233435271e-06, "loss": 0.0004, "step": 1640 }, { "epoch": 0.4181424385272009, "grad_norm": 1.1387280225753784, "learning_rate": 5.818042813455657e-06, "loss": 0.0005, "step": 1641 }, { "epoch": 0.41839724805707734, "grad_norm": 0.8424913287162781, "learning_rate": 5.815494393476045e-06, "loss": 0.0007, "step": 1642 }, { "epoch": 0.41865205758695373, "grad_norm": 1.8874963521957397, "learning_rate": 5.812945973496432e-06, "loss": 0.001, "step": 1643 }, { "epoch": 0.4189068671168302, "grad_norm": 146.2372589111328, "learning_rate": 5.81039755351682e-06, "loss": 0.1437, "step": 1644 }, { "epoch": 0.41916167664670656, "grad_norm": 1.0090858936309814, "learning_rate": 5.807849133537208e-06, "loss": 0.0007, "step": 1645 }, { "epoch": 0.419416486176583, "grad_norm": 1.113924503326416, "learning_rate": 5.805300713557595e-06, "loss": 0.0002, "step": 1646 }, { "epoch": 0.41967129570645945, "grad_norm": 1.7115830183029175, "learning_rate": 5.8027522935779825e-06, "loss": 0.0007, "step": 1647 }, { "epoch": 0.41992610523633583, "grad_norm": 2.4280920028686523, "learning_rate": 5.800203873598369e-06, "loss": 0.0007, "step": 1648 }, { "epoch": 0.4201809147662123, "grad_norm": 2.2553443908691406, "learning_rate": 5.797655453618757e-06, "loss": 0.0012, "step": 1649 }, { "epoch": 0.42043572429608866, "grad_norm": 1.6908833980560303, "learning_rate": 5.795107033639145e-06, "loss": 0.0004, "step": 1650 }, { "epoch": 0.4206905338259651, "grad_norm": 0.9557690024375916, "learning_rate": 5.792558613659531e-06, "loss": 0.0002, "step": 1651 }, { "epoch": 0.4209453433558415, "grad_norm": 7.552412033081055, "learning_rate": 5.790010193679919e-06, "loss": 0.0054, "step": 1652 }, { "epoch": 0.42120015288571794, "grad_norm": 6.378476619720459, "learning_rate": 5.787461773700306e-06, "loss": 0.0032, "step": 1653 }, { "epoch": 0.4214549624155943, "grad_norm": 4.117635250091553, "learning_rate": 5.784913353720694e-06, "loss": 0.0016, "step": 1654 }, { "epoch": 0.42170977194547077, "grad_norm": 0.9482070803642273, "learning_rate": 5.7823649337410815e-06, "loss": 0.0003, "step": 1655 }, { "epoch": 0.42196458147534716, "grad_norm": 0.5599589347839355, "learning_rate": 5.7798165137614684e-06, "loss": 0.0001, "step": 1656 }, { "epoch": 0.4222193910052236, "grad_norm": 2.975173234939575, "learning_rate": 5.777268093781856e-06, "loss": 0.0007, "step": 1657 }, { "epoch": 0.4224742005351, "grad_norm": 1.5107336044311523, "learning_rate": 5.774719673802242e-06, "loss": 0.0003, "step": 1658 }, { "epoch": 0.42272901006497643, "grad_norm": 0.2018614560365677, "learning_rate": 5.77217125382263e-06, "loss": 0.0002, "step": 1659 }, { "epoch": 0.4229838195948529, "grad_norm": 3.34543776512146, "learning_rate": 5.769622833843018e-06, "loss": 0.0009, "step": 1660 }, { "epoch": 0.42323862912472926, "grad_norm": 2.98475980758667, "learning_rate": 5.767074413863405e-06, "loss": 0.0008, "step": 1661 }, { "epoch": 0.4234934386546057, "grad_norm": 4.139312267303467, "learning_rate": 5.764525993883793e-06, "loss": 0.0014, "step": 1662 }, { "epoch": 0.4237482481844821, "grad_norm": 1.4464970827102661, "learning_rate": 5.76197757390418e-06, "loss": 0.001, "step": 1663 }, { "epoch": 0.42400305771435853, "grad_norm": 163.4978790283203, "learning_rate": 5.7594291539245674e-06, "loss": 0.0899, "step": 1664 }, { "epoch": 0.4242578672442349, "grad_norm": 2.693094491958618, "learning_rate": 5.756880733944955e-06, "loss": 0.0019, "step": 1665 }, { "epoch": 0.42451267677411136, "grad_norm": 3.1818740367889404, "learning_rate": 5.754332313965341e-06, "loss": 0.0011, "step": 1666 }, { "epoch": 0.42476748630398775, "grad_norm": 1.1174776554107666, "learning_rate": 5.751783893985729e-06, "loss": 0.0002, "step": 1667 }, { "epoch": 0.4250222958338642, "grad_norm": 3.3529539108276367, "learning_rate": 5.749235474006116e-06, "loss": 0.001, "step": 1668 }, { "epoch": 0.4252771053637406, "grad_norm": 1.123173475265503, "learning_rate": 5.746687054026504e-06, "loss": 0.0003, "step": 1669 }, { "epoch": 0.425531914893617, "grad_norm": 1.1210862398147583, "learning_rate": 5.744138634046892e-06, "loss": 0.0002, "step": 1670 }, { "epoch": 0.4257867244234934, "grad_norm": 3.1546542644500732, "learning_rate": 5.741590214067279e-06, "loss": 0.0007, "step": 1671 }, { "epoch": 0.42604153395336986, "grad_norm": 1.5032726526260376, "learning_rate": 5.7390417940876664e-06, "loss": 0.0004, "step": 1672 }, { "epoch": 0.4262963434832463, "grad_norm": 49.24574661254883, "learning_rate": 5.736493374108053e-06, "loss": 0.051, "step": 1673 }, { "epoch": 0.4265511530131227, "grad_norm": 0.3902004659175873, "learning_rate": 5.733944954128441e-06, "loss": 0.0002, "step": 1674 }, { "epoch": 0.42680596254299913, "grad_norm": 54.49069595336914, "learning_rate": 5.731396534148829e-06, "loss": 0.2147, "step": 1675 }, { "epoch": 0.4270607720728755, "grad_norm": 7.8239898681640625, "learning_rate": 5.728848114169215e-06, "loss": 0.0045, "step": 1676 }, { "epoch": 0.42731558160275196, "grad_norm": 7.8835577964782715, "learning_rate": 5.726299694189603e-06, "loss": 0.0041, "step": 1677 }, { "epoch": 0.42757039113262835, "grad_norm": 6.536893367767334, "learning_rate": 5.72375127420999e-06, "loss": 0.0032, "step": 1678 }, { "epoch": 0.4278252006625048, "grad_norm": 6.500173568725586, "learning_rate": 5.721202854230378e-06, "loss": 0.0031, "step": 1679 }, { "epoch": 0.4280800101923812, "grad_norm": 1.54367196559906, "learning_rate": 5.7186544342507654e-06, "loss": 0.0009, "step": 1680 }, { "epoch": 0.4283348197222576, "grad_norm": 1.6876887083053589, "learning_rate": 5.716106014271152e-06, "loss": 0.0003, "step": 1681 }, { "epoch": 0.428589629252134, "grad_norm": 4.094649314880371, "learning_rate": 5.71355759429154e-06, "loss": 0.0013, "step": 1682 }, { "epoch": 0.42884443878201045, "grad_norm": 56.062320709228516, "learning_rate": 5.711009174311926e-06, "loss": 0.7544, "step": 1683 }, { "epoch": 0.42909924831188684, "grad_norm": 6.535738945007324, "learning_rate": 5.708460754332314e-06, "loss": 0.0033, "step": 1684 }, { "epoch": 0.4293540578417633, "grad_norm": 5.9607648849487305, "learning_rate": 5.705912334352702e-06, "loss": 0.0036, "step": 1685 }, { "epoch": 0.4296088673716397, "grad_norm": 11.208115577697754, "learning_rate": 5.703363914373089e-06, "loss": 0.0066, "step": 1686 }, { "epoch": 0.4298636769015161, "grad_norm": 0.3790844678878784, "learning_rate": 5.700815494393477e-06, "loss": 0.0001, "step": 1687 }, { "epoch": 0.43011848643139255, "grad_norm": 4.083886623382568, "learning_rate": 5.698267074413864e-06, "loss": 0.0013, "step": 1688 }, { "epoch": 0.43037329596126894, "grad_norm": 5.5299763679504395, "learning_rate": 5.695718654434251e-06, "loss": 0.0026, "step": 1689 }, { "epoch": 0.4306281054911454, "grad_norm": 2.7973642349243164, "learning_rate": 5.693170234454639e-06, "loss": 0.001, "step": 1690 }, { "epoch": 0.4308829150210218, "grad_norm": 6.950647354125977, "learning_rate": 5.690621814475026e-06, "loss": 0.0047, "step": 1691 }, { "epoch": 0.4311377245508982, "grad_norm": 1.881259560585022, "learning_rate": 5.688073394495414e-06, "loss": 0.0009, "step": 1692 }, { "epoch": 0.4313925340807746, "grad_norm": 2.061190128326416, "learning_rate": 5.6855249745158e-06, "loss": 0.0006, "step": 1693 }, { "epoch": 0.43164734361065105, "grad_norm": 1.4948614835739136, "learning_rate": 5.682976554536188e-06, "loss": 0.0003, "step": 1694 }, { "epoch": 0.43190215314052743, "grad_norm": 0.6588423252105713, "learning_rate": 5.680428134556576e-06, "loss": 0.0008, "step": 1695 }, { "epoch": 0.4321569626704039, "grad_norm": 0.43550851941108704, "learning_rate": 5.677879714576963e-06, "loss": 0.0012, "step": 1696 }, { "epoch": 0.43241177220028026, "grad_norm": 1.6846463680267334, "learning_rate": 5.67533129459735e-06, "loss": 0.0004, "step": 1697 }, { "epoch": 0.4326665817301567, "grad_norm": 1.8501757383346558, "learning_rate": 5.672782874617737e-06, "loss": 0.0004, "step": 1698 }, { "epoch": 0.43292139126003315, "grad_norm": 2.6103711128234863, "learning_rate": 5.670234454638125e-06, "loss": 0.0008, "step": 1699 }, { "epoch": 0.43317620078990954, "grad_norm": 3.347325325012207, "learning_rate": 5.667686034658511e-06, "loss": 0.0009, "step": 1700 }, { "epoch": 0.433431010319786, "grad_norm": 3.934476137161255, "learning_rate": 5.665137614678899e-06, "loss": 0.0014, "step": 1701 }, { "epoch": 0.43368581984966237, "grad_norm": 1.4950400590896606, "learning_rate": 5.662589194699287e-06, "loss": 0.0005, "step": 1702 }, { "epoch": 0.4339406293795388, "grad_norm": 2.632445812225342, "learning_rate": 5.660040774719674e-06, "loss": 0.0013, "step": 1703 }, { "epoch": 0.4341954389094152, "grad_norm": 8.349006652832031, "learning_rate": 5.657492354740062e-06, "loss": 0.0066, "step": 1704 }, { "epoch": 0.43445024843929164, "grad_norm": 2.2700817584991455, "learning_rate": 5.654943934760449e-06, "loss": 0.0007, "step": 1705 }, { "epoch": 0.43470505796916803, "grad_norm": 0.802450954914093, "learning_rate": 5.652395514780836e-06, "loss": 0.0006, "step": 1706 }, { "epoch": 0.4349598674990445, "grad_norm": 1.8666948080062866, "learning_rate": 5.649847094801224e-06, "loss": 0.0004, "step": 1707 }, { "epoch": 0.43521467702892086, "grad_norm": 43.050209045410156, "learning_rate": 5.647298674821611e-06, "loss": 0.1147, "step": 1708 }, { "epoch": 0.4354694865587973, "grad_norm": 1.3033443689346313, "learning_rate": 5.644750254841999e-06, "loss": 0.0002, "step": 1709 }, { "epoch": 0.4357242960886737, "grad_norm": 5.425432205200195, "learning_rate": 5.642201834862385e-06, "loss": 0.0022, "step": 1710 }, { "epoch": 0.43597910561855013, "grad_norm": 3.1706137657165527, "learning_rate": 5.639653414882773e-06, "loss": 0.001, "step": 1711 }, { "epoch": 0.4362339151484266, "grad_norm": 92.9125747680664, "learning_rate": 5.637104994903161e-06, "loss": 4.0626, "step": 1712 }, { "epoch": 0.43648872467830296, "grad_norm": 7.007720470428467, "learning_rate": 5.634556574923548e-06, "loss": 0.0074, "step": 1713 }, { "epoch": 0.4367435342081794, "grad_norm": 3.717979669570923, "learning_rate": 5.632008154943935e-06, "loss": 0.0013, "step": 1714 }, { "epoch": 0.4369983437380558, "grad_norm": 7.391612529754639, "learning_rate": 5.629459734964322e-06, "loss": 0.0037, "step": 1715 }, { "epoch": 0.43725315326793224, "grad_norm": 4.138318061828613, "learning_rate": 5.62691131498471e-06, "loss": 0.0014, "step": 1716 }, { "epoch": 0.4375079627978086, "grad_norm": 2.268493890762329, "learning_rate": 5.624362895005098e-06, "loss": 0.0008, "step": 1717 }, { "epoch": 0.43776277232768507, "grad_norm": 4.471026420593262, "learning_rate": 5.621814475025484e-06, "loss": 0.002, "step": 1718 }, { "epoch": 0.43801758185756146, "grad_norm": 2.8111894130706787, "learning_rate": 5.619266055045872e-06, "loss": 0.0011, "step": 1719 }, { "epoch": 0.4382723913874379, "grad_norm": 6.714369297027588, "learning_rate": 5.616717635066259e-06, "loss": 0.0034, "step": 1720 }, { "epoch": 0.4385272009173143, "grad_norm": 2.9774277210235596, "learning_rate": 5.614169215086647e-06, "loss": 0.0011, "step": 1721 }, { "epoch": 0.43878201044719073, "grad_norm": 3.9297642707824707, "learning_rate": 5.611620795107034e-06, "loss": 0.0014, "step": 1722 }, { "epoch": 0.4390368199770671, "grad_norm": 0.16693729162216187, "learning_rate": 5.609072375127421e-06, "loss": 0.0005, "step": 1723 }, { "epoch": 0.43929162950694356, "grad_norm": 4.952264308929443, "learning_rate": 5.606523955147809e-06, "loss": 0.0021, "step": 1724 }, { "epoch": 0.43954643903682, "grad_norm": 2.4788973331451416, "learning_rate": 5.603975535168196e-06, "loss": 0.003, "step": 1725 }, { "epoch": 0.4398012485666964, "grad_norm": 6.522562503814697, "learning_rate": 5.601427115188584e-06, "loss": 0.0036, "step": 1726 }, { "epoch": 0.44005605809657283, "grad_norm": 9.143147468566895, "learning_rate": 5.598878695208972e-06, "loss": 0.0062, "step": 1727 }, { "epoch": 0.4403108676264492, "grad_norm": 4.491147041320801, "learning_rate": 5.596330275229358e-06, "loss": 0.0019, "step": 1728 }, { "epoch": 0.44056567715632566, "grad_norm": 1.0278325080871582, "learning_rate": 5.593781855249746e-06, "loss": 0.0005, "step": 1729 }, { "epoch": 0.44082048668620205, "grad_norm": 0.3271080553531647, "learning_rate": 5.5912334352701326e-06, "loss": 0.0009, "step": 1730 }, { "epoch": 0.4410752962160785, "grad_norm": 2.153355836868286, "learning_rate": 5.58868501529052e-06, "loss": 0.001, "step": 1731 }, { "epoch": 0.4413301057459549, "grad_norm": 0.469472199678421, "learning_rate": 5.586136595310908e-06, "loss": 0.0004, "step": 1732 }, { "epoch": 0.4415849152758313, "grad_norm": 4.960258960723877, "learning_rate": 5.583588175331295e-06, "loss": 0.0022, "step": 1733 }, { "epoch": 0.4418397248057077, "grad_norm": 5.581376552581787, "learning_rate": 5.581039755351683e-06, "loss": 0.0027, "step": 1734 }, { "epoch": 0.44209453433558416, "grad_norm": 0.3146277666091919, "learning_rate": 5.578491335372069e-06, "loss": 0.001, "step": 1735 }, { "epoch": 0.44234934386546054, "grad_norm": 0.2018270194530487, "learning_rate": 5.575942915392457e-06, "loss": 0.0004, "step": 1736 }, { "epoch": 0.442604153395337, "grad_norm": 1.1285618543624878, "learning_rate": 5.573394495412845e-06, "loss": 0.0006, "step": 1737 }, { "epoch": 0.44285896292521343, "grad_norm": 302.4094543457031, "learning_rate": 5.5708460754332316e-06, "loss": 0.2385, "step": 1738 }, { "epoch": 0.4431137724550898, "grad_norm": 5.408868789672852, "learning_rate": 5.568297655453619e-06, "loss": 0.0024, "step": 1739 }, { "epoch": 0.44336858198496626, "grad_norm": 0.24494542181491852, "learning_rate": 5.565749235474006e-06, "loss": 0.0006, "step": 1740 }, { "epoch": 0.44362339151484265, "grad_norm": 0.7539666295051575, "learning_rate": 5.563200815494394e-06, "loss": 0.0005, "step": 1741 }, { "epoch": 0.4438782010447191, "grad_norm": 1.3013606071472168, "learning_rate": 5.560652395514782e-06, "loss": 0.0004, "step": 1742 }, { "epoch": 0.4441330105745955, "grad_norm": 0.9468243718147278, "learning_rate": 5.558103975535169e-06, "loss": 0.0005, "step": 1743 }, { "epoch": 0.4443878201044719, "grad_norm": 62.11100769042969, "learning_rate": 5.555555555555557e-06, "loss": 0.0161, "step": 1744 }, { "epoch": 0.4446426296343483, "grad_norm": 1.3458471298217773, "learning_rate": 5.553007135575943e-06, "loss": 0.0011, "step": 1745 }, { "epoch": 0.44489743916422475, "grad_norm": 1.6671853065490723, "learning_rate": 5.5504587155963306e-06, "loss": 0.0005, "step": 1746 }, { "epoch": 0.44515224869410114, "grad_norm": 0.05719710513949394, "learning_rate": 5.547910295616718e-06, "loss": 0.0003, "step": 1747 }, { "epoch": 0.4454070582239776, "grad_norm": 1.6961263418197632, "learning_rate": 5.545361875637105e-06, "loss": 0.0007, "step": 1748 }, { "epoch": 0.44566186775385397, "grad_norm": 0.5770021677017212, "learning_rate": 5.542813455657493e-06, "loss": 0.0003, "step": 1749 }, { "epoch": 0.4459166772837304, "grad_norm": 0.35618388652801514, "learning_rate": 5.54026503567788e-06, "loss": 0.0012, "step": 1750 }, { "epoch": 0.44617148681360685, "grad_norm": 1.426931381225586, "learning_rate": 5.537716615698268e-06, "loss": 0.0023, "step": 1751 }, { "epoch": 0.44642629634348324, "grad_norm": 8.438220977783203, "learning_rate": 5.535168195718656e-06, "loss": 0.0075, "step": 1752 }, { "epoch": 0.4466811058733597, "grad_norm": 1.4973064661026, "learning_rate": 5.532619775739042e-06, "loss": 0.0004, "step": 1753 }, { "epoch": 0.4469359154032361, "grad_norm": 0.7498091459274292, "learning_rate": 5.53007135575943e-06, "loss": 0.0003, "step": 1754 }, { "epoch": 0.4471907249331125, "grad_norm": 0.7559434175491333, "learning_rate": 5.5275229357798165e-06, "loss": 0.0009, "step": 1755 }, { "epoch": 0.4474455344629889, "grad_norm": 2.803091049194336, "learning_rate": 5.524974515800204e-06, "loss": 0.0008, "step": 1756 }, { "epoch": 0.44770034399286535, "grad_norm": 2.23598313331604, "learning_rate": 5.522426095820592e-06, "loss": 0.0007, "step": 1757 }, { "epoch": 0.44795515352274173, "grad_norm": 0.2693818211555481, "learning_rate": 5.519877675840979e-06, "loss": 0.0003, "step": 1758 }, { "epoch": 0.4482099630526182, "grad_norm": 3.180786609649658, "learning_rate": 5.517329255861367e-06, "loss": 0.0013, "step": 1759 }, { "epoch": 0.44846477258249456, "grad_norm": 2.426375389099121, "learning_rate": 5.514780835881753e-06, "loss": 0.0006, "step": 1760 }, { "epoch": 0.448719582112371, "grad_norm": 3.9202020168304443, "learning_rate": 5.512232415902141e-06, "loss": 0.0015, "step": 1761 }, { "epoch": 0.4489743916422474, "grad_norm": 3.7358641624450684, "learning_rate": 5.509683995922529e-06, "loss": 0.0012, "step": 1762 }, { "epoch": 0.44922920117212384, "grad_norm": 2.7986788749694824, "learning_rate": 5.5071355759429155e-06, "loss": 0.0007, "step": 1763 }, { "epoch": 0.4494840107020003, "grad_norm": 3.3854451179504395, "learning_rate": 5.504587155963303e-06, "loss": 0.0012, "step": 1764 }, { "epoch": 0.44973882023187667, "grad_norm": 4.87045955657959, "learning_rate": 5.50203873598369e-06, "loss": 0.0022, "step": 1765 }, { "epoch": 0.4499936297617531, "grad_norm": 3.9898715019226074, "learning_rate": 5.499490316004078e-06, "loss": 0.0014, "step": 1766 }, { "epoch": 0.4502484392916295, "grad_norm": 4.509585857391357, "learning_rate": 5.496941896024466e-06, "loss": 0.0028, "step": 1767 }, { "epoch": 0.45050324882150594, "grad_norm": 2.050980806350708, "learning_rate": 5.494393476044853e-06, "loss": 0.0006, "step": 1768 }, { "epoch": 0.45075805835138233, "grad_norm": 3.9172730445861816, "learning_rate": 5.491845056065241e-06, "loss": 0.0016, "step": 1769 }, { "epoch": 0.45101286788125877, "grad_norm": 646.243896484375, "learning_rate": 5.489296636085627e-06, "loss": 0.2061, "step": 1770 }, { "epoch": 0.45126767741113516, "grad_norm": 1383.3048095703125, "learning_rate": 5.4867482161060146e-06, "loss": 2.267, "step": 1771 }, { "epoch": 0.4515224869410116, "grad_norm": 5.224039077758789, "learning_rate": 5.484199796126402e-06, "loss": 0.0022, "step": 1772 }, { "epoch": 0.451777296470888, "grad_norm": 114.83236694335938, "learning_rate": 5.481651376146789e-06, "loss": 0.0101, "step": 1773 }, { "epoch": 0.45203210600076443, "grad_norm": 9.31849479675293, "learning_rate": 5.479102956167177e-06, "loss": 0.0064, "step": 1774 }, { "epoch": 0.4522869155306408, "grad_norm": 11.519116401672363, "learning_rate": 5.476554536187564e-06, "loss": 0.01, "step": 1775 }, { "epoch": 0.45254172506051726, "grad_norm": 65.38799285888672, "learning_rate": 5.474006116207952e-06, "loss": 0.7693, "step": 1776 }, { "epoch": 0.4527965345903937, "grad_norm": 5.604466438293457, "learning_rate": 5.471457696228338e-06, "loss": 0.0024, "step": 1777 }, { "epoch": 0.4530513441202701, "grad_norm": 2.258413553237915, "learning_rate": 5.468909276248726e-06, "loss": 0.0006, "step": 1778 }, { "epoch": 0.45330615365014654, "grad_norm": 1.3071491718292236, "learning_rate": 5.4663608562691136e-06, "loss": 0.0003, "step": 1779 }, { "epoch": 0.4535609631800229, "grad_norm": 91.42700958251953, "learning_rate": 5.4638124362895005e-06, "loss": 3.9071, "step": 1780 }, { "epoch": 0.45381577270989937, "grad_norm": 7.728638648986816, "learning_rate": 5.461264016309888e-06, "loss": 0.0044, "step": 1781 }, { "epoch": 0.45407058223977576, "grad_norm": 6.680112838745117, "learning_rate": 5.458715596330275e-06, "loss": 0.0035, "step": 1782 }, { "epoch": 0.4543253917696522, "grad_norm": 1.5623997449874878, "learning_rate": 5.456167176350663e-06, "loss": 0.0018, "step": 1783 }, { "epoch": 0.4545802012995286, "grad_norm": 2.4258010387420654, "learning_rate": 5.453618756371051e-06, "loss": 0.001, "step": 1784 }, { "epoch": 0.45483501082940503, "grad_norm": 1.4268710613250732, "learning_rate": 5.451070336391438e-06, "loss": 0.0007, "step": 1785 }, { "epoch": 0.4550898203592814, "grad_norm": 1.6032658815383911, "learning_rate": 5.448521916411826e-06, "loss": 0.0006, "step": 1786 }, { "epoch": 0.45534462988915786, "grad_norm": 5.20622444152832, "learning_rate": 5.445973496432212e-06, "loss": 0.0024, "step": 1787 }, { "epoch": 0.45559943941903425, "grad_norm": 3.5571699142456055, "learning_rate": 5.4434250764525995e-06, "loss": 0.0014, "step": 1788 }, { "epoch": 0.4558542489489107, "grad_norm": 3.6580069065093994, "learning_rate": 5.440876656472987e-06, "loss": 0.0016, "step": 1789 }, { "epoch": 0.45610905847878713, "grad_norm": 2.7264435291290283, "learning_rate": 5.438328236493374e-06, "loss": 0.0011, "step": 1790 }, { "epoch": 0.4563638680086635, "grad_norm": 0.4166838824748993, "learning_rate": 5.435779816513762e-06, "loss": 0.001, "step": 1791 }, { "epoch": 0.45661867753853996, "grad_norm": 2.712650775909424, "learning_rate": 5.433231396534149e-06, "loss": 0.001, "step": 1792 }, { "epoch": 0.45687348706841635, "grad_norm": 3.544492244720459, "learning_rate": 5.430682976554537e-06, "loss": 0.0016, "step": 1793 }, { "epoch": 0.4571282965982928, "grad_norm": 6.933889865875244, "learning_rate": 5.428134556574925e-06, "loss": 0.0042, "step": 1794 }, { "epoch": 0.4573831061281692, "grad_norm": 0.4755961298942566, "learning_rate": 5.425586136595311e-06, "loss": 0.0013, "step": 1795 }, { "epoch": 0.4576379156580456, "grad_norm": 0.6592304706573486, "learning_rate": 5.4230377166156985e-06, "loss": 0.0004, "step": 1796 }, { "epoch": 0.457892725187922, "grad_norm": 0.08212131261825562, "learning_rate": 5.4204892966360855e-06, "loss": 0.0004, "step": 1797 }, { "epoch": 0.45814753471779845, "grad_norm": 1.4097470045089722, "learning_rate": 5.417940876656473e-06, "loss": 0.0008, "step": 1798 }, { "epoch": 0.45840234424767484, "grad_norm": 0.604625403881073, "learning_rate": 5.415392456676861e-06, "loss": 0.0008, "step": 1799 }, { "epoch": 0.4586571537775513, "grad_norm": 3.1725618839263916, "learning_rate": 5.412844036697248e-06, "loss": 0.0014, "step": 1800 }, { "epoch": 0.4589119633074277, "grad_norm": 0.9300532341003418, "learning_rate": 5.410295616717636e-06, "loss": 0.0004, "step": 1801 }, { "epoch": 0.4591667728373041, "grad_norm": 1.0304688215255737, "learning_rate": 5.407747196738023e-06, "loss": 0.0004, "step": 1802 }, { "epoch": 0.45942158236718056, "grad_norm": 2.241234540939331, "learning_rate": 5.405198776758411e-06, "loss": 0.001, "step": 1803 }, { "epoch": 0.45967639189705695, "grad_norm": 0.9437415599822998, "learning_rate": 5.402650356778798e-06, "loss": 0.0005, "step": 1804 }, { "epoch": 0.4599312014269334, "grad_norm": 1.87328040599823, "learning_rate": 5.4001019367991845e-06, "loss": 0.0009, "step": 1805 }, { "epoch": 0.4601860109568098, "grad_norm": 301.96319580078125, "learning_rate": 5.397553516819572e-06, "loss": 0.3674, "step": 1806 }, { "epoch": 0.4604408204866862, "grad_norm": 2.6327433586120605, "learning_rate": 5.395005096839959e-06, "loss": 0.0009, "step": 1807 }, { "epoch": 0.4606956300165626, "grad_norm": 4.587804317474365, "learning_rate": 5.392456676860347e-06, "loss": 0.0028, "step": 1808 }, { "epoch": 0.46095043954643905, "grad_norm": 2.172637939453125, "learning_rate": 5.389908256880735e-06, "loss": 0.0009, "step": 1809 }, { "epoch": 0.46120524907631544, "grad_norm": 1.1225390434265137, "learning_rate": 5.387359836901122e-06, "loss": 0.0005, "step": 1810 }, { "epoch": 0.4614600586061919, "grad_norm": 1.6767244338989258, "learning_rate": 5.38481141692151e-06, "loss": 0.0004, "step": 1811 }, { "epoch": 0.46171486813606827, "grad_norm": 5.302921772003174, "learning_rate": 5.382262996941896e-06, "loss": 0.0022, "step": 1812 }, { "epoch": 0.4619696776659447, "grad_norm": 3.256617546081543, "learning_rate": 5.3797145769622835e-06, "loss": 0.0011, "step": 1813 }, { "epoch": 0.4622244871958211, "grad_norm": 1.7951229810714722, "learning_rate": 5.377166156982671e-06, "loss": 0.0009, "step": 1814 }, { "epoch": 0.46247929672569754, "grad_norm": 2.815565824508667, "learning_rate": 5.374617737003058e-06, "loss": 0.0014, "step": 1815 }, { "epoch": 0.462734106255574, "grad_norm": 2.992102861404419, "learning_rate": 5.372069317023446e-06, "loss": 0.001, "step": 1816 }, { "epoch": 0.4629889157854504, "grad_norm": 4.5454535484313965, "learning_rate": 5.369520897043833e-06, "loss": 0.0028, "step": 1817 }, { "epoch": 0.4632437253153268, "grad_norm": 6.910698413848877, "learning_rate": 5.366972477064221e-06, "loss": 0.0038, "step": 1818 }, { "epoch": 0.4634985348452032, "grad_norm": 0.20612062513828278, "learning_rate": 5.364424057084609e-06, "loss": 0.0003, "step": 1819 }, { "epoch": 0.46375334437507965, "grad_norm": 2.056131601333618, "learning_rate": 5.3618756371049956e-06, "loss": 0.0008, "step": 1820 }, { "epoch": 0.46400815390495603, "grad_norm": 3.2509450912475586, "learning_rate": 5.359327217125383e-06, "loss": 0.001, "step": 1821 }, { "epoch": 0.4642629634348325, "grad_norm": 2450.322021484375, "learning_rate": 5.3567787971457695e-06, "loss": 3.1256, "step": 1822 }, { "epoch": 0.46451777296470886, "grad_norm": 0.38157162070274353, "learning_rate": 5.354230377166157e-06, "loss": 0.0007, "step": 1823 }, { "epoch": 0.4647725824945853, "grad_norm": 110.28287506103516, "learning_rate": 5.351681957186545e-06, "loss": 0.6174, "step": 1824 }, { "epoch": 0.4650273920244617, "grad_norm": 0.7495670914649963, "learning_rate": 5.349133537206932e-06, "loss": 0.0003, "step": 1825 }, { "epoch": 0.46528220155433814, "grad_norm": 1.5047953128814697, "learning_rate": 5.34658511722732e-06, "loss": 0.0004, "step": 1826 }, { "epoch": 0.4655370110842145, "grad_norm": 0.4719366729259491, "learning_rate": 5.344036697247707e-06, "loss": 0.0005, "step": 1827 }, { "epoch": 0.46579182061409097, "grad_norm": 0.5615885853767395, "learning_rate": 5.3414882772680946e-06, "loss": 0.0002, "step": 1828 }, { "epoch": 0.4660466301439674, "grad_norm": 0.5813694596290588, "learning_rate": 5.338939857288482e-06, "loss": 0.0004, "step": 1829 }, { "epoch": 0.4663014396738438, "grad_norm": 61.373294830322266, "learning_rate": 5.3363914373088685e-06, "loss": 0.6377, "step": 1830 }, { "epoch": 0.46655624920372024, "grad_norm": 0.1980016529560089, "learning_rate": 5.333843017329256e-06, "loss": 0.0002, "step": 1831 }, { "epoch": 0.46681105873359663, "grad_norm": 2.2371208667755127, "learning_rate": 5.331294597349643e-06, "loss": 0.0005, "step": 1832 }, { "epoch": 0.46706586826347307, "grad_norm": 203.7059326171875, "learning_rate": 5.328746177370031e-06, "loss": 0.5318, "step": 1833 }, { "epoch": 0.46732067779334946, "grad_norm": 3.932008743286133, "learning_rate": 5.326197757390419e-06, "loss": 0.0013, "step": 1834 }, { "epoch": 0.4675754873232259, "grad_norm": 0.1380815953016281, "learning_rate": 5.323649337410806e-06, "loss": 0.0003, "step": 1835 }, { "epoch": 0.4678302968531023, "grad_norm": 0.08435986936092377, "learning_rate": 5.3211009174311936e-06, "loss": 0.0003, "step": 1836 }, { "epoch": 0.46808510638297873, "grad_norm": 1.6793779134750366, "learning_rate": 5.31855249745158e-06, "loss": 0.0004, "step": 1837 }, { "epoch": 0.4683399159128551, "grad_norm": 0.9402336478233337, "learning_rate": 5.3160040774719675e-06, "loss": 0.0003, "step": 1838 }, { "epoch": 0.46859472544273156, "grad_norm": 0.9381915926933289, "learning_rate": 5.313455657492356e-06, "loss": 0.0003, "step": 1839 }, { "epoch": 0.46884953497260795, "grad_norm": 3.544792890548706, "learning_rate": 5.310907237512742e-06, "loss": 0.0011, "step": 1840 }, { "epoch": 0.4691043445024844, "grad_norm": 1.121354103088379, "learning_rate": 5.30835881753313e-06, "loss": 0.0003, "step": 1841 }, { "epoch": 0.46935915403236084, "grad_norm": 0.9372594952583313, "learning_rate": 5.305810397553517e-06, "loss": 0.0019, "step": 1842 }, { "epoch": 0.4696139635622372, "grad_norm": 1.313752293586731, "learning_rate": 5.303261977573905e-06, "loss": 0.0006, "step": 1843 }, { "epoch": 0.46986877309211367, "grad_norm": 0.5612956285476685, "learning_rate": 5.300713557594293e-06, "loss": 0.0002, "step": 1844 }, { "epoch": 0.47012358262199005, "grad_norm": 5.781057357788086, "learning_rate": 5.2981651376146795e-06, "loss": 0.003, "step": 1845 }, { "epoch": 0.4703783921518665, "grad_norm": 1.4867254495620728, "learning_rate": 5.295616717635067e-06, "loss": 0.0003, "step": 1846 }, { "epoch": 0.4706332016817429, "grad_norm": 2.254363536834717, "learning_rate": 5.2930682976554534e-06, "loss": 0.0005, "step": 1847 }, { "epoch": 0.47088801121161933, "grad_norm": 3.355599880218506, "learning_rate": 5.290519877675841e-06, "loss": 0.0013, "step": 1848 }, { "epoch": 0.4711428207414957, "grad_norm": 0.393839567899704, "learning_rate": 5.287971457696228e-06, "loss": 0.0003, "step": 1849 }, { "epoch": 0.47139763027137216, "grad_norm": 2.4298477172851562, "learning_rate": 5.285423037716616e-06, "loss": 0.0006, "step": 1850 }, { "epoch": 0.47165243980124855, "grad_norm": 1.8622450828552246, "learning_rate": 5.282874617737004e-06, "loss": 0.0004, "step": 1851 }, { "epoch": 0.471907249331125, "grad_norm": 3.531547784805298, "learning_rate": 5.280326197757391e-06, "loss": 0.0011, "step": 1852 }, { "epoch": 0.4721620588610014, "grad_norm": 64.3506851196289, "learning_rate": 5.2777777777777785e-06, "loss": 0.766, "step": 1853 }, { "epoch": 0.4724168683908778, "grad_norm": 3.1619162559509277, "learning_rate": 5.275229357798165e-06, "loss": 0.001, "step": 1854 }, { "epoch": 0.47267167792075426, "grad_norm": 2.0537874698638916, "learning_rate": 5.2726809378185524e-06, "loss": 0.0006, "step": 1855 }, { "epoch": 0.47292648745063065, "grad_norm": 1.3336927890777588, "learning_rate": 5.27013251783894e-06, "loss": 0.0006, "step": 1856 }, { "epoch": 0.4731812969805071, "grad_norm": 0.37693628668785095, "learning_rate": 5.267584097859327e-06, "loss": 0.0002, "step": 1857 }, { "epoch": 0.4734361065103835, "grad_norm": 2.419309616088867, "learning_rate": 5.265035677879715e-06, "loss": 0.0007, "step": 1858 }, { "epoch": 0.4736909160402599, "grad_norm": 1.5648800134658813, "learning_rate": 5.262487257900102e-06, "loss": 0.0007, "step": 1859 }, { "epoch": 0.4739457255701363, "grad_norm": 17.46479606628418, "learning_rate": 5.25993883792049e-06, "loss": 0.0066, "step": 1860 }, { "epoch": 0.47420053510001275, "grad_norm": 5.241545677185059, "learning_rate": 5.2573904179408775e-06, "loss": 0.0021, "step": 1861 }, { "epoch": 0.47445534462988914, "grad_norm": 1.4879193305969238, "learning_rate": 5.2548419979612645e-06, "loss": 0.0004, "step": 1862 }, { "epoch": 0.4747101541597656, "grad_norm": 2.2414450645446777, "learning_rate": 5.252293577981652e-06, "loss": 0.0005, "step": 1863 }, { "epoch": 0.474964963689642, "grad_norm": 1.676661491394043, "learning_rate": 5.249745158002038e-06, "loss": 0.0004, "step": 1864 }, { "epoch": 0.4752197732195184, "grad_norm": 2.066896438598633, "learning_rate": 5.247196738022426e-06, "loss": 0.0008, "step": 1865 }, { "epoch": 0.4754745827493948, "grad_norm": 5.9513468742370605, "learning_rate": 5.244648318042814e-06, "loss": 0.0027, "step": 1866 }, { "epoch": 0.47572939227927125, "grad_norm": 8.696468353271484, "learning_rate": 5.242099898063201e-06, "loss": 0.0056, "step": 1867 }, { "epoch": 0.4759842018091477, "grad_norm": 8.069467544555664, "learning_rate": 5.239551478083589e-06, "loss": 0.0048, "step": 1868 }, { "epoch": 0.4762390113390241, "grad_norm": 9.2771635055542, "learning_rate": 5.237003058103976e-06, "loss": 0.0065, "step": 1869 }, { "epoch": 0.4764938208689005, "grad_norm": 7.634064197540283, "learning_rate": 5.2344546381243635e-06, "loss": 0.0043, "step": 1870 }, { "epoch": 0.4767486303987769, "grad_norm": 4.286162376403809, "learning_rate": 5.231906218144751e-06, "loss": 0.0015, "step": 1871 }, { "epoch": 0.47700343992865335, "grad_norm": 1.2807914018630981, "learning_rate": 5.229357798165137e-06, "loss": 0.0014, "step": 1872 }, { "epoch": 0.47725824945852974, "grad_norm": 2.606138229370117, "learning_rate": 5.226809378185525e-06, "loss": 0.0007, "step": 1873 }, { "epoch": 0.4775130589884062, "grad_norm": 87.72962188720703, "learning_rate": 5.224260958205912e-06, "loss": 4.0627, "step": 1874 }, { "epoch": 0.47776786851828257, "grad_norm": 2.6040046215057373, "learning_rate": 5.2217125382263e-06, "loss": 0.0007, "step": 1875 }, { "epoch": 0.478022678048159, "grad_norm": 5.034156799316406, "learning_rate": 5.219164118246688e-06, "loss": 0.0021, "step": 1876 }, { "epoch": 0.4782774875780354, "grad_norm": 6.499547004699707, "learning_rate": 5.216615698267075e-06, "loss": 0.004, "step": 1877 }, { "epoch": 0.47853229710791184, "grad_norm": 1.9629807472229004, "learning_rate": 5.2140672782874625e-06, "loss": 0.0007, "step": 1878 }, { "epoch": 0.47878710663778823, "grad_norm": 5.692201614379883, "learning_rate": 5.2115188583078495e-06, "loss": 0.0028, "step": 1879 }, { "epoch": 0.47904191616766467, "grad_norm": 1.220809817314148, "learning_rate": 5.208970438328237e-06, "loss": 0.0007, "step": 1880 }, { "epoch": 0.4792967256975411, "grad_norm": 0.850121796131134, "learning_rate": 5.206422018348625e-06, "loss": 0.0006, "step": 1881 }, { "epoch": 0.4795515352274175, "grad_norm": 1.4858002662658691, "learning_rate": 5.203873598369011e-06, "loss": 0.0006, "step": 1882 }, { "epoch": 0.47980634475729395, "grad_norm": 0.5604867935180664, "learning_rate": 5.201325178389399e-06, "loss": 0.0005, "step": 1883 }, { "epoch": 0.48006115428717033, "grad_norm": 0.2294938564300537, "learning_rate": 5.198776758409786e-06, "loss": 0.0005, "step": 1884 }, { "epoch": 0.4803159638170468, "grad_norm": 0.8444663882255554, "learning_rate": 5.196228338430174e-06, "loss": 0.0004, "step": 1885 }, { "epoch": 0.48057077334692316, "grad_norm": 1.4006803035736084, "learning_rate": 5.1936799184505615e-06, "loss": 0.0007, "step": 1886 }, { "epoch": 0.4808255828767996, "grad_norm": 0.3829963505268097, "learning_rate": 5.1911314984709485e-06, "loss": 0.0005, "step": 1887 }, { "epoch": 0.481080392406676, "grad_norm": 1.5948599576950073, "learning_rate": 5.188583078491336e-06, "loss": 0.0009, "step": 1888 }, { "epoch": 0.48133520193655244, "grad_norm": 4.727985382080078, "learning_rate": 5.186034658511722e-06, "loss": 0.0019, "step": 1889 }, { "epoch": 0.4815900114664288, "grad_norm": 0.4733056426048279, "learning_rate": 5.18348623853211e-06, "loss": 0.0005, "step": 1890 }, { "epoch": 0.48184482099630527, "grad_norm": 234.17141723632812, "learning_rate": 5.180937818552498e-06, "loss": 0.1605, "step": 1891 }, { "epoch": 0.48209963052618166, "grad_norm": 0.3743118643760681, "learning_rate": 5.178389398572885e-06, "loss": 0.0004, "step": 1892 }, { "epoch": 0.4823544400560581, "grad_norm": 4.283950328826904, "learning_rate": 5.175840978593273e-06, "loss": 0.0017, "step": 1893 }, { "epoch": 0.48260924958593454, "grad_norm": 1.8542336225509644, "learning_rate": 5.17329255861366e-06, "loss": 0.0007, "step": 1894 }, { "epoch": 0.48286405911581093, "grad_norm": 49.049171447753906, "learning_rate": 5.1707441386340475e-06, "loss": 0.602, "step": 1895 }, { "epoch": 0.48311886864568737, "grad_norm": 1.1327205896377563, "learning_rate": 5.168195718654435e-06, "loss": 0.0006, "step": 1896 }, { "epoch": 0.48337367817556376, "grad_norm": 3.1620125770568848, "learning_rate": 5.165647298674822e-06, "loss": 0.001, "step": 1897 }, { "epoch": 0.4836284877054402, "grad_norm": 5.082614421844482, "learning_rate": 5.16309887869521e-06, "loss": 0.0031, "step": 1898 }, { "epoch": 0.4838832972353166, "grad_norm": 2.53269362449646, "learning_rate": 5.160550458715596e-06, "loss": 0.0015, "step": 1899 }, { "epoch": 0.48413810676519303, "grad_norm": 1.9290642738342285, "learning_rate": 5.158002038735984e-06, "loss": 0.0016, "step": 1900 }, { "epoch": 0.4843929162950694, "grad_norm": 1.590286135673523, "learning_rate": 5.155453618756372e-06, "loss": 0.0006, "step": 1901 }, { "epoch": 0.48464772582494586, "grad_norm": 1.215433955192566, "learning_rate": 5.152905198776759e-06, "loss": 0.0007, "step": 1902 }, { "epoch": 0.48490253535482225, "grad_norm": 2.88985013961792, "learning_rate": 5.1503567787971465e-06, "loss": 0.001, "step": 1903 }, { "epoch": 0.4851573448846987, "grad_norm": 4.564399242401123, "learning_rate": 5.1478083588175334e-06, "loss": 0.002, "step": 1904 }, { "epoch": 0.4854121544145751, "grad_norm": 1.2991440296173096, "learning_rate": 5.145259938837921e-06, "loss": 0.0004, "step": 1905 }, { "epoch": 0.4856669639444515, "grad_norm": 1.2852940559387207, "learning_rate": 5.142711518858309e-06, "loss": 0.002, "step": 1906 }, { "epoch": 0.48592177347432797, "grad_norm": 1.1164758205413818, "learning_rate": 5.140163098878695e-06, "loss": 0.0004, "step": 1907 }, { "epoch": 0.48617658300420435, "grad_norm": 3.1520469188690186, "learning_rate": 5.137614678899083e-06, "loss": 0.0012, "step": 1908 }, { "epoch": 0.4864313925340808, "grad_norm": 3.436539888381958, "learning_rate": 5.13506625891947e-06, "loss": 0.0014, "step": 1909 }, { "epoch": 0.4866862020639572, "grad_norm": 1.863572359085083, "learning_rate": 5.132517838939858e-06, "loss": 0.0008, "step": 1910 }, { "epoch": 0.48694101159383363, "grad_norm": 4.635220527648926, "learning_rate": 5.1299694189602455e-06, "loss": 0.0019, "step": 1911 }, { "epoch": 0.48719582112371, "grad_norm": 8.292279243469238, "learning_rate": 5.1274209989806325e-06, "loss": 0.0052, "step": 1912 }, { "epoch": 0.48745063065358646, "grad_norm": 0.37757331132888794, "learning_rate": 5.12487257900102e-06, "loss": 0.0003, "step": 1913 }, { "epoch": 0.48770544018346285, "grad_norm": 2.9825873374938965, "learning_rate": 5.122324159021407e-06, "loss": 0.0017, "step": 1914 }, { "epoch": 0.4879602497133393, "grad_norm": 4.004375457763672, "learning_rate": 5.119775739041795e-06, "loss": 0.0016, "step": 1915 }, { "epoch": 0.4882150592432157, "grad_norm": 17.207448959350586, "learning_rate": 5.117227319062183e-06, "loss": 0.0145, "step": 1916 }, { "epoch": 0.4884698687730921, "grad_norm": 2.242169141769409, "learning_rate": 5.114678899082569e-06, "loss": 0.0007, "step": 1917 }, { "epoch": 0.4887246783029685, "grad_norm": 1.4852194786071777, "learning_rate": 5.112130479102957e-06, "loss": 0.0005, "step": 1918 }, { "epoch": 0.48897948783284495, "grad_norm": 2.3501126766204834, "learning_rate": 5.109582059123344e-06, "loss": 0.0007, "step": 1919 }, { "epoch": 0.4892342973627214, "grad_norm": 0.3805984854698181, "learning_rate": 5.1070336391437315e-06, "loss": 0.0002, "step": 1920 }, { "epoch": 0.4894891068925978, "grad_norm": 2.7900705337524414, "learning_rate": 5.104485219164119e-06, "loss": 0.0008, "step": 1921 }, { "epoch": 0.4897439164224742, "grad_norm": 1.298135757446289, "learning_rate": 5.101936799184506e-06, "loss": 0.0003, "step": 1922 }, { "epoch": 0.4899987259523506, "grad_norm": 1.4971226453781128, "learning_rate": 5.099388379204894e-06, "loss": 0.0004, "step": 1923 }, { "epoch": 0.49025353548222705, "grad_norm": 1.1257603168487549, "learning_rate": 5.09683995922528e-06, "loss": 0.0003, "step": 1924 }, { "epoch": 0.49050834501210344, "grad_norm": 0.9541558623313904, "learning_rate": 5.094291539245668e-06, "loss": 0.0005, "step": 1925 }, { "epoch": 0.4907631545419799, "grad_norm": 2.7082555294036865, "learning_rate": 5.091743119266055e-06, "loss": 0.0012, "step": 1926 }, { "epoch": 0.49101796407185627, "grad_norm": 0.46790364384651184, "learning_rate": 5.089194699286443e-06, "loss": 0.0006, "step": 1927 }, { "epoch": 0.4912727736017327, "grad_norm": 0.9636201858520508, "learning_rate": 5.0866462793068305e-06, "loss": 0.0006, "step": 1928 }, { "epoch": 0.4915275831316091, "grad_norm": 0.41990524530410767, "learning_rate": 5.084097859327217e-06, "loss": 0.0003, "step": 1929 }, { "epoch": 0.49178239266148555, "grad_norm": 57.345298767089844, "learning_rate": 5.081549439347605e-06, "loss": 0.6449, "step": 1930 }, { "epoch": 0.49203720219136193, "grad_norm": 0.3780934810638428, "learning_rate": 5.079001019367991e-06, "loss": 0.0002, "step": 1931 }, { "epoch": 0.4922920117212384, "grad_norm": 2.420551061630249, "learning_rate": 5.076452599388379e-06, "loss": 0.0008, "step": 1932 }, { "epoch": 0.4925468212511148, "grad_norm": 2.426827907562256, "learning_rate": 5.073904179408767e-06, "loss": 0.0006, "step": 1933 }, { "epoch": 0.4928016307809912, "grad_norm": 0.9302656054496765, "learning_rate": 5.071355759429154e-06, "loss": 0.0003, "step": 1934 }, { "epoch": 0.49305644031086765, "grad_norm": 2.449612855911255, "learning_rate": 5.068807339449542e-06, "loss": 0.0008, "step": 1935 }, { "epoch": 0.49331124984074404, "grad_norm": 2.415536642074585, "learning_rate": 5.066258919469929e-06, "loss": 0.0007, "step": 1936 }, { "epoch": 0.4935660593706205, "grad_norm": 0.7524986267089844, "learning_rate": 5.0637104994903164e-06, "loss": 0.0003, "step": 1937 }, { "epoch": 0.49382086890049687, "grad_norm": 6.301401615142822, "learning_rate": 5.061162079510704e-06, "loss": 0.003, "step": 1938 }, { "epoch": 0.4940756784303733, "grad_norm": 0.3019289970397949, "learning_rate": 5.058613659531091e-06, "loss": 0.0004, "step": 1939 }, { "epoch": 0.4943304879602497, "grad_norm": 1.133628487586975, "learning_rate": 5.056065239551479e-06, "loss": 0.0004, "step": 1940 }, { "epoch": 0.49458529749012614, "grad_norm": 0.1894025057554245, "learning_rate": 5.053516819571865e-06, "loss": 0.0002, "step": 1941 }, { "epoch": 0.49484010702000253, "grad_norm": 1.5067275762557983, "learning_rate": 5.050968399592253e-06, "loss": 0.0005, "step": 1942 }, { "epoch": 0.49509491654987897, "grad_norm": 0.3758997321128845, "learning_rate": 5.048419979612641e-06, "loss": 0.0002, "step": 1943 }, { "epoch": 0.49534972607975536, "grad_norm": 0.3725154399871826, "learning_rate": 5.045871559633028e-06, "loss": 0.0002, "step": 1944 }, { "epoch": 0.4956045356096318, "grad_norm": 3.712613105773926, "learning_rate": 5.0433231396534154e-06, "loss": 0.0012, "step": 1945 }, { "epoch": 0.49585934513950825, "grad_norm": 4.487174987792969, "learning_rate": 5.040774719673802e-06, "loss": 0.0017, "step": 1946 }, { "epoch": 0.49611415466938463, "grad_norm": 5.396717071533203, "learning_rate": 5.03822629969419e-06, "loss": 0.0023, "step": 1947 }, { "epoch": 0.4963689641992611, "grad_norm": 4.289586067199707, "learning_rate": 5.035677879714578e-06, "loss": 0.0014, "step": 1948 }, { "epoch": 0.49662377372913746, "grad_norm": 2.851330518722534, "learning_rate": 5.033129459734964e-06, "loss": 0.002, "step": 1949 }, { "epoch": 0.4968785832590139, "grad_norm": 0.7889652252197266, "learning_rate": 5.030581039755352e-06, "loss": 0.0007, "step": 1950 }, { "epoch": 0.4971333927888903, "grad_norm": 0.19827410578727722, "learning_rate": 5.028032619775739e-06, "loss": 0.0002, "step": 1951 }, { "epoch": 0.49738820231876674, "grad_norm": 2.7954399585723877, "learning_rate": 5.025484199796127e-06, "loss": 0.0009, "step": 1952 }, { "epoch": 0.4976430118486431, "grad_norm": 1.699053406715393, "learning_rate": 5.0229357798165144e-06, "loss": 0.0006, "step": 1953 }, { "epoch": 0.49789782137851957, "grad_norm": 0.19835276901721954, "learning_rate": 5.020387359836901e-06, "loss": 0.0001, "step": 1954 }, { "epoch": 0.49815263090839595, "grad_norm": 0.9320805668830872, "learning_rate": 5.017838939857289e-06, "loss": 0.0002, "step": 1955 }, { "epoch": 0.4984074404382724, "grad_norm": 0.7726566791534424, "learning_rate": 5.015290519877676e-06, "loss": 0.0005, "step": 1956 }, { "epoch": 0.4986622499681488, "grad_norm": 0.5069624781608582, "learning_rate": 5.012742099898064e-06, "loss": 0.0006, "step": 1957 }, { "epoch": 0.49891705949802523, "grad_norm": 3.535125494003296, "learning_rate": 5.010193679918452e-06, "loss": 0.0011, "step": 1958 }, { "epoch": 0.49917186902790167, "grad_norm": 2.5980911254882812, "learning_rate": 5.007645259938838e-06, "loss": 0.0006, "step": 1959 }, { "epoch": 0.49942667855777806, "grad_norm": 0.19417963922023773, "learning_rate": 5.005096839959226e-06, "loss": 0.0001, "step": 1960 }, { "epoch": 0.4996814880876545, "grad_norm": 0.5637388825416565, "learning_rate": 5.002548419979613e-06, "loss": 0.0002, "step": 1961 }, { "epoch": 0.4999362976175309, "grad_norm": 7.2049736976623535, "learning_rate": 5e-06, "loss": 0.0045, "step": 1962 }, { "epoch": 0.5001911071474073, "grad_norm": 1.679073691368103, "learning_rate": 4.997451580020387e-06, "loss": 0.0003, "step": 1963 }, { "epoch": 0.5004459166772838, "grad_norm": 1.5048179626464844, "learning_rate": 4.994903160040775e-06, "loss": 0.0017, "step": 1964 }, { "epoch": 0.5007007262071601, "grad_norm": 2.9840424060821533, "learning_rate": 4.992354740061163e-06, "loss": 0.0011, "step": 1965 }, { "epoch": 0.5009555357370366, "grad_norm": 3.335458993911743, "learning_rate": 4.98980632008155e-06, "loss": 0.0009, "step": 1966 }, { "epoch": 0.501210345266913, "grad_norm": 0.9266287088394165, "learning_rate": 4.987257900101937e-06, "loss": 0.0002, "step": 1967 }, { "epoch": 0.5014651547967894, "grad_norm": 2.0393118858337402, "learning_rate": 4.984709480122325e-06, "loss": 0.0004, "step": 1968 }, { "epoch": 0.5017199643266658, "grad_norm": 0.3696623742580414, "learning_rate": 4.982161060142712e-06, "loss": 0.0001, "step": 1969 }, { "epoch": 0.5019747738565422, "grad_norm": 61.7398681640625, "learning_rate": 4.979612640163099e-06, "loss": 0.0228, "step": 1970 }, { "epoch": 0.5022295833864187, "grad_norm": 0.6225137710571289, "learning_rate": 4.977064220183486e-06, "loss": 0.0005, "step": 1971 }, { "epoch": 0.5024843929162951, "grad_norm": 2.0583667755126953, "learning_rate": 4.974515800203874e-06, "loss": 0.0008, "step": 1972 }, { "epoch": 0.5027392024461715, "grad_norm": 0.962236225605011, "learning_rate": 4.971967380224261e-06, "loss": 0.0004, "step": 1973 }, { "epoch": 0.5029940119760479, "grad_norm": 1.9414763450622559, "learning_rate": 4.969418960244649e-06, "loss": 0.0015, "step": 1974 }, { "epoch": 0.5032488215059243, "grad_norm": 2.0413289070129395, "learning_rate": 4.966870540265037e-06, "loss": 0.0004, "step": 1975 }, { "epoch": 0.5035036310358008, "grad_norm": 7.40861177444458, "learning_rate": 4.964322120285424e-06, "loss": 0.004, "step": 1976 }, { "epoch": 0.5037584405656772, "grad_norm": 9.413419723510742, "learning_rate": 4.961773700305811e-06, "loss": 0.0054, "step": 1977 }, { "epoch": 0.5040132500955535, "grad_norm": 44.27852249145508, "learning_rate": 4.9592252803261984e-06, "loss": 0.1124, "step": 1978 }, { "epoch": 0.50426805962543, "grad_norm": 1.6779927015304565, "learning_rate": 4.956676860346585e-06, "loss": 0.0003, "step": 1979 }, { "epoch": 0.5045228691553064, "grad_norm": 99.8744888305664, "learning_rate": 4.954128440366973e-06, "loss": 4.5628, "step": 1980 }, { "epoch": 0.5047776786851829, "grad_norm": 2.242044687271118, "learning_rate": 4.95158002038736e-06, "loss": 0.0005, "step": 1981 }, { "epoch": 0.5050324882150592, "grad_norm": 0.2653028666973114, "learning_rate": 4.949031600407748e-06, "loss": 0.0003, "step": 1982 }, { "epoch": 0.5052872977449356, "grad_norm": 2.615671396255493, "learning_rate": 4.946483180428135e-06, "loss": 0.001, "step": 1983 }, { "epoch": 0.5055421072748121, "grad_norm": 0.9290345907211304, "learning_rate": 4.943934760448522e-06, "loss": 0.0002, "step": 1984 }, { "epoch": 0.5057969168046885, "grad_norm": 2.3978826999664307, "learning_rate": 4.94138634046891e-06, "loss": 0.0006, "step": 1985 }, { "epoch": 0.506051726334565, "grad_norm": 1.165925145149231, "learning_rate": 4.9388379204892974e-06, "loss": 0.0009, "step": 1986 }, { "epoch": 0.5063065358644413, "grad_norm": 3.253103256225586, "learning_rate": 4.936289500509684e-06, "loss": 0.0008, "step": 1987 }, { "epoch": 0.5065613453943177, "grad_norm": 0.12025487422943115, "learning_rate": 4.933741080530071e-06, "loss": 0.0003, "step": 1988 }, { "epoch": 0.5068161549241942, "grad_norm": 7.433753490447998, "learning_rate": 4.931192660550459e-06, "loss": 0.0046, "step": 1989 }, { "epoch": 0.5070709644540706, "grad_norm": 4.7298383712768555, "learning_rate": 4.928644240570846e-06, "loss": 0.0055, "step": 1990 }, { "epoch": 0.507325773983947, "grad_norm": 0.64979487657547, "learning_rate": 4.926095820591234e-06, "loss": 0.0006, "step": 1991 }, { "epoch": 0.5075805835138234, "grad_norm": 2.790837287902832, "learning_rate": 4.923547400611622e-06, "loss": 0.0008, "step": 1992 }, { "epoch": 0.5078353930436998, "grad_norm": 2.0388126373291016, "learning_rate": 4.920998980632009e-06, "loss": 0.0005, "step": 1993 }, { "epoch": 0.5080902025735763, "grad_norm": 1.5009732246398926, "learning_rate": 4.918450560652396e-06, "loss": 0.0005, "step": 1994 }, { "epoch": 0.5083450121034526, "grad_norm": 0.862346351146698, "learning_rate": 4.915902140672783e-06, "loss": 0.0006, "step": 1995 }, { "epoch": 0.5085998216333291, "grad_norm": 0.3927450180053711, "learning_rate": 4.913353720693171e-06, "loss": 0.0003, "step": 1996 }, { "epoch": 0.5088546311632055, "grad_norm": 1.6754134893417358, "learning_rate": 4.910805300713558e-06, "loss": 0.0005, "step": 1997 }, { "epoch": 0.509109440693082, "grad_norm": 0.6678480505943298, "learning_rate": 4.908256880733945e-06, "loss": 0.0007, "step": 1998 }, { "epoch": 0.5093642502229584, "grad_norm": 7.111430644989014, "learning_rate": 4.905708460754333e-06, "loss": 0.0039, "step": 1999 }, { "epoch": 0.5096190597528347, "grad_norm": 4.5647873878479, "learning_rate": 4.90316004077472e-06, "loss": 0.0017, "step": 2000 }, { "epoch": 0.5098738692827112, "grad_norm": 0.4081718325614929, "learning_rate": 4.900611620795108e-06, "loss": 0.0004, "step": 2001 }, { "epoch": 0.5101286788125876, "grad_norm": 2.014273166656494, "learning_rate": 4.898063200815495e-06, "loss": 0.0005, "step": 2002 }, { "epoch": 0.510383488342464, "grad_norm": 0.3960637152194977, "learning_rate": 4.895514780835882e-06, "loss": 0.0003, "step": 2003 }, { "epoch": 0.5106382978723404, "grad_norm": 0.1975594013929367, "learning_rate": 4.892966360856269e-06, "loss": 0.0002, "step": 2004 }, { "epoch": 0.5108931074022168, "grad_norm": 1.306412935256958, "learning_rate": 4.890417940876656e-06, "loss": 0.0004, "step": 2005 }, { "epoch": 0.5111479169320933, "grad_norm": 700.3057861328125, "learning_rate": 4.887869520897044e-06, "loss": 0.9143, "step": 2006 }, { "epoch": 0.5114027264619697, "grad_norm": 2.22855806350708, "learning_rate": 4.885321100917432e-06, "loss": 0.0006, "step": 2007 }, { "epoch": 0.511657535991846, "grad_norm": 4.6322712898254395, "learning_rate": 4.882772680937819e-06, "loss": 0.0017, "step": 2008 }, { "epoch": 0.5119123455217225, "grad_norm": 2.0384812355041504, "learning_rate": 4.880224260958206e-06, "loss": 0.0005, "step": 2009 }, { "epoch": 0.5121671550515989, "grad_norm": 0.48533859848976135, "learning_rate": 4.877675840978594e-06, "loss": 0.0005, "step": 2010 }, { "epoch": 0.5124219645814754, "grad_norm": 0.4003552794456482, "learning_rate": 4.875127420998981e-06, "loss": 0.0003, "step": 2011 }, { "epoch": 0.5126767741113518, "grad_norm": 2.3445980548858643, "learning_rate": 4.872579001019368e-06, "loss": 0.0009, "step": 2012 }, { "epoch": 0.5129315836412282, "grad_norm": 0.7376776933670044, "learning_rate": 4.870030581039756e-06, "loss": 0.0002, "step": 2013 }, { "epoch": 0.5131863931711046, "grad_norm": 3.894028902053833, "learning_rate": 4.867482161060143e-06, "loss": 0.0013, "step": 2014 }, { "epoch": 0.513441202700981, "grad_norm": 0.1944580078125, "learning_rate": 4.86493374108053e-06, "loss": 0.0002, "step": 2015 }, { "epoch": 0.5136960122308575, "grad_norm": 1.47898268699646, "learning_rate": 4.862385321100918e-06, "loss": 0.0003, "step": 2016 }, { "epoch": 0.5139508217607338, "grad_norm": 1.304489254951477, "learning_rate": 4.859836901121306e-06, "loss": 0.0003, "step": 2017 }, { "epoch": 0.5142056312906103, "grad_norm": 0.24402308464050293, "learning_rate": 4.857288481141693e-06, "loss": 0.0005, "step": 2018 }, { "epoch": 0.5144604408204867, "grad_norm": 0.575560450553894, "learning_rate": 4.8547400611620796e-06, "loss": 0.0003, "step": 2019 }, { "epoch": 0.5147152503503631, "grad_norm": 886.7788696289062, "learning_rate": 4.852191641182467e-06, "loss": 3.9377, "step": 2020 }, { "epoch": 0.5149700598802395, "grad_norm": 0.4525306224822998, "learning_rate": 4.849643221202854e-06, "loss": 0.0005, "step": 2021 }, { "epoch": 0.5152248694101159, "grad_norm": 1.4240411520004272, "learning_rate": 4.847094801223242e-06, "loss": 0.0023, "step": 2022 }, { "epoch": 0.5154796789399924, "grad_norm": 1.329408049583435, "learning_rate": 4.844546381243629e-06, "loss": 0.0005, "step": 2023 }, { "epoch": 0.5157344884698688, "grad_norm": 0.07646592706441879, "learning_rate": 4.841997961264017e-06, "loss": 0.0002, "step": 2024 }, { "epoch": 0.5159892979997452, "grad_norm": 1.1182199716567993, "learning_rate": 4.839449541284404e-06, "loss": 0.0003, "step": 2025 }, { "epoch": 0.5162441075296216, "grad_norm": 2.242506742477417, "learning_rate": 4.836901121304791e-06, "loss": 0.0007, "step": 2026 }, { "epoch": 0.516498917059498, "grad_norm": 1.688819408416748, "learning_rate": 4.834352701325179e-06, "loss": 0.0006, "step": 2027 }, { "epoch": 0.5167537265893745, "grad_norm": 2.7785444259643555, "learning_rate": 4.831804281345566e-06, "loss": 0.0009, "step": 2028 }, { "epoch": 0.5170085361192509, "grad_norm": 2.595635414123535, "learning_rate": 4.829255861365953e-06, "loss": 0.0006, "step": 2029 }, { "epoch": 0.5172633456491272, "grad_norm": 0.4357888996601105, "learning_rate": 4.826707441386341e-06, "loss": 0.0004, "step": 2030 }, { "epoch": 0.5175181551790037, "grad_norm": 2.96460223197937, "learning_rate": 4.824159021406728e-06, "loss": 0.0007, "step": 2031 }, { "epoch": 0.5177729647088801, "grad_norm": 0.3751264810562134, "learning_rate": 4.821610601427116e-06, "loss": 0.0001, "step": 2032 }, { "epoch": 0.5180277742387566, "grad_norm": 5.690130710601807, "learning_rate": 4.819062181447503e-06, "loss": 0.0021, "step": 2033 }, { "epoch": 0.5182825837686329, "grad_norm": 3.552065849304199, "learning_rate": 4.816513761467891e-06, "loss": 0.001, "step": 2034 }, { "epoch": 0.5185373932985093, "grad_norm": 1.8493150472640991, "learning_rate": 4.813965341488278e-06, "loss": 0.0004, "step": 2035 }, { "epoch": 0.5187922028283858, "grad_norm": 0.45771241188049316, "learning_rate": 4.8114169215086645e-06, "loss": 0.0008, "step": 2036 }, { "epoch": 0.5190470123582622, "grad_norm": 0.5698198080062866, "learning_rate": 4.808868501529052e-06, "loss": 0.0002, "step": 2037 }, { "epoch": 0.5193018218881387, "grad_norm": 6.495179653167725, "learning_rate": 4.80632008154944e-06, "loss": 0.0032, "step": 2038 }, { "epoch": 0.519556631418015, "grad_norm": 2.7803425788879395, "learning_rate": 4.803771661569827e-06, "loss": 0.0007, "step": 2039 }, { "epoch": 0.5198114409478914, "grad_norm": 1.573736310005188, "learning_rate": 4.801223241590214e-06, "loss": 0.0017, "step": 2040 }, { "epoch": 0.5200662504777679, "grad_norm": 5.195489406585693, "learning_rate": 4.798674821610602e-06, "loss": 0.002, "step": 2041 }, { "epoch": 0.5203210600076443, "grad_norm": 1.4930917024612427, "learning_rate": 4.79612640163099e-06, "loss": 0.0003, "step": 2042 }, { "epoch": 0.5205758695375207, "grad_norm": 3.1579320430755615, "learning_rate": 4.793577981651377e-06, "loss": 0.0009, "step": 2043 }, { "epoch": 0.5208306790673971, "grad_norm": 4.299086570739746, "learning_rate": 4.7910295616717635e-06, "loss": 0.0014, "step": 2044 }, { "epoch": 0.5210854885972735, "grad_norm": 6.7569050788879395, "learning_rate": 4.788481141692151e-06, "loss": 0.0036, "step": 2045 }, { "epoch": 0.52134029812715, "grad_norm": 3.3590805530548096, "learning_rate": 4.785932721712538e-06, "loss": 0.001, "step": 2046 }, { "epoch": 0.5215951076570263, "grad_norm": 4.291635990142822, "learning_rate": 4.783384301732926e-06, "loss": 0.0015, "step": 2047 }, { "epoch": 0.5218499171869028, "grad_norm": 1.3137575387954712, "learning_rate": 4.780835881753314e-06, "loss": 0.0003, "step": 2048 }, { "epoch": 0.5221047267167792, "grad_norm": 0.37419018149375916, "learning_rate": 4.778287461773701e-06, "loss": 0.0001, "step": 2049 }, { "epoch": 0.5223595362466557, "grad_norm": 0.22756938636302948, "learning_rate": 4.775739041794088e-06, "loss": 0.0002, "step": 2050 }, { "epoch": 0.5226143457765321, "grad_norm": 2.2536797523498535, "learning_rate": 4.773190621814476e-06, "loss": 0.0008, "step": 2051 }, { "epoch": 0.5228691553064084, "grad_norm": 2.7589664459228516, "learning_rate": 4.770642201834863e-06, "loss": 0.0007, "step": 2052 }, { "epoch": 0.5231239648362849, "grad_norm": 0.19846907258033752, "learning_rate": 4.76809378185525e-06, "loss": 0.0001, "step": 2053 }, { "epoch": 0.5233787743661613, "grad_norm": 1.334179162979126, "learning_rate": 4.765545361875637e-06, "loss": 0.0008, "step": 2054 }, { "epoch": 0.5236335838960378, "grad_norm": 1.0168486833572388, "learning_rate": 4.762996941896025e-06, "loss": 0.0008, "step": 2055 }, { "epoch": 0.5238883934259141, "grad_norm": 4.057284832000732, "learning_rate": 4.760448521916412e-06, "loss": 0.0013, "step": 2056 }, { "epoch": 0.5241432029557905, "grad_norm": 4.646595001220703, "learning_rate": 4.757900101936799e-06, "loss": 0.0018, "step": 2057 }, { "epoch": 0.524398012485667, "grad_norm": 1.5021088123321533, "learning_rate": 4.755351681957187e-06, "loss": 0.0003, "step": 2058 }, { "epoch": 0.5246528220155434, "grad_norm": 3.1355948448181152, "learning_rate": 4.752803261977575e-06, "loss": 0.0008, "step": 2059 }, { "epoch": 0.5249076315454198, "grad_norm": 1.4995309114456177, "learning_rate": 4.7502548419979616e-06, "loss": 0.0003, "step": 2060 }, { "epoch": 0.5251624410752962, "grad_norm": 103.95832061767578, "learning_rate": 4.7477064220183485e-06, "loss": 0.6016, "step": 2061 }, { "epoch": 0.5254172506051726, "grad_norm": 113.96194458007812, "learning_rate": 4.745158002038736e-06, "loss": 0.7266, "step": 2062 }, { "epoch": 0.5256720601350491, "grad_norm": 1.6926488876342773, "learning_rate": 4.742609582059124e-06, "loss": 0.0006, "step": 2063 }, { "epoch": 0.5259268696649255, "grad_norm": 2.4197802543640137, "learning_rate": 4.740061162079511e-06, "loss": 0.0005, "step": 2064 }, { "epoch": 0.5261816791948019, "grad_norm": 0.29397743940353394, "learning_rate": 4.737512742099898e-06, "loss": 0.0006, "step": 2065 }, { "epoch": 0.5264364887246783, "grad_norm": 0.7479109168052673, "learning_rate": 4.734964322120286e-06, "loss": 0.0002, "step": 2066 }, { "epoch": 0.5266912982545547, "grad_norm": 0.7608151435852051, "learning_rate": 4.732415902140673e-06, "loss": 0.0003, "step": 2067 }, { "epoch": 0.5269461077844312, "grad_norm": 3.3613014221191406, "learning_rate": 4.7298674821610606e-06, "loss": 0.0011, "step": 2068 }, { "epoch": 0.5272009173143075, "grad_norm": 0.4046708047389984, "learning_rate": 4.727319062181448e-06, "loss": 0.0003, "step": 2069 }, { "epoch": 0.527455726844184, "grad_norm": 0.08232639729976654, "learning_rate": 4.724770642201835e-06, "loss": 0.0001, "step": 2070 }, { "epoch": 0.5277105363740604, "grad_norm": 0.9303514361381531, "learning_rate": 4.722222222222222e-06, "loss": 0.0002, "step": 2071 }, { "epoch": 0.5279653459039368, "grad_norm": 1.1146578788757324, "learning_rate": 4.71967380224261e-06, "loss": 0.0002, "step": 2072 }, { "epoch": 0.5282201554338132, "grad_norm": 2.036505699157715, "learning_rate": 4.717125382262998e-06, "loss": 0.0005, "step": 2073 }, { "epoch": 0.5284749649636896, "grad_norm": 0.4686679244041443, "learning_rate": 4.714576962283385e-06, "loss": 0.0005, "step": 2074 }, { "epoch": 0.5287297744935661, "grad_norm": 0.20530755817890167, "learning_rate": 4.712028542303772e-06, "loss": 0.0002, "step": 2075 }, { "epoch": 0.5289845840234425, "grad_norm": 2.035407781600952, "learning_rate": 4.70948012232416e-06, "loss": 0.0005, "step": 2076 }, { "epoch": 0.529239393553319, "grad_norm": 0.2493685632944107, "learning_rate": 4.7069317023445465e-06, "loss": 0.0002, "step": 2077 }, { "epoch": 0.5294942030831953, "grad_norm": 0.8635119199752808, "learning_rate": 4.704383282364934e-06, "loss": 0.0008, "step": 2078 }, { "epoch": 0.5297490126130717, "grad_norm": 56.41252899169922, "learning_rate": 4.701834862385321e-06, "loss": 0.6114, "step": 2079 }, { "epoch": 0.5300038221429482, "grad_norm": 2.4202818870544434, "learning_rate": 4.699286442405709e-06, "loss": 0.0007, "step": 2080 }, { "epoch": 0.5302586316728246, "grad_norm": 2.6099660396575928, "learning_rate": 4.696738022426096e-06, "loss": 0.0006, "step": 2081 }, { "epoch": 0.5305134412027009, "grad_norm": 2.9957082271575928, "learning_rate": 4.694189602446483e-06, "loss": 0.0011, "step": 2082 }, { "epoch": 0.5307682507325774, "grad_norm": 2.2293829917907715, "learning_rate": 4.691641182466871e-06, "loss": 0.0006, "step": 2083 }, { "epoch": 0.5310230602624538, "grad_norm": 3.742457628250122, "learning_rate": 4.689092762487259e-06, "loss": 0.0016, "step": 2084 }, { "epoch": 0.5312778697923303, "grad_norm": 2.9808499813079834, "learning_rate": 4.6865443425076455e-06, "loss": 0.0008, "step": 2085 }, { "epoch": 0.5315326793222066, "grad_norm": 4.80433464050293, "learning_rate": 4.683995922528033e-06, "loss": 0.0018, "step": 2086 }, { "epoch": 0.531787488852083, "grad_norm": 4.638812065124512, "learning_rate": 4.68144750254842e-06, "loss": 0.0017, "step": 2087 }, { "epoch": 0.5320422983819595, "grad_norm": 0.04327725991606712, "learning_rate": 4.678899082568808e-06, "loss": 0.0001, "step": 2088 }, { "epoch": 0.5322971079118359, "grad_norm": 4.888421058654785, "learning_rate": 4.676350662589195e-06, "loss": 0.0019, "step": 2089 }, { "epoch": 0.5325519174417124, "grad_norm": 0.566041111946106, "learning_rate": 4.673802242609583e-06, "loss": 0.0002, "step": 2090 }, { "epoch": 0.5328067269715887, "grad_norm": 3.887883424758911, "learning_rate": 4.67125382262997e-06, "loss": 0.0014, "step": 2091 }, { "epoch": 0.5330615365014651, "grad_norm": 2.043649673461914, "learning_rate": 4.668705402650357e-06, "loss": 0.0004, "step": 2092 }, { "epoch": 0.5333163460313416, "grad_norm": 2.233943223953247, "learning_rate": 4.6661569826707446e-06, "loss": 0.0006, "step": 2093 }, { "epoch": 0.533571155561218, "grad_norm": 2.4147796630859375, "learning_rate": 4.663608562691132e-06, "loss": 0.0008, "step": 2094 }, { "epoch": 0.5338259650910944, "grad_norm": 2.036471366882324, "learning_rate": 4.661060142711519e-06, "loss": 0.0004, "step": 2095 }, { "epoch": 0.5340807746209708, "grad_norm": 2.2328858375549316, "learning_rate": 4.658511722731906e-06, "loss": 0.0005, "step": 2096 }, { "epoch": 0.5343355841508473, "grad_norm": 1.6710283756256104, "learning_rate": 4.655963302752294e-06, "loss": 0.0005, "step": 2097 }, { "epoch": 0.5345903936807237, "grad_norm": 5.391232013702393, "learning_rate": 4.653414882772681e-06, "loss": 0.0022, "step": 2098 }, { "epoch": 0.5348452032106, "grad_norm": 3.460827589035034, "learning_rate": 4.650866462793069e-06, "loss": 0.001, "step": 2099 }, { "epoch": 0.5351000127404765, "grad_norm": 377.0355224609375, "learning_rate": 4.648318042813456e-06, "loss": 0.5633, "step": 2100 }, { "epoch": 0.5353548222703529, "grad_norm": 0.5595435500144958, "learning_rate": 4.6457696228338436e-06, "loss": 0.0002, "step": 2101 }, { "epoch": 0.5356096318002294, "grad_norm": 1.296982765197754, "learning_rate": 4.6432212028542305e-06, "loss": 0.0002, "step": 2102 }, { "epoch": 0.5358644413301058, "grad_norm": 2.9686639308929443, "learning_rate": 4.6406727828746175e-06, "loss": 0.0008, "step": 2103 }, { "epoch": 0.5361192508599821, "grad_norm": 3.553356409072876, "learning_rate": 4.638124362895005e-06, "loss": 0.0016, "step": 2104 }, { "epoch": 0.5363740603898586, "grad_norm": 0.43943771719932556, "learning_rate": 4.635575942915393e-06, "loss": 0.0004, "step": 2105 }, { "epoch": 0.536628869919735, "grad_norm": 1.4844383001327515, "learning_rate": 4.63302752293578e-06, "loss": 0.0003, "step": 2106 }, { "epoch": 0.5368836794496115, "grad_norm": 5.185815334320068, "learning_rate": 4.630479102956168e-06, "loss": 0.0022, "step": 2107 }, { "epoch": 0.5371384889794878, "grad_norm": 4.282326698303223, "learning_rate": 4.627930682976555e-06, "loss": 0.0017, "step": 2108 }, { "epoch": 0.5373932985093642, "grad_norm": 3.6978988647460938, "learning_rate": 4.6253822629969426e-06, "loss": 0.0012, "step": 2109 }, { "epoch": 0.5376481080392407, "grad_norm": 3.3324716091156006, "learning_rate": 4.6228338430173295e-06, "loss": 0.001, "step": 2110 }, { "epoch": 0.5379029175691171, "grad_norm": 0.7542738914489746, "learning_rate": 4.620285423037717e-06, "loss": 0.0002, "step": 2111 }, { "epoch": 0.5381577270989935, "grad_norm": 2.7903409004211426, "learning_rate": 4.617737003058104e-06, "loss": 0.0007, "step": 2112 }, { "epoch": 0.5384125366288699, "grad_norm": 4.27747106552124, "learning_rate": 4.615188583078491e-06, "loss": 0.0021, "step": 2113 }, { "epoch": 0.5386673461587463, "grad_norm": 5.944787502288818, "learning_rate": 4.612640163098879e-06, "loss": 0.0026, "step": 2114 }, { "epoch": 0.5389221556886228, "grad_norm": 2.9644203186035156, "learning_rate": 4.610091743119267e-06, "loss": 0.0008, "step": 2115 }, { "epoch": 0.5391769652184992, "grad_norm": 2.2208988666534424, "learning_rate": 4.607543323139654e-06, "loss": 0.0005, "step": 2116 }, { "epoch": 0.5394317747483756, "grad_norm": 2.0289628505706787, "learning_rate": 4.604994903160041e-06, "loss": 0.0012, "step": 2117 }, { "epoch": 0.539686584278252, "grad_norm": 2.4704842567443848, "learning_rate": 4.6024464831804285e-06, "loss": 0.002, "step": 2118 }, { "epoch": 0.5399413938081284, "grad_norm": 4.8101701736450195, "learning_rate": 4.599898063200816e-06, "loss": 0.0032, "step": 2119 }, { "epoch": 0.5401962033380049, "grad_norm": 8.497100830078125, "learning_rate": 4.597349643221203e-06, "loss": 0.0055, "step": 2120 }, { "epoch": 0.5404510128678812, "grad_norm": 6.4519500732421875, "learning_rate": 4.59480122324159e-06, "loss": 0.0031, "step": 2121 }, { "epoch": 0.5407058223977577, "grad_norm": 9.26585865020752, "learning_rate": 4.592252803261978e-06, "loss": 0.0062, "step": 2122 }, { "epoch": 0.5409606319276341, "grad_norm": 7.974853038787842, "learning_rate": 4.589704383282365e-06, "loss": 0.0048, "step": 2123 }, { "epoch": 0.5412154414575105, "grad_norm": 3.7047524452209473, "learning_rate": 4.587155963302753e-06, "loss": 0.0011, "step": 2124 }, { "epoch": 0.5414702509873869, "grad_norm": 1.8634837865829468, "learning_rate": 4.584607543323141e-06, "loss": 0.0006, "step": 2125 }, { "epoch": 0.5417250605172633, "grad_norm": 3.4326705932617188, "learning_rate": 4.5820591233435275e-06, "loss": 0.0016, "step": 2126 }, { "epoch": 0.5419798700471398, "grad_norm": 1.6715738773345947, "learning_rate": 4.5795107033639145e-06, "loss": 0.0003, "step": 2127 }, { "epoch": 0.5422346795770162, "grad_norm": 1.1205034255981445, "learning_rate": 4.576962283384302e-06, "loss": 0.0003, "step": 2128 }, { "epoch": 0.5424894891068927, "grad_norm": 6.486885070800781, "learning_rate": 4.57441386340469e-06, "loss": 0.0037, "step": 2129 }, { "epoch": 0.542744298636769, "grad_norm": 2.9709038734436035, "learning_rate": 4.571865443425077e-06, "loss": 0.0008, "step": 2130 }, { "epoch": 0.5429991081666454, "grad_norm": 1.4817205667495728, "learning_rate": 4.569317023445464e-06, "loss": 0.0003, "step": 2131 }, { "epoch": 0.5432539176965219, "grad_norm": 1.6749323606491089, "learning_rate": 4.566768603465852e-06, "loss": 0.0004, "step": 2132 }, { "epoch": 0.5435087272263983, "grad_norm": 2.790853977203369, "learning_rate": 4.564220183486239e-06, "loss": 0.0011, "step": 2133 }, { "epoch": 0.5437635367562746, "grad_norm": 2.5927655696868896, "learning_rate": 4.561671763506626e-06, "loss": 0.0007, "step": 2134 }, { "epoch": 0.5440183462861511, "grad_norm": 2.227834463119507, "learning_rate": 4.5591233435270135e-06, "loss": 0.0007, "step": 2135 }, { "epoch": 0.5442731558160275, "grad_norm": 1.2996338605880737, "learning_rate": 4.556574923547401e-06, "loss": 0.0003, "step": 2136 }, { "epoch": 0.544527965345904, "grad_norm": 0.5696894526481628, "learning_rate": 4.554026503567788e-06, "loss": 0.0002, "step": 2137 }, { "epoch": 0.5447827748757803, "grad_norm": 0.7679224610328674, "learning_rate": 4.551478083588175e-06, "loss": 0.0004, "step": 2138 }, { "epoch": 0.5450375844056567, "grad_norm": 1.6790560483932495, "learning_rate": 4.548929663608563e-06, "loss": 0.0004, "step": 2139 }, { "epoch": 0.5452923939355332, "grad_norm": 5.155569076538086, "learning_rate": 4.546381243628951e-06, "loss": 0.003, "step": 2140 }, { "epoch": 0.5455472034654096, "grad_norm": 3.164137363433838, "learning_rate": 4.543832823649338e-06, "loss": 0.0009, "step": 2141 }, { "epoch": 0.5458020129952861, "grad_norm": 2.245990037918091, "learning_rate": 4.541284403669725e-06, "loss": 0.0006, "step": 2142 }, { "epoch": 0.5460568225251624, "grad_norm": 2.035088539123535, "learning_rate": 4.5387359836901125e-06, "loss": 0.0006, "step": 2143 }, { "epoch": 0.5463116320550389, "grad_norm": 3.7098677158355713, "learning_rate": 4.5361875637104995e-06, "loss": 0.0013, "step": 2144 }, { "epoch": 0.5465664415849153, "grad_norm": 3.1632730960845947, "learning_rate": 4.533639143730887e-06, "loss": 0.0012, "step": 2145 }, { "epoch": 0.5468212511147917, "grad_norm": 0.7905943989753723, "learning_rate": 4.531090723751275e-06, "loss": 0.0006, "step": 2146 }, { "epoch": 0.5470760606446681, "grad_norm": 1.1191346645355225, "learning_rate": 4.528542303771662e-06, "loss": 0.0003, "step": 2147 }, { "epoch": 0.5473308701745445, "grad_norm": 4.437350273132324, "learning_rate": 4.525993883792049e-06, "loss": 0.0018, "step": 2148 }, { "epoch": 0.547585679704421, "grad_norm": 5.587113380432129, "learning_rate": 4.523445463812437e-06, "loss": 0.0023, "step": 2149 }, { "epoch": 0.5478404892342974, "grad_norm": 6.526993751525879, "learning_rate": 4.5208970438328246e-06, "loss": 0.0044, "step": 2150 }, { "epoch": 0.5480952987641737, "grad_norm": 5.730679512023926, "learning_rate": 4.5183486238532115e-06, "loss": 0.0025, "step": 2151 }, { "epoch": 0.5483501082940502, "grad_norm": 2.6262145042419434, "learning_rate": 4.5158002038735985e-06, "loss": 0.0011, "step": 2152 }, { "epoch": 0.5486049178239266, "grad_norm": 1.6719919443130493, "learning_rate": 4.513251783893986e-06, "loss": 0.0003, "step": 2153 }, { "epoch": 0.5488597273538031, "grad_norm": 4.778537750244141, "learning_rate": 4.510703363914373e-06, "loss": 0.0019, "step": 2154 }, { "epoch": 0.5491145368836795, "grad_norm": 1.8787422180175781, "learning_rate": 4.508154943934761e-06, "loss": 0.001, "step": 2155 }, { "epoch": 0.5493693464135558, "grad_norm": 8.512125015258789, "learning_rate": 4.505606523955148e-06, "loss": 0.0056, "step": 2156 }, { "epoch": 0.5496241559434323, "grad_norm": 7.241018772125244, "learning_rate": 4.503058103975536e-06, "loss": 0.0038, "step": 2157 }, { "epoch": 0.5498789654733087, "grad_norm": 9.149073600769043, "learning_rate": 4.500509683995923e-06, "loss": 0.007, "step": 2158 }, { "epoch": 0.5501337750031852, "grad_norm": 5.040731430053711, "learning_rate": 4.49796126401631e-06, "loss": 0.0025, "step": 2159 }, { "epoch": 0.5503885845330615, "grad_norm": 5.0077223777771, "learning_rate": 4.4954128440366975e-06, "loss": 0.0024, "step": 2160 }, { "epoch": 0.5506433940629379, "grad_norm": 4.63879919052124, "learning_rate": 4.492864424057085e-06, "loss": 0.002, "step": 2161 }, { "epoch": 0.5508982035928144, "grad_norm": 1.4768898487091064, "learning_rate": 4.490316004077472e-06, "loss": 0.0003, "step": 2162 }, { "epoch": 0.5511530131226908, "grad_norm": 0.11629445850849152, "learning_rate": 4.48776758409786e-06, "loss": 0.0002, "step": 2163 }, { "epoch": 0.5514078226525672, "grad_norm": 7.415570259094238, "learning_rate": 4.485219164118247e-06, "loss": 0.0041, "step": 2164 }, { "epoch": 0.5516626321824436, "grad_norm": 6.103084564208984, "learning_rate": 4.482670744138635e-06, "loss": 0.003, "step": 2165 }, { "epoch": 0.55191744171232, "grad_norm": 7.851794242858887, "learning_rate": 4.480122324159022e-06, "loss": 0.0056, "step": 2166 }, { "epoch": 0.5521722512421965, "grad_norm": 5.927475452423096, "learning_rate": 4.4775739041794095e-06, "loss": 0.0026, "step": 2167 }, { "epoch": 0.5524270607720729, "grad_norm": 7.619122505187988, "learning_rate": 4.4750254841997965e-06, "loss": 0.0043, "step": 2168 }, { "epoch": 0.5526818703019493, "grad_norm": 6.116474151611328, "learning_rate": 4.4724770642201834e-06, "loss": 0.0028, "step": 2169 }, { "epoch": 0.5529366798318257, "grad_norm": 1.2924548387527466, "learning_rate": 4.469928644240571e-06, "loss": 0.0003, "step": 2170 }, { "epoch": 0.5531914893617021, "grad_norm": 0.3791806995868683, "learning_rate": 4.467380224260959e-06, "loss": 0.0007, "step": 2171 }, { "epoch": 0.5534462988915786, "grad_norm": 4.832940101623535, "learning_rate": 4.464831804281346e-06, "loss": 0.0018, "step": 2172 }, { "epoch": 0.5537011084214549, "grad_norm": 2.9749600887298584, "learning_rate": 4.462283384301733e-06, "loss": 0.0007, "step": 2173 }, { "epoch": 0.5539559179513314, "grad_norm": 5.397824287414551, "learning_rate": 4.459734964322121e-06, "loss": 0.0023, "step": 2174 }, { "epoch": 0.5542107274812078, "grad_norm": 4.078503608703613, "learning_rate": 4.457186544342508e-06, "loss": 0.0013, "step": 2175 }, { "epoch": 0.5544655370110843, "grad_norm": 4.635170936584473, "learning_rate": 4.4546381243628955e-06, "loss": 0.0019, "step": 2176 }, { "epoch": 0.5547203465409606, "grad_norm": 0.21522541344165802, "learning_rate": 4.4520897043832824e-06, "loss": 0.0001, "step": 2177 }, { "epoch": 0.554975156070837, "grad_norm": 4.06574010848999, "learning_rate": 4.44954128440367e-06, "loss": 0.0013, "step": 2178 }, { "epoch": 0.5552299656007135, "grad_norm": 2.4247355461120605, "learning_rate": 4.446992864424057e-06, "loss": 0.0006, "step": 2179 }, { "epoch": 0.5554847751305899, "grad_norm": 3.9214885234832764, "learning_rate": 4.444444444444444e-06, "loss": 0.0017, "step": 2180 }, { "epoch": 0.5557395846604662, "grad_norm": 1.3099327087402344, "learning_rate": 4.441896024464832e-06, "loss": 0.0006, "step": 2181 }, { "epoch": 0.5559943941903427, "grad_norm": 1.1128416061401367, "learning_rate": 4.43934760448522e-06, "loss": 0.0002, "step": 2182 }, { "epoch": 0.5562492037202191, "grad_norm": 0.37972885370254517, "learning_rate": 4.436799184505607e-06, "loss": 0.0001, "step": 2183 }, { "epoch": 0.5565040132500956, "grad_norm": 2.4147191047668457, "learning_rate": 4.4342507645259945e-06, "loss": 0.0006, "step": 2184 }, { "epoch": 0.556758822779972, "grad_norm": 3.141909122467041, "learning_rate": 4.4317023445463815e-06, "loss": 0.0008, "step": 2185 }, { "epoch": 0.5570136323098483, "grad_norm": 0.7747337818145752, "learning_rate": 4.429153924566769e-06, "loss": 0.0004, "step": 2186 }, { "epoch": 0.5572684418397248, "grad_norm": 2.0370426177978516, "learning_rate": 4.426605504587156e-06, "loss": 0.0004, "step": 2187 }, { "epoch": 0.5575232513696012, "grad_norm": 6.404489040374756, "learning_rate": 4.424057084607544e-06, "loss": 0.0043, "step": 2188 }, { "epoch": 0.5577780608994777, "grad_norm": 1.1552035808563232, "learning_rate": 4.421508664627931e-06, "loss": 0.0007, "step": 2189 }, { "epoch": 0.558032870429354, "grad_norm": 2.3922691345214844, "learning_rate": 4.418960244648318e-06, "loss": 0.0006, "step": 2190 }, { "epoch": 0.5582876799592305, "grad_norm": 2.7628674507141113, "learning_rate": 4.416411824668706e-06, "loss": 0.0006, "step": 2191 }, { "epoch": 0.5585424894891069, "grad_norm": 3.7111754417419434, "learning_rate": 4.4138634046890935e-06, "loss": 0.0012, "step": 2192 }, { "epoch": 0.5587972990189833, "grad_norm": 4.2451581954956055, "learning_rate": 4.4113149847094805e-06, "loss": 0.0014, "step": 2193 }, { "epoch": 0.5590521085488597, "grad_norm": 2.7735683917999268, "learning_rate": 4.408766564729867e-06, "loss": 0.0007, "step": 2194 }, { "epoch": 0.5593069180787361, "grad_norm": 1.6674529314041138, "learning_rate": 4.406218144750255e-06, "loss": 0.0003, "step": 2195 }, { "epoch": 0.5595617276086126, "grad_norm": 1.1236127614974976, "learning_rate": 4.403669724770643e-06, "loss": 0.0004, "step": 2196 }, { "epoch": 0.559816537138489, "grad_norm": 1.8560782670974731, "learning_rate": 4.40112130479103e-06, "loss": 0.0003, "step": 2197 }, { "epoch": 0.5600713466683654, "grad_norm": 1.1128484010696411, "learning_rate": 4.398572884811417e-06, "loss": 0.0002, "step": 2198 }, { "epoch": 0.5603261561982418, "grad_norm": 4.633818626403809, "learning_rate": 4.396024464831805e-06, "loss": 0.0016, "step": 2199 }, { "epoch": 0.5605809657281182, "grad_norm": 1.1067262887954712, "learning_rate": 4.393476044852192e-06, "loss": 0.0002, "step": 2200 }, { "epoch": 0.5608357752579947, "grad_norm": 1.6753857135772705, "learning_rate": 4.3909276248725795e-06, "loss": 0.0004, "step": 2201 }, { "epoch": 0.5610905847878711, "grad_norm": 6.153477191925049, "learning_rate": 4.388379204892967e-06, "loss": 0.002, "step": 2202 }, { "epoch": 0.5613453943177474, "grad_norm": 4.445742607116699, "learning_rate": 4.385830784913354e-06, "loss": 0.0016, "step": 2203 }, { "epoch": 0.5616002038476239, "grad_norm": 4.0734429359436035, "learning_rate": 4.383282364933741e-06, "loss": 0.0013, "step": 2204 }, { "epoch": 0.5618550133775003, "grad_norm": 5.294697284698486, "learning_rate": 4.380733944954129e-06, "loss": 0.002, "step": 2205 }, { "epoch": 0.5621098229073768, "grad_norm": 2.4139885902404785, "learning_rate": 4.378185524974517e-06, "loss": 0.0005, "step": 2206 }, { "epoch": 0.5623646324372531, "grad_norm": 1.8484218120574951, "learning_rate": 4.375637104994904e-06, "loss": 0.0004, "step": 2207 }, { "epoch": 0.5626194419671295, "grad_norm": 2.2170727252960205, "learning_rate": 4.373088685015291e-06, "loss": 0.0004, "step": 2208 }, { "epoch": 0.562874251497006, "grad_norm": 2.413266658782959, "learning_rate": 4.3705402650356785e-06, "loss": 0.0009, "step": 2209 }, { "epoch": 0.5631290610268824, "grad_norm": 4.079699993133545, "learning_rate": 4.3679918450560654e-06, "loss": 0.0019, "step": 2210 }, { "epoch": 0.5633838705567589, "grad_norm": 4.807590007781982, "learning_rate": 4.365443425076452e-06, "loss": 0.0019, "step": 2211 }, { "epoch": 0.5636386800866352, "grad_norm": 1.309122920036316, "learning_rate": 4.36289500509684e-06, "loss": 0.0005, "step": 2212 }, { "epoch": 0.5638934896165116, "grad_norm": 0.9325226545333862, "learning_rate": 4.360346585117228e-06, "loss": 0.0003, "step": 2213 }, { "epoch": 0.5641482991463881, "grad_norm": 1.4828156232833862, "learning_rate": 4.357798165137615e-06, "loss": 0.0003, "step": 2214 }, { "epoch": 0.5644031086762645, "grad_norm": 0.7536486387252808, "learning_rate": 4.355249745158002e-06, "loss": 0.0003, "step": 2215 }, { "epoch": 0.5646579182061409, "grad_norm": 0.09784235060214996, "learning_rate": 4.35270132517839e-06, "loss": 0.0002, "step": 2216 }, { "epoch": 0.5649127277360173, "grad_norm": 0.5945402383804321, "learning_rate": 4.3501529051987775e-06, "loss": 0.0004, "step": 2217 }, { "epoch": 0.5651675372658937, "grad_norm": 0.5622185468673706, "learning_rate": 4.3476044852191644e-06, "loss": 0.0001, "step": 2218 }, { "epoch": 0.5654223467957702, "grad_norm": 0.05401952192187309, "learning_rate": 4.345056065239551e-06, "loss": 0.0001, "step": 2219 }, { "epoch": 0.5656771563256465, "grad_norm": 2.4015913009643555, "learning_rate": 4.342507645259939e-06, "loss": 0.0006, "step": 2220 }, { "epoch": 0.565931965855523, "grad_norm": 1.6936700344085693, "learning_rate": 4.339959225280326e-06, "loss": 0.0007, "step": 2221 }, { "epoch": 0.5661867753853994, "grad_norm": 0.7544029355049133, "learning_rate": 4.337410805300714e-06, "loss": 0.0001, "step": 2222 }, { "epoch": 0.5664415849152759, "grad_norm": 2.0528478622436523, "learning_rate": 4.334862385321102e-06, "loss": 0.0008, "step": 2223 }, { "epoch": 0.5666963944451523, "grad_norm": 2.223581314086914, "learning_rate": 4.332313965341489e-06, "loss": 0.0005, "step": 2224 }, { "epoch": 0.5669512039750286, "grad_norm": 2.2281875610351562, "learning_rate": 4.329765545361876e-06, "loss": 0.0005, "step": 2225 }, { "epoch": 0.5672060135049051, "grad_norm": 3.342020273208618, "learning_rate": 4.3272171253822634e-06, "loss": 0.0011, "step": 2226 }, { "epoch": 0.5674608230347815, "grad_norm": 0.943690299987793, "learning_rate": 4.324668705402651e-06, "loss": 0.0002, "step": 2227 }, { "epoch": 0.567715632564658, "grad_norm": 0.22586815059185028, "learning_rate": 4.322120285423038e-06, "loss": 0.0003, "step": 2228 }, { "epoch": 0.5679704420945343, "grad_norm": 2.394345283508301, "learning_rate": 4.319571865443425e-06, "loss": 0.0005, "step": 2229 }, { "epoch": 0.5682252516244107, "grad_norm": 5.186666488647461, "learning_rate": 4.317023445463813e-06, "loss": 0.002, "step": 2230 }, { "epoch": 0.5684800611542872, "grad_norm": 3.152087688446045, "learning_rate": 4.3144750254842e-06, "loss": 0.0009, "step": 2231 }, { "epoch": 0.5687348706841636, "grad_norm": 0.9352667331695557, "learning_rate": 4.311926605504588e-06, "loss": 0.0002, "step": 2232 }, { "epoch": 0.56898968021404, "grad_norm": 1.306598424911499, "learning_rate": 4.309378185524975e-06, "loss": 0.0003, "step": 2233 }, { "epoch": 0.5692444897439164, "grad_norm": 0.7512192130088806, "learning_rate": 4.3068297655453625e-06, "loss": 0.0012, "step": 2234 }, { "epoch": 0.5694992992737928, "grad_norm": 2.5998876094818115, "learning_rate": 4.304281345565749e-06, "loss": 0.0005, "step": 2235 }, { "epoch": 0.5697541088036693, "grad_norm": 0.9304439425468445, "learning_rate": 4.301732925586136e-06, "loss": 0.0002, "step": 2236 }, { "epoch": 0.5700089183335457, "grad_norm": 1.686388611793518, "learning_rate": 4.299184505606524e-06, "loss": 0.0004, "step": 2237 }, { "epoch": 0.570263727863422, "grad_norm": 1.8504163026809692, "learning_rate": 4.296636085626912e-06, "loss": 0.0003, "step": 2238 }, { "epoch": 0.5705185373932985, "grad_norm": 2.4114675521850586, "learning_rate": 4.294087665647299e-06, "loss": 0.0005, "step": 2239 }, { "epoch": 0.5707733469231749, "grad_norm": 3.7059383392333984, "learning_rate": 4.291539245667687e-06, "loss": 0.001, "step": 2240 }, { "epoch": 0.5710281564530514, "grad_norm": 0.8809105753898621, "learning_rate": 4.288990825688074e-06, "loss": 0.0009, "step": 2241 }, { "epoch": 0.5712829659829277, "grad_norm": 2.611163854598999, "learning_rate": 4.2864424057084615e-06, "loss": 0.0005, "step": 2242 }, { "epoch": 0.5715377755128042, "grad_norm": 0.74530029296875, "learning_rate": 4.283893985728848e-06, "loss": 0.0002, "step": 2243 }, { "epoch": 0.5717925850426806, "grad_norm": 0.2606738805770874, "learning_rate": 4.281345565749236e-06, "loss": 0.0003, "step": 2244 }, { "epoch": 0.572047394572557, "grad_norm": 1.8652862310409546, "learning_rate": 4.278797145769623e-06, "loss": 0.0005, "step": 2245 }, { "epoch": 0.5723022041024334, "grad_norm": 1.315298080444336, "learning_rate": 4.27624872579001e-06, "loss": 0.0002, "step": 2246 }, { "epoch": 0.5725570136323098, "grad_norm": 66.17877960205078, "learning_rate": 4.273700305810398e-06, "loss": 0.7188, "step": 2247 }, { "epoch": 0.5728118231621863, "grad_norm": 1.1166601181030273, "learning_rate": 4.271151885830786e-06, "loss": 0.0002, "step": 2248 }, { "epoch": 0.5730666326920627, "grad_norm": 1.126755952835083, "learning_rate": 4.268603465851173e-06, "loss": 0.0003, "step": 2249 }, { "epoch": 0.5733214422219391, "grad_norm": 1.665537714958191, "learning_rate": 4.26605504587156e-06, "loss": 0.0003, "step": 2250 }, { "epoch": 0.5735762517518155, "grad_norm": 1.5019487142562866, "learning_rate": 4.2635066258919474e-06, "loss": 0.0004, "step": 2251 }, { "epoch": 0.5738310612816919, "grad_norm": 3.1872949600219727, "learning_rate": 4.260958205912334e-06, "loss": 0.0009, "step": 2252 }, { "epoch": 0.5740858708115684, "grad_norm": 0.7459023594856262, "learning_rate": 4.258409785932722e-06, "loss": 0.0008, "step": 2253 }, { "epoch": 0.5743406803414448, "grad_norm": 22.446624755859375, "learning_rate": 4.255861365953109e-06, "loss": 0.0179, "step": 2254 }, { "epoch": 0.5745954898713211, "grad_norm": 2.668041229248047, "learning_rate": 4.253312945973497e-06, "loss": 0.0012, "step": 2255 }, { "epoch": 0.5748502994011976, "grad_norm": 6.294149398803711, "learning_rate": 4.250764525993884e-06, "loss": 0.0029, "step": 2256 }, { "epoch": 0.575105108931074, "grad_norm": 4.280332088470459, "learning_rate": 4.248216106014272e-06, "loss": 0.0016, "step": 2257 }, { "epoch": 0.5753599184609505, "grad_norm": 3.3555712699890137, "learning_rate": 4.2456676860346595e-06, "loss": 0.0009, "step": 2258 }, { "epoch": 0.5756147279908268, "grad_norm": 2.4290499687194824, "learning_rate": 4.2431192660550464e-06, "loss": 0.0008, "step": 2259 }, { "epoch": 0.5758695375207032, "grad_norm": 0.9243206977844238, "learning_rate": 4.240570846075433e-06, "loss": 0.0001, "step": 2260 }, { "epoch": 0.5761243470505797, "grad_norm": 0.20836898684501648, "learning_rate": 4.238022426095821e-06, "loss": 0.0002, "step": 2261 }, { "epoch": 0.5763791565804561, "grad_norm": 3.733273983001709, "learning_rate": 4.235474006116208e-06, "loss": 0.0013, "step": 2262 }, { "epoch": 0.5766339661103326, "grad_norm": 0.60247802734375, "learning_rate": 4.232925586136596e-06, "loss": 0.0002, "step": 2263 }, { "epoch": 0.5768887756402089, "grad_norm": 3.7255702018737793, "learning_rate": 4.230377166156983e-06, "loss": 0.0011, "step": 2264 }, { "epoch": 0.5771435851700853, "grad_norm": 1.6709717512130737, "learning_rate": 4.227828746177371e-06, "loss": 0.0003, "step": 2265 }, { "epoch": 0.5773983946999618, "grad_norm": 2.6622424125671387, "learning_rate": 4.225280326197758e-06, "loss": 0.0014, "step": 2266 }, { "epoch": 0.5776532042298382, "grad_norm": 0.38276562094688416, "learning_rate": 4.222731906218145e-06, "loss": 0.0002, "step": 2267 }, { "epoch": 0.5779080137597146, "grad_norm": 4.7639031410217285, "learning_rate": 4.220183486238532e-06, "loss": 0.0023, "step": 2268 }, { "epoch": 0.578162823289591, "grad_norm": 2.609205484390259, "learning_rate": 4.21763506625892e-06, "loss": 0.0006, "step": 2269 }, { "epoch": 0.5784176328194675, "grad_norm": 4.625356197357178, "learning_rate": 4.215086646279307e-06, "loss": 0.0016, "step": 2270 }, { "epoch": 0.5786724423493439, "grad_norm": 8.73829174041748, "learning_rate": 4.212538226299694e-06, "loss": 0.0055, "step": 2271 }, { "epoch": 0.5789272518792202, "grad_norm": 4.148978233337402, "learning_rate": 4.209989806320082e-06, "loss": 0.0032, "step": 2272 }, { "epoch": 0.5791820614090967, "grad_norm": 5.03248929977417, "learning_rate": 4.20744138634047e-06, "loss": 0.0019, "step": 2273 }, { "epoch": 0.5794368709389731, "grad_norm": 4.0161590576171875, "learning_rate": 4.204892966360857e-06, "loss": 0.0025, "step": 2274 }, { "epoch": 0.5796916804688496, "grad_norm": 1.4828299283981323, "learning_rate": 4.202344546381244e-06, "loss": 0.0002, "step": 2275 }, { "epoch": 0.579946489998726, "grad_norm": 2.7841086387634277, "learning_rate": 4.199796126401631e-06, "loss": 0.0007, "step": 2276 }, { "epoch": 0.5802012995286023, "grad_norm": 4.846079349517822, "learning_rate": 4.197247706422018e-06, "loss": 0.0018, "step": 2277 }, { "epoch": 0.5804561090584788, "grad_norm": 4.821611404418945, "learning_rate": 4.194699286442406e-06, "loss": 0.0019, "step": 2278 }, { "epoch": 0.5807109185883552, "grad_norm": 3.91159725189209, "learning_rate": 4.192150866462794e-06, "loss": 0.0012, "step": 2279 }, { "epoch": 0.5809657281182317, "grad_norm": 5.631570339202881, "learning_rate": 4.189602446483181e-06, "loss": 0.0023, "step": 2280 }, { "epoch": 0.581220537648108, "grad_norm": 4.5287089347839355, "learning_rate": 4.187054026503568e-06, "loss": 0.0015, "step": 2281 }, { "epoch": 0.5814753471779844, "grad_norm": 0.5639739036560059, "learning_rate": 4.184505606523956e-06, "loss": 0.0002, "step": 2282 }, { "epoch": 0.5817301567078609, "grad_norm": 0.5617520809173584, "learning_rate": 4.181957186544343e-06, "loss": 0.0002, "step": 2283 }, { "epoch": 0.5819849662377373, "grad_norm": 1.1141971349716187, "learning_rate": 4.17940876656473e-06, "loss": 0.0002, "step": 2284 }, { "epoch": 0.5822397757676137, "grad_norm": 1.5059318542480469, "learning_rate": 4.176860346585117e-06, "loss": 0.0007, "step": 2285 }, { "epoch": 0.5824945852974901, "grad_norm": 3.5396738052368164, "learning_rate": 4.174311926605505e-06, "loss": 0.0012, "step": 2286 }, { "epoch": 0.5827493948273665, "grad_norm": 1.1281332969665527, "learning_rate": 4.171763506625892e-06, "loss": 0.0004, "step": 2287 }, { "epoch": 0.583004204357243, "grad_norm": 1.8752360343933105, "learning_rate": 4.169215086646279e-06, "loss": 0.0005, "step": 2288 }, { "epoch": 0.5832590138871194, "grad_norm": 4.797525882720947, "learning_rate": 4.166666666666667e-06, "loss": 0.0029, "step": 2289 }, { "epoch": 0.5835138234169958, "grad_norm": 1.4806455373764038, "learning_rate": 4.164118246687055e-06, "loss": 0.0003, "step": 2290 }, { "epoch": 0.5837686329468722, "grad_norm": 2.231436014175415, "learning_rate": 4.161569826707442e-06, "loss": 0.0004, "step": 2291 }, { "epoch": 0.5840234424767486, "grad_norm": 3.5266528129577637, "learning_rate": 4.1590214067278286e-06, "loss": 0.001, "step": 2292 }, { "epoch": 0.5842782520066251, "grad_norm": 2.792703151702881, "learning_rate": 4.156472986748216e-06, "loss": 0.0007, "step": 2293 }, { "epoch": 0.5845330615365014, "grad_norm": 0.044058818370103836, "learning_rate": 4.153924566768604e-06, "loss": 0.0001, "step": 2294 }, { "epoch": 0.5847878710663779, "grad_norm": 0.9580600261688232, "learning_rate": 4.151376146788991e-06, "loss": 0.0004, "step": 2295 }, { "epoch": 0.5850426805962543, "grad_norm": 0.13467200100421906, "learning_rate": 4.148827726809379e-06, "loss": 0.0002, "step": 2296 }, { "epoch": 0.5852974901261307, "grad_norm": 2.6114330291748047, "learning_rate": 4.146279306829766e-06, "loss": 0.0006, "step": 2297 }, { "epoch": 0.5855522996560071, "grad_norm": 2.040454626083374, "learning_rate": 4.143730886850153e-06, "loss": 0.0004, "step": 2298 }, { "epoch": 0.5858071091858835, "grad_norm": 3.524815082550049, "learning_rate": 4.141182466870541e-06, "loss": 0.0015, "step": 2299 }, { "epoch": 0.58606191871576, "grad_norm": 3.1466660499572754, "learning_rate": 4.1386340468909284e-06, "loss": 0.0009, "step": 2300 }, { "epoch": 0.5863167282456364, "grad_norm": 2.2467739582061768, "learning_rate": 4.136085626911315e-06, "loss": 0.0005, "step": 2301 }, { "epoch": 0.5865715377755129, "grad_norm": 1.1308525800704956, "learning_rate": 4.133537206931702e-06, "loss": 0.0003, "step": 2302 }, { "epoch": 0.5868263473053892, "grad_norm": 0.5633962154388428, "learning_rate": 4.13098878695209e-06, "loss": 0.0002, "step": 2303 }, { "epoch": 0.5870811568352656, "grad_norm": 0.03511561080813408, "learning_rate": 4.128440366972478e-06, "loss": 0.0001, "step": 2304 }, { "epoch": 0.5873359663651421, "grad_norm": 0.7403897047042847, "learning_rate": 4.125891946992865e-06, "loss": 0.0001, "step": 2305 }, { "epoch": 0.5875907758950185, "grad_norm": 0.5676377415657043, "learning_rate": 4.123343527013252e-06, "loss": 0.0002, "step": 2306 }, { "epoch": 0.5878455854248948, "grad_norm": 0.5914222002029419, "learning_rate": 4.12079510703364e-06, "loss": 0.0003, "step": 2307 }, { "epoch": 0.5881003949547713, "grad_norm": 0.7503479719161987, "learning_rate": 4.118246687054027e-06, "loss": 0.0002, "step": 2308 }, { "epoch": 0.5883552044846477, "grad_norm": 2.2366456985473633, "learning_rate": 4.115698267074414e-06, "loss": 0.0005, "step": 2309 }, { "epoch": 0.5886100140145242, "grad_norm": 1.1234204769134521, "learning_rate": 4.113149847094801e-06, "loss": 0.0002, "step": 2310 }, { "epoch": 0.5888648235444005, "grad_norm": 0.7812642455101013, "learning_rate": 4.110601427115189e-06, "loss": 0.0004, "step": 2311 }, { "epoch": 0.589119633074277, "grad_norm": 0.7518934607505798, "learning_rate": 4.108053007135576e-06, "loss": 0.0001, "step": 2312 }, { "epoch": 0.5893744426041534, "grad_norm": 1.1526548862457275, "learning_rate": 4.105504587155963e-06, "loss": 0.0006, "step": 2313 }, { "epoch": 0.5896292521340298, "grad_norm": 60.3428955078125, "learning_rate": 4.102956167176351e-06, "loss": 0.7502, "step": 2314 }, { "epoch": 0.5898840616639063, "grad_norm": 60.48334884643555, "learning_rate": 4.100407747196739e-06, "loss": 0.7735, "step": 2315 }, { "epoch": 0.5901388711937826, "grad_norm": 2.7673099040985107, "learning_rate": 4.097859327217126e-06, "loss": 0.0006, "step": 2316 }, { "epoch": 0.590393680723659, "grad_norm": 2.593674421310425, "learning_rate": 4.095310907237513e-06, "loss": 0.0006, "step": 2317 }, { "epoch": 0.5906484902535355, "grad_norm": 3.8885467052459717, "learning_rate": 4.0927624872579e-06, "loss": 0.0012, "step": 2318 }, { "epoch": 0.5909032997834119, "grad_norm": 0.5743475556373596, "learning_rate": 4.090214067278288e-06, "loss": 0.0002, "step": 2319 }, { "epoch": 0.5911581093132883, "grad_norm": 3.8770623207092285, "learning_rate": 4.087665647298675e-06, "loss": 0.0014, "step": 2320 }, { "epoch": 0.5914129188431647, "grad_norm": 0.7497850656509399, "learning_rate": 4.085117227319063e-06, "loss": 0.0001, "step": 2321 }, { "epoch": 0.5916677283730412, "grad_norm": 1.5317955017089844, "learning_rate": 4.08256880733945e-06, "loss": 0.0006, "step": 2322 }, { "epoch": 0.5919225379029176, "grad_norm": 2.789416551589966, "learning_rate": 4.080020387359837e-06, "loss": 0.0007, "step": 2323 }, { "epoch": 0.5921773474327939, "grad_norm": 1.3038485050201416, "learning_rate": 4.077471967380225e-06, "loss": 0.0003, "step": 2324 }, { "epoch": 0.5924321569626704, "grad_norm": 0.7804918885231018, "learning_rate": 4.074923547400612e-06, "loss": 0.0004, "step": 2325 }, { "epoch": 0.5926869664925468, "grad_norm": 2.0448672771453857, "learning_rate": 4.072375127420999e-06, "loss": 0.0004, "step": 2326 }, { "epoch": 0.5929417760224233, "grad_norm": 2.9679973125457764, "learning_rate": 4.069826707441386e-06, "loss": 0.0009, "step": 2327 }, { "epoch": 0.5931965855522997, "grad_norm": 4.272449970245361, "learning_rate": 4.067278287461774e-06, "loss": 0.0016, "step": 2328 }, { "epoch": 0.593451395082176, "grad_norm": 4.095143795013428, "learning_rate": 4.064729867482161e-06, "loss": 0.0014, "step": 2329 }, { "epoch": 0.5937062046120525, "grad_norm": 0.2110641449689865, "learning_rate": 4.062181447502549e-06, "loss": 0.0002, "step": 2330 }, { "epoch": 0.5939610141419289, "grad_norm": 0.3745718002319336, "learning_rate": 4.059633027522936e-06, "loss": 0.0001, "step": 2331 }, { "epoch": 0.5942158236718054, "grad_norm": 1.108538269996643, "learning_rate": 4.057084607543324e-06, "loss": 0.0002, "step": 2332 }, { "epoch": 0.5944706332016817, "grad_norm": 2.974191665649414, "learning_rate": 4.0545361875637106e-06, "loss": 0.0012, "step": 2333 }, { "epoch": 0.5947254427315581, "grad_norm": 1.6697813272476196, "learning_rate": 4.051987767584098e-06, "loss": 0.0004, "step": 2334 }, { "epoch": 0.5949802522614346, "grad_norm": 0.19538511335849762, "learning_rate": 4.049439347604486e-06, "loss": 0.0001, "step": 2335 }, { "epoch": 0.595235061791311, "grad_norm": 4.460114002227783, "learning_rate": 4.046890927624873e-06, "loss": 0.0015, "step": 2336 }, { "epoch": 0.5954898713211874, "grad_norm": 2.0396974086761475, "learning_rate": 4.04434250764526e-06, "loss": 0.0005, "step": 2337 }, { "epoch": 0.5957446808510638, "grad_norm": 0.9442929029464722, "learning_rate": 4.041794087665648e-06, "loss": 0.0004, "step": 2338 }, { "epoch": 0.5959994903809402, "grad_norm": 0.3810424208641052, "learning_rate": 4.039245667686035e-06, "loss": 0.0002, "step": 2339 }, { "epoch": 0.5962542999108167, "grad_norm": 0.37041351199150085, "learning_rate": 4.036697247706423e-06, "loss": 0.0001, "step": 2340 }, { "epoch": 0.5965091094406931, "grad_norm": 2.783022165298462, "learning_rate": 4.0341488277268096e-06, "loss": 0.0009, "step": 2341 }, { "epoch": 0.5967639189705695, "grad_norm": 4.623509407043457, "learning_rate": 4.031600407747197e-06, "loss": 0.0017, "step": 2342 }, { "epoch": 0.5970187285004459, "grad_norm": 2.9703402519226074, "learning_rate": 4.029051987767584e-06, "loss": 0.0012, "step": 2343 }, { "epoch": 0.5972735380303223, "grad_norm": 3.5295510292053223, "learning_rate": 4.026503567787971e-06, "loss": 0.0014, "step": 2344 }, { "epoch": 0.5975283475601988, "grad_norm": 1.8534736633300781, "learning_rate": 4.023955147808359e-06, "loss": 0.0006, "step": 2345 }, { "epoch": 0.5977831570900751, "grad_norm": 0.19460131227970123, "learning_rate": 4.021406727828747e-06, "loss": 0.0002, "step": 2346 }, { "epoch": 0.5980379666199516, "grad_norm": 1.8473585844039917, "learning_rate": 4.018858307849134e-06, "loss": 0.0003, "step": 2347 }, { "epoch": 0.598292776149828, "grad_norm": 0.4301207363605499, "learning_rate": 4.016309887869521e-06, "loss": 0.0003, "step": 2348 }, { "epoch": 0.5985475856797045, "grad_norm": 2.0428214073181152, "learning_rate": 4.013761467889909e-06, "loss": 0.0004, "step": 2349 }, { "epoch": 0.5988023952095808, "grad_norm": 1.8440130949020386, "learning_rate": 4.011213047910296e-06, "loss": 0.0003, "step": 2350 }, { "epoch": 0.5990572047394572, "grad_norm": 3.083749294281006, "learning_rate": 4.008664627930683e-06, "loss": 0.0006, "step": 2351 }, { "epoch": 0.5993120142693337, "grad_norm": 1.114251971244812, "learning_rate": 4.00611620795107e-06, "loss": 0.0002, "step": 2352 }, { "epoch": 0.5995668237992101, "grad_norm": 1.482119083404541, "learning_rate": 4.003567787971458e-06, "loss": 0.0003, "step": 2353 }, { "epoch": 0.5998216333290866, "grad_norm": 0.5669883489608765, "learning_rate": 4.001019367991845e-06, "loss": 0.0002, "step": 2354 }, { "epoch": 0.6000764428589629, "grad_norm": 0.7571197748184204, "learning_rate": 3.998470948012233e-06, "loss": 0.0003, "step": 2355 }, { "epoch": 0.6003312523888393, "grad_norm": 0.7789928913116455, "learning_rate": 3.995922528032621e-06, "loss": 0.0005, "step": 2356 }, { "epoch": 0.6005860619187158, "grad_norm": 2.403995990753174, "learning_rate": 3.993374108053008e-06, "loss": 0.0009, "step": 2357 }, { "epoch": 0.6008408714485922, "grad_norm": 1.6603052616119385, "learning_rate": 3.9908256880733945e-06, "loss": 0.0003, "step": 2358 }, { "epoch": 0.6010956809784685, "grad_norm": 2.380953788757324, "learning_rate": 3.988277268093782e-06, "loss": 0.0019, "step": 2359 }, { "epoch": 0.601350490508345, "grad_norm": 0.7474303841590881, "learning_rate": 3.985728848114169e-06, "loss": 0.0002, "step": 2360 }, { "epoch": 0.6016053000382214, "grad_norm": 1.2990285158157349, "learning_rate": 3.983180428134557e-06, "loss": 0.0003, "step": 2361 }, { "epoch": 0.6018601095680979, "grad_norm": 0.7410891652107239, "learning_rate": 3.980632008154944e-06, "loss": 0.0001, "step": 2362 }, { "epoch": 0.6021149190979742, "grad_norm": 4.590795993804932, "learning_rate": 3.978083588175332e-06, "loss": 0.0017, "step": 2363 }, { "epoch": 0.6023697286278507, "grad_norm": 4.073731899261475, "learning_rate": 3.975535168195719e-06, "loss": 0.0013, "step": 2364 }, { "epoch": 0.6026245381577271, "grad_norm": 3.3232181072235107, "learning_rate": 3.972986748216106e-06, "loss": 0.001, "step": 2365 }, { "epoch": 0.6028793476876035, "grad_norm": 1.1192880868911743, "learning_rate": 3.9704383282364936e-06, "loss": 0.0003, "step": 2366 }, { "epoch": 0.60313415721748, "grad_norm": 2.049832344055176, "learning_rate": 3.967889908256881e-06, "loss": 0.0004, "step": 2367 }, { "epoch": 0.6033889667473563, "grad_norm": 2.2197911739349365, "learning_rate": 3.965341488277268e-06, "loss": 0.0005, "step": 2368 }, { "epoch": 0.6036437762772328, "grad_norm": 2.969033718109131, "learning_rate": 3.962793068297655e-06, "loss": 0.0008, "step": 2369 }, { "epoch": 0.6038985858071092, "grad_norm": 2.2287144660949707, "learning_rate": 3.960244648318043e-06, "loss": 0.0007, "step": 2370 }, { "epoch": 0.6041533953369856, "grad_norm": 2.5897645950317383, "learning_rate": 3.957696228338431e-06, "loss": 0.0006, "step": 2371 }, { "epoch": 0.604408204866862, "grad_norm": 3.8767244815826416, "learning_rate": 3.955147808358818e-06, "loss": 0.0012, "step": 2372 }, { "epoch": 0.6046630143967384, "grad_norm": 3.139017105102539, "learning_rate": 3.952599388379206e-06, "loss": 0.0008, "step": 2373 }, { "epoch": 0.6049178239266149, "grad_norm": 2.4043192863464355, "learning_rate": 3.9500509683995926e-06, "loss": 0.0006, "step": 2374 }, { "epoch": 0.6051726334564913, "grad_norm": 1.6636638641357422, "learning_rate": 3.9475025484199795e-06, "loss": 0.0006, "step": 2375 }, { "epoch": 0.6054274429863676, "grad_norm": 2.406338930130005, "learning_rate": 3.944954128440367e-06, "loss": 0.0005, "step": 2376 }, { "epoch": 0.6056822525162441, "grad_norm": 1.479410171508789, "learning_rate": 3.942405708460755e-06, "loss": 0.0002, "step": 2377 }, { "epoch": 0.6059370620461205, "grad_norm": 6.845188617706299, "learning_rate": 3.939857288481142e-06, "loss": 0.0067, "step": 2378 }, { "epoch": 0.606191871575997, "grad_norm": 4.26790714263916, "learning_rate": 3.937308868501529e-06, "loss": 0.0016, "step": 2379 }, { "epoch": 0.6064466811058734, "grad_norm": 2.5892741680145264, "learning_rate": 3.934760448521917e-06, "loss": 0.0006, "step": 2380 }, { "epoch": 0.6067014906357497, "grad_norm": 1.85635244846344, "learning_rate": 3.932212028542305e-06, "loss": 0.0003, "step": 2381 }, { "epoch": 0.6069563001656262, "grad_norm": 0.5533831119537354, "learning_rate": 3.9296636085626916e-06, "loss": 0.0001, "step": 2382 }, { "epoch": 0.6072111096955026, "grad_norm": 1.1137992143630981, "learning_rate": 3.9271151885830785e-06, "loss": 0.0002, "step": 2383 }, { "epoch": 0.6074659192253791, "grad_norm": 1.6686269044876099, "learning_rate": 3.924566768603466e-06, "loss": 0.0005, "step": 2384 }, { "epoch": 0.6077207287552554, "grad_norm": 1.8620216846466064, "learning_rate": 3.922018348623853e-06, "loss": 0.0004, "step": 2385 }, { "epoch": 0.6079755382851318, "grad_norm": 2.9598629474639893, "learning_rate": 3.919469928644241e-06, "loss": 0.0007, "step": 2386 }, { "epoch": 0.6082303478150083, "grad_norm": 2.6153676509857178, "learning_rate": 3.916921508664628e-06, "loss": 0.0008, "step": 2387 }, { "epoch": 0.6084851573448847, "grad_norm": 1.3382700681686401, "learning_rate": 3.914373088685016e-06, "loss": 0.0008, "step": 2388 }, { "epoch": 0.6087399668747611, "grad_norm": 3.3619799613952637, "learning_rate": 3.911824668705403e-06, "loss": 0.0012, "step": 2389 }, { "epoch": 0.6089947764046375, "grad_norm": 1.8488291501998901, "learning_rate": 3.90927624872579e-06, "loss": 0.0004, "step": 2390 }, { "epoch": 0.609249585934514, "grad_norm": 0.30609461665153503, "learning_rate": 3.906727828746178e-06, "loss": 0.0007, "step": 2391 }, { "epoch": 0.6095043954643904, "grad_norm": 3.340071678161621, "learning_rate": 3.904179408766565e-06, "loss": 0.0009, "step": 2392 }, { "epoch": 0.6097592049942668, "grad_norm": 3.1412105560302734, "learning_rate": 3.901630988786952e-06, "loss": 0.0008, "step": 2393 }, { "epoch": 0.6100140145241432, "grad_norm": 1.8590391874313354, "learning_rate": 3.89908256880734e-06, "loss": 0.0004, "step": 2394 }, { "epoch": 0.6102688240540196, "grad_norm": 2.422524929046631, "learning_rate": 3.896534148827727e-06, "loss": 0.0009, "step": 2395 }, { "epoch": 0.610523633583896, "grad_norm": 2.0391900539398193, "learning_rate": 3.893985728848114e-06, "loss": 0.0004, "step": 2396 }, { "epoch": 0.6107784431137725, "grad_norm": 1.6723039150238037, "learning_rate": 3.891437308868502e-06, "loss": 0.0003, "step": 2397 }, { "epoch": 0.6110332526436488, "grad_norm": 1.2025357484817505, "learning_rate": 3.88888888888889e-06, "loss": 0.0004, "step": 2398 }, { "epoch": 0.6112880621735253, "grad_norm": 0.23118740320205688, "learning_rate": 3.8863404689092765e-06, "loss": 0.0002, "step": 2399 }, { "epoch": 0.6115428717034017, "grad_norm": 0.7660293579101562, "learning_rate": 3.8837920489296635e-06, "loss": 0.0005, "step": 2400 }, { "epoch": 0.6117976812332782, "grad_norm": 1.4722241163253784, "learning_rate": 3.881243628950051e-06, "loss": 0.0002, "step": 2401 }, { "epoch": 0.6120524907631545, "grad_norm": 2.0396158695220947, "learning_rate": 3.878695208970439e-06, "loss": 0.0004, "step": 2402 }, { "epoch": 0.6123073002930309, "grad_norm": 2.396202564239502, "learning_rate": 3.876146788990826e-06, "loss": 0.0005, "step": 2403 }, { "epoch": 0.6125621098229074, "grad_norm": 3.3268539905548096, "learning_rate": 3.873598369011213e-06, "loss": 0.001, "step": 2404 }, { "epoch": 0.6128169193527838, "grad_norm": 0.946640133857727, "learning_rate": 3.871049949031601e-06, "loss": 0.0004, "step": 2405 }, { "epoch": 0.6130717288826603, "grad_norm": 1.480912446975708, "learning_rate": 3.868501529051988e-06, "loss": 0.0003, "step": 2406 }, { "epoch": 0.6133265384125366, "grad_norm": 0.5659340620040894, "learning_rate": 3.8659531090723755e-06, "loss": 0.0001, "step": 2407 }, { "epoch": 0.613581347942413, "grad_norm": 1.4747209548950195, "learning_rate": 3.8634046890927625e-06, "loss": 0.0003, "step": 2408 }, { "epoch": 0.6138361574722895, "grad_norm": 0.7378814816474915, "learning_rate": 3.86085626911315e-06, "loss": 0.0001, "step": 2409 }, { "epoch": 0.6140909670021659, "grad_norm": 1.1071144342422485, "learning_rate": 3.858307849133537e-06, "loss": 0.0002, "step": 2410 }, { "epoch": 0.6143457765320423, "grad_norm": 6.031867027282715, "learning_rate": 3.855759429153925e-06, "loss": 0.0028, "step": 2411 }, { "epoch": 0.6146005860619187, "grad_norm": 3.1443819999694824, "learning_rate": 3.853211009174313e-06, "loss": 0.0009, "step": 2412 }, { "epoch": 0.6148553955917951, "grad_norm": 0.9373631477355957, "learning_rate": 3.8506625891947e-06, "loss": 0.0003, "step": 2413 }, { "epoch": 0.6151102051216716, "grad_norm": 0.20452579855918884, "learning_rate": 3.848114169215087e-06, "loss": 0.0003, "step": 2414 }, { "epoch": 0.6153650146515479, "grad_norm": 2.0391433238983154, "learning_rate": 3.8455657492354746e-06, "loss": 0.0008, "step": 2415 }, { "epoch": 0.6156198241814244, "grad_norm": 1.3009730577468872, "learning_rate": 3.8430173292558615e-06, "loss": 0.0004, "step": 2416 }, { "epoch": 0.6158746337113008, "grad_norm": 1.2927988767623901, "learning_rate": 3.840468909276249e-06, "loss": 0.0002, "step": 2417 }, { "epoch": 0.6161294432411772, "grad_norm": 0.7378872036933899, "learning_rate": 3.837920489296636e-06, "loss": 0.0001, "step": 2418 }, { "epoch": 0.6163842527710537, "grad_norm": 2.411531686782837, "learning_rate": 3.835372069317024e-06, "loss": 0.0005, "step": 2419 }, { "epoch": 0.61663906230093, "grad_norm": 1.296330213546753, "learning_rate": 3.832823649337411e-06, "loss": 0.0002, "step": 2420 }, { "epoch": 0.6168938718308065, "grad_norm": 1.6930468082427979, "learning_rate": 3.830275229357798e-06, "loss": 0.0006, "step": 2421 }, { "epoch": 0.6171486813606829, "grad_norm": 2.603588104248047, "learning_rate": 3.827726809378186e-06, "loss": 0.0005, "step": 2422 }, { "epoch": 0.6174034908905593, "grad_norm": 2.951341152191162, "learning_rate": 3.8251783893985736e-06, "loss": 0.0007, "step": 2423 }, { "epoch": 0.6176583004204357, "grad_norm": 1.4910796880722046, "learning_rate": 3.8226299694189605e-06, "loss": 0.0004, "step": 2424 }, { "epoch": 0.6179131099503121, "grad_norm": 0.9220238924026489, "learning_rate": 3.8200815494393475e-06, "loss": 0.0001, "step": 2425 }, { "epoch": 0.6181679194801886, "grad_norm": 2.8125534057617188, "learning_rate": 3.817533129459735e-06, "loss": 0.0019, "step": 2426 }, { "epoch": 0.618422729010065, "grad_norm": 0.1877705454826355, "learning_rate": 3.814984709480123e-06, "loss": 0.0001, "step": 2427 }, { "epoch": 0.6186775385399413, "grad_norm": 2.6116199493408203, "learning_rate": 3.81243628950051e-06, "loss": 0.0009, "step": 2428 }, { "epoch": 0.6189323480698178, "grad_norm": 2.9677298069000244, "learning_rate": 3.8098878695208974e-06, "loss": 0.0007, "step": 2429 }, { "epoch": 0.6191871575996942, "grad_norm": 1.8617652654647827, "learning_rate": 3.8073394495412848e-06, "loss": 0.0008, "step": 2430 }, { "epoch": 0.6194419671295707, "grad_norm": 3.7008213996887207, "learning_rate": 3.8047910295616717e-06, "loss": 0.0014, "step": 2431 }, { "epoch": 0.6196967766594471, "grad_norm": 0.29059505462646484, "learning_rate": 3.8022426095820595e-06, "loss": 0.0004, "step": 2432 }, { "epoch": 0.6199515861893234, "grad_norm": 2.579146146774292, "learning_rate": 3.799694189602447e-06, "loss": 0.0005, "step": 2433 }, { "epoch": 0.6202063957191999, "grad_norm": 1.661944031715393, "learning_rate": 3.7971457696228343e-06, "loss": 0.0005, "step": 2434 }, { "epoch": 0.6204612052490763, "grad_norm": 1.853400707244873, "learning_rate": 3.7945973496432216e-06, "loss": 0.0003, "step": 2435 }, { "epoch": 0.6207160147789528, "grad_norm": 2.798421859741211, "learning_rate": 3.7920489296636086e-06, "loss": 0.0012, "step": 2436 }, { "epoch": 0.6209708243088291, "grad_norm": 2.6067824363708496, "learning_rate": 3.789500509683996e-06, "loss": 0.0005, "step": 2437 }, { "epoch": 0.6212256338387055, "grad_norm": 0.1929057538509369, "learning_rate": 3.7869520897043838e-06, "loss": 0.0001, "step": 2438 }, { "epoch": 0.621480443368582, "grad_norm": 2.817121744155884, "learning_rate": 3.784403669724771e-06, "loss": 0.0015, "step": 2439 }, { "epoch": 0.6217352528984584, "grad_norm": 0.7427836656570435, "learning_rate": 3.781855249745158e-06, "loss": 0.0001, "step": 2440 }, { "epoch": 0.6219900624283348, "grad_norm": 4.792870998382568, "learning_rate": 3.7793068297655455e-06, "loss": 0.0018, "step": 2441 }, { "epoch": 0.6222448719582112, "grad_norm": 3.124422788619995, "learning_rate": 3.776758409785933e-06, "loss": 0.0008, "step": 2442 }, { "epoch": 0.6224996814880877, "grad_norm": 1.1072521209716797, "learning_rate": 3.7742099898063207e-06, "loss": 0.0001, "step": 2443 }, { "epoch": 0.6227544910179641, "grad_norm": 2.7650883197784424, "learning_rate": 3.771661569826708e-06, "loss": 0.0007, "step": 2444 }, { "epoch": 0.6230093005478405, "grad_norm": 0.5847465991973877, "learning_rate": 3.769113149847095e-06, "loss": 0.0003, "step": 2445 }, { "epoch": 0.6232641100777169, "grad_norm": 3.1365325450897217, "learning_rate": 3.7665647298674824e-06, "loss": 0.0008, "step": 2446 }, { "epoch": 0.6235189196075933, "grad_norm": 1.6601247787475586, "learning_rate": 3.7640163098878697e-06, "loss": 0.0003, "step": 2447 }, { "epoch": 0.6237737291374698, "grad_norm": 0.46633318066596985, "learning_rate": 3.7614678899082575e-06, "loss": 0.0004, "step": 2448 }, { "epoch": 0.6240285386673462, "grad_norm": 2.4132702350616455, "learning_rate": 3.7589194699286445e-06, "loss": 0.0006, "step": 2449 }, { "epoch": 0.6242833481972225, "grad_norm": 1.7377135753631592, "learning_rate": 3.756371049949032e-06, "loss": 0.0007, "step": 2450 }, { "epoch": 0.624538157727099, "grad_norm": 1.4701309204101562, "learning_rate": 3.7538226299694192e-06, "loss": 0.0002, "step": 2451 }, { "epoch": 0.6247929672569754, "grad_norm": 1.6590487957000732, "learning_rate": 3.751274209989806e-06, "loss": 0.0003, "step": 2452 }, { "epoch": 0.6250477767868519, "grad_norm": 1.8626036643981934, "learning_rate": 3.7487257900101944e-06, "loss": 0.0003, "step": 2453 }, { "epoch": 0.6253025863167282, "grad_norm": 1.8405342102050781, "learning_rate": 3.7461773700305814e-06, "loss": 0.0003, "step": 2454 }, { "epoch": 0.6255573958466046, "grad_norm": 4.625308036804199, "learning_rate": 3.7436289500509687e-06, "loss": 0.0017, "step": 2455 }, { "epoch": 0.6258122053764811, "grad_norm": 5.5427470207214355, "learning_rate": 3.741080530071356e-06, "loss": 0.0023, "step": 2456 }, { "epoch": 0.6260670149063575, "grad_norm": 4.060410976409912, "learning_rate": 3.738532110091743e-06, "loss": 0.0014, "step": 2457 }, { "epoch": 0.626321824436234, "grad_norm": 3.500692367553711, "learning_rate": 3.735983690112131e-06, "loss": 0.0015, "step": 2458 }, { "epoch": 0.6265766339661103, "grad_norm": 6.493718147277832, "learning_rate": 3.7334352701325183e-06, "loss": 0.004, "step": 2459 }, { "epoch": 0.6268314434959867, "grad_norm": 3.319047451019287, "learning_rate": 3.7308868501529056e-06, "loss": 0.0008, "step": 2460 }, { "epoch": 0.6270862530258632, "grad_norm": 3.888899803161621, "learning_rate": 3.7283384301732926e-06, "loss": 0.0013, "step": 2461 }, { "epoch": 0.6273410625557396, "grad_norm": 2.5673248767852783, "learning_rate": 3.72579001019368e-06, "loss": 0.0005, "step": 2462 }, { "epoch": 0.627595872085616, "grad_norm": 2.7893145084381104, "learning_rate": 3.7232415902140678e-06, "loss": 0.001, "step": 2463 }, { "epoch": 0.6278506816154924, "grad_norm": 1.8727754354476929, "learning_rate": 3.720693170234455e-06, "loss": 0.0003, "step": 2464 }, { "epoch": 0.6281054911453688, "grad_norm": 3.3169803619384766, "learning_rate": 3.7181447502548425e-06, "loss": 0.0009, "step": 2465 }, { "epoch": 0.6283603006752453, "grad_norm": 1.1085058450698853, "learning_rate": 3.7155963302752295e-06, "loss": 0.0002, "step": 2466 }, { "epoch": 0.6286151102051216, "grad_norm": 0.9337425231933594, "learning_rate": 3.713047910295617e-06, "loss": 0.0003, "step": 2467 }, { "epoch": 0.6288699197349981, "grad_norm": 0.9253475666046143, "learning_rate": 3.7104994903160046e-06, "loss": 0.0001, "step": 2468 }, { "epoch": 0.6291247292648745, "grad_norm": 5.3378729820251465, "learning_rate": 3.707951070336392e-06, "loss": 0.0022, "step": 2469 }, { "epoch": 0.629379538794751, "grad_norm": 7.924640655517578, "learning_rate": 3.705402650356779e-06, "loss": 0.0046, "step": 2470 }, { "epoch": 0.6296343483246274, "grad_norm": 1.3014293909072876, "learning_rate": 3.7028542303771663e-06, "loss": 0.0002, "step": 2471 }, { "epoch": 0.6298891578545037, "grad_norm": 2.0351827144622803, "learning_rate": 3.7003058103975537e-06, "loss": 0.0006, "step": 2472 }, { "epoch": 0.6301439673843802, "grad_norm": 2.5894312858581543, "learning_rate": 3.697757390417941e-06, "loss": 0.0006, "step": 2473 }, { "epoch": 0.6303987769142566, "grad_norm": 1.1979798078536987, "learning_rate": 3.695208970438329e-06, "loss": 0.0009, "step": 2474 }, { "epoch": 0.630653586444133, "grad_norm": 4.80360746383667, "learning_rate": 3.692660550458716e-06, "loss": 0.0029, "step": 2475 }, { "epoch": 0.6309083959740094, "grad_norm": 3.1403584480285645, "learning_rate": 3.6901121304791032e-06, "loss": 0.0008, "step": 2476 }, { "epoch": 0.6311632055038858, "grad_norm": 61.60186004638672, "learning_rate": 3.6875637104994906e-06, "loss": 0.6876, "step": 2477 }, { "epoch": 0.6314180150337623, "grad_norm": 4.8126349449157715, "learning_rate": 3.6850152905198775e-06, "loss": 0.0017, "step": 2478 }, { "epoch": 0.6316728245636387, "grad_norm": 3.502526044845581, "learning_rate": 3.6824668705402653e-06, "loss": 0.0009, "step": 2479 }, { "epoch": 0.631927634093515, "grad_norm": 6.116466045379639, "learning_rate": 3.6799184505606527e-06, "loss": 0.003, "step": 2480 }, { "epoch": 0.6321824436233915, "grad_norm": 3.696560859680176, "learning_rate": 3.67737003058104e-06, "loss": 0.0011, "step": 2481 }, { "epoch": 0.6324372531532679, "grad_norm": 0.5656196475028992, "learning_rate": 3.6748216106014275e-06, "loss": 0.0004, "step": 2482 }, { "epoch": 0.6326920626831444, "grad_norm": 0.3782546818256378, "learning_rate": 3.6722731906218144e-06, "loss": 0.0002, "step": 2483 }, { "epoch": 0.6329468722130208, "grad_norm": 2.6063952445983887, "learning_rate": 3.6697247706422022e-06, "loss": 0.0007, "step": 2484 }, { "epoch": 0.6332016817428971, "grad_norm": 3.691690444946289, "learning_rate": 3.6671763506625896e-06, "loss": 0.0011, "step": 2485 }, { "epoch": 0.6334564912727736, "grad_norm": 3.500302791595459, "learning_rate": 3.664627930682977e-06, "loss": 0.001, "step": 2486 }, { "epoch": 0.63371130080265, "grad_norm": 2.234415292739868, "learning_rate": 3.662079510703364e-06, "loss": 0.0005, "step": 2487 }, { "epoch": 0.6339661103325265, "grad_norm": 0.20916098356246948, "learning_rate": 3.6595310907237513e-06, "loss": 0.0003, "step": 2488 }, { "epoch": 0.6342209198624028, "grad_norm": 0.7413625121116638, "learning_rate": 3.656982670744139e-06, "loss": 0.0001, "step": 2489 }, { "epoch": 0.6344757293922793, "grad_norm": 52.4476318359375, "learning_rate": 3.6544342507645265e-06, "loss": 0.5944, "step": 2490 }, { "epoch": 0.6347305389221557, "grad_norm": 2.2397546768188477, "learning_rate": 3.651885830784914e-06, "loss": 0.0008, "step": 2491 }, { "epoch": 0.6349853484520321, "grad_norm": 5.166646957397461, "learning_rate": 3.649337410805301e-06, "loss": 0.002, "step": 2492 }, { "epoch": 0.6352401579819085, "grad_norm": 2.765261650085449, "learning_rate": 3.646788990825688e-06, "loss": 0.0007, "step": 2493 }, { "epoch": 0.6354949675117849, "grad_norm": 3.5002236366271973, "learning_rate": 3.644240570846076e-06, "loss": 0.001, "step": 2494 }, { "epoch": 0.6357497770416614, "grad_norm": 1.6665757894515991, "learning_rate": 3.6416921508664634e-06, "loss": 0.0003, "step": 2495 }, { "epoch": 0.6360045865715378, "grad_norm": 1.8433152437210083, "learning_rate": 3.6391437308868503e-06, "loss": 0.0003, "step": 2496 }, { "epoch": 0.6362593961014142, "grad_norm": 0.9552200436592102, "learning_rate": 3.6365953109072377e-06, "loss": 0.0003, "step": 2497 }, { "epoch": 0.6365142056312906, "grad_norm": 3.330857992172241, "learning_rate": 3.634046890927625e-06, "loss": 0.0009, "step": 2498 }, { "epoch": 0.636769015161167, "grad_norm": 1.6678745746612549, "learning_rate": 3.631498470948013e-06, "loss": 0.0004, "step": 2499 }, { "epoch": 0.6370238246910435, "grad_norm": 3.690056085586548, "learning_rate": 3.6289500509684e-06, "loss": 0.0012, "step": 2500 }, { "epoch": 0.6372786342209199, "grad_norm": 0.42482659220695496, "learning_rate": 3.626401630988787e-06, "loss": 0.0005, "step": 2501 }, { "epoch": 0.6375334437507962, "grad_norm": 209.5989227294922, "learning_rate": 3.6238532110091746e-06, "loss": 0.1524, "step": 2502 }, { "epoch": 0.6377882532806727, "grad_norm": 2.0245494842529297, "learning_rate": 3.621304791029562e-06, "loss": 0.0004, "step": 2503 }, { "epoch": 0.6380430628105491, "grad_norm": 2.4044411182403564, "learning_rate": 3.6187563710499497e-06, "loss": 0.0005, "step": 2504 }, { "epoch": 0.6382978723404256, "grad_norm": 0.18036282062530518, "learning_rate": 3.6162079510703367e-06, "loss": 0.0003, "step": 2505 }, { "epoch": 0.6385526818703019, "grad_norm": 1.08353853225708, "learning_rate": 3.613659531090724e-06, "loss": 0.0028, "step": 2506 }, { "epoch": 0.6388074914001783, "grad_norm": 5.1686224937438965, "learning_rate": 3.6111111111111115e-06, "loss": 0.0022, "step": 2507 }, { "epoch": 0.6390623009300548, "grad_norm": 6.108651161193848, "learning_rate": 3.6085626911314984e-06, "loss": 0.0029, "step": 2508 }, { "epoch": 0.6393171104599312, "grad_norm": 8.515315055847168, "learning_rate": 3.606014271151886e-06, "loss": 0.0053, "step": 2509 }, { "epoch": 0.6395719199898077, "grad_norm": 3.3299942016601562, "learning_rate": 3.6034658511722736e-06, "loss": 0.0017, "step": 2510 }, { "epoch": 0.639826729519684, "grad_norm": 3.6874537467956543, "learning_rate": 3.600917431192661e-06, "loss": 0.0012, "step": 2511 }, { "epoch": 0.6400815390495604, "grad_norm": 2.0285892486572266, "learning_rate": 3.5983690112130483e-06, "loss": 0.0005, "step": 2512 }, { "epoch": 0.6403363485794369, "grad_norm": 0.201388418674469, "learning_rate": 3.5958205912334353e-06, "loss": 0.0001, "step": 2513 }, { "epoch": 0.6405911581093133, "grad_norm": 2.781156063079834, "learning_rate": 3.5932721712538227e-06, "loss": 0.0008, "step": 2514 }, { "epoch": 0.6408459676391897, "grad_norm": 2.3884706497192383, "learning_rate": 3.5907237512742105e-06, "loss": 0.0006, "step": 2515 }, { "epoch": 0.6411007771690661, "grad_norm": 4.421568870544434, "learning_rate": 3.588175331294598e-06, "loss": 0.0015, "step": 2516 }, { "epoch": 0.6413555866989425, "grad_norm": 4.798062324523926, "learning_rate": 3.585626911314985e-06, "loss": 0.0017, "step": 2517 }, { "epoch": 0.641610396228819, "grad_norm": 2.7754719257354736, "learning_rate": 3.583078491335372e-06, "loss": 0.0007, "step": 2518 }, { "epoch": 0.6418652057586953, "grad_norm": 2.20998215675354, "learning_rate": 3.5805300713557595e-06, "loss": 0.0005, "step": 2519 }, { "epoch": 0.6421200152885718, "grad_norm": 1.3104406595230103, "learning_rate": 3.5779816513761473e-06, "loss": 0.0004, "step": 2520 }, { "epoch": 0.6423748248184482, "grad_norm": 0.5615341663360596, "learning_rate": 3.5754332313965347e-06, "loss": 0.0002, "step": 2521 }, { "epoch": 0.6426296343483247, "grad_norm": 1.8510180711746216, "learning_rate": 3.5728848114169217e-06, "loss": 0.0004, "step": 2522 }, { "epoch": 0.6428844438782011, "grad_norm": 0.5563147664070129, "learning_rate": 3.570336391437309e-06, "loss": 0.0001, "step": 2523 }, { "epoch": 0.6431392534080774, "grad_norm": 1.4796444177627563, "learning_rate": 3.5677879714576964e-06, "loss": 0.0003, "step": 2524 }, { "epoch": 0.6433940629379539, "grad_norm": 1.1180038452148438, "learning_rate": 3.5652395514780842e-06, "loss": 0.0003, "step": 2525 }, { "epoch": 0.6436488724678303, "grad_norm": 1.1200588941574097, "learning_rate": 3.562691131498471e-06, "loss": 0.0003, "step": 2526 }, { "epoch": 0.6439036819977068, "grad_norm": 1.3142926692962646, "learning_rate": 3.5601427115188586e-06, "loss": 0.0006, "step": 2527 }, { "epoch": 0.6441584915275831, "grad_norm": 2.9606964588165283, "learning_rate": 3.557594291539246e-06, "loss": 0.0008, "step": 2528 }, { "epoch": 0.6444133010574595, "grad_norm": 2.8020715713500977, "learning_rate": 3.5550458715596333e-06, "loss": 0.0011, "step": 2529 }, { "epoch": 0.644668110587336, "grad_norm": 1.1050934791564941, "learning_rate": 3.552497451580021e-06, "loss": 0.0002, "step": 2530 }, { "epoch": 0.6449229201172124, "grad_norm": 1.4872971773147583, "learning_rate": 3.549949031600408e-06, "loss": 0.0005, "step": 2531 }, { "epoch": 0.6451777296470887, "grad_norm": 3.6974549293518066, "learning_rate": 3.5474006116207954e-06, "loss": 0.0011, "step": 2532 }, { "epoch": 0.6454325391769652, "grad_norm": 4.450881481170654, "learning_rate": 3.544852191641183e-06, "loss": 0.0015, "step": 2533 }, { "epoch": 0.6456873487068416, "grad_norm": 6.829345226287842, "learning_rate": 3.5423037716615698e-06, "loss": 0.0034, "step": 2534 }, { "epoch": 0.6459421582367181, "grad_norm": 4.616754531860352, "learning_rate": 3.5397553516819576e-06, "loss": 0.002, "step": 2535 }, { "epoch": 0.6461969677665945, "grad_norm": 4.234733581542969, "learning_rate": 3.537206931702345e-06, "loss": 0.0014, "step": 2536 }, { "epoch": 0.6464517772964709, "grad_norm": 1.4784519672393799, "learning_rate": 3.5346585117227323e-06, "loss": 0.0003, "step": 2537 }, { "epoch": 0.6467065868263473, "grad_norm": 0.5584776997566223, "learning_rate": 3.5321100917431193e-06, "loss": 0.0001, "step": 2538 }, { "epoch": 0.6469613963562237, "grad_norm": 58.19026184082031, "learning_rate": 3.5295616717635066e-06, "loss": 0.6525, "step": 2539 }, { "epoch": 0.6472162058861002, "grad_norm": 1.2985326051712036, "learning_rate": 3.5270132517838944e-06, "loss": 0.0002, "step": 2540 }, { "epoch": 0.6474710154159765, "grad_norm": 5.004227638244629, "learning_rate": 3.524464831804282e-06, "loss": 0.0019, "step": 2541 }, { "epoch": 0.647725824945853, "grad_norm": 0.9364778995513916, "learning_rate": 3.521916411824669e-06, "loss": 0.0005, "step": 2542 }, { "epoch": 0.6479806344757294, "grad_norm": 0.9256999492645264, "learning_rate": 3.519367991845056e-06, "loss": 0.0002, "step": 2543 }, { "epoch": 0.6482354440056058, "grad_norm": 2.405369520187378, "learning_rate": 3.5168195718654435e-06, "loss": 0.0006, "step": 2544 }, { "epoch": 0.6484902535354822, "grad_norm": 2.436110258102417, "learning_rate": 3.5142711518858313e-06, "loss": 0.001, "step": 2545 }, { "epoch": 0.6487450630653586, "grad_norm": 1.3004530668258667, "learning_rate": 3.5117227319062187e-06, "loss": 0.0004, "step": 2546 }, { "epoch": 0.6489998725952351, "grad_norm": 0.9197123646736145, "learning_rate": 3.5091743119266056e-06, "loss": 0.0001, "step": 2547 }, { "epoch": 0.6492546821251115, "grad_norm": 1.8419049978256226, "learning_rate": 3.506625891946993e-06, "loss": 0.0003, "step": 2548 }, { "epoch": 0.649509491654988, "grad_norm": 0.37618303298950195, "learning_rate": 3.5040774719673804e-06, "loss": 0.0001, "step": 2549 }, { "epoch": 0.6497643011848643, "grad_norm": 0.11108661442995071, "learning_rate": 3.5015290519877678e-06, "loss": 0.0002, "step": 2550 }, { "epoch": 0.6500191107147407, "grad_norm": 1.302750825881958, "learning_rate": 3.4989806320081556e-06, "loss": 0.0002, "step": 2551 }, { "epoch": 0.6502739202446172, "grad_norm": 3.3123939037323, "learning_rate": 3.4964322120285425e-06, "loss": 0.0009, "step": 2552 }, { "epoch": 0.6505287297744936, "grad_norm": 1.8411625623703003, "learning_rate": 3.49388379204893e-06, "loss": 0.0004, "step": 2553 }, { "epoch": 0.6507835393043699, "grad_norm": 3.139802932739258, "learning_rate": 3.4913353720693173e-06, "loss": 0.0013, "step": 2554 }, { "epoch": 0.6510383488342464, "grad_norm": 0.1927756518125534, "learning_rate": 3.4887869520897042e-06, "loss": 0.0004, "step": 2555 }, { "epoch": 0.6512931583641228, "grad_norm": 1.6647663116455078, "learning_rate": 3.486238532110092e-06, "loss": 0.0003, "step": 2556 }, { "epoch": 0.6515479678939993, "grad_norm": 3.309102773666382, "learning_rate": 3.4836901121304794e-06, "loss": 0.0009, "step": 2557 }, { "epoch": 0.6518027774238756, "grad_norm": 0.1879521757364273, "learning_rate": 3.4811416921508668e-06, "loss": 0.0001, "step": 2558 }, { "epoch": 0.652057586953752, "grad_norm": 0.5572286248207092, "learning_rate": 3.478593272171254e-06, "loss": 0.0001, "step": 2559 }, { "epoch": 0.6523123964836285, "grad_norm": 1.1049433946609497, "learning_rate": 3.476044852191641e-06, "loss": 0.0002, "step": 2560 }, { "epoch": 0.6525672060135049, "grad_norm": 1.6647289991378784, "learning_rate": 3.473496432212029e-06, "loss": 0.0003, "step": 2561 }, { "epoch": 0.6528220155433814, "grad_norm": 0.5797125697135925, "learning_rate": 3.4709480122324163e-06, "loss": 0.0003, "step": 2562 }, { "epoch": 0.6530768250732577, "grad_norm": 0.5981654524803162, "learning_rate": 3.4683995922528037e-06, "loss": 0.0004, "step": 2563 }, { "epoch": 0.6533316346031341, "grad_norm": 3.4874136447906494, "learning_rate": 3.4658511722731906e-06, "loss": 0.001, "step": 2564 }, { "epoch": 0.6535864441330106, "grad_norm": 2.7622294425964355, "learning_rate": 3.463302752293578e-06, "loss": 0.0008, "step": 2565 }, { "epoch": 0.653841253662887, "grad_norm": 1.6705995798110962, "learning_rate": 3.460754332313966e-06, "loss": 0.0003, "step": 2566 }, { "epoch": 0.6540960631927634, "grad_norm": 4.619772434234619, "learning_rate": 3.458205912334353e-06, "loss": 0.0016, "step": 2567 }, { "epoch": 0.6543508727226398, "grad_norm": 0.9272757172584534, "learning_rate": 3.4556574923547405e-06, "loss": 0.0002, "step": 2568 }, { "epoch": 0.6546056822525163, "grad_norm": 1.1054797172546387, "learning_rate": 3.4531090723751275e-06, "loss": 0.0002, "step": 2569 }, { "epoch": 0.6548604917823927, "grad_norm": 1.8539330959320068, "learning_rate": 3.450560652395515e-06, "loss": 0.0005, "step": 2570 }, { "epoch": 0.655115301312269, "grad_norm": 1.6660480499267578, "learning_rate": 3.4480122324159027e-06, "loss": 0.0003, "step": 2571 }, { "epoch": 0.6553701108421455, "grad_norm": 2.0373616218566895, "learning_rate": 3.44546381243629e-06, "loss": 0.0004, "step": 2572 }, { "epoch": 0.6556249203720219, "grad_norm": 0.7408754825592041, "learning_rate": 3.442915392456677e-06, "loss": 0.0002, "step": 2573 }, { "epoch": 0.6558797299018984, "grad_norm": 1.1052600145339966, "learning_rate": 3.4403669724770644e-06, "loss": 0.0002, "step": 2574 }, { "epoch": 0.6561345394317748, "grad_norm": 1.2900328636169434, "learning_rate": 3.4378185524974518e-06, "loss": 0.0002, "step": 2575 }, { "epoch": 0.6563893489616511, "grad_norm": 3.325385570526123, "learning_rate": 3.4352701325178396e-06, "loss": 0.0009, "step": 2576 }, { "epoch": 0.6566441584915276, "grad_norm": 62.122894287109375, "learning_rate": 3.432721712538227e-06, "loss": 0.6412, "step": 2577 }, { "epoch": 0.656898968021404, "grad_norm": 2.5853054523468018, "learning_rate": 3.430173292558614e-06, "loss": 0.0006, "step": 2578 }, { "epoch": 0.6571537775512805, "grad_norm": 2.226316452026367, "learning_rate": 3.4276248725790013e-06, "loss": 0.0007, "step": 2579 }, { "epoch": 0.6574085870811568, "grad_norm": 0.21526135504245758, "learning_rate": 3.4250764525993886e-06, "loss": 0.0001, "step": 2580 }, { "epoch": 0.6576633966110332, "grad_norm": 0.9277816414833069, "learning_rate": 3.4225280326197764e-06, "loss": 0.0002, "step": 2581 }, { "epoch": 0.6579182061409097, "grad_norm": 2.956005334854126, "learning_rate": 3.4199796126401634e-06, "loss": 0.001, "step": 2582 }, { "epoch": 0.6581730156707861, "grad_norm": 1.479380488395691, "learning_rate": 3.4174311926605508e-06, "loss": 0.0003, "step": 2583 }, { "epoch": 0.6584278252006625, "grad_norm": 1.2899665832519531, "learning_rate": 3.414882772680938e-06, "loss": 0.0003, "step": 2584 }, { "epoch": 0.6586826347305389, "grad_norm": 2.76055908203125, "learning_rate": 3.412334352701325e-06, "loss": 0.0007, "step": 2585 }, { "epoch": 0.6589374442604153, "grad_norm": 1.6565109491348267, "learning_rate": 3.4097859327217125e-06, "loss": 0.0003, "step": 2586 }, { "epoch": 0.6591922537902918, "grad_norm": 1.8425507545471191, "learning_rate": 3.4072375127421003e-06, "loss": 0.0003, "step": 2587 }, { "epoch": 0.6594470633201682, "grad_norm": 0.9266570210456848, "learning_rate": 3.4046890927624876e-06, "loss": 0.0002, "step": 2588 }, { "epoch": 0.6597018728500446, "grad_norm": 4.606771469116211, "learning_rate": 3.402140672782875e-06, "loss": 0.0016, "step": 2589 }, { "epoch": 0.659956682379921, "grad_norm": 2.9582157135009766, "learning_rate": 3.399592252803262e-06, "loss": 0.0009, "step": 2590 }, { "epoch": 0.6602114919097974, "grad_norm": 0.7377771139144897, "learning_rate": 3.3970438328236493e-06, "loss": 0.0003, "step": 2591 }, { "epoch": 0.6604663014396739, "grad_norm": 4.936312675476074, "learning_rate": 3.394495412844037e-06, "loss": 0.0017, "step": 2592 }, { "epoch": 0.6607211109695502, "grad_norm": 3.325269937515259, "learning_rate": 3.3919469928644245e-06, "loss": 0.0009, "step": 2593 }, { "epoch": 0.6609759204994267, "grad_norm": 350.6297302246094, "learning_rate": 3.3893985728848115e-06, "loss": 0.131, "step": 2594 }, { "epoch": 0.6612307300293031, "grad_norm": 2.2238998413085938, "learning_rate": 3.386850152905199e-06, "loss": 0.0005, "step": 2595 }, { "epoch": 0.6614855395591795, "grad_norm": 0.370735228061676, "learning_rate": 3.3843017329255862e-06, "loss": 0.0001, "step": 2596 }, { "epoch": 0.6617403490890559, "grad_norm": 1.8494324684143066, "learning_rate": 3.381753312945974e-06, "loss": 0.0005, "step": 2597 }, { "epoch": 0.6619951586189323, "grad_norm": 1.8429814577102661, "learning_rate": 3.3792048929663614e-06, "loss": 0.0003, "step": 2598 }, { "epoch": 0.6622499681488088, "grad_norm": 1.3152450323104858, "learning_rate": 3.3766564729867484e-06, "loss": 0.0003, "step": 2599 }, { "epoch": 0.6625047776786852, "grad_norm": 0.9481335878372192, "learning_rate": 3.3741080530071357e-06, "loss": 0.0007, "step": 2600 }, { "epoch": 0.6627595872085617, "grad_norm": 0.9246453642845154, "learning_rate": 3.371559633027523e-06, "loss": 0.0002, "step": 2601 }, { "epoch": 0.663014396738438, "grad_norm": 0.7420879006385803, "learning_rate": 3.369011213047911e-06, "loss": 0.0002, "step": 2602 }, { "epoch": 0.6632692062683144, "grad_norm": 1.8446993827819824, "learning_rate": 3.366462793068298e-06, "loss": 0.0004, "step": 2603 }, { "epoch": 0.6635240157981909, "grad_norm": 0.9205240607261658, "learning_rate": 3.3639143730886852e-06, "loss": 0.0002, "step": 2604 }, { "epoch": 0.6637788253280673, "grad_norm": 2.4065017700195312, "learning_rate": 3.3613659531090726e-06, "loss": 0.0005, "step": 2605 }, { "epoch": 0.6640336348579436, "grad_norm": 0.19366832077503204, "learning_rate": 3.35881753312946e-06, "loss": 0.0001, "step": 2606 }, { "epoch": 0.6642884443878201, "grad_norm": 0.738715648651123, "learning_rate": 3.3562691131498478e-06, "loss": 0.0001, "step": 2607 }, { "epoch": 0.6645432539176965, "grad_norm": 3.8630173206329346, "learning_rate": 3.3537206931702347e-06, "loss": 0.0012, "step": 2608 }, { "epoch": 0.664798063447573, "grad_norm": 3.1329448223114014, "learning_rate": 3.351172273190622e-06, "loss": 0.0008, "step": 2609 }, { "epoch": 0.6650528729774493, "grad_norm": 3.1373934745788574, "learning_rate": 3.3486238532110095e-06, "loss": 0.0008, "step": 2610 }, { "epoch": 0.6653076825073257, "grad_norm": 4.230576515197754, "learning_rate": 3.3460754332313964e-06, "loss": 0.0016, "step": 2611 }, { "epoch": 0.6655624920372022, "grad_norm": 3.5054819583892822, "learning_rate": 3.3435270132517842e-06, "loss": 0.001, "step": 2612 }, { "epoch": 0.6658173015670786, "grad_norm": 0.5565316081047058, "learning_rate": 3.3409785932721716e-06, "loss": 0.0001, "step": 2613 }, { "epoch": 0.6660721110969551, "grad_norm": 0.10795643925666809, "learning_rate": 3.338430173292559e-06, "loss": 0.0002, "step": 2614 }, { "epoch": 0.6663269206268314, "grad_norm": 0.9273853302001953, "learning_rate": 3.3358817533129464e-06, "loss": 0.0003, "step": 2615 }, { "epoch": 0.6665817301567079, "grad_norm": 2.954559326171875, "learning_rate": 3.3333333333333333e-06, "loss": 0.0008, "step": 2616 }, { "epoch": 0.6668365396865843, "grad_norm": 1.1079092025756836, "learning_rate": 3.330784913353721e-06, "loss": 0.0002, "step": 2617 }, { "epoch": 0.6670913492164607, "grad_norm": 1.213457703590393, "learning_rate": 3.3282364933741085e-06, "loss": 0.0006, "step": 2618 }, { "epoch": 0.6673461587463371, "grad_norm": 1.6553798913955688, "learning_rate": 3.325688073394496e-06, "loss": 0.0003, "step": 2619 }, { "epoch": 0.6676009682762135, "grad_norm": 22.050243377685547, "learning_rate": 3.323139653414883e-06, "loss": 0.0064, "step": 2620 }, { "epoch": 0.66785577780609, "grad_norm": 0.18599623441696167, "learning_rate": 3.32059123343527e-06, "loss": 0.0001, "step": 2621 }, { "epoch": 0.6681105873359664, "grad_norm": 0.011354401707649231, "learning_rate": 3.318042813455658e-06, "loss": 0.0001, "step": 2622 }, { "epoch": 0.6683653968658427, "grad_norm": 1.49033522605896, "learning_rate": 3.3154943934760454e-06, "loss": 0.0006, "step": 2623 }, { "epoch": 0.6686202063957192, "grad_norm": 1.4768147468566895, "learning_rate": 3.3129459734964323e-06, "loss": 0.0003, "step": 2624 }, { "epoch": 0.6688750159255956, "grad_norm": 6.619290351867676, "learning_rate": 3.3103975535168197e-06, "loss": 0.0067, "step": 2625 }, { "epoch": 0.6691298254554721, "grad_norm": 0.36993342638015747, "learning_rate": 3.307849133537207e-06, "loss": 0.0001, "step": 2626 }, { "epoch": 0.6693846349853485, "grad_norm": 2.764841079711914, "learning_rate": 3.3053007135575945e-06, "loss": 0.0006, "step": 2627 }, { "epoch": 0.6696394445152248, "grad_norm": 54.851844787597656, "learning_rate": 3.3027522935779823e-06, "loss": 0.6213, "step": 2628 }, { "epoch": 0.6698942540451013, "grad_norm": 2.4016246795654297, "learning_rate": 3.3002038735983692e-06, "loss": 0.0005, "step": 2629 }, { "epoch": 0.6701490635749777, "grad_norm": 4.627614974975586, "learning_rate": 3.2976554536187566e-06, "loss": 0.002, "step": 2630 }, { "epoch": 0.6704038731048542, "grad_norm": 1.8456484079360962, "learning_rate": 3.295107033639144e-06, "loss": 0.0003, "step": 2631 }, { "epoch": 0.6706586826347305, "grad_norm": 1.0748481750488281, "learning_rate": 3.292558613659531e-06, "loss": 0.001, "step": 2632 }, { "epoch": 0.6709134921646069, "grad_norm": 0.19113419950008392, "learning_rate": 3.2900101936799187e-06, "loss": 0.0001, "step": 2633 }, { "epoch": 0.6711683016944834, "grad_norm": 0.7425107955932617, "learning_rate": 3.287461773700306e-06, "loss": 0.0002, "step": 2634 }, { "epoch": 0.6714231112243598, "grad_norm": 1.658348798751831, "learning_rate": 3.2849133537206935e-06, "loss": 0.0005, "step": 2635 }, { "epoch": 0.6716779207542362, "grad_norm": 0.7451016306877136, "learning_rate": 3.282364933741081e-06, "loss": 0.0002, "step": 2636 }, { "epoch": 0.6719327302841126, "grad_norm": 1.658584713935852, "learning_rate": 3.279816513761468e-06, "loss": 0.0003, "step": 2637 }, { "epoch": 0.672187539813989, "grad_norm": 0.562913179397583, "learning_rate": 3.2772680937818556e-06, "loss": 0.0002, "step": 2638 }, { "epoch": 0.6724423493438655, "grad_norm": 1.2910484075546265, "learning_rate": 3.274719673802243e-06, "loss": 0.0002, "step": 2639 }, { "epoch": 0.6726971588737419, "grad_norm": 1.2888448238372803, "learning_rate": 3.2721712538226303e-06, "loss": 0.0002, "step": 2640 }, { "epoch": 0.6729519684036183, "grad_norm": 1.1650270223617554, "learning_rate": 3.2696228338430173e-06, "loss": 0.0006, "step": 2641 }, { "epoch": 0.6732067779334947, "grad_norm": 1.2900813817977905, "learning_rate": 3.2670744138634047e-06, "loss": 0.0002, "step": 2642 }, { "epoch": 0.6734615874633711, "grad_norm": 1.8655569553375244, "learning_rate": 3.2645259938837925e-06, "loss": 0.0007, "step": 2643 }, { "epoch": 0.6737163969932476, "grad_norm": 0.03959796205163002, "learning_rate": 3.26197757390418e-06, "loss": 0.0001, "step": 2644 }, { "epoch": 0.6739712065231239, "grad_norm": 0.1998073309659958, "learning_rate": 3.2594291539245672e-06, "loss": 0.0001, "step": 2645 }, { "epoch": 0.6742260160530004, "grad_norm": 0.7559089660644531, "learning_rate": 3.256880733944954e-06, "loss": 0.0003, "step": 2646 }, { "epoch": 0.6744808255828768, "grad_norm": 1.1100208759307861, "learning_rate": 3.2543323139653416e-06, "loss": 0.0002, "step": 2647 }, { "epoch": 0.6747356351127533, "grad_norm": 4.223999500274658, "learning_rate": 3.2517838939857294e-06, "loss": 0.0014, "step": 2648 }, { "epoch": 0.6749904446426296, "grad_norm": 4.351830959320068, "learning_rate": 3.2492354740061167e-06, "loss": 0.0015, "step": 2649 }, { "epoch": 0.675245254172506, "grad_norm": 3.31461501121521, "learning_rate": 3.2466870540265037e-06, "loss": 0.0009, "step": 2650 }, { "epoch": 0.6755000637023825, "grad_norm": 0.16628028452396393, "learning_rate": 3.244138634046891e-06, "loss": 0.0003, "step": 2651 }, { "epoch": 0.6757548732322589, "grad_norm": 0.3781527280807495, "learning_rate": 3.2415902140672784e-06, "loss": 0.0001, "step": 2652 }, { "epoch": 0.6760096827621354, "grad_norm": 0.3601059317588806, "learning_rate": 3.2390417940876662e-06, "loss": 0.0005, "step": 2653 }, { "epoch": 0.6762644922920117, "grad_norm": 1.4903794527053833, "learning_rate": 3.2364933741080536e-06, "loss": 0.0004, "step": 2654 }, { "epoch": 0.6765193018218881, "grad_norm": 1.8437386751174927, "learning_rate": 3.2339449541284406e-06, "loss": 0.0004, "step": 2655 }, { "epoch": 0.6767741113517646, "grad_norm": 0.7409384250640869, "learning_rate": 3.231396534148828e-06, "loss": 0.0001, "step": 2656 }, { "epoch": 0.677028920881641, "grad_norm": 0.37725725769996643, "learning_rate": 3.2288481141692153e-06, "loss": 0.0002, "step": 2657 }, { "epoch": 0.6772837304115173, "grad_norm": 0.5533708333969116, "learning_rate": 3.226299694189603e-06, "loss": 0.0001, "step": 2658 }, { "epoch": 0.6775385399413938, "grad_norm": 3.4999425411224365, "learning_rate": 3.22375127420999e-06, "loss": 0.0011, "step": 2659 }, { "epoch": 0.6777933494712702, "grad_norm": 2.9672083854675293, "learning_rate": 3.2212028542303774e-06, "loss": 0.0009, "step": 2660 }, { "epoch": 0.6780481590011467, "grad_norm": 0.3699132204055786, "learning_rate": 3.218654434250765e-06, "loss": 0.0001, "step": 2661 }, { "epoch": 0.678302968531023, "grad_norm": 1.481250286102295, "learning_rate": 3.216106014271152e-06, "loss": 0.0002, "step": 2662 }, { "epoch": 0.6785577780608995, "grad_norm": 1.4766701459884644, "learning_rate": 3.213557594291539e-06, "loss": 0.0002, "step": 2663 }, { "epoch": 0.6788125875907759, "grad_norm": 2.04038667678833, "learning_rate": 3.211009174311927e-06, "loss": 0.0007, "step": 2664 }, { "epoch": 0.6790673971206523, "grad_norm": 3.886486768722534, "learning_rate": 3.2084607543323143e-06, "loss": 0.0013, "step": 2665 }, { "epoch": 0.6793222066505288, "grad_norm": 2.760331869125366, "learning_rate": 3.2059123343527017e-06, "loss": 0.0008, "step": 2666 }, { "epoch": 0.6795770161804051, "grad_norm": 2.956921339035034, "learning_rate": 3.2033639143730887e-06, "loss": 0.0008, "step": 2667 }, { "epoch": 0.6798318257102816, "grad_norm": 0.42820999026298523, "learning_rate": 3.200815494393476e-06, "loss": 0.0003, "step": 2668 }, { "epoch": 0.680086635240158, "grad_norm": 0.5600118041038513, "learning_rate": 3.198267074413864e-06, "loss": 0.0001, "step": 2669 }, { "epoch": 0.6803414447700344, "grad_norm": 0.009283067658543587, "learning_rate": 3.195718654434251e-06, "loss": 0.0, "step": 2670 }, { "epoch": 0.6805962542999108, "grad_norm": 2.9366159439086914, "learning_rate": 3.193170234454638e-06, "loss": 0.0007, "step": 2671 }, { "epoch": 0.6808510638297872, "grad_norm": 1.2994444370269775, "learning_rate": 3.1906218144750255e-06, "loss": 0.0002, "step": 2672 }, { "epoch": 0.6811058733596637, "grad_norm": 4.853669166564941, "learning_rate": 3.188073394495413e-06, "loss": 0.0024, "step": 2673 }, { "epoch": 0.6813606828895401, "grad_norm": 0.5524427890777588, "learning_rate": 3.1855249745158007e-06, "loss": 0.0001, "step": 2674 }, { "epoch": 0.6816154924194164, "grad_norm": 0.7540398240089417, "learning_rate": 3.182976554536188e-06, "loss": 0.0002, "step": 2675 }, { "epoch": 0.6818703019492929, "grad_norm": 2.2180075645446777, "learning_rate": 3.180428134556575e-06, "loss": 0.0005, "step": 2676 }, { "epoch": 0.6821251114791693, "grad_norm": 1.5934773683547974, "learning_rate": 3.1778797145769624e-06, "loss": 0.0006, "step": 2677 }, { "epoch": 0.6823799210090458, "grad_norm": 2.0472419261932373, "learning_rate": 3.17533129459735e-06, "loss": 0.0009, "step": 2678 }, { "epoch": 0.6826347305389222, "grad_norm": 1.3177955150604248, "learning_rate": 3.1727828746177376e-06, "loss": 0.0002, "step": 2679 }, { "epoch": 0.6828895400687985, "grad_norm": 0.7373382449150085, "learning_rate": 3.1702344546381245e-06, "loss": 0.0001, "step": 2680 }, { "epoch": 0.683144349598675, "grad_norm": 0.36746442317962646, "learning_rate": 3.167686034658512e-06, "loss": 0.0001, "step": 2681 }, { "epoch": 0.6833991591285514, "grad_norm": 118.81938934326172, "learning_rate": 3.1651376146788993e-06, "loss": 0.3127, "step": 2682 }, { "epoch": 0.6836539686584279, "grad_norm": 0.9219802021980286, "learning_rate": 3.1625891946992867e-06, "loss": 0.0001, "step": 2683 }, { "epoch": 0.6839087781883042, "grad_norm": 1.4919533729553223, "learning_rate": 3.1600407747196745e-06, "loss": 0.0003, "step": 2684 }, { "epoch": 0.6841635877181806, "grad_norm": 2.769233226776123, "learning_rate": 3.1574923547400614e-06, "loss": 0.0007, "step": 2685 }, { "epoch": 0.6844183972480571, "grad_norm": 57.49580001831055, "learning_rate": 3.154943934760449e-06, "loss": 0.7152, "step": 2686 }, { "epoch": 0.6846732067779335, "grad_norm": 1.4782750606536865, "learning_rate": 3.152395514780836e-06, "loss": 0.0004, "step": 2687 }, { "epoch": 0.6849280163078099, "grad_norm": 1.1041438579559326, "learning_rate": 3.149847094801223e-06, "loss": 0.0002, "step": 2688 }, { "epoch": 0.6851828258376863, "grad_norm": 0.7670364379882812, "learning_rate": 3.147298674821611e-06, "loss": 0.0002, "step": 2689 }, { "epoch": 0.6854376353675627, "grad_norm": 0.3709690570831299, "learning_rate": 3.1447502548419983e-06, "loss": 0.0001, "step": 2690 }, { "epoch": 0.6856924448974392, "grad_norm": 0.554283857345581, "learning_rate": 3.1422018348623857e-06, "loss": 0.0001, "step": 2691 }, { "epoch": 0.6859472544273156, "grad_norm": 2.587843656539917, "learning_rate": 3.139653414882773e-06, "loss": 0.0007, "step": 2692 }, { "epoch": 0.686202063957192, "grad_norm": 0.7511634826660156, "learning_rate": 3.13710499490316e-06, "loss": 0.0003, "step": 2693 }, { "epoch": 0.6864568734870684, "grad_norm": 2.0288054943084717, "learning_rate": 3.134556574923548e-06, "loss": 0.0004, "step": 2694 }, { "epoch": 0.6867116830169449, "grad_norm": 2.7760891914367676, "learning_rate": 3.132008154943935e-06, "loss": 0.0006, "step": 2695 }, { "epoch": 0.6869664925468213, "grad_norm": 1.840688943862915, "learning_rate": 3.1294597349643226e-06, "loss": 0.0003, "step": 2696 }, { "epoch": 0.6872213020766976, "grad_norm": 1.4751856327056885, "learning_rate": 3.1269113149847095e-06, "loss": 0.0003, "step": 2697 }, { "epoch": 0.6874761116065741, "grad_norm": 0.1984991878271103, "learning_rate": 3.124362895005097e-06, "loss": 0.0001, "step": 2698 }, { "epoch": 0.6877309211364505, "grad_norm": 3.7009692192077637, "learning_rate": 3.1218144750254843e-06, "loss": 0.0015, "step": 2699 }, { "epoch": 0.687985730666327, "grad_norm": 4.64009952545166, "learning_rate": 3.119266055045872e-06, "loss": 0.0025, "step": 2700 }, { "epoch": 0.6882405401962033, "grad_norm": 2.3885374069213867, "learning_rate": 3.1167176350662594e-06, "loss": 0.0006, "step": 2701 }, { "epoch": 0.6884953497260797, "grad_norm": 0.9373103380203247, "learning_rate": 3.1141692150866464e-06, "loss": 0.0004, "step": 2702 }, { "epoch": 0.6887501592559562, "grad_norm": 3.1349761486053467, "learning_rate": 3.1116207951070338e-06, "loss": 0.0008, "step": 2703 }, { "epoch": 0.6890049687858326, "grad_norm": 0.38697826862335205, "learning_rate": 3.109072375127421e-06, "loss": 0.0002, "step": 2704 }, { "epoch": 0.6892597783157091, "grad_norm": 1.666070580482483, "learning_rate": 3.106523955147809e-06, "loss": 0.0003, "step": 2705 }, { "epoch": 0.6895145878455854, "grad_norm": 2.5812876224517822, "learning_rate": 3.103975535168196e-06, "loss": 0.0006, "step": 2706 }, { "epoch": 0.6897693973754618, "grad_norm": 0.5598768591880798, "learning_rate": 3.1014271151885833e-06, "loss": 0.0002, "step": 2707 }, { "epoch": 0.6900242069053383, "grad_norm": 0.08400752395391464, "learning_rate": 3.0988786952089706e-06, "loss": 0.0001, "step": 2708 }, { "epoch": 0.6902790164352147, "grad_norm": 2.5887529850006104, "learning_rate": 3.0963302752293576e-06, "loss": 0.0006, "step": 2709 }, { "epoch": 0.690533825965091, "grad_norm": 2.584056854248047, "learning_rate": 3.093781855249746e-06, "loss": 0.0006, "step": 2710 }, { "epoch": 0.6907886354949675, "grad_norm": 2.2141060829162598, "learning_rate": 3.0912334352701328e-06, "loss": 0.0006, "step": 2711 }, { "epoch": 0.6910434450248439, "grad_norm": 0.7676209807395935, "learning_rate": 3.08868501529052e-06, "loss": 0.0005, "step": 2712 }, { "epoch": 0.6912982545547204, "grad_norm": 1.299635887145996, "learning_rate": 3.0861365953109075e-06, "loss": 0.0002, "step": 2713 }, { "epoch": 0.6915530640845967, "grad_norm": 2.5878262519836426, "learning_rate": 3.0835881753312945e-06, "loss": 0.0007, "step": 2714 }, { "epoch": 0.6918078736144732, "grad_norm": 1.4709899425506592, "learning_rate": 3.0810397553516823e-06, "loss": 0.0002, "step": 2715 }, { "epoch": 0.6920626831443496, "grad_norm": 0.7559602856636047, "learning_rate": 3.0784913353720697e-06, "loss": 0.0003, "step": 2716 }, { "epoch": 0.692317492674226, "grad_norm": 2.5897469520568848, "learning_rate": 3.075942915392457e-06, "loss": 0.0006, "step": 2717 }, { "epoch": 0.6925723022041025, "grad_norm": 0.19124576449394226, "learning_rate": 3.073394495412844e-06, "loss": 0.0001, "step": 2718 }, { "epoch": 0.6928271117339788, "grad_norm": 1.1169166564941406, "learning_rate": 3.0708460754332314e-06, "loss": 0.0004, "step": 2719 }, { "epoch": 0.6930819212638553, "grad_norm": 0.5644445419311523, "learning_rate": 3.068297655453619e-06, "loss": 0.0002, "step": 2720 }, { "epoch": 0.6933367307937317, "grad_norm": 3.1347005367279053, "learning_rate": 3.0657492354740065e-06, "loss": 0.0008, "step": 2721 }, { "epoch": 0.6935915403236081, "grad_norm": 0.5748453140258789, "learning_rate": 3.063200815494394e-06, "loss": 0.0002, "step": 2722 }, { "epoch": 0.6938463498534845, "grad_norm": 0.18492399156093597, "learning_rate": 3.060652395514781e-06, "loss": 0.0001, "step": 2723 }, { "epoch": 0.6941011593833609, "grad_norm": 0.31273406744003296, "learning_rate": 3.0581039755351682e-06, "loss": 0.0002, "step": 2724 }, { "epoch": 0.6943559689132374, "grad_norm": 1.1066782474517822, "learning_rate": 3.055555555555556e-06, "loss": 0.0003, "step": 2725 }, { "epoch": 0.6946107784431138, "grad_norm": 1.2931748628616333, "learning_rate": 3.0530071355759434e-06, "loss": 0.0002, "step": 2726 }, { "epoch": 0.6948655879729901, "grad_norm": 0.3696155548095703, "learning_rate": 3.0504587155963304e-06, "loss": 0.0001, "step": 2727 }, { "epoch": 0.6951203975028666, "grad_norm": 0.5549303293228149, "learning_rate": 3.0479102956167177e-06, "loss": 0.0001, "step": 2728 }, { "epoch": 0.695375207032743, "grad_norm": 1.6683541536331177, "learning_rate": 3.045361875637105e-06, "loss": 0.0004, "step": 2729 }, { "epoch": 0.6956300165626195, "grad_norm": 1.6685220003128052, "learning_rate": 3.042813455657493e-06, "loss": 0.0004, "step": 2730 }, { "epoch": 0.6958848260924959, "grad_norm": 0.7968393564224243, "learning_rate": 3.0402650356778803e-06, "loss": 0.0006, "step": 2731 }, { "epoch": 0.6961396356223722, "grad_norm": 1.105386734008789, "learning_rate": 3.0377166156982672e-06, "loss": 0.0002, "step": 2732 }, { "epoch": 0.6963944451522487, "grad_norm": 1.6599414348602295, "learning_rate": 3.0351681957186546e-06, "loss": 0.0004, "step": 2733 }, { "epoch": 0.6966492546821251, "grad_norm": 1.1044732332229614, "learning_rate": 3.032619775739042e-06, "loss": 0.0002, "step": 2734 }, { "epoch": 0.6969040642120016, "grad_norm": 3.31355619430542, "learning_rate": 3.03007135575943e-06, "loss": 0.0009, "step": 2735 }, { "epoch": 0.6971588737418779, "grad_norm": 55.649696350097656, "learning_rate": 3.0275229357798168e-06, "loss": 0.5392, "step": 2736 }, { "epoch": 0.6974136832717543, "grad_norm": 0.9163509607315063, "learning_rate": 3.024974515800204e-06, "loss": 0.0001, "step": 2737 }, { "epoch": 0.6976684928016308, "grad_norm": 0.5595971345901489, "learning_rate": 3.0224260958205915e-06, "loss": 0.0001, "step": 2738 }, { "epoch": 0.6979233023315072, "grad_norm": 1.1276156902313232, "learning_rate": 3.019877675840979e-06, "loss": 0.0003, "step": 2739 }, { "epoch": 0.6981781118613836, "grad_norm": 0.3703216314315796, "learning_rate": 3.017329255861366e-06, "loss": 0.0001, "step": 2740 }, { "epoch": 0.69843292139126, "grad_norm": 2.2044413089752197, "learning_rate": 3.0147808358817536e-06, "loss": 0.0004, "step": 2741 }, { "epoch": 0.6986877309211365, "grad_norm": 0.7358995676040649, "learning_rate": 3.012232415902141e-06, "loss": 0.0001, "step": 2742 }, { "epoch": 0.6989425404510129, "grad_norm": 2.7574915885925293, "learning_rate": 3.0096839959225284e-06, "loss": 0.0006, "step": 2743 }, { "epoch": 0.6991973499808893, "grad_norm": 2.2073373794555664, "learning_rate": 3.0071355759429153e-06, "loss": 0.0005, "step": 2744 }, { "epoch": 0.6994521595107657, "grad_norm": 0.752642035484314, "learning_rate": 3.0045871559633027e-06, "loss": 0.0003, "step": 2745 }, { "epoch": 0.6997069690406421, "grad_norm": 0.558823823928833, "learning_rate": 3.0020387359836905e-06, "loss": 0.0001, "step": 2746 }, { "epoch": 0.6999617785705186, "grad_norm": 2.2159934043884277, "learning_rate": 2.999490316004078e-06, "loss": 0.0006, "step": 2747 }, { "epoch": 0.700216588100395, "grad_norm": 0.9394686818122864, "learning_rate": 2.9969418960244653e-06, "loss": 0.0003, "step": 2748 }, { "epoch": 0.7004713976302713, "grad_norm": 1.1257686614990234, "learning_rate": 2.9943934760448522e-06, "loss": 0.0004, "step": 2749 }, { "epoch": 0.7007262071601478, "grad_norm": 1.104080080986023, "learning_rate": 2.9918450560652396e-06, "loss": 0.0002, "step": 2750 }, { "epoch": 0.7009810166900242, "grad_norm": 0.921023428440094, "learning_rate": 2.9892966360856274e-06, "loss": 0.0001, "step": 2751 }, { "epoch": 0.7012358262199007, "grad_norm": 2.2061266899108887, "learning_rate": 2.9867482161060148e-06, "loss": 0.0004, "step": 2752 }, { "epoch": 0.701490635749777, "grad_norm": 3.5059266090393066, "learning_rate": 2.9841997961264017e-06, "loss": 0.001, "step": 2753 }, { "epoch": 0.7017454452796534, "grad_norm": 2.9614109992980957, "learning_rate": 2.981651376146789e-06, "loss": 0.0009, "step": 2754 }, { "epoch": 0.7020002548095299, "grad_norm": 3.6852972507476807, "learning_rate": 2.9791029561671765e-06, "loss": 0.0011, "step": 2755 }, { "epoch": 0.7022550643394063, "grad_norm": 2.7681446075439453, "learning_rate": 2.9765545361875643e-06, "loss": 0.0007, "step": 2756 }, { "epoch": 0.7025098738692828, "grad_norm": 0.22670501470565796, "learning_rate": 2.9740061162079512e-06, "loss": 0.0002, "step": 2757 }, { "epoch": 0.7027646833991591, "grad_norm": 0.3732622265815735, "learning_rate": 2.9714576962283386e-06, "loss": 0.0001, "step": 2758 }, { "epoch": 0.7030194929290355, "grad_norm": 0.949038028717041, "learning_rate": 2.968909276248726e-06, "loss": 0.0004, "step": 2759 }, { "epoch": 0.703274302458912, "grad_norm": 0.746753990650177, "learning_rate": 2.9663608562691134e-06, "loss": 0.0002, "step": 2760 }, { "epoch": 0.7035291119887884, "grad_norm": 1.3040014505386353, "learning_rate": 2.963812436289501e-06, "loss": 0.0003, "step": 2761 }, { "epoch": 0.7037839215186648, "grad_norm": 1.484684944152832, "learning_rate": 2.961264016309888e-06, "loss": 0.0003, "step": 2762 }, { "epoch": 0.7040387310485412, "grad_norm": 0.38727301359176636, "learning_rate": 2.9587155963302755e-06, "loss": 0.0002, "step": 2763 }, { "epoch": 0.7042935405784176, "grad_norm": 2.2234420776367188, "learning_rate": 2.956167176350663e-06, "loss": 0.0004, "step": 2764 }, { "epoch": 0.7045483501082941, "grad_norm": 3.4916739463806152, "learning_rate": 2.95361875637105e-06, "loss": 0.0014, "step": 2765 }, { "epoch": 0.7048031596381704, "grad_norm": 4.052647590637207, "learning_rate": 2.9510703363914376e-06, "loss": 0.0013, "step": 2766 }, { "epoch": 0.7050579691680469, "grad_norm": 0.39844444394111633, "learning_rate": 2.948521916411825e-06, "loss": 0.0002, "step": 2767 }, { "epoch": 0.7053127786979233, "grad_norm": 1.6797020435333252, "learning_rate": 2.9459734964322124e-06, "loss": 0.0006, "step": 2768 }, { "epoch": 0.7055675882277997, "grad_norm": 2.0313525199890137, "learning_rate": 2.9434250764525997e-06, "loss": 0.0004, "step": 2769 }, { "epoch": 0.7058223977576762, "grad_norm": 0.9212663769721985, "learning_rate": 2.9408766564729867e-06, "loss": 0.0001, "step": 2770 }, { "epoch": 0.7060772072875525, "grad_norm": 0.01346538495272398, "learning_rate": 2.9383282364933745e-06, "loss": 0.0001, "step": 2771 }, { "epoch": 0.706332016817429, "grad_norm": 0.925010621547699, "learning_rate": 2.935779816513762e-06, "loss": 0.0002, "step": 2772 }, { "epoch": 0.7065868263473054, "grad_norm": 0.19160839915275574, "learning_rate": 2.9332313965341492e-06, "loss": 0.0001, "step": 2773 }, { "epoch": 0.7068416358771819, "grad_norm": 0.37146055698394775, "learning_rate": 2.930682976554536e-06, "loss": 0.0001, "step": 2774 }, { "epoch": 0.7070964454070582, "grad_norm": 0.05651962757110596, "learning_rate": 2.9281345565749236e-06, "loss": 0.0001, "step": 2775 }, { "epoch": 0.7073512549369346, "grad_norm": 0.55997633934021, "learning_rate": 2.925586136595311e-06, "loss": 0.0001, "step": 2776 }, { "epoch": 0.7076060644668111, "grad_norm": 0.7598770260810852, "learning_rate": 2.9230377166156987e-06, "loss": 0.0003, "step": 2777 }, { "epoch": 0.7078608739966875, "grad_norm": 1.6579657793045044, "learning_rate": 2.920489296636086e-06, "loss": 0.0004, "step": 2778 }, { "epoch": 0.7081156835265638, "grad_norm": 3.135667562484741, "learning_rate": 2.917940876656473e-06, "loss": 0.0008, "step": 2779 }, { "epoch": 0.7083704930564403, "grad_norm": 1.8459022045135498, "learning_rate": 2.9153924566768605e-06, "loss": 0.0003, "step": 2780 }, { "epoch": 0.7086253025863167, "grad_norm": 2.4255831241607666, "learning_rate": 2.912844036697248e-06, "loss": 0.0005, "step": 2781 }, { "epoch": 0.7088801121161932, "grad_norm": 3.5106241703033447, "learning_rate": 2.9102956167176356e-06, "loss": 0.0013, "step": 2782 }, { "epoch": 0.7091349216460696, "grad_norm": 2.754584789276123, "learning_rate": 2.9077471967380226e-06, "loss": 0.0006, "step": 2783 }, { "epoch": 0.709389731175946, "grad_norm": 2.389066219329834, "learning_rate": 2.90519877675841e-06, "loss": 0.0006, "step": 2784 }, { "epoch": 0.7096445407058224, "grad_norm": 2.958122968673706, "learning_rate": 2.9026503567787973e-06, "loss": 0.0007, "step": 2785 }, { "epoch": 0.7098993502356988, "grad_norm": 1.4069024324417114, "learning_rate": 2.9001019367991847e-06, "loss": 0.0015, "step": 2786 }, { "epoch": 0.7101541597655753, "grad_norm": 2.4190478324890137, "learning_rate": 2.8975535168195725e-06, "loss": 0.0007, "step": 2787 }, { "epoch": 0.7104089692954516, "grad_norm": 1.6684207916259766, "learning_rate": 2.8950050968399595e-06, "loss": 0.0004, "step": 2788 }, { "epoch": 0.710663778825328, "grad_norm": 0.5515109896659851, "learning_rate": 2.892456676860347e-06, "loss": 0.0001, "step": 2789 }, { "epoch": 0.7109185883552045, "grad_norm": 2.213075637817383, "learning_rate": 2.8899082568807342e-06, "loss": 0.0004, "step": 2790 }, { "epoch": 0.7111733978850809, "grad_norm": 1.1036356687545776, "learning_rate": 2.887359836901121e-06, "loss": 0.0001, "step": 2791 }, { "epoch": 0.7114282074149573, "grad_norm": 0.7384249567985535, "learning_rate": 2.884811416921509e-06, "loss": 0.0001, "step": 2792 }, { "epoch": 0.7116830169448337, "grad_norm": 0.921043872833252, "learning_rate": 2.8822629969418963e-06, "loss": 0.0001, "step": 2793 }, { "epoch": 0.7119378264747102, "grad_norm": 0.36829832196235657, "learning_rate": 2.8797145769622837e-06, "loss": 0.0001, "step": 2794 }, { "epoch": 0.7121926360045866, "grad_norm": 1.6664100885391235, "learning_rate": 2.8771661569826707e-06, "loss": 0.0003, "step": 2795 }, { "epoch": 0.712447445534463, "grad_norm": 3.6864166259765625, "learning_rate": 2.874617737003058e-06, "loss": 0.0014, "step": 2796 }, { "epoch": 0.7127022550643394, "grad_norm": 3.1386606693267822, "learning_rate": 2.872069317023446e-06, "loss": 0.0012, "step": 2797 }, { "epoch": 0.7129570645942158, "grad_norm": 2.392627716064453, "learning_rate": 2.8695208970438332e-06, "loss": 0.0005, "step": 2798 }, { "epoch": 0.7132118741240923, "grad_norm": 1.4772202968597412, "learning_rate": 2.8669724770642206e-06, "loss": 0.0003, "step": 2799 }, { "epoch": 0.7134666836539687, "grad_norm": 0.9212695956230164, "learning_rate": 2.8644240570846075e-06, "loss": 0.0001, "step": 2800 }, { "epoch": 0.713721493183845, "grad_norm": 0.3752349615097046, "learning_rate": 2.861875637104995e-06, "loss": 0.0001, "step": 2801 }, { "epoch": 0.7139763027137215, "grad_norm": 1.4683953523635864, "learning_rate": 2.8593272171253827e-06, "loss": 0.0002, "step": 2802 }, { "epoch": 0.7142311122435979, "grad_norm": 0.7395980954170227, "learning_rate": 2.85677879714577e-06, "loss": 0.0001, "step": 2803 }, { "epoch": 0.7144859217734744, "grad_norm": 4.0501017570495605, "learning_rate": 2.854230377166157e-06, "loss": 0.0013, "step": 2804 }, { "epoch": 0.7147407313033507, "grad_norm": 0.9227868318557739, "learning_rate": 2.8516819571865444e-06, "loss": 0.0002, "step": 2805 }, { "epoch": 0.7149955408332271, "grad_norm": 0.9242676496505737, "learning_rate": 2.849133537206932e-06, "loss": 0.0002, "step": 2806 }, { "epoch": 0.7152503503631036, "grad_norm": 2.0333008766174316, "learning_rate": 2.8465851172273196e-06, "loss": 0.0004, "step": 2807 }, { "epoch": 0.71550515989298, "grad_norm": 1.4711722135543823, "learning_rate": 2.844036697247707e-06, "loss": 0.0003, "step": 2808 }, { "epoch": 0.7157599694228565, "grad_norm": 0.18508605659008026, "learning_rate": 2.841488277268094e-06, "loss": 0.0001, "step": 2809 }, { "epoch": 0.7160147789527328, "grad_norm": 1.1068778038024902, "learning_rate": 2.8389398572884813e-06, "loss": 0.0002, "step": 2810 }, { "epoch": 0.7162695884826092, "grad_norm": 0.9313662052154541, "learning_rate": 2.8363914373088687e-06, "loss": 0.0002, "step": 2811 }, { "epoch": 0.7165243980124857, "grad_norm": 1.65117609500885, "learning_rate": 2.8338430173292556e-06, "loss": 0.0002, "step": 2812 }, { "epoch": 0.7167792075423621, "grad_norm": 2.5818111896514893, "learning_rate": 2.8312945973496434e-06, "loss": 0.0005, "step": 2813 }, { "epoch": 0.7170340170722385, "grad_norm": 1.4785382747650146, "learning_rate": 2.828746177370031e-06, "loss": 0.0003, "step": 2814 }, { "epoch": 0.7172888266021149, "grad_norm": 1.6905827522277832, "learning_rate": 2.826197757390418e-06, "loss": 0.001, "step": 2815 }, { "epoch": 0.7175436361319913, "grad_norm": 1.8455923795700073, "learning_rate": 2.8236493374108056e-06, "loss": 0.0004, "step": 2816 }, { "epoch": 0.7177984456618678, "grad_norm": 3.129417657852173, "learning_rate": 2.8211009174311925e-06, "loss": 0.0008, "step": 2817 }, { "epoch": 0.7180532551917441, "grad_norm": 2.205747604370117, "learning_rate": 2.8185524974515803e-06, "loss": 0.0005, "step": 2818 }, { "epoch": 0.7183080647216206, "grad_norm": 4.218335151672363, "learning_rate": 2.8160040774719677e-06, "loss": 0.0014, "step": 2819 }, { "epoch": 0.718562874251497, "grad_norm": 1.8455759286880493, "learning_rate": 2.813455657492355e-06, "loss": 0.0005, "step": 2820 }, { "epoch": 0.7188176837813735, "grad_norm": 1.1049762964248657, "learning_rate": 2.810907237512742e-06, "loss": 0.0001, "step": 2821 }, { "epoch": 0.7190724933112499, "grad_norm": 0.18661470711231232, "learning_rate": 2.8083588175331294e-06, "loss": 0.0, "step": 2822 }, { "epoch": 0.7193273028411262, "grad_norm": 58.37765121459961, "learning_rate": 2.805810397553517e-06, "loss": 0.7502, "step": 2823 }, { "epoch": 0.7195821123710027, "grad_norm": 4.975430011749268, "learning_rate": 2.8032619775739046e-06, "loss": 0.0019, "step": 2824 }, { "epoch": 0.7198369219008791, "grad_norm": 4.971774578094482, "learning_rate": 2.800713557594292e-06, "loss": 0.0019, "step": 2825 }, { "epoch": 0.7200917314307556, "grad_norm": 6.822556495666504, "learning_rate": 2.798165137614679e-06, "loss": 0.0037, "step": 2826 }, { "epoch": 0.7203465409606319, "grad_norm": 4.412660598754883, "learning_rate": 2.7956167176350663e-06, "loss": 0.0015, "step": 2827 }, { "epoch": 0.7206013504905083, "grad_norm": 4.251497745513916, "learning_rate": 2.793068297655454e-06, "loss": 0.0014, "step": 2828 }, { "epoch": 0.7208561600203848, "grad_norm": 3.4951605796813965, "learning_rate": 2.7905198776758415e-06, "loss": 0.0011, "step": 2829 }, { "epoch": 0.7211109695502612, "grad_norm": 1.4690347909927368, "learning_rate": 2.7879714576962284e-06, "loss": 0.0002, "step": 2830 }, { "epoch": 0.7213657790801375, "grad_norm": 1.8456023931503296, "learning_rate": 2.7854230377166158e-06, "loss": 0.0004, "step": 2831 }, { "epoch": 0.721620588610014, "grad_norm": 1.4784091711044312, "learning_rate": 2.782874617737003e-06, "loss": 0.0004, "step": 2832 }, { "epoch": 0.7218753981398904, "grad_norm": 0.5581380128860474, "learning_rate": 2.780326197757391e-06, "loss": 0.0001, "step": 2833 }, { "epoch": 0.7221302076697669, "grad_norm": 1.2869272232055664, "learning_rate": 2.7777777777777783e-06, "loss": 0.0002, "step": 2834 }, { "epoch": 0.7223850171996433, "grad_norm": 0.5524535775184631, "learning_rate": 2.7752293577981653e-06, "loss": 0.0001, "step": 2835 }, { "epoch": 0.7226398267295197, "grad_norm": 0.5572689771652222, "learning_rate": 2.7726809378185527e-06, "loss": 0.0001, "step": 2836 }, { "epoch": 0.7228946362593961, "grad_norm": 1.1166706085205078, "learning_rate": 2.77013251783894e-06, "loss": 0.0005, "step": 2837 }, { "epoch": 0.7231494457892725, "grad_norm": 0.27378368377685547, "learning_rate": 2.767584097859328e-06, "loss": 0.0004, "step": 2838 }, { "epoch": 0.723404255319149, "grad_norm": 0.555590808391571, "learning_rate": 2.765035677879715e-06, "loss": 0.0001, "step": 2839 }, { "epoch": 0.7236590648490253, "grad_norm": 0.5574231743812561, "learning_rate": 2.762487257900102e-06, "loss": 0.0001, "step": 2840 }, { "epoch": 0.7239138743789018, "grad_norm": 0.5523794293403625, "learning_rate": 2.7599388379204895e-06, "loss": 0.0001, "step": 2841 }, { "epoch": 0.7241686839087782, "grad_norm": 0.7662056088447571, "learning_rate": 2.7573904179408765e-06, "loss": 0.0004, "step": 2842 }, { "epoch": 0.7244234934386546, "grad_norm": 0.3738636076450348, "learning_rate": 2.7548419979612643e-06, "loss": 0.0001, "step": 2843 }, { "epoch": 0.724678302968531, "grad_norm": 2.5651469230651855, "learning_rate": 2.7522935779816517e-06, "loss": 0.0006, "step": 2844 }, { "epoch": 0.7249331124984074, "grad_norm": 0.5618276596069336, "learning_rate": 2.749745158002039e-06, "loss": 0.0001, "step": 2845 }, { "epoch": 0.7251879220282839, "grad_norm": 443.3177795410156, "learning_rate": 2.7471967380224264e-06, "loss": 4.4063, "step": 2846 }, { "epoch": 0.7254427315581603, "grad_norm": 0.37396496534347534, "learning_rate": 2.7446483180428134e-06, "loss": 0.0001, "step": 2847 }, { "epoch": 0.7256975410880367, "grad_norm": 2.760106325149536, "learning_rate": 2.742099898063201e-06, "loss": 0.0007, "step": 2848 }, { "epoch": 0.7259523506179131, "grad_norm": 2.0315518379211426, "learning_rate": 2.7395514780835886e-06, "loss": 0.0006, "step": 2849 }, { "epoch": 0.7262071601477895, "grad_norm": 0.9222696423530579, "learning_rate": 2.737003058103976e-06, "loss": 0.0002, "step": 2850 }, { "epoch": 0.726461969677666, "grad_norm": 0.3667704463005066, "learning_rate": 2.734454638124363e-06, "loss": 0.0001, "step": 2851 }, { "epoch": 0.7267167792075424, "grad_norm": 3.132768392562866, "learning_rate": 2.7319062181447503e-06, "loss": 0.0009, "step": 2852 }, { "epoch": 0.7269715887374187, "grad_norm": 0.19122470915317535, "learning_rate": 2.7293577981651376e-06, "loss": 0.0001, "step": 2853 }, { "epoch": 0.7272263982672952, "grad_norm": 1.6647090911865234, "learning_rate": 2.7268093781855254e-06, "loss": 0.0003, "step": 2854 }, { "epoch": 0.7274812077971716, "grad_norm": 2.3837170600891113, "learning_rate": 2.724260958205913e-06, "loss": 0.0005, "step": 2855 }, { "epoch": 0.7277360173270481, "grad_norm": 54.08827590942383, "learning_rate": 2.7217125382262998e-06, "loss": 0.6565, "step": 2856 }, { "epoch": 0.7279908268569244, "grad_norm": 2.0189263820648193, "learning_rate": 2.719164118246687e-06, "loss": 0.0004, "step": 2857 }, { "epoch": 0.7282456363868008, "grad_norm": 1.2824220657348633, "learning_rate": 2.7166156982670745e-06, "loss": 0.0002, "step": 2858 }, { "epoch": 0.7285004459166773, "grad_norm": 0.36930757761001587, "learning_rate": 2.7140672782874623e-06, "loss": 0.0001, "step": 2859 }, { "epoch": 0.7287552554465537, "grad_norm": 0.3694204092025757, "learning_rate": 2.7115188583078493e-06, "loss": 0.0001, "step": 2860 }, { "epoch": 0.7290100649764302, "grad_norm": 0.7343830466270447, "learning_rate": 2.7089704383282366e-06, "loss": 0.0001, "step": 2861 }, { "epoch": 0.7292648745063065, "grad_norm": 0.1177668645977974, "learning_rate": 2.706422018348624e-06, "loss": 0.0002, "step": 2862 }, { "epoch": 0.729519684036183, "grad_norm": 0.23775821924209595, "learning_rate": 2.7038735983690114e-06, "loss": 0.0002, "step": 2863 }, { "epoch": 0.7297744935660594, "grad_norm": 4.235169887542725, "learning_rate": 2.701325178389399e-06, "loss": 0.0016, "step": 2864 }, { "epoch": 0.7300293030959358, "grad_norm": 3.8508594036102295, "learning_rate": 2.698776758409786e-06, "loss": 0.0012, "step": 2865 }, { "epoch": 0.7302841126258122, "grad_norm": 0.6042445302009583, "learning_rate": 2.6962283384301735e-06, "loss": 0.0004, "step": 2866 }, { "epoch": 0.7305389221556886, "grad_norm": 2.3874621391296387, "learning_rate": 2.693679918450561e-06, "loss": 0.0005, "step": 2867 }, { "epoch": 0.730793731685565, "grad_norm": 1.2880795001983643, "learning_rate": 2.691131498470948e-06, "loss": 0.0002, "step": 2868 }, { "epoch": 0.7310485412154415, "grad_norm": 1.2832401990890503, "learning_rate": 2.6885830784913356e-06, "loss": 0.0002, "step": 2869 }, { "epoch": 0.7313033507453178, "grad_norm": 0.576019823551178, "learning_rate": 2.686034658511723e-06, "loss": 0.0003, "step": 2870 }, { "epoch": 0.7315581602751943, "grad_norm": 1.2901780605316162, "learning_rate": 2.6834862385321104e-06, "loss": 0.0002, "step": 2871 }, { "epoch": 0.7318129698050707, "grad_norm": 1.2904059886932373, "learning_rate": 2.6809378185524978e-06, "loss": 0.0002, "step": 2872 }, { "epoch": 0.7320677793349472, "grad_norm": 1.8383139371871948, "learning_rate": 2.6783893985728847e-06, "loss": 0.0003, "step": 2873 }, { "epoch": 0.7323225888648236, "grad_norm": 0.9187649488449097, "learning_rate": 2.6758409785932725e-06, "loss": 0.0001, "step": 2874 }, { "epoch": 0.7325773983946999, "grad_norm": 2.2032926082611084, "learning_rate": 2.67329255861366e-06, "loss": 0.0005, "step": 2875 }, { "epoch": 0.7328322079245764, "grad_norm": 1.8315978050231934, "learning_rate": 2.6707441386340473e-06, "loss": 0.0003, "step": 2876 }, { "epoch": 0.7330870174544528, "grad_norm": 3.678623676300049, "learning_rate": 2.6681957186544342e-06, "loss": 0.0011, "step": 2877 }, { "epoch": 0.7333418269843293, "grad_norm": 2.5791358947753906, "learning_rate": 2.6656472986748216e-06, "loss": 0.0006, "step": 2878 }, { "epoch": 0.7335966365142056, "grad_norm": 0.2540728747844696, "learning_rate": 2.6630988786952094e-06, "loss": 0.0006, "step": 2879 }, { "epoch": 0.733851446044082, "grad_norm": 1.4802099466323853, "learning_rate": 2.6605504587155968e-06, "loss": 0.0004, "step": 2880 }, { "epoch": 0.7341062555739585, "grad_norm": 0.3211107850074768, "learning_rate": 2.6580020387359837e-06, "loss": 0.0004, "step": 2881 }, { "epoch": 0.7343610651038349, "grad_norm": 1.110300064086914, "learning_rate": 2.655453618756371e-06, "loss": 0.0003, "step": 2882 }, { "epoch": 0.7346158746337113, "grad_norm": 0.1909116804599762, "learning_rate": 2.6529051987767585e-06, "loss": 0.0001, "step": 2883 }, { "epoch": 0.7348706841635877, "grad_norm": 1.052180528640747, "learning_rate": 2.6503567787971463e-06, "loss": 0.0008, "step": 2884 }, { "epoch": 0.7351254936934641, "grad_norm": 2.3939321041107178, "learning_rate": 2.6478083588175337e-06, "loss": 0.0006, "step": 2885 }, { "epoch": 0.7353803032233406, "grad_norm": 1.2943540811538696, "learning_rate": 2.6452599388379206e-06, "loss": 0.0005, "step": 2886 }, { "epoch": 0.735635112753217, "grad_norm": 0.3745018541812897, "learning_rate": 2.642711518858308e-06, "loss": 0.0001, "step": 2887 }, { "epoch": 0.7358899222830934, "grad_norm": 1.4891620874404907, "learning_rate": 2.6401630988786954e-06, "loss": 0.0004, "step": 2888 }, { "epoch": 0.7361447318129698, "grad_norm": 0.1840047389268875, "learning_rate": 2.6376146788990823e-06, "loss": 0.0001, "step": 2889 }, { "epoch": 0.7363995413428462, "grad_norm": 2.5771896839141846, "learning_rate": 2.63506625891947e-06, "loss": 0.0006, "step": 2890 }, { "epoch": 0.7366543508727227, "grad_norm": 0.2193981409072876, "learning_rate": 2.6325178389398575e-06, "loss": 0.0002, "step": 2891 }, { "epoch": 0.736909160402599, "grad_norm": 0.5495812892913818, "learning_rate": 2.629969418960245e-06, "loss": 0.0001, "step": 2892 }, { "epoch": 0.7371639699324755, "grad_norm": 2.398303270339966, "learning_rate": 2.6274209989806322e-06, "loss": 0.0006, "step": 2893 }, { "epoch": 0.7374187794623519, "grad_norm": 1.4743530750274658, "learning_rate": 2.624872579001019e-06, "loss": 0.0002, "step": 2894 }, { "epoch": 0.7376735889922283, "grad_norm": 0.7372743487358093, "learning_rate": 2.622324159021407e-06, "loss": 0.0001, "step": 2895 }, { "epoch": 0.7379283985221047, "grad_norm": 0.1515139490365982, "learning_rate": 2.6197757390417944e-06, "loss": 0.0001, "step": 2896 }, { "epoch": 0.7381832080519811, "grad_norm": 5.204218864440918, "learning_rate": 2.6172273190621818e-06, "loss": 0.0023, "step": 2897 }, { "epoch": 0.7384380175818576, "grad_norm": 3.172452449798584, "learning_rate": 2.6146788990825687e-06, "loss": 0.0009, "step": 2898 }, { "epoch": 0.738692827111734, "grad_norm": 0.9168519377708435, "learning_rate": 2.612130479102956e-06, "loss": 0.0001, "step": 2899 }, { "epoch": 0.7389476366416105, "grad_norm": 0.9299772381782532, "learning_rate": 2.609582059123344e-06, "loss": 0.0002, "step": 2900 }, { "epoch": 0.7392024461714868, "grad_norm": 2.9403746128082275, "learning_rate": 2.6070336391437313e-06, "loss": 0.0007, "step": 2901 }, { "epoch": 0.7394572557013632, "grad_norm": 0.013145185075700283, "learning_rate": 2.6044852191641186e-06, "loss": 0.0001, "step": 2902 }, { "epoch": 0.7397120652312397, "grad_norm": 0.3695197105407715, "learning_rate": 2.6019367991845056e-06, "loss": 0.0001, "step": 2903 }, { "epoch": 0.7399668747611161, "grad_norm": 95.12290954589844, "learning_rate": 2.599388379204893e-06, "loss": 4.25, "step": 2904 }, { "epoch": 0.7402216842909924, "grad_norm": 0.9193333387374878, "learning_rate": 2.5968399592252808e-06, "loss": 0.0001, "step": 2905 }, { "epoch": 0.7404764938208689, "grad_norm": 0.5541448593139648, "learning_rate": 2.594291539245668e-06, "loss": 0.0001, "step": 2906 }, { "epoch": 0.7407313033507453, "grad_norm": 0.7347001433372498, "learning_rate": 2.591743119266055e-06, "loss": 0.0001, "step": 2907 }, { "epoch": 0.7409861128806218, "grad_norm": 1.6537377834320068, "learning_rate": 2.5891946992864425e-06, "loss": 0.0008, "step": 2908 }, { "epoch": 0.7412409224104981, "grad_norm": 0.37726157903671265, "learning_rate": 2.58664627930683e-06, "loss": 0.0008, "step": 2909 }, { "epoch": 0.7414957319403745, "grad_norm": 0.1954270750284195, "learning_rate": 2.5840978593272176e-06, "loss": 0.0001, "step": 2910 }, { "epoch": 0.741750541470251, "grad_norm": 56.63234329223633, "learning_rate": 2.581549439347605e-06, "loss": 0.6485, "step": 2911 }, { "epoch": 0.7420053510001274, "grad_norm": 1.2987490892410278, "learning_rate": 2.579001019367992e-06, "loss": 0.0004, "step": 2912 }, { "epoch": 0.7422601605300039, "grad_norm": 0.18614645302295685, "learning_rate": 2.5764525993883793e-06, "loss": 0.0001, "step": 2913 }, { "epoch": 0.7425149700598802, "grad_norm": 1.465818166732788, "learning_rate": 2.5739041794087667e-06, "loss": 0.0003, "step": 2914 }, { "epoch": 0.7427697795897567, "grad_norm": 1.4714927673339844, "learning_rate": 2.5713557594291545e-06, "loss": 0.0003, "step": 2915 }, { "epoch": 0.7430245891196331, "grad_norm": 4.248830318450928, "learning_rate": 2.5688073394495415e-06, "loss": 0.0016, "step": 2916 }, { "epoch": 0.7432793986495095, "grad_norm": 0.736788809299469, "learning_rate": 2.566258919469929e-06, "loss": 0.0002, "step": 2917 }, { "epoch": 0.7435342081793859, "grad_norm": 1.841267704963684, "learning_rate": 2.5637104994903162e-06, "loss": 0.0004, "step": 2918 }, { "epoch": 0.7437890177092623, "grad_norm": 1.1017861366271973, "learning_rate": 2.5611620795107036e-06, "loss": 0.0002, "step": 2919 }, { "epoch": 0.7440438272391388, "grad_norm": 2.223085641860962, "learning_rate": 2.5586136595310914e-06, "loss": 0.0007, "step": 2920 }, { "epoch": 0.7442986367690152, "grad_norm": 1.4740897417068481, "learning_rate": 2.5560652395514784e-06, "loss": 0.0003, "step": 2921 }, { "epoch": 0.7445534462988915, "grad_norm": 1.2832075357437134, "learning_rate": 2.5535168195718657e-06, "loss": 0.0003, "step": 2922 }, { "epoch": 0.744808255828768, "grad_norm": 1.1047935485839844, "learning_rate": 2.550968399592253e-06, "loss": 0.0003, "step": 2923 }, { "epoch": 0.7450630653586444, "grad_norm": 0.368584543466568, "learning_rate": 2.54841997961264e-06, "loss": 0.0001, "step": 2924 }, { "epoch": 0.7453178748885209, "grad_norm": 0.7612758278846741, "learning_rate": 2.5458715596330274e-06, "loss": 0.0006, "step": 2925 }, { "epoch": 0.7455726844183973, "grad_norm": 0.2084607630968094, "learning_rate": 2.5433231396534152e-06, "loss": 0.0002, "step": 2926 }, { "epoch": 0.7458274939482736, "grad_norm": 2.2052268981933594, "learning_rate": 2.5407747196738026e-06, "loss": 0.0005, "step": 2927 }, { "epoch": 0.7460823034781501, "grad_norm": 2.2168970108032227, "learning_rate": 2.5382262996941896e-06, "loss": 0.0006, "step": 2928 }, { "epoch": 0.7463371130080265, "grad_norm": 0.3744010031223297, "learning_rate": 2.535677879714577e-06, "loss": 0.0002, "step": 2929 }, { "epoch": 0.746591922537903, "grad_norm": 1.294124960899353, "learning_rate": 2.5331294597349643e-06, "loss": 0.0004, "step": 2930 }, { "epoch": 0.7468467320677793, "grad_norm": 1.1139129400253296, "learning_rate": 2.530581039755352e-06, "loss": 0.0003, "step": 2931 }, { "epoch": 0.7471015415976557, "grad_norm": 0.918247640132904, "learning_rate": 2.5280326197757395e-06, "loss": 0.0002, "step": 2932 }, { "epoch": 0.7473563511275322, "grad_norm": 1.4757192134857178, "learning_rate": 2.5254841997961264e-06, "loss": 0.0003, "step": 2933 }, { "epoch": 0.7476111606574086, "grad_norm": 2.590879201889038, "learning_rate": 2.522935779816514e-06, "loss": 0.0006, "step": 2934 }, { "epoch": 0.747865970187285, "grad_norm": 0.7568041086196899, "learning_rate": 2.520387359836901e-06, "loss": 0.0004, "step": 2935 }, { "epoch": 0.7481207797171614, "grad_norm": 0.582392156124115, "learning_rate": 2.517838939857289e-06, "loss": 0.0006, "step": 2936 }, { "epoch": 0.7483755892470378, "grad_norm": 0.3702266216278076, "learning_rate": 2.515290519877676e-06, "loss": 0.0002, "step": 2937 }, { "epoch": 0.7486303987769143, "grad_norm": 5.431333065032959, "learning_rate": 2.5127420998980633e-06, "loss": 0.0031, "step": 2938 }, { "epoch": 0.7488852083067907, "grad_norm": 1.316063404083252, "learning_rate": 2.5101936799184507e-06, "loss": 0.0004, "step": 2939 }, { "epoch": 0.7491400178366671, "grad_norm": 0.5510691404342651, "learning_rate": 2.507645259938838e-06, "loss": 0.0001, "step": 2940 }, { "epoch": 0.7493948273665435, "grad_norm": 1.2979425191879272, "learning_rate": 2.505096839959226e-06, "loss": 0.0003, "step": 2941 }, { "epoch": 0.74964963689642, "grad_norm": 0.3698081374168396, "learning_rate": 2.502548419979613e-06, "loss": 0.0001, "step": 2942 }, { "epoch": 0.7499044464262964, "grad_norm": 0.1959816962480545, "learning_rate": 2.5e-06, "loss": 0.0002, "step": 2943 }, { "epoch": 0.7501592559561727, "grad_norm": 0.5953224301338196, "learning_rate": 2.4974515800203876e-06, "loss": 0.0006, "step": 2944 }, { "epoch": 0.7504140654860492, "grad_norm": 0.7531657218933105, "learning_rate": 2.494903160040775e-06, "loss": 0.0003, "step": 2945 }, { "epoch": 0.7506688750159256, "grad_norm": 0.3827698230743408, "learning_rate": 2.4923547400611623e-06, "loss": 0.0002, "step": 2946 }, { "epoch": 0.750923684545802, "grad_norm": 0.18646171689033508, "learning_rate": 2.4898063200815497e-06, "loss": 0.0004, "step": 2947 }, { "epoch": 0.7511784940756784, "grad_norm": 0.1913936734199524, "learning_rate": 2.487257900101937e-06, "loss": 0.0001, "step": 2948 }, { "epoch": 0.7514333036055548, "grad_norm": 1.4748153686523438, "learning_rate": 2.4847094801223245e-06, "loss": 0.0003, "step": 2949 }, { "epoch": 0.7516881131354313, "grad_norm": 2.578857898712158, "learning_rate": 2.482161060142712e-06, "loss": 0.0007, "step": 2950 }, { "epoch": 0.7519429226653077, "grad_norm": 0.7434878945350647, "learning_rate": 2.4796126401630992e-06, "loss": 0.0003, "step": 2951 }, { "epoch": 0.7521977321951842, "grad_norm": 0.03783367574214935, "learning_rate": 2.4770642201834866e-06, "loss": 0.0001, "step": 2952 }, { "epoch": 0.7524525417250605, "grad_norm": 2.0161778926849365, "learning_rate": 2.474515800203874e-06, "loss": 0.0004, "step": 2953 }, { "epoch": 0.7527073512549369, "grad_norm": 0.7692095637321472, "learning_rate": 2.471967380224261e-06, "loss": 0.0005, "step": 2954 }, { "epoch": 0.7529621607848134, "grad_norm": 0.7638339400291443, "learning_rate": 2.4694189602446487e-06, "loss": 0.0003, "step": 2955 }, { "epoch": 0.7532169703146898, "grad_norm": 0.3704288601875305, "learning_rate": 2.4668705402650357e-06, "loss": 0.0001, "step": 2956 }, { "epoch": 0.7534717798445661, "grad_norm": 0.36944642663002014, "learning_rate": 2.464322120285423e-06, "loss": 0.0001, "step": 2957 }, { "epoch": 0.7537265893744426, "grad_norm": 1.101725697517395, "learning_rate": 2.461773700305811e-06, "loss": 0.0002, "step": 2958 }, { "epoch": 0.753981398904319, "grad_norm": 1.109573483467102, "learning_rate": 2.459225280326198e-06, "loss": 0.0002, "step": 2959 }, { "epoch": 0.7542362084341955, "grad_norm": 0.3683070242404938, "learning_rate": 2.4566768603465856e-06, "loss": 0.0001, "step": 2960 }, { "epoch": 0.7544910179640718, "grad_norm": 1.4731336832046509, "learning_rate": 2.4541284403669725e-06, "loss": 0.0003, "step": 2961 }, { "epoch": 0.7547458274939483, "grad_norm": 0.5522871613502502, "learning_rate": 2.45158002038736e-06, "loss": 0.0001, "step": 2962 }, { "epoch": 0.7550006370238247, "grad_norm": 1.1140395402908325, "learning_rate": 2.4490316004077473e-06, "loss": 0.0003, "step": 2963 }, { "epoch": 0.7552554465537011, "grad_norm": 0.5744396448135376, "learning_rate": 2.4464831804281347e-06, "loss": 0.0003, "step": 2964 }, { "epoch": 0.7555102560835776, "grad_norm": 62.92434310913086, "learning_rate": 2.443934760448522e-06, "loss": 0.6641, "step": 2965 }, { "epoch": 0.7557650656134539, "grad_norm": 2.2020983695983887, "learning_rate": 2.4413863404689094e-06, "loss": 0.0005, "step": 2966 }, { "epoch": 0.7560198751433304, "grad_norm": 0.22011248767375946, "learning_rate": 2.438837920489297e-06, "loss": 0.0002, "step": 2967 }, { "epoch": 0.7562746846732068, "grad_norm": 1.100399374961853, "learning_rate": 2.436289500509684e-06, "loss": 0.0002, "step": 2968 }, { "epoch": 0.7565294942030832, "grad_norm": 1.4552977085113525, "learning_rate": 2.4337410805300716e-06, "loss": 0.0007, "step": 2969 }, { "epoch": 0.7567843037329596, "grad_norm": 0.9229510426521301, "learning_rate": 2.431192660550459e-06, "loss": 0.0002, "step": 2970 }, { "epoch": 0.757039113262836, "grad_norm": 1.1055141687393188, "learning_rate": 2.4286442405708463e-06, "loss": 0.0003, "step": 2971 }, { "epoch": 0.7572939227927125, "grad_norm": 1.6613211631774902, "learning_rate": 2.4260958205912337e-06, "loss": 0.0004, "step": 2972 }, { "epoch": 0.7575487323225889, "grad_norm": 0.37615856528282166, "learning_rate": 2.423547400611621e-06, "loss": 0.0002, "step": 2973 }, { "epoch": 0.7578035418524652, "grad_norm": 1.475975751876831, "learning_rate": 2.4209989806320084e-06, "loss": 0.0003, "step": 2974 }, { "epoch": 0.7580583513823417, "grad_norm": 0.9288283586502075, "learning_rate": 2.4184505606523954e-06, "loss": 0.0003, "step": 2975 }, { "epoch": 0.7583131609122181, "grad_norm": 0.21787209808826447, "learning_rate": 2.415902140672783e-06, "loss": 0.0003, "step": 2976 }, { "epoch": 0.7585679704420946, "grad_norm": 0.37320828437805176, "learning_rate": 2.4133537206931706e-06, "loss": 0.0001, "step": 2977 }, { "epoch": 0.758822779971971, "grad_norm": 0.368941068649292, "learning_rate": 2.410805300713558e-06, "loss": 0.0001, "step": 2978 }, { "epoch": 0.7590775895018473, "grad_norm": 0.36953192949295044, "learning_rate": 2.4082568807339453e-06, "loss": 0.0001, "step": 2979 }, { "epoch": 0.7593323990317238, "grad_norm": 1.654251217842102, "learning_rate": 2.4057084607543323e-06, "loss": 0.0005, "step": 2980 }, { "epoch": 0.7595872085616002, "grad_norm": 2.0320000648498535, "learning_rate": 2.40316004077472e-06, "loss": 0.0004, "step": 2981 }, { "epoch": 0.7598420180914767, "grad_norm": 2.3983585834503174, "learning_rate": 2.400611620795107e-06, "loss": 0.0005, "step": 2982 }, { "epoch": 0.760096827621353, "grad_norm": 0.19238144159317017, "learning_rate": 2.398063200815495e-06, "loss": 0.0001, "step": 2983 }, { "epoch": 0.7603516371512294, "grad_norm": 108.23665618896484, "learning_rate": 2.3955147808358818e-06, "loss": 4.1571, "step": 2984 }, { "epoch": 0.7606064466811059, "grad_norm": 2.204516887664795, "learning_rate": 2.392966360856269e-06, "loss": 0.0005, "step": 2985 }, { "epoch": 0.7608612562109823, "grad_norm": 4.243253231048584, "learning_rate": 2.390417940876657e-06, "loss": 0.0014, "step": 2986 }, { "epoch": 0.7611160657408587, "grad_norm": 2.7714288234710693, "learning_rate": 2.387869520897044e-06, "loss": 0.0007, "step": 2987 }, { "epoch": 0.7613708752707351, "grad_norm": 0.3666797876358032, "learning_rate": 2.3853211009174317e-06, "loss": 0.0001, "step": 2988 }, { "epoch": 0.7616256848006115, "grad_norm": 0.5708041191101074, "learning_rate": 2.3827726809378187e-06, "loss": 0.0003, "step": 2989 }, { "epoch": 0.761880494330488, "grad_norm": 1.5041930675506592, "learning_rate": 2.380224260958206e-06, "loss": 0.0009, "step": 2990 }, { "epoch": 0.7621353038603644, "grad_norm": 0.745023787021637, "learning_rate": 2.3776758409785934e-06, "loss": 0.0003, "step": 2991 }, { "epoch": 0.7623901133902408, "grad_norm": 1.3089368343353271, "learning_rate": 2.3751274209989808e-06, "loss": 0.0006, "step": 2992 }, { "epoch": 0.7626449229201172, "grad_norm": 0.5585309863090515, "learning_rate": 2.372579001019368e-06, "loss": 0.0002, "step": 2993 }, { "epoch": 0.7628997324499937, "grad_norm": 0.7453469038009644, "learning_rate": 2.3700305810397555e-06, "loss": 0.0003, "step": 2994 }, { "epoch": 0.7631545419798701, "grad_norm": 0.5882524847984314, "learning_rate": 2.367482161060143e-06, "loss": 0.0003, "step": 2995 }, { "epoch": 0.7634093515097464, "grad_norm": 0.1922614872455597, "learning_rate": 2.3649337410805303e-06, "loss": 0.0002, "step": 2996 }, { "epoch": 0.7636641610396229, "grad_norm": 0.9255902767181396, "learning_rate": 2.3623853211009177e-06, "loss": 0.0002, "step": 2997 }, { "epoch": 0.7639189705694993, "grad_norm": 1.1023064851760864, "learning_rate": 2.359836901121305e-06, "loss": 0.0002, "step": 2998 }, { "epoch": 0.7641737800993758, "grad_norm": 3.140293836593628, "learning_rate": 2.3572884811416924e-06, "loss": 0.0009, "step": 2999 }, { "epoch": 0.7644285896292521, "grad_norm": 3.515413522720337, "learning_rate": 2.35474006116208e-06, "loss": 0.0011, "step": 3000 }, { "epoch": 0.7646833991591285, "grad_norm": 2.802678346633911, "learning_rate": 2.352191641182467e-06, "loss": 0.0014, "step": 3001 }, { "epoch": 0.764938208689005, "grad_norm": 1.4748337268829346, "learning_rate": 2.3496432212028545e-06, "loss": 0.0003, "step": 3002 }, { "epoch": 0.7651930182188814, "grad_norm": 0.07164131104946136, "learning_rate": 2.3470948012232415e-06, "loss": 0.0002, "step": 3003 }, { "epoch": 0.7654478277487579, "grad_norm": 1.6511785984039307, "learning_rate": 2.3445463812436293e-06, "loss": 0.0003, "step": 3004 }, { "epoch": 0.7657026372786342, "grad_norm": 2.5743348598480225, "learning_rate": 2.3419979612640167e-06, "loss": 0.0006, "step": 3005 }, { "epoch": 0.7659574468085106, "grad_norm": 0.7361340522766113, "learning_rate": 2.339449541284404e-06, "loss": 0.0002, "step": 3006 }, { "epoch": 0.7662122563383871, "grad_norm": 3.124058961868286, "learning_rate": 2.3369011213047914e-06, "loss": 0.0009, "step": 3007 }, { "epoch": 0.7664670658682635, "grad_norm": 1.8547073602676392, "learning_rate": 2.3343527013251784e-06, "loss": 0.0006, "step": 3008 }, { "epoch": 0.7667218753981399, "grad_norm": 1.6609312295913696, "learning_rate": 2.331804281345566e-06, "loss": 0.0005, "step": 3009 }, { "epoch": 0.7669766849280163, "grad_norm": 0.9261508584022522, "learning_rate": 2.329255861365953e-06, "loss": 0.0002, "step": 3010 }, { "epoch": 0.7672314944578927, "grad_norm": 0.2232576459646225, "learning_rate": 2.3267074413863405e-06, "loss": 0.0002, "step": 3011 }, { "epoch": 0.7674863039877692, "grad_norm": 1.4785889387130737, "learning_rate": 2.324159021406728e-06, "loss": 0.0003, "step": 3012 }, { "epoch": 0.7677411135176455, "grad_norm": 2.0189552307128906, "learning_rate": 2.3216106014271153e-06, "loss": 0.0005, "step": 3013 }, { "epoch": 0.767995923047522, "grad_norm": 0.7414259910583496, "learning_rate": 2.3190621814475026e-06, "loss": 0.0002, "step": 3014 }, { "epoch": 0.7682507325773984, "grad_norm": 1.4758479595184326, "learning_rate": 2.31651376146789e-06, "loss": 0.0003, "step": 3015 }, { "epoch": 0.7685055421072748, "grad_norm": 0.5570059418678284, "learning_rate": 2.3139653414882774e-06, "loss": 0.0002, "step": 3016 }, { "epoch": 0.7687603516371513, "grad_norm": 0.7477070689201355, "learning_rate": 2.3114169215086648e-06, "loss": 0.0003, "step": 3017 }, { "epoch": 0.7690151611670276, "grad_norm": 0.737244188785553, "learning_rate": 2.308868501529052e-06, "loss": 0.0002, "step": 3018 }, { "epoch": 0.7692699706969041, "grad_norm": 1.2888957262039185, "learning_rate": 2.3063200815494395e-06, "loss": 0.0002, "step": 3019 }, { "epoch": 0.7695247802267805, "grad_norm": 1.48014497756958, "learning_rate": 2.303771661569827e-06, "loss": 0.0003, "step": 3020 }, { "epoch": 0.769779589756657, "grad_norm": 2.0348713397979736, "learning_rate": 2.3012232415902143e-06, "loss": 0.0005, "step": 3021 }, { "epoch": 0.7700343992865333, "grad_norm": 0.19871680438518524, "learning_rate": 2.2986748216106016e-06, "loss": 0.0001, "step": 3022 }, { "epoch": 0.7702892088164097, "grad_norm": 2.2015397548675537, "learning_rate": 2.296126401630989e-06, "loss": 0.0005, "step": 3023 }, { "epoch": 0.7705440183462862, "grad_norm": 1.1119638681411743, "learning_rate": 2.2935779816513764e-06, "loss": 0.0003, "step": 3024 }, { "epoch": 0.7707988278761626, "grad_norm": 2.9494826793670654, "learning_rate": 2.2910295616717638e-06, "loss": 0.001, "step": 3025 }, { "epoch": 0.7710536374060389, "grad_norm": 0.7421502470970154, "learning_rate": 2.288481141692151e-06, "loss": 0.0003, "step": 3026 }, { "epoch": 0.7713084469359154, "grad_norm": 0.37487971782684326, "learning_rate": 2.2859327217125385e-06, "loss": 0.0001, "step": 3027 }, { "epoch": 0.7715632564657918, "grad_norm": 0.7652550935745239, "learning_rate": 2.283384301732926e-06, "loss": 0.0004, "step": 3028 }, { "epoch": 0.7718180659956683, "grad_norm": 1.301180124282837, "learning_rate": 2.280835881753313e-06, "loss": 0.0003, "step": 3029 }, { "epoch": 0.7720728755255447, "grad_norm": 1.8387938737869263, "learning_rate": 2.2782874617737006e-06, "loss": 0.0004, "step": 3030 }, { "epoch": 0.772327685055421, "grad_norm": 3.1387088298797607, "learning_rate": 2.2757390417940876e-06, "loss": 0.0008, "step": 3031 }, { "epoch": 0.7725824945852975, "grad_norm": 1.110465407371521, "learning_rate": 2.2731906218144754e-06, "loss": 0.0002, "step": 3032 }, { "epoch": 0.7728373041151739, "grad_norm": 0.36917853355407715, "learning_rate": 2.2706422018348624e-06, "loss": 0.0001, "step": 3033 }, { "epoch": 0.7730921136450504, "grad_norm": 0.38017919659614563, "learning_rate": 2.2680937818552497e-06, "loss": 0.0002, "step": 3034 }, { "epoch": 0.7733469231749267, "grad_norm": 0.37009185552597046, "learning_rate": 2.2655453618756375e-06, "loss": 0.0001, "step": 3035 }, { "epoch": 0.7736017327048031, "grad_norm": 1.29048752784729, "learning_rate": 2.2629969418960245e-06, "loss": 0.0003, "step": 3036 }, { "epoch": 0.7738565422346796, "grad_norm": 0.17175383865833282, "learning_rate": 2.2604485219164123e-06, "loss": 0.0004, "step": 3037 }, { "epoch": 0.774111351764556, "grad_norm": 0.5501797199249268, "learning_rate": 2.2579001019367992e-06, "loss": 0.0001, "step": 3038 }, { "epoch": 0.7743661612944324, "grad_norm": 1.1173862218856812, "learning_rate": 2.2553516819571866e-06, "loss": 0.0003, "step": 3039 }, { "epoch": 0.7746209708243088, "grad_norm": 2.3959147930145264, "learning_rate": 2.252803261977574e-06, "loss": 0.0006, "step": 3040 }, { "epoch": 0.7748757803541853, "grad_norm": 0.9191679358482361, "learning_rate": 2.2502548419979614e-06, "loss": 0.0002, "step": 3041 }, { "epoch": 0.7751305898840617, "grad_norm": 0.5517916679382324, "learning_rate": 2.2477064220183487e-06, "loss": 0.0001, "step": 3042 }, { "epoch": 0.7753853994139381, "grad_norm": 0.36952465772628784, "learning_rate": 2.245158002038736e-06, "loss": 0.0001, "step": 3043 }, { "epoch": 0.7756402089438145, "grad_norm": 0.037521183490753174, "learning_rate": 2.2426095820591235e-06, "loss": 0.0001, "step": 3044 }, { "epoch": 0.7758950184736909, "grad_norm": 0.924656867980957, "learning_rate": 2.240061162079511e-06, "loss": 0.0003, "step": 3045 }, { "epoch": 0.7761498280035674, "grad_norm": 1.1634819507598877, "learning_rate": 2.2375127420998982e-06, "loss": 0.0007, "step": 3046 }, { "epoch": 0.7764046375334438, "grad_norm": 1.288515567779541, "learning_rate": 2.2349643221202856e-06, "loss": 0.0002, "step": 3047 }, { "epoch": 0.7766594470633201, "grad_norm": 1.8481618165969849, "learning_rate": 2.232415902140673e-06, "loss": 0.0005, "step": 3048 }, { "epoch": 0.7769142565931966, "grad_norm": 0.555514931678772, "learning_rate": 2.2298674821610604e-06, "loss": 0.0001, "step": 3049 }, { "epoch": 0.777169066123073, "grad_norm": 1.4796017408370972, "learning_rate": 2.2273190621814477e-06, "loss": 0.0005, "step": 3050 }, { "epoch": 0.7774238756529495, "grad_norm": 0.1852571666240692, "learning_rate": 2.224770642201835e-06, "loss": 0.0001, "step": 3051 }, { "epoch": 0.7776786851828258, "grad_norm": 1.8484970331192017, "learning_rate": 2.222222222222222e-06, "loss": 0.0004, "step": 3052 }, { "epoch": 0.7779334947127022, "grad_norm": 1.6637530326843262, "learning_rate": 2.21967380224261e-06, "loss": 0.0003, "step": 3053 }, { "epoch": 0.7781883042425787, "grad_norm": 1.2883572578430176, "learning_rate": 2.2171253822629973e-06, "loss": 0.0003, "step": 3054 }, { "epoch": 0.7784431137724551, "grad_norm": 0.18615072965621948, "learning_rate": 2.2145769622833846e-06, "loss": 0.0001, "step": 3055 }, { "epoch": 0.7786979233023315, "grad_norm": 2.397702932357788, "learning_rate": 2.212028542303772e-06, "loss": 0.0006, "step": 3056 }, { "epoch": 0.7789527328322079, "grad_norm": 1.4900476932525635, "learning_rate": 2.209480122324159e-06, "loss": 0.0004, "step": 3057 }, { "epoch": 0.7792075423620843, "grad_norm": 2.2055234909057617, "learning_rate": 2.2069317023445468e-06, "loss": 0.0006, "step": 3058 }, { "epoch": 0.7794623518919608, "grad_norm": 1.6860849857330322, "learning_rate": 2.2043832823649337e-06, "loss": 0.0007, "step": 3059 }, { "epoch": 0.7797171614218372, "grad_norm": 0.37468862533569336, "learning_rate": 2.2018348623853215e-06, "loss": 0.0001, "step": 3060 }, { "epoch": 0.7799719709517136, "grad_norm": 0.37304404377937317, "learning_rate": 2.1992864424057085e-06, "loss": 0.0001, "step": 3061 }, { "epoch": 0.78022678048159, "grad_norm": 2.2035555839538574, "learning_rate": 2.196738022426096e-06, "loss": 0.0004, "step": 3062 }, { "epoch": 0.7804815900114664, "grad_norm": 2.5715222358703613, "learning_rate": 2.1941896024464836e-06, "loss": 0.0006, "step": 3063 }, { "epoch": 0.7807363995413429, "grad_norm": 3.3188183307647705, "learning_rate": 2.1916411824668706e-06, "loss": 0.0009, "step": 3064 }, { "epoch": 0.7809912090712192, "grad_norm": 3.5022199153900146, "learning_rate": 2.1890927624872584e-06, "loss": 0.0012, "step": 3065 }, { "epoch": 0.7812460186010957, "grad_norm": 1.6629393100738525, "learning_rate": 2.1865443425076453e-06, "loss": 0.0003, "step": 3066 }, { "epoch": 0.7815008281309721, "grad_norm": 1.123888373374939, "learning_rate": 2.1839959225280327e-06, "loss": 0.0005, "step": 3067 }, { "epoch": 0.7817556376608485, "grad_norm": 0.743690550327301, "learning_rate": 2.18144750254842e-06, "loss": 0.0002, "step": 3068 }, { "epoch": 0.7820104471907249, "grad_norm": 1.2854650020599365, "learning_rate": 2.1788990825688075e-06, "loss": 0.0002, "step": 3069 }, { "epoch": 0.7822652567206013, "grad_norm": 2.4080731868743896, "learning_rate": 2.176350662589195e-06, "loss": 0.0006, "step": 3070 }, { "epoch": 0.7825200662504778, "grad_norm": 0.9194686412811279, "learning_rate": 2.1738022426095822e-06, "loss": 0.0002, "step": 3071 }, { "epoch": 0.7827748757803542, "grad_norm": 1.470223307609558, "learning_rate": 2.1712538226299696e-06, "loss": 0.0003, "step": 3072 }, { "epoch": 0.7830296853102306, "grad_norm": 1.2900716066360474, "learning_rate": 2.168705402650357e-06, "loss": 0.0002, "step": 3073 }, { "epoch": 0.783284494840107, "grad_norm": 1.84135901927948, "learning_rate": 2.1661569826707443e-06, "loss": 0.0003, "step": 3074 }, { "epoch": 0.7835393043699834, "grad_norm": 1.1012203693389893, "learning_rate": 2.1636085626911317e-06, "loss": 0.0002, "step": 3075 }, { "epoch": 0.7837941138998599, "grad_norm": 1.8298717737197876, "learning_rate": 2.161060142711519e-06, "loss": 0.0003, "step": 3076 }, { "epoch": 0.7840489234297363, "grad_norm": 1.485879898071289, "learning_rate": 2.1585117227319065e-06, "loss": 0.0006, "step": 3077 }, { "epoch": 0.7843037329596126, "grad_norm": 2.2039592266082764, "learning_rate": 2.155963302752294e-06, "loss": 0.0004, "step": 3078 }, { "epoch": 0.7845585424894891, "grad_norm": 3.137876272201538, "learning_rate": 2.1534148827726812e-06, "loss": 0.0008, "step": 3079 }, { "epoch": 0.7848133520193655, "grad_norm": 2.2064528465270996, "learning_rate": 2.150866462793068e-06, "loss": 0.0005, "step": 3080 }, { "epoch": 0.785068161549242, "grad_norm": 1.2864198684692383, "learning_rate": 2.148318042813456e-06, "loss": 0.0002, "step": 3081 }, { "epoch": 0.7853229710791183, "grad_norm": 1.1074917316436768, "learning_rate": 2.1457696228338434e-06, "loss": 0.0002, "step": 3082 }, { "epoch": 0.7855777806089947, "grad_norm": 0.6225212216377258, "learning_rate": 2.1432212028542307e-06, "loss": 0.0005, "step": 3083 }, { "epoch": 0.7858325901388712, "grad_norm": 2.022770404815674, "learning_rate": 2.140672782874618e-06, "loss": 0.0004, "step": 3084 }, { "epoch": 0.7860873996687476, "grad_norm": 2.5819709300994873, "learning_rate": 2.138124362895005e-06, "loss": 0.0006, "step": 3085 }, { "epoch": 0.7863422091986241, "grad_norm": 0.7374182343482971, "learning_rate": 2.135575942915393e-06, "loss": 0.0001, "step": 3086 }, { "epoch": 0.7865970187285004, "grad_norm": 0.15653327107429504, "learning_rate": 2.13302752293578e-06, "loss": 0.0002, "step": 3087 }, { "epoch": 0.7868518282583769, "grad_norm": 0.7334911227226257, "learning_rate": 2.130479102956167e-06, "loss": 0.0001, "step": 3088 }, { "epoch": 0.7871066377882533, "grad_norm": 0.43113094568252563, "learning_rate": 2.1279306829765546e-06, "loss": 0.0004, "step": 3089 }, { "epoch": 0.7873614473181297, "grad_norm": 1.2868423461914062, "learning_rate": 2.125382262996942e-06, "loss": 0.0002, "step": 3090 }, { "epoch": 0.7876162568480061, "grad_norm": 2158.853759765625, "learning_rate": 2.1228338430173297e-06, "loss": 2.6407, "step": 3091 }, { "epoch": 0.7878710663778825, "grad_norm": 1.1046279668807983, "learning_rate": 2.1202854230377167e-06, "loss": 0.0002, "step": 3092 }, { "epoch": 0.788125875907759, "grad_norm": 0.7357399463653564, "learning_rate": 2.117737003058104e-06, "loss": 0.0001, "step": 3093 }, { "epoch": 0.7883806854376354, "grad_norm": 0.7394788265228271, "learning_rate": 2.1151885830784914e-06, "loss": 0.0001, "step": 3094 }, { "epoch": 0.7886354949675117, "grad_norm": 0.2019885927438736, "learning_rate": 2.112640163098879e-06, "loss": 0.0001, "step": 3095 }, { "epoch": 0.7888903044973882, "grad_norm": 0.3683953583240509, "learning_rate": 2.110091743119266e-06, "loss": 0.0001, "step": 3096 }, { "epoch": 0.7891451140272646, "grad_norm": 2.0194709300994873, "learning_rate": 2.1075433231396536e-06, "loss": 0.0005, "step": 3097 }, { "epoch": 0.7893999235571411, "grad_norm": 1.6536400318145752, "learning_rate": 2.104994903160041e-06, "loss": 0.0003, "step": 3098 }, { "epoch": 0.7896547330870175, "grad_norm": 2.662055253982544, "learning_rate": 2.1024464831804283e-06, "loss": 0.0007, "step": 3099 }, { "epoch": 0.7899095426168938, "grad_norm": 2.758090019226074, "learning_rate": 2.0998980632008157e-06, "loss": 0.0006, "step": 3100 }, { "epoch": 0.7901643521467703, "grad_norm": 1.4700638055801392, "learning_rate": 2.097349643221203e-06, "loss": 0.0002, "step": 3101 }, { "epoch": 0.7904191616766467, "grad_norm": 0.919804573059082, "learning_rate": 2.0948012232415905e-06, "loss": 0.0001, "step": 3102 }, { "epoch": 0.7906739712065232, "grad_norm": 0.5511453747749329, "learning_rate": 2.092252803261978e-06, "loss": 0.0001, "step": 3103 }, { "epoch": 0.7909287807363995, "grad_norm": 0.3740653991699219, "learning_rate": 2.089704383282365e-06, "loss": 0.0001, "step": 3104 }, { "epoch": 0.7911835902662759, "grad_norm": 1.6558029651641846, "learning_rate": 2.0871559633027526e-06, "loss": 0.0003, "step": 3105 }, { "epoch": 0.7914383997961524, "grad_norm": 0.9363110661506653, "learning_rate": 2.0846075433231395e-06, "loss": 0.0003, "step": 3106 }, { "epoch": 0.7916932093260288, "grad_norm": 0.424748957157135, "learning_rate": 2.0820591233435273e-06, "loss": 0.0006, "step": 3107 }, { "epoch": 0.7919480188559052, "grad_norm": 0.5660373568534851, "learning_rate": 2.0795107033639143e-06, "loss": 0.0002, "step": 3108 }, { "epoch": 0.7922028283857816, "grad_norm": 57.526668548583984, "learning_rate": 2.076962283384302e-06, "loss": 0.6758, "step": 3109 }, { "epoch": 0.792457637915658, "grad_norm": 0.7420459985733032, "learning_rate": 2.0744138634046895e-06, "loss": 0.0002, "step": 3110 }, { "epoch": 0.7927124474455345, "grad_norm": 1.140906810760498, "learning_rate": 2.0718654434250764e-06, "loss": 0.0004, "step": 3111 }, { "epoch": 0.7929672569754109, "grad_norm": 1.4739915132522583, "learning_rate": 2.0693170234454642e-06, "loss": 0.0002, "step": 3112 }, { "epoch": 0.7932220665052873, "grad_norm": 1.8460501432418823, "learning_rate": 2.066768603465851e-06, "loss": 0.0004, "step": 3113 }, { "epoch": 0.7934768760351637, "grad_norm": 4.182898044586182, "learning_rate": 2.064220183486239e-06, "loss": 0.0019, "step": 3114 }, { "epoch": 0.7937316855650401, "grad_norm": 0.5987021327018738, "learning_rate": 2.061671763506626e-06, "loss": 0.0001, "step": 3115 }, { "epoch": 0.7939864950949166, "grad_norm": 1.1361603736877441, "learning_rate": 2.0591233435270133e-06, "loss": 0.0003, "step": 3116 }, { "epoch": 0.7942413046247929, "grad_norm": 3.3224737644195557, "learning_rate": 2.0565749235474007e-06, "loss": 0.0009, "step": 3117 }, { "epoch": 0.7944961141546694, "grad_norm": 2.2047245502471924, "learning_rate": 2.054026503567788e-06, "loss": 0.0005, "step": 3118 }, { "epoch": 0.7947509236845458, "grad_norm": 4.214397430419922, "learning_rate": 2.0514780835881754e-06, "loss": 0.0016, "step": 3119 }, { "epoch": 0.7950057332144222, "grad_norm": 1.837996482849121, "learning_rate": 2.048929663608563e-06, "loss": 0.0003, "step": 3120 }, { "epoch": 0.7952605427442986, "grad_norm": 2.957571506500244, "learning_rate": 2.04638124362895e-06, "loss": 0.0007, "step": 3121 }, { "epoch": 0.795515352274175, "grad_norm": 8.33478832244873, "learning_rate": 2.0438328236493375e-06, "loss": 0.0027, "step": 3122 }, { "epoch": 0.7957701618040515, "grad_norm": 0.7338762879371643, "learning_rate": 2.041284403669725e-06, "loss": 0.0001, "step": 3123 }, { "epoch": 0.7960249713339279, "grad_norm": 0.050715625286102295, "learning_rate": 2.0387359836901123e-06, "loss": 0.0001, "step": 3124 }, { "epoch": 0.7962797808638044, "grad_norm": 0.754950761795044, "learning_rate": 2.0361875637104997e-06, "loss": 0.0003, "step": 3125 }, { "epoch": 0.7965345903936807, "grad_norm": 1.2899516820907593, "learning_rate": 2.033639143730887e-06, "loss": 0.0003, "step": 3126 }, { "epoch": 0.7967893999235571, "grad_norm": 1.2918493747711182, "learning_rate": 2.0310907237512744e-06, "loss": 0.0003, "step": 3127 }, { "epoch": 0.7970442094534336, "grad_norm": 0.20089885592460632, "learning_rate": 2.028542303771662e-06, "loss": 0.0001, "step": 3128 }, { "epoch": 0.79729901898331, "grad_norm": 1.2876601219177246, "learning_rate": 2.025993883792049e-06, "loss": 0.0002, "step": 3129 }, { "epoch": 0.7975538285131863, "grad_norm": 1.291759967803955, "learning_rate": 2.0234454638124366e-06, "loss": 0.0004, "step": 3130 }, { "epoch": 0.7978086380430628, "grad_norm": 1.4760452508926392, "learning_rate": 2.020897043832824e-06, "loss": 0.0003, "step": 3131 }, { "epoch": 0.7980634475729392, "grad_norm": 1.1037983894348145, "learning_rate": 2.0183486238532113e-06, "loss": 0.0002, "step": 3132 }, { "epoch": 0.7983182571028157, "grad_norm": 1.1021745204925537, "learning_rate": 2.0158002038735987e-06, "loss": 0.0002, "step": 3133 }, { "epoch": 0.798573066632692, "grad_norm": 2.2037570476531982, "learning_rate": 2.0132517838939856e-06, "loss": 0.0005, "step": 3134 }, { "epoch": 0.7988278761625685, "grad_norm": 1.6555317640304565, "learning_rate": 2.0107033639143734e-06, "loss": 0.0003, "step": 3135 }, { "epoch": 0.7990826856924449, "grad_norm": 0.73773193359375, "learning_rate": 2.0081549439347604e-06, "loss": 0.0001, "step": 3136 }, { "epoch": 0.7993374952223213, "grad_norm": 0.4190981090068817, "learning_rate": 2.005606523955148e-06, "loss": 0.0004, "step": 3137 }, { "epoch": 0.7995923047521978, "grad_norm": 1.6767370700836182, "learning_rate": 2.003058103975535e-06, "loss": 0.0007, "step": 3138 }, { "epoch": 0.7998471142820741, "grad_norm": 1.47402024269104, "learning_rate": 2.0005096839959225e-06, "loss": 0.0005, "step": 3139 }, { "epoch": 0.8001019238119506, "grad_norm": 0.19999223947525024, "learning_rate": 1.9979612640163103e-06, "loss": 0.0001, "step": 3140 }, { "epoch": 0.800356733341827, "grad_norm": 2.22971773147583, "learning_rate": 1.9954128440366973e-06, "loss": 0.0008, "step": 3141 }, { "epoch": 0.8006115428717034, "grad_norm": 2.205775737762451, "learning_rate": 1.9928644240570846e-06, "loss": 0.0004, "step": 3142 }, { "epoch": 0.8008663524015798, "grad_norm": 1.6534258127212524, "learning_rate": 1.990316004077472e-06, "loss": 0.0003, "step": 3143 }, { "epoch": 0.8011211619314562, "grad_norm": 0.5521608591079712, "learning_rate": 1.9877675840978594e-06, "loss": 0.0001, "step": 3144 }, { "epoch": 0.8013759714613327, "grad_norm": 0.20115043222904205, "learning_rate": 1.9852191641182468e-06, "loss": 0.0001, "step": 3145 }, { "epoch": 0.8016307809912091, "grad_norm": 2.0202057361602783, "learning_rate": 1.982670744138634e-06, "loss": 0.0004, "step": 3146 }, { "epoch": 0.8018855905210854, "grad_norm": 0.5503075122833252, "learning_rate": 1.9801223241590215e-06, "loss": 0.0001, "step": 3147 }, { "epoch": 0.8021404000509619, "grad_norm": 0.739199161529541, "learning_rate": 1.977573904179409e-06, "loss": 0.0001, "step": 3148 }, { "epoch": 0.8023952095808383, "grad_norm": 1.6604827642440796, "learning_rate": 1.9750254841997963e-06, "loss": 0.0003, "step": 3149 }, { "epoch": 0.8026500191107148, "grad_norm": 0.1898457258939743, "learning_rate": 1.9724770642201837e-06, "loss": 0.0001, "step": 3150 }, { "epoch": 0.8029048286405912, "grad_norm": 0.7365040183067322, "learning_rate": 1.969928644240571e-06, "loss": 0.0001, "step": 3151 }, { "epoch": 0.8031596381704675, "grad_norm": 2.2229647636413574, "learning_rate": 1.9673802242609584e-06, "loss": 0.0005, "step": 3152 }, { "epoch": 0.803414447700344, "grad_norm": 0.3707945644855499, "learning_rate": 1.9648318042813458e-06, "loss": 0.0001, "step": 3153 }, { "epoch": 0.8036692572302204, "grad_norm": 0.786823034286499, "learning_rate": 1.962283384301733e-06, "loss": 0.0006, "step": 3154 }, { "epoch": 0.8039240667600969, "grad_norm": 1.5073155164718628, "learning_rate": 1.9597349643221205e-06, "loss": 0.0006, "step": 3155 }, { "epoch": 0.8041788762899732, "grad_norm": 1.8491262197494507, "learning_rate": 1.957186544342508e-06, "loss": 0.0004, "step": 3156 }, { "epoch": 0.8044336858198496, "grad_norm": 0.7426828145980835, "learning_rate": 1.954638124362895e-06, "loss": 0.0002, "step": 3157 }, { "epoch": 0.8046884953497261, "grad_norm": 0.5512661337852478, "learning_rate": 1.9520897043832827e-06, "loss": 0.0001, "step": 3158 }, { "epoch": 0.8049433048796025, "grad_norm": 2.2039318084716797, "learning_rate": 1.94954128440367e-06, "loss": 0.0004, "step": 3159 }, { "epoch": 0.8051981144094789, "grad_norm": 1.6569502353668213, "learning_rate": 1.946992864424057e-06, "loss": 0.0003, "step": 3160 }, { "epoch": 0.8054529239393553, "grad_norm": 0.15307430922985077, "learning_rate": 1.944444444444445e-06, "loss": 0.0002, "step": 3161 }, { "epoch": 0.8057077334692317, "grad_norm": 0.5563902854919434, "learning_rate": 1.9418960244648317e-06, "loss": 0.0001, "step": 3162 }, { "epoch": 0.8059625429991082, "grad_norm": 0.5530635118484497, "learning_rate": 1.9393476044852195e-06, "loss": 0.0001, "step": 3163 }, { "epoch": 0.8062173525289846, "grad_norm": 0.19095957279205322, "learning_rate": 1.9367991845056065e-06, "loss": 0.0001, "step": 3164 }, { "epoch": 0.806472162058861, "grad_norm": 2.02547287940979, "learning_rate": 1.934250764525994e-06, "loss": 0.0004, "step": 3165 }, { "epoch": 0.8067269715887374, "grad_norm": 0.37043601274490356, "learning_rate": 1.9317023445463812e-06, "loss": 0.0001, "step": 3166 }, { "epoch": 0.8069817811186139, "grad_norm": 0.3848607838153839, "learning_rate": 1.9291539245667686e-06, "loss": 0.0002, "step": 3167 }, { "epoch": 0.8072365906484903, "grad_norm": 0.9350706338882446, "learning_rate": 1.9266055045871564e-06, "loss": 0.0003, "step": 3168 }, { "epoch": 0.8074914001783666, "grad_norm": 0.18497370183467865, "learning_rate": 1.9240570846075434e-06, "loss": 0.0001, "step": 3169 }, { "epoch": 0.8077462097082431, "grad_norm": 2.0273544788360596, "learning_rate": 1.9215086646279308e-06, "loss": 0.0004, "step": 3170 }, { "epoch": 0.8080010192381195, "grad_norm": 1.655487298965454, "learning_rate": 1.918960244648318e-06, "loss": 0.0003, "step": 3171 }, { "epoch": 0.808255828767996, "grad_norm": 64.3198013305664, "learning_rate": 1.9164118246687055e-06, "loss": 0.6056, "step": 3172 }, { "epoch": 0.8085106382978723, "grad_norm": 0.6008839011192322, "learning_rate": 1.913863404689093e-06, "loss": 0.0005, "step": 3173 }, { "epoch": 0.8087654478277487, "grad_norm": 2.757523536682129, "learning_rate": 1.9113149847094803e-06, "loss": 0.0007, "step": 3174 }, { "epoch": 0.8090202573576252, "grad_norm": 0.739566445350647, "learning_rate": 1.9087665647298676e-06, "loss": 0.0002, "step": 3175 }, { "epoch": 0.8092750668875016, "grad_norm": 2.024294853210449, "learning_rate": 1.906218144750255e-06, "loss": 0.0004, "step": 3176 }, { "epoch": 0.8095298764173781, "grad_norm": 1.314105749130249, "learning_rate": 1.9036697247706424e-06, "loss": 0.0005, "step": 3177 }, { "epoch": 0.8097846859472544, "grad_norm": 0.012134388089179993, "learning_rate": 1.9011213047910298e-06, "loss": 0.0001, "step": 3178 }, { "epoch": 0.8100394954771308, "grad_norm": 1.2883307933807373, "learning_rate": 1.8985728848114171e-06, "loss": 0.0002, "step": 3179 }, { "epoch": 0.8102943050070073, "grad_norm": 1.2898865938186646, "learning_rate": 1.8960244648318043e-06, "loss": 0.0002, "step": 3180 }, { "epoch": 0.8105491145368837, "grad_norm": 1.465761661529541, "learning_rate": 1.8934760448521919e-06, "loss": 0.0002, "step": 3181 }, { "epoch": 0.81080392406676, "grad_norm": 0.19191209971904755, "learning_rate": 1.890927624872579e-06, "loss": 0.0001, "step": 3182 }, { "epoch": 0.8110587335966365, "grad_norm": 0.1985073983669281, "learning_rate": 1.8883792048929664e-06, "loss": 0.0001, "step": 3183 }, { "epoch": 0.8113135431265129, "grad_norm": 1.1032836437225342, "learning_rate": 1.885830784913354e-06, "loss": 0.0002, "step": 3184 }, { "epoch": 0.8115683526563894, "grad_norm": 1.2925753593444824, "learning_rate": 1.8832823649337412e-06, "loss": 0.0002, "step": 3185 }, { "epoch": 0.8118231621862657, "grad_norm": 1.1011825799942017, "learning_rate": 1.8807339449541288e-06, "loss": 0.0002, "step": 3186 }, { "epoch": 0.8120779717161422, "grad_norm": 2.93947434425354, "learning_rate": 1.878185524974516e-06, "loss": 0.0008, "step": 3187 }, { "epoch": 0.8123327812460186, "grad_norm": 3.3971755504608154, "learning_rate": 1.875637104994903e-06, "loss": 0.002, "step": 3188 }, { "epoch": 0.812587590775895, "grad_norm": 0.9299476742744446, "learning_rate": 1.8730886850152907e-06, "loss": 0.0002, "step": 3189 }, { "epoch": 0.8128424003057715, "grad_norm": 2.2103867530822754, "learning_rate": 1.870540265035678e-06, "loss": 0.0008, "step": 3190 }, { "epoch": 0.8130972098356478, "grad_norm": 2.4134817123413086, "learning_rate": 1.8679918450560654e-06, "loss": 0.0008, "step": 3191 }, { "epoch": 0.8133520193655243, "grad_norm": 0.7384933829307556, "learning_rate": 1.8654434250764528e-06, "loss": 0.0001, "step": 3192 }, { "epoch": 0.8136068288954007, "grad_norm": 0.551109254360199, "learning_rate": 1.86289500509684e-06, "loss": 0.0001, "step": 3193 }, { "epoch": 0.8138616384252771, "grad_norm": 1.8400739431381226, "learning_rate": 1.8603465851172276e-06, "loss": 0.0003, "step": 3194 }, { "epoch": 0.8141164479551535, "grad_norm": 1.1163463592529297, "learning_rate": 1.8577981651376147e-06, "loss": 0.0003, "step": 3195 }, { "epoch": 0.8143712574850299, "grad_norm": 0.03568291664123535, "learning_rate": 1.8552497451580023e-06, "loss": 0.0001, "step": 3196 }, { "epoch": 0.8146260670149064, "grad_norm": 2.759697914123535, "learning_rate": 1.8527013251783895e-06, "loss": 0.0007, "step": 3197 }, { "epoch": 0.8148808765447828, "grad_norm": 0.18526646494865417, "learning_rate": 1.8501529051987769e-06, "loss": 0.0001, "step": 3198 }, { "epoch": 0.8151356860746591, "grad_norm": 3.3116796016693115, "learning_rate": 1.8476044852191644e-06, "loss": 0.0009, "step": 3199 }, { "epoch": 0.8153904956045356, "grad_norm": 0.39239633083343506, "learning_rate": 1.8450560652395516e-06, "loss": 0.0002, "step": 3200 }, { "epoch": 0.815645305134412, "grad_norm": 0.1849568635225296, "learning_rate": 1.8425076452599388e-06, "loss": 0.0001, "step": 3201 }, { "epoch": 0.8159001146642885, "grad_norm": 0.5536384582519531, "learning_rate": 1.8399592252803264e-06, "loss": 0.0001, "step": 3202 }, { "epoch": 0.8161549241941649, "grad_norm": 2.0222973823547363, "learning_rate": 1.8374108053007137e-06, "loss": 0.0005, "step": 3203 }, { "epoch": 0.8164097337240412, "grad_norm": 1.1048551797866821, "learning_rate": 1.8348623853211011e-06, "loss": 0.0002, "step": 3204 }, { "epoch": 0.8166645432539177, "grad_norm": 0.7348305583000183, "learning_rate": 1.8323139653414885e-06, "loss": 0.0001, "step": 3205 }, { "epoch": 0.8169193527837941, "grad_norm": 1.4712138175964355, "learning_rate": 1.8297655453618757e-06, "loss": 0.0003, "step": 3206 }, { "epoch": 0.8171741623136706, "grad_norm": 0.9228633046150208, "learning_rate": 1.8272171253822632e-06, "loss": 0.0001, "step": 3207 }, { "epoch": 0.8174289718435469, "grad_norm": 0.7343900203704834, "learning_rate": 1.8246687054026504e-06, "loss": 0.0001, "step": 3208 }, { "epoch": 0.8176837813734233, "grad_norm": 0.3699757754802704, "learning_rate": 1.822120285423038e-06, "loss": 0.0001, "step": 3209 }, { "epoch": 0.8179385909032998, "grad_norm": 1.6553899049758911, "learning_rate": 1.8195718654434252e-06, "loss": 0.0003, "step": 3210 }, { "epoch": 0.8181934004331762, "grad_norm": 0.5553399324417114, "learning_rate": 1.8170234454638125e-06, "loss": 0.0001, "step": 3211 }, { "epoch": 0.8184482099630526, "grad_norm": 64.83710479736328, "learning_rate": 1.8144750254842e-06, "loss": 0.7422, "step": 3212 }, { "epoch": 0.818703019492929, "grad_norm": 2.2023842334747314, "learning_rate": 1.8119266055045873e-06, "loss": 0.0004, "step": 3213 }, { "epoch": 0.8189578290228055, "grad_norm": 0.267941415309906, "learning_rate": 1.8093781855249749e-06, "loss": 0.0003, "step": 3214 }, { "epoch": 0.8192126385526819, "grad_norm": 0.5682148933410645, "learning_rate": 1.806829765545362e-06, "loss": 0.0002, "step": 3215 }, { "epoch": 0.8194674480825583, "grad_norm": 0.3921974003314972, "learning_rate": 1.8042813455657492e-06, "loss": 0.0002, "step": 3216 }, { "epoch": 0.8197222576124347, "grad_norm": 1.2882508039474487, "learning_rate": 1.8017329255861368e-06, "loss": 0.0002, "step": 3217 }, { "epoch": 0.8199770671423111, "grad_norm": 0.5511212944984436, "learning_rate": 1.7991845056065242e-06, "loss": 0.0001, "step": 3218 }, { "epoch": 0.8202318766721876, "grad_norm": 0.5552588105201721, "learning_rate": 1.7966360856269113e-06, "loss": 0.0001, "step": 3219 }, { "epoch": 0.820486686202064, "grad_norm": 0.5487279891967773, "learning_rate": 1.794087665647299e-06, "loss": 0.0001, "step": 3220 }, { "epoch": 0.8207414957319403, "grad_norm": 0.36682966351509094, "learning_rate": 1.791539245667686e-06, "loss": 0.0001, "step": 3221 }, { "epoch": 0.8209963052618168, "grad_norm": 0.9217358231544495, "learning_rate": 1.7889908256880737e-06, "loss": 0.0002, "step": 3222 }, { "epoch": 0.8212511147916932, "grad_norm": 0.5514655113220215, "learning_rate": 1.7864424057084608e-06, "loss": 0.0001, "step": 3223 }, { "epoch": 0.8215059243215697, "grad_norm": 1.4825053215026855, "learning_rate": 1.7838939857288482e-06, "loss": 0.0003, "step": 3224 }, { "epoch": 0.821760733851446, "grad_norm": 57.09859848022461, "learning_rate": 1.7813455657492356e-06, "loss": 0.6485, "step": 3225 }, { "epoch": 0.8220155433813224, "grad_norm": 1.2931084632873535, "learning_rate": 1.778797145769623e-06, "loss": 0.0005, "step": 3226 }, { "epoch": 0.8222703529111989, "grad_norm": 1.6532317399978638, "learning_rate": 1.7762487257900106e-06, "loss": 0.0003, "step": 3227 }, { "epoch": 0.8225251624410753, "grad_norm": 1.6523853540420532, "learning_rate": 1.7737003058103977e-06, "loss": 0.0003, "step": 3228 }, { "epoch": 0.8227799719709518, "grad_norm": 0.19479314982891083, "learning_rate": 1.7711518858307849e-06, "loss": 0.0001, "step": 3229 }, { "epoch": 0.8230347815008281, "grad_norm": 0.7435193061828613, "learning_rate": 1.7686034658511725e-06, "loss": 0.0003, "step": 3230 }, { "epoch": 0.8232895910307045, "grad_norm": 0.9208630919456482, "learning_rate": 1.7660550458715596e-06, "loss": 0.0002, "step": 3231 }, { "epoch": 0.823544400560581, "grad_norm": 1.2899810075759888, "learning_rate": 1.7635066258919472e-06, "loss": 0.0002, "step": 3232 }, { "epoch": 0.8237992100904574, "grad_norm": 0.38151517510414124, "learning_rate": 1.7609582059123346e-06, "loss": 0.0002, "step": 3233 }, { "epoch": 0.8240540196203338, "grad_norm": 0.07929843664169312, "learning_rate": 1.7584097859327218e-06, "loss": 0.0001, "step": 3234 }, { "epoch": 0.8243088291502102, "grad_norm": 0.9264568090438843, "learning_rate": 1.7558613659531093e-06, "loss": 0.0002, "step": 3235 }, { "epoch": 0.8245636386800866, "grad_norm": 2.005134105682373, "learning_rate": 1.7533129459734965e-06, "loss": 0.0005, "step": 3236 }, { "epoch": 0.8248184482099631, "grad_norm": 2.757439136505127, "learning_rate": 1.7507645259938839e-06, "loss": 0.0008, "step": 3237 }, { "epoch": 0.8250732577398394, "grad_norm": 0.9121221899986267, "learning_rate": 1.7482161060142713e-06, "loss": 0.0001, "step": 3238 }, { "epoch": 0.8253280672697159, "grad_norm": 4.598016738891602, "learning_rate": 1.7456676860346586e-06, "loss": 0.0016, "step": 3239 }, { "epoch": 0.8255828767995923, "grad_norm": 1.4669568538665771, "learning_rate": 1.743119266055046e-06, "loss": 0.0002, "step": 3240 }, { "epoch": 0.8258376863294687, "grad_norm": 3.8571181297302246, "learning_rate": 1.7405708460754334e-06, "loss": 0.0013, "step": 3241 }, { "epoch": 0.8260924958593452, "grad_norm": 0.5512069463729858, "learning_rate": 1.7380224260958206e-06, "loss": 0.0001, "step": 3242 }, { "epoch": 0.8263473053892215, "grad_norm": 0.5765921473503113, "learning_rate": 1.7354740061162081e-06, "loss": 0.0003, "step": 3243 }, { "epoch": 0.826602114919098, "grad_norm": 1.2912657260894775, "learning_rate": 1.7329255861365953e-06, "loss": 0.0003, "step": 3244 }, { "epoch": 0.8268569244489744, "grad_norm": 2.576601505279541, "learning_rate": 1.730377166156983e-06, "loss": 0.0006, "step": 3245 }, { "epoch": 0.8271117339788508, "grad_norm": 0.5548754334449768, "learning_rate": 1.7278287461773703e-06, "loss": 0.0002, "step": 3246 }, { "epoch": 0.8273665435087272, "grad_norm": 2.023921251296997, "learning_rate": 1.7252803261977574e-06, "loss": 0.0004, "step": 3247 }, { "epoch": 0.8276213530386036, "grad_norm": 0.18888024985790253, "learning_rate": 1.722731906218145e-06, "loss": 0.0001, "step": 3248 }, { "epoch": 0.8278761625684801, "grad_norm": 2.2184112071990967, "learning_rate": 1.7201834862385322e-06, "loss": 0.0004, "step": 3249 }, { "epoch": 0.8281309720983565, "grad_norm": 1.4673421382904053, "learning_rate": 1.7176350662589198e-06, "loss": 0.0002, "step": 3250 }, { "epoch": 0.8283857816282328, "grad_norm": 0.553364098072052, "learning_rate": 1.715086646279307e-06, "loss": 0.0001, "step": 3251 }, { "epoch": 0.8286405911581093, "grad_norm": 0.7361369729042053, "learning_rate": 1.7125382262996943e-06, "loss": 0.0001, "step": 3252 }, { "epoch": 0.8288954006879857, "grad_norm": 2.031947374343872, "learning_rate": 1.7099898063200817e-06, "loss": 0.0005, "step": 3253 }, { "epoch": 0.8291502102178622, "grad_norm": 1.845070242881775, "learning_rate": 1.707441386340469e-06, "loss": 0.0005, "step": 3254 }, { "epoch": 0.8294050197477386, "grad_norm": 1.660022497177124, "learning_rate": 1.7048929663608562e-06, "loss": 0.0005, "step": 3255 }, { "epoch": 0.829659829277615, "grad_norm": 0.928957998752594, "learning_rate": 1.7023445463812438e-06, "loss": 0.0002, "step": 3256 }, { "epoch": 0.8299146388074914, "grad_norm": 1.8400131464004517, "learning_rate": 1.699796126401631e-06, "loss": 0.0004, "step": 3257 }, { "epoch": 0.8301694483373678, "grad_norm": 1.1030590534210205, "learning_rate": 1.6972477064220186e-06, "loss": 0.0002, "step": 3258 }, { "epoch": 0.8304242578672443, "grad_norm": 1.4760640859603882, "learning_rate": 1.6946992864424057e-06, "loss": 0.0003, "step": 3259 }, { "epoch": 0.8306790673971206, "grad_norm": 0.37755170464515686, "learning_rate": 1.6921508664627931e-06, "loss": 0.0002, "step": 3260 }, { "epoch": 0.830933876926997, "grad_norm": 0.9198454022407532, "learning_rate": 1.6896024464831807e-06, "loss": 0.0002, "step": 3261 }, { "epoch": 0.8311886864568735, "grad_norm": 0.18647943437099457, "learning_rate": 1.6870540265035679e-06, "loss": 0.0001, "step": 3262 }, { "epoch": 0.8314434959867499, "grad_norm": 0.25740620493888855, "learning_rate": 1.6845056065239555e-06, "loss": 0.0003, "step": 3263 }, { "epoch": 0.8316983055166263, "grad_norm": 1.2813047170639038, "learning_rate": 1.6819571865443426e-06, "loss": 0.0002, "step": 3264 }, { "epoch": 0.8319531150465027, "grad_norm": 0.18361376225948334, "learning_rate": 1.67940876656473e-06, "loss": 0.0001, "step": 3265 }, { "epoch": 0.8322079245763792, "grad_norm": 1.8545080423355103, "learning_rate": 1.6768603465851174e-06, "loss": 0.0004, "step": 3266 }, { "epoch": 0.8324627341062556, "grad_norm": 0.7587059736251831, "learning_rate": 1.6743119266055047e-06, "loss": 0.0005, "step": 3267 }, { "epoch": 0.832717543636132, "grad_norm": 1.4740618467330933, "learning_rate": 1.6717635066258921e-06, "loss": 0.0005, "step": 3268 }, { "epoch": 0.8329723531660084, "grad_norm": 0.03907667472958565, "learning_rate": 1.6692150866462795e-06, "loss": 0.0001, "step": 3269 }, { "epoch": 0.8332271626958848, "grad_norm": 1.101696491241455, "learning_rate": 1.6666666666666667e-06, "loss": 0.0002, "step": 3270 }, { "epoch": 0.8334819722257613, "grad_norm": 1.7670366764068604, "learning_rate": 1.6641182466870542e-06, "loss": 0.001, "step": 3271 }, { "epoch": 0.8337367817556377, "grad_norm": 0.3264486491680145, "learning_rate": 1.6615698267074414e-06, "loss": 0.0006, "step": 3272 }, { "epoch": 0.833991591285514, "grad_norm": 0.5560122132301331, "learning_rate": 1.659021406727829e-06, "loss": 0.0001, "step": 3273 }, { "epoch": 0.8342464008153905, "grad_norm": 0.9217104911804199, "learning_rate": 1.6564729867482162e-06, "loss": 0.0002, "step": 3274 }, { "epoch": 0.8345012103452669, "grad_norm": 0.36759257316589355, "learning_rate": 1.6539245667686035e-06, "loss": 0.0001, "step": 3275 }, { "epoch": 0.8347560198751434, "grad_norm": 0.5748932361602783, "learning_rate": 1.6513761467889911e-06, "loss": 0.0003, "step": 3276 }, { "epoch": 0.8350108294050197, "grad_norm": 1.2879806756973267, "learning_rate": 1.6488277268093783e-06, "loss": 0.0002, "step": 3277 }, { "epoch": 0.8352656389348961, "grad_norm": 1.8688651323318481, "learning_rate": 1.6462793068297655e-06, "loss": 0.0008, "step": 3278 }, { "epoch": 0.8355204484647726, "grad_norm": 0.10898048430681229, "learning_rate": 1.643730886850153e-06, "loss": 0.0001, "step": 3279 }, { "epoch": 0.835775257994649, "grad_norm": 0.743310272693634, "learning_rate": 1.6411824668705404e-06, "loss": 0.0002, "step": 3280 }, { "epoch": 0.8360300675245255, "grad_norm": 0.18447287380695343, "learning_rate": 1.6386340468909278e-06, "loss": 0.0001, "step": 3281 }, { "epoch": 0.8362848770544018, "grad_norm": 1.839355230331421, "learning_rate": 1.6360856269113152e-06, "loss": 0.0003, "step": 3282 }, { "epoch": 0.8365396865842782, "grad_norm": 0.9253600835800171, "learning_rate": 1.6335372069317023e-06, "loss": 0.0002, "step": 3283 }, { "epoch": 0.8367944961141547, "grad_norm": 0.2202572375535965, "learning_rate": 1.63098878695209e-06, "loss": 0.0002, "step": 3284 }, { "epoch": 0.8370493056440311, "grad_norm": 0.3682425618171692, "learning_rate": 1.628440366972477e-06, "loss": 0.0001, "step": 3285 }, { "epoch": 0.8373041151739075, "grad_norm": 0.22030451893806458, "learning_rate": 1.6258919469928647e-06, "loss": 0.0002, "step": 3286 }, { "epoch": 0.8375589247037839, "grad_norm": 1.8501321077346802, "learning_rate": 1.6233435270132518e-06, "loss": 0.0005, "step": 3287 }, { "epoch": 0.8378137342336603, "grad_norm": 1.1010669469833374, "learning_rate": 1.6207951070336392e-06, "loss": 0.0002, "step": 3288 }, { "epoch": 0.8380685437635368, "grad_norm": 0.5532807111740112, "learning_rate": 1.6182466870540268e-06, "loss": 0.0001, "step": 3289 }, { "epoch": 0.8383233532934131, "grad_norm": 2.194368839263916, "learning_rate": 1.615698267074414e-06, "loss": 0.0005, "step": 3290 }, { "epoch": 0.8385781628232896, "grad_norm": 2.582592248916626, "learning_rate": 1.6131498470948016e-06, "loss": 0.0007, "step": 3291 }, { "epoch": 0.838832972353166, "grad_norm": 0.2496633529663086, "learning_rate": 1.6106014271151887e-06, "loss": 0.0003, "step": 3292 }, { "epoch": 0.8390877818830424, "grad_norm": 0.009337597526609898, "learning_rate": 1.608053007135576e-06, "loss": 0.0001, "step": 3293 }, { "epoch": 0.8393425914129189, "grad_norm": 1.8306231498718262, "learning_rate": 1.6055045871559635e-06, "loss": 0.0003, "step": 3294 }, { "epoch": 0.8395974009427952, "grad_norm": 0.3713855743408203, "learning_rate": 1.6029561671763509e-06, "loss": 0.0001, "step": 3295 }, { "epoch": 0.8398522104726717, "grad_norm": 1.3309533596038818, "learning_rate": 1.600407747196738e-06, "loss": 0.0002, "step": 3296 }, { "epoch": 0.8401070200025481, "grad_norm": 0.1837642937898636, "learning_rate": 1.5978593272171256e-06, "loss": 0.0001, "step": 3297 }, { "epoch": 0.8403618295324246, "grad_norm": 1.3016551733016968, "learning_rate": 1.5953109072375128e-06, "loss": 0.0007, "step": 3298 }, { "epoch": 0.8406166390623009, "grad_norm": 0.7412087917327881, "learning_rate": 1.5927624872579004e-06, "loss": 0.0001, "step": 3299 }, { "epoch": 0.8408714485921773, "grad_norm": 1.1021308898925781, "learning_rate": 1.5902140672782875e-06, "loss": 0.0001, "step": 3300 }, { "epoch": 0.8411262581220538, "grad_norm": 0.20390921831130981, "learning_rate": 1.587665647298675e-06, "loss": 0.0001, "step": 3301 }, { "epoch": 0.8413810676519302, "grad_norm": 0.5533671975135803, "learning_rate": 1.5851172273190623e-06, "loss": 0.0001, "step": 3302 }, { "epoch": 0.8416358771818065, "grad_norm": 0.5548416376113892, "learning_rate": 1.5825688073394496e-06, "loss": 0.0001, "step": 3303 }, { "epoch": 0.841890686711683, "grad_norm": 1.101852297782898, "learning_rate": 1.5800203873598372e-06, "loss": 0.0002, "step": 3304 }, { "epoch": 0.8421454962415594, "grad_norm": 0.7332857847213745, "learning_rate": 1.5774719673802244e-06, "loss": 0.0001, "step": 3305 }, { "epoch": 0.8424003057714359, "grad_norm": 1.28891921043396, "learning_rate": 1.5749235474006116e-06, "loss": 0.0002, "step": 3306 }, { "epoch": 0.8426551153013123, "grad_norm": 1.6532511711120605, "learning_rate": 1.5723751274209992e-06, "loss": 0.0002, "step": 3307 }, { "epoch": 0.8429099248311887, "grad_norm": 0.18971222639083862, "learning_rate": 1.5698267074413865e-06, "loss": 0.0001, "step": 3308 }, { "epoch": 0.8431647343610651, "grad_norm": 1.6553388833999634, "learning_rate": 1.567278287461774e-06, "loss": 0.0002, "step": 3309 }, { "epoch": 0.8434195438909415, "grad_norm": 0.09118817746639252, "learning_rate": 1.5647298674821613e-06, "loss": 0.0001, "step": 3310 }, { "epoch": 0.843674353420818, "grad_norm": 1.1467056274414062, "learning_rate": 1.5621814475025484e-06, "loss": 0.0005, "step": 3311 }, { "epoch": 0.8439291629506943, "grad_norm": 1.2888351678848267, "learning_rate": 1.559633027522936e-06, "loss": 0.0002, "step": 3312 }, { "epoch": 0.8441839724805708, "grad_norm": 2.5776126384735107, "learning_rate": 1.5570846075433232e-06, "loss": 0.0007, "step": 3313 }, { "epoch": 0.8444387820104472, "grad_norm": 0.9240690469741821, "learning_rate": 1.5545361875637106e-06, "loss": 0.0001, "step": 3314 }, { "epoch": 0.8446935915403236, "grad_norm": 1.106943130493164, "learning_rate": 1.551987767584098e-06, "loss": 0.0002, "step": 3315 }, { "epoch": 0.8449484010702, "grad_norm": 1.4647514820098877, "learning_rate": 1.5494393476044853e-06, "loss": 0.0002, "step": 3316 }, { "epoch": 0.8452032106000764, "grad_norm": 2.3805205821990967, "learning_rate": 1.546890927624873e-06, "loss": 0.0005, "step": 3317 }, { "epoch": 0.8454580201299529, "grad_norm": 2.578885078430176, "learning_rate": 1.54434250764526e-06, "loss": 0.0005, "step": 3318 }, { "epoch": 0.8457128296598293, "grad_norm": 0.10772942006587982, "learning_rate": 1.5417940876656472e-06, "loss": 0.0001, "step": 3319 }, { "epoch": 0.8459676391897057, "grad_norm": 1.103499412536621, "learning_rate": 1.5392456676860348e-06, "loss": 0.0001, "step": 3320 }, { "epoch": 0.8462224487195821, "grad_norm": 1.4658657312393188, "learning_rate": 1.536697247706422e-06, "loss": 0.0002, "step": 3321 }, { "epoch": 0.8464772582494585, "grad_norm": 1.2925177812576294, "learning_rate": 1.5341488277268096e-06, "loss": 0.0004, "step": 3322 }, { "epoch": 0.846732067779335, "grad_norm": 0.9189571738243103, "learning_rate": 1.531600407747197e-06, "loss": 0.0001, "step": 3323 }, { "epoch": 0.8469868773092114, "grad_norm": 0.9240712523460388, "learning_rate": 1.5290519877675841e-06, "loss": 0.0002, "step": 3324 }, { "epoch": 0.8472416868390877, "grad_norm": 0.5506331324577332, "learning_rate": 1.5265035677879717e-06, "loss": 0.0001, "step": 3325 }, { "epoch": 0.8474964963689642, "grad_norm": 0.36803001165390015, "learning_rate": 1.5239551478083589e-06, "loss": 0.0001, "step": 3326 }, { "epoch": 0.8477513058988406, "grad_norm": 1.2910338640213013, "learning_rate": 1.5214067278287465e-06, "loss": 0.0003, "step": 3327 }, { "epoch": 0.8480061154287171, "grad_norm": 0.2201465517282486, "learning_rate": 1.5188583078491336e-06, "loss": 0.0002, "step": 3328 }, { "epoch": 0.8482609249585934, "grad_norm": 2.7562382221221924, "learning_rate": 1.516309887869521e-06, "loss": 0.0008, "step": 3329 }, { "epoch": 0.8485157344884698, "grad_norm": 2.5622830390930176, "learning_rate": 1.5137614678899084e-06, "loss": 0.0006, "step": 3330 }, { "epoch": 0.8487705440183463, "grad_norm": 0.7392947673797607, "learning_rate": 1.5112130479102958e-06, "loss": 0.0001, "step": 3331 }, { "epoch": 0.8490253535482227, "grad_norm": 1.2865701913833618, "learning_rate": 1.508664627930683e-06, "loss": 0.0002, "step": 3332 }, { "epoch": 0.8492801630780992, "grad_norm": 0.9199346899986267, "learning_rate": 1.5061162079510705e-06, "loss": 0.0001, "step": 3333 }, { "epoch": 0.8495349726079755, "grad_norm": 2.0948424339294434, "learning_rate": 1.5035677879714577e-06, "loss": 0.0005, "step": 3334 }, { "epoch": 0.849789782137852, "grad_norm": 0.5550519824028015, "learning_rate": 1.5010193679918453e-06, "loss": 0.0001, "step": 3335 }, { "epoch": 0.8500445916677284, "grad_norm": 1.1170141696929932, "learning_rate": 1.4984709480122326e-06, "loss": 0.0002, "step": 3336 }, { "epoch": 0.8502994011976048, "grad_norm": 0.5519346594810486, "learning_rate": 1.4959225280326198e-06, "loss": 0.0001, "step": 3337 }, { "epoch": 0.8505542107274812, "grad_norm": 2.2096643447875977, "learning_rate": 1.4933741080530074e-06, "loss": 0.0006, "step": 3338 }, { "epoch": 0.8508090202573576, "grad_norm": 1.4617160558700562, "learning_rate": 1.4908256880733945e-06, "loss": 0.0002, "step": 3339 }, { "epoch": 0.851063829787234, "grad_norm": 0.932553231716156, "learning_rate": 1.4882772680937821e-06, "loss": 0.0004, "step": 3340 }, { "epoch": 0.8513186393171105, "grad_norm": 0.3854021430015564, "learning_rate": 1.4857288481141693e-06, "loss": 0.0001, "step": 3341 }, { "epoch": 0.8515734488469868, "grad_norm": 2.0369412899017334, "learning_rate": 1.4831804281345567e-06, "loss": 0.0008, "step": 3342 }, { "epoch": 0.8518282583768633, "grad_norm": 1.8336182832717896, "learning_rate": 1.480632008154944e-06, "loss": 0.0003, "step": 3343 }, { "epoch": 0.8520830679067397, "grad_norm": 0.008914090692996979, "learning_rate": 1.4780835881753314e-06, "loss": 0.0, "step": 3344 }, { "epoch": 0.8523378774366162, "grad_norm": 0.5752094388008118, "learning_rate": 1.4755351681957188e-06, "loss": 0.0004, "step": 3345 }, { "epoch": 0.8525926869664926, "grad_norm": 1.6630806922912598, "learning_rate": 1.4729867482161062e-06, "loss": 0.0005, "step": 3346 }, { "epoch": 0.8528474964963689, "grad_norm": 2.7396888732910156, "learning_rate": 1.4704383282364933e-06, "loss": 0.0006, "step": 3347 }, { "epoch": 0.8531023060262454, "grad_norm": 1.1074622869491577, "learning_rate": 1.467889908256881e-06, "loss": 0.0002, "step": 3348 }, { "epoch": 0.8533571155561218, "grad_norm": 1.651811122894287, "learning_rate": 1.465341488277268e-06, "loss": 0.0003, "step": 3349 }, { "epoch": 0.8536119250859983, "grad_norm": 0.554191529750824, "learning_rate": 1.4627930682976555e-06, "loss": 0.0001, "step": 3350 }, { "epoch": 0.8538667346158746, "grad_norm": 0.5724309086799622, "learning_rate": 1.460244648318043e-06, "loss": 0.0002, "step": 3351 }, { "epoch": 0.854121544145751, "grad_norm": 1.1007170677185059, "learning_rate": 1.4576962283384302e-06, "loss": 0.0001, "step": 3352 }, { "epoch": 0.8543763536756275, "grad_norm": 2.3844776153564453, "learning_rate": 1.4551478083588178e-06, "loss": 0.0005, "step": 3353 }, { "epoch": 0.8546311632055039, "grad_norm": 0.9176234602928162, "learning_rate": 1.452599388379205e-06, "loss": 0.0001, "step": 3354 }, { "epoch": 0.8548859727353803, "grad_norm": 1.4674813747406006, "learning_rate": 1.4500509683995924e-06, "loss": 0.0002, "step": 3355 }, { "epoch": 0.8551407822652567, "grad_norm": 1.2819050550460815, "learning_rate": 1.4475025484199797e-06, "loss": 0.0002, "step": 3356 }, { "epoch": 0.8553955917951331, "grad_norm": 0.9179772734642029, "learning_rate": 1.4449541284403671e-06, "loss": 0.0002, "step": 3357 }, { "epoch": 0.8556504013250096, "grad_norm": 4.955930233001709, "learning_rate": 1.4424057084607545e-06, "loss": 0.0019, "step": 3358 }, { "epoch": 0.855905210854886, "grad_norm": 4.762359619140625, "learning_rate": 1.4398572884811419e-06, "loss": 0.0018, "step": 3359 }, { "epoch": 0.8561600203847624, "grad_norm": 3.1344704627990723, "learning_rate": 1.437308868501529e-06, "loss": 0.0009, "step": 3360 }, { "epoch": 0.8564148299146388, "grad_norm": 2.3914613723754883, "learning_rate": 1.4347604485219166e-06, "loss": 0.0009, "step": 3361 }, { "epoch": 0.8566696394445152, "grad_norm": 0.7597498297691345, "learning_rate": 1.4322120285423038e-06, "loss": 0.0002, "step": 3362 }, { "epoch": 0.8569244489743917, "grad_norm": 3.844191312789917, "learning_rate": 1.4296636085626914e-06, "loss": 0.0011, "step": 3363 }, { "epoch": 0.857179258504268, "grad_norm": 2.0254764556884766, "learning_rate": 1.4271151885830785e-06, "loss": 0.0004, "step": 3364 }, { "epoch": 0.8574340680341445, "grad_norm": 2.017852783203125, "learning_rate": 1.424566768603466e-06, "loss": 0.0004, "step": 3365 }, { "epoch": 0.8576888775640209, "grad_norm": 1.8731733560562134, "learning_rate": 1.4220183486238535e-06, "loss": 0.0005, "step": 3366 }, { "epoch": 0.8579436870938973, "grad_norm": 0.3666713833808899, "learning_rate": 1.4194699286442407e-06, "loss": 0.0001, "step": 3367 }, { "epoch": 0.8581984966237737, "grad_norm": 0.9811946153640747, "learning_rate": 1.4169215086646278e-06, "loss": 0.0004, "step": 3368 }, { "epoch": 0.8584533061536501, "grad_norm": 0.9157223105430603, "learning_rate": 1.4143730886850154e-06, "loss": 0.0001, "step": 3369 }, { "epoch": 0.8587081156835266, "grad_norm": 0.8225432634353638, "learning_rate": 1.4118246687054028e-06, "loss": 0.0004, "step": 3370 }, { "epoch": 0.858962925213403, "grad_norm": 2.5615131855010986, "learning_rate": 1.4092762487257902e-06, "loss": 0.0005, "step": 3371 }, { "epoch": 0.8592177347432794, "grad_norm": 0.7401790618896484, "learning_rate": 1.4067278287461775e-06, "loss": 0.0001, "step": 3372 }, { "epoch": 0.8594725442731558, "grad_norm": 2.3896090984344482, "learning_rate": 1.4041794087665647e-06, "loss": 0.0007, "step": 3373 }, { "epoch": 0.8597273538030322, "grad_norm": 1.2895820140838623, "learning_rate": 1.4016309887869523e-06, "loss": 0.0002, "step": 3374 }, { "epoch": 0.8599821633329087, "grad_norm": 0.7338542938232422, "learning_rate": 1.3990825688073395e-06, "loss": 0.0001, "step": 3375 }, { "epoch": 0.8602369728627851, "grad_norm": 0.008764375001192093, "learning_rate": 1.396534148827727e-06, "loss": 0.0, "step": 3376 }, { "epoch": 0.8604917823926614, "grad_norm": 0.7372742295265198, "learning_rate": 1.3939857288481142e-06, "loss": 0.0002, "step": 3377 }, { "epoch": 0.8607465919225379, "grad_norm": 1.8388150930404663, "learning_rate": 1.3914373088685016e-06, "loss": 0.0003, "step": 3378 }, { "epoch": 0.8610014014524143, "grad_norm": 2.955885648727417, "learning_rate": 1.3888888888888892e-06, "loss": 0.0012, "step": 3379 }, { "epoch": 0.8612562109822908, "grad_norm": 59.71873092651367, "learning_rate": 1.3863404689092763e-06, "loss": 0.6915, "step": 3380 }, { "epoch": 0.8615110205121671, "grad_norm": 1.3065950870513916, "learning_rate": 1.383792048929664e-06, "loss": 0.0005, "step": 3381 }, { "epoch": 0.8617658300420435, "grad_norm": 0.18345652520656586, "learning_rate": 1.381243628950051e-06, "loss": 0.0001, "step": 3382 }, { "epoch": 0.86202063957192, "grad_norm": 2.0169460773468018, "learning_rate": 1.3786952089704382e-06, "loss": 0.0004, "step": 3383 }, { "epoch": 0.8622754491017964, "grad_norm": 2.0195398330688477, "learning_rate": 1.3761467889908258e-06, "loss": 0.0004, "step": 3384 }, { "epoch": 0.8625302586316729, "grad_norm": 0.9190673828125, "learning_rate": 1.3735983690112132e-06, "loss": 0.0001, "step": 3385 }, { "epoch": 0.8627850681615492, "grad_norm": 0.7413526773452759, "learning_rate": 1.3710499490316006e-06, "loss": 0.0002, "step": 3386 }, { "epoch": 0.8630398776914256, "grad_norm": 1.4667162895202637, "learning_rate": 1.368501529051988e-06, "loss": 0.0002, "step": 3387 }, { "epoch": 0.8632946872213021, "grad_norm": 3.6718366146087646, "learning_rate": 1.3659531090723751e-06, "loss": 0.0011, "step": 3388 }, { "epoch": 0.8635494967511785, "grad_norm": 1.6534829139709473, "learning_rate": 1.3634046890927627e-06, "loss": 0.0003, "step": 3389 }, { "epoch": 0.8638043062810549, "grad_norm": 0.18489569425582886, "learning_rate": 1.3608562691131499e-06, "loss": 0.0001, "step": 3390 }, { "epoch": 0.8640591158109313, "grad_norm": 0.9200904965400696, "learning_rate": 1.3583078491335373e-06, "loss": 0.0001, "step": 3391 }, { "epoch": 0.8643139253408078, "grad_norm": 1.1040948629379272, "learning_rate": 1.3557594291539246e-06, "loss": 0.0002, "step": 3392 }, { "epoch": 0.8645687348706842, "grad_norm": 0.3662554621696472, "learning_rate": 1.353211009174312e-06, "loss": 0.0001, "step": 3393 }, { "epoch": 0.8648235444005605, "grad_norm": 0.150706484913826, "learning_rate": 1.3506625891946996e-06, "loss": 0.0003, "step": 3394 }, { "epoch": 0.865078353930437, "grad_norm": 0.061559055000543594, "learning_rate": 1.3481141692150868e-06, "loss": 0.0001, "step": 3395 }, { "epoch": 0.8653331634603134, "grad_norm": 0.47618454694747925, "learning_rate": 1.345565749235474e-06, "loss": 0.0008, "step": 3396 }, { "epoch": 0.8655879729901899, "grad_norm": 0.18645165860652924, "learning_rate": 1.3430173292558615e-06, "loss": 0.0001, "step": 3397 }, { "epoch": 0.8658427825200663, "grad_norm": 0.36757388710975647, "learning_rate": 1.3404689092762489e-06, "loss": 0.0001, "step": 3398 }, { "epoch": 0.8660975920499426, "grad_norm": 0.5595375895500183, "learning_rate": 1.3379204892966363e-06, "loss": 0.0001, "step": 3399 }, { "epoch": 0.8663524015798191, "grad_norm": 0.21293781697750092, "learning_rate": 1.3353720693170236e-06, "loss": 0.0002, "step": 3400 }, { "epoch": 0.8666072111096955, "grad_norm": 2.2046139240264893, "learning_rate": 1.3328236493374108e-06, "loss": 0.0004, "step": 3401 }, { "epoch": 0.866862020639572, "grad_norm": 2.4015700817108154, "learning_rate": 1.3302752293577984e-06, "loss": 0.0005, "step": 3402 }, { "epoch": 0.8671168301694483, "grad_norm": 1.6640211343765259, "learning_rate": 1.3277268093781856e-06, "loss": 0.0004, "step": 3403 }, { "epoch": 0.8673716396993247, "grad_norm": 1.0997145175933838, "learning_rate": 1.3251783893985731e-06, "loss": 0.0002, "step": 3404 }, { "epoch": 0.8676264492292012, "grad_norm": 1.8387995958328247, "learning_rate": 1.3226299694189603e-06, "loss": 0.0003, "step": 3405 }, { "epoch": 0.8678812587590776, "grad_norm": 2.01757550239563, "learning_rate": 1.3200815494393477e-06, "loss": 0.0005, "step": 3406 }, { "epoch": 0.868136068288954, "grad_norm": 2.2193567752838135, "learning_rate": 1.317533129459735e-06, "loss": 0.0005, "step": 3407 }, { "epoch": 0.8683908778188304, "grad_norm": 2.561945915222168, "learning_rate": 1.3149847094801224e-06, "loss": 0.0008, "step": 3408 }, { "epoch": 0.8686456873487068, "grad_norm": 3.110769033432007, "learning_rate": 1.3124362895005096e-06, "loss": 0.0008, "step": 3409 }, { "epoch": 0.8689004968785833, "grad_norm": 2.378373861312866, "learning_rate": 1.3098878695208972e-06, "loss": 0.0005, "step": 3410 }, { "epoch": 0.8691553064084597, "grad_norm": 1.502359390258789, "learning_rate": 1.3073394495412844e-06, "loss": 0.0005, "step": 3411 }, { "epoch": 0.8694101159383361, "grad_norm": 1.2891746759414673, "learning_rate": 1.304791029561672e-06, "loss": 0.0002, "step": 3412 }, { "epoch": 0.8696649254682125, "grad_norm": 0.007398359943181276, "learning_rate": 1.3022426095820593e-06, "loss": 0.0, "step": 3413 }, { "epoch": 0.869919734998089, "grad_norm": 0.377372682094574, "learning_rate": 1.2996941896024465e-06, "loss": 0.0001, "step": 3414 }, { "epoch": 0.8701745445279654, "grad_norm": 2.7547638416290283, "learning_rate": 1.297145769622834e-06, "loss": 0.0006, "step": 3415 }, { "epoch": 0.8704293540578417, "grad_norm": 3.6623358726501465, "learning_rate": 1.2945973496432212e-06, "loss": 0.0011, "step": 3416 }, { "epoch": 0.8706841635877182, "grad_norm": 2.003265857696533, "learning_rate": 1.2920489296636088e-06, "loss": 0.0004, "step": 3417 }, { "epoch": 0.8709389731175946, "grad_norm": 3.6613471508026123, "learning_rate": 1.289500509683996e-06, "loss": 0.001, "step": 3418 }, { "epoch": 0.871193782647471, "grad_norm": 4.0462422370910645, "learning_rate": 1.2869520897043834e-06, "loss": 0.0013, "step": 3419 }, { "epoch": 0.8714485921773474, "grad_norm": 2.3822827339172363, "learning_rate": 1.2844036697247707e-06, "loss": 0.0005, "step": 3420 }, { "epoch": 0.8717034017072238, "grad_norm": 2.5745174884796143, "learning_rate": 1.2818552497451581e-06, "loss": 0.0005, "step": 3421 }, { "epoch": 0.8719582112371003, "grad_norm": 2.2087485790252686, "learning_rate": 1.2793068297655457e-06, "loss": 0.0004, "step": 3422 }, { "epoch": 0.8722130207669767, "grad_norm": 2.5609025955200195, "learning_rate": 1.2767584097859329e-06, "loss": 0.0006, "step": 3423 }, { "epoch": 0.8724678302968532, "grad_norm": 0.36718735098838806, "learning_rate": 1.27420998980632e-06, "loss": 0.0001, "step": 3424 }, { "epoch": 0.8727226398267295, "grad_norm": 1.2889671325683594, "learning_rate": 1.2716615698267076e-06, "loss": 0.0002, "step": 3425 }, { "epoch": 0.8729774493566059, "grad_norm": 0.4406125247478485, "learning_rate": 1.2691131498470948e-06, "loss": 0.0007, "step": 3426 }, { "epoch": 0.8732322588864824, "grad_norm": 1.2893283367156982, "learning_rate": 1.2665647298674822e-06, "loss": 0.0002, "step": 3427 }, { "epoch": 0.8734870684163588, "grad_norm": 1.654122233390808, "learning_rate": 1.2640163098878697e-06, "loss": 0.0004, "step": 3428 }, { "epoch": 0.8737418779462351, "grad_norm": 0.9196105003356934, "learning_rate": 1.261467889908257e-06, "loss": 0.0001, "step": 3429 }, { "epoch": 0.8739966874761116, "grad_norm": 0.3743404150009155, "learning_rate": 1.2589194699286445e-06, "loss": 0.0001, "step": 3430 }, { "epoch": 0.874251497005988, "grad_norm": 0.367385596036911, "learning_rate": 1.2563710499490317e-06, "loss": 0.0001, "step": 3431 }, { "epoch": 0.8745063065358645, "grad_norm": 1.4750473499298096, "learning_rate": 1.253822629969419e-06, "loss": 0.0003, "step": 3432 }, { "epoch": 0.8747611160657408, "grad_norm": 2.013272762298584, "learning_rate": 1.2512742099898064e-06, "loss": 0.0003, "step": 3433 }, { "epoch": 0.8750159255956172, "grad_norm": 0.3919350504875183, "learning_rate": 1.2487257900101938e-06, "loss": 0.0002, "step": 3434 }, { "epoch": 0.8752707351254937, "grad_norm": 1.8309710025787354, "learning_rate": 1.2461773700305812e-06, "loss": 0.0004, "step": 3435 }, { "epoch": 0.8755255446553701, "grad_norm": 1.831384301185608, "learning_rate": 1.2436289500509685e-06, "loss": 0.0003, "step": 3436 }, { "epoch": 0.8757803541852466, "grad_norm": 0.5510997772216797, "learning_rate": 1.241080530071356e-06, "loss": 0.0001, "step": 3437 }, { "epoch": 0.8760351637151229, "grad_norm": 0.9151842594146729, "learning_rate": 1.2385321100917433e-06, "loss": 0.0001, "step": 3438 }, { "epoch": 0.8762899732449994, "grad_norm": 0.19439759850502014, "learning_rate": 1.2359836901121305e-06, "loss": 0.0001, "step": 3439 }, { "epoch": 0.8765447827748758, "grad_norm": 0.553834855556488, "learning_rate": 1.2334352701325178e-06, "loss": 0.0001, "step": 3440 }, { "epoch": 0.8767995923047522, "grad_norm": 63.51747131347656, "learning_rate": 1.2308868501529054e-06, "loss": 0.711, "step": 3441 }, { "epoch": 0.8770544018346286, "grad_norm": 0.36934083700180054, "learning_rate": 1.2283384301732928e-06, "loss": 0.0001, "step": 3442 }, { "epoch": 0.877309211364505, "grad_norm": 1.840773344039917, "learning_rate": 1.22579001019368e-06, "loss": 0.0003, "step": 3443 }, { "epoch": 0.8775640208943815, "grad_norm": 2.4072346687316895, "learning_rate": 1.2232415902140673e-06, "loss": 0.0009, "step": 3444 }, { "epoch": 0.8778188304242579, "grad_norm": 1.1013978719711304, "learning_rate": 1.2206931702344547e-06, "loss": 0.0001, "step": 3445 }, { "epoch": 0.8780736399541342, "grad_norm": 0.9450958967208862, "learning_rate": 1.218144750254842e-06, "loss": 0.0005, "step": 3446 }, { "epoch": 0.8783284494840107, "grad_norm": 1.1038814783096313, "learning_rate": 1.2155963302752295e-06, "loss": 0.0002, "step": 3447 }, { "epoch": 0.8785832590138871, "grad_norm": 1.2945983409881592, "learning_rate": 1.2130479102956168e-06, "loss": 0.0003, "step": 3448 }, { "epoch": 0.8788380685437636, "grad_norm": 2.2063963413238525, "learning_rate": 1.2104994903160042e-06, "loss": 0.0004, "step": 3449 }, { "epoch": 0.87909287807364, "grad_norm": 0.8547044396400452, "learning_rate": 1.2079510703363916e-06, "loss": 0.0007, "step": 3450 }, { "epoch": 0.8793476876035163, "grad_norm": 2.0164003372192383, "learning_rate": 1.205402650356779e-06, "loss": 0.0003, "step": 3451 }, { "epoch": 0.8796024971333928, "grad_norm": 2.3811895847320557, "learning_rate": 1.2028542303771661e-06, "loss": 0.0005, "step": 3452 }, { "epoch": 0.8798573066632692, "grad_norm": 1.4640616178512573, "learning_rate": 1.2003058103975535e-06, "loss": 0.0002, "step": 3453 }, { "epoch": 0.8801121161931457, "grad_norm": 1.2901147603988647, "learning_rate": 1.1977573904179409e-06, "loss": 0.0002, "step": 3454 }, { "epoch": 0.880366925723022, "grad_norm": 2.027597665786743, "learning_rate": 1.1952089704383285e-06, "loss": 0.0005, "step": 3455 }, { "epoch": 0.8806217352528984, "grad_norm": 1.1103754043579102, "learning_rate": 1.1926605504587159e-06, "loss": 0.0004, "step": 3456 }, { "epoch": 0.8808765447827749, "grad_norm": 0.369667649269104, "learning_rate": 1.190112130479103e-06, "loss": 0.0001, "step": 3457 }, { "epoch": 0.8811313543126513, "grad_norm": 0.1212221309542656, "learning_rate": 1.1875637104994904e-06, "loss": 0.0002, "step": 3458 }, { "epoch": 0.8813861638425277, "grad_norm": 2.7822916507720947, "learning_rate": 1.1850152905198778e-06, "loss": 0.0007, "step": 3459 }, { "epoch": 0.8816409733724041, "grad_norm": 0.19800235331058502, "learning_rate": 1.1824668705402651e-06, "loss": 0.0001, "step": 3460 }, { "epoch": 0.8818957829022805, "grad_norm": 2.7544901371002197, "learning_rate": 1.1799184505606525e-06, "loss": 0.0006, "step": 3461 }, { "epoch": 0.882150592432157, "grad_norm": 1.4682608842849731, "learning_rate": 1.17737003058104e-06, "loss": 0.0002, "step": 3462 }, { "epoch": 0.8824054019620334, "grad_norm": 0.00898990873247385, "learning_rate": 1.1748216106014273e-06, "loss": 0.0, "step": 3463 }, { "epoch": 0.8826602114919098, "grad_norm": 1.1018623113632202, "learning_rate": 1.1722731906218146e-06, "loss": 0.0002, "step": 3464 }, { "epoch": 0.8829150210217862, "grad_norm": 0.5961185097694397, "learning_rate": 1.169724770642202e-06, "loss": 0.0004, "step": 3465 }, { "epoch": 0.8831698305516626, "grad_norm": 0.5504558682441711, "learning_rate": 1.1671763506625892e-06, "loss": 0.0001, "step": 3466 }, { "epoch": 0.8834246400815391, "grad_norm": 0.12546858191490173, "learning_rate": 1.1646279306829766e-06, "loss": 0.0001, "step": 3467 }, { "epoch": 0.8836794496114154, "grad_norm": 1.8494154214859009, "learning_rate": 1.162079510703364e-06, "loss": 0.0003, "step": 3468 }, { "epoch": 0.8839342591412919, "grad_norm": 0.18920619785785675, "learning_rate": 1.1595310907237513e-06, "loss": 0.0001, "step": 3469 }, { "epoch": 0.8841890686711683, "grad_norm": 0.5740501284599304, "learning_rate": 1.1569826707441387e-06, "loss": 0.0003, "step": 3470 }, { "epoch": 0.8844438782010448, "grad_norm": 0.18954558670520782, "learning_rate": 1.154434250764526e-06, "loss": 0.0001, "step": 3471 }, { "epoch": 0.8846986877309211, "grad_norm": 0.918994665145874, "learning_rate": 1.1518858307849134e-06, "loss": 0.0001, "step": 3472 }, { "epoch": 0.8849534972607975, "grad_norm": 0.9200156331062317, "learning_rate": 1.1493374108053008e-06, "loss": 0.0001, "step": 3473 }, { "epoch": 0.885208306790674, "grad_norm": 0.7377496361732483, "learning_rate": 1.1467889908256882e-06, "loss": 0.0001, "step": 3474 }, { "epoch": 0.8854631163205504, "grad_norm": 1.1188920736312866, "learning_rate": 1.1442405708460756e-06, "loss": 0.0005, "step": 3475 }, { "epoch": 0.8857179258504269, "grad_norm": 1.3024511337280273, "learning_rate": 1.141692150866463e-06, "loss": 0.0003, "step": 3476 }, { "epoch": 0.8859727353803032, "grad_norm": 0.38165345788002014, "learning_rate": 1.1391437308868503e-06, "loss": 0.0002, "step": 3477 }, { "epoch": 0.8862275449101796, "grad_norm": 1.2965563535690308, "learning_rate": 1.1365953109072377e-06, "loss": 0.0004, "step": 3478 }, { "epoch": 0.8864823544400561, "grad_norm": 0.3798123300075531, "learning_rate": 1.1340468909276249e-06, "loss": 0.0001, "step": 3479 }, { "epoch": 0.8867371639699325, "grad_norm": 0.2869892716407776, "learning_rate": 1.1314984709480122e-06, "loss": 0.0005, "step": 3480 }, { "epoch": 0.8869919734998089, "grad_norm": 0.1279313713312149, "learning_rate": 1.1289500509683996e-06, "loss": 0.0002, "step": 3481 }, { "epoch": 0.8872467830296853, "grad_norm": 0.18515996634960175, "learning_rate": 1.126401630988787e-06, "loss": 0.0001, "step": 3482 }, { "epoch": 0.8875015925595617, "grad_norm": 1.1042380332946777, "learning_rate": 1.1238532110091744e-06, "loss": 0.0002, "step": 3483 }, { "epoch": 0.8877564020894382, "grad_norm": 1.840084195137024, "learning_rate": 1.1213047910295617e-06, "loss": 0.0003, "step": 3484 }, { "epoch": 0.8880112116193145, "grad_norm": 0.5487528443336487, "learning_rate": 1.1187563710499491e-06, "loss": 0.0001, "step": 3485 }, { "epoch": 0.888266021149191, "grad_norm": 2.932152509689331, "learning_rate": 1.1162079510703365e-06, "loss": 0.0007, "step": 3486 }, { "epoch": 0.8885208306790674, "grad_norm": 0.9225159287452698, "learning_rate": 1.1136595310907239e-06, "loss": 0.0002, "step": 3487 }, { "epoch": 0.8887756402089438, "grad_norm": 0.3672446310520172, "learning_rate": 1.111111111111111e-06, "loss": 0.0001, "step": 3488 }, { "epoch": 0.8890304497388203, "grad_norm": 1.6510754823684692, "learning_rate": 1.1085626911314986e-06, "loss": 0.0003, "step": 3489 }, { "epoch": 0.8892852592686966, "grad_norm": 0.1848938763141632, "learning_rate": 1.106014271151886e-06, "loss": 0.0001, "step": 3490 }, { "epoch": 0.8895400687985731, "grad_norm": 0.9340513944625854, "learning_rate": 1.1034658511722734e-06, "loss": 0.0005, "step": 3491 }, { "epoch": 0.8897948783284495, "grad_norm": 59.85841369628906, "learning_rate": 1.1009174311926608e-06, "loss": 0.5821, "step": 3492 }, { "epoch": 0.8900496878583259, "grad_norm": 0.3730149567127228, "learning_rate": 1.098369011213048e-06, "loss": 0.0001, "step": 3493 }, { "epoch": 0.8903044973882023, "grad_norm": 1.1145740747451782, "learning_rate": 1.0958205912334353e-06, "loss": 0.0005, "step": 3494 }, { "epoch": 0.8905593069180787, "grad_norm": 0.5576505661010742, "learning_rate": 1.0932721712538227e-06, "loss": 0.0001, "step": 3495 }, { "epoch": 0.8908141164479552, "grad_norm": 1.8412446975708008, "learning_rate": 1.09072375127421e-06, "loss": 0.0003, "step": 3496 }, { "epoch": 0.8910689259778316, "grad_norm": 0.25445348024368286, "learning_rate": 1.0881753312945974e-06, "loss": 0.0003, "step": 3497 }, { "epoch": 0.8913237355077079, "grad_norm": 1.1015740633010864, "learning_rate": 1.0856269113149848e-06, "loss": 0.0001, "step": 3498 }, { "epoch": 0.8915785450375844, "grad_norm": 0.1845940798521042, "learning_rate": 1.0830784913353722e-06, "loss": 0.0001, "step": 3499 }, { "epoch": 0.8918333545674608, "grad_norm": 2.2050302028656006, "learning_rate": 1.0805300713557595e-06, "loss": 0.0004, "step": 3500 }, { "epoch": 0.8920881640973373, "grad_norm": 0.7409982681274414, "learning_rate": 1.077981651376147e-06, "loss": 0.0002, "step": 3501 }, { "epoch": 0.8923429736272137, "grad_norm": 0.41030049324035645, "learning_rate": 1.075433231396534e-06, "loss": 0.0002, "step": 3502 }, { "epoch": 0.89259778315709, "grad_norm": 2.2039127349853516, "learning_rate": 1.0728848114169217e-06, "loss": 0.0004, "step": 3503 }, { "epoch": 0.8928525926869665, "grad_norm": 0.18471427261829376, "learning_rate": 1.070336391437309e-06, "loss": 0.0001, "step": 3504 }, { "epoch": 0.8931074022168429, "grad_norm": 1.6455799341201782, "learning_rate": 1.0677879714576964e-06, "loss": 0.0003, "step": 3505 }, { "epoch": 0.8933622117467194, "grad_norm": 0.7338398098945618, "learning_rate": 1.0652395514780836e-06, "loss": 0.0001, "step": 3506 }, { "epoch": 0.8936170212765957, "grad_norm": 0.3746280074119568, "learning_rate": 1.062691131498471e-06, "loss": 0.0001, "step": 3507 }, { "epoch": 0.8938718308064721, "grad_norm": 0.9231210947036743, "learning_rate": 1.0601427115188583e-06, "loss": 0.0001, "step": 3508 }, { "epoch": 0.8941266403363486, "grad_norm": 0.3669900894165039, "learning_rate": 1.0575942915392457e-06, "loss": 0.0001, "step": 3509 }, { "epoch": 0.894381449866225, "grad_norm": 1.1063251495361328, "learning_rate": 1.055045871559633e-06, "loss": 0.0002, "step": 3510 }, { "epoch": 0.8946362593961014, "grad_norm": 0.9209025502204895, "learning_rate": 1.0524974515800205e-06, "loss": 0.0001, "step": 3511 }, { "epoch": 0.8948910689259778, "grad_norm": 0.3668091893196106, "learning_rate": 1.0499490316004079e-06, "loss": 0.0001, "step": 3512 }, { "epoch": 0.8951458784558542, "grad_norm": 2.38006591796875, "learning_rate": 1.0474006116207952e-06, "loss": 0.0005, "step": 3513 }, { "epoch": 0.8954006879857307, "grad_norm": 1.472967505455017, "learning_rate": 1.0448521916411826e-06, "loss": 0.0002, "step": 3514 }, { "epoch": 0.8956554975156071, "grad_norm": 1.6586750745773315, "learning_rate": 1.0423037716615698e-06, "loss": 0.0003, "step": 3515 }, { "epoch": 0.8959103070454835, "grad_norm": 0.37253957986831665, "learning_rate": 1.0397553516819571e-06, "loss": 0.0001, "step": 3516 }, { "epoch": 0.8961651165753599, "grad_norm": 2.0197091102600098, "learning_rate": 1.0372069317023447e-06, "loss": 0.0005, "step": 3517 }, { "epoch": 0.8964199261052364, "grad_norm": 1.8275729417800903, "learning_rate": 1.0346585117227321e-06, "loss": 0.0004, "step": 3518 }, { "epoch": 0.8966747356351128, "grad_norm": 0.7480601668357849, "learning_rate": 1.0321100917431195e-06, "loss": 0.0003, "step": 3519 }, { "epoch": 0.8969295451649891, "grad_norm": 1.2876546382904053, "learning_rate": 1.0295616717635066e-06, "loss": 0.0002, "step": 3520 }, { "epoch": 0.8971843546948656, "grad_norm": 2.3853037357330322, "learning_rate": 1.027013251783894e-06, "loss": 0.0005, "step": 3521 }, { "epoch": 0.897439164224742, "grad_norm": 2.5691308975219727, "learning_rate": 1.0244648318042814e-06, "loss": 0.0006, "step": 3522 }, { "epoch": 0.8976939737546185, "grad_norm": 1.6624302864074707, "learning_rate": 1.0219164118246688e-06, "loss": 0.0003, "step": 3523 }, { "epoch": 0.8979487832844948, "grad_norm": 2.0202295780181885, "learning_rate": 1.0193679918450562e-06, "loss": 0.0003, "step": 3524 }, { "epoch": 0.8982035928143712, "grad_norm": 0.7332234382629395, "learning_rate": 1.0168195718654435e-06, "loss": 0.0001, "step": 3525 }, { "epoch": 0.8984584023442477, "grad_norm": 2.0199265480041504, "learning_rate": 1.014271151885831e-06, "loss": 0.0005, "step": 3526 }, { "epoch": 0.8987132118741241, "grad_norm": 0.5531265139579773, "learning_rate": 1.0117227319062183e-06, "loss": 0.0001, "step": 3527 }, { "epoch": 0.8989680214040006, "grad_norm": 0.7395689487457275, "learning_rate": 1.0091743119266057e-06, "loss": 0.0002, "step": 3528 }, { "epoch": 0.8992228309338769, "grad_norm": 0.528448760509491, "learning_rate": 1.0066258919469928e-06, "loss": 0.0005, "step": 3529 }, { "epoch": 0.8994776404637533, "grad_norm": 0.9517723321914673, "learning_rate": 1.0040774719673802e-06, "loss": 0.0003, "step": 3530 }, { "epoch": 0.8997324499936298, "grad_norm": 0.5510780811309814, "learning_rate": 1.0015290519877676e-06, "loss": 0.0001, "step": 3531 }, { "epoch": 0.8999872595235062, "grad_norm": 0.36876794695854187, "learning_rate": 9.989806320081552e-07, "loss": 0.0001, "step": 3532 }, { "epoch": 0.9002420690533826, "grad_norm": 0.5510498881340027, "learning_rate": 9.964322120285423e-07, "loss": 0.0001, "step": 3533 }, { "epoch": 0.900496878583259, "grad_norm": 0.23947674036026, "learning_rate": 9.938837920489297e-07, "loss": 0.0003, "step": 3534 }, { "epoch": 0.9007516881131354, "grad_norm": 1.467177391052246, "learning_rate": 9.91335372069317e-07, "loss": 0.0002, "step": 3535 }, { "epoch": 0.9010064976430119, "grad_norm": 1.1011815071105957, "learning_rate": 9.887869520897045e-07, "loss": 0.0001, "step": 3536 }, { "epoch": 0.9012613071728882, "grad_norm": 1.10175621509552, "learning_rate": 9.862385321100918e-07, "loss": 0.0002, "step": 3537 }, { "epoch": 0.9015161167027647, "grad_norm": 0.18445050716400146, "learning_rate": 9.836901121304792e-07, "loss": 0.0001, "step": 3538 }, { "epoch": 0.9017709262326411, "grad_norm": 1.2893038988113403, "learning_rate": 9.811416921508666e-07, "loss": 0.0002, "step": 3539 }, { "epoch": 0.9020257357625175, "grad_norm": 0.5511295795440674, "learning_rate": 9.78593272171254e-07, "loss": 0.0001, "step": 3540 }, { "epoch": 0.902280545292394, "grad_norm": 0.19097451865673065, "learning_rate": 9.760448521916413e-07, "loss": 0.0001, "step": 3541 }, { "epoch": 0.9025353548222703, "grad_norm": 1.1009514331817627, "learning_rate": 9.734964322120285e-07, "loss": 0.0001, "step": 3542 }, { "epoch": 0.9027901643521468, "grad_norm": 1.8350906372070312, "learning_rate": 9.709480122324159e-07, "loss": 0.0003, "step": 3543 }, { "epoch": 0.9030449738820232, "grad_norm": 0.5983842611312866, "learning_rate": 9.683995922528032e-07, "loss": 0.0003, "step": 3544 }, { "epoch": 0.9032997834118996, "grad_norm": 3.1220011711120605, "learning_rate": 9.658511722731906e-07, "loss": 0.0008, "step": 3545 }, { "epoch": 0.903554592941776, "grad_norm": 0.054486777633428574, "learning_rate": 9.633027522935782e-07, "loss": 0.0001, "step": 3546 }, { "epoch": 0.9038094024716524, "grad_norm": 0.7359999418258667, "learning_rate": 9.607543323139654e-07, "loss": 0.0001, "step": 3547 }, { "epoch": 0.9040642120015289, "grad_norm": 2.201531171798706, "learning_rate": 9.582059123343528e-07, "loss": 0.0004, "step": 3548 }, { "epoch": 0.9043190215314053, "grad_norm": 0.5616910457611084, "learning_rate": 9.556574923547401e-07, "loss": 0.0002, "step": 3549 }, { "epoch": 0.9045738310612816, "grad_norm": 0.5519056916236877, "learning_rate": 9.531090723751275e-07, "loss": 0.0001, "step": 3550 }, { "epoch": 0.9048286405911581, "grad_norm": 1.8309035301208496, "learning_rate": 9.505606523955149e-07, "loss": 0.0003, "step": 3551 }, { "epoch": 0.9050834501210345, "grad_norm": 2.7570908069610596, "learning_rate": 9.480122324159022e-07, "loss": 0.0006, "step": 3552 }, { "epoch": 0.905338259650911, "grad_norm": 0.7335793375968933, "learning_rate": 9.454638124362895e-07, "loss": 0.0001, "step": 3553 }, { "epoch": 0.9055930691807874, "grad_norm": 1.6531035900115967, "learning_rate": 9.42915392456677e-07, "loss": 0.0003, "step": 3554 }, { "epoch": 0.9058478787106637, "grad_norm": 0.9283350110054016, "learning_rate": 9.403669724770644e-07, "loss": 0.0002, "step": 3555 }, { "epoch": 0.9061026882405402, "grad_norm": 0.7547029852867126, "learning_rate": 9.378185524974515e-07, "loss": 0.0003, "step": 3556 }, { "epoch": 0.9063574977704166, "grad_norm": 0.9188807606697083, "learning_rate": 9.35270132517839e-07, "loss": 0.0002, "step": 3557 }, { "epoch": 0.9066123073002931, "grad_norm": 0.16861118376255035, "learning_rate": 9.327217125382264e-07, "loss": 0.0002, "step": 3558 }, { "epoch": 0.9068671168301694, "grad_norm": 61.749210357666016, "learning_rate": 9.301732925586138e-07, "loss": 0.586, "step": 3559 }, { "epoch": 0.9071219263600458, "grad_norm": 2.381141185760498, "learning_rate": 9.276248725790012e-07, "loss": 0.0005, "step": 3560 }, { "epoch": 0.9073767358899223, "grad_norm": 2.9304709434509277, "learning_rate": 9.250764525993884e-07, "loss": 0.0007, "step": 3561 }, { "epoch": 0.9076315454197987, "grad_norm": 2.3800735473632812, "learning_rate": 9.225280326197758e-07, "loss": 0.0005, "step": 3562 }, { "epoch": 0.9078863549496751, "grad_norm": 3.1233887672424316, "learning_rate": 9.199796126401632e-07, "loss": 0.0009, "step": 3563 }, { "epoch": 0.9081411644795515, "grad_norm": 2.565709114074707, "learning_rate": 9.174311926605506e-07, "loss": 0.0006, "step": 3564 }, { "epoch": 0.908395974009428, "grad_norm": 1.6661815643310547, "learning_rate": 9.148827726809378e-07, "loss": 0.0003, "step": 3565 }, { "epoch": 0.9086507835393044, "grad_norm": 1.6520096063613892, "learning_rate": 9.123343527013252e-07, "loss": 0.0003, "step": 3566 }, { "epoch": 0.9089055930691808, "grad_norm": 1.2874633073806763, "learning_rate": 9.097859327217126e-07, "loss": 0.0002, "step": 3567 }, { "epoch": 0.9091604025990572, "grad_norm": 0.3818194568157196, "learning_rate": 9.072375127421e-07, "loss": 0.0002, "step": 3568 }, { "epoch": 0.9094152121289336, "grad_norm": 1.28807532787323, "learning_rate": 9.046890927624874e-07, "loss": 0.0002, "step": 3569 }, { "epoch": 0.9096700216588101, "grad_norm": 1.2884348630905151, "learning_rate": 9.021406727828746e-07, "loss": 0.0002, "step": 3570 }, { "epoch": 0.9099248311886865, "grad_norm": 2.018805742263794, "learning_rate": 8.995922528032621e-07, "loss": 0.0004, "step": 3571 }, { "epoch": 0.9101796407185628, "grad_norm": 0.6590033769607544, "learning_rate": 8.970438328236495e-07, "loss": 0.0015, "step": 3572 }, { "epoch": 0.9104344502484393, "grad_norm": 1.6621391773223877, "learning_rate": 8.944954128440368e-07, "loss": 0.0003, "step": 3573 }, { "epoch": 0.9106892597783157, "grad_norm": 0.20120427012443542, "learning_rate": 8.919469928644241e-07, "loss": 0.0001, "step": 3574 }, { "epoch": 0.9109440693081922, "grad_norm": 0.372779905796051, "learning_rate": 8.893985728848115e-07, "loss": 0.0001, "step": 3575 }, { "epoch": 0.9111988788380685, "grad_norm": 0.37185341119766235, "learning_rate": 8.868501529051989e-07, "loss": 0.0001, "step": 3576 }, { "epoch": 0.9114536883679449, "grad_norm": 0.03727300092577934, "learning_rate": 8.843017329255862e-07, "loss": 0.0001, "step": 3577 }, { "epoch": 0.9117084978978214, "grad_norm": 2.2089006900787354, "learning_rate": 8.817533129459736e-07, "loss": 0.0004, "step": 3578 }, { "epoch": 0.9119633074276978, "grad_norm": 1.4783239364624023, "learning_rate": 8.792048929663609e-07, "loss": 0.0003, "step": 3579 }, { "epoch": 0.9122181169575743, "grad_norm": 2.756650924682617, "learning_rate": 8.766564729867483e-07, "loss": 0.0007, "step": 3580 }, { "epoch": 0.9124729264874506, "grad_norm": 1.4654980897903442, "learning_rate": 8.741080530071356e-07, "loss": 0.0002, "step": 3581 }, { "epoch": 0.912727736017327, "grad_norm": 0.2029205560684204, "learning_rate": 8.71559633027523e-07, "loss": 0.0001, "step": 3582 }, { "epoch": 0.9129825455472035, "grad_norm": 1.1016254425048828, "learning_rate": 8.690112130479103e-07, "loss": 0.0001, "step": 3583 }, { "epoch": 0.9132373550770799, "grad_norm": 0.9202356338500977, "learning_rate": 8.664627930682977e-07, "loss": 0.0001, "step": 3584 }, { "epoch": 0.9134921646069563, "grad_norm": 0.9170317053794861, "learning_rate": 8.639143730886851e-07, "loss": 0.0001, "step": 3585 }, { "epoch": 0.9137469741368327, "grad_norm": 0.1419130265712738, "learning_rate": 8.613659531090725e-07, "loss": 0.0002, "step": 3586 }, { "epoch": 0.9140017836667091, "grad_norm": 1.1392321586608887, "learning_rate": 8.588175331294599e-07, "loss": 0.0008, "step": 3587 }, { "epoch": 0.9142565931965856, "grad_norm": 1.1216070652008057, "learning_rate": 8.562691131498472e-07, "loss": 0.0004, "step": 3588 }, { "epoch": 0.9145114027264619, "grad_norm": 0.19959479570388794, "learning_rate": 8.537206931702345e-07, "loss": 0.0001, "step": 3589 }, { "epoch": 0.9147662122563384, "grad_norm": 0.38144397735595703, "learning_rate": 8.511722731906219e-07, "loss": 0.0001, "step": 3590 }, { "epoch": 0.9150210217862148, "grad_norm": 0.18390347063541412, "learning_rate": 8.486238532110093e-07, "loss": 0.0001, "step": 3591 }, { "epoch": 0.9152758313160912, "grad_norm": 0.3781943619251251, "learning_rate": 8.460754332313966e-07, "loss": 0.0001, "step": 3592 }, { "epoch": 0.9155306408459677, "grad_norm": 2.01956844329834, "learning_rate": 8.435270132517839e-07, "loss": 0.0005, "step": 3593 }, { "epoch": 0.915785450375844, "grad_norm": 0.5508185625076294, "learning_rate": 8.409785932721713e-07, "loss": 0.0001, "step": 3594 }, { "epoch": 0.9160402599057205, "grad_norm": 0.7331763505935669, "learning_rate": 8.384301732925587e-07, "loss": 0.0001, "step": 3595 }, { "epoch": 0.9162950694355969, "grad_norm": 0.37098750472068787, "learning_rate": 8.358817533129461e-07, "loss": 0.0001, "step": 3596 }, { "epoch": 0.9165498789654734, "grad_norm": 0.7666352391242981, "learning_rate": 8.333333333333333e-07, "loss": 0.0003, "step": 3597 }, { "epoch": 0.9168046884953497, "grad_norm": 1.6507714986801147, "learning_rate": 8.307849133537207e-07, "loss": 0.0003, "step": 3598 }, { "epoch": 0.9170594980252261, "grad_norm": 0.7332517504692078, "learning_rate": 8.282364933741081e-07, "loss": 0.0001, "step": 3599 }, { "epoch": 0.9173143075551026, "grad_norm": 0.37251508235931396, "learning_rate": 8.256880733944956e-07, "loss": 0.0001, "step": 3600 }, { "epoch": 0.917569117084979, "grad_norm": 3.11468505859375, "learning_rate": 8.231396534148827e-07, "loss": 0.0008, "step": 3601 }, { "epoch": 0.9178239266148553, "grad_norm": 2.2069485187530518, "learning_rate": 8.205912334352702e-07, "loss": 0.0004, "step": 3602 }, { "epoch": 0.9180787361447318, "grad_norm": 56.63701629638672, "learning_rate": 8.180428134556576e-07, "loss": 0.6069, "step": 3603 }, { "epoch": 0.9183335456746082, "grad_norm": 0.5525765419006348, "learning_rate": 8.15494393476045e-07, "loss": 0.0001, "step": 3604 }, { "epoch": 0.9185883552044847, "grad_norm": 0.19270236790180206, "learning_rate": 8.129459734964323e-07, "loss": 0.0001, "step": 3605 }, { "epoch": 0.9188431647343611, "grad_norm": 0.5537660121917725, "learning_rate": 8.103975535168196e-07, "loss": 0.0001, "step": 3606 }, { "epoch": 0.9190979742642374, "grad_norm": 1.1016510725021362, "learning_rate": 8.07849133537207e-07, "loss": 0.0002, "step": 3607 }, { "epoch": 0.9193527837941139, "grad_norm": 1.2803144454956055, "learning_rate": 8.053007135575944e-07, "loss": 0.0002, "step": 3608 }, { "epoch": 0.9196075933239903, "grad_norm": 2.3835971355438232, "learning_rate": 8.027522935779817e-07, "loss": 0.0005, "step": 3609 }, { "epoch": 0.9198624028538668, "grad_norm": 1.6487327814102173, "learning_rate": 8.00203873598369e-07, "loss": 0.0004, "step": 3610 }, { "epoch": 0.9201172123837431, "grad_norm": 0.7352177500724792, "learning_rate": 7.976554536187564e-07, "loss": 0.0001, "step": 3611 }, { "epoch": 0.9203720219136196, "grad_norm": 1.4914618730545044, "learning_rate": 7.951070336391438e-07, "loss": 0.0006, "step": 3612 }, { "epoch": 0.920626831443496, "grad_norm": 0.5603355169296265, "learning_rate": 7.925586136595311e-07, "loss": 0.0001, "step": 3613 }, { "epoch": 0.9208816409733724, "grad_norm": 0.5508754253387451, "learning_rate": 7.900101936799186e-07, "loss": 0.0001, "step": 3614 }, { "epoch": 0.9211364505032488, "grad_norm": 55.66609191894531, "learning_rate": 7.874617737003058e-07, "loss": 0.5821, "step": 3615 }, { "epoch": 0.9213912600331252, "grad_norm": 2.0176587104797363, "learning_rate": 7.849133537206933e-07, "loss": 0.0003, "step": 3616 }, { "epoch": 0.9216460695630017, "grad_norm": 0.10484116524457932, "learning_rate": 7.823649337410806e-07, "loss": 0.0001, "step": 3617 }, { "epoch": 0.9219008790928781, "grad_norm": 0.3686692714691162, "learning_rate": 7.79816513761468e-07, "loss": 0.0001, "step": 3618 }, { "epoch": 0.9221556886227545, "grad_norm": 1.2829900979995728, "learning_rate": 7.772680937818553e-07, "loss": 0.0002, "step": 3619 }, { "epoch": 0.9224104981526309, "grad_norm": 1.1046204566955566, "learning_rate": 7.747196738022427e-07, "loss": 0.0002, "step": 3620 }, { "epoch": 0.9226653076825073, "grad_norm": 59.104270935058594, "learning_rate": 7.7217125382263e-07, "loss": 0.711, "step": 3621 }, { "epoch": 0.9229201172123838, "grad_norm": 1.8374758958816528, "learning_rate": 7.696228338430174e-07, "loss": 0.0003, "step": 3622 }, { "epoch": 0.9231749267422602, "grad_norm": 1.4758732318878174, "learning_rate": 7.670744138634048e-07, "loss": 0.0003, "step": 3623 }, { "epoch": 0.9234297362721365, "grad_norm": 2.2047922611236572, "learning_rate": 7.645259938837921e-07, "loss": 0.0005, "step": 3624 }, { "epoch": 0.923684545802013, "grad_norm": 0.5563812851905823, "learning_rate": 7.619775739041794e-07, "loss": 0.0001, "step": 3625 }, { "epoch": 0.9239393553318894, "grad_norm": 0.3870163857936859, "learning_rate": 7.594291539245668e-07, "loss": 0.0002, "step": 3626 }, { "epoch": 0.9241941648617659, "grad_norm": 1.6636908054351807, "learning_rate": 7.568807339449542e-07, "loss": 0.0003, "step": 3627 }, { "epoch": 0.9244489743916422, "grad_norm": 0.11806201934814453, "learning_rate": 7.543323139653415e-07, "loss": 0.0002, "step": 3628 }, { "epoch": 0.9247037839215186, "grad_norm": 1.49205482006073, "learning_rate": 7.517838939857288e-07, "loss": 0.0006, "step": 3629 }, { "epoch": 0.9249585934513951, "grad_norm": 0.567326545715332, "learning_rate": 7.492354740061163e-07, "loss": 0.0002, "step": 3630 }, { "epoch": 0.9252134029812715, "grad_norm": 1.1035710573196411, "learning_rate": 7.466870540265037e-07, "loss": 0.0001, "step": 3631 }, { "epoch": 0.925468212511148, "grad_norm": 2.0169641971588135, "learning_rate": 7.441386340468911e-07, "loss": 0.0003, "step": 3632 }, { "epoch": 0.9257230220410243, "grad_norm": 0.7376835942268372, "learning_rate": 7.415902140672783e-07, "loss": 0.0001, "step": 3633 }, { "epoch": 0.9259778315709007, "grad_norm": 1.6542859077453613, "learning_rate": 7.390417940876657e-07, "loss": 0.0003, "step": 3634 }, { "epoch": 0.9262326411007772, "grad_norm": 2.5657899379730225, "learning_rate": 7.364933741080531e-07, "loss": 0.0006, "step": 3635 }, { "epoch": 0.9264874506306536, "grad_norm": 0.010379470884799957, "learning_rate": 7.339449541284405e-07, "loss": 0.0001, "step": 3636 }, { "epoch": 0.92674226016053, "grad_norm": 2.2039690017700195, "learning_rate": 7.313965341488277e-07, "loss": 0.0004, "step": 3637 }, { "epoch": 0.9269970696904064, "grad_norm": 0.552245020866394, "learning_rate": 7.288481141692151e-07, "loss": 0.0001, "step": 3638 }, { "epoch": 0.9272518792202828, "grad_norm": 0.15703481435775757, "learning_rate": 7.262996941896025e-07, "loss": 0.0003, "step": 3639 }, { "epoch": 0.9275066887501593, "grad_norm": 2.9995486736297607, "learning_rate": 7.237512742099899e-07, "loss": 0.0018, "step": 3640 }, { "epoch": 0.9277614982800356, "grad_norm": 2.3847758769989014, "learning_rate": 7.212028542303772e-07, "loss": 0.0005, "step": 3641 }, { "epoch": 0.9280163078099121, "grad_norm": 1.8379788398742676, "learning_rate": 7.186544342507645e-07, "loss": 0.0003, "step": 3642 }, { "epoch": 0.9282711173397885, "grad_norm": 0.9223796129226685, "learning_rate": 7.161060142711519e-07, "loss": 0.0002, "step": 3643 }, { "epoch": 0.928525926869665, "grad_norm": 0.559033989906311, "learning_rate": 7.135575942915393e-07, "loss": 0.0001, "step": 3644 }, { "epoch": 0.9287807363995414, "grad_norm": 1.2915287017822266, "learning_rate": 7.110091743119267e-07, "loss": 0.0002, "step": 3645 }, { "epoch": 0.9290355459294177, "grad_norm": 0.5537595748901367, "learning_rate": 7.084607543323139e-07, "loss": 0.0001, "step": 3646 }, { "epoch": 0.9292903554592942, "grad_norm": 60.09255599975586, "learning_rate": 7.059123343527014e-07, "loss": 0.586, "step": 3647 }, { "epoch": 0.9295451649891706, "grad_norm": 1.6716148853302002, "learning_rate": 7.033639143730888e-07, "loss": 0.0008, "step": 3648 }, { "epoch": 0.9297999745190471, "grad_norm": 1.104021430015564, "learning_rate": 7.008154943934761e-07, "loss": 0.0002, "step": 3649 }, { "epoch": 0.9300547840489234, "grad_norm": 1.6517330408096313, "learning_rate": 6.982670744138635e-07, "loss": 0.0003, "step": 3650 }, { "epoch": 0.9303095935787998, "grad_norm": 0.7332671880722046, "learning_rate": 6.957186544342508e-07, "loss": 0.0001, "step": 3651 }, { "epoch": 0.9305644031086763, "grad_norm": 1.299636721611023, "learning_rate": 6.931702344546382e-07, "loss": 0.0003, "step": 3652 }, { "epoch": 0.9308192126385527, "grad_norm": 1.121782898902893, "learning_rate": 6.906218144750255e-07, "loss": 0.0005, "step": 3653 }, { "epoch": 0.931074022168429, "grad_norm": 0.5990380644798279, "learning_rate": 6.880733944954129e-07, "loss": 0.0002, "step": 3654 }, { "epoch": 0.9313288316983055, "grad_norm": 1.103664755821228, "learning_rate": 6.855249745158003e-07, "loss": 0.0002, "step": 3655 }, { "epoch": 0.9315836412281819, "grad_norm": 0.1841956526041031, "learning_rate": 6.829765545361876e-07, "loss": 0.0001, "step": 3656 }, { "epoch": 0.9318384507580584, "grad_norm": 0.5553373098373413, "learning_rate": 6.804281345565749e-07, "loss": 0.0001, "step": 3657 }, { "epoch": 0.9320932602879348, "grad_norm": 0.02965780347585678, "learning_rate": 6.778797145769623e-07, "loss": 0.0001, "step": 3658 }, { "epoch": 0.9323480698178112, "grad_norm": 2.3831307888031006, "learning_rate": 6.753312945973498e-07, "loss": 0.0005, "step": 3659 }, { "epoch": 0.9326028793476876, "grad_norm": 0.09807921946048737, "learning_rate": 6.72782874617737e-07, "loss": 0.0002, "step": 3660 }, { "epoch": 0.932857688877564, "grad_norm": 2.0480518341064453, "learning_rate": 6.702344546381244e-07, "loss": 0.0008, "step": 3661 }, { "epoch": 0.9331124984074405, "grad_norm": 3.3785347938537598, "learning_rate": 6.676860346585118e-07, "loss": 0.0017, "step": 3662 }, { "epoch": 0.9333673079373168, "grad_norm": 0.36845359206199646, "learning_rate": 6.651376146788992e-07, "loss": 0.0001, "step": 3663 }, { "epoch": 0.9336221174671933, "grad_norm": 0.3703159689903259, "learning_rate": 6.625891946992866e-07, "loss": 0.0001, "step": 3664 }, { "epoch": 0.9338769269970697, "grad_norm": 53.1001091003418, "learning_rate": 6.600407747196738e-07, "loss": 0.5628, "step": 3665 }, { "epoch": 0.9341317365269461, "grad_norm": 0.22094665467739105, "learning_rate": 6.574923547400612e-07, "loss": 0.0002, "step": 3666 }, { "epoch": 0.9343865460568225, "grad_norm": 0.7406710982322693, "learning_rate": 6.549439347604486e-07, "loss": 0.0002, "step": 3667 }, { "epoch": 0.9346413555866989, "grad_norm": 1.1161631345748901, "learning_rate": 6.52395514780836e-07, "loss": 0.0003, "step": 3668 }, { "epoch": 0.9348961651165754, "grad_norm": 1.844944715499878, "learning_rate": 6.498470948012232e-07, "loss": 0.0003, "step": 3669 }, { "epoch": 0.9351509746464518, "grad_norm": 1.6511539220809937, "learning_rate": 6.472986748216106e-07, "loss": 0.0003, "step": 3670 }, { "epoch": 0.9354057841763282, "grad_norm": 0.3692387044429779, "learning_rate": 6.44750254841998e-07, "loss": 0.0001, "step": 3671 }, { "epoch": 0.9356605937062046, "grad_norm": 0.7415934801101685, "learning_rate": 6.422018348623854e-07, "loss": 0.0003, "step": 3672 }, { "epoch": 0.935915403236081, "grad_norm": 0.5524946451187134, "learning_rate": 6.396534148827729e-07, "loss": 0.0001, "step": 3673 }, { "epoch": 0.9361702127659575, "grad_norm": 0.1841777116060257, "learning_rate": 6.3710499490316e-07, "loss": 0.0001, "step": 3674 }, { "epoch": 0.9364250222958339, "grad_norm": 0.7374027967453003, "learning_rate": 6.345565749235474e-07, "loss": 0.0001, "step": 3675 }, { "epoch": 0.9366798318257102, "grad_norm": 0.7337672114372253, "learning_rate": 6.320081549439349e-07, "loss": 0.0001, "step": 3676 }, { "epoch": 0.9369346413555867, "grad_norm": 1.2896963357925415, "learning_rate": 6.294597349643222e-07, "loss": 0.0002, "step": 3677 }, { "epoch": 0.9371894508854631, "grad_norm": 0.18994960188865662, "learning_rate": 6.269113149847095e-07, "loss": 0.0001, "step": 3678 }, { "epoch": 0.9374442604153396, "grad_norm": 0.3817441463470459, "learning_rate": 6.243628950050969e-07, "loss": 0.0002, "step": 3679 }, { "epoch": 0.9376990699452159, "grad_norm": 54.50048828125, "learning_rate": 6.218144750254843e-07, "loss": 0.5235, "step": 3680 }, { "epoch": 0.9379538794750923, "grad_norm": 1.1020861864089966, "learning_rate": 6.192660550458716e-07, "loss": 0.0002, "step": 3681 }, { "epoch": 0.9382086890049688, "grad_norm": 1.4685746431350708, "learning_rate": 6.167176350662589e-07, "loss": 0.0003, "step": 3682 }, { "epoch": 0.9384634985348452, "grad_norm": 2.0248804092407227, "learning_rate": 6.141692150866464e-07, "loss": 0.0004, "step": 3683 }, { "epoch": 0.9387183080647217, "grad_norm": 0.20111745595932007, "learning_rate": 6.116207951070337e-07, "loss": 0.0001, "step": 3684 }, { "epoch": 0.938973117594598, "grad_norm": 3.113308906555176, "learning_rate": 6.09072375127421e-07, "loss": 0.0008, "step": 3685 }, { "epoch": 0.9392279271244744, "grad_norm": 1.65528404712677, "learning_rate": 6.065239551478084e-07, "loss": 0.0004, "step": 3686 }, { "epoch": 0.9394827366543509, "grad_norm": 0.22185440361499786, "learning_rate": 6.039755351681958e-07, "loss": 0.0002, "step": 3687 }, { "epoch": 0.9397375461842273, "grad_norm": 0.3923008441925049, "learning_rate": 6.014271151885831e-07, "loss": 0.0002, "step": 3688 }, { "epoch": 0.9399923557141037, "grad_norm": 0.5528542399406433, "learning_rate": 5.988786952089704e-07, "loss": 0.0001, "step": 3689 }, { "epoch": 0.9402471652439801, "grad_norm": 1.6547032594680786, "learning_rate": 5.963302752293579e-07, "loss": 0.0005, "step": 3690 }, { "epoch": 0.9405019747738566, "grad_norm": 2.0305702686309814, "learning_rate": 5.937818552497452e-07, "loss": 0.0006, "step": 3691 }, { "epoch": 0.940756784303733, "grad_norm": 1.4827885627746582, "learning_rate": 5.912334352701326e-07, "loss": 0.0004, "step": 3692 }, { "epoch": 0.9410115938336093, "grad_norm": 1.8385858535766602, "learning_rate": 5.8868501529052e-07, "loss": 0.0003, "step": 3693 }, { "epoch": 0.9412664033634858, "grad_norm": 2.0162012577056885, "learning_rate": 5.861365953109073e-07, "loss": 0.0004, "step": 3694 }, { "epoch": 0.9415212128933622, "grad_norm": 0.1848587840795517, "learning_rate": 5.835881753312946e-07, "loss": 0.0001, "step": 3695 }, { "epoch": 0.9417760224232387, "grad_norm": 1.2872227430343628, "learning_rate": 5.81039755351682e-07, "loss": 0.0002, "step": 3696 }, { "epoch": 0.9420308319531151, "grad_norm": 0.5505845546722412, "learning_rate": 5.784913353720693e-07, "loss": 0.0001, "step": 3697 }, { "epoch": 0.9422856414829914, "grad_norm": 0.5525186657905579, "learning_rate": 5.759429153924567e-07, "loss": 0.0001, "step": 3698 }, { "epoch": 0.9425404510128679, "grad_norm": 0.5650697946548462, "learning_rate": 5.733944954128441e-07, "loss": 0.0002, "step": 3699 }, { "epoch": 0.9427952605427443, "grad_norm": 0.7322360873222351, "learning_rate": 5.708460754332315e-07, "loss": 0.0001, "step": 3700 }, { "epoch": 0.9430500700726208, "grad_norm": 0.7346574664115906, "learning_rate": 5.682976554536189e-07, "loss": 0.0001, "step": 3701 }, { "epoch": 0.9433048796024971, "grad_norm": 1.6505340337753296, "learning_rate": 5.657492354740061e-07, "loss": 0.0003, "step": 3702 }, { "epoch": 0.9435596891323735, "grad_norm": 0.2167109102010727, "learning_rate": 5.632008154943935e-07, "loss": 0.0002, "step": 3703 }, { "epoch": 0.94381449866225, "grad_norm": 1.3108876943588257, "learning_rate": 5.606523955147809e-07, "loss": 0.0007, "step": 3704 }, { "epoch": 0.9440693081921264, "grad_norm": 0.9200971722602844, "learning_rate": 5.581039755351682e-07, "loss": 0.0001, "step": 3705 }, { "epoch": 0.9443241177220028, "grad_norm": 0.9422057867050171, "learning_rate": 5.555555555555555e-07, "loss": 0.0002, "step": 3706 }, { "epoch": 0.9445789272518792, "grad_norm": 1.102460503578186, "learning_rate": 5.53007135575943e-07, "loss": 0.0002, "step": 3707 }, { "epoch": 0.9448337367817556, "grad_norm": 0.5528345704078674, "learning_rate": 5.504587155963304e-07, "loss": 0.0001, "step": 3708 }, { "epoch": 0.9450885463116321, "grad_norm": 60.501380920410156, "learning_rate": 5.479102956167176e-07, "loss": 0.75, "step": 3709 }, { "epoch": 0.9453433558415085, "grad_norm": 1.2873963117599487, "learning_rate": 5.45361875637105e-07, "loss": 0.0002, "step": 3710 }, { "epoch": 0.9455981653713849, "grad_norm": 0.19462937116622925, "learning_rate": 5.428134556574924e-07, "loss": 0.0001, "step": 3711 }, { "epoch": 0.9458529749012613, "grad_norm": 2.0236527919769287, "learning_rate": 5.402650356778798e-07, "loss": 0.0007, "step": 3712 }, { "epoch": 0.9461077844311377, "grad_norm": 0.19565679132938385, "learning_rate": 5.37716615698267e-07, "loss": 0.0001, "step": 3713 }, { "epoch": 0.9463625939610142, "grad_norm": 0.5555315017700195, "learning_rate": 5.351681957186545e-07, "loss": 0.0002, "step": 3714 }, { "epoch": 0.9466174034908905, "grad_norm": 0.18437345325946808, "learning_rate": 5.326197757390418e-07, "loss": 0.0001, "step": 3715 }, { "epoch": 0.946872213020767, "grad_norm": 0.7423838973045349, "learning_rate": 5.300713557594292e-07, "loss": 0.0002, "step": 3716 }, { "epoch": 0.9471270225506434, "grad_norm": 0.7352287769317627, "learning_rate": 5.275229357798165e-07, "loss": 0.0001, "step": 3717 }, { "epoch": 0.9473818320805198, "grad_norm": 0.18915493786334991, "learning_rate": 5.249745158002039e-07, "loss": 0.0001, "step": 3718 }, { "epoch": 0.9476366416103962, "grad_norm": 0.37028759717941284, "learning_rate": 5.224260958205913e-07, "loss": 0.0001, "step": 3719 }, { "epoch": 0.9478914511402726, "grad_norm": 1.8464672565460205, "learning_rate": 5.198776758409786e-07, "loss": 0.0003, "step": 3720 }, { "epoch": 0.9481462606701491, "grad_norm": 0.19003406167030334, "learning_rate": 5.173292558613661e-07, "loss": 0.0001, "step": 3721 }, { "epoch": 0.9484010702000255, "grad_norm": 0.19309401512145996, "learning_rate": 5.147808358817533e-07, "loss": 0.0001, "step": 3722 }, { "epoch": 0.948655879729902, "grad_norm": 1.6562715768814087, "learning_rate": 5.122324159021407e-07, "loss": 0.0003, "step": 3723 }, { "epoch": 0.9489106892597783, "grad_norm": 0.7846118211746216, "learning_rate": 5.096839959225281e-07, "loss": 0.0007, "step": 3724 }, { "epoch": 0.9491654987896547, "grad_norm": 1.464313268661499, "learning_rate": 5.071355759429155e-07, "loss": 0.0002, "step": 3725 }, { "epoch": 0.9494203083195312, "grad_norm": 1.4689186811447144, "learning_rate": 5.045871559633028e-07, "loss": 0.0005, "step": 3726 }, { "epoch": 0.9496751178494076, "grad_norm": 1.2801268100738525, "learning_rate": 5.020387359836901e-07, "loss": 0.0002, "step": 3727 }, { "epoch": 0.949929927379284, "grad_norm": 0.9254714250564575, "learning_rate": 4.994903160040776e-07, "loss": 0.0002, "step": 3728 }, { "epoch": 0.9501847369091604, "grad_norm": 1.1325124502182007, "learning_rate": 4.969418960244648e-07, "loss": 0.0002, "step": 3729 }, { "epoch": 0.9504395464390368, "grad_norm": 0.7386569976806641, "learning_rate": 4.943934760448522e-07, "loss": 0.0001, "step": 3730 }, { "epoch": 0.9506943559689133, "grad_norm": 2.023291826248169, "learning_rate": 4.918450560652396e-07, "loss": 0.0004, "step": 3731 }, { "epoch": 0.9509491654987896, "grad_norm": 0.7436145544052124, "learning_rate": 4.89296636085627e-07, "loss": 0.0001, "step": 3732 }, { "epoch": 0.951203975028666, "grad_norm": 0.74481600522995, "learning_rate": 4.867482161060142e-07, "loss": 0.0001, "step": 3733 }, { "epoch": 0.9514587845585425, "grad_norm": 1.2989559173583984, "learning_rate": 4.841997961264016e-07, "loss": 0.0003, "step": 3734 }, { "epoch": 0.9517135940884189, "grad_norm": 1.1238503456115723, "learning_rate": 4.816513761467891e-07, "loss": 0.0003, "step": 3735 }, { "epoch": 0.9519684036182954, "grad_norm": 0.3678167462348938, "learning_rate": 4.791029561671764e-07, "loss": 0.0001, "step": 3736 }, { "epoch": 0.9522232131481717, "grad_norm": 1.477127194404602, "learning_rate": 4.7655453618756375e-07, "loss": 0.0003, "step": 3737 }, { "epoch": 0.9524780226780482, "grad_norm": 0.18923471868038177, "learning_rate": 4.740061162079511e-07, "loss": 0.0001, "step": 3738 }, { "epoch": 0.9527328322079246, "grad_norm": 1.1015632152557373, "learning_rate": 4.714576962283385e-07, "loss": 0.0002, "step": 3739 }, { "epoch": 0.952987641737801, "grad_norm": 0.03339569643139839, "learning_rate": 4.689092762487258e-07, "loss": 0.0001, "step": 3740 }, { "epoch": 0.9532424512676774, "grad_norm": 0.36655592918395996, "learning_rate": 4.663608562691132e-07, "loss": 0.0001, "step": 3741 }, { "epoch": 0.9534972607975538, "grad_norm": 0.15017320215702057, "learning_rate": 4.638124362895006e-07, "loss": 0.0002, "step": 3742 }, { "epoch": 0.9537520703274303, "grad_norm": 3.130861759185791, "learning_rate": 4.612640163098879e-07, "loss": 0.0011, "step": 3743 }, { "epoch": 0.9540068798573067, "grad_norm": 0.7334142327308655, "learning_rate": 4.587155963302753e-07, "loss": 0.0001, "step": 3744 }, { "epoch": 0.954261689387183, "grad_norm": 0.23245546221733093, "learning_rate": 4.561671763506626e-07, "loss": 0.0003, "step": 3745 }, { "epoch": 0.9545164989170595, "grad_norm": 1.1058969497680664, "learning_rate": 4.5361875637105e-07, "loss": 0.0002, "step": 3746 }, { "epoch": 0.9547713084469359, "grad_norm": 0.22090157866477966, "learning_rate": 4.510703363914373e-07, "loss": 0.0002, "step": 3747 }, { "epoch": 0.9550261179768124, "grad_norm": 0.7332428693771362, "learning_rate": 4.4852191641182473e-07, "loss": 0.0001, "step": 3748 }, { "epoch": 0.9552809275066888, "grad_norm": 0.5513398051261902, "learning_rate": 4.4597349643221205e-07, "loss": 0.0001, "step": 3749 }, { "epoch": 0.9555357370365651, "grad_norm": 0.7395011186599731, "learning_rate": 4.4342507645259943e-07, "loss": 0.0001, "step": 3750 }, { "epoch": 0.9557905465664416, "grad_norm": 0.551510751247406, "learning_rate": 4.408766564729868e-07, "loss": 0.0001, "step": 3751 }, { "epoch": 0.956045356096318, "grad_norm": 2.203521490097046, "learning_rate": 4.3832823649337413e-07, "loss": 0.0004, "step": 3752 }, { "epoch": 0.9563001656261945, "grad_norm": 0.5519979000091553, "learning_rate": 4.357798165137615e-07, "loss": 0.0001, "step": 3753 }, { "epoch": 0.9565549751560708, "grad_norm": 1.2864341735839844, "learning_rate": 4.3323139653414883e-07, "loss": 0.0002, "step": 3754 }, { "epoch": 0.9568097846859472, "grad_norm": 0.18399474024772644, "learning_rate": 4.3068297655453626e-07, "loss": 0.0001, "step": 3755 }, { "epoch": 0.9570645942158237, "grad_norm": 56.12962341308594, "learning_rate": 4.281345565749236e-07, "loss": 0.5078, "step": 3756 }, { "epoch": 0.9573194037457001, "grad_norm": 1.8431013822555542, "learning_rate": 4.2558613659531096e-07, "loss": 0.0004, "step": 3757 }, { "epoch": 0.9575742132755765, "grad_norm": 0.9164131879806519, "learning_rate": 4.230377166156983e-07, "loss": 0.0001, "step": 3758 }, { "epoch": 0.9578290228054529, "grad_norm": 0.010440192185342312, "learning_rate": 4.2048929663608565e-07, "loss": 0.0001, "step": 3759 }, { "epoch": 0.9580838323353293, "grad_norm": 0.550505518913269, "learning_rate": 4.1794087665647303e-07, "loss": 0.0001, "step": 3760 }, { "epoch": 0.9583386418652058, "grad_norm": 0.18508228659629822, "learning_rate": 4.1539245667686035e-07, "loss": 0.0001, "step": 3761 }, { "epoch": 0.9585934513950822, "grad_norm": 0.9193313717842102, "learning_rate": 4.128440366972478e-07, "loss": 0.0001, "step": 3762 }, { "epoch": 0.9588482609249586, "grad_norm": 0.2543248236179352, "learning_rate": 4.102956167176351e-07, "loss": 0.0003, "step": 3763 }, { "epoch": 0.959103070454835, "grad_norm": 0.37186697125434875, "learning_rate": 4.077471967380225e-07, "loss": 0.0001, "step": 3764 }, { "epoch": 0.9593578799847114, "grad_norm": 0.37072545289993286, "learning_rate": 4.051987767584098e-07, "loss": 0.0001, "step": 3765 }, { "epoch": 0.9596126895145879, "grad_norm": 0.7336541414260864, "learning_rate": 4.026503567787972e-07, "loss": 0.0001, "step": 3766 }, { "epoch": 0.9598674990444642, "grad_norm": 0.11445502191781998, "learning_rate": 4.001019367991845e-07, "loss": 0.0002, "step": 3767 }, { "epoch": 0.9601223085743407, "grad_norm": 1.1023070812225342, "learning_rate": 3.975535168195719e-07, "loss": 0.0002, "step": 3768 }, { "epoch": 0.9603771181042171, "grad_norm": 1.1116125583648682, "learning_rate": 3.950050968399593e-07, "loss": 0.0002, "step": 3769 }, { "epoch": 0.9606319276340936, "grad_norm": 1.651186227798462, "learning_rate": 3.9245667686034663e-07, "loss": 0.0004, "step": 3770 }, { "epoch": 0.9608867371639699, "grad_norm": 0.915917158126831, "learning_rate": 3.89908256880734e-07, "loss": 0.0001, "step": 3771 }, { "epoch": 0.9611415466938463, "grad_norm": 0.2000689059495926, "learning_rate": 3.8735983690112133e-07, "loss": 0.0001, "step": 3772 }, { "epoch": 0.9613963562237228, "grad_norm": 0.05438242852687836, "learning_rate": 3.848114169215087e-07, "loss": 0.0001, "step": 3773 }, { "epoch": 0.9616511657535992, "grad_norm": 0.37352022528648376, "learning_rate": 3.8226299694189603e-07, "loss": 0.0001, "step": 3774 }, { "epoch": 0.9619059752834757, "grad_norm": 0.20354032516479492, "learning_rate": 3.797145769622834e-07, "loss": 0.0001, "step": 3775 }, { "epoch": 0.962160784813352, "grad_norm": 1.4687559604644775, "learning_rate": 3.7716615698267073e-07, "loss": 0.0002, "step": 3776 }, { "epoch": 0.9624155943432284, "grad_norm": 0.06890498846769333, "learning_rate": 3.7461773700305816e-07, "loss": 0.0001, "step": 3777 }, { "epoch": 0.9626704038731049, "grad_norm": 58.888160705566406, "learning_rate": 3.7206931702344553e-07, "loss": 0.5352, "step": 3778 }, { "epoch": 0.9629252134029813, "grad_norm": 0.2002619355916977, "learning_rate": 3.6952089704383286e-07, "loss": 0.0001, "step": 3779 }, { "epoch": 0.9631800229328576, "grad_norm": 0.189313143491745, "learning_rate": 3.6697247706422023e-07, "loss": 0.0001, "step": 3780 }, { "epoch": 0.9634348324627341, "grad_norm": 1.1032707691192627, "learning_rate": 3.6442405708460756e-07, "loss": 0.0002, "step": 3781 }, { "epoch": 0.9636896419926105, "grad_norm": 0.9212062358856201, "learning_rate": 3.6187563710499493e-07, "loss": 0.0002, "step": 3782 }, { "epoch": 0.963944451522487, "grad_norm": 0.19264507293701172, "learning_rate": 3.5932721712538226e-07, "loss": 0.0001, "step": 3783 }, { "epoch": 0.9641992610523633, "grad_norm": 0.9237861633300781, "learning_rate": 3.5677879714576963e-07, "loss": 0.0002, "step": 3784 }, { "epoch": 0.9644540705822398, "grad_norm": 0.2253534495830536, "learning_rate": 3.5423037716615695e-07, "loss": 0.0002, "step": 3785 }, { "epoch": 0.9647088801121162, "grad_norm": 0.9342687726020813, "learning_rate": 3.516819571865444e-07, "loss": 0.0003, "step": 3786 }, { "epoch": 0.9649636896419926, "grad_norm": 1.1009222269058228, "learning_rate": 3.4913353720693176e-07, "loss": 0.0002, "step": 3787 }, { "epoch": 0.9652184991718691, "grad_norm": 0.5500348806381226, "learning_rate": 3.465851172273191e-07, "loss": 0.0001, "step": 3788 }, { "epoch": 0.9654733087017454, "grad_norm": 0.1910192370414734, "learning_rate": 3.4403669724770646e-07, "loss": 0.0001, "step": 3789 }, { "epoch": 0.9657281182316219, "grad_norm": 0.9151273965835571, "learning_rate": 3.414882772680938e-07, "loss": 0.0001, "step": 3790 }, { "epoch": 0.9659829277614983, "grad_norm": 0.5819462537765503, "learning_rate": 3.3893985728848116e-07, "loss": 0.0003, "step": 3791 }, { "epoch": 0.9662377372913747, "grad_norm": 0.18471595644950867, "learning_rate": 3.363914373088685e-07, "loss": 0.0001, "step": 3792 }, { "epoch": 0.9664925468212511, "grad_norm": 0.19817444682121277, "learning_rate": 3.338430173292559e-07, "loss": 0.0001, "step": 3793 }, { "epoch": 0.9667473563511275, "grad_norm": 0.3673633337020874, "learning_rate": 3.312945973496433e-07, "loss": 0.0001, "step": 3794 }, { "epoch": 0.967002165881004, "grad_norm": 0.18777716159820557, "learning_rate": 3.287461773700306e-07, "loss": 0.0001, "step": 3795 }, { "epoch": 0.9672569754108804, "grad_norm": 0.3684002757072449, "learning_rate": 3.26197757390418e-07, "loss": 0.0001, "step": 3796 }, { "epoch": 0.9675117849407567, "grad_norm": 0.39651867747306824, "learning_rate": 3.236493374108053e-07, "loss": 0.0002, "step": 3797 }, { "epoch": 0.9677665944706332, "grad_norm": 0.9205226898193359, "learning_rate": 3.211009174311927e-07, "loss": 0.0001, "step": 3798 }, { "epoch": 0.9680214040005096, "grad_norm": 0.5718783140182495, "learning_rate": 3.1855249745158e-07, "loss": 0.0003, "step": 3799 }, { "epoch": 0.9682762135303861, "grad_norm": 0.09947971254587173, "learning_rate": 3.1600407747196744e-07, "loss": 0.0002, "step": 3800 }, { "epoch": 0.9685310230602625, "grad_norm": 0.4778984785079956, "learning_rate": 3.1345565749235476e-07, "loss": 0.0006, "step": 3801 }, { "epoch": 0.9687858325901388, "grad_norm": 0.5512447953224182, "learning_rate": 3.1090723751274214e-07, "loss": 0.0001, "step": 3802 }, { "epoch": 0.9690406421200153, "grad_norm": 1.2865618467330933, "learning_rate": 3.0835881753312946e-07, "loss": 0.0002, "step": 3803 }, { "epoch": 0.9692954516498917, "grad_norm": 0.3697473108768463, "learning_rate": 3.0581039755351683e-07, "loss": 0.0001, "step": 3804 }, { "epoch": 0.9695502611797682, "grad_norm": 0.9247238039970398, "learning_rate": 3.032619775739042e-07, "loss": 0.0002, "step": 3805 }, { "epoch": 0.9698050707096445, "grad_norm": 0.42829933762550354, "learning_rate": 3.0071355759429153e-07, "loss": 0.0004, "step": 3806 }, { "epoch": 0.9700598802395209, "grad_norm": 0.5637423396110535, "learning_rate": 2.9816513761467896e-07, "loss": 0.0002, "step": 3807 }, { "epoch": 0.9703146897693974, "grad_norm": 0.9276854991912842, "learning_rate": 2.956167176350663e-07, "loss": 0.0002, "step": 3808 }, { "epoch": 0.9705694992992738, "grad_norm": 1.4756884574890137, "learning_rate": 2.9306829765545366e-07, "loss": 0.0003, "step": 3809 }, { "epoch": 0.9708243088291502, "grad_norm": 1.290398359298706, "learning_rate": 2.90519877675841e-07, "loss": 0.0002, "step": 3810 }, { "epoch": 0.9710791183590266, "grad_norm": 0.18827484548091888, "learning_rate": 2.8797145769622836e-07, "loss": 0.0001, "step": 3811 }, { "epoch": 0.971333927888903, "grad_norm": 0.2011178433895111, "learning_rate": 2.8542303771661574e-07, "loss": 0.0001, "step": 3812 }, { "epoch": 0.9715887374187795, "grad_norm": 1.6506861448287964, "learning_rate": 2.8287461773700306e-07, "loss": 0.0003, "step": 3813 }, { "epoch": 0.9718435469486559, "grad_norm": 1.1028997898101807, "learning_rate": 2.8032619775739044e-07, "loss": 0.0002, "step": 3814 }, { "epoch": 0.9720983564785323, "grad_norm": 1.2872145175933838, "learning_rate": 2.7777777777777776e-07, "loss": 0.0002, "step": 3815 }, { "epoch": 0.9723531660084087, "grad_norm": 0.5577318668365479, "learning_rate": 2.752293577981652e-07, "loss": 0.0001, "step": 3816 }, { "epoch": 0.9726079755382852, "grad_norm": 0.918602705001831, "learning_rate": 2.726809378185525e-07, "loss": 0.0001, "step": 3817 }, { "epoch": 0.9728627850681616, "grad_norm": 0.3811383843421936, "learning_rate": 2.701325178389399e-07, "loss": 0.0002, "step": 3818 }, { "epoch": 0.9731175945980379, "grad_norm": 1.2923508882522583, "learning_rate": 2.6758409785932726e-07, "loss": 0.0003, "step": 3819 }, { "epoch": 0.9733724041279144, "grad_norm": 0.18442682921886444, "learning_rate": 2.650356778797146e-07, "loss": 0.0001, "step": 3820 }, { "epoch": 0.9736272136577908, "grad_norm": 0.745742678642273, "learning_rate": 2.6248725790010196e-07, "loss": 0.0002, "step": 3821 }, { "epoch": 0.9738820231876673, "grad_norm": 0.18980544805526733, "learning_rate": 2.599388379204893e-07, "loss": 0.0001, "step": 3822 }, { "epoch": 0.9741368327175436, "grad_norm": 0.18997935950756073, "learning_rate": 2.5739041794087666e-07, "loss": 0.0001, "step": 3823 }, { "epoch": 0.97439164224742, "grad_norm": 1.652574062347412, "learning_rate": 2.5484199796126404e-07, "loss": 0.0003, "step": 3824 }, { "epoch": 0.9746464517772965, "grad_norm": 0.9218937754631042, "learning_rate": 2.522935779816514e-07, "loss": 0.0001, "step": 3825 }, { "epoch": 0.9749012613071729, "grad_norm": 1.1046475172042847, "learning_rate": 2.497451580020388e-07, "loss": 0.0002, "step": 3826 }, { "epoch": 0.9751560708370494, "grad_norm": 0.5520105957984924, "learning_rate": 2.471967380224261e-07, "loss": 0.0001, "step": 3827 }, { "epoch": 0.9754108803669257, "grad_norm": 0.3665338158607483, "learning_rate": 2.446483180428135e-07, "loss": 0.0001, "step": 3828 }, { "epoch": 0.9756656898968021, "grad_norm": 0.5515462756156921, "learning_rate": 2.420998980632008e-07, "loss": 0.0001, "step": 3829 }, { "epoch": 0.9759204994266786, "grad_norm": 0.3731294274330139, "learning_rate": 2.395514780835882e-07, "loss": 0.0001, "step": 3830 }, { "epoch": 0.976175308956555, "grad_norm": 1.6541179418563843, "learning_rate": 2.3700305810397554e-07, "loss": 0.0003, "step": 3831 }, { "epoch": 0.9764301184864314, "grad_norm": 0.7323623299598694, "learning_rate": 2.344546381243629e-07, "loss": 0.0001, "step": 3832 }, { "epoch": 0.9766849280163078, "grad_norm": 0.3717745244503021, "learning_rate": 2.319062181447503e-07, "loss": 0.0001, "step": 3833 }, { "epoch": 0.9769397375461842, "grad_norm": 0.5531226396560669, "learning_rate": 2.2935779816513764e-07, "loss": 0.0001, "step": 3834 }, { "epoch": 0.9771945470760607, "grad_norm": 0.1897035837173462, "learning_rate": 2.26809378185525e-07, "loss": 0.0001, "step": 3835 }, { "epoch": 0.977449356605937, "grad_norm": 0.7341553568840027, "learning_rate": 2.2426095820591236e-07, "loss": 0.0001, "step": 3836 }, { "epoch": 0.9777041661358135, "grad_norm": 0.9198134541511536, "learning_rate": 2.2171253822629971e-07, "loss": 0.0002, "step": 3837 }, { "epoch": 0.9779589756656899, "grad_norm": 0.5516262054443359, "learning_rate": 2.1916411824668706e-07, "loss": 0.0001, "step": 3838 }, { "epoch": 0.9782137851955663, "grad_norm": 0.5621144771575928, "learning_rate": 2.1661569826707441e-07, "loss": 0.0002, "step": 3839 }, { "epoch": 0.9784685947254428, "grad_norm": 0.5517663955688477, "learning_rate": 2.140672782874618e-07, "loss": 0.0001, "step": 3840 }, { "epoch": 0.9787234042553191, "grad_norm": 1.8406730890274048, "learning_rate": 2.1151885830784914e-07, "loss": 0.0004, "step": 3841 }, { "epoch": 0.9789782137851956, "grad_norm": 0.7382583022117615, "learning_rate": 2.0897043832823652e-07, "loss": 0.0001, "step": 3842 }, { "epoch": 0.979233023315072, "grad_norm": 0.22377431392669678, "learning_rate": 2.064220183486239e-07, "loss": 0.0002, "step": 3843 }, { "epoch": 0.9794878328449484, "grad_norm": 1.1030762195587158, "learning_rate": 2.0387359836901124e-07, "loss": 0.0002, "step": 3844 }, { "epoch": 0.9797426423748248, "grad_norm": 0.7464997172355652, "learning_rate": 2.013251783893986e-07, "loss": 0.0003, "step": 3845 }, { "epoch": 0.9799974519047012, "grad_norm": 0.7324725985527039, "learning_rate": 1.9877675840978594e-07, "loss": 0.0001, "step": 3846 }, { "epoch": 0.9802522614345777, "grad_norm": 0.5596605539321899, "learning_rate": 1.9622833843017332e-07, "loss": 0.0002, "step": 3847 }, { "epoch": 0.9805070709644541, "grad_norm": 1.4782193899154663, "learning_rate": 1.9367991845056067e-07, "loss": 0.0003, "step": 3848 }, { "epoch": 0.9807618804943304, "grad_norm": 0.7434229254722595, "learning_rate": 1.9113149847094802e-07, "loss": 0.0002, "step": 3849 }, { "epoch": 0.9810166900242069, "grad_norm": 1.3027501106262207, "learning_rate": 1.8858307849133536e-07, "loss": 0.0004, "step": 3850 }, { "epoch": 0.9812714995540833, "grad_norm": 0.5509017109870911, "learning_rate": 1.8603465851172277e-07, "loss": 0.0001, "step": 3851 }, { "epoch": 0.9815263090839598, "grad_norm": 2.390479564666748, "learning_rate": 1.8348623853211012e-07, "loss": 0.0008, "step": 3852 }, { "epoch": 0.9817811186138362, "grad_norm": 0.3693541884422302, "learning_rate": 1.8093781855249747e-07, "loss": 0.0001, "step": 3853 }, { "epoch": 0.9820359281437125, "grad_norm": 4.120201587677002, "learning_rate": 1.7838939857288482e-07, "loss": 0.0024, "step": 3854 }, { "epoch": 0.982290737673589, "grad_norm": 0.26092061400413513, "learning_rate": 1.758409785932722e-07, "loss": 0.0004, "step": 3855 }, { "epoch": 0.9825455472034654, "grad_norm": 0.3670749366283417, "learning_rate": 1.7329255861365954e-07, "loss": 0.0001, "step": 3856 }, { "epoch": 0.9828003567333419, "grad_norm": 1.100860357284546, "learning_rate": 1.707441386340469e-07, "loss": 0.0002, "step": 3857 }, { "epoch": 0.9830551662632182, "grad_norm": 0.42142683267593384, "learning_rate": 1.6819571865443424e-07, "loss": 0.0005, "step": 3858 }, { "epoch": 0.9833099757930946, "grad_norm": 0.9160265922546387, "learning_rate": 1.6564729867482164e-07, "loss": 0.0001, "step": 3859 }, { "epoch": 0.9835647853229711, "grad_norm": 0.1890031397342682, "learning_rate": 1.63098878695209e-07, "loss": 0.0001, "step": 3860 }, { "epoch": 0.9838195948528475, "grad_norm": 0.733819305896759, "learning_rate": 1.6055045871559634e-07, "loss": 0.0001, "step": 3861 }, { "epoch": 0.9840744043827239, "grad_norm": 0.9313333630561829, "learning_rate": 1.5800203873598372e-07, "loss": 0.0003, "step": 3862 }, { "epoch": 0.9843292139126003, "grad_norm": 1.2893306016921997, "learning_rate": 1.5545361875637107e-07, "loss": 0.0002, "step": 3863 }, { "epoch": 0.9845840234424768, "grad_norm": 0.5506952404975891, "learning_rate": 1.5290519877675842e-07, "loss": 0.0001, "step": 3864 }, { "epoch": 0.9848388329723532, "grad_norm": 0.07853566855192184, "learning_rate": 1.5035677879714577e-07, "loss": 0.0001, "step": 3865 }, { "epoch": 0.9850936425022296, "grad_norm": 0.1898023784160614, "learning_rate": 1.4780835881753314e-07, "loss": 0.0001, "step": 3866 }, { "epoch": 0.985348452032106, "grad_norm": 1.4640926122665405, "learning_rate": 1.452599388379205e-07, "loss": 0.0002, "step": 3867 }, { "epoch": 0.9856032615619824, "grad_norm": 0.04004644230008125, "learning_rate": 1.4271151885830787e-07, "loss": 0.0001, "step": 3868 }, { "epoch": 0.9858580710918589, "grad_norm": 2.22625994682312, "learning_rate": 1.4016309887869522e-07, "loss": 0.0005, "step": 3869 }, { "epoch": 0.9861128806217353, "grad_norm": 2.2046256065368652, "learning_rate": 1.376146788990826e-07, "loss": 0.0004, "step": 3870 }, { "epoch": 0.9863676901516116, "grad_norm": 0.5550075173377991, "learning_rate": 1.3506625891946994e-07, "loss": 0.0001, "step": 3871 }, { "epoch": 0.9866224996814881, "grad_norm": 1.103770136833191, "learning_rate": 1.325178389398573e-07, "loss": 0.0002, "step": 3872 }, { "epoch": 0.9868773092113645, "grad_norm": 1.8320190906524658, "learning_rate": 1.2996941896024464e-07, "loss": 0.0003, "step": 3873 }, { "epoch": 0.987132118741241, "grad_norm": 1.2883317470550537, "learning_rate": 1.2742099898063202e-07, "loss": 0.0002, "step": 3874 }, { "epoch": 0.9873869282711173, "grad_norm": 1.4664080142974854, "learning_rate": 1.248725790010194e-07, "loss": 0.0002, "step": 3875 }, { "epoch": 0.9876417378009937, "grad_norm": 1.4671244621276855, "learning_rate": 1.2232415902140674e-07, "loss": 0.0002, "step": 3876 }, { "epoch": 0.9878965473308702, "grad_norm": 2.0268144607543945, "learning_rate": 1.197757390417941e-07, "loss": 0.0004, "step": 3877 }, { "epoch": 0.9881513568607466, "grad_norm": 0.37395355105400085, "learning_rate": 1.1722731906218144e-07, "loss": 0.0001, "step": 3878 }, { "epoch": 0.9884061663906231, "grad_norm": 0.7288954854011536, "learning_rate": 1.1467889908256882e-07, "loss": 0.0001, "step": 3879 }, { "epoch": 0.9886609759204994, "grad_norm": 2.3840157985687256, "learning_rate": 1.1213047910295618e-07, "loss": 0.0006, "step": 3880 }, { "epoch": 0.9889157854503758, "grad_norm": 0.7919071316719055, "learning_rate": 1.0958205912334353e-07, "loss": 0.0004, "step": 3881 }, { "epoch": 0.9891705949802523, "grad_norm": 1.8465124368667603, "learning_rate": 1.070336391437309e-07, "loss": 0.0003, "step": 3882 }, { "epoch": 0.9894254045101287, "grad_norm": 0.22556130588054657, "learning_rate": 1.0448521916411826e-07, "loss": 0.0002, "step": 3883 }, { "epoch": 0.9896802140400051, "grad_norm": 0.3957367241382599, "learning_rate": 1.0193679918450562e-07, "loss": 0.0002, "step": 3884 }, { "epoch": 0.9899350235698815, "grad_norm": 0.806275486946106, "learning_rate": 9.938837920489297e-08, "loss": 0.0007, "step": 3885 }, { "epoch": 0.9901898330997579, "grad_norm": 0.37404167652130127, "learning_rate": 9.683995922528033e-08, "loss": 0.0001, "step": 3886 }, { "epoch": 0.9904446426296344, "grad_norm": 0.9226627349853516, "learning_rate": 9.429153924566768e-08, "loss": 0.0002, "step": 3887 }, { "epoch": 0.9906994521595107, "grad_norm": 0.3666894733905792, "learning_rate": 9.174311926605506e-08, "loss": 0.0001, "step": 3888 }, { "epoch": 0.9909542616893872, "grad_norm": 0.13045720756053925, "learning_rate": 8.919469928644241e-08, "loss": 0.0002, "step": 3889 }, { "epoch": 0.9912090712192636, "grad_norm": 2.2070822715759277, "learning_rate": 8.664627930682977e-08, "loss": 0.0006, "step": 3890 }, { "epoch": 0.99146388074914, "grad_norm": 0.5533890128135681, "learning_rate": 8.409785932721712e-08, "loss": 0.0001, "step": 3891 }, { "epoch": 0.9917186902790165, "grad_norm": 0.937330961227417, "learning_rate": 8.15494393476045e-08, "loss": 0.0005, "step": 3892 }, { "epoch": 0.9919734998088928, "grad_norm": 2.189314126968384, "learning_rate": 7.900101936799186e-08, "loss": 0.0005, "step": 3893 }, { "epoch": 0.9922283093387693, "grad_norm": 0.5669463872909546, "learning_rate": 7.645259938837921e-08, "loss": 0.0001, "step": 3894 }, { "epoch": 0.9924831188686457, "grad_norm": 0.376699835062027, "learning_rate": 7.390417940876657e-08, "loss": 0.0002, "step": 3895 }, { "epoch": 0.9927379283985222, "grad_norm": 1.1021572351455688, "learning_rate": 7.135575942915393e-08, "loss": 0.0002, "step": 3896 }, { "epoch": 0.9929927379283985, "grad_norm": 0.5572252869606018, "learning_rate": 6.88073394495413e-08, "loss": 0.0002, "step": 3897 }, { "epoch": 0.9932475474582749, "grad_norm": 0.3750152289867401, "learning_rate": 6.625891946992865e-08, "loss": 0.0001, "step": 3898 }, { "epoch": 0.9935023569881514, "grad_norm": 0.7416457533836365, "learning_rate": 6.371049949031601e-08, "loss": 0.0001, "step": 3899 }, { "epoch": 0.9937571665180278, "grad_norm": 0.7376946210861206, "learning_rate": 6.116207951070337e-08, "loss": 0.0001, "step": 3900 }, { "epoch": 0.9940119760479041, "grad_norm": 0.19126904010772705, "learning_rate": 5.861365953109072e-08, "loss": 0.0001, "step": 3901 }, { "epoch": 0.9942667855777806, "grad_norm": 0.37203383445739746, "learning_rate": 5.606523955147809e-08, "loss": 0.0001, "step": 3902 }, { "epoch": 0.994521595107657, "grad_norm": 0.5565810799598694, "learning_rate": 5.351681957186545e-08, "loss": 0.0001, "step": 3903 }, { "epoch": 0.9947764046375335, "grad_norm": 0.3669120967388153, "learning_rate": 5.096839959225281e-08, "loss": 0.0001, "step": 3904 }, { "epoch": 0.9950312141674099, "grad_norm": 0.5587328672409058, "learning_rate": 4.8419979612640166e-08, "loss": 0.0002, "step": 3905 }, { "epoch": 0.9952860236972862, "grad_norm": 0.06684297323226929, "learning_rate": 4.587155963302753e-08, "loss": 0.0001, "step": 3906 }, { "epoch": 0.9955408332271627, "grad_norm": 1.1027741432189941, "learning_rate": 4.3323139653414885e-08, "loss": 0.0002, "step": 3907 }, { "epoch": 0.9957956427570391, "grad_norm": 3.2358102798461914, "learning_rate": 4.077471967380225e-08, "loss": 0.0013, "step": 3908 }, { "epoch": 0.9960504522869156, "grad_norm": 1.4757479429244995, "learning_rate": 3.8226299694189604e-08, "loss": 0.0002, "step": 3909 }, { "epoch": 0.9963052618167919, "grad_norm": 2.573387384414673, "learning_rate": 3.567787971457697e-08, "loss": 0.0005, "step": 3910 }, { "epoch": 0.9965600713466684, "grad_norm": 0.36739271879196167, "learning_rate": 3.3129459734964323e-08, "loss": 0.0001, "step": 3911 }, { "epoch": 0.9968148808765448, "grad_norm": 0.7394177913665771, "learning_rate": 3.0581039755351686e-08, "loss": 0.0001, "step": 3912 }, { "epoch": 0.9970696904064212, "grad_norm": 0.7411102652549744, "learning_rate": 2.8032619775739046e-08, "loss": 0.0002, "step": 3913 }, { "epoch": 0.9973244999362976, "grad_norm": 0.18536195158958435, "learning_rate": 2.5484199796126405e-08, "loss": 0.0001, "step": 3914 }, { "epoch": 0.997579309466174, "grad_norm": 0.7476856112480164, "learning_rate": 2.2935779816513765e-08, "loss": 0.0002, "step": 3915 }, { "epoch": 0.9978341189960505, "grad_norm": 0.12512223422527313, "learning_rate": 2.0387359836901124e-08, "loss": 0.0001, "step": 3916 }, { "epoch": 0.9980889285259269, "grad_norm": 0.20698951184749603, "learning_rate": 1.7838939857288484e-08, "loss": 0.0001, "step": 3917 }, { "epoch": 0.9983437380558033, "grad_norm": 1.2934640645980835, "learning_rate": 1.5290519877675843e-08, "loss": 0.0003, "step": 3918 }, { "epoch": 0.9985985475856797, "grad_norm": 0.5509880781173706, "learning_rate": 1.2742099898063203e-08, "loss": 0.0001, "step": 3919 }, { "epoch": 0.9988533571155561, "grad_norm": 0.7463032007217407, "learning_rate": 1.0193679918450562e-08, "loss": 0.0003, "step": 3920 }, { "epoch": 0.9991081666454326, "grad_norm": 0.19136406481266022, "learning_rate": 7.645259938837922e-09, "loss": 0.0001, "step": 3921 }, { "epoch": 0.999362976175309, "grad_norm": 0.4179600477218628, "learning_rate": 5.096839959225281e-09, "loss": 0.0005, "step": 3922 }, { "epoch": 0.9996177857051853, "grad_norm": 1.3107470273971558, "learning_rate": 2.5484199796126405e-09, "loss": 0.0004, "step": 3923 }, { "epoch": 0.9998725952350618, "grad_norm": 1.1066763401031494, "learning_rate": 0.0, "loss": 0.0002, "step": 3924 }, { "epoch": 0.9998725952350618, "eval_accuracy": 0.604494382022472, "eval_loss": 2.4827723503112793, "eval_runtime": 775.1537, "eval_samples_per_second": 1.156, "eval_steps_per_second": 1.156, "step": 3924 } ], "logging_steps": 1, "max_steps": 3924, "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 }