{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 6192, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00016149870801033592, "grad_norm": 76.79558563232422, "learning_rate": 9.998385012919898e-06, "loss": 1.3398, "step": 1 }, { "epoch": 0.00032299741602067185, "grad_norm": 70.90087890625, "learning_rate": 9.996770025839794e-06, "loss": 1.2188, "step": 2 }, { "epoch": 0.00048449612403100775, "grad_norm": 59.89802932739258, "learning_rate": 9.995155038759691e-06, "loss": 1.8398, "step": 3 }, { "epoch": 0.0006459948320413437, "grad_norm": 52.260406494140625, "learning_rate": 9.993540051679587e-06, "loss": 1.3594, "step": 4 }, { "epoch": 0.0008074935400516795, "grad_norm": 66.69401550292969, "learning_rate": 9.991925064599484e-06, "loss": 1.3281, "step": 5 }, { "epoch": 0.0009689922480620155, "grad_norm": 44.09524917602539, "learning_rate": 9.99031007751938e-06, "loss": 1.7578, "step": 6 }, { "epoch": 0.0011304909560723513, "grad_norm": 46.15640640258789, "learning_rate": 9.988695090439278e-06, "loss": 1.3145, "step": 7 }, { "epoch": 0.0012919896640826874, "grad_norm": 41.68496322631836, "learning_rate": 9.987080103359173e-06, "loss": 1.2695, "step": 8 }, { "epoch": 0.0014534883720930232, "grad_norm": 48.76371383666992, "learning_rate": 9.98546511627907e-06, "loss": 1.6133, "step": 9 }, { "epoch": 0.001614987080103359, "grad_norm": 46.967586517333984, "learning_rate": 9.983850129198968e-06, "loss": 1.2383, "step": 10 }, { "epoch": 0.0017764857881136951, "grad_norm": 59.64161682128906, "learning_rate": 9.982235142118864e-06, "loss": 1.6445, "step": 11 }, { "epoch": 0.001937984496124031, "grad_norm": 34.23611831665039, "learning_rate": 9.980620155038761e-06, "loss": 1.1953, "step": 12 }, { "epoch": 0.002099483204134367, "grad_norm": 58.193199157714844, "learning_rate": 9.979005167958657e-06, "loss": 2.1289, "step": 13 }, { "epoch": 0.0022609819121447027, "grad_norm": 41.048709869384766, "learning_rate": 9.977390180878553e-06, "loss": 0.9414, "step": 14 }, { "epoch": 0.0024224806201550387, "grad_norm": 44.809226989746094, "learning_rate": 9.97577519379845e-06, "loss": 1.3223, "step": 15 }, { "epoch": 0.002583979328165375, "grad_norm": 59.034183502197266, "learning_rate": 9.974160206718347e-06, "loss": 1.8398, "step": 16 }, { "epoch": 0.0027454780361757104, "grad_norm": 30.590232849121094, "learning_rate": 9.972545219638243e-06, "loss": 0.9746, "step": 17 }, { "epoch": 0.0029069767441860465, "grad_norm": 55.80778503417969, "learning_rate": 9.97093023255814e-06, "loss": 1.6797, "step": 18 }, { "epoch": 0.0030684754521963825, "grad_norm": 31.201833724975586, "learning_rate": 9.969315245478038e-06, "loss": 0.6533, "step": 19 }, { "epoch": 0.003229974160206718, "grad_norm": 51.043094635009766, "learning_rate": 9.967700258397934e-06, "loss": 1.6484, "step": 20 }, { "epoch": 0.003391472868217054, "grad_norm": 34.70353698730469, "learning_rate": 9.96608527131783e-06, "loss": 0.9922, "step": 21 }, { "epoch": 0.0035529715762273903, "grad_norm": 57.59230041503906, "learning_rate": 9.964470284237727e-06, "loss": 1.6953, "step": 22 }, { "epoch": 0.003714470284237726, "grad_norm": 45.892845153808594, "learning_rate": 9.962855297157623e-06, "loss": 1.6406, "step": 23 }, { "epoch": 0.003875968992248062, "grad_norm": 59.297691345214844, "learning_rate": 9.96124031007752e-06, "loss": 1.7383, "step": 24 }, { "epoch": 0.004037467700258398, "grad_norm": 76.88815307617188, "learning_rate": 9.959625322997417e-06, "loss": 1.9961, "step": 25 }, { "epoch": 0.004198966408268734, "grad_norm": 64.25715637207031, "learning_rate": 9.958010335917313e-06, "loss": 1.6074, "step": 26 }, { "epoch": 0.00436046511627907, "grad_norm": 48.399169921875, "learning_rate": 9.95639534883721e-06, "loss": 1.1992, "step": 27 }, { "epoch": 0.004521963824289405, "grad_norm": 44.927249908447266, "learning_rate": 9.954780361757108e-06, "loss": 1.1602, "step": 28 }, { "epoch": 0.004683462532299742, "grad_norm": 60.98359680175781, "learning_rate": 9.953165374677004e-06, "loss": 1.7148, "step": 29 }, { "epoch": 0.0048449612403100775, "grad_norm": 39.24678039550781, "learning_rate": 9.9515503875969e-06, "loss": 1.0723, "step": 30 }, { "epoch": 0.005006459948320413, "grad_norm": 42.5152473449707, "learning_rate": 9.949935400516797e-06, "loss": 1.2812, "step": 31 }, { "epoch": 0.00516795865633075, "grad_norm": 63.15961456298828, "learning_rate": 9.948320413436692e-06, "loss": 1.7578, "step": 32 }, { "epoch": 0.005329457364341085, "grad_norm": 66.47870635986328, "learning_rate": 9.94670542635659e-06, "loss": 1.1855, "step": 33 }, { "epoch": 0.005490956072351421, "grad_norm": 133.15084838867188, "learning_rate": 9.945090439276487e-06, "loss": 2.7031, "step": 34 }, { "epoch": 0.005652454780361757, "grad_norm": 44.101192474365234, "learning_rate": 9.943475452196383e-06, "loss": 1.6016, "step": 35 }, { "epoch": 0.005813953488372093, "grad_norm": 34.814064025878906, "learning_rate": 9.94186046511628e-06, "loss": 0.877, "step": 36 }, { "epoch": 0.005975452196382429, "grad_norm": 49.1778564453125, "learning_rate": 9.940245478036176e-06, "loss": 1.3203, "step": 37 }, { "epoch": 0.006136950904392765, "grad_norm": 54.12604522705078, "learning_rate": 9.938630490956073e-06, "loss": 1.9062, "step": 38 }, { "epoch": 0.006298449612403101, "grad_norm": 41.2735710144043, "learning_rate": 9.93701550387597e-06, "loss": 1.293, "step": 39 }, { "epoch": 0.006459948320413436, "grad_norm": 41.26066589355469, "learning_rate": 9.935400516795867e-06, "loss": 1.0039, "step": 40 }, { "epoch": 0.006621447028423773, "grad_norm": 57.77825927734375, "learning_rate": 9.933785529715762e-06, "loss": 1.7969, "step": 41 }, { "epoch": 0.006782945736434108, "grad_norm": 30.121261596679688, "learning_rate": 9.93217054263566e-06, "loss": 0.7305, "step": 42 }, { "epoch": 0.006944444444444444, "grad_norm": 27.481191635131836, "learning_rate": 9.930555555555557e-06, "loss": 0.4541, "step": 43 }, { "epoch": 0.0071059431524547806, "grad_norm": 59.36357116699219, "learning_rate": 9.928940568475453e-06, "loss": 1.2852, "step": 44 }, { "epoch": 0.007267441860465116, "grad_norm": 56.57508087158203, "learning_rate": 9.92732558139535e-06, "loss": 1.3428, "step": 45 }, { "epoch": 0.007428940568475452, "grad_norm": 56.65838623046875, "learning_rate": 9.925710594315246e-06, "loss": 1.1445, "step": 46 }, { "epoch": 0.007590439276485788, "grad_norm": 76.87432098388672, "learning_rate": 9.924095607235142e-06, "loss": 1.4102, "step": 47 }, { "epoch": 0.007751937984496124, "grad_norm": 48.070579528808594, "learning_rate": 9.922480620155039e-06, "loss": 1.3477, "step": 48 }, { "epoch": 0.00791343669250646, "grad_norm": 59.253841400146484, "learning_rate": 9.920865633074936e-06, "loss": 1.3477, "step": 49 }, { "epoch": 0.008074935400516795, "grad_norm": 101.00814056396484, "learning_rate": 9.919250645994832e-06, "loss": 2.3477, "step": 50 }, { "epoch": 0.008236434108527133, "grad_norm": 34.48191833496094, "learning_rate": 9.91763565891473e-06, "loss": 0.8965, "step": 51 }, { "epoch": 0.008397932816537468, "grad_norm": 36.63728713989258, "learning_rate": 9.916020671834627e-06, "loss": 1.002, "step": 52 }, { "epoch": 0.008559431524547804, "grad_norm": 30.57077980041504, "learning_rate": 9.914405684754523e-06, "loss": 0.9336, "step": 53 }, { "epoch": 0.00872093023255814, "grad_norm": 81.01558685302734, "learning_rate": 9.912790697674418e-06, "loss": 2.1758, "step": 54 }, { "epoch": 0.008882428940568475, "grad_norm": 82.27029418945312, "learning_rate": 9.911175710594316e-06, "loss": 2.8633, "step": 55 }, { "epoch": 0.00904392764857881, "grad_norm": 65.06700134277344, "learning_rate": 9.909560723514212e-06, "loss": 2.3086, "step": 56 }, { "epoch": 0.009205426356589148, "grad_norm": 30.321643829345703, "learning_rate": 9.907945736434109e-06, "loss": 0.959, "step": 57 }, { "epoch": 0.009366925064599484, "grad_norm": 50.278812408447266, "learning_rate": 9.906330749354006e-06, "loss": 1.7227, "step": 58 }, { "epoch": 0.00952842377260982, "grad_norm": 65.57909393310547, "learning_rate": 9.904715762273902e-06, "loss": 2.6953, "step": 59 }, { "epoch": 0.009689922480620155, "grad_norm": 52.525386810302734, "learning_rate": 9.9031007751938e-06, "loss": 2.0859, "step": 60 }, { "epoch": 0.00985142118863049, "grad_norm": 28.37175941467285, "learning_rate": 9.901485788113697e-06, "loss": 0.7969, "step": 61 }, { "epoch": 0.010012919896640826, "grad_norm": 33.81599807739258, "learning_rate": 9.899870801033593e-06, "loss": 1.2812, "step": 62 }, { "epoch": 0.010174418604651164, "grad_norm": 31.664100646972656, "learning_rate": 9.898255813953488e-06, "loss": 1.1016, "step": 63 }, { "epoch": 0.0103359173126615, "grad_norm": 29.604244232177734, "learning_rate": 9.896640826873386e-06, "loss": 0.8652, "step": 64 }, { "epoch": 0.010497416020671835, "grad_norm": 24.585224151611328, "learning_rate": 9.895025839793283e-06, "loss": 0.6709, "step": 65 }, { "epoch": 0.01065891472868217, "grad_norm": 31.173246383666992, "learning_rate": 9.893410852713179e-06, "loss": 1.0059, "step": 66 }, { "epoch": 0.010820413436692506, "grad_norm": 24.578813552856445, "learning_rate": 9.891795865633076e-06, "loss": 0.9102, "step": 67 }, { "epoch": 0.010981912144702842, "grad_norm": 42.838035583496094, "learning_rate": 9.890180878552974e-06, "loss": 1.7539, "step": 68 }, { "epoch": 0.011143410852713179, "grad_norm": 26.492156982421875, "learning_rate": 9.88856589147287e-06, "loss": 1.2227, "step": 69 }, { "epoch": 0.011304909560723515, "grad_norm": 24.39004898071289, "learning_rate": 9.886950904392765e-06, "loss": 0.9551, "step": 70 }, { "epoch": 0.01146640826873385, "grad_norm": 39.708492279052734, "learning_rate": 9.885335917312663e-06, "loss": 1.8203, "step": 71 }, { "epoch": 0.011627906976744186, "grad_norm": 55.32390594482422, "learning_rate": 9.883720930232558e-06, "loss": 2.2949, "step": 72 }, { "epoch": 0.011789405684754522, "grad_norm": 49.11772155761719, "learning_rate": 9.882105943152456e-06, "loss": 2.1797, "step": 73 }, { "epoch": 0.011950904392764857, "grad_norm": 30.340133666992188, "learning_rate": 9.880490956072353e-06, "loss": 1.293, "step": 74 }, { "epoch": 0.012112403100775194, "grad_norm": 31.55632781982422, "learning_rate": 9.878875968992249e-06, "loss": 0.7695, "step": 75 }, { "epoch": 0.01227390180878553, "grad_norm": 33.18221664428711, "learning_rate": 9.877260981912146e-06, "loss": 1.4766, "step": 76 }, { "epoch": 0.012435400516795866, "grad_norm": 52.166202545166016, "learning_rate": 9.875645994832042e-06, "loss": 1.6992, "step": 77 }, { "epoch": 0.012596899224806201, "grad_norm": 51.99414825439453, "learning_rate": 9.87403100775194e-06, "loss": 2.3594, "step": 78 }, { "epoch": 0.012758397932816537, "grad_norm": 32.5830192565918, "learning_rate": 9.872416020671835e-06, "loss": 0.959, "step": 79 }, { "epoch": 0.012919896640826873, "grad_norm": 67.59732818603516, "learning_rate": 9.870801033591732e-06, "loss": 2.1719, "step": 80 }, { "epoch": 0.01308139534883721, "grad_norm": 29.857242584228516, "learning_rate": 9.869186046511628e-06, "loss": 1.5195, "step": 81 }, { "epoch": 0.013242894056847546, "grad_norm": 38.38104248046875, "learning_rate": 9.867571059431526e-06, "loss": 1.8008, "step": 82 }, { "epoch": 0.013404392764857881, "grad_norm": 34.26373291015625, "learning_rate": 9.865956072351423e-06, "loss": 1.2539, "step": 83 }, { "epoch": 0.013565891472868217, "grad_norm": 53.08757781982422, "learning_rate": 9.864341085271319e-06, "loss": 2.2266, "step": 84 }, { "epoch": 0.013727390180878552, "grad_norm": 24.870634078979492, "learning_rate": 9.862726098191216e-06, "loss": 0.7461, "step": 85 }, { "epoch": 0.013888888888888888, "grad_norm": 35.57952117919922, "learning_rate": 9.861111111111112e-06, "loss": 1.6758, "step": 86 }, { "epoch": 0.014050387596899225, "grad_norm": 24.261667251586914, "learning_rate": 9.859496124031007e-06, "loss": 0.959, "step": 87 }, { "epoch": 0.014211886304909561, "grad_norm": 42.23406982421875, "learning_rate": 9.857881136950905e-06, "loss": 1.5625, "step": 88 }, { "epoch": 0.014373385012919897, "grad_norm": 34.1858024597168, "learning_rate": 9.856266149870802e-06, "loss": 1.3672, "step": 89 }, { "epoch": 0.014534883720930232, "grad_norm": 28.48642349243164, "learning_rate": 9.854651162790698e-06, "loss": 1.4805, "step": 90 }, { "epoch": 0.014696382428940568, "grad_norm": 24.300155639648438, "learning_rate": 9.853036175710595e-06, "loss": 0.9785, "step": 91 }, { "epoch": 0.014857881136950904, "grad_norm": 26.877588272094727, "learning_rate": 9.851421188630493e-06, "loss": 0.9863, "step": 92 }, { "epoch": 0.015019379844961241, "grad_norm": 26.83998680114746, "learning_rate": 9.849806201550389e-06, "loss": 1.3594, "step": 93 }, { "epoch": 0.015180878552971577, "grad_norm": 29.78314781188965, "learning_rate": 9.848191214470286e-06, "loss": 1.5039, "step": 94 }, { "epoch": 0.015342377260981912, "grad_norm": 28.145736694335938, "learning_rate": 9.846576227390182e-06, "loss": 1.2148, "step": 95 }, { "epoch": 0.015503875968992248, "grad_norm": 36.24938201904297, "learning_rate": 9.844961240310077e-06, "loss": 0.9727, "step": 96 }, { "epoch": 0.015665374677002585, "grad_norm": 29.262731552124023, "learning_rate": 9.843346253229975e-06, "loss": 1.3477, "step": 97 }, { "epoch": 0.01582687338501292, "grad_norm": 36.81642150878906, "learning_rate": 9.841731266149872e-06, "loss": 1.8789, "step": 98 }, { "epoch": 0.015988372093023256, "grad_norm": 33.66748809814453, "learning_rate": 9.840116279069768e-06, "loss": 1.5781, "step": 99 }, { "epoch": 0.01614987080103359, "grad_norm": 27.626798629760742, "learning_rate": 9.838501291989665e-06, "loss": 1.0957, "step": 100 }, { "epoch": 0.016311369509043928, "grad_norm": 37.41068649291992, "learning_rate": 9.836886304909563e-06, "loss": 0.9727, "step": 101 }, { "epoch": 0.016472868217054265, "grad_norm": 49.581050872802734, "learning_rate": 9.835271317829458e-06, "loss": 1.8242, "step": 102 }, { "epoch": 0.0166343669250646, "grad_norm": 48.01289749145508, "learning_rate": 9.833656330749354e-06, "loss": 1.8711, "step": 103 }, { "epoch": 0.016795865633074936, "grad_norm": 23.125659942626953, "learning_rate": 9.832041343669252e-06, "loss": 0.9219, "step": 104 }, { "epoch": 0.01695736434108527, "grad_norm": 34.77952194213867, "learning_rate": 9.830426356589147e-06, "loss": 1.2812, "step": 105 }, { "epoch": 0.017118863049095608, "grad_norm": 43.532840728759766, "learning_rate": 9.828811369509045e-06, "loss": 0.916, "step": 106 }, { "epoch": 0.01728036175710594, "grad_norm": 26.733367919921875, "learning_rate": 9.827196382428942e-06, "loss": 1.252, "step": 107 }, { "epoch": 0.01744186046511628, "grad_norm": 36.78434753417969, "learning_rate": 9.825581395348838e-06, "loss": 1.7695, "step": 108 }, { "epoch": 0.017603359173126616, "grad_norm": 37.09992599487305, "learning_rate": 9.823966408268735e-06, "loss": 1.2539, "step": 109 }, { "epoch": 0.01776485788113695, "grad_norm": 27.969463348388672, "learning_rate": 9.822351421188631e-06, "loss": 1.1211, "step": 110 }, { "epoch": 0.017926356589147287, "grad_norm": 27.1507568359375, "learning_rate": 9.820736434108528e-06, "loss": 1.2266, "step": 111 }, { "epoch": 0.01808785529715762, "grad_norm": 25.665510177612305, "learning_rate": 9.819121447028424e-06, "loss": 1.2598, "step": 112 }, { "epoch": 0.01824935400516796, "grad_norm": 29.078083038330078, "learning_rate": 9.817506459948321e-06, "loss": 1.293, "step": 113 }, { "epoch": 0.018410852713178296, "grad_norm": 46.223209381103516, "learning_rate": 9.815891472868217e-06, "loss": 1.1055, "step": 114 }, { "epoch": 0.01857235142118863, "grad_norm": 29.93814468383789, "learning_rate": 9.814276485788115e-06, "loss": 0.6885, "step": 115 }, { "epoch": 0.018733850129198967, "grad_norm": 73.2942123413086, "learning_rate": 9.812661498708012e-06, "loss": 3.2734, "step": 116 }, { "epoch": 0.0188953488372093, "grad_norm": 33.6378173828125, "learning_rate": 9.811046511627908e-06, "loss": 1.4648, "step": 117 }, { "epoch": 0.01905684754521964, "grad_norm": 31.302404403686523, "learning_rate": 9.809431524547805e-06, "loss": 1.5469, "step": 118 }, { "epoch": 0.019218346253229972, "grad_norm": 24.657724380493164, "learning_rate": 9.8078165374677e-06, "loss": 1.2305, "step": 119 }, { "epoch": 0.01937984496124031, "grad_norm": 31.328521728515625, "learning_rate": 9.806201550387598e-06, "loss": 1.3984, "step": 120 }, { "epoch": 0.019541343669250647, "grad_norm": 54.305084228515625, "learning_rate": 9.804586563307494e-06, "loss": 2.2617, "step": 121 }, { "epoch": 0.01970284237726098, "grad_norm": 25.969654083251953, "learning_rate": 9.802971576227391e-06, "loss": 0.9688, "step": 122 }, { "epoch": 0.01986434108527132, "grad_norm": 33.52921676635742, "learning_rate": 9.801356589147287e-06, "loss": 1.3242, "step": 123 }, { "epoch": 0.020025839793281652, "grad_norm": 28.040876388549805, "learning_rate": 9.799741602067184e-06, "loss": 1.5391, "step": 124 }, { "epoch": 0.02018733850129199, "grad_norm": 31.96416664123535, "learning_rate": 9.798126614987082e-06, "loss": 1.3242, "step": 125 }, { "epoch": 0.020348837209302327, "grad_norm": 23.03655433654785, "learning_rate": 9.796511627906978e-06, "loss": 0.7705, "step": 126 }, { "epoch": 0.02051033591731266, "grad_norm": 32.219482421875, "learning_rate": 9.794896640826875e-06, "loss": 1.5234, "step": 127 }, { "epoch": 0.020671834625323, "grad_norm": 27.610105514526367, "learning_rate": 9.79328165374677e-06, "loss": 1.4219, "step": 128 }, { "epoch": 0.020833333333333332, "grad_norm": 35.61982345581055, "learning_rate": 9.791666666666666e-06, "loss": 1.4805, "step": 129 }, { "epoch": 0.02099483204134367, "grad_norm": 32.32982635498047, "learning_rate": 9.790051679586564e-06, "loss": 1.75, "step": 130 }, { "epoch": 0.021156330749354007, "grad_norm": 21.430744171142578, "learning_rate": 9.788436692506461e-06, "loss": 0.8242, "step": 131 }, { "epoch": 0.02131782945736434, "grad_norm": 79.58429718017578, "learning_rate": 9.786821705426357e-06, "loss": 2.7578, "step": 132 }, { "epoch": 0.021479328165374678, "grad_norm": 29.776683807373047, "learning_rate": 9.785206718346254e-06, "loss": 0.8027, "step": 133 }, { "epoch": 0.021640826873385012, "grad_norm": 29.280120849609375, "learning_rate": 9.783591731266152e-06, "loss": 1.1094, "step": 134 }, { "epoch": 0.02180232558139535, "grad_norm": 26.895191192626953, "learning_rate": 9.781976744186047e-06, "loss": 0.7734, "step": 135 }, { "epoch": 0.021963824289405683, "grad_norm": 24.683116912841797, "learning_rate": 9.780361757105943e-06, "loss": 1.0098, "step": 136 }, { "epoch": 0.02212532299741602, "grad_norm": 32.72072219848633, "learning_rate": 9.77874677002584e-06, "loss": 1.4121, "step": 137 }, { "epoch": 0.022286821705426358, "grad_norm": 20.95906639099121, "learning_rate": 9.777131782945736e-06, "loss": 0.627, "step": 138 }, { "epoch": 0.022448320413436692, "grad_norm": 32.715301513671875, "learning_rate": 9.775516795865634e-06, "loss": 1.207, "step": 139 }, { "epoch": 0.02260981912144703, "grad_norm": 29.991962432861328, "learning_rate": 9.773901808785531e-06, "loss": 1.0039, "step": 140 }, { "epoch": 0.022771317829457363, "grad_norm": 33.30686569213867, "learning_rate": 9.772286821705427e-06, "loss": 1.2676, "step": 141 }, { "epoch": 0.0229328165374677, "grad_norm": 37.170806884765625, "learning_rate": 9.770671834625324e-06, "loss": 1.2461, "step": 142 }, { "epoch": 0.023094315245478038, "grad_norm": 22.080219268798828, "learning_rate": 9.769056847545222e-06, "loss": 0.6953, "step": 143 }, { "epoch": 0.023255813953488372, "grad_norm": 46.28438949584961, "learning_rate": 9.767441860465117e-06, "loss": 1.1172, "step": 144 }, { "epoch": 0.02341731266149871, "grad_norm": 79.79766845703125, "learning_rate": 9.765826873385013e-06, "loss": 2.4102, "step": 145 }, { "epoch": 0.023578811369509043, "grad_norm": 29.170948028564453, "learning_rate": 9.76421188630491e-06, "loss": 0.8076, "step": 146 }, { "epoch": 0.02374031007751938, "grad_norm": 23.113372802734375, "learning_rate": 9.762596899224806e-06, "loss": 0.5166, "step": 147 }, { "epoch": 0.023901808785529714, "grad_norm": 39.49415969848633, "learning_rate": 9.760981912144704e-06, "loss": 1.25, "step": 148 }, { "epoch": 0.02406330749354005, "grad_norm": 47.61834716796875, "learning_rate": 9.759366925064601e-06, "loss": 1.3477, "step": 149 }, { "epoch": 0.02422480620155039, "grad_norm": 45.54127502441406, "learning_rate": 9.757751937984497e-06, "loss": 1.4653, "step": 150 }, { "epoch": 0.024386304909560723, "grad_norm": 28.23267364501953, "learning_rate": 9.756136950904394e-06, "loss": 0.8354, "step": 151 }, { "epoch": 0.02454780361757106, "grad_norm": 43.325748443603516, "learning_rate": 9.75452196382429e-06, "loss": 1.1328, "step": 152 }, { "epoch": 0.024709302325581394, "grad_norm": 46.78194046020508, "learning_rate": 9.752906976744187e-06, "loss": 0.7402, "step": 153 }, { "epoch": 0.02487080103359173, "grad_norm": 42.832969665527344, "learning_rate": 9.751291989664083e-06, "loss": 1.4453, "step": 154 }, { "epoch": 0.02503229974160207, "grad_norm": 58.71327590942383, "learning_rate": 9.74967700258398e-06, "loss": 1.2383, "step": 155 }, { "epoch": 0.025193798449612403, "grad_norm": 26.87215232849121, "learning_rate": 9.748062015503876e-06, "loss": 0.5137, "step": 156 }, { "epoch": 0.02535529715762274, "grad_norm": 56.98737335205078, "learning_rate": 9.746447028423773e-06, "loss": 1.6445, "step": 157 }, { "epoch": 0.025516795865633074, "grad_norm": 32.82681655883789, "learning_rate": 9.744832041343671e-06, "loss": 0.3047, "step": 158 }, { "epoch": 0.02567829457364341, "grad_norm": 51.175025939941406, "learning_rate": 9.743217054263567e-06, "loss": 1.3379, "step": 159 }, { "epoch": 0.025839793281653745, "grad_norm": 50.284664154052734, "learning_rate": 9.741602067183464e-06, "loss": 0.627, "step": 160 }, { "epoch": 0.026001291989664083, "grad_norm": 38.82826232910156, "learning_rate": 9.73998708010336e-06, "loss": 0.9941, "step": 161 }, { "epoch": 0.02616279069767442, "grad_norm": 43.58761215209961, "learning_rate": 9.738372093023255e-06, "loss": 1.1484, "step": 162 }, { "epoch": 0.026324289405684754, "grad_norm": 46.85894775390625, "learning_rate": 9.736757105943153e-06, "loss": 0.4387, "step": 163 }, { "epoch": 0.02648578811369509, "grad_norm": 48.67445755004883, "learning_rate": 9.73514211886305e-06, "loss": 1.2314, "step": 164 }, { "epoch": 0.026647286821705425, "grad_norm": 91.58438110351562, "learning_rate": 9.733527131782946e-06, "loss": 4.1562, "step": 165 }, { "epoch": 0.026808785529715762, "grad_norm": 41.113773345947266, "learning_rate": 9.731912144702843e-06, "loss": 1.3477, "step": 166 }, { "epoch": 0.0269702842377261, "grad_norm": 51.14543533325195, "learning_rate": 9.73029715762274e-06, "loss": 1.2812, "step": 167 }, { "epoch": 0.027131782945736434, "grad_norm": 52.804569244384766, "learning_rate": 9.728682170542636e-06, "loss": 1.3906, "step": 168 }, { "epoch": 0.02729328165374677, "grad_norm": 45.17402267456055, "learning_rate": 9.727067183462532e-06, "loss": 1.4492, "step": 169 }, { "epoch": 0.027454780361757105, "grad_norm": 53.575931549072266, "learning_rate": 9.72545219638243e-06, "loss": 1.8281, "step": 170 }, { "epoch": 0.027616279069767442, "grad_norm": 103.90090942382812, "learning_rate": 9.723837209302325e-06, "loss": 3.3828, "step": 171 }, { "epoch": 0.027777777777777776, "grad_norm": 50.44481658935547, "learning_rate": 9.722222222222223e-06, "loss": 1.373, "step": 172 }, { "epoch": 0.027939276485788114, "grad_norm": 61.84478759765625, "learning_rate": 9.72060723514212e-06, "loss": 1.6719, "step": 173 }, { "epoch": 0.02810077519379845, "grad_norm": 44.471527099609375, "learning_rate": 9.718992248062016e-06, "loss": 0.8203, "step": 174 }, { "epoch": 0.028262273901808785, "grad_norm": 66.40045928955078, "learning_rate": 9.717377260981913e-06, "loss": 2.2227, "step": 175 }, { "epoch": 0.028423772609819122, "grad_norm": 79.96929931640625, "learning_rate": 9.71576227390181e-06, "loss": 3.0312, "step": 176 }, { "epoch": 0.028585271317829456, "grad_norm": 82.55130767822266, "learning_rate": 9.714147286821706e-06, "loss": 2.8984, "step": 177 }, { "epoch": 0.028746770025839793, "grad_norm": 41.79607009887695, "learning_rate": 9.712532299741602e-06, "loss": 1.3828, "step": 178 }, { "epoch": 0.02890826873385013, "grad_norm": 47.33134841918945, "learning_rate": 9.7109173126615e-06, "loss": 1.5898, "step": 179 }, { "epoch": 0.029069767441860465, "grad_norm": 31.636274337768555, "learning_rate": 9.709302325581395e-06, "loss": 1.1836, "step": 180 }, { "epoch": 0.029231266149870802, "grad_norm": 39.921592712402344, "learning_rate": 9.707687338501293e-06, "loss": 1.1187, "step": 181 }, { "epoch": 0.029392764857881136, "grad_norm": 53.35190200805664, "learning_rate": 9.70607235142119e-06, "loss": 1.5469, "step": 182 }, { "epoch": 0.029554263565891473, "grad_norm": 79.85649871826172, "learning_rate": 9.704457364341086e-06, "loss": 2.7305, "step": 183 }, { "epoch": 0.029715762273901807, "grad_norm": 44.54948425292969, "learning_rate": 9.702842377260983e-06, "loss": 1.2773, "step": 184 }, { "epoch": 0.029877260981912145, "grad_norm": 18.34815216064453, "learning_rate": 9.701227390180879e-06, "loss": 0.4849, "step": 185 }, { "epoch": 0.030038759689922482, "grad_norm": 28.623353958129883, "learning_rate": 9.699612403100776e-06, "loss": 1.1719, "step": 186 }, { "epoch": 0.030200258397932816, "grad_norm": 37.74894714355469, "learning_rate": 9.697997416020672e-06, "loss": 1.4062, "step": 187 }, { "epoch": 0.030361757105943153, "grad_norm": 29.39565658569336, "learning_rate": 9.69638242894057e-06, "loss": 1.1445, "step": 188 }, { "epoch": 0.030523255813953487, "grad_norm": 64.43157958984375, "learning_rate": 9.694767441860465e-06, "loss": 2.5938, "step": 189 }, { "epoch": 0.030684754521963824, "grad_norm": 54.83304214477539, "learning_rate": 9.693152454780362e-06, "loss": 2.3164, "step": 190 }, { "epoch": 0.030846253229974162, "grad_norm": 56.47328186035156, "learning_rate": 9.69153746770026e-06, "loss": 1.8945, "step": 191 }, { "epoch": 0.031007751937984496, "grad_norm": 32.101932525634766, "learning_rate": 9.689922480620156e-06, "loss": 1.4102, "step": 192 }, { "epoch": 0.031169250645994833, "grad_norm": 31.57245635986328, "learning_rate": 9.688307493540053e-06, "loss": 1.0068, "step": 193 }, { "epoch": 0.03133074935400517, "grad_norm": 21.364885330200195, "learning_rate": 9.686692506459949e-06, "loss": 0.8535, "step": 194 }, { "epoch": 0.031492248062015504, "grad_norm": 27.309587478637695, "learning_rate": 9.685077519379846e-06, "loss": 1.3008, "step": 195 }, { "epoch": 0.03165374677002584, "grad_norm": 35.116661071777344, "learning_rate": 9.683462532299742e-06, "loss": 2.0078, "step": 196 }, { "epoch": 0.03181524547803618, "grad_norm": 58.79365158081055, "learning_rate": 9.68184754521964e-06, "loss": 1.8555, "step": 197 }, { "epoch": 0.03197674418604651, "grad_norm": 24.708566665649414, "learning_rate": 9.680232558139537e-06, "loss": 1.4082, "step": 198 }, { "epoch": 0.03213824289405685, "grad_norm": 21.00739860534668, "learning_rate": 9.678617571059432e-06, "loss": 0.6836, "step": 199 }, { "epoch": 0.03229974160206718, "grad_norm": 30.47132682800293, "learning_rate": 9.67700258397933e-06, "loss": 1.7891, "step": 200 }, { "epoch": 0.03246124031007752, "grad_norm": 20.127620697021484, "learning_rate": 9.675387596899225e-06, "loss": 1.0391, "step": 201 }, { "epoch": 0.032622739018087855, "grad_norm": 38.66702651977539, "learning_rate": 9.673772609819123e-06, "loss": 0.9688, "step": 202 }, { "epoch": 0.03278423772609819, "grad_norm": 46.424007415771484, "learning_rate": 9.672157622739019e-06, "loss": 1.9531, "step": 203 }, { "epoch": 0.03294573643410853, "grad_norm": 35.09517288208008, "learning_rate": 9.670542635658916e-06, "loss": 1.8594, "step": 204 }, { "epoch": 0.033107235142118864, "grad_norm": 30.953365325927734, "learning_rate": 9.668927648578812e-06, "loss": 1.6719, "step": 205 }, { "epoch": 0.0332687338501292, "grad_norm": 31.368059158325195, "learning_rate": 9.667312661498709e-06, "loss": 1.4961, "step": 206 }, { "epoch": 0.03343023255813953, "grad_norm": 30.21981430053711, "learning_rate": 9.665697674418607e-06, "loss": 1.5898, "step": 207 }, { "epoch": 0.03359173126614987, "grad_norm": 25.168724060058594, "learning_rate": 9.664082687338502e-06, "loss": 1.1523, "step": 208 }, { "epoch": 0.03375322997416021, "grad_norm": 44.599727630615234, "learning_rate": 9.6624677002584e-06, "loss": 1.8594, "step": 209 }, { "epoch": 0.03391472868217054, "grad_norm": 31.627300262451172, "learning_rate": 9.660852713178295e-06, "loss": 1.6641, "step": 210 }, { "epoch": 0.03407622739018088, "grad_norm": 23.522380828857422, "learning_rate": 9.659237726098191e-06, "loss": 1.3203, "step": 211 }, { "epoch": 0.034237726098191215, "grad_norm": 26.411663055419922, "learning_rate": 9.657622739018088e-06, "loss": 1.4258, "step": 212 }, { "epoch": 0.03439922480620155, "grad_norm": 25.119327545166016, "learning_rate": 9.656007751937986e-06, "loss": 1.4883, "step": 213 }, { "epoch": 0.03456072351421188, "grad_norm": 21.356550216674805, "learning_rate": 9.654392764857882e-06, "loss": 1.25, "step": 214 }, { "epoch": 0.034722222222222224, "grad_norm": 29.246829986572266, "learning_rate": 9.652777777777779e-06, "loss": 1.2676, "step": 215 }, { "epoch": 0.03488372093023256, "grad_norm": 26.871627807617188, "learning_rate": 9.651162790697676e-06, "loss": 1.5391, "step": 216 }, { "epoch": 0.03504521963824289, "grad_norm": 33.735572814941406, "learning_rate": 9.649547803617572e-06, "loss": 1.375, "step": 217 }, { "epoch": 0.03520671834625323, "grad_norm": 37.9691047668457, "learning_rate": 9.647932816537468e-06, "loss": 1.8672, "step": 218 }, { "epoch": 0.035368217054263566, "grad_norm": 18.380769729614258, "learning_rate": 9.646317829457365e-06, "loss": 1.0352, "step": 219 }, { "epoch": 0.0355297157622739, "grad_norm": 18.796905517578125, "learning_rate": 9.644702842377261e-06, "loss": 1.1602, "step": 220 }, { "epoch": 0.03569121447028424, "grad_norm": 21.1939640045166, "learning_rate": 9.643087855297158e-06, "loss": 1.4961, "step": 221 }, { "epoch": 0.035852713178294575, "grad_norm": 19.483434677124023, "learning_rate": 9.641472868217056e-06, "loss": 1.1875, "step": 222 }, { "epoch": 0.03601421188630491, "grad_norm": 23.49736213684082, "learning_rate": 9.639857881136951e-06, "loss": 1.2617, "step": 223 }, { "epoch": 0.03617571059431524, "grad_norm": 21.91901397705078, "learning_rate": 9.638242894056849e-06, "loss": 1.3125, "step": 224 }, { "epoch": 0.036337209302325583, "grad_norm": 31.84812355041504, "learning_rate": 9.636627906976745e-06, "loss": 1.8867, "step": 225 }, { "epoch": 0.03649870801033592, "grad_norm": 30.72222328186035, "learning_rate": 9.635012919896642e-06, "loss": 1.707, "step": 226 }, { "epoch": 0.03666020671834625, "grad_norm": 29.555177688598633, "learning_rate": 9.633397932816538e-06, "loss": 0.832, "step": 227 }, { "epoch": 0.03682170542635659, "grad_norm": 23.892263412475586, "learning_rate": 9.631782945736435e-06, "loss": 1.1211, "step": 228 }, { "epoch": 0.036983204134366926, "grad_norm": 26.997596740722656, "learning_rate": 9.63016795865633e-06, "loss": 0.959, "step": 229 }, { "epoch": 0.03714470284237726, "grad_norm": 21.590991973876953, "learning_rate": 9.628552971576228e-06, "loss": 0.9434, "step": 230 }, { "epoch": 0.037306201550387594, "grad_norm": 19.47920799255371, "learning_rate": 9.626937984496126e-06, "loss": 1.1406, "step": 231 }, { "epoch": 0.037467700258397935, "grad_norm": 29.124311447143555, "learning_rate": 9.625322997416021e-06, "loss": 1.3594, "step": 232 }, { "epoch": 0.03762919896640827, "grad_norm": 22.828418731689453, "learning_rate": 9.623708010335919e-06, "loss": 1.0078, "step": 233 }, { "epoch": 0.0377906976744186, "grad_norm": 27.632158279418945, "learning_rate": 9.622093023255814e-06, "loss": 0.9375, "step": 234 }, { "epoch": 0.03795219638242894, "grad_norm": 21.554283142089844, "learning_rate": 9.620478036175712e-06, "loss": 1.2324, "step": 235 }, { "epoch": 0.03811369509043928, "grad_norm": 25.902626037597656, "learning_rate": 9.618863049095608e-06, "loss": 1.3633, "step": 236 }, { "epoch": 0.03827519379844961, "grad_norm": 30.330768585205078, "learning_rate": 9.617248062015505e-06, "loss": 1.1973, "step": 237 }, { "epoch": 0.038436692506459945, "grad_norm": 22.04692840576172, "learning_rate": 9.6156330749354e-06, "loss": 0.9697, "step": 238 }, { "epoch": 0.038598191214470286, "grad_norm": 14.584917068481445, "learning_rate": 9.614018087855298e-06, "loss": 0.4326, "step": 239 }, { "epoch": 0.03875968992248062, "grad_norm": 28.635988235473633, "learning_rate": 9.612403100775196e-06, "loss": 1.5254, "step": 240 }, { "epoch": 0.038921188630490954, "grad_norm": 41.21638488769531, "learning_rate": 9.610788113695091e-06, "loss": 1.6562, "step": 241 }, { "epoch": 0.039082687338501294, "grad_norm": 27.903766632080078, "learning_rate": 9.609173126614989e-06, "loss": 0.8247, "step": 242 }, { "epoch": 0.03924418604651163, "grad_norm": 35.776039123535156, "learning_rate": 9.607558139534884e-06, "loss": 1.5195, "step": 243 }, { "epoch": 0.03940568475452196, "grad_norm": 31.803220748901367, "learning_rate": 9.60594315245478e-06, "loss": 1.4258, "step": 244 }, { "epoch": 0.0395671834625323, "grad_norm": 23.998266220092773, "learning_rate": 9.604328165374677e-06, "loss": 1.4102, "step": 245 }, { "epoch": 0.03972868217054264, "grad_norm": 34.48735046386719, "learning_rate": 9.602713178294575e-06, "loss": 1.5781, "step": 246 }, { "epoch": 0.03989018087855297, "grad_norm": 32.193031311035156, "learning_rate": 9.60109819121447e-06, "loss": 1.1797, "step": 247 }, { "epoch": 0.040051679586563305, "grad_norm": 53.25713348388672, "learning_rate": 9.599483204134368e-06, "loss": 1.6328, "step": 248 }, { "epoch": 0.040213178294573645, "grad_norm": 39.01054763793945, "learning_rate": 9.597868217054265e-06, "loss": 1.168, "step": 249 }, { "epoch": 0.04037467700258398, "grad_norm": 37.28520584106445, "learning_rate": 9.596253229974161e-06, "loss": 1.3203, "step": 250 }, { "epoch": 0.04053617571059431, "grad_norm": 53.89059829711914, "learning_rate": 9.594638242894057e-06, "loss": 1.2578, "step": 251 }, { "epoch": 0.040697674418604654, "grad_norm": 36.912925720214844, "learning_rate": 9.593023255813954e-06, "loss": 1.2656, "step": 252 }, { "epoch": 0.04085917312661499, "grad_norm": 53.50445556640625, "learning_rate": 9.59140826873385e-06, "loss": 1.6836, "step": 253 }, { "epoch": 0.04102067183462532, "grad_norm": 57.46610641479492, "learning_rate": 9.589793281653747e-06, "loss": 2.7383, "step": 254 }, { "epoch": 0.041182170542635656, "grad_norm": 41.02031707763672, "learning_rate": 9.588178294573645e-06, "loss": 0.8613, "step": 255 }, { "epoch": 0.041343669250646, "grad_norm": 29.971975326538086, "learning_rate": 9.58656330749354e-06, "loss": 1.3887, "step": 256 }, { "epoch": 0.04150516795865633, "grad_norm": 36.6939697265625, "learning_rate": 9.584948320413438e-06, "loss": 0.8633, "step": 257 }, { "epoch": 0.041666666666666664, "grad_norm": 40.25952911376953, "learning_rate": 9.583333333333335e-06, "loss": 1.8906, "step": 258 }, { "epoch": 0.041828165374677005, "grad_norm": 44.906063079833984, "learning_rate": 9.581718346253231e-06, "loss": 0.7012, "step": 259 }, { "epoch": 0.04198966408268734, "grad_norm": 32.54353332519531, "learning_rate": 9.580103359173127e-06, "loss": 0.8359, "step": 260 }, { "epoch": 0.04215116279069767, "grad_norm": 42.065406799316406, "learning_rate": 9.578488372093024e-06, "loss": 1.1562, "step": 261 }, { "epoch": 0.042312661498708014, "grad_norm": 68.24365997314453, "learning_rate": 9.57687338501292e-06, "loss": 2.3828, "step": 262 }, { "epoch": 0.04247416020671835, "grad_norm": 23.46645164489746, "learning_rate": 9.575258397932817e-06, "loss": 0.7334, "step": 263 }, { "epoch": 0.04263565891472868, "grad_norm": 40.1244010925293, "learning_rate": 9.573643410852715e-06, "loss": 1.293, "step": 264 }, { "epoch": 0.042797157622739015, "grad_norm": 43.3118896484375, "learning_rate": 9.57202842377261e-06, "loss": 1.3516, "step": 265 }, { "epoch": 0.042958656330749356, "grad_norm": 53.284847259521484, "learning_rate": 9.570413436692508e-06, "loss": 1.5078, "step": 266 }, { "epoch": 0.04312015503875969, "grad_norm": 35.91545486450195, "learning_rate": 9.568798449612404e-06, "loss": 0.918, "step": 267 }, { "epoch": 0.043281653746770024, "grad_norm": 30.897903442382812, "learning_rate": 9.567183462532301e-06, "loss": 0.7568, "step": 268 }, { "epoch": 0.043443152454780365, "grad_norm": 45.199127197265625, "learning_rate": 9.565568475452197e-06, "loss": 1.7344, "step": 269 }, { "epoch": 0.0436046511627907, "grad_norm": 30.830127716064453, "learning_rate": 9.563953488372094e-06, "loss": 1.1797, "step": 270 }, { "epoch": 0.04376614987080103, "grad_norm": 46.216796875, "learning_rate": 9.56233850129199e-06, "loss": 1.397, "step": 271 }, { "epoch": 0.04392764857881137, "grad_norm": 32.46525192260742, "learning_rate": 9.560723514211887e-06, "loss": 0.6699, "step": 272 }, { "epoch": 0.04408914728682171, "grad_norm": 38.431400299072266, "learning_rate": 9.559108527131785e-06, "loss": 0.877, "step": 273 }, { "epoch": 0.04425064599483204, "grad_norm": 56.346282958984375, "learning_rate": 9.55749354005168e-06, "loss": 1.9961, "step": 274 }, { "epoch": 0.044412144702842375, "grad_norm": 52.36452102661133, "learning_rate": 9.555878552971578e-06, "loss": 1.2324, "step": 275 }, { "epoch": 0.044573643410852716, "grad_norm": 43.434547424316406, "learning_rate": 9.554263565891473e-06, "loss": 0.916, "step": 276 }, { "epoch": 0.04473514211886305, "grad_norm": 27.748117446899414, "learning_rate": 9.552648578811369e-06, "loss": 0.9346, "step": 277 }, { "epoch": 0.044896640826873384, "grad_norm": 31.430925369262695, "learning_rate": 9.551033591731267e-06, "loss": 1.0361, "step": 278 }, { "epoch": 0.04505813953488372, "grad_norm": 28.606788635253906, "learning_rate": 9.549418604651164e-06, "loss": 0.5391, "step": 279 }, { "epoch": 0.04521963824289406, "grad_norm": 30.86676025390625, "learning_rate": 9.54780361757106e-06, "loss": 0.6543, "step": 280 }, { "epoch": 0.04538113695090439, "grad_norm": 60.929256439208984, "learning_rate": 9.546188630490957e-06, "loss": 1.7324, "step": 281 }, { "epoch": 0.045542635658914726, "grad_norm": 49.21186065673828, "learning_rate": 9.544573643410854e-06, "loss": 2.0, "step": 282 }, { "epoch": 0.04570413436692507, "grad_norm": 31.7922420501709, "learning_rate": 9.54295865633075e-06, "loss": 0.8633, "step": 283 }, { "epoch": 0.0458656330749354, "grad_norm": 82.986083984375, "learning_rate": 9.541343669250646e-06, "loss": 2.4316, "step": 284 }, { "epoch": 0.046027131782945735, "grad_norm": 57.66022491455078, "learning_rate": 9.539728682170543e-06, "loss": 1.1084, "step": 285 }, { "epoch": 0.046188630490956076, "grad_norm": 47.75569152832031, "learning_rate": 9.538113695090439e-06, "loss": 1.166, "step": 286 }, { "epoch": 0.04635012919896641, "grad_norm": 47.829463958740234, "learning_rate": 9.536498708010336e-06, "loss": 1.9023, "step": 287 }, { "epoch": 0.046511627906976744, "grad_norm": 33.73856735229492, "learning_rate": 9.534883720930234e-06, "loss": 0.6484, "step": 288 }, { "epoch": 0.04667312661498708, "grad_norm": 103.53948974609375, "learning_rate": 9.53326873385013e-06, "loss": 2.4141, "step": 289 }, { "epoch": 0.04683462532299742, "grad_norm": 128.87644958496094, "learning_rate": 9.531653746770027e-06, "loss": 2.4023, "step": 290 }, { "epoch": 0.04699612403100775, "grad_norm": 41.125125885009766, "learning_rate": 9.530038759689924e-06, "loss": 1.0156, "step": 291 }, { "epoch": 0.047157622739018086, "grad_norm": 74.97247314453125, "learning_rate": 9.52842377260982e-06, "loss": 1.4844, "step": 292 }, { "epoch": 0.04731912144702843, "grad_norm": 48.1612434387207, "learning_rate": 9.526808785529716e-06, "loss": 0.4854, "step": 293 }, { "epoch": 0.04748062015503876, "grad_norm": 51.62887191772461, "learning_rate": 9.525193798449613e-06, "loss": 1.4121, "step": 294 }, { "epoch": 0.047642118863049095, "grad_norm": 34.96765899658203, "learning_rate": 9.523578811369509e-06, "loss": 0.5889, "step": 295 }, { "epoch": 0.04780361757105943, "grad_norm": 79.19114685058594, "learning_rate": 9.521963824289406e-06, "loss": 1.3574, "step": 296 }, { "epoch": 0.04796511627906977, "grad_norm": 43.7443733215332, "learning_rate": 9.520348837209304e-06, "loss": 1.2695, "step": 297 }, { "epoch": 0.0481266149870801, "grad_norm": 34.54993438720703, "learning_rate": 9.5187338501292e-06, "loss": 0.2529, "step": 298 }, { "epoch": 0.04828811369509044, "grad_norm": 36.50247573852539, "learning_rate": 9.517118863049097e-06, "loss": 0.6328, "step": 299 }, { "epoch": 0.04844961240310078, "grad_norm": 82.16258239746094, "learning_rate": 9.515503875968993e-06, "loss": 1.5977, "step": 300 }, { "epoch": 0.04861111111111111, "grad_norm": 97.453125, "learning_rate": 9.51388888888889e-06, "loss": 2.8086, "step": 301 }, { "epoch": 0.048772609819121446, "grad_norm": 34.92945861816406, "learning_rate": 9.512273901808786e-06, "loss": 0.7383, "step": 302 }, { "epoch": 0.04893410852713178, "grad_norm": 81.74836730957031, "learning_rate": 9.510658914728683e-06, "loss": 1.8203, "step": 303 }, { "epoch": 0.04909560723514212, "grad_norm": 48.100990295410156, "learning_rate": 9.509043927648579e-06, "loss": 1.167, "step": 304 }, { "epoch": 0.049257105943152454, "grad_norm": 45.8168830871582, "learning_rate": 9.507428940568476e-06, "loss": 0.9473, "step": 305 }, { "epoch": 0.04941860465116279, "grad_norm": 28.06601333618164, "learning_rate": 9.505813953488374e-06, "loss": 0.6113, "step": 306 }, { "epoch": 0.04958010335917313, "grad_norm": 37.84730911254883, "learning_rate": 9.50419896640827e-06, "loss": 1.4141, "step": 307 }, { "epoch": 0.04974160206718346, "grad_norm": 45.61326217651367, "learning_rate": 9.502583979328167e-06, "loss": 0.5869, "step": 308 }, { "epoch": 0.0499031007751938, "grad_norm": 91.35755920410156, "learning_rate": 9.500968992248062e-06, "loss": 3.0234, "step": 309 }, { "epoch": 0.05006459948320414, "grad_norm": 36.85948944091797, "learning_rate": 9.499354005167958e-06, "loss": 0.8535, "step": 310 }, { "epoch": 0.05022609819121447, "grad_norm": 36.11427307128906, "learning_rate": 9.497739018087856e-06, "loss": 0.4507, "step": 311 }, { "epoch": 0.050387596899224806, "grad_norm": 48.58005142211914, "learning_rate": 9.496124031007753e-06, "loss": 0.9355, "step": 312 }, { "epoch": 0.05054909560723514, "grad_norm": 87.75407409667969, "learning_rate": 9.494509043927649e-06, "loss": 2.4199, "step": 313 }, { "epoch": 0.05071059431524548, "grad_norm": 32.83212661743164, "learning_rate": 9.492894056847546e-06, "loss": 0.8252, "step": 314 }, { "epoch": 0.050872093023255814, "grad_norm": 41.517578125, "learning_rate": 9.491279069767443e-06, "loss": 1.2031, "step": 315 }, { "epoch": 0.05103359173126615, "grad_norm": 35.6826286315918, "learning_rate": 9.48966408268734e-06, "loss": 0.7285, "step": 316 }, { "epoch": 0.05119509043927649, "grad_norm": 39.52707290649414, "learning_rate": 9.488049095607237e-06, "loss": 0.7031, "step": 317 }, { "epoch": 0.05135658914728682, "grad_norm": 36.70552444458008, "learning_rate": 9.486434108527132e-06, "loss": 0.5063, "step": 318 }, { "epoch": 0.05151808785529716, "grad_norm": 80.50111389160156, "learning_rate": 9.484819121447028e-06, "loss": 2.2715, "step": 319 }, { "epoch": 0.05167958656330749, "grad_norm": 56.98918533325195, "learning_rate": 9.483204134366925e-06, "loss": 1.5469, "step": 320 }, { "epoch": 0.05184108527131783, "grad_norm": 32.921653747558594, "learning_rate": 9.481589147286823e-06, "loss": 1.1484, "step": 321 }, { "epoch": 0.052002583979328165, "grad_norm": 22.78489112854004, "learning_rate": 9.479974160206719e-06, "loss": 0.4697, "step": 322 }, { "epoch": 0.0521640826873385, "grad_norm": 69.04683685302734, "learning_rate": 9.478359173126616e-06, "loss": 1.6641, "step": 323 }, { "epoch": 0.05232558139534884, "grad_norm": 50.072391510009766, "learning_rate": 9.476744186046513e-06, "loss": 1.2148, "step": 324 }, { "epoch": 0.052487080103359174, "grad_norm": 44.03664779663086, "learning_rate": 9.475129198966409e-06, "loss": 0.709, "step": 325 }, { "epoch": 0.05264857881136951, "grad_norm": 48.81203842163086, "learning_rate": 9.473514211886305e-06, "loss": 1.0801, "step": 326 }, { "epoch": 0.05281007751937984, "grad_norm": 46.95895767211914, "learning_rate": 9.471899224806202e-06, "loss": 1.4922, "step": 327 }, { "epoch": 0.05297157622739018, "grad_norm": 37.54174041748047, "learning_rate": 9.4702842377261e-06, "loss": 0.6025, "step": 328 }, { "epoch": 0.053133074935400516, "grad_norm": 63.62781524658203, "learning_rate": 9.468669250645995e-06, "loss": 1.2246, "step": 329 }, { "epoch": 0.05329457364341085, "grad_norm": 61.5670166015625, "learning_rate": 9.467054263565893e-06, "loss": 1.4688, "step": 330 }, { "epoch": 0.05345607235142119, "grad_norm": 45.76570129394531, "learning_rate": 9.46543927648579e-06, "loss": 1.1709, "step": 331 }, { "epoch": 0.053617571059431525, "grad_norm": 44.48666000366211, "learning_rate": 9.463824289405686e-06, "loss": 0.9272, "step": 332 }, { "epoch": 0.05377906976744186, "grad_norm": 56.33510971069336, "learning_rate": 9.462209302325582e-06, "loss": 1.1094, "step": 333 }, { "epoch": 0.0539405684754522, "grad_norm": 94.06155395507812, "learning_rate": 9.460594315245479e-06, "loss": 1.8848, "step": 334 }, { "epoch": 0.054102067183462534, "grad_norm": 103.11168670654297, "learning_rate": 9.458979328165375e-06, "loss": 1.4431, "step": 335 }, { "epoch": 0.05426356589147287, "grad_norm": 61.9290657043457, "learning_rate": 9.457364341085272e-06, "loss": 0.8596, "step": 336 }, { "epoch": 0.0544250645994832, "grad_norm": 23.29356575012207, "learning_rate": 9.45574935400517e-06, "loss": 0.4929, "step": 337 }, { "epoch": 0.05458656330749354, "grad_norm": 59.13797378540039, "learning_rate": 9.454134366925065e-06, "loss": 1.3066, "step": 338 }, { "epoch": 0.054748062015503876, "grad_norm": 49.24556350708008, "learning_rate": 9.452519379844963e-06, "loss": 0.8496, "step": 339 }, { "epoch": 0.05490956072351421, "grad_norm": 17.406583786010742, "learning_rate": 9.450904392764858e-06, "loss": 0.2368, "step": 340 }, { "epoch": 0.05507105943152455, "grad_norm": 48.29962158203125, "learning_rate": 9.449289405684756e-06, "loss": 1.4043, "step": 341 }, { "epoch": 0.055232558139534885, "grad_norm": 45.57868576049805, "learning_rate": 9.447674418604651e-06, "loss": 1.1406, "step": 342 }, { "epoch": 0.05539405684754522, "grad_norm": 28.645864486694336, "learning_rate": 9.446059431524549e-06, "loss": 0.3105, "step": 343 }, { "epoch": 0.05555555555555555, "grad_norm": 80.31922149658203, "learning_rate": 9.444444444444445e-06, "loss": 1.3906, "step": 344 }, { "epoch": 0.05571705426356589, "grad_norm": 89.99958801269531, "learning_rate": 9.442829457364342e-06, "loss": 0.9727, "step": 345 }, { "epoch": 0.05587855297157623, "grad_norm": 95.4963607788086, "learning_rate": 9.44121447028424e-06, "loss": 1.9961, "step": 346 }, { "epoch": 0.05604005167958656, "grad_norm": 100.7720718383789, "learning_rate": 9.439599483204135e-06, "loss": 2.3848, "step": 347 }, { "epoch": 0.0562015503875969, "grad_norm": 57.150814056396484, "learning_rate": 9.437984496124032e-06, "loss": 1.1055, "step": 348 }, { "epoch": 0.056363049095607236, "grad_norm": 55.154781341552734, "learning_rate": 9.436369509043928e-06, "loss": 1.0244, "step": 349 }, { "epoch": 0.05652454780361757, "grad_norm": 26.17344856262207, "learning_rate": 9.434754521963826e-06, "loss": 0.4761, "step": 350 }, { "epoch": 0.056686046511627904, "grad_norm": 13.104007720947266, "learning_rate": 9.433139534883721e-06, "loss": 0.1033, "step": 351 }, { "epoch": 0.056847545219638244, "grad_norm": 35.44514083862305, "learning_rate": 9.431524547803619e-06, "loss": 0.6549, "step": 352 }, { "epoch": 0.05700904392764858, "grad_norm": 75.32524871826172, "learning_rate": 9.429909560723514e-06, "loss": 2.4531, "step": 353 }, { "epoch": 0.05717054263565891, "grad_norm": 85.32984924316406, "learning_rate": 9.428294573643412e-06, "loss": 3.3691, "step": 354 }, { "epoch": 0.05733204134366925, "grad_norm": 57.67392349243164, "learning_rate": 9.42667958656331e-06, "loss": 0.8047, "step": 355 }, { "epoch": 0.05749354005167959, "grad_norm": 56.84392166137695, "learning_rate": 9.425064599483205e-06, "loss": 0.7363, "step": 356 }, { "epoch": 0.05765503875968992, "grad_norm": 46.05898666381836, "learning_rate": 9.423449612403102e-06, "loss": 0.8672, "step": 357 }, { "epoch": 0.05781653746770026, "grad_norm": 56.180931091308594, "learning_rate": 9.421834625322998e-06, "loss": 1.1855, "step": 358 }, { "epoch": 0.057978036175710596, "grad_norm": 29.742061614990234, "learning_rate": 9.420219638242894e-06, "loss": 0.4736, "step": 359 }, { "epoch": 0.05813953488372093, "grad_norm": 24.45988655090332, "learning_rate": 9.418604651162791e-06, "loss": 0.4561, "step": 360 }, { "epoch": 0.05830103359173126, "grad_norm": 104.51405334472656, "learning_rate": 9.416989664082689e-06, "loss": 3.3184, "step": 361 }, { "epoch": 0.058462532299741604, "grad_norm": 60.7552375793457, "learning_rate": 9.415374677002584e-06, "loss": 1.46, "step": 362 }, { "epoch": 0.05862403100775194, "grad_norm": 48.536869049072266, "learning_rate": 9.413759689922482e-06, "loss": 1.293, "step": 363 }, { "epoch": 0.05878552971576227, "grad_norm": 124.3392562866211, "learning_rate": 9.412144702842379e-06, "loss": 2.9375, "step": 364 }, { "epoch": 0.05894702842377261, "grad_norm": 35.472930908203125, "learning_rate": 9.410529715762275e-06, "loss": 0.7705, "step": 365 }, { "epoch": 0.05910852713178295, "grad_norm": 44.34809494018555, "learning_rate": 9.40891472868217e-06, "loss": 0.488, "step": 366 }, { "epoch": 0.05927002583979328, "grad_norm": 40.80113220214844, "learning_rate": 9.407299741602068e-06, "loss": 1.1206, "step": 367 }, { "epoch": 0.059431524547803614, "grad_norm": 58.20582580566406, "learning_rate": 9.405684754521964e-06, "loss": 1.1406, "step": 368 }, { "epoch": 0.059593023255813955, "grad_norm": 31.242523193359375, "learning_rate": 9.404069767441861e-06, "loss": 0.3252, "step": 369 }, { "epoch": 0.05975452196382429, "grad_norm": 37.64901351928711, "learning_rate": 9.402454780361759e-06, "loss": 0.5801, "step": 370 }, { "epoch": 0.05991602067183462, "grad_norm": 42.750457763671875, "learning_rate": 9.400839793281654e-06, "loss": 1.2773, "step": 371 }, { "epoch": 0.060077519379844964, "grad_norm": 55.502323150634766, "learning_rate": 9.399224806201552e-06, "loss": 1.7031, "step": 372 }, { "epoch": 0.0602390180878553, "grad_norm": 35.980838775634766, "learning_rate": 9.397609819121449e-06, "loss": 0.7852, "step": 373 }, { "epoch": 0.06040051679586563, "grad_norm": 63.23392105102539, "learning_rate": 9.395994832041345e-06, "loss": 2.7764, "step": 374 }, { "epoch": 0.060562015503875966, "grad_norm": 37.408790588378906, "learning_rate": 9.39437984496124e-06, "loss": 0.6035, "step": 375 }, { "epoch": 0.060723514211886306, "grad_norm": 47.88526153564453, "learning_rate": 9.392764857881138e-06, "loss": 0.6636, "step": 376 }, { "epoch": 0.06088501291989664, "grad_norm": 48.52799606323242, "learning_rate": 9.391149870801034e-06, "loss": 1.5898, "step": 377 }, { "epoch": 0.061046511627906974, "grad_norm": 52.54782485961914, "learning_rate": 9.389534883720931e-06, "loss": 1.7852, "step": 378 }, { "epoch": 0.061208010335917315, "grad_norm": 81.26190185546875, "learning_rate": 9.387919896640828e-06, "loss": 1.7393, "step": 379 }, { "epoch": 0.06136950904392765, "grad_norm": 47.55427551269531, "learning_rate": 9.386304909560724e-06, "loss": 1.4121, "step": 380 }, { "epoch": 0.06153100775193798, "grad_norm": 35.2093505859375, "learning_rate": 9.384689922480622e-06, "loss": 0.6196, "step": 381 }, { "epoch": 0.061692506459948324, "grad_norm": 35.474666595458984, "learning_rate": 9.383074935400517e-06, "loss": 0.8047, "step": 382 }, { "epoch": 0.06185400516795866, "grad_norm": 18.937114715576172, "learning_rate": 9.381459948320415e-06, "loss": 0.6797, "step": 383 }, { "epoch": 0.06201550387596899, "grad_norm": 44.77072525024414, "learning_rate": 9.37984496124031e-06, "loss": 1.3345, "step": 384 }, { "epoch": 0.062177002583979325, "grad_norm": 32.076751708984375, "learning_rate": 9.378229974160208e-06, "loss": 0.9316, "step": 385 }, { "epoch": 0.062338501291989666, "grad_norm": 29.621992111206055, "learning_rate": 9.376614987080103e-06, "loss": 1.0703, "step": 386 }, { "epoch": 0.0625, "grad_norm": 34.96480178833008, "learning_rate": 9.375000000000001e-06, "loss": 1.0781, "step": 387 }, { "epoch": 0.06266149870801034, "grad_norm": 72.98635864257812, "learning_rate": 9.373385012919898e-06, "loss": 2.3359, "step": 388 }, { "epoch": 0.06282299741602067, "grad_norm": 38.61865997314453, "learning_rate": 9.371770025839794e-06, "loss": 1.3242, "step": 389 }, { "epoch": 0.06298449612403101, "grad_norm": 49.454647064208984, "learning_rate": 9.370155038759691e-06, "loss": 1.002, "step": 390 }, { "epoch": 0.06314599483204135, "grad_norm": 62.96749496459961, "learning_rate": 9.368540051679587e-06, "loss": 1.9805, "step": 391 }, { "epoch": 0.06330749354005168, "grad_norm": 48.340293884277344, "learning_rate": 9.366925064599483e-06, "loss": 1.1328, "step": 392 }, { "epoch": 0.06346899224806202, "grad_norm": 44.795047760009766, "learning_rate": 9.36531007751938e-06, "loss": 1.0918, "step": 393 }, { "epoch": 0.06363049095607236, "grad_norm": 52.10526657104492, "learning_rate": 9.363695090439278e-06, "loss": 1.6328, "step": 394 }, { "epoch": 0.06379198966408269, "grad_norm": 26.709774017333984, "learning_rate": 9.362080103359173e-06, "loss": 0.5293, "step": 395 }, { "epoch": 0.06395348837209303, "grad_norm": 31.561174392700195, "learning_rate": 9.36046511627907e-06, "loss": 0.752, "step": 396 }, { "epoch": 0.06411498708010335, "grad_norm": 116.9555892944336, "learning_rate": 9.358850129198968e-06, "loss": 1.998, "step": 397 }, { "epoch": 0.0642764857881137, "grad_norm": 35.43762969970703, "learning_rate": 9.357235142118864e-06, "loss": 1.1328, "step": 398 }, { "epoch": 0.06443798449612403, "grad_norm": 66.63896942138672, "learning_rate": 9.35562015503876e-06, "loss": 1.9258, "step": 399 }, { "epoch": 0.06459948320413436, "grad_norm": 73.59046173095703, "learning_rate": 9.354005167958657e-06, "loss": 2.0391, "step": 400 }, { "epoch": 0.0647609819121447, "grad_norm": 67.44512939453125, "learning_rate": 9.352390180878553e-06, "loss": 1.3086, "step": 401 }, { "epoch": 0.06492248062015504, "grad_norm": 39.174400329589844, "learning_rate": 9.35077519379845e-06, "loss": 2.1133, "step": 402 }, { "epoch": 0.06508397932816537, "grad_norm": 39.34491729736328, "learning_rate": 9.349160206718348e-06, "loss": 1.5664, "step": 403 }, { "epoch": 0.06524547803617571, "grad_norm": 28.21921730041504, "learning_rate": 9.347545219638243e-06, "loss": 1.043, "step": 404 }, { "epoch": 0.06540697674418605, "grad_norm": 70.06619262695312, "learning_rate": 9.34593023255814e-06, "loss": 2.0645, "step": 405 }, { "epoch": 0.06556847545219638, "grad_norm": 42.695858001708984, "learning_rate": 9.344315245478038e-06, "loss": 1.7969, "step": 406 }, { "epoch": 0.06572997416020672, "grad_norm": 48.17440414428711, "learning_rate": 9.342700258397934e-06, "loss": 1.1074, "step": 407 }, { "epoch": 0.06589147286821706, "grad_norm": 27.659908294677734, "learning_rate": 9.34108527131783e-06, "loss": 1.0449, "step": 408 }, { "epoch": 0.06605297157622739, "grad_norm": 32.70599365234375, "learning_rate": 9.339470284237727e-06, "loss": 1.3203, "step": 409 }, { "epoch": 0.06621447028423773, "grad_norm": 49.15275192260742, "learning_rate": 9.337855297157623e-06, "loss": 1.7656, "step": 410 }, { "epoch": 0.06637596899224807, "grad_norm": 39.97230911254883, "learning_rate": 9.33624031007752e-06, "loss": 1.3477, "step": 411 }, { "epoch": 0.0665374677002584, "grad_norm": 35.465126037597656, "learning_rate": 9.334625322997417e-06, "loss": 0.7285, "step": 412 }, { "epoch": 0.06669896640826874, "grad_norm": 32.8124885559082, "learning_rate": 9.333010335917313e-06, "loss": 1.1328, "step": 413 }, { "epoch": 0.06686046511627906, "grad_norm": 54.195343017578125, "learning_rate": 9.33139534883721e-06, "loss": 1.3438, "step": 414 }, { "epoch": 0.0670219638242894, "grad_norm": 17.94105339050293, "learning_rate": 9.329780361757106e-06, "loss": 0.498, "step": 415 }, { "epoch": 0.06718346253229975, "grad_norm": 17.813133239746094, "learning_rate": 9.328165374677004e-06, "loss": 0.3184, "step": 416 }, { "epoch": 0.06734496124031007, "grad_norm": 25.107379913330078, "learning_rate": 9.3265503875969e-06, "loss": 0.5254, "step": 417 }, { "epoch": 0.06750645994832041, "grad_norm": 36.548946380615234, "learning_rate": 9.324935400516797e-06, "loss": 1.1309, "step": 418 }, { "epoch": 0.06766795865633075, "grad_norm": 43.827537536621094, "learning_rate": 9.323320413436692e-06, "loss": 1.7969, "step": 419 }, { "epoch": 0.06782945736434108, "grad_norm": 23.29400062561035, "learning_rate": 9.32170542635659e-06, "loss": 0.4795, "step": 420 }, { "epoch": 0.06799095607235142, "grad_norm": 46.04701232910156, "learning_rate": 9.320090439276487e-06, "loss": 2.1797, "step": 421 }, { "epoch": 0.06815245478036176, "grad_norm": 29.428955078125, "learning_rate": 9.318475452196383e-06, "loss": 1.1562, "step": 422 }, { "epoch": 0.06831395348837209, "grad_norm": 11.082924842834473, "learning_rate": 9.31686046511628e-06, "loss": 0.2241, "step": 423 }, { "epoch": 0.06847545219638243, "grad_norm": 56.92595291137695, "learning_rate": 9.315245478036176e-06, "loss": 1.96, "step": 424 }, { "epoch": 0.06863695090439277, "grad_norm": 31.91712188720703, "learning_rate": 9.313630490956072e-06, "loss": 0.748, "step": 425 }, { "epoch": 0.0687984496124031, "grad_norm": 17.214500427246094, "learning_rate": 9.31201550387597e-06, "loss": 0.1191, "step": 426 }, { "epoch": 0.06895994832041344, "grad_norm": 42.616424560546875, "learning_rate": 9.310400516795867e-06, "loss": 1.1035, "step": 427 }, { "epoch": 0.06912144702842377, "grad_norm": 27.877328872680664, "learning_rate": 9.308785529715762e-06, "loss": 0.5801, "step": 428 }, { "epoch": 0.0692829457364341, "grad_norm": 57.3867301940918, "learning_rate": 9.30717054263566e-06, "loss": 1.3418, "step": 429 }, { "epoch": 0.06944444444444445, "grad_norm": 25.54570960998535, "learning_rate": 9.305555555555557e-06, "loss": 0.5234, "step": 430 }, { "epoch": 0.06960594315245477, "grad_norm": 65.53132629394531, "learning_rate": 9.303940568475453e-06, "loss": 1.2998, "step": 431 }, { "epoch": 0.06976744186046512, "grad_norm": 69.33045959472656, "learning_rate": 9.30232558139535e-06, "loss": 1.5762, "step": 432 }, { "epoch": 0.06992894056847546, "grad_norm": 74.90953063964844, "learning_rate": 9.300710594315246e-06, "loss": 1.9453, "step": 433 }, { "epoch": 0.07009043927648578, "grad_norm": 79.57827758789062, "learning_rate": 9.299095607235142e-06, "loss": 2.6631, "step": 434 }, { "epoch": 0.07025193798449612, "grad_norm": 47.0899658203125, "learning_rate": 9.297480620155039e-06, "loss": 1.2148, "step": 435 }, { "epoch": 0.07041343669250646, "grad_norm": 37.28929901123047, "learning_rate": 9.295865633074937e-06, "loss": 0.6074, "step": 436 }, { "epoch": 0.07057493540051679, "grad_norm": 95.96952056884766, "learning_rate": 9.294250645994832e-06, "loss": 2.7305, "step": 437 }, { "epoch": 0.07073643410852713, "grad_norm": 47.56111526489258, "learning_rate": 9.29263565891473e-06, "loss": 1.1816, "step": 438 }, { "epoch": 0.07089793281653747, "grad_norm": 51.389835357666016, "learning_rate": 9.291020671834627e-06, "loss": 1.8447, "step": 439 }, { "epoch": 0.0710594315245478, "grad_norm": 57.404014587402344, "learning_rate": 9.289405684754523e-06, "loss": 1.1523, "step": 440 }, { "epoch": 0.07122093023255814, "grad_norm": 74.69640350341797, "learning_rate": 9.287790697674419e-06, "loss": 1.6836, "step": 441 }, { "epoch": 0.07138242894056848, "grad_norm": 27.343976974487305, "learning_rate": 9.286175710594316e-06, "loss": 0.7593, "step": 442 }, { "epoch": 0.07154392764857881, "grad_norm": 48.86415100097656, "learning_rate": 9.284560723514212e-06, "loss": 0.5547, "step": 443 }, { "epoch": 0.07170542635658915, "grad_norm": 84.32017517089844, "learning_rate": 9.282945736434109e-06, "loss": 1.8477, "step": 444 }, { "epoch": 0.07186692506459948, "grad_norm": 40.04664611816406, "learning_rate": 9.281330749354006e-06, "loss": 0.5508, "step": 445 }, { "epoch": 0.07202842377260982, "grad_norm": 37.74442672729492, "learning_rate": 9.279715762273902e-06, "loss": 0.4443, "step": 446 }, { "epoch": 0.07218992248062016, "grad_norm": 121.65870666503906, "learning_rate": 9.2781007751938e-06, "loss": 1.0479, "step": 447 }, { "epoch": 0.07235142118863049, "grad_norm": 55.33258819580078, "learning_rate": 9.276485788113695e-06, "loss": 1.0503, "step": 448 }, { "epoch": 0.07251291989664083, "grad_norm": 25.8602352142334, "learning_rate": 9.274870801033593e-06, "loss": 0.6055, "step": 449 }, { "epoch": 0.07267441860465117, "grad_norm": 109.41778564453125, "learning_rate": 9.273255813953488e-06, "loss": 5.0859, "step": 450 }, { "epoch": 0.0728359173126615, "grad_norm": 32.19736099243164, "learning_rate": 9.271640826873386e-06, "loss": 0.9355, "step": 451 }, { "epoch": 0.07299741602067183, "grad_norm": 38.984886169433594, "learning_rate": 9.270025839793283e-06, "loss": 1.2715, "step": 452 }, { "epoch": 0.07315891472868218, "grad_norm": 39.770137786865234, "learning_rate": 9.268410852713179e-06, "loss": 1.2383, "step": 453 }, { "epoch": 0.0733204134366925, "grad_norm": 41.6244010925293, "learning_rate": 9.266795865633076e-06, "loss": 1.7188, "step": 454 }, { "epoch": 0.07348191214470284, "grad_norm": 37.79920196533203, "learning_rate": 9.265180878552972e-06, "loss": 0.8672, "step": 455 }, { "epoch": 0.07364341085271318, "grad_norm": 66.85942077636719, "learning_rate": 9.26356589147287e-06, "loss": 3.1875, "step": 456 }, { "epoch": 0.07380490956072351, "grad_norm": 28.475011825561523, "learning_rate": 9.261950904392765e-06, "loss": 0.8618, "step": 457 }, { "epoch": 0.07396640826873385, "grad_norm": 76.79132080078125, "learning_rate": 9.260335917312663e-06, "loss": 3.25, "step": 458 }, { "epoch": 0.07412790697674419, "grad_norm": 11.481006622314453, "learning_rate": 9.258720930232558e-06, "loss": 0.2271, "step": 459 }, { "epoch": 0.07428940568475452, "grad_norm": 19.52233123779297, "learning_rate": 9.257105943152456e-06, "loss": 0.3118, "step": 460 }, { "epoch": 0.07445090439276486, "grad_norm": 39.92616653442383, "learning_rate": 9.255490956072353e-06, "loss": 1.123, "step": 461 }, { "epoch": 0.07461240310077519, "grad_norm": 20.58506202697754, "learning_rate": 9.253875968992249e-06, "loss": 0.7363, "step": 462 }, { "epoch": 0.07477390180878553, "grad_norm": 16.45789337158203, "learning_rate": 9.252260981912146e-06, "loss": 0.1885, "step": 463 }, { "epoch": 0.07493540051679587, "grad_norm": 62.099552154541016, "learning_rate": 9.250645994832042e-06, "loss": 1.9141, "step": 464 }, { "epoch": 0.0750968992248062, "grad_norm": 33.513999938964844, "learning_rate": 9.24903100775194e-06, "loss": 1.2168, "step": 465 }, { "epoch": 0.07525839793281654, "grad_norm": 69.66895294189453, "learning_rate": 9.247416020671835e-06, "loss": 2.0273, "step": 466 }, { "epoch": 0.07541989664082688, "grad_norm": 34.18638229370117, "learning_rate": 9.245801033591732e-06, "loss": 1.248, "step": 467 }, { "epoch": 0.0755813953488372, "grad_norm": 20.840694427490234, "learning_rate": 9.244186046511628e-06, "loss": 0.3887, "step": 468 }, { "epoch": 0.07574289405684755, "grad_norm": 33.8105354309082, "learning_rate": 9.242571059431526e-06, "loss": 1.5391, "step": 469 }, { "epoch": 0.07590439276485789, "grad_norm": 15.862746238708496, "learning_rate": 9.240956072351423e-06, "loss": 0.4863, "step": 470 }, { "epoch": 0.07606589147286821, "grad_norm": 27.460105895996094, "learning_rate": 9.239341085271319e-06, "loss": 0.5928, "step": 471 }, { "epoch": 0.07622739018087855, "grad_norm": 11.633153915405273, "learning_rate": 9.237726098191216e-06, "loss": 0.1909, "step": 472 }, { "epoch": 0.0763888888888889, "grad_norm": 60.626976013183594, "learning_rate": 9.236111111111112e-06, "loss": 2.3672, "step": 473 }, { "epoch": 0.07655038759689922, "grad_norm": 95.63785552978516, "learning_rate": 9.234496124031008e-06, "loss": 3.6211, "step": 474 }, { "epoch": 0.07671188630490956, "grad_norm": 32.22743225097656, "learning_rate": 9.232881136950905e-06, "loss": 0.7988, "step": 475 }, { "epoch": 0.07687338501291989, "grad_norm": 46.9025993347168, "learning_rate": 9.231266149870802e-06, "loss": 1.418, "step": 476 }, { "epoch": 0.07703488372093023, "grad_norm": 22.55260467529297, "learning_rate": 9.229651162790698e-06, "loss": 0.5288, "step": 477 }, { "epoch": 0.07719638242894057, "grad_norm": 21.301546096801758, "learning_rate": 9.228036175710595e-06, "loss": 0.1241, "step": 478 }, { "epoch": 0.0773578811369509, "grad_norm": 45.44717025756836, "learning_rate": 9.226421188630493e-06, "loss": 1.793, "step": 479 }, { "epoch": 0.07751937984496124, "grad_norm": 30.089771270751953, "learning_rate": 9.224806201550389e-06, "loss": 0.5273, "step": 480 }, { "epoch": 0.07768087855297158, "grad_norm": 29.668230056762695, "learning_rate": 9.223191214470284e-06, "loss": 0.4438, "step": 481 }, { "epoch": 0.07784237726098191, "grad_norm": 9.294238090515137, "learning_rate": 9.221576227390182e-06, "loss": 0.0994, "step": 482 }, { "epoch": 0.07800387596899225, "grad_norm": 26.416683197021484, "learning_rate": 9.219961240310077e-06, "loss": 0.5752, "step": 483 }, { "epoch": 0.07816537467700259, "grad_norm": 78.75829315185547, "learning_rate": 9.218346253229975e-06, "loss": 1.4688, "step": 484 }, { "epoch": 0.07832687338501292, "grad_norm": 7.906054973602295, "learning_rate": 9.216731266149872e-06, "loss": 0.0669, "step": 485 }, { "epoch": 0.07848837209302326, "grad_norm": 43.29100036621094, "learning_rate": 9.215116279069768e-06, "loss": 1.5234, "step": 486 }, { "epoch": 0.0786498708010336, "grad_norm": 88.10053253173828, "learning_rate": 9.213501291989665e-06, "loss": 2.4609, "step": 487 }, { "epoch": 0.07881136950904392, "grad_norm": 80.28191375732422, "learning_rate": 9.211886304909563e-06, "loss": 1.4355, "step": 488 }, { "epoch": 0.07897286821705427, "grad_norm": 37.90756607055664, "learning_rate": 9.210271317829458e-06, "loss": 0.6953, "step": 489 }, { "epoch": 0.0791343669250646, "grad_norm": 64.09827423095703, "learning_rate": 9.208656330749354e-06, "loss": 0.5894, "step": 490 }, { "epoch": 0.07929586563307493, "grad_norm": 47.282596588134766, "learning_rate": 9.207041343669252e-06, "loss": 0.2251, "step": 491 }, { "epoch": 0.07945736434108527, "grad_norm": 70.66152954101562, "learning_rate": 9.205426356589147e-06, "loss": 2.2188, "step": 492 }, { "epoch": 0.0796188630490956, "grad_norm": 39.63001251220703, "learning_rate": 9.203811369509045e-06, "loss": 0.8848, "step": 493 }, { "epoch": 0.07978036175710594, "grad_norm": 112.05986022949219, "learning_rate": 9.202196382428942e-06, "loss": 1.7358, "step": 494 }, { "epoch": 0.07994186046511628, "grad_norm": 36.11805725097656, "learning_rate": 9.200581395348838e-06, "loss": 0.5347, "step": 495 }, { "epoch": 0.08010335917312661, "grad_norm": 25.606605529785156, "learning_rate": 9.198966408268735e-06, "loss": 0.5636, "step": 496 }, { "epoch": 0.08026485788113695, "grad_norm": 39.319583892822266, "learning_rate": 9.197351421188631e-06, "loss": 0.5574, "step": 497 }, { "epoch": 0.08042635658914729, "grad_norm": 38.66518783569336, "learning_rate": 9.195736434108528e-06, "loss": 0.4824, "step": 498 }, { "epoch": 0.08058785529715762, "grad_norm": 100.18243408203125, "learning_rate": 9.194121447028424e-06, "loss": 4.7109, "step": 499 }, { "epoch": 0.08074935400516796, "grad_norm": 40.0194091796875, "learning_rate": 9.192506459948321e-06, "loss": 0.4419, "step": 500 }, { "epoch": 0.0809108527131783, "grad_norm": 30.686321258544922, "learning_rate": 9.190891472868217e-06, "loss": 0.6213, "step": 501 }, { "epoch": 0.08107235142118863, "grad_norm": 41.951419830322266, "learning_rate": 9.189276485788115e-06, "loss": 0.9297, "step": 502 }, { "epoch": 0.08123385012919897, "grad_norm": 26.307300567626953, "learning_rate": 9.187661498708012e-06, "loss": 0.2377, "step": 503 }, { "epoch": 0.08139534883720931, "grad_norm": 82.95475006103516, "learning_rate": 9.186046511627908e-06, "loss": 4.0801, "step": 504 }, { "epoch": 0.08155684754521964, "grad_norm": 40.22134017944336, "learning_rate": 9.184431524547805e-06, "loss": 1.0898, "step": 505 }, { "epoch": 0.08171834625322998, "grad_norm": 73.591796875, "learning_rate": 9.1828165374677e-06, "loss": 2.375, "step": 506 }, { "epoch": 0.08187984496124032, "grad_norm": 61.785247802734375, "learning_rate": 9.181201550387597e-06, "loss": 1.4883, "step": 507 }, { "epoch": 0.08204134366925064, "grad_norm": 18.712125778198242, "learning_rate": 9.179586563307494e-06, "loss": 0.2295, "step": 508 }, { "epoch": 0.08220284237726098, "grad_norm": 51.85026550292969, "learning_rate": 9.177971576227391e-06, "loss": 1.1572, "step": 509 }, { "epoch": 0.08236434108527131, "grad_norm": 46.660301208496094, "learning_rate": 9.176356589147287e-06, "loss": 1.4277, "step": 510 }, { "epoch": 0.08252583979328165, "grad_norm": 98.59659576416016, "learning_rate": 9.174741602067184e-06, "loss": 4.2266, "step": 511 }, { "epoch": 0.082687338501292, "grad_norm": 38.266056060791016, "learning_rate": 9.173126614987082e-06, "loss": 0.4277, "step": 512 }, { "epoch": 0.08284883720930232, "grad_norm": 25.113510131835938, "learning_rate": 9.171511627906978e-06, "loss": 0.5215, "step": 513 }, { "epoch": 0.08301033591731266, "grad_norm": 68.15489959716797, "learning_rate": 9.169896640826873e-06, "loss": 1.6938, "step": 514 }, { "epoch": 0.083171834625323, "grad_norm": 48.816734313964844, "learning_rate": 9.16828165374677e-06, "loss": 2.125, "step": 515 }, { "epoch": 0.08333333333333333, "grad_norm": 29.036781311035156, "learning_rate": 9.166666666666666e-06, "loss": 0.9551, "step": 516 }, { "epoch": 0.08349483204134367, "grad_norm": 58.68679428100586, "learning_rate": 9.165051679586564e-06, "loss": 2.2417, "step": 517 }, { "epoch": 0.08365633074935401, "grad_norm": 64.78298950195312, "learning_rate": 9.163436692506461e-06, "loss": 1.9429, "step": 518 }, { "epoch": 0.08381782945736434, "grad_norm": 27.06208038330078, "learning_rate": 9.161821705426357e-06, "loss": 0.6914, "step": 519 }, { "epoch": 0.08397932816537468, "grad_norm": 30.04177474975586, "learning_rate": 9.160206718346254e-06, "loss": 1.041, "step": 520 }, { "epoch": 0.08414082687338502, "grad_norm": 15.607624053955078, "learning_rate": 9.158591731266152e-06, "loss": 0.4253, "step": 521 }, { "epoch": 0.08430232558139535, "grad_norm": 45.514373779296875, "learning_rate": 9.156976744186047e-06, "loss": 1.3262, "step": 522 }, { "epoch": 0.08446382428940569, "grad_norm": 43.325748443603516, "learning_rate": 9.155361757105943e-06, "loss": 1.7285, "step": 523 }, { "epoch": 0.08462532299741603, "grad_norm": 15.6883544921875, "learning_rate": 9.15374677002584e-06, "loss": 0.3965, "step": 524 }, { "epoch": 0.08478682170542635, "grad_norm": 22.71440887451172, "learning_rate": 9.152131782945736e-06, "loss": 0.6035, "step": 525 }, { "epoch": 0.0849483204134367, "grad_norm": 49.13056945800781, "learning_rate": 9.150516795865634e-06, "loss": 1.6855, "step": 526 }, { "epoch": 0.08510981912144702, "grad_norm": 19.646501541137695, "learning_rate": 9.148901808785531e-06, "loss": 0.6274, "step": 527 }, { "epoch": 0.08527131782945736, "grad_norm": 19.901626586914062, "learning_rate": 9.147286821705427e-06, "loss": 0.4868, "step": 528 }, { "epoch": 0.0854328165374677, "grad_norm": 28.370525360107422, "learning_rate": 9.145671834625324e-06, "loss": 1.0098, "step": 529 }, { "epoch": 0.08559431524547803, "grad_norm": 39.339996337890625, "learning_rate": 9.14405684754522e-06, "loss": 1.7578, "step": 530 }, { "epoch": 0.08575581395348837, "grad_norm": 57.73760986328125, "learning_rate": 9.142441860465117e-06, "loss": 1.6602, "step": 531 }, { "epoch": 0.08591731266149871, "grad_norm": 32.78150177001953, "learning_rate": 9.140826873385013e-06, "loss": 1.0312, "step": 532 }, { "epoch": 0.08607881136950904, "grad_norm": 8.090594291687012, "learning_rate": 9.13921188630491e-06, "loss": 0.1633, "step": 533 }, { "epoch": 0.08624031007751938, "grad_norm": 46.22599411010742, "learning_rate": 9.137596899224806e-06, "loss": 1.457, "step": 534 }, { "epoch": 0.08640180878552972, "grad_norm": 67.69791412353516, "learning_rate": 9.135981912144704e-06, "loss": 2.0898, "step": 535 }, { "epoch": 0.08656330749354005, "grad_norm": 47.39612579345703, "learning_rate": 9.134366925064601e-06, "loss": 1.334, "step": 536 }, { "epoch": 0.08672480620155039, "grad_norm": 72.16323852539062, "learning_rate": 9.132751937984497e-06, "loss": 2.2773, "step": 537 }, { "epoch": 0.08688630490956073, "grad_norm": 43.25524139404297, "learning_rate": 9.131136950904394e-06, "loss": 2.1445, "step": 538 }, { "epoch": 0.08704780361757106, "grad_norm": 50.8029670715332, "learning_rate": 9.12952196382429e-06, "loss": 2.5859, "step": 539 }, { "epoch": 0.0872093023255814, "grad_norm": 39.96868896484375, "learning_rate": 9.127906976744186e-06, "loss": 0.9814, "step": 540 }, { "epoch": 0.08737080103359172, "grad_norm": 34.395416259765625, "learning_rate": 9.126291989664083e-06, "loss": 1.543, "step": 541 }, { "epoch": 0.08753229974160207, "grad_norm": 45.53892135620117, "learning_rate": 9.12467700258398e-06, "loss": 2.1211, "step": 542 }, { "epoch": 0.0876937984496124, "grad_norm": 38.29220199584961, "learning_rate": 9.123062015503876e-06, "loss": 1.449, "step": 543 }, { "epoch": 0.08785529715762273, "grad_norm": 23.146509170532227, "learning_rate": 9.121447028423773e-06, "loss": 0.5127, "step": 544 }, { "epoch": 0.08801679586563307, "grad_norm": 48.36539077758789, "learning_rate": 9.119832041343671e-06, "loss": 1.3906, "step": 545 }, { "epoch": 0.08817829457364341, "grad_norm": 21.01957893371582, "learning_rate": 9.118217054263567e-06, "loss": 0.4912, "step": 546 }, { "epoch": 0.08833979328165374, "grad_norm": 41.54667663574219, "learning_rate": 9.116602067183464e-06, "loss": 1.3125, "step": 547 }, { "epoch": 0.08850129198966408, "grad_norm": 39.7177734375, "learning_rate": 9.11498708010336e-06, "loss": 0.6572, "step": 548 }, { "epoch": 0.08866279069767442, "grad_norm": 29.422260284423828, "learning_rate": 9.113372093023255e-06, "loss": 0.5498, "step": 549 }, { "epoch": 0.08882428940568475, "grad_norm": 20.286436080932617, "learning_rate": 9.111757105943153e-06, "loss": 0.7795, "step": 550 }, { "epoch": 0.08898578811369509, "grad_norm": 47.73656463623047, "learning_rate": 9.11014211886305e-06, "loss": 1.5742, "step": 551 }, { "epoch": 0.08914728682170543, "grad_norm": 58.252132415771484, "learning_rate": 9.108527131782946e-06, "loss": 2.3828, "step": 552 }, { "epoch": 0.08930878552971576, "grad_norm": 23.325889587402344, "learning_rate": 9.106912144702843e-06, "loss": 0.6426, "step": 553 }, { "epoch": 0.0894702842377261, "grad_norm": 37.148468017578125, "learning_rate": 9.10529715762274e-06, "loss": 1.6133, "step": 554 }, { "epoch": 0.08963178294573644, "grad_norm": 62.1334228515625, "learning_rate": 9.103682170542637e-06, "loss": 2.668, "step": 555 }, { "epoch": 0.08979328165374677, "grad_norm": 37.15915298461914, "learning_rate": 9.102067183462532e-06, "loss": 1.2227, "step": 556 }, { "epoch": 0.08995478036175711, "grad_norm": 24.174026489257812, "learning_rate": 9.10045219638243e-06, "loss": 0.5742, "step": 557 }, { "epoch": 0.09011627906976744, "grad_norm": 32.47032928466797, "learning_rate": 9.098837209302325e-06, "loss": 0.8613, "step": 558 }, { "epoch": 0.09027777777777778, "grad_norm": 41.50307083129883, "learning_rate": 9.097222222222223e-06, "loss": 1.791, "step": 559 }, { "epoch": 0.09043927648578812, "grad_norm": 31.593172073364258, "learning_rate": 9.09560723514212e-06, "loss": 0.4976, "step": 560 }, { "epoch": 0.09060077519379844, "grad_norm": 41.82628631591797, "learning_rate": 9.093992248062016e-06, "loss": 1.0215, "step": 561 }, { "epoch": 0.09076227390180878, "grad_norm": 59.07633590698242, "learning_rate": 9.092377260981913e-06, "loss": 0.752, "step": 562 }, { "epoch": 0.09092377260981913, "grad_norm": 35.98068618774414, "learning_rate": 9.090762273901809e-06, "loss": 1.3496, "step": 563 }, { "epoch": 0.09108527131782945, "grad_norm": 28.34218406677246, "learning_rate": 9.089147286821706e-06, "loss": 1.2051, "step": 564 }, { "epoch": 0.0912467700258398, "grad_norm": 55.721717834472656, "learning_rate": 9.087532299741602e-06, "loss": 1.4414, "step": 565 }, { "epoch": 0.09140826873385013, "grad_norm": 25.085033416748047, "learning_rate": 9.0859173126615e-06, "loss": 0.6621, "step": 566 }, { "epoch": 0.09156976744186046, "grad_norm": 16.089649200439453, "learning_rate": 9.084302325581395e-06, "loss": 0.3828, "step": 567 }, { "epoch": 0.0917312661498708, "grad_norm": 45.689613342285156, "learning_rate": 9.082687338501293e-06, "loss": 1.6445, "step": 568 }, { "epoch": 0.09189276485788114, "grad_norm": 57.20183563232422, "learning_rate": 9.08107235142119e-06, "loss": 1.8652, "step": 569 }, { "epoch": 0.09205426356589147, "grad_norm": 25.351097106933594, "learning_rate": 9.079457364341086e-06, "loss": 0.3428, "step": 570 }, { "epoch": 0.09221576227390181, "grad_norm": 45.26689910888672, "learning_rate": 9.077842377260983e-06, "loss": 1.3223, "step": 571 }, { "epoch": 0.09237726098191215, "grad_norm": 29.020069122314453, "learning_rate": 9.076227390180879e-06, "loss": 0.3672, "step": 572 }, { "epoch": 0.09253875968992248, "grad_norm": 57.11550521850586, "learning_rate": 9.074612403100775e-06, "loss": 1.5391, "step": 573 }, { "epoch": 0.09270025839793282, "grad_norm": 34.9595832824707, "learning_rate": 9.072997416020672e-06, "loss": 0.5771, "step": 574 }, { "epoch": 0.09286175710594315, "grad_norm": 39.040916442871094, "learning_rate": 9.07138242894057e-06, "loss": 1.2148, "step": 575 }, { "epoch": 0.09302325581395349, "grad_norm": 24.6822566986084, "learning_rate": 9.069767441860465e-06, "loss": 0.7979, "step": 576 }, { "epoch": 0.09318475452196383, "grad_norm": 50.80165100097656, "learning_rate": 9.068152454780363e-06, "loss": 1.4766, "step": 577 }, { "epoch": 0.09334625322997415, "grad_norm": 23.38970375061035, "learning_rate": 9.06653746770026e-06, "loss": 0.458, "step": 578 }, { "epoch": 0.0935077519379845, "grad_norm": 49.772666931152344, "learning_rate": 9.064922480620156e-06, "loss": 1.3438, "step": 579 }, { "epoch": 0.09366925064599484, "grad_norm": 38.44062423706055, "learning_rate": 9.063307493540053e-06, "loss": 0.8423, "step": 580 }, { "epoch": 0.09383074935400516, "grad_norm": 34.09309387207031, "learning_rate": 9.061692506459949e-06, "loss": 0.7983, "step": 581 }, { "epoch": 0.0939922480620155, "grad_norm": 43.086387634277344, "learning_rate": 9.060077519379846e-06, "loss": 0.7754, "step": 582 }, { "epoch": 0.09415374677002585, "grad_norm": 30.14906120300293, "learning_rate": 9.058462532299742e-06, "loss": 0.5288, "step": 583 }, { "epoch": 0.09431524547803617, "grad_norm": 26.234106063842773, "learning_rate": 9.05684754521964e-06, "loss": 0.822, "step": 584 }, { "epoch": 0.09447674418604651, "grad_norm": 60.70266342163086, "learning_rate": 9.055232558139537e-06, "loss": 1.9424, "step": 585 }, { "epoch": 0.09463824289405685, "grad_norm": 15.322680473327637, "learning_rate": 9.053617571059432e-06, "loss": 0.1454, "step": 586 }, { "epoch": 0.09479974160206718, "grad_norm": 60.260292053222656, "learning_rate": 9.05200258397933e-06, "loss": 1.7031, "step": 587 }, { "epoch": 0.09496124031007752, "grad_norm": 32.79610824584961, "learning_rate": 9.050387596899226e-06, "loss": 0.9355, "step": 588 }, { "epoch": 0.09512273901808785, "grad_norm": 52.81801986694336, "learning_rate": 9.048772609819121e-06, "loss": 2.6562, "step": 589 }, { "epoch": 0.09528423772609819, "grad_norm": 11.446856498718262, "learning_rate": 9.047157622739019e-06, "loss": 0.1343, "step": 590 }, { "epoch": 0.09544573643410853, "grad_norm": 43.930206298828125, "learning_rate": 9.045542635658916e-06, "loss": 0.9902, "step": 591 }, { "epoch": 0.09560723514211886, "grad_norm": 47.64548110961914, "learning_rate": 9.043927648578812e-06, "loss": 0.9512, "step": 592 }, { "epoch": 0.0957687338501292, "grad_norm": 30.563987731933594, "learning_rate": 9.04231266149871e-06, "loss": 0.79, "step": 593 }, { "epoch": 0.09593023255813954, "grad_norm": 47.8863410949707, "learning_rate": 9.040697674418607e-06, "loss": 1.0493, "step": 594 }, { "epoch": 0.09609173126614987, "grad_norm": 20.698150634765625, "learning_rate": 9.039082687338502e-06, "loss": 0.3506, "step": 595 }, { "epoch": 0.0962532299741602, "grad_norm": 34.667999267578125, "learning_rate": 9.037467700258398e-06, "loss": 1.1094, "step": 596 }, { "epoch": 0.09641472868217055, "grad_norm": 86.57431030273438, "learning_rate": 9.035852713178295e-06, "loss": 3.4746, "step": 597 }, { "epoch": 0.09657622739018087, "grad_norm": 41.94252395629883, "learning_rate": 9.034237726098191e-06, "loss": 1.2344, "step": 598 }, { "epoch": 0.09673772609819122, "grad_norm": 60.8990478515625, "learning_rate": 9.032622739018089e-06, "loss": 2.0938, "step": 599 }, { "epoch": 0.09689922480620156, "grad_norm": 18.876256942749023, "learning_rate": 9.031007751937986e-06, "loss": 0.1594, "step": 600 }, { "epoch": 0.09706072351421188, "grad_norm": 39.76816177368164, "learning_rate": 9.029392764857882e-06, "loss": 1.25, "step": 601 }, { "epoch": 0.09722222222222222, "grad_norm": 56.813541412353516, "learning_rate": 9.027777777777779e-06, "loss": 1.543, "step": 602 }, { "epoch": 0.09738372093023256, "grad_norm": 56.65843963623047, "learning_rate": 9.026162790697676e-06, "loss": 1.3789, "step": 603 }, { "epoch": 0.09754521963824289, "grad_norm": 25.1234188079834, "learning_rate": 9.024547803617572e-06, "loss": 0.7258, "step": 604 }, { "epoch": 0.09770671834625323, "grad_norm": 69.47755432128906, "learning_rate": 9.022932816537468e-06, "loss": 3.3594, "step": 605 }, { "epoch": 0.09786821705426356, "grad_norm": 40.490478515625, "learning_rate": 9.021317829457365e-06, "loss": 1.2432, "step": 606 }, { "epoch": 0.0980297157622739, "grad_norm": 20.77425765991211, "learning_rate": 9.019702842377261e-06, "loss": 0.4668, "step": 607 }, { "epoch": 0.09819121447028424, "grad_norm": 21.385360717773438, "learning_rate": 9.018087855297158e-06, "loss": 0.8145, "step": 608 }, { "epoch": 0.09835271317829457, "grad_norm": 32.194786071777344, "learning_rate": 9.016472868217056e-06, "loss": 0.8838, "step": 609 }, { "epoch": 0.09851421188630491, "grad_norm": 33.370574951171875, "learning_rate": 9.014857881136952e-06, "loss": 1.2983, "step": 610 }, { "epoch": 0.09867571059431525, "grad_norm": 22.168067932128906, "learning_rate": 9.013242894056849e-06, "loss": 0.4102, "step": 611 }, { "epoch": 0.09883720930232558, "grad_norm": 81.25538635253906, "learning_rate": 9.011627906976745e-06, "loss": 3.2686, "step": 612 }, { "epoch": 0.09899870801033592, "grad_norm": 64.50383758544922, "learning_rate": 9.010012919896642e-06, "loss": 2.1445, "step": 613 }, { "epoch": 0.09916020671834626, "grad_norm": 18.526737213134766, "learning_rate": 9.008397932816538e-06, "loss": 0.3447, "step": 614 }, { "epoch": 0.09932170542635659, "grad_norm": 58.99739456176758, "learning_rate": 9.006782945736435e-06, "loss": 1.2148, "step": 615 }, { "epoch": 0.09948320413436693, "grad_norm": 19.787412643432617, "learning_rate": 9.005167958656331e-06, "loss": 0.3955, "step": 616 }, { "epoch": 0.09964470284237727, "grad_norm": 30.68622398376465, "learning_rate": 9.003552971576228e-06, "loss": 0.4551, "step": 617 }, { "epoch": 0.0998062015503876, "grad_norm": 24.581241607666016, "learning_rate": 9.001937984496126e-06, "loss": 0.6924, "step": 618 }, { "epoch": 0.09996770025839793, "grad_norm": 93.69095611572266, "learning_rate": 9.000322997416021e-06, "loss": 3.1172, "step": 619 }, { "epoch": 0.10012919896640828, "grad_norm": 42.77409744262695, "learning_rate": 8.998708010335919e-06, "loss": 0.7632, "step": 620 }, { "epoch": 0.1002906976744186, "grad_norm": 43.8558235168457, "learning_rate": 8.997093023255815e-06, "loss": 1.2461, "step": 621 }, { "epoch": 0.10045219638242894, "grad_norm": 58.30179214477539, "learning_rate": 8.99547803617571e-06, "loss": 1.7773, "step": 622 }, { "epoch": 0.10061369509043927, "grad_norm": 53.39860916137695, "learning_rate": 8.993863049095608e-06, "loss": 2.2656, "step": 623 }, { "epoch": 0.10077519379844961, "grad_norm": 20.622909545898438, "learning_rate": 8.992248062015505e-06, "loss": 0.7988, "step": 624 }, { "epoch": 0.10093669250645995, "grad_norm": 29.949777603149414, "learning_rate": 8.9906330749354e-06, "loss": 1.3594, "step": 625 }, { "epoch": 0.10109819121447028, "grad_norm": 51.681941986083984, "learning_rate": 8.989018087855298e-06, "loss": 0.9707, "step": 626 }, { "epoch": 0.10125968992248062, "grad_norm": 40.776329040527344, "learning_rate": 8.987403100775196e-06, "loss": 1.2578, "step": 627 }, { "epoch": 0.10142118863049096, "grad_norm": 55.418155670166016, "learning_rate": 8.985788113695091e-06, "loss": 2.1196, "step": 628 }, { "epoch": 0.10158268733850129, "grad_norm": 19.359556198120117, "learning_rate": 8.984173126614987e-06, "loss": 0.6033, "step": 629 }, { "epoch": 0.10174418604651163, "grad_norm": 19.37401008605957, "learning_rate": 8.982558139534884e-06, "loss": 0.5127, "step": 630 }, { "epoch": 0.10190568475452197, "grad_norm": 58.90756607055664, "learning_rate": 8.98094315245478e-06, "loss": 1.2266, "step": 631 }, { "epoch": 0.1020671834625323, "grad_norm": 37.25347900390625, "learning_rate": 8.979328165374678e-06, "loss": 1.3086, "step": 632 }, { "epoch": 0.10222868217054264, "grad_norm": 93.2126235961914, "learning_rate": 8.977713178294575e-06, "loss": 4.3359, "step": 633 }, { "epoch": 0.10239018087855298, "grad_norm": 24.16040802001953, "learning_rate": 8.97609819121447e-06, "loss": 0.7124, "step": 634 }, { "epoch": 0.1025516795865633, "grad_norm": 36.59452819824219, "learning_rate": 8.974483204134368e-06, "loss": 0.7803, "step": 635 }, { "epoch": 0.10271317829457365, "grad_norm": 25.3646183013916, "learning_rate": 8.972868217054265e-06, "loss": 0.4482, "step": 636 }, { "epoch": 0.10287467700258397, "grad_norm": 40.50971603393555, "learning_rate": 8.971253229974161e-06, "loss": 0.9951, "step": 637 }, { "epoch": 0.10303617571059431, "grad_norm": 113.4276123046875, "learning_rate": 8.969638242894057e-06, "loss": 4.9531, "step": 638 }, { "epoch": 0.10319767441860465, "grad_norm": 52.017459869384766, "learning_rate": 8.968023255813954e-06, "loss": 1.0605, "step": 639 }, { "epoch": 0.10335917312661498, "grad_norm": 67.75501251220703, "learning_rate": 8.96640826873385e-06, "loss": 2.4297, "step": 640 }, { "epoch": 0.10352067183462532, "grad_norm": 57.31987762451172, "learning_rate": 8.964793281653747e-06, "loss": 2.1953, "step": 641 }, { "epoch": 0.10368217054263566, "grad_norm": 44.04484939575195, "learning_rate": 8.963178294573645e-06, "loss": 1.8945, "step": 642 }, { "epoch": 0.10384366925064599, "grad_norm": 41.63870620727539, "learning_rate": 8.96156330749354e-06, "loss": 1.1191, "step": 643 }, { "epoch": 0.10400516795865633, "grad_norm": 50.96044158935547, "learning_rate": 8.959948320413438e-06, "loss": 2.0664, "step": 644 }, { "epoch": 0.10416666666666667, "grad_norm": 44.11690139770508, "learning_rate": 8.958333333333334e-06, "loss": 1.6484, "step": 645 }, { "epoch": 0.104328165374677, "grad_norm": 39.61762237548828, "learning_rate": 8.956718346253231e-06, "loss": 1.9883, "step": 646 }, { "epoch": 0.10448966408268734, "grad_norm": 49.37284469604492, "learning_rate": 8.955103359173127e-06, "loss": 1.7031, "step": 647 }, { "epoch": 0.10465116279069768, "grad_norm": 76.31446075439453, "learning_rate": 8.953488372093024e-06, "loss": 2.6836, "step": 648 }, { "epoch": 0.104812661498708, "grad_norm": 40.10307312011719, "learning_rate": 8.95187338501292e-06, "loss": 1.0312, "step": 649 }, { "epoch": 0.10497416020671835, "grad_norm": 41.13298797607422, "learning_rate": 8.950258397932817e-06, "loss": 1.3926, "step": 650 }, { "epoch": 0.10513565891472869, "grad_norm": 13.760480880737305, "learning_rate": 8.948643410852715e-06, "loss": 0.292, "step": 651 }, { "epoch": 0.10529715762273902, "grad_norm": 40.85681915283203, "learning_rate": 8.94702842377261e-06, "loss": 1.918, "step": 652 }, { "epoch": 0.10545865633074936, "grad_norm": 35.59889221191406, "learning_rate": 8.945413436692508e-06, "loss": 1.0254, "step": 653 }, { "epoch": 0.10562015503875968, "grad_norm": 41.90641784667969, "learning_rate": 8.943798449612404e-06, "loss": 1.6719, "step": 654 }, { "epoch": 0.10578165374677002, "grad_norm": 25.487030029296875, "learning_rate": 8.9421834625323e-06, "loss": 1.0469, "step": 655 }, { "epoch": 0.10594315245478036, "grad_norm": 26.41151237487793, "learning_rate": 8.940568475452197e-06, "loss": 0.8535, "step": 656 }, { "epoch": 0.10610465116279069, "grad_norm": 37.90562057495117, "learning_rate": 8.938953488372094e-06, "loss": 1.5703, "step": 657 }, { "epoch": 0.10626614987080103, "grad_norm": 30.40921401977539, "learning_rate": 8.93733850129199e-06, "loss": 0.9863, "step": 658 }, { "epoch": 0.10642764857881137, "grad_norm": 32.93626022338867, "learning_rate": 8.935723514211887e-06, "loss": 1.1172, "step": 659 }, { "epoch": 0.1065891472868217, "grad_norm": 32.05014419555664, "learning_rate": 8.934108527131785e-06, "loss": 1.2363, "step": 660 }, { "epoch": 0.10675064599483204, "grad_norm": 22.535696029663086, "learning_rate": 8.93249354005168e-06, "loss": 0.7471, "step": 661 }, { "epoch": 0.10691214470284238, "grad_norm": 40.582393646240234, "learning_rate": 8.930878552971578e-06, "loss": 1.1562, "step": 662 }, { "epoch": 0.10707364341085271, "grad_norm": 28.213638305664062, "learning_rate": 8.929263565891473e-06, "loss": 1.0264, "step": 663 }, { "epoch": 0.10723514211886305, "grad_norm": 34.64348220825195, "learning_rate": 8.92764857881137e-06, "loss": 1.7734, "step": 664 }, { "epoch": 0.10739664082687339, "grad_norm": 46.57194137573242, "learning_rate": 8.926033591731267e-06, "loss": 1.3008, "step": 665 }, { "epoch": 0.10755813953488372, "grad_norm": 39.15975570678711, "learning_rate": 8.924418604651164e-06, "loss": 1.6836, "step": 666 }, { "epoch": 0.10771963824289406, "grad_norm": 41.886165618896484, "learning_rate": 8.92280361757106e-06, "loss": 0.8984, "step": 667 }, { "epoch": 0.1078811369509044, "grad_norm": 35.46730041503906, "learning_rate": 8.921188630490957e-06, "loss": 0.9736, "step": 668 }, { "epoch": 0.10804263565891473, "grad_norm": 34.02943801879883, "learning_rate": 8.919573643410855e-06, "loss": 1.4297, "step": 669 }, { "epoch": 0.10820413436692507, "grad_norm": 22.211050033569336, "learning_rate": 8.91795865633075e-06, "loss": 0.5605, "step": 670 }, { "epoch": 0.1083656330749354, "grad_norm": 38.63640594482422, "learning_rate": 8.916343669250646e-06, "loss": 0.834, "step": 671 }, { "epoch": 0.10852713178294573, "grad_norm": 46.643760681152344, "learning_rate": 8.914728682170543e-06, "loss": 1.9502, "step": 672 }, { "epoch": 0.10868863049095608, "grad_norm": 30.743080139160156, "learning_rate": 8.913113695090439e-06, "loss": 0.8301, "step": 673 }, { "epoch": 0.1088501291989664, "grad_norm": 102.85321807861328, "learning_rate": 8.911498708010336e-06, "loss": 4.3594, "step": 674 }, { "epoch": 0.10901162790697674, "grad_norm": 22.991905212402344, "learning_rate": 8.909883720930234e-06, "loss": 0.5869, "step": 675 }, { "epoch": 0.10917312661498708, "grad_norm": 52.669498443603516, "learning_rate": 8.90826873385013e-06, "loss": 2.4102, "step": 676 }, { "epoch": 0.10933462532299741, "grad_norm": 60.64424514770508, "learning_rate": 8.906653746770027e-06, "loss": 1.6299, "step": 677 }, { "epoch": 0.10949612403100775, "grad_norm": 30.213382720947266, "learning_rate": 8.905038759689923e-06, "loss": 0.9463, "step": 678 }, { "epoch": 0.10965762273901809, "grad_norm": 33.178829193115234, "learning_rate": 8.90342377260982e-06, "loss": 1.4219, "step": 679 }, { "epoch": 0.10981912144702842, "grad_norm": 50.380374908447266, "learning_rate": 8.901808785529716e-06, "loss": 1.3125, "step": 680 }, { "epoch": 0.10998062015503876, "grad_norm": 34.22032928466797, "learning_rate": 8.900193798449613e-06, "loss": 0.9785, "step": 681 }, { "epoch": 0.1101421188630491, "grad_norm": 30.236793518066406, "learning_rate": 8.898578811369509e-06, "loss": 0.9229, "step": 682 }, { "epoch": 0.11030361757105943, "grad_norm": 52.37091064453125, "learning_rate": 8.896963824289406e-06, "loss": 2.1211, "step": 683 }, { "epoch": 0.11046511627906977, "grad_norm": 40.92532730102539, "learning_rate": 8.895348837209304e-06, "loss": 1.5273, "step": 684 }, { "epoch": 0.11062661498708011, "grad_norm": 44.73381042480469, "learning_rate": 8.8937338501292e-06, "loss": 1.5117, "step": 685 }, { "epoch": 0.11078811369509044, "grad_norm": 29.483234405517578, "learning_rate": 8.892118863049097e-06, "loss": 1.0312, "step": 686 }, { "epoch": 0.11094961240310078, "grad_norm": 59.080810546875, "learning_rate": 8.890503875968993e-06, "loss": 1.5176, "step": 687 }, { "epoch": 0.1111111111111111, "grad_norm": 74.19731140136719, "learning_rate": 8.888888888888888e-06, "loss": 2.1504, "step": 688 }, { "epoch": 0.11127260981912145, "grad_norm": 45.683326721191406, "learning_rate": 8.887273901808786e-06, "loss": 1.1836, "step": 689 }, { "epoch": 0.11143410852713179, "grad_norm": 41.623497009277344, "learning_rate": 8.885658914728683e-06, "loss": 1.3496, "step": 690 }, { "epoch": 0.11159560723514211, "grad_norm": 29.183673858642578, "learning_rate": 8.884043927648579e-06, "loss": 1.3633, "step": 691 }, { "epoch": 0.11175710594315245, "grad_norm": 18.083141326904297, "learning_rate": 8.882428940568476e-06, "loss": 0.376, "step": 692 }, { "epoch": 0.1119186046511628, "grad_norm": 53.1865348815918, "learning_rate": 8.880813953488374e-06, "loss": 1.9297, "step": 693 }, { "epoch": 0.11208010335917312, "grad_norm": 20.108680725097656, "learning_rate": 8.87919896640827e-06, "loss": 0.6895, "step": 694 }, { "epoch": 0.11224160206718346, "grad_norm": 21.477779388427734, "learning_rate": 8.877583979328167e-06, "loss": 0.4595, "step": 695 }, { "epoch": 0.1124031007751938, "grad_norm": 57.02824401855469, "learning_rate": 8.875968992248062e-06, "loss": 1.8281, "step": 696 }, { "epoch": 0.11256459948320413, "grad_norm": 23.3840274810791, "learning_rate": 8.874354005167958e-06, "loss": 0.5488, "step": 697 }, { "epoch": 0.11272609819121447, "grad_norm": 82.3414077758789, "learning_rate": 8.872739018087856e-06, "loss": 1.2188, "step": 698 }, { "epoch": 0.11288759689922481, "grad_norm": 41.98219680786133, "learning_rate": 8.871124031007753e-06, "loss": 1.5742, "step": 699 }, { "epoch": 0.11304909560723514, "grad_norm": 40.88731002807617, "learning_rate": 8.869509043927649e-06, "loss": 1.1758, "step": 700 }, { "epoch": 0.11321059431524548, "grad_norm": 27.121116638183594, "learning_rate": 8.867894056847546e-06, "loss": 0.6797, "step": 701 }, { "epoch": 0.11337209302325581, "grad_norm": 38.00214385986328, "learning_rate": 8.866279069767444e-06, "loss": 1.1875, "step": 702 }, { "epoch": 0.11353359173126615, "grad_norm": 26.646303176879883, "learning_rate": 8.86466408268734e-06, "loss": 0.8867, "step": 703 }, { "epoch": 0.11369509043927649, "grad_norm": 31.83124351501465, "learning_rate": 8.863049095607235e-06, "loss": 0.9287, "step": 704 }, { "epoch": 0.11385658914728682, "grad_norm": 44.705501556396484, "learning_rate": 8.861434108527132e-06, "loss": 1.3633, "step": 705 }, { "epoch": 0.11401808785529716, "grad_norm": 32.57560348510742, "learning_rate": 8.859819121447028e-06, "loss": 0.3975, "step": 706 }, { "epoch": 0.1141795865633075, "grad_norm": 18.21328353881836, "learning_rate": 8.858204134366925e-06, "loss": 0.2622, "step": 707 }, { "epoch": 0.11434108527131782, "grad_norm": 10.766865730285645, "learning_rate": 8.856589147286823e-06, "loss": 0.2192, "step": 708 }, { "epoch": 0.11450258397932817, "grad_norm": 36.91117477416992, "learning_rate": 8.854974160206719e-06, "loss": 0.9238, "step": 709 }, { "epoch": 0.1146640826873385, "grad_norm": 11.334389686584473, "learning_rate": 8.853359173126616e-06, "loss": 0.2529, "step": 710 }, { "epoch": 0.11482558139534883, "grad_norm": 28.13345718383789, "learning_rate": 8.851744186046512e-06, "loss": 0.5981, "step": 711 }, { "epoch": 0.11498708010335917, "grad_norm": 36.84641647338867, "learning_rate": 8.850129198966409e-06, "loss": 1.2314, "step": 712 }, { "epoch": 0.11514857881136951, "grad_norm": 54.057525634765625, "learning_rate": 8.848514211886305e-06, "loss": 1.6172, "step": 713 }, { "epoch": 0.11531007751937984, "grad_norm": 37.03998565673828, "learning_rate": 8.846899224806202e-06, "loss": 0.5908, "step": 714 }, { "epoch": 0.11547157622739018, "grad_norm": 25.081220626831055, "learning_rate": 8.8452842377261e-06, "loss": 0.8027, "step": 715 }, { "epoch": 0.11563307493540052, "grad_norm": 22.298616409301758, "learning_rate": 8.843669250645995e-06, "loss": 0.6211, "step": 716 }, { "epoch": 0.11579457364341085, "grad_norm": 46.22538757324219, "learning_rate": 8.842054263565893e-06, "loss": 1.8379, "step": 717 }, { "epoch": 0.11595607235142119, "grad_norm": 50.396671295166016, "learning_rate": 8.84043927648579e-06, "loss": 1.2734, "step": 718 }, { "epoch": 0.11611757105943152, "grad_norm": 55.76021194458008, "learning_rate": 8.838824289405686e-06, "loss": 0.8315, "step": 719 }, { "epoch": 0.11627906976744186, "grad_norm": 32.62904739379883, "learning_rate": 8.837209302325582e-06, "loss": 0.5791, "step": 720 }, { "epoch": 0.1164405684754522, "grad_norm": 93.80036163330078, "learning_rate": 8.835594315245479e-06, "loss": 1.9922, "step": 721 }, { "epoch": 0.11660206718346253, "grad_norm": 77.91983795166016, "learning_rate": 8.833979328165375e-06, "loss": 1.6152, "step": 722 }, { "epoch": 0.11676356589147287, "grad_norm": 90.53694915771484, "learning_rate": 8.832364341085272e-06, "loss": 2.1211, "step": 723 }, { "epoch": 0.11692506459948321, "grad_norm": 69.48345947265625, "learning_rate": 8.83074935400517e-06, "loss": 1.9922, "step": 724 }, { "epoch": 0.11708656330749354, "grad_norm": 51.00074768066406, "learning_rate": 8.829134366925065e-06, "loss": 0.9941, "step": 725 }, { "epoch": 0.11724806201550388, "grad_norm": 45.6805419921875, "learning_rate": 8.827519379844963e-06, "loss": 0.8582, "step": 726 }, { "epoch": 0.11740956072351422, "grad_norm": 39.89131546020508, "learning_rate": 8.825904392764858e-06, "loss": 1.2412, "step": 727 }, { "epoch": 0.11757105943152454, "grad_norm": 81.89569854736328, "learning_rate": 8.824289405684756e-06, "loss": 1.7344, "step": 728 }, { "epoch": 0.11773255813953488, "grad_norm": 77.31897735595703, "learning_rate": 8.822674418604651e-06, "loss": 1.7734, "step": 729 }, { "epoch": 0.11789405684754523, "grad_norm": 67.0735855102539, "learning_rate": 8.821059431524549e-06, "loss": 1.7266, "step": 730 }, { "epoch": 0.11805555555555555, "grad_norm": 52.87324905395508, "learning_rate": 8.819444444444445e-06, "loss": 0.957, "step": 731 }, { "epoch": 0.1182170542635659, "grad_norm": 57.01503372192383, "learning_rate": 8.817829457364342e-06, "loss": 1.3887, "step": 732 }, { "epoch": 0.11837855297157623, "grad_norm": 37.3969841003418, "learning_rate": 8.81621447028424e-06, "loss": 0.667, "step": 733 }, { "epoch": 0.11854005167958656, "grad_norm": 32.19352722167969, "learning_rate": 8.814599483204135e-06, "loss": 0.9805, "step": 734 }, { "epoch": 0.1187015503875969, "grad_norm": 52.74599075317383, "learning_rate": 8.812984496124033e-06, "loss": 0.583, "step": 735 }, { "epoch": 0.11886304909560723, "grad_norm": 40.48023986816406, "learning_rate": 8.811369509043928e-06, "loss": 0.8403, "step": 736 }, { "epoch": 0.11902454780361757, "grad_norm": 27.237735748291016, "learning_rate": 8.809754521963824e-06, "loss": 0.7261, "step": 737 }, { "epoch": 0.11918604651162791, "grad_norm": 8.578248023986816, "learning_rate": 8.808139534883721e-06, "loss": 0.0477, "step": 738 }, { "epoch": 0.11934754521963824, "grad_norm": 50.25552749633789, "learning_rate": 8.806524547803619e-06, "loss": 1.4414, "step": 739 }, { "epoch": 0.11950904392764858, "grad_norm": 77.23401641845703, "learning_rate": 8.804909560723514e-06, "loss": 1.0762, "step": 740 }, { "epoch": 0.11967054263565892, "grad_norm": 47.67634201049805, "learning_rate": 8.803294573643412e-06, "loss": 0.7812, "step": 741 }, { "epoch": 0.11983204134366925, "grad_norm": 79.73262023925781, "learning_rate": 8.80167958656331e-06, "loss": 1.7178, "step": 742 }, { "epoch": 0.11999354005167959, "grad_norm": 71.42972564697266, "learning_rate": 8.800064599483205e-06, "loss": 3.3867, "step": 743 }, { "epoch": 0.12015503875968993, "grad_norm": 45.1870231628418, "learning_rate": 8.7984496124031e-06, "loss": 0.3838, "step": 744 }, { "epoch": 0.12031653746770025, "grad_norm": 79.06161499023438, "learning_rate": 8.796834625322998e-06, "loss": 0.9297, "step": 745 }, { "epoch": 0.1204780361757106, "grad_norm": 41.40586853027344, "learning_rate": 8.795219638242894e-06, "loss": 0.833, "step": 746 }, { "epoch": 0.12063953488372094, "grad_norm": 47.45283889770508, "learning_rate": 8.793604651162791e-06, "loss": 1.2129, "step": 747 }, { "epoch": 0.12080103359173126, "grad_norm": 64.88446807861328, "learning_rate": 8.791989664082689e-06, "loss": 0.9365, "step": 748 }, { "epoch": 0.1209625322997416, "grad_norm": 54.36955642700195, "learning_rate": 8.790374677002584e-06, "loss": 1.0498, "step": 749 }, { "epoch": 0.12112403100775193, "grad_norm": 67.10396575927734, "learning_rate": 8.788759689922482e-06, "loss": 0.918, "step": 750 }, { "epoch": 0.12128552971576227, "grad_norm": 30.812772750854492, "learning_rate": 8.78714470284238e-06, "loss": 0.582, "step": 751 }, { "epoch": 0.12144702842377261, "grad_norm": 36.85850524902344, "learning_rate": 8.785529715762275e-06, "loss": 0.957, "step": 752 }, { "epoch": 0.12160852713178294, "grad_norm": 25.42339324951172, "learning_rate": 8.78391472868217e-06, "loss": 0.2705, "step": 753 }, { "epoch": 0.12177002583979328, "grad_norm": 81.67982482910156, "learning_rate": 8.782299741602068e-06, "loss": 1.7051, "step": 754 }, { "epoch": 0.12193152454780362, "grad_norm": 78.97907257080078, "learning_rate": 8.780684754521964e-06, "loss": 3.2227, "step": 755 }, { "epoch": 0.12209302325581395, "grad_norm": 51.93388366699219, "learning_rate": 8.779069767441861e-06, "loss": 0.6299, "step": 756 }, { "epoch": 0.12225452196382429, "grad_norm": 39.547637939453125, "learning_rate": 8.777454780361759e-06, "loss": 1.002, "step": 757 }, { "epoch": 0.12241602067183463, "grad_norm": 82.73979187011719, "learning_rate": 8.775839793281654e-06, "loss": 2.1055, "step": 758 }, { "epoch": 0.12257751937984496, "grad_norm": 51.43307113647461, "learning_rate": 8.774224806201552e-06, "loss": 1.7344, "step": 759 }, { "epoch": 0.1227390180878553, "grad_norm": 50.14860916137695, "learning_rate": 8.772609819121447e-06, "loss": 1.25, "step": 760 }, { "epoch": 0.12290051679586564, "grad_norm": 63.143775939941406, "learning_rate": 8.770994832041345e-06, "loss": 1.7148, "step": 761 }, { "epoch": 0.12306201550387597, "grad_norm": 36.6978759765625, "learning_rate": 8.76937984496124e-06, "loss": 1.0605, "step": 762 }, { "epoch": 0.1232235142118863, "grad_norm": 46.56044387817383, "learning_rate": 8.767764857881138e-06, "loss": 1.3027, "step": 763 }, { "epoch": 0.12338501291989665, "grad_norm": 23.373727798461914, "learning_rate": 8.766149870801034e-06, "loss": 0.3789, "step": 764 }, { "epoch": 0.12354651162790697, "grad_norm": 74.12035369873047, "learning_rate": 8.764534883720931e-06, "loss": 1.4697, "step": 765 }, { "epoch": 0.12370801033591731, "grad_norm": 28.39664649963379, "learning_rate": 8.762919896640828e-06, "loss": 0.8613, "step": 766 }, { "epoch": 0.12386950904392764, "grad_norm": 35.79790496826172, "learning_rate": 8.761304909560724e-06, "loss": 0.4531, "step": 767 }, { "epoch": 0.12403100775193798, "grad_norm": 70.27489471435547, "learning_rate": 8.759689922480622e-06, "loss": 3.0769, "step": 768 }, { "epoch": 0.12419250645994832, "grad_norm": 36.79070281982422, "learning_rate": 8.758074935400517e-06, "loss": 1.2109, "step": 769 }, { "epoch": 0.12435400516795865, "grad_norm": 50.579227447509766, "learning_rate": 8.756459948320413e-06, "loss": 0.6558, "step": 770 }, { "epoch": 0.12451550387596899, "grad_norm": 32.89299774169922, "learning_rate": 8.75484496124031e-06, "loss": 0.9302, "step": 771 }, { "epoch": 0.12467700258397933, "grad_norm": 45.505775451660156, "learning_rate": 8.753229974160208e-06, "loss": 0.7637, "step": 772 }, { "epoch": 0.12483850129198966, "grad_norm": 14.72292709350586, "learning_rate": 8.751614987080104e-06, "loss": 0.1587, "step": 773 }, { "epoch": 0.125, "grad_norm": 34.698692321777344, "learning_rate": 8.750000000000001e-06, "loss": 0.73, "step": 774 }, { "epoch": 0.12516149870801033, "grad_norm": 29.625825881958008, "learning_rate": 8.748385012919898e-06, "loss": 0.5044, "step": 775 }, { "epoch": 0.12532299741602068, "grad_norm": 41.60588455200195, "learning_rate": 8.746770025839794e-06, "loss": 0.8337, "step": 776 }, { "epoch": 0.125484496124031, "grad_norm": 114.0472640991211, "learning_rate": 8.745155038759691e-06, "loss": 3.3884, "step": 777 }, { "epoch": 0.12564599483204134, "grad_norm": 24.560091018676758, "learning_rate": 8.743540051679587e-06, "loss": 0.7874, "step": 778 }, { "epoch": 0.1258074935400517, "grad_norm": 76.49118041992188, "learning_rate": 8.741925064599483e-06, "loss": 1.584, "step": 779 }, { "epoch": 0.12596899224806202, "grad_norm": 24.624483108520508, "learning_rate": 8.74031007751938e-06, "loss": 0.4678, "step": 780 }, { "epoch": 0.12613049095607234, "grad_norm": 29.1342716217041, "learning_rate": 8.738695090439278e-06, "loss": 0.7847, "step": 781 }, { "epoch": 0.1262919896640827, "grad_norm": 32.180240631103516, "learning_rate": 8.737080103359173e-06, "loss": 0.6211, "step": 782 }, { "epoch": 0.12645348837209303, "grad_norm": 73.40406799316406, "learning_rate": 8.73546511627907e-06, "loss": 1.1836, "step": 783 }, { "epoch": 0.12661498708010335, "grad_norm": 38.969512939453125, "learning_rate": 8.733850129198968e-06, "loss": 1.2539, "step": 784 }, { "epoch": 0.1267764857881137, "grad_norm": 43.46171569824219, "learning_rate": 8.732235142118864e-06, "loss": 0.5957, "step": 785 }, { "epoch": 0.12693798449612403, "grad_norm": 55.332698822021484, "learning_rate": 8.73062015503876e-06, "loss": 1.0508, "step": 786 }, { "epoch": 0.12709948320413436, "grad_norm": 63.27558517456055, "learning_rate": 8.729005167958657e-06, "loss": 1.1318, "step": 787 }, { "epoch": 0.12726098191214472, "grad_norm": 61.36309051513672, "learning_rate": 8.727390180878553e-06, "loss": 1.6875, "step": 788 }, { "epoch": 0.12742248062015504, "grad_norm": 32.93703079223633, "learning_rate": 8.72577519379845e-06, "loss": 0.7441, "step": 789 }, { "epoch": 0.12758397932816537, "grad_norm": 51.93931579589844, "learning_rate": 8.724160206718348e-06, "loss": 1.8179, "step": 790 }, { "epoch": 0.1277454780361757, "grad_norm": 35.7287483215332, "learning_rate": 8.722545219638243e-06, "loss": 0.3008, "step": 791 }, { "epoch": 0.12790697674418605, "grad_norm": 58.37565612792969, "learning_rate": 8.72093023255814e-06, "loss": 0.9785, "step": 792 }, { "epoch": 0.12806847545219638, "grad_norm": 52.21208190917969, "learning_rate": 8.719315245478036e-06, "loss": 1.3438, "step": 793 }, { "epoch": 0.1282299741602067, "grad_norm": 24.61114501953125, "learning_rate": 8.717700258397934e-06, "loss": 0.4856, "step": 794 }, { "epoch": 0.12839147286821706, "grad_norm": 68.9398422241211, "learning_rate": 8.71608527131783e-06, "loss": 1.3164, "step": 795 }, { "epoch": 0.1285529715762274, "grad_norm": 34.79998016357422, "learning_rate": 8.714470284237727e-06, "loss": 0.9951, "step": 796 }, { "epoch": 0.12871447028423771, "grad_norm": 27.788633346557617, "learning_rate": 8.712855297157623e-06, "loss": 0.8203, "step": 797 }, { "epoch": 0.12887596899224807, "grad_norm": 43.59756088256836, "learning_rate": 8.71124031007752e-06, "loss": 0.3721, "step": 798 }, { "epoch": 0.1290374677002584, "grad_norm": 167.70127868652344, "learning_rate": 8.709625322997417e-06, "loss": 1.8164, "step": 799 }, { "epoch": 0.12919896640826872, "grad_norm": 73.72618865966797, "learning_rate": 8.708010335917313e-06, "loss": 1.4609, "step": 800 }, { "epoch": 0.12936046511627908, "grad_norm": 53.75882339477539, "learning_rate": 8.70639534883721e-06, "loss": 0.8965, "step": 801 }, { "epoch": 0.1295219638242894, "grad_norm": 68.44160461425781, "learning_rate": 8.704780361757106e-06, "loss": 1.3594, "step": 802 }, { "epoch": 0.12968346253229973, "grad_norm": 22.37550926208496, "learning_rate": 8.703165374677002e-06, "loss": 0.1616, "step": 803 }, { "epoch": 0.1298449612403101, "grad_norm": 24.374011993408203, "learning_rate": 8.7015503875969e-06, "loss": 0.6758, "step": 804 }, { "epoch": 0.1300064599483204, "grad_norm": 52.35784149169922, "learning_rate": 8.699935400516797e-06, "loss": 1.5254, "step": 805 }, { "epoch": 0.13016795865633074, "grad_norm": 39.97738265991211, "learning_rate": 8.698320413436693e-06, "loss": 0.8223, "step": 806 }, { "epoch": 0.1303294573643411, "grad_norm": 42.60114669799805, "learning_rate": 8.69670542635659e-06, "loss": 0.7637, "step": 807 }, { "epoch": 0.13049095607235142, "grad_norm": 20.965957641601562, "learning_rate": 8.695090439276487e-06, "loss": 0.7031, "step": 808 }, { "epoch": 0.13065245478036175, "grad_norm": 44.42530822753906, "learning_rate": 8.693475452196383e-06, "loss": 0.5146, "step": 809 }, { "epoch": 0.1308139534883721, "grad_norm": 75.33453369140625, "learning_rate": 8.69186046511628e-06, "loss": 1.582, "step": 810 }, { "epoch": 0.13097545219638243, "grad_norm": 50.13288497924805, "learning_rate": 8.690245478036176e-06, "loss": 1.6602, "step": 811 }, { "epoch": 0.13113695090439276, "grad_norm": 7.748682498931885, "learning_rate": 8.688630490956072e-06, "loss": 0.1587, "step": 812 }, { "epoch": 0.1312984496124031, "grad_norm": 86.00690460205078, "learning_rate": 8.68701550387597e-06, "loss": 3.0703, "step": 813 }, { "epoch": 0.13145994832041344, "grad_norm": 10.89064884185791, "learning_rate": 8.685400516795867e-06, "loss": 0.0588, "step": 814 }, { "epoch": 0.13162144702842377, "grad_norm": 129.22694396972656, "learning_rate": 8.683785529715762e-06, "loss": 2.8945, "step": 815 }, { "epoch": 0.13178294573643412, "grad_norm": 38.683799743652344, "learning_rate": 8.68217054263566e-06, "loss": 0.9868, "step": 816 }, { "epoch": 0.13194444444444445, "grad_norm": 72.82988739013672, "learning_rate": 8.680555555555557e-06, "loss": 1.6484, "step": 817 }, { "epoch": 0.13210594315245477, "grad_norm": 29.498945236206055, "learning_rate": 8.678940568475453e-06, "loss": 0.623, "step": 818 }, { "epoch": 0.13226744186046513, "grad_norm": 73.11285400390625, "learning_rate": 8.677325581395349e-06, "loss": 1.7637, "step": 819 }, { "epoch": 0.13242894056847546, "grad_norm": 20.74608612060547, "learning_rate": 8.675710594315246e-06, "loss": 0.3027, "step": 820 }, { "epoch": 0.13259043927648578, "grad_norm": 7.844263553619385, "learning_rate": 8.674095607235142e-06, "loss": 0.1362, "step": 821 }, { "epoch": 0.13275193798449614, "grad_norm": 24.878570556640625, "learning_rate": 8.67248062015504e-06, "loss": 0.5625, "step": 822 }, { "epoch": 0.13291343669250646, "grad_norm": 34.17182922363281, "learning_rate": 8.670865633074937e-06, "loss": 0.2822, "step": 823 }, { "epoch": 0.1330749354005168, "grad_norm": 31.654932022094727, "learning_rate": 8.669250645994832e-06, "loss": 1.0146, "step": 824 }, { "epoch": 0.13323643410852712, "grad_norm": 24.829345703125, "learning_rate": 8.66763565891473e-06, "loss": 0.2764, "step": 825 }, { "epoch": 0.13339793281653747, "grad_norm": 32.80287170410156, "learning_rate": 8.666020671834625e-06, "loss": 0.2405, "step": 826 }, { "epoch": 0.1335594315245478, "grad_norm": 67.26468658447266, "learning_rate": 8.664405684754523e-06, "loss": 1.4434, "step": 827 }, { "epoch": 0.13372093023255813, "grad_norm": 25.82058334350586, "learning_rate": 8.662790697674419e-06, "loss": 0.4207, "step": 828 }, { "epoch": 0.13388242894056848, "grad_norm": 28.167680740356445, "learning_rate": 8.661175710594316e-06, "loss": 0.4482, "step": 829 }, { "epoch": 0.1340439276485788, "grad_norm": 110.12085723876953, "learning_rate": 8.659560723514212e-06, "loss": 4.2988, "step": 830 }, { "epoch": 0.13420542635658914, "grad_norm": 46.11558151245117, "learning_rate": 8.657945736434109e-06, "loss": 0.5723, "step": 831 }, { "epoch": 0.1343669250645995, "grad_norm": 30.405736923217773, "learning_rate": 8.656330749354006e-06, "loss": 0.5305, "step": 832 }, { "epoch": 0.13452842377260982, "grad_norm": 23.320201873779297, "learning_rate": 8.654715762273902e-06, "loss": 0.2163, "step": 833 }, { "epoch": 0.13468992248062014, "grad_norm": 30.99766731262207, "learning_rate": 8.6531007751938e-06, "loss": 1.0195, "step": 834 }, { "epoch": 0.1348514211886305, "grad_norm": 183.38009643554688, "learning_rate": 8.651485788113695e-06, "loss": 2.5625, "step": 835 }, { "epoch": 0.13501291989664083, "grad_norm": 107.14524841308594, "learning_rate": 8.649870801033593e-06, "loss": 2.8672, "step": 836 }, { "epoch": 0.13517441860465115, "grad_norm": 33.26909637451172, "learning_rate": 8.648255813953488e-06, "loss": 0.3907, "step": 837 }, { "epoch": 0.1353359173126615, "grad_norm": 59.6794319152832, "learning_rate": 8.646640826873386e-06, "loss": 1.1445, "step": 838 }, { "epoch": 0.13549741602067183, "grad_norm": 63.08851623535156, "learning_rate": 8.645025839793283e-06, "loss": 2.2578, "step": 839 }, { "epoch": 0.13565891472868216, "grad_norm": 79.2103271484375, "learning_rate": 8.643410852713179e-06, "loss": 0.873, "step": 840 }, { "epoch": 0.13582041343669252, "grad_norm": 70.1295166015625, "learning_rate": 8.641795865633076e-06, "loss": 0.75, "step": 841 }, { "epoch": 0.13598191214470284, "grad_norm": 58.27149963378906, "learning_rate": 8.640180878552972e-06, "loss": 1.4844, "step": 842 }, { "epoch": 0.13614341085271317, "grad_norm": 4.817265033721924, "learning_rate": 8.63856589147287e-06, "loss": 0.0295, "step": 843 }, { "epoch": 0.13630490956072353, "grad_norm": 115.09516143798828, "learning_rate": 8.636950904392765e-06, "loss": 2.3047, "step": 844 }, { "epoch": 0.13646640826873385, "grad_norm": 12.758674621582031, "learning_rate": 8.635335917312663e-06, "loss": 0.2002, "step": 845 }, { "epoch": 0.13662790697674418, "grad_norm": 77.86034393310547, "learning_rate": 8.633720930232558e-06, "loss": 1.6602, "step": 846 }, { "epoch": 0.13678940568475453, "grad_norm": 42.078277587890625, "learning_rate": 8.632105943152456e-06, "loss": 0.4529, "step": 847 }, { "epoch": 0.13695090439276486, "grad_norm": 30.533781051635742, "learning_rate": 8.630490956072353e-06, "loss": 0.8281, "step": 848 }, { "epoch": 0.1371124031007752, "grad_norm": 69.99637603759766, "learning_rate": 8.628875968992249e-06, "loss": 1.2676, "step": 849 }, { "epoch": 0.13727390180878554, "grad_norm": 17.370838165283203, "learning_rate": 8.627260981912146e-06, "loss": 0.291, "step": 850 }, { "epoch": 0.13743540051679587, "grad_norm": 84.30978393554688, "learning_rate": 8.625645994832042e-06, "loss": 4.5098, "step": 851 }, { "epoch": 0.1375968992248062, "grad_norm": 53.68305969238281, "learning_rate": 8.624031007751938e-06, "loss": 0.6821, "step": 852 }, { "epoch": 0.13775839793281655, "grad_norm": 6.44879150390625, "learning_rate": 8.622416020671835e-06, "loss": 0.0392, "step": 853 }, { "epoch": 0.13791989664082688, "grad_norm": 109.45848846435547, "learning_rate": 8.620801033591733e-06, "loss": 4.2969, "step": 854 }, { "epoch": 0.1380813953488372, "grad_norm": 41.61216354370117, "learning_rate": 8.619186046511628e-06, "loss": 0.5903, "step": 855 }, { "epoch": 0.13824289405684753, "grad_norm": 12.719937324523926, "learning_rate": 8.617571059431526e-06, "loss": 0.0453, "step": 856 }, { "epoch": 0.1384043927648579, "grad_norm": 59.71858215332031, "learning_rate": 8.615956072351423e-06, "loss": 1.5801, "step": 857 }, { "epoch": 0.1385658914728682, "grad_norm": 29.056201934814453, "learning_rate": 8.614341085271319e-06, "loss": 1.2109, "step": 858 }, { "epoch": 0.13872739018087854, "grad_norm": 52.36174774169922, "learning_rate": 8.612726098191214e-06, "loss": 1.5391, "step": 859 }, { "epoch": 0.1388888888888889, "grad_norm": 68.38326263427734, "learning_rate": 8.611111111111112e-06, "loss": 1.4297, "step": 860 }, { "epoch": 0.13905038759689922, "grad_norm": 25.679651260375977, "learning_rate": 8.609496124031008e-06, "loss": 0.7255, "step": 861 }, { "epoch": 0.13921188630490955, "grad_norm": 44.334476470947266, "learning_rate": 8.607881136950905e-06, "loss": 0.5117, "step": 862 }, { "epoch": 0.1393733850129199, "grad_norm": 43.3869743347168, "learning_rate": 8.606266149870802e-06, "loss": 0.7217, "step": 863 }, { "epoch": 0.13953488372093023, "grad_norm": 37.96360397338867, "learning_rate": 8.604651162790698e-06, "loss": 0.6523, "step": 864 }, { "epoch": 0.13969638242894056, "grad_norm": 40.131595611572266, "learning_rate": 8.603036175710596e-06, "loss": 0.875, "step": 865 }, { "epoch": 0.1398578811369509, "grad_norm": 26.72825813293457, "learning_rate": 8.601421188630493e-06, "loss": 0.6557, "step": 866 }, { "epoch": 0.14001937984496124, "grad_norm": 31.863630294799805, "learning_rate": 8.599806201550389e-06, "loss": 0.9033, "step": 867 }, { "epoch": 0.14018087855297157, "grad_norm": 78.94849395751953, "learning_rate": 8.598191214470284e-06, "loss": 2.8926, "step": 868 }, { "epoch": 0.14034237726098192, "grad_norm": 102.6404800415039, "learning_rate": 8.596576227390182e-06, "loss": 1.5391, "step": 869 }, { "epoch": 0.14050387596899225, "grad_norm": 47.84922409057617, "learning_rate": 8.594961240310077e-06, "loss": 0.5708, "step": 870 }, { "epoch": 0.14066537467700257, "grad_norm": 48.66322708129883, "learning_rate": 8.593346253229975e-06, "loss": 0.978, "step": 871 }, { "epoch": 0.14082687338501293, "grad_norm": 31.856632232666016, "learning_rate": 8.591731266149872e-06, "loss": 1.1333, "step": 872 }, { "epoch": 0.14098837209302326, "grad_norm": 50.07192611694336, "learning_rate": 8.590116279069768e-06, "loss": 1.2539, "step": 873 }, { "epoch": 0.14114987080103358, "grad_norm": 83.96635437011719, "learning_rate": 8.588501291989665e-06, "loss": 2.9375, "step": 874 }, { "epoch": 0.14131136950904394, "grad_norm": 90.38880920410156, "learning_rate": 8.586886304909561e-06, "loss": 3.6211, "step": 875 }, { "epoch": 0.14147286821705427, "grad_norm": 22.720197677612305, "learning_rate": 8.585271317829459e-06, "loss": 0.5583, "step": 876 }, { "epoch": 0.1416343669250646, "grad_norm": 21.485769271850586, "learning_rate": 8.583656330749354e-06, "loss": 0.4343, "step": 877 }, { "epoch": 0.14179586563307495, "grad_norm": 31.130905151367188, "learning_rate": 8.582041343669252e-06, "loss": 0.647, "step": 878 }, { "epoch": 0.14195736434108527, "grad_norm": 33.46270751953125, "learning_rate": 8.580426356589147e-06, "loss": 0.8105, "step": 879 }, { "epoch": 0.1421188630490956, "grad_norm": 16.381654739379883, "learning_rate": 8.578811369509045e-06, "loss": 0.2339, "step": 880 }, { "epoch": 0.14228036175710596, "grad_norm": 19.950204849243164, "learning_rate": 8.577196382428942e-06, "loss": 0.3989, "step": 881 }, { "epoch": 0.14244186046511628, "grad_norm": 76.01622772216797, "learning_rate": 8.575581395348838e-06, "loss": 1.7539, "step": 882 }, { "epoch": 0.1426033591731266, "grad_norm": 22.553823471069336, "learning_rate": 8.573966408268735e-06, "loss": 0.4902, "step": 883 }, { "epoch": 0.14276485788113696, "grad_norm": 81.66170501708984, "learning_rate": 8.572351421188631e-06, "loss": 1.668, "step": 884 }, { "epoch": 0.1429263565891473, "grad_norm": 61.78046417236328, "learning_rate": 8.570736434108527e-06, "loss": 2.1523, "step": 885 }, { "epoch": 0.14308785529715762, "grad_norm": 46.03974151611328, "learning_rate": 8.569121447028424e-06, "loss": 0.7363, "step": 886 }, { "epoch": 0.14324935400516794, "grad_norm": 85.55069732666016, "learning_rate": 8.567506459948322e-06, "loss": 2.5078, "step": 887 }, { "epoch": 0.1434108527131783, "grad_norm": 86.97715759277344, "learning_rate": 8.565891472868217e-06, "loss": 1.8457, "step": 888 }, { "epoch": 0.14357235142118863, "grad_norm": 25.91556167602539, "learning_rate": 8.564276485788115e-06, "loss": 0.2065, "step": 889 }, { "epoch": 0.14373385012919895, "grad_norm": 33.33233642578125, "learning_rate": 8.562661498708012e-06, "loss": 0.3716, "step": 890 }, { "epoch": 0.1438953488372093, "grad_norm": 40.501461029052734, "learning_rate": 8.561046511627908e-06, "loss": 0.8809, "step": 891 }, { "epoch": 0.14405684754521964, "grad_norm": 31.085453033447266, "learning_rate": 8.559431524547805e-06, "loss": 0.5289, "step": 892 }, { "epoch": 0.14421834625322996, "grad_norm": 36.218467712402344, "learning_rate": 8.557816537467701e-06, "loss": 0.9375, "step": 893 }, { "epoch": 0.14437984496124032, "grad_norm": 38.86641311645508, "learning_rate": 8.556201550387597e-06, "loss": 0.5762, "step": 894 }, { "epoch": 0.14454134366925064, "grad_norm": 82.25772094726562, "learning_rate": 8.554586563307494e-06, "loss": 4.1406, "step": 895 }, { "epoch": 0.14470284237726097, "grad_norm": 22.28309440612793, "learning_rate": 8.552971576227391e-06, "loss": 0.3423, "step": 896 }, { "epoch": 0.14486434108527133, "grad_norm": 62.8947639465332, "learning_rate": 8.551356589147287e-06, "loss": 1.5039, "step": 897 }, { "epoch": 0.14502583979328165, "grad_norm": 46.06399917602539, "learning_rate": 8.549741602067185e-06, "loss": 0.8877, "step": 898 }, { "epoch": 0.14518733850129198, "grad_norm": 43.268062591552734, "learning_rate": 8.548126614987082e-06, "loss": 1.123, "step": 899 }, { "epoch": 0.14534883720930233, "grad_norm": 45.368865966796875, "learning_rate": 8.546511627906978e-06, "loss": 1.373, "step": 900 }, { "epoch": 0.14551033591731266, "grad_norm": 68.20630645751953, "learning_rate": 8.544896640826873e-06, "loss": 1.1641, "step": 901 }, { "epoch": 0.145671834625323, "grad_norm": 59.25798034667969, "learning_rate": 8.54328165374677e-06, "loss": 1.2891, "step": 902 }, { "epoch": 0.14583333333333334, "grad_norm": 22.551111221313477, "learning_rate": 8.541666666666666e-06, "loss": 0.2383, "step": 903 }, { "epoch": 0.14599483204134367, "grad_norm": 73.62442779541016, "learning_rate": 8.540051679586564e-06, "loss": 1.1777, "step": 904 }, { "epoch": 0.146156330749354, "grad_norm": 28.47213363647461, "learning_rate": 8.538436692506461e-06, "loss": 0.9111, "step": 905 }, { "epoch": 0.14631782945736435, "grad_norm": 35.14766311645508, "learning_rate": 8.536821705426357e-06, "loss": 0.9883, "step": 906 }, { "epoch": 0.14647932816537468, "grad_norm": 35.13622283935547, "learning_rate": 8.535206718346254e-06, "loss": 0.918, "step": 907 }, { "epoch": 0.146640826873385, "grad_norm": 25.8448543548584, "learning_rate": 8.53359173126615e-06, "loss": 0.8047, "step": 908 }, { "epoch": 0.14680232558139536, "grad_norm": 11.456949234008789, "learning_rate": 8.531976744186048e-06, "loss": 0.2378, "step": 909 }, { "epoch": 0.1469638242894057, "grad_norm": 24.72621726989746, "learning_rate": 8.530361757105943e-06, "loss": 0.6084, "step": 910 }, { "epoch": 0.147125322997416, "grad_norm": 21.183181762695312, "learning_rate": 8.52874677002584e-06, "loss": 0.3481, "step": 911 }, { "epoch": 0.14728682170542637, "grad_norm": 99.80609893798828, "learning_rate": 8.527131782945736e-06, "loss": 2.9961, "step": 912 }, { "epoch": 0.1474483204134367, "grad_norm": 42.41456985473633, "learning_rate": 8.525516795865634e-06, "loss": 1.0039, "step": 913 }, { "epoch": 0.14760981912144702, "grad_norm": 14.95228385925293, "learning_rate": 8.523901808785531e-06, "loss": 0.1791, "step": 914 }, { "epoch": 0.14777131782945738, "grad_norm": 70.10316467285156, "learning_rate": 8.522286821705427e-06, "loss": 1.3994, "step": 915 }, { "epoch": 0.1479328165374677, "grad_norm": 44.92292022705078, "learning_rate": 8.520671834625324e-06, "loss": 1.543, "step": 916 }, { "epoch": 0.14809431524547803, "grad_norm": 80.7549819946289, "learning_rate": 8.51905684754522e-06, "loss": 2.074, "step": 917 }, { "epoch": 0.14825581395348839, "grad_norm": 19.863338470458984, "learning_rate": 8.517441860465116e-06, "loss": 0.2597, "step": 918 }, { "epoch": 0.1484173126614987, "grad_norm": 34.63666915893555, "learning_rate": 8.515826873385013e-06, "loss": 0.4482, "step": 919 }, { "epoch": 0.14857881136950904, "grad_norm": 57.111915588378906, "learning_rate": 8.51421188630491e-06, "loss": 0.8184, "step": 920 }, { "epoch": 0.14874031007751937, "grad_norm": 64.1720199584961, "learning_rate": 8.512596899224806e-06, "loss": 1.3467, "step": 921 }, { "epoch": 0.14890180878552972, "grad_norm": 68.93275451660156, "learning_rate": 8.510981912144704e-06, "loss": 1.3604, "step": 922 }, { "epoch": 0.14906330749354005, "grad_norm": 62.441261291503906, "learning_rate": 8.509366925064601e-06, "loss": 1.0371, "step": 923 }, { "epoch": 0.14922480620155038, "grad_norm": 66.2093734741211, "learning_rate": 8.507751937984497e-06, "loss": 1.5723, "step": 924 }, { "epoch": 0.14938630490956073, "grad_norm": 7.329293251037598, "learning_rate": 8.506136950904394e-06, "loss": 0.0861, "step": 925 }, { "epoch": 0.14954780361757106, "grad_norm": 121.60498809814453, "learning_rate": 8.50452196382429e-06, "loss": 2.5938, "step": 926 }, { "epoch": 0.14970930232558138, "grad_norm": 43.44725036621094, "learning_rate": 8.502906976744186e-06, "loss": 1.1016, "step": 927 }, { "epoch": 0.14987080103359174, "grad_norm": 59.67887878417969, "learning_rate": 8.501291989664083e-06, "loss": 0.6824, "step": 928 }, { "epoch": 0.15003229974160207, "grad_norm": 11.613261222839355, "learning_rate": 8.49967700258398e-06, "loss": 0.0812, "step": 929 }, { "epoch": 0.1501937984496124, "grad_norm": 75.4846420288086, "learning_rate": 8.498062015503876e-06, "loss": 1.8477, "step": 930 }, { "epoch": 0.15035529715762275, "grad_norm": 18.77307891845703, "learning_rate": 8.496447028423774e-06, "loss": 0.2612, "step": 931 }, { "epoch": 0.15051679586563307, "grad_norm": 78.94052124023438, "learning_rate": 8.494832041343671e-06, "loss": 1.4199, "step": 932 }, { "epoch": 0.1506782945736434, "grad_norm": 62.93824005126953, "learning_rate": 8.493217054263567e-06, "loss": 1.8477, "step": 933 }, { "epoch": 0.15083979328165376, "grad_norm": 163.20071411132812, "learning_rate": 8.491602067183462e-06, "loss": 4.5371, "step": 934 }, { "epoch": 0.15100129198966408, "grad_norm": 54.06083297729492, "learning_rate": 8.48998708010336e-06, "loss": 1.1445, "step": 935 }, { "epoch": 0.1511627906976744, "grad_norm": 102.92623138427734, "learning_rate": 8.488372093023256e-06, "loss": 3.0292, "step": 936 }, { "epoch": 0.15132428940568476, "grad_norm": 22.777381896972656, "learning_rate": 8.486757105943153e-06, "loss": 0.2829, "step": 937 }, { "epoch": 0.1514857881136951, "grad_norm": 103.8198471069336, "learning_rate": 8.48514211886305e-06, "loss": 1.9023, "step": 938 }, { "epoch": 0.15164728682170542, "grad_norm": 37.59333038330078, "learning_rate": 8.483527131782946e-06, "loss": 0.4785, "step": 939 }, { "epoch": 0.15180878552971577, "grad_norm": 22.4121150970459, "learning_rate": 8.481912144702843e-06, "loss": 0.3772, "step": 940 }, { "epoch": 0.1519702842377261, "grad_norm": 72.42255401611328, "learning_rate": 8.480297157622739e-06, "loss": 0.9541, "step": 941 }, { "epoch": 0.15213178294573643, "grad_norm": 93.4955062866211, "learning_rate": 8.478682170542637e-06, "loss": 1.1816, "step": 942 }, { "epoch": 0.15229328165374678, "grad_norm": 88.41499328613281, "learning_rate": 8.477067183462532e-06, "loss": 1.418, "step": 943 }, { "epoch": 0.1524547803617571, "grad_norm": 36.08567428588867, "learning_rate": 8.47545219638243e-06, "loss": 0.6025, "step": 944 }, { "epoch": 0.15261627906976744, "grad_norm": 79.35089874267578, "learning_rate": 8.473837209302325e-06, "loss": 2.0156, "step": 945 }, { "epoch": 0.1527777777777778, "grad_norm": 203.92776489257812, "learning_rate": 8.472222222222223e-06, "loss": 2.6875, "step": 946 }, { "epoch": 0.15293927648578812, "grad_norm": 109.68254852294922, "learning_rate": 8.47060723514212e-06, "loss": 0.9824, "step": 947 }, { "epoch": 0.15310077519379844, "grad_norm": 59.15589904785156, "learning_rate": 8.468992248062016e-06, "loss": 0.5181, "step": 948 }, { "epoch": 0.1532622739018088, "grad_norm": 137.33578491210938, "learning_rate": 8.467377260981913e-06, "loss": 1.8296, "step": 949 }, { "epoch": 0.15342377260981913, "grad_norm": 56.63798522949219, "learning_rate": 8.465762273901809e-06, "loss": 1.2148, "step": 950 }, { "epoch": 0.15358527131782945, "grad_norm": 173.36917114257812, "learning_rate": 8.464147286821706e-06, "loss": 4.2617, "step": 951 }, { "epoch": 0.15374677002583978, "grad_norm": 119.68536376953125, "learning_rate": 8.462532299741602e-06, "loss": 3.1074, "step": 952 }, { "epoch": 0.15390826873385013, "grad_norm": 32.524044036865234, "learning_rate": 8.4609173126615e-06, "loss": 0.7114, "step": 953 }, { "epoch": 0.15406976744186046, "grad_norm": 29.56294822692871, "learning_rate": 8.459302325581395e-06, "loss": 0.4607, "step": 954 }, { "epoch": 0.1542312661498708, "grad_norm": 37.17157745361328, "learning_rate": 8.457687338501293e-06, "loss": 0.4668, "step": 955 }, { "epoch": 0.15439276485788114, "grad_norm": 103.07756042480469, "learning_rate": 8.45607235142119e-06, "loss": 2.3359, "step": 956 }, { "epoch": 0.15455426356589147, "grad_norm": 47.8453483581543, "learning_rate": 8.454457364341086e-06, "loss": 0.8867, "step": 957 }, { "epoch": 0.1547157622739018, "grad_norm": 53.981266021728516, "learning_rate": 8.452842377260983e-06, "loss": 1.168, "step": 958 }, { "epoch": 0.15487726098191215, "grad_norm": 45.03995895385742, "learning_rate": 8.451227390180879e-06, "loss": 1.5391, "step": 959 }, { "epoch": 0.15503875968992248, "grad_norm": 39.93153762817383, "learning_rate": 8.449612403100775e-06, "loss": 0.5127, "step": 960 }, { "epoch": 0.1552002583979328, "grad_norm": 53.34650421142578, "learning_rate": 8.447997416020672e-06, "loss": 0.998, "step": 961 }, { "epoch": 0.15536175710594316, "grad_norm": 64.08987426757812, "learning_rate": 8.44638242894057e-06, "loss": 0.9648, "step": 962 }, { "epoch": 0.1555232558139535, "grad_norm": 78.42278289794922, "learning_rate": 8.444767441860465e-06, "loss": 0.6035, "step": 963 }, { "epoch": 0.15568475452196381, "grad_norm": 71.65522003173828, "learning_rate": 8.443152454780363e-06, "loss": 1.2788, "step": 964 }, { "epoch": 0.15584625322997417, "grad_norm": 72.0644302368164, "learning_rate": 8.44153746770026e-06, "loss": 2.4766, "step": 965 }, { "epoch": 0.1560077519379845, "grad_norm": 106.22745513916016, "learning_rate": 8.439922480620156e-06, "loss": 2.8579, "step": 966 }, { "epoch": 0.15616925064599482, "grad_norm": 42.12954330444336, "learning_rate": 8.438307493540051e-06, "loss": 0.6992, "step": 967 }, { "epoch": 0.15633074935400518, "grad_norm": 103.01502990722656, "learning_rate": 8.436692506459949e-06, "loss": 3.4766, "step": 968 }, { "epoch": 0.1564922480620155, "grad_norm": 44.8524055480957, "learning_rate": 8.435077519379846e-06, "loss": 0.876, "step": 969 }, { "epoch": 0.15665374677002583, "grad_norm": 26.248300552368164, "learning_rate": 8.433462532299742e-06, "loss": 0.3545, "step": 970 }, { "epoch": 0.15681524547803619, "grad_norm": 52.66579818725586, "learning_rate": 8.43184754521964e-06, "loss": 0.5742, "step": 971 }, { "epoch": 0.1569767441860465, "grad_norm": 45.19504928588867, "learning_rate": 8.430232558139537e-06, "loss": 0.7598, "step": 972 }, { "epoch": 0.15713824289405684, "grad_norm": 17.147306442260742, "learning_rate": 8.428617571059432e-06, "loss": 0.2097, "step": 973 }, { "epoch": 0.1572997416020672, "grad_norm": 40.7651252746582, "learning_rate": 8.42700258397933e-06, "loss": 0.6895, "step": 974 }, { "epoch": 0.15746124031007752, "grad_norm": 37.7332649230957, "learning_rate": 8.425387596899226e-06, "loss": 1.0645, "step": 975 }, { "epoch": 0.15762273901808785, "grad_norm": 66.53743743896484, "learning_rate": 8.423772609819121e-06, "loss": 0.8945, "step": 976 }, { "epoch": 0.1577842377260982, "grad_norm": 96.39051055908203, "learning_rate": 8.422157622739019e-06, "loss": 2.0547, "step": 977 }, { "epoch": 0.15794573643410853, "grad_norm": 43.71870422363281, "learning_rate": 8.420542635658916e-06, "loss": 0.6934, "step": 978 }, { "epoch": 0.15810723514211886, "grad_norm": 26.401655197143555, "learning_rate": 8.418927648578812e-06, "loss": 0.3613, "step": 979 }, { "epoch": 0.1582687338501292, "grad_norm": 45.2999267578125, "learning_rate": 8.41731266149871e-06, "loss": 0.5398, "step": 980 }, { "epoch": 0.15843023255813954, "grad_norm": 117.33164978027344, "learning_rate": 8.415697674418607e-06, "loss": 3.2188, "step": 981 }, { "epoch": 0.15859173126614987, "grad_norm": 112.82902526855469, "learning_rate": 8.414082687338502e-06, "loss": 1.9844, "step": 982 }, { "epoch": 0.15875322997416022, "grad_norm": 40.30073165893555, "learning_rate": 8.412467700258398e-06, "loss": 0.4355, "step": 983 }, { "epoch": 0.15891472868217055, "grad_norm": 31.270366668701172, "learning_rate": 8.410852713178295e-06, "loss": 0.3438, "step": 984 }, { "epoch": 0.15907622739018087, "grad_norm": 12.926656723022461, "learning_rate": 8.409237726098191e-06, "loss": 0.0549, "step": 985 }, { "epoch": 0.1592377260981912, "grad_norm": 26.275423049926758, "learning_rate": 8.407622739018089e-06, "loss": 0.3516, "step": 986 }, { "epoch": 0.15939922480620156, "grad_norm": 64.01627349853516, "learning_rate": 8.406007751937986e-06, "loss": 2.2891, "step": 987 }, { "epoch": 0.15956072351421188, "grad_norm": 29.18509864807129, "learning_rate": 8.404392764857882e-06, "loss": 0.2559, "step": 988 }, { "epoch": 0.1597222222222222, "grad_norm": 79.02012634277344, "learning_rate": 8.402777777777779e-06, "loss": 1.6309, "step": 989 }, { "epoch": 0.15988372093023256, "grad_norm": 48.89326095581055, "learning_rate": 8.401162790697675e-06, "loss": 1.3887, "step": 990 }, { "epoch": 0.1600452196382429, "grad_norm": 28.176010131835938, "learning_rate": 8.399547803617572e-06, "loss": 0.3564, "step": 991 }, { "epoch": 0.16020671834625322, "grad_norm": 59.33149337768555, "learning_rate": 8.397932816537468e-06, "loss": 0.5581, "step": 992 }, { "epoch": 0.16036821705426357, "grad_norm": 60.41129684448242, "learning_rate": 8.396317829457365e-06, "loss": 0.5576, "step": 993 }, { "epoch": 0.1605297157622739, "grad_norm": 64.7509765625, "learning_rate": 8.394702842377261e-06, "loss": 0.6426, "step": 994 }, { "epoch": 0.16069121447028423, "grad_norm": 100.13131713867188, "learning_rate": 8.393087855297158e-06, "loss": 2.2676, "step": 995 }, { "epoch": 0.16085271317829458, "grad_norm": 84.02515411376953, "learning_rate": 8.391472868217056e-06, "loss": 1.2644, "step": 996 }, { "epoch": 0.1610142118863049, "grad_norm": 94.68416595458984, "learning_rate": 8.389857881136952e-06, "loss": 1.5625, "step": 997 }, { "epoch": 0.16117571059431524, "grad_norm": 57.60042190551758, "learning_rate": 8.388242894056849e-06, "loss": 0.7471, "step": 998 }, { "epoch": 0.1613372093023256, "grad_norm": 26.174930572509766, "learning_rate": 8.386627906976745e-06, "loss": 0.2186, "step": 999 }, { "epoch": 0.16149870801033592, "grad_norm": 81.27520751953125, "learning_rate": 8.38501291989664e-06, "loss": 1.4062, "step": 1000 }, { "epoch": 0.16166020671834624, "grad_norm": 145.95660400390625, "learning_rate": 8.383397932816538e-06, "loss": 3.5391, "step": 1001 }, { "epoch": 0.1618217054263566, "grad_norm": 96.8889389038086, "learning_rate": 8.381782945736435e-06, "loss": 1.4336, "step": 1002 }, { "epoch": 0.16198320413436693, "grad_norm": 96.52655029296875, "learning_rate": 8.380167958656331e-06, "loss": 3.4072, "step": 1003 }, { "epoch": 0.16214470284237725, "grad_norm": 84.01363372802734, "learning_rate": 8.378552971576228e-06, "loss": 1.7607, "step": 1004 }, { "epoch": 0.1623062015503876, "grad_norm": 21.17816925048828, "learning_rate": 8.376937984496126e-06, "loss": 0.1306, "step": 1005 }, { "epoch": 0.16246770025839793, "grad_norm": 127.9647445678711, "learning_rate": 8.375322997416021e-06, "loss": 2.8301, "step": 1006 }, { "epoch": 0.16262919896640826, "grad_norm": 11.973234176635742, "learning_rate": 8.373708010335919e-06, "loss": 0.1322, "step": 1007 }, { "epoch": 0.16279069767441862, "grad_norm": 64.7510986328125, "learning_rate": 8.372093023255815e-06, "loss": 1.0781, "step": 1008 }, { "epoch": 0.16295219638242894, "grad_norm": 126.1361312866211, "learning_rate": 8.37047803617571e-06, "loss": 2.1836, "step": 1009 }, { "epoch": 0.16311369509043927, "grad_norm": 48.303260803222656, "learning_rate": 8.368863049095608e-06, "loss": 0.6992, "step": 1010 }, { "epoch": 0.16327519379844962, "grad_norm": 90.31510925292969, "learning_rate": 8.367248062015505e-06, "loss": 1.2646, "step": 1011 }, { "epoch": 0.16343669250645995, "grad_norm": 57.295021057128906, "learning_rate": 8.3656330749354e-06, "loss": 1.6523, "step": 1012 }, { "epoch": 0.16359819121447028, "grad_norm": 37.875709533691406, "learning_rate": 8.364018087855298e-06, "loss": 0.416, "step": 1013 }, { "epoch": 0.16375968992248063, "grad_norm": 75.42929077148438, "learning_rate": 8.362403100775196e-06, "loss": 1.3604, "step": 1014 }, { "epoch": 0.16392118863049096, "grad_norm": 59.81015396118164, "learning_rate": 8.360788113695091e-06, "loss": 1.5469, "step": 1015 }, { "epoch": 0.1640826873385013, "grad_norm": 56.49078369140625, "learning_rate": 8.359173126614987e-06, "loss": 1.0684, "step": 1016 }, { "epoch": 0.16424418604651161, "grad_norm": 100.14282989501953, "learning_rate": 8.357558139534884e-06, "loss": 1.6361, "step": 1017 }, { "epoch": 0.16440568475452197, "grad_norm": 74.19847106933594, "learning_rate": 8.35594315245478e-06, "loss": 1.4531, "step": 1018 }, { "epoch": 0.1645671834625323, "grad_norm": 20.422298431396484, "learning_rate": 8.354328165374678e-06, "loss": 0.3545, "step": 1019 }, { "epoch": 0.16472868217054262, "grad_norm": 36.45651626586914, "learning_rate": 8.352713178294575e-06, "loss": 0.6484, "step": 1020 }, { "epoch": 0.16489018087855298, "grad_norm": 100.89750671386719, "learning_rate": 8.35109819121447e-06, "loss": 0.9517, "step": 1021 }, { "epoch": 0.1650516795865633, "grad_norm": 50.36946105957031, "learning_rate": 8.349483204134368e-06, "loss": 0.8906, "step": 1022 }, { "epoch": 0.16521317829457363, "grad_norm": 69.49468231201172, "learning_rate": 8.347868217054264e-06, "loss": 0.7002, "step": 1023 }, { "epoch": 0.165374677002584, "grad_norm": 30.106094360351562, "learning_rate": 8.346253229974161e-06, "loss": 0.4014, "step": 1024 }, { "epoch": 0.1655361757105943, "grad_norm": 63.210819244384766, "learning_rate": 8.344638242894057e-06, "loss": 0.8408, "step": 1025 }, { "epoch": 0.16569767441860464, "grad_norm": 64.72061157226562, "learning_rate": 8.343023255813954e-06, "loss": 1.1934, "step": 1026 }, { "epoch": 0.165859173126615, "grad_norm": 21.038755416870117, "learning_rate": 8.34140826873385e-06, "loss": 0.2681, "step": 1027 }, { "epoch": 0.16602067183462532, "grad_norm": 48.31766128540039, "learning_rate": 8.339793281653747e-06, "loss": 1.0625, "step": 1028 }, { "epoch": 0.16618217054263565, "grad_norm": 80.5725326538086, "learning_rate": 8.338178294573645e-06, "loss": 1.4629, "step": 1029 }, { "epoch": 0.166343669250646, "grad_norm": 18.819332122802734, "learning_rate": 8.33656330749354e-06, "loss": 0.1354, "step": 1030 }, { "epoch": 0.16650516795865633, "grad_norm": 63.42562484741211, "learning_rate": 8.334948320413438e-06, "loss": 1.5332, "step": 1031 }, { "epoch": 0.16666666666666666, "grad_norm": 99.59076690673828, "learning_rate": 8.333333333333334e-06, "loss": 0.9846, "step": 1032 }, { "epoch": 0.166828165374677, "grad_norm": 46.907493591308594, "learning_rate": 8.33171834625323e-06, "loss": 0.3506, "step": 1033 }, { "epoch": 0.16698966408268734, "grad_norm": 76.60377502441406, "learning_rate": 8.330103359173127e-06, "loss": 1.3545, "step": 1034 }, { "epoch": 0.16715116279069767, "grad_norm": 71.3504867553711, "learning_rate": 8.328488372093024e-06, "loss": 1.3359, "step": 1035 }, { "epoch": 0.16731266149870802, "grad_norm": 55.619136810302734, "learning_rate": 8.32687338501292e-06, "loss": 0.4141, "step": 1036 }, { "epoch": 0.16747416020671835, "grad_norm": 58.61499786376953, "learning_rate": 8.325258397932817e-06, "loss": 1.1016, "step": 1037 }, { "epoch": 0.16763565891472867, "grad_norm": 88.58052825927734, "learning_rate": 8.323643410852715e-06, "loss": 1.1562, "step": 1038 }, { "epoch": 0.16779715762273903, "grad_norm": 80.89569854736328, "learning_rate": 8.32202842377261e-06, "loss": 1.9375, "step": 1039 }, { "epoch": 0.16795865633074936, "grad_norm": 131.7953643798828, "learning_rate": 8.320413436692508e-06, "loss": 1.7188, "step": 1040 }, { "epoch": 0.16812015503875968, "grad_norm": 150.26368713378906, "learning_rate": 8.318798449612404e-06, "loss": 3.0703, "step": 1041 }, { "epoch": 0.16828165374677004, "grad_norm": 18.878639221191406, "learning_rate": 8.3171834625323e-06, "loss": 0.2845, "step": 1042 }, { "epoch": 0.16844315245478036, "grad_norm": 95.6979751586914, "learning_rate": 8.315568475452197e-06, "loss": 1.3516, "step": 1043 }, { "epoch": 0.1686046511627907, "grad_norm": 37.22861099243164, "learning_rate": 8.313953488372094e-06, "loss": 0.8477, "step": 1044 }, { "epoch": 0.16876614987080105, "grad_norm": 56.40743637084961, "learning_rate": 8.31233850129199e-06, "loss": 0.2434, "step": 1045 }, { "epoch": 0.16892764857881137, "grad_norm": 44.17649459838867, "learning_rate": 8.310723514211887e-06, "loss": 0.3008, "step": 1046 }, { "epoch": 0.1690891472868217, "grad_norm": 88.33065795898438, "learning_rate": 8.309108527131785e-06, "loss": 1.3242, "step": 1047 }, { "epoch": 0.16925064599483206, "grad_norm": 90.5719985961914, "learning_rate": 8.30749354005168e-06, "loss": 1.625, "step": 1048 }, { "epoch": 0.16941214470284238, "grad_norm": 17.777629852294922, "learning_rate": 8.305878552971576e-06, "loss": 0.0984, "step": 1049 }, { "epoch": 0.1695736434108527, "grad_norm": 57.507503509521484, "learning_rate": 8.304263565891474e-06, "loss": 1.4062, "step": 1050 }, { "epoch": 0.16973514211886304, "grad_norm": 31.74405288696289, "learning_rate": 8.30264857881137e-06, "loss": 0.6309, "step": 1051 }, { "epoch": 0.1698966408268734, "grad_norm": 51.454471588134766, "learning_rate": 8.301033591731267e-06, "loss": 0.8994, "step": 1052 }, { "epoch": 0.17005813953488372, "grad_norm": 71.44012451171875, "learning_rate": 8.299418604651164e-06, "loss": 2.6797, "step": 1053 }, { "epoch": 0.17021963824289404, "grad_norm": 84.42403411865234, "learning_rate": 8.29780361757106e-06, "loss": 1.2871, "step": 1054 }, { "epoch": 0.1703811369509044, "grad_norm": 26.714061737060547, "learning_rate": 8.296188630490957e-06, "loss": 0.3127, "step": 1055 }, { "epoch": 0.17054263565891473, "grad_norm": 49.88245391845703, "learning_rate": 8.294573643410853e-06, "loss": 0.7754, "step": 1056 }, { "epoch": 0.17070413436692505, "grad_norm": 25.611328125, "learning_rate": 8.29295865633075e-06, "loss": 0.2227, "step": 1057 }, { "epoch": 0.1708656330749354, "grad_norm": 90.32389068603516, "learning_rate": 8.291343669250646e-06, "loss": 3.3525, "step": 1058 }, { "epoch": 0.17102713178294573, "grad_norm": 148.87698364257812, "learning_rate": 8.289728682170543e-06, "loss": 2.4219, "step": 1059 }, { "epoch": 0.17118863049095606, "grad_norm": 67.22480773925781, "learning_rate": 8.288113695090439e-06, "loss": 1.3555, "step": 1060 }, { "epoch": 0.17135012919896642, "grad_norm": 37.7390022277832, "learning_rate": 8.286498708010337e-06, "loss": 0.5498, "step": 1061 }, { "epoch": 0.17151162790697674, "grad_norm": 30.8333683013916, "learning_rate": 8.284883720930234e-06, "loss": 0.3594, "step": 1062 }, { "epoch": 0.17167312661498707, "grad_norm": 25.481460571289062, "learning_rate": 8.28326873385013e-06, "loss": 0.2283, "step": 1063 }, { "epoch": 0.17183462532299743, "grad_norm": 19.315372467041016, "learning_rate": 8.281653746770027e-06, "loss": 0.3811, "step": 1064 }, { "epoch": 0.17199612403100775, "grad_norm": 125.23221588134766, "learning_rate": 8.280038759689923e-06, "loss": 1.6133, "step": 1065 }, { "epoch": 0.17215762273901808, "grad_norm": 68.45337677001953, "learning_rate": 8.27842377260982e-06, "loss": 0.7461, "step": 1066 }, { "epoch": 0.17231912144702843, "grad_norm": 28.03350257873535, "learning_rate": 8.276808785529716e-06, "loss": 0.1724, "step": 1067 }, { "epoch": 0.17248062015503876, "grad_norm": 83.58008575439453, "learning_rate": 8.275193798449613e-06, "loss": 1.3984, "step": 1068 }, { "epoch": 0.1726421188630491, "grad_norm": 63.83644104003906, "learning_rate": 8.273578811369509e-06, "loss": 0.541, "step": 1069 }, { "epoch": 0.17280361757105944, "grad_norm": 69.45263671875, "learning_rate": 8.271963824289406e-06, "loss": 1.3398, "step": 1070 }, { "epoch": 0.17296511627906977, "grad_norm": 92.15190124511719, "learning_rate": 8.270348837209304e-06, "loss": 0.7393, "step": 1071 }, { "epoch": 0.1731266149870801, "grad_norm": 134.89024353027344, "learning_rate": 8.2687338501292e-06, "loss": 3.8999, "step": 1072 }, { "epoch": 0.17328811369509045, "grad_norm": 29.491758346557617, "learning_rate": 8.267118863049097e-06, "loss": 0.8047, "step": 1073 }, { "epoch": 0.17344961240310078, "grad_norm": 27.143888473510742, "learning_rate": 8.265503875968993e-06, "loss": 0.1929, "step": 1074 }, { "epoch": 0.1736111111111111, "grad_norm": 54.87688446044922, "learning_rate": 8.263888888888888e-06, "loss": 0.751, "step": 1075 }, { "epoch": 0.17377260981912146, "grad_norm": 6.311633110046387, "learning_rate": 8.262273901808786e-06, "loss": 0.0578, "step": 1076 }, { "epoch": 0.1739341085271318, "grad_norm": 39.64244079589844, "learning_rate": 8.260658914728683e-06, "loss": 0.9873, "step": 1077 }, { "epoch": 0.1740956072351421, "grad_norm": 45.246273040771484, "learning_rate": 8.259043927648579e-06, "loss": 0.7612, "step": 1078 }, { "epoch": 0.17425710594315247, "grad_norm": 80.18894958496094, "learning_rate": 8.257428940568476e-06, "loss": 0.873, "step": 1079 }, { "epoch": 0.1744186046511628, "grad_norm": 31.75733757019043, "learning_rate": 8.255813953488374e-06, "loss": 0.1436, "step": 1080 }, { "epoch": 0.17458010335917312, "grad_norm": 63.6485595703125, "learning_rate": 8.25419896640827e-06, "loss": 0.9531, "step": 1081 }, { "epoch": 0.17474160206718345, "grad_norm": 36.78181076049805, "learning_rate": 8.252583979328165e-06, "loss": 0.6533, "step": 1082 }, { "epoch": 0.1749031007751938, "grad_norm": 147.215576171875, "learning_rate": 8.250968992248063e-06, "loss": 5.9219, "step": 1083 }, { "epoch": 0.17506459948320413, "grad_norm": 39.71880340576172, "learning_rate": 8.249354005167958e-06, "loss": 0.7788, "step": 1084 }, { "epoch": 0.17522609819121446, "grad_norm": 29.733613967895508, "learning_rate": 8.247739018087856e-06, "loss": 0.2686, "step": 1085 }, { "epoch": 0.1753875968992248, "grad_norm": 69.01448059082031, "learning_rate": 8.246124031007753e-06, "loss": 1.313, "step": 1086 }, { "epoch": 0.17554909560723514, "grad_norm": 83.42897033691406, "learning_rate": 8.244509043927649e-06, "loss": 3.0156, "step": 1087 }, { "epoch": 0.17571059431524547, "grad_norm": 61.033599853515625, "learning_rate": 8.242894056847546e-06, "loss": 1.543, "step": 1088 }, { "epoch": 0.17587209302325582, "grad_norm": 33.62605285644531, "learning_rate": 8.241279069767444e-06, "loss": 0.3789, "step": 1089 }, { "epoch": 0.17603359173126615, "grad_norm": 80.73255157470703, "learning_rate": 8.23966408268734e-06, "loss": 1.1602, "step": 1090 }, { "epoch": 0.17619509043927647, "grad_norm": 118.55643463134766, "learning_rate": 8.238049095607235e-06, "loss": 2.0, "step": 1091 }, { "epoch": 0.17635658914728683, "grad_norm": 64.01731872558594, "learning_rate": 8.236434108527132e-06, "loss": 0.6133, "step": 1092 }, { "epoch": 0.17651808785529716, "grad_norm": 135.8675994873047, "learning_rate": 8.234819121447028e-06, "loss": 4.75, "step": 1093 }, { "epoch": 0.17667958656330748, "grad_norm": 60.96310043334961, "learning_rate": 8.233204134366926e-06, "loss": 0.8887, "step": 1094 }, { "epoch": 0.17684108527131784, "grad_norm": 114.74048614501953, "learning_rate": 8.231589147286823e-06, "loss": 2.7344, "step": 1095 }, { "epoch": 0.17700258397932817, "grad_norm": 10.496928215026855, "learning_rate": 8.229974160206719e-06, "loss": 0.0836, "step": 1096 }, { "epoch": 0.1771640826873385, "grad_norm": 31.24321937561035, "learning_rate": 8.228359173126616e-06, "loss": 0.5752, "step": 1097 }, { "epoch": 0.17732558139534885, "grad_norm": 38.07929992675781, "learning_rate": 8.226744186046512e-06, "loss": 0.6162, "step": 1098 }, { "epoch": 0.17748708010335917, "grad_norm": 16.67603302001953, "learning_rate": 8.22512919896641e-06, "loss": 0.0679, "step": 1099 }, { "epoch": 0.1776485788113695, "grad_norm": 56.1229133605957, "learning_rate": 8.223514211886305e-06, "loss": 0.9102, "step": 1100 }, { "epoch": 0.17781007751937986, "grad_norm": 19.870948791503906, "learning_rate": 8.221899224806202e-06, "loss": 0.3025, "step": 1101 }, { "epoch": 0.17797157622739018, "grad_norm": 33.743858337402344, "learning_rate": 8.2202842377261e-06, "loss": 0.3613, "step": 1102 }, { "epoch": 0.1781330749354005, "grad_norm": 53.37968444824219, "learning_rate": 8.218669250645995e-06, "loss": 1.002, "step": 1103 }, { "epoch": 0.17829457364341086, "grad_norm": 84.48668670654297, "learning_rate": 8.217054263565893e-06, "loss": 2.0273, "step": 1104 }, { "epoch": 0.1784560723514212, "grad_norm": 46.83168029785156, "learning_rate": 8.215439276485789e-06, "loss": 1.2637, "step": 1105 }, { "epoch": 0.17861757105943152, "grad_norm": 45.08584976196289, "learning_rate": 8.213824289405686e-06, "loss": 0.9307, "step": 1106 }, { "epoch": 0.17877906976744187, "grad_norm": 52.67623519897461, "learning_rate": 8.212209302325582e-06, "loss": 1.6929, "step": 1107 }, { "epoch": 0.1789405684754522, "grad_norm": 22.88245964050293, "learning_rate": 8.210594315245479e-06, "loss": 0.3198, "step": 1108 }, { "epoch": 0.17910206718346253, "grad_norm": 39.21488571166992, "learning_rate": 8.208979328165375e-06, "loss": 0.8315, "step": 1109 }, { "epoch": 0.17926356589147288, "grad_norm": 21.501873016357422, "learning_rate": 8.207364341085272e-06, "loss": 0.2881, "step": 1110 }, { "epoch": 0.1794250645994832, "grad_norm": 24.468610763549805, "learning_rate": 8.20574935400517e-06, "loss": 0.1401, "step": 1111 }, { "epoch": 0.17958656330749354, "grad_norm": 43.0170783996582, "learning_rate": 8.204134366925065e-06, "loss": 1.1748, "step": 1112 }, { "epoch": 0.17974806201550386, "grad_norm": 62.999961853027344, "learning_rate": 8.202519379844963e-06, "loss": 1.1064, "step": 1113 }, { "epoch": 0.17990956072351422, "grad_norm": 153.3686981201172, "learning_rate": 8.200904392764858e-06, "loss": 1.9141, "step": 1114 }, { "epoch": 0.18007105943152454, "grad_norm": 46.967777252197266, "learning_rate": 8.199289405684754e-06, "loss": 1.1523, "step": 1115 }, { "epoch": 0.18023255813953487, "grad_norm": 25.00326919555664, "learning_rate": 8.197674418604652e-06, "loss": 0.5889, "step": 1116 }, { "epoch": 0.18039405684754523, "grad_norm": 21.602590560913086, "learning_rate": 8.196059431524549e-06, "loss": 0.1323, "step": 1117 }, { "epoch": 0.18055555555555555, "grad_norm": 32.24546432495117, "learning_rate": 8.194444444444445e-06, "loss": 0.75, "step": 1118 }, { "epoch": 0.18071705426356588, "grad_norm": 40.122867584228516, "learning_rate": 8.192829457364342e-06, "loss": 1.0983, "step": 1119 }, { "epoch": 0.18087855297157623, "grad_norm": 78.9273452758789, "learning_rate": 8.19121447028424e-06, "loss": 2.9668, "step": 1120 }, { "epoch": 0.18104005167958656, "grad_norm": 27.646282196044922, "learning_rate": 8.189599483204135e-06, "loss": 1.064, "step": 1121 }, { "epoch": 0.1812015503875969, "grad_norm": 55.786617279052734, "learning_rate": 8.187984496124033e-06, "loss": 1.1758, "step": 1122 }, { "epoch": 0.18136304909560724, "grad_norm": 63.80012130737305, "learning_rate": 8.186369509043928e-06, "loss": 1.5859, "step": 1123 }, { "epoch": 0.18152454780361757, "grad_norm": 23.720266342163086, "learning_rate": 8.184754521963824e-06, "loss": 0.324, "step": 1124 }, { "epoch": 0.1816860465116279, "grad_norm": 24.35289764404297, "learning_rate": 8.183139534883721e-06, "loss": 0.7949, "step": 1125 }, { "epoch": 0.18184754521963825, "grad_norm": 92.19154357910156, "learning_rate": 8.181524547803619e-06, "loss": 1.7891, "step": 1126 }, { "epoch": 0.18200904392764858, "grad_norm": 60.93740463256836, "learning_rate": 8.179909560723515e-06, "loss": 1.0684, "step": 1127 }, { "epoch": 0.1821705426356589, "grad_norm": 156.5941162109375, "learning_rate": 8.178294573643412e-06, "loss": 3.9375, "step": 1128 }, { "epoch": 0.18233204134366926, "grad_norm": 123.10037994384766, "learning_rate": 8.17667958656331e-06, "loss": 2.6094, "step": 1129 }, { "epoch": 0.1824935400516796, "grad_norm": 45.46022415161133, "learning_rate": 8.175064599483205e-06, "loss": 0.707, "step": 1130 }, { "epoch": 0.1826550387596899, "grad_norm": 61.49689865112305, "learning_rate": 8.1734496124031e-06, "loss": 0.894, "step": 1131 }, { "epoch": 0.18281653746770027, "grad_norm": 58.076560974121094, "learning_rate": 8.171834625322998e-06, "loss": 1.1587, "step": 1132 }, { "epoch": 0.1829780361757106, "grad_norm": 63.141475677490234, "learning_rate": 8.170219638242894e-06, "loss": 1.0283, "step": 1133 }, { "epoch": 0.18313953488372092, "grad_norm": 50.88545608520508, "learning_rate": 8.168604651162791e-06, "loss": 0.624, "step": 1134 }, { "epoch": 0.18330103359173128, "grad_norm": 61.50099182128906, "learning_rate": 8.166989664082689e-06, "loss": 0.557, "step": 1135 }, { "epoch": 0.1834625322997416, "grad_norm": 35.634605407714844, "learning_rate": 8.165374677002584e-06, "loss": 0.967, "step": 1136 }, { "epoch": 0.18362403100775193, "grad_norm": 30.759782791137695, "learning_rate": 8.163759689922482e-06, "loss": 0.4424, "step": 1137 }, { "epoch": 0.18378552971576229, "grad_norm": 45.813636779785156, "learning_rate": 8.162144702842378e-06, "loss": 1.5, "step": 1138 }, { "epoch": 0.1839470284237726, "grad_norm": 41.94113540649414, "learning_rate": 8.160529715762275e-06, "loss": 1.1727, "step": 1139 }, { "epoch": 0.18410852713178294, "grad_norm": 22.321548461914062, "learning_rate": 8.15891472868217e-06, "loss": 0.5083, "step": 1140 }, { "epoch": 0.1842700258397933, "grad_norm": 61.083892822265625, "learning_rate": 8.157299741602068e-06, "loss": 1.8281, "step": 1141 }, { "epoch": 0.18443152454780362, "grad_norm": 33.52189636230469, "learning_rate": 8.155684754521964e-06, "loss": 0.4512, "step": 1142 }, { "epoch": 0.18459302325581395, "grad_norm": 16.621994018554688, "learning_rate": 8.154069767441861e-06, "loss": 0.4741, "step": 1143 }, { "epoch": 0.1847545219638243, "grad_norm": 40.99005889892578, "learning_rate": 8.152454780361759e-06, "loss": 0.3965, "step": 1144 }, { "epoch": 0.18491602067183463, "grad_norm": 81.02011108398438, "learning_rate": 8.150839793281654e-06, "loss": 2.1875, "step": 1145 }, { "epoch": 0.18507751937984496, "grad_norm": 37.229217529296875, "learning_rate": 8.149224806201552e-06, "loss": 0.2559, "step": 1146 }, { "epoch": 0.18523901808785528, "grad_norm": 32.0142822265625, "learning_rate": 8.147609819121447e-06, "loss": 0.207, "step": 1147 }, { "epoch": 0.18540051679586564, "grad_norm": 120.59060668945312, "learning_rate": 8.145994832041343e-06, "loss": 3.584, "step": 1148 }, { "epoch": 0.18556201550387597, "grad_norm": 68.94578552246094, "learning_rate": 8.14437984496124e-06, "loss": 1.5039, "step": 1149 }, { "epoch": 0.1857235142118863, "grad_norm": 39.66535568237305, "learning_rate": 8.142764857881138e-06, "loss": 0.8057, "step": 1150 }, { "epoch": 0.18588501291989665, "grad_norm": 71.03946685791016, "learning_rate": 8.141149870801034e-06, "loss": 1.7148, "step": 1151 }, { "epoch": 0.18604651162790697, "grad_norm": 106.71699523925781, "learning_rate": 8.139534883720931e-06, "loss": 1.7656, "step": 1152 }, { "epoch": 0.1862080103359173, "grad_norm": 66.77527618408203, "learning_rate": 8.137919896640829e-06, "loss": 1.4805, "step": 1153 }, { "epoch": 0.18636950904392766, "grad_norm": 22.55129623413086, "learning_rate": 8.136304909560724e-06, "loss": 0.5991, "step": 1154 }, { "epoch": 0.18653100775193798, "grad_norm": 90.7267074584961, "learning_rate": 8.134689922480622e-06, "loss": 2.6436, "step": 1155 }, { "epoch": 0.1866925064599483, "grad_norm": 60.05411911010742, "learning_rate": 8.133074935400517e-06, "loss": 0.9219, "step": 1156 }, { "epoch": 0.18685400516795866, "grad_norm": 14.998563766479492, "learning_rate": 8.131459948320413e-06, "loss": 0.1958, "step": 1157 }, { "epoch": 0.187015503875969, "grad_norm": 31.981733322143555, "learning_rate": 8.12984496124031e-06, "loss": 0.6421, "step": 1158 }, { "epoch": 0.18717700258397932, "grad_norm": 95.19621276855469, "learning_rate": 8.128229974160208e-06, "loss": 2.6406, "step": 1159 }, { "epoch": 0.18733850129198967, "grad_norm": 130.15013122558594, "learning_rate": 8.126614987080104e-06, "loss": 3.6562, "step": 1160 }, { "epoch": 0.1875, "grad_norm": 66.19631958007812, "learning_rate": 8.125000000000001e-06, "loss": 1.3711, "step": 1161 }, { "epoch": 0.18766149870801033, "grad_norm": 62.52029037475586, "learning_rate": 8.123385012919898e-06, "loss": 2.5352, "step": 1162 }, { "epoch": 0.18782299741602068, "grad_norm": 83.79479217529297, "learning_rate": 8.121770025839794e-06, "loss": 1.8982, "step": 1163 }, { "epoch": 0.187984496124031, "grad_norm": 80.17743682861328, "learning_rate": 8.12015503875969e-06, "loss": 0.7986, "step": 1164 }, { "epoch": 0.18814599483204134, "grad_norm": 43.12319564819336, "learning_rate": 8.118540051679587e-06, "loss": 0.9697, "step": 1165 }, { "epoch": 0.1883074935400517, "grad_norm": 35.84133529663086, "learning_rate": 8.116925064599483e-06, "loss": 0.9805, "step": 1166 }, { "epoch": 0.18846899224806202, "grad_norm": 21.093555450439453, "learning_rate": 8.11531007751938e-06, "loss": 0.3765, "step": 1167 }, { "epoch": 0.18863049095607234, "grad_norm": 64.2374038696289, "learning_rate": 8.113695090439278e-06, "loss": 1.2197, "step": 1168 }, { "epoch": 0.1887919896640827, "grad_norm": 30.25702476501465, "learning_rate": 8.112080103359173e-06, "loss": 1.1836, "step": 1169 }, { "epoch": 0.18895348837209303, "grad_norm": 66.5123519897461, "learning_rate": 8.110465116279071e-06, "loss": 2.1172, "step": 1170 }, { "epoch": 0.18911498708010335, "grad_norm": 56.58976745605469, "learning_rate": 8.108850129198967e-06, "loss": 1.1719, "step": 1171 }, { "epoch": 0.1892764857881137, "grad_norm": 40.82367706298828, "learning_rate": 8.107235142118864e-06, "loss": 0.8633, "step": 1172 }, { "epoch": 0.18943798449612403, "grad_norm": 66.20960235595703, "learning_rate": 8.10562015503876e-06, "loss": 1.3008, "step": 1173 }, { "epoch": 0.18959948320413436, "grad_norm": 93.07907104492188, "learning_rate": 8.104005167958657e-06, "loss": 2.2344, "step": 1174 }, { "epoch": 0.18976098191214472, "grad_norm": 41.96046447753906, "learning_rate": 8.102390180878553e-06, "loss": 1.1406, "step": 1175 }, { "epoch": 0.18992248062015504, "grad_norm": 29.56212615966797, "learning_rate": 8.10077519379845e-06, "loss": 0.8887, "step": 1176 }, { "epoch": 0.19008397932816537, "grad_norm": 25.21943473815918, "learning_rate": 8.099160206718348e-06, "loss": 0.2388, "step": 1177 }, { "epoch": 0.1902454780361757, "grad_norm": 18.228487014770508, "learning_rate": 8.097545219638243e-06, "loss": 0.25, "step": 1178 }, { "epoch": 0.19040697674418605, "grad_norm": 59.69008255004883, "learning_rate": 8.09593023255814e-06, "loss": 1.4609, "step": 1179 }, { "epoch": 0.19056847545219638, "grad_norm": 35.02397918701172, "learning_rate": 8.094315245478036e-06, "loss": 0.2549, "step": 1180 }, { "epoch": 0.1907299741602067, "grad_norm": 71.45579528808594, "learning_rate": 8.092700258397934e-06, "loss": 1.5625, "step": 1181 }, { "epoch": 0.19089147286821706, "grad_norm": 37.15359878540039, "learning_rate": 8.09108527131783e-06, "loss": 1.0469, "step": 1182 }, { "epoch": 0.1910529715762274, "grad_norm": 25.438526153564453, "learning_rate": 8.089470284237727e-06, "loss": 0.5835, "step": 1183 }, { "epoch": 0.19121447028423771, "grad_norm": 91.36073303222656, "learning_rate": 8.087855297157623e-06, "loss": 3.3262, "step": 1184 }, { "epoch": 0.19137596899224807, "grad_norm": 37.14643859863281, "learning_rate": 8.08624031007752e-06, "loss": 0.9043, "step": 1185 }, { "epoch": 0.1915374677002584, "grad_norm": 41.95316696166992, "learning_rate": 8.084625322997418e-06, "loss": 0.5923, "step": 1186 }, { "epoch": 0.19169896640826872, "grad_norm": 53.05824661254883, "learning_rate": 8.083010335917313e-06, "loss": 0.9727, "step": 1187 }, { "epoch": 0.19186046511627908, "grad_norm": 36.97319793701172, "learning_rate": 8.08139534883721e-06, "loss": 0.8828, "step": 1188 }, { "epoch": 0.1920219638242894, "grad_norm": 40.71122360229492, "learning_rate": 8.079780361757106e-06, "loss": 1.1875, "step": 1189 }, { "epoch": 0.19218346253229973, "grad_norm": 48.688682556152344, "learning_rate": 8.078165374677002e-06, "loss": 1.394, "step": 1190 }, { "epoch": 0.1923449612403101, "grad_norm": 53.7146110534668, "learning_rate": 8.0765503875969e-06, "loss": 0.9668, "step": 1191 }, { "epoch": 0.1925064599483204, "grad_norm": 35.543212890625, "learning_rate": 8.074935400516797e-06, "loss": 0.5293, "step": 1192 }, { "epoch": 0.19266795865633074, "grad_norm": 45.186588287353516, "learning_rate": 8.073320413436693e-06, "loss": 0.6719, "step": 1193 }, { "epoch": 0.1928294573643411, "grad_norm": 34.65067672729492, "learning_rate": 8.07170542635659e-06, "loss": 0.7402, "step": 1194 }, { "epoch": 0.19299095607235142, "grad_norm": 65.64677429199219, "learning_rate": 8.070090439276487e-06, "loss": 1.3672, "step": 1195 }, { "epoch": 0.19315245478036175, "grad_norm": 50.57604217529297, "learning_rate": 8.068475452196383e-06, "loss": 1.041, "step": 1196 }, { "epoch": 0.1933139534883721, "grad_norm": 50.494544982910156, "learning_rate": 8.066860465116279e-06, "loss": 0.9902, "step": 1197 }, { "epoch": 0.19347545219638243, "grad_norm": 40.95868682861328, "learning_rate": 8.065245478036176e-06, "loss": 0.959, "step": 1198 }, { "epoch": 0.19363695090439276, "grad_norm": 48.237152099609375, "learning_rate": 8.063630490956072e-06, "loss": 0.4961, "step": 1199 }, { "epoch": 0.1937984496124031, "grad_norm": 22.979984283447266, "learning_rate": 8.06201550387597e-06, "loss": 0.374, "step": 1200 }, { "epoch": 0.19395994832041344, "grad_norm": 29.24559783935547, "learning_rate": 8.060400516795867e-06, "loss": 0.4387, "step": 1201 }, { "epoch": 0.19412144702842377, "grad_norm": 99.08175659179688, "learning_rate": 8.058785529715762e-06, "loss": 2.4062, "step": 1202 }, { "epoch": 0.19428294573643412, "grad_norm": 43.962120056152344, "learning_rate": 8.05717054263566e-06, "loss": 1.3398, "step": 1203 }, { "epoch": 0.19444444444444445, "grad_norm": 35.819435119628906, "learning_rate": 8.055555555555557e-06, "loss": 0.9855, "step": 1204 }, { "epoch": 0.19460594315245477, "grad_norm": 15.962981224060059, "learning_rate": 8.053940568475453e-06, "loss": 0.0657, "step": 1205 }, { "epoch": 0.19476744186046513, "grad_norm": 30.909862518310547, "learning_rate": 8.052325581395349e-06, "loss": 0.3259, "step": 1206 }, { "epoch": 0.19492894056847546, "grad_norm": 62.636383056640625, "learning_rate": 8.050710594315246e-06, "loss": 0.9111, "step": 1207 }, { "epoch": 0.19509043927648578, "grad_norm": 61.68458938598633, "learning_rate": 8.049095607235142e-06, "loss": 1.5664, "step": 1208 }, { "epoch": 0.19525193798449614, "grad_norm": 42.89532470703125, "learning_rate": 8.04748062015504e-06, "loss": 0.6855, "step": 1209 }, { "epoch": 0.19541343669250646, "grad_norm": 45.40447998046875, "learning_rate": 8.045865633074937e-06, "loss": 0.8672, "step": 1210 }, { "epoch": 0.1955749354005168, "grad_norm": 80.853515625, "learning_rate": 8.044250645994832e-06, "loss": 0.7285, "step": 1211 }, { "epoch": 0.19573643410852712, "grad_norm": 38.23036575317383, "learning_rate": 8.04263565891473e-06, "loss": 0.7773, "step": 1212 }, { "epoch": 0.19589793281653747, "grad_norm": 96.30510711669922, "learning_rate": 8.041020671834625e-06, "loss": 2.4414, "step": 1213 }, { "epoch": 0.1960594315245478, "grad_norm": 42.27883529663086, "learning_rate": 8.039405684754523e-06, "loss": 0.519, "step": 1214 }, { "epoch": 0.19622093023255813, "grad_norm": 25.302928924560547, "learning_rate": 8.037790697674419e-06, "loss": 0.2395, "step": 1215 }, { "epoch": 0.19638242894056848, "grad_norm": 62.73643493652344, "learning_rate": 8.036175710594316e-06, "loss": 1.6641, "step": 1216 }, { "epoch": 0.1965439276485788, "grad_norm": 20.237260818481445, "learning_rate": 8.034560723514212e-06, "loss": 0.157, "step": 1217 }, { "epoch": 0.19670542635658914, "grad_norm": 33.50948715209961, "learning_rate": 8.032945736434109e-06, "loss": 0.7881, "step": 1218 }, { "epoch": 0.1968669250645995, "grad_norm": 173.59532165527344, "learning_rate": 8.031330749354007e-06, "loss": 4.6367, "step": 1219 }, { "epoch": 0.19702842377260982, "grad_norm": 156.19952392578125, "learning_rate": 8.029715762273902e-06, "loss": 3.998, "step": 1220 }, { "epoch": 0.19718992248062014, "grad_norm": 137.50320434570312, "learning_rate": 8.0281007751938e-06, "loss": 3.8555, "step": 1221 }, { "epoch": 0.1973514211886305, "grad_norm": 59.760780334472656, "learning_rate": 8.026485788113695e-06, "loss": 2.0908, "step": 1222 }, { "epoch": 0.19751291989664083, "grad_norm": 28.317174911499023, "learning_rate": 8.024870801033591e-06, "loss": 0.3662, "step": 1223 }, { "epoch": 0.19767441860465115, "grad_norm": 51.57847595214844, "learning_rate": 8.023255813953488e-06, "loss": 0.9648, "step": 1224 }, { "epoch": 0.1978359173126615, "grad_norm": 21.217079162597656, "learning_rate": 8.021640826873386e-06, "loss": 0.1382, "step": 1225 }, { "epoch": 0.19799741602067183, "grad_norm": 147.22415161132812, "learning_rate": 8.020025839793283e-06, "loss": 1.8359, "step": 1226 }, { "epoch": 0.19815891472868216, "grad_norm": 45.24002456665039, "learning_rate": 8.018410852713179e-06, "loss": 0.6914, "step": 1227 }, { "epoch": 0.19832041343669252, "grad_norm": 56.47598648071289, "learning_rate": 8.016795865633076e-06, "loss": 0.6328, "step": 1228 }, { "epoch": 0.19848191214470284, "grad_norm": 14.53053092956543, "learning_rate": 8.015180878552972e-06, "loss": 0.2905, "step": 1229 }, { "epoch": 0.19864341085271317, "grad_norm": 34.692649841308594, "learning_rate": 8.013565891472868e-06, "loss": 0.4995, "step": 1230 }, { "epoch": 0.19880490956072353, "grad_norm": 28.291248321533203, "learning_rate": 8.011950904392765e-06, "loss": 0.6332, "step": 1231 }, { "epoch": 0.19896640826873385, "grad_norm": 68.05488586425781, "learning_rate": 8.010335917312663e-06, "loss": 1.8115, "step": 1232 }, { "epoch": 0.19912790697674418, "grad_norm": 73.76227569580078, "learning_rate": 8.008720930232558e-06, "loss": 2.1953, "step": 1233 }, { "epoch": 0.19928940568475453, "grad_norm": 51.32095718383789, "learning_rate": 8.007105943152456e-06, "loss": 0.8672, "step": 1234 }, { "epoch": 0.19945090439276486, "grad_norm": 23.968076705932617, "learning_rate": 8.005490956072353e-06, "loss": 0.2598, "step": 1235 }, { "epoch": 0.1996124031007752, "grad_norm": 41.50014877319336, "learning_rate": 8.003875968992249e-06, "loss": 0.5103, "step": 1236 }, { "epoch": 0.19977390180878554, "grad_norm": 6.9409661293029785, "learning_rate": 8.002260981912146e-06, "loss": 0.0812, "step": 1237 }, { "epoch": 0.19993540051679587, "grad_norm": 76.75154876708984, "learning_rate": 8.000645994832042e-06, "loss": 1.5566, "step": 1238 }, { "epoch": 0.2000968992248062, "grad_norm": 19.476953506469727, "learning_rate": 7.999031007751938e-06, "loss": 0.2656, "step": 1239 }, { "epoch": 0.20025839793281655, "grad_norm": 37.173763275146484, "learning_rate": 7.997416020671835e-06, "loss": 0.5977, "step": 1240 }, { "epoch": 0.20041989664082688, "grad_norm": 27.764719009399414, "learning_rate": 7.995801033591733e-06, "loss": 0.7959, "step": 1241 }, { "epoch": 0.2005813953488372, "grad_norm": 33.042850494384766, "learning_rate": 7.994186046511628e-06, "loss": 0.9194, "step": 1242 }, { "epoch": 0.20074289405684753, "grad_norm": 35.08023452758789, "learning_rate": 7.992571059431526e-06, "loss": 1.0471, "step": 1243 }, { "epoch": 0.2009043927648579, "grad_norm": 50.08639907836914, "learning_rate": 7.990956072351423e-06, "loss": 1.0527, "step": 1244 }, { "epoch": 0.2010658914728682, "grad_norm": 37.68444061279297, "learning_rate": 7.989341085271319e-06, "loss": 0.7559, "step": 1245 }, { "epoch": 0.20122739018087854, "grad_norm": 57.319610595703125, "learning_rate": 7.987726098191215e-06, "loss": 1.2812, "step": 1246 }, { "epoch": 0.2013888888888889, "grad_norm": 35.170475006103516, "learning_rate": 7.986111111111112e-06, "loss": 0.8867, "step": 1247 }, { "epoch": 0.20155038759689922, "grad_norm": 58.89297866821289, "learning_rate": 7.984496124031008e-06, "loss": 0.8188, "step": 1248 }, { "epoch": 0.20171188630490955, "grad_norm": 41.576175689697266, "learning_rate": 7.982881136950905e-06, "loss": 0.7402, "step": 1249 }, { "epoch": 0.2018733850129199, "grad_norm": 40.48714065551758, "learning_rate": 7.981266149870802e-06, "loss": 0.3975, "step": 1250 }, { "epoch": 0.20203488372093023, "grad_norm": 49.96556091308594, "learning_rate": 7.979651162790698e-06, "loss": 0.5779, "step": 1251 }, { "epoch": 0.20219638242894056, "grad_norm": 124.23779296875, "learning_rate": 7.978036175710596e-06, "loss": 3.2969, "step": 1252 }, { "epoch": 0.2023578811369509, "grad_norm": 72.87438201904297, "learning_rate": 7.976421188630491e-06, "loss": 1.293, "step": 1253 }, { "epoch": 0.20251937984496124, "grad_norm": 174.10812377929688, "learning_rate": 7.974806201550389e-06, "loss": 2.543, "step": 1254 }, { "epoch": 0.20268087855297157, "grad_norm": 43.12869644165039, "learning_rate": 7.973191214470284e-06, "loss": 1.2012, "step": 1255 }, { "epoch": 0.20284237726098192, "grad_norm": 19.551301956176758, "learning_rate": 7.971576227390182e-06, "loss": 0.1807, "step": 1256 }, { "epoch": 0.20300387596899225, "grad_norm": 48.587806701660156, "learning_rate": 7.969961240310078e-06, "loss": 0.6094, "step": 1257 }, { "epoch": 0.20316537467700257, "grad_norm": 24.470178604125977, "learning_rate": 7.968346253229975e-06, "loss": 0.1758, "step": 1258 }, { "epoch": 0.20332687338501293, "grad_norm": 71.4449234008789, "learning_rate": 7.966731266149872e-06, "loss": 1.1523, "step": 1259 }, { "epoch": 0.20348837209302326, "grad_norm": 35.61811828613281, "learning_rate": 7.965116279069768e-06, "loss": 0.4688, "step": 1260 }, { "epoch": 0.20364987080103358, "grad_norm": 37.27864074707031, "learning_rate": 7.963501291989665e-06, "loss": 0.5947, "step": 1261 }, { "epoch": 0.20381136950904394, "grad_norm": 17.65334129333496, "learning_rate": 7.961886304909561e-06, "loss": 0.1849, "step": 1262 }, { "epoch": 0.20397286821705427, "grad_norm": 42.678890228271484, "learning_rate": 7.960271317829457e-06, "loss": 0.8008, "step": 1263 }, { "epoch": 0.2041343669250646, "grad_norm": 53.14448547363281, "learning_rate": 7.958656330749354e-06, "loss": 1.1738, "step": 1264 }, { "epoch": 0.20429586563307495, "grad_norm": 33.393856048583984, "learning_rate": 7.957041343669252e-06, "loss": 0.6045, "step": 1265 }, { "epoch": 0.20445736434108527, "grad_norm": 46.46134567260742, "learning_rate": 7.955426356589147e-06, "loss": 1.2646, "step": 1266 }, { "epoch": 0.2046188630490956, "grad_norm": 110.01546478271484, "learning_rate": 7.953811369509045e-06, "loss": 2.582, "step": 1267 }, { "epoch": 0.20478036175710596, "grad_norm": 80.14481353759766, "learning_rate": 7.952196382428942e-06, "loss": 1.5264, "step": 1268 }, { "epoch": 0.20494186046511628, "grad_norm": 87.41683197021484, "learning_rate": 7.950581395348838e-06, "loss": 1.479, "step": 1269 }, { "epoch": 0.2051033591731266, "grad_norm": 13.897855758666992, "learning_rate": 7.948966408268735e-06, "loss": 0.1543, "step": 1270 }, { "epoch": 0.20526485788113696, "grad_norm": 21.28791618347168, "learning_rate": 7.947351421188631e-06, "loss": 0.5249, "step": 1271 }, { "epoch": 0.2054263565891473, "grad_norm": 59.33272933959961, "learning_rate": 7.945736434108527e-06, "loss": 0.5635, "step": 1272 }, { "epoch": 0.20558785529715762, "grad_norm": 41.77866744995117, "learning_rate": 7.944121447028424e-06, "loss": 1.4629, "step": 1273 }, { "epoch": 0.20574935400516794, "grad_norm": 90.01881408691406, "learning_rate": 7.942506459948322e-06, "loss": 1.6704, "step": 1274 }, { "epoch": 0.2059108527131783, "grad_norm": 45.44303512573242, "learning_rate": 7.940891472868217e-06, "loss": 1.2808, "step": 1275 }, { "epoch": 0.20607235142118863, "grad_norm": 42.90483856201172, "learning_rate": 7.939276485788115e-06, "loss": 1.123, "step": 1276 }, { "epoch": 0.20623385012919895, "grad_norm": 11.939225196838379, "learning_rate": 7.937661498708012e-06, "loss": 0.0827, "step": 1277 }, { "epoch": 0.2063953488372093, "grad_norm": 115.77215576171875, "learning_rate": 7.936046511627908e-06, "loss": 2.1289, "step": 1278 }, { "epoch": 0.20655684754521964, "grad_norm": 69.2236099243164, "learning_rate": 7.934431524547804e-06, "loss": 0.7969, "step": 1279 }, { "epoch": 0.20671834625322996, "grad_norm": 47.13788604736328, "learning_rate": 7.932816537467701e-06, "loss": 0.8262, "step": 1280 }, { "epoch": 0.20687984496124032, "grad_norm": 92.15480041503906, "learning_rate": 7.931201550387597e-06, "loss": 1.8359, "step": 1281 }, { "epoch": 0.20704134366925064, "grad_norm": 60.39603805541992, "learning_rate": 7.929586563307494e-06, "loss": 1.1875, "step": 1282 }, { "epoch": 0.20720284237726097, "grad_norm": 27.237842559814453, "learning_rate": 7.927971576227391e-06, "loss": 0.667, "step": 1283 }, { "epoch": 0.20736434108527133, "grad_norm": 117.6595230102539, "learning_rate": 7.926356589147287e-06, "loss": 3.3477, "step": 1284 }, { "epoch": 0.20752583979328165, "grad_norm": 30.436199188232422, "learning_rate": 7.924741602067185e-06, "loss": 0.4248, "step": 1285 }, { "epoch": 0.20768733850129198, "grad_norm": 42.81721878051758, "learning_rate": 7.92312661498708e-06, "loss": 0.4067, "step": 1286 }, { "epoch": 0.20784883720930233, "grad_norm": 29.237070083618164, "learning_rate": 7.921511627906978e-06, "loss": 0.4014, "step": 1287 }, { "epoch": 0.20801033591731266, "grad_norm": 27.257287979125977, "learning_rate": 7.919896640826873e-06, "loss": 0.5156, "step": 1288 }, { "epoch": 0.208171834625323, "grad_norm": 33.59557342529297, "learning_rate": 7.91828165374677e-06, "loss": 1.1055, "step": 1289 }, { "epoch": 0.20833333333333334, "grad_norm": 38.88725662231445, "learning_rate": 7.916666666666667e-06, "loss": 0.4277, "step": 1290 }, { "epoch": 0.20849483204134367, "grad_norm": 40.15574645996094, "learning_rate": 7.915051679586564e-06, "loss": 0.5918, "step": 1291 }, { "epoch": 0.208656330749354, "grad_norm": 89.15220642089844, "learning_rate": 7.913436692506461e-06, "loss": 1.832, "step": 1292 }, { "epoch": 0.20881782945736435, "grad_norm": 15.842670440673828, "learning_rate": 7.911821705426357e-06, "loss": 0.3271, "step": 1293 }, { "epoch": 0.20897932816537468, "grad_norm": 38.856380462646484, "learning_rate": 7.910206718346254e-06, "loss": 1.0859, "step": 1294 }, { "epoch": 0.209140826873385, "grad_norm": 92.00164794921875, "learning_rate": 7.90859173126615e-06, "loss": 1.1807, "step": 1295 }, { "epoch": 0.20930232558139536, "grad_norm": 67.41541290283203, "learning_rate": 7.906976744186048e-06, "loss": 0.9531, "step": 1296 }, { "epoch": 0.2094638242894057, "grad_norm": 112.64704132080078, "learning_rate": 7.905361757105943e-06, "loss": 1.8555, "step": 1297 }, { "epoch": 0.209625322997416, "grad_norm": 28.67453956604004, "learning_rate": 7.90374677002584e-06, "loss": 0.9033, "step": 1298 }, { "epoch": 0.20978682170542637, "grad_norm": 29.991031646728516, "learning_rate": 7.902131782945736e-06, "loss": 0.7085, "step": 1299 }, { "epoch": 0.2099483204134367, "grad_norm": 46.27861022949219, "learning_rate": 7.900516795865634e-06, "loss": 0.4385, "step": 1300 }, { "epoch": 0.21010981912144702, "grad_norm": 72.03160858154297, "learning_rate": 7.898901808785531e-06, "loss": 1.7347, "step": 1301 }, { "epoch": 0.21027131782945738, "grad_norm": 53.12786865234375, "learning_rate": 7.897286821705427e-06, "loss": 0.4736, "step": 1302 }, { "epoch": 0.2104328165374677, "grad_norm": 21.128873825073242, "learning_rate": 7.895671834625324e-06, "loss": 0.1262, "step": 1303 }, { "epoch": 0.21059431524547803, "grad_norm": 24.772499084472656, "learning_rate": 7.89405684754522e-06, "loss": 0.3079, "step": 1304 }, { "epoch": 0.21075581395348839, "grad_norm": 36.538333892822266, "learning_rate": 7.892441860465116e-06, "loss": 0.8276, "step": 1305 }, { "epoch": 0.2109173126614987, "grad_norm": 36.32126235961914, "learning_rate": 7.890826873385013e-06, "loss": 0.6484, "step": 1306 }, { "epoch": 0.21107881136950904, "grad_norm": 54.84724044799805, "learning_rate": 7.88921188630491e-06, "loss": 0.8228, "step": 1307 }, { "epoch": 0.21124031007751937, "grad_norm": 65.41783905029297, "learning_rate": 7.887596899224806e-06, "loss": 1.502, "step": 1308 }, { "epoch": 0.21140180878552972, "grad_norm": 37.1949348449707, "learning_rate": 7.885981912144704e-06, "loss": 0.6094, "step": 1309 }, { "epoch": 0.21156330749354005, "grad_norm": 49.11770248413086, "learning_rate": 7.884366925064601e-06, "loss": 0.8984, "step": 1310 }, { "epoch": 0.21172480620155038, "grad_norm": 39.44390869140625, "learning_rate": 7.882751937984497e-06, "loss": 0.5711, "step": 1311 }, { "epoch": 0.21188630490956073, "grad_norm": 49.7438850402832, "learning_rate": 7.881136950904393e-06, "loss": 1.0295, "step": 1312 }, { "epoch": 0.21204780361757106, "grad_norm": 15.86960220336914, "learning_rate": 7.87952196382429e-06, "loss": 0.208, "step": 1313 }, { "epoch": 0.21220930232558138, "grad_norm": 38.89486312866211, "learning_rate": 7.877906976744186e-06, "loss": 1.3203, "step": 1314 }, { "epoch": 0.21237080103359174, "grad_norm": 62.54158401489258, "learning_rate": 7.876291989664083e-06, "loss": 1.5156, "step": 1315 }, { "epoch": 0.21253229974160207, "grad_norm": 78.18499755859375, "learning_rate": 7.87467700258398e-06, "loss": 1.5029, "step": 1316 }, { "epoch": 0.2126937984496124, "grad_norm": 23.167755126953125, "learning_rate": 7.873062015503876e-06, "loss": 0.4121, "step": 1317 }, { "epoch": 0.21285529715762275, "grad_norm": 39.086978912353516, "learning_rate": 7.871447028423774e-06, "loss": 0.9707, "step": 1318 }, { "epoch": 0.21301679586563307, "grad_norm": 98.23711395263672, "learning_rate": 7.869832041343671e-06, "loss": 2.7349, "step": 1319 }, { "epoch": 0.2131782945736434, "grad_norm": 32.52608871459961, "learning_rate": 7.868217054263567e-06, "loss": 0.6826, "step": 1320 }, { "epoch": 0.21333979328165376, "grad_norm": 33.887264251708984, "learning_rate": 7.866602067183462e-06, "loss": 0.2802, "step": 1321 }, { "epoch": 0.21350129198966408, "grad_norm": 30.62562370300293, "learning_rate": 7.86498708010336e-06, "loss": 0.4663, "step": 1322 }, { "epoch": 0.2136627906976744, "grad_norm": 41.92716979980469, "learning_rate": 7.863372093023256e-06, "loss": 0.5176, "step": 1323 }, { "epoch": 0.21382428940568476, "grad_norm": 61.835044860839844, "learning_rate": 7.861757105943153e-06, "loss": 0.824, "step": 1324 }, { "epoch": 0.2139857881136951, "grad_norm": 50.35405731201172, "learning_rate": 7.86014211886305e-06, "loss": 0.748, "step": 1325 }, { "epoch": 0.21414728682170542, "grad_norm": 18.418636322021484, "learning_rate": 7.858527131782946e-06, "loss": 0.6304, "step": 1326 }, { "epoch": 0.21430878552971577, "grad_norm": 25.511079788208008, "learning_rate": 7.856912144702843e-06, "loss": 0.5605, "step": 1327 }, { "epoch": 0.2144702842377261, "grad_norm": 32.37224197387695, "learning_rate": 7.85529715762274e-06, "loss": 0.7129, "step": 1328 }, { "epoch": 0.21463178294573643, "grad_norm": 32.19040298461914, "learning_rate": 7.853682170542637e-06, "loss": 0.3291, "step": 1329 }, { "epoch": 0.21479328165374678, "grad_norm": 54.172332763671875, "learning_rate": 7.852067183462532e-06, "loss": 0.7026, "step": 1330 }, { "epoch": 0.2149547803617571, "grad_norm": 93.69122314453125, "learning_rate": 7.85045219638243e-06, "loss": 2.1875, "step": 1331 }, { "epoch": 0.21511627906976744, "grad_norm": 82.51332092285156, "learning_rate": 7.848837209302325e-06, "loss": 1.418, "step": 1332 }, { "epoch": 0.2152777777777778, "grad_norm": 83.20160675048828, "learning_rate": 7.847222222222223e-06, "loss": 1.2793, "step": 1333 }, { "epoch": 0.21543927648578812, "grad_norm": 39.613895416259766, "learning_rate": 7.84560723514212e-06, "loss": 0.627, "step": 1334 }, { "epoch": 0.21560077519379844, "grad_norm": 30.036100387573242, "learning_rate": 7.843992248062016e-06, "loss": 0.671, "step": 1335 }, { "epoch": 0.2157622739018088, "grad_norm": 41.91242218017578, "learning_rate": 7.842377260981913e-06, "loss": 1.2559, "step": 1336 }, { "epoch": 0.21592377260981913, "grad_norm": 23.244606018066406, "learning_rate": 7.840762273901809e-06, "loss": 0.3755, "step": 1337 }, { "epoch": 0.21608527131782945, "grad_norm": 27.81515884399414, "learning_rate": 7.839147286821705e-06, "loss": 0.1787, "step": 1338 }, { "epoch": 0.21624677002583978, "grad_norm": 25.648439407348633, "learning_rate": 7.837532299741602e-06, "loss": 0.2141, "step": 1339 }, { "epoch": 0.21640826873385013, "grad_norm": 9.746564865112305, "learning_rate": 7.8359173126615e-06, "loss": 0.103, "step": 1340 }, { "epoch": 0.21656976744186046, "grad_norm": 17.197507858276367, "learning_rate": 7.834302325581395e-06, "loss": 0.3457, "step": 1341 }, { "epoch": 0.2167312661498708, "grad_norm": 47.29166793823242, "learning_rate": 7.832687338501293e-06, "loss": 0.6162, "step": 1342 }, { "epoch": 0.21689276485788114, "grad_norm": 57.55278396606445, "learning_rate": 7.83107235142119e-06, "loss": 1.3789, "step": 1343 }, { "epoch": 0.21705426356589147, "grad_norm": 61.294410705566406, "learning_rate": 7.829457364341086e-06, "loss": 1.3457, "step": 1344 }, { "epoch": 0.2172157622739018, "grad_norm": 62.340457916259766, "learning_rate": 7.827842377260982e-06, "loss": 1.4995, "step": 1345 }, { "epoch": 0.21737726098191215, "grad_norm": 36.110511779785156, "learning_rate": 7.826227390180879e-06, "loss": 0.1685, "step": 1346 }, { "epoch": 0.21753875968992248, "grad_norm": 56.896995544433594, "learning_rate": 7.824612403100775e-06, "loss": 1.123, "step": 1347 }, { "epoch": 0.2177002583979328, "grad_norm": 40.98909378051758, "learning_rate": 7.822997416020672e-06, "loss": 0.3818, "step": 1348 }, { "epoch": 0.21786175710594316, "grad_norm": 44.0789794921875, "learning_rate": 7.82138242894057e-06, "loss": 0.7207, "step": 1349 }, { "epoch": 0.2180232558139535, "grad_norm": 34.21632766723633, "learning_rate": 7.819767441860465e-06, "loss": 1.1332, "step": 1350 }, { "epoch": 0.21818475452196381, "grad_norm": 38.77276611328125, "learning_rate": 7.818152454780363e-06, "loss": 1.6454, "step": 1351 }, { "epoch": 0.21834625322997417, "grad_norm": 47.62874221801758, "learning_rate": 7.81653746770026e-06, "loss": 0.5469, "step": 1352 }, { "epoch": 0.2185077519379845, "grad_norm": 71.51105499267578, "learning_rate": 7.814922480620156e-06, "loss": 1.1812, "step": 1353 }, { "epoch": 0.21866925064599482, "grad_norm": 10.122574806213379, "learning_rate": 7.813307493540051e-06, "loss": 0.3398, "step": 1354 }, { "epoch": 0.21883074935400518, "grad_norm": 23.063068389892578, "learning_rate": 7.811692506459949e-06, "loss": 0.6143, "step": 1355 }, { "epoch": 0.2189922480620155, "grad_norm": 85.43409729003906, "learning_rate": 7.810077519379846e-06, "loss": 1.2012, "step": 1356 }, { "epoch": 0.21915374677002583, "grad_norm": 33.77708053588867, "learning_rate": 7.808462532299742e-06, "loss": 0.467, "step": 1357 }, { "epoch": 0.21931524547803619, "grad_norm": 37.367252349853516, "learning_rate": 7.80684754521964e-06, "loss": 0.3872, "step": 1358 }, { "epoch": 0.2194767441860465, "grad_norm": 146.06304931640625, "learning_rate": 7.805232558139537e-06, "loss": 2.6514, "step": 1359 }, { "epoch": 0.21963824289405684, "grad_norm": 40.23634719848633, "learning_rate": 7.803617571059433e-06, "loss": 0.5527, "step": 1360 }, { "epoch": 0.2197997416020672, "grad_norm": 9.62831974029541, "learning_rate": 7.802002583979328e-06, "loss": 0.2056, "step": 1361 }, { "epoch": 0.21996124031007752, "grad_norm": 62.042274475097656, "learning_rate": 7.800387596899226e-06, "loss": 2.6875, "step": 1362 }, { "epoch": 0.22012273901808785, "grad_norm": 45.44324493408203, "learning_rate": 7.798772609819121e-06, "loss": 0.6104, "step": 1363 }, { "epoch": 0.2202842377260982, "grad_norm": 49.0070686340332, "learning_rate": 7.797157622739019e-06, "loss": 1.3145, "step": 1364 }, { "epoch": 0.22044573643410853, "grad_norm": 77.9384765625, "learning_rate": 7.795542635658916e-06, "loss": 2.2041, "step": 1365 }, { "epoch": 0.22060723514211886, "grad_norm": 45.292964935302734, "learning_rate": 7.793927648578812e-06, "loss": 0.748, "step": 1366 }, { "epoch": 0.2207687338501292, "grad_norm": 63.299930572509766, "learning_rate": 7.79231266149871e-06, "loss": 1.1069, "step": 1367 }, { "epoch": 0.22093023255813954, "grad_norm": 27.412677764892578, "learning_rate": 7.790697674418605e-06, "loss": 0.769, "step": 1368 }, { "epoch": 0.22109173126614987, "grad_norm": 20.129804611206055, "learning_rate": 7.789082687338502e-06, "loss": 0.2949, "step": 1369 }, { "epoch": 0.22125322997416022, "grad_norm": 42.75400924682617, "learning_rate": 7.787467700258398e-06, "loss": 0.4409, "step": 1370 }, { "epoch": 0.22141472868217055, "grad_norm": 72.54290771484375, "learning_rate": 7.785852713178296e-06, "loss": 1.6719, "step": 1371 }, { "epoch": 0.22157622739018087, "grad_norm": 36.29075622558594, "learning_rate": 7.784237726098191e-06, "loss": 0.4404, "step": 1372 }, { "epoch": 0.2217377260981912, "grad_norm": 143.77635192871094, "learning_rate": 7.782622739018089e-06, "loss": 3.2334, "step": 1373 }, { "epoch": 0.22189922480620156, "grad_norm": 83.1842269897461, "learning_rate": 7.781007751937986e-06, "loss": 2.0234, "step": 1374 }, { "epoch": 0.22206072351421188, "grad_norm": 29.28056526184082, "learning_rate": 7.779392764857882e-06, "loss": 0.3926, "step": 1375 }, { "epoch": 0.2222222222222222, "grad_norm": 33.14296340942383, "learning_rate": 7.77777777777778e-06, "loss": 1.0176, "step": 1376 }, { "epoch": 0.22238372093023256, "grad_norm": 27.708356857299805, "learning_rate": 7.776162790697675e-06, "loss": 0.7383, "step": 1377 }, { "epoch": 0.2225452196382429, "grad_norm": 36.7325325012207, "learning_rate": 7.774547803617572e-06, "loss": 0.8989, "step": 1378 }, { "epoch": 0.22270671834625322, "grad_norm": 37.081512451171875, "learning_rate": 7.772932816537468e-06, "loss": 0.4685, "step": 1379 }, { "epoch": 0.22286821705426357, "grad_norm": 67.60771179199219, "learning_rate": 7.771317829457365e-06, "loss": 1.2852, "step": 1380 }, { "epoch": 0.2230297157622739, "grad_norm": 22.773408889770508, "learning_rate": 7.769702842377261e-06, "loss": 0.3691, "step": 1381 }, { "epoch": 0.22319121447028423, "grad_norm": 40.81745529174805, "learning_rate": 7.768087855297159e-06, "loss": 0.7349, "step": 1382 }, { "epoch": 0.22335271317829458, "grad_norm": 39.970272064208984, "learning_rate": 7.766472868217056e-06, "loss": 0.4746, "step": 1383 }, { "epoch": 0.2235142118863049, "grad_norm": 57.969905853271484, "learning_rate": 7.764857881136952e-06, "loss": 1.1133, "step": 1384 }, { "epoch": 0.22367571059431524, "grad_norm": 83.32913970947266, "learning_rate": 7.763242894056849e-06, "loss": 2.0635, "step": 1385 }, { "epoch": 0.2238372093023256, "grad_norm": 71.51854705810547, "learning_rate": 7.761627906976745e-06, "loss": 1.1045, "step": 1386 }, { "epoch": 0.22399870801033592, "grad_norm": 44.0305061340332, "learning_rate": 7.76001291989664e-06, "loss": 0.8945, "step": 1387 }, { "epoch": 0.22416020671834624, "grad_norm": 219.11114501953125, "learning_rate": 7.758397932816538e-06, "loss": 1.834, "step": 1388 }, { "epoch": 0.2243217054263566, "grad_norm": 35.94410705566406, "learning_rate": 7.756782945736435e-06, "loss": 0.7734, "step": 1389 }, { "epoch": 0.22448320413436693, "grad_norm": 36.378265380859375, "learning_rate": 7.755167958656331e-06, "loss": 0.5332, "step": 1390 }, { "epoch": 0.22464470284237725, "grad_norm": 74.21299743652344, "learning_rate": 7.753552971576228e-06, "loss": 1.5479, "step": 1391 }, { "epoch": 0.2248062015503876, "grad_norm": 82.30271911621094, "learning_rate": 7.751937984496126e-06, "loss": 1.8242, "step": 1392 }, { "epoch": 0.22496770025839793, "grad_norm": 25.0136775970459, "learning_rate": 7.750322997416022e-06, "loss": 0.2468, "step": 1393 }, { "epoch": 0.22512919896640826, "grad_norm": 19.025049209594727, "learning_rate": 7.748708010335917e-06, "loss": 0.4038, "step": 1394 }, { "epoch": 0.22529069767441862, "grad_norm": 30.616962432861328, "learning_rate": 7.747093023255815e-06, "loss": 0.3926, "step": 1395 }, { "epoch": 0.22545219638242894, "grad_norm": 17.288589477539062, "learning_rate": 7.74547803617571e-06, "loss": 0.2544, "step": 1396 }, { "epoch": 0.22561369509043927, "grad_norm": 12.515443801879883, "learning_rate": 7.743863049095608e-06, "loss": 0.0535, "step": 1397 }, { "epoch": 0.22577519379844962, "grad_norm": 22.183902740478516, "learning_rate": 7.742248062015505e-06, "loss": 0.6729, "step": 1398 }, { "epoch": 0.22593669250645995, "grad_norm": 34.869140625, "learning_rate": 7.740633074935401e-06, "loss": 1.1348, "step": 1399 }, { "epoch": 0.22609819121447028, "grad_norm": 32.28328323364258, "learning_rate": 7.739018087855298e-06, "loss": 0.5522, "step": 1400 }, { "epoch": 0.22625968992248063, "grad_norm": 18.104890823364258, "learning_rate": 7.737403100775194e-06, "loss": 0.3574, "step": 1401 }, { "epoch": 0.22642118863049096, "grad_norm": 26.968860626220703, "learning_rate": 7.735788113695091e-06, "loss": 1.0381, "step": 1402 }, { "epoch": 0.2265826873385013, "grad_norm": 61.75414276123047, "learning_rate": 7.734173126614987e-06, "loss": 1.0488, "step": 1403 }, { "epoch": 0.22674418604651161, "grad_norm": 66.36134338378906, "learning_rate": 7.732558139534885e-06, "loss": 2.3164, "step": 1404 }, { "epoch": 0.22690568475452197, "grad_norm": 32.33060836791992, "learning_rate": 7.73094315245478e-06, "loss": 1.0664, "step": 1405 }, { "epoch": 0.2270671834625323, "grad_norm": 91.71162414550781, "learning_rate": 7.729328165374678e-06, "loss": 1.9756, "step": 1406 }, { "epoch": 0.22722868217054262, "grad_norm": 48.950836181640625, "learning_rate": 7.727713178294575e-06, "loss": 1.0557, "step": 1407 }, { "epoch": 0.22739018087855298, "grad_norm": 47.297157287597656, "learning_rate": 7.72609819121447e-06, "loss": 1.2148, "step": 1408 }, { "epoch": 0.2275516795865633, "grad_norm": 48.324737548828125, "learning_rate": 7.724483204134368e-06, "loss": 0.3369, "step": 1409 }, { "epoch": 0.22771317829457363, "grad_norm": 51.385704040527344, "learning_rate": 7.722868217054264e-06, "loss": 1.043, "step": 1410 }, { "epoch": 0.227874677002584, "grad_norm": 103.96475982666016, "learning_rate": 7.721253229974161e-06, "loss": 2.1309, "step": 1411 }, { "epoch": 0.2280361757105943, "grad_norm": 21.037492752075195, "learning_rate": 7.719638242894057e-06, "loss": 0.3215, "step": 1412 }, { "epoch": 0.22819767441860464, "grad_norm": 27.883092880249023, "learning_rate": 7.718023255813954e-06, "loss": 0.4673, "step": 1413 }, { "epoch": 0.228359173126615, "grad_norm": 71.40032958984375, "learning_rate": 7.71640826873385e-06, "loss": 1.9062, "step": 1414 }, { "epoch": 0.22852067183462532, "grad_norm": 62.36619567871094, "learning_rate": 7.714793281653748e-06, "loss": 1.7734, "step": 1415 }, { "epoch": 0.22868217054263565, "grad_norm": 15.942386627197266, "learning_rate": 7.713178294573645e-06, "loss": 0.0996, "step": 1416 }, { "epoch": 0.228843669250646, "grad_norm": 20.729780197143555, "learning_rate": 7.71156330749354e-06, "loss": 0.5696, "step": 1417 }, { "epoch": 0.22900516795865633, "grad_norm": 72.73824310302734, "learning_rate": 7.709948320413438e-06, "loss": 3.2617, "step": 1418 }, { "epoch": 0.22916666666666666, "grad_norm": 42.88822555541992, "learning_rate": 7.708333333333334e-06, "loss": 0.9629, "step": 1419 }, { "epoch": 0.229328165374677, "grad_norm": 40.650062561035156, "learning_rate": 7.70671834625323e-06, "loss": 1.334, "step": 1420 }, { "epoch": 0.22948966408268734, "grad_norm": 93.28089141845703, "learning_rate": 7.705103359173127e-06, "loss": 2.5566, "step": 1421 }, { "epoch": 0.22965116279069767, "grad_norm": 76.81897735595703, "learning_rate": 7.703488372093024e-06, "loss": 3.0389, "step": 1422 }, { "epoch": 0.22981266149870802, "grad_norm": 38.399288177490234, "learning_rate": 7.70187338501292e-06, "loss": 1.0894, "step": 1423 }, { "epoch": 0.22997416020671835, "grad_norm": 13.102630615234375, "learning_rate": 7.700258397932817e-06, "loss": 0.2324, "step": 1424 }, { "epoch": 0.23013565891472867, "grad_norm": 19.009546279907227, "learning_rate": 7.698643410852715e-06, "loss": 0.2451, "step": 1425 }, { "epoch": 0.23029715762273903, "grad_norm": 40.1020622253418, "learning_rate": 7.69702842377261e-06, "loss": 0.5332, "step": 1426 }, { "epoch": 0.23045865633074936, "grad_norm": 48.66688537597656, "learning_rate": 7.695413436692506e-06, "loss": 1.4023, "step": 1427 }, { "epoch": 0.23062015503875968, "grad_norm": 5.575917720794678, "learning_rate": 7.693798449612404e-06, "loss": 0.0701, "step": 1428 }, { "epoch": 0.23078165374677004, "grad_norm": 48.65315246582031, "learning_rate": 7.6921834625323e-06, "loss": 0.4756, "step": 1429 }, { "epoch": 0.23094315245478036, "grad_norm": 38.59938430786133, "learning_rate": 7.690568475452197e-06, "loss": 1.0811, "step": 1430 }, { "epoch": 0.2311046511627907, "grad_norm": 29.833311080932617, "learning_rate": 7.688953488372094e-06, "loss": 0.5547, "step": 1431 }, { "epoch": 0.23126614987080105, "grad_norm": 21.28484535217285, "learning_rate": 7.68733850129199e-06, "loss": 0.3496, "step": 1432 }, { "epoch": 0.23142764857881137, "grad_norm": 44.01904296875, "learning_rate": 7.685723514211887e-06, "loss": 0.9824, "step": 1433 }, { "epoch": 0.2315891472868217, "grad_norm": 6.661888122558594, "learning_rate": 7.684108527131785e-06, "loss": 0.1299, "step": 1434 }, { "epoch": 0.23175064599483206, "grad_norm": 40.548824310302734, "learning_rate": 7.68249354005168e-06, "loss": 0.6631, "step": 1435 }, { "epoch": 0.23191214470284238, "grad_norm": 7.880975723266602, "learning_rate": 7.680878552971576e-06, "loss": 0.0784, "step": 1436 }, { "epoch": 0.2320736434108527, "grad_norm": 27.503170013427734, "learning_rate": 7.679263565891474e-06, "loss": 0.2998, "step": 1437 }, { "epoch": 0.23223514211886304, "grad_norm": 57.53742599487305, "learning_rate": 7.67764857881137e-06, "loss": 0.7344, "step": 1438 }, { "epoch": 0.2323966408268734, "grad_norm": 40.37491989135742, "learning_rate": 7.676033591731267e-06, "loss": 0.2559, "step": 1439 }, { "epoch": 0.23255813953488372, "grad_norm": 94.28919219970703, "learning_rate": 7.674418604651164e-06, "loss": 2.8828, "step": 1440 }, { "epoch": 0.23271963824289404, "grad_norm": 69.62620544433594, "learning_rate": 7.67280361757106e-06, "loss": 1.1973, "step": 1441 }, { "epoch": 0.2328811369509044, "grad_norm": 26.4761962890625, "learning_rate": 7.671188630490957e-06, "loss": 0.7295, "step": 1442 }, { "epoch": 0.23304263565891473, "grad_norm": 60.63497543334961, "learning_rate": 7.669573643410853e-06, "loss": 1.2539, "step": 1443 }, { "epoch": 0.23320413436692505, "grad_norm": 47.06913757324219, "learning_rate": 7.66795865633075e-06, "loss": 1.1514, "step": 1444 }, { "epoch": 0.2333656330749354, "grad_norm": 33.64692306518555, "learning_rate": 7.666343669250646e-06, "loss": 0.344, "step": 1445 }, { "epoch": 0.23352713178294573, "grad_norm": 36.5890998840332, "learning_rate": 7.664728682170543e-06, "loss": 0.5342, "step": 1446 }, { "epoch": 0.23368863049095606, "grad_norm": 12.553521156311035, "learning_rate": 7.66311369509044e-06, "loss": 0.2471, "step": 1447 }, { "epoch": 0.23385012919896642, "grad_norm": 33.860931396484375, "learning_rate": 7.661498708010337e-06, "loss": 0.3975, "step": 1448 }, { "epoch": 0.23401162790697674, "grad_norm": 13.172255516052246, "learning_rate": 7.659883720930234e-06, "loss": 0.0698, "step": 1449 }, { "epoch": 0.23417312661498707, "grad_norm": 42.6441764831543, "learning_rate": 7.65826873385013e-06, "loss": 1.291, "step": 1450 }, { "epoch": 0.23433462532299743, "grad_norm": 12.520383834838867, "learning_rate": 7.656653746770027e-06, "loss": 0.0714, "step": 1451 }, { "epoch": 0.23449612403100775, "grad_norm": 86.29460144042969, "learning_rate": 7.655038759689923e-06, "loss": 3.0015, "step": 1452 }, { "epoch": 0.23465762273901808, "grad_norm": 28.363445281982422, "learning_rate": 7.653423772609819e-06, "loss": 0.5076, "step": 1453 }, { "epoch": 0.23481912144702843, "grad_norm": 111.24663543701172, "learning_rate": 7.651808785529716e-06, "loss": 2.2734, "step": 1454 }, { "epoch": 0.23498062015503876, "grad_norm": 24.180387496948242, "learning_rate": 7.650193798449613e-06, "loss": 0.2729, "step": 1455 }, { "epoch": 0.2351421188630491, "grad_norm": 104.8469009399414, "learning_rate": 7.648578811369509e-06, "loss": 3.2026, "step": 1456 }, { "epoch": 0.23530361757105944, "grad_norm": 51.2978401184082, "learning_rate": 7.646963824289406e-06, "loss": 0.6533, "step": 1457 }, { "epoch": 0.23546511627906977, "grad_norm": 62.86526870727539, "learning_rate": 7.645348837209304e-06, "loss": 1.0742, "step": 1458 }, { "epoch": 0.2356266149870801, "grad_norm": 22.736347198486328, "learning_rate": 7.6437338501292e-06, "loss": 0.2471, "step": 1459 }, { "epoch": 0.23578811369509045, "grad_norm": 36.42643356323242, "learning_rate": 7.642118863049095e-06, "loss": 0.5547, "step": 1460 }, { "epoch": 0.23594961240310078, "grad_norm": 78.02777862548828, "learning_rate": 7.640503875968993e-06, "loss": 2.2695, "step": 1461 }, { "epoch": 0.2361111111111111, "grad_norm": 69.48680114746094, "learning_rate": 7.638888888888888e-06, "loss": 1.3408, "step": 1462 }, { "epoch": 0.23627260981912146, "grad_norm": 56.30437469482422, "learning_rate": 7.637273901808786e-06, "loss": 1.2441, "step": 1463 }, { "epoch": 0.2364341085271318, "grad_norm": 44.01201629638672, "learning_rate": 7.635658914728683e-06, "loss": 1.7148, "step": 1464 }, { "epoch": 0.2365956072351421, "grad_norm": 27.562721252441406, "learning_rate": 7.634043927648579e-06, "loss": 0.4559, "step": 1465 }, { "epoch": 0.23675710594315247, "grad_norm": 29.493894577026367, "learning_rate": 7.632428940568476e-06, "loss": 0.6019, "step": 1466 }, { "epoch": 0.2369186046511628, "grad_norm": 70.38337707519531, "learning_rate": 7.630813953488374e-06, "loss": 0.7615, "step": 1467 }, { "epoch": 0.23708010335917312, "grad_norm": 30.228424072265625, "learning_rate": 7.629198966408269e-06, "loss": 1.0796, "step": 1468 }, { "epoch": 0.23724160206718345, "grad_norm": 37.03933334350586, "learning_rate": 7.627583979328166e-06, "loss": 0.709, "step": 1469 }, { "epoch": 0.2374031007751938, "grad_norm": 33.65873336791992, "learning_rate": 7.625968992248063e-06, "loss": 0.3291, "step": 1470 }, { "epoch": 0.23756459948320413, "grad_norm": 111.3636474609375, "learning_rate": 7.624354005167959e-06, "loss": 4.1289, "step": 1471 }, { "epoch": 0.23772609819121446, "grad_norm": 49.77165985107422, "learning_rate": 7.622739018087856e-06, "loss": 0.4131, "step": 1472 }, { "epoch": 0.2378875968992248, "grad_norm": 50.58515930175781, "learning_rate": 7.621124031007753e-06, "loss": 0.8311, "step": 1473 }, { "epoch": 0.23804909560723514, "grad_norm": 30.86918067932129, "learning_rate": 7.619509043927649e-06, "loss": 0.5537, "step": 1474 }, { "epoch": 0.23821059431524547, "grad_norm": 38.423004150390625, "learning_rate": 7.617894056847545e-06, "loss": 0.8477, "step": 1475 }, { "epoch": 0.23837209302325582, "grad_norm": 32.84397506713867, "learning_rate": 7.616279069767443e-06, "loss": 0.5518, "step": 1476 }, { "epoch": 0.23853359173126615, "grad_norm": 19.012439727783203, "learning_rate": 7.6146640826873385e-06, "loss": 0.418, "step": 1477 }, { "epoch": 0.23869509043927647, "grad_norm": 30.7504825592041, "learning_rate": 7.613049095607236e-06, "loss": 0.6436, "step": 1478 }, { "epoch": 0.23885658914728683, "grad_norm": 25.225990295410156, "learning_rate": 7.6114341085271325e-06, "loss": 0.4995, "step": 1479 }, { "epoch": 0.23901808785529716, "grad_norm": 58.456783294677734, "learning_rate": 7.609819121447028e-06, "loss": 1.4883, "step": 1480 }, { "epoch": 0.23917958656330748, "grad_norm": 16.492753982543945, "learning_rate": 7.608204134366926e-06, "loss": 0.1567, "step": 1481 }, { "epoch": 0.23934108527131784, "grad_norm": 61.44706344604492, "learning_rate": 7.606589147286822e-06, "loss": 0.6406, "step": 1482 }, { "epoch": 0.23950258397932817, "grad_norm": 87.31669616699219, "learning_rate": 7.604974160206719e-06, "loss": 1.4043, "step": 1483 }, { "epoch": 0.2396640826873385, "grad_norm": 81.10541534423828, "learning_rate": 7.603359173126615e-06, "loss": 2.1406, "step": 1484 }, { "epoch": 0.23982558139534885, "grad_norm": 24.32609748840332, "learning_rate": 7.601744186046513e-06, "loss": 0.4023, "step": 1485 }, { "epoch": 0.23998708010335917, "grad_norm": 20.596948623657227, "learning_rate": 7.600129198966409e-06, "loss": 0.2031, "step": 1486 }, { "epoch": 0.2401485788113695, "grad_norm": 83.36646270751953, "learning_rate": 7.598514211886305e-06, "loss": 2.6445, "step": 1487 }, { "epoch": 0.24031007751937986, "grad_norm": 12.121856689453125, "learning_rate": 7.596899224806202e-06, "loss": 0.068, "step": 1488 }, { "epoch": 0.24047157622739018, "grad_norm": 42.41205978393555, "learning_rate": 7.5952842377261e-06, "loss": 1.2476, "step": 1489 }, { "epoch": 0.2406330749354005, "grad_norm": 84.05561828613281, "learning_rate": 7.5936692506459955e-06, "loss": 1.8594, "step": 1490 }, { "epoch": 0.24079457364341086, "grad_norm": 49.03755187988281, "learning_rate": 7.592054263565892e-06, "loss": 0.8057, "step": 1491 }, { "epoch": 0.2409560723514212, "grad_norm": 31.32419204711914, "learning_rate": 7.5904392764857894e-06, "loss": 0.1704, "step": 1492 }, { "epoch": 0.24111757105943152, "grad_norm": 59.3359260559082, "learning_rate": 7.588824289405685e-06, "loss": 1.0149, "step": 1493 }, { "epoch": 0.24127906976744187, "grad_norm": 32.459964752197266, "learning_rate": 7.5872093023255826e-06, "loss": 0.1963, "step": 1494 }, { "epoch": 0.2414405684754522, "grad_norm": 82.92269134521484, "learning_rate": 7.585594315245479e-06, "loss": 2.4023, "step": 1495 }, { "epoch": 0.24160206718346253, "grad_norm": 16.072635650634766, "learning_rate": 7.583979328165375e-06, "loss": 0.313, "step": 1496 }, { "epoch": 0.24176356589147288, "grad_norm": 42.90665817260742, "learning_rate": 7.582364341085272e-06, "loss": 0.9775, "step": 1497 }, { "epoch": 0.2419250645994832, "grad_norm": 38.81245040893555, "learning_rate": 7.580749354005169e-06, "loss": 0.5098, "step": 1498 }, { "epoch": 0.24208656330749354, "grad_norm": 7.880475044250488, "learning_rate": 7.579134366925065e-06, "loss": 0.1094, "step": 1499 }, { "epoch": 0.24224806201550386, "grad_norm": 73.48716735839844, "learning_rate": 7.577519379844962e-06, "loss": 1.3594, "step": 1500 }, { "epoch": 0.24240956072351422, "grad_norm": 9.310516357421875, "learning_rate": 7.575904392764859e-06, "loss": 0.097, "step": 1501 }, { "epoch": 0.24257105943152454, "grad_norm": 21.959806442260742, "learning_rate": 7.574289405684755e-06, "loss": 0.6309, "step": 1502 }, { "epoch": 0.24273255813953487, "grad_norm": 73.15780639648438, "learning_rate": 7.572674418604652e-06, "loss": 1.2695, "step": 1503 }, { "epoch": 0.24289405684754523, "grad_norm": 64.1723861694336, "learning_rate": 7.571059431524549e-06, "loss": 0.877, "step": 1504 }, { "epoch": 0.24305555555555555, "grad_norm": 48.86443328857422, "learning_rate": 7.569444444444445e-06, "loss": 0.7109, "step": 1505 }, { "epoch": 0.24321705426356588, "grad_norm": 13.258976936340332, "learning_rate": 7.567829457364342e-06, "loss": 0.2502, "step": 1506 }, { "epoch": 0.24337855297157623, "grad_norm": 66.7729263305664, "learning_rate": 7.566214470284239e-06, "loss": 0.9766, "step": 1507 }, { "epoch": 0.24354005167958656, "grad_norm": 75.30908966064453, "learning_rate": 7.564599483204134e-06, "loss": 1.7148, "step": 1508 }, { "epoch": 0.2437015503875969, "grad_norm": 50.63945770263672, "learning_rate": 7.562984496124032e-06, "loss": 1.6035, "step": 1509 }, { "epoch": 0.24386304909560724, "grad_norm": 19.811004638671875, "learning_rate": 7.561369509043928e-06, "loss": 0.146, "step": 1510 }, { "epoch": 0.24402454780361757, "grad_norm": 24.155254364013672, "learning_rate": 7.559754521963825e-06, "loss": 0.075, "step": 1511 }, { "epoch": 0.2441860465116279, "grad_norm": 203.2283172607422, "learning_rate": 7.5581395348837215e-06, "loss": 2.585, "step": 1512 }, { "epoch": 0.24434754521963825, "grad_norm": 48.83116149902344, "learning_rate": 7.556524547803619e-06, "loss": 0.3701, "step": 1513 }, { "epoch": 0.24450904392764858, "grad_norm": 122.99716186523438, "learning_rate": 7.554909560723515e-06, "loss": 2.3242, "step": 1514 }, { "epoch": 0.2446705426356589, "grad_norm": 56.44098663330078, "learning_rate": 7.553294573643411e-06, "loss": 1.0752, "step": 1515 }, { "epoch": 0.24483204134366926, "grad_norm": 44.55864715576172, "learning_rate": 7.5516795865633086e-06, "loss": 0.9697, "step": 1516 }, { "epoch": 0.2449935400516796, "grad_norm": 130.7926025390625, "learning_rate": 7.550064599483204e-06, "loss": 2.3672, "step": 1517 }, { "epoch": 0.2451550387596899, "grad_norm": 53.723026275634766, "learning_rate": 7.548449612403102e-06, "loss": 0.6992, "step": 1518 }, { "epoch": 0.24531653746770027, "grad_norm": 34.9411506652832, "learning_rate": 7.546834625322998e-06, "loss": 0.9136, "step": 1519 }, { "epoch": 0.2454780361757106, "grad_norm": 79.3532485961914, "learning_rate": 7.545219638242894e-06, "loss": 0.856, "step": 1520 }, { "epoch": 0.24563953488372092, "grad_norm": 54.11784362792969, "learning_rate": 7.543604651162791e-06, "loss": 1.1348, "step": 1521 }, { "epoch": 0.24580103359173128, "grad_norm": 22.323917388916016, "learning_rate": 7.541989664082689e-06, "loss": 0.0889, "step": 1522 }, { "epoch": 0.2459625322997416, "grad_norm": 17.2332706451416, "learning_rate": 7.5403746770025845e-06, "loss": 0.0797, "step": 1523 }, { "epoch": 0.24612403100775193, "grad_norm": 22.291173934936523, "learning_rate": 7.538759689922481e-06, "loss": 0.1387, "step": 1524 }, { "epoch": 0.24628552971576229, "grad_norm": 55.60903549194336, "learning_rate": 7.5371447028423785e-06, "loss": 1.1367, "step": 1525 }, { "epoch": 0.2464470284237726, "grad_norm": 23.123779296875, "learning_rate": 7.535529715762274e-06, "loss": 0.29, "step": 1526 }, { "epoch": 0.24660852713178294, "grad_norm": 43.292537689208984, "learning_rate": 7.533914728682172e-06, "loss": 0.634, "step": 1527 }, { "epoch": 0.2467700258397933, "grad_norm": 53.47548294067383, "learning_rate": 7.532299741602068e-06, "loss": 0.5435, "step": 1528 }, { "epoch": 0.24693152454780362, "grad_norm": 87.8198013305664, "learning_rate": 7.530684754521964e-06, "loss": 2.4004, "step": 1529 }, { "epoch": 0.24709302325581395, "grad_norm": 45.11891174316406, "learning_rate": 7.529069767441861e-06, "loss": 0.4805, "step": 1530 }, { "epoch": 0.2472545219638243, "grad_norm": 8.563511848449707, "learning_rate": 7.527454780361758e-06, "loss": 0.1465, "step": 1531 }, { "epoch": 0.24741602067183463, "grad_norm": 50.22437286376953, "learning_rate": 7.525839793281654e-06, "loss": 0.9346, "step": 1532 }, { "epoch": 0.24757751937984496, "grad_norm": 61.246456146240234, "learning_rate": 7.524224806201551e-06, "loss": 0.834, "step": 1533 }, { "epoch": 0.24773901808785528, "grad_norm": 41.899147033691406, "learning_rate": 7.522609819121448e-06, "loss": 0.7646, "step": 1534 }, { "epoch": 0.24790051679586564, "grad_norm": 41.11418533325195, "learning_rate": 7.520994832041344e-06, "loss": 0.5332, "step": 1535 }, { "epoch": 0.24806201550387597, "grad_norm": 101.59305572509766, "learning_rate": 7.519379844961241e-06, "loss": 2.9609, "step": 1536 }, { "epoch": 0.2482235142118863, "grad_norm": 23.62383460998535, "learning_rate": 7.517764857881138e-06, "loss": 0.209, "step": 1537 }, { "epoch": 0.24838501291989665, "grad_norm": 110.04724884033203, "learning_rate": 7.516149870801034e-06, "loss": 2.3438, "step": 1538 }, { "epoch": 0.24854651162790697, "grad_norm": 64.42703247070312, "learning_rate": 7.514534883720931e-06, "loss": 1.2734, "step": 1539 }, { "epoch": 0.2487080103359173, "grad_norm": 30.895784378051758, "learning_rate": 7.512919896640828e-06, "loss": 0.1846, "step": 1540 }, { "epoch": 0.24886950904392766, "grad_norm": 49.11211013793945, "learning_rate": 7.511304909560723e-06, "loss": 0.873, "step": 1541 }, { "epoch": 0.24903100775193798, "grad_norm": 24.981231689453125, "learning_rate": 7.509689922480621e-06, "loss": 0.6437, "step": 1542 }, { "epoch": 0.2491925064599483, "grad_norm": 40.55112838745117, "learning_rate": 7.508074935400517e-06, "loss": 0.9258, "step": 1543 }, { "epoch": 0.24935400516795866, "grad_norm": 41.29207229614258, "learning_rate": 7.506459948320414e-06, "loss": 0.8291, "step": 1544 }, { "epoch": 0.249515503875969, "grad_norm": 45.55941390991211, "learning_rate": 7.5048449612403105e-06, "loss": 0.5098, "step": 1545 }, { "epoch": 0.24967700258397932, "grad_norm": 74.80687713623047, "learning_rate": 7.503229974160208e-06, "loss": 1.3711, "step": 1546 }, { "epoch": 0.24983850129198967, "grad_norm": 109.1178970336914, "learning_rate": 7.501614987080104e-06, "loss": 3.3711, "step": 1547 }, { "epoch": 0.25, "grad_norm": 53.844482421875, "learning_rate": 7.500000000000001e-06, "loss": 0.6396, "step": 1548 }, { "epoch": 0.25016149870801035, "grad_norm": 97.71902465820312, "learning_rate": 7.498385012919898e-06, "loss": 2.0742, "step": 1549 }, { "epoch": 0.25032299741602065, "grad_norm": 14.075937271118164, "learning_rate": 7.496770025839793e-06, "loss": 0.2124, "step": 1550 }, { "epoch": 0.250484496124031, "grad_norm": 48.60476303100586, "learning_rate": 7.495155038759691e-06, "loss": 1.1836, "step": 1551 }, { "epoch": 0.25064599483204136, "grad_norm": 42.08123779296875, "learning_rate": 7.493540051679587e-06, "loss": 0.7002, "step": 1552 }, { "epoch": 0.25080749354005166, "grad_norm": 76.89168548583984, "learning_rate": 7.491925064599484e-06, "loss": 1.0122, "step": 1553 }, { "epoch": 0.250968992248062, "grad_norm": 26.936723709106445, "learning_rate": 7.49031007751938e-06, "loss": 0.2916, "step": 1554 }, { "epoch": 0.25113049095607237, "grad_norm": 42.392578125, "learning_rate": 7.488695090439278e-06, "loss": 1.1992, "step": 1555 }, { "epoch": 0.25129198966408267, "grad_norm": 22.11261558532715, "learning_rate": 7.4870801033591735e-06, "loss": 0.2158, "step": 1556 }, { "epoch": 0.251453488372093, "grad_norm": 37.34724426269531, "learning_rate": 7.48546511627907e-06, "loss": 0.8291, "step": 1557 }, { "epoch": 0.2516149870801034, "grad_norm": 70.635498046875, "learning_rate": 7.4838501291989675e-06, "loss": 1.3354, "step": 1558 }, { "epoch": 0.2517764857881137, "grad_norm": 36.09482955932617, "learning_rate": 7.482235142118863e-06, "loss": 0.5273, "step": 1559 }, { "epoch": 0.25193798449612403, "grad_norm": 10.163454055786133, "learning_rate": 7.480620155038761e-06, "loss": 0.0791, "step": 1560 }, { "epoch": 0.2520994832041344, "grad_norm": 73.91993713378906, "learning_rate": 7.479005167958657e-06, "loss": 1.5703, "step": 1561 }, { "epoch": 0.2522609819121447, "grad_norm": 21.709627151489258, "learning_rate": 7.477390180878553e-06, "loss": 0.3152, "step": 1562 }, { "epoch": 0.25242248062015504, "grad_norm": 18.47602653503418, "learning_rate": 7.47577519379845e-06, "loss": 0.1882, "step": 1563 }, { "epoch": 0.2525839793281654, "grad_norm": 46.4912223815918, "learning_rate": 7.474160206718347e-06, "loss": 0.4951, "step": 1564 }, { "epoch": 0.2527454780361757, "grad_norm": 33.21100616455078, "learning_rate": 7.472545219638243e-06, "loss": 0.2371, "step": 1565 }, { "epoch": 0.25290697674418605, "grad_norm": 82.29865264892578, "learning_rate": 7.47093023255814e-06, "loss": 1.0938, "step": 1566 }, { "epoch": 0.2530684754521964, "grad_norm": 37.906349182128906, "learning_rate": 7.469315245478037e-06, "loss": 0.9419, "step": 1567 }, { "epoch": 0.2532299741602067, "grad_norm": 34.43241882324219, "learning_rate": 7.467700258397933e-06, "loss": 0.6416, "step": 1568 }, { "epoch": 0.25339147286821706, "grad_norm": 67.82628631591797, "learning_rate": 7.46608527131783e-06, "loss": 1.0586, "step": 1569 }, { "epoch": 0.2535529715762274, "grad_norm": 36.665733337402344, "learning_rate": 7.464470284237727e-06, "loss": 0.498, "step": 1570 }, { "epoch": 0.2537144702842377, "grad_norm": 47.601924896240234, "learning_rate": 7.462855297157623e-06, "loss": 0.6621, "step": 1571 }, { "epoch": 0.25387596899224807, "grad_norm": 25.646297454833984, "learning_rate": 7.46124031007752e-06, "loss": 0.5405, "step": 1572 }, { "epoch": 0.2540374677002584, "grad_norm": 51.51289367675781, "learning_rate": 7.459625322997417e-06, "loss": 1.4937, "step": 1573 }, { "epoch": 0.2541989664082687, "grad_norm": 4.793233871459961, "learning_rate": 7.4580103359173124e-06, "loss": 0.049, "step": 1574 }, { "epoch": 0.2543604651162791, "grad_norm": 106.8423843383789, "learning_rate": 7.45639534883721e-06, "loss": 1.7612, "step": 1575 }, { "epoch": 0.25452196382428943, "grad_norm": 36.61439895629883, "learning_rate": 7.454780361757107e-06, "loss": 1.0176, "step": 1576 }, { "epoch": 0.25468346253229973, "grad_norm": 72.64775085449219, "learning_rate": 7.453165374677003e-06, "loss": 1.6777, "step": 1577 }, { "epoch": 0.2548449612403101, "grad_norm": 19.87989044189453, "learning_rate": 7.4515503875968995e-06, "loss": 0.1223, "step": 1578 }, { "epoch": 0.25500645994832044, "grad_norm": 20.288211822509766, "learning_rate": 7.449935400516797e-06, "loss": 0.386, "step": 1579 }, { "epoch": 0.25516795865633074, "grad_norm": 92.08747100830078, "learning_rate": 7.448320413436693e-06, "loss": 1.6602, "step": 1580 }, { "epoch": 0.2553294573643411, "grad_norm": 73.29782104492188, "learning_rate": 7.44670542635659e-06, "loss": 0.5427, "step": 1581 }, { "epoch": 0.2554909560723514, "grad_norm": 45.2894401550293, "learning_rate": 7.445090439276487e-06, "loss": 0.9395, "step": 1582 }, { "epoch": 0.25565245478036175, "grad_norm": 21.001693725585938, "learning_rate": 7.443475452196382e-06, "loss": 0.3516, "step": 1583 }, { "epoch": 0.2558139534883721, "grad_norm": 30.446367263793945, "learning_rate": 7.44186046511628e-06, "loss": 0.5879, "step": 1584 }, { "epoch": 0.2559754521963824, "grad_norm": 19.628450393676758, "learning_rate": 7.440245478036176e-06, "loss": 0.3184, "step": 1585 }, { "epoch": 0.25613695090439276, "grad_norm": 40.857547760009766, "learning_rate": 7.438630490956073e-06, "loss": 0.5269, "step": 1586 }, { "epoch": 0.2562984496124031, "grad_norm": 42.718048095703125, "learning_rate": 7.437015503875969e-06, "loss": 0.6221, "step": 1587 }, { "epoch": 0.2564599483204134, "grad_norm": 110.18079376220703, "learning_rate": 7.435400516795867e-06, "loss": 3.625, "step": 1588 }, { "epoch": 0.25662144702842377, "grad_norm": 71.64636993408203, "learning_rate": 7.4337855297157625e-06, "loss": 1.1914, "step": 1589 }, { "epoch": 0.2567829457364341, "grad_norm": 156.99021911621094, "learning_rate": 7.432170542635659e-06, "loss": 2.8779, "step": 1590 }, { "epoch": 0.2569444444444444, "grad_norm": 32.75139236450195, "learning_rate": 7.4305555555555565e-06, "loss": 0.2522, "step": 1591 }, { "epoch": 0.2571059431524548, "grad_norm": 28.629714965820312, "learning_rate": 7.428940568475452e-06, "loss": 0.4541, "step": 1592 }, { "epoch": 0.25726744186046513, "grad_norm": 106.71478271484375, "learning_rate": 7.42732558139535e-06, "loss": 1.6348, "step": 1593 }, { "epoch": 0.25742894056847543, "grad_norm": 47.931644439697266, "learning_rate": 7.425710594315246e-06, "loss": 1.3379, "step": 1594 }, { "epoch": 0.2575904392764858, "grad_norm": 230.74879455566406, "learning_rate": 7.424095607235142e-06, "loss": 2.1705, "step": 1595 }, { "epoch": 0.25775193798449614, "grad_norm": 96.60572052001953, "learning_rate": 7.422480620155039e-06, "loss": 2.2949, "step": 1596 }, { "epoch": 0.25791343669250644, "grad_norm": 92.2607192993164, "learning_rate": 7.420865633074936e-06, "loss": 1.6367, "step": 1597 }, { "epoch": 0.2580749354005168, "grad_norm": 24.63920021057129, "learning_rate": 7.4192506459948324e-06, "loss": 0.6182, "step": 1598 }, { "epoch": 0.25823643410852715, "grad_norm": 49.558414459228516, "learning_rate": 7.417635658914729e-06, "loss": 0.6221, "step": 1599 }, { "epoch": 0.25839793281653745, "grad_norm": 54.05058670043945, "learning_rate": 7.416020671834626e-06, "loss": 1.1484, "step": 1600 }, { "epoch": 0.2585594315245478, "grad_norm": 46.72705841064453, "learning_rate": 7.414405684754522e-06, "loss": 0.7769, "step": 1601 }, { "epoch": 0.25872093023255816, "grad_norm": 95.9456558227539, "learning_rate": 7.412790697674419e-06, "loss": 1.6602, "step": 1602 }, { "epoch": 0.25888242894056845, "grad_norm": 55.71812438964844, "learning_rate": 7.411175710594316e-06, "loss": 1.3516, "step": 1603 }, { "epoch": 0.2590439276485788, "grad_norm": 23.909238815307617, "learning_rate": 7.409560723514212e-06, "loss": 0.416, "step": 1604 }, { "epoch": 0.25920542635658916, "grad_norm": 16.71084976196289, "learning_rate": 7.407945736434109e-06, "loss": 0.4756, "step": 1605 }, { "epoch": 0.25936692506459946, "grad_norm": 6.303974151611328, "learning_rate": 7.406330749354006e-06, "loss": 0.0886, "step": 1606 }, { "epoch": 0.2595284237726098, "grad_norm": 22.24675750732422, "learning_rate": 7.4047157622739015e-06, "loss": 0.2725, "step": 1607 }, { "epoch": 0.2596899224806202, "grad_norm": 54.61298751831055, "learning_rate": 7.403100775193799e-06, "loss": 0.6367, "step": 1608 }, { "epoch": 0.25985142118863047, "grad_norm": 44.91133117675781, "learning_rate": 7.401485788113696e-06, "loss": 0.4402, "step": 1609 }, { "epoch": 0.2600129198966408, "grad_norm": 79.78364562988281, "learning_rate": 7.399870801033592e-06, "loss": 2.2344, "step": 1610 }, { "epoch": 0.2601744186046512, "grad_norm": 92.68379974365234, "learning_rate": 7.3982558139534886e-06, "loss": 1.8333, "step": 1611 }, { "epoch": 0.2603359173126615, "grad_norm": 92.49256896972656, "learning_rate": 7.396640826873386e-06, "loss": 1.3333, "step": 1612 }, { "epoch": 0.26049741602067183, "grad_norm": 17.898632049560547, "learning_rate": 7.3950258397932825e-06, "loss": 0.1268, "step": 1613 }, { "epoch": 0.2606589147286822, "grad_norm": 27.523426055908203, "learning_rate": 7.393410852713179e-06, "loss": 0.3208, "step": 1614 }, { "epoch": 0.2608204134366925, "grad_norm": 61.259395599365234, "learning_rate": 7.391795865633076e-06, "loss": 0.7373, "step": 1615 }, { "epoch": 0.26098191214470284, "grad_norm": 75.45567321777344, "learning_rate": 7.390180878552973e-06, "loss": 1.8242, "step": 1616 }, { "epoch": 0.2611434108527132, "grad_norm": 19.548673629760742, "learning_rate": 7.388565891472869e-06, "loss": 0.3806, "step": 1617 }, { "epoch": 0.2613049095607235, "grad_norm": 27.8178768157959, "learning_rate": 7.386950904392765e-06, "loss": 0.657, "step": 1618 }, { "epoch": 0.26146640826873385, "grad_norm": 46.203285217285156, "learning_rate": 7.385335917312663e-06, "loss": 0.4414, "step": 1619 }, { "epoch": 0.2616279069767442, "grad_norm": 88.16699981689453, "learning_rate": 7.3837209302325584e-06, "loss": 1.8555, "step": 1620 }, { "epoch": 0.2617894056847545, "grad_norm": 35.555110931396484, "learning_rate": 7.382105943152456e-06, "loss": 1.1296, "step": 1621 }, { "epoch": 0.26195090439276486, "grad_norm": 53.27534484863281, "learning_rate": 7.380490956072352e-06, "loss": 0.7324, "step": 1622 }, { "epoch": 0.2621124031007752, "grad_norm": 24.93277359008789, "learning_rate": 7.378875968992248e-06, "loss": 0.3948, "step": 1623 }, { "epoch": 0.2622739018087855, "grad_norm": 32.98225784301758, "learning_rate": 7.3772609819121455e-06, "loss": 0.4944, "step": 1624 }, { "epoch": 0.26243540051679587, "grad_norm": 122.42716979980469, "learning_rate": 7.375645994832042e-06, "loss": 1.9854, "step": 1625 }, { "epoch": 0.2625968992248062, "grad_norm": 7.843042373657227, "learning_rate": 7.374031007751939e-06, "loss": 0.0404, "step": 1626 }, { "epoch": 0.2627583979328165, "grad_norm": 48.16130828857422, "learning_rate": 7.372416020671835e-06, "loss": 1.03, "step": 1627 }, { "epoch": 0.2629198966408269, "grad_norm": 44.1708869934082, "learning_rate": 7.370801033591733e-06, "loss": 0.3877, "step": 1628 }, { "epoch": 0.26308139534883723, "grad_norm": 59.45792770385742, "learning_rate": 7.369186046511628e-06, "loss": 0.6465, "step": 1629 }, { "epoch": 0.26324289405684753, "grad_norm": 10.161492347717285, "learning_rate": 7.367571059431525e-06, "loss": 0.0312, "step": 1630 }, { "epoch": 0.2634043927648579, "grad_norm": 126.29415893554688, "learning_rate": 7.365956072351422e-06, "loss": 1.7798, "step": 1631 }, { "epoch": 0.26356589147286824, "grad_norm": 70.10379791259766, "learning_rate": 7.364341085271318e-06, "loss": 1.0923, "step": 1632 }, { "epoch": 0.26372739018087854, "grad_norm": 40.291297912597656, "learning_rate": 7.362726098191215e-06, "loss": 0.4551, "step": 1633 }, { "epoch": 0.2638888888888889, "grad_norm": 86.81749725341797, "learning_rate": 7.361111111111112e-06, "loss": 2.7734, "step": 1634 }, { "epoch": 0.26405038759689925, "grad_norm": 19.33092498779297, "learning_rate": 7.359496124031008e-06, "loss": 0.1367, "step": 1635 }, { "epoch": 0.26421188630490955, "grad_norm": 43.10274887084961, "learning_rate": 7.357881136950905e-06, "loss": 0.6895, "step": 1636 }, { "epoch": 0.2643733850129199, "grad_norm": 72.59715270996094, "learning_rate": 7.3562661498708025e-06, "loss": 1.0977, "step": 1637 }, { "epoch": 0.26453488372093026, "grad_norm": 25.7235050201416, "learning_rate": 7.354651162790698e-06, "loss": 0.0994, "step": 1638 }, { "epoch": 0.26469638242894056, "grad_norm": 26.17191505432129, "learning_rate": 7.353036175710595e-06, "loss": 0.2302, "step": 1639 }, { "epoch": 0.2648578811369509, "grad_norm": 86.33165740966797, "learning_rate": 7.351421188630492e-06, "loss": 1.4399, "step": 1640 }, { "epoch": 0.26501937984496127, "grad_norm": 15.23220443725586, "learning_rate": 7.349806201550388e-06, "loss": 0.0509, "step": 1641 }, { "epoch": 0.26518087855297157, "grad_norm": 67.87796020507812, "learning_rate": 7.348191214470285e-06, "loss": 1.6836, "step": 1642 }, { "epoch": 0.2653423772609819, "grad_norm": 51.76834487915039, "learning_rate": 7.346576227390182e-06, "loss": 1.0518, "step": 1643 }, { "epoch": 0.2655038759689923, "grad_norm": 5.578802585601807, "learning_rate": 7.3449612403100776e-06, "loss": 0.0363, "step": 1644 }, { "epoch": 0.2656653746770026, "grad_norm": 124.83919525146484, "learning_rate": 7.343346253229975e-06, "loss": 1.8057, "step": 1645 }, { "epoch": 0.26582687338501293, "grad_norm": 27.389787673950195, "learning_rate": 7.3417312661498715e-06, "loss": 0.748, "step": 1646 }, { "epoch": 0.26598837209302323, "grad_norm": 110.11502838134766, "learning_rate": 7.340116279069768e-06, "loss": 1.7188, "step": 1647 }, { "epoch": 0.2661498708010336, "grad_norm": 40.96629333496094, "learning_rate": 7.338501291989665e-06, "loss": 0.9666, "step": 1648 }, { "epoch": 0.26631136950904394, "grad_norm": 44.2090950012207, "learning_rate": 7.336886304909562e-06, "loss": 1.1617, "step": 1649 }, { "epoch": 0.26647286821705424, "grad_norm": 52.914371490478516, "learning_rate": 7.335271317829458e-06, "loss": 0.6005, "step": 1650 }, { "epoch": 0.2666343669250646, "grad_norm": 58.14436721801758, "learning_rate": 7.333656330749354e-06, "loss": 1.502, "step": 1651 }, { "epoch": 0.26679586563307495, "grad_norm": 55.496429443359375, "learning_rate": 7.332041343669252e-06, "loss": 0.8643, "step": 1652 }, { "epoch": 0.26695736434108525, "grad_norm": 38.13867950439453, "learning_rate": 7.3304263565891475e-06, "loss": 0.5596, "step": 1653 }, { "epoch": 0.2671188630490956, "grad_norm": 53.62562561035156, "learning_rate": 7.328811369509045e-06, "loss": 1.0781, "step": 1654 }, { "epoch": 0.26728036175710596, "grad_norm": 9.78215503692627, "learning_rate": 7.3271963824289414e-06, "loss": 0.1377, "step": 1655 }, { "epoch": 0.26744186046511625, "grad_norm": 113.40748596191406, "learning_rate": 7.325581395348837e-06, "loss": 1.9977, "step": 1656 }, { "epoch": 0.2676033591731266, "grad_norm": 104.40174102783203, "learning_rate": 7.3239664082687345e-06, "loss": 3.8579, "step": 1657 }, { "epoch": 0.26776485788113696, "grad_norm": 37.04096984863281, "learning_rate": 7.322351421188631e-06, "loss": 0.6406, "step": 1658 }, { "epoch": 0.26792635658914726, "grad_norm": 169.8207550048828, "learning_rate": 7.320736434108528e-06, "loss": 2.9277, "step": 1659 }, { "epoch": 0.2680878552971576, "grad_norm": 78.29705810546875, "learning_rate": 7.319121447028424e-06, "loss": 0.8721, "step": 1660 }, { "epoch": 0.268249354005168, "grad_norm": 48.73546600341797, "learning_rate": 7.317506459948322e-06, "loss": 0.4551, "step": 1661 }, { "epoch": 0.26841085271317827, "grad_norm": 69.51815032958984, "learning_rate": 7.315891472868217e-06, "loss": 1.1611, "step": 1662 }, { "epoch": 0.2685723514211886, "grad_norm": 38.30005645751953, "learning_rate": 7.314276485788115e-06, "loss": 0.4834, "step": 1663 }, { "epoch": 0.268733850129199, "grad_norm": 41.19718551635742, "learning_rate": 7.312661498708011e-06, "loss": 1.1631, "step": 1664 }, { "epoch": 0.2688953488372093, "grad_norm": 53.0683479309082, "learning_rate": 7.311046511627907e-06, "loss": 1.2148, "step": 1665 }, { "epoch": 0.26905684754521964, "grad_norm": 45.367218017578125, "learning_rate": 7.3094315245478044e-06, "loss": 0.9639, "step": 1666 }, { "epoch": 0.26921834625323, "grad_norm": 110.08916473388672, "learning_rate": 7.307816537467701e-06, "loss": 2.2109, "step": 1667 }, { "epoch": 0.2693798449612403, "grad_norm": 21.040565490722656, "learning_rate": 7.3062015503875976e-06, "loss": 0.1982, "step": 1668 }, { "epoch": 0.26954134366925064, "grad_norm": 57.54045867919922, "learning_rate": 7.304586563307494e-06, "loss": 1.3672, "step": 1669 }, { "epoch": 0.269702842377261, "grad_norm": 14.467737197875977, "learning_rate": 7.3029715762273915e-06, "loss": 0.1118, "step": 1670 }, { "epoch": 0.2698643410852713, "grad_norm": 84.63352966308594, "learning_rate": 7.301356589147287e-06, "loss": 1.6133, "step": 1671 }, { "epoch": 0.27002583979328165, "grad_norm": 43.8784294128418, "learning_rate": 7.299741602067184e-06, "loss": 0.4282, "step": 1672 }, { "epoch": 0.270187338501292, "grad_norm": 40.24197769165039, "learning_rate": 7.298126614987081e-06, "loss": 0.7939, "step": 1673 }, { "epoch": 0.2703488372093023, "grad_norm": 39.14079666137695, "learning_rate": 7.296511627906977e-06, "loss": 0.5371, "step": 1674 }, { "epoch": 0.27051033591731266, "grad_norm": 36.13401412963867, "learning_rate": 7.294896640826874e-06, "loss": 0.4521, "step": 1675 }, { "epoch": 0.270671834625323, "grad_norm": 4.803767204284668, "learning_rate": 7.293281653746771e-06, "loss": 0.0408, "step": 1676 }, { "epoch": 0.2708333333333333, "grad_norm": 88.42391204833984, "learning_rate": 7.291666666666667e-06, "loss": 1.7653, "step": 1677 }, { "epoch": 0.27099483204134367, "grad_norm": 54.4057502746582, "learning_rate": 7.290051679586564e-06, "loss": 1.1191, "step": 1678 }, { "epoch": 0.271156330749354, "grad_norm": 88.80107116699219, "learning_rate": 7.2884366925064606e-06, "loss": 2.6003, "step": 1679 }, { "epoch": 0.2713178294573643, "grad_norm": 65.08838653564453, "learning_rate": 7.286821705426357e-06, "loss": 1.4531, "step": 1680 }, { "epoch": 0.2714793281653747, "grad_norm": 39.006378173828125, "learning_rate": 7.285206718346254e-06, "loss": 0.5615, "step": 1681 }, { "epoch": 0.27164082687338503, "grad_norm": 39.737491607666016, "learning_rate": 7.283591731266151e-06, "loss": 1.0684, "step": 1682 }, { "epoch": 0.27180232558139533, "grad_norm": 26.86434555053711, "learning_rate": 7.281976744186047e-06, "loss": 0.2356, "step": 1683 }, { "epoch": 0.2719638242894057, "grad_norm": 15.335992813110352, "learning_rate": 7.280361757105943e-06, "loss": 0.0955, "step": 1684 }, { "epoch": 0.27212532299741604, "grad_norm": 250.72808837890625, "learning_rate": 7.278746770025841e-06, "loss": 3.6094, "step": 1685 }, { "epoch": 0.27228682170542634, "grad_norm": 56.65821838378906, "learning_rate": 7.2771317829457365e-06, "loss": 1.097, "step": 1686 }, { "epoch": 0.2724483204134367, "grad_norm": 101.79039001464844, "learning_rate": 7.275516795865634e-06, "loss": 1.625, "step": 1687 }, { "epoch": 0.27260981912144705, "grad_norm": 15.330724716186523, "learning_rate": 7.2739018087855304e-06, "loss": 0.0657, "step": 1688 }, { "epoch": 0.27277131782945735, "grad_norm": 22.942428588867188, "learning_rate": 7.272286821705426e-06, "loss": 0.1348, "step": 1689 }, { "epoch": 0.2729328165374677, "grad_norm": 43.707096099853516, "learning_rate": 7.2706718346253236e-06, "loss": 0.873, "step": 1690 }, { "epoch": 0.27309431524547806, "grad_norm": 66.86373138427734, "learning_rate": 7.269056847545221e-06, "loss": 0.7487, "step": 1691 }, { "epoch": 0.27325581395348836, "grad_norm": 58.27635192871094, "learning_rate": 7.267441860465117e-06, "loss": 0.979, "step": 1692 }, { "epoch": 0.2734173126614987, "grad_norm": 118.57364654541016, "learning_rate": 7.265826873385013e-06, "loss": 1.1095, "step": 1693 }, { "epoch": 0.27357881136950907, "grad_norm": 76.18675231933594, "learning_rate": 7.264211886304911e-06, "loss": 0.8359, "step": 1694 }, { "epoch": 0.27374031007751937, "grad_norm": 120.90357971191406, "learning_rate": 7.262596899224806e-06, "loss": 2.1621, "step": 1695 }, { "epoch": 0.2739018087855297, "grad_norm": 26.37372589111328, "learning_rate": 7.260981912144704e-06, "loss": 0.554, "step": 1696 }, { "epoch": 0.2740633074935401, "grad_norm": 63.16633987426758, "learning_rate": 7.2593669250646e-06, "loss": 1.582, "step": 1697 }, { "epoch": 0.2742248062015504, "grad_norm": 24.600250244140625, "learning_rate": 7.257751937984496e-06, "loss": 0.076, "step": 1698 }, { "epoch": 0.27438630490956073, "grad_norm": 108.1574478149414, "learning_rate": 7.2561369509043935e-06, "loss": 3.4375, "step": 1699 }, { "epoch": 0.2745478036175711, "grad_norm": 87.62215423583984, "learning_rate": 7.25452196382429e-06, "loss": 0.9341, "step": 1700 }, { "epoch": 0.2747093023255814, "grad_norm": 33.97563552856445, "learning_rate": 7.2529069767441866e-06, "loss": 0.5264, "step": 1701 }, { "epoch": 0.27487080103359174, "grad_norm": 30.105792999267578, "learning_rate": 7.251291989664083e-06, "loss": 0.5088, "step": 1702 }, { "epoch": 0.2750322997416021, "grad_norm": 29.19022560119629, "learning_rate": 7.2496770025839805e-06, "loss": 0.4824, "step": 1703 }, { "epoch": 0.2751937984496124, "grad_norm": 35.73456954956055, "learning_rate": 7.248062015503876e-06, "loss": 0.6877, "step": 1704 }, { "epoch": 0.27535529715762275, "grad_norm": 36.49464797973633, "learning_rate": 7.246447028423773e-06, "loss": 0.2603, "step": 1705 }, { "epoch": 0.2755167958656331, "grad_norm": 109.6068115234375, "learning_rate": 7.24483204134367e-06, "loss": 2.4082, "step": 1706 }, { "epoch": 0.2756782945736434, "grad_norm": 68.66720581054688, "learning_rate": 7.243217054263566e-06, "loss": 1.5142, "step": 1707 }, { "epoch": 0.27583979328165376, "grad_norm": 61.994083404541016, "learning_rate": 7.241602067183463e-06, "loss": 0.8242, "step": 1708 }, { "epoch": 0.2760012919896641, "grad_norm": 76.85298919677734, "learning_rate": 7.23998708010336e-06, "loss": 1.772, "step": 1709 }, { "epoch": 0.2761627906976744, "grad_norm": 76.50523376464844, "learning_rate": 7.238372093023256e-06, "loss": 1.6602, "step": 1710 }, { "epoch": 0.27632428940568476, "grad_norm": 48.73271560668945, "learning_rate": 7.236757105943153e-06, "loss": 0.5776, "step": 1711 }, { "epoch": 0.27648578811369506, "grad_norm": 64.90419006347656, "learning_rate": 7.23514211886305e-06, "loss": 1.2285, "step": 1712 }, { "epoch": 0.2766472868217054, "grad_norm": 84.64593505859375, "learning_rate": 7.233527131782946e-06, "loss": 2.9841, "step": 1713 }, { "epoch": 0.2768087855297158, "grad_norm": 14.720380783081055, "learning_rate": 7.231912144702843e-06, "loss": 0.1865, "step": 1714 }, { "epoch": 0.27697028423772607, "grad_norm": 26.56454086303711, "learning_rate": 7.23029715762274e-06, "loss": 0.2402, "step": 1715 }, { "epoch": 0.2771317829457364, "grad_norm": 61.47307586669922, "learning_rate": 7.228682170542636e-06, "loss": 1.5312, "step": 1716 }, { "epoch": 0.2772932816537468, "grad_norm": 55.85887145996094, "learning_rate": 7.227067183462532e-06, "loss": 1.0605, "step": 1717 }, { "epoch": 0.2774547803617571, "grad_norm": 45.96915817260742, "learning_rate": 7.22545219638243e-06, "loss": 1.2461, "step": 1718 }, { "epoch": 0.27761627906976744, "grad_norm": 30.662334442138672, "learning_rate": 7.2238372093023255e-06, "loss": 0.1831, "step": 1719 }, { "epoch": 0.2777777777777778, "grad_norm": 26.42164421081543, "learning_rate": 7.222222222222223e-06, "loss": 0.4541, "step": 1720 }, { "epoch": 0.2779392764857881, "grad_norm": 26.208337783813477, "learning_rate": 7.2206072351421195e-06, "loss": 0.4893, "step": 1721 }, { "epoch": 0.27810077519379844, "grad_norm": 136.0435791015625, "learning_rate": 7.218992248062015e-06, "loss": 2.5625, "step": 1722 }, { "epoch": 0.2782622739018088, "grad_norm": 34.085174560546875, "learning_rate": 7.217377260981913e-06, "loss": 0.2, "step": 1723 }, { "epoch": 0.2784237726098191, "grad_norm": 90.33343505859375, "learning_rate": 7.21576227390181e-06, "loss": 0.9453, "step": 1724 }, { "epoch": 0.27858527131782945, "grad_norm": 66.16089630126953, "learning_rate": 7.214147286821706e-06, "loss": 0.9141, "step": 1725 }, { "epoch": 0.2787467700258398, "grad_norm": 100.14765167236328, "learning_rate": 7.212532299741602e-06, "loss": 1.418, "step": 1726 }, { "epoch": 0.2789082687338501, "grad_norm": 80.2387466430664, "learning_rate": 7.2109173126615e-06, "loss": 1.5859, "step": 1727 }, { "epoch": 0.27906976744186046, "grad_norm": 20.431764602661133, "learning_rate": 7.209302325581395e-06, "loss": 0.2012, "step": 1728 }, { "epoch": 0.2792312661498708, "grad_norm": 79.5293960571289, "learning_rate": 7.207687338501293e-06, "loss": 0.9714, "step": 1729 }, { "epoch": 0.2793927648578811, "grad_norm": 76.45832824707031, "learning_rate": 7.206072351421189e-06, "loss": 1.625, "step": 1730 }, { "epoch": 0.27955426356589147, "grad_norm": 45.6080322265625, "learning_rate": 7.204457364341085e-06, "loss": 0.6465, "step": 1731 }, { "epoch": 0.2797157622739018, "grad_norm": 91.06903839111328, "learning_rate": 7.2028423772609825e-06, "loss": 1.1133, "step": 1732 }, { "epoch": 0.2798772609819121, "grad_norm": 45.83683395385742, "learning_rate": 7.201227390180879e-06, "loss": 0.3945, "step": 1733 }, { "epoch": 0.2800387596899225, "grad_norm": 47.21257781982422, "learning_rate": 7.199612403100776e-06, "loss": 0.959, "step": 1734 }, { "epoch": 0.28020025839793283, "grad_norm": 77.15673828125, "learning_rate": 7.197997416020672e-06, "loss": 2.168, "step": 1735 }, { "epoch": 0.28036175710594313, "grad_norm": 65.50760650634766, "learning_rate": 7.1963824289405696e-06, "loss": 1.8291, "step": 1736 }, { "epoch": 0.2805232558139535, "grad_norm": 62.37920379638672, "learning_rate": 7.194767441860465e-06, "loss": 1.5059, "step": 1737 }, { "epoch": 0.28068475452196384, "grad_norm": 23.949031829833984, "learning_rate": 7.193152454780362e-06, "loss": 0.3586, "step": 1738 }, { "epoch": 0.28084625322997414, "grad_norm": 104.66152954101562, "learning_rate": 7.191537467700259e-06, "loss": 2.6094, "step": 1739 }, { "epoch": 0.2810077519379845, "grad_norm": 28.66046905517578, "learning_rate": 7.189922480620155e-06, "loss": 0.3657, "step": 1740 }, { "epoch": 0.28116925064599485, "grad_norm": 25.560867309570312, "learning_rate": 7.188307493540052e-06, "loss": 0.3359, "step": 1741 }, { "epoch": 0.28133074935400515, "grad_norm": 31.785097122192383, "learning_rate": 7.186692506459949e-06, "loss": 0.4834, "step": 1742 }, { "epoch": 0.2814922480620155, "grad_norm": 95.06989288330078, "learning_rate": 7.185077519379846e-06, "loss": 1.478, "step": 1743 }, { "epoch": 0.28165374677002586, "grad_norm": 39.211036682128906, "learning_rate": 7.183462532299742e-06, "loss": 1.0576, "step": 1744 }, { "epoch": 0.28181524547803616, "grad_norm": 107.34898376464844, "learning_rate": 7.181847545219639e-06, "loss": 1.9609, "step": 1745 }, { "epoch": 0.2819767441860465, "grad_norm": 85.26769256591797, "learning_rate": 7.180232558139536e-06, "loss": 1.3994, "step": 1746 }, { "epoch": 0.28213824289405687, "grad_norm": 14.131573677062988, "learning_rate": 7.178617571059432e-06, "loss": 0.1309, "step": 1747 }, { "epoch": 0.28229974160206717, "grad_norm": 66.3228530883789, "learning_rate": 7.177002583979329e-06, "loss": 0.714, "step": 1748 }, { "epoch": 0.2824612403100775, "grad_norm": 15.563180923461914, "learning_rate": 7.175387596899226e-06, "loss": 0.1587, "step": 1749 }, { "epoch": 0.2826227390180879, "grad_norm": 53.19793701171875, "learning_rate": 7.173772609819122e-06, "loss": 0.9863, "step": 1750 }, { "epoch": 0.2827842377260982, "grad_norm": 40.40842056274414, "learning_rate": 7.172157622739019e-06, "loss": 1.2006, "step": 1751 }, { "epoch": 0.28294573643410853, "grad_norm": 7.892530918121338, "learning_rate": 7.170542635658916e-06, "loss": 0.0741, "step": 1752 }, { "epoch": 0.2831072351421189, "grad_norm": 95.11578369140625, "learning_rate": 7.168927648578812e-06, "loss": 1.5791, "step": 1753 }, { "epoch": 0.2832687338501292, "grad_norm": 119.7568359375, "learning_rate": 7.1673126614987085e-06, "loss": 3.0, "step": 1754 }, { "epoch": 0.28343023255813954, "grad_norm": 22.4869327545166, "learning_rate": 7.165697674418606e-06, "loss": 0.1785, "step": 1755 }, { "epoch": 0.2835917312661499, "grad_norm": 30.10750961303711, "learning_rate": 7.164082687338502e-06, "loss": 0.7705, "step": 1756 }, { "epoch": 0.2837532299741602, "grad_norm": 39.708953857421875, "learning_rate": 7.162467700258399e-06, "loss": 1.002, "step": 1757 }, { "epoch": 0.28391472868217055, "grad_norm": 76.10417938232422, "learning_rate": 7.1608527131782956e-06, "loss": 1.3398, "step": 1758 }, { "epoch": 0.2840762273901809, "grad_norm": 18.772615432739258, "learning_rate": 7.159237726098191e-06, "loss": 0.1945, "step": 1759 }, { "epoch": 0.2842377260981912, "grad_norm": 104.90581512451172, "learning_rate": 7.157622739018089e-06, "loss": 1.3945, "step": 1760 }, { "epoch": 0.28439922480620156, "grad_norm": 86.28683471679688, "learning_rate": 7.156007751937985e-06, "loss": 1.7188, "step": 1761 }, { "epoch": 0.2845607235142119, "grad_norm": 71.64920806884766, "learning_rate": 7.154392764857882e-06, "loss": 1.1445, "step": 1762 }, { "epoch": 0.2847222222222222, "grad_norm": 55.64582824707031, "learning_rate": 7.152777777777778e-06, "loss": 1.2051, "step": 1763 }, { "epoch": 0.28488372093023256, "grad_norm": 40.71453094482422, "learning_rate": 7.151162790697676e-06, "loss": 0.525, "step": 1764 }, { "epoch": 0.2850452196382429, "grad_norm": 143.46139526367188, "learning_rate": 7.1495478036175715e-06, "loss": 2.0703, "step": 1765 }, { "epoch": 0.2852067183462532, "grad_norm": 104.31547546386719, "learning_rate": 7.147932816537468e-06, "loss": 1.8711, "step": 1766 }, { "epoch": 0.2853682170542636, "grad_norm": 38.62446975708008, "learning_rate": 7.1463178294573655e-06, "loss": 0.5137, "step": 1767 }, { "epoch": 0.28552971576227393, "grad_norm": 85.4679183959961, "learning_rate": 7.144702842377261e-06, "loss": 1.3008, "step": 1768 }, { "epoch": 0.2856912144702842, "grad_norm": 30.712034225463867, "learning_rate": 7.143087855297159e-06, "loss": 1.043, "step": 1769 }, { "epoch": 0.2858527131782946, "grad_norm": 42.59610366821289, "learning_rate": 7.141472868217055e-06, "loss": 1.0869, "step": 1770 }, { "epoch": 0.28601421188630494, "grad_norm": 27.88468360900879, "learning_rate": 7.139857881136951e-06, "loss": 0.7266, "step": 1771 }, { "epoch": 0.28617571059431524, "grad_norm": 24.426782608032227, "learning_rate": 7.138242894056848e-06, "loss": 0.231, "step": 1772 }, { "epoch": 0.2863372093023256, "grad_norm": 42.45061492919922, "learning_rate": 7.136627906976745e-06, "loss": 0.3442, "step": 1773 }, { "epoch": 0.2864987080103359, "grad_norm": 44.567405700683594, "learning_rate": 7.135012919896641e-06, "loss": 0.353, "step": 1774 }, { "epoch": 0.28666020671834624, "grad_norm": 55.67676544189453, "learning_rate": 7.133397932816538e-06, "loss": 0.9531, "step": 1775 }, { "epoch": 0.2868217054263566, "grad_norm": 11.52689266204834, "learning_rate": 7.131782945736435e-06, "loss": 0.2856, "step": 1776 }, { "epoch": 0.2869832041343669, "grad_norm": 42.20951843261719, "learning_rate": 7.130167958656331e-06, "loss": 0.3833, "step": 1777 }, { "epoch": 0.28714470284237725, "grad_norm": 56.86880111694336, "learning_rate": 7.1285529715762285e-06, "loss": 1.4375, "step": 1778 }, { "epoch": 0.2873062015503876, "grad_norm": 23.908851623535156, "learning_rate": 7.126937984496125e-06, "loss": 0.9092, "step": 1779 }, { "epoch": 0.2874677002583979, "grad_norm": 28.7761287689209, "learning_rate": 7.125322997416021e-06, "loss": 0.2236, "step": 1780 }, { "epoch": 0.28762919896640826, "grad_norm": 71.99066162109375, "learning_rate": 7.123708010335918e-06, "loss": 1.7363, "step": 1781 }, { "epoch": 0.2877906976744186, "grad_norm": 42.100006103515625, "learning_rate": 7.122093023255815e-06, "loss": 1.3281, "step": 1782 }, { "epoch": 0.2879521963824289, "grad_norm": 53.975379943847656, "learning_rate": 7.120478036175711e-06, "loss": 0.814, "step": 1783 }, { "epoch": 0.28811369509043927, "grad_norm": 33.316463470458984, "learning_rate": 7.118863049095608e-06, "loss": 0.4023, "step": 1784 }, { "epoch": 0.2882751937984496, "grad_norm": 49.51335525512695, "learning_rate": 7.117248062015505e-06, "loss": 1.043, "step": 1785 }, { "epoch": 0.2884366925064599, "grad_norm": 83.04106140136719, "learning_rate": 7.115633074935401e-06, "loss": 2.2227, "step": 1786 }, { "epoch": 0.2885981912144703, "grad_norm": 9.65747356414795, "learning_rate": 7.1140180878552975e-06, "loss": 0.0977, "step": 1787 }, { "epoch": 0.28875968992248063, "grad_norm": 26.30281639099121, "learning_rate": 7.112403100775195e-06, "loss": 0.4282, "step": 1788 }, { "epoch": 0.28892118863049093, "grad_norm": 51.33134460449219, "learning_rate": 7.110788113695091e-06, "loss": 0.3208, "step": 1789 }, { "epoch": 0.2890826873385013, "grad_norm": 64.96759796142578, "learning_rate": 7.109173126614988e-06, "loss": 1.1289, "step": 1790 }, { "epoch": 0.28924418604651164, "grad_norm": 43.337642669677734, "learning_rate": 7.107558139534885e-06, "loss": 0.8301, "step": 1791 }, { "epoch": 0.28940568475452194, "grad_norm": 74.6241455078125, "learning_rate": 7.10594315245478e-06, "loss": 1.1074, "step": 1792 }, { "epoch": 0.2895671834625323, "grad_norm": 27.67166519165039, "learning_rate": 7.104328165374678e-06, "loss": 0.4453, "step": 1793 }, { "epoch": 0.28972868217054265, "grad_norm": 54.22337341308594, "learning_rate": 7.102713178294574e-06, "loss": 0.519, "step": 1794 }, { "epoch": 0.28989018087855295, "grad_norm": 58.55338668823242, "learning_rate": 7.101098191214471e-06, "loss": 1.0947, "step": 1795 }, { "epoch": 0.2900516795865633, "grad_norm": 48.263580322265625, "learning_rate": 7.099483204134367e-06, "loss": 1.0166, "step": 1796 }, { "epoch": 0.29021317829457366, "grad_norm": 46.92389678955078, "learning_rate": 7.097868217054265e-06, "loss": 0.5994, "step": 1797 }, { "epoch": 0.29037467700258396, "grad_norm": 67.99052429199219, "learning_rate": 7.0962532299741605e-06, "loss": 1.6777, "step": 1798 }, { "epoch": 0.2905361757105943, "grad_norm": 44.25467300415039, "learning_rate": 7.094638242894057e-06, "loss": 0.6699, "step": 1799 }, { "epoch": 0.29069767441860467, "grad_norm": 42.91510009765625, "learning_rate": 7.0930232558139545e-06, "loss": 1.0078, "step": 1800 }, { "epoch": 0.29085917312661497, "grad_norm": 72.06844329833984, "learning_rate": 7.09140826873385e-06, "loss": 1.2578, "step": 1801 }, { "epoch": 0.2910206718346253, "grad_norm": 20.78049659729004, "learning_rate": 7.089793281653748e-06, "loss": 0.1049, "step": 1802 }, { "epoch": 0.2911821705426357, "grad_norm": 26.464828491210938, "learning_rate": 7.088178294573644e-06, "loss": 0.0999, "step": 1803 }, { "epoch": 0.291343669250646, "grad_norm": 197.3726043701172, "learning_rate": 7.08656330749354e-06, "loss": 5.5781, "step": 1804 }, { "epoch": 0.29150516795865633, "grad_norm": 34.46311569213867, "learning_rate": 7.084948320413437e-06, "loss": 0.3848, "step": 1805 }, { "epoch": 0.2916666666666667, "grad_norm": 64.03428649902344, "learning_rate": 7.083333333333335e-06, "loss": 0.3167, "step": 1806 }, { "epoch": 0.291828165374677, "grad_norm": 36.43134307861328, "learning_rate": 7.08171834625323e-06, "loss": 0.918, "step": 1807 }, { "epoch": 0.29198966408268734, "grad_norm": 31.917436599731445, "learning_rate": 7.080103359173127e-06, "loss": 0.439, "step": 1808 }, { "epoch": 0.2921511627906977, "grad_norm": 46.015052795410156, "learning_rate": 7.078488372093024e-06, "loss": 0.374, "step": 1809 }, { "epoch": 0.292312661498708, "grad_norm": 32.40153884887695, "learning_rate": 7.07687338501292e-06, "loss": 0.4414, "step": 1810 }, { "epoch": 0.29247416020671835, "grad_norm": 25.735963821411133, "learning_rate": 7.0752583979328175e-06, "loss": 0.2927, "step": 1811 }, { "epoch": 0.2926356589147287, "grad_norm": 51.048301696777344, "learning_rate": 7.073643410852714e-06, "loss": 1.1523, "step": 1812 }, { "epoch": 0.292797157622739, "grad_norm": 30.062055587768555, "learning_rate": 7.07202842377261e-06, "loss": 0.2412, "step": 1813 }, { "epoch": 0.29295865633074936, "grad_norm": 31.425628662109375, "learning_rate": 7.070413436692507e-06, "loss": 0.2388, "step": 1814 }, { "epoch": 0.2931201550387597, "grad_norm": 51.20408248901367, "learning_rate": 7.068798449612404e-06, "loss": 1.3086, "step": 1815 }, { "epoch": 0.29328165374677, "grad_norm": 68.68212127685547, "learning_rate": 7.0671834625323e-06, "loss": 1.1299, "step": 1816 }, { "epoch": 0.29344315245478036, "grad_norm": 18.515283584594727, "learning_rate": 7.065568475452197e-06, "loss": 0.2324, "step": 1817 }, { "epoch": 0.2936046511627907, "grad_norm": 5.517800331115723, "learning_rate": 7.063953488372094e-06, "loss": 0.0408, "step": 1818 }, { "epoch": 0.293766149870801, "grad_norm": 44.88121032714844, "learning_rate": 7.06233850129199e-06, "loss": 1.082, "step": 1819 }, { "epoch": 0.2939276485788114, "grad_norm": 69.43482971191406, "learning_rate": 7.0607235142118865e-06, "loss": 1.5273, "step": 1820 }, { "epoch": 0.29408914728682173, "grad_norm": 107.53089904785156, "learning_rate": 7.059108527131784e-06, "loss": 1.7422, "step": 1821 }, { "epoch": 0.294250645994832, "grad_norm": 37.66817855834961, "learning_rate": 7.05749354005168e-06, "loss": 0.6958, "step": 1822 }, { "epoch": 0.2944121447028424, "grad_norm": 69.76068878173828, "learning_rate": 7.055878552971577e-06, "loss": 0.7788, "step": 1823 }, { "epoch": 0.29457364341085274, "grad_norm": 49.711429595947266, "learning_rate": 7.054263565891474e-06, "loss": 0.5449, "step": 1824 }, { "epoch": 0.29473514211886304, "grad_norm": 51.278175354003906, "learning_rate": 7.052648578811369e-06, "loss": 1.3203, "step": 1825 }, { "epoch": 0.2948966408268734, "grad_norm": 33.68355941772461, "learning_rate": 7.051033591731267e-06, "loss": 0.6797, "step": 1826 }, { "epoch": 0.29505813953488375, "grad_norm": 21.207246780395508, "learning_rate": 7.049418604651163e-06, "loss": 0.1841, "step": 1827 }, { "epoch": 0.29521963824289404, "grad_norm": 67.90666961669922, "learning_rate": 7.04780361757106e-06, "loss": 0.5801, "step": 1828 }, { "epoch": 0.2953811369509044, "grad_norm": 94.74952697753906, "learning_rate": 7.046188630490956e-06, "loss": 1.748, "step": 1829 }, { "epoch": 0.29554263565891475, "grad_norm": 40.78789520263672, "learning_rate": 7.044573643410854e-06, "loss": 0.9055, "step": 1830 }, { "epoch": 0.29570413436692505, "grad_norm": 75.45520782470703, "learning_rate": 7.0429586563307495e-06, "loss": 1.1636, "step": 1831 }, { "epoch": 0.2958656330749354, "grad_norm": 19.949264526367188, "learning_rate": 7.041343669250646e-06, "loss": 0.2327, "step": 1832 }, { "epoch": 0.29602713178294576, "grad_norm": 27.352827072143555, "learning_rate": 7.0397286821705435e-06, "loss": 0.2969, "step": 1833 }, { "epoch": 0.29618863049095606, "grad_norm": 16.580039978027344, "learning_rate": 7.038113695090439e-06, "loss": 0.1494, "step": 1834 }, { "epoch": 0.2963501291989664, "grad_norm": 27.295547485351562, "learning_rate": 7.036498708010337e-06, "loss": 0.5596, "step": 1835 }, { "epoch": 0.29651162790697677, "grad_norm": 118.9788589477539, "learning_rate": 7.034883720930233e-06, "loss": 2.6133, "step": 1836 }, { "epoch": 0.29667312661498707, "grad_norm": 36.44071578979492, "learning_rate": 7.033268733850129e-06, "loss": 0.75, "step": 1837 }, { "epoch": 0.2968346253229974, "grad_norm": 75.65033721923828, "learning_rate": 7.031653746770026e-06, "loss": 0.7207, "step": 1838 }, { "epoch": 0.2969961240310077, "grad_norm": 43.93791580200195, "learning_rate": 7.030038759689924e-06, "loss": 0.334, "step": 1839 }, { "epoch": 0.2971576227390181, "grad_norm": 13.322487831115723, "learning_rate": 7.028423772609819e-06, "loss": 0.1365, "step": 1840 }, { "epoch": 0.29731912144702843, "grad_norm": 34.916873931884766, "learning_rate": 7.026808785529716e-06, "loss": 0.3043, "step": 1841 }, { "epoch": 0.29748062015503873, "grad_norm": 32.79376983642578, "learning_rate": 7.025193798449613e-06, "loss": 0.7207, "step": 1842 }, { "epoch": 0.2976421188630491, "grad_norm": 50.650848388671875, "learning_rate": 7.023578811369509e-06, "loss": 0.4504, "step": 1843 }, { "epoch": 0.29780361757105944, "grad_norm": 78.09847259521484, "learning_rate": 7.0219638242894065e-06, "loss": 0.5342, "step": 1844 }, { "epoch": 0.29796511627906974, "grad_norm": 28.018232345581055, "learning_rate": 7.020348837209303e-06, "loss": 0.4526, "step": 1845 }, { "epoch": 0.2981266149870801, "grad_norm": 123.45020294189453, "learning_rate": 7.018733850129199e-06, "loss": 1.563, "step": 1846 }, { "epoch": 0.29828811369509045, "grad_norm": 80.57960510253906, "learning_rate": 7.017118863049096e-06, "loss": 1.3857, "step": 1847 }, { "epoch": 0.29844961240310075, "grad_norm": 47.72007369995117, "learning_rate": 7.015503875968993e-06, "loss": 0.2529, "step": 1848 }, { "epoch": 0.2986111111111111, "grad_norm": 32.86091232299805, "learning_rate": 7.013888888888889e-06, "loss": 0.2476, "step": 1849 }, { "epoch": 0.29877260981912146, "grad_norm": 43.69801330566406, "learning_rate": 7.012273901808786e-06, "loss": 0.6699, "step": 1850 }, { "epoch": 0.29893410852713176, "grad_norm": 40.416709899902344, "learning_rate": 7.010658914728683e-06, "loss": 0.4785, "step": 1851 }, { "epoch": 0.2990956072351421, "grad_norm": 56.49475860595703, "learning_rate": 7.009043927648579e-06, "loss": 0.916, "step": 1852 }, { "epoch": 0.29925710594315247, "grad_norm": 61.357330322265625, "learning_rate": 7.0074289405684755e-06, "loss": 0.9121, "step": 1853 }, { "epoch": 0.29941860465116277, "grad_norm": 19.04566764831543, "learning_rate": 7.005813953488373e-06, "loss": 0.1974, "step": 1854 }, { "epoch": 0.2995801033591731, "grad_norm": 65.32856750488281, "learning_rate": 7.004198966408269e-06, "loss": 0.9556, "step": 1855 }, { "epoch": 0.2997416020671835, "grad_norm": 29.8968448638916, "learning_rate": 7.002583979328166e-06, "loss": 0.29, "step": 1856 }, { "epoch": 0.2999031007751938, "grad_norm": 28.869380950927734, "learning_rate": 7.000968992248063e-06, "loss": 0.4338, "step": 1857 }, { "epoch": 0.30006459948320413, "grad_norm": 55.923892974853516, "learning_rate": 6.999354005167958e-06, "loss": 0.8369, "step": 1858 }, { "epoch": 0.3002260981912145, "grad_norm": 50.09220886230469, "learning_rate": 6.997739018087856e-06, "loss": 0.3584, "step": 1859 }, { "epoch": 0.3003875968992248, "grad_norm": 29.897836685180664, "learning_rate": 6.996124031007752e-06, "loss": 0.3638, "step": 1860 }, { "epoch": 0.30054909560723514, "grad_norm": 8.9434175491333, "learning_rate": 6.994509043927649e-06, "loss": 0.0575, "step": 1861 }, { "epoch": 0.3007105943152455, "grad_norm": 26.40554428100586, "learning_rate": 6.9928940568475454e-06, "loss": 0.3213, "step": 1862 }, { "epoch": 0.3008720930232558, "grad_norm": 29.134578704833984, "learning_rate": 6.991279069767443e-06, "loss": 0.1284, "step": 1863 }, { "epoch": 0.30103359173126615, "grad_norm": 47.02405548095703, "learning_rate": 6.9896640826873386e-06, "loss": 0.371, "step": 1864 }, { "epoch": 0.3011950904392765, "grad_norm": 29.01561737060547, "learning_rate": 6.988049095607236e-06, "loss": 0.2661, "step": 1865 }, { "epoch": 0.3013565891472868, "grad_norm": 58.802940368652344, "learning_rate": 6.9864341085271325e-06, "loss": 0.8613, "step": 1866 }, { "epoch": 0.30151808785529716, "grad_norm": 59.5710334777832, "learning_rate": 6.984819121447028e-06, "loss": 0.9824, "step": 1867 }, { "epoch": 0.3016795865633075, "grad_norm": 263.0526123046875, "learning_rate": 6.983204134366926e-06, "loss": 3.8984, "step": 1868 }, { "epoch": 0.3018410852713178, "grad_norm": 49.72529602050781, "learning_rate": 6.981589147286822e-06, "loss": 0.7861, "step": 1869 }, { "epoch": 0.30200258397932817, "grad_norm": 24.781641006469727, "learning_rate": 6.979974160206719e-06, "loss": 0.2544, "step": 1870 }, { "epoch": 0.3021640826873385, "grad_norm": 113.59662628173828, "learning_rate": 6.978359173126615e-06, "loss": 2.4863, "step": 1871 }, { "epoch": 0.3023255813953488, "grad_norm": 134.10372924804688, "learning_rate": 6.976744186046513e-06, "loss": 4.4434, "step": 1872 }, { "epoch": 0.3024870801033592, "grad_norm": 38.6744384765625, "learning_rate": 6.975129198966409e-06, "loss": 0.1973, "step": 1873 }, { "epoch": 0.30264857881136953, "grad_norm": 33.98035430908203, "learning_rate": 6.973514211886305e-06, "loss": 0.6357, "step": 1874 }, { "epoch": 0.3028100775193798, "grad_norm": 54.309024810791016, "learning_rate": 6.971899224806202e-06, "loss": 0.7359, "step": 1875 }, { "epoch": 0.3029715762273902, "grad_norm": 34.54389572143555, "learning_rate": 6.970284237726099e-06, "loss": 0.3413, "step": 1876 }, { "epoch": 0.30313307493540054, "grad_norm": 65.57372283935547, "learning_rate": 6.9686692506459955e-06, "loss": 0.915, "step": 1877 }, { "epoch": 0.30329457364341084, "grad_norm": 42.025577545166016, "learning_rate": 6.967054263565892e-06, "loss": 0.23, "step": 1878 }, { "epoch": 0.3034560723514212, "grad_norm": 61.017173767089844, "learning_rate": 6.9654392764857895e-06, "loss": 1.1719, "step": 1879 }, { "epoch": 0.30361757105943155, "grad_norm": 135.14718627929688, "learning_rate": 6.963824289405685e-06, "loss": 2.1182, "step": 1880 }, { "epoch": 0.30377906976744184, "grad_norm": 73.15462493896484, "learning_rate": 6.962209302325582e-06, "loss": 1.7266, "step": 1881 }, { "epoch": 0.3039405684754522, "grad_norm": 27.804895401000977, "learning_rate": 6.960594315245479e-06, "loss": 0.1919, "step": 1882 }, { "epoch": 0.30410206718346255, "grad_norm": 33.68471908569336, "learning_rate": 6.958979328165375e-06, "loss": 0.2114, "step": 1883 }, { "epoch": 0.30426356589147285, "grad_norm": 56.55443572998047, "learning_rate": 6.957364341085272e-06, "loss": 1.1055, "step": 1884 }, { "epoch": 0.3044250645994832, "grad_norm": 159.1809539794922, "learning_rate": 6.955749354005169e-06, "loss": 2.0195, "step": 1885 }, { "epoch": 0.30458656330749356, "grad_norm": 28.243824005126953, "learning_rate": 6.9541343669250646e-06, "loss": 0.1519, "step": 1886 }, { "epoch": 0.30474806201550386, "grad_norm": 74.42101287841797, "learning_rate": 6.952519379844962e-06, "loss": 1.2891, "step": 1887 }, { "epoch": 0.3049095607235142, "grad_norm": 21.617429733276367, "learning_rate": 6.9509043927648585e-06, "loss": 0.0785, "step": 1888 }, { "epoch": 0.30507105943152457, "grad_norm": 48.34174346923828, "learning_rate": 6.949289405684755e-06, "loss": 0.6945, "step": 1889 }, { "epoch": 0.30523255813953487, "grad_norm": 8.95276927947998, "learning_rate": 6.947674418604652e-06, "loss": 0.0452, "step": 1890 }, { "epoch": 0.3053940568475452, "grad_norm": 17.18292236328125, "learning_rate": 6.946059431524549e-06, "loss": 0.2061, "step": 1891 }, { "epoch": 0.3055555555555556, "grad_norm": 135.3762969970703, "learning_rate": 6.944444444444445e-06, "loss": 2.3486, "step": 1892 }, { "epoch": 0.3057170542635659, "grad_norm": 82.20623016357422, "learning_rate": 6.942829457364342e-06, "loss": 0.9883, "step": 1893 }, { "epoch": 0.30587855297157623, "grad_norm": 43.875972747802734, "learning_rate": 6.941214470284239e-06, "loss": 0.2798, "step": 1894 }, { "epoch": 0.3060400516795866, "grad_norm": 173.05955505371094, "learning_rate": 6.9395994832041345e-06, "loss": 3.6172, "step": 1895 }, { "epoch": 0.3062015503875969, "grad_norm": 106.66667175292969, "learning_rate": 6.937984496124032e-06, "loss": 1.6582, "step": 1896 }, { "epoch": 0.30636304909560724, "grad_norm": 55.12664031982422, "learning_rate": 6.9363695090439284e-06, "loss": 0.7451, "step": 1897 }, { "epoch": 0.3065245478036176, "grad_norm": 66.6450424194336, "learning_rate": 6.934754521963825e-06, "loss": 0.7832, "step": 1898 }, { "epoch": 0.3066860465116279, "grad_norm": 135.15826416015625, "learning_rate": 6.9331395348837215e-06, "loss": 1.9414, "step": 1899 }, { "epoch": 0.30684754521963825, "grad_norm": 23.547832489013672, "learning_rate": 6.931524547803619e-06, "loss": 0.3428, "step": 1900 }, { "epoch": 0.3070090439276486, "grad_norm": 92.42605590820312, "learning_rate": 6.929909560723515e-06, "loss": 1.418, "step": 1901 }, { "epoch": 0.3071705426356589, "grad_norm": 68.27194213867188, "learning_rate": 6.928294573643411e-06, "loss": 0.8223, "step": 1902 }, { "epoch": 0.30733204134366926, "grad_norm": 66.73298645019531, "learning_rate": 6.926679586563309e-06, "loss": 0.844, "step": 1903 }, { "epoch": 0.30749354005167956, "grad_norm": 43.315399169921875, "learning_rate": 6.925064599483204e-06, "loss": 0.641, "step": 1904 }, { "epoch": 0.3076550387596899, "grad_norm": 110.92855072021484, "learning_rate": 6.923449612403102e-06, "loss": 1.9297, "step": 1905 }, { "epoch": 0.30781653746770027, "grad_norm": 53.115623474121094, "learning_rate": 6.921834625322998e-06, "loss": 0.8459, "step": 1906 }, { "epoch": 0.30797803617571057, "grad_norm": 65.67390441894531, "learning_rate": 6.920219638242894e-06, "loss": 0.9874, "step": 1907 }, { "epoch": 0.3081395348837209, "grad_norm": 69.5621337890625, "learning_rate": 6.9186046511627914e-06, "loss": 1.2363, "step": 1908 }, { "epoch": 0.3083010335917313, "grad_norm": 15.151824951171875, "learning_rate": 6.916989664082688e-06, "loss": 0.1072, "step": 1909 }, { "epoch": 0.3084625322997416, "grad_norm": 171.50827026367188, "learning_rate": 6.9153746770025846e-06, "loss": 0.7832, "step": 1910 }, { "epoch": 0.30862403100775193, "grad_norm": 141.78695678710938, "learning_rate": 6.913759689922481e-06, "loss": 2.4648, "step": 1911 }, { "epoch": 0.3087855297157623, "grad_norm": 176.72988891601562, "learning_rate": 6.9121447028423785e-06, "loss": 3.9766, "step": 1912 }, { "epoch": 0.3089470284237726, "grad_norm": 16.662755966186523, "learning_rate": 6.910529715762274e-06, "loss": 0.114, "step": 1913 }, { "epoch": 0.30910852713178294, "grad_norm": 68.98399353027344, "learning_rate": 6.908914728682171e-06, "loss": 0.8931, "step": 1914 }, { "epoch": 0.3092700258397933, "grad_norm": 89.8611831665039, "learning_rate": 6.907299741602068e-06, "loss": 1.0298, "step": 1915 }, { "epoch": 0.3094315245478036, "grad_norm": 40.72753143310547, "learning_rate": 6.905684754521964e-06, "loss": 0.9844, "step": 1916 }, { "epoch": 0.30959302325581395, "grad_norm": 8.283044815063477, "learning_rate": 6.904069767441861e-06, "loss": 0.0612, "step": 1917 }, { "epoch": 0.3097545219638243, "grad_norm": 149.00035095214844, "learning_rate": 6.902454780361758e-06, "loss": 2.5, "step": 1918 }, { "epoch": 0.3099160206718346, "grad_norm": 41.53738784790039, "learning_rate": 6.900839793281654e-06, "loss": 0.4736, "step": 1919 }, { "epoch": 0.31007751937984496, "grad_norm": 41.433963775634766, "learning_rate": 6.899224806201551e-06, "loss": 0.2192, "step": 1920 }, { "epoch": 0.3102390180878553, "grad_norm": 106.40824127197266, "learning_rate": 6.897609819121448e-06, "loss": 1.9336, "step": 1921 }, { "epoch": 0.3104005167958656, "grad_norm": 17.29876136779785, "learning_rate": 6.895994832041344e-06, "loss": 0.2251, "step": 1922 }, { "epoch": 0.31056201550387597, "grad_norm": 63.65379333496094, "learning_rate": 6.894379844961241e-06, "loss": 0.8633, "step": 1923 }, { "epoch": 0.3107235142118863, "grad_norm": 179.72337341308594, "learning_rate": 6.892764857881138e-06, "loss": 4.1914, "step": 1924 }, { "epoch": 0.3108850129198966, "grad_norm": 58.75829315185547, "learning_rate": 6.891149870801034e-06, "loss": 0.6951, "step": 1925 }, { "epoch": 0.311046511627907, "grad_norm": 147.05589294433594, "learning_rate": 6.889534883720931e-06, "loss": 3.3203, "step": 1926 }, { "epoch": 0.31120801033591733, "grad_norm": 47.10170364379883, "learning_rate": 6.887919896640828e-06, "loss": 0.9941, "step": 1927 }, { "epoch": 0.31136950904392763, "grad_norm": 9.469071388244629, "learning_rate": 6.8863049095607235e-06, "loss": 0.0996, "step": 1928 }, { "epoch": 0.311531007751938, "grad_norm": 107.0784912109375, "learning_rate": 6.884689922480621e-06, "loss": 1.428, "step": 1929 }, { "epoch": 0.31169250645994834, "grad_norm": 16.219287872314453, "learning_rate": 6.8830749354005174e-06, "loss": 0.1561, "step": 1930 }, { "epoch": 0.31185400516795864, "grad_norm": 134.1500701904297, "learning_rate": 6.881459948320414e-06, "loss": 2.0251, "step": 1931 }, { "epoch": 0.312015503875969, "grad_norm": 88.30909729003906, "learning_rate": 6.8798449612403106e-06, "loss": 2.8672, "step": 1932 }, { "epoch": 0.31217700258397935, "grad_norm": 165.33364868164062, "learning_rate": 6.878229974160208e-06, "loss": 3.834, "step": 1933 }, { "epoch": 0.31233850129198965, "grad_norm": 42.94999694824219, "learning_rate": 6.876614987080104e-06, "loss": 0.6074, "step": 1934 }, { "epoch": 0.3125, "grad_norm": 86.44923400878906, "learning_rate": 6.875e-06, "loss": 1.0781, "step": 1935 }, { "epoch": 0.31266149870801035, "grad_norm": 24.767749786376953, "learning_rate": 6.873385012919898e-06, "loss": 0.332, "step": 1936 }, { "epoch": 0.31282299741602065, "grad_norm": 21.6710205078125, "learning_rate": 6.871770025839793e-06, "loss": 0.5576, "step": 1937 }, { "epoch": 0.312984496124031, "grad_norm": 14.479226112365723, "learning_rate": 6.870155038759691e-06, "loss": 0.0502, "step": 1938 }, { "epoch": 0.31314599483204136, "grad_norm": 23.135818481445312, "learning_rate": 6.868540051679587e-06, "loss": 0.2375, "step": 1939 }, { "epoch": 0.31330749354005166, "grad_norm": 76.46595001220703, "learning_rate": 6.866925064599483e-06, "loss": 1.543, "step": 1940 }, { "epoch": 0.313468992248062, "grad_norm": 56.946533203125, "learning_rate": 6.8653100775193805e-06, "loss": 0.7479, "step": 1941 }, { "epoch": 0.31363049095607237, "grad_norm": 32.547542572021484, "learning_rate": 6.863695090439277e-06, "loss": 0.3447, "step": 1942 }, { "epoch": 0.31379198966408267, "grad_norm": 71.56258392333984, "learning_rate": 6.8620801033591736e-06, "loss": 1.3203, "step": 1943 }, { "epoch": 0.313953488372093, "grad_norm": 36.5561408996582, "learning_rate": 6.86046511627907e-06, "loss": 0.2855, "step": 1944 }, { "epoch": 0.3141149870801034, "grad_norm": 15.945393562316895, "learning_rate": 6.8588501291989675e-06, "loss": 0.0422, "step": 1945 }, { "epoch": 0.3142764857881137, "grad_norm": 74.89093017578125, "learning_rate": 6.857235142118863e-06, "loss": 1.2773, "step": 1946 }, { "epoch": 0.31443798449612403, "grad_norm": 55.78115463256836, "learning_rate": 6.85562015503876e-06, "loss": 0.791, "step": 1947 }, { "epoch": 0.3145994832041344, "grad_norm": 108.41162872314453, "learning_rate": 6.854005167958657e-06, "loss": 1.7422, "step": 1948 }, { "epoch": 0.3147609819121447, "grad_norm": 77.88683319091797, "learning_rate": 6.852390180878553e-06, "loss": 1.6406, "step": 1949 }, { "epoch": 0.31492248062015504, "grad_norm": 10.026765823364258, "learning_rate": 6.85077519379845e-06, "loss": 0.1323, "step": 1950 }, { "epoch": 0.3150839793281654, "grad_norm": 22.87284278869629, "learning_rate": 6.849160206718347e-06, "loss": 0.1135, "step": 1951 }, { "epoch": 0.3152454780361757, "grad_norm": 8.822253227233887, "learning_rate": 6.8475452196382435e-06, "loss": 0.0812, "step": 1952 }, { "epoch": 0.31540697674418605, "grad_norm": 81.95968627929688, "learning_rate": 6.84593023255814e-06, "loss": 1.425, "step": 1953 }, { "epoch": 0.3155684754521964, "grad_norm": 55.47211456298828, "learning_rate": 6.8443152454780374e-06, "loss": 0.6729, "step": 1954 }, { "epoch": 0.3157299741602067, "grad_norm": 53.12609100341797, "learning_rate": 6.842700258397933e-06, "loss": 0.5913, "step": 1955 }, { "epoch": 0.31589147286821706, "grad_norm": 79.87676239013672, "learning_rate": 6.84108527131783e-06, "loss": 0.8789, "step": 1956 }, { "epoch": 0.3160529715762274, "grad_norm": 51.16328430175781, "learning_rate": 6.839470284237727e-06, "loss": 0.5656, "step": 1957 }, { "epoch": 0.3162144702842377, "grad_norm": 143.1739959716797, "learning_rate": 6.837855297157623e-06, "loss": 3.2324, "step": 1958 }, { "epoch": 0.31637596899224807, "grad_norm": 112.82102966308594, "learning_rate": 6.83624031007752e-06, "loss": 1.8008, "step": 1959 }, { "epoch": 0.3165374677002584, "grad_norm": 49.56686019897461, "learning_rate": 6.834625322997417e-06, "loss": 0.5337, "step": 1960 }, { "epoch": 0.3166989664082687, "grad_norm": 48.21661376953125, "learning_rate": 6.8330103359173125e-06, "loss": 0.6729, "step": 1961 }, { "epoch": 0.3168604651162791, "grad_norm": 53.74493408203125, "learning_rate": 6.83139534883721e-06, "loss": 0.5166, "step": 1962 }, { "epoch": 0.31702196382428943, "grad_norm": 81.4117202758789, "learning_rate": 6.8297803617571065e-06, "loss": 0.9453, "step": 1963 }, { "epoch": 0.31718346253229973, "grad_norm": 35.93918228149414, "learning_rate": 6.828165374677003e-06, "loss": 0.4111, "step": 1964 }, { "epoch": 0.3173449612403101, "grad_norm": 27.32982063293457, "learning_rate": 6.8265503875969e-06, "loss": 0.2144, "step": 1965 }, { "epoch": 0.31750645994832044, "grad_norm": 45.310211181640625, "learning_rate": 6.824935400516797e-06, "loss": 0.5259, "step": 1966 }, { "epoch": 0.31766795865633074, "grad_norm": 29.650379180908203, "learning_rate": 6.823320413436693e-06, "loss": 0.3348, "step": 1967 }, { "epoch": 0.3178294573643411, "grad_norm": 84.3740463256836, "learning_rate": 6.821705426356589e-06, "loss": 1.5195, "step": 1968 }, { "epoch": 0.3179909560723514, "grad_norm": 20.7628173828125, "learning_rate": 6.820090439276487e-06, "loss": 0.3262, "step": 1969 }, { "epoch": 0.31815245478036175, "grad_norm": 45.86277770996094, "learning_rate": 6.818475452196382e-06, "loss": 0.6064, "step": 1970 }, { "epoch": 0.3183139534883721, "grad_norm": 25.72551918029785, "learning_rate": 6.81686046511628e-06, "loss": 0.3752, "step": 1971 }, { "epoch": 0.3184754521963824, "grad_norm": 75.27384185791016, "learning_rate": 6.815245478036176e-06, "loss": 0.7588, "step": 1972 }, { "epoch": 0.31863695090439276, "grad_norm": 6.354292392730713, "learning_rate": 6.813630490956072e-06, "loss": 0.053, "step": 1973 }, { "epoch": 0.3187984496124031, "grad_norm": 7.068199634552002, "learning_rate": 6.8120155038759695e-06, "loss": 0.0366, "step": 1974 }, { "epoch": 0.3189599483204134, "grad_norm": 109.64950561523438, "learning_rate": 6.810400516795866e-06, "loss": 2.0156, "step": 1975 }, { "epoch": 0.31912144702842377, "grad_norm": 42.20124053955078, "learning_rate": 6.808785529715763e-06, "loss": 0.4946, "step": 1976 }, { "epoch": 0.3192829457364341, "grad_norm": 60.85858917236328, "learning_rate": 6.807170542635659e-06, "loss": 1.1459, "step": 1977 }, { "epoch": 0.3194444444444444, "grad_norm": 152.02610778808594, "learning_rate": 6.8055555555555566e-06, "loss": 2.6953, "step": 1978 }, { "epoch": 0.3196059431524548, "grad_norm": 19.77166748046875, "learning_rate": 6.803940568475452e-06, "loss": 0.1912, "step": 1979 }, { "epoch": 0.31976744186046513, "grad_norm": 51.19384002685547, "learning_rate": 6.80232558139535e-06, "loss": 0.7363, "step": 1980 }, { "epoch": 0.31992894056847543, "grad_norm": 130.8761444091797, "learning_rate": 6.800710594315246e-06, "loss": 1.8145, "step": 1981 }, { "epoch": 0.3200904392764858, "grad_norm": 100.26598358154297, "learning_rate": 6.799095607235142e-06, "loss": 1.5, "step": 1982 }, { "epoch": 0.32025193798449614, "grad_norm": 169.0968780517578, "learning_rate": 6.797480620155039e-06, "loss": 2.5029, "step": 1983 }, { "epoch": 0.32041343669250644, "grad_norm": 157.45211791992188, "learning_rate": 6.795865633074936e-06, "loss": 4.7891, "step": 1984 }, { "epoch": 0.3205749354005168, "grad_norm": 83.78268432617188, "learning_rate": 6.7942506459948325e-06, "loss": 0.9297, "step": 1985 }, { "epoch": 0.32073643410852715, "grad_norm": 99.8552474975586, "learning_rate": 6.792635658914729e-06, "loss": 1.4453, "step": 1986 }, { "epoch": 0.32089793281653745, "grad_norm": 41.288299560546875, "learning_rate": 6.7910206718346264e-06, "loss": 0.5894, "step": 1987 }, { "epoch": 0.3210594315245478, "grad_norm": 115.97752380371094, "learning_rate": 6.789405684754522e-06, "loss": 2.0918, "step": 1988 }, { "epoch": 0.32122093023255816, "grad_norm": 89.96470642089844, "learning_rate": 6.787790697674419e-06, "loss": 1.2524, "step": 1989 }, { "epoch": 0.32138242894056845, "grad_norm": 98.7801742553711, "learning_rate": 6.786175710594316e-06, "loss": 0.8237, "step": 1990 }, { "epoch": 0.3215439276485788, "grad_norm": 143.7722930908203, "learning_rate": 6.784560723514212e-06, "loss": 3.543, "step": 1991 }, { "epoch": 0.32170542635658916, "grad_norm": 104.00434875488281, "learning_rate": 6.782945736434109e-06, "loss": 1.4609, "step": 1992 }, { "epoch": 0.32186692506459946, "grad_norm": 177.62721252441406, "learning_rate": 6.781330749354006e-06, "loss": 5.875, "step": 1993 }, { "epoch": 0.3220284237726098, "grad_norm": 74.1453857421875, "learning_rate": 6.7797157622739015e-06, "loss": 1.5396, "step": 1994 }, { "epoch": 0.3221899224806202, "grad_norm": 39.4672966003418, "learning_rate": 6.778100775193799e-06, "loss": 0.3871, "step": 1995 }, { "epoch": 0.32235142118863047, "grad_norm": 4.97456169128418, "learning_rate": 6.7764857881136955e-06, "loss": 0.0381, "step": 1996 }, { "epoch": 0.3225129198966408, "grad_norm": 29.70588493347168, "learning_rate": 6.774870801033592e-06, "loss": 0.2617, "step": 1997 }, { "epoch": 0.3226744186046512, "grad_norm": 100.20330047607422, "learning_rate": 6.773255813953489e-06, "loss": 1.8633, "step": 1998 }, { "epoch": 0.3228359173126615, "grad_norm": 51.29451370239258, "learning_rate": 6.771640826873386e-06, "loss": 0.7332, "step": 1999 }, { "epoch": 0.32299741602067183, "grad_norm": 37.90533447265625, "learning_rate": 6.7700258397932826e-06, "loss": 0.3936, "step": 2000 }, { "epoch": 0.3231589147286822, "grad_norm": 85.71157836914062, "learning_rate": 6.768410852713178e-06, "loss": 2.0703, "step": 2001 }, { "epoch": 0.3233204134366925, "grad_norm": 24.215185165405273, "learning_rate": 6.766795865633076e-06, "loss": 0.2451, "step": 2002 }, { "epoch": 0.32348191214470284, "grad_norm": 84.853271484375, "learning_rate": 6.765180878552972e-06, "loss": 1.6289, "step": 2003 }, { "epoch": 0.3236434108527132, "grad_norm": 58.15114212036133, "learning_rate": 6.763565891472869e-06, "loss": 0.5686, "step": 2004 }, { "epoch": 0.3238049095607235, "grad_norm": 48.43271255493164, "learning_rate": 6.761950904392765e-06, "loss": 0.8179, "step": 2005 }, { "epoch": 0.32396640826873385, "grad_norm": 23.407773971557617, "learning_rate": 6.760335917312663e-06, "loss": 0.3022, "step": 2006 }, { "epoch": 0.3241279069767442, "grad_norm": 10.161870956420898, "learning_rate": 6.7587209302325585e-06, "loss": 0.0218, "step": 2007 }, { "epoch": 0.3242894056847545, "grad_norm": 75.86894226074219, "learning_rate": 6.757105943152456e-06, "loss": 1.1055, "step": 2008 }, { "epoch": 0.32445090439276486, "grad_norm": 63.70458984375, "learning_rate": 6.7554909560723525e-06, "loss": 1.0137, "step": 2009 }, { "epoch": 0.3246124031007752, "grad_norm": 43.84260177612305, "learning_rate": 6.753875968992248e-06, "loss": 0.3096, "step": 2010 }, { "epoch": 0.3247739018087855, "grad_norm": 15.796850204467773, "learning_rate": 6.752260981912146e-06, "loss": 0.1909, "step": 2011 }, { "epoch": 0.32493540051679587, "grad_norm": 109.86540985107422, "learning_rate": 6.750645994832042e-06, "loss": 3.0547, "step": 2012 }, { "epoch": 0.3250968992248062, "grad_norm": 66.19121551513672, "learning_rate": 6.749031007751939e-06, "loss": 1.6729, "step": 2013 }, { "epoch": 0.3252583979328165, "grad_norm": 51.02176284790039, "learning_rate": 6.747416020671835e-06, "loss": 0.439, "step": 2014 }, { "epoch": 0.3254198966408269, "grad_norm": 19.08128547668457, "learning_rate": 6.745801033591733e-06, "loss": 0.1206, "step": 2015 }, { "epoch": 0.32558139534883723, "grad_norm": 44.648807525634766, "learning_rate": 6.744186046511628e-06, "loss": 0.4482, "step": 2016 }, { "epoch": 0.32574289405684753, "grad_norm": 10.709989547729492, "learning_rate": 6.742571059431525e-06, "loss": 0.1234, "step": 2017 }, { "epoch": 0.3259043927648579, "grad_norm": 29.76681137084961, "learning_rate": 6.740956072351422e-06, "loss": 0.1367, "step": 2018 }, { "epoch": 0.32606589147286824, "grad_norm": 33.57713317871094, "learning_rate": 6.739341085271318e-06, "loss": 0.3809, "step": 2019 }, { "epoch": 0.32622739018087854, "grad_norm": 22.195234298706055, "learning_rate": 6.7377260981912155e-06, "loss": 0.1997, "step": 2020 }, { "epoch": 0.3263888888888889, "grad_norm": 69.84056854248047, "learning_rate": 6.736111111111112e-06, "loss": 1.0371, "step": 2021 }, { "epoch": 0.32655038759689925, "grad_norm": 77.24095916748047, "learning_rate": 6.734496124031008e-06, "loss": 1.5508, "step": 2022 }, { "epoch": 0.32671188630490955, "grad_norm": 36.31396484375, "learning_rate": 6.732881136950905e-06, "loss": 0.3867, "step": 2023 }, { "epoch": 0.3268733850129199, "grad_norm": 166.25279235839844, "learning_rate": 6.731266149870802e-06, "loss": 2.3252, "step": 2024 }, { "epoch": 0.32703488372093026, "grad_norm": 47.27819061279297, "learning_rate": 6.729651162790698e-06, "loss": 0.6561, "step": 2025 }, { "epoch": 0.32719638242894056, "grad_norm": 22.491483688354492, "learning_rate": 6.728036175710595e-06, "loss": 0.2467, "step": 2026 }, { "epoch": 0.3273578811369509, "grad_norm": 36.46189498901367, "learning_rate": 6.726421188630492e-06, "loss": 0.2021, "step": 2027 }, { "epoch": 0.32751937984496127, "grad_norm": 18.836444854736328, "learning_rate": 6.724806201550388e-06, "loss": 0.0994, "step": 2028 }, { "epoch": 0.32768087855297157, "grad_norm": 50.10405349731445, "learning_rate": 6.7231912144702845e-06, "loss": 0.3271, "step": 2029 }, { "epoch": 0.3278423772609819, "grad_norm": 52.21334457397461, "learning_rate": 6.721576227390182e-06, "loss": 1.2231, "step": 2030 }, { "epoch": 0.3280038759689923, "grad_norm": 19.673784255981445, "learning_rate": 6.719961240310078e-06, "loss": 0.3237, "step": 2031 }, { "epoch": 0.3281653746770026, "grad_norm": 53.939266204833984, "learning_rate": 6.718346253229975e-06, "loss": 1.0391, "step": 2032 }, { "epoch": 0.32832687338501293, "grad_norm": 74.3012466430664, "learning_rate": 6.716731266149872e-06, "loss": 1.3438, "step": 2033 }, { "epoch": 0.32848837209302323, "grad_norm": 23.073619842529297, "learning_rate": 6.715116279069767e-06, "loss": 0.1755, "step": 2034 }, { "epoch": 0.3286498708010336, "grad_norm": 81.28337097167969, "learning_rate": 6.713501291989665e-06, "loss": 1.4141, "step": 2035 }, { "epoch": 0.32881136950904394, "grad_norm": 22.869285583496094, "learning_rate": 6.711886304909562e-06, "loss": 0.1807, "step": 2036 }, { "epoch": 0.32897286821705424, "grad_norm": 16.732894897460938, "learning_rate": 6.710271317829458e-06, "loss": 0.1965, "step": 2037 }, { "epoch": 0.3291343669250646, "grad_norm": 48.616477966308594, "learning_rate": 6.708656330749354e-06, "loss": 0.2244, "step": 2038 }, { "epoch": 0.32929586563307495, "grad_norm": 38.13424301147461, "learning_rate": 6.707041343669252e-06, "loss": 0.2939, "step": 2039 }, { "epoch": 0.32945736434108525, "grad_norm": 37.94100570678711, "learning_rate": 6.7054263565891475e-06, "loss": 1.0088, "step": 2040 }, { "epoch": 0.3296188630490956, "grad_norm": 90.02474975585938, "learning_rate": 6.703811369509045e-06, "loss": 2.0117, "step": 2041 }, { "epoch": 0.32978036175710596, "grad_norm": 10.616349220275879, "learning_rate": 6.7021963824289415e-06, "loss": 0.0349, "step": 2042 }, { "epoch": 0.32994186046511625, "grad_norm": 127.21553802490234, "learning_rate": 6.700581395348837e-06, "loss": 1.6953, "step": 2043 }, { "epoch": 0.3301033591731266, "grad_norm": 116.30426788330078, "learning_rate": 6.698966408268735e-06, "loss": 2.4399, "step": 2044 }, { "epoch": 0.33026485788113696, "grad_norm": 107.34989166259766, "learning_rate": 6.697351421188631e-06, "loss": 1.8594, "step": 2045 }, { "epoch": 0.33042635658914726, "grad_norm": 35.942893981933594, "learning_rate": 6.695736434108528e-06, "loss": 0.7798, "step": 2046 }, { "epoch": 0.3305878552971576, "grad_norm": 20.315004348754883, "learning_rate": 6.694121447028424e-06, "loss": 0.1612, "step": 2047 }, { "epoch": 0.330749354005168, "grad_norm": 94.37885284423828, "learning_rate": 6.692506459948322e-06, "loss": 0.9746, "step": 2048 }, { "epoch": 0.33091085271317827, "grad_norm": 46.34102249145508, "learning_rate": 6.690891472868217e-06, "loss": 0.7402, "step": 2049 }, { "epoch": 0.3310723514211886, "grad_norm": 21.481342315673828, "learning_rate": 6.689276485788114e-06, "loss": 0.0594, "step": 2050 }, { "epoch": 0.331233850129199, "grad_norm": 38.20833969116211, "learning_rate": 6.687661498708011e-06, "loss": 0.4307, "step": 2051 }, { "epoch": 0.3313953488372093, "grad_norm": 73.77997589111328, "learning_rate": 6.686046511627907e-06, "loss": 0.9429, "step": 2052 }, { "epoch": 0.33155684754521964, "grad_norm": 16.254547119140625, "learning_rate": 6.6844315245478045e-06, "loss": 0.0863, "step": 2053 }, { "epoch": 0.33171834625323, "grad_norm": 58.528411865234375, "learning_rate": 6.682816537467701e-06, "loss": 1.2734, "step": 2054 }, { "epoch": 0.3318798449612403, "grad_norm": 32.12617111206055, "learning_rate": 6.681201550387597e-06, "loss": 0.1367, "step": 2055 }, { "epoch": 0.33204134366925064, "grad_norm": 12.527823448181152, "learning_rate": 6.679586563307494e-06, "loss": 0.1238, "step": 2056 }, { "epoch": 0.332202842377261, "grad_norm": 44.77598571777344, "learning_rate": 6.677971576227391e-06, "loss": 0.7837, "step": 2057 }, { "epoch": 0.3323643410852713, "grad_norm": 32.59293746948242, "learning_rate": 6.676356589147287e-06, "loss": 0.23, "step": 2058 }, { "epoch": 0.33252583979328165, "grad_norm": 3.6759188175201416, "learning_rate": 6.674741602067184e-06, "loss": 0.0315, "step": 2059 }, { "epoch": 0.332687338501292, "grad_norm": 21.219085693359375, "learning_rate": 6.673126614987081e-06, "loss": 0.1914, "step": 2060 }, { "epoch": 0.3328488372093023, "grad_norm": 20.61033821105957, "learning_rate": 6.671511627906977e-06, "loss": 0.2156, "step": 2061 }, { "epoch": 0.33301033591731266, "grad_norm": 43.402164459228516, "learning_rate": 6.6698966408268735e-06, "loss": 0.9883, "step": 2062 }, { "epoch": 0.333171834625323, "grad_norm": 120.4224853515625, "learning_rate": 6.668281653746771e-06, "loss": 2.1516, "step": 2063 }, { "epoch": 0.3333333333333333, "grad_norm": 20.071672439575195, "learning_rate": 6.666666666666667e-06, "loss": 0.0624, "step": 2064 }, { "epoch": 0.33349483204134367, "grad_norm": 121.2310562133789, "learning_rate": 6.665051679586564e-06, "loss": 2.7031, "step": 2065 }, { "epoch": 0.333656330749354, "grad_norm": 2.6366517543792725, "learning_rate": 6.663436692506461e-06, "loss": 0.0231, "step": 2066 }, { "epoch": 0.3338178294573643, "grad_norm": 92.55900573730469, "learning_rate": 6.661821705426357e-06, "loss": 1.8398, "step": 2067 }, { "epoch": 0.3339793281653747, "grad_norm": 74.49494171142578, "learning_rate": 6.660206718346254e-06, "loss": 1.5166, "step": 2068 }, { "epoch": 0.33414082687338503, "grad_norm": 108.40795135498047, "learning_rate": 6.658591731266151e-06, "loss": 2.1865, "step": 2069 }, { "epoch": 0.33430232558139533, "grad_norm": 53.70155334472656, "learning_rate": 6.656976744186047e-06, "loss": 0.3428, "step": 2070 }, { "epoch": 0.3344638242894057, "grad_norm": 54.728050231933594, "learning_rate": 6.655361757105943e-06, "loss": 0.8242, "step": 2071 }, { "epoch": 0.33462532299741604, "grad_norm": 165.19314575195312, "learning_rate": 6.653746770025841e-06, "loss": 2.9619, "step": 2072 }, { "epoch": 0.33478682170542634, "grad_norm": 8.796479225158691, "learning_rate": 6.6521317829457365e-06, "loss": 0.0624, "step": 2073 }, { "epoch": 0.3349483204134367, "grad_norm": 54.960411071777344, "learning_rate": 6.650516795865634e-06, "loss": 1.0137, "step": 2074 }, { "epoch": 0.33510981912144705, "grad_norm": 24.493669509887695, "learning_rate": 6.6489018087855305e-06, "loss": 0.2524, "step": 2075 }, { "epoch": 0.33527131782945735, "grad_norm": 20.42629623413086, "learning_rate": 6.647286821705426e-06, "loss": 0.3579, "step": 2076 }, { "epoch": 0.3354328165374677, "grad_norm": 50.00072479248047, "learning_rate": 6.645671834625324e-06, "loss": 0.5291, "step": 2077 }, { "epoch": 0.33559431524547806, "grad_norm": 93.80459594726562, "learning_rate": 6.64405684754522e-06, "loss": 1.1504, "step": 2078 }, { "epoch": 0.33575581395348836, "grad_norm": 10.913341522216797, "learning_rate": 6.642441860465117e-06, "loss": 0.1138, "step": 2079 }, { "epoch": 0.3359173126614987, "grad_norm": 22.96479034423828, "learning_rate": 6.640826873385013e-06, "loss": 0.3203, "step": 2080 }, { "epoch": 0.33607881136950907, "grad_norm": 70.95831298828125, "learning_rate": 6.639211886304911e-06, "loss": 0.7578, "step": 2081 }, { "epoch": 0.33624031007751937, "grad_norm": 47.869728088378906, "learning_rate": 6.637596899224806e-06, "loss": 0.8477, "step": 2082 }, { "epoch": 0.3364018087855297, "grad_norm": 148.49267578125, "learning_rate": 6.635981912144703e-06, "loss": 3.0, "step": 2083 }, { "epoch": 0.3365633074935401, "grad_norm": 54.326683044433594, "learning_rate": 6.6343669250646e-06, "loss": 0.312, "step": 2084 }, { "epoch": 0.3367248062015504, "grad_norm": 66.99519348144531, "learning_rate": 6.632751937984496e-06, "loss": 0.5891, "step": 2085 }, { "epoch": 0.33688630490956073, "grad_norm": 94.1563720703125, "learning_rate": 6.6311369509043935e-06, "loss": 1.2158, "step": 2086 }, { "epoch": 0.3370478036175711, "grad_norm": 10.36342716217041, "learning_rate": 6.62952196382429e-06, "loss": 0.1812, "step": 2087 }, { "epoch": 0.3372093023255814, "grad_norm": 88.93258666992188, "learning_rate": 6.627906976744186e-06, "loss": 1.6832, "step": 2088 }, { "epoch": 0.33737080103359174, "grad_norm": 56.74596405029297, "learning_rate": 6.626291989664083e-06, "loss": 0.8154, "step": 2089 }, { "epoch": 0.3375322997416021, "grad_norm": 50.461875915527344, "learning_rate": 6.62467700258398e-06, "loss": 0.4116, "step": 2090 }, { "epoch": 0.3376937984496124, "grad_norm": 125.23414611816406, "learning_rate": 6.623062015503876e-06, "loss": 1.5273, "step": 2091 }, { "epoch": 0.33785529715762275, "grad_norm": 51.25056838989258, "learning_rate": 6.621447028423773e-06, "loss": 0.7964, "step": 2092 }, { "epoch": 0.3380167958656331, "grad_norm": 47.59447479248047, "learning_rate": 6.61983204134367e-06, "loss": 0.6128, "step": 2093 }, { "epoch": 0.3381782945736434, "grad_norm": 34.25441360473633, "learning_rate": 6.618217054263566e-06, "loss": 0.5283, "step": 2094 }, { "epoch": 0.33833979328165376, "grad_norm": 70.68153381347656, "learning_rate": 6.616602067183463e-06, "loss": 1.8672, "step": 2095 }, { "epoch": 0.3385012919896641, "grad_norm": 171.3063507080078, "learning_rate": 6.61498708010336e-06, "loss": 3.3623, "step": 2096 }, { "epoch": 0.3386627906976744, "grad_norm": 41.76528549194336, "learning_rate": 6.613372093023256e-06, "loss": 0.22, "step": 2097 }, { "epoch": 0.33882428940568476, "grad_norm": 36.826786041259766, "learning_rate": 6.611757105943153e-06, "loss": 0.2344, "step": 2098 }, { "epoch": 0.33898578811369506, "grad_norm": 48.718833923339844, "learning_rate": 6.61014211886305e-06, "loss": 0.5891, "step": 2099 }, { "epoch": 0.3391472868217054, "grad_norm": 16.48591423034668, "learning_rate": 6.608527131782946e-06, "loss": 0.0723, "step": 2100 }, { "epoch": 0.3393087855297158, "grad_norm": 178.64901733398438, "learning_rate": 6.606912144702843e-06, "loss": 4.7188, "step": 2101 }, { "epoch": 0.33947028423772607, "grad_norm": 91.61321258544922, "learning_rate": 6.60529715762274e-06, "loss": 1.3394, "step": 2102 }, { "epoch": 0.3396317829457364, "grad_norm": 72.93411254882812, "learning_rate": 6.603682170542636e-06, "loss": 1.4678, "step": 2103 }, { "epoch": 0.3397932816537468, "grad_norm": 23.772119522094727, "learning_rate": 6.6020671834625324e-06, "loss": 0.1211, "step": 2104 }, { "epoch": 0.3399547803617571, "grad_norm": 34.73065948486328, "learning_rate": 6.60045219638243e-06, "loss": 0.5117, "step": 2105 }, { "epoch": 0.34011627906976744, "grad_norm": 12.533705711364746, "learning_rate": 6.5988372093023256e-06, "loss": 0.0559, "step": 2106 }, { "epoch": 0.3402777777777778, "grad_norm": 38.47639083862305, "learning_rate": 6.597222222222223e-06, "loss": 0.5804, "step": 2107 }, { "epoch": 0.3404392764857881, "grad_norm": 95.83769226074219, "learning_rate": 6.5956072351421195e-06, "loss": 2.75, "step": 2108 }, { "epoch": 0.34060077519379844, "grad_norm": 91.43690490722656, "learning_rate": 6.593992248062015e-06, "loss": 0.7983, "step": 2109 }, { "epoch": 0.3407622739018088, "grad_norm": 43.82231903076172, "learning_rate": 6.592377260981913e-06, "loss": 0.3877, "step": 2110 }, { "epoch": 0.3409237726098191, "grad_norm": 62.19240951538086, "learning_rate": 6.590762273901809e-06, "loss": 0.8848, "step": 2111 }, { "epoch": 0.34108527131782945, "grad_norm": 48.57513427734375, "learning_rate": 6.589147286821706e-06, "loss": 0.4824, "step": 2112 }, { "epoch": 0.3412467700258398, "grad_norm": 35.93342590332031, "learning_rate": 6.587532299741602e-06, "loss": 0.7051, "step": 2113 }, { "epoch": 0.3414082687338501, "grad_norm": 34.247703552246094, "learning_rate": 6.5859173126615e-06, "loss": 0.2456, "step": 2114 }, { "epoch": 0.34156976744186046, "grad_norm": 31.153087615966797, "learning_rate": 6.5843023255813954e-06, "loss": 0.5762, "step": 2115 }, { "epoch": 0.3417312661498708, "grad_norm": 46.14760971069336, "learning_rate": 6.582687338501292e-06, "loss": 0.7578, "step": 2116 }, { "epoch": 0.3418927648578811, "grad_norm": 51.395851135253906, "learning_rate": 6.581072351421189e-06, "loss": 0.6611, "step": 2117 }, { "epoch": 0.34205426356589147, "grad_norm": 45.40790557861328, "learning_rate": 6.579457364341085e-06, "loss": 0.4089, "step": 2118 }, { "epoch": 0.3422157622739018, "grad_norm": 29.982330322265625, "learning_rate": 6.5778423772609825e-06, "loss": 0.2588, "step": 2119 }, { "epoch": 0.3423772609819121, "grad_norm": 4.039842128753662, "learning_rate": 6.576227390180879e-06, "loss": 0.0298, "step": 2120 }, { "epoch": 0.3425387596899225, "grad_norm": 107.01802062988281, "learning_rate": 6.574612403100775e-06, "loss": 2.0039, "step": 2121 }, { "epoch": 0.34270025839793283, "grad_norm": 58.076175689697266, "learning_rate": 6.572997416020672e-06, "loss": 1.0615, "step": 2122 }, { "epoch": 0.34286175710594313, "grad_norm": 50.138648986816406, "learning_rate": 6.57138242894057e-06, "loss": 0.8232, "step": 2123 }, { "epoch": 0.3430232558139535, "grad_norm": 34.71946716308594, "learning_rate": 6.569767441860465e-06, "loss": 0.5024, "step": 2124 }, { "epoch": 0.34318475452196384, "grad_norm": 47.22257614135742, "learning_rate": 6.568152454780362e-06, "loss": 0.613, "step": 2125 }, { "epoch": 0.34334625322997414, "grad_norm": 49.699928283691406, "learning_rate": 6.566537467700259e-06, "loss": 1.0547, "step": 2126 }, { "epoch": 0.3435077519379845, "grad_norm": 25.809627532958984, "learning_rate": 6.564922480620155e-06, "loss": 0.1552, "step": 2127 }, { "epoch": 0.34366925064599485, "grad_norm": 69.52291870117188, "learning_rate": 6.563307493540052e-06, "loss": 0.8066, "step": 2128 }, { "epoch": 0.34383074935400515, "grad_norm": 49.231937408447266, "learning_rate": 6.561692506459949e-06, "loss": 0.791, "step": 2129 }, { "epoch": 0.3439922480620155, "grad_norm": 122.18907165527344, "learning_rate": 6.560077519379846e-06, "loss": 1.1875, "step": 2130 }, { "epoch": 0.34415374677002586, "grad_norm": 38.80152893066406, "learning_rate": 6.558462532299742e-06, "loss": 0.229, "step": 2131 }, { "epoch": 0.34431524547803616, "grad_norm": 156.11276245117188, "learning_rate": 6.556847545219639e-06, "loss": 2.5391, "step": 2132 }, { "epoch": 0.3444767441860465, "grad_norm": 48.26968765258789, "learning_rate": 6.555232558139536e-06, "loss": 0.6772, "step": 2133 }, { "epoch": 0.34463824289405687, "grad_norm": 14.054708480834961, "learning_rate": 6.553617571059432e-06, "loss": 0.1251, "step": 2134 }, { "epoch": 0.34479974160206717, "grad_norm": 104.88960266113281, "learning_rate": 6.552002583979329e-06, "loss": 1.2153, "step": 2135 }, { "epoch": 0.3449612403100775, "grad_norm": 113.71329498291016, "learning_rate": 6.550387596899226e-06, "loss": 2.0146, "step": 2136 }, { "epoch": 0.3451227390180879, "grad_norm": 41.94515609741211, "learning_rate": 6.5487726098191215e-06, "loss": 0.7703, "step": 2137 }, { "epoch": 0.3452842377260982, "grad_norm": 86.61846160888672, "learning_rate": 6.547157622739019e-06, "loss": 1.0233, "step": 2138 }, { "epoch": 0.34544573643410853, "grad_norm": 94.51773834228516, "learning_rate": 6.545542635658915e-06, "loss": 2.0234, "step": 2139 }, { "epoch": 0.3456072351421189, "grad_norm": 17.436166763305664, "learning_rate": 6.543927648578812e-06, "loss": 0.1309, "step": 2140 }, { "epoch": 0.3457687338501292, "grad_norm": 18.480934143066406, "learning_rate": 6.5423126614987085e-06, "loss": 0.1587, "step": 2141 }, { "epoch": 0.34593023255813954, "grad_norm": 13.871764183044434, "learning_rate": 6.540697674418606e-06, "loss": 0.0748, "step": 2142 }, { "epoch": 0.3460917312661499, "grad_norm": 27.472896575927734, "learning_rate": 6.539082687338502e-06, "loss": 0.3137, "step": 2143 }, { "epoch": 0.3462532299741602, "grad_norm": 60.13277053833008, "learning_rate": 6.537467700258398e-06, "loss": 1.1133, "step": 2144 }, { "epoch": 0.34641472868217055, "grad_norm": 61.159671783447266, "learning_rate": 6.535852713178296e-06, "loss": 0.4971, "step": 2145 }, { "epoch": 0.3465762273901809, "grad_norm": 137.22671508789062, "learning_rate": 6.534237726098191e-06, "loss": 1.5649, "step": 2146 }, { "epoch": 0.3467377260981912, "grad_norm": 82.50495910644531, "learning_rate": 6.532622739018089e-06, "loss": 0.834, "step": 2147 }, { "epoch": 0.34689922480620156, "grad_norm": 17.611793518066406, "learning_rate": 6.531007751937985e-06, "loss": 0.2598, "step": 2148 }, { "epoch": 0.3470607235142119, "grad_norm": 24.724042892456055, "learning_rate": 6.529392764857881e-06, "loss": 0.1431, "step": 2149 }, { "epoch": 0.3472222222222222, "grad_norm": 123.74290466308594, "learning_rate": 6.5277777777777784e-06, "loss": 3.9541, "step": 2150 }, { "epoch": 0.34738372093023256, "grad_norm": 36.68781280517578, "learning_rate": 6.526162790697676e-06, "loss": 0.3584, "step": 2151 }, { "epoch": 0.3475452196382429, "grad_norm": 67.90443420410156, "learning_rate": 6.5245478036175715e-06, "loss": 1.6149, "step": 2152 }, { "epoch": 0.3477067183462532, "grad_norm": 20.547327041625977, "learning_rate": 6.522932816537468e-06, "loss": 0.098, "step": 2153 }, { "epoch": 0.3478682170542636, "grad_norm": 29.507766723632812, "learning_rate": 6.5213178294573655e-06, "loss": 0.2483, "step": 2154 }, { "epoch": 0.34802971576227393, "grad_norm": 152.54139709472656, "learning_rate": 6.519702842377261e-06, "loss": 2.0918, "step": 2155 }, { "epoch": 0.3481912144702842, "grad_norm": 54.031654357910156, "learning_rate": 6.518087855297159e-06, "loss": 0.6572, "step": 2156 }, { "epoch": 0.3483527131782946, "grad_norm": 72.52198791503906, "learning_rate": 6.516472868217055e-06, "loss": 1.0898, "step": 2157 }, { "epoch": 0.34851421188630494, "grad_norm": 71.91697692871094, "learning_rate": 6.514857881136951e-06, "loss": 1.0156, "step": 2158 }, { "epoch": 0.34867571059431524, "grad_norm": 40.0117073059082, "learning_rate": 6.513242894056848e-06, "loss": 0.4609, "step": 2159 }, { "epoch": 0.3488372093023256, "grad_norm": 12.827279090881348, "learning_rate": 6.511627906976745e-06, "loss": 0.1411, "step": 2160 }, { "epoch": 0.3489987080103359, "grad_norm": 109.57475280761719, "learning_rate": 6.5100129198966414e-06, "loss": 3.0156, "step": 2161 }, { "epoch": 0.34916020671834624, "grad_norm": 52.31642150878906, "learning_rate": 6.508397932816538e-06, "loss": 0.5002, "step": 2162 }, { "epoch": 0.3493217054263566, "grad_norm": 151.90626525878906, "learning_rate": 6.506782945736435e-06, "loss": 1.9902, "step": 2163 }, { "epoch": 0.3494832041343669, "grad_norm": 29.767269134521484, "learning_rate": 6.505167958656331e-06, "loss": 0.4211, "step": 2164 }, { "epoch": 0.34964470284237725, "grad_norm": 19.963050842285156, "learning_rate": 6.503552971576228e-06, "loss": 0.1206, "step": 2165 }, { "epoch": 0.3498062015503876, "grad_norm": 29.935867309570312, "learning_rate": 6.501937984496125e-06, "loss": 0.5869, "step": 2166 }, { "epoch": 0.3499677002583979, "grad_norm": 28.76565933227539, "learning_rate": 6.500322997416021e-06, "loss": 0.179, "step": 2167 }, { "epoch": 0.35012919896640826, "grad_norm": 70.62157440185547, "learning_rate": 6.498708010335918e-06, "loss": 1.0898, "step": 2168 }, { "epoch": 0.3502906976744186, "grad_norm": 38.928245544433594, "learning_rate": 6.497093023255815e-06, "loss": 0.583, "step": 2169 }, { "epoch": 0.3504521963824289, "grad_norm": 38.12440872192383, "learning_rate": 6.4954780361757105e-06, "loss": 0.4009, "step": 2170 }, { "epoch": 0.35061369509043927, "grad_norm": 16.79375648498535, "learning_rate": 6.493863049095608e-06, "loss": 0.0906, "step": 2171 }, { "epoch": 0.3507751937984496, "grad_norm": 11.67273235321045, "learning_rate": 6.4922480620155044e-06, "loss": 0.0976, "step": 2172 }, { "epoch": 0.3509366925064599, "grad_norm": 37.579593658447266, "learning_rate": 6.490633074935401e-06, "loss": 0.3282, "step": 2173 }, { "epoch": 0.3510981912144703, "grad_norm": 48.48765563964844, "learning_rate": 6.4890180878552976e-06, "loss": 0.3262, "step": 2174 }, { "epoch": 0.35125968992248063, "grad_norm": 70.97402954101562, "learning_rate": 6.487403100775195e-06, "loss": 0.9229, "step": 2175 }, { "epoch": 0.35142118863049093, "grad_norm": 139.8714141845703, "learning_rate": 6.485788113695091e-06, "loss": 3.0302, "step": 2176 }, { "epoch": 0.3515826873385013, "grad_norm": 34.83955001831055, "learning_rate": 6.484173126614987e-06, "loss": 0.307, "step": 2177 }, { "epoch": 0.35174418604651164, "grad_norm": 46.11766815185547, "learning_rate": 6.482558139534885e-06, "loss": 0.3955, "step": 2178 }, { "epoch": 0.35190568475452194, "grad_norm": 124.6083755493164, "learning_rate": 6.48094315245478e-06, "loss": 1.4498, "step": 2179 }, { "epoch": 0.3520671834625323, "grad_norm": 59.55065155029297, "learning_rate": 6.479328165374678e-06, "loss": 0.4102, "step": 2180 }, { "epoch": 0.35222868217054265, "grad_norm": 45.3599739074707, "learning_rate": 6.477713178294574e-06, "loss": 0.6245, "step": 2181 }, { "epoch": 0.35239018087855295, "grad_norm": 42.9569206237793, "learning_rate": 6.476098191214471e-06, "loss": 0.3442, "step": 2182 }, { "epoch": 0.3525516795865633, "grad_norm": 51.507328033447266, "learning_rate": 6.4744832041343675e-06, "loss": 0.635, "step": 2183 }, { "epoch": 0.35271317829457366, "grad_norm": 67.08146667480469, "learning_rate": 6.472868217054265e-06, "loss": 0.7441, "step": 2184 }, { "epoch": 0.35287467700258396, "grad_norm": 113.331787109375, "learning_rate": 6.4712532299741606e-06, "loss": 1.4033, "step": 2185 }, { "epoch": 0.3530361757105943, "grad_norm": 13.33896255493164, "learning_rate": 6.469638242894057e-06, "loss": 0.0654, "step": 2186 }, { "epoch": 0.35319767441860467, "grad_norm": 102.17230224609375, "learning_rate": 6.4680232558139545e-06, "loss": 1.4199, "step": 2187 }, { "epoch": 0.35335917312661497, "grad_norm": 56.782745361328125, "learning_rate": 6.46640826873385e-06, "loss": 0.6403, "step": 2188 }, { "epoch": 0.3535206718346253, "grad_norm": 191.45404052734375, "learning_rate": 6.464793281653748e-06, "loss": 4.1367, "step": 2189 }, { "epoch": 0.3536821705426357, "grad_norm": 92.96929931640625, "learning_rate": 6.463178294573644e-06, "loss": 1.2643, "step": 2190 }, { "epoch": 0.353843669250646, "grad_norm": 146.17962646484375, "learning_rate": 6.46156330749354e-06, "loss": 1.4844, "step": 2191 }, { "epoch": 0.35400516795865633, "grad_norm": 23.20313835144043, "learning_rate": 6.459948320413437e-06, "loss": 0.1313, "step": 2192 }, { "epoch": 0.3541666666666667, "grad_norm": 42.677833557128906, "learning_rate": 6.458333333333334e-06, "loss": 1.1436, "step": 2193 }, { "epoch": 0.354328165374677, "grad_norm": 45.51493453979492, "learning_rate": 6.4567183462532305e-06, "loss": 0.3057, "step": 2194 }, { "epoch": 0.35448966408268734, "grad_norm": 71.79712677001953, "learning_rate": 6.455103359173127e-06, "loss": 1.1816, "step": 2195 }, { "epoch": 0.3546511627906977, "grad_norm": 61.666259765625, "learning_rate": 6.4534883720930244e-06, "loss": 0.8066, "step": 2196 }, { "epoch": 0.354812661498708, "grad_norm": 63.234066009521484, "learning_rate": 6.45187338501292e-06, "loss": 1.2119, "step": 2197 }, { "epoch": 0.35497416020671835, "grad_norm": 127.9188232421875, "learning_rate": 6.450258397932817e-06, "loss": 1.7578, "step": 2198 }, { "epoch": 0.3551356589147287, "grad_norm": 9.57427978515625, "learning_rate": 6.448643410852714e-06, "loss": 0.0757, "step": 2199 }, { "epoch": 0.355297157622739, "grad_norm": 62.152122497558594, "learning_rate": 6.44702842377261e-06, "loss": 0.8984, "step": 2200 }, { "epoch": 0.35545865633074936, "grad_norm": 19.293489456176758, "learning_rate": 6.445413436692507e-06, "loss": 0.078, "step": 2201 }, { "epoch": 0.3556201550387597, "grad_norm": 130.18963623046875, "learning_rate": 6.443798449612404e-06, "loss": 2.3984, "step": 2202 }, { "epoch": 0.35578165374677, "grad_norm": 40.46199417114258, "learning_rate": 6.4421834625322995e-06, "loss": 0.564, "step": 2203 }, { "epoch": 0.35594315245478036, "grad_norm": 22.899137496948242, "learning_rate": 6.440568475452197e-06, "loss": 0.2318, "step": 2204 }, { "epoch": 0.3561046511627907, "grad_norm": 76.42522430419922, "learning_rate": 6.4389534883720935e-06, "loss": 0.5972, "step": 2205 }, { "epoch": 0.356266149870801, "grad_norm": 15.175477981567383, "learning_rate": 6.43733850129199e-06, "loss": 0.2283, "step": 2206 }, { "epoch": 0.3564276485788114, "grad_norm": 33.937774658203125, "learning_rate": 6.435723514211887e-06, "loss": 0.2998, "step": 2207 }, { "epoch": 0.35658914728682173, "grad_norm": 44.387210845947266, "learning_rate": 6.434108527131784e-06, "loss": 0.4116, "step": 2208 }, { "epoch": 0.356750645994832, "grad_norm": 31.9627685546875, "learning_rate": 6.43249354005168e-06, "loss": 0.9219, "step": 2209 }, { "epoch": 0.3569121447028424, "grad_norm": 54.90498352050781, "learning_rate": 6.430878552971577e-06, "loss": 1.4355, "step": 2210 }, { "epoch": 0.35707364341085274, "grad_norm": 41.65192413330078, "learning_rate": 6.429263565891474e-06, "loss": 0.2065, "step": 2211 }, { "epoch": 0.35723514211886304, "grad_norm": 92.64990234375, "learning_rate": 6.427648578811369e-06, "loss": 1.4297, "step": 2212 }, { "epoch": 0.3573966408268734, "grad_norm": 81.29757690429688, "learning_rate": 6.426033591731267e-06, "loss": 0.957, "step": 2213 }, { "epoch": 0.35755813953488375, "grad_norm": 72.34320831298828, "learning_rate": 6.424418604651163e-06, "loss": 1.5645, "step": 2214 }, { "epoch": 0.35771963824289404, "grad_norm": 36.4597282409668, "learning_rate": 6.42280361757106e-06, "loss": 0.6875, "step": 2215 }, { "epoch": 0.3578811369509044, "grad_norm": 26.778974533081055, "learning_rate": 6.4211886304909565e-06, "loss": 0.442, "step": 2216 }, { "epoch": 0.35804263565891475, "grad_norm": 44.4640007019043, "learning_rate": 6.419573643410854e-06, "loss": 1.5586, "step": 2217 }, { "epoch": 0.35820413436692505, "grad_norm": 24.459383010864258, "learning_rate": 6.41795865633075e-06, "loss": 0.1741, "step": 2218 }, { "epoch": 0.3583656330749354, "grad_norm": 62.49545669555664, "learning_rate": 6.416343669250646e-06, "loss": 1.0898, "step": 2219 }, { "epoch": 0.35852713178294576, "grad_norm": 35.13343048095703, "learning_rate": 6.4147286821705436e-06, "loss": 0.1968, "step": 2220 }, { "epoch": 0.35868863049095606, "grad_norm": 22.102718353271484, "learning_rate": 6.413113695090439e-06, "loss": 0.3159, "step": 2221 }, { "epoch": 0.3588501291989664, "grad_norm": 24.63499641418457, "learning_rate": 6.411498708010337e-06, "loss": 0.2695, "step": 2222 }, { "epoch": 0.35901162790697677, "grad_norm": 50.570350646972656, "learning_rate": 6.409883720930233e-06, "loss": 0.4814, "step": 2223 }, { "epoch": 0.35917312661498707, "grad_norm": 143.31524658203125, "learning_rate": 6.408268733850129e-06, "loss": 5.5547, "step": 2224 }, { "epoch": 0.3593346253229974, "grad_norm": 62.573768615722656, "learning_rate": 6.406653746770026e-06, "loss": 0.7334, "step": 2225 }, { "epoch": 0.3594961240310077, "grad_norm": 84.23763275146484, "learning_rate": 6.405038759689923e-06, "loss": 1.0859, "step": 2226 }, { "epoch": 0.3596576227390181, "grad_norm": 60.15433120727539, "learning_rate": 6.4034237726098195e-06, "loss": 0.7871, "step": 2227 }, { "epoch": 0.35981912144702843, "grad_norm": 48.741943359375, "learning_rate": 6.401808785529716e-06, "loss": 0.9055, "step": 2228 }, { "epoch": 0.35998062015503873, "grad_norm": 62.8432502746582, "learning_rate": 6.4001937984496134e-06, "loss": 1.0391, "step": 2229 }, { "epoch": 0.3601421188630491, "grad_norm": 15.378423690795898, "learning_rate": 6.398578811369509e-06, "loss": 0.2383, "step": 2230 }, { "epoch": 0.36030361757105944, "grad_norm": 83.70903015136719, "learning_rate": 6.396963824289406e-06, "loss": 1.6484, "step": 2231 }, { "epoch": 0.36046511627906974, "grad_norm": 131.3368682861328, "learning_rate": 6.395348837209303e-06, "loss": 2.6602, "step": 2232 }, { "epoch": 0.3606266149870801, "grad_norm": 28.16475486755371, "learning_rate": 6.393733850129199e-06, "loss": 0.5253, "step": 2233 }, { "epoch": 0.36078811369509045, "grad_norm": 141.62921142578125, "learning_rate": 6.392118863049096e-06, "loss": 5.3906, "step": 2234 }, { "epoch": 0.36094961240310075, "grad_norm": 126.03339385986328, "learning_rate": 6.390503875968993e-06, "loss": 1.6406, "step": 2235 }, { "epoch": 0.3611111111111111, "grad_norm": 29.496482849121094, "learning_rate": 6.3888888888888885e-06, "loss": 0.2266, "step": 2236 }, { "epoch": 0.36127260981912146, "grad_norm": 42.397125244140625, "learning_rate": 6.387273901808786e-06, "loss": 0.3477, "step": 2237 }, { "epoch": 0.36143410852713176, "grad_norm": 104.88844299316406, "learning_rate": 6.385658914728683e-06, "loss": 1.9102, "step": 2238 }, { "epoch": 0.3615956072351421, "grad_norm": 135.28094482421875, "learning_rate": 6.384043927648579e-06, "loss": 3.875, "step": 2239 }, { "epoch": 0.36175710594315247, "grad_norm": 84.85322570800781, "learning_rate": 6.382428940568476e-06, "loss": 1.7031, "step": 2240 }, { "epoch": 0.36191860465116277, "grad_norm": 19.218765258789062, "learning_rate": 6.380813953488373e-06, "loss": 0.2566, "step": 2241 }, { "epoch": 0.3620801033591731, "grad_norm": 58.6144905090332, "learning_rate": 6.379198966408269e-06, "loss": 0.8462, "step": 2242 }, { "epoch": 0.3622416020671835, "grad_norm": 127.36767578125, "learning_rate": 6.377583979328166e-06, "loss": 2.1182, "step": 2243 }, { "epoch": 0.3624031007751938, "grad_norm": 62.37018966674805, "learning_rate": 6.375968992248063e-06, "loss": 0.4307, "step": 2244 }, { "epoch": 0.36256459948320413, "grad_norm": 107.26371002197266, "learning_rate": 6.374354005167958e-06, "loss": 1.7192, "step": 2245 }, { "epoch": 0.3627260981912145, "grad_norm": 102.11228942871094, "learning_rate": 6.372739018087856e-06, "loss": 2.416, "step": 2246 }, { "epoch": 0.3628875968992248, "grad_norm": 57.42615509033203, "learning_rate": 6.371124031007752e-06, "loss": 1.0322, "step": 2247 }, { "epoch": 0.36304909560723514, "grad_norm": 38.941654205322266, "learning_rate": 6.369509043927649e-06, "loss": 0.4082, "step": 2248 }, { "epoch": 0.3632105943152455, "grad_norm": 87.67349243164062, "learning_rate": 6.3678940568475455e-06, "loss": 1.5664, "step": 2249 }, { "epoch": 0.3633720930232558, "grad_norm": 22.131786346435547, "learning_rate": 6.366279069767443e-06, "loss": 0.2988, "step": 2250 }, { "epoch": 0.36353359173126615, "grad_norm": 121.31550598144531, "learning_rate": 6.364664082687339e-06, "loss": 1.8851, "step": 2251 }, { "epoch": 0.3636950904392765, "grad_norm": 32.06058883666992, "learning_rate": 6.363049095607235e-06, "loss": 0.6182, "step": 2252 }, { "epoch": 0.3638565891472868, "grad_norm": 34.88571548461914, "learning_rate": 6.361434108527133e-06, "loss": 0.2812, "step": 2253 }, { "epoch": 0.36401808785529716, "grad_norm": 6.403223991394043, "learning_rate": 6.359819121447028e-06, "loss": 0.0659, "step": 2254 }, { "epoch": 0.3641795865633075, "grad_norm": 58.08830261230469, "learning_rate": 6.358204134366926e-06, "loss": 0.9404, "step": 2255 }, { "epoch": 0.3643410852713178, "grad_norm": 36.23179244995117, "learning_rate": 6.356589147286822e-06, "loss": 0.9146, "step": 2256 }, { "epoch": 0.36450258397932817, "grad_norm": 23.684425354003906, "learning_rate": 6.354974160206718e-06, "loss": 0.314, "step": 2257 }, { "epoch": 0.3646640826873385, "grad_norm": 21.60593032836914, "learning_rate": 6.353359173126615e-06, "loss": 0.2087, "step": 2258 }, { "epoch": 0.3648255813953488, "grad_norm": 9.06408977508545, "learning_rate": 6.351744186046512e-06, "loss": 0.1094, "step": 2259 }, { "epoch": 0.3649870801033592, "grad_norm": 68.18714904785156, "learning_rate": 6.350129198966409e-06, "loss": 0.7012, "step": 2260 }, { "epoch": 0.36514857881136953, "grad_norm": 11.942947387695312, "learning_rate": 6.348514211886305e-06, "loss": 0.092, "step": 2261 }, { "epoch": 0.3653100775193798, "grad_norm": 93.76091766357422, "learning_rate": 6.3468992248062025e-06, "loss": 1.6602, "step": 2262 }, { "epoch": 0.3654715762273902, "grad_norm": 18.955768585205078, "learning_rate": 6.345284237726099e-06, "loss": 0.0591, "step": 2263 }, { "epoch": 0.36563307493540054, "grad_norm": 96.43118286132812, "learning_rate": 6.343669250645995e-06, "loss": 1.0161, "step": 2264 }, { "epoch": 0.36579457364341084, "grad_norm": 61.498756408691406, "learning_rate": 6.342054263565892e-06, "loss": 0.9727, "step": 2265 }, { "epoch": 0.3659560723514212, "grad_norm": 55.35329055786133, "learning_rate": 6.3404392764857896e-06, "loss": 0.7378, "step": 2266 }, { "epoch": 0.36611757105943155, "grad_norm": 30.058311462402344, "learning_rate": 6.338824289405685e-06, "loss": 0.2202, "step": 2267 }, { "epoch": 0.36627906976744184, "grad_norm": 57.40587615966797, "learning_rate": 6.337209302325582e-06, "loss": 0.8013, "step": 2268 }, { "epoch": 0.3664405684754522, "grad_norm": 54.36941909790039, "learning_rate": 6.335594315245479e-06, "loss": 0.291, "step": 2269 }, { "epoch": 0.36660206718346255, "grad_norm": 38.66282653808594, "learning_rate": 6.333979328165375e-06, "loss": 0.4009, "step": 2270 }, { "epoch": 0.36676356589147285, "grad_norm": 20.69089698791504, "learning_rate": 6.332364341085272e-06, "loss": 0.0642, "step": 2271 }, { "epoch": 0.3669250645994832, "grad_norm": 14.264996528625488, "learning_rate": 6.330749354005169e-06, "loss": 0.0538, "step": 2272 }, { "epoch": 0.36708656330749356, "grad_norm": 125.76214599609375, "learning_rate": 6.329134366925065e-06, "loss": 1.9727, "step": 2273 }, { "epoch": 0.36724806201550386, "grad_norm": 53.89142990112305, "learning_rate": 6.327519379844962e-06, "loss": 0.7603, "step": 2274 }, { "epoch": 0.3674095607235142, "grad_norm": 14.001379013061523, "learning_rate": 6.325904392764859e-06, "loss": 0.1003, "step": 2275 }, { "epoch": 0.36757105943152457, "grad_norm": 128.32122802734375, "learning_rate": 6.324289405684755e-06, "loss": 1.1172, "step": 2276 }, { "epoch": 0.36773255813953487, "grad_norm": 99.20806884765625, "learning_rate": 6.322674418604652e-06, "loss": 0.8535, "step": 2277 }, { "epoch": 0.3678940568475452, "grad_norm": 26.597768783569336, "learning_rate": 6.321059431524549e-06, "loss": 0.3257, "step": 2278 }, { "epoch": 0.3680555555555556, "grad_norm": 131.13429260253906, "learning_rate": 6.319444444444445e-06, "loss": 2.3945, "step": 2279 }, { "epoch": 0.3682170542635659, "grad_norm": 9.328413009643555, "learning_rate": 6.317829457364341e-06, "loss": 0.08, "step": 2280 }, { "epoch": 0.36837855297157623, "grad_norm": 118.53802490234375, "learning_rate": 6.316214470284239e-06, "loss": 1.3086, "step": 2281 }, { "epoch": 0.3685400516795866, "grad_norm": 47.35447692871094, "learning_rate": 6.3145994832041345e-06, "loss": 0.5566, "step": 2282 }, { "epoch": 0.3687015503875969, "grad_norm": 59.008419036865234, "learning_rate": 6.312984496124032e-06, "loss": 0.8291, "step": 2283 }, { "epoch": 0.36886304909560724, "grad_norm": 20.49928092956543, "learning_rate": 6.3113695090439285e-06, "loss": 0.1932, "step": 2284 }, { "epoch": 0.3690245478036176, "grad_norm": 79.3552474975586, "learning_rate": 6.309754521963824e-06, "loss": 1.0706, "step": 2285 }, { "epoch": 0.3691860465116279, "grad_norm": 145.65638732910156, "learning_rate": 6.308139534883722e-06, "loss": 2.1729, "step": 2286 }, { "epoch": 0.36934754521963825, "grad_norm": 29.756568908691406, "learning_rate": 6.306524547803618e-06, "loss": 0.2061, "step": 2287 }, { "epoch": 0.3695090439276486, "grad_norm": 46.71205520629883, "learning_rate": 6.304909560723515e-06, "loss": 0.4192, "step": 2288 }, { "epoch": 0.3696705426356589, "grad_norm": 36.01832580566406, "learning_rate": 6.303294573643411e-06, "loss": 0.2363, "step": 2289 }, { "epoch": 0.36983204134366926, "grad_norm": 37.851341247558594, "learning_rate": 6.301679586563309e-06, "loss": 0.2095, "step": 2290 }, { "epoch": 0.36999354005167956, "grad_norm": 50.895782470703125, "learning_rate": 6.300064599483204e-06, "loss": 0.3115, "step": 2291 }, { "epoch": 0.3701550387596899, "grad_norm": 43.22266387939453, "learning_rate": 6.298449612403101e-06, "loss": 0.416, "step": 2292 }, { "epoch": 0.37031653746770027, "grad_norm": 142.91273498535156, "learning_rate": 6.296834625322998e-06, "loss": 3.3047, "step": 2293 }, { "epoch": 0.37047803617571057, "grad_norm": 18.22423553466797, "learning_rate": 6.295219638242894e-06, "loss": 0.0807, "step": 2294 }, { "epoch": 0.3706395348837209, "grad_norm": 29.512462615966797, "learning_rate": 6.2936046511627915e-06, "loss": 0.196, "step": 2295 }, { "epoch": 0.3708010335917313, "grad_norm": 71.12730407714844, "learning_rate": 6.291989664082688e-06, "loss": 0.8125, "step": 2296 }, { "epoch": 0.3709625322997416, "grad_norm": 44.226566314697266, "learning_rate": 6.290374677002585e-06, "loss": 0.4004, "step": 2297 }, { "epoch": 0.37112403100775193, "grad_norm": 170.79151916503906, "learning_rate": 6.288759689922481e-06, "loss": 4.2773, "step": 2298 }, { "epoch": 0.3712855297157623, "grad_norm": 144.667236328125, "learning_rate": 6.2871447028423786e-06, "loss": 0.581, "step": 2299 }, { "epoch": 0.3714470284237726, "grad_norm": 70.02876281738281, "learning_rate": 6.285529715762274e-06, "loss": 1.416, "step": 2300 }, { "epoch": 0.37160852713178294, "grad_norm": 158.53726196289062, "learning_rate": 6.283914728682171e-06, "loss": 1.4581, "step": 2301 }, { "epoch": 0.3717700258397933, "grad_norm": 52.35844039916992, "learning_rate": 6.282299741602068e-06, "loss": 0.3999, "step": 2302 }, { "epoch": 0.3719315245478036, "grad_norm": 81.39667510986328, "learning_rate": 6.280684754521964e-06, "loss": 0.9023, "step": 2303 }, { "epoch": 0.37209302325581395, "grad_norm": 91.2499008178711, "learning_rate": 6.279069767441861e-06, "loss": 1.1138, "step": 2304 }, { "epoch": 0.3722545219638243, "grad_norm": 58.07764434814453, "learning_rate": 6.277454780361758e-06, "loss": 0.7246, "step": 2305 }, { "epoch": 0.3724160206718346, "grad_norm": 13.374954223632812, "learning_rate": 6.275839793281654e-06, "loss": 0.0696, "step": 2306 }, { "epoch": 0.37257751937984496, "grad_norm": 14.730239868164062, "learning_rate": 6.274224806201551e-06, "loss": 0.0591, "step": 2307 }, { "epoch": 0.3727390180878553, "grad_norm": 132.43272399902344, "learning_rate": 6.272609819121448e-06, "loss": 2.1311, "step": 2308 }, { "epoch": 0.3729005167958656, "grad_norm": 99.2610092163086, "learning_rate": 6.270994832041344e-06, "loss": 1.5117, "step": 2309 }, { "epoch": 0.37306201550387597, "grad_norm": 57.71018600463867, "learning_rate": 6.269379844961241e-06, "loss": 0.5195, "step": 2310 }, { "epoch": 0.3732235142118863, "grad_norm": 50.4673957824707, "learning_rate": 6.267764857881138e-06, "loss": 0.6484, "step": 2311 }, { "epoch": 0.3733850129198966, "grad_norm": 64.82830810546875, "learning_rate": 6.266149870801034e-06, "loss": 0.8496, "step": 2312 }, { "epoch": 0.373546511627907, "grad_norm": 67.81531524658203, "learning_rate": 6.26453488372093e-06, "loss": 0.6348, "step": 2313 }, { "epoch": 0.37370801033591733, "grad_norm": 65.57093811035156, "learning_rate": 6.262919896640828e-06, "loss": 0.5771, "step": 2314 }, { "epoch": 0.37386950904392763, "grad_norm": 76.50149536132812, "learning_rate": 6.2613049095607235e-06, "loss": 1.0625, "step": 2315 }, { "epoch": 0.374031007751938, "grad_norm": 129.7333221435547, "learning_rate": 6.259689922480621e-06, "loss": 1.4111, "step": 2316 }, { "epoch": 0.37419250645994834, "grad_norm": 34.44402313232422, "learning_rate": 6.2580749354005175e-06, "loss": 0.1456, "step": 2317 }, { "epoch": 0.37435400516795864, "grad_norm": 79.90435791015625, "learning_rate": 6.256459948320413e-06, "loss": 0.6318, "step": 2318 }, { "epoch": 0.374515503875969, "grad_norm": 87.6480941772461, "learning_rate": 6.254844961240311e-06, "loss": 1.291, "step": 2319 }, { "epoch": 0.37467700258397935, "grad_norm": 67.98696899414062, "learning_rate": 6.253229974160208e-06, "loss": 0.5234, "step": 2320 }, { "epoch": 0.37483850129198965, "grad_norm": 103.469482421875, "learning_rate": 6.251614987080104e-06, "loss": 1.0186, "step": 2321 }, { "epoch": 0.375, "grad_norm": 77.45037078857422, "learning_rate": 6.25e-06, "loss": 0.577, "step": 2322 }, { "epoch": 0.37516149870801035, "grad_norm": 112.79857635498047, "learning_rate": 6.248385012919898e-06, "loss": 1.8867, "step": 2323 }, { "epoch": 0.37532299741602065, "grad_norm": 150.65420532226562, "learning_rate": 6.246770025839793e-06, "loss": 2.0234, "step": 2324 }, { "epoch": 0.375484496124031, "grad_norm": 20.78785514831543, "learning_rate": 6.245155038759691e-06, "loss": 0.2533, "step": 2325 }, { "epoch": 0.37564599483204136, "grad_norm": 47.25862121582031, "learning_rate": 6.243540051679587e-06, "loss": 0.6382, "step": 2326 }, { "epoch": 0.37580749354005166, "grad_norm": 69.11727142333984, "learning_rate": 6.241925064599483e-06, "loss": 0.5762, "step": 2327 }, { "epoch": 0.375968992248062, "grad_norm": 28.755525588989258, "learning_rate": 6.2403100775193805e-06, "loss": 0.2213, "step": 2328 }, { "epoch": 0.37613049095607237, "grad_norm": 69.28251647949219, "learning_rate": 6.238695090439277e-06, "loss": 0.7031, "step": 2329 }, { "epoch": 0.37629198966408267, "grad_norm": 26.297225952148438, "learning_rate": 6.237080103359174e-06, "loss": 0.3345, "step": 2330 }, { "epoch": 0.376453488372093, "grad_norm": 94.30638885498047, "learning_rate": 6.23546511627907e-06, "loss": 1.8398, "step": 2331 }, { "epoch": 0.3766149870801034, "grad_norm": 126.664794921875, "learning_rate": 6.233850129198968e-06, "loss": 0.916, "step": 2332 }, { "epoch": 0.3767764857881137, "grad_norm": 94.12515258789062, "learning_rate": 6.232235142118863e-06, "loss": 1.8828, "step": 2333 }, { "epoch": 0.37693798449612403, "grad_norm": 31.670703887939453, "learning_rate": 6.23062015503876e-06, "loss": 0.1562, "step": 2334 }, { "epoch": 0.3770994832041344, "grad_norm": 84.6046371459961, "learning_rate": 6.229005167958657e-06, "loss": 1.2227, "step": 2335 }, { "epoch": 0.3772609819121447, "grad_norm": 30.03165626525879, "learning_rate": 6.227390180878553e-06, "loss": 0.2607, "step": 2336 }, { "epoch": 0.37742248062015504, "grad_norm": 65.00210571289062, "learning_rate": 6.22577519379845e-06, "loss": 0.6152, "step": 2337 }, { "epoch": 0.3775839793281654, "grad_norm": 32.08143997192383, "learning_rate": 6.224160206718347e-06, "loss": 0.4224, "step": 2338 }, { "epoch": 0.3777454780361757, "grad_norm": 27.94882583618164, "learning_rate": 6.222545219638243e-06, "loss": 0.2354, "step": 2339 }, { "epoch": 0.37790697674418605, "grad_norm": 133.85980224609375, "learning_rate": 6.22093023255814e-06, "loss": 1.7588, "step": 2340 }, { "epoch": 0.3780684754521964, "grad_norm": 54.39912796020508, "learning_rate": 6.219315245478037e-06, "loss": 0.3108, "step": 2341 }, { "epoch": 0.3782299741602067, "grad_norm": 41.89564514160156, "learning_rate": 6.217700258397933e-06, "loss": 0.4824, "step": 2342 }, { "epoch": 0.37839147286821706, "grad_norm": 67.69287872314453, "learning_rate": 6.21608527131783e-06, "loss": 0.6067, "step": 2343 }, { "epoch": 0.3785529715762274, "grad_norm": 84.6852798461914, "learning_rate": 6.214470284237727e-06, "loss": 1.3555, "step": 2344 }, { "epoch": 0.3787144702842377, "grad_norm": 7.361351013183594, "learning_rate": 6.212855297157623e-06, "loss": 0.0228, "step": 2345 }, { "epoch": 0.37887596899224807, "grad_norm": 37.64358139038086, "learning_rate": 6.2112403100775194e-06, "loss": 0.4072, "step": 2346 }, { "epoch": 0.3790374677002584, "grad_norm": 44.3711051940918, "learning_rate": 6.209625322997417e-06, "loss": 0.2363, "step": 2347 }, { "epoch": 0.3791989664082687, "grad_norm": 47.20055389404297, "learning_rate": 6.2080103359173126e-06, "loss": 0.3936, "step": 2348 }, { "epoch": 0.3793604651162791, "grad_norm": 150.37977600097656, "learning_rate": 6.20639534883721e-06, "loss": 3.7383, "step": 2349 }, { "epoch": 0.37952196382428943, "grad_norm": 112.72986602783203, "learning_rate": 6.2047803617571065e-06, "loss": 1.332, "step": 2350 }, { "epoch": 0.37968346253229973, "grad_norm": 60.49261474609375, "learning_rate": 6.203165374677002e-06, "loss": 0.6416, "step": 2351 }, { "epoch": 0.3798449612403101, "grad_norm": 107.0880126953125, "learning_rate": 6.2015503875969e-06, "loss": 1.6016, "step": 2352 }, { "epoch": 0.38000645994832044, "grad_norm": 50.133148193359375, "learning_rate": 6.199935400516797e-06, "loss": 0.5601, "step": 2353 }, { "epoch": 0.38016795865633074, "grad_norm": 55.10967254638672, "learning_rate": 6.198320413436693e-06, "loss": 0.5083, "step": 2354 }, { "epoch": 0.3803294573643411, "grad_norm": 23.452510833740234, "learning_rate": 6.196705426356589e-06, "loss": 0.1987, "step": 2355 }, { "epoch": 0.3804909560723514, "grad_norm": 56.72646713256836, "learning_rate": 6.195090439276487e-06, "loss": 0.6768, "step": 2356 }, { "epoch": 0.38065245478036175, "grad_norm": 22.616031646728516, "learning_rate": 6.1934754521963824e-06, "loss": 0.2332, "step": 2357 }, { "epoch": 0.3808139534883721, "grad_norm": 5.636739253997803, "learning_rate": 6.19186046511628e-06, "loss": 0.0343, "step": 2358 }, { "epoch": 0.3809754521963824, "grad_norm": 48.24698257446289, "learning_rate": 6.190245478036176e-06, "loss": 0.519, "step": 2359 }, { "epoch": 0.38113695090439276, "grad_norm": 111.98689270019531, "learning_rate": 6.188630490956072e-06, "loss": 1.3535, "step": 2360 }, { "epoch": 0.3812984496124031, "grad_norm": 17.021041870117188, "learning_rate": 6.1870155038759695e-06, "loss": 0.1387, "step": 2361 }, { "epoch": 0.3814599483204134, "grad_norm": 34.11575698852539, "learning_rate": 6.185400516795866e-06, "loss": 0.2048, "step": 2362 }, { "epoch": 0.38162144702842377, "grad_norm": 77.74372100830078, "learning_rate": 6.183785529715763e-06, "loss": 0.9766, "step": 2363 }, { "epoch": 0.3817829457364341, "grad_norm": 118.050537109375, "learning_rate": 6.182170542635659e-06, "loss": 2.2998, "step": 2364 }, { "epoch": 0.3819444444444444, "grad_norm": 41.114925384521484, "learning_rate": 6.180555555555557e-06, "loss": 0.4219, "step": 2365 }, { "epoch": 0.3821059431524548, "grad_norm": 25.750274658203125, "learning_rate": 6.178940568475452e-06, "loss": 0.105, "step": 2366 }, { "epoch": 0.38226744186046513, "grad_norm": 108.90617370605469, "learning_rate": 6.177325581395349e-06, "loss": 1.2607, "step": 2367 }, { "epoch": 0.38242894056847543, "grad_norm": 26.878446578979492, "learning_rate": 6.175710594315246e-06, "loss": 0.1591, "step": 2368 }, { "epoch": 0.3825904392764858, "grad_norm": 142.5294647216797, "learning_rate": 6.174095607235142e-06, "loss": 1.6295, "step": 2369 }, { "epoch": 0.38275193798449614, "grad_norm": 78.7647705078125, "learning_rate": 6.172480620155039e-06, "loss": 0.8994, "step": 2370 }, { "epoch": 0.38291343669250644, "grad_norm": 98.38365936279297, "learning_rate": 6.170865633074936e-06, "loss": 0.8833, "step": 2371 }, { "epoch": 0.3830749354005168, "grad_norm": 82.21161651611328, "learning_rate": 6.169250645994832e-06, "loss": 1.8971, "step": 2372 }, { "epoch": 0.38323643410852715, "grad_norm": 196.63372802734375, "learning_rate": 6.167635658914729e-06, "loss": 1.9058, "step": 2373 }, { "epoch": 0.38339793281653745, "grad_norm": 9.043557167053223, "learning_rate": 6.166020671834626e-06, "loss": 0.0284, "step": 2374 }, { "epoch": 0.3835594315245478, "grad_norm": 79.17250061035156, "learning_rate": 6.164405684754522e-06, "loss": 1.3535, "step": 2375 }, { "epoch": 0.38372093023255816, "grad_norm": 20.2280216217041, "learning_rate": 6.162790697674419e-06, "loss": 0.1895, "step": 2376 }, { "epoch": 0.38388242894056845, "grad_norm": 63.95960998535156, "learning_rate": 6.161175710594316e-06, "loss": 0.7139, "step": 2377 }, { "epoch": 0.3840439276485788, "grad_norm": 98.18208312988281, "learning_rate": 6.159560723514212e-06, "loss": 1.286, "step": 2378 }, { "epoch": 0.38420542635658916, "grad_norm": 27.242843627929688, "learning_rate": 6.1579457364341085e-06, "loss": 0.2207, "step": 2379 }, { "epoch": 0.38436692506459946, "grad_norm": 28.757747650146484, "learning_rate": 6.156330749354006e-06, "loss": 0.144, "step": 2380 }, { "epoch": 0.3845284237726098, "grad_norm": 45.07590866088867, "learning_rate": 6.1547157622739016e-06, "loss": 0.5253, "step": 2381 }, { "epoch": 0.3846899224806202, "grad_norm": 44.4776496887207, "learning_rate": 6.153100775193799e-06, "loss": 0.5837, "step": 2382 }, { "epoch": 0.38485142118863047, "grad_norm": 148.5565948486328, "learning_rate": 6.1514857881136955e-06, "loss": 4.0137, "step": 2383 }, { "epoch": 0.3850129198966408, "grad_norm": 58.90546417236328, "learning_rate": 6.149870801033592e-06, "loss": 0.4062, "step": 2384 }, { "epoch": 0.3851744186046512, "grad_norm": 92.93370056152344, "learning_rate": 6.148255813953489e-06, "loss": 1.1328, "step": 2385 }, { "epoch": 0.3853359173126615, "grad_norm": 70.53105926513672, "learning_rate": 6.146640826873386e-06, "loss": 0.5996, "step": 2386 }, { "epoch": 0.38549741602067183, "grad_norm": 165.81126403808594, "learning_rate": 6.145025839793283e-06, "loss": 4.1895, "step": 2387 }, { "epoch": 0.3856589147286822, "grad_norm": 23.291622161865234, "learning_rate": 6.143410852713178e-06, "loss": 0.1607, "step": 2388 }, { "epoch": 0.3858204134366925, "grad_norm": 139.04144287109375, "learning_rate": 6.141795865633076e-06, "loss": 1.9297, "step": 2389 }, { "epoch": 0.38598191214470284, "grad_norm": 31.255062103271484, "learning_rate": 6.140180878552972e-06, "loss": 0.1399, "step": 2390 }, { "epoch": 0.3861434108527132, "grad_norm": 42.937557220458984, "learning_rate": 6.138565891472869e-06, "loss": 0.4199, "step": 2391 }, { "epoch": 0.3863049095607235, "grad_norm": 132.39210510253906, "learning_rate": 6.1369509043927654e-06, "loss": 1.4434, "step": 2392 }, { "epoch": 0.38646640826873385, "grad_norm": 25.461706161499023, "learning_rate": 6.135335917312663e-06, "loss": 0.1528, "step": 2393 }, { "epoch": 0.3866279069767442, "grad_norm": 39.53515625, "learning_rate": 6.1337209302325585e-06, "loss": 0.3442, "step": 2394 }, { "epoch": 0.3867894056847545, "grad_norm": 16.839935302734375, "learning_rate": 6.132105943152455e-06, "loss": 0.0779, "step": 2395 }, { "epoch": 0.38695090439276486, "grad_norm": 143.7227325439453, "learning_rate": 6.1304909560723525e-06, "loss": 1.8953, "step": 2396 }, { "epoch": 0.3871124031007752, "grad_norm": 45.27322769165039, "learning_rate": 6.128875968992248e-06, "loss": 0.6323, "step": 2397 }, { "epoch": 0.3872739018087855, "grad_norm": 77.30289459228516, "learning_rate": 6.127260981912146e-06, "loss": 0.8926, "step": 2398 }, { "epoch": 0.38743540051679587, "grad_norm": 43.715576171875, "learning_rate": 6.125645994832042e-06, "loss": 0.4901, "step": 2399 }, { "epoch": 0.3875968992248062, "grad_norm": 132.46498107910156, "learning_rate": 6.124031007751938e-06, "loss": 2.8359, "step": 2400 }, { "epoch": 0.3877583979328165, "grad_norm": 39.33356857299805, "learning_rate": 6.122416020671835e-06, "loss": 0.6113, "step": 2401 }, { "epoch": 0.3879198966408269, "grad_norm": 135.18789672851562, "learning_rate": 6.120801033591732e-06, "loss": 1.4911, "step": 2402 }, { "epoch": 0.38808139534883723, "grad_norm": 140.8626708984375, "learning_rate": 6.1191860465116284e-06, "loss": 1.6638, "step": 2403 }, { "epoch": 0.38824289405684753, "grad_norm": 25.98055648803711, "learning_rate": 6.117571059431525e-06, "loss": 0.1993, "step": 2404 }, { "epoch": 0.3884043927648579, "grad_norm": 50.105159759521484, "learning_rate": 6.115956072351422e-06, "loss": 0.391, "step": 2405 }, { "epoch": 0.38856589147286824, "grad_norm": 102.81242370605469, "learning_rate": 6.114341085271318e-06, "loss": 1.9431, "step": 2406 }, { "epoch": 0.38872739018087854, "grad_norm": 111.9253158569336, "learning_rate": 6.112726098191215e-06, "loss": 1.3418, "step": 2407 }, { "epoch": 0.3888888888888889, "grad_norm": 37.60805892944336, "learning_rate": 6.111111111111112e-06, "loss": 0.3677, "step": 2408 }, { "epoch": 0.38905038759689925, "grad_norm": 61.363868713378906, "learning_rate": 6.109496124031008e-06, "loss": 0.6106, "step": 2409 }, { "epoch": 0.38921188630490955, "grad_norm": 125.94910430908203, "learning_rate": 6.107881136950905e-06, "loss": 1.7832, "step": 2410 }, { "epoch": 0.3893733850129199, "grad_norm": 55.010860443115234, "learning_rate": 6.106266149870802e-06, "loss": 0.4727, "step": 2411 }, { "epoch": 0.38953488372093026, "grad_norm": 172.85438537597656, "learning_rate": 6.104651162790698e-06, "loss": 1.9062, "step": 2412 }, { "epoch": 0.38969638242894056, "grad_norm": 37.93182373046875, "learning_rate": 6.103036175710595e-06, "loss": 0.417, "step": 2413 }, { "epoch": 0.3898578811369509, "grad_norm": 104.81250762939453, "learning_rate": 6.101421188630492e-06, "loss": 0.9902, "step": 2414 }, { "epoch": 0.39001937984496127, "grad_norm": 15.04330825805664, "learning_rate": 6.099806201550388e-06, "loss": 0.1631, "step": 2415 }, { "epoch": 0.39018087855297157, "grad_norm": 81.66114807128906, "learning_rate": 6.0981912144702846e-06, "loss": 0.8196, "step": 2416 }, { "epoch": 0.3903423772609819, "grad_norm": 17.401391983032227, "learning_rate": 6.096576227390182e-06, "loss": 0.093, "step": 2417 }, { "epoch": 0.3905038759689923, "grad_norm": 50.54694366455078, "learning_rate": 6.094961240310078e-06, "loss": 0.7563, "step": 2418 }, { "epoch": 0.3906653746770026, "grad_norm": 167.948486328125, "learning_rate": 6.093346253229975e-06, "loss": 2.625, "step": 2419 }, { "epoch": 0.39082687338501293, "grad_norm": 20.744850158691406, "learning_rate": 6.091731266149872e-06, "loss": 0.1166, "step": 2420 }, { "epoch": 0.39098837209302323, "grad_norm": 56.37360382080078, "learning_rate": 6.090116279069767e-06, "loss": 0.4082, "step": 2421 }, { "epoch": 0.3911498708010336, "grad_norm": 24.644161224365234, "learning_rate": 6.088501291989665e-06, "loss": 0.3037, "step": 2422 }, { "epoch": 0.39131136950904394, "grad_norm": 100.43226623535156, "learning_rate": 6.086886304909561e-06, "loss": 1.3711, "step": 2423 }, { "epoch": 0.39147286821705424, "grad_norm": 37.40611267089844, "learning_rate": 6.085271317829458e-06, "loss": 0.2944, "step": 2424 }, { "epoch": 0.3916343669250646, "grad_norm": 64.39725494384766, "learning_rate": 6.0836563307493544e-06, "loss": 0.5586, "step": 2425 }, { "epoch": 0.39179586563307495, "grad_norm": 59.3597297668457, "learning_rate": 6.082041343669252e-06, "loss": 0.7632, "step": 2426 }, { "epoch": 0.39195736434108525, "grad_norm": 52.55398178100586, "learning_rate": 6.0804263565891476e-06, "loss": 0.3818, "step": 2427 }, { "epoch": 0.3921188630490956, "grad_norm": 69.6267318725586, "learning_rate": 6.078811369509044e-06, "loss": 0.7461, "step": 2428 }, { "epoch": 0.39228036175710596, "grad_norm": 68.41471862792969, "learning_rate": 6.0771963824289415e-06, "loss": 0.6289, "step": 2429 }, { "epoch": 0.39244186046511625, "grad_norm": 12.815414428710938, "learning_rate": 6.075581395348837e-06, "loss": 0.0702, "step": 2430 }, { "epoch": 0.3926033591731266, "grad_norm": 79.00912475585938, "learning_rate": 6.073966408268735e-06, "loss": 0.9219, "step": 2431 }, { "epoch": 0.39276485788113696, "grad_norm": 11.449596405029297, "learning_rate": 6.072351421188631e-06, "loss": 0.0615, "step": 2432 }, { "epoch": 0.39292635658914726, "grad_norm": 91.27648162841797, "learning_rate": 6.070736434108527e-06, "loss": 0.791, "step": 2433 }, { "epoch": 0.3930878552971576, "grad_norm": 20.404582977294922, "learning_rate": 6.069121447028424e-06, "loss": 0.1495, "step": 2434 }, { "epoch": 0.393249354005168, "grad_norm": 18.469287872314453, "learning_rate": 6.067506459948322e-06, "loss": 0.0748, "step": 2435 }, { "epoch": 0.39341085271317827, "grad_norm": 77.5099105834961, "learning_rate": 6.0658914728682175e-06, "loss": 0.4897, "step": 2436 }, { "epoch": 0.3935723514211886, "grad_norm": 148.3916473388672, "learning_rate": 6.064276485788114e-06, "loss": 1.9941, "step": 2437 }, { "epoch": 0.393733850129199, "grad_norm": 180.06056213378906, "learning_rate": 6.062661498708011e-06, "loss": 4.1377, "step": 2438 }, { "epoch": 0.3938953488372093, "grad_norm": 23.554039001464844, "learning_rate": 6.061046511627907e-06, "loss": 0.1455, "step": 2439 }, { "epoch": 0.39405684754521964, "grad_norm": 66.94406127929688, "learning_rate": 6.0594315245478045e-06, "loss": 0.5459, "step": 2440 }, { "epoch": 0.39421834625323, "grad_norm": 77.78428649902344, "learning_rate": 6.057816537467701e-06, "loss": 0.8633, "step": 2441 }, { "epoch": 0.3943798449612403, "grad_norm": 91.0423355102539, "learning_rate": 6.056201550387597e-06, "loss": 0.7012, "step": 2442 }, { "epoch": 0.39454134366925064, "grad_norm": 38.27851104736328, "learning_rate": 6.054586563307494e-06, "loss": 0.3091, "step": 2443 }, { "epoch": 0.394702842377261, "grad_norm": 50.515830993652344, "learning_rate": 6.052971576227391e-06, "loss": 0.3569, "step": 2444 }, { "epoch": 0.3948643410852713, "grad_norm": 119.08203125, "learning_rate": 6.051356589147287e-06, "loss": 0.9753, "step": 2445 }, { "epoch": 0.39502583979328165, "grad_norm": 53.364559173583984, "learning_rate": 6.049741602067184e-06, "loss": 0.9482, "step": 2446 }, { "epoch": 0.395187338501292, "grad_norm": 82.78591918945312, "learning_rate": 6.048126614987081e-06, "loss": 0.627, "step": 2447 }, { "epoch": 0.3953488372093023, "grad_norm": 188.51315307617188, "learning_rate": 6.046511627906977e-06, "loss": 4.5625, "step": 2448 }, { "epoch": 0.39551033591731266, "grad_norm": 185.71043395996094, "learning_rate": 6.044896640826874e-06, "loss": 1.5518, "step": 2449 }, { "epoch": 0.395671834625323, "grad_norm": 110.61734008789062, "learning_rate": 6.043281653746771e-06, "loss": 0.8965, "step": 2450 }, { "epoch": 0.3958333333333333, "grad_norm": 73.70136260986328, "learning_rate": 6.041666666666667e-06, "loss": 0.856, "step": 2451 }, { "epoch": 0.39599483204134367, "grad_norm": 6.217300891876221, "learning_rate": 6.040051679586564e-06, "loss": 0.0319, "step": 2452 }, { "epoch": 0.396156330749354, "grad_norm": 112.72502899169922, "learning_rate": 6.038436692506461e-06, "loss": 1.7534, "step": 2453 }, { "epoch": 0.3963178294573643, "grad_norm": 25.407123565673828, "learning_rate": 6.036821705426356e-06, "loss": 0.2612, "step": 2454 }, { "epoch": 0.3964793281653747, "grad_norm": 4.605186939239502, "learning_rate": 6.035206718346254e-06, "loss": 0.0228, "step": 2455 }, { "epoch": 0.39664082687338503, "grad_norm": 112.5196304321289, "learning_rate": 6.03359173126615e-06, "loss": 0.5496, "step": 2456 }, { "epoch": 0.39680232558139533, "grad_norm": 67.32876586914062, "learning_rate": 6.031976744186047e-06, "loss": 0.6262, "step": 2457 }, { "epoch": 0.3969638242894057, "grad_norm": 61.45144271850586, "learning_rate": 6.0303617571059435e-06, "loss": 0.6782, "step": 2458 }, { "epoch": 0.39712532299741604, "grad_norm": 125.98107147216797, "learning_rate": 6.028746770025841e-06, "loss": 0.8975, "step": 2459 }, { "epoch": 0.39728682170542634, "grad_norm": 153.21017456054688, "learning_rate": 6.027131782945737e-06, "loss": 2.0137, "step": 2460 }, { "epoch": 0.3974483204134367, "grad_norm": 106.51406860351562, "learning_rate": 6.025516795865633e-06, "loss": 2.4023, "step": 2461 }, { "epoch": 0.39760981912144705, "grad_norm": 84.03296661376953, "learning_rate": 6.0239018087855306e-06, "loss": 0.875, "step": 2462 }, { "epoch": 0.39777131782945735, "grad_norm": 31.665346145629883, "learning_rate": 6.022286821705426e-06, "loss": 0.2346, "step": 2463 }, { "epoch": 0.3979328165374677, "grad_norm": 60.950462341308594, "learning_rate": 6.020671834625324e-06, "loss": 0.707, "step": 2464 }, { "epoch": 0.39809431524547806, "grad_norm": 151.6940155029297, "learning_rate": 6.01905684754522e-06, "loss": 2.4688, "step": 2465 }, { "epoch": 0.39825581395348836, "grad_norm": 19.830198287963867, "learning_rate": 6.017441860465116e-06, "loss": 0.0715, "step": 2466 }, { "epoch": 0.3984173126614987, "grad_norm": 68.9081039428711, "learning_rate": 6.015826873385013e-06, "loss": 0.5975, "step": 2467 }, { "epoch": 0.39857881136950907, "grad_norm": 101.43030548095703, "learning_rate": 6.014211886304911e-06, "loss": 1.4058, "step": 2468 }, { "epoch": 0.39874031007751937, "grad_norm": 43.01805114746094, "learning_rate": 6.0125968992248065e-06, "loss": 0.3313, "step": 2469 }, { "epoch": 0.3989018087855297, "grad_norm": 58.436241149902344, "learning_rate": 6.010981912144703e-06, "loss": 0.4407, "step": 2470 }, { "epoch": 0.3990633074935401, "grad_norm": 75.8067855834961, "learning_rate": 6.0093669250646004e-06, "loss": 0.6592, "step": 2471 }, { "epoch": 0.3992248062015504, "grad_norm": 120.4783935546875, "learning_rate": 6.007751937984496e-06, "loss": 0.6393, "step": 2472 }, { "epoch": 0.39938630490956073, "grad_norm": 134.87875366210938, "learning_rate": 6.0061369509043936e-06, "loss": 1.2998, "step": 2473 }, { "epoch": 0.3995478036175711, "grad_norm": 15.112757682800293, "learning_rate": 6.00452196382429e-06, "loss": 0.078, "step": 2474 }, { "epoch": 0.3997093023255814, "grad_norm": 92.59982299804688, "learning_rate": 6.002906976744186e-06, "loss": 1.083, "step": 2475 }, { "epoch": 0.39987080103359174, "grad_norm": 35.18867874145508, "learning_rate": 6.001291989664083e-06, "loss": 0.1738, "step": 2476 }, { "epoch": 0.4000322997416021, "grad_norm": 10.08859634399414, "learning_rate": 5.99967700258398e-06, "loss": 0.0776, "step": 2477 }, { "epoch": 0.4001937984496124, "grad_norm": 91.72712707519531, "learning_rate": 5.998062015503876e-06, "loss": 0.6777, "step": 2478 }, { "epoch": 0.40035529715762275, "grad_norm": 123.24082946777344, "learning_rate": 5.996447028423773e-06, "loss": 1.7812, "step": 2479 }, { "epoch": 0.4005167958656331, "grad_norm": 31.00568199157715, "learning_rate": 5.99483204134367e-06, "loss": 0.3236, "step": 2480 }, { "epoch": 0.4006782945736434, "grad_norm": 35.732269287109375, "learning_rate": 5.993217054263566e-06, "loss": 0.3213, "step": 2481 }, { "epoch": 0.40083979328165376, "grad_norm": 20.442201614379883, "learning_rate": 5.991602067183463e-06, "loss": 0.1646, "step": 2482 }, { "epoch": 0.4010012919896641, "grad_norm": 16.692636489868164, "learning_rate": 5.98998708010336e-06, "loss": 0.09, "step": 2483 }, { "epoch": 0.4011627906976744, "grad_norm": 61.35157012939453, "learning_rate": 5.988372093023256e-06, "loss": 0.7168, "step": 2484 }, { "epoch": 0.40132428940568476, "grad_norm": 171.9945526123047, "learning_rate": 5.986757105943153e-06, "loss": 2.6289, "step": 2485 }, { "epoch": 0.40148578811369506, "grad_norm": 58.18051528930664, "learning_rate": 5.98514211886305e-06, "loss": 0.4922, "step": 2486 }, { "epoch": 0.4016472868217054, "grad_norm": 83.94271087646484, "learning_rate": 5.983527131782945e-06, "loss": 0.5938, "step": 2487 }, { "epoch": 0.4018087855297158, "grad_norm": 107.27928924560547, "learning_rate": 5.981912144702843e-06, "loss": 1.3711, "step": 2488 }, { "epoch": 0.40197028423772607, "grad_norm": 31.867412567138672, "learning_rate": 5.980297157622739e-06, "loss": 0.1812, "step": 2489 }, { "epoch": 0.4021317829457364, "grad_norm": 50.724185943603516, "learning_rate": 5.978682170542636e-06, "loss": 0.3164, "step": 2490 }, { "epoch": 0.4022932816537468, "grad_norm": 60.40257263183594, "learning_rate": 5.9770671834625325e-06, "loss": 0.543, "step": 2491 }, { "epoch": 0.4024547803617571, "grad_norm": 41.541561126708984, "learning_rate": 5.97545219638243e-06, "loss": 0.2207, "step": 2492 }, { "epoch": 0.40261627906976744, "grad_norm": 71.24761962890625, "learning_rate": 5.973837209302326e-06, "loss": 0.4297, "step": 2493 }, { "epoch": 0.4027777777777778, "grad_norm": 38.65147018432617, "learning_rate": 5.972222222222222e-06, "loss": 0.4697, "step": 2494 }, { "epoch": 0.4029392764857881, "grad_norm": 6.964886665344238, "learning_rate": 5.97060723514212e-06, "loss": 0.0164, "step": 2495 }, { "epoch": 0.40310077519379844, "grad_norm": 52.04367446899414, "learning_rate": 5.968992248062015e-06, "loss": 0.625, "step": 2496 }, { "epoch": 0.4032622739018088, "grad_norm": 108.1500244140625, "learning_rate": 5.967377260981913e-06, "loss": 1.5859, "step": 2497 }, { "epoch": 0.4034237726098191, "grad_norm": 54.0458869934082, "learning_rate": 5.965762273901809e-06, "loss": 0.6289, "step": 2498 }, { "epoch": 0.40358527131782945, "grad_norm": 85.59051513671875, "learning_rate": 5.964147286821706e-06, "loss": 1.3242, "step": 2499 }, { "epoch": 0.4037467700258398, "grad_norm": 31.49103355407715, "learning_rate": 5.962532299741602e-06, "loss": 0.1226, "step": 2500 }, { "epoch": 0.4039082687338501, "grad_norm": 85.03257751464844, "learning_rate": 5.9609173126615e-06, "loss": 1.0107, "step": 2501 }, { "epoch": 0.40406976744186046, "grad_norm": 44.62438201904297, "learning_rate": 5.9593023255813955e-06, "loss": 0.3286, "step": 2502 }, { "epoch": 0.4042312661498708, "grad_norm": 53.467952728271484, "learning_rate": 5.957687338501292e-06, "loss": 0.342, "step": 2503 }, { "epoch": 0.4043927648578811, "grad_norm": 109.16583251953125, "learning_rate": 5.9560723514211895e-06, "loss": 1.2266, "step": 2504 }, { "epoch": 0.40455426356589147, "grad_norm": 64.15372467041016, "learning_rate": 5.954457364341085e-06, "loss": 1.2754, "step": 2505 }, { "epoch": 0.4047157622739018, "grad_norm": 5.101788520812988, "learning_rate": 5.952842377260983e-06, "loss": 0.0422, "step": 2506 }, { "epoch": 0.4048772609819121, "grad_norm": 25.10938835144043, "learning_rate": 5.951227390180879e-06, "loss": 0.1211, "step": 2507 }, { "epoch": 0.4050387596899225, "grad_norm": 74.49404907226562, "learning_rate": 5.949612403100775e-06, "loss": 1.125, "step": 2508 }, { "epoch": 0.40520025839793283, "grad_norm": 15.913512229919434, "learning_rate": 5.947997416020672e-06, "loss": 0.0667, "step": 2509 }, { "epoch": 0.40536175710594313, "grad_norm": 96.850341796875, "learning_rate": 5.946382428940569e-06, "loss": 1.957, "step": 2510 }, { "epoch": 0.4055232558139535, "grad_norm": 5.815565586090088, "learning_rate": 5.944767441860465e-06, "loss": 0.0275, "step": 2511 }, { "epoch": 0.40568475452196384, "grad_norm": 98.45035552978516, "learning_rate": 5.943152454780362e-06, "loss": 1.2012, "step": 2512 }, { "epoch": 0.40584625322997414, "grad_norm": 134.37417602539062, "learning_rate": 5.941537467700259e-06, "loss": 1.6211, "step": 2513 }, { "epoch": 0.4060077519379845, "grad_norm": 11.095293045043945, "learning_rate": 5.939922480620155e-06, "loss": 0.0571, "step": 2514 }, { "epoch": 0.40616925064599485, "grad_norm": 5.30712366104126, "learning_rate": 5.938307493540052e-06, "loss": 0.0262, "step": 2515 }, { "epoch": 0.40633074935400515, "grad_norm": 40.64237594604492, "learning_rate": 5.936692506459949e-06, "loss": 0.2495, "step": 2516 }, { "epoch": 0.4064922480620155, "grad_norm": 45.697444915771484, "learning_rate": 5.935077519379846e-06, "loss": 0.2218, "step": 2517 }, { "epoch": 0.40665374677002586, "grad_norm": 142.01806640625, "learning_rate": 5.933462532299742e-06, "loss": 2.2141, "step": 2518 }, { "epoch": 0.40681524547803616, "grad_norm": 119.9623794555664, "learning_rate": 5.931847545219639e-06, "loss": 1.6797, "step": 2519 }, { "epoch": 0.4069767441860465, "grad_norm": 39.813907623291016, "learning_rate": 5.930232558139536e-06, "loss": 0.3467, "step": 2520 }, { "epoch": 0.40713824289405687, "grad_norm": 138.6341552734375, "learning_rate": 5.928617571059432e-06, "loss": 1.1973, "step": 2521 }, { "epoch": 0.40729974160206717, "grad_norm": 22.118175506591797, "learning_rate": 5.927002583979329e-06, "loss": 0.2168, "step": 2522 }, { "epoch": 0.4074612403100775, "grad_norm": 128.98194885253906, "learning_rate": 5.925387596899226e-06, "loss": 1.7812, "step": 2523 }, { "epoch": 0.4076227390180879, "grad_norm": 120.39082336425781, "learning_rate": 5.9237726098191215e-06, "loss": 1.5322, "step": 2524 }, { "epoch": 0.4077842377260982, "grad_norm": 60.167327880859375, "learning_rate": 5.922157622739019e-06, "loss": 1.0, "step": 2525 }, { "epoch": 0.40794573643410853, "grad_norm": 82.4023208618164, "learning_rate": 5.9205426356589155e-06, "loss": 1.0695, "step": 2526 }, { "epoch": 0.4081072351421189, "grad_norm": 30.824411392211914, "learning_rate": 5.918927648578812e-06, "loss": 0.4551, "step": 2527 }, { "epoch": 0.4082687338501292, "grad_norm": 38.478790283203125, "learning_rate": 5.917312661498709e-06, "loss": 0.4419, "step": 2528 }, { "epoch": 0.40843023255813954, "grad_norm": 33.42477798461914, "learning_rate": 5.915697674418606e-06, "loss": 0.3564, "step": 2529 }, { "epoch": 0.4085917312661499, "grad_norm": 116.6594009399414, "learning_rate": 5.914082687338502e-06, "loss": 0.8696, "step": 2530 }, { "epoch": 0.4087532299741602, "grad_norm": 50.38743209838867, "learning_rate": 5.912467700258398e-06, "loss": 0.2944, "step": 2531 }, { "epoch": 0.40891472868217055, "grad_norm": 51.32328796386719, "learning_rate": 5.910852713178296e-06, "loss": 0.5886, "step": 2532 }, { "epoch": 0.4090762273901809, "grad_norm": 169.71034240722656, "learning_rate": 5.909237726098191e-06, "loss": 2.5156, "step": 2533 }, { "epoch": 0.4092377260981912, "grad_norm": 11.17376708984375, "learning_rate": 5.907622739018089e-06, "loss": 0.0355, "step": 2534 }, { "epoch": 0.40939922480620156, "grad_norm": 86.80624389648438, "learning_rate": 5.906007751937985e-06, "loss": 0.6403, "step": 2535 }, { "epoch": 0.4095607235142119, "grad_norm": 99.90137481689453, "learning_rate": 5.904392764857881e-06, "loss": 1.3254, "step": 2536 }, { "epoch": 0.4097222222222222, "grad_norm": 31.1072998046875, "learning_rate": 5.9027777777777785e-06, "loss": 0.458, "step": 2537 }, { "epoch": 0.40988372093023256, "grad_norm": 32.52107238769531, "learning_rate": 5.901162790697675e-06, "loss": 0.1846, "step": 2538 }, { "epoch": 0.4100452196382429, "grad_norm": 120.39974975585938, "learning_rate": 5.899547803617572e-06, "loss": 1.7988, "step": 2539 }, { "epoch": 0.4102067183462532, "grad_norm": 54.42860412597656, "learning_rate": 5.897932816537468e-06, "loss": 0.6367, "step": 2540 }, { "epoch": 0.4103682170542636, "grad_norm": 17.611026763916016, "learning_rate": 5.8963178294573656e-06, "loss": 0.1163, "step": 2541 }, { "epoch": 0.41052971576227393, "grad_norm": 99.0682144165039, "learning_rate": 5.894702842377261e-06, "loss": 1.1185, "step": 2542 }, { "epoch": 0.4106912144702842, "grad_norm": 32.16516876220703, "learning_rate": 5.893087855297158e-06, "loss": 0.2217, "step": 2543 }, { "epoch": 0.4108527131782946, "grad_norm": 107.47667694091797, "learning_rate": 5.891472868217055e-06, "loss": 1.7461, "step": 2544 }, { "epoch": 0.41101421188630494, "grad_norm": 14.51154899597168, "learning_rate": 5.889857881136951e-06, "loss": 0.1433, "step": 2545 }, { "epoch": 0.41117571059431524, "grad_norm": 17.54587173461914, "learning_rate": 5.888242894056848e-06, "loss": 0.1227, "step": 2546 }, { "epoch": 0.4113372093023256, "grad_norm": 198.02099609375, "learning_rate": 5.886627906976745e-06, "loss": 2.0566, "step": 2547 }, { "epoch": 0.4114987080103359, "grad_norm": 19.02878761291504, "learning_rate": 5.885012919896641e-06, "loss": 0.0485, "step": 2548 }, { "epoch": 0.41166020671834624, "grad_norm": 49.69482421875, "learning_rate": 5.883397932816538e-06, "loss": 0.4209, "step": 2549 }, { "epoch": 0.4118217054263566, "grad_norm": 120.85791015625, "learning_rate": 5.8817829457364355e-06, "loss": 2.2188, "step": 2550 }, { "epoch": 0.4119832041343669, "grad_norm": 30.426055908203125, "learning_rate": 5.880167958656331e-06, "loss": 0.2576, "step": 2551 }, { "epoch": 0.41214470284237725, "grad_norm": 34.50377655029297, "learning_rate": 5.878552971576228e-06, "loss": 0.2375, "step": 2552 }, { "epoch": 0.4123062015503876, "grad_norm": 57.65916061401367, "learning_rate": 5.876937984496125e-06, "loss": 0.6191, "step": 2553 }, { "epoch": 0.4124677002583979, "grad_norm": 9.993672370910645, "learning_rate": 5.875322997416021e-06, "loss": 0.0529, "step": 2554 }, { "epoch": 0.41262919896640826, "grad_norm": 36.24388122558594, "learning_rate": 5.873708010335918e-06, "loss": 0.1935, "step": 2555 }, { "epoch": 0.4127906976744186, "grad_norm": 145.62405395507812, "learning_rate": 5.872093023255815e-06, "loss": 1.7969, "step": 2556 }, { "epoch": 0.4129521963824289, "grad_norm": 15.325505256652832, "learning_rate": 5.8704780361757105e-06, "loss": 0.0388, "step": 2557 }, { "epoch": 0.41311369509043927, "grad_norm": 26.97499656677246, "learning_rate": 5.868863049095608e-06, "loss": 0.1318, "step": 2558 }, { "epoch": 0.4132751937984496, "grad_norm": 33.46246337890625, "learning_rate": 5.8672480620155045e-06, "loss": 0.2761, "step": 2559 }, { "epoch": 0.4134366925064599, "grad_norm": 97.66792297363281, "learning_rate": 5.865633074935401e-06, "loss": 1.0215, "step": 2560 }, { "epoch": 0.4135981912144703, "grad_norm": 12.993112564086914, "learning_rate": 5.864018087855298e-06, "loss": 0.0898, "step": 2561 }, { "epoch": 0.41375968992248063, "grad_norm": 41.066104888916016, "learning_rate": 5.862403100775195e-06, "loss": 0.252, "step": 2562 }, { "epoch": 0.41392118863049093, "grad_norm": 59.68684005737305, "learning_rate": 5.860788113695091e-06, "loss": 0.416, "step": 2563 }, { "epoch": 0.4140826873385013, "grad_norm": 5.561979293823242, "learning_rate": 5.859173126614987e-06, "loss": 0.0434, "step": 2564 }, { "epoch": 0.41424418604651164, "grad_norm": 10.528552055358887, "learning_rate": 5.857558139534885e-06, "loss": 0.0544, "step": 2565 }, { "epoch": 0.41440568475452194, "grad_norm": 30.05979347229004, "learning_rate": 5.85594315245478e-06, "loss": 0.2334, "step": 2566 }, { "epoch": 0.4145671834625323, "grad_norm": 82.17698669433594, "learning_rate": 5.854328165374678e-06, "loss": 0.5352, "step": 2567 }, { "epoch": 0.41472868217054265, "grad_norm": 38.33406066894531, "learning_rate": 5.852713178294574e-06, "loss": 0.3611, "step": 2568 }, { "epoch": 0.41489018087855295, "grad_norm": 87.36537170410156, "learning_rate": 5.85109819121447e-06, "loss": 0.8652, "step": 2569 }, { "epoch": 0.4150516795865633, "grad_norm": 56.22511291503906, "learning_rate": 5.8494832041343675e-06, "loss": 0.4209, "step": 2570 }, { "epoch": 0.41521317829457366, "grad_norm": 50.894012451171875, "learning_rate": 5.847868217054264e-06, "loss": 0.2648, "step": 2571 }, { "epoch": 0.41537467700258396, "grad_norm": 286.1044616699219, "learning_rate": 5.846253229974161e-06, "loss": 2.1973, "step": 2572 }, { "epoch": 0.4155361757105943, "grad_norm": 40.90626907348633, "learning_rate": 5.844638242894057e-06, "loss": 0.2826, "step": 2573 }, { "epoch": 0.41569767441860467, "grad_norm": 9.627572059631348, "learning_rate": 5.843023255813955e-06, "loss": 0.0723, "step": 2574 }, { "epoch": 0.41585917312661497, "grad_norm": 149.58551025390625, "learning_rate": 5.84140826873385e-06, "loss": 1.7837, "step": 2575 }, { "epoch": 0.4160206718346253, "grad_norm": 9.64654541015625, "learning_rate": 5.839793281653747e-06, "loss": 0.0125, "step": 2576 }, { "epoch": 0.4161821705426357, "grad_norm": 108.85932922363281, "learning_rate": 5.838178294573644e-06, "loss": 2.3828, "step": 2577 }, { "epoch": 0.416343669250646, "grad_norm": 50.89369583129883, "learning_rate": 5.83656330749354e-06, "loss": 0.416, "step": 2578 }, { "epoch": 0.41650516795865633, "grad_norm": 32.8580322265625, "learning_rate": 5.834948320413437e-06, "loss": 0.1587, "step": 2579 }, { "epoch": 0.4166666666666667, "grad_norm": 151.43331909179688, "learning_rate": 5.833333333333334e-06, "loss": 6.8535, "step": 2580 }, { "epoch": 0.416828165374677, "grad_norm": 64.51286315917969, "learning_rate": 5.83171834625323e-06, "loss": 1.1694, "step": 2581 }, { "epoch": 0.41698966408268734, "grad_norm": 86.70548248291016, "learning_rate": 5.830103359173127e-06, "loss": 1.1094, "step": 2582 }, { "epoch": 0.4171511627906977, "grad_norm": 13.269782066345215, "learning_rate": 5.8284883720930245e-06, "loss": 0.0501, "step": 2583 }, { "epoch": 0.417312661498708, "grad_norm": 65.90884399414062, "learning_rate": 5.82687338501292e-06, "loss": 0.8535, "step": 2584 }, { "epoch": 0.41747416020671835, "grad_norm": 83.37985229492188, "learning_rate": 5.825258397932817e-06, "loss": 1.4639, "step": 2585 }, { "epoch": 0.4176356589147287, "grad_norm": 62.28909683227539, "learning_rate": 5.823643410852714e-06, "loss": 0.4463, "step": 2586 }, { "epoch": 0.417797157622739, "grad_norm": 61.95939254760742, "learning_rate": 5.82202842377261e-06, "loss": 0.4712, "step": 2587 }, { "epoch": 0.41795865633074936, "grad_norm": 9.472983360290527, "learning_rate": 5.820413436692507e-06, "loss": 0.0597, "step": 2588 }, { "epoch": 0.4181201550387597, "grad_norm": 52.92585754394531, "learning_rate": 5.818798449612404e-06, "loss": 0.4692, "step": 2589 }, { "epoch": 0.41828165374677, "grad_norm": 26.8128719329834, "learning_rate": 5.8171834625322996e-06, "loss": 0.1206, "step": 2590 }, { "epoch": 0.41844315245478036, "grad_norm": 18.013254165649414, "learning_rate": 5.815568475452197e-06, "loss": 0.1063, "step": 2591 }, { "epoch": 0.4186046511627907, "grad_norm": 227.40493774414062, "learning_rate": 5.8139534883720935e-06, "loss": 2.3164, "step": 2592 }, { "epoch": 0.418766149870801, "grad_norm": 23.66376495361328, "learning_rate": 5.81233850129199e-06, "loss": 0.2361, "step": 2593 }, { "epoch": 0.4189276485788114, "grad_norm": 28.09423828125, "learning_rate": 5.810723514211887e-06, "loss": 0.2656, "step": 2594 }, { "epoch": 0.41908914728682173, "grad_norm": 38.037376403808594, "learning_rate": 5.809108527131784e-06, "loss": 0.3369, "step": 2595 }, { "epoch": 0.419250645994832, "grad_norm": 65.79281616210938, "learning_rate": 5.80749354005168e-06, "loss": 0.5938, "step": 2596 }, { "epoch": 0.4194121447028424, "grad_norm": 5.39930534362793, "learning_rate": 5.805878552971576e-06, "loss": 0.0213, "step": 2597 }, { "epoch": 0.41957364341085274, "grad_norm": 16.870223999023438, "learning_rate": 5.804263565891474e-06, "loss": 0.0996, "step": 2598 }, { "epoch": 0.41973514211886304, "grad_norm": 60.406402587890625, "learning_rate": 5.8026485788113694e-06, "loss": 0.6699, "step": 2599 }, { "epoch": 0.4198966408268734, "grad_norm": 83.88275146484375, "learning_rate": 5.801033591731267e-06, "loss": 0.5385, "step": 2600 }, { "epoch": 0.42005813953488375, "grad_norm": 70.79623413085938, "learning_rate": 5.799418604651163e-06, "loss": 1.1709, "step": 2601 }, { "epoch": 0.42021963824289404, "grad_norm": 35.1534423828125, "learning_rate": 5.797803617571059e-06, "loss": 0.2521, "step": 2602 }, { "epoch": 0.4203811369509044, "grad_norm": 112.985595703125, "learning_rate": 5.7961886304909565e-06, "loss": 0.6875, "step": 2603 }, { "epoch": 0.42054263565891475, "grad_norm": 60.67313766479492, "learning_rate": 5.794573643410853e-06, "loss": 0.5713, "step": 2604 }, { "epoch": 0.42070413436692505, "grad_norm": 15.556225776672363, "learning_rate": 5.79295865633075e-06, "loss": 0.1191, "step": 2605 }, { "epoch": 0.4208656330749354, "grad_norm": 37.33353805541992, "learning_rate": 5.791343669250646e-06, "loss": 0.2998, "step": 2606 }, { "epoch": 0.42102713178294576, "grad_norm": 22.490921020507812, "learning_rate": 5.789728682170544e-06, "loss": 0.2642, "step": 2607 }, { "epoch": 0.42118863049095606, "grad_norm": 121.75713348388672, "learning_rate": 5.788113695090439e-06, "loss": 1.4155, "step": 2608 }, { "epoch": 0.4213501291989664, "grad_norm": 180.88368225097656, "learning_rate": 5.786498708010336e-06, "loss": 3.8789, "step": 2609 }, { "epoch": 0.42151162790697677, "grad_norm": 26.623754501342773, "learning_rate": 5.784883720930233e-06, "loss": 0.2114, "step": 2610 }, { "epoch": 0.42167312661498707, "grad_norm": 231.95216369628906, "learning_rate": 5.783268733850129e-06, "loss": 2.6145, "step": 2611 }, { "epoch": 0.4218346253229974, "grad_norm": 45.85734558105469, "learning_rate": 5.781653746770026e-06, "loss": 0.498, "step": 2612 }, { "epoch": 0.4219961240310077, "grad_norm": 6.218127250671387, "learning_rate": 5.780038759689923e-06, "loss": 0.0244, "step": 2613 }, { "epoch": 0.4221576227390181, "grad_norm": 76.33898162841797, "learning_rate": 5.7784237726098195e-06, "loss": 0.5156, "step": 2614 }, { "epoch": 0.42231912144702843, "grad_norm": 85.26900482177734, "learning_rate": 5.776808785529716e-06, "loss": 0.967, "step": 2615 }, { "epoch": 0.42248062015503873, "grad_norm": 82.35603332519531, "learning_rate": 5.7751937984496135e-06, "loss": 0.6785, "step": 2616 }, { "epoch": 0.4226421188630491, "grad_norm": 67.21611785888672, "learning_rate": 5.773578811369509e-06, "loss": 0.4326, "step": 2617 }, { "epoch": 0.42280361757105944, "grad_norm": 31.22662925720215, "learning_rate": 5.771963824289406e-06, "loss": 0.2007, "step": 2618 }, { "epoch": 0.42296511627906974, "grad_norm": 264.9603271484375, "learning_rate": 5.770348837209303e-06, "loss": 3.166, "step": 2619 }, { "epoch": 0.4231266149870801, "grad_norm": 29.917146682739258, "learning_rate": 5.768733850129199e-06, "loss": 0.1768, "step": 2620 }, { "epoch": 0.42328811369509045, "grad_norm": 70.43795013427734, "learning_rate": 5.767118863049096e-06, "loss": 0.8379, "step": 2621 }, { "epoch": 0.42344961240310075, "grad_norm": 138.7112274169922, "learning_rate": 5.765503875968993e-06, "loss": 1.5415, "step": 2622 }, { "epoch": 0.4236111111111111, "grad_norm": 11.941529273986816, "learning_rate": 5.7638888888888886e-06, "loss": 0.0629, "step": 2623 }, { "epoch": 0.42377260981912146, "grad_norm": 179.71876525878906, "learning_rate": 5.762273901808786e-06, "loss": 1.5186, "step": 2624 }, { "epoch": 0.42393410852713176, "grad_norm": 41.316001892089844, "learning_rate": 5.7606589147286825e-06, "loss": 0.3047, "step": 2625 }, { "epoch": 0.4240956072351421, "grad_norm": 148.6783905029297, "learning_rate": 5.759043927648579e-06, "loss": 2.2588, "step": 2626 }, { "epoch": 0.42425710594315247, "grad_norm": 31.33850860595703, "learning_rate": 5.757428940568476e-06, "loss": 0.2734, "step": 2627 }, { "epoch": 0.42441860465116277, "grad_norm": 113.26164245605469, "learning_rate": 5.755813953488373e-06, "loss": 1.4043, "step": 2628 }, { "epoch": 0.4245801033591731, "grad_norm": 18.08074951171875, "learning_rate": 5.754198966408269e-06, "loss": 0.13, "step": 2629 }, { "epoch": 0.4247416020671835, "grad_norm": 54.08904266357422, "learning_rate": 5.752583979328165e-06, "loss": 0.5298, "step": 2630 }, { "epoch": 0.4249031007751938, "grad_norm": 151.16880798339844, "learning_rate": 5.750968992248063e-06, "loss": 2.1777, "step": 2631 }, { "epoch": 0.42506459948320413, "grad_norm": 52.14460372924805, "learning_rate": 5.7493540051679585e-06, "loss": 0.3391, "step": 2632 }, { "epoch": 0.4252260981912145, "grad_norm": 119.04143524169922, "learning_rate": 5.747739018087856e-06, "loss": 1.1445, "step": 2633 }, { "epoch": 0.4253875968992248, "grad_norm": 74.93216705322266, "learning_rate": 5.7461240310077524e-06, "loss": 1.1367, "step": 2634 }, { "epoch": 0.42554909560723514, "grad_norm": 86.63346862792969, "learning_rate": 5.744509043927648e-06, "loss": 0.5605, "step": 2635 }, { "epoch": 0.4257105943152455, "grad_norm": 217.16806030273438, "learning_rate": 5.7428940568475455e-06, "loss": 6.125, "step": 2636 }, { "epoch": 0.4258720930232558, "grad_norm": 27.858095169067383, "learning_rate": 5.741279069767443e-06, "loss": 0.2236, "step": 2637 }, { "epoch": 0.42603359173126615, "grad_norm": 101.27275085449219, "learning_rate": 5.739664082687339e-06, "loss": 1.6562, "step": 2638 }, { "epoch": 0.4261950904392765, "grad_norm": 51.279876708984375, "learning_rate": 5.738049095607235e-06, "loss": 0.3999, "step": 2639 }, { "epoch": 0.4263565891472868, "grad_norm": 40.1328010559082, "learning_rate": 5.736434108527133e-06, "loss": 0.2578, "step": 2640 }, { "epoch": 0.42651808785529716, "grad_norm": 15.733963966369629, "learning_rate": 5.734819121447028e-06, "loss": 0.1382, "step": 2641 }, { "epoch": 0.4266795865633075, "grad_norm": 101.49366760253906, "learning_rate": 5.733204134366926e-06, "loss": 0.8274, "step": 2642 }, { "epoch": 0.4268410852713178, "grad_norm": 62.52422332763672, "learning_rate": 5.731589147286822e-06, "loss": 0.5186, "step": 2643 }, { "epoch": 0.42700258397932817, "grad_norm": 65.30613708496094, "learning_rate": 5.729974160206718e-06, "loss": 0.5098, "step": 2644 }, { "epoch": 0.4271640826873385, "grad_norm": 195.80569458007812, "learning_rate": 5.7283591731266154e-06, "loss": 3.7617, "step": 2645 }, { "epoch": 0.4273255813953488, "grad_norm": 60.419837951660156, "learning_rate": 5.726744186046512e-06, "loss": 0.7227, "step": 2646 }, { "epoch": 0.4274870801033592, "grad_norm": 44.91365051269531, "learning_rate": 5.725129198966409e-06, "loss": 0.3574, "step": 2647 }, { "epoch": 0.42764857881136953, "grad_norm": 41.603477478027344, "learning_rate": 5.723514211886305e-06, "loss": 0.4629, "step": 2648 }, { "epoch": 0.4278100775193798, "grad_norm": 127.09058380126953, "learning_rate": 5.7218992248062025e-06, "loss": 2.8345, "step": 2649 }, { "epoch": 0.4279715762273902, "grad_norm": 12.165583610534668, "learning_rate": 5.720284237726099e-06, "loss": 0.0668, "step": 2650 }, { "epoch": 0.42813307493540054, "grad_norm": 179.0957489013672, "learning_rate": 5.718669250645995e-06, "loss": 3.6504, "step": 2651 }, { "epoch": 0.42829457364341084, "grad_norm": 67.56720733642578, "learning_rate": 5.717054263565892e-06, "loss": 1.2529, "step": 2652 }, { "epoch": 0.4284560723514212, "grad_norm": 14.326777458190918, "learning_rate": 5.715439276485789e-06, "loss": 0.0874, "step": 2653 }, { "epoch": 0.42861757105943155, "grad_norm": 104.58984375, "learning_rate": 5.713824289405685e-06, "loss": 0.6191, "step": 2654 }, { "epoch": 0.42877906976744184, "grad_norm": 20.316843032836914, "learning_rate": 5.712209302325582e-06, "loss": 0.2178, "step": 2655 }, { "epoch": 0.4289405684754522, "grad_norm": 56.27986526489258, "learning_rate": 5.710594315245479e-06, "loss": 0.6428, "step": 2656 }, { "epoch": 0.42910206718346255, "grad_norm": 86.3016357421875, "learning_rate": 5.708979328165375e-06, "loss": 1.166, "step": 2657 }, { "epoch": 0.42926356589147285, "grad_norm": 63.34148406982422, "learning_rate": 5.7073643410852716e-06, "loss": 0.7588, "step": 2658 }, { "epoch": 0.4294250645994832, "grad_norm": 99.07940673828125, "learning_rate": 5.705749354005169e-06, "loss": 1.7947, "step": 2659 }, { "epoch": 0.42958656330749356, "grad_norm": 27.056663513183594, "learning_rate": 5.704134366925065e-06, "loss": 0.1497, "step": 2660 }, { "epoch": 0.42974806201550386, "grad_norm": 125.88761138916016, "learning_rate": 5.702519379844962e-06, "loss": 0.7686, "step": 2661 }, { "epoch": 0.4299095607235142, "grad_norm": 36.285606384277344, "learning_rate": 5.700904392764859e-06, "loss": 0.4304, "step": 2662 }, { "epoch": 0.43007105943152457, "grad_norm": 92.62466430664062, "learning_rate": 5.699289405684754e-06, "loss": 0.7969, "step": 2663 }, { "epoch": 0.43023255813953487, "grad_norm": 78.8375015258789, "learning_rate": 5.697674418604652e-06, "loss": 0.4688, "step": 2664 }, { "epoch": 0.4303940568475452, "grad_norm": 48.1276741027832, "learning_rate": 5.696059431524549e-06, "loss": 0.4492, "step": 2665 }, { "epoch": 0.4305555555555556, "grad_norm": 60.68540954589844, "learning_rate": 5.694444444444445e-06, "loss": 0.6729, "step": 2666 }, { "epoch": 0.4307170542635659, "grad_norm": 59.73825454711914, "learning_rate": 5.6928294573643414e-06, "loss": 0.5906, "step": 2667 }, { "epoch": 0.43087855297157623, "grad_norm": 109.76298522949219, "learning_rate": 5.691214470284239e-06, "loss": 1.2896, "step": 2668 }, { "epoch": 0.4310400516795866, "grad_norm": 157.18214416503906, "learning_rate": 5.6895994832041346e-06, "loss": 2.3691, "step": 2669 }, { "epoch": 0.4312015503875969, "grad_norm": 27.551240921020508, "learning_rate": 5.687984496124032e-06, "loss": 0.2617, "step": 2670 }, { "epoch": 0.43136304909560724, "grad_norm": 65.7902603149414, "learning_rate": 5.6863695090439285e-06, "loss": 0.6477, "step": 2671 }, { "epoch": 0.4315245478036176, "grad_norm": 35.27558135986328, "learning_rate": 5.684754521963824e-06, "loss": 0.2438, "step": 2672 }, { "epoch": 0.4316860465116279, "grad_norm": 59.256988525390625, "learning_rate": 5.683139534883722e-06, "loss": 0.5946, "step": 2673 }, { "epoch": 0.43184754521963825, "grad_norm": 58.290462493896484, "learning_rate": 5.681524547803618e-06, "loss": 0.2529, "step": 2674 }, { "epoch": 0.4320090439276486, "grad_norm": 229.3851776123047, "learning_rate": 5.679909560723515e-06, "loss": 5.6641, "step": 2675 }, { "epoch": 0.4321705426356589, "grad_norm": 22.85926628112793, "learning_rate": 5.678294573643411e-06, "loss": 0.1348, "step": 2676 }, { "epoch": 0.43233204134366926, "grad_norm": 164.8119354248047, "learning_rate": 5.676679586563309e-06, "loss": 2.6816, "step": 2677 }, { "epoch": 0.43249354005167956, "grad_norm": 62.62114715576172, "learning_rate": 5.6750645994832045e-06, "loss": 0.6597, "step": 2678 }, { "epoch": 0.4326550387596899, "grad_norm": 98.42538452148438, "learning_rate": 5.673449612403101e-06, "loss": 1.7402, "step": 2679 }, { "epoch": 0.43281653746770027, "grad_norm": 33.97970962524414, "learning_rate": 5.671834625322998e-06, "loss": 0.3198, "step": 2680 }, { "epoch": 0.43297803617571057, "grad_norm": 28.164020538330078, "learning_rate": 5.670219638242894e-06, "loss": 0.1469, "step": 2681 }, { "epoch": 0.4331395348837209, "grad_norm": 17.759729385375977, "learning_rate": 5.6686046511627915e-06, "loss": 0.1509, "step": 2682 }, { "epoch": 0.4333010335917313, "grad_norm": 73.95191192626953, "learning_rate": 5.666989664082688e-06, "loss": 0.8052, "step": 2683 }, { "epoch": 0.4334625322997416, "grad_norm": 59.483802795410156, "learning_rate": 5.665374677002584e-06, "loss": 0.4688, "step": 2684 }, { "epoch": 0.43362403100775193, "grad_norm": 48.07333755493164, "learning_rate": 5.663759689922481e-06, "loss": 0.5522, "step": 2685 }, { "epoch": 0.4337855297157623, "grad_norm": 39.1097526550293, "learning_rate": 5.662144702842378e-06, "loss": 0.2393, "step": 2686 }, { "epoch": 0.4339470284237726, "grad_norm": 110.52162170410156, "learning_rate": 5.660529715762274e-06, "loss": 1.2236, "step": 2687 }, { "epoch": 0.43410852713178294, "grad_norm": 31.377765655517578, "learning_rate": 5.658914728682171e-06, "loss": 0.2939, "step": 2688 }, { "epoch": 0.4342700258397933, "grad_norm": 82.67131805419922, "learning_rate": 5.657299741602068e-06, "loss": 0.8502, "step": 2689 }, { "epoch": 0.4344315245478036, "grad_norm": 43.958290100097656, "learning_rate": 5.655684754521964e-06, "loss": 0.2899, "step": 2690 }, { "epoch": 0.43459302325581395, "grad_norm": 75.38506317138672, "learning_rate": 5.654069767441861e-06, "loss": 1.5977, "step": 2691 }, { "epoch": 0.4347545219638243, "grad_norm": 99.18727111816406, "learning_rate": 5.652454780361758e-06, "loss": 0.5669, "step": 2692 }, { "epoch": 0.4349160206718346, "grad_norm": 82.10910034179688, "learning_rate": 5.650839793281654e-06, "loss": 1.2412, "step": 2693 }, { "epoch": 0.43507751937984496, "grad_norm": 59.13995361328125, "learning_rate": 5.649224806201551e-06, "loss": 0.6826, "step": 2694 }, { "epoch": 0.4352390180878553, "grad_norm": 55.84537124633789, "learning_rate": 5.647609819121448e-06, "loss": 0.3674, "step": 2695 }, { "epoch": 0.4354005167958656, "grad_norm": 33.950687408447266, "learning_rate": 5.645994832041343e-06, "loss": 0.3848, "step": 2696 }, { "epoch": 0.43556201550387597, "grad_norm": 111.0342788696289, "learning_rate": 5.644379844961241e-06, "loss": 1.0918, "step": 2697 }, { "epoch": 0.4357235142118863, "grad_norm": 57.90147018432617, "learning_rate": 5.642764857881138e-06, "loss": 0.541, "step": 2698 }, { "epoch": 0.4358850129198966, "grad_norm": 22.863309860229492, "learning_rate": 5.641149870801034e-06, "loss": 0.3988, "step": 2699 }, { "epoch": 0.436046511627907, "grad_norm": 34.53969192504883, "learning_rate": 5.6395348837209305e-06, "loss": 0.2959, "step": 2700 }, { "epoch": 0.43620801033591733, "grad_norm": 13.236416816711426, "learning_rate": 5.637919896640828e-06, "loss": 0.0698, "step": 2701 }, { "epoch": 0.43636950904392763, "grad_norm": 40.301673889160156, "learning_rate": 5.636304909560724e-06, "loss": 0.458, "step": 2702 }, { "epoch": 0.436531007751938, "grad_norm": 21.007858276367188, "learning_rate": 5.634689922480621e-06, "loss": 0.0767, "step": 2703 }, { "epoch": 0.43669250645994834, "grad_norm": 32.22655487060547, "learning_rate": 5.6330749354005176e-06, "loss": 0.1694, "step": 2704 }, { "epoch": 0.43685400516795864, "grad_norm": 10.99466609954834, "learning_rate": 5.631459948320413e-06, "loss": 0.0447, "step": 2705 }, { "epoch": 0.437015503875969, "grad_norm": 46.96470260620117, "learning_rate": 5.629844961240311e-06, "loss": 0.3828, "step": 2706 }, { "epoch": 0.43717700258397935, "grad_norm": 17.17325210571289, "learning_rate": 5.628229974160207e-06, "loss": 0.177, "step": 2707 }, { "epoch": 0.43733850129198965, "grad_norm": 102.96915435791016, "learning_rate": 5.626614987080104e-06, "loss": 1.3359, "step": 2708 }, { "epoch": 0.4375, "grad_norm": 357.7171325683594, "learning_rate": 5.625e-06, "loss": 4.4766, "step": 2709 }, { "epoch": 0.43766149870801035, "grad_norm": 69.86155700683594, "learning_rate": 5.623385012919898e-06, "loss": 0.8164, "step": 2710 }, { "epoch": 0.43782299741602065, "grad_norm": 199.0330352783203, "learning_rate": 5.6217700258397935e-06, "loss": 2.8906, "step": 2711 }, { "epoch": 0.437984496124031, "grad_norm": 43.50978469848633, "learning_rate": 5.62015503875969e-06, "loss": 0.2578, "step": 2712 }, { "epoch": 0.43814599483204136, "grad_norm": 76.44007110595703, "learning_rate": 5.6185400516795874e-06, "loss": 1.1165, "step": 2713 }, { "epoch": 0.43830749354005166, "grad_norm": 67.00987243652344, "learning_rate": 5.616925064599483e-06, "loss": 0.7661, "step": 2714 }, { "epoch": 0.438468992248062, "grad_norm": 30.534055709838867, "learning_rate": 5.6153100775193806e-06, "loss": 0.1429, "step": 2715 }, { "epoch": 0.43863049095607237, "grad_norm": 47.127384185791016, "learning_rate": 5.613695090439277e-06, "loss": 0.5695, "step": 2716 }, { "epoch": 0.43879198966408267, "grad_norm": 31.076704025268555, "learning_rate": 5.612080103359173e-06, "loss": 0.2539, "step": 2717 }, { "epoch": 0.438953488372093, "grad_norm": 121.65422058105469, "learning_rate": 5.61046511627907e-06, "loss": 1.3242, "step": 2718 }, { "epoch": 0.4391149870801034, "grad_norm": 69.62686920166016, "learning_rate": 5.608850129198967e-06, "loss": 0.9487, "step": 2719 }, { "epoch": 0.4392764857881137, "grad_norm": 41.3511848449707, "learning_rate": 5.607235142118863e-06, "loss": 0.2162, "step": 2720 }, { "epoch": 0.43943798449612403, "grad_norm": 45.18160629272461, "learning_rate": 5.60562015503876e-06, "loss": 1.1874, "step": 2721 }, { "epoch": 0.4395994832041344, "grad_norm": 51.574256896972656, "learning_rate": 5.604005167958657e-06, "loss": 0.3838, "step": 2722 }, { "epoch": 0.4397609819121447, "grad_norm": 40.69700622558594, "learning_rate": 5.602390180878553e-06, "loss": 0.2744, "step": 2723 }, { "epoch": 0.43992248062015504, "grad_norm": 67.82152557373047, "learning_rate": 5.6007751937984504e-06, "loss": 0.499, "step": 2724 }, { "epoch": 0.4400839793281654, "grad_norm": 17.890544891357422, "learning_rate": 5.599160206718347e-06, "loss": 0.0669, "step": 2725 }, { "epoch": 0.4402454780361757, "grad_norm": 117.67720031738281, "learning_rate": 5.597545219638243e-06, "loss": 1.7148, "step": 2726 }, { "epoch": 0.44040697674418605, "grad_norm": 45.67194747924805, "learning_rate": 5.59593023255814e-06, "loss": 0.3452, "step": 2727 }, { "epoch": 0.4405684754521964, "grad_norm": 128.33270263671875, "learning_rate": 5.594315245478037e-06, "loss": 1.8145, "step": 2728 }, { "epoch": 0.4407299741602067, "grad_norm": 123.3648681640625, "learning_rate": 5.592700258397933e-06, "loss": 0.9746, "step": 2729 }, { "epoch": 0.44089147286821706, "grad_norm": 10.30591869354248, "learning_rate": 5.59108527131783e-06, "loss": 0.084, "step": 2730 }, { "epoch": 0.4410529715762274, "grad_norm": 96.37708282470703, "learning_rate": 5.589470284237727e-06, "loss": 1.1084, "step": 2731 }, { "epoch": 0.4412144702842377, "grad_norm": 39.22614288330078, "learning_rate": 5.587855297157623e-06, "loss": 0.3848, "step": 2732 }, { "epoch": 0.44137596899224807, "grad_norm": 80.07492065429688, "learning_rate": 5.5862403100775195e-06, "loss": 1.1221, "step": 2733 }, { "epoch": 0.4415374677002584, "grad_norm": 56.7581672668457, "learning_rate": 5.584625322997417e-06, "loss": 0.4473, "step": 2734 }, { "epoch": 0.4416989664082687, "grad_norm": 154.04908752441406, "learning_rate": 5.583010335917313e-06, "loss": 1.8086, "step": 2735 }, { "epoch": 0.4418604651162791, "grad_norm": 33.54295349121094, "learning_rate": 5.58139534883721e-06, "loss": 0.2925, "step": 2736 }, { "epoch": 0.44202196382428943, "grad_norm": 48.65817642211914, "learning_rate": 5.5797803617571066e-06, "loss": 0.7798, "step": 2737 }, { "epoch": 0.44218346253229973, "grad_norm": 72.09832000732422, "learning_rate": 5.578165374677002e-06, "loss": 0.5811, "step": 2738 }, { "epoch": 0.4423449612403101, "grad_norm": 180.7467803955078, "learning_rate": 5.5765503875969e-06, "loss": 3.0449, "step": 2739 }, { "epoch": 0.44250645994832044, "grad_norm": 59.180137634277344, "learning_rate": 5.574935400516796e-06, "loss": 0.6416, "step": 2740 }, { "epoch": 0.44266795865633074, "grad_norm": 42.476436614990234, "learning_rate": 5.573320413436693e-06, "loss": 0.5447, "step": 2741 }, { "epoch": 0.4428294573643411, "grad_norm": 20.845211029052734, "learning_rate": 5.571705426356589e-06, "loss": 0.1582, "step": 2742 }, { "epoch": 0.4429909560723514, "grad_norm": 64.75459289550781, "learning_rate": 5.570090439276487e-06, "loss": 1.1382, "step": 2743 }, { "epoch": 0.44315245478036175, "grad_norm": 243.51893615722656, "learning_rate": 5.5684754521963825e-06, "loss": 5.4844, "step": 2744 }, { "epoch": 0.4433139534883721, "grad_norm": 121.30289459228516, "learning_rate": 5.566860465116279e-06, "loss": 2.5664, "step": 2745 }, { "epoch": 0.4434754521963824, "grad_norm": 213.9838104248047, "learning_rate": 5.5652454780361765e-06, "loss": 2.7104, "step": 2746 }, { "epoch": 0.44363695090439276, "grad_norm": 133.67288208007812, "learning_rate": 5.563630490956072e-06, "loss": 1.5625, "step": 2747 }, { "epoch": 0.4437984496124031, "grad_norm": 69.64422607421875, "learning_rate": 5.56201550387597e-06, "loss": 0.4277, "step": 2748 }, { "epoch": 0.4439599483204134, "grad_norm": 92.37046813964844, "learning_rate": 5.560400516795866e-06, "loss": 2.0, "step": 2749 }, { "epoch": 0.44412144702842377, "grad_norm": 25.21763038635254, "learning_rate": 5.558785529715762e-06, "loss": 0.2993, "step": 2750 }, { "epoch": 0.4442829457364341, "grad_norm": 57.84855651855469, "learning_rate": 5.557170542635659e-06, "loss": 0.6885, "step": 2751 }, { "epoch": 0.4444444444444444, "grad_norm": 57.46376037597656, "learning_rate": 5.555555555555557e-06, "loss": 1.043, "step": 2752 }, { "epoch": 0.4446059431524548, "grad_norm": 49.34068298339844, "learning_rate": 5.553940568475452e-06, "loss": 0.5059, "step": 2753 }, { "epoch": 0.44476744186046513, "grad_norm": 26.834522247314453, "learning_rate": 5.552325581395349e-06, "loss": 0.2471, "step": 2754 }, { "epoch": 0.44492894056847543, "grad_norm": 143.69992065429688, "learning_rate": 5.550710594315246e-06, "loss": 2.376, "step": 2755 }, { "epoch": 0.4450904392764858, "grad_norm": 120.82210540771484, "learning_rate": 5.549095607235142e-06, "loss": 1.7637, "step": 2756 }, { "epoch": 0.44525193798449614, "grad_norm": 29.48112678527832, "learning_rate": 5.5474806201550395e-06, "loss": 0.1948, "step": 2757 }, { "epoch": 0.44541343669250644, "grad_norm": 136.90948486328125, "learning_rate": 5.545865633074936e-06, "loss": 2.6006, "step": 2758 }, { "epoch": 0.4455749354005168, "grad_norm": 37.59646987915039, "learning_rate": 5.544250645994832e-06, "loss": 0.3589, "step": 2759 }, { "epoch": 0.44573643410852715, "grad_norm": 155.86778259277344, "learning_rate": 5.542635658914729e-06, "loss": 1.3896, "step": 2760 }, { "epoch": 0.44589793281653745, "grad_norm": 26.32891082763672, "learning_rate": 5.541020671834626e-06, "loss": 0.4922, "step": 2761 }, { "epoch": 0.4460594315245478, "grad_norm": 62.23179626464844, "learning_rate": 5.539405684754522e-06, "loss": 0.6162, "step": 2762 }, { "epoch": 0.44622093023255816, "grad_norm": 35.503173828125, "learning_rate": 5.537790697674419e-06, "loss": 0.3936, "step": 2763 }, { "epoch": 0.44638242894056845, "grad_norm": 34.5741081237793, "learning_rate": 5.536175710594316e-06, "loss": 0.166, "step": 2764 }, { "epoch": 0.4465439276485788, "grad_norm": 67.51404571533203, "learning_rate": 5.534560723514212e-06, "loss": 0.6406, "step": 2765 }, { "epoch": 0.44670542635658916, "grad_norm": 94.64474487304688, "learning_rate": 5.5329457364341085e-06, "loss": 1.0039, "step": 2766 }, { "epoch": 0.44686692506459946, "grad_norm": 55.607093811035156, "learning_rate": 5.531330749354006e-06, "loss": 0.4741, "step": 2767 }, { "epoch": 0.4470284237726098, "grad_norm": 12.049602508544922, "learning_rate": 5.529715762273902e-06, "loss": 0.0564, "step": 2768 }, { "epoch": 0.4471899224806202, "grad_norm": 107.52879333496094, "learning_rate": 5.528100775193799e-06, "loss": 1.7769, "step": 2769 }, { "epoch": 0.44735142118863047, "grad_norm": 102.69499969482422, "learning_rate": 5.526485788113696e-06, "loss": 0.7085, "step": 2770 }, { "epoch": 0.4475129198966408, "grad_norm": 67.53001403808594, "learning_rate": 5.524870801033591e-06, "loss": 0.7715, "step": 2771 }, { "epoch": 0.4476744186046512, "grad_norm": 37.20490264892578, "learning_rate": 5.523255813953489e-06, "loss": 0.3494, "step": 2772 }, { "epoch": 0.4478359173126615, "grad_norm": 99.24800872802734, "learning_rate": 5.521640826873385e-06, "loss": 1.002, "step": 2773 }, { "epoch": 0.44799741602067183, "grad_norm": 86.64488983154297, "learning_rate": 5.520025839793283e-06, "loss": 1.4814, "step": 2774 }, { "epoch": 0.4481589147286822, "grad_norm": 85.18621063232422, "learning_rate": 5.518410852713178e-06, "loss": 0.5791, "step": 2775 }, { "epoch": 0.4483204134366925, "grad_norm": 186.7127685546875, "learning_rate": 5.516795865633076e-06, "loss": 5.0703, "step": 2776 }, { "epoch": 0.44848191214470284, "grad_norm": 51.79637908935547, "learning_rate": 5.515180878552972e-06, "loss": 0.3599, "step": 2777 }, { "epoch": 0.4486434108527132, "grad_norm": 49.4703483581543, "learning_rate": 5.513565891472868e-06, "loss": 0.2939, "step": 2778 }, { "epoch": 0.4488049095607235, "grad_norm": 56.29851150512695, "learning_rate": 5.5119509043927655e-06, "loss": 1.0625, "step": 2779 }, { "epoch": 0.44896640826873385, "grad_norm": 60.414100646972656, "learning_rate": 5.510335917312663e-06, "loss": 0.4219, "step": 2780 }, { "epoch": 0.4491279069767442, "grad_norm": 43.13254165649414, "learning_rate": 5.508720930232559e-06, "loss": 0.4531, "step": 2781 }, { "epoch": 0.4492894056847545, "grad_norm": 50.55438995361328, "learning_rate": 5.507105943152455e-06, "loss": 0.6855, "step": 2782 }, { "epoch": 0.44945090439276486, "grad_norm": 170.85757446289062, "learning_rate": 5.5054909560723526e-06, "loss": 1.9336, "step": 2783 }, { "epoch": 0.4496124031007752, "grad_norm": 21.028400421142578, "learning_rate": 5.503875968992248e-06, "loss": 0.0552, "step": 2784 }, { "epoch": 0.4497739018087855, "grad_norm": 60.43437194824219, "learning_rate": 5.502260981912146e-06, "loss": 0.5533, "step": 2785 }, { "epoch": 0.44993540051679587, "grad_norm": 95.90222930908203, "learning_rate": 5.500645994832042e-06, "loss": 0.6171, "step": 2786 }, { "epoch": 0.4500968992248062, "grad_norm": 7.48720645904541, "learning_rate": 5.499031007751938e-06, "loss": 0.0238, "step": 2787 }, { "epoch": 0.4502583979328165, "grad_norm": 142.13632202148438, "learning_rate": 5.497416020671835e-06, "loss": 1.6211, "step": 2788 }, { "epoch": 0.4504198966408269, "grad_norm": 79.17487335205078, "learning_rate": 5.495801033591732e-06, "loss": 0.9531, "step": 2789 }, { "epoch": 0.45058139534883723, "grad_norm": 22.560665130615234, "learning_rate": 5.4941860465116285e-06, "loss": 0.1387, "step": 2790 }, { "epoch": 0.45074289405684753, "grad_norm": 22.392601013183594, "learning_rate": 5.492571059431525e-06, "loss": 0.1628, "step": 2791 }, { "epoch": 0.4509043927648579, "grad_norm": 64.92548370361328, "learning_rate": 5.4909560723514225e-06, "loss": 0.3057, "step": 2792 }, { "epoch": 0.45106589147286824, "grad_norm": 84.05113983154297, "learning_rate": 5.489341085271318e-06, "loss": 0.9001, "step": 2793 }, { "epoch": 0.45122739018087854, "grad_norm": 27.306396484375, "learning_rate": 5.487726098191215e-06, "loss": 0.2153, "step": 2794 }, { "epoch": 0.4513888888888889, "grad_norm": 228.0685272216797, "learning_rate": 5.486111111111112e-06, "loss": 4.3359, "step": 2795 }, { "epoch": 0.45155038759689925, "grad_norm": 30.323780059814453, "learning_rate": 5.484496124031008e-06, "loss": 0.1274, "step": 2796 }, { "epoch": 0.45171188630490955, "grad_norm": 17.911890029907227, "learning_rate": 5.482881136950905e-06, "loss": 0.0811, "step": 2797 }, { "epoch": 0.4518733850129199, "grad_norm": 106.43614196777344, "learning_rate": 5.481266149870802e-06, "loss": 1.6992, "step": 2798 }, { "epoch": 0.45203488372093026, "grad_norm": 42.45252227783203, "learning_rate": 5.4796511627906975e-06, "loss": 0.502, "step": 2799 }, { "epoch": 0.45219638242894056, "grad_norm": 113.4515380859375, "learning_rate": 5.478036175710595e-06, "loss": 0.885, "step": 2800 }, { "epoch": 0.4523578811369509, "grad_norm": 56.1331672668457, "learning_rate": 5.4764211886304915e-06, "loss": 0.9943, "step": 2801 }, { "epoch": 0.45251937984496127, "grad_norm": 126.35726165771484, "learning_rate": 5.474806201550388e-06, "loss": 1.4053, "step": 2802 }, { "epoch": 0.45268087855297157, "grad_norm": 26.85645294189453, "learning_rate": 5.473191214470285e-06, "loss": 0.1587, "step": 2803 }, { "epoch": 0.4528423772609819, "grad_norm": 48.28738784790039, "learning_rate": 5.471576227390182e-06, "loss": 0.3503, "step": 2804 }, { "epoch": 0.4530038759689923, "grad_norm": 81.14542388916016, "learning_rate": 5.469961240310078e-06, "loss": 0.611, "step": 2805 }, { "epoch": 0.4531653746770026, "grad_norm": 110.8840560913086, "learning_rate": 5.468346253229974e-06, "loss": 0.9165, "step": 2806 }, { "epoch": 0.45332687338501293, "grad_norm": 95.56580352783203, "learning_rate": 5.466731266149872e-06, "loss": 0.6318, "step": 2807 }, { "epoch": 0.45348837209302323, "grad_norm": 72.30294799804688, "learning_rate": 5.465116279069767e-06, "loss": 0.6582, "step": 2808 }, { "epoch": 0.4536498708010336, "grad_norm": 65.44771575927734, "learning_rate": 5.463501291989665e-06, "loss": 0.7095, "step": 2809 }, { "epoch": 0.45381136950904394, "grad_norm": 13.16952896118164, "learning_rate": 5.461886304909561e-06, "loss": 0.0565, "step": 2810 }, { "epoch": 0.45397286821705424, "grad_norm": 11.730345726013184, "learning_rate": 5.460271317829457e-06, "loss": 0.0981, "step": 2811 }, { "epoch": 0.4541343669250646, "grad_norm": 77.01202392578125, "learning_rate": 5.4586563307493545e-06, "loss": 1.2207, "step": 2812 }, { "epoch": 0.45429586563307495, "grad_norm": 34.51966094970703, "learning_rate": 5.457041343669252e-06, "loss": 0.3213, "step": 2813 }, { "epoch": 0.45445736434108525, "grad_norm": 16.92664337158203, "learning_rate": 5.455426356589148e-06, "loss": 0.1091, "step": 2814 }, { "epoch": 0.4546188630490956, "grad_norm": 165.73106384277344, "learning_rate": 5.453811369509044e-06, "loss": 1.4683, "step": 2815 }, { "epoch": 0.45478036175710596, "grad_norm": 74.65623474121094, "learning_rate": 5.452196382428942e-06, "loss": 0.7334, "step": 2816 }, { "epoch": 0.45494186046511625, "grad_norm": 122.85989379882812, "learning_rate": 5.450581395348837e-06, "loss": 1.8281, "step": 2817 }, { "epoch": 0.4551033591731266, "grad_norm": 11.609135627746582, "learning_rate": 5.448966408268735e-06, "loss": 0.0579, "step": 2818 }, { "epoch": 0.45526485788113696, "grad_norm": 139.9883575439453, "learning_rate": 5.447351421188631e-06, "loss": 1.7624, "step": 2819 }, { "epoch": 0.45542635658914726, "grad_norm": 89.18506622314453, "learning_rate": 5.445736434108527e-06, "loss": 0.3443, "step": 2820 }, { "epoch": 0.4555878552971576, "grad_norm": 147.088623046875, "learning_rate": 5.444121447028424e-06, "loss": 1.021, "step": 2821 }, { "epoch": 0.455749354005168, "grad_norm": 176.72067260742188, "learning_rate": 5.442506459948321e-06, "loss": 1.5747, "step": 2822 }, { "epoch": 0.45591085271317827, "grad_norm": 3.3258821964263916, "learning_rate": 5.4408914728682175e-06, "loss": 0.0251, "step": 2823 }, { "epoch": 0.4560723514211886, "grad_norm": 60.119564056396484, "learning_rate": 5.439276485788114e-06, "loss": 0.4604, "step": 2824 }, { "epoch": 0.456233850129199, "grad_norm": 27.268556594848633, "learning_rate": 5.4376614987080115e-06, "loss": 0.2847, "step": 2825 }, { "epoch": 0.4563953488372093, "grad_norm": 28.21610450744629, "learning_rate": 5.436046511627907e-06, "loss": 0.3229, "step": 2826 }, { "epoch": 0.45655684754521964, "grad_norm": 61.94053649902344, "learning_rate": 5.434431524547804e-06, "loss": 0.5117, "step": 2827 }, { "epoch": 0.45671834625323, "grad_norm": 14.843710899353027, "learning_rate": 5.432816537467701e-06, "loss": 0.1182, "step": 2828 }, { "epoch": 0.4568798449612403, "grad_norm": 41.8660774230957, "learning_rate": 5.431201550387597e-06, "loss": 0.4438, "step": 2829 }, { "epoch": 0.45704134366925064, "grad_norm": 164.47206115722656, "learning_rate": 5.429586563307494e-06, "loss": 1.4736, "step": 2830 }, { "epoch": 0.457202842377261, "grad_norm": 51.69850540161133, "learning_rate": 5.427971576227391e-06, "loss": 0.5059, "step": 2831 }, { "epoch": 0.4573643410852713, "grad_norm": 17.740493774414062, "learning_rate": 5.4263565891472865e-06, "loss": 0.1353, "step": 2832 }, { "epoch": 0.45752583979328165, "grad_norm": 34.56638717651367, "learning_rate": 5.424741602067184e-06, "loss": 0.5293, "step": 2833 }, { "epoch": 0.457687338501292, "grad_norm": 38.482208251953125, "learning_rate": 5.4231266149870805e-06, "loss": 0.2654, "step": 2834 }, { "epoch": 0.4578488372093023, "grad_norm": 51.07622146606445, "learning_rate": 5.421511627906977e-06, "loss": 0.3291, "step": 2835 }, { "epoch": 0.45801033591731266, "grad_norm": 51.81575012207031, "learning_rate": 5.419896640826874e-06, "loss": 0.2969, "step": 2836 }, { "epoch": 0.458171834625323, "grad_norm": 22.815805435180664, "learning_rate": 5.418281653746771e-06, "loss": 0.1367, "step": 2837 }, { "epoch": 0.4583333333333333, "grad_norm": 56.2890510559082, "learning_rate": 5.416666666666667e-06, "loss": 1.0571, "step": 2838 }, { "epoch": 0.45849483204134367, "grad_norm": 21.303510665893555, "learning_rate": 5.415051679586564e-06, "loss": 0.1836, "step": 2839 }, { "epoch": 0.458656330749354, "grad_norm": 54.82783889770508, "learning_rate": 5.413436692506461e-06, "loss": 0.7979, "step": 2840 }, { "epoch": 0.4588178294573643, "grad_norm": 103.34622192382812, "learning_rate": 5.4118217054263564e-06, "loss": 1.6094, "step": 2841 }, { "epoch": 0.4589793281653747, "grad_norm": 43.473358154296875, "learning_rate": 5.410206718346254e-06, "loss": 0.2339, "step": 2842 }, { "epoch": 0.45914082687338503, "grad_norm": 192.4974822998047, "learning_rate": 5.40859173126615e-06, "loss": 1.5151, "step": 2843 }, { "epoch": 0.45930232558139533, "grad_norm": 22.516748428344727, "learning_rate": 5.406976744186047e-06, "loss": 0.0835, "step": 2844 }, { "epoch": 0.4594638242894057, "grad_norm": 51.865997314453125, "learning_rate": 5.4053617571059435e-06, "loss": 0.2622, "step": 2845 }, { "epoch": 0.45962532299741604, "grad_norm": 78.69941711425781, "learning_rate": 5.403746770025841e-06, "loss": 1.0762, "step": 2846 }, { "epoch": 0.45978682170542634, "grad_norm": 95.21492767333984, "learning_rate": 5.402131782945737e-06, "loss": 1.3984, "step": 2847 }, { "epoch": 0.4599483204134367, "grad_norm": 79.17472839355469, "learning_rate": 5.400516795865633e-06, "loss": 1.7305, "step": 2848 }, { "epoch": 0.46010981912144705, "grad_norm": 134.0528106689453, "learning_rate": 5.398901808785531e-06, "loss": 0.6426, "step": 2849 }, { "epoch": 0.46027131782945735, "grad_norm": 79.64128875732422, "learning_rate": 5.397286821705426e-06, "loss": 1.2388, "step": 2850 }, { "epoch": 0.4604328165374677, "grad_norm": 285.6505126953125, "learning_rate": 5.395671834625324e-06, "loss": 2.043, "step": 2851 }, { "epoch": 0.46059431524547806, "grad_norm": 88.79026794433594, "learning_rate": 5.39405684754522e-06, "loss": 0.7324, "step": 2852 }, { "epoch": 0.46075581395348836, "grad_norm": 86.75531005859375, "learning_rate": 5.392441860465116e-06, "loss": 0.8036, "step": 2853 }, { "epoch": 0.4609173126614987, "grad_norm": 191.40274047851562, "learning_rate": 5.390826873385013e-06, "loss": 2.9922, "step": 2854 }, { "epoch": 0.46107881136950907, "grad_norm": 24.191513061523438, "learning_rate": 5.38921188630491e-06, "loss": 0.1012, "step": 2855 }, { "epoch": 0.46124031007751937, "grad_norm": 17.09734535217285, "learning_rate": 5.3875968992248065e-06, "loss": 0.1276, "step": 2856 }, { "epoch": 0.4614018087855297, "grad_norm": 81.67154693603516, "learning_rate": 5.385981912144703e-06, "loss": 1.9741, "step": 2857 }, { "epoch": 0.4615633074935401, "grad_norm": 134.97576904296875, "learning_rate": 5.3843669250646005e-06, "loss": 0.9551, "step": 2858 }, { "epoch": 0.4617248062015504, "grad_norm": 161.3875732421875, "learning_rate": 5.382751937984496e-06, "loss": 2.1916, "step": 2859 }, { "epoch": 0.46188630490956073, "grad_norm": 65.13124084472656, "learning_rate": 5.381136950904393e-06, "loss": 0.4742, "step": 2860 }, { "epoch": 0.4620478036175711, "grad_norm": 11.442261695861816, "learning_rate": 5.37952196382429e-06, "loss": 0.0906, "step": 2861 }, { "epoch": 0.4622093023255814, "grad_norm": 9.8009672164917, "learning_rate": 5.377906976744186e-06, "loss": 0.0564, "step": 2862 }, { "epoch": 0.46237080103359174, "grad_norm": 24.55092430114746, "learning_rate": 5.376291989664083e-06, "loss": 0.2329, "step": 2863 }, { "epoch": 0.4625322997416021, "grad_norm": 106.36900329589844, "learning_rate": 5.37467700258398e-06, "loss": 0.7676, "step": 2864 }, { "epoch": 0.4626937984496124, "grad_norm": 98.00723266601562, "learning_rate": 5.3730620155038756e-06, "loss": 0.9893, "step": 2865 }, { "epoch": 0.46285529715762275, "grad_norm": 17.59671401977539, "learning_rate": 5.371447028423773e-06, "loss": 0.0514, "step": 2866 }, { "epoch": 0.4630167958656331, "grad_norm": 125.75530242919922, "learning_rate": 5.36983204134367e-06, "loss": 2.2036, "step": 2867 }, { "epoch": 0.4631782945736434, "grad_norm": 48.41181945800781, "learning_rate": 5.368217054263566e-06, "loss": 0.542, "step": 2868 }, { "epoch": 0.46333979328165376, "grad_norm": 47.97621154785156, "learning_rate": 5.366602067183463e-06, "loss": 0.3135, "step": 2869 }, { "epoch": 0.4635012919896641, "grad_norm": 12.66353988647461, "learning_rate": 5.36498708010336e-06, "loss": 0.0677, "step": 2870 }, { "epoch": 0.4636627906976744, "grad_norm": 81.64176177978516, "learning_rate": 5.363372093023256e-06, "loss": 0.4412, "step": 2871 }, { "epoch": 0.46382428940568476, "grad_norm": 294.8971252441406, "learning_rate": 5.361757105943153e-06, "loss": 1.9492, "step": 2872 }, { "epoch": 0.46398578811369506, "grad_norm": 169.3515167236328, "learning_rate": 5.36014211886305e-06, "loss": 0.7944, "step": 2873 }, { "epoch": 0.4641472868217054, "grad_norm": 47.74367141723633, "learning_rate": 5.3585271317829455e-06, "loss": 0.458, "step": 2874 }, { "epoch": 0.4643087855297158, "grad_norm": 26.147918701171875, "learning_rate": 5.356912144702843e-06, "loss": 0.1632, "step": 2875 }, { "epoch": 0.46447028423772607, "grad_norm": 107.15824890136719, "learning_rate": 5.3552971576227394e-06, "loss": 1.0317, "step": 2876 }, { "epoch": 0.4646317829457364, "grad_norm": 19.426713943481445, "learning_rate": 5.353682170542636e-06, "loss": 0.1097, "step": 2877 }, { "epoch": 0.4647932816537468, "grad_norm": 72.13561248779297, "learning_rate": 5.3520671834625325e-06, "loss": 0.5959, "step": 2878 }, { "epoch": 0.4649547803617571, "grad_norm": 132.27130126953125, "learning_rate": 5.35045219638243e-06, "loss": 2.2236, "step": 2879 }, { "epoch": 0.46511627906976744, "grad_norm": 146.1820068359375, "learning_rate": 5.348837209302326e-06, "loss": 1.1431, "step": 2880 }, { "epoch": 0.4652777777777778, "grad_norm": 198.44081115722656, "learning_rate": 5.347222222222222e-06, "loss": 1.3594, "step": 2881 }, { "epoch": 0.4654392764857881, "grad_norm": 112.10215759277344, "learning_rate": 5.34560723514212e-06, "loss": 0.6704, "step": 2882 }, { "epoch": 0.46560077519379844, "grad_norm": 34.764984130859375, "learning_rate": 5.343992248062015e-06, "loss": 0.4312, "step": 2883 }, { "epoch": 0.4657622739018088, "grad_norm": 56.545143127441406, "learning_rate": 5.342377260981913e-06, "loss": 0.5637, "step": 2884 }, { "epoch": 0.4659237726098191, "grad_norm": 134.6724090576172, "learning_rate": 5.340762273901809e-06, "loss": 1.0205, "step": 2885 }, { "epoch": 0.46608527131782945, "grad_norm": 110.9711685180664, "learning_rate": 5.339147286821705e-06, "loss": 2.1523, "step": 2886 }, { "epoch": 0.4662467700258398, "grad_norm": 26.529932022094727, "learning_rate": 5.3375322997416024e-06, "loss": 0.1452, "step": 2887 }, { "epoch": 0.4664082687338501, "grad_norm": 22.408628463745117, "learning_rate": 5.335917312661499e-06, "loss": 0.1467, "step": 2888 }, { "epoch": 0.46656976744186046, "grad_norm": 77.17118072509766, "learning_rate": 5.3343023255813956e-06, "loss": 1.3576, "step": 2889 }, { "epoch": 0.4667312661498708, "grad_norm": 56.90502166748047, "learning_rate": 5.332687338501292e-06, "loss": 0.5537, "step": 2890 }, { "epoch": 0.4668927648578811, "grad_norm": 20.57225227355957, "learning_rate": 5.3310723514211895e-06, "loss": 0.1631, "step": 2891 }, { "epoch": 0.46705426356589147, "grad_norm": 110.2259521484375, "learning_rate": 5.329457364341085e-06, "loss": 1.1703, "step": 2892 }, { "epoch": 0.4672157622739018, "grad_norm": 22.965938568115234, "learning_rate": 5.327842377260982e-06, "loss": 0.1642, "step": 2893 }, { "epoch": 0.4673772609819121, "grad_norm": 31.455350875854492, "learning_rate": 5.326227390180879e-06, "loss": 0.3864, "step": 2894 }, { "epoch": 0.4675387596899225, "grad_norm": 79.8438720703125, "learning_rate": 5.324612403100775e-06, "loss": 1.5469, "step": 2895 }, { "epoch": 0.46770025839793283, "grad_norm": 41.191612243652344, "learning_rate": 5.322997416020672e-06, "loss": 0.2129, "step": 2896 }, { "epoch": 0.46786175710594313, "grad_norm": 71.36924743652344, "learning_rate": 5.321382428940569e-06, "loss": 0.3544, "step": 2897 }, { "epoch": 0.4680232558139535, "grad_norm": 117.86170196533203, "learning_rate": 5.319767441860465e-06, "loss": 1.7202, "step": 2898 }, { "epoch": 0.46818475452196384, "grad_norm": 62.301334381103516, "learning_rate": 5.318152454780362e-06, "loss": 0.5742, "step": 2899 }, { "epoch": 0.46834625322997414, "grad_norm": 23.115556716918945, "learning_rate": 5.316537467700259e-06, "loss": 0.0752, "step": 2900 }, { "epoch": 0.4685077519379845, "grad_norm": 56.57586669921875, "learning_rate": 5.314922480620155e-06, "loss": 0.895, "step": 2901 }, { "epoch": 0.46866925064599485, "grad_norm": 94.52606201171875, "learning_rate": 5.313307493540052e-06, "loss": 1.3848, "step": 2902 }, { "epoch": 0.46883074935400515, "grad_norm": 198.8121795654297, "learning_rate": 5.311692506459949e-06, "loss": 2.1348, "step": 2903 }, { "epoch": 0.4689922480620155, "grad_norm": 124.70975494384766, "learning_rate": 5.310077519379846e-06, "loss": 2.2219, "step": 2904 }, { "epoch": 0.46915374677002586, "grad_norm": 20.461669921875, "learning_rate": 5.308462532299742e-06, "loss": 0.0604, "step": 2905 }, { "epoch": 0.46931524547803616, "grad_norm": 19.123920440673828, "learning_rate": 5.306847545219639e-06, "loss": 0.1259, "step": 2906 }, { "epoch": 0.4694767441860465, "grad_norm": 52.50466537475586, "learning_rate": 5.305232558139536e-06, "loss": 0.4431, "step": 2907 }, { "epoch": 0.46963824289405687, "grad_norm": 5.227933406829834, "learning_rate": 5.303617571059432e-06, "loss": 0.0238, "step": 2908 }, { "epoch": 0.46979974160206717, "grad_norm": 75.44810485839844, "learning_rate": 5.3020025839793284e-06, "loss": 1.2448, "step": 2909 }, { "epoch": 0.4699612403100775, "grad_norm": 19.39260482788086, "learning_rate": 5.300387596899226e-06, "loss": 0.0854, "step": 2910 }, { "epoch": 0.4701227390180879, "grad_norm": 124.99665069580078, "learning_rate": 5.2987726098191216e-06, "loss": 2.2344, "step": 2911 }, { "epoch": 0.4702842377260982, "grad_norm": 11.214142799377441, "learning_rate": 5.297157622739019e-06, "loss": 0.0539, "step": 2912 }, { "epoch": 0.47044573643410853, "grad_norm": 87.58059692382812, "learning_rate": 5.2955426356589155e-06, "loss": 0.7146, "step": 2913 }, { "epoch": 0.4706072351421189, "grad_norm": 60.85387420654297, "learning_rate": 5.293927648578811e-06, "loss": 0.8076, "step": 2914 }, { "epoch": 0.4707687338501292, "grad_norm": 80.63970184326172, "learning_rate": 5.292312661498709e-06, "loss": 0.6101, "step": 2915 }, { "epoch": 0.47093023255813954, "grad_norm": 8.868745803833008, "learning_rate": 5.290697674418605e-06, "loss": 0.056, "step": 2916 }, { "epoch": 0.4710917312661499, "grad_norm": 49.71299362182617, "learning_rate": 5.289082687338502e-06, "loss": 0.2542, "step": 2917 }, { "epoch": 0.4712532299741602, "grad_norm": 30.80664825439453, "learning_rate": 5.287467700258398e-06, "loss": 0.1357, "step": 2918 }, { "epoch": 0.47141472868217055, "grad_norm": 145.52716064453125, "learning_rate": 5.285852713178296e-06, "loss": 4.2933, "step": 2919 }, { "epoch": 0.4715762273901809, "grad_norm": 51.50445556640625, "learning_rate": 5.2842377260981915e-06, "loss": 0.6992, "step": 2920 }, { "epoch": 0.4717377260981912, "grad_norm": 59.603370666503906, "learning_rate": 5.282622739018088e-06, "loss": 0.4479, "step": 2921 }, { "epoch": 0.47189922480620156, "grad_norm": 157.6669464111328, "learning_rate": 5.281007751937985e-06, "loss": 2.8184, "step": 2922 }, { "epoch": 0.4720607235142119, "grad_norm": 19.67925262451172, "learning_rate": 5.279392764857881e-06, "loss": 0.0948, "step": 2923 }, { "epoch": 0.4722222222222222, "grad_norm": 75.21385955810547, "learning_rate": 5.2777777777777785e-06, "loss": 0.9766, "step": 2924 }, { "epoch": 0.47238372093023256, "grad_norm": 9.762616157531738, "learning_rate": 5.276162790697675e-06, "loss": 0.074, "step": 2925 }, { "epoch": 0.4725452196382429, "grad_norm": 88.53866577148438, "learning_rate": 5.274547803617572e-06, "loss": 0.8125, "step": 2926 }, { "epoch": 0.4727067183462532, "grad_norm": 19.382190704345703, "learning_rate": 5.272932816537468e-06, "loss": 0.116, "step": 2927 }, { "epoch": 0.4728682170542636, "grad_norm": 66.05052947998047, "learning_rate": 5.271317829457366e-06, "loss": 0.7422, "step": 2928 }, { "epoch": 0.47302971576227393, "grad_norm": 31.336549758911133, "learning_rate": 5.269702842377261e-06, "loss": 0.209, "step": 2929 }, { "epoch": 0.4731912144702842, "grad_norm": 13.37606143951416, "learning_rate": 5.268087855297158e-06, "loss": 0.0473, "step": 2930 }, { "epoch": 0.4733527131782946, "grad_norm": 52.45757293701172, "learning_rate": 5.266472868217055e-06, "loss": 0.5342, "step": 2931 }, { "epoch": 0.47351421188630494, "grad_norm": 114.01138305664062, "learning_rate": 5.264857881136951e-06, "loss": 1.2578, "step": 2932 }, { "epoch": 0.47367571059431524, "grad_norm": 24.851755142211914, "learning_rate": 5.2632428940568484e-06, "loss": 0.1396, "step": 2933 }, { "epoch": 0.4738372093023256, "grad_norm": 95.54053497314453, "learning_rate": 5.261627906976745e-06, "loss": 1.0732, "step": 2934 }, { "epoch": 0.4739987080103359, "grad_norm": 30.95374298095703, "learning_rate": 5.260012919896641e-06, "loss": 0.1388, "step": 2935 }, { "epoch": 0.47416020671834624, "grad_norm": 13.248644828796387, "learning_rate": 5.258397932816538e-06, "loss": 0.0811, "step": 2936 }, { "epoch": 0.4743217054263566, "grad_norm": 72.90837097167969, "learning_rate": 5.256782945736435e-06, "loss": 0.8687, "step": 2937 }, { "epoch": 0.4744832041343669, "grad_norm": 14.175899505615234, "learning_rate": 5.255167958656331e-06, "loss": 0.0677, "step": 2938 }, { "epoch": 0.47464470284237725, "grad_norm": 69.50324249267578, "learning_rate": 5.253552971576228e-06, "loss": 0.464, "step": 2939 }, { "epoch": 0.4748062015503876, "grad_norm": 6.158231258392334, "learning_rate": 5.251937984496125e-06, "loss": 0.0302, "step": 2940 }, { "epoch": 0.4749677002583979, "grad_norm": 31.34061050415039, "learning_rate": 5.250322997416021e-06, "loss": 0.377, "step": 2941 }, { "epoch": 0.47512919896640826, "grad_norm": 39.205223083496094, "learning_rate": 5.2487080103359175e-06, "loss": 0.2764, "step": 2942 }, { "epoch": 0.4752906976744186, "grad_norm": 114.41816711425781, "learning_rate": 5.247093023255815e-06, "loss": 1.9492, "step": 2943 }, { "epoch": 0.4754521963824289, "grad_norm": 296.87908935546875, "learning_rate": 5.245478036175711e-06, "loss": 1.7036, "step": 2944 }, { "epoch": 0.47561369509043927, "grad_norm": 42.44519805908203, "learning_rate": 5.243863049095608e-06, "loss": 0.4376, "step": 2945 }, { "epoch": 0.4757751937984496, "grad_norm": 30.087003707885742, "learning_rate": 5.2422480620155046e-06, "loss": 0.1475, "step": 2946 }, { "epoch": 0.4759366925064599, "grad_norm": 22.827754974365234, "learning_rate": 5.2406330749354e-06, "loss": 0.2056, "step": 2947 }, { "epoch": 0.4760981912144703, "grad_norm": 85.5960693359375, "learning_rate": 5.239018087855298e-06, "loss": 1.3818, "step": 2948 }, { "epoch": 0.47625968992248063, "grad_norm": 13.64590072631836, "learning_rate": 5.237403100775194e-06, "loss": 0.0439, "step": 2949 }, { "epoch": 0.47642118863049093, "grad_norm": 103.01564025878906, "learning_rate": 5.235788113695091e-06, "loss": 1.4316, "step": 2950 }, { "epoch": 0.4765826873385013, "grad_norm": 93.493408203125, "learning_rate": 5.234173126614987e-06, "loss": 0.5605, "step": 2951 }, { "epoch": 0.47674418604651164, "grad_norm": 55.22138595581055, "learning_rate": 5.232558139534885e-06, "loss": 0.4712, "step": 2952 }, { "epoch": 0.47690568475452194, "grad_norm": 15.505147933959961, "learning_rate": 5.2309431524547805e-06, "loss": 0.0305, "step": 2953 }, { "epoch": 0.4770671834625323, "grad_norm": 214.59359741210938, "learning_rate": 5.229328165374678e-06, "loss": 4.4883, "step": 2954 }, { "epoch": 0.47722868217054265, "grad_norm": 53.40773391723633, "learning_rate": 5.2277131782945744e-06, "loss": 0.3181, "step": 2955 }, { "epoch": 0.47739018087855295, "grad_norm": 67.78229522705078, "learning_rate": 5.22609819121447e-06, "loss": 0.5469, "step": 2956 }, { "epoch": 0.4775516795865633, "grad_norm": 82.67569732666016, "learning_rate": 5.2244832041343676e-06, "loss": 1.2363, "step": 2957 }, { "epoch": 0.47771317829457366, "grad_norm": 139.4439697265625, "learning_rate": 5.222868217054264e-06, "loss": 1.123, "step": 2958 }, { "epoch": 0.47787467700258396, "grad_norm": 6.885078430175781, "learning_rate": 5.221253229974161e-06, "loss": 0.0322, "step": 2959 }, { "epoch": 0.4780361757105943, "grad_norm": 31.31545639038086, "learning_rate": 5.219638242894057e-06, "loss": 0.2598, "step": 2960 }, { "epoch": 0.47819767441860467, "grad_norm": 161.013671875, "learning_rate": 5.218023255813955e-06, "loss": 2.5508, "step": 2961 }, { "epoch": 0.47835917312661497, "grad_norm": 57.826393127441406, "learning_rate": 5.21640826873385e-06, "loss": 0.7793, "step": 2962 }, { "epoch": 0.4785206718346253, "grad_norm": 54.27024459838867, "learning_rate": 5.214793281653747e-06, "loss": 0.412, "step": 2963 }, { "epoch": 0.4786821705426357, "grad_norm": 30.085718154907227, "learning_rate": 5.213178294573644e-06, "loss": 0.1216, "step": 2964 }, { "epoch": 0.478843669250646, "grad_norm": 67.959228515625, "learning_rate": 5.21156330749354e-06, "loss": 0.8862, "step": 2965 }, { "epoch": 0.47900516795865633, "grad_norm": 34.68132019042969, "learning_rate": 5.2099483204134374e-06, "loss": 0.231, "step": 2966 }, { "epoch": 0.4791666666666667, "grad_norm": 22.1279239654541, "learning_rate": 5.208333333333334e-06, "loss": 0.2671, "step": 2967 }, { "epoch": 0.479328165374677, "grad_norm": 76.8531723022461, "learning_rate": 5.20671834625323e-06, "loss": 1.373, "step": 2968 }, { "epoch": 0.47948966408268734, "grad_norm": 84.25572967529297, "learning_rate": 5.205103359173127e-06, "loss": 0.7114, "step": 2969 }, { "epoch": 0.4796511627906977, "grad_norm": 78.49954223632812, "learning_rate": 5.203488372093024e-06, "loss": 1.1787, "step": 2970 }, { "epoch": 0.479812661498708, "grad_norm": 86.37480163574219, "learning_rate": 5.20187338501292e-06, "loss": 0.7729, "step": 2971 }, { "epoch": 0.47997416020671835, "grad_norm": 45.890384674072266, "learning_rate": 5.200258397932817e-06, "loss": 0.4014, "step": 2972 }, { "epoch": 0.4801356589147287, "grad_norm": 142.0817108154297, "learning_rate": 5.198643410852714e-06, "loss": 3.1582, "step": 2973 }, { "epoch": 0.480297157622739, "grad_norm": 151.17677307128906, "learning_rate": 5.19702842377261e-06, "loss": 5.4834, "step": 2974 }, { "epoch": 0.48045865633074936, "grad_norm": 94.518798828125, "learning_rate": 5.1954134366925065e-06, "loss": 1.127, "step": 2975 }, { "epoch": 0.4806201550387597, "grad_norm": 57.8267822265625, "learning_rate": 5.193798449612404e-06, "loss": 0.4033, "step": 2976 }, { "epoch": 0.48078165374677, "grad_norm": 263.2840576171875, "learning_rate": 5.1921834625323e-06, "loss": 1.0098, "step": 2977 }, { "epoch": 0.48094315245478036, "grad_norm": 190.06964111328125, "learning_rate": 5.190568475452197e-06, "loss": 2.4141, "step": 2978 }, { "epoch": 0.4811046511627907, "grad_norm": 47.769622802734375, "learning_rate": 5.1889534883720936e-06, "loss": 0.2666, "step": 2979 }, { "epoch": 0.481266149870801, "grad_norm": 36.1462516784668, "learning_rate": 5.187338501291989e-06, "loss": 0.2368, "step": 2980 }, { "epoch": 0.4814276485788114, "grad_norm": 137.0074920654297, "learning_rate": 5.185723514211887e-06, "loss": 1.084, "step": 2981 }, { "epoch": 0.48158914728682173, "grad_norm": 89.51133728027344, "learning_rate": 5.184108527131784e-06, "loss": 1.2227, "step": 2982 }, { "epoch": 0.481750645994832, "grad_norm": 60.28077697753906, "learning_rate": 5.18249354005168e-06, "loss": 0.4381, "step": 2983 }, { "epoch": 0.4819121447028424, "grad_norm": 137.47775268554688, "learning_rate": 5.180878552971576e-06, "loss": 2.1328, "step": 2984 }, { "epoch": 0.48207364341085274, "grad_norm": 51.151466369628906, "learning_rate": 5.179263565891474e-06, "loss": 0.418, "step": 2985 }, { "epoch": 0.48223514211886304, "grad_norm": 69.11027526855469, "learning_rate": 5.1776485788113695e-06, "loss": 0.835, "step": 2986 }, { "epoch": 0.4823966408268734, "grad_norm": 8.555811882019043, "learning_rate": 5.176033591731267e-06, "loss": 0.0662, "step": 2987 }, { "epoch": 0.48255813953488375, "grad_norm": 29.879030227661133, "learning_rate": 5.1744186046511635e-06, "loss": 0.1451, "step": 2988 }, { "epoch": 0.48271963824289404, "grad_norm": 57.314720153808594, "learning_rate": 5.172803617571059e-06, "loss": 0.5098, "step": 2989 }, { "epoch": 0.4828811369509044, "grad_norm": 61.742191314697266, "learning_rate": 5.171188630490957e-06, "loss": 0.8774, "step": 2990 }, { "epoch": 0.48304263565891475, "grad_norm": 32.07488250732422, "learning_rate": 5.169573643410853e-06, "loss": 0.1797, "step": 2991 }, { "epoch": 0.48320413436692505, "grad_norm": 109.33070373535156, "learning_rate": 5.16795865633075e-06, "loss": 2.6826, "step": 2992 }, { "epoch": 0.4833656330749354, "grad_norm": 46.198204040527344, "learning_rate": 5.166343669250646e-06, "loss": 0.3594, "step": 2993 }, { "epoch": 0.48352713178294576, "grad_norm": 23.26021957397461, "learning_rate": 5.164728682170544e-06, "loss": 0.1858, "step": 2994 }, { "epoch": 0.48368863049095606, "grad_norm": 102.1323013305664, "learning_rate": 5.163113695090439e-06, "loss": 0.8516, "step": 2995 }, { "epoch": 0.4838501291989664, "grad_norm": 13.366844177246094, "learning_rate": 5.161498708010336e-06, "loss": 0.0708, "step": 2996 }, { "epoch": 0.48401162790697677, "grad_norm": 26.863061904907227, "learning_rate": 5.159883720930233e-06, "loss": 0.2827, "step": 2997 }, { "epoch": 0.48417312661498707, "grad_norm": 32.8065299987793, "learning_rate": 5.158268733850129e-06, "loss": 0.2744, "step": 2998 }, { "epoch": 0.4843346253229974, "grad_norm": 32.428810119628906, "learning_rate": 5.1566537467700265e-06, "loss": 0.3031, "step": 2999 }, { "epoch": 0.4844961240310077, "grad_norm": 131.9278564453125, "learning_rate": 5.155038759689923e-06, "loss": 2.8135, "step": 3000 }, { "epoch": 0.4846576227390181, "grad_norm": 56.98029708862305, "learning_rate": 5.153423772609819e-06, "loss": 0.7646, "step": 3001 }, { "epoch": 0.48481912144702843, "grad_norm": 8.85546875, "learning_rate": 5.151808785529716e-06, "loss": 0.0204, "step": 3002 }, { "epoch": 0.48498062015503873, "grad_norm": 19.46196746826172, "learning_rate": 5.150193798449613e-06, "loss": 0.1692, "step": 3003 }, { "epoch": 0.4851421188630491, "grad_norm": 55.93352508544922, "learning_rate": 5.148578811369509e-06, "loss": 0.3204, "step": 3004 }, { "epoch": 0.48530361757105944, "grad_norm": 173.95220947265625, "learning_rate": 5.146963824289406e-06, "loss": 1.6709, "step": 3005 }, { "epoch": 0.48546511627906974, "grad_norm": 45.91129684448242, "learning_rate": 5.145348837209303e-06, "loss": 0.2559, "step": 3006 }, { "epoch": 0.4856266149870801, "grad_norm": 90.38096618652344, "learning_rate": 5.143733850129199e-06, "loss": 0.3594, "step": 3007 }, { "epoch": 0.48578811369509045, "grad_norm": 48.200294494628906, "learning_rate": 5.1421188630490955e-06, "loss": 0.7986, "step": 3008 }, { "epoch": 0.48594961240310075, "grad_norm": 61.09331130981445, "learning_rate": 5.140503875968993e-06, "loss": 0.8237, "step": 3009 }, { "epoch": 0.4861111111111111, "grad_norm": 9.297157287597656, "learning_rate": 5.138888888888889e-06, "loss": 0.0782, "step": 3010 }, { "epoch": 0.48627260981912146, "grad_norm": 16.610342025756836, "learning_rate": 5.137273901808786e-06, "loss": 0.0835, "step": 3011 }, { "epoch": 0.48643410852713176, "grad_norm": 29.04210090637207, "learning_rate": 5.135658914728683e-06, "loss": 0.439, "step": 3012 }, { "epoch": 0.4865956072351421, "grad_norm": 31.1036376953125, "learning_rate": 5.134043927648578e-06, "loss": 0.1692, "step": 3013 }, { "epoch": 0.48675710594315247, "grad_norm": 108.5325698852539, "learning_rate": 5.132428940568476e-06, "loss": 1.0801, "step": 3014 }, { "epoch": 0.48691860465116277, "grad_norm": 8.241636276245117, "learning_rate": 5.130813953488373e-06, "loss": 0.0693, "step": 3015 }, { "epoch": 0.4870801033591731, "grad_norm": 20.977956771850586, "learning_rate": 5.129198966408269e-06, "loss": 0.0677, "step": 3016 }, { "epoch": 0.4872416020671835, "grad_norm": 61.31210708618164, "learning_rate": 5.127583979328165e-06, "loss": 0.7871, "step": 3017 }, { "epoch": 0.4874031007751938, "grad_norm": 134.6963348388672, "learning_rate": 5.125968992248063e-06, "loss": 1.8359, "step": 3018 }, { "epoch": 0.48756459948320413, "grad_norm": 82.44365692138672, "learning_rate": 5.1243540051679585e-06, "loss": 0.9902, "step": 3019 }, { "epoch": 0.4877260981912145, "grad_norm": 84.40813446044922, "learning_rate": 5.122739018087856e-06, "loss": 0.8613, "step": 3020 }, { "epoch": 0.4878875968992248, "grad_norm": 46.1293830871582, "learning_rate": 5.1211240310077525e-06, "loss": 0.2712, "step": 3021 }, { "epoch": 0.48804909560723514, "grad_norm": 56.168853759765625, "learning_rate": 5.119509043927648e-06, "loss": 0.3857, "step": 3022 }, { "epoch": 0.4882105943152455, "grad_norm": 73.0849609375, "learning_rate": 5.117894056847546e-06, "loss": 0.4471, "step": 3023 }, { "epoch": 0.4883720930232558, "grad_norm": 177.76132202148438, "learning_rate": 5.116279069767442e-06, "loss": 0.9367, "step": 3024 }, { "epoch": 0.48853359173126615, "grad_norm": 3.325263261795044, "learning_rate": 5.114664082687339e-06, "loss": 0.0151, "step": 3025 }, { "epoch": 0.4886950904392765, "grad_norm": 106.35871887207031, "learning_rate": 5.113049095607235e-06, "loss": 1.3184, "step": 3026 }, { "epoch": 0.4888565891472868, "grad_norm": 118.22447967529297, "learning_rate": 5.111434108527133e-06, "loss": 0.5691, "step": 3027 }, { "epoch": 0.48901808785529716, "grad_norm": 77.03599548339844, "learning_rate": 5.109819121447028e-06, "loss": 0.8896, "step": 3028 }, { "epoch": 0.4891795865633075, "grad_norm": 39.549293518066406, "learning_rate": 5.108204134366925e-06, "loss": 0.3617, "step": 3029 }, { "epoch": 0.4893410852713178, "grad_norm": 88.43897247314453, "learning_rate": 5.106589147286822e-06, "loss": 1.875, "step": 3030 }, { "epoch": 0.48950258397932817, "grad_norm": 147.4815673828125, "learning_rate": 5.104974160206718e-06, "loss": 1.4604, "step": 3031 }, { "epoch": 0.4896640826873385, "grad_norm": 114.95828247070312, "learning_rate": 5.1033591731266155e-06, "loss": 1.375, "step": 3032 }, { "epoch": 0.4898255813953488, "grad_norm": 42.89358901977539, "learning_rate": 5.101744186046512e-06, "loss": 0.1992, "step": 3033 }, { "epoch": 0.4899870801033592, "grad_norm": 38.17774963378906, "learning_rate": 5.1001291989664095e-06, "loss": 0.5137, "step": 3034 }, { "epoch": 0.49014857881136953, "grad_norm": 21.88650894165039, "learning_rate": 5.098514211886305e-06, "loss": 0.1498, "step": 3035 }, { "epoch": 0.4903100775193798, "grad_norm": 58.414913177490234, "learning_rate": 5.096899224806202e-06, "loss": 0.3672, "step": 3036 }, { "epoch": 0.4904715762273902, "grad_norm": 76.1167984008789, "learning_rate": 5.095284237726099e-06, "loss": 0.8843, "step": 3037 }, { "epoch": 0.49063307493540054, "grad_norm": 12.489971160888672, "learning_rate": 5.093669250645995e-06, "loss": 0.0396, "step": 3038 }, { "epoch": 0.49079457364341084, "grad_norm": 48.283172607421875, "learning_rate": 5.092054263565892e-06, "loss": 0.3711, "step": 3039 }, { "epoch": 0.4909560723514212, "grad_norm": 26.216875076293945, "learning_rate": 5.090439276485789e-06, "loss": 0.1301, "step": 3040 }, { "epoch": 0.49111757105943155, "grad_norm": 209.70872497558594, "learning_rate": 5.088824289405685e-06, "loss": 0.375, "step": 3041 }, { "epoch": 0.49127906976744184, "grad_norm": 47.38846206665039, "learning_rate": 5.087209302325582e-06, "loss": 0.4018, "step": 3042 }, { "epoch": 0.4914405684754522, "grad_norm": 48.05839920043945, "learning_rate": 5.085594315245479e-06, "loss": 0.7686, "step": 3043 }, { "epoch": 0.49160206718346255, "grad_norm": 43.24413299560547, "learning_rate": 5.083979328165375e-06, "loss": 0.4541, "step": 3044 }, { "epoch": 0.49176356589147285, "grad_norm": 149.9031219482422, "learning_rate": 5.082364341085272e-06, "loss": 1.1367, "step": 3045 }, { "epoch": 0.4919250645994832, "grad_norm": 30.622507095336914, "learning_rate": 5.080749354005169e-06, "loss": 0.1577, "step": 3046 }, { "epoch": 0.49208656330749356, "grad_norm": 55.63866424560547, "learning_rate": 5.079134366925065e-06, "loss": 0.4668, "step": 3047 }, { "epoch": 0.49224806201550386, "grad_norm": 48.56483459472656, "learning_rate": 5.077519379844962e-06, "loss": 0.408, "step": 3048 }, { "epoch": 0.4924095607235142, "grad_norm": 708.3572998046875, "learning_rate": 5.075904392764859e-06, "loss": 1.3399, "step": 3049 }, { "epoch": 0.49257105943152457, "grad_norm": 154.25558471679688, "learning_rate": 5.074289405684754e-06, "loss": 2.2489, "step": 3050 }, { "epoch": 0.49273255813953487, "grad_norm": 56.549896240234375, "learning_rate": 5.072674418604652e-06, "loss": 0.4734, "step": 3051 }, { "epoch": 0.4928940568475452, "grad_norm": 70.61193084716797, "learning_rate": 5.071059431524548e-06, "loss": 0.418, "step": 3052 }, { "epoch": 0.4930555555555556, "grad_norm": 123.42379760742188, "learning_rate": 5.069444444444445e-06, "loss": 1.9492, "step": 3053 }, { "epoch": 0.4932170542635659, "grad_norm": 100.7396011352539, "learning_rate": 5.0678294573643415e-06, "loss": 1.418, "step": 3054 }, { "epoch": 0.49337855297157623, "grad_norm": 7.131004810333252, "learning_rate": 5.066214470284239e-06, "loss": 0.0312, "step": 3055 }, { "epoch": 0.4935400516795866, "grad_norm": 103.04524993896484, "learning_rate": 5.064599483204135e-06, "loss": 0.8022, "step": 3056 }, { "epoch": 0.4937015503875969, "grad_norm": 47.73760986328125, "learning_rate": 5.062984496124031e-06, "loss": 0.2891, "step": 3057 }, { "epoch": 0.49386304909560724, "grad_norm": 40.34502410888672, "learning_rate": 5.061369509043929e-06, "loss": 0.4199, "step": 3058 }, { "epoch": 0.4940245478036176, "grad_norm": 144.0771484375, "learning_rate": 5.059754521963824e-06, "loss": 4.083, "step": 3059 }, { "epoch": 0.4941860465116279, "grad_norm": 82.38703155517578, "learning_rate": 5.058139534883722e-06, "loss": 0.7908, "step": 3060 }, { "epoch": 0.49434754521963825, "grad_norm": 11.57528305053711, "learning_rate": 5.056524547803618e-06, "loss": 0.0649, "step": 3061 }, { "epoch": 0.4945090439276486, "grad_norm": 17.676294326782227, "learning_rate": 5.054909560723514e-06, "loss": 0.0736, "step": 3062 }, { "epoch": 0.4946705426356589, "grad_norm": 184.14407348632812, "learning_rate": 5.053294573643411e-06, "loss": 2.1582, "step": 3063 }, { "epoch": 0.49483204134366926, "grad_norm": 184.09996032714844, "learning_rate": 5.051679586563308e-06, "loss": 0.7019, "step": 3064 }, { "epoch": 0.49499354005167956, "grad_norm": 53.48347854614258, "learning_rate": 5.0500645994832045e-06, "loss": 0.3359, "step": 3065 }, { "epoch": 0.4951550387596899, "grad_norm": 97.33977508544922, "learning_rate": 5.048449612403101e-06, "loss": 0.8691, "step": 3066 }, { "epoch": 0.49531653746770027, "grad_norm": 17.367551803588867, "learning_rate": 5.0468346253229985e-06, "loss": 0.0739, "step": 3067 }, { "epoch": 0.49547803617571057, "grad_norm": 46.484336853027344, "learning_rate": 5.045219638242894e-06, "loss": 0.4047, "step": 3068 }, { "epoch": 0.4956395348837209, "grad_norm": 153.64906311035156, "learning_rate": 5.043604651162792e-06, "loss": 1.043, "step": 3069 }, { "epoch": 0.4958010335917313, "grad_norm": 79.61328887939453, "learning_rate": 5.041989664082688e-06, "loss": 0.7422, "step": 3070 }, { "epoch": 0.4959625322997416, "grad_norm": 73.27643585205078, "learning_rate": 5.040374677002584e-06, "loss": 0.7656, "step": 3071 }, { "epoch": 0.49612403100775193, "grad_norm": 75.16970825195312, "learning_rate": 5.038759689922481e-06, "loss": 0.7781, "step": 3072 }, { "epoch": 0.4962855297157623, "grad_norm": 80.0980224609375, "learning_rate": 5.037144702842378e-06, "loss": 0.8945, "step": 3073 }, { "epoch": 0.4964470284237726, "grad_norm": 77.16100311279297, "learning_rate": 5.035529715762274e-06, "loss": 0.8945, "step": 3074 }, { "epoch": 0.49660852713178294, "grad_norm": 96.5589828491211, "learning_rate": 5.033914728682171e-06, "loss": 1.3516, "step": 3075 }, { "epoch": 0.4967700258397933, "grad_norm": 46.6983757019043, "learning_rate": 5.032299741602068e-06, "loss": 0.4023, "step": 3076 }, { "epoch": 0.4969315245478036, "grad_norm": 148.31658935546875, "learning_rate": 5.030684754521964e-06, "loss": 1.625, "step": 3077 }, { "epoch": 0.49709302325581395, "grad_norm": 56.12321090698242, "learning_rate": 5.029069767441861e-06, "loss": 0.3787, "step": 3078 }, { "epoch": 0.4972545219638243, "grad_norm": 88.94259643554688, "learning_rate": 5.027454780361758e-06, "loss": 1.2173, "step": 3079 }, { "epoch": 0.4974160206718346, "grad_norm": 24.028989791870117, "learning_rate": 5.025839793281654e-06, "loss": 0.1289, "step": 3080 }, { "epoch": 0.49757751937984496, "grad_norm": 12.273305892944336, "learning_rate": 5.024224806201551e-06, "loss": 0.0779, "step": 3081 }, { "epoch": 0.4977390180878553, "grad_norm": 95.7322998046875, "learning_rate": 5.022609819121448e-06, "loss": 0.4668, "step": 3082 }, { "epoch": 0.4979005167958656, "grad_norm": 54.38605499267578, "learning_rate": 5.0209948320413434e-06, "loss": 0.5422, "step": 3083 }, { "epoch": 0.49806201550387597, "grad_norm": 28.32834815979004, "learning_rate": 5.019379844961241e-06, "loss": 0.2473, "step": 3084 }, { "epoch": 0.4982235142118863, "grad_norm": 53.16819381713867, "learning_rate": 5.017764857881137e-06, "loss": 0.3467, "step": 3085 }, { "epoch": 0.4983850129198966, "grad_norm": 140.04005432128906, "learning_rate": 5.016149870801034e-06, "loss": 2.2891, "step": 3086 }, { "epoch": 0.498546511627907, "grad_norm": 102.84977722167969, "learning_rate": 5.0145348837209305e-06, "loss": 1.4414, "step": 3087 }, { "epoch": 0.49870801033591733, "grad_norm": 30.02459144592285, "learning_rate": 5.012919896640828e-06, "loss": 0.1602, "step": 3088 }, { "epoch": 0.49886950904392763, "grad_norm": 100.29472351074219, "learning_rate": 5.011304909560724e-06, "loss": 0.8242, "step": 3089 }, { "epoch": 0.499031007751938, "grad_norm": 56.615352630615234, "learning_rate": 5.00968992248062e-06, "loss": 0.4199, "step": 3090 }, { "epoch": 0.49919250645994834, "grad_norm": 73.65623474121094, "learning_rate": 5.008074935400518e-06, "loss": 0.7383, "step": 3091 }, { "epoch": 0.49935400516795864, "grad_norm": 243.367431640625, "learning_rate": 5.006459948320413e-06, "loss": 4.8815, "step": 3092 }, { "epoch": 0.499515503875969, "grad_norm": 9.969426155090332, "learning_rate": 5.004844961240311e-06, "loss": 0.0311, "step": 3093 }, { "epoch": 0.49967700258397935, "grad_norm": 46.2801628112793, "learning_rate": 5.003229974160207e-06, "loss": 0.4879, "step": 3094 }, { "epoch": 0.49983850129198965, "grad_norm": 62.52739334106445, "learning_rate": 5.001614987080103e-06, "loss": 0.4756, "step": 3095 }, { "epoch": 0.5, "grad_norm": 9.038115501403809, "learning_rate": 5e-06, "loss": 0.0302, "step": 3096 }, { "epoch": 0.5001614987080103, "grad_norm": 26.92881965637207, "learning_rate": 4.998385012919897e-06, "loss": 0.1919, "step": 3097 }, { "epoch": 0.5003229974160207, "grad_norm": 27.584131240844727, "learning_rate": 4.9967700258397935e-06, "loss": 0.1735, "step": 3098 }, { "epoch": 0.500484496124031, "grad_norm": 25.789724349975586, "learning_rate": 4.99515503875969e-06, "loss": 0.1271, "step": 3099 }, { "epoch": 0.5006459948320413, "grad_norm": 29.030914306640625, "learning_rate": 4.993540051679587e-06, "loss": 0.1572, "step": 3100 }, { "epoch": 0.5008074935400517, "grad_norm": 38.2061767578125, "learning_rate": 4.991925064599484e-06, "loss": 0.2064, "step": 3101 }, { "epoch": 0.500968992248062, "grad_norm": 41.92595291137695, "learning_rate": 4.990310077519381e-06, "loss": 0.2145, "step": 3102 }, { "epoch": 0.5011304909560723, "grad_norm": 32.00626754760742, "learning_rate": 4.988695090439276e-06, "loss": 0.3306, "step": 3103 }, { "epoch": 0.5012919896640827, "grad_norm": 136.30039978027344, "learning_rate": 4.987080103359174e-06, "loss": 1.2168, "step": 3104 }, { "epoch": 0.501453488372093, "grad_norm": 35.14351272583008, "learning_rate": 4.98546511627907e-06, "loss": 0.209, "step": 3105 }, { "epoch": 0.5016149870801033, "grad_norm": 121.10633087158203, "learning_rate": 4.983850129198967e-06, "loss": 1.3516, "step": 3106 }, { "epoch": 0.5017764857881137, "grad_norm": 58.526756286621094, "learning_rate": 4.982235142118863e-06, "loss": 0.2927, "step": 3107 }, { "epoch": 0.501937984496124, "grad_norm": 29.151161193847656, "learning_rate": 4.98062015503876e-06, "loss": 0.1284, "step": 3108 }, { "epoch": 0.5020994832041343, "grad_norm": 79.79387664794922, "learning_rate": 4.9790051679586565e-06, "loss": 0.4756, "step": 3109 }, { "epoch": 0.5022609819121447, "grad_norm": 24.824419021606445, "learning_rate": 4.977390180878554e-06, "loss": 0.1182, "step": 3110 }, { "epoch": 0.502422480620155, "grad_norm": 123.62291717529297, "learning_rate": 4.97577519379845e-06, "loss": 2.9102, "step": 3111 }, { "epoch": 0.5025839793281653, "grad_norm": 34.30845642089844, "learning_rate": 4.974160206718346e-06, "loss": 0.1868, "step": 3112 }, { "epoch": 0.5027454780361758, "grad_norm": 44.38898849487305, "learning_rate": 4.972545219638244e-06, "loss": 0.3632, "step": 3113 }, { "epoch": 0.502906976744186, "grad_norm": 134.23239135742188, "learning_rate": 4.97093023255814e-06, "loss": 1.3208, "step": 3114 }, { "epoch": 0.5030684754521964, "grad_norm": 100.45787048339844, "learning_rate": 4.969315245478037e-06, "loss": 1.4697, "step": 3115 }, { "epoch": 0.5032299741602068, "grad_norm": 51.75249481201172, "learning_rate": 4.967700258397933e-06, "loss": 0.4526, "step": 3116 }, { "epoch": 0.5033914728682171, "grad_norm": 19.128868103027344, "learning_rate": 4.96608527131783e-06, "loss": 0.1052, "step": 3117 }, { "epoch": 0.5035529715762274, "grad_norm": 157.80209350585938, "learning_rate": 4.964470284237726e-06, "loss": 1.4492, "step": 3118 }, { "epoch": 0.5037144702842378, "grad_norm": 172.3326416015625, "learning_rate": 4.962855297157623e-06, "loss": 1.8906, "step": 3119 }, { "epoch": 0.5038759689922481, "grad_norm": 12.958535194396973, "learning_rate": 4.9612403100775195e-06, "loss": 0.0779, "step": 3120 }, { "epoch": 0.5040374677002584, "grad_norm": 38.56460189819336, "learning_rate": 4.959625322997416e-06, "loss": 0.3701, "step": 3121 }, { "epoch": 0.5041989664082688, "grad_norm": 57.62569808959961, "learning_rate": 4.9580103359173135e-06, "loss": 0.5176, "step": 3122 }, { "epoch": 0.5043604651162791, "grad_norm": 69.64543151855469, "learning_rate": 4.956395348837209e-06, "loss": 0.5049, "step": 3123 }, { "epoch": 0.5045219638242894, "grad_norm": 16.94476318359375, "learning_rate": 4.954780361757106e-06, "loss": 0.0933, "step": 3124 }, { "epoch": 0.5046834625322998, "grad_norm": 17.834148406982422, "learning_rate": 4.953165374677003e-06, "loss": 0.0872, "step": 3125 }, { "epoch": 0.5048449612403101, "grad_norm": 37.65559005737305, "learning_rate": 4.9515503875969e-06, "loss": 0.3828, "step": 3126 }, { "epoch": 0.5050064599483204, "grad_norm": 134.00640869140625, "learning_rate": 4.949935400516796e-06, "loss": 0.8384, "step": 3127 }, { "epoch": 0.5051679586563308, "grad_norm": 18.48673439025879, "learning_rate": 4.948320413436693e-06, "loss": 0.1252, "step": 3128 }, { "epoch": 0.5053294573643411, "grad_norm": 173.31588745117188, "learning_rate": 4.9467054263565894e-06, "loss": 1.4922, "step": 3129 }, { "epoch": 0.5054909560723514, "grad_norm": 13.505708694458008, "learning_rate": 4.945090439276487e-06, "loss": 0.0659, "step": 3130 }, { "epoch": 0.5056524547803618, "grad_norm": 48.145320892333984, "learning_rate": 4.9434754521963825e-06, "loss": 0.5688, "step": 3131 }, { "epoch": 0.5058139534883721, "grad_norm": 4.400786399841309, "learning_rate": 4.941860465116279e-06, "loss": 0.0316, "step": 3132 }, { "epoch": 0.5059754521963824, "grad_norm": 36.83575439453125, "learning_rate": 4.9402454780361765e-06, "loss": 0.3092, "step": 3133 }, { "epoch": 0.5061369509043928, "grad_norm": 34.95946502685547, "learning_rate": 4.938630490956073e-06, "loss": 0.2275, "step": 3134 }, { "epoch": 0.5062984496124031, "grad_norm": 120.81349182128906, "learning_rate": 4.93701550387597e-06, "loss": 1.1055, "step": 3135 }, { "epoch": 0.5064599483204134, "grad_norm": 15.081911087036133, "learning_rate": 4.935400516795866e-06, "loss": 0.0742, "step": 3136 }, { "epoch": 0.5066214470284238, "grad_norm": 65.820068359375, "learning_rate": 4.933785529715763e-06, "loss": 0.4238, "step": 3137 }, { "epoch": 0.5067829457364341, "grad_norm": 19.15339469909668, "learning_rate": 4.932170542635659e-06, "loss": 0.0543, "step": 3138 }, { "epoch": 0.5069444444444444, "grad_norm": 14.720766067504883, "learning_rate": 4.930555555555556e-06, "loss": 0.0346, "step": 3139 }, { "epoch": 0.5071059431524548, "grad_norm": 51.327754974365234, "learning_rate": 4.9289405684754524e-06, "loss": 0.2144, "step": 3140 }, { "epoch": 0.5072674418604651, "grad_norm": 27.57261085510254, "learning_rate": 4.927325581395349e-06, "loss": 0.1464, "step": 3141 }, { "epoch": 0.5074289405684754, "grad_norm": 116.62384796142578, "learning_rate": 4.925710594315246e-06, "loss": 0.9316, "step": 3142 }, { "epoch": 0.5075904392764858, "grad_norm": 20.066104888916016, "learning_rate": 4.924095607235143e-06, "loss": 0.0569, "step": 3143 }, { "epoch": 0.5077519379844961, "grad_norm": 98.053466796875, "learning_rate": 4.922480620155039e-06, "loss": 0.9385, "step": 3144 }, { "epoch": 0.5079134366925064, "grad_norm": 29.90731430053711, "learning_rate": 4.920865633074936e-06, "loss": 0.127, "step": 3145 }, { "epoch": 0.5080749354005168, "grad_norm": 91.60749053955078, "learning_rate": 4.919250645994833e-06, "loss": 0.3144, "step": 3146 }, { "epoch": 0.5082364341085271, "grad_norm": 39.19000244140625, "learning_rate": 4.917635658914729e-06, "loss": 0.2507, "step": 3147 }, { "epoch": 0.5083979328165374, "grad_norm": 29.918710708618164, "learning_rate": 4.916020671834626e-06, "loss": 0.1704, "step": 3148 }, { "epoch": 0.5085594315245479, "grad_norm": 171.10032653808594, "learning_rate": 4.914405684754522e-06, "loss": 5.0801, "step": 3149 }, { "epoch": 0.5087209302325582, "grad_norm": 7.23101282119751, "learning_rate": 4.912790697674419e-06, "loss": 0.0257, "step": 3150 }, { "epoch": 0.5088824289405685, "grad_norm": 86.18959045410156, "learning_rate": 4.9111757105943154e-06, "loss": 0.5172, "step": 3151 }, { "epoch": 0.5090439276485789, "grad_norm": 66.01463317871094, "learning_rate": 4.909560723514212e-06, "loss": 0.3828, "step": 3152 }, { "epoch": 0.5092054263565892, "grad_norm": 32.52999496459961, "learning_rate": 4.9079457364341086e-06, "loss": 0.1461, "step": 3153 }, { "epoch": 0.5093669250645995, "grad_norm": 68.60380554199219, "learning_rate": 4.906330749354006e-06, "loss": 0.5679, "step": 3154 }, { "epoch": 0.5095284237726099, "grad_norm": 57.209999084472656, "learning_rate": 4.9047157622739025e-06, "loss": 0.4507, "step": 3155 }, { "epoch": 0.5096899224806202, "grad_norm": 192.057861328125, "learning_rate": 4.903100775193799e-06, "loss": 1.9609, "step": 3156 }, { "epoch": 0.5098514211886305, "grad_norm": 55.243560791015625, "learning_rate": 4.901485788113696e-06, "loss": 0.3066, "step": 3157 }, { "epoch": 0.5100129198966409, "grad_norm": 15.824583053588867, "learning_rate": 4.899870801033592e-06, "loss": 0.1497, "step": 3158 }, { "epoch": 0.5101744186046512, "grad_norm": 136.13946533203125, "learning_rate": 4.898255813953489e-06, "loss": 1.6992, "step": 3159 }, { "epoch": 0.5103359173126615, "grad_norm": 19.352031707763672, "learning_rate": 4.896640826873385e-06, "loss": 0.0782, "step": 3160 }, { "epoch": 0.5104974160206718, "grad_norm": 19.090736389160156, "learning_rate": 4.895025839793282e-06, "loss": 0.0476, "step": 3161 }, { "epoch": 0.5106589147286822, "grad_norm": 132.46299743652344, "learning_rate": 4.8934108527131785e-06, "loss": 1.0889, "step": 3162 }, { "epoch": 0.5108204134366925, "grad_norm": 6.781325817108154, "learning_rate": 4.891795865633076e-06, "loss": 0.0308, "step": 3163 }, { "epoch": 0.5109819121447028, "grad_norm": 26.527347564697266, "learning_rate": 4.8901808785529716e-06, "loss": 0.1961, "step": 3164 }, { "epoch": 0.5111434108527132, "grad_norm": 32.078678131103516, "learning_rate": 4.888565891472868e-06, "loss": 0.2583, "step": 3165 }, { "epoch": 0.5113049095607235, "grad_norm": 92.67670440673828, "learning_rate": 4.8869509043927655e-06, "loss": 0.7822, "step": 3166 }, { "epoch": 0.5114664082687338, "grad_norm": 9.900257110595703, "learning_rate": 4.885335917312662e-06, "loss": 0.0455, "step": 3167 }, { "epoch": 0.5116279069767442, "grad_norm": 77.57841491699219, "learning_rate": 4.883720930232559e-06, "loss": 0.6455, "step": 3168 }, { "epoch": 0.5117894056847545, "grad_norm": 34.12131118774414, "learning_rate": 4.882105943152455e-06, "loss": 0.1885, "step": 3169 }, { "epoch": 0.5119509043927648, "grad_norm": 65.49581146240234, "learning_rate": 4.880490956072352e-06, "loss": 0.6616, "step": 3170 }, { "epoch": 0.5121124031007752, "grad_norm": 25.27524757385254, "learning_rate": 4.878875968992248e-06, "loss": 0.0931, "step": 3171 }, { "epoch": 0.5122739018087855, "grad_norm": 21.76163101196289, "learning_rate": 4.877260981912145e-06, "loss": 0.1067, "step": 3172 }, { "epoch": 0.5124354005167958, "grad_norm": 132.9540252685547, "learning_rate": 4.8756459948320415e-06, "loss": 1.2751, "step": 3173 }, { "epoch": 0.5125968992248062, "grad_norm": 32.013343811035156, "learning_rate": 4.874031007751938e-06, "loss": 0.2881, "step": 3174 }, { "epoch": 0.5127583979328165, "grad_norm": 47.95347595214844, "learning_rate": 4.8724160206718354e-06, "loss": 0.4052, "step": 3175 }, { "epoch": 0.5129198966408268, "grad_norm": 200.51466369628906, "learning_rate": 4.870801033591732e-06, "loss": 1.7578, "step": 3176 }, { "epoch": 0.5130813953488372, "grad_norm": 81.18551635742188, "learning_rate": 4.869186046511628e-06, "loss": 0.6279, "step": 3177 }, { "epoch": 0.5132428940568475, "grad_norm": 59.75703048706055, "learning_rate": 4.867571059431525e-06, "loss": 0.7936, "step": 3178 }, { "epoch": 0.5134043927648578, "grad_norm": 72.08639526367188, "learning_rate": 4.865956072351422e-06, "loss": 0.5576, "step": 3179 }, { "epoch": 0.5135658914728682, "grad_norm": 43.909828186035156, "learning_rate": 4.864341085271318e-06, "loss": 0.2788, "step": 3180 }, { "epoch": 0.5137273901808785, "grad_norm": 92.96397399902344, "learning_rate": 4.862726098191215e-06, "loss": 1.323, "step": 3181 }, { "epoch": 0.5138888888888888, "grad_norm": 11.179069519042969, "learning_rate": 4.861111111111111e-06, "loss": 0.0425, "step": 3182 }, { "epoch": 0.5140503875968992, "grad_norm": 57.73335266113281, "learning_rate": 4.859496124031008e-06, "loss": 0.3984, "step": 3183 }, { "epoch": 0.5142118863049095, "grad_norm": 125.29632568359375, "learning_rate": 4.857881136950905e-06, "loss": 1.0759, "step": 3184 }, { "epoch": 0.5143733850129198, "grad_norm": 72.84188079833984, "learning_rate": 4.856266149870801e-06, "loss": 0.3218, "step": 3185 }, { "epoch": 0.5145348837209303, "grad_norm": 66.58265686035156, "learning_rate": 4.854651162790698e-06, "loss": 0.3496, "step": 3186 }, { "epoch": 0.5146963824289406, "grad_norm": 16.62007713317871, "learning_rate": 4.853036175710595e-06, "loss": 0.0545, "step": 3187 }, { "epoch": 0.5148578811369509, "grad_norm": 10.277118682861328, "learning_rate": 4.8514211886304916e-06, "loss": 0.045, "step": 3188 }, { "epoch": 0.5150193798449613, "grad_norm": 135.35870361328125, "learning_rate": 4.849806201550388e-06, "loss": 0.8613, "step": 3189 }, { "epoch": 0.5151808785529716, "grad_norm": 15.020659446716309, "learning_rate": 4.848191214470285e-06, "loss": 0.0317, "step": 3190 }, { "epoch": 0.5153423772609819, "grad_norm": 49.485355377197266, "learning_rate": 4.846576227390181e-06, "loss": 0.2378, "step": 3191 }, { "epoch": 0.5155038759689923, "grad_norm": 104.67500305175781, "learning_rate": 4.844961240310078e-06, "loss": 0.8604, "step": 3192 }, { "epoch": 0.5156653746770026, "grad_norm": 84.14425659179688, "learning_rate": 4.843346253229974e-06, "loss": 1.3242, "step": 3193 }, { "epoch": 0.5158268733850129, "grad_norm": 46.89265441894531, "learning_rate": 4.841731266149871e-06, "loss": 0.2524, "step": 3194 }, { "epoch": 0.5159883720930233, "grad_norm": 16.216310501098633, "learning_rate": 4.840116279069768e-06, "loss": 0.1045, "step": 3195 }, { "epoch": 0.5161498708010336, "grad_norm": 222.35948181152344, "learning_rate": 4.838501291989665e-06, "loss": 5.2266, "step": 3196 }, { "epoch": 0.5163113695090439, "grad_norm": 153.93695068359375, "learning_rate": 4.8368863049095614e-06, "loss": 1.7375, "step": 3197 }, { "epoch": 0.5164728682170543, "grad_norm": 158.06707763671875, "learning_rate": 4.835271317829458e-06, "loss": 3.8555, "step": 3198 }, { "epoch": 0.5166343669250646, "grad_norm": 17.119720458984375, "learning_rate": 4.8336563307493546e-06, "loss": 0.0414, "step": 3199 }, { "epoch": 0.5167958656330749, "grad_norm": 137.98924255371094, "learning_rate": 4.832041343669251e-06, "loss": 1.7539, "step": 3200 }, { "epoch": 0.5169573643410853, "grad_norm": 41.39362716674805, "learning_rate": 4.830426356589148e-06, "loss": 0.3973, "step": 3201 }, { "epoch": 0.5171188630490956, "grad_norm": 11.405296325683594, "learning_rate": 4.828811369509044e-06, "loss": 0.0721, "step": 3202 }, { "epoch": 0.5172803617571059, "grad_norm": 69.26708984375, "learning_rate": 4.827196382428941e-06, "loss": 0.6736, "step": 3203 }, { "epoch": 0.5174418604651163, "grad_norm": 13.478693962097168, "learning_rate": 4.825581395348838e-06, "loss": 0.05, "step": 3204 }, { "epoch": 0.5176033591731266, "grad_norm": 26.326175689697266, "learning_rate": 4.823966408268734e-06, "loss": 0.2258, "step": 3205 }, { "epoch": 0.5177648578811369, "grad_norm": 59.499122619628906, "learning_rate": 4.8223514211886305e-06, "loss": 0.6348, "step": 3206 }, { "epoch": 0.5179263565891473, "grad_norm": 25.292390823364258, "learning_rate": 4.820736434108528e-06, "loss": 0.1781, "step": 3207 }, { "epoch": 0.5180878552971576, "grad_norm": 95.37432098388672, "learning_rate": 4.8191214470284244e-06, "loss": 0.6743, "step": 3208 }, { "epoch": 0.5182493540051679, "grad_norm": 58.79099655151367, "learning_rate": 4.817506459948321e-06, "loss": 0.543, "step": 3209 }, { "epoch": 0.5184108527131783, "grad_norm": 110.6507568359375, "learning_rate": 4.8158914728682176e-06, "loss": 1.043, "step": 3210 }, { "epoch": 0.5185723514211886, "grad_norm": 15.199731826782227, "learning_rate": 4.814276485788114e-06, "loss": 0.0959, "step": 3211 }, { "epoch": 0.5187338501291989, "grad_norm": 84.52584838867188, "learning_rate": 4.812661498708011e-06, "loss": 0.0967, "step": 3212 }, { "epoch": 0.5188953488372093, "grad_norm": 92.46298217773438, "learning_rate": 4.811046511627907e-06, "loss": 0.666, "step": 3213 }, { "epoch": 0.5190568475452196, "grad_norm": 282.18634033203125, "learning_rate": 4.809431524547804e-06, "loss": 3.0703, "step": 3214 }, { "epoch": 0.5192183462532299, "grad_norm": 67.38877868652344, "learning_rate": 4.8078165374677e-06, "loss": 0.7991, "step": 3215 }, { "epoch": 0.5193798449612403, "grad_norm": 67.27672576904297, "learning_rate": 4.806201550387598e-06, "loss": 0.4312, "step": 3216 }, { "epoch": 0.5195413436692506, "grad_norm": 64.688232421875, "learning_rate": 4.804586563307494e-06, "loss": 0.458, "step": 3217 }, { "epoch": 0.5197028423772609, "grad_norm": 115.49765014648438, "learning_rate": 4.80297157622739e-06, "loss": 0.9957, "step": 3218 }, { "epoch": 0.5198643410852714, "grad_norm": 21.248186111450195, "learning_rate": 4.8013565891472875e-06, "loss": 0.1174, "step": 3219 }, { "epoch": 0.5200258397932817, "grad_norm": 91.23150634765625, "learning_rate": 4.799741602067184e-06, "loss": 0.9467, "step": 3220 }, { "epoch": 0.520187338501292, "grad_norm": 14.59495735168457, "learning_rate": 4.7981266149870806e-06, "loss": 0.0807, "step": 3221 }, { "epoch": 0.5203488372093024, "grad_norm": 37.418575286865234, "learning_rate": 4.796511627906977e-06, "loss": 0.1943, "step": 3222 }, { "epoch": 0.5205103359173127, "grad_norm": 54.99848175048828, "learning_rate": 4.794896640826874e-06, "loss": 0.426, "step": 3223 }, { "epoch": 0.520671834625323, "grad_norm": 25.076854705810547, "learning_rate": 4.79328165374677e-06, "loss": 0.0848, "step": 3224 }, { "epoch": 0.5208333333333334, "grad_norm": 51.20491409301758, "learning_rate": 4.791666666666668e-06, "loss": 0.3276, "step": 3225 }, { "epoch": 0.5209948320413437, "grad_norm": 6.554235458374023, "learning_rate": 4.790051679586563e-06, "loss": 0.0239, "step": 3226 }, { "epoch": 0.521156330749354, "grad_norm": 95.15316772460938, "learning_rate": 4.78843669250646e-06, "loss": 0.4081, "step": 3227 }, { "epoch": 0.5213178294573644, "grad_norm": 96.29405212402344, "learning_rate": 4.786821705426357e-06, "loss": 0.9869, "step": 3228 }, { "epoch": 0.5214793281653747, "grad_norm": 6.38887357711792, "learning_rate": 4.785206718346254e-06, "loss": 0.0212, "step": 3229 }, { "epoch": 0.521640826873385, "grad_norm": 86.56709289550781, "learning_rate": 4.7835917312661505e-06, "loss": 0.5674, "step": 3230 }, { "epoch": 0.5218023255813954, "grad_norm": 10.58948802947998, "learning_rate": 4.781976744186047e-06, "loss": 0.0599, "step": 3231 }, { "epoch": 0.5219638242894057, "grad_norm": 17.851285934448242, "learning_rate": 4.780361757105944e-06, "loss": 0.0441, "step": 3232 }, { "epoch": 0.522125322997416, "grad_norm": 31.490718841552734, "learning_rate": 4.77874677002584e-06, "loss": 0.145, "step": 3233 }, { "epoch": 0.5222868217054264, "grad_norm": 17.919227600097656, "learning_rate": 4.777131782945737e-06, "loss": 0.0582, "step": 3234 }, { "epoch": 0.5224483204134367, "grad_norm": 13.788688659667969, "learning_rate": 4.775516795865633e-06, "loss": 0.1078, "step": 3235 }, { "epoch": 0.522609819121447, "grad_norm": 3.262014150619507, "learning_rate": 4.77390180878553e-06, "loss": 0.0068, "step": 3236 }, { "epoch": 0.5227713178294574, "grad_norm": 67.18985748291016, "learning_rate": 4.772286821705427e-06, "loss": 0.4409, "step": 3237 }, { "epoch": 0.5229328165374677, "grad_norm": 115.5505599975586, "learning_rate": 4.770671834625323e-06, "loss": 1.0469, "step": 3238 }, { "epoch": 0.523094315245478, "grad_norm": 102.99254608154297, "learning_rate": 4.7690568475452195e-06, "loss": 1.0547, "step": 3239 }, { "epoch": 0.5232558139534884, "grad_norm": 58.95829772949219, "learning_rate": 4.767441860465117e-06, "loss": 0.5503, "step": 3240 }, { "epoch": 0.5234173126614987, "grad_norm": 174.47447204589844, "learning_rate": 4.7658268733850135e-06, "loss": 3.3203, "step": 3241 }, { "epoch": 0.523578811369509, "grad_norm": 14.548066139221191, "learning_rate": 4.76421188630491e-06, "loss": 0.0279, "step": 3242 }, { "epoch": 0.5237403100775194, "grad_norm": 47.631736755371094, "learning_rate": 4.762596899224807e-06, "loss": 0.4614, "step": 3243 }, { "epoch": 0.5239018087855297, "grad_norm": 119.13699340820312, "learning_rate": 4.760981912144703e-06, "loss": 0.9688, "step": 3244 }, { "epoch": 0.52406330749354, "grad_norm": 13.47996711730957, "learning_rate": 4.7593669250646e-06, "loss": 0.0323, "step": 3245 }, { "epoch": 0.5242248062015504, "grad_norm": 194.7312774658203, "learning_rate": 4.757751937984496e-06, "loss": 2.9355, "step": 3246 }, { "epoch": 0.5243863049095607, "grad_norm": 61.13215637207031, "learning_rate": 4.756136950904393e-06, "loss": 0.4766, "step": 3247 }, { "epoch": 0.524547803617571, "grad_norm": 84.43917083740234, "learning_rate": 4.754521963824289e-06, "loss": 0.5026, "step": 3248 }, { "epoch": 0.5247093023255814, "grad_norm": 421.2065124511719, "learning_rate": 4.752906976744187e-06, "loss": 1.6445, "step": 3249 }, { "epoch": 0.5248708010335917, "grad_norm": 22.701152801513672, "learning_rate": 4.751291989664083e-06, "loss": 0.0718, "step": 3250 }, { "epoch": 0.525032299741602, "grad_norm": 14.09085464477539, "learning_rate": 4.749677002583979e-06, "loss": 0.0775, "step": 3251 }, { "epoch": 0.5251937984496124, "grad_norm": 165.6508331298828, "learning_rate": 4.7480620155038765e-06, "loss": 2.1543, "step": 3252 }, { "epoch": 0.5253552971576227, "grad_norm": 76.27933502197266, "learning_rate": 4.746447028423773e-06, "loss": 1.5381, "step": 3253 }, { "epoch": 0.525516795865633, "grad_norm": 14.460216522216797, "learning_rate": 4.74483204134367e-06, "loss": 0.0543, "step": 3254 }, { "epoch": 0.5256782945736435, "grad_norm": 68.79252624511719, "learning_rate": 4.743217054263566e-06, "loss": 0.5015, "step": 3255 }, { "epoch": 0.5258397932816538, "grad_norm": 194.0127410888672, "learning_rate": 4.741602067183463e-06, "loss": 5.1816, "step": 3256 }, { "epoch": 0.526001291989664, "grad_norm": 27.324499130249023, "learning_rate": 4.739987080103359e-06, "loss": 0.0978, "step": 3257 }, { "epoch": 0.5261627906976745, "grad_norm": 66.11825561523438, "learning_rate": 4.738372093023257e-06, "loss": 0.7324, "step": 3258 }, { "epoch": 0.5263242894056848, "grad_norm": 51.78776168823242, "learning_rate": 4.736757105943152e-06, "loss": 0.4675, "step": 3259 }, { "epoch": 0.5264857881136951, "grad_norm": 24.403554916381836, "learning_rate": 4.73514211886305e-06, "loss": 0.0629, "step": 3260 }, { "epoch": 0.5266472868217055, "grad_norm": 14.2928466796875, "learning_rate": 4.733527131782946e-06, "loss": 0.0376, "step": 3261 }, { "epoch": 0.5268087855297158, "grad_norm": 56.892173767089844, "learning_rate": 4.731912144702843e-06, "loss": 0.5063, "step": 3262 }, { "epoch": 0.5269702842377261, "grad_norm": 105.70085906982422, "learning_rate": 4.7302971576227395e-06, "loss": 0.6289, "step": 3263 }, { "epoch": 0.5271317829457365, "grad_norm": 9.376158714294434, "learning_rate": 4.728682170542636e-06, "loss": 0.0233, "step": 3264 }, { "epoch": 0.5272932816537468, "grad_norm": 27.741899490356445, "learning_rate": 4.727067183462533e-06, "loss": 0.1284, "step": 3265 }, { "epoch": 0.5274547803617571, "grad_norm": 47.62760543823242, "learning_rate": 4.725452196382429e-06, "loss": 0.4009, "step": 3266 }, { "epoch": 0.5276162790697675, "grad_norm": 156.98797607421875, "learning_rate": 4.723837209302326e-06, "loss": 2.1611, "step": 3267 }, { "epoch": 0.5277777777777778, "grad_norm": 213.14767456054688, "learning_rate": 4.722222222222222e-06, "loss": 0.873, "step": 3268 }, { "epoch": 0.5279392764857881, "grad_norm": 19.52714729309082, "learning_rate": 4.72060723514212e-06, "loss": 0.0652, "step": 3269 }, { "epoch": 0.5281007751937985, "grad_norm": 37.040626525878906, "learning_rate": 4.718992248062016e-06, "loss": 0.1815, "step": 3270 }, { "epoch": 0.5282622739018088, "grad_norm": 42.34280014038086, "learning_rate": 4.717377260981913e-06, "loss": 0.2148, "step": 3271 }, { "epoch": 0.5284237726098191, "grad_norm": 134.55252075195312, "learning_rate": 4.715762273901809e-06, "loss": 1.9922, "step": 3272 }, { "epoch": 0.5285852713178295, "grad_norm": 19.20001983642578, "learning_rate": 4.714147286821706e-06, "loss": 0.1024, "step": 3273 }, { "epoch": 0.5287467700258398, "grad_norm": 51.79813003540039, "learning_rate": 4.7125322997416025e-06, "loss": 0.4072, "step": 3274 }, { "epoch": 0.5289082687338501, "grad_norm": 98.01510620117188, "learning_rate": 4.710917312661499e-06, "loss": 0.832, "step": 3275 }, { "epoch": 0.5290697674418605, "grad_norm": 43.49917221069336, "learning_rate": 4.709302325581396e-06, "loss": 0.3094, "step": 3276 }, { "epoch": 0.5292312661498708, "grad_norm": 13.763124465942383, "learning_rate": 4.707687338501292e-06, "loss": 0.1021, "step": 3277 }, { "epoch": 0.5293927648578811, "grad_norm": 49.20454788208008, "learning_rate": 4.7060723514211896e-06, "loss": 0.5781, "step": 3278 }, { "epoch": 0.5295542635658915, "grad_norm": 200.97796630859375, "learning_rate": 4.704457364341085e-06, "loss": 2.4268, "step": 3279 }, { "epoch": 0.5297157622739018, "grad_norm": 36.67306900024414, "learning_rate": 4.702842377260982e-06, "loss": 0.1465, "step": 3280 }, { "epoch": 0.5298772609819121, "grad_norm": 11.99143123626709, "learning_rate": 4.701227390180879e-06, "loss": 0.0426, "step": 3281 }, { "epoch": 0.5300387596899225, "grad_norm": 17.471858978271484, "learning_rate": 4.699612403100776e-06, "loss": 0.043, "step": 3282 }, { "epoch": 0.5302002583979328, "grad_norm": 51.85344696044922, "learning_rate": 4.697997416020672e-06, "loss": 0.2324, "step": 3283 }, { "epoch": 0.5303617571059431, "grad_norm": 44.582366943359375, "learning_rate": 4.696382428940569e-06, "loss": 0.2598, "step": 3284 }, { "epoch": 0.5305232558139535, "grad_norm": 53.86716842651367, "learning_rate": 4.6947674418604655e-06, "loss": 0.7271, "step": 3285 }, { "epoch": 0.5306847545219638, "grad_norm": 25.92633628845215, "learning_rate": 4.693152454780362e-06, "loss": 0.165, "step": 3286 }, { "epoch": 0.5308462532299741, "grad_norm": 27.32276725769043, "learning_rate": 4.691537467700259e-06, "loss": 0.2181, "step": 3287 }, { "epoch": 0.5310077519379846, "grad_norm": 119.16604614257812, "learning_rate": 4.689922480620155e-06, "loss": 1.4766, "step": 3288 }, { "epoch": 0.5311692506459949, "grad_norm": 148.1122283935547, "learning_rate": 4.688307493540052e-06, "loss": 0.748, "step": 3289 }, { "epoch": 0.5313307493540051, "grad_norm": 153.95962524414062, "learning_rate": 4.686692506459949e-06, "loss": 1.2393, "step": 3290 }, { "epoch": 0.5314922480620154, "grad_norm": 37.11846160888672, "learning_rate": 4.685077519379846e-06, "loss": 0.4746, "step": 3291 }, { "epoch": 0.5316537467700259, "grad_norm": 1.249842882156372, "learning_rate": 4.683462532299741e-06, "loss": 0.0066, "step": 3292 }, { "epoch": 0.5318152454780362, "grad_norm": 89.70956420898438, "learning_rate": 4.681847545219639e-06, "loss": 0.6445, "step": 3293 }, { "epoch": 0.5319767441860465, "grad_norm": 29.422258377075195, "learning_rate": 4.680232558139535e-06, "loss": 0.2222, "step": 3294 }, { "epoch": 0.5321382428940569, "grad_norm": 12.994044303894043, "learning_rate": 4.678617571059432e-06, "loss": 0.0579, "step": 3295 }, { "epoch": 0.5322997416020672, "grad_norm": 48.40529251098633, "learning_rate": 4.6770025839793285e-06, "loss": 0.3628, "step": 3296 }, { "epoch": 0.5324612403100775, "grad_norm": 30.70733642578125, "learning_rate": 4.675387596899225e-06, "loss": 0.147, "step": 3297 }, { "epoch": 0.5326227390180879, "grad_norm": 5.069884777069092, "learning_rate": 4.673772609819122e-06, "loss": 0.0292, "step": 3298 }, { "epoch": 0.5327842377260982, "grad_norm": 117.31562042236328, "learning_rate": 4.672157622739019e-06, "loss": 0.6516, "step": 3299 }, { "epoch": 0.5329457364341085, "grad_norm": 17.22684097290039, "learning_rate": 4.670542635658915e-06, "loss": 0.08, "step": 3300 }, { "epoch": 0.5331072351421189, "grad_norm": 29.266990661621094, "learning_rate": 4.668927648578811e-06, "loss": 0.145, "step": 3301 }, { "epoch": 0.5332687338501292, "grad_norm": 93.14310455322266, "learning_rate": 4.667312661498709e-06, "loss": 0.8936, "step": 3302 }, { "epoch": 0.5334302325581395, "grad_norm": 11.651708602905273, "learning_rate": 4.665697674418605e-06, "loss": 0.0811, "step": 3303 }, { "epoch": 0.5335917312661499, "grad_norm": 21.40642738342285, "learning_rate": 4.664082687338502e-06, "loss": 0.1436, "step": 3304 }, { "epoch": 0.5337532299741602, "grad_norm": 68.87507629394531, "learning_rate": 4.662467700258398e-06, "loss": 0.4614, "step": 3305 }, { "epoch": 0.5339147286821705, "grad_norm": 221.23031616210938, "learning_rate": 4.660852713178295e-06, "loss": 1.6963, "step": 3306 }, { "epoch": 0.5340762273901809, "grad_norm": 107.56267547607422, "learning_rate": 4.6592377260981915e-06, "loss": 0.5026, "step": 3307 }, { "epoch": 0.5342377260981912, "grad_norm": 11.92765998840332, "learning_rate": 4.657622739018088e-06, "loss": 0.0642, "step": 3308 }, { "epoch": 0.5343992248062015, "grad_norm": 11.871277809143066, "learning_rate": 4.656007751937985e-06, "loss": 0.0648, "step": 3309 }, { "epoch": 0.5345607235142119, "grad_norm": 76.00794982910156, "learning_rate": 4.654392764857881e-06, "loss": 1.009, "step": 3310 }, { "epoch": 0.5347222222222222, "grad_norm": 168.94647216796875, "learning_rate": 4.652777777777779e-06, "loss": 1.6396, "step": 3311 }, { "epoch": 0.5348837209302325, "grad_norm": 96.15467834472656, "learning_rate": 4.651162790697675e-06, "loss": 0.8789, "step": 3312 }, { "epoch": 0.5350452196382429, "grad_norm": 59.989864349365234, "learning_rate": 4.649547803617571e-06, "loss": 0.4336, "step": 3313 }, { "epoch": 0.5352067183462532, "grad_norm": 72.37677764892578, "learning_rate": 4.647932816537468e-06, "loss": 0.793, "step": 3314 }, { "epoch": 0.5353682170542635, "grad_norm": 109.11213684082031, "learning_rate": 4.646317829457365e-06, "loss": 0.8887, "step": 3315 }, { "epoch": 0.5355297157622739, "grad_norm": 38.61005401611328, "learning_rate": 4.644702842377261e-06, "loss": 0.355, "step": 3316 }, { "epoch": 0.5356912144702842, "grad_norm": 44.01213836669922, "learning_rate": 4.643087855297158e-06, "loss": 0.3951, "step": 3317 }, { "epoch": 0.5358527131782945, "grad_norm": 128.79067993164062, "learning_rate": 4.6414728682170545e-06, "loss": 2.4683, "step": 3318 }, { "epoch": 0.5360142118863049, "grad_norm": 19.259336471557617, "learning_rate": 4.639857881136951e-06, "loss": 0.1521, "step": 3319 }, { "epoch": 0.5361757105943152, "grad_norm": 55.14324951171875, "learning_rate": 4.638242894056848e-06, "loss": 0.2517, "step": 3320 }, { "epoch": 0.5363372093023255, "grad_norm": 21.599918365478516, "learning_rate": 4.636627906976744e-06, "loss": 0.0708, "step": 3321 }, { "epoch": 0.536498708010336, "grad_norm": 98.98516082763672, "learning_rate": 4.635012919896642e-06, "loss": 0.5156, "step": 3322 }, { "epoch": 0.5366602067183462, "grad_norm": 147.421142578125, "learning_rate": 4.633397932816538e-06, "loss": 1.8789, "step": 3323 }, { "epoch": 0.5368217054263565, "grad_norm": 27.04939079284668, "learning_rate": 4.631782945736435e-06, "loss": 0.1564, "step": 3324 }, { "epoch": 0.536983204134367, "grad_norm": 130.88372802734375, "learning_rate": 4.630167958656331e-06, "loss": 0.9966, "step": 3325 }, { "epoch": 0.5371447028423773, "grad_norm": 64.1499252319336, "learning_rate": 4.628552971576228e-06, "loss": 0.8066, "step": 3326 }, { "epoch": 0.5373062015503876, "grad_norm": 41.65611267089844, "learning_rate": 4.626937984496124e-06, "loss": 0.2279, "step": 3327 }, { "epoch": 0.537467700258398, "grad_norm": 60.55613708496094, "learning_rate": 4.625322997416021e-06, "loss": 0.626, "step": 3328 }, { "epoch": 0.5376291989664083, "grad_norm": 55.70313262939453, "learning_rate": 4.6237080103359175e-06, "loss": 0.3723, "step": 3329 }, { "epoch": 0.5377906976744186, "grad_norm": 95.59894561767578, "learning_rate": 4.622093023255814e-06, "loss": 0.7715, "step": 3330 }, { "epoch": 0.537952196382429, "grad_norm": 34.848365783691406, "learning_rate": 4.6204780361757115e-06, "loss": 0.0831, "step": 3331 }, { "epoch": 0.5381136950904393, "grad_norm": 72.97459411621094, "learning_rate": 4.618863049095608e-06, "loss": 0.3408, "step": 3332 }, { "epoch": 0.5382751937984496, "grad_norm": 171.3072967529297, "learning_rate": 4.617248062015504e-06, "loss": 0.6115, "step": 3333 }, { "epoch": 0.53843669250646, "grad_norm": 28.200420379638672, "learning_rate": 4.615633074935401e-06, "loss": 0.1544, "step": 3334 }, { "epoch": 0.5385981912144703, "grad_norm": 64.2269058227539, "learning_rate": 4.614018087855298e-06, "loss": 0.1411, "step": 3335 }, { "epoch": 0.5387596899224806, "grad_norm": 78.39427947998047, "learning_rate": 4.612403100775194e-06, "loss": 0.5304, "step": 3336 }, { "epoch": 0.538921188630491, "grad_norm": 49.281410217285156, "learning_rate": 4.610788113695091e-06, "loss": 0.582, "step": 3337 }, { "epoch": 0.5390826873385013, "grad_norm": 15.996264457702637, "learning_rate": 4.609173126614987e-06, "loss": 0.0721, "step": 3338 }, { "epoch": 0.5392441860465116, "grad_norm": 88.71039581298828, "learning_rate": 4.607558139534884e-06, "loss": 0.9961, "step": 3339 }, { "epoch": 0.539405684754522, "grad_norm": 10.80495548248291, "learning_rate": 4.605943152454781e-06, "loss": 0.0598, "step": 3340 }, { "epoch": 0.5395671834625323, "grad_norm": 32.62382888793945, "learning_rate": 4.604328165374677e-06, "loss": 0.1158, "step": 3341 }, { "epoch": 0.5397286821705426, "grad_norm": 70.0945053100586, "learning_rate": 4.602713178294574e-06, "loss": 0.1608, "step": 3342 }, { "epoch": 0.539890180878553, "grad_norm": 3.7786529064178467, "learning_rate": 4.601098191214471e-06, "loss": 0.0077, "step": 3343 }, { "epoch": 0.5400516795865633, "grad_norm": 65.2151870727539, "learning_rate": 4.599483204134368e-06, "loss": 0.5103, "step": 3344 }, { "epoch": 0.5402131782945736, "grad_norm": 9.960440635681152, "learning_rate": 4.597868217054264e-06, "loss": 0.0369, "step": 3345 }, { "epoch": 0.540374677002584, "grad_norm": 25.28655242919922, "learning_rate": 4.596253229974161e-06, "loss": 0.0727, "step": 3346 }, { "epoch": 0.5405361757105943, "grad_norm": 35.07341384887695, "learning_rate": 4.594638242894057e-06, "loss": 0.1938, "step": 3347 }, { "epoch": 0.5406976744186046, "grad_norm": 138.60012817382812, "learning_rate": 4.593023255813954e-06, "loss": 2.2344, "step": 3348 }, { "epoch": 0.540859173126615, "grad_norm": 44.58335494995117, "learning_rate": 4.59140826873385e-06, "loss": 0.2596, "step": 3349 }, { "epoch": 0.5410206718346253, "grad_norm": 35.812103271484375, "learning_rate": 4.589793281653747e-06, "loss": 0.1409, "step": 3350 }, { "epoch": 0.5411821705426356, "grad_norm": 16.228178024291992, "learning_rate": 4.5881782945736435e-06, "loss": 0.0462, "step": 3351 }, { "epoch": 0.541343669250646, "grad_norm": 57.4921760559082, "learning_rate": 4.586563307493541e-06, "loss": 0.2591, "step": 3352 }, { "epoch": 0.5415051679586563, "grad_norm": 23.823728561401367, "learning_rate": 4.584948320413437e-06, "loss": 0.1519, "step": 3353 }, { "epoch": 0.5416666666666666, "grad_norm": 158.8123016357422, "learning_rate": 4.583333333333333e-06, "loss": 2.1133, "step": 3354 }, { "epoch": 0.541828165374677, "grad_norm": 348.30133056640625, "learning_rate": 4.581718346253231e-06, "loss": 2.5215, "step": 3355 }, { "epoch": 0.5419896640826873, "grad_norm": 61.19696044921875, "learning_rate": 4.580103359173127e-06, "loss": 0.7041, "step": 3356 }, { "epoch": 0.5421511627906976, "grad_norm": 45.25624465942383, "learning_rate": 4.578488372093024e-06, "loss": 0.4174, "step": 3357 }, { "epoch": 0.542312661498708, "grad_norm": 94.52564239501953, "learning_rate": 4.57687338501292e-06, "loss": 0.687, "step": 3358 }, { "epoch": 0.5424741602067183, "grad_norm": 36.83036422729492, "learning_rate": 4.575258397932817e-06, "loss": 0.2266, "step": 3359 }, { "epoch": 0.5426356589147286, "grad_norm": 86.54500579833984, "learning_rate": 4.573643410852713e-06, "loss": 0.3276, "step": 3360 }, { "epoch": 0.5427971576227391, "grad_norm": 29.032243728637695, "learning_rate": 4.57202842377261e-06, "loss": 0.3848, "step": 3361 }, { "epoch": 0.5429586563307494, "grad_norm": 6.0796403884887695, "learning_rate": 4.5704134366925065e-06, "loss": 0.0161, "step": 3362 }, { "epoch": 0.5431201550387597, "grad_norm": 166.669921875, "learning_rate": 4.568798449612403e-06, "loss": 1.3613, "step": 3363 }, { "epoch": 0.5432816537467701, "grad_norm": 281.16046142578125, "learning_rate": 4.5671834625323005e-06, "loss": 4.6172, "step": 3364 }, { "epoch": 0.5434431524547804, "grad_norm": 77.44474029541016, "learning_rate": 4.565568475452197e-06, "loss": 0.9556, "step": 3365 }, { "epoch": 0.5436046511627907, "grad_norm": 92.287841796875, "learning_rate": 4.563953488372093e-06, "loss": 1.0762, "step": 3366 }, { "epoch": 0.5437661498708011, "grad_norm": 130.02964782714844, "learning_rate": 4.56233850129199e-06, "loss": 1.5195, "step": 3367 }, { "epoch": 0.5439276485788114, "grad_norm": 26.397775650024414, "learning_rate": 4.560723514211887e-06, "loss": 0.0846, "step": 3368 }, { "epoch": 0.5440891472868217, "grad_norm": 51.313785552978516, "learning_rate": 4.559108527131783e-06, "loss": 0.311, "step": 3369 }, { "epoch": 0.5442506459948321, "grad_norm": 8.846057891845703, "learning_rate": 4.55749354005168e-06, "loss": 0.0546, "step": 3370 }, { "epoch": 0.5444121447028424, "grad_norm": 150.3097381591797, "learning_rate": 4.5558785529715764e-06, "loss": 1.1992, "step": 3371 }, { "epoch": 0.5445736434108527, "grad_norm": 87.87647247314453, "learning_rate": 4.554263565891473e-06, "loss": 0.3865, "step": 3372 }, { "epoch": 0.5447351421188631, "grad_norm": 95.44005584716797, "learning_rate": 4.55264857881137e-06, "loss": 0.5298, "step": 3373 }, { "epoch": 0.5448966408268734, "grad_norm": 165.783935546875, "learning_rate": 4.551033591731266e-06, "loss": 1.748, "step": 3374 }, { "epoch": 0.5450581395348837, "grad_norm": 187.95436096191406, "learning_rate": 4.549418604651163e-06, "loss": 2.4775, "step": 3375 }, { "epoch": 0.5452196382428941, "grad_norm": 33.71186065673828, "learning_rate": 4.54780361757106e-06, "loss": 0.4517, "step": 3376 }, { "epoch": 0.5453811369509044, "grad_norm": 49.20877456665039, "learning_rate": 4.546188630490957e-06, "loss": 0.5078, "step": 3377 }, { "epoch": 0.5455426356589147, "grad_norm": 80.92903137207031, "learning_rate": 4.544573643410853e-06, "loss": 0.8088, "step": 3378 }, { "epoch": 0.5457041343669251, "grad_norm": 16.748706817626953, "learning_rate": 4.54295865633075e-06, "loss": 0.1045, "step": 3379 }, { "epoch": 0.5458656330749354, "grad_norm": 42.86079025268555, "learning_rate": 4.541343669250646e-06, "loss": 0.2295, "step": 3380 }, { "epoch": 0.5460271317829457, "grad_norm": 193.4375762939453, "learning_rate": 4.539728682170543e-06, "loss": 1.0625, "step": 3381 }, { "epoch": 0.5461886304909561, "grad_norm": 214.58648681640625, "learning_rate": 4.5381136950904394e-06, "loss": 3.5, "step": 3382 }, { "epoch": 0.5463501291989664, "grad_norm": 22.237281799316406, "learning_rate": 4.536498708010336e-06, "loss": 0.1044, "step": 3383 }, { "epoch": 0.5465116279069767, "grad_norm": 23.001487731933594, "learning_rate": 4.5348837209302326e-06, "loss": 0.1143, "step": 3384 }, { "epoch": 0.5466731266149871, "grad_norm": 56.90681457519531, "learning_rate": 4.53326873385013e-06, "loss": 0.3516, "step": 3385 }, { "epoch": 0.5468346253229974, "grad_norm": 200.552978515625, "learning_rate": 4.5316537467700265e-06, "loss": 3.6143, "step": 3386 }, { "epoch": 0.5469961240310077, "grad_norm": 59.17742919921875, "learning_rate": 4.530038759689923e-06, "loss": 1.1826, "step": 3387 }, { "epoch": 0.5471576227390181, "grad_norm": 69.56758880615234, "learning_rate": 4.52842377260982e-06, "loss": 0.6589, "step": 3388 }, { "epoch": 0.5473191214470284, "grad_norm": 225.65797424316406, "learning_rate": 4.526808785529716e-06, "loss": 3.6315, "step": 3389 }, { "epoch": 0.5474806201550387, "grad_norm": 17.226320266723633, "learning_rate": 4.525193798449613e-06, "loss": 0.0605, "step": 3390 }, { "epoch": 0.5476421188630491, "grad_norm": 28.564016342163086, "learning_rate": 4.523578811369509e-06, "loss": 0.1454, "step": 3391 }, { "epoch": 0.5478036175710594, "grad_norm": 88.12540435791016, "learning_rate": 4.521963824289406e-06, "loss": 0.6816, "step": 3392 }, { "epoch": 0.5479651162790697, "grad_norm": 28.429101943969727, "learning_rate": 4.520348837209303e-06, "loss": 0.145, "step": 3393 }, { "epoch": 0.5481266149870802, "grad_norm": 17.5483455657959, "learning_rate": 4.518733850129199e-06, "loss": 0.063, "step": 3394 }, { "epoch": 0.5482881136950905, "grad_norm": 22.734643936157227, "learning_rate": 4.5171188630490956e-06, "loss": 0.1084, "step": 3395 }, { "epoch": 0.5484496124031008, "grad_norm": 24.243791580200195, "learning_rate": 4.515503875968993e-06, "loss": 0.1783, "step": 3396 }, { "epoch": 0.5486111111111112, "grad_norm": 14.224390029907227, "learning_rate": 4.5138888888888895e-06, "loss": 0.1072, "step": 3397 }, { "epoch": 0.5487726098191215, "grad_norm": 176.5659942626953, "learning_rate": 4.512273901808786e-06, "loss": 2.7306, "step": 3398 }, { "epoch": 0.5489341085271318, "grad_norm": 96.75839233398438, "learning_rate": 4.510658914728683e-06, "loss": 0.3387, "step": 3399 }, { "epoch": 0.5490956072351422, "grad_norm": 46.24894714355469, "learning_rate": 4.509043927648579e-06, "loss": 0.2695, "step": 3400 }, { "epoch": 0.5492571059431525, "grad_norm": 66.8849868774414, "learning_rate": 4.507428940568476e-06, "loss": 0.6523, "step": 3401 }, { "epoch": 0.5494186046511628, "grad_norm": 6.591047763824463, "learning_rate": 4.505813953488372e-06, "loss": 0.0389, "step": 3402 }, { "epoch": 0.5495801033591732, "grad_norm": 79.73851013183594, "learning_rate": 4.504198966408269e-06, "loss": 0.5293, "step": 3403 }, { "epoch": 0.5497416020671835, "grad_norm": 29.41836929321289, "learning_rate": 4.5025839793281654e-06, "loss": 0.2167, "step": 3404 }, { "epoch": 0.5499031007751938, "grad_norm": 13.365338325500488, "learning_rate": 4.500968992248063e-06, "loss": 0.0925, "step": 3405 }, { "epoch": 0.5500645994832042, "grad_norm": 36.35478591918945, "learning_rate": 4.499354005167959e-06, "loss": 0.1704, "step": 3406 }, { "epoch": 0.5502260981912145, "grad_norm": 45.96434783935547, "learning_rate": 4.497739018087855e-06, "loss": 0.4395, "step": 3407 }, { "epoch": 0.5503875968992248, "grad_norm": 48.25083923339844, "learning_rate": 4.4961240310077525e-06, "loss": 0.5801, "step": 3408 }, { "epoch": 0.5505490956072352, "grad_norm": 16.36669921875, "learning_rate": 4.494509043927649e-06, "loss": 0.046, "step": 3409 }, { "epoch": 0.5507105943152455, "grad_norm": 106.65106201171875, "learning_rate": 4.492894056847546e-06, "loss": 1.1855, "step": 3410 }, { "epoch": 0.5508720930232558, "grad_norm": 35.901485443115234, "learning_rate": 4.491279069767442e-06, "loss": 0.6768, "step": 3411 }, { "epoch": 0.5510335917312662, "grad_norm": 34.36276626586914, "learning_rate": 4.489664082687339e-06, "loss": 0.2189, "step": 3412 }, { "epoch": 0.5511950904392765, "grad_norm": 33.169742584228516, "learning_rate": 4.488049095607235e-06, "loss": 0.2409, "step": 3413 }, { "epoch": 0.5513565891472868, "grad_norm": 5.345489025115967, "learning_rate": 4.486434108527133e-06, "loss": 0.0152, "step": 3414 }, { "epoch": 0.5515180878552972, "grad_norm": 176.1214141845703, "learning_rate": 4.4848191214470285e-06, "loss": 1.4634, "step": 3415 }, { "epoch": 0.5516795865633075, "grad_norm": 71.69219970703125, "learning_rate": 4.483204134366925e-06, "loss": 1.3418, "step": 3416 }, { "epoch": 0.5518410852713178, "grad_norm": 51.80827713012695, "learning_rate": 4.481589147286822e-06, "loss": 0.4152, "step": 3417 }, { "epoch": 0.5520025839793282, "grad_norm": 179.29627990722656, "learning_rate": 4.479974160206719e-06, "loss": 2.8359, "step": 3418 }, { "epoch": 0.5521640826873385, "grad_norm": 145.1573028564453, "learning_rate": 4.4783591731266155e-06, "loss": 1.1682, "step": 3419 }, { "epoch": 0.5523255813953488, "grad_norm": 45.15314483642578, "learning_rate": 4.476744186046512e-06, "loss": 0.4927, "step": 3420 }, { "epoch": 0.5524870801033591, "grad_norm": 25.18459701538086, "learning_rate": 4.475129198966409e-06, "loss": 0.0918, "step": 3421 }, { "epoch": 0.5526485788113695, "grad_norm": 159.40660095214844, "learning_rate": 4.473514211886305e-06, "loss": 2.5469, "step": 3422 }, { "epoch": 0.5528100775193798, "grad_norm": 108.60037994384766, "learning_rate": 4.471899224806202e-06, "loss": 0.9885, "step": 3423 }, { "epoch": 0.5529715762273901, "grad_norm": 227.83685302734375, "learning_rate": 4.470284237726098e-06, "loss": 3.0469, "step": 3424 }, { "epoch": 0.5531330749354005, "grad_norm": 111.4163589477539, "learning_rate": 4.468669250645995e-06, "loss": 1.1185, "step": 3425 }, { "epoch": 0.5532945736434108, "grad_norm": 90.115478515625, "learning_rate": 4.467054263565892e-06, "loss": 0.382, "step": 3426 }, { "epoch": 0.5534560723514211, "grad_norm": 37.95322036743164, "learning_rate": 4.465439276485789e-06, "loss": 0.325, "step": 3427 }, { "epoch": 0.5536175710594315, "grad_norm": 54.50164794921875, "learning_rate": 4.463824289405685e-06, "loss": 0.3711, "step": 3428 }, { "epoch": 0.5537790697674418, "grad_norm": 110.09818267822266, "learning_rate": 4.462209302325582e-06, "loss": 1.3099, "step": 3429 }, { "epoch": 0.5539405684754521, "grad_norm": 120.131103515625, "learning_rate": 4.4605943152454785e-06, "loss": 1.1562, "step": 3430 }, { "epoch": 0.5541020671834626, "grad_norm": 38.928707122802734, "learning_rate": 4.458979328165375e-06, "loss": 0.2256, "step": 3431 }, { "epoch": 0.5542635658914729, "grad_norm": 108.06352996826172, "learning_rate": 4.457364341085272e-06, "loss": 0.7983, "step": 3432 }, { "epoch": 0.5544250645994832, "grad_norm": 110.16450500488281, "learning_rate": 4.455749354005168e-06, "loss": 0.8545, "step": 3433 }, { "epoch": 0.5545865633074936, "grad_norm": 18.19721221923828, "learning_rate": 4.454134366925065e-06, "loss": 0.0701, "step": 3434 }, { "epoch": 0.5547480620155039, "grad_norm": 96.33490753173828, "learning_rate": 4.452519379844961e-06, "loss": 1.7031, "step": 3435 }, { "epoch": 0.5549095607235142, "grad_norm": 94.19821166992188, "learning_rate": 4.450904392764858e-06, "loss": 0.9637, "step": 3436 }, { "epoch": 0.5550710594315246, "grad_norm": 84.15779113769531, "learning_rate": 4.4492894056847545e-06, "loss": 0.5288, "step": 3437 }, { "epoch": 0.5552325581395349, "grad_norm": 27.786380767822266, "learning_rate": 4.447674418604652e-06, "loss": 0.1643, "step": 3438 }, { "epoch": 0.5553940568475452, "grad_norm": 34.837100982666016, "learning_rate": 4.4460594315245484e-06, "loss": 0.1633, "step": 3439 }, { "epoch": 0.5555555555555556, "grad_norm": 6.493173122406006, "learning_rate": 4.444444444444444e-06, "loss": 0.0406, "step": 3440 }, { "epoch": 0.5557170542635659, "grad_norm": 91.83944702148438, "learning_rate": 4.4428294573643416e-06, "loss": 1.1414, "step": 3441 }, { "epoch": 0.5558785529715762, "grad_norm": 80.22872924804688, "learning_rate": 4.441214470284238e-06, "loss": 0.6035, "step": 3442 }, { "epoch": 0.5560400516795866, "grad_norm": 45.17259979248047, "learning_rate": 4.439599483204135e-06, "loss": 0.2142, "step": 3443 }, { "epoch": 0.5562015503875969, "grad_norm": 63.509193420410156, "learning_rate": 4.437984496124031e-06, "loss": 0.6003, "step": 3444 }, { "epoch": 0.5563630490956072, "grad_norm": 153.1083526611328, "learning_rate": 4.436369509043928e-06, "loss": 1.9197, "step": 3445 }, { "epoch": 0.5565245478036176, "grad_norm": 6.753037929534912, "learning_rate": 4.434754521963824e-06, "loss": 0.0317, "step": 3446 }, { "epoch": 0.5566860465116279, "grad_norm": 63.83512496948242, "learning_rate": 4.433139534883722e-06, "loss": 0.6753, "step": 3447 }, { "epoch": 0.5568475452196382, "grad_norm": 71.00296020507812, "learning_rate": 4.4315245478036175e-06, "loss": 0.6309, "step": 3448 }, { "epoch": 0.5570090439276486, "grad_norm": 55.41192626953125, "learning_rate": 4.429909560723514e-06, "loss": 0.5225, "step": 3449 }, { "epoch": 0.5571705426356589, "grad_norm": 60.91825485229492, "learning_rate": 4.4282945736434114e-06, "loss": 0.582, "step": 3450 }, { "epoch": 0.5573320413436692, "grad_norm": 131.72344970703125, "learning_rate": 4.426679586563308e-06, "loss": 1.3984, "step": 3451 }, { "epoch": 0.5574935400516796, "grad_norm": 81.45997619628906, "learning_rate": 4.4250645994832046e-06, "loss": 0.6438, "step": 3452 }, { "epoch": 0.5576550387596899, "grad_norm": 56.323360443115234, "learning_rate": 4.423449612403101e-06, "loss": 0.2969, "step": 3453 }, { "epoch": 0.5578165374677002, "grad_norm": 37.99010467529297, "learning_rate": 4.421834625322998e-06, "loss": 0.3203, "step": 3454 }, { "epoch": 0.5579780361757106, "grad_norm": 140.5020751953125, "learning_rate": 4.420219638242895e-06, "loss": 1.7246, "step": 3455 }, { "epoch": 0.5581395348837209, "grad_norm": 33.87836837768555, "learning_rate": 4.418604651162791e-06, "loss": 0.1968, "step": 3456 }, { "epoch": 0.5583010335917312, "grad_norm": 28.220848083496094, "learning_rate": 4.416989664082687e-06, "loss": 0.2724, "step": 3457 }, { "epoch": 0.5584625322997416, "grad_norm": 76.46290588378906, "learning_rate": 4.415374677002585e-06, "loss": 0.7852, "step": 3458 }, { "epoch": 0.5586240310077519, "grad_norm": 57.87002182006836, "learning_rate": 4.413759689922481e-06, "loss": 0.3422, "step": 3459 }, { "epoch": 0.5587855297157622, "grad_norm": 112.09387969970703, "learning_rate": 4.412144702842378e-06, "loss": 0.9529, "step": 3460 }, { "epoch": 0.5589470284237726, "grad_norm": 52.305572509765625, "learning_rate": 4.4105297157622745e-06, "loss": 0.2666, "step": 3461 }, { "epoch": 0.5591085271317829, "grad_norm": 3.7542598247528076, "learning_rate": 4.408914728682171e-06, "loss": 0.0209, "step": 3462 }, { "epoch": 0.5592700258397932, "grad_norm": 25.473033905029297, "learning_rate": 4.4072997416020676e-06, "loss": 0.1836, "step": 3463 }, { "epoch": 0.5594315245478036, "grad_norm": 100.34222412109375, "learning_rate": 4.405684754521964e-06, "loss": 1.0048, "step": 3464 }, { "epoch": 0.559593023255814, "grad_norm": 46.31234359741211, "learning_rate": 4.404069767441861e-06, "loss": 0.13, "step": 3465 }, { "epoch": 0.5597545219638242, "grad_norm": 45.2562141418457, "learning_rate": 4.402454780361757e-06, "loss": 0.5469, "step": 3466 }, { "epoch": 0.5599160206718347, "grad_norm": 94.882568359375, "learning_rate": 4.400839793281655e-06, "loss": 0.8965, "step": 3467 }, { "epoch": 0.560077519379845, "grad_norm": 62.143959045410156, "learning_rate": 4.39922480620155e-06, "loss": 0.4331, "step": 3468 }, { "epoch": 0.5602390180878553, "grad_norm": 58.40436553955078, "learning_rate": 4.397609819121447e-06, "loss": 0.2856, "step": 3469 }, { "epoch": 0.5604005167958657, "grad_norm": 90.04257202148438, "learning_rate": 4.395994832041344e-06, "loss": 0.4518, "step": 3470 }, { "epoch": 0.560562015503876, "grad_norm": 52.388240814208984, "learning_rate": 4.394379844961241e-06, "loss": 0.2534, "step": 3471 }, { "epoch": 0.5607235142118863, "grad_norm": 17.691822052001953, "learning_rate": 4.3927648578811375e-06, "loss": 0.0419, "step": 3472 }, { "epoch": 0.5608850129198967, "grad_norm": 172.499755859375, "learning_rate": 4.391149870801034e-06, "loss": 1.066, "step": 3473 }, { "epoch": 0.561046511627907, "grad_norm": 174.7133026123047, "learning_rate": 4.389534883720931e-06, "loss": 3.1641, "step": 3474 }, { "epoch": 0.5612080103359173, "grad_norm": 45.5996208190918, "learning_rate": 4.387919896640827e-06, "loss": 0.2744, "step": 3475 }, { "epoch": 0.5613695090439277, "grad_norm": 28.269817352294922, "learning_rate": 4.386304909560724e-06, "loss": 0.1271, "step": 3476 }, { "epoch": 0.561531007751938, "grad_norm": 5.732726573944092, "learning_rate": 4.38468992248062e-06, "loss": 0.0256, "step": 3477 }, { "epoch": 0.5616925064599483, "grad_norm": 63.40727615356445, "learning_rate": 4.383074935400517e-06, "loss": 0.6524, "step": 3478 }, { "epoch": 0.5618540051679587, "grad_norm": 11.566207885742188, "learning_rate": 4.381459948320414e-06, "loss": 0.0807, "step": 3479 }, { "epoch": 0.562015503875969, "grad_norm": 74.70191955566406, "learning_rate": 4.379844961240311e-06, "loss": 1.3779, "step": 3480 }, { "epoch": 0.5621770025839793, "grad_norm": 20.250215530395508, "learning_rate": 4.3782299741602065e-06, "loss": 0.0579, "step": 3481 }, { "epoch": 0.5623385012919897, "grad_norm": 10.989693641662598, "learning_rate": 4.376614987080104e-06, "loss": 0.0319, "step": 3482 }, { "epoch": 0.5625, "grad_norm": 6.219689846038818, "learning_rate": 4.3750000000000005e-06, "loss": 0.0466, "step": 3483 }, { "epoch": 0.5626614987080103, "grad_norm": 74.85267639160156, "learning_rate": 4.373385012919897e-06, "loss": 0.5879, "step": 3484 }, { "epoch": 0.5628229974160207, "grad_norm": 81.99671173095703, "learning_rate": 4.371770025839794e-06, "loss": 1.4307, "step": 3485 }, { "epoch": 0.562984496124031, "grad_norm": 23.280752182006836, "learning_rate": 4.37015503875969e-06, "loss": 0.1016, "step": 3486 }, { "epoch": 0.5631459948320413, "grad_norm": 41.52714920043945, "learning_rate": 4.368540051679587e-06, "loss": 0.2302, "step": 3487 }, { "epoch": 0.5633074935400517, "grad_norm": 89.4996109008789, "learning_rate": 4.366925064599484e-06, "loss": 0.2711, "step": 3488 }, { "epoch": 0.563468992248062, "grad_norm": 53.06291198730469, "learning_rate": 4.36531007751938e-06, "loss": 0.3428, "step": 3489 }, { "epoch": 0.5636304909560723, "grad_norm": 8.069273948669434, "learning_rate": 4.363695090439276e-06, "loss": 0.0228, "step": 3490 }, { "epoch": 0.5637919896640827, "grad_norm": 12.107958793640137, "learning_rate": 4.362080103359174e-06, "loss": 0.0806, "step": 3491 }, { "epoch": 0.563953488372093, "grad_norm": 17.079660415649414, "learning_rate": 4.36046511627907e-06, "loss": 0.2275, "step": 3492 }, { "epoch": 0.5641149870801033, "grad_norm": 44.012062072753906, "learning_rate": 4.358850129198967e-06, "loss": 0.595, "step": 3493 }, { "epoch": 0.5642764857881137, "grad_norm": 52.271728515625, "learning_rate": 4.3572351421188635e-06, "loss": 0.2915, "step": 3494 }, { "epoch": 0.564437984496124, "grad_norm": 59.360538482666016, "learning_rate": 4.35562015503876e-06, "loss": 0.3804, "step": 3495 }, { "epoch": 0.5645994832041343, "grad_norm": 144.15789794921875, "learning_rate": 4.354005167958657e-06, "loss": 1.0488, "step": 3496 }, { "epoch": 0.5647609819121447, "grad_norm": 205.38072204589844, "learning_rate": 4.352390180878553e-06, "loss": 2.3175, "step": 3497 }, { "epoch": 0.564922480620155, "grad_norm": 35.74675369262695, "learning_rate": 4.35077519379845e-06, "loss": 0.1619, "step": 3498 }, { "epoch": 0.5650839793281653, "grad_norm": 111.92705535888672, "learning_rate": 4.349160206718346e-06, "loss": 1.0151, "step": 3499 }, { "epoch": 0.5652454780361758, "grad_norm": 30.44804573059082, "learning_rate": 4.347545219638244e-06, "loss": 0.1448, "step": 3500 }, { "epoch": 0.565406976744186, "grad_norm": 53.425376892089844, "learning_rate": 4.34593023255814e-06, "loss": 0.3406, "step": 3501 }, { "epoch": 0.5655684754521964, "grad_norm": 125.66132354736328, "learning_rate": 4.344315245478036e-06, "loss": 1.2383, "step": 3502 }, { "epoch": 0.5657299741602068, "grad_norm": 73.50841522216797, "learning_rate": 4.342700258397933e-06, "loss": 0.3672, "step": 3503 }, { "epoch": 0.5658914728682171, "grad_norm": 36.42827224731445, "learning_rate": 4.34108527131783e-06, "loss": 0.2212, "step": 3504 }, { "epoch": 0.5660529715762274, "grad_norm": 8.972089767456055, "learning_rate": 4.3394702842377265e-06, "loss": 0.0188, "step": 3505 }, { "epoch": 0.5662144702842378, "grad_norm": 101.65872192382812, "learning_rate": 4.337855297157623e-06, "loss": 1.3206, "step": 3506 }, { "epoch": 0.5663759689922481, "grad_norm": 22.878173828125, "learning_rate": 4.33624031007752e-06, "loss": 0.1445, "step": 3507 }, { "epoch": 0.5665374677002584, "grad_norm": 93.19624328613281, "learning_rate": 4.334625322997416e-06, "loss": 1.1128, "step": 3508 }, { "epoch": 0.5666989664082688, "grad_norm": 45.891441345214844, "learning_rate": 4.333010335917313e-06, "loss": 0.4141, "step": 3509 }, { "epoch": 0.5668604651162791, "grad_norm": 81.15577697753906, "learning_rate": 4.331395348837209e-06, "loss": 0.7461, "step": 3510 }, { "epoch": 0.5670219638242894, "grad_norm": 18.031494140625, "learning_rate": 4.329780361757106e-06, "loss": 0.0519, "step": 3511 }, { "epoch": 0.5671834625322998, "grad_norm": 55.224735260009766, "learning_rate": 4.328165374677003e-06, "loss": 0.3528, "step": 3512 }, { "epoch": 0.5673449612403101, "grad_norm": 72.41143035888672, "learning_rate": 4.3265503875969e-06, "loss": 0.4843, "step": 3513 }, { "epoch": 0.5675064599483204, "grad_norm": 59.9742546081543, "learning_rate": 4.324935400516796e-06, "loss": 0.3657, "step": 3514 }, { "epoch": 0.5676679586563308, "grad_norm": 79.61295318603516, "learning_rate": 4.323320413436693e-06, "loss": 0.4014, "step": 3515 }, { "epoch": 0.5678294573643411, "grad_norm": 77.90160369873047, "learning_rate": 4.3217054263565895e-06, "loss": 0.5294, "step": 3516 }, { "epoch": 0.5679909560723514, "grad_norm": 106.06012725830078, "learning_rate": 4.320090439276486e-06, "loss": 1.293, "step": 3517 }, { "epoch": 0.5681524547803618, "grad_norm": 271.3138732910156, "learning_rate": 4.318475452196383e-06, "loss": 3.8047, "step": 3518 }, { "epoch": 0.5683139534883721, "grad_norm": 23.76103973388672, "learning_rate": 4.316860465116279e-06, "loss": 0.074, "step": 3519 }, { "epoch": 0.5684754521963824, "grad_norm": 56.557228088378906, "learning_rate": 4.3152454780361766e-06, "loss": 0.4346, "step": 3520 }, { "epoch": 0.5686369509043928, "grad_norm": 28.01567268371582, "learning_rate": 4.313630490956073e-06, "loss": 0.0832, "step": 3521 }, { "epoch": 0.5687984496124031, "grad_norm": 143.83786010742188, "learning_rate": 4.312015503875969e-06, "loss": 1.2319, "step": 3522 }, { "epoch": 0.5689599483204134, "grad_norm": 54.46408462524414, "learning_rate": 4.310400516795866e-06, "loss": 0.5986, "step": 3523 }, { "epoch": 0.5691214470284238, "grad_norm": 232.71453857421875, "learning_rate": 4.308785529715763e-06, "loss": 3.0566, "step": 3524 }, { "epoch": 0.5692829457364341, "grad_norm": 46.17728042602539, "learning_rate": 4.307170542635659e-06, "loss": 0.2427, "step": 3525 }, { "epoch": 0.5694444444444444, "grad_norm": 114.66761779785156, "learning_rate": 4.305555555555556e-06, "loss": 1.0854, "step": 3526 }, { "epoch": 0.5696059431524548, "grad_norm": 126.3352279663086, "learning_rate": 4.3039405684754525e-06, "loss": 1.0273, "step": 3527 }, { "epoch": 0.5697674418604651, "grad_norm": 8.26089859008789, "learning_rate": 4.302325581395349e-06, "loss": 0.0449, "step": 3528 }, { "epoch": 0.5699289405684754, "grad_norm": 183.63665771484375, "learning_rate": 4.3007105943152465e-06, "loss": 3.8281, "step": 3529 }, { "epoch": 0.5700904392764858, "grad_norm": 87.56319427490234, "learning_rate": 4.299095607235142e-06, "loss": 0.604, "step": 3530 }, { "epoch": 0.5702519379844961, "grad_norm": 29.71417236328125, "learning_rate": 4.297480620155039e-06, "loss": 0.1558, "step": 3531 }, { "epoch": 0.5704134366925064, "grad_norm": 264.7254333496094, "learning_rate": 4.295865633074936e-06, "loss": 3.0156, "step": 3532 }, { "epoch": 0.5705749354005168, "grad_norm": 6.663449764251709, "learning_rate": 4.294250645994833e-06, "loss": 0.0135, "step": 3533 }, { "epoch": 0.5707364341085271, "grad_norm": 191.12490844726562, "learning_rate": 4.292635658914729e-06, "loss": 2.3164, "step": 3534 }, { "epoch": 0.5708979328165374, "grad_norm": 27.67169189453125, "learning_rate": 4.291020671834626e-06, "loss": 0.1842, "step": 3535 }, { "epoch": 0.5710594315245479, "grad_norm": 166.17161560058594, "learning_rate": 4.289405684754522e-06, "loss": 1.3232, "step": 3536 }, { "epoch": 0.5712209302325582, "grad_norm": 27.88707160949707, "learning_rate": 4.287790697674419e-06, "loss": 0.158, "step": 3537 }, { "epoch": 0.5713824289405685, "grad_norm": 88.8681411743164, "learning_rate": 4.2861757105943155e-06, "loss": 0.5762, "step": 3538 }, { "epoch": 0.5715439276485789, "grad_norm": 26.175668716430664, "learning_rate": 4.284560723514212e-06, "loss": 0.1582, "step": 3539 }, { "epoch": 0.5717054263565892, "grad_norm": 14.221001625061035, "learning_rate": 4.282945736434109e-06, "loss": 0.0676, "step": 3540 }, { "epoch": 0.5718669250645995, "grad_norm": 116.20561981201172, "learning_rate": 4.281330749354006e-06, "loss": 1.6035, "step": 3541 }, { "epoch": 0.5720284237726099, "grad_norm": 30.108556747436523, "learning_rate": 4.279715762273903e-06, "loss": 0.1404, "step": 3542 }, { "epoch": 0.5721899224806202, "grad_norm": 46.02999496459961, "learning_rate": 4.278100775193798e-06, "loss": 0.2998, "step": 3543 }, { "epoch": 0.5723514211886305, "grad_norm": 42.66814422607422, "learning_rate": 4.276485788113696e-06, "loss": 0.5107, "step": 3544 }, { "epoch": 0.5725129198966409, "grad_norm": 9.018303871154785, "learning_rate": 4.274870801033592e-06, "loss": 0.0293, "step": 3545 }, { "epoch": 0.5726744186046512, "grad_norm": 237.8684844970703, "learning_rate": 4.273255813953489e-06, "loss": 2.4175, "step": 3546 }, { "epoch": 0.5728359173126615, "grad_norm": 13.527363777160645, "learning_rate": 4.271640826873385e-06, "loss": 0.0774, "step": 3547 }, { "epoch": 0.5729974160206718, "grad_norm": 162.14956665039062, "learning_rate": 4.270025839793282e-06, "loss": 0.7422, "step": 3548 }, { "epoch": 0.5731589147286822, "grad_norm": 47.00349044799805, "learning_rate": 4.2684108527131785e-06, "loss": 0.2588, "step": 3549 }, { "epoch": 0.5733204134366925, "grad_norm": 28.78444480895996, "learning_rate": 4.266795865633075e-06, "loss": 0.3047, "step": 3550 }, { "epoch": 0.5734819121447028, "grad_norm": 64.32102966308594, "learning_rate": 4.265180878552972e-06, "loss": 0.5166, "step": 3551 }, { "epoch": 0.5736434108527132, "grad_norm": 34.00917434692383, "learning_rate": 4.263565891472868e-06, "loss": 0.1566, "step": 3552 }, { "epoch": 0.5738049095607235, "grad_norm": 171.07765197753906, "learning_rate": 4.261950904392766e-06, "loss": 1.3877, "step": 3553 }, { "epoch": 0.5739664082687338, "grad_norm": 42.092159271240234, "learning_rate": 4.260335917312662e-06, "loss": 0.2305, "step": 3554 }, { "epoch": 0.5741279069767442, "grad_norm": 95.8816909790039, "learning_rate": 4.258720930232558e-06, "loss": 0.6411, "step": 3555 }, { "epoch": 0.5742894056847545, "grad_norm": 27.30411148071289, "learning_rate": 4.257105943152455e-06, "loss": 0.1179, "step": 3556 }, { "epoch": 0.5744509043927648, "grad_norm": 274.2612609863281, "learning_rate": 4.255490956072352e-06, "loss": 2.3672, "step": 3557 }, { "epoch": 0.5746124031007752, "grad_norm": 58.824214935302734, "learning_rate": 4.253875968992248e-06, "loss": 1.0662, "step": 3558 }, { "epoch": 0.5747739018087855, "grad_norm": 85.97480773925781, "learning_rate": 4.252260981912145e-06, "loss": 0.5898, "step": 3559 }, { "epoch": 0.5749354005167958, "grad_norm": 62.41580581665039, "learning_rate": 4.2506459948320415e-06, "loss": 0.8857, "step": 3560 }, { "epoch": 0.5750968992248062, "grad_norm": 12.958106994628906, "learning_rate": 4.249031007751938e-06, "loss": 0.0553, "step": 3561 }, { "epoch": 0.5752583979328165, "grad_norm": 179.70883178710938, "learning_rate": 4.2474160206718355e-06, "loss": 1.9375, "step": 3562 }, { "epoch": 0.5754198966408268, "grad_norm": 82.39899444580078, "learning_rate": 4.245801033591731e-06, "loss": 0.3257, "step": 3563 }, { "epoch": 0.5755813953488372, "grad_norm": 77.59927368164062, "learning_rate": 4.244186046511628e-06, "loss": 0.439, "step": 3564 }, { "epoch": 0.5757428940568475, "grad_norm": 59.232017517089844, "learning_rate": 4.242571059431525e-06, "loss": 0.6766, "step": 3565 }, { "epoch": 0.5759043927648578, "grad_norm": 16.32615852355957, "learning_rate": 4.240956072351422e-06, "loss": 0.0371, "step": 3566 }, { "epoch": 0.5760658914728682, "grad_norm": 164.08660888671875, "learning_rate": 4.239341085271318e-06, "loss": 0.9427, "step": 3567 }, { "epoch": 0.5762273901808785, "grad_norm": 38.29515838623047, "learning_rate": 4.237726098191215e-06, "loss": 0.1816, "step": 3568 }, { "epoch": 0.5763888888888888, "grad_norm": 51.59413146972656, "learning_rate": 4.236111111111111e-06, "loss": 0.752, "step": 3569 }, { "epoch": 0.5765503875968992, "grad_norm": 272.0995178222656, "learning_rate": 4.234496124031008e-06, "loss": 0.5967, "step": 3570 }, { "epoch": 0.5767118863049095, "grad_norm": 135.67579650878906, "learning_rate": 4.2328811369509045e-06, "loss": 1.4727, "step": 3571 }, { "epoch": 0.5768733850129198, "grad_norm": 8.11788558959961, "learning_rate": 4.231266149870801e-06, "loss": 0.0342, "step": 3572 }, { "epoch": 0.5770348837209303, "grad_norm": 40.8809814453125, "learning_rate": 4.229651162790698e-06, "loss": 0.4087, "step": 3573 }, { "epoch": 0.5771963824289406, "grad_norm": 6.083901405334473, "learning_rate": 4.228036175710595e-06, "loss": 0.0318, "step": 3574 }, { "epoch": 0.5773578811369509, "grad_norm": 41.833648681640625, "learning_rate": 4.226421188630492e-06, "loss": 0.2065, "step": 3575 }, { "epoch": 0.5775193798449613, "grad_norm": 87.1464614868164, "learning_rate": 4.224806201550387e-06, "loss": 1.1249, "step": 3576 }, { "epoch": 0.5776808785529716, "grad_norm": 49.14838790893555, "learning_rate": 4.223191214470285e-06, "loss": 0.3921, "step": 3577 }, { "epoch": 0.5778423772609819, "grad_norm": 7.708846092224121, "learning_rate": 4.221576227390181e-06, "loss": 0.0565, "step": 3578 }, { "epoch": 0.5780038759689923, "grad_norm": 47.20134735107422, "learning_rate": 4.219961240310078e-06, "loss": 0.3164, "step": 3579 }, { "epoch": 0.5781653746770026, "grad_norm": 7.7031779289245605, "learning_rate": 4.218346253229974e-06, "loss": 0.0329, "step": 3580 }, { "epoch": 0.5783268733850129, "grad_norm": 90.06170654296875, "learning_rate": 4.216731266149871e-06, "loss": 0.6045, "step": 3581 }, { "epoch": 0.5784883720930233, "grad_norm": 24.707752227783203, "learning_rate": 4.215116279069768e-06, "loss": 0.1162, "step": 3582 }, { "epoch": 0.5786498708010336, "grad_norm": 111.80924224853516, "learning_rate": 4.213501291989665e-06, "loss": 0.7959, "step": 3583 }, { "epoch": 0.5788113695090439, "grad_norm": 49.499759674072266, "learning_rate": 4.211886304909561e-06, "loss": 0.3691, "step": 3584 }, { "epoch": 0.5789728682170543, "grad_norm": 40.57290267944336, "learning_rate": 4.210271317829458e-06, "loss": 0.137, "step": 3585 }, { "epoch": 0.5791343669250646, "grad_norm": 83.45358276367188, "learning_rate": 4.208656330749355e-06, "loss": 1.0381, "step": 3586 }, { "epoch": 0.5792958656330749, "grad_norm": 19.092552185058594, "learning_rate": 4.207041343669251e-06, "loss": 0.0768, "step": 3587 }, { "epoch": 0.5794573643410853, "grad_norm": 64.5089111328125, "learning_rate": 4.205426356589148e-06, "loss": 0.5127, "step": 3588 }, { "epoch": 0.5796188630490956, "grad_norm": 17.420141220092773, "learning_rate": 4.203811369509044e-06, "loss": 0.1383, "step": 3589 }, { "epoch": 0.5797803617571059, "grad_norm": 75.27214813232422, "learning_rate": 4.202196382428941e-06, "loss": 0.8782, "step": 3590 }, { "epoch": 0.5799418604651163, "grad_norm": 8.168603897094727, "learning_rate": 4.200581395348837e-06, "loss": 0.0248, "step": 3591 }, { "epoch": 0.5801033591731266, "grad_norm": 134.6168975830078, "learning_rate": 4.198966408268734e-06, "loss": 0.9707, "step": 3592 }, { "epoch": 0.5802648578811369, "grad_norm": 127.55052185058594, "learning_rate": 4.1973514211886305e-06, "loss": 2.8691, "step": 3593 }, { "epoch": 0.5804263565891473, "grad_norm": 19.826419830322266, "learning_rate": 4.195736434108528e-06, "loss": 0.0636, "step": 3594 }, { "epoch": 0.5805878552971576, "grad_norm": 5.812010765075684, "learning_rate": 4.1941214470284245e-06, "loss": 0.0339, "step": 3595 }, { "epoch": 0.5807493540051679, "grad_norm": 7.210159778594971, "learning_rate": 4.19250645994832e-06, "loss": 0.0468, "step": 3596 }, { "epoch": 0.5809108527131783, "grad_norm": 49.724483489990234, "learning_rate": 4.190891472868218e-06, "loss": 0.7871, "step": 3597 }, { "epoch": 0.5810723514211886, "grad_norm": 25.44877052307129, "learning_rate": 4.189276485788114e-06, "loss": 0.0889, "step": 3598 }, { "epoch": 0.5812338501291989, "grad_norm": 245.6687469482422, "learning_rate": 4.187661498708011e-06, "loss": 1.7934, "step": 3599 }, { "epoch": 0.5813953488372093, "grad_norm": 43.539302825927734, "learning_rate": 4.186046511627907e-06, "loss": 0.2048, "step": 3600 }, { "epoch": 0.5815568475452196, "grad_norm": 24.94220542907715, "learning_rate": 4.184431524547804e-06, "loss": 0.1934, "step": 3601 }, { "epoch": 0.5817183462532299, "grad_norm": 68.84149169921875, "learning_rate": 4.1828165374677e-06, "loss": 0.4609, "step": 3602 }, { "epoch": 0.5818798449612403, "grad_norm": 31.561864852905273, "learning_rate": 4.181201550387598e-06, "loss": 0.2991, "step": 3603 }, { "epoch": 0.5820413436692506, "grad_norm": 88.60340881347656, "learning_rate": 4.1795865633074935e-06, "loss": 1.351, "step": 3604 }, { "epoch": 0.5822028423772609, "grad_norm": 25.468738555908203, "learning_rate": 4.17797157622739e-06, "loss": 0.0634, "step": 3605 }, { "epoch": 0.5823643410852714, "grad_norm": 301.46636962890625, "learning_rate": 4.1763565891472875e-06, "loss": 1.5625, "step": 3606 }, { "epoch": 0.5825258397932817, "grad_norm": 51.45703887939453, "learning_rate": 4.174741602067184e-06, "loss": 0.3616, "step": 3607 }, { "epoch": 0.582687338501292, "grad_norm": 62.11928176879883, "learning_rate": 4.173126614987081e-06, "loss": 0.5912, "step": 3608 }, { "epoch": 0.5828488372093024, "grad_norm": 22.286291122436523, "learning_rate": 4.171511627906977e-06, "loss": 0.0653, "step": 3609 }, { "epoch": 0.5830103359173127, "grad_norm": 136.10665893554688, "learning_rate": 4.169896640826874e-06, "loss": 0.8625, "step": 3610 }, { "epoch": 0.583171834625323, "grad_norm": 67.49446105957031, "learning_rate": 4.16828165374677e-06, "loss": 1.3623, "step": 3611 }, { "epoch": 0.5833333333333334, "grad_norm": 79.45510864257812, "learning_rate": 4.166666666666667e-06, "loss": 1.2134, "step": 3612 }, { "epoch": 0.5834948320413437, "grad_norm": 169.30064392089844, "learning_rate": 4.1650516795865634e-06, "loss": 1.2834, "step": 3613 }, { "epoch": 0.583656330749354, "grad_norm": 151.99655151367188, "learning_rate": 4.16343669250646e-06, "loss": 0.5126, "step": 3614 }, { "epoch": 0.5838178294573644, "grad_norm": 25.873340606689453, "learning_rate": 4.161821705426357e-06, "loss": 0.1873, "step": 3615 }, { "epoch": 0.5839793281653747, "grad_norm": 72.14543914794922, "learning_rate": 4.160206718346254e-06, "loss": 0.6681, "step": 3616 }, { "epoch": 0.584140826873385, "grad_norm": 183.09942626953125, "learning_rate": 4.15859173126615e-06, "loss": 2.1152, "step": 3617 }, { "epoch": 0.5843023255813954, "grad_norm": 52.68362045288086, "learning_rate": 4.156976744186047e-06, "loss": 0.8545, "step": 3618 }, { "epoch": 0.5844638242894057, "grad_norm": 53.78880310058594, "learning_rate": 4.155361757105944e-06, "loss": 0.7528, "step": 3619 }, { "epoch": 0.584625322997416, "grad_norm": 74.61196899414062, "learning_rate": 4.15374677002584e-06, "loss": 1.3032, "step": 3620 }, { "epoch": 0.5847868217054264, "grad_norm": 9.4596586227417, "learning_rate": 4.152131782945737e-06, "loss": 0.0266, "step": 3621 }, { "epoch": 0.5849483204134367, "grad_norm": 112.80204772949219, "learning_rate": 4.150516795865633e-06, "loss": 0.8223, "step": 3622 }, { "epoch": 0.585109819121447, "grad_norm": 90.36541748046875, "learning_rate": 4.14890180878553e-06, "loss": 1.0336, "step": 3623 }, { "epoch": 0.5852713178294574, "grad_norm": 35.338497161865234, "learning_rate": 4.1472868217054264e-06, "loss": 0.1807, "step": 3624 }, { "epoch": 0.5854328165374677, "grad_norm": 68.54187774658203, "learning_rate": 4.145671834625323e-06, "loss": 0.7129, "step": 3625 }, { "epoch": 0.585594315245478, "grad_norm": 52.184478759765625, "learning_rate": 4.1440568475452196e-06, "loss": 0.4741, "step": 3626 }, { "epoch": 0.5857558139534884, "grad_norm": 39.123329162597656, "learning_rate": 4.142441860465117e-06, "loss": 0.3921, "step": 3627 }, { "epoch": 0.5859173126614987, "grad_norm": 155.32064819335938, "learning_rate": 4.1408268733850135e-06, "loss": 2.25, "step": 3628 }, { "epoch": 0.586078811369509, "grad_norm": 79.74119567871094, "learning_rate": 4.13921188630491e-06, "loss": 0.8711, "step": 3629 }, { "epoch": 0.5862403100775194, "grad_norm": 22.14918327331543, "learning_rate": 4.137596899224807e-06, "loss": 0.0618, "step": 3630 }, { "epoch": 0.5864018087855297, "grad_norm": 91.99484252929688, "learning_rate": 4.135981912144703e-06, "loss": 0.3577, "step": 3631 }, { "epoch": 0.58656330749354, "grad_norm": 159.45751953125, "learning_rate": 4.1343669250646e-06, "loss": 0.9429, "step": 3632 }, { "epoch": 0.5867248062015504, "grad_norm": 73.6822280883789, "learning_rate": 4.132751937984496e-06, "loss": 0.584, "step": 3633 }, { "epoch": 0.5868863049095607, "grad_norm": 47.41387176513672, "learning_rate": 4.131136950904393e-06, "loss": 0.353, "step": 3634 }, { "epoch": 0.587047803617571, "grad_norm": 117.94673156738281, "learning_rate": 4.1295219638242894e-06, "loss": 1.1407, "step": 3635 }, { "epoch": 0.5872093023255814, "grad_norm": 57.61438751220703, "learning_rate": 4.127906976744187e-06, "loss": 0.3323, "step": 3636 }, { "epoch": 0.5873708010335917, "grad_norm": 55.711917877197266, "learning_rate": 4.1262919896640826e-06, "loss": 0.5205, "step": 3637 }, { "epoch": 0.587532299741602, "grad_norm": 59.37240219116211, "learning_rate": 4.124677002583979e-06, "loss": 0.5728, "step": 3638 }, { "epoch": 0.5876937984496124, "grad_norm": 62.49786376953125, "learning_rate": 4.1230620155038765e-06, "loss": 0.335, "step": 3639 }, { "epoch": 0.5878552971576227, "grad_norm": 133.3717803955078, "learning_rate": 4.121447028423773e-06, "loss": 1.0879, "step": 3640 }, { "epoch": 0.588016795865633, "grad_norm": 39.5881233215332, "learning_rate": 4.11983204134367e-06, "loss": 0.3099, "step": 3641 }, { "epoch": 0.5881782945736435, "grad_norm": 6.87169885635376, "learning_rate": 4.118217054263566e-06, "loss": 0.0397, "step": 3642 }, { "epoch": 0.5883397932816538, "grad_norm": 7.2825493812561035, "learning_rate": 4.116602067183463e-06, "loss": 0.0505, "step": 3643 }, { "epoch": 0.588501291989664, "grad_norm": 99.3143539428711, "learning_rate": 4.114987080103359e-06, "loss": 1.1201, "step": 3644 }, { "epoch": 0.5886627906976745, "grad_norm": 18.852479934692383, "learning_rate": 4.113372093023256e-06, "loss": 0.0947, "step": 3645 }, { "epoch": 0.5888242894056848, "grad_norm": 28.505382537841797, "learning_rate": 4.1117571059431524e-06, "loss": 0.2567, "step": 3646 }, { "epoch": 0.5889857881136951, "grad_norm": 29.84832000732422, "learning_rate": 4.11014211886305e-06, "loss": 0.1616, "step": 3647 }, { "epoch": 0.5891472868217055, "grad_norm": 40.46078872680664, "learning_rate": 4.108527131782946e-06, "loss": 0.1206, "step": 3648 }, { "epoch": 0.5893087855297158, "grad_norm": 112.89787292480469, "learning_rate": 4.106912144702843e-06, "loss": 0.7949, "step": 3649 }, { "epoch": 0.5894702842377261, "grad_norm": 39.876094818115234, "learning_rate": 4.1052971576227395e-06, "loss": 0.3396, "step": 3650 }, { "epoch": 0.5896317829457365, "grad_norm": 13.156928062438965, "learning_rate": 4.103682170542636e-06, "loss": 0.0751, "step": 3651 }, { "epoch": 0.5897932816537468, "grad_norm": 232.07015991210938, "learning_rate": 4.102067183462533e-06, "loss": 2.4763, "step": 3652 }, { "epoch": 0.5899547803617571, "grad_norm": 62.63560104370117, "learning_rate": 4.100452196382429e-06, "loss": 0.6785, "step": 3653 }, { "epoch": 0.5901162790697675, "grad_norm": 288.0491027832031, "learning_rate": 4.098837209302326e-06, "loss": 2.1777, "step": 3654 }, { "epoch": 0.5902777777777778, "grad_norm": 93.59358978271484, "learning_rate": 4.097222222222222e-06, "loss": 1.1084, "step": 3655 }, { "epoch": 0.5904392764857881, "grad_norm": 26.445816040039062, "learning_rate": 4.09560723514212e-06, "loss": 0.1787, "step": 3656 }, { "epoch": 0.5906007751937985, "grad_norm": 232.35423278808594, "learning_rate": 4.093992248062016e-06, "loss": 2.0742, "step": 3657 }, { "epoch": 0.5907622739018088, "grad_norm": 85.7178955078125, "learning_rate": 4.092377260981912e-06, "loss": 0.9355, "step": 3658 }, { "epoch": 0.5909237726098191, "grad_norm": 174.11622619628906, "learning_rate": 4.090762273901809e-06, "loss": 2.9495, "step": 3659 }, { "epoch": 0.5910852713178295, "grad_norm": 78.30540466308594, "learning_rate": 4.089147286821706e-06, "loss": 1.0696, "step": 3660 }, { "epoch": 0.5912467700258398, "grad_norm": 43.30282211303711, "learning_rate": 4.0875322997416025e-06, "loss": 0.2671, "step": 3661 }, { "epoch": 0.5914082687338501, "grad_norm": 13.04956340789795, "learning_rate": 4.085917312661499e-06, "loss": 0.0327, "step": 3662 }, { "epoch": 0.5915697674418605, "grad_norm": 45.423622131347656, "learning_rate": 4.084302325581396e-06, "loss": 0.2268, "step": 3663 }, { "epoch": 0.5917312661498708, "grad_norm": 527.6284790039062, "learning_rate": 4.082687338501292e-06, "loss": 2.6035, "step": 3664 }, { "epoch": 0.5918927648578811, "grad_norm": 188.45245361328125, "learning_rate": 4.081072351421189e-06, "loss": 2.082, "step": 3665 }, { "epoch": 0.5920542635658915, "grad_norm": 137.1211700439453, "learning_rate": 4.079457364341085e-06, "loss": 1.2461, "step": 3666 }, { "epoch": 0.5922157622739018, "grad_norm": 113.75174713134766, "learning_rate": 4.077842377260982e-06, "loss": 1.1293, "step": 3667 }, { "epoch": 0.5923772609819121, "grad_norm": 107.96170043945312, "learning_rate": 4.076227390180879e-06, "loss": 1.3047, "step": 3668 }, { "epoch": 0.5925387596899225, "grad_norm": 26.944597244262695, "learning_rate": 4.074612403100776e-06, "loss": 0.3007, "step": 3669 }, { "epoch": 0.5927002583979328, "grad_norm": 13.726917266845703, "learning_rate": 4.072997416020672e-06, "loss": 0.0992, "step": 3670 }, { "epoch": 0.5928617571059431, "grad_norm": 99.25923156738281, "learning_rate": 4.071382428940569e-06, "loss": 0.532, "step": 3671 }, { "epoch": 0.5930232558139535, "grad_norm": 57.156986236572266, "learning_rate": 4.0697674418604655e-06, "loss": 0.3856, "step": 3672 }, { "epoch": 0.5931847545219638, "grad_norm": 132.2756805419922, "learning_rate": 4.068152454780362e-06, "loss": 1.0825, "step": 3673 }, { "epoch": 0.5933462532299741, "grad_norm": 58.769371032714844, "learning_rate": 4.066537467700259e-06, "loss": 0.5098, "step": 3674 }, { "epoch": 0.5935077519379846, "grad_norm": 44.3773193359375, "learning_rate": 4.064922480620155e-06, "loss": 0.2185, "step": 3675 }, { "epoch": 0.5936692506459949, "grad_norm": 92.07686614990234, "learning_rate": 4.063307493540052e-06, "loss": 0.9746, "step": 3676 }, { "epoch": 0.5938307493540051, "grad_norm": 221.8712615966797, "learning_rate": 4.061692506459949e-06, "loss": 1.4805, "step": 3677 }, { "epoch": 0.5939922480620154, "grad_norm": 16.709932327270508, "learning_rate": 4.060077519379845e-06, "loss": 0.0594, "step": 3678 }, { "epoch": 0.5941537467700259, "grad_norm": 89.1037368774414, "learning_rate": 4.0584625322997415e-06, "loss": 0.75, "step": 3679 }, { "epoch": 0.5943152454780362, "grad_norm": 82.308349609375, "learning_rate": 4.056847545219639e-06, "loss": 0.949, "step": 3680 }, { "epoch": 0.5944767441860465, "grad_norm": 43.63930130004883, "learning_rate": 4.0552325581395354e-06, "loss": 0.4041, "step": 3681 }, { "epoch": 0.5946382428940569, "grad_norm": 21.151731491088867, "learning_rate": 4.053617571059432e-06, "loss": 0.0991, "step": 3682 }, { "epoch": 0.5947997416020672, "grad_norm": 5.309004306793213, "learning_rate": 4.0520025839793286e-06, "loss": 0.037, "step": 3683 }, { "epoch": 0.5949612403100775, "grad_norm": 22.3919620513916, "learning_rate": 4.050387596899225e-06, "loss": 0.1204, "step": 3684 }, { "epoch": 0.5951227390180879, "grad_norm": 71.88214111328125, "learning_rate": 4.048772609819122e-06, "loss": 0.7546, "step": 3685 }, { "epoch": 0.5952842377260982, "grad_norm": 70.18851470947266, "learning_rate": 4.047157622739018e-06, "loss": 0.5186, "step": 3686 }, { "epoch": 0.5954457364341085, "grad_norm": 16.69082260131836, "learning_rate": 4.045542635658915e-06, "loss": 0.1107, "step": 3687 }, { "epoch": 0.5956072351421189, "grad_norm": 74.24593353271484, "learning_rate": 4.043927648578811e-06, "loss": 0.6162, "step": 3688 }, { "epoch": 0.5957687338501292, "grad_norm": 22.91489028930664, "learning_rate": 4.042312661498709e-06, "loss": 0.0806, "step": 3689 }, { "epoch": 0.5959302325581395, "grad_norm": 73.07402038574219, "learning_rate": 4.040697674418605e-06, "loss": 0.2441, "step": 3690 }, { "epoch": 0.5960917312661499, "grad_norm": 94.20911407470703, "learning_rate": 4.039082687338501e-06, "loss": 0.6823, "step": 3691 }, { "epoch": 0.5962532299741602, "grad_norm": 78.80321502685547, "learning_rate": 4.0374677002583984e-06, "loss": 0.6592, "step": 3692 }, { "epoch": 0.5964147286821705, "grad_norm": 4.786534786224365, "learning_rate": 4.035852713178295e-06, "loss": 0.0302, "step": 3693 }, { "epoch": 0.5965762273901809, "grad_norm": 44.688941955566406, "learning_rate": 4.0342377260981916e-06, "loss": 0.2794, "step": 3694 }, { "epoch": 0.5967377260981912, "grad_norm": 77.40940856933594, "learning_rate": 4.032622739018088e-06, "loss": 0.709, "step": 3695 }, { "epoch": 0.5968992248062015, "grad_norm": 7.566969871520996, "learning_rate": 4.031007751937985e-06, "loss": 0.0219, "step": 3696 }, { "epoch": 0.5970607235142119, "grad_norm": 49.386234283447266, "learning_rate": 4.029392764857881e-06, "loss": 0.2559, "step": 3697 }, { "epoch": 0.5972222222222222, "grad_norm": 269.1136474609375, "learning_rate": 4.027777777777779e-06, "loss": 0.9907, "step": 3698 }, { "epoch": 0.5973837209302325, "grad_norm": 67.67627716064453, "learning_rate": 4.026162790697674e-06, "loss": 0.4375, "step": 3699 }, { "epoch": 0.5975452196382429, "grad_norm": 27.159259796142578, "learning_rate": 4.024547803617571e-06, "loss": 0.1235, "step": 3700 }, { "epoch": 0.5977067183462532, "grad_norm": 51.589962005615234, "learning_rate": 4.022932816537468e-06, "loss": 0.193, "step": 3701 }, { "epoch": 0.5978682170542635, "grad_norm": 165.777587890625, "learning_rate": 4.021317829457365e-06, "loss": 0.7483, "step": 3702 }, { "epoch": 0.5980297157622739, "grad_norm": 48.09364700317383, "learning_rate": 4.0197028423772614e-06, "loss": 0.2894, "step": 3703 }, { "epoch": 0.5981912144702842, "grad_norm": 124.79264068603516, "learning_rate": 4.018087855297158e-06, "loss": 0.8959, "step": 3704 }, { "epoch": 0.5983527131782945, "grad_norm": 55.34177780151367, "learning_rate": 4.0164728682170546e-06, "loss": 0.3249, "step": 3705 }, { "epoch": 0.5985142118863049, "grad_norm": 20.955852508544922, "learning_rate": 4.014857881136951e-06, "loss": 0.0927, "step": 3706 }, { "epoch": 0.5986757105943152, "grad_norm": 33.41099166870117, "learning_rate": 4.013242894056848e-06, "loss": 0.1223, "step": 3707 }, { "epoch": 0.5988372093023255, "grad_norm": 28.888200759887695, "learning_rate": 4.011627906976744e-06, "loss": 0.2349, "step": 3708 }, { "epoch": 0.598998708010336, "grad_norm": 23.68356704711914, "learning_rate": 4.010012919896642e-06, "loss": 0.0859, "step": 3709 }, { "epoch": 0.5991602067183462, "grad_norm": 96.87800598144531, "learning_rate": 4.008397932816538e-06, "loss": 0.5599, "step": 3710 }, { "epoch": 0.5993217054263565, "grad_norm": 23.314559936523438, "learning_rate": 4.006782945736434e-06, "loss": 0.134, "step": 3711 }, { "epoch": 0.599483204134367, "grad_norm": 323.1916809082031, "learning_rate": 4.005167958656331e-06, "loss": 2.9355, "step": 3712 }, { "epoch": 0.5996447028423773, "grad_norm": 188.50315856933594, "learning_rate": 4.003552971576228e-06, "loss": 1.6846, "step": 3713 }, { "epoch": 0.5998062015503876, "grad_norm": 29.092201232910156, "learning_rate": 4.0019379844961245e-06, "loss": 0.0981, "step": 3714 }, { "epoch": 0.599967700258398, "grad_norm": 48.60337829589844, "learning_rate": 4.000322997416021e-06, "loss": 0.2388, "step": 3715 }, { "epoch": 0.6001291989664083, "grad_norm": 25.5988712310791, "learning_rate": 3.9987080103359176e-06, "loss": 0.1174, "step": 3716 }, { "epoch": 0.6002906976744186, "grad_norm": 101.56603240966797, "learning_rate": 3.997093023255814e-06, "loss": 0.438, "step": 3717 }, { "epoch": 0.600452196382429, "grad_norm": 85.47632598876953, "learning_rate": 3.9954780361757115e-06, "loss": 0.5283, "step": 3718 }, { "epoch": 0.6006136950904393, "grad_norm": 37.83102035522461, "learning_rate": 3.993863049095607e-06, "loss": 0.2119, "step": 3719 }, { "epoch": 0.6007751937984496, "grad_norm": 84.66171264648438, "learning_rate": 3.992248062015504e-06, "loss": 0.6621, "step": 3720 }, { "epoch": 0.60093669250646, "grad_norm": 27.5589542388916, "learning_rate": 3.990633074935401e-06, "loss": 0.0925, "step": 3721 }, { "epoch": 0.6010981912144703, "grad_norm": 55.878623962402344, "learning_rate": 3.989018087855298e-06, "loss": 0.4826, "step": 3722 }, { "epoch": 0.6012596899224806, "grad_norm": 42.78911590576172, "learning_rate": 3.987403100775194e-06, "loss": 0.3606, "step": 3723 }, { "epoch": 0.601421188630491, "grad_norm": 23.06690788269043, "learning_rate": 3.985788113695091e-06, "loss": 0.0807, "step": 3724 }, { "epoch": 0.6015826873385013, "grad_norm": 36.2921257019043, "learning_rate": 3.9841731266149875e-06, "loss": 0.1394, "step": 3725 }, { "epoch": 0.6017441860465116, "grad_norm": 192.49417114257812, "learning_rate": 3.982558139534884e-06, "loss": 1.6445, "step": 3726 }, { "epoch": 0.601905684754522, "grad_norm": 57.58213806152344, "learning_rate": 3.980943152454781e-06, "loss": 0.27, "step": 3727 }, { "epoch": 0.6020671834625323, "grad_norm": 10.848184585571289, "learning_rate": 3.979328165374677e-06, "loss": 0.0779, "step": 3728 }, { "epoch": 0.6022286821705426, "grad_norm": 44.33887481689453, "learning_rate": 3.977713178294574e-06, "loss": 0.2202, "step": 3729 }, { "epoch": 0.602390180878553, "grad_norm": 67.87969970703125, "learning_rate": 3.976098191214471e-06, "loss": 0.9961, "step": 3730 }, { "epoch": 0.6025516795865633, "grad_norm": 27.02927589416504, "learning_rate": 3.974483204134368e-06, "loss": 0.2031, "step": 3731 }, { "epoch": 0.6027131782945736, "grad_norm": 14.668182373046875, "learning_rate": 3.972868217054263e-06, "loss": 0.0337, "step": 3732 }, { "epoch": 0.602874677002584, "grad_norm": 31.44286346435547, "learning_rate": 3.971253229974161e-06, "loss": 0.0924, "step": 3733 }, { "epoch": 0.6030361757105943, "grad_norm": 24.853254318237305, "learning_rate": 3.969638242894057e-06, "loss": 0.1233, "step": 3734 }, { "epoch": 0.6031976744186046, "grad_norm": 48.661407470703125, "learning_rate": 3.968023255813954e-06, "loss": 0.2515, "step": 3735 }, { "epoch": 0.603359173126615, "grad_norm": 28.954147338867188, "learning_rate": 3.9664082687338505e-06, "loss": 0.123, "step": 3736 }, { "epoch": 0.6035206718346253, "grad_norm": 18.825294494628906, "learning_rate": 3.964793281653747e-06, "loss": 0.0728, "step": 3737 }, { "epoch": 0.6036821705426356, "grad_norm": 106.48941040039062, "learning_rate": 3.963178294573644e-06, "loss": 0.623, "step": 3738 }, { "epoch": 0.603843669250646, "grad_norm": 13.996826171875, "learning_rate": 3.96156330749354e-06, "loss": 0.1318, "step": 3739 }, { "epoch": 0.6040051679586563, "grad_norm": 79.6875991821289, "learning_rate": 3.959948320413437e-06, "loss": 1.0918, "step": 3740 }, { "epoch": 0.6041666666666666, "grad_norm": 96.93777465820312, "learning_rate": 3.958333333333333e-06, "loss": 1.6582, "step": 3741 }, { "epoch": 0.604328165374677, "grad_norm": 32.058170318603516, "learning_rate": 3.956718346253231e-06, "loss": 0.1741, "step": 3742 }, { "epoch": 0.6044896640826873, "grad_norm": 151.6114044189453, "learning_rate": 3.955103359173127e-06, "loss": 0.7911, "step": 3743 }, { "epoch": 0.6046511627906976, "grad_norm": 117.32527923583984, "learning_rate": 3.953488372093024e-06, "loss": 1.0684, "step": 3744 }, { "epoch": 0.604812661498708, "grad_norm": 195.44473266601562, "learning_rate": 3.95187338501292e-06, "loss": 3.3945, "step": 3745 }, { "epoch": 0.6049741602067183, "grad_norm": 62.94993209838867, "learning_rate": 3.950258397932817e-06, "loss": 0.9668, "step": 3746 }, { "epoch": 0.6051356589147286, "grad_norm": 5.679162979125977, "learning_rate": 3.9486434108527135e-06, "loss": 0.0275, "step": 3747 }, { "epoch": 0.6052971576227391, "grad_norm": 24.431659698486328, "learning_rate": 3.94702842377261e-06, "loss": 0.1978, "step": 3748 }, { "epoch": 0.6054586563307494, "grad_norm": 154.14427185058594, "learning_rate": 3.945413436692507e-06, "loss": 0.8281, "step": 3749 }, { "epoch": 0.6056201550387597, "grad_norm": 15.159942626953125, "learning_rate": 3.943798449612403e-06, "loss": 0.0944, "step": 3750 }, { "epoch": 0.6057816537467701, "grad_norm": 52.136234283447266, "learning_rate": 3.9421834625323006e-06, "loss": 0.2463, "step": 3751 }, { "epoch": 0.6059431524547804, "grad_norm": 68.10514068603516, "learning_rate": 3.940568475452196e-06, "loss": 0.3037, "step": 3752 }, { "epoch": 0.6061046511627907, "grad_norm": 95.24928283691406, "learning_rate": 3.938953488372093e-06, "loss": 1.123, "step": 3753 }, { "epoch": 0.6062661498708011, "grad_norm": 62.17962646484375, "learning_rate": 3.93733850129199e-06, "loss": 0.2039, "step": 3754 }, { "epoch": 0.6064276485788114, "grad_norm": 26.85736846923828, "learning_rate": 3.935723514211887e-06, "loss": 0.16, "step": 3755 }, { "epoch": 0.6065891472868217, "grad_norm": 54.92530822753906, "learning_rate": 3.934108527131783e-06, "loss": 0.218, "step": 3756 }, { "epoch": 0.6067506459948321, "grad_norm": 149.6497039794922, "learning_rate": 3.93249354005168e-06, "loss": 1.1348, "step": 3757 }, { "epoch": 0.6069121447028424, "grad_norm": 15.189355850219727, "learning_rate": 3.9308785529715765e-06, "loss": 0.0496, "step": 3758 }, { "epoch": 0.6070736434108527, "grad_norm": 18.877653121948242, "learning_rate": 3.929263565891473e-06, "loss": 0.0676, "step": 3759 }, { "epoch": 0.6072351421188631, "grad_norm": 60.07136535644531, "learning_rate": 3.92764857881137e-06, "loss": 0.1812, "step": 3760 }, { "epoch": 0.6073966408268734, "grad_norm": 45.475746154785156, "learning_rate": 3.926033591731266e-06, "loss": 0.3828, "step": 3761 }, { "epoch": 0.6075581395348837, "grad_norm": 25.627592086791992, "learning_rate": 3.924418604651163e-06, "loss": 0.1104, "step": 3762 }, { "epoch": 0.6077196382428941, "grad_norm": 95.85137939453125, "learning_rate": 3.92280361757106e-06, "loss": 0.7207, "step": 3763 }, { "epoch": 0.6078811369509044, "grad_norm": 115.45411682128906, "learning_rate": 3.921188630490957e-06, "loss": 0.4374, "step": 3764 }, { "epoch": 0.6080426356589147, "grad_norm": 51.254966735839844, "learning_rate": 3.919573643410852e-06, "loss": 0.2366, "step": 3765 }, { "epoch": 0.6082041343669251, "grad_norm": 151.52932739257812, "learning_rate": 3.91795865633075e-06, "loss": 1.1768, "step": 3766 }, { "epoch": 0.6083656330749354, "grad_norm": 29.99671745300293, "learning_rate": 3.916343669250646e-06, "loss": 0.1406, "step": 3767 }, { "epoch": 0.6085271317829457, "grad_norm": 19.333101272583008, "learning_rate": 3.914728682170543e-06, "loss": 0.0919, "step": 3768 }, { "epoch": 0.6086886304909561, "grad_norm": 226.72299194335938, "learning_rate": 3.9131136950904395e-06, "loss": 3.2476, "step": 3769 }, { "epoch": 0.6088501291989664, "grad_norm": 29.290794372558594, "learning_rate": 3.911498708010336e-06, "loss": 0.1449, "step": 3770 }, { "epoch": 0.6090116279069767, "grad_norm": 30.943758010864258, "learning_rate": 3.909883720930233e-06, "loss": 0.1797, "step": 3771 }, { "epoch": 0.6091731266149871, "grad_norm": 110.11788940429688, "learning_rate": 3.90826873385013e-06, "loss": 1.7017, "step": 3772 }, { "epoch": 0.6093346253229974, "grad_norm": 14.431543350219727, "learning_rate": 3.906653746770026e-06, "loss": 0.0257, "step": 3773 }, { "epoch": 0.6094961240310077, "grad_norm": 54.00700759887695, "learning_rate": 3.905038759689923e-06, "loss": 0.3779, "step": 3774 }, { "epoch": 0.6096576227390181, "grad_norm": 51.905094146728516, "learning_rate": 3.90342377260982e-06, "loss": 0.3389, "step": 3775 }, { "epoch": 0.6098191214470284, "grad_norm": 91.5354995727539, "learning_rate": 3.901808785529716e-06, "loss": 0.7402, "step": 3776 }, { "epoch": 0.6099806201550387, "grad_norm": 83.16058349609375, "learning_rate": 3.900193798449613e-06, "loss": 0.5062, "step": 3777 }, { "epoch": 0.6101421188630491, "grad_norm": 163.1096649169922, "learning_rate": 3.898578811369509e-06, "loss": 2.0391, "step": 3778 }, { "epoch": 0.6103036175710594, "grad_norm": 28.36846160888672, "learning_rate": 3.896963824289406e-06, "loss": 0.0664, "step": 3779 }, { "epoch": 0.6104651162790697, "grad_norm": 132.49942016601562, "learning_rate": 3.8953488372093025e-06, "loss": 0.9609, "step": 3780 }, { "epoch": 0.6106266149870802, "grad_norm": 125.95689392089844, "learning_rate": 3.893733850129199e-06, "loss": 0.6812, "step": 3781 }, { "epoch": 0.6107881136950905, "grad_norm": 40.73847198486328, "learning_rate": 3.892118863049096e-06, "loss": 0.1692, "step": 3782 }, { "epoch": 0.6109496124031008, "grad_norm": 19.074947357177734, "learning_rate": 3.890503875968993e-06, "loss": 0.0994, "step": 3783 }, { "epoch": 0.6111111111111112, "grad_norm": 50.976173400878906, "learning_rate": 3.88888888888889e-06, "loss": 0.1696, "step": 3784 }, { "epoch": 0.6112726098191215, "grad_norm": 210.4456787109375, "learning_rate": 3.887273901808786e-06, "loss": 1.2427, "step": 3785 }, { "epoch": 0.6114341085271318, "grad_norm": 44.570919036865234, "learning_rate": 3.885658914728683e-06, "loss": 0.1689, "step": 3786 }, { "epoch": 0.6115956072351422, "grad_norm": 66.7671890258789, "learning_rate": 3.884043927648579e-06, "loss": 0.2505, "step": 3787 }, { "epoch": 0.6117571059431525, "grad_norm": 241.4447784423828, "learning_rate": 3.882428940568476e-06, "loss": 3.3515, "step": 3788 }, { "epoch": 0.6119186046511628, "grad_norm": 15.975622177124023, "learning_rate": 3.880813953488372e-06, "loss": 0.0526, "step": 3789 }, { "epoch": 0.6120801033591732, "grad_norm": 119.1011962890625, "learning_rate": 3.879198966408269e-06, "loss": 1.0865, "step": 3790 }, { "epoch": 0.6122416020671835, "grad_norm": 20.90439224243164, "learning_rate": 3.8775839793281655e-06, "loss": 0.1367, "step": 3791 }, { "epoch": 0.6124031007751938, "grad_norm": 68.41986083984375, "learning_rate": 3.875968992248063e-06, "loss": 0.41, "step": 3792 }, { "epoch": 0.6125645994832042, "grad_norm": 102.14627075195312, "learning_rate": 3.874354005167959e-06, "loss": 1.6126, "step": 3793 }, { "epoch": 0.6127260981912145, "grad_norm": 79.72941589355469, "learning_rate": 3.872739018087855e-06, "loss": 0.51, "step": 3794 }, { "epoch": 0.6128875968992248, "grad_norm": 15.727285385131836, "learning_rate": 3.871124031007753e-06, "loss": 0.0881, "step": 3795 }, { "epoch": 0.6130490956072352, "grad_norm": 43.69721603393555, "learning_rate": 3.869509043927649e-06, "loss": 0.3574, "step": 3796 }, { "epoch": 0.6132105943152455, "grad_norm": 134.50985717773438, "learning_rate": 3.867894056847546e-06, "loss": 1.0465, "step": 3797 }, { "epoch": 0.6133720930232558, "grad_norm": 99.63007354736328, "learning_rate": 3.866279069767442e-06, "loss": 1.6348, "step": 3798 }, { "epoch": 0.6135335917312662, "grad_norm": 47.47903060913086, "learning_rate": 3.864664082687339e-06, "loss": 0.2571, "step": 3799 }, { "epoch": 0.6136950904392765, "grad_norm": 50.436153411865234, "learning_rate": 3.863049095607235e-06, "loss": 0.3505, "step": 3800 }, { "epoch": 0.6138565891472868, "grad_norm": 9.860955238342285, "learning_rate": 3.861434108527132e-06, "loss": 0.0481, "step": 3801 }, { "epoch": 0.6140180878552972, "grad_norm": 129.80982971191406, "learning_rate": 3.8598191214470285e-06, "loss": 0.9062, "step": 3802 }, { "epoch": 0.6141795865633075, "grad_norm": 72.97607421875, "learning_rate": 3.858204134366925e-06, "loss": 0.3975, "step": 3803 }, { "epoch": 0.6143410852713178, "grad_norm": 129.48922729492188, "learning_rate": 3.8565891472868225e-06, "loss": 0.8511, "step": 3804 }, { "epoch": 0.6145025839793282, "grad_norm": 61.35108947753906, "learning_rate": 3.854974160206719e-06, "loss": 0.4924, "step": 3805 }, { "epoch": 0.6146640826873385, "grad_norm": 23.328506469726562, "learning_rate": 3.853359173126615e-06, "loss": 0.1436, "step": 3806 }, { "epoch": 0.6148255813953488, "grad_norm": 19.82598304748535, "learning_rate": 3.851744186046512e-06, "loss": 0.1831, "step": 3807 }, { "epoch": 0.6149870801033591, "grad_norm": 17.316848754882812, "learning_rate": 3.850129198966409e-06, "loss": 0.0698, "step": 3808 }, { "epoch": 0.6151485788113695, "grad_norm": 14.599233627319336, "learning_rate": 3.848514211886305e-06, "loss": 0.1287, "step": 3809 }, { "epoch": 0.6153100775193798, "grad_norm": 48.88003921508789, "learning_rate": 3.846899224806202e-06, "loss": 0.2988, "step": 3810 }, { "epoch": 0.6154715762273901, "grad_norm": 80.54173278808594, "learning_rate": 3.845284237726098e-06, "loss": 0.8442, "step": 3811 }, { "epoch": 0.6156330749354005, "grad_norm": 52.817569732666016, "learning_rate": 3.843669250645995e-06, "loss": 0.4316, "step": 3812 }, { "epoch": 0.6157945736434108, "grad_norm": 82.0170669555664, "learning_rate": 3.842054263565892e-06, "loss": 0.6985, "step": 3813 }, { "epoch": 0.6159560723514211, "grad_norm": 16.207761764526367, "learning_rate": 3.840439276485788e-06, "loss": 0.0796, "step": 3814 }, { "epoch": 0.6161175710594315, "grad_norm": 47.77348709106445, "learning_rate": 3.838824289405685e-06, "loss": 0.259, "step": 3815 }, { "epoch": 0.6162790697674418, "grad_norm": 64.80634307861328, "learning_rate": 3.837209302325582e-06, "loss": 0.3521, "step": 3816 }, { "epoch": 0.6164405684754521, "grad_norm": 46.9179801940918, "learning_rate": 3.835594315245479e-06, "loss": 0.2443, "step": 3817 }, { "epoch": 0.6166020671834626, "grad_norm": 137.60513305664062, "learning_rate": 3.833979328165375e-06, "loss": 1.1914, "step": 3818 }, { "epoch": 0.6167635658914729, "grad_norm": 51.73308563232422, "learning_rate": 3.832364341085272e-06, "loss": 0.7361, "step": 3819 }, { "epoch": 0.6169250645994832, "grad_norm": 37.940643310546875, "learning_rate": 3.830749354005168e-06, "loss": 0.1929, "step": 3820 }, { "epoch": 0.6170865633074936, "grad_norm": 55.351654052734375, "learning_rate": 3.829134366925065e-06, "loss": 0.1604, "step": 3821 }, { "epoch": 0.6172480620155039, "grad_norm": 19.681690216064453, "learning_rate": 3.827519379844961e-06, "loss": 0.0613, "step": 3822 }, { "epoch": 0.6174095607235142, "grad_norm": 24.662641525268555, "learning_rate": 3.825904392764858e-06, "loss": 0.1674, "step": 3823 }, { "epoch": 0.6175710594315246, "grad_norm": 8.845675468444824, "learning_rate": 3.8242894056847545e-06, "loss": 0.076, "step": 3824 }, { "epoch": 0.6177325581395349, "grad_norm": 148.3900146484375, "learning_rate": 3.822674418604652e-06, "loss": 0.793, "step": 3825 }, { "epoch": 0.6178940568475452, "grad_norm": 84.08940887451172, "learning_rate": 3.821059431524548e-06, "loss": 0.6005, "step": 3826 }, { "epoch": 0.6180555555555556, "grad_norm": 15.158218383789062, "learning_rate": 3.819444444444444e-06, "loss": 0.0602, "step": 3827 }, { "epoch": 0.6182170542635659, "grad_norm": 188.44508361816406, "learning_rate": 3.817829457364342e-06, "loss": 1.3134, "step": 3828 }, { "epoch": 0.6183785529715762, "grad_norm": 82.73466491699219, "learning_rate": 3.816214470284238e-06, "loss": 0.8828, "step": 3829 }, { "epoch": 0.6185400516795866, "grad_norm": 252.54701232910156, "learning_rate": 3.8145994832041343e-06, "loss": 3.1543, "step": 3830 }, { "epoch": 0.6187015503875969, "grad_norm": 102.06999969482422, "learning_rate": 3.8129844961240313e-06, "loss": 0.7417, "step": 3831 }, { "epoch": 0.6188630490956072, "grad_norm": 125.34610748291016, "learning_rate": 3.811369509043928e-06, "loss": 1.832, "step": 3832 }, { "epoch": 0.6190245478036176, "grad_norm": 136.7899932861328, "learning_rate": 3.8097545219638244e-06, "loss": 0.609, "step": 3833 }, { "epoch": 0.6191860465116279, "grad_norm": 45.90282440185547, "learning_rate": 3.8081395348837214e-06, "loss": 0.1914, "step": 3834 }, { "epoch": 0.6193475452196382, "grad_norm": 52.281063079833984, "learning_rate": 3.806524547803618e-06, "loss": 0.376, "step": 3835 }, { "epoch": 0.6195090439276486, "grad_norm": 142.592529296875, "learning_rate": 3.804909560723514e-06, "loss": 3.3003, "step": 3836 }, { "epoch": 0.6196705426356589, "grad_norm": 13.220987319946289, "learning_rate": 3.803294573643411e-06, "loss": 0.0388, "step": 3837 }, { "epoch": 0.6198320413436692, "grad_norm": 29.280302047729492, "learning_rate": 3.8016795865633076e-06, "loss": 0.1567, "step": 3838 }, { "epoch": 0.6199935400516796, "grad_norm": 31.457565307617188, "learning_rate": 3.8000645994832046e-06, "loss": 0.066, "step": 3839 }, { "epoch": 0.6201550387596899, "grad_norm": 15.595020294189453, "learning_rate": 3.798449612403101e-06, "loss": 0.0534, "step": 3840 }, { "epoch": 0.6203165374677002, "grad_norm": 148.37400817871094, "learning_rate": 3.7968346253229977e-06, "loss": 0.7032, "step": 3841 }, { "epoch": 0.6204780361757106, "grad_norm": 25.690956115722656, "learning_rate": 3.7952196382428947e-06, "loss": 0.0726, "step": 3842 }, { "epoch": 0.6206395348837209, "grad_norm": 11.986808776855469, "learning_rate": 3.7936046511627913e-06, "loss": 0.0359, "step": 3843 }, { "epoch": 0.6208010335917312, "grad_norm": 56.16572570800781, "learning_rate": 3.7919896640826874e-06, "loss": 0.2518, "step": 3844 }, { "epoch": 0.6209625322997416, "grad_norm": 65.65409851074219, "learning_rate": 3.7903746770025844e-06, "loss": 0.5244, "step": 3845 }, { "epoch": 0.6211240310077519, "grad_norm": 53.06861877441406, "learning_rate": 3.788759689922481e-06, "loss": 0.4908, "step": 3846 }, { "epoch": 0.6212855297157622, "grad_norm": 39.645530700683594, "learning_rate": 3.7871447028423775e-06, "loss": 0.2456, "step": 3847 }, { "epoch": 0.6214470284237726, "grad_norm": 22.828218460083008, "learning_rate": 3.7855297157622745e-06, "loss": 0.2216, "step": 3848 }, { "epoch": 0.6216085271317829, "grad_norm": 39.7410888671875, "learning_rate": 3.783914728682171e-06, "loss": 0.1824, "step": 3849 }, { "epoch": 0.6217700258397932, "grad_norm": 27.42284393310547, "learning_rate": 3.782299741602067e-06, "loss": 0.115, "step": 3850 }, { "epoch": 0.6219315245478036, "grad_norm": 84.97794342041016, "learning_rate": 3.780684754521964e-06, "loss": 1.0215, "step": 3851 }, { "epoch": 0.622093023255814, "grad_norm": 36.80080795288086, "learning_rate": 3.7790697674418607e-06, "loss": 0.0994, "step": 3852 }, { "epoch": 0.6222545219638242, "grad_norm": 68.02147674560547, "learning_rate": 3.7774547803617573e-06, "loss": 0.721, "step": 3853 }, { "epoch": 0.6224160206718347, "grad_norm": 18.770526885986328, "learning_rate": 3.7758397932816543e-06, "loss": 0.1035, "step": 3854 }, { "epoch": 0.622577519379845, "grad_norm": 36.30936813354492, "learning_rate": 3.774224806201551e-06, "loss": 0.1172, "step": 3855 }, { "epoch": 0.6227390180878553, "grad_norm": 63.454071044921875, "learning_rate": 3.772609819121447e-06, "loss": 0.4123, "step": 3856 }, { "epoch": 0.6229005167958657, "grad_norm": 176.5922088623047, "learning_rate": 3.7709948320413444e-06, "loss": 1.4246, "step": 3857 }, { "epoch": 0.623062015503876, "grad_norm": 15.252528190612793, "learning_rate": 3.7693798449612405e-06, "loss": 0.0735, "step": 3858 }, { "epoch": 0.6232235142118863, "grad_norm": 133.10414123535156, "learning_rate": 3.767764857881137e-06, "loss": 1.1621, "step": 3859 }, { "epoch": 0.6233850129198967, "grad_norm": 7.288905143737793, "learning_rate": 3.766149870801034e-06, "loss": 0.0215, "step": 3860 }, { "epoch": 0.623546511627907, "grad_norm": 51.20869445800781, "learning_rate": 3.7645348837209306e-06, "loss": 0.1661, "step": 3861 }, { "epoch": 0.6237080103359173, "grad_norm": 118.65644073486328, "learning_rate": 3.762919896640827e-06, "loss": 0.5102, "step": 3862 }, { "epoch": 0.6238695090439277, "grad_norm": 125.31889343261719, "learning_rate": 3.761304909560724e-06, "loss": 0.7645, "step": 3863 }, { "epoch": 0.624031007751938, "grad_norm": 64.5816650390625, "learning_rate": 3.7596899224806203e-06, "loss": 0.4854, "step": 3864 }, { "epoch": 0.6241925064599483, "grad_norm": 56.78852462768555, "learning_rate": 3.758074935400517e-06, "loss": 0.2148, "step": 3865 }, { "epoch": 0.6243540051679587, "grad_norm": 4.997383117675781, "learning_rate": 3.756459948320414e-06, "loss": 0.0356, "step": 3866 }, { "epoch": 0.624515503875969, "grad_norm": 17.626434326171875, "learning_rate": 3.7548449612403104e-06, "loss": 0.0925, "step": 3867 }, { "epoch": 0.6246770025839793, "grad_norm": 42.06584167480469, "learning_rate": 3.753229974160207e-06, "loss": 0.2119, "step": 3868 }, { "epoch": 0.6248385012919897, "grad_norm": 22.887115478515625, "learning_rate": 3.751614987080104e-06, "loss": 0.1057, "step": 3869 }, { "epoch": 0.625, "grad_norm": 77.91349792480469, "learning_rate": 3.7500000000000005e-06, "loss": 0.4847, "step": 3870 }, { "epoch": 0.6251614987080103, "grad_norm": 54.32097244262695, "learning_rate": 3.7483850129198967e-06, "loss": 0.2646, "step": 3871 }, { "epoch": 0.6253229974160207, "grad_norm": 60.152889251708984, "learning_rate": 3.7467700258397936e-06, "loss": 1.0662, "step": 3872 }, { "epoch": 0.625484496124031, "grad_norm": 78.50951385498047, "learning_rate": 3.74515503875969e-06, "loss": 0.5964, "step": 3873 }, { "epoch": 0.6256459948320413, "grad_norm": 55.76995086669922, "learning_rate": 3.7435400516795868e-06, "loss": 0.4995, "step": 3874 }, { "epoch": 0.6258074935400517, "grad_norm": 130.31997680664062, "learning_rate": 3.7419250645994837e-06, "loss": 1.1582, "step": 3875 }, { "epoch": 0.625968992248062, "grad_norm": 48.6399040222168, "learning_rate": 3.7403100775193803e-06, "loss": 0.2997, "step": 3876 }, { "epoch": 0.6261304909560723, "grad_norm": 180.32443237304688, "learning_rate": 3.7386950904392764e-06, "loss": 1.4551, "step": 3877 }, { "epoch": 0.6262919896640827, "grad_norm": 12.644453048706055, "learning_rate": 3.7370801033591734e-06, "loss": 0.0925, "step": 3878 }, { "epoch": 0.626453488372093, "grad_norm": 76.44670104980469, "learning_rate": 3.73546511627907e-06, "loss": 0.2988, "step": 3879 }, { "epoch": 0.6266149870801033, "grad_norm": 20.59844970703125, "learning_rate": 3.7338501291989665e-06, "loss": 0.1034, "step": 3880 }, { "epoch": 0.6267764857881137, "grad_norm": 29.009920120239258, "learning_rate": 3.7322351421188635e-06, "loss": 0.1134, "step": 3881 }, { "epoch": 0.626937984496124, "grad_norm": 15.811054229736328, "learning_rate": 3.73062015503876e-06, "loss": 0.0386, "step": 3882 }, { "epoch": 0.6270994832041343, "grad_norm": 73.79634094238281, "learning_rate": 3.7290051679586562e-06, "loss": 1.1792, "step": 3883 }, { "epoch": 0.6272609819121447, "grad_norm": 120.78900146484375, "learning_rate": 3.7273901808785536e-06, "loss": 0.9531, "step": 3884 }, { "epoch": 0.627422480620155, "grad_norm": 24.727197647094727, "learning_rate": 3.7257751937984498e-06, "loss": 0.1528, "step": 3885 }, { "epoch": 0.6275839793281653, "grad_norm": 15.999601364135742, "learning_rate": 3.7241602067183463e-06, "loss": 0.0662, "step": 3886 }, { "epoch": 0.6277454780361758, "grad_norm": 69.75029754638672, "learning_rate": 3.7225452196382433e-06, "loss": 0.559, "step": 3887 }, { "epoch": 0.627906976744186, "grad_norm": 30.234445571899414, "learning_rate": 3.72093023255814e-06, "loss": 0.199, "step": 3888 }, { "epoch": 0.6280684754521964, "grad_norm": 33.610443115234375, "learning_rate": 3.7193152454780364e-06, "loss": 0.1694, "step": 3889 }, { "epoch": 0.6282299741602068, "grad_norm": 78.95763397216797, "learning_rate": 3.7177002583979334e-06, "loss": 0.3594, "step": 3890 }, { "epoch": 0.6283914728682171, "grad_norm": 45.23805618286133, "learning_rate": 3.7160852713178295e-06, "loss": 0.2443, "step": 3891 }, { "epoch": 0.6285529715762274, "grad_norm": 131.04563903808594, "learning_rate": 3.714470284237726e-06, "loss": 0.8359, "step": 3892 }, { "epoch": 0.6287144702842378, "grad_norm": 7.481390953063965, "learning_rate": 3.712855297157623e-06, "loss": 0.0237, "step": 3893 }, { "epoch": 0.6288759689922481, "grad_norm": 12.176234245300293, "learning_rate": 3.7112403100775197e-06, "loss": 0.0511, "step": 3894 }, { "epoch": 0.6290374677002584, "grad_norm": 9.656280517578125, "learning_rate": 3.7096253229974162e-06, "loss": 0.0541, "step": 3895 }, { "epoch": 0.6291989664082688, "grad_norm": 76.52851867675781, "learning_rate": 3.708010335917313e-06, "loss": 0.3997, "step": 3896 }, { "epoch": 0.6293604651162791, "grad_norm": 9.346433639526367, "learning_rate": 3.7063953488372093e-06, "loss": 0.03, "step": 3897 }, { "epoch": 0.6295219638242894, "grad_norm": 78.83812713623047, "learning_rate": 3.704780361757106e-06, "loss": 0.5713, "step": 3898 }, { "epoch": 0.6296834625322998, "grad_norm": 16.59564781188965, "learning_rate": 3.703165374677003e-06, "loss": 0.0547, "step": 3899 }, { "epoch": 0.6298449612403101, "grad_norm": 18.392526626586914, "learning_rate": 3.7015503875968994e-06, "loss": 0.0796, "step": 3900 }, { "epoch": 0.6300064599483204, "grad_norm": 210.839599609375, "learning_rate": 3.699935400516796e-06, "loss": 2.8047, "step": 3901 }, { "epoch": 0.6301679586563308, "grad_norm": 130.3399200439453, "learning_rate": 3.698320413436693e-06, "loss": 0.6982, "step": 3902 }, { "epoch": 0.6303294573643411, "grad_norm": 303.5307922363281, "learning_rate": 3.6967054263565895e-06, "loss": 1.8633, "step": 3903 }, { "epoch": 0.6304909560723514, "grad_norm": 149.63272094726562, "learning_rate": 3.6950904392764865e-06, "loss": 0.9219, "step": 3904 }, { "epoch": 0.6306524547803618, "grad_norm": 4.935088634490967, "learning_rate": 3.6934754521963827e-06, "loss": 0.0212, "step": 3905 }, { "epoch": 0.6308139534883721, "grad_norm": 209.00205993652344, "learning_rate": 3.6918604651162792e-06, "loss": 1.6902, "step": 3906 }, { "epoch": 0.6309754521963824, "grad_norm": 124.33704376220703, "learning_rate": 3.690245478036176e-06, "loss": 1.2666, "step": 3907 }, { "epoch": 0.6311369509043928, "grad_norm": 48.1355094909668, "learning_rate": 3.6886304909560728e-06, "loss": 0.2192, "step": 3908 }, { "epoch": 0.6312984496124031, "grad_norm": 20.690265655517578, "learning_rate": 3.6870155038759693e-06, "loss": 0.0577, "step": 3909 }, { "epoch": 0.6314599483204134, "grad_norm": 27.40618896484375, "learning_rate": 3.6854005167958663e-06, "loss": 0.2207, "step": 3910 }, { "epoch": 0.6316214470284238, "grad_norm": 33.004005432128906, "learning_rate": 3.6837855297157624e-06, "loss": 0.1294, "step": 3911 }, { "epoch": 0.6317829457364341, "grad_norm": 115.29752349853516, "learning_rate": 3.682170542635659e-06, "loss": 0.6646, "step": 3912 }, { "epoch": 0.6319444444444444, "grad_norm": 52.49881362915039, "learning_rate": 3.680555555555556e-06, "loss": 0.2344, "step": 3913 }, { "epoch": 0.6321059431524548, "grad_norm": 49.26408767700195, "learning_rate": 3.6789405684754525e-06, "loss": 0.3228, "step": 3914 }, { "epoch": 0.6322674418604651, "grad_norm": 178.55007934570312, "learning_rate": 3.677325581395349e-06, "loss": 0.5931, "step": 3915 }, { "epoch": 0.6324289405684754, "grad_norm": 161.49232482910156, "learning_rate": 3.675710594315246e-06, "loss": 0.6763, "step": 3916 }, { "epoch": 0.6325904392764858, "grad_norm": 35.55043029785156, "learning_rate": 3.6740956072351426e-06, "loss": 0.1235, "step": 3917 }, { "epoch": 0.6327519379844961, "grad_norm": 47.52241516113281, "learning_rate": 3.6724806201550388e-06, "loss": 0.3218, "step": 3918 }, { "epoch": 0.6329134366925064, "grad_norm": 20.994098663330078, "learning_rate": 3.6708656330749358e-06, "loss": 0.1042, "step": 3919 }, { "epoch": 0.6330749354005168, "grad_norm": 5.014932632446289, "learning_rate": 3.6692506459948323e-06, "loss": 0.0304, "step": 3920 }, { "epoch": 0.6332364341085271, "grad_norm": 62.5321044921875, "learning_rate": 3.667635658914729e-06, "loss": 1.1074, "step": 3921 }, { "epoch": 0.6333979328165374, "grad_norm": 68.84416198730469, "learning_rate": 3.666020671834626e-06, "loss": 0.4238, "step": 3922 }, { "epoch": 0.6335594315245479, "grad_norm": 212.5950927734375, "learning_rate": 3.6644056847545224e-06, "loss": 0.6719, "step": 3923 }, { "epoch": 0.6337209302325582, "grad_norm": 90.4771728515625, "learning_rate": 3.6627906976744186e-06, "loss": 1.4077, "step": 3924 }, { "epoch": 0.6338824289405685, "grad_norm": 149.25926208496094, "learning_rate": 3.6611757105943156e-06, "loss": 0.4734, "step": 3925 }, { "epoch": 0.6340439276485789, "grad_norm": 96.79281616210938, "learning_rate": 3.659560723514212e-06, "loss": 0.825, "step": 3926 }, { "epoch": 0.6342054263565892, "grad_norm": 183.99412536621094, "learning_rate": 3.6579457364341087e-06, "loss": 1.541, "step": 3927 }, { "epoch": 0.6343669250645995, "grad_norm": 3.967503786087036, "learning_rate": 3.6563307493540057e-06, "loss": 0.0134, "step": 3928 }, { "epoch": 0.6345284237726099, "grad_norm": 35.911346435546875, "learning_rate": 3.6547157622739022e-06, "loss": 0.1941, "step": 3929 }, { "epoch": 0.6346899224806202, "grad_norm": 133.1095733642578, "learning_rate": 3.6531007751937988e-06, "loss": 0.8601, "step": 3930 }, { "epoch": 0.6348514211886305, "grad_norm": 138.3555145263672, "learning_rate": 3.6514857881136958e-06, "loss": 1.2949, "step": 3931 }, { "epoch": 0.6350129198966409, "grad_norm": 63.983604431152344, "learning_rate": 3.649870801033592e-06, "loss": 0.5519, "step": 3932 }, { "epoch": 0.6351744186046512, "grad_norm": 88.42891693115234, "learning_rate": 3.6482558139534885e-06, "loss": 0.5181, "step": 3933 }, { "epoch": 0.6353359173126615, "grad_norm": 138.55752563476562, "learning_rate": 3.6466408268733854e-06, "loss": 0.6523, "step": 3934 }, { "epoch": 0.6354974160206718, "grad_norm": 18.17194366455078, "learning_rate": 3.645025839793282e-06, "loss": 0.0746, "step": 3935 }, { "epoch": 0.6356589147286822, "grad_norm": 118.2284164428711, "learning_rate": 3.6434108527131786e-06, "loss": 1.0664, "step": 3936 }, { "epoch": 0.6358204134366925, "grad_norm": 255.8965606689453, "learning_rate": 3.6417958656330755e-06, "loss": 2.3184, "step": 3937 }, { "epoch": 0.6359819121447028, "grad_norm": 19.739303588867188, "learning_rate": 3.6401808785529717e-06, "loss": 0.0872, "step": 3938 }, { "epoch": 0.6361434108527132, "grad_norm": 88.56903076171875, "learning_rate": 3.6385658914728682e-06, "loss": 0.4199, "step": 3939 }, { "epoch": 0.6363049095607235, "grad_norm": 10.888155937194824, "learning_rate": 3.6369509043927652e-06, "loss": 0.0255, "step": 3940 }, { "epoch": 0.6364664082687338, "grad_norm": 84.21771240234375, "learning_rate": 3.6353359173126618e-06, "loss": 0.4392, "step": 3941 }, { "epoch": 0.6366279069767442, "grad_norm": 203.7976837158203, "learning_rate": 3.6337209302325583e-06, "loss": 2.4106, "step": 3942 }, { "epoch": 0.6367894056847545, "grad_norm": 19.0809383392334, "learning_rate": 3.6321059431524553e-06, "loss": 0.0537, "step": 3943 }, { "epoch": 0.6369509043927648, "grad_norm": 74.1391372680664, "learning_rate": 3.630490956072352e-06, "loss": 0.4705, "step": 3944 }, { "epoch": 0.6371124031007752, "grad_norm": 173.68307495117188, "learning_rate": 3.628875968992248e-06, "loss": 0.5977, "step": 3945 }, { "epoch": 0.6372739018087855, "grad_norm": 119.98329162597656, "learning_rate": 3.627260981912145e-06, "loss": 0.5505, "step": 3946 }, { "epoch": 0.6374354005167958, "grad_norm": 226.3469696044922, "learning_rate": 3.6256459948320416e-06, "loss": 1.4883, "step": 3947 }, { "epoch": 0.6375968992248062, "grad_norm": 45.735939025878906, "learning_rate": 3.624031007751938e-06, "loss": 0.1753, "step": 3948 }, { "epoch": 0.6377583979328165, "grad_norm": 24.28514289855957, "learning_rate": 3.622416020671835e-06, "loss": 0.1311, "step": 3949 }, { "epoch": 0.6379198966408268, "grad_norm": 32.338653564453125, "learning_rate": 3.6208010335917317e-06, "loss": 0.1091, "step": 3950 }, { "epoch": 0.6380813953488372, "grad_norm": 42.667808532714844, "learning_rate": 3.619186046511628e-06, "loss": 0.2264, "step": 3951 }, { "epoch": 0.6382428940568475, "grad_norm": 59.33751678466797, "learning_rate": 3.617571059431525e-06, "loss": 0.1929, "step": 3952 }, { "epoch": 0.6384043927648578, "grad_norm": 23.585525512695312, "learning_rate": 3.6159560723514214e-06, "loss": 0.1145, "step": 3953 }, { "epoch": 0.6385658914728682, "grad_norm": 77.24742126464844, "learning_rate": 3.614341085271318e-06, "loss": 0.9004, "step": 3954 }, { "epoch": 0.6387273901808785, "grad_norm": 196.90444946289062, "learning_rate": 3.612726098191215e-06, "loss": 0.6436, "step": 3955 }, { "epoch": 0.6388888888888888, "grad_norm": 40.32563781738281, "learning_rate": 3.6111111111111115e-06, "loss": 0.1562, "step": 3956 }, { "epoch": 0.6390503875968992, "grad_norm": 110.74758911132812, "learning_rate": 3.6094961240310076e-06, "loss": 1.0322, "step": 3957 }, { "epoch": 0.6392118863049095, "grad_norm": 20.171184539794922, "learning_rate": 3.607881136950905e-06, "loss": 0.0443, "step": 3958 }, { "epoch": 0.6393733850129198, "grad_norm": 7.494832515716553, "learning_rate": 3.606266149870801e-06, "loss": 0.013, "step": 3959 }, { "epoch": 0.6395348837209303, "grad_norm": 30.702531814575195, "learning_rate": 3.6046511627906977e-06, "loss": 0.0937, "step": 3960 }, { "epoch": 0.6396963824289406, "grad_norm": 67.68385314941406, "learning_rate": 3.6030361757105947e-06, "loss": 0.2596, "step": 3961 }, { "epoch": 0.6398578811369509, "grad_norm": 25.847536087036133, "learning_rate": 3.6014211886304912e-06, "loss": 0.0663, "step": 3962 }, { "epoch": 0.6400193798449613, "grad_norm": 70.3006591796875, "learning_rate": 3.599806201550388e-06, "loss": 0.2939, "step": 3963 }, { "epoch": 0.6401808785529716, "grad_norm": 29.13813591003418, "learning_rate": 3.5981912144702848e-06, "loss": 0.1206, "step": 3964 }, { "epoch": 0.6403423772609819, "grad_norm": 75.55599975585938, "learning_rate": 3.596576227390181e-06, "loss": 0.8237, "step": 3965 }, { "epoch": 0.6405038759689923, "grad_norm": 21.589969635009766, "learning_rate": 3.5949612403100775e-06, "loss": 0.1038, "step": 3966 }, { "epoch": 0.6406653746770026, "grad_norm": 78.87311553955078, "learning_rate": 3.5933462532299745e-06, "loss": 0.4087, "step": 3967 }, { "epoch": 0.6408268733850129, "grad_norm": 75.84004211425781, "learning_rate": 3.591731266149871e-06, "loss": 0.2998, "step": 3968 }, { "epoch": 0.6409883720930233, "grad_norm": 179.49530029296875, "learning_rate": 3.590116279069768e-06, "loss": 0.8701, "step": 3969 }, { "epoch": 0.6411498708010336, "grad_norm": 16.86842918395996, "learning_rate": 3.5885012919896646e-06, "loss": 0.062, "step": 3970 }, { "epoch": 0.6413113695090439, "grad_norm": 26.066864013671875, "learning_rate": 3.586886304909561e-06, "loss": 0.1126, "step": 3971 }, { "epoch": 0.6414728682170543, "grad_norm": 55.662940979003906, "learning_rate": 3.585271317829458e-06, "loss": 0.7264, "step": 3972 }, { "epoch": 0.6416343669250646, "grad_norm": 120.59918975830078, "learning_rate": 3.5836563307493542e-06, "loss": 1.2976, "step": 3973 }, { "epoch": 0.6417958656330749, "grad_norm": 5.664034843444824, "learning_rate": 3.582041343669251e-06, "loss": 0.0113, "step": 3974 }, { "epoch": 0.6419573643410853, "grad_norm": 116.4485855102539, "learning_rate": 3.5804263565891478e-06, "loss": 0.4111, "step": 3975 }, { "epoch": 0.6421188630490956, "grad_norm": 128.89971923828125, "learning_rate": 3.5788113695090443e-06, "loss": 1.0703, "step": 3976 }, { "epoch": 0.6422803617571059, "grad_norm": 51.338104248046875, "learning_rate": 3.577196382428941e-06, "loss": 0.3467, "step": 3977 }, { "epoch": 0.6424418604651163, "grad_norm": 62.95273208618164, "learning_rate": 3.575581395348838e-06, "loss": 0.8589, "step": 3978 }, { "epoch": 0.6426033591731266, "grad_norm": 65.38751220703125, "learning_rate": 3.573966408268734e-06, "loss": 0.3242, "step": 3979 }, { "epoch": 0.6427648578811369, "grad_norm": 68.52928161621094, "learning_rate": 3.5723514211886306e-06, "loss": 0.2933, "step": 3980 }, { "epoch": 0.6429263565891473, "grad_norm": 75.3767318725586, "learning_rate": 3.5707364341085276e-06, "loss": 0.4403, "step": 3981 }, { "epoch": 0.6430878552971576, "grad_norm": 30.455158233642578, "learning_rate": 3.569121447028424e-06, "loss": 0.1536, "step": 3982 }, { "epoch": 0.6432493540051679, "grad_norm": 23.001615524291992, "learning_rate": 3.5675064599483207e-06, "loss": 0.0689, "step": 3983 }, { "epoch": 0.6434108527131783, "grad_norm": 50.514183044433594, "learning_rate": 3.5658914728682177e-06, "loss": 0.4289, "step": 3984 }, { "epoch": 0.6435723514211886, "grad_norm": 17.794776916503906, "learning_rate": 3.5642764857881142e-06, "loss": 0.0662, "step": 3985 }, { "epoch": 0.6437338501291989, "grad_norm": 15.16141414642334, "learning_rate": 3.5626614987080104e-06, "loss": 0.067, "step": 3986 }, { "epoch": 0.6438953488372093, "grad_norm": 7.178042888641357, "learning_rate": 3.5610465116279074e-06, "loss": 0.0218, "step": 3987 }, { "epoch": 0.6440568475452196, "grad_norm": 128.74696350097656, "learning_rate": 3.559431524547804e-06, "loss": 2.291, "step": 3988 }, { "epoch": 0.6442183462532299, "grad_norm": 11.102397918701172, "learning_rate": 3.5578165374677005e-06, "loss": 0.0732, "step": 3989 }, { "epoch": 0.6443798449612403, "grad_norm": 43.47677230834961, "learning_rate": 3.5562015503875975e-06, "loss": 0.2445, "step": 3990 }, { "epoch": 0.6445413436692506, "grad_norm": 49.272090911865234, "learning_rate": 3.554586563307494e-06, "loss": 0.2659, "step": 3991 }, { "epoch": 0.6447028423772609, "grad_norm": 13.073284149169922, "learning_rate": 3.55297157622739e-06, "loss": 0.0324, "step": 3992 }, { "epoch": 0.6448643410852714, "grad_norm": 90.48981475830078, "learning_rate": 3.551356589147287e-06, "loss": 1.5166, "step": 3993 }, { "epoch": 0.6450258397932817, "grad_norm": 105.48208618164062, "learning_rate": 3.5497416020671837e-06, "loss": 0.7246, "step": 3994 }, { "epoch": 0.645187338501292, "grad_norm": 22.054311752319336, "learning_rate": 3.5481266149870803e-06, "loss": 0.2324, "step": 3995 }, { "epoch": 0.6453488372093024, "grad_norm": 70.68492889404297, "learning_rate": 3.5465116279069772e-06, "loss": 0.9443, "step": 3996 }, { "epoch": 0.6455103359173127, "grad_norm": 76.44707489013672, "learning_rate": 3.544896640826874e-06, "loss": 0.7344, "step": 3997 }, { "epoch": 0.645671834625323, "grad_norm": 43.82598114013672, "learning_rate": 3.54328165374677e-06, "loss": 0.2238, "step": 3998 }, { "epoch": 0.6458333333333334, "grad_norm": 91.01209259033203, "learning_rate": 3.5416666666666673e-06, "loss": 0.7776, "step": 3999 }, { "epoch": 0.6459948320413437, "grad_norm": 21.81119728088379, "learning_rate": 3.5400516795865635e-06, "loss": 0.0688, "step": 4000 }, { "epoch": 0.646156330749354, "grad_norm": 122.74840545654297, "learning_rate": 3.53843669250646e-06, "loss": 0.6367, "step": 4001 }, { "epoch": 0.6463178294573644, "grad_norm": 84.87368774414062, "learning_rate": 3.536821705426357e-06, "loss": 0.6035, "step": 4002 }, { "epoch": 0.6464793281653747, "grad_norm": 85.5205307006836, "learning_rate": 3.5352067183462536e-06, "loss": 0.8286, "step": 4003 }, { "epoch": 0.646640826873385, "grad_norm": 87.75443267822266, "learning_rate": 3.53359173126615e-06, "loss": 0.7678, "step": 4004 }, { "epoch": 0.6468023255813954, "grad_norm": 34.90277099609375, "learning_rate": 3.531976744186047e-06, "loss": 0.165, "step": 4005 }, { "epoch": 0.6469638242894057, "grad_norm": 224.765869140625, "learning_rate": 3.5303617571059433e-06, "loss": 1.1153, "step": 4006 }, { "epoch": 0.647125322997416, "grad_norm": 24.480682373046875, "learning_rate": 3.52874677002584e-06, "loss": 0.0787, "step": 4007 }, { "epoch": 0.6472868217054264, "grad_norm": 60.38601303100586, "learning_rate": 3.527131782945737e-06, "loss": 0.3749, "step": 4008 }, { "epoch": 0.6474483204134367, "grad_norm": 15.772540092468262, "learning_rate": 3.5255167958656334e-06, "loss": 0.0913, "step": 4009 }, { "epoch": 0.647609819121447, "grad_norm": 24.77691650390625, "learning_rate": 3.52390180878553e-06, "loss": 0.1008, "step": 4010 }, { "epoch": 0.6477713178294574, "grad_norm": 83.1695327758789, "learning_rate": 3.522286821705427e-06, "loss": 0.5137, "step": 4011 }, { "epoch": 0.6479328165374677, "grad_norm": 48.443504333496094, "learning_rate": 3.520671834625323e-06, "loss": 0.3276, "step": 4012 }, { "epoch": 0.648094315245478, "grad_norm": 23.960769653320312, "learning_rate": 3.5190568475452196e-06, "loss": 0.1189, "step": 4013 }, { "epoch": 0.6482558139534884, "grad_norm": 37.157649993896484, "learning_rate": 3.5174418604651166e-06, "loss": 0.1306, "step": 4014 }, { "epoch": 0.6484173126614987, "grad_norm": 9.350452423095703, "learning_rate": 3.515826873385013e-06, "loss": 0.0386, "step": 4015 }, { "epoch": 0.648578811369509, "grad_norm": 23.5723876953125, "learning_rate": 3.5142118863049097e-06, "loss": 0.0745, "step": 4016 }, { "epoch": 0.6487403100775194, "grad_norm": 155.0591583251953, "learning_rate": 3.5125968992248067e-06, "loss": 1.0101, "step": 4017 }, { "epoch": 0.6489018087855297, "grad_norm": 69.18051147460938, "learning_rate": 3.5109819121447033e-06, "loss": 0.1863, "step": 4018 }, { "epoch": 0.64906330749354, "grad_norm": 19.3592586517334, "learning_rate": 3.5093669250645994e-06, "loss": 0.0715, "step": 4019 }, { "epoch": 0.6492248062015504, "grad_norm": 16.516002655029297, "learning_rate": 3.5077519379844964e-06, "loss": 0.0508, "step": 4020 }, { "epoch": 0.6493863049095607, "grad_norm": 65.60688018798828, "learning_rate": 3.506136950904393e-06, "loss": 0.5145, "step": 4021 }, { "epoch": 0.649547803617571, "grad_norm": 50.68901062011719, "learning_rate": 3.5045219638242895e-06, "loss": 0.379, "step": 4022 }, { "epoch": 0.6497093023255814, "grad_norm": 118.41240692138672, "learning_rate": 3.5029069767441865e-06, "loss": 1.1211, "step": 4023 }, { "epoch": 0.6498708010335917, "grad_norm": 97.37555694580078, "learning_rate": 3.501291989664083e-06, "loss": 0.3848, "step": 4024 }, { "epoch": 0.650032299741602, "grad_norm": 44.662166595458984, "learning_rate": 3.499677002583979e-06, "loss": 0.21, "step": 4025 }, { "epoch": 0.6501937984496124, "grad_norm": 127.15914154052734, "learning_rate": 3.498062015503876e-06, "loss": 0.5381, "step": 4026 }, { "epoch": 0.6503552971576227, "grad_norm": 21.11202621459961, "learning_rate": 3.4964470284237727e-06, "loss": 0.1685, "step": 4027 }, { "epoch": 0.650516795865633, "grad_norm": 84.88455963134766, "learning_rate": 3.4948320413436693e-06, "loss": 0.4658, "step": 4028 }, { "epoch": 0.6506782945736435, "grad_norm": 27.044187545776367, "learning_rate": 3.4932170542635663e-06, "loss": 0.114, "step": 4029 }, { "epoch": 0.6508397932816538, "grad_norm": 64.17813873291016, "learning_rate": 3.491602067183463e-06, "loss": 0.3924, "step": 4030 }, { "epoch": 0.651001291989664, "grad_norm": 68.14521026611328, "learning_rate": 3.4899870801033594e-06, "loss": 0.1958, "step": 4031 }, { "epoch": 0.6511627906976745, "grad_norm": 19.187883377075195, "learning_rate": 3.4883720930232564e-06, "loss": 0.0999, "step": 4032 }, { "epoch": 0.6513242894056848, "grad_norm": 18.88838768005371, "learning_rate": 3.4867571059431525e-06, "loss": 0.0697, "step": 4033 }, { "epoch": 0.6514857881136951, "grad_norm": 12.139350891113281, "learning_rate": 3.4851421188630495e-06, "loss": 0.048, "step": 4034 }, { "epoch": 0.6516472868217055, "grad_norm": 59.62579345703125, "learning_rate": 3.483527131782946e-06, "loss": 0.2458, "step": 4035 }, { "epoch": 0.6518087855297158, "grad_norm": 115.59537506103516, "learning_rate": 3.4819121447028426e-06, "loss": 0.5054, "step": 4036 }, { "epoch": 0.6519702842377261, "grad_norm": 31.546192169189453, "learning_rate": 3.4802971576227396e-06, "loss": 0.146, "step": 4037 }, { "epoch": 0.6521317829457365, "grad_norm": 9.820865631103516, "learning_rate": 3.478682170542636e-06, "loss": 0.0273, "step": 4038 }, { "epoch": 0.6522932816537468, "grad_norm": 27.803892135620117, "learning_rate": 3.4770671834625323e-06, "loss": 0.1797, "step": 4039 }, { "epoch": 0.6524547803617571, "grad_norm": 19.246217727661133, "learning_rate": 3.4754521963824293e-06, "loss": 0.0666, "step": 4040 }, { "epoch": 0.6526162790697675, "grad_norm": 130.17726135253906, "learning_rate": 3.473837209302326e-06, "loss": 1.1953, "step": 4041 }, { "epoch": 0.6527777777777778, "grad_norm": 179.5092315673828, "learning_rate": 3.4722222222222224e-06, "loss": 0.5859, "step": 4042 }, { "epoch": 0.6529392764857881, "grad_norm": 39.0302619934082, "learning_rate": 3.4706072351421194e-06, "loss": 0.3452, "step": 4043 }, { "epoch": 0.6531007751937985, "grad_norm": 133.1734161376953, "learning_rate": 3.468992248062016e-06, "loss": 0.7725, "step": 4044 }, { "epoch": 0.6532622739018088, "grad_norm": 154.24237060546875, "learning_rate": 3.4673772609819125e-06, "loss": 1.4746, "step": 4045 }, { "epoch": 0.6534237726098191, "grad_norm": 59.90147018432617, "learning_rate": 3.4657622739018095e-06, "loss": 0.2839, "step": 4046 }, { "epoch": 0.6535852713178295, "grad_norm": 106.09088134765625, "learning_rate": 3.4641472868217056e-06, "loss": 0.6523, "step": 4047 }, { "epoch": 0.6537467700258398, "grad_norm": 227.05255126953125, "learning_rate": 3.462532299741602e-06, "loss": 3.5156, "step": 4048 }, { "epoch": 0.6539082687338501, "grad_norm": 28.476972579956055, "learning_rate": 3.460917312661499e-06, "loss": 0.0658, "step": 4049 }, { "epoch": 0.6540697674418605, "grad_norm": 268.4936218261719, "learning_rate": 3.4593023255813957e-06, "loss": 1.8564, "step": 4050 }, { "epoch": 0.6542312661498708, "grad_norm": 182.0825653076172, "learning_rate": 3.4576873385012923e-06, "loss": 1.6289, "step": 4051 }, { "epoch": 0.6543927648578811, "grad_norm": 7.148982524871826, "learning_rate": 3.4560723514211893e-06, "loss": 0.0387, "step": 4052 }, { "epoch": 0.6545542635658915, "grad_norm": 121.92388153076172, "learning_rate": 3.4544573643410854e-06, "loss": 0.6455, "step": 4053 }, { "epoch": 0.6547157622739018, "grad_norm": 33.01392364501953, "learning_rate": 3.452842377260982e-06, "loss": 0.2165, "step": 4054 }, { "epoch": 0.6548772609819121, "grad_norm": 25.150083541870117, "learning_rate": 3.451227390180879e-06, "loss": 0.1128, "step": 4055 }, { "epoch": 0.6550387596899225, "grad_norm": 18.238985061645508, "learning_rate": 3.4496124031007755e-06, "loss": 0.092, "step": 4056 }, { "epoch": 0.6552002583979328, "grad_norm": 58.39970016479492, "learning_rate": 3.447997416020672e-06, "loss": 0.4531, "step": 4057 }, { "epoch": 0.6553617571059431, "grad_norm": 24.142118453979492, "learning_rate": 3.446382428940569e-06, "loss": 0.0743, "step": 4058 }, { "epoch": 0.6555232558139535, "grad_norm": 132.02120971679688, "learning_rate": 3.4447674418604656e-06, "loss": 1.3718, "step": 4059 }, { "epoch": 0.6556847545219638, "grad_norm": 30.029844284057617, "learning_rate": 3.4431524547803617e-06, "loss": 0.0854, "step": 4060 }, { "epoch": 0.6558462532299741, "grad_norm": 106.54950714111328, "learning_rate": 3.4415374677002587e-06, "loss": 1.2758, "step": 4061 }, { "epoch": 0.6560077519379846, "grad_norm": 10.890609741210938, "learning_rate": 3.4399224806201553e-06, "loss": 0.0366, "step": 4062 }, { "epoch": 0.6561692506459949, "grad_norm": 8.759427070617676, "learning_rate": 3.438307493540052e-06, "loss": 0.0372, "step": 4063 }, { "epoch": 0.6563307493540051, "grad_norm": 52.23380661010742, "learning_rate": 3.436692506459949e-06, "loss": 0.2397, "step": 4064 }, { "epoch": 0.6564922480620154, "grad_norm": 24.927715301513672, "learning_rate": 3.4350775193798454e-06, "loss": 0.0812, "step": 4065 }, { "epoch": 0.6566537467700259, "grad_norm": 256.84765625, "learning_rate": 3.4334625322997415e-06, "loss": 2.2911, "step": 4066 }, { "epoch": 0.6568152454780362, "grad_norm": 22.64592742919922, "learning_rate": 3.4318475452196385e-06, "loss": 0.0635, "step": 4067 }, { "epoch": 0.6569767441860465, "grad_norm": 92.90618896484375, "learning_rate": 3.430232558139535e-06, "loss": 0.7859, "step": 4068 }, { "epoch": 0.6571382428940569, "grad_norm": 41.94186782836914, "learning_rate": 3.4286175710594316e-06, "loss": 0.3057, "step": 4069 }, { "epoch": 0.6572997416020672, "grad_norm": 300.4344482421875, "learning_rate": 3.4270025839793286e-06, "loss": 2.2344, "step": 4070 }, { "epoch": 0.6574612403100775, "grad_norm": 151.1815643310547, "learning_rate": 3.425387596899225e-06, "loss": 0.7017, "step": 4071 }, { "epoch": 0.6576227390180879, "grad_norm": 95.16336822509766, "learning_rate": 3.4237726098191217e-06, "loss": 0.8892, "step": 4072 }, { "epoch": 0.6577842377260982, "grad_norm": 134.1562042236328, "learning_rate": 3.4221576227390187e-06, "loss": 0.8477, "step": 4073 }, { "epoch": 0.6579457364341085, "grad_norm": 83.32737731933594, "learning_rate": 3.420542635658915e-06, "loss": 0.752, "step": 4074 }, { "epoch": 0.6581072351421189, "grad_norm": 125.16368865966797, "learning_rate": 3.4189276485788114e-06, "loss": 0.7656, "step": 4075 }, { "epoch": 0.6582687338501292, "grad_norm": 168.53121948242188, "learning_rate": 3.4173126614987084e-06, "loss": 1.3027, "step": 4076 }, { "epoch": 0.6584302325581395, "grad_norm": 8.020203590393066, "learning_rate": 3.415697674418605e-06, "loss": 0.0344, "step": 4077 }, { "epoch": 0.6585917312661499, "grad_norm": 38.875755310058594, "learning_rate": 3.4140826873385015e-06, "loss": 0.1292, "step": 4078 }, { "epoch": 0.6587532299741602, "grad_norm": 432.4205627441406, "learning_rate": 3.4124677002583985e-06, "loss": 1.4102, "step": 4079 }, { "epoch": 0.6589147286821705, "grad_norm": 15.261824607849121, "learning_rate": 3.4108527131782946e-06, "loss": 0.0392, "step": 4080 }, { "epoch": 0.6590762273901809, "grad_norm": 136.78253173828125, "learning_rate": 3.409237726098191e-06, "loss": 0.9824, "step": 4081 }, { "epoch": 0.6592377260981912, "grad_norm": 28.49374771118164, "learning_rate": 3.407622739018088e-06, "loss": 0.1317, "step": 4082 }, { "epoch": 0.6593992248062015, "grad_norm": 20.083337783813477, "learning_rate": 3.4060077519379847e-06, "loss": 0.0954, "step": 4083 }, { "epoch": 0.6595607235142119, "grad_norm": 6.2153472900390625, "learning_rate": 3.4043927648578813e-06, "loss": 0.0442, "step": 4084 }, { "epoch": 0.6597222222222222, "grad_norm": 60.77741622924805, "learning_rate": 3.4027777777777783e-06, "loss": 0.2466, "step": 4085 }, { "epoch": 0.6598837209302325, "grad_norm": 69.50476837158203, "learning_rate": 3.401162790697675e-06, "loss": 0.7153, "step": 4086 }, { "epoch": 0.6600452196382429, "grad_norm": 15.357965469360352, "learning_rate": 3.399547803617571e-06, "loss": 0.0377, "step": 4087 }, { "epoch": 0.6602067183462532, "grad_norm": 21.997787475585938, "learning_rate": 3.397932816537468e-06, "loss": 0.0957, "step": 4088 }, { "epoch": 0.6603682170542635, "grad_norm": 55.65455627441406, "learning_rate": 3.3963178294573645e-06, "loss": 0.4275, "step": 4089 }, { "epoch": 0.6605297157622739, "grad_norm": 87.42382049560547, "learning_rate": 3.394702842377261e-06, "loss": 0.7358, "step": 4090 }, { "epoch": 0.6606912144702842, "grad_norm": 5.945248603820801, "learning_rate": 3.393087855297158e-06, "loss": 0.0219, "step": 4091 }, { "epoch": 0.6608527131782945, "grad_norm": 277.4178466796875, "learning_rate": 3.3914728682170546e-06, "loss": 2.0781, "step": 4092 }, { "epoch": 0.6610142118863049, "grad_norm": 40.419288635253906, "learning_rate": 3.3898578811369508e-06, "loss": 0.4434, "step": 4093 }, { "epoch": 0.6611757105943152, "grad_norm": 8.64141845703125, "learning_rate": 3.3882428940568477e-06, "loss": 0.0387, "step": 4094 }, { "epoch": 0.6613372093023255, "grad_norm": 95.42656707763672, "learning_rate": 3.3866279069767443e-06, "loss": 0.9326, "step": 4095 }, { "epoch": 0.661498708010336, "grad_norm": 9.114093780517578, "learning_rate": 3.3850129198966413e-06, "loss": 0.0341, "step": 4096 }, { "epoch": 0.6616602067183462, "grad_norm": 37.111026763916016, "learning_rate": 3.383397932816538e-06, "loss": 0.1782, "step": 4097 }, { "epoch": 0.6618217054263565, "grad_norm": 65.79957580566406, "learning_rate": 3.3817829457364344e-06, "loss": 0.4581, "step": 4098 }, { "epoch": 0.661983204134367, "grad_norm": 20.521034240722656, "learning_rate": 3.3801679586563314e-06, "loss": 0.0569, "step": 4099 }, { "epoch": 0.6621447028423773, "grad_norm": 19.633899688720703, "learning_rate": 3.378552971576228e-06, "loss": 0.0408, "step": 4100 }, { "epoch": 0.6623062015503876, "grad_norm": 103.48189544677734, "learning_rate": 3.376937984496124e-06, "loss": 1.3633, "step": 4101 }, { "epoch": 0.662467700258398, "grad_norm": 150.4292449951172, "learning_rate": 3.375322997416021e-06, "loss": 1.1523, "step": 4102 }, { "epoch": 0.6626291989664083, "grad_norm": 183.40809631347656, "learning_rate": 3.3737080103359176e-06, "loss": 1.0078, "step": 4103 }, { "epoch": 0.6627906976744186, "grad_norm": 330.03265380859375, "learning_rate": 3.372093023255814e-06, "loss": 4.4512, "step": 4104 }, { "epoch": 0.662952196382429, "grad_norm": 4.124258518218994, "learning_rate": 3.370478036175711e-06, "loss": 0.0169, "step": 4105 }, { "epoch": 0.6631136950904393, "grad_norm": 39.024330139160156, "learning_rate": 3.3688630490956077e-06, "loss": 0.1865, "step": 4106 }, { "epoch": 0.6632751937984496, "grad_norm": 43.55690383911133, "learning_rate": 3.367248062015504e-06, "loss": 0.5283, "step": 4107 }, { "epoch": 0.66343669250646, "grad_norm": 59.88417053222656, "learning_rate": 3.365633074935401e-06, "loss": 0.3271, "step": 4108 }, { "epoch": 0.6635981912144703, "grad_norm": 298.41558837890625, "learning_rate": 3.3640180878552974e-06, "loss": 3.791, "step": 4109 }, { "epoch": 0.6637596899224806, "grad_norm": 114.29387664794922, "learning_rate": 3.362403100775194e-06, "loss": 1.8643, "step": 4110 }, { "epoch": 0.663921188630491, "grad_norm": 67.30947875976562, "learning_rate": 3.360788113695091e-06, "loss": 0.7559, "step": 4111 }, { "epoch": 0.6640826873385013, "grad_norm": 87.20885467529297, "learning_rate": 3.3591731266149875e-06, "loss": 0.3649, "step": 4112 }, { "epoch": 0.6642441860465116, "grad_norm": 17.694272994995117, "learning_rate": 3.3575581395348837e-06, "loss": 0.0722, "step": 4113 }, { "epoch": 0.664405684754522, "grad_norm": 124.30349731445312, "learning_rate": 3.355943152454781e-06, "loss": 0.7266, "step": 4114 }, { "epoch": 0.6645671834625323, "grad_norm": 9.665562629699707, "learning_rate": 3.354328165374677e-06, "loss": 0.0281, "step": 4115 }, { "epoch": 0.6647286821705426, "grad_norm": 15.252203941345215, "learning_rate": 3.3527131782945738e-06, "loss": 0.071, "step": 4116 }, { "epoch": 0.664890180878553, "grad_norm": 51.66657638549805, "learning_rate": 3.3510981912144707e-06, "loss": 0.3848, "step": 4117 }, { "epoch": 0.6650516795865633, "grad_norm": 271.1101989746094, "learning_rate": 3.3494832041343673e-06, "loss": 3.793, "step": 4118 }, { "epoch": 0.6652131782945736, "grad_norm": 49.083290100097656, "learning_rate": 3.347868217054264e-06, "loss": 0.4003, "step": 4119 }, { "epoch": 0.665374677002584, "grad_norm": 84.14884185791016, "learning_rate": 3.346253229974161e-06, "loss": 0.6436, "step": 4120 }, { "epoch": 0.6655361757105943, "grad_norm": 27.37792205810547, "learning_rate": 3.344638242894057e-06, "loss": 0.1177, "step": 4121 }, { "epoch": 0.6656976744186046, "grad_norm": 209.76976013183594, "learning_rate": 3.3430232558139535e-06, "loss": 0.6958, "step": 4122 }, { "epoch": 0.665859173126615, "grad_norm": 41.370201110839844, "learning_rate": 3.3414082687338505e-06, "loss": 0.1553, "step": 4123 }, { "epoch": 0.6660206718346253, "grad_norm": 3.215097427368164, "learning_rate": 3.339793281653747e-06, "loss": 0.0134, "step": 4124 }, { "epoch": 0.6661821705426356, "grad_norm": 22.340993881225586, "learning_rate": 3.3381782945736436e-06, "loss": 0.0746, "step": 4125 }, { "epoch": 0.666343669250646, "grad_norm": 60.89006042480469, "learning_rate": 3.3365633074935406e-06, "loss": 0.4297, "step": 4126 }, { "epoch": 0.6665051679586563, "grad_norm": 65.83216094970703, "learning_rate": 3.3349483204134368e-06, "loss": 0.1055, "step": 4127 }, { "epoch": 0.6666666666666666, "grad_norm": 103.43423461914062, "learning_rate": 3.3333333333333333e-06, "loss": 0.3352, "step": 4128 }, { "epoch": 0.666828165374677, "grad_norm": 48.43952178955078, "learning_rate": 3.3317183462532303e-06, "loss": 0.2954, "step": 4129 }, { "epoch": 0.6669896640826873, "grad_norm": 34.26655960083008, "learning_rate": 3.330103359173127e-06, "loss": 0.1246, "step": 4130 }, { "epoch": 0.6671511627906976, "grad_norm": 232.73028564453125, "learning_rate": 3.3284883720930234e-06, "loss": 1.0703, "step": 4131 }, { "epoch": 0.667312661498708, "grad_norm": 12.061991691589355, "learning_rate": 3.3268733850129204e-06, "loss": 0.0615, "step": 4132 }, { "epoch": 0.6674741602067183, "grad_norm": 75.71231842041016, "learning_rate": 3.325258397932817e-06, "loss": 0.7148, "step": 4133 }, { "epoch": 0.6676356589147286, "grad_norm": 181.4583282470703, "learning_rate": 3.323643410852713e-06, "loss": 2.1914, "step": 4134 }, { "epoch": 0.6677971576227391, "grad_norm": 45.250701904296875, "learning_rate": 3.32202842377261e-06, "loss": 0.2161, "step": 4135 }, { "epoch": 0.6679586563307494, "grad_norm": 13.58615493774414, "learning_rate": 3.3204134366925067e-06, "loss": 0.027, "step": 4136 }, { "epoch": 0.6681201550387597, "grad_norm": 22.69464874267578, "learning_rate": 3.318798449612403e-06, "loss": 0.0732, "step": 4137 }, { "epoch": 0.6682816537467701, "grad_norm": 43.24041748046875, "learning_rate": 3.3171834625323e-06, "loss": 0.3439, "step": 4138 }, { "epoch": 0.6684431524547804, "grad_norm": 16.160778045654297, "learning_rate": 3.3155684754521968e-06, "loss": 0.0442, "step": 4139 }, { "epoch": 0.6686046511627907, "grad_norm": 30.476835250854492, "learning_rate": 3.313953488372093e-06, "loss": 0.0955, "step": 4140 }, { "epoch": 0.6687661498708011, "grad_norm": 105.7751693725586, "learning_rate": 3.31233850129199e-06, "loss": 0.3385, "step": 4141 }, { "epoch": 0.6689276485788114, "grad_norm": 45.271095275878906, "learning_rate": 3.3107235142118864e-06, "loss": 0.2922, "step": 4142 }, { "epoch": 0.6690891472868217, "grad_norm": 58.43315887451172, "learning_rate": 3.309108527131783e-06, "loss": 0.2612, "step": 4143 }, { "epoch": 0.6692506459948321, "grad_norm": 54.70867156982422, "learning_rate": 3.30749354005168e-06, "loss": 0.4077, "step": 4144 }, { "epoch": 0.6694121447028424, "grad_norm": 16.640724182128906, "learning_rate": 3.3058785529715765e-06, "loss": 0.027, "step": 4145 }, { "epoch": 0.6695736434108527, "grad_norm": 106.15853118896484, "learning_rate": 3.304263565891473e-06, "loss": 0.542, "step": 4146 }, { "epoch": 0.6697351421188631, "grad_norm": 84.04988861083984, "learning_rate": 3.30264857881137e-06, "loss": 0.6345, "step": 4147 }, { "epoch": 0.6698966408268734, "grad_norm": 15.668933868408203, "learning_rate": 3.3010335917312662e-06, "loss": 0.1055, "step": 4148 }, { "epoch": 0.6700581395348837, "grad_norm": 22.54230308532715, "learning_rate": 3.2994186046511628e-06, "loss": 0.0616, "step": 4149 }, { "epoch": 0.6702196382428941, "grad_norm": 126.85286712646484, "learning_rate": 3.2978036175710598e-06, "loss": 0.394, "step": 4150 }, { "epoch": 0.6703811369509044, "grad_norm": 284.0020751953125, "learning_rate": 3.2961886304909563e-06, "loss": 2.2969, "step": 4151 }, { "epoch": 0.6705426356589147, "grad_norm": 114.4161376953125, "learning_rate": 3.294573643410853e-06, "loss": 0.7116, "step": 4152 }, { "epoch": 0.6707041343669251, "grad_norm": 63.88688659667969, "learning_rate": 3.29295865633075e-06, "loss": 0.388, "step": 4153 }, { "epoch": 0.6708656330749354, "grad_norm": 52.91446304321289, "learning_rate": 3.291343669250646e-06, "loss": 0.2131, "step": 4154 }, { "epoch": 0.6710271317829457, "grad_norm": 4.589268207550049, "learning_rate": 3.2897286821705426e-06, "loss": 0.0262, "step": 4155 }, { "epoch": 0.6711886304909561, "grad_norm": 172.34988403320312, "learning_rate": 3.2881136950904395e-06, "loss": 0.6797, "step": 4156 }, { "epoch": 0.6713501291989664, "grad_norm": 144.4352569580078, "learning_rate": 3.286498708010336e-06, "loss": 0.6226, "step": 4157 }, { "epoch": 0.6715116279069767, "grad_norm": 69.2967758178711, "learning_rate": 3.2848837209302327e-06, "loss": 0.3838, "step": 4158 }, { "epoch": 0.6716731266149871, "grad_norm": 86.44055938720703, "learning_rate": 3.2832687338501296e-06, "loss": 0.3362, "step": 4159 }, { "epoch": 0.6718346253229974, "grad_norm": 63.48957443237305, "learning_rate": 3.281653746770026e-06, "loss": 0.162, "step": 4160 }, { "epoch": 0.6719961240310077, "grad_norm": 114.60958099365234, "learning_rate": 3.280038759689923e-06, "loss": 0.3296, "step": 4161 }, { "epoch": 0.6721576227390181, "grad_norm": 41.0926513671875, "learning_rate": 3.2784237726098193e-06, "loss": 0.2366, "step": 4162 }, { "epoch": 0.6723191214470284, "grad_norm": 333.74200439453125, "learning_rate": 3.276808785529716e-06, "loss": 2.0952, "step": 4163 }, { "epoch": 0.6724806201550387, "grad_norm": 68.67015838623047, "learning_rate": 3.275193798449613e-06, "loss": 0.2339, "step": 4164 }, { "epoch": 0.6726421188630491, "grad_norm": 64.46421813964844, "learning_rate": 3.2735788113695094e-06, "loss": 0.3047, "step": 4165 }, { "epoch": 0.6728036175710594, "grad_norm": 249.10699462890625, "learning_rate": 3.271963824289406e-06, "loss": 1.7744, "step": 4166 }, { "epoch": 0.6729651162790697, "grad_norm": 92.62248229980469, "learning_rate": 3.270348837209303e-06, "loss": 1.289, "step": 4167 }, { "epoch": 0.6731266149870802, "grad_norm": 77.3644027709961, "learning_rate": 3.268733850129199e-06, "loss": 0.2366, "step": 4168 }, { "epoch": 0.6732881136950905, "grad_norm": 202.244384765625, "learning_rate": 3.2671188630490957e-06, "loss": 0.7305, "step": 4169 }, { "epoch": 0.6734496124031008, "grad_norm": 74.12789916992188, "learning_rate": 3.2655038759689927e-06, "loss": 0.6163, "step": 4170 }, { "epoch": 0.6736111111111112, "grad_norm": 10.558059692382812, "learning_rate": 3.2638888888888892e-06, "loss": 0.0322, "step": 4171 }, { "epoch": 0.6737726098191215, "grad_norm": 99.56394958496094, "learning_rate": 3.2622739018087858e-06, "loss": 1.7969, "step": 4172 }, { "epoch": 0.6739341085271318, "grad_norm": 242.79879760742188, "learning_rate": 3.2606589147286828e-06, "loss": 4.8691, "step": 4173 }, { "epoch": 0.6740956072351422, "grad_norm": 358.5068664550781, "learning_rate": 3.2590439276485793e-06, "loss": 4.1406, "step": 4174 }, { "epoch": 0.6742571059431525, "grad_norm": 110.85617065429688, "learning_rate": 3.2574289405684755e-06, "loss": 0.1352, "step": 4175 }, { "epoch": 0.6744186046511628, "grad_norm": 182.3379364013672, "learning_rate": 3.2558139534883724e-06, "loss": 1.9895, "step": 4176 }, { "epoch": 0.6745801033591732, "grad_norm": 13.433889389038086, "learning_rate": 3.254198966408269e-06, "loss": 0.0457, "step": 4177 }, { "epoch": 0.6747416020671835, "grad_norm": 22.452089309692383, "learning_rate": 3.2525839793281656e-06, "loss": 0.1384, "step": 4178 }, { "epoch": 0.6749031007751938, "grad_norm": 30.990312576293945, "learning_rate": 3.2509689922480625e-06, "loss": 0.1343, "step": 4179 }, { "epoch": 0.6750645994832042, "grad_norm": 56.60266876220703, "learning_rate": 3.249354005167959e-06, "loss": 0.2207, "step": 4180 }, { "epoch": 0.6752260981912145, "grad_norm": 157.00294494628906, "learning_rate": 3.2477390180878552e-06, "loss": 0.415, "step": 4181 }, { "epoch": 0.6753875968992248, "grad_norm": 121.42749786376953, "learning_rate": 3.2461240310077522e-06, "loss": 0.9561, "step": 4182 }, { "epoch": 0.6755490956072352, "grad_norm": 93.87796783447266, "learning_rate": 3.2445090439276488e-06, "loss": 0.8018, "step": 4183 }, { "epoch": 0.6757105943152455, "grad_norm": 45.255550384521484, "learning_rate": 3.2428940568475453e-06, "loss": 0.4892, "step": 4184 }, { "epoch": 0.6758720930232558, "grad_norm": 137.9204864501953, "learning_rate": 3.2412790697674423e-06, "loss": 1.0039, "step": 4185 }, { "epoch": 0.6760335917312662, "grad_norm": 64.00491333007812, "learning_rate": 3.239664082687339e-06, "loss": 0.314, "step": 4186 }, { "epoch": 0.6761950904392765, "grad_norm": 379.3918151855469, "learning_rate": 3.2380490956072354e-06, "loss": 1.8028, "step": 4187 }, { "epoch": 0.6763565891472868, "grad_norm": 12.59438705444336, "learning_rate": 3.2364341085271324e-06, "loss": 0.0453, "step": 4188 }, { "epoch": 0.6765180878552972, "grad_norm": 107.7799072265625, "learning_rate": 3.2348191214470286e-06, "loss": 0.9863, "step": 4189 }, { "epoch": 0.6766795865633075, "grad_norm": 140.57427978515625, "learning_rate": 3.233204134366925e-06, "loss": 0.6045, "step": 4190 }, { "epoch": 0.6768410852713178, "grad_norm": 17.366025924682617, "learning_rate": 3.231589147286822e-06, "loss": 0.0237, "step": 4191 }, { "epoch": 0.6770025839793282, "grad_norm": 76.19950866699219, "learning_rate": 3.2299741602067187e-06, "loss": 0.8438, "step": 4192 }, { "epoch": 0.6771640826873385, "grad_norm": 57.78287887573242, "learning_rate": 3.2283591731266152e-06, "loss": 0.5265, "step": 4193 }, { "epoch": 0.6773255813953488, "grad_norm": 33.08108901977539, "learning_rate": 3.2267441860465122e-06, "loss": 0.2476, "step": 4194 }, { "epoch": 0.6774870801033591, "grad_norm": 10.125252723693848, "learning_rate": 3.2251291989664083e-06, "loss": 0.0745, "step": 4195 }, { "epoch": 0.6776485788113695, "grad_norm": 69.54145050048828, "learning_rate": 3.223514211886305e-06, "loss": 0.2344, "step": 4196 }, { "epoch": 0.6778100775193798, "grad_norm": 34.18072509765625, "learning_rate": 3.221899224806202e-06, "loss": 0.2354, "step": 4197 }, { "epoch": 0.6779715762273901, "grad_norm": 8.690866470336914, "learning_rate": 3.2202842377260985e-06, "loss": 0.0306, "step": 4198 }, { "epoch": 0.6781330749354005, "grad_norm": 52.242958068847656, "learning_rate": 3.218669250645995e-06, "loss": 0.3521, "step": 4199 }, { "epoch": 0.6782945736434108, "grad_norm": 133.71926879882812, "learning_rate": 3.217054263565892e-06, "loss": 0.7739, "step": 4200 }, { "epoch": 0.6784560723514211, "grad_norm": 57.64510726928711, "learning_rate": 3.2154392764857886e-06, "loss": 0.2056, "step": 4201 }, { "epoch": 0.6786175710594315, "grad_norm": 69.16544342041016, "learning_rate": 3.2138242894056847e-06, "loss": 0.542, "step": 4202 }, { "epoch": 0.6787790697674418, "grad_norm": 70.96756744384766, "learning_rate": 3.2122093023255817e-06, "loss": 1.106, "step": 4203 }, { "epoch": 0.6789405684754521, "grad_norm": 56.256465911865234, "learning_rate": 3.2105943152454782e-06, "loss": 0.5635, "step": 4204 }, { "epoch": 0.6791020671834626, "grad_norm": 13.18403434753418, "learning_rate": 3.208979328165375e-06, "loss": 0.0939, "step": 4205 }, { "epoch": 0.6792635658914729, "grad_norm": 15.586018562316895, "learning_rate": 3.2073643410852718e-06, "loss": 0.0611, "step": 4206 }, { "epoch": 0.6794250645994832, "grad_norm": 85.45006561279297, "learning_rate": 3.2057493540051683e-06, "loss": 0.8765, "step": 4207 }, { "epoch": 0.6795865633074936, "grad_norm": 43.690284729003906, "learning_rate": 3.2041343669250645e-06, "loss": 0.2034, "step": 4208 }, { "epoch": 0.6797480620155039, "grad_norm": 37.228641510009766, "learning_rate": 3.2025193798449615e-06, "loss": 0.2964, "step": 4209 }, { "epoch": 0.6799095607235142, "grad_norm": 8.67248821258545, "learning_rate": 3.200904392764858e-06, "loss": 0.0154, "step": 4210 }, { "epoch": 0.6800710594315246, "grad_norm": 239.28155517578125, "learning_rate": 3.1992894056847546e-06, "loss": 2.0645, "step": 4211 }, { "epoch": 0.6802325581395349, "grad_norm": 79.99638366699219, "learning_rate": 3.1976744186046516e-06, "loss": 0.4717, "step": 4212 }, { "epoch": 0.6803940568475452, "grad_norm": 13.583005905151367, "learning_rate": 3.196059431524548e-06, "loss": 0.0425, "step": 4213 }, { "epoch": 0.6805555555555556, "grad_norm": 34.9345817565918, "learning_rate": 3.1944444444444443e-06, "loss": 0.1213, "step": 4214 }, { "epoch": 0.6807170542635659, "grad_norm": 6.314236164093018, "learning_rate": 3.1928294573643417e-06, "loss": 0.0064, "step": 4215 }, { "epoch": 0.6808785529715762, "grad_norm": 130.48153686523438, "learning_rate": 3.191214470284238e-06, "loss": 0.3396, "step": 4216 }, { "epoch": 0.6810400516795866, "grad_norm": 33.925689697265625, "learning_rate": 3.1895994832041344e-06, "loss": 0.149, "step": 4217 }, { "epoch": 0.6812015503875969, "grad_norm": 62.862525939941406, "learning_rate": 3.1879844961240313e-06, "loss": 0.2419, "step": 4218 }, { "epoch": 0.6813630490956072, "grad_norm": 7.152085304260254, "learning_rate": 3.186369509043928e-06, "loss": 0.0318, "step": 4219 }, { "epoch": 0.6815245478036176, "grad_norm": 143.87017822265625, "learning_rate": 3.1847545219638245e-06, "loss": 0.504, "step": 4220 }, { "epoch": 0.6816860465116279, "grad_norm": 133.09051513671875, "learning_rate": 3.1831395348837214e-06, "loss": 1.4102, "step": 4221 }, { "epoch": 0.6818475452196382, "grad_norm": 51.946571350097656, "learning_rate": 3.1815245478036176e-06, "loss": 0.666, "step": 4222 }, { "epoch": 0.6820090439276486, "grad_norm": 19.13260841369629, "learning_rate": 3.179909560723514e-06, "loss": 0.1453, "step": 4223 }, { "epoch": 0.6821705426356589, "grad_norm": 30.896350860595703, "learning_rate": 3.178294573643411e-06, "loss": 0.0901, "step": 4224 }, { "epoch": 0.6823320413436692, "grad_norm": 14.9568452835083, "learning_rate": 3.1766795865633077e-06, "loss": 0.0504, "step": 4225 }, { "epoch": 0.6824935400516796, "grad_norm": 98.3282470703125, "learning_rate": 3.1750645994832047e-06, "loss": 0.6465, "step": 4226 }, { "epoch": 0.6826550387596899, "grad_norm": 112.7584457397461, "learning_rate": 3.1734496124031012e-06, "loss": 0.5305, "step": 4227 }, { "epoch": 0.6828165374677002, "grad_norm": 56.272315979003906, "learning_rate": 3.1718346253229974e-06, "loss": 0.2687, "step": 4228 }, { "epoch": 0.6829780361757106, "grad_norm": 67.32918548583984, "learning_rate": 3.1702196382428948e-06, "loss": 0.2407, "step": 4229 }, { "epoch": 0.6831395348837209, "grad_norm": 26.998205184936523, "learning_rate": 3.168604651162791e-06, "loss": 0.0669, "step": 4230 }, { "epoch": 0.6833010335917312, "grad_norm": 23.699615478515625, "learning_rate": 3.1669896640826875e-06, "loss": 0.0994, "step": 4231 }, { "epoch": 0.6834625322997416, "grad_norm": 57.949462890625, "learning_rate": 3.1653746770025845e-06, "loss": 0.2427, "step": 4232 }, { "epoch": 0.6836240310077519, "grad_norm": 135.9049072265625, "learning_rate": 3.163759689922481e-06, "loss": 0.8184, "step": 4233 }, { "epoch": 0.6837855297157622, "grad_norm": 206.87327575683594, "learning_rate": 3.1621447028423776e-06, "loss": 1.7773, "step": 4234 }, { "epoch": 0.6839470284237726, "grad_norm": 25.668964385986328, "learning_rate": 3.1605297157622746e-06, "loss": 0.082, "step": 4235 }, { "epoch": 0.6841085271317829, "grad_norm": 13.200111389160156, "learning_rate": 3.1589147286821707e-06, "loss": 0.1138, "step": 4236 }, { "epoch": 0.6842700258397932, "grad_norm": 124.53187561035156, "learning_rate": 3.1572997416020673e-06, "loss": 1.4316, "step": 4237 }, { "epoch": 0.6844315245478036, "grad_norm": 39.65747833251953, "learning_rate": 3.1556847545219642e-06, "loss": 0.1556, "step": 4238 }, { "epoch": 0.684593023255814, "grad_norm": 190.40084838867188, "learning_rate": 3.154069767441861e-06, "loss": 1.6348, "step": 4239 }, { "epoch": 0.6847545219638242, "grad_norm": 69.83956909179688, "learning_rate": 3.1524547803617574e-06, "loss": 0.6378, "step": 4240 }, { "epoch": 0.6849160206718347, "grad_norm": 27.234479904174805, "learning_rate": 3.1508397932816543e-06, "loss": 0.0908, "step": 4241 }, { "epoch": 0.685077519379845, "grad_norm": 83.7167739868164, "learning_rate": 3.1492248062015505e-06, "loss": 0.5713, "step": 4242 }, { "epoch": 0.6852390180878553, "grad_norm": 18.056598663330078, "learning_rate": 3.147609819121447e-06, "loss": 0.0508, "step": 4243 }, { "epoch": 0.6854005167958657, "grad_norm": 221.33432006835938, "learning_rate": 3.145994832041344e-06, "loss": 1.5859, "step": 4244 }, { "epoch": 0.685562015503876, "grad_norm": 310.1512756347656, "learning_rate": 3.1443798449612406e-06, "loss": 1.8125, "step": 4245 }, { "epoch": 0.6857235142118863, "grad_norm": 9.762502670288086, "learning_rate": 3.142764857881137e-06, "loss": 0.0164, "step": 4246 }, { "epoch": 0.6858850129198967, "grad_norm": 100.84883117675781, "learning_rate": 3.141149870801034e-06, "loss": 0.8984, "step": 4247 }, { "epoch": 0.686046511627907, "grad_norm": 9.677896499633789, "learning_rate": 3.1395348837209307e-06, "loss": 0.0626, "step": 4248 }, { "epoch": 0.6862080103359173, "grad_norm": 18.200355529785156, "learning_rate": 3.137919896640827e-06, "loss": 0.0436, "step": 4249 }, { "epoch": 0.6863695090439277, "grad_norm": 79.07083892822266, "learning_rate": 3.136304909560724e-06, "loss": 0.5103, "step": 4250 }, { "epoch": 0.686531007751938, "grad_norm": 19.159469604492188, "learning_rate": 3.1346899224806204e-06, "loss": 0.1057, "step": 4251 }, { "epoch": 0.6866925064599483, "grad_norm": 109.02156066894531, "learning_rate": 3.133074935400517e-06, "loss": 0.5056, "step": 4252 }, { "epoch": 0.6868540051679587, "grad_norm": 12.355266571044922, "learning_rate": 3.131459948320414e-06, "loss": 0.029, "step": 4253 }, { "epoch": 0.687015503875969, "grad_norm": 47.43425750732422, "learning_rate": 3.1298449612403105e-06, "loss": 0.1588, "step": 4254 }, { "epoch": 0.6871770025839793, "grad_norm": 102.82567596435547, "learning_rate": 3.1282299741602066e-06, "loss": 1.1425, "step": 4255 }, { "epoch": 0.6873385012919897, "grad_norm": 92.8921127319336, "learning_rate": 3.126614987080104e-06, "loss": 0.8459, "step": 4256 }, { "epoch": 0.6875, "grad_norm": 49.4465217590332, "learning_rate": 3.125e-06, "loss": 0.2325, "step": 4257 }, { "epoch": 0.6876614987080103, "grad_norm": 24.448284149169922, "learning_rate": 3.1233850129198967e-06, "loss": 0.1262, "step": 4258 }, { "epoch": 0.6878229974160207, "grad_norm": 14.649596214294434, "learning_rate": 3.1217700258397937e-06, "loss": 0.0374, "step": 4259 }, { "epoch": 0.687984496124031, "grad_norm": 67.73892211914062, "learning_rate": 3.1201550387596903e-06, "loss": 0.3521, "step": 4260 }, { "epoch": 0.6881459948320413, "grad_norm": 190.56198120117188, "learning_rate": 3.118540051679587e-06, "loss": 0.5859, "step": 4261 }, { "epoch": 0.6883074935400517, "grad_norm": 59.845027923583984, "learning_rate": 3.116925064599484e-06, "loss": 0.4169, "step": 4262 }, { "epoch": 0.688468992248062, "grad_norm": 11.889394760131836, "learning_rate": 3.11531007751938e-06, "loss": 0.0458, "step": 4263 }, { "epoch": 0.6886304909560723, "grad_norm": 23.250776290893555, "learning_rate": 3.1136950904392765e-06, "loss": 0.0727, "step": 4264 }, { "epoch": 0.6887919896640827, "grad_norm": 62.42424392700195, "learning_rate": 3.1120801033591735e-06, "loss": 0.6178, "step": 4265 }, { "epoch": 0.688953488372093, "grad_norm": 17.150169372558594, "learning_rate": 3.11046511627907e-06, "loss": 0.1221, "step": 4266 }, { "epoch": 0.6891149870801033, "grad_norm": 132.66578674316406, "learning_rate": 3.1088501291989666e-06, "loss": 0.5979, "step": 4267 }, { "epoch": 0.6892764857881137, "grad_norm": 98.10713958740234, "learning_rate": 3.1072351421188636e-06, "loss": 0.792, "step": 4268 }, { "epoch": 0.689437984496124, "grad_norm": 37.0681037902832, "learning_rate": 3.1056201550387597e-06, "loss": 0.1044, "step": 4269 }, { "epoch": 0.6895994832041343, "grad_norm": 64.42164611816406, "learning_rate": 3.1040051679586563e-06, "loss": 0.6406, "step": 4270 }, { "epoch": 0.6897609819121447, "grad_norm": 152.35325622558594, "learning_rate": 3.1023901808785533e-06, "loss": 0.6348, "step": 4271 }, { "epoch": 0.689922480620155, "grad_norm": 25.112716674804688, "learning_rate": 3.10077519379845e-06, "loss": 0.08, "step": 4272 }, { "epoch": 0.6900839793281653, "grad_norm": 254.3738555908203, "learning_rate": 3.0991602067183464e-06, "loss": 4.0084, "step": 4273 }, { "epoch": 0.6902454780361758, "grad_norm": 193.38230895996094, "learning_rate": 3.0975452196382434e-06, "loss": 0.8877, "step": 4274 }, { "epoch": 0.690406976744186, "grad_norm": 92.52515411376953, "learning_rate": 3.09593023255814e-06, "loss": 0.7072, "step": 4275 }, { "epoch": 0.6905684754521964, "grad_norm": 282.1804504394531, "learning_rate": 3.094315245478036e-06, "loss": 2.7422, "step": 4276 }, { "epoch": 0.6907299741602068, "grad_norm": 244.97508239746094, "learning_rate": 3.092700258397933e-06, "loss": 4.2471, "step": 4277 }, { "epoch": 0.6908914728682171, "grad_norm": 15.337565422058105, "learning_rate": 3.0910852713178296e-06, "loss": 0.093, "step": 4278 }, { "epoch": 0.6910529715762274, "grad_norm": 27.62666130065918, "learning_rate": 3.089470284237726e-06, "loss": 0.2097, "step": 4279 }, { "epoch": 0.6912144702842378, "grad_norm": 35.762481689453125, "learning_rate": 3.087855297157623e-06, "loss": 0.2158, "step": 4280 }, { "epoch": 0.6913759689922481, "grad_norm": 47.76983642578125, "learning_rate": 3.0862403100775197e-06, "loss": 0.5883, "step": 4281 }, { "epoch": 0.6915374677002584, "grad_norm": 35.64093780517578, "learning_rate": 3.084625322997416e-06, "loss": 0.2316, "step": 4282 }, { "epoch": 0.6916989664082688, "grad_norm": 149.04689025878906, "learning_rate": 3.083010335917313e-06, "loss": 1.1409, "step": 4283 }, { "epoch": 0.6918604651162791, "grad_norm": 34.48567581176758, "learning_rate": 3.0813953488372094e-06, "loss": 0.2302, "step": 4284 }, { "epoch": 0.6920219638242894, "grad_norm": 35.54165267944336, "learning_rate": 3.079780361757106e-06, "loss": 0.0922, "step": 4285 }, { "epoch": 0.6921834625322998, "grad_norm": 18.638687133789062, "learning_rate": 3.078165374677003e-06, "loss": 0.0829, "step": 4286 }, { "epoch": 0.6923449612403101, "grad_norm": 76.75836944580078, "learning_rate": 3.0765503875968995e-06, "loss": 0.4219, "step": 4287 }, { "epoch": 0.6925064599483204, "grad_norm": 49.297306060791016, "learning_rate": 3.074935400516796e-06, "loss": 0.1913, "step": 4288 }, { "epoch": 0.6926679586563308, "grad_norm": 69.4686508178711, "learning_rate": 3.073320413436693e-06, "loss": 0.4229, "step": 4289 }, { "epoch": 0.6928294573643411, "grad_norm": 14.871185302734375, "learning_rate": 3.071705426356589e-06, "loss": 0.0841, "step": 4290 }, { "epoch": 0.6929909560723514, "grad_norm": 7.020259857177734, "learning_rate": 3.070090439276486e-06, "loss": 0.0334, "step": 4291 }, { "epoch": 0.6931524547803618, "grad_norm": 34.840797424316406, "learning_rate": 3.0684754521963827e-06, "loss": 0.1521, "step": 4292 }, { "epoch": 0.6933139534883721, "grad_norm": 19.114307403564453, "learning_rate": 3.0668604651162793e-06, "loss": 0.0689, "step": 4293 }, { "epoch": 0.6934754521963824, "grad_norm": 111.00659942626953, "learning_rate": 3.0652454780361763e-06, "loss": 0.5273, "step": 4294 }, { "epoch": 0.6936369509043928, "grad_norm": 61.414634704589844, "learning_rate": 3.063630490956073e-06, "loss": 0.3892, "step": 4295 }, { "epoch": 0.6937984496124031, "grad_norm": 23.858858108520508, "learning_rate": 3.062015503875969e-06, "loss": 0.0759, "step": 4296 }, { "epoch": 0.6939599483204134, "grad_norm": 196.48590087890625, "learning_rate": 3.060400516795866e-06, "loss": 3.0332, "step": 4297 }, { "epoch": 0.6941214470284238, "grad_norm": 1.7059434652328491, "learning_rate": 3.0587855297157625e-06, "loss": 0.0057, "step": 4298 }, { "epoch": 0.6942829457364341, "grad_norm": 13.128500938415527, "learning_rate": 3.057170542635659e-06, "loss": 0.0608, "step": 4299 }, { "epoch": 0.6944444444444444, "grad_norm": 89.0051040649414, "learning_rate": 3.055555555555556e-06, "loss": 0.5121, "step": 4300 }, { "epoch": 0.6946059431524548, "grad_norm": 38.931453704833984, "learning_rate": 3.0539405684754526e-06, "loss": 0.2007, "step": 4301 }, { "epoch": 0.6947674418604651, "grad_norm": 106.29132843017578, "learning_rate": 3.052325581395349e-06, "loss": 1.1094, "step": 4302 }, { "epoch": 0.6949289405684754, "grad_norm": 71.260986328125, "learning_rate": 3.050710594315246e-06, "loss": 0.4072, "step": 4303 }, { "epoch": 0.6950904392764858, "grad_norm": 162.3622589111328, "learning_rate": 3.0490956072351423e-06, "loss": 0.6847, "step": 4304 }, { "epoch": 0.6952519379844961, "grad_norm": 70.7803955078125, "learning_rate": 3.047480620155039e-06, "loss": 0.5486, "step": 4305 }, { "epoch": 0.6954134366925064, "grad_norm": 170.8458251953125, "learning_rate": 3.045865633074936e-06, "loss": 1.2849, "step": 4306 }, { "epoch": 0.6955749354005168, "grad_norm": 157.71038818359375, "learning_rate": 3.0442506459948324e-06, "loss": 0.6567, "step": 4307 }, { "epoch": 0.6957364341085271, "grad_norm": 267.8586120605469, "learning_rate": 3.042635658914729e-06, "loss": 3.0647, "step": 4308 }, { "epoch": 0.6958979328165374, "grad_norm": 41.73794174194336, "learning_rate": 3.041020671834626e-06, "loss": 0.0898, "step": 4309 }, { "epoch": 0.6960594315245479, "grad_norm": 5.889359951019287, "learning_rate": 3.039405684754522e-06, "loss": 0.025, "step": 4310 }, { "epoch": 0.6962209302325582, "grad_norm": 155.33326721191406, "learning_rate": 3.0377906976744186e-06, "loss": 1.0685, "step": 4311 }, { "epoch": 0.6963824289405685, "grad_norm": 16.566730499267578, "learning_rate": 3.0361757105943156e-06, "loss": 0.0801, "step": 4312 }, { "epoch": 0.6965439276485789, "grad_norm": 63.626304626464844, "learning_rate": 3.034560723514212e-06, "loss": 0.2161, "step": 4313 }, { "epoch": 0.6967054263565892, "grad_norm": 45.45939636230469, "learning_rate": 3.0329457364341087e-06, "loss": 0.1157, "step": 4314 }, { "epoch": 0.6968669250645995, "grad_norm": 87.69593811035156, "learning_rate": 3.0313307493540057e-06, "loss": 0.3914, "step": 4315 }, { "epoch": 0.6970284237726099, "grad_norm": 19.231761932373047, "learning_rate": 3.0297157622739023e-06, "loss": 0.067, "step": 4316 }, { "epoch": 0.6971899224806202, "grad_norm": 66.109130859375, "learning_rate": 3.0281007751937984e-06, "loss": 0.2441, "step": 4317 }, { "epoch": 0.6973514211886305, "grad_norm": 94.11692810058594, "learning_rate": 3.0264857881136954e-06, "loss": 0.7734, "step": 4318 }, { "epoch": 0.6975129198966409, "grad_norm": 47.68772888183594, "learning_rate": 3.024870801033592e-06, "loss": 0.3647, "step": 4319 }, { "epoch": 0.6976744186046512, "grad_norm": 8.07285213470459, "learning_rate": 3.0232558139534885e-06, "loss": 0.0321, "step": 4320 }, { "epoch": 0.6978359173126615, "grad_norm": 288.4978942871094, "learning_rate": 3.0216408268733855e-06, "loss": 2.4456, "step": 4321 }, { "epoch": 0.6979974160206718, "grad_norm": 8.563328742980957, "learning_rate": 3.020025839793282e-06, "loss": 0.0247, "step": 4322 }, { "epoch": 0.6981589147286822, "grad_norm": 79.41490936279297, "learning_rate": 3.018410852713178e-06, "loss": 0.3413, "step": 4323 }, { "epoch": 0.6983204134366925, "grad_norm": 56.999393463134766, "learning_rate": 3.016795865633075e-06, "loss": 0.3201, "step": 4324 }, { "epoch": 0.6984819121447028, "grad_norm": 29.833532333374023, "learning_rate": 3.0151808785529717e-06, "loss": 0.2495, "step": 4325 }, { "epoch": 0.6986434108527132, "grad_norm": 98.08021545410156, "learning_rate": 3.0135658914728683e-06, "loss": 0.9165, "step": 4326 }, { "epoch": 0.6988049095607235, "grad_norm": 98.63410186767578, "learning_rate": 3.0119509043927653e-06, "loss": 0.3965, "step": 4327 }, { "epoch": 0.6989664082687338, "grad_norm": 161.39805603027344, "learning_rate": 3.010335917312662e-06, "loss": 1.5977, "step": 4328 }, { "epoch": 0.6991279069767442, "grad_norm": 107.70547485351562, "learning_rate": 3.008720930232558e-06, "loss": 0.6328, "step": 4329 }, { "epoch": 0.6992894056847545, "grad_norm": 161.40333557128906, "learning_rate": 3.0071059431524554e-06, "loss": 0.7031, "step": 4330 }, { "epoch": 0.6994509043927648, "grad_norm": 77.22137451171875, "learning_rate": 3.0054909560723515e-06, "loss": 0.3384, "step": 4331 }, { "epoch": 0.6996124031007752, "grad_norm": 29.572914123535156, "learning_rate": 3.003875968992248e-06, "loss": 0.1185, "step": 4332 }, { "epoch": 0.6997739018087855, "grad_norm": 290.2196044921875, "learning_rate": 3.002260981912145e-06, "loss": 1.0518, "step": 4333 }, { "epoch": 0.6999354005167958, "grad_norm": 453.12384033203125, "learning_rate": 3.0006459948320416e-06, "loss": 0.8603, "step": 4334 }, { "epoch": 0.7000968992248062, "grad_norm": 23.14675521850586, "learning_rate": 2.999031007751938e-06, "loss": 0.1372, "step": 4335 }, { "epoch": 0.7002583979328165, "grad_norm": 229.93023681640625, "learning_rate": 2.997416020671835e-06, "loss": 2.4894, "step": 4336 }, { "epoch": 0.7004198966408268, "grad_norm": 18.16876220703125, "learning_rate": 2.9958010335917313e-06, "loss": 0.0569, "step": 4337 }, { "epoch": 0.7005813953488372, "grad_norm": 519.5552978515625, "learning_rate": 2.994186046511628e-06, "loss": 1.8848, "step": 4338 }, { "epoch": 0.7007428940568475, "grad_norm": 13.604511260986328, "learning_rate": 2.992571059431525e-06, "loss": 0.0507, "step": 4339 }, { "epoch": 0.7009043927648578, "grad_norm": 108.14942932128906, "learning_rate": 2.9909560723514214e-06, "loss": 0.6675, "step": 4340 }, { "epoch": 0.7010658914728682, "grad_norm": 33.668148040771484, "learning_rate": 2.989341085271318e-06, "loss": 0.1332, "step": 4341 }, { "epoch": 0.7012273901808785, "grad_norm": 18.20671272277832, "learning_rate": 2.987726098191215e-06, "loss": 0.0981, "step": 4342 }, { "epoch": 0.7013888888888888, "grad_norm": 23.285160064697266, "learning_rate": 2.986111111111111e-06, "loss": 0.1409, "step": 4343 }, { "epoch": 0.7015503875968992, "grad_norm": 42.026004791259766, "learning_rate": 2.9844961240310076e-06, "loss": 0.2852, "step": 4344 }, { "epoch": 0.7017118863049095, "grad_norm": 60.68056869506836, "learning_rate": 2.9828811369509046e-06, "loss": 0.2324, "step": 4345 }, { "epoch": 0.7018733850129198, "grad_norm": 12.298482894897461, "learning_rate": 2.981266149870801e-06, "loss": 0.025, "step": 4346 }, { "epoch": 0.7020348837209303, "grad_norm": 19.269502639770508, "learning_rate": 2.9796511627906977e-06, "loss": 0.0602, "step": 4347 }, { "epoch": 0.7021963824289406, "grad_norm": 26.552886962890625, "learning_rate": 2.9780361757105947e-06, "loss": 0.0884, "step": 4348 }, { "epoch": 0.7023578811369509, "grad_norm": 19.407907485961914, "learning_rate": 2.9764211886304913e-06, "loss": 0.0802, "step": 4349 }, { "epoch": 0.7025193798449613, "grad_norm": 194.30409240722656, "learning_rate": 2.9748062015503874e-06, "loss": 0.6216, "step": 4350 }, { "epoch": 0.7026808785529716, "grad_norm": 59.734031677246094, "learning_rate": 2.9731912144702844e-06, "loss": 0.355, "step": 4351 }, { "epoch": 0.7028423772609819, "grad_norm": 128.62384033203125, "learning_rate": 2.971576227390181e-06, "loss": 0.6467, "step": 4352 }, { "epoch": 0.7030038759689923, "grad_norm": 23.18964195251465, "learning_rate": 2.9699612403100775e-06, "loss": 0.1167, "step": 4353 }, { "epoch": 0.7031653746770026, "grad_norm": 117.10677337646484, "learning_rate": 2.9683462532299745e-06, "loss": 0.7842, "step": 4354 }, { "epoch": 0.7033268733850129, "grad_norm": 30.127517700195312, "learning_rate": 2.966731266149871e-06, "loss": 0.0925, "step": 4355 }, { "epoch": 0.7034883720930233, "grad_norm": 29.920793533325195, "learning_rate": 2.965116279069768e-06, "loss": 0.168, "step": 4356 }, { "epoch": 0.7036498708010336, "grad_norm": 25.43878746032715, "learning_rate": 2.9635012919896646e-06, "loss": 0.1029, "step": 4357 }, { "epoch": 0.7038113695090439, "grad_norm": 24.267250061035156, "learning_rate": 2.9618863049095608e-06, "loss": 0.0859, "step": 4358 }, { "epoch": 0.7039728682170543, "grad_norm": 430.69390869140625, "learning_rate": 2.9602713178294577e-06, "loss": 4.4375, "step": 4359 }, { "epoch": 0.7041343669250646, "grad_norm": 28.100357055664062, "learning_rate": 2.9586563307493543e-06, "loss": 0.1389, "step": 4360 }, { "epoch": 0.7042958656330749, "grad_norm": 93.97061920166016, "learning_rate": 2.957041343669251e-06, "loss": 0.218, "step": 4361 }, { "epoch": 0.7044573643410853, "grad_norm": 52.822059631347656, "learning_rate": 2.955426356589148e-06, "loss": 0.2389, "step": 4362 }, { "epoch": 0.7046188630490956, "grad_norm": 11.755165100097656, "learning_rate": 2.9538113695090444e-06, "loss": 0.0297, "step": 4363 }, { "epoch": 0.7047803617571059, "grad_norm": 71.40790557861328, "learning_rate": 2.9521963824289405e-06, "loss": 0.363, "step": 4364 }, { "epoch": 0.7049418604651163, "grad_norm": 147.75381469726562, "learning_rate": 2.9505813953488375e-06, "loss": 1.5176, "step": 4365 }, { "epoch": 0.7051033591731266, "grad_norm": 30.314319610595703, "learning_rate": 2.948966408268734e-06, "loss": 0.1238, "step": 4366 }, { "epoch": 0.7052648578811369, "grad_norm": 18.31490707397461, "learning_rate": 2.9473514211886306e-06, "loss": 0.0485, "step": 4367 }, { "epoch": 0.7054263565891473, "grad_norm": 119.41642761230469, "learning_rate": 2.9457364341085276e-06, "loss": 0.9619, "step": 4368 }, { "epoch": 0.7055878552971576, "grad_norm": 7.383219242095947, "learning_rate": 2.944121447028424e-06, "loss": 0.0269, "step": 4369 }, { "epoch": 0.7057493540051679, "grad_norm": 22.46985626220703, "learning_rate": 2.9425064599483203e-06, "loss": 0.1373, "step": 4370 }, { "epoch": 0.7059108527131783, "grad_norm": 62.880699157714844, "learning_rate": 2.9408914728682177e-06, "loss": 0.4209, "step": 4371 }, { "epoch": 0.7060723514211886, "grad_norm": 81.2223129272461, "learning_rate": 2.939276485788114e-06, "loss": 0.8174, "step": 4372 }, { "epoch": 0.7062338501291989, "grad_norm": 86.22067260742188, "learning_rate": 2.9376614987080104e-06, "loss": 0.4812, "step": 4373 }, { "epoch": 0.7063953488372093, "grad_norm": 61.65596008300781, "learning_rate": 2.9360465116279074e-06, "loss": 0.1787, "step": 4374 }, { "epoch": 0.7065568475452196, "grad_norm": 291.08807373046875, "learning_rate": 2.934431524547804e-06, "loss": 3.0195, "step": 4375 }, { "epoch": 0.7067183462532299, "grad_norm": 11.534256935119629, "learning_rate": 2.9328165374677005e-06, "loss": 0.0555, "step": 4376 }, { "epoch": 0.7068798449612403, "grad_norm": 92.5316162109375, "learning_rate": 2.9312015503875975e-06, "loss": 0.4297, "step": 4377 }, { "epoch": 0.7070413436692506, "grad_norm": 3.5349724292755127, "learning_rate": 2.9295865633074936e-06, "loss": 0.0228, "step": 4378 }, { "epoch": 0.7072028423772609, "grad_norm": 99.6095199584961, "learning_rate": 2.92797157622739e-06, "loss": 0.1711, "step": 4379 }, { "epoch": 0.7073643410852714, "grad_norm": 69.36312103271484, "learning_rate": 2.926356589147287e-06, "loss": 0.5566, "step": 4380 }, { "epoch": 0.7075258397932817, "grad_norm": 188.49032592773438, "learning_rate": 2.9247416020671838e-06, "loss": 0.9102, "step": 4381 }, { "epoch": 0.707687338501292, "grad_norm": 89.35177612304688, "learning_rate": 2.9231266149870803e-06, "loss": 0.9546, "step": 4382 }, { "epoch": 0.7078488372093024, "grad_norm": 24.381671905517578, "learning_rate": 2.9215116279069773e-06, "loss": 0.0761, "step": 4383 }, { "epoch": 0.7080103359173127, "grad_norm": 18.496774673461914, "learning_rate": 2.9198966408268734e-06, "loss": 0.0711, "step": 4384 }, { "epoch": 0.708171834625323, "grad_norm": 58.3271484375, "learning_rate": 2.91828165374677e-06, "loss": 0.272, "step": 4385 }, { "epoch": 0.7083333333333334, "grad_norm": 22.007232666015625, "learning_rate": 2.916666666666667e-06, "loss": 0.0724, "step": 4386 }, { "epoch": 0.7084948320413437, "grad_norm": 381.25738525390625, "learning_rate": 2.9150516795865635e-06, "loss": 2.8934, "step": 4387 }, { "epoch": 0.708656330749354, "grad_norm": 308.6488952636719, "learning_rate": 2.91343669250646e-06, "loss": 2.1836, "step": 4388 }, { "epoch": 0.7088178294573644, "grad_norm": 227.53900146484375, "learning_rate": 2.911821705426357e-06, "loss": 1.9006, "step": 4389 }, { "epoch": 0.7089793281653747, "grad_norm": 106.8840560913086, "learning_rate": 2.9102067183462536e-06, "loss": 0.3579, "step": 4390 }, { "epoch": 0.709140826873385, "grad_norm": 168.0601806640625, "learning_rate": 2.9085917312661498e-06, "loss": 1.4971, "step": 4391 }, { "epoch": 0.7093023255813954, "grad_norm": 231.19326782226562, "learning_rate": 2.9069767441860468e-06, "loss": 1.1961, "step": 4392 }, { "epoch": 0.7094638242894057, "grad_norm": 297.70672607421875, "learning_rate": 2.9053617571059433e-06, "loss": 0.5767, "step": 4393 }, { "epoch": 0.709625322997416, "grad_norm": 38.17879867553711, "learning_rate": 2.90374677002584e-06, "loss": 0.186, "step": 4394 }, { "epoch": 0.7097868217054264, "grad_norm": 7.0714263916015625, "learning_rate": 2.902131782945737e-06, "loss": 0.0267, "step": 4395 }, { "epoch": 0.7099483204134367, "grad_norm": 67.23233032226562, "learning_rate": 2.9005167958656334e-06, "loss": 0.3906, "step": 4396 }, { "epoch": 0.710109819121447, "grad_norm": 139.89419555664062, "learning_rate": 2.8989018087855296e-06, "loss": 1.4141, "step": 4397 }, { "epoch": 0.7102713178294574, "grad_norm": 11.643963813781738, "learning_rate": 2.8972868217054265e-06, "loss": 0.0392, "step": 4398 }, { "epoch": 0.7104328165374677, "grad_norm": 296.3921813964844, "learning_rate": 2.895671834625323e-06, "loss": 3.2321, "step": 4399 }, { "epoch": 0.710594315245478, "grad_norm": 90.9906234741211, "learning_rate": 2.8940568475452197e-06, "loss": 0.2573, "step": 4400 }, { "epoch": 0.7107558139534884, "grad_norm": 49.47856903076172, "learning_rate": 2.8924418604651166e-06, "loss": 0.1941, "step": 4401 }, { "epoch": 0.7109173126614987, "grad_norm": 34.82453536987305, "learning_rate": 2.890826873385013e-06, "loss": 0.1052, "step": 4402 }, { "epoch": 0.711078811369509, "grad_norm": 243.3920135498047, "learning_rate": 2.8892118863049098e-06, "loss": 2.8828, "step": 4403 }, { "epoch": 0.7112403100775194, "grad_norm": 17.40386390686035, "learning_rate": 2.8875968992248067e-06, "loss": 0.0912, "step": 4404 }, { "epoch": 0.7114018087855297, "grad_norm": 462.8196105957031, "learning_rate": 2.885981912144703e-06, "loss": 0.709, "step": 4405 }, { "epoch": 0.71156330749354, "grad_norm": 7.5483174324035645, "learning_rate": 2.8843669250645994e-06, "loss": 0.0278, "step": 4406 }, { "epoch": 0.7117248062015504, "grad_norm": 53.27341079711914, "learning_rate": 2.8827519379844964e-06, "loss": 0.2596, "step": 4407 }, { "epoch": 0.7118863049095607, "grad_norm": 15.988765716552734, "learning_rate": 2.881136950904393e-06, "loss": 0.0273, "step": 4408 }, { "epoch": 0.712047803617571, "grad_norm": 45.0598258972168, "learning_rate": 2.8795219638242895e-06, "loss": 0.3071, "step": 4409 }, { "epoch": 0.7122093023255814, "grad_norm": 253.65484619140625, "learning_rate": 2.8779069767441865e-06, "loss": 0.8867, "step": 4410 }, { "epoch": 0.7123708010335917, "grad_norm": 52.518310546875, "learning_rate": 2.8762919896640827e-06, "loss": 0.2699, "step": 4411 }, { "epoch": 0.712532299741602, "grad_norm": 87.06906127929688, "learning_rate": 2.8746770025839792e-06, "loss": 0.4714, "step": 4412 }, { "epoch": 0.7126937984496124, "grad_norm": 21.88828468322754, "learning_rate": 2.8730620155038762e-06, "loss": 0.0476, "step": 4413 }, { "epoch": 0.7128552971576227, "grad_norm": 160.96458435058594, "learning_rate": 2.8714470284237728e-06, "loss": 1.63, "step": 4414 }, { "epoch": 0.713016795865633, "grad_norm": 215.90670776367188, "learning_rate": 2.8698320413436693e-06, "loss": 1.8281, "step": 4415 }, { "epoch": 0.7131782945736435, "grad_norm": 26.23227882385254, "learning_rate": 2.8682170542635663e-06, "loss": 0.0726, "step": 4416 }, { "epoch": 0.7133397932816538, "grad_norm": 56.609825134277344, "learning_rate": 2.866602067183463e-06, "loss": 0.7947, "step": 4417 }, { "epoch": 0.713501291989664, "grad_norm": 63.92579650878906, "learning_rate": 2.864987080103359e-06, "loss": 0.365, "step": 4418 }, { "epoch": 0.7136627906976745, "grad_norm": 174.55337524414062, "learning_rate": 2.863372093023256e-06, "loss": 1.7725, "step": 4419 }, { "epoch": 0.7138242894056848, "grad_norm": 57.821529388427734, "learning_rate": 2.8617571059431526e-06, "loss": 0.3857, "step": 4420 }, { "epoch": 0.7139857881136951, "grad_norm": 3.356942653656006, "learning_rate": 2.8601421188630495e-06, "loss": 0.0174, "step": 4421 }, { "epoch": 0.7141472868217055, "grad_norm": 103.34500122070312, "learning_rate": 2.858527131782946e-06, "loss": 1.1313, "step": 4422 }, { "epoch": 0.7143087855297158, "grad_norm": 53.83638000488281, "learning_rate": 2.8569121447028427e-06, "loss": 0.4204, "step": 4423 }, { "epoch": 0.7144702842377261, "grad_norm": 37.2034797668457, "learning_rate": 2.8552971576227396e-06, "loss": 0.2424, "step": 4424 }, { "epoch": 0.7146317829457365, "grad_norm": 56.60368347167969, "learning_rate": 2.8536821705426358e-06, "loss": 0.1765, "step": 4425 }, { "epoch": 0.7147932816537468, "grad_norm": 16.4691219329834, "learning_rate": 2.8520671834625323e-06, "loss": 0.0785, "step": 4426 }, { "epoch": 0.7149547803617571, "grad_norm": 160.1484375, "learning_rate": 2.8504521963824293e-06, "loss": 1.6631, "step": 4427 }, { "epoch": 0.7151162790697675, "grad_norm": 103.40705108642578, "learning_rate": 2.848837209302326e-06, "loss": 1.1562, "step": 4428 }, { "epoch": 0.7152777777777778, "grad_norm": 23.452144622802734, "learning_rate": 2.8472222222222224e-06, "loss": 0.0826, "step": 4429 }, { "epoch": 0.7154392764857881, "grad_norm": 108.86074829101562, "learning_rate": 2.8456072351421194e-06, "loss": 0.6797, "step": 4430 }, { "epoch": 0.7156007751937985, "grad_norm": 42.815208435058594, "learning_rate": 2.843992248062016e-06, "loss": 0.3062, "step": 4431 }, { "epoch": 0.7157622739018088, "grad_norm": 101.0250473022461, "learning_rate": 2.842377260981912e-06, "loss": 0.692, "step": 4432 }, { "epoch": 0.7159237726098191, "grad_norm": 50.2115478515625, "learning_rate": 2.840762273901809e-06, "loss": 0.2344, "step": 4433 }, { "epoch": 0.7160852713178295, "grad_norm": 53.74542999267578, "learning_rate": 2.8391472868217057e-06, "loss": 0.544, "step": 4434 }, { "epoch": 0.7162467700258398, "grad_norm": 55.10050582885742, "learning_rate": 2.8375322997416022e-06, "loss": 0.4929, "step": 4435 }, { "epoch": 0.7164082687338501, "grad_norm": 13.150368690490723, "learning_rate": 2.835917312661499e-06, "loss": 0.0642, "step": 4436 }, { "epoch": 0.7165697674418605, "grad_norm": 4.595853328704834, "learning_rate": 2.8343023255813958e-06, "loss": 0.0267, "step": 4437 }, { "epoch": 0.7167312661498708, "grad_norm": 70.97452545166016, "learning_rate": 2.832687338501292e-06, "loss": 0.752, "step": 4438 }, { "epoch": 0.7168927648578811, "grad_norm": 97.20576477050781, "learning_rate": 2.831072351421189e-06, "loss": 0.999, "step": 4439 }, { "epoch": 0.7170542635658915, "grad_norm": 133.35684204101562, "learning_rate": 2.8294573643410855e-06, "loss": 1.5603, "step": 4440 }, { "epoch": 0.7172157622739018, "grad_norm": 55.832820892333984, "learning_rate": 2.827842377260982e-06, "loss": 0.2172, "step": 4441 }, { "epoch": 0.7173772609819121, "grad_norm": 56.67898941040039, "learning_rate": 2.826227390180879e-06, "loss": 0.3811, "step": 4442 }, { "epoch": 0.7175387596899225, "grad_norm": 17.655900955200195, "learning_rate": 2.8246124031007756e-06, "loss": 0.0424, "step": 4443 }, { "epoch": 0.7177002583979328, "grad_norm": 184.95623779296875, "learning_rate": 2.8229974160206717e-06, "loss": 1.7485, "step": 4444 }, { "epoch": 0.7178617571059431, "grad_norm": 23.694345474243164, "learning_rate": 2.821382428940569e-06, "loss": 0.1445, "step": 4445 }, { "epoch": 0.7180232558139535, "grad_norm": 94.63401794433594, "learning_rate": 2.8197674418604652e-06, "loss": 0.6133, "step": 4446 }, { "epoch": 0.7181847545219638, "grad_norm": 12.953531265258789, "learning_rate": 2.818152454780362e-06, "loss": 0.0681, "step": 4447 }, { "epoch": 0.7183462532299741, "grad_norm": 10.419922828674316, "learning_rate": 2.8165374677002588e-06, "loss": 0.0517, "step": 4448 }, { "epoch": 0.7185077519379846, "grad_norm": 18.070222854614258, "learning_rate": 2.8149224806201553e-06, "loss": 0.0807, "step": 4449 }, { "epoch": 0.7186692506459949, "grad_norm": 74.12086486816406, "learning_rate": 2.813307493540052e-06, "loss": 0.5688, "step": 4450 }, { "epoch": 0.7188307493540051, "grad_norm": 48.677730560302734, "learning_rate": 2.811692506459949e-06, "loss": 0.1897, "step": 4451 }, { "epoch": 0.7189922480620154, "grad_norm": 153.01962280273438, "learning_rate": 2.810077519379845e-06, "loss": 0.6368, "step": 4452 }, { "epoch": 0.7191537467700259, "grad_norm": 202.3713836669922, "learning_rate": 2.8084625322997416e-06, "loss": 1.8628, "step": 4453 }, { "epoch": 0.7193152454780362, "grad_norm": 176.05540466308594, "learning_rate": 2.8068475452196386e-06, "loss": 0.8408, "step": 4454 }, { "epoch": 0.7194767441860465, "grad_norm": 9.239829063415527, "learning_rate": 2.805232558139535e-06, "loss": 0.0214, "step": 4455 }, { "epoch": 0.7196382428940569, "grad_norm": 34.63663101196289, "learning_rate": 2.8036175710594317e-06, "loss": 0.1267, "step": 4456 }, { "epoch": 0.7197997416020672, "grad_norm": 16.200159072875977, "learning_rate": 2.8020025839793287e-06, "loss": 0.1035, "step": 4457 }, { "epoch": 0.7199612403100775, "grad_norm": 304.7335205078125, "learning_rate": 2.8003875968992252e-06, "loss": 1.4926, "step": 4458 }, { "epoch": 0.7201227390180879, "grad_norm": 77.24150085449219, "learning_rate": 2.7987726098191214e-06, "loss": 0.1824, "step": 4459 }, { "epoch": 0.7202842377260982, "grad_norm": 68.31204223632812, "learning_rate": 2.7971576227390183e-06, "loss": 0.6514, "step": 4460 }, { "epoch": 0.7204457364341085, "grad_norm": 267.4267272949219, "learning_rate": 2.795542635658915e-06, "loss": 1.6943, "step": 4461 }, { "epoch": 0.7206072351421189, "grad_norm": 23.031530380249023, "learning_rate": 2.7939276485788115e-06, "loss": 0.0798, "step": 4462 }, { "epoch": 0.7207687338501292, "grad_norm": 110.31086730957031, "learning_rate": 2.7923126614987084e-06, "loss": 0.9438, "step": 4463 }, { "epoch": 0.7209302325581395, "grad_norm": 27.4280948638916, "learning_rate": 2.790697674418605e-06, "loss": 0.2185, "step": 4464 }, { "epoch": 0.7210917312661499, "grad_norm": 163.7159423828125, "learning_rate": 2.789082687338501e-06, "loss": 1.3052, "step": 4465 }, { "epoch": 0.7212532299741602, "grad_norm": 61.83344650268555, "learning_rate": 2.787467700258398e-06, "loss": 0.2249, "step": 4466 }, { "epoch": 0.7214147286821705, "grad_norm": 203.45484924316406, "learning_rate": 2.7858527131782947e-06, "loss": 2.2617, "step": 4467 }, { "epoch": 0.7215762273901809, "grad_norm": 1.4778090715408325, "learning_rate": 2.7842377260981912e-06, "loss": 0.0088, "step": 4468 }, { "epoch": 0.7217377260981912, "grad_norm": 7.532685279846191, "learning_rate": 2.7826227390180882e-06, "loss": 0.026, "step": 4469 }, { "epoch": 0.7218992248062015, "grad_norm": 57.24382400512695, "learning_rate": 2.781007751937985e-06, "loss": 0.3992, "step": 4470 }, { "epoch": 0.7220607235142119, "grad_norm": 10.39846420288086, "learning_rate": 2.779392764857881e-06, "loss": 0.0482, "step": 4471 }, { "epoch": 0.7222222222222222, "grad_norm": 36.69139099121094, "learning_rate": 2.7777777777777783e-06, "loss": 0.1221, "step": 4472 }, { "epoch": 0.7223837209302325, "grad_norm": 61.501277923583984, "learning_rate": 2.7761627906976745e-06, "loss": 0.2007, "step": 4473 }, { "epoch": 0.7225452196382429, "grad_norm": 33.74900817871094, "learning_rate": 2.774547803617571e-06, "loss": 0.1448, "step": 4474 }, { "epoch": 0.7227067183462532, "grad_norm": 40.227684020996094, "learning_rate": 2.772932816537468e-06, "loss": 0.2236, "step": 4475 }, { "epoch": 0.7228682170542635, "grad_norm": 242.88571166992188, "learning_rate": 2.7713178294573646e-06, "loss": 0.9297, "step": 4476 }, { "epoch": 0.7230297157622739, "grad_norm": 33.36329650878906, "learning_rate": 2.769702842377261e-06, "loss": 0.1251, "step": 4477 }, { "epoch": 0.7231912144702842, "grad_norm": 150.80714416503906, "learning_rate": 2.768087855297158e-06, "loss": 0.9258, "step": 4478 }, { "epoch": 0.7233527131782945, "grad_norm": 63.5515251159668, "learning_rate": 2.7664728682170543e-06, "loss": 0.2954, "step": 4479 }, { "epoch": 0.7235142118863049, "grad_norm": 82.65245819091797, "learning_rate": 2.764857881136951e-06, "loss": 0.6222, "step": 4480 }, { "epoch": 0.7236757105943152, "grad_norm": 117.69886016845703, "learning_rate": 2.763242894056848e-06, "loss": 0.4385, "step": 4481 }, { "epoch": 0.7238372093023255, "grad_norm": 145.64663696289062, "learning_rate": 2.7616279069767444e-06, "loss": 2.0117, "step": 4482 }, { "epoch": 0.723998708010336, "grad_norm": 45.07558059692383, "learning_rate": 2.7600129198966413e-06, "loss": 0.2798, "step": 4483 }, { "epoch": 0.7241602067183462, "grad_norm": 6.470263481140137, "learning_rate": 2.758397932816538e-06, "loss": 0.028, "step": 4484 }, { "epoch": 0.7243217054263565, "grad_norm": 136.0615997314453, "learning_rate": 2.756782945736434e-06, "loss": 1.1807, "step": 4485 }, { "epoch": 0.724483204134367, "grad_norm": 18.711217880249023, "learning_rate": 2.7551679586563314e-06, "loss": 0.0518, "step": 4486 }, { "epoch": 0.7246447028423773, "grad_norm": 44.361968994140625, "learning_rate": 2.7535529715762276e-06, "loss": 0.1675, "step": 4487 }, { "epoch": 0.7248062015503876, "grad_norm": 20.753856658935547, "learning_rate": 2.751937984496124e-06, "loss": 0.0625, "step": 4488 }, { "epoch": 0.724967700258398, "grad_norm": 17.52935218811035, "learning_rate": 2.750322997416021e-06, "loss": 0.0759, "step": 4489 }, { "epoch": 0.7251291989664083, "grad_norm": 65.38661193847656, "learning_rate": 2.7487080103359177e-06, "loss": 0.7197, "step": 4490 }, { "epoch": 0.7252906976744186, "grad_norm": 14.677460670471191, "learning_rate": 2.7470930232558142e-06, "loss": 0.1283, "step": 4491 }, { "epoch": 0.725452196382429, "grad_norm": 7.757281303405762, "learning_rate": 2.7454780361757112e-06, "loss": 0.0275, "step": 4492 }, { "epoch": 0.7256136950904393, "grad_norm": 124.87124633789062, "learning_rate": 2.7438630490956074e-06, "loss": 0.6797, "step": 4493 }, { "epoch": 0.7257751937984496, "grad_norm": 14.609213829040527, "learning_rate": 2.742248062015504e-06, "loss": 0.1182, "step": 4494 }, { "epoch": 0.72593669250646, "grad_norm": 47.60532760620117, "learning_rate": 2.740633074935401e-06, "loss": 0.2423, "step": 4495 }, { "epoch": 0.7260981912144703, "grad_norm": 147.5388641357422, "learning_rate": 2.7390180878552975e-06, "loss": 1.0459, "step": 4496 }, { "epoch": 0.7262596899224806, "grad_norm": 68.54842376708984, "learning_rate": 2.737403100775194e-06, "loss": 0.2488, "step": 4497 }, { "epoch": 0.726421188630491, "grad_norm": 91.24524688720703, "learning_rate": 2.735788113695091e-06, "loss": 0.49, "step": 4498 }, { "epoch": 0.7265826873385013, "grad_norm": 15.883292198181152, "learning_rate": 2.734173126614987e-06, "loss": 0.0559, "step": 4499 }, { "epoch": 0.7267441860465116, "grad_norm": 48.52610778808594, "learning_rate": 2.7325581395348837e-06, "loss": 0.1558, "step": 4500 }, { "epoch": 0.726905684754522, "grad_norm": 31.211708068847656, "learning_rate": 2.7309431524547807e-06, "loss": 0.1221, "step": 4501 }, { "epoch": 0.7270671834625323, "grad_norm": 10.383772850036621, "learning_rate": 2.7293281653746773e-06, "loss": 0.0269, "step": 4502 }, { "epoch": 0.7272286821705426, "grad_norm": 92.22789001464844, "learning_rate": 2.727713178294574e-06, "loss": 0.6777, "step": 4503 }, { "epoch": 0.727390180878553, "grad_norm": 134.8096466064453, "learning_rate": 2.726098191214471e-06, "loss": 1.1738, "step": 4504 }, { "epoch": 0.7275516795865633, "grad_norm": 182.59521484375, "learning_rate": 2.7244832041343674e-06, "loss": 0.7539, "step": 4505 }, { "epoch": 0.7277131782945736, "grad_norm": 16.694873809814453, "learning_rate": 2.7228682170542635e-06, "loss": 0.0378, "step": 4506 }, { "epoch": 0.727874677002584, "grad_norm": 19.78574562072754, "learning_rate": 2.7212532299741605e-06, "loss": 0.0699, "step": 4507 }, { "epoch": 0.7280361757105943, "grad_norm": 34.920528411865234, "learning_rate": 2.719638242894057e-06, "loss": 0.1123, "step": 4508 }, { "epoch": 0.7281976744186046, "grad_norm": 41.120460510253906, "learning_rate": 2.7180232558139536e-06, "loss": 0.2467, "step": 4509 }, { "epoch": 0.728359173126615, "grad_norm": 120.14308166503906, "learning_rate": 2.7164082687338506e-06, "loss": 0.6646, "step": 4510 }, { "epoch": 0.7285206718346253, "grad_norm": 72.77070617675781, "learning_rate": 2.714793281653747e-06, "loss": 0.3317, "step": 4511 }, { "epoch": 0.7286821705426356, "grad_norm": 282.5746154785156, "learning_rate": 2.7131782945736433e-06, "loss": 1.4653, "step": 4512 }, { "epoch": 0.728843669250646, "grad_norm": 41.06638717651367, "learning_rate": 2.7115633074935403e-06, "loss": 0.1858, "step": 4513 }, { "epoch": 0.7290051679586563, "grad_norm": 29.992237091064453, "learning_rate": 2.709948320413437e-06, "loss": 0.0669, "step": 4514 }, { "epoch": 0.7291666666666666, "grad_norm": 21.191312789916992, "learning_rate": 2.7083333333333334e-06, "loss": 0.0481, "step": 4515 }, { "epoch": 0.729328165374677, "grad_norm": 5.504312038421631, "learning_rate": 2.7067183462532304e-06, "loss": 0.0116, "step": 4516 }, { "epoch": 0.7294896640826873, "grad_norm": 34.66682052612305, "learning_rate": 2.705103359173127e-06, "loss": 0.1141, "step": 4517 }, { "epoch": 0.7296511627906976, "grad_norm": 208.63055419921875, "learning_rate": 2.7034883720930235e-06, "loss": 2.7935, "step": 4518 }, { "epoch": 0.729812661498708, "grad_norm": 262.45721435546875, "learning_rate": 2.7018733850129205e-06, "loss": 0.6407, "step": 4519 }, { "epoch": 0.7299741602067183, "grad_norm": 16.35857391357422, "learning_rate": 2.7002583979328166e-06, "loss": 0.0271, "step": 4520 }, { "epoch": 0.7301356589147286, "grad_norm": 311.4580383300781, "learning_rate": 2.698643410852713e-06, "loss": 2.4321, "step": 4521 }, { "epoch": 0.7302971576227391, "grad_norm": 8.835404396057129, "learning_rate": 2.69702842377261e-06, "loss": 0.0293, "step": 4522 }, { "epoch": 0.7304586563307494, "grad_norm": 136.50563049316406, "learning_rate": 2.6954134366925067e-06, "loss": 0.3757, "step": 4523 }, { "epoch": 0.7306201550387597, "grad_norm": 64.41693115234375, "learning_rate": 2.6937984496124033e-06, "loss": 0.4824, "step": 4524 }, { "epoch": 0.7307816537467701, "grad_norm": 13.159382820129395, "learning_rate": 2.6921834625323002e-06, "loss": 0.0615, "step": 4525 }, { "epoch": 0.7309431524547804, "grad_norm": 30.352121353149414, "learning_rate": 2.6905684754521964e-06, "loss": 0.1042, "step": 4526 }, { "epoch": 0.7311046511627907, "grad_norm": 89.5604019165039, "learning_rate": 2.688953488372093e-06, "loss": 0.6191, "step": 4527 }, { "epoch": 0.7312661498708011, "grad_norm": 26.072490692138672, "learning_rate": 2.68733850129199e-06, "loss": 0.0779, "step": 4528 }, { "epoch": 0.7314276485788114, "grad_norm": 37.660797119140625, "learning_rate": 2.6857235142118865e-06, "loss": 0.2411, "step": 4529 }, { "epoch": 0.7315891472868217, "grad_norm": 109.0886459350586, "learning_rate": 2.684108527131783e-06, "loss": 0.5247, "step": 4530 }, { "epoch": 0.7317506459948321, "grad_norm": 38.32889175415039, "learning_rate": 2.68249354005168e-06, "loss": 0.2891, "step": 4531 }, { "epoch": 0.7319121447028424, "grad_norm": 45.202880859375, "learning_rate": 2.6808785529715766e-06, "loss": 0.1849, "step": 4532 }, { "epoch": 0.7320736434108527, "grad_norm": 50.770931243896484, "learning_rate": 2.6792635658914727e-06, "loss": 0.281, "step": 4533 }, { "epoch": 0.7322351421188631, "grad_norm": 109.72393798828125, "learning_rate": 2.6776485788113697e-06, "loss": 0.5186, "step": 4534 }, { "epoch": 0.7323966408268734, "grad_norm": 24.434659957885742, "learning_rate": 2.6760335917312663e-06, "loss": 0.0935, "step": 4535 }, { "epoch": 0.7325581395348837, "grad_norm": 11.82715892791748, "learning_rate": 2.674418604651163e-06, "loss": 0.0376, "step": 4536 }, { "epoch": 0.7327196382428941, "grad_norm": 113.91094970703125, "learning_rate": 2.67280361757106e-06, "loss": 0.3066, "step": 4537 }, { "epoch": 0.7328811369509044, "grad_norm": 264.6119079589844, "learning_rate": 2.6711886304909564e-06, "loss": 2.8667, "step": 4538 }, { "epoch": 0.7330426356589147, "grad_norm": 165.69223022460938, "learning_rate": 2.6695736434108525e-06, "loss": 0.583, "step": 4539 }, { "epoch": 0.7332041343669251, "grad_norm": 83.13922119140625, "learning_rate": 2.6679586563307495e-06, "loss": 0.4033, "step": 4540 }, { "epoch": 0.7333656330749354, "grad_norm": 55.35585021972656, "learning_rate": 2.666343669250646e-06, "loss": 0.522, "step": 4541 }, { "epoch": 0.7335271317829457, "grad_norm": 23.401704788208008, "learning_rate": 2.6647286821705426e-06, "loss": 0.0845, "step": 4542 }, { "epoch": 0.7336886304909561, "grad_norm": 49.4188232421875, "learning_rate": 2.6631136950904396e-06, "loss": 0.1987, "step": 4543 }, { "epoch": 0.7338501291989664, "grad_norm": 20.64641571044922, "learning_rate": 2.661498708010336e-06, "loss": 0.1125, "step": 4544 }, { "epoch": 0.7340116279069767, "grad_norm": 16.165822982788086, "learning_rate": 2.6598837209302323e-06, "loss": 0.0533, "step": 4545 }, { "epoch": 0.7341731266149871, "grad_norm": 144.7584228515625, "learning_rate": 2.6582687338501297e-06, "loss": 0.6938, "step": 4546 }, { "epoch": 0.7343346253229974, "grad_norm": 94.8850326538086, "learning_rate": 2.656653746770026e-06, "loss": 0.3408, "step": 4547 }, { "epoch": 0.7344961240310077, "grad_norm": 45.92573547363281, "learning_rate": 2.655038759689923e-06, "loss": 0.2329, "step": 4548 }, { "epoch": 0.7346576227390181, "grad_norm": 37.953556060791016, "learning_rate": 2.6534237726098194e-06, "loss": 0.1631, "step": 4549 }, { "epoch": 0.7348191214470284, "grad_norm": 164.74502563476562, "learning_rate": 2.651808785529716e-06, "loss": 0.8672, "step": 4550 }, { "epoch": 0.7349806201550387, "grad_norm": 307.3532409667969, "learning_rate": 2.650193798449613e-06, "loss": 2.0123, "step": 4551 }, { "epoch": 0.7351421188630491, "grad_norm": 48.31825637817383, "learning_rate": 2.6485788113695095e-06, "loss": 0.3044, "step": 4552 }, { "epoch": 0.7353036175710594, "grad_norm": 28.000219345092773, "learning_rate": 2.6469638242894056e-06, "loss": 0.2415, "step": 4553 }, { "epoch": 0.7354651162790697, "grad_norm": 8.281302452087402, "learning_rate": 2.6453488372093026e-06, "loss": 0.0133, "step": 4554 }, { "epoch": 0.7356266149870802, "grad_norm": 86.86200714111328, "learning_rate": 2.643733850129199e-06, "loss": 0.4077, "step": 4555 }, { "epoch": 0.7357881136950905, "grad_norm": 354.5458679199219, "learning_rate": 2.6421188630490957e-06, "loss": 2.3555, "step": 4556 }, { "epoch": 0.7359496124031008, "grad_norm": 32.396156311035156, "learning_rate": 2.6405038759689927e-06, "loss": 0.1638, "step": 4557 }, { "epoch": 0.7361111111111112, "grad_norm": 211.52220153808594, "learning_rate": 2.6388888888888893e-06, "loss": 0.8125, "step": 4558 }, { "epoch": 0.7362726098191215, "grad_norm": 50.451744079589844, "learning_rate": 2.637273901808786e-06, "loss": 0.2188, "step": 4559 }, { "epoch": 0.7364341085271318, "grad_norm": 62.433982849121094, "learning_rate": 2.635658914728683e-06, "loss": 0.7062, "step": 4560 }, { "epoch": 0.7365956072351422, "grad_norm": 12.89474105834961, "learning_rate": 2.634043927648579e-06, "loss": 0.0869, "step": 4561 }, { "epoch": 0.7367571059431525, "grad_norm": 106.73463439941406, "learning_rate": 2.6324289405684755e-06, "loss": 0.6387, "step": 4562 }, { "epoch": 0.7369186046511628, "grad_norm": 47.693363189697266, "learning_rate": 2.6308139534883725e-06, "loss": 0.1953, "step": 4563 }, { "epoch": 0.7370801033591732, "grad_norm": 19.067115783691406, "learning_rate": 2.629198966408269e-06, "loss": 0.0533, "step": 4564 }, { "epoch": 0.7372416020671835, "grad_norm": 66.7677230834961, "learning_rate": 2.6275839793281656e-06, "loss": 0.5503, "step": 4565 }, { "epoch": 0.7374031007751938, "grad_norm": 27.596630096435547, "learning_rate": 2.6259689922480626e-06, "loss": 0.0906, "step": 4566 }, { "epoch": 0.7375645994832042, "grad_norm": 103.34076690673828, "learning_rate": 2.6243540051679587e-06, "loss": 1.0098, "step": 4567 }, { "epoch": 0.7377260981912145, "grad_norm": 10.884857177734375, "learning_rate": 2.6227390180878553e-06, "loss": 0.0285, "step": 4568 }, { "epoch": 0.7378875968992248, "grad_norm": 34.00555419921875, "learning_rate": 2.6211240310077523e-06, "loss": 0.1145, "step": 4569 }, { "epoch": 0.7380490956072352, "grad_norm": 113.58333587646484, "learning_rate": 2.619509043927649e-06, "loss": 0.6394, "step": 4570 }, { "epoch": 0.7382105943152455, "grad_norm": 52.0593147277832, "learning_rate": 2.6178940568475454e-06, "loss": 0.2559, "step": 4571 }, { "epoch": 0.7383720930232558, "grad_norm": 59.37482833862305, "learning_rate": 2.6162790697674424e-06, "loss": 0.2573, "step": 4572 }, { "epoch": 0.7385335917312662, "grad_norm": 141.4961700439453, "learning_rate": 2.614664082687339e-06, "loss": 0.6738, "step": 4573 }, { "epoch": 0.7386950904392765, "grad_norm": 57.133140563964844, "learning_rate": 2.613049095607235e-06, "loss": 0.4082, "step": 4574 }, { "epoch": 0.7388565891472868, "grad_norm": 32.22671890258789, "learning_rate": 2.611434108527132e-06, "loss": 0.1204, "step": 4575 }, { "epoch": 0.7390180878552972, "grad_norm": 197.86708068847656, "learning_rate": 2.6098191214470286e-06, "loss": 2.0312, "step": 4576 }, { "epoch": 0.7391795865633075, "grad_norm": 241.0409393310547, "learning_rate": 2.608204134366925e-06, "loss": 1.9834, "step": 4577 }, { "epoch": 0.7393410852713178, "grad_norm": 411.1399841308594, "learning_rate": 2.606589147286822e-06, "loss": 0.9451, "step": 4578 }, { "epoch": 0.7395025839793282, "grad_norm": 43.29279327392578, "learning_rate": 2.6049741602067187e-06, "loss": 0.207, "step": 4579 }, { "epoch": 0.7396640826873385, "grad_norm": 160.22128295898438, "learning_rate": 2.603359173126615e-06, "loss": 0.9219, "step": 4580 }, { "epoch": 0.7398255813953488, "grad_norm": 44.76338195800781, "learning_rate": 2.601744186046512e-06, "loss": 0.2767, "step": 4581 }, { "epoch": 0.7399870801033591, "grad_norm": 79.35201263427734, "learning_rate": 2.6001291989664084e-06, "loss": 0.1957, "step": 4582 }, { "epoch": 0.7401485788113695, "grad_norm": 7.9296698570251465, "learning_rate": 2.598514211886305e-06, "loss": 0.0294, "step": 4583 }, { "epoch": 0.7403100775193798, "grad_norm": 26.614185333251953, "learning_rate": 2.596899224806202e-06, "loss": 0.0887, "step": 4584 }, { "epoch": 0.7404715762273901, "grad_norm": 135.3283233642578, "learning_rate": 2.5952842377260985e-06, "loss": 0.314, "step": 4585 }, { "epoch": 0.7406330749354005, "grad_norm": 44.94691848754883, "learning_rate": 2.5936692506459946e-06, "loss": 0.2422, "step": 4586 }, { "epoch": 0.7407945736434108, "grad_norm": 21.55805015563965, "learning_rate": 2.592054263565892e-06, "loss": 0.1128, "step": 4587 }, { "epoch": 0.7409560723514211, "grad_norm": 14.960688591003418, "learning_rate": 2.590439276485788e-06, "loss": 0.0469, "step": 4588 }, { "epoch": 0.7411175710594315, "grad_norm": 6.763660430908203, "learning_rate": 2.5888242894056847e-06, "loss": 0.021, "step": 4589 }, { "epoch": 0.7412790697674418, "grad_norm": 58.50383377075195, "learning_rate": 2.5872093023255817e-06, "loss": 0.2648, "step": 4590 }, { "epoch": 0.7414405684754521, "grad_norm": 74.0493392944336, "learning_rate": 2.5855943152454783e-06, "loss": 0.4082, "step": 4591 }, { "epoch": 0.7416020671834626, "grad_norm": 68.94171905517578, "learning_rate": 2.583979328165375e-06, "loss": 1.2344, "step": 4592 }, { "epoch": 0.7417635658914729, "grad_norm": 164.55776977539062, "learning_rate": 2.582364341085272e-06, "loss": 0.9519, "step": 4593 }, { "epoch": 0.7419250645994832, "grad_norm": 40.41731643676758, "learning_rate": 2.580749354005168e-06, "loss": 0.2498, "step": 4594 }, { "epoch": 0.7420865633074936, "grad_norm": 23.1707820892334, "learning_rate": 2.5791343669250645e-06, "loss": 0.0878, "step": 4595 }, { "epoch": 0.7422480620155039, "grad_norm": 4.4262261390686035, "learning_rate": 2.5775193798449615e-06, "loss": 0.0141, "step": 4596 }, { "epoch": 0.7424095607235142, "grad_norm": 82.21753692626953, "learning_rate": 2.575904392764858e-06, "loss": 0.7875, "step": 4597 }, { "epoch": 0.7425710594315246, "grad_norm": 23.557222366333008, "learning_rate": 2.5742894056847546e-06, "loss": 0.0807, "step": 4598 }, { "epoch": 0.7427325581395349, "grad_norm": 14.413225173950195, "learning_rate": 2.5726744186046516e-06, "loss": 0.0505, "step": 4599 }, { "epoch": 0.7428940568475452, "grad_norm": 186.40301513671875, "learning_rate": 2.5710594315245478e-06, "loss": 0.4951, "step": 4600 }, { "epoch": 0.7430555555555556, "grad_norm": 66.81478118896484, "learning_rate": 2.5694444444444443e-06, "loss": 0.1656, "step": 4601 }, { "epoch": 0.7432170542635659, "grad_norm": 21.298492431640625, "learning_rate": 2.5678294573643413e-06, "loss": 0.0464, "step": 4602 }, { "epoch": 0.7433785529715762, "grad_norm": 144.33460998535156, "learning_rate": 2.566214470284238e-06, "loss": 0.6917, "step": 4603 }, { "epoch": 0.7435400516795866, "grad_norm": 36.68936538696289, "learning_rate": 2.5645994832041344e-06, "loss": 0.1758, "step": 4604 }, { "epoch": 0.7437015503875969, "grad_norm": 133.87698364257812, "learning_rate": 2.5629844961240314e-06, "loss": 0.5947, "step": 4605 }, { "epoch": 0.7438630490956072, "grad_norm": 54.03537368774414, "learning_rate": 2.561369509043928e-06, "loss": 0.3038, "step": 4606 }, { "epoch": 0.7440245478036176, "grad_norm": 23.0835018157959, "learning_rate": 2.559754521963824e-06, "loss": 0.075, "step": 4607 }, { "epoch": 0.7441860465116279, "grad_norm": 52.34504699707031, "learning_rate": 2.558139534883721e-06, "loss": 0.395, "step": 4608 }, { "epoch": 0.7443475452196382, "grad_norm": 51.30753707885742, "learning_rate": 2.5565245478036176e-06, "loss": 0.1169, "step": 4609 }, { "epoch": 0.7445090439276486, "grad_norm": 69.48959350585938, "learning_rate": 2.554909560723514e-06, "loss": 0.2379, "step": 4610 }, { "epoch": 0.7446705426356589, "grad_norm": 47.21634292602539, "learning_rate": 2.553294573643411e-06, "loss": 0.2373, "step": 4611 }, { "epoch": 0.7448320413436692, "grad_norm": 300.41302490234375, "learning_rate": 2.5516795865633077e-06, "loss": 0.8724, "step": 4612 }, { "epoch": 0.7449935400516796, "grad_norm": 250.43234252929688, "learning_rate": 2.5500645994832047e-06, "loss": 2.4282, "step": 4613 }, { "epoch": 0.7451550387596899, "grad_norm": 81.41476440429688, "learning_rate": 2.548449612403101e-06, "loss": 0.5322, "step": 4614 }, { "epoch": 0.7453165374677002, "grad_norm": 58.60771560668945, "learning_rate": 2.5468346253229974e-06, "loss": 0.3689, "step": 4615 }, { "epoch": 0.7454780361757106, "grad_norm": 12.150517463684082, "learning_rate": 2.5452196382428944e-06, "loss": 0.0426, "step": 4616 }, { "epoch": 0.7456395348837209, "grad_norm": 235.12594604492188, "learning_rate": 2.543604651162791e-06, "loss": 2.1353, "step": 4617 }, { "epoch": 0.7458010335917312, "grad_norm": 107.59249114990234, "learning_rate": 2.5419896640826875e-06, "loss": 1.2781, "step": 4618 }, { "epoch": 0.7459625322997416, "grad_norm": 24.57741928100586, "learning_rate": 2.5403746770025845e-06, "loss": 0.124, "step": 4619 }, { "epoch": 0.7461240310077519, "grad_norm": 36.93639373779297, "learning_rate": 2.538759689922481e-06, "loss": 0.146, "step": 4620 }, { "epoch": 0.7462855297157622, "grad_norm": 15.35484504699707, "learning_rate": 2.537144702842377e-06, "loss": 0.0287, "step": 4621 }, { "epoch": 0.7464470284237726, "grad_norm": 43.90385055541992, "learning_rate": 2.535529715762274e-06, "loss": 0.198, "step": 4622 }, { "epoch": 0.7466085271317829, "grad_norm": 66.2145004272461, "learning_rate": 2.5339147286821708e-06, "loss": 0.6113, "step": 4623 }, { "epoch": 0.7467700258397932, "grad_norm": 73.63436889648438, "learning_rate": 2.5322997416020673e-06, "loss": 0.4258, "step": 4624 }, { "epoch": 0.7469315245478036, "grad_norm": 18.20196533203125, "learning_rate": 2.5306847545219643e-06, "loss": 0.0521, "step": 4625 }, { "epoch": 0.747093023255814, "grad_norm": 45.86671829223633, "learning_rate": 2.529069767441861e-06, "loss": 0.2212, "step": 4626 }, { "epoch": 0.7472545219638242, "grad_norm": 259.7723693847656, "learning_rate": 2.527454780361757e-06, "loss": 3.9475, "step": 4627 }, { "epoch": 0.7474160206718347, "grad_norm": 12.737114906311035, "learning_rate": 2.525839793281654e-06, "loss": 0.0382, "step": 4628 }, { "epoch": 0.747577519379845, "grad_norm": 42.84809875488281, "learning_rate": 2.5242248062015505e-06, "loss": 0.1855, "step": 4629 }, { "epoch": 0.7477390180878553, "grad_norm": 152.43154907226562, "learning_rate": 2.522609819121447e-06, "loss": 0.8021, "step": 4630 }, { "epoch": 0.7479005167958657, "grad_norm": 51.69551467895508, "learning_rate": 2.520994832041344e-06, "loss": 0.2075, "step": 4631 }, { "epoch": 0.748062015503876, "grad_norm": 77.45441436767578, "learning_rate": 2.5193798449612406e-06, "loss": 0.2935, "step": 4632 }, { "epoch": 0.7482235142118863, "grad_norm": 33.80318069458008, "learning_rate": 2.517764857881137e-06, "loss": 0.1122, "step": 4633 }, { "epoch": 0.7483850129198967, "grad_norm": 105.41830444335938, "learning_rate": 2.516149870801034e-06, "loss": 0.4648, "step": 4634 }, { "epoch": 0.748546511627907, "grad_norm": 67.00370025634766, "learning_rate": 2.5145348837209303e-06, "loss": 0.3367, "step": 4635 }, { "epoch": 0.7487080103359173, "grad_norm": 6.426934242248535, "learning_rate": 2.512919896640827e-06, "loss": 0.0234, "step": 4636 }, { "epoch": 0.7488695090439277, "grad_norm": 31.43288803100586, "learning_rate": 2.511304909560724e-06, "loss": 0.1084, "step": 4637 }, { "epoch": 0.749031007751938, "grad_norm": 30.345008850097656, "learning_rate": 2.5096899224806204e-06, "loss": 0.0935, "step": 4638 }, { "epoch": 0.7491925064599483, "grad_norm": 155.99209594726562, "learning_rate": 2.508074935400517e-06, "loss": 1.0879, "step": 4639 }, { "epoch": 0.7493540051679587, "grad_norm": 92.58602142333984, "learning_rate": 2.506459948320414e-06, "loss": 0.5928, "step": 4640 }, { "epoch": 0.749515503875969, "grad_norm": 26.715534210205078, "learning_rate": 2.50484496124031e-06, "loss": 0.0727, "step": 4641 }, { "epoch": 0.7496770025839793, "grad_norm": 8.545639038085938, "learning_rate": 2.5032299741602067e-06, "loss": 0.0259, "step": 4642 }, { "epoch": 0.7498385012919897, "grad_norm": 477.3057556152344, "learning_rate": 2.5016149870801036e-06, "loss": 2.5117, "step": 4643 }, { "epoch": 0.75, "grad_norm": 24.409032821655273, "learning_rate": 2.5e-06, "loss": 0.0895, "step": 4644 }, { "epoch": 0.7501614987080103, "grad_norm": 192.2982635498047, "learning_rate": 2.4983850129198968e-06, "loss": 1.1543, "step": 4645 }, { "epoch": 0.7503229974160207, "grad_norm": 42.18426513671875, "learning_rate": 2.4967700258397933e-06, "loss": 0.4755, "step": 4646 }, { "epoch": 0.750484496124031, "grad_norm": 5.806551456451416, "learning_rate": 2.4951550387596903e-06, "loss": 0.0172, "step": 4647 }, { "epoch": 0.7506459948320413, "grad_norm": 65.6973648071289, "learning_rate": 2.493540051679587e-06, "loss": 0.4229, "step": 4648 }, { "epoch": 0.7508074935400517, "grad_norm": 120.10063171386719, "learning_rate": 2.4919250645994834e-06, "loss": 0.5078, "step": 4649 }, { "epoch": 0.750968992248062, "grad_norm": 94.50465393066406, "learning_rate": 2.49031007751938e-06, "loss": 0.4609, "step": 4650 }, { "epoch": 0.7511304909560723, "grad_norm": 117.86177825927734, "learning_rate": 2.488695090439277e-06, "loss": 0.9473, "step": 4651 }, { "epoch": 0.7512919896640827, "grad_norm": 56.74190902709961, "learning_rate": 2.487080103359173e-06, "loss": 0.229, "step": 4652 }, { "epoch": 0.751453488372093, "grad_norm": 101.5741195678711, "learning_rate": 2.48546511627907e-06, "loss": 0.6602, "step": 4653 }, { "epoch": 0.7516149870801033, "grad_norm": 153.61172485351562, "learning_rate": 2.4838501291989667e-06, "loss": 0.9404, "step": 4654 }, { "epoch": 0.7517764857881137, "grad_norm": 114.00481414794922, "learning_rate": 2.482235142118863e-06, "loss": 0.3643, "step": 4655 }, { "epoch": 0.751937984496124, "grad_norm": 101.15873718261719, "learning_rate": 2.4806201550387598e-06, "loss": 0.6953, "step": 4656 }, { "epoch": 0.7520994832041343, "grad_norm": 70.08769226074219, "learning_rate": 2.4790051679586568e-06, "loss": 0.3838, "step": 4657 }, { "epoch": 0.7522609819121447, "grad_norm": 87.68262481689453, "learning_rate": 2.477390180878553e-06, "loss": 0.7676, "step": 4658 }, { "epoch": 0.752422480620155, "grad_norm": 33.807125091552734, "learning_rate": 2.47577519379845e-06, "loss": 0.1184, "step": 4659 }, { "epoch": 0.7525839793281653, "grad_norm": 77.08302307128906, "learning_rate": 2.4741602067183464e-06, "loss": 0.2605, "step": 4660 }, { "epoch": 0.7527454780361758, "grad_norm": 38.71072006225586, "learning_rate": 2.4725452196382434e-06, "loss": 0.1877, "step": 4661 }, { "epoch": 0.752906976744186, "grad_norm": 53.696136474609375, "learning_rate": 2.4709302325581396e-06, "loss": 0.1536, "step": 4662 }, { "epoch": 0.7530684754521964, "grad_norm": 271.8394470214844, "learning_rate": 2.4693152454780365e-06, "loss": 0.8474, "step": 4663 }, { "epoch": 0.7532299741602068, "grad_norm": 191.79830932617188, "learning_rate": 2.467700258397933e-06, "loss": 1.918, "step": 4664 }, { "epoch": 0.7533914728682171, "grad_norm": 350.0113830566406, "learning_rate": 2.4660852713178297e-06, "loss": 2.1504, "step": 4665 }, { "epoch": 0.7535529715762274, "grad_norm": 13.348382949829102, "learning_rate": 2.4644702842377262e-06, "loss": 0.0525, "step": 4666 }, { "epoch": 0.7537144702842378, "grad_norm": 191.5743408203125, "learning_rate": 2.462855297157623e-06, "loss": 0.752, "step": 4667 }, { "epoch": 0.7538759689922481, "grad_norm": 149.28379821777344, "learning_rate": 2.4612403100775193e-06, "loss": 0.8807, "step": 4668 }, { "epoch": 0.7540374677002584, "grad_norm": 41.73699951171875, "learning_rate": 2.4596253229974163e-06, "loss": 0.1636, "step": 4669 }, { "epoch": 0.7541989664082688, "grad_norm": 31.565832138061523, "learning_rate": 2.458010335917313e-06, "loss": 0.0956, "step": 4670 }, { "epoch": 0.7543604651162791, "grad_norm": 55.95780563354492, "learning_rate": 2.4563953488372094e-06, "loss": 0.1698, "step": 4671 }, { "epoch": 0.7545219638242894, "grad_norm": 114.53193664550781, "learning_rate": 2.454780361757106e-06, "loss": 1.2832, "step": 4672 }, { "epoch": 0.7546834625322998, "grad_norm": 40.843345642089844, "learning_rate": 2.453165374677003e-06, "loss": 0.3521, "step": 4673 }, { "epoch": 0.7548449612403101, "grad_norm": 18.90746307373047, "learning_rate": 2.4515503875968995e-06, "loss": 0.2083, "step": 4674 }, { "epoch": 0.7550064599483204, "grad_norm": 81.44945526123047, "learning_rate": 2.449935400516796e-06, "loss": 0.7844, "step": 4675 }, { "epoch": 0.7551679586563308, "grad_norm": 9.384523391723633, "learning_rate": 2.4483204134366927e-06, "loss": 0.0392, "step": 4676 }, { "epoch": 0.7553294573643411, "grad_norm": 17.106157302856445, "learning_rate": 2.4467054263565892e-06, "loss": 0.044, "step": 4677 }, { "epoch": 0.7554909560723514, "grad_norm": 36.69807815551758, "learning_rate": 2.4450904392764858e-06, "loss": 0.2261, "step": 4678 }, { "epoch": 0.7556524547803618, "grad_norm": 86.8497543334961, "learning_rate": 2.4434754521963828e-06, "loss": 0.5771, "step": 4679 }, { "epoch": 0.7558139534883721, "grad_norm": 77.9181137084961, "learning_rate": 2.4418604651162793e-06, "loss": 0.3291, "step": 4680 }, { "epoch": 0.7559754521963824, "grad_norm": 108.59313201904297, "learning_rate": 2.440245478036176e-06, "loss": 0.3578, "step": 4681 }, { "epoch": 0.7561369509043928, "grad_norm": 113.31842041015625, "learning_rate": 2.4386304909560724e-06, "loss": 0.5663, "step": 4682 }, { "epoch": 0.7562984496124031, "grad_norm": 9.884767532348633, "learning_rate": 2.437015503875969e-06, "loss": 0.0331, "step": 4683 }, { "epoch": 0.7564599483204134, "grad_norm": 26.368396759033203, "learning_rate": 2.435400516795866e-06, "loss": 0.052, "step": 4684 }, { "epoch": 0.7566214470284238, "grad_norm": 12.582874298095703, "learning_rate": 2.4337855297157626e-06, "loss": 0.0518, "step": 4685 }, { "epoch": 0.7567829457364341, "grad_norm": 208.97018432617188, "learning_rate": 2.432170542635659e-06, "loss": 1.125, "step": 4686 }, { "epoch": 0.7569444444444444, "grad_norm": 112.29371643066406, "learning_rate": 2.4305555555555557e-06, "loss": 1.0869, "step": 4687 }, { "epoch": 0.7571059431524548, "grad_norm": 22.668834686279297, "learning_rate": 2.4289405684754527e-06, "loss": 0.0573, "step": 4688 }, { "epoch": 0.7572674418604651, "grad_norm": 23.971134185791016, "learning_rate": 2.427325581395349e-06, "loss": 0.1606, "step": 4689 }, { "epoch": 0.7574289405684754, "grad_norm": 180.35911560058594, "learning_rate": 2.4257105943152458e-06, "loss": 0.7517, "step": 4690 }, { "epoch": 0.7575904392764858, "grad_norm": 21.01690673828125, "learning_rate": 2.4240956072351423e-06, "loss": 0.0612, "step": 4691 }, { "epoch": 0.7577519379844961, "grad_norm": 37.786643981933594, "learning_rate": 2.422480620155039e-06, "loss": 0.1936, "step": 4692 }, { "epoch": 0.7579134366925064, "grad_norm": 49.97019577026367, "learning_rate": 2.4208656330749355e-06, "loss": 0.1904, "step": 4693 }, { "epoch": 0.7580749354005168, "grad_norm": 240.7152099609375, "learning_rate": 2.4192506459948324e-06, "loss": 1.2349, "step": 4694 }, { "epoch": 0.7582364341085271, "grad_norm": 79.94078063964844, "learning_rate": 2.417635658914729e-06, "loss": 0.2332, "step": 4695 }, { "epoch": 0.7583979328165374, "grad_norm": 207.9928436279297, "learning_rate": 2.4160206718346256e-06, "loss": 1.2325, "step": 4696 }, { "epoch": 0.7585594315245479, "grad_norm": 7.8856635093688965, "learning_rate": 2.414405684754522e-06, "loss": 0.0353, "step": 4697 }, { "epoch": 0.7587209302325582, "grad_norm": 200.14991760253906, "learning_rate": 2.412790697674419e-06, "loss": 1.4062, "step": 4698 }, { "epoch": 0.7588824289405685, "grad_norm": 81.56059265136719, "learning_rate": 2.4111757105943152e-06, "loss": 1.0996, "step": 4699 }, { "epoch": 0.7590439276485789, "grad_norm": 194.05726623535156, "learning_rate": 2.4095607235142122e-06, "loss": 1.3742, "step": 4700 }, { "epoch": 0.7592054263565892, "grad_norm": 31.858821868896484, "learning_rate": 2.4079457364341088e-06, "loss": 0.1404, "step": 4701 }, { "epoch": 0.7593669250645995, "grad_norm": 52.0834846496582, "learning_rate": 2.4063307493540053e-06, "loss": 0.2828, "step": 4702 }, { "epoch": 0.7595284237726099, "grad_norm": 125.6553955078125, "learning_rate": 2.404715762273902e-06, "loss": 0.5463, "step": 4703 }, { "epoch": 0.7596899224806202, "grad_norm": 9.654336929321289, "learning_rate": 2.403100775193799e-06, "loss": 0.0285, "step": 4704 }, { "epoch": 0.7598514211886305, "grad_norm": 16.655458450317383, "learning_rate": 2.401485788113695e-06, "loss": 0.0525, "step": 4705 }, { "epoch": 0.7600129198966409, "grad_norm": 5.889278411865234, "learning_rate": 2.399870801033592e-06, "loss": 0.0347, "step": 4706 }, { "epoch": 0.7601744186046512, "grad_norm": 19.64615821838379, "learning_rate": 2.3982558139534886e-06, "loss": 0.0339, "step": 4707 }, { "epoch": 0.7603359173126615, "grad_norm": 303.638916015625, "learning_rate": 2.396640826873385e-06, "loss": 3.437, "step": 4708 }, { "epoch": 0.7604974160206718, "grad_norm": 34.737640380859375, "learning_rate": 2.3950258397932817e-06, "loss": 0.1038, "step": 4709 }, { "epoch": 0.7606589147286822, "grad_norm": 71.31887817382812, "learning_rate": 2.3934108527131787e-06, "loss": 0.6401, "step": 4710 }, { "epoch": 0.7608204134366925, "grad_norm": 35.29633712768555, "learning_rate": 2.3917958656330752e-06, "loss": 0.1094, "step": 4711 }, { "epoch": 0.7609819121447028, "grad_norm": 343.8380432128906, "learning_rate": 2.390180878552972e-06, "loss": 3.0859, "step": 4712 }, { "epoch": 0.7611434108527132, "grad_norm": 259.1978759765625, "learning_rate": 2.3885658914728683e-06, "loss": 2.5078, "step": 4713 }, { "epoch": 0.7613049095607235, "grad_norm": 120.19506072998047, "learning_rate": 2.386950904392765e-06, "loss": 0.7222, "step": 4714 }, { "epoch": 0.7614664082687338, "grad_norm": 104.7490463256836, "learning_rate": 2.3853359173126615e-06, "loss": 1.0656, "step": 4715 }, { "epoch": 0.7616279069767442, "grad_norm": 30.64553451538086, "learning_rate": 2.3837209302325585e-06, "loss": 0.2256, "step": 4716 }, { "epoch": 0.7617894056847545, "grad_norm": 70.60340881347656, "learning_rate": 2.382105943152455e-06, "loss": 0.3322, "step": 4717 }, { "epoch": 0.7619509043927648, "grad_norm": 168.075439453125, "learning_rate": 2.3804909560723516e-06, "loss": 1.7686, "step": 4718 }, { "epoch": 0.7621124031007752, "grad_norm": 12.830155372619629, "learning_rate": 2.378875968992248e-06, "loss": 0.0549, "step": 4719 }, { "epoch": 0.7622739018087855, "grad_norm": 100.10289001464844, "learning_rate": 2.3772609819121447e-06, "loss": 0.8623, "step": 4720 }, { "epoch": 0.7624354005167958, "grad_norm": 50.75680160522461, "learning_rate": 2.3756459948320417e-06, "loss": 0.1582, "step": 4721 }, { "epoch": 0.7625968992248062, "grad_norm": 100.56391906738281, "learning_rate": 2.3740310077519382e-06, "loss": 0.9107, "step": 4722 }, { "epoch": 0.7627583979328165, "grad_norm": 57.46937561035156, "learning_rate": 2.372416020671835e-06, "loss": 0.226, "step": 4723 }, { "epoch": 0.7629198966408268, "grad_norm": 13.870665550231934, "learning_rate": 2.3708010335917314e-06, "loss": 0.0334, "step": 4724 }, { "epoch": 0.7630813953488372, "grad_norm": 80.56944274902344, "learning_rate": 2.3691860465116283e-06, "loss": 0.3105, "step": 4725 }, { "epoch": 0.7632428940568475, "grad_norm": 203.909423828125, "learning_rate": 2.367571059431525e-06, "loss": 1.3633, "step": 4726 }, { "epoch": 0.7634043927648578, "grad_norm": 127.46913146972656, "learning_rate": 2.3659560723514215e-06, "loss": 1.9021, "step": 4727 }, { "epoch": 0.7635658914728682, "grad_norm": 57.774444580078125, "learning_rate": 2.364341085271318e-06, "loss": 0.3325, "step": 4728 }, { "epoch": 0.7637273901808785, "grad_norm": 38.10487365722656, "learning_rate": 2.3627260981912146e-06, "loss": 0.1707, "step": 4729 }, { "epoch": 0.7638888888888888, "grad_norm": 140.07325744628906, "learning_rate": 2.361111111111111e-06, "loss": 1.2227, "step": 4730 }, { "epoch": 0.7640503875968992, "grad_norm": 47.12117004394531, "learning_rate": 2.359496124031008e-06, "loss": 0.2339, "step": 4731 }, { "epoch": 0.7642118863049095, "grad_norm": 44.44609832763672, "learning_rate": 2.3578811369509047e-06, "loss": 0.1083, "step": 4732 }, { "epoch": 0.7643733850129198, "grad_norm": 15.141770362854004, "learning_rate": 2.3562661498708012e-06, "loss": 0.0347, "step": 4733 }, { "epoch": 0.7645348837209303, "grad_norm": 21.882797241210938, "learning_rate": 2.354651162790698e-06, "loss": 0.0833, "step": 4734 }, { "epoch": 0.7646963824289406, "grad_norm": 48.273685455322266, "learning_rate": 2.3530361757105948e-06, "loss": 0.5044, "step": 4735 }, { "epoch": 0.7648578811369509, "grad_norm": 201.9748992919922, "learning_rate": 2.351421188630491e-06, "loss": 1.084, "step": 4736 }, { "epoch": 0.7650193798449613, "grad_norm": 2.6010446548461914, "learning_rate": 2.349806201550388e-06, "loss": 0.0077, "step": 4737 }, { "epoch": 0.7651808785529716, "grad_norm": 217.6674346923828, "learning_rate": 2.3481912144702845e-06, "loss": 0.7637, "step": 4738 }, { "epoch": 0.7653423772609819, "grad_norm": 21.473976135253906, "learning_rate": 2.346576227390181e-06, "loss": 0.0497, "step": 4739 }, { "epoch": 0.7655038759689923, "grad_norm": 9.239020347595215, "learning_rate": 2.3449612403100776e-06, "loss": 0.0265, "step": 4740 }, { "epoch": 0.7656653746770026, "grad_norm": 27.024593353271484, "learning_rate": 2.3433462532299746e-06, "loss": 0.1279, "step": 4741 }, { "epoch": 0.7658268733850129, "grad_norm": 72.47074127197266, "learning_rate": 2.3417312661498707e-06, "loss": 0.3652, "step": 4742 }, { "epoch": 0.7659883720930233, "grad_norm": 83.99581146240234, "learning_rate": 2.3401162790697677e-06, "loss": 0.436, "step": 4743 }, { "epoch": 0.7661498708010336, "grad_norm": 7.061821460723877, "learning_rate": 2.3385012919896642e-06, "loss": 0.0174, "step": 4744 }, { "epoch": 0.7663113695090439, "grad_norm": 276.6634826660156, "learning_rate": 2.336886304909561e-06, "loss": 3.2764, "step": 4745 }, { "epoch": 0.7664728682170543, "grad_norm": 189.76889038085938, "learning_rate": 2.3352713178294574e-06, "loss": 2.876, "step": 4746 }, { "epoch": 0.7666343669250646, "grad_norm": 49.79312515258789, "learning_rate": 2.3336563307493544e-06, "loss": 0.3364, "step": 4747 }, { "epoch": 0.7667958656330749, "grad_norm": 195.07708740234375, "learning_rate": 2.332041343669251e-06, "loss": 0.507, "step": 4748 }, { "epoch": 0.7669573643410853, "grad_norm": 141.68077087402344, "learning_rate": 2.3304263565891475e-06, "loss": 0.5561, "step": 4749 }, { "epoch": 0.7671188630490956, "grad_norm": 97.89805603027344, "learning_rate": 2.328811369509044e-06, "loss": 1.4297, "step": 4750 }, { "epoch": 0.7672803617571059, "grad_norm": 4.4171905517578125, "learning_rate": 2.3271963824289406e-06, "loss": 0.0071, "step": 4751 }, { "epoch": 0.7674418604651163, "grad_norm": 27.36895751953125, "learning_rate": 2.3255813953488376e-06, "loss": 0.0958, "step": 4752 }, { "epoch": 0.7676033591731266, "grad_norm": 43.67045211791992, "learning_rate": 2.323966408268734e-06, "loss": 0.3389, "step": 4753 }, { "epoch": 0.7677648578811369, "grad_norm": 19.346965789794922, "learning_rate": 2.3223514211886307e-06, "loss": 0.0778, "step": 4754 }, { "epoch": 0.7679263565891473, "grad_norm": 65.39983367919922, "learning_rate": 2.3207364341085273e-06, "loss": 0.2324, "step": 4755 }, { "epoch": 0.7680878552971576, "grad_norm": 424.2745056152344, "learning_rate": 2.319121447028424e-06, "loss": 2.2314, "step": 4756 }, { "epoch": 0.7682493540051679, "grad_norm": 21.329483032226562, "learning_rate": 2.317506459948321e-06, "loss": 0.0974, "step": 4757 }, { "epoch": 0.7684108527131783, "grad_norm": 80.34711456298828, "learning_rate": 2.3158914728682174e-06, "loss": 0.501, "step": 4758 }, { "epoch": 0.7685723514211886, "grad_norm": 65.03469848632812, "learning_rate": 2.314276485788114e-06, "loss": 0.2764, "step": 4759 }, { "epoch": 0.7687338501291989, "grad_norm": 60.084693908691406, "learning_rate": 2.3126614987080105e-06, "loss": 0.1798, "step": 4760 }, { "epoch": 0.7688953488372093, "grad_norm": 49.79558181762695, "learning_rate": 2.311046511627907e-06, "loss": 0.2444, "step": 4761 }, { "epoch": 0.7690568475452196, "grad_norm": 83.61820220947266, "learning_rate": 2.309431524547804e-06, "loss": 0.375, "step": 4762 }, { "epoch": 0.7692183462532299, "grad_norm": 302.6371154785156, "learning_rate": 2.3078165374677006e-06, "loss": 1.752, "step": 4763 }, { "epoch": 0.7693798449612403, "grad_norm": 74.86998748779297, "learning_rate": 2.306201550387597e-06, "loss": 0.3828, "step": 4764 }, { "epoch": 0.7695413436692506, "grad_norm": 23.841354370117188, "learning_rate": 2.3045865633074937e-06, "loss": 0.0793, "step": 4765 }, { "epoch": 0.7697028423772609, "grad_norm": 28.41211700439453, "learning_rate": 2.3029715762273907e-06, "loss": 0.0824, "step": 4766 }, { "epoch": 0.7698643410852714, "grad_norm": 19.70965003967285, "learning_rate": 2.301356589147287e-06, "loss": 0.0498, "step": 4767 }, { "epoch": 0.7700258397932817, "grad_norm": 114.00862884521484, "learning_rate": 2.299741602067184e-06, "loss": 0.9517, "step": 4768 }, { "epoch": 0.770187338501292, "grad_norm": 314.31390380859375, "learning_rate": 2.2981266149870804e-06, "loss": 1.9858, "step": 4769 }, { "epoch": 0.7703488372093024, "grad_norm": 2.7007040977478027, "learning_rate": 2.296511627906977e-06, "loss": 0.0078, "step": 4770 }, { "epoch": 0.7705103359173127, "grad_norm": 16.970993041992188, "learning_rate": 2.2948966408268735e-06, "loss": 0.0569, "step": 4771 }, { "epoch": 0.770671834625323, "grad_norm": 28.24833869934082, "learning_rate": 2.2932816537467705e-06, "loss": 0.1561, "step": 4772 }, { "epoch": 0.7708333333333334, "grad_norm": 39.5369873046875, "learning_rate": 2.2916666666666666e-06, "loss": 0.1663, "step": 4773 }, { "epoch": 0.7709948320413437, "grad_norm": 43.425479888916016, "learning_rate": 2.2900516795865636e-06, "loss": 0.3311, "step": 4774 }, { "epoch": 0.771156330749354, "grad_norm": 118.38552856445312, "learning_rate": 2.28843669250646e-06, "loss": 0.9453, "step": 4775 }, { "epoch": 0.7713178294573644, "grad_norm": 64.61836242675781, "learning_rate": 2.2868217054263567e-06, "loss": 0.3242, "step": 4776 }, { "epoch": 0.7714793281653747, "grad_norm": 335.3985290527344, "learning_rate": 2.2852067183462533e-06, "loss": 3.5908, "step": 4777 }, { "epoch": 0.771640826873385, "grad_norm": 32.462799072265625, "learning_rate": 2.2835917312661503e-06, "loss": 0.1108, "step": 4778 }, { "epoch": 0.7718023255813954, "grad_norm": 31.163284301757812, "learning_rate": 2.2819767441860464e-06, "loss": 0.1516, "step": 4779 }, { "epoch": 0.7719638242894057, "grad_norm": 188.57009887695312, "learning_rate": 2.2803617571059434e-06, "loss": 1.6133, "step": 4780 }, { "epoch": 0.772125322997416, "grad_norm": 110.83251953125, "learning_rate": 2.27874677002584e-06, "loss": 0.7637, "step": 4781 }, { "epoch": 0.7722868217054264, "grad_norm": 227.88961791992188, "learning_rate": 2.2771317829457365e-06, "loss": 0.9463, "step": 4782 }, { "epoch": 0.7724483204134367, "grad_norm": 35.006744384765625, "learning_rate": 2.275516795865633e-06, "loss": 0.1991, "step": 4783 }, { "epoch": 0.772609819121447, "grad_norm": 56.28898239135742, "learning_rate": 2.27390180878553e-06, "loss": 0.1358, "step": 4784 }, { "epoch": 0.7727713178294574, "grad_norm": 80.11046600341797, "learning_rate": 2.2722868217054266e-06, "loss": 0.3323, "step": 4785 }, { "epoch": 0.7729328165374677, "grad_norm": 384.9432067871094, "learning_rate": 2.270671834625323e-06, "loss": 0.7344, "step": 4786 }, { "epoch": 0.773094315245478, "grad_norm": 15.168935775756836, "learning_rate": 2.2690568475452197e-06, "loss": 0.0311, "step": 4787 }, { "epoch": 0.7732558139534884, "grad_norm": 28.780277252197266, "learning_rate": 2.2674418604651163e-06, "loss": 0.1075, "step": 4788 }, { "epoch": 0.7734173126614987, "grad_norm": 75.05648803710938, "learning_rate": 2.2658268733850133e-06, "loss": 0.6434, "step": 4789 }, { "epoch": 0.773578811369509, "grad_norm": 32.03690719604492, "learning_rate": 2.26421188630491e-06, "loss": 0.1626, "step": 4790 }, { "epoch": 0.7737403100775194, "grad_norm": 32.107967376708984, "learning_rate": 2.2625968992248064e-06, "loss": 0.0736, "step": 4791 }, { "epoch": 0.7739018087855297, "grad_norm": 104.05818176269531, "learning_rate": 2.260981912144703e-06, "loss": 0.8672, "step": 4792 }, { "epoch": 0.77406330749354, "grad_norm": 52.19904327392578, "learning_rate": 2.2593669250645995e-06, "loss": 0.1227, "step": 4793 }, { "epoch": 0.7742248062015504, "grad_norm": 25.722990036010742, "learning_rate": 2.2577519379844965e-06, "loss": 0.0919, "step": 4794 }, { "epoch": 0.7743863049095607, "grad_norm": 133.29502868652344, "learning_rate": 2.256136950904393e-06, "loss": 1.7656, "step": 4795 }, { "epoch": 0.774547803617571, "grad_norm": 48.925846099853516, "learning_rate": 2.2545219638242896e-06, "loss": 0.1904, "step": 4796 }, { "epoch": 0.7747093023255814, "grad_norm": 79.78763580322266, "learning_rate": 2.252906976744186e-06, "loss": 0.2734, "step": 4797 }, { "epoch": 0.7748708010335917, "grad_norm": 9.241923332214355, "learning_rate": 2.2512919896640827e-06, "loss": 0.046, "step": 4798 }, { "epoch": 0.775032299741602, "grad_norm": 53.75716781616211, "learning_rate": 2.2496770025839797e-06, "loss": 0.2891, "step": 4799 }, { "epoch": 0.7751937984496124, "grad_norm": 41.58349609375, "learning_rate": 2.2480620155038763e-06, "loss": 0.3171, "step": 4800 }, { "epoch": 0.7753552971576227, "grad_norm": 50.68268966674805, "learning_rate": 2.246447028423773e-06, "loss": 0.208, "step": 4801 }, { "epoch": 0.775516795865633, "grad_norm": 3.785435914993286, "learning_rate": 2.2448320413436694e-06, "loss": 0.0158, "step": 4802 }, { "epoch": 0.7756782945736435, "grad_norm": 73.06945037841797, "learning_rate": 2.2432170542635664e-06, "loss": 0.7254, "step": 4803 }, { "epoch": 0.7758397932816538, "grad_norm": 108.71514129638672, "learning_rate": 2.2416020671834625e-06, "loss": 0.4814, "step": 4804 }, { "epoch": 0.776001291989664, "grad_norm": 193.34934997558594, "learning_rate": 2.2399870801033595e-06, "loss": 0.4729, "step": 4805 }, { "epoch": 0.7761627906976745, "grad_norm": 41.10222625732422, "learning_rate": 2.238372093023256e-06, "loss": 0.2676, "step": 4806 }, { "epoch": 0.7763242894056848, "grad_norm": 54.58225631713867, "learning_rate": 2.2367571059431526e-06, "loss": 0.2487, "step": 4807 }, { "epoch": 0.7764857881136951, "grad_norm": 8.801992416381836, "learning_rate": 2.235142118863049e-06, "loss": 0.012, "step": 4808 }, { "epoch": 0.7766472868217055, "grad_norm": 77.87613677978516, "learning_rate": 2.233527131782946e-06, "loss": 0.8982, "step": 4809 }, { "epoch": 0.7768087855297158, "grad_norm": 16.047435760498047, "learning_rate": 2.2319121447028423e-06, "loss": 0.0592, "step": 4810 }, { "epoch": 0.7769702842377261, "grad_norm": 4.215124607086182, "learning_rate": 2.2302971576227393e-06, "loss": 0.0161, "step": 4811 }, { "epoch": 0.7771317829457365, "grad_norm": 335.80267333984375, "learning_rate": 2.228682170542636e-06, "loss": 1.9648, "step": 4812 }, { "epoch": 0.7772932816537468, "grad_norm": 92.8466796875, "learning_rate": 2.2270671834625324e-06, "loss": 0.2861, "step": 4813 }, { "epoch": 0.7774547803617571, "grad_norm": 52.55376434326172, "learning_rate": 2.225452196382429e-06, "loss": 0.1139, "step": 4814 }, { "epoch": 0.7776162790697675, "grad_norm": 285.23480224609375, "learning_rate": 2.223837209302326e-06, "loss": 1.2344, "step": 4815 }, { "epoch": 0.7777777777777778, "grad_norm": 56.292842864990234, "learning_rate": 2.222222222222222e-06, "loss": 0.3588, "step": 4816 }, { "epoch": 0.7779392764857881, "grad_norm": 8.782157897949219, "learning_rate": 2.220607235142119e-06, "loss": 0.0193, "step": 4817 }, { "epoch": 0.7781007751937985, "grad_norm": 34.575233459472656, "learning_rate": 2.2189922480620156e-06, "loss": 0.1017, "step": 4818 }, { "epoch": 0.7782622739018088, "grad_norm": 69.61238861083984, "learning_rate": 2.217377260981912e-06, "loss": 0.4097, "step": 4819 }, { "epoch": 0.7784237726098191, "grad_norm": 16.509761810302734, "learning_rate": 2.2157622739018087e-06, "loss": 0.085, "step": 4820 }, { "epoch": 0.7785852713178295, "grad_norm": 304.322265625, "learning_rate": 2.2141472868217057e-06, "loss": 1.1323, "step": 4821 }, { "epoch": 0.7787467700258398, "grad_norm": 57.83958435058594, "learning_rate": 2.2125322997416023e-06, "loss": 0.1312, "step": 4822 }, { "epoch": 0.7789082687338501, "grad_norm": 52.24795150756836, "learning_rate": 2.210917312661499e-06, "loss": 0.365, "step": 4823 }, { "epoch": 0.7790697674418605, "grad_norm": 316.74066162109375, "learning_rate": 2.2093023255813954e-06, "loss": 2.0918, "step": 4824 }, { "epoch": 0.7792312661498708, "grad_norm": 39.28800964355469, "learning_rate": 2.2076873385012924e-06, "loss": 0.1154, "step": 4825 }, { "epoch": 0.7793927648578811, "grad_norm": 95.8288803100586, "learning_rate": 2.206072351421189e-06, "loss": 0.4561, "step": 4826 }, { "epoch": 0.7795542635658915, "grad_norm": 5.051427364349365, "learning_rate": 2.2044573643410855e-06, "loss": 0.0179, "step": 4827 }, { "epoch": 0.7797157622739018, "grad_norm": 63.149131774902344, "learning_rate": 2.202842377260982e-06, "loss": 0.3242, "step": 4828 }, { "epoch": 0.7798772609819121, "grad_norm": 284.4638366699219, "learning_rate": 2.2012273901808786e-06, "loss": 2.9469, "step": 4829 }, { "epoch": 0.7800387596899225, "grad_norm": 132.73329162597656, "learning_rate": 2.199612403100775e-06, "loss": 0.8256, "step": 4830 }, { "epoch": 0.7802002583979328, "grad_norm": 205.921875, "learning_rate": 2.197997416020672e-06, "loss": 1.6044, "step": 4831 }, { "epoch": 0.7803617571059431, "grad_norm": 196.865478515625, "learning_rate": 2.1963824289405687e-06, "loss": 1.4353, "step": 4832 }, { "epoch": 0.7805232558139535, "grad_norm": 216.59228515625, "learning_rate": 2.1947674418604653e-06, "loss": 1.2539, "step": 4833 }, { "epoch": 0.7806847545219638, "grad_norm": 3.1218149662017822, "learning_rate": 2.193152454780362e-06, "loss": 0.0082, "step": 4834 }, { "epoch": 0.7808462532299741, "grad_norm": 96.96613311767578, "learning_rate": 2.1915374677002584e-06, "loss": 1.1208, "step": 4835 }, { "epoch": 0.7810077519379846, "grad_norm": 39.366485595703125, "learning_rate": 2.1899224806201554e-06, "loss": 0.1233, "step": 4836 }, { "epoch": 0.7811692506459949, "grad_norm": 168.6717071533203, "learning_rate": 2.188307493540052e-06, "loss": 1.0254, "step": 4837 }, { "epoch": 0.7813307493540051, "grad_norm": 74.12130737304688, "learning_rate": 2.1866925064599485e-06, "loss": 0.5361, "step": 4838 }, { "epoch": 0.7814922480620154, "grad_norm": 43.0799674987793, "learning_rate": 2.185077519379845e-06, "loss": 0.1376, "step": 4839 }, { "epoch": 0.7816537467700259, "grad_norm": 188.06036376953125, "learning_rate": 2.183462532299742e-06, "loss": 0.9167, "step": 4840 }, { "epoch": 0.7818152454780362, "grad_norm": 20.861839294433594, "learning_rate": 2.181847545219638e-06, "loss": 0.0558, "step": 4841 }, { "epoch": 0.7819767441860465, "grad_norm": 4.446837902069092, "learning_rate": 2.180232558139535e-06, "loss": 0.0097, "step": 4842 }, { "epoch": 0.7821382428940569, "grad_norm": 47.251861572265625, "learning_rate": 2.1786175710594317e-06, "loss": 0.1711, "step": 4843 }, { "epoch": 0.7822997416020672, "grad_norm": 10.64998722076416, "learning_rate": 2.1770025839793283e-06, "loss": 0.0203, "step": 4844 }, { "epoch": 0.7824612403100775, "grad_norm": 56.57632064819336, "learning_rate": 2.175387596899225e-06, "loss": 0.1616, "step": 4845 }, { "epoch": 0.7826227390180879, "grad_norm": 268.8695068359375, "learning_rate": 2.173772609819122e-06, "loss": 1.0531, "step": 4846 }, { "epoch": 0.7827842377260982, "grad_norm": 85.87916564941406, "learning_rate": 2.172157622739018e-06, "loss": 0.2911, "step": 4847 }, { "epoch": 0.7829457364341085, "grad_norm": 38.59315872192383, "learning_rate": 2.170542635658915e-06, "loss": 0.1296, "step": 4848 }, { "epoch": 0.7831072351421189, "grad_norm": 14.140521049499512, "learning_rate": 2.1689276485788115e-06, "loss": 0.0565, "step": 4849 }, { "epoch": 0.7832687338501292, "grad_norm": 72.00340270996094, "learning_rate": 2.167312661498708e-06, "loss": 0.5483, "step": 4850 }, { "epoch": 0.7834302325581395, "grad_norm": 85.5381851196289, "learning_rate": 2.1656976744186046e-06, "loss": 0.7484, "step": 4851 }, { "epoch": 0.7835917312661499, "grad_norm": 40.91469955444336, "learning_rate": 2.1640826873385016e-06, "loss": 0.3253, "step": 4852 }, { "epoch": 0.7837532299741602, "grad_norm": 64.05117797851562, "learning_rate": 2.162467700258398e-06, "loss": 0.2863, "step": 4853 }, { "epoch": 0.7839147286821705, "grad_norm": 12.7196044921875, "learning_rate": 2.1608527131782947e-06, "loss": 0.0359, "step": 4854 }, { "epoch": 0.7840762273901809, "grad_norm": 70.28707122802734, "learning_rate": 2.1592377260981913e-06, "loss": 0.3525, "step": 4855 }, { "epoch": 0.7842377260981912, "grad_norm": 117.88401794433594, "learning_rate": 2.1576227390180883e-06, "loss": 0.8242, "step": 4856 }, { "epoch": 0.7843992248062015, "grad_norm": 15.54667854309082, "learning_rate": 2.1560077519379844e-06, "loss": 0.0558, "step": 4857 }, { "epoch": 0.7845607235142119, "grad_norm": 124.49594116210938, "learning_rate": 2.1543927648578814e-06, "loss": 0.7963, "step": 4858 }, { "epoch": 0.7847222222222222, "grad_norm": 122.44487762451172, "learning_rate": 2.152777777777778e-06, "loss": 1.0657, "step": 4859 }, { "epoch": 0.7848837209302325, "grad_norm": 91.1722640991211, "learning_rate": 2.1511627906976745e-06, "loss": 0.7266, "step": 4860 }, { "epoch": 0.7850452196382429, "grad_norm": 37.53163528442383, "learning_rate": 2.149547803617571e-06, "loss": 0.1592, "step": 4861 }, { "epoch": 0.7852067183462532, "grad_norm": 106.61930084228516, "learning_rate": 2.147932816537468e-06, "loss": 0.3596, "step": 4862 }, { "epoch": 0.7853682170542635, "grad_norm": 153.29161071777344, "learning_rate": 2.1463178294573646e-06, "loss": 0.6611, "step": 4863 }, { "epoch": 0.7855297157622739, "grad_norm": 615.6929321289062, "learning_rate": 2.144702842377261e-06, "loss": 2.7617, "step": 4864 }, { "epoch": 0.7856912144702842, "grad_norm": 11.244659423828125, "learning_rate": 2.1430878552971577e-06, "loss": 0.0469, "step": 4865 }, { "epoch": 0.7858527131782945, "grad_norm": 38.181461334228516, "learning_rate": 2.1414728682170543e-06, "loss": 0.1489, "step": 4866 }, { "epoch": 0.7860142118863049, "grad_norm": 303.90576171875, "learning_rate": 2.1398578811369513e-06, "loss": 2.5292, "step": 4867 }, { "epoch": 0.7861757105943152, "grad_norm": 119.77725219726562, "learning_rate": 2.138242894056848e-06, "loss": 0.2827, "step": 4868 }, { "epoch": 0.7863372093023255, "grad_norm": 325.73284912109375, "learning_rate": 2.1366279069767444e-06, "loss": 1.1201, "step": 4869 }, { "epoch": 0.786498708010336, "grad_norm": 108.21077728271484, "learning_rate": 2.135012919896641e-06, "loss": 0.4023, "step": 4870 }, { "epoch": 0.7866602067183462, "grad_norm": 9.1973237991333, "learning_rate": 2.1333979328165375e-06, "loss": 0.0235, "step": 4871 }, { "epoch": 0.7868217054263565, "grad_norm": 28.27455711364746, "learning_rate": 2.131782945736434e-06, "loss": 0.2515, "step": 4872 }, { "epoch": 0.786983204134367, "grad_norm": 5.576756954193115, "learning_rate": 2.130167958656331e-06, "loss": 0.0178, "step": 4873 }, { "epoch": 0.7871447028423773, "grad_norm": 13.05694580078125, "learning_rate": 2.1285529715762276e-06, "loss": 0.0566, "step": 4874 }, { "epoch": 0.7873062015503876, "grad_norm": 21.752782821655273, "learning_rate": 2.126937984496124e-06, "loss": 0.0803, "step": 4875 }, { "epoch": 0.787467700258398, "grad_norm": 18.34182357788086, "learning_rate": 2.1253229974160208e-06, "loss": 0.0801, "step": 4876 }, { "epoch": 0.7876291989664083, "grad_norm": 20.342424392700195, "learning_rate": 2.1237080103359177e-06, "loss": 0.0945, "step": 4877 }, { "epoch": 0.7877906976744186, "grad_norm": 36.61065673828125, "learning_rate": 2.122093023255814e-06, "loss": 0.1072, "step": 4878 }, { "epoch": 0.787952196382429, "grad_norm": 54.44866943359375, "learning_rate": 2.120478036175711e-06, "loss": 0.5571, "step": 4879 }, { "epoch": 0.7881136950904393, "grad_norm": 63.970821380615234, "learning_rate": 2.1188630490956074e-06, "loss": 0.28, "step": 4880 }, { "epoch": 0.7882751937984496, "grad_norm": 184.9656982421875, "learning_rate": 2.117248062015504e-06, "loss": 0.8528, "step": 4881 }, { "epoch": 0.78843669250646, "grad_norm": 10.743674278259277, "learning_rate": 2.1156330749354005e-06, "loss": 0.03, "step": 4882 }, { "epoch": 0.7885981912144703, "grad_norm": 113.29145050048828, "learning_rate": 2.1140180878552975e-06, "loss": 1.0767, "step": 4883 }, { "epoch": 0.7887596899224806, "grad_norm": 31.005168914794922, "learning_rate": 2.1124031007751937e-06, "loss": 0.1499, "step": 4884 }, { "epoch": 0.788921188630491, "grad_norm": 43.49369430541992, "learning_rate": 2.1107881136950906e-06, "loss": 0.1392, "step": 4885 }, { "epoch": 0.7890826873385013, "grad_norm": 77.65272521972656, "learning_rate": 2.109173126614987e-06, "loss": 0.2052, "step": 4886 }, { "epoch": 0.7892441860465116, "grad_norm": 15.126131057739258, "learning_rate": 2.107558139534884e-06, "loss": 0.0414, "step": 4887 }, { "epoch": 0.789405684754522, "grad_norm": 109.55886840820312, "learning_rate": 2.1059431524547803e-06, "loss": 0.3284, "step": 4888 }, { "epoch": 0.7895671834625323, "grad_norm": 257.93475341796875, "learning_rate": 2.1043281653746773e-06, "loss": 0.9463, "step": 4889 }, { "epoch": 0.7897286821705426, "grad_norm": 75.31468963623047, "learning_rate": 2.102713178294574e-06, "loss": 0.4897, "step": 4890 }, { "epoch": 0.789890180878553, "grad_norm": 93.75696563720703, "learning_rate": 2.1010981912144704e-06, "loss": 0.4502, "step": 4891 }, { "epoch": 0.7900516795865633, "grad_norm": 45.78736877441406, "learning_rate": 2.099483204134367e-06, "loss": 0.1831, "step": 4892 }, { "epoch": 0.7902131782945736, "grad_norm": 36.31535720825195, "learning_rate": 2.097868217054264e-06, "loss": 0.1073, "step": 4893 }, { "epoch": 0.790374677002584, "grad_norm": 24.793062210083008, "learning_rate": 2.09625322997416e-06, "loss": 0.0941, "step": 4894 }, { "epoch": 0.7905361757105943, "grad_norm": 12.649964332580566, "learning_rate": 2.094638242894057e-06, "loss": 0.0461, "step": 4895 }, { "epoch": 0.7906976744186046, "grad_norm": 11.669432640075684, "learning_rate": 2.0930232558139536e-06, "loss": 0.028, "step": 4896 }, { "epoch": 0.790859173126615, "grad_norm": 105.10661315917969, "learning_rate": 2.09140826873385e-06, "loss": 0.6702, "step": 4897 }, { "epoch": 0.7910206718346253, "grad_norm": 54.16647720336914, "learning_rate": 2.0897932816537468e-06, "loss": 0.4167, "step": 4898 }, { "epoch": 0.7911821705426356, "grad_norm": 197.67599487304688, "learning_rate": 2.0881782945736438e-06, "loss": 1.7495, "step": 4899 }, { "epoch": 0.791343669250646, "grad_norm": 31.8839168548584, "learning_rate": 2.0865633074935403e-06, "loss": 0.0779, "step": 4900 }, { "epoch": 0.7915051679586563, "grad_norm": 111.21747589111328, "learning_rate": 2.084948320413437e-06, "loss": 0.2183, "step": 4901 }, { "epoch": 0.7916666666666666, "grad_norm": 143.6334686279297, "learning_rate": 2.0833333333333334e-06, "loss": 0.5134, "step": 4902 }, { "epoch": 0.791828165374677, "grad_norm": 8.14763069152832, "learning_rate": 2.08171834625323e-06, "loss": 0.015, "step": 4903 }, { "epoch": 0.7919896640826873, "grad_norm": 34.24874496459961, "learning_rate": 2.080103359173127e-06, "loss": 0.148, "step": 4904 }, { "epoch": 0.7921511627906976, "grad_norm": 24.558202743530273, "learning_rate": 2.0784883720930235e-06, "loss": 0.1144, "step": 4905 }, { "epoch": 0.792312661498708, "grad_norm": 254.3067626953125, "learning_rate": 2.07687338501292e-06, "loss": 1.5713, "step": 4906 }, { "epoch": 0.7924741602067183, "grad_norm": 20.803388595581055, "learning_rate": 2.0752583979328167e-06, "loss": 0.0673, "step": 4907 }, { "epoch": 0.7926356589147286, "grad_norm": 8.941951751708984, "learning_rate": 2.0736434108527132e-06, "loss": 0.0454, "step": 4908 }, { "epoch": 0.7927971576227391, "grad_norm": 17.662189483642578, "learning_rate": 2.0720284237726098e-06, "loss": 0.0535, "step": 4909 }, { "epoch": 0.7929586563307494, "grad_norm": 40.05923843383789, "learning_rate": 2.0704134366925068e-06, "loss": 0.1667, "step": 4910 }, { "epoch": 0.7931201550387597, "grad_norm": 258.247314453125, "learning_rate": 2.0687984496124033e-06, "loss": 1.158, "step": 4911 }, { "epoch": 0.7932816537467701, "grad_norm": 214.7875518798828, "learning_rate": 2.0671834625323e-06, "loss": 0.8083, "step": 4912 }, { "epoch": 0.7934431524547804, "grad_norm": 62.92426300048828, "learning_rate": 2.0655684754521964e-06, "loss": 0.459, "step": 4913 }, { "epoch": 0.7936046511627907, "grad_norm": 5.29307222366333, "learning_rate": 2.0639534883720934e-06, "loss": 0.0264, "step": 4914 }, { "epoch": 0.7937661498708011, "grad_norm": 248.65245056152344, "learning_rate": 2.0623385012919896e-06, "loss": 1.1577, "step": 4915 }, { "epoch": 0.7939276485788114, "grad_norm": 34.752708435058594, "learning_rate": 2.0607235142118865e-06, "loss": 0.1367, "step": 4916 }, { "epoch": 0.7940891472868217, "grad_norm": 94.43002319335938, "learning_rate": 2.059108527131783e-06, "loss": 1.2477, "step": 4917 }, { "epoch": 0.7942506459948321, "grad_norm": 4.7836432456970215, "learning_rate": 2.0574935400516797e-06, "loss": 0.0105, "step": 4918 }, { "epoch": 0.7944121447028424, "grad_norm": 63.1376838684082, "learning_rate": 2.0558785529715762e-06, "loss": 0.7822, "step": 4919 }, { "epoch": 0.7945736434108527, "grad_norm": 13.660503387451172, "learning_rate": 2.054263565891473e-06, "loss": 0.0546, "step": 4920 }, { "epoch": 0.7947351421188631, "grad_norm": 24.146554946899414, "learning_rate": 2.0526485788113698e-06, "loss": 0.0591, "step": 4921 }, { "epoch": 0.7948966408268734, "grad_norm": 344.479248046875, "learning_rate": 2.0510335917312663e-06, "loss": 1.957, "step": 4922 }, { "epoch": 0.7950581395348837, "grad_norm": 11.9397554397583, "learning_rate": 2.049418604651163e-06, "loss": 0.0923, "step": 4923 }, { "epoch": 0.7952196382428941, "grad_norm": 62.34480667114258, "learning_rate": 2.04780361757106e-06, "loss": 0.3838, "step": 4924 }, { "epoch": 0.7953811369509044, "grad_norm": 19.861942291259766, "learning_rate": 2.046188630490956e-06, "loss": 0.0726, "step": 4925 }, { "epoch": 0.7955426356589147, "grad_norm": 6.42310094833374, "learning_rate": 2.044573643410853e-06, "loss": 0.0087, "step": 4926 }, { "epoch": 0.7957041343669251, "grad_norm": 53.35761260986328, "learning_rate": 2.0429586563307496e-06, "loss": 0.2571, "step": 4927 }, { "epoch": 0.7958656330749354, "grad_norm": 25.912199020385742, "learning_rate": 2.041343669250646e-06, "loss": 0.1494, "step": 4928 }, { "epoch": 0.7960271317829457, "grad_norm": 26.161949157714844, "learning_rate": 2.0397286821705427e-06, "loss": 0.0616, "step": 4929 }, { "epoch": 0.7961886304909561, "grad_norm": 93.5312271118164, "learning_rate": 2.0381136950904397e-06, "loss": 0.8406, "step": 4930 }, { "epoch": 0.7963501291989664, "grad_norm": 4.932405948638916, "learning_rate": 2.036498708010336e-06, "loss": 0.0268, "step": 4931 }, { "epoch": 0.7965116279069767, "grad_norm": 37.304779052734375, "learning_rate": 2.0348837209302328e-06, "loss": 0.2725, "step": 4932 }, { "epoch": 0.7966731266149871, "grad_norm": 138.488525390625, "learning_rate": 2.0332687338501293e-06, "loss": 0.498, "step": 4933 }, { "epoch": 0.7968346253229974, "grad_norm": 71.78416442871094, "learning_rate": 2.031653746770026e-06, "loss": 0.5688, "step": 4934 }, { "epoch": 0.7969961240310077, "grad_norm": 44.98928451538086, "learning_rate": 2.0300387596899225e-06, "loss": 0.166, "step": 4935 }, { "epoch": 0.7971576227390181, "grad_norm": 29.596891403198242, "learning_rate": 2.0284237726098194e-06, "loss": 0.0743, "step": 4936 }, { "epoch": 0.7973191214470284, "grad_norm": 44.10761642456055, "learning_rate": 2.026808785529716e-06, "loss": 0.3184, "step": 4937 }, { "epoch": 0.7974806201550387, "grad_norm": 160.49172973632812, "learning_rate": 2.0251937984496126e-06, "loss": 0.5908, "step": 4938 }, { "epoch": 0.7976421188630491, "grad_norm": 67.79480743408203, "learning_rate": 2.023578811369509e-06, "loss": 0.2065, "step": 4939 }, { "epoch": 0.7978036175710594, "grad_norm": 11.800934791564941, "learning_rate": 2.0219638242894057e-06, "loss": 0.0559, "step": 4940 }, { "epoch": 0.7979651162790697, "grad_norm": 38.147003173828125, "learning_rate": 2.0203488372093027e-06, "loss": 0.1233, "step": 4941 }, { "epoch": 0.7981266149870802, "grad_norm": 113.6938705444336, "learning_rate": 2.0187338501291992e-06, "loss": 1.1465, "step": 4942 }, { "epoch": 0.7982881136950905, "grad_norm": 33.40287780761719, "learning_rate": 2.0171188630490958e-06, "loss": 0.1558, "step": 4943 }, { "epoch": 0.7984496124031008, "grad_norm": 128.17405700683594, "learning_rate": 2.0155038759689923e-06, "loss": 0.5991, "step": 4944 }, { "epoch": 0.7986111111111112, "grad_norm": 43.08134078979492, "learning_rate": 2.0138888888888893e-06, "loss": 0.1287, "step": 4945 }, { "epoch": 0.7987726098191215, "grad_norm": 25.946430206298828, "learning_rate": 2.0122739018087855e-06, "loss": 0.0771, "step": 4946 }, { "epoch": 0.7989341085271318, "grad_norm": 34.23838424682617, "learning_rate": 2.0106589147286824e-06, "loss": 0.2236, "step": 4947 }, { "epoch": 0.7990956072351422, "grad_norm": 17.855573654174805, "learning_rate": 2.009043927648579e-06, "loss": 0.0391, "step": 4948 }, { "epoch": 0.7992571059431525, "grad_norm": 145.8179473876953, "learning_rate": 2.0074289405684756e-06, "loss": 1.2667, "step": 4949 }, { "epoch": 0.7994186046511628, "grad_norm": 109.45867156982422, "learning_rate": 2.005813953488372e-06, "loss": 0.7025, "step": 4950 }, { "epoch": 0.7995801033591732, "grad_norm": 27.90595817565918, "learning_rate": 2.004198966408269e-06, "loss": 0.0969, "step": 4951 }, { "epoch": 0.7997416020671835, "grad_norm": 65.41680145263672, "learning_rate": 2.0025839793281657e-06, "loss": 0.2024, "step": 4952 }, { "epoch": 0.7999031007751938, "grad_norm": 22.89280128479004, "learning_rate": 2.0009689922480622e-06, "loss": 0.1065, "step": 4953 }, { "epoch": 0.8000645994832042, "grad_norm": 55.551597595214844, "learning_rate": 1.9993540051679588e-06, "loss": 0.34, "step": 4954 }, { "epoch": 0.8002260981912145, "grad_norm": 72.671875, "learning_rate": 1.9977390180878558e-06, "loss": 0.5898, "step": 4955 }, { "epoch": 0.8003875968992248, "grad_norm": 233.9957733154297, "learning_rate": 1.996124031007752e-06, "loss": 1.3228, "step": 4956 }, { "epoch": 0.8005490956072352, "grad_norm": 96.9635009765625, "learning_rate": 1.994509043927649e-06, "loss": 0.2729, "step": 4957 }, { "epoch": 0.8007105943152455, "grad_norm": 3.9587841033935547, "learning_rate": 1.9928940568475455e-06, "loss": 0.0208, "step": 4958 }, { "epoch": 0.8008720930232558, "grad_norm": 26.33218002319336, "learning_rate": 1.991279069767442e-06, "loss": 0.0889, "step": 4959 }, { "epoch": 0.8010335917312662, "grad_norm": 10.71048355102539, "learning_rate": 1.9896640826873386e-06, "loss": 0.0315, "step": 4960 }, { "epoch": 0.8011950904392765, "grad_norm": 121.90252685546875, "learning_rate": 1.9880490956072356e-06, "loss": 0.7621, "step": 4961 }, { "epoch": 0.8013565891472868, "grad_norm": 3.4159719944000244, "learning_rate": 1.9864341085271317e-06, "loss": 0.0101, "step": 4962 }, { "epoch": 0.8015180878552972, "grad_norm": 57.24496078491211, "learning_rate": 1.9848191214470287e-06, "loss": 0.5283, "step": 4963 }, { "epoch": 0.8016795865633075, "grad_norm": 15.52929401397705, "learning_rate": 1.9832041343669252e-06, "loss": 0.0656, "step": 4964 }, { "epoch": 0.8018410852713178, "grad_norm": 113.65632629394531, "learning_rate": 1.981589147286822e-06, "loss": 0.3352, "step": 4965 }, { "epoch": 0.8020025839793282, "grad_norm": 109.35179901123047, "learning_rate": 1.9799741602067184e-06, "loss": 0.2335, "step": 4966 }, { "epoch": 0.8021640826873385, "grad_norm": 28.189855575561523, "learning_rate": 1.9783591731266153e-06, "loss": 0.0938, "step": 4967 }, { "epoch": 0.8023255813953488, "grad_norm": 16.00963592529297, "learning_rate": 1.976744186046512e-06, "loss": 0.0815, "step": 4968 }, { "epoch": 0.8024870801033591, "grad_norm": 8.472941398620605, "learning_rate": 1.9751291989664085e-06, "loss": 0.0305, "step": 4969 }, { "epoch": 0.8026485788113695, "grad_norm": 30.26272964477539, "learning_rate": 1.973514211886305e-06, "loss": 0.1226, "step": 4970 }, { "epoch": 0.8028100775193798, "grad_norm": 71.69397735595703, "learning_rate": 1.9718992248062016e-06, "loss": 0.3271, "step": 4971 }, { "epoch": 0.8029715762273901, "grad_norm": 157.73106384277344, "learning_rate": 1.970284237726098e-06, "loss": 2.9403, "step": 4972 }, { "epoch": 0.8031330749354005, "grad_norm": 26.93714141845703, "learning_rate": 1.968669250645995e-06, "loss": 0.0659, "step": 4973 }, { "epoch": 0.8032945736434108, "grad_norm": 150.6220703125, "learning_rate": 1.9670542635658917e-06, "loss": 1.0532, "step": 4974 }, { "epoch": 0.8034560723514211, "grad_norm": 28.75718879699707, "learning_rate": 1.9654392764857882e-06, "loss": 0.1577, "step": 4975 }, { "epoch": 0.8036175710594315, "grad_norm": 125.84843444824219, "learning_rate": 1.963824289405685e-06, "loss": 0.9609, "step": 4976 }, { "epoch": 0.8037790697674418, "grad_norm": 5.389484882354736, "learning_rate": 1.9622093023255814e-06, "loss": 0.0193, "step": 4977 }, { "epoch": 0.8039405684754521, "grad_norm": 6.944935321807861, "learning_rate": 1.9605943152454783e-06, "loss": 0.0157, "step": 4978 }, { "epoch": 0.8041020671834626, "grad_norm": 159.7758026123047, "learning_rate": 1.958979328165375e-06, "loss": 1.6025, "step": 4979 }, { "epoch": 0.8042635658914729, "grad_norm": 46.29662322998047, "learning_rate": 1.9573643410852715e-06, "loss": 0.0924, "step": 4980 }, { "epoch": 0.8044250645994832, "grad_norm": 42.7935905456543, "learning_rate": 1.955749354005168e-06, "loss": 0.153, "step": 4981 }, { "epoch": 0.8045865633074936, "grad_norm": 56.26274871826172, "learning_rate": 1.954134366925065e-06, "loss": 0.8509, "step": 4982 }, { "epoch": 0.8047480620155039, "grad_norm": 50.670066833496094, "learning_rate": 1.9525193798449616e-06, "loss": 0.3257, "step": 4983 }, { "epoch": 0.8049095607235142, "grad_norm": 105.56961822509766, "learning_rate": 1.950904392764858e-06, "loss": 0.6001, "step": 4984 }, { "epoch": 0.8050710594315246, "grad_norm": 4.609029293060303, "learning_rate": 1.9492894056847547e-06, "loss": 0.0134, "step": 4985 }, { "epoch": 0.8052325581395349, "grad_norm": 92.57161712646484, "learning_rate": 1.9476744186046512e-06, "loss": 1.3242, "step": 4986 }, { "epoch": 0.8053940568475452, "grad_norm": 54.69158172607422, "learning_rate": 1.946059431524548e-06, "loss": 0.1797, "step": 4987 }, { "epoch": 0.8055555555555556, "grad_norm": 132.4959259033203, "learning_rate": 1.944444444444445e-06, "loss": 0.6211, "step": 4988 }, { "epoch": 0.8057170542635659, "grad_norm": 17.725540161132812, "learning_rate": 1.9428294573643414e-06, "loss": 0.0437, "step": 4989 }, { "epoch": 0.8058785529715762, "grad_norm": 39.38172149658203, "learning_rate": 1.941214470284238e-06, "loss": 0.1157, "step": 4990 }, { "epoch": 0.8060400516795866, "grad_norm": 124.32270050048828, "learning_rate": 1.9395994832041345e-06, "loss": 0.6626, "step": 4991 }, { "epoch": 0.8062015503875969, "grad_norm": 29.16582489013672, "learning_rate": 1.9379844961240315e-06, "loss": 0.135, "step": 4992 }, { "epoch": 0.8063630490956072, "grad_norm": 17.973939895629883, "learning_rate": 1.9363695090439276e-06, "loss": 0.0565, "step": 4993 }, { "epoch": 0.8065245478036176, "grad_norm": 65.42113494873047, "learning_rate": 1.9347545219638246e-06, "loss": 0.4501, "step": 4994 }, { "epoch": 0.8066860465116279, "grad_norm": 54.12174606323242, "learning_rate": 1.933139534883721e-06, "loss": 0.3134, "step": 4995 }, { "epoch": 0.8068475452196382, "grad_norm": 11.172935485839844, "learning_rate": 1.9315245478036177e-06, "loss": 0.0306, "step": 4996 }, { "epoch": 0.8070090439276486, "grad_norm": 19.713125228881836, "learning_rate": 1.9299095607235143e-06, "loss": 0.0477, "step": 4997 }, { "epoch": 0.8071705426356589, "grad_norm": 75.91326141357422, "learning_rate": 1.9282945736434112e-06, "loss": 0.1659, "step": 4998 }, { "epoch": 0.8073320413436692, "grad_norm": 107.55035400390625, "learning_rate": 1.9266795865633074e-06, "loss": 0.3123, "step": 4999 }, { "epoch": 0.8074935400516796, "grad_norm": 4.975345134735107, "learning_rate": 1.9250645994832044e-06, "loss": 0.0072, "step": 5000 }, { "epoch": 0.8076550387596899, "grad_norm": 95.22245025634766, "learning_rate": 1.923449612403101e-06, "loss": 0.7637, "step": 5001 }, { "epoch": 0.8078165374677002, "grad_norm": 86.39625549316406, "learning_rate": 1.9218346253229975e-06, "loss": 1.0215, "step": 5002 }, { "epoch": 0.8079780361757106, "grad_norm": 102.9690170288086, "learning_rate": 1.920219638242894e-06, "loss": 0.5039, "step": 5003 }, { "epoch": 0.8081395348837209, "grad_norm": 12.876277923583984, "learning_rate": 1.918604651162791e-06, "loss": 0.063, "step": 5004 }, { "epoch": 0.8083010335917312, "grad_norm": 179.21304321289062, "learning_rate": 1.9169896640826876e-06, "loss": 1.5664, "step": 5005 }, { "epoch": 0.8084625322997416, "grad_norm": 118.8301773071289, "learning_rate": 1.915374677002584e-06, "loss": 1.8909, "step": 5006 }, { "epoch": 0.8086240310077519, "grad_norm": 144.02691650390625, "learning_rate": 1.9137596899224807e-06, "loss": 0.6246, "step": 5007 }, { "epoch": 0.8087855297157622, "grad_norm": 60.60038757324219, "learning_rate": 1.9121447028423773e-06, "loss": 0.4294, "step": 5008 }, { "epoch": 0.8089470284237726, "grad_norm": 125.37600708007812, "learning_rate": 1.910529715762274e-06, "loss": 1.0781, "step": 5009 }, { "epoch": 0.8091085271317829, "grad_norm": 72.30793762207031, "learning_rate": 1.908914728682171e-06, "loss": 0.1857, "step": 5010 }, { "epoch": 0.8092700258397932, "grad_norm": 10.21880054473877, "learning_rate": 1.9072997416020672e-06, "loss": 0.0195, "step": 5011 }, { "epoch": 0.8094315245478036, "grad_norm": 170.46592712402344, "learning_rate": 1.905684754521964e-06, "loss": 0.7064, "step": 5012 }, { "epoch": 0.809593023255814, "grad_norm": 119.24199676513672, "learning_rate": 1.9040697674418607e-06, "loss": 0.4863, "step": 5013 }, { "epoch": 0.8097545219638242, "grad_norm": 49.46800994873047, "learning_rate": 1.902454780361757e-06, "loss": 0.3359, "step": 5014 }, { "epoch": 0.8099160206718347, "grad_norm": 23.978553771972656, "learning_rate": 1.9008397932816538e-06, "loss": 0.1033, "step": 5015 }, { "epoch": 0.810077519379845, "grad_norm": 67.77281188964844, "learning_rate": 1.8992248062015506e-06, "loss": 0.5483, "step": 5016 }, { "epoch": 0.8102390180878553, "grad_norm": 15.032849311828613, "learning_rate": 1.8976098191214474e-06, "loss": 0.0649, "step": 5017 }, { "epoch": 0.8104005167958657, "grad_norm": 20.43928337097168, "learning_rate": 1.8959948320413437e-06, "loss": 0.0523, "step": 5018 }, { "epoch": 0.810562015503876, "grad_norm": 7.867697715759277, "learning_rate": 1.8943798449612405e-06, "loss": 0.0253, "step": 5019 }, { "epoch": 0.8107235142118863, "grad_norm": 24.357641220092773, "learning_rate": 1.8927648578811373e-06, "loss": 0.0808, "step": 5020 }, { "epoch": 0.8108850129198967, "grad_norm": 211.29928588867188, "learning_rate": 1.8911498708010336e-06, "loss": 0.9474, "step": 5021 }, { "epoch": 0.811046511627907, "grad_norm": 13.772377014160156, "learning_rate": 1.8895348837209304e-06, "loss": 0.0468, "step": 5022 }, { "epoch": 0.8112080103359173, "grad_norm": 11.217918395996094, "learning_rate": 1.8879198966408271e-06, "loss": 0.0319, "step": 5023 }, { "epoch": 0.8113695090439277, "grad_norm": 30.06627082824707, "learning_rate": 1.8863049095607235e-06, "loss": 0.0981, "step": 5024 }, { "epoch": 0.811531007751938, "grad_norm": 45.13410568237305, "learning_rate": 1.8846899224806203e-06, "loss": 0.1323, "step": 5025 }, { "epoch": 0.8116925064599483, "grad_norm": 95.2502670288086, "learning_rate": 1.883074935400517e-06, "loss": 0.4834, "step": 5026 }, { "epoch": 0.8118540051679587, "grad_norm": 41.166648864746094, "learning_rate": 1.8814599483204136e-06, "loss": 0.1471, "step": 5027 }, { "epoch": 0.812015503875969, "grad_norm": 156.6622772216797, "learning_rate": 1.8798449612403102e-06, "loss": 1.4326, "step": 5028 }, { "epoch": 0.8121770025839793, "grad_norm": 37.39023208618164, "learning_rate": 1.878229974160207e-06, "loss": 0.3276, "step": 5029 }, { "epoch": 0.8123385012919897, "grad_norm": 36.01860427856445, "learning_rate": 1.8766149870801035e-06, "loss": 0.0742, "step": 5030 }, { "epoch": 0.8125, "grad_norm": 40.79506301879883, "learning_rate": 1.8750000000000003e-06, "loss": 0.1624, "step": 5031 }, { "epoch": 0.8126614987080103, "grad_norm": 182.32337951660156, "learning_rate": 1.8733850129198968e-06, "loss": 0.4929, "step": 5032 }, { "epoch": 0.8128229974160207, "grad_norm": 40.041568756103516, "learning_rate": 1.8717700258397934e-06, "loss": 0.1177, "step": 5033 }, { "epoch": 0.812984496124031, "grad_norm": 103.96536254882812, "learning_rate": 1.8701550387596902e-06, "loss": 0.833, "step": 5034 }, { "epoch": 0.8131459948320413, "grad_norm": 351.1397399902344, "learning_rate": 1.8685400516795867e-06, "loss": 3.167, "step": 5035 }, { "epoch": 0.8133074935400517, "grad_norm": 36.77400207519531, "learning_rate": 1.8669250645994833e-06, "loss": 0.0927, "step": 5036 }, { "epoch": 0.813468992248062, "grad_norm": 84.6493911743164, "learning_rate": 1.86531007751938e-06, "loss": 0.4561, "step": 5037 }, { "epoch": 0.8136304909560723, "grad_norm": 28.365638732910156, "learning_rate": 1.8636950904392768e-06, "loss": 0.0958, "step": 5038 }, { "epoch": 0.8137919896640827, "grad_norm": 104.63352966308594, "learning_rate": 1.8620801033591732e-06, "loss": 0.9473, "step": 5039 }, { "epoch": 0.813953488372093, "grad_norm": 67.47439575195312, "learning_rate": 1.86046511627907e-06, "loss": 0.2554, "step": 5040 }, { "epoch": 0.8141149870801033, "grad_norm": 24.751127243041992, "learning_rate": 1.8588501291989667e-06, "loss": 0.0891, "step": 5041 }, { "epoch": 0.8142764857881137, "grad_norm": 233.26686096191406, "learning_rate": 1.857235142118863e-06, "loss": 2.093, "step": 5042 }, { "epoch": 0.814437984496124, "grad_norm": 19.622983932495117, "learning_rate": 1.8556201550387598e-06, "loss": 0.0594, "step": 5043 }, { "epoch": 0.8145994832041343, "grad_norm": 13.59741497039795, "learning_rate": 1.8540051679586566e-06, "loss": 0.042, "step": 5044 }, { "epoch": 0.8147609819121447, "grad_norm": 119.6760482788086, "learning_rate": 1.852390180878553e-06, "loss": 0.8394, "step": 5045 }, { "epoch": 0.814922480620155, "grad_norm": 118.66425323486328, "learning_rate": 1.8507751937984497e-06, "loss": 0.8071, "step": 5046 }, { "epoch": 0.8150839793281653, "grad_norm": 3.784381151199341, "learning_rate": 1.8491602067183465e-06, "loss": 0.0118, "step": 5047 }, { "epoch": 0.8152454780361758, "grad_norm": 25.87114906311035, "learning_rate": 1.8475452196382433e-06, "loss": 0.1119, "step": 5048 }, { "epoch": 0.815406976744186, "grad_norm": 6.863203048706055, "learning_rate": 1.8459302325581396e-06, "loss": 0.0114, "step": 5049 }, { "epoch": 0.8155684754521964, "grad_norm": 107.41139221191406, "learning_rate": 1.8443152454780364e-06, "loss": 0.6997, "step": 5050 }, { "epoch": 0.8157299741602068, "grad_norm": 28.990583419799805, "learning_rate": 1.8427002583979332e-06, "loss": 0.091, "step": 5051 }, { "epoch": 0.8158914728682171, "grad_norm": 9.072525024414062, "learning_rate": 1.8410852713178295e-06, "loss": 0.0151, "step": 5052 }, { "epoch": 0.8160529715762274, "grad_norm": 84.69747924804688, "learning_rate": 1.8394702842377263e-06, "loss": 0.178, "step": 5053 }, { "epoch": 0.8162144702842378, "grad_norm": 23.02843475341797, "learning_rate": 1.837855297157623e-06, "loss": 0.1256, "step": 5054 }, { "epoch": 0.8163759689922481, "grad_norm": 65.01831817626953, "learning_rate": 1.8362403100775194e-06, "loss": 0.2979, "step": 5055 }, { "epoch": 0.8165374677002584, "grad_norm": 47.84524917602539, "learning_rate": 1.8346253229974162e-06, "loss": 0.3061, "step": 5056 }, { "epoch": 0.8166989664082688, "grad_norm": 101.72400665283203, "learning_rate": 1.833010335917313e-06, "loss": 0.3182, "step": 5057 }, { "epoch": 0.8168604651162791, "grad_norm": 18.531177520751953, "learning_rate": 1.8313953488372093e-06, "loss": 0.1094, "step": 5058 }, { "epoch": 0.8170219638242894, "grad_norm": 40.48749542236328, "learning_rate": 1.829780361757106e-06, "loss": 0.2922, "step": 5059 }, { "epoch": 0.8171834625322998, "grad_norm": 32.945228576660156, "learning_rate": 1.8281653746770028e-06, "loss": 0.1602, "step": 5060 }, { "epoch": 0.8173449612403101, "grad_norm": 41.88206100463867, "learning_rate": 1.8265503875968994e-06, "loss": 0.1696, "step": 5061 }, { "epoch": 0.8175064599483204, "grad_norm": 7.044802665710449, "learning_rate": 1.824935400516796e-06, "loss": 0.0182, "step": 5062 }, { "epoch": 0.8176679586563308, "grad_norm": 89.76817321777344, "learning_rate": 1.8233204134366927e-06, "loss": 1.3652, "step": 5063 }, { "epoch": 0.8178294573643411, "grad_norm": 65.7735366821289, "learning_rate": 1.8217054263565893e-06, "loss": 0.4563, "step": 5064 }, { "epoch": 0.8179909560723514, "grad_norm": 499.4571228027344, "learning_rate": 1.8200904392764858e-06, "loss": 1.7979, "step": 5065 }, { "epoch": 0.8181524547803618, "grad_norm": 19.91293716430664, "learning_rate": 1.8184754521963826e-06, "loss": 0.1664, "step": 5066 }, { "epoch": 0.8183139534883721, "grad_norm": 432.7888488769531, "learning_rate": 1.8168604651162792e-06, "loss": 3.3474, "step": 5067 }, { "epoch": 0.8184754521963824, "grad_norm": 46.36208724975586, "learning_rate": 1.815245478036176e-06, "loss": 0.293, "step": 5068 }, { "epoch": 0.8186369509043928, "grad_norm": 155.6041717529297, "learning_rate": 1.8136304909560725e-06, "loss": 0.7856, "step": 5069 }, { "epoch": 0.8187984496124031, "grad_norm": 110.1429214477539, "learning_rate": 1.812015503875969e-06, "loss": 0.592, "step": 5070 }, { "epoch": 0.8189599483204134, "grad_norm": 80.23467254638672, "learning_rate": 1.8104005167958658e-06, "loss": 1.0859, "step": 5071 }, { "epoch": 0.8191214470284238, "grad_norm": 115.07328033447266, "learning_rate": 1.8087855297157624e-06, "loss": 0.6895, "step": 5072 }, { "epoch": 0.8192829457364341, "grad_norm": 187.86666870117188, "learning_rate": 1.807170542635659e-06, "loss": 4.3965, "step": 5073 }, { "epoch": 0.8194444444444444, "grad_norm": 176.30618286132812, "learning_rate": 1.8055555555555557e-06, "loss": 0.6804, "step": 5074 }, { "epoch": 0.8196059431524548, "grad_norm": 8.361760139465332, "learning_rate": 1.8039405684754525e-06, "loss": 0.0304, "step": 5075 }, { "epoch": 0.8197674418604651, "grad_norm": 255.37078857421875, "learning_rate": 1.8023255813953488e-06, "loss": 3.7242, "step": 5076 }, { "epoch": 0.8199289405684754, "grad_norm": 182.4302520751953, "learning_rate": 1.8007105943152456e-06, "loss": 0.5176, "step": 5077 }, { "epoch": 0.8200904392764858, "grad_norm": 44.10890579223633, "learning_rate": 1.7990956072351424e-06, "loss": 0.1343, "step": 5078 }, { "epoch": 0.8202519379844961, "grad_norm": 82.66439819335938, "learning_rate": 1.7974806201550387e-06, "loss": 0.4683, "step": 5079 }, { "epoch": 0.8204134366925064, "grad_norm": 6.79079008102417, "learning_rate": 1.7958656330749355e-06, "loss": 0.0289, "step": 5080 }, { "epoch": 0.8205749354005168, "grad_norm": 23.145429611206055, "learning_rate": 1.7942506459948323e-06, "loss": 0.0327, "step": 5081 }, { "epoch": 0.8207364341085271, "grad_norm": 2.8210904598236084, "learning_rate": 1.792635658914729e-06, "loss": 0.0076, "step": 5082 }, { "epoch": 0.8208979328165374, "grad_norm": 59.037227630615234, "learning_rate": 1.7910206718346254e-06, "loss": 0.5699, "step": 5083 }, { "epoch": 0.8210594315245479, "grad_norm": 6.923072338104248, "learning_rate": 1.7894056847545222e-06, "loss": 0.0209, "step": 5084 }, { "epoch": 0.8212209302325582, "grad_norm": 183.7178497314453, "learning_rate": 1.787790697674419e-06, "loss": 2.5977, "step": 5085 }, { "epoch": 0.8213824289405685, "grad_norm": 97.9183120727539, "learning_rate": 1.7861757105943153e-06, "loss": 0.8418, "step": 5086 }, { "epoch": 0.8215439276485789, "grad_norm": 132.8774871826172, "learning_rate": 1.784560723514212e-06, "loss": 1.3516, "step": 5087 }, { "epoch": 0.8217054263565892, "grad_norm": 11.050450325012207, "learning_rate": 1.7829457364341088e-06, "loss": 0.0258, "step": 5088 }, { "epoch": 0.8218669250645995, "grad_norm": 61.384517669677734, "learning_rate": 1.7813307493540052e-06, "loss": 0.291, "step": 5089 }, { "epoch": 0.8220284237726099, "grad_norm": 238.8555450439453, "learning_rate": 1.779715762273902e-06, "loss": 0.7676, "step": 5090 }, { "epoch": 0.8221899224806202, "grad_norm": 44.22385787963867, "learning_rate": 1.7781007751937987e-06, "loss": 0.3633, "step": 5091 }, { "epoch": 0.8223514211886305, "grad_norm": 30.481361389160156, "learning_rate": 1.776485788113695e-06, "loss": 0.1206, "step": 5092 }, { "epoch": 0.8225129198966409, "grad_norm": 32.68006896972656, "learning_rate": 1.7748708010335918e-06, "loss": 0.0589, "step": 5093 }, { "epoch": 0.8226744186046512, "grad_norm": 125.90088653564453, "learning_rate": 1.7732558139534886e-06, "loss": 0.4913, "step": 5094 }, { "epoch": 0.8228359173126615, "grad_norm": 37.768184661865234, "learning_rate": 1.771640826873385e-06, "loss": 0.1116, "step": 5095 }, { "epoch": 0.8229974160206718, "grad_norm": 15.048967361450195, "learning_rate": 1.7700258397932817e-06, "loss": 0.0379, "step": 5096 }, { "epoch": 0.8231589147286822, "grad_norm": 20.276769638061523, "learning_rate": 1.7684108527131785e-06, "loss": 0.0945, "step": 5097 }, { "epoch": 0.8233204134366925, "grad_norm": 7.334293842315674, "learning_rate": 1.766795865633075e-06, "loss": 0.0396, "step": 5098 }, { "epoch": 0.8234819121447028, "grad_norm": 137.06179809570312, "learning_rate": 1.7651808785529716e-06, "loss": 0.2532, "step": 5099 }, { "epoch": 0.8236434108527132, "grad_norm": 12.903780937194824, "learning_rate": 1.7635658914728684e-06, "loss": 0.0379, "step": 5100 }, { "epoch": 0.8238049095607235, "grad_norm": 80.9657974243164, "learning_rate": 1.761950904392765e-06, "loss": 0.2724, "step": 5101 }, { "epoch": 0.8239664082687338, "grad_norm": 76.53411102294922, "learning_rate": 1.7603359173126615e-06, "loss": 1.0151, "step": 5102 }, { "epoch": 0.8241279069767442, "grad_norm": 93.75109100341797, "learning_rate": 1.7587209302325583e-06, "loss": 0.23, "step": 5103 }, { "epoch": 0.8242894056847545, "grad_norm": 66.07174682617188, "learning_rate": 1.7571059431524549e-06, "loss": 0.2656, "step": 5104 }, { "epoch": 0.8244509043927648, "grad_norm": 33.49079132080078, "learning_rate": 1.7554909560723516e-06, "loss": 0.0698, "step": 5105 }, { "epoch": 0.8246124031007752, "grad_norm": 166.82008361816406, "learning_rate": 1.7538759689922482e-06, "loss": 0.9642, "step": 5106 }, { "epoch": 0.8247739018087855, "grad_norm": 33.5915641784668, "learning_rate": 1.7522609819121447e-06, "loss": 0.1497, "step": 5107 }, { "epoch": 0.8249354005167958, "grad_norm": 14.441758155822754, "learning_rate": 1.7506459948320415e-06, "loss": 0.0573, "step": 5108 }, { "epoch": 0.8250968992248062, "grad_norm": 52.30284881591797, "learning_rate": 1.749031007751938e-06, "loss": 0.3352, "step": 5109 }, { "epoch": 0.8252583979328165, "grad_norm": 35.851871490478516, "learning_rate": 1.7474160206718346e-06, "loss": 0.1113, "step": 5110 }, { "epoch": 0.8254198966408268, "grad_norm": 128.93020629882812, "learning_rate": 1.7458010335917314e-06, "loss": 0.7824, "step": 5111 }, { "epoch": 0.8255813953488372, "grad_norm": 15.509716033935547, "learning_rate": 1.7441860465116282e-06, "loss": 0.051, "step": 5112 }, { "epoch": 0.8257428940568475, "grad_norm": 52.62364196777344, "learning_rate": 1.7425710594315247e-06, "loss": 0.1124, "step": 5113 }, { "epoch": 0.8259043927648578, "grad_norm": 28.76874542236328, "learning_rate": 1.7409560723514213e-06, "loss": 0.1443, "step": 5114 }, { "epoch": 0.8260658914728682, "grad_norm": 37.187252044677734, "learning_rate": 1.739341085271318e-06, "loss": 0.1484, "step": 5115 }, { "epoch": 0.8262273901808785, "grad_norm": 390.60430908203125, "learning_rate": 1.7377260981912146e-06, "loss": 2.1172, "step": 5116 }, { "epoch": 0.8263888888888888, "grad_norm": 26.964895248413086, "learning_rate": 1.7361111111111112e-06, "loss": 0.0692, "step": 5117 }, { "epoch": 0.8265503875968992, "grad_norm": 425.0381164550781, "learning_rate": 1.734496124031008e-06, "loss": 5.1406, "step": 5118 }, { "epoch": 0.8267118863049095, "grad_norm": 62.9305305480957, "learning_rate": 1.7328811369509047e-06, "loss": 0.252, "step": 5119 }, { "epoch": 0.8268733850129198, "grad_norm": 8.622968673706055, "learning_rate": 1.731266149870801e-06, "loss": 0.0311, "step": 5120 }, { "epoch": 0.8270348837209303, "grad_norm": 280.46148681640625, "learning_rate": 1.7296511627906979e-06, "loss": 1.0476, "step": 5121 }, { "epoch": 0.8271963824289406, "grad_norm": 40.21464157104492, "learning_rate": 1.7280361757105946e-06, "loss": 0.1567, "step": 5122 }, { "epoch": 0.8273578811369509, "grad_norm": 205.3932342529297, "learning_rate": 1.726421188630491e-06, "loss": 0.7285, "step": 5123 }, { "epoch": 0.8275193798449613, "grad_norm": 16.276779174804688, "learning_rate": 1.7248062015503877e-06, "loss": 0.0343, "step": 5124 }, { "epoch": 0.8276808785529716, "grad_norm": 160.56814575195312, "learning_rate": 1.7231912144702845e-06, "loss": 0.7637, "step": 5125 }, { "epoch": 0.8278423772609819, "grad_norm": 95.78390502929688, "learning_rate": 1.7215762273901809e-06, "loss": 0.2796, "step": 5126 }, { "epoch": 0.8280038759689923, "grad_norm": 161.30274963378906, "learning_rate": 1.7199612403100776e-06, "loss": 1.1758, "step": 5127 }, { "epoch": 0.8281653746770026, "grad_norm": 38.677181243896484, "learning_rate": 1.7183462532299744e-06, "loss": 0.2012, "step": 5128 }, { "epoch": 0.8283268733850129, "grad_norm": 17.18783187866211, "learning_rate": 1.7167312661498708e-06, "loss": 0.0933, "step": 5129 }, { "epoch": 0.8284883720930233, "grad_norm": 33.16604995727539, "learning_rate": 1.7151162790697675e-06, "loss": 0.1577, "step": 5130 }, { "epoch": 0.8286498708010336, "grad_norm": 14.027776718139648, "learning_rate": 1.7135012919896643e-06, "loss": 0.0426, "step": 5131 }, { "epoch": 0.8288113695090439, "grad_norm": 103.43582916259766, "learning_rate": 1.7118863049095609e-06, "loss": 0.4834, "step": 5132 }, { "epoch": 0.8289728682170543, "grad_norm": 61.270450592041016, "learning_rate": 1.7102713178294574e-06, "loss": 0.4314, "step": 5133 }, { "epoch": 0.8291343669250646, "grad_norm": 148.840087890625, "learning_rate": 1.7086563307493542e-06, "loss": 1.5859, "step": 5134 }, { "epoch": 0.8292958656330749, "grad_norm": 51.54362869262695, "learning_rate": 1.7070413436692508e-06, "loss": 0.1699, "step": 5135 }, { "epoch": 0.8294573643410853, "grad_norm": 21.42789649963379, "learning_rate": 1.7054263565891473e-06, "loss": 0.0806, "step": 5136 }, { "epoch": 0.8296188630490956, "grad_norm": 29.32561683654785, "learning_rate": 1.703811369509044e-06, "loss": 0.1741, "step": 5137 }, { "epoch": 0.8297803617571059, "grad_norm": 19.428741455078125, "learning_rate": 1.7021963824289406e-06, "loss": 0.094, "step": 5138 }, { "epoch": 0.8299418604651163, "grad_norm": 102.41084289550781, "learning_rate": 1.7005813953488374e-06, "loss": 0.2922, "step": 5139 }, { "epoch": 0.8301033591731266, "grad_norm": 31.493349075317383, "learning_rate": 1.698966408268734e-06, "loss": 0.1726, "step": 5140 }, { "epoch": 0.8302648578811369, "grad_norm": 33.540748596191406, "learning_rate": 1.6973514211886305e-06, "loss": 0.2061, "step": 5141 }, { "epoch": 0.8304263565891473, "grad_norm": 288.02801513671875, "learning_rate": 1.6957364341085273e-06, "loss": 0.4648, "step": 5142 }, { "epoch": 0.8305878552971576, "grad_norm": 33.223533630371094, "learning_rate": 1.6941214470284239e-06, "loss": 0.1855, "step": 5143 }, { "epoch": 0.8307493540051679, "grad_norm": 286.9012451171875, "learning_rate": 1.6925064599483206e-06, "loss": 0.656, "step": 5144 }, { "epoch": 0.8309108527131783, "grad_norm": 63.51129150390625, "learning_rate": 1.6908914728682172e-06, "loss": 0.2804, "step": 5145 }, { "epoch": 0.8310723514211886, "grad_norm": 3.9150049686431885, "learning_rate": 1.689276485788114e-06, "loss": 0.0084, "step": 5146 }, { "epoch": 0.8312338501291989, "grad_norm": 30.3436279296875, "learning_rate": 1.6876614987080105e-06, "loss": 0.3271, "step": 5147 }, { "epoch": 0.8313953488372093, "grad_norm": 156.75531005859375, "learning_rate": 1.686046511627907e-06, "loss": 1.1997, "step": 5148 }, { "epoch": 0.8315568475452196, "grad_norm": 270.3425598144531, "learning_rate": 1.6844315245478039e-06, "loss": 2.2656, "step": 5149 }, { "epoch": 0.8317183462532299, "grad_norm": 242.82940673828125, "learning_rate": 1.6828165374677004e-06, "loss": 1.5352, "step": 5150 }, { "epoch": 0.8318798449612403, "grad_norm": 54.41334533691406, "learning_rate": 1.681201550387597e-06, "loss": 0.201, "step": 5151 }, { "epoch": 0.8320413436692506, "grad_norm": 198.01268005371094, "learning_rate": 1.6795865633074938e-06, "loss": 0.7012, "step": 5152 }, { "epoch": 0.8322028423772609, "grad_norm": 38.71493911743164, "learning_rate": 1.6779715762273905e-06, "loss": 0.1603, "step": 5153 }, { "epoch": 0.8323643410852714, "grad_norm": 9.772201538085938, "learning_rate": 1.6763565891472869e-06, "loss": 0.0275, "step": 5154 }, { "epoch": 0.8325258397932817, "grad_norm": 7.224889278411865, "learning_rate": 1.6747416020671837e-06, "loss": 0.0176, "step": 5155 }, { "epoch": 0.832687338501292, "grad_norm": 77.9144287109375, "learning_rate": 1.6731266149870804e-06, "loss": 0.3503, "step": 5156 }, { "epoch": 0.8328488372093024, "grad_norm": 97.67327117919922, "learning_rate": 1.6715116279069768e-06, "loss": 1.2517, "step": 5157 }, { "epoch": 0.8330103359173127, "grad_norm": 21.802867889404297, "learning_rate": 1.6698966408268735e-06, "loss": 0.1181, "step": 5158 }, { "epoch": 0.833171834625323, "grad_norm": 29.893251419067383, "learning_rate": 1.6682816537467703e-06, "loss": 0.0916, "step": 5159 }, { "epoch": 0.8333333333333334, "grad_norm": 48.05184555053711, "learning_rate": 1.6666666666666667e-06, "loss": 0.2417, "step": 5160 }, { "epoch": 0.8334948320413437, "grad_norm": 9.809296607971191, "learning_rate": 1.6650516795865634e-06, "loss": 0.0282, "step": 5161 }, { "epoch": 0.833656330749354, "grad_norm": 365.6551208496094, "learning_rate": 1.6634366925064602e-06, "loss": 0.5434, "step": 5162 }, { "epoch": 0.8338178294573644, "grad_norm": 47.05787658691406, "learning_rate": 1.6618217054263566e-06, "loss": 0.2319, "step": 5163 }, { "epoch": 0.8339793281653747, "grad_norm": 7.3359599113464355, "learning_rate": 1.6602067183462533e-06, "loss": 0.0259, "step": 5164 }, { "epoch": 0.834140826873385, "grad_norm": 77.57971954345703, "learning_rate": 1.65859173126615e-06, "loss": 0.1899, "step": 5165 }, { "epoch": 0.8343023255813954, "grad_norm": 8.963828086853027, "learning_rate": 1.6569767441860464e-06, "loss": 0.0198, "step": 5166 }, { "epoch": 0.8344638242894057, "grad_norm": 175.8299560546875, "learning_rate": 1.6553617571059432e-06, "loss": 0.6519, "step": 5167 }, { "epoch": 0.834625322997416, "grad_norm": 57.555389404296875, "learning_rate": 1.65374677002584e-06, "loss": 0.1588, "step": 5168 }, { "epoch": 0.8347868217054264, "grad_norm": 16.47736167907715, "learning_rate": 1.6521317829457365e-06, "loss": 0.0316, "step": 5169 }, { "epoch": 0.8349483204134367, "grad_norm": 23.457372665405273, "learning_rate": 1.6505167958656331e-06, "loss": 0.1139, "step": 5170 }, { "epoch": 0.835109819121447, "grad_norm": 49.20857238769531, "learning_rate": 1.6489018087855299e-06, "loss": 0.3787, "step": 5171 }, { "epoch": 0.8352713178294574, "grad_norm": 52.27215576171875, "learning_rate": 1.6472868217054264e-06, "loss": 0.1812, "step": 5172 }, { "epoch": 0.8354328165374677, "grad_norm": 130.7491455078125, "learning_rate": 1.645671834625323e-06, "loss": 0.7566, "step": 5173 }, { "epoch": 0.835594315245478, "grad_norm": 17.885095596313477, "learning_rate": 1.6440568475452198e-06, "loss": 0.056, "step": 5174 }, { "epoch": 0.8357558139534884, "grad_norm": 124.66873931884766, "learning_rate": 1.6424418604651163e-06, "loss": 1.0679, "step": 5175 }, { "epoch": 0.8359173126614987, "grad_norm": 64.41826629638672, "learning_rate": 1.640826873385013e-06, "loss": 0.3185, "step": 5176 }, { "epoch": 0.836078811369509, "grad_norm": 15.336202621459961, "learning_rate": 1.6392118863049097e-06, "loss": 0.0304, "step": 5177 }, { "epoch": 0.8362403100775194, "grad_norm": 127.72416687011719, "learning_rate": 1.6375968992248064e-06, "loss": 1.3252, "step": 5178 }, { "epoch": 0.8364018087855297, "grad_norm": 164.83045959472656, "learning_rate": 1.635981912144703e-06, "loss": 1.5728, "step": 5179 }, { "epoch": 0.83656330749354, "grad_norm": 130.27601623535156, "learning_rate": 1.6343669250645996e-06, "loss": 0.4788, "step": 5180 }, { "epoch": 0.8367248062015504, "grad_norm": 28.759841918945312, "learning_rate": 1.6327519379844963e-06, "loss": 0.0603, "step": 5181 }, { "epoch": 0.8368863049095607, "grad_norm": 21.539466857910156, "learning_rate": 1.6311369509043929e-06, "loss": 0.089, "step": 5182 }, { "epoch": 0.837047803617571, "grad_norm": 225.9366455078125, "learning_rate": 1.6295219638242897e-06, "loss": 0.6875, "step": 5183 }, { "epoch": 0.8372093023255814, "grad_norm": 92.6092300415039, "learning_rate": 1.6279069767441862e-06, "loss": 0.7173, "step": 5184 }, { "epoch": 0.8373708010335917, "grad_norm": 4.685151100158691, "learning_rate": 1.6262919896640828e-06, "loss": 0.0085, "step": 5185 }, { "epoch": 0.837532299741602, "grad_norm": 32.899620056152344, "learning_rate": 1.6246770025839796e-06, "loss": 0.1628, "step": 5186 }, { "epoch": 0.8376937984496124, "grad_norm": 54.628761291503906, "learning_rate": 1.6230620155038761e-06, "loss": 0.2927, "step": 5187 }, { "epoch": 0.8378552971576227, "grad_norm": 105.55955505371094, "learning_rate": 1.6214470284237727e-06, "loss": 0.5737, "step": 5188 }, { "epoch": 0.838016795865633, "grad_norm": 42.251251220703125, "learning_rate": 1.6198320413436694e-06, "loss": 0.1389, "step": 5189 }, { "epoch": 0.8381782945736435, "grad_norm": 15.421539306640625, "learning_rate": 1.6182170542635662e-06, "loss": 0.0394, "step": 5190 }, { "epoch": 0.8383397932816538, "grad_norm": 208.64666748046875, "learning_rate": 1.6166020671834626e-06, "loss": 0.626, "step": 5191 }, { "epoch": 0.838501291989664, "grad_norm": 152.017822265625, "learning_rate": 1.6149870801033593e-06, "loss": 1.644, "step": 5192 }, { "epoch": 0.8386627906976745, "grad_norm": 99.24620056152344, "learning_rate": 1.6133720930232561e-06, "loss": 0.6311, "step": 5193 }, { "epoch": 0.8388242894056848, "grad_norm": 12.174884796142578, "learning_rate": 1.6117571059431525e-06, "loss": 0.0245, "step": 5194 }, { "epoch": 0.8389857881136951, "grad_norm": 2.0342867374420166, "learning_rate": 1.6101421188630492e-06, "loss": 0.0077, "step": 5195 }, { "epoch": 0.8391472868217055, "grad_norm": 66.54890441894531, "learning_rate": 1.608527131782946e-06, "loss": 0.3398, "step": 5196 }, { "epoch": 0.8393087855297158, "grad_norm": 57.63433837890625, "learning_rate": 1.6069121447028423e-06, "loss": 0.272, "step": 5197 }, { "epoch": 0.8394702842377261, "grad_norm": 39.86158752441406, "learning_rate": 1.6052971576227391e-06, "loss": 0.1611, "step": 5198 }, { "epoch": 0.8396317829457365, "grad_norm": 83.3412857055664, "learning_rate": 1.6036821705426359e-06, "loss": 0.6856, "step": 5199 }, { "epoch": 0.8397932816537468, "grad_norm": 8.739989280700684, "learning_rate": 1.6020671834625322e-06, "loss": 0.0222, "step": 5200 }, { "epoch": 0.8399547803617571, "grad_norm": 5.289313793182373, "learning_rate": 1.600452196382429e-06, "loss": 0.014, "step": 5201 }, { "epoch": 0.8401162790697675, "grad_norm": 73.92407989501953, "learning_rate": 1.5988372093023258e-06, "loss": 0.4674, "step": 5202 }, { "epoch": 0.8402777777777778, "grad_norm": 130.8268585205078, "learning_rate": 1.5972222222222221e-06, "loss": 1.1371, "step": 5203 }, { "epoch": 0.8404392764857881, "grad_norm": 20.93178939819336, "learning_rate": 1.595607235142119e-06, "loss": 0.0907, "step": 5204 }, { "epoch": 0.8406007751937985, "grad_norm": 126.18985748291016, "learning_rate": 1.5939922480620157e-06, "loss": 0.2422, "step": 5205 }, { "epoch": 0.8407622739018088, "grad_norm": 2.832980155944824, "learning_rate": 1.5923772609819122e-06, "loss": 0.011, "step": 5206 }, { "epoch": 0.8409237726098191, "grad_norm": 212.90760803222656, "learning_rate": 1.5907622739018088e-06, "loss": 2.1641, "step": 5207 }, { "epoch": 0.8410852713178295, "grad_norm": 53.4998664855957, "learning_rate": 1.5891472868217056e-06, "loss": 0.121, "step": 5208 }, { "epoch": 0.8412467700258398, "grad_norm": 33.67383575439453, "learning_rate": 1.5875322997416023e-06, "loss": 0.3265, "step": 5209 }, { "epoch": 0.8414082687338501, "grad_norm": 102.73135375976562, "learning_rate": 1.5859173126614987e-06, "loss": 0.2307, "step": 5210 }, { "epoch": 0.8415697674418605, "grad_norm": 24.59977149963379, "learning_rate": 1.5843023255813955e-06, "loss": 0.1244, "step": 5211 }, { "epoch": 0.8417312661498708, "grad_norm": 13.838600158691406, "learning_rate": 1.5826873385012922e-06, "loss": 0.0352, "step": 5212 }, { "epoch": 0.8418927648578811, "grad_norm": 285.7311096191406, "learning_rate": 1.5810723514211888e-06, "loss": 0.6481, "step": 5213 }, { "epoch": 0.8420542635658915, "grad_norm": 100.25279235839844, "learning_rate": 1.5794573643410853e-06, "loss": 0.4863, "step": 5214 }, { "epoch": 0.8422157622739018, "grad_norm": 42.155548095703125, "learning_rate": 1.5778423772609821e-06, "loss": 0.095, "step": 5215 }, { "epoch": 0.8423772609819121, "grad_norm": 26.51299476623535, "learning_rate": 1.5762273901808787e-06, "loss": 0.0752, "step": 5216 }, { "epoch": 0.8425387596899225, "grad_norm": 5.190813064575195, "learning_rate": 1.5746124031007752e-06, "loss": 0.0212, "step": 5217 }, { "epoch": 0.8427002583979328, "grad_norm": 28.469690322875977, "learning_rate": 1.572997416020672e-06, "loss": 0.1292, "step": 5218 }, { "epoch": 0.8428617571059431, "grad_norm": 31.545589447021484, "learning_rate": 1.5713824289405686e-06, "loss": 0.0643, "step": 5219 }, { "epoch": 0.8430232558139535, "grad_norm": 14.074322700500488, "learning_rate": 1.5697674418604653e-06, "loss": 0.0442, "step": 5220 }, { "epoch": 0.8431847545219638, "grad_norm": 32.92824172973633, "learning_rate": 1.568152454780362e-06, "loss": 0.1217, "step": 5221 }, { "epoch": 0.8433462532299741, "grad_norm": 81.08119201660156, "learning_rate": 1.5665374677002585e-06, "loss": 0.4448, "step": 5222 }, { "epoch": 0.8435077519379846, "grad_norm": 14.78087329864502, "learning_rate": 1.5649224806201552e-06, "loss": 0.0323, "step": 5223 }, { "epoch": 0.8436692506459949, "grad_norm": 15.806533813476562, "learning_rate": 1.563307493540052e-06, "loss": 0.0472, "step": 5224 }, { "epoch": 0.8438307493540051, "grad_norm": 15.371112823486328, "learning_rate": 1.5616925064599484e-06, "loss": 0.044, "step": 5225 }, { "epoch": 0.8439922480620154, "grad_norm": 74.38631439208984, "learning_rate": 1.5600775193798451e-06, "loss": 0.2133, "step": 5226 }, { "epoch": 0.8441537467700259, "grad_norm": 57.69782257080078, "learning_rate": 1.558462532299742e-06, "loss": 0.1217, "step": 5227 }, { "epoch": 0.8443152454780362, "grad_norm": 21.661502838134766, "learning_rate": 1.5568475452196382e-06, "loss": 0.0505, "step": 5228 }, { "epoch": 0.8444767441860465, "grad_norm": 26.225112915039062, "learning_rate": 1.555232558139535e-06, "loss": 0.0932, "step": 5229 }, { "epoch": 0.8446382428940569, "grad_norm": 98.43328857421875, "learning_rate": 1.5536175710594318e-06, "loss": 0.7796, "step": 5230 }, { "epoch": 0.8447997416020672, "grad_norm": 72.72222900390625, "learning_rate": 1.5520025839793281e-06, "loss": 0.2456, "step": 5231 }, { "epoch": 0.8449612403100775, "grad_norm": 18.67633056640625, "learning_rate": 1.550387596899225e-06, "loss": 0.0485, "step": 5232 }, { "epoch": 0.8451227390180879, "grad_norm": 122.23152160644531, "learning_rate": 1.5487726098191217e-06, "loss": 0.5489, "step": 5233 }, { "epoch": 0.8452842377260982, "grad_norm": 28.37898826599121, "learning_rate": 1.547157622739018e-06, "loss": 0.2002, "step": 5234 }, { "epoch": 0.8454457364341085, "grad_norm": 303.9608154296875, "learning_rate": 1.5455426356589148e-06, "loss": 2.3047, "step": 5235 }, { "epoch": 0.8456072351421189, "grad_norm": 15.378299713134766, "learning_rate": 1.5439276485788116e-06, "loss": 0.0396, "step": 5236 }, { "epoch": 0.8457687338501292, "grad_norm": 186.79449462890625, "learning_rate": 1.542312661498708e-06, "loss": 1.4043, "step": 5237 }, { "epoch": 0.8459302325581395, "grad_norm": 11.13858413696289, "learning_rate": 1.5406976744186047e-06, "loss": 0.0157, "step": 5238 }, { "epoch": 0.8460917312661499, "grad_norm": 118.04178619384766, "learning_rate": 1.5390826873385015e-06, "loss": 1.2139, "step": 5239 }, { "epoch": 0.8462532299741602, "grad_norm": 22.682661056518555, "learning_rate": 1.537467700258398e-06, "loss": 0.0793, "step": 5240 }, { "epoch": 0.8464147286821705, "grad_norm": 11.22689437866211, "learning_rate": 1.5358527131782946e-06, "loss": 0.0272, "step": 5241 }, { "epoch": 0.8465762273901809, "grad_norm": 175.69142150878906, "learning_rate": 1.5342377260981914e-06, "loss": 1.7578, "step": 5242 }, { "epoch": 0.8467377260981912, "grad_norm": 39.970359802246094, "learning_rate": 1.5326227390180881e-06, "loss": 0.1558, "step": 5243 }, { "epoch": 0.8468992248062015, "grad_norm": 36.63422393798828, "learning_rate": 1.5310077519379845e-06, "loss": 0.2614, "step": 5244 }, { "epoch": 0.8470607235142119, "grad_norm": 126.61371612548828, "learning_rate": 1.5293927648578812e-06, "loss": 0.4824, "step": 5245 }, { "epoch": 0.8472222222222222, "grad_norm": 10.29622745513916, "learning_rate": 1.527777777777778e-06, "loss": 0.0376, "step": 5246 }, { "epoch": 0.8473837209302325, "grad_norm": 75.46931457519531, "learning_rate": 1.5261627906976746e-06, "loss": 0.7017, "step": 5247 }, { "epoch": 0.8475452196382429, "grad_norm": 23.22909164428711, "learning_rate": 1.5245478036175711e-06, "loss": 0.0701, "step": 5248 }, { "epoch": 0.8477067183462532, "grad_norm": 92.05826568603516, "learning_rate": 1.522932816537468e-06, "loss": 0.2092, "step": 5249 }, { "epoch": 0.8478682170542635, "grad_norm": 69.85411071777344, "learning_rate": 1.5213178294573645e-06, "loss": 0.1625, "step": 5250 }, { "epoch": 0.8480297157622739, "grad_norm": 66.0870361328125, "learning_rate": 1.519702842377261e-06, "loss": 0.1875, "step": 5251 }, { "epoch": 0.8481912144702842, "grad_norm": 11.725933074951172, "learning_rate": 1.5180878552971578e-06, "loss": 0.0176, "step": 5252 }, { "epoch": 0.8483527131782945, "grad_norm": 101.13117980957031, "learning_rate": 1.5164728682170544e-06, "loss": 0.2144, "step": 5253 }, { "epoch": 0.8485142118863049, "grad_norm": 93.76177215576172, "learning_rate": 1.5148578811369511e-06, "loss": 0.7503, "step": 5254 }, { "epoch": 0.8486757105943152, "grad_norm": 482.6470642089844, "learning_rate": 1.5132428940568477e-06, "loss": 1.5273, "step": 5255 }, { "epoch": 0.8488372093023255, "grad_norm": 73.9581069946289, "learning_rate": 1.5116279069767443e-06, "loss": 0.2263, "step": 5256 }, { "epoch": 0.848998708010336, "grad_norm": 243.14743041992188, "learning_rate": 1.510012919896641e-06, "loss": 0.875, "step": 5257 }, { "epoch": 0.8491602067183462, "grad_norm": 17.302627563476562, "learning_rate": 1.5083979328165376e-06, "loss": 0.0343, "step": 5258 }, { "epoch": 0.8493217054263565, "grad_norm": 120.18231201171875, "learning_rate": 1.5067829457364341e-06, "loss": 0.2993, "step": 5259 }, { "epoch": 0.849483204134367, "grad_norm": 20.51936149597168, "learning_rate": 1.505167958656331e-06, "loss": 0.0515, "step": 5260 }, { "epoch": 0.8496447028423773, "grad_norm": 45.846195220947266, "learning_rate": 1.5035529715762277e-06, "loss": 0.1761, "step": 5261 }, { "epoch": 0.8498062015503876, "grad_norm": 9.638501167297363, "learning_rate": 1.501937984496124e-06, "loss": 0.014, "step": 5262 }, { "epoch": 0.849967700258398, "grad_norm": 43.852046966552734, "learning_rate": 1.5003229974160208e-06, "loss": 0.1407, "step": 5263 }, { "epoch": 0.8501291989664083, "grad_norm": 21.882160186767578, "learning_rate": 1.4987080103359176e-06, "loss": 0.1765, "step": 5264 }, { "epoch": 0.8502906976744186, "grad_norm": 106.09591674804688, "learning_rate": 1.497093023255814e-06, "loss": 0.355, "step": 5265 }, { "epoch": 0.850452196382429, "grad_norm": 24.483112335205078, "learning_rate": 1.4954780361757107e-06, "loss": 0.0767, "step": 5266 }, { "epoch": 0.8506136950904393, "grad_norm": 19.616920471191406, "learning_rate": 1.4938630490956075e-06, "loss": 0.0729, "step": 5267 }, { "epoch": 0.8507751937984496, "grad_norm": 6.446753025054932, "learning_rate": 1.4922480620155038e-06, "loss": 0.0181, "step": 5268 }, { "epoch": 0.85093669250646, "grad_norm": 82.42825317382812, "learning_rate": 1.4906330749354006e-06, "loss": 0.5811, "step": 5269 }, { "epoch": 0.8510981912144703, "grad_norm": 25.599681854248047, "learning_rate": 1.4890180878552974e-06, "loss": 0.1184, "step": 5270 }, { "epoch": 0.8512596899224806, "grad_norm": 12.943626403808594, "learning_rate": 1.4874031007751937e-06, "loss": 0.0441, "step": 5271 }, { "epoch": 0.851421188630491, "grad_norm": 156.73477172851562, "learning_rate": 1.4857881136950905e-06, "loss": 1.6831, "step": 5272 }, { "epoch": 0.8515826873385013, "grad_norm": 82.68549346923828, "learning_rate": 1.4841731266149873e-06, "loss": 0.3312, "step": 5273 }, { "epoch": 0.8517441860465116, "grad_norm": 56.43813705444336, "learning_rate": 1.482558139534884e-06, "loss": 0.1897, "step": 5274 }, { "epoch": 0.851905684754522, "grad_norm": 41.080055236816406, "learning_rate": 1.4809431524547804e-06, "loss": 0.1675, "step": 5275 }, { "epoch": 0.8520671834625323, "grad_norm": 10.192876815795898, "learning_rate": 1.4793281653746771e-06, "loss": 0.0304, "step": 5276 }, { "epoch": 0.8522286821705426, "grad_norm": 7.473336219787598, "learning_rate": 1.477713178294574e-06, "loss": 0.0258, "step": 5277 }, { "epoch": 0.852390180878553, "grad_norm": 175.55711364746094, "learning_rate": 1.4760981912144703e-06, "loss": 0.5935, "step": 5278 }, { "epoch": 0.8525516795865633, "grad_norm": 19.9418888092041, "learning_rate": 1.474483204134367e-06, "loss": 0.0779, "step": 5279 }, { "epoch": 0.8527131782945736, "grad_norm": 122.4635009765625, "learning_rate": 1.4728682170542638e-06, "loss": 1.3127, "step": 5280 }, { "epoch": 0.852874677002584, "grad_norm": 54.208518981933594, "learning_rate": 1.4712532299741602e-06, "loss": 0.2754, "step": 5281 }, { "epoch": 0.8530361757105943, "grad_norm": 404.2509765625, "learning_rate": 1.469638242894057e-06, "loss": 2.2809, "step": 5282 }, { "epoch": 0.8531976744186046, "grad_norm": 3.091623544692993, "learning_rate": 1.4680232558139537e-06, "loss": 0.0075, "step": 5283 }, { "epoch": 0.853359173126615, "grad_norm": 12.366004943847656, "learning_rate": 1.4664082687338503e-06, "loss": 0.0267, "step": 5284 }, { "epoch": 0.8535206718346253, "grad_norm": 19.261642456054688, "learning_rate": 1.4647932816537468e-06, "loss": 0.0542, "step": 5285 }, { "epoch": 0.8536821705426356, "grad_norm": 37.553165435791016, "learning_rate": 1.4631782945736436e-06, "loss": 0.1534, "step": 5286 }, { "epoch": 0.853843669250646, "grad_norm": 104.30696868896484, "learning_rate": 1.4615633074935402e-06, "loss": 0.5537, "step": 5287 }, { "epoch": 0.8540051679586563, "grad_norm": 4.296767234802246, "learning_rate": 1.4599483204134367e-06, "loss": 0.0057, "step": 5288 }, { "epoch": 0.8541666666666666, "grad_norm": 25.817174911499023, "learning_rate": 1.4583333333333335e-06, "loss": 0.1533, "step": 5289 }, { "epoch": 0.854328165374677, "grad_norm": 73.53780364990234, "learning_rate": 1.45671834625323e-06, "loss": 0.4375, "step": 5290 }, { "epoch": 0.8544896640826873, "grad_norm": 6.828434944152832, "learning_rate": 1.4551033591731268e-06, "loss": 0.0235, "step": 5291 }, { "epoch": 0.8546511627906976, "grad_norm": 30.343748092651367, "learning_rate": 1.4534883720930234e-06, "loss": 0.0996, "step": 5292 }, { "epoch": 0.854812661498708, "grad_norm": 42.30915069580078, "learning_rate": 1.45187338501292e-06, "loss": 0.2078, "step": 5293 }, { "epoch": 0.8549741602067183, "grad_norm": 13.076082229614258, "learning_rate": 1.4502583979328167e-06, "loss": 0.0331, "step": 5294 }, { "epoch": 0.8551356589147286, "grad_norm": 34.68808364868164, "learning_rate": 1.4486434108527133e-06, "loss": 0.106, "step": 5295 }, { "epoch": 0.8552971576227391, "grad_norm": 108.249755859375, "learning_rate": 1.4470284237726098e-06, "loss": 0.7352, "step": 5296 }, { "epoch": 0.8554586563307494, "grad_norm": 26.745790481567383, "learning_rate": 1.4454134366925066e-06, "loss": 0.144, "step": 5297 }, { "epoch": 0.8556201550387597, "grad_norm": 9.263345718383789, "learning_rate": 1.4437984496124034e-06, "loss": 0.0352, "step": 5298 }, { "epoch": 0.8557816537467701, "grad_norm": 44.691226959228516, "learning_rate": 1.4421834625322997e-06, "loss": 0.2412, "step": 5299 }, { "epoch": 0.8559431524547804, "grad_norm": 15.607769012451172, "learning_rate": 1.4405684754521965e-06, "loss": 0.0249, "step": 5300 }, { "epoch": 0.8561046511627907, "grad_norm": 50.034332275390625, "learning_rate": 1.4389534883720933e-06, "loss": 0.5322, "step": 5301 }, { "epoch": 0.8562661498708011, "grad_norm": 38.64585494995117, "learning_rate": 1.4373385012919896e-06, "loss": 0.1541, "step": 5302 }, { "epoch": 0.8564276485788114, "grad_norm": 15.711409568786621, "learning_rate": 1.4357235142118864e-06, "loss": 0.0459, "step": 5303 }, { "epoch": 0.8565891472868217, "grad_norm": 36.37254333496094, "learning_rate": 1.4341085271317832e-06, "loss": 0.2222, "step": 5304 }, { "epoch": 0.8567506459948321, "grad_norm": 83.06704711914062, "learning_rate": 1.4324935400516795e-06, "loss": 0.2966, "step": 5305 }, { "epoch": 0.8569121447028424, "grad_norm": 11.704944610595703, "learning_rate": 1.4308785529715763e-06, "loss": 0.0275, "step": 5306 }, { "epoch": 0.8570736434108527, "grad_norm": 46.796268463134766, "learning_rate": 1.429263565891473e-06, "loss": 0.2156, "step": 5307 }, { "epoch": 0.8572351421188631, "grad_norm": 16.543867111206055, "learning_rate": 1.4276485788113698e-06, "loss": 0.0643, "step": 5308 }, { "epoch": 0.8573966408268734, "grad_norm": 36.041534423828125, "learning_rate": 1.4260335917312662e-06, "loss": 0.1294, "step": 5309 }, { "epoch": 0.8575581395348837, "grad_norm": 22.7963924407959, "learning_rate": 1.424418604651163e-06, "loss": 0.1405, "step": 5310 }, { "epoch": 0.8577196382428941, "grad_norm": 19.06656265258789, "learning_rate": 1.4228036175710597e-06, "loss": 0.047, "step": 5311 }, { "epoch": 0.8578811369509044, "grad_norm": 45.8640251159668, "learning_rate": 1.421188630490956e-06, "loss": 0.1245, "step": 5312 }, { "epoch": 0.8580426356589147, "grad_norm": 15.01877212524414, "learning_rate": 1.4195736434108528e-06, "loss": 0.0354, "step": 5313 }, { "epoch": 0.8582041343669251, "grad_norm": 14.248077392578125, "learning_rate": 1.4179586563307496e-06, "loss": 0.0287, "step": 5314 }, { "epoch": 0.8583656330749354, "grad_norm": 57.34133529663086, "learning_rate": 1.416343669250646e-06, "loss": 0.1379, "step": 5315 }, { "epoch": 0.8585271317829457, "grad_norm": 7.102544784545898, "learning_rate": 1.4147286821705427e-06, "loss": 0.008, "step": 5316 }, { "epoch": 0.8586886304909561, "grad_norm": 88.28673553466797, "learning_rate": 1.4131136950904395e-06, "loss": 1.2026, "step": 5317 }, { "epoch": 0.8588501291989664, "grad_norm": 47.66817855834961, "learning_rate": 1.4114987080103358e-06, "loss": 0.1904, "step": 5318 }, { "epoch": 0.8590116279069767, "grad_norm": 13.797920227050781, "learning_rate": 1.4098837209302326e-06, "loss": 0.0308, "step": 5319 }, { "epoch": 0.8591731266149871, "grad_norm": 325.1897888183594, "learning_rate": 1.4082687338501294e-06, "loss": 0.9327, "step": 5320 }, { "epoch": 0.8593346253229974, "grad_norm": 133.5941925048828, "learning_rate": 1.406653746770026e-06, "loss": 1.8921, "step": 5321 }, { "epoch": 0.8594961240310077, "grad_norm": 68.02499389648438, "learning_rate": 1.4050387596899225e-06, "loss": 0.27, "step": 5322 }, { "epoch": 0.8596576227390181, "grad_norm": 38.322635650634766, "learning_rate": 1.4034237726098193e-06, "loss": 0.0922, "step": 5323 }, { "epoch": 0.8598191214470284, "grad_norm": 94.09502410888672, "learning_rate": 1.4018087855297158e-06, "loss": 0.2783, "step": 5324 }, { "epoch": 0.8599806201550387, "grad_norm": 54.05680847167969, "learning_rate": 1.4001937984496126e-06, "loss": 0.2067, "step": 5325 }, { "epoch": 0.8601421188630491, "grad_norm": 814.4915771484375, "learning_rate": 1.3985788113695092e-06, "loss": 1.4609, "step": 5326 }, { "epoch": 0.8603036175710594, "grad_norm": 214.06484985351562, "learning_rate": 1.3969638242894057e-06, "loss": 2.5625, "step": 5327 }, { "epoch": 0.8604651162790697, "grad_norm": 60.447044372558594, "learning_rate": 1.3953488372093025e-06, "loss": 0.3531, "step": 5328 }, { "epoch": 0.8606266149870802, "grad_norm": 11.683518409729004, "learning_rate": 1.393733850129199e-06, "loss": 0.019, "step": 5329 }, { "epoch": 0.8607881136950905, "grad_norm": 4.19122314453125, "learning_rate": 1.3921188630490956e-06, "loss": 0.0042, "step": 5330 }, { "epoch": 0.8609496124031008, "grad_norm": 85.99114990234375, "learning_rate": 1.3905038759689924e-06, "loss": 0.3828, "step": 5331 }, { "epoch": 0.8611111111111112, "grad_norm": 30.939300537109375, "learning_rate": 1.3888888888888892e-06, "loss": 0.175, "step": 5332 }, { "epoch": 0.8612726098191215, "grad_norm": 19.6437931060791, "learning_rate": 1.3872739018087855e-06, "loss": 0.0901, "step": 5333 }, { "epoch": 0.8614341085271318, "grad_norm": 93.73191833496094, "learning_rate": 1.3856589147286823e-06, "loss": 0.2928, "step": 5334 }, { "epoch": 0.8615956072351422, "grad_norm": 263.9039306640625, "learning_rate": 1.384043927648579e-06, "loss": 3.8872, "step": 5335 }, { "epoch": 0.8617571059431525, "grad_norm": 106.76187133789062, "learning_rate": 1.3824289405684754e-06, "loss": 0.3257, "step": 5336 }, { "epoch": 0.8619186046511628, "grad_norm": 84.90724182128906, "learning_rate": 1.3808139534883722e-06, "loss": 1.0238, "step": 5337 }, { "epoch": 0.8620801033591732, "grad_norm": 47.93739318847656, "learning_rate": 1.379198966408269e-06, "loss": 0.1028, "step": 5338 }, { "epoch": 0.8622416020671835, "grad_norm": 87.66361999511719, "learning_rate": 1.3775839793281657e-06, "loss": 0.3525, "step": 5339 }, { "epoch": 0.8624031007751938, "grad_norm": 62.23148727416992, "learning_rate": 1.375968992248062e-06, "loss": 0.1931, "step": 5340 }, { "epoch": 0.8625645994832042, "grad_norm": 91.81607055664062, "learning_rate": 1.3743540051679588e-06, "loss": 1.0259, "step": 5341 }, { "epoch": 0.8627260981912145, "grad_norm": 71.13398742675781, "learning_rate": 1.3727390180878556e-06, "loss": 0.7793, "step": 5342 }, { "epoch": 0.8628875968992248, "grad_norm": 59.69212341308594, "learning_rate": 1.371124031007752e-06, "loss": 0.7155, "step": 5343 }, { "epoch": 0.8630490956072352, "grad_norm": 69.8660659790039, "learning_rate": 1.3695090439276487e-06, "loss": 0.1881, "step": 5344 }, { "epoch": 0.8632105943152455, "grad_norm": 297.15966796875, "learning_rate": 1.3678940568475455e-06, "loss": 2.2374, "step": 5345 }, { "epoch": 0.8633720930232558, "grad_norm": 129.2286376953125, "learning_rate": 1.3662790697674419e-06, "loss": 1.0977, "step": 5346 }, { "epoch": 0.8635335917312662, "grad_norm": 75.1009521484375, "learning_rate": 1.3646640826873386e-06, "loss": 0.3079, "step": 5347 }, { "epoch": 0.8636950904392765, "grad_norm": 27.377681732177734, "learning_rate": 1.3630490956072354e-06, "loss": 0.0622, "step": 5348 }, { "epoch": 0.8638565891472868, "grad_norm": 9.112333297729492, "learning_rate": 1.3614341085271317e-06, "loss": 0.0204, "step": 5349 }, { "epoch": 0.8640180878552972, "grad_norm": 57.467647552490234, "learning_rate": 1.3598191214470285e-06, "loss": 0.4186, "step": 5350 }, { "epoch": 0.8641795865633075, "grad_norm": 136.8479766845703, "learning_rate": 1.3582041343669253e-06, "loss": 0.8535, "step": 5351 }, { "epoch": 0.8643410852713178, "grad_norm": 20.638891220092773, "learning_rate": 1.3565891472868216e-06, "loss": 0.0508, "step": 5352 }, { "epoch": 0.8645025839793282, "grad_norm": 6.905615329742432, "learning_rate": 1.3549741602067184e-06, "loss": 0.0163, "step": 5353 }, { "epoch": 0.8646640826873385, "grad_norm": 1022.177490234375, "learning_rate": 1.3533591731266152e-06, "loss": 2.0894, "step": 5354 }, { "epoch": 0.8648255813953488, "grad_norm": 25.994503021240234, "learning_rate": 1.3517441860465117e-06, "loss": 0.0948, "step": 5355 }, { "epoch": 0.8649870801033591, "grad_norm": 66.72040557861328, "learning_rate": 1.3501291989664083e-06, "loss": 0.4886, "step": 5356 }, { "epoch": 0.8651485788113695, "grad_norm": 17.01051902770996, "learning_rate": 1.348514211886305e-06, "loss": 0.1084, "step": 5357 }, { "epoch": 0.8653100775193798, "grad_norm": 114.10592651367188, "learning_rate": 1.3468992248062016e-06, "loss": 0.9531, "step": 5358 }, { "epoch": 0.8654715762273901, "grad_norm": 52.86206817626953, "learning_rate": 1.3452842377260982e-06, "loss": 0.1875, "step": 5359 }, { "epoch": 0.8656330749354005, "grad_norm": 61.42433166503906, "learning_rate": 1.343669250645995e-06, "loss": 0.3545, "step": 5360 }, { "epoch": 0.8657945736434108, "grad_norm": 71.51271057128906, "learning_rate": 1.3420542635658915e-06, "loss": 0.3047, "step": 5361 }, { "epoch": 0.8659560723514211, "grad_norm": 55.36314010620117, "learning_rate": 1.3404392764857883e-06, "loss": 0.2637, "step": 5362 }, { "epoch": 0.8661175710594315, "grad_norm": 21.409547805786133, "learning_rate": 1.3388242894056849e-06, "loss": 0.094, "step": 5363 }, { "epoch": 0.8662790697674418, "grad_norm": 100.23452758789062, "learning_rate": 1.3372093023255814e-06, "loss": 0.3643, "step": 5364 }, { "epoch": 0.8664405684754521, "grad_norm": 31.101219177246094, "learning_rate": 1.3355943152454782e-06, "loss": 0.2172, "step": 5365 }, { "epoch": 0.8666020671834626, "grad_norm": 13.629120826721191, "learning_rate": 1.3339793281653747e-06, "loss": 0.0599, "step": 5366 }, { "epoch": 0.8667635658914729, "grad_norm": 77.29702758789062, "learning_rate": 1.3323643410852713e-06, "loss": 0.3264, "step": 5367 }, { "epoch": 0.8669250645994832, "grad_norm": 7.352380275726318, "learning_rate": 1.330749354005168e-06, "loss": 0.0111, "step": 5368 }, { "epoch": 0.8670865633074936, "grad_norm": 149.35398864746094, "learning_rate": 1.3291343669250649e-06, "loss": 0.4697, "step": 5369 }, { "epoch": 0.8672480620155039, "grad_norm": 85.9087905883789, "learning_rate": 1.3275193798449614e-06, "loss": 0.1411, "step": 5370 }, { "epoch": 0.8674095607235142, "grad_norm": 31.258350372314453, "learning_rate": 1.325904392764858e-06, "loss": 0.1605, "step": 5371 }, { "epoch": 0.8675710594315246, "grad_norm": 21.476598739624023, "learning_rate": 1.3242894056847547e-06, "loss": 0.0449, "step": 5372 }, { "epoch": 0.8677325581395349, "grad_norm": 39.12784194946289, "learning_rate": 1.3226744186046513e-06, "loss": 0.0638, "step": 5373 }, { "epoch": 0.8678940568475452, "grad_norm": 11.819829940795898, "learning_rate": 1.3210594315245479e-06, "loss": 0.0363, "step": 5374 }, { "epoch": 0.8680555555555556, "grad_norm": 4.61168909072876, "learning_rate": 1.3194444444444446e-06, "loss": 0.0157, "step": 5375 }, { "epoch": 0.8682170542635659, "grad_norm": 148.84092712402344, "learning_rate": 1.3178294573643414e-06, "loss": 0.7773, "step": 5376 }, { "epoch": 0.8683785529715762, "grad_norm": 191.99134826660156, "learning_rate": 1.3162144702842378e-06, "loss": 1.0609, "step": 5377 }, { "epoch": 0.8685400516795866, "grad_norm": 112.68553161621094, "learning_rate": 1.3145994832041345e-06, "loss": 0.4771, "step": 5378 }, { "epoch": 0.8687015503875969, "grad_norm": 7.272781848907471, "learning_rate": 1.3129844961240313e-06, "loss": 0.0261, "step": 5379 }, { "epoch": 0.8688630490956072, "grad_norm": 249.9340057373047, "learning_rate": 1.3113695090439276e-06, "loss": 0.8545, "step": 5380 }, { "epoch": 0.8690245478036176, "grad_norm": 553.8522338867188, "learning_rate": 1.3097545219638244e-06, "loss": 2.9609, "step": 5381 }, { "epoch": 0.8691860465116279, "grad_norm": 79.06165313720703, "learning_rate": 1.3081395348837212e-06, "loss": 0.7266, "step": 5382 }, { "epoch": 0.8693475452196382, "grad_norm": 301.6236572265625, "learning_rate": 1.3065245478036175e-06, "loss": 3.373, "step": 5383 }, { "epoch": 0.8695090439276486, "grad_norm": 80.80294036865234, "learning_rate": 1.3049095607235143e-06, "loss": 0.5176, "step": 5384 }, { "epoch": 0.8696705426356589, "grad_norm": 33.720680236816406, "learning_rate": 1.303294573643411e-06, "loss": 0.1071, "step": 5385 }, { "epoch": 0.8698320413436692, "grad_norm": 83.69221496582031, "learning_rate": 1.3016795865633074e-06, "loss": 0.7114, "step": 5386 }, { "epoch": 0.8699935400516796, "grad_norm": 53.87630844116211, "learning_rate": 1.3000645994832042e-06, "loss": 0.2427, "step": 5387 }, { "epoch": 0.8701550387596899, "grad_norm": 26.434846878051758, "learning_rate": 1.298449612403101e-06, "loss": 0.0718, "step": 5388 }, { "epoch": 0.8703165374677002, "grad_norm": 43.179931640625, "learning_rate": 1.2968346253229973e-06, "loss": 0.3633, "step": 5389 }, { "epoch": 0.8704780361757106, "grad_norm": 25.51143455505371, "learning_rate": 1.295219638242894e-06, "loss": 0.0946, "step": 5390 }, { "epoch": 0.8706395348837209, "grad_norm": 49.10210418701172, "learning_rate": 1.2936046511627909e-06, "loss": 0.1448, "step": 5391 }, { "epoch": 0.8708010335917312, "grad_norm": 73.79601287841797, "learning_rate": 1.2919896640826874e-06, "loss": 0.4043, "step": 5392 }, { "epoch": 0.8709625322997416, "grad_norm": 297.4090576171875, "learning_rate": 1.290374677002584e-06, "loss": 2.3789, "step": 5393 }, { "epoch": 0.8711240310077519, "grad_norm": 190.9123077392578, "learning_rate": 1.2887596899224808e-06, "loss": 1.5072, "step": 5394 }, { "epoch": 0.8712855297157622, "grad_norm": 81.8473129272461, "learning_rate": 1.2871447028423773e-06, "loss": 0.7932, "step": 5395 }, { "epoch": 0.8714470284237726, "grad_norm": 10.181602478027344, "learning_rate": 1.2855297157622739e-06, "loss": 0.0328, "step": 5396 }, { "epoch": 0.8716085271317829, "grad_norm": 53.377620697021484, "learning_rate": 1.2839147286821706e-06, "loss": 0.2354, "step": 5397 }, { "epoch": 0.8717700258397932, "grad_norm": 32.62125015258789, "learning_rate": 1.2822997416020672e-06, "loss": 0.1486, "step": 5398 }, { "epoch": 0.8719315245478036, "grad_norm": 149.4861602783203, "learning_rate": 1.280684754521964e-06, "loss": 0.7294, "step": 5399 }, { "epoch": 0.872093023255814, "grad_norm": 21.993371963500977, "learning_rate": 1.2790697674418605e-06, "loss": 0.0507, "step": 5400 }, { "epoch": 0.8722545219638242, "grad_norm": 16.38791275024414, "learning_rate": 1.277454780361757e-06, "loss": 0.0366, "step": 5401 }, { "epoch": 0.8724160206718347, "grad_norm": 79.76123046875, "learning_rate": 1.2758397932816539e-06, "loss": 0.4556, "step": 5402 }, { "epoch": 0.872577519379845, "grad_norm": 12.802877426147461, "learning_rate": 1.2742248062015504e-06, "loss": 0.026, "step": 5403 }, { "epoch": 0.8727390180878553, "grad_norm": 60.236122131347656, "learning_rate": 1.2726098191214472e-06, "loss": 0.4659, "step": 5404 }, { "epoch": 0.8729005167958657, "grad_norm": 14.140619277954102, "learning_rate": 1.2709948320413438e-06, "loss": 0.0389, "step": 5405 }, { "epoch": 0.873062015503876, "grad_norm": 72.56256866455078, "learning_rate": 1.2693798449612405e-06, "loss": 0.4961, "step": 5406 }, { "epoch": 0.8732235142118863, "grad_norm": 42.46687316894531, "learning_rate": 1.267764857881137e-06, "loss": 0.1404, "step": 5407 }, { "epoch": 0.8733850129198967, "grad_norm": 196.6951904296875, "learning_rate": 1.2661498708010337e-06, "loss": 1.0938, "step": 5408 }, { "epoch": 0.873546511627907, "grad_norm": 82.17710876464844, "learning_rate": 1.2645348837209304e-06, "loss": 0.4956, "step": 5409 }, { "epoch": 0.8737080103359173, "grad_norm": 17.52728843688965, "learning_rate": 1.262919896640827e-06, "loss": 0.0428, "step": 5410 }, { "epoch": 0.8738695090439277, "grad_norm": 90.73661041259766, "learning_rate": 1.2613049095607235e-06, "loss": 0.75, "step": 5411 }, { "epoch": 0.874031007751938, "grad_norm": 316.5418395996094, "learning_rate": 1.2596899224806203e-06, "loss": 1.3083, "step": 5412 }, { "epoch": 0.8741925064599483, "grad_norm": 38.94427490234375, "learning_rate": 1.258074935400517e-06, "loss": 0.1155, "step": 5413 }, { "epoch": 0.8743540051679587, "grad_norm": 207.49688720703125, "learning_rate": 1.2564599483204134e-06, "loss": 0.7743, "step": 5414 }, { "epoch": 0.874515503875969, "grad_norm": 95.19855499267578, "learning_rate": 1.2548449612403102e-06, "loss": 1.2832, "step": 5415 }, { "epoch": 0.8746770025839793, "grad_norm": 53.939640045166016, "learning_rate": 1.253229974160207e-06, "loss": 0.2115, "step": 5416 }, { "epoch": 0.8748385012919897, "grad_norm": 207.6700897216797, "learning_rate": 1.2516149870801033e-06, "loss": 0.496, "step": 5417 }, { "epoch": 0.875, "grad_norm": 218.03382873535156, "learning_rate": 1.25e-06, "loss": 2.8906, "step": 5418 }, { "epoch": 0.8751614987080103, "grad_norm": 109.64811706542969, "learning_rate": 1.2483850129198967e-06, "loss": 1.2404, "step": 5419 }, { "epoch": 0.8753229974160207, "grad_norm": 124.64183807373047, "learning_rate": 1.2467700258397934e-06, "loss": 1.3506, "step": 5420 }, { "epoch": 0.875484496124031, "grad_norm": 51.96189498901367, "learning_rate": 1.24515503875969e-06, "loss": 0.3254, "step": 5421 }, { "epoch": 0.8756459948320413, "grad_norm": 147.68316650390625, "learning_rate": 1.2435400516795866e-06, "loss": 1.7969, "step": 5422 }, { "epoch": 0.8758074935400517, "grad_norm": 4.355051040649414, "learning_rate": 1.2419250645994833e-06, "loss": 0.0239, "step": 5423 }, { "epoch": 0.875968992248062, "grad_norm": 67.8182601928711, "learning_rate": 1.2403100775193799e-06, "loss": 0.1953, "step": 5424 }, { "epoch": 0.8761304909560723, "grad_norm": 69.93702697753906, "learning_rate": 1.2386950904392764e-06, "loss": 0.3252, "step": 5425 }, { "epoch": 0.8762919896640827, "grad_norm": 17.933290481567383, "learning_rate": 1.2370801033591732e-06, "loss": 0.0564, "step": 5426 }, { "epoch": 0.876453488372093, "grad_norm": 11.351962089538574, "learning_rate": 1.2354651162790698e-06, "loss": 0.028, "step": 5427 }, { "epoch": 0.8766149870801033, "grad_norm": 145.76771545410156, "learning_rate": 1.2338501291989665e-06, "loss": 0.713, "step": 5428 }, { "epoch": 0.8767764857881137, "grad_norm": 211.31057739257812, "learning_rate": 1.2322351421188631e-06, "loss": 1.6211, "step": 5429 }, { "epoch": 0.876937984496124, "grad_norm": 13.92141342163086, "learning_rate": 1.2306201550387597e-06, "loss": 0.037, "step": 5430 }, { "epoch": 0.8770994832041343, "grad_norm": 422.5655212402344, "learning_rate": 1.2290051679586564e-06, "loss": 3.2993, "step": 5431 }, { "epoch": 0.8772609819121447, "grad_norm": 26.26375961303711, "learning_rate": 1.227390180878553e-06, "loss": 0.0977, "step": 5432 }, { "epoch": 0.877422480620155, "grad_norm": 98.71892547607422, "learning_rate": 1.2257751937984498e-06, "loss": 0.5369, "step": 5433 }, { "epoch": 0.8775839793281653, "grad_norm": 107.69895935058594, "learning_rate": 1.2241602067183463e-06, "loss": 0.5801, "step": 5434 }, { "epoch": 0.8777454780361758, "grad_norm": 10.519105911254883, "learning_rate": 1.2225452196382429e-06, "loss": 0.0389, "step": 5435 }, { "epoch": 0.877906976744186, "grad_norm": 100.11698913574219, "learning_rate": 1.2209302325581397e-06, "loss": 0.1892, "step": 5436 }, { "epoch": 0.8780684754521964, "grad_norm": 55.714927673339844, "learning_rate": 1.2193152454780362e-06, "loss": 0.1552, "step": 5437 }, { "epoch": 0.8782299741602068, "grad_norm": 18.089134216308594, "learning_rate": 1.217700258397933e-06, "loss": 0.048, "step": 5438 }, { "epoch": 0.8783914728682171, "grad_norm": 23.76468276977539, "learning_rate": 1.2160852713178296e-06, "loss": 0.1016, "step": 5439 }, { "epoch": 0.8785529715762274, "grad_norm": 9.099820137023926, "learning_rate": 1.2144702842377263e-06, "loss": 0.0188, "step": 5440 }, { "epoch": 0.8787144702842378, "grad_norm": 28.447053909301758, "learning_rate": 1.2128552971576229e-06, "loss": 0.0745, "step": 5441 }, { "epoch": 0.8788759689922481, "grad_norm": 19.0067195892334, "learning_rate": 1.2112403100775194e-06, "loss": 0.0595, "step": 5442 }, { "epoch": 0.8790374677002584, "grad_norm": 74.96520233154297, "learning_rate": 1.2096253229974162e-06, "loss": 0.3716, "step": 5443 }, { "epoch": 0.8791989664082688, "grad_norm": 163.74630737304688, "learning_rate": 1.2080103359173128e-06, "loss": 0.5281, "step": 5444 }, { "epoch": 0.8793604651162791, "grad_norm": 3.8423831462860107, "learning_rate": 1.2063953488372096e-06, "loss": 0.0103, "step": 5445 }, { "epoch": 0.8795219638242894, "grad_norm": 121.86266326904297, "learning_rate": 1.2047803617571061e-06, "loss": 0.2991, "step": 5446 }, { "epoch": 0.8796834625322998, "grad_norm": 7.242310047149658, "learning_rate": 1.2031653746770027e-06, "loss": 0.0255, "step": 5447 }, { "epoch": 0.8798449612403101, "grad_norm": 27.368602752685547, "learning_rate": 1.2015503875968994e-06, "loss": 0.0558, "step": 5448 }, { "epoch": 0.8800064599483204, "grad_norm": 7.540249347686768, "learning_rate": 1.199935400516796e-06, "loss": 0.0264, "step": 5449 }, { "epoch": 0.8801679586563308, "grad_norm": 11.058555603027344, "learning_rate": 1.1983204134366926e-06, "loss": 0.0402, "step": 5450 }, { "epoch": 0.8803294573643411, "grad_norm": 192.75848388671875, "learning_rate": 1.1967054263565893e-06, "loss": 1.3613, "step": 5451 }, { "epoch": 0.8804909560723514, "grad_norm": 68.71719360351562, "learning_rate": 1.195090439276486e-06, "loss": 0.6825, "step": 5452 }, { "epoch": 0.8806524547803618, "grad_norm": 18.359127044677734, "learning_rate": 1.1934754521963825e-06, "loss": 0.0535, "step": 5453 }, { "epoch": 0.8808139534883721, "grad_norm": 104.12110137939453, "learning_rate": 1.1918604651162792e-06, "loss": 1.127, "step": 5454 }, { "epoch": 0.8809754521963824, "grad_norm": 72.20308685302734, "learning_rate": 1.1902454780361758e-06, "loss": 0.2884, "step": 5455 }, { "epoch": 0.8811369509043928, "grad_norm": 34.59365463256836, "learning_rate": 1.1886304909560723e-06, "loss": 0.1436, "step": 5456 }, { "epoch": 0.8812984496124031, "grad_norm": 40.599891662597656, "learning_rate": 1.1870155038759691e-06, "loss": 0.2305, "step": 5457 }, { "epoch": 0.8814599483204134, "grad_norm": 10.259800910949707, "learning_rate": 1.1854005167958657e-06, "loss": 0.0488, "step": 5458 }, { "epoch": 0.8816214470284238, "grad_norm": 7.9085187911987305, "learning_rate": 1.1837855297157624e-06, "loss": 0.0167, "step": 5459 }, { "epoch": 0.8817829457364341, "grad_norm": 68.67396545410156, "learning_rate": 1.182170542635659e-06, "loss": 0.3828, "step": 5460 }, { "epoch": 0.8819444444444444, "grad_norm": 306.1863098144531, "learning_rate": 1.1805555555555556e-06, "loss": 0.2578, "step": 5461 }, { "epoch": 0.8821059431524548, "grad_norm": 86.7605972290039, "learning_rate": 1.1789405684754523e-06, "loss": 0.6855, "step": 5462 }, { "epoch": 0.8822674418604651, "grad_norm": 37.06996154785156, "learning_rate": 1.177325581395349e-06, "loss": 0.1572, "step": 5463 }, { "epoch": 0.8824289405684754, "grad_norm": 29.824241638183594, "learning_rate": 1.1757105943152455e-06, "loss": 0.0801, "step": 5464 }, { "epoch": 0.8825904392764858, "grad_norm": 11.443422317504883, "learning_rate": 1.1740956072351422e-06, "loss": 0.0311, "step": 5465 }, { "epoch": 0.8827519379844961, "grad_norm": 115.12870025634766, "learning_rate": 1.1724806201550388e-06, "loss": 0.2522, "step": 5466 }, { "epoch": 0.8829134366925064, "grad_norm": 6.452846527099609, "learning_rate": 1.1708656330749354e-06, "loss": 0.0129, "step": 5467 }, { "epoch": 0.8830749354005168, "grad_norm": 166.258544921875, "learning_rate": 1.1692506459948321e-06, "loss": 2.2449, "step": 5468 }, { "epoch": 0.8832364341085271, "grad_norm": 178.9996795654297, "learning_rate": 1.1676356589147287e-06, "loss": 0.3565, "step": 5469 }, { "epoch": 0.8833979328165374, "grad_norm": 171.8509063720703, "learning_rate": 1.1660206718346255e-06, "loss": 0.7783, "step": 5470 }, { "epoch": 0.8835594315245479, "grad_norm": 51.24929428100586, "learning_rate": 1.164405684754522e-06, "loss": 0.208, "step": 5471 }, { "epoch": 0.8837209302325582, "grad_norm": 48.12455368041992, "learning_rate": 1.1627906976744188e-06, "loss": 0.3311, "step": 5472 }, { "epoch": 0.8838824289405685, "grad_norm": 3.9195311069488525, "learning_rate": 1.1611757105943153e-06, "loss": 0.0127, "step": 5473 }, { "epoch": 0.8840439276485789, "grad_norm": 179.1058349609375, "learning_rate": 1.159560723514212e-06, "loss": 1.6133, "step": 5474 }, { "epoch": 0.8842054263565892, "grad_norm": 134.2257843017578, "learning_rate": 1.1579457364341087e-06, "loss": 1.2188, "step": 5475 }, { "epoch": 0.8843669250645995, "grad_norm": 628.86865234375, "learning_rate": 1.1563307493540052e-06, "loss": 1.4, "step": 5476 }, { "epoch": 0.8845284237726099, "grad_norm": 25.94464111328125, "learning_rate": 1.154715762273902e-06, "loss": 0.112, "step": 5477 }, { "epoch": 0.8846899224806202, "grad_norm": 164.12509155273438, "learning_rate": 1.1531007751937986e-06, "loss": 0.6299, "step": 5478 }, { "epoch": 0.8848514211886305, "grad_norm": 62.36351776123047, "learning_rate": 1.1514857881136953e-06, "loss": 0.1904, "step": 5479 }, { "epoch": 0.8850129198966409, "grad_norm": 156.33407592773438, "learning_rate": 1.149870801033592e-06, "loss": 0.2935, "step": 5480 }, { "epoch": 0.8851744186046512, "grad_norm": 507.5437316894531, "learning_rate": 1.1482558139534885e-06, "loss": 1.0394, "step": 5481 }, { "epoch": 0.8853359173126615, "grad_norm": 20.664997100830078, "learning_rate": 1.1466408268733852e-06, "loss": 0.0457, "step": 5482 }, { "epoch": 0.8854974160206718, "grad_norm": 56.87749099731445, "learning_rate": 1.1450258397932818e-06, "loss": 0.2925, "step": 5483 }, { "epoch": 0.8856589147286822, "grad_norm": 177.930908203125, "learning_rate": 1.1434108527131784e-06, "loss": 0.9845, "step": 5484 }, { "epoch": 0.8858204134366925, "grad_norm": 128.80918884277344, "learning_rate": 1.1417958656330751e-06, "loss": 0.3694, "step": 5485 }, { "epoch": 0.8859819121447028, "grad_norm": 34.0142822265625, "learning_rate": 1.1401808785529717e-06, "loss": 0.1387, "step": 5486 }, { "epoch": 0.8861434108527132, "grad_norm": 47.855201721191406, "learning_rate": 1.1385658914728682e-06, "loss": 0.6182, "step": 5487 }, { "epoch": 0.8863049095607235, "grad_norm": 244.4276123046875, "learning_rate": 1.136950904392765e-06, "loss": 4.0156, "step": 5488 }, { "epoch": 0.8864664082687338, "grad_norm": 74.3651123046875, "learning_rate": 1.1353359173126616e-06, "loss": 0.5391, "step": 5489 }, { "epoch": 0.8866279069767442, "grad_norm": 99.54730224609375, "learning_rate": 1.1337209302325581e-06, "loss": 0.7998, "step": 5490 }, { "epoch": 0.8867894056847545, "grad_norm": 56.058650970458984, "learning_rate": 1.132105943152455e-06, "loss": 0.2847, "step": 5491 }, { "epoch": 0.8869509043927648, "grad_norm": 78.41580200195312, "learning_rate": 1.1304909560723515e-06, "loss": 0.3823, "step": 5492 }, { "epoch": 0.8871124031007752, "grad_norm": 61.3240852355957, "learning_rate": 1.1288759689922482e-06, "loss": 0.2493, "step": 5493 }, { "epoch": 0.8872739018087855, "grad_norm": 70.86451721191406, "learning_rate": 1.1272609819121448e-06, "loss": 0.2112, "step": 5494 }, { "epoch": 0.8874354005167958, "grad_norm": 82.25988006591797, "learning_rate": 1.1256459948320414e-06, "loss": 0.5532, "step": 5495 }, { "epoch": 0.8875968992248062, "grad_norm": 194.7060089111328, "learning_rate": 1.1240310077519381e-06, "loss": 0.3999, "step": 5496 }, { "epoch": 0.8877583979328165, "grad_norm": 290.92620849609375, "learning_rate": 1.1224160206718347e-06, "loss": 2.3494, "step": 5497 }, { "epoch": 0.8879198966408268, "grad_norm": 12.323009490966797, "learning_rate": 1.1208010335917313e-06, "loss": 0.0654, "step": 5498 }, { "epoch": 0.8880813953488372, "grad_norm": 14.41379451751709, "learning_rate": 1.119186046511628e-06, "loss": 0.0488, "step": 5499 }, { "epoch": 0.8882428940568475, "grad_norm": 61.789947509765625, "learning_rate": 1.1175710594315246e-06, "loss": 0.459, "step": 5500 }, { "epoch": 0.8884043927648578, "grad_norm": 22.855745315551758, "learning_rate": 1.1159560723514211e-06, "loss": 0.074, "step": 5501 }, { "epoch": 0.8885658914728682, "grad_norm": 31.713687896728516, "learning_rate": 1.114341085271318e-06, "loss": 0.1504, "step": 5502 }, { "epoch": 0.8887273901808785, "grad_norm": 54.31528091430664, "learning_rate": 1.1127260981912145e-06, "loss": 0.4023, "step": 5503 }, { "epoch": 0.8888888888888888, "grad_norm": 38.66396713256836, "learning_rate": 1.111111111111111e-06, "loss": 0.1982, "step": 5504 }, { "epoch": 0.8890503875968992, "grad_norm": 63.51042938232422, "learning_rate": 1.1094961240310078e-06, "loss": 0.2698, "step": 5505 }, { "epoch": 0.8892118863049095, "grad_norm": 44.90509033203125, "learning_rate": 1.1078811369509044e-06, "loss": 0.2983, "step": 5506 }, { "epoch": 0.8893733850129198, "grad_norm": 114.71578979492188, "learning_rate": 1.1062661498708011e-06, "loss": 0.6829, "step": 5507 }, { "epoch": 0.8895348837209303, "grad_norm": 11.764612197875977, "learning_rate": 1.1046511627906977e-06, "loss": 0.0276, "step": 5508 }, { "epoch": 0.8896963824289406, "grad_norm": 119.11175537109375, "learning_rate": 1.1030361757105945e-06, "loss": 0.3557, "step": 5509 }, { "epoch": 0.8898578811369509, "grad_norm": 26.882360458374023, "learning_rate": 1.101421188630491e-06, "loss": 0.099, "step": 5510 }, { "epoch": 0.8900193798449613, "grad_norm": 43.05479049682617, "learning_rate": 1.0998062015503876e-06, "loss": 0.188, "step": 5511 }, { "epoch": 0.8901808785529716, "grad_norm": 8.610702514648438, "learning_rate": 1.0981912144702844e-06, "loss": 0.0248, "step": 5512 }, { "epoch": 0.8903423772609819, "grad_norm": 36.03791809082031, "learning_rate": 1.096576227390181e-06, "loss": 0.1606, "step": 5513 }, { "epoch": 0.8905038759689923, "grad_norm": 21.61526870727539, "learning_rate": 1.0949612403100777e-06, "loss": 0.0616, "step": 5514 }, { "epoch": 0.8906653746770026, "grad_norm": 253.64634704589844, "learning_rate": 1.0933462532299743e-06, "loss": 0.7573, "step": 5515 }, { "epoch": 0.8908268733850129, "grad_norm": 43.448814392089844, "learning_rate": 1.091731266149871e-06, "loss": 0.0477, "step": 5516 }, { "epoch": 0.8909883720930233, "grad_norm": 15.534643173217773, "learning_rate": 1.0901162790697676e-06, "loss": 0.0334, "step": 5517 }, { "epoch": 0.8911498708010336, "grad_norm": 80.02925109863281, "learning_rate": 1.0885012919896641e-06, "loss": 0.2822, "step": 5518 }, { "epoch": 0.8913113695090439, "grad_norm": 49.753204345703125, "learning_rate": 1.086886304909561e-06, "loss": 0.1176, "step": 5519 }, { "epoch": 0.8914728682170543, "grad_norm": 59.663719177246094, "learning_rate": 1.0852713178294575e-06, "loss": 0.2627, "step": 5520 }, { "epoch": 0.8916343669250646, "grad_norm": 180.5288543701172, "learning_rate": 1.083656330749354e-06, "loss": 0.9621, "step": 5521 }, { "epoch": 0.8917958656330749, "grad_norm": 33.34419250488281, "learning_rate": 1.0820413436692508e-06, "loss": 0.1646, "step": 5522 }, { "epoch": 0.8919573643410853, "grad_norm": 227.38723754882812, "learning_rate": 1.0804263565891474e-06, "loss": 2.9688, "step": 5523 }, { "epoch": 0.8921188630490956, "grad_norm": 114.14816284179688, "learning_rate": 1.0788113695090441e-06, "loss": 0.7007, "step": 5524 }, { "epoch": 0.8922803617571059, "grad_norm": 25.234098434448242, "learning_rate": 1.0771963824289407e-06, "loss": 0.1242, "step": 5525 }, { "epoch": 0.8924418604651163, "grad_norm": 113.11649322509766, "learning_rate": 1.0755813953488373e-06, "loss": 0.2872, "step": 5526 }, { "epoch": 0.8926033591731266, "grad_norm": 5.397861003875732, "learning_rate": 1.073966408268734e-06, "loss": 0.018, "step": 5527 }, { "epoch": 0.8927648578811369, "grad_norm": 143.72525024414062, "learning_rate": 1.0723514211886306e-06, "loss": 1.022, "step": 5528 }, { "epoch": 0.8929263565891473, "grad_norm": 23.446535110473633, "learning_rate": 1.0707364341085272e-06, "loss": 0.0787, "step": 5529 }, { "epoch": 0.8930878552971576, "grad_norm": 32.44103240966797, "learning_rate": 1.069121447028424e-06, "loss": 0.1282, "step": 5530 }, { "epoch": 0.8932493540051679, "grad_norm": 69.28693389892578, "learning_rate": 1.0675064599483205e-06, "loss": 0.2237, "step": 5531 }, { "epoch": 0.8934108527131783, "grad_norm": 48.93045425415039, "learning_rate": 1.065891472868217e-06, "loss": 0.1929, "step": 5532 }, { "epoch": 0.8935723514211886, "grad_norm": 180.97528076171875, "learning_rate": 1.0642764857881138e-06, "loss": 2.6875, "step": 5533 }, { "epoch": 0.8937338501291989, "grad_norm": 4.166284084320068, "learning_rate": 1.0626614987080104e-06, "loss": 0.0053, "step": 5534 }, { "epoch": 0.8938953488372093, "grad_norm": 38.532997131347656, "learning_rate": 1.061046511627907e-06, "loss": 0.0926, "step": 5535 }, { "epoch": 0.8940568475452196, "grad_norm": 141.07298278808594, "learning_rate": 1.0594315245478037e-06, "loss": 1.8354, "step": 5536 }, { "epoch": 0.8942183462532299, "grad_norm": 5.547150611877441, "learning_rate": 1.0578165374677003e-06, "loss": 0.012, "step": 5537 }, { "epoch": 0.8943798449612403, "grad_norm": 319.310302734375, "learning_rate": 1.0562015503875968e-06, "loss": 1.0527, "step": 5538 }, { "epoch": 0.8945413436692506, "grad_norm": 10.777496337890625, "learning_rate": 1.0545865633074936e-06, "loss": 0.0256, "step": 5539 }, { "epoch": 0.8947028423772609, "grad_norm": 297.1497802734375, "learning_rate": 1.0529715762273902e-06, "loss": 1.3035, "step": 5540 }, { "epoch": 0.8948643410852714, "grad_norm": 102.10917663574219, "learning_rate": 1.051356589147287e-06, "loss": 0.7605, "step": 5541 }, { "epoch": 0.8950258397932817, "grad_norm": 120.01045227050781, "learning_rate": 1.0497416020671835e-06, "loss": 0.8716, "step": 5542 }, { "epoch": 0.895187338501292, "grad_norm": 18.11654281616211, "learning_rate": 1.04812661498708e-06, "loss": 0.0402, "step": 5543 }, { "epoch": 0.8953488372093024, "grad_norm": 36.68653869628906, "learning_rate": 1.0465116279069768e-06, "loss": 0.0906, "step": 5544 }, { "epoch": 0.8955103359173127, "grad_norm": 19.912717819213867, "learning_rate": 1.0448966408268734e-06, "loss": 0.0589, "step": 5545 }, { "epoch": 0.895671834625323, "grad_norm": 15.041481018066406, "learning_rate": 1.0432816537467702e-06, "loss": 0.0486, "step": 5546 }, { "epoch": 0.8958333333333334, "grad_norm": 68.54440307617188, "learning_rate": 1.0416666666666667e-06, "loss": 0.2881, "step": 5547 }, { "epoch": 0.8959948320413437, "grad_norm": 16.30152702331543, "learning_rate": 1.0400516795865635e-06, "loss": 0.0455, "step": 5548 }, { "epoch": 0.896156330749354, "grad_norm": 137.7530975341797, "learning_rate": 1.03843669250646e-06, "loss": 0.3707, "step": 5549 }, { "epoch": 0.8963178294573644, "grad_norm": 67.6007080078125, "learning_rate": 1.0368217054263566e-06, "loss": 0.1946, "step": 5550 }, { "epoch": 0.8964793281653747, "grad_norm": 48.94307327270508, "learning_rate": 1.0352067183462534e-06, "loss": 0.1064, "step": 5551 }, { "epoch": 0.896640826873385, "grad_norm": 152.73193359375, "learning_rate": 1.03359173126615e-06, "loss": 1.0796, "step": 5552 }, { "epoch": 0.8968023255813954, "grad_norm": 54.918739318847656, "learning_rate": 1.0319767441860467e-06, "loss": 0.028, "step": 5553 }, { "epoch": 0.8969638242894057, "grad_norm": 31.815876007080078, "learning_rate": 1.0303617571059433e-06, "loss": 0.0739, "step": 5554 }, { "epoch": 0.897125322997416, "grad_norm": 61.72735595703125, "learning_rate": 1.0287467700258398e-06, "loss": 0.3789, "step": 5555 }, { "epoch": 0.8972868217054264, "grad_norm": 14.773497581481934, "learning_rate": 1.0271317829457366e-06, "loss": 0.0246, "step": 5556 }, { "epoch": 0.8974483204134367, "grad_norm": 55.07961654663086, "learning_rate": 1.0255167958656332e-06, "loss": 0.2803, "step": 5557 }, { "epoch": 0.897609819121447, "grad_norm": 374.8088684082031, "learning_rate": 1.02390180878553e-06, "loss": 1.959, "step": 5558 }, { "epoch": 0.8977713178294574, "grad_norm": 35.6600227355957, "learning_rate": 1.0222868217054265e-06, "loss": 0.1182, "step": 5559 }, { "epoch": 0.8979328165374677, "grad_norm": 7.413965225219727, "learning_rate": 1.020671834625323e-06, "loss": 0.0193, "step": 5560 }, { "epoch": 0.898094315245478, "grad_norm": 74.99874877929688, "learning_rate": 1.0190568475452198e-06, "loss": 0.2529, "step": 5561 }, { "epoch": 0.8982558139534884, "grad_norm": 111.4134750366211, "learning_rate": 1.0174418604651164e-06, "loss": 0.4333, "step": 5562 }, { "epoch": 0.8984173126614987, "grad_norm": 124.23050689697266, "learning_rate": 1.015826873385013e-06, "loss": 0.531, "step": 5563 }, { "epoch": 0.898578811369509, "grad_norm": 47.961002349853516, "learning_rate": 1.0142118863049097e-06, "loss": 0.1443, "step": 5564 }, { "epoch": 0.8987403100775194, "grad_norm": 142.36978149414062, "learning_rate": 1.0125968992248063e-06, "loss": 0.6789, "step": 5565 }, { "epoch": 0.8989018087855297, "grad_norm": 155.42434692382812, "learning_rate": 1.0109819121447028e-06, "loss": 1.4736, "step": 5566 }, { "epoch": 0.89906330749354, "grad_norm": 8.306867599487305, "learning_rate": 1.0093669250645996e-06, "loss": 0.0164, "step": 5567 }, { "epoch": 0.8992248062015504, "grad_norm": 87.76748657226562, "learning_rate": 1.0077519379844962e-06, "loss": 0.5117, "step": 5568 }, { "epoch": 0.8993863049095607, "grad_norm": 67.52802276611328, "learning_rate": 1.0061369509043927e-06, "loss": 0.2466, "step": 5569 }, { "epoch": 0.899547803617571, "grad_norm": 5.667248725891113, "learning_rate": 1.0045219638242895e-06, "loss": 0.0156, "step": 5570 }, { "epoch": 0.8997093023255814, "grad_norm": 22.5109806060791, "learning_rate": 1.002906976744186e-06, "loss": 0.0886, "step": 5571 }, { "epoch": 0.8998708010335917, "grad_norm": 37.11650466918945, "learning_rate": 1.0012919896640828e-06, "loss": 0.0844, "step": 5572 }, { "epoch": 0.900032299741602, "grad_norm": 112.23880767822266, "learning_rate": 9.996770025839794e-07, "loss": 0.3325, "step": 5573 }, { "epoch": 0.9001937984496124, "grad_norm": 17.70612907409668, "learning_rate": 9.98062015503876e-07, "loss": 0.0411, "step": 5574 }, { "epoch": 0.9003552971576227, "grad_norm": 16.321887969970703, "learning_rate": 9.964470284237727e-07, "loss": 0.167, "step": 5575 }, { "epoch": 0.900516795865633, "grad_norm": 34.63011169433594, "learning_rate": 9.948320413436693e-07, "loss": 0.0896, "step": 5576 }, { "epoch": 0.9006782945736435, "grad_norm": 151.36448669433594, "learning_rate": 9.932170542635658e-07, "loss": 1.0, "step": 5577 }, { "epoch": 0.9008397932816538, "grad_norm": 169.1396026611328, "learning_rate": 9.916020671834626e-07, "loss": 0.8774, "step": 5578 }, { "epoch": 0.901001291989664, "grad_norm": 4.782687664031982, "learning_rate": 9.899870801033592e-07, "loss": 0.0132, "step": 5579 }, { "epoch": 0.9011627906976745, "grad_norm": 6.394525527954102, "learning_rate": 9.88372093023256e-07, "loss": 0.0171, "step": 5580 }, { "epoch": 0.9013242894056848, "grad_norm": 64.12237548828125, "learning_rate": 9.867571059431525e-07, "loss": 0.3726, "step": 5581 }, { "epoch": 0.9014857881136951, "grad_norm": 14.500175476074219, "learning_rate": 9.85142118863049e-07, "loss": 0.0522, "step": 5582 }, { "epoch": 0.9016472868217055, "grad_norm": 91.33767700195312, "learning_rate": 9.835271317829458e-07, "loss": 0.4488, "step": 5583 }, { "epoch": 0.9018087855297158, "grad_norm": 21.675395965576172, "learning_rate": 9.819121447028424e-07, "loss": 0.0886, "step": 5584 }, { "epoch": 0.9019702842377261, "grad_norm": 26.655527114868164, "learning_rate": 9.802971576227392e-07, "loss": 0.084, "step": 5585 }, { "epoch": 0.9021317829457365, "grad_norm": 21.315690994262695, "learning_rate": 9.786821705426357e-07, "loss": 0.0767, "step": 5586 }, { "epoch": 0.9022932816537468, "grad_norm": 74.84884643554688, "learning_rate": 9.770671834625325e-07, "loss": 0.207, "step": 5587 }, { "epoch": 0.9024547803617571, "grad_norm": 74.8124771118164, "learning_rate": 9.75452196382429e-07, "loss": 0.1953, "step": 5588 }, { "epoch": 0.9026162790697675, "grad_norm": 56.89448547363281, "learning_rate": 9.738372093023256e-07, "loss": 0.2773, "step": 5589 }, { "epoch": 0.9027777777777778, "grad_norm": 227.14878845214844, "learning_rate": 9.722222222222224e-07, "loss": 0.7131, "step": 5590 }, { "epoch": 0.9029392764857881, "grad_norm": 323.2142028808594, "learning_rate": 9.70607235142119e-07, "loss": 2.0984, "step": 5591 }, { "epoch": 0.9031007751937985, "grad_norm": 60.331607818603516, "learning_rate": 9.689922480620157e-07, "loss": 0.3501, "step": 5592 }, { "epoch": 0.9032622739018088, "grad_norm": 11.730805397033691, "learning_rate": 9.673772609819123e-07, "loss": 0.0246, "step": 5593 }, { "epoch": 0.9034237726098191, "grad_norm": 28.223453521728516, "learning_rate": 9.657622739018088e-07, "loss": 0.1252, "step": 5594 }, { "epoch": 0.9035852713178295, "grad_norm": 442.29644775390625, "learning_rate": 9.641472868217056e-07, "loss": 2.5742, "step": 5595 }, { "epoch": 0.9037467700258398, "grad_norm": 99.2081069946289, "learning_rate": 9.625322997416022e-07, "loss": 0.2693, "step": 5596 }, { "epoch": 0.9039082687338501, "grad_norm": 141.11322021484375, "learning_rate": 9.609173126614987e-07, "loss": 0.8242, "step": 5597 }, { "epoch": 0.9040697674418605, "grad_norm": 8.984481811523438, "learning_rate": 9.593023255813955e-07, "loss": 0.0352, "step": 5598 }, { "epoch": 0.9042312661498708, "grad_norm": 14.177014350891113, "learning_rate": 9.57687338501292e-07, "loss": 0.0435, "step": 5599 }, { "epoch": 0.9043927648578811, "grad_norm": 177.7064666748047, "learning_rate": 9.560723514211886e-07, "loss": 0.3115, "step": 5600 }, { "epoch": 0.9045542635658915, "grad_norm": 174.68392944335938, "learning_rate": 9.544573643410854e-07, "loss": 1.8929, "step": 5601 }, { "epoch": 0.9047157622739018, "grad_norm": 61.61103057861328, "learning_rate": 9.52842377260982e-07, "loss": 0.233, "step": 5602 }, { "epoch": 0.9048772609819121, "grad_norm": 258.70257568359375, "learning_rate": 9.512273901808785e-07, "loss": 0.873, "step": 5603 }, { "epoch": 0.9050387596899225, "grad_norm": 104.03192138671875, "learning_rate": 9.496124031007753e-07, "loss": 1.317, "step": 5604 }, { "epoch": 0.9052002583979328, "grad_norm": 9.759635925292969, "learning_rate": 9.479974160206719e-07, "loss": 0.0241, "step": 5605 }, { "epoch": 0.9053617571059431, "grad_norm": 36.75540542602539, "learning_rate": 9.463824289405686e-07, "loss": 0.1589, "step": 5606 }, { "epoch": 0.9055232558139535, "grad_norm": 20.945756912231445, "learning_rate": 9.447674418604652e-07, "loss": 0.0703, "step": 5607 }, { "epoch": 0.9056847545219638, "grad_norm": 88.43724822998047, "learning_rate": 9.431524547803617e-07, "loss": 0.27, "step": 5608 }, { "epoch": 0.9058462532299741, "grad_norm": 12.937784194946289, "learning_rate": 9.415374677002585e-07, "loss": 0.0247, "step": 5609 }, { "epoch": 0.9060077519379846, "grad_norm": 9.660589218139648, "learning_rate": 9.399224806201551e-07, "loss": 0.0256, "step": 5610 }, { "epoch": 0.9061692506459949, "grad_norm": 24.68714714050293, "learning_rate": 9.383074935400517e-07, "loss": 0.0493, "step": 5611 }, { "epoch": 0.9063307493540051, "grad_norm": 9.350950241088867, "learning_rate": 9.366925064599484e-07, "loss": 0.0448, "step": 5612 }, { "epoch": 0.9064922480620154, "grad_norm": 59.028499603271484, "learning_rate": 9.350775193798451e-07, "loss": 0.1663, "step": 5613 }, { "epoch": 0.9066537467700259, "grad_norm": 21.5670108795166, "learning_rate": 9.334625322997416e-07, "loss": 0.0547, "step": 5614 }, { "epoch": 0.9068152454780362, "grad_norm": 183.21885681152344, "learning_rate": 9.318475452196384e-07, "loss": 1.6279, "step": 5615 }, { "epoch": 0.9069767441860465, "grad_norm": 159.51828002929688, "learning_rate": 9.30232558139535e-07, "loss": 0.7094, "step": 5616 }, { "epoch": 0.9071382428940569, "grad_norm": 6.159058570861816, "learning_rate": 9.286175710594315e-07, "loss": 0.0225, "step": 5617 }, { "epoch": 0.9072997416020672, "grad_norm": 66.51811218261719, "learning_rate": 9.270025839793283e-07, "loss": 0.4385, "step": 5618 }, { "epoch": 0.9074612403100775, "grad_norm": 48.182437896728516, "learning_rate": 9.253875968992249e-07, "loss": 0.0713, "step": 5619 }, { "epoch": 0.9076227390180879, "grad_norm": 17.04886817932129, "learning_rate": 9.237726098191216e-07, "loss": 0.0571, "step": 5620 }, { "epoch": 0.9077842377260982, "grad_norm": 27.615848541259766, "learning_rate": 9.221576227390182e-07, "loss": 0.1057, "step": 5621 }, { "epoch": 0.9079457364341085, "grad_norm": 45.716644287109375, "learning_rate": 9.205426356589148e-07, "loss": 0.1538, "step": 5622 }, { "epoch": 0.9081072351421189, "grad_norm": 82.45057678222656, "learning_rate": 9.189276485788115e-07, "loss": 0.5, "step": 5623 }, { "epoch": 0.9082687338501292, "grad_norm": 14.440678596496582, "learning_rate": 9.173126614987081e-07, "loss": 0.056, "step": 5624 }, { "epoch": 0.9084302325581395, "grad_norm": 35.301048278808594, "learning_rate": 9.156976744186046e-07, "loss": 0.1208, "step": 5625 }, { "epoch": 0.9085917312661499, "grad_norm": 27.547697067260742, "learning_rate": 9.140826873385014e-07, "loss": 0.0997, "step": 5626 }, { "epoch": 0.9087532299741602, "grad_norm": 14.65160083770752, "learning_rate": 9.12467700258398e-07, "loss": 0.0259, "step": 5627 }, { "epoch": 0.9089147286821705, "grad_norm": 196.7369384765625, "learning_rate": 9.108527131782946e-07, "loss": 0.6113, "step": 5628 }, { "epoch": 0.9090762273901809, "grad_norm": 20.711149215698242, "learning_rate": 9.092377260981913e-07, "loss": 0.0858, "step": 5629 }, { "epoch": 0.9092377260981912, "grad_norm": 13.181986808776855, "learning_rate": 9.07622739018088e-07, "loss": 0.0476, "step": 5630 }, { "epoch": 0.9093992248062015, "grad_norm": 10.864059448242188, "learning_rate": 9.060077519379845e-07, "loss": 0.0134, "step": 5631 }, { "epoch": 0.9095607235142119, "grad_norm": 35.697914123535156, "learning_rate": 9.043927648578812e-07, "loss": 0.2228, "step": 5632 }, { "epoch": 0.9097222222222222, "grad_norm": 10.847173690795898, "learning_rate": 9.027777777777779e-07, "loss": 0.0366, "step": 5633 }, { "epoch": 0.9098837209302325, "grad_norm": 70.46647644042969, "learning_rate": 9.011627906976744e-07, "loss": 0.6888, "step": 5634 }, { "epoch": 0.9100452196382429, "grad_norm": 7.5652008056640625, "learning_rate": 8.995478036175712e-07, "loss": 0.0074, "step": 5635 }, { "epoch": 0.9102067183462532, "grad_norm": 48.147674560546875, "learning_rate": 8.979328165374678e-07, "loss": 0.2782, "step": 5636 }, { "epoch": 0.9103682170542635, "grad_norm": 66.90872955322266, "learning_rate": 8.963178294573645e-07, "loss": 0.3384, "step": 5637 }, { "epoch": 0.9105297157622739, "grad_norm": 140.52212524414062, "learning_rate": 8.947028423772611e-07, "loss": 0.5601, "step": 5638 }, { "epoch": 0.9106912144702842, "grad_norm": 22.94031524658203, "learning_rate": 8.930878552971576e-07, "loss": 0.083, "step": 5639 }, { "epoch": 0.9108527131782945, "grad_norm": 65.79000854492188, "learning_rate": 8.914728682170544e-07, "loss": 0.25, "step": 5640 }, { "epoch": 0.9110142118863049, "grad_norm": 13.968158721923828, "learning_rate": 8.89857881136951e-07, "loss": 0.0404, "step": 5641 }, { "epoch": 0.9111757105943152, "grad_norm": 14.273292541503906, "learning_rate": 8.882428940568475e-07, "loss": 0.0369, "step": 5642 }, { "epoch": 0.9113372093023255, "grad_norm": 28.64579200744629, "learning_rate": 8.866279069767443e-07, "loss": 0.0385, "step": 5643 }, { "epoch": 0.911498708010336, "grad_norm": 285.78656005859375, "learning_rate": 8.850129198966409e-07, "loss": 1.0225, "step": 5644 }, { "epoch": 0.9116602067183462, "grad_norm": 53.69416427612305, "learning_rate": 8.833979328165375e-07, "loss": 0.1451, "step": 5645 }, { "epoch": 0.9118217054263565, "grad_norm": 126.4889144897461, "learning_rate": 8.817829457364342e-07, "loss": 2.1668, "step": 5646 }, { "epoch": 0.911983204134367, "grad_norm": 20.10920524597168, "learning_rate": 8.801679586563308e-07, "loss": 0.044, "step": 5647 }, { "epoch": 0.9121447028423773, "grad_norm": 11.285922050476074, "learning_rate": 8.785529715762274e-07, "loss": 0.051, "step": 5648 }, { "epoch": 0.9123062015503876, "grad_norm": 50.47505187988281, "learning_rate": 8.769379844961241e-07, "loss": 0.4309, "step": 5649 }, { "epoch": 0.912467700258398, "grad_norm": 298.9360046386719, "learning_rate": 8.753229974160208e-07, "loss": 1.1453, "step": 5650 }, { "epoch": 0.9126291989664083, "grad_norm": 25.28203582763672, "learning_rate": 8.737080103359173e-07, "loss": 0.0486, "step": 5651 }, { "epoch": 0.9127906976744186, "grad_norm": 73.84022521972656, "learning_rate": 8.720930232558141e-07, "loss": 0.2537, "step": 5652 }, { "epoch": 0.912952196382429, "grad_norm": 62.35209274291992, "learning_rate": 8.704780361757107e-07, "loss": 0.4536, "step": 5653 }, { "epoch": 0.9131136950904393, "grad_norm": 20.067859649658203, "learning_rate": 8.688630490956073e-07, "loss": 0.1128, "step": 5654 }, { "epoch": 0.9132751937984496, "grad_norm": 4.892362117767334, "learning_rate": 8.67248062015504e-07, "loss": 0.0121, "step": 5655 }, { "epoch": 0.91343669250646, "grad_norm": 15.729026794433594, "learning_rate": 8.656330749354005e-07, "loss": 0.0438, "step": 5656 }, { "epoch": 0.9135981912144703, "grad_norm": 13.781485557556152, "learning_rate": 8.640180878552973e-07, "loss": 0.0446, "step": 5657 }, { "epoch": 0.9137596899224806, "grad_norm": 28.305063247680664, "learning_rate": 8.624031007751939e-07, "loss": 0.0925, "step": 5658 }, { "epoch": 0.913921188630491, "grad_norm": 157.0335693359375, "learning_rate": 8.607881136950904e-07, "loss": 1.2305, "step": 5659 }, { "epoch": 0.9140826873385013, "grad_norm": 30.38142204284668, "learning_rate": 8.591731266149872e-07, "loss": 0.1407, "step": 5660 }, { "epoch": 0.9142441860465116, "grad_norm": 34.12215042114258, "learning_rate": 8.575581395348838e-07, "loss": 0.0876, "step": 5661 }, { "epoch": 0.914405684754522, "grad_norm": 133.4530029296875, "learning_rate": 8.559431524547804e-07, "loss": 0.9888, "step": 5662 }, { "epoch": 0.9145671834625323, "grad_norm": 21.668380737304688, "learning_rate": 8.543281653746771e-07, "loss": 0.0332, "step": 5663 }, { "epoch": 0.9147286821705426, "grad_norm": 15.149330139160156, "learning_rate": 8.527131782945737e-07, "loss": 0.0287, "step": 5664 }, { "epoch": 0.914890180878553, "grad_norm": 126.99005889892578, "learning_rate": 8.510981912144703e-07, "loss": 1.3789, "step": 5665 }, { "epoch": 0.9150516795865633, "grad_norm": 54.978336334228516, "learning_rate": 8.49483204134367e-07, "loss": 0.1699, "step": 5666 }, { "epoch": 0.9152131782945736, "grad_norm": 68.11266326904297, "learning_rate": 8.478682170542637e-07, "loss": 0.1983, "step": 5667 }, { "epoch": 0.915374677002584, "grad_norm": 53.80091094970703, "learning_rate": 8.462532299741603e-07, "loss": 0.291, "step": 5668 }, { "epoch": 0.9155361757105943, "grad_norm": 31.974937438964844, "learning_rate": 8.44638242894057e-07, "loss": 0.1152, "step": 5669 }, { "epoch": 0.9156976744186046, "grad_norm": 7.817755222320557, "learning_rate": 8.430232558139535e-07, "loss": 0.045, "step": 5670 }, { "epoch": 0.915859173126615, "grad_norm": 86.04725646972656, "learning_rate": 8.414082687338502e-07, "loss": 0.2224, "step": 5671 }, { "epoch": 0.9160206718346253, "grad_norm": 10.725769996643066, "learning_rate": 8.397932816537469e-07, "loss": 0.0572, "step": 5672 }, { "epoch": 0.9161821705426356, "grad_norm": 21.117828369140625, "learning_rate": 8.381782945736434e-07, "loss": 0.113, "step": 5673 }, { "epoch": 0.916343669250646, "grad_norm": 72.22320556640625, "learning_rate": 8.365633074935402e-07, "loss": 0.3057, "step": 5674 }, { "epoch": 0.9165051679586563, "grad_norm": 11.200927734375, "learning_rate": 8.349483204134368e-07, "loss": 0.0221, "step": 5675 }, { "epoch": 0.9166666666666666, "grad_norm": 11.350613594055176, "learning_rate": 8.333333333333333e-07, "loss": 0.035, "step": 5676 }, { "epoch": 0.916828165374677, "grad_norm": 94.54927825927734, "learning_rate": 8.317183462532301e-07, "loss": 0.8105, "step": 5677 }, { "epoch": 0.9169896640826873, "grad_norm": 28.585285186767578, "learning_rate": 8.301033591731267e-07, "loss": 0.0556, "step": 5678 }, { "epoch": 0.9171511627906976, "grad_norm": 145.20968627929688, "learning_rate": 8.284883720930232e-07, "loss": 1.2471, "step": 5679 }, { "epoch": 0.917312661498708, "grad_norm": 20.792646408081055, "learning_rate": 8.2687338501292e-07, "loss": 0.1226, "step": 5680 }, { "epoch": 0.9174741602067183, "grad_norm": 17.226470947265625, "learning_rate": 8.252583979328166e-07, "loss": 0.0564, "step": 5681 }, { "epoch": 0.9176356589147286, "grad_norm": 83.02423095703125, "learning_rate": 8.236434108527132e-07, "loss": 1.1541, "step": 5682 }, { "epoch": 0.9177971576227391, "grad_norm": 191.15264892578125, "learning_rate": 8.220284237726099e-07, "loss": 1.6898, "step": 5683 }, { "epoch": 0.9179586563307494, "grad_norm": 86.12100982666016, "learning_rate": 8.204134366925066e-07, "loss": 0.5996, "step": 5684 }, { "epoch": 0.9181201550387597, "grad_norm": 46.8677864074707, "learning_rate": 8.187984496124032e-07, "loss": 0.1488, "step": 5685 }, { "epoch": 0.9182816537467701, "grad_norm": 28.42079734802246, "learning_rate": 8.171834625322998e-07, "loss": 0.0953, "step": 5686 }, { "epoch": 0.9184431524547804, "grad_norm": 47.3350944519043, "learning_rate": 8.155684754521964e-07, "loss": 0.2686, "step": 5687 }, { "epoch": 0.9186046511627907, "grad_norm": 231.41932678222656, "learning_rate": 8.139534883720931e-07, "loss": 1.9199, "step": 5688 }, { "epoch": 0.9187661498708011, "grad_norm": 37.968055725097656, "learning_rate": 8.123385012919898e-07, "loss": 0.1752, "step": 5689 }, { "epoch": 0.9189276485788114, "grad_norm": 428.0679016113281, "learning_rate": 8.107235142118863e-07, "loss": 1.8608, "step": 5690 }, { "epoch": 0.9190891472868217, "grad_norm": 18.76347541809082, "learning_rate": 8.091085271317831e-07, "loss": 0.0831, "step": 5691 }, { "epoch": 0.9192506459948321, "grad_norm": 40.21015167236328, "learning_rate": 8.074935400516797e-07, "loss": 0.1494, "step": 5692 }, { "epoch": 0.9194121447028424, "grad_norm": 364.2669982910156, "learning_rate": 8.058785529715762e-07, "loss": 2.1406, "step": 5693 }, { "epoch": 0.9195736434108527, "grad_norm": 79.65684509277344, "learning_rate": 8.04263565891473e-07, "loss": 1.037, "step": 5694 }, { "epoch": 0.9197351421188631, "grad_norm": 152.80169677734375, "learning_rate": 8.026485788113696e-07, "loss": 1.5596, "step": 5695 }, { "epoch": 0.9198966408268734, "grad_norm": 94.7992935180664, "learning_rate": 8.010335917312661e-07, "loss": 0.6299, "step": 5696 }, { "epoch": 0.9200581395348837, "grad_norm": 14.247613906860352, "learning_rate": 7.994186046511629e-07, "loss": 0.0317, "step": 5697 }, { "epoch": 0.9202196382428941, "grad_norm": 120.69756317138672, "learning_rate": 7.978036175710595e-07, "loss": 0.5641, "step": 5698 }, { "epoch": 0.9203811369509044, "grad_norm": 50.99098205566406, "learning_rate": 7.961886304909561e-07, "loss": 0.561, "step": 5699 }, { "epoch": 0.9205426356589147, "grad_norm": 10.521161079406738, "learning_rate": 7.945736434108528e-07, "loss": 0.0355, "step": 5700 }, { "epoch": 0.9207041343669251, "grad_norm": 112.46849060058594, "learning_rate": 7.929586563307493e-07, "loss": 0.481, "step": 5701 }, { "epoch": 0.9208656330749354, "grad_norm": 121.0560531616211, "learning_rate": 7.913436692506461e-07, "loss": 0.8486, "step": 5702 }, { "epoch": 0.9210271317829457, "grad_norm": 183.33963012695312, "learning_rate": 7.897286821705427e-07, "loss": 2.3022, "step": 5703 }, { "epoch": 0.9211886304909561, "grad_norm": 102.93045806884766, "learning_rate": 7.881136950904393e-07, "loss": 0.6861, "step": 5704 }, { "epoch": 0.9213501291989664, "grad_norm": 29.958545684814453, "learning_rate": 7.86498708010336e-07, "loss": 0.1395, "step": 5705 }, { "epoch": 0.9215116279069767, "grad_norm": 12.921989440917969, "learning_rate": 7.848837209302327e-07, "loss": 0.0353, "step": 5706 }, { "epoch": 0.9216731266149871, "grad_norm": 272.8630676269531, "learning_rate": 7.832687338501292e-07, "loss": 3.2545, "step": 5707 }, { "epoch": 0.9218346253229974, "grad_norm": 106.26055908203125, "learning_rate": 7.81653746770026e-07, "loss": 0.708, "step": 5708 }, { "epoch": 0.9219961240310077, "grad_norm": 315.0635681152344, "learning_rate": 7.800387596899226e-07, "loss": 1.2219, "step": 5709 }, { "epoch": 0.9221576227390181, "grad_norm": 48.60681915283203, "learning_rate": 7.784237726098191e-07, "loss": 0.1216, "step": 5710 }, { "epoch": 0.9223191214470284, "grad_norm": 47.4992561340332, "learning_rate": 7.768087855297159e-07, "loss": 0.128, "step": 5711 }, { "epoch": 0.9224806201550387, "grad_norm": 29.7581844329834, "learning_rate": 7.751937984496125e-07, "loss": 0.0382, "step": 5712 }, { "epoch": 0.9226421188630491, "grad_norm": 76.53771209716797, "learning_rate": 7.73578811369509e-07, "loss": 0.4541, "step": 5713 }, { "epoch": 0.9228036175710594, "grad_norm": 19.059823989868164, "learning_rate": 7.719638242894058e-07, "loss": 0.0496, "step": 5714 }, { "epoch": 0.9229651162790697, "grad_norm": 92.95579528808594, "learning_rate": 7.703488372093023e-07, "loss": 0.4934, "step": 5715 }, { "epoch": 0.9231266149870802, "grad_norm": 20.964824676513672, "learning_rate": 7.68733850129199e-07, "loss": 0.05, "step": 5716 }, { "epoch": 0.9232881136950905, "grad_norm": 22.626964569091797, "learning_rate": 7.671188630490957e-07, "loss": 0.0807, "step": 5717 }, { "epoch": 0.9234496124031008, "grad_norm": 126.17631530761719, "learning_rate": 7.655038759689922e-07, "loss": 0.7402, "step": 5718 }, { "epoch": 0.9236111111111112, "grad_norm": 110.9935302734375, "learning_rate": 7.63888888888889e-07, "loss": 0.4482, "step": 5719 }, { "epoch": 0.9237726098191215, "grad_norm": 104.26549530029297, "learning_rate": 7.622739018087856e-07, "loss": 0.2578, "step": 5720 }, { "epoch": 0.9239341085271318, "grad_norm": 238.70668029785156, "learning_rate": 7.606589147286822e-07, "loss": 1.1206, "step": 5721 }, { "epoch": 0.9240956072351422, "grad_norm": 167.36572265625, "learning_rate": 7.590439276485789e-07, "loss": 0.4734, "step": 5722 }, { "epoch": 0.9242571059431525, "grad_norm": 32.67929458618164, "learning_rate": 7.574289405684756e-07, "loss": 0.0424, "step": 5723 }, { "epoch": 0.9244186046511628, "grad_norm": 15.68513298034668, "learning_rate": 7.558139534883721e-07, "loss": 0.0585, "step": 5724 }, { "epoch": 0.9245801033591732, "grad_norm": 294.3316955566406, "learning_rate": 7.541989664082688e-07, "loss": 0.9932, "step": 5725 }, { "epoch": 0.9247416020671835, "grad_norm": 47.63167190551758, "learning_rate": 7.525839793281655e-07, "loss": 0.2189, "step": 5726 }, { "epoch": 0.9249031007751938, "grad_norm": 7.551253795623779, "learning_rate": 7.50968992248062e-07, "loss": 0.018, "step": 5727 }, { "epoch": 0.9250645994832042, "grad_norm": 20.324636459350586, "learning_rate": 7.493540051679588e-07, "loss": 0.0685, "step": 5728 }, { "epoch": 0.9252260981912145, "grad_norm": 48.524497985839844, "learning_rate": 7.477390180878554e-07, "loss": 0.1768, "step": 5729 }, { "epoch": 0.9253875968992248, "grad_norm": 20.903602600097656, "learning_rate": 7.461240310077519e-07, "loss": 0.0432, "step": 5730 }, { "epoch": 0.9255490956072352, "grad_norm": 29.497100830078125, "learning_rate": 7.445090439276487e-07, "loss": 0.0647, "step": 5731 }, { "epoch": 0.9257105943152455, "grad_norm": 56.16791915893555, "learning_rate": 7.428940568475452e-07, "loss": 0.1511, "step": 5732 }, { "epoch": 0.9258720930232558, "grad_norm": 60.54449462890625, "learning_rate": 7.41279069767442e-07, "loss": 0.2319, "step": 5733 }, { "epoch": 0.9260335917312662, "grad_norm": 53.655338287353516, "learning_rate": 7.396640826873386e-07, "loss": 0.2187, "step": 5734 }, { "epoch": 0.9261950904392765, "grad_norm": 31.506555557250977, "learning_rate": 7.380490956072351e-07, "loss": 0.1067, "step": 5735 }, { "epoch": 0.9263565891472868, "grad_norm": 215.4676971435547, "learning_rate": 7.364341085271319e-07, "loss": 0.6465, "step": 5736 }, { "epoch": 0.9265180878552972, "grad_norm": 3.9029839038848877, "learning_rate": 7.348191214470285e-07, "loss": 0.0133, "step": 5737 }, { "epoch": 0.9266795865633075, "grad_norm": 19.784839630126953, "learning_rate": 7.332041343669251e-07, "loss": 0.0757, "step": 5738 }, { "epoch": 0.9268410852713178, "grad_norm": 26.771188735961914, "learning_rate": 7.315891472868218e-07, "loss": 0.0581, "step": 5739 }, { "epoch": 0.9270025839793282, "grad_norm": 87.23326873779297, "learning_rate": 7.299741602067184e-07, "loss": 0.4606, "step": 5740 }, { "epoch": 0.9271640826873385, "grad_norm": 110.8184585571289, "learning_rate": 7.28359173126615e-07, "loss": 1.1978, "step": 5741 }, { "epoch": 0.9273255813953488, "grad_norm": 19.96223258972168, "learning_rate": 7.267441860465117e-07, "loss": 0.0874, "step": 5742 }, { "epoch": 0.9274870801033591, "grad_norm": 44.66018295288086, "learning_rate": 7.251291989664084e-07, "loss": 0.0637, "step": 5743 }, { "epoch": 0.9276485788113695, "grad_norm": 19.885534286499023, "learning_rate": 7.235142118863049e-07, "loss": 0.0476, "step": 5744 }, { "epoch": 0.9278100775193798, "grad_norm": 17.70598793029785, "learning_rate": 7.218992248062017e-07, "loss": 0.094, "step": 5745 }, { "epoch": 0.9279715762273901, "grad_norm": 34.76509094238281, "learning_rate": 7.202842377260982e-07, "loss": 0.1234, "step": 5746 }, { "epoch": 0.9281330749354005, "grad_norm": 114.21424865722656, "learning_rate": 7.186692506459948e-07, "loss": 0.8203, "step": 5747 }, { "epoch": 0.9282945736434108, "grad_norm": 52.23902893066406, "learning_rate": 7.170542635658916e-07, "loss": 0.2368, "step": 5748 }, { "epoch": 0.9284560723514211, "grad_norm": 33.00517654418945, "learning_rate": 7.154392764857881e-07, "loss": 0.1718, "step": 5749 }, { "epoch": 0.9286175710594315, "grad_norm": 127.0212631225586, "learning_rate": 7.138242894056849e-07, "loss": 0.123, "step": 5750 }, { "epoch": 0.9287790697674418, "grad_norm": 22.06593132019043, "learning_rate": 7.122093023255815e-07, "loss": 0.0941, "step": 5751 }, { "epoch": 0.9289405684754521, "grad_norm": 80.70235443115234, "learning_rate": 7.10594315245478e-07, "loss": 0.542, "step": 5752 }, { "epoch": 0.9291020671834626, "grad_norm": 46.6558952331543, "learning_rate": 7.089793281653748e-07, "loss": 0.3174, "step": 5753 }, { "epoch": 0.9292635658914729, "grad_norm": 82.10566711425781, "learning_rate": 7.073643410852714e-07, "loss": 0.5693, "step": 5754 }, { "epoch": 0.9294250645994832, "grad_norm": 246.78323364257812, "learning_rate": 7.057493540051679e-07, "loss": 0.8876, "step": 5755 }, { "epoch": 0.9295865633074936, "grad_norm": 13.577047348022461, "learning_rate": 7.041343669250647e-07, "loss": 0.0335, "step": 5756 }, { "epoch": 0.9297480620155039, "grad_norm": 18.769790649414062, "learning_rate": 7.025193798449613e-07, "loss": 0.0345, "step": 5757 }, { "epoch": 0.9299095607235142, "grad_norm": 9.914336204528809, "learning_rate": 7.009043927648579e-07, "loss": 0.0329, "step": 5758 }, { "epoch": 0.9300710594315246, "grad_norm": 17.053422927856445, "learning_rate": 6.992894056847546e-07, "loss": 0.0435, "step": 5759 }, { "epoch": 0.9302325581395349, "grad_norm": 61.951683044433594, "learning_rate": 6.976744186046513e-07, "loss": 0.1151, "step": 5760 }, { "epoch": 0.9303940568475452, "grad_norm": 37.07258605957031, "learning_rate": 6.960594315245478e-07, "loss": 0.0962, "step": 5761 }, { "epoch": 0.9305555555555556, "grad_norm": 243.86912536621094, "learning_rate": 6.944444444444446e-07, "loss": 1.1782, "step": 5762 }, { "epoch": 0.9307170542635659, "grad_norm": 34.466121673583984, "learning_rate": 6.928294573643411e-07, "loss": 0.1347, "step": 5763 }, { "epoch": 0.9308785529715762, "grad_norm": 18.821928024291992, "learning_rate": 6.912144702842377e-07, "loss": 0.0493, "step": 5764 }, { "epoch": 0.9310400516795866, "grad_norm": 53.58086395263672, "learning_rate": 6.895994832041345e-07, "loss": 0.2323, "step": 5765 }, { "epoch": 0.9312015503875969, "grad_norm": 38.45888900756836, "learning_rate": 6.87984496124031e-07, "loss": 0.2334, "step": 5766 }, { "epoch": 0.9313630490956072, "grad_norm": 27.516387939453125, "learning_rate": 6.863695090439278e-07, "loss": 0.1431, "step": 5767 }, { "epoch": 0.9315245478036176, "grad_norm": 18.792890548706055, "learning_rate": 6.847545219638244e-07, "loss": 0.0427, "step": 5768 }, { "epoch": 0.9316860465116279, "grad_norm": 51.37224578857422, "learning_rate": 6.831395348837209e-07, "loss": 0.239, "step": 5769 }, { "epoch": 0.9318475452196382, "grad_norm": 31.17441177368164, "learning_rate": 6.815245478036177e-07, "loss": 0.1558, "step": 5770 }, { "epoch": 0.9320090439276486, "grad_norm": 39.74771499633789, "learning_rate": 6.799095607235143e-07, "loss": 0.1497, "step": 5771 }, { "epoch": 0.9321705426356589, "grad_norm": 36.982242584228516, "learning_rate": 6.782945736434108e-07, "loss": 0.0947, "step": 5772 }, { "epoch": 0.9323320413436692, "grad_norm": 144.94602966308594, "learning_rate": 6.766795865633076e-07, "loss": 0.6289, "step": 5773 }, { "epoch": 0.9324935400516796, "grad_norm": 2.9076435565948486, "learning_rate": 6.750645994832042e-07, "loss": 0.0062, "step": 5774 }, { "epoch": 0.9326550387596899, "grad_norm": 27.255281448364258, "learning_rate": 6.734496124031008e-07, "loss": 0.0633, "step": 5775 }, { "epoch": 0.9328165374677002, "grad_norm": 25.552738189697266, "learning_rate": 6.718346253229975e-07, "loss": 0.0592, "step": 5776 }, { "epoch": 0.9329780361757106, "grad_norm": 111.24880981445312, "learning_rate": 6.702196382428941e-07, "loss": 0.4626, "step": 5777 }, { "epoch": 0.9331395348837209, "grad_norm": 21.795461654663086, "learning_rate": 6.686046511627907e-07, "loss": 0.0652, "step": 5778 }, { "epoch": 0.9333010335917312, "grad_norm": 58.90971374511719, "learning_rate": 6.669896640826874e-07, "loss": 0.1923, "step": 5779 }, { "epoch": 0.9334625322997416, "grad_norm": 9.21917724609375, "learning_rate": 6.65374677002584e-07, "loss": 0.0336, "step": 5780 }, { "epoch": 0.9336240310077519, "grad_norm": 140.50926208496094, "learning_rate": 6.637596899224807e-07, "loss": 0.4404, "step": 5781 }, { "epoch": 0.9337855297157622, "grad_norm": 11.414021492004395, "learning_rate": 6.621447028423774e-07, "loss": 0.0256, "step": 5782 }, { "epoch": 0.9339470284237726, "grad_norm": 47.478328704833984, "learning_rate": 6.605297157622739e-07, "loss": 0.1992, "step": 5783 }, { "epoch": 0.9341085271317829, "grad_norm": 67.67344665527344, "learning_rate": 6.589147286821707e-07, "loss": 0.2729, "step": 5784 }, { "epoch": 0.9342700258397932, "grad_norm": 3.3343844413757324, "learning_rate": 6.572997416020673e-07, "loss": 0.011, "step": 5785 }, { "epoch": 0.9344315245478036, "grad_norm": 27.34969711303711, "learning_rate": 6.556847545219638e-07, "loss": 0.0636, "step": 5786 }, { "epoch": 0.934593023255814, "grad_norm": 79.119384765625, "learning_rate": 6.540697674418606e-07, "loss": 0.4454, "step": 5787 }, { "epoch": 0.9347545219638242, "grad_norm": 34.43779754638672, "learning_rate": 6.524547803617572e-07, "loss": 0.1097, "step": 5788 }, { "epoch": 0.9349160206718347, "grad_norm": 17.87405014038086, "learning_rate": 6.508397932816537e-07, "loss": 0.0447, "step": 5789 }, { "epoch": 0.935077519379845, "grad_norm": 30.783367156982422, "learning_rate": 6.492248062015505e-07, "loss": 0.1074, "step": 5790 }, { "epoch": 0.9352390180878553, "grad_norm": 179.9691925048828, "learning_rate": 6.47609819121447e-07, "loss": 0.9639, "step": 5791 }, { "epoch": 0.9354005167958657, "grad_norm": 349.13116455078125, "learning_rate": 6.459948320413437e-07, "loss": 0.5789, "step": 5792 }, { "epoch": 0.935562015503876, "grad_norm": 27.345462799072266, "learning_rate": 6.443798449612404e-07, "loss": 0.0851, "step": 5793 }, { "epoch": 0.9357235142118863, "grad_norm": 88.18682098388672, "learning_rate": 6.427648578811369e-07, "loss": 0.2217, "step": 5794 }, { "epoch": 0.9358850129198967, "grad_norm": 194.77706909179688, "learning_rate": 6.411498708010336e-07, "loss": 0.5957, "step": 5795 }, { "epoch": 0.936046511627907, "grad_norm": 239.37559509277344, "learning_rate": 6.395348837209303e-07, "loss": 1.2422, "step": 5796 }, { "epoch": 0.9362080103359173, "grad_norm": 7.860161304473877, "learning_rate": 6.379198966408269e-07, "loss": 0.0168, "step": 5797 }, { "epoch": 0.9363695090439277, "grad_norm": 93.60726928710938, "learning_rate": 6.363049095607236e-07, "loss": 0.4399, "step": 5798 }, { "epoch": 0.936531007751938, "grad_norm": 2.8117856979370117, "learning_rate": 6.346899224806203e-07, "loss": 0.0095, "step": 5799 }, { "epoch": 0.9366925064599483, "grad_norm": 126.10808563232422, "learning_rate": 6.330749354005168e-07, "loss": 0.4658, "step": 5800 }, { "epoch": 0.9368540051679587, "grad_norm": 37.2819709777832, "learning_rate": 6.314599483204135e-07, "loss": 0.1387, "step": 5801 }, { "epoch": 0.937015503875969, "grad_norm": 76.89386749267578, "learning_rate": 6.298449612403102e-07, "loss": 0.4424, "step": 5802 }, { "epoch": 0.9371770025839793, "grad_norm": 22.589820861816406, "learning_rate": 6.282299741602067e-07, "loss": 0.0415, "step": 5803 }, { "epoch": 0.9373385012919897, "grad_norm": 18.860679626464844, "learning_rate": 6.266149870801035e-07, "loss": 0.051, "step": 5804 }, { "epoch": 0.9375, "grad_norm": 40.369361877441406, "learning_rate": 6.25e-07, "loss": 0.0895, "step": 5805 }, { "epoch": 0.9376614987080103, "grad_norm": 29.911056518554688, "learning_rate": 6.233850129198967e-07, "loss": 0.0707, "step": 5806 }, { "epoch": 0.9378229974160207, "grad_norm": 8.859254837036133, "learning_rate": 6.217700258397933e-07, "loss": 0.0375, "step": 5807 }, { "epoch": 0.937984496124031, "grad_norm": 373.68511962890625, "learning_rate": 6.201550387596899e-07, "loss": 2.7368, "step": 5808 }, { "epoch": 0.9381459948320413, "grad_norm": 98.12491607666016, "learning_rate": 6.185400516795866e-07, "loss": 0.4248, "step": 5809 }, { "epoch": 0.9383074935400517, "grad_norm": 76.76329803466797, "learning_rate": 6.169250645994833e-07, "loss": 0.2996, "step": 5810 }, { "epoch": 0.938468992248062, "grad_norm": 118.49623107910156, "learning_rate": 6.153100775193798e-07, "loss": 0.5205, "step": 5811 }, { "epoch": 0.9386304909560723, "grad_norm": 87.0689468383789, "learning_rate": 6.136950904392765e-07, "loss": 0.2129, "step": 5812 }, { "epoch": 0.9387919896640827, "grad_norm": 34.25596237182617, "learning_rate": 6.120801033591732e-07, "loss": 0.1272, "step": 5813 }, { "epoch": 0.938953488372093, "grad_norm": 102.79739379882812, "learning_rate": 6.104651162790698e-07, "loss": 0.3295, "step": 5814 }, { "epoch": 0.9391149870801033, "grad_norm": 16.21625328063965, "learning_rate": 6.088501291989665e-07, "loss": 0.0336, "step": 5815 }, { "epoch": 0.9392764857881137, "grad_norm": 441.0455017089844, "learning_rate": 6.072351421188632e-07, "loss": 1.772, "step": 5816 }, { "epoch": 0.939437984496124, "grad_norm": 9.385475158691406, "learning_rate": 6.056201550387597e-07, "loss": 0.0253, "step": 5817 }, { "epoch": 0.9395994832041343, "grad_norm": 87.60160064697266, "learning_rate": 6.040051679586564e-07, "loss": 0.4111, "step": 5818 }, { "epoch": 0.9397609819121447, "grad_norm": 35.83944320678711, "learning_rate": 6.023901808785531e-07, "loss": 0.1626, "step": 5819 }, { "epoch": 0.939922480620155, "grad_norm": 12.658404350280762, "learning_rate": 6.007751937984497e-07, "loss": 0.0235, "step": 5820 }, { "epoch": 0.9400839793281653, "grad_norm": 24.796470642089844, "learning_rate": 5.991602067183463e-07, "loss": 0.0746, "step": 5821 }, { "epoch": 0.9402454780361758, "grad_norm": 6.761043548583984, "learning_rate": 5.97545219638243e-07, "loss": 0.0151, "step": 5822 }, { "epoch": 0.940406976744186, "grad_norm": 23.13376808166504, "learning_rate": 5.959302325581396e-07, "loss": 0.0547, "step": 5823 }, { "epoch": 0.9405684754521964, "grad_norm": 3.6765098571777344, "learning_rate": 5.943152454780362e-07, "loss": 0.0216, "step": 5824 }, { "epoch": 0.9407299741602068, "grad_norm": 74.19940185546875, "learning_rate": 5.927002583979328e-07, "loss": 0.3572, "step": 5825 }, { "epoch": 0.9408914728682171, "grad_norm": 138.96212768554688, "learning_rate": 5.910852713178295e-07, "loss": 0.5747, "step": 5826 }, { "epoch": 0.9410529715762274, "grad_norm": 162.65586853027344, "learning_rate": 5.894702842377262e-07, "loss": 0.7339, "step": 5827 }, { "epoch": 0.9412144702842378, "grad_norm": 230.9801483154297, "learning_rate": 5.878552971576227e-07, "loss": 2.8345, "step": 5828 }, { "epoch": 0.9413759689922481, "grad_norm": 13.591398239135742, "learning_rate": 5.862403100775194e-07, "loss": 0.0417, "step": 5829 }, { "epoch": 0.9415374677002584, "grad_norm": 145.98178100585938, "learning_rate": 5.846253229974161e-07, "loss": 0.459, "step": 5830 }, { "epoch": 0.9416989664082688, "grad_norm": 120.96601104736328, "learning_rate": 5.830103359173127e-07, "loss": 0.5234, "step": 5831 }, { "epoch": 0.9418604651162791, "grad_norm": 33.9947624206543, "learning_rate": 5.813953488372094e-07, "loss": 0.1433, "step": 5832 }, { "epoch": 0.9420219638242894, "grad_norm": 133.96995544433594, "learning_rate": 5.79780361757106e-07, "loss": 0.2871, "step": 5833 }, { "epoch": 0.9421834625322998, "grad_norm": 23.027610778808594, "learning_rate": 5.781653746770026e-07, "loss": 0.0702, "step": 5834 }, { "epoch": 0.9423449612403101, "grad_norm": 234.86216735839844, "learning_rate": 5.765503875968993e-07, "loss": 0.5682, "step": 5835 }, { "epoch": 0.9425064599483204, "grad_norm": 16.395225524902344, "learning_rate": 5.74935400516796e-07, "loss": 0.0359, "step": 5836 }, { "epoch": 0.9426679586563308, "grad_norm": 56.66997528076172, "learning_rate": 5.733204134366926e-07, "loss": 0.3109, "step": 5837 }, { "epoch": 0.9428294573643411, "grad_norm": 110.46503448486328, "learning_rate": 5.717054263565892e-07, "loss": 0.7772, "step": 5838 }, { "epoch": 0.9429909560723514, "grad_norm": 49.941768646240234, "learning_rate": 5.700904392764858e-07, "loss": 0.0808, "step": 5839 }, { "epoch": 0.9431524547803618, "grad_norm": 12.038908958435059, "learning_rate": 5.684754521963825e-07, "loss": 0.0286, "step": 5840 }, { "epoch": 0.9433139534883721, "grad_norm": 102.65005493164062, "learning_rate": 5.668604651162791e-07, "loss": 0.4664, "step": 5841 }, { "epoch": 0.9434754521963824, "grad_norm": 352.8171691894531, "learning_rate": 5.652454780361757e-07, "loss": 1.7889, "step": 5842 }, { "epoch": 0.9436369509043928, "grad_norm": 8.421989440917969, "learning_rate": 5.636304909560724e-07, "loss": 0.0359, "step": 5843 }, { "epoch": 0.9437984496124031, "grad_norm": 40.65839767456055, "learning_rate": 5.620155038759691e-07, "loss": 0.1187, "step": 5844 }, { "epoch": 0.9439599483204134, "grad_norm": 381.61004638671875, "learning_rate": 5.604005167958656e-07, "loss": 1.9014, "step": 5845 }, { "epoch": 0.9441214470284238, "grad_norm": 92.86785125732422, "learning_rate": 5.587855297157623e-07, "loss": 0.5202, "step": 5846 }, { "epoch": 0.9442829457364341, "grad_norm": 5.4394211769104, "learning_rate": 5.57170542635659e-07, "loss": 0.018, "step": 5847 }, { "epoch": 0.9444444444444444, "grad_norm": 9.983192443847656, "learning_rate": 5.555555555555555e-07, "loss": 0.0201, "step": 5848 }, { "epoch": 0.9446059431524548, "grad_norm": 128.94638061523438, "learning_rate": 5.539405684754522e-07, "loss": 0.8374, "step": 5849 }, { "epoch": 0.9447674418604651, "grad_norm": 87.2095947265625, "learning_rate": 5.523255813953489e-07, "loss": 0.376, "step": 5850 }, { "epoch": 0.9449289405684754, "grad_norm": 99.29019927978516, "learning_rate": 5.507105943152455e-07, "loss": 0.8965, "step": 5851 }, { "epoch": 0.9450904392764858, "grad_norm": 35.43587112426758, "learning_rate": 5.490956072351422e-07, "loss": 0.0695, "step": 5852 }, { "epoch": 0.9452519379844961, "grad_norm": 141.41583251953125, "learning_rate": 5.474806201550388e-07, "loss": 0.6562, "step": 5853 }, { "epoch": 0.9454134366925064, "grad_norm": 45.40204620361328, "learning_rate": 5.458656330749355e-07, "loss": 0.1065, "step": 5854 }, { "epoch": 0.9455749354005168, "grad_norm": 16.233295440673828, "learning_rate": 5.442506459948321e-07, "loss": 0.0283, "step": 5855 }, { "epoch": 0.9457364341085271, "grad_norm": 26.53812026977539, "learning_rate": 5.426356589147287e-07, "loss": 0.1012, "step": 5856 }, { "epoch": 0.9458979328165374, "grad_norm": 11.803215980529785, "learning_rate": 5.410206718346254e-07, "loss": 0.0114, "step": 5857 }, { "epoch": 0.9460594315245479, "grad_norm": 24.844175338745117, "learning_rate": 5.394056847545221e-07, "loss": 0.0675, "step": 5858 }, { "epoch": 0.9462209302325582, "grad_norm": 51.118534088134766, "learning_rate": 5.377906976744186e-07, "loss": 0.3159, "step": 5859 }, { "epoch": 0.9463824289405685, "grad_norm": 33.93734359741211, "learning_rate": 5.361757105943153e-07, "loss": 0.1638, "step": 5860 }, { "epoch": 0.9465439276485789, "grad_norm": 8.93449878692627, "learning_rate": 5.34560723514212e-07, "loss": 0.0278, "step": 5861 }, { "epoch": 0.9467054263565892, "grad_norm": 81.6197509765625, "learning_rate": 5.329457364341085e-07, "loss": 0.6475, "step": 5862 }, { "epoch": 0.9468669250645995, "grad_norm": 17.954893112182617, "learning_rate": 5.313307493540052e-07, "loss": 0.0891, "step": 5863 }, { "epoch": 0.9470284237726099, "grad_norm": 72.99421691894531, "learning_rate": 5.297157622739019e-07, "loss": 0.348, "step": 5864 }, { "epoch": 0.9471899224806202, "grad_norm": 10.571310043334961, "learning_rate": 5.281007751937984e-07, "loss": 0.0389, "step": 5865 }, { "epoch": 0.9473514211886305, "grad_norm": 24.280691146850586, "learning_rate": 5.264857881136951e-07, "loss": 0.0627, "step": 5866 }, { "epoch": 0.9475129198966409, "grad_norm": 101.54000854492188, "learning_rate": 5.248708010335917e-07, "loss": 0.3112, "step": 5867 }, { "epoch": 0.9476744186046512, "grad_norm": 41.313453674316406, "learning_rate": 5.232558139534884e-07, "loss": 0.2197, "step": 5868 }, { "epoch": 0.9478359173126615, "grad_norm": 18.4427490234375, "learning_rate": 5.216408268733851e-07, "loss": 0.0605, "step": 5869 }, { "epoch": 0.9479974160206718, "grad_norm": 9.342970848083496, "learning_rate": 5.200258397932817e-07, "loss": 0.0474, "step": 5870 }, { "epoch": 0.9481589147286822, "grad_norm": 3.7514119148254395, "learning_rate": 5.184108527131783e-07, "loss": 0.0141, "step": 5871 }, { "epoch": 0.9483204134366925, "grad_norm": 229.212890625, "learning_rate": 5.16795865633075e-07, "loss": 1.7109, "step": 5872 }, { "epoch": 0.9484819121447028, "grad_norm": 24.41523551940918, "learning_rate": 5.151808785529716e-07, "loss": 0.0629, "step": 5873 }, { "epoch": 0.9486434108527132, "grad_norm": 12.102143287658691, "learning_rate": 5.135658914728683e-07, "loss": 0.0386, "step": 5874 }, { "epoch": 0.9488049095607235, "grad_norm": 163.27621459960938, "learning_rate": 5.11950904392765e-07, "loss": 2.0605, "step": 5875 }, { "epoch": 0.9489664082687338, "grad_norm": 98.11262512207031, "learning_rate": 5.103359173126615e-07, "loss": 0.8048, "step": 5876 }, { "epoch": 0.9491279069767442, "grad_norm": 33.72016906738281, "learning_rate": 5.087209302325582e-07, "loss": 0.2354, "step": 5877 }, { "epoch": 0.9492894056847545, "grad_norm": 241.86595153808594, "learning_rate": 5.071059431524549e-07, "loss": 0.53, "step": 5878 }, { "epoch": 0.9494509043927648, "grad_norm": 8.330443382263184, "learning_rate": 5.054909560723514e-07, "loss": 0.0184, "step": 5879 }, { "epoch": 0.9496124031007752, "grad_norm": 102.32926940917969, "learning_rate": 5.038759689922481e-07, "loss": 0.3855, "step": 5880 }, { "epoch": 0.9497739018087855, "grad_norm": 130.3855438232422, "learning_rate": 5.022609819121448e-07, "loss": 0.2874, "step": 5881 }, { "epoch": 0.9499354005167958, "grad_norm": 180.7582244873047, "learning_rate": 5.006459948320414e-07, "loss": 1.1499, "step": 5882 }, { "epoch": 0.9500968992248062, "grad_norm": 55.85865783691406, "learning_rate": 4.99031007751938e-07, "loss": 0.1067, "step": 5883 }, { "epoch": 0.9502583979328165, "grad_norm": 8.67647933959961, "learning_rate": 4.974160206718346e-07, "loss": 0.0275, "step": 5884 }, { "epoch": 0.9504198966408268, "grad_norm": 17.75741958618164, "learning_rate": 4.958010335917313e-07, "loss": 0.0309, "step": 5885 }, { "epoch": 0.9505813953488372, "grad_norm": 76.8814926147461, "learning_rate": 4.94186046511628e-07, "loss": 0.5486, "step": 5886 }, { "epoch": 0.9507428940568475, "grad_norm": 3.5083138942718506, "learning_rate": 4.925710594315245e-07, "loss": 0.014, "step": 5887 }, { "epoch": 0.9509043927648578, "grad_norm": 60.44778823852539, "learning_rate": 4.909560723514212e-07, "loss": 0.8633, "step": 5888 }, { "epoch": 0.9510658914728682, "grad_norm": 13.47667121887207, "learning_rate": 4.893410852713179e-07, "loss": 0.0472, "step": 5889 }, { "epoch": 0.9512273901808785, "grad_norm": 20.217866897583008, "learning_rate": 4.877260981912145e-07, "loss": 0.0713, "step": 5890 }, { "epoch": 0.9513888888888888, "grad_norm": 31.165283203125, "learning_rate": 4.861111111111112e-07, "loss": 0.0776, "step": 5891 }, { "epoch": 0.9515503875968992, "grad_norm": 53.7841682434082, "learning_rate": 4.844961240310079e-07, "loss": 0.1604, "step": 5892 }, { "epoch": 0.9517118863049095, "grad_norm": 61.480224609375, "learning_rate": 4.828811369509044e-07, "loss": 0.1238, "step": 5893 }, { "epoch": 0.9518733850129198, "grad_norm": 176.0103302001953, "learning_rate": 4.812661498708011e-07, "loss": 0.4844, "step": 5894 }, { "epoch": 0.9520348837209303, "grad_norm": 503.0207214355469, "learning_rate": 4.796511627906978e-07, "loss": 2.627, "step": 5895 }, { "epoch": 0.9521963824289406, "grad_norm": 163.5226287841797, "learning_rate": 4.780361757105943e-07, "loss": 0.7227, "step": 5896 }, { "epoch": 0.9523578811369509, "grad_norm": 192.2053985595703, "learning_rate": 4.76421188630491e-07, "loss": 1.134, "step": 5897 }, { "epoch": 0.9525193798449613, "grad_norm": 119.38048553466797, "learning_rate": 4.7480620155038765e-07, "loss": 0.9282, "step": 5898 }, { "epoch": 0.9526808785529716, "grad_norm": 135.47232055664062, "learning_rate": 4.731912144702843e-07, "loss": 0.6289, "step": 5899 }, { "epoch": 0.9528423772609819, "grad_norm": 28.449060440063477, "learning_rate": 4.7157622739018087e-07, "loss": 0.2117, "step": 5900 }, { "epoch": 0.9530038759689923, "grad_norm": 16.401592254638672, "learning_rate": 4.6996124031007754e-07, "loss": 0.0358, "step": 5901 }, { "epoch": 0.9531653746770026, "grad_norm": 35.21873474121094, "learning_rate": 4.683462532299742e-07, "loss": 0.1453, "step": 5902 }, { "epoch": 0.9533268733850129, "grad_norm": 45.679718017578125, "learning_rate": 4.667312661498708e-07, "loss": 0.1216, "step": 5903 }, { "epoch": 0.9534883720930233, "grad_norm": 22.327739715576172, "learning_rate": 4.651162790697675e-07, "loss": 0.1375, "step": 5904 }, { "epoch": 0.9536498708010336, "grad_norm": 40.21857833862305, "learning_rate": 4.6350129198966415e-07, "loss": 0.1545, "step": 5905 }, { "epoch": 0.9538113695090439, "grad_norm": 33.91249084472656, "learning_rate": 4.618863049095608e-07, "loss": 0.0675, "step": 5906 }, { "epoch": 0.9539728682170543, "grad_norm": 23.569355010986328, "learning_rate": 4.602713178294574e-07, "loss": 0.0618, "step": 5907 }, { "epoch": 0.9541343669250646, "grad_norm": 32.967525482177734, "learning_rate": 4.5865633074935404e-07, "loss": 0.0544, "step": 5908 }, { "epoch": 0.9542958656330749, "grad_norm": 122.98624420166016, "learning_rate": 4.570413436692507e-07, "loss": 0.3887, "step": 5909 }, { "epoch": 0.9544573643410853, "grad_norm": 361.03314208984375, "learning_rate": 4.554263565891473e-07, "loss": 0.8336, "step": 5910 }, { "epoch": 0.9546188630490956, "grad_norm": 39.33354949951172, "learning_rate": 4.53811369509044e-07, "loss": 0.2017, "step": 5911 }, { "epoch": 0.9547803617571059, "grad_norm": 640.5374145507812, "learning_rate": 4.521963824289406e-07, "loss": 2.332, "step": 5912 }, { "epoch": 0.9549418604651163, "grad_norm": 105.46931457519531, "learning_rate": 4.505813953488372e-07, "loss": 0.8887, "step": 5913 }, { "epoch": 0.9551033591731266, "grad_norm": 8.882101058959961, "learning_rate": 4.489664082687339e-07, "loss": 0.0257, "step": 5914 }, { "epoch": 0.9552648578811369, "grad_norm": 18.509010314941406, "learning_rate": 4.4735142118863054e-07, "loss": 0.052, "step": 5915 }, { "epoch": 0.9554263565891473, "grad_norm": 27.90040397644043, "learning_rate": 4.457364341085272e-07, "loss": 0.0543, "step": 5916 }, { "epoch": 0.9555878552971576, "grad_norm": 32.514461517333984, "learning_rate": 4.4412144702842377e-07, "loss": 0.1384, "step": 5917 }, { "epoch": 0.9557493540051679, "grad_norm": 81.29673767089844, "learning_rate": 4.4250645994832044e-07, "loss": 0.1131, "step": 5918 }, { "epoch": 0.9559108527131783, "grad_norm": 13.219732284545898, "learning_rate": 4.408914728682171e-07, "loss": 0.0314, "step": 5919 }, { "epoch": 0.9560723514211886, "grad_norm": 154.9604034423828, "learning_rate": 4.392764857881137e-07, "loss": 0.8291, "step": 5920 }, { "epoch": 0.9562338501291989, "grad_norm": 47.889461517333984, "learning_rate": 4.376614987080104e-07, "loss": 0.2612, "step": 5921 }, { "epoch": 0.9563953488372093, "grad_norm": 7.761568069458008, "learning_rate": 4.3604651162790705e-07, "loss": 0.0374, "step": 5922 }, { "epoch": 0.9565568475452196, "grad_norm": 40.18340301513672, "learning_rate": 4.3443152454780366e-07, "loss": 0.1218, "step": 5923 }, { "epoch": 0.9567183462532299, "grad_norm": 8.51230525970459, "learning_rate": 4.3281653746770027e-07, "loss": 0.0206, "step": 5924 }, { "epoch": 0.9568798449612403, "grad_norm": 27.16123390197754, "learning_rate": 4.3120155038759694e-07, "loss": 0.0691, "step": 5925 }, { "epoch": 0.9570413436692506, "grad_norm": 8.385797500610352, "learning_rate": 4.295865633074936e-07, "loss": 0.021, "step": 5926 }, { "epoch": 0.9572028423772609, "grad_norm": 19.037874221801758, "learning_rate": 4.279715762273902e-07, "loss": 0.0619, "step": 5927 }, { "epoch": 0.9573643410852714, "grad_norm": 24.63016128540039, "learning_rate": 4.2635658914728683e-07, "loss": 0.0519, "step": 5928 }, { "epoch": 0.9575258397932817, "grad_norm": 101.54861450195312, "learning_rate": 4.247416020671835e-07, "loss": 0.343, "step": 5929 }, { "epoch": 0.957687338501292, "grad_norm": 26.302398681640625, "learning_rate": 4.2312661498708016e-07, "loss": 0.1073, "step": 5930 }, { "epoch": 0.9578488372093024, "grad_norm": 49.424217224121094, "learning_rate": 4.215116279069768e-07, "loss": 0.1582, "step": 5931 }, { "epoch": 0.9580103359173127, "grad_norm": 21.570518493652344, "learning_rate": 4.1989664082687344e-07, "loss": 0.0528, "step": 5932 }, { "epoch": 0.958171834625323, "grad_norm": 95.95063781738281, "learning_rate": 4.182816537467701e-07, "loss": 0.3833, "step": 5933 }, { "epoch": 0.9583333333333334, "grad_norm": 91.2730941772461, "learning_rate": 4.1666666666666667e-07, "loss": 0.2941, "step": 5934 }, { "epoch": 0.9584948320413437, "grad_norm": 12.306938171386719, "learning_rate": 4.1505167958656333e-07, "loss": 0.0525, "step": 5935 }, { "epoch": 0.958656330749354, "grad_norm": 17.130550384521484, "learning_rate": 4.1343669250646e-07, "loss": 0.039, "step": 5936 }, { "epoch": 0.9588178294573644, "grad_norm": 7.3586602210998535, "learning_rate": 4.118217054263566e-07, "loss": 0.0168, "step": 5937 }, { "epoch": 0.9589793281653747, "grad_norm": 8.508479118347168, "learning_rate": 4.102067183462533e-07, "loss": 0.0121, "step": 5938 }, { "epoch": 0.959140826873385, "grad_norm": 57.3703727722168, "learning_rate": 4.085917312661499e-07, "loss": 0.1646, "step": 5939 }, { "epoch": 0.9593023255813954, "grad_norm": 8.535802841186523, "learning_rate": 4.0697674418604655e-07, "loss": 0.0294, "step": 5940 }, { "epoch": 0.9594638242894057, "grad_norm": 56.14381790161133, "learning_rate": 4.0536175710594317e-07, "loss": 0.321, "step": 5941 }, { "epoch": 0.959625322997416, "grad_norm": 58.537113189697266, "learning_rate": 4.0374677002583983e-07, "loss": 0.1953, "step": 5942 }, { "epoch": 0.9597868217054264, "grad_norm": 16.33776092529297, "learning_rate": 4.021317829457365e-07, "loss": 0.049, "step": 5943 }, { "epoch": 0.9599483204134367, "grad_norm": 37.489505767822266, "learning_rate": 4.0051679586563306e-07, "loss": 0.1431, "step": 5944 }, { "epoch": 0.960109819121447, "grad_norm": 13.525137901306152, "learning_rate": 3.989018087855297e-07, "loss": 0.0377, "step": 5945 }, { "epoch": 0.9602713178294574, "grad_norm": 22.018136978149414, "learning_rate": 3.972868217054264e-07, "loss": 0.0698, "step": 5946 }, { "epoch": 0.9604328165374677, "grad_norm": 16.30061149597168, "learning_rate": 3.9567183462532306e-07, "loss": 0.027, "step": 5947 }, { "epoch": 0.960594315245478, "grad_norm": 261.9815673828125, "learning_rate": 3.9405684754521967e-07, "loss": 0.6758, "step": 5948 }, { "epoch": 0.9607558139534884, "grad_norm": 6.367702007293701, "learning_rate": 3.9244186046511634e-07, "loss": 0.0154, "step": 5949 }, { "epoch": 0.9609173126614987, "grad_norm": 16.771738052368164, "learning_rate": 3.90826873385013e-07, "loss": 0.0247, "step": 5950 }, { "epoch": 0.961078811369509, "grad_norm": 22.449987411499023, "learning_rate": 3.8921188630490956e-07, "loss": 0.0645, "step": 5951 }, { "epoch": 0.9612403100775194, "grad_norm": 27.06707000732422, "learning_rate": 3.8759689922480623e-07, "loss": 0.1199, "step": 5952 }, { "epoch": 0.9614018087855297, "grad_norm": 9.100019454956055, "learning_rate": 3.859819121447029e-07, "loss": 0.0152, "step": 5953 }, { "epoch": 0.96156330749354, "grad_norm": 127.91116333007812, "learning_rate": 3.843669250645995e-07, "loss": 0.3301, "step": 5954 }, { "epoch": 0.9617248062015504, "grad_norm": 101.90176391601562, "learning_rate": 3.827519379844961e-07, "loss": 0.5056, "step": 5955 }, { "epoch": 0.9618863049095607, "grad_norm": 19.791610717773438, "learning_rate": 3.811369509043928e-07, "loss": 0.0359, "step": 5956 }, { "epoch": 0.962047803617571, "grad_norm": 61.713844299316406, "learning_rate": 3.7952196382428945e-07, "loss": 0.1647, "step": 5957 }, { "epoch": 0.9622093023255814, "grad_norm": 179.87034606933594, "learning_rate": 3.7790697674418606e-07, "loss": 0.6094, "step": 5958 }, { "epoch": 0.9623708010335917, "grad_norm": 10.813329696655273, "learning_rate": 3.7629198966408273e-07, "loss": 0.0331, "step": 5959 }, { "epoch": 0.962532299741602, "grad_norm": 189.31927490234375, "learning_rate": 3.746770025839794e-07, "loss": 1.6731, "step": 5960 }, { "epoch": 0.9626937984496124, "grad_norm": 181.38160705566406, "learning_rate": 3.7306201550387596e-07, "loss": 0.7422, "step": 5961 }, { "epoch": 0.9628552971576227, "grad_norm": 13.749191284179688, "learning_rate": 3.714470284237726e-07, "loss": 0.0185, "step": 5962 }, { "epoch": 0.963016795865633, "grad_norm": 4.739109992980957, "learning_rate": 3.698320413436693e-07, "loss": 0.0108, "step": 5963 }, { "epoch": 0.9631782945736435, "grad_norm": 14.526257514953613, "learning_rate": 3.6821705426356595e-07, "loss": 0.0524, "step": 5964 }, { "epoch": 0.9633397932816538, "grad_norm": 11.838703155517578, "learning_rate": 3.6660206718346257e-07, "loss": 0.0231, "step": 5965 }, { "epoch": 0.963501291989664, "grad_norm": 77.66172790527344, "learning_rate": 3.649870801033592e-07, "loss": 0.3783, "step": 5966 }, { "epoch": 0.9636627906976745, "grad_norm": 28.58131980895996, "learning_rate": 3.6337209302325584e-07, "loss": 0.1073, "step": 5967 }, { "epoch": 0.9638242894056848, "grad_norm": 52.87177276611328, "learning_rate": 3.6175710594315246e-07, "loss": 0.2158, "step": 5968 }, { "epoch": 0.9639857881136951, "grad_norm": 47.76152801513672, "learning_rate": 3.601421188630491e-07, "loss": 0.139, "step": 5969 }, { "epoch": 0.9641472868217055, "grad_norm": 50.92583084106445, "learning_rate": 3.585271317829458e-07, "loss": 0.1133, "step": 5970 }, { "epoch": 0.9643087855297158, "grad_norm": 16.13580322265625, "learning_rate": 3.5691214470284246e-07, "loss": 0.0543, "step": 5971 }, { "epoch": 0.9644702842377261, "grad_norm": 145.45059204101562, "learning_rate": 3.55297157622739e-07, "loss": 0.4603, "step": 5972 }, { "epoch": 0.9646317829457365, "grad_norm": 166.92018127441406, "learning_rate": 3.536821705426357e-07, "loss": 1.3594, "step": 5973 }, { "epoch": 0.9647932816537468, "grad_norm": 70.45912170410156, "learning_rate": 3.5206718346253235e-07, "loss": 0.2219, "step": 5974 }, { "epoch": 0.9649547803617571, "grad_norm": 16.971927642822266, "learning_rate": 3.5045219638242896e-07, "loss": 0.0318, "step": 5975 }, { "epoch": 0.9651162790697675, "grad_norm": 240.7974395751953, "learning_rate": 3.488372093023256e-07, "loss": 0.9395, "step": 5976 }, { "epoch": 0.9652777777777778, "grad_norm": 17.776275634765625, "learning_rate": 3.472222222222223e-07, "loss": 0.0458, "step": 5977 }, { "epoch": 0.9654392764857881, "grad_norm": 118.42121124267578, "learning_rate": 3.4560723514211885e-07, "loss": 0.2627, "step": 5978 }, { "epoch": 0.9656007751937985, "grad_norm": 59.449520111083984, "learning_rate": 3.439922480620155e-07, "loss": 0.1082, "step": 5979 }, { "epoch": 0.9657622739018088, "grad_norm": 225.0883331298828, "learning_rate": 3.423772609819122e-07, "loss": 1.1465, "step": 5980 }, { "epoch": 0.9659237726098191, "grad_norm": 97.1311264038086, "learning_rate": 3.4076227390180885e-07, "loss": 0.7677, "step": 5981 }, { "epoch": 0.9660852713178295, "grad_norm": 163.85484313964844, "learning_rate": 3.391472868217054e-07, "loss": 1.4766, "step": 5982 }, { "epoch": 0.9662467700258398, "grad_norm": 81.18598937988281, "learning_rate": 3.375322997416021e-07, "loss": 0.4453, "step": 5983 }, { "epoch": 0.9664082687338501, "grad_norm": 13.828380584716797, "learning_rate": 3.3591731266149874e-07, "loss": 0.0251, "step": 5984 }, { "epoch": 0.9665697674418605, "grad_norm": 18.900442123413086, "learning_rate": 3.3430232558139535e-07, "loss": 0.0505, "step": 5985 }, { "epoch": 0.9667312661498708, "grad_norm": 18.440109252929688, "learning_rate": 3.32687338501292e-07, "loss": 0.0586, "step": 5986 }, { "epoch": 0.9668927648578811, "grad_norm": 147.3574676513672, "learning_rate": 3.310723514211887e-07, "loss": 0.7051, "step": 5987 }, { "epoch": 0.9670542635658915, "grad_norm": 42.81232452392578, "learning_rate": 3.2945736434108535e-07, "loss": 0.1353, "step": 5988 }, { "epoch": 0.9672157622739018, "grad_norm": 8.933058738708496, "learning_rate": 3.278423772609819e-07, "loss": 0.0404, "step": 5989 }, { "epoch": 0.9673772609819121, "grad_norm": 4.605691909790039, "learning_rate": 3.262273901808786e-07, "loss": 0.0048, "step": 5990 }, { "epoch": 0.9675387596899225, "grad_norm": 81.10779571533203, "learning_rate": 3.2461240310077524e-07, "loss": 0.1836, "step": 5991 }, { "epoch": 0.9677002583979328, "grad_norm": 6.786738872528076, "learning_rate": 3.2299741602067186e-07, "loss": 0.0306, "step": 5992 }, { "epoch": 0.9678617571059431, "grad_norm": 8.859600067138672, "learning_rate": 3.2138242894056847e-07, "loss": 0.0182, "step": 5993 }, { "epoch": 0.9680232558139535, "grad_norm": 13.530834197998047, "learning_rate": 3.1976744186046514e-07, "loss": 0.035, "step": 5994 }, { "epoch": 0.9681847545219638, "grad_norm": 34.444610595703125, "learning_rate": 3.181524547803618e-07, "loss": 0.0579, "step": 5995 }, { "epoch": 0.9683462532299741, "grad_norm": 131.32350158691406, "learning_rate": 3.165374677002584e-07, "loss": 1.0928, "step": 5996 }, { "epoch": 0.9685077519379846, "grad_norm": 141.5932159423828, "learning_rate": 3.149224806201551e-07, "loss": 0.2603, "step": 5997 }, { "epoch": 0.9686692506459949, "grad_norm": 69.53764343261719, "learning_rate": 3.1330749354005175e-07, "loss": 0.8428, "step": 5998 }, { "epoch": 0.9688307493540051, "grad_norm": 66.76309204101562, "learning_rate": 3.1169250645994836e-07, "loss": 0.3906, "step": 5999 }, { "epoch": 0.9689922480620154, "grad_norm": 66.64408111572266, "learning_rate": 3.1007751937984497e-07, "loss": 0.3286, "step": 6000 }, { "epoch": 0.9691537467700259, "grad_norm": 80.536865234375, "learning_rate": 3.0846253229974164e-07, "loss": 0.1456, "step": 6001 }, { "epoch": 0.9693152454780362, "grad_norm": 71.23542022705078, "learning_rate": 3.0684754521963825e-07, "loss": 0.3381, "step": 6002 }, { "epoch": 0.9694767441860465, "grad_norm": 75.1663818359375, "learning_rate": 3.052325581395349e-07, "loss": 0.1839, "step": 6003 }, { "epoch": 0.9696382428940569, "grad_norm": 58.55707550048828, "learning_rate": 3.036175710594316e-07, "loss": 0.0986, "step": 6004 }, { "epoch": 0.9697997416020672, "grad_norm": 30.454853057861328, "learning_rate": 3.020025839793282e-07, "loss": 0.0503, "step": 6005 }, { "epoch": 0.9699612403100775, "grad_norm": 24.14276885986328, "learning_rate": 3.0038759689922486e-07, "loss": 0.083, "step": 6006 }, { "epoch": 0.9701227390180879, "grad_norm": 71.0398941040039, "learning_rate": 2.987726098191215e-07, "loss": 0.5823, "step": 6007 }, { "epoch": 0.9702842377260982, "grad_norm": 19.32884979248047, "learning_rate": 2.971576227390181e-07, "loss": 0.0529, "step": 6008 }, { "epoch": 0.9704457364341085, "grad_norm": 54.57669448852539, "learning_rate": 2.9554263565891475e-07, "loss": 0.1081, "step": 6009 }, { "epoch": 0.9706072351421189, "grad_norm": 158.21835327148438, "learning_rate": 2.9392764857881137e-07, "loss": 1.3572, "step": 6010 }, { "epoch": 0.9707687338501292, "grad_norm": 31.587993621826172, "learning_rate": 2.9231266149870803e-07, "loss": 0.1082, "step": 6011 }, { "epoch": 0.9709302325581395, "grad_norm": 60.3181266784668, "learning_rate": 2.906976744186047e-07, "loss": 0.231, "step": 6012 }, { "epoch": 0.9710917312661499, "grad_norm": 252.97137451171875, "learning_rate": 2.890826873385013e-07, "loss": 1.9884, "step": 6013 }, { "epoch": 0.9712532299741602, "grad_norm": 56.6864128112793, "learning_rate": 2.87467700258398e-07, "loss": 0.1912, "step": 6014 }, { "epoch": 0.9714147286821705, "grad_norm": 49.92947006225586, "learning_rate": 2.858527131782946e-07, "loss": 0.2972, "step": 6015 }, { "epoch": 0.9715762273901809, "grad_norm": 165.99996948242188, "learning_rate": 2.8423772609819125e-07, "loss": 1.0098, "step": 6016 }, { "epoch": 0.9717377260981912, "grad_norm": 80.4512939453125, "learning_rate": 2.8262273901808787e-07, "loss": 0.881, "step": 6017 }, { "epoch": 0.9718992248062015, "grad_norm": 30.6920223236084, "learning_rate": 2.8100775193798453e-07, "loss": 0.0828, "step": 6018 }, { "epoch": 0.9720607235142119, "grad_norm": 24.68262481689453, "learning_rate": 2.7939276485788115e-07, "loss": 0.056, "step": 6019 }, { "epoch": 0.9722222222222222, "grad_norm": 36.20486068725586, "learning_rate": 2.7777777777777776e-07, "loss": 0.0751, "step": 6020 }, { "epoch": 0.9723837209302325, "grad_norm": 45.735992431640625, "learning_rate": 2.761627906976744e-07, "loss": 0.1749, "step": 6021 }, { "epoch": 0.9725452196382429, "grad_norm": 273.4228515625, "learning_rate": 2.745478036175711e-07, "loss": 3.5732, "step": 6022 }, { "epoch": 0.9727067183462532, "grad_norm": 12.538921356201172, "learning_rate": 2.7293281653746776e-07, "loss": 0.0383, "step": 6023 }, { "epoch": 0.9728682170542635, "grad_norm": 13.481090545654297, "learning_rate": 2.7131782945736437e-07, "loss": 0.0252, "step": 6024 }, { "epoch": 0.9730297157622739, "grad_norm": 25.5665283203125, "learning_rate": 2.6970284237726104e-07, "loss": 0.0544, "step": 6025 }, { "epoch": 0.9731912144702842, "grad_norm": 20.65190315246582, "learning_rate": 2.6808785529715765e-07, "loss": 0.0745, "step": 6026 }, { "epoch": 0.9733527131782945, "grad_norm": 110.8478775024414, "learning_rate": 2.6647286821705426e-07, "loss": 0.7524, "step": 6027 }, { "epoch": 0.9735142118863049, "grad_norm": 19.46284294128418, "learning_rate": 2.6485788113695093e-07, "loss": 0.0502, "step": 6028 }, { "epoch": 0.9736757105943152, "grad_norm": 46.6978645324707, "learning_rate": 2.6324289405684754e-07, "loss": 0.0918, "step": 6029 }, { "epoch": 0.9738372093023255, "grad_norm": 13.658269882202148, "learning_rate": 2.616279069767442e-07, "loss": 0.0367, "step": 6030 }, { "epoch": 0.973998708010336, "grad_norm": 33.874088287353516, "learning_rate": 2.6001291989664087e-07, "loss": 0.0983, "step": 6031 }, { "epoch": 0.9741602067183462, "grad_norm": 113.6613998413086, "learning_rate": 2.583979328165375e-07, "loss": 0.326, "step": 6032 }, { "epoch": 0.9743217054263565, "grad_norm": 62.18171310424805, "learning_rate": 2.5678294573643415e-07, "loss": 0.1715, "step": 6033 }, { "epoch": 0.974483204134367, "grad_norm": 33.18284606933594, "learning_rate": 2.5516795865633076e-07, "loss": 0.1566, "step": 6034 }, { "epoch": 0.9746447028423773, "grad_norm": 23.34901237487793, "learning_rate": 2.5355297157622743e-07, "loss": 0.0547, "step": 6035 }, { "epoch": 0.9748062015503876, "grad_norm": 3.601874828338623, "learning_rate": 2.5193798449612404e-07, "loss": 0.0096, "step": 6036 }, { "epoch": 0.974967700258398, "grad_norm": 20.716747283935547, "learning_rate": 2.503229974160207e-07, "loss": 0.0345, "step": 6037 }, { "epoch": 0.9751291989664083, "grad_norm": 35.32859802246094, "learning_rate": 2.487080103359173e-07, "loss": 0.0972, "step": 6038 }, { "epoch": 0.9752906976744186, "grad_norm": 39.365840911865234, "learning_rate": 2.47093023255814e-07, "loss": 0.1597, "step": 6039 }, { "epoch": 0.975452196382429, "grad_norm": 44.06831359863281, "learning_rate": 2.454780361757106e-07, "loss": 0.1245, "step": 6040 }, { "epoch": 0.9756136950904393, "grad_norm": 30.187681198120117, "learning_rate": 2.4386304909560727e-07, "loss": 0.0709, "step": 6041 }, { "epoch": 0.9757751937984496, "grad_norm": 35.96669387817383, "learning_rate": 2.4224806201550393e-07, "loss": 0.147, "step": 6042 }, { "epoch": 0.97593669250646, "grad_norm": 67.02538299560547, "learning_rate": 2.4063307493540054e-07, "loss": 0.186, "step": 6043 }, { "epoch": 0.9760981912144703, "grad_norm": 18.241317749023438, "learning_rate": 2.3901808785529716e-07, "loss": 0.0695, "step": 6044 }, { "epoch": 0.9762596899224806, "grad_norm": 402.9166259765625, "learning_rate": 2.3740310077519382e-07, "loss": 0.7932, "step": 6045 }, { "epoch": 0.976421188630491, "grad_norm": 13.027029037475586, "learning_rate": 2.3578811369509044e-07, "loss": 0.034, "step": 6046 }, { "epoch": 0.9765826873385013, "grad_norm": 36.91437911987305, "learning_rate": 2.341731266149871e-07, "loss": 0.1167, "step": 6047 }, { "epoch": 0.9767441860465116, "grad_norm": 27.035125732421875, "learning_rate": 2.3255813953488374e-07, "loss": 0.056, "step": 6048 }, { "epoch": 0.976905684754522, "grad_norm": 3.2554919719696045, "learning_rate": 2.309431524547804e-07, "loss": 0.0056, "step": 6049 }, { "epoch": 0.9770671834625323, "grad_norm": 187.34359741210938, "learning_rate": 2.2932816537467702e-07, "loss": 0.8627, "step": 6050 }, { "epoch": 0.9772286821705426, "grad_norm": 39.80254364013672, "learning_rate": 2.2771317829457366e-07, "loss": 0.4321, "step": 6051 }, { "epoch": 0.977390180878553, "grad_norm": 33.931636810302734, "learning_rate": 2.260981912144703e-07, "loss": 0.1418, "step": 6052 }, { "epoch": 0.9775516795865633, "grad_norm": 61.71797180175781, "learning_rate": 2.2448320413436694e-07, "loss": 0.2617, "step": 6053 }, { "epoch": 0.9777131782945736, "grad_norm": 25.454374313354492, "learning_rate": 2.228682170542636e-07, "loss": 0.1436, "step": 6054 }, { "epoch": 0.977874677002584, "grad_norm": 21.925527572631836, "learning_rate": 2.2125322997416022e-07, "loss": 0.0445, "step": 6055 }, { "epoch": 0.9780361757105943, "grad_norm": 294.2434997558594, "learning_rate": 2.1963824289405686e-07, "loss": 0.9233, "step": 6056 }, { "epoch": 0.9781976744186046, "grad_norm": 178.77684020996094, "learning_rate": 2.1802325581395352e-07, "loss": 1.0625, "step": 6057 }, { "epoch": 0.978359173126615, "grad_norm": 69.69813537597656, "learning_rate": 2.1640826873385014e-07, "loss": 0.5923, "step": 6058 }, { "epoch": 0.9785206718346253, "grad_norm": 23.052345275878906, "learning_rate": 2.147932816537468e-07, "loss": 0.0583, "step": 6059 }, { "epoch": 0.9786821705426356, "grad_norm": 98.25874328613281, "learning_rate": 2.1317829457364341e-07, "loss": 0.1335, "step": 6060 }, { "epoch": 0.978843669250646, "grad_norm": 28.886390686035156, "learning_rate": 2.1156330749354008e-07, "loss": 0.1376, "step": 6061 }, { "epoch": 0.9790051679586563, "grad_norm": 265.7860107421875, "learning_rate": 2.0994832041343672e-07, "loss": 1.7538, "step": 6062 }, { "epoch": 0.9791666666666666, "grad_norm": 64.69391632080078, "learning_rate": 2.0833333333333333e-07, "loss": 0.3389, "step": 6063 }, { "epoch": 0.979328165374677, "grad_norm": 34.53584289550781, "learning_rate": 2.0671834625323e-07, "loss": 0.2314, "step": 6064 }, { "epoch": 0.9794896640826873, "grad_norm": 42.15909194946289, "learning_rate": 2.0510335917312664e-07, "loss": 0.1807, "step": 6065 }, { "epoch": 0.9796511627906976, "grad_norm": 111.05843353271484, "learning_rate": 2.0348837209302328e-07, "loss": 0.8315, "step": 6066 }, { "epoch": 0.979812661498708, "grad_norm": 125.92597961425781, "learning_rate": 2.0187338501291992e-07, "loss": 1.3446, "step": 6067 }, { "epoch": 0.9799741602067183, "grad_norm": 24.590932846069336, "learning_rate": 2.0025839793281653e-07, "loss": 0.0578, "step": 6068 }, { "epoch": 0.9801356589147286, "grad_norm": 47.58061981201172, "learning_rate": 1.986434108527132e-07, "loss": 0.2141, "step": 6069 }, { "epoch": 0.9802971576227391, "grad_norm": 5.332834243774414, "learning_rate": 1.9702842377260983e-07, "loss": 0.0049, "step": 6070 }, { "epoch": 0.9804586563307494, "grad_norm": 63.11848449707031, "learning_rate": 1.954134366925065e-07, "loss": 0.0958, "step": 6071 }, { "epoch": 0.9806201550387597, "grad_norm": 175.2263641357422, "learning_rate": 1.9379844961240311e-07, "loss": 1.084, "step": 6072 }, { "epoch": 0.9807816537467701, "grad_norm": 65.55241394042969, "learning_rate": 1.9218346253229975e-07, "loss": 0.3848, "step": 6073 }, { "epoch": 0.9809431524547804, "grad_norm": 301.0293273925781, "learning_rate": 1.905684754521964e-07, "loss": 1.3691, "step": 6074 }, { "epoch": 0.9811046511627907, "grad_norm": 59.88888931274414, "learning_rate": 1.8895348837209303e-07, "loss": 0.095, "step": 6075 }, { "epoch": 0.9812661498708011, "grad_norm": 24.229415893554688, "learning_rate": 1.873385012919897e-07, "loss": 0.1172, "step": 6076 }, { "epoch": 0.9814276485788114, "grad_norm": 5.093415260314941, "learning_rate": 1.857235142118863e-07, "loss": 0.0106, "step": 6077 }, { "epoch": 0.9815891472868217, "grad_norm": 3.9818248748779297, "learning_rate": 1.8410852713178298e-07, "loss": 0.0052, "step": 6078 }, { "epoch": 0.9817506459948321, "grad_norm": 7.6025471687316895, "learning_rate": 1.824935400516796e-07, "loss": 0.0223, "step": 6079 }, { "epoch": 0.9819121447028424, "grad_norm": 8.02210521697998, "learning_rate": 1.8087855297157623e-07, "loss": 0.0151, "step": 6080 }, { "epoch": 0.9820736434108527, "grad_norm": 100.23951721191406, "learning_rate": 1.792635658914729e-07, "loss": 0.2821, "step": 6081 }, { "epoch": 0.9822351421188631, "grad_norm": 28.14502716064453, "learning_rate": 1.776485788113695e-07, "loss": 0.1015, "step": 6082 }, { "epoch": 0.9823966408268734, "grad_norm": 54.18645477294922, "learning_rate": 1.7603359173126617e-07, "loss": 0.1424, "step": 6083 }, { "epoch": 0.9825581395348837, "grad_norm": 15.15038013458252, "learning_rate": 1.744186046511628e-07, "loss": 0.0187, "step": 6084 }, { "epoch": 0.9827196382428941, "grad_norm": 92.88373565673828, "learning_rate": 1.7280361757105943e-07, "loss": 0.8303, "step": 6085 }, { "epoch": 0.9828811369509044, "grad_norm": 70.06646728515625, "learning_rate": 1.711886304909561e-07, "loss": 0.3063, "step": 6086 }, { "epoch": 0.9830426356589147, "grad_norm": 50.002174377441406, "learning_rate": 1.695736434108527e-07, "loss": 0.1685, "step": 6087 }, { "epoch": 0.9832041343669251, "grad_norm": 195.73631286621094, "learning_rate": 1.6795865633074937e-07, "loss": 1.5371, "step": 6088 }, { "epoch": 0.9833656330749354, "grad_norm": 211.422607421875, "learning_rate": 1.66343669250646e-07, "loss": 1.1727, "step": 6089 }, { "epoch": 0.9835271317829457, "grad_norm": 107.04815673828125, "learning_rate": 1.6472868217054268e-07, "loss": 0.3933, "step": 6090 }, { "epoch": 0.9836886304909561, "grad_norm": 206.78659057617188, "learning_rate": 1.631136950904393e-07, "loss": 1.1052, "step": 6091 }, { "epoch": 0.9838501291989664, "grad_norm": 90.01934814453125, "learning_rate": 1.6149870801033593e-07, "loss": 0.3035, "step": 6092 }, { "epoch": 0.9840116279069767, "grad_norm": 11.465340614318848, "learning_rate": 1.5988372093023257e-07, "loss": 0.0402, "step": 6093 }, { "epoch": 0.9841731266149871, "grad_norm": 24.45577621459961, "learning_rate": 1.582687338501292e-07, "loss": 0.0722, "step": 6094 }, { "epoch": 0.9843346253229974, "grad_norm": 72.83013153076172, "learning_rate": 1.5665374677002587e-07, "loss": 0.2798, "step": 6095 }, { "epoch": 0.9844961240310077, "grad_norm": 52.44400405883789, "learning_rate": 1.5503875968992249e-07, "loss": 0.1586, "step": 6096 }, { "epoch": 0.9846576227390181, "grad_norm": 44.1508674621582, "learning_rate": 1.5342377260981913e-07, "loss": 0.2746, "step": 6097 }, { "epoch": 0.9848191214470284, "grad_norm": 26.840829849243164, "learning_rate": 1.518087855297158e-07, "loss": 0.0957, "step": 6098 }, { "epoch": 0.9849806201550387, "grad_norm": 20.31291961669922, "learning_rate": 1.5019379844961243e-07, "loss": 0.0562, "step": 6099 }, { "epoch": 0.9851421188630491, "grad_norm": 4.967879295349121, "learning_rate": 1.4857881136950904e-07, "loss": 0.0109, "step": 6100 }, { "epoch": 0.9853036175710594, "grad_norm": 132.1717529296875, "learning_rate": 1.4696382428940568e-07, "loss": 0.4873, "step": 6101 }, { "epoch": 0.9854651162790697, "grad_norm": 31.378807067871094, "learning_rate": 1.4534883720930235e-07, "loss": 0.1205, "step": 6102 }, { "epoch": 0.9856266149870802, "grad_norm": 30.17380142211914, "learning_rate": 1.43733850129199e-07, "loss": 0.1058, "step": 6103 }, { "epoch": 0.9857881136950905, "grad_norm": 18.349069595336914, "learning_rate": 1.4211886304909563e-07, "loss": 0.1091, "step": 6104 }, { "epoch": 0.9859496124031008, "grad_norm": 152.4226531982422, "learning_rate": 1.4050387596899227e-07, "loss": 0.2168, "step": 6105 }, { "epoch": 0.9861111111111112, "grad_norm": 615.1142578125, "learning_rate": 1.3888888888888888e-07, "loss": 3.1588, "step": 6106 }, { "epoch": 0.9862726098191215, "grad_norm": 13.081950187683105, "learning_rate": 1.3727390180878555e-07, "loss": 0.0391, "step": 6107 }, { "epoch": 0.9864341085271318, "grad_norm": 105.10308074951172, "learning_rate": 1.3565891472868218e-07, "loss": 1.2861, "step": 6108 }, { "epoch": 0.9865956072351422, "grad_norm": 32.74763107299805, "learning_rate": 1.3404392764857882e-07, "loss": 0.1641, "step": 6109 }, { "epoch": 0.9867571059431525, "grad_norm": 50.808597564697266, "learning_rate": 1.3242894056847546e-07, "loss": 0.0707, "step": 6110 }, { "epoch": 0.9869186046511628, "grad_norm": 66.56394958496094, "learning_rate": 1.308139534883721e-07, "loss": 0.1914, "step": 6111 }, { "epoch": 0.9870801033591732, "grad_norm": 32.812049865722656, "learning_rate": 1.2919896640826874e-07, "loss": 0.1611, "step": 6112 }, { "epoch": 0.9872416020671835, "grad_norm": 139.99351501464844, "learning_rate": 1.2758397932816538e-07, "loss": 0.4468, "step": 6113 }, { "epoch": 0.9874031007751938, "grad_norm": 189.92471313476562, "learning_rate": 1.2596899224806202e-07, "loss": 0.2765, "step": 6114 }, { "epoch": 0.9875645994832042, "grad_norm": 19.926427841186523, "learning_rate": 1.2435400516795866e-07, "loss": 0.0681, "step": 6115 }, { "epoch": 0.9877260981912145, "grad_norm": 56.77025604248047, "learning_rate": 1.227390180878553e-07, "loss": 0.3079, "step": 6116 }, { "epoch": 0.9878875968992248, "grad_norm": 163.92141723632812, "learning_rate": 1.2112403100775197e-07, "loss": 0.7149, "step": 6117 }, { "epoch": 0.9880490956072352, "grad_norm": 68.1237564086914, "learning_rate": 1.1950904392764858e-07, "loss": 0.2108, "step": 6118 }, { "epoch": 0.9882105943152455, "grad_norm": 43.72234344482422, "learning_rate": 1.1789405684754522e-07, "loss": 0.078, "step": 6119 }, { "epoch": 0.9883720930232558, "grad_norm": 87.38377380371094, "learning_rate": 1.1627906976744187e-07, "loss": 0.4512, "step": 6120 }, { "epoch": 0.9885335917312662, "grad_norm": 14.73243236541748, "learning_rate": 1.1466408268733851e-07, "loss": 0.0452, "step": 6121 }, { "epoch": 0.9886950904392765, "grad_norm": 332.3315734863281, "learning_rate": 1.1304909560723515e-07, "loss": 1.6196, "step": 6122 }, { "epoch": 0.9888565891472868, "grad_norm": 114.82734680175781, "learning_rate": 1.114341085271318e-07, "loss": 0.54, "step": 6123 }, { "epoch": 0.9890180878552972, "grad_norm": 20.10335922241211, "learning_rate": 1.0981912144702843e-07, "loss": 0.0385, "step": 6124 }, { "epoch": 0.9891795865633075, "grad_norm": 10.591322898864746, "learning_rate": 1.0820413436692507e-07, "loss": 0.025, "step": 6125 }, { "epoch": 0.9893410852713178, "grad_norm": 2.44689679145813, "learning_rate": 1.0658914728682171e-07, "loss": 0.0104, "step": 6126 }, { "epoch": 0.9895025839793282, "grad_norm": 55.400962829589844, "learning_rate": 1.0497416020671836e-07, "loss": 0.1012, "step": 6127 }, { "epoch": 0.9896640826873385, "grad_norm": 56.7547721862793, "learning_rate": 1.03359173126615e-07, "loss": 0.1456, "step": 6128 }, { "epoch": 0.9898255813953488, "grad_norm": 15.37308406829834, "learning_rate": 1.0174418604651164e-07, "loss": 0.0444, "step": 6129 }, { "epoch": 0.9899870801033591, "grad_norm": 24.38033103942871, "learning_rate": 1.0012919896640826e-07, "loss": 0.0488, "step": 6130 }, { "epoch": 0.9901485788113695, "grad_norm": 15.484797477722168, "learning_rate": 9.851421188630492e-08, "loss": 0.0491, "step": 6131 }, { "epoch": 0.9903100775193798, "grad_norm": 4.475937843322754, "learning_rate": 9.689922480620156e-08, "loss": 0.009, "step": 6132 }, { "epoch": 0.9904715762273901, "grad_norm": 43.04434585571289, "learning_rate": 9.52842377260982e-08, "loss": 0.1347, "step": 6133 }, { "epoch": 0.9906330749354005, "grad_norm": 65.20465087890625, "learning_rate": 9.366925064599485e-08, "loss": 0.3247, "step": 6134 }, { "epoch": 0.9907945736434108, "grad_norm": 213.53363037109375, "learning_rate": 9.205426356589149e-08, "loss": 0.3497, "step": 6135 }, { "epoch": 0.9909560723514211, "grad_norm": 24.973560333251953, "learning_rate": 9.043927648578811e-08, "loss": 0.0869, "step": 6136 }, { "epoch": 0.9911175710594315, "grad_norm": 78.48967742919922, "learning_rate": 8.882428940568475e-08, "loss": 0.3079, "step": 6137 }, { "epoch": 0.9912790697674418, "grad_norm": 33.55525588989258, "learning_rate": 8.72093023255814e-08, "loss": 0.0459, "step": 6138 }, { "epoch": 0.9914405684754521, "grad_norm": 5.055386066436768, "learning_rate": 8.559431524547805e-08, "loss": 0.0112, "step": 6139 }, { "epoch": 0.9916020671834626, "grad_norm": 11.181694030761719, "learning_rate": 8.397932816537469e-08, "loss": 0.0493, "step": 6140 }, { "epoch": 0.9917635658914729, "grad_norm": 8.049049377441406, "learning_rate": 8.236434108527134e-08, "loss": 0.0277, "step": 6141 }, { "epoch": 0.9919250645994832, "grad_norm": 77.82860565185547, "learning_rate": 8.074935400516796e-08, "loss": 0.5059, "step": 6142 }, { "epoch": 0.9920865633074936, "grad_norm": 97.01065063476562, "learning_rate": 7.91343669250646e-08, "loss": 0.9307, "step": 6143 }, { "epoch": 0.9922480620155039, "grad_norm": 40.52311325073242, "learning_rate": 7.751937984496124e-08, "loss": 0.1217, "step": 6144 }, { "epoch": 0.9924095607235142, "grad_norm": 107.60467529296875, "learning_rate": 7.59043927648579e-08, "loss": 0.7393, "step": 6145 }, { "epoch": 0.9925710594315246, "grad_norm": 144.07559204101562, "learning_rate": 7.428940568475452e-08, "loss": 0.707, "step": 6146 }, { "epoch": 0.9927325581395349, "grad_norm": 3.017810344696045, "learning_rate": 7.267441860465117e-08, "loss": 0.0073, "step": 6147 }, { "epoch": 0.9928940568475452, "grad_norm": 21.350479125976562, "learning_rate": 7.105943152454781e-08, "loss": 0.1057, "step": 6148 }, { "epoch": 0.9930555555555556, "grad_norm": 94.33572387695312, "learning_rate": 6.944444444444444e-08, "loss": 0.3603, "step": 6149 }, { "epoch": 0.9932170542635659, "grad_norm": 14.886467933654785, "learning_rate": 6.782945736434109e-08, "loss": 0.0374, "step": 6150 }, { "epoch": 0.9933785529715762, "grad_norm": 55.838523864746094, "learning_rate": 6.621447028423773e-08, "loss": 0.3496, "step": 6151 }, { "epoch": 0.9935400516795866, "grad_norm": 12.420098304748535, "learning_rate": 6.459948320413437e-08, "loss": 0.0197, "step": 6152 }, { "epoch": 0.9937015503875969, "grad_norm": 24.905553817749023, "learning_rate": 6.298449612403101e-08, "loss": 0.118, "step": 6153 }, { "epoch": 0.9938630490956072, "grad_norm": 50.43333435058594, "learning_rate": 6.136950904392765e-08, "loss": 0.2786, "step": 6154 }, { "epoch": 0.9940245478036176, "grad_norm": 11.85419750213623, "learning_rate": 5.975452196382429e-08, "loss": 0.0252, "step": 6155 }, { "epoch": 0.9941860465116279, "grad_norm": 31.68519401550293, "learning_rate": 5.8139534883720935e-08, "loss": 0.0506, "step": 6156 }, { "epoch": 0.9943475452196382, "grad_norm": 26.75107192993164, "learning_rate": 5.6524547803617575e-08, "loss": 0.0607, "step": 6157 }, { "epoch": 0.9945090439276486, "grad_norm": 190.4609375, "learning_rate": 5.4909560723514214e-08, "loss": 0.6611, "step": 6158 }, { "epoch": 0.9946705426356589, "grad_norm": 15.154743194580078, "learning_rate": 5.3294573643410854e-08, "loss": 0.0383, "step": 6159 }, { "epoch": 0.9948320413436692, "grad_norm": 24.97726058959961, "learning_rate": 5.16795865633075e-08, "loss": 0.0969, "step": 6160 }, { "epoch": 0.9949935400516796, "grad_norm": 9.467305183410645, "learning_rate": 5.006459948320413e-08, "loss": 0.032, "step": 6161 }, { "epoch": 0.9951550387596899, "grad_norm": 44.3709602355957, "learning_rate": 4.844961240310078e-08, "loss": 0.2687, "step": 6162 }, { "epoch": 0.9953165374677002, "grad_norm": 22.7031192779541, "learning_rate": 4.6834625322997424e-08, "loss": 0.0583, "step": 6163 }, { "epoch": 0.9954780361757106, "grad_norm": 323.58184814453125, "learning_rate": 4.521963824289406e-08, "loss": 1.4033, "step": 6164 }, { "epoch": 0.9956395348837209, "grad_norm": 30.811851501464844, "learning_rate": 4.36046511627907e-08, "loss": 0.0471, "step": 6165 }, { "epoch": 0.9958010335917312, "grad_norm": 21.7785701751709, "learning_rate": 4.198966408268734e-08, "loss": 0.0808, "step": 6166 }, { "epoch": 0.9959625322997416, "grad_norm": 26.37322235107422, "learning_rate": 4.037467700258398e-08, "loss": 0.0708, "step": 6167 }, { "epoch": 0.9961240310077519, "grad_norm": 34.58024597167969, "learning_rate": 3.875968992248062e-08, "loss": 0.1028, "step": 6168 }, { "epoch": 0.9962855297157622, "grad_norm": 8.340499877929688, "learning_rate": 3.714470284237726e-08, "loss": 0.0127, "step": 6169 }, { "epoch": 0.9964470284237726, "grad_norm": 31.56998634338379, "learning_rate": 3.552971576227391e-08, "loss": 0.0728, "step": 6170 }, { "epoch": 0.9966085271317829, "grad_norm": 37.289756774902344, "learning_rate": 3.3914728682170546e-08, "loss": 0.1149, "step": 6171 }, { "epoch": 0.9967700258397932, "grad_norm": 89.28128814697266, "learning_rate": 3.2299741602067186e-08, "loss": 0.5635, "step": 6172 }, { "epoch": 0.9969315245478036, "grad_norm": 163.85816955566406, "learning_rate": 3.0684754521963825e-08, "loss": 1.7153, "step": 6173 }, { "epoch": 0.997093023255814, "grad_norm": 11.887969970703125, "learning_rate": 2.9069767441860468e-08, "loss": 0.0342, "step": 6174 }, { "epoch": 0.9972545219638242, "grad_norm": 92.30889892578125, "learning_rate": 2.7454780361757107e-08, "loss": 0.417, "step": 6175 }, { "epoch": 0.9974160206718347, "grad_norm": 25.6446475982666, "learning_rate": 2.583979328165375e-08, "loss": 0.1184, "step": 6176 }, { "epoch": 0.997577519379845, "grad_norm": 66.2801284790039, "learning_rate": 2.422480620155039e-08, "loss": 0.1594, "step": 6177 }, { "epoch": 0.9977390180878553, "grad_norm": 7.203554630279541, "learning_rate": 2.260981912144703e-08, "loss": 0.0253, "step": 6178 }, { "epoch": 0.9979005167958657, "grad_norm": 162.26394653320312, "learning_rate": 2.099483204134367e-08, "loss": 0.7177, "step": 6179 }, { "epoch": 0.998062015503876, "grad_norm": 135.93544006347656, "learning_rate": 1.937984496124031e-08, "loss": 0.3242, "step": 6180 }, { "epoch": 0.9982235142118863, "grad_norm": 66.27774810791016, "learning_rate": 1.7764857881136953e-08, "loss": 0.1218, "step": 6181 }, { "epoch": 0.9983850129198967, "grad_norm": 21.901283264160156, "learning_rate": 1.6149870801033593e-08, "loss": 0.0659, "step": 6182 }, { "epoch": 0.998546511627907, "grad_norm": 93.11935424804688, "learning_rate": 1.4534883720930234e-08, "loss": 0.2312, "step": 6183 }, { "epoch": 0.9987080103359173, "grad_norm": 36.21965408325195, "learning_rate": 1.2919896640826875e-08, "loss": 0.124, "step": 6184 }, { "epoch": 0.9988695090439277, "grad_norm": 13.620026588439941, "learning_rate": 1.1304909560723514e-08, "loss": 0.0204, "step": 6185 }, { "epoch": 0.999031007751938, "grad_norm": 2.9136569499969482, "learning_rate": 9.689922480620155e-09, "loss": 0.0044, "step": 6186 }, { "epoch": 0.9991925064599483, "grad_norm": 36.06045913696289, "learning_rate": 8.074935400516796e-09, "loss": 0.1724, "step": 6187 }, { "epoch": 0.9993540051679587, "grad_norm": 116.08317565917969, "learning_rate": 6.4599483204134375e-09, "loss": 0.4273, "step": 6188 }, { "epoch": 0.999515503875969, "grad_norm": 18.431859970092773, "learning_rate": 4.844961240310078e-09, "loss": 0.0515, "step": 6189 }, { "epoch": 0.9996770025839793, "grad_norm": 79.60527038574219, "learning_rate": 3.2299741602067187e-09, "loss": 0.6063, "step": 6190 }, { "epoch": 0.9998385012919897, "grad_norm": 35.0000114440918, "learning_rate": 1.6149870801033594e-09, "loss": 0.1494, "step": 6191 }, { "epoch": 1.0, "grad_norm": 74.45528411865234, "learning_rate": 0.0, "loss": 0.1729, "step": 6192 } ], "logging_steps": 1, "max_steps": 6192, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 1548, "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 }