{ "best_metric": null, "best_model_checkpoint": null, "epoch": 100.0, "eval_steps": 200.0, "global_step": 1700, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.058823529411764705, "grad_norm": 43.639549255371094, "learning_rate": 3.921568627450981e-07, "loss": 6.4004, "step": 1 }, { "epoch": 0.11764705882352941, "grad_norm": 36.49709701538086, "learning_rate": 7.843137254901962e-07, "loss": 6.248, "step": 2 }, { "epoch": 0.17647058823529413, "grad_norm": 55.4442138671875, "learning_rate": 1.1764705882352942e-06, "loss": 6.3535, "step": 3 }, { "epoch": 0.23529411764705882, "grad_norm": 40.5142707824707, "learning_rate": 1.5686274509803923e-06, "loss": 6.5273, "step": 4 }, { "epoch": 0.29411764705882354, "grad_norm": 90.17640686035156, "learning_rate": 1.96078431372549e-06, "loss": 6.4883, "step": 5 }, { "epoch": 0.35294117647058826, "grad_norm": 51.15266036987305, "learning_rate": 2.3529411764705885e-06, "loss": 6.6074, "step": 6 }, { "epoch": 0.4117647058823529, "grad_norm": 40.98651885986328, "learning_rate": 2.7450980392156867e-06, "loss": 6.3516, "step": 7 }, { "epoch": 0.47058823529411764, "grad_norm": 52.092647552490234, "learning_rate": 3.1372549019607846e-06, "loss": 6.2383, "step": 8 }, { "epoch": 0.5294117647058824, "grad_norm": 34.8509635925293, "learning_rate": 3.529411764705883e-06, "loss": 6.1133, "step": 9 }, { "epoch": 0.5882352941176471, "grad_norm": 52.24551010131836, "learning_rate": 3.92156862745098e-06, "loss": 6.0977, "step": 10 }, { "epoch": 0.6470588235294118, "grad_norm": 33.72833251953125, "learning_rate": 4.313725490196079e-06, "loss": 5.5703, "step": 11 }, { "epoch": 0.7058823529411765, "grad_norm": 47.9254264831543, "learning_rate": 4.705882352941177e-06, "loss": 5.4883, "step": 12 }, { "epoch": 0.7647058823529411, "grad_norm": 41.007904052734375, "learning_rate": 5.098039215686274e-06, "loss": 5.3359, "step": 13 }, { "epoch": 0.8235294117647058, "grad_norm": 33.39987564086914, "learning_rate": 5.4901960784313735e-06, "loss": 5.3086, "step": 14 }, { "epoch": 0.8823529411764706, "grad_norm": 28.169689178466797, "learning_rate": 5.882352941176471e-06, "loss": 4.8301, "step": 15 }, { "epoch": 0.9411764705882353, "grad_norm": 37.265716552734375, "learning_rate": 6.274509803921569e-06, "loss": 4.792, "step": 16 }, { "epoch": 1.0, "grad_norm": 23.017866134643555, "learning_rate": 6.666666666666667e-06, "loss": 4.9023, "step": 17 }, { "epoch": 1.0588235294117647, "grad_norm": 24.191726684570312, "learning_rate": 7.058823529411766e-06, "loss": 4.4531, "step": 18 }, { "epoch": 1.1176470588235294, "grad_norm": 18.19198989868164, "learning_rate": 7.450980392156863e-06, "loss": 4.3857, "step": 19 }, { "epoch": 1.1764705882352942, "grad_norm": 16.217632293701172, "learning_rate": 7.84313725490196e-06, "loss": 4.0947, "step": 20 }, { "epoch": 1.2352941176470589, "grad_norm": 21.044132232666016, "learning_rate": 8.23529411764706e-06, "loss": 4.29, "step": 21 }, { "epoch": 1.2941176470588236, "grad_norm": 14.460254669189453, "learning_rate": 8.627450980392157e-06, "loss": 4.1104, "step": 22 }, { "epoch": 1.3529411764705883, "grad_norm": 43.94008255004883, "learning_rate": 9.019607843137256e-06, "loss": 3.9961, "step": 23 }, { "epoch": 1.4117647058823528, "grad_norm": 21.906930923461914, "learning_rate": 9.411764705882354e-06, "loss": 4.0176, "step": 24 }, { "epoch": 1.4705882352941178, "grad_norm": 10.889321327209473, "learning_rate": 9.803921568627451e-06, "loss": 3.7256, "step": 25 }, { "epoch": 1.5294117647058822, "grad_norm": 12.525795936584473, "learning_rate": 1.0196078431372549e-05, "loss": 3.8711, "step": 26 }, { "epoch": 1.5882352941176472, "grad_norm": 14.274972915649414, "learning_rate": 1.0588235294117648e-05, "loss": 3.7842, "step": 27 }, { "epoch": 1.6470588235294117, "grad_norm": 16.792036056518555, "learning_rate": 1.0980392156862747e-05, "loss": 3.6436, "step": 28 }, { "epoch": 1.7058823529411766, "grad_norm": 10.332293510437012, "learning_rate": 1.1372549019607844e-05, "loss": 3.5625, "step": 29 }, { "epoch": 1.7647058823529411, "grad_norm": 16.79983139038086, "learning_rate": 1.1764705882352942e-05, "loss": 3.4717, "step": 30 }, { "epoch": 1.8235294117647058, "grad_norm": 17.554670333862305, "learning_rate": 1.215686274509804e-05, "loss": 3.4209, "step": 31 }, { "epoch": 1.8823529411764706, "grad_norm": 25.51144027709961, "learning_rate": 1.2549019607843138e-05, "loss": 3.6309, "step": 32 }, { "epoch": 1.9411764705882353, "grad_norm": 75.74697875976562, "learning_rate": 1.2941176470588238e-05, "loss": 3.418, "step": 33 }, { "epoch": 2.0, "grad_norm": 51.40288162231445, "learning_rate": 1.3333333333333333e-05, "loss": 3.3623, "step": 34 }, { "epoch": 2.0588235294117645, "grad_norm": 15.71206283569336, "learning_rate": 1.3725490196078432e-05, "loss": 3.3682, "step": 35 }, { "epoch": 2.1176470588235294, "grad_norm": 27.537328720092773, "learning_rate": 1.4117647058823532e-05, "loss": 3.3906, "step": 36 }, { "epoch": 2.176470588235294, "grad_norm": 19.966129302978516, "learning_rate": 1.4509803921568629e-05, "loss": 3.3281, "step": 37 }, { "epoch": 2.235294117647059, "grad_norm": 21.453075408935547, "learning_rate": 1.4901960784313726e-05, "loss": 3.375, "step": 38 }, { "epoch": 2.2941176470588234, "grad_norm": 40.88306427001953, "learning_rate": 1.5294117647058822e-05, "loss": 3.292, "step": 39 }, { "epoch": 2.3529411764705883, "grad_norm": 82.29407501220703, "learning_rate": 1.568627450980392e-05, "loss": 3.3311, "step": 40 }, { "epoch": 2.411764705882353, "grad_norm": 15.576737403869629, "learning_rate": 1.607843137254902e-05, "loss": 3.167, "step": 41 }, { "epoch": 2.4705882352941178, "grad_norm": 20.998512268066406, "learning_rate": 1.647058823529412e-05, "loss": 3.2744, "step": 42 }, { "epoch": 2.5294117647058822, "grad_norm": 24.492145538330078, "learning_rate": 1.686274509803922e-05, "loss": 3.126, "step": 43 }, { "epoch": 2.588235294117647, "grad_norm": 14.534441947937012, "learning_rate": 1.7254901960784314e-05, "loss": 3.1475, "step": 44 }, { "epoch": 2.6470588235294117, "grad_norm": 16.832632064819336, "learning_rate": 1.7647058823529414e-05, "loss": 3.2148, "step": 45 }, { "epoch": 2.7058823529411766, "grad_norm": 19.044008255004883, "learning_rate": 1.8039215686274513e-05, "loss": 3.249, "step": 46 }, { "epoch": 2.764705882352941, "grad_norm": 23.00897979736328, "learning_rate": 1.843137254901961e-05, "loss": 2.9961, "step": 47 }, { "epoch": 2.8235294117647056, "grad_norm": 14.287858963012695, "learning_rate": 1.8823529411764708e-05, "loss": 3.2646, "step": 48 }, { "epoch": 2.8823529411764706, "grad_norm": 10.992258071899414, "learning_rate": 1.9215686274509807e-05, "loss": 3.1191, "step": 49 }, { "epoch": 2.9411764705882355, "grad_norm": 15.143692970275879, "learning_rate": 1.9607843137254903e-05, "loss": 3.1514, "step": 50 }, { "epoch": 3.0, "grad_norm": 16.56675910949707, "learning_rate": 2e-05, "loss": 3.0977, "step": 51 }, { "epoch": 3.0588235294117645, "grad_norm": 9.419181823730469, "learning_rate": 1.9999981852019372e-05, "loss": 3.1914, "step": 52 }, { "epoch": 3.1176470588235294, "grad_norm": 13.587262153625488, "learning_rate": 1.9999927408143346e-05, "loss": 3.0107, "step": 53 }, { "epoch": 3.176470588235294, "grad_norm": 10.331709861755371, "learning_rate": 1.9999836668569542e-05, "loss": 2.9424, "step": 54 }, { "epoch": 3.235294117647059, "grad_norm": 23.011505126953125, "learning_rate": 1.9999709633627305e-05, "loss": 3.0273, "step": 55 }, { "epoch": 3.2941176470588234, "grad_norm": 25.459199905395508, "learning_rate": 1.9999546303777714e-05, "loss": 2.9492, "step": 56 }, { "epoch": 3.3529411764705883, "grad_norm": 22.28142738342285, "learning_rate": 1.9999346679613598e-05, "loss": 2.9268, "step": 57 }, { "epoch": 3.411764705882353, "grad_norm": 12.388641357421875, "learning_rate": 1.999911076185951e-05, "loss": 2.9961, "step": 58 }, { "epoch": 3.4705882352941178, "grad_norm": 12.241747856140137, "learning_rate": 1.9998838551371733e-05, "loss": 2.8477, "step": 59 }, { "epoch": 3.5294117647058822, "grad_norm": 9.25222396850586, "learning_rate": 1.9998530049138286e-05, "loss": 2.9873, "step": 60 }, { "epoch": 3.588235294117647, "grad_norm": 10.952873229980469, "learning_rate": 1.9998185256278905e-05, "loss": 3.0303, "step": 61 }, { "epoch": 3.6470588235294117, "grad_norm": 7.141952037811279, "learning_rate": 1.999780417404505e-05, "loss": 2.9053, "step": 62 }, { "epoch": 3.7058823529411766, "grad_norm": 8.604562759399414, "learning_rate": 1.9997386803819892e-05, "loss": 3.0557, "step": 63 }, { "epoch": 3.764705882352941, "grad_norm": 20.612958908081055, "learning_rate": 1.9996933147118322e-05, "loss": 2.8672, "step": 64 }, { "epoch": 3.8235294117647056, "grad_norm": 6.484840393066406, "learning_rate": 1.9996443205586926e-05, "loss": 2.8359, "step": 65 }, { "epoch": 3.8823529411764706, "grad_norm": 10.652170181274414, "learning_rate": 1.9995916981003995e-05, "loss": 2.957, "step": 66 }, { "epoch": 3.9411764705882355, "grad_norm": 12.776596069335938, "learning_rate": 1.9995354475279517e-05, "loss": 2.9785, "step": 67 }, { "epoch": 4.0, "grad_norm": 9.438146591186523, "learning_rate": 1.9994755690455154e-05, "loss": 2.9287, "step": 68 }, { "epoch": 4.0588235294117645, "grad_norm": 9.608416557312012, "learning_rate": 1.999412062870425e-05, "loss": 2.8223, "step": 69 }, { "epoch": 4.117647058823529, "grad_norm": 8.964274406433105, "learning_rate": 1.999344929233183e-05, "loss": 2.7734, "step": 70 }, { "epoch": 4.176470588235294, "grad_norm": 10.980924606323242, "learning_rate": 1.999274168377457e-05, "loss": 2.7324, "step": 71 }, { "epoch": 4.235294117647059, "grad_norm": 9.79592227935791, "learning_rate": 1.999199780560081e-05, "loss": 2.8359, "step": 72 }, { "epoch": 4.294117647058823, "grad_norm": 7.801930904388428, "learning_rate": 1.9991217660510518e-05, "loss": 2.7275, "step": 73 }, { "epoch": 4.352941176470588, "grad_norm": 9.107315063476562, "learning_rate": 1.9990401251335313e-05, "loss": 2.833, "step": 74 }, { "epoch": 4.411764705882353, "grad_norm": 9.951732635498047, "learning_rate": 1.9989548581038425e-05, "loss": 2.6797, "step": 75 }, { "epoch": 4.470588235294118, "grad_norm": 8.400995254516602, "learning_rate": 1.9988659652714704e-05, "loss": 2.8604, "step": 76 }, { "epoch": 4.529411764705882, "grad_norm": 7.7693047523498535, "learning_rate": 1.9987734469590605e-05, "loss": 2.7773, "step": 77 }, { "epoch": 4.588235294117647, "grad_norm": 9.412419319152832, "learning_rate": 1.9986773035024164e-05, "loss": 2.7451, "step": 78 }, { "epoch": 4.647058823529412, "grad_norm": 8.995473861694336, "learning_rate": 1.9985775352505e-05, "loss": 2.5996, "step": 79 }, { "epoch": 4.705882352941177, "grad_norm": 7.539700984954834, "learning_rate": 1.9984741425654308e-05, "loss": 2.7227, "step": 80 }, { "epoch": 4.764705882352941, "grad_norm": 7.847495079040527, "learning_rate": 1.9983671258224808e-05, "loss": 2.6602, "step": 81 }, { "epoch": 4.823529411764706, "grad_norm": 8.1257963180542, "learning_rate": 1.9982564854100786e-05, "loss": 2.6885, "step": 82 }, { "epoch": 4.882352941176471, "grad_norm": 7.572000503540039, "learning_rate": 1.998142221729804e-05, "loss": 2.7471, "step": 83 }, { "epoch": 4.9411764705882355, "grad_norm": 9.618082046508789, "learning_rate": 1.9980243351963887e-05, "loss": 2.791, "step": 84 }, { "epoch": 5.0, "grad_norm": 19.0474796295166, "learning_rate": 1.997902826237712e-05, "loss": 2.6582, "step": 85 }, { "epoch": 5.0588235294117645, "grad_norm": 11.013533592224121, "learning_rate": 1.997777695294803e-05, "loss": 2.5728, "step": 86 }, { "epoch": 5.117647058823529, "grad_norm": 13.179403305053711, "learning_rate": 1.997648942821836e-05, "loss": 2.7881, "step": 87 }, { "epoch": 5.176470588235294, "grad_norm": 9.207347869873047, "learning_rate": 1.9975165692861315e-05, "loss": 2.5684, "step": 88 }, { "epoch": 5.235294117647059, "grad_norm": 8.816306114196777, "learning_rate": 1.997380575168151e-05, "loss": 2.6992, "step": 89 }, { "epoch": 5.294117647058823, "grad_norm": 10.574613571166992, "learning_rate": 1.9972409609614986e-05, "loss": 2.5063, "step": 90 }, { "epoch": 5.352941176470588, "grad_norm": 8.78278636932373, "learning_rate": 1.9970977271729172e-05, "loss": 2.6348, "step": 91 }, { "epoch": 5.411764705882353, "grad_norm": 8.737688064575195, "learning_rate": 1.996950874322288e-05, "loss": 2.5508, "step": 92 }, { "epoch": 5.470588235294118, "grad_norm": 10.0756196975708, "learning_rate": 1.9968004029426277e-05, "loss": 2.5508, "step": 93 }, { "epoch": 5.529411764705882, "grad_norm": 8.524308204650879, "learning_rate": 1.996646313580086e-05, "loss": 2.5791, "step": 94 }, { "epoch": 5.588235294117647, "grad_norm": 12.276015281677246, "learning_rate": 1.9964886067939453e-05, "loss": 2.5264, "step": 95 }, { "epoch": 5.647058823529412, "grad_norm": 9.499804496765137, "learning_rate": 1.996327283156618e-05, "loss": 2.4424, "step": 96 }, { "epoch": 5.705882352941177, "grad_norm": 9.503997802734375, "learning_rate": 1.996162343253643e-05, "loss": 2.5679, "step": 97 }, { "epoch": 5.764705882352941, "grad_norm": 7.600337982177734, "learning_rate": 1.9959937876836858e-05, "loss": 2.582, "step": 98 }, { "epoch": 5.823529411764706, "grad_norm": 9.360596656799316, "learning_rate": 1.9958216170585356e-05, "loss": 2.6553, "step": 99 }, { "epoch": 5.882352941176471, "grad_norm": 14.96545696258545, "learning_rate": 1.9956458320031016e-05, "loss": 2.4741, "step": 100 }, { "epoch": 5.9411764705882355, "grad_norm": 8.164722442626953, "learning_rate": 1.9954664331554126e-05, "loss": 2.5068, "step": 101 }, { "epoch": 6.0, "grad_norm": 12.244168281555176, "learning_rate": 1.995283421166614e-05, "loss": 2.5293, "step": 102 }, { "epoch": 6.0588235294117645, "grad_norm": 9.12121868133545, "learning_rate": 1.9950967967009657e-05, "loss": 2.2798, "step": 103 }, { "epoch": 6.117647058823529, "grad_norm": 8.4320650100708, "learning_rate": 1.994906560435839e-05, "loss": 2.4385, "step": 104 }, { "epoch": 6.176470588235294, "grad_norm": 10.35110855102539, "learning_rate": 1.9947127130617144e-05, "loss": 2.4038, "step": 105 }, { "epoch": 6.235294117647059, "grad_norm": 8.76579475402832, "learning_rate": 1.99451525528218e-05, "loss": 2.4707, "step": 106 }, { "epoch": 6.294117647058823, "grad_norm": 8.972146034240723, "learning_rate": 1.9943141878139274e-05, "loss": 2.335, "step": 107 }, { "epoch": 6.352941176470588, "grad_norm": 11.090405464172363, "learning_rate": 1.9941095113867507e-05, "loss": 2.4834, "step": 108 }, { "epoch": 6.411764705882353, "grad_norm": 10.663291931152344, "learning_rate": 1.9939012267435426e-05, "loss": 2.3755, "step": 109 }, { "epoch": 6.470588235294118, "grad_norm": 11.804580688476562, "learning_rate": 1.993689334640292e-05, "loss": 2.3931, "step": 110 }, { "epoch": 6.529411764705882, "grad_norm": 10.242867469787598, "learning_rate": 1.9934738358460818e-05, "loss": 2.3486, "step": 111 }, { "epoch": 6.588235294117647, "grad_norm": 13.05301284790039, "learning_rate": 1.9932547311430858e-05, "loss": 2.3389, "step": 112 }, { "epoch": 6.647058823529412, "grad_norm": 13.161552429199219, "learning_rate": 1.993032021326565e-05, "loss": 2.4028, "step": 113 }, { "epoch": 6.705882352941177, "grad_norm": 12.055183410644531, "learning_rate": 1.9928057072048662e-05, "loss": 2.4219, "step": 114 }, { "epoch": 6.764705882352941, "grad_norm": 9.412785530090332, "learning_rate": 1.992575789599419e-05, "loss": 2.2676, "step": 115 }, { "epoch": 6.823529411764706, "grad_norm": 9.314119338989258, "learning_rate": 1.9923422693447307e-05, "loss": 2.4082, "step": 116 }, { "epoch": 6.882352941176471, "grad_norm": 9.606888771057129, "learning_rate": 1.992105147288386e-05, "loss": 2.4521, "step": 117 }, { "epoch": 6.9411764705882355, "grad_norm": 10.189269065856934, "learning_rate": 1.9918644242910418e-05, "loss": 2.2969, "step": 118 }, { "epoch": 7.0, "grad_norm": 9.221379280090332, "learning_rate": 1.9916201012264255e-05, "loss": 2.3062, "step": 119 }, { "epoch": 7.0588235294117645, "grad_norm": 7.627739429473877, "learning_rate": 1.991372178981331e-05, "loss": 2.2065, "step": 120 }, { "epoch": 7.117647058823529, "grad_norm": 9.953232765197754, "learning_rate": 1.991120658455616e-05, "loss": 2.2441, "step": 121 }, { "epoch": 7.176470588235294, "grad_norm": 9.432150840759277, "learning_rate": 1.990865540562199e-05, "loss": 2.2773, "step": 122 }, { "epoch": 7.235294117647059, "grad_norm": 9.045056343078613, "learning_rate": 1.990606826227054e-05, "loss": 2.2261, "step": 123 }, { "epoch": 7.294117647058823, "grad_norm": 13.119576454162598, "learning_rate": 1.99034451638921e-05, "loss": 2.0693, "step": 124 }, { "epoch": 7.352941176470588, "grad_norm": 11.607037544250488, "learning_rate": 1.990078612000746e-05, "loss": 2.1821, "step": 125 }, { "epoch": 7.411764705882353, "grad_norm": 10.091217041015625, "learning_rate": 1.989809114026787e-05, "loss": 2.2764, "step": 126 }, { "epoch": 7.470588235294118, "grad_norm": 9.865386962890625, "learning_rate": 1.9895360234455023e-05, "loss": 2.2534, "step": 127 }, { "epoch": 7.529411764705882, "grad_norm": 10.750882148742676, "learning_rate": 1.9892593412481004e-05, "loss": 2.2388, "step": 128 }, { "epoch": 7.588235294117647, "grad_norm": 10.1782808303833, "learning_rate": 1.9889790684388252e-05, "loss": 2.2251, "step": 129 }, { "epoch": 7.647058823529412, "grad_norm": 10.850335121154785, "learning_rate": 1.9886952060349546e-05, "loss": 2.2954, "step": 130 }, { "epoch": 7.705882352941177, "grad_norm": 7.938097953796387, "learning_rate": 1.9884077550667942e-05, "loss": 2.3062, "step": 131 }, { "epoch": 7.764705882352941, "grad_norm": 10.520757675170898, "learning_rate": 1.988116716577675e-05, "loss": 2.3076, "step": 132 }, { "epoch": 7.823529411764706, "grad_norm": 10.912127494812012, "learning_rate": 1.987822091623949e-05, "loss": 2.1436, "step": 133 }, { "epoch": 7.882352941176471, "grad_norm": 7.468573093414307, "learning_rate": 1.987523881274986e-05, "loss": 2.2178, "step": 134 }, { "epoch": 7.9411764705882355, "grad_norm": 9.595022201538086, "learning_rate": 1.987222086613169e-05, "loss": 2.228, "step": 135 }, { "epoch": 8.0, "grad_norm": 9.515213966369629, "learning_rate": 1.9869167087338908e-05, "loss": 2.1108, "step": 136 }, { "epoch": 8.058823529411764, "grad_norm": 39.20478439331055, "learning_rate": 1.9866077487455497e-05, "loss": 2.0034, "step": 137 }, { "epoch": 8.117647058823529, "grad_norm": 8.61018180847168, "learning_rate": 1.9862952077695457e-05, "loss": 2.1016, "step": 138 }, { "epoch": 8.176470588235293, "grad_norm": 10.327203750610352, "learning_rate": 1.9859790869402764e-05, "loss": 2.1602, "step": 139 }, { "epoch": 8.235294117647058, "grad_norm": 9.717497825622559, "learning_rate": 1.9856593874051326e-05, "loss": 2.1914, "step": 140 }, { "epoch": 8.294117647058824, "grad_norm": 9.128399848937988, "learning_rate": 1.9853361103244943e-05, "loss": 1.9214, "step": 141 }, { "epoch": 8.352941176470589, "grad_norm": 9.169191360473633, "learning_rate": 1.9850092568717273e-05, "loss": 1.999, "step": 142 }, { "epoch": 8.411764705882353, "grad_norm": 25.799766540527344, "learning_rate": 1.9846788282331773e-05, "loss": 2.1455, "step": 143 }, { "epoch": 8.470588235294118, "grad_norm": 11.617899894714355, "learning_rate": 1.984344825608167e-05, "loss": 2.1113, "step": 144 }, { "epoch": 8.529411764705882, "grad_norm": 18.497943878173828, "learning_rate": 1.9840072502089906e-05, "loss": 2.2246, "step": 145 }, { "epoch": 8.588235294117647, "grad_norm": 12.115826606750488, "learning_rate": 1.9836661032609104e-05, "loss": 2.0659, "step": 146 }, { "epoch": 8.647058823529411, "grad_norm": 13.62680721282959, "learning_rate": 1.983321386002153e-05, "loss": 2.1636, "step": 147 }, { "epoch": 8.705882352941176, "grad_norm": 13.071581840515137, "learning_rate": 1.982973099683902e-05, "loss": 2.0742, "step": 148 }, { "epoch": 8.764705882352942, "grad_norm": 8.82536792755127, "learning_rate": 1.9826212455702963e-05, "loss": 2.1265, "step": 149 }, { "epoch": 8.823529411764707, "grad_norm": 10.402623176574707, "learning_rate": 1.9822658249384237e-05, "loss": 1.9976, "step": 150 }, { "epoch": 8.882352941176471, "grad_norm": 9.288298606872559, "learning_rate": 1.9819068390783184e-05, "loss": 1.9019, "step": 151 }, { "epoch": 8.941176470588236, "grad_norm": 10.52776050567627, "learning_rate": 1.981544289292954e-05, "loss": 1.9355, "step": 152 }, { "epoch": 9.0, "grad_norm": 12.23874568939209, "learning_rate": 1.9811781768982392e-05, "loss": 2.0439, "step": 153 }, { "epoch": 9.058823529411764, "grad_norm": 11.252086639404297, "learning_rate": 1.9808085032230146e-05, "loss": 2.0132, "step": 154 }, { "epoch": 9.117647058823529, "grad_norm": 12.236040115356445, "learning_rate": 1.980435269609046e-05, "loss": 1.9243, "step": 155 }, { "epoch": 9.176470588235293, "grad_norm": 10.86466121673584, "learning_rate": 1.9800584774110208e-05, "loss": 1.9824, "step": 156 }, { "epoch": 9.235294117647058, "grad_norm": 7.931847095489502, "learning_rate": 1.979678127996543e-05, "loss": 1.9546, "step": 157 }, { "epoch": 9.294117647058824, "grad_norm": 10.230725288391113, "learning_rate": 1.9792942227461265e-05, "loss": 2.002, "step": 158 }, { "epoch": 9.352941176470589, "grad_norm": 12.577863693237305, "learning_rate": 1.9789067630531932e-05, "loss": 1.854, "step": 159 }, { "epoch": 9.411764705882353, "grad_norm": 12.590815544128418, "learning_rate": 1.9785157503240645e-05, "loss": 1.8306, "step": 160 }, { "epoch": 9.470588235294118, "grad_norm": 10.349114418029785, "learning_rate": 1.9781211859779593e-05, "loss": 1.9888, "step": 161 }, { "epoch": 9.529411764705882, "grad_norm": 8.35677719116211, "learning_rate": 1.977723071446986e-05, "loss": 1.9297, "step": 162 }, { "epoch": 9.588235294117647, "grad_norm": 11.489068984985352, "learning_rate": 1.9773214081761404e-05, "loss": 1.8018, "step": 163 }, { "epoch": 9.647058823529411, "grad_norm": 11.149316787719727, "learning_rate": 1.976916197623298e-05, "loss": 1.9102, "step": 164 }, { "epoch": 9.705882352941176, "grad_norm": 12.661219596862793, "learning_rate": 1.9765074412592087e-05, "loss": 1.9863, "step": 165 }, { "epoch": 9.764705882352942, "grad_norm": 11.165590286254883, "learning_rate": 1.9760951405674936e-05, "loss": 1.9067, "step": 166 }, { "epoch": 9.823529411764707, "grad_norm": 8.715911865234375, "learning_rate": 1.9756792970446375e-05, "loss": 1.9146, "step": 167 }, { "epoch": 9.882352941176471, "grad_norm": 8.804875373840332, "learning_rate": 1.9752599121999845e-05, "loss": 1.8481, "step": 168 }, { "epoch": 9.941176470588236, "grad_norm": 10.330653190612793, "learning_rate": 1.974836987555732e-05, "loss": 1.8354, "step": 169 }, { "epoch": 10.0, "grad_norm": 8.241497993469238, "learning_rate": 1.9744105246469264e-05, "loss": 1.8472, "step": 170 }, { "epoch": 10.058823529411764, "grad_norm": 15.061513900756836, "learning_rate": 1.9739805250214546e-05, "loss": 1.8555, "step": 171 }, { "epoch": 10.117647058823529, "grad_norm": 8.238058090209961, "learning_rate": 1.9735469902400424e-05, "loss": 1.7466, "step": 172 }, { "epoch": 10.176470588235293, "grad_norm": 8.91250991821289, "learning_rate": 1.973109921876246e-05, "loss": 1.6704, "step": 173 }, { "epoch": 10.235294117647058, "grad_norm": 10.190889358520508, "learning_rate": 1.9726693215164464e-05, "loss": 1.8906, "step": 174 }, { "epoch": 10.294117647058824, "grad_norm": 9.007473945617676, "learning_rate": 1.9722251907598458e-05, "loss": 1.8066, "step": 175 }, { "epoch": 10.352941176470589, "grad_norm": 7.6829833984375, "learning_rate": 1.9717775312184592e-05, "loss": 1.8516, "step": 176 }, { "epoch": 10.411764705882353, "grad_norm": 8.435174942016602, "learning_rate": 1.971326344517109e-05, "loss": 1.8306, "step": 177 }, { "epoch": 10.470588235294118, "grad_norm": 9.037138938903809, "learning_rate": 1.9708716322934222e-05, "loss": 1.9478, "step": 178 }, { "epoch": 10.529411764705882, "grad_norm": 11.691319465637207, "learning_rate": 1.9704133961978195e-05, "loss": 1.7207, "step": 179 }, { "epoch": 10.588235294117647, "grad_norm": 9.262616157531738, "learning_rate": 1.9699516378935134e-05, "loss": 1.8604, "step": 180 }, { "epoch": 10.647058823529411, "grad_norm": 7.748818874359131, "learning_rate": 1.9694863590564997e-05, "loss": 1.8018, "step": 181 }, { "epoch": 10.705882352941176, "grad_norm": 11.156909942626953, "learning_rate": 1.969017561375553e-05, "loss": 1.6694, "step": 182 }, { "epoch": 10.764705882352942, "grad_norm": 8.987791061401367, "learning_rate": 1.968545246552219e-05, "loss": 1.7593, "step": 183 }, { "epoch": 10.823529411764707, "grad_norm": 10.345136642456055, "learning_rate": 1.9680694163008102e-05, "loss": 1.7886, "step": 184 }, { "epoch": 10.882352941176471, "grad_norm": 10.792691230773926, "learning_rate": 1.9675900723483985e-05, "loss": 1.6472, "step": 185 }, { "epoch": 10.941176470588236, "grad_norm": 8.306448936462402, "learning_rate": 1.967107216434808e-05, "loss": 1.6306, "step": 186 }, { "epoch": 11.0, "grad_norm": 10.899456977844238, "learning_rate": 1.9666208503126115e-05, "loss": 1.6299, "step": 187 }, { "epoch": 11.058823529411764, "grad_norm": 9.003738403320312, "learning_rate": 1.966130975747121e-05, "loss": 1.5857, "step": 188 }, { "epoch": 11.117647058823529, "grad_norm": 8.796257019042969, "learning_rate": 1.9656375945163834e-05, "loss": 1.6631, "step": 189 }, { "epoch": 11.176470588235293, "grad_norm": 12.069965362548828, "learning_rate": 1.965140708411174e-05, "loss": 1.6538, "step": 190 }, { "epoch": 11.235294117647058, "grad_norm": 9.124354362487793, "learning_rate": 1.9646403192349876e-05, "loss": 1.7578, "step": 191 }, { "epoch": 11.294117647058824, "grad_norm": 10.62402629852295, "learning_rate": 1.964136428804036e-05, "loss": 1.6204, "step": 192 }, { "epoch": 11.352941176470589, "grad_norm": 10.819178581237793, "learning_rate": 1.963629038947237e-05, "loss": 1.6392, "step": 193 }, { "epoch": 11.411764705882353, "grad_norm": 9.56255054473877, "learning_rate": 1.9631181515062108e-05, "loss": 1.7715, "step": 194 }, { "epoch": 11.470588235294118, "grad_norm": 18.483970642089844, "learning_rate": 1.962603768335273e-05, "loss": 1.6162, "step": 195 }, { "epoch": 11.529411764705882, "grad_norm": 12.61978530883789, "learning_rate": 1.9620858913014268e-05, "loss": 1.7207, "step": 196 }, { "epoch": 11.588235294117647, "grad_norm": 10.203033447265625, "learning_rate": 1.9615645222843566e-05, "loss": 1.5764, "step": 197 }, { "epoch": 11.647058823529411, "grad_norm": 8.901611328125, "learning_rate": 1.961039663176421e-05, "loss": 1.5552, "step": 198 }, { "epoch": 11.705882352941176, "grad_norm": 14.010965347290039, "learning_rate": 1.960511315882647e-05, "loss": 1.6226, "step": 199 }, { "epoch": 11.764705882352942, "grad_norm": 9.723356246948242, "learning_rate": 1.9599794823207217e-05, "loss": 1.6362, "step": 200 }, { "epoch": 11.823529411764707, "grad_norm": 11.106191635131836, "learning_rate": 1.9594441644209863e-05, "loss": 1.6699, "step": 201 }, { "epoch": 11.882352941176471, "grad_norm": 10.819182395935059, "learning_rate": 1.958905364126428e-05, "loss": 1.748, "step": 202 }, { "epoch": 11.941176470588236, "grad_norm": 14.87402057647705, "learning_rate": 1.958363083392675e-05, "loss": 1.5684, "step": 203 }, { "epoch": 12.0, "grad_norm": 9.130367279052734, "learning_rate": 1.957817324187987e-05, "loss": 1.6221, "step": 204 }, { "epoch": 12.058823529411764, "grad_norm": 11.417094230651855, "learning_rate": 1.9572680884932496e-05, "loss": 1.6489, "step": 205 }, { "epoch": 12.117647058823529, "grad_norm": 15.244763374328613, "learning_rate": 1.956715378301967e-05, "loss": 1.4326, "step": 206 }, { "epoch": 12.176470588235293, "grad_norm": 10.178763389587402, "learning_rate": 1.956159195620253e-05, "loss": 1.4785, "step": 207 }, { "epoch": 12.235294117647058, "grad_norm": 13.306642532348633, "learning_rate": 1.9555995424668266e-05, "loss": 1.5862, "step": 208 }, { "epoch": 12.294117647058824, "grad_norm": 13.248449325561523, "learning_rate": 1.9550364208730026e-05, "loss": 1.4546, "step": 209 }, { "epoch": 12.352941176470589, "grad_norm": 9.095641136169434, "learning_rate": 1.954469832882685e-05, "loss": 1.439, "step": 210 }, { "epoch": 12.411764705882353, "grad_norm": 11.951369285583496, "learning_rate": 1.9538997805523595e-05, "loss": 1.5576, "step": 211 }, { "epoch": 12.470588235294118, "grad_norm": 9.02932357788086, "learning_rate": 1.9533262659510858e-05, "loss": 1.5671, "step": 212 }, { "epoch": 12.529411764705882, "grad_norm": 10.480745315551758, "learning_rate": 1.9527492911604905e-05, "loss": 1.5908, "step": 213 }, { "epoch": 12.588235294117647, "grad_norm": 10.89215087890625, "learning_rate": 1.9521688582747585e-05, "loss": 1.645, "step": 214 }, { "epoch": 12.647058823529411, "grad_norm": 9.34056568145752, "learning_rate": 1.9515849694006272e-05, "loss": 1.3662, "step": 215 }, { "epoch": 12.705882352941176, "grad_norm": 11.367390632629395, "learning_rate": 1.950997626657377e-05, "loss": 1.4558, "step": 216 }, { "epoch": 12.764705882352942, "grad_norm": 10.650875091552734, "learning_rate": 1.950406832176825e-05, "loss": 1.4001, "step": 217 }, { "epoch": 12.823529411764707, "grad_norm": 9.907959938049316, "learning_rate": 1.949812588103317e-05, "loss": 1.502, "step": 218 }, { "epoch": 12.882352941176471, "grad_norm": 10.009119033813477, "learning_rate": 1.9492148965937187e-05, "loss": 1.5288, "step": 219 }, { "epoch": 12.941176470588236, "grad_norm": 9.250482559204102, "learning_rate": 1.9486137598174085e-05, "loss": 1.5981, "step": 220 }, { "epoch": 13.0, "grad_norm": 8.99602222442627, "learning_rate": 1.9480091799562706e-05, "loss": 1.5469, "step": 221 }, { "epoch": 13.058823529411764, "grad_norm": 13.221899032592773, "learning_rate": 1.9474011592046855e-05, "loss": 1.5469, "step": 222 }, { "epoch": 13.117647058823529, "grad_norm": 10.545061111450195, "learning_rate": 1.9467896997695232e-05, "loss": 1.3787, "step": 223 }, { "epoch": 13.176470588235293, "grad_norm": 8.502839088439941, "learning_rate": 1.9461748038701344e-05, "loss": 1.3899, "step": 224 }, { "epoch": 13.235294117647058, "grad_norm": 10.502857208251953, "learning_rate": 1.9455564737383423e-05, "loss": 1.6831, "step": 225 }, { "epoch": 13.294117647058824, "grad_norm": 10.673855781555176, "learning_rate": 1.9449347116184364e-05, "loss": 1.3857, "step": 226 }, { "epoch": 13.352941176470589, "grad_norm": 9.116896629333496, "learning_rate": 1.9443095197671615e-05, "loss": 1.2139, "step": 227 }, { "epoch": 13.411764705882353, "grad_norm": 9.33806324005127, "learning_rate": 1.943680900453712e-05, "loss": 1.4514, "step": 228 }, { "epoch": 13.470588235294118, "grad_norm": 8.920222282409668, "learning_rate": 1.943048855959722e-05, "loss": 1.4624, "step": 229 }, { "epoch": 13.529411764705882, "grad_norm": 9.52849006652832, "learning_rate": 1.9424133885792575e-05, "loss": 1.3933, "step": 230 }, { "epoch": 13.588235294117647, "grad_norm": 9.704018592834473, "learning_rate": 1.9417745006188082e-05, "loss": 1.3381, "step": 231 }, { "epoch": 13.647058823529411, "grad_norm": 9.277046203613281, "learning_rate": 1.94113219439728e-05, "loss": 1.4097, "step": 232 }, { "epoch": 13.705882352941176, "grad_norm": 8.906158447265625, "learning_rate": 1.9404864722459852e-05, "loss": 1.5029, "step": 233 }, { "epoch": 13.764705882352942, "grad_norm": 9.329651832580566, "learning_rate": 1.9398373365086335e-05, "loss": 1.2419, "step": 234 }, { "epoch": 13.823529411764707, "grad_norm": 10.241909980773926, "learning_rate": 1.939184789541326e-05, "loss": 1.3358, "step": 235 }, { "epoch": 13.882352941176471, "grad_norm": 9.222643852233887, "learning_rate": 1.9385288337125448e-05, "loss": 1.261, "step": 236 }, { "epoch": 13.941176470588236, "grad_norm": 8.16667652130127, "learning_rate": 1.9378694714031442e-05, "loss": 1.3579, "step": 237 }, { "epoch": 14.0, "grad_norm": 9.825841903686523, "learning_rate": 1.937206705006344e-05, "loss": 1.3169, "step": 238 }, { "epoch": 14.058823529411764, "grad_norm": 9.070094108581543, "learning_rate": 1.9365405369277172e-05, "loss": 1.2903, "step": 239 }, { "epoch": 14.117647058823529, "grad_norm": 8.340785026550293, "learning_rate": 1.935870969585186e-05, "loss": 1.2512, "step": 240 }, { "epoch": 14.176470588235293, "grad_norm": 8.328289985656738, "learning_rate": 1.9351980054090086e-05, "loss": 1.3027, "step": 241 }, { "epoch": 14.235294117647058, "grad_norm": 9.469372749328613, "learning_rate": 1.9345216468417737e-05, "loss": 1.2715, "step": 242 }, { "epoch": 14.294117647058824, "grad_norm": 8.60491943359375, "learning_rate": 1.93384189633839e-05, "loss": 1.2234, "step": 243 }, { "epoch": 14.352941176470589, "grad_norm": 7.4619245529174805, "learning_rate": 1.9331587563660766e-05, "loss": 1.4731, "step": 244 }, { "epoch": 14.411764705882353, "grad_norm": 9.062922477722168, "learning_rate": 1.932472229404356e-05, "loss": 1.2881, "step": 245 }, { "epoch": 14.470588235294118, "grad_norm": 7.204548358917236, "learning_rate": 1.9317823179450438e-05, "loss": 1.2804, "step": 246 }, { "epoch": 14.529411764705882, "grad_norm": 8.042075157165527, "learning_rate": 1.93108902449224e-05, "loss": 1.4238, "step": 247 }, { "epoch": 14.588235294117647, "grad_norm": 10.834410667419434, "learning_rate": 1.93039235156232e-05, "loss": 1.2344, "step": 248 }, { "epoch": 14.647058823529411, "grad_norm": 8.23576831817627, "learning_rate": 1.9296923016839243e-05, "loss": 1.2053, "step": 249 }, { "epoch": 14.705882352941176, "grad_norm": 8.060479164123535, "learning_rate": 1.9289888773979524e-05, "loss": 1.4006, "step": 250 }, { "epoch": 14.764705882352942, "grad_norm": 10.602763175964355, "learning_rate": 1.9282820812575497e-05, "loss": 1.3438, "step": 251 }, { "epoch": 14.823529411764707, "grad_norm": 7.457324504852295, "learning_rate": 1.927571915828101e-05, "loss": 1.2528, "step": 252 }, { "epoch": 14.882352941176471, "grad_norm": 8.067986488342285, "learning_rate": 1.9268583836872198e-05, "loss": 1.1621, "step": 253 }, { "epoch": 14.941176470588236, "grad_norm": 7.343730449676514, "learning_rate": 1.9261414874247394e-05, "loss": 1.1804, "step": 254 }, { "epoch": 15.0, "grad_norm": 10.232341766357422, "learning_rate": 1.9254212296427043e-05, "loss": 1.2881, "step": 255 }, { "epoch": 15.058823529411764, "grad_norm": 8.354348182678223, "learning_rate": 1.924697612955359e-05, "loss": 1.2898, "step": 256 }, { "epoch": 15.117647058823529, "grad_norm": 6.956756114959717, "learning_rate": 1.9239706399891398e-05, "loss": 1.1243, "step": 257 }, { "epoch": 15.176470588235293, "grad_norm": 8.421477317810059, "learning_rate": 1.923240313382665e-05, "loss": 1.1506, "step": 258 }, { "epoch": 15.235294117647058, "grad_norm": 8.68120002746582, "learning_rate": 1.922506635786725e-05, "loss": 1.072, "step": 259 }, { "epoch": 15.294117647058824, "grad_norm": 8.12938404083252, "learning_rate": 1.9217696098642736e-05, "loss": 1.1599, "step": 260 }, { "epoch": 15.352941176470589, "grad_norm": 7.2845306396484375, "learning_rate": 1.921029238290417e-05, "loss": 1.2502, "step": 261 }, { "epoch": 15.411764705882353, "grad_norm": 7.830509185791016, "learning_rate": 1.9202855237524053e-05, "loss": 0.9521, "step": 262 }, { "epoch": 15.470588235294118, "grad_norm": 10.00294017791748, "learning_rate": 1.9195384689496212e-05, "loss": 1.168, "step": 263 }, { "epoch": 15.529411764705882, "grad_norm": 8.725485801696777, "learning_rate": 1.9187880765935724e-05, "loss": 1.2007, "step": 264 }, { "epoch": 15.588235294117647, "grad_norm": 6.980824947357178, "learning_rate": 1.91803434940788e-05, "loss": 1.053, "step": 265 }, { "epoch": 15.647058823529411, "grad_norm": 7.76743221282959, "learning_rate": 1.9172772901282694e-05, "loss": 1.0919, "step": 266 }, { "epoch": 15.705882352941176, "grad_norm": 9.755024909973145, "learning_rate": 1.91651690150256e-05, "loss": 1.2671, "step": 267 }, { "epoch": 15.764705882352942, "grad_norm": 9.392074584960938, "learning_rate": 1.9157531862906557e-05, "loss": 1.121, "step": 268 }, { "epoch": 15.823529411764707, "grad_norm": 8.740220069885254, "learning_rate": 1.914986147264533e-05, "loss": 1.2759, "step": 269 }, { "epoch": 15.882352941176471, "grad_norm": 8.985494613647461, "learning_rate": 1.9142157872082354e-05, "loss": 1.3005, "step": 270 }, { "epoch": 15.941176470588236, "grad_norm": 8.223883628845215, "learning_rate": 1.9134421089178578e-05, "loss": 1.1006, "step": 271 }, { "epoch": 16.0, "grad_norm": 8.415818214416504, "learning_rate": 1.9126651152015404e-05, "loss": 1.2512, "step": 272 }, { "epoch": 16.058823529411764, "grad_norm": 7.3380656242370605, "learning_rate": 1.911884808879456e-05, "loss": 1.0537, "step": 273 }, { "epoch": 16.11764705882353, "grad_norm": 8.667423248291016, "learning_rate": 1.9111011927838024e-05, "loss": 0.9004, "step": 274 }, { "epoch": 16.176470588235293, "grad_norm": 8.700292587280273, "learning_rate": 1.9103142697587884e-05, "loss": 1.1348, "step": 275 }, { "epoch": 16.235294117647058, "grad_norm": 8.624340057373047, "learning_rate": 1.9095240426606273e-05, "loss": 1.0067, "step": 276 }, { "epoch": 16.294117647058822, "grad_norm": 10.75705337524414, "learning_rate": 1.9087305143575245e-05, "loss": 1.1282, "step": 277 }, { "epoch": 16.352941176470587, "grad_norm": 8.72046947479248, "learning_rate": 1.907933687729667e-05, "loss": 1.0543, "step": 278 }, { "epoch": 16.41176470588235, "grad_norm": 9.421009063720703, "learning_rate": 1.907133565669214e-05, "loss": 1.1357, "step": 279 }, { "epoch": 16.470588235294116, "grad_norm": 8.6544828414917, "learning_rate": 1.9063301510802846e-05, "loss": 0.9988, "step": 280 }, { "epoch": 16.529411764705884, "grad_norm": 7.3404083251953125, "learning_rate": 1.90552344687895e-05, "loss": 1.1375, "step": 281 }, { "epoch": 16.58823529411765, "grad_norm": 10.884233474731445, "learning_rate": 1.9047134559932203e-05, "loss": 1.0933, "step": 282 }, { "epoch": 16.647058823529413, "grad_norm": 8.752082824707031, "learning_rate": 1.9039001813630357e-05, "loss": 1.0898, "step": 283 }, { "epoch": 16.705882352941178, "grad_norm": 8.246918678283691, "learning_rate": 1.9030836259402545e-05, "loss": 0.9935, "step": 284 }, { "epoch": 16.764705882352942, "grad_norm": 9.318117141723633, "learning_rate": 1.9022637926886427e-05, "loss": 1.1075, "step": 285 }, { "epoch": 16.823529411764707, "grad_norm": 9.169379234313965, "learning_rate": 1.9014406845838643e-05, "loss": 1.1204, "step": 286 }, { "epoch": 16.88235294117647, "grad_norm": 8.516507148742676, "learning_rate": 1.900614304613469e-05, "loss": 1.0875, "step": 287 }, { "epoch": 16.941176470588236, "grad_norm": 8.264412879943848, "learning_rate": 1.899784655776883e-05, "loss": 1.0369, "step": 288 }, { "epoch": 17.0, "grad_norm": 10.037005424499512, "learning_rate": 1.8989517410853956e-05, "loss": 1.1218, "step": 289 }, { "epoch": 17.058823529411764, "grad_norm": 7.080535411834717, "learning_rate": 1.898115563562151e-05, "loss": 1.075, "step": 290 }, { "epoch": 17.11764705882353, "grad_norm": 8.170382499694824, "learning_rate": 1.8972761262421363e-05, "loss": 1.0338, "step": 291 }, { "epoch": 17.176470588235293, "grad_norm": 8.794394493103027, "learning_rate": 1.8964334321721698e-05, "loss": 1.093, "step": 292 }, { "epoch": 17.235294117647058, "grad_norm": 7.711061477661133, "learning_rate": 1.8955874844108903e-05, "loss": 1.0306, "step": 293 }, { "epoch": 17.294117647058822, "grad_norm": 7.933358192443848, "learning_rate": 1.894738286028747e-05, "loss": 0.9176, "step": 294 }, { "epoch": 17.352941176470587, "grad_norm": 7.066795825958252, "learning_rate": 1.8938858401079866e-05, "loss": 1.0131, "step": 295 }, { "epoch": 17.41176470588235, "grad_norm": 7.483026504516602, "learning_rate": 1.8930301497426434e-05, "loss": 0.8979, "step": 296 }, { "epoch": 17.470588235294116, "grad_norm": 9.090031623840332, "learning_rate": 1.8921712180385287e-05, "loss": 0.9585, "step": 297 }, { "epoch": 17.529411764705884, "grad_norm": 8.832393646240234, "learning_rate": 1.891309048113217e-05, "loss": 0.9143, "step": 298 }, { "epoch": 17.58823529411765, "grad_norm": 7.829322338104248, "learning_rate": 1.890443643096037e-05, "loss": 1.0073, "step": 299 }, { "epoch": 17.647058823529413, "grad_norm": 14.34745979309082, "learning_rate": 1.889575006128059e-05, "loss": 1.075, "step": 300 }, { "epoch": 17.705882352941178, "grad_norm": 8.352957725524902, "learning_rate": 1.888703140362085e-05, "loss": 0.9565, "step": 301 }, { "epoch": 17.764705882352942, "grad_norm": 9.477667808532715, "learning_rate": 1.887828048962636e-05, "loss": 1.0778, "step": 302 }, { "epoch": 17.823529411764707, "grad_norm": 9.908851623535156, "learning_rate": 1.8869497351059393e-05, "loss": 0.8929, "step": 303 }, { "epoch": 17.88235294117647, "grad_norm": 8.893075942993164, "learning_rate": 1.8860682019799203e-05, "loss": 1.0883, "step": 304 }, { "epoch": 17.941176470588236, "grad_norm": 9.700044631958008, "learning_rate": 1.8851834527841877e-05, "loss": 1.052, "step": 305 }, { "epoch": 18.0, "grad_norm": 7.838306427001953, "learning_rate": 1.8842954907300236e-05, "loss": 0.9701, "step": 306 }, { "epoch": 18.058823529411764, "grad_norm": 10.048843383789062, "learning_rate": 1.8834043190403727e-05, "loss": 0.8104, "step": 307 }, { "epoch": 18.11764705882353, "grad_norm": 8.519925117492676, "learning_rate": 1.882509940949827e-05, "loss": 1.1232, "step": 308 }, { "epoch": 18.176470588235293, "grad_norm": 7.925708293914795, "learning_rate": 1.8816123597046183e-05, "loss": 1.052, "step": 309 }, { "epoch": 18.235294117647058, "grad_norm": 7.492148399353027, "learning_rate": 1.8807115785626043e-05, "loss": 0.8915, "step": 310 }, { "epoch": 18.294117647058822, "grad_norm": 9.188785552978516, "learning_rate": 1.879807600793256e-05, "loss": 1.0044, "step": 311 }, { "epoch": 18.352941176470587, "grad_norm": 8.188364028930664, "learning_rate": 1.878900429677649e-05, "loss": 0.7354, "step": 312 }, { "epoch": 18.41176470588235, "grad_norm": 12.29259204864502, "learning_rate": 1.8779900685084462e-05, "loss": 1.0007, "step": 313 }, { "epoch": 18.470588235294116, "grad_norm": 8.938858985900879, "learning_rate": 1.8770765205898925e-05, "loss": 0.8465, "step": 314 }, { "epoch": 18.529411764705884, "grad_norm": 8.054702758789062, "learning_rate": 1.876159789237797e-05, "loss": 0.9232, "step": 315 }, { "epoch": 18.58823529411765, "grad_norm": 7.835578918457031, "learning_rate": 1.875239877779525e-05, "loss": 0.8226, "step": 316 }, { "epoch": 18.647058823529413, "grad_norm": 13.629203796386719, "learning_rate": 1.8743167895539827e-05, "loss": 0.7832, "step": 317 }, { "epoch": 18.705882352941178, "grad_norm": 8.966890335083008, "learning_rate": 1.8733905279116083e-05, "loss": 0.9501, "step": 318 }, { "epoch": 18.764705882352942, "grad_norm": 8.845645904541016, "learning_rate": 1.872461096214357e-05, "loss": 1.0684, "step": 319 }, { "epoch": 18.823529411764707, "grad_norm": 7.386916637420654, "learning_rate": 1.8715284978356908e-05, "loss": 0.8246, "step": 320 }, { "epoch": 18.88235294117647, "grad_norm": 12.755834579467773, "learning_rate": 1.870592736160565e-05, "loss": 0.8374, "step": 321 }, { "epoch": 18.941176470588236, "grad_norm": 7.769658088684082, "learning_rate": 1.869653814585416e-05, "loss": 1.053, "step": 322 }, { "epoch": 19.0, "grad_norm": 8.368332862854004, "learning_rate": 1.8687117365181514e-05, "loss": 0.7628, "step": 323 }, { "epoch": 19.058823529411764, "grad_norm": 8.025890350341797, "learning_rate": 1.8677665053781324e-05, "loss": 0.8486, "step": 324 }, { "epoch": 19.11764705882353, "grad_norm": 8.173334121704102, "learning_rate": 1.8668181245961678e-05, "loss": 0.8848, "step": 325 }, { "epoch": 19.176470588235293, "grad_norm": 9.85013484954834, "learning_rate": 1.8658665976144955e-05, "loss": 0.7292, "step": 326 }, { "epoch": 19.235294117647058, "grad_norm": 6.760746955871582, "learning_rate": 1.864911927886775e-05, "loss": 0.8156, "step": 327 }, { "epoch": 19.294117647058822, "grad_norm": 6.198314189910889, "learning_rate": 1.8639541188780712e-05, "loss": 0.7821, "step": 328 }, { "epoch": 19.352941176470587, "grad_norm": 9.005284309387207, "learning_rate": 1.8629931740648446e-05, "loss": 0.7618, "step": 329 }, { "epoch": 19.41176470588235, "grad_norm": 9.195181846618652, "learning_rate": 1.8620290969349365e-05, "loss": 0.9202, "step": 330 }, { "epoch": 19.470588235294116, "grad_norm": 9.459832191467285, "learning_rate": 1.8610618909875575e-05, "loss": 1.0681, "step": 331 }, { "epoch": 19.529411764705884, "grad_norm": 6.775201320648193, "learning_rate": 1.8600915597332748e-05, "loss": 0.9631, "step": 332 }, { "epoch": 19.58823529411765, "grad_norm": 9.778947830200195, "learning_rate": 1.8591181066939984e-05, "loss": 0.8514, "step": 333 }, { "epoch": 19.647058823529413, "grad_norm": 7.898700714111328, "learning_rate": 1.85814153540297e-05, "loss": 0.8159, "step": 334 }, { "epoch": 19.705882352941178, "grad_norm": 9.369712829589844, "learning_rate": 1.8571618494047494e-05, "loss": 0.7833, "step": 335 }, { "epoch": 19.764705882352942, "grad_norm": 7.676392555236816, "learning_rate": 1.8561790522552005e-05, "loss": 0.8095, "step": 336 }, { "epoch": 19.823529411764707, "grad_norm": 6.871535778045654, "learning_rate": 1.85519314752148e-05, "loss": 0.8762, "step": 337 }, { "epoch": 19.88235294117647, "grad_norm": 9.259310722351074, "learning_rate": 1.8542041387820244e-05, "loss": 0.8231, "step": 338 }, { "epoch": 19.941176470588236, "grad_norm": 6.931603908538818, "learning_rate": 1.853212029626536e-05, "loss": 0.7739, "step": 339 }, { "epoch": 20.0, "grad_norm": 8.068831443786621, "learning_rate": 1.8522168236559693e-05, "loss": 0.8663, "step": 340 }, { "epoch": 20.058823529411764, "grad_norm": 6.260993957519531, "learning_rate": 1.8512185244825215e-05, "loss": 0.7877, "step": 341 }, { "epoch": 20.11764705882353, "grad_norm": 6.125193119049072, "learning_rate": 1.8502171357296144e-05, "loss": 0.7488, "step": 342 }, { "epoch": 20.176470588235293, "grad_norm": 6.586556434631348, "learning_rate": 1.849212661031885e-05, "loss": 0.5836, "step": 343 }, { "epoch": 20.235294117647058, "grad_norm": 8.132612228393555, "learning_rate": 1.848205104035171e-05, "loss": 0.7032, "step": 344 }, { "epoch": 20.294117647058822, "grad_norm": 19.203317642211914, "learning_rate": 1.8471944683964972e-05, "loss": 0.6055, "step": 345 }, { "epoch": 20.352941176470587, "grad_norm": 9.117682456970215, "learning_rate": 1.8461807577840628e-05, "loss": 0.821, "step": 346 }, { "epoch": 20.41176470588235, "grad_norm": 7.463310718536377, "learning_rate": 1.8451639758772275e-05, "loss": 0.746, "step": 347 }, { "epoch": 20.470588235294116, "grad_norm": 6.011261940002441, "learning_rate": 1.8441441263665e-05, "loss": 0.7736, "step": 348 }, { "epoch": 20.529411764705884, "grad_norm": 8.924156188964844, "learning_rate": 1.843121212953521e-05, "loss": 0.9622, "step": 349 }, { "epoch": 20.58823529411765, "grad_norm": 11.159602165222168, "learning_rate": 1.842095239351054e-05, "loss": 0.679, "step": 350 }, { "epoch": 20.647058823529413, "grad_norm": 7.689438343048096, "learning_rate": 1.841066209282968e-05, "loss": 0.8718, "step": 351 }, { "epoch": 20.705882352941178, "grad_norm": 5.8976731300354, "learning_rate": 1.8400341264842276e-05, "loss": 0.8431, "step": 352 }, { "epoch": 20.764705882352942, "grad_norm": 6.383970737457275, "learning_rate": 1.8389989947008753e-05, "loss": 0.9081, "step": 353 }, { "epoch": 20.823529411764707, "grad_norm": 8.230146408081055, "learning_rate": 1.837960817690022e-05, "loss": 0.759, "step": 354 }, { "epoch": 20.88235294117647, "grad_norm": 7.361762046813965, "learning_rate": 1.836919599219831e-05, "loss": 0.7211, "step": 355 }, { "epoch": 20.941176470588236, "grad_norm": 7.112512111663818, "learning_rate": 1.8358753430695047e-05, "loss": 0.59, "step": 356 }, { "epoch": 21.0, "grad_norm": 7.234877586364746, "learning_rate": 1.8348280530292712e-05, "loss": 0.8176, "step": 357 }, { "epoch": 21.058823529411764, "grad_norm": 6.833087921142578, "learning_rate": 1.8337777329003707e-05, "loss": 0.8085, "step": 358 }, { "epoch": 21.11764705882353, "grad_norm": 6.714828014373779, "learning_rate": 1.8327243864950405e-05, "loss": 0.5746, "step": 359 }, { "epoch": 21.176470588235293, "grad_norm": 7.178631782531738, "learning_rate": 1.8316680176365032e-05, "loss": 0.8804, "step": 360 }, { "epoch": 21.235294117647058, "grad_norm": 8.206107139587402, "learning_rate": 1.8306086301589505e-05, "loss": 0.7128, "step": 361 }, { "epoch": 21.294117647058822, "grad_norm": 7.00532865524292, "learning_rate": 1.8295462279075317e-05, "loss": 0.6833, "step": 362 }, { "epoch": 21.352941176470587, "grad_norm": 7.006269931793213, "learning_rate": 1.8284808147383374e-05, "loss": 0.6337, "step": 363 }, { "epoch": 21.41176470588235, "grad_norm": 6.161952018737793, "learning_rate": 1.8274123945183874e-05, "loss": 0.6311, "step": 364 }, { "epoch": 21.470588235294116, "grad_norm": 6.1048479080200195, "learning_rate": 1.8263409711256158e-05, "loss": 0.6677, "step": 365 }, { "epoch": 21.529411764705884, "grad_norm": 7.036968231201172, "learning_rate": 1.8252665484488558e-05, "loss": 0.6809, "step": 366 }, { "epoch": 21.58823529411765, "grad_norm": 6.070561408996582, "learning_rate": 1.8241891303878292e-05, "loss": 0.6641, "step": 367 }, { "epoch": 21.647058823529413, "grad_norm": 6.6453657150268555, "learning_rate": 1.8231087208531272e-05, "loss": 0.7266, "step": 368 }, { "epoch": 21.705882352941178, "grad_norm": 6.062489986419678, "learning_rate": 1.822025323766201e-05, "loss": 0.6642, "step": 369 }, { "epoch": 21.764705882352942, "grad_norm": 6.50183629989624, "learning_rate": 1.820938943059344e-05, "loss": 0.6376, "step": 370 }, { "epoch": 21.823529411764707, "grad_norm": 7.810697555541992, "learning_rate": 1.819849582675679e-05, "loss": 0.4641, "step": 371 }, { "epoch": 21.88235294117647, "grad_norm": 4.946179389953613, "learning_rate": 1.818757246569145e-05, "loss": 0.8298, "step": 372 }, { "epoch": 21.941176470588236, "grad_norm": 10.589685440063477, "learning_rate": 1.8176619387044807e-05, "loss": 0.5784, "step": 373 }, { "epoch": 22.0, "grad_norm": 8.507018089294434, "learning_rate": 1.816563663057211e-05, "loss": 0.777, "step": 374 }, { "epoch": 22.058823529411764, "grad_norm": 5.676335334777832, "learning_rate": 1.8154624236136333e-05, "loss": 0.7692, "step": 375 }, { "epoch": 22.11764705882353, "grad_norm": 5.446448802947998, "learning_rate": 1.8143582243708016e-05, "loss": 0.6194, "step": 376 }, { "epoch": 22.176470588235293, "grad_norm": 5.533574104309082, "learning_rate": 1.8132510693365135e-05, "loss": 0.5524, "step": 377 }, { "epoch": 22.235294117647058, "grad_norm": 5.63781213760376, "learning_rate": 1.8121409625292946e-05, "loss": 0.5985, "step": 378 }, { "epoch": 22.294117647058822, "grad_norm": 9.048171997070312, "learning_rate": 1.811027907978384e-05, "loss": 0.6218, "step": 379 }, { "epoch": 22.352941176470587, "grad_norm": 6.552326679229736, "learning_rate": 1.8099119097237207e-05, "loss": 0.6094, "step": 380 }, { "epoch": 22.41176470588235, "grad_norm": 6.26932430267334, "learning_rate": 1.808792971815927e-05, "loss": 0.7028, "step": 381 }, { "epoch": 22.470588235294116, "grad_norm": 7.909559726715088, "learning_rate": 1.8076710983162963e-05, "loss": 0.6774, "step": 382 }, { "epoch": 22.529411764705884, "grad_norm": 5.711240768432617, "learning_rate": 1.8065462932967754e-05, "loss": 0.6434, "step": 383 }, { "epoch": 22.58823529411765, "grad_norm": 5.687758445739746, "learning_rate": 1.8054185608399534e-05, "loss": 0.7344, "step": 384 }, { "epoch": 22.647058823529413, "grad_norm": 5.714588642120361, "learning_rate": 1.8042879050390426e-05, "loss": 0.5451, "step": 385 }, { "epoch": 22.705882352941178, "grad_norm": 7.995841026306152, "learning_rate": 1.8031543299978673e-05, "loss": 0.6644, "step": 386 }, { "epoch": 22.764705882352942, "grad_norm": 10.562273979187012, "learning_rate": 1.8020178398308475e-05, "loss": 0.598, "step": 387 }, { "epoch": 22.823529411764707, "grad_norm": 5.781543731689453, "learning_rate": 1.8008784386629828e-05, "loss": 0.6333, "step": 388 }, { "epoch": 22.88235294117647, "grad_norm": 6.832375526428223, "learning_rate": 1.7997361306298398e-05, "loss": 0.6072, "step": 389 }, { "epoch": 22.941176470588236, "grad_norm": 5.10422945022583, "learning_rate": 1.798590919877535e-05, "loss": 0.5154, "step": 390 }, { "epoch": 23.0, "grad_norm": 4.947280406951904, "learning_rate": 1.797442810562721e-05, "loss": 0.538, "step": 391 }, { "epoch": 23.058823529411764, "grad_norm": 5.977565288543701, "learning_rate": 1.796291806852571e-05, "loss": 0.6868, "step": 392 }, { "epoch": 23.11764705882353, "grad_norm": 5.926088809967041, "learning_rate": 1.7951379129247634e-05, "loss": 0.4657, "step": 393 }, { "epoch": 23.176470588235293, "grad_norm": 5.755044937133789, "learning_rate": 1.7939811329674673e-05, "loss": 0.5793, "step": 394 }, { "epoch": 23.235294117647058, "grad_norm": 5.695695877075195, "learning_rate": 1.7928214711793272e-05, "loss": 0.7013, "step": 395 }, { "epoch": 23.294117647058822, "grad_norm": 6.363724708557129, "learning_rate": 1.7916589317694463e-05, "loss": 0.5943, "step": 396 }, { "epoch": 23.352941176470587, "grad_norm": 5.796786308288574, "learning_rate": 1.7904935189573733e-05, "loss": 0.4719, "step": 397 }, { "epoch": 23.41176470588235, "grad_norm": 5.725315093994141, "learning_rate": 1.789325236973086e-05, "loss": 0.5222, "step": 398 }, { "epoch": 23.470588235294116, "grad_norm": 5.31943941116333, "learning_rate": 1.7881540900569767e-05, "loss": 0.3718, "step": 399 }, { "epoch": 23.529411764705884, "grad_norm": 5.5251054763793945, "learning_rate": 1.7869800824598347e-05, "loss": 0.7034, "step": 400 }, { "epoch": 23.58823529411765, "grad_norm": 5.74472188949585, "learning_rate": 1.7858032184428348e-05, "loss": 0.516, "step": 401 }, { "epoch": 23.647058823529413, "grad_norm": 5.440629005432129, "learning_rate": 1.7846235022775165e-05, "loss": 0.6029, "step": 402 }, { "epoch": 23.705882352941178, "grad_norm": 6.5425705909729, "learning_rate": 1.7834409382457745e-05, "loss": 0.6733, "step": 403 }, { "epoch": 23.764705882352942, "grad_norm": 5.712334156036377, "learning_rate": 1.7822555306398378e-05, "loss": 0.4878, "step": 404 }, { "epoch": 23.823529411764707, "grad_norm": 5.677436828613281, "learning_rate": 1.7810672837622573e-05, "loss": 0.8976, "step": 405 }, { "epoch": 23.88235294117647, "grad_norm": 5.427934646606445, "learning_rate": 1.7798762019258892e-05, "loss": 0.4551, "step": 406 }, { "epoch": 23.941176470588236, "grad_norm": 5.941454887390137, "learning_rate": 1.77868228945388e-05, "loss": 0.5134, "step": 407 }, { "epoch": 24.0, "grad_norm": 5.640189170837402, "learning_rate": 1.7774855506796497e-05, "loss": 0.6156, "step": 408 }, { "epoch": 24.058823529411764, "grad_norm": 4.917799472808838, "learning_rate": 1.776285989946876e-05, "loss": 0.5391, "step": 409 }, { "epoch": 24.11764705882353, "grad_norm": 4.497367858886719, "learning_rate": 1.7750836116094805e-05, "loss": 0.3932, "step": 410 }, { "epoch": 24.176470588235293, "grad_norm": 4.95587682723999, "learning_rate": 1.773878420031611e-05, "loss": 0.3559, "step": 411 }, { "epoch": 24.235294117647058, "grad_norm": 5.035000324249268, "learning_rate": 1.772670419587626e-05, "loss": 0.3825, "step": 412 }, { "epoch": 24.294117647058822, "grad_norm": 4.602354526519775, "learning_rate": 1.7714596146620794e-05, "loss": 0.5561, "step": 413 }, { "epoch": 24.352941176470587, "grad_norm": 5.243723392486572, "learning_rate": 1.7702460096497035e-05, "loss": 0.5418, "step": 414 }, { "epoch": 24.41176470588235, "grad_norm": 5.492788314819336, "learning_rate": 1.7690296089553952e-05, "loss": 0.5236, "step": 415 }, { "epoch": 24.470588235294116, "grad_norm": 5.047097206115723, "learning_rate": 1.7678104169941972e-05, "loss": 0.5447, "step": 416 }, { "epoch": 24.529411764705884, "grad_norm": 5.079856872558594, "learning_rate": 1.766588438191284e-05, "loss": 0.5696, "step": 417 }, { "epoch": 24.58823529411765, "grad_norm": 6.489667892456055, "learning_rate": 1.765363676981945e-05, "loss": 0.4746, "step": 418 }, { "epoch": 24.647058823529413, "grad_norm": 5.933997631072998, "learning_rate": 1.764136137811569e-05, "loss": 0.6847, "step": 419 }, { "epoch": 24.705882352941178, "grad_norm": 5.186289310455322, "learning_rate": 1.7629058251356277e-05, "loss": 0.59, "step": 420 }, { "epoch": 24.764705882352942, "grad_norm": 4.73969030380249, "learning_rate": 1.7616727434196587e-05, "loss": 0.5649, "step": 421 }, { "epoch": 24.823529411764707, "grad_norm": 5.3309550285339355, "learning_rate": 1.7604368971392508e-05, "loss": 0.4822, "step": 422 }, { "epoch": 24.88235294117647, "grad_norm": 6.195250511169434, "learning_rate": 1.7591982907800264e-05, "loss": 0.5919, "step": 423 }, { "epoch": 24.941176470588236, "grad_norm": 5.542773246765137, "learning_rate": 1.7579569288376276e-05, "loss": 0.4364, "step": 424 }, { "epoch": 25.0, "grad_norm": 4.799807548522949, "learning_rate": 1.7567128158176955e-05, "loss": 0.5948, "step": 425 }, { "epoch": 25.058823529411764, "grad_norm": 6.300415992736816, "learning_rate": 1.7554659562358585e-05, "loss": 0.4458, "step": 426 }, { "epoch": 25.11764705882353, "grad_norm": 4.52308988571167, "learning_rate": 1.7542163546177136e-05, "loss": 0.3868, "step": 427 }, { "epoch": 25.176470588235293, "grad_norm": 4.874640464782715, "learning_rate": 1.7529640154988095e-05, "loss": 0.6342, "step": 428 }, { "epoch": 25.235294117647058, "grad_norm": 5.02623176574707, "learning_rate": 1.7517089434246316e-05, "loss": 0.4317, "step": 429 }, { "epoch": 25.294117647058822, "grad_norm": 5.959002494812012, "learning_rate": 1.7504511429505845e-05, "loss": 0.4948, "step": 430 }, { "epoch": 25.352941176470587, "grad_norm": 5.981856822967529, "learning_rate": 1.7491906186419762e-05, "loss": 0.5035, "step": 431 }, { "epoch": 25.41176470588235, "grad_norm": 6.132950782775879, "learning_rate": 1.7479273750740006e-05, "loss": 0.4714, "step": 432 }, { "epoch": 25.470588235294116, "grad_norm": 5.518557071685791, "learning_rate": 1.746661416831722e-05, "loss": 0.4595, "step": 433 }, { "epoch": 25.529411764705884, "grad_norm": 5.1790666580200195, "learning_rate": 1.7453927485100573e-05, "loss": 0.4614, "step": 434 }, { "epoch": 25.58823529411765, "grad_norm": 4.946079254150391, "learning_rate": 1.74412137471376e-05, "loss": 0.4657, "step": 435 }, { "epoch": 25.647058823529413, "grad_norm": 5.158217906951904, "learning_rate": 1.7428473000574038e-05, "loss": 0.4881, "step": 436 }, { "epoch": 25.705882352941178, "grad_norm": 5.005776882171631, "learning_rate": 1.7415705291653647e-05, "loss": 0.4714, "step": 437 }, { "epoch": 25.764705882352942, "grad_norm": 8.655595779418945, "learning_rate": 1.740291066671806e-05, "loss": 0.4083, "step": 438 }, { "epoch": 25.823529411764707, "grad_norm": 5.880197525024414, "learning_rate": 1.7390089172206594e-05, "loss": 0.5511, "step": 439 }, { "epoch": 25.88235294117647, "grad_norm": 4.598823070526123, "learning_rate": 1.7377240854656094e-05, "loss": 0.4329, "step": 440 }, { "epoch": 25.941176470588236, "grad_norm": 6.179619312286377, "learning_rate": 1.736436576070077e-05, "loss": 0.5088, "step": 441 }, { "epoch": 26.0, "grad_norm": 4.657426357269287, "learning_rate": 1.7351463937072008e-05, "loss": 0.4644, "step": 442 }, { "epoch": 26.058823529411764, "grad_norm": 4.670167922973633, "learning_rate": 1.7338535430598217e-05, "loss": 0.5668, "step": 443 }, { "epoch": 26.11764705882353, "grad_norm": 4.65447473526001, "learning_rate": 1.7325580288204652e-05, "loss": 0.4241, "step": 444 }, { "epoch": 26.176470588235293, "grad_norm": 6.5170207023620605, "learning_rate": 1.7312598556913255e-05, "loss": 0.424, "step": 445 }, { "epoch": 26.235294117647058, "grad_norm": 5.808255672454834, "learning_rate": 1.7299590283842467e-05, "loss": 0.4447, "step": 446 }, { "epoch": 26.294117647058822, "grad_norm": 4.468023300170898, "learning_rate": 1.7286555516207054e-05, "loss": 0.5769, "step": 447 }, { "epoch": 26.352941176470587, "grad_norm": 4.931513786315918, "learning_rate": 1.727349430131797e-05, "loss": 0.4418, "step": 448 }, { "epoch": 26.41176470588235, "grad_norm": 4.942298889160156, "learning_rate": 1.726040668658215e-05, "loss": 0.3018, "step": 449 }, { "epoch": 26.470588235294116, "grad_norm": 4.987689971923828, "learning_rate": 1.7247292719502343e-05, "loss": 0.3271, "step": 450 }, { "epoch": 26.529411764705884, "grad_norm": 5.471391201019287, "learning_rate": 1.7234152447676957e-05, "loss": 0.4529, "step": 451 }, { "epoch": 26.58823529411765, "grad_norm": 7.287616729736328, "learning_rate": 1.722098591879987e-05, "loss": 0.4799, "step": 452 }, { "epoch": 26.647058823529413, "grad_norm": 7.6732659339904785, "learning_rate": 1.7207793180660265e-05, "loss": 0.5143, "step": 453 }, { "epoch": 26.705882352941178, "grad_norm": 6.390470027923584, "learning_rate": 1.7194574281142455e-05, "loss": 0.3737, "step": 454 }, { "epoch": 26.764705882352942, "grad_norm": 4.518446922302246, "learning_rate": 1.7181329268225702e-05, "loss": 0.448, "step": 455 }, { "epoch": 26.823529411764707, "grad_norm": 8.37795639038086, "learning_rate": 1.716805818998406e-05, "loss": 0.5129, "step": 456 }, { "epoch": 26.88235294117647, "grad_norm": 5.824464321136475, "learning_rate": 1.7154761094586177e-05, "loss": 0.4365, "step": 457 }, { "epoch": 26.941176470588236, "grad_norm": 5.571789264678955, "learning_rate": 1.714143803029514e-05, "loss": 0.4615, "step": 458 }, { "epoch": 27.0, "grad_norm": 4.98638916015625, "learning_rate": 1.7128089045468294e-05, "loss": 0.5366, "step": 459 }, { "epoch": 27.058823529411764, "grad_norm": 4.353346824645996, "learning_rate": 1.711471418855706e-05, "loss": 0.4491, "step": 460 }, { "epoch": 27.11764705882353, "grad_norm": 5.481398105621338, "learning_rate": 1.710131350810677e-05, "loss": 0.3391, "step": 461 }, { "epoch": 27.176470588235293, "grad_norm": 5.00620174407959, "learning_rate": 1.7087887052756485e-05, "loss": 0.5167, "step": 462 }, { "epoch": 27.235294117647058, "grad_norm": 5.270884037017822, "learning_rate": 1.7074434871238804e-05, "loss": 0.2617, "step": 463 }, { "epoch": 27.294117647058822, "grad_norm": 5.8420844078063965, "learning_rate": 1.7060957012379722e-05, "loss": 0.4455, "step": 464 }, { "epoch": 27.352941176470587, "grad_norm": 6.637415885925293, "learning_rate": 1.704745352509842e-05, "loss": 0.4454, "step": 465 }, { "epoch": 27.41176470588235, "grad_norm": 4.822769641876221, "learning_rate": 1.7033924458407108e-05, "loss": 0.5365, "step": 466 }, { "epoch": 27.470588235294116, "grad_norm": 4.963967800140381, "learning_rate": 1.7020369861410826e-05, "loss": 0.4895, "step": 467 }, { "epoch": 27.529411764705884, "grad_norm": 4.569235324859619, "learning_rate": 1.700678978330729e-05, "loss": 0.4047, "step": 468 }, { "epoch": 27.58823529411765, "grad_norm": 82.57929229736328, "learning_rate": 1.6993184273386702e-05, "loss": 0.5427, "step": 469 }, { "epoch": 27.647058823529413, "grad_norm": 5.166533470153809, "learning_rate": 1.6979553381031565e-05, "loss": 0.3767, "step": 470 }, { "epoch": 27.705882352941178, "grad_norm": 6.35746431350708, "learning_rate": 1.6965897155716512e-05, "loss": 0.4034, "step": 471 }, { "epoch": 27.764705882352942, "grad_norm": 8.027579307556152, "learning_rate": 1.6952215647008134e-05, "loss": 0.5413, "step": 472 }, { "epoch": 27.823529411764707, "grad_norm": 8.453990936279297, "learning_rate": 1.693850890456477e-05, "loss": 0.3765, "step": 473 }, { "epoch": 27.88235294117647, "grad_norm": 6.1160173416137695, "learning_rate": 1.6924776978136363e-05, "loss": 0.3539, "step": 474 }, { "epoch": 27.941176470588236, "grad_norm": 6.21852445602417, "learning_rate": 1.6911019917564263e-05, "loss": 0.343, "step": 475 }, { "epoch": 28.0, "grad_norm": 5.944028854370117, "learning_rate": 1.6897237772781046e-05, "loss": 0.244, "step": 476 }, { "epoch": 28.058823529411764, "grad_norm": 5.899088382720947, "learning_rate": 1.688343059381032e-05, "loss": 0.3178, "step": 477 }, { "epoch": 28.11764705882353, "grad_norm": 5.763691425323486, "learning_rate": 1.6869598430766584e-05, "loss": 0.3676, "step": 478 }, { "epoch": 28.176470588235293, "grad_norm": 5.194077968597412, "learning_rate": 1.685574133385499e-05, "loss": 0.3947, "step": 479 }, { "epoch": 28.235294117647058, "grad_norm": 4.816842079162598, "learning_rate": 1.6841859353371212e-05, "loss": 0.3585, "step": 480 }, { "epoch": 28.294117647058822, "grad_norm": 6.153573036193848, "learning_rate": 1.682795253970123e-05, "loss": 0.4632, "step": 481 }, { "epoch": 28.352941176470587, "grad_norm": 5.5739665031433105, "learning_rate": 1.6814020943321156e-05, "loss": 0.3545, "step": 482 }, { "epoch": 28.41176470588235, "grad_norm": 4.992555618286133, "learning_rate": 1.6800064614797066e-05, "loss": 0.4654, "step": 483 }, { "epoch": 28.470588235294116, "grad_norm": 9.034517288208008, "learning_rate": 1.678608360478479e-05, "loss": 0.4031, "step": 484 }, { "epoch": 28.529411764705884, "grad_norm": 4.654303073883057, "learning_rate": 1.677207796402975e-05, "loss": 0.368, "step": 485 }, { "epoch": 28.58823529411765, "grad_norm": 6.624362945556641, "learning_rate": 1.675804774336677e-05, "loss": 0.4517, "step": 486 }, { "epoch": 28.647058823529413, "grad_norm": 4.689206600189209, "learning_rate": 1.6743992993719876e-05, "loss": 0.5558, "step": 487 }, { "epoch": 28.705882352941178, "grad_norm": 5.843018054962158, "learning_rate": 1.672991376610214e-05, "loss": 0.4532, "step": 488 }, { "epoch": 28.764705882352942, "grad_norm": 5.425209999084473, "learning_rate": 1.6715810111615467e-05, "loss": 0.2998, "step": 489 }, { "epoch": 28.823529411764707, "grad_norm": 6.044084548950195, "learning_rate": 1.6701682081450428e-05, "loss": 0.3483, "step": 490 }, { "epoch": 28.88235294117647, "grad_norm": 4.699377536773682, "learning_rate": 1.668752972688607e-05, "loss": 0.2673, "step": 491 }, { "epoch": 28.941176470588236, "grad_norm": 7.615293502807617, "learning_rate": 1.667335309928972e-05, "loss": 0.4274, "step": 492 }, { "epoch": 29.0, "grad_norm": 4.812665939331055, "learning_rate": 1.665915225011681e-05, "loss": 0.2913, "step": 493 }, { "epoch": 29.058823529411764, "grad_norm": 3.704082727432251, "learning_rate": 1.6644927230910693e-05, "loss": 0.3455, "step": 494 }, { "epoch": 29.11764705882353, "grad_norm": 5.657630920410156, "learning_rate": 1.663067809330244e-05, "loss": 0.4772, "step": 495 }, { "epoch": 29.176470588235293, "grad_norm": 5.027153015136719, "learning_rate": 1.6616404889010667e-05, "loss": 0.416, "step": 496 }, { "epoch": 29.235294117647058, "grad_norm": 5.494652271270752, "learning_rate": 1.660210766984134e-05, "loss": 0.4636, "step": 497 }, { "epoch": 29.294117647058822, "grad_norm": 5.205593109130859, "learning_rate": 1.6587786487687584e-05, "loss": 0.4618, "step": 498 }, { "epoch": 29.352941176470587, "grad_norm": 5.225488662719727, "learning_rate": 1.6573441394529517e-05, "loss": 0.4828, "step": 499 }, { "epoch": 29.41176470588235, "grad_norm": 4.562355995178223, "learning_rate": 1.6559072442434025e-05, "loss": 0.2441, "step": 500 }, { "epoch": 29.470588235294116, "grad_norm": 5.462298393249512, "learning_rate": 1.654467968355461e-05, "loss": 0.2423, "step": 501 }, { "epoch": 29.529411764705884, "grad_norm": 5.196276664733887, "learning_rate": 1.6530263170131164e-05, "loss": 0.2843, "step": 502 }, { "epoch": 29.58823529411765, "grad_norm": 4.699471950531006, "learning_rate": 1.6515822954489816e-05, "loss": 0.3852, "step": 503 }, { "epoch": 29.647058823529413, "grad_norm": 5.0923333168029785, "learning_rate": 1.650135908904271e-05, "loss": 0.3743, "step": 504 }, { "epoch": 29.705882352941178, "grad_norm": 4.132554531097412, "learning_rate": 1.6486871626287844e-05, "loss": 0.3204, "step": 505 }, { "epoch": 29.764705882352942, "grad_norm": 4.677626609802246, "learning_rate": 1.647236061880885e-05, "loss": 0.3707, "step": 506 }, { "epoch": 29.823529411764707, "grad_norm": 6.070128917694092, "learning_rate": 1.6457826119274827e-05, "loss": 0.358, "step": 507 }, { "epoch": 29.88235294117647, "grad_norm": 7.002991676330566, "learning_rate": 1.644326818044014e-05, "loss": 0.2097, "step": 508 }, { "epoch": 29.941176470588236, "grad_norm": 4.7061543464660645, "learning_rate": 1.6428686855144222e-05, "loss": 0.3562, "step": 509 }, { "epoch": 30.0, "grad_norm": 4.965969562530518, "learning_rate": 1.6414082196311402e-05, "loss": 0.4122, "step": 510 }, { "epoch": 30.058823529411764, "grad_norm": 5.196765422821045, "learning_rate": 1.6399454256950686e-05, "loss": 0.1957, "step": 511 }, { "epoch": 30.11764705882353, "grad_norm": 4.171016693115234, "learning_rate": 1.638480309015559e-05, "loss": 0.2851, "step": 512 }, { "epoch": 30.176470588235293, "grad_norm": 4.746505260467529, "learning_rate": 1.6370128749103932e-05, "loss": 0.242, "step": 513 }, { "epoch": 30.235294117647058, "grad_norm": 4.313458442687988, "learning_rate": 1.6355431287057644e-05, "loss": 0.3441, "step": 514 }, { "epoch": 30.294117647058822, "grad_norm": 4.274465084075928, "learning_rate": 1.6340710757362575e-05, "loss": 0.316, "step": 515 }, { "epoch": 30.352941176470587, "grad_norm": 5.252163410186768, "learning_rate": 1.6325967213448307e-05, "loss": 0.3383, "step": 516 }, { "epoch": 30.41176470588235, "grad_norm": 4.5995306968688965, "learning_rate": 1.6311200708827942e-05, "loss": 0.2223, "step": 517 }, { "epoch": 30.470588235294116, "grad_norm": 5.642678737640381, "learning_rate": 1.6296411297097937e-05, "loss": 0.4463, "step": 518 }, { "epoch": 30.529411764705884, "grad_norm": 6.400211334228516, "learning_rate": 1.628159903193788e-05, "loss": 0.5015, "step": 519 }, { "epoch": 30.58823529411765, "grad_norm": 4.0197062492370605, "learning_rate": 1.6266763967110307e-05, "loss": 0.3344, "step": 520 }, { "epoch": 30.647058823529413, "grad_norm": 4.563940525054932, "learning_rate": 1.6251906156460517e-05, "loss": 0.3654, "step": 521 }, { "epoch": 30.705882352941178, "grad_norm": 5.676427364349365, "learning_rate": 1.623702565391636e-05, "loss": 0.3536, "step": 522 }, { "epoch": 30.764705882352942, "grad_norm": 4.913084983825684, "learning_rate": 1.6222122513488048e-05, "loss": 0.4175, "step": 523 }, { "epoch": 30.823529411764707, "grad_norm": 4.055754661560059, "learning_rate": 1.620719678926797e-05, "loss": 0.2366, "step": 524 }, { "epoch": 30.88235294117647, "grad_norm": 5.574784278869629, "learning_rate": 1.6192248535430467e-05, "loss": 0.3025, "step": 525 }, { "epoch": 30.941176470588236, "grad_norm": 3.9612314701080322, "learning_rate": 1.617727780623167e-05, "loss": 0.2953, "step": 526 }, { "epoch": 31.0, "grad_norm": 4.898277759552002, "learning_rate": 1.6162284656009276e-05, "loss": 0.3519, "step": 527 }, { "epoch": 31.058823529411764, "grad_norm": 4.5056633949279785, "learning_rate": 1.6147269139182366e-05, "loss": 0.1923, "step": 528 }, { "epoch": 31.11764705882353, "grad_norm": 4.63279914855957, "learning_rate": 1.61322313102512e-05, "loss": 0.1866, "step": 529 }, { "epoch": 31.176470588235293, "grad_norm": 8.00849437713623, "learning_rate": 1.6117171223797027e-05, "loss": 0.2388, "step": 530 }, { "epoch": 31.235294117647058, "grad_norm": 3.9407408237457275, "learning_rate": 1.6102088934481882e-05, "loss": 0.3958, "step": 531 }, { "epoch": 31.294117647058822, "grad_norm": 5.983728408813477, "learning_rate": 1.6086984497048373e-05, "loss": 0.469, "step": 532 }, { "epoch": 31.352941176470587, "grad_norm": 4.270047664642334, "learning_rate": 1.6071857966319517e-05, "loss": 0.2697, "step": 533 }, { "epoch": 31.41176470588235, "grad_norm": 3.8567914962768555, "learning_rate": 1.6056709397198506e-05, "loss": 0.2929, "step": 534 }, { "epoch": 31.470588235294116, "grad_norm": 5.830452919006348, "learning_rate": 1.604153884466853e-05, "loss": 0.3716, "step": 535 }, { "epoch": 31.529411764705884, "grad_norm": 5.621644020080566, "learning_rate": 1.6026346363792565e-05, "loss": 0.4453, "step": 536 }, { "epoch": 31.58823529411765, "grad_norm": 3.465603828430176, "learning_rate": 1.6011132009713185e-05, "loss": 0.2151, "step": 537 }, { "epoch": 31.647058823529413, "grad_norm": 3.9897196292877197, "learning_rate": 1.599589583765235e-05, "loss": 0.2405, "step": 538 }, { "epoch": 31.705882352941178, "grad_norm": 5.2384233474731445, "learning_rate": 1.5980637902911208e-05, "loss": 0.3637, "step": 539 }, { "epoch": 31.764705882352942, "grad_norm": 4.915337085723877, "learning_rate": 1.5965358260869905e-05, "loss": 0.4701, "step": 540 }, { "epoch": 31.823529411764707, "grad_norm": 4.876436233520508, "learning_rate": 1.5950056966987363e-05, "loss": 0.2766, "step": 541 }, { "epoch": 31.88235294117647, "grad_norm": 4.610793590545654, "learning_rate": 1.5934734076801102e-05, "loss": 0.3495, "step": 542 }, { "epoch": 31.941176470588236, "grad_norm": 4.224476337432861, "learning_rate": 1.591938964592703e-05, "loss": 0.2113, "step": 543 }, { "epoch": 32.0, "grad_norm": 4.074953079223633, "learning_rate": 1.5904023730059227e-05, "loss": 0.1963, "step": 544 }, { "epoch": 32.05882352941177, "grad_norm": 4.13562536239624, "learning_rate": 1.5888636384969767e-05, "loss": 0.3495, "step": 545 }, { "epoch": 32.11764705882353, "grad_norm": 4.5545735359191895, "learning_rate": 1.5873227666508494e-05, "loss": 0.3795, "step": 546 }, { "epoch": 32.1764705882353, "grad_norm": 9.258374214172363, "learning_rate": 1.5857797630602837e-05, "loss": 0.1658, "step": 547 }, { "epoch": 32.23529411764706, "grad_norm": 8.656449317932129, "learning_rate": 1.584234633325759e-05, "loss": 0.2542, "step": 548 }, { "epoch": 32.294117647058826, "grad_norm": 6.23453426361084, "learning_rate": 1.5826873830554723e-05, "loss": 0.3903, "step": 549 }, { "epoch": 32.35294117647059, "grad_norm": 4.170133113861084, "learning_rate": 1.5811380178653175e-05, "loss": 0.2707, "step": 550 }, { "epoch": 32.411764705882355, "grad_norm": 4.32157564163208, "learning_rate": 1.5795865433788642e-05, "loss": 0.3526, "step": 551 }, { "epoch": 32.470588235294116, "grad_norm": 3.6629137992858887, "learning_rate": 1.5780329652273386e-05, "loss": 0.3269, "step": 552 }, { "epoch": 32.529411764705884, "grad_norm": 4.425212860107422, "learning_rate": 1.576477289049601e-05, "loss": 0.2798, "step": 553 }, { "epoch": 32.588235294117645, "grad_norm": 5.334105968475342, "learning_rate": 1.5749195204921286e-05, "loss": 0.438, "step": 554 }, { "epoch": 32.64705882352941, "grad_norm": 5.580854892730713, "learning_rate": 1.5733596652089914e-05, "loss": 0.2063, "step": 555 }, { "epoch": 32.705882352941174, "grad_norm": 4.344142436981201, "learning_rate": 1.571797728861835e-05, "loss": 0.15, "step": 556 }, { "epoch": 32.76470588235294, "grad_norm": 4.815372467041016, "learning_rate": 1.570233717119857e-05, "loss": 0.1983, "step": 557 }, { "epoch": 32.8235294117647, "grad_norm": 4.637873649597168, "learning_rate": 1.568667635659788e-05, "loss": 0.3187, "step": 558 }, { "epoch": 32.88235294117647, "grad_norm": 4.572510242462158, "learning_rate": 1.5670994901658715e-05, "loss": 0.2953, "step": 559 }, { "epoch": 32.94117647058823, "grad_norm": 3.892313241958618, "learning_rate": 1.5655292863298427e-05, "loss": 0.2724, "step": 560 }, { "epoch": 33.0, "grad_norm": 3.8735897541046143, "learning_rate": 1.5639570298509067e-05, "loss": 0.3566, "step": 561 }, { "epoch": 33.05882352941177, "grad_norm": 4.2727179527282715, "learning_rate": 1.5623827264357196e-05, "loss": 0.2533, "step": 562 }, { "epoch": 33.11764705882353, "grad_norm": 4.017293453216553, "learning_rate": 1.5608063817983677e-05, "loss": 0.2967, "step": 563 }, { "epoch": 33.1764705882353, "grad_norm": 3.6584606170654297, "learning_rate": 1.5592280016603447e-05, "loss": 0.195, "step": 564 }, { "epoch": 33.23529411764706, "grad_norm": 4.944647789001465, "learning_rate": 1.5576475917505334e-05, "loss": 0.2338, "step": 565 }, { "epoch": 33.294117647058826, "grad_norm": 3.6905765533447266, "learning_rate": 1.5560651578051832e-05, "loss": 0.1309, "step": 566 }, { "epoch": 33.35294117647059, "grad_norm": 5.513548851013184, "learning_rate": 1.55448070556789e-05, "loss": 0.2909, "step": 567 }, { "epoch": 33.411764705882355, "grad_norm": 3.742478609085083, "learning_rate": 1.552894240789576e-05, "loss": 0.1861, "step": 568 }, { "epoch": 33.470588235294116, "grad_norm": 5.1291279792785645, "learning_rate": 1.5513057692284678e-05, "loss": 0.3064, "step": 569 }, { "epoch": 33.529411764705884, "grad_norm": 4.592774391174316, "learning_rate": 1.549715296650075e-05, "loss": 0.3073, "step": 570 }, { "epoch": 33.588235294117645, "grad_norm": 4.3770575523376465, "learning_rate": 1.5481228288271712e-05, "loss": 0.261, "step": 571 }, { "epoch": 33.64705882352941, "grad_norm": 5.329732894897461, "learning_rate": 1.546528371539771e-05, "loss": 0.2621, "step": 572 }, { "epoch": 33.705882352941174, "grad_norm": 37.09141159057617, "learning_rate": 1.5449319305751108e-05, "loss": 0.3212, "step": 573 }, { "epoch": 33.76470588235294, "grad_norm": 4.064065933227539, "learning_rate": 1.543333511727626e-05, "loss": 0.3447, "step": 574 }, { "epoch": 33.8235294117647, "grad_norm": 3.8767833709716797, "learning_rate": 1.541733120798932e-05, "loss": 0.2397, "step": 575 }, { "epoch": 33.88235294117647, "grad_norm": 7.212404251098633, "learning_rate": 1.5401307635978016e-05, "loss": 0.4142, "step": 576 }, { "epoch": 33.94117647058823, "grad_norm": 4.509988784790039, "learning_rate": 1.5385264459401436e-05, "loss": 0.3455, "step": 577 }, { "epoch": 34.0, "grad_norm": 4.719954490661621, "learning_rate": 1.536920173648984e-05, "loss": 0.1841, "step": 578 }, { "epoch": 34.05882352941177, "grad_norm": 4.27302885055542, "learning_rate": 1.5353119525544417e-05, "loss": 0.2643, "step": 579 }, { "epoch": 34.11764705882353, "grad_norm": 3.6128334999084473, "learning_rate": 1.53370178849371e-05, "loss": 0.2271, "step": 580 }, { "epoch": 34.1764705882353, "grad_norm": 4.146597385406494, "learning_rate": 1.532089687311035e-05, "loss": 0.3118, "step": 581 }, { "epoch": 34.23529411764706, "grad_norm": 11.791861534118652, "learning_rate": 1.5304756548576913e-05, "loss": 0.2415, "step": 582 }, { "epoch": 34.294117647058826, "grad_norm": 3.8106744289398193, "learning_rate": 1.5288596969919664e-05, "loss": 0.247, "step": 583 }, { "epoch": 34.35294117647059, "grad_norm": 5.933708667755127, "learning_rate": 1.5272418195791336e-05, "loss": 0.2404, "step": 584 }, { "epoch": 34.411764705882355, "grad_norm": 4.407947063446045, "learning_rate": 1.525622028491435e-05, "loss": 0.1923, "step": 585 }, { "epoch": 34.470588235294116, "grad_norm": 3.8750479221343994, "learning_rate": 1.5240003296080578e-05, "loss": 0.209, "step": 586 }, { "epoch": 34.529411764705884, "grad_norm": 5.120879173278809, "learning_rate": 1.5223767288151145e-05, "loss": 0.2529, "step": 587 }, { "epoch": 34.588235294117645, "grad_norm": 4.206233024597168, "learning_rate": 1.5207512320056198e-05, "loss": 0.3183, "step": 588 }, { "epoch": 34.64705882352941, "grad_norm": 3.9487502574920654, "learning_rate": 1.5191238450794706e-05, "loss": 0.2541, "step": 589 }, { "epoch": 34.705882352941174, "grad_norm": 5.9402241706848145, "learning_rate": 1.5174945739434244e-05, "loss": 0.2019, "step": 590 }, { "epoch": 34.76470588235294, "grad_norm": 4.436616897583008, "learning_rate": 1.5158634245110771e-05, "loss": 0.2347, "step": 591 }, { "epoch": 34.8235294117647, "grad_norm": 4.340810775756836, "learning_rate": 1.5142304027028427e-05, "loss": 0.3582, "step": 592 }, { "epoch": 34.88235294117647, "grad_norm": 3.9307312965393066, "learning_rate": 1.5125955144459308e-05, "loss": 0.1599, "step": 593 }, { "epoch": 34.94117647058823, "grad_norm": 4.346500873565674, "learning_rate": 1.5109587656743251e-05, "loss": 0.441, "step": 594 }, { "epoch": 35.0, "grad_norm": 4.473658561706543, "learning_rate": 1.5093201623287631e-05, "loss": 0.2399, "step": 595 }, { "epoch": 35.05882352941177, "grad_norm": 3.3592734336853027, "learning_rate": 1.5076797103567131e-05, "loss": 0.1241, "step": 596 }, { "epoch": 35.11764705882353, "grad_norm": 4.331708908081055, "learning_rate": 1.5060374157123527e-05, "loss": 0.3073, "step": 597 }, { "epoch": 35.1764705882353, "grad_norm": 3.371270179748535, "learning_rate": 1.5043932843565485e-05, "loss": 0.2841, "step": 598 }, { "epoch": 35.23529411764706, "grad_norm": 3.6429834365844727, "learning_rate": 1.5027473222568336e-05, "loss": 0.3034, "step": 599 }, { "epoch": 35.294117647058826, "grad_norm": 3.636260747909546, "learning_rate": 1.5010995353873852e-05, "loss": 0.1549, "step": 600 }, { "epoch": 35.35294117647059, "grad_norm": 3.7058522701263428, "learning_rate": 1.4994499297290046e-05, "loss": 0.2394, "step": 601 }, { "epoch": 35.411764705882355, "grad_norm": 3.8854284286499023, "learning_rate": 1.4977985112690936e-05, "loss": 0.1912, "step": 602 }, { "epoch": 35.470588235294116, "grad_norm": 3.282867431640625, "learning_rate": 1.4961452860016347e-05, "loss": 0.3225, "step": 603 }, { "epoch": 35.529411764705884, "grad_norm": 3.587818145751953, "learning_rate": 1.4944902599271675e-05, "loss": 0.2312, "step": 604 }, { "epoch": 35.588235294117645, "grad_norm": 3.0545263290405273, "learning_rate": 1.4928334390527683e-05, "loss": 0.2369, "step": 605 }, { "epoch": 35.64705882352941, "grad_norm": 3.800175666809082, "learning_rate": 1.4911748293920281e-05, "loss": 0.2782, "step": 606 }, { "epoch": 35.705882352941174, "grad_norm": 3.427614212036133, "learning_rate": 1.4895144369650299e-05, "loss": 0.1412, "step": 607 }, { "epoch": 35.76470588235294, "grad_norm": 4.640096664428711, "learning_rate": 1.4878522677983277e-05, "loss": 0.2326, "step": 608 }, { "epoch": 35.8235294117647, "grad_norm": 4.529035568237305, "learning_rate": 1.486188327924924e-05, "loss": 0.2534, "step": 609 }, { "epoch": 35.88235294117647, "grad_norm": 5.117388725280762, "learning_rate": 1.4845226233842487e-05, "loss": 0.2369, "step": 610 }, { "epoch": 35.94117647058823, "grad_norm": 5.048950672149658, "learning_rate": 1.4828551602221364e-05, "loss": 0.3405, "step": 611 }, { "epoch": 36.0, "grad_norm": 10.957911491394043, "learning_rate": 1.4811859444908053e-05, "loss": 0.2218, "step": 612 }, { "epoch": 36.05882352941177, "grad_norm": 4.336139678955078, "learning_rate": 1.4795149822488337e-05, "loss": 0.2045, "step": 613 }, { "epoch": 36.11764705882353, "grad_norm": 3.457240343093872, "learning_rate": 1.4778422795611404e-05, "loss": 0.2466, "step": 614 }, { "epoch": 36.1764705882353, "grad_norm": 3.9533274173736572, "learning_rate": 1.4761678424989606e-05, "loss": 0.2043, "step": 615 }, { "epoch": 36.23529411764706, "grad_norm": 3.224555730819702, "learning_rate": 1.4744916771398237e-05, "loss": 0.216, "step": 616 }, { "epoch": 36.294117647058826, "grad_norm": 8.987381935119629, "learning_rate": 1.4728137895675339e-05, "loss": 0.2998, "step": 617 }, { "epoch": 36.35294117647059, "grad_norm": 3.4630818367004395, "learning_rate": 1.4711341858721448e-05, "loss": 0.1846, "step": 618 }, { "epoch": 36.411764705882355, "grad_norm": 3.012936592102051, "learning_rate": 1.46945287214994e-05, "loss": 0.1472, "step": 619 }, { "epoch": 36.470588235294116, "grad_norm": 3.5640037059783936, "learning_rate": 1.4677698545034087e-05, "loss": 0.1681, "step": 620 }, { "epoch": 36.529411764705884, "grad_norm": 3.9296510219573975, "learning_rate": 1.466085139041226e-05, "loss": 0.2553, "step": 621 }, { "epoch": 36.588235294117645, "grad_norm": 4.233677387237549, "learning_rate": 1.464398731878228e-05, "loss": 0.2387, "step": 622 }, { "epoch": 36.64705882352941, "grad_norm": 3.1741223335266113, "learning_rate": 1.4627106391353918e-05, "loss": 0.138, "step": 623 }, { "epoch": 36.705882352941174, "grad_norm": 3.2893216609954834, "learning_rate": 1.461020866939812e-05, "loss": 0.2366, "step": 624 }, { "epoch": 36.76470588235294, "grad_norm": 4.569036483764648, "learning_rate": 1.4593294214246798e-05, "loss": 0.2088, "step": 625 }, { "epoch": 36.8235294117647, "grad_norm": 3.6140382289886475, "learning_rate": 1.4576363087292587e-05, "loss": 0.2144, "step": 626 }, { "epoch": 36.88235294117647, "grad_norm": 3.5178050994873047, "learning_rate": 1.4559415349988643e-05, "loss": 0.3058, "step": 627 }, { "epoch": 36.94117647058823, "grad_norm": 3.602147102355957, "learning_rate": 1.4542451063848411e-05, "loss": 0.4041, "step": 628 }, { "epoch": 37.0, "grad_norm": 4.408696174621582, "learning_rate": 1.4525470290445392e-05, "loss": 0.1579, "step": 629 }, { "epoch": 37.05882352941177, "grad_norm": 3.156903028488159, "learning_rate": 1.4508473091412935e-05, "loss": 0.2741, "step": 630 }, { "epoch": 37.11764705882353, "grad_norm": 3.8585472106933594, "learning_rate": 1.4491459528444011e-05, "loss": 0.2459, "step": 631 }, { "epoch": 37.1764705882353, "grad_norm": 3.894097089767456, "learning_rate": 1.4474429663290983e-05, "loss": 0.308, "step": 632 }, { "epoch": 37.23529411764706, "grad_norm": 4.551866054534912, "learning_rate": 1.4457383557765385e-05, "loss": 0.2238, "step": 633 }, { "epoch": 37.294117647058826, "grad_norm": 4.380842685699463, "learning_rate": 1.444032127373769e-05, "loss": 0.3234, "step": 634 }, { "epoch": 37.35294117647059, "grad_norm": 3.871978998184204, "learning_rate": 1.44232428731371e-05, "loss": 0.2245, "step": 635 }, { "epoch": 37.411764705882355, "grad_norm": 3.5235583782196045, "learning_rate": 1.4406148417951315e-05, "loss": 0.1887, "step": 636 }, { "epoch": 37.470588235294116, "grad_norm": 3.483611583709717, "learning_rate": 1.4389037970226299e-05, "loss": 0.1759, "step": 637 }, { "epoch": 37.529411764705884, "grad_norm": 4.1474690437316895, "learning_rate": 1.437191159206607e-05, "loss": 0.1764, "step": 638 }, { "epoch": 37.588235294117645, "grad_norm": 3.341463804244995, "learning_rate": 1.435476934563246e-05, "loss": 0.1015, "step": 639 }, { "epoch": 37.64705882352941, "grad_norm": 3.206871271133423, "learning_rate": 1.4337611293144908e-05, "loss": 0.175, "step": 640 }, { "epoch": 37.705882352941174, "grad_norm": 3.6282169818878174, "learning_rate": 1.4320437496880207e-05, "loss": 0.1729, "step": 641 }, { "epoch": 37.76470588235294, "grad_norm": 3.1944828033447266, "learning_rate": 1.4303248019172303e-05, "loss": 0.1908, "step": 642 }, { "epoch": 37.8235294117647, "grad_norm": 3.699207305908203, "learning_rate": 1.4286042922412058e-05, "loss": 0.1939, "step": 643 }, { "epoch": 37.88235294117647, "grad_norm": 24.917417526245117, "learning_rate": 1.4268822269047027e-05, "loss": 0.2807, "step": 644 }, { "epoch": 37.94117647058823, "grad_norm": 4.009507656097412, "learning_rate": 1.4251586121581223e-05, "loss": 0.1555, "step": 645 }, { "epoch": 38.0, "grad_norm": 3.62250018119812, "learning_rate": 1.4234334542574906e-05, "loss": 0.3549, "step": 646 }, { "epoch": 38.05882352941177, "grad_norm": 2.898505687713623, "learning_rate": 1.4217067594644332e-05, "loss": 0.1119, "step": 647 }, { "epoch": 38.11764705882353, "grad_norm": 3.335519552230835, "learning_rate": 1.4199785340461554e-05, "loss": 0.1921, "step": 648 }, { "epoch": 38.1764705882353, "grad_norm": 3.5431151390075684, "learning_rate": 1.4182487842754173e-05, "loss": 0.1925, "step": 649 }, { "epoch": 38.23529411764706, "grad_norm": 3.1161081790924072, "learning_rate": 1.4165175164305121e-05, "loss": 0.2345, "step": 650 }, { "epoch": 38.294117647058826, "grad_norm": 4.612059116363525, "learning_rate": 1.4147847367952427e-05, "loss": 0.1866, "step": 651 }, { "epoch": 38.35294117647059, "grad_norm": 3.239628314971924, "learning_rate": 1.4130504516588997e-05, "loss": 0.3085, "step": 652 }, { "epoch": 38.411764705882355, "grad_norm": 3.7442405223846436, "learning_rate": 1.4113146673162375e-05, "loss": 0.2696, "step": 653 }, { "epoch": 38.470588235294116, "grad_norm": 5.512420177459717, "learning_rate": 1.4095773900674518e-05, "loss": 0.2321, "step": 654 }, { "epoch": 38.529411764705884, "grad_norm": 3.77240252494812, "learning_rate": 1.4078386262181579e-05, "loss": 0.1648, "step": 655 }, { "epoch": 38.588235294117645, "grad_norm": 4.740634918212891, "learning_rate": 1.406098382079366e-05, "loss": 0.1506, "step": 656 }, { "epoch": 38.64705882352941, "grad_norm": 2.805983304977417, "learning_rate": 1.4043566639674597e-05, "loss": 0.2052, "step": 657 }, { "epoch": 38.705882352941174, "grad_norm": 4.29941463470459, "learning_rate": 1.4026134782041728e-05, "loss": 0.2502, "step": 658 }, { "epoch": 38.76470588235294, "grad_norm": 3.9916114807128906, "learning_rate": 1.4008688311165646e-05, "loss": 0.1914, "step": 659 }, { "epoch": 38.8235294117647, "grad_norm": 2.982227087020874, "learning_rate": 1.3991227290369997e-05, "loss": 0.1967, "step": 660 }, { "epoch": 38.88235294117647, "grad_norm": 4.20482063293457, "learning_rate": 1.397375178303124e-05, "loss": 0.238, "step": 661 }, { "epoch": 38.94117647058823, "grad_norm": 5.036376953125, "learning_rate": 1.3956261852578403e-05, "loss": 0.1915, "step": 662 }, { "epoch": 39.0, "grad_norm": 3.5503692626953125, "learning_rate": 1.3938757562492873e-05, "loss": 0.1453, "step": 663 }, { "epoch": 39.05882352941177, "grad_norm": 3.7864937782287598, "learning_rate": 1.3921238976308151e-05, "loss": 0.2633, "step": 664 }, { "epoch": 39.11764705882353, "grad_norm": 2.848403215408325, "learning_rate": 1.3903706157609633e-05, "loss": 0.2141, "step": 665 }, { "epoch": 39.1764705882353, "grad_norm": 3.2796549797058105, "learning_rate": 1.3886159170034365e-05, "loss": 0.1454, "step": 666 }, { "epoch": 39.23529411764706, "grad_norm": 3.843458414077759, "learning_rate": 1.3868598077270829e-05, "loss": 0.25, "step": 667 }, { "epoch": 39.294117647058826, "grad_norm": 2.7658538818359375, "learning_rate": 1.3851022943058698e-05, "loss": 0.1524, "step": 668 }, { "epoch": 39.35294117647059, "grad_norm": 3.7598276138305664, "learning_rate": 1.3833433831188612e-05, "loss": 0.1693, "step": 669 }, { "epoch": 39.411764705882355, "grad_norm": 2.957955837249756, "learning_rate": 1.3815830805501942e-05, "loss": 0.1882, "step": 670 }, { "epoch": 39.470588235294116, "grad_norm": 7.283099174499512, "learning_rate": 1.3798213929890562e-05, "loss": 0.1844, "step": 671 }, { "epoch": 39.529411764705884, "grad_norm": 4.275941371917725, "learning_rate": 1.3780583268296617e-05, "loss": 0.1792, "step": 672 }, { "epoch": 39.588235294117645, "grad_norm": 3.928623676300049, "learning_rate": 1.3762938884712285e-05, "loss": 0.2135, "step": 673 }, { "epoch": 39.64705882352941, "grad_norm": 2.9317753314971924, "learning_rate": 1.3745280843179554e-05, "loss": 0.1924, "step": 674 }, { "epoch": 39.705882352941174, "grad_norm": 3.715703010559082, "learning_rate": 1.3727609207789982e-05, "loss": 0.235, "step": 675 }, { "epoch": 39.76470588235294, "grad_norm": 2.9217491149902344, "learning_rate": 1.3709924042684473e-05, "loss": 0.1335, "step": 676 }, { "epoch": 39.8235294117647, "grad_norm": 3.7185444831848145, "learning_rate": 1.3692225412053027e-05, "loss": 0.2034, "step": 677 }, { "epoch": 39.88235294117647, "grad_norm": 3.603661060333252, "learning_rate": 1.3674513380134532e-05, "loss": 0.2376, "step": 678 }, { "epoch": 39.94117647058823, "grad_norm": 3.837702751159668, "learning_rate": 1.3656788011216503e-05, "loss": 0.1862, "step": 679 }, { "epoch": 40.0, "grad_norm": 3.1096138954162598, "learning_rate": 1.3639049369634878e-05, "loss": 0.198, "step": 680 }, { "epoch": 40.05882352941177, "grad_norm": 3.917692184448242, "learning_rate": 1.3621297519773755e-05, "loss": 0.1597, "step": 681 }, { "epoch": 40.11764705882353, "grad_norm": 2.4184134006500244, "learning_rate": 1.3603532526065184e-05, "loss": 0.138, "step": 682 }, { "epoch": 40.1764705882353, "grad_norm": 4.614278316497803, "learning_rate": 1.3585754452988914e-05, "loss": 0.3387, "step": 683 }, { "epoch": 40.23529411764706, "grad_norm": 3.1114695072174072, "learning_rate": 1.3567963365072174e-05, "loss": 0.1845, "step": 684 }, { "epoch": 40.294117647058826, "grad_norm": 2.923638343811035, "learning_rate": 1.3550159326889426e-05, "loss": 0.2191, "step": 685 }, { "epoch": 40.35294117647059, "grad_norm": 3.0505127906799316, "learning_rate": 1.3532342403062138e-05, "loss": 0.1179, "step": 686 }, { "epoch": 40.411764705882355, "grad_norm": 2.8519704341888428, "learning_rate": 1.3514512658258545e-05, "loss": 0.2626, "step": 687 }, { "epoch": 40.470588235294116, "grad_norm": 3.1810107231140137, "learning_rate": 1.3496670157193425e-05, "loss": 0.1454, "step": 688 }, { "epoch": 40.529411764705884, "grad_norm": 4.166629791259766, "learning_rate": 1.3478814964627846e-05, "loss": 0.2383, "step": 689 }, { "epoch": 40.588235294117645, "grad_norm": 4.075011253356934, "learning_rate": 1.346094714536895e-05, "loss": 0.1079, "step": 690 }, { "epoch": 40.64705882352941, "grad_norm": 4.179084777832031, "learning_rate": 1.34430667642697e-05, "loss": 0.1194, "step": 691 }, { "epoch": 40.705882352941174, "grad_norm": 3.5015757083892822, "learning_rate": 1.3425173886228664e-05, "loss": 0.1668, "step": 692 }, { "epoch": 40.76470588235294, "grad_norm": 2.46016788482666, "learning_rate": 1.3407268576189756e-05, "loss": 0.0984, "step": 693 }, { "epoch": 40.8235294117647, "grad_norm": 3.1524672508239746, "learning_rate": 1.3389350899142022e-05, "loss": 0.163, "step": 694 }, { "epoch": 40.88235294117647, "grad_norm": 3.164720058441162, "learning_rate": 1.3371420920119395e-05, "loss": 0.2075, "step": 695 }, { "epoch": 40.94117647058823, "grad_norm": 3.6302990913391113, "learning_rate": 1.3353478704200458e-05, "loss": 0.1534, "step": 696 }, { "epoch": 41.0, "grad_norm": 3.584653377532959, "learning_rate": 1.3335524316508208e-05, "loss": 0.1742, "step": 697 }, { "epoch": 41.05882352941177, "grad_norm": 3.02028751373291, "learning_rate": 1.3317557822209818e-05, "loss": 0.2419, "step": 698 }, { "epoch": 41.11764705882353, "grad_norm": 3.3798162937164307, "learning_rate": 1.3299579286516409e-05, "loss": 0.1523, "step": 699 }, { "epoch": 41.1764705882353, "grad_norm": 2.9079980850219727, "learning_rate": 1.3281588774682804e-05, "loss": 0.2293, "step": 700 }, { "epoch": 41.23529411764706, "grad_norm": 3.9510550498962402, "learning_rate": 1.3263586352007294e-05, "loss": 0.1539, "step": 701 }, { "epoch": 41.294117647058826, "grad_norm": 3.4161131381988525, "learning_rate": 1.3245572083831402e-05, "loss": 0.1538, "step": 702 }, { "epoch": 41.35294117647059, "grad_norm": 14.705596923828125, "learning_rate": 1.3227546035539652e-05, "loss": 0.2016, "step": 703 }, { "epoch": 41.411764705882355, "grad_norm": 4.979067325592041, "learning_rate": 1.320950827255931e-05, "loss": 0.1125, "step": 704 }, { "epoch": 41.470588235294116, "grad_norm": 2.5300991535186768, "learning_rate": 1.3191458860360176e-05, "loss": 0.1395, "step": 705 }, { "epoch": 41.529411764705884, "grad_norm": 4.824472427368164, "learning_rate": 1.3173397864454325e-05, "loss": 0.2028, "step": 706 }, { "epoch": 41.588235294117645, "grad_norm": 3.3634023666381836, "learning_rate": 1.315532535039588e-05, "loss": 0.1232, "step": 707 }, { "epoch": 41.64705882352941, "grad_norm": 3.29884934425354, "learning_rate": 1.3137241383780766e-05, "loss": 0.1619, "step": 708 }, { "epoch": 41.705882352941174, "grad_norm": 3.1915836334228516, "learning_rate": 1.3119146030246476e-05, "loss": 0.2622, "step": 709 }, { "epoch": 41.76470588235294, "grad_norm": 11.620856285095215, "learning_rate": 1.310103935547184e-05, "loss": 0.156, "step": 710 }, { "epoch": 41.8235294117647, "grad_norm": 2.7301971912384033, "learning_rate": 1.3082921425176774e-05, "loss": 0.1486, "step": 711 }, { "epoch": 41.88235294117647, "grad_norm": 10.548784255981445, "learning_rate": 1.3064792305122042e-05, "loss": 0.2195, "step": 712 }, { "epoch": 41.94117647058823, "grad_norm": 2.8721768856048584, "learning_rate": 1.3046652061109035e-05, "loss": 0.162, "step": 713 }, { "epoch": 42.0, "grad_norm": 2.8617031574249268, "learning_rate": 1.3028500758979507e-05, "loss": 0.1498, "step": 714 }, { "epoch": 42.05882352941177, "grad_norm": 2.9735164642333984, "learning_rate": 1.301033846461536e-05, "loss": 0.215, "step": 715 }, { "epoch": 42.11764705882353, "grad_norm": 3.0693187713623047, "learning_rate": 1.2992165243938382e-05, "loss": 0.1024, "step": 716 }, { "epoch": 42.1764705882353, "grad_norm": 4.039045810699463, "learning_rate": 1.2973981162910022e-05, "loss": 0.1342, "step": 717 }, { "epoch": 42.23529411764706, "grad_norm": 2.8853251934051514, "learning_rate": 1.2955786287531159e-05, "loss": 0.1201, "step": 718 }, { "epoch": 42.294117647058826, "grad_norm": 3.8681468963623047, "learning_rate": 1.2937580683841833e-05, "loss": 0.2448, "step": 719 }, { "epoch": 42.35294117647059, "grad_norm": 4.289619445800781, "learning_rate": 1.2919364417921037e-05, "loss": 0.2314, "step": 720 }, { "epoch": 42.411764705882355, "grad_norm": 3.081756114959717, "learning_rate": 1.2901137555886458e-05, "loss": 0.19, "step": 721 }, { "epoch": 42.470588235294116, "grad_norm": 2.942272186279297, "learning_rate": 1.288290016389425e-05, "loss": 0.1879, "step": 722 }, { "epoch": 42.529411764705884, "grad_norm": 3.345478057861328, "learning_rate": 1.286465230813877e-05, "loss": 0.1525, "step": 723 }, { "epoch": 42.588235294117645, "grad_norm": 3.389369010925293, "learning_rate": 1.2846394054852371e-05, "loss": 0.1288, "step": 724 }, { "epoch": 42.64705882352941, "grad_norm": 4.5362772941589355, "learning_rate": 1.2828125470305137e-05, "loss": 0.2826, "step": 725 }, { "epoch": 42.705882352941174, "grad_norm": 3.73178768157959, "learning_rate": 1.2809846620804652e-05, "loss": 0.2115, "step": 726 }, { "epoch": 42.76470588235294, "grad_norm": 4.66780948638916, "learning_rate": 1.2791557572695756e-05, "loss": 0.1279, "step": 727 }, { "epoch": 42.8235294117647, "grad_norm": 3.3652167320251465, "learning_rate": 1.277325839236031e-05, "loss": 0.1598, "step": 728 }, { "epoch": 42.88235294117647, "grad_norm": 6.168628215789795, "learning_rate": 1.2754949146216946e-05, "loss": 0.1214, "step": 729 }, { "epoch": 42.94117647058823, "grad_norm": 3.570805072784424, "learning_rate": 1.2736629900720832e-05, "loss": 0.1207, "step": 730 }, { "epoch": 43.0, "grad_norm": 3.403740406036377, "learning_rate": 1.2718300722363431e-05, "loss": 0.1995, "step": 731 }, { "epoch": 43.05882352941177, "grad_norm": 3.3914742469787598, "learning_rate": 1.2699961677672259e-05, "loss": 0.1348, "step": 732 }, { "epoch": 43.11764705882353, "grad_norm": 4.28464412689209, "learning_rate": 1.2681612833210642e-05, "loss": 0.125, "step": 733 }, { "epoch": 43.1764705882353, "grad_norm": 2.766533851623535, "learning_rate": 1.266325425557747e-05, "loss": 0.1462, "step": 734 }, { "epoch": 43.23529411764706, "grad_norm": 2.826220989227295, "learning_rate": 1.2644886011406972e-05, "loss": 0.1534, "step": 735 }, { "epoch": 43.294117647058826, "grad_norm": 3.130544662475586, "learning_rate": 1.2626508167368452e-05, "loss": 0.1013, "step": 736 }, { "epoch": 43.35294117647059, "grad_norm": 12.231790542602539, "learning_rate": 1.2608120790166062e-05, "loss": 0.246, "step": 737 }, { "epoch": 43.411764705882355, "grad_norm": 3.840749979019165, "learning_rate": 1.2589723946538556e-05, "loss": 0.2015, "step": 738 }, { "epoch": 43.470588235294116, "grad_norm": 3.0935707092285156, "learning_rate": 1.2571317703259048e-05, "loss": 0.144, "step": 739 }, { "epoch": 43.529411764705884, "grad_norm": 3.8420681953430176, "learning_rate": 1.2552902127134765e-05, "loss": 0.148, "step": 740 }, { "epoch": 43.588235294117645, "grad_norm": 3.537487268447876, "learning_rate": 1.2534477285006807e-05, "loss": 0.1996, "step": 741 }, { "epoch": 43.64705882352941, "grad_norm": 5.062342166900635, "learning_rate": 1.2516043243749918e-05, "loss": 0.2358, "step": 742 }, { "epoch": 43.705882352941174, "grad_norm": 4.727156162261963, "learning_rate": 1.2497600070272214e-05, "loss": 0.0998, "step": 743 }, { "epoch": 43.76470588235294, "grad_norm": 3.5581183433532715, "learning_rate": 1.2479147831514972e-05, "loss": 0.2186, "step": 744 }, { "epoch": 43.8235294117647, "grad_norm": 2.820336103439331, "learning_rate": 1.2460686594452365e-05, "loss": 0.1274, "step": 745 }, { "epoch": 43.88235294117647, "grad_norm": 3.7229201793670654, "learning_rate": 1.2442216426091226e-05, "loss": 0.1918, "step": 746 }, { "epoch": 43.94117647058823, "grad_norm": 2.613948106765747, "learning_rate": 1.242373739347081e-05, "loss": 0.1221, "step": 747 }, { "epoch": 44.0, "grad_norm": 13.36833667755127, "learning_rate": 1.2405249563662539e-05, "loss": 0.1524, "step": 748 }, { "epoch": 44.05882352941177, "grad_norm": 5.536099910736084, "learning_rate": 1.2386753003769767e-05, "loss": 0.1867, "step": 749 }, { "epoch": 44.11764705882353, "grad_norm": 3.1664369106292725, "learning_rate": 1.236824778092754e-05, "loss": 0.1768, "step": 750 }, { "epoch": 44.1764705882353, "grad_norm": 18.664636611938477, "learning_rate": 1.2349733962302342e-05, "loss": 0.1936, "step": 751 }, { "epoch": 44.23529411764706, "grad_norm": 2.8296773433685303, "learning_rate": 1.2331211615091857e-05, "loss": 0.1758, "step": 752 }, { "epoch": 44.294117647058826, "grad_norm": 2.6604557037353516, "learning_rate": 1.2312680806524722e-05, "loss": 0.0895, "step": 753 }, { "epoch": 44.35294117647059, "grad_norm": 9.638259887695312, "learning_rate": 1.2294141603860295e-05, "loss": 0.1657, "step": 754 }, { "epoch": 44.411764705882355, "grad_norm": 6.132715702056885, "learning_rate": 1.2275594074388387e-05, "loss": 0.1009, "step": 755 }, { "epoch": 44.470588235294116, "grad_norm": 4.741987228393555, "learning_rate": 1.2257038285429041e-05, "loss": 0.2758, "step": 756 }, { "epoch": 44.529411764705884, "grad_norm": 3.7568678855895996, "learning_rate": 1.2238474304332277e-05, "loss": 0.2515, "step": 757 }, { "epoch": 44.588235294117645, "grad_norm": 3.5362603664398193, "learning_rate": 1.2219902198477851e-05, "loss": 0.1166, "step": 758 }, { "epoch": 44.64705882352941, "grad_norm": 3.214996337890625, "learning_rate": 1.2201322035275005e-05, "loss": 0.1258, "step": 759 }, { "epoch": 44.705882352941174, "grad_norm": 3.9752237796783447, "learning_rate": 1.2182733882162227e-05, "loss": 0.1675, "step": 760 }, { "epoch": 44.76470588235294, "grad_norm": 3.5385589599609375, "learning_rate": 1.2164137806607006e-05, "loss": 0.1929, "step": 761 }, { "epoch": 44.8235294117647, "grad_norm": 4.829566478729248, "learning_rate": 1.2145533876105586e-05, "loss": 0.1443, "step": 762 }, { "epoch": 44.88235294117647, "grad_norm": 3.451157331466675, "learning_rate": 1.2126922158182721e-05, "loss": 0.0984, "step": 763 }, { "epoch": 44.94117647058823, "grad_norm": 3.46059250831604, "learning_rate": 1.210830272039143e-05, "loss": 0.1598, "step": 764 }, { "epoch": 45.0, "grad_norm": 2.716735601425171, "learning_rate": 1.2089675630312755e-05, "loss": 0.1143, "step": 765 }, { "epoch": 45.05882352941177, "grad_norm": 2.971785068511963, "learning_rate": 1.2071040955555503e-05, "loss": 0.1551, "step": 766 }, { "epoch": 45.11764705882353, "grad_norm": 2.9231116771698, "learning_rate": 1.2052398763756022e-05, "loss": 0.1731, "step": 767 }, { "epoch": 45.1764705882353, "grad_norm": 2.712475538253784, "learning_rate": 1.2033749122577938e-05, "loss": 0.1192, "step": 768 }, { "epoch": 45.23529411764706, "grad_norm": 3.0825273990631104, "learning_rate": 1.2015092099711918e-05, "loss": 0.1614, "step": 769 }, { "epoch": 45.294117647058826, "grad_norm": 3.441948890686035, "learning_rate": 1.199642776287542e-05, "loss": 0.1874, "step": 770 }, { "epoch": 45.35294117647059, "grad_norm": 3.8023180961608887, "learning_rate": 1.1977756179812447e-05, "loss": 0.1723, "step": 771 }, { "epoch": 45.411764705882355, "grad_norm": 3.400852918624878, "learning_rate": 1.1959077418293308e-05, "loss": 0.2466, "step": 772 }, { "epoch": 45.470588235294116, "grad_norm": 3.1584246158599854, "learning_rate": 1.1940391546114358e-05, "loss": 0.0902, "step": 773 }, { "epoch": 45.529411764705884, "grad_norm": 3.406317710876465, "learning_rate": 1.1921698631097768e-05, "loss": 0.133, "step": 774 }, { "epoch": 45.588235294117645, "grad_norm": 2.968200206756592, "learning_rate": 1.1902998741091271e-05, "loss": 0.0826, "step": 775 }, { "epoch": 45.64705882352941, "grad_norm": 3.8358709812164307, "learning_rate": 1.1884291943967919e-05, "loss": 0.2094, "step": 776 }, { "epoch": 45.705882352941174, "grad_norm": 2.3897249698638916, "learning_rate": 1.1865578307625822e-05, "loss": 0.138, "step": 777 }, { "epoch": 45.76470588235294, "grad_norm": 3.935864210128784, "learning_rate": 1.1846857899987928e-05, "loss": 0.128, "step": 778 }, { "epoch": 45.8235294117647, "grad_norm": 4.2306904792785645, "learning_rate": 1.1828130789001758e-05, "loss": 0.1497, "step": 779 }, { "epoch": 45.88235294117647, "grad_norm": 3.393631935119629, "learning_rate": 1.1809397042639155e-05, "loss": 0.2226, "step": 780 }, { "epoch": 45.94117647058823, "grad_norm": 3.6321964263916016, "learning_rate": 1.1790656728896054e-05, "loss": 0.1248, "step": 781 }, { "epoch": 46.0, "grad_norm": 4.7908124923706055, "learning_rate": 1.177190991579223e-05, "loss": 0.1926, "step": 782 }, { "epoch": 46.05882352941177, "grad_norm": 3.0268146991729736, "learning_rate": 1.1753156671371037e-05, "loss": 0.1203, "step": 783 }, { "epoch": 46.11764705882353, "grad_norm": 2.901962995529175, "learning_rate": 1.173439706369918e-05, "loss": 0.0748, "step": 784 }, { "epoch": 46.1764705882353, "grad_norm": 3.1827709674835205, "learning_rate": 1.1715631160866466e-05, "loss": 0.128, "step": 785 }, { "epoch": 46.23529411764706, "grad_norm": 2.6724817752838135, "learning_rate": 1.1696859030985528e-05, "loss": 0.2332, "step": 786 }, { "epoch": 46.294117647058826, "grad_norm": 2.707611560821533, "learning_rate": 1.1678080742191628e-05, "loss": 0.1201, "step": 787 }, { "epoch": 46.35294117647059, "grad_norm": 3.3216845989227295, "learning_rate": 1.1659296362642367e-05, "loss": 0.1351, "step": 788 }, { "epoch": 46.411764705882355, "grad_norm": 2.614147663116455, "learning_rate": 1.1640505960517456e-05, "loss": 0.1799, "step": 789 }, { "epoch": 46.470588235294116, "grad_norm": 2.959559679031372, "learning_rate": 1.1621709604018465e-05, "loss": 0.0839, "step": 790 }, { "epoch": 46.529411764705884, "grad_norm": 2.9060754776000977, "learning_rate": 1.160290736136858e-05, "loss": 0.1795, "step": 791 }, { "epoch": 46.588235294117645, "grad_norm": 6.090054512023926, "learning_rate": 1.1584099300812345e-05, "loss": 0.156, "step": 792 }, { "epoch": 46.64705882352941, "grad_norm": 2.6058173179626465, "learning_rate": 1.1565285490615423e-05, "loss": 0.0874, "step": 793 }, { "epoch": 46.705882352941174, "grad_norm": 3.3022027015686035, "learning_rate": 1.1546465999064346e-05, "loss": 0.1701, "step": 794 }, { "epoch": 46.76470588235294, "grad_norm": 3.2136588096618652, "learning_rate": 1.1527640894466273e-05, "loss": 0.1379, "step": 795 }, { "epoch": 46.8235294117647, "grad_norm": 3.339705467224121, "learning_rate": 1.1508810245148724e-05, "loss": 0.257, "step": 796 }, { "epoch": 46.88235294117647, "grad_norm": 2.7446775436401367, "learning_rate": 1.148997411945936e-05, "loss": 0.1727, "step": 797 }, { "epoch": 46.94117647058823, "grad_norm": 2.235184669494629, "learning_rate": 1.14711325857657e-05, "loss": 0.0863, "step": 798 }, { "epoch": 47.0, "grad_norm": 2.5053062438964844, "learning_rate": 1.1452285712454905e-05, "loss": 0.106, "step": 799 }, { "epoch": 47.05882352941177, "grad_norm": 2.8649399280548096, "learning_rate": 1.1433433567933514e-05, "loss": 0.1474, "step": 800 }, { "epoch": 47.11764705882353, "grad_norm": 2.613492250442505, "learning_rate": 1.1414576220627199e-05, "loss": 0.0796, "step": 801 }, { "epoch": 47.1764705882353, "grad_norm": 4.319946765899658, "learning_rate": 1.1395713738980512e-05, "loss": 0.1325, "step": 802 }, { "epoch": 47.23529411764706, "grad_norm": 3.135833978652954, "learning_rate": 1.1376846191456646e-05, "loss": 0.0994, "step": 803 }, { "epoch": 47.294117647058826, "grad_norm": 2.3916053771972656, "learning_rate": 1.1357973646537177e-05, "loss": 0.0774, "step": 804 }, { "epoch": 47.35294117647059, "grad_norm": 2.9093377590179443, "learning_rate": 1.1339096172721823e-05, "loss": 0.3367, "step": 805 }, { "epoch": 47.411764705882355, "grad_norm": 5.196231365203857, "learning_rate": 1.1320213838528185e-05, "loss": 0.1945, "step": 806 }, { "epoch": 47.470588235294116, "grad_norm": 2.7416510581970215, "learning_rate": 1.1301326712491512e-05, "loss": 0.1142, "step": 807 }, { "epoch": 47.529411764705884, "grad_norm": 2.3996877670288086, "learning_rate": 1.1282434863164446e-05, "loss": 0.177, "step": 808 }, { "epoch": 47.588235294117645, "grad_norm": 2.0464253425598145, "learning_rate": 1.1263538359116767e-05, "loss": 0.0891, "step": 809 }, { "epoch": 47.64705882352941, "grad_norm": 3.025827646255493, "learning_rate": 1.1244637268935157e-05, "loss": 0.1175, "step": 810 }, { "epoch": 47.705882352941174, "grad_norm": 2.9891061782836914, "learning_rate": 1.1225731661222938e-05, "loss": 0.1169, "step": 811 }, { "epoch": 47.76470588235294, "grad_norm": 2.234125852584839, "learning_rate": 1.1206821604599827e-05, "loss": 0.0611, "step": 812 }, { "epoch": 47.8235294117647, "grad_norm": 3.1017377376556396, "learning_rate": 1.1187907167701694e-05, "loss": 0.1966, "step": 813 }, { "epoch": 47.88235294117647, "grad_norm": 2.629161834716797, "learning_rate": 1.1168988419180308e-05, "loss": 0.1153, "step": 814 }, { "epoch": 47.94117647058823, "grad_norm": 3.037599802017212, "learning_rate": 1.1150065427703088e-05, "loss": 0.0883, "step": 815 }, { "epoch": 48.0, "grad_norm": 3.4654269218444824, "learning_rate": 1.1131138261952845e-05, "loss": 0.1504, "step": 816 }, { "epoch": 48.05882352941177, "grad_norm": 2.4275951385498047, "learning_rate": 1.1112206990627547e-05, "loss": 0.0643, "step": 817 }, { "epoch": 48.11764705882353, "grad_norm": 2.9709553718566895, "learning_rate": 1.1093271682440063e-05, "loss": 0.1085, "step": 818 }, { "epoch": 48.1764705882353, "grad_norm": 2.2968692779541016, "learning_rate": 1.1074332406117918e-05, "loss": 0.1145, "step": 819 }, { "epoch": 48.23529411764706, "grad_norm": 2.6494407653808594, "learning_rate": 1.105538923040303e-05, "loss": 0.1084, "step": 820 }, { "epoch": 48.294117647058826, "grad_norm": 2.823833703994751, "learning_rate": 1.1036442224051485e-05, "loss": 0.1538, "step": 821 }, { "epoch": 48.35294117647059, "grad_norm": 8.525276184082031, "learning_rate": 1.1017491455833258e-05, "loss": 0.0769, "step": 822 }, { "epoch": 48.411764705882355, "grad_norm": 2.670389413833618, "learning_rate": 1.0998536994531982e-05, "loss": 0.1881, "step": 823 }, { "epoch": 48.470588235294116, "grad_norm": 2.584061861038208, "learning_rate": 1.0979578908944698e-05, "loss": 0.1723, "step": 824 }, { "epoch": 48.529411764705884, "grad_norm": 2.8650758266448975, "learning_rate": 1.0960617267881601e-05, "loss": 0.0903, "step": 825 }, { "epoch": 48.588235294117645, "grad_norm": 2.3102641105651855, "learning_rate": 1.094165214016579e-05, "loss": 0.0612, "step": 826 }, { "epoch": 48.64705882352941, "grad_norm": 3.305549144744873, "learning_rate": 1.092268359463302e-05, "loss": 0.2436, "step": 827 }, { "epoch": 48.705882352941174, "grad_norm": 2.0116915702819824, "learning_rate": 1.090371170013145e-05, "loss": 0.0768, "step": 828 }, { "epoch": 48.76470588235294, "grad_norm": 2.708873748779297, "learning_rate": 1.0884736525521395e-05, "loss": 0.2963, "step": 829 }, { "epoch": 48.8235294117647, "grad_norm": 3.181830883026123, "learning_rate": 1.0865758139675071e-05, "loss": 0.133, "step": 830 }, { "epoch": 48.88235294117647, "grad_norm": 2.6251277923583984, "learning_rate": 1.084677661147636e-05, "loss": 0.0956, "step": 831 }, { "epoch": 48.94117647058823, "grad_norm": 3.6014316082000732, "learning_rate": 1.0827792009820537e-05, "loss": 0.1773, "step": 832 }, { "epoch": 49.0, "grad_norm": 2.3700881004333496, "learning_rate": 1.0808804403614044e-05, "loss": 0.0623, "step": 833 }, { "epoch": 49.05882352941177, "grad_norm": 2.4597833156585693, "learning_rate": 1.078981386177422e-05, "loss": 0.1033, "step": 834 }, { "epoch": 49.11764705882353, "grad_norm": 2.8126020431518555, "learning_rate": 1.0770820453229063e-05, "loss": 0.1729, "step": 835 }, { "epoch": 49.1764705882353, "grad_norm": 2.968871831893921, "learning_rate": 1.0751824246916975e-05, "loss": 0.1347, "step": 836 }, { "epoch": 49.23529411764706, "grad_norm": 5.173304557800293, "learning_rate": 1.0732825311786512e-05, "loss": 0.1559, "step": 837 }, { "epoch": 49.294117647058826, "grad_norm": 2.734658718109131, "learning_rate": 1.0713823716796136e-05, "loss": 0.1195, "step": 838 }, { "epoch": 49.35294117647059, "grad_norm": 2.529216766357422, "learning_rate": 1.0694819530913965e-05, "loss": 0.1226, "step": 839 }, { "epoch": 49.411764705882355, "grad_norm": 3.1976747512817383, "learning_rate": 1.0675812823117516e-05, "loss": 0.1336, "step": 840 }, { "epoch": 49.470588235294116, "grad_norm": 2.4332756996154785, "learning_rate": 1.0656803662393462e-05, "loss": 0.0766, "step": 841 }, { "epoch": 49.529411764705884, "grad_norm": 3.8745784759521484, "learning_rate": 1.0637792117737378e-05, "loss": 0.0951, "step": 842 }, { "epoch": 49.588235294117645, "grad_norm": 2.5881199836730957, "learning_rate": 1.0618778258153495e-05, "loss": 0.1273, "step": 843 }, { "epoch": 49.64705882352941, "grad_norm": 3.071343421936035, "learning_rate": 1.0599762152654442e-05, "loss": 0.1928, "step": 844 }, { "epoch": 49.705882352941174, "grad_norm": 3.365490436553955, "learning_rate": 1.0580743870261006e-05, "loss": 0.0965, "step": 845 }, { "epoch": 49.76470588235294, "grad_norm": 2.515998125076294, "learning_rate": 1.0561723480001864e-05, "loss": 0.0802, "step": 846 }, { "epoch": 49.8235294117647, "grad_norm": 3.8120810985565186, "learning_rate": 1.0542701050913357e-05, "loss": 0.1314, "step": 847 }, { "epoch": 49.88235294117647, "grad_norm": 2.9535303115844727, "learning_rate": 1.052367665203922e-05, "loss": 0.1629, "step": 848 }, { "epoch": 49.94117647058823, "grad_norm": 3.178980827331543, "learning_rate": 1.0504650352430334e-05, "loss": 0.0768, "step": 849 }, { "epoch": 50.0, "grad_norm": 2.8479862213134766, "learning_rate": 1.0485622221144485e-05, "loss": 0.1834, "step": 850 }, { "epoch": 50.05882352941177, "grad_norm": 2.7964913845062256, "learning_rate": 1.0466592327246101e-05, "loss": 0.1981, "step": 851 }, { "epoch": 50.11764705882353, "grad_norm": 2.8469254970550537, "learning_rate": 1.0447560739806013e-05, "loss": 0.1377, "step": 852 }, { "epoch": 50.1764705882353, "grad_norm": 1.9806376695632935, "learning_rate": 1.04285275279012e-05, "loss": 0.0725, "step": 853 }, { "epoch": 50.23529411764706, "grad_norm": 3.231048583984375, "learning_rate": 1.0409492760614532e-05, "loss": 0.1712, "step": 854 }, { "epoch": 50.294117647058826, "grad_norm": 2.3866496086120605, "learning_rate": 1.0390456507034525e-05, "loss": 0.0842, "step": 855 }, { "epoch": 50.35294117647059, "grad_norm": 2.558290958404541, "learning_rate": 1.0371418836255091e-05, "loss": 0.0981, "step": 856 }, { "epoch": 50.411764705882355, "grad_norm": 2.239016056060791, "learning_rate": 1.0352379817375292e-05, "loss": 0.0505, "step": 857 }, { "epoch": 50.470588235294116, "grad_norm": 2.5159480571746826, "learning_rate": 1.0333339519499069e-05, "loss": 0.1056, "step": 858 }, { "epoch": 50.529411764705884, "grad_norm": 2.3022494316101074, "learning_rate": 1.0314298011735018e-05, "loss": 0.1514, "step": 859 }, { "epoch": 50.588235294117645, "grad_norm": 2.855541467666626, "learning_rate": 1.0295255363196121e-05, "loss": 0.1365, "step": 860 }, { "epoch": 50.64705882352941, "grad_norm": 3.373917579650879, "learning_rate": 1.0276211642999505e-05, "loss": 0.1141, "step": 861 }, { "epoch": 50.705882352941174, "grad_norm": 5.038883209228516, "learning_rate": 1.0257166920266177e-05, "loss": 0.1089, "step": 862 }, { "epoch": 50.76470588235294, "grad_norm": 2.6901161670684814, "learning_rate": 1.0238121264120789e-05, "loss": 0.088, "step": 863 }, { "epoch": 50.8235294117647, "grad_norm": 7.013632774353027, "learning_rate": 1.0219074743691384e-05, "loss": 0.0812, "step": 864 }, { "epoch": 50.88235294117647, "grad_norm": 4.17929220199585, "learning_rate": 1.020002742810914e-05, "loss": 0.2187, "step": 865 }, { "epoch": 50.94117647058823, "grad_norm": 2.702371597290039, "learning_rate": 1.0180979386508119e-05, "loss": 0.0915, "step": 866 }, { "epoch": 51.0, "grad_norm": 2.1229047775268555, "learning_rate": 1.0161930688025018e-05, "loss": 0.1684, "step": 867 }, { "epoch": 51.05882352941177, "grad_norm": 2.256157398223877, "learning_rate": 1.0142881401798917e-05, "loss": 0.1175, "step": 868 }, { "epoch": 51.11764705882353, "grad_norm": 2.8414218425750732, "learning_rate": 1.0123831596971033e-05, "loss": 0.0875, "step": 869 }, { "epoch": 51.1764705882353, "grad_norm": 2.473802328109741, "learning_rate": 1.0104781342684465e-05, "loss": 0.1639, "step": 870 }, { "epoch": 51.23529411764706, "grad_norm": 9.966187477111816, "learning_rate": 1.008573070808394e-05, "loss": 0.2173, "step": 871 }, { "epoch": 51.294117647058826, "grad_norm": 2.6679494380950928, "learning_rate": 1.006667976231557e-05, "loss": 0.0577, "step": 872 }, { "epoch": 51.35294117647059, "grad_norm": 3.305358409881592, "learning_rate": 1.0047628574526595e-05, "loss": 0.1101, "step": 873 }, { "epoch": 51.411764705882355, "grad_norm": 3.8286852836608887, "learning_rate": 1.002857721386513e-05, "loss": 0.0578, "step": 874 }, { "epoch": 51.470588235294116, "grad_norm": 3.1898412704467773, "learning_rate": 1.0009525749479918e-05, "loss": 0.0948, "step": 875 }, { "epoch": 51.529411764705884, "grad_norm": 3.1662368774414062, "learning_rate": 9.990474250520085e-06, "loss": 0.2324, "step": 876 }, { "epoch": 51.588235294117645, "grad_norm": 2.51600980758667, "learning_rate": 9.971422786134874e-06, "loss": 0.1114, "step": 877 }, { "epoch": 51.64705882352941, "grad_norm": 2.0254783630371094, "learning_rate": 9.952371425473406e-06, "loss": 0.0701, "step": 878 }, { "epoch": 51.705882352941174, "grad_norm": 2.6197547912597656, "learning_rate": 9.933320237684433e-06, "loss": 0.1162, "step": 879 }, { "epoch": 51.76470588235294, "grad_norm": 2.833610773086548, "learning_rate": 9.914269291916063e-06, "loss": 0.0648, "step": 880 }, { "epoch": 51.8235294117647, "grad_norm": 3.0622050762176514, "learning_rate": 9.89521865731554e-06, "loss": 0.1316, "step": 881 }, { "epoch": 51.88235294117647, "grad_norm": 18.00055503845215, "learning_rate": 9.87616840302897e-06, "loss": 0.0857, "step": 882 }, { "epoch": 51.94117647058823, "grad_norm": 3.162602663040161, "learning_rate": 9.857118598201088e-06, "loss": 0.0939, "step": 883 }, { "epoch": 52.0, "grad_norm": 3.6887407302856445, "learning_rate": 9.838069311974986e-06, "loss": 0.1642, "step": 884 }, { "epoch": 52.05882352941177, "grad_norm": 2.8375205993652344, "learning_rate": 9.819020613491883e-06, "loss": 0.149, "step": 885 }, { "epoch": 52.11764705882353, "grad_norm": 2.4975876808166504, "learning_rate": 9.799972571890862e-06, "loss": 0.0699, "step": 886 }, { "epoch": 52.1764705882353, "grad_norm": 3.1931276321411133, "learning_rate": 9.780925256308617e-06, "loss": 0.11, "step": 887 }, { "epoch": 52.23529411764706, "grad_norm": 2.8182597160339355, "learning_rate": 9.761878735879214e-06, "loss": 0.0952, "step": 888 }, { "epoch": 52.294117647058826, "grad_norm": 3.484889268875122, "learning_rate": 9.742833079733828e-06, "loss": 0.1595, "step": 889 }, { "epoch": 52.35294117647059, "grad_norm": 2.686429023742676, "learning_rate": 9.7237883570005e-06, "loss": 0.0971, "step": 890 }, { "epoch": 52.411764705882355, "grad_norm": 2.6333348751068115, "learning_rate": 9.704744636803882e-06, "loss": 0.1806, "step": 891 }, { "epoch": 52.470588235294116, "grad_norm": 2.8333513736724854, "learning_rate": 9.685701988264985e-06, "loss": 0.082, "step": 892 }, { "epoch": 52.529411764705884, "grad_norm": 2.3150947093963623, "learning_rate": 9.666660480500936e-06, "loss": 0.0743, "step": 893 }, { "epoch": 52.588235294117645, "grad_norm": 1.831231951713562, "learning_rate": 9.647620182624715e-06, "loss": 0.0849, "step": 894 }, { "epoch": 52.64705882352941, "grad_norm": 2.423409938812256, "learning_rate": 9.628581163744912e-06, "loss": 0.1782, "step": 895 }, { "epoch": 52.705882352941174, "grad_norm": 1.944344401359558, "learning_rate": 9.60954349296548e-06, "loss": 0.0782, "step": 896 }, { "epoch": 52.76470588235294, "grad_norm": 2.649003744125366, "learning_rate": 9.590507239385473e-06, "loss": 0.1263, "step": 897 }, { "epoch": 52.8235294117647, "grad_norm": 2.4523630142211914, "learning_rate": 9.571472472098804e-06, "loss": 0.0867, "step": 898 }, { "epoch": 52.88235294117647, "grad_norm": 1.9579745531082153, "learning_rate": 9.552439260193992e-06, "loss": 0.0984, "step": 899 }, { "epoch": 52.94117647058823, "grad_norm": 2.6260666847229004, "learning_rate": 9.533407672753904e-06, "loss": 0.0705, "step": 900 }, { "epoch": 53.0, "grad_norm": 2.3615245819091797, "learning_rate": 9.514377778855521e-06, "loss": 0.1131, "step": 901 }, { "epoch": 53.05882352941177, "grad_norm": 2.516961097717285, "learning_rate": 9.495349647569666e-06, "loss": 0.0565, "step": 902 }, { "epoch": 53.11764705882353, "grad_norm": 5.055416107177734, "learning_rate": 9.476323347960781e-06, "loss": 0.148, "step": 903 }, { "epoch": 53.1764705882353, "grad_norm": 2.724700689315796, "learning_rate": 9.457298949086643e-06, "loss": 0.1385, "step": 904 }, { "epoch": 53.23529411764706, "grad_norm": 2.0691821575164795, "learning_rate": 9.438276519998136e-06, "loss": 0.064, "step": 905 }, { "epoch": 53.294117647058826, "grad_norm": 2.2804641723632812, "learning_rate": 9.419256129738996e-06, "loss": 0.0473, "step": 906 }, { "epoch": 53.35294117647059, "grad_norm": 2.1843438148498535, "learning_rate": 9.400237847345558e-06, "loss": 0.0932, "step": 907 }, { "epoch": 53.411764705882355, "grad_norm": 2.9909870624542236, "learning_rate": 9.381221741846505e-06, "loss": 0.1739, "step": 908 }, { "epoch": 53.470588235294116, "grad_norm": 2.1224942207336426, "learning_rate": 9.362207882262622e-06, "loss": 0.0824, "step": 909 }, { "epoch": 53.529411764705884, "grad_norm": 2.7031216621398926, "learning_rate": 9.343196337606541e-06, "loss": 0.1072, "step": 910 }, { "epoch": 53.588235294117645, "grad_norm": 2.525592088699341, "learning_rate": 9.324187176882486e-06, "loss": 0.1311, "step": 911 }, { "epoch": 53.64705882352941, "grad_norm": 2.3430416584014893, "learning_rate": 9.305180469086035e-06, "loss": 0.0937, "step": 912 }, { "epoch": 53.705882352941174, "grad_norm": 2.8973207473754883, "learning_rate": 9.286176283203864e-06, "loss": 0.2203, "step": 913 }, { "epoch": 53.76470588235294, "grad_norm": 2.5040533542633057, "learning_rate": 9.267174688213488e-06, "loss": 0.0985, "step": 914 }, { "epoch": 53.8235294117647, "grad_norm": 2.7054920196533203, "learning_rate": 9.248175753083026e-06, "loss": 0.0986, "step": 915 }, { "epoch": 53.88235294117647, "grad_norm": 2.5714826583862305, "learning_rate": 9.22917954677094e-06, "loss": 0.0696, "step": 916 }, { "epoch": 53.94117647058823, "grad_norm": 2.7210867404937744, "learning_rate": 9.210186138225784e-06, "loss": 0.1002, "step": 917 }, { "epoch": 54.0, "grad_norm": 2.274341583251953, "learning_rate": 9.19119559638596e-06, "loss": 0.1185, "step": 918 }, { "epoch": 54.05882352941177, "grad_norm": 2.5188066959381104, "learning_rate": 9.172207990179466e-06, "loss": 0.0881, "step": 919 }, { "epoch": 54.11764705882353, "grad_norm": 2.449145555496216, "learning_rate": 9.153223388523644e-06, "loss": 0.1546, "step": 920 }, { "epoch": 54.1764705882353, "grad_norm": 2.1233668327331543, "learning_rate": 9.134241860324932e-06, "loss": 0.1245, "step": 921 }, { "epoch": 54.23529411764706, "grad_norm": 3.3754565715789795, "learning_rate": 9.115263474478608e-06, "loss": 0.0845, "step": 922 }, { "epoch": 54.294117647058826, "grad_norm": 1.5214611291885376, "learning_rate": 9.096288299868552e-06, "loss": 0.0401, "step": 923 }, { "epoch": 54.35294117647059, "grad_norm": 2.3150076866149902, "learning_rate": 9.07731640536698e-06, "loss": 0.1251, "step": 924 }, { "epoch": 54.411764705882355, "grad_norm": 2.274306297302246, "learning_rate": 9.058347859834212e-06, "loss": 0.0911, "step": 925 }, { "epoch": 54.470588235294116, "grad_norm": 2.6436767578125, "learning_rate": 9.0393827321184e-06, "loss": 0.1057, "step": 926 }, { "epoch": 54.529411764705884, "grad_norm": 2.241518974304199, "learning_rate": 9.020421091055305e-06, "loss": 0.0841, "step": 927 }, { "epoch": 54.588235294117645, "grad_norm": 2.238372802734375, "learning_rate": 9.001463005468022e-06, "loss": 0.083, "step": 928 }, { "epoch": 54.64705882352941, "grad_norm": 4.0823493003845215, "learning_rate": 8.982508544166747e-06, "loss": 0.0771, "step": 929 }, { "epoch": 54.705882352941174, "grad_norm": 5.275056838989258, "learning_rate": 8.96355777594852e-06, "loss": 0.2249, "step": 930 }, { "epoch": 54.76470588235294, "grad_norm": 2.1949234008789062, "learning_rate": 8.944610769596971e-06, "loss": 0.0935, "step": 931 }, { "epoch": 54.8235294117647, "grad_norm": 1.8113642930984497, "learning_rate": 8.925667593882085e-06, "loss": 0.0483, "step": 932 }, { "epoch": 54.88235294117647, "grad_norm": 2.4045565128326416, "learning_rate": 8.906728317559939e-06, "loss": 0.1361, "step": 933 }, { "epoch": 54.94117647058823, "grad_norm": 2.5111749172210693, "learning_rate": 8.887793009372456e-06, "loss": 0.1078, "step": 934 }, { "epoch": 55.0, "grad_norm": 1.8530802726745605, "learning_rate": 8.868861738047158e-06, "loss": 0.0535, "step": 935 }, { "epoch": 55.05882352941177, "grad_norm": 2.447056293487549, "learning_rate": 8.849934572296916e-06, "loss": 0.1037, "step": 936 }, { "epoch": 55.11764705882353, "grad_norm": 2.8879566192626953, "learning_rate": 8.831011580819693e-06, "loss": 0.1017, "step": 937 }, { "epoch": 55.1764705882353, "grad_norm": 2.379065752029419, "learning_rate": 8.812092832298307e-06, "loss": 0.0623, "step": 938 }, { "epoch": 55.23529411764706, "grad_norm": 2.9746365547180176, "learning_rate": 8.793178395400178e-06, "loss": 0.0713, "step": 939 }, { "epoch": 55.294117647058826, "grad_norm": 2.582214593887329, "learning_rate": 8.774268338777067e-06, "loss": 0.1117, "step": 940 }, { "epoch": 55.35294117647059, "grad_norm": 2.242445230484009, "learning_rate": 8.755362731064845e-06, "loss": 0.1449, "step": 941 }, { "epoch": 55.411764705882355, "grad_norm": 3.213926315307617, "learning_rate": 8.736461640883234e-06, "loss": 0.0807, "step": 942 }, { "epoch": 55.470588235294116, "grad_norm": 1.6993956565856934, "learning_rate": 8.717565136835559e-06, "loss": 0.1421, "step": 943 }, { "epoch": 55.529411764705884, "grad_norm": 4.0843281745910645, "learning_rate": 8.698673287508493e-06, "loss": 0.1549, "step": 944 }, { "epoch": 55.588235294117645, "grad_norm": 2.0110442638397217, "learning_rate": 8.679786161471821e-06, "loss": 0.0922, "step": 945 }, { "epoch": 55.64705882352941, "grad_norm": 1.8739012479782104, "learning_rate": 8.660903827278184e-06, "loss": 0.0966, "step": 946 }, { "epoch": 55.705882352941174, "grad_norm": 2.0270133018493652, "learning_rate": 8.642026353462828e-06, "loss": 0.1097, "step": 947 }, { "epoch": 55.76470588235294, "grad_norm": 2.424452304840088, "learning_rate": 8.623153808543359e-06, "loss": 0.1367, "step": 948 }, { "epoch": 55.8235294117647, "grad_norm": 2.5616672039031982, "learning_rate": 8.604286261019491e-06, "loss": 0.123, "step": 949 }, { "epoch": 55.88235294117647, "grad_norm": 2.7459635734558105, "learning_rate": 8.585423779372806e-06, "loss": 0.0929, "step": 950 }, { "epoch": 55.94117647058823, "grad_norm": 2.596480131149292, "learning_rate": 8.566566432066491e-06, "loss": 0.0664, "step": 951 }, { "epoch": 56.0, "grad_norm": 2.277029514312744, "learning_rate": 8.5477142875451e-06, "loss": 0.0561, "step": 952 }, { "epoch": 56.05882352941177, "grad_norm": 1.754686713218689, "learning_rate": 8.528867414234302e-06, "loss": 0.0596, "step": 953 }, { "epoch": 56.11764705882353, "grad_norm": 2.6427359580993652, "learning_rate": 8.510025880540644e-06, "loss": 0.1357, "step": 954 }, { "epoch": 56.1764705882353, "grad_norm": 1.7537795305252075, "learning_rate": 8.491189754851274e-06, "loss": 0.0454, "step": 955 }, { "epoch": 56.23529411764706, "grad_norm": 3.2595224380493164, "learning_rate": 8.472359105533727e-06, "loss": 0.1158, "step": 956 }, { "epoch": 56.294117647058826, "grad_norm": 2.455657482147217, "learning_rate": 8.453534000935652e-06, "loss": 0.075, "step": 957 }, { "epoch": 56.35294117647059, "grad_norm": 1.9798626899719238, "learning_rate": 8.434714509384578e-06, "loss": 0.0937, "step": 958 }, { "epoch": 56.411764705882355, "grad_norm": 2.4683902263641357, "learning_rate": 8.415900699187656e-06, "loss": 0.0755, "step": 959 }, { "epoch": 56.470588235294116, "grad_norm": 2.6059443950653076, "learning_rate": 8.397092638631422e-06, "loss": 0.106, "step": 960 }, { "epoch": 56.529411764705884, "grad_norm": 2.0211246013641357, "learning_rate": 8.378290395981535e-06, "loss": 0.139, "step": 961 }, { "epoch": 56.588235294117645, "grad_norm": 2.3349804878234863, "learning_rate": 8.359494039482544e-06, "loss": 0.1542, "step": 962 }, { "epoch": 56.64705882352941, "grad_norm": 2.147113800048828, "learning_rate": 8.340703637357633e-06, "loss": 0.1185, "step": 963 }, { "epoch": 56.705882352941174, "grad_norm": 2.9251534938812256, "learning_rate": 8.321919257808372e-06, "loss": 0.2033, "step": 964 }, { "epoch": 56.76470588235294, "grad_norm": 2.801241159439087, "learning_rate": 8.303140969014473e-06, "loss": 0.078, "step": 965 }, { "epoch": 56.8235294117647, "grad_norm": 2.2328317165374756, "learning_rate": 8.284368839133539e-06, "loss": 0.0673, "step": 966 }, { "epoch": 56.88235294117647, "grad_norm": 1.9924618005752563, "learning_rate": 8.265602936300821e-06, "loss": 0.0688, "step": 967 }, { "epoch": 56.94117647058823, "grad_norm": 2.1585426330566406, "learning_rate": 8.246843328628966e-06, "loss": 0.0845, "step": 968 }, { "epoch": 57.0, "grad_norm": 2.2520289421081543, "learning_rate": 8.228090084207773e-06, "loss": 0.0894, "step": 969 }, { "epoch": 57.05882352941177, "grad_norm": 2.4182403087615967, "learning_rate": 8.209343271103947e-06, "loss": 0.1332, "step": 970 }, { "epoch": 57.11764705882353, "grad_norm": 2.377816915512085, "learning_rate": 8.190602957360848e-06, "loss": 0.079, "step": 971 }, { "epoch": 57.1764705882353, "grad_norm": 2.20737624168396, "learning_rate": 8.171869210998244e-06, "loss": 0.1327, "step": 972 }, { "epoch": 57.23529411764706, "grad_norm": 2.3041000366210938, "learning_rate": 8.153142100012073e-06, "loss": 0.0901, "step": 973 }, { "epoch": 57.294117647058826, "grad_norm": 2.5677459239959717, "learning_rate": 8.134421692374181e-06, "loss": 0.1443, "step": 974 }, { "epoch": 57.35294117647059, "grad_norm": 1.8329119682312012, "learning_rate": 8.115708056032086e-06, "loss": 0.0495, "step": 975 }, { "epoch": 57.411764705882355, "grad_norm": 2.298112630844116, "learning_rate": 8.09700125890873e-06, "loss": 0.1301, "step": 976 }, { "epoch": 57.470588235294116, "grad_norm": 2.5729730129241943, "learning_rate": 8.078301368902233e-06, "loss": 0.0691, "step": 977 }, { "epoch": 57.529411764705884, "grad_norm": 2.709261894226074, "learning_rate": 8.059608453885646e-06, "loss": 0.075, "step": 978 }, { "epoch": 57.588235294117645, "grad_norm": 2.2083306312561035, "learning_rate": 8.040922581706695e-06, "loss": 0.0824, "step": 979 }, { "epoch": 57.64705882352941, "grad_norm": 1.972137451171875, "learning_rate": 8.022243820187554e-06, "loss": 0.0462, "step": 980 }, { "epoch": 57.705882352941174, "grad_norm": 1.7183464765548706, "learning_rate": 8.003572237124582e-06, "loss": 0.0499, "step": 981 }, { "epoch": 57.76470588235294, "grad_norm": 2.1274566650390625, "learning_rate": 7.984907900288083e-06, "loss": 0.1096, "step": 982 }, { "epoch": 57.8235294117647, "grad_norm": 2.106963634490967, "learning_rate": 7.966250877422064e-06, "loss": 0.064, "step": 983 }, { "epoch": 57.88235294117647, "grad_norm": 2.9204659461975098, "learning_rate": 7.947601236243982e-06, "loss": 0.1783, "step": 984 }, { "epoch": 57.94117647058823, "grad_norm": 2.1676111221313477, "learning_rate": 7.9289590444445e-06, "loss": 0.1201, "step": 985 }, { "epoch": 58.0, "grad_norm": 2.48429536819458, "learning_rate": 7.91032436968725e-06, "loss": 0.1145, "step": 986 }, { "epoch": 58.05882352941177, "grad_norm": 1.5226337909698486, "learning_rate": 7.891697279608574e-06, "loss": 0.0407, "step": 987 }, { "epoch": 58.11764705882353, "grad_norm": 2.073188066482544, "learning_rate": 7.873077841817282e-06, "loss": 0.0822, "step": 988 }, { "epoch": 58.1764705882353, "grad_norm": 2.2245726585388184, "learning_rate": 7.854466123894416e-06, "loss": 0.1909, "step": 989 }, { "epoch": 58.23529411764706, "grad_norm": 3.3995211124420166, "learning_rate": 7.835862193392996e-06, "loss": 0.0917, "step": 990 }, { "epoch": 58.294117647058826, "grad_norm": 2.321596384048462, "learning_rate": 7.817266117837774e-06, "loss": 0.0816, "step": 991 }, { "epoch": 58.35294117647059, "grad_norm": 1.5946742296218872, "learning_rate": 7.798677964724996e-06, "loss": 0.0397, "step": 992 }, { "epoch": 58.411764705882355, "grad_norm": 1.9393513202667236, "learning_rate": 7.780097801522154e-06, "loss": 0.1606, "step": 993 }, { "epoch": 58.470588235294116, "grad_norm": 3.1876089572906494, "learning_rate": 7.761525695667726e-06, "loss": 0.0901, "step": 994 }, { "epoch": 58.529411764705884, "grad_norm": 1.9649244546890259, "learning_rate": 7.742961714570964e-06, "loss": 0.0644, "step": 995 }, { "epoch": 58.588235294117645, "grad_norm": 1.6281061172485352, "learning_rate": 7.724405925611618e-06, "loss": 0.0462, "step": 996 }, { "epoch": 58.64705882352941, "grad_norm": 1.9317772388458252, "learning_rate": 7.705858396139712e-06, "loss": 0.0627, "step": 997 }, { "epoch": 58.705882352941174, "grad_norm": 1.2906898260116577, "learning_rate": 7.687319193475282e-06, "loss": 0.0446, "step": 998 }, { "epoch": 58.76470588235294, "grad_norm": 2.3061625957489014, "learning_rate": 7.668788384908148e-06, "loss": 0.0534, "step": 999 }, { "epoch": 58.8235294117647, "grad_norm": 2.230212688446045, "learning_rate": 7.650266037697664e-06, "loss": 0.1127, "step": 1000 }, { "epoch": 58.88235294117647, "grad_norm": 7.071537971496582, "learning_rate": 7.631752219072465e-06, "loss": 0.215, "step": 1001 }, { "epoch": 58.94117647058823, "grad_norm": 1.7719978094100952, "learning_rate": 7.613246996230238e-06, "loss": 0.0495, "step": 1002 }, { "epoch": 59.0, "grad_norm": 3.1264519691467285, "learning_rate": 7.594750436337467e-06, "loss": 0.0899, "step": 1003 }, { "epoch": 59.05882352941177, "grad_norm": 2.447171211242676, "learning_rate": 7.576262606529195e-06, "loss": 0.0936, "step": 1004 }, { "epoch": 59.11764705882353, "grad_norm": 1.754564881324768, "learning_rate": 7.557783573908774e-06, "loss": 0.0479, "step": 1005 }, { "epoch": 59.1764705882353, "grad_norm": 2.4242193698883057, "learning_rate": 7.539313405547636e-06, "loss": 0.1814, "step": 1006 }, { "epoch": 59.23529411764706, "grad_norm": 2.7452683448791504, "learning_rate": 7.520852168485029e-06, "loss": 0.0723, "step": 1007 }, { "epoch": 59.294117647058826, "grad_norm": 2.674380302429199, "learning_rate": 7.502399929727786e-06, "loss": 0.0941, "step": 1008 }, { "epoch": 59.35294117647059, "grad_norm": 2.6913845539093018, "learning_rate": 7.483956756250084e-06, "loss": 0.1087, "step": 1009 }, { "epoch": 59.411764705882355, "grad_norm": 1.9957925081253052, "learning_rate": 7.465522714993194e-06, "loss": 0.0599, "step": 1010 }, { "epoch": 59.470588235294116, "grad_norm": 2.693538188934326, "learning_rate": 7.447097872865238e-06, "loss": 0.05, "step": 1011 }, { "epoch": 59.529411764705884, "grad_norm": 2.107944965362549, "learning_rate": 7.428682296740953e-06, "loss": 0.059, "step": 1012 }, { "epoch": 59.588235294117645, "grad_norm": 1.795035481452942, "learning_rate": 7.410276053461443e-06, "loss": 0.0497, "step": 1013 }, { "epoch": 59.64705882352941, "grad_norm": 2.269866943359375, "learning_rate": 7.391879209833938e-06, "loss": 0.0767, "step": 1014 }, { "epoch": 59.705882352941174, "grad_norm": 2.4396181106567383, "learning_rate": 7.3734918326315476e-06, "loss": 0.0574, "step": 1015 }, { "epoch": 59.76470588235294, "grad_norm": 1.860007882118225, "learning_rate": 7.355113988593028e-06, "loss": 0.1145, "step": 1016 }, { "epoch": 59.8235294117647, "grad_norm": 1.638607382774353, "learning_rate": 7.336745744422532e-06, "loss": 0.0963, "step": 1017 }, { "epoch": 59.88235294117647, "grad_norm": 3.9875805377960205, "learning_rate": 7.318387166789362e-06, "loss": 0.0923, "step": 1018 }, { "epoch": 59.94117647058823, "grad_norm": 1.7986887693405151, "learning_rate": 7.300038322327745e-06, "loss": 0.0981, "step": 1019 }, { "epoch": 60.0, "grad_norm": 2.3788816928863525, "learning_rate": 7.2816992776365714e-06, "loss": 0.212, "step": 1020 }, { "epoch": 60.05882352941177, "grad_norm": 2.2918646335601807, "learning_rate": 7.263370099279173e-06, "loss": 0.1561, "step": 1021 }, { "epoch": 60.11764705882353, "grad_norm": 1.6800100803375244, "learning_rate": 7.245050853783058e-06, "loss": 0.0462, "step": 1022 }, { "epoch": 60.1764705882353, "grad_norm": 1.7822667360305786, "learning_rate": 7.2267416076396924e-06, "loss": 0.0431, "step": 1023 }, { "epoch": 60.23529411764706, "grad_norm": 2.1274075508117676, "learning_rate": 7.208442427304248e-06, "loss": 0.1185, "step": 1024 }, { "epoch": 60.294117647058826, "grad_norm": 2.4499032497406006, "learning_rate": 7.190153379195352e-06, "loss": 0.0567, "step": 1025 }, { "epoch": 60.35294117647059, "grad_norm": 2.0662641525268555, "learning_rate": 7.171874529694867e-06, "loss": 0.0996, "step": 1026 }, { "epoch": 60.411764705882355, "grad_norm": 2.213526964187622, "learning_rate": 7.153605945147631e-06, "loss": 0.0798, "step": 1027 }, { "epoch": 60.470588235294116, "grad_norm": 5.366666316986084, "learning_rate": 7.135347691861232e-06, "loss": 0.1559, "step": 1028 }, { "epoch": 60.529411764705884, "grad_norm": 1.992749810218811, "learning_rate": 7.1170998361057544e-06, "loss": 0.0556, "step": 1029 }, { "epoch": 60.588235294117645, "grad_norm": 2.7120440006256104, "learning_rate": 7.0988624441135435e-06, "loss": 0.1087, "step": 1030 }, { "epoch": 60.64705882352941, "grad_norm": 1.7108114957809448, "learning_rate": 7.0806355820789665e-06, "loss": 0.0436, "step": 1031 }, { "epoch": 60.705882352941174, "grad_norm": 2.421257734298706, "learning_rate": 7.062419316158171e-06, "loss": 0.1483, "step": 1032 }, { "epoch": 60.76470588235294, "grad_norm": 1.471854567527771, "learning_rate": 7.044213712468846e-06, "loss": 0.0437, "step": 1033 }, { "epoch": 60.8235294117647, "grad_norm": 1.96552574634552, "learning_rate": 7.026018837089981e-06, "loss": 0.1107, "step": 1034 }, { "epoch": 60.88235294117647, "grad_norm": 2.239698886871338, "learning_rate": 7.007834756061622e-06, "loss": 0.077, "step": 1035 }, { "epoch": 60.94117647058823, "grad_norm": 3.901865005493164, "learning_rate": 6.989661535384642e-06, "loss": 0.0849, "step": 1036 }, { "epoch": 61.0, "grad_norm": 1.6884169578552246, "learning_rate": 6.971499241020495e-06, "loss": 0.0917, "step": 1037 }, { "epoch": 61.05882352941177, "grad_norm": 3.328171491622925, "learning_rate": 6.9533479388909686e-06, "loss": 0.1027, "step": 1038 }, { "epoch": 61.11764705882353, "grad_norm": 1.8833792209625244, "learning_rate": 6.935207694877959e-06, "loss": 0.1299, "step": 1039 }, { "epoch": 61.1764705882353, "grad_norm": 1.80158269405365, "learning_rate": 6.9170785748232294e-06, "loss": 0.0482, "step": 1040 }, { "epoch": 61.23529411764706, "grad_norm": 1.975502371788025, "learning_rate": 6.898960644528161e-06, "loss": 0.0926, "step": 1041 }, { "epoch": 61.294117647058826, "grad_norm": 2.4963951110839844, "learning_rate": 6.880853969753525e-06, "loss": 0.1713, "step": 1042 }, { "epoch": 61.35294117647059, "grad_norm": 1.9869033098220825, "learning_rate": 6.86275861621924e-06, "loss": 0.059, "step": 1043 }, { "epoch": 61.411764705882355, "grad_norm": 2.0113096237182617, "learning_rate": 6.844674649604125e-06, "loss": 0.0839, "step": 1044 }, { "epoch": 61.470588235294116, "grad_norm": 2.24310564994812, "learning_rate": 6.826602135545679e-06, "loss": 0.1247, "step": 1045 }, { "epoch": 61.529411764705884, "grad_norm": 1.9205106496810913, "learning_rate": 6.808541139639828e-06, "loss": 0.0773, "step": 1046 }, { "epoch": 61.588235294117645, "grad_norm": 1.752769112586975, "learning_rate": 6.790491727440695e-06, "loss": 0.0512, "step": 1047 }, { "epoch": 61.64705882352941, "grad_norm": 2.0939979553222656, "learning_rate": 6.772453964460353e-06, "loss": 0.0944, "step": 1048 }, { "epoch": 61.705882352941174, "grad_norm": 1.5341358184814453, "learning_rate": 6.7544279161686e-06, "loss": 0.0404, "step": 1049 }, { "epoch": 61.76470588235294, "grad_norm": 1.6492493152618408, "learning_rate": 6.736413647992712e-06, "loss": 0.0841, "step": 1050 }, { "epoch": 61.8235294117647, "grad_norm": 2.349644899368286, "learning_rate": 6.718411225317201e-06, "loss": 0.0396, "step": 1051 }, { "epoch": 61.88235294117647, "grad_norm": 2.166228771209717, "learning_rate": 6.700420713483595e-06, "loss": 0.0734, "step": 1052 }, { "epoch": 61.94117647058823, "grad_norm": 18.314407348632812, "learning_rate": 6.682442177790186e-06, "loss": 0.109, "step": 1053 }, { "epoch": 62.0, "grad_norm": 2.4039740562438965, "learning_rate": 6.664475683491797e-06, "loss": 0.0583, "step": 1054 }, { "epoch": 62.05882352941177, "grad_norm": 2.3474650382995605, "learning_rate": 6.646521295799544e-06, "loss": 0.0679, "step": 1055 }, { "epoch": 62.11764705882353, "grad_norm": 1.7737869024276733, "learning_rate": 6.628579079880605e-06, "loss": 0.0675, "step": 1056 }, { "epoch": 62.1764705882353, "grad_norm": 2.7261910438537598, "learning_rate": 6.610649100857979e-06, "loss": 0.0864, "step": 1057 }, { "epoch": 62.23529411764706, "grad_norm": 2.16615629196167, "learning_rate": 6.592731423810246e-06, "loss": 0.0976, "step": 1058 }, { "epoch": 62.294117647058826, "grad_norm": 2.1358823776245117, "learning_rate": 6.574826113771337e-06, "loss": 0.079, "step": 1059 }, { "epoch": 62.35294117647059, "grad_norm": 2.5131914615631104, "learning_rate": 6.5569332357302985e-06, "loss": 0.084, "step": 1060 }, { "epoch": 62.411764705882355, "grad_norm": 2.123958110809326, "learning_rate": 6.5390528546310515e-06, "loss": 0.1301, "step": 1061 }, { "epoch": 62.470588235294116, "grad_norm": 2.0222392082214355, "learning_rate": 6.521185035372155e-06, "loss": 0.1082, "step": 1062 }, { "epoch": 62.529411764705884, "grad_norm": 2.0828986167907715, "learning_rate": 6.503329842806576e-06, "loss": 0.0448, "step": 1063 }, { "epoch": 62.588235294117645, "grad_norm": 1.7876306772232056, "learning_rate": 6.485487341741455e-06, "loss": 0.0738, "step": 1064 }, { "epoch": 62.64705882352941, "grad_norm": 1.6102685928344727, "learning_rate": 6.467657596937864e-06, "loss": 0.0506, "step": 1065 }, { "epoch": 62.705882352941174, "grad_norm": 2.040708541870117, "learning_rate": 6.449840673110574e-06, "loss": 0.1482, "step": 1066 }, { "epoch": 62.76470588235294, "grad_norm": 1.68242609500885, "learning_rate": 6.432036634927828e-06, "loss": 0.0454, "step": 1067 }, { "epoch": 62.8235294117647, "grad_norm": 6.075501441955566, "learning_rate": 6.414245547011087e-06, "loss": 0.065, "step": 1068 }, { "epoch": 62.88235294117647, "grad_norm": 1.7385658025741577, "learning_rate": 6.396467473934818e-06, "loss": 0.0777, "step": 1069 }, { "epoch": 62.94117647058823, "grad_norm": 2.7751076221466064, "learning_rate": 6.378702480226248e-06, "loss": 0.1404, "step": 1070 }, { "epoch": 63.0, "grad_norm": 2.381718158721924, "learning_rate": 6.360950630365126e-06, "loss": 0.0461, "step": 1071 }, { "epoch": 63.05882352941177, "grad_norm": 5.089624881744385, "learning_rate": 6.343211988783497e-06, "loss": 0.0938, "step": 1072 }, { "epoch": 63.11764705882353, "grad_norm": 1.9434514045715332, "learning_rate": 6.325486619865471e-06, "loss": 0.0455, "step": 1073 }, { "epoch": 63.1764705882353, "grad_norm": 2.260155200958252, "learning_rate": 6.3077745879469754e-06, "loss": 0.091, "step": 1074 }, { "epoch": 63.23529411764706, "grad_norm": 1.9531822204589844, "learning_rate": 6.290075957315531e-06, "loss": 0.0997, "step": 1075 }, { "epoch": 63.294117647058826, "grad_norm": 2.474963903427124, "learning_rate": 6.272390792210019e-06, "loss": 0.0623, "step": 1076 }, { "epoch": 63.35294117647059, "grad_norm": 2.22385311126709, "learning_rate": 6.254719156820449e-06, "loss": 0.1231, "step": 1077 }, { "epoch": 63.411764705882355, "grad_norm": 2.48574161529541, "learning_rate": 6.237061115287718e-06, "loss": 0.0922, "step": 1078 }, { "epoch": 63.470588235294116, "grad_norm": 2.557159423828125, "learning_rate": 6.219416731703386e-06, "loss": 0.1051, "step": 1079 }, { "epoch": 63.529411764705884, "grad_norm": 1.697761058807373, "learning_rate": 6.201786070109441e-06, "loss": 0.0585, "step": 1080 }, { "epoch": 63.588235294117645, "grad_norm": 2.52803373336792, "learning_rate": 6.184169194498062e-06, "loss": 0.127, "step": 1081 }, { "epoch": 63.64705882352941, "grad_norm": 1.7810182571411133, "learning_rate": 6.1665661688113894e-06, "loss": 0.0411, "step": 1082 }, { "epoch": 63.705882352941174, "grad_norm": 2.4112768173217773, "learning_rate": 6.148977056941303e-06, "loss": 0.1114, "step": 1083 }, { "epoch": 63.76470588235294, "grad_norm": 2.278695583343506, "learning_rate": 6.131401922729172e-06, "loss": 0.0564, "step": 1084 }, { "epoch": 63.8235294117647, "grad_norm": 1.6721426248550415, "learning_rate": 6.113840829965636e-06, "loss": 0.1015, "step": 1085 }, { "epoch": 63.88235294117647, "grad_norm": 1.929262638092041, "learning_rate": 6.096293842390369e-06, "loss": 0.1012, "step": 1086 }, { "epoch": 63.94117647058823, "grad_norm": 1.92971670627594, "learning_rate": 6.078761023691851e-06, "loss": 0.033, "step": 1087 }, { "epoch": 64.0, "grad_norm": 2.2830886840820312, "learning_rate": 6.061242437507131e-06, "loss": 0.1071, "step": 1088 }, { "epoch": 64.05882352941177, "grad_norm": 2.0567493438720703, "learning_rate": 6.0437381474215995e-06, "loss": 0.0537, "step": 1089 }, { "epoch": 64.11764705882354, "grad_norm": 2.5466954708099365, "learning_rate": 6.026248216968763e-06, "loss": 0.0589, "step": 1090 }, { "epoch": 64.17647058823529, "grad_norm": 1.4295732975006104, "learning_rate": 6.0087727096300044e-06, "loss": 0.0316, "step": 1091 }, { "epoch": 64.23529411764706, "grad_norm": 1.9535611867904663, "learning_rate": 5.991311688834358e-06, "loss": 0.0854, "step": 1092 }, { "epoch": 64.29411764705883, "grad_norm": 1.8694779872894287, "learning_rate": 5.973865217958277e-06, "loss": 0.0905, "step": 1093 }, { "epoch": 64.3529411764706, "grad_norm": 5.704258441925049, "learning_rate": 5.956433360325406e-06, "loss": 0.2117, "step": 1094 }, { "epoch": 64.41176470588235, "grad_norm": 1.540730357170105, "learning_rate": 5.939016179206343e-06, "loss": 0.0454, "step": 1095 }, { "epoch": 64.47058823529412, "grad_norm": 2.030562400817871, "learning_rate": 5.921613737818426e-06, "loss": 0.0703, "step": 1096 }, { "epoch": 64.52941176470588, "grad_norm": 2.1824512481689453, "learning_rate": 5.904226099325488e-06, "loss": 0.0723, "step": 1097 }, { "epoch": 64.58823529411765, "grad_norm": 2.589940309524536, "learning_rate": 5.8868533268376315e-06, "loss": 0.1241, "step": 1098 }, { "epoch": 64.6470588235294, "grad_norm": 1.7525533437728882, "learning_rate": 5.869495483411006e-06, "loss": 0.08, "step": 1099 }, { "epoch": 64.70588235294117, "grad_norm": 1.7038687467575073, "learning_rate": 5.852152632047577e-06, "loss": 0.0594, "step": 1100 }, { "epoch": 64.76470588235294, "grad_norm": 1.9226654767990112, "learning_rate": 5.834824835694881e-06, "loss": 0.0976, "step": 1101 }, { "epoch": 64.82352941176471, "grad_norm": 1.3111071586608887, "learning_rate": 5.817512157245828e-06, "loss": 0.0623, "step": 1102 }, { "epoch": 64.88235294117646, "grad_norm": 1.7544695138931274, "learning_rate": 5.800214659538448e-06, "loss": 0.0471, "step": 1103 }, { "epoch": 64.94117647058823, "grad_norm": 1.7723764181137085, "learning_rate": 5.782932405355675e-06, "loss": 0.0479, "step": 1104 }, { "epoch": 65.0, "grad_norm": 2.1918625831604004, "learning_rate": 5.765665457425102e-06, "loss": 0.1242, "step": 1105 }, { "epoch": 65.05882352941177, "grad_norm": 1.6214667558670044, "learning_rate": 5.748413878418781e-06, "loss": 0.0398, "step": 1106 }, { "epoch": 65.11764705882354, "grad_norm": 1.771479606628418, "learning_rate": 5.73117773095298e-06, "loss": 0.077, "step": 1107 }, { "epoch": 65.17647058823529, "grad_norm": 1.7498199939727783, "learning_rate": 5.713957077587944e-06, "loss": 0.0858, "step": 1108 }, { "epoch": 65.23529411764706, "grad_norm": 2.4708049297332764, "learning_rate": 5.6967519808276995e-06, "loss": 0.1106, "step": 1109 }, { "epoch": 65.29411764705883, "grad_norm": 2.046297311782837, "learning_rate": 5.679562503119796e-06, "loss": 0.1387, "step": 1110 }, { "epoch": 65.3529411764706, "grad_norm": 2.1072499752044678, "learning_rate": 5.662388706855093e-06, "loss": 0.0635, "step": 1111 }, { "epoch": 65.41176470588235, "grad_norm": 1.6686792373657227, "learning_rate": 5.645230654367537e-06, "loss": 0.0555, "step": 1112 }, { "epoch": 65.47058823529412, "grad_norm": 1.5782941579818726, "learning_rate": 5.6280884079339296e-06, "loss": 0.0336, "step": 1113 }, { "epoch": 65.52941176470588, "grad_norm": 1.4799176454544067, "learning_rate": 5.610962029773703e-06, "loss": 0.0344, "step": 1114 }, { "epoch": 65.58823529411765, "grad_norm": 1.8770173788070679, "learning_rate": 5.593851582048688e-06, "loss": 0.0676, "step": 1115 }, { "epoch": 65.6470588235294, "grad_norm": 3.5710809230804443, "learning_rate": 5.5767571268629015e-06, "loss": 0.1042, "step": 1116 }, { "epoch": 65.70588235294117, "grad_norm": 1.8977634906768799, "learning_rate": 5.559678726262313e-06, "loss": 0.0538, "step": 1117 }, { "epoch": 65.76470588235294, "grad_norm": 2.393293857574463, "learning_rate": 5.542616442234618e-06, "loss": 0.1016, "step": 1118 }, { "epoch": 65.82352941176471, "grad_norm": 1.9154714345932007, "learning_rate": 5.5255703367090175e-06, "loss": 0.0801, "step": 1119 }, { "epoch": 65.88235294117646, "grad_norm": 1.4917478561401367, "learning_rate": 5.50854047155599e-06, "loss": 0.1235, "step": 1120 }, { "epoch": 65.94117647058823, "grad_norm": 3.041322708129883, "learning_rate": 5.491526908587067e-06, "loss": 0.0888, "step": 1121 }, { "epoch": 66.0, "grad_norm": 1.5157771110534668, "learning_rate": 5.4745297095546125e-06, "loss": 0.0442, "step": 1122 }, { "epoch": 66.05882352941177, "grad_norm": 2.1035757064819336, "learning_rate": 5.457548936151591e-06, "loss": 0.0512, "step": 1123 }, { "epoch": 66.11764705882354, "grad_norm": 2.0161569118499756, "learning_rate": 5.440584650011355e-06, "loss": 0.0386, "step": 1124 }, { "epoch": 66.17647058823529, "grad_norm": 2.3808257579803467, "learning_rate": 5.423636912707413e-06, "loss": 0.0802, "step": 1125 }, { "epoch": 66.23529411764706, "grad_norm": 2.3906702995300293, "learning_rate": 5.406705785753203e-06, "loss": 0.0742, "step": 1126 }, { "epoch": 66.29411764705883, "grad_norm": 1.872725486755371, "learning_rate": 5.389791330601883e-06, "loss": 0.0696, "step": 1127 }, { "epoch": 66.3529411764706, "grad_norm": 1.812416911125183, "learning_rate": 5.372893608646088e-06, "loss": 0.1112, "step": 1128 }, { "epoch": 66.41176470588235, "grad_norm": 1.8651305437088013, "learning_rate": 5.3560126812177245e-06, "loss": 0.1011, "step": 1129 }, { "epoch": 66.47058823529412, "grad_norm": 2.672950506210327, "learning_rate": 5.339148609587744e-06, "loss": 0.1508, "step": 1130 }, { "epoch": 66.52941176470588, "grad_norm": 2.594212055206299, "learning_rate": 5.322301454965914e-06, "loss": 0.1022, "step": 1131 }, { "epoch": 66.58823529411765, "grad_norm": 2.898777484893799, "learning_rate": 5.305471278500605e-06, "loss": 0.1145, "step": 1132 }, { "epoch": 66.6470588235294, "grad_norm": 2.3946032524108887, "learning_rate": 5.288658141278556e-06, "loss": 0.126, "step": 1133 }, { "epoch": 66.70588235294117, "grad_norm": 1.6420799493789673, "learning_rate": 5.271862104324665e-06, "loss": 0.0451, "step": 1134 }, { "epoch": 66.76470588235294, "grad_norm": 1.8424899578094482, "learning_rate": 5.255083228601765e-06, "loss": 0.0427, "step": 1135 }, { "epoch": 66.82352941176471, "grad_norm": 2.032358169555664, "learning_rate": 5.2383215750103985e-06, "loss": 0.0716, "step": 1136 }, { "epoch": 66.88235294117646, "grad_norm": 1.690839409828186, "learning_rate": 5.221577204388597e-06, "loss": 0.0516, "step": 1137 }, { "epoch": 66.94117647058823, "grad_norm": 1.4845377206802368, "learning_rate": 5.204850177511662e-06, "loss": 0.0324, "step": 1138 }, { "epoch": 67.0, "grad_norm": 1.849793553352356, "learning_rate": 5.18814055509195e-06, "loss": 0.0744, "step": 1139 }, { "epoch": 67.05882352941177, "grad_norm": 1.4078534841537476, "learning_rate": 5.171448397778637e-06, "loss": 0.0294, "step": 1140 }, { "epoch": 67.11764705882354, "grad_norm": 2.576775550842285, "learning_rate": 5.1547737661575194e-06, "loss": 0.1429, "step": 1141 }, { "epoch": 67.17647058823529, "grad_norm": 2.1244189739227295, "learning_rate": 5.138116720750766e-06, "loss": 0.0439, "step": 1142 }, { "epoch": 67.23529411764706, "grad_norm": 1.767438530921936, "learning_rate": 5.121477322016728e-06, "loss": 0.0547, "step": 1143 }, { "epoch": 67.29411764705883, "grad_norm": 1.8662372827529907, "learning_rate": 5.1048556303497046e-06, "loss": 0.0928, "step": 1144 }, { "epoch": 67.3529411764706, "grad_norm": 2.144318103790283, "learning_rate": 5.088251706079721e-06, "loss": 0.1045, "step": 1145 }, { "epoch": 67.41176470588235, "grad_norm": 1.3838058710098267, "learning_rate": 5.0716656094723184e-06, "loss": 0.0371, "step": 1146 }, { "epoch": 67.47058823529412, "grad_norm": 1.9961042404174805, "learning_rate": 5.055097400728328e-06, "loss": 0.0954, "step": 1147 }, { "epoch": 67.52941176470588, "grad_norm": 1.3730318546295166, "learning_rate": 5.038547139983657e-06, "loss": 0.052, "step": 1148 }, { "epoch": 67.58823529411765, "grad_norm": 16.746097564697266, "learning_rate": 5.022014887309067e-06, "loss": 0.0493, "step": 1149 }, { "epoch": 67.6470588235294, "grad_norm": 1.8709453344345093, "learning_rate": 5.005500702709958e-06, "loss": 0.0472, "step": 1150 }, { "epoch": 67.70588235294117, "grad_norm": 1.3543012142181396, "learning_rate": 4.98900464612615e-06, "loss": 0.0357, "step": 1151 }, { "epoch": 67.76470588235294, "grad_norm": 2.045917510986328, "learning_rate": 4.972526777431667e-06, "loss": 0.0466, "step": 1152 }, { "epoch": 67.82352941176471, "grad_norm": 1.9513531923294067, "learning_rate": 4.956067156434517e-06, "loss": 0.1224, "step": 1153 }, { "epoch": 67.88235294117646, "grad_norm": 2.677604913711548, "learning_rate": 4.939625842876479e-06, "loss": 0.1765, "step": 1154 }, { "epoch": 67.94117647058823, "grad_norm": 1.663493275642395, "learning_rate": 4.923202896432876e-06, "loss": 0.0854, "step": 1155 }, { "epoch": 68.0, "grad_norm": 1.3894809484481812, "learning_rate": 4.9067983767123736e-06, "loss": 0.0503, "step": 1156 }, { "epoch": 68.05882352941177, "grad_norm": 1.9582152366638184, "learning_rate": 4.890412343256753e-06, "loss": 0.0724, "step": 1157 }, { "epoch": 68.11764705882354, "grad_norm": 1.9910277128219604, "learning_rate": 4.8740448555406976e-06, "loss": 0.1723, "step": 1158 }, { "epoch": 68.17647058823529, "grad_norm": 2.205488920211792, "learning_rate": 4.857695972971573e-06, "loss": 0.0742, "step": 1159 }, { "epoch": 68.23529411764706, "grad_norm": 1.3329463005065918, "learning_rate": 4.841365754889229e-06, "loss": 0.0319, "step": 1160 }, { "epoch": 68.29411764705883, "grad_norm": 2.0146613121032715, "learning_rate": 4.8250542605657565e-06, "loss": 0.0838, "step": 1161 }, { "epoch": 68.3529411764706, "grad_norm": 3.302537679672241, "learning_rate": 4.808761549205294e-06, "loss": 0.1065, "step": 1162 }, { "epoch": 68.41176470588235, "grad_norm": 1.9014934301376343, "learning_rate": 4.7924876799438015e-06, "loss": 0.0347, "step": 1163 }, { "epoch": 68.47058823529412, "grad_norm": 1.803194284439087, "learning_rate": 4.776232711848857e-06, "loss": 0.0879, "step": 1164 }, { "epoch": 68.52941176470588, "grad_norm": 1.8186376094818115, "learning_rate": 4.759996703919423e-06, "loss": 0.0634, "step": 1165 }, { "epoch": 68.58823529411765, "grad_norm": 1.0831764936447144, "learning_rate": 4.743779715085652e-06, "loss": 0.0281, "step": 1166 }, { "epoch": 68.6470588235294, "grad_norm": 2.1017637252807617, "learning_rate": 4.7275818042086665e-06, "loss": 0.0988, "step": 1167 }, { "epoch": 68.70588235294117, "grad_norm": 2.5162010192871094, "learning_rate": 4.711403030080338e-06, "loss": 0.0964, "step": 1168 }, { "epoch": 68.76470588235294, "grad_norm": 1.8956377506256104, "learning_rate": 4.695243451423086e-06, "loss": 0.0828, "step": 1169 }, { "epoch": 68.82352941176471, "grad_norm": 1.629029631614685, "learning_rate": 4.679103126889653e-06, "loss": 0.0789, "step": 1170 }, { "epoch": 68.88235294117646, "grad_norm": 2.241673231124878, "learning_rate": 4.662982115062899e-06, "loss": 0.0664, "step": 1171 }, { "epoch": 68.94117647058823, "grad_norm": 1.2547017335891724, "learning_rate": 4.646880474455584e-06, "loss": 0.0367, "step": 1172 }, { "epoch": 69.0, "grad_norm": 2.338430404663086, "learning_rate": 4.630798263510162e-06, "loss": 0.0423, "step": 1173 }, { "epoch": 69.05882352941177, "grad_norm": 1.8219914436340332, "learning_rate": 4.614735540598564e-06, "loss": 0.0434, "step": 1174 }, { "epoch": 69.11764705882354, "grad_norm": 2.6298296451568604, "learning_rate": 4.598692364021985e-06, "loss": 0.1203, "step": 1175 }, { "epoch": 69.17647058823529, "grad_norm": 1.4541008472442627, "learning_rate": 4.582668792010679e-06, "loss": 0.0677, "step": 1176 }, { "epoch": 69.23529411764706, "grad_norm": 1.6225227117538452, "learning_rate": 4.566664882723739e-06, "loss": 0.081, "step": 1177 }, { "epoch": 69.29411764705883, "grad_norm": 1.3577585220336914, "learning_rate": 4.550680694248897e-06, "loss": 0.0343, "step": 1178 }, { "epoch": 69.3529411764706, "grad_norm": 2.228172540664673, "learning_rate": 4.534716284602294e-06, "loss": 0.0934, "step": 1179 }, { "epoch": 69.41176470588235, "grad_norm": 1.8923990726470947, "learning_rate": 4.518771711728293e-06, "loss": 0.1023, "step": 1180 }, { "epoch": 69.47058823529412, "grad_norm": 2.4779233932495117, "learning_rate": 4.502847033499252e-06, "loss": 0.0556, "step": 1181 }, { "epoch": 69.52941176470588, "grad_norm": 1.6250523328781128, "learning_rate": 4.486942307715325e-06, "loss": 0.0311, "step": 1182 }, { "epoch": 69.58823529411765, "grad_norm": 1.8590589761734009, "learning_rate": 4.47105759210424e-06, "loss": 0.0635, "step": 1183 }, { "epoch": 69.6470588235294, "grad_norm": 2.300161838531494, "learning_rate": 4.455192944321101e-06, "loss": 0.1783, "step": 1184 }, { "epoch": 69.70588235294117, "grad_norm": 2.7942347526550293, "learning_rate": 4.439348421948172e-06, "loss": 0.0332, "step": 1185 }, { "epoch": 69.76470588235294, "grad_norm": 1.598071813583374, "learning_rate": 4.423524082494669e-06, "loss": 0.044, "step": 1186 }, { "epoch": 69.82352941176471, "grad_norm": 1.5257189273834229, "learning_rate": 4.407719983396555e-06, "loss": 0.0478, "step": 1187 }, { "epoch": 69.88235294117646, "grad_norm": 2.225369453430176, "learning_rate": 4.3919361820163244e-06, "loss": 0.1063, "step": 1188 }, { "epoch": 69.94117647058823, "grad_norm": 2.1886301040649414, "learning_rate": 4.376172735642803e-06, "loss": 0.0935, "step": 1189 }, { "epoch": 70.0, "grad_norm": 1.5718376636505127, "learning_rate": 4.360429701490935e-06, "loss": 0.0514, "step": 1190 }, { "epoch": 70.05882352941177, "grad_norm": 2.1521449089050293, "learning_rate": 4.344707136701579e-06, "loss": 0.095, "step": 1191 }, { "epoch": 70.11764705882354, "grad_norm": 1.5383597612380981, "learning_rate": 4.329005098341288e-06, "loss": 0.0825, "step": 1192 }, { "epoch": 70.17647058823529, "grad_norm": 1.4976763725280762, "learning_rate": 4.313323643402124e-06, "loss": 0.0499, "step": 1193 }, { "epoch": 70.23529411764706, "grad_norm": 2.4064793586730957, "learning_rate": 4.297662828801434e-06, "loss": 0.057, "step": 1194 }, { "epoch": 70.29411764705883, "grad_norm": 1.7607786655426025, "learning_rate": 4.2820227113816525e-06, "loss": 0.1061, "step": 1195 }, { "epoch": 70.3529411764706, "grad_norm": 2.0553324222564697, "learning_rate": 4.266403347910086e-06, "loss": 0.2041, "step": 1196 }, { "epoch": 70.41176470588235, "grad_norm": 1.5513496398925781, "learning_rate": 4.250804795078717e-06, "loss": 0.037, "step": 1197 }, { "epoch": 70.47058823529412, "grad_norm": 1.1724457740783691, "learning_rate": 4.235227109503992e-06, "loss": 0.0226, "step": 1198 }, { "epoch": 70.52941176470588, "grad_norm": 1.627712368965149, "learning_rate": 4.2196703477266185e-06, "loss": 0.0838, "step": 1199 }, { "epoch": 70.58823529411765, "grad_norm": 1.4119625091552734, "learning_rate": 4.20413456621136e-06, "loss": 0.0317, "step": 1200 }, { "epoch": 70.6470588235294, "grad_norm": 1.1626073122024536, "learning_rate": 4.188619821346828e-06, "loss": 0.0297, "step": 1201 }, { "epoch": 70.70588235294117, "grad_norm": 2.311382532119751, "learning_rate": 4.173126169445279e-06, "loss": 0.0764, "step": 1202 }, { "epoch": 70.76470588235294, "grad_norm": 1.930677890777588, "learning_rate": 4.157653666742413e-06, "loss": 0.0482, "step": 1203 }, { "epoch": 70.82352941176471, "grad_norm": 2.024998426437378, "learning_rate": 4.142202369397169e-06, "loss": 0.1183, "step": 1204 }, { "epoch": 70.88235294117646, "grad_norm": 1.8030592203140259, "learning_rate": 4.12677233349151e-06, "loss": 0.0624, "step": 1205 }, { "epoch": 70.94117647058823, "grad_norm": 1.577624797821045, "learning_rate": 4.111363615030238e-06, "loss": 0.0825, "step": 1206 }, { "epoch": 71.0, "grad_norm": 1.8432421684265137, "learning_rate": 4.095976269940777e-06, "loss": 0.0667, "step": 1207 }, { "epoch": 71.05882352941177, "grad_norm": 1.5795820951461792, "learning_rate": 4.080610354072975e-06, "loss": 0.0306, "step": 1208 }, { "epoch": 71.11764705882354, "grad_norm": 1.5941683053970337, "learning_rate": 4.065265923198901e-06, "loss": 0.0356, "step": 1209 }, { "epoch": 71.17647058823529, "grad_norm": 2.042832851409912, "learning_rate": 4.0499430330126425e-06, "loss": 0.1845, "step": 1210 }, { "epoch": 71.23529411764706, "grad_norm": 1.471020221710205, "learning_rate": 4.0346417391300985e-06, "loss": 0.0395, "step": 1211 }, { "epoch": 71.29411764705883, "grad_norm": 1.9737622737884521, "learning_rate": 4.019362097088793e-06, "loss": 0.0639, "step": 1212 }, { "epoch": 71.3529411764706, "grad_norm": 1.542229175567627, "learning_rate": 4.00410416234765e-06, "loss": 0.094, "step": 1213 }, { "epoch": 71.41176470588235, "grad_norm": 3.9951534271240234, "learning_rate": 3.988867990286814e-06, "loss": 0.0864, "step": 1214 }, { "epoch": 71.47058823529412, "grad_norm": 2.132652759552002, "learning_rate": 3.973653636207437e-06, "loss": 0.0357, "step": 1215 }, { "epoch": 71.52941176470588, "grad_norm": 1.830737829208374, "learning_rate": 3.958461155331475e-06, "loss": 0.0899, "step": 1216 }, { "epoch": 71.58823529411765, "grad_norm": 1.3480134010314941, "learning_rate": 3.943290602801499e-06, "loss": 0.0583, "step": 1217 }, { "epoch": 71.6470588235294, "grad_norm": 1.6048173904418945, "learning_rate": 3.928142033680487e-06, "loss": 0.0818, "step": 1218 }, { "epoch": 71.70588235294117, "grad_norm": 1.838568925857544, "learning_rate": 3.913015502951629e-06, "loss": 0.0397, "step": 1219 }, { "epoch": 71.76470588235294, "grad_norm": 1.7728995084762573, "learning_rate": 3.897911065518122e-06, "loss": 0.0413, "step": 1220 }, { "epoch": 71.82352941176471, "grad_norm": 1.8261284828186035, "learning_rate": 3.8828287762029725e-06, "loss": 0.088, "step": 1221 }, { "epoch": 71.88235294117646, "grad_norm": 1.8425546884536743, "learning_rate": 3.8677686897488e-06, "loss": 0.0303, "step": 1222 }, { "epoch": 71.94117647058823, "grad_norm": 1.7914022207260132, "learning_rate": 3.852730860817636e-06, "loss": 0.1049, "step": 1223 }, { "epoch": 72.0, "grad_norm": 1.736825704574585, "learning_rate": 3.837715343990727e-06, "loss": 0.0901, "step": 1224 }, { "epoch": 72.05882352941177, "grad_norm": 1.5525181293487549, "learning_rate": 3.822722193768334e-06, "loss": 0.0343, "step": 1225 }, { "epoch": 72.11764705882354, "grad_norm": 1.4110966920852661, "learning_rate": 3.8077514645695347e-06, "loss": 0.0324, "step": 1226 }, { "epoch": 72.17647058823529, "grad_norm": 1.9844331741333008, "learning_rate": 3.7928032107320333e-06, "loss": 0.1027, "step": 1227 }, { "epoch": 72.23529411764706, "grad_norm": 1.6649786233901978, "learning_rate": 3.777877486511955e-06, "loss": 0.1189, "step": 1228 }, { "epoch": 72.29411764705883, "grad_norm": 1.8845796585083008, "learning_rate": 3.7629743460836455e-06, "loss": 0.0994, "step": 1229 }, { "epoch": 72.3529411764706, "grad_norm": 2.4299497604370117, "learning_rate": 3.748093843539489e-06, "loss": 0.0939, "step": 1230 }, { "epoch": 72.41176470588235, "grad_norm": 1.6283212900161743, "learning_rate": 3.7332360328896975e-06, "loss": 0.0924, "step": 1231 }, { "epoch": 72.47058823529412, "grad_norm": 2.2074055671691895, "learning_rate": 3.7184009680621256e-06, "loss": 0.033, "step": 1232 }, { "epoch": 72.52941176470588, "grad_norm": 1.463223934173584, "learning_rate": 3.703588702902066e-06, "loss": 0.041, "step": 1233 }, { "epoch": 72.58823529411765, "grad_norm": 1.794360637664795, "learning_rate": 3.6887992911720593e-06, "loss": 0.0729, "step": 1234 }, { "epoch": 72.6470588235294, "grad_norm": 2.0354204177856445, "learning_rate": 3.6740327865516967e-06, "loss": 0.0778, "step": 1235 }, { "epoch": 72.70588235294117, "grad_norm": 1.702148675918579, "learning_rate": 3.659289242637426e-06, "loss": 0.035, "step": 1236 }, { "epoch": 72.76470588235294, "grad_norm": 1.955505132675171, "learning_rate": 3.644568712942358e-06, "loss": 0.0506, "step": 1237 }, { "epoch": 72.82352941176471, "grad_norm": 2.317246675491333, "learning_rate": 3.629871250896069e-06, "loss": 0.1168, "step": 1238 }, { "epoch": 72.88235294117646, "grad_norm": 2.1656157970428467, "learning_rate": 3.6151969098444108e-06, "loss": 0.0726, "step": 1239 }, { "epoch": 72.94117647058823, "grad_norm": 1.7713035345077515, "learning_rate": 3.600545743049315e-06, "loss": 0.1032, "step": 1240 }, { "epoch": 73.0, "grad_norm": 1.4963555335998535, "learning_rate": 3.585917803688603e-06, "loss": 0.0348, "step": 1241 }, { "epoch": 73.05882352941177, "grad_norm": 1.6327489614486694, "learning_rate": 3.5713131448557814e-06, "loss": 0.074, "step": 1242 }, { "epoch": 73.11764705882354, "grad_norm": 1.5044529438018799, "learning_rate": 3.5567318195598643e-06, "loss": 0.0747, "step": 1243 }, { "epoch": 73.17647058823529, "grad_norm": 2.0979466438293457, "learning_rate": 3.542173880725176e-06, "loss": 0.0897, "step": 1244 }, { "epoch": 73.23529411764706, "grad_norm": 6.597360610961914, "learning_rate": 3.5276393811911524e-06, "loss": 0.0599, "step": 1245 }, { "epoch": 73.29411764705883, "grad_norm": 1.9030721187591553, "learning_rate": 3.5131283737121592e-06, "loss": 0.0734, "step": 1246 }, { "epoch": 73.3529411764706, "grad_norm": 1.4432328939437866, "learning_rate": 3.4986409109572906e-06, "loss": 0.0372, "step": 1247 }, { "epoch": 73.41176470588235, "grad_norm": 1.6196798086166382, "learning_rate": 3.4841770455101875e-06, "loss": 0.1331, "step": 1248 }, { "epoch": 73.47058823529412, "grad_norm": 1.7272660732269287, "learning_rate": 3.4697368298688374e-06, "loss": 0.0525, "step": 1249 }, { "epoch": 73.52941176470588, "grad_norm": 1.8266502618789673, "learning_rate": 3.455320316445393e-06, "loss": 0.0869, "step": 1250 }, { "epoch": 73.58823529411765, "grad_norm": 1.6594314575195312, "learning_rate": 3.4409275575659763e-06, "loss": 0.0396, "step": 1251 }, { "epoch": 73.6470588235294, "grad_norm": 1.5694966316223145, "learning_rate": 3.4265586054704857e-06, "loss": 0.0837, "step": 1252 }, { "epoch": 73.70588235294117, "grad_norm": 1.5278329849243164, "learning_rate": 3.412213512312418e-06, "loss": 0.0403, "step": 1253 }, { "epoch": 73.76470588235294, "grad_norm": 1.3905272483825684, "learning_rate": 3.3978923301586654e-06, "loss": 0.0485, "step": 1254 }, { "epoch": 73.82352941176471, "grad_norm": 1.9646095037460327, "learning_rate": 3.3835951109893395e-06, "loss": 0.1128, "step": 1255 }, { "epoch": 73.88235294117646, "grad_norm": 2.0984232425689697, "learning_rate": 3.3693219066975635e-06, "loss": 0.0581, "step": 1256 }, { "epoch": 73.94117647058823, "grad_norm": 1.7349494695663452, "learning_rate": 3.3550727690893104e-06, "loss": 0.0428, "step": 1257 }, { "epoch": 74.0, "grad_norm": 1.8324089050292969, "learning_rate": 3.3408477498831917e-06, "loss": 0.0719, "step": 1258 }, { "epoch": 74.05882352941177, "grad_norm": 1.4679962396621704, "learning_rate": 3.326646900710284e-06, "loss": 0.0383, "step": 1259 }, { "epoch": 74.11764705882354, "grad_norm": 1.291725993156433, "learning_rate": 3.3124702731139345e-06, "loss": 0.034, "step": 1260 }, { "epoch": 74.17647058823529, "grad_norm": 1.7393850088119507, "learning_rate": 3.298317918549575e-06, "loss": 0.0423, "step": 1261 }, { "epoch": 74.23529411764706, "grad_norm": 2.5063884258270264, "learning_rate": 3.2841898883845335e-06, "loss": 0.077, "step": 1262 }, { "epoch": 74.29411764705883, "grad_norm": 1.4715299606323242, "learning_rate": 3.270086233897861e-06, "loss": 0.0278, "step": 1263 }, { "epoch": 74.3529411764706, "grad_norm": 1.73170006275177, "learning_rate": 3.256007006280124e-06, "loss": 0.0556, "step": 1264 }, { "epoch": 74.41176470588235, "grad_norm": 1.4847959280014038, "learning_rate": 3.241952256633234e-06, "loss": 0.0592, "step": 1265 }, { "epoch": 74.47058823529412, "grad_norm": 2.001354217529297, "learning_rate": 3.2279220359702513e-06, "loss": 0.1242, "step": 1266 }, { "epoch": 74.52941176470588, "grad_norm": 1.7341153621673584, "learning_rate": 3.2139163952152117e-06, "loss": 0.0553, "step": 1267 }, { "epoch": 74.58823529411765, "grad_norm": 1.276216745376587, "learning_rate": 3.199935385202937e-06, "loss": 0.0365, "step": 1268 }, { "epoch": 74.6470588235294, "grad_norm": 1.2654128074645996, "learning_rate": 3.1859790566788464e-06, "loss": 0.0322, "step": 1269 }, { "epoch": 74.70588235294117, "grad_norm": 2.0905582904815674, "learning_rate": 3.172047460298774e-06, "loss": 0.0867, "step": 1270 }, { "epoch": 74.76470588235294, "grad_norm": 1.9313535690307617, "learning_rate": 3.15814064662879e-06, "loss": 0.0958, "step": 1271 }, { "epoch": 74.82352941176471, "grad_norm": 1.910239815711975, "learning_rate": 3.1442586661450102e-06, "loss": 0.2195, "step": 1272 }, { "epoch": 74.88235294117646, "grad_norm": 2.0258758068084717, "learning_rate": 3.130401569233418e-06, "loss": 0.0687, "step": 1273 }, { "epoch": 74.94117647058823, "grad_norm": 2.1090497970581055, "learning_rate": 3.1165694061896788e-06, "loss": 0.0319, "step": 1274 }, { "epoch": 75.0, "grad_norm": 1.0852181911468506, "learning_rate": 3.1027622272189572e-06, "loss": 0.0389, "step": 1275 }, { "epoch": 75.05882352941177, "grad_norm": 4.0622148513793945, "learning_rate": 3.0889800824357374e-06, "loss": 0.0649, "step": 1276 }, { "epoch": 75.11764705882354, "grad_norm": 1.7910910844802856, "learning_rate": 3.075223021863638e-06, "loss": 0.1512, "step": 1277 }, { "epoch": 75.17647058823529, "grad_norm": 6.50875186920166, "learning_rate": 3.0614910954352363e-06, "loss": 0.0879, "step": 1278 }, { "epoch": 75.23529411764706, "grad_norm": 1.5753756761550903, "learning_rate": 3.047784352991873e-06, "loss": 0.0546, "step": 1279 }, { "epoch": 75.29411764705883, "grad_norm": 1.6199556589126587, "learning_rate": 3.0341028442834897e-06, "loss": 0.0475, "step": 1280 }, { "epoch": 75.3529411764706, "grad_norm": 1.9946000576019287, "learning_rate": 3.020446618968438e-06, "loss": 0.0566, "step": 1281 }, { "epoch": 75.41176470588235, "grad_norm": 1.9650324583053589, "learning_rate": 3.0068157266133004e-06, "loss": 0.1299, "step": 1282 }, { "epoch": 75.47058823529412, "grad_norm": 1.4884922504425049, "learning_rate": 2.9932102166927114e-06, "loss": 0.0375, "step": 1283 }, { "epoch": 75.52941176470588, "grad_norm": 1.035851240158081, "learning_rate": 2.9796301385891767e-06, "loss": 0.0266, "step": 1284 }, { "epoch": 75.58823529411765, "grad_norm": 1.4468518495559692, "learning_rate": 2.9660755415928955e-06, "loss": 0.0349, "step": 1285 }, { "epoch": 75.6470588235294, "grad_norm": 1.5248000621795654, "learning_rate": 2.9525464749015797e-06, "loss": 0.0794, "step": 1286 }, { "epoch": 75.70588235294117, "grad_norm": 1.2782536745071411, "learning_rate": 2.9390429876202788e-06, "loss": 0.04, "step": 1287 }, { "epoch": 75.76470588235294, "grad_norm": 1.728056788444519, "learning_rate": 2.9255651287611964e-06, "loss": 0.0624, "step": 1288 }, { "epoch": 75.82352941176471, "grad_norm": 1.6487969160079956, "learning_rate": 2.9121129472435183e-06, "loss": 0.0332, "step": 1289 }, { "epoch": 75.88235294117646, "grad_norm": 1.7404379844665527, "learning_rate": 2.898686491893229e-06, "loss": 0.0605, "step": 1290 }, { "epoch": 75.94117647058823, "grad_norm": 2.5000126361846924, "learning_rate": 2.885285811442938e-06, "loss": 0.0689, "step": 1291 }, { "epoch": 76.0, "grad_norm": 1.353662371635437, "learning_rate": 2.8719109545317102e-06, "loss": 0.0696, "step": 1292 }, { "epoch": 76.05882352941177, "grad_norm": 1.4691040515899658, "learning_rate": 2.858561969704864e-06, "loss": 0.0683, "step": 1293 }, { "epoch": 76.11764705882354, "grad_norm": 1.7759066820144653, "learning_rate": 2.8452389054138284e-06, "loss": 0.1016, "step": 1294 }, { "epoch": 76.17647058823529, "grad_norm": 1.5680325031280518, "learning_rate": 2.8319418100159456e-06, "loss": 0.0427, "step": 1295 }, { "epoch": 76.23529411764706, "grad_norm": 1.451442837715149, "learning_rate": 2.8186707317743e-06, "loss": 0.0585, "step": 1296 }, { "epoch": 76.29411764705883, "grad_norm": 1.9467896223068237, "learning_rate": 2.8054257188575486e-06, "loss": 0.1222, "step": 1297 }, { "epoch": 76.3529411764706, "grad_norm": 1.577833652496338, "learning_rate": 2.792206819339737e-06, "loss": 0.083, "step": 1298 }, { "epoch": 76.41176470588235, "grad_norm": 3.209770679473877, "learning_rate": 2.779014081200132e-06, "loss": 0.0759, "step": 1299 }, { "epoch": 76.47058823529412, "grad_norm": 1.6940313577651978, "learning_rate": 2.7658475523230444e-06, "loss": 0.0696, "step": 1300 }, { "epoch": 76.52941176470588, "grad_norm": 1.5390088558197021, "learning_rate": 2.7527072804976586e-06, "loss": 0.0564, "step": 1301 }, { "epoch": 76.58823529411765, "grad_norm": 1.52656090259552, "learning_rate": 2.739593313417851e-06, "loss": 0.0375, "step": 1302 }, { "epoch": 76.6470588235294, "grad_norm": 1.3220809698104858, "learning_rate": 2.726505698682028e-06, "loss": 0.0396, "step": 1303 }, { "epoch": 76.70588235294117, "grad_norm": 1.095950961112976, "learning_rate": 2.713444483792945e-06, "loss": 0.0613, "step": 1304 }, { "epoch": 76.76470588235294, "grad_norm": 1.8308571577072144, "learning_rate": 2.7004097161575393e-06, "loss": 0.0657, "step": 1305 }, { "epoch": 76.82352941176471, "grad_norm": 1.3979910612106323, "learning_rate": 2.6874014430867478e-06, "loss": 0.0357, "step": 1306 }, { "epoch": 76.88235294117646, "grad_norm": 1.3929507732391357, "learning_rate": 2.6744197117953506e-06, "loss": 0.0253, "step": 1307 }, { "epoch": 76.94117647058823, "grad_norm": 2.0179033279418945, "learning_rate": 2.661464569401789e-06, "loss": 0.0898, "step": 1308 }, { "epoch": 77.0, "grad_norm": 2.591480255126953, "learning_rate": 2.648536062927999e-06, "loss": 0.1177, "step": 1309 }, { "epoch": 77.05882352941177, "grad_norm": 2.1703028678894043, "learning_rate": 2.635634239299235e-06, "loss": 0.0617, "step": 1310 }, { "epoch": 77.11764705882354, "grad_norm": 1.9598915576934814, "learning_rate": 2.6227591453439094e-06, "loss": 0.0951, "step": 1311 }, { "epoch": 77.17647058823529, "grad_norm": 2.195244312286377, "learning_rate": 2.6099108277934105e-06, "loss": 0.1163, "step": 1312 }, { "epoch": 77.23529411764706, "grad_norm": 1.311383843421936, "learning_rate": 2.5970893332819436e-06, "loss": 0.0396, "step": 1313 }, { "epoch": 77.29411764705883, "grad_norm": 1.6745401620864868, "learning_rate": 2.584294708346352e-06, "loss": 0.0677, "step": 1314 }, { "epoch": 77.3529411764706, "grad_norm": 1.4427839517593384, "learning_rate": 2.571526999425965e-06, "loss": 0.0311, "step": 1315 }, { "epoch": 77.41176470588235, "grad_norm": 1.2203447818756104, "learning_rate": 2.558786252862402e-06, "loss": 0.0333, "step": 1316 }, { "epoch": 77.47058823529412, "grad_norm": 1.6183561086654663, "learning_rate": 2.5460725148994292e-06, "loss": 0.0602, "step": 1317 }, { "epoch": 77.52941176470588, "grad_norm": 1.4406027793884277, "learning_rate": 2.533385831682782e-06, "loss": 0.0851, "step": 1318 }, { "epoch": 77.58823529411765, "grad_norm": 1.628800630569458, "learning_rate": 2.5207262492599947e-06, "loss": 0.0849, "step": 1319 }, { "epoch": 77.6470588235294, "grad_norm": 1.5525051355361938, "learning_rate": 2.50809381358024e-06, "loss": 0.034, "step": 1320 }, { "epoch": 77.70588235294117, "grad_norm": 1.8759832382202148, "learning_rate": 2.495488570494157e-06, "loss": 0.0985, "step": 1321 }, { "epoch": 77.76470588235294, "grad_norm": 3.2046866416931152, "learning_rate": 2.4829105657536866e-06, "loss": 0.0976, "step": 1322 }, { "epoch": 77.82352941176471, "grad_norm": 2.035099744796753, "learning_rate": 2.470359845011908e-06, "loss": 0.0507, "step": 1323 }, { "epoch": 77.88235294117646, "grad_norm": 1.7447682619094849, "learning_rate": 2.457836453822867e-06, "loss": 0.0822, "step": 1324 }, { "epoch": 77.94117647058823, "grad_norm": 2.1366026401519775, "learning_rate": 2.445340437641416e-06, "loss": 0.046, "step": 1325 }, { "epoch": 78.0, "grad_norm": 1.96588933467865, "learning_rate": 2.432871841823047e-06, "loss": 0.0796, "step": 1326 }, { "epoch": 78.05882352941177, "grad_norm": 1.6314164400100708, "learning_rate": 2.420430711623727e-06, "loss": 0.0857, "step": 1327 }, { "epoch": 78.11764705882354, "grad_norm": 1.3423418998718262, "learning_rate": 2.408017092199735e-06, "loss": 0.0308, "step": 1328 }, { "epoch": 78.17647058823529, "grad_norm": 1.8659029006958008, "learning_rate": 2.395631028607497e-06, "loss": 0.0851, "step": 1329 }, { "epoch": 78.23529411764706, "grad_norm": 1.9361073970794678, "learning_rate": 2.3832725658034163e-06, "loss": 0.0441, "step": 1330 }, { "epoch": 78.29411764705883, "grad_norm": 1.5438425540924072, "learning_rate": 2.3709417486437246e-06, "loss": 0.0403, "step": 1331 }, { "epoch": 78.3529411764706, "grad_norm": 1.7671531438827515, "learning_rate": 2.3586386218843096e-06, "loss": 0.0979, "step": 1332 }, { "epoch": 78.41176470588235, "grad_norm": 13.083359718322754, "learning_rate": 2.3463632301805505e-06, "loss": 0.0932, "step": 1333 }, { "epoch": 78.47058823529412, "grad_norm": 1.6415494680404663, "learning_rate": 2.3341156180871626e-06, "loss": 0.0819, "step": 1334 }, { "epoch": 78.52941176470588, "grad_norm": 2.312978744506836, "learning_rate": 2.3218958300580307e-06, "loss": 0.1749, "step": 1335 }, { "epoch": 78.58823529411765, "grad_norm": 1.8194575309753418, "learning_rate": 2.3097039104460506e-06, "loss": 0.0617, "step": 1336 }, { "epoch": 78.6470588235294, "grad_norm": 1.676568627357483, "learning_rate": 2.2975399035029657e-06, "loss": 0.0476, "step": 1337 }, { "epoch": 78.70588235294117, "grad_norm": 1.986046552658081, "learning_rate": 2.285403853379209e-06, "loss": 0.0544, "step": 1338 }, { "epoch": 78.76470588235294, "grad_norm": 1.2377493381500244, "learning_rate": 2.2732958041237417e-06, "loss": 0.0468, "step": 1339 }, { "epoch": 78.82352941176471, "grad_norm": 1.8677095174789429, "learning_rate": 2.2612157996838913e-06, "loss": 0.0596, "step": 1340 }, { "epoch": 78.88235294117646, "grad_norm": 1.8034330606460571, "learning_rate": 2.2491638839051953e-06, "loss": 0.0588, "step": 1341 }, { "epoch": 78.94117647058823, "grad_norm": 1.5654281377792358, "learning_rate": 2.237140100531244e-06, "loss": 0.0498, "step": 1342 }, { "epoch": 79.0, "grad_norm": 1.3761240243911743, "learning_rate": 2.2251444932035094e-06, "loss": 0.04, "step": 1343 }, { "epoch": 79.05882352941177, "grad_norm": 1.6407970190048218, "learning_rate": 2.2131771054612028e-06, "loss": 0.1157, "step": 1344 }, { "epoch": 79.11764705882354, "grad_norm": 1.581857681274414, "learning_rate": 2.20123798074111e-06, "loss": 0.0479, "step": 1345 }, { "epoch": 79.17647058823529, "grad_norm": 1.7815468311309814, "learning_rate": 2.1893271623774303e-06, "loss": 0.0589, "step": 1346 }, { "epoch": 79.23529411764706, "grad_norm": 1.942599892616272, "learning_rate": 2.177444693601626e-06, "loss": 0.0913, "step": 1347 }, { "epoch": 79.29411764705883, "grad_norm": 2.5264101028442383, "learning_rate": 2.165590617542258e-06, "loss": 0.1203, "step": 1348 }, { "epoch": 79.3529411764706, "grad_norm": 1.4449893236160278, "learning_rate": 2.153764977224836e-06, "loss": 0.0635, "step": 1349 }, { "epoch": 79.41176470588235, "grad_norm": 1.6911612749099731, "learning_rate": 2.141967815571656e-06, "loss": 0.0247, "step": 1350 }, { "epoch": 79.47058823529412, "grad_norm": 1.4627761840820312, "learning_rate": 2.1301991754016537e-06, "loss": 0.1078, "step": 1351 }, { "epoch": 79.52941176470588, "grad_norm": 2.4630050659179688, "learning_rate": 2.118459099430237e-06, "loss": 0.0512, "step": 1352 }, { "epoch": 79.58823529411765, "grad_norm": 1.50606369972229, "learning_rate": 2.1067476302691416e-06, "loss": 0.0285, "step": 1353 }, { "epoch": 79.6470588235294, "grad_norm": 1.4844212532043457, "learning_rate": 2.0950648104262694e-06, "loss": 0.0303, "step": 1354 }, { "epoch": 79.70588235294117, "grad_norm": 2.4108893871307373, "learning_rate": 2.0834106823055432e-06, "loss": 0.0974, "step": 1355 }, { "epoch": 79.76470588235294, "grad_norm": 1.1597278118133545, "learning_rate": 2.0717852882067334e-06, "loss": 0.0266, "step": 1356 }, { "epoch": 79.82352941176471, "grad_norm": 1.9524753093719482, "learning_rate": 2.0601886703253284e-06, "loss": 0.0828, "step": 1357 }, { "epoch": 79.88235294117646, "grad_norm": 2.2617900371551514, "learning_rate": 2.0486208707523692e-06, "loss": 0.0705, "step": 1358 }, { "epoch": 79.94117647058823, "grad_norm": 2.7905962467193604, "learning_rate": 2.037081931474294e-06, "loss": 0.039, "step": 1359 }, { "epoch": 80.0, "grad_norm": 2.040579319000244, "learning_rate": 2.025571894372794e-06, "loss": 0.0769, "step": 1360 }, { "epoch": 80.05882352941177, "grad_norm": 1.709731936454773, "learning_rate": 2.014090801224654e-06, "loss": 0.0597, "step": 1361 }, { "epoch": 80.11764705882354, "grad_norm": 1.5002096891403198, "learning_rate": 2.002638693701606e-06, "loss": 0.0319, "step": 1362 }, { "epoch": 80.17647058823529, "grad_norm": 1.9935553073883057, "learning_rate": 1.9912156133701753e-06, "loss": 0.0575, "step": 1363 }, { "epoch": 80.23529411764706, "grad_norm": 1.7752723693847656, "learning_rate": 1.9798216016915295e-06, "loss": 0.0431, "step": 1364 }, { "epoch": 80.29411764705883, "grad_norm": 1.8962852954864502, "learning_rate": 1.9684567000213295e-06, "loss": 0.0469, "step": 1365 }, { "epoch": 80.3529411764706, "grad_norm": 1.930747151374817, "learning_rate": 1.957120949609578e-06, "loss": 0.0681, "step": 1366 }, { "epoch": 80.41176470588235, "grad_norm": 1.3271692991256714, "learning_rate": 1.9458143916004713e-06, "loss": 0.0754, "step": 1367 }, { "epoch": 80.47058823529412, "grad_norm": 1.3746063709259033, "learning_rate": 1.934537067032247e-06, "loss": 0.0326, "step": 1368 }, { "epoch": 80.52941176470588, "grad_norm": 1.4813778400421143, "learning_rate": 1.923289016837041e-06, "loss": 0.0694, "step": 1369 }, { "epoch": 80.58823529411765, "grad_norm": 1.5063945055007935, "learning_rate": 1.9120702818407322e-06, "loss": 0.0307, "step": 1370 }, { "epoch": 80.6470588235294, "grad_norm": 2.3628652095794678, "learning_rate": 1.9008809027627962e-06, "loss": 0.0639, "step": 1371 }, { "epoch": 80.70588235294117, "grad_norm": 1.6431548595428467, "learning_rate": 1.8897209202161615e-06, "loss": 0.066, "step": 1372 }, { "epoch": 80.76470588235294, "grad_norm": 1.572918176651001, "learning_rate": 1.8785903747070566e-06, "loss": 0.0806, "step": 1373 }, { "epoch": 80.82352941176471, "grad_norm": 1.7661786079406738, "learning_rate": 1.8674893066348665e-06, "loss": 0.0997, "step": 1374 }, { "epoch": 80.88235294117646, "grad_norm": 1.7099701166152954, "learning_rate": 1.856417756291985e-06, "loss": 0.0778, "step": 1375 }, { "epoch": 80.94117647058823, "grad_norm": 1.5696171522140503, "learning_rate": 1.8453757638636683e-06, "loss": 0.0529, "step": 1376 }, { "epoch": 81.0, "grad_norm": 1.9857321977615356, "learning_rate": 1.8343633694278895e-06, "loss": 0.0684, "step": 1377 }, { "epoch": 81.05882352941177, "grad_norm": 1.3047956228256226, "learning_rate": 1.8233806129551935e-06, "loss": 0.0686, "step": 1378 }, { "epoch": 81.11764705882354, "grad_norm": 1.4102740287780762, "learning_rate": 1.8124275343085518e-06, "loss": 0.0523, "step": 1379 }, { "epoch": 81.17647058823529, "grad_norm": 1.1121536493301392, "learning_rate": 1.8015041732432116e-06, "loss": 0.0298, "step": 1380 }, { "epoch": 81.23529411764706, "grad_norm": 1.3087522983551025, "learning_rate": 1.7906105694065646e-06, "loss": 0.0554, "step": 1381 }, { "epoch": 81.29411764705883, "grad_norm": 1.5270016193389893, "learning_rate": 1.7797467623379927e-06, "loss": 0.0378, "step": 1382 }, { "epoch": 81.3529411764706, "grad_norm": 1.758679747581482, "learning_rate": 1.7689127914687288e-06, "loss": 0.1009, "step": 1383 }, { "epoch": 81.41176470588235, "grad_norm": 1.480966567993164, "learning_rate": 1.758108696121711e-06, "loss": 0.0525, "step": 1384 }, { "epoch": 81.47058823529412, "grad_norm": 2.3634719848632812, "learning_rate": 1.7473345155114429e-06, "loss": 0.0432, "step": 1385 }, { "epoch": 81.52941176470588, "grad_norm": 1.4587624073028564, "learning_rate": 1.736590288743847e-06, "loss": 0.0363, "step": 1386 }, { "epoch": 81.58823529411765, "grad_norm": 1.8332316875457764, "learning_rate": 1.725876054816128e-06, "loss": 0.1525, "step": 1387 }, { "epoch": 81.6470588235294, "grad_norm": 1.7769784927368164, "learning_rate": 1.7151918526166277e-06, "loss": 0.0501, "step": 1388 }, { "epoch": 81.70588235294117, "grad_norm": 1.7982438802719116, "learning_rate": 1.7045377209246849e-06, "loss": 0.0426, "step": 1389 }, { "epoch": 81.76470588235294, "grad_norm": 2.3104188442230225, "learning_rate": 1.6939136984104954e-06, "loss": 0.0471, "step": 1390 }, { "epoch": 81.82352941176471, "grad_norm": 1.403760313987732, "learning_rate": 1.68331982363497e-06, "loss": 0.0933, "step": 1391 }, { "epoch": 81.88235294117646, "grad_norm": 1.9614943265914917, "learning_rate": 1.672756135049598e-06, "loss": 0.0492, "step": 1392 }, { "epoch": 81.94117647058823, "grad_norm": 1.7119088172912598, "learning_rate": 1.6622226709962963e-06, "loss": 0.0699, "step": 1393 }, { "epoch": 82.0, "grad_norm": 1.7040654420852661, "learning_rate": 1.6517194697072903e-06, "loss": 0.079, "step": 1394 }, { "epoch": 82.05882352941177, "grad_norm": 1.3292866945266724, "learning_rate": 1.6412465693049561e-06, "loss": 0.0508, "step": 1395 }, { "epoch": 82.11764705882354, "grad_norm": 1.4891904592514038, "learning_rate": 1.630804007801694e-06, "loss": 0.0759, "step": 1396 }, { "epoch": 82.17647058823529, "grad_norm": 1.287721872329712, "learning_rate": 1.620391823099784e-06, "loss": 0.0242, "step": 1397 }, { "epoch": 82.23529411764706, "grad_norm": 1.380759835243225, "learning_rate": 1.610010052991251e-06, "loss": 0.0345, "step": 1398 }, { "epoch": 82.29411764705883, "grad_norm": 1.7552822828292847, "learning_rate": 1.5996587351577296e-06, "loss": 0.024, "step": 1399 }, { "epoch": 82.3529411764706, "grad_norm": 1.5083329677581787, "learning_rate": 1.5893379071703207e-06, "loss": 0.1177, "step": 1400 }, { "epoch": 82.41176470588235, "grad_norm": 1.2917335033416748, "learning_rate": 1.5790476064894622e-06, "loss": 0.036, "step": 1401 }, { "epoch": 82.47058823529412, "grad_norm": 1.6364712715148926, "learning_rate": 1.5687878704647907e-06, "loss": 0.0754, "step": 1402 }, { "epoch": 82.52941176470588, "grad_norm": 6.825064182281494, "learning_rate": 1.558558736335003e-06, "loss": 0.046, "step": 1403 }, { "epoch": 82.58823529411765, "grad_norm": 1.3741040229797363, "learning_rate": 1.5483602412277254e-06, "loss": 0.0267, "step": 1404 }, { "epoch": 82.6470588235294, "grad_norm": 1.7312531471252441, "learning_rate": 1.5381924221593781e-06, "loss": 0.0924, "step": 1405 }, { "epoch": 82.70588235294117, "grad_norm": 1.789716362953186, "learning_rate": 1.528055316035033e-06, "loss": 0.088, "step": 1406 }, { "epoch": 82.76470588235294, "grad_norm": 1.7512542009353638, "learning_rate": 1.5179489596482944e-06, "loss": 0.0413, "step": 1407 }, { "epoch": 82.82352941176471, "grad_norm": 1.9579652547836304, "learning_rate": 1.5078733896811537e-06, "loss": 0.1418, "step": 1408 }, { "epoch": 82.88235294117646, "grad_norm": 2.1275405883789062, "learning_rate": 1.4978286427038602e-06, "loss": 0.0401, "step": 1409 }, { "epoch": 82.94117647058823, "grad_norm": 1.3847870826721191, "learning_rate": 1.4878147551747901e-06, "loss": 0.0644, "step": 1410 }, { "epoch": 83.0, "grad_norm": 2.1133224964141846, "learning_rate": 1.4778317634403082e-06, "loss": 0.0805, "step": 1411 }, { "epoch": 83.05882352941177, "grad_norm": 1.5740872621536255, "learning_rate": 1.467879703734646e-06, "loss": 0.0824, "step": 1412 }, { "epoch": 83.11764705882354, "grad_norm": 1.3392335176467896, "learning_rate": 1.4579586121797585e-06, "loss": 0.0749, "step": 1413 }, { "epoch": 83.17647058823529, "grad_norm": 1.533657193183899, "learning_rate": 1.4480685247852011e-06, "loss": 0.0983, "step": 1414 }, { "epoch": 83.23529411764706, "grad_norm": 1.7777667045593262, "learning_rate": 1.4382094774479983e-06, "loss": 0.0533, "step": 1415 }, { "epoch": 83.29411764705883, "grad_norm": 1.5754491090774536, "learning_rate": 1.4283815059525085e-06, "loss": 0.0706, "step": 1416 }, { "epoch": 83.3529411764706, "grad_norm": 1.481256365776062, "learning_rate": 1.4185846459703012e-06, "loss": 0.0808, "step": 1417 }, { "epoch": 83.41176470588235, "grad_norm": 2.0089190006256104, "learning_rate": 1.4088189330600177e-06, "loss": 0.079, "step": 1418 }, { "epoch": 83.47058823529412, "grad_norm": 1.9939610958099365, "learning_rate": 1.3990844026672557e-06, "loss": 0.0297, "step": 1419 }, { "epoch": 83.52941176470588, "grad_norm": 1.8444596529006958, "learning_rate": 1.3893810901244264e-06, "loss": 0.0542, "step": 1420 }, { "epoch": 83.58823529411765, "grad_norm": 1.3461389541625977, "learning_rate": 1.3797090306506368e-06, "loss": 0.0721, "step": 1421 }, { "epoch": 83.6470588235294, "grad_norm": 1.3731071949005127, "learning_rate": 1.3700682593515557e-06, "loss": 0.0377, "step": 1422 }, { "epoch": 83.70588235294117, "grad_norm": 0.9588115215301514, "learning_rate": 1.3604588112192895e-06, "loss": 0.0261, "step": 1423 }, { "epoch": 83.76470588235294, "grad_norm": 1.8610379695892334, "learning_rate": 1.350880721132254e-06, "loss": 0.0438, "step": 1424 }, { "epoch": 83.82352941176471, "grad_norm": 1.5652433633804321, "learning_rate": 1.3413340238550476e-06, "loss": 0.0381, "step": 1425 }, { "epoch": 83.88235294117646, "grad_norm": 1.9390555620193481, "learning_rate": 1.3318187540383254e-06, "loss": 0.0519, "step": 1426 }, { "epoch": 83.94117647058823, "grad_norm": 2.0639357566833496, "learning_rate": 1.3223349462186753e-06, "loss": 0.0951, "step": 1427 }, { "epoch": 84.0, "grad_norm": 1.7024593353271484, "learning_rate": 1.3128826348184886e-06, "loss": 0.0896, "step": 1428 }, { "epoch": 84.05882352941177, "grad_norm": 1.727534294128418, "learning_rate": 1.30346185414584e-06, "loss": 0.061, "step": 1429 }, { "epoch": 84.11764705882354, "grad_norm": 1.680250644683838, "learning_rate": 1.2940726383943547e-06, "loss": 0.0583, "step": 1430 }, { "epoch": 84.17647058823529, "grad_norm": 1.9366788864135742, "learning_rate": 1.284715021643096e-06, "loss": 0.0406, "step": 1431 }, { "epoch": 84.23529411764706, "grad_norm": 1.685917615890503, "learning_rate": 1.2753890378564327e-06, "loss": 0.0654, "step": 1432 }, { "epoch": 84.29411764705883, "grad_norm": 2.0447378158569336, "learning_rate": 1.2660947208839203e-06, "loss": 0.1281, "step": 1433 }, { "epoch": 84.3529411764706, "grad_norm": 1.436333417892456, "learning_rate": 1.256832104460175e-06, "loss": 0.0494, "step": 1434 }, { "epoch": 84.41176470588235, "grad_norm": 1.3438456058502197, "learning_rate": 1.247601222204753e-06, "loss": 0.0321, "step": 1435 }, { "epoch": 84.47058823529412, "grad_norm": 1.6440610885620117, "learning_rate": 1.238402107622031e-06, "loss": 0.0603, "step": 1436 }, { "epoch": 84.52941176470588, "grad_norm": 1.5294479131698608, "learning_rate": 1.2292347941010773e-06, "loss": 0.0929, "step": 1437 }, { "epoch": 84.58823529411765, "grad_norm": 3.550767660140991, "learning_rate": 1.2200993149155382e-06, "loss": 0.0869, "step": 1438 }, { "epoch": 84.6470588235294, "grad_norm": 1.1512681245803833, "learning_rate": 1.2109957032235143e-06, "loss": 0.0301, "step": 1439 }, { "epoch": 84.70588235294117, "grad_norm": 1.294584035873413, "learning_rate": 1.2019239920674385e-06, "loss": 0.0689, "step": 1440 }, { "epoch": 84.76470588235294, "grad_norm": 0.9377285242080688, "learning_rate": 1.1928842143739595e-06, "loss": 0.0234, "step": 1441 }, { "epoch": 84.82352941176471, "grad_norm": 1.7522201538085938, "learning_rate": 1.1838764029538207e-06, "loss": 0.0399, "step": 1442 }, { "epoch": 84.88235294117646, "grad_norm": 1.301515817642212, "learning_rate": 1.1749005905017353e-06, "loss": 0.0414, "step": 1443 }, { "epoch": 84.94117647058823, "grad_norm": 1.943636417388916, "learning_rate": 1.1659568095962793e-06, "loss": 0.096, "step": 1444 }, { "epoch": 85.0, "grad_norm": 1.7821638584136963, "learning_rate": 1.1570450926997657e-06, "loss": 0.0889, "step": 1445 }, { "epoch": 85.05882352941177, "grad_norm": 3.0800695419311523, "learning_rate": 1.148165472158127e-06, "loss": 0.0952, "step": 1446 }, { "epoch": 85.11764705882354, "grad_norm": 1.349579095840454, "learning_rate": 1.1393179802008002e-06, "loss": 0.0628, "step": 1447 }, { "epoch": 85.17647058823529, "grad_norm": 1.607113003730774, "learning_rate": 1.130502648940608e-06, "loss": 0.0697, "step": 1448 }, { "epoch": 85.23529411764706, "grad_norm": 1.0676943063735962, "learning_rate": 1.1217195103736422e-06, "loss": 0.0226, "step": 1449 }, { "epoch": 85.29411764705883, "grad_norm": 2.583757162094116, "learning_rate": 1.1129685963791493e-06, "loss": 0.028, "step": 1450 }, { "epoch": 85.3529411764706, "grad_norm": 1.7796918153762817, "learning_rate": 1.1042499387194105e-06, "loss": 0.1059, "step": 1451 }, { "epoch": 85.41176470588235, "grad_norm": 1.0145155191421509, "learning_rate": 1.0955635690396348e-06, "loss": 0.0252, "step": 1452 }, { "epoch": 85.47058823529412, "grad_norm": 1.4165287017822266, "learning_rate": 1.0869095188678347e-06, "loss": 0.0315, "step": 1453 }, { "epoch": 85.52941176470588, "grad_norm": 1.3284443616867065, "learning_rate": 1.0782878196147162e-06, "loss": 0.0538, "step": 1454 }, { "epoch": 85.58823529411765, "grad_norm": 1.3321113586425781, "learning_rate": 1.0696985025735662e-06, "loss": 0.0346, "step": 1455 }, { "epoch": 85.6470588235294, "grad_norm": 1.3598536252975464, "learning_rate": 1.061141598920139e-06, "loss": 0.0385, "step": 1456 }, { "epoch": 85.70588235294117, "grad_norm": 1.9543812274932861, "learning_rate": 1.052617139712535e-06, "loss": 0.0406, "step": 1457 }, { "epoch": 85.76470588235294, "grad_norm": 1.9540636539459229, "learning_rate": 1.0441251558911002e-06, "loss": 0.0801, "step": 1458 }, { "epoch": 85.82352941176471, "grad_norm": 0.9536816477775574, "learning_rate": 1.035665678278306e-06, "loss": 0.023, "step": 1459 }, { "epoch": 85.88235294117646, "grad_norm": 1.6425310373306274, "learning_rate": 1.0272387375786397e-06, "loss": 0.0697, "step": 1460 }, { "epoch": 85.94117647058823, "grad_norm": 2.0008504390716553, "learning_rate": 1.0188443643784917e-06, "loss": 0.1544, "step": 1461 }, { "epoch": 86.0, "grad_norm": 1.6319931745529175, "learning_rate": 1.010482589146048e-06, "loss": 0.0967, "step": 1462 }, { "epoch": 86.05882352941177, "grad_norm": 1.1225669384002686, "learning_rate": 1.0021534422311741e-06, "loss": 0.03, "step": 1463 }, { "epoch": 86.11764705882354, "grad_norm": 1.6801974773406982, "learning_rate": 9.938569538653108e-07, "loss": 0.0853, "step": 1464 }, { "epoch": 86.17647058823529, "grad_norm": 1.230038046836853, "learning_rate": 9.855931541613594e-07, "loss": 0.0371, "step": 1465 }, { "epoch": 86.23529411764706, "grad_norm": 1.7456353902816772, "learning_rate": 9.773620731135747e-07, "loss": 0.0741, "step": 1466 }, { "epoch": 86.29411764705883, "grad_norm": 0.9525942802429199, "learning_rate": 9.691637405974574e-07, "loss": 0.0306, "step": 1467 }, { "epoch": 86.3529411764706, "grad_norm": 1.7127455472946167, "learning_rate": 9.609981863696439e-07, "loss": 0.0357, "step": 1468 }, { "epoch": 86.41176470588235, "grad_norm": 1.5985796451568604, "learning_rate": 9.528654400677973e-07, "loss": 0.0553, "step": 1469 }, { "epoch": 86.47058823529412, "grad_norm": 1.440968632698059, "learning_rate": 9.447655312105019e-07, "loss": 0.0706, "step": 1470 }, { "epoch": 86.52941176470588, "grad_norm": 1.2967031002044678, "learning_rate": 9.366984891971564e-07, "loss": 0.0432, "step": 1471 }, { "epoch": 86.58823529411765, "grad_norm": 1.702025055885315, "learning_rate": 9.286643433078646e-07, "loss": 0.1008, "step": 1472 }, { "epoch": 86.6470588235294, "grad_norm": 1.2131925821304321, "learning_rate": 9.206631227033314e-07, "loss": 0.0309, "step": 1473 }, { "epoch": 86.70588235294117, "grad_norm": 1.435332179069519, "learning_rate": 9.126948564247561e-07, "loss": 0.0767, "step": 1474 }, { "epoch": 86.76470588235294, "grad_norm": 1.769909381866455, "learning_rate": 9.047595733937276e-07, "loss": 0.0804, "step": 1475 }, { "epoch": 86.82352941176471, "grad_norm": 2.4295907020568848, "learning_rate": 8.968573024121185e-07, "loss": 0.0853, "step": 1476 }, { "epoch": 86.88235294117646, "grad_norm": 1.5631027221679688, "learning_rate": 8.889880721619793e-07, "loss": 0.0628, "step": 1477 }, { "epoch": 86.94117647058823, "grad_norm": 1.7708773612976074, "learning_rate": 8.81151911205439e-07, "loss": 0.0628, "step": 1478 }, { "epoch": 87.0, "grad_norm": 1.4804843664169312, "learning_rate": 8.733488479845997e-07, "loss": 0.0594, "step": 1479 }, { "epoch": 87.05882352941177, "grad_norm": 1.4921247959136963, "learning_rate": 8.655789108214241e-07, "loss": 0.0349, "step": 1480 }, { "epoch": 87.11764705882354, "grad_norm": 1.271370768547058, "learning_rate": 8.578421279176485e-07, "loss": 0.0308, "step": 1481 }, { "epoch": 87.17647058823529, "grad_norm": 2.413344383239746, "learning_rate": 8.501385273546703e-07, "loss": 0.0819, "step": 1482 }, { "epoch": 87.23529411764706, "grad_norm": 1.3129096031188965, "learning_rate": 8.424681370934484e-07, "loss": 0.0346, "step": 1483 }, { "epoch": 87.29411764705883, "grad_norm": 2.2763307094573975, "learning_rate": 8.348309849744007e-07, "loss": 0.0869, "step": 1484 }, { "epoch": 87.3529411764706, "grad_norm": 3.4722578525543213, "learning_rate": 8.272270987173059e-07, "loss": 0.1061, "step": 1485 }, { "epoch": 87.41176470588235, "grad_norm": 1.7079938650131226, "learning_rate": 8.196565059212003e-07, "loss": 0.0987, "step": 1486 }, { "epoch": 87.47058823529412, "grad_norm": 1.881752371788025, "learning_rate": 8.121192340642769e-07, "loss": 0.0845, "step": 1487 }, { "epoch": 87.52941176470588, "grad_norm": 1.4184015989303589, "learning_rate": 8.046153105037902e-07, "loss": 0.0958, "step": 1488 }, { "epoch": 87.58823529411765, "grad_norm": 1.839342713356018, "learning_rate": 7.971447624759498e-07, "loss": 0.1184, "step": 1489 }, { "epoch": 87.6470588235294, "grad_norm": 1.359444260597229, "learning_rate": 7.897076170958307e-07, "loss": 0.0331, "step": 1490 }, { "epoch": 87.70588235294117, "grad_norm": 1.3871406316757202, "learning_rate": 7.823039013572642e-07, "loss": 0.0303, "step": 1491 }, { "epoch": 87.76470588235294, "grad_norm": 1.0865353345870972, "learning_rate": 7.749336421327502e-07, "loss": 0.028, "step": 1492 }, { "epoch": 87.82352941176471, "grad_norm": 1.8688945770263672, "learning_rate": 7.675968661733535e-07, "loss": 0.0653, "step": 1493 }, { "epoch": 87.88235294117646, "grad_norm": 1.5749787092208862, "learning_rate": 7.602936001086058e-07, "loss": 0.0491, "step": 1494 }, { "epoch": 87.94117647058823, "grad_norm": 0.9939689636230469, "learning_rate": 7.530238704464133e-07, "loss": 0.0391, "step": 1495 }, { "epoch": 88.0, "grad_norm": 1.4181450605392456, "learning_rate": 7.457877035729588e-07, "loss": 0.0248, "step": 1496 }, { "epoch": 88.05882352941177, "grad_norm": 2.515380859375, "learning_rate": 7.385851257526066e-07, "loss": 0.0429, "step": 1497 }, { "epoch": 88.11764705882354, "grad_norm": 1.6168322563171387, "learning_rate": 7.31416163127805e-07, "loss": 0.0308, "step": 1498 }, { "epoch": 88.17647058823529, "grad_norm": 1.5306034088134766, "learning_rate": 7.242808417189929e-07, "loss": 0.054, "step": 1499 }, { "epoch": 88.23529411764706, "grad_norm": 1.642764925956726, "learning_rate": 7.171791874245048e-07, "loss": 0.0349, "step": 1500 }, { "epoch": 88.29411764705883, "grad_norm": 1.3398083448410034, "learning_rate": 7.101112260204779e-07, "loss": 0.0397, "step": 1501 }, { "epoch": 88.3529411764706, "grad_norm": 1.4987317323684692, "learning_rate": 7.030769831607576e-07, "loss": 0.0683, "step": 1502 }, { "epoch": 88.41176470588235, "grad_norm": 2.484175443649292, "learning_rate": 6.960764843768053e-07, "loss": 0.0538, "step": 1503 }, { "epoch": 88.47058823529412, "grad_norm": 1.733365535736084, "learning_rate": 6.891097550776027e-07, "loss": 0.09, "step": 1504 }, { "epoch": 88.52941176470588, "grad_norm": 1.5073997974395752, "learning_rate": 6.821768205495638e-07, "loss": 0.067, "step": 1505 }, { "epoch": 88.58823529411765, "grad_norm": 1.6284509897232056, "learning_rate": 6.752777059564431e-07, "loss": 0.1071, "step": 1506 }, { "epoch": 88.6470588235294, "grad_norm": 1.9007198810577393, "learning_rate": 6.684124363392375e-07, "loss": 0.0887, "step": 1507 }, { "epoch": 88.70588235294117, "grad_norm": 0.8960374593734741, "learning_rate": 6.615810366161035e-07, "loss": 0.025, "step": 1508 }, { "epoch": 88.76470588235294, "grad_norm": 1.859602928161621, "learning_rate": 6.547835315822637e-07, "loss": 0.0531, "step": 1509 }, { "epoch": 88.82352941176471, "grad_norm": 1.4542087316513062, "learning_rate": 6.480199459099157e-07, "loss": 0.0651, "step": 1510 }, { "epoch": 88.88235294117646, "grad_norm": 1.770760178565979, "learning_rate": 6.412903041481433e-07, "loss": 0.0986, "step": 1511 }, { "epoch": 88.94117647058823, "grad_norm": 1.3468950986862183, "learning_rate": 6.345946307228301e-07, "loss": 0.0456, "step": 1512 }, { "epoch": 89.0, "grad_norm": 1.732003092765808, "learning_rate": 6.279329499365649e-07, "loss": 0.0781, "step": 1513 }, { "epoch": 89.05882352941177, "grad_norm": 1.6535214185714722, "learning_rate": 6.213052859685575e-07, "loss": 0.0983, "step": 1514 }, { "epoch": 89.11764705882354, "grad_norm": 1.8872981071472168, "learning_rate": 6.147116628745542e-07, "loss": 0.047, "step": 1515 }, { "epoch": 89.17647058823529, "grad_norm": 1.7581654787063599, "learning_rate": 6.081521045867422e-07, "loss": 0.0339, "step": 1516 }, { "epoch": 89.23529411764706, "grad_norm": 1.932322382926941, "learning_rate": 6.016266349136679e-07, "loss": 0.1236, "step": 1517 }, { "epoch": 89.29411764705883, "grad_norm": 1.8379243612289429, "learning_rate": 5.951352775401519e-07, "loss": 0.0565, "step": 1518 }, { "epoch": 89.3529411764706, "grad_norm": 7.357274055480957, "learning_rate": 5.88678056027202e-07, "loss": 0.0728, "step": 1519 }, { "epoch": 89.41176470588235, "grad_norm": 2.9142088890075684, "learning_rate": 5.822549938119193e-07, "loss": 0.0452, "step": 1520 }, { "epoch": 89.47058823529412, "grad_norm": 1.317084789276123, "learning_rate": 5.758661142074285e-07, "loss": 0.0305, "step": 1521 }, { "epoch": 89.52941176470588, "grad_norm": 1.4337207078933716, "learning_rate": 5.695114404027824e-07, "loss": 0.0686, "step": 1522 }, { "epoch": 89.58823529411765, "grad_norm": 1.5704351663589478, "learning_rate": 5.631909954628812e-07, "loss": 0.0539, "step": 1523 }, { "epoch": 89.6470588235294, "grad_norm": 1.3327590227127075, "learning_rate": 5.569048023283852e-07, "loss": 0.0528, "step": 1524 }, { "epoch": 89.70588235294117, "grad_norm": 2.355527400970459, "learning_rate": 5.506528838156377e-07, "loss": 0.1134, "step": 1525 }, { "epoch": 89.76470588235294, "grad_norm": 1.6319234371185303, "learning_rate": 5.444352626165783e-07, "loss": 0.0642, "step": 1526 }, { "epoch": 89.82352941176471, "grad_norm": 1.3746941089630127, "learning_rate": 5.382519612986592e-07, "loss": 0.0289, "step": 1527 }, { "epoch": 89.88235294117646, "grad_norm": 1.4701186418533325, "learning_rate": 5.321030023047691e-07, "loss": 0.0707, "step": 1528 }, { "epoch": 89.94117647058823, "grad_norm": 1.394299030303955, "learning_rate": 5.259884079531452e-07, "loss": 0.0337, "step": 1529 }, { "epoch": 90.0, "grad_norm": 1.189337134361267, "learning_rate": 5.199082004372958e-07, "loss": 0.0566, "step": 1530 }, { "epoch": 90.05882352941177, "grad_norm": 1.1999685764312744, "learning_rate": 5.138624018259164e-07, "loss": 0.0212, "step": 1531 }, { "epoch": 90.11764705882354, "grad_norm": 1.835867166519165, "learning_rate": 5.078510340628163e-07, "loss": 0.043, "step": 1532 }, { "epoch": 90.17647058823529, "grad_norm": 1.2612080574035645, "learning_rate": 5.018741189668308e-07, "loss": 0.0251, "step": 1533 }, { "epoch": 90.23529411764706, "grad_norm": 1.6388763189315796, "learning_rate": 4.959316782317502e-07, "loss": 0.0699, "step": 1534 }, { "epoch": 90.29411764705883, "grad_norm": 3.3105485439300537, "learning_rate": 4.900237334262326e-07, "loss": 0.0308, "step": 1535 }, { "epoch": 90.3529411764706, "grad_norm": 1.683754563331604, "learning_rate": 4.841503059937313e-07, "loss": 0.1327, "step": 1536 }, { "epoch": 90.41176470588235, "grad_norm": 1.507497787475586, "learning_rate": 4.783114172524173e-07, "loss": 0.1079, "step": 1537 }, { "epoch": 90.47058823529412, "grad_norm": 1.3952748775482178, "learning_rate": 4.7250708839509774e-07, "loss": 0.0688, "step": 1538 }, { "epoch": 90.52941176470588, "grad_norm": 1.850264310836792, "learning_rate": 4.667373404891429e-07, "loss": 0.0434, "step": 1539 }, { "epoch": 90.58823529411765, "grad_norm": 1.505570650100708, "learning_rate": 4.610021944764054e-07, "loss": 0.0567, "step": 1540 }, { "epoch": 90.6470588235294, "grad_norm": 1.3680875301361084, "learning_rate": 4.553016711731506e-07, "loss": 0.054, "step": 1541 }, { "epoch": 90.70588235294117, "grad_norm": 1.8724169731140137, "learning_rate": 4.496357912699767e-07, "loss": 0.0487, "step": 1542 }, { "epoch": 90.76470588235294, "grad_norm": 1.5032649040222168, "learning_rate": 4.440045753317379e-07, "loss": 0.0628, "step": 1543 }, { "epoch": 90.82352941176471, "grad_norm": 1.5360603332519531, "learning_rate": 4.384080437974736e-07, "loss": 0.0396, "step": 1544 }, { "epoch": 90.88235294117646, "grad_norm": 1.5431989431381226, "learning_rate": 4.328462169803338e-07, "loss": 0.0502, "step": 1545 }, { "epoch": 90.94117647058823, "grad_norm": 1.6763865947723389, "learning_rate": 4.2731911506750377e-07, "loss": 0.0381, "step": 1546 }, { "epoch": 91.0, "grad_norm": 2.0274996757507324, "learning_rate": 4.218267581201296e-07, "loss": 0.129, "step": 1547 }, { "epoch": 91.05882352941177, "grad_norm": 1.7646598815917969, "learning_rate": 4.163691660732505e-07, "loss": 0.0857, "step": 1548 }, { "epoch": 91.11764705882354, "grad_norm": 4.4834370613098145, "learning_rate": 4.109463587357199e-07, "loss": 0.0407, "step": 1549 }, { "epoch": 91.17647058823529, "grad_norm": 1.971574068069458, "learning_rate": 4.055583557901399e-07, "loss": 0.0575, "step": 1550 }, { "epoch": 91.23529411764706, "grad_norm": 1.3554234504699707, "learning_rate": 4.00205176792785e-07, "loss": 0.1005, "step": 1551 }, { "epoch": 91.29411764705883, "grad_norm": 1.315651297569275, "learning_rate": 3.9488684117353184e-07, "loss": 0.0395, "step": 1552 }, { "epoch": 91.3529411764706, "grad_norm": 1.0303173065185547, "learning_rate": 3.8960336823579135e-07, "loss": 0.0266, "step": 1553 }, { "epoch": 91.41176470588235, "grad_norm": 1.923319697380066, "learning_rate": 3.843547771564371e-07, "loss": 0.068, "step": 1554 }, { "epoch": 91.47058823529412, "grad_norm": 1.6142544746398926, "learning_rate": 3.791410869857326e-07, "loss": 0.0849, "step": 1555 }, { "epoch": 91.52941176470588, "grad_norm": 1.0580295324325562, "learning_rate": 3.739623166472717e-07, "loss": 0.0284, "step": 1556 }, { "epoch": 91.58823529411765, "grad_norm": 1.3730387687683105, "learning_rate": 3.6881848493789507e-07, "loss": 0.0422, "step": 1557 }, { "epoch": 91.6470588235294, "grad_norm": 1.1691946983337402, "learning_rate": 3.6370961052763607e-07, "loss": 0.0284, "step": 1558 }, { "epoch": 91.70588235294117, "grad_norm": 1.73823082447052, "learning_rate": 3.5863571195964507e-07, "loss": 0.1202, "step": 1559 }, { "epoch": 91.76470588235294, "grad_norm": 1.641265630722046, "learning_rate": 3.5359680765012505e-07, "loss": 0.0323, "step": 1560 }, { "epoch": 91.82352941176471, "grad_norm": 1.3079763650894165, "learning_rate": 3.4859291588826393e-07, "loss": 0.0654, "step": 1561 }, { "epoch": 91.88235294117646, "grad_norm": 1.5614491701126099, "learning_rate": 3.436240548361669e-07, "loss": 0.0651, "step": 1562 }, { "epoch": 91.94117647058823, "grad_norm": 1.097105622291565, "learning_rate": 3.3869024252879304e-07, "loss": 0.0271, "step": 1563 }, { "epoch": 92.0, "grad_norm": 1.7284287214279175, "learning_rate": 3.3379149687388866e-07, "loss": 0.0882, "step": 1564 }, { "epoch": 92.05882352941177, "grad_norm": 1.3286806344985962, "learning_rate": 3.289278356519221e-07, "loss": 0.0451, "step": 1565 }, { "epoch": 92.11764705882354, "grad_norm": 1.38875412940979, "learning_rate": 3.240992765160189e-07, "loss": 0.0637, "step": 1566 }, { "epoch": 92.17647058823529, "grad_norm": 1.5926504135131836, "learning_rate": 3.193058369918989e-07, "loss": 0.0412, "step": 1567 }, { "epoch": 92.23529411764706, "grad_norm": 1.6081769466400146, "learning_rate": 3.145475344778115e-07, "loss": 0.0738, "step": 1568 }, { "epoch": 92.29411764705883, "grad_norm": 1.765147089958191, "learning_rate": 3.0982438624447275e-07, "loss": 0.068, "step": 1569 }, { "epoch": 92.3529411764706, "grad_norm": 2.0561461448669434, "learning_rate": 3.051364094350051e-07, "loss": 0.0896, "step": 1570 }, { "epoch": 92.41176470588235, "grad_norm": 1.015868067741394, "learning_rate": 3.004836210648687e-07, "loss": 0.0345, "step": 1571 }, { "epoch": 92.47058823529412, "grad_norm": 2.530893325805664, "learning_rate": 2.958660380218048e-07, "loss": 0.0801, "step": 1572 }, { "epoch": 92.52941176470588, "grad_norm": 0.8780486583709717, "learning_rate": 2.91283677065779e-07, "loss": 0.0236, "step": 1573 }, { "epoch": 92.58823529411765, "grad_norm": 1.7259502410888672, "learning_rate": 2.867365548289092e-07, "loss": 0.0451, "step": 1574 }, { "epoch": 92.6470588235294, "grad_norm": 1.0677293539047241, "learning_rate": 2.8222468781541226e-07, "loss": 0.02, "step": 1575 }, { "epoch": 92.70588235294117, "grad_norm": 1.2796434164047241, "learning_rate": 2.77748092401543e-07, "loss": 0.0313, "step": 1576 }, { "epoch": 92.76470588235294, "grad_norm": 1.7719359397888184, "learning_rate": 2.733067848355353e-07, "loss": 0.0994, "step": 1577 }, { "epoch": 92.82352941176471, "grad_norm": 1.4592092037200928, "learning_rate": 2.6890078123754214e-07, "loss": 0.023, "step": 1578 }, { "epoch": 92.88235294117646, "grad_norm": 2.015256643295288, "learning_rate": 2.645300975995768e-07, "loss": 0.1103, "step": 1579 }, { "epoch": 92.94117647058823, "grad_norm": 1.2016096115112305, "learning_rate": 2.601947497854562e-07, "loss": 0.0265, "step": 1580 }, { "epoch": 93.0, "grad_norm": 1.6638706922531128, "learning_rate": 2.5589475353073987e-07, "loss": 0.1046, "step": 1581 }, { "epoch": 93.05882352941177, "grad_norm": 1.3327665328979492, "learning_rate": 2.5163012444267997e-07, "loss": 0.0373, "step": 1582 }, { "epoch": 93.11764705882354, "grad_norm": 1.1112209558486938, "learning_rate": 2.4740087800015576e-07, "loss": 0.0327, "step": 1583 }, { "epoch": 93.17647058823529, "grad_norm": 1.576121211051941, "learning_rate": 2.4320702955362596e-07, "loss": 0.0917, "step": 1584 }, { "epoch": 93.23529411764706, "grad_norm": 1.5937718152999878, "learning_rate": 2.3904859432506535e-07, "loss": 0.0832, "step": 1585 }, { "epoch": 93.29411764705883, "grad_norm": 1.379560112953186, "learning_rate": 2.349255874079137e-07, "loss": 0.0535, "step": 1586 }, { "epoch": 93.3529411764706, "grad_norm": 1.5485881567001343, "learning_rate": 2.308380237670227e-07, "loss": 0.0975, "step": 1587 }, { "epoch": 93.41176470588235, "grad_norm": 1.4052072763442993, "learning_rate": 2.2678591823859563e-07, "loss": 0.1123, "step": 1588 }, { "epoch": 93.47058823529412, "grad_norm": 1.2261841297149658, "learning_rate": 2.2276928553013998e-07, "loss": 0.0232, "step": 1589 }, { "epoch": 93.52941176470588, "grad_norm": 1.558768391609192, "learning_rate": 2.1878814022041062e-07, "loss": 0.0499, "step": 1590 }, { "epoch": 93.58823529411765, "grad_norm": 1.6721471548080444, "learning_rate": 2.1484249675935653e-07, "loss": 0.0702, "step": 1591 }, { "epoch": 93.6470588235294, "grad_norm": 1.2389872074127197, "learning_rate": 2.1093236946807095e-07, "loss": 0.0214, "step": 1592 }, { "epoch": 93.70588235294117, "grad_norm": 2.034205675125122, "learning_rate": 2.070577725387357e-07, "loss": 0.0537, "step": 1593 }, { "epoch": 93.76470588235294, "grad_norm": 1.462477207183838, "learning_rate": 2.0321872003457255e-07, "loss": 0.034, "step": 1594 }, { "epoch": 93.82352941176471, "grad_norm": 1.078531265258789, "learning_rate": 1.994152258897919e-07, "loss": 0.0258, "step": 1595 }, { "epoch": 93.88235294117646, "grad_norm": 1.8153966665267944, "learning_rate": 1.956473039095419e-07, "loss": 0.0986, "step": 1596 }, { "epoch": 93.94117647058823, "grad_norm": 1.461488127708435, "learning_rate": 1.919149677698562e-07, "loss": 0.0719, "step": 1597 }, { "epoch": 94.0, "grad_norm": 1.9292405843734741, "learning_rate": 1.8821823101760949e-07, "loss": 0.0385, "step": 1598 }, { "epoch": 94.05882352941177, "grad_norm": 1.3086336851119995, "learning_rate": 1.8455710707046215e-07, "loss": 0.0679, "step": 1599 }, { "epoch": 94.11764705882354, "grad_norm": 1.411601185798645, "learning_rate": 1.8093160921681564e-07, "loss": 0.0295, "step": 1600 }, { "epoch": 94.17647058823529, "grad_norm": 1.1931512355804443, "learning_rate": 1.7734175061576264e-07, "loss": 0.0312, "step": 1601 }, { "epoch": 94.23529411764706, "grad_norm": 1.776712417602539, "learning_rate": 1.7378754429704049e-07, "loss": 0.0569, "step": 1602 }, { "epoch": 94.29411764705883, "grad_norm": 1.221267580986023, "learning_rate": 1.7026900316098217e-07, "loss": 0.0602, "step": 1603 }, { "epoch": 94.3529411764706, "grad_norm": 1.6622343063354492, "learning_rate": 1.6678613997847094e-07, "loss": 0.0579, "step": 1604 }, { "epoch": 94.41176470588235, "grad_norm": 1.5090763568878174, "learning_rate": 1.633389673908947e-07, "loss": 0.0755, "step": 1605 }, { "epoch": 94.47058823529412, "grad_norm": 1.8792388439178467, "learning_rate": 1.5992749791009622e-07, "loss": 0.1474, "step": 1606 }, { "epoch": 94.52941176470588, "grad_norm": 1.3404693603515625, "learning_rate": 1.5655174391833394e-07, "loss": 0.03, "step": 1607 }, { "epoch": 94.58823529411765, "grad_norm": 1.1935615539550781, "learning_rate": 1.5321171766822796e-07, "loss": 0.0392, "step": 1608 }, { "epoch": 94.6470588235294, "grad_norm": 1.8770685195922852, "learning_rate": 1.4990743128272755e-07, "loss": 0.0898, "step": 1609 }, { "epoch": 94.70588235294117, "grad_norm": 2.1046628952026367, "learning_rate": 1.4663889675505694e-07, "loss": 0.0352, "step": 1610 }, { "epoch": 94.76470588235294, "grad_norm": 1.2156071662902832, "learning_rate": 1.434061259486763e-07, "loss": 0.0363, "step": 1611 }, { "epoch": 94.82352941176471, "grad_norm": 2.1262800693511963, "learning_rate": 1.4020913059723862e-07, "loss": 0.0822, "step": 1612 }, { "epoch": 94.88235294117646, "grad_norm": 1.7029160261154175, "learning_rate": 1.37047922304544e-07, "loss": 0.0893, "step": 1613 }, { "epoch": 94.94117647058823, "grad_norm": 1.429010272026062, "learning_rate": 1.3392251254450428e-07, "loss": 0.0386, "step": 1614 }, { "epoch": 95.0, "grad_norm": 1.4527175426483154, "learning_rate": 1.30832912661093e-07, "loss": 0.026, "step": 1615 }, { "epoch": 95.05882352941177, "grad_norm": 1.5953952074050903, "learning_rate": 1.2777913386831097e-07, "loss": 0.0927, "step": 1616 }, { "epoch": 95.11764705882354, "grad_norm": 1.455911636352539, "learning_rate": 1.2476118725014085e-07, "loss": 0.0631, "step": 1617 }, { "epoch": 95.17647058823529, "grad_norm": 1.6352274417877197, "learning_rate": 1.217790837605104e-07, "loss": 0.1013, "step": 1618 }, { "epoch": 95.23529411764706, "grad_norm": 11.379027366638184, "learning_rate": 1.1883283422325142e-07, "loss": 0.0351, "step": 1619 }, { "epoch": 95.29411764705883, "grad_norm": 1.1606031656265259, "learning_rate": 1.1592244933205988e-07, "loss": 0.0447, "step": 1620 }, { "epoch": 95.3529411764706, "grad_norm": 1.2983261346817017, "learning_rate": 1.130479396504558e-07, "loss": 0.0246, "step": 1621 }, { "epoch": 95.41176470588235, "grad_norm": 1.2248892784118652, "learning_rate": 1.1020931561174897e-07, "loss": 0.0356, "step": 1622 }, { "epoch": 95.47058823529412, "grad_norm": 1.297864317893982, "learning_rate": 1.0740658751899891e-07, "loss": 0.1103, "step": 1623 }, { "epoch": 95.52941176470588, "grad_norm": 1.479166865348816, "learning_rate": 1.0463976554497823e-07, "loss": 0.0406, "step": 1624 }, { "epoch": 95.58823529411765, "grad_norm": 1.5235199928283691, "learning_rate": 1.019088597321305e-07, "loss": 0.0269, "step": 1625 }, { "epoch": 95.6470588235294, "grad_norm": 1.4718759059906006, "learning_rate": 9.921387999254128e-08, "loss": 0.096, "step": 1626 }, { "epoch": 95.70588235294117, "grad_norm": 1.435036540031433, "learning_rate": 9.65548361079005e-08, "loss": 0.0337, "step": 1627 }, { "epoch": 95.76470588235294, "grad_norm": 2.9721319675445557, "learning_rate": 9.393173772946129e-08, "loss": 0.0785, "step": 1628 }, { "epoch": 95.82352941176471, "grad_norm": 1.6321048736572266, "learning_rate": 9.134459437801225e-08, "loss": 0.0599, "step": 1629 }, { "epoch": 95.88235294117646, "grad_norm": 1.5033143758773804, "learning_rate": 8.87934154438408e-08, "loss": 0.0652, "step": 1630 }, { "epoch": 95.94117647058823, "grad_norm": 1.7039252519607544, "learning_rate": 8.627821018669213e-08, "loss": 0.0437, "step": 1631 }, { "epoch": 96.0, "grad_norm": 1.3960661888122559, "learning_rate": 8.379898773574924e-08, "loss": 0.0519, "step": 1632 }, { "epoch": 96.05882352941177, "grad_norm": 1.78780996799469, "learning_rate": 8.135575708958621e-08, "loss": 0.0382, "step": 1633 }, { "epoch": 96.11764705882354, "grad_norm": 1.5066839456558228, "learning_rate": 7.894852711614275e-08, "loss": 0.0555, "step": 1634 }, { "epoch": 96.17647058823529, "grad_norm": 1.5087800025939941, "learning_rate": 7.657730655269424e-08, "loss": 0.0739, "step": 1635 }, { "epoch": 96.23529411764706, "grad_norm": 1.83529531955719, "learning_rate": 7.424210400581167e-08, "loss": 0.1016, "step": 1636 }, { "epoch": 96.29411764705883, "grad_norm": 1.5357167720794678, "learning_rate": 7.19429279513384e-08, "loss": 0.0443, "step": 1637 }, { "epoch": 96.3529411764706, "grad_norm": 1.514906644821167, "learning_rate": 6.96797867343535e-08, "loss": 0.0612, "step": 1638 }, { "epoch": 96.41176470588235, "grad_norm": 1.625988483428955, "learning_rate": 6.745268856914622e-08, "loss": 0.0392, "step": 1639 }, { "epoch": 96.47058823529412, "grad_norm": 1.2151917219161987, "learning_rate": 6.526164153918269e-08, "loss": 0.0284, "step": 1640 }, { "epoch": 96.52941176470588, "grad_norm": 1.8283357620239258, "learning_rate": 6.310665359708035e-08, "loss": 0.0659, "step": 1641 }, { "epoch": 96.58823529411765, "grad_norm": 1.7530206441879272, "learning_rate": 6.098773256457468e-08, "loss": 0.0291, "step": 1642 }, { "epoch": 96.6470588235294, "grad_norm": 2.2579822540283203, "learning_rate": 5.890488613249257e-08, "loss": 0.0678, "step": 1643 }, { "epoch": 96.70588235294117, "grad_norm": 1.8850383758544922, "learning_rate": 5.685812186072559e-08, "loss": 0.0953, "step": 1644 }, { "epoch": 96.76470588235294, "grad_norm": 1.521911859512329, "learning_rate": 5.484744717820123e-08, "loss": 0.0303, "step": 1645 }, { "epoch": 96.82352941176471, "grad_norm": 1.9838532209396362, "learning_rate": 5.287286938285729e-08, "loss": 0.0592, "step": 1646 }, { "epoch": 96.88235294117646, "grad_norm": 1.3336255550384521, "learning_rate": 5.0934395641611954e-08, "loss": 0.0352, "step": 1647 }, { "epoch": 96.94117647058823, "grad_norm": 1.6650339365005493, "learning_rate": 4.903203299034376e-08, "loss": 0.1237, "step": 1648 }, { "epoch": 97.0, "grad_norm": 1.8853039741516113, "learning_rate": 4.716578833386054e-08, "loss": 0.0882, "step": 1649 }, { "epoch": 97.05882352941177, "grad_norm": 1.545451283454895, "learning_rate": 4.5335668445876115e-08, "loss": 0.0296, "step": 1650 }, { "epoch": 97.11764705882354, "grad_norm": 1.390011191368103, "learning_rate": 4.354167996898584e-08, "loss": 0.0556, "step": 1651 }, { "epoch": 97.17647058823529, "grad_norm": 1.4995321035385132, "learning_rate": 4.178382941464554e-08, "loss": 0.0553, "step": 1652 }, { "epoch": 97.23529411764706, "grad_norm": 2.1895222663879395, "learning_rate": 4.006212316314151e-08, "loss": 0.0395, "step": 1653 }, { "epoch": 97.29411764705883, "grad_norm": 1.4705277681350708, "learning_rate": 3.8376567463571654e-08, "loss": 0.0659, "step": 1654 }, { "epoch": 97.3529411764706, "grad_norm": 1.5410383939743042, "learning_rate": 3.672716843382218e-08, "loss": 0.0794, "step": 1655 }, { "epoch": 97.41176470588235, "grad_norm": 1.1986891031265259, "learning_rate": 3.511393206054759e-08, "loss": 0.0276, "step": 1656 }, { "epoch": 97.47058823529412, "grad_norm": 1.730941891670227, "learning_rate": 3.3536864199141814e-08, "loss": 0.0405, "step": 1657 }, { "epoch": 97.52941176470588, "grad_norm": 1.7975687980651855, "learning_rate": 3.1995970573726055e-08, "loss": 0.0682, "step": 1658 }, { "epoch": 97.58823529411765, "grad_norm": 1.4790481328964233, "learning_rate": 3.049125677711984e-08, "loss": 0.0311, "step": 1659 }, { "epoch": 97.6470588235294, "grad_norm": 1.908799171447754, "learning_rate": 2.902272827082886e-08, "loss": 0.0519, "step": 1660 }, { "epoch": 97.70588235294117, "grad_norm": 1.7086906433105469, "learning_rate": 2.7590390385017208e-08, "loss": 0.0826, "step": 1661 }, { "epoch": 97.76470588235294, "grad_norm": 1.4691822528839111, "learning_rate": 2.619424831849293e-08, "loss": 0.0681, "step": 1662 }, { "epoch": 97.82352941176471, "grad_norm": 1.6778796911239624, "learning_rate": 2.4834307138688064e-08, "loss": 0.0784, "step": 1663 }, { "epoch": 97.88235294117646, "grad_norm": 1.885629415512085, "learning_rate": 2.3510571781640846e-08, "loss": 0.1065, "step": 1664 }, { "epoch": 97.94117647058823, "grad_norm": 1.5108624696731567, "learning_rate": 2.2223047051973533e-08, "loss": 0.0471, "step": 1665 }, { "epoch": 98.0, "grad_norm": 2.00940203666687, "learning_rate": 2.0971737622883515e-08, "loss": 0.0899, "step": 1666 }, { "epoch": 98.05882352941177, "grad_norm": 2.4714226722717285, "learning_rate": 1.9756648036116654e-08, "loss": 0.0549, "step": 1667 }, { "epoch": 98.11764705882354, "grad_norm": 1.6516335010528564, "learning_rate": 1.8577782701959536e-08, "loss": 0.0438, "step": 1668 }, { "epoch": 98.17647058823529, "grad_norm": 2.0235722064971924, "learning_rate": 1.743514589921502e-08, "loss": 0.114, "step": 1669 }, { "epoch": 98.23529411764706, "grad_norm": 2.0519068241119385, "learning_rate": 1.6328741775194502e-08, "loss": 0.0654, "step": 1670 }, { "epoch": 98.29411764705883, "grad_norm": 1.8690237998962402, "learning_rate": 1.5258574345696775e-08, "loss": 0.0834, "step": 1671 }, { "epoch": 98.3529411764706, "grad_norm": 1.7942980527877808, "learning_rate": 1.4224647494999189e-08, "loss": 0.0395, "step": 1672 }, { "epoch": 98.41176470588235, "grad_norm": 1.9478211402893066, "learning_rate": 1.3226964975837641e-08, "loss": 0.1109, "step": 1673 }, { "epoch": 98.47058823529412, "grad_norm": 1.2691676616668701, "learning_rate": 1.2265530409397708e-08, "loss": 0.0339, "step": 1674 }, { "epoch": 98.52941176470588, "grad_norm": 1.386500597000122, "learning_rate": 1.134034728529687e-08, "loss": 0.0361, "step": 1675 }, { "epoch": 98.58823529411765, "grad_norm": 1.3231767416000366, "learning_rate": 1.0451418961576754e-08, "loss": 0.0374, "step": 1676 }, { "epoch": 98.6470588235294, "grad_norm": 0.9929178357124329, "learning_rate": 9.598748664688683e-09, "loss": 0.023, "step": 1677 }, { "epoch": 98.70588235294117, "grad_norm": 1.6882702112197876, "learning_rate": 8.782339489482595e-09, "loss": 0.1198, "step": 1678 }, { "epoch": 98.76470588235294, "grad_norm": 1.374429702758789, "learning_rate": 8.002194399191476e-09, "loss": 0.0348, "step": 1679 }, { "epoch": 98.82352941176471, "grad_norm": 1.2156238555908203, "learning_rate": 7.2583162254280526e-09, "loss": 0.0246, "step": 1680 }, { "epoch": 98.88235294117646, "grad_norm": 1.4856112003326416, "learning_rate": 6.550707668171452e-09, "loss": 0.0619, "step": 1681 }, { "epoch": 98.94117647058823, "grad_norm": 4.639705181121826, "learning_rate": 5.879371295751668e-09, "loss": 0.0713, "step": 1682 }, { "epoch": 99.0, "grad_norm": 15.063031196594238, "learning_rate": 5.2443095448506674e-09, "loss": 0.0636, "step": 1683 }, { "epoch": 99.05882352941177, "grad_norm": 1.4607163667678833, "learning_rate": 4.645524720485739e-09, "loss": 0.0519, "step": 1684 }, { "epoch": 99.11764705882354, "grad_norm": 1.459328293800354, "learning_rate": 4.0830189960050505e-09, "loss": 0.0457, "step": 1685 }, { "epoch": 99.17647058823529, "grad_norm": 1.3032056093215942, "learning_rate": 3.556794413074327e-09, "loss": 0.0341, "step": 1686 }, { "epoch": 99.23529411764706, "grad_norm": 1.1812529563903809, "learning_rate": 3.066852881679072e-09, "loss": 0.028, "step": 1687 }, { "epoch": 99.29411764705883, "grad_norm": 1.7535244226455688, "learning_rate": 2.613196180107913e-09, "loss": 0.0434, "step": 1688 }, { "epoch": 99.3529411764706, "grad_norm": 1.4377093315124512, "learning_rate": 2.1958259549514914e-09, "loss": 0.057, "step": 1689 }, { "epoch": 99.41176470588235, "grad_norm": 1.2816752195358276, "learning_rate": 1.8147437210958018e-09, "loss": 0.0282, "step": 1690 }, { "epoch": 99.47058823529412, "grad_norm": 1.5607774257659912, "learning_rate": 1.4699508617144199e-09, "loss": 0.0547, "step": 1691 }, { "epoch": 99.52941176470588, "grad_norm": 1.57564377784729, "learning_rate": 1.161448628267392e-09, "loss": 0.1208, "step": 1692 }, { "epoch": 99.58823529411765, "grad_norm": 1.4121248722076416, "learning_rate": 8.892381404923545e-10, "loss": 0.0459, "step": 1693 }, { "epoch": 99.6470588235294, "grad_norm": 1.1185537576675415, "learning_rate": 6.533203864034221e-10, "loss": 0.0608, "step": 1694 }, { "epoch": 99.70588235294117, "grad_norm": 1.5961196422576904, "learning_rate": 4.5369622228674805e-10, "loss": 0.0324, "step": 1695 }, { "epoch": 99.76470588235294, "grad_norm": 2.089541435241699, "learning_rate": 2.9036637269830303e-10, "loss": 0.1189, "step": 1696 }, { "epoch": 99.82352941176471, "grad_norm": 1.7632555961608887, "learning_rate": 1.6333143045832445e-10, "loss": 0.0738, "step": 1697 }, { "epoch": 99.88235294117646, "grad_norm": 1.152005672454834, "learning_rate": 7.259185665353663e-11, "loss": 0.0307, "step": 1698 }, { "epoch": 99.94117647058823, "grad_norm": 1.5251919031143188, "learning_rate": 1.8147980630489613e-11, "loss": 0.1163, "step": 1699 }, { "epoch": 100.0, "grad_norm": 2.0700817108154297, "learning_rate": 0.0, "loss": 0.0512, "step": 1700 }, { "epoch": 100.0, "step": 1700, "total_flos": 3969891224780800.0, "train_loss": 0.5549032051423016, "train_runtime": 1008.6063, "train_samples_per_second": 52.746, "train_steps_per_second": 1.685 } ], "logging_steps": 1.0, "max_steps": 1700, "num_input_tokens_seen": 0, "num_train_epochs": 100, "save_steps": 50000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 3969891224780800.0, "train_batch_size": 2, "trial_name": null, "trial_params": null }