{ "best_metric": 0.3466954827308655, "best_model_checkpoint": "miner_id_24/checkpoint-400", "epoch": 0.3137254901960784, "eval_steps": 100, "global_step": 400, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.000784313725490196, "grad_norm": 13.154260635375977, "learning_rate": 7e-06, "loss": 4.2676, "step": 1 }, { "epoch": 0.000784313725490196, "eval_loss": 1.2241708040237427, "eval_runtime": 167.7852, "eval_samples_per_second": 12.802, "eval_steps_per_second": 3.201, "step": 1 }, { "epoch": 0.001568627450980392, "grad_norm": 15.154508590698242, "learning_rate": 1.4e-05, "loss": 2.7549, "step": 2 }, { "epoch": 0.002352941176470588, "grad_norm": 14.48685073852539, "learning_rate": 2.1e-05, "loss": 2.7878, "step": 3 }, { "epoch": 0.003137254901960784, "grad_norm": 13.144845008850098, "learning_rate": 2.8e-05, "loss": 2.3061, "step": 4 }, { "epoch": 0.00392156862745098, "grad_norm": 9.391257286071777, "learning_rate": 3.5e-05, "loss": 1.9517, "step": 5 }, { "epoch": 0.004705882352941176, "grad_norm": 8.726778984069824, "learning_rate": 4.2e-05, "loss": 2.1089, "step": 6 }, { "epoch": 0.005490196078431373, "grad_norm": 7.819472789764404, "learning_rate": 4.899999999999999e-05, "loss": 1.7353, "step": 7 }, { "epoch": 0.006274509803921568, "grad_norm": 7.163968086242676, "learning_rate": 5.6e-05, "loss": 1.7352, "step": 8 }, { "epoch": 0.007058823529411765, "grad_norm": 7.747889995574951, "learning_rate": 6.3e-05, "loss": 2.1403, "step": 9 }, { "epoch": 0.00784313725490196, "grad_norm": 7.355112075805664, "learning_rate": 7e-05, "loss": 1.7824, "step": 10 }, { "epoch": 0.008627450980392156, "grad_norm": 6.8065619468688965, "learning_rate": 6.999886445007847e-05, "loss": 1.7958, "step": 11 }, { "epoch": 0.009411764705882352, "grad_norm": 7.497786045074463, "learning_rate": 6.999545787399811e-05, "loss": 1.8139, "step": 12 }, { "epoch": 0.01019607843137255, "grad_norm": 8.02735710144043, "learning_rate": 6.998978049280675e-05, "loss": 1.7139, "step": 13 }, { "epoch": 0.010980392156862745, "grad_norm": 7.836209297180176, "learning_rate": 6.998183267490152e-05, "loss": 1.7008, "step": 14 }, { "epoch": 0.011764705882352941, "grad_norm": 6.723918914794922, "learning_rate": 6.997161493600493e-05, "loss": 1.5095, "step": 15 }, { "epoch": 0.012549019607843137, "grad_norm": 7.266279220581055, "learning_rate": 6.995912793913143e-05, "loss": 1.7114, "step": 16 }, { "epoch": 0.013333333333333334, "grad_norm": 8.910981178283691, "learning_rate": 6.994437249454433e-05, "loss": 1.8847, "step": 17 }, { "epoch": 0.01411764705882353, "grad_norm": 7.561214923858643, "learning_rate": 6.992734955970331e-05, "loss": 1.4478, "step": 18 }, { "epoch": 0.014901960784313726, "grad_norm": 11.097957611083984, "learning_rate": 6.990806023920219e-05, "loss": 1.8268, "step": 19 }, { "epoch": 0.01568627450980392, "grad_norm": 8.547746658325195, "learning_rate": 6.988650578469734e-05, "loss": 1.9715, "step": 20 }, { "epoch": 0.01647058823529412, "grad_norm": 7.6952643394470215, "learning_rate": 6.986268759482644e-05, "loss": 1.3908, "step": 21 }, { "epoch": 0.017254901960784313, "grad_norm": 9.783441543579102, "learning_rate": 6.983660721511769e-05, "loss": 1.6528, "step": 22 }, { "epoch": 0.01803921568627451, "grad_norm": 8.756048202514648, "learning_rate": 6.980826633788956e-05, "loss": 1.3471, "step": 23 }, { "epoch": 0.018823529411764704, "grad_norm": 10.131261825561523, "learning_rate": 6.977766680214095e-05, "loss": 1.6056, "step": 24 }, { "epoch": 0.0196078431372549, "grad_norm": 9.784903526306152, "learning_rate": 6.974481059343188e-05, "loss": 1.349, "step": 25 }, { "epoch": 0.0203921568627451, "grad_norm": 10.655691146850586, "learning_rate": 6.970969984375466e-05, "loss": 1.0902, "step": 26 }, { "epoch": 0.021176470588235293, "grad_norm": 10.72864055633545, "learning_rate": 6.967233683139552e-05, "loss": 1.4288, "step": 27 }, { "epoch": 0.02196078431372549, "grad_norm": 7.463002681732178, "learning_rate": 6.963272398078678e-05, "loss": 1.0672, "step": 28 }, { "epoch": 0.022745098039215685, "grad_norm": 11.380589485168457, "learning_rate": 6.959086386234956e-05, "loss": 1.3336, "step": 29 }, { "epoch": 0.023529411764705882, "grad_norm": 9.633968353271484, "learning_rate": 6.954675919232694e-05, "loss": 0.9633, "step": 30 }, { "epoch": 0.02431372549019608, "grad_norm": 8.88882827758789, "learning_rate": 6.950041283260778e-05, "loss": 1.0786, "step": 31 }, { "epoch": 0.025098039215686273, "grad_norm": 14.510672569274902, "learning_rate": 6.945182779054092e-05, "loss": 1.4122, "step": 32 }, { "epoch": 0.02588235294117647, "grad_norm": 15.632862091064453, "learning_rate": 6.940100721874013e-05, "loss": 2.0142, "step": 33 }, { "epoch": 0.02666666666666667, "grad_norm": 12.684489250183105, "learning_rate": 6.934795441487947e-05, "loss": 1.6836, "step": 34 }, { "epoch": 0.027450980392156862, "grad_norm": 13.649406433105469, "learning_rate": 6.929267282147936e-05, "loss": 1.7083, "step": 35 }, { "epoch": 0.02823529411764706, "grad_norm": 10.714078903198242, "learning_rate": 6.923516602568319e-05, "loss": 1.6407, "step": 36 }, { "epoch": 0.029019607843137254, "grad_norm": 10.981186866760254, "learning_rate": 6.917543775902452e-05, "loss": 1.483, "step": 37 }, { "epoch": 0.02980392156862745, "grad_norm": 12.960887908935547, "learning_rate": 6.911349189718497e-05, "loss": 1.7072, "step": 38 }, { "epoch": 0.03058823529411765, "grad_norm": 11.793508529663086, "learning_rate": 6.904933245974274e-05, "loss": 2.088, "step": 39 }, { "epoch": 0.03137254901960784, "grad_norm": 14.335042953491211, "learning_rate": 6.898296360991182e-05, "loss": 1.5573, "step": 40 }, { "epoch": 0.03215686274509804, "grad_norm": 19.681434631347656, "learning_rate": 6.891438965427172e-05, "loss": 1.7118, "step": 41 }, { "epoch": 0.03294117647058824, "grad_norm": 14.27527904510498, "learning_rate": 6.884361504248818e-05, "loss": 2.44, "step": 42 }, { "epoch": 0.03372549019607843, "grad_norm": 16.58424186706543, "learning_rate": 6.87706443670243e-05, "loss": 1.6355, "step": 43 }, { "epoch": 0.034509803921568626, "grad_norm": 20.70366096496582, "learning_rate": 6.869548236284267e-05, "loss": 2.2336, "step": 44 }, { "epoch": 0.03529411764705882, "grad_norm": 13.909195899963379, "learning_rate": 6.861813390709803e-05, "loss": 1.6299, "step": 45 }, { "epoch": 0.03607843137254902, "grad_norm": 18.837871551513672, "learning_rate": 6.853860401882081e-05, "loss": 2.6585, "step": 46 }, { "epoch": 0.03686274509803922, "grad_norm": 32.49602127075195, "learning_rate": 6.845689785859151e-05, "loss": 2.2312, "step": 47 }, { "epoch": 0.03764705882352941, "grad_norm": 22.12024688720703, "learning_rate": 6.837302072820576e-05, "loss": 2.5807, "step": 48 }, { "epoch": 0.038431372549019606, "grad_norm": 32.349609375, "learning_rate": 6.828697807033038e-05, "loss": 2.6884, "step": 49 }, { "epoch": 0.0392156862745098, "grad_norm": 38.483177185058594, "learning_rate": 6.819877546815009e-05, "loss": 3.4595, "step": 50 }, { "epoch": 0.04, "grad_norm": 11.971654891967773, "learning_rate": 6.810841864500537e-05, "loss": 3.9855, "step": 51 }, { "epoch": 0.0407843137254902, "grad_norm": 5.677441120147705, "learning_rate": 6.8015913464021e-05, "loss": 2.2597, "step": 52 }, { "epoch": 0.04156862745098039, "grad_norm": 3.608100652694702, "learning_rate": 6.79212659277256e-05, "loss": 1.8872, "step": 53 }, { "epoch": 0.042352941176470586, "grad_norm": 3.213096857070923, "learning_rate": 6.782448217766216e-05, "loss": 1.4016, "step": 54 }, { "epoch": 0.043137254901960784, "grad_norm": 4.3245744705200195, "learning_rate": 6.772556849398952e-05, "loss": 1.5969, "step": 55 }, { "epoch": 0.04392156862745098, "grad_norm": 4.572765350341797, "learning_rate": 6.762453129507487e-05, "loss": 1.5966, "step": 56 }, { "epoch": 0.04470588235294118, "grad_norm": 2.8827388286590576, "learning_rate": 6.752137713707726e-05, "loss": 1.0312, "step": 57 }, { "epoch": 0.04549019607843137, "grad_norm": 3.4303178787231445, "learning_rate": 6.741611271352217e-05, "loss": 1.3693, "step": 58 }, { "epoch": 0.04627450980392157, "grad_norm": 3.742342948913574, "learning_rate": 6.730874485486721e-05, "loss": 1.2792, "step": 59 }, { "epoch": 0.047058823529411764, "grad_norm": 3.946136713027954, "learning_rate": 6.719928052805884e-05, "loss": 1.4547, "step": 60 }, { "epoch": 0.04784313725490196, "grad_norm": 3.579709768295288, "learning_rate": 6.708772683608038e-05, "loss": 1.265, "step": 61 }, { "epoch": 0.04862745098039216, "grad_norm": 3.937344551086426, "learning_rate": 6.697409101749102e-05, "loss": 1.475, "step": 62 }, { "epoch": 0.04941176470588235, "grad_norm": 4.387577056884766, "learning_rate": 6.685838044595621e-05, "loss": 1.5258, "step": 63 }, { "epoch": 0.05019607843137255, "grad_norm": 5.537537574768066, "learning_rate": 6.674060262976909e-05, "loss": 1.426, "step": 64 }, { "epoch": 0.050980392156862744, "grad_norm": 4.949011325836182, "learning_rate": 6.662076521136337e-05, "loss": 1.6677, "step": 65 }, { "epoch": 0.05176470588235294, "grad_norm": 5.1415114402771, "learning_rate": 6.649887596681742e-05, "loss": 1.5576, "step": 66 }, { "epoch": 0.05254901960784314, "grad_norm": 4.477947235107422, "learning_rate": 6.637494280534962e-05, "loss": 1.4792, "step": 67 }, { "epoch": 0.05333333333333334, "grad_norm": 5.3575758934021, "learning_rate": 6.624897376880523e-05, "loss": 1.6056, "step": 68 }, { "epoch": 0.05411764705882353, "grad_norm": 5.173808574676514, "learning_rate": 6.612097703113452e-05, "loss": 1.1905, "step": 69 }, { "epoch": 0.054901960784313725, "grad_norm": 5.39131498336792, "learning_rate": 6.599096089786233e-05, "loss": 1.5183, "step": 70 }, { "epoch": 0.05568627450980392, "grad_norm": 6.076783180236816, "learning_rate": 6.585893380554929e-05, "loss": 1.0184, "step": 71 }, { "epoch": 0.05647058823529412, "grad_norm": 7.40262508392334, "learning_rate": 6.572490432124416e-05, "loss": 1.6043, "step": 72 }, { "epoch": 0.05725490196078432, "grad_norm": 6.334597587585449, "learning_rate": 6.558888114192812e-05, "loss": 1.3284, "step": 73 }, { "epoch": 0.05803921568627451, "grad_norm": 6.603546619415283, "learning_rate": 6.545087309395037e-05, "loss": 1.4609, "step": 74 }, { "epoch": 0.058823529411764705, "grad_norm": 10.423970222473145, "learning_rate": 6.531088913245535e-05, "loss": 1.5566, "step": 75 }, { "epoch": 0.0596078431372549, "grad_norm": 7.711127758026123, "learning_rate": 6.516893834080172e-05, "loss": 1.525, "step": 76 }, { "epoch": 0.0603921568627451, "grad_norm": 6.958899974822998, "learning_rate": 6.502502992997292e-05, "loss": 1.6027, "step": 77 }, { "epoch": 0.0611764705882353, "grad_norm": 8.05106258392334, "learning_rate": 6.487917323797953e-05, "loss": 1.5831, "step": 78 }, { "epoch": 0.06196078431372549, "grad_norm": 5.530167579650879, "learning_rate": 6.473137772925323e-05, "loss": 1.0131, "step": 79 }, { "epoch": 0.06274509803921569, "grad_norm": 6.529147624969482, "learning_rate": 6.458165299403281e-05, "loss": 1.026, "step": 80 }, { "epoch": 0.06352941176470588, "grad_norm": 4.573212623596191, "learning_rate": 6.443000874774177e-05, "loss": 0.6019, "step": 81 }, { "epoch": 0.06431372549019608, "grad_norm": 6.415572166442871, "learning_rate": 6.427645483035793e-05, "loss": 1.2427, "step": 82 }, { "epoch": 0.06509803921568627, "grad_norm": 12.587393760681152, "learning_rate": 6.412100120577495e-05, "loss": 0.8412, "step": 83 }, { "epoch": 0.06588235294117648, "grad_norm": 8.243964195251465, "learning_rate": 6.396365796115574e-05, "loss": 1.4314, "step": 84 }, { "epoch": 0.06666666666666667, "grad_norm": 10.298484802246094, "learning_rate": 6.380443530627796e-05, "loss": 1.471, "step": 85 }, { "epoch": 0.06745098039215686, "grad_norm": 7.470930099487305, "learning_rate": 6.364334357287154e-05, "loss": 1.6067, "step": 86 }, { "epoch": 0.06823529411764706, "grad_norm": 7.880890846252441, "learning_rate": 6.348039321394819e-05, "loss": 1.7093, "step": 87 }, { "epoch": 0.06901960784313725, "grad_norm": 8.50384521484375, "learning_rate": 6.331559480312315e-05, "loss": 1.5957, "step": 88 }, { "epoch": 0.06980392156862746, "grad_norm": 11.584599494934082, "learning_rate": 6.314895903392917e-05, "loss": 2.2822, "step": 89 }, { "epoch": 0.07058823529411765, "grad_norm": 8.506998062133789, "learning_rate": 6.298049671912254e-05, "loss": 1.3157, "step": 90 }, { "epoch": 0.07137254901960784, "grad_norm": 8.96076488494873, "learning_rate": 6.281021878998139e-05, "loss": 1.399, "step": 91 }, { "epoch": 0.07215686274509804, "grad_norm": 7.781743049621582, "learning_rate": 6.263813629559657e-05, "loss": 1.3365, "step": 92 }, { "epoch": 0.07294117647058823, "grad_norm": 11.301823616027832, "learning_rate": 6.246426040215451e-05, "loss": 1.931, "step": 93 }, { "epoch": 0.07372549019607844, "grad_norm": 14.643918991088867, "learning_rate": 6.228860239221277e-05, "loss": 1.949, "step": 94 }, { "epoch": 0.07450980392156863, "grad_norm": 12.800115585327148, "learning_rate": 6.21111736639679e-05, "loss": 1.9397, "step": 95 }, { "epoch": 0.07529411764705882, "grad_norm": 15.508123397827148, "learning_rate": 6.193198573051581e-05, "loss": 2.5479, "step": 96 }, { "epoch": 0.07607843137254902, "grad_norm": 22.137767791748047, "learning_rate": 6.175105021910468e-05, "loss": 2.0942, "step": 97 }, { "epoch": 0.07686274509803921, "grad_norm": 17.05497169494629, "learning_rate": 6.156837887038059e-05, "loss": 2.3568, "step": 98 }, { "epoch": 0.07764705882352942, "grad_norm": 18.51692771911621, "learning_rate": 6.138398353762557e-05, "loss": 2.4071, "step": 99 }, { "epoch": 0.0784313725490196, "grad_norm": 18.45566177368164, "learning_rate": 6.119787618598853e-05, "loss": 1.7988, "step": 100 }, { "epoch": 0.0784313725490196, "eval_loss": 0.4173365831375122, "eval_runtime": 169.3484, "eval_samples_per_second": 12.684, "eval_steps_per_second": 3.171, "step": 100 }, { "epoch": 0.0792156862745098, "grad_norm": 4.533246040344238, "learning_rate": 6.101006889170879e-05, "loss": 3.0858, "step": 101 }, { "epoch": 0.08, "grad_norm": 3.2407262325286865, "learning_rate": 6.0820573841332556e-05, "loss": 1.9589, "step": 102 }, { "epoch": 0.08078431372549019, "grad_norm": 2.6930501461029053, "learning_rate": 6.062940333092207e-05, "loss": 1.269, "step": 103 }, { "epoch": 0.0815686274509804, "grad_norm": 2.588240146636963, "learning_rate": 6.043656976525781e-05, "loss": 1.2504, "step": 104 }, { "epoch": 0.08235294117647059, "grad_norm": 3.165269374847412, "learning_rate": 6.0242085657033506e-05, "loss": 1.6896, "step": 105 }, { "epoch": 0.08313725490196078, "grad_norm": 2.964270830154419, "learning_rate": 6.004596362604423e-05, "loss": 1.2404, "step": 106 }, { "epoch": 0.08392156862745098, "grad_norm": 3.3895821571350098, "learning_rate": 5.9848216398367535e-05, "loss": 1.2462, "step": 107 }, { "epoch": 0.08470588235294117, "grad_norm": 3.5753889083862305, "learning_rate": 5.964885680553763e-05, "loss": 1.4962, "step": 108 }, { "epoch": 0.08549019607843138, "grad_norm": 3.2786760330200195, "learning_rate": 5.94478977837128e-05, "loss": 1.5371, "step": 109 }, { "epoch": 0.08627450980392157, "grad_norm": 3.7601850032806396, "learning_rate": 5.9245352372835974e-05, "loss": 1.1844, "step": 110 }, { "epoch": 0.08705882352941176, "grad_norm": 3.1973512172698975, "learning_rate": 5.9041233715788624e-05, "loss": 1.1938, "step": 111 }, { "epoch": 0.08784313725490196, "grad_norm": 3.321074962615967, "learning_rate": 5.88355550575379e-05, "loss": 1.3623, "step": 112 }, { "epoch": 0.08862745098039215, "grad_norm": 3.9187910556793213, "learning_rate": 5.862832974427716e-05, "loss": 1.2505, "step": 113 }, { "epoch": 0.08941176470588236, "grad_norm": 3.5713284015655518, "learning_rate": 5.8419571222560034e-05, "loss": 1.2605, "step": 114 }, { "epoch": 0.09019607843137255, "grad_norm": 3.8370423316955566, "learning_rate": 5.8209293038427824e-05, "loss": 1.1408, "step": 115 }, { "epoch": 0.09098039215686274, "grad_norm": 4.057117938995361, "learning_rate": 5.7997508836530576e-05, "loss": 1.4296, "step": 116 }, { "epoch": 0.09176470588235294, "grad_norm": 5.31479024887085, "learning_rate": 5.778423235924163e-05, "loss": 1.3345, "step": 117 }, { "epoch": 0.09254901960784313, "grad_norm": 4.572357177734375, "learning_rate": 5.7569477445765964e-05, "loss": 1.1513, "step": 118 }, { "epoch": 0.09333333333333334, "grad_norm": 5.681028366088867, "learning_rate": 5.735325803124217e-05, "loss": 1.4338, "step": 119 }, { "epoch": 0.09411764705882353, "grad_norm": 4.767769813537598, "learning_rate": 5.71355881458382e-05, "loss": 1.3424, "step": 120 }, { "epoch": 0.09490196078431372, "grad_norm": 6.746960639953613, "learning_rate": 5.6916481913840975e-05, "loss": 1.3405, "step": 121 }, { "epoch": 0.09568627450980392, "grad_norm": 4.564548492431641, "learning_rate": 5.669595355273991e-05, "loss": 1.2797, "step": 122 }, { "epoch": 0.09647058823529411, "grad_norm": 4.359226703643799, "learning_rate": 5.64740173723043e-05, "loss": 1.006, "step": 123 }, { "epoch": 0.09725490196078432, "grad_norm": 5.169023036956787, "learning_rate": 5.625068777365489e-05, "loss": 1.1459, "step": 124 }, { "epoch": 0.09803921568627451, "grad_norm": 5.223190784454346, "learning_rate": 5.6025979248329255e-05, "loss": 1.2089, "step": 125 }, { "epoch": 0.0988235294117647, "grad_norm": 6.859619617462158, "learning_rate": 5.579990637734161e-05, "loss": 0.9527, "step": 126 }, { "epoch": 0.0996078431372549, "grad_norm": 4.413240432739258, "learning_rate": 5.557248383023655e-05, "loss": 0.9772, "step": 127 }, { "epoch": 0.1003921568627451, "grad_norm": 6.224684238433838, "learning_rate": 5.5343726364137276e-05, "loss": 1.0011, "step": 128 }, { "epoch": 0.1011764705882353, "grad_norm": 5.601330280303955, "learning_rate": 5.511364882278793e-05, "loss": 1.249, "step": 129 }, { "epoch": 0.10196078431372549, "grad_norm": 5.103614807128906, "learning_rate": 5.488226613559045e-05, "loss": 1.1634, "step": 130 }, { "epoch": 0.1027450980392157, "grad_norm": 7.077230453491211, "learning_rate": 5.464959331663583e-05, "loss": 1.5048, "step": 131 }, { "epoch": 0.10352941176470588, "grad_norm": 5.399044990539551, "learning_rate": 5.4415645463729855e-05, "loss": 1.0851, "step": 132 }, { "epoch": 0.10431372549019607, "grad_norm": 7.850146770477295, "learning_rate": 5.4180437757413455e-05, "loss": 1.1617, "step": 133 }, { "epoch": 0.10509803921568628, "grad_norm": 5.438139915466309, "learning_rate": 5.394398545997759e-05, "loss": 0.9825, "step": 134 }, { "epoch": 0.10588235294117647, "grad_norm": 5.945227146148682, "learning_rate": 5.370630391447303e-05, "loss": 1.165, "step": 135 }, { "epoch": 0.10666666666666667, "grad_norm": 7.722451686859131, "learning_rate": 5.346740854371464e-05, "loss": 1.9443, "step": 136 }, { "epoch": 0.10745098039215686, "grad_norm": 7.847848892211914, "learning_rate": 5.322731484928068e-05, "loss": 1.4477, "step": 137 }, { "epoch": 0.10823529411764705, "grad_norm": 8.356393814086914, "learning_rate": 5.29860384105069e-05, "loss": 1.6004, "step": 138 }, { "epoch": 0.10901960784313726, "grad_norm": 8.85940170288086, "learning_rate": 5.2743594883475664e-05, "loss": 1.4381, "step": 139 }, { "epoch": 0.10980392156862745, "grad_norm": 8.563812255859375, "learning_rate": 5.2499999999999995e-05, "loss": 1.5443, "step": 140 }, { "epoch": 0.11058823529411765, "grad_norm": 8.796785354614258, "learning_rate": 5.225526956660281e-05, "loss": 1.5716, "step": 141 }, { "epoch": 0.11137254901960784, "grad_norm": 11.239343643188477, "learning_rate": 5.200941946349121e-05, "loss": 2.2336, "step": 142 }, { "epoch": 0.11215686274509803, "grad_norm": 10.173134803771973, "learning_rate": 5.176246564352607e-05, "loss": 1.8621, "step": 143 }, { "epoch": 0.11294117647058824, "grad_norm": 7.808456897735596, "learning_rate": 5.1514424131186865e-05, "loss": 1.3717, "step": 144 }, { "epoch": 0.11372549019607843, "grad_norm": 11.386785507202148, "learning_rate": 5.12653110215319e-05, "loss": 1.9377, "step": 145 }, { "epoch": 0.11450980392156863, "grad_norm": 11.134692192077637, "learning_rate": 5.101514247915384e-05, "loss": 2.1051, "step": 146 }, { "epoch": 0.11529411764705882, "grad_norm": 11.531853675842285, "learning_rate": 5.0763934737130886e-05, "loss": 2.1194, "step": 147 }, { "epoch": 0.11607843137254902, "grad_norm": 12.823151588439941, "learning_rate": 5.0511704095973445e-05, "loss": 2.008, "step": 148 }, { "epoch": 0.11686274509803922, "grad_norm": 14.165589332580566, "learning_rate": 5.0258466922566345e-05, "loss": 2.579, "step": 149 }, { "epoch": 0.11764705882352941, "grad_norm": 22.23053741455078, "learning_rate": 5.000423964910689e-05, "loss": 2.1285, "step": 150 }, { "epoch": 0.11843137254901961, "grad_norm": 3.6735522747039795, "learning_rate": 4.974903877203852e-05, "loss": 3.2373, "step": 151 }, { "epoch": 0.1192156862745098, "grad_norm": 2.631013870239258, "learning_rate": 4.949288085098041e-05, "loss": 1.7498, "step": 152 }, { "epoch": 0.12, "grad_norm": 2.442901372909546, "learning_rate": 4.9235782507653006e-05, "loss": 1.7117, "step": 153 }, { "epoch": 0.1207843137254902, "grad_norm": 2.393778085708618, "learning_rate": 4.897776042479934e-05, "loss": 1.3329, "step": 154 }, { "epoch": 0.12156862745098039, "grad_norm": 2.4069385528564453, "learning_rate": 4.8718831345102625e-05, "loss": 1.3254, "step": 155 }, { "epoch": 0.1223529411764706, "grad_norm": 2.7312748432159424, "learning_rate": 4.845901207009978e-05, "loss": 1.1983, "step": 156 }, { "epoch": 0.12313725490196079, "grad_norm": 2.8144845962524414, "learning_rate": 4.819831945909121e-05, "loss": 1.1755, "step": 157 }, { "epoch": 0.12392156862745098, "grad_norm": 3.241198778152466, "learning_rate": 4.793677042804688e-05, "loss": 1.1232, "step": 158 }, { "epoch": 0.12470588235294118, "grad_norm": 3.7568423748016357, "learning_rate": 4.767438194850858e-05, "loss": 1.4093, "step": 159 }, { "epoch": 0.12549019607843137, "grad_norm": 3.8611888885498047, "learning_rate": 4.741117104648874e-05, "loss": 1.3189, "step": 160 }, { "epoch": 0.12627450980392158, "grad_norm": 3.7197554111480713, "learning_rate": 4.71471548013656e-05, "loss": 1.4241, "step": 161 }, { "epoch": 0.12705882352941175, "grad_norm": 3.673598527908325, "learning_rate": 4.688235034477495e-05, "loss": 1.2979, "step": 162 }, { "epoch": 0.12784313725490196, "grad_norm": 3.9346017837524414, "learning_rate": 4.6616774859498493e-05, "loss": 1.4983, "step": 163 }, { "epoch": 0.12862745098039216, "grad_norm": 3.7387478351593018, "learning_rate": 4.635044557834889e-05, "loss": 1.5969, "step": 164 }, { "epoch": 0.12941176470588237, "grad_norm": 4.286586284637451, "learning_rate": 4.6083379783051544e-05, "loss": 1.1618, "step": 165 }, { "epoch": 0.13019607843137254, "grad_norm": 3.6195132732391357, "learning_rate": 4.581559480312316e-05, "loss": 1.3013, "step": 166 }, { "epoch": 0.13098039215686275, "grad_norm": 3.9469423294067383, "learning_rate": 4.554710801474736e-05, "loss": 1.2563, "step": 167 }, { "epoch": 0.13176470588235295, "grad_norm": 3.9838595390319824, "learning_rate": 4.527793683964708e-05, "loss": 1.084, "step": 168 }, { "epoch": 0.13254901960784313, "grad_norm": 4.2628021240234375, "learning_rate": 4.5008098743954114e-05, "loss": 1.1541, "step": 169 }, { "epoch": 0.13333333333333333, "grad_norm": 4.49942684173584, "learning_rate": 4.473761123707584e-05, "loss": 1.4258, "step": 170 }, { "epoch": 0.13411764705882354, "grad_norm": 5.674098968505859, "learning_rate": 4.4466491870558936e-05, "loss": 1.8754, "step": 171 }, { "epoch": 0.1349019607843137, "grad_norm": 5.979763984680176, "learning_rate": 4.419475823695054e-05, "loss": 1.3147, "step": 172 }, { "epoch": 0.13568627450980392, "grad_norm": 4.832103252410889, "learning_rate": 4.392242796865676e-05, "loss": 1.2014, "step": 173 }, { "epoch": 0.13647058823529412, "grad_norm": 4.375277996063232, "learning_rate": 4.364951873679843e-05, "loss": 1.1941, "step": 174 }, { "epoch": 0.13725490196078433, "grad_norm": 4.7891526222229, "learning_rate": 4.337604825006452e-05, "loss": 1.1638, "step": 175 }, { "epoch": 0.1380392156862745, "grad_norm": 5.445965766906738, "learning_rate": 4.310203425356301e-05, "loss": 1.054, "step": 176 }, { "epoch": 0.1388235294117647, "grad_norm": 4.09723424911499, "learning_rate": 4.282749452766945e-05, "loss": 0.6077, "step": 177 }, { "epoch": 0.1396078431372549, "grad_norm": 5.525692939758301, "learning_rate": 4.255244688687321e-05, "loss": 1.0198, "step": 178 }, { "epoch": 0.1403921568627451, "grad_norm": 5.302620887756348, "learning_rate": 4.227690917862157e-05, "loss": 0.9727, "step": 179 }, { "epoch": 0.1411764705882353, "grad_norm": 4.293851375579834, "learning_rate": 4.200089928216155e-05, "loss": 0.7823, "step": 180 }, { "epoch": 0.1419607843137255, "grad_norm": 6.371180057525635, "learning_rate": 4.1724435107379795e-05, "loss": 1.2332, "step": 181 }, { "epoch": 0.14274509803921567, "grad_norm": 7.856477737426758, "learning_rate": 4.14475345936404e-05, "loss": 1.3729, "step": 182 }, { "epoch": 0.14352941176470588, "grad_norm": 4.569979190826416, "learning_rate": 4.1170215708620915e-05, "loss": 0.8416, "step": 183 }, { "epoch": 0.14431372549019608, "grad_norm": 6.07737398147583, "learning_rate": 4.089249644714634e-05, "loss": 1.4574, "step": 184 }, { "epoch": 0.1450980392156863, "grad_norm": 6.413763523101807, "learning_rate": 4.0614394830021604e-05, "loss": 1.5253, "step": 185 }, { "epoch": 0.14588235294117646, "grad_norm": 5.022174835205078, "learning_rate": 4.0335928902862086e-05, "loss": 0.9277, "step": 186 }, { "epoch": 0.14666666666666667, "grad_norm": 7.028177261352539, "learning_rate": 4.005711673492274e-05, "loss": 1.6335, "step": 187 }, { "epoch": 0.14745098039215687, "grad_norm": 7.276490211486816, "learning_rate": 3.977797641792561e-05, "loss": 1.3358, "step": 188 }, { "epoch": 0.14823529411764705, "grad_norm": 8.936336517333984, "learning_rate": 3.9498526064885806e-05, "loss": 1.8587, "step": 189 }, { "epoch": 0.14901960784313725, "grad_norm": 9.971092224121094, "learning_rate": 3.92187838089363e-05, "loss": 1.9026, "step": 190 }, { "epoch": 0.14980392156862746, "grad_norm": 8.014878273010254, "learning_rate": 3.893876780215119e-05, "loss": 1.6085, "step": 191 }, { "epoch": 0.15058823529411763, "grad_norm": 9.771968841552734, "learning_rate": 3.865849621436787e-05, "loss": 1.8793, "step": 192 }, { "epoch": 0.15137254901960784, "grad_norm": 11.60059928894043, "learning_rate": 3.8377987232008025e-05, "loss": 1.8012, "step": 193 }, { "epoch": 0.15215686274509804, "grad_norm": 9.739047050476074, "learning_rate": 3.809725905689754e-05, "loss": 1.9012, "step": 194 }, { "epoch": 0.15294117647058825, "grad_norm": 12.654723167419434, "learning_rate": 3.781632990508541e-05, "loss": 2.3507, "step": 195 }, { "epoch": 0.15372549019607842, "grad_norm": 11.12818717956543, "learning_rate": 3.7535218005661685e-05, "loss": 1.877, "step": 196 }, { "epoch": 0.15450980392156863, "grad_norm": 13.460821151733398, "learning_rate": 3.7253941599574675e-05, "loss": 2.0457, "step": 197 }, { "epoch": 0.15529411764705883, "grad_norm": 16.934213638305664, "learning_rate": 3.6972518938447306e-05, "loss": 2.5687, "step": 198 }, { "epoch": 0.156078431372549, "grad_norm": 17.341718673706055, "learning_rate": 3.6690968283392746e-05, "loss": 2.6656, "step": 199 }, { "epoch": 0.1568627450980392, "grad_norm": 13.909801483154297, "learning_rate": 3.640930790382953e-05, "loss": 2.0502, "step": 200 }, { "epoch": 0.1568627450980392, "eval_loss": 0.36784711480140686, "eval_runtime": 169.7291, "eval_samples_per_second": 12.655, "eval_steps_per_second": 3.164, "step": 200 }, { "epoch": 0.15764705882352942, "grad_norm": 2.9938836097717285, "learning_rate": 3.6127556076296055e-05, "loss": 3.0403, "step": 201 }, { "epoch": 0.1584313725490196, "grad_norm": 2.702965497970581, "learning_rate": 3.5845731083264626e-05, "loss": 1.9497, "step": 202 }, { "epoch": 0.1592156862745098, "grad_norm": 2.5172975063323975, "learning_rate": 3.5563851211955195e-05, "loss": 1.5842, "step": 203 }, { "epoch": 0.16, "grad_norm": 2.603217363357544, "learning_rate": 3.528193475314864e-05, "loss": 1.5727, "step": 204 }, { "epoch": 0.1607843137254902, "grad_norm": 2.615779399871826, "learning_rate": 3.5e-05, "loss": 1.3275, "step": 205 }, { "epoch": 0.16156862745098038, "grad_norm": 2.509752035140991, "learning_rate": 3.4718065246851354e-05, "loss": 1.2942, "step": 206 }, { "epoch": 0.1623529411764706, "grad_norm": 2.902695655822754, "learning_rate": 3.4436148788044806e-05, "loss": 1.1889, "step": 207 }, { "epoch": 0.1631372549019608, "grad_norm": 3.608950138092041, "learning_rate": 3.4154268916735375e-05, "loss": 1.4452, "step": 208 }, { "epoch": 0.16392156862745097, "grad_norm": 2.8737595081329346, "learning_rate": 3.387244392370395e-05, "loss": 1.2904, "step": 209 }, { "epoch": 0.16470588235294117, "grad_norm": 3.235945463180542, "learning_rate": 3.3590692096170476e-05, "loss": 1.2737, "step": 210 }, { "epoch": 0.16549019607843138, "grad_norm": 3.6993987560272217, "learning_rate": 3.3309031716607255e-05, "loss": 1.3786, "step": 211 }, { "epoch": 0.16627450980392156, "grad_norm": 3.325028419494629, "learning_rate": 3.302748106155268e-05, "loss": 1.1542, "step": 212 }, { "epoch": 0.16705882352941176, "grad_norm": 2.829763650894165, "learning_rate": 3.274605840042531e-05, "loss": 0.9143, "step": 213 }, { "epoch": 0.16784313725490196, "grad_norm": 4.127065181732178, "learning_rate": 3.2464781994338315e-05, "loss": 1.4021, "step": 214 }, { "epoch": 0.16862745098039217, "grad_norm": 3.2039248943328857, "learning_rate": 3.2183670094914595e-05, "loss": 1.1642, "step": 215 }, { "epoch": 0.16941176470588235, "grad_norm": 3.7868335247039795, "learning_rate": 3.190274094310245e-05, "loss": 1.3842, "step": 216 }, { "epoch": 0.17019607843137255, "grad_norm": 3.718687057495117, "learning_rate": 3.162201276799197e-05, "loss": 1.4267, "step": 217 }, { "epoch": 0.17098039215686275, "grad_norm": 3.7910690307617188, "learning_rate": 3.134150378563213e-05, "loss": 1.2455, "step": 218 }, { "epoch": 0.17176470588235293, "grad_norm": 3.5929932594299316, "learning_rate": 3.1061232197848805e-05, "loss": 1.2007, "step": 219 }, { "epoch": 0.17254901960784313, "grad_norm": 4.848661422729492, "learning_rate": 3.07812161910637e-05, "loss": 1.5243, "step": 220 }, { "epoch": 0.17333333333333334, "grad_norm": 4.240635395050049, "learning_rate": 3.050147393511419e-05, "loss": 1.2182, "step": 221 }, { "epoch": 0.17411764705882352, "grad_norm": 4.414323806762695, "learning_rate": 3.0222023582074396e-05, "loss": 1.3929, "step": 222 }, { "epoch": 0.17490196078431372, "grad_norm": 4.253678321838379, "learning_rate": 2.994288326507726e-05, "loss": 1.0231, "step": 223 }, { "epoch": 0.17568627450980392, "grad_norm": 4.7787089347839355, "learning_rate": 2.9664071097137908e-05, "loss": 1.3696, "step": 224 }, { "epoch": 0.17647058823529413, "grad_norm": 6.1173248291015625, "learning_rate": 2.9385605169978387e-05, "loss": 1.3486, "step": 225 }, { "epoch": 0.1772549019607843, "grad_norm": 5.057071685791016, "learning_rate": 2.9107503552853648e-05, "loss": 1.1463, "step": 226 }, { "epoch": 0.1780392156862745, "grad_norm": 4.796481132507324, "learning_rate": 2.8829784291379082e-05, "loss": 1.0947, "step": 227 }, { "epoch": 0.17882352941176471, "grad_norm": 4.333375930786133, "learning_rate": 2.8552465406359593e-05, "loss": 1.0557, "step": 228 }, { "epoch": 0.1796078431372549, "grad_norm": 4.937939167022705, "learning_rate": 2.8275564892620202e-05, "loss": 0.8979, "step": 229 }, { "epoch": 0.1803921568627451, "grad_norm": 5.160710334777832, "learning_rate": 2.799910071783845e-05, "loss": 0.9403, "step": 230 }, { "epoch": 0.1811764705882353, "grad_norm": 7.218094348907471, "learning_rate": 2.7723090821378425e-05, "loss": 0.899, "step": 231 }, { "epoch": 0.18196078431372548, "grad_norm": 4.6421332359313965, "learning_rate": 2.7447553113126786e-05, "loss": 0.934, "step": 232 }, { "epoch": 0.18274509803921568, "grad_norm": 7.592020034790039, "learning_rate": 2.7172505472330564e-05, "loss": 0.9639, "step": 233 }, { "epoch": 0.18352941176470589, "grad_norm": 6.882318496704102, "learning_rate": 2.6897965746436997e-05, "loss": 1.3508, "step": 234 }, { "epoch": 0.1843137254901961, "grad_norm": 6.335398197174072, "learning_rate": 2.6623951749935487e-05, "loss": 1.2642, "step": 235 }, { "epoch": 0.18509803921568627, "grad_norm": 5.449366569519043, "learning_rate": 2.635048126320156e-05, "loss": 1.4236, "step": 236 }, { "epoch": 0.18588235294117647, "grad_norm": 8.11070728302002, "learning_rate": 2.6077572031343227e-05, "loss": 1.3364, "step": 237 }, { "epoch": 0.18666666666666668, "grad_norm": 6.608270168304443, "learning_rate": 2.5805241763049455e-05, "loss": 1.4484, "step": 238 }, { "epoch": 0.18745098039215685, "grad_norm": 7.942669868469238, "learning_rate": 2.553350812944107e-05, "loss": 1.4957, "step": 239 }, { "epoch": 0.18823529411764706, "grad_norm": 7.074437618255615, "learning_rate": 2.5262388762924156e-05, "loss": 1.3078, "step": 240 }, { "epoch": 0.18901960784313726, "grad_norm": 7.799503326416016, "learning_rate": 2.499190125604588e-05, "loss": 1.5845, "step": 241 }, { "epoch": 0.18980392156862744, "grad_norm": 9.548646926879883, "learning_rate": 2.4722063160352926e-05, "loss": 2.075, "step": 242 }, { "epoch": 0.19058823529411764, "grad_norm": 9.75729751586914, "learning_rate": 2.4452891985252647e-05, "loss": 2.0484, "step": 243 }, { "epoch": 0.19137254901960785, "grad_norm": 8.452042579650879, "learning_rate": 2.4184405196876842e-05, "loss": 1.6378, "step": 244 }, { "epoch": 0.19215686274509805, "grad_norm": 7.993230819702148, "learning_rate": 2.3916620216948467e-05, "loss": 1.8914, "step": 245 }, { "epoch": 0.19294117647058823, "grad_norm": 20.330001831054688, "learning_rate": 2.3649554421651106e-05, "loss": 2.7448, "step": 246 }, { "epoch": 0.19372549019607843, "grad_norm": 11.52759838104248, "learning_rate": 2.338322514050151e-05, "loss": 1.4114, "step": 247 }, { "epoch": 0.19450980392156864, "grad_norm": 12.445416450500488, "learning_rate": 2.3117649655225048e-05, "loss": 2.3627, "step": 248 }, { "epoch": 0.1952941176470588, "grad_norm": 16.47171401977539, "learning_rate": 2.2852845198634395e-05, "loss": 3.235, "step": 249 }, { "epoch": 0.19607843137254902, "grad_norm": 14.980457305908203, "learning_rate": 2.258882895351125e-05, "loss": 2.4173, "step": 250 }, { "epoch": 0.19686274509803922, "grad_norm": 2.359992742538452, "learning_rate": 2.2325618051491415e-05, "loss": 3.2603, "step": 251 }, { "epoch": 0.1976470588235294, "grad_norm": 1.96599280834198, "learning_rate": 2.2063229571953114e-05, "loss": 1.6396, "step": 252 }, { "epoch": 0.1984313725490196, "grad_norm": 2.0115535259246826, "learning_rate": 2.1801680540908783e-05, "loss": 1.2482, "step": 253 }, { "epoch": 0.1992156862745098, "grad_norm": 2.1881866455078125, "learning_rate": 2.154098792990022e-05, "loss": 1.1678, "step": 254 }, { "epoch": 0.2, "grad_norm": 2.3712048530578613, "learning_rate": 2.1281168654897375e-05, "loss": 1.134, "step": 255 }, { "epoch": 0.2007843137254902, "grad_norm": 2.554553985595703, "learning_rate": 2.102223957520066e-05, "loss": 1.2021, "step": 256 }, { "epoch": 0.2015686274509804, "grad_norm": 2.6500658988952637, "learning_rate": 2.0764217492346998e-05, "loss": 1.0142, "step": 257 }, { "epoch": 0.2023529411764706, "grad_norm": 3.106962203979492, "learning_rate": 2.0507119149019585e-05, "loss": 1.3034, "step": 258 }, { "epoch": 0.20313725490196077, "grad_norm": 2.9345974922180176, "learning_rate": 2.0250961227961486e-05, "loss": 0.9994, "step": 259 }, { "epoch": 0.20392156862745098, "grad_norm": 3.326606035232544, "learning_rate": 1.9995760350893095e-05, "loss": 1.2394, "step": 260 }, { "epoch": 0.20470588235294118, "grad_norm": 2.8550965785980225, "learning_rate": 1.9741533077433642e-05, "loss": 1.1, "step": 261 }, { "epoch": 0.2054901960784314, "grad_norm": 3.312087297439575, "learning_rate": 1.9488295904026556e-05, "loss": 1.1038, "step": 262 }, { "epoch": 0.20627450980392156, "grad_norm": 3.6784210205078125, "learning_rate": 1.9236065262869107e-05, "loss": 1.1669, "step": 263 }, { "epoch": 0.20705882352941177, "grad_norm": 4.164744853973389, "learning_rate": 1.8984857520846164e-05, "loss": 1.5447, "step": 264 }, { "epoch": 0.20784313725490197, "grad_norm": 3.8596198558807373, "learning_rate": 1.87346889784681e-05, "loss": 1.3421, "step": 265 }, { "epoch": 0.20862745098039215, "grad_norm": 4.981409549713135, "learning_rate": 1.848557586881313e-05, "loss": 1.5625, "step": 266 }, { "epoch": 0.20941176470588235, "grad_norm": 5.043278694152832, "learning_rate": 1.823753435647393e-05, "loss": 1.1419, "step": 267 }, { "epoch": 0.21019607843137256, "grad_norm": 5.383882999420166, "learning_rate": 1.7990580536508787e-05, "loss": 1.6431, "step": 268 }, { "epoch": 0.21098039215686273, "grad_norm": 3.967205286026001, "learning_rate": 1.774473043339719e-05, "loss": 1.053, "step": 269 }, { "epoch": 0.21176470588235294, "grad_norm": 4.5731072425842285, "learning_rate": 1.7500000000000005e-05, "loss": 1.435, "step": 270 }, { "epoch": 0.21254901960784314, "grad_norm": 4.236204147338867, "learning_rate": 1.7256405116524344e-05, "loss": 0.9911, "step": 271 }, { "epoch": 0.21333333333333335, "grad_norm": 4.109267711639404, "learning_rate": 1.7013961589493098e-05, "loss": 0.8052, "step": 272 }, { "epoch": 0.21411764705882352, "grad_norm": 4.672120571136475, "learning_rate": 1.6772685150719313e-05, "loss": 1.013, "step": 273 }, { "epoch": 0.21490196078431373, "grad_norm": 6.016241550445557, "learning_rate": 1.653259145628535e-05, "loss": 1.5547, "step": 274 }, { "epoch": 0.21568627450980393, "grad_norm": 4.318469524383545, "learning_rate": 1.6293696085526958e-05, "loss": 1.0291, "step": 275 }, { "epoch": 0.2164705882352941, "grad_norm": 5.033075332641602, "learning_rate": 1.6056014540022405e-05, "loss": 0.8338, "step": 276 }, { "epoch": 0.2172549019607843, "grad_norm": 5.945756912231445, "learning_rate": 1.5819562242586556e-05, "loss": 1.2347, "step": 277 }, { "epoch": 0.21803921568627452, "grad_norm": 5.846539497375488, "learning_rate": 1.5584354536270135e-05, "loss": 0.9788, "step": 278 }, { "epoch": 0.2188235294117647, "grad_norm": 4.702358245849609, "learning_rate": 1.535040668336417e-05, "loss": 0.9638, "step": 279 }, { "epoch": 0.2196078431372549, "grad_norm": 5.319607734680176, "learning_rate": 1.511773386440955e-05, "loss": 1.1117, "step": 280 }, { "epoch": 0.2203921568627451, "grad_norm": 4.019665241241455, "learning_rate": 1.4886351177212074e-05, "loss": 0.6134, "step": 281 }, { "epoch": 0.2211764705882353, "grad_norm": 5.152523994445801, "learning_rate": 1.4656273635862728e-05, "loss": 1.053, "step": 282 }, { "epoch": 0.22196078431372548, "grad_norm": 6.206692218780518, "learning_rate": 1.4427516169763444e-05, "loss": 1.6169, "step": 283 }, { "epoch": 0.2227450980392157, "grad_norm": 5.860468864440918, "learning_rate": 1.4200093622658394e-05, "loss": 1.1937, "step": 284 }, { "epoch": 0.2235294117647059, "grad_norm": 5.6560773849487305, "learning_rate": 1.3974020751670732e-05, "loss": 1.4664, "step": 285 }, { "epoch": 0.22431372549019607, "grad_norm": 7.365577220916748, "learning_rate": 1.3749312226345108e-05, "loss": 1.4524, "step": 286 }, { "epoch": 0.22509803921568627, "grad_norm": 6.791868686676025, "learning_rate": 1.3525982627695691e-05, "loss": 0.9192, "step": 287 }, { "epoch": 0.22588235294117648, "grad_norm": 7.1987080574035645, "learning_rate": 1.3304046447260085e-05, "loss": 1.5558, "step": 288 }, { "epoch": 0.22666666666666666, "grad_norm": 8.44493293762207, "learning_rate": 1.308351808615902e-05, "loss": 1.6203, "step": 289 }, { "epoch": 0.22745098039215686, "grad_norm": 7.552881240844727, "learning_rate": 1.2864411854161799e-05, "loss": 1.6813, "step": 290 }, { "epoch": 0.22823529411764706, "grad_norm": 9.73569393157959, "learning_rate": 1.2646741968757827e-05, "loss": 1.6688, "step": 291 }, { "epoch": 0.22901960784313727, "grad_norm": 8.195075035095215, "learning_rate": 1.2430522554234036e-05, "loss": 1.8058, "step": 292 }, { "epoch": 0.22980392156862745, "grad_norm": 10.075333595275879, "learning_rate": 1.2215767640758367e-05, "loss": 1.983, "step": 293 }, { "epoch": 0.23058823529411765, "grad_norm": 9.949265480041504, "learning_rate": 1.2002491163469428e-05, "loss": 1.9262, "step": 294 }, { "epoch": 0.23137254901960785, "grad_norm": 9.431818962097168, "learning_rate": 1.1790706961572175e-05, "loss": 2.0562, "step": 295 }, { "epoch": 0.23215686274509803, "grad_norm": 21.495634078979492, "learning_rate": 1.1580428777439972e-05, "loss": 1.9353, "step": 296 }, { "epoch": 0.23294117647058823, "grad_norm": 9.102306365966797, "learning_rate": 1.1371670255722838e-05, "loss": 1.6257, "step": 297 }, { "epoch": 0.23372549019607844, "grad_norm": 14.28601360321045, "learning_rate": 1.1164444942462095e-05, "loss": 2.8148, "step": 298 }, { "epoch": 0.23450980392156862, "grad_norm": 18.514936447143555, "learning_rate": 1.0958766284211363e-05, "loss": 2.4219, "step": 299 }, { "epoch": 0.23529411764705882, "grad_norm": 17.440731048583984, "learning_rate": 1.0754647627164022e-05, "loss": 2.8099, "step": 300 }, { "epoch": 0.23529411764705882, "eval_loss": 0.35413047671318054, "eval_runtime": 168.8442, "eval_samples_per_second": 12.722, "eval_steps_per_second": 3.18, "step": 300 }, { "epoch": 0.23607843137254902, "grad_norm": 2.4137773513793945, "learning_rate": 1.0552102216287209e-05, "loss": 3.5898, "step": 301 }, { "epoch": 0.23686274509803923, "grad_norm": 2.2105584144592285, "learning_rate": 1.0351143194462377e-05, "loss": 1.8317, "step": 302 }, { "epoch": 0.2376470588235294, "grad_norm": 2.3090407848358154, "learning_rate": 1.0151783601632461e-05, "loss": 1.5182, "step": 303 }, { "epoch": 0.2384313725490196, "grad_norm": 2.195463180541992, "learning_rate": 9.954036373955764e-06, "loss": 1.2783, "step": 304 }, { "epoch": 0.23921568627450981, "grad_norm": 2.6486854553222656, "learning_rate": 9.757914342966493e-06, "loss": 1.4084, "step": 305 }, { "epoch": 0.24, "grad_norm": 3.396476984024048, "learning_rate": 9.563430234742194e-06, "loss": 1.3099, "step": 306 }, { "epoch": 0.2407843137254902, "grad_norm": 2.8271055221557617, "learning_rate": 9.370596669077936e-06, "loss": 1.3866, "step": 307 }, { "epoch": 0.2415686274509804, "grad_norm": 2.6021790504455566, "learning_rate": 9.179426158667445e-06, "loss": 1.1333, "step": 308 }, { "epoch": 0.24235294117647058, "grad_norm": 2.9245004653930664, "learning_rate": 8.989931108291197e-06, "loss": 1.0958, "step": 309 }, { "epoch": 0.24313725490196078, "grad_norm": 2.7905077934265137, "learning_rate": 8.802123814011457e-06, "loss": 1.0016, "step": 310 }, { "epoch": 0.24392156862745099, "grad_norm": 3.4618003368377686, "learning_rate": 8.616016462374414e-06, "loss": 1.2777, "step": 311 }, { "epoch": 0.2447058823529412, "grad_norm": 4.318086624145508, "learning_rate": 8.431621129619399e-06, "loss": 1.139, "step": 312 }, { "epoch": 0.24549019607843137, "grad_norm": 4.268404483795166, "learning_rate": 8.248949780895306e-06, "loss": 1.6534, "step": 313 }, { "epoch": 0.24627450980392157, "grad_norm": 3.4030003547668457, "learning_rate": 8.068014269484189e-06, "loss": 1.146, "step": 314 }, { "epoch": 0.24705882352941178, "grad_norm": 10.152856826782227, "learning_rate": 7.888826336032093e-06, "loss": 1.53, "step": 315 }, { "epoch": 0.24784313725490195, "grad_norm": 3.9469947814941406, "learning_rate": 7.71139760778722e-06, "loss": 1.47, "step": 316 }, { "epoch": 0.24862745098039216, "grad_norm": 4.744470596313477, "learning_rate": 7.535739597845489e-06, "loss": 1.4464, "step": 317 }, { "epoch": 0.24941176470588236, "grad_norm": 5.53031063079834, "learning_rate": 7.361863704403427e-06, "loss": 1.1041, "step": 318 }, { "epoch": 0.25019607843137254, "grad_norm": 4.436324596405029, "learning_rate": 7.189781210018604e-06, "loss": 1.3396, "step": 319 }, { "epoch": 0.25098039215686274, "grad_norm": 5.226994037628174, "learning_rate": 7.019503280877466e-06, "loss": 1.4027, "step": 320 }, { "epoch": 0.25176470588235295, "grad_norm": 4.095127582550049, "learning_rate": 6.851040966070809e-06, "loss": 1.0458, "step": 321 }, { "epoch": 0.25254901960784315, "grad_norm": 3.7459771633148193, "learning_rate": 6.684405196876842e-06, "loss": 0.7628, "step": 322 }, { "epoch": 0.25333333333333335, "grad_norm": 4.650934219360352, "learning_rate": 6.519606786051812e-06, "loss": 0.9564, "step": 323 }, { "epoch": 0.2541176470588235, "grad_norm": 5.133500576019287, "learning_rate": 6.356656427128452e-06, "loss": 1.198, "step": 324 }, { "epoch": 0.2549019607843137, "grad_norm": 4.67864990234375, "learning_rate": 6.195564693722027e-06, "loss": 1.1566, "step": 325 }, { "epoch": 0.2556862745098039, "grad_norm": 6.105125427246094, "learning_rate": 6.0363420388442584e-06, "loss": 1.2762, "step": 326 }, { "epoch": 0.2564705882352941, "grad_norm": 5.678111553192139, "learning_rate": 5.878998794225053e-06, "loss": 0.9718, "step": 327 }, { "epoch": 0.2572549019607843, "grad_norm": 7.232873916625977, "learning_rate": 5.723545169642064e-06, "loss": 1.1458, "step": 328 }, { "epoch": 0.2580392156862745, "grad_norm": 4.871210098266602, "learning_rate": 5.569991252258227e-06, "loss": 0.7986, "step": 329 }, { "epoch": 0.25882352941176473, "grad_norm": 4.941523551940918, "learning_rate": 5.418347005967188e-06, "loss": 0.9478, "step": 330 }, { "epoch": 0.2596078431372549, "grad_norm": 5.843031406402588, "learning_rate": 5.268622270746769e-06, "loss": 1.3708, "step": 331 }, { "epoch": 0.2603921568627451, "grad_norm": 5.538339614868164, "learning_rate": 5.120826762020479e-06, "loss": 1.1192, "step": 332 }, { "epoch": 0.2611764705882353, "grad_norm": 7.33754825592041, "learning_rate": 4.974970070027069e-06, "loss": 1.2714, "step": 333 }, { "epoch": 0.2619607843137255, "grad_norm": 5.050201892852783, "learning_rate": 4.8310616591982745e-06, "loss": 0.9628, "step": 334 }, { "epoch": 0.2627450980392157, "grad_norm": 5.330129623413086, "learning_rate": 4.689110867544645e-06, "loss": 0.9378, "step": 335 }, { "epoch": 0.2635294117647059, "grad_norm": 5.494550704956055, "learning_rate": 4.549126906049629e-06, "loss": 1.3844, "step": 336 }, { "epoch": 0.2643137254901961, "grad_norm": 7.918943405151367, "learning_rate": 4.411118858071873e-06, "loss": 1.6752, "step": 337 }, { "epoch": 0.26509803921568625, "grad_norm": 8.494551658630371, "learning_rate": 4.275095678755838e-06, "loss": 1.3365, "step": 338 }, { "epoch": 0.26588235294117646, "grad_norm": 6.204376697540283, "learning_rate": 4.141066194450712e-06, "loss": 1.2237, "step": 339 }, { "epoch": 0.26666666666666666, "grad_norm": 10.700553894042969, "learning_rate": 4.009039102137657e-06, "loss": 2.0144, "step": 340 }, { "epoch": 0.26745098039215687, "grad_norm": 8.845378875732422, "learning_rate": 3.879022968865484e-06, "loss": 2.1293, "step": 341 }, { "epoch": 0.26823529411764707, "grad_norm": 8.239388465881348, "learning_rate": 3.751026231194761e-06, "loss": 1.779, "step": 342 }, { "epoch": 0.2690196078431373, "grad_norm": 9.209283828735352, "learning_rate": 3.6250571946503677e-06, "loss": 1.6962, "step": 343 }, { "epoch": 0.2698039215686274, "grad_norm": 9.209612846374512, "learning_rate": 3.5011240331825784e-06, "loss": 1.7011, "step": 344 }, { "epoch": 0.27058823529411763, "grad_norm": 10.101373672485352, "learning_rate": 3.379234788636626e-06, "loss": 1.532, "step": 345 }, { "epoch": 0.27137254901960783, "grad_norm": 11.359880447387695, "learning_rate": 3.2593973702309106e-06, "loss": 2.0318, "step": 346 }, { "epoch": 0.27215686274509804, "grad_norm": 10.647089958190918, "learning_rate": 3.1416195540437896e-06, "loss": 1.9065, "step": 347 }, { "epoch": 0.27294117647058824, "grad_norm": 15.720868110656738, "learning_rate": 3.0259089825089655e-06, "loss": 1.8493, "step": 348 }, { "epoch": 0.27372549019607845, "grad_norm": 13.961613655090332, "learning_rate": 2.9122731639196154e-06, "loss": 2.564, "step": 349 }, { "epoch": 0.27450980392156865, "grad_norm": 12.990523338317871, "learning_rate": 2.800719471941152e-06, "loss": 1.8791, "step": 350 }, { "epoch": 0.2752941176470588, "grad_norm": 2.013960123062134, "learning_rate": 2.691255145132793e-06, "loss": 2.7899, "step": 351 }, { "epoch": 0.276078431372549, "grad_norm": 1.9573743343353271, "learning_rate": 2.5838872864778276e-06, "loss": 1.9973, "step": 352 }, { "epoch": 0.2768627450980392, "grad_norm": 2.0770833492279053, "learning_rate": 2.478622862922738e-06, "loss": 1.5119, "step": 353 }, { "epoch": 0.2776470588235294, "grad_norm": 2.2673606872558594, "learning_rate": 2.375468704925131e-06, "loss": 1.473, "step": 354 }, { "epoch": 0.2784313725490196, "grad_norm": 2.035940408706665, "learning_rate": 2.2744315060104845e-06, "loss": 1.1183, "step": 355 }, { "epoch": 0.2792156862745098, "grad_norm": 2.1854915618896484, "learning_rate": 2.175517822337849e-06, "loss": 1.1662, "step": 356 }, { "epoch": 0.28, "grad_norm": 2.2467715740203857, "learning_rate": 2.0787340722744e-06, "loss": 0.9999, "step": 357 }, { "epoch": 0.2807843137254902, "grad_norm": 2.530546188354492, "learning_rate": 1.9840865359789894e-06, "loss": 1.1862, "step": 358 }, { "epoch": 0.2815686274509804, "grad_norm": 3.1195712089538574, "learning_rate": 1.891581354994618e-06, "loss": 1.2027, "step": 359 }, { "epoch": 0.2823529411764706, "grad_norm": 3.0070858001708984, "learning_rate": 1.801224531849908e-06, "loss": 1.0958, "step": 360 }, { "epoch": 0.2831372549019608, "grad_norm": 3.387432813644409, "learning_rate": 1.7130219296696263e-06, "loss": 1.3352, "step": 361 }, { "epoch": 0.283921568627451, "grad_norm": 3.0927343368530273, "learning_rate": 1.6269792717942316e-06, "loss": 1.222, "step": 362 }, { "epoch": 0.2847058823529412, "grad_norm": 3.4728779792785645, "learning_rate": 1.5431021414084833e-06, "loss": 1.4163, "step": 363 }, { "epoch": 0.28549019607843135, "grad_norm": 3.3711791038513184, "learning_rate": 1.461395981179182e-06, "loss": 1.1301, "step": 364 }, { "epoch": 0.28627450980392155, "grad_norm": 3.495633602142334, "learning_rate": 1.3818660929019714e-06, "loss": 1.2015, "step": 365 }, { "epoch": 0.28705882352941176, "grad_norm": 4.38746452331543, "learning_rate": 1.3045176371573263e-06, "loss": 1.3248, "step": 366 }, { "epoch": 0.28784313725490196, "grad_norm": 4.259886741638184, "learning_rate": 1.2293556329757026e-06, "loss": 1.512, "step": 367 }, { "epoch": 0.28862745098039216, "grad_norm": 3.8518102169036865, "learning_rate": 1.1563849575118322e-06, "loss": 1.1679, "step": 368 }, { "epoch": 0.28941176470588237, "grad_norm": 3.8261094093322754, "learning_rate": 1.08561034572828e-06, "loss": 1.1916, "step": 369 }, { "epoch": 0.2901960784313726, "grad_norm": 5.102038383483887, "learning_rate": 1.0170363900881796e-06, "loss": 1.8872, "step": 370 }, { "epoch": 0.2909803921568627, "grad_norm": 5.184679985046387, "learning_rate": 9.506675402572445e-07, "loss": 1.4387, "step": 371 }, { "epoch": 0.2917647058823529, "grad_norm": 4.591920375823975, "learning_rate": 8.865081028150279e-07, "loss": 0.9937, "step": 372 }, { "epoch": 0.29254901960784313, "grad_norm": 5.405058860778809, "learning_rate": 8.24562240975476e-07, "loss": 1.3768, "step": 373 }, { "epoch": 0.29333333333333333, "grad_norm": 4.751702308654785, "learning_rate": 7.648339743168009e-07, "loss": 1.1257, "step": 374 }, { "epoch": 0.29411764705882354, "grad_norm": 5.307697772979736, "learning_rate": 7.073271785206313e-07, "loss": 1.3546, "step": 375 }, { "epoch": 0.29490196078431374, "grad_norm": 5.108140468597412, "learning_rate": 6.520455851205298e-07, "loss": 0.9153, "step": 376 }, { "epoch": 0.29568627450980395, "grad_norm": 4.52064847946167, "learning_rate": 5.98992781259876e-07, "loss": 0.8798, "step": 377 }, { "epoch": 0.2964705882352941, "grad_norm": 3.916222095489502, "learning_rate": 5.48172209459079e-07, "loss": 0.7, "step": 378 }, { "epoch": 0.2972549019607843, "grad_norm": 4.668052673339844, "learning_rate": 4.995871673922213e-07, "loss": 1.0176, "step": 379 }, { "epoch": 0.2980392156862745, "grad_norm": 4.989783763885498, "learning_rate": 4.532408076730504e-07, "loss": 1.0162, "step": 380 }, { "epoch": 0.2988235294117647, "grad_norm": 3.6178388595581055, "learning_rate": 4.091361376504382e-07, "loss": 0.746, "step": 381 }, { "epoch": 0.2996078431372549, "grad_norm": 4.038928985595703, "learning_rate": 3.672760192132107e-07, "loss": 0.8829, "step": 382 }, { "epoch": 0.3003921568627451, "grad_norm": 5.406976699829102, "learning_rate": 3.276631686044734e-07, "loss": 1.3866, "step": 383 }, { "epoch": 0.30117647058823527, "grad_norm": 4.995997905731201, "learning_rate": 2.903001562453289e-07, "loss": 0.8837, "step": 384 }, { "epoch": 0.30196078431372547, "grad_norm": 6.531996250152588, "learning_rate": 2.5518940656811094e-07, "loss": 1.3823, "step": 385 }, { "epoch": 0.3027450980392157, "grad_norm": 6.122218132019043, "learning_rate": 2.2233319785904537e-07, "loss": 1.251, "step": 386 }, { "epoch": 0.3035294117647059, "grad_norm": 6.061277389526367, "learning_rate": 1.9173366211043484e-07, "loss": 1.4177, "step": 387 }, { "epoch": 0.3043137254901961, "grad_norm": 5.348453521728516, "learning_rate": 1.6339278488229802e-07, "loss": 1.3636, "step": 388 }, { "epoch": 0.3050980392156863, "grad_norm": 5.866225719451904, "learning_rate": 1.3731240517354802e-07, "loss": 1.1487, "step": 389 }, { "epoch": 0.3058823529411765, "grad_norm": 5.9875688552856445, "learning_rate": 1.1349421530265245e-07, "loss": 1.3415, "step": 390 }, { "epoch": 0.30666666666666664, "grad_norm": 7.915737628936768, "learning_rate": 9.19397607978134e-08, "loss": 1.225, "step": 391 }, { "epoch": 0.30745098039215685, "grad_norm": 8.340410232543945, "learning_rate": 7.265044029669486e-08, "loss": 1.9622, "step": 392 }, { "epoch": 0.30823529411764705, "grad_norm": 10.420660018920898, "learning_rate": 5.562750545566652e-08, "loss": 1.9871, "step": 393 }, { "epoch": 0.30901960784313726, "grad_norm": 9.027931213378906, "learning_rate": 4.087206086856754e-08, "loss": 1.585, "step": 394 }, { "epoch": 0.30980392156862746, "grad_norm": 10.239818572998047, "learning_rate": 2.838506399506446e-08, "loss": 1.9573, "step": 395 }, { "epoch": 0.31058823529411766, "grad_norm": 7.999627590179443, "learning_rate": 1.8167325098478735e-08, "loss": 1.6673, "step": 396 }, { "epoch": 0.31137254901960787, "grad_norm": 9.10187816619873, "learning_rate": 1.0219507193247045e-08, "loss": 1.6721, "step": 397 }, { "epoch": 0.312156862745098, "grad_norm": 13.277602195739746, "learning_rate": 4.542126001886303e-09, "loss": 3.0342, "step": 398 }, { "epoch": 0.3129411764705882, "grad_norm": 13.699840545654297, "learning_rate": 1.1355499215215391e-09, "loss": 1.7288, "step": 399 }, { "epoch": 0.3137254901960784, "grad_norm": 11.232860565185547, "learning_rate": 0.0, "loss": 2.2469, "step": 400 }, { "epoch": 0.3137254901960784, "eval_loss": 0.3466954827308655, "eval_runtime": 169.3978, "eval_samples_per_second": 12.68, "eval_steps_per_second": 3.17, "step": 400 } ], "logging_steps": 1, "max_steps": 400, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 100, "stateful_callbacks": { "EarlyStoppingCallback": { "args": { "early_stopping_patience": 4, "early_stopping_threshold": 0.0 }, "attributes": { "early_stopping_patience_counter": 0 } }, "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 6.053128390909624e+17, "train_batch_size": 8, "trial_name": null, "trial_params": null }