{ "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 978, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.003067484662576687, "grad_norm": 0.7997869645290624, "learning_rate": 1.0204081632653061e-07, "loss": 0.0303, "step": 1 }, { "epoch": 0.006134969325153374, "grad_norm": 1.3916303874119316, "learning_rate": 2.0408163265306121e-07, "loss": 0.0855, "step": 2 }, { "epoch": 0.009202453987730062, "grad_norm": 0.9776667246072426, "learning_rate": 3.0612244897959183e-07, "loss": 0.0244, "step": 3 }, { "epoch": 0.012269938650306749, "grad_norm": 0.9337405630277921, "learning_rate": 4.0816326530612243e-07, "loss": 0.0266, "step": 4 }, { "epoch": 0.015337423312883436, "grad_norm": 0.6813063564646261, "learning_rate": 5.102040816326531e-07, "loss": 0.0138, "step": 5 }, { "epoch": 0.018404907975460124, "grad_norm": 0.7694825948696452, "learning_rate": 6.122448979591837e-07, "loss": 0.0277, "step": 6 }, { "epoch": 0.02147239263803681, "grad_norm": 1.019918445629934, "learning_rate": 7.142857142857143e-07, "loss": 0.0431, "step": 7 }, { "epoch": 0.024539877300613498, "grad_norm": 0.8356666628019843, "learning_rate": 8.163265306122449e-07, "loss": 0.0376, "step": 8 }, { "epoch": 0.027607361963190184, "grad_norm": 0.9661685002153698, "learning_rate": 9.183673469387756e-07, "loss": 0.0252, "step": 9 }, { "epoch": 0.03067484662576687, "grad_norm": 1.0057781682964735, "learning_rate": 1.0204081632653063e-06, "loss": 0.0551, "step": 10 }, { "epoch": 0.03374233128834356, "grad_norm": 1.27237502100699, "learning_rate": 1.122448979591837e-06, "loss": 0.0616, "step": 11 }, { "epoch": 0.03680981595092025, "grad_norm": 0.8635209118844634, "learning_rate": 1.2244897959183673e-06, "loss": 0.0356, "step": 12 }, { "epoch": 0.03987730061349693, "grad_norm": 1.0917475029682462, "learning_rate": 1.3265306122448982e-06, "loss": 0.0517, "step": 13 }, { "epoch": 0.04294478527607362, "grad_norm": 0.9971259570435305, "learning_rate": 1.4285714285714286e-06, "loss": 0.0388, "step": 14 }, { "epoch": 0.046012269938650305, "grad_norm": 1.18589108762414, "learning_rate": 1.5306122448979593e-06, "loss": 0.0411, "step": 15 }, { "epoch": 0.049079754601226995, "grad_norm": 0.7687072641757458, "learning_rate": 1.6326530612244897e-06, "loss": 0.0298, "step": 16 }, { "epoch": 0.05214723926380368, "grad_norm": 0.7583676893025098, "learning_rate": 1.7346938775510206e-06, "loss": 0.0341, "step": 17 }, { "epoch": 0.05521472392638037, "grad_norm": 1.1470980781176612, "learning_rate": 1.8367346938775512e-06, "loss": 0.0382, "step": 18 }, { "epoch": 0.05828220858895705, "grad_norm": 1.0121900682885199, "learning_rate": 1.938775510204082e-06, "loss": 0.038, "step": 19 }, { "epoch": 0.06134969325153374, "grad_norm": 1.0122959695964604, "learning_rate": 2.0408163265306125e-06, "loss": 0.0312, "step": 20 }, { "epoch": 0.06441717791411043, "grad_norm": 0.8672537023810776, "learning_rate": 2.1428571428571427e-06, "loss": 0.0343, "step": 21 }, { "epoch": 0.06748466257668712, "grad_norm": 0.8464502422934749, "learning_rate": 2.244897959183674e-06, "loss": 0.0322, "step": 22 }, { "epoch": 0.0705521472392638, "grad_norm": 0.9014580063188392, "learning_rate": 2.3469387755102044e-06, "loss": 0.0517, "step": 23 }, { "epoch": 0.0736196319018405, "grad_norm": 0.644281280221218, "learning_rate": 2.4489795918367347e-06, "loss": 0.0196, "step": 24 }, { "epoch": 0.07668711656441718, "grad_norm": 1.153639389526131, "learning_rate": 2.5510204081632657e-06, "loss": 0.0575, "step": 25 }, { "epoch": 0.07975460122699386, "grad_norm": 0.5976101616752824, "learning_rate": 2.6530612244897964e-06, "loss": 0.0229, "step": 26 }, { "epoch": 0.08282208588957055, "grad_norm": 0.6774405227259789, "learning_rate": 2.7551020408163266e-06, "loss": 0.0269, "step": 27 }, { "epoch": 0.08588957055214724, "grad_norm": 0.8858054392922697, "learning_rate": 2.8571428571428573e-06, "loss": 0.0346, "step": 28 }, { "epoch": 0.08895705521472393, "grad_norm": 0.6961058532345435, "learning_rate": 2.959183673469388e-06, "loss": 0.0439, "step": 29 }, { "epoch": 0.09202453987730061, "grad_norm": 0.5514138971509586, "learning_rate": 3.0612244897959185e-06, "loss": 0.0142, "step": 30 }, { "epoch": 0.0950920245398773, "grad_norm": 0.9140038253698272, "learning_rate": 3.1632653061224496e-06, "loss": 0.0314, "step": 31 }, { "epoch": 0.09815950920245399, "grad_norm": 0.7760922802051613, "learning_rate": 3.2653061224489794e-06, "loss": 0.0268, "step": 32 }, { "epoch": 0.10122699386503067, "grad_norm": 0.6792779586916694, "learning_rate": 3.3673469387755105e-06, "loss": 0.0271, "step": 33 }, { "epoch": 0.10429447852760736, "grad_norm": 0.6969367770150492, "learning_rate": 3.469387755102041e-06, "loss": 0.028, "step": 34 }, { "epoch": 0.10736196319018405, "grad_norm": 0.86611814489453, "learning_rate": 3.5714285714285718e-06, "loss": 0.0388, "step": 35 }, { "epoch": 0.11042944785276074, "grad_norm": 0.6484808223290501, "learning_rate": 3.6734693877551024e-06, "loss": 0.0335, "step": 36 }, { "epoch": 0.11349693251533742, "grad_norm": 0.5965376825578463, "learning_rate": 3.7755102040816327e-06, "loss": 0.0252, "step": 37 }, { "epoch": 0.1165644171779141, "grad_norm": 0.6268545591800431, "learning_rate": 3.877551020408164e-06, "loss": 0.0142, "step": 38 }, { "epoch": 0.1196319018404908, "grad_norm": 0.6668195357079019, "learning_rate": 3.979591836734694e-06, "loss": 0.0217, "step": 39 }, { "epoch": 0.12269938650306748, "grad_norm": 0.5086913557218375, "learning_rate": 4.081632653061225e-06, "loss": 0.0253, "step": 40 }, { "epoch": 0.12576687116564417, "grad_norm": 0.730907605009217, "learning_rate": 4.183673469387755e-06, "loss": 0.0228, "step": 41 }, { "epoch": 0.12883435582822086, "grad_norm": 0.7102463889462425, "learning_rate": 4.2857142857142855e-06, "loss": 0.0293, "step": 42 }, { "epoch": 0.13190184049079753, "grad_norm": 0.4909211588758295, "learning_rate": 4.3877551020408165e-06, "loss": 0.017, "step": 43 }, { "epoch": 0.13496932515337423, "grad_norm": 0.32680440879911227, "learning_rate": 4.489795918367348e-06, "loss": 0.0161, "step": 44 }, { "epoch": 0.13803680981595093, "grad_norm": 0.4833196352907147, "learning_rate": 4.591836734693878e-06, "loss": 0.0353, "step": 45 }, { "epoch": 0.1411042944785276, "grad_norm": 0.36928803194700305, "learning_rate": 4.693877551020409e-06, "loss": 0.02, "step": 46 }, { "epoch": 0.1441717791411043, "grad_norm": 0.4472383224420318, "learning_rate": 4.795918367346939e-06, "loss": 0.0131, "step": 47 }, { "epoch": 0.147239263803681, "grad_norm": 0.36087285993792934, "learning_rate": 4.897959183673469e-06, "loss": 0.0159, "step": 48 }, { "epoch": 0.15030674846625766, "grad_norm": 0.5556656554696969, "learning_rate": 5e-06, "loss": 0.0268, "step": 49 }, { "epoch": 0.15337423312883436, "grad_norm": 0.8497844272956676, "learning_rate": 5.1020408163265315e-06, "loss": 0.0454, "step": 50 }, { "epoch": 0.15644171779141106, "grad_norm": 0.4536793707964894, "learning_rate": 5.204081632653062e-06, "loss": 0.0201, "step": 51 }, { "epoch": 0.15950920245398773, "grad_norm": 0.7743465371861137, "learning_rate": 5.306122448979593e-06, "loss": 0.0386, "step": 52 }, { "epoch": 0.16257668711656442, "grad_norm": 0.49482528494719774, "learning_rate": 5.408163265306123e-06, "loss": 0.0209, "step": 53 }, { "epoch": 0.1656441717791411, "grad_norm": 0.43390985077433325, "learning_rate": 5.510204081632653e-06, "loss": 0.0154, "step": 54 }, { "epoch": 0.1687116564417178, "grad_norm": 0.19867010422132667, "learning_rate": 5.6122448979591834e-06, "loss": 0.0063, "step": 55 }, { "epoch": 0.17177914110429449, "grad_norm": 0.42254705168474815, "learning_rate": 5.7142857142857145e-06, "loss": 0.0156, "step": 56 }, { "epoch": 0.17484662576687116, "grad_norm": 0.483259025167169, "learning_rate": 5.816326530612246e-06, "loss": 0.0234, "step": 57 }, { "epoch": 0.17791411042944785, "grad_norm": 0.48205583406482455, "learning_rate": 5.918367346938776e-06, "loss": 0.0181, "step": 58 }, { "epoch": 0.18098159509202455, "grad_norm": 0.5452087947152122, "learning_rate": 6.020408163265307e-06, "loss": 0.0239, "step": 59 }, { "epoch": 0.18404907975460122, "grad_norm": 0.47352939498338353, "learning_rate": 6.122448979591837e-06, "loss": 0.0207, "step": 60 }, { "epoch": 0.18711656441717792, "grad_norm": 0.5722083241556664, "learning_rate": 6.224489795918368e-06, "loss": 0.0346, "step": 61 }, { "epoch": 0.1901840490797546, "grad_norm": 0.35217938893247586, "learning_rate": 6.326530612244899e-06, "loss": 0.0145, "step": 62 }, { "epoch": 0.19325153374233128, "grad_norm": 0.4140072480123712, "learning_rate": 6.4285714285714295e-06, "loss": 0.0152, "step": 63 }, { "epoch": 0.19631901840490798, "grad_norm": 0.49720539052583107, "learning_rate": 6.530612244897959e-06, "loss": 0.0296, "step": 64 }, { "epoch": 0.19938650306748465, "grad_norm": 0.469616800045718, "learning_rate": 6.63265306122449e-06, "loss": 0.0174, "step": 65 }, { "epoch": 0.20245398773006135, "grad_norm": 0.618218528026767, "learning_rate": 6.734693877551021e-06, "loss": 0.0344, "step": 66 }, { "epoch": 0.20552147239263804, "grad_norm": 0.7394292500503313, "learning_rate": 6.836734693877551e-06, "loss": 0.0324, "step": 67 }, { "epoch": 0.2085889570552147, "grad_norm": 0.5210994325814557, "learning_rate": 6.938775510204082e-06, "loss": 0.0228, "step": 68 }, { "epoch": 0.2116564417177914, "grad_norm": 0.4487340952497001, "learning_rate": 7.0408163265306125e-06, "loss": 0.0227, "step": 69 }, { "epoch": 0.2147239263803681, "grad_norm": 0.49919865989262474, "learning_rate": 7.1428571428571436e-06, "loss": 0.0149, "step": 70 }, { "epoch": 0.21779141104294478, "grad_norm": 0.42206279707783706, "learning_rate": 7.244897959183675e-06, "loss": 0.0151, "step": 71 }, { "epoch": 0.22085889570552147, "grad_norm": 0.31366279555645177, "learning_rate": 7.346938775510205e-06, "loss": 0.0094, "step": 72 }, { "epoch": 0.22392638036809817, "grad_norm": 0.5571009715110524, "learning_rate": 7.448979591836736e-06, "loss": 0.0259, "step": 73 }, { "epoch": 0.22699386503067484, "grad_norm": 0.541523890264921, "learning_rate": 7.551020408163265e-06, "loss": 0.0157, "step": 74 }, { "epoch": 0.23006134969325154, "grad_norm": 0.7290444815058128, "learning_rate": 7.653061224489796e-06, "loss": 0.0417, "step": 75 }, { "epoch": 0.2331288343558282, "grad_norm": 0.36222020238590813, "learning_rate": 7.755102040816327e-06, "loss": 0.0162, "step": 76 }, { "epoch": 0.2361963190184049, "grad_norm": 0.6725897974035238, "learning_rate": 7.857142857142858e-06, "loss": 0.0367, "step": 77 }, { "epoch": 0.2392638036809816, "grad_norm": 0.47377840842133173, "learning_rate": 7.959183673469388e-06, "loss": 0.0194, "step": 78 }, { "epoch": 0.24233128834355827, "grad_norm": 0.6367586297561662, "learning_rate": 8.06122448979592e-06, "loss": 0.0367, "step": 79 }, { "epoch": 0.24539877300613497, "grad_norm": 0.5733239067596057, "learning_rate": 8.16326530612245e-06, "loss": 0.0266, "step": 80 }, { "epoch": 0.24846625766871167, "grad_norm": 0.3752422775055063, "learning_rate": 8.26530612244898e-06, "loss": 0.0165, "step": 81 }, { "epoch": 0.25153374233128833, "grad_norm": 0.8041178386540389, "learning_rate": 8.36734693877551e-06, "loss": 0.0482, "step": 82 }, { "epoch": 0.254601226993865, "grad_norm": 0.48929300164772, "learning_rate": 8.469387755102042e-06, "loss": 0.0242, "step": 83 }, { "epoch": 0.25766871165644173, "grad_norm": 0.38718573399562517, "learning_rate": 8.571428571428571e-06, "loss": 0.0213, "step": 84 }, { "epoch": 0.2607361963190184, "grad_norm": 0.4376639244558999, "learning_rate": 8.673469387755103e-06, "loss": 0.019, "step": 85 }, { "epoch": 0.26380368098159507, "grad_norm": 0.5268233065102302, "learning_rate": 8.775510204081633e-06, "loss": 0.0247, "step": 86 }, { "epoch": 0.2668711656441718, "grad_norm": 0.7284585489566475, "learning_rate": 8.877551020408163e-06, "loss": 0.0498, "step": 87 }, { "epoch": 0.26993865030674846, "grad_norm": 0.548627684829585, "learning_rate": 8.979591836734695e-06, "loss": 0.0399, "step": 88 }, { "epoch": 0.27300613496932513, "grad_norm": 0.49417006334168573, "learning_rate": 9.081632653061225e-06, "loss": 0.0241, "step": 89 }, { "epoch": 0.27607361963190186, "grad_norm": 0.35327219241848984, "learning_rate": 9.183673469387756e-06, "loss": 0.012, "step": 90 }, { "epoch": 0.2791411042944785, "grad_norm": 0.5651567661019409, "learning_rate": 9.285714285714288e-06, "loss": 0.0306, "step": 91 }, { "epoch": 0.2822085889570552, "grad_norm": 0.4977883329448499, "learning_rate": 9.387755102040818e-06, "loss": 0.0331, "step": 92 }, { "epoch": 0.2852760736196319, "grad_norm": 0.4742291191076062, "learning_rate": 9.489795918367348e-06, "loss": 0.0264, "step": 93 }, { "epoch": 0.2883435582822086, "grad_norm": 0.4979060862617067, "learning_rate": 9.591836734693878e-06, "loss": 0.0196, "step": 94 }, { "epoch": 0.29141104294478526, "grad_norm": 0.4618772300579882, "learning_rate": 9.693877551020408e-06, "loss": 0.0312, "step": 95 }, { "epoch": 0.294478527607362, "grad_norm": 0.5244902559593818, "learning_rate": 9.795918367346939e-06, "loss": 0.0135, "step": 96 }, { "epoch": 0.29754601226993865, "grad_norm": 0.42514924793603937, "learning_rate": 9.89795918367347e-06, "loss": 0.0134, "step": 97 }, { "epoch": 0.3006134969325153, "grad_norm": 0.6303178022610803, "learning_rate": 1e-05, "loss": 0.0431, "step": 98 }, { "epoch": 0.30368098159509205, "grad_norm": 0.43154235410273956, "learning_rate": 9.999968137932856e-06, "loss": 0.0256, "step": 99 }, { "epoch": 0.3067484662576687, "grad_norm": 0.3913466056843226, "learning_rate": 9.999872552137497e-06, "loss": 0.0209, "step": 100 }, { "epoch": 0.3098159509202454, "grad_norm": 0.3803054399869341, "learning_rate": 9.999713243832148e-06, "loss": 0.0176, "step": 101 }, { "epoch": 0.3128834355828221, "grad_norm": 0.560817485298493, "learning_rate": 9.999490215047167e-06, "loss": 0.034, "step": 102 }, { "epoch": 0.3159509202453988, "grad_norm": 0.5606212067152239, "learning_rate": 9.999203468625017e-06, "loss": 0.0223, "step": 103 }, { "epoch": 0.31901840490797545, "grad_norm": 0.3078230944385983, "learning_rate": 9.998853008220231e-06, "loss": 0.0102, "step": 104 }, { "epoch": 0.3220858895705521, "grad_norm": 0.2936441456513863, "learning_rate": 9.998438838299366e-06, "loss": 0.0131, "step": 105 }, { "epoch": 0.32515337423312884, "grad_norm": 0.545125785404182, "learning_rate": 9.997960964140946e-06, "loss": 0.0225, "step": 106 }, { "epoch": 0.3282208588957055, "grad_norm": 0.443159450605464, "learning_rate": 9.997419391835397e-06, "loss": 0.014, "step": 107 }, { "epoch": 0.3312883435582822, "grad_norm": 0.5171297711383881, "learning_rate": 9.99681412828496e-06, "loss": 0.0336, "step": 108 }, { "epoch": 0.3343558282208589, "grad_norm": 0.3541766364039508, "learning_rate": 9.996145181203616e-06, "loss": 0.0155, "step": 109 }, { "epoch": 0.3374233128834356, "grad_norm": 0.5566728911251606, "learning_rate": 9.995412559116979e-06, "loss": 0.0184, "step": 110 }, { "epoch": 0.34049079754601225, "grad_norm": 0.43277308181190083, "learning_rate": 9.994616271362193e-06, "loss": 0.0217, "step": 111 }, { "epoch": 0.34355828220858897, "grad_norm": 1.0421460915519953, "learning_rate": 9.993756328087806e-06, "loss": 0.0997, "step": 112 }, { "epoch": 0.34662576687116564, "grad_norm": 0.5075832442140273, "learning_rate": 9.992832740253646e-06, "loss": 0.0168, "step": 113 }, { "epoch": 0.3496932515337423, "grad_norm": 0.3720333544524573, "learning_rate": 9.991845519630679e-06, "loss": 0.0145, "step": 114 }, { "epoch": 0.35276073619631904, "grad_norm": 0.3602385407355057, "learning_rate": 9.990794678800863e-06, "loss": 0.0154, "step": 115 }, { "epoch": 0.3558282208588957, "grad_norm": 0.4542869951307087, "learning_rate": 9.989680231156983e-06, "loss": 0.026, "step": 116 }, { "epoch": 0.3588957055214724, "grad_norm": 0.4351733531466369, "learning_rate": 9.988502190902478e-06, "loss": 0.0279, "step": 117 }, { "epoch": 0.3619631901840491, "grad_norm": 0.46867639008666345, "learning_rate": 9.987260573051268e-06, "loss": 0.0216, "step": 118 }, { "epoch": 0.36503067484662577, "grad_norm": 0.4387734488332606, "learning_rate": 9.98595539342756e-06, "loss": 0.0233, "step": 119 }, { "epoch": 0.36809815950920244, "grad_norm": 0.539494122809603, "learning_rate": 9.984586668665641e-06, "loss": 0.02, "step": 120 }, { "epoch": 0.37116564417177916, "grad_norm": 0.5140548435409961, "learning_rate": 9.98315441620967e-06, "loss": 0.0315, "step": 121 }, { "epoch": 0.37423312883435583, "grad_norm": 0.5589763772928042, "learning_rate": 9.981658654313458e-06, "loss": 0.0303, "step": 122 }, { "epoch": 0.3773006134969325, "grad_norm": 0.5761892222444737, "learning_rate": 9.980099402040231e-06, "loss": 0.0378, "step": 123 }, { "epoch": 0.3803680981595092, "grad_norm": 0.6156623359330066, "learning_rate": 9.978476679262387e-06, "loss": 0.0384, "step": 124 }, { "epoch": 0.3834355828220859, "grad_norm": 0.551665717386314, "learning_rate": 9.976790506661252e-06, "loss": 0.0274, "step": 125 }, { "epoch": 0.38650306748466257, "grad_norm": 0.5955360132547092, "learning_rate": 9.975040905726799e-06, "loss": 0.0231, "step": 126 }, { "epoch": 0.3895705521472393, "grad_norm": 0.44370422954795896, "learning_rate": 9.97322789875739e-06, "loss": 0.0223, "step": 127 }, { "epoch": 0.39263803680981596, "grad_norm": 0.5206926401483173, "learning_rate": 9.971351508859488e-06, "loss": 0.0254, "step": 128 }, { "epoch": 0.39570552147239263, "grad_norm": 0.6756035296339351, "learning_rate": 9.969411759947354e-06, "loss": 0.1297, "step": 129 }, { "epoch": 0.3987730061349693, "grad_norm": 0.42001192691622796, "learning_rate": 9.96740867674275e-06, "loss": 0.0198, "step": 130 }, { "epoch": 0.401840490797546, "grad_norm": 0.5471786768275381, "learning_rate": 9.965342284774633e-06, "loss": 0.0407, "step": 131 }, { "epoch": 0.4049079754601227, "grad_norm": 0.46505959754331605, "learning_rate": 9.963212610378804e-06, "loss": 0.0203, "step": 132 }, { "epoch": 0.40797546012269936, "grad_norm": 0.5716077696919392, "learning_rate": 9.961019680697593e-06, "loss": 0.0241, "step": 133 }, { "epoch": 0.4110429447852761, "grad_norm": 0.49711227219120874, "learning_rate": 9.958763523679515e-06, "loss": 0.0173, "step": 134 }, { "epoch": 0.41411042944785276, "grad_norm": 0.47418803366323253, "learning_rate": 9.956444168078896e-06, "loss": 0.0191, "step": 135 }, { "epoch": 0.4171779141104294, "grad_norm": 0.4847709417484094, "learning_rate": 9.954061643455524e-06, "loss": 0.0262, "step": 136 }, { "epoch": 0.42024539877300615, "grad_norm": 0.49885884477945064, "learning_rate": 9.951615980174262e-06, "loss": 0.0354, "step": 137 }, { "epoch": 0.4233128834355828, "grad_norm": 0.5163880173448544, "learning_rate": 9.949107209404664e-06, "loss": 0.026, "step": 138 }, { "epoch": 0.4263803680981595, "grad_norm": 0.6209064580342998, "learning_rate": 9.946535363120581e-06, "loss": 0.0373, "step": 139 }, { "epoch": 0.4294478527607362, "grad_norm": 0.5706790350480078, "learning_rate": 9.943900474099748e-06, "loss": 0.0193, "step": 140 }, { "epoch": 0.4325153374233129, "grad_norm": 0.6595953727064474, "learning_rate": 9.94120257592337e-06, "loss": 0.0299, "step": 141 }, { "epoch": 0.43558282208588955, "grad_norm": 0.5292023247761327, "learning_rate": 9.938441702975689e-06, "loss": 0.0514, "step": 142 }, { "epoch": 0.4386503067484663, "grad_norm": 0.460766980739092, "learning_rate": 9.935617890443557e-06, "loss": 0.0152, "step": 143 }, { "epoch": 0.44171779141104295, "grad_norm": 0.4778170956102627, "learning_rate": 9.932731174315973e-06, "loss": 0.0219, "step": 144 }, { "epoch": 0.4447852760736196, "grad_norm": 0.468431790732579, "learning_rate": 9.929781591383634e-06, "loss": 0.0199, "step": 145 }, { "epoch": 0.44785276073619634, "grad_norm": 0.4136064992794228, "learning_rate": 9.926769179238467e-06, "loss": 0.0287, "step": 146 }, { "epoch": 0.450920245398773, "grad_norm": 0.48063176334483343, "learning_rate": 9.923693976273138e-06, "loss": 0.0253, "step": 147 }, { "epoch": 0.4539877300613497, "grad_norm": 0.4390033441811796, "learning_rate": 9.92055602168058e-06, "loss": 0.0228, "step": 148 }, { "epoch": 0.4570552147239264, "grad_norm": 0.6077680223170788, "learning_rate": 9.917355355453481e-06, "loss": 0.044, "step": 149 }, { "epoch": 0.4601226993865031, "grad_norm": 0.7649894761219165, "learning_rate": 9.914092018383778e-06, "loss": 0.0527, "step": 150 }, { "epoch": 0.46319018404907975, "grad_norm": 0.5196433338574888, "learning_rate": 9.910766052062137e-06, "loss": 0.0389, "step": 151 }, { "epoch": 0.4662576687116564, "grad_norm": 0.5592360599484748, "learning_rate": 9.90737749887742e-06, "loss": 0.0379, "step": 152 }, { "epoch": 0.46932515337423314, "grad_norm": 0.6311825130544745, "learning_rate": 9.903926402016153e-06, "loss": 0.0328, "step": 153 }, { "epoch": 0.4723926380368098, "grad_norm": 0.6765533980399042, "learning_rate": 9.900412805461968e-06, "loss": 0.0274, "step": 154 }, { "epoch": 0.4754601226993865, "grad_norm": 0.4182044283444094, "learning_rate": 9.896836753995043e-06, "loss": 0.0174, "step": 155 }, { "epoch": 0.4785276073619632, "grad_norm": 0.4494240910276348, "learning_rate": 9.893198293191539e-06, "loss": 0.0203, "step": 156 }, { "epoch": 0.4815950920245399, "grad_norm": 0.7211159726447208, "learning_rate": 9.889497469423004e-06, "loss": 0.0543, "step": 157 }, { "epoch": 0.48466257668711654, "grad_norm": 0.47130058401261765, "learning_rate": 9.885734329855798e-06, "loss": 0.016, "step": 158 }, { "epoch": 0.48773006134969327, "grad_norm": 0.523537473814604, "learning_rate": 9.881908922450485e-06, "loss": 0.0421, "step": 159 }, { "epoch": 0.49079754601226994, "grad_norm": 0.5459938744286107, "learning_rate": 9.878021295961218e-06, "loss": 0.032, "step": 160 }, { "epoch": 0.4938650306748466, "grad_norm": 0.565898967516583, "learning_rate": 9.874071499935124e-06, "loss": 0.0414, "step": 161 }, { "epoch": 0.49693251533742333, "grad_norm": 0.48527683019849255, "learning_rate": 9.870059584711668e-06, "loss": 0.0352, "step": 162 }, { "epoch": 0.5, "grad_norm": 0.49346103186322066, "learning_rate": 9.865985601422018e-06, "loss": 0.0189, "step": 163 }, { "epoch": 0.5030674846625767, "grad_norm": 0.4121625641879102, "learning_rate": 9.861849601988384e-06, "loss": 0.0213, "step": 164 }, { "epoch": 0.5061349693251533, "grad_norm": 0.4158207874245154, "learning_rate": 9.857651639123362e-06, "loss": 0.0186, "step": 165 }, { "epoch": 0.50920245398773, "grad_norm": 0.3274693773098448, "learning_rate": 9.853391766329264e-06, "loss": 0.0151, "step": 166 }, { "epoch": 0.5122699386503068, "grad_norm": 0.4958803903962215, "learning_rate": 9.849070037897428e-06, "loss": 0.0284, "step": 167 }, { "epoch": 0.5153374233128835, "grad_norm": 0.6475651680085904, "learning_rate": 9.844686508907538e-06, "loss": 0.0425, "step": 168 }, { "epoch": 0.5184049079754601, "grad_norm": 0.5682638967858453, "learning_rate": 9.840241235226909e-06, "loss": 0.0266, "step": 169 }, { "epoch": 0.5214723926380368, "grad_norm": 0.5740486936349286, "learning_rate": 9.835734273509787e-06, "loss": 0.0482, "step": 170 }, { "epoch": 0.5245398773006135, "grad_norm": 0.43454823069684023, "learning_rate": 9.831165681196618e-06, "loss": 0.02, "step": 171 }, { "epoch": 0.5276073619631901, "grad_norm": 0.5008900995831749, "learning_rate": 9.826535516513318e-06, "loss": 0.0345, "step": 172 }, { "epoch": 0.5306748466257669, "grad_norm": 0.6963240546508572, "learning_rate": 9.821843838470536e-06, "loss": 0.025, "step": 173 }, { "epoch": 0.5337423312883436, "grad_norm": 0.4675708697803615, "learning_rate": 9.817090706862895e-06, "loss": 0.0223, "step": 174 }, { "epoch": 0.5368098159509203, "grad_norm": 0.3991135320601108, "learning_rate": 9.812276182268236e-06, "loss": 0.0192, "step": 175 }, { "epoch": 0.5398773006134969, "grad_norm": 0.6264372022418258, "learning_rate": 9.807400326046843e-06, "loss": 0.0407, "step": 176 }, { "epoch": 0.5429447852760736, "grad_norm": 0.5713577007790391, "learning_rate": 9.802463200340655e-06, "loss": 0.0263, "step": 177 }, { "epoch": 0.5460122699386503, "grad_norm": 0.5221974991626641, "learning_rate": 9.797464868072489e-06, "loss": 0.0257, "step": 178 }, { "epoch": 0.549079754601227, "grad_norm": 0.6380655188875212, "learning_rate": 9.792405392945218e-06, "loss": 0.0446, "step": 179 }, { "epoch": 0.5521472392638037, "grad_norm": 0.37992384347324565, "learning_rate": 9.787284839440982e-06, "loss": 0.0167, "step": 180 }, { "epoch": 0.5552147239263804, "grad_norm": 0.43784118394068366, "learning_rate": 9.78210327282035e-06, "loss": 0.024, "step": 181 }, { "epoch": 0.558282208588957, "grad_norm": 0.6888326299119037, "learning_rate": 9.776860759121485e-06, "loss": 0.0382, "step": 182 }, { "epoch": 0.5613496932515337, "grad_norm": 0.713129424822473, "learning_rate": 9.77155736515932e-06, "loss": 0.0551, "step": 183 }, { "epoch": 0.5644171779141104, "grad_norm": 0.5242216640303207, "learning_rate": 9.766193158524693e-06, "loss": 0.0279, "step": 184 }, { "epoch": 0.5674846625766872, "grad_norm": 0.5021676250184067, "learning_rate": 9.76076820758349e-06, "loss": 0.0199, "step": 185 }, { "epoch": 0.5705521472392638, "grad_norm": 0.53833739733463, "learning_rate": 9.755282581475769e-06, "loss": 0.0306, "step": 186 }, { "epoch": 0.5736196319018405, "grad_norm": 0.36310921117767025, "learning_rate": 9.749736350114886e-06, "loss": 0.013, "step": 187 }, { "epoch": 0.5766871165644172, "grad_norm": 0.612018294089201, "learning_rate": 9.744129584186599e-06, "loss": 0.0302, "step": 188 }, { "epoch": 0.5797546012269938, "grad_norm": 0.6183883573878961, "learning_rate": 9.73846235514817e-06, "loss": 0.0373, "step": 189 }, { "epoch": 0.5828220858895705, "grad_norm": 0.5335161630334854, "learning_rate": 9.73273473522745e-06, "loss": 0.0406, "step": 190 }, { "epoch": 0.5858895705521472, "grad_norm": 0.43028825452090924, "learning_rate": 9.726946797421966e-06, "loss": 0.0185, "step": 191 }, { "epoch": 0.588957055214724, "grad_norm": 0.3607144910335978, "learning_rate": 9.72109861549798e-06, "loss": 0.019, "step": 192 }, { "epoch": 0.5920245398773006, "grad_norm": 0.6082644030068691, "learning_rate": 9.715190263989562e-06, "loss": 0.0679, "step": 193 }, { "epoch": 0.5950920245398773, "grad_norm": 0.6346813811659229, "learning_rate": 9.709221818197626e-06, "loss": 0.0352, "step": 194 }, { "epoch": 0.598159509202454, "grad_norm": 0.7228677414909738, "learning_rate": 9.703193354188978e-06, "loss": 0.0489, "step": 195 }, { "epoch": 0.6012269938650306, "grad_norm": 0.6070162862574225, "learning_rate": 9.697104948795353e-06, "loss": 0.0292, "step": 196 }, { "epoch": 0.6042944785276073, "grad_norm": 0.6333886635907163, "learning_rate": 9.690956679612422e-06, "loss": 0.0347, "step": 197 }, { "epoch": 0.6073619631901841, "grad_norm": 0.5604386523878644, "learning_rate": 9.68474862499881e-06, "loss": 0.1063, "step": 198 }, { "epoch": 0.6104294478527608, "grad_norm": 0.4877317760711139, "learning_rate": 9.678480864075099e-06, "loss": 0.0351, "step": 199 }, { "epoch": 0.6134969325153374, "grad_norm": 0.73365826110845, "learning_rate": 9.672153476722816e-06, "loss": 0.0509, "step": 200 }, { "epoch": 0.6165644171779141, "grad_norm": 0.4167148427493197, "learning_rate": 9.665766543583418e-06, "loss": 0.0301, "step": 201 }, { "epoch": 0.6196319018404908, "grad_norm": 0.37456090728363056, "learning_rate": 9.659320146057263e-06, "loss": 0.0137, "step": 202 }, { "epoch": 0.6226993865030674, "grad_norm": 0.6047651481834828, "learning_rate": 9.65281436630257e-06, "loss": 0.044, "step": 203 }, { "epoch": 0.6257668711656442, "grad_norm": 0.6874182637171276, "learning_rate": 9.646249287234375e-06, "loss": 0.0327, "step": 204 }, { "epoch": 0.6288343558282209, "grad_norm": 0.460131137033222, "learning_rate": 9.639624992523474e-06, "loss": 0.0294, "step": 205 }, { "epoch": 0.6319018404907976, "grad_norm": 0.5745062676923193, "learning_rate": 9.632941566595357e-06, "loss": 0.0431, "step": 206 }, { "epoch": 0.6349693251533742, "grad_norm": 0.40736187331439816, "learning_rate": 9.626199094629132e-06, "loss": 0.0213, "step": 207 }, { "epoch": 0.6380368098159509, "grad_norm": 0.4237280861879652, "learning_rate": 9.619397662556434e-06, "loss": 0.0209, "step": 208 }, { "epoch": 0.6411042944785276, "grad_norm": 0.4262103667083273, "learning_rate": 9.612537357060338e-06, "loss": 0.0227, "step": 209 }, { "epoch": 0.6441717791411042, "grad_norm": 0.4164271188520456, "learning_rate": 9.60561826557425e-06, "loss": 0.0179, "step": 210 }, { "epoch": 0.647239263803681, "grad_norm": 0.4610603132848594, "learning_rate": 9.598640476280795e-06, "loss": 0.0239, "step": 211 }, { "epoch": 0.6503067484662577, "grad_norm": 0.4431083869802891, "learning_rate": 9.591604078110686e-06, "loss": 0.0174, "step": 212 }, { "epoch": 0.6533742331288344, "grad_norm": 0.34132375743490156, "learning_rate": 9.584509160741599e-06, "loss": 0.0134, "step": 213 }, { "epoch": 0.656441717791411, "grad_norm": 0.5748219630943767, "learning_rate": 9.577355814597031e-06, "loss": 0.0386, "step": 214 }, { "epoch": 0.6595092024539877, "grad_norm": 0.5478545394367712, "learning_rate": 9.570144130845139e-06, "loss": 0.0299, "step": 215 }, { "epoch": 0.6625766871165644, "grad_norm": 0.6412397075388798, "learning_rate": 9.56287420139758e-06, "loss": 0.0353, "step": 216 }, { "epoch": 0.6656441717791411, "grad_norm": 0.3984095445625212, "learning_rate": 9.555546118908351e-06, "loss": 0.0249, "step": 217 }, { "epoch": 0.6687116564417178, "grad_norm": 0.46212114811871957, "learning_rate": 9.548159976772593e-06, "loss": 0.0194, "step": 218 }, { "epoch": 0.6717791411042945, "grad_norm": 0.712044108676838, "learning_rate": 9.540715869125407e-06, "loss": 0.0512, "step": 219 }, { "epoch": 0.6748466257668712, "grad_norm": 0.6344419761933424, "learning_rate": 9.533213890840657e-06, "loss": 0.0443, "step": 220 }, { "epoch": 0.6779141104294478, "grad_norm": 0.5387333699270996, "learning_rate": 9.52565413752976e-06, "loss": 0.0221, "step": 221 }, { "epoch": 0.6809815950920245, "grad_norm": 0.4509061477504429, "learning_rate": 9.518036705540459e-06, "loss": 0.0195, "step": 222 }, { "epoch": 0.6840490797546013, "grad_norm": 0.49381501765327435, "learning_rate": 9.510361691955607e-06, "loss": 0.0226, "step": 223 }, { "epoch": 0.6871165644171779, "grad_norm": 0.7919318387287834, "learning_rate": 9.502629194591925e-06, "loss": 0.0734, "step": 224 }, { "epoch": 0.6901840490797546, "grad_norm": 0.45378402430883924, "learning_rate": 9.494839311998753e-06, "loss": 0.0258, "step": 225 }, { "epoch": 0.6932515337423313, "grad_norm": 0.4537878089824592, "learning_rate": 9.486992143456792e-06, "loss": 0.0242, "step": 226 }, { "epoch": 0.696319018404908, "grad_norm": 0.5028749745217802, "learning_rate": 9.47908778897685e-06, "loss": 0.023, "step": 227 }, { "epoch": 0.6993865030674846, "grad_norm": 0.6478779043026598, "learning_rate": 9.471126349298557e-06, "loss": 0.0438, "step": 228 }, { "epoch": 0.7024539877300614, "grad_norm": 0.5588980543929863, "learning_rate": 9.46310792588908e-06, "loss": 0.0293, "step": 229 }, { "epoch": 0.7055214723926381, "grad_norm": 0.3667360725898723, "learning_rate": 9.45503262094184e-06, "loss": 0.0168, "step": 230 }, { "epoch": 0.7085889570552147, "grad_norm": 0.48407267064607673, "learning_rate": 9.446900537375199e-06, "loss": 0.025, "step": 231 }, { "epoch": 0.7116564417177914, "grad_norm": 0.5343532675542083, "learning_rate": 9.438711778831154e-06, "loss": 0.0879, "step": 232 }, { "epoch": 0.7147239263803681, "grad_norm": 0.5280162128110422, "learning_rate": 9.430466449674014e-06, "loss": 0.0296, "step": 233 }, { "epoch": 0.7177914110429447, "grad_norm": 0.3758038349070859, "learning_rate": 9.422164654989073e-06, "loss": 0.0136, "step": 234 }, { "epoch": 0.7208588957055214, "grad_norm": 0.5117526267420697, "learning_rate": 9.413806500581267e-06, "loss": 0.0261, "step": 235 }, { "epoch": 0.7239263803680982, "grad_norm": 0.3981771203856743, "learning_rate": 9.405392092973824e-06, "loss": 0.0154, "step": 236 }, { "epoch": 0.7269938650306749, "grad_norm": 0.7223827510240209, "learning_rate": 9.396921539406916e-06, "loss": 0.0353, "step": 237 }, { "epoch": 0.7300613496932515, "grad_norm": 0.5167528334171827, "learning_rate": 9.388394947836278e-06, "loss": 0.0227, "step": 238 }, { "epoch": 0.7331288343558282, "grad_norm": 0.45907180962386196, "learning_rate": 9.379812426931847e-06, "loss": 0.0243, "step": 239 }, { "epoch": 0.7361963190184049, "grad_norm": 0.47959151249699833, "learning_rate": 9.371174086076364e-06, "loss": 0.0277, "step": 240 }, { "epoch": 0.7392638036809815, "grad_norm": 0.6484335074593419, "learning_rate": 9.362480035363987e-06, "loss": 0.035, "step": 241 }, { "epoch": 0.7423312883435583, "grad_norm": 0.788011620141102, "learning_rate": 9.353730385598887e-06, "loss": 0.046, "step": 242 }, { "epoch": 0.745398773006135, "grad_norm": 0.49585961794822014, "learning_rate": 9.344925248293837e-06, "loss": 0.0368, "step": 243 }, { "epoch": 0.7484662576687117, "grad_norm": 0.5556890213184674, "learning_rate": 9.336064735668784e-06, "loss": 0.0316, "step": 244 }, { "epoch": 0.7515337423312883, "grad_norm": 0.5654657395691955, "learning_rate": 9.327148960649431e-06, "loss": 0.0379, "step": 245 }, { "epoch": 0.754601226993865, "grad_norm": 0.6404905959487527, "learning_rate": 9.318178036865786e-06, "loss": 0.0458, "step": 246 }, { "epoch": 0.7576687116564417, "grad_norm": 0.47153130888171024, "learning_rate": 9.309152078650718e-06, "loss": 0.0236, "step": 247 }, { "epoch": 0.7607361963190185, "grad_norm": 0.6946471966054896, "learning_rate": 9.300071201038503e-06, "loss": 0.0351, "step": 248 }, { "epoch": 0.7638036809815951, "grad_norm": 0.7452826487701208, "learning_rate": 9.290935519763354e-06, "loss": 0.036, "step": 249 }, { "epoch": 0.7668711656441718, "grad_norm": 0.39427567747880354, "learning_rate": 9.281745151257946e-06, "loss": 0.0218, "step": 250 }, { "epoch": 0.7699386503067485, "grad_norm": 0.27722944056563503, "learning_rate": 9.272500212651934e-06, "loss": 0.0144, "step": 251 }, { "epoch": 0.7730061349693251, "grad_norm": 0.5591852306028331, "learning_rate": 9.263200821770462e-06, "loss": 0.0275, "step": 252 }, { "epoch": 0.7760736196319018, "grad_norm": 0.4621394066240241, "learning_rate": 9.253847097132656e-06, "loss": 0.0233, "step": 253 }, { "epoch": 0.7791411042944786, "grad_norm": 0.5194878891553444, "learning_rate": 9.244439157950114e-06, "loss": 0.0344, "step": 254 }, { "epoch": 0.7822085889570553, "grad_norm": 0.49348339338246816, "learning_rate": 9.234977124125395e-06, "loss": 0.0285, "step": 255 }, { "epoch": 0.7852760736196319, "grad_norm": 0.4850343759514216, "learning_rate": 9.225461116250483e-06, "loss": 0.0218, "step": 256 }, { "epoch": 0.7883435582822086, "grad_norm": 0.5770601668132987, "learning_rate": 9.215891255605248e-06, "loss": 0.0359, "step": 257 }, { "epoch": 0.7914110429447853, "grad_norm": 0.42625771720843614, "learning_rate": 9.206267664155906e-06, "loss": 0.0156, "step": 258 }, { "epoch": 0.7944785276073619, "grad_norm": 0.3426717684737494, "learning_rate": 9.196590464553467e-06, "loss": 0.011, "step": 259 }, { "epoch": 0.7975460122699386, "grad_norm": 0.5865425048649809, "learning_rate": 9.186859780132165e-06, "loss": 0.0376, "step": 260 }, { "epoch": 0.8006134969325154, "grad_norm": 0.37273506776238424, "learning_rate": 9.177075734907883e-06, "loss": 0.0234, "step": 261 }, { "epoch": 0.803680981595092, "grad_norm": 0.5585965373567147, "learning_rate": 9.167238453576589e-06, "loss": 0.0426, "step": 262 }, { "epoch": 0.8067484662576687, "grad_norm": 0.3651968044742629, "learning_rate": 9.157348061512728e-06, "loss": 0.0248, "step": 263 }, { "epoch": 0.8098159509202454, "grad_norm": 0.5081337037835217, "learning_rate": 9.147404684767632e-06, "loss": 0.0266, "step": 264 }, { "epoch": 0.8128834355828221, "grad_norm": 0.5073465299835719, "learning_rate": 9.137408450067921e-06, "loss": 0.0272, "step": 265 }, { "epoch": 0.8159509202453987, "grad_norm": 0.5174001449539224, "learning_rate": 9.12735948481387e-06, "loss": 0.0266, "step": 266 }, { "epoch": 0.8190184049079755, "grad_norm": 0.6040688463652676, "learning_rate": 9.117257917077806e-06, "loss": 0.0314, "step": 267 }, { "epoch": 0.8220858895705522, "grad_norm": 0.6431020172095245, "learning_rate": 9.107103875602458e-06, "loss": 0.0404, "step": 268 }, { "epoch": 0.8251533742331288, "grad_norm": 0.4570395520776521, "learning_rate": 9.096897489799329e-06, "loss": 0.0245, "step": 269 }, { "epoch": 0.8282208588957055, "grad_norm": 0.5176485694689812, "learning_rate": 9.086638889747034e-06, "loss": 0.0318, "step": 270 }, { "epoch": 0.8312883435582822, "grad_norm": 0.4030417312195923, "learning_rate": 9.076328206189662e-06, "loss": 0.0186, "step": 271 }, { "epoch": 0.8343558282208589, "grad_norm": 0.7912092430093546, "learning_rate": 9.065965570535083e-06, "loss": 0.0211, "step": 272 }, { "epoch": 0.8374233128834356, "grad_norm": 0.3882727172008151, "learning_rate": 9.055551114853296e-06, "loss": 0.0201, "step": 273 }, { "epoch": 0.8404907975460123, "grad_norm": 0.6169335765533825, "learning_rate": 9.045084971874738e-06, "loss": 0.0457, "step": 274 }, { "epoch": 0.843558282208589, "grad_norm": 0.484075907018457, "learning_rate": 9.034567274988585e-06, "loss": 0.0266, "step": 275 }, { "epoch": 0.8466257668711656, "grad_norm": 0.7402879740038205, "learning_rate": 9.023998158241067e-06, "loss": 0.0412, "step": 276 }, { "epoch": 0.8496932515337423, "grad_norm": 0.8200807227847277, "learning_rate": 9.013377756333745e-06, "loss": 0.024, "step": 277 }, { "epoch": 0.852760736196319, "grad_norm": 0.47917185258179273, "learning_rate": 9.002706204621802e-06, "loss": 0.0354, "step": 278 }, { "epoch": 0.8558282208588958, "grad_norm": 0.5686281464689263, "learning_rate": 8.991983639112318e-06, "loss": 0.0463, "step": 279 }, { "epoch": 0.8588957055214724, "grad_norm": 0.5035593732198547, "learning_rate": 8.981210196462533e-06, "loss": 0.024, "step": 280 }, { "epoch": 0.8619631901840491, "grad_norm": 0.40871215414302914, "learning_rate": 8.970386013978108e-06, "loss": 0.026, "step": 281 }, { "epoch": 0.8650306748466258, "grad_norm": 0.5136824340178904, "learning_rate": 8.959511229611377e-06, "loss": 0.0248, "step": 282 }, { "epoch": 0.8680981595092024, "grad_norm": 0.530309289976694, "learning_rate": 8.94858598195958e-06, "loss": 0.0249, "step": 283 }, { "epoch": 0.8711656441717791, "grad_norm": 0.6333632955147461, "learning_rate": 8.93761041026311e-06, "loss": 0.0417, "step": 284 }, { "epoch": 0.8742331288343558, "grad_norm": 0.3971198296026937, "learning_rate": 8.926584654403725e-06, "loss": 0.027, "step": 285 }, { "epoch": 0.8773006134969326, "grad_norm": 0.40732995997319704, "learning_rate": 8.915508854902778e-06, "loss": 0.0225, "step": 286 }, { "epoch": 0.8803680981595092, "grad_norm": 0.6090138795408967, "learning_rate": 8.904383152919414e-06, "loss": 0.018, "step": 287 }, { "epoch": 0.8834355828220859, "grad_norm": 0.7007910043919408, "learning_rate": 8.893207690248776e-06, "loss": 0.0367, "step": 288 }, { "epoch": 0.8865030674846626, "grad_norm": 0.5110317128121923, "learning_rate": 8.881982609320206e-06, "loss": 0.0242, "step": 289 }, { "epoch": 0.8895705521472392, "grad_norm": 0.6970377701125607, "learning_rate": 8.870708053195414e-06, "loss": 0.0483, "step": 290 }, { "epoch": 0.8926380368098159, "grad_norm": 0.5436387432201235, "learning_rate": 8.859384165566665e-06, "loss": 0.031, "step": 291 }, { "epoch": 0.8957055214723927, "grad_norm": 0.48318114304921483, "learning_rate": 8.848011090754946e-06, "loss": 0.0208, "step": 292 }, { "epoch": 0.8987730061349694, "grad_norm": 0.5078098207586927, "learning_rate": 8.836588973708129e-06, "loss": 0.0242, "step": 293 }, { "epoch": 0.901840490797546, "grad_norm": 0.43132547768511786, "learning_rate": 8.825117959999117e-06, "loss": 0.0209, "step": 294 }, { "epoch": 0.9049079754601227, "grad_norm": 0.4592204692663985, "learning_rate": 8.813598195823991e-06, "loss": 0.0141, "step": 295 }, { "epoch": 0.9079754601226994, "grad_norm": 0.8293697180348152, "learning_rate": 8.802029828000157e-06, "loss": 0.0341, "step": 296 }, { "epoch": 0.911042944785276, "grad_norm": 0.6353969688740292, "learning_rate": 8.790413003964452e-06, "loss": 0.0531, "step": 297 }, { "epoch": 0.9141104294478528, "grad_norm": 0.5135592242814184, "learning_rate": 8.778747871771293e-06, "loss": 0.0299, "step": 298 }, { "epoch": 0.9171779141104295, "grad_norm": 0.6244937792417771, "learning_rate": 8.767034580090767e-06, "loss": 0.0251, "step": 299 }, { "epoch": 0.9202453987730062, "grad_norm": 0.4900368286386423, "learning_rate": 8.75527327820675e-06, "loss": 0.0303, "step": 300 }, { "epoch": 0.9233128834355828, "grad_norm": 0.7676655526039488, "learning_rate": 8.743464116014998e-06, "loss": 0.055, "step": 301 }, { "epoch": 0.9263803680981595, "grad_norm": 0.48627457126119966, "learning_rate": 8.731607244021236e-06, "loss": 0.0264, "step": 302 }, { "epoch": 0.9294478527607362, "grad_norm": 0.433812749109046, "learning_rate": 8.719702813339248e-06, "loss": 0.0259, "step": 303 }, { "epoch": 0.9325153374233128, "grad_norm": 0.8758544568172416, "learning_rate": 8.70775097568894e-06, "loss": 0.045, "step": 304 }, { "epoch": 0.9355828220858896, "grad_norm": 0.45179382392506523, "learning_rate": 8.695751883394415e-06, "loss": 0.0243, "step": 305 }, { "epoch": 0.9386503067484663, "grad_norm": 0.4958191093808722, "learning_rate": 8.683705689382025e-06, "loss": 0.04, "step": 306 }, { "epoch": 0.941717791411043, "grad_norm": 0.44166098382760377, "learning_rate": 8.671612547178428e-06, "loss": 0.027, "step": 307 }, { "epoch": 0.9447852760736196, "grad_norm": 0.5911491565988687, "learning_rate": 8.659472610908628e-06, "loss": 0.0191, "step": 308 }, { "epoch": 0.9478527607361963, "grad_norm": 0.4765718802258442, "learning_rate": 8.647286035294009e-06, "loss": 0.0264, "step": 309 }, { "epoch": 0.950920245398773, "grad_norm": 0.5852473574381952, "learning_rate": 8.63505297565037e-06, "loss": 0.0294, "step": 310 }, { "epoch": 0.9539877300613497, "grad_norm": 0.484982934872736, "learning_rate": 8.622773587885935e-06, "loss": 0.0311, "step": 311 }, { "epoch": 0.9570552147239264, "grad_norm": 0.4847238416638495, "learning_rate": 8.610448028499377e-06, "loss": 0.0295, "step": 312 }, { "epoch": 0.9601226993865031, "grad_norm": 0.47605607743918293, "learning_rate": 8.598076454577815e-06, "loss": 0.0192, "step": 313 }, { "epoch": 0.9631901840490797, "grad_norm": 0.42846762697517876, "learning_rate": 8.585659023794818e-06, "loss": 0.016, "step": 314 }, { "epoch": 0.9662576687116564, "grad_norm": 0.6802283196669912, "learning_rate": 8.57319589440839e-06, "loss": 0.0617, "step": 315 }, { "epoch": 0.9693251533742331, "grad_norm": 0.6454169454682009, "learning_rate": 8.56068722525896e-06, "loss": 0.038, "step": 316 }, { "epoch": 0.9723926380368099, "grad_norm": 0.428982591104215, "learning_rate": 8.548133175767345e-06, "loss": 0.0236, "step": 317 }, { "epoch": 0.9754601226993865, "grad_norm": 0.5625492848129335, "learning_rate": 8.535533905932739e-06, "loss": 0.0271, "step": 318 }, { "epoch": 0.9785276073619632, "grad_norm": 0.4561896126978212, "learning_rate": 8.522889576330649e-06, "loss": 0.0256, "step": 319 }, { "epoch": 0.9815950920245399, "grad_norm": 0.49887714884464673, "learning_rate": 8.510200348110868e-06, "loss": 0.0244, "step": 320 }, { "epoch": 0.9846625766871165, "grad_norm": 0.3420730257876837, "learning_rate": 8.497466382995413e-06, "loss": 0.0164, "step": 321 }, { "epoch": 0.9877300613496932, "grad_norm": 0.5755461353996324, "learning_rate": 8.48468784327647e-06, "loss": 0.0326, "step": 322 }, { "epoch": 0.99079754601227, "grad_norm": 0.800649841232001, "learning_rate": 8.471864891814304e-06, "loss": 0.0431, "step": 323 }, { "epoch": 0.9938650306748467, "grad_norm": 0.7338066801521459, "learning_rate": 8.45899769203522e-06, "loss": 0.046, "step": 324 }, { "epoch": 0.9969325153374233, "grad_norm": 0.6907151900519473, "learning_rate": 8.446086407929448e-06, "loss": 0.0382, "step": 325 }, { "epoch": 1.0, "grad_norm": 0.6407879693661741, "learning_rate": 8.433131204049067e-06, "loss": 0.0356, "step": 326 }, { "epoch": 1.0030674846625767, "grad_norm": 0.3297580524046458, "learning_rate": 8.420132245505912e-06, "loss": 0.0111, "step": 327 }, { "epoch": 1.0061349693251533, "grad_norm": 0.9160618795214253, "learning_rate": 8.407089697969458e-06, "loss": 0.0068, "step": 328 }, { "epoch": 1.00920245398773, "grad_norm": 0.4355224927337849, "learning_rate": 8.39400372766471e-06, "loss": 0.0129, "step": 329 }, { "epoch": 1.0122699386503067, "grad_norm": 0.25041904255355624, "learning_rate": 8.380874501370098e-06, "loss": 0.0051, "step": 330 }, { "epoch": 1.0153374233128833, "grad_norm": 0.3627103002389334, "learning_rate": 8.367702186415338e-06, "loss": 0.019, "step": 331 }, { "epoch": 1.01840490797546, "grad_norm": 0.38724283256482883, "learning_rate": 8.354486950679301e-06, "loss": 0.0139, "step": 332 }, { "epoch": 1.021472392638037, "grad_norm": 0.2420700649978279, "learning_rate": 8.341228962587881e-06, "loss": 0.0102, "step": 333 }, { "epoch": 1.0245398773006136, "grad_norm": 0.4873618568078472, "learning_rate": 8.327928391111841e-06, "loss": 0.0182, "step": 334 }, { "epoch": 1.0276073619631902, "grad_norm": 0.4856636301987984, "learning_rate": 8.31458540576466e-06, "loss": 0.0188, "step": 335 }, { "epoch": 1.030674846625767, "grad_norm": 0.27927576098631274, "learning_rate": 8.301200176600375e-06, "loss": 0.0086, "step": 336 }, { "epoch": 1.0337423312883436, "grad_norm": 0.4700940197211149, "learning_rate": 8.287772874211414e-06, "loss": 0.0109, "step": 337 }, { "epoch": 1.0368098159509203, "grad_norm": 0.2741293793378236, "learning_rate": 8.274303669726427e-06, "loss": 0.0086, "step": 338 }, { "epoch": 1.039877300613497, "grad_norm": 0.47118341110657486, "learning_rate": 8.260792734808085e-06, "loss": 0.0115, "step": 339 }, { "epoch": 1.0429447852760736, "grad_norm": 0.19194865035728823, "learning_rate": 8.247240241650918e-06, "loss": 0.0053, "step": 340 }, { "epoch": 1.0460122699386503, "grad_norm": 0.37141621100709526, "learning_rate": 8.233646362979106e-06, "loss": 0.0178, "step": 341 }, { "epoch": 1.049079754601227, "grad_norm": 0.3513576345329715, "learning_rate": 8.220011272044276e-06, "loss": 0.0091, "step": 342 }, { "epoch": 1.0521472392638036, "grad_norm": 0.4198658600158762, "learning_rate": 8.206335142623305e-06, "loss": 0.0122, "step": 343 }, { "epoch": 1.0552147239263803, "grad_norm": 0.4461226377098048, "learning_rate": 8.192618149016092e-06, "loss": 0.0142, "step": 344 }, { "epoch": 1.058282208588957, "grad_norm": 0.2883795931974432, "learning_rate": 8.178860466043344e-06, "loss": 0.0084, "step": 345 }, { "epoch": 1.0613496932515338, "grad_norm": 0.2659747508038468, "learning_rate": 8.165062269044353e-06, "loss": 0.0059, "step": 346 }, { "epoch": 1.0644171779141105, "grad_norm": 0.4052660141940983, "learning_rate": 8.151223733874747e-06, "loss": 0.0077, "step": 347 }, { "epoch": 1.0674846625766872, "grad_norm": 0.43345686528418303, "learning_rate": 8.13734503690426e-06, "loss": 0.0047, "step": 348 }, { "epoch": 1.0705521472392638, "grad_norm": 0.5189604081089867, "learning_rate": 8.123426355014485e-06, "loss": 0.0059, "step": 349 }, { "epoch": 1.0736196319018405, "grad_norm": 0.5168669787962032, "learning_rate": 8.109467865596612e-06, "loss": 0.0102, "step": 350 }, { "epoch": 1.0766871165644172, "grad_norm": 0.6184876506284773, "learning_rate": 8.095469746549172e-06, "loss": 0.0155, "step": 351 }, { "epoch": 1.0797546012269938, "grad_norm": 0.4523533733127926, "learning_rate": 8.081432176275765e-06, "loss": 0.012, "step": 352 }, { "epoch": 1.0828220858895705, "grad_norm": 0.6042328026836615, "learning_rate": 8.067355333682799e-06, "loss": 0.0099, "step": 353 }, { "epoch": 1.0858895705521472, "grad_norm": 0.3712026513508595, "learning_rate": 8.053239398177191e-06, "loss": 0.0064, "step": 354 }, { "epoch": 1.0889570552147239, "grad_norm": 0.20180786491281724, "learning_rate": 8.039084549664098e-06, "loss": 0.0027, "step": 355 }, { "epoch": 1.0920245398773005, "grad_norm": 0.3787474646578074, "learning_rate": 8.024890968544614e-06, "loss": 0.0059, "step": 356 }, { "epoch": 1.0950920245398774, "grad_norm": 0.4859890722014772, "learning_rate": 8.01065883571347e-06, "loss": 0.0196, "step": 357 }, { "epoch": 1.098159509202454, "grad_norm": 0.8325992603907834, "learning_rate": 7.996388332556735e-06, "loss": 0.0213, "step": 358 }, { "epoch": 1.1012269938650308, "grad_norm": 0.7990116238977559, "learning_rate": 7.982079640949504e-06, "loss": 0.0167, "step": 359 }, { "epoch": 1.1042944785276074, "grad_norm": 0.3820950122221644, "learning_rate": 7.967732943253572e-06, "loss": 0.007, "step": 360 }, { "epoch": 1.107361963190184, "grad_norm": 0.43830665202404445, "learning_rate": 7.953348422315115e-06, "loss": 0.0123, "step": 361 }, { "epoch": 1.1104294478527608, "grad_norm": 0.26105895388858136, "learning_rate": 7.938926261462366e-06, "loss": 0.0065, "step": 362 }, { "epoch": 1.1134969325153374, "grad_norm": 0.395098975702821, "learning_rate": 7.924466644503265e-06, "loss": 0.0063, "step": 363 }, { "epoch": 1.116564417177914, "grad_norm": 0.49236578269029757, "learning_rate": 7.90996975572313e-06, "loss": 0.0102, "step": 364 }, { "epoch": 1.1196319018404908, "grad_norm": 0.2920551819139753, "learning_rate": 7.895435779882294e-06, "loss": 0.0047, "step": 365 }, { "epoch": 1.1226993865030674, "grad_norm": 0.652075843621035, "learning_rate": 7.880864902213765e-06, "loss": 0.0121, "step": 366 }, { "epoch": 1.1257668711656441, "grad_norm": 0.35616324665939764, "learning_rate": 7.866257308420858e-06, "loss": 0.0054, "step": 367 }, { "epoch": 1.1288343558282208, "grad_norm": 0.4669198553674882, "learning_rate": 7.851613184674821e-06, "loss": 0.0115, "step": 368 }, { "epoch": 1.1319018404907975, "grad_norm": 0.26000909056545185, "learning_rate": 7.836932717612481e-06, "loss": 0.006, "step": 369 }, { "epoch": 1.1349693251533743, "grad_norm": 0.4669430776716613, "learning_rate": 7.822216094333847e-06, "loss": 0.0099, "step": 370 }, { "epoch": 1.138036809815951, "grad_norm": 0.6134594985097804, "learning_rate": 7.807463502399735e-06, "loss": 0.0152, "step": 371 }, { "epoch": 1.1411042944785277, "grad_norm": 0.3822828619584183, "learning_rate": 7.792675129829374e-06, "loss": 0.0111, "step": 372 }, { "epoch": 1.1441717791411044, "grad_norm": 0.41327016906829583, "learning_rate": 7.777851165098012e-06, "loss": 0.0173, "step": 373 }, { "epoch": 1.147239263803681, "grad_norm": 0.2714658959849174, "learning_rate": 7.762991797134513e-06, "loss": 0.0107, "step": 374 }, { "epoch": 1.1503067484662577, "grad_norm": 0.5544800061588814, "learning_rate": 7.748097215318951e-06, "loss": 0.0118, "step": 375 }, { "epoch": 1.1533742331288344, "grad_norm": 0.5040932416460946, "learning_rate": 7.73316760948019e-06, "loss": 0.0099, "step": 376 }, { "epoch": 1.156441717791411, "grad_norm": 0.3685884655615084, "learning_rate": 7.718203169893472e-06, "loss": 0.0086, "step": 377 }, { "epoch": 1.1595092024539877, "grad_norm": 0.6030394172186191, "learning_rate": 7.703204087277989e-06, "loss": 0.0141, "step": 378 }, { "epoch": 1.1625766871165644, "grad_norm": 0.5895133158335721, "learning_rate": 7.68817055279445e-06, "loss": 0.013, "step": 379 }, { "epoch": 1.165644171779141, "grad_norm": 0.41300680548579227, "learning_rate": 7.673102758042653e-06, "loss": 0.0192, "step": 380 }, { "epoch": 1.1687116564417177, "grad_norm": 0.3374801828046708, "learning_rate": 7.65800089505903e-06, "loss": 0.0088, "step": 381 }, { "epoch": 1.1717791411042944, "grad_norm": 0.3454554679129908, "learning_rate": 7.64286515631421e-06, "loss": 0.0063, "step": 382 }, { "epoch": 1.1748466257668713, "grad_norm": 0.41181297072167894, "learning_rate": 7.627695734710565e-06, "loss": 0.0071, "step": 383 }, { "epoch": 1.177914110429448, "grad_norm": 0.2719651060060069, "learning_rate": 7.612492823579744e-06, "loss": 0.0069, "step": 384 }, { "epoch": 1.1809815950920246, "grad_norm": 0.7134985643754311, "learning_rate": 7.597256616680219e-06, "loss": 0.0128, "step": 385 }, { "epoch": 1.1840490797546013, "grad_norm": 0.41654763808421014, "learning_rate": 7.5819873081948105e-06, "loss": 0.0123, "step": 386 }, { "epoch": 1.187116564417178, "grad_norm": 0.387643208763756, "learning_rate": 7.566685092728207e-06, "loss": 0.0079, "step": 387 }, { "epoch": 1.1901840490797546, "grad_norm": 0.7361696877729214, "learning_rate": 7.5513501653045e-06, "loss": 0.0248, "step": 388 }, { "epoch": 1.1932515337423313, "grad_norm": 0.4404703959236721, "learning_rate": 7.535982721364679e-06, "loss": 0.0122, "step": 389 }, { "epoch": 1.196319018404908, "grad_norm": 0.4151059096070281, "learning_rate": 7.52058295676416e-06, "loss": 0.0109, "step": 390 }, { "epoch": 1.1993865030674846, "grad_norm": 0.34795956449739657, "learning_rate": 7.505151067770277e-06, "loss": 0.0093, "step": 391 }, { "epoch": 1.2024539877300613, "grad_norm": 0.33091849762828235, "learning_rate": 7.48968725105978e-06, "loss": 0.0066, "step": 392 }, { "epoch": 1.205521472392638, "grad_norm": 0.5219288124176215, "learning_rate": 7.474191703716339e-06, "loss": 0.0129, "step": 393 }, { "epoch": 1.2085889570552146, "grad_norm": 0.5725267558019617, "learning_rate": 7.45866462322802e-06, "loss": 0.0144, "step": 394 }, { "epoch": 1.2116564417177913, "grad_norm": 0.47084724943390577, "learning_rate": 7.443106207484776e-06, "loss": 0.0078, "step": 395 }, { "epoch": 1.2147239263803682, "grad_norm": 0.563293178774316, "learning_rate": 7.427516654775921e-06, "loss": 0.0097, "step": 396 }, { "epoch": 1.2177914110429449, "grad_norm": 0.45467942711747367, "learning_rate": 7.411896163787607e-06, "loss": 0.0234, "step": 397 }, { "epoch": 1.2208588957055215, "grad_norm": 0.607816846697092, "learning_rate": 7.396244933600285e-06, "loss": 0.0249, "step": 398 }, { "epoch": 1.2239263803680982, "grad_norm": 0.4194689019423566, "learning_rate": 7.380563163686174e-06, "loss": 0.0125, "step": 399 }, { "epoch": 1.2269938650306749, "grad_norm": 0.36586881186412595, "learning_rate": 7.364851053906719e-06, "loss": 0.013, "step": 400 }, { "epoch": 1.2300613496932515, "grad_norm": 0.5550225458831591, "learning_rate": 7.3491088045100365e-06, "loss": 0.0743, "step": 401 }, { "epoch": 1.2331288343558282, "grad_norm": 0.4466491951675168, "learning_rate": 7.333336616128369e-06, "loss": 0.0111, "step": 402 }, { "epoch": 1.2361963190184049, "grad_norm": 0.19051013453254942, "learning_rate": 7.317534689775527e-06, "loss": 0.003, "step": 403 }, { "epoch": 1.2392638036809815, "grad_norm": 0.2592362774390213, "learning_rate": 7.301703226844328e-06, "loss": 0.0044, "step": 404 }, { "epoch": 1.2423312883435582, "grad_norm": 0.4406593596675372, "learning_rate": 7.285842429104023e-06, "loss": 0.011, "step": 405 }, { "epoch": 1.2453987730061349, "grad_norm": 0.36125048050407316, "learning_rate": 7.269952498697734e-06, "loss": 0.0063, "step": 406 }, { "epoch": 1.2484662576687118, "grad_norm": 0.7328229395413601, "learning_rate": 7.2540336381398745e-06, "loss": 0.0181, "step": 407 }, { "epoch": 1.2515337423312882, "grad_norm": 0.46059726711050014, "learning_rate": 7.238086050313563e-06, "loss": 0.0138, "step": 408 }, { "epoch": 1.2546012269938651, "grad_norm": 0.32550746217650556, "learning_rate": 7.2221099384680485e-06, "loss": 0.0075, "step": 409 }, { "epoch": 1.2576687116564418, "grad_norm": 0.5498065211034909, "learning_rate": 7.206105506216107e-06, "loss": 0.01, "step": 410 }, { "epoch": 1.2607361963190185, "grad_norm": 0.49119461834253564, "learning_rate": 7.19007295753146e-06, "loss": 0.0093, "step": 411 }, { "epoch": 1.2638036809815951, "grad_norm": 0.5480841951866141, "learning_rate": 7.174012496746161e-06, "loss": 0.0117, "step": 412 }, { "epoch": 1.2668711656441718, "grad_norm": 0.46105615708600756, "learning_rate": 7.157924328548003e-06, "loss": 0.0116, "step": 413 }, { "epoch": 1.2699386503067485, "grad_norm": 0.3678598691625631, "learning_rate": 7.1418086579779075e-06, "loss": 0.006, "step": 414 }, { "epoch": 1.2730061349693251, "grad_norm": 0.3960979904562928, "learning_rate": 7.125665690427301e-06, "loss": 0.0087, "step": 415 }, { "epoch": 1.2760736196319018, "grad_norm": 0.43312438511224055, "learning_rate": 7.109495631635512e-06, "loss": 0.0071, "step": 416 }, { "epoch": 1.2791411042944785, "grad_norm": 0.4306674783353921, "learning_rate": 7.093298687687141e-06, "loss": 0.0104, "step": 417 }, { "epoch": 1.2822085889570551, "grad_norm": 0.42081355809821863, "learning_rate": 7.0770750650094335e-06, "loss": 0.0122, "step": 418 }, { "epoch": 1.2852760736196318, "grad_norm": 0.2781000974050112, "learning_rate": 7.060824970369651e-06, "loss": 0.0029, "step": 419 }, { "epoch": 1.2883435582822087, "grad_norm": 0.442937301778717, "learning_rate": 7.044548610872435e-06, "loss": 0.0086, "step": 420 }, { "epoch": 1.2914110429447851, "grad_norm": 0.37495174577866447, "learning_rate": 7.028246193957171e-06, "loss": 0.0104, "step": 421 }, { "epoch": 1.294478527607362, "grad_norm": 0.5023190671289056, "learning_rate": 7.011917927395341e-06, "loss": 0.0108, "step": 422 }, { "epoch": 1.2975460122699387, "grad_norm": 0.23654582360722132, "learning_rate": 6.9955640192878705e-06, "loss": 0.0042, "step": 423 }, { "epoch": 1.3006134969325154, "grad_norm": 0.3855334572363615, "learning_rate": 6.979184678062493e-06, "loss": 0.0095, "step": 424 }, { "epoch": 1.303680981595092, "grad_norm": 0.4576634484879509, "learning_rate": 6.962780112471073e-06, "loss": 0.0086, "step": 425 }, { "epoch": 1.3067484662576687, "grad_norm": 0.42556417794111534, "learning_rate": 6.946350531586959e-06, "loss": 0.0087, "step": 426 }, { "epoch": 1.3098159509202454, "grad_norm": 0.3237784652171552, "learning_rate": 6.929896144802314e-06, "loss": 0.0083, "step": 427 }, { "epoch": 1.312883435582822, "grad_norm": 0.3524888807864575, "learning_rate": 6.913417161825449e-06, "loss": 0.0086, "step": 428 }, { "epoch": 1.3159509202453987, "grad_norm": 0.4553142705420359, "learning_rate": 6.896913792678151e-06, "loss": 0.0116, "step": 429 }, { "epoch": 1.3190184049079754, "grad_norm": 0.40859472374578365, "learning_rate": 6.880386247692999e-06, "loss": 0.0111, "step": 430 }, { "epoch": 1.322085889570552, "grad_norm": 0.4777916998093413, "learning_rate": 6.863834737510695e-06, "loss": 0.0145, "step": 431 }, { "epoch": 1.3251533742331287, "grad_norm": 0.5288188669572824, "learning_rate": 6.84725947307737e-06, "loss": 0.0098, "step": 432 }, { "epoch": 1.3282208588957056, "grad_norm": 0.33681749772954367, "learning_rate": 6.830660665641898e-06, "loss": 0.0074, "step": 433 }, { "epoch": 1.331288343558282, "grad_norm": 0.380221021862379, "learning_rate": 6.814038526753205e-06, "loss": 0.0074, "step": 434 }, { "epoch": 1.334355828220859, "grad_norm": 0.47839600855959413, "learning_rate": 6.797393268257577e-06, "loss": 0.0111, "step": 435 }, { "epoch": 1.3374233128834356, "grad_norm": 0.4962504965337215, "learning_rate": 6.780725102295949e-06, "loss": 0.008, "step": 436 }, { "epoch": 1.3404907975460123, "grad_norm": 0.4665720038905597, "learning_rate": 6.764034241301208e-06, "loss": 0.01, "step": 437 }, { "epoch": 1.343558282208589, "grad_norm": 0.5222835885924834, "learning_rate": 6.747320897995493e-06, "loss": 0.0171, "step": 438 }, { "epoch": 1.3466257668711656, "grad_norm": 0.38877477124771365, "learning_rate": 6.730585285387465e-06, "loss": 0.0073, "step": 439 }, { "epoch": 1.3496932515337423, "grad_norm": 0.4599037526704758, "learning_rate": 6.7138276167696135e-06, "loss": 0.0082, "step": 440 }, { "epoch": 1.352760736196319, "grad_norm": 0.3746438265537782, "learning_rate": 6.697048105715521e-06, "loss": 0.0092, "step": 441 }, { "epoch": 1.3558282208588956, "grad_norm": 0.44845911397633703, "learning_rate": 6.680246966077151e-06, "loss": 0.0162, "step": 442 }, { "epoch": 1.3588957055214723, "grad_norm": 0.6221158471942455, "learning_rate": 6.663424411982121e-06, "loss": 0.0121, "step": 443 }, { "epoch": 1.3619631901840492, "grad_norm": 0.4239501149738595, "learning_rate": 6.646580657830967e-06, "loss": 0.0096, "step": 444 }, { "epoch": 1.3650306748466257, "grad_norm": 0.5037673671165694, "learning_rate": 6.629715918294422e-06, "loss": 0.012, "step": 445 }, { "epoch": 1.3680981595092025, "grad_norm": 0.5349443172718803, "learning_rate": 6.612830408310671e-06, "loss": 0.0137, "step": 446 }, { "epoch": 1.3711656441717792, "grad_norm": 0.5684159361020573, "learning_rate": 6.595924343082613e-06, "loss": 0.0096, "step": 447 }, { "epoch": 1.3742331288343559, "grad_norm": 0.47959961882205193, "learning_rate": 6.578997938075126e-06, "loss": 0.0126, "step": 448 }, { "epoch": 1.3773006134969326, "grad_norm": 0.3894016180198823, "learning_rate": 6.562051409012308e-06, "loss": 0.0086, "step": 449 }, { "epoch": 1.3803680981595092, "grad_norm": 0.5650260235219143, "learning_rate": 6.545084971874738e-06, "loss": 0.0065, "step": 450 }, { "epoch": 1.383435582822086, "grad_norm": 0.5097608717889676, "learning_rate": 6.528098842896721e-06, "loss": 0.0081, "step": 451 }, { "epoch": 1.3865030674846626, "grad_norm": 0.39970265760625523, "learning_rate": 6.51109323856353e-06, "loss": 0.0178, "step": 452 }, { "epoch": 1.3895705521472392, "grad_norm": 0.31936637132172235, "learning_rate": 6.494068375608646e-06, "loss": 0.006, "step": 453 }, { "epoch": 1.392638036809816, "grad_norm": 0.48256810572064407, "learning_rate": 6.477024471011001e-06, "loss": 0.0137, "step": 454 }, { "epoch": 1.3957055214723926, "grad_norm": 0.3268496682428815, "learning_rate": 6.459961741992208e-06, "loss": 0.0101, "step": 455 }, { "epoch": 1.3987730061349692, "grad_norm": 0.27522272221212113, "learning_rate": 6.442880406013795e-06, "loss": 0.0061, "step": 456 }, { "epoch": 1.4018404907975461, "grad_norm": 0.3967353171926939, "learning_rate": 6.425780680774429e-06, "loss": 0.0084, "step": 457 }, { "epoch": 1.4049079754601226, "grad_norm": 0.4825910261512796, "learning_rate": 6.408662784207149e-06, "loss": 0.0186, "step": 458 }, { "epoch": 1.4079754601226995, "grad_norm": 0.3756408674022022, "learning_rate": 6.391526934476584e-06, "loss": 0.0099, "step": 459 }, { "epoch": 1.4110429447852761, "grad_norm": 0.3003980819124125, "learning_rate": 6.374373349976169e-06, "loss": 0.0031, "step": 460 }, { "epoch": 1.4141104294478528, "grad_norm": 0.32353782969275396, "learning_rate": 6.3572022493253715e-06, "loss": 0.0073, "step": 461 }, { "epoch": 1.4171779141104295, "grad_norm": 0.4094722519435924, "learning_rate": 6.3400138513668955e-06, "loss": 0.0073, "step": 462 }, { "epoch": 1.4202453987730062, "grad_norm": 0.3620868544363379, "learning_rate": 6.322808375163896e-06, "loss": 0.0087, "step": 463 }, { "epoch": 1.4233128834355828, "grad_norm": 0.33066038137309844, "learning_rate": 6.305586039997188e-06, "loss": 0.0105, "step": 464 }, { "epoch": 1.4263803680981595, "grad_norm": 0.46544639219707007, "learning_rate": 6.2883470653624525e-06, "loss": 0.006, "step": 465 }, { "epoch": 1.4294478527607362, "grad_norm": 0.42293823165805505, "learning_rate": 6.271091670967437e-06, "loss": 0.0133, "step": 466 }, { "epoch": 1.4325153374233128, "grad_norm": 0.631453029409068, "learning_rate": 6.253820076729151e-06, "loss": 0.0165, "step": 467 }, { "epoch": 1.4355828220858895, "grad_norm": 0.40562333528166167, "learning_rate": 6.236532502771078e-06, "loss": 0.0115, "step": 468 }, { "epoch": 1.4386503067484662, "grad_norm": 0.6375756159004377, "learning_rate": 6.219229169420354e-06, "loss": 0.0183, "step": 469 }, { "epoch": 1.441717791411043, "grad_norm": 0.23763952949099515, "learning_rate": 6.201910297204963e-06, "loss": 0.0064, "step": 470 }, { "epoch": 1.4447852760736195, "grad_norm": 0.34910826558617847, "learning_rate": 6.1845761068509365e-06, "loss": 0.0063, "step": 471 }, { "epoch": 1.4478527607361964, "grad_norm": 0.3946649165234159, "learning_rate": 6.1672268192795285e-06, "loss": 0.0067, "step": 472 }, { "epoch": 1.450920245398773, "grad_norm": 0.3762470480826563, "learning_rate": 6.149862655604404e-06, "loss": 0.0056, "step": 473 }, { "epoch": 1.4539877300613497, "grad_norm": 0.32139974596711196, "learning_rate": 6.132483837128823e-06, "loss": 0.0081, "step": 474 }, { "epoch": 1.4570552147239264, "grad_norm": 0.49808493719028013, "learning_rate": 6.115090585342818e-06, "loss": 0.0135, "step": 475 }, { "epoch": 1.460122699386503, "grad_norm": 0.2557083632490389, "learning_rate": 6.097683121920373e-06, "loss": 0.0064, "step": 476 }, { "epoch": 1.4631901840490797, "grad_norm": 0.27134111617865514, "learning_rate": 6.080261668716594e-06, "loss": 0.0059, "step": 477 }, { "epoch": 1.4662576687116564, "grad_norm": 0.33147639324012906, "learning_rate": 6.062826447764883e-06, "loss": 0.0057, "step": 478 }, { "epoch": 1.469325153374233, "grad_norm": 0.6381027833440677, "learning_rate": 6.045377681274118e-06, "loss": 0.0171, "step": 479 }, { "epoch": 1.4723926380368098, "grad_norm": 0.4862891083399746, "learning_rate": 6.027915591625804e-06, "loss": 0.0088, "step": 480 }, { "epoch": 1.4754601226993864, "grad_norm": 0.4199767355964214, "learning_rate": 6.010440401371249e-06, "loss": 0.0137, "step": 481 }, { "epoch": 1.478527607361963, "grad_norm": 0.36036542279401557, "learning_rate": 5.9929523332287275e-06, "loss": 0.009, "step": 482 }, { "epoch": 1.48159509202454, "grad_norm": 0.48796334911431527, "learning_rate": 5.975451610080643e-06, "loss": 0.0119, "step": 483 }, { "epoch": 1.4846625766871164, "grad_norm": 0.5634036649823058, "learning_rate": 5.957938454970678e-06, "loss": 0.0252, "step": 484 }, { "epoch": 1.4877300613496933, "grad_norm": 0.40399169151476966, "learning_rate": 5.940413091100963e-06, "loss": 0.0095, "step": 485 }, { "epoch": 1.49079754601227, "grad_norm": 0.4399277205497462, "learning_rate": 5.922875741829227e-06, "loss": 0.0112, "step": 486 }, { "epoch": 1.4938650306748467, "grad_norm": 0.5288214595932401, "learning_rate": 5.905326630665952e-06, "loss": 0.0081, "step": 487 }, { "epoch": 1.4969325153374233, "grad_norm": 0.49760115052816156, "learning_rate": 5.887765981271518e-06, "loss": 0.0102, "step": 488 }, { "epoch": 1.5, "grad_norm": 0.41652107889647155, "learning_rate": 5.87019401745336e-06, "loss": 0.0049, "step": 489 }, { "epoch": 1.5030674846625767, "grad_norm": 0.2597678001433525, "learning_rate": 5.85261096316312e-06, "loss": 0.0073, "step": 490 }, { "epoch": 1.5061349693251533, "grad_norm": 0.38080455678840186, "learning_rate": 5.835017042493776e-06, "loss": 0.0096, "step": 491 }, { "epoch": 1.50920245398773, "grad_norm": 0.43224653405380203, "learning_rate": 5.817412479676801e-06, "loss": 0.0057, "step": 492 }, { "epoch": 1.5122699386503067, "grad_norm": 0.3262531946722451, "learning_rate": 5.799797499079301e-06, "loss": 0.0087, "step": 493 }, { "epoch": 1.5153374233128836, "grad_norm": 0.3016558145931953, "learning_rate": 5.782172325201155e-06, "loss": 0.0056, "step": 494 }, { "epoch": 1.51840490797546, "grad_norm": 0.37233224370927953, "learning_rate": 5.764537182672151e-06, "loss": 0.0083, "step": 495 }, { "epoch": 1.521472392638037, "grad_norm": 0.37132800727522514, "learning_rate": 5.746892296249126e-06, "loss": 0.011, "step": 496 }, { "epoch": 1.5245398773006134, "grad_norm": 0.40201041032133183, "learning_rate": 5.729237890813104e-06, "loss": 0.009, "step": 497 }, { "epoch": 1.5276073619631902, "grad_norm": 0.40733654757757093, "learning_rate": 5.711574191366427e-06, "loss": 0.0131, "step": 498 }, { "epoch": 1.530674846625767, "grad_norm": 0.6727885144806748, "learning_rate": 5.693901423029882e-06, "loss": 0.0152, "step": 499 }, { "epoch": 1.5337423312883436, "grad_norm": 1.000739650786129, "learning_rate": 5.676219811039844e-06, "loss": 0.0241, "step": 500 }, { "epoch": 1.5368098159509203, "grad_norm": 0.5235874249664593, "learning_rate": 5.658529580745399e-06, "loss": 0.0132, "step": 501 }, { "epoch": 1.539877300613497, "grad_norm": 0.5350650961027892, "learning_rate": 5.640830957605466e-06, "loss": 0.0127, "step": 502 }, { "epoch": 1.5429447852760736, "grad_norm": 0.3817664984754992, "learning_rate": 5.62312416718593e-06, "loss": 0.0114, "step": 503 }, { "epoch": 1.5460122699386503, "grad_norm": 0.3825385427474814, "learning_rate": 5.605409435156774e-06, "loss": 0.0138, "step": 504 }, { "epoch": 1.5490797546012272, "grad_norm": 0.21138154531244147, "learning_rate": 5.587686987289189e-06, "loss": 0.0035, "step": 505 }, { "epoch": 1.5521472392638036, "grad_norm": 0.38123503529777064, "learning_rate": 5.569957049452703e-06, "loss": 0.0089, "step": 506 }, { "epoch": 1.5552147239263805, "grad_norm": 0.600172423644381, "learning_rate": 5.552219847612307e-06, "loss": 0.0152, "step": 507 }, { "epoch": 1.558282208588957, "grad_norm": 0.4589414710140213, "learning_rate": 5.534475607825566e-06, "loss": 0.013, "step": 508 }, { "epoch": 1.5613496932515338, "grad_norm": 0.3995615128331652, "learning_rate": 5.516724556239741e-06, "loss": 0.0106, "step": 509 }, { "epoch": 1.5644171779141103, "grad_norm": 0.5902773538723448, "learning_rate": 5.498966919088914e-06, "loss": 0.0121, "step": 510 }, { "epoch": 1.5674846625766872, "grad_norm": 0.33586208745792945, "learning_rate": 5.481202922691097e-06, "loss": 0.0059, "step": 511 }, { "epoch": 1.5705521472392638, "grad_norm": 0.29125387626531496, "learning_rate": 5.4634327934453444e-06, "loss": 0.0055, "step": 512 }, { "epoch": 1.5736196319018405, "grad_norm": 0.4066095991071624, "learning_rate": 5.44565675782888e-06, "loss": 0.011, "step": 513 }, { "epoch": 1.5766871165644172, "grad_norm": 0.32808822216637573, "learning_rate": 5.4278750423942e-06, "loss": 0.0078, "step": 514 }, { "epoch": 1.5797546012269938, "grad_norm": 0.3314484003720461, "learning_rate": 5.410087873766187e-06, "loss": 0.0083, "step": 515 }, { "epoch": 1.5828220858895705, "grad_norm": 0.3336629272045388, "learning_rate": 5.392295478639226e-06, "loss": 0.0088, "step": 516 }, { "epoch": 1.5858895705521472, "grad_norm": 0.3628271904977443, "learning_rate": 5.3744980837743125e-06, "loss": 0.0105, "step": 517 }, { "epoch": 1.588957055214724, "grad_norm": 0.31955150321101644, "learning_rate": 5.356695915996162e-06, "loss": 0.0084, "step": 518 }, { "epoch": 1.5920245398773005, "grad_norm": 0.5210028446504106, "learning_rate": 5.338889202190322e-06, "loss": 0.0058, "step": 519 }, { "epoch": 1.5950920245398774, "grad_norm": 0.759371015661337, "learning_rate": 5.321078169300275e-06, "loss": 0.0613, "step": 520 }, { "epoch": 1.5981595092024539, "grad_norm": 0.3113390592301875, "learning_rate": 5.303263044324556e-06, "loss": 0.0041, "step": 521 }, { "epoch": 1.6012269938650308, "grad_norm": 0.4287829629260837, "learning_rate": 5.285444054313841e-06, "loss": 0.0087, "step": 522 }, { "epoch": 1.6042944785276072, "grad_norm": 0.46055387526192015, "learning_rate": 5.267621426368076e-06, "loss": 0.0082, "step": 523 }, { "epoch": 1.607361963190184, "grad_norm": 0.3797840805599973, "learning_rate": 5.249795387633571e-06, "loss": 0.0065, "step": 524 }, { "epoch": 1.6104294478527608, "grad_norm": 0.3930675190109252, "learning_rate": 5.231966165300099e-06, "loss": 0.0103, "step": 525 }, { "epoch": 1.6134969325153374, "grad_norm": 0.3442820596582778, "learning_rate": 5.214133986598014e-06, "loss": 0.0082, "step": 526 }, { "epoch": 1.616564417177914, "grad_norm": 0.3050671116162007, "learning_rate": 5.1962990787953436e-06, "loss": 0.006, "step": 527 }, { "epoch": 1.6196319018404908, "grad_norm": 0.3536096571221858, "learning_rate": 5.178461669194903e-06, "loss": 0.0065, "step": 528 }, { "epoch": 1.6226993865030674, "grad_norm": 0.4060284135099058, "learning_rate": 5.160621985131389e-06, "loss": 0.0257, "step": 529 }, { "epoch": 1.6257668711656441, "grad_norm": 0.25927903674544184, "learning_rate": 5.142780253968481e-06, "loss": 0.0066, "step": 530 }, { "epoch": 1.628834355828221, "grad_norm": 0.2838865350019555, "learning_rate": 5.124936703095961e-06, "loss": 0.0099, "step": 531 }, { "epoch": 1.6319018404907975, "grad_norm": 0.3504714577452807, "learning_rate": 5.1070915599267914e-06, "loss": 0.0069, "step": 532 }, { "epoch": 1.6349693251533743, "grad_norm": 0.3550050266957588, "learning_rate": 5.089245051894231e-06, "loss": 0.0111, "step": 533 }, { "epoch": 1.6380368098159508, "grad_norm": 0.35728707924870967, "learning_rate": 5.071397406448937e-06, "loss": 0.0099, "step": 534 }, { "epoch": 1.6411042944785277, "grad_norm": 0.3773540116151394, "learning_rate": 5.05354885105606e-06, "loss": 0.0058, "step": 535 }, { "epoch": 1.6441717791411041, "grad_norm": 0.3949338659980868, "learning_rate": 5.035699613192348e-06, "loss": 0.0082, "step": 536 }, { "epoch": 1.647239263803681, "grad_norm": 0.4646573686558157, "learning_rate": 5.017849920343246e-06, "loss": 0.01, "step": 537 }, { "epoch": 1.6503067484662577, "grad_norm": 0.5642128335140358, "learning_rate": 5e-06, "loss": 0.0143, "step": 538 }, { "epoch": 1.6533742331288344, "grad_norm": 0.2479092759047113, "learning_rate": 4.9821500796567565e-06, "loss": 0.0032, "step": 539 }, { "epoch": 1.656441717791411, "grad_norm": 0.42550957150659624, "learning_rate": 4.964300386807653e-06, "loss": 0.0089, "step": 540 }, { "epoch": 1.6595092024539877, "grad_norm": 0.36622590607369476, "learning_rate": 4.946451148943941e-06, "loss": 0.0064, "step": 541 }, { "epoch": 1.6625766871165644, "grad_norm": 0.4546077208760806, "learning_rate": 4.928602593551065e-06, "loss": 0.0113, "step": 542 }, { "epoch": 1.665644171779141, "grad_norm": 0.3196662160157007, "learning_rate": 4.91075494810577e-06, "loss": 0.0038, "step": 543 }, { "epoch": 1.668711656441718, "grad_norm": 0.1781642502108262, "learning_rate": 4.89290844007321e-06, "loss": 0.0026, "step": 544 }, { "epoch": 1.6717791411042944, "grad_norm": 0.44145521346579086, "learning_rate": 4.875063296904042e-06, "loss": 0.0147, "step": 545 }, { "epoch": 1.6748466257668713, "grad_norm": 0.2982215293082941, "learning_rate": 4.85721974603152e-06, "loss": 0.0034, "step": 546 }, { "epoch": 1.6779141104294477, "grad_norm": 0.5088058908733735, "learning_rate": 4.8393780148686135e-06, "loss": 0.0106, "step": 547 }, { "epoch": 1.6809815950920246, "grad_norm": 0.2439455722582877, "learning_rate": 4.821538330805098e-06, "loss": 0.0047, "step": 548 }, { "epoch": 1.6840490797546013, "grad_norm": 0.48092459587570097, "learning_rate": 4.803700921204659e-06, "loss": 0.0071, "step": 549 }, { "epoch": 1.687116564417178, "grad_norm": 0.43594614747810895, "learning_rate": 4.785866013401989e-06, "loss": 0.0097, "step": 550 }, { "epoch": 1.6901840490797546, "grad_norm": 0.2719151732829852, "learning_rate": 4.768033834699903e-06, "loss": 0.0051, "step": 551 }, { "epoch": 1.6932515337423313, "grad_norm": 0.2671115953727491, "learning_rate": 4.750204612366432e-06, "loss": 0.0059, "step": 552 }, { "epoch": 1.696319018404908, "grad_norm": 0.5253601733548036, "learning_rate": 4.7323785736319246e-06, "loss": 0.0081, "step": 553 }, { "epoch": 1.6993865030674846, "grad_norm": 0.3830793685344685, "learning_rate": 4.71455594568616e-06, "loss": 0.0083, "step": 554 }, { "epoch": 1.7024539877300615, "grad_norm": 0.31034852969587023, "learning_rate": 4.6967369556754476e-06, "loss": 0.0054, "step": 555 }, { "epoch": 1.705521472392638, "grad_norm": 0.4396568468476409, "learning_rate": 4.678921830699724e-06, "loss": 0.0085, "step": 556 }, { "epoch": 1.7085889570552149, "grad_norm": 0.40764526944989854, "learning_rate": 4.661110797809679e-06, "loss": 0.0084, "step": 557 }, { "epoch": 1.7116564417177913, "grad_norm": 0.4813930042673, "learning_rate": 4.643304084003839e-06, "loss": 0.0078, "step": 558 }, { "epoch": 1.7147239263803682, "grad_norm": 0.38796762918199773, "learning_rate": 4.62550191622569e-06, "loss": 0.0093, "step": 559 }, { "epoch": 1.7177914110429446, "grad_norm": 0.26689306148678504, "learning_rate": 4.6077045213607765e-06, "loss": 0.0077, "step": 560 }, { "epoch": 1.7208588957055215, "grad_norm": 0.40170358200898043, "learning_rate": 4.589912126233815e-06, "loss": 0.0087, "step": 561 }, { "epoch": 1.7239263803680982, "grad_norm": 0.2382315472763123, "learning_rate": 4.572124957605803e-06, "loss": 0.0044, "step": 562 }, { "epoch": 1.7269938650306749, "grad_norm": 0.7026811275146098, "learning_rate": 4.554343242171121e-06, "loss": 0.0095, "step": 563 }, { "epoch": 1.7300613496932515, "grad_norm": 0.35813866462271615, "learning_rate": 4.536567206554656e-06, "loss": 0.0082, "step": 564 }, { "epoch": 1.7331288343558282, "grad_norm": 0.461673612521973, "learning_rate": 4.5187970773089054e-06, "loss": 0.0173, "step": 565 }, { "epoch": 1.7361963190184049, "grad_norm": 0.2649325350539358, "learning_rate": 4.501033080911087e-06, "loss": 0.0067, "step": 566 }, { "epoch": 1.7392638036809815, "grad_norm": 0.36931292694664386, "learning_rate": 4.483275443760261e-06, "loss": 0.0071, "step": 567 }, { "epoch": 1.7423312883435584, "grad_norm": 0.46353780643994597, "learning_rate": 4.465524392174437e-06, "loss": 0.013, "step": 568 }, { "epoch": 1.7453987730061349, "grad_norm": 0.14701431312044777, "learning_rate": 4.447780152387694e-06, "loss": 0.0018, "step": 569 }, { "epoch": 1.7484662576687118, "grad_norm": 0.24576972288599763, "learning_rate": 4.430042950547298e-06, "loss": 0.0044, "step": 570 }, { "epoch": 1.7515337423312882, "grad_norm": 0.5453432857666575, "learning_rate": 4.4123130127108125e-06, "loss": 0.0196, "step": 571 }, { "epoch": 1.7546012269938651, "grad_norm": 0.3874868229083472, "learning_rate": 4.394590564843226e-06, "loss": 0.0054, "step": 572 }, { "epoch": 1.7576687116564416, "grad_norm": 0.3535835403255216, "learning_rate": 4.376875832814071e-06, "loss": 0.0053, "step": 573 }, { "epoch": 1.7607361963190185, "grad_norm": 0.2749489766791634, "learning_rate": 4.359169042394537e-06, "loss": 0.0058, "step": 574 }, { "epoch": 1.7638036809815951, "grad_norm": 0.27719491743372315, "learning_rate": 4.341470419254603e-06, "loss": 0.0056, "step": 575 }, { "epoch": 1.7668711656441718, "grad_norm": 0.42628419434925463, "learning_rate": 4.323780188960156e-06, "loss": 0.0082, "step": 576 }, { "epoch": 1.7699386503067485, "grad_norm": 0.37149048561678855, "learning_rate": 4.306098576970119e-06, "loss": 0.0073, "step": 577 }, { "epoch": 1.7730061349693251, "grad_norm": 0.4347325087122037, "learning_rate": 4.2884258086335755e-06, "loss": 0.0081, "step": 578 }, { "epoch": 1.7760736196319018, "grad_norm": 0.4011537380642771, "learning_rate": 4.270762109186896e-06, "loss": 0.007, "step": 579 }, { "epoch": 1.7791411042944785, "grad_norm": 0.3626109586279271, "learning_rate": 4.253107703750875e-06, "loss": 0.0069, "step": 580 }, { "epoch": 1.7822085889570554, "grad_norm": 0.27490585927193734, "learning_rate": 4.235462817327851e-06, "loss": 0.0053, "step": 581 }, { "epoch": 1.7852760736196318, "grad_norm": 0.4947399469938696, "learning_rate": 4.217827674798845e-06, "loss": 0.0058, "step": 582 }, { "epoch": 1.7883435582822087, "grad_norm": 0.4136276570729244, "learning_rate": 4.2002025009207e-06, "loss": 0.0117, "step": 583 }, { "epoch": 1.7914110429447851, "grad_norm": 0.38559944723198136, "learning_rate": 4.1825875203232015e-06, "loss": 0.0088, "step": 584 }, { "epoch": 1.794478527607362, "grad_norm": 0.4626653370701894, "learning_rate": 4.164982957506226e-06, "loss": 0.0078, "step": 585 }, { "epoch": 1.7975460122699385, "grad_norm": 0.24447282240138413, "learning_rate": 4.147389036836881e-06, "loss": 0.0026, "step": 586 }, { "epoch": 1.8006134969325154, "grad_norm": 0.3519436504460508, "learning_rate": 4.129805982546641e-06, "loss": 0.0085, "step": 587 }, { "epoch": 1.803680981595092, "grad_norm": 0.28482524916262597, "learning_rate": 4.1122340187284845e-06, "loss": 0.0033, "step": 588 }, { "epoch": 1.8067484662576687, "grad_norm": 0.29892032636140786, "learning_rate": 4.09467336933405e-06, "loss": 0.0042, "step": 589 }, { "epoch": 1.8098159509202454, "grad_norm": 0.4407796725334618, "learning_rate": 4.077124258170774e-06, "loss": 0.009, "step": 590 }, { "epoch": 1.812883435582822, "grad_norm": 0.4583133431318533, "learning_rate": 4.059586908899039e-06, "loss": 0.0076, "step": 591 }, { "epoch": 1.8159509202453987, "grad_norm": 0.8039166331283459, "learning_rate": 4.042061545029323e-06, "loss": 0.0089, "step": 592 }, { "epoch": 1.8190184049079754, "grad_norm": 0.3667345238233017, "learning_rate": 4.02454838991936e-06, "loss": 0.011, "step": 593 }, { "epoch": 1.8220858895705523, "grad_norm": 0.208178307389126, "learning_rate": 4.007047666771274e-06, "loss": 0.0022, "step": 594 }, { "epoch": 1.8251533742331287, "grad_norm": 0.5198316863113215, "learning_rate": 3.989559598628753e-06, "loss": 0.0117, "step": 595 }, { "epoch": 1.8282208588957056, "grad_norm": 0.4442650765755917, "learning_rate": 3.972084408374198e-06, "loss": 0.0101, "step": 596 }, { "epoch": 1.831288343558282, "grad_norm": 0.48936961648999217, "learning_rate": 3.954622318725884e-06, "loss": 0.0068, "step": 597 }, { "epoch": 1.834355828220859, "grad_norm": 0.20528461066058848, "learning_rate": 3.937173552235117e-06, "loss": 0.0045, "step": 598 }, { "epoch": 1.8374233128834356, "grad_norm": 0.29341413585659293, "learning_rate": 3.919738331283408e-06, "loss": 0.0057, "step": 599 }, { "epoch": 1.8404907975460123, "grad_norm": 0.2557977944149396, "learning_rate": 3.902316878079629e-06, "loss": 0.0041, "step": 600 }, { "epoch": 1.843558282208589, "grad_norm": 0.4227990072533373, "learning_rate": 3.8849094146571836e-06, "loss": 0.0119, "step": 601 }, { "epoch": 1.8466257668711656, "grad_norm": 0.5214661055541566, "learning_rate": 3.867516162871177e-06, "loss": 0.0141, "step": 602 }, { "epoch": 1.8496932515337423, "grad_norm": 0.37635698001587115, "learning_rate": 3.850137344395598e-06, "loss": 0.0082, "step": 603 }, { "epoch": 1.852760736196319, "grad_norm": 0.23128959245280112, "learning_rate": 3.832773180720475e-06, "loss": 0.0056, "step": 604 }, { "epoch": 1.8558282208588959, "grad_norm": 0.38717364607066673, "learning_rate": 3.815423893149065e-06, "loss": 0.0135, "step": 605 }, { "epoch": 1.8588957055214723, "grad_norm": 0.27375167119044574, "learning_rate": 3.798089702795038e-06, "loss": 0.0066, "step": 606 }, { "epoch": 1.8619631901840492, "grad_norm": 0.3375505038267758, "learning_rate": 3.7807708305796494e-06, "loss": 0.0076, "step": 607 }, { "epoch": 1.8650306748466257, "grad_norm": 0.5530104418841859, "learning_rate": 3.7634674972289227e-06, "loss": 0.0146, "step": 608 }, { "epoch": 1.8680981595092025, "grad_norm": 0.41041915581422406, "learning_rate": 3.7461799232708497e-06, "loss": 0.0051, "step": 609 }, { "epoch": 1.871165644171779, "grad_norm": 0.3713305823986346, "learning_rate": 3.7289083290325668e-06, "loss": 0.0102, "step": 610 }, { "epoch": 1.8742331288343559, "grad_norm": 0.46230186432718484, "learning_rate": 3.7116529346375487e-06, "loss": 0.0098, "step": 611 }, { "epoch": 1.8773006134969326, "grad_norm": 0.16535019806482476, "learning_rate": 3.694413960002814e-06, "loss": 0.003, "step": 612 }, { "epoch": 1.8803680981595092, "grad_norm": 0.2930581875533518, "learning_rate": 3.677191624836106e-06, "loss": 0.0032, "step": 613 }, { "epoch": 1.883435582822086, "grad_norm": 0.373888513580928, "learning_rate": 3.6599861486331074e-06, "loss": 0.0059, "step": 614 }, { "epoch": 1.8865030674846626, "grad_norm": 0.6381390350325629, "learning_rate": 3.6427977506746293e-06, "loss": 0.0122, "step": 615 }, { "epoch": 1.8895705521472392, "grad_norm": 0.41045091382297616, "learning_rate": 3.6256266500238312e-06, "loss": 0.0129, "step": 616 }, { "epoch": 1.892638036809816, "grad_norm": 0.49828625275146515, "learning_rate": 3.608473065523419e-06, "loss": 0.01, "step": 617 }, { "epoch": 1.8957055214723928, "grad_norm": 0.3956832540966562, "learning_rate": 3.5913372157928515e-06, "loss": 0.0062, "step": 618 }, { "epoch": 1.8987730061349692, "grad_norm": 0.3672630309971258, "learning_rate": 3.5742193192255724e-06, "loss": 0.0061, "step": 619 }, { "epoch": 1.9018404907975461, "grad_norm": 0.3145272087003987, "learning_rate": 3.557119593986208e-06, "loss": 0.0059, "step": 620 }, { "epoch": 1.9049079754601226, "grad_norm": 0.30910831663414223, "learning_rate": 3.5400382580077923e-06, "loss": 0.0033, "step": 621 }, { "epoch": 1.9079754601226995, "grad_norm": 0.1959441278445393, "learning_rate": 3.5229755289890005e-06, "loss": 0.0043, "step": 622 }, { "epoch": 1.911042944785276, "grad_norm": 0.5277004855106132, "learning_rate": 3.5059316243913556e-06, "loss": 0.0106, "step": 623 }, { "epoch": 1.9141104294478528, "grad_norm": 0.21771914977302229, "learning_rate": 3.4889067614364714e-06, "loss": 0.0041, "step": 624 }, { "epoch": 1.9171779141104295, "grad_norm": 0.39360915150568093, "learning_rate": 3.47190115710328e-06, "loss": 0.0073, "step": 625 }, { "epoch": 1.9202453987730062, "grad_norm": 0.3485481083987293, "learning_rate": 3.4549150281252635e-06, "loss": 0.0136, "step": 626 }, { "epoch": 1.9233128834355828, "grad_norm": 0.24255517433539572, "learning_rate": 3.437948590987693e-06, "loss": 0.0051, "step": 627 }, { "epoch": 1.9263803680981595, "grad_norm": 0.29586738644042815, "learning_rate": 3.4210020619248762e-06, "loss": 0.0089, "step": 628 }, { "epoch": 1.9294478527607362, "grad_norm": 0.4186524656839354, "learning_rate": 3.404075656917388e-06, "loss": 0.0074, "step": 629 }, { "epoch": 1.9325153374233128, "grad_norm": 0.30619441436870193, "learning_rate": 3.3871695916893314e-06, "loss": 0.0096, "step": 630 }, { "epoch": 1.9355828220858897, "grad_norm": 0.578307433564674, "learning_rate": 3.370284081705579e-06, "loss": 0.0075, "step": 631 }, { "epoch": 1.9386503067484662, "grad_norm": 0.44719452467917326, "learning_rate": 3.3534193421690348e-06, "loss": 0.0108, "step": 632 }, { "epoch": 1.941717791411043, "grad_norm": 0.26118371726805234, "learning_rate": 3.336575588017881e-06, "loss": 0.0052, "step": 633 }, { "epoch": 1.9447852760736195, "grad_norm": 0.48800768739543726, "learning_rate": 3.319753033922849e-06, "loss": 0.0086, "step": 634 }, { "epoch": 1.9478527607361964, "grad_norm": 0.4006588666794408, "learning_rate": 3.302951894284481e-06, "loss": 0.0114, "step": 635 }, { "epoch": 1.9509202453987728, "grad_norm": 0.2289092092840238, "learning_rate": 3.286172383230388e-06, "loss": 0.0024, "step": 636 }, { "epoch": 1.9539877300613497, "grad_norm": 0.4127045861094555, "learning_rate": 3.269414714612534e-06, "loss": 0.0087, "step": 637 }, { "epoch": 1.9570552147239264, "grad_norm": 0.3163153860623139, "learning_rate": 3.252679102004509e-06, "loss": 0.0065, "step": 638 }, { "epoch": 1.960122699386503, "grad_norm": 0.2603842753498582, "learning_rate": 3.2359657586987934e-06, "loss": 0.0067, "step": 639 }, { "epoch": 1.9631901840490797, "grad_norm": 0.557283191260559, "learning_rate": 3.2192748977040535e-06, "loss": 0.0097, "step": 640 }, { "epoch": 1.9662576687116564, "grad_norm": 0.34901479007130604, "learning_rate": 3.2026067317424244e-06, "loss": 0.0096, "step": 641 }, { "epoch": 1.969325153374233, "grad_norm": 0.45136631101338426, "learning_rate": 3.1859614732467957e-06, "loss": 0.0135, "step": 642 }, { "epoch": 1.9723926380368098, "grad_norm": 0.23384558728197918, "learning_rate": 3.169339334358105e-06, "loss": 0.0054, "step": 643 }, { "epoch": 1.9754601226993866, "grad_norm": 0.14084876331113608, "learning_rate": 3.152740526922631e-06, "loss": 0.0017, "step": 644 }, { "epoch": 1.978527607361963, "grad_norm": 0.4962557161989476, "learning_rate": 3.1361652624893065e-06, "loss": 0.0534, "step": 645 }, { "epoch": 1.98159509202454, "grad_norm": 0.34403517138193335, "learning_rate": 3.119613752307002e-06, "loss": 0.0041, "step": 646 }, { "epoch": 1.9846625766871164, "grad_norm": 0.15306880270008735, "learning_rate": 3.1030862073218493e-06, "loss": 0.0023, "step": 647 }, { "epoch": 1.9877300613496933, "grad_norm": 0.33113489538489377, "learning_rate": 3.0865828381745515e-06, "loss": 0.0071, "step": 648 }, { "epoch": 1.99079754601227, "grad_norm": 0.5136944684210311, "learning_rate": 3.0701038551976888e-06, "loss": 0.0127, "step": 649 }, { "epoch": 1.9938650306748467, "grad_norm": 0.636270989600524, "learning_rate": 3.053649468413043e-06, "loss": 0.0163, "step": 650 }, { "epoch": 1.9969325153374233, "grad_norm": 0.4511697907324155, "learning_rate": 3.037219887528928e-06, "loss": 0.0075, "step": 651 }, { "epoch": 2.0, "grad_norm": 0.3185377520844948, "learning_rate": 3.020815321937509e-06, "loss": 0.0052, "step": 652 }, { "epoch": 2.003067484662577, "grad_norm": 0.1568715672001099, "learning_rate": 3.0044359807121295e-06, "loss": 0.0029, "step": 653 }, { "epoch": 2.0061349693251533, "grad_norm": 0.08384719456802235, "learning_rate": 2.9880820726046613e-06, "loss": 0.0009, "step": 654 }, { "epoch": 2.0092024539877302, "grad_norm": 0.2688822844440017, "learning_rate": 2.9717538060428307e-06, "loss": 0.0057, "step": 655 }, { "epoch": 2.0122699386503067, "grad_norm": 0.156419801199432, "learning_rate": 2.955451389127567e-06, "loss": 0.0027, "step": 656 }, { "epoch": 2.0153374233128836, "grad_norm": 0.13118375650446254, "learning_rate": 2.9391750296303506e-06, "loss": 0.0009, "step": 657 }, { "epoch": 2.01840490797546, "grad_norm": 0.2776210187070672, "learning_rate": 2.9229249349905686e-06, "loss": 0.003, "step": 658 }, { "epoch": 2.021472392638037, "grad_norm": 0.23522515094248445, "learning_rate": 2.906701312312861e-06, "loss": 0.0042, "step": 659 }, { "epoch": 2.0245398773006134, "grad_norm": 0.2197503271341387, "learning_rate": 2.8905043683644875e-06, "loss": 0.0022, "step": 660 }, { "epoch": 2.0276073619631902, "grad_norm": 0.45337971579413167, "learning_rate": 2.8743343095727006e-06, "loss": 0.0059, "step": 661 }, { "epoch": 2.0306748466257667, "grad_norm": 0.18232608740756975, "learning_rate": 2.8581913420220954e-06, "loss": 0.0017, "step": 662 }, { "epoch": 2.0337423312883436, "grad_norm": 0.10602348834285584, "learning_rate": 2.842075671451996e-06, "loss": 0.0027, "step": 663 }, { "epoch": 2.03680981595092, "grad_norm": 0.24863800686061804, "learning_rate": 2.8259875032538407e-06, "loss": 0.0053, "step": 664 }, { "epoch": 2.039877300613497, "grad_norm": 0.24148419846802188, "learning_rate": 2.8099270424685427e-06, "loss": 0.0036, "step": 665 }, { "epoch": 2.042944785276074, "grad_norm": 0.18602273100673639, "learning_rate": 2.7938944937838924e-06, "loss": 0.0031, "step": 666 }, { "epoch": 2.0460122699386503, "grad_norm": 0.22353235398611618, "learning_rate": 2.7778900615319536e-06, "loss": 0.0018, "step": 667 }, { "epoch": 2.049079754601227, "grad_norm": 0.22727585190425154, "learning_rate": 2.761913949686438e-06, "loss": 0.0023, "step": 668 }, { "epoch": 2.0521472392638036, "grad_norm": 0.18334423927687288, "learning_rate": 2.745966361860127e-06, "loss": 0.0023, "step": 669 }, { "epoch": 2.0552147239263805, "grad_norm": 0.18018696754304367, "learning_rate": 2.7300475013022666e-06, "loss": 0.0033, "step": 670 }, { "epoch": 2.058282208588957, "grad_norm": 0.2365128870735293, "learning_rate": 2.7141575708959777e-06, "loss": 0.0025, "step": 671 }, { "epoch": 2.061349693251534, "grad_norm": 0.24627552484555662, "learning_rate": 2.698296773155673e-06, "loss": 0.0024, "step": 672 }, { "epoch": 2.0644171779141103, "grad_norm": 0.3428729630059143, "learning_rate": 2.682465310224473e-06, "loss": 0.0096, "step": 673 }, { "epoch": 2.067484662576687, "grad_norm": 0.21165758715609964, "learning_rate": 2.6666633838716317e-06, "loss": 0.0043, "step": 674 }, { "epoch": 2.0705521472392636, "grad_norm": 0.12204006380648764, "learning_rate": 2.6508911954899648e-06, "loss": 0.0014, "step": 675 }, { "epoch": 2.0736196319018405, "grad_norm": 0.1545359743955643, "learning_rate": 2.6351489460932815e-06, "loss": 0.0038, "step": 676 }, { "epoch": 2.0766871165644174, "grad_norm": 0.27334965743962947, "learning_rate": 2.6194368363138258e-06, "loss": 0.0039, "step": 677 }, { "epoch": 2.079754601226994, "grad_norm": 0.14500449963412065, "learning_rate": 2.603755066399718e-06, "loss": 0.0026, "step": 678 }, { "epoch": 2.0828220858895707, "grad_norm": 0.25173969209235253, "learning_rate": 2.5881038362123937e-06, "loss": 0.004, "step": 679 }, { "epoch": 2.085889570552147, "grad_norm": 0.23000782863325714, "learning_rate": 2.5724833452240796e-06, "loss": 0.0045, "step": 680 }, { "epoch": 2.088957055214724, "grad_norm": 0.3760891599612169, "learning_rate": 2.5568937925152272e-06, "loss": 0.0071, "step": 681 }, { "epoch": 2.0920245398773005, "grad_norm": 0.4376412312537201, "learning_rate": 2.5413353767719805e-06, "loss": 0.0043, "step": 682 }, { "epoch": 2.0950920245398774, "grad_norm": 0.05798098025158203, "learning_rate": 2.5258082962836616e-06, "loss": 0.0007, "step": 683 }, { "epoch": 2.098159509202454, "grad_norm": 0.2054750633268483, "learning_rate": 2.510312748940222e-06, "loss": 0.0027, "step": 684 }, { "epoch": 2.1012269938650308, "grad_norm": 0.11726586713716912, "learning_rate": 2.4948489322297258e-06, "loss": 0.0016, "step": 685 }, { "epoch": 2.104294478527607, "grad_norm": 0.12698443724257102, "learning_rate": 2.4794170432358415e-06, "loss": 0.0031, "step": 686 }, { "epoch": 2.107361963190184, "grad_norm": 0.22795055728578995, "learning_rate": 2.464017278635323e-06, "loss": 0.0032, "step": 687 }, { "epoch": 2.1104294478527605, "grad_norm": 0.1957131964875298, "learning_rate": 2.448649834695503e-06, "loss": 0.0033, "step": 688 }, { "epoch": 2.1134969325153374, "grad_norm": 0.19210852074104182, "learning_rate": 2.433314907271794e-06, "loss": 0.0027, "step": 689 }, { "epoch": 2.116564417177914, "grad_norm": 0.04233819240534726, "learning_rate": 2.418012691805191e-06, "loss": 0.0005, "step": 690 }, { "epoch": 2.1196319018404908, "grad_norm": 0.26464285040782537, "learning_rate": 2.402743383319781e-06, "loss": 0.0039, "step": 691 }, { "epoch": 2.1226993865030677, "grad_norm": 0.11980748517310268, "learning_rate": 2.387507176420256e-06, "loss": 0.0021, "step": 692 }, { "epoch": 2.125766871165644, "grad_norm": 0.3334739663015842, "learning_rate": 2.372304265289436e-06, "loss": 0.0037, "step": 693 }, { "epoch": 2.128834355828221, "grad_norm": 0.6850456912727723, "learning_rate": 2.3571348436857906e-06, "loss": 0.0028, "step": 694 }, { "epoch": 2.1319018404907975, "grad_norm": 0.28144676205745234, "learning_rate": 2.3419991049409713e-06, "loss": 0.0036, "step": 695 }, { "epoch": 2.1349693251533743, "grad_norm": 0.108927883546715, "learning_rate": 2.326897241957348e-06, "loss": 0.0018, "step": 696 }, { "epoch": 2.138036809815951, "grad_norm": 0.38148279786228145, "learning_rate": 2.3118294472055497e-06, "loss": 0.0019, "step": 697 }, { "epoch": 2.1411042944785277, "grad_norm": 0.1970691641521866, "learning_rate": 2.296795912722014e-06, "loss": 0.0041, "step": 698 }, { "epoch": 2.144171779141104, "grad_norm": 0.15517303717415623, "learning_rate": 2.2817968301065284e-06, "loss": 0.002, "step": 699 }, { "epoch": 2.147239263803681, "grad_norm": 0.14207658565717568, "learning_rate": 2.266832390519811e-06, "loss": 0.0027, "step": 700 }, { "epoch": 2.1503067484662575, "grad_norm": 0.15125238624427667, "learning_rate": 2.2519027846810517e-06, "loss": 0.0019, "step": 701 }, { "epoch": 2.1533742331288344, "grad_norm": 0.7896376793637964, "learning_rate": 2.2370082028654867e-06, "loss": 0.0032, "step": 702 }, { "epoch": 2.1564417177914113, "grad_norm": 0.10686878822853349, "learning_rate": 2.2221488349019903e-06, "loss": 0.0009, "step": 703 }, { "epoch": 2.1595092024539877, "grad_norm": 0.22123439350597152, "learning_rate": 2.207324870170629e-06, "loss": 0.0052, "step": 704 }, { "epoch": 2.1625766871165646, "grad_norm": 0.2728382879459407, "learning_rate": 2.192536497600266e-06, "loss": 0.0032, "step": 705 }, { "epoch": 2.165644171779141, "grad_norm": 0.10959838407771286, "learning_rate": 2.1777839056661555e-06, "loss": 0.0026, "step": 706 }, { "epoch": 2.168711656441718, "grad_norm": 0.09379423416997926, "learning_rate": 2.1630672823875207e-06, "loss": 0.0013, "step": 707 }, { "epoch": 2.1717791411042944, "grad_norm": 0.24830346437651182, "learning_rate": 2.148386815325179e-06, "loss": 0.0027, "step": 708 }, { "epoch": 2.1748466257668713, "grad_norm": 0.29344159057905606, "learning_rate": 2.133742691579145e-06, "loss": 0.0009, "step": 709 }, { "epoch": 2.1779141104294477, "grad_norm": 0.30344248949976294, "learning_rate": 2.1191350977862362e-06, "loss": 0.0009, "step": 710 }, { "epoch": 2.1809815950920246, "grad_norm": 0.2217780773980183, "learning_rate": 2.1045642201177075e-06, "loss": 0.0033, "step": 711 }, { "epoch": 2.184049079754601, "grad_norm": 0.07121103642716385, "learning_rate": 2.0900302442768717e-06, "loss": 0.0009, "step": 712 }, { "epoch": 2.187116564417178, "grad_norm": 0.3321155284383894, "learning_rate": 2.075533355496735e-06, "loss": 0.0045, "step": 713 }, { "epoch": 2.190184049079755, "grad_norm": 0.40941768481145857, "learning_rate": 2.061073738537635e-06, "loss": 0.0073, "step": 714 }, { "epoch": 2.1932515337423313, "grad_norm": 0.21360420814522896, "learning_rate": 2.0466515776848857e-06, "loss": 0.004, "step": 715 }, { "epoch": 2.196319018404908, "grad_norm": 0.22144562584721872, "learning_rate": 2.0322670567464304e-06, "loss": 0.0034, "step": 716 }, { "epoch": 2.1993865030674846, "grad_norm": 0.09113523991255253, "learning_rate": 2.0179203590504976e-06, "loss": 0.0008, "step": 717 }, { "epoch": 2.2024539877300615, "grad_norm": 0.21551251871892765, "learning_rate": 2.0036116674432653e-06, "loss": 0.0029, "step": 718 }, { "epoch": 2.205521472392638, "grad_norm": 0.1251050643902028, "learning_rate": 1.9893411642865317e-06, "loss": 0.0008, "step": 719 }, { "epoch": 2.208588957055215, "grad_norm": 0.12424583137608142, "learning_rate": 1.975109031455388e-06, "loss": 0.002, "step": 720 }, { "epoch": 2.2116564417177913, "grad_norm": 0.21560055643150958, "learning_rate": 1.9609154503359024e-06, "loss": 0.0031, "step": 721 }, { "epoch": 2.214723926380368, "grad_norm": 0.2651282013461321, "learning_rate": 1.946760601822809e-06, "loss": 0.007, "step": 722 }, { "epoch": 2.2177914110429446, "grad_norm": 0.20960322018223657, "learning_rate": 1.9326446663172036e-06, "loss": 0.0046, "step": 723 }, { "epoch": 2.2208588957055215, "grad_norm": 0.1515234364008256, "learning_rate": 1.9185678237242373e-06, "loss": 0.001, "step": 724 }, { "epoch": 2.223926380368098, "grad_norm": 0.027001817306581417, "learning_rate": 1.9045302534508298e-06, "loss": 0.0003, "step": 725 }, { "epoch": 2.226993865030675, "grad_norm": 0.3772275262465891, "learning_rate": 1.89053213440339e-06, "loss": 0.0027, "step": 726 }, { "epoch": 2.2300613496932513, "grad_norm": 0.16416633898381086, "learning_rate": 1.8765736449855165e-06, "loss": 0.003, "step": 727 }, { "epoch": 2.233128834355828, "grad_norm": 0.24149851337878356, "learning_rate": 1.8626549630957397e-06, "loss": 0.0043, "step": 728 }, { "epoch": 2.236196319018405, "grad_norm": 0.15235662451937504, "learning_rate": 1.8487762661252557e-06, "loss": 0.0021, "step": 729 }, { "epoch": 2.2392638036809815, "grad_norm": 0.1793857007355656, "learning_rate": 1.8349377309556487e-06, "loss": 0.0032, "step": 730 }, { "epoch": 2.2423312883435584, "grad_norm": 0.1250791648939321, "learning_rate": 1.8211395339566561e-06, "loss": 0.0016, "step": 731 }, { "epoch": 2.245398773006135, "grad_norm": 0.17768178437092705, "learning_rate": 1.8073818509839097e-06, "loss": 0.0022, "step": 732 }, { "epoch": 2.2484662576687118, "grad_norm": 0.36063973591837933, "learning_rate": 1.7936648573766956e-06, "loss": 0.0027, "step": 733 }, { "epoch": 2.2515337423312882, "grad_norm": 0.23032103099080775, "learning_rate": 1.7799887279557238e-06, "loss": 0.0041, "step": 734 }, { "epoch": 2.254601226993865, "grad_norm": 0.44527941027974083, "learning_rate": 1.766353637020895e-06, "loss": 0.0051, "step": 735 }, { "epoch": 2.2576687116564416, "grad_norm": 0.24421891863384448, "learning_rate": 1.7527597583490825e-06, "loss": 0.0019, "step": 736 }, { "epoch": 2.2607361963190185, "grad_norm": 0.17699844104396198, "learning_rate": 1.7392072651919163e-06, "loss": 0.0022, "step": 737 }, { "epoch": 2.263803680981595, "grad_norm": 0.09239369823066165, "learning_rate": 1.7256963302735752e-06, "loss": 0.0011, "step": 738 }, { "epoch": 2.266871165644172, "grad_norm": 0.2896983984065453, "learning_rate": 1.7122271257885853e-06, "loss": 0.0045, "step": 739 }, { "epoch": 2.2699386503067487, "grad_norm": 0.48822736310943793, "learning_rate": 1.6987998233996279e-06, "loss": 0.0029, "step": 740 }, { "epoch": 2.273006134969325, "grad_norm": 0.09163689641484868, "learning_rate": 1.6854145942353417e-06, "loss": 0.0007, "step": 741 }, { "epoch": 2.276073619631902, "grad_norm": 0.08173136485172053, "learning_rate": 1.6720716088881595e-06, "loss": 0.0014, "step": 742 }, { "epoch": 2.2791411042944785, "grad_norm": 0.2545572168918258, "learning_rate": 1.6587710374121203e-06, "loss": 0.0037, "step": 743 }, { "epoch": 2.2822085889570554, "grad_norm": 0.1336031235680743, "learning_rate": 1.645513049320699e-06, "loss": 0.0016, "step": 744 }, { "epoch": 2.285276073619632, "grad_norm": 0.12057113363078809, "learning_rate": 1.6322978135846628e-06, "loss": 0.0015, "step": 745 }, { "epoch": 2.2883435582822087, "grad_norm": 0.4840801469818166, "learning_rate": 1.6191254986299044e-06, "loss": 0.004, "step": 746 }, { "epoch": 2.291411042944785, "grad_norm": 0.07732626484388591, "learning_rate": 1.6059962723352912e-06, "loss": 0.0009, "step": 747 }, { "epoch": 2.294478527607362, "grad_norm": 0.8607829885785011, "learning_rate": 1.5929103020305441e-06, "loss": 0.0088, "step": 748 }, { "epoch": 2.2975460122699385, "grad_norm": 0.12352491257367763, "learning_rate": 1.5798677544940894e-06, "loss": 0.0015, "step": 749 }, { "epoch": 2.3006134969325154, "grad_norm": 0.11613033648370168, "learning_rate": 1.566868795950932e-06, "loss": 0.0018, "step": 750 }, { "epoch": 2.3036809815950923, "grad_norm": 0.12377281598855457, "learning_rate": 1.5539135920705545e-06, "loss": 0.0018, "step": 751 }, { "epoch": 2.3067484662576687, "grad_norm": 0.1545665427345137, "learning_rate": 1.5410023079647823e-06, "loss": 0.0017, "step": 752 }, { "epoch": 2.309815950920245, "grad_norm": 0.17077252050239153, "learning_rate": 1.5281351081856977e-06, "loss": 0.0027, "step": 753 }, { "epoch": 2.312883435582822, "grad_norm": 0.158419257234868, "learning_rate": 1.5153121567235334e-06, "loss": 0.0033, "step": 754 }, { "epoch": 2.315950920245399, "grad_norm": 0.0742824026432101, "learning_rate": 1.5025336170045861e-06, "loss": 0.0005, "step": 755 }, { "epoch": 2.3190184049079754, "grad_norm": 0.34483635411603586, "learning_rate": 1.4897996518891328e-06, "loss": 0.0022, "step": 756 }, { "epoch": 2.3220858895705523, "grad_norm": 0.06803963933434524, "learning_rate": 1.4771104236693524e-06, "loss": 0.0008, "step": 757 }, { "epoch": 2.3251533742331287, "grad_norm": 0.24863338473945598, "learning_rate": 1.4644660940672628e-06, "loss": 0.0026, "step": 758 }, { "epoch": 2.3282208588957056, "grad_norm": 0.11453250446971239, "learning_rate": 1.451866824232655e-06, "loss": 0.0015, "step": 759 }, { "epoch": 2.331288343558282, "grad_norm": 0.14650541046670912, "learning_rate": 1.439312774741042e-06, "loss": 0.003, "step": 760 }, { "epoch": 2.334355828220859, "grad_norm": 0.1648695133277369, "learning_rate": 1.4268041055916104e-06, "loss": 0.0033, "step": 761 }, { "epoch": 2.3374233128834354, "grad_norm": 0.13479759235284597, "learning_rate": 1.4143409762051829e-06, "loss": 0.0024, "step": 762 }, { "epoch": 2.3404907975460123, "grad_norm": 0.2810401943509781, "learning_rate": 1.4019235454221858e-06, "loss": 0.0055, "step": 763 }, { "epoch": 2.3435582822085887, "grad_norm": 0.16871442265684417, "learning_rate": 1.389551971500624e-06, "loss": 0.0015, "step": 764 }, { "epoch": 2.3466257668711656, "grad_norm": 0.05570036580769489, "learning_rate": 1.3772264121140661e-06, "loss": 0.0009, "step": 765 }, { "epoch": 2.3496932515337425, "grad_norm": 0.1683059834869823, "learning_rate": 1.3649470243496327e-06, "loss": 0.0037, "step": 766 }, { "epoch": 2.352760736196319, "grad_norm": 0.17630906022011067, "learning_rate": 1.3527139647059912e-06, "loss": 0.0016, "step": 767 }, { "epoch": 2.355828220858896, "grad_norm": 0.19859929329947976, "learning_rate": 1.340527389091374e-06, "loss": 0.0048, "step": 768 }, { "epoch": 2.3588957055214723, "grad_norm": 0.13967179062236237, "learning_rate": 1.3283874528215735e-06, "loss": 0.0045, "step": 769 }, { "epoch": 2.361963190184049, "grad_norm": 0.0984015071068125, "learning_rate": 1.3162943106179748e-06, "loss": 0.0011, "step": 770 }, { "epoch": 2.3650306748466257, "grad_norm": 0.2649561326989602, "learning_rate": 1.304248116605587e-06, "loss": 0.0034, "step": 771 }, { "epoch": 2.3680981595092025, "grad_norm": 0.20608647696978882, "learning_rate": 1.2922490243110613e-06, "loss": 0.0021, "step": 772 }, { "epoch": 2.371165644171779, "grad_norm": 0.14261322619364292, "learning_rate": 1.2802971866607522e-06, "loss": 0.001, "step": 773 }, { "epoch": 2.374233128834356, "grad_norm": 0.2809342227158972, "learning_rate": 1.2683927559787657e-06, "loss": 0.006, "step": 774 }, { "epoch": 2.3773006134969323, "grad_norm": 0.14128256049100985, "learning_rate": 1.256535883985005e-06, "loss": 0.0022, "step": 775 }, { "epoch": 2.3803680981595092, "grad_norm": 0.17658685558545897, "learning_rate": 1.2447267217932508e-06, "loss": 0.002, "step": 776 }, { "epoch": 2.383435582822086, "grad_norm": 0.16558249370135808, "learning_rate": 1.232965419909235e-06, "loss": 0.0038, "step": 777 }, { "epoch": 2.3865030674846626, "grad_norm": 0.3563889958841285, "learning_rate": 1.2212521282287093e-06, "loss": 0.005, "step": 778 }, { "epoch": 2.3895705521472395, "grad_norm": 0.17260337058148664, "learning_rate": 1.2095869960355494e-06, "loss": 0.003, "step": 779 }, { "epoch": 2.392638036809816, "grad_norm": 0.1256214537945841, "learning_rate": 1.1979701719998454e-06, "loss": 0.0019, "step": 780 }, { "epoch": 2.395705521472393, "grad_norm": 0.20072911308799066, "learning_rate": 1.186401804176009e-06, "loss": 0.0026, "step": 781 }, { "epoch": 2.3987730061349692, "grad_norm": 0.12122821617966971, "learning_rate": 1.1748820400008843e-06, "loss": 0.0014, "step": 782 }, { "epoch": 2.401840490797546, "grad_norm": 0.12587338917377983, "learning_rate": 1.1634110262918718e-06, "loss": 0.0041, "step": 783 }, { "epoch": 2.4049079754601226, "grad_norm": 0.3656333842091313, "learning_rate": 1.1519889092450543e-06, "loss": 0.0018, "step": 784 }, { "epoch": 2.4079754601226995, "grad_norm": 0.17872994295712075, "learning_rate": 1.1406158344333367e-06, "loss": 0.002, "step": 785 }, { "epoch": 2.411042944785276, "grad_norm": 0.1619892096996537, "learning_rate": 1.1292919468045876e-06, "loss": 0.0023, "step": 786 }, { "epoch": 2.414110429447853, "grad_norm": 0.19439110501138088, "learning_rate": 1.1180173906797947e-06, "loss": 0.0025, "step": 787 }, { "epoch": 2.4171779141104293, "grad_norm": 0.10302567991744616, "learning_rate": 1.1067923097512256e-06, "loss": 0.0016, "step": 788 }, { "epoch": 2.420245398773006, "grad_norm": 0.2911971113981604, "learning_rate": 1.0956168470805877e-06, "loss": 0.0069, "step": 789 }, { "epoch": 2.4233128834355826, "grad_norm": 0.06255863079734855, "learning_rate": 1.0844911450972228e-06, "loss": 0.0009, "step": 790 }, { "epoch": 2.4263803680981595, "grad_norm": 0.08073513643263174, "learning_rate": 1.0734153455962765e-06, "loss": 0.0011, "step": 791 }, { "epoch": 2.4294478527607364, "grad_norm": 0.10514179663335815, "learning_rate": 1.0623895897368914e-06, "loss": 0.0015, "step": 792 }, { "epoch": 2.432515337423313, "grad_norm": 0.1666940116725245, "learning_rate": 1.0514140180404202e-06, "loss": 0.003, "step": 793 }, { "epoch": 2.4355828220858897, "grad_norm": 0.3297566953046385, "learning_rate": 1.0404887703886252e-06, "loss": 0.0044, "step": 794 }, { "epoch": 2.438650306748466, "grad_norm": 0.17531363640577552, "learning_rate": 1.0296139860218928e-06, "loss": 0.0024, "step": 795 }, { "epoch": 2.441717791411043, "grad_norm": 0.16868562336567175, "learning_rate": 1.0187898035374683e-06, "loss": 0.0016, "step": 796 }, { "epoch": 2.4447852760736195, "grad_norm": 0.04584091385348708, "learning_rate": 1.008016360887683e-06, "loss": 0.0004, "step": 797 }, { "epoch": 2.4478527607361964, "grad_norm": 0.18449622203743876, "learning_rate": 9.972937953781985e-07, "loss": 0.0042, "step": 798 }, { "epoch": 2.450920245398773, "grad_norm": 0.13526404943688417, "learning_rate": 9.86622243666256e-07, "loss": 0.0016, "step": 799 }, { "epoch": 2.4539877300613497, "grad_norm": 0.25252415406049705, "learning_rate": 9.760018417589333e-07, "loss": 0.0034, "step": 800 }, { "epoch": 2.457055214723926, "grad_norm": 0.215518358226254, "learning_rate": 9.654327250114147e-07, "loss": 0.0037, "step": 801 }, { "epoch": 2.460122699386503, "grad_norm": 0.24277757923937995, "learning_rate": 9.549150281252633e-07, "loss": 0.0023, "step": 802 }, { "epoch": 2.46319018404908, "grad_norm": 0.35989818534447426, "learning_rate": 9.444488851467043e-07, "loss": 0.003, "step": 803 }, { "epoch": 2.4662576687116564, "grad_norm": 0.21107732092261233, "learning_rate": 9.340344294649184e-07, "loss": 0.0015, "step": 804 }, { "epoch": 2.4693251533742333, "grad_norm": 0.16593048874828475, "learning_rate": 9.236717938103396e-07, "loss": 0.0021, "step": 805 }, { "epoch": 2.4723926380368098, "grad_norm": 0.07842110293160545, "learning_rate": 9.133611102529655e-07, "loss": 0.001, "step": 806 }, { "epoch": 2.4754601226993866, "grad_norm": 0.0736410449630101, "learning_rate": 9.031025102006725e-07, "loss": 0.0007, "step": 807 }, { "epoch": 2.478527607361963, "grad_norm": 0.15374285119898826, "learning_rate": 8.928961243975437e-07, "loss": 0.0009, "step": 808 }, { "epoch": 2.48159509202454, "grad_norm": 0.09763968706400328, "learning_rate": 8.827420829221944e-07, "loss": 0.0009, "step": 809 }, { "epoch": 2.4846625766871164, "grad_norm": 0.5363611758679326, "learning_rate": 8.7264051518613e-07, "loss": 0.0054, "step": 810 }, { "epoch": 2.4877300613496933, "grad_norm": 0.09587721531791853, "learning_rate": 8.625915499320819e-07, "loss": 0.001, "step": 811 }, { "epoch": 2.4907975460122698, "grad_norm": 0.07047828098328363, "learning_rate": 8.525953152323685e-07, "loss": 0.0009, "step": 812 }, { "epoch": 2.4938650306748467, "grad_norm": 0.18694536936682596, "learning_rate": 8.426519384872733e-07, "loss": 0.0024, "step": 813 }, { "epoch": 2.4969325153374236, "grad_norm": 0.0805331719571376, "learning_rate": 8.327615464234129e-07, "loss": 0.0006, "step": 814 }, { "epoch": 2.5, "grad_norm": 0.07186649864490034, "learning_rate": 8.229242650921166e-07, "loss": 0.0008, "step": 815 }, { "epoch": 2.5030674846625764, "grad_norm": 0.12548189563671605, "learning_rate": 8.131402198678373e-07, "loss": 0.0027, "step": 816 }, { "epoch": 2.5061349693251533, "grad_norm": 0.2623745485957233, "learning_rate": 8.034095354465337e-07, "loss": 0.0028, "step": 817 }, { "epoch": 2.5092024539877302, "grad_norm": 0.11779343034917025, "learning_rate": 7.937323358440935e-07, "loss": 0.001, "step": 818 }, { "epoch": 2.5122699386503067, "grad_norm": 0.09129107157387073, "learning_rate": 7.841087443947543e-07, "loss": 0.0015, "step": 819 }, { "epoch": 2.5153374233128836, "grad_norm": 0.0603800063286766, "learning_rate": 7.745388837495188e-07, "loss": 0.0006, "step": 820 }, { "epoch": 2.51840490797546, "grad_norm": 0.1546472643676182, "learning_rate": 7.650228758746059e-07, "loss": 0.0024, "step": 821 }, { "epoch": 2.521472392638037, "grad_norm": 0.3094740167503855, "learning_rate": 7.555608420498872e-07, "loss": 0.0051, "step": 822 }, { "epoch": 2.5245398773006134, "grad_norm": 0.28126279193166526, "learning_rate": 7.461529028673465e-07, "loss": 0.0197, "step": 823 }, { "epoch": 2.5276073619631902, "grad_norm": 0.0954026314786444, "learning_rate": 7.367991782295392e-07, "loss": 0.001, "step": 824 }, { "epoch": 2.530674846625767, "grad_norm": 0.08452856435235079, "learning_rate": 7.274997873480666e-07, "loss": 0.0011, "step": 825 }, { "epoch": 2.5337423312883436, "grad_norm": 0.12493971123469698, "learning_rate": 7.182548487420555e-07, "loss": 0.0019, "step": 826 }, { "epoch": 2.53680981595092, "grad_norm": 0.21679833329538098, "learning_rate": 7.090644802366469e-07, "loss": 0.0027, "step": 827 }, { "epoch": 2.539877300613497, "grad_norm": 0.1801329143229757, "learning_rate": 6.999287989614972e-07, "loss": 0.0014, "step": 828 }, { "epoch": 2.542944785276074, "grad_norm": 0.22834022050034936, "learning_rate": 6.908479213492825e-07, "loss": 0.0011, "step": 829 }, { "epoch": 2.5460122699386503, "grad_norm": 0.2595629467564202, "learning_rate": 6.818219631342149e-07, "loss": 0.0039, "step": 830 }, { "epoch": 2.549079754601227, "grad_norm": 0.2836220482178321, "learning_rate": 6.728510393505694e-07, "loss": 0.0026, "step": 831 }, { "epoch": 2.5521472392638036, "grad_norm": 0.06453238892514226, "learning_rate": 6.639352643312164e-07, "loss": 0.0004, "step": 832 }, { "epoch": 2.5552147239263805, "grad_norm": 0.20852970651359395, "learning_rate": 6.550747517061656e-07, "loss": 0.001, "step": 833 }, { "epoch": 2.558282208588957, "grad_norm": 0.2841174821202171, "learning_rate": 6.462696144011149e-07, "loss": 0.0014, "step": 834 }, { "epoch": 2.561349693251534, "grad_norm": 0.2581856484121087, "learning_rate": 6.375199646360142e-07, "loss": 0.0022, "step": 835 }, { "epoch": 2.5644171779141103, "grad_norm": 0.2834383185441278, "learning_rate": 6.28825913923638e-07, "loss": 0.0047, "step": 836 }, { "epoch": 2.567484662576687, "grad_norm": 0.5107731515234254, "learning_rate": 6.201875730681544e-07, "loss": 0.0093, "step": 837 }, { "epoch": 2.5705521472392636, "grad_norm": 0.27824373108650474, "learning_rate": 6.116050521637218e-07, "loss": 0.0049, "step": 838 }, { "epoch": 2.5736196319018405, "grad_norm": 0.2579706167604278, "learning_rate": 6.030784605930856e-07, "loss": 0.0039, "step": 839 }, { "epoch": 2.5766871165644174, "grad_norm": 0.12452116042145692, "learning_rate": 5.946079070261773e-07, "loss": 0.0024, "step": 840 }, { "epoch": 2.579754601226994, "grad_norm": 0.2858948864429028, "learning_rate": 5.861934994187357e-07, "loss": 0.008, "step": 841 }, { "epoch": 2.5828220858895703, "grad_norm": 0.237981754274074, "learning_rate": 5.778353450109286e-07, "loss": 0.0029, "step": 842 }, { "epoch": 2.585889570552147, "grad_norm": 0.02606381443790744, "learning_rate": 5.69533550325988e-07, "loss": 0.0004, "step": 843 }, { "epoch": 2.588957055214724, "grad_norm": 0.14042042098956603, "learning_rate": 5.61288221168848e-07, "loss": 0.0031, "step": 844 }, { "epoch": 2.5920245398773005, "grad_norm": 0.1407516693295058, "learning_rate": 5.530994626248026e-07, "loss": 0.001, "step": 845 }, { "epoch": 2.5950920245398774, "grad_norm": 0.10839729984992401, "learning_rate": 5.449673790581611e-07, "loss": 0.0022, "step": 846 }, { "epoch": 2.598159509202454, "grad_norm": 0.39504221516607024, "learning_rate": 5.368920741109202e-07, "loss": 0.005, "step": 847 }, { "epoch": 2.6012269938650308, "grad_norm": 0.16569517436714892, "learning_rate": 5.288736507014436e-07, "loss": 0.0023, "step": 848 }, { "epoch": 2.604294478527607, "grad_norm": 0.10691366249706033, "learning_rate": 5.209122110231496e-07, "loss": 0.0015, "step": 849 }, { "epoch": 2.607361963190184, "grad_norm": 0.08708747519441029, "learning_rate": 5.130078565432089e-07, "loss": 0.0015, "step": 850 }, { "epoch": 2.610429447852761, "grad_norm": 0.16248278133516086, "learning_rate": 5.051606880012483e-07, "loss": 0.0014, "step": 851 }, { "epoch": 2.6134969325153374, "grad_norm": 0.15115304316641817, "learning_rate": 4.97370805408075e-07, "loss": 0.0014, "step": 852 }, { "epoch": 2.616564417177914, "grad_norm": 0.17910642709470473, "learning_rate": 4.896383080443934e-07, "loss": 0.0019, "step": 853 }, { "epoch": 2.6196319018404908, "grad_norm": 0.10629488236344448, "learning_rate": 4.819632944595415e-07, "loss": 0.0018, "step": 854 }, { "epoch": 2.6226993865030677, "grad_norm": 0.3160945524607155, "learning_rate": 4.7434586247024075e-07, "loss": 0.003, "step": 855 }, { "epoch": 2.625766871165644, "grad_norm": 0.1796109375846772, "learning_rate": 4.667861091593434e-07, "loss": 0.002, "step": 856 }, { "epoch": 2.628834355828221, "grad_norm": 0.09805120856646073, "learning_rate": 4.5928413087459325e-07, "loss": 0.0015, "step": 857 }, { "epoch": 2.6319018404907975, "grad_norm": 0.06512120496498192, "learning_rate": 4.5184002322740784e-07, "loss": 0.0009, "step": 858 }, { "epoch": 2.6349693251533743, "grad_norm": 0.08736046400576912, "learning_rate": 4.444538810916499e-07, "loss": 0.0005, "step": 859 }, { "epoch": 2.638036809815951, "grad_norm": 0.11903062986503751, "learning_rate": 4.371257986024202e-07, "loss": 0.0014, "step": 860 }, { "epoch": 2.6411042944785277, "grad_norm": 0.1738579960227546, "learning_rate": 4.298558691548638e-07, "loss": 0.0039, "step": 861 }, { "epoch": 2.644171779141104, "grad_norm": 0.23024694590361497, "learning_rate": 4.2264418540297e-07, "loss": 0.0042, "step": 862 }, { "epoch": 2.647239263803681, "grad_norm": 0.09629140238564009, "learning_rate": 4.1549083925840174e-07, "loss": 0.0007, "step": 863 }, { "epoch": 2.6503067484662575, "grad_norm": 0.17847211521709486, "learning_rate": 4.083959218893158e-07, "loss": 0.0021, "step": 864 }, { "epoch": 2.6533742331288344, "grad_norm": 0.14244846012513432, "learning_rate": 4.0135952371920636e-07, "loss": 0.0018, "step": 865 }, { "epoch": 2.6564417177914113, "grad_norm": 1.1152167570510674, "learning_rate": 3.9438173442575e-07, "loss": 0.0059, "step": 866 }, { "epoch": 2.6595092024539877, "grad_norm": 0.07020182370714435, "learning_rate": 3.874626429396622e-07, "loss": 0.0009, "step": 867 }, { "epoch": 2.662576687116564, "grad_norm": 0.18577423108935284, "learning_rate": 3.8060233744356634e-07, "loss": 0.0062, "step": 868 }, { "epoch": 2.665644171779141, "grad_norm": 0.4435684511619857, "learning_rate": 3.7380090537086853e-07, "loss": 0.0022, "step": 869 }, { "epoch": 2.668711656441718, "grad_norm": 0.2716721122126474, "learning_rate": 3.6705843340464287e-07, "loss": 0.0211, "step": 870 }, { "epoch": 2.6717791411042944, "grad_norm": 0.1978025404729548, "learning_rate": 3.6037500747652665e-07, "loss": 0.0011, "step": 871 }, { "epoch": 2.6748466257668713, "grad_norm": 0.2832440687817866, "learning_rate": 3.53750712765627e-07, "loss": 0.0025, "step": 872 }, { "epoch": 2.6779141104294477, "grad_norm": 0.24595271118155285, "learning_rate": 3.4718563369743217e-07, "loss": 0.0058, "step": 873 }, { "epoch": 2.6809815950920246, "grad_norm": 0.09084236194044377, "learning_rate": 3.406798539427386e-07, "loss": 0.0013, "step": 874 }, { "epoch": 2.684049079754601, "grad_norm": 0.06503950468432222, "learning_rate": 3.3423345641658275e-07, "loss": 0.0013, "step": 875 }, { "epoch": 2.687116564417178, "grad_norm": 0.3314300890265219, "learning_rate": 3.2784652327718544e-07, "loss": 0.0031, "step": 876 }, { "epoch": 2.690184049079755, "grad_norm": 0.1396336301171356, "learning_rate": 3.2151913592490256e-07, "loss": 0.0034, "step": 877 }, { "epoch": 2.6932515337423313, "grad_norm": 0.5653766171978473, "learning_rate": 3.1525137500119207e-07, "loss": 0.0045, "step": 878 }, { "epoch": 2.6963190184049077, "grad_norm": 0.10599295362951303, "learning_rate": 3.0904332038757977e-07, "loss": 0.0013, "step": 879 }, { "epoch": 2.6993865030674846, "grad_norm": 0.23406302898617593, "learning_rate": 3.0289505120464743e-07, "loss": 0.0021, "step": 880 }, { "epoch": 2.7024539877300615, "grad_norm": 0.06314907207577944, "learning_rate": 2.96806645811023e-07, "loss": 0.0012, "step": 881 }, { "epoch": 2.705521472392638, "grad_norm": 0.2275280025600531, "learning_rate": 2.9077818180237693e-07, "loss": 0.0062, "step": 882 }, { "epoch": 2.708588957055215, "grad_norm": 0.21241533924947312, "learning_rate": 2.848097360104396e-07, "loss": 0.0044, "step": 883 }, { "epoch": 2.7116564417177913, "grad_norm": 0.31301496464316575, "learning_rate": 2.7890138450202053e-07, "loss": 0.0059, "step": 884 }, { "epoch": 2.714723926380368, "grad_norm": 0.12241935449577525, "learning_rate": 2.7305320257803524e-07, "loss": 0.0014, "step": 885 }, { "epoch": 2.7177914110429446, "grad_norm": 0.3673718302258716, "learning_rate": 2.6726526477254986e-07, "loss": 0.0039, "step": 886 }, { "epoch": 2.7208588957055215, "grad_norm": 0.2702232084402326, "learning_rate": 2.615376448518314e-07, "loss": 0.0037, "step": 887 }, { "epoch": 2.7239263803680984, "grad_norm": 0.4714333984588122, "learning_rate": 2.5587041581340235e-07, "loss": 0.004, "step": 888 }, { "epoch": 2.726993865030675, "grad_norm": 0.07117338469739014, "learning_rate": 2.502636498851152e-07, "loss": 0.0009, "step": 889 }, { "epoch": 2.7300613496932513, "grad_norm": 0.13772813511574597, "learning_rate": 2.447174185242324e-07, "loss": 0.0007, "step": 890 }, { "epoch": 2.733128834355828, "grad_norm": 0.14647291454327963, "learning_rate": 2.392317924165116e-07, "loss": 0.0029, "step": 891 }, { "epoch": 2.736196319018405, "grad_norm": 0.1033756801336396, "learning_rate": 2.338068414753075e-07, "loss": 0.0034, "step": 892 }, { "epoch": 2.7392638036809815, "grad_norm": 0.13310767902357046, "learning_rate": 2.2844263484068097e-07, "loss": 0.0017, "step": 893 }, { "epoch": 2.7423312883435584, "grad_norm": 0.11225484608545211, "learning_rate": 2.2313924087851657e-07, "loss": 0.0016, "step": 894 }, { "epoch": 2.745398773006135, "grad_norm": 0.1581141376006445, "learning_rate": 2.1789672717965227e-07, "loss": 0.0036, "step": 895 }, { "epoch": 2.7484662576687118, "grad_norm": 0.13380182051018658, "learning_rate": 2.1271516055901774e-07, "loss": 0.0015, "step": 896 }, { "epoch": 2.7515337423312882, "grad_norm": 0.30152265206371165, "learning_rate": 2.0759460705478186e-07, "loss": 0.0046, "step": 897 }, { "epoch": 2.754601226993865, "grad_norm": 0.2248407912430735, "learning_rate": 2.0253513192751374e-07, "loss": 0.0055, "step": 898 }, { "epoch": 2.7576687116564416, "grad_norm": 0.2815675146560431, "learning_rate": 1.9753679965934526e-07, "loss": 0.004, "step": 899 }, { "epoch": 2.7607361963190185, "grad_norm": 0.32432774333312003, "learning_rate": 1.925996739531577e-07, "loss": 0.0016, "step": 900 }, { "epoch": 2.763803680981595, "grad_norm": 0.12189177547164826, "learning_rate": 1.8772381773176417e-07, "loss": 0.0019, "step": 901 }, { "epoch": 2.766871165644172, "grad_norm": 0.16085814495182016, "learning_rate": 1.8290929313710514e-07, "loss": 0.0018, "step": 902 }, { "epoch": 2.7699386503067487, "grad_norm": 0.23030017813530398, "learning_rate": 1.7815616152946523e-07, "loss": 0.0033, "step": 903 }, { "epoch": 2.773006134969325, "grad_norm": 0.2560418398994958, "learning_rate": 1.7346448348668443e-07, "loss": 0.0031, "step": 904 }, { "epoch": 2.7760736196319016, "grad_norm": 0.17074392845672137, "learning_rate": 1.6883431880338473e-07, "loss": 0.001, "step": 905 }, { "epoch": 2.7791411042944785, "grad_norm": 0.2075421386457064, "learning_rate": 1.6426572649021477e-07, "loss": 0.003, "step": 906 }, { "epoch": 2.7822085889570554, "grad_norm": 0.2904212772947595, "learning_rate": 1.5975876477309193e-07, "loss": 0.0031, "step": 907 }, { "epoch": 2.785276073619632, "grad_norm": 0.1008864192169771, "learning_rate": 1.5531349109246364e-07, "loss": 0.0015, "step": 908 }, { "epoch": 2.7883435582822087, "grad_norm": 0.10209292364608329, "learning_rate": 1.5092996210257281e-07, "loss": 0.0004, "step": 909 }, { "epoch": 2.791411042944785, "grad_norm": 0.12376916447693681, "learning_rate": 1.466082336707375e-07, "loss": 0.0022, "step": 910 }, { "epoch": 2.794478527607362, "grad_norm": 0.25821128489268347, "learning_rate": 1.423483608766385e-07, "loss": 0.0029, "step": 911 }, { "epoch": 2.7975460122699385, "grad_norm": 0.1579943616682322, "learning_rate": 1.3815039801161723e-07, "loss": 0.0036, "step": 912 }, { "epoch": 2.8006134969325154, "grad_norm": 0.3539618179713858, "learning_rate": 1.3401439857798292e-07, "loss": 0.005, "step": 913 }, { "epoch": 2.8036809815950923, "grad_norm": 0.11487729898070202, "learning_rate": 1.2994041528833267e-07, "loss": 0.001, "step": 914 }, { "epoch": 2.8067484662576687, "grad_norm": 0.20363041791420816, "learning_rate": 1.2592850006487801e-07, "loss": 0.0046, "step": 915 }, { "epoch": 2.809815950920245, "grad_norm": 0.18459789759276166, "learning_rate": 1.2197870403878375e-07, "loss": 0.0017, "step": 916 }, { "epoch": 2.812883435582822, "grad_norm": 0.2409014981906113, "learning_rate": 1.1809107754951643e-07, "loss": 0.0019, "step": 917 }, { "epoch": 2.815950920245399, "grad_norm": 0.13585872733436774, "learning_rate": 1.1426567014420297e-07, "loss": 0.0018, "step": 918 }, { "epoch": 2.8190184049079754, "grad_norm": 0.11734644264645676, "learning_rate": 1.105025305769969e-07, "loss": 0.0021, "step": 919 }, { "epoch": 2.8220858895705523, "grad_norm": 0.15393567968740215, "learning_rate": 1.0680170680846259e-07, "loss": 0.0035, "step": 920 }, { "epoch": 2.8251533742331287, "grad_norm": 0.5560090616250628, "learning_rate": 1.0316324600495697e-07, "loss": 0.0476, "step": 921 }, { "epoch": 2.8282208588957056, "grad_norm": 0.09066848667102294, "learning_rate": 9.958719453803278e-08, "loss": 0.0008, "step": 922 }, { "epoch": 2.831288343558282, "grad_norm": 0.10536567544481347, "learning_rate": 9.607359798384785e-08, "loss": 0.0015, "step": 923 }, { "epoch": 2.834355828220859, "grad_norm": 0.12467104732377465, "learning_rate": 9.26225011225812e-08, "loss": 0.001, "step": 924 }, { "epoch": 2.837423312883436, "grad_norm": 0.2808585290483472, "learning_rate": 8.923394793786455e-08, "loss": 0.0018, "step": 925 }, { "epoch": 2.8404907975460123, "grad_norm": 0.27131651565465975, "learning_rate": 8.590798161622227e-08, "loss": 0.0032, "step": 926 }, { "epoch": 2.8435582822085887, "grad_norm": 0.23536713230244996, "learning_rate": 8.264464454651843e-08, "loss": 0.0023, "step": 927 }, { "epoch": 2.8466257668711656, "grad_norm": 0.3065758572901688, "learning_rate": 7.944397831941952e-08, "loss": 0.0056, "step": 928 }, { "epoch": 2.8496932515337425, "grad_norm": 0.0762524485953778, "learning_rate": 7.630602372686258e-08, "loss": 0.0017, "step": 929 }, { "epoch": 2.852760736196319, "grad_norm": 0.23912997282686557, "learning_rate": 7.32308207615351e-08, "loss": 0.003, "step": 930 }, { "epoch": 2.855828220858896, "grad_norm": 0.17423640557655332, "learning_rate": 7.021840861636652e-08, "loss": 0.0017, "step": 931 }, { "epoch": 2.8588957055214723, "grad_norm": 0.1868208824915922, "learning_rate": 6.72688256840287e-08, "loss": 0.0027, "step": 932 }, { "epoch": 2.861963190184049, "grad_norm": 0.1894809084876684, "learning_rate": 6.438210955644453e-08, "loss": 0.0009, "step": 933 }, { "epoch": 2.8650306748466257, "grad_norm": 0.46134304784124597, "learning_rate": 6.15582970243117e-08, "loss": 0.0029, "step": 934 }, { "epoch": 2.8680981595092025, "grad_norm": 0.4497782763650271, "learning_rate": 5.879742407663147e-08, "loss": 0.0028, "step": 935 }, { "epoch": 2.871165644171779, "grad_norm": 0.047072709464050136, "learning_rate": 5.609952590025225e-08, "loss": 0.0005, "step": 936 }, { "epoch": 2.874233128834356, "grad_norm": 0.07165859203606187, "learning_rate": 5.3464636879419496e-08, "loss": 0.001, "step": 937 }, { "epoch": 2.8773006134969323, "grad_norm": 0.09661223755215212, "learning_rate": 5.089279059533658e-08, "loss": 0.0023, "step": 938 }, { "epoch": 2.8803680981595092, "grad_norm": 0.09171222486373179, "learning_rate": 4.8384019825739036e-08, "loss": 0.0009, "step": 939 }, { "epoch": 2.883435582822086, "grad_norm": 0.2299910781565878, "learning_rate": 4.593835654447709e-08, "loss": 0.0049, "step": 940 }, { "epoch": 2.8865030674846626, "grad_norm": 0.09951233304726008, "learning_rate": 4.3555831921104905e-08, "loss": 0.0004, "step": 941 }, { "epoch": 2.889570552147239, "grad_norm": 0.13293849716173955, "learning_rate": 4.123647632048644e-08, "loss": 0.0019, "step": 942 }, { "epoch": 2.892638036809816, "grad_norm": 0.17984529573061198, "learning_rate": 3.8980319302407976e-08, "loss": 0.0022, "step": 943 }, { "epoch": 2.895705521472393, "grad_norm": 0.23956821016046442, "learning_rate": 3.678738962119899e-08, "loss": 0.0032, "step": 944 }, { "epoch": 2.8987730061349692, "grad_norm": 0.1195679361937312, "learning_rate": 3.465771522536854e-08, "loss": 0.0014, "step": 945 }, { "epoch": 2.901840490797546, "grad_norm": 0.06300492295803174, "learning_rate": 3.25913232572489e-08, "loss": 0.0003, "step": 946 }, { "epoch": 2.9049079754601226, "grad_norm": 0.1186283025014797, "learning_rate": 3.058824005264805e-08, "loss": 0.0008, "step": 947 }, { "epoch": 2.9079754601226995, "grad_norm": 0.2913471755472139, "learning_rate": 2.8648491140513267e-08, "loss": 0.0027, "step": 948 }, { "epoch": 2.911042944785276, "grad_norm": 0.37708500248147947, "learning_rate": 2.6772101242610316e-08, "loss": 0.0033, "step": 949 }, { "epoch": 2.914110429447853, "grad_norm": 0.10203749592807297, "learning_rate": 2.495909427320198e-08, "loss": 0.0011, "step": 950 }, { "epoch": 2.9171779141104297, "grad_norm": 0.04844387099370729, "learning_rate": 2.3209493338749444e-08, "loss": 0.0008, "step": 951 }, { "epoch": 2.920245398773006, "grad_norm": 0.14844116201176166, "learning_rate": 2.1523320737613095e-08, "loss": 0.0023, "step": 952 }, { "epoch": 2.9233128834355826, "grad_norm": 0.06648586707215681, "learning_rate": 1.9900597959770506e-08, "loss": 0.001, "step": 953 }, { "epoch": 2.9263803680981595, "grad_norm": 0.06323384833319899, "learning_rate": 1.834134568654333e-08, "loss": 0.0009, "step": 954 }, { "epoch": 2.9294478527607364, "grad_norm": 0.21617483164150297, "learning_rate": 1.6845583790330855e-08, "loss": 0.0053, "step": 955 }, { "epoch": 2.932515337423313, "grad_norm": 0.46122917062522734, "learning_rate": 1.541333133436018e-08, "loss": 0.0061, "step": 956 }, { "epoch": 2.9355828220858897, "grad_norm": 0.03574910339123186, "learning_rate": 1.404460657244089e-08, "loss": 0.0002, "step": 957 }, { "epoch": 2.938650306748466, "grad_norm": 0.39754534061164976, "learning_rate": 1.2739426948732426e-08, "loss": 0.0057, "step": 958 }, { "epoch": 2.941717791411043, "grad_norm": 0.22679275982144256, "learning_rate": 1.1497809097523737e-08, "loss": 0.0018, "step": 959 }, { "epoch": 2.9447852760736195, "grad_norm": 0.320664640430231, "learning_rate": 1.0319768843018995e-08, "loss": 0.0038, "step": 960 }, { "epoch": 2.9478527607361964, "grad_norm": 0.2519872312593028, "learning_rate": 9.20532119913775e-09, "loss": 0.0065, "step": 961 }, { "epoch": 2.950920245398773, "grad_norm": 0.11284270513200542, "learning_rate": 8.15448036932176e-09, "loss": 0.0006, "step": 962 }, { "epoch": 2.9539877300613497, "grad_norm": 0.10208352343863136, "learning_rate": 7.167259746355681e-09, "loss": 0.0011, "step": 963 }, { "epoch": 2.957055214723926, "grad_norm": 0.25701217063452914, "learning_rate": 6.243671912194993e-09, "loss": 0.002, "step": 964 }, { "epoch": 2.960122699386503, "grad_norm": 0.0764572908354291, "learning_rate": 5.383728637807784e-09, "loss": 0.0015, "step": 965 }, { "epoch": 2.96319018404908, "grad_norm": 0.07471945840486285, "learning_rate": 4.587440883021543e-09, "loss": 0.0008, "step": 966 }, { "epoch": 2.9662576687116564, "grad_norm": 0.2099586583717868, "learning_rate": 3.854818796385495e-09, "loss": 0.0045, "step": 967 }, { "epoch": 2.969325153374233, "grad_norm": 0.24637085011322238, "learning_rate": 3.1858717150412554e-09, "loss": 0.0019, "step": 968 }, { "epoch": 2.9723926380368098, "grad_norm": 0.05193551709164074, "learning_rate": 2.580608164604592e-09, "loss": 0.0003, "step": 969 }, { "epoch": 2.9754601226993866, "grad_norm": 0.04547774488809871, "learning_rate": 2.0390358590538507e-09, "loss": 0.0004, "step": 970 }, { "epoch": 2.978527607361963, "grad_norm": 0.7184876930985127, "learning_rate": 1.5611617006344725e-09, "loss": 0.006, "step": 971 }, { "epoch": 2.98159509202454, "grad_norm": 0.14243992488733678, "learning_rate": 1.1469917797696238e-09, "loss": 0.0017, "step": 972 }, { "epoch": 2.9846625766871164, "grad_norm": 0.16006164134574483, "learning_rate": 7.96531374983589e-10, "loss": 0.0043, "step": 973 }, { "epoch": 2.9877300613496933, "grad_norm": 0.09809638018672716, "learning_rate": 5.09784952833492e-10, "loss": 0.0016, "step": 974 }, { "epoch": 2.9907975460122698, "grad_norm": 0.2412737011340249, "learning_rate": 2.8675616785212023e-10, "loss": 0.0062, "step": 975 }, { "epoch": 2.9938650306748467, "grad_norm": 0.17763972988409896, "learning_rate": 1.2744786250407092e-10, "loss": 0.0017, "step": 976 }, { "epoch": 2.9969325153374236, "grad_norm": 0.21922904263222162, "learning_rate": 3.186206714522744e-11, "loss": 0.0036, "step": 977 }, { "epoch": 3.0, "grad_norm": 0.04875957060132452, "learning_rate": 0.0, "loss": 0.0005, "step": 978 }, { "epoch": 3.0, "step": 978, "total_flos": 34026430169088.0, "train_loss": 0.014223355380586132, "train_runtime": 5099.3334, "train_samples_per_second": 6.131, "train_steps_per_second": 0.192 } ], "logging_steps": 1, "max_steps": 978, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 783, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 34026430169088.0, "train_batch_size": 4, "trial_name": null, "trial_params": null }