{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 2181, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0004585052728106373, "grad_norm": 89.6312026977539, "learning_rate": 1.5151515151515153e-05, "loss": 8.9079, "step": 1 }, { "epoch": 0.0009170105456212746, "grad_norm": 90.3139877319336, "learning_rate": 3.0303030303030306e-05, "loss": 8.9139, "step": 2 }, { "epoch": 0.001375515818431912, "grad_norm": 71.63379669189453, "learning_rate": 4.545454545454546e-05, "loss": 8.2583, "step": 3 }, { "epoch": 0.0018340210912425492, "grad_norm": 21.473169326782227, "learning_rate": 6.060606060606061e-05, "loss": 6.5858, "step": 4 }, { "epoch": 0.0022925263640531865, "grad_norm": 16.133493423461914, "learning_rate": 7.575757575757576e-05, "loss": 6.1093, "step": 5 }, { "epoch": 0.002751031636863824, "grad_norm": 13.74282455444336, "learning_rate": 9.090909090909092e-05, "loss": 5.6927, "step": 6 }, { "epoch": 0.0032095369096744614, "grad_norm": 9.239209175109863, "learning_rate": 0.00010606060606060606, "loss": 5.287, "step": 7 }, { "epoch": 0.0036680421824850985, "grad_norm": 6.510136604309082, "learning_rate": 0.00012121212121212122, "loss": 4.8858, "step": 8 }, { "epoch": 0.0041265474552957355, "grad_norm": 4.54268741607666, "learning_rate": 0.00013636363636363637, "loss": 4.6652, "step": 9 }, { "epoch": 0.004585052728106373, "grad_norm": 3.5991032123565674, "learning_rate": 0.00015151515151515152, "loss": 4.5218, "step": 10 }, { "epoch": 0.0050435580009170105, "grad_norm": 3.2578372955322266, "learning_rate": 0.00016666666666666666, "loss": 4.3768, "step": 11 }, { "epoch": 0.005502063273727648, "grad_norm": 2.957754135131836, "learning_rate": 0.00018181818181818183, "loss": 4.4946, "step": 12 }, { "epoch": 0.005960568546538285, "grad_norm": 2.702655076980591, "learning_rate": 0.00019696969696969695, "loss": 4.4287, "step": 13 }, { "epoch": 0.006419073819348923, "grad_norm": 2.379763603210449, "learning_rate": 0.00021212121212121213, "loss": 4.3313, "step": 14 }, { "epoch": 0.0068775790921595595, "grad_norm": 1.9976245164871216, "learning_rate": 0.00022727272727272727, "loss": 4.2346, "step": 15 }, { "epoch": 0.007336084364970197, "grad_norm": 1.6521345376968384, "learning_rate": 0.00024242424242424245, "loss": 4.2122, "step": 16 }, { "epoch": 0.007794589637780834, "grad_norm": 1.2080203294754028, "learning_rate": 0.00025757575757575756, "loss": 4.0583, "step": 17 }, { "epoch": 0.008253094910591471, "grad_norm": 1.0779513120651245, "learning_rate": 0.00027272727272727274, "loss": 4.0719, "step": 18 }, { "epoch": 0.00871160018340211, "grad_norm": 1.0318020582199097, "learning_rate": 0.0002878787878787879, "loss": 4.0886, "step": 19 }, { "epoch": 0.009170105456212746, "grad_norm": 0.860767662525177, "learning_rate": 0.00030303030303030303, "loss": 4.2221, "step": 20 }, { "epoch": 0.009628610729023384, "grad_norm": 0.7610700726509094, "learning_rate": 0.0003181818181818182, "loss": 4.2007, "step": 21 }, { "epoch": 0.010087116001834021, "grad_norm": 0.9166486859321594, "learning_rate": 0.0003333333333333333, "loss": 4.0741, "step": 22 }, { "epoch": 0.01054562127464466, "grad_norm": 0.8143223524093628, "learning_rate": 0.0003484848484848485, "loss": 3.9995, "step": 23 }, { "epoch": 0.011004126547455296, "grad_norm": 0.707416296005249, "learning_rate": 0.00036363636363636367, "loss": 4.0924, "step": 24 }, { "epoch": 0.011462631820265932, "grad_norm": 0.7235525846481323, "learning_rate": 0.0003787878787878788, "loss": 4.0107, "step": 25 }, { "epoch": 0.01192113709307657, "grad_norm": 0.6550369262695312, "learning_rate": 0.0003939393939393939, "loss": 3.9407, "step": 26 }, { "epoch": 0.012379642365887207, "grad_norm": 0.6669556498527527, "learning_rate": 0.00040909090909090913, "loss": 4.0254, "step": 27 }, { "epoch": 0.012838147638697846, "grad_norm": 0.6668184995651245, "learning_rate": 0.00042424242424242425, "loss": 3.9598, "step": 28 }, { "epoch": 0.013296652911508482, "grad_norm": 0.6219844222068787, "learning_rate": 0.0004393939393939394, "loss": 4.0614, "step": 29 }, { "epoch": 0.013755158184319119, "grad_norm": 0.5886954069137573, "learning_rate": 0.00045454545454545455, "loss": 3.9803, "step": 30 }, { "epoch": 0.014213663457129757, "grad_norm": 0.5552184581756592, "learning_rate": 0.0004696969696969697, "loss": 3.9302, "step": 31 }, { "epoch": 0.014672168729940394, "grad_norm": 0.4866012930870056, "learning_rate": 0.0004848484848484849, "loss": 3.9852, "step": 32 }, { "epoch": 0.015130674002751032, "grad_norm": 0.5210157632827759, "learning_rate": 0.0005, "loss": 3.9284, "step": 33 }, { "epoch": 0.015589179275561669, "grad_norm": 0.4954196512699127, "learning_rate": 0.0005151515151515151, "loss": 3.9632, "step": 34 }, { "epoch": 0.016047684548372305, "grad_norm": 0.5996412634849548, "learning_rate": 0.0005303030303030302, "loss": 3.8995, "step": 35 }, { "epoch": 0.016506189821182942, "grad_norm": 0.5500408411026001, "learning_rate": 0.0005454545454545455, "loss": 3.7987, "step": 36 }, { "epoch": 0.016964695093993582, "grad_norm": 0.5862100124359131, "learning_rate": 0.0005606060606060606, "loss": 3.8804, "step": 37 }, { "epoch": 0.01742320036680422, "grad_norm": 0.5667603015899658, "learning_rate": 0.0005757575757575758, "loss": 3.8612, "step": 38 }, { "epoch": 0.017881705639614855, "grad_norm": 0.5688158273696899, "learning_rate": 0.0005909090909090909, "loss": 3.8771, "step": 39 }, { "epoch": 0.018340210912425492, "grad_norm": 0.548868715763092, "learning_rate": 0.0006060606060606061, "loss": 3.8187, "step": 40 }, { "epoch": 0.01879871618523613, "grad_norm": 0.6249773502349854, "learning_rate": 0.0006212121212121212, "loss": 3.7284, "step": 41 }, { "epoch": 0.01925722145804677, "grad_norm": 0.6058098673820496, "learning_rate": 0.0006363636363636364, "loss": 3.7446, "step": 42 }, { "epoch": 0.019715726730857405, "grad_norm": 0.6619426012039185, "learning_rate": 0.0006515151515151515, "loss": 3.8003, "step": 43 }, { "epoch": 0.020174232003668042, "grad_norm": 0.7826529145240784, "learning_rate": 0.0006666666666666666, "loss": 3.6795, "step": 44 }, { "epoch": 0.02063273727647868, "grad_norm": 0.7745411396026611, "learning_rate": 0.0006818181818181818, "loss": 3.6255, "step": 45 }, { "epoch": 0.02109124254928932, "grad_norm": 0.7803659439086914, "learning_rate": 0.000696969696969697, "loss": 3.723, "step": 46 }, { "epoch": 0.021549747822099955, "grad_norm": 0.6924562454223633, "learning_rate": 0.0007121212121212122, "loss": 3.605, "step": 47 }, { "epoch": 0.02200825309491059, "grad_norm": 0.7772483825683594, "learning_rate": 0.0007272727272727273, "loss": 3.5976, "step": 48 }, { "epoch": 0.02246675836772123, "grad_norm": 0.7926068902015686, "learning_rate": 0.0007424242424242425, "loss": 3.7015, "step": 49 }, { "epoch": 0.022925263640531865, "grad_norm": 0.8191553950309753, "learning_rate": 0.0007575757575757576, "loss": 3.6405, "step": 50 }, { "epoch": 0.023383768913342505, "grad_norm": 0.8465298414230347, "learning_rate": 0.0007727272727272727, "loss": 3.5958, "step": 51 }, { "epoch": 0.02384227418615314, "grad_norm": 0.7837768197059631, "learning_rate": 0.0007878787878787878, "loss": 3.6304, "step": 52 }, { "epoch": 0.024300779458963778, "grad_norm": 0.8417014479637146, "learning_rate": 0.000803030303030303, "loss": 3.643, "step": 53 }, { "epoch": 0.024759284731774415, "grad_norm": 0.7814474105834961, "learning_rate": 0.0008181818181818183, "loss": 3.5402, "step": 54 }, { "epoch": 0.02521779000458505, "grad_norm": 0.8083184361457825, "learning_rate": 0.0008333333333333334, "loss": 3.5062, "step": 55 }, { "epoch": 0.02567629527739569, "grad_norm": 0.795933187007904, "learning_rate": 0.0008484848484848485, "loss": 3.5752, "step": 56 }, { "epoch": 0.026134800550206328, "grad_norm": 0.7559511661529541, "learning_rate": 0.0008636363636363636, "loss": 3.4936, "step": 57 }, { "epoch": 0.026593305823016965, "grad_norm": 0.8491300940513611, "learning_rate": 0.0008787878787878789, "loss": 3.6088, "step": 58 }, { "epoch": 0.0270518110958276, "grad_norm": 0.8369011282920837, "learning_rate": 0.000893939393939394, "loss": 3.4466, "step": 59 }, { "epoch": 0.027510316368638238, "grad_norm": 0.8128252625465393, "learning_rate": 0.0009090909090909091, "loss": 3.4184, "step": 60 }, { "epoch": 0.027968821641448878, "grad_norm": 0.8722922205924988, "learning_rate": 0.0009242424242424242, "loss": 3.3046, "step": 61 }, { "epoch": 0.028427326914259515, "grad_norm": 0.8812612295150757, "learning_rate": 0.0009393939393939394, "loss": 3.404, "step": 62 }, { "epoch": 0.02888583218707015, "grad_norm": 0.829634964466095, "learning_rate": 0.0009545454545454546, "loss": 3.3831, "step": 63 }, { "epoch": 0.029344337459880788, "grad_norm": 1.0163425207138062, "learning_rate": 0.0009696969696969698, "loss": 3.4588, "step": 64 }, { "epoch": 0.029802842732691424, "grad_norm": 0.8942959904670715, "learning_rate": 0.000984848484848485, "loss": 3.2784, "step": 65 }, { "epoch": 0.030261348005502064, "grad_norm": 0.8079606890678406, "learning_rate": 0.001, "loss": 3.3797, "step": 66 }, { "epoch": 0.0307198532783127, "grad_norm": 0.8794964551925659, "learning_rate": 0.0009999994484067654, "loss": 3.372, "step": 67 }, { "epoch": 0.031178358551123338, "grad_norm": 0.8519138097763062, "learning_rate": 0.0009999977936282788, "loss": 3.2568, "step": 68 }, { "epoch": 0.03163686382393398, "grad_norm": 0.8422295451164246, "learning_rate": 0.0009999950356681913, "loss": 3.2995, "step": 69 }, { "epoch": 0.03209536909674461, "grad_norm": 0.8314495086669922, "learning_rate": 0.0009999911745325876, "loss": 3.2213, "step": 70 }, { "epoch": 0.03255387436955525, "grad_norm": 0.8072673678398132, "learning_rate": 0.0009999862102299873, "loss": 3.2592, "step": 71 }, { "epoch": 0.033012379642365884, "grad_norm": 0.9119292497634888, "learning_rate": 0.0009999801427713433, "loss": 3.2598, "step": 72 }, { "epoch": 0.033470884915176524, "grad_norm": 0.9167115092277527, "learning_rate": 0.0009999729721700424, "loss": 3.1282, "step": 73 }, { "epoch": 0.033929390187987164, "grad_norm": 0.8548741936683655, "learning_rate": 0.000999964698441906, "loss": 3.1922, "step": 74 }, { "epoch": 0.0343878954607978, "grad_norm": 0.9006883502006531, "learning_rate": 0.0009999553216051892, "loss": 3.2273, "step": 75 }, { "epoch": 0.03484640073360844, "grad_norm": 0.811022937297821, "learning_rate": 0.00099994484168058, "loss": 3.148, "step": 76 }, { "epoch": 0.03530490600641907, "grad_norm": 0.7904940843582153, "learning_rate": 0.0009999332586912019, "loss": 3.1502, "step": 77 }, { "epoch": 0.03576341127922971, "grad_norm": 0.8493460416793823, "learning_rate": 0.0009999205726626108, "loss": 3.1535, "step": 78 }, { "epoch": 0.03622191655204035, "grad_norm": 0.8164473176002502, "learning_rate": 0.000999906783622797, "loss": 3.1067, "step": 79 }, { "epoch": 0.036680421824850984, "grad_norm": 0.9152378439903259, "learning_rate": 0.0009998918916021842, "loss": 3.1901, "step": 80 }, { "epoch": 0.037138927097661624, "grad_norm": 0.8219537734985352, "learning_rate": 0.0009998758966336297, "loss": 3.0321, "step": 81 }, { "epoch": 0.03759743237047226, "grad_norm": 0.8656907677650452, "learning_rate": 0.0009998587987524242, "loss": 3.0422, "step": 82 }, { "epoch": 0.0380559376432829, "grad_norm": 0.9810438752174377, "learning_rate": 0.0009998405979962926, "loss": 3.099, "step": 83 }, { "epoch": 0.03851444291609354, "grad_norm": 0.8791067004203796, "learning_rate": 0.000999821294405392, "loss": 3.1102, "step": 84 }, { "epoch": 0.03897294818890417, "grad_norm": 1.073812484741211, "learning_rate": 0.0009998008880223134, "loss": 3.1489, "step": 85 }, { "epoch": 0.03943145346171481, "grad_norm": 0.9411929845809937, "learning_rate": 0.000999779378892081, "loss": 3.0799, "step": 86 }, { "epoch": 0.039889958734525444, "grad_norm": 0.8400881290435791, "learning_rate": 0.0009997567670621522, "loss": 2.9361, "step": 87 }, { "epoch": 0.040348464007336084, "grad_norm": 1.0118547677993774, "learning_rate": 0.0009997330525824165, "loss": 3.0461, "step": 88 }, { "epoch": 0.040806969280146724, "grad_norm": 0.8789583444595337, "learning_rate": 0.0009997082355051976, "loss": 3.0606, "step": 89 }, { "epoch": 0.04126547455295736, "grad_norm": 0.8469064235687256, "learning_rate": 0.000999682315885251, "loss": 3.1012, "step": 90 }, { "epoch": 0.041723979825768, "grad_norm": 1.022143840789795, "learning_rate": 0.0009996552937797645, "loss": 2.9711, "step": 91 }, { "epoch": 0.04218248509857864, "grad_norm": 0.8260385990142822, "learning_rate": 0.0009996271692483596, "loss": 3.0158, "step": 92 }, { "epoch": 0.04264099037138927, "grad_norm": 0.9323636889457703, "learning_rate": 0.0009995979423530893, "loss": 2.9659, "step": 93 }, { "epoch": 0.04309949564419991, "grad_norm": 0.8418369293212891, "learning_rate": 0.000999567613158439, "loss": 3.0072, "step": 94 }, { "epoch": 0.04355800091701054, "grad_norm": 0.791230320930481, "learning_rate": 0.0009995361817313263, "loss": 2.9942, "step": 95 }, { "epoch": 0.04401650618982118, "grad_norm": 0.8927780985832214, "learning_rate": 0.0009995036481411004, "loss": 3.0584, "step": 96 }, { "epoch": 0.044475011462631824, "grad_norm": 0.7939008474349976, "learning_rate": 0.0009994700124595429, "loss": 3.0909, "step": 97 }, { "epoch": 0.04493351673544246, "grad_norm": 0.8703867197036743, "learning_rate": 0.0009994352747608663, "loss": 3.0062, "step": 98 }, { "epoch": 0.0453920220082531, "grad_norm": 0.9740915298461914, "learning_rate": 0.0009993994351217151, "loss": 3.011, "step": 99 }, { "epoch": 0.04585052728106373, "grad_norm": 0.8462305665016174, "learning_rate": 0.000999362493621165, "loss": 3.0195, "step": 100 }, { "epoch": 0.04630903255387437, "grad_norm": 0.8924987316131592, "learning_rate": 0.0009993244503407226, "loss": 2.9981, "step": 101 }, { "epoch": 0.04676753782668501, "grad_norm": 0.853411078453064, "learning_rate": 0.0009992853053643258, "loss": 2.8726, "step": 102 }, { "epoch": 0.04722604309949564, "grad_norm": 0.750439465045929, "learning_rate": 0.0009992450587783426, "loss": 2.8993, "step": 103 }, { "epoch": 0.04768454837230628, "grad_norm": 0.8093625903129578, "learning_rate": 0.000999203710671572, "loss": 2.9556, "step": 104 }, { "epoch": 0.048143053645116916, "grad_norm": 0.9567016363143921, "learning_rate": 0.0009991612611352438, "loss": 2.9536, "step": 105 }, { "epoch": 0.048601558917927556, "grad_norm": 0.8183872103691101, "learning_rate": 0.0009991177102630173, "loss": 2.982, "step": 106 }, { "epoch": 0.049060064190738197, "grad_norm": 1.0702154636383057, "learning_rate": 0.0009990730581509817, "loss": 2.8494, "step": 107 }, { "epoch": 0.04951856946354883, "grad_norm": 0.8388418555259705, "learning_rate": 0.0009990273048976566, "loss": 2.8346, "step": 108 }, { "epoch": 0.04997707473635947, "grad_norm": 0.9252877235412598, "learning_rate": 0.0009989804506039905, "loss": 2.9098, "step": 109 }, { "epoch": 0.0504355800091701, "grad_norm": 0.995561957359314, "learning_rate": 0.0009989324953733614, "loss": 2.9821, "step": 110 }, { "epoch": 0.05089408528198074, "grad_norm": 0.909844160079956, "learning_rate": 0.0009988834393115766, "loss": 2.9853, "step": 111 }, { "epoch": 0.05135259055479138, "grad_norm": 0.7798081040382385, "learning_rate": 0.000998833282526872, "loss": 2.9241, "step": 112 }, { "epoch": 0.051811095827602016, "grad_norm": 0.9033353924751282, "learning_rate": 0.0009987820251299122, "loss": 2.8293, "step": 113 }, { "epoch": 0.052269601100412656, "grad_norm": 0.7936144471168518, "learning_rate": 0.00099872966723379, "loss": 2.8975, "step": 114 }, { "epoch": 0.05272810637322329, "grad_norm": 0.7796331644058228, "learning_rate": 0.0009986762089540266, "loss": 2.9079, "step": 115 }, { "epoch": 0.05318661164603393, "grad_norm": 0.860529363155365, "learning_rate": 0.0009986216504085709, "loss": 2.935, "step": 116 }, { "epoch": 0.05364511691884457, "grad_norm": 0.8514927625656128, "learning_rate": 0.0009985659917177991, "loss": 2.9902, "step": 117 }, { "epoch": 0.0541036221916552, "grad_norm": 0.7927408218383789, "learning_rate": 0.0009985092330045155, "loss": 2.8841, "step": 118 }, { "epoch": 0.05456212746446584, "grad_norm": 0.8137592077255249, "learning_rate": 0.0009984513743939508, "loss": 2.9013, "step": 119 }, { "epoch": 0.055020632737276476, "grad_norm": 0.8622753024101257, "learning_rate": 0.0009983924160137626, "loss": 2.8821, "step": 120 }, { "epoch": 0.055479138010087116, "grad_norm": 0.8287363648414612, "learning_rate": 0.000998332357994035, "loss": 2.7775, "step": 121 }, { "epoch": 0.055937643282897756, "grad_norm": 0.8615383505821228, "learning_rate": 0.0009982712004672786, "loss": 2.8636, "step": 122 }, { "epoch": 0.05639614855570839, "grad_norm": 0.811129629611969, "learning_rate": 0.0009982089435684295, "loss": 2.8062, "step": 123 }, { "epoch": 0.05685465382851903, "grad_norm": 0.8522815108299255, "learning_rate": 0.0009981455874348499, "loss": 2.8244, "step": 124 }, { "epoch": 0.05731315910132966, "grad_norm": 0.840760350227356, "learning_rate": 0.0009980811322063269, "loss": 2.8017, "step": 125 }, { "epoch": 0.0577716643741403, "grad_norm": 0.8835216164588928, "learning_rate": 0.0009980155780250728, "loss": 2.7901, "step": 126 }, { "epoch": 0.05823016964695094, "grad_norm": 0.812961757183075, "learning_rate": 0.0009979489250357243, "loss": 2.7502, "step": 127 }, { "epoch": 0.058688674919761576, "grad_norm": 0.8224567770957947, "learning_rate": 0.0009978811733853431, "loss": 2.7531, "step": 128 }, { "epoch": 0.059147180192572216, "grad_norm": 0.8505994081497192, "learning_rate": 0.0009978123232234147, "loss": 2.7798, "step": 129 }, { "epoch": 0.05960568546538285, "grad_norm": 0.835339367389679, "learning_rate": 0.000997742374701848, "loss": 2.7393, "step": 130 }, { "epoch": 0.06006419073819349, "grad_norm": 0.7374517917633057, "learning_rate": 0.0009976713279749754, "loss": 2.6671, "step": 131 }, { "epoch": 0.06052269601100413, "grad_norm": 0.824495255947113, "learning_rate": 0.0009975991831995528, "loss": 2.7388, "step": 132 }, { "epoch": 0.06098120128381476, "grad_norm": 0.794761598110199, "learning_rate": 0.0009975259405347581, "loss": 2.7582, "step": 133 }, { "epoch": 0.0614397065566254, "grad_norm": 0.8175957798957825, "learning_rate": 0.0009974516001421926, "loss": 2.7647, "step": 134 }, { "epoch": 0.061898211829436035, "grad_norm": 0.8196097612380981, "learning_rate": 0.000997376162185878, "loss": 2.6773, "step": 135 }, { "epoch": 0.062356717102246675, "grad_norm": 0.7816139459609985, "learning_rate": 0.0009972996268322594, "loss": 2.7403, "step": 136 }, { "epoch": 0.06281522237505731, "grad_norm": 0.8296950459480286, "learning_rate": 0.0009972219942502017, "loss": 2.763, "step": 137 }, { "epoch": 0.06327372764786796, "grad_norm": 0.7842466235160828, "learning_rate": 0.0009971432646109918, "loss": 2.7441, "step": 138 }, { "epoch": 0.06373223292067859, "grad_norm": 0.7252905368804932, "learning_rate": 0.0009970634380883365, "loss": 2.7317, "step": 139 }, { "epoch": 0.06419073819348922, "grad_norm": 0.7634514570236206, "learning_rate": 0.0009969825148583627, "loss": 2.7659, "step": 140 }, { "epoch": 0.06464924346629987, "grad_norm": 0.7778244614601135, "learning_rate": 0.0009969004950996173, "loss": 2.6597, "step": 141 }, { "epoch": 0.0651077487391105, "grad_norm": 0.8146045207977295, "learning_rate": 0.0009968173789930668, "loss": 2.7723, "step": 142 }, { "epoch": 0.06556625401192114, "grad_norm": 0.7817020416259766, "learning_rate": 0.0009967331667220958, "loss": 2.6466, "step": 143 }, { "epoch": 0.06602475928473177, "grad_norm": 0.8090150356292725, "learning_rate": 0.0009966478584725086, "loss": 2.5969, "step": 144 }, { "epoch": 0.06648326455754242, "grad_norm": 0.7618467211723328, "learning_rate": 0.0009965614544325263, "loss": 2.7322, "step": 145 }, { "epoch": 0.06694176983035305, "grad_norm": 0.8167206048965454, "learning_rate": 0.000996473954792789, "loss": 2.7905, "step": 146 }, { "epoch": 0.06740027510316368, "grad_norm": 0.791314423084259, "learning_rate": 0.0009963853597463532, "loss": 2.7007, "step": 147 }, { "epoch": 0.06785878037597433, "grad_norm": 0.8050386905670166, "learning_rate": 0.000996295669488693, "loss": 2.6895, "step": 148 }, { "epoch": 0.06831728564878496, "grad_norm": 0.7936955690383911, "learning_rate": 0.0009962048842176979, "loss": 2.7382, "step": 149 }, { "epoch": 0.0687757909215956, "grad_norm": 0.7279852032661438, "learning_rate": 0.0009961130041336748, "loss": 2.7085, "step": 150 }, { "epoch": 0.06923429619440624, "grad_norm": 0.7519203424453735, "learning_rate": 0.0009960200294393449, "loss": 2.7585, "step": 151 }, { "epoch": 0.06969280146721687, "grad_norm": 0.7762235999107361, "learning_rate": 0.0009959259603398453, "loss": 2.7362, "step": 152 }, { "epoch": 0.07015130674002751, "grad_norm": 0.7462303042411804, "learning_rate": 0.0009958307970427275, "loss": 2.5587, "step": 153 }, { "epoch": 0.07060981201283814, "grad_norm": 0.7439447045326233, "learning_rate": 0.0009957345397579572, "loss": 2.6598, "step": 154 }, { "epoch": 0.07106831728564879, "grad_norm": 0.7570639252662659, "learning_rate": 0.0009956371886979138, "loss": 2.6196, "step": 155 }, { "epoch": 0.07152682255845942, "grad_norm": 0.7781119346618652, "learning_rate": 0.00099553874407739, "loss": 2.7402, "step": 156 }, { "epoch": 0.07198532783127005, "grad_norm": 0.7740487456321716, "learning_rate": 0.0009954392061135916, "loss": 2.6426, "step": 157 }, { "epoch": 0.0724438331040807, "grad_norm": 0.7941224575042725, "learning_rate": 0.0009953385750261364, "loss": 2.6514, "step": 158 }, { "epoch": 0.07290233837689133, "grad_norm": 0.7795929908752441, "learning_rate": 0.0009952368510370538, "loss": 2.6252, "step": 159 }, { "epoch": 0.07336084364970197, "grad_norm": 0.7422616481781006, "learning_rate": 0.0009951340343707852, "loss": 2.6687, "step": 160 }, { "epoch": 0.07381934892251261, "grad_norm": 0.7390515208244324, "learning_rate": 0.0009950301252541823, "loss": 2.6709, "step": 161 }, { "epoch": 0.07427785419532325, "grad_norm": 0.7419459223747253, "learning_rate": 0.0009949251239165075, "loss": 2.6245, "step": 162 }, { "epoch": 0.07473635946813388, "grad_norm": 0.6776374578475952, "learning_rate": 0.000994819030589433, "loss": 2.6948, "step": 163 }, { "epoch": 0.07519486474094451, "grad_norm": 0.7132704854011536, "learning_rate": 0.00099471184550704, "loss": 2.5985, "step": 164 }, { "epoch": 0.07565337001375516, "grad_norm": 0.7755231261253357, "learning_rate": 0.0009946035689058189, "loss": 2.6483, "step": 165 }, { "epoch": 0.0761118752865658, "grad_norm": 0.7836331725120544, "learning_rate": 0.0009944942010246681, "loss": 2.711, "step": 166 }, { "epoch": 0.07657038055937643, "grad_norm": 0.7065374255180359, "learning_rate": 0.0009943837421048942, "loss": 2.6456, "step": 167 }, { "epoch": 0.07702888583218707, "grad_norm": 0.7108990550041199, "learning_rate": 0.0009942721923902106, "loss": 2.582, "step": 168 }, { "epoch": 0.07748739110499771, "grad_norm": 0.6788803935050964, "learning_rate": 0.0009941595521267377, "loss": 2.563, "step": 169 }, { "epoch": 0.07794589637780834, "grad_norm": 0.7292799353599548, "learning_rate": 0.0009940458215630017, "loss": 2.6864, "step": 170 }, { "epoch": 0.07840440165061899, "grad_norm": 0.7661619782447815, "learning_rate": 0.0009939310009499348, "loss": 2.714, "step": 171 }, { "epoch": 0.07886290692342962, "grad_norm": 0.7363874912261963, "learning_rate": 0.000993815090540874, "loss": 2.7387, "step": 172 }, { "epoch": 0.07932141219624025, "grad_norm": 0.6720567941665649, "learning_rate": 0.000993698090591561, "loss": 2.714, "step": 173 }, { "epoch": 0.07977991746905089, "grad_norm": 0.719196617603302, "learning_rate": 0.0009935800013601416, "loss": 2.6246, "step": 174 }, { "epoch": 0.08023842274186153, "grad_norm": 0.7657513618469238, "learning_rate": 0.000993460823107164, "loss": 2.6453, "step": 175 }, { "epoch": 0.08069692801467217, "grad_norm": 0.7313987612724304, "learning_rate": 0.0009933405560955803, "loss": 2.6592, "step": 176 }, { "epoch": 0.0811554332874828, "grad_norm": 0.7516104578971863, "learning_rate": 0.0009932192005907446, "loss": 2.6589, "step": 177 }, { "epoch": 0.08161393856029345, "grad_norm": 0.6854733228683472, "learning_rate": 0.0009930967568604118, "loss": 2.745, "step": 178 }, { "epoch": 0.08207244383310408, "grad_norm": 0.7707527875900269, "learning_rate": 0.000992973225174739, "loss": 2.5888, "step": 179 }, { "epoch": 0.08253094910591471, "grad_norm": 0.6960265636444092, "learning_rate": 0.0009928486058062827, "loss": 2.6258, "step": 180 }, { "epoch": 0.08298945437872536, "grad_norm": 0.7211937308311462, "learning_rate": 0.0009927228990299999, "loss": 2.7127, "step": 181 }, { "epoch": 0.083447959651536, "grad_norm": 0.7599518895149231, "learning_rate": 0.0009925961051232468, "loss": 2.7358, "step": 182 }, { "epoch": 0.08390646492434663, "grad_norm": 0.74045330286026, "learning_rate": 0.000992468224365778, "loss": 2.6317, "step": 183 }, { "epoch": 0.08436497019715727, "grad_norm": 0.8253055810928345, "learning_rate": 0.000992339257039746, "loss": 2.6695, "step": 184 }, { "epoch": 0.08482347546996791, "grad_norm": 0.7280474305152893, "learning_rate": 0.0009922092034297006, "loss": 2.5948, "step": 185 }, { "epoch": 0.08528198074277854, "grad_norm": 0.7211241722106934, "learning_rate": 0.0009920780638225891, "loss": 2.5493, "step": 186 }, { "epoch": 0.08574048601558917, "grad_norm": 0.7805038094520569, "learning_rate": 0.0009919458385077538, "loss": 2.5393, "step": 187 }, { "epoch": 0.08619899128839982, "grad_norm": 0.6865206956863403, "learning_rate": 0.0009918125277769336, "loss": 2.6097, "step": 188 }, { "epoch": 0.08665749656121045, "grad_norm": 0.6762561798095703, "learning_rate": 0.0009916781319242614, "loss": 2.6053, "step": 189 }, { "epoch": 0.08711600183402109, "grad_norm": 0.6718788146972656, "learning_rate": 0.0009915426512462646, "loss": 2.6598, "step": 190 }, { "epoch": 0.08757450710683173, "grad_norm": 0.7050771713256836, "learning_rate": 0.0009914060860418644, "loss": 2.6023, "step": 191 }, { "epoch": 0.08803301237964237, "grad_norm": 0.7779918313026428, "learning_rate": 0.000991268436612374, "loss": 2.6915, "step": 192 }, { "epoch": 0.088491517652453, "grad_norm": 0.7435088753700256, "learning_rate": 0.0009911297032614997, "loss": 2.715, "step": 193 }, { "epoch": 0.08895002292526365, "grad_norm": 0.7455050349235535, "learning_rate": 0.000990989886295339, "loss": 2.5431, "step": 194 }, { "epoch": 0.08940852819807428, "grad_norm": 0.7284430265426636, "learning_rate": 0.0009908489860223804, "loss": 2.6722, "step": 195 }, { "epoch": 0.08986703347088491, "grad_norm": 0.6531606316566467, "learning_rate": 0.000990707002753502, "loss": 2.6069, "step": 196 }, { "epoch": 0.09032553874369555, "grad_norm": 0.6464143991470337, "learning_rate": 0.0009905639368019724, "loss": 2.6202, "step": 197 }, { "epoch": 0.0907840440165062, "grad_norm": 0.7029107809066772, "learning_rate": 0.0009904197884834482, "loss": 2.5886, "step": 198 }, { "epoch": 0.09124254928931683, "grad_norm": 0.643893301486969, "learning_rate": 0.0009902745581159742, "loss": 2.6557, "step": 199 }, { "epoch": 0.09170105456212746, "grad_norm": 0.6850451231002808, "learning_rate": 0.0009901282460199829, "loss": 2.7011, "step": 200 }, { "epoch": 0.0921595598349381, "grad_norm": 0.7384447455406189, "learning_rate": 0.0009899808525182935, "loss": 2.655, "step": 201 }, { "epoch": 0.09261806510774874, "grad_norm": 0.659737765789032, "learning_rate": 0.0009898323779361107, "loss": 2.6042, "step": 202 }, { "epoch": 0.09307657038055937, "grad_norm": 0.750080406665802, "learning_rate": 0.000989682822601025, "loss": 2.5655, "step": 203 }, { "epoch": 0.09353507565337002, "grad_norm": 0.7366781234741211, "learning_rate": 0.0009895321868430113, "loss": 2.6045, "step": 204 }, { "epoch": 0.09399358092618065, "grad_norm": 0.6554088592529297, "learning_rate": 0.0009893804709944281, "loss": 2.6122, "step": 205 }, { "epoch": 0.09445208619899129, "grad_norm": 0.723312258720398, "learning_rate": 0.0009892276753900174, "loss": 2.645, "step": 206 }, { "epoch": 0.09491059147180192, "grad_norm": 0.6729600429534912, "learning_rate": 0.0009890738003669028, "loss": 2.6302, "step": 207 }, { "epoch": 0.09536909674461257, "grad_norm": 0.6689825057983398, "learning_rate": 0.0009889188462645904, "loss": 2.6286, "step": 208 }, { "epoch": 0.0958276020174232, "grad_norm": 0.7905515432357788, "learning_rate": 0.0009887628134249667, "loss": 2.6964, "step": 209 }, { "epoch": 0.09628610729023383, "grad_norm": 0.6612294316291809, "learning_rate": 0.0009886057021922983, "loss": 2.6925, "step": 210 }, { "epoch": 0.09674461256304448, "grad_norm": 0.7880565524101257, "learning_rate": 0.0009884475129132311, "loss": 2.6375, "step": 211 }, { "epoch": 0.09720311783585511, "grad_norm": 0.6407657265663147, "learning_rate": 0.0009882882459367897, "loss": 2.4613, "step": 212 }, { "epoch": 0.09766162310866575, "grad_norm": 0.683295726776123, "learning_rate": 0.0009881279016143766, "loss": 2.4862, "step": 213 }, { "epoch": 0.09812012838147639, "grad_norm": 0.7771915793418884, "learning_rate": 0.0009879664802997707, "loss": 2.5901, "step": 214 }, { "epoch": 0.09857863365428703, "grad_norm": 0.7148461937904358, "learning_rate": 0.000987803982349128, "loss": 2.6909, "step": 215 }, { "epoch": 0.09903713892709766, "grad_norm": 0.6359391212463379, "learning_rate": 0.0009876404081209796, "loss": 2.6596, "step": 216 }, { "epoch": 0.09949564419990829, "grad_norm": 0.6708654761314392, "learning_rate": 0.000987475757976231, "loss": 2.5812, "step": 217 }, { "epoch": 0.09995414947271894, "grad_norm": 0.6446850895881653, "learning_rate": 0.000987310032278162, "loss": 2.5798, "step": 218 }, { "epoch": 0.10041265474552957, "grad_norm": 0.6731483936309814, "learning_rate": 0.0009871432313924254, "loss": 2.5845, "step": 219 }, { "epoch": 0.1008711600183402, "grad_norm": 0.8195900917053223, "learning_rate": 0.000986975355687046, "loss": 2.6453, "step": 220 }, { "epoch": 0.10132966529115085, "grad_norm": 0.6609553694725037, "learning_rate": 0.0009868064055324204, "loss": 2.6425, "step": 221 }, { "epoch": 0.10178817056396149, "grad_norm": 0.7818976044654846, "learning_rate": 0.0009866363813013153, "loss": 2.6354, "step": 222 }, { "epoch": 0.10224667583677212, "grad_norm": 0.7008618116378784, "learning_rate": 0.0009864652833688676, "loss": 2.6305, "step": 223 }, { "epoch": 0.10270518110958277, "grad_norm": 0.6296684145927429, "learning_rate": 0.0009862931121125836, "loss": 2.5535, "step": 224 }, { "epoch": 0.1031636863823934, "grad_norm": 0.7166476845741272, "learning_rate": 0.000986119867912337, "loss": 2.5976, "step": 225 }, { "epoch": 0.10362219165520403, "grad_norm": 0.7142142057418823, "learning_rate": 0.000985945551150369, "loss": 2.5857, "step": 226 }, { "epoch": 0.10408069692801467, "grad_norm": 0.65469890832901, "learning_rate": 0.0009857701622112876, "loss": 2.6271, "step": 227 }, { "epoch": 0.10453920220082531, "grad_norm": 0.7255174517631531, "learning_rate": 0.000985593701482066, "loss": 2.6798, "step": 228 }, { "epoch": 0.10499770747363595, "grad_norm": 0.7853558659553528, "learning_rate": 0.0009854161693520424, "loss": 2.5619, "step": 229 }, { "epoch": 0.10545621274644658, "grad_norm": 0.6279813051223755, "learning_rate": 0.0009852375662129194, "loss": 2.6152, "step": 230 }, { "epoch": 0.10591471801925723, "grad_norm": 0.6759238839149475, "learning_rate": 0.0009850578924587613, "loss": 2.5836, "step": 231 }, { "epoch": 0.10637322329206786, "grad_norm": 0.7071142196655273, "learning_rate": 0.000984877148485996, "loss": 2.6634, "step": 232 }, { "epoch": 0.10683172856487849, "grad_norm": 0.6480698585510254, "learning_rate": 0.000984695334693412, "loss": 2.7005, "step": 233 }, { "epoch": 0.10729023383768914, "grad_norm": 0.6557847857475281, "learning_rate": 0.000984512451482158, "loss": 2.6101, "step": 234 }, { "epoch": 0.10774873911049977, "grad_norm": 0.7439600825309753, "learning_rate": 0.0009843284992557431, "loss": 2.5876, "step": 235 }, { "epoch": 0.1082072443833104, "grad_norm": 0.6205833554267883, "learning_rate": 0.000984143478420034, "loss": 2.6273, "step": 236 }, { "epoch": 0.10866574965612105, "grad_norm": 0.7119700312614441, "learning_rate": 0.0009839573893832563, "loss": 2.5416, "step": 237 }, { "epoch": 0.10912425492893169, "grad_norm": 0.8242546916007996, "learning_rate": 0.000983770232555991, "loss": 2.6923, "step": 238 }, { "epoch": 0.10958276020174232, "grad_norm": 0.6326975226402283, "learning_rate": 0.0009835820083511765, "loss": 2.5317, "step": 239 }, { "epoch": 0.11004126547455295, "grad_norm": 0.5985181927680969, "learning_rate": 0.0009833927171841055, "loss": 2.6491, "step": 240 }, { "epoch": 0.1104997707473636, "grad_norm": 0.6098171472549438, "learning_rate": 0.0009832023594724246, "loss": 2.6397, "step": 241 }, { "epoch": 0.11095827602017423, "grad_norm": 0.7078850865364075, "learning_rate": 0.0009830109356361344, "loss": 2.6152, "step": 242 }, { "epoch": 0.11141678129298486, "grad_norm": 0.5823879837989807, "learning_rate": 0.0009828184460975867, "loss": 2.6173, "step": 243 }, { "epoch": 0.11187528656579551, "grad_norm": 0.6380863785743713, "learning_rate": 0.0009826248912814855, "loss": 2.5891, "step": 244 }, { "epoch": 0.11233379183860615, "grad_norm": 0.6562332510948181, "learning_rate": 0.0009824302716148847, "loss": 2.5794, "step": 245 }, { "epoch": 0.11279229711141678, "grad_norm": 0.6000638604164124, "learning_rate": 0.0009822345875271884, "loss": 2.5179, "step": 246 }, { "epoch": 0.11325080238422743, "grad_norm": 0.6230500340461731, "learning_rate": 0.0009820378394501481, "loss": 2.5751, "step": 247 }, { "epoch": 0.11370930765703806, "grad_norm": 0.7061066031455994, "learning_rate": 0.0009818400278178636, "loss": 2.483, "step": 248 }, { "epoch": 0.11416781292984869, "grad_norm": 0.5904055237770081, "learning_rate": 0.0009816411530667814, "loss": 2.5582, "step": 249 }, { "epoch": 0.11462631820265932, "grad_norm": 0.6341516375541687, "learning_rate": 0.000981441215635693, "loss": 2.6526, "step": 250 }, { "epoch": 0.11508482347546997, "grad_norm": 0.6906593441963196, "learning_rate": 0.0009812402159657353, "loss": 2.6754, "step": 251 }, { "epoch": 0.1155433287482806, "grad_norm": 0.6036014556884766, "learning_rate": 0.000981038154500388, "loss": 2.6298, "step": 252 }, { "epoch": 0.11600183402109124, "grad_norm": 0.6857396364212036, "learning_rate": 0.0009808350316854746, "loss": 2.5398, "step": 253 }, { "epoch": 0.11646033929390188, "grad_norm": 0.6498702764511108, "learning_rate": 0.0009806308479691594, "loss": 2.6634, "step": 254 }, { "epoch": 0.11691884456671252, "grad_norm": 0.6380029916763306, "learning_rate": 0.0009804256038019482, "loss": 2.6105, "step": 255 }, { "epoch": 0.11737734983952315, "grad_norm": 0.6056024432182312, "learning_rate": 0.0009802192996366857, "loss": 2.6103, "step": 256 }, { "epoch": 0.1178358551123338, "grad_norm": 0.5880225896835327, "learning_rate": 0.0009800119359285563, "loss": 2.5289, "step": 257 }, { "epoch": 0.11829436038514443, "grad_norm": 0.6208682656288147, "learning_rate": 0.0009798035131350813, "loss": 2.584, "step": 258 }, { "epoch": 0.11875286565795506, "grad_norm": 0.6291673183441162, "learning_rate": 0.0009795940317161194, "loss": 2.5443, "step": 259 }, { "epoch": 0.1192113709307657, "grad_norm": 0.6266107559204102, "learning_rate": 0.0009793834921338646, "loss": 2.6463, "step": 260 }, { "epoch": 0.11966987620357634, "grad_norm": 0.6030829548835754, "learning_rate": 0.0009791718948528457, "loss": 2.6109, "step": 261 }, { "epoch": 0.12012838147638698, "grad_norm": 0.6148271560668945, "learning_rate": 0.0009789592403399252, "loss": 2.5514, "step": 262 }, { "epoch": 0.12058688674919761, "grad_norm": 0.6307891011238098, "learning_rate": 0.0009787455290642985, "loss": 2.5693, "step": 263 }, { "epoch": 0.12104539202200826, "grad_norm": 0.62510746717453, "learning_rate": 0.000978530761497492, "loss": 2.5906, "step": 264 }, { "epoch": 0.12150389729481889, "grad_norm": 0.6534302830696106, "learning_rate": 0.0009783149381133633, "loss": 2.5485, "step": 265 }, { "epoch": 0.12196240256762952, "grad_norm": 0.6074642539024353, "learning_rate": 0.0009780980593880992, "loss": 2.5587, "step": 266 }, { "epoch": 0.12242090784044017, "grad_norm": 0.5880428552627563, "learning_rate": 0.0009778801258002153, "loss": 2.5815, "step": 267 }, { "epoch": 0.1228794131132508, "grad_norm": 0.5763751864433289, "learning_rate": 0.000977661137830554, "loss": 2.5364, "step": 268 }, { "epoch": 0.12333791838606144, "grad_norm": 0.6359049677848816, "learning_rate": 0.0009774410959622845, "loss": 2.6077, "step": 269 }, { "epoch": 0.12379642365887207, "grad_norm": 0.6145782470703125, "learning_rate": 0.000977220000680901, "loss": 2.473, "step": 270 }, { "epoch": 0.12425492893168272, "grad_norm": 0.6185532808303833, "learning_rate": 0.000976997852474223, "loss": 2.4808, "step": 271 }, { "epoch": 0.12471343420449335, "grad_norm": 0.6106144189834595, "learning_rate": 0.0009767746518323914, "loss": 2.51, "step": 272 }, { "epoch": 0.12517193947730398, "grad_norm": 0.6200641393661499, "learning_rate": 0.0009765503992478704, "loss": 2.522, "step": 273 }, { "epoch": 0.12563044475011462, "grad_norm": 0.6676740646362305, "learning_rate": 0.0009763250952154449, "loss": 2.5583, "step": 274 }, { "epoch": 0.12608895002292525, "grad_norm": 0.6546444892883301, "learning_rate": 0.0009760987402322195, "loss": 2.5564, "step": 275 }, { "epoch": 0.1265474552957359, "grad_norm": 0.6739364862442017, "learning_rate": 0.0009758713347976178, "loss": 2.6407, "step": 276 }, { "epoch": 0.12700596056854654, "grad_norm": 0.5954195261001587, "learning_rate": 0.000975642879413381, "loss": 2.6025, "step": 277 }, { "epoch": 0.12746446584135718, "grad_norm": 0.5663610696792603, "learning_rate": 0.0009754133745835665, "loss": 2.5694, "step": 278 }, { "epoch": 0.1279229711141678, "grad_norm": 0.6913132667541504, "learning_rate": 0.0009751828208145482, "loss": 2.5711, "step": 279 }, { "epoch": 0.12838147638697844, "grad_norm": 0.6230430006980896, "learning_rate": 0.0009749512186150131, "loss": 2.5844, "step": 280 }, { "epoch": 0.12883998165978908, "grad_norm": 0.6440043449401855, "learning_rate": 0.0009747185684959625, "loss": 2.5813, "step": 281 }, { "epoch": 0.12929848693259974, "grad_norm": 0.7036247253417969, "learning_rate": 0.000974484870970709, "loss": 2.5894, "step": 282 }, { "epoch": 0.12975699220541037, "grad_norm": 0.6413722634315491, "learning_rate": 0.0009742501265548767, "loss": 2.5696, "step": 283 }, { "epoch": 0.130215497478221, "grad_norm": 0.6380294561386108, "learning_rate": 0.0009740143357663993, "loss": 2.6822, "step": 284 }, { "epoch": 0.13067400275103164, "grad_norm": 0.6950642466545105, "learning_rate": 0.000973777499125519, "loss": 2.5697, "step": 285 }, { "epoch": 0.13113250802384227, "grad_norm": 0.6845468878746033, "learning_rate": 0.0009735396171547859, "loss": 2.5002, "step": 286 }, { "epoch": 0.1315910132966529, "grad_norm": 0.6598009467124939, "learning_rate": 0.0009733006903790564, "loss": 2.6371, "step": 287 }, { "epoch": 0.13204951856946354, "grad_norm": 0.7019338607788086, "learning_rate": 0.0009730607193254922, "loss": 2.5826, "step": 288 }, { "epoch": 0.1325080238422742, "grad_norm": 0.5928915739059448, "learning_rate": 0.0009728197045235585, "loss": 2.6142, "step": 289 }, { "epoch": 0.13296652911508483, "grad_norm": 0.6495969295501709, "learning_rate": 0.0009725776465050242, "loss": 2.5416, "step": 290 }, { "epoch": 0.13342503438789546, "grad_norm": 0.6135349869728088, "learning_rate": 0.0009723345458039594, "loss": 2.5496, "step": 291 }, { "epoch": 0.1338835396607061, "grad_norm": 0.5801755785942078, "learning_rate": 0.000972090402956735, "loss": 2.6089, "step": 292 }, { "epoch": 0.13434204493351673, "grad_norm": 0.5726706981658936, "learning_rate": 0.0009718452185020212, "loss": 2.5342, "step": 293 }, { "epoch": 0.13480055020632736, "grad_norm": 0.6335543990135193, "learning_rate": 0.0009715989929807862, "loss": 2.5169, "step": 294 }, { "epoch": 0.13525905547913802, "grad_norm": 0.6074330806732178, "learning_rate": 0.0009713517269362955, "loss": 2.5953, "step": 295 }, { "epoch": 0.13571756075194866, "grad_norm": 0.6056231260299683, "learning_rate": 0.0009711034209141101, "loss": 2.5398, "step": 296 }, { "epoch": 0.1361760660247593, "grad_norm": 0.6114384531974792, "learning_rate": 0.0009708540754620856, "loss": 2.6214, "step": 297 }, { "epoch": 0.13663457129756992, "grad_norm": 0.588083028793335, "learning_rate": 0.0009706036911303713, "loss": 2.6616, "step": 298 }, { "epoch": 0.13709307657038056, "grad_norm": 0.6131693124771118, "learning_rate": 0.0009703522684714083, "loss": 2.4941, "step": 299 }, { "epoch": 0.1375515818431912, "grad_norm": 0.6044536828994751, "learning_rate": 0.0009700998080399286, "loss": 2.5836, "step": 300 }, { "epoch": 0.13801008711600182, "grad_norm": 0.5578015446662903, "learning_rate": 0.0009698463103929542, "loss": 2.6024, "step": 301 }, { "epoch": 0.13846859238881248, "grad_norm": 0.5918693542480469, "learning_rate": 0.0009695917760897954, "loss": 2.5565, "step": 302 }, { "epoch": 0.13892709766162312, "grad_norm": 0.5961376428604126, "learning_rate": 0.0009693362056920501, "loss": 2.6136, "step": 303 }, { "epoch": 0.13938560293443375, "grad_norm": 0.590192973613739, "learning_rate": 0.0009690795997636015, "loss": 2.5528, "step": 304 }, { "epoch": 0.13984410820724438, "grad_norm": 0.5912079811096191, "learning_rate": 0.0009688219588706179, "loss": 2.5489, "step": 305 }, { "epoch": 0.14030261348005502, "grad_norm": 0.5485712885856628, "learning_rate": 0.0009685632835815518, "loss": 2.5233, "step": 306 }, { "epoch": 0.14076111875286565, "grad_norm": 0.5716344118118286, "learning_rate": 0.0009683035744671367, "loss": 2.5201, "step": 307 }, { "epoch": 0.14121962402567628, "grad_norm": 0.6181029081344604, "learning_rate": 0.0009680428321003883, "loss": 2.4859, "step": 308 }, { "epoch": 0.14167812929848694, "grad_norm": 0.58668452501297, "learning_rate": 0.000967781057056601, "loss": 2.5452, "step": 309 }, { "epoch": 0.14213663457129758, "grad_norm": 0.5966657996177673, "learning_rate": 0.0009675182499133485, "loss": 2.5951, "step": 310 }, { "epoch": 0.1425951398441082, "grad_norm": 0.5770857334136963, "learning_rate": 0.0009672544112504813, "loss": 2.5431, "step": 311 }, { "epoch": 0.14305364511691884, "grad_norm": 0.6413736343383789, "learning_rate": 0.0009669895416501257, "loss": 2.4969, "step": 312 }, { "epoch": 0.14351215038972948, "grad_norm": 0.6049234867095947, "learning_rate": 0.0009667236416966833, "loss": 2.6225, "step": 313 }, { "epoch": 0.1439706556625401, "grad_norm": 0.5999007225036621, "learning_rate": 0.0009664567119768281, "loss": 2.5356, "step": 314 }, { "epoch": 0.14442916093535077, "grad_norm": 0.622349202632904, "learning_rate": 0.0009661887530795067, "loss": 2.5123, "step": 315 }, { "epoch": 0.1448876662081614, "grad_norm": 0.6049217581748962, "learning_rate": 0.0009659197655959365, "loss": 2.5464, "step": 316 }, { "epoch": 0.14534617148097204, "grad_norm": 0.5926075577735901, "learning_rate": 0.000965649750119604, "loss": 2.5565, "step": 317 }, { "epoch": 0.14580467675378267, "grad_norm": 0.6536279320716858, "learning_rate": 0.0009653787072462643, "loss": 2.5748, "step": 318 }, { "epoch": 0.1462631820265933, "grad_norm": 0.597464382648468, "learning_rate": 0.0009651066375739388, "loss": 2.6439, "step": 319 }, { "epoch": 0.14672168729940394, "grad_norm": 0.6035195589065552, "learning_rate": 0.000964833541702915, "loss": 2.5044, "step": 320 }, { "epoch": 0.14718019257221457, "grad_norm": 0.5538439750671387, "learning_rate": 0.0009645594202357438, "loss": 2.555, "step": 321 }, { "epoch": 0.14763869784502523, "grad_norm": 0.5793418884277344, "learning_rate": 0.0009642842737772397, "loss": 2.5241, "step": 322 }, { "epoch": 0.14809720311783586, "grad_norm": 0.5895296931266785, "learning_rate": 0.0009640081029344782, "loss": 2.5751, "step": 323 }, { "epoch": 0.1485557083906465, "grad_norm": 0.6034107804298401, "learning_rate": 0.0009637309083167956, "loss": 2.6122, "step": 324 }, { "epoch": 0.14901421366345713, "grad_norm": 0.5846843719482422, "learning_rate": 0.0009634526905357859, "loss": 2.5936, "step": 325 }, { "epoch": 0.14947271893626776, "grad_norm": 0.6000763773918152, "learning_rate": 0.000963173450205302, "loss": 2.6445, "step": 326 }, { "epoch": 0.1499312242090784, "grad_norm": 0.604120671749115, "learning_rate": 0.0009628931879414517, "loss": 2.4909, "step": 327 }, { "epoch": 0.15038972948188903, "grad_norm": 0.5964603424072266, "learning_rate": 0.0009626119043625983, "loss": 2.5436, "step": 328 }, { "epoch": 0.1508482347546997, "grad_norm": 0.6051590442657471, "learning_rate": 0.0009623296000893582, "loss": 2.5208, "step": 329 }, { "epoch": 0.15130674002751032, "grad_norm": 0.5707788467407227, "learning_rate": 0.0009620462757446, "loss": 2.5448, "step": 330 }, { "epoch": 0.15176524530032096, "grad_norm": 0.5382934808731079, "learning_rate": 0.0009617619319534428, "loss": 2.5824, "step": 331 }, { "epoch": 0.1522237505731316, "grad_norm": 0.5656911730766296, "learning_rate": 0.000961476569343255, "loss": 2.569, "step": 332 }, { "epoch": 0.15268225584594222, "grad_norm": 0.547869086265564, "learning_rate": 0.0009611901885436529, "loss": 2.6067, "step": 333 }, { "epoch": 0.15314076111875286, "grad_norm": 0.5131890177726746, "learning_rate": 0.0009609027901864996, "loss": 2.5036, "step": 334 }, { "epoch": 0.15359926639156352, "grad_norm": 0.597594141960144, "learning_rate": 0.0009606143749059029, "loss": 2.5716, "step": 335 }, { "epoch": 0.15405777166437415, "grad_norm": 0.5916976928710938, "learning_rate": 0.0009603249433382144, "loss": 2.5884, "step": 336 }, { "epoch": 0.15451627693718478, "grad_norm": 0.5927761793136597, "learning_rate": 0.0009600344961220282, "loss": 2.545, "step": 337 }, { "epoch": 0.15497478220999542, "grad_norm": 0.5937693119049072, "learning_rate": 0.0009597430338981791, "loss": 2.5918, "step": 338 }, { "epoch": 0.15543328748280605, "grad_norm": 0.5822310447692871, "learning_rate": 0.0009594505573097414, "loss": 2.5711, "step": 339 }, { "epoch": 0.15589179275561668, "grad_norm": 0.5900925993919373, "learning_rate": 0.0009591570670020277, "loss": 2.5832, "step": 340 }, { "epoch": 0.15635029802842731, "grad_norm": 0.5826736092567444, "learning_rate": 0.0009588625636225871, "loss": 2.5488, "step": 341 }, { "epoch": 0.15680880330123798, "grad_norm": 0.5694997906684875, "learning_rate": 0.0009585670478212036, "loss": 2.3951, "step": 342 }, { "epoch": 0.1572673085740486, "grad_norm": 0.5512198209762573, "learning_rate": 0.0009582705202498956, "loss": 2.5534, "step": 343 }, { "epoch": 0.15772581384685924, "grad_norm": 0.5626025795936584, "learning_rate": 0.0009579729815629133, "loss": 2.5993, "step": 344 }, { "epoch": 0.15818431911966988, "grad_norm": 0.5917795300483704, "learning_rate": 0.0009576744324167379, "loss": 2.6416, "step": 345 }, { "epoch": 0.1586428243924805, "grad_norm": 0.5532258152961731, "learning_rate": 0.0009573748734700804, "loss": 2.4818, "step": 346 }, { "epoch": 0.15910132966529114, "grad_norm": 0.5765784978866577, "learning_rate": 0.0009570743053838796, "loss": 2.5898, "step": 347 }, { "epoch": 0.15955983493810177, "grad_norm": 0.5838421583175659, "learning_rate": 0.0009567727288213005, "loss": 2.4944, "step": 348 }, { "epoch": 0.16001834021091244, "grad_norm": 0.583063542842865, "learning_rate": 0.0009564701444477337, "loss": 2.6087, "step": 349 }, { "epoch": 0.16047684548372307, "grad_norm": 0.5646284222602844, "learning_rate": 0.000956166552930793, "loss": 2.518, "step": 350 }, { "epoch": 0.1609353507565337, "grad_norm": 0.5773508548736572, "learning_rate": 0.0009558619549403147, "loss": 2.6355, "step": 351 }, { "epoch": 0.16139385602934433, "grad_norm": 0.5680062174797058, "learning_rate": 0.0009555563511483555, "loss": 2.5323, "step": 352 }, { "epoch": 0.16185236130215497, "grad_norm": 0.5783677101135254, "learning_rate": 0.0009552497422291912, "loss": 2.4898, "step": 353 }, { "epoch": 0.1623108665749656, "grad_norm": 0.5482762455940247, "learning_rate": 0.0009549421288593157, "loss": 2.5687, "step": 354 }, { "epoch": 0.16276937184777626, "grad_norm": 0.5588790774345398, "learning_rate": 0.0009546335117174385, "loss": 2.5811, "step": 355 }, { "epoch": 0.1632278771205869, "grad_norm": 0.5619999170303345, "learning_rate": 0.0009543238914844843, "loss": 2.6016, "step": 356 }, { "epoch": 0.16368638239339753, "grad_norm": 0.6477952599525452, "learning_rate": 0.0009540132688435907, "loss": 2.5737, "step": 357 }, { "epoch": 0.16414488766620816, "grad_norm": 0.6029506921768188, "learning_rate": 0.0009537016444801074, "loss": 2.4718, "step": 358 }, { "epoch": 0.1646033929390188, "grad_norm": 0.6382081508636475, "learning_rate": 0.0009533890190815935, "loss": 2.6751, "step": 359 }, { "epoch": 0.16506189821182943, "grad_norm": 0.6095901727676392, "learning_rate": 0.0009530753933378173, "loss": 2.5304, "step": 360 }, { "epoch": 0.16552040348464006, "grad_norm": 0.5480243563652039, "learning_rate": 0.0009527607679407545, "loss": 2.5603, "step": 361 }, { "epoch": 0.16597890875745072, "grad_norm": 0.5135402679443359, "learning_rate": 0.0009524451435845857, "loss": 2.5814, "step": 362 }, { "epoch": 0.16643741403026135, "grad_norm": 0.5636144280433655, "learning_rate": 0.0009521285209656963, "loss": 2.5117, "step": 363 }, { "epoch": 0.166895919303072, "grad_norm": 0.5645924806594849, "learning_rate": 0.0009518109007826734, "loss": 2.4748, "step": 364 }, { "epoch": 0.16735442457588262, "grad_norm": 0.5755054950714111, "learning_rate": 0.0009514922837363059, "loss": 2.577, "step": 365 }, { "epoch": 0.16781292984869325, "grad_norm": 0.5164499878883362, "learning_rate": 0.0009511726705295817, "loss": 2.5744, "step": 366 }, { "epoch": 0.1682714351215039, "grad_norm": 0.5702193975448608, "learning_rate": 0.000950852061867687, "loss": 2.5495, "step": 367 }, { "epoch": 0.16872994039431455, "grad_norm": 0.5885387063026428, "learning_rate": 0.0009505304584580038, "loss": 2.5279, "step": 368 }, { "epoch": 0.16918844566712518, "grad_norm": 0.575178325176239, "learning_rate": 0.0009502078610101092, "loss": 2.6664, "step": 369 }, { "epoch": 0.16964695093993581, "grad_norm": 0.5362281203269958, "learning_rate": 0.0009498842702357736, "loss": 2.6107, "step": 370 }, { "epoch": 0.17010545621274645, "grad_norm": 0.609125554561615, "learning_rate": 0.0009495596868489587, "loss": 2.5695, "step": 371 }, { "epoch": 0.17056396148555708, "grad_norm": 0.5674661993980408, "learning_rate": 0.0009492341115658167, "loss": 2.5744, "step": 372 }, { "epoch": 0.1710224667583677, "grad_norm": 0.5334498882293701, "learning_rate": 0.0009489075451046879, "loss": 2.5198, "step": 373 }, { "epoch": 0.17148097203117835, "grad_norm": 0.5488100051879883, "learning_rate": 0.0009485799881861, "loss": 2.4441, "step": 374 }, { "epoch": 0.171939477303989, "grad_norm": 0.5720640420913696, "learning_rate": 0.0009482514415327654, "loss": 2.5804, "step": 375 }, { "epoch": 0.17239798257679964, "grad_norm": 0.5555192828178406, "learning_rate": 0.000947921905869581, "loss": 2.5472, "step": 376 }, { "epoch": 0.17285648784961027, "grad_norm": 0.610795259475708, "learning_rate": 0.0009475913819236248, "loss": 2.6648, "step": 377 }, { "epoch": 0.1733149931224209, "grad_norm": 0.5508929491043091, "learning_rate": 0.0009472598704241561, "loss": 2.5945, "step": 378 }, { "epoch": 0.17377349839523154, "grad_norm": 0.6799100637435913, "learning_rate": 0.0009469273721026131, "loss": 2.4822, "step": 379 }, { "epoch": 0.17423200366804217, "grad_norm": 0.5537827014923096, "learning_rate": 0.0009465938876926111, "loss": 2.4574, "step": 380 }, { "epoch": 0.1746905089408528, "grad_norm": 0.7075497508049011, "learning_rate": 0.0009462594179299406, "loss": 2.7207, "step": 381 }, { "epoch": 0.17514901421366347, "grad_norm": 0.6364816427230835, "learning_rate": 0.0009459239635525672, "loss": 2.5628, "step": 382 }, { "epoch": 0.1756075194864741, "grad_norm": 0.5640485286712646, "learning_rate": 0.0009455875253006281, "loss": 2.5683, "step": 383 }, { "epoch": 0.17606602475928473, "grad_norm": 0.7373421788215637, "learning_rate": 0.0009452501039164315, "loss": 2.659, "step": 384 }, { "epoch": 0.17652453003209537, "grad_norm": 0.5834930539131165, "learning_rate": 0.0009449117001444549, "loss": 2.4739, "step": 385 }, { "epoch": 0.176983035304906, "grad_norm": 0.5889790654182434, "learning_rate": 0.0009445723147313433, "loss": 2.5964, "step": 386 }, { "epoch": 0.17744154057771663, "grad_norm": 0.6707600951194763, "learning_rate": 0.0009442319484259074, "loss": 2.5548, "step": 387 }, { "epoch": 0.1779000458505273, "grad_norm": 0.6110239028930664, "learning_rate": 0.0009438906019791222, "loss": 2.6153, "step": 388 }, { "epoch": 0.17835855112333793, "grad_norm": 0.5651149749755859, "learning_rate": 0.0009435482761441251, "loss": 2.5925, "step": 389 }, { "epoch": 0.17881705639614856, "grad_norm": 0.5925535559654236, "learning_rate": 0.000943204971676215, "loss": 2.6464, "step": 390 }, { "epoch": 0.1792755616689592, "grad_norm": 0.5651217699050903, "learning_rate": 0.0009428606893328493, "loss": 2.4937, "step": 391 }, { "epoch": 0.17973406694176983, "grad_norm": 0.5302489399909973, "learning_rate": 0.0009425154298736432, "loss": 2.5789, "step": 392 }, { "epoch": 0.18019257221458046, "grad_norm": 0.6552790999412537, "learning_rate": 0.0009421691940603678, "loss": 2.5294, "step": 393 }, { "epoch": 0.1806510774873911, "grad_norm": 0.6294134855270386, "learning_rate": 0.0009418219826569488, "loss": 2.6865, "step": 394 }, { "epoch": 0.18110958276020175, "grad_norm": 0.5430222749710083, "learning_rate": 0.0009414737964294635, "loss": 2.4537, "step": 395 }, { "epoch": 0.1815680880330124, "grad_norm": 0.648733377456665, "learning_rate": 0.000941124636146141, "loss": 2.5384, "step": 396 }, { "epoch": 0.18202659330582302, "grad_norm": 0.5951704978942871, "learning_rate": 0.0009407745025773589, "loss": 2.4399, "step": 397 }, { "epoch": 0.18248509857863365, "grad_norm": 0.539134681224823, "learning_rate": 0.0009404233964956423, "loss": 2.3742, "step": 398 }, { "epoch": 0.1829436038514443, "grad_norm": 0.599242627620697, "learning_rate": 0.0009400713186756625, "loss": 2.5025, "step": 399 }, { "epoch": 0.18340210912425492, "grad_norm": 0.6046454310417175, "learning_rate": 0.0009397182698942342, "loss": 2.5384, "step": 400 }, { "epoch": 0.18386061439706555, "grad_norm": 0.49774590134620667, "learning_rate": 0.0009393642509303149, "loss": 2.5237, "step": 401 }, { "epoch": 0.1843191196698762, "grad_norm": 0.6292476058006287, "learning_rate": 0.0009390092625650023, "loss": 2.5562, "step": 402 }, { "epoch": 0.18477762494268685, "grad_norm": 0.5865110754966736, "learning_rate": 0.0009386533055815332, "loss": 2.5005, "step": 403 }, { "epoch": 0.18523613021549748, "grad_norm": 0.5399919152259827, "learning_rate": 0.0009382963807652813, "loss": 2.5091, "step": 404 }, { "epoch": 0.1856946354883081, "grad_norm": 0.5731890797615051, "learning_rate": 0.000937938488903756, "loss": 2.5694, "step": 405 }, { "epoch": 0.18615314076111875, "grad_norm": 0.5865057706832886, "learning_rate": 0.0009375796307866003, "loss": 2.6005, "step": 406 }, { "epoch": 0.18661164603392938, "grad_norm": 0.5505797266960144, "learning_rate": 0.0009372198072055888, "loss": 2.5809, "step": 407 }, { "epoch": 0.18707015130674004, "grad_norm": 0.596774697303772, "learning_rate": 0.0009368590189546268, "loss": 2.4929, "step": 408 }, { "epoch": 0.18752865657955067, "grad_norm": 0.5412145256996155, "learning_rate": 0.0009364972668297474, "loss": 2.5476, "step": 409 }, { "epoch": 0.1879871618523613, "grad_norm": 0.5468689203262329, "learning_rate": 0.0009361345516291111, "loss": 2.5157, "step": 410 }, { "epoch": 0.18844566712517194, "grad_norm": 0.5968159437179565, "learning_rate": 0.0009357708741530024, "loss": 2.5796, "step": 411 }, { "epoch": 0.18890417239798257, "grad_norm": 0.5981835722923279, "learning_rate": 0.00093540623520383, "loss": 2.4987, "step": 412 }, { "epoch": 0.1893626776707932, "grad_norm": 0.5652632713317871, "learning_rate": 0.000935040635586123, "loss": 2.5418, "step": 413 }, { "epoch": 0.18982118294360384, "grad_norm": 0.5192498564720154, "learning_rate": 0.0009346740761065305, "loss": 2.5079, "step": 414 }, { "epoch": 0.1902796882164145, "grad_norm": 0.5427750945091248, "learning_rate": 0.0009343065575738197, "loss": 2.484, "step": 415 }, { "epoch": 0.19073819348922513, "grad_norm": 0.5291280746459961, "learning_rate": 0.0009339380807988733, "loss": 2.5574, "step": 416 }, { "epoch": 0.19119669876203577, "grad_norm": 0.5242193341255188, "learning_rate": 0.0009335686465946887, "loss": 2.5714, "step": 417 }, { "epoch": 0.1916552040348464, "grad_norm": 0.5996776223182678, "learning_rate": 0.0009331982557763754, "loss": 2.6054, "step": 418 }, { "epoch": 0.19211370930765703, "grad_norm": 0.535576343536377, "learning_rate": 0.0009328269091611537, "loss": 2.4219, "step": 419 }, { "epoch": 0.19257221458046767, "grad_norm": 0.5485090613365173, "learning_rate": 0.0009324546075683524, "loss": 2.6273, "step": 420 }, { "epoch": 0.19303071985327833, "grad_norm": 0.5857663750648499, "learning_rate": 0.0009320813518194083, "loss": 2.5626, "step": 421 }, { "epoch": 0.19348922512608896, "grad_norm": 0.543411135673523, "learning_rate": 0.0009317071427378624, "loss": 2.4671, "step": 422 }, { "epoch": 0.1939477303988996, "grad_norm": 0.5718532800674438, "learning_rate": 0.0009313319811493594, "loss": 2.4983, "step": 423 }, { "epoch": 0.19440623567171023, "grad_norm": 0.6025922298431396, "learning_rate": 0.000930955867881646, "loss": 2.6215, "step": 424 }, { "epoch": 0.19486474094452086, "grad_norm": 0.5608757734298706, "learning_rate": 0.0009305788037645681, "loss": 2.4778, "step": 425 }, { "epoch": 0.1953232462173315, "grad_norm": 0.5591913461685181, "learning_rate": 0.0009302007896300697, "loss": 2.573, "step": 426 }, { "epoch": 0.19578175149014213, "grad_norm": 0.5269467234611511, "learning_rate": 0.0009298218263121911, "loss": 2.5272, "step": 427 }, { "epoch": 0.19624025676295279, "grad_norm": 0.5396514534950256, "learning_rate": 0.0009294419146470668, "loss": 2.5629, "step": 428 }, { "epoch": 0.19669876203576342, "grad_norm": 0.5113781094551086, "learning_rate": 0.0009290610554729234, "loss": 2.5149, "step": 429 }, { "epoch": 0.19715726730857405, "grad_norm": 0.5556777715682983, "learning_rate": 0.0009286792496300784, "loss": 2.5821, "step": 430 }, { "epoch": 0.19761577258138469, "grad_norm": 0.5331014394760132, "learning_rate": 0.0009282964979609379, "loss": 2.5139, "step": 431 }, { "epoch": 0.19807427785419532, "grad_norm": 0.5384894609451294, "learning_rate": 0.0009279128013099947, "loss": 2.4829, "step": 432 }, { "epoch": 0.19853278312700595, "grad_norm": 0.5790057182312012, "learning_rate": 0.0009275281605238268, "loss": 2.5405, "step": 433 }, { "epoch": 0.19899128839981658, "grad_norm": 0.5751925110816956, "learning_rate": 0.0009271425764510953, "loss": 2.6007, "step": 434 }, { "epoch": 0.19944979367262725, "grad_norm": 0.5120574831962585, "learning_rate": 0.0009267560499425423, "loss": 2.4147, "step": 435 }, { "epoch": 0.19990829894543788, "grad_norm": 0.5448221564292908, "learning_rate": 0.0009263685818509895, "loss": 2.4435, "step": 436 }, { "epoch": 0.2003668042182485, "grad_norm": 0.5894081592559814, "learning_rate": 0.000925980173031336, "loss": 2.5307, "step": 437 }, { "epoch": 0.20082530949105915, "grad_norm": 0.5890480875968933, "learning_rate": 0.0009255908243405567, "loss": 2.488, "step": 438 }, { "epoch": 0.20128381476386978, "grad_norm": 0.5528947710990906, "learning_rate": 0.0009252005366376996, "loss": 2.6474, "step": 439 }, { "epoch": 0.2017423200366804, "grad_norm": 0.6188745498657227, "learning_rate": 0.0009248093107838852, "loss": 2.5252, "step": 440 }, { "epoch": 0.20220082530949107, "grad_norm": 0.519440770149231, "learning_rate": 0.0009244171476423036, "loss": 2.517, "step": 441 }, { "epoch": 0.2026593305823017, "grad_norm": 0.5096246004104614, "learning_rate": 0.0009240240480782129, "loss": 2.436, "step": 442 }, { "epoch": 0.20311783585511234, "grad_norm": 0.5687729716300964, "learning_rate": 0.0009236300129589376, "loss": 2.6328, "step": 443 }, { "epoch": 0.20357634112792297, "grad_norm": 0.550613284111023, "learning_rate": 0.0009232350431538657, "loss": 2.6133, "step": 444 }, { "epoch": 0.2040348464007336, "grad_norm": 0.5400554537773132, "learning_rate": 0.0009228391395344482, "loss": 2.4954, "step": 445 }, { "epoch": 0.20449335167354424, "grad_norm": 0.4928134083747864, "learning_rate": 0.000922442302974196, "loss": 2.5963, "step": 446 }, { "epoch": 0.20495185694635487, "grad_norm": 0.5219937562942505, "learning_rate": 0.0009220445343486785, "loss": 2.5325, "step": 447 }, { "epoch": 0.20541036221916553, "grad_norm": 0.5145986080169678, "learning_rate": 0.0009216458345355217, "loss": 2.5354, "step": 448 }, { "epoch": 0.20586886749197617, "grad_norm": 0.5352473258972168, "learning_rate": 0.0009212462044144061, "loss": 2.592, "step": 449 }, { "epoch": 0.2063273727647868, "grad_norm": 0.5252971053123474, "learning_rate": 0.0009208456448670648, "loss": 2.5023, "step": 450 }, { "epoch": 0.20678587803759743, "grad_norm": 0.5064184069633484, "learning_rate": 0.0009204441567772816, "loss": 2.5748, "step": 451 }, { "epoch": 0.20724438331040806, "grad_norm": 0.5311024188995361, "learning_rate": 0.0009200417410308888, "loss": 2.51, "step": 452 }, { "epoch": 0.2077028885832187, "grad_norm": 0.5557432174682617, "learning_rate": 0.0009196383985157656, "loss": 2.5555, "step": 453 }, { "epoch": 0.20816139385602933, "grad_norm": 0.5421399474143982, "learning_rate": 0.000919234130121836, "loss": 2.4679, "step": 454 }, { "epoch": 0.20861989912884, "grad_norm": 0.553537905216217, "learning_rate": 0.0009188289367410672, "loss": 2.4899, "step": 455 }, { "epoch": 0.20907840440165062, "grad_norm": 0.5620190501213074, "learning_rate": 0.0009184228192674666, "loss": 2.4643, "step": 456 }, { "epoch": 0.20953690967446126, "grad_norm": 0.5662469863891602, "learning_rate": 0.0009180157785970808, "loss": 2.615, "step": 457 }, { "epoch": 0.2099954149472719, "grad_norm": 0.5100295543670654, "learning_rate": 0.0009176078156279932, "loss": 2.4623, "step": 458 }, { "epoch": 0.21045392022008252, "grad_norm": 0.5270982980728149, "learning_rate": 0.0009171989312603226, "loss": 2.5047, "step": 459 }, { "epoch": 0.21091242549289316, "grad_norm": 0.5312187671661377, "learning_rate": 0.0009167891263962202, "loss": 2.5426, "step": 460 }, { "epoch": 0.21137093076570382, "grad_norm": 0.5393531918525696, "learning_rate": 0.0009163784019398685, "loss": 2.544, "step": 461 }, { "epoch": 0.21182943603851445, "grad_norm": 0.5723551511764526, "learning_rate": 0.0009159667587974785, "loss": 2.6384, "step": 462 }, { "epoch": 0.21228794131132508, "grad_norm": 0.5142784714698792, "learning_rate": 0.0009155541978772887, "loss": 2.4493, "step": 463 }, { "epoch": 0.21274644658413572, "grad_norm": 0.5554559826850891, "learning_rate": 0.0009151407200895625, "loss": 2.4119, "step": 464 }, { "epoch": 0.21320495185694635, "grad_norm": 0.5520978569984436, "learning_rate": 0.000914726326346586, "loss": 2.465, "step": 465 }, { "epoch": 0.21366345712975698, "grad_norm": 0.5879413485527039, "learning_rate": 0.0009143110175626661, "loss": 2.5952, "step": 466 }, { "epoch": 0.21412196240256762, "grad_norm": 0.5624451637268066, "learning_rate": 0.0009138947946541291, "loss": 2.635, "step": 467 }, { "epoch": 0.21458046767537828, "grad_norm": 0.535145103931427, "learning_rate": 0.0009134776585393181, "loss": 2.5061, "step": 468 }, { "epoch": 0.2150389729481889, "grad_norm": 0.5354058146476746, "learning_rate": 0.0009130596101385906, "loss": 2.5643, "step": 469 }, { "epoch": 0.21549747822099954, "grad_norm": 0.4816924035549164, "learning_rate": 0.0009126406503743174, "loss": 2.5257, "step": 470 }, { "epoch": 0.21595598349381018, "grad_norm": 0.5290162563323975, "learning_rate": 0.0009122207801708802, "loss": 2.4185, "step": 471 }, { "epoch": 0.2164144887666208, "grad_norm": 0.5376173257827759, "learning_rate": 0.0009118000004546689, "loss": 2.5575, "step": 472 }, { "epoch": 0.21687299403943144, "grad_norm": 0.5183917284011841, "learning_rate": 0.0009113783121540807, "loss": 2.4625, "step": 473 }, { "epoch": 0.2173314993122421, "grad_norm": 0.5453611016273499, "learning_rate": 0.0009109557161995172, "loss": 2.5781, "step": 474 }, { "epoch": 0.21779000458505274, "grad_norm": 0.4953596293926239, "learning_rate": 0.0009105322135233828, "loss": 2.5689, "step": 475 }, { "epoch": 0.21824850985786337, "grad_norm": 0.620176374912262, "learning_rate": 0.0009101078050600821, "loss": 2.4945, "step": 476 }, { "epoch": 0.218707015130674, "grad_norm": 0.5294750928878784, "learning_rate": 0.0009096824917460186, "loss": 2.5496, "step": 477 }, { "epoch": 0.21916552040348464, "grad_norm": 0.5557565689086914, "learning_rate": 0.0009092562745195921, "loss": 2.4857, "step": 478 }, { "epoch": 0.21962402567629527, "grad_norm": 0.5819594264030457, "learning_rate": 0.0009088291543211967, "loss": 2.5284, "step": 479 }, { "epoch": 0.2200825309491059, "grad_norm": 0.5852700471878052, "learning_rate": 0.0009084011320932188, "loss": 2.5141, "step": 480 }, { "epoch": 0.22054103622191656, "grad_norm": 0.5862120985984802, "learning_rate": 0.0009079722087800352, "loss": 2.4728, "step": 481 }, { "epoch": 0.2209995414947272, "grad_norm": 0.5905824303627014, "learning_rate": 0.0009075423853280106, "loss": 2.5512, "step": 482 }, { "epoch": 0.22145804676753783, "grad_norm": 0.5095152854919434, "learning_rate": 0.0009071116626854958, "loss": 2.5253, "step": 483 }, { "epoch": 0.22191655204034846, "grad_norm": 0.588388979434967, "learning_rate": 0.0009066800418028256, "loss": 2.5964, "step": 484 }, { "epoch": 0.2223750573131591, "grad_norm": 0.508466899394989, "learning_rate": 0.0009062475236323168, "loss": 2.5408, "step": 485 }, { "epoch": 0.22283356258596973, "grad_norm": 0.5260148644447327, "learning_rate": 0.0009058141091282656, "loss": 2.494, "step": 486 }, { "epoch": 0.22329206785878036, "grad_norm": 0.5795764327049255, "learning_rate": 0.0009053797992469461, "loss": 2.5431, "step": 487 }, { "epoch": 0.22375057313159102, "grad_norm": 0.5151141881942749, "learning_rate": 0.0009049445949466078, "loss": 2.5106, "step": 488 }, { "epoch": 0.22420907840440166, "grad_norm": 0.5401947498321533, "learning_rate": 0.0009045084971874737, "loss": 2.481, "step": 489 }, { "epoch": 0.2246675836772123, "grad_norm": 0.6477903127670288, "learning_rate": 0.0009040715069317382, "loss": 2.456, "step": 490 }, { "epoch": 0.22512608895002292, "grad_norm": 0.5201408863067627, "learning_rate": 0.0009036336251435648, "loss": 2.5049, "step": 491 }, { "epoch": 0.22558459422283356, "grad_norm": 0.5675253868103027, "learning_rate": 0.0009031948527890839, "loss": 2.4634, "step": 492 }, { "epoch": 0.2260430994956442, "grad_norm": 0.5311404466629028, "learning_rate": 0.000902755190836391, "loss": 2.5346, "step": 493 }, { "epoch": 0.22650160476845485, "grad_norm": 0.538214385509491, "learning_rate": 0.0009023146402555442, "loss": 2.5081, "step": 494 }, { "epoch": 0.22696011004126548, "grad_norm": 0.5664333701133728, "learning_rate": 0.0009018732020185624, "loss": 2.4891, "step": 495 }, { "epoch": 0.22741861531407612, "grad_norm": 0.5465412139892578, "learning_rate": 0.0009014308770994235, "loss": 2.5475, "step": 496 }, { "epoch": 0.22787712058688675, "grad_norm": 0.5316890478134155, "learning_rate": 0.0009009876664740605, "loss": 2.593, "step": 497 }, { "epoch": 0.22833562585969738, "grad_norm": 0.5560000538825989, "learning_rate": 0.0009005435711203618, "loss": 2.5972, "step": 498 }, { "epoch": 0.22879413113250802, "grad_norm": 0.5882994532585144, "learning_rate": 0.000900098592018167, "loss": 2.5436, "step": 499 }, { "epoch": 0.22925263640531865, "grad_norm": 0.5104226469993591, "learning_rate": 0.0008996527301492663, "loss": 2.4333, "step": 500 }, { "epoch": 0.2297111416781293, "grad_norm": 0.5770875811576843, "learning_rate": 0.0008992059864973972, "loss": 2.482, "step": 501 }, { "epoch": 0.23016964695093994, "grad_norm": 0.5763022899627686, "learning_rate": 0.0008987583620482427, "loss": 2.4023, "step": 502 }, { "epoch": 0.23062815222375058, "grad_norm": 0.601789116859436, "learning_rate": 0.0008983098577894292, "loss": 2.5238, "step": 503 }, { "epoch": 0.2310866574965612, "grad_norm": 0.5488766431808472, "learning_rate": 0.0008978604747105246, "loss": 2.4631, "step": 504 }, { "epoch": 0.23154516276937184, "grad_norm": 0.5194130539894104, "learning_rate": 0.0008974102138030354, "loss": 2.6442, "step": 505 }, { "epoch": 0.23200366804218248, "grad_norm": 0.7235816717147827, "learning_rate": 0.000896959076060405, "loss": 2.4816, "step": 506 }, { "epoch": 0.2324621733149931, "grad_norm": 0.5202719569206238, "learning_rate": 0.0008965070624780116, "loss": 2.4796, "step": 507 }, { "epoch": 0.23292067858780377, "grad_norm": 0.5158277750015259, "learning_rate": 0.0008960541740531658, "loss": 2.4521, "step": 508 }, { "epoch": 0.2333791838606144, "grad_norm": 0.6098846793174744, "learning_rate": 0.0008956004117851083, "loss": 2.4258, "step": 509 }, { "epoch": 0.23383768913342504, "grad_norm": 0.6221984028816223, "learning_rate": 0.0008951457766750079, "loss": 2.6042, "step": 510 }, { "epoch": 0.23429619440623567, "grad_norm": 0.5208629965782166, "learning_rate": 0.0008946902697259593, "loss": 2.4595, "step": 511 }, { "epoch": 0.2347546996790463, "grad_norm": 0.5231128931045532, "learning_rate": 0.0008942338919429805, "loss": 2.4762, "step": 512 }, { "epoch": 0.23521320495185694, "grad_norm": 0.6004589200019836, "learning_rate": 0.0008937766443330113, "loss": 2.5031, "step": 513 }, { "epoch": 0.2356717102246676, "grad_norm": 0.5673144459724426, "learning_rate": 0.0008933185279049103, "loss": 2.5718, "step": 514 }, { "epoch": 0.23613021549747823, "grad_norm": 0.5157675743103027, "learning_rate": 0.0008928595436694532, "loss": 2.4659, "step": 515 }, { "epoch": 0.23658872077028886, "grad_norm": 0.5752490758895874, "learning_rate": 0.0008923996926393305, "loss": 2.5159, "step": 516 }, { "epoch": 0.2370472260430995, "grad_norm": 0.5071921348571777, "learning_rate": 0.0008919389758291449, "loss": 2.3861, "step": 517 }, { "epoch": 0.23750573131591013, "grad_norm": 0.4951089918613434, "learning_rate": 0.0008914773942554098, "loss": 2.5325, "step": 518 }, { "epoch": 0.23796423658872076, "grad_norm": 0.5067383050918579, "learning_rate": 0.000891014948936546, "loss": 2.529, "step": 519 }, { "epoch": 0.2384227418615314, "grad_norm": 0.5437703132629395, "learning_rate": 0.0008905516408928804, "loss": 2.544, "step": 520 }, { "epoch": 0.23888124713434206, "grad_norm": 0.559182345867157, "learning_rate": 0.0008900874711466434, "loss": 2.5697, "step": 521 }, { "epoch": 0.2393397524071527, "grad_norm": 0.6132202744483948, "learning_rate": 0.0008896224407219666, "loss": 2.4351, "step": 522 }, { "epoch": 0.23979825767996332, "grad_norm": 0.5843932032585144, "learning_rate": 0.0008891565506448804, "loss": 2.4587, "step": 523 }, { "epoch": 0.24025676295277396, "grad_norm": 0.5678679943084717, "learning_rate": 0.0008886898019433122, "loss": 2.4215, "step": 524 }, { "epoch": 0.2407152682255846, "grad_norm": 0.5783699154853821, "learning_rate": 0.0008882221956470836, "loss": 2.5597, "step": 525 }, { "epoch": 0.24117377349839522, "grad_norm": 0.5550544261932373, "learning_rate": 0.0008877537327879086, "loss": 2.492, "step": 526 }, { "epoch": 0.24163227877120586, "grad_norm": 0.5496196150779724, "learning_rate": 0.0008872844143993908, "loss": 2.4319, "step": 527 }, { "epoch": 0.24209078404401652, "grad_norm": 0.523337185382843, "learning_rate": 0.0008868142415170218, "loss": 2.4529, "step": 528 }, { "epoch": 0.24254928931682715, "grad_norm": 0.5119760036468506, "learning_rate": 0.0008863432151781781, "loss": 2.5248, "step": 529 }, { "epoch": 0.24300779458963778, "grad_norm": 0.529503345489502, "learning_rate": 0.0008858713364221195, "loss": 2.48, "step": 530 }, { "epoch": 0.24346629986244842, "grad_norm": 0.5714029669761658, "learning_rate": 0.0008853986062899868, "loss": 2.5274, "step": 531 }, { "epoch": 0.24392480513525905, "grad_norm": 0.5238893032073975, "learning_rate": 0.0008849250258247986, "loss": 2.5938, "step": 532 }, { "epoch": 0.24438331040806968, "grad_norm": 0.5141520500183105, "learning_rate": 0.0008844505960714503, "loss": 2.4108, "step": 533 }, { "epoch": 0.24484181568088034, "grad_norm": 0.538997232913971, "learning_rate": 0.0008839753180767108, "loss": 2.5305, "step": 534 }, { "epoch": 0.24530032095369098, "grad_norm": 0.5396038293838501, "learning_rate": 0.0008834991928892204, "loss": 2.5989, "step": 535 }, { "epoch": 0.2457588262265016, "grad_norm": 0.5229591131210327, "learning_rate": 0.000883022221559489, "loss": 2.5276, "step": 536 }, { "epoch": 0.24621733149931224, "grad_norm": 0.5420697927474976, "learning_rate": 0.0008825444051398934, "loss": 2.6101, "step": 537 }, { "epoch": 0.24667583677212288, "grad_norm": 0.5033102035522461, "learning_rate": 0.0008820657446846745, "loss": 2.5574, "step": 538 }, { "epoch": 0.2471343420449335, "grad_norm": 0.47626206278800964, "learning_rate": 0.000881586241249936, "loss": 2.3981, "step": 539 }, { "epoch": 0.24759284731774414, "grad_norm": 0.4836049973964691, "learning_rate": 0.0008811058958936411, "loss": 2.4967, "step": 540 }, { "epoch": 0.2480513525905548, "grad_norm": 0.5031061172485352, "learning_rate": 0.000880624709675611, "loss": 2.3832, "step": 541 }, { "epoch": 0.24850985786336544, "grad_norm": 0.47544097900390625, "learning_rate": 0.000880142683657522, "loss": 2.5574, "step": 542 }, { "epoch": 0.24896836313617607, "grad_norm": 0.47325387597084045, "learning_rate": 0.0008796598189029029, "loss": 2.4341, "step": 543 }, { "epoch": 0.2494268684089867, "grad_norm": 0.5193619132041931, "learning_rate": 0.0008791761164771338, "loss": 2.4623, "step": 544 }, { "epoch": 0.24988537368179733, "grad_norm": 0.5026707649230957, "learning_rate": 0.0008786915774474424, "loss": 2.5141, "step": 545 }, { "epoch": 0.25034387895460797, "grad_norm": 0.5422954559326172, "learning_rate": 0.0008782062028829027, "loss": 2.6304, "step": 546 }, { "epoch": 0.25080238422741863, "grad_norm": 0.4968990385532379, "learning_rate": 0.0008777199938544318, "loss": 2.4465, "step": 547 }, { "epoch": 0.25126088950022923, "grad_norm": 0.47817763686180115, "learning_rate": 0.0008772329514347883, "loss": 2.436, "step": 548 }, { "epoch": 0.2517193947730399, "grad_norm": 0.5454387664794922, "learning_rate": 0.0008767450766985694, "loss": 2.586, "step": 549 }, { "epoch": 0.2521779000458505, "grad_norm": 0.4939558804035187, "learning_rate": 0.0008762563707222086, "loss": 2.4664, "step": 550 }, { "epoch": 0.25263640531866116, "grad_norm": 0.52228844165802, "learning_rate": 0.0008757668345839738, "loss": 2.5376, "step": 551 }, { "epoch": 0.2530949105914718, "grad_norm": 0.5036587119102478, "learning_rate": 0.0008752764693639638, "loss": 2.5275, "step": 552 }, { "epoch": 0.2535534158642824, "grad_norm": 0.5437069535255432, "learning_rate": 0.0008747852761441078, "loss": 2.6074, "step": 553 }, { "epoch": 0.2540119211370931, "grad_norm": 0.5377963185310364, "learning_rate": 0.0008742932560081607, "loss": 2.5496, "step": 554 }, { "epoch": 0.2544704264099037, "grad_norm": 0.5116319060325623, "learning_rate": 0.0008738004100417025, "loss": 2.5223, "step": 555 }, { "epoch": 0.25492893168271435, "grad_norm": 0.5447186827659607, "learning_rate": 0.0008733067393321355, "loss": 2.5327, "step": 556 }, { "epoch": 0.255387436955525, "grad_norm": 0.48498019576072693, "learning_rate": 0.000872812244968681, "loss": 2.523, "step": 557 }, { "epoch": 0.2558459422283356, "grad_norm": 0.5081266760826111, "learning_rate": 0.0008723169280423783, "loss": 2.5681, "step": 558 }, { "epoch": 0.2563044475011463, "grad_norm": 0.5406227111816406, "learning_rate": 0.0008718207896460811, "loss": 2.5185, "step": 559 }, { "epoch": 0.2567629527739569, "grad_norm": 0.5006083250045776, "learning_rate": 0.0008713238308744557, "loss": 2.4695, "step": 560 }, { "epoch": 0.25722145804676755, "grad_norm": 0.5412190556526184, "learning_rate": 0.0008708260528239789, "loss": 2.4488, "step": 561 }, { "epoch": 0.25767996331957815, "grad_norm": 0.5273689031600952, "learning_rate": 0.000870327456592934, "loss": 2.4553, "step": 562 }, { "epoch": 0.2581384685923888, "grad_norm": 0.5407193303108215, "learning_rate": 0.0008698280432814107, "loss": 2.548, "step": 563 }, { "epoch": 0.2585969738651995, "grad_norm": 0.5686867237091064, "learning_rate": 0.000869327813991301, "loss": 2.5687, "step": 564 }, { "epoch": 0.2590554791380101, "grad_norm": 0.5542526245117188, "learning_rate": 0.0008688267698262971, "loss": 2.5387, "step": 565 }, { "epoch": 0.25951398441082074, "grad_norm": 0.5438316464424133, "learning_rate": 0.0008683249118918894, "loss": 2.5002, "step": 566 }, { "epoch": 0.25997248968363135, "grad_norm": 0.5591607093811035, "learning_rate": 0.0008678222412953637, "loss": 2.4857, "step": 567 }, { "epoch": 0.260430994956442, "grad_norm": 0.5316394567489624, "learning_rate": 0.0008673187591457987, "loss": 2.6189, "step": 568 }, { "epoch": 0.2608895002292526, "grad_norm": 0.5961428880691528, "learning_rate": 0.0008668144665540639, "loss": 2.5333, "step": 569 }, { "epoch": 0.2613480055020633, "grad_norm": 0.5136201977729797, "learning_rate": 0.0008663093646328167, "loss": 2.5578, "step": 570 }, { "epoch": 0.26180651077487394, "grad_norm": 0.5163459181785583, "learning_rate": 0.0008658034544965003, "loss": 2.3764, "step": 571 }, { "epoch": 0.26226501604768454, "grad_norm": 0.5454079508781433, "learning_rate": 0.0008652967372613412, "loss": 2.4726, "step": 572 }, { "epoch": 0.2627235213204952, "grad_norm": 0.5121014714241028, "learning_rate": 0.0008647892140453466, "loss": 2.3997, "step": 573 }, { "epoch": 0.2631820265933058, "grad_norm": 0.48196157813072205, "learning_rate": 0.0008642808859683021, "loss": 2.4361, "step": 574 }, { "epoch": 0.26364053186611647, "grad_norm": 0.5400364398956299, "learning_rate": 0.0008637717541517689, "loss": 2.5478, "step": 575 }, { "epoch": 0.2640990371389271, "grad_norm": 0.5700699090957642, "learning_rate": 0.0008632618197190816, "loss": 2.5599, "step": 576 }, { "epoch": 0.26455754241173773, "grad_norm": 0.5675872564315796, "learning_rate": 0.0008627510837953458, "loss": 2.5549, "step": 577 }, { "epoch": 0.2650160476845484, "grad_norm": 0.4928607940673828, "learning_rate": 0.0008622395475074355, "loss": 2.5317, "step": 578 }, { "epoch": 0.265474552957359, "grad_norm": 0.4981940686702728, "learning_rate": 0.0008617272119839903, "loss": 2.5033, "step": 579 }, { "epoch": 0.26593305823016966, "grad_norm": 0.5268711447715759, "learning_rate": 0.0008612140783554136, "loss": 2.6055, "step": 580 }, { "epoch": 0.26639156350298027, "grad_norm": 0.5228254795074463, "learning_rate": 0.0008607001477538696, "loss": 2.4605, "step": 581 }, { "epoch": 0.2668500687757909, "grad_norm": 0.5025787949562073, "learning_rate": 0.0008601854213132807, "loss": 2.573, "step": 582 }, { "epoch": 0.26730857404860153, "grad_norm": 0.4954626262187958, "learning_rate": 0.0008596699001693256, "loss": 2.4877, "step": 583 }, { "epoch": 0.2677670793214122, "grad_norm": 0.5518064498901367, "learning_rate": 0.000859153585459436, "loss": 2.4844, "step": 584 }, { "epoch": 0.26822558459422285, "grad_norm": 0.5043811798095703, "learning_rate": 0.0008586364783227949, "loss": 2.5106, "step": 585 }, { "epoch": 0.26868408986703346, "grad_norm": 0.4880671799182892, "learning_rate": 0.0008581185799003332, "loss": 2.5365, "step": 586 }, { "epoch": 0.2691425951398441, "grad_norm": 0.5009052157402039, "learning_rate": 0.0008575998913347283, "loss": 2.6103, "step": 587 }, { "epoch": 0.2696011004126547, "grad_norm": 0.4757317304611206, "learning_rate": 0.0008570804137704004, "loss": 2.4808, "step": 588 }, { "epoch": 0.2700596056854654, "grad_norm": 0.489135205745697, "learning_rate": 0.0008565601483535108, "loss": 2.4607, "step": 589 }, { "epoch": 0.27051811095827605, "grad_norm": 0.4782392382621765, "learning_rate": 0.0008560390962319591, "loss": 2.4786, "step": 590 }, { "epoch": 0.27097661623108665, "grad_norm": 0.5060367584228516, "learning_rate": 0.0008555172585553804, "loss": 2.5401, "step": 591 }, { "epoch": 0.2714351215038973, "grad_norm": 0.5133999586105347, "learning_rate": 0.0008549946364751435, "loss": 2.4367, "step": 592 }, { "epoch": 0.2718936267767079, "grad_norm": 0.5095196962356567, "learning_rate": 0.0008544712311443475, "loss": 2.5186, "step": 593 }, { "epoch": 0.2723521320495186, "grad_norm": 0.5159398913383484, "learning_rate": 0.0008539470437178196, "loss": 2.4631, "step": 594 }, { "epoch": 0.2728106373223292, "grad_norm": 0.4829711318016052, "learning_rate": 0.000853422075352113, "loss": 2.5025, "step": 595 }, { "epoch": 0.27326914259513985, "grad_norm": 0.46904894709587097, "learning_rate": 0.0008528963272055035, "loss": 2.5288, "step": 596 }, { "epoch": 0.2737276478679505, "grad_norm": 0.49332237243652344, "learning_rate": 0.0008523698004379877, "loss": 2.4629, "step": 597 }, { "epoch": 0.2741861531407611, "grad_norm": 0.5167871117591858, "learning_rate": 0.00085184249621128, "loss": 2.4264, "step": 598 }, { "epoch": 0.2746446584135718, "grad_norm": 0.5271046161651611, "learning_rate": 0.0008513144156888101, "loss": 2.4878, "step": 599 }, { "epoch": 0.2751031636863824, "grad_norm": 0.4967330992221832, "learning_rate": 0.0008507855600357207, "loss": 2.5739, "step": 600 }, { "epoch": 0.27556166895919304, "grad_norm": 0.5140023231506348, "learning_rate": 0.0008502559304188644, "loss": 2.4245, "step": 601 }, { "epoch": 0.27602017423200365, "grad_norm": 0.6132800579071045, "learning_rate": 0.0008497255280068019, "loss": 2.5463, "step": 602 }, { "epoch": 0.2764786795048143, "grad_norm": 0.5084259510040283, "learning_rate": 0.0008491943539697986, "loss": 2.5026, "step": 603 }, { "epoch": 0.27693718477762497, "grad_norm": 0.5507619976997375, "learning_rate": 0.0008486624094798226, "loss": 2.4659, "step": 604 }, { "epoch": 0.2773956900504356, "grad_norm": 0.521020233631134, "learning_rate": 0.0008481296957105417, "loss": 2.5316, "step": 605 }, { "epoch": 0.27785419532324623, "grad_norm": 0.5280611515045166, "learning_rate": 0.0008475962138373213, "loss": 2.5962, "step": 606 }, { "epoch": 0.27831270059605684, "grad_norm": 0.4935886263847351, "learning_rate": 0.0008470619650372211, "loss": 2.4605, "step": 607 }, { "epoch": 0.2787712058688675, "grad_norm": 0.5141266584396362, "learning_rate": 0.0008465269504889934, "loss": 2.475, "step": 608 }, { "epoch": 0.2792297111416781, "grad_norm": 0.4820075035095215, "learning_rate": 0.0008459911713730799, "loss": 2.4896, "step": 609 }, { "epoch": 0.27968821641448877, "grad_norm": 0.5103136897087097, "learning_rate": 0.0008454546288716089, "loss": 2.4875, "step": 610 }, { "epoch": 0.2801467216872994, "grad_norm": 0.5118184685707092, "learning_rate": 0.0008449173241683935, "loss": 2.4422, "step": 611 }, { "epoch": 0.28060522696011003, "grad_norm": 0.563139796257019, "learning_rate": 0.0008443792584489281, "loss": 2.6486, "step": 612 }, { "epoch": 0.2810637322329207, "grad_norm": 0.4854675233364105, "learning_rate": 0.0008438404329003863, "loss": 2.4904, "step": 613 }, { "epoch": 0.2815222375057313, "grad_norm": 0.5297295451164246, "learning_rate": 0.0008433008487116183, "loss": 2.4449, "step": 614 }, { "epoch": 0.28198074277854196, "grad_norm": 0.5569421052932739, "learning_rate": 0.0008427605070731481, "loss": 2.4043, "step": 615 }, { "epoch": 0.28243924805135256, "grad_norm": 0.4893357455730438, "learning_rate": 0.0008422194091771708, "loss": 2.4769, "step": 616 }, { "epoch": 0.2828977533241632, "grad_norm": 0.5305376052856445, "learning_rate": 0.0008416775562175503, "loss": 2.5417, "step": 617 }, { "epoch": 0.2833562585969739, "grad_norm": 0.5125036239624023, "learning_rate": 0.000841134949389816, "loss": 2.4686, "step": 618 }, { "epoch": 0.2838147638697845, "grad_norm": 0.5312923192977905, "learning_rate": 0.0008405915898911611, "loss": 2.5346, "step": 619 }, { "epoch": 0.28427326914259515, "grad_norm": 0.48293235898017883, "learning_rate": 0.0008400474789204396, "loss": 2.5168, "step": 620 }, { "epoch": 0.28473177441540576, "grad_norm": 0.48212724924087524, "learning_rate": 0.0008395026176781626, "loss": 2.4494, "step": 621 }, { "epoch": 0.2851902796882164, "grad_norm": 0.5107107162475586, "learning_rate": 0.0008389570073664976, "loss": 2.5329, "step": 622 }, { "epoch": 0.285648784961027, "grad_norm": 0.49845534563064575, "learning_rate": 0.0008384106491892642, "loss": 2.593, "step": 623 }, { "epoch": 0.2861072902338377, "grad_norm": 0.509881854057312, "learning_rate": 0.0008378635443519327, "loss": 2.4979, "step": 624 }, { "epoch": 0.28656579550664835, "grad_norm": 0.5022458434104919, "learning_rate": 0.0008373156940616199, "loss": 2.4741, "step": 625 }, { "epoch": 0.28702430077945895, "grad_norm": 0.5109903812408447, "learning_rate": 0.0008367670995270882, "loss": 2.5694, "step": 626 }, { "epoch": 0.2874828060522696, "grad_norm": 0.4845179319381714, "learning_rate": 0.0008362177619587416, "loss": 2.4514, "step": 627 }, { "epoch": 0.2879413113250802, "grad_norm": 0.499578595161438, "learning_rate": 0.0008356676825686238, "loss": 2.4534, "step": 628 }, { "epoch": 0.2883998165978909, "grad_norm": 0.4788621962070465, "learning_rate": 0.0008351168625704147, "loss": 2.5912, "step": 629 }, { "epoch": 0.28885832187070154, "grad_norm": 0.5032369494438171, "learning_rate": 0.0008345653031794292, "loss": 2.5132, "step": 630 }, { "epoch": 0.28931682714351215, "grad_norm": 0.5019371509552002, "learning_rate": 0.0008340130056126125, "loss": 2.4244, "step": 631 }, { "epoch": 0.2897753324163228, "grad_norm": 0.4773755669593811, "learning_rate": 0.0008334599710885394, "loss": 2.4869, "step": 632 }, { "epoch": 0.2902338376891334, "grad_norm": 0.5088169574737549, "learning_rate": 0.0008329062008274098, "loss": 2.4016, "step": 633 }, { "epoch": 0.2906923429619441, "grad_norm": 0.5326182842254639, "learning_rate": 0.000832351696051048, "loss": 2.4973, "step": 634 }, { "epoch": 0.2911508482347547, "grad_norm": 0.4539983570575714, "learning_rate": 0.000831796457982898, "loss": 2.4961, "step": 635 }, { "epoch": 0.29160935350756534, "grad_norm": 0.5001047253608704, "learning_rate": 0.0008312404878480222, "loss": 2.5025, "step": 636 }, { "epoch": 0.292067858780376, "grad_norm": 0.4994875192642212, "learning_rate": 0.0008306837868730979, "loss": 2.5162, "step": 637 }, { "epoch": 0.2925263640531866, "grad_norm": 0.48875173926353455, "learning_rate": 0.0008301263562864152, "loss": 2.5136, "step": 638 }, { "epoch": 0.29298486932599727, "grad_norm": 0.531602680683136, "learning_rate": 0.0008295681973178737, "loss": 2.5558, "step": 639 }, { "epoch": 0.29344337459880787, "grad_norm": 0.49160489439964294, "learning_rate": 0.0008290093111989804, "loss": 2.4778, "step": 640 }, { "epoch": 0.29390187987161853, "grad_norm": 0.520937442779541, "learning_rate": 0.0008284496991628465, "loss": 2.462, "step": 641 }, { "epoch": 0.29436038514442914, "grad_norm": 0.48621174693107605, "learning_rate": 0.0008278893624441847, "loss": 2.514, "step": 642 }, { "epoch": 0.2948188904172398, "grad_norm": 0.5071423649787903, "learning_rate": 0.000827328302279307, "loss": 2.4213, "step": 643 }, { "epoch": 0.29527739569005046, "grad_norm": 0.49237170815467834, "learning_rate": 0.0008267665199061211, "loss": 2.4531, "step": 644 }, { "epoch": 0.29573590096286106, "grad_norm": 0.4938497245311737, "learning_rate": 0.0008262040165641288, "loss": 2.5185, "step": 645 }, { "epoch": 0.2961944062356717, "grad_norm": 0.47884687781333923, "learning_rate": 0.0008256407934944219, "loss": 2.4961, "step": 646 }, { "epoch": 0.29665291150848233, "grad_norm": 0.4599781930446625, "learning_rate": 0.0008250768519396807, "loss": 2.4285, "step": 647 }, { "epoch": 0.297111416781293, "grad_norm": 0.4715670645236969, "learning_rate": 0.0008245121931441706, "loss": 2.5279, "step": 648 }, { "epoch": 0.2975699220541036, "grad_norm": 0.4889383316040039, "learning_rate": 0.0008239468183537393, "loss": 2.5586, "step": 649 }, { "epoch": 0.29802842732691426, "grad_norm": 0.5213057994842529, "learning_rate": 0.0008233807288158146, "loss": 2.4556, "step": 650 }, { "epoch": 0.2984869325997249, "grad_norm": 0.48586586117744446, "learning_rate": 0.0008228139257794012, "loss": 2.5455, "step": 651 }, { "epoch": 0.2989454378725355, "grad_norm": 0.5012825131416321, "learning_rate": 0.0008222464104950778, "loss": 2.5004, "step": 652 }, { "epoch": 0.2994039431453462, "grad_norm": 0.5063830018043518, "learning_rate": 0.000821678184214995, "loss": 2.4515, "step": 653 }, { "epoch": 0.2998624484181568, "grad_norm": 0.5231145620346069, "learning_rate": 0.0008211092481928716, "loss": 2.4454, "step": 654 }, { "epoch": 0.30032095369096745, "grad_norm": 0.5180516839027405, "learning_rate": 0.0008205396036839927, "loss": 2.5232, "step": 655 }, { "epoch": 0.30077945896377806, "grad_norm": 0.505483090877533, "learning_rate": 0.0008199692519452069, "loss": 2.5915, "step": 656 }, { "epoch": 0.3012379642365887, "grad_norm": 0.5549357533454895, "learning_rate": 0.0008193981942349224, "loss": 2.4484, "step": 657 }, { "epoch": 0.3016964695093994, "grad_norm": 0.49482086300849915, "learning_rate": 0.0008188264318131056, "loss": 2.4702, "step": 658 }, { "epoch": 0.30215497478221, "grad_norm": 0.503358006477356, "learning_rate": 0.0008182539659412776, "loss": 2.4329, "step": 659 }, { "epoch": 0.30261348005502064, "grad_norm": 0.5417336821556091, "learning_rate": 0.0008176807978825118, "loss": 2.4397, "step": 660 }, { "epoch": 0.30307198532783125, "grad_norm": 0.571476936340332, "learning_rate": 0.0008171069289014306, "loss": 2.449, "step": 661 }, { "epoch": 0.3035304906006419, "grad_norm": 0.4716215431690216, "learning_rate": 0.0008165323602642028, "loss": 2.5638, "step": 662 }, { "epoch": 0.30398899587345257, "grad_norm": 0.5062218308448792, "learning_rate": 0.0008159570932385414, "loss": 2.4999, "step": 663 }, { "epoch": 0.3044475011462632, "grad_norm": 0.5160056948661804, "learning_rate": 0.0008153811290936999, "loss": 2.5432, "step": 664 }, { "epoch": 0.30490600641907384, "grad_norm": 0.4765881299972534, "learning_rate": 0.0008148044691004698, "loss": 2.5176, "step": 665 }, { "epoch": 0.30536451169188444, "grad_norm": 0.47617971897125244, "learning_rate": 0.0008142271145311783, "loss": 2.5195, "step": 666 }, { "epoch": 0.3058230169646951, "grad_norm": 0.523962676525116, "learning_rate": 0.000813649066659685, "loss": 2.5325, "step": 667 }, { "epoch": 0.3062815222375057, "grad_norm": 0.4904899597167969, "learning_rate": 0.0008130703267613787, "loss": 2.4488, "step": 668 }, { "epoch": 0.30674002751031637, "grad_norm": 0.5108135342597961, "learning_rate": 0.0008124908961131759, "loss": 2.513, "step": 669 }, { "epoch": 0.30719853278312703, "grad_norm": 0.4817577600479126, "learning_rate": 0.0008119107759935163, "loss": 2.4451, "step": 670 }, { "epoch": 0.30765703805593764, "grad_norm": 0.4813890755176544, "learning_rate": 0.0008113299676823615, "loss": 2.4962, "step": 671 }, { "epoch": 0.3081155433287483, "grad_norm": 0.5025249123573303, "learning_rate": 0.0008107484724611911, "loss": 2.4543, "step": 672 }, { "epoch": 0.3085740486015589, "grad_norm": 0.4799831211566925, "learning_rate": 0.0008101662916130006, "loss": 2.5263, "step": 673 }, { "epoch": 0.30903255387436956, "grad_norm": 0.5337442755699158, "learning_rate": 0.0008095834264222979, "loss": 2.4869, "step": 674 }, { "epoch": 0.30949105914718017, "grad_norm": 0.4819645285606384, "learning_rate": 0.0008089998781751009, "loss": 2.5954, "step": 675 }, { "epoch": 0.30994956441999083, "grad_norm": 0.4648689031600952, "learning_rate": 0.0008084156481589349, "loss": 2.5083, "step": 676 }, { "epoch": 0.3104080696928015, "grad_norm": 0.48743513226509094, "learning_rate": 0.0008078307376628291, "loss": 2.4552, "step": 677 }, { "epoch": 0.3108665749656121, "grad_norm": 0.4816006124019623, "learning_rate": 0.0008072451479773143, "loss": 2.5533, "step": 678 }, { "epoch": 0.31132508023842276, "grad_norm": 0.46881452202796936, "learning_rate": 0.0008066588803944195, "loss": 2.4642, "step": 679 }, { "epoch": 0.31178358551123336, "grad_norm": 0.4834159016609192, "learning_rate": 0.0008060719362076697, "loss": 2.5469, "step": 680 }, { "epoch": 0.312242090784044, "grad_norm": 0.4680463969707489, "learning_rate": 0.0008054843167120826, "loss": 2.4666, "step": 681 }, { "epoch": 0.31270059605685463, "grad_norm": 0.47788766026496887, "learning_rate": 0.0008048960232041663, "loss": 2.548, "step": 682 }, { "epoch": 0.3131591013296653, "grad_norm": 0.5565969347953796, "learning_rate": 0.0008043070569819153, "loss": 2.4356, "step": 683 }, { "epoch": 0.31361760660247595, "grad_norm": 0.4910687804222107, "learning_rate": 0.0008037174193448089, "loss": 2.4896, "step": 684 }, { "epoch": 0.31407611187528656, "grad_norm": 0.5090759992599487, "learning_rate": 0.0008031271115938077, "loss": 2.5016, "step": 685 }, { "epoch": 0.3145346171480972, "grad_norm": 0.5579278469085693, "learning_rate": 0.0008025361350313505, "loss": 2.5602, "step": 686 }, { "epoch": 0.3149931224209078, "grad_norm": 0.4940716028213501, "learning_rate": 0.0008019444909613523, "loss": 2.501, "step": 687 }, { "epoch": 0.3154516276937185, "grad_norm": 0.4923720955848694, "learning_rate": 0.0008013521806892003, "loss": 2.4022, "step": 688 }, { "epoch": 0.3159101329665291, "grad_norm": 0.4736354947090149, "learning_rate": 0.000800759205521752, "loss": 2.5449, "step": 689 }, { "epoch": 0.31636863823933975, "grad_norm": 0.48131924867630005, "learning_rate": 0.0008001655667673318, "loss": 2.5848, "step": 690 }, { "epoch": 0.3168271435121504, "grad_norm": 0.4661370515823364, "learning_rate": 0.0007995712657357279, "loss": 2.5015, "step": 691 }, { "epoch": 0.317285648784961, "grad_norm": 0.475654661655426, "learning_rate": 0.0007989763037381904, "loss": 2.661, "step": 692 }, { "epoch": 0.3177441540577717, "grad_norm": 0.49579954147338867, "learning_rate": 0.0007983806820874271, "loss": 2.4729, "step": 693 }, { "epoch": 0.3182026593305823, "grad_norm": 0.48459377884864807, "learning_rate": 0.0007977844020976016, "loss": 2.4392, "step": 694 }, { "epoch": 0.31866116460339294, "grad_norm": 0.5070955157279968, "learning_rate": 0.00079718746508433, "loss": 2.5467, "step": 695 }, { "epoch": 0.31911966987620355, "grad_norm": 0.486790269613266, "learning_rate": 0.0007965898723646776, "loss": 2.5048, "step": 696 }, { "epoch": 0.3195781751490142, "grad_norm": 0.4794185161590576, "learning_rate": 0.0007959916252571573, "loss": 2.4623, "step": 697 }, { "epoch": 0.32003668042182487, "grad_norm": 0.48720309138298035, "learning_rate": 0.000795392725081725, "loss": 2.4015, "step": 698 }, { "epoch": 0.3204951856946355, "grad_norm": 0.5142336487770081, "learning_rate": 0.000794793173159778, "loss": 2.5411, "step": 699 }, { "epoch": 0.32095369096744614, "grad_norm": 0.5075454711914062, "learning_rate": 0.0007941929708141513, "loss": 2.584, "step": 700 }, { "epoch": 0.32141219624025674, "grad_norm": 0.48563024401664734, "learning_rate": 0.0007935921193691153, "loss": 2.4649, "step": 701 }, { "epoch": 0.3218707015130674, "grad_norm": 0.4740475118160248, "learning_rate": 0.0007929906201503722, "loss": 2.4656, "step": 702 }, { "epoch": 0.32232920678587806, "grad_norm": 0.49214744567871094, "learning_rate": 0.0007923884744850536, "loss": 2.545, "step": 703 }, { "epoch": 0.32278771205868867, "grad_norm": 0.5218258500099182, "learning_rate": 0.0007917856837017176, "loss": 2.4607, "step": 704 }, { "epoch": 0.32324621733149933, "grad_norm": 0.5280850529670715, "learning_rate": 0.0007911822491303452, "loss": 2.5321, "step": 705 }, { "epoch": 0.32370472260430994, "grad_norm": 0.5054512023925781, "learning_rate": 0.0007905781721023382, "loss": 2.5318, "step": 706 }, { "epoch": 0.3241632278771206, "grad_norm": 0.4976559281349182, "learning_rate": 0.000789973453950516, "loss": 2.3988, "step": 707 }, { "epoch": 0.3246217331499312, "grad_norm": 0.4690145254135132, "learning_rate": 0.000789368096009112, "loss": 2.5212, "step": 708 }, { "epoch": 0.32508023842274186, "grad_norm": 0.46754175424575806, "learning_rate": 0.0007887620996137721, "loss": 2.4128, "step": 709 }, { "epoch": 0.3255387436955525, "grad_norm": 0.4910288155078888, "learning_rate": 0.0007881554661015497, "loss": 2.3914, "step": 710 }, { "epoch": 0.32599724896836313, "grad_norm": 0.5294497609138489, "learning_rate": 0.0007875481968109051, "loss": 2.4985, "step": 711 }, { "epoch": 0.3264557542411738, "grad_norm": 0.4969954192638397, "learning_rate": 0.0007869402930817007, "loss": 2.5555, "step": 712 }, { "epoch": 0.3269142595139844, "grad_norm": 0.47155246138572693, "learning_rate": 0.0007863317562551987, "loss": 2.3934, "step": 713 }, { "epoch": 0.32737276478679506, "grad_norm": 0.4808892607688904, "learning_rate": 0.0007857225876740584, "loss": 2.4498, "step": 714 }, { "epoch": 0.32783127005960566, "grad_norm": 0.4955471456050873, "learning_rate": 0.0007851127886823327, "loss": 2.3767, "step": 715 }, { "epoch": 0.3282897753324163, "grad_norm": 0.5097181797027588, "learning_rate": 0.0007845023606254658, "loss": 2.4626, "step": 716 }, { "epoch": 0.328748280605227, "grad_norm": 0.47530293464660645, "learning_rate": 0.0007838913048502894, "loss": 2.4194, "step": 717 }, { "epoch": 0.3292067858780376, "grad_norm": 0.5498925447463989, "learning_rate": 0.0007832796227050208, "loss": 2.6397, "step": 718 }, { "epoch": 0.32966529115084825, "grad_norm": 0.5018754601478577, "learning_rate": 0.0007826673155392587, "loss": 2.5185, "step": 719 }, { "epoch": 0.33012379642365886, "grad_norm": 0.4556347727775574, "learning_rate": 0.000782054384703981, "loss": 2.3909, "step": 720 }, { "epoch": 0.3305823016964695, "grad_norm": 0.4710378348827362, "learning_rate": 0.0007814408315515418, "loss": 2.4726, "step": 721 }, { "epoch": 0.3310408069692801, "grad_norm": 0.4738503098487854, "learning_rate": 0.0007808266574356683, "loss": 2.5344, "step": 722 }, { "epoch": 0.3314993122420908, "grad_norm": 0.49838986992836, "learning_rate": 0.0007802118637114573, "loss": 2.4397, "step": 723 }, { "epoch": 0.33195781751490144, "grad_norm": 0.5398308634757996, "learning_rate": 0.0007795964517353734, "loss": 2.6203, "step": 724 }, { "epoch": 0.33241632278771205, "grad_norm": 0.47457432746887207, "learning_rate": 0.0007789804228652449, "loss": 2.4597, "step": 725 }, { "epoch": 0.3328748280605227, "grad_norm": 0.5179492831230164, "learning_rate": 0.0007783637784602609, "loss": 2.4176, "step": 726 }, { "epoch": 0.3333333333333333, "grad_norm": 0.46739235520362854, "learning_rate": 0.0007777465198809692, "loss": 2.4798, "step": 727 }, { "epoch": 0.333791838606144, "grad_norm": 0.5625951290130615, "learning_rate": 0.0007771286484892722, "loss": 2.6307, "step": 728 }, { "epoch": 0.3342503438789546, "grad_norm": 0.4784482717514038, "learning_rate": 0.000776510165648425, "loss": 2.5378, "step": 729 }, { "epoch": 0.33470884915176524, "grad_norm": 0.4844149947166443, "learning_rate": 0.0007758910727230311, "loss": 2.4275, "step": 730 }, { "epoch": 0.3351673544245759, "grad_norm": 0.5376376509666443, "learning_rate": 0.0007752713710790404, "loss": 2.461, "step": 731 }, { "epoch": 0.3356258596973865, "grad_norm": 0.532508909702301, "learning_rate": 0.0007746510620837459, "loss": 2.5051, "step": 732 }, { "epoch": 0.33608436497019717, "grad_norm": 0.506984293460846, "learning_rate": 0.0007740301471057807, "loss": 2.4333, "step": 733 }, { "epoch": 0.3365428702430078, "grad_norm": 0.5040539503097534, "learning_rate": 0.0007734086275151146, "loss": 2.4733, "step": 734 }, { "epoch": 0.33700137551581844, "grad_norm": 0.4616777300834656, "learning_rate": 0.0007727865046830517, "loss": 2.3643, "step": 735 }, { "epoch": 0.3374598807886291, "grad_norm": 0.4922655522823334, "learning_rate": 0.0007721637799822269, "loss": 2.5044, "step": 736 }, { "epoch": 0.3379183860614397, "grad_norm": 0.4682179391384125, "learning_rate": 0.0007715404547866032, "loss": 2.4599, "step": 737 }, { "epoch": 0.33837689133425036, "grad_norm": 0.47201722860336304, "learning_rate": 0.0007709165304714685, "loss": 2.5636, "step": 738 }, { "epoch": 0.33883539660706097, "grad_norm": 0.47104138135910034, "learning_rate": 0.0007702920084134324, "loss": 2.5365, "step": 739 }, { "epoch": 0.33929390187987163, "grad_norm": 0.4874970018863678, "learning_rate": 0.0007696668899904236, "loss": 2.4153, "step": 740 }, { "epoch": 0.33975240715268223, "grad_norm": 0.5016466975212097, "learning_rate": 0.0007690411765816864, "loss": 2.4924, "step": 741 }, { "epoch": 0.3402109124254929, "grad_norm": 0.505591869354248, "learning_rate": 0.0007684148695677778, "loss": 2.3876, "step": 742 }, { "epoch": 0.34066941769830356, "grad_norm": 0.5239207744598389, "learning_rate": 0.000767787970330565, "loss": 2.4288, "step": 743 }, { "epoch": 0.34112792297111416, "grad_norm": 0.46158549189567566, "learning_rate": 0.000767160480253221, "loss": 2.5593, "step": 744 }, { "epoch": 0.3415864282439248, "grad_norm": 0.5225353240966797, "learning_rate": 0.0007665324007202235, "loss": 2.4067, "step": 745 }, { "epoch": 0.3420449335167354, "grad_norm": 0.4816958010196686, "learning_rate": 0.0007659037331173498, "loss": 2.4674, "step": 746 }, { "epoch": 0.3425034387895461, "grad_norm": 0.47204354405403137, "learning_rate": 0.0007652744788316752, "loss": 2.399, "step": 747 }, { "epoch": 0.3429619440623567, "grad_norm": 0.4828701615333557, "learning_rate": 0.0007646446392515692, "loss": 2.4663, "step": 748 }, { "epoch": 0.34342044933516735, "grad_norm": 0.5179896354675293, "learning_rate": 0.000764014215766693, "loss": 2.5088, "step": 749 }, { "epoch": 0.343878954607978, "grad_norm": 0.5135321617126465, "learning_rate": 0.0007633832097679958, "loss": 2.4802, "step": 750 }, { "epoch": 0.3443374598807886, "grad_norm": 0.5362501740455627, "learning_rate": 0.0007627516226477122, "loss": 2.4735, "step": 751 }, { "epoch": 0.3447959651535993, "grad_norm": 0.5061324238777161, "learning_rate": 0.0007621194557993589, "loss": 2.581, "step": 752 }, { "epoch": 0.3452544704264099, "grad_norm": 0.4922947287559509, "learning_rate": 0.0007614867106177319, "loss": 2.4955, "step": 753 }, { "epoch": 0.34571297569922055, "grad_norm": 0.49132925271987915, "learning_rate": 0.0007608533884989029, "loss": 2.4957, "step": 754 }, { "epoch": 0.34617148097203115, "grad_norm": 0.5040573477745056, "learning_rate": 0.0007602194908402166, "loss": 2.4473, "step": 755 }, { "epoch": 0.3466299862448418, "grad_norm": 0.46081608533859253, "learning_rate": 0.0007595850190402877, "loss": 2.5242, "step": 756 }, { "epoch": 0.3470884915176525, "grad_norm": 0.495818555355072, "learning_rate": 0.0007589499744989976, "loss": 2.625, "step": 757 }, { "epoch": 0.3475469967904631, "grad_norm": 0.48972395062446594, "learning_rate": 0.0007583143586174916, "loss": 2.4808, "step": 758 }, { "epoch": 0.34800550206327374, "grad_norm": 0.4666306674480438, "learning_rate": 0.000757678172798175, "loss": 2.4996, "step": 759 }, { "epoch": 0.34846400733608435, "grad_norm": 0.4716382324695587, "learning_rate": 0.0007570414184447112, "loss": 2.4877, "step": 760 }, { "epoch": 0.348922512608895, "grad_norm": 0.493770956993103, "learning_rate": 0.0007564040969620179, "loss": 2.5016, "step": 761 }, { "epoch": 0.3493810178817056, "grad_norm": 0.4789999723434448, "learning_rate": 0.0007557662097562636, "loss": 2.3408, "step": 762 }, { "epoch": 0.3498395231545163, "grad_norm": 0.5204689502716064, "learning_rate": 0.0007551277582348658, "loss": 2.5568, "step": 763 }, { "epoch": 0.35029802842732694, "grad_norm": 0.47120845317840576, "learning_rate": 0.0007544887438064862, "loss": 2.4185, "step": 764 }, { "epoch": 0.35075653370013754, "grad_norm": 0.4945816099643707, "learning_rate": 0.0007538491678810294, "loss": 2.3618, "step": 765 }, { "epoch": 0.3512150389729482, "grad_norm": 0.46599090099334717, "learning_rate": 0.0007532090318696381, "loss": 2.5101, "step": 766 }, { "epoch": 0.3516735442457588, "grad_norm": 0.46806642413139343, "learning_rate": 0.0007525683371846913, "loss": 2.504, "step": 767 }, { "epoch": 0.35213204951856947, "grad_norm": 0.456985741853714, "learning_rate": 0.0007519270852398001, "loss": 2.5044, "step": 768 }, { "epoch": 0.35259055479138013, "grad_norm": 0.45167818665504456, "learning_rate": 0.000751285277449806, "loss": 2.4831, "step": 769 }, { "epoch": 0.35304906006419073, "grad_norm": 0.4702685475349426, "learning_rate": 0.0007506429152307756, "loss": 2.4824, "step": 770 }, { "epoch": 0.3535075653370014, "grad_norm": 0.4711180329322815, "learning_rate": 0.00075, "loss": 2.4052, "step": 771 }, { "epoch": 0.353966070609812, "grad_norm": 0.4505963921546936, "learning_rate": 0.00074935653317599, "loss": 2.4639, "step": 772 }, { "epoch": 0.35442457588262266, "grad_norm": 0.49853184819221497, "learning_rate": 0.000748712516178473, "loss": 2.4825, "step": 773 }, { "epoch": 0.35488308115543327, "grad_norm": 0.49328115582466125, "learning_rate": 0.0007480679504283911, "loss": 2.4983, "step": 774 }, { "epoch": 0.3553415864282439, "grad_norm": 0.4945114254951477, "learning_rate": 0.0007474228373478964, "loss": 2.4863, "step": 775 }, { "epoch": 0.3558000917010546, "grad_norm": 0.47543656826019287, "learning_rate": 0.0007467771783603492, "loss": 2.4044, "step": 776 }, { "epoch": 0.3562585969738652, "grad_norm": 0.44838574528694153, "learning_rate": 0.0007461309748903138, "loss": 2.5309, "step": 777 }, { "epoch": 0.35671710224667585, "grad_norm": 0.47382208704948425, "learning_rate": 0.0007454842283635562, "loss": 2.4845, "step": 778 }, { "epoch": 0.35717560751948646, "grad_norm": 0.49823275208473206, "learning_rate": 0.0007448369402070404, "loss": 2.3463, "step": 779 }, { "epoch": 0.3576341127922971, "grad_norm": 0.5136720538139343, "learning_rate": 0.0007441891118489254, "loss": 2.538, "step": 780 }, { "epoch": 0.3580926180651077, "grad_norm": 0.4927549958229065, "learning_rate": 0.0007435407447185622, "loss": 2.4879, "step": 781 }, { "epoch": 0.3585511233379184, "grad_norm": 0.49229538440704346, "learning_rate": 0.0007428918402464908, "loss": 2.4666, "step": 782 }, { "epoch": 0.35900962861072905, "grad_norm": 0.49309495091438293, "learning_rate": 0.0007422423998644359, "loss": 2.3957, "step": 783 }, { "epoch": 0.35946813388353965, "grad_norm": 0.4603478014469147, "learning_rate": 0.0007415924250053055, "loss": 2.4968, "step": 784 }, { "epoch": 0.3599266391563503, "grad_norm": 0.5007213354110718, "learning_rate": 0.0007409419171031865, "loss": 2.4727, "step": 785 }, { "epoch": 0.3603851444291609, "grad_norm": 0.47564634680747986, "learning_rate": 0.0007402908775933419, "loss": 2.409, "step": 786 }, { "epoch": 0.3608436497019716, "grad_norm": 0.5003230571746826, "learning_rate": 0.0007396393079122077, "loss": 2.3448, "step": 787 }, { "epoch": 0.3613021549747822, "grad_norm": 0.46794670820236206, "learning_rate": 0.0007389872094973896, "loss": 2.4367, "step": 788 }, { "epoch": 0.36176066024759285, "grad_norm": 0.4499052166938782, "learning_rate": 0.00073833458378766, "loss": 2.4575, "step": 789 }, { "epoch": 0.3622191655204035, "grad_norm": 0.4952113628387451, "learning_rate": 0.0007376814322229544, "loss": 2.5595, "step": 790 }, { "epoch": 0.3626776707932141, "grad_norm": 0.5037971138954163, "learning_rate": 0.0007370277562443688, "loss": 2.4422, "step": 791 }, { "epoch": 0.3631361760660248, "grad_norm": 0.4730939269065857, "learning_rate": 0.0007363735572941564, "loss": 2.5036, "step": 792 }, { "epoch": 0.3635946813388354, "grad_norm": 0.5195810198783875, "learning_rate": 0.0007357188368157236, "loss": 2.5925, "step": 793 }, { "epoch": 0.36405318661164604, "grad_norm": 0.4948427975177765, "learning_rate": 0.0007350635962536284, "loss": 2.4551, "step": 794 }, { "epoch": 0.36451169188445665, "grad_norm": 0.5114973783493042, "learning_rate": 0.0007344078370535756, "loss": 2.3749, "step": 795 }, { "epoch": 0.3649701971572673, "grad_norm": 0.46566635370254517, "learning_rate": 0.0007337515606624148, "loss": 2.5044, "step": 796 }, { "epoch": 0.36542870243007797, "grad_norm": 0.4537060260772705, "learning_rate": 0.0007330947685281362, "loss": 2.3792, "step": 797 }, { "epoch": 0.3658872077028886, "grad_norm": 0.4674481451511383, "learning_rate": 0.0007324374620998682, "loss": 2.4823, "step": 798 }, { "epoch": 0.36634571297569923, "grad_norm": 0.5132829546928406, "learning_rate": 0.000731779642827874, "loss": 2.4064, "step": 799 }, { "epoch": 0.36680421824850984, "grad_norm": 0.5076897740364075, "learning_rate": 0.0007311213121635483, "loss": 2.4335, "step": 800 }, { "epoch": 0.3672627235213205, "grad_norm": 0.49934977293014526, "learning_rate": 0.0007304624715594139, "loss": 2.4472, "step": 801 }, { "epoch": 0.3677212287941311, "grad_norm": 0.46451807022094727, "learning_rate": 0.0007298031224691193, "loss": 2.4729, "step": 802 }, { "epoch": 0.36817973406694177, "grad_norm": 0.4751005172729492, "learning_rate": 0.0007291432663474339, "loss": 2.4174, "step": 803 }, { "epoch": 0.3686382393397524, "grad_norm": 0.4676305949687958, "learning_rate": 0.0007284829046502467, "loss": 2.475, "step": 804 }, { "epoch": 0.36909674461256303, "grad_norm": 0.49120116233825684, "learning_rate": 0.0007278220388345619, "loss": 2.4476, "step": 805 }, { "epoch": 0.3695552498853737, "grad_norm": 0.4533020257949829, "learning_rate": 0.0007271606703584958, "loss": 2.433, "step": 806 }, { "epoch": 0.3700137551581843, "grad_norm": 0.4544160068035126, "learning_rate": 0.000726498800681274, "loss": 2.4188, "step": 807 }, { "epoch": 0.37047226043099496, "grad_norm": 0.5088750123977661, "learning_rate": 0.0007258364312632279, "loss": 2.4699, "step": 808 }, { "epoch": 0.3709307657038056, "grad_norm": 0.47790804505348206, "learning_rate": 0.0007251735635657915, "loss": 2.4486, "step": 809 }, { "epoch": 0.3713892709766162, "grad_norm": 0.4787186086177826, "learning_rate": 0.000724510199051498, "loss": 2.458, "step": 810 }, { "epoch": 0.3718477762494269, "grad_norm": 0.5019228458404541, "learning_rate": 0.0007238463391839769, "loss": 2.4447, "step": 811 }, { "epoch": 0.3723062815222375, "grad_norm": 0.473299115896225, "learning_rate": 0.0007231819854279508, "loss": 2.5319, "step": 812 }, { "epoch": 0.37276478679504815, "grad_norm": 0.47398027777671814, "learning_rate": 0.0007225171392492316, "loss": 2.3914, "step": 813 }, { "epoch": 0.37322329206785876, "grad_norm": 0.5190762877464294, "learning_rate": 0.0007218518021147182, "loss": 2.5304, "step": 814 }, { "epoch": 0.3736817973406694, "grad_norm": 0.4927901029586792, "learning_rate": 0.0007211859754923923, "loss": 2.4589, "step": 815 }, { "epoch": 0.3741403026134801, "grad_norm": 0.48509424924850464, "learning_rate": 0.0007205196608513158, "loss": 2.4001, "step": 816 }, { "epoch": 0.3745988078862907, "grad_norm": 0.4533082842826843, "learning_rate": 0.0007198528596616272, "loss": 2.5176, "step": 817 }, { "epoch": 0.37505731315910135, "grad_norm": 0.48269736766815186, "learning_rate": 0.0007191855733945387, "loss": 2.3752, "step": 818 }, { "epoch": 0.37551581843191195, "grad_norm": 0.46667489409446716, "learning_rate": 0.0007185178035223327, "loss": 2.4096, "step": 819 }, { "epoch": 0.3759743237047226, "grad_norm": 0.4753306210041046, "learning_rate": 0.0007178495515183583, "loss": 2.4633, "step": 820 }, { "epoch": 0.3764328289775332, "grad_norm": 0.49039340019226074, "learning_rate": 0.000717180818857029, "loss": 2.4581, "step": 821 }, { "epoch": 0.3768913342503439, "grad_norm": 0.4823123812675476, "learning_rate": 0.0007165116070138182, "loss": 2.3596, "step": 822 }, { "epoch": 0.37734983952315454, "grad_norm": 0.4582519233226776, "learning_rate": 0.0007158419174652569, "loss": 2.3281, "step": 823 }, { "epoch": 0.37780834479596515, "grad_norm": 0.5021100044250488, "learning_rate": 0.00071517175168893, "loss": 2.4392, "step": 824 }, { "epoch": 0.3782668500687758, "grad_norm": 0.45531928539276123, "learning_rate": 0.0007145011111634732, "loss": 2.5255, "step": 825 }, { "epoch": 0.3787253553415864, "grad_norm": 0.4993340075016022, "learning_rate": 0.0007138299973685694, "loss": 2.5717, "step": 826 }, { "epoch": 0.3791838606143971, "grad_norm": 0.5079259276390076, "learning_rate": 0.0007131584117849459, "loss": 2.4618, "step": 827 }, { "epoch": 0.3796423658872077, "grad_norm": 0.5084542632102966, "learning_rate": 0.0007124863558943713, "loss": 2.4861, "step": 828 }, { "epoch": 0.38010087116001834, "grad_norm": 0.4595532715320587, "learning_rate": 0.0007118138311796514, "loss": 2.5331, "step": 829 }, { "epoch": 0.380559376432829, "grad_norm": 0.5431782007217407, "learning_rate": 0.0007111408391246262, "loss": 2.4112, "step": 830 }, { "epoch": 0.3810178817056396, "grad_norm": 0.47534796595573425, "learning_rate": 0.0007104673812141675, "loss": 2.539, "step": 831 }, { "epoch": 0.38147638697845027, "grad_norm": 0.5211132168769836, "learning_rate": 0.0007097934589341745, "loss": 2.5511, "step": 832 }, { "epoch": 0.38193489225126087, "grad_norm": 0.46920859813690186, "learning_rate": 0.0007091190737715711, "loss": 2.5069, "step": 833 }, { "epoch": 0.38239339752407153, "grad_norm": 0.45785412192344666, "learning_rate": 0.0007084442272143026, "loss": 2.4536, "step": 834 }, { "epoch": 0.38285190279688214, "grad_norm": 0.45247727632522583, "learning_rate": 0.000707768920751332, "loss": 2.5169, "step": 835 }, { "epoch": 0.3833104080696928, "grad_norm": 0.4721006453037262, "learning_rate": 0.0007070931558726373, "loss": 2.5134, "step": 836 }, { "epoch": 0.38376891334250346, "grad_norm": 0.44979557394981384, "learning_rate": 0.0007064169340692076, "loss": 2.4313, "step": 837 }, { "epoch": 0.38422741861531406, "grad_norm": 0.4435129761695862, "learning_rate": 0.0007057402568330407, "loss": 2.3829, "step": 838 }, { "epoch": 0.3846859238881247, "grad_norm": 0.4309985637664795, "learning_rate": 0.0007050631256571389, "loss": 2.3851, "step": 839 }, { "epoch": 0.38514442916093533, "grad_norm": 0.46195662021636963, "learning_rate": 0.000704385542035506, "loss": 2.4414, "step": 840 }, { "epoch": 0.385602934433746, "grad_norm": 0.4740036427974701, "learning_rate": 0.000703707507463144, "loss": 2.3973, "step": 841 }, { "epoch": 0.38606143970655665, "grad_norm": 0.47478780150413513, "learning_rate": 0.0007030290234360505, "loss": 2.5736, "step": 842 }, { "epoch": 0.38651994497936726, "grad_norm": 0.4838266372680664, "learning_rate": 0.0007023500914512139, "loss": 2.4748, "step": 843 }, { "epoch": 0.3869784502521779, "grad_norm": 0.45548906922340393, "learning_rate": 0.0007016707130066116, "loss": 2.407, "step": 844 }, { "epoch": 0.3874369555249885, "grad_norm": 0.4546142816543579, "learning_rate": 0.0007009908896012055, "loss": 2.4641, "step": 845 }, { "epoch": 0.3878954607977992, "grad_norm": 0.45099204778671265, "learning_rate": 0.0007003106227349399, "loss": 2.5704, "step": 846 }, { "epoch": 0.3883539660706098, "grad_norm": 0.4800565540790558, "learning_rate": 0.000699629913908737, "loss": 2.47, "step": 847 }, { "epoch": 0.38881247134342045, "grad_norm": 0.486941933631897, "learning_rate": 0.0006989487646244943, "loss": 2.49, "step": 848 }, { "epoch": 0.3892709766162311, "grad_norm": 0.49665388464927673, "learning_rate": 0.0006982671763850814, "loss": 2.4348, "step": 849 }, { "epoch": 0.3897294818890417, "grad_norm": 0.4610441327095032, "learning_rate": 0.0006975851506943359, "loss": 2.3933, "step": 850 }, { "epoch": 0.3901879871618524, "grad_norm": 0.5300979018211365, "learning_rate": 0.0006969026890570611, "loss": 2.5115, "step": 851 }, { "epoch": 0.390646492434663, "grad_norm": 0.48305070400238037, "learning_rate": 0.0006962197929790216, "loss": 2.3862, "step": 852 }, { "epoch": 0.39110499770747365, "grad_norm": 0.457640677690506, "learning_rate": 0.0006955364639669409, "loss": 2.555, "step": 853 }, { "epoch": 0.39156350298028425, "grad_norm": 0.4472521245479584, "learning_rate": 0.0006948527035284978, "loss": 2.4858, "step": 854 }, { "epoch": 0.3920220082530949, "grad_norm": 0.45382168889045715, "learning_rate": 0.0006941685131723225, "loss": 2.5047, "step": 855 }, { "epoch": 0.39248051352590557, "grad_norm": 0.47845160961151123, "learning_rate": 0.0006934838944079943, "loss": 2.4995, "step": 856 }, { "epoch": 0.3929390187987162, "grad_norm": 0.46178799867630005, "learning_rate": 0.0006927988487460378, "loss": 2.4398, "step": 857 }, { "epoch": 0.39339752407152684, "grad_norm": 0.45385468006134033, "learning_rate": 0.0006921133776979186, "loss": 2.4679, "step": 858 }, { "epoch": 0.39385602934433744, "grad_norm": 0.4846494197845459, "learning_rate": 0.0006914274827760418, "loss": 2.3779, "step": 859 }, { "epoch": 0.3943145346171481, "grad_norm": 0.48416373133659363, "learning_rate": 0.0006907411654937475, "loss": 2.5133, "step": 860 }, { "epoch": 0.3947730398899587, "grad_norm": 0.4826957881450653, "learning_rate": 0.0006900544273653075, "loss": 2.5815, "step": 861 }, { "epoch": 0.39523154516276937, "grad_norm": 0.4939316213130951, "learning_rate": 0.000689367269905922, "loss": 2.4994, "step": 862 }, { "epoch": 0.39569005043558003, "grad_norm": 0.458729088306427, "learning_rate": 0.0006886796946317168, "loss": 2.4185, "step": 863 }, { "epoch": 0.39614855570839064, "grad_norm": 0.45166710019111633, "learning_rate": 0.0006879917030597397, "loss": 2.4394, "step": 864 }, { "epoch": 0.3966070609812013, "grad_norm": 0.4646100699901581, "learning_rate": 0.0006873032967079561, "loss": 2.389, "step": 865 }, { "epoch": 0.3970655662540119, "grad_norm": 0.47953346371650696, "learning_rate": 0.0006866144770952474, "loss": 2.4571, "step": 866 }, { "epoch": 0.39752407152682256, "grad_norm": 0.48801127076148987, "learning_rate": 0.0006859252457414067, "loss": 2.5315, "step": 867 }, { "epoch": 0.39798257679963317, "grad_norm": 0.46776098012924194, "learning_rate": 0.0006852356041671351, "loss": 2.3368, "step": 868 }, { "epoch": 0.39844108207244383, "grad_norm": 0.484793484210968, "learning_rate": 0.0006845455538940394, "loss": 2.4994, "step": 869 }, { "epoch": 0.3988995873452545, "grad_norm": 0.46986517310142517, "learning_rate": 0.0006838550964446276, "loss": 2.4132, "step": 870 }, { "epoch": 0.3993580926180651, "grad_norm": 0.4764801561832428, "learning_rate": 0.0006831642333423067, "loss": 2.5467, "step": 871 }, { "epoch": 0.39981659789087576, "grad_norm": 0.553652286529541, "learning_rate": 0.000682472966111378, "loss": 2.4435, "step": 872 }, { "epoch": 0.40027510316368636, "grad_norm": 0.47316744923591614, "learning_rate": 0.0006817812962770348, "loss": 2.4848, "step": 873 }, { "epoch": 0.400733608436497, "grad_norm": 0.48192256689071655, "learning_rate": 0.0006810892253653589, "loss": 2.4118, "step": 874 }, { "epoch": 0.40119211370930763, "grad_norm": 0.4988864064216614, "learning_rate": 0.0006803967549033167, "loss": 2.4768, "step": 875 }, { "epoch": 0.4016506189821183, "grad_norm": 0.48958611488342285, "learning_rate": 0.0006797038864187564, "loss": 2.5105, "step": 876 }, { "epoch": 0.40210912425492895, "grad_norm": 0.49673572182655334, "learning_rate": 0.0006790106214404043, "loss": 2.5441, "step": 877 }, { "epoch": 0.40256762952773956, "grad_norm": 0.4830143451690674, "learning_rate": 0.0006783169614978614, "loss": 2.471, "step": 878 }, { "epoch": 0.4030261348005502, "grad_norm": 0.47977522015571594, "learning_rate": 0.0006776229081216001, "loss": 2.4184, "step": 879 }, { "epoch": 0.4034846400733608, "grad_norm": 0.454280287027359, "learning_rate": 0.0006769284628429611, "loss": 2.4069, "step": 880 }, { "epoch": 0.4039431453461715, "grad_norm": 0.5290543437004089, "learning_rate": 0.0006762336271941498, "loss": 2.4698, "step": 881 }, { "epoch": 0.40440165061898214, "grad_norm": 0.488118439912796, "learning_rate": 0.0006755384027082326, "loss": 2.418, "step": 882 }, { "epoch": 0.40486015589179275, "grad_norm": 0.48343366384506226, "learning_rate": 0.0006748427909191342, "loss": 2.4084, "step": 883 }, { "epoch": 0.4053186611646034, "grad_norm": 0.5353105664253235, "learning_rate": 0.0006741467933616335, "loss": 2.4354, "step": 884 }, { "epoch": 0.405777166437414, "grad_norm": 0.47047802805900574, "learning_rate": 0.0006734504115713604, "loss": 2.4262, "step": 885 }, { "epoch": 0.4062356717102247, "grad_norm": 0.4728517234325409, "learning_rate": 0.0006727536470847932, "loss": 2.4333, "step": 886 }, { "epoch": 0.4066941769830353, "grad_norm": 0.48541244864463806, "learning_rate": 0.000672056501439254, "loss": 2.3859, "step": 887 }, { "epoch": 0.40715268225584594, "grad_norm": 0.4519893527030945, "learning_rate": 0.0006713589761729063, "loss": 2.4222, "step": 888 }, { "epoch": 0.4076111875286566, "grad_norm": 0.5226224660873413, "learning_rate": 0.0006706610728247508, "loss": 2.5111, "step": 889 }, { "epoch": 0.4080696928014672, "grad_norm": 0.49453553557395935, "learning_rate": 0.0006699627929346227, "loss": 2.5517, "step": 890 }, { "epoch": 0.40852819807427787, "grad_norm": 0.5406020879745483, "learning_rate": 0.0006692641380431879, "loss": 2.4568, "step": 891 }, { "epoch": 0.4089867033470885, "grad_norm": 0.47185781598091125, "learning_rate": 0.0006685651096919393, "loss": 2.4471, "step": 892 }, { "epoch": 0.40944520861989914, "grad_norm": 0.5062583088874817, "learning_rate": 0.0006678657094231944, "loss": 2.257, "step": 893 }, { "epoch": 0.40990371389270974, "grad_norm": 0.49273499846458435, "learning_rate": 0.0006671659387800909, "loss": 2.5593, "step": 894 }, { "epoch": 0.4103622191655204, "grad_norm": 0.490082323551178, "learning_rate": 0.000666465799306584, "loss": 2.4549, "step": 895 }, { "epoch": 0.41082072443833106, "grad_norm": 0.4863387644290924, "learning_rate": 0.0006657652925474423, "loss": 2.4125, "step": 896 }, { "epoch": 0.41127922971114167, "grad_norm": 0.5017125010490417, "learning_rate": 0.000665064420048245, "loss": 2.5133, "step": 897 }, { "epoch": 0.41173773498395233, "grad_norm": 0.4683160185813904, "learning_rate": 0.0006643631833553785, "loss": 2.3856, "step": 898 }, { "epoch": 0.41219624025676294, "grad_norm": 0.46449413895606995, "learning_rate": 0.000663661584016032, "loss": 2.4116, "step": 899 }, { "epoch": 0.4126547455295736, "grad_norm": 0.45589032769203186, "learning_rate": 0.0006629596235781957, "loss": 2.5296, "step": 900 }, { "epoch": 0.4131132508023842, "grad_norm": 0.4699907600879669, "learning_rate": 0.0006622573035906556, "loss": 2.4336, "step": 901 }, { "epoch": 0.41357175607519486, "grad_norm": 0.48073622584342957, "learning_rate": 0.0006615546256029921, "loss": 2.5015, "step": 902 }, { "epoch": 0.4140302613480055, "grad_norm": 0.4875558912754059, "learning_rate": 0.0006608515911655743, "loss": 2.6259, "step": 903 }, { "epoch": 0.41448876662081613, "grad_norm": 0.47009405493736267, "learning_rate": 0.0006601482018295591, "loss": 2.5145, "step": 904 }, { "epoch": 0.4149472718936268, "grad_norm": 0.45319053530693054, "learning_rate": 0.0006594444591468851, "loss": 2.4829, "step": 905 }, { "epoch": 0.4154057771664374, "grad_norm": 0.462557315826416, "learning_rate": 0.0006587403646702713, "loss": 2.4812, "step": 906 }, { "epoch": 0.41586428243924806, "grad_norm": 0.4525872468948364, "learning_rate": 0.0006580359199532126, "loss": 2.4031, "step": 907 }, { "epoch": 0.41632278771205866, "grad_norm": 0.4792379140853882, "learning_rate": 0.000657331126549977, "loss": 2.4192, "step": 908 }, { "epoch": 0.4167812929848693, "grad_norm": 0.4633989930152893, "learning_rate": 0.0006566259860156014, "loss": 2.4234, "step": 909 }, { "epoch": 0.41723979825768, "grad_norm": 0.5089423060417175, "learning_rate": 0.0006559204999058888, "loss": 2.5071, "step": 910 }, { "epoch": 0.4176983035304906, "grad_norm": 0.5051239132881165, "learning_rate": 0.0006552146697774049, "loss": 2.4914, "step": 911 }, { "epoch": 0.41815680880330125, "grad_norm": 0.4888673722743988, "learning_rate": 0.0006545084971874737, "loss": 2.4738, "step": 912 }, { "epoch": 0.41861531407611186, "grad_norm": 0.4959743916988373, "learning_rate": 0.0006538019836941758, "loss": 2.4748, "step": 913 }, { "epoch": 0.4190738193489225, "grad_norm": 0.47992533445358276, "learning_rate": 0.0006530951308563431, "loss": 2.5229, "step": 914 }, { "epoch": 0.4195323246217332, "grad_norm": 0.44094118475914, "learning_rate": 0.0006523879402335567, "loss": 2.4706, "step": 915 }, { "epoch": 0.4199908298945438, "grad_norm": 0.46255770325660706, "learning_rate": 0.0006516804133861429, "loss": 2.4637, "step": 916 }, { "epoch": 0.42044933516735444, "grad_norm": 0.4518481492996216, "learning_rate": 0.0006509725518751698, "loss": 2.3934, "step": 917 }, { "epoch": 0.42090784044016505, "grad_norm": 0.5264530181884766, "learning_rate": 0.0006502643572624438, "loss": 2.462, "step": 918 }, { "epoch": 0.4213663457129757, "grad_norm": 0.465501070022583, "learning_rate": 0.0006495558311105064, "loss": 2.3683, "step": 919 }, { "epoch": 0.4218248509857863, "grad_norm": 0.45718127489089966, "learning_rate": 0.0006488469749826305, "loss": 2.452, "step": 920 }, { "epoch": 0.422283356258597, "grad_norm": 0.47971111536026, "learning_rate": 0.000648137790442817, "loss": 2.4773, "step": 921 }, { "epoch": 0.42274186153140764, "grad_norm": 0.4839756488800049, "learning_rate": 0.0006474282790557916, "loss": 2.3887, "step": 922 }, { "epoch": 0.42320036680421824, "grad_norm": 0.4580536484718323, "learning_rate": 0.000646718442387001, "loss": 2.426, "step": 923 }, { "epoch": 0.4236588720770289, "grad_norm": 0.4854689836502075, "learning_rate": 0.0006460082820026094, "loss": 2.4082, "step": 924 }, { "epoch": 0.4241173773498395, "grad_norm": 0.48722681403160095, "learning_rate": 0.0006452977994694959, "loss": 2.3589, "step": 925 }, { "epoch": 0.42457588262265017, "grad_norm": 0.4968544840812683, "learning_rate": 0.0006445869963552496, "loss": 2.3903, "step": 926 }, { "epoch": 0.4250343878954608, "grad_norm": 0.4614403247833252, "learning_rate": 0.0006438758742281672, "loss": 2.4531, "step": 927 }, { "epoch": 0.42549289316827144, "grad_norm": 0.4920521378517151, "learning_rate": 0.0006431644346572495, "loss": 2.4392, "step": 928 }, { "epoch": 0.4259513984410821, "grad_norm": 0.5021727085113525, "learning_rate": 0.0006424526792121974, "loss": 2.3472, "step": 929 }, { "epoch": 0.4264099037138927, "grad_norm": 0.4522426128387451, "learning_rate": 0.0006417406094634089, "loss": 2.5383, "step": 930 }, { "epoch": 0.42686840898670336, "grad_norm": 0.4554641544818878, "learning_rate": 0.0006410282269819756, "loss": 2.4153, "step": 931 }, { "epoch": 0.42732691425951397, "grad_norm": 0.5178560614585876, "learning_rate": 0.0006403155333396787, "loss": 2.4917, "step": 932 }, { "epoch": 0.42778541953232463, "grad_norm": 0.4553394913673401, "learning_rate": 0.0006396025301089863, "loss": 2.4867, "step": 933 }, { "epoch": 0.42824392480513523, "grad_norm": 0.4692284166812897, "learning_rate": 0.0006388892188630493, "loss": 2.392, "step": 934 }, { "epoch": 0.4287024300779459, "grad_norm": 0.45819544792175293, "learning_rate": 0.0006381756011756982, "loss": 2.4416, "step": 935 }, { "epoch": 0.42916093535075656, "grad_norm": 0.4594976007938385, "learning_rate": 0.0006374616786214403, "loss": 2.5459, "step": 936 }, { "epoch": 0.42961944062356716, "grad_norm": 0.44818904995918274, "learning_rate": 0.0006367474527754544, "loss": 2.4687, "step": 937 }, { "epoch": 0.4300779458963778, "grad_norm": 0.4488353729248047, "learning_rate": 0.0006360329252135894, "loss": 2.4685, "step": 938 }, { "epoch": 0.4305364511691884, "grad_norm": 0.4578348398208618, "learning_rate": 0.0006353180975123595, "loss": 2.4983, "step": 939 }, { "epoch": 0.4309949564419991, "grad_norm": 0.4648081660270691, "learning_rate": 0.0006346029712489413, "loss": 2.4364, "step": 940 }, { "epoch": 0.4314534617148097, "grad_norm": 0.46719297766685486, "learning_rate": 0.0006338875480011698, "loss": 2.5372, "step": 941 }, { "epoch": 0.43191196698762035, "grad_norm": 0.45542722940444946, "learning_rate": 0.0006331718293475357, "loss": 2.3282, "step": 942 }, { "epoch": 0.432370472260431, "grad_norm": 0.45853570103645325, "learning_rate": 0.0006324558168671811, "loss": 2.4385, "step": 943 }, { "epoch": 0.4328289775332416, "grad_norm": 0.46841874718666077, "learning_rate": 0.0006317395121398968, "loss": 2.5204, "step": 944 }, { "epoch": 0.4332874828060523, "grad_norm": 0.46844637393951416, "learning_rate": 0.0006310229167461179, "loss": 2.4275, "step": 945 }, { "epoch": 0.4337459880788629, "grad_norm": 0.5104042887687683, "learning_rate": 0.0006303060322669214, "loss": 2.3023, "step": 946 }, { "epoch": 0.43420449335167355, "grad_norm": 0.5155215263366699, "learning_rate": 0.0006295888602840214, "loss": 2.6016, "step": 947 }, { "epoch": 0.4346629986244842, "grad_norm": 0.48236551880836487, "learning_rate": 0.0006288714023797671, "loss": 2.4455, "step": 948 }, { "epoch": 0.4351215038972948, "grad_norm": 0.5006770491600037, "learning_rate": 0.000628153660137138, "loss": 2.5658, "step": 949 }, { "epoch": 0.4355800091701055, "grad_norm": 0.4774892330169678, "learning_rate": 0.0006274356351397413, "loss": 2.454, "step": 950 }, { "epoch": 0.4360385144429161, "grad_norm": 0.4677017033100128, "learning_rate": 0.0006267173289718079, "loss": 2.3864, "step": 951 }, { "epoch": 0.43649701971572674, "grad_norm": 0.4866916835308075, "learning_rate": 0.000625998743218189, "loss": 2.5003, "step": 952 }, { "epoch": 0.43695552498853735, "grad_norm": 0.4758725166320801, "learning_rate": 0.000625279879464353, "loss": 2.4211, "step": 953 }, { "epoch": 0.437414030261348, "grad_norm": 0.4572947919368744, "learning_rate": 0.000624560739296381, "loss": 2.4555, "step": 954 }, { "epoch": 0.43787253553415867, "grad_norm": 0.48687025904655457, "learning_rate": 0.0006238413243009648, "loss": 2.3416, "step": 955 }, { "epoch": 0.4383310408069693, "grad_norm": 0.47105884552001953, "learning_rate": 0.000623121636065402, "loss": 2.4547, "step": 956 }, { "epoch": 0.43878954607977994, "grad_norm": 0.4704273045063019, "learning_rate": 0.0006224016761775933, "loss": 2.4132, "step": 957 }, { "epoch": 0.43924805135259054, "grad_norm": 0.5046764016151428, "learning_rate": 0.0006216814462260386, "loss": 2.528, "step": 958 }, { "epoch": 0.4397065566254012, "grad_norm": 0.4965061545372009, "learning_rate": 0.0006209609477998338, "loss": 2.4553, "step": 959 }, { "epoch": 0.4401650618982118, "grad_norm": 0.48382359743118286, "learning_rate": 0.0006202401824886674, "loss": 2.4613, "step": 960 }, { "epoch": 0.44062356717102247, "grad_norm": 0.5026668310165405, "learning_rate": 0.0006195191518828162, "loss": 2.4834, "step": 961 }, { "epoch": 0.44108207244383313, "grad_norm": 0.5016959309577942, "learning_rate": 0.0006187978575731427, "loss": 2.5209, "step": 962 }, { "epoch": 0.44154057771664373, "grad_norm": 0.4857417941093445, "learning_rate": 0.0006180763011510911, "loss": 2.4773, "step": 963 }, { "epoch": 0.4419990829894544, "grad_norm": 0.4661930501461029, "learning_rate": 0.000617354484208684, "loss": 2.3761, "step": 964 }, { "epoch": 0.442457588262265, "grad_norm": 0.48591357469558716, "learning_rate": 0.0006166324083385189, "loss": 2.4507, "step": 965 }, { "epoch": 0.44291609353507566, "grad_norm": 0.4785884618759155, "learning_rate": 0.0006159100751337642, "loss": 2.4586, "step": 966 }, { "epoch": 0.44337459880788627, "grad_norm": 0.457625150680542, "learning_rate": 0.0006151874861881565, "loss": 2.4977, "step": 967 }, { "epoch": 0.4438331040806969, "grad_norm": 0.44921568036079407, "learning_rate": 0.0006144646430959964, "loss": 2.4276, "step": 968 }, { "epoch": 0.4442916093535076, "grad_norm": 0.4560396373271942, "learning_rate": 0.0006137415474521454, "loss": 2.4125, "step": 969 }, { "epoch": 0.4447501146263182, "grad_norm": 0.4628419578075409, "learning_rate": 0.0006130182008520222, "loss": 2.4414, "step": 970 }, { "epoch": 0.44520861989912885, "grad_norm": 0.4394832253456116, "learning_rate": 0.000612294604891599, "loss": 2.431, "step": 971 }, { "epoch": 0.44566712517193946, "grad_norm": 0.47777193784713745, "learning_rate": 0.0006115707611673986, "loss": 2.4192, "step": 972 }, { "epoch": 0.4461256304447501, "grad_norm": 0.4992137551307678, "learning_rate": 0.0006108466712764902, "loss": 2.5682, "step": 973 }, { "epoch": 0.4465841357175607, "grad_norm": 0.4614331126213074, "learning_rate": 0.0006101223368164858, "loss": 2.493, "step": 974 }, { "epoch": 0.4470426409903714, "grad_norm": 0.48823127150535583, "learning_rate": 0.0006093977593855375, "loss": 2.551, "step": 975 }, { "epoch": 0.44750114626318205, "grad_norm": 0.4704797565937042, "learning_rate": 0.0006086729405823335, "loss": 2.4198, "step": 976 }, { "epoch": 0.44795965153599265, "grad_norm": 0.4921104311943054, "learning_rate": 0.0006079478820060943, "loss": 2.4577, "step": 977 }, { "epoch": 0.4484181568088033, "grad_norm": 0.43696120381355286, "learning_rate": 0.0006072225852565695, "loss": 2.4418, "step": 978 }, { "epoch": 0.4488766620816139, "grad_norm": 0.4465458393096924, "learning_rate": 0.0006064970519340341, "loss": 2.4247, "step": 979 }, { "epoch": 0.4493351673544246, "grad_norm": 0.4515821039676666, "learning_rate": 0.0006057712836392856, "loss": 2.4818, "step": 980 }, { "epoch": 0.4497936726272352, "grad_norm": 0.4465532600879669, "learning_rate": 0.0006050452819736389, "loss": 2.3885, "step": 981 }, { "epoch": 0.45025217790004585, "grad_norm": 0.47944334149360657, "learning_rate": 0.000604319048538925, "loss": 2.4207, "step": 982 }, { "epoch": 0.4507106831728565, "grad_norm": 0.4609314501285553, "learning_rate": 0.0006035925849374855, "loss": 2.4157, "step": 983 }, { "epoch": 0.4511691884456671, "grad_norm": 0.4673696458339691, "learning_rate": 0.0006028658927721697, "loss": 2.4653, "step": 984 }, { "epoch": 0.4516276937184778, "grad_norm": 0.43880653381347656, "learning_rate": 0.0006021389736463321, "loss": 2.4166, "step": 985 }, { "epoch": 0.4520861989912884, "grad_norm": 0.4618498682975769, "learning_rate": 0.0006014118291638271, "loss": 2.3959, "step": 986 }, { "epoch": 0.45254470426409904, "grad_norm": 0.46034321188926697, "learning_rate": 0.0006006844609290065, "loss": 2.4727, "step": 987 }, { "epoch": 0.4530032095369097, "grad_norm": 0.4597344398498535, "learning_rate": 0.0005999568705467161, "loss": 2.461, "step": 988 }, { "epoch": 0.4534617148097203, "grad_norm": 0.4357251822948456, "learning_rate": 0.0005992290596222915, "loss": 2.3844, "step": 989 }, { "epoch": 0.45392022008253097, "grad_norm": 0.46407681703567505, "learning_rate": 0.0005985010297615551, "loss": 2.3583, "step": 990 }, { "epoch": 0.4543787253553416, "grad_norm": 0.514202356338501, "learning_rate": 0.0005977727825708123, "loss": 2.5153, "step": 991 }, { "epoch": 0.45483723062815223, "grad_norm": 0.45113158226013184, "learning_rate": 0.0005970443196568478, "loss": 2.4833, "step": 992 }, { "epoch": 0.45529573590096284, "grad_norm": 0.5029706954956055, "learning_rate": 0.0005963156426269227, "loss": 2.5131, "step": 993 }, { "epoch": 0.4557542411737735, "grad_norm": 0.481144517660141, "learning_rate": 0.0005955867530887702, "loss": 2.3842, "step": 994 }, { "epoch": 0.45621274644658416, "grad_norm": 0.47905784845352173, "learning_rate": 0.0005948576526505923, "loss": 2.4382, "step": 995 }, { "epoch": 0.45667125171939477, "grad_norm": 0.4525410830974579, "learning_rate": 0.0005941283429210568, "loss": 2.3317, "step": 996 }, { "epoch": 0.4571297569922054, "grad_norm": 0.49851951003074646, "learning_rate": 0.0005933988255092926, "loss": 2.4802, "step": 997 }, { "epoch": 0.45758826226501603, "grad_norm": 0.4817736744880676, "learning_rate": 0.0005926691020248874, "loss": 2.443, "step": 998 }, { "epoch": 0.4580467675378267, "grad_norm": 0.46093061566352844, "learning_rate": 0.0005919391740778833, "loss": 2.382, "step": 999 }, { "epoch": 0.4585052728106373, "grad_norm": 0.4627176821231842, "learning_rate": 0.0005912090432787736, "loss": 2.4612, "step": 1000 }, { "epoch": 0.45896377808344796, "grad_norm": 0.4464676082134247, "learning_rate": 0.000590478711238499, "loss": 2.4594, "step": 1001 }, { "epoch": 0.4594222833562586, "grad_norm": 0.4653976857662201, "learning_rate": 0.0005897481795684446, "loss": 2.4113, "step": 1002 }, { "epoch": 0.4598807886290692, "grad_norm": 0.4493559002876282, "learning_rate": 0.0005890174498804355, "loss": 2.4361, "step": 1003 }, { "epoch": 0.4603392939018799, "grad_norm": 0.4464796483516693, "learning_rate": 0.0005882865237867339, "loss": 2.3945, "step": 1004 }, { "epoch": 0.4607977991746905, "grad_norm": 0.4635982811450958, "learning_rate": 0.0005875554029000353, "loss": 2.4734, "step": 1005 }, { "epoch": 0.46125630444750115, "grad_norm": 0.42866286635398865, "learning_rate": 0.0005868240888334653, "loss": 2.4003, "step": 1006 }, { "epoch": 0.46171480972031176, "grad_norm": 0.4710012972354889, "learning_rate": 0.0005860925832005753, "loss": 2.4992, "step": 1007 }, { "epoch": 0.4621733149931224, "grad_norm": 0.458265095949173, "learning_rate": 0.0005853608876153395, "loss": 2.4818, "step": 1008 }, { "epoch": 0.4626318202659331, "grad_norm": 0.483789324760437, "learning_rate": 0.0005846290036921512, "loss": 2.3972, "step": 1009 }, { "epoch": 0.4630903255387437, "grad_norm": 0.4705009460449219, "learning_rate": 0.0005838969330458195, "loss": 2.4477, "step": 1010 }, { "epoch": 0.46354883081155435, "grad_norm": 0.49005094170570374, "learning_rate": 0.0005831646772915651, "loss": 2.4343, "step": 1011 }, { "epoch": 0.46400733608436495, "grad_norm": 0.46670791506767273, "learning_rate": 0.0005824322380450173, "loss": 2.4384, "step": 1012 }, { "epoch": 0.4644658413571756, "grad_norm": 0.48711952567100525, "learning_rate": 0.0005816996169222102, "loss": 2.4468, "step": 1013 }, { "epoch": 0.4649243466299862, "grad_norm": 0.461687833070755, "learning_rate": 0.0005809668155395793, "loss": 2.4781, "step": 1014 }, { "epoch": 0.4653828519027969, "grad_norm": 0.46107563376426697, "learning_rate": 0.0005802338355139578, "loss": 2.4307, "step": 1015 }, { "epoch": 0.46584135717560754, "grad_norm": 0.47539234161376953, "learning_rate": 0.0005795006784625728, "loss": 2.4838, "step": 1016 }, { "epoch": 0.46629986244841815, "grad_norm": 0.48568326234817505, "learning_rate": 0.0005787673460030423, "loss": 2.4843, "step": 1017 }, { "epoch": 0.4667583677212288, "grad_norm": 0.4536236524581909, "learning_rate": 0.000578033839753371, "loss": 2.5152, "step": 1018 }, { "epoch": 0.4672168729940394, "grad_norm": 0.4997485876083374, "learning_rate": 0.0005773001613319476, "loss": 2.3419, "step": 1019 }, { "epoch": 0.4676753782668501, "grad_norm": 0.4623103439807892, "learning_rate": 0.00057656631235754, "loss": 2.4912, "step": 1020 }, { "epoch": 0.46813388353966073, "grad_norm": 0.45869773626327515, "learning_rate": 0.0005758322944492929, "loss": 2.4813, "step": 1021 }, { "epoch": 0.46859238881247134, "grad_norm": 0.47725677490234375, "learning_rate": 0.0005750981092267237, "loss": 2.4381, "step": 1022 }, { "epoch": 0.469050894085282, "grad_norm": 0.4883020222187042, "learning_rate": 0.0005743637583097183, "loss": 2.3519, "step": 1023 }, { "epoch": 0.4695093993580926, "grad_norm": 0.48052722215652466, "learning_rate": 0.0005736292433185291, "loss": 2.403, "step": 1024 }, { "epoch": 0.46996790463090327, "grad_norm": 0.4742206931114197, "learning_rate": 0.0005728945658737699, "loss": 2.5376, "step": 1025 }, { "epoch": 0.47042640990371387, "grad_norm": 0.4480728209018707, "learning_rate": 0.0005721597275964133, "loss": 2.4595, "step": 1026 }, { "epoch": 0.47088491517652453, "grad_norm": 0.4462078809738159, "learning_rate": 0.0005714247301077865, "loss": 2.4514, "step": 1027 }, { "epoch": 0.4713434204493352, "grad_norm": 0.4276384711265564, "learning_rate": 0.0005706895750295682, "loss": 2.4875, "step": 1028 }, { "epoch": 0.4718019257221458, "grad_norm": 0.4370797574520111, "learning_rate": 0.0005699542639837844, "loss": 2.4078, "step": 1029 }, { "epoch": 0.47226043099495646, "grad_norm": 0.44996994733810425, "learning_rate": 0.0005692187985928055, "loss": 2.5216, "step": 1030 }, { "epoch": 0.47271893626776706, "grad_norm": 0.4885231852531433, "learning_rate": 0.0005684831804793427, "loss": 2.6886, "step": 1031 }, { "epoch": 0.4731774415405777, "grad_norm": 0.46301454305648804, "learning_rate": 0.0005677474112664438, "loss": 2.4105, "step": 1032 }, { "epoch": 0.47363594681338833, "grad_norm": 0.4885183870792389, "learning_rate": 0.0005670114925774899, "loss": 2.4913, "step": 1033 }, { "epoch": 0.474094452086199, "grad_norm": 0.46368205547332764, "learning_rate": 0.0005662754260361924, "loss": 2.4222, "step": 1034 }, { "epoch": 0.47455295735900965, "grad_norm": 0.47101131081581116, "learning_rate": 0.0005655392132665884, "loss": 2.4514, "step": 1035 }, { "epoch": 0.47501146263182026, "grad_norm": 0.5082679390907288, "learning_rate": 0.000564802855893038, "loss": 2.4437, "step": 1036 }, { "epoch": 0.4754699679046309, "grad_norm": 0.471524715423584, "learning_rate": 0.0005640663555402198, "loss": 2.5195, "step": 1037 }, { "epoch": 0.4759284731774415, "grad_norm": 0.4893777370452881, "learning_rate": 0.0005633297138331285, "loss": 2.5181, "step": 1038 }, { "epoch": 0.4763869784502522, "grad_norm": 0.4702717661857605, "learning_rate": 0.0005625929323970705, "loss": 2.6414, "step": 1039 }, { "epoch": 0.4768454837230628, "grad_norm": 0.4625989496707916, "learning_rate": 0.0005618560128576603, "loss": 2.469, "step": 1040 }, { "epoch": 0.47730398899587345, "grad_norm": 0.475857138633728, "learning_rate": 0.0005611189568408173, "loss": 2.482, "step": 1041 }, { "epoch": 0.4777624942686841, "grad_norm": 0.43992286920547485, "learning_rate": 0.0005603817659727619, "loss": 2.4178, "step": 1042 }, { "epoch": 0.4782209995414947, "grad_norm": 0.47170722484588623, "learning_rate": 0.0005596444418800121, "loss": 2.4069, "step": 1043 }, { "epoch": 0.4786795048143054, "grad_norm": 0.47044554352760315, "learning_rate": 0.0005589069861893798, "loss": 2.4163, "step": 1044 }, { "epoch": 0.479138010087116, "grad_norm": 0.45441001653671265, "learning_rate": 0.0005581694005279673, "loss": 2.4088, "step": 1045 }, { "epoch": 0.47959651535992665, "grad_norm": 0.4475735127925873, "learning_rate": 0.0005574316865231637, "loss": 2.3783, "step": 1046 }, { "epoch": 0.48005502063273725, "grad_norm": 0.48642420768737793, "learning_rate": 0.0005566938458026411, "loss": 2.4135, "step": 1047 }, { "epoch": 0.4805135259055479, "grad_norm": 0.49303334951400757, "learning_rate": 0.0005559558799943514, "loss": 2.3929, "step": 1048 }, { "epoch": 0.48097203117835857, "grad_norm": 0.48339930176734924, "learning_rate": 0.0005552177907265223, "loss": 2.2873, "step": 1049 }, { "epoch": 0.4814305364511692, "grad_norm": 0.46865010261535645, "learning_rate": 0.000554479579627654, "loss": 2.389, "step": 1050 }, { "epoch": 0.48188904172397984, "grad_norm": 0.5025343894958496, "learning_rate": 0.0005537412483265157, "loss": 2.4853, "step": 1051 }, { "epoch": 0.48234754699679044, "grad_norm": 0.4903334975242615, "learning_rate": 0.0005530027984521413, "loss": 2.4409, "step": 1052 }, { "epoch": 0.4828060522696011, "grad_norm": 0.47356104850769043, "learning_rate": 0.0005522642316338268, "loss": 2.3737, "step": 1053 }, { "epoch": 0.4832645575424117, "grad_norm": 0.5065352320671082, "learning_rate": 0.0005515255495011259, "loss": 2.4407, "step": 1054 }, { "epoch": 0.48372306281522237, "grad_norm": 0.4597373604774475, "learning_rate": 0.0005507867536838472, "loss": 2.3811, "step": 1055 }, { "epoch": 0.48418156808803303, "grad_norm": 0.470963716506958, "learning_rate": 0.0005500478458120492, "loss": 2.3519, "step": 1056 }, { "epoch": 0.48464007336084364, "grad_norm": 0.485066294670105, "learning_rate": 0.0005493088275160387, "loss": 2.4019, "step": 1057 }, { "epoch": 0.4850985786336543, "grad_norm": 0.4818322956562042, "learning_rate": 0.0005485697004263657, "loss": 2.46, "step": 1058 }, { "epoch": 0.4855570839064649, "grad_norm": 0.4334261119365692, "learning_rate": 0.0005478304661738199, "loss": 2.4967, "step": 1059 }, { "epoch": 0.48601558917927556, "grad_norm": 0.4701922535896301, "learning_rate": 0.0005470911263894279, "loss": 2.4032, "step": 1060 }, { "epoch": 0.4864740944520862, "grad_norm": 0.4770032465457916, "learning_rate": 0.0005463516827044491, "loss": 2.4518, "step": 1061 }, { "epoch": 0.48693259972489683, "grad_norm": 0.4394857585430145, "learning_rate": 0.000545612136750372, "loss": 2.3963, "step": 1062 }, { "epoch": 0.4873911049977075, "grad_norm": 0.4644238352775574, "learning_rate": 0.0005448724901589107, "loss": 2.5978, "step": 1063 }, { "epoch": 0.4878496102705181, "grad_norm": 0.464792400598526, "learning_rate": 0.0005441327445620014, "loss": 2.4464, "step": 1064 }, { "epoch": 0.48830811554332876, "grad_norm": 0.4908622205257416, "learning_rate": 0.0005433929015917988, "loss": 2.4152, "step": 1065 }, { "epoch": 0.48876662081613936, "grad_norm": 0.43696606159210205, "learning_rate": 0.0005426529628806724, "loss": 2.3594, "step": 1066 }, { "epoch": 0.48922512608895, "grad_norm": 0.4452190399169922, "learning_rate": 0.0005419129300612029, "loss": 2.4478, "step": 1067 }, { "epoch": 0.4896836313617607, "grad_norm": 0.4619537591934204, "learning_rate": 0.000541172804766179, "loss": 2.5209, "step": 1068 }, { "epoch": 0.4901421366345713, "grad_norm": 0.45636606216430664, "learning_rate": 0.0005404325886285927, "loss": 2.4155, "step": 1069 }, { "epoch": 0.49060064190738195, "grad_norm": 0.4738315939903259, "learning_rate": 0.000539692283281637, "loss": 2.4781, "step": 1070 }, { "epoch": 0.49105914718019256, "grad_norm": 0.4536193907260895, "learning_rate": 0.0005389518903587017, "loss": 2.4328, "step": 1071 }, { "epoch": 0.4915176524530032, "grad_norm": 0.4794519543647766, "learning_rate": 0.0005382114114933695, "loss": 2.3854, "step": 1072 }, { "epoch": 0.4919761577258138, "grad_norm": 0.46174702048301697, "learning_rate": 0.0005374708483194132, "loss": 2.4355, "step": 1073 }, { "epoch": 0.4924346629986245, "grad_norm": 0.4814574122428894, "learning_rate": 0.000536730202470791, "loss": 2.3966, "step": 1074 }, { "epoch": 0.49289316827143514, "grad_norm": 0.48740851879119873, "learning_rate": 0.0005359894755816443, "loss": 2.4252, "step": 1075 }, { "epoch": 0.49335167354424575, "grad_norm": 0.45988935232162476, "learning_rate": 0.0005352486692862926, "loss": 2.462, "step": 1076 }, { "epoch": 0.4938101788170564, "grad_norm": 0.45411917567253113, "learning_rate": 0.0005345077852192307, "loss": 2.4788, "step": 1077 }, { "epoch": 0.494268684089867, "grad_norm": 0.4728260040283203, "learning_rate": 0.0005337668250151254, "loss": 2.4183, "step": 1078 }, { "epoch": 0.4947271893626777, "grad_norm": 0.44663146138191223, "learning_rate": 0.0005330257903088111, "loss": 2.4012, "step": 1079 }, { "epoch": 0.4951856946354883, "grad_norm": 0.4989568293094635, "learning_rate": 0.000532284682735287, "loss": 2.3327, "step": 1080 }, { "epoch": 0.49564419990829894, "grad_norm": 0.4485569894313812, "learning_rate": 0.0005315435039297124, "loss": 2.4913, "step": 1081 }, { "epoch": 0.4961027051811096, "grad_norm": 0.43407323956489563, "learning_rate": 0.0005308022555274046, "loss": 2.3984, "step": 1082 }, { "epoch": 0.4965612104539202, "grad_norm": 0.4361623525619507, "learning_rate": 0.0005300609391638336, "loss": 2.3418, "step": 1083 }, { "epoch": 0.49701971572673087, "grad_norm": 0.4912613332271576, "learning_rate": 0.0005293195564746201, "loss": 2.401, "step": 1084 }, { "epoch": 0.4974782209995415, "grad_norm": 0.5030926465988159, "learning_rate": 0.0005285781090955304, "loss": 2.4456, "step": 1085 }, { "epoch": 0.49793672627235214, "grad_norm": 0.4453772008419037, "learning_rate": 0.0005278365986624743, "loss": 2.374, "step": 1086 }, { "epoch": 0.49839523154516274, "grad_norm": 0.45499685406684875, "learning_rate": 0.0005270950268115001, "loss": 2.3859, "step": 1087 }, { "epoch": 0.4988537368179734, "grad_norm": 0.45490923523902893, "learning_rate": 0.0005263533951787919, "loss": 2.4459, "step": 1088 }, { "epoch": 0.49931224209078406, "grad_norm": 0.4579101800918579, "learning_rate": 0.000525611705400666, "loss": 2.4912, "step": 1089 }, { "epoch": 0.49977074736359467, "grad_norm": 0.430134117603302, "learning_rate": 0.0005248699591135664, "loss": 2.4094, "step": 1090 }, { "epoch": 0.5002292526364053, "grad_norm": 0.48007792234420776, "learning_rate": 0.0005241281579540618, "loss": 2.4179, "step": 1091 }, { "epoch": 0.5006877579092159, "grad_norm": 0.5402202606201172, "learning_rate": 0.0005233863035588427, "loss": 2.4971, "step": 1092 }, { "epoch": 0.5011462631820266, "grad_norm": 0.5043858885765076, "learning_rate": 0.0005226443975647161, "loss": 2.5098, "step": 1093 }, { "epoch": 0.5016047684548373, "grad_norm": 0.4517482817173004, "learning_rate": 0.0005219024416086036, "loss": 2.3777, "step": 1094 }, { "epoch": 0.5020632737276479, "grad_norm": 0.45503416657447815, "learning_rate": 0.0005211604373275366, "loss": 2.4683, "step": 1095 }, { "epoch": 0.5025217790004585, "grad_norm": 0.4734433889389038, "learning_rate": 0.0005204183863586533, "loss": 2.3438, "step": 1096 }, { "epoch": 0.5029802842732691, "grad_norm": 0.5009045600891113, "learning_rate": 0.0005196762903391951, "loss": 2.4718, "step": 1097 }, { "epoch": 0.5034387895460798, "grad_norm": 0.47185084223747253, "learning_rate": 0.0005189341509065023, "loss": 2.4445, "step": 1098 }, { "epoch": 0.5038972948188905, "grad_norm": 0.4606148600578308, "learning_rate": 0.0005181919696980112, "loss": 2.4616, "step": 1099 }, { "epoch": 0.504355800091701, "grad_norm": 0.47614261507987976, "learning_rate": 0.0005174497483512506, "loss": 2.452, "step": 1100 }, { "epoch": 0.5048143053645117, "grad_norm": 0.48151203989982605, "learning_rate": 0.0005167074885038374, "loss": 2.3429, "step": 1101 }, { "epoch": 0.5052728106373223, "grad_norm": 0.444448858499527, "learning_rate": 0.0005159651917934735, "loss": 2.3807, "step": 1102 }, { "epoch": 0.505731315910133, "grad_norm": 0.4722660183906555, "learning_rate": 0.0005152228598579428, "loss": 2.493, "step": 1103 }, { "epoch": 0.5061898211829436, "grad_norm": 0.4876071810722351, "learning_rate": 0.000514480494335106, "loss": 2.4245, "step": 1104 }, { "epoch": 0.5066483264557542, "grad_norm": 0.4380664527416229, "learning_rate": 0.0005137380968628983, "loss": 2.4385, "step": 1105 }, { "epoch": 0.5071068317285649, "grad_norm": 0.46031010150909424, "learning_rate": 0.0005129956690793255, "loss": 2.3949, "step": 1106 }, { "epoch": 0.5075653370013755, "grad_norm": 0.4446025490760803, "learning_rate": 0.0005122532126224601, "loss": 2.4544, "step": 1107 }, { "epoch": 0.5080238422741862, "grad_norm": 0.46066731214523315, "learning_rate": 0.0005115107291304378, "loss": 2.4608, "step": 1108 }, { "epoch": 0.5084823475469968, "grad_norm": 0.45820853114128113, "learning_rate": 0.0005107682202414544, "loss": 2.4482, "step": 1109 }, { "epoch": 0.5089408528198074, "grad_norm": 0.45146963000297546, "learning_rate": 0.0005100256875937613, "loss": 2.3832, "step": 1110 }, { "epoch": 0.509399358092618, "grad_norm": 0.47720983624458313, "learning_rate": 0.0005092831328256625, "loss": 2.4941, "step": 1111 }, { "epoch": 0.5098578633654287, "grad_norm": 0.4544755220413208, "learning_rate": 0.0005085405575755105, "loss": 2.4298, "step": 1112 }, { "epoch": 0.5103163686382394, "grad_norm": 0.465750128030777, "learning_rate": 0.0005077979634817034, "loss": 2.3637, "step": 1113 }, { "epoch": 0.51077487391105, "grad_norm": 0.46850112080574036, "learning_rate": 0.0005070553521826808, "loss": 2.406, "step": 1114 }, { "epoch": 0.5112333791838606, "grad_norm": 0.46745410561561584, "learning_rate": 0.00050631272531692, "loss": 2.5051, "step": 1115 }, { "epoch": 0.5116918844566712, "grad_norm": 0.4481866955757141, "learning_rate": 0.0005055700845229327, "loss": 2.3857, "step": 1116 }, { "epoch": 0.5121503897294819, "grad_norm": 0.4744948446750641, "learning_rate": 0.000504827431439262, "loss": 2.4969, "step": 1117 }, { "epoch": 0.5126088950022926, "grad_norm": 0.45592689514160156, "learning_rate": 0.000504084767704477, "loss": 2.4568, "step": 1118 }, { "epoch": 0.5130674002751031, "grad_norm": 0.46528419852256775, "learning_rate": 0.0005033420949571712, "loss": 2.3909, "step": 1119 }, { "epoch": 0.5135259055479138, "grad_norm": 0.4565544128417969, "learning_rate": 0.0005025994148359574, "loss": 2.4317, "step": 1120 }, { "epoch": 0.5139844108207244, "grad_norm": 0.499233603477478, "learning_rate": 0.0005018567289794651, "loss": 2.3184, "step": 1121 }, { "epoch": 0.5144429160935351, "grad_norm": 0.5171236991882324, "learning_rate": 0.0005011140390263362, "loss": 2.4826, "step": 1122 }, { "epoch": 0.5149014213663458, "grad_norm": 0.4707772135734558, "learning_rate": 0.0005003713466152218, "loss": 2.4064, "step": 1123 }, { "epoch": 0.5153599266391563, "grad_norm": 0.47825464606285095, "learning_rate": 0.0004996286533847783, "loss": 2.4719, "step": 1124 }, { "epoch": 0.515818431911967, "grad_norm": 0.46984946727752686, "learning_rate": 0.000498885960973664, "loss": 2.4367, "step": 1125 }, { "epoch": 0.5162769371847776, "grad_norm": 0.48586684465408325, "learning_rate": 0.000498143271020535, "loss": 2.4077, "step": 1126 }, { "epoch": 0.5167354424575883, "grad_norm": 0.4825684130191803, "learning_rate": 0.0004974005851640428, "loss": 2.5067, "step": 1127 }, { "epoch": 0.517193947730399, "grad_norm": 0.48276928067207336, "learning_rate": 0.000496657905042829, "loss": 2.4601, "step": 1128 }, { "epoch": 0.5176524530032095, "grad_norm": 0.47224485874176025, "learning_rate": 0.0004959152322955232, "loss": 2.4052, "step": 1129 }, { "epoch": 0.5181109582760202, "grad_norm": 0.49594688415527344, "learning_rate": 0.0004951725685607382, "loss": 2.5322, "step": 1130 }, { "epoch": 0.5185694635488308, "grad_norm": 0.45508846640586853, "learning_rate": 0.0004944299154770673, "loss": 2.353, "step": 1131 }, { "epoch": 0.5190279688216415, "grad_norm": 0.46311867237091064, "learning_rate": 0.0004936872746830802, "loss": 2.3963, "step": 1132 }, { "epoch": 0.519486474094452, "grad_norm": 0.4325368106365204, "learning_rate": 0.0004929446478173195, "loss": 2.4252, "step": 1133 }, { "epoch": 0.5199449793672627, "grad_norm": 0.46993157267570496, "learning_rate": 0.0004922020365182968, "loss": 2.396, "step": 1134 }, { "epoch": 0.5204034846400734, "grad_norm": 0.4870544672012329, "learning_rate": 0.0004914594424244897, "loss": 2.4251, "step": 1135 }, { "epoch": 0.520861989912884, "grad_norm": 0.4807164669036865, "learning_rate": 0.0004907168671743376, "loss": 2.4154, "step": 1136 }, { "epoch": 0.5213204951856947, "grad_norm": 0.4675467908382416, "learning_rate": 0.0004899743124062387, "loss": 2.4241, "step": 1137 }, { "epoch": 0.5217790004585052, "grad_norm": 0.4891907572746277, "learning_rate": 0.0004892317797585456, "loss": 2.3654, "step": 1138 }, { "epoch": 0.5222375057313159, "grad_norm": 0.4500700533390045, "learning_rate": 0.0004884892708695623, "loss": 2.3159, "step": 1139 }, { "epoch": 0.5226960110041265, "grad_norm": 0.4609105885028839, "learning_rate": 0.0004877467873775402, "loss": 2.4362, "step": 1140 }, { "epoch": 0.5231545162769372, "grad_norm": 0.4727395176887512, "learning_rate": 0.00048700433092067473, "loss": 2.4942, "step": 1141 }, { "epoch": 0.5236130215497479, "grad_norm": 0.48160243034362793, "learning_rate": 0.0004862619031371019, "loss": 2.42, "step": 1142 }, { "epoch": 0.5240715268225584, "grad_norm": 0.4488910138607025, "learning_rate": 0.0004855195056648942, "loss": 2.3754, "step": 1143 }, { "epoch": 0.5245300320953691, "grad_norm": 0.45488911867141724, "learning_rate": 0.00048477714014205734, "loss": 2.4101, "step": 1144 }, { "epoch": 0.5249885373681797, "grad_norm": 0.4609996974468231, "learning_rate": 0.00048403480820652644, "loss": 2.4767, "step": 1145 }, { "epoch": 0.5254470426409904, "grad_norm": 0.4567868411540985, "learning_rate": 0.0004832925114961629, "loss": 2.4219, "step": 1146 }, { "epoch": 0.5259055479138011, "grad_norm": 0.4514508843421936, "learning_rate": 0.0004825502516487497, "loss": 2.439, "step": 1147 }, { "epoch": 0.5263640531866116, "grad_norm": 0.45852819085121155, "learning_rate": 0.00048180803030198896, "loss": 2.5136, "step": 1148 }, { "epoch": 0.5268225584594223, "grad_norm": 0.468732088804245, "learning_rate": 0.0004810658490934979, "loss": 2.4528, "step": 1149 }, { "epoch": 0.5272810637322329, "grad_norm": 0.43933671712875366, "learning_rate": 0.000480323709660805, "loss": 2.5201, "step": 1150 }, { "epoch": 0.5277395690050436, "grad_norm": 0.4712791442871094, "learning_rate": 0.0004795816136413467, "loss": 2.3253, "step": 1151 }, { "epoch": 0.5281980742778541, "grad_norm": 0.43904587626457214, "learning_rate": 0.00047883956267246353, "loss": 2.4494, "step": 1152 }, { "epoch": 0.5286565795506648, "grad_norm": 0.4487941861152649, "learning_rate": 0.00047809755839139657, "loss": 2.5005, "step": 1153 }, { "epoch": 0.5291150848234755, "grad_norm": 0.48183509707450867, "learning_rate": 0.0004773556024352841, "loss": 2.4496, "step": 1154 }, { "epoch": 0.5295735900962861, "grad_norm": 0.44479262828826904, "learning_rate": 0.00047661369644115754, "loss": 2.305, "step": 1155 }, { "epoch": 0.5300320953690968, "grad_norm": 0.45462387800216675, "learning_rate": 0.0004758718420459383, "loss": 2.3372, "step": 1156 }, { "epoch": 0.5304906006419073, "grad_norm": 0.4659642279148102, "learning_rate": 0.0004751300408864339, "loss": 2.3758, "step": 1157 }, { "epoch": 0.530949105914718, "grad_norm": 0.4309777617454529, "learning_rate": 0.00047438829459933414, "loss": 2.4386, "step": 1158 }, { "epoch": 0.5314076111875287, "grad_norm": 0.4634147882461548, "learning_rate": 0.0004736466048212082, "loss": 2.4704, "step": 1159 }, { "epoch": 0.5318661164603393, "grad_norm": 0.4872623085975647, "learning_rate": 0.0004729049731885002, "loss": 2.3949, "step": 1160 }, { "epoch": 0.53232462173315, "grad_norm": 0.4506804049015045, "learning_rate": 0.000472163401337526, "loss": 2.3752, "step": 1161 }, { "epoch": 0.5327831270059605, "grad_norm": 0.4434259831905365, "learning_rate": 0.00047142189090446985, "loss": 2.3686, "step": 1162 }, { "epoch": 0.5332416322787712, "grad_norm": 0.46903085708618164, "learning_rate": 0.0004706804435253802, "loss": 2.5143, "step": 1163 }, { "epoch": 0.5337001375515819, "grad_norm": 0.45211631059646606, "learning_rate": 0.0004699390608361665, "loss": 2.5507, "step": 1164 }, { "epoch": 0.5341586428243925, "grad_norm": 0.45304545760154724, "learning_rate": 0.0004691977444725955, "loss": 2.3912, "step": 1165 }, { "epoch": 0.5346171480972031, "grad_norm": 0.45792579650878906, "learning_rate": 0.0004684564960702877, "loss": 2.4349, "step": 1166 }, { "epoch": 0.5350756533700137, "grad_norm": 0.4905306398868561, "learning_rate": 0.0004677153172647131, "loss": 2.4195, "step": 1167 }, { "epoch": 0.5355341586428244, "grad_norm": 0.48034757375717163, "learning_rate": 0.00046697420969118894, "loss": 2.5028, "step": 1168 }, { "epoch": 0.535992663915635, "grad_norm": 0.47530466318130493, "learning_rate": 0.00046623317498487466, "loss": 2.3807, "step": 1169 }, { "epoch": 0.5364511691884457, "grad_norm": 0.4725816249847412, "learning_rate": 0.0004654922147807694, "loss": 2.5301, "step": 1170 }, { "epoch": 0.5369096744612563, "grad_norm": 0.44743382930755615, "learning_rate": 0.00046475133071370757, "loss": 2.3646, "step": 1171 }, { "epoch": 0.5373681797340669, "grad_norm": 0.4968271851539612, "learning_rate": 0.00046401052441835574, "loss": 2.4177, "step": 1172 }, { "epoch": 0.5378266850068776, "grad_norm": 0.4540119171142578, "learning_rate": 0.000463269797529209, "loss": 2.313, "step": 1173 }, { "epoch": 0.5382851902796882, "grad_norm": 0.44216349720954895, "learning_rate": 0.00046252915168058697, "loss": 2.4743, "step": 1174 }, { "epoch": 0.5387436955524989, "grad_norm": 0.48889845609664917, "learning_rate": 0.0004617885885066305, "loss": 2.3639, "step": 1175 }, { "epoch": 0.5392022008253095, "grad_norm": 0.464162677526474, "learning_rate": 0.0004610481096412984, "loss": 2.4231, "step": 1176 }, { "epoch": 0.5396607060981201, "grad_norm": 0.4708946645259857, "learning_rate": 0.000460307716718363, "loss": 2.4376, "step": 1177 }, { "epoch": 0.5401192113709308, "grad_norm": 0.47956669330596924, "learning_rate": 0.0004595674113714074, "loss": 2.3855, "step": 1178 }, { "epoch": 0.5405777166437414, "grad_norm": 0.4446319043636322, "learning_rate": 0.0004588271952338212, "loss": 2.4343, "step": 1179 }, { "epoch": 0.5410362219165521, "grad_norm": 0.4957142770290375, "learning_rate": 0.00045808706993879714, "loss": 2.5348, "step": 1180 }, { "epoch": 0.5414947271893626, "grad_norm": 0.46579092741012573, "learning_rate": 0.00045734703711932767, "loss": 2.4138, "step": 1181 }, { "epoch": 0.5419532324621733, "grad_norm": 0.4367072284221649, "learning_rate": 0.0004566070984082013, "loss": 2.3411, "step": 1182 }, { "epoch": 0.542411737734984, "grad_norm": 0.463439017534256, "learning_rate": 0.00045586725543799865, "loss": 2.3358, "step": 1183 }, { "epoch": 0.5428702430077946, "grad_norm": 0.46225494146347046, "learning_rate": 0.00045512750984108937, "loss": 2.4695, "step": 1184 }, { "epoch": 0.5433287482806052, "grad_norm": 0.44770553708076477, "learning_rate": 0.000454387863249628, "loss": 2.531, "step": 1185 }, { "epoch": 0.5437872535534158, "grad_norm": 0.47738131880760193, "learning_rate": 0.00045364831729555096, "loss": 2.3792, "step": 1186 }, { "epoch": 0.5442457588262265, "grad_norm": 0.484330952167511, "learning_rate": 0.0004529088736105721, "loss": 2.483, "step": 1187 }, { "epoch": 0.5447042640990372, "grad_norm": 0.4538387954235077, "learning_rate": 0.0004521695338261802, "loss": 2.4824, "step": 1188 }, { "epoch": 0.5451627693718478, "grad_norm": 0.4453546702861786, "learning_rate": 0.0004514302995736344, "loss": 2.4409, "step": 1189 }, { "epoch": 0.5456212746446584, "grad_norm": 0.4559042453765869, "learning_rate": 0.0004506911724839613, "loss": 2.4223, "step": 1190 }, { "epoch": 0.546079779917469, "grad_norm": 0.4692598879337311, "learning_rate": 0.0004499521541879508, "loss": 2.3858, "step": 1191 }, { "epoch": 0.5465382851902797, "grad_norm": 0.46065378189086914, "learning_rate": 0.00044921324631615303, "loss": 2.4233, "step": 1192 }, { "epoch": 0.5469967904630904, "grad_norm": 0.4415675699710846, "learning_rate": 0.0004484744504988742, "loss": 2.3955, "step": 1193 }, { "epoch": 0.547455295735901, "grad_norm": 0.5197876691818237, "learning_rate": 0.00044773576836617336, "loss": 2.3873, "step": 1194 }, { "epoch": 0.5479138010087116, "grad_norm": 0.4390837550163269, "learning_rate": 0.0004469972015478588, "loss": 2.3617, "step": 1195 }, { "epoch": 0.5483723062815222, "grad_norm": 0.45728355646133423, "learning_rate": 0.0004462587516734844, "loss": 2.4087, "step": 1196 }, { "epoch": 0.5488308115543329, "grad_norm": 0.44679996371269226, "learning_rate": 0.00044552042037234596, "loss": 2.435, "step": 1197 }, { "epoch": 0.5492893168271435, "grad_norm": 0.4571087062358856, "learning_rate": 0.00044478220927347774, "loss": 2.4544, "step": 1198 }, { "epoch": 0.5497478220999541, "grad_norm": 0.43499836325645447, "learning_rate": 0.00044404412000564875, "loss": 2.3437, "step": 1199 }, { "epoch": 0.5502063273727648, "grad_norm": 0.49141645431518555, "learning_rate": 0.000443306154197359, "loss": 2.426, "step": 1200 }, { "epoch": 0.5506648326455754, "grad_norm": 0.46360987424850464, "learning_rate": 0.00044256831347683646, "loss": 2.3723, "step": 1201 }, { "epoch": 0.5511233379183861, "grad_norm": 0.5013500452041626, "learning_rate": 0.0004418305994720328, "loss": 2.4431, "step": 1202 }, { "epoch": 0.5515818431911967, "grad_norm": 0.4450450539588928, "learning_rate": 0.0004410930138106203, "loss": 2.3876, "step": 1203 }, { "epoch": 0.5520403484640073, "grad_norm": 0.4502420425415039, "learning_rate": 0.000440355558119988, "loss": 2.4738, "step": 1204 }, { "epoch": 0.552498853736818, "grad_norm": 0.4705275297164917, "learning_rate": 0.00043961823402723814, "loss": 2.397, "step": 1205 }, { "epoch": 0.5529573590096286, "grad_norm": 0.4538814127445221, "learning_rate": 0.0004388810431591829, "loss": 2.5244, "step": 1206 }, { "epoch": 0.5534158642824393, "grad_norm": 0.47246885299682617, "learning_rate": 0.0004381439871423398, "loss": 2.3954, "step": 1207 }, { "epoch": 0.5538743695552499, "grad_norm": 0.43410706520080566, "learning_rate": 0.00043740706760292966, "loss": 2.3405, "step": 1208 }, { "epoch": 0.5543328748280605, "grad_norm": 0.44099968671798706, "learning_rate": 0.0004366702861668716, "loss": 2.465, "step": 1209 }, { "epoch": 0.5547913801008711, "grad_norm": 0.4514848291873932, "learning_rate": 0.00043593364445978036, "loss": 2.4718, "step": 1210 }, { "epoch": 0.5552498853736818, "grad_norm": 0.500238835811615, "learning_rate": 0.0004351971441069622, "loss": 2.3342, "step": 1211 }, { "epoch": 0.5557083906464925, "grad_norm": 0.46396932005882263, "learning_rate": 0.0004344607867334116, "loss": 2.3553, "step": 1212 }, { "epoch": 0.556166895919303, "grad_norm": 0.45996665954589844, "learning_rate": 0.00043372457396380766, "loss": 2.342, "step": 1213 }, { "epoch": 0.5566254011921137, "grad_norm": 0.5037462115287781, "learning_rate": 0.00043298850742251013, "loss": 2.5729, "step": 1214 }, { "epoch": 0.5570839064649243, "grad_norm": 0.47675326466560364, "learning_rate": 0.0004322525887335563, "loss": 2.3878, "step": 1215 }, { "epoch": 0.557542411737735, "grad_norm": 0.47130027413368225, "learning_rate": 0.00043151681952065734, "loss": 2.4985, "step": 1216 }, { "epoch": 0.5580009170105457, "grad_norm": 0.49527254700660706, "learning_rate": 0.00043078120140719456, "loss": 2.4946, "step": 1217 }, { "epoch": 0.5584594222833562, "grad_norm": 0.4859922230243683, "learning_rate": 0.0004300457360162158, "loss": 2.4128, "step": 1218 }, { "epoch": 0.5589179275561669, "grad_norm": 0.47722741961479187, "learning_rate": 0.0004293104249704319, "loss": 2.3616, "step": 1219 }, { "epoch": 0.5593764328289775, "grad_norm": 0.4716128706932068, "learning_rate": 0.00042857526989221355, "loss": 2.4979, "step": 1220 }, { "epoch": 0.5598349381017882, "grad_norm": 0.4866699278354645, "learning_rate": 0.00042784027240358674, "loss": 2.4578, "step": 1221 }, { "epoch": 0.5602934433745989, "grad_norm": 0.5027143359184265, "learning_rate": 0.0004271054341262301, "loss": 2.3978, "step": 1222 }, { "epoch": 0.5607519486474094, "grad_norm": 0.45892077684402466, "learning_rate": 0.000426370756681471, "loss": 2.4042, "step": 1223 }, { "epoch": 0.5612104539202201, "grad_norm": 0.46669384837150574, "learning_rate": 0.0004256362416902817, "loss": 2.4514, "step": 1224 }, { "epoch": 0.5616689591930307, "grad_norm": 0.492447167634964, "learning_rate": 0.00042490189077327637, "loss": 2.384, "step": 1225 }, { "epoch": 0.5621274644658414, "grad_norm": 0.49825209379196167, "learning_rate": 0.00042416770555070703, "loss": 2.4668, "step": 1226 }, { "epoch": 0.562585969738652, "grad_norm": 0.4538726508617401, "learning_rate": 0.00042343368764246, "loss": 2.2988, "step": 1227 }, { "epoch": 0.5630444750114626, "grad_norm": 0.4654730260372162, "learning_rate": 0.0004226998386680524, "loss": 2.5026, "step": 1228 }, { "epoch": 0.5635029802842733, "grad_norm": 0.4798143208026886, "learning_rate": 0.000421966160246629, "loss": 2.5101, "step": 1229 }, { "epoch": 0.5639614855570839, "grad_norm": 0.47309383749961853, "learning_rate": 0.00042123265399695783, "loss": 2.4078, "step": 1230 }, { "epoch": 0.5644199908298946, "grad_norm": 0.4469056725502014, "learning_rate": 0.0004204993215374273, "loss": 2.5544, "step": 1231 }, { "epoch": 0.5648784961027051, "grad_norm": 0.4581888020038605, "learning_rate": 0.00041976616448604226, "loss": 2.4836, "step": 1232 }, { "epoch": 0.5653370013755158, "grad_norm": 0.46397706866264343, "learning_rate": 0.00041903318446042076, "loss": 2.4301, "step": 1233 }, { "epoch": 0.5657955066483265, "grad_norm": 0.45031216740608215, "learning_rate": 0.00041830038307778984, "loss": 2.4747, "step": 1234 }, { "epoch": 0.5662540119211371, "grad_norm": 0.47439906001091003, "learning_rate": 0.0004175677619549828, "loss": 2.5083, "step": 1235 }, { "epoch": 0.5667125171939478, "grad_norm": 0.4676271677017212, "learning_rate": 0.000416835322708435, "loss": 2.3313, "step": 1236 }, { "epoch": 0.5671710224667583, "grad_norm": 0.4627252519130707, "learning_rate": 0.00041610306695418056, "loss": 2.3922, "step": 1237 }, { "epoch": 0.567629527739569, "grad_norm": 0.5063115954399109, "learning_rate": 0.0004153709963078488, "loss": 2.4749, "step": 1238 }, { "epoch": 0.5680880330123796, "grad_norm": 0.4810315668582916, "learning_rate": 0.0004146391123846606, "loss": 2.4316, "step": 1239 }, { "epoch": 0.5685465382851903, "grad_norm": 0.44339436292648315, "learning_rate": 0.0004139074167994249, "loss": 2.4508, "step": 1240 }, { "epoch": 0.569005043558001, "grad_norm": 0.45793694257736206, "learning_rate": 0.00041317591116653486, "loss": 2.4327, "step": 1241 }, { "epoch": 0.5694635488308115, "grad_norm": 0.4341580867767334, "learning_rate": 0.0004124445970999648, "loss": 2.4488, "step": 1242 }, { "epoch": 0.5699220541036222, "grad_norm": 0.5226630568504333, "learning_rate": 0.00041171347621326627, "loss": 2.4174, "step": 1243 }, { "epoch": 0.5703805593764328, "grad_norm": 0.4769025146961212, "learning_rate": 0.00041098255011956465, "loss": 2.4799, "step": 1244 }, { "epoch": 0.5708390646492435, "grad_norm": 0.46113377809524536, "learning_rate": 0.00041025182043155547, "loss": 2.4859, "step": 1245 }, { "epoch": 0.571297569922054, "grad_norm": 0.49879127740859985, "learning_rate": 0.000409521288761501, "loss": 2.4945, "step": 1246 }, { "epoch": 0.5717560751948647, "grad_norm": 0.49018150568008423, "learning_rate": 0.00040879095672122646, "loss": 2.3806, "step": 1247 }, { "epoch": 0.5722145804676754, "grad_norm": 0.4963681995868683, "learning_rate": 0.0004080608259221167, "loss": 2.4494, "step": 1248 }, { "epoch": 0.572673085740486, "grad_norm": 0.4578735828399658, "learning_rate": 0.0004073308979751126, "loss": 2.418, "step": 1249 }, { "epoch": 0.5731315910132967, "grad_norm": 0.48955196142196655, "learning_rate": 0.0004066011744907074, "loss": 2.319, "step": 1250 }, { "epoch": 0.5735900962861072, "grad_norm": 0.47079023718833923, "learning_rate": 0.00040587165707894326, "loss": 2.4179, "step": 1251 }, { "epoch": 0.5740486015589179, "grad_norm": 0.44356846809387207, "learning_rate": 0.0004051423473494076, "loss": 2.3753, "step": 1252 }, { "epoch": 0.5745071068317286, "grad_norm": 0.45391392707824707, "learning_rate": 0.0004044132469112299, "loss": 2.4423, "step": 1253 }, { "epoch": 0.5749656121045392, "grad_norm": 0.48760169744491577, "learning_rate": 0.00040368435737307733, "loss": 2.4986, "step": 1254 }, { "epoch": 0.5754241173773499, "grad_norm": 0.4667738378047943, "learning_rate": 0.00040295568034315224, "loss": 2.3722, "step": 1255 }, { "epoch": 0.5758826226501604, "grad_norm": 0.4450877606868744, "learning_rate": 0.0004022272174291878, "loss": 2.4672, "step": 1256 }, { "epoch": 0.5763411279229711, "grad_norm": 0.4708748757839203, "learning_rate": 0.0004014989702384449, "loss": 2.3707, "step": 1257 }, { "epoch": 0.5767996331957818, "grad_norm": 0.44671207666397095, "learning_rate": 0.00040077094037770843, "loss": 2.4252, "step": 1258 }, { "epoch": 0.5772581384685924, "grad_norm": 0.4748927652835846, "learning_rate": 0.0004000431294532838, "loss": 2.3899, "step": 1259 }, { "epoch": 0.5777166437414031, "grad_norm": 0.45330965518951416, "learning_rate": 0.0003993155390709935, "loss": 2.4282, "step": 1260 }, { "epoch": 0.5781751490142136, "grad_norm": 0.42812666296958923, "learning_rate": 0.0003985881708361729, "loss": 2.4749, "step": 1261 }, { "epoch": 0.5786336542870243, "grad_norm": 0.4438706040382385, "learning_rate": 0.00039786102635366784, "loss": 2.3517, "step": 1262 }, { "epoch": 0.579092159559835, "grad_norm": 0.44825753569602966, "learning_rate": 0.0003971341072278302, "loss": 2.4501, "step": 1263 }, { "epoch": 0.5795506648326456, "grad_norm": 0.45674365758895874, "learning_rate": 0.00039640741506251457, "loss": 2.4858, "step": 1264 }, { "epoch": 0.5800091701054562, "grad_norm": 0.4897169768810272, "learning_rate": 0.00039568095146107495, "loss": 2.3867, "step": 1265 }, { "epoch": 0.5804676753782668, "grad_norm": 0.4745970070362091, "learning_rate": 0.00039495471802636096, "loss": 2.569, "step": 1266 }, { "epoch": 0.5809261806510775, "grad_norm": 0.4774334132671356, "learning_rate": 0.0003942287163607145, "loss": 2.3512, "step": 1267 }, { "epoch": 0.5813846859238881, "grad_norm": 0.47877055406570435, "learning_rate": 0.0003935029480659658, "loss": 2.4466, "step": 1268 }, { "epoch": 0.5818431911966988, "grad_norm": 0.4327148497104645, "learning_rate": 0.00039277741474343054, "loss": 2.5017, "step": 1269 }, { "epoch": 0.5823016964695094, "grad_norm": 0.46775880455970764, "learning_rate": 0.0003920521179939057, "loss": 2.4385, "step": 1270 }, { "epoch": 0.58276020174232, "grad_norm": 0.4751521050930023, "learning_rate": 0.00039132705941766644, "loss": 2.5192, "step": 1271 }, { "epoch": 0.5832187070151307, "grad_norm": 0.4553193747997284, "learning_rate": 0.0003906022406144624, "loss": 2.3327, "step": 1272 }, { "epoch": 0.5836772122879413, "grad_norm": 0.4566701650619507, "learning_rate": 0.0003898776631835143, "loss": 2.4655, "step": 1273 }, { "epoch": 0.584135717560752, "grad_norm": 0.48396119475364685, "learning_rate": 0.00038915332872350994, "loss": 2.3514, "step": 1274 }, { "epoch": 0.5845942228335625, "grad_norm": 0.43196284770965576, "learning_rate": 0.00038842923883260135, "loss": 2.3358, "step": 1275 }, { "epoch": 0.5850527281063732, "grad_norm": 0.4765409827232361, "learning_rate": 0.00038770539510840093, "loss": 2.4924, "step": 1276 }, { "epoch": 0.5855112333791839, "grad_norm": 0.441554456949234, "learning_rate": 0.00038698179914797783, "loss": 2.4306, "step": 1277 }, { "epoch": 0.5859697386519945, "grad_norm": 0.4550929665565491, "learning_rate": 0.0003862584525478545, "loss": 2.4724, "step": 1278 }, { "epoch": 0.5864282439248051, "grad_norm": 0.4530290961265564, "learning_rate": 0.00038553535690400353, "loss": 2.4821, "step": 1279 }, { "epoch": 0.5868867491976157, "grad_norm": 0.4560193419456482, "learning_rate": 0.00038481251381184355, "loss": 2.3897, "step": 1280 }, { "epoch": 0.5873452544704264, "grad_norm": 0.4298976957798004, "learning_rate": 0.00038408992486623584, "loss": 2.4385, "step": 1281 }, { "epoch": 0.5878037597432371, "grad_norm": 0.46507301926612854, "learning_rate": 0.00038336759166148117, "loss": 2.5216, "step": 1282 }, { "epoch": 0.5882622650160477, "grad_norm": 0.4396106004714966, "learning_rate": 0.0003826455157913159, "loss": 2.3097, "step": 1283 }, { "epoch": 0.5887207702888583, "grad_norm": 0.4433026611804962, "learning_rate": 0.00038192369884890886, "loss": 2.4573, "step": 1284 }, { "epoch": 0.5891792755616689, "grad_norm": 0.4513545334339142, "learning_rate": 0.00038120214242685723, "loss": 2.3623, "step": 1285 }, { "epoch": 0.5896377808344796, "grad_norm": 0.457552433013916, "learning_rate": 0.00038048084811718373, "loss": 2.3951, "step": 1286 }, { "epoch": 0.5900962861072903, "grad_norm": 0.4743898808956146, "learning_rate": 0.0003797598175113327, "loss": 2.3952, "step": 1287 }, { "epoch": 0.5905547913801009, "grad_norm": 0.4375896751880646, "learning_rate": 0.0003790390522001662, "loss": 2.3618, "step": 1288 }, { "epoch": 0.5910132966529115, "grad_norm": 0.4473514258861542, "learning_rate": 0.0003783185537739615, "loss": 2.4702, "step": 1289 }, { "epoch": 0.5914718019257221, "grad_norm": 0.47537970542907715, "learning_rate": 0.00037759832382240697, "loss": 2.448, "step": 1290 }, { "epoch": 0.5919303071985328, "grad_norm": 0.451750785112381, "learning_rate": 0.00037687836393459826, "loss": 2.4142, "step": 1291 }, { "epoch": 0.5923888124713435, "grad_norm": 0.5145715475082397, "learning_rate": 0.0003761586756990354, "loss": 2.4232, "step": 1292 }, { "epoch": 0.5928473177441541, "grad_norm": 0.46456849575042725, "learning_rate": 0.0003754392607036191, "loss": 2.4775, "step": 1293 }, { "epoch": 0.5933058230169647, "grad_norm": 0.4662027359008789, "learning_rate": 0.0003747201205356472, "loss": 2.4511, "step": 1294 }, { "epoch": 0.5937643282897753, "grad_norm": 0.5032756924629211, "learning_rate": 0.0003740012567818111, "loss": 2.561, "step": 1295 }, { "epoch": 0.594222833562586, "grad_norm": 0.4580008387565613, "learning_rate": 0.0003732826710281922, "loss": 2.3591, "step": 1296 }, { "epoch": 0.5946813388353966, "grad_norm": 0.4460383653640747, "learning_rate": 0.0003725643648602588, "loss": 2.3869, "step": 1297 }, { "epoch": 0.5951398441082072, "grad_norm": 0.47869575023651123, "learning_rate": 0.0003718463398628621, "loss": 2.4588, "step": 1298 }, { "epoch": 0.5955983493810179, "grad_norm": 0.4519692361354828, "learning_rate": 0.0003711285976202331, "loss": 2.4468, "step": 1299 }, { "epoch": 0.5960568546538285, "grad_norm": 0.4564054012298584, "learning_rate": 0.0003704111397159787, "loss": 2.3795, "step": 1300 }, { "epoch": 0.5965153599266392, "grad_norm": 0.42385393381118774, "learning_rate": 0.0003696939677330788, "loss": 2.3911, "step": 1301 }, { "epoch": 0.5969738651994498, "grad_norm": 0.4492715299129486, "learning_rate": 0.00036897708325388213, "loss": 2.3603, "step": 1302 }, { "epoch": 0.5974323704722604, "grad_norm": 0.47204527258872986, "learning_rate": 0.0003682604878601034, "loss": 2.3753, "step": 1303 }, { "epoch": 0.597890875745071, "grad_norm": 0.47396522760391235, "learning_rate": 0.000367544183132819, "loss": 2.3591, "step": 1304 }, { "epoch": 0.5983493810178817, "grad_norm": 0.4689193367958069, "learning_rate": 0.0003668281706524645, "loss": 2.3235, "step": 1305 }, { "epoch": 0.5988078862906924, "grad_norm": 0.44321396946907043, "learning_rate": 0.0003661124519988304, "loss": 2.4394, "step": 1306 }, { "epoch": 0.599266391563503, "grad_norm": 0.4379924237728119, "learning_rate": 0.00036539702875105893, "loss": 2.3897, "step": 1307 }, { "epoch": 0.5997248968363136, "grad_norm": 0.4558914005756378, "learning_rate": 0.0003646819024876406, "loss": 2.3995, "step": 1308 }, { "epoch": 0.6001834021091242, "grad_norm": 0.4861031770706177, "learning_rate": 0.0003639670747864107, "loss": 2.4806, "step": 1309 }, { "epoch": 0.6006419073819349, "grad_norm": 0.43611377477645874, "learning_rate": 0.00036325254722454584, "loss": 2.4578, "step": 1310 }, { "epoch": 0.6011004126547456, "grad_norm": 0.4894067049026489, "learning_rate": 0.00036253832137855997, "loss": 2.4325, "step": 1311 }, { "epoch": 0.6015589179275561, "grad_norm": 0.4633631706237793, "learning_rate": 0.00036182439882430183, "loss": 2.4169, "step": 1312 }, { "epoch": 0.6020174232003668, "grad_norm": 0.44995003938674927, "learning_rate": 0.00036111078113695096, "loss": 2.3138, "step": 1313 }, { "epoch": 0.6024759284731774, "grad_norm": 0.4887736737728119, "learning_rate": 0.0003603974698910139, "loss": 2.2985, "step": 1314 }, { "epoch": 0.6029344337459881, "grad_norm": 0.46004626154899597, "learning_rate": 0.0003596844666603214, "loss": 2.3478, "step": 1315 }, { "epoch": 0.6033929390187988, "grad_norm": 0.45907652378082275, "learning_rate": 0.0003589717730180245, "loss": 2.4462, "step": 1316 }, { "epoch": 0.6038514442916093, "grad_norm": 0.45249494910240173, "learning_rate": 0.00035825939053659117, "loss": 2.5165, "step": 1317 }, { "epoch": 0.60430994956442, "grad_norm": 0.45186901092529297, "learning_rate": 0.00035754732078780273, "loss": 2.3298, "step": 1318 }, { "epoch": 0.6047684548372306, "grad_norm": 0.4635755717754364, "learning_rate": 0.00035683556534275076, "loss": 2.5313, "step": 1319 }, { "epoch": 0.6052269601100413, "grad_norm": 0.47196030616760254, "learning_rate": 0.00035612412577183303, "loss": 2.3763, "step": 1320 }, { "epoch": 0.605685465382852, "grad_norm": 0.47723764181137085, "learning_rate": 0.00035541300364475063, "loss": 2.3738, "step": 1321 }, { "epoch": 0.6061439706556625, "grad_norm": 0.44803789258003235, "learning_rate": 0.0003547022005305043, "loss": 2.4799, "step": 1322 }, { "epoch": 0.6066024759284732, "grad_norm": 0.45631173253059387, "learning_rate": 0.0003539917179973907, "loss": 2.4577, "step": 1323 }, { "epoch": 0.6070609812012838, "grad_norm": 0.42769426107406616, "learning_rate": 0.00035328155761299917, "loss": 2.4155, "step": 1324 }, { "epoch": 0.6075194864740945, "grad_norm": 0.4650164842605591, "learning_rate": 0.0003525717209442085, "loss": 2.3152, "step": 1325 }, { "epoch": 0.6079779917469051, "grad_norm": 0.4264846444129944, "learning_rate": 0.00035186220955718306, "loss": 2.2693, "step": 1326 }, { "epoch": 0.6084364970197157, "grad_norm": 0.5137759447097778, "learning_rate": 0.0003511530250173696, "loss": 2.4369, "step": 1327 }, { "epoch": 0.6088950022925264, "grad_norm": 0.44778910279273987, "learning_rate": 0.00035044416888949364, "loss": 2.3674, "step": 1328 }, { "epoch": 0.609353507565337, "grad_norm": 0.4681735038757324, "learning_rate": 0.0003497356427375562, "loss": 2.4377, "step": 1329 }, { "epoch": 0.6098120128381477, "grad_norm": 0.44314271211624146, "learning_rate": 0.00034902744812483034, "loss": 2.4386, "step": 1330 }, { "epoch": 0.6102705181109582, "grad_norm": 0.4682645797729492, "learning_rate": 0.00034831958661385714, "loss": 2.4847, "step": 1331 }, { "epoch": 0.6107290233837689, "grad_norm": 0.4649456739425659, "learning_rate": 0.0003476120597664434, "loss": 2.366, "step": 1332 }, { "epoch": 0.6111875286565795, "grad_norm": 0.447383850812912, "learning_rate": 0.00034690486914365704, "loss": 2.4463, "step": 1333 }, { "epoch": 0.6116460339293902, "grad_norm": 0.47792181372642517, "learning_rate": 0.00034619801630582435, "loss": 2.5176, "step": 1334 }, { "epoch": 0.6121045392022009, "grad_norm": 0.43545645475387573, "learning_rate": 0.00034549150281252633, "loss": 2.3365, "step": 1335 }, { "epoch": 0.6125630444750114, "grad_norm": 0.456363707780838, "learning_rate": 0.0003447853302225952, "loss": 2.4181, "step": 1336 }, { "epoch": 0.6130215497478221, "grad_norm": 0.4809851348400116, "learning_rate": 0.00034407950009411126, "loss": 2.2869, "step": 1337 }, { "epoch": 0.6134800550206327, "grad_norm": 0.4695688784122467, "learning_rate": 0.00034337401398439873, "loss": 2.3475, "step": 1338 }, { "epoch": 0.6139385602934434, "grad_norm": 0.46897369623184204, "learning_rate": 0.00034266887345002305, "loss": 2.3574, "step": 1339 }, { "epoch": 0.6143970655662541, "grad_norm": 0.442411869764328, "learning_rate": 0.0003419640800467874, "loss": 2.4509, "step": 1340 }, { "epoch": 0.6148555708390646, "grad_norm": 0.4758976995944977, "learning_rate": 0.0003412596353297288, "loss": 2.4626, "step": 1341 }, { "epoch": 0.6153140761118753, "grad_norm": 0.46862444281578064, "learning_rate": 0.00034055554085311493, "loss": 2.4015, "step": 1342 }, { "epoch": 0.6157725813846859, "grad_norm": 0.44300785660743713, "learning_rate": 0.00033985179817044105, "loss": 2.4553, "step": 1343 }, { "epoch": 0.6162310866574966, "grad_norm": 0.44375336170196533, "learning_rate": 0.0003391484088344257, "loss": 2.3966, "step": 1344 }, { "epoch": 0.6166895919303071, "grad_norm": 0.43962815403938293, "learning_rate": 0.00033844537439700807, "loss": 2.5086, "step": 1345 }, { "epoch": 0.6171480972031178, "grad_norm": 0.45856571197509766, "learning_rate": 0.00033774269640934445, "loss": 2.4511, "step": 1346 }, { "epoch": 0.6176066024759285, "grad_norm": 0.4313191771507263, "learning_rate": 0.0003370403764218045, "loss": 2.3665, "step": 1347 }, { "epoch": 0.6180651077487391, "grad_norm": 0.42421382665634155, "learning_rate": 0.000336338415983968, "loss": 2.3469, "step": 1348 }, { "epoch": 0.6185236130215498, "grad_norm": 0.4381328523159027, "learning_rate": 0.00033563681664462155, "loss": 2.3473, "step": 1349 }, { "epoch": 0.6189821182943603, "grad_norm": 0.5023850798606873, "learning_rate": 0.000334935579951755, "loss": 2.5855, "step": 1350 }, { "epoch": 0.619440623567171, "grad_norm": 0.4589841663837433, "learning_rate": 0.0003342347074525578, "loss": 2.3668, "step": 1351 }, { "epoch": 0.6198991288399817, "grad_norm": 0.44768619537353516, "learning_rate": 0.0003335342006934161, "loss": 2.4598, "step": 1352 }, { "epoch": 0.6203576341127923, "grad_norm": 0.4619479477405548, "learning_rate": 0.00033283406121990914, "loss": 2.4776, "step": 1353 }, { "epoch": 0.620816139385603, "grad_norm": 0.4636728763580322, "learning_rate": 0.0003321342905768057, "loss": 2.3522, "step": 1354 }, { "epoch": 0.6212746446584135, "grad_norm": 0.47313642501831055, "learning_rate": 0.00033143489030806086, "loss": 2.4201, "step": 1355 }, { "epoch": 0.6217331499312242, "grad_norm": 0.44106101989746094, "learning_rate": 0.00033073586195681227, "loss": 2.3913, "step": 1356 }, { "epoch": 0.6221916552040349, "grad_norm": 0.47318825125694275, "learning_rate": 0.00033003720706537736, "loss": 2.2988, "step": 1357 }, { "epoch": 0.6226501604768455, "grad_norm": 0.47347259521484375, "learning_rate": 0.0003293389271752492, "loss": 2.3917, "step": 1358 }, { "epoch": 0.6231086657496562, "grad_norm": 0.47882598638534546, "learning_rate": 0.00032864102382709374, "loss": 2.5046, "step": 1359 }, { "epoch": 0.6235671710224667, "grad_norm": 0.46698710322380066, "learning_rate": 0.000327943498560746, "loss": 2.4145, "step": 1360 }, { "epoch": 0.6240256762952774, "grad_norm": 0.4433773159980774, "learning_rate": 0.00032724635291520694, "loss": 2.4187, "step": 1361 }, { "epoch": 0.624484181568088, "grad_norm": 0.4827621877193451, "learning_rate": 0.00032654958842863967, "loss": 2.4101, "step": 1362 }, { "epoch": 0.6249426868408987, "grad_norm": 0.45228174328804016, "learning_rate": 0.0003258532066383667, "loss": 2.2961, "step": 1363 }, { "epoch": 0.6254011921137093, "grad_norm": 0.4570356011390686, "learning_rate": 0.000325157209080866, "loss": 2.3502, "step": 1364 }, { "epoch": 0.6258596973865199, "grad_norm": 0.45143771171569824, "learning_rate": 0.00032446159729176743, "loss": 2.4537, "step": 1365 }, { "epoch": 0.6263182026593306, "grad_norm": 0.48375990986824036, "learning_rate": 0.0003237663728058502, "loss": 2.321, "step": 1366 }, { "epoch": 0.6267767079321412, "grad_norm": 0.47481873631477356, "learning_rate": 0.0003230715371570389, "loss": 2.5128, "step": 1367 }, { "epoch": 0.6272352132049519, "grad_norm": 0.445255845785141, "learning_rate": 0.00032237709187839996, "loss": 2.3367, "step": 1368 }, { "epoch": 0.6276937184777625, "grad_norm": 0.4269105792045593, "learning_rate": 0.0003216830385021388, "loss": 2.3214, "step": 1369 }, { "epoch": 0.6281522237505731, "grad_norm": 0.4767317473888397, "learning_rate": 0.0003209893785595959, "loss": 2.5377, "step": 1370 }, { "epoch": 0.6286107290233838, "grad_norm": 0.45423510670661926, "learning_rate": 0.00032029611358124366, "loss": 2.3513, "step": 1371 }, { "epoch": 0.6290692342961944, "grad_norm": 0.48342275619506836, "learning_rate": 0.00031960324509668336, "loss": 2.3708, "step": 1372 }, { "epoch": 0.6295277395690051, "grad_norm": 0.4805603325366974, "learning_rate": 0.0003189107746346412, "loss": 2.4425, "step": 1373 }, { "epoch": 0.6299862448418156, "grad_norm": 0.45650094747543335, "learning_rate": 0.0003182187037229653, "loss": 2.3927, "step": 1374 }, { "epoch": 0.6304447501146263, "grad_norm": 0.4815146327018738, "learning_rate": 0.0003175270338886221, "loss": 2.5213, "step": 1375 }, { "epoch": 0.630903255387437, "grad_norm": 0.4830882251262665, "learning_rate": 0.00031683576665769345, "loss": 2.434, "step": 1376 }, { "epoch": 0.6313617606602476, "grad_norm": 0.4881056845188141, "learning_rate": 0.0003161449035553724, "loss": 2.3159, "step": 1377 }, { "epoch": 0.6318202659330582, "grad_norm": 0.4908444583415985, "learning_rate": 0.00031545444610596077, "loss": 2.452, "step": 1378 }, { "epoch": 0.6322787712058688, "grad_norm": 0.46427658200263977, "learning_rate": 0.000314764395832865, "loss": 2.4214, "step": 1379 }, { "epoch": 0.6327372764786795, "grad_norm": 0.4554003179073334, "learning_rate": 0.0003140747542585934, "loss": 2.365, "step": 1380 }, { "epoch": 0.6331957817514902, "grad_norm": 0.47302570939064026, "learning_rate": 0.00031338552290475266, "loss": 2.41, "step": 1381 }, { "epoch": 0.6336542870243008, "grad_norm": 0.4516913592815399, "learning_rate": 0.00031269670329204396, "loss": 2.3468, "step": 1382 }, { "epoch": 0.6341127922971114, "grad_norm": 0.4696624279022217, "learning_rate": 0.0003120082969402604, "loss": 2.4567, "step": 1383 }, { "epoch": 0.634571297569922, "grad_norm": 0.4718548655509949, "learning_rate": 0.00031132030536828314, "loss": 2.522, "step": 1384 }, { "epoch": 0.6350298028427327, "grad_norm": 0.471697598695755, "learning_rate": 0.00031063273009407805, "loss": 2.4033, "step": 1385 }, { "epoch": 0.6354883081155434, "grad_norm": 0.4939471483230591, "learning_rate": 0.00030994557263469265, "loss": 2.3848, "step": 1386 }, { "epoch": 0.635946813388354, "grad_norm": 0.445483535528183, "learning_rate": 0.0003092588345062526, "loss": 2.3644, "step": 1387 }, { "epoch": 0.6364053186611646, "grad_norm": 0.49615082144737244, "learning_rate": 0.0003085725172239582, "loss": 2.4974, "step": 1388 }, { "epoch": 0.6368638239339752, "grad_norm": 0.4554246664047241, "learning_rate": 0.0003078866223020815, "loss": 2.406, "step": 1389 }, { "epoch": 0.6373223292067859, "grad_norm": 0.4667593836784363, "learning_rate": 0.0003072011512539624, "loss": 2.3862, "step": 1390 }, { "epoch": 0.6377808344795965, "grad_norm": 0.458993136882782, "learning_rate": 0.00030651610559200574, "loss": 2.4849, "step": 1391 }, { "epoch": 0.6382393397524071, "grad_norm": 0.47354239225387573, "learning_rate": 0.00030583148682767757, "loss": 2.3941, "step": 1392 }, { "epoch": 0.6386978450252178, "grad_norm": 0.475159227848053, "learning_rate": 0.00030514729647150243, "loss": 2.3998, "step": 1393 }, { "epoch": 0.6391563502980284, "grad_norm": 0.44674867391586304, "learning_rate": 0.0003044635360330592, "loss": 2.3715, "step": 1394 }, { "epoch": 0.6396148555708391, "grad_norm": 0.4451742172241211, "learning_rate": 0.00030378020702097845, "loss": 2.4173, "step": 1395 }, { "epoch": 0.6400733608436497, "grad_norm": 0.49074965715408325, "learning_rate": 0.000303097310942939, "loss": 2.5232, "step": 1396 }, { "epoch": 0.6405318661164603, "grad_norm": 0.45490139722824097, "learning_rate": 0.0003024148493056641, "loss": 2.3977, "step": 1397 }, { "epoch": 0.640990371389271, "grad_norm": 0.48554107546806335, "learning_rate": 0.00030173282361491865, "loss": 2.3743, "step": 1398 }, { "epoch": 0.6414488766620816, "grad_norm": 0.4747178554534912, "learning_rate": 0.0003010512353755057, "loss": 2.3579, "step": 1399 }, { "epoch": 0.6419073819348923, "grad_norm": 0.4759728014469147, "learning_rate": 0.00030037008609126313, "loss": 2.4067, "step": 1400 }, { "epoch": 0.6423658872077029, "grad_norm": 0.4702655076980591, "learning_rate": 0.0002996893772650602, "loss": 2.3684, "step": 1401 }, { "epoch": 0.6428243924805135, "grad_norm": 0.4432293772697449, "learning_rate": 0.0002990091103987945, "loss": 2.4882, "step": 1402 }, { "epoch": 0.6432828977533241, "grad_norm": 0.4768392741680145, "learning_rate": 0.0002983292869933886, "loss": 2.4868, "step": 1403 }, { "epoch": 0.6437414030261348, "grad_norm": 0.4883682131767273, "learning_rate": 0.0002976499085487862, "loss": 2.428, "step": 1404 }, { "epoch": 0.6441999082989455, "grad_norm": 0.45367422699928284, "learning_rate": 0.00029697097656394963, "loss": 2.406, "step": 1405 }, { "epoch": 0.6446584135717561, "grad_norm": 0.4636426270008087, "learning_rate": 0.00029629249253685595, "loss": 2.3591, "step": 1406 }, { "epoch": 0.6451169188445667, "grad_norm": 0.4620175361633301, "learning_rate": 0.00029561445796449416, "loss": 2.4865, "step": 1407 }, { "epoch": 0.6455754241173773, "grad_norm": 0.46177515387535095, "learning_rate": 0.0002949368743428612, "loss": 2.3548, "step": 1408 }, { "epoch": 0.646033929390188, "grad_norm": 0.44374755024909973, "learning_rate": 0.0002942597431669593, "loss": 2.3764, "step": 1409 }, { "epoch": 0.6464924346629987, "grad_norm": 0.4824047088623047, "learning_rate": 0.0002935830659307924, "loss": 2.4188, "step": 1410 }, { "epoch": 0.6469509399358092, "grad_norm": 0.48756924271583557, "learning_rate": 0.0002929068441273629, "loss": 2.4239, "step": 1411 }, { "epoch": 0.6474094452086199, "grad_norm": 0.48978307843208313, "learning_rate": 0.0002922310792486681, "loss": 2.397, "step": 1412 }, { "epoch": 0.6478679504814305, "grad_norm": 0.4528706967830658, "learning_rate": 0.00029155577278569745, "loss": 2.3568, "step": 1413 }, { "epoch": 0.6483264557542412, "grad_norm": 0.48085030913352966, "learning_rate": 0.00029088092622842895, "loss": 2.4586, "step": 1414 }, { "epoch": 0.6487849610270519, "grad_norm": 0.44443458318710327, "learning_rate": 0.00029020654106582544, "loss": 2.4626, "step": 1415 }, { "epoch": 0.6492434662998624, "grad_norm": 0.46039775013923645, "learning_rate": 0.0002895326187858326, "loss": 2.3736, "step": 1416 }, { "epoch": 0.6497019715726731, "grad_norm": 0.48584696650505066, "learning_rate": 0.00028885916087537377, "loss": 2.4325, "step": 1417 }, { "epoch": 0.6501604768454837, "grad_norm": 0.47679242491722107, "learning_rate": 0.00028818616882034877, "loss": 2.4603, "step": 1418 }, { "epoch": 0.6506189821182944, "grad_norm": 0.46679049730300903, "learning_rate": 0.0002875136441056286, "loss": 2.3147, "step": 1419 }, { "epoch": 0.651077487391105, "grad_norm": 0.4910959303379059, "learning_rate": 0.000286841588215054, "loss": 2.4787, "step": 1420 }, { "epoch": 0.6515359926639156, "grad_norm": 0.48436638712882996, "learning_rate": 0.0002861700026314308, "loss": 2.4409, "step": 1421 }, { "epoch": 0.6519944979367263, "grad_norm": 0.4860231876373291, "learning_rate": 0.00028549888883652686, "loss": 2.3664, "step": 1422 }, { "epoch": 0.6524530032095369, "grad_norm": 0.46070000529289246, "learning_rate": 0.00028482824831107, "loss": 2.4536, "step": 1423 }, { "epoch": 0.6529115084823476, "grad_norm": 0.49830228090286255, "learning_rate": 0.000284158082534743, "loss": 2.428, "step": 1424 }, { "epoch": 0.6533700137551581, "grad_norm": 0.4679667353630066, "learning_rate": 0.00028348839298618177, "loss": 2.355, "step": 1425 }, { "epoch": 0.6538285190279688, "grad_norm": 0.46012330055236816, "learning_rate": 0.0002828191811429709, "loss": 2.3948, "step": 1426 }, { "epoch": 0.6542870243007795, "grad_norm": 0.48015737533569336, "learning_rate": 0.00028215044848164164, "loss": 2.4344, "step": 1427 }, { "epoch": 0.6547455295735901, "grad_norm": 0.5161628723144531, "learning_rate": 0.00028148219647766747, "loss": 2.5749, "step": 1428 }, { "epoch": 0.6552040348464008, "grad_norm": 0.4648689329624176, "learning_rate": 0.00028081442660546124, "loss": 2.3356, "step": 1429 }, { "epoch": 0.6556625401192113, "grad_norm": 0.4703148305416107, "learning_rate": 0.0002801471403383728, "loss": 2.3477, "step": 1430 }, { "epoch": 0.656121045392022, "grad_norm": 0.44403961300849915, "learning_rate": 0.00027948033914868415, "loss": 2.3788, "step": 1431 }, { "epoch": 0.6565795506648326, "grad_norm": 0.4660142660140991, "learning_rate": 0.00027881402450760775, "loss": 2.4316, "step": 1432 }, { "epoch": 0.6570380559376433, "grad_norm": 0.4710667133331299, "learning_rate": 0.00027814819788528165, "loss": 2.4574, "step": 1433 }, { "epoch": 0.657496561210454, "grad_norm": 0.4691605567932129, "learning_rate": 0.00027748286075076836, "loss": 2.3598, "step": 1434 }, { "epoch": 0.6579550664832645, "grad_norm": 0.4762023985385895, "learning_rate": 0.00027681801457204937, "loss": 2.36, "step": 1435 }, { "epoch": 0.6584135717560752, "grad_norm": 0.4633997678756714, "learning_rate": 0.00027615366081602306, "loss": 2.5081, "step": 1436 }, { "epoch": 0.6588720770288858, "grad_norm": 0.4507983922958374, "learning_rate": 0.0002754898009485021, "loss": 2.3207, "step": 1437 }, { "epoch": 0.6593305823016965, "grad_norm": 0.4341924786567688, "learning_rate": 0.0002748264364342085, "loss": 2.3463, "step": 1438 }, { "epoch": 0.6597890875745072, "grad_norm": 0.47592583298683167, "learning_rate": 0.00027416356873677204, "loss": 2.3281, "step": 1439 }, { "epoch": 0.6602475928473177, "grad_norm": 0.459052711725235, "learning_rate": 0.0002735011993187258, "loss": 2.3632, "step": 1440 }, { "epoch": 0.6607060981201284, "grad_norm": 0.4733355641365051, "learning_rate": 0.0002728393296415042, "loss": 2.4833, "step": 1441 }, { "epoch": 0.661164603392939, "grad_norm": 0.4617414176464081, "learning_rate": 0.00027217796116543817, "loss": 2.4222, "step": 1442 }, { "epoch": 0.6616231086657497, "grad_norm": 0.47972747683525085, "learning_rate": 0.0002715170953497532, "loss": 2.3269, "step": 1443 }, { "epoch": 0.6620816139385602, "grad_norm": 0.47127601504325867, "learning_rate": 0.00027085673365256614, "loss": 2.3392, "step": 1444 }, { "epoch": 0.6625401192113709, "grad_norm": 0.44934481382369995, "learning_rate": 0.00027019687753088075, "loss": 2.4611, "step": 1445 }, { "epoch": 0.6629986244841816, "grad_norm": 0.46069222688674927, "learning_rate": 0.00026953752844058597, "loss": 2.4089, "step": 1446 }, { "epoch": 0.6634571297569922, "grad_norm": 0.46655499935150146, "learning_rate": 0.0002688786878364516, "loss": 2.4463, "step": 1447 }, { "epoch": 0.6639156350298029, "grad_norm": 0.4818456470966339, "learning_rate": 0.00026822035717212597, "loss": 2.4921, "step": 1448 }, { "epoch": 0.6643741403026134, "grad_norm": 0.4498130977153778, "learning_rate": 0.00026756253790013193, "loss": 2.3087, "step": 1449 }, { "epoch": 0.6648326455754241, "grad_norm": 0.46364718675613403, "learning_rate": 0.0002669052314718641, "loss": 2.349, "step": 1450 }, { "epoch": 0.6652911508482348, "grad_norm": 0.4375849962234497, "learning_rate": 0.0002662484393375855, "loss": 2.3132, "step": 1451 }, { "epoch": 0.6657496561210454, "grad_norm": 0.4975627064704895, "learning_rate": 0.00026559216294642446, "loss": 2.3818, "step": 1452 }, { "epoch": 0.6662081613938561, "grad_norm": 0.45664745569229126, "learning_rate": 0.0002649364037463718, "loss": 2.4251, "step": 1453 }, { "epoch": 0.6666666666666666, "grad_norm": 0.4572896361351013, "learning_rate": 0.0002642811631842764, "loss": 2.3915, "step": 1454 }, { "epoch": 0.6671251719394773, "grad_norm": 0.4568662643432617, "learning_rate": 0.0002636264427058439, "loss": 2.346, "step": 1455 }, { "epoch": 0.667583677212288, "grad_norm": 0.4677676856517792, "learning_rate": 0.00026297224375563123, "loss": 2.4018, "step": 1456 }, { "epoch": 0.6680421824850986, "grad_norm": 0.48621970415115356, "learning_rate": 0.00026231856777704575, "loss": 2.4287, "step": 1457 }, { "epoch": 0.6685006877579092, "grad_norm": 0.4599752426147461, "learning_rate": 0.00026166541621234026, "loss": 2.3217, "step": 1458 }, { "epoch": 0.6689591930307198, "grad_norm": 0.4585476517677307, "learning_rate": 0.00026101279050261045, "loss": 2.2719, "step": 1459 }, { "epoch": 0.6694176983035305, "grad_norm": 0.4880422055721283, "learning_rate": 0.00026036069208779247, "loss": 2.4268, "step": 1460 }, { "epoch": 0.6698762035763411, "grad_norm": 0.48888155817985535, "learning_rate": 0.0002597091224066581, "loss": 2.3795, "step": 1461 }, { "epoch": 0.6703347088491518, "grad_norm": 0.46095946431159973, "learning_rate": 0.00025905808289681365, "loss": 2.3573, "step": 1462 }, { "epoch": 0.6707932141219624, "grad_norm": 0.48185041546821594, "learning_rate": 0.0002584075749946946, "loss": 2.407, "step": 1463 }, { "epoch": 0.671251719394773, "grad_norm": 0.4735257625579834, "learning_rate": 0.00025775760013556424, "loss": 2.416, "step": 1464 }, { "epoch": 0.6717102246675837, "grad_norm": 0.48655927181243896, "learning_rate": 0.0002571081597535095, "loss": 2.4579, "step": 1465 }, { "epoch": 0.6721687299403943, "grad_norm": 0.49890032410621643, "learning_rate": 0.00025645925528143776, "loss": 2.4697, "step": 1466 }, { "epoch": 0.672627235213205, "grad_norm": 0.5235516428947449, "learning_rate": 0.0002558108881510747, "loss": 2.5134, "step": 1467 }, { "epoch": 0.6730857404860155, "grad_norm": 0.4801587462425232, "learning_rate": 0.00025516305979295963, "loss": 2.3311, "step": 1468 }, { "epoch": 0.6735442457588262, "grad_norm": 0.48125141859054565, "learning_rate": 0.0002545157716364439, "loss": 2.4166, "step": 1469 }, { "epoch": 0.6740027510316369, "grad_norm": 0.4776366949081421, "learning_rate": 0.00025386902510968624, "loss": 2.3184, "step": 1470 }, { "epoch": 0.6744612563044475, "grad_norm": 0.449143648147583, "learning_rate": 0.00025322282163965095, "loss": 2.4595, "step": 1471 }, { "epoch": 0.6749197615772582, "grad_norm": 0.4788149297237396, "learning_rate": 0.00025257716265210384, "loss": 2.3687, "step": 1472 }, { "epoch": 0.6753782668500687, "grad_norm": 0.48083752393722534, "learning_rate": 0.0002519320495716091, "loss": 2.3298, "step": 1473 }, { "epoch": 0.6758367721228794, "grad_norm": 0.5280911326408386, "learning_rate": 0.00025128748382152716, "loss": 2.5122, "step": 1474 }, { "epoch": 0.6762952773956901, "grad_norm": 0.4556063115596771, "learning_rate": 0.00025064346682401016, "loss": 2.3713, "step": 1475 }, { "epoch": 0.6767537826685007, "grad_norm": 0.5112197399139404, "learning_rate": 0.0002500000000000001, "loss": 2.3813, "step": 1476 }, { "epoch": 0.6772122879413113, "grad_norm": 0.4736495316028595, "learning_rate": 0.0002493570847692246, "loss": 2.3712, "step": 1477 }, { "epoch": 0.6776707932141219, "grad_norm": 0.46050602197647095, "learning_rate": 0.00024871472255019424, "loss": 2.3367, "step": 1478 }, { "epoch": 0.6781292984869326, "grad_norm": 0.4678228795528412, "learning_rate": 0.00024807291476019994, "loss": 2.4062, "step": 1479 }, { "epoch": 0.6785878037597433, "grad_norm": 0.4708285331726074, "learning_rate": 0.00024743166281530877, "loss": 2.3553, "step": 1480 }, { "epoch": 0.6790463090325539, "grad_norm": 0.45587509870529175, "learning_rate": 0.000246790968130362, "loss": 2.4443, "step": 1481 }, { "epoch": 0.6795048143053645, "grad_norm": 0.4796959161758423, "learning_rate": 0.0002461508321189706, "loss": 2.3662, "step": 1482 }, { "epoch": 0.6799633195781751, "grad_norm": 0.4432728588581085, "learning_rate": 0.00024551125619351385, "loss": 2.374, "step": 1483 }, { "epoch": 0.6804218248509858, "grad_norm": 0.48211321234703064, "learning_rate": 0.00024487224176513453, "loss": 2.3756, "step": 1484 }, { "epoch": 0.6808803301237965, "grad_norm": 0.46660488843917847, "learning_rate": 0.0002442337902437365, "loss": 2.4781, "step": 1485 }, { "epoch": 0.6813388353966071, "grad_norm": 0.44977471232414246, "learning_rate": 0.0002435959030379824, "loss": 2.3293, "step": 1486 }, { "epoch": 0.6817973406694177, "grad_norm": 0.4333764612674713, "learning_rate": 0.00024295858155528888, "loss": 2.3558, "step": 1487 }, { "epoch": 0.6822558459422283, "grad_norm": 0.4899200201034546, "learning_rate": 0.00024232182720182523, "loss": 2.4432, "step": 1488 }, { "epoch": 0.682714351215039, "grad_norm": 0.479081928730011, "learning_rate": 0.00024168564138250855, "loss": 2.4118, "step": 1489 }, { "epoch": 0.6831728564878496, "grad_norm": 0.46026358008384705, "learning_rate": 0.00024105002550100246, "loss": 2.4613, "step": 1490 }, { "epoch": 0.6836313617606602, "grad_norm": 0.4476301074028015, "learning_rate": 0.00024041498095971254, "loss": 2.404, "step": 1491 }, { "epoch": 0.6840898670334709, "grad_norm": 0.4779755771160126, "learning_rate": 0.0002397805091597835, "loss": 2.4687, "step": 1492 }, { "epoch": 0.6845483723062815, "grad_norm": 0.49555936455726624, "learning_rate": 0.0002391466115010973, "loss": 2.4268, "step": 1493 }, { "epoch": 0.6850068775790922, "grad_norm": 0.46483707427978516, "learning_rate": 0.00023851328938226808, "loss": 2.439, "step": 1494 }, { "epoch": 0.6854653828519028, "grad_norm": 0.4958858788013458, "learning_rate": 0.00023788054420064109, "loss": 2.3656, "step": 1495 }, { "epoch": 0.6859238881247134, "grad_norm": 0.47024157643318176, "learning_rate": 0.00023724837735228773, "loss": 2.3589, "step": 1496 }, { "epoch": 0.686382393397524, "grad_norm": 0.5271995663642883, "learning_rate": 0.00023661679023200422, "loss": 2.4388, "step": 1497 }, { "epoch": 0.6868408986703347, "grad_norm": 0.4739536643028259, "learning_rate": 0.00023598578423330714, "loss": 2.3675, "step": 1498 }, { "epoch": 0.6872994039431454, "grad_norm": 0.4843970835208893, "learning_rate": 0.00023535536074843083, "loss": 2.3619, "step": 1499 }, { "epoch": 0.687757909215956, "grad_norm": 0.4583735167980194, "learning_rate": 0.00023472552116832502, "loss": 2.4384, "step": 1500 }, { "epoch": 0.6882164144887666, "grad_norm": 0.4560268819332123, "learning_rate": 0.0002340962668826503, "loss": 2.4342, "step": 1501 }, { "epoch": 0.6886749197615772, "grad_norm": 0.45700451731681824, "learning_rate": 0.00023346759927977663, "loss": 2.4638, "step": 1502 }, { "epoch": 0.6891334250343879, "grad_norm": 0.468904972076416, "learning_rate": 0.0002328395197467789, "loss": 2.4423, "step": 1503 }, { "epoch": 0.6895919303071986, "grad_norm": 0.4594002068042755, "learning_rate": 0.00023221202966943515, "loss": 2.4442, "step": 1504 }, { "epoch": 0.6900504355800092, "grad_norm": 0.4791922867298126, "learning_rate": 0.0002315851304322223, "loss": 2.4169, "step": 1505 }, { "epoch": 0.6905089408528198, "grad_norm": 0.4642763137817383, "learning_rate": 0.0002309588234183137, "loss": 2.4691, "step": 1506 }, { "epoch": 0.6909674461256304, "grad_norm": 0.4908992350101471, "learning_rate": 0.00023033311000957653, "loss": 2.3584, "step": 1507 }, { "epoch": 0.6914259513984411, "grad_norm": 0.4913836419582367, "learning_rate": 0.00022970799158656758, "loss": 2.4394, "step": 1508 }, { "epoch": 0.6918844566712518, "grad_norm": 0.4789724051952362, "learning_rate": 0.0002290834695285316, "loss": 2.3685, "step": 1509 }, { "epoch": 0.6923429619440623, "grad_norm": 0.5153934955596924, "learning_rate": 0.00022845954521339678, "loss": 2.5064, "step": 1510 }, { "epoch": 0.692801467216873, "grad_norm": 0.4637894630432129, "learning_rate": 0.0002278362200177732, "loss": 2.4078, "step": 1511 }, { "epoch": 0.6932599724896836, "grad_norm": 0.48305970430374146, "learning_rate": 0.00022721349531694852, "loss": 2.3608, "step": 1512 }, { "epoch": 0.6937184777624943, "grad_norm": 0.46188053488731384, "learning_rate": 0.0002265913724848855, "loss": 2.3841, "step": 1513 }, { "epoch": 0.694176983035305, "grad_norm": 0.5064974427223206, "learning_rate": 0.00022596985289421946, "loss": 2.3878, "step": 1514 }, { "epoch": 0.6946354883081155, "grad_norm": 0.4825488030910492, "learning_rate": 0.00022534893791625405, "loss": 2.3757, "step": 1515 }, { "epoch": 0.6950939935809262, "grad_norm": 0.4669504463672638, "learning_rate": 0.00022472862892095968, "loss": 2.4082, "step": 1516 }, { "epoch": 0.6955524988537368, "grad_norm": 0.5110418796539307, "learning_rate": 0.00022410892727696896, "loss": 2.5491, "step": 1517 }, { "epoch": 0.6960110041265475, "grad_norm": 0.45133909583091736, "learning_rate": 0.0002234898343515751, "loss": 2.4145, "step": 1518 }, { "epoch": 0.6964695093993581, "grad_norm": 0.46629998087882996, "learning_rate": 0.00022287135151072792, "loss": 2.3841, "step": 1519 }, { "epoch": 0.6969280146721687, "grad_norm": 0.4554193615913391, "learning_rate": 0.00022225348011903096, "loss": 2.404, "step": 1520 }, { "epoch": 0.6973865199449794, "grad_norm": 0.4570068120956421, "learning_rate": 0.0002216362215397393, "loss": 2.3504, "step": 1521 }, { "epoch": 0.69784502521779, "grad_norm": 0.45473039150238037, "learning_rate": 0.00022101957713475522, "loss": 2.3801, "step": 1522 }, { "epoch": 0.6983035304906007, "grad_norm": 0.4494574964046478, "learning_rate": 0.00022040354826462666, "loss": 2.3847, "step": 1523 }, { "epoch": 0.6987620357634112, "grad_norm": 0.4449474513530731, "learning_rate": 0.0002197881362885426, "loss": 2.3773, "step": 1524 }, { "epoch": 0.6992205410362219, "grad_norm": 0.47448229789733887, "learning_rate": 0.0002191733425643318, "loss": 2.2861, "step": 1525 }, { "epoch": 0.6996790463090325, "grad_norm": 0.4743362069129944, "learning_rate": 0.00021855916844845826, "loss": 2.3492, "step": 1526 }, { "epoch": 0.7001375515818432, "grad_norm": 0.45902037620544434, "learning_rate": 0.00021794561529601898, "loss": 2.3512, "step": 1527 }, { "epoch": 0.7005960568546539, "grad_norm": 0.4840593934059143, "learning_rate": 0.00021733268446074138, "loss": 2.3464, "step": 1528 }, { "epoch": 0.7010545621274644, "grad_norm": 0.4833325147628784, "learning_rate": 0.00021672037729497917, "loss": 2.3537, "step": 1529 }, { "epoch": 0.7015130674002751, "grad_norm": 0.4720504879951477, "learning_rate": 0.0002161086951497106, "loss": 2.4958, "step": 1530 }, { "epoch": 0.7019715726730857, "grad_norm": 0.4806617796421051, "learning_rate": 0.00021549763937453442, "loss": 2.5034, "step": 1531 }, { "epoch": 0.7024300779458964, "grad_norm": 0.45898449420928955, "learning_rate": 0.00021488721131766736, "loss": 2.4468, "step": 1532 }, { "epoch": 0.7028885832187071, "grad_norm": 0.45827844738960266, "learning_rate": 0.00021427741232594183, "loss": 2.3657, "step": 1533 }, { "epoch": 0.7033470884915176, "grad_norm": 0.4752943813800812, "learning_rate": 0.0002136682437448013, "loss": 2.4339, "step": 1534 }, { "epoch": 0.7038055937643283, "grad_norm": 0.5323183536529541, "learning_rate": 0.0002130597069182994, "loss": 2.3287, "step": 1535 }, { "epoch": 0.7042640990371389, "grad_norm": 0.4645003378391266, "learning_rate": 0.0002124518031890948, "loss": 2.3881, "step": 1536 }, { "epoch": 0.7047226043099496, "grad_norm": 0.4567202925682068, "learning_rate": 0.0002118445338984502, "loss": 2.4037, "step": 1537 }, { "epoch": 0.7051811095827603, "grad_norm": 0.5198760032653809, "learning_rate": 0.00021123790038622808, "loss": 2.4432, "step": 1538 }, { "epoch": 0.7056396148555708, "grad_norm": 0.522384762763977, "learning_rate": 0.0002106319039908879, "loss": 2.5214, "step": 1539 }, { "epoch": 0.7060981201283815, "grad_norm": 0.4659295976161957, "learning_rate": 0.00021002654604948412, "loss": 2.4033, "step": 1540 }, { "epoch": 0.7065566254011921, "grad_norm": 0.481618195772171, "learning_rate": 0.00020942182789766172, "loss": 2.4042, "step": 1541 }, { "epoch": 0.7070151306740028, "grad_norm": 0.45138195157051086, "learning_rate": 0.00020881775086965492, "loss": 2.4026, "step": 1542 }, { "epoch": 0.7074736359468133, "grad_norm": 0.4814971685409546, "learning_rate": 0.00020821431629828246, "loss": 2.4038, "step": 1543 }, { "epoch": 0.707932141219624, "grad_norm": 0.4779079556465149, "learning_rate": 0.00020761152551494643, "loss": 2.3733, "step": 1544 }, { "epoch": 0.7083906464924347, "grad_norm": 0.4480411410331726, "learning_rate": 0.00020700937984962798, "loss": 2.3333, "step": 1545 }, { "epoch": 0.7088491517652453, "grad_norm": 0.4749911427497864, "learning_rate": 0.0002064078806308848, "loss": 2.4901, "step": 1546 }, { "epoch": 0.709307657038056, "grad_norm": 0.45038041472435, "learning_rate": 0.00020580702918584882, "loss": 2.4898, "step": 1547 }, { "epoch": 0.7097661623108665, "grad_norm": 0.4517367482185364, "learning_rate": 0.000205206826840222, "loss": 2.2482, "step": 1548 }, { "epoch": 0.7102246675836772, "grad_norm": 0.47881585359573364, "learning_rate": 0.0002046072749182751, "loss": 2.4657, "step": 1549 }, { "epoch": 0.7106831728564879, "grad_norm": 0.46140679717063904, "learning_rate": 0.00020400837474284273, "loss": 2.4062, "step": 1550 }, { "epoch": 0.7111416781292985, "grad_norm": 0.4819934666156769, "learning_rate": 0.0002034101276353224, "loss": 2.4085, "step": 1551 }, { "epoch": 0.7116001834021092, "grad_norm": 0.45329567790031433, "learning_rate": 0.00020281253491567027, "loss": 2.4324, "step": 1552 }, { "epoch": 0.7120586886749197, "grad_norm": 0.45584434270858765, "learning_rate": 0.0002022155979023984, "loss": 2.3598, "step": 1553 }, { "epoch": 0.7125171939477304, "grad_norm": 0.46209490299224854, "learning_rate": 0.000201619317912573, "loss": 2.4088, "step": 1554 }, { "epoch": 0.712975699220541, "grad_norm": 0.47079604864120483, "learning_rate": 0.00020102369626180962, "loss": 2.3555, "step": 1555 }, { "epoch": 0.7134342044933517, "grad_norm": 0.48297640681266785, "learning_rate": 0.0002004287342642721, "loss": 2.3803, "step": 1556 }, { "epoch": 0.7138927097661623, "grad_norm": 0.4741036891937256, "learning_rate": 0.00019983443323266824, "loss": 2.5149, "step": 1557 }, { "epoch": 0.7143512150389729, "grad_norm": 0.47223299741744995, "learning_rate": 0.00019924079447824805, "loss": 2.3884, "step": 1558 }, { "epoch": 0.7148097203117836, "grad_norm": 0.4834589660167694, "learning_rate": 0.00019864781931079977, "loss": 2.3045, "step": 1559 }, { "epoch": 0.7152682255845942, "grad_norm": 0.4536362290382385, "learning_rate": 0.00019805550903864773, "loss": 2.394, "step": 1560 }, { "epoch": 0.7157267308574049, "grad_norm": 0.4755149185657501, "learning_rate": 0.00019746386496864948, "loss": 2.3724, "step": 1561 }, { "epoch": 0.7161852361302155, "grad_norm": 0.478840708732605, "learning_rate": 0.00019687288840619226, "loss": 2.3675, "step": 1562 }, { "epoch": 0.7166437414030261, "grad_norm": 0.48134663701057434, "learning_rate": 0.0001962825806551911, "loss": 2.3124, "step": 1563 }, { "epoch": 0.7171022466758368, "grad_norm": 0.4729054272174835, "learning_rate": 0.0001956929430180846, "loss": 2.3217, "step": 1564 }, { "epoch": 0.7175607519486474, "grad_norm": 0.471104234457016, "learning_rate": 0.00019510397679583374, "loss": 2.3988, "step": 1565 }, { "epoch": 0.7180192572214581, "grad_norm": 0.4671911895275116, "learning_rate": 0.0001945156832879174, "loss": 2.3104, "step": 1566 }, { "epoch": 0.7184777624942686, "grad_norm": 0.4741764962673187, "learning_rate": 0.00019392806379233036, "loss": 2.3675, "step": 1567 }, { "epoch": 0.7189362677670793, "grad_norm": 0.46789681911468506, "learning_rate": 0.00019334111960558065, "loss": 2.2808, "step": 1568 }, { "epoch": 0.71939477303989, "grad_norm": 0.48799723386764526, "learning_rate": 0.00019275485202268573, "loss": 2.4905, "step": 1569 }, { "epoch": 0.7198532783127006, "grad_norm": 0.48220503330230713, "learning_rate": 0.00019216926233717085, "loss": 2.2922, "step": 1570 }, { "epoch": 0.7203117835855112, "grad_norm": 0.45209670066833496, "learning_rate": 0.00019158435184106498, "loss": 2.4137, "step": 1571 }, { "epoch": 0.7207702888583218, "grad_norm": 0.4845699369907379, "learning_rate": 0.00019100012182489905, "loss": 2.3892, "step": 1572 }, { "epoch": 0.7212287941311325, "grad_norm": 0.4703586995601654, "learning_rate": 0.00019041657357770226, "loss": 2.4073, "step": 1573 }, { "epoch": 0.7216872994039432, "grad_norm": 0.4493473172187805, "learning_rate": 0.00018983370838699943, "loss": 2.3792, "step": 1574 }, { "epoch": 0.7221458046767538, "grad_norm": 0.45864227414131165, "learning_rate": 0.00018925152753880892, "loss": 2.3027, "step": 1575 }, { "epoch": 0.7226043099495644, "grad_norm": 0.44898003339767456, "learning_rate": 0.00018867003231763847, "loss": 2.4348, "step": 1576 }, { "epoch": 0.723062815222375, "grad_norm": 0.47078225016593933, "learning_rate": 0.00018808922400648375, "loss": 2.3739, "step": 1577 }, { "epoch": 0.7235213204951857, "grad_norm": 0.45881637930870056, "learning_rate": 0.00018750910388682428, "loss": 2.3915, "step": 1578 }, { "epoch": 0.7239798257679964, "grad_norm": 0.46436411142349243, "learning_rate": 0.00018692967323862125, "loss": 2.3411, "step": 1579 }, { "epoch": 0.724438331040807, "grad_norm": 0.4778837561607361, "learning_rate": 0.00018635093334031517, "loss": 2.3587, "step": 1580 }, { "epoch": 0.7248968363136176, "grad_norm": 0.46819111704826355, "learning_rate": 0.00018577288546882165, "loss": 2.3155, "step": 1581 }, { "epoch": 0.7253553415864282, "grad_norm": 0.4855227470397949, "learning_rate": 0.00018519553089953023, "loss": 2.3881, "step": 1582 }, { "epoch": 0.7258138468592389, "grad_norm": 0.47348669171333313, "learning_rate": 0.0001846188709063001, "loss": 2.4681, "step": 1583 }, { "epoch": 0.7262723521320495, "grad_norm": 0.49490469694137573, "learning_rate": 0.00018404290676145857, "loss": 2.4758, "step": 1584 }, { "epoch": 0.7267308574048602, "grad_norm": 0.46061837673187256, "learning_rate": 0.00018346763973579722, "loss": 2.3561, "step": 1585 }, { "epoch": 0.7271893626776708, "grad_norm": 0.5045784711837769, "learning_rate": 0.00018289307109856939, "loss": 2.325, "step": 1586 }, { "epoch": 0.7276478679504814, "grad_norm": 0.4891712963581085, "learning_rate": 0.0001823192021174882, "loss": 2.4275, "step": 1587 }, { "epoch": 0.7281063732232921, "grad_norm": 0.4761686623096466, "learning_rate": 0.0001817460340587223, "loss": 2.4068, "step": 1588 }, { "epoch": 0.7285648784961027, "grad_norm": 0.4657875895500183, "learning_rate": 0.00018117356818689445, "loss": 2.2808, "step": 1589 }, { "epoch": 0.7290233837689133, "grad_norm": 0.47082680463790894, "learning_rate": 0.00018060180576507756, "loss": 2.4342, "step": 1590 }, { "epoch": 0.729481889041724, "grad_norm": 0.48209020495414734, "learning_rate": 0.00018003074805479313, "loss": 2.4524, "step": 1591 }, { "epoch": 0.7299403943145346, "grad_norm": 0.4968235194683075, "learning_rate": 0.00017946039631600724, "loss": 2.3437, "step": 1592 }, { "epoch": 0.7303988995873453, "grad_norm": 0.4835037589073181, "learning_rate": 0.00017889075180712837, "loss": 2.4379, "step": 1593 }, { "epoch": 0.7308574048601559, "grad_norm": 0.475197434425354, "learning_rate": 0.00017832181578500512, "loss": 2.3867, "step": 1594 }, { "epoch": 0.7313159101329665, "grad_norm": 0.4952372610569, "learning_rate": 0.0001777535895049221, "loss": 2.3165, "step": 1595 }, { "epoch": 0.7317744154057771, "grad_norm": 0.4982004165649414, "learning_rate": 0.0001771860742205988, "loss": 2.4232, "step": 1596 }, { "epoch": 0.7322329206785878, "grad_norm": 0.4895860254764557, "learning_rate": 0.00017661927118418525, "loss": 2.3438, "step": 1597 }, { "epoch": 0.7326914259513985, "grad_norm": 0.4653266370296478, "learning_rate": 0.00017605318164626066, "loss": 2.3421, "step": 1598 }, { "epoch": 0.7331499312242091, "grad_norm": 0.45577967166900635, "learning_rate": 0.00017548780685582949, "loss": 2.4568, "step": 1599 }, { "epoch": 0.7336084364970197, "grad_norm": 0.4608478546142578, "learning_rate": 0.00017492314806031922, "loss": 2.5223, "step": 1600 }, { "epoch": 0.7340669417698303, "grad_norm": 0.505902886390686, "learning_rate": 0.00017435920650557806, "loss": 2.2809, "step": 1601 }, { "epoch": 0.734525447042641, "grad_norm": 0.4569036662578583, "learning_rate": 0.00017379598343587112, "loss": 2.3818, "step": 1602 }, { "epoch": 0.7349839523154517, "grad_norm": 0.45701631903648376, "learning_rate": 0.00017323348009387878, "loss": 2.4531, "step": 1603 }, { "epoch": 0.7354424575882622, "grad_norm": 0.47347304224967957, "learning_rate": 0.0001726716977206929, "loss": 2.3762, "step": 1604 }, { "epoch": 0.7359009628610729, "grad_norm": 0.4660072326660156, "learning_rate": 0.00017211063755581525, "loss": 2.3489, "step": 1605 }, { "epoch": 0.7363594681338835, "grad_norm": 0.47608378529548645, "learning_rate": 0.0001715503008371536, "loss": 2.3767, "step": 1606 }, { "epoch": 0.7368179734066942, "grad_norm": 0.4724350571632385, "learning_rate": 0.0001709906888010196, "loss": 2.4395, "step": 1607 }, { "epoch": 0.7372764786795049, "grad_norm": 0.4696725606918335, "learning_rate": 0.00017043180268212638, "loss": 2.3738, "step": 1608 }, { "epoch": 0.7377349839523154, "grad_norm": 0.4906388819217682, "learning_rate": 0.00016987364371358481, "loss": 2.3865, "step": 1609 }, { "epoch": 0.7381934892251261, "grad_norm": 0.48294582962989807, "learning_rate": 0.00016931621312690214, "loss": 2.4458, "step": 1610 }, { "epoch": 0.7386519944979367, "grad_norm": 0.47015276551246643, "learning_rate": 0.00016875951215197777, "loss": 2.4199, "step": 1611 }, { "epoch": 0.7391104997707474, "grad_norm": 0.48591873049736023, "learning_rate": 0.00016820354201710214, "loss": 2.3624, "step": 1612 }, { "epoch": 0.739569005043558, "grad_norm": 0.47013258934020996, "learning_rate": 0.00016764830394895203, "loss": 2.4075, "step": 1613 }, { "epoch": 0.7400275103163686, "grad_norm": 0.4634890854358673, "learning_rate": 0.00016709379917259027, "loss": 2.3926, "step": 1614 }, { "epoch": 0.7404860155891793, "grad_norm": 0.4429205656051636, "learning_rate": 0.00016654002891146091, "loss": 2.3476, "step": 1615 }, { "epoch": 0.7409445208619899, "grad_norm": 0.4862567186355591, "learning_rate": 0.00016598699438738764, "loss": 2.3741, "step": 1616 }, { "epoch": 0.7414030261348006, "grad_norm": 0.5004334449768066, "learning_rate": 0.00016543469682057105, "loss": 2.4199, "step": 1617 }, { "epoch": 0.7418615314076112, "grad_norm": 0.4711209237575531, "learning_rate": 0.00016488313742958526, "loss": 2.4231, "step": 1618 }, { "epoch": 0.7423200366804218, "grad_norm": 0.4766780734062195, "learning_rate": 0.00016433231743137646, "loss": 2.428, "step": 1619 }, { "epoch": 0.7427785419532325, "grad_norm": 0.464592844247818, "learning_rate": 0.0001637822380412584, "loss": 2.3458, "step": 1620 }, { "epoch": 0.7432370472260431, "grad_norm": 0.48016810417175293, "learning_rate": 0.00016323290047291195, "loss": 2.3249, "step": 1621 }, { "epoch": 0.7436955524988538, "grad_norm": 0.47811198234558105, "learning_rate": 0.0001626843059383803, "loss": 2.4497, "step": 1622 }, { "epoch": 0.7441540577716643, "grad_norm": 0.47525420784950256, "learning_rate": 0.00016213645564806752, "loss": 2.4762, "step": 1623 }, { "epoch": 0.744612563044475, "grad_norm": 0.4698454439640045, "learning_rate": 0.0001615893508107359, "loss": 2.2558, "step": 1624 }, { "epoch": 0.7450710683172856, "grad_norm": 0.47121182084083557, "learning_rate": 0.00016104299263350252, "loss": 2.4076, "step": 1625 }, { "epoch": 0.7455295735900963, "grad_norm": 0.4660525619983673, "learning_rate": 0.00016049738232183758, "loss": 2.3568, "step": 1626 }, { "epoch": 0.745988078862907, "grad_norm": 0.46998924016952515, "learning_rate": 0.0001599525210795606, "loss": 2.2751, "step": 1627 }, { "epoch": 0.7464465841357175, "grad_norm": 0.43794572353363037, "learning_rate": 0.00015940841010883889, "loss": 2.383, "step": 1628 }, { "epoch": 0.7469050894085282, "grad_norm": 0.4752472937107086, "learning_rate": 0.00015886505061018413, "loss": 2.37, "step": 1629 }, { "epoch": 0.7473635946813388, "grad_norm": 0.4532284140586853, "learning_rate": 0.0001583224437824498, "loss": 2.3423, "step": 1630 }, { "epoch": 0.7478220999541495, "grad_norm": 0.48096874356269836, "learning_rate": 0.0001577805908228293, "loss": 2.4491, "step": 1631 }, { "epoch": 0.7482806052269602, "grad_norm": 0.4905768036842346, "learning_rate": 0.00015723949292685191, "loss": 2.3277, "step": 1632 }, { "epoch": 0.7487391104997707, "grad_norm": 0.49635952711105347, "learning_rate": 0.0001566991512883818, "loss": 2.3436, "step": 1633 }, { "epoch": 0.7491976157725814, "grad_norm": 0.48239338397979736, "learning_rate": 0.00015615956709961378, "loss": 2.506, "step": 1634 }, { "epoch": 0.749656121045392, "grad_norm": 0.4737139344215393, "learning_rate": 0.00015562074155107215, "loss": 2.3906, "step": 1635 }, { "epoch": 0.7501146263182027, "grad_norm": 0.48808157444000244, "learning_rate": 0.0001550826758316068, "loss": 2.4622, "step": 1636 }, { "epoch": 0.7505731315910132, "grad_norm": 0.49779441952705383, "learning_rate": 0.00015454537112839122, "loss": 2.4779, "step": 1637 }, { "epoch": 0.7510316368638239, "grad_norm": 0.46816036105155945, "learning_rate": 0.00015400882862692033, "loss": 2.4207, "step": 1638 }, { "epoch": 0.7514901421366346, "grad_norm": 0.47931739687919617, "learning_rate": 0.00015347304951100665, "loss": 2.4057, "step": 1639 }, { "epoch": 0.7519486474094452, "grad_norm": 0.4722326695919037, "learning_rate": 0.00015293803496277907, "loss": 2.4537, "step": 1640 }, { "epoch": 0.7524071526822559, "grad_norm": 0.4773688316345215, "learning_rate": 0.00015240378616267886, "loss": 2.4405, "step": 1641 }, { "epoch": 0.7528656579550664, "grad_norm": 0.4539555311203003, "learning_rate": 0.00015187030428945843, "loss": 2.258, "step": 1642 }, { "epoch": 0.7533241632278771, "grad_norm": 0.4743576943874359, "learning_rate": 0.0001513375905201776, "loss": 2.4136, "step": 1643 }, { "epoch": 0.7537826685006878, "grad_norm": 0.48475944995880127, "learning_rate": 0.00015080564603020142, "loss": 2.3943, "step": 1644 }, { "epoch": 0.7542411737734984, "grad_norm": 0.4910453259944916, "learning_rate": 0.0001502744719931982, "loss": 2.3059, "step": 1645 }, { "epoch": 0.7546996790463091, "grad_norm": 0.48358917236328125, "learning_rate": 0.00014974406958113558, "loss": 2.4159, "step": 1646 }, { "epoch": 0.7551581843191196, "grad_norm": 0.4878486692905426, "learning_rate": 0.00014921443996427947, "loss": 2.4109, "step": 1647 }, { "epoch": 0.7556166895919303, "grad_norm": 0.4827605187892914, "learning_rate": 0.0001486855843111901, "loss": 2.4798, "step": 1648 }, { "epoch": 0.756075194864741, "grad_norm": 0.4724687933921814, "learning_rate": 0.0001481575037887201, "loss": 2.4302, "step": 1649 }, { "epoch": 0.7565337001375516, "grad_norm": 0.45011085271835327, "learning_rate": 0.00014763019956201253, "loss": 2.4324, "step": 1650 }, { "epoch": 0.7569922054103623, "grad_norm": 0.478814035654068, "learning_rate": 0.0001471036727944966, "loss": 2.3392, "step": 1651 }, { "epoch": 0.7574507106831728, "grad_norm": 0.457099586725235, "learning_rate": 0.0001465779246478872, "loss": 2.3201, "step": 1652 }, { "epoch": 0.7579092159559835, "grad_norm": 0.4550655782222748, "learning_rate": 0.00014605295628218045, "loss": 2.2798, "step": 1653 }, { "epoch": 0.7583677212287941, "grad_norm": 0.47658994793891907, "learning_rate": 0.0001455287688556527, "loss": 2.3809, "step": 1654 }, { "epoch": 0.7588262265016048, "grad_norm": 0.45358383655548096, "learning_rate": 0.00014500536352485673, "loss": 2.324, "step": 1655 }, { "epoch": 0.7592847317744154, "grad_norm": 0.46572446823120117, "learning_rate": 0.00014448274144461965, "loss": 2.3895, "step": 1656 }, { "epoch": 0.759743237047226, "grad_norm": 0.46312031149864197, "learning_rate": 0.00014396090376804112, "loss": 2.4155, "step": 1657 }, { "epoch": 0.7602017423200367, "grad_norm": 0.4597212076187134, "learning_rate": 0.00014343985164648926, "loss": 2.2804, "step": 1658 }, { "epoch": 0.7606602475928473, "grad_norm": 0.46114569902420044, "learning_rate": 0.00014291958622959973, "loss": 2.5432, "step": 1659 }, { "epoch": 0.761118752865658, "grad_norm": 0.4833245873451233, "learning_rate": 0.00014240010866527176, "loss": 2.4022, "step": 1660 }, { "epoch": 0.7615772581384685, "grad_norm": 0.4721977710723877, "learning_rate": 0.00014188142009966686, "loss": 2.3949, "step": 1661 }, { "epoch": 0.7620357634112792, "grad_norm": 0.4890829622745514, "learning_rate": 0.0001413635216772053, "loss": 2.3322, "step": 1662 }, { "epoch": 0.7624942686840899, "grad_norm": 0.5047375559806824, "learning_rate": 0.000140846414540564, "loss": 2.4191, "step": 1663 }, { "epoch": 0.7629527739569005, "grad_norm": 0.46636152267456055, "learning_rate": 0.00014033009983067452, "loss": 2.4087, "step": 1664 }, { "epoch": 0.7634112792297112, "grad_norm": 0.45077434182167053, "learning_rate": 0.00013981457868671927, "loss": 2.3147, "step": 1665 }, { "epoch": 0.7638697845025217, "grad_norm": 0.46734321117401123, "learning_rate": 0.0001392998522461305, "loss": 2.3598, "step": 1666 }, { "epoch": 0.7643282897753324, "grad_norm": 0.5054630041122437, "learning_rate": 0.00013878592164458635, "loss": 2.4345, "step": 1667 }, { "epoch": 0.7647867950481431, "grad_norm": 0.46657657623291016, "learning_rate": 0.00013827278801600978, "loss": 2.3601, "step": 1668 }, { "epoch": 0.7652453003209537, "grad_norm": 0.4739791750907898, "learning_rate": 0.0001377604524925647, "loss": 2.3903, "step": 1669 }, { "epoch": 0.7657038055937643, "grad_norm": 0.47624868154525757, "learning_rate": 0.00013724891620465424, "loss": 2.3224, "step": 1670 }, { "epoch": 0.7661623108665749, "grad_norm": 0.49205082654953003, "learning_rate": 0.0001367381802809185, "loss": 2.4229, "step": 1671 }, { "epoch": 0.7666208161393856, "grad_norm": 0.49947649240493774, "learning_rate": 0.00013622824584823113, "loss": 2.3743, "step": 1672 }, { "epoch": 0.7670793214121963, "grad_norm": 0.4628148376941681, "learning_rate": 0.00013571911403169795, "loss": 2.3834, "step": 1673 }, { "epoch": 0.7675378266850069, "grad_norm": 0.47582125663757324, "learning_rate": 0.0001352107859546533, "loss": 2.4211, "step": 1674 }, { "epoch": 0.7679963319578175, "grad_norm": 0.45759981870651245, "learning_rate": 0.00013470326273865886, "loss": 2.3979, "step": 1675 }, { "epoch": 0.7684548372306281, "grad_norm": 0.4718741774559021, "learning_rate": 0.00013419654550349985, "loss": 2.3265, "step": 1676 }, { "epoch": 0.7689133425034388, "grad_norm": 0.4663947522640228, "learning_rate": 0.00013369063536718346, "loss": 2.3176, "step": 1677 }, { "epoch": 0.7693718477762495, "grad_norm": 0.4780106246471405, "learning_rate": 0.00013318553344593632, "loss": 2.3343, "step": 1678 }, { "epoch": 0.7698303530490601, "grad_norm": 0.5050276517868042, "learning_rate": 0.00013268124085420136, "loss": 2.3559, "step": 1679 }, { "epoch": 0.7702888583218707, "grad_norm": 0.46845418214797974, "learning_rate": 0.0001321777587046364, "loss": 2.3707, "step": 1680 }, { "epoch": 0.7707473635946813, "grad_norm": 0.49548783898353577, "learning_rate": 0.00013167508810811059, "loss": 2.3549, "step": 1681 }, { "epoch": 0.771205868867492, "grad_norm": 0.4831186532974243, "learning_rate": 0.0001311732301737029, "loss": 2.4144, "step": 1682 }, { "epoch": 0.7716643741403026, "grad_norm": 0.46997007727622986, "learning_rate": 0.0001306721860086991, "loss": 2.4072, "step": 1683 }, { "epoch": 0.7721228794131133, "grad_norm": 0.4688609540462494, "learning_rate": 0.00013017195671858928, "loss": 2.3078, "step": 1684 }, { "epoch": 0.7725813846859239, "grad_norm": 0.44723641872406006, "learning_rate": 0.0001296725434070661, "loss": 2.3179, "step": 1685 }, { "epoch": 0.7730398899587345, "grad_norm": 0.4652445614337921, "learning_rate": 0.00012917394717602121, "loss": 2.3785, "step": 1686 }, { "epoch": 0.7734983952315452, "grad_norm": 0.5763563513755798, "learning_rate": 0.00012867616912554426, "loss": 2.5041, "step": 1687 }, { "epoch": 0.7739569005043558, "grad_norm": 0.47587093710899353, "learning_rate": 0.00012817921035391882, "loss": 2.3852, "step": 1688 }, { "epoch": 0.7744154057771664, "grad_norm": 0.4751850366592407, "learning_rate": 0.00012768307195762168, "loss": 2.4434, "step": 1689 }, { "epoch": 0.774873911049977, "grad_norm": 0.48266640305519104, "learning_rate": 0.00012718775503131908, "loss": 2.3216, "step": 1690 }, { "epoch": 0.7753324163227877, "grad_norm": 0.4559253752231598, "learning_rate": 0.0001266932606678646, "loss": 2.3158, "step": 1691 }, { "epoch": 0.7757909215955984, "grad_norm": 0.46752840280532837, "learning_rate": 0.00012619958995829756, "loss": 2.3616, "step": 1692 }, { "epoch": 0.776249426868409, "grad_norm": 0.4706156551837921, "learning_rate": 0.0001257067439918394, "loss": 2.328, "step": 1693 }, { "epoch": 0.7767079321412196, "grad_norm": 0.46821895241737366, "learning_rate": 0.00012521472385589234, "loss": 2.4262, "step": 1694 }, { "epoch": 0.7771664374140302, "grad_norm": 0.4758947789669037, "learning_rate": 0.00012472353063603626, "loss": 2.3768, "step": 1695 }, { "epoch": 0.7776249426868409, "grad_norm": 0.48505285382270813, "learning_rate": 0.0001242331654160263, "loss": 2.3749, "step": 1696 }, { "epoch": 0.7780834479596516, "grad_norm": 0.4897075891494751, "learning_rate": 0.0001237436292777914, "loss": 2.4665, "step": 1697 }, { "epoch": 0.7785419532324622, "grad_norm": 0.481426477432251, "learning_rate": 0.00012325492330143061, "loss": 2.429, "step": 1698 }, { "epoch": 0.7790004585052728, "grad_norm": 0.48033830523490906, "learning_rate": 0.00012276704856521175, "loss": 2.4965, "step": 1699 }, { "epoch": 0.7794589637780834, "grad_norm": 0.47155362367630005, "learning_rate": 0.00012228000614556816, "loss": 2.3133, "step": 1700 }, { "epoch": 0.7799174690508941, "grad_norm": 0.46856945753097534, "learning_rate": 0.00012179379711709738, "loss": 2.2483, "step": 1701 }, { "epoch": 0.7803759743237048, "grad_norm": 0.4639533460140228, "learning_rate": 0.0001213084225525577, "loss": 2.5019, "step": 1702 }, { "epoch": 0.7808344795965153, "grad_norm": 0.4520926773548126, "learning_rate": 0.00012082388352286627, "loss": 2.411, "step": 1703 }, { "epoch": 0.781292984869326, "grad_norm": 0.45112207531929016, "learning_rate": 0.00012034018109709716, "loss": 2.3825, "step": 1704 }, { "epoch": 0.7817514901421366, "grad_norm": 0.4551258087158203, "learning_rate": 0.00011985731634247809, "loss": 2.4252, "step": 1705 }, { "epoch": 0.7822099954149473, "grad_norm": 0.4593683183193207, "learning_rate": 0.00011937529032438904, "loss": 2.367, "step": 1706 }, { "epoch": 0.782668500687758, "grad_norm": 0.4732966125011444, "learning_rate": 0.00011889410410635887, "loss": 2.3612, "step": 1707 }, { "epoch": 0.7831270059605685, "grad_norm": 0.4681137204170227, "learning_rate": 0.0001184137587500641, "loss": 2.423, "step": 1708 }, { "epoch": 0.7835855112333792, "grad_norm": 0.49649733304977417, "learning_rate": 0.00011793425531532564, "loss": 2.3485, "step": 1709 }, { "epoch": 0.7840440165061898, "grad_norm": 0.4702542722225189, "learning_rate": 0.00011745559486010671, "loss": 2.4986, "step": 1710 }, { "epoch": 0.7845025217790005, "grad_norm": 0.4661259651184082, "learning_rate": 0.00011697777844051105, "loss": 2.3154, "step": 1711 }, { "epoch": 0.7849610270518111, "grad_norm": 0.4549489915370941, "learning_rate": 0.00011650080711077964, "loss": 2.3849, "step": 1712 }, { "epoch": 0.7854195323246217, "grad_norm": 0.4829387664794922, "learning_rate": 0.00011602468192328936, "loss": 2.3757, "step": 1713 }, { "epoch": 0.7858780375974324, "grad_norm": 0.4641295075416565, "learning_rate": 0.00011554940392854973, "loss": 2.3573, "step": 1714 }, { "epoch": 0.786336542870243, "grad_norm": 0.4821907877922058, "learning_rate": 0.00011507497417520146, "loss": 2.4427, "step": 1715 }, { "epoch": 0.7867950481430537, "grad_norm": 0.48769611120224, "learning_rate": 0.00011460139371001339, "loss": 2.4476, "step": 1716 }, { "epoch": 0.7872535534158643, "grad_norm": 0.44480088353157043, "learning_rate": 0.00011412866357788049, "loss": 2.3402, "step": 1717 }, { "epoch": 0.7877120586886749, "grad_norm": 0.4654664397239685, "learning_rate": 0.00011365678482182207, "loss": 2.4033, "step": 1718 }, { "epoch": 0.7881705639614855, "grad_norm": 0.4757126271724701, "learning_rate": 0.0001131857584829783, "loss": 2.3994, "step": 1719 }, { "epoch": 0.7886290692342962, "grad_norm": 0.4959290325641632, "learning_rate": 0.0001127155856006093, "loss": 2.344, "step": 1720 }, { "epoch": 0.7890875745071069, "grad_norm": 0.462208092212677, "learning_rate": 0.00011224626721209141, "loss": 2.3793, "step": 1721 }, { "epoch": 0.7895460797799174, "grad_norm": 0.47644922137260437, "learning_rate": 0.0001117778043529164, "loss": 2.4288, "step": 1722 }, { "epoch": 0.7900045850527281, "grad_norm": 0.4765971601009369, "learning_rate": 0.0001113101980566879, "loss": 2.4143, "step": 1723 }, { "epoch": 0.7904630903255387, "grad_norm": 0.4794289767742157, "learning_rate": 0.00011084344935511958, "loss": 2.3318, "step": 1724 }, { "epoch": 0.7909215955983494, "grad_norm": 0.4636545479297638, "learning_rate": 0.00011037755927803345, "loss": 2.338, "step": 1725 }, { "epoch": 0.7913801008711601, "grad_norm": 0.4522865414619446, "learning_rate": 0.00010991252885335651, "loss": 2.3727, "step": 1726 }, { "epoch": 0.7918386061439706, "grad_norm": 0.4832746088504791, "learning_rate": 0.00010944835910711958, "loss": 2.3507, "step": 1727 }, { "epoch": 0.7922971114167813, "grad_norm": 0.4782036244869232, "learning_rate": 0.00010898505106345396, "loss": 2.4603, "step": 1728 }, { "epoch": 0.7927556166895919, "grad_norm": 0.48222312331199646, "learning_rate": 0.00010852260574459022, "loss": 2.3128, "step": 1729 }, { "epoch": 0.7932141219624026, "grad_norm": 0.4763835370540619, "learning_rate": 0.00010806102417085512, "loss": 2.3833, "step": 1730 }, { "epoch": 0.7936726272352133, "grad_norm": 0.46209847927093506, "learning_rate": 0.00010760030736066951, "loss": 2.4494, "step": 1731 }, { "epoch": 0.7941311325080238, "grad_norm": 0.46913397312164307, "learning_rate": 0.00010714045633054687, "loss": 2.3916, "step": 1732 }, { "epoch": 0.7945896377808345, "grad_norm": 0.46026411652565, "learning_rate": 0.00010668147209508971, "loss": 2.2989, "step": 1733 }, { "epoch": 0.7950481430536451, "grad_norm": 0.48411527276039124, "learning_rate": 0.00010622335566698877, "loss": 2.554, "step": 1734 }, { "epoch": 0.7955066483264558, "grad_norm": 0.47193488478660583, "learning_rate": 0.00010576610805701942, "loss": 2.4309, "step": 1735 }, { "epoch": 0.7959651535992663, "grad_norm": 0.5021685361862183, "learning_rate": 0.00010530973027404073, "loss": 2.3604, "step": 1736 }, { "epoch": 0.796423658872077, "grad_norm": 0.5073973536491394, "learning_rate": 0.00010485422332499212, "loss": 2.3956, "step": 1737 }, { "epoch": 0.7968821641448877, "grad_norm": 0.47074732184410095, "learning_rate": 0.00010439958821489165, "loss": 2.4275, "step": 1738 }, { "epoch": 0.7973406694176983, "grad_norm": 0.46118077635765076, "learning_rate": 0.00010394582594683428, "loss": 2.4972, "step": 1739 }, { "epoch": 0.797799174690509, "grad_norm": 0.4576266407966614, "learning_rate": 0.0001034929375219884, "loss": 2.3367, "step": 1740 }, { "epoch": 0.7982576799633195, "grad_norm": 0.4880698323249817, "learning_rate": 0.00010304092393959514, "loss": 2.3059, "step": 1741 }, { "epoch": 0.7987161852361302, "grad_norm": 0.45509767532348633, "learning_rate": 0.00010258978619696468, "loss": 2.3849, "step": 1742 }, { "epoch": 0.7991746905089409, "grad_norm": 0.46991655230522156, "learning_rate": 0.00010213952528947551, "loss": 2.4419, "step": 1743 }, { "epoch": 0.7996331957817515, "grad_norm": 0.4786435067653656, "learning_rate": 0.00010169014221057089, "loss": 2.3362, "step": 1744 }, { "epoch": 0.8000917010545622, "grad_norm": 0.4890434145927429, "learning_rate": 0.00010124163795175734, "loss": 2.383, "step": 1745 }, { "epoch": 0.8005502063273727, "grad_norm": 0.4533989727497101, "learning_rate": 0.00010079401350260287, "loss": 2.3929, "step": 1746 }, { "epoch": 0.8010087116001834, "grad_norm": 0.4767332077026367, "learning_rate": 0.00010034726985073362, "loss": 2.4437, "step": 1747 }, { "epoch": 0.801467216872994, "grad_norm": 0.4689553678035736, "learning_rate": 9.9901407981833e-05, "loss": 2.4194, "step": 1748 }, { "epoch": 0.8019257221458047, "grad_norm": 0.4735293686389923, "learning_rate": 9.94564288796384e-05, "loss": 2.3093, "step": 1749 }, { "epoch": 0.8023842274186153, "grad_norm": 0.4508242607116699, "learning_rate": 9.901233352593953e-05, "loss": 2.4041, "step": 1750 }, { "epoch": 0.8028427326914259, "grad_norm": 0.48054811358451843, "learning_rate": 9.856912290057668e-05, "loss": 2.3963, "step": 1751 }, { "epoch": 0.8033012379642366, "grad_norm": 0.469817191362381, "learning_rate": 9.812679798143748e-05, "loss": 2.4032, "step": 1752 }, { "epoch": 0.8037597432370472, "grad_norm": 0.49669188261032104, "learning_rate": 9.768535974445586e-05, "loss": 2.3936, "step": 1753 }, { "epoch": 0.8042182485098579, "grad_norm": 0.4737200438976288, "learning_rate": 9.724480916360906e-05, "loss": 2.4004, "step": 1754 }, { "epoch": 0.8046767537826685, "grad_norm": 0.5013248324394226, "learning_rate": 9.68051472109162e-05, "loss": 2.5, "step": 1755 }, { "epoch": 0.8051352590554791, "grad_norm": 0.47001272439956665, "learning_rate": 9.636637485643529e-05, "loss": 2.3159, "step": 1756 }, { "epoch": 0.8055937643282898, "grad_norm": 0.46256357431411743, "learning_rate": 9.592849306826174e-05, "loss": 2.3619, "step": 1757 }, { "epoch": 0.8060522696011004, "grad_norm": 0.4967513382434845, "learning_rate": 9.549150281252633e-05, "loss": 2.4955, "step": 1758 }, { "epoch": 0.8065107748739111, "grad_norm": 0.48486965894699097, "learning_rate": 9.505540505339223e-05, "loss": 2.2931, "step": 1759 }, { "epoch": 0.8069692801467216, "grad_norm": 0.4766705334186554, "learning_rate": 9.4620200753054e-05, "loss": 2.4093, "step": 1760 }, { "epoch": 0.8074277854195323, "grad_norm": 0.4748774468898773, "learning_rate": 9.418589087173441e-05, "loss": 2.3832, "step": 1761 }, { "epoch": 0.807886290692343, "grad_norm": 0.47227227687835693, "learning_rate": 9.375247636768325e-05, "loss": 2.3007, "step": 1762 }, { "epoch": 0.8083447959651536, "grad_norm": 0.4767397344112396, "learning_rate": 9.331995819717443e-05, "loss": 2.3938, "step": 1763 }, { "epoch": 0.8088033012379643, "grad_norm": 0.4610329866409302, "learning_rate": 9.288833731450419e-05, "loss": 2.353, "step": 1764 }, { "epoch": 0.8092618065107748, "grad_norm": 0.472724974155426, "learning_rate": 9.245761467198948e-05, "loss": 2.3671, "step": 1765 }, { "epoch": 0.8097203117835855, "grad_norm": 0.4860416054725647, "learning_rate": 9.20277912199648e-05, "loss": 2.2732, "step": 1766 }, { "epoch": 0.8101788170563962, "grad_norm": 0.4729323387145996, "learning_rate": 9.159886790678123e-05, "loss": 2.4128, "step": 1767 }, { "epoch": 0.8106373223292068, "grad_norm": 0.45774078369140625, "learning_rate": 9.11708456788033e-05, "loss": 2.332, "step": 1768 }, { "epoch": 0.8110958276020174, "grad_norm": 0.47525468468666077, "learning_rate": 9.074372548040793e-05, "loss": 2.3243, "step": 1769 }, { "epoch": 0.811554332874828, "grad_norm": 0.45800334215164185, "learning_rate": 9.031750825398145e-05, "loss": 2.4182, "step": 1770 }, { "epoch": 0.8120128381476387, "grad_norm": 0.47021520137786865, "learning_rate": 8.98921949399179e-05, "loss": 2.4281, "step": 1771 }, { "epoch": 0.8124713434204494, "grad_norm": 0.4702485203742981, "learning_rate": 8.94677864766173e-05, "loss": 2.3194, "step": 1772 }, { "epoch": 0.81292984869326, "grad_norm": 0.49320077896118164, "learning_rate": 8.904428380048269e-05, "loss": 2.407, "step": 1773 }, { "epoch": 0.8133883539660706, "grad_norm": 0.4607609808444977, "learning_rate": 8.862168784591929e-05, "loss": 2.3917, "step": 1774 }, { "epoch": 0.8138468592388812, "grad_norm": 0.47798216342926025, "learning_rate": 8.819999954533115e-05, "loss": 2.3464, "step": 1775 }, { "epoch": 0.8143053645116919, "grad_norm": 0.5091061592102051, "learning_rate": 8.777921982911996e-05, "loss": 2.2699, "step": 1776 }, { "epoch": 0.8147638697845025, "grad_norm": 0.4729434847831726, "learning_rate": 8.735934962568253e-05, "loss": 2.3369, "step": 1777 }, { "epoch": 0.8152223750573132, "grad_norm": 0.48135906457901, "learning_rate": 8.694038986140945e-05, "loss": 2.387, "step": 1778 }, { "epoch": 0.8156808803301238, "grad_norm": 0.49116405844688416, "learning_rate": 8.652234146068206e-05, "loss": 2.3728, "step": 1779 }, { "epoch": 0.8161393856029344, "grad_norm": 0.4826553761959076, "learning_rate": 8.610520534587086e-05, "loss": 2.3249, "step": 1780 }, { "epoch": 0.8165978908757451, "grad_norm": 0.4927981495857239, "learning_rate": 8.568898243733397e-05, "loss": 2.355, "step": 1781 }, { "epoch": 0.8170563961485557, "grad_norm": 0.5061519145965576, "learning_rate": 8.527367365341409e-05, "loss": 2.4352, "step": 1782 }, { "epoch": 0.8175149014213663, "grad_norm": 0.485474556684494, "learning_rate": 8.485927991043757e-05, "loss": 2.4715, "step": 1783 }, { "epoch": 0.817973406694177, "grad_norm": 0.496390700340271, "learning_rate": 8.444580212271125e-05, "loss": 2.4139, "step": 1784 }, { "epoch": 0.8184319119669876, "grad_norm": 0.4663461446762085, "learning_rate": 8.403324120252159e-05, "loss": 2.3988, "step": 1785 }, { "epoch": 0.8188904172397983, "grad_norm": 0.4826914668083191, "learning_rate": 8.362159806013175e-05, "loss": 2.3349, "step": 1786 }, { "epoch": 0.8193489225126089, "grad_norm": 0.4732806384563446, "learning_rate": 8.321087360377988e-05, "loss": 2.3253, "step": 1787 }, { "epoch": 0.8198074277854195, "grad_norm": 0.4882335960865021, "learning_rate": 8.280106873967752e-05, "loss": 2.4848, "step": 1788 }, { "epoch": 0.8202659330582301, "grad_norm": 0.4560544192790985, "learning_rate": 8.239218437200679e-05, "loss": 2.3753, "step": 1789 }, { "epoch": 0.8207244383310408, "grad_norm": 0.4739437401294708, "learning_rate": 8.198422140291939e-05, "loss": 2.3146, "step": 1790 }, { "epoch": 0.8211829436038515, "grad_norm": 0.47248467803001404, "learning_rate": 8.157718073253351e-05, "loss": 2.4481, "step": 1791 }, { "epoch": 0.8216414488766621, "grad_norm": 0.48067110776901245, "learning_rate": 8.117106325893287e-05, "loss": 2.3155, "step": 1792 }, { "epoch": 0.8220999541494727, "grad_norm": 0.4883652329444885, "learning_rate": 8.076586987816404e-05, "loss": 2.3755, "step": 1793 }, { "epoch": 0.8225584594222833, "grad_norm": 0.48846834897994995, "learning_rate": 8.036160148423449e-05, "loss": 2.3761, "step": 1794 }, { "epoch": 0.823016964695094, "grad_norm": 0.48132023215293884, "learning_rate": 7.995825896911141e-05, "loss": 2.3592, "step": 1795 }, { "epoch": 0.8234754699679047, "grad_norm": 0.45197194814682007, "learning_rate": 7.955584322271853e-05, "loss": 2.3763, "step": 1796 }, { "epoch": 0.8239339752407153, "grad_norm": 0.4643998444080353, "learning_rate": 7.915435513293523e-05, "loss": 2.3807, "step": 1797 }, { "epoch": 0.8243924805135259, "grad_norm": 0.46614041924476624, "learning_rate": 7.875379558559387e-05, "loss": 2.2151, "step": 1798 }, { "epoch": 0.8248509857863365, "grad_norm": 0.47697725892066956, "learning_rate": 7.835416546447838e-05, "loss": 2.3796, "step": 1799 }, { "epoch": 0.8253094910591472, "grad_norm": 0.4705403447151184, "learning_rate": 7.795546565132167e-05, "loss": 2.3889, "step": 1800 }, { "epoch": 0.8257679963319579, "grad_norm": 0.5231021642684937, "learning_rate": 7.755769702580412e-05, "loss": 2.3821, "step": 1801 }, { "epoch": 0.8262265016047684, "grad_norm": 0.4732932448387146, "learning_rate": 7.716086046555193e-05, "loss": 2.3595, "step": 1802 }, { "epoch": 0.8266850068775791, "grad_norm": 0.4723265767097473, "learning_rate": 7.676495684613432e-05, "loss": 2.3862, "step": 1803 }, { "epoch": 0.8271435121503897, "grad_norm": 0.4961133599281311, "learning_rate": 7.636998704106252e-05, "loss": 2.3011, "step": 1804 }, { "epoch": 0.8276020174232004, "grad_norm": 0.48994192481040955, "learning_rate": 7.597595192178702e-05, "loss": 2.3088, "step": 1805 }, { "epoch": 0.828060522696011, "grad_norm": 0.47138819098472595, "learning_rate": 7.558285235769646e-05, "loss": 2.3274, "step": 1806 }, { "epoch": 0.8285190279688216, "grad_norm": 0.4779699444770813, "learning_rate": 7.519068921611494e-05, "loss": 2.4083, "step": 1807 }, { "epoch": 0.8289775332416323, "grad_norm": 0.513308584690094, "learning_rate": 7.479946336230047e-05, "loss": 2.3832, "step": 1808 }, { "epoch": 0.8294360385144429, "grad_norm": 0.49236026406288147, "learning_rate": 7.440917565944349e-05, "loss": 2.4219, "step": 1809 }, { "epoch": 0.8298945437872536, "grad_norm": 0.4522506296634674, "learning_rate": 7.4019826968664e-05, "loss": 2.3809, "step": 1810 }, { "epoch": 0.8303530490600642, "grad_norm": 0.485207200050354, "learning_rate": 7.363141814901053e-05, "loss": 2.3935, "step": 1811 }, { "epoch": 0.8308115543328748, "grad_norm": 0.44153255224227905, "learning_rate": 7.32439500574577e-05, "loss": 2.3662, "step": 1812 }, { "epoch": 0.8312700596056855, "grad_norm": 0.45327600836753845, "learning_rate": 7.285742354890473e-05, "loss": 2.3492, "step": 1813 }, { "epoch": 0.8317285648784961, "grad_norm": 0.5051504373550415, "learning_rate": 7.247183947617325e-05, "loss": 2.3595, "step": 1814 }, { "epoch": 0.8321870701513068, "grad_norm": 0.47452110052108765, "learning_rate": 7.20871986900053e-05, "loss": 2.3214, "step": 1815 }, { "epoch": 0.8326455754241173, "grad_norm": 0.46808671951293945, "learning_rate": 7.170350203906218e-05, "loss": 2.325, "step": 1816 }, { "epoch": 0.833104080696928, "grad_norm": 0.4625532627105713, "learning_rate": 7.132075036992158e-05, "loss": 2.3774, "step": 1817 }, { "epoch": 0.8335625859697386, "grad_norm": 0.5878528356552124, "learning_rate": 7.093894452707666e-05, "loss": 2.3256, "step": 1818 }, { "epoch": 0.8340210912425493, "grad_norm": 0.4643569588661194, "learning_rate": 7.055808535293334e-05, "loss": 2.3127, "step": 1819 }, { "epoch": 0.83447959651536, "grad_norm": 0.5026644468307495, "learning_rate": 7.017817368780888e-05, "loss": 2.4113, "step": 1820 }, { "epoch": 0.8349381017881705, "grad_norm": 0.449619323015213, "learning_rate": 6.979921036993042e-05, "loss": 2.3616, "step": 1821 }, { "epoch": 0.8353966070609812, "grad_norm": 0.5110437273979187, "learning_rate": 6.942119623543202e-05, "loss": 2.4742, "step": 1822 }, { "epoch": 0.8358551123337918, "grad_norm": 0.45372867584228516, "learning_rate": 6.904413211835414e-05, "loss": 2.3534, "step": 1823 }, { "epoch": 0.8363136176066025, "grad_norm": 0.4746163487434387, "learning_rate": 6.866801885064056e-05, "loss": 2.3868, "step": 1824 }, { "epoch": 0.8367721228794132, "grad_norm": 0.4805297553539276, "learning_rate": 6.829285726213769e-05, "loss": 2.3307, "step": 1825 }, { "epoch": 0.8372306281522237, "grad_norm": 0.4842601716518402, "learning_rate": 6.79186481805918e-05, "loss": 2.3733, "step": 1826 }, { "epoch": 0.8376891334250344, "grad_norm": 0.4564161002635956, "learning_rate": 6.754539243164754e-05, "loss": 2.3771, "step": 1827 }, { "epoch": 0.838147638697845, "grad_norm": 0.4593990445137024, "learning_rate": 6.717309083884654e-05, "loss": 2.4028, "step": 1828 }, { "epoch": 0.8386061439706557, "grad_norm": 0.4829443097114563, "learning_rate": 6.680174422362468e-05, "loss": 2.2797, "step": 1829 }, { "epoch": 0.8390646492434664, "grad_norm": 0.46747028827667236, "learning_rate": 6.643135340531136e-05, "loss": 2.4158, "step": 1830 }, { "epoch": 0.8395231545162769, "grad_norm": 0.47184044122695923, "learning_rate": 6.606191920112664e-05, "loss": 2.4592, "step": 1831 }, { "epoch": 0.8399816597890876, "grad_norm": 0.47340887784957886, "learning_rate": 6.569344242618036e-05, "loss": 2.321, "step": 1832 }, { "epoch": 0.8404401650618982, "grad_norm": 0.48284947872161865, "learning_rate": 6.532592389346958e-05, "loss": 2.4392, "step": 1833 }, { "epoch": 0.8408986703347089, "grad_norm": 0.5048288702964783, "learning_rate": 6.495936441387713e-05, "loss": 2.3329, "step": 1834 }, { "epoch": 0.8413571756075194, "grad_norm": 0.47203776240348816, "learning_rate": 6.459376479617013e-05, "loss": 2.3761, "step": 1835 }, { "epoch": 0.8418156808803301, "grad_norm": 0.4729713201522827, "learning_rate": 6.422912584699752e-05, "loss": 2.4178, "step": 1836 }, { "epoch": 0.8422741861531408, "grad_norm": 0.4563446342945099, "learning_rate": 6.386544837088904e-05, "loss": 2.3095, "step": 1837 }, { "epoch": 0.8427326914259514, "grad_norm": 0.4515555799007416, "learning_rate": 6.350273317025251e-05, "loss": 2.2634, "step": 1838 }, { "epoch": 0.8431911966987621, "grad_norm": 0.4782533049583435, "learning_rate": 6.314098104537324e-05, "loss": 2.4055, "step": 1839 }, { "epoch": 0.8436497019715726, "grad_norm": 0.4826701581478119, "learning_rate": 6.278019279441122e-05, "loss": 2.4029, "step": 1840 }, { "epoch": 0.8441082072443833, "grad_norm": 0.4790927767753601, "learning_rate": 6.242036921339972e-05, "loss": 2.3569, "step": 1841 }, { "epoch": 0.844566712517194, "grad_norm": 0.5025001168251038, "learning_rate": 6.206151109624402e-05, "loss": 2.365, "step": 1842 }, { "epoch": 0.8450252177900046, "grad_norm": 0.48382657766342163, "learning_rate": 6.170361923471868e-05, "loss": 2.3001, "step": 1843 }, { "epoch": 0.8454837230628153, "grad_norm": 0.4546387195587158, "learning_rate": 6.134669441846691e-05, "loss": 2.3543, "step": 1844 }, { "epoch": 0.8459422283356258, "grad_norm": 0.4845122694969177, "learning_rate": 6.099073743499772e-05, "loss": 2.4106, "step": 1845 }, { "epoch": 0.8464007336084365, "grad_norm": 0.4919211268424988, "learning_rate": 6.063574906968511e-05, "loss": 2.3241, "step": 1846 }, { "epoch": 0.8468592388812471, "grad_norm": 0.517510712146759, "learning_rate": 6.028173010576582e-05, "loss": 2.2799, "step": 1847 }, { "epoch": 0.8473177441540578, "grad_norm": 0.5019983649253845, "learning_rate": 5.9928681324337544e-05, "loss": 2.4, "step": 1848 }, { "epoch": 0.8477762494268684, "grad_norm": 0.4892749786376953, "learning_rate": 5.957660350435773e-05, "loss": 2.324, "step": 1849 }, { "epoch": 0.848234754699679, "grad_norm": 0.4781142771244049, "learning_rate": 5.922549742264122e-05, "loss": 2.3708, "step": 1850 }, { "epoch": 0.8486932599724897, "grad_norm": 0.46425941586494446, "learning_rate": 5.8875363853859166e-05, "loss": 2.3446, "step": 1851 }, { "epoch": 0.8491517652453003, "grad_norm": 0.46367934346199036, "learning_rate": 5.852620357053651e-05, "loss": 2.4269, "step": 1852 }, { "epoch": 0.849610270518111, "grad_norm": 0.49848610162734985, "learning_rate": 5.8178017343051336e-05, "loss": 2.2976, "step": 1853 }, { "epoch": 0.8500687757909215, "grad_norm": 0.4877791106700897, "learning_rate": 5.783080593963219e-05, "loss": 2.3662, "step": 1854 }, { "epoch": 0.8505272810637322, "grad_norm": 0.4964427053928375, "learning_rate": 5.748457012635683e-05, "loss": 2.3948, "step": 1855 }, { "epoch": 0.8509857863365429, "grad_norm": 0.44984158873558044, "learning_rate": 5.713931066715078e-05, "loss": 2.2547, "step": 1856 }, { "epoch": 0.8514442916093535, "grad_norm": 0.44088122248649597, "learning_rate": 5.679502832378497e-05, "loss": 2.4265, "step": 1857 }, { "epoch": 0.8519027968821642, "grad_norm": 0.4668961763381958, "learning_rate": 5.645172385587482e-05, "loss": 2.2956, "step": 1858 }, { "epoch": 0.8523613021549747, "grad_norm": 0.4477249085903168, "learning_rate": 5.6109398020877834e-05, "loss": 2.2758, "step": 1859 }, { "epoch": 0.8528198074277854, "grad_norm": 0.48009973764419556, "learning_rate": 5.576805157409265e-05, "loss": 2.3286, "step": 1860 }, { "epoch": 0.8532783127005961, "grad_norm": 0.4786818325519562, "learning_rate": 5.542768526865677e-05, "loss": 2.3257, "step": 1861 }, { "epoch": 0.8537368179734067, "grad_norm": 0.493378609418869, "learning_rate": 5.508829985554509e-05, "loss": 2.4363, "step": 1862 }, { "epoch": 0.8541953232462174, "grad_norm": 0.47787952423095703, "learning_rate": 5.474989608356856e-05, "loss": 2.4303, "step": 1863 }, { "epoch": 0.8546538285190279, "grad_norm": 0.44499900937080383, "learning_rate": 5.441247469937194e-05, "loss": 2.298, "step": 1864 }, { "epoch": 0.8551123337918386, "grad_norm": 0.4724040627479553, "learning_rate": 5.407603644743286e-05, "loss": 2.2943, "step": 1865 }, { "epoch": 0.8555708390646493, "grad_norm": 0.4571748375892639, "learning_rate": 5.374058207005944e-05, "loss": 2.2418, "step": 1866 }, { "epoch": 0.8560293443374599, "grad_norm": 0.478316068649292, "learning_rate": 5.3406112307389066e-05, "loss": 2.3526, "step": 1867 }, { "epoch": 0.8564878496102705, "grad_norm": 0.47959041595458984, "learning_rate": 5.3072627897386926e-05, "loss": 2.387, "step": 1868 }, { "epoch": 0.8569463548830811, "grad_norm": 0.48476940393447876, "learning_rate": 5.27401295758439e-05, "loss": 2.3475, "step": 1869 }, { "epoch": 0.8574048601558918, "grad_norm": 0.4861583411693573, "learning_rate": 5.2408618076375315e-05, "loss": 2.4762, "step": 1870 }, { "epoch": 0.8578633654287025, "grad_norm": 0.46123820543289185, "learning_rate": 5.207809413041914e-05, "loss": 2.4509, "step": 1871 }, { "epoch": 0.8583218707015131, "grad_norm": 0.47313305735588074, "learning_rate": 5.174855846723459e-05, "loss": 2.3092, "step": 1872 }, { "epoch": 0.8587803759743237, "grad_norm": 0.4784107208251953, "learning_rate": 5.1420011813900104e-05, "loss": 2.3721, "step": 1873 }, { "epoch": 0.8592388812471343, "grad_norm": 0.4807412624359131, "learning_rate": 5.109245489531211e-05, "loss": 2.3602, "step": 1874 }, { "epoch": 0.859697386519945, "grad_norm": 0.4669828414916992, "learning_rate": 5.0765888434183446e-05, "loss": 2.2598, "step": 1875 }, { "epoch": 0.8601558917927556, "grad_norm": 0.47123876214027405, "learning_rate": 5.0440313151041364e-05, "loss": 2.3947, "step": 1876 }, { "epoch": 0.8606143970655663, "grad_norm": 0.46684589982032776, "learning_rate": 5.011572976422657e-05, "loss": 2.3126, "step": 1877 }, { "epoch": 0.8610729023383769, "grad_norm": 0.47263461351394653, "learning_rate": 4.9792138989890825e-05, "loss": 2.3539, "step": 1878 }, { "epoch": 0.8615314076111875, "grad_norm": 0.48494839668273926, "learning_rate": 4.9469541541996234e-05, "loss": 2.3295, "step": 1879 }, { "epoch": 0.8619899128839982, "grad_norm": 0.456073135137558, "learning_rate": 4.914793813231305e-05, "loss": 2.3892, "step": 1880 }, { "epoch": 0.8624484181568088, "grad_norm": 0.49641165137290955, "learning_rate": 4.882732947041818e-05, "loss": 2.3973, "step": 1881 }, { "epoch": 0.8629069234296194, "grad_norm": 0.47957807779312134, "learning_rate": 4.850771626369416e-05, "loss": 2.3905, "step": 1882 }, { "epoch": 0.86336542870243, "grad_norm": 0.49232178926467896, "learning_rate": 4.818909921732662e-05, "loss": 2.3508, "step": 1883 }, { "epoch": 0.8638239339752407, "grad_norm": 0.48334717750549316, "learning_rate": 4.787147903430383e-05, "loss": 2.324, "step": 1884 }, { "epoch": 0.8642824392480514, "grad_norm": 0.47511133551597595, "learning_rate": 4.755485641541424e-05, "loss": 2.3462, "step": 1885 }, { "epoch": 0.864740944520862, "grad_norm": 0.4969499409198761, "learning_rate": 4.723923205924557e-05, "loss": 2.3317, "step": 1886 }, { "epoch": 0.8651994497936726, "grad_norm": 0.47647926211357117, "learning_rate": 4.6924606662182736e-05, "loss": 2.4401, "step": 1887 }, { "epoch": 0.8656579550664832, "grad_norm": 0.4681853950023651, "learning_rate": 4.6610980918406596e-05, "loss": 2.3435, "step": 1888 }, { "epoch": 0.8661164603392939, "grad_norm": 0.47125011682510376, "learning_rate": 4.629835551989276e-05, "loss": 2.4018, "step": 1889 }, { "epoch": 0.8665749656121046, "grad_norm": 0.47393670678138733, "learning_rate": 4.5986731156409224e-05, "loss": 2.3616, "step": 1890 }, { "epoch": 0.8670334708849152, "grad_norm": 0.490223228931427, "learning_rate": 4.567610851551568e-05, "loss": 2.4863, "step": 1891 }, { "epoch": 0.8674919761577258, "grad_norm": 0.46028193831443787, "learning_rate": 4.536648828256146e-05, "loss": 2.3433, "step": 1892 }, { "epoch": 0.8679504814305364, "grad_norm": 0.4749680459499359, "learning_rate": 4.505787114068433e-05, "loss": 2.3522, "step": 1893 }, { "epoch": 0.8684089867033471, "grad_norm": 0.473898321390152, "learning_rate": 4.4750257770808764e-05, "loss": 2.4078, "step": 1894 }, { "epoch": 0.8688674919761578, "grad_norm": 0.4911269545555115, "learning_rate": 4.444364885164448e-05, "loss": 2.3364, "step": 1895 }, { "epoch": 0.8693259972489684, "grad_norm": 0.48834091424942017, "learning_rate": 4.413804505968533e-05, "loss": 2.3812, "step": 1896 }, { "epoch": 0.869784502521779, "grad_norm": 0.48702865839004517, "learning_rate": 4.3833447069206944e-05, "loss": 2.4674, "step": 1897 }, { "epoch": 0.8702430077945896, "grad_norm": 0.5065435767173767, "learning_rate": 4.352985555226635e-05, "loss": 2.4098, "step": 1898 }, { "epoch": 0.8707015130674003, "grad_norm": 0.4655471444129944, "learning_rate": 4.322727117869951e-05, "loss": 2.3338, "step": 1899 }, { "epoch": 0.871160018340211, "grad_norm": 0.45795220136642456, "learning_rate": 4.29256946161205e-05, "loss": 2.3154, "step": 1900 }, { "epoch": 0.8716185236130215, "grad_norm": 0.4808029532432556, "learning_rate": 4.262512652991968e-05, "loss": 2.2821, "step": 1901 }, { "epoch": 0.8720770288858322, "grad_norm": 0.4701562523841858, "learning_rate": 4.2325567583262113e-05, "loss": 2.3175, "step": 1902 }, { "epoch": 0.8725355341586428, "grad_norm": 0.49831125140190125, "learning_rate": 4.2027018437086895e-05, "loss": 2.3381, "step": 1903 }, { "epoch": 0.8729940394314535, "grad_norm": 0.4758774936199188, "learning_rate": 4.172947975010449e-05, "loss": 2.2934, "step": 1904 }, { "epoch": 0.8734525447042641, "grad_norm": 0.5021149516105652, "learning_rate": 4.143295217879645e-05, "loss": 2.4847, "step": 1905 }, { "epoch": 0.8739110499770747, "grad_norm": 0.48954278230667114, "learning_rate": 4.113743637741296e-05, "loss": 2.3532, "step": 1906 }, { "epoch": 0.8743695552498854, "grad_norm": 0.4654124677181244, "learning_rate": 4.084293299797226e-05, "loss": 2.3821, "step": 1907 }, { "epoch": 0.874828060522696, "grad_norm": 0.4871256649494171, "learning_rate": 4.054944269025862e-05, "loss": 2.2988, "step": 1908 }, { "epoch": 0.8752865657955067, "grad_norm": 0.49786219000816345, "learning_rate": 4.025696610182095e-05, "loss": 2.3406, "step": 1909 }, { "epoch": 0.8757450710683173, "grad_norm": 0.4974231421947479, "learning_rate": 3.996550387797187e-05, "loss": 2.2789, "step": 1910 }, { "epoch": 0.8762035763411279, "grad_norm": 0.4873538017272949, "learning_rate": 3.9675056661785556e-05, "loss": 2.2528, "step": 1911 }, { "epoch": 0.8766620816139385, "grad_norm": 0.4641670882701874, "learning_rate": 3.9385625094097154e-05, "loss": 2.4131, "step": 1912 }, { "epoch": 0.8771205868867492, "grad_norm": 0.46530961990356445, "learning_rate": 3.909720981350034e-05, "loss": 2.2619, "step": 1913 }, { "epoch": 0.8775790921595599, "grad_norm": 0.4874619245529175, "learning_rate": 3.880981145634704e-05, "loss": 2.4263, "step": 1914 }, { "epoch": 0.8780375974323704, "grad_norm": 0.5033944845199585, "learning_rate": 3.852343065674507e-05, "loss": 2.3895, "step": 1915 }, { "epoch": 0.8784961027051811, "grad_norm": 0.4707699120044708, "learning_rate": 3.8238068046557276e-05, "loss": 2.3368, "step": 1916 }, { "epoch": 0.8789546079779917, "grad_norm": 0.48432862758636475, "learning_rate": 3.795372425540006e-05, "loss": 2.3038, "step": 1917 }, { "epoch": 0.8794131132508024, "grad_norm": 0.4839876890182495, "learning_rate": 3.76703999106418e-05, "loss": 2.2991, "step": 1918 }, { "epoch": 0.8798716185236131, "grad_norm": 0.45670172572135925, "learning_rate": 3.7388095637401754e-05, "loss": 2.2613, "step": 1919 }, { "epoch": 0.8803301237964236, "grad_norm": 0.4480465352535248, "learning_rate": 3.7106812058548376e-05, "loss": 2.3455, "step": 1920 }, { "epoch": 0.8807886290692343, "grad_norm": 0.4902710020542145, "learning_rate": 3.682654979469807e-05, "loss": 2.3629, "step": 1921 }, { "epoch": 0.8812471343420449, "grad_norm": 0.4531223177909851, "learning_rate": 3.654730946421403e-05, "loss": 2.4105, "step": 1922 }, { "epoch": 0.8817056396148556, "grad_norm": 0.47030436992645264, "learning_rate": 3.6269091683204466e-05, "loss": 2.3729, "step": 1923 }, { "epoch": 0.8821641448876663, "grad_norm": 0.45522019267082214, "learning_rate": 3.5991897065521693e-05, "loss": 2.3972, "step": 1924 }, { "epoch": 0.8826226501604768, "grad_norm": 0.48063862323760986, "learning_rate": 3.571572622276026e-05, "loss": 2.3854, "step": 1925 }, { "epoch": 0.8830811554332875, "grad_norm": 0.5169646739959717, "learning_rate": 3.544057976425619e-05, "loss": 2.4861, "step": 1926 }, { "epoch": 0.8835396607060981, "grad_norm": 0.4662131369113922, "learning_rate": 3.5166458297085146e-05, "loss": 2.4328, "step": 1927 }, { "epoch": 0.8839981659789088, "grad_norm": 0.4809960722923279, "learning_rate": 3.489336242606111e-05, "loss": 2.3651, "step": 1928 }, { "epoch": 0.8844566712517193, "grad_norm": 0.45175400376319885, "learning_rate": 3.462129275373577e-05, "loss": 2.3362, "step": 1929 }, { "epoch": 0.88491517652453, "grad_norm": 0.45470207929611206, "learning_rate": 3.4350249880395924e-05, "loss": 2.4086, "step": 1930 }, { "epoch": 0.8853736817973407, "grad_norm": 0.5061991810798645, "learning_rate": 3.408023440406355e-05, "loss": 2.3365, "step": 1931 }, { "epoch": 0.8858321870701513, "grad_norm": 0.4702357351779938, "learning_rate": 3.381124692049331e-05, "loss": 2.4354, "step": 1932 }, { "epoch": 0.886290692342962, "grad_norm": 0.4478657841682434, "learning_rate": 3.354328802317197e-05, "loss": 2.324, "step": 1933 }, { "epoch": 0.8867491976157725, "grad_norm": 0.48024627566337585, "learning_rate": 3.327635830331677e-05, "loss": 2.2865, "step": 1934 }, { "epoch": 0.8872077028885832, "grad_norm": 0.46708500385284424, "learning_rate": 3.3010458349874206e-05, "loss": 2.345, "step": 1935 }, { "epoch": 0.8876662081613939, "grad_norm": 0.46397748589515686, "learning_rate": 3.2745588749518775e-05, "loss": 2.4169, "step": 1936 }, { "epoch": 0.8881247134342045, "grad_norm": 0.4819386601448059, "learning_rate": 3.248175008665161e-05, "loss": 2.3297, "step": 1937 }, { "epoch": 0.8885832187070152, "grad_norm": 0.48014214634895325, "learning_rate": 3.221894294339911e-05, "loss": 2.3714, "step": 1938 }, { "epoch": 0.8890417239798257, "grad_norm": 0.4755948483943939, "learning_rate": 3.1957167899611836e-05, "loss": 2.3141, "step": 1939 }, { "epoch": 0.8895002292526364, "grad_norm": 0.48372912406921387, "learning_rate": 3.169642553286334e-05, "loss": 2.4048, "step": 1940 }, { "epoch": 0.889958734525447, "grad_norm": 0.4716033339500427, "learning_rate": 3.143671641844831e-05, "loss": 2.3678, "step": 1941 }, { "epoch": 0.8904172397982577, "grad_norm": 0.48180660605430603, "learning_rate": 3.117804112938205e-05, "loss": 2.2784, "step": 1942 }, { "epoch": 0.8908757450710684, "grad_norm": 0.4782445728778839, "learning_rate": 3.092040023639869e-05, "loss": 2.4428, "step": 1943 }, { "epoch": 0.8913342503438789, "grad_norm": 0.4744698405265808, "learning_rate": 3.066379430795002e-05, "loss": 2.3702, "step": 1944 }, { "epoch": 0.8917927556166896, "grad_norm": 0.47320812940597534, "learning_rate": 3.040822391020459e-05, "loss": 2.3847, "step": 1945 }, { "epoch": 0.8922512608895002, "grad_norm": 0.46369266510009766, "learning_rate": 3.0153689607045842e-05, "loss": 2.3085, "step": 1946 }, { "epoch": 0.8927097661623109, "grad_norm": 0.4676467776298523, "learning_rate": 2.9900191960071545e-05, "loss": 2.3229, "step": 1947 }, { "epoch": 0.8931682714351215, "grad_norm": 0.4937705397605896, "learning_rate": 2.9647731528591848e-05, "loss": 2.3872, "step": 1948 }, { "epoch": 0.8936267767079321, "grad_norm": 0.46484488248825073, "learning_rate": 2.9396308869628795e-05, "loss": 2.3632, "step": 1949 }, { "epoch": 0.8940852819807428, "grad_norm": 0.49699831008911133, "learning_rate": 2.914592453791448e-05, "loss": 2.3519, "step": 1950 }, { "epoch": 0.8945437872535534, "grad_norm": 0.48416319489479065, "learning_rate": 2.8896579085889994e-05, "loss": 2.3928, "step": 1951 }, { "epoch": 0.8950022925263641, "grad_norm": 0.4549061357975006, "learning_rate": 2.86482730637046e-05, "loss": 2.3479, "step": 1952 }, { "epoch": 0.8954607977991746, "grad_norm": 0.47029629349708557, "learning_rate": 2.840100701921383e-05, "loss": 2.4225, "step": 1953 }, { "epoch": 0.8959193030719853, "grad_norm": 0.4657432436943054, "learning_rate": 2.8154781497978898e-05, "loss": 2.3911, "step": 1954 }, { "epoch": 0.896377808344796, "grad_norm": 0.4901619255542755, "learning_rate": 2.7909597043265013e-05, "loss": 2.3641, "step": 1955 }, { "epoch": 0.8968363136176066, "grad_norm": 0.47901731729507446, "learning_rate": 2.7665454196040662e-05, "loss": 2.3826, "step": 1956 }, { "epoch": 0.8972948188904173, "grad_norm": 0.4648403823375702, "learning_rate": 2.7422353494975905e-05, "loss": 2.3626, "step": 1957 }, { "epoch": 0.8977533241632278, "grad_norm": 0.48162761330604553, "learning_rate": 2.7180295476441573e-05, "loss": 2.3483, "step": 1958 }, { "epoch": 0.8982118294360385, "grad_norm": 0.4870343506336212, "learning_rate": 2.6939280674508016e-05, "loss": 2.3609, "step": 1959 }, { "epoch": 0.8986703347088492, "grad_norm": 0.4528430104255676, "learning_rate": 2.669930962094358e-05, "loss": 2.3613, "step": 1960 }, { "epoch": 0.8991288399816598, "grad_norm": 0.4723399877548218, "learning_rate": 2.6460382845214126e-05, "loss": 2.4474, "step": 1961 }, { "epoch": 0.8995873452544704, "grad_norm": 0.46861281991004944, "learning_rate": 2.6222500874481025e-05, "loss": 2.4065, "step": 1962 }, { "epoch": 0.900045850527281, "grad_norm": 0.4868358075618744, "learning_rate": 2.5985664233600827e-05, "loss": 2.3844, "step": 1963 }, { "epoch": 0.9005043558000917, "grad_norm": 0.4830412268638611, "learning_rate": 2.574987344512336e-05, "loss": 2.3525, "step": 1964 }, { "epoch": 0.9009628610729024, "grad_norm": 0.48201480507850647, "learning_rate": 2.5515129029290984e-05, "loss": 2.3682, "step": 1965 }, { "epoch": 0.901421366345713, "grad_norm": 0.5042434334754944, "learning_rate": 2.5281431504037556e-05, "loss": 2.3296, "step": 1966 }, { "epoch": 0.9018798716185236, "grad_norm": 0.4726353585720062, "learning_rate": 2.504878138498684e-05, "loss": 2.3079, "step": 1967 }, { "epoch": 0.9023383768913342, "grad_norm": 0.47002536058425903, "learning_rate": 2.48171791854519e-05, "loss": 2.418, "step": 1968 }, { "epoch": 0.9027968821641449, "grad_norm": 0.473162442445755, "learning_rate": 2.4586625416433473e-05, "loss": 2.366, "step": 1969 }, { "epoch": 0.9032553874369555, "grad_norm": 0.464281290769577, "learning_rate": 2.435712058661921e-05, "loss": 2.3685, "step": 1970 }, { "epoch": 0.9037138927097662, "grad_norm": 0.4963575601577759, "learning_rate": 2.4128665202382327e-05, "loss": 2.3245, "step": 1971 }, { "epoch": 0.9041723979825768, "grad_norm": 0.46388179063796997, "learning_rate": 2.3901259767780515e-05, "loss": 2.3351, "step": 1972 }, { "epoch": 0.9046309032553874, "grad_norm": 0.4835372865200043, "learning_rate": 2.367490478455514e-05, "loss": 2.448, "step": 1973 }, { "epoch": 0.9050894085281981, "grad_norm": 0.46585556864738464, "learning_rate": 2.3449600752129597e-05, "loss": 2.352, "step": 1974 }, { "epoch": 0.9055479138010087, "grad_norm": 0.47851839661598206, "learning_rate": 2.3225348167608685e-05, "loss": 2.3227, "step": 1975 }, { "epoch": 0.9060064190738194, "grad_norm": 0.47982197999954224, "learning_rate": 2.3002147525777118e-05, "loss": 2.4497, "step": 1976 }, { "epoch": 0.90646492434663, "grad_norm": 0.49924060702323914, "learning_rate": 2.2779999319098856e-05, "loss": 2.3387, "step": 1977 }, { "epoch": 0.9069234296194406, "grad_norm": 0.4782711863517761, "learning_rate": 2.255890403771571e-05, "loss": 2.346, "step": 1978 }, { "epoch": 0.9073819348922513, "grad_norm": 0.463890016078949, "learning_rate": 2.233886216944614e-05, "loss": 2.2717, "step": 1979 }, { "epoch": 0.9078404401650619, "grad_norm": 0.4686962366104126, "learning_rate": 2.211987419978484e-05, "loss": 2.423, "step": 1980 }, { "epoch": 0.9082989454378725, "grad_norm": 0.4939550459384918, "learning_rate": 2.1901940611900705e-05, "loss": 2.3895, "step": 1981 }, { "epoch": 0.9087574507106831, "grad_norm": 0.47162336111068726, "learning_rate": 2.168506188663666e-05, "loss": 2.391, "step": 1982 }, { "epoch": 0.9092159559834938, "grad_norm": 0.46975088119506836, "learning_rate": 2.1469238502507925e-05, "loss": 2.3569, "step": 1983 }, { "epoch": 0.9096744612563045, "grad_norm": 0.455267071723938, "learning_rate": 2.125447093570154e-05, "loss": 2.3214, "step": 1984 }, { "epoch": 0.9101329665291151, "grad_norm": 0.478328138589859, "learning_rate": 2.1040759660074793e-05, "loss": 2.3643, "step": 1985 }, { "epoch": 0.9105914718019257, "grad_norm": 0.49209243059158325, "learning_rate": 2.0828105147154273e-05, "loss": 2.3636, "step": 1986 }, { "epoch": 0.9110499770747363, "grad_norm": 0.4820910692214966, "learning_rate": 2.061650786613545e-05, "loss": 2.3721, "step": 1987 }, { "epoch": 0.911508482347547, "grad_norm": 0.516726016998291, "learning_rate": 2.040596828388058e-05, "loss": 2.3686, "step": 1988 }, { "epoch": 0.9119669876203577, "grad_norm": 0.4976004362106323, "learning_rate": 2.019648686491865e-05, "loss": 2.4055, "step": 1989 }, { "epoch": 0.9124254928931683, "grad_norm": 0.4833821654319763, "learning_rate": 1.9988064071443767e-05, "loss": 2.2983, "step": 1990 }, { "epoch": 0.9128839981659789, "grad_norm": 0.48727431893348694, "learning_rate": 1.9780700363314253e-05, "loss": 2.376, "step": 1991 }, { "epoch": 0.9133425034387895, "grad_norm": 0.45935913920402527, "learning_rate": 1.957439619805196e-05, "loss": 2.4051, "step": 1992 }, { "epoch": 0.9138010087116002, "grad_norm": 0.4744124710559845, "learning_rate": 1.9369152030840554e-05, "loss": 2.2662, "step": 1993 }, { "epoch": 0.9142595139844109, "grad_norm": 0.479186475276947, "learning_rate": 1.916496831452552e-05, "loss": 2.4016, "step": 1994 }, { "epoch": 0.9147180192572214, "grad_norm": 0.49834144115448, "learning_rate": 1.8961845499611998e-05, "loss": 2.239, "step": 1995 }, { "epoch": 0.9151765245300321, "grad_norm": 0.4834553003311157, "learning_rate": 1.8759784034264925e-05, "loss": 2.4194, "step": 1996 }, { "epoch": 0.9156350298028427, "grad_norm": 0.47225508093833923, "learning_rate": 1.855878436430708e-05, "loss": 2.4505, "step": 1997 }, { "epoch": 0.9160935350756534, "grad_norm": 0.46168768405914307, "learning_rate": 1.835884693321871e-05, "loss": 2.3338, "step": 1998 }, { "epoch": 0.916552040348464, "grad_norm": 0.47755667567253113, "learning_rate": 1.8159972182136386e-05, "loss": 2.4539, "step": 1999 }, { "epoch": 0.9170105456212746, "grad_norm": 0.4648977518081665, "learning_rate": 1.7962160549851945e-05, "loss": 2.3264, "step": 2000 }, { "epoch": 0.9174690508940853, "grad_norm": 0.4774320423603058, "learning_rate": 1.776541247281177e-05, "loss": 2.4589, "step": 2001 }, { "epoch": 0.9179275561668959, "grad_norm": 0.45641762018203735, "learning_rate": 1.7569728385115224e-05, "loss": 2.2914, "step": 2002 }, { "epoch": 0.9183860614397066, "grad_norm": 0.4717657268047333, "learning_rate": 1.7375108718514665e-05, "loss": 2.3102, "step": 2003 }, { "epoch": 0.9188445667125172, "grad_norm": 0.47568273544311523, "learning_rate": 1.7181553902413438e-05, "loss": 2.3939, "step": 2004 }, { "epoch": 0.9193030719853278, "grad_norm": 0.4740332067012787, "learning_rate": 1.698906436386577e-05, "loss": 2.3424, "step": 2005 }, { "epoch": 0.9197615772581385, "grad_norm": 0.47005197405815125, "learning_rate": 1.679764052757532e-05, "loss": 2.452, "step": 2006 }, { "epoch": 0.9202200825309491, "grad_norm": 0.46999478340148926, "learning_rate": 1.6607282815894464e-05, "loss": 2.3183, "step": 2007 }, { "epoch": 0.9206785878037598, "grad_norm": 0.4675203263759613, "learning_rate": 1.6417991648823405e-05, "loss": 2.3188, "step": 2008 }, { "epoch": 0.9211370930765704, "grad_norm": 0.4566081166267395, "learning_rate": 1.6229767444008835e-05, "loss": 2.3338, "step": 2009 }, { "epoch": 0.921595598349381, "grad_norm": 0.4811600148677826, "learning_rate": 1.604261061674378e-05, "loss": 2.4468, "step": 2010 }, { "epoch": 0.9220541036221916, "grad_norm": 0.48928824067115784, "learning_rate": 1.5856521579965865e-05, "loss": 2.3145, "step": 2011 }, { "epoch": 0.9225126088950023, "grad_norm": 0.4850217401981354, "learning_rate": 1.5671500744256938e-05, "loss": 2.3988, "step": 2012 }, { "epoch": 0.922971114167813, "grad_norm": 0.48682504892349243, "learning_rate": 1.5487548517841953e-05, "loss": 2.3188, "step": 2013 }, { "epoch": 0.9234296194406235, "grad_norm": 0.4918004274368286, "learning_rate": 1.530466530658814e-05, "loss": 2.4216, "step": 2014 }, { "epoch": 0.9238881247134342, "grad_norm": 0.5126110315322876, "learning_rate": 1.5122851514004054e-05, "loss": 2.4669, "step": 2015 }, { "epoch": 0.9243466299862448, "grad_norm": 0.5169333219528198, "learning_rate": 1.4942107541238703e-05, "loss": 2.3953, "step": 2016 }, { "epoch": 0.9248051352590555, "grad_norm": 0.46691831946372986, "learning_rate": 1.4762433787080809e-05, "loss": 2.2847, "step": 2017 }, { "epoch": 0.9252636405318662, "grad_norm": 0.5075626373291016, "learning_rate": 1.4583830647957541e-05, "loss": 2.4404, "step": 2018 }, { "epoch": 0.9257221458046767, "grad_norm": 0.4711140990257263, "learning_rate": 1.4406298517934068e-05, "loss": 2.2292, "step": 2019 }, { "epoch": 0.9261806510774874, "grad_norm": 0.47954049706459045, "learning_rate": 1.4229837788712562e-05, "loss": 2.3033, "step": 2020 }, { "epoch": 0.926639156350298, "grad_norm": 0.46203911304473877, "learning_rate": 1.4054448849631085e-05, "loss": 2.4063, "step": 2021 }, { "epoch": 0.9270976616231087, "grad_norm": 0.4905766546726227, "learning_rate": 1.3880132087663145e-05, "loss": 2.3194, "step": 2022 }, { "epoch": 0.9275561668959194, "grad_norm": 0.45643240213394165, "learning_rate": 1.3706887887416419e-05, "loss": 2.3448, "step": 2023 }, { "epoch": 0.9280146721687299, "grad_norm": 0.4675247371196747, "learning_rate": 1.3534716631132316e-05, "loss": 2.2802, "step": 2024 }, { "epoch": 0.9284731774415406, "grad_norm": 0.4858630895614624, "learning_rate": 1.3363618698684853e-05, "loss": 2.3127, "step": 2025 }, { "epoch": 0.9289316827143512, "grad_norm": 0.4426301419734955, "learning_rate": 1.3193594467579728e-05, "loss": 2.3426, "step": 2026 }, { "epoch": 0.9293901879871619, "grad_norm": 0.4646582305431366, "learning_rate": 1.3024644312954026e-05, "loss": 2.3387, "step": 2027 }, { "epoch": 0.9298486932599724, "grad_norm": 0.4898240268230438, "learning_rate": 1.2856768607574564e-05, "loss": 2.3483, "step": 2028 }, { "epoch": 0.9303071985327831, "grad_norm": 0.4754076898097992, "learning_rate": 1.2689967721837947e-05, "loss": 2.4057, "step": 2029 }, { "epoch": 0.9307657038055938, "grad_norm": 0.503186821937561, "learning_rate": 1.2524242023769006e-05, "loss": 2.455, "step": 2030 }, { "epoch": 0.9312242090784044, "grad_norm": 0.4813970923423767, "learning_rate": 1.2359591879020526e-05, "loss": 2.4104, "step": 2031 }, { "epoch": 0.9316827143512151, "grad_norm": 0.4884364902973175, "learning_rate": 1.2196017650872081e-05, "loss": 2.371, "step": 2032 }, { "epoch": 0.9321412196240256, "grad_norm": 0.4841212034225464, "learning_rate": 1.2033519700229367e-05, "loss": 2.2649, "step": 2033 }, { "epoch": 0.9325997248968363, "grad_norm": 0.48509421944618225, "learning_rate": 1.1872098385623586e-05, "loss": 2.4238, "step": 2034 }, { "epoch": 0.933058230169647, "grad_norm": 0.4741576910018921, "learning_rate": 1.1711754063210289e-05, "loss": 2.383, "step": 2035 }, { "epoch": 0.9335167354424576, "grad_norm": 0.4869562089443207, "learning_rate": 1.155248708676887e-05, "loss": 2.3211, "step": 2036 }, { "epoch": 0.9339752407152683, "grad_norm": 0.4907585382461548, "learning_rate": 1.1394297807701737e-05, "loss": 2.3389, "step": 2037 }, { "epoch": 0.9344337459880788, "grad_norm": 0.4715602397918701, "learning_rate": 1.1237186575033254e-05, "loss": 2.4203, "step": 2038 }, { "epoch": 0.9348922512608895, "grad_norm": 0.4907388985157013, "learning_rate": 1.1081153735409522e-05, "loss": 2.3357, "step": 2039 }, { "epoch": 0.9353507565337001, "grad_norm": 0.4925310015678406, "learning_rate": 1.0926199633097156e-05, "loss": 2.3981, "step": 2040 }, { "epoch": 0.9358092618065108, "grad_norm": 0.4800892472267151, "learning_rate": 1.0772324609982787e-05, "loss": 2.3717, "step": 2041 }, { "epoch": 0.9362677670793215, "grad_norm": 0.48686203360557556, "learning_rate": 1.0619529005571893e-05, "loss": 2.4121, "step": 2042 }, { "epoch": 0.936726272352132, "grad_norm": 0.5056031942367554, "learning_rate": 1.0467813156988748e-05, "loss": 2.4305, "step": 2043 }, { "epoch": 0.9371847776249427, "grad_norm": 0.45925596356391907, "learning_rate": 1.0317177398975031e-05, "loss": 2.366, "step": 2044 }, { "epoch": 0.9376432828977533, "grad_norm": 0.4684313237667084, "learning_rate": 1.0167622063889326e-05, "loss": 2.2825, "step": 2045 }, { "epoch": 0.938101788170564, "grad_norm": 0.47126510739326477, "learning_rate": 1.0019147481706625e-05, "loss": 2.3525, "step": 2046 }, { "epoch": 0.9385602934433745, "grad_norm": 0.4954504370689392, "learning_rate": 9.871753980017051e-06, "loss": 2.3489, "step": 2047 }, { "epoch": 0.9390187987161852, "grad_norm": 0.4975089430809021, "learning_rate": 9.725441884025855e-06, "loss": 2.4198, "step": 2048 }, { "epoch": 0.9394773039889959, "grad_norm": 0.47014597058296204, "learning_rate": 9.580211516551862e-06, "loss": 2.3305, "step": 2049 }, { "epoch": 0.9399358092618065, "grad_norm": 0.4635334312915802, "learning_rate": 9.436063198027589e-06, "loss": 2.3654, "step": 2050 }, { "epoch": 0.9403943145346172, "grad_norm": 0.4737807512283325, "learning_rate": 9.292997246497959e-06, "loss": 2.3878, "step": 2051 }, { "epoch": 0.9408528198074277, "grad_norm": 0.4857623279094696, "learning_rate": 9.151013977619693e-06, "loss": 2.2445, "step": 2052 }, { "epoch": 0.9413113250802384, "grad_norm": 0.464293509721756, "learning_rate": 9.010113704661038e-06, "loss": 2.3035, "step": 2053 }, { "epoch": 0.9417698303530491, "grad_norm": 0.46866562962532043, "learning_rate": 8.870296738500316e-06, "loss": 2.3518, "step": 2054 }, { "epoch": 0.9422283356258597, "grad_norm": 0.46806198358535767, "learning_rate": 8.731563387626095e-06, "loss": 2.3846, "step": 2055 }, { "epoch": 0.9426868408986704, "grad_norm": 0.4767438769340515, "learning_rate": 8.59391395813569e-06, "loss": 2.3318, "step": 2056 }, { "epoch": 0.9431453461714809, "grad_norm": 0.4628843665122986, "learning_rate": 8.457348753735328e-06, "loss": 2.3338, "step": 2057 }, { "epoch": 0.9436038514442916, "grad_norm": 0.46587151288986206, "learning_rate": 8.321868075738593e-06, "loss": 2.3977, "step": 2058 }, { "epoch": 0.9440623567171023, "grad_norm": 0.4854567050933838, "learning_rate": 8.187472223066371e-06, "loss": 2.3708, "step": 2059 }, { "epoch": 0.9445208619899129, "grad_norm": 0.4826723039150238, "learning_rate": 8.054161492246136e-06, "loss": 2.3355, "step": 2060 }, { "epoch": 0.9449793672627235, "grad_norm": 0.48855116963386536, "learning_rate": 7.921936177411049e-06, "loss": 2.4133, "step": 2061 }, { "epoch": 0.9454378725355341, "grad_norm": 0.4957791566848755, "learning_rate": 7.790796570299463e-06, "loss": 2.3543, "step": 2062 }, { "epoch": 0.9458963778083448, "grad_norm": 0.47234705090522766, "learning_rate": 7.660742960254207e-06, "loss": 2.3812, "step": 2063 }, { "epoch": 0.9463548830811555, "grad_norm": 0.4760810434818268, "learning_rate": 7.531775634222138e-06, "loss": 2.3332, "step": 2064 }, { "epoch": 0.9468133883539661, "grad_norm": 0.4722563624382019, "learning_rate": 7.403894876753192e-06, "loss": 2.305, "step": 2065 }, { "epoch": 0.9472718936267767, "grad_norm": 0.47263360023498535, "learning_rate": 7.277100970000061e-06, "loss": 2.3191, "step": 2066 }, { "epoch": 0.9477303988995873, "grad_norm": 0.4748469293117523, "learning_rate": 7.151394193717408e-06, "loss": 2.3564, "step": 2067 }, { "epoch": 0.948188904172398, "grad_norm": 0.4817005693912506, "learning_rate": 7.026774825261151e-06, "loss": 2.3989, "step": 2068 }, { "epoch": 0.9486474094452086, "grad_norm": 0.4597158432006836, "learning_rate": 6.903243139588233e-06, "loss": 2.3044, "step": 2069 }, { "epoch": 0.9491059147180193, "grad_norm": 0.4780486226081848, "learning_rate": 6.780799409255522e-06, "loss": 2.3947, "step": 2070 }, { "epoch": 0.9495644199908299, "grad_norm": 0.47961628437042236, "learning_rate": 6.659443904419637e-06, "loss": 2.4927, "step": 2071 }, { "epoch": 0.9500229252636405, "grad_norm": 0.5020300149917603, "learning_rate": 6.539176892836008e-06, "loss": 2.2911, "step": 2072 }, { "epoch": 0.9504814305364512, "grad_norm": 0.46948087215423584, "learning_rate": 6.4199986398585375e-06, "loss": 2.4536, "step": 2073 }, { "epoch": 0.9509399358092618, "grad_norm": 0.4876088798046112, "learning_rate": 6.3019094084388884e-06, "loss": 2.4013, "step": 2074 }, { "epoch": 0.9513984410820725, "grad_norm": 0.49027249217033386, "learning_rate": 6.18490945912592e-06, "loss": 2.3119, "step": 2075 }, { "epoch": 0.951856946354883, "grad_norm": 0.46236616373062134, "learning_rate": 6.068999050065249e-06, "loss": 2.3192, "step": 2076 }, { "epoch": 0.9523154516276937, "grad_norm": 0.46928614377975464, "learning_rate": 5.9541784369983586e-06, "loss": 2.2629, "step": 2077 }, { "epoch": 0.9527739569005044, "grad_norm": 0.48047155141830444, "learning_rate": 5.840447873262433e-06, "loss": 2.3683, "step": 2078 }, { "epoch": 0.953232462173315, "grad_norm": 0.45896923542022705, "learning_rate": 5.727807609789471e-06, "loss": 2.3225, "step": 2079 }, { "epoch": 0.9536909674461256, "grad_norm": 0.4745423495769501, "learning_rate": 5.616257895105892e-06, "loss": 2.3619, "step": 2080 }, { "epoch": 0.9541494727189362, "grad_norm": 0.46333083510398865, "learning_rate": 5.505798975331933e-06, "loss": 2.4048, "step": 2081 }, { "epoch": 0.9546079779917469, "grad_norm": 0.5127224326133728, "learning_rate": 5.396431094181198e-06, "loss": 2.4897, "step": 2082 }, { "epoch": 0.9550664832645576, "grad_norm": 0.4808500409126282, "learning_rate": 5.288154492960107e-06, "loss": 2.3709, "step": 2083 }, { "epoch": 0.9555249885373682, "grad_norm": 0.485986590385437, "learning_rate": 5.1809694105671155e-06, "loss": 2.2434, "step": 2084 }, { "epoch": 0.9559834938101788, "grad_norm": 0.46171513199806213, "learning_rate": 5.074876083492441e-06, "loss": 2.3014, "step": 2085 }, { "epoch": 0.9564419990829894, "grad_norm": 0.5012249946594238, "learning_rate": 4.96987474581767e-06, "loss": 2.417, "step": 2086 }, { "epoch": 0.9569005043558001, "grad_norm": 0.495924174785614, "learning_rate": 4.865965629214819e-06, "loss": 2.3252, "step": 2087 }, { "epoch": 0.9573590096286108, "grad_norm": 0.46497678756713867, "learning_rate": 4.763148962946218e-06, "loss": 2.3851, "step": 2088 }, { "epoch": 0.9578175149014214, "grad_norm": 0.516520619392395, "learning_rate": 4.661424973863681e-06, "loss": 2.4582, "step": 2089 }, { "epoch": 0.958276020174232, "grad_norm": 0.47401997447013855, "learning_rate": 4.560793886408398e-06, "loss": 2.3284, "step": 2090 }, { "epoch": 0.9587345254470426, "grad_norm": 0.5004991292953491, "learning_rate": 4.461255922609986e-06, "loss": 2.41, "step": 2091 }, { "epoch": 0.9591930307198533, "grad_norm": 0.47961175441741943, "learning_rate": 4.362811302086267e-06, "loss": 2.2748, "step": 2092 }, { "epoch": 0.959651535992664, "grad_norm": 0.47103151679039, "learning_rate": 4.265460242042885e-06, "loss": 2.3902, "step": 2093 }, { "epoch": 0.9601100412654745, "grad_norm": 0.4732281565666199, "learning_rate": 4.169202957272522e-06, "loss": 2.329, "step": 2094 }, { "epoch": 0.9605685465382852, "grad_norm": 0.4769635498523712, "learning_rate": 4.074039660154738e-06, "loss": 2.3035, "step": 2095 }, { "epoch": 0.9610270518110958, "grad_norm": 0.45755961537361145, "learning_rate": 3.9799705606551325e-06, "loss": 2.3992, "step": 2096 }, { "epoch": 0.9614855570839065, "grad_norm": 0.4819487929344177, "learning_rate": 3.886995866325294e-06, "loss": 2.3653, "step": 2097 }, { "epoch": 0.9619440623567171, "grad_norm": 0.48199397325515747, "learning_rate": 3.795115782302072e-06, "loss": 2.3439, "step": 2098 }, { "epoch": 0.9624025676295277, "grad_norm": 0.47428032755851746, "learning_rate": 3.704330511307197e-06, "loss": 2.3421, "step": 2099 }, { "epoch": 0.9628610729023384, "grad_norm": 0.47649797797203064, "learning_rate": 3.614640253646828e-06, "loss": 2.3335, "step": 2100 }, { "epoch": 0.963319578175149, "grad_norm": 0.4819425046443939, "learning_rate": 3.5260452072110594e-06, "loss": 2.4141, "step": 2101 }, { "epoch": 0.9637780834479597, "grad_norm": 0.4818284213542938, "learning_rate": 3.4385455674737498e-06, "loss": 2.3784, "step": 2102 }, { "epoch": 0.9642365887207703, "grad_norm": 0.4618454873561859, "learning_rate": 3.3521415274915256e-06, "loss": 2.3257, "step": 2103 }, { "epoch": 0.9646950939935809, "grad_norm": 0.4991169273853302, "learning_rate": 3.2668332779041133e-06, "loss": 2.3598, "step": 2104 }, { "epoch": 0.9651535992663915, "grad_norm": 0.4991302788257599, "learning_rate": 3.1826210069332838e-06, "loss": 2.3793, "step": 2105 }, { "epoch": 0.9656121045392022, "grad_norm": 0.4879521131515503, "learning_rate": 3.0995049003826324e-06, "loss": 2.4627, "step": 2106 }, { "epoch": 0.9660706098120129, "grad_norm": 0.473728746175766, "learning_rate": 3.017485141637355e-06, "loss": 2.2648, "step": 2107 }, { "epoch": 0.9665291150848234, "grad_norm": 0.47358986735343933, "learning_rate": 2.9365619116636376e-06, "loss": 2.2861, "step": 2108 }, { "epoch": 0.9669876203576341, "grad_norm": 0.47346189618110657, "learning_rate": 2.856735389008269e-06, "loss": 2.3598, "step": 2109 }, { "epoch": 0.9674461256304447, "grad_norm": 0.45880797505378723, "learning_rate": 2.778005749798307e-06, "loss": 2.3233, "step": 2110 }, { "epoch": 0.9679046309032554, "grad_norm": 0.4664134383201599, "learning_rate": 2.700373167740744e-06, "loss": 2.3676, "step": 2111 }, { "epoch": 0.9683631361760661, "grad_norm": 0.4710712730884552, "learning_rate": 2.62383781412201e-06, "loss": 2.3353, "step": 2112 }, { "epoch": 0.9688216414488766, "grad_norm": 0.4799773693084717, "learning_rate": 2.5483998578076373e-06, "loss": 2.3796, "step": 2113 }, { "epoch": 0.9692801467216873, "grad_norm": 0.4813452959060669, "learning_rate": 2.4740594652418736e-06, "loss": 2.3986, "step": 2114 }, { "epoch": 0.9697386519944979, "grad_norm": 0.47783249616622925, "learning_rate": 2.4008168004472917e-06, "loss": 2.3779, "step": 2115 }, { "epoch": 0.9701971572673086, "grad_norm": 0.46109652519226074, "learning_rate": 2.3286720250246253e-06, "loss": 2.349, "step": 2116 }, { "epoch": 0.9706556625401193, "grad_norm": 0.48212599754333496, "learning_rate": 2.2576252981520994e-06, "loss": 2.3454, "step": 2117 }, { "epoch": 0.9711141678129298, "grad_norm": 0.4877486228942871, "learning_rate": 2.1876767765853233e-06, "loss": 2.3203, "step": 2118 }, { "epoch": 0.9715726730857405, "grad_norm": 0.456173300743103, "learning_rate": 2.118826614656788e-06, "loss": 2.3465, "step": 2119 }, { "epoch": 0.9720311783585511, "grad_norm": 0.47518691420555115, "learning_rate": 2.051074964275701e-06, "loss": 2.3305, "step": 2120 }, { "epoch": 0.9724896836313618, "grad_norm": 0.4810916483402252, "learning_rate": 1.984421974927375e-06, "loss": 2.3383, "step": 2121 }, { "epoch": 0.9729481889041725, "grad_norm": 0.4805362820625305, "learning_rate": 1.9188677936731734e-06, "loss": 2.4283, "step": 2122 }, { "epoch": 0.973406694176983, "grad_norm": 0.4700576364994049, "learning_rate": 1.8544125651501208e-06, "loss": 2.363, "step": 2123 }, { "epoch": 0.9738651994497937, "grad_norm": 0.5001873970031738, "learning_rate": 1.7910564315704035e-06, "loss": 2.4136, "step": 2124 }, { "epoch": 0.9743237047226043, "grad_norm": 0.475168377161026, "learning_rate": 1.7287995327214257e-06, "loss": 2.3867, "step": 2125 }, { "epoch": 0.974782209995415, "grad_norm": 0.4826975464820862, "learning_rate": 1.6676420059649754e-06, "loss": 2.3914, "step": 2126 }, { "epoch": 0.9752407152682255, "grad_norm": 0.47891345620155334, "learning_rate": 1.6075839862374485e-06, "loss": 2.2831, "step": 2127 }, { "epoch": 0.9756992205410362, "grad_norm": 0.4682932496070862, "learning_rate": 1.5486256060492366e-06, "loss": 2.3204, "step": 2128 }, { "epoch": 0.9761577258138469, "grad_norm": 0.4810064733028412, "learning_rate": 1.4907669954844495e-06, "loss": 2.32, "step": 2129 }, { "epoch": 0.9766162310866575, "grad_norm": 0.4906843602657318, "learning_rate": 1.434008282200805e-06, "loss": 2.3685, "step": 2130 }, { "epoch": 0.9770747363594682, "grad_norm": 0.49314752221107483, "learning_rate": 1.3783495914291844e-06, "loss": 2.4069, "step": 2131 }, { "epoch": 0.9775332416322787, "grad_norm": 0.4861360192298889, "learning_rate": 1.3237910459734104e-06, "loss": 2.4377, "step": 2132 }, { "epoch": 0.9779917469050894, "grad_norm": 0.4766589105129242, "learning_rate": 1.270332766210025e-06, "loss": 2.4298, "step": 2133 }, { "epoch": 0.9784502521779, "grad_norm": 0.47296690940856934, "learning_rate": 1.2179748700879012e-06, "loss": 2.4099, "step": 2134 }, { "epoch": 0.9789087574507107, "grad_norm": 0.47536540031433105, "learning_rate": 1.1667174731280205e-06, "loss": 2.4331, "step": 2135 }, { "epoch": 0.9793672627235214, "grad_norm": 0.4549230635166168, "learning_rate": 1.1165606884234182e-06, "loss": 2.3356, "step": 2136 }, { "epoch": 0.9798257679963319, "grad_norm": 0.4830435812473297, "learning_rate": 1.0675046266386268e-06, "loss": 2.378, "step": 2137 }, { "epoch": 0.9802842732691426, "grad_norm": 0.4945777952671051, "learning_rate": 1.019549396009567e-06, "loss": 2.4428, "step": 2138 }, { "epoch": 0.9807427785419532, "grad_norm": 0.4767206311225891, "learning_rate": 9.726951023434348e-07, "loss": 2.3569, "step": 2139 }, { "epoch": 0.9812012838147639, "grad_norm": 0.4845016598701477, "learning_rate": 9.269418490182591e-07, "loss": 2.4441, "step": 2140 }, { "epoch": 0.9816597890875745, "grad_norm": 0.4859788119792938, "learning_rate": 8.822897369827332e-07, "loss": 2.3523, "step": 2141 }, { "epoch": 0.9821182943603851, "grad_norm": 0.456930547952652, "learning_rate": 8.387388647561611e-07, "loss": 2.2958, "step": 2142 }, { "epoch": 0.9825767996331958, "grad_norm": 0.47970232367515564, "learning_rate": 7.962893284279016e-07, "loss": 2.234, "step": 2143 }, { "epoch": 0.9830353049060064, "grad_norm": 0.5030495524406433, "learning_rate": 7.549412216574791e-07, "loss": 2.4444, "step": 2144 }, { "epoch": 0.9834938101788171, "grad_norm": 0.4940820336341858, "learning_rate": 7.146946356743067e-07, "loss": 2.286, "step": 2145 }, { "epoch": 0.9839523154516276, "grad_norm": 0.5099005103111267, "learning_rate": 6.755496592773524e-07, "loss": 2.3379, "step": 2146 }, { "epoch": 0.9844108207244383, "grad_norm": 0.5327973365783691, "learning_rate": 6.375063788349733e-07, "loss": 2.4046, "step": 2147 }, { "epoch": 0.984869325997249, "grad_norm": 0.4852864444255829, "learning_rate": 6.005648782848594e-07, "loss": 2.3405, "step": 2148 }, { "epoch": 0.9853278312700596, "grad_norm": 0.4743252694606781, "learning_rate": 5.647252391337565e-07, "loss": 2.4208, "step": 2149 }, { "epoch": 0.9857863365428703, "grad_norm": 0.48365768790245056, "learning_rate": 5.299875404572441e-07, "loss": 2.3663, "step": 2150 }, { "epoch": 0.9862448418156808, "grad_norm": 0.4659525752067566, "learning_rate": 4.963518588996796e-07, "loss": 2.3225, "step": 2151 }, { "epoch": 0.9867033470884915, "grad_norm": 0.4686768352985382, "learning_rate": 4.638182686738657e-07, "loss": 2.3517, "step": 2152 }, { "epoch": 0.9871618523613022, "grad_norm": 0.47299620509147644, "learning_rate": 4.3238684156110543e-07, "loss": 2.3108, "step": 2153 }, { "epoch": 0.9876203576341128, "grad_norm": 0.48831701278686523, "learning_rate": 4.020576469108139e-07, "loss": 2.2795, "step": 2154 }, { "epoch": 0.9880788629069235, "grad_norm": 0.4862913191318512, "learning_rate": 3.7283075164046274e-07, "loss": 2.3832, "step": 2155 }, { "epoch": 0.988537368179734, "grad_norm": 0.4799465537071228, "learning_rate": 3.4470622023557995e-07, "loss": 2.3276, "step": 2156 }, { "epoch": 0.9889958734525447, "grad_norm": 0.4858568012714386, "learning_rate": 3.176841147492504e-07, "loss": 2.3095, "step": 2157 }, { "epoch": 0.9894543787253554, "grad_norm": 0.4689309597015381, "learning_rate": 2.9176449480244895e-07, "loss": 2.2731, "step": 2158 }, { "epoch": 0.989912883998166, "grad_norm": 0.47506242990493774, "learning_rate": 2.6694741758342967e-07, "loss": 2.4119, "step": 2159 }, { "epoch": 0.9903713892709766, "grad_norm": 0.45982369780540466, "learning_rate": 2.432329378478926e-07, "loss": 2.3542, "step": 2160 }, { "epoch": 0.9908298945437872, "grad_norm": 0.4976836144924164, "learning_rate": 2.2062110791892798e-07, "loss": 2.3304, "step": 2161 }, { "epoch": 0.9912883998165979, "grad_norm": 0.48377615213394165, "learning_rate": 1.9911197768662792e-07, "loss": 2.341, "step": 2162 }, { "epoch": 0.9917469050894085, "grad_norm": 0.4718770980834961, "learning_rate": 1.7870559460814173e-07, "loss": 2.4521, "step": 2163 }, { "epoch": 0.9922054103622192, "grad_norm": 0.4846175014972687, "learning_rate": 1.5940200370750947e-07, "loss": 2.3761, "step": 2164 }, { "epoch": 0.9926639156350298, "grad_norm": 0.5086680054664612, "learning_rate": 1.4120124757577291e-07, "loss": 2.4762, "step": 2165 }, { "epoch": 0.9931224209078404, "grad_norm": 0.49406152963638306, "learning_rate": 1.2410336637047603e-07, "loss": 2.2716, "step": 2166 }, { "epoch": 0.9935809261806511, "grad_norm": 0.4673479497432709, "learning_rate": 1.081083978159425e-07, "loss": 2.4138, "step": 2167 }, { "epoch": 0.9940394314534617, "grad_norm": 0.4535638988018036, "learning_rate": 9.321637720310915e-08, "loss": 2.3038, "step": 2168 }, { "epoch": 0.9944979367262724, "grad_norm": 0.49148187041282654, "learning_rate": 7.942733738924845e-08, "loss": 2.4346, "step": 2169 }, { "epoch": 0.994956441999083, "grad_norm": 0.47544974088668823, "learning_rate": 6.6741308798135e-08, "loss": 2.3448, "step": 2170 }, { "epoch": 0.9954149472718936, "grad_norm": 0.4825502634048462, "learning_rate": 5.5158319419934546e-08, "loss": 2.4021, "step": 2171 }, { "epoch": 0.9958734525447043, "grad_norm": 0.4873143434524536, "learning_rate": 4.4678394810981906e-08, "loss": 2.35, "step": 2172 }, { "epoch": 0.9963319578175149, "grad_norm": 0.4692324995994568, "learning_rate": 3.5301558093947527e-08, "loss": 2.4214, "step": 2173 }, { "epoch": 0.9967904630903255, "grad_norm": 0.46082159876823425, "learning_rate": 2.7027829957559925e-08, "loss": 2.354, "step": 2174 }, { "epoch": 0.9972489683631361, "grad_norm": 0.4857215881347656, "learning_rate": 1.985722865682771e-08, "loss": 2.3665, "step": 2175 }, { "epoch": 0.9977074736359468, "grad_norm": 0.4665515124797821, "learning_rate": 1.3789770012762048e-08, "loss": 2.2713, "step": 2176 }, { "epoch": 0.9981659789087575, "grad_norm": 0.462049663066864, "learning_rate": 8.825467412376665e-09, "loss": 2.3626, "step": 2177 }, { "epoch": 0.9986244841815681, "grad_norm": 0.46807149052619934, "learning_rate": 4.96433180879885e-09, "loss": 2.3748, "step": 2178 }, { "epoch": 0.9990829894543787, "grad_norm": 0.4730471968650818, "learning_rate": 2.206371721158451e-09, "loss": 2.3749, "step": 2179 }, { "epoch": 0.9995414947271893, "grad_norm": 0.4668895900249481, "learning_rate": 5.515932345323549e-10, "loss": 2.3001, "step": 2180 }, { "epoch": 1.0, "grad_norm": 0.448803573846817, "learning_rate": 0.0, "loss": 2.3944, "step": 2181 }, { "epoch": 1.0, "step": 2181, "total_flos": 6.473631625254011e+17, "train_loss": 2.517149598267252, "train_runtime": 5969.0084, "train_samples_per_second": 93.504, "train_steps_per_second": 0.365 } ], "logging_steps": 1.0, "max_steps": 2181, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 24000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": false, "should_training_stop": false }, "attributes": {} } }, "total_flos": 6.473631625254011e+17, "train_batch_size": 16, "trial_name": null, "trial_params": null }