Text Generation
Transformers
Safetensors
qwen3
conversational
text-generation-inference
davidanugraha's picture
Upload folder using huggingface_hub
b5ada77 verified
Raw
History Blame Contribute Delete
170 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 975,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.003078699249567058,
"grad_norm": 7.6762862397276015,
"learning_rate": 0.0,
"loss": 1.6081,
"step": 1
},
{
"epoch": 0.006157398499134116,
"grad_norm": 7.2553568740644305,
"learning_rate": 1.0204081632653061e-07,
"loss": 1.5436,
"step": 2
},
{
"epoch": 0.009236097748701174,
"grad_norm": 7.004828991301673,
"learning_rate": 2.0408163265306121e-07,
"loss": 1.5148,
"step": 3
},
{
"epoch": 0.012314796998268231,
"grad_norm": 6.935837625449577,
"learning_rate": 3.0612244897959183e-07,
"loss": 1.5312,
"step": 4
},
{
"epoch": 0.01539349624783529,
"grad_norm": 6.33455034747555,
"learning_rate": 4.0816326530612243e-07,
"loss": 1.4897,
"step": 5
},
{
"epoch": 0.01847219549740235,
"grad_norm": 7.0616985246304385,
"learning_rate": 5.102040816326531e-07,
"loss": 1.5687,
"step": 6
},
{
"epoch": 0.021550894746969404,
"grad_norm": 7.089005103944213,
"learning_rate": 6.122448979591837e-07,
"loss": 1.5059,
"step": 7
},
{
"epoch": 0.024629593996536463,
"grad_norm": 6.521309836236844,
"learning_rate": 7.142857142857143e-07,
"loss": 1.5089,
"step": 8
},
{
"epoch": 0.02770829324610352,
"grad_norm": 6.59306463046276,
"learning_rate": 8.163265306122449e-07,
"loss": 1.5407,
"step": 9
},
{
"epoch": 0.03078699249567058,
"grad_norm": 6.233701312970518,
"learning_rate": 9.183673469387756e-07,
"loss": 1.4822,
"step": 10
},
{
"epoch": 0.03386569174523764,
"grad_norm": 6.074480871814344,
"learning_rate": 1.0204081632653063e-06,
"loss": 1.5503,
"step": 11
},
{
"epoch": 0.0369443909948047,
"grad_norm": 5.678340026182479,
"learning_rate": 1.122448979591837e-06,
"loss": 1.5037,
"step": 12
},
{
"epoch": 0.040023090244371756,
"grad_norm": 5.602335495459839,
"learning_rate": 1.2244897959183673e-06,
"loss": 1.4823,
"step": 13
},
{
"epoch": 0.04310178949393881,
"grad_norm": 4.041056484458271,
"learning_rate": 1.3265306122448982e-06,
"loss": 1.3712,
"step": 14
},
{
"epoch": 0.04618048874350587,
"grad_norm": 4.134762915666753,
"learning_rate": 1.4285714285714286e-06,
"loss": 1.4473,
"step": 15
},
{
"epoch": 0.049259187993072925,
"grad_norm": 3.8490814865325262,
"learning_rate": 1.5306122448979593e-06,
"loss": 1.4376,
"step": 16
},
{
"epoch": 0.052337887242639984,
"grad_norm": 3.7213326992417635,
"learning_rate": 1.6326530612244897e-06,
"loss": 1.4244,
"step": 17
},
{
"epoch": 0.05541658649220704,
"grad_norm": 3.3983290221271916,
"learning_rate": 1.7346938775510206e-06,
"loss": 1.3874,
"step": 18
},
{
"epoch": 0.0584952857417741,
"grad_norm": 2.043538704895783,
"learning_rate": 1.8367346938775512e-06,
"loss": 1.3658,
"step": 19
},
{
"epoch": 0.06157398499134116,
"grad_norm": 1.523707142183616,
"learning_rate": 1.938775510204082e-06,
"loss": 1.2845,
"step": 20
},
{
"epoch": 0.06465268424090821,
"grad_norm": 1.4012583912181333,
"learning_rate": 2.0408163265306125e-06,
"loss": 1.312,
"step": 21
},
{
"epoch": 0.06773138349047528,
"grad_norm": 1.438236087785234,
"learning_rate": 2.1428571428571427e-06,
"loss": 1.3546,
"step": 22
},
{
"epoch": 0.07081008274004233,
"grad_norm": 1.2356993856460103,
"learning_rate": 2.244897959183674e-06,
"loss": 1.3006,
"step": 23
},
{
"epoch": 0.0738887819896094,
"grad_norm": 1.2848569993664685,
"learning_rate": 2.3469387755102044e-06,
"loss": 1.3404,
"step": 24
},
{
"epoch": 0.07696748123917645,
"grad_norm": 1.1447371224658784,
"learning_rate": 2.4489795918367347e-06,
"loss": 1.3251,
"step": 25
},
{
"epoch": 0.08004618048874351,
"grad_norm": 1.2943325496057256,
"learning_rate": 2.5510204081632657e-06,
"loss": 1.2858,
"step": 26
},
{
"epoch": 0.08312487973831056,
"grad_norm": 1.3803064787973331,
"learning_rate": 2.6530612244897964e-06,
"loss": 1.2365,
"step": 27
},
{
"epoch": 0.08620357898787762,
"grad_norm": 1.3710239312771144,
"learning_rate": 2.7551020408163266e-06,
"loss": 1.2489,
"step": 28
},
{
"epoch": 0.08928227823744468,
"grad_norm": 1.214324239044426,
"learning_rate": 2.8571428571428573e-06,
"loss": 1.202,
"step": 29
},
{
"epoch": 0.09236097748701173,
"grad_norm": 1.056364977100255,
"learning_rate": 2.959183673469388e-06,
"loss": 1.2126,
"step": 30
},
{
"epoch": 0.0954396767365788,
"grad_norm": 1.044476433899929,
"learning_rate": 3.0612244897959185e-06,
"loss": 1.2288,
"step": 31
},
{
"epoch": 0.09851837598614585,
"grad_norm": 0.8269850081580365,
"learning_rate": 3.1632653061224496e-06,
"loss": 1.1956,
"step": 32
},
{
"epoch": 0.10159707523571292,
"grad_norm": 0.6960425678816088,
"learning_rate": 3.2653061224489794e-06,
"loss": 1.1707,
"step": 33
},
{
"epoch": 0.10467577448527997,
"grad_norm": 0.6240461576700048,
"learning_rate": 3.3673469387755105e-06,
"loss": 1.1815,
"step": 34
},
{
"epoch": 0.10775447373484703,
"grad_norm": 0.6109843060555834,
"learning_rate": 3.469387755102041e-06,
"loss": 1.1927,
"step": 35
},
{
"epoch": 0.11083317298441409,
"grad_norm": 0.6741412542318923,
"learning_rate": 3.5714285714285718e-06,
"loss": 1.1703,
"step": 36
},
{
"epoch": 0.11391187223398114,
"grad_norm": 0.575094776064826,
"learning_rate": 3.6734693877551024e-06,
"loss": 1.1244,
"step": 37
},
{
"epoch": 0.1169905714835482,
"grad_norm": 0.8571866963793218,
"learning_rate": 3.7755102040816327e-06,
"loss": 1.1211,
"step": 38
},
{
"epoch": 0.12006927073311525,
"grad_norm": 0.5345222397532509,
"learning_rate": 3.877551020408164e-06,
"loss": 1.122,
"step": 39
},
{
"epoch": 0.12314796998268232,
"grad_norm": 0.514410824645149,
"learning_rate": 3.979591836734694e-06,
"loss": 1.1157,
"step": 40
},
{
"epoch": 0.12622666923224937,
"grad_norm": 0.48053423461875816,
"learning_rate": 4.081632653061225e-06,
"loss": 1.1461,
"step": 41
},
{
"epoch": 0.12930536848181642,
"grad_norm": 0.4111548142660251,
"learning_rate": 4.183673469387755e-06,
"loss": 1.0792,
"step": 42
},
{
"epoch": 0.1323840677313835,
"grad_norm": 0.42697229877433346,
"learning_rate": 4.2857142857142855e-06,
"loss": 1.1164,
"step": 43
},
{
"epoch": 0.13546276698095056,
"grad_norm": 0.42411836990069424,
"learning_rate": 4.3877551020408165e-06,
"loss": 1.1278,
"step": 44
},
{
"epoch": 0.1385414662305176,
"grad_norm": 0.5501794736719339,
"learning_rate": 4.489795918367348e-06,
"loss": 1.1246,
"step": 45
},
{
"epoch": 0.14162016548008466,
"grad_norm": 0.40670830755674436,
"learning_rate": 4.591836734693878e-06,
"loss": 1.1332,
"step": 46
},
{
"epoch": 0.1446988647296517,
"grad_norm": 0.3704308825927558,
"learning_rate": 4.693877551020409e-06,
"loss": 1.0771,
"step": 47
},
{
"epoch": 0.1477775639792188,
"grad_norm": 0.3544794413677721,
"learning_rate": 4.795918367346939e-06,
"loss": 1.0809,
"step": 48
},
{
"epoch": 0.15085626322878584,
"grad_norm": 0.3604622537315224,
"learning_rate": 4.897959183673469e-06,
"loss": 1.0653,
"step": 49
},
{
"epoch": 0.1539349624783529,
"grad_norm": 0.34269989653881605,
"learning_rate": 5e-06,
"loss": 1.1156,
"step": 50
},
{
"epoch": 0.15701366172791995,
"grad_norm": 0.33319411582371367,
"learning_rate": 5.1020408163265315e-06,
"loss": 1.0303,
"step": 51
},
{
"epoch": 0.16009236097748702,
"grad_norm": 0.3207920288009826,
"learning_rate": 5.204081632653062e-06,
"loss": 1.1027,
"step": 52
},
{
"epoch": 0.16317106022705408,
"grad_norm": 0.3344702787960009,
"learning_rate": 5.306122448979593e-06,
"loss": 1.1327,
"step": 53
},
{
"epoch": 0.16624975947662113,
"grad_norm": 0.3295457427156366,
"learning_rate": 5.408163265306123e-06,
"loss": 1.0925,
"step": 54
},
{
"epoch": 0.16932845872618818,
"grad_norm": 0.3152967559990598,
"learning_rate": 5.510204081632653e-06,
"loss": 1.0847,
"step": 55
},
{
"epoch": 0.17240715797575523,
"grad_norm": 0.2970385911768163,
"learning_rate": 5.6122448979591834e-06,
"loss": 1.0574,
"step": 56
},
{
"epoch": 0.1754858572253223,
"grad_norm": 0.2787460665168615,
"learning_rate": 5.7142857142857145e-06,
"loss": 1.0385,
"step": 57
},
{
"epoch": 0.17856455647488936,
"grad_norm": 0.297865096636938,
"learning_rate": 5.816326530612246e-06,
"loss": 1.0553,
"step": 58
},
{
"epoch": 0.18164325572445641,
"grad_norm": 0.27595486379996903,
"learning_rate": 5.918367346938776e-06,
"loss": 1.0336,
"step": 59
},
{
"epoch": 0.18472195497402347,
"grad_norm": 0.3080206447223133,
"learning_rate": 6.020408163265307e-06,
"loss": 1.0815,
"step": 60
},
{
"epoch": 0.18780065422359055,
"grad_norm": 0.271203527490175,
"learning_rate": 6.122448979591837e-06,
"loss": 1.0388,
"step": 61
},
{
"epoch": 0.1908793534731576,
"grad_norm": 0.2528253583678469,
"learning_rate": 6.224489795918368e-06,
"loss": 1.0301,
"step": 62
},
{
"epoch": 0.19395805272272465,
"grad_norm": 0.2810411289740686,
"learning_rate": 6.326530612244899e-06,
"loss": 1.0378,
"step": 63
},
{
"epoch": 0.1970367519722917,
"grad_norm": 0.2712413656418754,
"learning_rate": 6.4285714285714295e-06,
"loss": 1.0195,
"step": 64
},
{
"epoch": 0.20011545122185875,
"grad_norm": 0.2533356500499567,
"learning_rate": 6.530612244897959e-06,
"loss": 0.9852,
"step": 65
},
{
"epoch": 0.20319415047142583,
"grad_norm": 0.24883416826064644,
"learning_rate": 6.63265306122449e-06,
"loss": 1.0095,
"step": 66
},
{
"epoch": 0.20627284972099288,
"grad_norm": 0.26004865364865526,
"learning_rate": 6.734693877551021e-06,
"loss": 1.0399,
"step": 67
},
{
"epoch": 0.20935154897055994,
"grad_norm": 0.26709702511054434,
"learning_rate": 6.836734693877551e-06,
"loss": 1.0515,
"step": 68
},
{
"epoch": 0.212430248220127,
"grad_norm": 0.2448770500748951,
"learning_rate": 6.938775510204082e-06,
"loss": 0.9825,
"step": 69
},
{
"epoch": 0.21550894746969407,
"grad_norm": 0.2518407465293782,
"learning_rate": 7.0408163265306125e-06,
"loss": 1.0595,
"step": 70
},
{
"epoch": 0.21858764671926112,
"grad_norm": 0.2583530558367604,
"learning_rate": 7.1428571428571436e-06,
"loss": 0.9966,
"step": 71
},
{
"epoch": 0.22166634596882817,
"grad_norm": 0.2624475563823761,
"learning_rate": 7.244897959183675e-06,
"loss": 0.9993,
"step": 72
},
{
"epoch": 0.22474504521839522,
"grad_norm": 0.26075170855200724,
"learning_rate": 7.346938775510205e-06,
"loss": 1.0441,
"step": 73
},
{
"epoch": 0.22782374446796227,
"grad_norm": 0.25911405395375126,
"learning_rate": 7.448979591836736e-06,
"loss": 1.0146,
"step": 74
},
{
"epoch": 0.23090244371752935,
"grad_norm": 0.23791804237751532,
"learning_rate": 7.551020408163265e-06,
"loss": 1.0289,
"step": 75
},
{
"epoch": 0.2339811429670964,
"grad_norm": 0.24296561717731074,
"learning_rate": 7.653061224489796e-06,
"loss": 1.0135,
"step": 76
},
{
"epoch": 0.23705984221666346,
"grad_norm": 0.2438416896054916,
"learning_rate": 7.755102040816327e-06,
"loss": 1.0232,
"step": 77
},
{
"epoch": 0.2401385414662305,
"grad_norm": 0.24844353728508622,
"learning_rate": 7.857142857142858e-06,
"loss": 0.9932,
"step": 78
},
{
"epoch": 0.2432172407157976,
"grad_norm": 0.2529237287765586,
"learning_rate": 7.959183673469388e-06,
"loss": 1.0252,
"step": 79
},
{
"epoch": 0.24629593996536464,
"grad_norm": 0.23780413810684603,
"learning_rate": 8.06122448979592e-06,
"loss": 1.0067,
"step": 80
},
{
"epoch": 0.2493746392149317,
"grad_norm": 0.2411610878426367,
"learning_rate": 8.16326530612245e-06,
"loss": 0.9789,
"step": 81
},
{
"epoch": 0.25245333846449874,
"grad_norm": 0.25553671929345567,
"learning_rate": 8.26530612244898e-06,
"loss": 1.0157,
"step": 82
},
{
"epoch": 0.2555320377140658,
"grad_norm": 0.2381647908430099,
"learning_rate": 8.36734693877551e-06,
"loss": 1.0119,
"step": 83
},
{
"epoch": 0.25861073696363285,
"grad_norm": 0.2423853873300254,
"learning_rate": 8.469387755102042e-06,
"loss": 1.0114,
"step": 84
},
{
"epoch": 0.2616894362131999,
"grad_norm": 0.25843530934063885,
"learning_rate": 8.571428571428571e-06,
"loss": 0.9987,
"step": 85
},
{
"epoch": 0.264768135462767,
"grad_norm": 0.22567077188718443,
"learning_rate": 8.673469387755103e-06,
"loss": 0.9813,
"step": 86
},
{
"epoch": 0.26784683471233406,
"grad_norm": 0.23194020150989653,
"learning_rate": 8.775510204081633e-06,
"loss": 0.9818,
"step": 87
},
{
"epoch": 0.2709255339619011,
"grad_norm": 0.23140476492973205,
"learning_rate": 8.877551020408163e-06,
"loss": 1.0077,
"step": 88
},
{
"epoch": 0.27400423321146816,
"grad_norm": 0.2516969181166962,
"learning_rate": 8.979591836734695e-06,
"loss": 1.0149,
"step": 89
},
{
"epoch": 0.2770829324610352,
"grad_norm": 0.22633366959665108,
"learning_rate": 9.081632653061225e-06,
"loss": 0.973,
"step": 90
},
{
"epoch": 0.28016163171060227,
"grad_norm": 0.24495995743449453,
"learning_rate": 9.183673469387756e-06,
"loss": 1.0085,
"step": 91
},
{
"epoch": 0.2832403309601693,
"grad_norm": 0.2340654521530156,
"learning_rate": 9.285714285714288e-06,
"loss": 0.9765,
"step": 92
},
{
"epoch": 0.28631903020973637,
"grad_norm": 0.2433140206421755,
"learning_rate": 9.387755102040818e-06,
"loss": 0.9996,
"step": 93
},
{
"epoch": 0.2893977294593034,
"grad_norm": 0.24756935957940238,
"learning_rate": 9.489795918367348e-06,
"loss": 1.0007,
"step": 94
},
{
"epoch": 0.29247642870887053,
"grad_norm": 0.23527080976211437,
"learning_rate": 9.591836734693878e-06,
"loss": 0.9647,
"step": 95
},
{
"epoch": 0.2955551279584376,
"grad_norm": 0.2369643971671493,
"learning_rate": 9.693877551020408e-06,
"loss": 0.9429,
"step": 96
},
{
"epoch": 0.29863382720800463,
"grad_norm": 0.324700148498595,
"learning_rate": 9.795918367346939e-06,
"loss": 1.0074,
"step": 97
},
{
"epoch": 0.3017125264575717,
"grad_norm": 0.23264017758720767,
"learning_rate": 9.89795918367347e-06,
"loss": 0.9572,
"step": 98
},
{
"epoch": 0.30479122570713874,
"grad_norm": 0.24383935729881154,
"learning_rate": 1e-05,
"loss": 1.0285,
"step": 99
},
{
"epoch": 0.3078699249567058,
"grad_norm": 0.24296539560019878,
"learning_rate": 9.999967919575756e-06,
"loss": 1.0349,
"step": 100
},
{
"epoch": 0.31094862420627284,
"grad_norm": 0.23995576223683912,
"learning_rate": 9.99987167871469e-06,
"loss": 0.9814,
"step": 101
},
{
"epoch": 0.3140273234558399,
"grad_norm": 0.2317403837961078,
"learning_rate": 9.999711278651776e-06,
"loss": 0.9578,
"step": 102
},
{
"epoch": 0.31710602270540694,
"grad_norm": 0.23429221676377304,
"learning_rate": 9.9994867214453e-06,
"loss": 0.9806,
"step": 103
},
{
"epoch": 0.32018472195497405,
"grad_norm": 0.2328805812132316,
"learning_rate": 9.999198009976813e-06,
"loss": 0.9693,
"step": 104
},
{
"epoch": 0.3232634212045411,
"grad_norm": 0.23479256880784025,
"learning_rate": 9.998845147951112e-06,
"loss": 0.9466,
"step": 105
},
{
"epoch": 0.32634212045410815,
"grad_norm": 0.24996167875984182,
"learning_rate": 9.998428139896184e-06,
"loss": 0.9961,
"step": 106
},
{
"epoch": 0.3294208197036752,
"grad_norm": 0.22616882269510602,
"learning_rate": 9.997946991163144e-06,
"loss": 0.9449,
"step": 107
},
{
"epoch": 0.33249951895324226,
"grad_norm": 0.24787489578053426,
"learning_rate": 9.997401707926176e-06,
"loss": 0.9566,
"step": 108
},
{
"epoch": 0.3355782182028093,
"grad_norm": 0.2465212224514598,
"learning_rate": 9.996792297182447e-06,
"loss": 1.0131,
"step": 109
},
{
"epoch": 0.33865691745237636,
"grad_norm": 0.23883477045743723,
"learning_rate": 9.996118766752018e-06,
"loss": 0.9515,
"step": 110
},
{
"epoch": 0.3417356167019434,
"grad_norm": 0.24345627967957567,
"learning_rate": 9.995381125277747e-06,
"loss": 0.9755,
"step": 111
},
{
"epoch": 0.34481431595151046,
"grad_norm": 0.2331125228495404,
"learning_rate": 9.994579382225176e-06,
"loss": 0.9277,
"step": 112
},
{
"epoch": 0.34789301520107757,
"grad_norm": 0.2391370697573781,
"learning_rate": 9.993713547882404e-06,
"loss": 0.9522,
"step": 113
},
{
"epoch": 0.3509717144506446,
"grad_norm": 0.24740880361920184,
"learning_rate": 9.992783633359967e-06,
"loss": 0.9615,
"step": 114
},
{
"epoch": 0.3540504137002117,
"grad_norm": 0.25149103063931966,
"learning_rate": 9.991789650590687e-06,
"loss": 0.9657,
"step": 115
},
{
"epoch": 0.3571291129497787,
"grad_norm": 0.24419151684184173,
"learning_rate": 9.990731612329517e-06,
"loss": 0.9847,
"step": 116
},
{
"epoch": 0.3602078121993458,
"grad_norm": 0.24127522003686444,
"learning_rate": 9.989609532153382e-06,
"loss": 0.96,
"step": 117
},
{
"epoch": 0.36328651144891283,
"grad_norm": 0.2348266746584977,
"learning_rate": 9.988423424461009e-06,
"loss": 0.9699,
"step": 118
},
{
"epoch": 0.3663652106984799,
"grad_norm": 0.24333113125447497,
"learning_rate": 9.987173304472733e-06,
"loss": 0.9369,
"step": 119
},
{
"epoch": 0.36944390994804693,
"grad_norm": 0.24165122428746325,
"learning_rate": 9.985859188230302e-06,
"loss": 0.9511,
"step": 120
},
{
"epoch": 0.372522609197614,
"grad_norm": 0.23648975169176661,
"learning_rate": 9.984481092596683e-06,
"loss": 0.9947,
"step": 121
},
{
"epoch": 0.3756013084471811,
"grad_norm": 0.24133576956010333,
"learning_rate": 9.983039035255831e-06,
"loss": 0.927,
"step": 122
},
{
"epoch": 0.37868000769674814,
"grad_norm": 0.2314965558585553,
"learning_rate": 9.981533034712472e-06,
"loss": 0.928,
"step": 123
},
{
"epoch": 0.3817587069463152,
"grad_norm": 0.24017330770409864,
"learning_rate": 9.979963110291857e-06,
"loss": 0.9429,
"step": 124
},
{
"epoch": 0.38483740619588225,
"grad_norm": 0.24749917112976713,
"learning_rate": 9.978329282139526e-06,
"loss": 0.957,
"step": 125
},
{
"epoch": 0.3879161054454493,
"grad_norm": 0.2456622314573542,
"learning_rate": 9.976631571221039e-06,
"loss": 0.9741,
"step": 126
},
{
"epoch": 0.39099480469501635,
"grad_norm": 0.249166135910459,
"learning_rate": 9.974869999321707e-06,
"loss": 1.0026,
"step": 127
},
{
"epoch": 0.3940735039445834,
"grad_norm": 0.24495419707101526,
"learning_rate": 9.973044589046325e-06,
"loss": 0.953,
"step": 128
},
{
"epoch": 0.39715220319415045,
"grad_norm": 0.23746216180701943,
"learning_rate": 9.971155363818863e-06,
"loss": 0.972,
"step": 129
},
{
"epoch": 0.4002309024437175,
"grad_norm": 0.24428141655546667,
"learning_rate": 9.969202347882181e-06,
"loss": 0.932,
"step": 130
},
{
"epoch": 0.4033096016932846,
"grad_norm": 0.23669551097133704,
"learning_rate": 9.967185566297713e-06,
"loss": 0.9052,
"step": 131
},
{
"epoch": 0.40638830094285167,
"grad_norm": 0.25239673142467994,
"learning_rate": 9.965105044945138e-06,
"loss": 0.9497,
"step": 132
},
{
"epoch": 0.4094670001924187,
"grad_norm": 0.2554209241299065,
"learning_rate": 9.962960810522065e-06,
"loss": 0.9546,
"step": 133
},
{
"epoch": 0.41254569944198577,
"grad_norm": 0.23986695886465362,
"learning_rate": 9.960752890543668e-06,
"loss": 0.963,
"step": 134
},
{
"epoch": 0.4156243986915528,
"grad_norm": 0.24905086158751802,
"learning_rate": 9.958481313342354e-06,
"loss": 0.9623,
"step": 135
},
{
"epoch": 0.4187030979411199,
"grad_norm": 0.24162306576835293,
"learning_rate": 9.956146108067388e-06,
"loss": 0.9555,
"step": 136
},
{
"epoch": 0.4217817971906869,
"grad_norm": 0.24341458494654228,
"learning_rate": 9.953747304684518e-06,
"loss": 0.9244,
"step": 137
},
{
"epoch": 0.424860496440254,
"grad_norm": 0.2480035125677915,
"learning_rate": 9.951284933975598e-06,
"loss": 1.0288,
"step": 138
},
{
"epoch": 0.42793919568982103,
"grad_norm": 0.24976612415556382,
"learning_rate": 9.948759027538186e-06,
"loss": 0.943,
"step": 139
},
{
"epoch": 0.43101789493938814,
"grad_norm": 0.25779540369691606,
"learning_rate": 9.946169617785142e-06,
"loss": 0.9715,
"step": 140
},
{
"epoch": 0.4340965941889552,
"grad_norm": 0.23968048105439946,
"learning_rate": 9.94351673794421e-06,
"loss": 0.9482,
"step": 141
},
{
"epoch": 0.43717529343852224,
"grad_norm": 0.23289948088220958,
"learning_rate": 9.940800422057596e-06,
"loss": 0.9426,
"step": 142
},
{
"epoch": 0.4402539926880893,
"grad_norm": 0.2297826883164788,
"learning_rate": 9.938020704981527e-06,
"loss": 0.9219,
"step": 143
},
{
"epoch": 0.44333269193765634,
"grad_norm": 0.22815439891547554,
"learning_rate": 9.935177622385805e-06,
"loss": 0.937,
"step": 144
},
{
"epoch": 0.4464113911872234,
"grad_norm": 0.23695635649265118,
"learning_rate": 9.932271210753344e-06,
"loss": 0.9282,
"step": 145
},
{
"epoch": 0.44949009043679045,
"grad_norm": 0.24123370893225643,
"learning_rate": 9.929301507379714e-06,
"loss": 0.9331,
"step": 146
},
{
"epoch": 0.4525687896863575,
"grad_norm": 0.24637077609339264,
"learning_rate": 9.926268550372654e-06,
"loss": 0.9636,
"step": 147
},
{
"epoch": 0.45564748893592455,
"grad_norm": 0.2373248906760509,
"learning_rate": 9.923172378651581e-06,
"loss": 0.9108,
"step": 148
},
{
"epoch": 0.45872618818549166,
"grad_norm": 0.24695655128882715,
"learning_rate": 9.920013031947097e-06,
"loss": 0.9567,
"step": 149
},
{
"epoch": 0.4618048874350587,
"grad_norm": 0.25601297506688836,
"learning_rate": 9.916790550800474e-06,
"loss": 0.9869,
"step": 150
},
{
"epoch": 0.46488358668462576,
"grad_norm": 0.2543434097134282,
"learning_rate": 9.913504976563138e-06,
"loss": 0.9202,
"step": 151
},
{
"epoch": 0.4679622859341928,
"grad_norm": 0.23080918546804705,
"learning_rate": 9.910156351396135e-06,
"loss": 0.8933,
"step": 152
},
{
"epoch": 0.47104098518375986,
"grad_norm": 0.24006356980764706,
"learning_rate": 9.90674471826959e-06,
"loss": 0.9773,
"step": 153
},
{
"epoch": 0.4741196844333269,
"grad_norm": 0.2471231406770028,
"learning_rate": 9.903270120962158e-06,
"loss": 0.9556,
"step": 154
},
{
"epoch": 0.47719838368289397,
"grad_norm": 0.23910469455718603,
"learning_rate": 9.899732604060464e-06,
"loss": 0.9496,
"step": 155
},
{
"epoch": 0.480277082932461,
"grad_norm": 0.23785026286160252,
"learning_rate": 9.896132212958524e-06,
"loss": 0.9575,
"step": 156
},
{
"epoch": 0.48335578218202807,
"grad_norm": 0.23172416549682723,
"learning_rate": 9.892468993857167e-06,
"loss": 0.944,
"step": 157
},
{
"epoch": 0.4864344814315952,
"grad_norm": 0.2665983629045336,
"learning_rate": 9.888742993763442e-06,
"loss": 0.9389,
"step": 158
},
{
"epoch": 0.48951318068116223,
"grad_norm": 0.2564131067418342,
"learning_rate": 9.884954260490018e-06,
"loss": 0.9467,
"step": 159
},
{
"epoch": 0.4925918799307293,
"grad_norm": 0.24180620982823195,
"learning_rate": 9.881102842654557e-06,
"loss": 0.9455,
"step": 160
},
{
"epoch": 0.49567057918029633,
"grad_norm": 0.2369315498501098,
"learning_rate": 9.877188789679111e-06,
"loss": 0.9315,
"step": 161
},
{
"epoch": 0.4987492784298634,
"grad_norm": 0.23523756440126883,
"learning_rate": 9.873212151789468e-06,
"loss": 0.9538,
"step": 162
},
{
"epoch": 0.5018279776794304,
"grad_norm": 0.23802345798653649,
"learning_rate": 9.869172980014525e-06,
"loss": 0.8587,
"step": 163
},
{
"epoch": 0.5049066769289975,
"grad_norm": 0.24502941484448904,
"learning_rate": 9.865071326185617e-06,
"loss": 0.9109,
"step": 164
},
{
"epoch": 0.5079853761785645,
"grad_norm": 0.24774268311207864,
"learning_rate": 9.860907242935863e-06,
"loss": 0.9478,
"step": 165
},
{
"epoch": 0.5110640754281316,
"grad_norm": 0.27071787962516364,
"learning_rate": 9.856680783699484e-06,
"loss": 0.9652,
"step": 166
},
{
"epoch": 0.5141427746776986,
"grad_norm": 0.2624307597340836,
"learning_rate": 9.852392002711126e-06,
"loss": 0.928,
"step": 167
},
{
"epoch": 0.5172214739272657,
"grad_norm": 0.2574343409060601,
"learning_rate": 9.848040955005149e-06,
"loss": 0.9847,
"step": 168
},
{
"epoch": 0.5203001731768327,
"grad_norm": 0.24446895028801205,
"learning_rate": 9.843627696414938e-06,
"loss": 0.9433,
"step": 169
},
{
"epoch": 0.5233788724263998,
"grad_norm": 0.26604221924628485,
"learning_rate": 9.839152283572178e-06,
"loss": 0.9492,
"step": 170
},
{
"epoch": 0.5264575716759669,
"grad_norm": 0.2418001779051891,
"learning_rate": 9.834614773906125e-06,
"loss": 0.9892,
"step": 171
},
{
"epoch": 0.529536270925534,
"grad_norm": 0.24107515038530022,
"learning_rate": 9.830015225642872e-06,
"loss": 0.8911,
"step": 172
},
{
"epoch": 0.5326149701751011,
"grad_norm": 0.24337229530629442,
"learning_rate": 9.825353697804603e-06,
"loss": 0.9312,
"step": 173
},
{
"epoch": 0.5356936694246681,
"grad_norm": 0.25241834169479355,
"learning_rate": 9.820630250208836e-06,
"loss": 0.925,
"step": 174
},
{
"epoch": 0.5387723686742352,
"grad_norm": 0.2533121094934043,
"learning_rate": 9.815844943467649e-06,
"loss": 0.9217,
"step": 175
},
{
"epoch": 0.5418510679238022,
"grad_norm": 0.26530879064762447,
"learning_rate": 9.810997838986913e-06,
"loss": 0.9595,
"step": 176
},
{
"epoch": 0.5449297671733693,
"grad_norm": 0.25217550960290724,
"learning_rate": 9.806088998965494e-06,
"loss": 0.9262,
"step": 177
},
{
"epoch": 0.5480084664229363,
"grad_norm": 0.2496243080999867,
"learning_rate": 9.801118486394461e-06,
"loss": 0.9143,
"step": 178
},
{
"epoch": 0.5510871656725034,
"grad_norm": 0.2478965277476841,
"learning_rate": 9.796086365056271e-06,
"loss": 0.946,
"step": 179
},
{
"epoch": 0.5541658649220704,
"grad_norm": 0.23746873455242137,
"learning_rate": 9.790992699523967e-06,
"loss": 0.933,
"step": 180
},
{
"epoch": 0.5572445641716375,
"grad_norm": 0.24137860089700794,
"learning_rate": 9.785837555160322e-06,
"loss": 0.926,
"step": 181
},
{
"epoch": 0.5603232634212045,
"grad_norm": 0.24573162361366074,
"learning_rate": 9.780620998117025e-06,
"loss": 0.9353,
"step": 182
},
{
"epoch": 0.5634019626707716,
"grad_norm": 0.25885091437855257,
"learning_rate": 9.775343095333825e-06,
"loss": 0.9763,
"step": 183
},
{
"epoch": 0.5664806619203386,
"grad_norm": 0.2448842992799257,
"learning_rate": 9.770003914537661e-06,
"loss": 0.9159,
"step": 184
},
{
"epoch": 0.5695593611699057,
"grad_norm": 0.2664737182786056,
"learning_rate": 9.76460352424181e-06,
"loss": 0.9414,
"step": 185
},
{
"epoch": 0.5726380604194727,
"grad_norm": 0.2533642857170274,
"learning_rate": 9.759141993744995e-06,
"loss": 0.931,
"step": 186
},
{
"epoch": 0.5757167596690398,
"grad_norm": 0.2565843996145463,
"learning_rate": 9.753619393130505e-06,
"loss": 0.9184,
"step": 187
},
{
"epoch": 0.5787954589186068,
"grad_norm": 0.25892991719114605,
"learning_rate": 9.748035793265284e-06,
"loss": 0.9587,
"step": 188
},
{
"epoch": 0.5818741581681739,
"grad_norm": 0.24511756119746253,
"learning_rate": 9.742391265799038e-06,
"loss": 0.9135,
"step": 189
},
{
"epoch": 0.5849528574177411,
"grad_norm": 0.2651622826517966,
"learning_rate": 9.736685883163297e-06,
"loss": 0.9306,
"step": 190
},
{
"epoch": 0.5880315566673081,
"grad_norm": 0.24021542694001807,
"learning_rate": 9.730919718570501e-06,
"loss": 0.9167,
"step": 191
},
{
"epoch": 0.5911102559168752,
"grad_norm": 0.25684912451312697,
"learning_rate": 9.725092846013053e-06,
"loss": 0.9099,
"step": 192
},
{
"epoch": 0.5941889551664422,
"grad_norm": 0.27246885305968543,
"learning_rate": 9.71920534026237e-06,
"loss": 0.9605,
"step": 193
},
{
"epoch": 0.5972676544160093,
"grad_norm": 0.23938025849898661,
"learning_rate": 9.713257276867924e-06,
"loss": 0.9081,
"step": 194
},
{
"epoch": 0.6003463536655763,
"grad_norm": 0.2521959731991198,
"learning_rate": 9.707248732156274e-06,
"loss": 0.937,
"step": 195
},
{
"epoch": 0.6034250529151434,
"grad_norm": 0.24892629508891376,
"learning_rate": 9.701179783230088e-06,
"loss": 0.9237,
"step": 196
},
{
"epoch": 0.6065037521647104,
"grad_norm": 0.2516558798889323,
"learning_rate": 9.695050507967145e-06,
"loss": 0.9223,
"step": 197
},
{
"epoch": 0.6095824514142775,
"grad_norm": 0.25081879986925115,
"learning_rate": 9.688860985019348e-06,
"loss": 0.9493,
"step": 198
},
{
"epoch": 0.6126611506638445,
"grad_norm": 0.27100048855836784,
"learning_rate": 9.682611293811704e-06,
"loss": 0.9355,
"step": 199
},
{
"epoch": 0.6157398499134116,
"grad_norm": 0.2731442288453324,
"learning_rate": 9.676301514541311e-06,
"loss": 0.9368,
"step": 200
},
{
"epoch": 0.6188185491629786,
"grad_norm": 0.2248830717063508,
"learning_rate": 9.669931728176329e-06,
"loss": 0.9137,
"step": 201
},
{
"epoch": 0.6218972484125457,
"grad_norm": 0.2673560732569597,
"learning_rate": 9.663502016454937e-06,
"loss": 0.9042,
"step": 202
},
{
"epoch": 0.6249759476621127,
"grad_norm": 0.251566160215187,
"learning_rate": 9.657012461884285e-06,
"loss": 0.9308,
"step": 203
},
{
"epoch": 0.6280546469116798,
"grad_norm": 0.25347001920892376,
"learning_rate": 9.650463147739442e-06,
"loss": 0.9447,
"step": 204
},
{
"epoch": 0.6311333461612468,
"grad_norm": 0.24723958201077512,
"learning_rate": 9.643854158062315e-06,
"loss": 0.855,
"step": 205
},
{
"epoch": 0.6342120454108139,
"grad_norm": 0.27243908559333607,
"learning_rate": 9.637185577660584e-06,
"loss": 0.9414,
"step": 206
},
{
"epoch": 0.6372907446603809,
"grad_norm": 0.26152941452151107,
"learning_rate": 9.630457492106602e-06,
"loss": 0.9395,
"step": 207
},
{
"epoch": 0.6403694439099481,
"grad_norm": 0.2679507120250142,
"learning_rate": 9.623669987736305e-06,
"loss": 0.9253,
"step": 208
},
{
"epoch": 0.6434481431595152,
"grad_norm": 0.26408938555095496,
"learning_rate": 9.616823151648102e-06,
"loss": 0.9409,
"step": 209
},
{
"epoch": 0.6465268424090822,
"grad_norm": 0.24037285760759936,
"learning_rate": 9.609917071701755e-06,
"loss": 0.9384,
"step": 210
},
{
"epoch": 0.6496055416586493,
"grad_norm": 0.2621085059959121,
"learning_rate": 9.602951836517255e-06,
"loss": 0.898,
"step": 211
},
{
"epoch": 0.6526842409082163,
"grad_norm": 0.25387811904429614,
"learning_rate": 9.59592753547368e-06,
"loss": 0.9294,
"step": 212
},
{
"epoch": 0.6557629401577834,
"grad_norm": 0.23831478348050672,
"learning_rate": 9.588844258708052e-06,
"loss": 0.9402,
"step": 213
},
{
"epoch": 0.6588416394073504,
"grad_norm": 0.2694438278229194,
"learning_rate": 9.581702097114185e-06,
"loss": 0.975,
"step": 214
},
{
"epoch": 0.6619203386569175,
"grad_norm": 0.2549062894598875,
"learning_rate": 9.574501142341504e-06,
"loss": 0.9138,
"step": 215
},
{
"epoch": 0.6649990379064845,
"grad_norm": 0.23398616679439604,
"learning_rate": 9.567241486793886e-06,
"loss": 0.9515,
"step": 216
},
{
"epoch": 0.6680777371560516,
"grad_norm": 0.2688284108472295,
"learning_rate": 9.55992322362846e-06,
"loss": 0.9292,
"step": 217
},
{
"epoch": 0.6711564364056186,
"grad_norm": 0.2595731722752511,
"learning_rate": 9.552546446754422e-06,
"loss": 0.9383,
"step": 218
},
{
"epoch": 0.6742351356551857,
"grad_norm": 0.22880020455914685,
"learning_rate": 9.545111250831824e-06,
"loss": 0.9084,
"step": 219
},
{
"epoch": 0.6773138349047527,
"grad_norm": 0.25604300001093033,
"learning_rate": 9.537617731270365e-06,
"loss": 0.8863,
"step": 220
},
{
"epoch": 0.6803925341543198,
"grad_norm": 0.2615729587570963,
"learning_rate": 9.530065984228156e-06,
"loss": 0.9623,
"step": 221
},
{
"epoch": 0.6834712334038868,
"grad_norm": 0.255036573776491,
"learning_rate": 9.522456106610498e-06,
"loss": 0.929,
"step": 222
},
{
"epoch": 0.6865499326534539,
"grad_norm": 0.25522828525519575,
"learning_rate": 9.514788196068632e-06,
"loss": 0.8734,
"step": 223
},
{
"epoch": 0.6896286319030209,
"grad_norm": 0.2402149556038119,
"learning_rate": 9.507062350998487e-06,
"loss": 0.8809,
"step": 224
},
{
"epoch": 0.692707331152588,
"grad_norm": 0.24506423616289832,
"learning_rate": 9.499278670539418e-06,
"loss": 0.9152,
"step": 225
},
{
"epoch": 0.6957860304021551,
"grad_norm": 0.26223634908604687,
"learning_rate": 9.491437254572933e-06,
"loss": 0.9155,
"step": 226
},
{
"epoch": 0.6988647296517222,
"grad_norm": 0.25465239662910116,
"learning_rate": 9.483538203721414e-06,
"loss": 0.9653,
"step": 227
},
{
"epoch": 0.7019434289012892,
"grad_norm": 0.23524142363511988,
"learning_rate": 9.47558161934682e-06,
"loss": 0.9127,
"step": 228
},
{
"epoch": 0.7050221281508563,
"grad_norm": 0.24530725696344688,
"learning_rate": 9.467567603549392e-06,
"loss": 0.9203,
"step": 229
},
{
"epoch": 0.7081008274004233,
"grad_norm": 0.2476471310174333,
"learning_rate": 9.459496259166343e-06,
"loss": 0.9494,
"step": 230
},
{
"epoch": 0.7111795266499904,
"grad_norm": 0.24577569618596184,
"learning_rate": 9.451367689770532e-06,
"loss": 0.9217,
"step": 231
},
{
"epoch": 0.7142582258995575,
"grad_norm": 0.2419821287707517,
"learning_rate": 9.443181999669143e-06,
"loss": 0.9246,
"step": 232
},
{
"epoch": 0.7173369251491245,
"grad_norm": 0.26511892010794064,
"learning_rate": 9.434939293902335e-06,
"loss": 0.9192,
"step": 233
},
{
"epoch": 0.7204156243986916,
"grad_norm": 0.24277890729406734,
"learning_rate": 9.426639678241914e-06,
"loss": 0.9207,
"step": 234
},
{
"epoch": 0.7234943236482586,
"grad_norm": 0.24300036864615013,
"learning_rate": 9.41828325918995e-06,
"loss": 0.911,
"step": 235
},
{
"epoch": 0.7265730228978257,
"grad_norm": 0.2541366329593059,
"learning_rate": 9.409870143977438e-06,
"loss": 0.9016,
"step": 236
},
{
"epoch": 0.7296517221473927,
"grad_norm": 0.24718483596416485,
"learning_rate": 9.401400440562893e-06,
"loss": 0.9274,
"step": 237
},
{
"epoch": 0.7327304213969598,
"grad_norm": 0.23585305648873262,
"learning_rate": 9.392874257630988e-06,
"loss": 0.9087,
"step": 238
},
{
"epoch": 0.7358091206465268,
"grad_norm": 0.2785680207281239,
"learning_rate": 9.384291704591153e-06,
"loss": 0.9256,
"step": 239
},
{
"epoch": 0.7388878198960939,
"grad_norm": 0.24557628095729012,
"learning_rate": 9.375652891576163e-06,
"loss": 0.9046,
"step": 240
},
{
"epoch": 0.7419665191456609,
"grad_norm": 0.2434854835037153,
"learning_rate": 9.366957929440731e-06,
"loss": 0.9469,
"step": 241
},
{
"epoch": 0.745045218395228,
"grad_norm": 0.26522439728838126,
"learning_rate": 9.358206929760088e-06,
"loss": 0.9401,
"step": 242
},
{
"epoch": 0.748123917644795,
"grad_norm": 0.24137846923199488,
"learning_rate": 9.349400004828548e-06,
"loss": 0.945,
"step": 243
},
{
"epoch": 0.7512026168943622,
"grad_norm": 0.25412041832010346,
"learning_rate": 9.340537267658064e-06,
"loss": 0.915,
"step": 244
},
{
"epoch": 0.7542813161439292,
"grad_norm": 0.24935296746994054,
"learning_rate": 9.331618831976782e-06,
"loss": 0.9136,
"step": 245
},
{
"epoch": 0.7573600153934963,
"grad_norm": 0.2470742269178257,
"learning_rate": 9.322644812227587e-06,
"loss": 0.9549,
"step": 246
},
{
"epoch": 0.7604387146430633,
"grad_norm": 0.2510380628672779,
"learning_rate": 9.313615323566621e-06,
"loss": 0.9045,
"step": 247
},
{
"epoch": 0.7635174138926304,
"grad_norm": 0.2447996805455651,
"learning_rate": 9.304530481861813e-06,
"loss": 0.9283,
"step": 248
},
{
"epoch": 0.7665961131421974,
"grad_norm": 0.2412718972768966,
"learning_rate": 9.295390403691395e-06,
"loss": 0.9155,
"step": 249
},
{
"epoch": 0.7696748123917645,
"grad_norm": 0.2605769253823783,
"learning_rate": 9.286195206342404e-06,
"loss": 0.9322,
"step": 250
},
{
"epoch": 0.7727535116413315,
"grad_norm": 0.24926945860104782,
"learning_rate": 9.276945007809167e-06,
"loss": 0.914,
"step": 251
},
{
"epoch": 0.7758322108908986,
"grad_norm": 0.24069286338357718,
"learning_rate": 9.267639926791805e-06,
"loss": 0.9134,
"step": 252
},
{
"epoch": 0.7789109101404657,
"grad_norm": 0.2505756133581526,
"learning_rate": 9.258280082694694e-06,
"loss": 0.9098,
"step": 253
},
{
"epoch": 0.7819896093900327,
"grad_norm": 0.2476662054875445,
"learning_rate": 9.248865595624947e-06,
"loss": 0.9508,
"step": 254
},
{
"epoch": 0.7850683086395998,
"grad_norm": 0.24459106386513568,
"learning_rate": 9.239396586390854e-06,
"loss": 0.9045,
"step": 255
},
{
"epoch": 0.7881470078891668,
"grad_norm": 0.2399233207613046,
"learning_rate": 9.229873176500352e-06,
"loss": 0.9277,
"step": 256
},
{
"epoch": 0.7912257071387339,
"grad_norm": 0.26417900861068605,
"learning_rate": 9.22029548815945e-06,
"loss": 0.9659,
"step": 257
},
{
"epoch": 0.7943044063883009,
"grad_norm": 0.2360832927748832,
"learning_rate": 9.210663644270672e-06,
"loss": 0.9257,
"step": 258
},
{
"epoch": 0.797383105637868,
"grad_norm": 0.23330199065591498,
"learning_rate": 9.200977768431474e-06,
"loss": 0.9687,
"step": 259
},
{
"epoch": 0.800461804887435,
"grad_norm": 0.24860171214517035,
"learning_rate": 9.191237984932656e-06,
"loss": 0.9282,
"step": 260
},
{
"epoch": 0.8035405041370021,
"grad_norm": 0.22767494399427152,
"learning_rate": 9.181444418756774e-06,
"loss": 0.9111,
"step": 261
},
{
"epoch": 0.8066192033865692,
"grad_norm": 0.24483794507894094,
"learning_rate": 9.171597195576532e-06,
"loss": 0.9286,
"step": 262
},
{
"epoch": 0.8096979026361363,
"grad_norm": 0.25502892315669484,
"learning_rate": 9.161696441753167e-06,
"loss": 0.914,
"step": 263
},
{
"epoch": 0.8127766018857033,
"grad_norm": 0.24781715946503066,
"learning_rate": 9.151742284334834e-06,
"loss": 0.9178,
"step": 264
},
{
"epoch": 0.8158553011352704,
"grad_norm": 0.24240328752026663,
"learning_rate": 9.141734851054969e-06,
"loss": 0.914,
"step": 265
},
{
"epoch": 0.8189340003848374,
"grad_norm": 0.24149521222843484,
"learning_rate": 9.131674270330653e-06,
"loss": 0.8769,
"step": 266
},
{
"epoch": 0.8220126996344045,
"grad_norm": 0.24626041826138415,
"learning_rate": 9.121560671260968e-06,
"loss": 0.9273,
"step": 267
},
{
"epoch": 0.8250913988839715,
"grad_norm": 0.24247596840475402,
"learning_rate": 9.111394183625333e-06,
"loss": 0.9188,
"step": 268
},
{
"epoch": 0.8281700981335386,
"grad_norm": 0.24162197389893814,
"learning_rate": 9.101174937881841e-06,
"loss": 0.918,
"step": 269
},
{
"epoch": 0.8312487973831056,
"grad_norm": 0.2676910419589146,
"learning_rate": 9.090903065165587e-06,
"loss": 0.9141,
"step": 270
},
{
"epoch": 0.8343274966326727,
"grad_norm": 0.23389115971701657,
"learning_rate": 9.080578697286986e-06,
"loss": 0.925,
"step": 271
},
{
"epoch": 0.8374061958822397,
"grad_norm": 0.26843163168505285,
"learning_rate": 9.07020196673008e-06,
"loss": 0.914,
"step": 272
},
{
"epoch": 0.8404848951318068,
"grad_norm": 0.2679719572105889,
"learning_rate": 9.059773006650832e-06,
"loss": 0.9356,
"step": 273
},
{
"epoch": 0.8435635943813738,
"grad_norm": 0.24019662302658948,
"learning_rate": 9.049291950875435e-06,
"loss": 0.9095,
"step": 274
},
{
"epoch": 0.8466422936309409,
"grad_norm": 0.255470845756434,
"learning_rate": 9.038758933898565e-06,
"loss": 0.9568,
"step": 275
},
{
"epoch": 0.849720992880508,
"grad_norm": 0.256075237040761,
"learning_rate": 9.028174090881694e-06,
"loss": 0.8859,
"step": 276
},
{
"epoch": 0.852799692130075,
"grad_norm": 0.23343452771104067,
"learning_rate": 9.017537557651316e-06,
"loss": 0.9119,
"step": 277
},
{
"epoch": 0.8558783913796421,
"grad_norm": 0.2616444815662004,
"learning_rate": 9.006849470697236e-06,
"loss": 0.9553,
"step": 278
},
{
"epoch": 0.8589570906292091,
"grad_norm": 0.24195580043829323,
"learning_rate": 8.996109967170796e-06,
"loss": 0.8643,
"step": 279
},
{
"epoch": 0.8620357898787763,
"grad_norm": 0.2447067333820804,
"learning_rate": 8.985319184883127e-06,
"loss": 0.8679,
"step": 280
},
{
"epoch": 0.8651144891283433,
"grad_norm": 0.2445135337424023,
"learning_rate": 8.974477262303382e-06,
"loss": 0.8961,
"step": 281
},
{
"epoch": 0.8681931883779104,
"grad_norm": 0.24033969492056376,
"learning_rate": 8.963584338556948e-06,
"loss": 0.8795,
"step": 282
},
{
"epoch": 0.8712718876274774,
"grad_norm": 0.23504632439717768,
"learning_rate": 8.952640553423673e-06,
"loss": 0.9369,
"step": 283
},
{
"epoch": 0.8743505868770445,
"grad_norm": 0.26348662667677963,
"learning_rate": 8.941646047336064e-06,
"loss": 0.9412,
"step": 284
},
{
"epoch": 0.8774292861266115,
"grad_norm": 0.2625380471901545,
"learning_rate": 8.93060096137749e-06,
"loss": 0.9093,
"step": 285
},
{
"epoch": 0.8805079853761786,
"grad_norm": 0.2522531299992897,
"learning_rate": 8.919505437280366e-06,
"loss": 0.8939,
"step": 286
},
{
"epoch": 0.8835866846257456,
"grad_norm": 0.2498527765567612,
"learning_rate": 8.908359617424345e-06,
"loss": 0.8899,
"step": 287
},
{
"epoch": 0.8866653838753127,
"grad_norm": 0.2559324978504737,
"learning_rate": 8.897163644834472e-06,
"loss": 0.9378,
"step": 288
},
{
"epoch": 0.8897440831248797,
"grad_norm": 0.2329641019045212,
"learning_rate": 8.885917663179371e-06,
"loss": 0.9091,
"step": 289
},
{
"epoch": 0.8928227823744468,
"grad_norm": 0.27100973316639154,
"learning_rate": 8.874621816769389e-06,
"loss": 0.9304,
"step": 290
},
{
"epoch": 0.8959014816240138,
"grad_norm": 0.26364969728483967,
"learning_rate": 8.86327625055474e-06,
"loss": 0.9165,
"step": 291
},
{
"epoch": 0.8989801808735809,
"grad_norm": 0.24358531724284999,
"learning_rate": 8.851881110123657e-06,
"loss": 0.9452,
"step": 292
},
{
"epoch": 0.9020588801231479,
"grad_norm": 0.24837138582806928,
"learning_rate": 8.840436541700515e-06,
"loss": 0.943,
"step": 293
},
{
"epoch": 0.905137579372715,
"grad_norm": 0.2646736009084181,
"learning_rate": 8.82894269214396e-06,
"loss": 0.9132,
"step": 294
},
{
"epoch": 0.908216278622282,
"grad_norm": 0.261926097716352,
"learning_rate": 8.817399708945018e-06,
"loss": 0.9039,
"step": 295
},
{
"epoch": 0.9112949778718491,
"grad_norm": 0.2501938695223563,
"learning_rate": 8.805807740225207e-06,
"loss": 0.9328,
"step": 296
},
{
"epoch": 0.9143736771214162,
"grad_norm": 0.2425731674872489,
"learning_rate": 8.79416693473464e-06,
"loss": 0.8444,
"step": 297
},
{
"epoch": 0.9174523763709833,
"grad_norm": 0.24339585760416285,
"learning_rate": 8.782477441850107e-06,
"loss": 0.9472,
"step": 298
},
{
"epoch": 0.9205310756205504,
"grad_norm": 0.24331436629555378,
"learning_rate": 8.770739411573164e-06,
"loss": 0.9357,
"step": 299
},
{
"epoch": 0.9236097748701174,
"grad_norm": 0.23517434497640016,
"learning_rate": 8.758952994528208e-06,
"loss": 0.8795,
"step": 300
},
{
"epoch": 0.9266884741196845,
"grad_norm": 0.26126800407154716,
"learning_rate": 8.747118341960542e-06,
"loss": 0.9356,
"step": 301
},
{
"epoch": 0.9297671733692515,
"grad_norm": 0.2395042827726231,
"learning_rate": 8.735235605734436e-06,
"loss": 0.917,
"step": 302
},
{
"epoch": 0.9328458726188186,
"grad_norm": 0.23987296153453452,
"learning_rate": 8.723304938331179e-06,
"loss": 0.9328,
"step": 303
},
{
"epoch": 0.9359245718683856,
"grad_norm": 0.2354958840068119,
"learning_rate": 8.711326492847118e-06,
"loss": 0.9062,
"step": 304
},
{
"epoch": 0.9390032711179527,
"grad_norm": 0.24458661570469958,
"learning_rate": 8.699300422991701e-06,
"loss": 0.9352,
"step": 305
},
{
"epoch": 0.9420819703675197,
"grad_norm": 0.2330187574455436,
"learning_rate": 8.687226883085493e-06,
"loss": 0.9148,
"step": 306
},
{
"epoch": 0.9451606696170868,
"grad_norm": 0.22939630899981844,
"learning_rate": 8.67510602805821e-06,
"loss": 0.8972,
"step": 307
},
{
"epoch": 0.9482393688666538,
"grad_norm": 0.25182153759362846,
"learning_rate": 8.66293801344672e-06,
"loss": 0.9299,
"step": 308
},
{
"epoch": 0.9513180681162209,
"grad_norm": 0.24672424752793506,
"learning_rate": 8.65072299539305e-06,
"loss": 0.9015,
"step": 309
},
{
"epoch": 0.9543967673657879,
"grad_norm": 0.25564364998716094,
"learning_rate": 8.638461130642386e-06,
"loss": 0.9269,
"step": 310
},
{
"epoch": 0.957475466615355,
"grad_norm": 0.26106061104069905,
"learning_rate": 8.626152576541059e-06,
"loss": 0.9323,
"step": 311
},
{
"epoch": 0.960554165864922,
"grad_norm": 0.2154522969322482,
"learning_rate": 8.613797491034519e-06,
"loss": 0.8467,
"step": 312
},
{
"epoch": 0.9636328651144891,
"grad_norm": 0.2516090261360356,
"learning_rate": 8.601396032665324e-06,
"loss": 0.9251,
"step": 313
},
{
"epoch": 0.9667115643640561,
"grad_norm": 0.2424330279727654,
"learning_rate": 8.58894836057109e-06,
"loss": 0.9029,
"step": 314
},
{
"epoch": 0.9697902636136232,
"grad_norm": 0.2665862929938747,
"learning_rate": 8.576454634482458e-06,
"loss": 0.9465,
"step": 315
},
{
"epoch": 0.9728689628631904,
"grad_norm": 0.24540550890749058,
"learning_rate": 8.563915014721045e-06,
"loss": 0.9172,
"step": 316
},
{
"epoch": 0.9759476621127574,
"grad_norm": 0.2579462768052683,
"learning_rate": 8.551329662197375e-06,
"loss": 0.9139,
"step": 317
},
{
"epoch": 0.9790263613623245,
"grad_norm": 0.23775188839748934,
"learning_rate": 8.538698738408828e-06,
"loss": 0.8793,
"step": 318
},
{
"epoch": 0.9821050606118915,
"grad_norm": 0.2559438515430679,
"learning_rate": 8.526022405437565e-06,
"loss": 0.9317,
"step": 319
},
{
"epoch": 0.9851837598614586,
"grad_norm": 0.2508159788351336,
"learning_rate": 8.513300825948436e-06,
"loss": 0.9154,
"step": 320
},
{
"epoch": 0.9882624591110256,
"grad_norm": 0.24311433081309117,
"learning_rate": 8.500534163186914e-06,
"loss": 0.8971,
"step": 321
},
{
"epoch": 0.9913411583605927,
"grad_norm": 0.25732947760140373,
"learning_rate": 8.487722580976977e-06,
"loss": 0.9161,
"step": 322
},
{
"epoch": 0.9944198576101597,
"grad_norm": 0.24071041461109438,
"learning_rate": 8.474866243719024e-06,
"loss": 0.928,
"step": 323
},
{
"epoch": 0.9974985568597268,
"grad_norm": 0.2666156544425345,
"learning_rate": 8.461965316387759e-06,
"loss": 0.9493,
"step": 324
},
{
"epoch": 1.0,
"grad_norm": 0.2666156544425345,
"learning_rate": 8.449019964530065e-06,
"loss": 0.8833,
"step": 325
},
{
"epoch": 1.003078699249567,
"grad_norm": 0.276759892676397,
"learning_rate": 8.4360303542629e-06,
"loss": 0.9416,
"step": 326
},
{
"epoch": 1.006157398499134,
"grad_norm": 0.27090854392048047,
"learning_rate": 8.422996652271143e-06,
"loss": 0.9122,
"step": 327
},
{
"epoch": 1.0092360977487012,
"grad_norm": 0.2968021684014539,
"learning_rate": 8.409919025805471e-06,
"loss": 0.9364,
"step": 328
},
{
"epoch": 1.0123147969982682,
"grad_norm": 0.26301093836859835,
"learning_rate": 8.396797642680205e-06,
"loss": 0.9032,
"step": 329
},
{
"epoch": 1.0153934962478353,
"grad_norm": 0.25622718341459877,
"learning_rate": 8.383632671271162e-06,
"loss": 0.8799,
"step": 330
},
{
"epoch": 1.0184721954974023,
"grad_norm": 0.26329095862965446,
"learning_rate": 8.37042428051349e-06,
"loss": 0.8607,
"step": 331
},
{
"epoch": 1.0215508947469694,
"grad_norm": 0.23872200571746977,
"learning_rate": 8.357172639899497e-06,
"loss": 0.8219,
"step": 332
},
{
"epoch": 1.0246295939965364,
"grad_norm": 0.24769122354136078,
"learning_rate": 8.343877919476486e-06,
"loss": 0.8664,
"step": 333
},
{
"epoch": 1.0277082932461035,
"grad_norm": 0.24927869216831874,
"learning_rate": 8.330540289844566e-06,
"loss": 0.8688,
"step": 334
},
{
"epoch": 1.0307869924956705,
"grad_norm": 0.2630434117089861,
"learning_rate": 8.317159922154463e-06,
"loss": 0.9166,
"step": 335
},
{
"epoch": 1.0338656917452376,
"grad_norm": 0.2501345489679188,
"learning_rate": 8.303736988105326e-06,
"loss": 0.8992,
"step": 336
},
{
"epoch": 1.0369443909948046,
"grad_norm": 0.2655982399906758,
"learning_rate": 8.290271659942524e-06,
"loss": 0.8811,
"step": 337
},
{
"epoch": 1.0400230902443717,
"grad_norm": 0.26817452173113604,
"learning_rate": 8.27676411045543e-06,
"loss": 0.9298,
"step": 338
},
{
"epoch": 1.0431017894939387,
"grad_norm": 0.25352374987893683,
"learning_rate": 8.263214512975214e-06,
"loss": 0.8751,
"step": 339
},
{
"epoch": 1.0461804887435058,
"grad_norm": 0.24456010559575803,
"learning_rate": 8.249623041372607e-06,
"loss": 0.8632,
"step": 340
},
{
"epoch": 1.0492591879930728,
"grad_norm": 0.24605085932290968,
"learning_rate": 8.235989870055684e-06,
"loss": 0.8581,
"step": 341
},
{
"epoch": 1.0523378872426399,
"grad_norm": 0.24548585907146017,
"learning_rate": 8.222315173967608e-06,
"loss": 0.8801,
"step": 342
},
{
"epoch": 1.055416586492207,
"grad_norm": 0.25243848179044737,
"learning_rate": 8.208599128584403e-06,
"loss": 0.8744,
"step": 343
},
{
"epoch": 1.0584952857417742,
"grad_norm": 0.2613742390485315,
"learning_rate": 8.19484190991269e-06,
"loss": 0.8822,
"step": 344
},
{
"epoch": 1.0615739849913413,
"grad_norm": 0.27640899653285855,
"learning_rate": 8.181043694487432e-06,
"loss": 0.8771,
"step": 345
},
{
"epoch": 1.0646526842409083,
"grad_norm": 0.2411328280122734,
"learning_rate": 8.167204659369672e-06,
"loss": 0.8723,
"step": 346
},
{
"epoch": 1.0677313834904754,
"grad_norm": 0.25503023111748185,
"learning_rate": 8.153324982144258e-06,
"loss": 0.8816,
"step": 347
},
{
"epoch": 1.0708100827400424,
"grad_norm": 0.24109726096376402,
"learning_rate": 8.139404840917562e-06,
"loss": 0.8826,
"step": 348
},
{
"epoch": 1.0738887819896095,
"grad_norm": 0.24384501105176618,
"learning_rate": 8.1254444143152e-06,
"loss": 0.8946,
"step": 349
},
{
"epoch": 1.0769674812391765,
"grad_norm": 0.252676337846711,
"learning_rate": 8.111443881479733e-06,
"loss": 0.8905,
"step": 350
},
{
"epoch": 1.0800461804887436,
"grad_norm": 0.24079271356089385,
"learning_rate": 8.097403422068377e-06,
"loss": 0.8565,
"step": 351
},
{
"epoch": 1.0831248797383106,
"grad_norm": 0.24833636824472707,
"learning_rate": 8.083323216250688e-06,
"loss": 0.8233,
"step": 352
},
{
"epoch": 1.0862035789878777,
"grad_norm": 0.24596066617557716,
"learning_rate": 8.069203444706257e-06,
"loss": 0.8778,
"step": 353
},
{
"epoch": 1.0892822782374447,
"grad_norm": 0.23310480888169238,
"learning_rate": 8.055044288622386e-06,
"loss": 0.839,
"step": 354
},
{
"epoch": 1.0923609774870118,
"grad_norm": 0.24370013295224682,
"learning_rate": 8.040845929691773e-06,
"loss": 0.8896,
"step": 355
},
{
"epoch": 1.0954396767365788,
"grad_norm": 0.25216464853384907,
"learning_rate": 8.026608550110166e-06,
"loss": 0.8647,
"step": 356
},
{
"epoch": 1.0985183759861459,
"grad_norm": 0.24717504718218447,
"learning_rate": 8.012332332574038e-06,
"loss": 0.9123,
"step": 357
},
{
"epoch": 1.101597075235713,
"grad_norm": 0.23606234331466158,
"learning_rate": 7.998017460278234e-06,
"loss": 0.8821,
"step": 358
},
{
"epoch": 1.10467577448528,
"grad_norm": 0.2513527804245468,
"learning_rate": 7.983664116913624e-06,
"loss": 0.8959,
"step": 359
},
{
"epoch": 1.107754473734847,
"grad_norm": 0.23910605915498864,
"learning_rate": 7.969272486664747e-06,
"loss": 0.8522,
"step": 360
},
{
"epoch": 1.110833172984414,
"grad_norm": 0.24578423029160462,
"learning_rate": 7.954842754207442e-06,
"loss": 0.8785,
"step": 361
},
{
"epoch": 1.1139118722339811,
"grad_norm": 0.2412660980323421,
"learning_rate": 7.940375104706489e-06,
"loss": 0.8631,
"step": 362
},
{
"epoch": 1.1169905714835482,
"grad_norm": 0.22734778957818583,
"learning_rate": 7.925869723813218e-06,
"loss": 0.8684,
"step": 363
},
{
"epoch": 1.1200692707331152,
"grad_norm": 0.23640200262600752,
"learning_rate": 7.911326797663138e-06,
"loss": 0.8347,
"step": 364
},
{
"epoch": 1.1231479699826823,
"grad_norm": 0.2516491175592187,
"learning_rate": 7.896746512873546e-06,
"loss": 0.8661,
"step": 365
},
{
"epoch": 1.1262266692322493,
"grad_norm": 0.24485818662669678,
"learning_rate": 7.882129056541133e-06,
"loss": 0.9021,
"step": 366
},
{
"epoch": 1.1293053684818164,
"grad_norm": 0.24486888166012932,
"learning_rate": 7.867474616239577e-06,
"loss": 0.8974,
"step": 367
},
{
"epoch": 1.1323840677313834,
"grad_norm": 0.3774973799331707,
"learning_rate": 7.852783380017141e-06,
"loss": 0.8867,
"step": 368
},
{
"epoch": 1.1354627669809505,
"grad_norm": 0.24662461903207444,
"learning_rate": 7.838055536394265e-06,
"loss": 0.8623,
"step": 369
},
{
"epoch": 1.1385414662305176,
"grad_norm": 0.2580890837937278,
"learning_rate": 7.823291274361136e-06,
"loss": 0.889,
"step": 370
},
{
"epoch": 1.1416201654800846,
"grad_norm": 0.2800472733025288,
"learning_rate": 7.80849078337527e-06,
"loss": 0.8906,
"step": 371
},
{
"epoch": 1.1446988647296517,
"grad_norm": 0.2551077932110902,
"learning_rate": 7.793654253359075e-06,
"loss": 0.8814,
"step": 372
},
{
"epoch": 1.1477775639792187,
"grad_norm": 0.2484613658437565,
"learning_rate": 7.77878187469743e-06,
"loss": 0.8642,
"step": 373
},
{
"epoch": 1.1508562632287858,
"grad_norm": 0.2397134768422901,
"learning_rate": 7.763873838235216e-06,
"loss": 0.8516,
"step": 374
},
{
"epoch": 1.1539349624783528,
"grad_norm": 0.26406158108618993,
"learning_rate": 7.748930335274887e-06,
"loss": 0.8997,
"step": 375
},
{
"epoch": 1.1570136617279199,
"grad_norm": 0.2587850138989264,
"learning_rate": 7.733951557574012e-06,
"loss": 0.8468,
"step": 376
},
{
"epoch": 1.1600923609774871,
"grad_norm": 0.22841683602304946,
"learning_rate": 7.718937697342802e-06,
"loss": 0.8557,
"step": 377
},
{
"epoch": 1.1631710602270542,
"grad_norm": 0.2683694073837402,
"learning_rate": 7.703888947241668e-06,
"loss": 0.8955,
"step": 378
},
{
"epoch": 1.1662497594766212,
"grad_norm": 0.24330472023039235,
"learning_rate": 7.68880550037872e-06,
"loss": 0.8686,
"step": 379
},
{
"epoch": 1.1693284587261883,
"grad_norm": 0.2331856448745946,
"learning_rate": 7.673687550307308e-06,
"loss": 0.8498,
"step": 380
},
{
"epoch": 1.1724071579757553,
"grad_norm": 0.2527940753364362,
"learning_rate": 7.658535291023533e-06,
"loss": 0.8742,
"step": 381
},
{
"epoch": 1.1754858572253224,
"grad_norm": 0.251692326508692,
"learning_rate": 7.643348916963755e-06,
"loss": 0.8949,
"step": 382
},
{
"epoch": 1.1785645564748894,
"grad_norm": 0.24410157430512422,
"learning_rate": 7.628128623002109e-06,
"loss": 0.8926,
"step": 383
},
{
"epoch": 1.1816432557244565,
"grad_norm": 0.2544986201563911,
"learning_rate": 7.612874604447985e-06,
"loss": 0.9003,
"step": 384
},
{
"epoch": 1.1847219549740236,
"grad_norm": 0.23760937967454676,
"learning_rate": 7.59758705704354e-06,
"loss": 0.8897,
"step": 385
},
{
"epoch": 1.1878006542235906,
"grad_norm": 0.22920066895791236,
"learning_rate": 7.582266176961175e-06,
"loss": 0.923,
"step": 386
},
{
"epoch": 1.1908793534731577,
"grad_norm": 0.37366931660455954,
"learning_rate": 7.566912160801023e-06,
"loss": 0.8293,
"step": 387
},
{
"epoch": 1.1939580527227247,
"grad_norm": 0.23886266375375065,
"learning_rate": 7.551525205588427e-06,
"loss": 0.923,
"step": 388
},
{
"epoch": 1.1970367519722918,
"grad_norm": 0.24355815696341995,
"learning_rate": 7.536105508771406e-06,
"loss": 0.8358,
"step": 389
},
{
"epoch": 1.2001154512218588,
"grad_norm": 0.24456303129864032,
"learning_rate": 7.520653268218128e-06,
"loss": 0.878,
"step": 390
},
{
"epoch": 1.2031941504714259,
"grad_norm": 0.23510010519030677,
"learning_rate": 7.5051686822143614e-06,
"loss": 0.851,
"step": 391
},
{
"epoch": 1.206272849720993,
"grad_norm": 0.24056264428587013,
"learning_rate": 7.489651949460946e-06,
"loss": 0.9,
"step": 392
},
{
"epoch": 1.20935154897056,
"grad_norm": 0.25045425993409975,
"learning_rate": 7.474103269071229e-06,
"loss": 0.8734,
"step": 393
},
{
"epoch": 1.212430248220127,
"grad_norm": 0.2361198205002829,
"learning_rate": 7.458522840568514e-06,
"loss": 0.8338,
"step": 394
},
{
"epoch": 1.215508947469694,
"grad_norm": 0.2347453401935896,
"learning_rate": 7.442910863883503e-06,
"loss": 0.9115,
"step": 395
},
{
"epoch": 1.2185876467192611,
"grad_norm": 0.2555871938784729,
"learning_rate": 7.427267539351733e-06,
"loss": 0.8555,
"step": 396
},
{
"epoch": 1.2216663459688282,
"grad_norm": 0.24177152473452024,
"learning_rate": 7.411593067710997e-06,
"loss": 0.8723,
"step": 397
},
{
"epoch": 1.2247450452183952,
"grad_norm": 0.2398290360619593,
"learning_rate": 7.395887650098776e-06,
"loss": 0.858,
"step": 398
},
{
"epoch": 1.2278237444679623,
"grad_norm": 0.2501157619952234,
"learning_rate": 7.3801514880496525e-06,
"loss": 0.854,
"step": 399
},
{
"epoch": 1.2309024437175293,
"grad_norm": 0.24350765665261256,
"learning_rate": 7.364384783492731e-06,
"loss": 0.8926,
"step": 400
},
{
"epoch": 1.2339811429670964,
"grad_norm": 0.2581239968022416,
"learning_rate": 7.348587738749037e-06,
"loss": 0.9007,
"step": 401
},
{
"epoch": 1.2370598422166634,
"grad_norm": 0.23500264563000425,
"learning_rate": 7.3327605565289305e-06,
"loss": 0.8751,
"step": 402
},
{
"epoch": 1.2401385414662305,
"grad_norm": 0.24674781755632724,
"learning_rate": 7.3169034399295006e-06,
"loss": 0.8579,
"step": 403
},
{
"epoch": 1.2432172407157975,
"grad_norm": 0.260794717177315,
"learning_rate": 7.301016592431956e-06,
"loss": 0.8577,
"step": 404
},
{
"epoch": 1.2462959399653646,
"grad_norm": 0.3422777320372874,
"learning_rate": 7.285100217899022e-06,
"loss": 0.8439,
"step": 405
},
{
"epoch": 1.2493746392149316,
"grad_norm": 0.2401581844807946,
"learning_rate": 7.269154520572318e-06,
"loss": 0.8597,
"step": 406
},
{
"epoch": 1.2524533384644987,
"grad_norm": 0.2802438278869203,
"learning_rate": 7.253179705069735e-06,
"loss": 0.8695,
"step": 407
},
{
"epoch": 1.2555320377140657,
"grad_norm": 0.2417004699919177,
"learning_rate": 7.23717597638282e-06,
"loss": 0.8651,
"step": 408
},
{
"epoch": 1.2586107369636328,
"grad_norm": 0.24895061369525212,
"learning_rate": 7.221143539874131e-06,
"loss": 0.8401,
"step": 409
},
{
"epoch": 1.2616894362131998,
"grad_norm": 0.24666450390315126,
"learning_rate": 7.20508260127462e-06,
"loss": 0.8763,
"step": 410
},
{
"epoch": 1.264768135462767,
"grad_norm": 0.2642714535987437,
"learning_rate": 7.18899336668097e-06,
"loss": 0.8723,
"step": 411
},
{
"epoch": 1.267846834712334,
"grad_norm": 0.2379693959579241,
"learning_rate": 7.172876042552974e-06,
"loss": 0.8346,
"step": 412
},
{
"epoch": 1.270925533961901,
"grad_norm": 0.24554776990714927,
"learning_rate": 7.156730835710868e-06,
"loss": 0.8685,
"step": 413
},
{
"epoch": 1.274004233211468,
"grad_norm": 0.2417424217609596,
"learning_rate": 7.1405579533326876e-06,
"loss": 0.8694,
"step": 414
},
{
"epoch": 1.277082932461035,
"grad_norm": 0.24935469137402416,
"learning_rate": 7.124357602951603e-06,
"loss": 0.8592,
"step": 415
},
{
"epoch": 1.2801616317106022,
"grad_norm": 0.23667109405552306,
"learning_rate": 7.108129992453261e-06,
"loss": 0.8867,
"step": 416
},
{
"epoch": 1.2832403309601692,
"grad_norm": 0.2425262476107213,
"learning_rate": 7.09187533007311e-06,
"loss": 0.8773,
"step": 417
},
{
"epoch": 1.2863190302097363,
"grad_norm": 0.24342150786412103,
"learning_rate": 7.075593824393741e-06,
"loss": 0.9117,
"step": 418
},
{
"epoch": 1.2893977294593033,
"grad_norm": 0.25491314971382445,
"learning_rate": 7.059285684342192e-06,
"loss": 0.8682,
"step": 419
},
{
"epoch": 1.2924764287088706,
"grad_norm": 0.2726178828428502,
"learning_rate": 7.042951119187288e-06,
"loss": 0.8512,
"step": 420
},
{
"epoch": 1.2955551279584376,
"grad_norm": 0.2516555626004353,
"learning_rate": 7.0265903385369385e-06,
"loss": 0.8644,
"step": 421
},
{
"epoch": 1.2986338272080047,
"grad_norm": 0.24242400828974608,
"learning_rate": 7.0102035523354585e-06,
"loss": 0.8751,
"step": 422
},
{
"epoch": 1.3017125264575717,
"grad_norm": 0.27869928444728226,
"learning_rate": 6.99379097086087e-06,
"loss": 0.9051,
"step": 423
},
{
"epoch": 1.3047912257071388,
"grad_norm": 0.260078269310079,
"learning_rate": 6.977352804722201e-06,
"loss": 0.8539,
"step": 424
},
{
"epoch": 1.3078699249567058,
"grad_norm": 0.2362581225364301,
"learning_rate": 6.960889264856791e-06,
"loss": 0.827,
"step": 425
},
{
"epoch": 1.310948624206273,
"grad_norm": 0.26039581350120705,
"learning_rate": 6.944400562527578e-06,
"loss": 0.8884,
"step": 426
},
{
"epoch": 1.31402732345584,
"grad_norm": 0.25825506352353667,
"learning_rate": 6.927886909320387e-06,
"loss": 0.8873,
"step": 427
},
{
"epoch": 1.317106022705407,
"grad_norm": 0.26252134534222793,
"learning_rate": 6.911348517141217e-06,
"loss": 0.8661,
"step": 428
},
{
"epoch": 1.320184721954974,
"grad_norm": 0.23127822617076133,
"learning_rate": 6.894785598213527e-06,
"loss": 0.8798,
"step": 429
},
{
"epoch": 1.323263421204541,
"grad_norm": 0.2594025831183444,
"learning_rate": 6.878198365075499e-06,
"loss": 0.8957,
"step": 430
},
{
"epoch": 1.3263421204541082,
"grad_norm": 0.25977921959733313,
"learning_rate": 6.861587030577326e-06,
"loss": 0.8459,
"step": 431
},
{
"epoch": 1.3294208197036752,
"grad_norm": 0.2420663106025989,
"learning_rate": 6.8449518078784705e-06,
"loss": 0.8804,
"step": 432
},
{
"epoch": 1.3324995189532423,
"grad_norm": 0.25897501182438815,
"learning_rate": 6.828292910444934e-06,
"loss": 0.8555,
"step": 433
},
{
"epoch": 1.3355782182028093,
"grad_norm": 0.24571682800330732,
"learning_rate": 6.8116105520465126e-06,
"loss": 0.8724,
"step": 434
},
{
"epoch": 1.3386569174523764,
"grad_norm": 0.24880251274901471,
"learning_rate": 6.794904946754064e-06,
"loss": 0.8471,
"step": 435
},
{
"epoch": 1.3417356167019434,
"grad_norm": 0.23621380512390353,
"learning_rate": 6.778176308936746e-06,
"loss": 0.8764,
"step": 436
},
{
"epoch": 1.3448143159515105,
"grad_norm": 0.25075266956968867,
"learning_rate": 6.761424853259283e-06,
"loss": 0.886,
"step": 437
},
{
"epoch": 1.3478930152010775,
"grad_norm": 0.22699315284125088,
"learning_rate": 6.7446507946791925e-06,
"loss": 0.8486,
"step": 438
},
{
"epoch": 1.3509717144506446,
"grad_norm": 0.23999154565690825,
"learning_rate": 6.727854348444043e-06,
"loss": 0.8532,
"step": 439
},
{
"epoch": 1.3540504137002116,
"grad_norm": 0.24046431813741648,
"learning_rate": 6.7110357300886805e-06,
"loss": 0.8727,
"step": 440
},
{
"epoch": 1.3571291129497787,
"grad_norm": 0.2385504751910788,
"learning_rate": 6.694195155432473e-06,
"loss": 0.8415,
"step": 441
},
{
"epoch": 1.3602078121993457,
"grad_norm": 0.24788130069993006,
"learning_rate": 6.6773328405765306e-06,
"loss": 0.8931,
"step": 442
},
{
"epoch": 1.3632865114489128,
"grad_norm": 0.24313251565685237,
"learning_rate": 6.660449001900939e-06,
"loss": 0.9209,
"step": 443
},
{
"epoch": 1.3663652106984798,
"grad_norm": 0.2588425558653781,
"learning_rate": 6.6435438560619815e-06,
"loss": 0.841,
"step": 444
},
{
"epoch": 1.3694439099480469,
"grad_norm": 0.23380607999181446,
"learning_rate": 6.626617619989358e-06,
"loss": 0.8768,
"step": 445
},
{
"epoch": 1.372522609197614,
"grad_norm": 0.23681671872611712,
"learning_rate": 6.6096705108834015e-06,
"loss": 0.8686,
"step": 446
},
{
"epoch": 1.3756013084471812,
"grad_norm": 0.3009839781407332,
"learning_rate": 6.592702746212295e-06,
"loss": 0.8981,
"step": 447
},
{
"epoch": 1.3786800076967483,
"grad_norm": 0.2421462987251025,
"learning_rate": 6.575714543709271e-06,
"loss": 0.9261,
"step": 448
},
{
"epoch": 1.3817587069463153,
"grad_norm": 0.24584487621733495,
"learning_rate": 6.558706121369827e-06,
"loss": 0.9026,
"step": 449
},
{
"epoch": 1.3848374061958824,
"grad_norm": 0.247985734035166,
"learning_rate": 6.541677697448926e-06,
"loss": 0.9009,
"step": 450
},
{
"epoch": 1.3879161054454494,
"grad_norm": 0.2509600346645963,
"learning_rate": 6.524629490458193e-06,
"loss": 0.8411,
"step": 451
},
{
"epoch": 1.3909948046950165,
"grad_norm": 0.2395699074316301,
"learning_rate": 6.507561719163111e-06,
"loss": 0.9004,
"step": 452
},
{
"epoch": 1.3940735039445835,
"grad_norm": 0.30989500337363945,
"learning_rate": 6.4904746025802234e-06,
"loss": 0.8447,
"step": 453
},
{
"epoch": 1.3971522031941506,
"grad_norm": 0.23394208187731705,
"learning_rate": 6.473368359974304e-06,
"loss": 0.86,
"step": 454
},
{
"epoch": 1.4002309024437176,
"grad_norm": 0.23406693839354947,
"learning_rate": 6.456243210855564e-06,
"loss": 0.8568,
"step": 455
},
{
"epoch": 1.4033096016932847,
"grad_norm": 0.24716674248727222,
"learning_rate": 6.439099374976821e-06,
"loss": 0.8816,
"step": 456
},
{
"epoch": 1.4063883009428517,
"grad_norm": 0.22958629427032134,
"learning_rate": 6.421937072330686e-06,
"loss": 0.9013,
"step": 457
},
{
"epoch": 1.4094670001924188,
"grad_norm": 0.2544822790940788,
"learning_rate": 6.4047565231467425e-06,
"loss": 0.8716,
"step": 458
},
{
"epoch": 1.4125456994419858,
"grad_norm": 0.24454391635771133,
"learning_rate": 6.387557947888708e-06,
"loss": 0.8492,
"step": 459
},
{
"epoch": 1.4156243986915529,
"grad_norm": 0.24131147663283706,
"learning_rate": 6.370341567251624e-06,
"loss": 0.8285,
"step": 460
},
{
"epoch": 1.41870309794112,
"grad_norm": 0.226963659812761,
"learning_rate": 6.353107602159005e-06,
"loss": 0.8719,
"step": 461
},
{
"epoch": 1.421781797190687,
"grad_norm": 0.24573614375186503,
"learning_rate": 6.335856273760016e-06,
"loss": 0.8799,
"step": 462
},
{
"epoch": 1.424860496440254,
"grad_norm": 0.2428767635572903,
"learning_rate": 6.3185878034266326e-06,
"loss": 0.8802,
"step": 463
},
{
"epoch": 1.427939195689821,
"grad_norm": 0.24499614078691787,
"learning_rate": 6.301302412750794e-06,
"loss": 0.8745,
"step": 464
},
{
"epoch": 1.4310178949393881,
"grad_norm": 0.24538110774401264,
"learning_rate": 6.284000323541566e-06,
"loss": 0.9203,
"step": 465
},
{
"epoch": 1.4340965941889552,
"grad_norm": 0.2530483639210842,
"learning_rate": 6.266681757822297e-06,
"loss": 0.8815,
"step": 466
},
{
"epoch": 1.4371752934385222,
"grad_norm": 0.2579369068858789,
"learning_rate": 6.249346937827759e-06,
"loss": 0.9187,
"step": 467
},
{
"epoch": 1.4402539926880893,
"grad_norm": 0.2575619166966343,
"learning_rate": 6.231996086001303e-06,
"loss": 0.8803,
"step": 468
},
{
"epoch": 1.4433326919376563,
"grad_norm": 0.23596783211373998,
"learning_rate": 6.214629424992007e-06,
"loss": 0.8791,
"step": 469
},
{
"epoch": 1.4464113911872234,
"grad_norm": 0.2511743368078798,
"learning_rate": 6.197247177651808e-06,
"loss": 0.8608,
"step": 470
},
{
"epoch": 1.4494900904367904,
"grad_norm": 0.24013481845493814,
"learning_rate": 6.179849567032658e-06,
"loss": 0.89,
"step": 471
},
{
"epoch": 1.4525687896863575,
"grad_norm": 0.2465400460862126,
"learning_rate": 6.162436816383646e-06,
"loss": 0.8844,
"step": 472
},
{
"epoch": 1.4556474889359245,
"grad_norm": 0.24200307935042567,
"learning_rate": 6.145009149148142e-06,
"loss": 0.8968,
"step": 473
},
{
"epoch": 1.4587261881854916,
"grad_norm": 0.24277985241327188,
"learning_rate": 6.127566788960935e-06,
"loss": 0.867,
"step": 474
},
{
"epoch": 1.4618048874350587,
"grad_norm": 0.26199481085375487,
"learning_rate": 6.110109959645344e-06,
"loss": 0.9029,
"step": 475
},
{
"epoch": 1.4648835866846257,
"grad_norm": 0.23873594126425465,
"learning_rate": 6.092638885210369e-06,
"loss": 0.8648,
"step": 476
},
{
"epoch": 1.4679622859341928,
"grad_norm": 0.2449147600308907,
"learning_rate": 6.075153789847801e-06,
"loss": 0.8731,
"step": 477
},
{
"epoch": 1.4710409851837598,
"grad_norm": 0.2373131733606055,
"learning_rate": 6.0576548979293525e-06,
"loss": 0.8741,
"step": 478
},
{
"epoch": 1.4741196844333269,
"grad_norm": 0.24544346533262942,
"learning_rate": 6.0401424340037705e-06,
"loss": 0.8867,
"step": 479
},
{
"epoch": 1.477198383682894,
"grad_norm": 0.24269570169470212,
"learning_rate": 6.022616622793969e-06,
"loss": 0.8401,
"step": 480
},
{
"epoch": 1.480277082932461,
"grad_norm": 0.24434127590898086,
"learning_rate": 6.0050776891941266e-06,
"loss": 0.8315,
"step": 481
},
{
"epoch": 1.483355782182028,
"grad_norm": 0.25012135006923913,
"learning_rate": 5.987525858266818e-06,
"loss": 0.8773,
"step": 482
},
{
"epoch": 1.486434481431595,
"grad_norm": 0.25176819102821424,
"learning_rate": 5.9699613552401145e-06,
"loss": 0.8804,
"step": 483
},
{
"epoch": 1.4895131806811621,
"grad_norm": 0.262185648234918,
"learning_rate": 5.952384405504703e-06,
"loss": 0.8952,
"step": 484
},
{
"epoch": 1.4925918799307292,
"grad_norm": 0.2500176630733897,
"learning_rate": 5.9347952346109824e-06,
"loss": 0.8968,
"step": 485
},
{
"epoch": 1.4956705791802962,
"grad_norm": 0.2549862952498234,
"learning_rate": 5.917194068266179e-06,
"loss": 0.8677,
"step": 486
},
{
"epoch": 1.4987492784298633,
"grad_norm": 0.2377822433309426,
"learning_rate": 5.899581132331446e-06,
"loss": 0.8744,
"step": 487
},
{
"epoch": 1.5018279776794303,
"grad_norm": 0.2417562885543654,
"learning_rate": 5.881956652818966e-06,
"loss": 0.8423,
"step": 488
},
{
"epoch": 1.5049066769289974,
"grad_norm": 0.23092627154601364,
"learning_rate": 5.8643208558890516e-06,
"loss": 0.8968,
"step": 489
},
{
"epoch": 1.5079853761785644,
"grad_norm": 0.25968304565833644,
"learning_rate": 5.8466739678472426e-06,
"loss": 0.8909,
"step": 490
},
{
"epoch": 1.5110640754281315,
"grad_norm": 0.24557476319425794,
"learning_rate": 5.8290162151414e-06,
"loss": 0.8707,
"step": 491
},
{
"epoch": 1.5141427746776985,
"grad_norm": 0.2452082579855338,
"learning_rate": 5.811347824358802e-06,
"loss": 0.9042,
"step": 492
},
{
"epoch": 1.5172214739272656,
"grad_norm": 0.25131797222944846,
"learning_rate": 5.793669022223239e-06,
"loss": 0.8763,
"step": 493
},
{
"epoch": 1.5203001731768326,
"grad_norm": 0.2638758092461222,
"learning_rate": 5.775980035592098e-06,
"loss": 0.9035,
"step": 494
},
{
"epoch": 1.5233788724263997,
"grad_norm": 0.25559996117246625,
"learning_rate": 5.758281091453461e-06,
"loss": 0.8907,
"step": 495
},
{
"epoch": 1.5264575716759667,
"grad_norm": 0.23072795601371907,
"learning_rate": 5.740572416923179e-06,
"loss": 0.8538,
"step": 496
},
{
"epoch": 1.529536270925534,
"grad_norm": 0.2535835882079426,
"learning_rate": 5.722854239241971e-06,
"loss": 0.9082,
"step": 497
},
{
"epoch": 1.532614970175101,
"grad_norm": 0.24800327861842403,
"learning_rate": 5.7051267857724965e-06,
"loss": 0.8402,
"step": 498
},
{
"epoch": 1.5356936694246681,
"grad_norm": 0.2508552282322579,
"learning_rate": 5.68739028399645e-06,
"loss": 0.7883,
"step": 499
},
{
"epoch": 1.5387723686742352,
"grad_norm": 0.23232439878442204,
"learning_rate": 5.6696449615116314e-06,
"loss": 0.8832,
"step": 500
},
{
"epoch": 1.5418510679238022,
"grad_norm": 0.24776471835592684,
"learning_rate": 5.651891046029031e-06,
"loss": 0.8611,
"step": 501
},
{
"epoch": 1.5449297671733693,
"grad_norm": 0.24717292660725904,
"learning_rate": 5.634128765369902e-06,
"loss": 0.8664,
"step": 502
},
{
"epoch": 1.5480084664229363,
"grad_norm": 0.2469179768790754,
"learning_rate": 5.616358347462847e-06,
"loss": 0.9128,
"step": 503
},
{
"epoch": 1.5510871656725034,
"grad_norm": 0.25637224771479067,
"learning_rate": 5.598580020340882e-06,
"loss": 0.8807,
"step": 504
},
{
"epoch": 1.5541658649220704,
"grad_norm": 0.23609502599794216,
"learning_rate": 5.5807940121385186e-06,
"loss": 0.8673,
"step": 505
},
{
"epoch": 1.5572445641716375,
"grad_norm": 0.2352207907434502,
"learning_rate": 5.563000551088833e-06,
"loss": 0.88,
"step": 506
},
{
"epoch": 1.5603232634212045,
"grad_norm": 0.24915747308495176,
"learning_rate": 5.545199865520536e-06,
"loss": 0.8818,
"step": 507
},
{
"epoch": 1.5634019626707716,
"grad_norm": 0.24539770056290555,
"learning_rate": 5.527392183855046e-06,
"loss": 0.8698,
"step": 508
},
{
"epoch": 1.5664806619203386,
"grad_norm": 0.24042631802911352,
"learning_rate": 5.509577734603555e-06,
"loss": 0.8913,
"step": 509
},
{
"epoch": 1.5695593611699057,
"grad_norm": 0.24676753632621698,
"learning_rate": 5.4917567463641e-06,
"loss": 0.8683,
"step": 510
},
{
"epoch": 1.5726380604194727,
"grad_norm": 0.24281166024392842,
"learning_rate": 5.473929447818628e-06,
"loss": 0.8374,
"step": 511
},
{
"epoch": 1.5757167596690398,
"grad_norm": 0.2336216249509978,
"learning_rate": 5.456096067730055e-06,
"loss": 0.8314,
"step": 512
},
{
"epoch": 1.5787954589186068,
"grad_norm": 0.2214822240534207,
"learning_rate": 5.438256834939342e-06,
"loss": 0.8556,
"step": 513
},
{
"epoch": 1.581874158168174,
"grad_norm": 0.24139899272130835,
"learning_rate": 5.420411978362552e-06,
"loss": 0.8271,
"step": 514
},
{
"epoch": 1.5849528574177412,
"grad_norm": 0.22981685176192268,
"learning_rate": 5.402561726987915e-06,
"loss": 0.802,
"step": 515
},
{
"epoch": 1.5880315566673082,
"grad_norm": 0.24845253985190824,
"learning_rate": 5.384706309872881e-06,
"loss": 0.9108,
"step": 516
},
{
"epoch": 1.5911102559168753,
"grad_norm": 0.23650374314411407,
"learning_rate": 5.366845956141198e-06,
"loss": 0.8297,
"step": 517
},
{
"epoch": 1.5941889551664423,
"grad_norm": 0.2369145684361407,
"learning_rate": 5.348980894979951e-06,
"loss": 0.8884,
"step": 518
},
{
"epoch": 1.5972676544160094,
"grad_norm": 0.23481629273009993,
"learning_rate": 5.3311113556366404e-06,
"loss": 0.8698,
"step": 519
},
{
"epoch": 1.6003463536655764,
"grad_norm": 0.2427789463974632,
"learning_rate": 5.313237567416224e-06,
"loss": 0.8809,
"step": 520
},
{
"epoch": 1.6034250529151435,
"grad_norm": 0.31248916737080856,
"learning_rate": 5.295359759678187e-06,
"loss": 0.9098,
"step": 521
},
{
"epoch": 1.6065037521647105,
"grad_norm": 0.24565475392977196,
"learning_rate": 5.277478161833594e-06,
"loss": 0.8692,
"step": 522
},
{
"epoch": 1.6095824514142776,
"grad_norm": 0.24144950582200592,
"learning_rate": 5.259593003342138e-06,
"loss": 0.8849,
"step": 523
},
{
"epoch": 1.6126611506638446,
"grad_norm": 0.23618064055210236,
"learning_rate": 5.241704513709213e-06,
"loss": 0.8904,
"step": 524
},
{
"epoch": 1.6157398499134117,
"grad_norm": 0.23192335315198373,
"learning_rate": 5.22381292248295e-06,
"loss": 0.9112,
"step": 525
},
{
"epoch": 1.6188185491629787,
"grad_norm": 0.24804915047390189,
"learning_rate": 5.2059184592512864e-06,
"loss": 0.8384,
"step": 526
},
{
"epoch": 1.6218972484125458,
"grad_norm": 0.23657591556142857,
"learning_rate": 5.18802135363901e-06,
"loss": 0.8896,
"step": 527
},
{
"epoch": 1.6249759476621128,
"grad_norm": 0.25661164289224825,
"learning_rate": 5.170121835304818e-06,
"loss": 0.825,
"step": 528
},
{
"epoch": 1.62805464691168,
"grad_norm": 0.22713739743316588,
"learning_rate": 5.152220133938364e-06,
"loss": 0.8752,
"step": 529
},
{
"epoch": 1.631133346161247,
"grad_norm": 0.2393388454446188,
"learning_rate": 5.1343164792573204e-06,
"loss": 0.9063,
"step": 530
},
{
"epoch": 1.634212045410814,
"grad_norm": 0.24195130131172482,
"learning_rate": 5.1164111010044225e-06,
"loss": 0.8426,
"step": 531
},
{
"epoch": 1.637290744660381,
"grad_norm": 0.2479064925915565,
"learning_rate": 5.098504228944522e-06,
"loss": 0.8821,
"step": 532
},
{
"epoch": 1.640369443909948,
"grad_norm": 0.2430787701253668,
"learning_rate": 5.080596092861641e-06,
"loss": 0.8928,
"step": 533
},
{
"epoch": 1.6434481431595152,
"grad_norm": 0.2524614819929656,
"learning_rate": 5.0626869225560174e-06,
"loss": 0.8479,
"step": 534
},
{
"epoch": 1.6465268424090822,
"grad_norm": 0.2372236816765915,
"learning_rate": 5.0447769478411654e-06,
"loss": 0.901,
"step": 535
},
{
"epoch": 1.6496055416586493,
"grad_norm": 0.23629154787657172,
"learning_rate": 5.0268663985409216e-06,
"loss": 0.8722,
"step": 536
},
{
"epoch": 1.6526842409082163,
"grad_norm": 0.23211360891084543,
"learning_rate": 5.008955504486491e-06,
"loss": 0.8565,
"step": 537
},
{
"epoch": 1.6557629401577834,
"grad_norm": 0.2680035889407426,
"learning_rate": 4.99104449551351e-06,
"loss": 0.86,
"step": 538
},
{
"epoch": 1.6588416394073504,
"grad_norm": 0.23142653603744853,
"learning_rate": 4.973133601459078e-06,
"loss": 0.9066,
"step": 539
},
{
"epoch": 1.6619203386569175,
"grad_norm": 0.2683607949064942,
"learning_rate": 4.955223052158835e-06,
"loss": 0.8352,
"step": 540
},
{
"epoch": 1.6649990379064845,
"grad_norm": 0.2314028530159739,
"learning_rate": 4.937313077443985e-06,
"loss": 0.9023,
"step": 541
},
{
"epoch": 1.6680777371560516,
"grad_norm": 0.24465655709081577,
"learning_rate": 4.919403907138362e-06,
"loss": 0.8455,
"step": 542
},
{
"epoch": 1.6711564364056186,
"grad_norm": 0.24221647311028913,
"learning_rate": 4.901495771055479e-06,
"loss": 0.859,
"step": 543
},
{
"epoch": 1.6742351356551857,
"grad_norm": 0.2438562351326107,
"learning_rate": 4.883588898995577e-06,
"loss": 0.8677,
"step": 544
},
{
"epoch": 1.6773138349047527,
"grad_norm": 0.22179954218294268,
"learning_rate": 4.86568352074268e-06,
"loss": 0.8345,
"step": 545
},
{
"epoch": 1.6803925341543198,
"grad_norm": 0.22486217782488802,
"learning_rate": 4.847779866061638e-06,
"loss": 0.856,
"step": 546
},
{
"epoch": 1.6834712334038868,
"grad_norm": 0.22459513926965324,
"learning_rate": 4.829878164695184e-06,
"loss": 0.8337,
"step": 547
},
{
"epoch": 1.6865499326534539,
"grad_norm": 0.25026926825281587,
"learning_rate": 4.811978646360991e-06,
"loss": 0.8884,
"step": 548
},
{
"epoch": 1.689628631903021,
"grad_norm": 0.2519507206705291,
"learning_rate": 4.7940815407487135e-06,
"loss": 0.8858,
"step": 549
},
{
"epoch": 1.692707331152588,
"grad_norm": 0.2416624320103976,
"learning_rate": 4.776187077517051e-06,
"loss": 0.9082,
"step": 550
},
{
"epoch": 1.695786030402155,
"grad_norm": 0.22886853595753984,
"learning_rate": 4.7582954862907896e-06,
"loss": 0.8651,
"step": 551
},
{
"epoch": 1.698864729651722,
"grad_norm": 0.2174759404387797,
"learning_rate": 4.740406996657863e-06,
"loss": 0.847,
"step": 552
},
{
"epoch": 1.7019434289012891,
"grad_norm": 0.2390584461884235,
"learning_rate": 4.72252183816641e-06,
"loss": 0.9128,
"step": 553
},
{
"epoch": 1.7050221281508562,
"grad_norm": 0.24068572667684823,
"learning_rate": 4.704640240321814e-06,
"loss": 0.8867,
"step": 554
},
{
"epoch": 1.7081008274004232,
"grad_norm": 0.25564639109351095,
"learning_rate": 4.686762432583778e-06,
"loss": 0.8205,
"step": 555
},
{
"epoch": 1.7111795266499903,
"grad_norm": 0.23442371346359767,
"learning_rate": 4.668888644363361e-06,
"loss": 0.8601,
"step": 556
},
{
"epoch": 1.7142582258995573,
"grad_norm": 0.23669424988320334,
"learning_rate": 4.6510191050200495e-06,
"loss": 0.8639,
"step": 557
},
{
"epoch": 1.7173369251491244,
"grad_norm": 0.24224283295855903,
"learning_rate": 4.633154043858804e-06,
"loss": 0.8655,
"step": 558
},
{
"epoch": 1.7204156243986914,
"grad_norm": 0.2472891047838766,
"learning_rate": 4.615293690127119e-06,
"loss": 0.8984,
"step": 559
},
{
"epoch": 1.7234943236482585,
"grad_norm": 0.25670654965763573,
"learning_rate": 4.597438273012087e-06,
"loss": 0.8858,
"step": 560
},
{
"epoch": 1.7265730228978255,
"grad_norm": 0.24585591404221188,
"learning_rate": 4.579588021637448e-06,
"loss": 0.8544,
"step": 561
},
{
"epoch": 1.7296517221473926,
"grad_norm": 0.23946263045870972,
"learning_rate": 4.561743165060659e-06,
"loss": 0.8725,
"step": 562
},
{
"epoch": 1.7327304213969597,
"grad_norm": 0.24934788523823928,
"learning_rate": 4.543903932269949e-06,
"loss": 0.8766,
"step": 563
},
{
"epoch": 1.7358091206465267,
"grad_norm": 0.22647992817986462,
"learning_rate": 4.526070552181374e-06,
"loss": 0.8763,
"step": 564
},
{
"epoch": 1.7388878198960938,
"grad_norm": 0.2239399825518789,
"learning_rate": 4.508243253635902e-06,
"loss": 0.8542,
"step": 565
},
{
"epoch": 1.7419665191456608,
"grad_norm": 0.24046313227548927,
"learning_rate": 4.4904222653964445e-06,
"loss": 0.8967,
"step": 566
},
{
"epoch": 1.7450452183952279,
"grad_norm": 0.24890549930398903,
"learning_rate": 4.472607816144956e-06,
"loss": 0.8448,
"step": 567
},
{
"epoch": 1.748123917644795,
"grad_norm": 0.2338921023931779,
"learning_rate": 4.454800134479466e-06,
"loss": 0.8223,
"step": 568
},
{
"epoch": 1.7512026168943622,
"grad_norm": 0.2209606159647853,
"learning_rate": 4.436999448911168e-06,
"loss": 0.8629,
"step": 569
},
{
"epoch": 1.7542813161439292,
"grad_norm": 0.24329770193799344,
"learning_rate": 4.419205987861483e-06,
"loss": 0.8368,
"step": 570
},
{
"epoch": 1.7573600153934963,
"grad_norm": 0.24016562567034758,
"learning_rate": 4.401419979659119e-06,
"loss": 0.8349,
"step": 571
},
{
"epoch": 1.7604387146430633,
"grad_norm": 0.22294522400811564,
"learning_rate": 4.383641652537155e-06,
"loss": 0.9005,
"step": 572
},
{
"epoch": 1.7635174138926304,
"grad_norm": 0.2511298514120106,
"learning_rate": 4.365871234630101e-06,
"loss": 0.8636,
"step": 573
},
{
"epoch": 1.7665961131421974,
"grad_norm": 0.24165576408568865,
"learning_rate": 4.3481089539709705e-06,
"loss": 0.8515,
"step": 574
},
{
"epoch": 1.7696748123917645,
"grad_norm": 0.23926532393325203,
"learning_rate": 4.33035503848837e-06,
"loss": 0.862,
"step": 575
},
{
"epoch": 1.7727535116413315,
"grad_norm": 0.2409141305177723,
"learning_rate": 4.31260971600355e-06,
"loss": 0.8578,
"step": 576
},
{
"epoch": 1.7758322108908986,
"grad_norm": 0.2490140427509988,
"learning_rate": 4.294873214227505e-06,
"loss": 0.8826,
"step": 577
},
{
"epoch": 1.7789109101404657,
"grad_norm": 0.2546884111216467,
"learning_rate": 4.277145760758032e-06,
"loss": 0.8795,
"step": 578
},
{
"epoch": 1.7819896093900327,
"grad_norm": 0.23359237129821397,
"learning_rate": 4.2594275830768225e-06,
"loss": 0.8722,
"step": 579
},
{
"epoch": 1.7850683086395998,
"grad_norm": 0.22529731028176653,
"learning_rate": 4.241718908546541e-06,
"loss": 0.8909,
"step": 580
},
{
"epoch": 1.7881470078891668,
"grad_norm": 0.23962117322820914,
"learning_rate": 4.224019964407902e-06,
"loss": 0.8282,
"step": 581
},
{
"epoch": 1.7912257071387339,
"grad_norm": 0.2400805130904326,
"learning_rate": 4.206330977776762e-06,
"loss": 0.8528,
"step": 582
},
{
"epoch": 1.794304406388301,
"grad_norm": 0.2807390854625172,
"learning_rate": 4.1886521756412e-06,
"loss": 0.849,
"step": 583
},
{
"epoch": 1.797383105637868,
"grad_norm": 0.2317008711241012,
"learning_rate": 4.170983784858602e-06,
"loss": 0.8706,
"step": 584
},
{
"epoch": 1.800461804887435,
"grad_norm": 0.22960222770075844,
"learning_rate": 4.153326032152758e-06,
"loss": 0.8979,
"step": 585
},
{
"epoch": 1.803540504137002,
"grad_norm": 0.24162402039603836,
"learning_rate": 4.135679144110948e-06,
"loss": 0.8559,
"step": 586
},
{
"epoch": 1.8066192033865693,
"grad_norm": 0.22925569420079514,
"learning_rate": 4.118043347181035e-06,
"loss": 0.8688,
"step": 587
},
{
"epoch": 1.8096979026361364,
"grad_norm": 0.23930420877726746,
"learning_rate": 4.100418867668557e-06,
"loss": 0.8758,
"step": 588
},
{
"epoch": 1.8127766018857034,
"grad_norm": 0.2357947069249543,
"learning_rate": 4.0828059317338225e-06,
"loss": 0.8645,
"step": 589
},
{
"epoch": 1.8158553011352705,
"grad_norm": 0.23834944044372317,
"learning_rate": 4.06520476538902e-06,
"loss": 0.8976,
"step": 590
},
{
"epoch": 1.8189340003848375,
"grad_norm": 0.2366894097096441,
"learning_rate": 4.047615594495299e-06,
"loss": 0.8818,
"step": 591
},
{
"epoch": 1.8220126996344046,
"grad_norm": 0.24185311662924638,
"learning_rate": 4.030038644759886e-06,
"loss": 0.8905,
"step": 592
},
{
"epoch": 1.8250913988839716,
"grad_norm": 0.22383397441968267,
"learning_rate": 4.0124741417331835e-06,
"loss": 0.8464,
"step": 593
},
{
"epoch": 1.8281700981335387,
"grad_norm": 0.24151175687615029,
"learning_rate": 3.994922310805875e-06,
"loss": 0.8864,
"step": 594
},
{
"epoch": 1.8312487973831058,
"grad_norm": 0.23483772929765256,
"learning_rate": 3.9773833772060335e-06,
"loss": 0.8949,
"step": 595
},
{
"epoch": 1.8343274966326728,
"grad_norm": 0.25580014522845945,
"learning_rate": 3.9598575659962294e-06,
"loss": 0.8682,
"step": 596
},
{
"epoch": 1.8374061958822399,
"grad_norm": 0.232008669889148,
"learning_rate": 3.942345102070649e-06,
"loss": 0.9072,
"step": 597
},
{
"epoch": 1.840484895131807,
"grad_norm": 0.24782917036175173,
"learning_rate": 3.924846210152199e-06,
"loss": 0.8534,
"step": 598
},
{
"epoch": 1.843563594381374,
"grad_norm": 0.2337162729144535,
"learning_rate": 3.9073611147896315e-06,
"loss": 0.8606,
"step": 599
},
{
"epoch": 1.846642293630941,
"grad_norm": 0.2707053335834137,
"learning_rate": 3.889890040354659e-06,
"loss": 0.8895,
"step": 600
},
{
"epoch": 1.849720992880508,
"grad_norm": 0.23601057297125144,
"learning_rate": 3.872433211039067e-06,
"loss": 0.8533,
"step": 601
},
{
"epoch": 1.8527996921300751,
"grad_norm": 0.24429702714104823,
"learning_rate": 3.854990850851859e-06,
"loss": 0.8508,
"step": 602
},
{
"epoch": 1.8558783913796422,
"grad_norm": 0.25405237483458065,
"learning_rate": 3.837563183616355e-06,
"loss": 0.85,
"step": 603
},
{
"epoch": 1.8589570906292092,
"grad_norm": 0.24884917032187737,
"learning_rate": 3.820150432967344e-06,
"loss": 0.8489,
"step": 604
},
{
"epoch": 1.8620357898787763,
"grad_norm": 0.23586436270911693,
"learning_rate": 3.8027528223481936e-06,
"loss": 0.8649,
"step": 605
},
{
"epoch": 1.8651144891283433,
"grad_norm": 0.2404408423189326,
"learning_rate": 3.785370575007995e-06,
"loss": 0.8851,
"step": 606
},
{
"epoch": 1.8681931883779104,
"grad_norm": 0.268904935779624,
"learning_rate": 3.7680039139986983e-06,
"loss": 0.8141,
"step": 607
},
{
"epoch": 1.8712718876274774,
"grad_norm": 0.2560963567716389,
"learning_rate": 3.7506530621722426e-06,
"loss": 0.8657,
"step": 608
},
{
"epoch": 1.8743505868770445,
"grad_norm": 0.22112776156733008,
"learning_rate": 3.7333182421777048e-06,
"loss": 0.8501,
"step": 609
},
{
"epoch": 1.8774292861266115,
"grad_norm": 0.23275010767847898,
"learning_rate": 3.715999676458435e-06,
"loss": 0.8816,
"step": 610
},
{
"epoch": 1.8805079853761786,
"grad_norm": 0.25273993228256275,
"learning_rate": 3.698697587249208e-06,
"loss": 0.8775,
"step": 611
},
{
"epoch": 1.8835866846257456,
"grad_norm": 0.2423685469164429,
"learning_rate": 3.6814121965733695e-06,
"loss": 0.8487,
"step": 612
},
{
"epoch": 1.8866653838753127,
"grad_norm": 0.23605380649283772,
"learning_rate": 3.664143726239984e-06,
"loss": 0.8794,
"step": 613
},
{
"epoch": 1.8897440831248797,
"grad_norm": 0.2469587924579458,
"learning_rate": 3.646892397840997e-06,
"loss": 0.8548,
"step": 614
},
{
"epoch": 1.8928227823744468,
"grad_norm": 0.24024619025426996,
"learning_rate": 3.6296584327483787e-06,
"loss": 0.872,
"step": 615
},
{
"epoch": 1.8959014816240138,
"grad_norm": 0.22280138707417765,
"learning_rate": 3.6124420521112925e-06,
"loss": 0.8745,
"step": 616
},
{
"epoch": 1.898980180873581,
"grad_norm": 0.2403955030516092,
"learning_rate": 3.59524347685326e-06,
"loss": 0.8393,
"step": 617
},
{
"epoch": 1.902058880123148,
"grad_norm": 0.24589501379847223,
"learning_rate": 3.578062927669315e-06,
"loss": 0.8962,
"step": 618
},
{
"epoch": 1.905137579372715,
"grad_norm": 0.26035048416138823,
"learning_rate": 3.5609006250231813e-06,
"loss": 0.8605,
"step": 619
},
{
"epoch": 1.908216278622282,
"grad_norm": 0.2283286805772837,
"learning_rate": 3.543756789144439e-06,
"loss": 0.8697,
"step": 620
},
{
"epoch": 1.911294977871849,
"grad_norm": 0.2869524320638206,
"learning_rate": 3.526631640025697e-06,
"loss": 0.8625,
"step": 621
},
{
"epoch": 1.9143736771214162,
"grad_norm": 0.22540815839644504,
"learning_rate": 3.5095253974197787e-06,
"loss": 0.8742,
"step": 622
},
{
"epoch": 1.9174523763709832,
"grad_norm": 0.23573684088976804,
"learning_rate": 3.492438280836888e-06,
"loss": 0.8605,
"step": 623
},
{
"epoch": 1.9205310756205503,
"grad_norm": 0.2371944413886398,
"learning_rate": 3.4753705095418087e-06,
"loss": 0.8502,
"step": 624
},
{
"epoch": 1.9236097748701173,
"grad_norm": 0.22352165862953732,
"learning_rate": 3.4583223025510748e-06,
"loss": 0.8653,
"step": 625
},
{
"epoch": 1.9266884741196844,
"grad_norm": 0.23071394069235382,
"learning_rate": 3.4412938786301743e-06,
"loss": 0.8456,
"step": 626
},
{
"epoch": 1.9297671733692514,
"grad_norm": 0.22887282102384657,
"learning_rate": 3.424285456290732e-06,
"loss": 0.8157,
"step": 627
},
{
"epoch": 1.9328458726188185,
"grad_norm": 0.23003295147118233,
"learning_rate": 3.407297253787706e-06,
"loss": 0.8401,
"step": 628
},
{
"epoch": 1.9359245718683855,
"grad_norm": 0.23096386439155586,
"learning_rate": 3.3903294891165993e-06,
"loss": 0.9029,
"step": 629
},
{
"epoch": 1.9390032711179526,
"grad_norm": 0.24726568870350749,
"learning_rate": 3.3733823800106436e-06,
"loss": 0.8687,
"step": 630
},
{
"epoch": 1.9420819703675196,
"grad_norm": 0.23225658883834968,
"learning_rate": 3.3564561439380206e-06,
"loss": 0.8994,
"step": 631
},
{
"epoch": 1.9451606696170867,
"grad_norm": 0.22724014543505516,
"learning_rate": 3.339550998099063e-06,
"loss": 0.8721,
"step": 632
},
{
"epoch": 1.9482393688666537,
"grad_norm": 0.22955713730375288,
"learning_rate": 3.3226671594234707e-06,
"loss": 0.8451,
"step": 633
},
{
"epoch": 1.9513180681162208,
"grad_norm": 0.21911164301268593,
"learning_rate": 3.3058048445675286e-06,
"loss": 0.8712,
"step": 634
},
{
"epoch": 1.9543967673657878,
"grad_norm": 0.24518883269084193,
"learning_rate": 3.28896426991132e-06,
"loss": 0.8289,
"step": 635
},
{
"epoch": 1.9574754666153549,
"grad_norm": 0.22903243220331604,
"learning_rate": 3.272145651555959e-06,
"loss": 0.8915,
"step": 636
},
{
"epoch": 1.960554165864922,
"grad_norm": 0.22781003304559164,
"learning_rate": 3.2553492053208104e-06,
"loss": 0.872,
"step": 637
},
{
"epoch": 1.963632865114489,
"grad_norm": 0.2559599933164248,
"learning_rate": 3.2385751467407177e-06,
"loss": 0.8663,
"step": 638
},
{
"epoch": 1.966711564364056,
"grad_norm": 0.2538425627066515,
"learning_rate": 3.221823691063255e-06,
"loss": 0.8611,
"step": 639
},
{
"epoch": 1.969790263613623,
"grad_norm": 0.23082655732735588,
"learning_rate": 3.205095053245937e-06,
"loss": 0.8567,
"step": 640
},
{
"epoch": 1.9728689628631904,
"grad_norm": 0.23771137314633606,
"learning_rate": 3.1883894479534882e-06,
"loss": 0.8723,
"step": 641
},
{
"epoch": 1.9759476621127574,
"grad_norm": 0.23687498653687258,
"learning_rate": 3.171707089555068e-06,
"loss": 0.8593,
"step": 642
},
{
"epoch": 1.9790263613623245,
"grad_norm": 0.2256607516119969,
"learning_rate": 3.1550481921215304e-06,
"loss": 0.8671,
"step": 643
},
{
"epoch": 1.9821050606118915,
"grad_norm": 0.2453331059907523,
"learning_rate": 3.138412969422675e-06,
"loss": 0.8931,
"step": 644
},
{
"epoch": 1.9851837598614586,
"grad_norm": 0.23340410771364534,
"learning_rate": 3.121801634924502e-06,
"loss": 0.8803,
"step": 645
},
{
"epoch": 1.9882624591110256,
"grad_norm": 0.23723686717177803,
"learning_rate": 3.1052144017864743e-06,
"loss": 0.8238,
"step": 646
},
{
"epoch": 1.9913411583605927,
"grad_norm": 0.22576732125488122,
"learning_rate": 3.088651482858783e-06,
"loss": 0.8135,
"step": 647
},
{
"epoch": 1.9944198576101597,
"grad_norm": 0.23842775007620237,
"learning_rate": 3.0721130906796135e-06,
"loss": 0.8811,
"step": 648
},
{
"epoch": 1.9974985568597268,
"grad_norm": 0.2403327128762041,
"learning_rate": 3.055599437472423e-06,
"loss": 0.8847,
"step": 649
},
{
"epoch": 2.0,
"grad_norm": 0.2506144289292039,
"learning_rate": 3.0391107351432086e-06,
"loss": 0.8222,
"step": 650
},
{
"epoch": 2.003078699249567,
"grad_norm": 0.2632065169611864,
"learning_rate": 3.022647195277799e-06,
"loss": 0.8065,
"step": 651
},
{
"epoch": 2.006157398499134,
"grad_norm": 0.25927642693308345,
"learning_rate": 3.0062090291391325e-06,
"loss": 0.8438,
"step": 652
},
{
"epoch": 2.009236097748701,
"grad_norm": 0.2560434045520623,
"learning_rate": 2.9897964476645423e-06,
"loss": 0.8562,
"step": 653
},
{
"epoch": 2.012314796998268,
"grad_norm": 0.24947055861111583,
"learning_rate": 2.9734096614630636e-06,
"loss": 0.8382,
"step": 654
},
{
"epoch": 2.0153934962478353,
"grad_norm": 0.23943830172013983,
"learning_rate": 2.9570488808127142e-06,
"loss": 0.8561,
"step": 655
},
{
"epoch": 2.0184721954974023,
"grad_norm": 0.22585017452636458,
"learning_rate": 2.94071431565781e-06,
"loss": 0.8263,
"step": 656
},
{
"epoch": 2.0215508947469694,
"grad_norm": 0.2440194261106326,
"learning_rate": 2.9244061756062613e-06,
"loss": 0.8519,
"step": 657
},
{
"epoch": 2.0246295939965364,
"grad_norm": 0.2414504325024437,
"learning_rate": 2.9081246699268905e-06,
"loss": 0.8336,
"step": 658
},
{
"epoch": 2.0277082932461035,
"grad_norm": 0.24845312219934826,
"learning_rate": 2.891870007546742e-06,
"loss": 0.8501,
"step": 659
},
{
"epoch": 2.0307869924956705,
"grad_norm": 0.242209505988487,
"learning_rate": 2.8756423970483965e-06,
"loss": 0.8493,
"step": 660
},
{
"epoch": 2.0338656917452376,
"grad_norm": 0.23265587872012475,
"learning_rate": 2.8594420466673133e-06,
"loss": 0.8204,
"step": 661
},
{
"epoch": 2.0369443909948046,
"grad_norm": 0.23193701706877973,
"learning_rate": 2.843269164289133e-06,
"loss": 0.8061,
"step": 662
},
{
"epoch": 2.0400230902443717,
"grad_norm": 0.23385036315420626,
"learning_rate": 2.827123957447029e-06,
"loss": 0.8437,
"step": 663
},
{
"epoch": 2.0431017894939387,
"grad_norm": 0.26168296109621303,
"learning_rate": 2.8110066333190317e-06,
"loss": 0.854,
"step": 664
},
{
"epoch": 2.0461804887435058,
"grad_norm": 0.2327042837624201,
"learning_rate": 2.794917398725382e-06,
"loss": 0.8292,
"step": 665
},
{
"epoch": 2.049259187993073,
"grad_norm": 0.21951870573757737,
"learning_rate": 2.7788564601258706e-06,
"loss": 0.8259,
"step": 666
},
{
"epoch": 2.05233788724264,
"grad_norm": 0.22725324672124128,
"learning_rate": 2.762824023617182e-06,
"loss": 0.8361,
"step": 667
},
{
"epoch": 2.055416586492207,
"grad_norm": 0.24195419914404015,
"learning_rate": 2.746820294930265e-06,
"loss": 0.8132,
"step": 668
},
{
"epoch": 2.058495285741774,
"grad_norm": 0.2337475544403441,
"learning_rate": 2.730845479427684e-06,
"loss": 0.826,
"step": 669
},
{
"epoch": 2.061573984991341,
"grad_norm": 0.23801223385548173,
"learning_rate": 2.714899782100978e-06,
"loss": 0.8345,
"step": 670
},
{
"epoch": 2.064652684240908,
"grad_norm": 0.23387831898708886,
"learning_rate": 2.6989834075680452e-06,
"loss": 0.8181,
"step": 671
},
{
"epoch": 2.067731383490475,
"grad_norm": 0.24730735083234054,
"learning_rate": 2.6830965600704994e-06,
"loss": 0.8429,
"step": 672
},
{
"epoch": 2.070810082740042,
"grad_norm": 0.23416493138125025,
"learning_rate": 2.6672394434710703e-06,
"loss": 0.8611,
"step": 673
},
{
"epoch": 2.0738887819896092,
"grad_norm": 0.23108511125796244,
"learning_rate": 2.6514122612509653e-06,
"loss": 0.8708,
"step": 674
},
{
"epoch": 2.0769674812391763,
"grad_norm": 0.30933139206643,
"learning_rate": 2.6356152165072715e-06,
"loss": 0.8413,
"step": 675
},
{
"epoch": 2.0800461804887433,
"grad_norm": 0.23551112909526886,
"learning_rate": 2.619848511950348e-06,
"loss": 0.8475,
"step": 676
},
{
"epoch": 2.0831248797383104,
"grad_norm": 0.22763481246202175,
"learning_rate": 2.6041123499012244e-06,
"loss": 0.8995,
"step": 677
},
{
"epoch": 2.0862035789878774,
"grad_norm": 0.2218634208580765,
"learning_rate": 2.588406932289005e-06,
"loss": 0.8617,
"step": 678
},
{
"epoch": 2.0892822782374445,
"grad_norm": 0.23578935381626395,
"learning_rate": 2.5727324606482686e-06,
"loss": 0.8824,
"step": 679
},
{
"epoch": 2.0923609774870116,
"grad_norm": 0.251317071852673,
"learning_rate": 2.5570891361164975e-06,
"loss": 0.8636,
"step": 680
},
{
"epoch": 2.0954396767365786,
"grad_norm": 0.21351519156304594,
"learning_rate": 2.541477159431489e-06,
"loss": 0.8313,
"step": 681
},
{
"epoch": 2.0985183759861457,
"grad_norm": 0.22042567396928395,
"learning_rate": 2.525896730928773e-06,
"loss": 0.8731,
"step": 682
},
{
"epoch": 2.1015970752357127,
"grad_norm": 0.23702502385432966,
"learning_rate": 2.5103480505390543e-06,
"loss": 0.8304,
"step": 683
},
{
"epoch": 2.1046757744852798,
"grad_norm": 0.2153167638703399,
"learning_rate": 2.4948313177856402e-06,
"loss": 0.9014,
"step": 684
},
{
"epoch": 2.107754473734847,
"grad_norm": 0.22893541920732943,
"learning_rate": 2.4793467317818744e-06,
"loss": 0.8145,
"step": 685
},
{
"epoch": 2.110833172984414,
"grad_norm": 0.2237005823190317,
"learning_rate": 2.4638944912285963e-06,
"loss": 0.8412,
"step": 686
},
{
"epoch": 2.113911872233981,
"grad_norm": 0.23742295476228925,
"learning_rate": 2.4484747944115743e-06,
"loss": 0.8611,
"step": 687
},
{
"epoch": 2.1169905714835484,
"grad_norm": 0.28485027911857475,
"learning_rate": 2.4330878391989774e-06,
"loss": 0.8692,
"step": 688
},
{
"epoch": 2.1200692707331155,
"grad_norm": 0.23378185328116932,
"learning_rate": 2.417733823038828e-06,
"loss": 0.8266,
"step": 689
},
{
"epoch": 2.1231479699826825,
"grad_norm": 0.23451353036464137,
"learning_rate": 2.4024129429564626e-06,
"loss": 0.8466,
"step": 690
},
{
"epoch": 2.1262266692322496,
"grad_norm": 0.23387657795722544,
"learning_rate": 2.3871253955520163e-06,
"loss": 0.8127,
"step": 691
},
{
"epoch": 2.1293053684818166,
"grad_norm": 0.23248402401947377,
"learning_rate": 2.3718713769978925e-06,
"loss": 0.8213,
"step": 692
},
{
"epoch": 2.1323840677313837,
"grad_norm": 0.23706520225974131,
"learning_rate": 2.356651083036246e-06,
"loss": 0.843,
"step": 693
},
{
"epoch": 2.1354627669809507,
"grad_norm": 0.23710487467001728,
"learning_rate": 2.3414647089764696e-06,
"loss": 0.821,
"step": 694
},
{
"epoch": 2.1385414662305178,
"grad_norm": 0.24405866825393452,
"learning_rate": 2.3263124496926936e-06,
"loss": 0.8596,
"step": 695
},
{
"epoch": 2.141620165480085,
"grad_norm": 0.23314402094899678,
"learning_rate": 2.311194499621282e-06,
"loss": 0.835,
"step": 696
},
{
"epoch": 2.144698864729652,
"grad_norm": 0.2339010792369486,
"learning_rate": 2.296111052758333e-06,
"loss": 0.8629,
"step": 697
},
{
"epoch": 2.147777563979219,
"grad_norm": 0.243022565409757,
"learning_rate": 2.2810623026571988e-06,
"loss": 0.8222,
"step": 698
},
{
"epoch": 2.150856263228786,
"grad_norm": 0.2251158138526969,
"learning_rate": 2.266048442425989e-06,
"loss": 0.8291,
"step": 699
},
{
"epoch": 2.153934962478353,
"grad_norm": 0.22995408661018263,
"learning_rate": 2.251069664725114e-06,
"loss": 0.8051,
"step": 700
},
{
"epoch": 2.15701366172792,
"grad_norm": 0.23786905754445278,
"learning_rate": 2.2361261617647873e-06,
"loss": 0.8593,
"step": 701
},
{
"epoch": 2.160092360977487,
"grad_norm": 0.22238897378407732,
"learning_rate": 2.221218125302572e-06,
"loss": 0.8475,
"step": 702
},
{
"epoch": 2.163171060227054,
"grad_norm": 0.2409376152515762,
"learning_rate": 2.206345746640925e-06,
"loss": 0.8194,
"step": 703
},
{
"epoch": 2.1662497594766212,
"grad_norm": 0.2345698398897385,
"learning_rate": 2.191509216624732e-06,
"loss": 0.8496,
"step": 704
},
{
"epoch": 2.1693284587261883,
"grad_norm": 0.23918981745393544,
"learning_rate": 2.176708725638866e-06,
"loss": 0.8008,
"step": 705
},
{
"epoch": 2.1724071579757553,
"grad_norm": 0.24380759059520382,
"learning_rate": 2.161944463605736e-06,
"loss": 0.8211,
"step": 706
},
{
"epoch": 2.1754858572253224,
"grad_norm": 0.2582019109695414,
"learning_rate": 2.147216619982859e-06,
"loss": 0.8455,
"step": 707
},
{
"epoch": 2.1785645564748894,
"grad_norm": 0.23783881576481114,
"learning_rate": 2.1325253837604253e-06,
"loss": 0.8691,
"step": 708
},
{
"epoch": 2.1816432557244565,
"grad_norm": 0.24492401624008797,
"learning_rate": 2.117870943458868e-06,
"loss": 0.8196,
"step": 709
},
{
"epoch": 2.1847219549740236,
"grad_norm": 0.2219368360623884,
"learning_rate": 2.1032534871264554e-06,
"loss": 0.8258,
"step": 710
},
{
"epoch": 2.1878006542235906,
"grad_norm": 0.22974008998618684,
"learning_rate": 2.0886732023368637e-06,
"loss": 0.904,
"step": 711
},
{
"epoch": 2.1908793534731577,
"grad_norm": 0.23537831681729535,
"learning_rate": 2.0741302761867838e-06,
"loss": 0.8606,
"step": 712
},
{
"epoch": 2.1939580527227247,
"grad_norm": 0.2327223246764362,
"learning_rate": 2.059624895293514e-06,
"loss": 0.8252,
"step": 713
},
{
"epoch": 2.1970367519722918,
"grad_norm": 0.23577099287636177,
"learning_rate": 2.0451572457925587e-06,
"loss": 0.8507,
"step": 714
},
{
"epoch": 2.200115451221859,
"grad_norm": 0.23366851480303227,
"learning_rate": 2.0307275133352544e-06,
"loss": 0.8891,
"step": 715
},
{
"epoch": 2.203194150471426,
"grad_norm": 0.24821022271145754,
"learning_rate": 2.016335883086378e-06,
"loss": 0.8761,
"step": 716
},
{
"epoch": 2.206272849720993,
"grad_norm": 0.2347888592730076,
"learning_rate": 2.0019825397217673e-06,
"loss": 0.7948,
"step": 717
},
{
"epoch": 2.20935154897056,
"grad_norm": 0.22379601791809228,
"learning_rate": 1.9876676674259625e-06,
"loss": 0.8346,
"step": 718
},
{
"epoch": 2.212430248220127,
"grad_norm": 0.23069642639206184,
"learning_rate": 1.9733914498898338e-06,
"loss": 0.8806,
"step": 719
},
{
"epoch": 2.215508947469694,
"grad_norm": 0.23688821473020255,
"learning_rate": 1.9591540703082284e-06,
"loss": 0.8737,
"step": 720
},
{
"epoch": 2.218587646719261,
"grad_norm": 0.24068569952704222,
"learning_rate": 1.9449557113776157e-06,
"loss": 0.8206,
"step": 721
},
{
"epoch": 2.221666345968828,
"grad_norm": 0.23171967551106035,
"learning_rate": 1.930796555293744e-06,
"loss": 0.8579,
"step": 722
},
{
"epoch": 2.2247450452183952,
"grad_norm": 0.24029473406952884,
"learning_rate": 1.9166767837493137e-06,
"loss": 0.8528,
"step": 723
},
{
"epoch": 2.2278237444679623,
"grad_norm": 0.2438079273110106,
"learning_rate": 1.9025965779316236e-06,
"loss": 0.8492,
"step": 724
},
{
"epoch": 2.2309024437175293,
"grad_norm": 0.23601558456196375,
"learning_rate": 1.8885561185202682e-06,
"loss": 0.8595,
"step": 725
},
{
"epoch": 2.2339811429670964,
"grad_norm": 0.22767399139253267,
"learning_rate": 1.8745555856848013e-06,
"loss": 0.8776,
"step": 726
},
{
"epoch": 2.2370598422166634,
"grad_norm": 0.23348060469006376,
"learning_rate": 1.8605951590824385e-06,
"loss": 0.8647,
"step": 727
},
{
"epoch": 2.2401385414662305,
"grad_norm": 0.22700844372163298,
"learning_rate": 1.8466750178557442e-06,
"loss": 0.8366,
"step": 728
},
{
"epoch": 2.2432172407157975,
"grad_norm": 0.2174522829564044,
"learning_rate": 1.832795340630329e-06,
"loss": 0.8009,
"step": 729
},
{
"epoch": 2.2462959399653646,
"grad_norm": 0.21346427567062592,
"learning_rate": 1.8189563055125691e-06,
"loss": 0.8423,
"step": 730
},
{
"epoch": 2.2493746392149316,
"grad_norm": 0.2374887094550808,
"learning_rate": 1.8051580900873112e-06,
"loss": 0.7801,
"step": 731
},
{
"epoch": 2.2524533384644987,
"grad_norm": 0.2250593612933212,
"learning_rate": 1.7914008714155984e-06,
"loss": 0.8034,
"step": 732
},
{
"epoch": 2.2555320377140657,
"grad_norm": 0.23293428960330465,
"learning_rate": 1.7776848260323937e-06,
"loss": 0.8983,
"step": 733
},
{
"epoch": 2.258610736963633,
"grad_norm": 0.24046965281537155,
"learning_rate": 1.7640101299443163e-06,
"loss": 0.8645,
"step": 734
},
{
"epoch": 2.2616894362132,
"grad_norm": 0.24414380255341103,
"learning_rate": 1.7503769586273933e-06,
"loss": 0.88,
"step": 735
},
{
"epoch": 2.264768135462767,
"grad_norm": 0.23249280363014194,
"learning_rate": 1.7367854870247874e-06,
"loss": 0.8454,
"step": 736
},
{
"epoch": 2.267846834712334,
"grad_norm": 0.22698421515363507,
"learning_rate": 1.7232358895445722e-06,
"loss": 0.8194,
"step": 737
},
{
"epoch": 2.270925533961901,
"grad_norm": 0.2424961229284742,
"learning_rate": 1.7097283400574782e-06,
"loss": 0.8784,
"step": 738
},
{
"epoch": 2.274004233211468,
"grad_norm": 0.29034676831693745,
"learning_rate": 1.6962630118946742e-06,
"loss": 0.8418,
"step": 739
},
{
"epoch": 2.277082932461035,
"grad_norm": 0.24062810782676536,
"learning_rate": 1.682840077845539e-06,
"loss": 0.8483,
"step": 740
},
{
"epoch": 2.280161631710602,
"grad_norm": 0.22180691188024756,
"learning_rate": 1.6694597101554354e-06,
"loss": 0.8572,
"step": 741
},
{
"epoch": 2.283240330960169,
"grad_norm": 0.24825655929259619,
"learning_rate": 1.6561220805235146e-06,
"loss": 0.8573,
"step": 742
},
{
"epoch": 2.2863190302097363,
"grad_norm": 0.22740220138114653,
"learning_rate": 1.6428273601005052e-06,
"loss": 0.8632,
"step": 743
},
{
"epoch": 2.2893977294593033,
"grad_norm": 0.24149014361913568,
"learning_rate": 1.629575719486512e-06,
"loss": 0.8572,
"step": 744
},
{
"epoch": 2.2924764287088704,
"grad_norm": 0.22379601628395732,
"learning_rate": 1.616367328728838e-06,
"loss": 0.8594,
"step": 745
},
{
"epoch": 2.2955551279584374,
"grad_norm": 0.23583028558555058,
"learning_rate": 1.6032023573197947e-06,
"loss": 0.8508,
"step": 746
},
{
"epoch": 2.2986338272080045,
"grad_norm": 0.22042850102418635,
"learning_rate": 1.5900809741945311e-06,
"loss": 0.8105,
"step": 747
},
{
"epoch": 2.3017125264575715,
"grad_norm": 0.23334172430216374,
"learning_rate": 1.5770033477288599e-06,
"loss": 0.8436,
"step": 748
},
{
"epoch": 2.3047912257071386,
"grad_norm": 0.2322225057102178,
"learning_rate": 1.5639696457371023e-06,
"loss": 0.8385,
"step": 749
},
{
"epoch": 2.3078699249567056,
"grad_norm": 0.22194847905746878,
"learning_rate": 1.5509800354699355e-06,
"loss": 0.8388,
"step": 750
},
{
"epoch": 2.3109486242062727,
"grad_norm": 0.23209512373356478,
"learning_rate": 1.5380346836122429e-06,
"loss": 0.8474,
"step": 751
},
{
"epoch": 2.3140273234558397,
"grad_norm": 0.2223477673870283,
"learning_rate": 1.525133756280977e-06,
"loss": 0.8749,
"step": 752
},
{
"epoch": 2.3171060227054068,
"grad_norm": 0.23141352847934166,
"learning_rate": 1.5122774190230243e-06,
"loss": 0.8853,
"step": 753
},
{
"epoch": 2.3201847219549743,
"grad_norm": 0.23062664210205389,
"learning_rate": 1.4994658368130876e-06,
"loss": 0.8346,
"step": 754
},
{
"epoch": 2.3232634212045413,
"grad_norm": 0.2215532209006794,
"learning_rate": 1.486699174051565e-06,
"loss": 0.8145,
"step": 755
},
{
"epoch": 2.3263421204541084,
"grad_norm": 0.22600723734150047,
"learning_rate": 1.4739775945624373e-06,
"loss": 0.8263,
"step": 756
},
{
"epoch": 2.3294208197036754,
"grad_norm": 0.22143336713406017,
"learning_rate": 1.461301261591172e-06,
"loss": 0.8114,
"step": 757
},
{
"epoch": 2.3324995189532425,
"grad_norm": 0.22271954812926723,
"learning_rate": 1.4486703378026256e-06,
"loss": 0.8369,
"step": 758
},
{
"epoch": 2.3355782182028095,
"grad_norm": 0.22658746997200865,
"learning_rate": 1.4360849852789565e-06,
"loss": 0.8171,
"step": 759
},
{
"epoch": 2.3386569174523766,
"grad_norm": 0.2162121277661999,
"learning_rate": 1.423545365517543e-06,
"loss": 0.8027,
"step": 760
},
{
"epoch": 2.3417356167019436,
"grad_norm": 0.2304719291746044,
"learning_rate": 1.4110516394289103e-06,
"loss": 0.8736,
"step": 761
},
{
"epoch": 2.3448143159515107,
"grad_norm": 0.23214626525469848,
"learning_rate": 1.3986039673346779e-06,
"loss": 0.8265,
"step": 762
},
{
"epoch": 2.3478930152010777,
"grad_norm": 0.2268381102532204,
"learning_rate": 1.3862025089654824e-06,
"loss": 0.8947,
"step": 763
},
{
"epoch": 2.350971714450645,
"grad_norm": 0.2385163483420689,
"learning_rate": 1.3738474234589443e-06,
"loss": 0.8331,
"step": 764
},
{
"epoch": 2.354050413700212,
"grad_norm": 0.2390331445172912,
"learning_rate": 1.3615388693576148e-06,
"loss": 0.8732,
"step": 765
},
{
"epoch": 2.357129112949779,
"grad_norm": 0.22955588916496036,
"learning_rate": 1.349277004606951e-06,
"loss": 0.8594,
"step": 766
},
{
"epoch": 2.360207812199346,
"grad_norm": 0.21970006090004243,
"learning_rate": 1.3370619865532824e-06,
"loss": 0.7861,
"step": 767
},
{
"epoch": 2.363286511448913,
"grad_norm": 0.230418857783953,
"learning_rate": 1.3248939719417914e-06,
"loss": 0.8268,
"step": 768
},
{
"epoch": 2.36636521069848,
"grad_norm": 0.21222149462576967,
"learning_rate": 1.312773116914508e-06,
"loss": 0.798,
"step": 769
},
{
"epoch": 2.369443909948047,
"grad_norm": 0.22262097093208535,
"learning_rate": 1.3006995770083015e-06,
"loss": 0.8427,
"step": 770
},
{
"epoch": 2.372522609197614,
"grad_norm": 0.22750342291942272,
"learning_rate": 1.2886735071528829e-06,
"loss": 0.8274,
"step": 771
},
{
"epoch": 2.375601308447181,
"grad_norm": 0.23680438793884293,
"learning_rate": 1.2766950616688233e-06,
"loss": 0.8417,
"step": 772
},
{
"epoch": 2.3786800076967483,
"grad_norm": 0.22306977725172839,
"learning_rate": 1.264764394265564e-06,
"loss": 0.8091,
"step": 773
},
{
"epoch": 2.3817587069463153,
"grad_norm": 0.22041606530713104,
"learning_rate": 1.2528816580394593e-06,
"loss": 0.8155,
"step": 774
},
{
"epoch": 2.3848374061958824,
"grad_norm": 0.2319982882257984,
"learning_rate": 1.2410470054717944e-06,
"loss": 0.8195,
"step": 775
},
{
"epoch": 2.3879161054454494,
"grad_norm": 0.23206812452151346,
"learning_rate": 1.2292605884268377e-06,
"loss": 0.846,
"step": 776
},
{
"epoch": 2.3909948046950165,
"grad_norm": 0.2253987855080877,
"learning_rate": 1.217522558149894e-06,
"loss": 0.8776,
"step": 777
},
{
"epoch": 2.3940735039445835,
"grad_norm": 0.23978736453951358,
"learning_rate": 1.2058330652653605e-06,
"loss": 0.8307,
"step": 778
},
{
"epoch": 2.3971522031941506,
"grad_norm": 0.22047620559443692,
"learning_rate": 1.194192259774794e-06,
"loss": 0.8522,
"step": 779
},
{
"epoch": 2.4002309024437176,
"grad_norm": 0.2255029516924654,
"learning_rate": 1.1826002910549839e-06,
"loss": 0.8092,
"step": 780
},
{
"epoch": 2.4033096016932847,
"grad_norm": 0.22748667467933123,
"learning_rate": 1.1710573078560406e-06,
"loss": 0.8773,
"step": 781
},
{
"epoch": 2.4063883009428517,
"grad_norm": 0.21978466898899587,
"learning_rate": 1.1595634582994857e-06,
"loss": 0.7762,
"step": 782
},
{
"epoch": 2.4094670001924188,
"grad_norm": 0.2123601020631904,
"learning_rate": 1.1481188898763435e-06,
"loss": 0.8186,
"step": 783
},
{
"epoch": 2.412545699441986,
"grad_norm": 0.2200140512131943,
"learning_rate": 1.1367237494452604e-06,
"loss": 0.8557,
"step": 784
},
{
"epoch": 2.415624398691553,
"grad_norm": 0.22718149738177595,
"learning_rate": 1.1253781832306126e-06,
"loss": 0.8212,
"step": 785
},
{
"epoch": 2.41870309794112,
"grad_norm": 0.2626494778904241,
"learning_rate": 1.1140823368206287e-06,
"loss": 0.8031,
"step": 786
},
{
"epoch": 2.421781797190687,
"grad_norm": 0.2304850629397168,
"learning_rate": 1.1028363551655296e-06,
"loss": 0.8436,
"step": 787
},
{
"epoch": 2.424860496440254,
"grad_norm": 0.23779144390206716,
"learning_rate": 1.0916403825756578e-06,
"loss": 0.8436,
"step": 788
},
{
"epoch": 2.427939195689821,
"grad_norm": 0.2253718147127126,
"learning_rate": 1.080494562719634e-06,
"loss": 0.8227,
"step": 789
},
{
"epoch": 2.431017894939388,
"grad_norm": 0.23899024287866094,
"learning_rate": 1.0693990386225118e-06,
"loss": 0.8424,
"step": 790
},
{
"epoch": 2.434096594188955,
"grad_norm": 0.24397315362125152,
"learning_rate": 1.058353952663937e-06,
"loss": 0.8648,
"step": 791
},
{
"epoch": 2.4371752934385222,
"grad_norm": 0.24301907266829076,
"learning_rate": 1.0473594465763282e-06,
"loss": 0.8245,
"step": 792
},
{
"epoch": 2.4402539926880893,
"grad_norm": 0.22689000896943365,
"learning_rate": 1.0364156614430525e-06,
"loss": 0.8565,
"step": 793
},
{
"epoch": 2.4433326919376563,
"grad_norm": 0.2309233420575025,
"learning_rate": 1.0255227376966198e-06,
"loss": 0.8081,
"step": 794
},
{
"epoch": 2.4464113911872234,
"grad_norm": 0.236711561978388,
"learning_rate": 1.0146808151168736e-06,
"loss": 0.816,
"step": 795
},
{
"epoch": 2.4494900904367904,
"grad_norm": 0.21931754203757395,
"learning_rate": 1.0038900328292051e-06,
"loss": 0.8272,
"step": 796
},
{
"epoch": 2.4525687896863575,
"grad_norm": 0.22527350770670657,
"learning_rate": 9.931505293027655e-07,
"loss": 0.8226,
"step": 797
},
{
"epoch": 2.4556474889359245,
"grad_norm": 0.21984307356386273,
"learning_rate": 9.824624423486833e-07,
"loss": 0.8281,
"step": 798
},
{
"epoch": 2.4587261881854916,
"grad_norm": 0.22204340349648355,
"learning_rate": 9.71825909118308e-07,
"loss": 0.9001,
"step": 799
},
{
"epoch": 2.4618048874350587,
"grad_norm": 0.23089658728762086,
"learning_rate": 9.612410661014338e-07,
"loss": 0.8391,
"step": 800
},
{
"epoch": 2.4648835866846257,
"grad_norm": 0.2278803378229771,
"learning_rate": 9.50708049124568e-07,
"loss": 0.8379,
"step": 801
},
{
"epoch": 2.4679622859341928,
"grad_norm": 0.2370178692049797,
"learning_rate": 9.402269933491692e-07,
"loss": 0.8439,
"step": 802
},
{
"epoch": 2.47104098518376,
"grad_norm": 0.2229543530878649,
"learning_rate": 9.29798033269923e-07,
"loss": 0.7776,
"step": 803
},
{
"epoch": 2.474119684433327,
"grad_norm": 0.24402092811363743,
"learning_rate": 9.194213027130155e-07,
"loss": 0.8279,
"step": 804
},
{
"epoch": 2.477198383682894,
"grad_norm": 0.23558787532961123,
"learning_rate": 9.090969348344147e-07,
"loss": 0.862,
"step": 805
},
{
"epoch": 2.480277082932461,
"grad_norm": 0.2799555949111056,
"learning_rate": 8.988250621181621e-07,
"loss": 0.8163,
"step": 806
},
{
"epoch": 2.483355782182028,
"grad_norm": 0.22449886610239306,
"learning_rate": 8.886058163746686e-07,
"loss": 0.8213,
"step": 807
},
{
"epoch": 2.486434481431595,
"grad_norm": 0.24364392856915754,
"learning_rate": 8.78439328739032e-07,
"loss": 0.821,
"step": 808
},
{
"epoch": 2.489513180681162,
"grad_norm": 0.2630991901598819,
"learning_rate": 8.683257296693476e-07,
"loss": 0.8751,
"step": 809
},
{
"epoch": 2.492591879930729,
"grad_norm": 0.23200611225140585,
"learning_rate": 8.582651489450328e-07,
"loss": 0.8125,
"step": 810
},
{
"epoch": 2.4956705791802962,
"grad_norm": 0.22437483085314536,
"learning_rate": 8.482577156651683e-07,
"loss": 0.8485,
"step": 811
},
{
"epoch": 2.4987492784298633,
"grad_norm": 0.2388118695272368,
"learning_rate": 8.383035582468341e-07,
"loss": 0.8811,
"step": 812
},
{
"epoch": 2.5018279776794303,
"grad_norm": 0.2436813114038449,
"learning_rate": 8.284028044234693e-07,
"loss": 0.8601,
"step": 813
},
{
"epoch": 2.5049066769289974,
"grad_norm": 0.21301942107742983,
"learning_rate": 8.185555812432272e-07,
"loss": 0.855,
"step": 814
},
{
"epoch": 2.5079853761785644,
"grad_norm": 0.21605275572181498,
"learning_rate": 8.087620150673453e-07,
"loss": 0.796,
"step": 815
},
{
"epoch": 2.5110640754281315,
"grad_norm": 0.22136048072982634,
"learning_rate": 7.99022231568527e-07,
"loss": 0.8318,
"step": 816
},
{
"epoch": 2.5141427746776985,
"grad_norm": 0.2278401120367128,
"learning_rate": 7.893363557293293e-07,
"loss": 0.8808,
"step": 817
},
{
"epoch": 2.5172214739272656,
"grad_norm": 0.23853145651181795,
"learning_rate": 7.797045118405511e-07,
"loss": 0.8114,
"step": 818
},
{
"epoch": 2.5203001731768326,
"grad_norm": 0.24139204959277535,
"learning_rate": 7.701268234996495e-07,
"loss": 0.8268,
"step": 819
},
{
"epoch": 2.5233788724263997,
"grad_norm": 0.21906838187313607,
"learning_rate": 7.606034136091461e-07,
"loss": 0.8515,
"step": 820
},
{
"epoch": 2.5264575716759667,
"grad_norm": 0.22353012243741543,
"learning_rate": 7.511344043750546e-07,
"loss": 0.8438,
"step": 821
},
{
"epoch": 2.529536270925534,
"grad_norm": 0.24113336047569772,
"learning_rate": 7.417199173053063e-07,
"loss": 0.8843,
"step": 822
},
{
"epoch": 2.532614970175101,
"grad_norm": 0.23028580739427149,
"learning_rate": 7.323600732081954e-07,
"loss": 0.8222,
"step": 823
},
{
"epoch": 2.535693669424668,
"grad_norm": 0.23065486789263445,
"learning_rate": 7.230549921908342e-07,
"loss": 0.8206,
"step": 824
},
{
"epoch": 2.538772368674235,
"grad_norm": 0.22279493349776192,
"learning_rate": 7.138047936575976e-07,
"loss": 0.8343,
"step": 825
},
{
"epoch": 2.541851067923802,
"grad_norm": 0.21826094656766837,
"learning_rate": 7.046095963086053e-07,
"loss": 0.8244,
"step": 826
},
{
"epoch": 2.544929767173369,
"grad_norm": 0.22164075701764316,
"learning_rate": 6.954695181381876e-07,
"loss": 0.8526,
"step": 827
},
{
"epoch": 2.548008466422936,
"grad_norm": 0.2240757412936755,
"learning_rate": 6.863846764333798e-07,
"loss": 0.8295,
"step": 828
},
{
"epoch": 2.551087165672503,
"grad_norm": 0.22239008279252742,
"learning_rate": 6.773551877724144e-07,
"loss": 0.8564,
"step": 829
},
{
"epoch": 2.55416586492207,
"grad_norm": 0.2278231158301186,
"learning_rate": 6.683811680232182e-07,
"loss": 0.7971,
"step": 830
},
{
"epoch": 2.5572445641716373,
"grad_norm": 0.22159668670917831,
"learning_rate": 6.594627323419383e-07,
"loss": 0.8752,
"step": 831
},
{
"epoch": 2.5603232634212043,
"grad_norm": 0.22750909281464063,
"learning_rate": 6.505999951714537e-07,
"loss": 0.8861,
"step": 832
},
{
"epoch": 2.5634019626707714,
"grad_norm": 0.22109666959494906,
"learning_rate": 6.417930702399133e-07,
"loss": 0.8468,
"step": 833
},
{
"epoch": 2.5664806619203384,
"grad_norm": 0.23112153878468647,
"learning_rate": 6.33042070559271e-07,
"loss": 0.8529,
"step": 834
},
{
"epoch": 2.5695593611699055,
"grad_norm": 0.2297458419700991,
"learning_rate": 6.243471084238378e-07,
"loss": 0.8689,
"step": 835
},
{
"epoch": 2.5726380604194725,
"grad_norm": 0.21578699434521503,
"learning_rate": 6.157082954088478e-07,
"loss": 0.7902,
"step": 836
},
{
"epoch": 2.5757167596690396,
"grad_norm": 0.21747841612156554,
"learning_rate": 6.071257423690119e-07,
"loss": 0.8691,
"step": 837
},
{
"epoch": 2.5787954589186066,
"grad_norm": 0.22455305402136413,
"learning_rate": 5.985995594371091e-07,
"loss": 0.821,
"step": 838
},
{
"epoch": 2.5818741581681737,
"grad_norm": 0.22253531062382764,
"learning_rate": 5.901298560225644e-07,
"loss": 0.8604,
"step": 839
},
{
"epoch": 2.584952857417741,
"grad_norm": 0.2318489532483918,
"learning_rate": 5.817167408100499e-07,
"loss": 0.8267,
"step": 840
},
{
"epoch": 2.588031556667308,
"grad_norm": 0.2317479826256508,
"learning_rate": 5.733603217580885e-07,
"loss": 0.8564,
"step": 841
},
{
"epoch": 2.5911102559168753,
"grad_norm": 0.22222098087172276,
"learning_rate": 5.650607060976659e-07,
"loss": 0.8366,
"step": 842
},
{
"epoch": 2.5941889551664423,
"grad_norm": 0.22558823326040037,
"learning_rate": 5.568180003308593e-07,
"loss": 0.8455,
"step": 843
},
{
"epoch": 2.5972676544160094,
"grad_norm": 0.22517318330162056,
"learning_rate": 5.486323102294689e-07,
"loss": 0.8864,
"step": 844
},
{
"epoch": 2.6003463536655764,
"grad_norm": 0.22018283187036372,
"learning_rate": 5.405037408336577e-07,
"loss": 0.8721,
"step": 845
},
{
"epoch": 2.6034250529151435,
"grad_norm": 0.22293725459227548,
"learning_rate": 5.324323964506078e-07,
"loss": 0.8652,
"step": 846
},
{
"epoch": 2.6065037521647105,
"grad_norm": 0.22327030606702958,
"learning_rate": 5.244183806531805e-07,
"loss": 0.8083,
"step": 847
},
{
"epoch": 2.6095824514142776,
"grad_norm": 0.22767394590167395,
"learning_rate": 5.164617962785878e-07,
"loss": 0.8109,
"step": 848
},
{
"epoch": 2.6126611506638446,
"grad_norm": 0.21169668836208091,
"learning_rate": 5.085627454270687e-07,
"loss": 0.8385,
"step": 849
},
{
"epoch": 2.6157398499134117,
"grad_norm": 0.22413803236048283,
"learning_rate": 5.007213294605839e-07,
"loss": 0.8352,
"step": 850
},
{
"epoch": 2.6188185491629787,
"grad_norm": 0.2214793922015388,
"learning_rate": 4.929376490015148e-07,
"loss": 0.8658,
"step": 851
},
{
"epoch": 2.621897248412546,
"grad_norm": 0.2216666271986793,
"learning_rate": 4.852118039313691e-07,
"loss": 0.796,
"step": 852
},
{
"epoch": 2.624975947662113,
"grad_norm": 0.22844670773372602,
"learning_rate": 4.775438933895032e-07,
"loss": 0.8491,
"step": 853
},
{
"epoch": 2.62805464691168,
"grad_norm": 0.22818017196392082,
"learning_rate": 4.6993401577184515e-07,
"loss": 0.8546,
"step": 854
},
{
"epoch": 2.631133346161247,
"grad_norm": 0.22838330421080386,
"learning_rate": 4.623822687296359e-07,
"loss": 0.846,
"step": 855
},
{
"epoch": 2.634212045410814,
"grad_norm": 0.2256160359726944,
"learning_rate": 4.5488874916817636e-07,
"loss": 0.8807,
"step": 856
},
{
"epoch": 2.637290744660381,
"grad_norm": 0.22297103683292707,
"learning_rate": 4.474535532455793e-07,
"loss": 0.8358,
"step": 857
},
{
"epoch": 2.640369443909948,
"grad_norm": 0.2332114290381529,
"learning_rate": 4.400767763715413e-07,
"loss": 0.8234,
"step": 858
},
{
"epoch": 2.643448143159515,
"grad_norm": 0.2331719300147976,
"learning_rate": 4.3275851320611517e-07,
"loss": 0.8141,
"step": 859
},
{
"epoch": 2.646526842409082,
"grad_norm": 0.22645691635776777,
"learning_rate": 4.2549885765849673e-07,
"loss": 0.8694,
"step": 860
},
{
"epoch": 2.6496055416586493,
"grad_norm": 0.21716184143730555,
"learning_rate": 4.1829790288581694e-07,
"loss": 0.8661,
"step": 861
},
{
"epoch": 2.6526842409082163,
"grad_norm": 0.23013336076460622,
"learning_rate": 4.111557412919476e-07,
"loss": 0.829,
"step": 862
},
{
"epoch": 2.6557629401577834,
"grad_norm": 0.22601089934921306,
"learning_rate": 4.0407246452632164e-07,
"loss": 0.8332,
"step": 863
},
{
"epoch": 2.6588416394073504,
"grad_norm": 0.22227844789679432,
"learning_rate": 3.9704816348274613e-07,
"loss": 0.8149,
"step": 864
},
{
"epoch": 2.6619203386569175,
"grad_norm": 0.2177118454088843,
"learning_rate": 3.90082928298246e-07,
"loss": 0.8018,
"step": 865
},
{
"epoch": 2.6649990379064845,
"grad_norm": 0.22056846315659762,
"learning_rate": 3.8317684835189885e-07,
"loss": 0.8415,
"step": 866
},
{
"epoch": 2.6680777371560516,
"grad_norm": 0.21614609644555607,
"learning_rate": 3.7633001226369594e-07,
"loss": 0.8174,
"step": 867
},
{
"epoch": 2.6711564364056186,
"grad_norm": 0.21551598064340127,
"learning_rate": 3.6954250789340006e-07,
"loss": 0.7983,
"step": 868
},
{
"epoch": 2.6742351356551857,
"grad_norm": 0.21140537734349923,
"learning_rate": 3.628144223394181e-07,
"loss": 0.8322,
"step": 869
},
{
"epoch": 2.6773138349047527,
"grad_norm": 0.20967093594161001,
"learning_rate": 3.561458419376862e-07,
"loss": 0.8532,
"step": 870
},
{
"epoch": 2.6803925341543198,
"grad_norm": 0.22990587789252084,
"learning_rate": 3.495368522605602e-07,
"loss": 0.8397,
"step": 871
},
{
"epoch": 2.683471233403887,
"grad_norm": 0.22231549579524137,
"learning_rate": 3.429875381157161e-07,
"loss": 0.8262,
"step": 872
},
{
"epoch": 2.686549932653454,
"grad_norm": 0.22886985448919456,
"learning_rate": 3.3649798354506545e-07,
"loss": 0.8871,
"step": 873
},
{
"epoch": 2.689628631903021,
"grad_norm": 0.22003259909250156,
"learning_rate": 3.3006827182367196e-07,
"loss": 0.8073,
"step": 874
},
{
"epoch": 2.692707331152588,
"grad_norm": 0.22742597600380163,
"learning_rate": 3.2369848545868997e-07,
"loss": 0.8661,
"step": 875
},
{
"epoch": 2.695786030402155,
"grad_norm": 0.22868877060486337,
"learning_rate": 3.17388706188298e-07,
"loss": 0.834,
"step": 876
},
{
"epoch": 2.698864729651722,
"grad_norm": 0.21787153465218684,
"learning_rate": 3.111390149806537e-07,
"loss": 0.7858,
"step": 877
},
{
"epoch": 2.701943428901289,
"grad_norm": 0.22570078190133194,
"learning_rate": 3.049494920328555e-07,
"loss": 0.826,
"step": 878
},
{
"epoch": 2.705022128150856,
"grad_norm": 0.2243914756159185,
"learning_rate": 2.988202167699128e-07,
"loss": 0.839,
"step": 879
},
{
"epoch": 2.7081008274004232,
"grad_norm": 0.22519826895866205,
"learning_rate": 2.927512678437267e-07,
"loss": 0.8237,
"step": 880
},
{
"epoch": 2.7111795266499903,
"grad_norm": 0.238902485686314,
"learning_rate": 2.867427231320774e-07,
"loss": 0.8559,
"step": 881
},
{
"epoch": 2.7142582258995573,
"grad_norm": 0.23042640260287536,
"learning_rate": 2.807946597376315e-07,
"loss": 0.882,
"step": 882
},
{
"epoch": 2.7173369251491244,
"grad_norm": 0.23089625608312397,
"learning_rate": 2.7490715398694856e-07,
"loss": 0.8594,
"step": 883
},
{
"epoch": 2.7204156243986914,
"grad_norm": 0.2246650825844926,
"learning_rate": 2.6908028142950016e-07,
"loss": 0.8657,
"step": 884
},
{
"epoch": 2.7234943236482585,
"grad_norm": 0.2258566049620716,
"learning_rate": 2.633141168367043e-07,
"loss": 0.8401,
"step": 885
},
{
"epoch": 2.7265730228978255,
"grad_norm": 0.21746790121867032,
"learning_rate": 2.5760873420096423e-07,
"loss": 0.8098,
"step": 886
},
{
"epoch": 2.7296517221473926,
"grad_norm": 0.22700354804834932,
"learning_rate": 2.5196420673471664e-07,
"loss": 0.843,
"step": 887
},
{
"epoch": 2.7327304213969597,
"grad_norm": 0.2266326370232206,
"learning_rate": 2.4638060686949736e-07,
"loss": 0.8824,
"step": 888
},
{
"epoch": 2.7358091206465267,
"grad_norm": 0.26185349410947084,
"learning_rate": 2.408580062550053e-07,
"loss": 0.8495,
"step": 889
},
{
"epoch": 2.7388878198960938,
"grad_norm": 0.22741750862384238,
"learning_rate": 2.3539647575819169e-07,
"loss": 0.8395,
"step": 890
},
{
"epoch": 2.741966519145661,
"grad_norm": 0.22615487096214848,
"learning_rate": 2.2999608546234063e-07,
"loss": 0.8012,
"step": 891
},
{
"epoch": 2.745045218395228,
"grad_norm": 0.21848490351475774,
"learning_rate": 2.246569046661773e-07,
"loss": 0.8692,
"step": 892
},
{
"epoch": 2.748123917644795,
"grad_norm": 0.21844617170173872,
"learning_rate": 2.1937900188297502e-07,
"loss": 0.8347,
"step": 893
},
{
"epoch": 2.7512026168943624,
"grad_norm": 0.2262540926446087,
"learning_rate": 2.1416244483967896e-07,
"loss": 0.8611,
"step": 894
},
{
"epoch": 2.7542813161439295,
"grad_norm": 0.2154952474239762,
"learning_rate": 2.0900730047603513e-07,
"loss": 0.7942,
"step": 895
},
{
"epoch": 2.7573600153934965,
"grad_norm": 0.21589356108725566,
"learning_rate": 2.0391363494372817e-07,
"loss": 0.8501,
"step": 896
},
{
"epoch": 2.7604387146430636,
"grad_norm": 0.2320250906884757,
"learning_rate": 1.9888151360554052e-07,
"loss": 0.9031,
"step": 897
},
{
"epoch": 2.7635174138926306,
"grad_norm": 0.23475684232788444,
"learning_rate": 1.9391100103450743e-07,
"loss": 0.8525,
"step": 898
},
{
"epoch": 2.7665961131421977,
"grad_norm": 0.21866007981336483,
"learning_rate": 1.890021610130882e-07,
"loss": 0.8189,
"step": 899
},
{
"epoch": 2.7696748123917647,
"grad_norm": 0.2174948225638527,
"learning_rate": 1.8415505653235244e-07,
"loss": 0.8376,
"step": 900
},
{
"epoch": 2.7727535116413318,
"grad_norm": 0.22006191115906404,
"learning_rate": 1.7936974979116505e-07,
"loss": 0.8366,
"step": 901
},
{
"epoch": 2.775832210890899,
"grad_norm": 0.2303616638680649,
"learning_rate": 1.7464630219539758e-07,
"loss": 0.8466,
"step": 902
},
{
"epoch": 2.778910910140466,
"grad_norm": 0.21719288352491736,
"learning_rate": 1.6998477435712923e-07,
"loss": 0.8265,
"step": 903
},
{
"epoch": 2.781989609390033,
"grad_norm": 0.22481604624243925,
"learning_rate": 1.6538522609387597e-07,
"loss": 0.8319,
"step": 904
},
{
"epoch": 2.7850683086396,
"grad_norm": 0.21429225139171987,
"learning_rate": 1.6084771642782214e-07,
"loss": 0.838,
"step": 905
},
{
"epoch": 2.788147007889167,
"grad_norm": 0.22194210468751915,
"learning_rate": 1.5637230358506173e-07,
"loss": 0.8247,
"step": 906
},
{
"epoch": 2.791225707138734,
"grad_norm": 0.23699639188325108,
"learning_rate": 1.5195904499485215e-07,
"loss": 0.8427,
"step": 907
},
{
"epoch": 2.794304406388301,
"grad_norm": 0.21126771897929203,
"learning_rate": 1.4760799728887553e-07,
"loss": 0.8107,
"step": 908
},
{
"epoch": 2.797383105637868,
"grad_norm": 0.2185312236043927,
"learning_rate": 1.4331921630051538e-07,
"loss": 0.873,
"step": 909
},
{
"epoch": 2.8004618048874352,
"grad_norm": 0.22162193710004144,
"learning_rate": 1.390927570641376e-07,
"loss": 0.8059,
"step": 910
},
{
"epoch": 2.8035405041370023,
"grad_norm": 0.22441107598451382,
"learning_rate": 1.349286738143829e-07,
"loss": 0.8195,
"step": 911
},
{
"epoch": 2.8066192033865693,
"grad_norm": 0.21557627624281156,
"learning_rate": 1.3082701998547498e-07,
"loss": 0.852,
"step": 912
},
{
"epoch": 2.8096979026361364,
"grad_norm": 0.21945269925130453,
"learning_rate": 1.267878482105317e-07,
"loss": 0.8746,
"step": 913
},
{
"epoch": 2.8127766018857034,
"grad_norm": 0.22357578889352,
"learning_rate": 1.2281121032089062e-07,
"loss": 0.8202,
"step": 914
},
{
"epoch": 2.8158553011352705,
"grad_norm": 0.22752559478727943,
"learning_rate": 1.1889715734544449e-07,
"loss": 0.8325,
"step": 915
},
{
"epoch": 2.8189340003848375,
"grad_norm": 0.2124940858260051,
"learning_rate": 1.150457395099841e-07,
"loss": 0.8434,
"step": 916
},
{
"epoch": 2.8220126996344046,
"grad_norm": 0.2157794578759078,
"learning_rate": 1.1125700623655756e-07,
"loss": 0.8062,
"step": 917
},
{
"epoch": 2.8250913988839716,
"grad_norm": 0.2181787040339238,
"learning_rate": 1.0753100614283362e-07,
"loss": 0.8343,
"step": 918
},
{
"epoch": 2.8281700981335387,
"grad_norm": 0.2074610780009133,
"learning_rate": 1.0386778704147671e-07,
"loss": 0.8169,
"step": 919
},
{
"epoch": 2.8312487973831058,
"grad_norm": 0.2204593136350251,
"learning_rate": 1.0026739593953616e-07,
"loss": 0.865,
"step": 920
},
{
"epoch": 2.834327496632673,
"grad_norm": 0.21907980416329806,
"learning_rate": 9.672987903784237e-08,
"loss": 0.845,
"step": 921
},
{
"epoch": 2.83740619588224,
"grad_norm": 0.22860904197624857,
"learning_rate": 9.32552817304122e-08,
"loss": 0.8636,
"step": 922
},
{
"epoch": 2.840484895131807,
"grad_norm": 0.21755262234247255,
"learning_rate": 8.98436486038673e-08,
"loss": 0.8195,
"step": 923
},
{
"epoch": 2.843563594381374,
"grad_norm": 0.22723056285347357,
"learning_rate": 8.649502343686333e-08,
"loss": 0.8193,
"step": 924
},
{
"epoch": 2.846642293630941,
"grad_norm": 0.2132875219805442,
"learning_rate": 8.320944919952723e-08,
"loss": 0.8316,
"step": 925
},
{
"epoch": 2.849720992880508,
"grad_norm": 0.22784897097532497,
"learning_rate": 7.998696805290419e-08,
"loss": 0.8299,
"step": 926
},
{
"epoch": 2.852799692130075,
"grad_norm": 0.22444636396859846,
"learning_rate": 7.682762134841981e-08,
"loss": 0.8844,
"step": 927
},
{
"epoch": 2.855878391379642,
"grad_norm": 0.22092497586912369,
"learning_rate": 7.373144962734613e-08,
"loss": 0.8129,
"step": 928
},
{
"epoch": 2.858957090629209,
"grad_norm": 0.2163075426109022,
"learning_rate": 7.069849262028583e-08,
"loss": 0.8215,
"step": 929
},
{
"epoch": 2.8620357898787763,
"grad_norm": 0.21570527806169992,
"learning_rate": 6.772878924665716e-08,
"loss": 0.8159,
"step": 930
},
{
"epoch": 2.8651144891283433,
"grad_norm": 0.22275170357879553,
"learning_rate": 6.482237761419652e-08,
"loss": 0.8371,
"step": 931
},
{
"epoch": 2.8681931883779104,
"grad_norm": 0.2245146292563083,
"learning_rate": 6.197929501847277e-08,
"loss": 0.8004,
"step": 932
},
{
"epoch": 2.8712718876274774,
"grad_norm": 0.22177505025695865,
"learning_rate": 5.919957794240372e-08,
"loss": 0.8086,
"step": 933
},
{
"epoch": 2.8743505868770445,
"grad_norm": 0.22084739593901448,
"learning_rate": 5.648326205579091e-08,
"loss": 0.8523,
"step": 934
},
{
"epoch": 2.8774292861266115,
"grad_norm": 0.22000923098831113,
"learning_rate": 5.383038221485948e-08,
"loss": 0.828,
"step": 935
},
{
"epoch": 2.8805079853761786,
"grad_norm": 0.23961042890787176,
"learning_rate": 5.124097246181514e-08,
"loss": 0.9008,
"step": 936
},
{
"epoch": 2.8835866846257456,
"grad_norm": 0.2640879608387801,
"learning_rate": 4.8715066024402904e-08,
"loss": 0.8572,
"step": 937
},
{
"epoch": 2.8866653838753127,
"grad_norm": 0.22459621574334815,
"learning_rate": 4.625269531548293e-08,
"loss": 0.8353,
"step": 938
},
{
"epoch": 2.8897440831248797,
"grad_norm": 0.20869331041823563,
"learning_rate": 4.385389193261314e-08,
"loss": 0.8119,
"step": 939
},
{
"epoch": 2.892822782374447,
"grad_norm": 0.2222946910903488,
"learning_rate": 4.151868665764669e-08,
"loss": 0.8178,
"step": 940
},
{
"epoch": 2.895901481624014,
"grad_norm": 0.21818676292089753,
"learning_rate": 3.9247109456332925e-08,
"loss": 0.8382,
"step": 941
},
{
"epoch": 2.898980180873581,
"grad_norm": 0.21920773908131513,
"learning_rate": 3.703918947793761e-08,
"loss": 0.8186,
"step": 942
},
{
"epoch": 2.902058880123148,
"grad_norm": 0.22806735460251343,
"learning_rate": 3.489495505486273e-08,
"loss": 0.8467,
"step": 943
},
{
"epoch": 2.905137579372715,
"grad_norm": 0.220005814740774,
"learning_rate": 3.281443370228898e-08,
"loss": 0.848,
"step": 944
},
{
"epoch": 2.908216278622282,
"grad_norm": 0.21756771263299057,
"learning_rate": 3.079765211781993e-08,
"loss": 0.8482,
"step": 945
},
{
"epoch": 2.911294977871849,
"grad_norm": 0.21937291441678433,
"learning_rate": 2.884463618113842e-08,
"loss": 0.8016,
"step": 946
},
{
"epoch": 2.914373677121416,
"grad_norm": 0.22429028671278778,
"learning_rate": 2.6955410953676265e-08,
"loss": 0.8391,
"step": 947
},
{
"epoch": 2.917452376370983,
"grad_norm": 0.2214591619964106,
"learning_rate": 2.5130000678292854e-08,
"loss": 0.8421,
"step": 948
},
{
"epoch": 2.9205310756205503,
"grad_norm": 0.2274969130277938,
"learning_rate": 2.3368428778962615e-08,
"loss": 0.8459,
"step": 949
},
{
"epoch": 2.9236097748701173,
"grad_norm": 0.22274956108684457,
"learning_rate": 2.16707178604747e-08,
"loss": 0.8018,
"step": 950
},
{
"epoch": 2.9266884741196844,
"grad_norm": 0.2226188548292508,
"learning_rate": 2.003688970814377e-08,
"loss": 0.8324,
"step": 951
},
{
"epoch": 2.9297671733692514,
"grad_norm": 0.2303807805787845,
"learning_rate": 1.846696528752967e-08,
"loss": 0.8283,
"step": 952
},
{
"epoch": 2.9328458726188185,
"grad_norm": 0.23285185453447485,
"learning_rate": 1.6960964744169306e-08,
"loss": 0.8341,
"step": 953
},
{
"epoch": 2.9359245718683855,
"grad_norm": 0.22737617282711922,
"learning_rate": 1.5518907403317406e-08,
"loss": 0.8572,
"step": 954
},
{
"epoch": 2.9390032711179526,
"grad_norm": 0.2535576923607803,
"learning_rate": 1.4140811769698392e-08,
"loss": 0.836,
"step": 955
},
{
"epoch": 2.9420819703675196,
"grad_norm": 0.2093206470793714,
"learning_rate": 1.2826695527268785e-08,
"loss": 0.7973,
"step": 956
},
{
"epoch": 2.9451606696170867,
"grad_norm": 0.21461959102738232,
"learning_rate": 1.157657553899183e-08,
"loss": 0.8083,
"step": 957
},
{
"epoch": 2.9482393688666537,
"grad_norm": 0.22128779738744017,
"learning_rate": 1.039046784661879e-08,
"loss": 0.8611,
"step": 958
},
{
"epoch": 2.9513180681162208,
"grad_norm": 0.22770935864286593,
"learning_rate": 9.268387670485213e-09,
"loss": 0.8165,
"step": 959
},
{
"epoch": 2.954396767365788,
"grad_norm": 0.23093405014015922,
"learning_rate": 8.210349409314422e-09,
"loss": 0.8421,
"step": 960
},
{
"epoch": 2.957475466615355,
"grad_norm": 0.22738585887837046,
"learning_rate": 7.216366640032668e-09,
"loss": 0.8409,
"step": 961
},
{
"epoch": 2.960554165864922,
"grad_norm": 0.224297760541658,
"learning_rate": 6.286452117595932e-09,
"loss": 0.8613,
"step": 962
},
{
"epoch": 2.963632865114489,
"grad_norm": 0.2128855509377922,
"learning_rate": 5.420617774824499e-09,
"loss": 0.7809,
"step": 963
},
{
"epoch": 2.966711564364056,
"grad_norm": 0.21661800052382058,
"learning_rate": 4.618874722252531e-09,
"loss": 0.8044,
"step": 964
},
{
"epoch": 2.969790263613623,
"grad_norm": 0.22262200467992604,
"learning_rate": 3.881233247982063e-09,
"loss": 0.8703,
"step": 965
},
{
"epoch": 2.97286896286319,
"grad_norm": 0.22502348135706718,
"learning_rate": 3.2077028175536706e-09,
"loss": 0.8169,
"step": 966
},
{
"epoch": 2.975947662112757,
"grad_norm": 0.22849132510040063,
"learning_rate": 2.5982920738248927e-09,
"loss": 0.8774,
"step": 967
},
{
"epoch": 2.9790263613623242,
"grad_norm": 0.22319933041136963,
"learning_rate": 2.0530088368569957e-09,
"loss": 0.7804,
"step": 968
},
{
"epoch": 2.9821050606118913,
"grad_norm": 0.24101643126984526,
"learning_rate": 1.5718601038178238e-09,
"loss": 0.8576,
"step": 969
},
{
"epoch": 2.9851837598614583,
"grad_norm": 0.22216270925310933,
"learning_rate": 1.154852048888544e-09,
"loss": 0.8476,
"step": 970
},
{
"epoch": 2.9882624591110254,
"grad_norm": 0.22704689144079418,
"learning_rate": 8.019900231881483e-10,
"loss": 0.8418,
"step": 971
},
{
"epoch": 2.9913411583605924,
"grad_norm": 0.22180565383914788,
"learning_rate": 5.132785547018459e-10,
"loss": 0.8555,
"step": 972
},
{
"epoch": 2.9944198576101595,
"grad_norm": 0.232096124052605,
"learning_rate": 2.88721348223886e-10,
"loss": 0.8316,
"step": 973
},
{
"epoch": 2.9974985568597265,
"grad_norm": 0.22593898743230423,
"learning_rate": 1.2832128531092837e-10,
"loss": 0.8627,
"step": 974
},
{
"epoch": 3.0,
"grad_norm": 0.24139180737433957,
"learning_rate": 3.208042424318603e-11,
"loss": 0.8179,
"step": 975
}
],
"logging_steps": 1,
"max_steps": 975,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1889486304706560.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}