Files
openthoughts3_100k_qwen25_1…/trainer_state.json
ModelHub XC 5994dc82b0 初始化项目,由ModelHub XC社区提供模型
Model: mlfoundations-dev/openthoughts3_100k_qwen25_1b_bsz256_lr4e5_epochs7
Source: Original Platform
2026-06-26 19:34:12 +08:00

19202 lines
465 KiB
JSON

{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 7.0,
"eval_steps": 500,
"global_step": 2737,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0025575447570332483,
"grad_norm": 2.9635707889439162,
"learning_rate": 1.4598540145985402e-07,
"loss": 1.5218,
"step": 1
},
{
"epoch": 0.005115089514066497,
"grad_norm": 2.957024717990724,
"learning_rate": 2.9197080291970804e-07,
"loss": 1.4755,
"step": 2
},
{
"epoch": 0.0076726342710997444,
"grad_norm": 3.015694338099643,
"learning_rate": 4.379562043795621e-07,
"loss": 1.4935,
"step": 3
},
{
"epoch": 0.010230179028132993,
"grad_norm": 2.950951521796724,
"learning_rate": 5.839416058394161e-07,
"loss": 1.4812,
"step": 4
},
{
"epoch": 0.01278772378516624,
"grad_norm": 2.9774099804734737,
"learning_rate": 7.299270072992701e-07,
"loss": 1.494,
"step": 5
},
{
"epoch": 0.015345268542199489,
"grad_norm": 2.9962593760046197,
"learning_rate": 8.759124087591242e-07,
"loss": 1.5061,
"step": 6
},
{
"epoch": 0.017902813299232736,
"grad_norm": 2.8615759185642897,
"learning_rate": 1.0218978102189781e-06,
"loss": 1.4765,
"step": 7
},
{
"epoch": 0.020460358056265986,
"grad_norm": 2.883215776492285,
"learning_rate": 1.1678832116788322e-06,
"loss": 1.4864,
"step": 8
},
{
"epoch": 0.023017902813299233,
"grad_norm": 2.659077744646757,
"learning_rate": 1.3138686131386864e-06,
"loss": 1.5167,
"step": 9
},
{
"epoch": 0.02557544757033248,
"grad_norm": 2.6459608432449833,
"learning_rate": 1.4598540145985402e-06,
"loss": 1.5092,
"step": 10
},
{
"epoch": 0.028132992327365727,
"grad_norm": 2.562232774245823,
"learning_rate": 1.6058394160583942e-06,
"loss": 1.4999,
"step": 11
},
{
"epoch": 0.030690537084398978,
"grad_norm": 2.0596527475508246,
"learning_rate": 1.7518248175182485e-06,
"loss": 1.4546,
"step": 12
},
{
"epoch": 0.03324808184143223,
"grad_norm": 1.990692995362292,
"learning_rate": 1.8978102189781023e-06,
"loss": 1.4966,
"step": 13
},
{
"epoch": 0.03580562659846547,
"grad_norm": 1.954244091451144,
"learning_rate": 2.0437956204379563e-06,
"loss": 1.4433,
"step": 14
},
{
"epoch": 0.03836317135549872,
"grad_norm": 1.8910060928671393,
"learning_rate": 2.1897810218978103e-06,
"loss": 1.4802,
"step": 15
},
{
"epoch": 0.04092071611253197,
"grad_norm": 1.4257423396886646,
"learning_rate": 2.3357664233576643e-06,
"loss": 1.4242,
"step": 16
},
{
"epoch": 0.043478260869565216,
"grad_norm": 1.685733393905784,
"learning_rate": 2.4817518248175183e-06,
"loss": 1.4457,
"step": 17
},
{
"epoch": 0.04603580562659847,
"grad_norm": 1.739309250439508,
"learning_rate": 2.627737226277373e-06,
"loss": 1.4517,
"step": 18
},
{
"epoch": 0.04859335038363171,
"grad_norm": 1.6697333213062746,
"learning_rate": 2.7737226277372264e-06,
"loss": 1.4402,
"step": 19
},
{
"epoch": 0.05115089514066496,
"grad_norm": 1.5016099487717074,
"learning_rate": 2.9197080291970804e-06,
"loss": 1.429,
"step": 20
},
{
"epoch": 0.05370843989769821,
"grad_norm": 1.3707698782204212,
"learning_rate": 3.065693430656935e-06,
"loss": 1.4112,
"step": 21
},
{
"epoch": 0.056265984654731455,
"grad_norm": 1.18341491640736,
"learning_rate": 3.2116788321167884e-06,
"loss": 1.4202,
"step": 22
},
{
"epoch": 0.058823529411764705,
"grad_norm": 1.3372875723371354,
"learning_rate": 3.3576642335766425e-06,
"loss": 1.3742,
"step": 23
},
{
"epoch": 0.061381074168797956,
"grad_norm": 1.351848069752737,
"learning_rate": 3.503649635036497e-06,
"loss": 1.3971,
"step": 24
},
{
"epoch": 0.0639386189258312,
"grad_norm": 1.292208492976159,
"learning_rate": 3.6496350364963505e-06,
"loss": 1.372,
"step": 25
},
{
"epoch": 0.06649616368286446,
"grad_norm": 1.110960602074978,
"learning_rate": 3.7956204379562045e-06,
"loss": 1.3751,
"step": 26
},
{
"epoch": 0.06905370843989769,
"grad_norm": 1.0056051422338692,
"learning_rate": 3.9416058394160585e-06,
"loss": 1.362,
"step": 27
},
{
"epoch": 0.07161125319693094,
"grad_norm": 0.8433663455791572,
"learning_rate": 4.0875912408759126e-06,
"loss": 1.384,
"step": 28
},
{
"epoch": 0.0741687979539642,
"grad_norm": 0.6761749715490012,
"learning_rate": 4.233576642335767e-06,
"loss": 1.3591,
"step": 29
},
{
"epoch": 0.07672634271099744,
"grad_norm": 0.827159716206725,
"learning_rate": 4.379562043795621e-06,
"loss": 1.3656,
"step": 30
},
{
"epoch": 0.0792838874680307,
"grad_norm": 0.9573359587436694,
"learning_rate": 4.525547445255475e-06,
"loss": 1.3598,
"step": 31
},
{
"epoch": 0.08184143222506395,
"grad_norm": 0.8695899842670742,
"learning_rate": 4.671532846715329e-06,
"loss": 1.3339,
"step": 32
},
{
"epoch": 0.08439897698209718,
"grad_norm": 0.7364805454796883,
"learning_rate": 4.8175182481751835e-06,
"loss": 1.3593,
"step": 33
},
{
"epoch": 0.08695652173913043,
"grad_norm": 0.6973875864304822,
"learning_rate": 4.963503649635037e-06,
"loss": 1.3734,
"step": 34
},
{
"epoch": 0.08951406649616368,
"grad_norm": 0.6501434750619707,
"learning_rate": 5.1094890510948916e-06,
"loss": 1.3171,
"step": 35
},
{
"epoch": 0.09207161125319693,
"grad_norm": 0.6110832361242909,
"learning_rate": 5.255474452554746e-06,
"loss": 1.3302,
"step": 36
},
{
"epoch": 0.09462915601023018,
"grad_norm": 0.6245992853725667,
"learning_rate": 5.401459854014599e-06,
"loss": 1.3449,
"step": 37
},
{
"epoch": 0.09718670076726342,
"grad_norm": 0.598041117546877,
"learning_rate": 5.547445255474453e-06,
"loss": 1.3153,
"step": 38
},
{
"epoch": 0.09974424552429667,
"grad_norm": 0.5716481030800229,
"learning_rate": 5.693430656934307e-06,
"loss": 1.3333,
"step": 39
},
{
"epoch": 0.10230179028132992,
"grad_norm": 0.5027719688463927,
"learning_rate": 5.839416058394161e-06,
"loss": 1.2736,
"step": 40
},
{
"epoch": 0.10485933503836317,
"grad_norm": 0.4531288485526633,
"learning_rate": 5.985401459854016e-06,
"loss": 1.3006,
"step": 41
},
{
"epoch": 0.10741687979539642,
"grad_norm": 0.41509553123439175,
"learning_rate": 6.13138686131387e-06,
"loss": 1.2979,
"step": 42
},
{
"epoch": 0.10997442455242967,
"grad_norm": 0.44762981508449373,
"learning_rate": 6.277372262773723e-06,
"loss": 1.271,
"step": 43
},
{
"epoch": 0.11253196930946291,
"grad_norm": 0.4861076235923236,
"learning_rate": 6.423357664233577e-06,
"loss": 1.2759,
"step": 44
},
{
"epoch": 0.11508951406649616,
"grad_norm": 0.43606428852527673,
"learning_rate": 6.569343065693431e-06,
"loss": 1.2784,
"step": 45
},
{
"epoch": 0.11764705882352941,
"grad_norm": 0.39180019328459315,
"learning_rate": 6.715328467153285e-06,
"loss": 1.2979,
"step": 46
},
{
"epoch": 0.12020460358056266,
"grad_norm": 0.3801825137267237,
"learning_rate": 6.86131386861314e-06,
"loss": 1.3004,
"step": 47
},
{
"epoch": 0.12276214833759591,
"grad_norm": 0.3552352437266276,
"learning_rate": 7.007299270072994e-06,
"loss": 1.2682,
"step": 48
},
{
"epoch": 0.12531969309462915,
"grad_norm": 0.34471939435704263,
"learning_rate": 7.153284671532848e-06,
"loss": 1.2824,
"step": 49
},
{
"epoch": 0.1278772378516624,
"grad_norm": 0.40797013792942205,
"learning_rate": 7.299270072992701e-06,
"loss": 1.2524,
"step": 50
},
{
"epoch": 0.13043478260869565,
"grad_norm": 0.32937016885557413,
"learning_rate": 7.445255474452555e-06,
"loss": 1.2566,
"step": 51
},
{
"epoch": 0.1329923273657289,
"grad_norm": 0.27877623434500126,
"learning_rate": 7.591240875912409e-06,
"loss": 1.2377,
"step": 52
},
{
"epoch": 0.13554987212276215,
"grad_norm": 0.30154653697862177,
"learning_rate": 7.737226277372264e-06,
"loss": 1.2347,
"step": 53
},
{
"epoch": 0.13810741687979539,
"grad_norm": 0.30706238768956157,
"learning_rate": 7.883211678832117e-06,
"loss": 1.2741,
"step": 54
},
{
"epoch": 0.14066496163682865,
"grad_norm": 0.30302402776641885,
"learning_rate": 8.029197080291972e-06,
"loss": 1.2862,
"step": 55
},
{
"epoch": 0.1432225063938619,
"grad_norm": 0.27222666982060817,
"learning_rate": 8.175182481751825e-06,
"loss": 1.2718,
"step": 56
},
{
"epoch": 0.14578005115089515,
"grad_norm": 0.2599088882708423,
"learning_rate": 8.32116788321168e-06,
"loss": 1.2575,
"step": 57
},
{
"epoch": 0.1483375959079284,
"grad_norm": 0.26793126679750795,
"learning_rate": 8.467153284671533e-06,
"loss": 1.2734,
"step": 58
},
{
"epoch": 0.15089514066496162,
"grad_norm": 0.27372518112672745,
"learning_rate": 8.613138686131386e-06,
"loss": 1.2346,
"step": 59
},
{
"epoch": 0.1534526854219949,
"grad_norm": 0.2705606078561125,
"learning_rate": 8.759124087591241e-06,
"loss": 1.2595,
"step": 60
},
{
"epoch": 0.15601023017902813,
"grad_norm": 0.2462898207533062,
"learning_rate": 8.905109489051096e-06,
"loss": 1.2479,
"step": 61
},
{
"epoch": 0.1585677749360614,
"grad_norm": 0.27734973837157667,
"learning_rate": 9.05109489051095e-06,
"loss": 1.2819,
"step": 62
},
{
"epoch": 0.16112531969309463,
"grad_norm": 0.2333355630940424,
"learning_rate": 9.197080291970804e-06,
"loss": 1.2596,
"step": 63
},
{
"epoch": 0.1636828644501279,
"grad_norm": 0.25390170276984986,
"learning_rate": 9.343065693430657e-06,
"loss": 1.2555,
"step": 64
},
{
"epoch": 0.16624040920716113,
"grad_norm": 0.23611526162550192,
"learning_rate": 9.48905109489051e-06,
"loss": 1.2493,
"step": 65
},
{
"epoch": 0.16879795396419436,
"grad_norm": 0.2080882700430996,
"learning_rate": 9.635036496350367e-06,
"loss": 1.2094,
"step": 66
},
{
"epoch": 0.17135549872122763,
"grad_norm": 0.21238380361161624,
"learning_rate": 9.78102189781022e-06,
"loss": 1.2302,
"step": 67
},
{
"epoch": 0.17391304347826086,
"grad_norm": 0.20495347093471789,
"learning_rate": 9.927007299270073e-06,
"loss": 1.2109,
"step": 68
},
{
"epoch": 0.17647058823529413,
"grad_norm": 0.23416256497223187,
"learning_rate": 1.0072992700729928e-05,
"loss": 1.195,
"step": 69
},
{
"epoch": 0.17902813299232737,
"grad_norm": 0.19544683543539945,
"learning_rate": 1.0218978102189783e-05,
"loss": 1.2051,
"step": 70
},
{
"epoch": 0.1815856777493606,
"grad_norm": 0.20193557183776129,
"learning_rate": 1.0364963503649636e-05,
"loss": 1.2155,
"step": 71
},
{
"epoch": 0.18414322250639387,
"grad_norm": 0.18843006705162113,
"learning_rate": 1.0510948905109491e-05,
"loss": 1.2299,
"step": 72
},
{
"epoch": 0.1867007672634271,
"grad_norm": 0.2019494475484416,
"learning_rate": 1.0656934306569344e-05,
"loss": 1.2623,
"step": 73
},
{
"epoch": 0.18925831202046037,
"grad_norm": 0.20249152976175708,
"learning_rate": 1.0802919708029198e-05,
"loss": 1.2084,
"step": 74
},
{
"epoch": 0.1918158567774936,
"grad_norm": 0.1908970556266133,
"learning_rate": 1.0948905109489052e-05,
"loss": 1.2069,
"step": 75
},
{
"epoch": 0.19437340153452684,
"grad_norm": 0.20686065176986063,
"learning_rate": 1.1094890510948906e-05,
"loss": 1.2127,
"step": 76
},
{
"epoch": 0.1969309462915601,
"grad_norm": 0.2137483908926262,
"learning_rate": 1.124087591240876e-05,
"loss": 1.2187,
"step": 77
},
{
"epoch": 0.19948849104859334,
"grad_norm": 0.20357397399438737,
"learning_rate": 1.1386861313868614e-05,
"loss": 1.2014,
"step": 78
},
{
"epoch": 0.2020460358056266,
"grad_norm": 0.1910729311244722,
"learning_rate": 1.1532846715328467e-05,
"loss": 1.195,
"step": 79
},
{
"epoch": 0.20460358056265984,
"grad_norm": 0.19104673049889764,
"learning_rate": 1.1678832116788322e-05,
"loss": 1.2109,
"step": 80
},
{
"epoch": 0.2071611253196931,
"grad_norm": 0.18843918828784803,
"learning_rate": 1.1824817518248176e-05,
"loss": 1.1945,
"step": 81
},
{
"epoch": 0.20971867007672634,
"grad_norm": 0.22237675262983145,
"learning_rate": 1.1970802919708031e-05,
"loss": 1.1943,
"step": 82
},
{
"epoch": 0.21227621483375958,
"grad_norm": 0.19585062326368735,
"learning_rate": 1.2116788321167885e-05,
"loss": 1.2067,
"step": 83
},
{
"epoch": 0.21483375959079284,
"grad_norm": 0.20796693138186426,
"learning_rate": 1.226277372262774e-05,
"loss": 1.2037,
"step": 84
},
{
"epoch": 0.21739130434782608,
"grad_norm": 0.2420142629013636,
"learning_rate": 1.2408759124087593e-05,
"loss": 1.2526,
"step": 85
},
{
"epoch": 0.21994884910485935,
"grad_norm": 0.19264306309885623,
"learning_rate": 1.2554744525547446e-05,
"loss": 1.2051,
"step": 86
},
{
"epoch": 0.22250639386189258,
"grad_norm": 0.23572212522126862,
"learning_rate": 1.27007299270073e-05,
"loss": 1.2105,
"step": 87
},
{
"epoch": 0.22506393861892582,
"grad_norm": 0.18676619235375888,
"learning_rate": 1.2846715328467154e-05,
"loss": 1.2331,
"step": 88
},
{
"epoch": 0.22762148337595908,
"grad_norm": 0.22263641910877627,
"learning_rate": 1.2992700729927009e-05,
"loss": 1.193,
"step": 89
},
{
"epoch": 0.23017902813299232,
"grad_norm": 0.20332847304472856,
"learning_rate": 1.3138686131386862e-05,
"loss": 1.2218,
"step": 90
},
{
"epoch": 0.23273657289002558,
"grad_norm": 0.19279934529354495,
"learning_rate": 1.3284671532846715e-05,
"loss": 1.2187,
"step": 91
},
{
"epoch": 0.23529411764705882,
"grad_norm": 0.18535284736119417,
"learning_rate": 1.343065693430657e-05,
"loss": 1.2233,
"step": 92
},
{
"epoch": 0.23785166240409208,
"grad_norm": 0.1922516164053042,
"learning_rate": 1.3576642335766423e-05,
"loss": 1.2045,
"step": 93
},
{
"epoch": 0.24040920716112532,
"grad_norm": 0.19869568512009167,
"learning_rate": 1.372262773722628e-05,
"loss": 1.228,
"step": 94
},
{
"epoch": 0.24296675191815856,
"grad_norm": 0.2126076133117338,
"learning_rate": 1.3868613138686133e-05,
"loss": 1.1727,
"step": 95
},
{
"epoch": 0.24552429667519182,
"grad_norm": 0.20707252702622744,
"learning_rate": 1.4014598540145988e-05,
"loss": 1.2123,
"step": 96
},
{
"epoch": 0.24808184143222506,
"grad_norm": 0.22477804993156936,
"learning_rate": 1.416058394160584e-05,
"loss": 1.2096,
"step": 97
},
{
"epoch": 0.2506393861892583,
"grad_norm": 0.22488754031476965,
"learning_rate": 1.4306569343065696e-05,
"loss": 1.1826,
"step": 98
},
{
"epoch": 0.2531969309462916,
"grad_norm": 0.21839730256494638,
"learning_rate": 1.4452554744525549e-05,
"loss": 1.1815,
"step": 99
},
{
"epoch": 0.2557544757033248,
"grad_norm": 0.22159986357787417,
"learning_rate": 1.4598540145985402e-05,
"loss": 1.1745,
"step": 100
},
{
"epoch": 0.25831202046035806,
"grad_norm": 0.21874152644245082,
"learning_rate": 1.4744525547445257e-05,
"loss": 1.1938,
"step": 101
},
{
"epoch": 0.2608695652173913,
"grad_norm": 0.25032119638816225,
"learning_rate": 1.489051094890511e-05,
"loss": 1.2036,
"step": 102
},
{
"epoch": 0.26342710997442453,
"grad_norm": 0.23926399305386825,
"learning_rate": 1.5036496350364965e-05,
"loss": 1.173,
"step": 103
},
{
"epoch": 0.2659846547314578,
"grad_norm": 0.2744133562234567,
"learning_rate": 1.5182481751824818e-05,
"loss": 1.1643,
"step": 104
},
{
"epoch": 0.26854219948849106,
"grad_norm": 0.2250964324002027,
"learning_rate": 1.5328467153284673e-05,
"loss": 1.1722,
"step": 105
},
{
"epoch": 0.2710997442455243,
"grad_norm": 0.26186788501699393,
"learning_rate": 1.5474452554744528e-05,
"loss": 1.1618,
"step": 106
},
{
"epoch": 0.27365728900255754,
"grad_norm": 0.21232956768327443,
"learning_rate": 1.5620437956204383e-05,
"loss": 1.1903,
"step": 107
},
{
"epoch": 0.27621483375959077,
"grad_norm": 0.2100552175310244,
"learning_rate": 1.5766423357664234e-05,
"loss": 1.1698,
"step": 108
},
{
"epoch": 0.27877237851662406,
"grad_norm": 0.24390785271475385,
"learning_rate": 1.591240875912409e-05,
"loss": 1.1925,
"step": 109
},
{
"epoch": 0.2813299232736573,
"grad_norm": 0.26134533420997114,
"learning_rate": 1.6058394160583944e-05,
"loss": 1.1934,
"step": 110
},
{
"epoch": 0.28388746803069054,
"grad_norm": 0.25275807079953994,
"learning_rate": 1.62043795620438e-05,
"loss": 1.1951,
"step": 111
},
{
"epoch": 0.2864450127877238,
"grad_norm": 0.2520419721199999,
"learning_rate": 1.635036496350365e-05,
"loss": 1.1832,
"step": 112
},
{
"epoch": 0.289002557544757,
"grad_norm": 0.2699916516897491,
"learning_rate": 1.6496350364963505e-05,
"loss": 1.1562,
"step": 113
},
{
"epoch": 0.2915601023017903,
"grad_norm": 0.28841556568198157,
"learning_rate": 1.664233576642336e-05,
"loss": 1.1821,
"step": 114
},
{
"epoch": 0.29411764705882354,
"grad_norm": 0.22822114978201002,
"learning_rate": 1.678832116788321e-05,
"loss": 1.1653,
"step": 115
},
{
"epoch": 0.2966751918158568,
"grad_norm": 0.26777871821383475,
"learning_rate": 1.6934306569343066e-05,
"loss": 1.1741,
"step": 116
},
{
"epoch": 0.29923273657289,
"grad_norm": 0.25676620734685335,
"learning_rate": 1.708029197080292e-05,
"loss": 1.1695,
"step": 117
},
{
"epoch": 0.30179028132992325,
"grad_norm": 0.27081965006009656,
"learning_rate": 1.7226277372262773e-05,
"loss": 1.2189,
"step": 118
},
{
"epoch": 0.30434782608695654,
"grad_norm": 0.23398961930258177,
"learning_rate": 1.737226277372263e-05,
"loss": 1.1799,
"step": 119
},
{
"epoch": 0.3069053708439898,
"grad_norm": 0.25981615353926174,
"learning_rate": 1.7518248175182482e-05,
"loss": 1.186,
"step": 120
},
{
"epoch": 0.309462915601023,
"grad_norm": 0.21881297550639361,
"learning_rate": 1.7664233576642337e-05,
"loss": 1.1996,
"step": 121
},
{
"epoch": 0.31202046035805625,
"grad_norm": 0.2394761659661747,
"learning_rate": 1.7810218978102192e-05,
"loss": 1.1524,
"step": 122
},
{
"epoch": 0.3145780051150895,
"grad_norm": 0.27029722802671036,
"learning_rate": 1.7956204379562047e-05,
"loss": 1.1726,
"step": 123
},
{
"epoch": 0.3171355498721228,
"grad_norm": 0.24326082746743632,
"learning_rate": 1.81021897810219e-05,
"loss": 1.1687,
"step": 124
},
{
"epoch": 0.319693094629156,
"grad_norm": 0.31089002912951674,
"learning_rate": 1.8248175182481753e-05,
"loss": 1.146,
"step": 125
},
{
"epoch": 0.32225063938618925,
"grad_norm": 0.32660720710403346,
"learning_rate": 1.8394160583941608e-05,
"loss": 1.2032,
"step": 126
},
{
"epoch": 0.3248081841432225,
"grad_norm": 0.35812683374401805,
"learning_rate": 1.854014598540146e-05,
"loss": 1.145,
"step": 127
},
{
"epoch": 0.3273657289002558,
"grad_norm": 0.3787635661888352,
"learning_rate": 1.8686131386861315e-05,
"loss": 1.1708,
"step": 128
},
{
"epoch": 0.329923273657289,
"grad_norm": 0.4152825867413198,
"learning_rate": 1.883211678832117e-05,
"loss": 1.1599,
"step": 129
},
{
"epoch": 0.33248081841432225,
"grad_norm": 0.4282035454252041,
"learning_rate": 1.897810218978102e-05,
"loss": 1.1881,
"step": 130
},
{
"epoch": 0.3350383631713555,
"grad_norm": 0.5541202351716032,
"learning_rate": 1.912408759124088e-05,
"loss": 1.1554,
"step": 131
},
{
"epoch": 0.3375959079283887,
"grad_norm": 0.6117341491356942,
"learning_rate": 1.9270072992700734e-05,
"loss": 1.2128,
"step": 132
},
{
"epoch": 0.340153452685422,
"grad_norm": 0.43596536411889675,
"learning_rate": 1.9416058394160586e-05,
"loss": 1.1841,
"step": 133
},
{
"epoch": 0.34271099744245526,
"grad_norm": 0.3011840555513657,
"learning_rate": 1.956204379562044e-05,
"loss": 1.1562,
"step": 134
},
{
"epoch": 0.3452685421994885,
"grad_norm": 0.2972812271508846,
"learning_rate": 1.9708029197080295e-05,
"loss": 1.2119,
"step": 135
},
{
"epoch": 0.34782608695652173,
"grad_norm": 0.3639730740685411,
"learning_rate": 1.9854014598540147e-05,
"loss": 1.1698,
"step": 136
},
{
"epoch": 0.35038363171355497,
"grad_norm": 0.4395398830915075,
"learning_rate": 2e-05,
"loss": 1.183,
"step": 137
},
{
"epoch": 0.35294117647058826,
"grad_norm": 0.4175787696771395,
"learning_rate": 2.0145985401459857e-05,
"loss": 1.1707,
"step": 138
},
{
"epoch": 0.3554987212276215,
"grad_norm": 0.3018791391904324,
"learning_rate": 2.0291970802919708e-05,
"loss": 1.147,
"step": 139
},
{
"epoch": 0.35805626598465473,
"grad_norm": 0.2913415095914862,
"learning_rate": 2.0437956204379566e-05,
"loss": 1.201,
"step": 140
},
{
"epoch": 0.36061381074168797,
"grad_norm": 0.25899928611353373,
"learning_rate": 2.0583941605839418e-05,
"loss": 1.1145,
"step": 141
},
{
"epoch": 0.3631713554987212,
"grad_norm": 0.27116705541573327,
"learning_rate": 2.0729927007299273e-05,
"loss": 1.1687,
"step": 142
},
{
"epoch": 0.3657289002557545,
"grad_norm": 0.42730907805551704,
"learning_rate": 2.0875912408759124e-05,
"loss": 1.1931,
"step": 143
},
{
"epoch": 0.36828644501278773,
"grad_norm": 0.4469179700471288,
"learning_rate": 2.1021897810218982e-05,
"loss": 1.1606,
"step": 144
},
{
"epoch": 0.37084398976982097,
"grad_norm": 0.3332148227022474,
"learning_rate": 2.1167883211678834e-05,
"loss": 1.1749,
"step": 145
},
{
"epoch": 0.3734015345268542,
"grad_norm": 0.34621699390104765,
"learning_rate": 2.131386861313869e-05,
"loss": 1.1617,
"step": 146
},
{
"epoch": 0.37595907928388744,
"grad_norm": 0.44938000603819805,
"learning_rate": 2.145985401459854e-05,
"loss": 1.1939,
"step": 147
},
{
"epoch": 0.37851662404092073,
"grad_norm": 0.37996197622756905,
"learning_rate": 2.1605839416058395e-05,
"loss": 1.1568,
"step": 148
},
{
"epoch": 0.38107416879795397,
"grad_norm": 0.28957096566777796,
"learning_rate": 2.1751824817518246e-05,
"loss": 1.1389,
"step": 149
},
{
"epoch": 0.3836317135549872,
"grad_norm": 0.37131922891602737,
"learning_rate": 2.1897810218978105e-05,
"loss": 1.1547,
"step": 150
},
{
"epoch": 0.38618925831202044,
"grad_norm": 0.4829052475341007,
"learning_rate": 2.204379562043796e-05,
"loss": 1.159,
"step": 151
},
{
"epoch": 0.3887468030690537,
"grad_norm": 0.5847679519882145,
"learning_rate": 2.218978102189781e-05,
"loss": 1.1627,
"step": 152
},
{
"epoch": 0.391304347826087,
"grad_norm": 0.7370600260219867,
"learning_rate": 2.2335766423357666e-05,
"loss": 1.1521,
"step": 153
},
{
"epoch": 0.3938618925831202,
"grad_norm": 0.9140897429060579,
"learning_rate": 2.248175182481752e-05,
"loss": 1.1563,
"step": 154
},
{
"epoch": 0.39641943734015345,
"grad_norm": 0.9263340717297044,
"learning_rate": 2.2627737226277376e-05,
"loss": 1.1701,
"step": 155
},
{
"epoch": 0.3989769820971867,
"grad_norm": 0.7327714722145657,
"learning_rate": 2.2773722627737227e-05,
"loss": 1.1746,
"step": 156
},
{
"epoch": 0.40153452685422,
"grad_norm": 0.5992586132942535,
"learning_rate": 2.2919708029197082e-05,
"loss": 1.1851,
"step": 157
},
{
"epoch": 0.4040920716112532,
"grad_norm": 0.5353409844690658,
"learning_rate": 2.3065693430656934e-05,
"loss": 1.1833,
"step": 158
},
{
"epoch": 0.40664961636828645,
"grad_norm": 0.5166800693827149,
"learning_rate": 2.3211678832116792e-05,
"loss": 1.1329,
"step": 159
},
{
"epoch": 0.4092071611253197,
"grad_norm": 0.5419472931985324,
"learning_rate": 2.3357664233576643e-05,
"loss": 1.1512,
"step": 160
},
{
"epoch": 0.4117647058823529,
"grad_norm": 0.5320546683048022,
"learning_rate": 2.3503649635036498e-05,
"loss": 1.1447,
"step": 161
},
{
"epoch": 0.4143222506393862,
"grad_norm": 0.5040432864125984,
"learning_rate": 2.3649635036496353e-05,
"loss": 1.1437,
"step": 162
},
{
"epoch": 0.41687979539641945,
"grad_norm": 0.4605512314669385,
"learning_rate": 2.3795620437956204e-05,
"loss": 1.1439,
"step": 163
},
{
"epoch": 0.4194373401534527,
"grad_norm": 0.4380520127279597,
"learning_rate": 2.3941605839416063e-05,
"loss": 1.1574,
"step": 164
},
{
"epoch": 0.4219948849104859,
"grad_norm": 0.6124557207640321,
"learning_rate": 2.4087591240875914e-05,
"loss": 1.1291,
"step": 165
},
{
"epoch": 0.42455242966751916,
"grad_norm": 0.7360038801117461,
"learning_rate": 2.423357664233577e-05,
"loss": 1.1555,
"step": 166
},
{
"epoch": 0.42710997442455245,
"grad_norm": 0.7671524886201543,
"learning_rate": 2.437956204379562e-05,
"loss": 1.1533,
"step": 167
},
{
"epoch": 0.4296675191815857,
"grad_norm": 0.6676800807341117,
"learning_rate": 2.452554744525548e-05,
"loss": 1.1576,
"step": 168
},
{
"epoch": 0.4322250639386189,
"grad_norm": 0.48987762251415845,
"learning_rate": 2.467153284671533e-05,
"loss": 1.1558,
"step": 169
},
{
"epoch": 0.43478260869565216,
"grad_norm": 0.422823186730046,
"learning_rate": 2.4817518248175185e-05,
"loss": 1.1398,
"step": 170
},
{
"epoch": 0.4373401534526854,
"grad_norm": 0.7156968300683684,
"learning_rate": 2.4963503649635037e-05,
"loss": 1.1502,
"step": 171
},
{
"epoch": 0.4398976982097187,
"grad_norm": 0.7774809728475114,
"learning_rate": 2.510948905109489e-05,
"loss": 1.1495,
"step": 172
},
{
"epoch": 0.4424552429667519,
"grad_norm": 0.5745026917019085,
"learning_rate": 2.5255474452554743e-05,
"loss": 1.1679,
"step": 173
},
{
"epoch": 0.44501278772378516,
"grad_norm": 0.4958368977342923,
"learning_rate": 2.54014598540146e-05,
"loss": 1.1745,
"step": 174
},
{
"epoch": 0.4475703324808184,
"grad_norm": 0.4043791753673972,
"learning_rate": 2.5547445255474456e-05,
"loss": 1.1494,
"step": 175
},
{
"epoch": 0.45012787723785164,
"grad_norm": 0.4349730560314562,
"learning_rate": 2.5693430656934308e-05,
"loss": 1.1779,
"step": 176
},
{
"epoch": 0.45268542199488493,
"grad_norm": 0.47480919267885613,
"learning_rate": 2.5839416058394166e-05,
"loss": 1.1507,
"step": 177
},
{
"epoch": 0.45524296675191817,
"grad_norm": 0.44257586581284925,
"learning_rate": 2.5985401459854017e-05,
"loss": 1.1471,
"step": 178
},
{
"epoch": 0.4578005115089514,
"grad_norm": 0.39395419005614024,
"learning_rate": 2.6131386861313872e-05,
"loss": 1.1338,
"step": 179
},
{
"epoch": 0.46035805626598464,
"grad_norm": 0.46197349263848936,
"learning_rate": 2.6277372262773724e-05,
"loss": 1.1413,
"step": 180
},
{
"epoch": 0.4629156010230179,
"grad_norm": 0.5467171204098674,
"learning_rate": 2.642335766423358e-05,
"loss": 1.1679,
"step": 181
},
{
"epoch": 0.46547314578005117,
"grad_norm": 0.4456855144725915,
"learning_rate": 2.656934306569343e-05,
"loss": 1.147,
"step": 182
},
{
"epoch": 0.4680306905370844,
"grad_norm": 0.35352947833928405,
"learning_rate": 2.6715328467153288e-05,
"loss": 1.1555,
"step": 183
},
{
"epoch": 0.47058823529411764,
"grad_norm": 0.37604438710478477,
"learning_rate": 2.686131386861314e-05,
"loss": 1.1527,
"step": 184
},
{
"epoch": 0.4731457800511509,
"grad_norm": 0.4017451379953799,
"learning_rate": 2.7007299270072995e-05,
"loss": 1.1562,
"step": 185
},
{
"epoch": 0.47570332480818417,
"grad_norm": 0.3355674535932529,
"learning_rate": 2.7153284671532846e-05,
"loss": 1.1575,
"step": 186
},
{
"epoch": 0.4782608695652174,
"grad_norm": 0.3130478459080379,
"learning_rate": 2.7299270072992704e-05,
"loss": 1.1373,
"step": 187
},
{
"epoch": 0.48081841432225064,
"grad_norm": 0.4446718712601563,
"learning_rate": 2.744525547445256e-05,
"loss": 1.1668,
"step": 188
},
{
"epoch": 0.4833759590792839,
"grad_norm": 0.540129263235713,
"learning_rate": 2.759124087591241e-05,
"loss": 1.1438,
"step": 189
},
{
"epoch": 0.4859335038363171,
"grad_norm": 0.5922167087130495,
"learning_rate": 2.7737226277372266e-05,
"loss": 1.1504,
"step": 190
},
{
"epoch": 0.4884910485933504,
"grad_norm": 0.5075063271222039,
"learning_rate": 2.7883211678832117e-05,
"loss": 1.0951,
"step": 191
},
{
"epoch": 0.49104859335038364,
"grad_norm": 0.5320045362227517,
"learning_rate": 2.8029197080291975e-05,
"loss": 1.1383,
"step": 192
},
{
"epoch": 0.4936061381074169,
"grad_norm": 0.5313524301827326,
"learning_rate": 2.8175182481751827e-05,
"loss": 1.1124,
"step": 193
},
{
"epoch": 0.4961636828644501,
"grad_norm": 0.6551170290782115,
"learning_rate": 2.832116788321168e-05,
"loss": 1.1078,
"step": 194
},
{
"epoch": 0.49872122762148335,
"grad_norm": 0.8904452819811364,
"learning_rate": 2.8467153284671533e-05,
"loss": 1.1865,
"step": 195
},
{
"epoch": 0.5012787723785166,
"grad_norm": 1.1119131548939938,
"learning_rate": 2.861313868613139e-05,
"loss": 1.1423,
"step": 196
},
{
"epoch": 0.5038363171355499,
"grad_norm": 0.7690481247344314,
"learning_rate": 2.8759124087591243e-05,
"loss": 1.1291,
"step": 197
},
{
"epoch": 0.5063938618925832,
"grad_norm": 0.6068355156021454,
"learning_rate": 2.8905109489051098e-05,
"loss": 1.1615,
"step": 198
},
{
"epoch": 0.5089514066496164,
"grad_norm": 0.6596363446983933,
"learning_rate": 2.905109489051095e-05,
"loss": 1.1256,
"step": 199
},
{
"epoch": 0.5115089514066496,
"grad_norm": 0.9114508644380089,
"learning_rate": 2.9197080291970804e-05,
"loss": 1.1418,
"step": 200
},
{
"epoch": 0.5140664961636828,
"grad_norm": 1.02427164464288,
"learning_rate": 2.9343065693430662e-05,
"loss": 1.124,
"step": 201
},
{
"epoch": 0.5166240409207161,
"grad_norm": 0.6998034282822818,
"learning_rate": 2.9489051094890514e-05,
"loss": 1.1161,
"step": 202
},
{
"epoch": 0.5191815856777494,
"grad_norm": 0.5104663961707757,
"learning_rate": 2.963503649635037e-05,
"loss": 1.1481,
"step": 203
},
{
"epoch": 0.5217391304347826,
"grad_norm": 0.6861277074435671,
"learning_rate": 2.978102189781022e-05,
"loss": 1.1198,
"step": 204
},
{
"epoch": 0.5242966751918159,
"grad_norm": 0.4895837987975406,
"learning_rate": 2.9927007299270075e-05,
"loss": 1.1408,
"step": 205
},
{
"epoch": 0.5268542199488491,
"grad_norm": 0.4750724119518469,
"learning_rate": 3.007299270072993e-05,
"loss": 1.1307,
"step": 206
},
{
"epoch": 0.5294117647058824,
"grad_norm": 0.5592069446689829,
"learning_rate": 3.0218978102189785e-05,
"loss": 1.148,
"step": 207
},
{
"epoch": 0.5319693094629157,
"grad_norm": 0.5740468040244172,
"learning_rate": 3.0364963503649636e-05,
"loss": 1.1325,
"step": 208
},
{
"epoch": 0.5345268542199488,
"grad_norm": 0.609122165091842,
"learning_rate": 3.051094890510949e-05,
"loss": 1.1265,
"step": 209
},
{
"epoch": 0.5370843989769821,
"grad_norm": 0.46915731333688765,
"learning_rate": 3.0656934306569346e-05,
"loss": 1.1158,
"step": 210
},
{
"epoch": 0.5396419437340153,
"grad_norm": 0.5250514439106148,
"learning_rate": 3.08029197080292e-05,
"loss": 1.1378,
"step": 211
},
{
"epoch": 0.5421994884910486,
"grad_norm": 0.5371453416786197,
"learning_rate": 3.0948905109489056e-05,
"loss": 1.1059,
"step": 212
},
{
"epoch": 0.5447570332480819,
"grad_norm": 0.46195810827900435,
"learning_rate": 3.109489051094891e-05,
"loss": 1.1469,
"step": 213
},
{
"epoch": 0.5473145780051151,
"grad_norm": 0.5190881716239564,
"learning_rate": 3.1240875912408765e-05,
"loss": 1.101,
"step": 214
},
{
"epoch": 0.5498721227621484,
"grad_norm": 0.45772493932270886,
"learning_rate": 3.138686131386862e-05,
"loss": 1.1064,
"step": 215
},
{
"epoch": 0.5524296675191815,
"grad_norm": 0.6418401078647901,
"learning_rate": 3.153284671532847e-05,
"loss": 1.1407,
"step": 216
},
{
"epoch": 0.5549872122762148,
"grad_norm": 0.807357254045679,
"learning_rate": 3.167883211678832e-05,
"loss": 1.114,
"step": 217
},
{
"epoch": 0.5575447570332481,
"grad_norm": 1.0774726302289037,
"learning_rate": 3.182481751824818e-05,
"loss": 1.0863,
"step": 218
},
{
"epoch": 0.5601023017902813,
"grad_norm": 0.9160636003608879,
"learning_rate": 3.197080291970803e-05,
"loss": 1.1124,
"step": 219
},
{
"epoch": 0.5626598465473146,
"grad_norm": 0.7863370319064351,
"learning_rate": 3.211678832116789e-05,
"loss": 1.1641,
"step": 220
},
{
"epoch": 0.5652173913043478,
"grad_norm": 0.6092991648146338,
"learning_rate": 3.226277372262774e-05,
"loss": 1.1063,
"step": 221
},
{
"epoch": 0.5677749360613811,
"grad_norm": 0.3355718391632478,
"learning_rate": 3.24087591240876e-05,
"loss": 1.1069,
"step": 222
},
{
"epoch": 0.5703324808184144,
"grad_norm": 0.3670653870960199,
"learning_rate": 3.255474452554745e-05,
"loss": 1.1118,
"step": 223
},
{
"epoch": 0.5728900255754475,
"grad_norm": 0.4368873141277494,
"learning_rate": 3.27007299270073e-05,
"loss": 1.1435,
"step": 224
},
{
"epoch": 0.5754475703324808,
"grad_norm": 0.5849998714298743,
"learning_rate": 3.284671532846716e-05,
"loss": 1.1391,
"step": 225
},
{
"epoch": 0.578005115089514,
"grad_norm": 0.6113062787295098,
"learning_rate": 3.299270072992701e-05,
"loss": 1.1127,
"step": 226
},
{
"epoch": 0.5805626598465473,
"grad_norm": 0.5541334978750299,
"learning_rate": 3.313868613138687e-05,
"loss": 1.125,
"step": 227
},
{
"epoch": 0.5831202046035806,
"grad_norm": 0.5895609310858135,
"learning_rate": 3.328467153284672e-05,
"loss": 1.11,
"step": 228
},
{
"epoch": 0.5856777493606138,
"grad_norm": 0.6247788324610238,
"learning_rate": 3.343065693430657e-05,
"loss": 1.1287,
"step": 229
},
{
"epoch": 0.5882352941176471,
"grad_norm": 0.668302342823751,
"learning_rate": 3.357664233576642e-05,
"loss": 1.1337,
"step": 230
},
{
"epoch": 0.5907928388746803,
"grad_norm": 0.7599150780492421,
"learning_rate": 3.372262773722628e-05,
"loss": 1.131,
"step": 231
},
{
"epoch": 0.5933503836317136,
"grad_norm": 0.9246619867056259,
"learning_rate": 3.386861313868613e-05,
"loss": 1.1193,
"step": 232
},
{
"epoch": 0.5959079283887468,
"grad_norm": 1.1508830666363266,
"learning_rate": 3.401459854014599e-05,
"loss": 1.1256,
"step": 233
},
{
"epoch": 0.59846547314578,
"grad_norm": 0.7850743807177226,
"learning_rate": 3.416058394160584e-05,
"loss": 1.0796,
"step": 234
},
{
"epoch": 0.6010230179028133,
"grad_norm": 0.5549530183464894,
"learning_rate": 3.4306569343065694e-05,
"loss": 1.1006,
"step": 235
},
{
"epoch": 0.6035805626598465,
"grad_norm": 0.647850374242059,
"learning_rate": 3.4452554744525545e-05,
"loss": 1.1289,
"step": 236
},
{
"epoch": 0.6061381074168798,
"grad_norm": 0.8091493107753931,
"learning_rate": 3.4598540145985404e-05,
"loss": 1.1101,
"step": 237
},
{
"epoch": 0.6086956521739131,
"grad_norm": 0.8110463140232488,
"learning_rate": 3.474452554744526e-05,
"loss": 1.0984,
"step": 238
},
{
"epoch": 0.6112531969309463,
"grad_norm": 0.7215947590292753,
"learning_rate": 3.4890510948905113e-05,
"loss": 1.1386,
"step": 239
},
{
"epoch": 0.6138107416879796,
"grad_norm": 0.6215758172373026,
"learning_rate": 3.5036496350364965e-05,
"loss": 1.106,
"step": 240
},
{
"epoch": 0.6163682864450127,
"grad_norm": 0.5875994134397073,
"learning_rate": 3.5182481751824816e-05,
"loss": 1.1025,
"step": 241
},
{
"epoch": 0.618925831202046,
"grad_norm": 0.564702048122142,
"learning_rate": 3.5328467153284675e-05,
"loss": 1.1263,
"step": 242
},
{
"epoch": 0.6214833759590793,
"grad_norm": 0.7446668779549663,
"learning_rate": 3.5474452554744526e-05,
"loss": 1.1236,
"step": 243
},
{
"epoch": 0.6240409207161125,
"grad_norm": 0.6482507318947761,
"learning_rate": 3.5620437956204384e-05,
"loss": 1.1692,
"step": 244
},
{
"epoch": 0.6265984654731458,
"grad_norm": 0.5279644882907087,
"learning_rate": 3.5766423357664236e-05,
"loss": 1.1288,
"step": 245
},
{
"epoch": 0.629156010230179,
"grad_norm": 0.5029060289429896,
"learning_rate": 3.5912408759124094e-05,
"loss": 1.1239,
"step": 246
},
{
"epoch": 0.6317135549872123,
"grad_norm": 0.5036950017578802,
"learning_rate": 3.6058394160583946e-05,
"loss": 1.1148,
"step": 247
},
{
"epoch": 0.6342710997442456,
"grad_norm": 0.625522981617417,
"learning_rate": 3.62043795620438e-05,
"loss": 1.1608,
"step": 248
},
{
"epoch": 0.6368286445012787,
"grad_norm": 0.7678220998472406,
"learning_rate": 3.635036496350365e-05,
"loss": 1.1425,
"step": 249
},
{
"epoch": 0.639386189258312,
"grad_norm": 0.9358561676391077,
"learning_rate": 3.649635036496351e-05,
"loss": 1.1213,
"step": 250
},
{
"epoch": 0.6419437340153452,
"grad_norm": 1.0699268072723376,
"learning_rate": 3.6642335766423365e-05,
"loss": 1.1299,
"step": 251
},
{
"epoch": 0.6445012787723785,
"grad_norm": 0.8217919971512252,
"learning_rate": 3.6788321167883217e-05,
"loss": 1.1333,
"step": 252
},
{
"epoch": 0.6470588235294118,
"grad_norm": 0.7864793534576517,
"learning_rate": 3.693430656934307e-05,
"loss": 1.1149,
"step": 253
},
{
"epoch": 0.649616368286445,
"grad_norm": 0.8677946139868593,
"learning_rate": 3.708029197080292e-05,
"loss": 1.1667,
"step": 254
},
{
"epoch": 0.6521739130434783,
"grad_norm": 0.7672177411758275,
"learning_rate": 3.722627737226278e-05,
"loss": 1.1167,
"step": 255
},
{
"epoch": 0.6547314578005116,
"grad_norm": 0.596809627301341,
"learning_rate": 3.737226277372263e-05,
"loss": 1.1074,
"step": 256
},
{
"epoch": 0.6572890025575447,
"grad_norm": 0.5515836958775989,
"learning_rate": 3.751824817518249e-05,
"loss": 1.1438,
"step": 257
},
{
"epoch": 0.659846547314578,
"grad_norm": 0.5902403172668593,
"learning_rate": 3.766423357664234e-05,
"loss": 1.1055,
"step": 258
},
{
"epoch": 0.6624040920716112,
"grad_norm": 0.5244457227108441,
"learning_rate": 3.781021897810219e-05,
"loss": 1.1379,
"step": 259
},
{
"epoch": 0.6649616368286445,
"grad_norm": 0.5815805254861676,
"learning_rate": 3.795620437956204e-05,
"loss": 1.1165,
"step": 260
},
{
"epoch": 0.6675191815856778,
"grad_norm": 0.8721710722425067,
"learning_rate": 3.81021897810219e-05,
"loss": 1.0881,
"step": 261
},
{
"epoch": 0.670076726342711,
"grad_norm": 1.0174681687817262,
"learning_rate": 3.824817518248176e-05,
"loss": 1.1032,
"step": 262
},
{
"epoch": 0.6726342710997443,
"grad_norm": 0.9222431658366145,
"learning_rate": 3.839416058394161e-05,
"loss": 1.1467,
"step": 263
},
{
"epoch": 0.6751918158567775,
"grad_norm": 0.7265180796009453,
"learning_rate": 3.854014598540147e-05,
"loss": 1.1259,
"step": 264
},
{
"epoch": 0.6777493606138107,
"grad_norm": 0.6510966546449701,
"learning_rate": 3.868613138686132e-05,
"loss": 1.1536,
"step": 265
},
{
"epoch": 0.680306905370844,
"grad_norm": 0.730970983337431,
"learning_rate": 3.883211678832117e-05,
"loss": 1.1209,
"step": 266
},
{
"epoch": 0.6828644501278772,
"grad_norm": 0.6800748870906258,
"learning_rate": 3.897810218978102e-05,
"loss": 1.0751,
"step": 267
},
{
"epoch": 0.6854219948849105,
"grad_norm": 0.7472352428364261,
"learning_rate": 3.912408759124088e-05,
"loss": 1.1167,
"step": 268
},
{
"epoch": 0.6879795396419437,
"grad_norm": 1.0783648866228714,
"learning_rate": 3.927007299270073e-05,
"loss": 1.0978,
"step": 269
},
{
"epoch": 0.690537084398977,
"grad_norm": 0.869855528937522,
"learning_rate": 3.941605839416059e-05,
"loss": 1.138,
"step": 270
},
{
"epoch": 0.6930946291560103,
"grad_norm": 0.7115291823282514,
"learning_rate": 3.956204379562044e-05,
"loss": 1.1344,
"step": 271
},
{
"epoch": 0.6956521739130435,
"grad_norm": 0.5781649216008872,
"learning_rate": 3.9708029197080294e-05,
"loss": 1.0946,
"step": 272
},
{
"epoch": 0.6982097186700768,
"grad_norm": 0.6149999448335685,
"learning_rate": 3.9854014598540145e-05,
"loss": 1.109,
"step": 273
},
{
"epoch": 0.7007672634271099,
"grad_norm": 0.6500237890032072,
"learning_rate": 4e-05,
"loss": 1.1289,
"step": 274
},
{
"epoch": 0.7033248081841432,
"grad_norm": 0.6330629822202228,
"learning_rate": 3.9999983730625253e-05,
"loss": 1.1067,
"step": 275
},
{
"epoch": 0.7058823529411765,
"grad_norm": 0.5619393490473983,
"learning_rate": 3.999993492252747e-05,
"loss": 1.1261,
"step": 276
},
{
"epoch": 0.7084398976982097,
"grad_norm": 0.6013535131894646,
"learning_rate": 3.9999853575786077e-05,
"loss": 1.1391,
"step": 277
},
{
"epoch": 0.710997442455243,
"grad_norm": 0.6020048009259275,
"learning_rate": 3.99997396905334e-05,
"loss": 1.1079,
"step": 278
},
{
"epoch": 0.7135549872122762,
"grad_norm": 0.7808302089293203,
"learning_rate": 3.999959326695472e-05,
"loss": 1.1393,
"step": 279
},
{
"epoch": 0.7161125319693095,
"grad_norm": 0.8356530146885535,
"learning_rate": 3.9999414305288286e-05,
"loss": 1.1042,
"step": 280
},
{
"epoch": 0.7186700767263428,
"grad_norm": 0.8070348321443147,
"learning_rate": 3.999920280582524e-05,
"loss": 1.0983,
"step": 281
},
{
"epoch": 0.7212276214833759,
"grad_norm": 0.8083589330430379,
"learning_rate": 3.9998958768909675e-05,
"loss": 1.126,
"step": 282
},
{
"epoch": 0.7237851662404092,
"grad_norm": 0.8153347627006287,
"learning_rate": 3.999868219493863e-05,
"loss": 1.055,
"step": 283
},
{
"epoch": 0.7263427109974424,
"grad_norm": 0.7023909850774769,
"learning_rate": 3.9998373084362075e-05,
"loss": 1.1187,
"step": 284
},
{
"epoch": 0.7289002557544757,
"grad_norm": 0.6139361690619497,
"learning_rate": 3.999803143768291e-05,
"loss": 1.1003,
"step": 285
},
{
"epoch": 0.731457800511509,
"grad_norm": 0.5948738493196901,
"learning_rate": 3.9997657255456965e-05,
"loss": 1.1236,
"step": 286
},
{
"epoch": 0.7340153452685422,
"grad_norm": 0.6882374035298456,
"learning_rate": 3.999725053829303e-05,
"loss": 1.1214,
"step": 287
},
{
"epoch": 0.7365728900255755,
"grad_norm": 0.8084417687106696,
"learning_rate": 3.9996811286852796e-05,
"loss": 1.0876,
"step": 288
},
{
"epoch": 0.7391304347826086,
"grad_norm": 0.7145649319149529,
"learning_rate": 3.99963395018509e-05,
"loss": 1.1189,
"step": 289
},
{
"epoch": 0.7416879795396419,
"grad_norm": 0.5788883111631697,
"learning_rate": 3.999583518405491e-05,
"loss": 1.1273,
"step": 290
},
{
"epoch": 0.7442455242966752,
"grad_norm": 0.47203390332488526,
"learning_rate": 3.999529833428531e-05,
"loss": 1.1117,
"step": 291
},
{
"epoch": 0.7468030690537084,
"grad_norm": 0.470532299489505,
"learning_rate": 3.999472895341553e-05,
"loss": 1.0768,
"step": 292
},
{
"epoch": 0.7493606138107417,
"grad_norm": 0.41407305757120705,
"learning_rate": 3.999412704237191e-05,
"loss": 1.1026,
"step": 293
},
{
"epoch": 0.7519181585677749,
"grad_norm": 0.46064032712582936,
"learning_rate": 3.9993492602133734e-05,
"loss": 1.0995,
"step": 294
},
{
"epoch": 0.7544757033248082,
"grad_norm": 0.5953117769099772,
"learning_rate": 3.999282563373318e-05,
"loss": 1.0905,
"step": 295
},
{
"epoch": 0.7570332480818415,
"grad_norm": 0.5639476868801947,
"learning_rate": 3.999212613825538e-05,
"loss": 1.0981,
"step": 296
},
{
"epoch": 0.7595907928388747,
"grad_norm": 0.47485520206266274,
"learning_rate": 3.999139411683836e-05,
"loss": 1.1376,
"step": 297
},
{
"epoch": 0.7621483375959079,
"grad_norm": 0.5417739162294339,
"learning_rate": 3.999062957067307e-05,
"loss": 1.1124,
"step": 298
},
{
"epoch": 0.7647058823529411,
"grad_norm": 0.6449384118110638,
"learning_rate": 3.998983250100338e-05,
"loss": 1.1284,
"step": 299
},
{
"epoch": 0.7672634271099744,
"grad_norm": 0.6676759068939218,
"learning_rate": 3.998900290912609e-05,
"loss": 1.0862,
"step": 300
},
{
"epoch": 0.7698209718670077,
"grad_norm": 0.6149128164393245,
"learning_rate": 3.9988140796390874e-05,
"loss": 1.1186,
"step": 301
},
{
"epoch": 0.7723785166240409,
"grad_norm": 0.5460371943479854,
"learning_rate": 3.998724616420034e-05,
"loss": 1.1183,
"step": 302
},
{
"epoch": 0.7749360613810742,
"grad_norm": 0.4642794113976264,
"learning_rate": 3.998631901401e-05,
"loss": 1.088,
"step": 303
},
{
"epoch": 0.7774936061381074,
"grad_norm": 0.5147582705198133,
"learning_rate": 3.998535934732827e-05,
"loss": 1.0933,
"step": 304
},
{
"epoch": 0.7800511508951407,
"grad_norm": 0.45988313713885,
"learning_rate": 3.9984367165716466e-05,
"loss": 1.1068,
"step": 305
},
{
"epoch": 0.782608695652174,
"grad_norm": 0.4437393811230941,
"learning_rate": 3.998334247078881e-05,
"loss": 1.1008,
"step": 306
},
{
"epoch": 0.7851662404092071,
"grad_norm": 0.3864667008156923,
"learning_rate": 3.998228526421241e-05,
"loss": 1.0668,
"step": 307
},
{
"epoch": 0.7877237851662404,
"grad_norm": 0.4067701945780569,
"learning_rate": 3.998119554770728e-05,
"loss": 1.0978,
"step": 308
},
{
"epoch": 0.7902813299232737,
"grad_norm": 0.4682003347143726,
"learning_rate": 3.9980073323046324e-05,
"loss": 1.1158,
"step": 309
},
{
"epoch": 0.7928388746803069,
"grad_norm": 0.5875404213822213,
"learning_rate": 3.997891859205532e-05,
"loss": 1.0839,
"step": 310
},
{
"epoch": 0.7953964194373402,
"grad_norm": 0.6236930337300663,
"learning_rate": 3.997773135661295e-05,
"loss": 1.1172,
"step": 311
},
{
"epoch": 0.7979539641943734,
"grad_norm": 0.7006486821550716,
"learning_rate": 3.997651161865078e-05,
"loss": 1.1302,
"step": 312
},
{
"epoch": 0.8005115089514067,
"grad_norm": 0.7609408124475948,
"learning_rate": 3.997525938015323e-05,
"loss": 1.0961,
"step": 313
},
{
"epoch": 0.80306905370844,
"grad_norm": 0.8008929510771413,
"learning_rate": 3.997397464315762e-05,
"loss": 1.1036,
"step": 314
},
{
"epoch": 0.8056265984654731,
"grad_norm": 0.865997392073584,
"learning_rate": 3.997265740975414e-05,
"loss": 1.0919,
"step": 315
},
{
"epoch": 0.8081841432225064,
"grad_norm": 0.9113738533485526,
"learning_rate": 3.997130768208585e-05,
"loss": 1.0899,
"step": 316
},
{
"epoch": 0.8107416879795396,
"grad_norm": 0.7530828577709697,
"learning_rate": 3.996992546234867e-05,
"loss": 1.1185,
"step": 317
},
{
"epoch": 0.8132992327365729,
"grad_norm": 0.5442852001885987,
"learning_rate": 3.996851075279138e-05,
"loss": 1.1084,
"step": 318
},
{
"epoch": 0.8158567774936062,
"grad_norm": 0.5345433204091397,
"learning_rate": 3.9967063555715634e-05,
"loss": 1.0699,
"step": 319
},
{
"epoch": 0.8184143222506394,
"grad_norm": 0.44557439869349674,
"learning_rate": 3.996558387347592e-05,
"loss": 1.0958,
"step": 320
},
{
"epoch": 0.8209718670076727,
"grad_norm": 0.44168314049542273,
"learning_rate": 3.996407170847959e-05,
"loss": 1.0763,
"step": 321
},
{
"epoch": 0.8235294117647058,
"grad_norm": 0.5354488125190239,
"learning_rate": 3.9962527063186845e-05,
"loss": 1.1032,
"step": 322
},
{
"epoch": 0.8260869565217391,
"grad_norm": 0.47506571108298457,
"learning_rate": 3.9960949940110734e-05,
"loss": 1.1235,
"step": 323
},
{
"epoch": 0.8286445012787724,
"grad_norm": 0.5186848412653702,
"learning_rate": 3.995934034181712e-05,
"loss": 1.1262,
"step": 324
},
{
"epoch": 0.8312020460358056,
"grad_norm": 0.42258629948636006,
"learning_rate": 3.9957698270924735e-05,
"loss": 1.0859,
"step": 325
},
{
"epoch": 0.8337595907928389,
"grad_norm": 0.32189689287468726,
"learning_rate": 3.995602373010511e-05,
"loss": 1.0937,
"step": 326
},
{
"epoch": 0.8363171355498721,
"grad_norm": 0.41525101801363073,
"learning_rate": 3.995431672208263e-05,
"loss": 1.1347,
"step": 327
},
{
"epoch": 0.8388746803069054,
"grad_norm": 0.5490574519978781,
"learning_rate": 3.99525772496345e-05,
"loss": 1.1073,
"step": 328
},
{
"epoch": 0.8414322250639387,
"grad_norm": 0.548723945317829,
"learning_rate": 3.995080531559071e-05,
"loss": 1.1465,
"step": 329
},
{
"epoch": 0.8439897698209718,
"grad_norm": 0.5105171842872774,
"learning_rate": 3.99490009228341e-05,
"loss": 1.1159,
"step": 330
},
{
"epoch": 0.8465473145780051,
"grad_norm": 0.5294530849314222,
"learning_rate": 3.99471640743003e-05,
"loss": 1.0895,
"step": 331
},
{
"epoch": 0.8491048593350383,
"grad_norm": 0.5691233488307068,
"learning_rate": 3.9945294772977747e-05,
"loss": 1.0813,
"step": 332
},
{
"epoch": 0.8516624040920716,
"grad_norm": 0.5160186742163387,
"learning_rate": 3.994339302190768e-05,
"loss": 1.1273,
"step": 333
},
{
"epoch": 0.8542199488491049,
"grad_norm": 0.5015301623074467,
"learning_rate": 3.994145882418413e-05,
"loss": 1.061,
"step": 334
},
{
"epoch": 0.8567774936061381,
"grad_norm": 0.5928619879817332,
"learning_rate": 3.993949218295391e-05,
"loss": 1.087,
"step": 335
},
{
"epoch": 0.8593350383631714,
"grad_norm": 0.5709725737358415,
"learning_rate": 3.9937493101416626e-05,
"loss": 1.0876,
"step": 336
},
{
"epoch": 0.8618925831202046,
"grad_norm": 0.6809732337488936,
"learning_rate": 3.993546158282466e-05,
"loss": 1.1086,
"step": 337
},
{
"epoch": 0.8644501278772379,
"grad_norm": 0.8551479096280965,
"learning_rate": 3.9933397630483165e-05,
"loss": 1.1151,
"step": 338
},
{
"epoch": 0.8670076726342711,
"grad_norm": 0.9652613252359079,
"learning_rate": 3.993130124775007e-05,
"loss": 1.0848,
"step": 339
},
{
"epoch": 0.8695652173913043,
"grad_norm": 1.155001072143155,
"learning_rate": 3.9929172438036035e-05,
"loss": 1.1238,
"step": 340
},
{
"epoch": 0.8721227621483376,
"grad_norm": 0.8067771786160456,
"learning_rate": 3.9927011204804526e-05,
"loss": 1.0784,
"step": 341
},
{
"epoch": 0.8746803069053708,
"grad_norm": 0.5357290897886735,
"learning_rate": 3.992481755157172e-05,
"loss": 1.1043,
"step": 342
},
{
"epoch": 0.8772378516624041,
"grad_norm": 0.4995847325007271,
"learning_rate": 3.992259148190656e-05,
"loss": 1.0989,
"step": 343
},
{
"epoch": 0.8797953964194374,
"grad_norm": 0.5959103672076177,
"learning_rate": 3.992033299943072e-05,
"loss": 1.1054,
"step": 344
},
{
"epoch": 0.8823529411764706,
"grad_norm": 0.6048861847403096,
"learning_rate": 3.991804210781861e-05,
"loss": 1.0895,
"step": 345
},
{
"epoch": 0.8849104859335039,
"grad_norm": 0.4988308431334708,
"learning_rate": 3.991571881079736e-05,
"loss": 1.1547,
"step": 346
},
{
"epoch": 0.887468030690537,
"grad_norm": 0.45970030620455815,
"learning_rate": 3.991336311214684e-05,
"loss": 1.0403,
"step": 347
},
{
"epoch": 0.8900255754475703,
"grad_norm": 0.49108796202887645,
"learning_rate": 3.991097501569963e-05,
"loss": 1.1019,
"step": 348
},
{
"epoch": 0.8925831202046036,
"grad_norm": 0.4519220772301018,
"learning_rate": 3.9908554525340995e-05,
"loss": 1.0814,
"step": 349
},
{
"epoch": 0.8951406649616368,
"grad_norm": 0.5721610282920665,
"learning_rate": 3.990610164500893e-05,
"loss": 1.1053,
"step": 350
},
{
"epoch": 0.8976982097186701,
"grad_norm": 0.5043216306258643,
"learning_rate": 3.990361637869412e-05,
"loss": 1.125,
"step": 351
},
{
"epoch": 0.9002557544757033,
"grad_norm": 0.33096250065683464,
"learning_rate": 3.990109873043994e-05,
"loss": 1.0797,
"step": 352
},
{
"epoch": 0.9028132992327366,
"grad_norm": 0.4118129543237625,
"learning_rate": 3.9898548704342437e-05,
"loss": 1.1007,
"step": 353
},
{
"epoch": 0.9053708439897699,
"grad_norm": 0.4405673266338336,
"learning_rate": 3.989596630455035e-05,
"loss": 1.0946,
"step": 354
},
{
"epoch": 0.907928388746803,
"grad_norm": 0.3544788596605096,
"learning_rate": 3.989335153526509e-05,
"loss": 1.0707,
"step": 355
},
{
"epoch": 0.9104859335038363,
"grad_norm": 0.45917996019960844,
"learning_rate": 3.9890704400740705e-05,
"loss": 1.0864,
"step": 356
},
{
"epoch": 0.9130434782608695,
"grad_norm": 0.45483794654383436,
"learning_rate": 3.988802490528393e-05,
"loss": 1.0984,
"step": 357
},
{
"epoch": 0.9156010230179028,
"grad_norm": 0.3718894204151617,
"learning_rate": 3.988531305325414e-05,
"loss": 1.1195,
"step": 358
},
{
"epoch": 0.9181585677749361,
"grad_norm": 0.34370568886087605,
"learning_rate": 3.9882568849063335e-05,
"loss": 1.1026,
"step": 359
},
{
"epoch": 0.9207161125319693,
"grad_norm": 0.36530582646575005,
"learning_rate": 3.987979229717617e-05,
"loss": 1.0672,
"step": 360
},
{
"epoch": 0.9232736572890026,
"grad_norm": 0.40309630305871075,
"learning_rate": 3.987698340210993e-05,
"loss": 1.1181,
"step": 361
},
{
"epoch": 0.9258312020460358,
"grad_norm": 0.4798685208427719,
"learning_rate": 3.987414216843451e-05,
"loss": 1.0414,
"step": 362
},
{
"epoch": 0.928388746803069,
"grad_norm": 0.5805121526300497,
"learning_rate": 3.98712686007724e-05,
"loss": 1.1068,
"step": 363
},
{
"epoch": 0.9309462915601023,
"grad_norm": 0.5198525814368523,
"learning_rate": 3.986836270379873e-05,
"loss": 1.0875,
"step": 364
},
{
"epoch": 0.9335038363171355,
"grad_norm": 0.4851619608742611,
"learning_rate": 3.986542448224121e-05,
"loss": 1.1071,
"step": 365
},
{
"epoch": 0.9360613810741688,
"grad_norm": 0.5175641724417654,
"learning_rate": 3.986245394088015e-05,
"loss": 1.1272,
"step": 366
},
{
"epoch": 0.9386189258312021,
"grad_norm": 0.531743371564595,
"learning_rate": 3.985945108454843e-05,
"loss": 1.1075,
"step": 367
},
{
"epoch": 0.9411764705882353,
"grad_norm": 0.5628299016470453,
"learning_rate": 3.98564159181315e-05,
"loss": 1.0985,
"step": 368
},
{
"epoch": 0.9437340153452686,
"grad_norm": 0.559691193164672,
"learning_rate": 3.98533484465674e-05,
"loss": 1.0924,
"step": 369
},
{
"epoch": 0.9462915601023018,
"grad_norm": 0.4603859918059814,
"learning_rate": 3.985024867484671e-05,
"loss": 1.0934,
"step": 370
},
{
"epoch": 0.948849104859335,
"grad_norm": 0.372253690862611,
"learning_rate": 3.9847116608012566e-05,
"loss": 1.1174,
"step": 371
},
{
"epoch": 0.9514066496163683,
"grad_norm": 0.39740382328437535,
"learning_rate": 3.984395225116064e-05,
"loss": 1.0973,
"step": 372
},
{
"epoch": 0.9539641943734015,
"grad_norm": 0.47816155505672137,
"learning_rate": 3.9840755609439145e-05,
"loss": 1.102,
"step": 373
},
{
"epoch": 0.9565217391304348,
"grad_norm": 0.488226938431436,
"learning_rate": 3.983752668804882e-05,
"loss": 1.0996,
"step": 374
},
{
"epoch": 0.959079283887468,
"grad_norm": 0.44913015107902343,
"learning_rate": 3.983426549224291e-05,
"loss": 1.1235,
"step": 375
},
{
"epoch": 0.9616368286445013,
"grad_norm": 0.4368874489723291,
"learning_rate": 3.983097202732719e-05,
"loss": 1.0745,
"step": 376
},
{
"epoch": 0.9641943734015346,
"grad_norm": 0.27691989976946607,
"learning_rate": 3.9827646298659904e-05,
"loss": 1.1046,
"step": 377
},
{
"epoch": 0.9667519181585678,
"grad_norm": 0.4259438151890329,
"learning_rate": 3.9824288311651825e-05,
"loss": 1.1107,
"step": 378
},
{
"epoch": 0.969309462915601,
"grad_norm": 0.4280199970904566,
"learning_rate": 3.982089807176617e-05,
"loss": 1.0801,
"step": 379
},
{
"epoch": 0.9718670076726342,
"grad_norm": 0.3841190320034587,
"learning_rate": 3.981747558451866e-05,
"loss": 1.0627,
"step": 380
},
{
"epoch": 0.9744245524296675,
"grad_norm": 0.47168053742941896,
"learning_rate": 3.981402085547746e-05,
"loss": 1.0912,
"step": 381
},
{
"epoch": 0.9769820971867008,
"grad_norm": 0.42170320584928817,
"learning_rate": 3.98105338902632e-05,
"loss": 1.097,
"step": 382
},
{
"epoch": 0.979539641943734,
"grad_norm": 0.4411555079409664,
"learning_rate": 3.9807014694548956e-05,
"loss": 1.1043,
"step": 383
},
{
"epoch": 0.9820971867007673,
"grad_norm": 0.37031687281577386,
"learning_rate": 3.9803463274060246e-05,
"loss": 1.0549,
"step": 384
},
{
"epoch": 0.9846547314578005,
"grad_norm": 0.32290380202286484,
"learning_rate": 3.979987963457499e-05,
"loss": 1.1365,
"step": 385
},
{
"epoch": 0.9872122762148338,
"grad_norm": 0.36230420930125995,
"learning_rate": 3.979626378192356e-05,
"loss": 1.0906,
"step": 386
},
{
"epoch": 0.989769820971867,
"grad_norm": 0.5648030723766292,
"learning_rate": 3.979261572198872e-05,
"loss": 1.104,
"step": 387
},
{
"epoch": 0.9923273657289002,
"grad_norm": 0.9078923326237621,
"learning_rate": 3.9788935460705635e-05,
"loss": 1.1182,
"step": 388
},
{
"epoch": 0.9948849104859335,
"grad_norm": 1.145057226173462,
"learning_rate": 3.9785223004061865e-05,
"loss": 1.0997,
"step": 389
},
{
"epoch": 0.9974424552429667,
"grad_norm": 0.7375732788047394,
"learning_rate": 3.978147835809733e-05,
"loss": 1.0825,
"step": 390
},
{
"epoch": 1.0,
"grad_norm": 0.46229445624378956,
"learning_rate": 3.977770152890435e-05,
"loss": 1.1115,
"step": 391
},
{
"epoch": 1.0025575447570332,
"grad_norm": 0.3818083283988484,
"learning_rate": 3.977389252262758e-05,
"loss": 1.0794,
"step": 392
},
{
"epoch": 1.0051150895140666,
"grad_norm": 0.3519209214464206,
"learning_rate": 3.977005134546405e-05,
"loss": 1.1013,
"step": 393
},
{
"epoch": 1.0076726342710998,
"grad_norm": 0.4368995926170067,
"learning_rate": 3.976617800366309e-05,
"loss": 1.0684,
"step": 394
},
{
"epoch": 1.010230179028133,
"grad_norm": 0.3900513578773351,
"learning_rate": 3.976227250352641e-05,
"loss": 1.0783,
"step": 395
},
{
"epoch": 1.0127877237851663,
"grad_norm": 0.38397819359948265,
"learning_rate": 3.9758334851407996e-05,
"loss": 1.0943,
"step": 396
},
{
"epoch": 1.0153452685421995,
"grad_norm": 0.4694313786910681,
"learning_rate": 3.975436505371417e-05,
"loss": 1.0958,
"step": 397
},
{
"epoch": 1.0179028132992327,
"grad_norm": 0.4380778755144381,
"learning_rate": 3.975036311690355e-05,
"loss": 1.0893,
"step": 398
},
{
"epoch": 1.020460358056266,
"grad_norm": 0.5165931599471566,
"learning_rate": 3.9746329047487034e-05,
"loss": 1.0594,
"step": 399
},
{
"epoch": 1.0230179028132993,
"grad_norm": 0.6600002793241831,
"learning_rate": 3.974226285202779e-05,
"loss": 1.1146,
"step": 400
},
{
"epoch": 1.0255754475703325,
"grad_norm": 0.6959171541898408,
"learning_rate": 3.9738164537141275e-05,
"loss": 1.0926,
"step": 401
},
{
"epoch": 1.0281329923273657,
"grad_norm": 0.7458040360013451,
"learning_rate": 3.9734034109495186e-05,
"loss": 1.1104,
"step": 402
},
{
"epoch": 1.030690537084399,
"grad_norm": 0.6861149758243119,
"learning_rate": 3.972987157580947e-05,
"loss": 1.0787,
"step": 403
},
{
"epoch": 1.0332480818414322,
"grad_norm": 0.5830714165709471,
"learning_rate": 3.972567694285632e-05,
"loss": 1.0816,
"step": 404
},
{
"epoch": 1.0358056265984654,
"grad_norm": 0.5271256219676713,
"learning_rate": 3.972145021746013e-05,
"loss": 1.0794,
"step": 405
},
{
"epoch": 1.0383631713554988,
"grad_norm": 0.5236958260896567,
"learning_rate": 3.971719140649751e-05,
"loss": 1.1137,
"step": 406
},
{
"epoch": 1.040920716112532,
"grad_norm": 0.5494101072555336,
"learning_rate": 3.97129005168973e-05,
"loss": 1.0481,
"step": 407
},
{
"epoch": 1.0434782608695652,
"grad_norm": 0.48929235700153795,
"learning_rate": 3.970857755564049e-05,
"loss": 1.0809,
"step": 408
},
{
"epoch": 1.0460358056265984,
"grad_norm": 0.36062713966221727,
"learning_rate": 3.970422252976029e-05,
"loss": 1.0573,
"step": 409
},
{
"epoch": 1.0485933503836318,
"grad_norm": 0.4058259876089094,
"learning_rate": 3.969983544634204e-05,
"loss": 1.0519,
"step": 410
},
{
"epoch": 1.051150895140665,
"grad_norm": 0.45284593193542,
"learning_rate": 3.9695416312523245e-05,
"loss": 1.0816,
"step": 411
},
{
"epoch": 1.0537084398976981,
"grad_norm": 0.43133392446394014,
"learning_rate": 3.9690965135493566e-05,
"loss": 1.0606,
"step": 412
},
{
"epoch": 1.0562659846547315,
"grad_norm": 0.45855353915390107,
"learning_rate": 3.9686481922494796e-05,
"loss": 1.0668,
"step": 413
},
{
"epoch": 1.0588235294117647,
"grad_norm": 0.44058365293490726,
"learning_rate": 3.968196668082083e-05,
"loss": 1.1167,
"step": 414
},
{
"epoch": 1.061381074168798,
"grad_norm": 0.4095755635190781,
"learning_rate": 3.9677419417817696e-05,
"loss": 1.0871,
"step": 415
},
{
"epoch": 1.0639386189258313,
"grad_norm": 0.43229315221585174,
"learning_rate": 3.96728401408835e-05,
"loss": 1.0659,
"step": 416
},
{
"epoch": 1.0664961636828645,
"grad_norm": 0.4220671808393113,
"learning_rate": 3.966822885746844e-05,
"loss": 1.0517,
"step": 417
},
{
"epoch": 1.0690537084398977,
"grad_norm": 0.48787808116716475,
"learning_rate": 3.966358557507478e-05,
"loss": 1.0637,
"step": 418
},
{
"epoch": 1.0716112531969308,
"grad_norm": 0.4869202553068813,
"learning_rate": 3.9658910301256864e-05,
"loss": 1.119,
"step": 419
},
{
"epoch": 1.0741687979539642,
"grad_norm": 0.477015439756663,
"learning_rate": 3.965420304362106e-05,
"loss": 1.0642,
"step": 420
},
{
"epoch": 1.0767263427109974,
"grad_norm": 0.5535930482823976,
"learning_rate": 3.964946380982578e-05,
"loss": 1.0797,
"step": 421
},
{
"epoch": 1.0792838874680306,
"grad_norm": 0.4716314569881298,
"learning_rate": 3.9644692607581464e-05,
"loss": 1.0724,
"step": 422
},
{
"epoch": 1.081841432225064,
"grad_norm": 0.4247435864037584,
"learning_rate": 3.9639889444650565e-05,
"loss": 1.0571,
"step": 423
},
{
"epoch": 1.0843989769820972,
"grad_norm": 0.41341878524947645,
"learning_rate": 3.963505432884752e-05,
"loss": 1.1009,
"step": 424
},
{
"epoch": 1.0869565217391304,
"grad_norm": 0.5547432895078889,
"learning_rate": 3.963018726803877e-05,
"loss": 1.0653,
"step": 425
},
{
"epoch": 1.0895140664961638,
"grad_norm": 0.6284375727571099,
"learning_rate": 3.9625288270142716e-05,
"loss": 1.0559,
"step": 426
},
{
"epoch": 1.092071611253197,
"grad_norm": 0.7131923849117704,
"learning_rate": 3.9620357343129706e-05,
"loss": 1.0826,
"step": 427
},
{
"epoch": 1.0946291560102301,
"grad_norm": 0.8228384514293502,
"learning_rate": 3.961539449502208e-05,
"loss": 1.0616,
"step": 428
},
{
"epoch": 1.0971867007672633,
"grad_norm": 0.9112887765645529,
"learning_rate": 3.961039973389405e-05,
"loss": 1.0915,
"step": 429
},
{
"epoch": 1.0997442455242967,
"grad_norm": 0.8995359209600274,
"learning_rate": 3.9605373067871806e-05,
"loss": 1.0689,
"step": 430
},
{
"epoch": 1.10230179028133,
"grad_norm": 0.7956471679247282,
"learning_rate": 3.9600314505133396e-05,
"loss": 1.0965,
"step": 431
},
{
"epoch": 1.104859335038363,
"grad_norm": 0.6524277555765188,
"learning_rate": 3.95952240539088e-05,
"loss": 1.1127,
"step": 432
},
{
"epoch": 1.1074168797953965,
"grad_norm": 0.576815762606571,
"learning_rate": 3.9590101722479864e-05,
"loss": 1.0923,
"step": 433
},
{
"epoch": 1.1099744245524297,
"grad_norm": 0.5179815700708057,
"learning_rate": 3.9584947519180296e-05,
"loss": 1.0817,
"step": 434
},
{
"epoch": 1.1125319693094629,
"grad_norm": 0.3669159640839577,
"learning_rate": 3.957976145239566e-05,
"loss": 1.071,
"step": 435
},
{
"epoch": 1.1150895140664963,
"grad_norm": 0.45175070196548184,
"learning_rate": 3.957454353056337e-05,
"loss": 1.0718,
"step": 436
},
{
"epoch": 1.1176470588235294,
"grad_norm": 0.42416776332872974,
"learning_rate": 3.9569293762172655e-05,
"loss": 1.0508,
"step": 437
},
{
"epoch": 1.1202046035805626,
"grad_norm": 0.40075568544449747,
"learning_rate": 3.9564012155764564e-05,
"loss": 1.063,
"step": 438
},
{
"epoch": 1.1227621483375958,
"grad_norm": 0.3301120148907268,
"learning_rate": 3.955869871993193e-05,
"loss": 1.0835,
"step": 439
},
{
"epoch": 1.1253196930946292,
"grad_norm": 0.31268221799407214,
"learning_rate": 3.955335346331939e-05,
"loss": 1.061,
"step": 440
},
{
"epoch": 1.1278772378516624,
"grad_norm": 0.421075328402457,
"learning_rate": 3.954797639462334e-05,
"loss": 1.0825,
"step": 441
},
{
"epoch": 1.1304347826086956,
"grad_norm": 0.44153527881913607,
"learning_rate": 3.954256752259193e-05,
"loss": 1.0619,
"step": 442
},
{
"epoch": 1.132992327365729,
"grad_norm": 0.48670244568231075,
"learning_rate": 3.953712685602507e-05,
"loss": 1.0865,
"step": 443
},
{
"epoch": 1.1355498721227621,
"grad_norm": 0.47736961048678445,
"learning_rate": 3.9531654403774366e-05,
"loss": 1.0542,
"step": 444
},
{
"epoch": 1.1381074168797953,
"grad_norm": 0.5007106635395288,
"learning_rate": 3.9526150174743166e-05,
"loss": 1.0763,
"step": 445
},
{
"epoch": 1.1406649616368287,
"grad_norm": 0.5171914985622378,
"learning_rate": 3.952061417788651e-05,
"loss": 1.0491,
"step": 446
},
{
"epoch": 1.143222506393862,
"grad_norm": 0.5295256590690459,
"learning_rate": 3.951504642221111e-05,
"loss": 1.0342,
"step": 447
},
{
"epoch": 1.145780051150895,
"grad_norm": 0.5407803467243536,
"learning_rate": 3.950944691677536e-05,
"loss": 1.0688,
"step": 448
},
{
"epoch": 1.1483375959079285,
"grad_norm": 0.6004919807649173,
"learning_rate": 3.950381567068931e-05,
"loss": 1.059,
"step": 449
},
{
"epoch": 1.1508951406649617,
"grad_norm": 0.5794538850051559,
"learning_rate": 3.9498152693114636e-05,
"loss": 1.1069,
"step": 450
},
{
"epoch": 1.1534526854219949,
"grad_norm": 0.5520342558428422,
"learning_rate": 3.949245799326465e-05,
"loss": 1.081,
"step": 451
},
{
"epoch": 1.156010230179028,
"grad_norm": 0.47999763589931316,
"learning_rate": 3.948673158040428e-05,
"loss": 1.0846,
"step": 452
},
{
"epoch": 1.1585677749360614,
"grad_norm": 0.5263956766545858,
"learning_rate": 3.9480973463850044e-05,
"loss": 1.0998,
"step": 453
},
{
"epoch": 1.1611253196930946,
"grad_norm": 0.5702817068691663,
"learning_rate": 3.9475183652970026e-05,
"loss": 1.0736,
"step": 454
},
{
"epoch": 1.1636828644501278,
"grad_norm": 0.3911434484207975,
"learning_rate": 3.9469362157183885e-05,
"loss": 1.0683,
"step": 455
},
{
"epoch": 1.1662404092071612,
"grad_norm": 0.4627795900857231,
"learning_rate": 3.9463508985962845e-05,
"loss": 1.0587,
"step": 456
},
{
"epoch": 1.1687979539641944,
"grad_norm": 0.4953160170615254,
"learning_rate": 3.9457624148829644e-05,
"loss": 1.0505,
"step": 457
},
{
"epoch": 1.1713554987212276,
"grad_norm": 0.5871232495665668,
"learning_rate": 3.945170765535853e-05,
"loss": 1.056,
"step": 458
},
{
"epoch": 1.1739130434782608,
"grad_norm": 0.5886126285412883,
"learning_rate": 3.9445759515175293e-05,
"loss": 1.0635,
"step": 459
},
{
"epoch": 1.1764705882352942,
"grad_norm": 0.5775521341870387,
"learning_rate": 3.9439779737957164e-05,
"loss": 1.0912,
"step": 460
},
{
"epoch": 1.1790281329923273,
"grad_norm": 0.6739517351837386,
"learning_rate": 3.943376833343288e-05,
"loss": 1.1136,
"step": 461
},
{
"epoch": 1.1815856777493605,
"grad_norm": 0.6752416771525206,
"learning_rate": 3.9427725311382603e-05,
"loss": 1.0575,
"step": 462
},
{
"epoch": 1.184143222506394,
"grad_norm": 0.5166111559005302,
"learning_rate": 3.942165068163798e-05,
"loss": 1.0496,
"step": 463
},
{
"epoch": 1.186700767263427,
"grad_norm": 0.4142639594837752,
"learning_rate": 3.941554445408202e-05,
"loss": 1.0733,
"step": 464
},
{
"epoch": 1.1892583120204603,
"grad_norm": 0.38169129443334754,
"learning_rate": 3.94094066386492e-05,
"loss": 1.0815,
"step": 465
},
{
"epoch": 1.1918158567774937,
"grad_norm": 0.3148962237527228,
"learning_rate": 3.940323724532536e-05,
"loss": 1.0823,
"step": 466
},
{
"epoch": 1.1943734015345269,
"grad_norm": 0.4350278655655421,
"learning_rate": 3.93970362841477e-05,
"loss": 1.058,
"step": 467
},
{
"epoch": 1.19693094629156,
"grad_norm": 0.570286594568403,
"learning_rate": 3.939080376520481e-05,
"loss": 1.1071,
"step": 468
},
{
"epoch": 1.1994884910485935,
"grad_norm": 0.5952820030721013,
"learning_rate": 3.9384539698636606e-05,
"loss": 1.0713,
"step": 469
},
{
"epoch": 1.2020460358056266,
"grad_norm": 0.5155080007829775,
"learning_rate": 3.937824409463433e-05,
"loss": 1.0927,
"step": 470
},
{
"epoch": 1.2046035805626598,
"grad_norm": 0.5511523293446119,
"learning_rate": 3.937191696344054e-05,
"loss": 1.0674,
"step": 471
},
{
"epoch": 1.207161125319693,
"grad_norm": 0.39744259853198133,
"learning_rate": 3.936555831534908e-05,
"loss": 1.0232,
"step": 472
},
{
"epoch": 1.2097186700767264,
"grad_norm": 0.39758229665361045,
"learning_rate": 3.9359168160705074e-05,
"loss": 1.0815,
"step": 473
},
{
"epoch": 1.2122762148337596,
"grad_norm": 0.34091042245371367,
"learning_rate": 3.93527465099049e-05,
"loss": 1.063,
"step": 474
},
{
"epoch": 1.2148337595907928,
"grad_norm": 0.35727280233292935,
"learning_rate": 3.9346293373396186e-05,
"loss": 1.0774,
"step": 475
},
{
"epoch": 1.2173913043478262,
"grad_norm": 0.42130715147814757,
"learning_rate": 3.9339808761677784e-05,
"loss": 1.0509,
"step": 476
},
{
"epoch": 1.2199488491048593,
"grad_norm": 0.4303854025874077,
"learning_rate": 3.933329268529974e-05,
"loss": 1.1019,
"step": 477
},
{
"epoch": 1.2225063938618925,
"grad_norm": 0.4227074766735461,
"learning_rate": 3.9326745154863326e-05,
"loss": 1.0866,
"step": 478
},
{
"epoch": 1.2250639386189257,
"grad_norm": 0.4543377415106031,
"learning_rate": 3.932016618102094e-05,
"loss": 1.0943,
"step": 479
},
{
"epoch": 1.227621483375959,
"grad_norm": 0.46441298894176936,
"learning_rate": 3.931355577447618e-05,
"loss": 1.0478,
"step": 480
},
{
"epoch": 1.2301790281329923,
"grad_norm": 0.35858169740835527,
"learning_rate": 3.930691394598375e-05,
"loss": 1.0552,
"step": 481
},
{
"epoch": 1.2327365728900257,
"grad_norm": 0.47326465081866653,
"learning_rate": 3.93002407063495e-05,
"loss": 1.074,
"step": 482
},
{
"epoch": 1.2352941176470589,
"grad_norm": 0.5275528761395574,
"learning_rate": 3.9293536066430366e-05,
"loss": 1.0658,
"step": 483
},
{
"epoch": 1.237851662404092,
"grad_norm": 0.5309767853756476,
"learning_rate": 3.928680003713439e-05,
"loss": 1.08,
"step": 484
},
{
"epoch": 1.2404092071611252,
"grad_norm": 0.4731807150894732,
"learning_rate": 3.9280032629420646e-05,
"loss": 1.086,
"step": 485
},
{
"epoch": 1.2429667519181586,
"grad_norm": 0.36291913005320176,
"learning_rate": 3.927323385429931e-05,
"loss": 1.0802,
"step": 486
},
{
"epoch": 1.2455242966751918,
"grad_norm": 0.3222852773031256,
"learning_rate": 3.9266403722831546e-05,
"loss": 1.0512,
"step": 487
},
{
"epoch": 1.248081841432225,
"grad_norm": 0.3234452956355875,
"learning_rate": 3.925954224612957e-05,
"loss": 1.0718,
"step": 488
},
{
"epoch": 1.2506393861892584,
"grad_norm": 0.3374439444139576,
"learning_rate": 3.925264943535655e-05,
"loss": 1.076,
"step": 489
},
{
"epoch": 1.2531969309462916,
"grad_norm": 0.3370539480110268,
"learning_rate": 3.924572530172668e-05,
"loss": 1.098,
"step": 490
},
{
"epoch": 1.2557544757033248,
"grad_norm": 0.4079348237297101,
"learning_rate": 3.923876985650508e-05,
"loss": 1.0685,
"step": 491
},
{
"epoch": 1.258312020460358,
"grad_norm": 0.48060128644514083,
"learning_rate": 3.923178311100783e-05,
"loss": 1.0942,
"step": 492
},
{
"epoch": 1.2608695652173914,
"grad_norm": 0.4953515156495457,
"learning_rate": 3.922476507660192e-05,
"loss": 1.0613,
"step": 493
},
{
"epoch": 1.2634271099744245,
"grad_norm": 0.5072252040490288,
"learning_rate": 3.921771576470527e-05,
"loss": 1.06,
"step": 494
},
{
"epoch": 1.265984654731458,
"grad_norm": 0.5120075860462384,
"learning_rate": 3.921063518678665e-05,
"loss": 1.0859,
"step": 495
},
{
"epoch": 1.2685421994884911,
"grad_norm": 0.5326911420263161,
"learning_rate": 3.9203523354365735e-05,
"loss": 1.0759,
"step": 496
},
{
"epoch": 1.2710997442455243,
"grad_norm": 0.41538511976408143,
"learning_rate": 3.919638027901302e-05,
"loss": 1.0668,
"step": 497
},
{
"epoch": 1.2736572890025575,
"grad_norm": 0.349238115710704,
"learning_rate": 3.918920597234985e-05,
"loss": 1.0564,
"step": 498
},
{
"epoch": 1.2762148337595907,
"grad_norm": 0.4185470229151538,
"learning_rate": 3.918200044604836e-05,
"loss": 1.0695,
"step": 499
},
{
"epoch": 1.278772378516624,
"grad_norm": 0.3385657068555429,
"learning_rate": 3.917476371183151e-05,
"loss": 1.0428,
"step": 500
},
{
"epoch": 1.2813299232736572,
"grad_norm": 0.31315327777452157,
"learning_rate": 3.9167495781473e-05,
"loss": 1.063,
"step": 501
},
{
"epoch": 1.2838874680306906,
"grad_norm": 0.3637592396215665,
"learning_rate": 3.91601966667973e-05,
"loss": 1.0691,
"step": 502
},
{
"epoch": 1.2864450127877238,
"grad_norm": 0.4037518802439983,
"learning_rate": 3.9152866379679614e-05,
"loss": 1.0866,
"step": 503
},
{
"epoch": 1.289002557544757,
"grad_norm": 0.4952112760927536,
"learning_rate": 3.914550493204586e-05,
"loss": 1.0612,
"step": 504
},
{
"epoch": 1.2915601023017902,
"grad_norm": 0.4785480977948807,
"learning_rate": 3.9138112335872663e-05,
"loss": 1.0828,
"step": 505
},
{
"epoch": 1.2941176470588236,
"grad_norm": 0.3926501789148642,
"learning_rate": 3.91306886031873e-05,
"loss": 1.0506,
"step": 506
},
{
"epoch": 1.2966751918158568,
"grad_norm": 0.4764425416390801,
"learning_rate": 3.912323374606773e-05,
"loss": 1.0855,
"step": 507
},
{
"epoch": 1.29923273657289,
"grad_norm": 0.6127677296964769,
"learning_rate": 3.911574777664254e-05,
"loss": 1.0483,
"step": 508
},
{
"epoch": 1.3017902813299234,
"grad_norm": 0.5035795375824405,
"learning_rate": 3.910823070709093e-05,
"loss": 1.0793,
"step": 509
},
{
"epoch": 1.3043478260869565,
"grad_norm": 0.4237858936852908,
"learning_rate": 3.9100682549642696e-05,
"loss": 1.0505,
"step": 510
},
{
"epoch": 1.3069053708439897,
"grad_norm": 0.37726853544904126,
"learning_rate": 3.909310331657823e-05,
"loss": 1.0574,
"step": 511
},
{
"epoch": 1.309462915601023,
"grad_norm": 0.33448809112540856,
"learning_rate": 3.908549302022846e-05,
"loss": 1.0913,
"step": 512
},
{
"epoch": 1.3120204603580563,
"grad_norm": 0.42579464916868914,
"learning_rate": 3.907785167297487e-05,
"loss": 1.0454,
"step": 513
},
{
"epoch": 1.3145780051150895,
"grad_norm": 0.3189151725630645,
"learning_rate": 3.907017928724946e-05,
"loss": 1.0661,
"step": 514
},
{
"epoch": 1.317135549872123,
"grad_norm": 0.43081732949042756,
"learning_rate": 3.9062475875534704e-05,
"loss": 1.0586,
"step": 515
},
{
"epoch": 1.319693094629156,
"grad_norm": 0.5425689904084623,
"learning_rate": 3.9054741450363586e-05,
"loss": 1.0547,
"step": 516
},
{
"epoch": 1.3222506393861893,
"grad_norm": 0.5527616448071511,
"learning_rate": 3.9046976024319524e-05,
"loss": 1.0497,
"step": 517
},
{
"epoch": 1.3248081841432224,
"grad_norm": 0.7020026841600744,
"learning_rate": 3.9039179610036374e-05,
"loss": 1.0674,
"step": 518
},
{
"epoch": 1.3273657289002558,
"grad_norm": 0.9523359783242552,
"learning_rate": 3.903135222019844e-05,
"loss": 1.0592,
"step": 519
},
{
"epoch": 1.329923273657289,
"grad_norm": 1.1983224650872797,
"learning_rate": 3.902349386754037e-05,
"loss": 1.0943,
"step": 520
},
{
"epoch": 1.3324808184143222,
"grad_norm": 0.5814615374526603,
"learning_rate": 3.9015604564847225e-05,
"loss": 1.0985,
"step": 521
},
{
"epoch": 1.3350383631713556,
"grad_norm": 0.42335954340111454,
"learning_rate": 3.900768432495441e-05,
"loss": 1.0786,
"step": 522
},
{
"epoch": 1.3375959079283888,
"grad_norm": 0.5702652210322368,
"learning_rate": 3.899973316074765e-05,
"loss": 1.0664,
"step": 523
},
{
"epoch": 1.340153452685422,
"grad_norm": 0.69026690823692,
"learning_rate": 3.8991751085163e-05,
"loss": 1.0944,
"step": 524
},
{
"epoch": 1.3427109974424551,
"grad_norm": 0.6823704528932245,
"learning_rate": 3.8983738111186784e-05,
"loss": 1.0658,
"step": 525
},
{
"epoch": 1.3452685421994885,
"grad_norm": 0.5857579808694497,
"learning_rate": 3.897569425185563e-05,
"loss": 1.0909,
"step": 526
},
{
"epoch": 1.3478260869565217,
"grad_norm": 0.45817099746254636,
"learning_rate": 3.896761952025638e-05,
"loss": 1.0411,
"step": 527
},
{
"epoch": 1.350383631713555,
"grad_norm": 0.43457162622492124,
"learning_rate": 3.8959513929526124e-05,
"loss": 1.1047,
"step": 528
},
{
"epoch": 1.3529411764705883,
"grad_norm": 0.4462558953821414,
"learning_rate": 3.895137749285215e-05,
"loss": 1.0651,
"step": 529
},
{
"epoch": 1.3554987212276215,
"grad_norm": 0.44466600769670217,
"learning_rate": 3.894321022347193e-05,
"loss": 1.0379,
"step": 530
},
{
"epoch": 1.3580562659846547,
"grad_norm": 0.3994567077314908,
"learning_rate": 3.893501213467311e-05,
"loss": 1.0734,
"step": 531
},
{
"epoch": 1.3606138107416879,
"grad_norm": 0.48941022994253963,
"learning_rate": 3.892678323979344e-05,
"loss": 1.0611,
"step": 532
},
{
"epoch": 1.3631713554987213,
"grad_norm": 0.43521790345589173,
"learning_rate": 3.891852355222085e-05,
"loss": 1.0991,
"step": 533
},
{
"epoch": 1.3657289002557544,
"grad_norm": 0.40678733530773103,
"learning_rate": 3.891023308539332e-05,
"loss": 1.0641,
"step": 534
},
{
"epoch": 1.3682864450127878,
"grad_norm": 0.4458758332186816,
"learning_rate": 3.890191185279892e-05,
"loss": 1.077,
"step": 535
},
{
"epoch": 1.370843989769821,
"grad_norm": 0.39950353909121245,
"learning_rate": 3.889355986797578e-05,
"loss": 1.1091,
"step": 536
},
{
"epoch": 1.3734015345268542,
"grad_norm": 0.41383737937408677,
"learning_rate": 3.888517714451206e-05,
"loss": 1.0649,
"step": 537
},
{
"epoch": 1.3759590792838874,
"grad_norm": 0.4203594606661656,
"learning_rate": 3.887676369604592e-05,
"loss": 1.0807,
"step": 538
},
{
"epoch": 1.3785166240409208,
"grad_norm": 0.4264859136650187,
"learning_rate": 3.8868319536265524e-05,
"loss": 1.0616,
"step": 539
},
{
"epoch": 1.381074168797954,
"grad_norm": 0.5010355720474308,
"learning_rate": 3.885984467890897e-05,
"loss": 1.073,
"step": 540
},
{
"epoch": 1.3836317135549872,
"grad_norm": 0.5462831330016021,
"learning_rate": 3.885133913776435e-05,
"loss": 1.0976,
"step": 541
},
{
"epoch": 1.3861892583120206,
"grad_norm": 0.5355278476013258,
"learning_rate": 3.884280292666962e-05,
"loss": 1.0858,
"step": 542
},
{
"epoch": 1.3887468030690537,
"grad_norm": 0.5051546421054438,
"learning_rate": 3.8834236059512684e-05,
"loss": 1.0442,
"step": 543
},
{
"epoch": 1.391304347826087,
"grad_norm": 0.4984221188937596,
"learning_rate": 3.8825638550231295e-05,
"loss": 1.0648,
"step": 544
},
{
"epoch": 1.39386189258312,
"grad_norm": 0.439116100414634,
"learning_rate": 3.881701041281305e-05,
"loss": 1.056,
"step": 545
},
{
"epoch": 1.3964194373401535,
"grad_norm": 0.4289743304279261,
"learning_rate": 3.8808351661295405e-05,
"loss": 1.0338,
"step": 546
},
{
"epoch": 1.3989769820971867,
"grad_norm": 0.4925906191844476,
"learning_rate": 3.87996623097656e-05,
"loss": 1.0731,
"step": 547
},
{
"epoch": 1.40153452685422,
"grad_norm": 0.42640490900227823,
"learning_rate": 3.879094237236066e-05,
"loss": 1.0675,
"step": 548
},
{
"epoch": 1.4040920716112533,
"grad_norm": 0.3505524354279509,
"learning_rate": 3.878219186326739e-05,
"loss": 1.0875,
"step": 549
},
{
"epoch": 1.4066496163682864,
"grad_norm": 0.4409832771467105,
"learning_rate": 3.877341079672231e-05,
"loss": 1.0258,
"step": 550
},
{
"epoch": 1.4092071611253196,
"grad_norm": 0.5108110971750124,
"learning_rate": 3.876459918701168e-05,
"loss": 1.0406,
"step": 551
},
{
"epoch": 1.4117647058823528,
"grad_norm": 0.48534856181582575,
"learning_rate": 3.875575704847142e-05,
"loss": 1.0862,
"step": 552
},
{
"epoch": 1.4143222506393862,
"grad_norm": 0.3234496742473989,
"learning_rate": 3.874688439548716e-05,
"loss": 1.069,
"step": 553
},
{
"epoch": 1.4168797953964194,
"grad_norm": 0.4261954118796616,
"learning_rate": 3.8737981242494124e-05,
"loss": 1.0803,
"step": 554
},
{
"epoch": 1.4194373401534528,
"grad_norm": 0.4746861957350986,
"learning_rate": 3.872904760397721e-05,
"loss": 1.0931,
"step": 555
},
{
"epoch": 1.421994884910486,
"grad_norm": 0.3832076298527735,
"learning_rate": 3.8720083494470874e-05,
"loss": 1.0741,
"step": 556
},
{
"epoch": 1.4245524296675192,
"grad_norm": 0.314397374193223,
"learning_rate": 3.8711088928559174e-05,
"loss": 1.0646,
"step": 557
},
{
"epoch": 1.4271099744245523,
"grad_norm": 0.48312590478689404,
"learning_rate": 3.8702063920875696e-05,
"loss": 1.0533,
"step": 558
},
{
"epoch": 1.4296675191815857,
"grad_norm": 0.4585966574259363,
"learning_rate": 3.869300848610356e-05,
"loss": 1.0516,
"step": 559
},
{
"epoch": 1.432225063938619,
"grad_norm": 0.42125111947509647,
"learning_rate": 3.868392263897541e-05,
"loss": 1.0385,
"step": 560
},
{
"epoch": 1.434782608695652,
"grad_norm": 0.4796650860678778,
"learning_rate": 3.8674806394273326e-05,
"loss": 1.0494,
"step": 561
},
{
"epoch": 1.4373401534526855,
"grad_norm": 0.5598045203907789,
"learning_rate": 3.866565976682889e-05,
"loss": 1.0664,
"step": 562
},
{
"epoch": 1.4398976982097187,
"grad_norm": 0.6544001428623203,
"learning_rate": 3.865648277152308e-05,
"loss": 1.0924,
"step": 563
},
{
"epoch": 1.4424552429667519,
"grad_norm": 0.8181707296322468,
"learning_rate": 3.8647275423286294e-05,
"loss": 1.0892,
"step": 564
},
{
"epoch": 1.445012787723785,
"grad_norm": 0.8806393358437705,
"learning_rate": 3.863803773709833e-05,
"loss": 1.0976,
"step": 565
},
{
"epoch": 1.4475703324808185,
"grad_norm": 0.7477270100938752,
"learning_rate": 3.86287697279883e-05,
"loss": 1.0602,
"step": 566
},
{
"epoch": 1.4501278772378516,
"grad_norm": 0.3569962075433497,
"learning_rate": 3.86194714110347e-05,
"loss": 1.0387,
"step": 567
},
{
"epoch": 1.452685421994885,
"grad_norm": 0.3543347546663264,
"learning_rate": 3.861014280136529e-05,
"loss": 1.0696,
"step": 568
},
{
"epoch": 1.4552429667519182,
"grad_norm": 0.4970664250979028,
"learning_rate": 3.860078391415713e-05,
"loss": 1.0609,
"step": 569
},
{
"epoch": 1.4578005115089514,
"grad_norm": 0.5782757311963443,
"learning_rate": 3.859139476463657e-05,
"loss": 1.0525,
"step": 570
},
{
"epoch": 1.4603580562659846,
"grad_norm": 0.5434048966033144,
"learning_rate": 3.858197536807915e-05,
"loss": 1.0739,
"step": 571
},
{
"epoch": 1.4629156010230178,
"grad_norm": 0.4601770813341688,
"learning_rate": 3.8572525739809654e-05,
"loss": 1.0758,
"step": 572
},
{
"epoch": 1.4654731457800512,
"grad_norm": 0.38722533469779286,
"learning_rate": 3.856304589520202e-05,
"loss": 1.0606,
"step": 573
},
{
"epoch": 1.4680306905370843,
"grad_norm": 0.49325063140199615,
"learning_rate": 3.8553535849679375e-05,
"loss": 1.0482,
"step": 574
},
{
"epoch": 1.4705882352941178,
"grad_norm": 0.4950426201265235,
"learning_rate": 3.854399561871396e-05,
"loss": 1.0816,
"step": 575
},
{
"epoch": 1.473145780051151,
"grad_norm": 0.543604743326878,
"learning_rate": 3.853442521782714e-05,
"loss": 1.0861,
"step": 576
},
{
"epoch": 1.4757033248081841,
"grad_norm": 0.5749592990637336,
"learning_rate": 3.8524824662589354e-05,
"loss": 1.0768,
"step": 577
},
{
"epoch": 1.4782608695652173,
"grad_norm": 0.49070761223130277,
"learning_rate": 3.851519396862011e-05,
"loss": 1.0692,
"step": 578
},
{
"epoch": 1.4808184143222507,
"grad_norm": 0.3943903108976757,
"learning_rate": 3.850553315158794e-05,
"loss": 1.0455,
"step": 579
},
{
"epoch": 1.4833759590792839,
"grad_norm": 0.4238248938295029,
"learning_rate": 3.8495842227210394e-05,
"loss": 1.0596,
"step": 580
},
{
"epoch": 1.485933503836317,
"grad_norm": 0.4291305895330894,
"learning_rate": 3.848612121125399e-05,
"loss": 1.0776,
"step": 581
},
{
"epoch": 1.4884910485933505,
"grad_norm": 0.4548690848762231,
"learning_rate": 3.847637011953423e-05,
"loss": 1.0691,
"step": 582
},
{
"epoch": 1.4910485933503836,
"grad_norm": 0.40843513739737003,
"learning_rate": 3.8466588967915515e-05,
"loss": 1.0919,
"step": 583
},
{
"epoch": 1.4936061381074168,
"grad_norm": 0.41803311955742734,
"learning_rate": 3.845677777231118e-05,
"loss": 1.1006,
"step": 584
},
{
"epoch": 1.49616368286445,
"grad_norm": 0.3719560495309354,
"learning_rate": 3.844693654868342e-05,
"loss": 1.0376,
"step": 585
},
{
"epoch": 1.4987212276214834,
"grad_norm": 0.4385376962732821,
"learning_rate": 3.843706531304328e-05,
"loss": 1.0528,
"step": 586
},
{
"epoch": 1.5012787723785166,
"grad_norm": 0.4067229405241637,
"learning_rate": 3.8427164081450666e-05,
"loss": 1.0606,
"step": 587
},
{
"epoch": 1.50383631713555,
"grad_norm": 0.5149031594008476,
"learning_rate": 3.841723287001425e-05,
"loss": 1.0897,
"step": 588
},
{
"epoch": 1.5063938618925832,
"grad_norm": 0.509311529817368,
"learning_rate": 3.8407271694891486e-05,
"loss": 1.0282,
"step": 589
},
{
"epoch": 1.5089514066496164,
"grad_norm": 0.48395383940774744,
"learning_rate": 3.839728057228859e-05,
"loss": 1.0564,
"step": 590
},
{
"epoch": 1.5115089514066495,
"grad_norm": 0.5149614538006095,
"learning_rate": 3.8387259518460504e-05,
"loss": 1.0313,
"step": 591
},
{
"epoch": 1.5140664961636827,
"grad_norm": 0.49838365767776116,
"learning_rate": 3.8377208549710835e-05,
"loss": 1.0868,
"step": 592
},
{
"epoch": 1.5166240409207161,
"grad_norm": 0.46244674081808324,
"learning_rate": 3.8367127682391895e-05,
"loss": 1.0145,
"step": 593
},
{
"epoch": 1.5191815856777495,
"grad_norm": 0.47731408176603424,
"learning_rate": 3.835701693290462e-05,
"loss": 1.0794,
"step": 594
},
{
"epoch": 1.5217391304347827,
"grad_norm": 0.5306996012817398,
"learning_rate": 3.834687631769858e-05,
"loss": 1.0506,
"step": 595
},
{
"epoch": 1.5242966751918159,
"grad_norm": 0.44457030960228416,
"learning_rate": 3.8336705853271896e-05,
"loss": 1.0631,
"step": 596
},
{
"epoch": 1.526854219948849,
"grad_norm": 0.3655113187023026,
"learning_rate": 3.832650555617129e-05,
"loss": 1.0712,
"step": 597
},
{
"epoch": 1.5294117647058822,
"grad_norm": 0.3712045140565929,
"learning_rate": 3.831627544299203e-05,
"loss": 1.0728,
"step": 598
},
{
"epoch": 1.5319693094629157,
"grad_norm": 0.40839038158163093,
"learning_rate": 3.8306015530377835e-05,
"loss": 1.0485,
"step": 599
},
{
"epoch": 1.5345268542199488,
"grad_norm": 0.4720669374696074,
"learning_rate": 3.829572583502096e-05,
"loss": 1.096,
"step": 600
},
{
"epoch": 1.5370843989769822,
"grad_norm": 0.5378463886339307,
"learning_rate": 3.828540637366209e-05,
"loss": 1.0499,
"step": 601
},
{
"epoch": 1.5396419437340154,
"grad_norm": 0.49538223098962153,
"learning_rate": 3.8275057163090344e-05,
"loss": 1.0452,
"step": 602
},
{
"epoch": 1.5421994884910486,
"grad_norm": 0.391483632406279,
"learning_rate": 3.8264678220143246e-05,
"loss": 1.0521,
"step": 603
},
{
"epoch": 1.5447570332480818,
"grad_norm": 0.3799757213480213,
"learning_rate": 3.8254269561706686e-05,
"loss": 1.0674,
"step": 604
},
{
"epoch": 1.547314578005115,
"grad_norm": 0.307050244020001,
"learning_rate": 3.824383120471489e-05,
"loss": 1.0531,
"step": 605
},
{
"epoch": 1.5498721227621484,
"grad_norm": 0.37142808309004094,
"learning_rate": 3.823336316615043e-05,
"loss": 1.0314,
"step": 606
},
{
"epoch": 1.5524296675191815,
"grad_norm": 0.5689453303808312,
"learning_rate": 3.8222865463044134e-05,
"loss": 1.0559,
"step": 607
},
{
"epoch": 1.554987212276215,
"grad_norm": 0.588920041491494,
"learning_rate": 3.821233811247512e-05,
"loss": 1.0773,
"step": 608
},
{
"epoch": 1.5575447570332481,
"grad_norm": 0.3941133829851967,
"learning_rate": 3.820178113157072e-05,
"loss": 1.0725,
"step": 609
},
{
"epoch": 1.5601023017902813,
"grad_norm": 0.3075596153991981,
"learning_rate": 3.8191194537506486e-05,
"loss": 1.0419,
"step": 610
},
{
"epoch": 1.5626598465473145,
"grad_norm": 0.47740829759275283,
"learning_rate": 3.8180578347506154e-05,
"loss": 1.0797,
"step": 611
},
{
"epoch": 1.5652173913043477,
"grad_norm": 0.4188396139473123,
"learning_rate": 3.816993257884158e-05,
"loss": 1.0685,
"step": 612
},
{
"epoch": 1.567774936061381,
"grad_norm": 0.2983620381432042,
"learning_rate": 3.815925724883278e-05,
"loss": 1.0567,
"step": 613
},
{
"epoch": 1.5703324808184145,
"grad_norm": 0.3846448207755022,
"learning_rate": 3.814855237484785e-05,
"loss": 1.0612,
"step": 614
},
{
"epoch": 1.5728900255754477,
"grad_norm": 0.3383910452858533,
"learning_rate": 3.813781797430294e-05,
"loss": 1.0829,
"step": 615
},
{
"epoch": 1.5754475703324808,
"grad_norm": 0.30585062273219626,
"learning_rate": 3.8127054064662255e-05,
"loss": 1.0372,
"step": 616
},
{
"epoch": 1.578005115089514,
"grad_norm": 0.3049880861371505,
"learning_rate": 3.8116260663438004e-05,
"loss": 1.0687,
"step": 617
},
{
"epoch": 1.5805626598465472,
"grad_norm": 0.31926543631621834,
"learning_rate": 3.8105437788190365e-05,
"loss": 1.061,
"step": 618
},
{
"epoch": 1.5831202046035806,
"grad_norm": 0.36285173950071564,
"learning_rate": 3.80945854565275e-05,
"loss": 1.0484,
"step": 619
},
{
"epoch": 1.5856777493606138,
"grad_norm": 0.436991227470107,
"learning_rate": 3.808370368610545e-05,
"loss": 1.034,
"step": 620
},
{
"epoch": 1.5882352941176472,
"grad_norm": 0.4707604093264852,
"learning_rate": 3.80727924946282e-05,
"loss": 1.067,
"step": 621
},
{
"epoch": 1.5907928388746804,
"grad_norm": 0.5629021561165096,
"learning_rate": 3.8061851899847554e-05,
"loss": 1.0777,
"step": 622
},
{
"epoch": 1.5933503836317136,
"grad_norm": 0.6393024589396724,
"learning_rate": 3.8050881919563186e-05,
"loss": 1.0511,
"step": 623
},
{
"epoch": 1.5959079283887467,
"grad_norm": 0.6433656555557394,
"learning_rate": 3.803988257162257e-05,
"loss": 1.0932,
"step": 624
},
{
"epoch": 1.59846547314578,
"grad_norm": 0.5833653013901674,
"learning_rate": 3.8028853873920954e-05,
"loss": 1.1167,
"step": 625
},
{
"epoch": 1.6010230179028133,
"grad_norm": 0.42039415674055763,
"learning_rate": 3.801779584440134e-05,
"loss": 1.0434,
"step": 626
},
{
"epoch": 1.6035805626598465,
"grad_norm": 0.3641526698392794,
"learning_rate": 3.800670850105445e-05,
"loss": 1.0519,
"step": 627
},
{
"epoch": 1.60613810741688,
"grad_norm": 0.3512419247266656,
"learning_rate": 3.79955918619187e-05,
"loss": 1.0876,
"step": 628
},
{
"epoch": 1.608695652173913,
"grad_norm": 0.3217191243315485,
"learning_rate": 3.798444594508017e-05,
"loss": 1.0419,
"step": 629
},
{
"epoch": 1.6112531969309463,
"grad_norm": 0.3245546370592062,
"learning_rate": 3.797327076867255e-05,
"loss": 1.0469,
"step": 630
},
{
"epoch": 1.6138107416879794,
"grad_norm": 0.48968806097265694,
"learning_rate": 3.7962066350877185e-05,
"loss": 1.0579,
"step": 631
},
{
"epoch": 1.6163682864450126,
"grad_norm": 0.5012555491541114,
"learning_rate": 3.795083270992294e-05,
"loss": 1.0528,
"step": 632
},
{
"epoch": 1.618925831202046,
"grad_norm": 0.5049460458037872,
"learning_rate": 3.793956986408626e-05,
"loss": 1.0327,
"step": 633
},
{
"epoch": 1.6214833759590794,
"grad_norm": 0.5454953965603144,
"learning_rate": 3.792827783169108e-05,
"loss": 1.0828,
"step": 634
},
{
"epoch": 1.6240409207161126,
"grad_norm": 0.41787536882414533,
"learning_rate": 3.791695663110883e-05,
"loss": 1.0342,
"step": 635
},
{
"epoch": 1.6265984654731458,
"grad_norm": 0.33423710341206375,
"learning_rate": 3.7905606280758396e-05,
"loss": 1.0832,
"step": 636
},
{
"epoch": 1.629156010230179,
"grad_norm": 0.3586135617262744,
"learning_rate": 3.78942267991061e-05,
"loss": 1.0747,
"step": 637
},
{
"epoch": 1.6317135549872122,
"grad_norm": 0.5226129887884744,
"learning_rate": 3.7882818204665635e-05,
"loss": 1.0587,
"step": 638
},
{
"epoch": 1.6342710997442456,
"grad_norm": 0.5955111313035846,
"learning_rate": 3.787138051599808e-05,
"loss": 1.0952,
"step": 639
},
{
"epoch": 1.6368286445012787,
"grad_norm": 0.5146503231598974,
"learning_rate": 3.7859913751711826e-05,
"loss": 1.0225,
"step": 640
},
{
"epoch": 1.6393861892583121,
"grad_norm": 0.4438470183700803,
"learning_rate": 3.78484179304626e-05,
"loss": 1.0463,
"step": 641
},
{
"epoch": 1.6419437340153453,
"grad_norm": 0.36541076123788924,
"learning_rate": 3.7836893070953366e-05,
"loss": 1.0497,
"step": 642
},
{
"epoch": 1.6445012787723785,
"grad_norm": 0.36819784176979126,
"learning_rate": 3.7825339191934364e-05,
"loss": 1.063,
"step": 643
},
{
"epoch": 1.6470588235294117,
"grad_norm": 0.4631448829560534,
"learning_rate": 3.781375631220302e-05,
"loss": 1.111,
"step": 644
},
{
"epoch": 1.6496163682864449,
"grad_norm": 0.5470400176800343,
"learning_rate": 3.7802144450603966e-05,
"loss": 1.0207,
"step": 645
},
{
"epoch": 1.6521739130434783,
"grad_norm": 0.6593324931081654,
"learning_rate": 3.779050362602897e-05,
"loss": 1.0769,
"step": 646
},
{
"epoch": 1.6547314578005117,
"grad_norm": 0.643056893494062,
"learning_rate": 3.777883385741693e-05,
"loss": 1.0364,
"step": 647
},
{
"epoch": 1.6572890025575449,
"grad_norm": 0.592773621179805,
"learning_rate": 3.7767135163753826e-05,
"loss": 1.0534,
"step": 648
},
{
"epoch": 1.659846547314578,
"grad_norm": 0.5422011969887595,
"learning_rate": 3.77554075640727e-05,
"loss": 1.0608,
"step": 649
},
{
"epoch": 1.6624040920716112,
"grad_norm": 0.35581962988515814,
"learning_rate": 3.774365107745363e-05,
"loss": 1.04,
"step": 650
},
{
"epoch": 1.6649616368286444,
"grad_norm": 0.35876113735996046,
"learning_rate": 3.773186572302367e-05,
"loss": 1.0665,
"step": 651
},
{
"epoch": 1.6675191815856778,
"grad_norm": 0.4195858599076735,
"learning_rate": 3.772005151995687e-05,
"loss": 1.0918,
"step": 652
},
{
"epoch": 1.670076726342711,
"grad_norm": 0.4851635116544184,
"learning_rate": 3.770820848747419e-05,
"loss": 1.0658,
"step": 653
},
{
"epoch": 1.6726342710997444,
"grad_norm": 0.5124007702917756,
"learning_rate": 3.769633664484351e-05,
"loss": 1.0703,
"step": 654
},
{
"epoch": 1.6751918158567776,
"grad_norm": 0.4667764210539717,
"learning_rate": 3.768443601137958e-05,
"loss": 1.0597,
"step": 655
},
{
"epoch": 1.6777493606138107,
"grad_norm": 0.42277177604865634,
"learning_rate": 3.767250660644397e-05,
"loss": 1.0461,
"step": 656
},
{
"epoch": 1.680306905370844,
"grad_norm": 0.36572119132682473,
"learning_rate": 3.766054844944509e-05,
"loss": 1.0478,
"step": 657
},
{
"epoch": 1.682864450127877,
"grad_norm": 0.3433213155201022,
"learning_rate": 3.764856155983811e-05,
"loss": 1.0714,
"step": 658
},
{
"epoch": 1.6854219948849105,
"grad_norm": 0.4152059805283714,
"learning_rate": 3.7636545957124944e-05,
"loss": 1.0732,
"step": 659
},
{
"epoch": 1.6879795396419437,
"grad_norm": 0.3758935920161636,
"learning_rate": 3.762450166085423e-05,
"loss": 1.0853,
"step": 660
},
{
"epoch": 1.690537084398977,
"grad_norm": 0.3880734497498912,
"learning_rate": 3.7612428690621296e-05,
"loss": 1.0668,
"step": 661
},
{
"epoch": 1.6930946291560103,
"grad_norm": 0.5056256120889562,
"learning_rate": 3.7600327066068096e-05,
"loss": 1.0572,
"step": 662
},
{
"epoch": 1.6956521739130435,
"grad_norm": 0.6074691686964009,
"learning_rate": 3.758819680688322e-05,
"loss": 1.0814,
"step": 663
},
{
"epoch": 1.6982097186700766,
"grad_norm": 0.7014228537086638,
"learning_rate": 3.7576037932801845e-05,
"loss": 1.0651,
"step": 664
},
{
"epoch": 1.7007672634271098,
"grad_norm": 0.6340797508686761,
"learning_rate": 3.75638504636057e-05,
"loss": 1.0691,
"step": 665
},
{
"epoch": 1.7033248081841432,
"grad_norm": 0.5045295207809493,
"learning_rate": 3.755163441912302e-05,
"loss": 1.0446,
"step": 666
},
{
"epoch": 1.7058823529411766,
"grad_norm": 0.40872647546205754,
"learning_rate": 3.753938981922856e-05,
"loss": 1.0586,
"step": 667
},
{
"epoch": 1.7084398976982098,
"grad_norm": 0.3016088995758064,
"learning_rate": 3.7527116683843524e-05,
"loss": 1.0511,
"step": 668
},
{
"epoch": 1.710997442455243,
"grad_norm": 0.4347317124746144,
"learning_rate": 3.7514815032935524e-05,
"loss": 1.0802,
"step": 669
},
{
"epoch": 1.7135549872122762,
"grad_norm": 0.5643117418167675,
"learning_rate": 3.750248488651857e-05,
"loss": 1.0558,
"step": 670
},
{
"epoch": 1.7161125319693094,
"grad_norm": 0.6777770246270608,
"learning_rate": 3.749012626465306e-05,
"loss": 1.0762,
"step": 671
},
{
"epoch": 1.7186700767263428,
"grad_norm": 0.6310745302523975,
"learning_rate": 3.747773918744568e-05,
"loss": 1.0734,
"step": 672
},
{
"epoch": 1.721227621483376,
"grad_norm": 0.46736979089705405,
"learning_rate": 3.7465323675049445e-05,
"loss": 1.051,
"step": 673
},
{
"epoch": 1.7237851662404093,
"grad_norm": 0.3500543240474948,
"learning_rate": 3.745287974766361e-05,
"loss": 1.0931,
"step": 674
},
{
"epoch": 1.7263427109974425,
"grad_norm": 0.505848854691756,
"learning_rate": 3.7440407425533666e-05,
"loss": 1.0642,
"step": 675
},
{
"epoch": 1.7289002557544757,
"grad_norm": 0.5636818680408255,
"learning_rate": 3.742790672895131e-05,
"loss": 1.0345,
"step": 676
},
{
"epoch": 1.7314578005115089,
"grad_norm": 0.5928460911654702,
"learning_rate": 3.741537767825437e-05,
"loss": 1.0504,
"step": 677
},
{
"epoch": 1.734015345268542,
"grad_norm": 0.5419470387318002,
"learning_rate": 3.7402820293826854e-05,
"loss": 1.0522,
"step": 678
},
{
"epoch": 1.7365728900255755,
"grad_norm": 0.4835417652807545,
"learning_rate": 3.739023459609884e-05,
"loss": 1.0523,
"step": 679
},
{
"epoch": 1.7391304347826086,
"grad_norm": 0.34601055558106963,
"learning_rate": 3.737762060554645e-05,
"loss": 1.0502,
"step": 680
},
{
"epoch": 1.741687979539642,
"grad_norm": 0.3990929385440922,
"learning_rate": 3.7364978342691884e-05,
"loss": 1.0416,
"step": 681
},
{
"epoch": 1.7442455242966752,
"grad_norm": 0.328247977498985,
"learning_rate": 3.7352307828103296e-05,
"loss": 1.0838,
"step": 682
},
{
"epoch": 1.7468030690537084,
"grad_norm": 0.38315845983995184,
"learning_rate": 3.733960908239482e-05,
"loss": 1.0802,
"step": 683
},
{
"epoch": 1.7493606138107416,
"grad_norm": 0.35637419921390134,
"learning_rate": 3.732688212622653e-05,
"loss": 1.0637,
"step": 684
},
{
"epoch": 1.7519181585677748,
"grad_norm": 0.2885616385723044,
"learning_rate": 3.7314126980304386e-05,
"loss": 1.058,
"step": 685
},
{
"epoch": 1.7544757033248082,
"grad_norm": 0.42891679221341744,
"learning_rate": 3.730134366538021e-05,
"loss": 1.0849,
"step": 686
},
{
"epoch": 1.7570332480818416,
"grad_norm": 0.49183310353900955,
"learning_rate": 3.728853220225166e-05,
"loss": 1.0587,
"step": 687
},
{
"epoch": 1.7595907928388748,
"grad_norm": 0.4243062837677566,
"learning_rate": 3.7275692611762185e-05,
"loss": 1.0496,
"step": 688
},
{
"epoch": 1.762148337595908,
"grad_norm": 0.378430373380162,
"learning_rate": 3.726282491480099e-05,
"loss": 1.0307,
"step": 689
},
{
"epoch": 1.7647058823529411,
"grad_norm": 0.37191234979894905,
"learning_rate": 3.724992913230302e-05,
"loss": 1.0596,
"step": 690
},
{
"epoch": 1.7672634271099743,
"grad_norm": 0.30363989857175677,
"learning_rate": 3.72370052852489e-05,
"loss": 1.0704,
"step": 691
},
{
"epoch": 1.7698209718670077,
"grad_norm": 0.3126467220971844,
"learning_rate": 3.722405339466493e-05,
"loss": 1.0612,
"step": 692
},
{
"epoch": 1.772378516624041,
"grad_norm": 0.2990769906959218,
"learning_rate": 3.721107348162302e-05,
"loss": 1.0739,
"step": 693
},
{
"epoch": 1.7749360613810743,
"grad_norm": 0.27834277371107397,
"learning_rate": 3.719806556724068e-05,
"loss": 1.0294,
"step": 694
},
{
"epoch": 1.7774936061381075,
"grad_norm": 0.3024194176783559,
"learning_rate": 3.718502967268097e-05,
"loss": 1.0427,
"step": 695
},
{
"epoch": 1.7800511508951407,
"grad_norm": 0.3935315617491812,
"learning_rate": 3.7171965819152485e-05,
"loss": 1.0757,
"step": 696
},
{
"epoch": 1.7826086956521738,
"grad_norm": 0.4189181625932617,
"learning_rate": 3.715887402790928e-05,
"loss": 1.0535,
"step": 697
},
{
"epoch": 1.785166240409207,
"grad_norm": 0.33692620931995027,
"learning_rate": 3.71457543202509e-05,
"loss": 1.072,
"step": 698
},
{
"epoch": 1.7877237851662404,
"grad_norm": 0.3554848164207662,
"learning_rate": 3.713260671752227e-05,
"loss": 1.055,
"step": 699
},
{
"epoch": 1.7902813299232738,
"grad_norm": 0.35182859532585814,
"learning_rate": 3.711943124111373e-05,
"loss": 1.0737,
"step": 700
},
{
"epoch": 1.792838874680307,
"grad_norm": 0.41031103785026257,
"learning_rate": 3.710622791246096e-05,
"loss": 1.0422,
"step": 701
},
{
"epoch": 1.7953964194373402,
"grad_norm": 0.39424300255839323,
"learning_rate": 3.709299675304494e-05,
"loss": 1.0616,
"step": 702
},
{
"epoch": 1.7979539641943734,
"grad_norm": 0.4555029754578195,
"learning_rate": 3.7079737784391945e-05,
"loss": 1.0631,
"step": 703
},
{
"epoch": 1.8005115089514065,
"grad_norm": 0.4705951467702596,
"learning_rate": 3.7066451028073485e-05,
"loss": 1.0448,
"step": 704
},
{
"epoch": 1.80306905370844,
"grad_norm": 0.44303616071257196,
"learning_rate": 3.705313650570628e-05,
"loss": 1.0726,
"step": 705
},
{
"epoch": 1.8056265984654731,
"grad_norm": 0.3465904737573362,
"learning_rate": 3.703979423895223e-05,
"loss": 1.0242,
"step": 706
},
{
"epoch": 1.8081841432225065,
"grad_norm": 0.30931641016024963,
"learning_rate": 3.702642424951837e-05,
"loss": 1.0309,
"step": 707
},
{
"epoch": 1.8107416879795397,
"grad_norm": 0.3459555084583964,
"learning_rate": 3.701302655915683e-05,
"loss": 1.0492,
"step": 708
},
{
"epoch": 1.813299232736573,
"grad_norm": 0.3207413041259264,
"learning_rate": 3.699960118966482e-05,
"loss": 1.0059,
"step": 709
},
{
"epoch": 1.815856777493606,
"grad_norm": 0.27335826678249586,
"learning_rate": 3.698614816288458e-05,
"loss": 1.0525,
"step": 710
},
{
"epoch": 1.8184143222506393,
"grad_norm": 0.3512185783619656,
"learning_rate": 3.6972667500703336e-05,
"loss": 1.1068,
"step": 711
},
{
"epoch": 1.8209718670076727,
"grad_norm": 0.3449388732709764,
"learning_rate": 3.6959159225053284e-05,
"loss": 1.0536,
"step": 712
},
{
"epoch": 1.8235294117647058,
"grad_norm": 0.34478662640081204,
"learning_rate": 3.6945623357911546e-05,
"loss": 1.0614,
"step": 713
},
{
"epoch": 1.8260869565217392,
"grad_norm": 0.32433295685155444,
"learning_rate": 3.693205992130013e-05,
"loss": 1.0526,
"step": 714
},
{
"epoch": 1.8286445012787724,
"grad_norm": 0.3007946947243801,
"learning_rate": 3.69184689372859e-05,
"loss": 1.0335,
"step": 715
},
{
"epoch": 1.8312020460358056,
"grad_norm": 0.3415063242165242,
"learning_rate": 3.690485042798054e-05,
"loss": 1.0506,
"step": 716
},
{
"epoch": 1.8337595907928388,
"grad_norm": 0.28639575010482055,
"learning_rate": 3.689120441554051e-05,
"loss": 1.0236,
"step": 717
},
{
"epoch": 1.836317135549872,
"grad_norm": 0.28918187194335754,
"learning_rate": 3.687753092216701e-05,
"loss": 1.0687,
"step": 718
},
{
"epoch": 1.8388746803069054,
"grad_norm": 0.3477206594131831,
"learning_rate": 3.686382997010598e-05,
"loss": 1.0683,
"step": 719
},
{
"epoch": 1.8414322250639388,
"grad_norm": 0.3540698474416019,
"learning_rate": 3.685010158164799e-05,
"loss": 1.0745,
"step": 720
},
{
"epoch": 1.843989769820972,
"grad_norm": 0.2819967874578661,
"learning_rate": 3.683634577912829e-05,
"loss": 1.0743,
"step": 721
},
{
"epoch": 1.8465473145780051,
"grad_norm": 0.34541512764581056,
"learning_rate": 3.682256258492669e-05,
"loss": 1.0331,
"step": 722
},
{
"epoch": 1.8491048593350383,
"grad_norm": 0.40318459343876106,
"learning_rate": 3.6808752021467604e-05,
"loss": 1.06,
"step": 723
},
{
"epoch": 1.8516624040920715,
"grad_norm": 0.4431616126680033,
"learning_rate": 3.6794914111219946e-05,
"loss": 1.0548,
"step": 724
},
{
"epoch": 1.854219948849105,
"grad_norm": 0.43095418190835205,
"learning_rate": 3.678104887669713e-05,
"loss": 1.0805,
"step": 725
},
{
"epoch": 1.856777493606138,
"grad_norm": 0.3359930310460609,
"learning_rate": 3.676715634045703e-05,
"loss": 1.0034,
"step": 726
},
{
"epoch": 1.8593350383631715,
"grad_norm": 0.2783122552140855,
"learning_rate": 3.675323652510193e-05,
"loss": 1.0313,
"step": 727
},
{
"epoch": 1.8618925831202047,
"grad_norm": 0.31914558241105906,
"learning_rate": 3.67392894532785e-05,
"loss": 1.0368,
"step": 728
},
{
"epoch": 1.8644501278772379,
"grad_norm": 0.31136562159969755,
"learning_rate": 3.672531514767775e-05,
"loss": 1.0933,
"step": 729
},
{
"epoch": 1.867007672634271,
"grad_norm": 0.3010147767390215,
"learning_rate": 3.671131363103501e-05,
"loss": 1.0495,
"step": 730
},
{
"epoch": 1.8695652173913042,
"grad_norm": 0.26268194061369654,
"learning_rate": 3.6697284926129875e-05,
"loss": 1.0197,
"step": 731
},
{
"epoch": 1.8721227621483376,
"grad_norm": 0.34555978577155866,
"learning_rate": 3.668322905578615e-05,
"loss": 1.0575,
"step": 732
},
{
"epoch": 1.8746803069053708,
"grad_norm": 0.38577665433145175,
"learning_rate": 3.666914604287188e-05,
"loss": 1.0656,
"step": 733
},
{
"epoch": 1.8772378516624042,
"grad_norm": 0.46281189605513456,
"learning_rate": 3.665503591029924e-05,
"loss": 1.0272,
"step": 734
},
{
"epoch": 1.8797953964194374,
"grad_norm": 0.5872098135377278,
"learning_rate": 3.664089868102452e-05,
"loss": 1.055,
"step": 735
},
{
"epoch": 1.8823529411764706,
"grad_norm": 0.7304569569675708,
"learning_rate": 3.662673437804813e-05,
"loss": 1.0874,
"step": 736
},
{
"epoch": 1.8849104859335037,
"grad_norm": 0.7801051737238067,
"learning_rate": 3.6612543024414484e-05,
"loss": 1.0627,
"step": 737
},
{
"epoch": 1.887468030690537,
"grad_norm": 0.64787713492266,
"learning_rate": 3.659832464321204e-05,
"loss": 1.0656,
"step": 738
},
{
"epoch": 1.8900255754475703,
"grad_norm": 0.41454153010254297,
"learning_rate": 3.658407925757322e-05,
"loss": 1.05,
"step": 739
},
{
"epoch": 1.8925831202046037,
"grad_norm": 0.3315606964806271,
"learning_rate": 3.656980689067437e-05,
"loss": 1.0223,
"step": 740
},
{
"epoch": 1.895140664961637,
"grad_norm": 0.3973251331292121,
"learning_rate": 3.655550756573573e-05,
"loss": 1.0356,
"step": 741
},
{
"epoch": 1.89769820971867,
"grad_norm": 0.42047840545039894,
"learning_rate": 3.654118130602141e-05,
"loss": 1.0408,
"step": 742
},
{
"epoch": 1.9002557544757033,
"grad_norm": 0.46061717483591413,
"learning_rate": 3.652682813483935e-05,
"loss": 1.0794,
"step": 743
},
{
"epoch": 1.9028132992327365,
"grad_norm": 0.42961953038043665,
"learning_rate": 3.6512448075541255e-05,
"loss": 1.031,
"step": 744
},
{
"epoch": 1.9053708439897699,
"grad_norm": 0.29462576072812346,
"learning_rate": 3.649804115152258e-05,
"loss": 1.0297,
"step": 745
},
{
"epoch": 1.907928388746803,
"grad_norm": 0.299083936768984,
"learning_rate": 3.648360738622249e-05,
"loss": 1.0534,
"step": 746
},
{
"epoch": 1.9104859335038364,
"grad_norm": 0.36689992823979556,
"learning_rate": 3.646914680312383e-05,
"loss": 1.0454,
"step": 747
},
{
"epoch": 1.9130434782608696,
"grad_norm": 0.4839112744409888,
"learning_rate": 3.645465942575305e-05,
"loss": 1.059,
"step": 748
},
{
"epoch": 1.9156010230179028,
"grad_norm": 0.5199137184375483,
"learning_rate": 3.644014527768021e-05,
"loss": 1.0825,
"step": 749
},
{
"epoch": 1.918158567774936,
"grad_norm": 0.4131122860589314,
"learning_rate": 3.642560438251893e-05,
"loss": 1.0359,
"step": 750
},
{
"epoch": 1.9207161125319692,
"grad_norm": 0.35274316648685045,
"learning_rate": 3.641103676392633e-05,
"loss": 1.0877,
"step": 751
},
{
"epoch": 1.9232736572890026,
"grad_norm": 0.2805128295164736,
"learning_rate": 3.639644244560301e-05,
"loss": 1.0383,
"step": 752
},
{
"epoch": 1.9258312020460358,
"grad_norm": 0.4403423533825957,
"learning_rate": 3.6381821451293024e-05,
"loss": 1.0558,
"step": 753
},
{
"epoch": 1.9283887468030692,
"grad_norm": 0.6064213238314878,
"learning_rate": 3.636717380478381e-05,
"loss": 1.0499,
"step": 754
},
{
"epoch": 1.9309462915601023,
"grad_norm": 0.7181613250015212,
"learning_rate": 3.6352499529906175e-05,
"loss": 1.0191,
"step": 755
},
{
"epoch": 1.9335038363171355,
"grad_norm": 0.7041352397077552,
"learning_rate": 3.6337798650534244e-05,
"loss": 1.0568,
"step": 756
},
{
"epoch": 1.9360613810741687,
"grad_norm": 0.5692276213781257,
"learning_rate": 3.6323071190585434e-05,
"loss": 1.0528,
"step": 757
},
{
"epoch": 1.938618925831202,
"grad_norm": 0.3767034547471204,
"learning_rate": 3.630831717402039e-05,
"loss": 1.0341,
"step": 758
},
{
"epoch": 1.9411764705882353,
"grad_norm": 0.3534327156140319,
"learning_rate": 3.629353662484299e-05,
"loss": 1.0837,
"step": 759
},
{
"epoch": 1.9437340153452687,
"grad_norm": 0.460660267811811,
"learning_rate": 3.6278729567100255e-05,
"loss": 1.0886,
"step": 760
},
{
"epoch": 1.9462915601023019,
"grad_norm": 0.43356116855747223,
"learning_rate": 3.626389602488234e-05,
"loss": 1.0867,
"step": 761
},
{
"epoch": 1.948849104859335,
"grad_norm": 0.37998176300793174,
"learning_rate": 3.624903602232249e-05,
"loss": 1.0599,
"step": 762
},
{
"epoch": 1.9514066496163682,
"grad_norm": 0.36690914711595024,
"learning_rate": 3.623414958359701e-05,
"loss": 1.0517,
"step": 763
},
{
"epoch": 1.9539641943734014,
"grad_norm": 0.4019781273317386,
"learning_rate": 3.6219236732925194e-05,
"loss": 1.0665,
"step": 764
},
{
"epoch": 1.9565217391304348,
"grad_norm": 0.41295666906168105,
"learning_rate": 3.6204297494569324e-05,
"loss": 0.9985,
"step": 765
},
{
"epoch": 1.959079283887468,
"grad_norm": 0.3802888856096745,
"learning_rate": 3.618933189283461e-05,
"loss": 1.0605,
"step": 766
},
{
"epoch": 1.9616368286445014,
"grad_norm": 0.3910690179319618,
"learning_rate": 3.6174339952069136e-05,
"loss": 1.0315,
"step": 767
},
{
"epoch": 1.9641943734015346,
"grad_norm": 0.40883673703654155,
"learning_rate": 3.615932169666386e-05,
"loss": 1.0586,
"step": 768
},
{
"epoch": 1.9667519181585678,
"grad_norm": 0.35792751864639316,
"learning_rate": 3.6144277151052554e-05,
"loss": 1.0662,
"step": 769
},
{
"epoch": 1.969309462915601,
"grad_norm": 0.4362924985149231,
"learning_rate": 3.612920633971174e-05,
"loss": 1.0465,
"step": 770
},
{
"epoch": 1.9718670076726341,
"grad_norm": 0.4459730188737693,
"learning_rate": 3.6114109287160695e-05,
"loss": 1.038,
"step": 771
},
{
"epoch": 1.9744245524296675,
"grad_norm": 0.43615247342723856,
"learning_rate": 3.609898601796137e-05,
"loss": 1.0547,
"step": 772
},
{
"epoch": 1.976982097186701,
"grad_norm": 0.3733557356515598,
"learning_rate": 3.608383655671839e-05,
"loss": 1.067,
"step": 773
},
{
"epoch": 1.979539641943734,
"grad_norm": 0.31470776756495644,
"learning_rate": 3.606866092807897e-05,
"loss": 1.035,
"step": 774
},
{
"epoch": 1.9820971867007673,
"grad_norm": 0.34167640349671424,
"learning_rate": 3.6053459156732914e-05,
"loss": 1.0439,
"step": 775
},
{
"epoch": 1.9846547314578005,
"grad_norm": 0.33790678920141903,
"learning_rate": 3.603823126741256e-05,
"loss": 1.0628,
"step": 776
},
{
"epoch": 1.9872122762148337,
"grad_norm": 0.3670022097816714,
"learning_rate": 3.602297728489272e-05,
"loss": 1.04,
"step": 777
},
{
"epoch": 1.989769820971867,
"grad_norm": 0.37533163713754947,
"learning_rate": 3.6007697233990666e-05,
"loss": 1.0831,
"step": 778
},
{
"epoch": 1.9923273657289002,
"grad_norm": 0.30077837214671677,
"learning_rate": 3.5992391139566106e-05,
"loss": 1.0627,
"step": 779
},
{
"epoch": 1.9948849104859336,
"grad_norm": 0.28000597836795454,
"learning_rate": 3.597705902652108e-05,
"loss": 1.0424,
"step": 780
},
{
"epoch": 1.9974424552429668,
"grad_norm": 0.3509850987209364,
"learning_rate": 3.5961700919799994e-05,
"loss": 1.0533,
"step": 781
},
{
"epoch": 2.0,
"grad_norm": 0.3682542878028121,
"learning_rate": 3.594631684438951e-05,
"loss": 1.0282,
"step": 782
},
{
"epoch": 2.002557544757033,
"grad_norm": 0.34476887243283794,
"learning_rate": 3.593090682531857e-05,
"loss": 1.03,
"step": 783
},
{
"epoch": 2.0051150895140664,
"grad_norm": 0.3631067518938324,
"learning_rate": 3.59154708876583e-05,
"loss": 1.0221,
"step": 784
},
{
"epoch": 2.0076726342710995,
"grad_norm": 0.2988045459730597,
"learning_rate": 3.590000905652202e-05,
"loss": 1.023,
"step": 785
},
{
"epoch": 2.010230179028133,
"grad_norm": 0.3780525945800409,
"learning_rate": 3.588452135706514e-05,
"loss": 1.1067,
"step": 786
},
{
"epoch": 2.0127877237851663,
"grad_norm": 0.506960936315167,
"learning_rate": 3.58690078144852e-05,
"loss": 1.0386,
"step": 787
},
{
"epoch": 2.0153452685421995,
"grad_norm": 0.5036095017149632,
"learning_rate": 3.585346845402175e-05,
"loss": 1.0535,
"step": 788
},
{
"epoch": 2.0179028132992327,
"grad_norm": 0.4363480984900362,
"learning_rate": 3.5837903300956376e-05,
"loss": 1.035,
"step": 789
},
{
"epoch": 2.020460358056266,
"grad_norm": 0.39557407720366333,
"learning_rate": 3.582231238061259e-05,
"loss": 1.0219,
"step": 790
},
{
"epoch": 2.023017902813299,
"grad_norm": 0.35542358260228324,
"learning_rate": 3.580669571835585e-05,
"loss": 1.023,
"step": 791
},
{
"epoch": 2.0255754475703327,
"grad_norm": 0.39211334875702714,
"learning_rate": 3.5791053339593494e-05,
"loss": 1.0281,
"step": 792
},
{
"epoch": 2.028132992327366,
"grad_norm": 0.4703716622817256,
"learning_rate": 3.577538526977469e-05,
"loss": 1.0183,
"step": 793
},
{
"epoch": 2.030690537084399,
"grad_norm": 0.44802012462862173,
"learning_rate": 3.575969153439041e-05,
"loss": 1.019,
"step": 794
},
{
"epoch": 2.0332480818414322,
"grad_norm": 0.4847202684592886,
"learning_rate": 3.574397215897338e-05,
"loss": 1.0269,
"step": 795
},
{
"epoch": 2.0358056265984654,
"grad_norm": 0.4174455108325541,
"learning_rate": 3.572822716909804e-05,
"loss": 1.0326,
"step": 796
},
{
"epoch": 2.0383631713554986,
"grad_norm": 0.29359506235186306,
"learning_rate": 3.57124565903805e-05,
"loss": 1.0571,
"step": 797
},
{
"epoch": 2.040920716112532,
"grad_norm": 0.4211894765567659,
"learning_rate": 3.569666044847851e-05,
"loss": 1.0306,
"step": 798
},
{
"epoch": 2.0434782608695654,
"grad_norm": 0.5593610935644447,
"learning_rate": 3.568083876909141e-05,
"loss": 1.0324,
"step": 799
},
{
"epoch": 2.0460358056265986,
"grad_norm": 0.5332435545150164,
"learning_rate": 3.566499157796007e-05,
"loss": 1.0755,
"step": 800
},
{
"epoch": 2.0485933503836318,
"grad_norm": 0.3510096438458946,
"learning_rate": 3.564911890086689e-05,
"loss": 1.0276,
"step": 801
},
{
"epoch": 2.051150895140665,
"grad_norm": 0.31958326640243073,
"learning_rate": 3.563322076363573e-05,
"loss": 1.0346,
"step": 802
},
{
"epoch": 2.053708439897698,
"grad_norm": 0.5313641729011047,
"learning_rate": 3.561729719213184e-05,
"loss": 1.0407,
"step": 803
},
{
"epoch": 2.0562659846547313,
"grad_norm": 0.5858378839343258,
"learning_rate": 3.56013482122619e-05,
"loss": 1.068,
"step": 804
},
{
"epoch": 2.0588235294117645,
"grad_norm": 0.5215568823069776,
"learning_rate": 3.558537384997389e-05,
"loss": 1.0185,
"step": 805
},
{
"epoch": 2.061381074168798,
"grad_norm": 0.678379428738595,
"learning_rate": 3.556937413125711e-05,
"loss": 1.0322,
"step": 806
},
{
"epoch": 2.0639386189258313,
"grad_norm": 0.8337275740784315,
"learning_rate": 3.555334908214208e-05,
"loss": 1.0734,
"step": 807
},
{
"epoch": 2.0664961636828645,
"grad_norm": 0.7261038633451826,
"learning_rate": 3.553729872870058e-05,
"loss": 1.0477,
"step": 808
},
{
"epoch": 2.0690537084398977,
"grad_norm": 0.41343717310926115,
"learning_rate": 3.552122309704551e-05,
"loss": 1.0238,
"step": 809
},
{
"epoch": 2.071611253196931,
"grad_norm": 0.42014708296303344,
"learning_rate": 3.5505122213330926e-05,
"loss": 1.029,
"step": 810
},
{
"epoch": 2.074168797953964,
"grad_norm": 0.42859887530027807,
"learning_rate": 3.548899610375197e-05,
"loss": 1.0507,
"step": 811
},
{
"epoch": 2.0767263427109977,
"grad_norm": 0.2837953461637091,
"learning_rate": 3.547284479454479e-05,
"loss": 1.0234,
"step": 812
},
{
"epoch": 2.079283887468031,
"grad_norm": 0.3108594231616873,
"learning_rate": 3.545666831198658e-05,
"loss": 1.025,
"step": 813
},
{
"epoch": 2.081841432225064,
"grad_norm": 0.3210796979622124,
"learning_rate": 3.544046668239545e-05,
"loss": 1.0855,
"step": 814
},
{
"epoch": 2.084398976982097,
"grad_norm": 0.35290042721284903,
"learning_rate": 3.542423993213045e-05,
"loss": 1.0127,
"step": 815
},
{
"epoch": 2.0869565217391304,
"grad_norm": 0.3061610233453768,
"learning_rate": 3.540798808759148e-05,
"loss": 1.0238,
"step": 816
},
{
"epoch": 2.0895140664961636,
"grad_norm": 0.3158641534135511,
"learning_rate": 3.5391711175219274e-05,
"loss": 1.0336,
"step": 817
},
{
"epoch": 2.0920716112531967,
"grad_norm": 0.320676408292332,
"learning_rate": 3.537540922149535e-05,
"loss": 1.0161,
"step": 818
},
{
"epoch": 2.0946291560102304,
"grad_norm": 0.2784942734777997,
"learning_rate": 3.535908225294198e-05,
"loss": 1.018,
"step": 819
},
{
"epoch": 2.0971867007672635,
"grad_norm": 0.27427752623715057,
"learning_rate": 3.534273029612211e-05,
"loss": 1.0198,
"step": 820
},
{
"epoch": 2.0997442455242967,
"grad_norm": 0.3424435250558866,
"learning_rate": 3.5326353377639346e-05,
"loss": 1.0307,
"step": 821
},
{
"epoch": 2.10230179028133,
"grad_norm": 0.40395744995559435,
"learning_rate": 3.530995152413792e-05,
"loss": 1.0167,
"step": 822
},
{
"epoch": 2.104859335038363,
"grad_norm": 0.42322881578827115,
"learning_rate": 3.529352476230262e-05,
"loss": 1.0521,
"step": 823
},
{
"epoch": 2.1074168797953963,
"grad_norm": 0.39403243993138826,
"learning_rate": 3.527707311885876e-05,
"loss": 1.0409,
"step": 824
},
{
"epoch": 2.10997442455243,
"grad_norm": 0.3399966279762583,
"learning_rate": 3.526059662057213e-05,
"loss": 1.0439,
"step": 825
},
{
"epoch": 2.112531969309463,
"grad_norm": 0.36391592870214196,
"learning_rate": 3.5244095294248976e-05,
"loss": 1.0374,
"step": 826
},
{
"epoch": 2.1150895140664963,
"grad_norm": 0.41613733319623397,
"learning_rate": 3.522756916673592e-05,
"loss": 1.0296,
"step": 827
},
{
"epoch": 2.1176470588235294,
"grad_norm": 0.4867739010480437,
"learning_rate": 3.521101826491992e-05,
"loss": 1.0367,
"step": 828
},
{
"epoch": 2.1202046035805626,
"grad_norm": 0.6013178320886087,
"learning_rate": 3.5194442615728285e-05,
"loss": 1.0456,
"step": 829
},
{
"epoch": 2.122762148337596,
"grad_norm": 0.6802290996954595,
"learning_rate": 3.5177842246128546e-05,
"loss": 1.0705,
"step": 830
},
{
"epoch": 2.125319693094629,
"grad_norm": 0.48182411903532013,
"learning_rate": 3.516121718312847e-05,
"loss": 1.0655,
"step": 831
},
{
"epoch": 2.1278772378516626,
"grad_norm": 0.43599863627347213,
"learning_rate": 3.514456745377598e-05,
"loss": 1.0592,
"step": 832
},
{
"epoch": 2.130434782608696,
"grad_norm": 0.5503409777534483,
"learning_rate": 3.512789308515917e-05,
"loss": 1.0568,
"step": 833
},
{
"epoch": 2.132992327365729,
"grad_norm": 0.6147183774365734,
"learning_rate": 3.5111194104406176e-05,
"loss": 1.0174,
"step": 834
},
{
"epoch": 2.135549872122762,
"grad_norm": 0.6422101075595786,
"learning_rate": 3.5094470538685204e-05,
"loss": 1.0498,
"step": 835
},
{
"epoch": 2.1381074168797953,
"grad_norm": 0.5579693240767665,
"learning_rate": 3.507772241520445e-05,
"loss": 1.0616,
"step": 836
},
{
"epoch": 2.1406649616368285,
"grad_norm": 0.37978695418623076,
"learning_rate": 3.506094976121206e-05,
"loss": 1.0003,
"step": 837
},
{
"epoch": 2.1432225063938617,
"grad_norm": 0.3944980589848742,
"learning_rate": 3.5044152603996095e-05,
"loss": 1.0644,
"step": 838
},
{
"epoch": 2.1457800511508953,
"grad_norm": 0.4110471908405651,
"learning_rate": 3.5027330970884476e-05,
"loss": 1.0408,
"step": 839
},
{
"epoch": 2.1483375959079285,
"grad_norm": 0.4748490112801617,
"learning_rate": 3.501048488924496e-05,
"loss": 1.0635,
"step": 840
},
{
"epoch": 2.1508951406649617,
"grad_norm": 0.4270454490764321,
"learning_rate": 3.4993614386485055e-05,
"loss": 1.0382,
"step": 841
},
{
"epoch": 2.153452685421995,
"grad_norm": 0.23066751027416876,
"learning_rate": 3.497671949005202e-05,
"loss": 1.0564,
"step": 842
},
{
"epoch": 2.156010230179028,
"grad_norm": 0.35749933474842516,
"learning_rate": 3.4959800227432795e-05,
"loss": 1.0287,
"step": 843
},
{
"epoch": 2.1585677749360612,
"grad_norm": 0.48448193801861017,
"learning_rate": 3.494285662615397e-05,
"loss": 1.0773,
"step": 844
},
{
"epoch": 2.1611253196930944,
"grad_norm": 0.44928477334103767,
"learning_rate": 3.4925888713781706e-05,
"loss": 1.0334,
"step": 845
},
{
"epoch": 2.163682864450128,
"grad_norm": 0.39158927023658086,
"learning_rate": 3.4908896517921754e-05,
"loss": 1.0306,
"step": 846
},
{
"epoch": 2.166240409207161,
"grad_norm": 0.41488260647215264,
"learning_rate": 3.489188006621934e-05,
"loss": 0.9938,
"step": 847
},
{
"epoch": 2.1687979539641944,
"grad_norm": 0.48278574906259986,
"learning_rate": 3.487483938635918e-05,
"loss": 1.0846,
"step": 848
},
{
"epoch": 2.1713554987212276,
"grad_norm": 0.43582210385725134,
"learning_rate": 3.485777450606539e-05,
"loss": 1.0433,
"step": 849
},
{
"epoch": 2.1739130434782608,
"grad_norm": 0.40707374169776855,
"learning_rate": 3.4840685453101456e-05,
"loss": 1.0445,
"step": 850
},
{
"epoch": 2.176470588235294,
"grad_norm": 0.30587971102368344,
"learning_rate": 3.48235722552702e-05,
"loss": 1.0458,
"step": 851
},
{
"epoch": 2.1790281329923276,
"grad_norm": 0.402736732310713,
"learning_rate": 3.480643494041374e-05,
"loss": 1.0638,
"step": 852
},
{
"epoch": 2.1815856777493607,
"grad_norm": 0.5135740283646704,
"learning_rate": 3.4789273536413404e-05,
"loss": 1.0018,
"step": 853
},
{
"epoch": 2.184143222506394,
"grad_norm": 0.4275618393848173,
"learning_rate": 3.477208807118972e-05,
"loss": 0.9836,
"step": 854
},
{
"epoch": 2.186700767263427,
"grad_norm": 0.3452717169194966,
"learning_rate": 3.475487857270237e-05,
"loss": 1.0369,
"step": 855
},
{
"epoch": 2.1892583120204603,
"grad_norm": 0.3407872175351053,
"learning_rate": 3.473764506895012e-05,
"loss": 1.0573,
"step": 856
},
{
"epoch": 2.1918158567774935,
"grad_norm": 0.4063848580515412,
"learning_rate": 3.472038758797084e-05,
"loss": 1.0744,
"step": 857
},
{
"epoch": 2.1943734015345266,
"grad_norm": 0.3624007949000717,
"learning_rate": 3.470310615784133e-05,
"loss": 1.0654,
"step": 858
},
{
"epoch": 2.1969309462915603,
"grad_norm": 0.28221976818130207,
"learning_rate": 3.468580080667742e-05,
"loss": 1.0613,
"step": 859
},
{
"epoch": 2.1994884910485935,
"grad_norm": 0.2876965443465958,
"learning_rate": 3.466847156263382e-05,
"loss": 1.0577,
"step": 860
},
{
"epoch": 2.2020460358056266,
"grad_norm": 0.4260872080561482,
"learning_rate": 3.465111845390414e-05,
"loss": 1.0485,
"step": 861
},
{
"epoch": 2.20460358056266,
"grad_norm": 0.38264859298873866,
"learning_rate": 3.46337415087208e-05,
"loss": 1.0222,
"step": 862
},
{
"epoch": 2.207161125319693,
"grad_norm": 0.3352129289323696,
"learning_rate": 3.4616340755355e-05,
"loss": 1.0431,
"step": 863
},
{
"epoch": 2.209718670076726,
"grad_norm": 0.3498625416338442,
"learning_rate": 3.459891622211667e-05,
"loss": 1.048,
"step": 864
},
{
"epoch": 2.21227621483376,
"grad_norm": 0.3993144499753061,
"learning_rate": 3.458146793735445e-05,
"loss": 1.0063,
"step": 865
},
{
"epoch": 2.214833759590793,
"grad_norm": 0.3423437911803536,
"learning_rate": 3.4563995929455605e-05,
"loss": 1.0348,
"step": 866
},
{
"epoch": 2.217391304347826,
"grad_norm": 0.28404682730352504,
"learning_rate": 3.4546500226846e-05,
"loss": 1.0788,
"step": 867
},
{
"epoch": 2.2199488491048593,
"grad_norm": 0.40531432541743023,
"learning_rate": 3.452898085799004e-05,
"loss": 1.0189,
"step": 868
},
{
"epoch": 2.2225063938618925,
"grad_norm": 0.34142009368516163,
"learning_rate": 3.451143785139066e-05,
"loss": 1.0552,
"step": 869
},
{
"epoch": 2.2250639386189257,
"grad_norm": 0.39017136195023944,
"learning_rate": 3.4493871235589217e-05,
"loss": 1.041,
"step": 870
},
{
"epoch": 2.227621483375959,
"grad_norm": 0.3115046179826426,
"learning_rate": 3.44762810391655e-05,
"loss": 1.0199,
"step": 871
},
{
"epoch": 2.2301790281329925,
"grad_norm": 0.3545970292998508,
"learning_rate": 3.445866729073767e-05,
"loss": 1.0406,
"step": 872
},
{
"epoch": 2.2327365728900257,
"grad_norm": 0.37922403147004347,
"learning_rate": 3.444103001896218e-05,
"loss": 1.0591,
"step": 873
},
{
"epoch": 2.235294117647059,
"grad_norm": 0.3658839323909771,
"learning_rate": 3.4423369252533774e-05,
"loss": 1.0461,
"step": 874
},
{
"epoch": 2.237851662404092,
"grad_norm": 0.37681775401843953,
"learning_rate": 3.440568502018541e-05,
"loss": 1.0177,
"step": 875
},
{
"epoch": 2.2404092071611252,
"grad_norm": 0.3250815268266776,
"learning_rate": 3.438797735068825e-05,
"loss": 1.0344,
"step": 876
},
{
"epoch": 2.2429667519181584,
"grad_norm": 0.2577522137129716,
"learning_rate": 3.437024627285155e-05,
"loss": 1.0231,
"step": 877
},
{
"epoch": 2.2455242966751916,
"grad_norm": 0.3062484956966101,
"learning_rate": 3.435249181552265e-05,
"loss": 1.0312,
"step": 878
},
{
"epoch": 2.2480818414322252,
"grad_norm": 0.40714785593643416,
"learning_rate": 3.433471400758697e-05,
"loss": 1.0314,
"step": 879
},
{
"epoch": 2.2506393861892584,
"grad_norm": 0.3881465741071191,
"learning_rate": 3.431691287796787e-05,
"loss": 1.005,
"step": 880
},
{
"epoch": 2.2531969309462916,
"grad_norm": 0.3159607748832716,
"learning_rate": 3.4299088455626685e-05,
"loss": 1.0251,
"step": 881
},
{
"epoch": 2.2557544757033248,
"grad_norm": 0.2785017524540126,
"learning_rate": 3.428124076956263e-05,
"loss": 0.9916,
"step": 882
},
{
"epoch": 2.258312020460358,
"grad_norm": 0.2753619492554733,
"learning_rate": 3.4263369848812784e-05,
"loss": 1.0309,
"step": 883
},
{
"epoch": 2.260869565217391,
"grad_norm": 0.3120510209959417,
"learning_rate": 3.424547572245201e-05,
"loss": 1.0282,
"step": 884
},
{
"epoch": 2.2634271099744243,
"grad_norm": 0.27056729525588585,
"learning_rate": 3.422755841959294e-05,
"loss": 1.0458,
"step": 885
},
{
"epoch": 2.265984654731458,
"grad_norm": 0.2863845589692127,
"learning_rate": 3.420961796938589e-05,
"loss": 1.0534,
"step": 886
},
{
"epoch": 2.268542199488491,
"grad_norm": 0.3210914318226761,
"learning_rate": 3.419165440101887e-05,
"loss": 1.0105,
"step": 887
},
{
"epoch": 2.2710997442455243,
"grad_norm": 0.41616857236392335,
"learning_rate": 3.417366774371747e-05,
"loss": 1.0152,
"step": 888
},
{
"epoch": 2.2736572890025575,
"grad_norm": 0.42605052861672543,
"learning_rate": 3.415565802674486e-05,
"loss": 1.0382,
"step": 889
},
{
"epoch": 2.2762148337595907,
"grad_norm": 0.31861481167521705,
"learning_rate": 3.4137625279401724e-05,
"loss": 1.0738,
"step": 890
},
{
"epoch": 2.2787723785166243,
"grad_norm": 0.27906706808206594,
"learning_rate": 3.411956953102621e-05,
"loss": 1.0517,
"step": 891
},
{
"epoch": 2.2813299232736575,
"grad_norm": 0.31056275044088116,
"learning_rate": 3.41014908109939e-05,
"loss": 1.0354,
"step": 892
},
{
"epoch": 2.2838874680306906,
"grad_norm": 0.26727159109094756,
"learning_rate": 3.408338914871774e-05,
"loss": 1.0407,
"step": 893
},
{
"epoch": 2.286445012787724,
"grad_norm": 0.3426272544798734,
"learning_rate": 3.406526457364799e-05,
"loss": 1.0598,
"step": 894
},
{
"epoch": 2.289002557544757,
"grad_norm": 0.31816460953552667,
"learning_rate": 3.404711711527222e-05,
"loss": 1.0466,
"step": 895
},
{
"epoch": 2.29156010230179,
"grad_norm": 0.2888074651731447,
"learning_rate": 3.40289468031152e-05,
"loss": 1.0363,
"step": 896
},
{
"epoch": 2.2941176470588234,
"grad_norm": 0.2602304741054429,
"learning_rate": 3.401075366673888e-05,
"loss": 1.0336,
"step": 897
},
{
"epoch": 2.296675191815857,
"grad_norm": 0.4052941105334914,
"learning_rate": 3.399253773574238e-05,
"loss": 1.0112,
"step": 898
},
{
"epoch": 2.29923273657289,
"grad_norm": 0.41837190737307517,
"learning_rate": 3.397429903976186e-05,
"loss": 1.0454,
"step": 899
},
{
"epoch": 2.3017902813299234,
"grad_norm": 0.39021994147720995,
"learning_rate": 3.3956037608470555e-05,
"loss": 1.0304,
"step": 900
},
{
"epoch": 2.3043478260869565,
"grad_norm": 0.3434091790542349,
"learning_rate": 3.393775347157865e-05,
"loss": 1.0231,
"step": 901
},
{
"epoch": 2.3069053708439897,
"grad_norm": 0.3243303846056098,
"learning_rate": 3.391944665883332e-05,
"loss": 1.0301,
"step": 902
},
{
"epoch": 2.309462915601023,
"grad_norm": 0.34434845885945947,
"learning_rate": 3.390111720001858e-05,
"loss": 1.0488,
"step": 903
},
{
"epoch": 2.312020460358056,
"grad_norm": 0.39625351389911473,
"learning_rate": 3.388276512495533e-05,
"loss": 0.9764,
"step": 904
},
{
"epoch": 2.3145780051150897,
"grad_norm": 0.3991059269330394,
"learning_rate": 3.386439046350123e-05,
"loss": 1.0104,
"step": 905
},
{
"epoch": 2.317135549872123,
"grad_norm": 0.3485099377649964,
"learning_rate": 3.3845993245550726e-05,
"loss": 1.0195,
"step": 906
},
{
"epoch": 2.319693094629156,
"grad_norm": 0.3760750744049425,
"learning_rate": 3.382757350103494e-05,
"loss": 1.0265,
"step": 907
},
{
"epoch": 2.3222506393861893,
"grad_norm": 0.3092275749291268,
"learning_rate": 3.380913125992163e-05,
"loss": 1.0647,
"step": 908
},
{
"epoch": 2.3248081841432224,
"grad_norm": 0.3407775851645365,
"learning_rate": 3.3790666552215177e-05,
"loss": 1.0016,
"step": 909
},
{
"epoch": 2.3273657289002556,
"grad_norm": 0.3986511554019992,
"learning_rate": 3.37721794079565e-05,
"loss": 1.0787,
"step": 910
},
{
"epoch": 2.329923273657289,
"grad_norm": 0.34347930226910395,
"learning_rate": 3.375366985722305e-05,
"loss": 1.0527,
"step": 911
},
{
"epoch": 2.3324808184143224,
"grad_norm": 0.37052724204327386,
"learning_rate": 3.373513793012867e-05,
"loss": 1.0117,
"step": 912
},
{
"epoch": 2.3350383631713556,
"grad_norm": 0.30051731142995297,
"learning_rate": 3.371658365682368e-05,
"loss": 1.0454,
"step": 913
},
{
"epoch": 2.337595907928389,
"grad_norm": 0.2983597250107576,
"learning_rate": 3.36980070674947e-05,
"loss": 1.0454,
"step": 914
},
{
"epoch": 2.340153452685422,
"grad_norm": 0.3812423666925809,
"learning_rate": 3.36794081923647e-05,
"loss": 1.0437,
"step": 915
},
{
"epoch": 2.342710997442455,
"grad_norm": 0.35314006080451715,
"learning_rate": 3.366078706169288e-05,
"loss": 1.0651,
"step": 916
},
{
"epoch": 2.3452685421994883,
"grad_norm": 0.32694659308054125,
"learning_rate": 3.3642143705774644e-05,
"loss": 1.007,
"step": 917
},
{
"epoch": 2.3478260869565215,
"grad_norm": 0.3447766010702551,
"learning_rate": 3.362347815494158e-05,
"loss": 1.0322,
"step": 918
},
{
"epoch": 2.350383631713555,
"grad_norm": 0.4206611305531367,
"learning_rate": 3.360479043956136e-05,
"loss": 1.0179,
"step": 919
},
{
"epoch": 2.3529411764705883,
"grad_norm": 0.3596401018341398,
"learning_rate": 3.358608059003774e-05,
"loss": 1.0204,
"step": 920
},
{
"epoch": 2.3554987212276215,
"grad_norm": 0.3342103360248012,
"learning_rate": 3.356734863681047e-05,
"loss": 1.0247,
"step": 921
},
{
"epoch": 2.3580562659846547,
"grad_norm": 0.3173545008322276,
"learning_rate": 3.354859461035526e-05,
"loss": 1.0223,
"step": 922
},
{
"epoch": 2.360613810741688,
"grad_norm": 0.3714760810500541,
"learning_rate": 3.352981854118375e-05,
"loss": 1.024,
"step": 923
},
{
"epoch": 2.363171355498721,
"grad_norm": 0.3762134358180157,
"learning_rate": 3.3511020459843426e-05,
"loss": 1.0317,
"step": 924
},
{
"epoch": 2.3657289002557547,
"grad_norm": 0.31386453038319984,
"learning_rate": 3.349220039691759e-05,
"loss": 1.0197,
"step": 925
},
{
"epoch": 2.368286445012788,
"grad_norm": 0.33667149744613917,
"learning_rate": 3.347335838302532e-05,
"loss": 0.9966,
"step": 926
},
{
"epoch": 2.370843989769821,
"grad_norm": 0.3656300585469283,
"learning_rate": 3.345449444882137e-05,
"loss": 1.0539,
"step": 927
},
{
"epoch": 2.373401534526854,
"grad_norm": 0.3580120336762772,
"learning_rate": 3.34356086249962e-05,
"loss": 1.061,
"step": 928
},
{
"epoch": 2.3759590792838874,
"grad_norm": 0.3937196727430575,
"learning_rate": 3.3416700942275854e-05,
"loss": 1.0314,
"step": 929
},
{
"epoch": 2.3785166240409206,
"grad_norm": 0.39949395345109756,
"learning_rate": 3.339777143142196e-05,
"loss": 1.0565,
"step": 930
},
{
"epoch": 2.381074168797954,
"grad_norm": 0.30311220929585764,
"learning_rate": 3.3378820123231646e-05,
"loss": 1.0204,
"step": 931
},
{
"epoch": 2.3836317135549874,
"grad_norm": 0.2829587252989885,
"learning_rate": 3.3359847048537496e-05,
"loss": 1.0437,
"step": 932
},
{
"epoch": 2.3861892583120206,
"grad_norm": 0.3348201829659415,
"learning_rate": 3.3340852238207525e-05,
"loss": 1.0278,
"step": 933
},
{
"epoch": 2.3887468030690537,
"grad_norm": 0.40401093974170393,
"learning_rate": 3.33218357231451e-05,
"loss": 1.0147,
"step": 934
},
{
"epoch": 2.391304347826087,
"grad_norm": 0.3390237942322822,
"learning_rate": 3.330279753428891e-05,
"loss": 1.012,
"step": 935
},
{
"epoch": 2.39386189258312,
"grad_norm": 0.35338253519082996,
"learning_rate": 3.328373770261288e-05,
"loss": 1.0361,
"step": 936
},
{
"epoch": 2.3964194373401533,
"grad_norm": 0.341101526298639,
"learning_rate": 3.3264656259126176e-05,
"loss": 1.0327,
"step": 937
},
{
"epoch": 2.398976982097187,
"grad_norm": 0.34588403748614843,
"learning_rate": 3.324555323487311e-05,
"loss": 1.0438,
"step": 938
},
{
"epoch": 2.40153452685422,
"grad_norm": 0.29090150555694466,
"learning_rate": 3.3226428660933115e-05,
"loss": 1.0269,
"step": 939
},
{
"epoch": 2.4040920716112533,
"grad_norm": 0.31151685510973315,
"learning_rate": 3.320728256842067e-05,
"loss": 1.0359,
"step": 940
},
{
"epoch": 2.4066496163682864,
"grad_norm": 0.32291787524173865,
"learning_rate": 3.318811498848527e-05,
"loss": 1.0281,
"step": 941
},
{
"epoch": 2.4092071611253196,
"grad_norm": 0.2630398718729125,
"learning_rate": 3.3168925952311365e-05,
"loss": 1.0335,
"step": 942
},
{
"epoch": 2.411764705882353,
"grad_norm": 0.35223837257888496,
"learning_rate": 3.3149715491118336e-05,
"loss": 1.0378,
"step": 943
},
{
"epoch": 2.414322250639386,
"grad_norm": 0.4546143470475912,
"learning_rate": 3.313048363616038e-05,
"loss": 1.0378,
"step": 944
},
{
"epoch": 2.4168797953964196,
"grad_norm": 0.585508072917306,
"learning_rate": 3.311123041872654e-05,
"loss": 0.9924,
"step": 945
},
{
"epoch": 2.419437340153453,
"grad_norm": 0.5913838503356222,
"learning_rate": 3.309195587014059e-05,
"loss": 1.0119,
"step": 946
},
{
"epoch": 2.421994884910486,
"grad_norm": 0.4289519968378196,
"learning_rate": 3.3072660021761005e-05,
"loss": 1.0628,
"step": 947
},
{
"epoch": 2.424552429667519,
"grad_norm": 0.3033649311346593,
"learning_rate": 3.305334290498094e-05,
"loss": 1.0203,
"step": 948
},
{
"epoch": 2.4271099744245523,
"grad_norm": 0.40540246453508627,
"learning_rate": 3.3034004551228124e-05,
"loss": 1.0351,
"step": 949
},
{
"epoch": 2.4296675191815855,
"grad_norm": 0.5264410731653394,
"learning_rate": 3.3014644991964866e-05,
"loss": 1.017,
"step": 950
},
{
"epoch": 2.4322250639386187,
"grad_norm": 0.4471282277894698,
"learning_rate": 3.299526425868793e-05,
"loss": 1.0315,
"step": 951
},
{
"epoch": 2.4347826086956523,
"grad_norm": 0.2893919281157759,
"learning_rate": 3.2975862382928586e-05,
"loss": 0.9981,
"step": 952
},
{
"epoch": 2.4373401534526855,
"grad_norm": 0.28677192037274934,
"learning_rate": 3.295643939625246e-05,
"loss": 1.0487,
"step": 953
},
{
"epoch": 2.4398976982097187,
"grad_norm": 0.4498390247184629,
"learning_rate": 3.2936995330259535e-05,
"loss": 1.0505,
"step": 954
},
{
"epoch": 2.442455242966752,
"grad_norm": 0.48741447826473155,
"learning_rate": 3.29175302165841e-05,
"loss": 1.0254,
"step": 955
},
{
"epoch": 2.445012787723785,
"grad_norm": 0.4583713318260062,
"learning_rate": 3.289804408689467e-05,
"loss": 1.0357,
"step": 956
},
{
"epoch": 2.4475703324808182,
"grad_norm": 0.3493148651212194,
"learning_rate": 3.2878536972893965e-05,
"loss": 1.0058,
"step": 957
},
{
"epoch": 2.4501278772378514,
"grad_norm": 0.3548246176699886,
"learning_rate": 3.285900890631882e-05,
"loss": 0.9997,
"step": 958
},
{
"epoch": 2.452685421994885,
"grad_norm": 0.43179552446813185,
"learning_rate": 3.2839459918940214e-05,
"loss": 1.0231,
"step": 959
},
{
"epoch": 2.455242966751918,
"grad_norm": 0.40280146595458866,
"learning_rate": 3.2819890042563094e-05,
"loss": 0.9963,
"step": 960
},
{
"epoch": 2.4578005115089514,
"grad_norm": 0.29104437869826166,
"learning_rate": 3.280029930902645e-05,
"loss": 1.065,
"step": 961
},
{
"epoch": 2.4603580562659846,
"grad_norm": 0.30411031126871707,
"learning_rate": 3.2780687750203163e-05,
"loss": 1.0403,
"step": 962
},
{
"epoch": 2.4629156010230178,
"grad_norm": 0.3261139611132537,
"learning_rate": 3.2761055398000024e-05,
"loss": 1.0497,
"step": 963
},
{
"epoch": 2.4654731457800514,
"grad_norm": 0.32078949369997534,
"learning_rate": 3.274140228435763e-05,
"loss": 1.0549,
"step": 964
},
{
"epoch": 2.4680306905370846,
"grad_norm": 0.38769485722519065,
"learning_rate": 3.272172844125038e-05,
"loss": 1.0071,
"step": 965
},
{
"epoch": 2.4705882352941178,
"grad_norm": 0.39011368252766637,
"learning_rate": 3.270203390068638e-05,
"loss": 1.0212,
"step": 966
},
{
"epoch": 2.473145780051151,
"grad_norm": 0.3559640784244842,
"learning_rate": 3.268231869470743e-05,
"loss": 1.0327,
"step": 967
},
{
"epoch": 2.475703324808184,
"grad_norm": 0.2549960248223228,
"learning_rate": 3.266258285538892e-05,
"loss": 1.0358,
"step": 968
},
{
"epoch": 2.4782608695652173,
"grad_norm": 0.3673854964338514,
"learning_rate": 3.264282641483983e-05,
"loss": 1.0364,
"step": 969
},
{
"epoch": 2.4808184143222505,
"grad_norm": 0.41284771075829285,
"learning_rate": 3.2623049405202656e-05,
"loss": 1.0033,
"step": 970
},
{
"epoch": 2.483375959079284,
"grad_norm": 0.4349758735991385,
"learning_rate": 3.2603251858653364e-05,
"loss": 1.0263,
"step": 971
},
{
"epoch": 2.4859335038363173,
"grad_norm": 0.3830161036052956,
"learning_rate": 3.258343380740131e-05,
"loss": 1.0492,
"step": 972
},
{
"epoch": 2.4884910485933505,
"grad_norm": 0.36058423037414133,
"learning_rate": 3.256359528368923e-05,
"loss": 1.0283,
"step": 973
},
{
"epoch": 2.4910485933503836,
"grad_norm": 0.3576003176406701,
"learning_rate": 3.2543736319793166e-05,
"loss": 1.0287,
"step": 974
},
{
"epoch": 2.493606138107417,
"grad_norm": 0.40371145203246356,
"learning_rate": 3.252385694802241e-05,
"loss": 1.0049,
"step": 975
},
{
"epoch": 2.49616368286445,
"grad_norm": 0.538524450709586,
"learning_rate": 3.2503957200719454e-05,
"loss": 1.0362,
"step": 976
},
{
"epoch": 2.498721227621483,
"grad_norm": 0.44608120620805436,
"learning_rate": 3.2484037110259936e-05,
"loss": 1.0257,
"step": 977
},
{
"epoch": 2.501278772378517,
"grad_norm": 0.334739016098985,
"learning_rate": 3.246409670905261e-05,
"loss": 1.0384,
"step": 978
},
{
"epoch": 2.50383631713555,
"grad_norm": 0.34494891498764113,
"learning_rate": 3.244413602953925e-05,
"loss": 1.0418,
"step": 979
},
{
"epoch": 2.506393861892583,
"grad_norm": 0.26027241767697906,
"learning_rate": 3.242415510419465e-05,
"loss": 0.9994,
"step": 980
},
{
"epoch": 2.5089514066496164,
"grad_norm": 0.35746746709039434,
"learning_rate": 3.2404153965526516e-05,
"loss": 1.0131,
"step": 981
},
{
"epoch": 2.5115089514066495,
"grad_norm": 0.38212998444391627,
"learning_rate": 3.2384132646075446e-05,
"loss": 0.9952,
"step": 982
},
{
"epoch": 2.5140664961636827,
"grad_norm": 0.3356711985262144,
"learning_rate": 3.236409117841488e-05,
"loss": 1.0071,
"step": 983
},
{
"epoch": 2.516624040920716,
"grad_norm": 0.3295389578061702,
"learning_rate": 3.2344029595151024e-05,
"loss": 1.0353,
"step": 984
},
{
"epoch": 2.5191815856777495,
"grad_norm": 0.3114755787433909,
"learning_rate": 3.232394792892284e-05,
"loss": 1.0312,
"step": 985
},
{
"epoch": 2.5217391304347827,
"grad_norm": 0.3102449437708034,
"learning_rate": 3.230384621240193e-05,
"loss": 1.0036,
"step": 986
},
{
"epoch": 2.524296675191816,
"grad_norm": 0.29165577216030847,
"learning_rate": 3.2283724478292525e-05,
"loss": 1.0192,
"step": 987
},
{
"epoch": 2.526854219948849,
"grad_norm": 0.3928864738184209,
"learning_rate": 3.226358275933144e-05,
"loss": 1.0421,
"step": 988
},
{
"epoch": 2.5294117647058822,
"grad_norm": 0.3921112106668738,
"learning_rate": 3.224342108828798e-05,
"loss": 1.041,
"step": 989
},
{
"epoch": 2.531969309462916,
"grad_norm": 0.3623859088885075,
"learning_rate": 3.2223239497963934e-05,
"loss": 1.0506,
"step": 990
},
{
"epoch": 2.5345268542199486,
"grad_norm": 0.25999017279470166,
"learning_rate": 3.220303802119348e-05,
"loss": 1.0326,
"step": 991
},
{
"epoch": 2.5370843989769822,
"grad_norm": 0.3333560821282555,
"learning_rate": 3.2182816690843165e-05,
"loss": 1.0509,
"step": 992
},
{
"epoch": 2.5396419437340154,
"grad_norm": 0.28544064631438576,
"learning_rate": 3.216257553981183e-05,
"loss": 1.0308,
"step": 993
},
{
"epoch": 2.5421994884910486,
"grad_norm": 0.3546457746118965,
"learning_rate": 3.214231460103055e-05,
"loss": 1.0126,
"step": 994
},
{
"epoch": 2.544757033248082,
"grad_norm": 0.35175717402269296,
"learning_rate": 3.2122033907462614e-05,
"loss": 1.0051,
"step": 995
},
{
"epoch": 2.547314578005115,
"grad_norm": 0.34233801481755133,
"learning_rate": 3.2101733492103443e-05,
"loss": 1.0249,
"step": 996
},
{
"epoch": 2.5498721227621486,
"grad_norm": 0.37474713636308743,
"learning_rate": 3.2081413387980544e-05,
"loss": 1.0263,
"step": 997
},
{
"epoch": 2.5524296675191813,
"grad_norm": 0.3086931923343721,
"learning_rate": 3.2061073628153443e-05,
"loss": 1.0257,
"step": 998
},
{
"epoch": 2.554987212276215,
"grad_norm": 0.3400532972469253,
"learning_rate": 3.204071424571368e-05,
"loss": 1.0194,
"step": 999
},
{
"epoch": 2.557544757033248,
"grad_norm": 0.4487394676938873,
"learning_rate": 3.2020335273784675e-05,
"loss": 1.0223,
"step": 1000
},
{
"epoch": 2.5601023017902813,
"grad_norm": 0.4294386959713973,
"learning_rate": 3.1999936745521764e-05,
"loss": 1.0261,
"step": 1001
},
{
"epoch": 2.5626598465473145,
"grad_norm": 0.3343666920948676,
"learning_rate": 3.1979518694112055e-05,
"loss": 1.0183,
"step": 1002
},
{
"epoch": 2.5652173913043477,
"grad_norm": 0.36057368605026,
"learning_rate": 3.195908115277445e-05,
"loss": 1.0424,
"step": 1003
},
{
"epoch": 2.5677749360613813,
"grad_norm": 0.31769846844741717,
"learning_rate": 3.193862415475956e-05,
"loss": 1.0485,
"step": 1004
},
{
"epoch": 2.5703324808184145,
"grad_norm": 0.3863540621960715,
"learning_rate": 3.1918147733349623e-05,
"loss": 1.0434,
"step": 1005
},
{
"epoch": 2.5728900255754477,
"grad_norm": 0.44896876151143816,
"learning_rate": 3.189765192185852e-05,
"loss": 0.9821,
"step": 1006
},
{
"epoch": 2.575447570332481,
"grad_norm": 0.358319196529918,
"learning_rate": 3.187713675363163e-05,
"loss": 1.0567,
"step": 1007
},
{
"epoch": 2.578005115089514,
"grad_norm": 0.28602159121664567,
"learning_rate": 3.1856602262045866e-05,
"loss": 1.0462,
"step": 1008
},
{
"epoch": 2.580562659846547,
"grad_norm": 0.33739843898479016,
"learning_rate": 3.1836048480509556e-05,
"loss": 1.0442,
"step": 1009
},
{
"epoch": 2.5831202046035804,
"grad_norm": 0.34857955840946614,
"learning_rate": 3.181547544246241e-05,
"loss": 1.026,
"step": 1010
},
{
"epoch": 2.585677749360614,
"grad_norm": 0.3509145527694494,
"learning_rate": 3.179488318137549e-05,
"loss": 1.033,
"step": 1011
},
{
"epoch": 2.588235294117647,
"grad_norm": 0.2703505958300998,
"learning_rate": 3.1774271730751105e-05,
"loss": 1.0358,
"step": 1012
},
{
"epoch": 2.5907928388746804,
"grad_norm": 0.30121591011845916,
"learning_rate": 3.1753641124122805e-05,
"loss": 1.0186,
"step": 1013
},
{
"epoch": 2.5933503836317136,
"grad_norm": 0.3580965839066195,
"learning_rate": 3.17329913950553e-05,
"loss": 1.0371,
"step": 1014
},
{
"epoch": 2.5959079283887467,
"grad_norm": 0.30214909367000026,
"learning_rate": 3.171232257714439e-05,
"loss": 1.017,
"step": 1015
},
{
"epoch": 2.59846547314578,
"grad_norm": 0.31121659495912707,
"learning_rate": 3.169163470401697e-05,
"loss": 1.0161,
"step": 1016
},
{
"epoch": 2.601023017902813,
"grad_norm": 0.2978595190084536,
"learning_rate": 3.16709278093309e-05,
"loss": 1.029,
"step": 1017
},
{
"epoch": 2.6035805626598467,
"grad_norm": 0.3303634417814776,
"learning_rate": 3.1650201926775e-05,
"loss": 1.0325,
"step": 1018
},
{
"epoch": 2.60613810741688,
"grad_norm": 0.3548149730706776,
"learning_rate": 3.1629457090069006e-05,
"loss": 1.0055,
"step": 1019
},
{
"epoch": 2.608695652173913,
"grad_norm": 0.3023258018004769,
"learning_rate": 3.1608693332963456e-05,
"loss": 1.0127,
"step": 1020
},
{
"epoch": 2.6112531969309463,
"grad_norm": 0.32951775793982324,
"learning_rate": 3.158791068923969e-05,
"loss": 1.025,
"step": 1021
},
{
"epoch": 2.6138107416879794,
"grad_norm": 0.33174873111472786,
"learning_rate": 3.156710919270977e-05,
"loss": 1.0183,
"step": 1022
},
{
"epoch": 2.6163682864450126,
"grad_norm": 0.32018868983340326,
"learning_rate": 3.154628887721642e-05,
"loss": 1.0409,
"step": 1023
},
{
"epoch": 2.618925831202046,
"grad_norm": 0.3440503079405171,
"learning_rate": 3.1525449776633e-05,
"loss": 1.0189,
"step": 1024
},
{
"epoch": 2.6214833759590794,
"grad_norm": 0.35118417695246185,
"learning_rate": 3.150459192486342e-05,
"loss": 1.0184,
"step": 1025
},
{
"epoch": 2.6240409207161126,
"grad_norm": 0.2740688133857465,
"learning_rate": 3.148371535584212e-05,
"loss": 1.025,
"step": 1026
},
{
"epoch": 2.626598465473146,
"grad_norm": 0.33857449772861686,
"learning_rate": 3.146282010353394e-05,
"loss": 1.032,
"step": 1027
},
{
"epoch": 2.629156010230179,
"grad_norm": 0.32550068961608813,
"learning_rate": 3.144190620193417e-05,
"loss": 1.04,
"step": 1028
},
{
"epoch": 2.631713554987212,
"grad_norm": 0.2749693403486734,
"learning_rate": 3.142097368506841e-05,
"loss": 1.0415,
"step": 1029
},
{
"epoch": 2.634271099744246,
"grad_norm": 0.34576697335791207,
"learning_rate": 3.140002258699257e-05,
"loss": 1.0286,
"step": 1030
},
{
"epoch": 2.6368286445012785,
"grad_norm": 0.31889362027193385,
"learning_rate": 3.137905294179276e-05,
"loss": 1.0341,
"step": 1031
},
{
"epoch": 2.639386189258312,
"grad_norm": 0.2695973703302828,
"learning_rate": 3.1358064783585296e-05,
"loss": 0.9978,
"step": 1032
},
{
"epoch": 2.6419437340153453,
"grad_norm": 0.2822299051978675,
"learning_rate": 3.1337058146516595e-05,
"loss": 1.0607,
"step": 1033
},
{
"epoch": 2.6445012787723785,
"grad_norm": 0.28303866798707605,
"learning_rate": 3.131603306476313e-05,
"loss": 1.0353,
"step": 1034
},
{
"epoch": 2.6470588235294117,
"grad_norm": 0.29878386565238557,
"learning_rate": 3.129498957253141e-05,
"loss": 1.0406,
"step": 1035
},
{
"epoch": 2.649616368286445,
"grad_norm": 0.300478772982156,
"learning_rate": 3.127392770405787e-05,
"loss": 1.0126,
"step": 1036
},
{
"epoch": 2.6521739130434785,
"grad_norm": 0.314433403781917,
"learning_rate": 3.1252847493608865e-05,
"loss": 1.0181,
"step": 1037
},
{
"epoch": 2.6547314578005117,
"grad_norm": 0.24980230416972485,
"learning_rate": 3.123174897548057e-05,
"loss": 1.0258,
"step": 1038
},
{
"epoch": 2.657289002557545,
"grad_norm": 0.2929645036492049,
"learning_rate": 3.121063218399896e-05,
"loss": 1.0329,
"step": 1039
},
{
"epoch": 2.659846547314578,
"grad_norm": 0.35855658072940827,
"learning_rate": 3.118949715351972e-05,
"loss": 1.0434,
"step": 1040
},
{
"epoch": 2.662404092071611,
"grad_norm": 0.3774594661066989,
"learning_rate": 3.116834391842825e-05,
"loss": 1.0203,
"step": 1041
},
{
"epoch": 2.6649616368286444,
"grad_norm": 0.4020638277985882,
"learning_rate": 3.114717251313952e-05,
"loss": 1.0534,
"step": 1042
},
{
"epoch": 2.6675191815856776,
"grad_norm": 0.25432170611182703,
"learning_rate": 3.112598297209809e-05,
"loss": 0.9873,
"step": 1043
},
{
"epoch": 2.670076726342711,
"grad_norm": 0.2911411839402648,
"learning_rate": 3.110477532977802e-05,
"loss": 1.0306,
"step": 1044
},
{
"epoch": 2.6726342710997444,
"grad_norm": 0.33081432069532596,
"learning_rate": 3.108354962068281e-05,
"loss": 1.0153,
"step": 1045
},
{
"epoch": 2.6751918158567776,
"grad_norm": 0.37331494126622766,
"learning_rate": 3.1062305879345366e-05,
"loss": 1.0007,
"step": 1046
},
{
"epoch": 2.6777493606138107,
"grad_norm": 0.3334210130985932,
"learning_rate": 3.1041044140327925e-05,
"loss": 1.0303,
"step": 1047
},
{
"epoch": 2.680306905370844,
"grad_norm": 0.31468912174046776,
"learning_rate": 3.101976443822202e-05,
"loss": 1.0341,
"step": 1048
},
{
"epoch": 2.682864450127877,
"grad_norm": 0.3785096648157943,
"learning_rate": 3.099846680764838e-05,
"loss": 1.0077,
"step": 1049
},
{
"epoch": 2.6854219948849103,
"grad_norm": 0.4081225981999444,
"learning_rate": 3.097715128325692e-05,
"loss": 1.0297,
"step": 1050
},
{
"epoch": 2.687979539641944,
"grad_norm": 0.3845735808866462,
"learning_rate": 3.095581789972667e-05,
"loss": 1.03,
"step": 1051
},
{
"epoch": 2.690537084398977,
"grad_norm": 0.3323964790307005,
"learning_rate": 3.0934466691765717e-05,
"loss": 1.0259,
"step": 1052
},
{
"epoch": 2.6930946291560103,
"grad_norm": 0.31599895535543004,
"learning_rate": 3.091309769411113e-05,
"loss": 1.0336,
"step": 1053
},
{
"epoch": 2.6956521739130435,
"grad_norm": 0.4390373728726494,
"learning_rate": 3.0891710941528934e-05,
"loss": 0.9985,
"step": 1054
},
{
"epoch": 2.6982097186700766,
"grad_norm": 0.5040811530705346,
"learning_rate": 3.0870306468814045e-05,
"loss": 1.012,
"step": 1055
},
{
"epoch": 2.70076726342711,
"grad_norm": 0.41193294311095624,
"learning_rate": 3.084888431079019e-05,
"loss": 1.0313,
"step": 1056
},
{
"epoch": 2.703324808184143,
"grad_norm": 0.364618078546927,
"learning_rate": 3.0827444502309893e-05,
"loss": 1.034,
"step": 1057
},
{
"epoch": 2.7058823529411766,
"grad_norm": 0.4380736628055229,
"learning_rate": 3.080598707825438e-05,
"loss": 1.0217,
"step": 1058
},
{
"epoch": 2.70843989769821,
"grad_norm": 0.47186442373029863,
"learning_rate": 3.0784512073533535e-05,
"loss": 1.0481,
"step": 1059
},
{
"epoch": 2.710997442455243,
"grad_norm": 0.3421100407629605,
"learning_rate": 3.076301952308585e-05,
"loss": 1.0056,
"step": 1060
},
{
"epoch": 2.713554987212276,
"grad_norm": 0.41290720356538774,
"learning_rate": 3.074150946187836e-05,
"loss": 1.0513,
"step": 1061
},
{
"epoch": 2.7161125319693094,
"grad_norm": 0.4145486411289134,
"learning_rate": 3.071998192490658e-05,
"loss": 1.0365,
"step": 1062
},
{
"epoch": 2.718670076726343,
"grad_norm": 0.4717833554832218,
"learning_rate": 3.069843694719449e-05,
"loss": 1.0727,
"step": 1063
},
{
"epoch": 2.7212276214833757,
"grad_norm": 0.32830135945624245,
"learning_rate": 3.067687456379439e-05,
"loss": 1.0319,
"step": 1064
},
{
"epoch": 2.7237851662404093,
"grad_norm": 0.31333709958852457,
"learning_rate": 3.065529480978696e-05,
"loss": 1.0009,
"step": 1065
},
{
"epoch": 2.7263427109974425,
"grad_norm": 0.3594825854654955,
"learning_rate": 3.063369772028109e-05,
"loss": 1.021,
"step": 1066
},
{
"epoch": 2.7289002557544757,
"grad_norm": 0.3928884507210811,
"learning_rate": 3.06120833304139e-05,
"loss": 1.0295,
"step": 1067
},
{
"epoch": 2.731457800511509,
"grad_norm": 0.4265361695022141,
"learning_rate": 3.059045167535066e-05,
"loss": 1.0475,
"step": 1068
},
{
"epoch": 2.734015345268542,
"grad_norm": 0.41103081466369895,
"learning_rate": 3.056880279028471e-05,
"loss": 1.0228,
"step": 1069
},
{
"epoch": 2.7365728900255757,
"grad_norm": 0.36278826604988385,
"learning_rate": 3.054713671043743e-05,
"loss": 1.0659,
"step": 1070
},
{
"epoch": 2.7391304347826084,
"grad_norm": 0.31147294066159364,
"learning_rate": 3.052545347105819e-05,
"loss": 1.0178,
"step": 1071
},
{
"epoch": 2.741687979539642,
"grad_norm": 0.3301997452385739,
"learning_rate": 3.0503753107424258e-05,
"loss": 1.0466,
"step": 1072
},
{
"epoch": 2.7442455242966752,
"grad_norm": 0.27563971344061766,
"learning_rate": 3.0482035654840758e-05,
"loss": 1.0284,
"step": 1073
},
{
"epoch": 2.7468030690537084,
"grad_norm": 0.2849217210734341,
"learning_rate": 3.046030114864064e-05,
"loss": 1.0171,
"step": 1074
},
{
"epoch": 2.7493606138107416,
"grad_norm": 0.3147294212009837,
"learning_rate": 3.0438549624184594e-05,
"loss": 1.0716,
"step": 1075
},
{
"epoch": 2.7519181585677748,
"grad_norm": 0.30539084667878436,
"learning_rate": 3.0416781116860973e-05,
"loss": 1.0262,
"step": 1076
},
{
"epoch": 2.7544757033248084,
"grad_norm": 0.2762716227457634,
"learning_rate": 3.0394995662085783e-05,
"loss": 1.0454,
"step": 1077
},
{
"epoch": 2.7570332480818416,
"grad_norm": 0.3405352419085539,
"learning_rate": 3.0373193295302597e-05,
"loss": 1.0101,
"step": 1078
},
{
"epoch": 2.7595907928388748,
"grad_norm": 0.2781256683879574,
"learning_rate": 3.0351374051982508e-05,
"loss": 0.9872,
"step": 1079
},
{
"epoch": 2.762148337595908,
"grad_norm": 0.2683963616416243,
"learning_rate": 3.0329537967624062e-05,
"loss": 1.0349,
"step": 1080
},
{
"epoch": 2.764705882352941,
"grad_norm": 0.32788496469480943,
"learning_rate": 3.030768507775319e-05,
"loss": 0.995,
"step": 1081
},
{
"epoch": 2.7672634271099743,
"grad_norm": 0.30589363777855844,
"learning_rate": 3.0285815417923188e-05,
"loss": 1.0315,
"step": 1082
},
{
"epoch": 2.7698209718670075,
"grad_norm": 0.2902794941527518,
"learning_rate": 3.0263929023714615e-05,
"loss": 1.0333,
"step": 1083
},
{
"epoch": 2.772378516624041,
"grad_norm": 0.27983178134129,
"learning_rate": 3.024202593073528e-05,
"loss": 1.0352,
"step": 1084
},
{
"epoch": 2.7749360613810743,
"grad_norm": 0.43231713661750476,
"learning_rate": 3.0220106174620133e-05,
"loss": 1.0105,
"step": 1085
},
{
"epoch": 2.7774936061381075,
"grad_norm": 0.41638636225917003,
"learning_rate": 3.0198169791031266e-05,
"loss": 1.063,
"step": 1086
},
{
"epoch": 2.7800511508951407,
"grad_norm": 0.2848807068316711,
"learning_rate": 3.0176216815657774e-05,
"loss": 0.9861,
"step": 1087
},
{
"epoch": 2.782608695652174,
"grad_norm": 0.2737423773735809,
"learning_rate": 3.0154247284215802e-05,
"loss": 1.013,
"step": 1088
},
{
"epoch": 2.785166240409207,
"grad_norm": 0.30223181565809215,
"learning_rate": 3.0132261232448388e-05,
"loss": 1.0355,
"step": 1089
},
{
"epoch": 2.78772378516624,
"grad_norm": 0.3354652437056132,
"learning_rate": 3.0110258696125466e-05,
"loss": 1.0042,
"step": 1090
},
{
"epoch": 2.790281329923274,
"grad_norm": 0.3333357227145358,
"learning_rate": 3.0088239711043797e-05,
"loss": 0.9789,
"step": 1091
},
{
"epoch": 2.792838874680307,
"grad_norm": 0.3232301816307233,
"learning_rate": 3.0066204313026885e-05,
"loss": 1.0393,
"step": 1092
},
{
"epoch": 2.79539641943734,
"grad_norm": 0.3423032506796668,
"learning_rate": 3.0044152537924948e-05,
"loss": 1.0404,
"step": 1093
},
{
"epoch": 2.7979539641943734,
"grad_norm": 0.27189725542176,
"learning_rate": 3.0022084421614832e-05,
"loss": 1.0256,
"step": 1094
},
{
"epoch": 2.8005115089514065,
"grad_norm": 0.27279646077432573,
"learning_rate": 3.0000000000000004e-05,
"loss": 1.0203,
"step": 1095
},
{
"epoch": 2.80306905370844,
"grad_norm": 0.2769934932225626,
"learning_rate": 2.9977899309010423e-05,
"loss": 1.0702,
"step": 1096
},
{
"epoch": 2.805626598465473,
"grad_norm": 0.3268249455928128,
"learning_rate": 2.9955782384602533e-05,
"loss": 1.0374,
"step": 1097
},
{
"epoch": 2.8081841432225065,
"grad_norm": 0.34238777711967283,
"learning_rate": 2.993364926275919e-05,
"loss": 1.0517,
"step": 1098
},
{
"epoch": 2.8107416879795397,
"grad_norm": 0.3204841187721774,
"learning_rate": 2.99114999794896e-05,
"loss": 1.038,
"step": 1099
},
{
"epoch": 2.813299232736573,
"grad_norm": 0.34347961192000986,
"learning_rate": 2.9889334570829257e-05,
"loss": 1.0159,
"step": 1100
},
{
"epoch": 2.815856777493606,
"grad_norm": 0.3317103877761027,
"learning_rate": 2.9867153072839903e-05,
"loss": 1.044,
"step": 1101
},
{
"epoch": 2.8184143222506393,
"grad_norm": 0.2879072677025752,
"learning_rate": 2.984495552160944e-05,
"loss": 1.0235,
"step": 1102
},
{
"epoch": 2.820971867007673,
"grad_norm": 0.2669347109236764,
"learning_rate": 2.9822741953251902e-05,
"loss": 1.0399,
"step": 1103
},
{
"epoch": 2.8235294117647056,
"grad_norm": 0.24592239142015593,
"learning_rate": 2.9800512403907373e-05,
"loss": 1.0152,
"step": 1104
},
{
"epoch": 2.8260869565217392,
"grad_norm": 0.250370053241817,
"learning_rate": 2.977826690974194e-05,
"loss": 1.0054,
"step": 1105
},
{
"epoch": 2.8286445012787724,
"grad_norm": 0.3276149908684206,
"learning_rate": 2.9756005506947628e-05,
"loss": 1.0344,
"step": 1106
},
{
"epoch": 2.8312020460358056,
"grad_norm": 0.3176615860514685,
"learning_rate": 2.973372823174236e-05,
"loss": 1.0302,
"step": 1107
},
{
"epoch": 2.833759590792839,
"grad_norm": 0.22420502657862215,
"learning_rate": 2.971143512036985e-05,
"loss": 1.0421,
"step": 1108
},
{
"epoch": 2.836317135549872,
"grad_norm": 0.257691804926265,
"learning_rate": 2.9689126209099615e-05,
"loss": 1.0452,
"step": 1109
},
{
"epoch": 2.8388746803069056,
"grad_norm": 0.2787031390424134,
"learning_rate": 2.966680153422685e-05,
"loss": 1.0273,
"step": 1110
},
{
"epoch": 2.8414322250639388,
"grad_norm": 0.29248480770836066,
"learning_rate": 2.96444611320724e-05,
"loss": 1.0277,
"step": 1111
},
{
"epoch": 2.843989769820972,
"grad_norm": 0.21675448901362926,
"learning_rate": 2.9622105038982715e-05,
"loss": 1.0371,
"step": 1112
},
{
"epoch": 2.846547314578005,
"grad_norm": 0.32868930558298726,
"learning_rate": 2.959973329132975e-05,
"loss": 1.025,
"step": 1113
},
{
"epoch": 2.8491048593350383,
"grad_norm": 0.39257087154953485,
"learning_rate": 2.9577345925510948e-05,
"loss": 1.023,
"step": 1114
},
{
"epoch": 2.8516624040920715,
"grad_norm": 0.3338238404066049,
"learning_rate": 2.9554942977949145e-05,
"loss": 1.0315,
"step": 1115
},
{
"epoch": 2.8542199488491047,
"grad_norm": 0.23129269445735415,
"learning_rate": 2.953252448509254e-05,
"loss": 1.0226,
"step": 1116
},
{
"epoch": 2.8567774936061383,
"grad_norm": 0.3137815601480366,
"learning_rate": 2.951009048341462e-05,
"loss": 1.0083,
"step": 1117
},
{
"epoch": 2.8593350383631715,
"grad_norm": 0.2757576513083469,
"learning_rate": 2.9487641009414108e-05,
"loss": 1.036,
"step": 1118
},
{
"epoch": 2.8618925831202047,
"grad_norm": 0.2968472409431522,
"learning_rate": 2.9465176099614884e-05,
"loss": 1.0103,
"step": 1119
},
{
"epoch": 2.864450127877238,
"grad_norm": 0.2570561810687309,
"learning_rate": 2.9442695790565955e-05,
"loss": 1.0684,
"step": 1120
},
{
"epoch": 2.867007672634271,
"grad_norm": 0.23768883327358178,
"learning_rate": 2.9420200118841386e-05,
"loss": 1.033,
"step": 1121
},
{
"epoch": 2.869565217391304,
"grad_norm": 0.2605177771848745,
"learning_rate": 2.9397689121040213e-05,
"loss": 1.0163,
"step": 1122
},
{
"epoch": 2.8721227621483374,
"grad_norm": 0.28576880527653575,
"learning_rate": 2.9375162833786435e-05,
"loss": 0.9875,
"step": 1123
},
{
"epoch": 2.874680306905371,
"grad_norm": 0.2742174274441978,
"learning_rate": 2.935262129372891e-05,
"loss": 1.0675,
"step": 1124
},
{
"epoch": 2.877237851662404,
"grad_norm": 0.2176000129273426,
"learning_rate": 2.9330064537541312e-05,
"loss": 1.0417,
"step": 1125
},
{
"epoch": 2.8797953964194374,
"grad_norm": 0.25530271559505907,
"learning_rate": 2.9307492601922075e-05,
"loss": 1.0371,
"step": 1126
},
{
"epoch": 2.8823529411764706,
"grad_norm": 0.25559105596494947,
"learning_rate": 2.9284905523594333e-05,
"loss": 1.0037,
"step": 1127
},
{
"epoch": 2.8849104859335037,
"grad_norm": 0.24325879489115165,
"learning_rate": 2.9262303339305834e-05,
"loss": 1.0101,
"step": 1128
},
{
"epoch": 2.887468030690537,
"grad_norm": 0.2547855688294357,
"learning_rate": 2.9239686085828933e-05,
"loss": 1.0139,
"step": 1129
},
{
"epoch": 2.89002557544757,
"grad_norm": 0.25736340993894663,
"learning_rate": 2.9217053799960477e-05,
"loss": 1.0385,
"step": 1130
},
{
"epoch": 2.8925831202046037,
"grad_norm": 0.27470797917303247,
"learning_rate": 2.9194406518521786e-05,
"loss": 1.0498,
"step": 1131
},
{
"epoch": 2.895140664961637,
"grad_norm": 0.25609194319865586,
"learning_rate": 2.9171744278358568e-05,
"loss": 0.987,
"step": 1132
},
{
"epoch": 2.89769820971867,
"grad_norm": 0.32083054500836145,
"learning_rate": 2.914906711634088e-05,
"loss": 1.0105,
"step": 1133
},
{
"epoch": 2.9002557544757033,
"grad_norm": 0.3387414632497515,
"learning_rate": 2.912637506936304e-05,
"loss": 1.0106,
"step": 1134
},
{
"epoch": 2.9028132992327365,
"grad_norm": 0.23634828743054007,
"learning_rate": 2.910366817434358e-05,
"loss": 1.0339,
"step": 1135
},
{
"epoch": 2.90537084398977,
"grad_norm": 0.26267061578943923,
"learning_rate": 2.9080946468225214e-05,
"loss": 1.0076,
"step": 1136
},
{
"epoch": 2.907928388746803,
"grad_norm": 0.33959297470156813,
"learning_rate": 2.9058209987974727e-05,
"loss": 1.0686,
"step": 1137
},
{
"epoch": 2.9104859335038364,
"grad_norm": 0.3409288928963959,
"learning_rate": 2.9035458770582957e-05,
"loss": 1.0047,
"step": 1138
},
{
"epoch": 2.9130434782608696,
"grad_norm": 0.29376753364345026,
"learning_rate": 2.9012692853064712e-05,
"loss": 1.0235,
"step": 1139
},
{
"epoch": 2.915601023017903,
"grad_norm": 0.27706936171216345,
"learning_rate": 2.8989912272458713e-05,
"loss": 1.0162,
"step": 1140
},
{
"epoch": 2.918158567774936,
"grad_norm": 0.3003940174076924,
"learning_rate": 2.896711706582754e-05,
"loss": 1.0297,
"step": 1141
},
{
"epoch": 2.920716112531969,
"grad_norm": 0.275991442696101,
"learning_rate": 2.8944307270257574e-05,
"loss": 1.04,
"step": 1142
},
{
"epoch": 2.923273657289003,
"grad_norm": 0.3005799680840971,
"learning_rate": 2.8921482922858915e-05,
"loss": 1.0399,
"step": 1143
},
{
"epoch": 2.9258312020460355,
"grad_norm": 0.33298520369377627,
"learning_rate": 2.8898644060765363e-05,
"loss": 1.0129,
"step": 1144
},
{
"epoch": 2.928388746803069,
"grad_norm": 0.356826250282169,
"learning_rate": 2.8875790721134312e-05,
"loss": 1.0218,
"step": 1145
},
{
"epoch": 2.9309462915601023,
"grad_norm": 0.3514210330691824,
"learning_rate": 2.8852922941146717e-05,
"loss": 1.011,
"step": 1146
},
{
"epoch": 2.9335038363171355,
"grad_norm": 0.28551238674571655,
"learning_rate": 2.8830040758007026e-05,
"loss": 1.0304,
"step": 1147
},
{
"epoch": 2.9360613810741687,
"grad_norm": 0.40598714937616354,
"learning_rate": 2.8807144208943117e-05,
"loss": 1.0262,
"step": 1148
},
{
"epoch": 2.938618925831202,
"grad_norm": 0.44197720286280806,
"learning_rate": 2.8784233331206245e-05,
"loss": 1.0429,
"step": 1149
},
{
"epoch": 2.9411764705882355,
"grad_norm": 0.2807856045804052,
"learning_rate": 2.8761308162070983e-05,
"loss": 1.0321,
"step": 1150
},
{
"epoch": 2.9437340153452687,
"grad_norm": 0.28981428483349336,
"learning_rate": 2.8738368738835146e-05,
"loss": 1.0448,
"step": 1151
},
{
"epoch": 2.946291560102302,
"grad_norm": 0.4281228928351924,
"learning_rate": 2.871541509881973e-05,
"loss": 1.0155,
"step": 1152
},
{
"epoch": 2.948849104859335,
"grad_norm": 0.3821568053436914,
"learning_rate": 2.869244727936889e-05,
"loss": 1.0383,
"step": 1153
},
{
"epoch": 2.9514066496163682,
"grad_norm": 0.3414471004435733,
"learning_rate": 2.8669465317849813e-05,
"loss": 1.0438,
"step": 1154
},
{
"epoch": 2.9539641943734014,
"grad_norm": 0.37130883161202133,
"learning_rate": 2.8646469251652726e-05,
"loss": 1.058,
"step": 1155
},
{
"epoch": 2.9565217391304346,
"grad_norm": 0.41156636505054933,
"learning_rate": 2.8623459118190782e-05,
"loss": 1.0337,
"step": 1156
},
{
"epoch": 2.959079283887468,
"grad_norm": 0.41565674281527526,
"learning_rate": 2.860043495490004e-05,
"loss": 1.0257,
"step": 1157
},
{
"epoch": 2.9616368286445014,
"grad_norm": 0.40871449911279334,
"learning_rate": 2.8577396799239365e-05,
"loss": 1.0185,
"step": 1158
},
{
"epoch": 2.9641943734015346,
"grad_norm": 0.2849309324589542,
"learning_rate": 2.8554344688690406e-05,
"loss": 1.0185,
"step": 1159
},
{
"epoch": 2.9667519181585678,
"grad_norm": 0.47729726597621686,
"learning_rate": 2.8531278660757502e-05,
"loss": 1.0393,
"step": 1160
},
{
"epoch": 2.969309462915601,
"grad_norm": 0.4614042987367875,
"learning_rate": 2.8508198752967628e-05,
"loss": 1.0081,
"step": 1161
},
{
"epoch": 2.971867007672634,
"grad_norm": 0.3718647311616514,
"learning_rate": 2.848510500287036e-05,
"loss": 1.0152,
"step": 1162
},
{
"epoch": 2.9744245524296673,
"grad_norm": 0.29753291832038586,
"learning_rate": 2.846199744803778e-05,
"loss": 1.0576,
"step": 1163
},
{
"epoch": 2.976982097186701,
"grad_norm": 0.3924125306306016,
"learning_rate": 2.8438876126064446e-05,
"loss": 1.0096,
"step": 1164
},
{
"epoch": 2.979539641943734,
"grad_norm": 0.4446013815402128,
"learning_rate": 2.841574107456729e-05,
"loss": 1.0224,
"step": 1165
},
{
"epoch": 2.9820971867007673,
"grad_norm": 0.3874020659909739,
"learning_rate": 2.8392592331185608e-05,
"loss": 1.0498,
"step": 1166
},
{
"epoch": 2.9846547314578005,
"grad_norm": 0.28410943497320607,
"learning_rate": 2.8369429933580943e-05,
"loss": 1.027,
"step": 1167
},
{
"epoch": 2.9872122762148337,
"grad_norm": 0.33967133885925593,
"learning_rate": 2.8346253919437084e-05,
"loss": 1.0437,
"step": 1168
},
{
"epoch": 2.9897698209718673,
"grad_norm": 0.40354986078593685,
"learning_rate": 2.832306432645994e-05,
"loss": 1.0168,
"step": 1169
},
{
"epoch": 2.9923273657289,
"grad_norm": 0.277670280736742,
"learning_rate": 2.8299861192377537e-05,
"loss": 1.0143,
"step": 1170
},
{
"epoch": 2.9948849104859336,
"grad_norm": 0.2660098824438029,
"learning_rate": 2.8276644554939924e-05,
"loss": 1.0107,
"step": 1171
},
{
"epoch": 2.997442455242967,
"grad_norm": 0.2950569911059824,
"learning_rate": 2.825341445191912e-05,
"loss": 1.0553,
"step": 1172
},
{
"epoch": 3.0,
"grad_norm": 0.3188511927228667,
"learning_rate": 2.823017092110905e-05,
"loss": 1.0467,
"step": 1173
},
{
"epoch": 3.002557544757033,
"grad_norm": 0.3465579988802434,
"learning_rate": 2.820691400032548e-05,
"loss": 1.0361,
"step": 1174
},
{
"epoch": 3.0051150895140664,
"grad_norm": 0.26005864548914087,
"learning_rate": 2.8183643727405966e-05,
"loss": 1.0344,
"step": 1175
},
{
"epoch": 3.0076726342710995,
"grad_norm": 0.27896249482968427,
"learning_rate": 2.8160360140209794e-05,
"loss": 1.0133,
"step": 1176
},
{
"epoch": 3.010230179028133,
"grad_norm": 0.32842298240082546,
"learning_rate": 2.81370632766179e-05,
"loss": 1.055,
"step": 1177
},
{
"epoch": 3.0127877237851663,
"grad_norm": 0.261456599862609,
"learning_rate": 2.8113753174532826e-05,
"loss": 1.0386,
"step": 1178
},
{
"epoch": 3.0153452685421995,
"grad_norm": 0.25749632862942085,
"learning_rate": 2.8090429871878647e-05,
"loss": 0.9899,
"step": 1179
},
{
"epoch": 3.0179028132992327,
"grad_norm": 0.28270560235783554,
"learning_rate": 2.806709340660093e-05,
"loss": 0.9978,
"step": 1180
},
{
"epoch": 3.020460358056266,
"grad_norm": 0.2649247738141889,
"learning_rate": 2.804374381666662e-05,
"loss": 1.0118,
"step": 1181
},
{
"epoch": 3.023017902813299,
"grad_norm": 0.29117414751237736,
"learning_rate": 2.802038114006407e-05,
"loss": 1.0075,
"step": 1182
},
{
"epoch": 3.0255754475703327,
"grad_norm": 0.28295961597799246,
"learning_rate": 2.7997005414802873e-05,
"loss": 1.0277,
"step": 1183
},
{
"epoch": 3.028132992327366,
"grad_norm": 0.24700574471226935,
"learning_rate": 2.797361667891388e-05,
"loss": 0.9938,
"step": 1184
},
{
"epoch": 3.030690537084399,
"grad_norm": 0.2852782791330361,
"learning_rate": 2.7950214970449104e-05,
"loss": 1.0089,
"step": 1185
},
{
"epoch": 3.0332480818414322,
"grad_norm": 0.29813911544916116,
"learning_rate": 2.7926800327481655e-05,
"loss": 0.9896,
"step": 1186
},
{
"epoch": 3.0358056265984654,
"grad_norm": 0.26709949962373675,
"learning_rate": 2.79033727881057e-05,
"loss": 1.0161,
"step": 1187
},
{
"epoch": 3.0383631713554986,
"grad_norm": 0.24668737784436573,
"learning_rate": 2.7879932390436376e-05,
"loss": 0.9788,
"step": 1188
},
{
"epoch": 3.040920716112532,
"grad_norm": 0.26602996861807326,
"learning_rate": 2.7856479172609745e-05,
"loss": 1.0142,
"step": 1189
},
{
"epoch": 3.0434782608695654,
"grad_norm": 0.2681256525652391,
"learning_rate": 2.783301317278273e-05,
"loss": 1.0036,
"step": 1190
},
{
"epoch": 3.0460358056265986,
"grad_norm": 0.281612488954363,
"learning_rate": 2.780953442913304e-05,
"loss": 0.9806,
"step": 1191
},
{
"epoch": 3.0485933503836318,
"grad_norm": 0.3505036641543003,
"learning_rate": 2.7786042979859127e-05,
"loss": 1.021,
"step": 1192
},
{
"epoch": 3.051150895140665,
"grad_norm": 0.2747489614975718,
"learning_rate": 2.7762538863180104e-05,
"loss": 1.0274,
"step": 1193
},
{
"epoch": 3.053708439897698,
"grad_norm": 0.23490353408491924,
"learning_rate": 2.773902211733571e-05,
"loss": 1.0001,
"step": 1194
},
{
"epoch": 3.0562659846547313,
"grad_norm": 0.28016312637880986,
"learning_rate": 2.771549278058621e-05,
"loss": 1.0055,
"step": 1195
},
{
"epoch": 3.0588235294117645,
"grad_norm": 0.280938697462226,
"learning_rate": 2.7691950891212368e-05,
"loss": 1.0195,
"step": 1196
},
{
"epoch": 3.061381074168798,
"grad_norm": 0.23857869209680105,
"learning_rate": 2.766839648751536e-05,
"loss": 1.0163,
"step": 1197
},
{
"epoch": 3.0639386189258313,
"grad_norm": 0.26204364870347474,
"learning_rate": 2.764482960781674e-05,
"loss": 1.0156,
"step": 1198
},
{
"epoch": 3.0664961636828645,
"grad_norm": 0.23886064848371663,
"learning_rate": 2.7621250290458347e-05,
"loss": 1.0323,
"step": 1199
},
{
"epoch": 3.0690537084398977,
"grad_norm": 0.2563194363976881,
"learning_rate": 2.7597658573802243e-05,
"loss": 0.9923,
"step": 1200
},
{
"epoch": 3.071611253196931,
"grad_norm": 0.2883302385934852,
"learning_rate": 2.7574054496230683e-05,
"loss": 1.0052,
"step": 1201
},
{
"epoch": 3.074168797953964,
"grad_norm": 0.2907041925619475,
"learning_rate": 2.7550438096146022e-05,
"loss": 1.0276,
"step": 1202
},
{
"epoch": 3.0767263427109977,
"grad_norm": 0.23933246763179727,
"learning_rate": 2.752680941197068e-05,
"loss": 1.0322,
"step": 1203
},
{
"epoch": 3.079283887468031,
"grad_norm": 0.2830386855523637,
"learning_rate": 2.7503168482147034e-05,
"loss": 1.0333,
"step": 1204
},
{
"epoch": 3.081841432225064,
"grad_norm": 0.23823606531998417,
"learning_rate": 2.7479515345137406e-05,
"loss": 1.0323,
"step": 1205
},
{
"epoch": 3.084398976982097,
"grad_norm": 0.24046955360343872,
"learning_rate": 2.7455850039423962e-05,
"loss": 1.0033,
"step": 1206
},
{
"epoch": 3.0869565217391304,
"grad_norm": 0.2746362063596697,
"learning_rate": 2.7432172603508686e-05,
"loss": 1.0243,
"step": 1207
},
{
"epoch": 3.0895140664961636,
"grad_norm": 0.25917469821293737,
"learning_rate": 2.7408483075913285e-05,
"loss": 1.0505,
"step": 1208
},
{
"epoch": 3.0920716112531967,
"grad_norm": 0.2433543657721417,
"learning_rate": 2.7384781495179127e-05,
"loss": 0.9982,
"step": 1209
},
{
"epoch": 3.0946291560102304,
"grad_norm": 0.26064559625253925,
"learning_rate": 2.7361067899867215e-05,
"loss": 1.0202,
"step": 1210
},
{
"epoch": 3.0971867007672635,
"grad_norm": 0.26578439287138117,
"learning_rate": 2.7337342328558078e-05,
"loss": 1.0219,
"step": 1211
},
{
"epoch": 3.0997442455242967,
"grad_norm": 0.2728863223482901,
"learning_rate": 2.7313604819851748e-05,
"loss": 0.9848,
"step": 1212
},
{
"epoch": 3.10230179028133,
"grad_norm": 0.21444045547718613,
"learning_rate": 2.7289855412367648e-05,
"loss": 1.0554,
"step": 1213
},
{
"epoch": 3.104859335038363,
"grad_norm": 0.2576501077446558,
"learning_rate": 2.7266094144744602e-05,
"loss": 1.0019,
"step": 1214
},
{
"epoch": 3.1074168797953963,
"grad_norm": 0.26949546760431214,
"learning_rate": 2.7242321055640685e-05,
"loss": 1.0446,
"step": 1215
},
{
"epoch": 3.10997442455243,
"grad_norm": 0.22328126171378793,
"learning_rate": 2.7218536183733243e-05,
"loss": 1.0006,
"step": 1216
},
{
"epoch": 3.112531969309463,
"grad_norm": 0.25342267846631883,
"learning_rate": 2.719473956771877e-05,
"loss": 1.0341,
"step": 1217
},
{
"epoch": 3.1150895140664963,
"grad_norm": 0.29087042190737855,
"learning_rate": 2.7170931246312867e-05,
"loss": 1.007,
"step": 1218
},
{
"epoch": 3.1176470588235294,
"grad_norm": 0.2570860397910995,
"learning_rate": 2.7147111258250195e-05,
"loss": 1.0237,
"step": 1219
},
{
"epoch": 3.1202046035805626,
"grad_norm": 0.23243232335156347,
"learning_rate": 2.7123279642284373e-05,
"loss": 1.0129,
"step": 1220
},
{
"epoch": 3.122762148337596,
"grad_norm": 0.21372237677978434,
"learning_rate": 2.7099436437187963e-05,
"loss": 1.0083,
"step": 1221
},
{
"epoch": 3.125319693094629,
"grad_norm": 0.2454112730798479,
"learning_rate": 2.7075581681752357e-05,
"loss": 0.9583,
"step": 1222
},
{
"epoch": 3.1278772378516626,
"grad_norm": 0.28376734542222376,
"learning_rate": 2.7051715414787757e-05,
"loss": 1.0202,
"step": 1223
},
{
"epoch": 3.130434782608696,
"grad_norm": 0.21760384063057203,
"learning_rate": 2.7027837675123088e-05,
"loss": 0.9905,
"step": 1224
},
{
"epoch": 3.132992327365729,
"grad_norm": 0.23090610817033624,
"learning_rate": 2.7003948501605934e-05,
"loss": 0.9947,
"step": 1225
},
{
"epoch": 3.135549872122762,
"grad_norm": 0.2588561823757637,
"learning_rate": 2.6980047933102497e-05,
"loss": 1.0122,
"step": 1226
},
{
"epoch": 3.1381074168797953,
"grad_norm": 0.3668657055433508,
"learning_rate": 2.6956136008497495e-05,
"loss": 1.011,
"step": 1227
},
{
"epoch": 3.1406649616368285,
"grad_norm": 0.32076098037586065,
"learning_rate": 2.693221276669414e-05,
"loss": 1.0302,
"step": 1228
},
{
"epoch": 3.1432225063938617,
"grad_norm": 0.23820970663533267,
"learning_rate": 2.6908278246614047e-05,
"loss": 1.0354,
"step": 1229
},
{
"epoch": 3.1457800511508953,
"grad_norm": 0.3543999426173227,
"learning_rate": 2.6884332487197197e-05,
"loss": 1.0212,
"step": 1230
},
{
"epoch": 3.1483375959079285,
"grad_norm": 0.33454500636512646,
"learning_rate": 2.686037552740183e-05,
"loss": 1.0284,
"step": 1231
},
{
"epoch": 3.1508951406649617,
"grad_norm": 0.2731580262255819,
"learning_rate": 2.683640740620442e-05,
"loss": 1.0082,
"step": 1232
},
{
"epoch": 3.153452685421995,
"grad_norm": 0.28605907387048835,
"learning_rate": 2.6812428162599613e-05,
"loss": 1.0474,
"step": 1233
},
{
"epoch": 3.156010230179028,
"grad_norm": 0.32001188674246095,
"learning_rate": 2.6788437835600127e-05,
"loss": 1.0285,
"step": 1234
},
{
"epoch": 3.1585677749360612,
"grad_norm": 0.25216645409492455,
"learning_rate": 2.6764436464236735e-05,
"loss": 1.0168,
"step": 1235
},
{
"epoch": 3.1611253196930944,
"grad_norm": 0.27148523945779013,
"learning_rate": 2.6740424087558156e-05,
"loss": 1.0225,
"step": 1236
},
{
"epoch": 3.163682864450128,
"grad_norm": 0.24623424939802263,
"learning_rate": 2.671640074463104e-05,
"loss": 1.013,
"step": 1237
},
{
"epoch": 3.166240409207161,
"grad_norm": 0.2558876015634447,
"learning_rate": 2.669236647453985e-05,
"loss": 1.015,
"step": 1238
},
{
"epoch": 3.1687979539641944,
"grad_norm": 0.2384429157026653,
"learning_rate": 2.666832131638685e-05,
"loss": 0.9808,
"step": 1239
},
{
"epoch": 3.1713554987212276,
"grad_norm": 0.26881517135856464,
"learning_rate": 2.6644265309292006e-05,
"loss": 1.0059,
"step": 1240
},
{
"epoch": 3.1739130434782608,
"grad_norm": 0.2661004928343902,
"learning_rate": 2.662019849239293e-05,
"loss": 0.997,
"step": 1241
},
{
"epoch": 3.176470588235294,
"grad_norm": 0.31273625345339795,
"learning_rate": 2.6596120904844842e-05,
"loss": 1.0017,
"step": 1242
},
{
"epoch": 3.1790281329923276,
"grad_norm": 0.2763431416816016,
"learning_rate": 2.6572032585820458e-05,
"loss": 1.0074,
"step": 1243
},
{
"epoch": 3.1815856777493607,
"grad_norm": 0.19959690279743475,
"learning_rate": 2.654793357450998e-05,
"loss": 0.9939,
"step": 1244
},
{
"epoch": 3.184143222506394,
"grad_norm": 0.2806692461477797,
"learning_rate": 2.6523823910120986e-05,
"loss": 0.9689,
"step": 1245
},
{
"epoch": 3.186700767263427,
"grad_norm": 0.3114396227227017,
"learning_rate": 2.6499703631878394e-05,
"loss": 1.0213,
"step": 1246
},
{
"epoch": 3.1892583120204603,
"grad_norm": 0.21393961329030795,
"learning_rate": 2.6475572779024382e-05,
"loss": 1.0286,
"step": 1247
},
{
"epoch": 3.1918158567774935,
"grad_norm": 0.2989195229499343,
"learning_rate": 2.6451431390818343e-05,
"loss": 1.0121,
"step": 1248
},
{
"epoch": 3.1943734015345266,
"grad_norm": 0.33928650026272444,
"learning_rate": 2.642727950653681e-05,
"loss": 1.0004,
"step": 1249
},
{
"epoch": 3.1969309462915603,
"grad_norm": 0.27702969437760705,
"learning_rate": 2.6403117165473386e-05,
"loss": 1.0096,
"step": 1250
},
{
"epoch": 3.1994884910485935,
"grad_norm": 0.26857956234275654,
"learning_rate": 2.6378944406938693e-05,
"loss": 0.9885,
"step": 1251
},
{
"epoch": 3.2020460358056266,
"grad_norm": 0.2413697222713823,
"learning_rate": 2.635476127026029e-05,
"loss": 1.0229,
"step": 1252
},
{
"epoch": 3.20460358056266,
"grad_norm": 0.3137737079157682,
"learning_rate": 2.633056779478263e-05,
"loss": 0.9842,
"step": 1253
},
{
"epoch": 3.207161125319693,
"grad_norm": 0.3256111903028223,
"learning_rate": 2.630636401986699e-05,
"loss": 1.0066,
"step": 1254
},
{
"epoch": 3.209718670076726,
"grad_norm": 0.21050448469654487,
"learning_rate": 2.628214998489139e-05,
"loss": 1.0065,
"step": 1255
},
{
"epoch": 3.21227621483376,
"grad_norm": 0.26214504371022174,
"learning_rate": 2.6257925729250563e-05,
"loss": 1.0004,
"step": 1256
},
{
"epoch": 3.214833759590793,
"grad_norm": 0.27808124446663696,
"learning_rate": 2.6233691292355846e-05,
"loss": 1.0044,
"step": 1257
},
{
"epoch": 3.217391304347826,
"grad_norm": 0.20926694725239744,
"learning_rate": 2.6209446713635164e-05,
"loss": 0.9985,
"step": 1258
},
{
"epoch": 3.2199488491048593,
"grad_norm": 0.23358927415865904,
"learning_rate": 2.618519203253292e-05,
"loss": 1.0465,
"step": 1259
},
{
"epoch": 3.2225063938618925,
"grad_norm": 0.2490544026035825,
"learning_rate": 2.6160927288509972e-05,
"loss": 1.016,
"step": 1260
},
{
"epoch": 3.2250639386189257,
"grad_norm": 0.23291930906613517,
"learning_rate": 2.6136652521043538e-05,
"loss": 0.9968,
"step": 1261
},
{
"epoch": 3.227621483375959,
"grad_norm": 0.2233307849320106,
"learning_rate": 2.6112367769627137e-05,
"loss": 0.9776,
"step": 1262
},
{
"epoch": 3.2301790281329925,
"grad_norm": 0.19641474386934185,
"learning_rate": 2.6088073073770567e-05,
"loss": 1.002,
"step": 1263
},
{
"epoch": 3.2327365728900257,
"grad_norm": 0.23326552455206,
"learning_rate": 2.6063768472999754e-05,
"loss": 1.0071,
"step": 1264
},
{
"epoch": 3.235294117647059,
"grad_norm": 0.2612917869030843,
"learning_rate": 2.6039454006856774e-05,
"loss": 1.0042,
"step": 1265
},
{
"epoch": 3.237851662404092,
"grad_norm": 0.27580568206648937,
"learning_rate": 2.601512971489974e-05,
"loss": 1.0068,
"step": 1266
},
{
"epoch": 3.2404092071611252,
"grad_norm": 0.2447849159729633,
"learning_rate": 2.5990795636702762e-05,
"loss": 0.9989,
"step": 1267
},
{
"epoch": 3.2429667519181584,
"grad_norm": 0.3214335918467597,
"learning_rate": 2.5966451811855847e-05,
"loss": 0.9643,
"step": 1268
},
{
"epoch": 3.2455242966751916,
"grad_norm": 0.31754029590347954,
"learning_rate": 2.5942098279964895e-05,
"loss": 1.0237,
"step": 1269
},
{
"epoch": 3.2480818414322252,
"grad_norm": 0.2758769397849654,
"learning_rate": 2.591773508065157e-05,
"loss": 1.0,
"step": 1270
},
{
"epoch": 3.2506393861892584,
"grad_norm": 0.31104718348852967,
"learning_rate": 2.589336225355328e-05,
"loss": 1.018,
"step": 1271
},
{
"epoch": 3.2531969309462916,
"grad_norm": 0.25672543976063106,
"learning_rate": 2.5868979838323082e-05,
"loss": 1.0007,
"step": 1272
},
{
"epoch": 3.2557544757033248,
"grad_norm": 0.29581853427049365,
"learning_rate": 2.584458787462964e-05,
"loss": 0.9987,
"step": 1273
},
{
"epoch": 3.258312020460358,
"grad_norm": 0.2845616080900404,
"learning_rate": 2.5820186402157168e-05,
"loss": 1.0121,
"step": 1274
},
{
"epoch": 3.260869565217391,
"grad_norm": 0.24715206377476634,
"learning_rate": 2.579577546060532e-05,
"loss": 1.0222,
"step": 1275
},
{
"epoch": 3.2634271099744243,
"grad_norm": 0.3161634596169807,
"learning_rate": 2.5771355089689184e-05,
"loss": 0.9791,
"step": 1276
},
{
"epoch": 3.265984654731458,
"grad_norm": 0.32065763178527046,
"learning_rate": 2.5746925329139166e-05,
"loss": 1.0343,
"step": 1277
},
{
"epoch": 3.268542199488491,
"grad_norm": 0.2588603962172159,
"learning_rate": 2.572248621870097e-05,
"loss": 0.9994,
"step": 1278
},
{
"epoch": 3.2710997442455243,
"grad_norm": 0.27339335287987476,
"learning_rate": 2.5698037798135492e-05,
"loss": 1.0497,
"step": 1279
},
{
"epoch": 3.2736572890025575,
"grad_norm": 0.3190511262163905,
"learning_rate": 2.5673580107218792e-05,
"loss": 1.0185,
"step": 1280
},
{
"epoch": 3.2762148337595907,
"grad_norm": 0.28364078627553185,
"learning_rate": 2.5649113185741986e-05,
"loss": 1.0257,
"step": 1281
},
{
"epoch": 3.2787723785166243,
"grad_norm": 0.23821259851981724,
"learning_rate": 2.5624637073511248e-05,
"loss": 1.0144,
"step": 1282
},
{
"epoch": 3.2813299232736575,
"grad_norm": 0.26870907881050227,
"learning_rate": 2.5600151810347668e-05,
"loss": 1.0146,
"step": 1283
},
{
"epoch": 3.2838874680306906,
"grad_norm": 0.22187022603191378,
"learning_rate": 2.5575657436087244e-05,
"loss": 1.0091,
"step": 1284
},
{
"epoch": 3.286445012787724,
"grad_norm": 0.24738678119043905,
"learning_rate": 2.5551153990580795e-05,
"loss": 1.0133,
"step": 1285
},
{
"epoch": 3.289002557544757,
"grad_norm": 0.25403939478591225,
"learning_rate": 2.552664151369388e-05,
"loss": 1.0104,
"step": 1286
},
{
"epoch": 3.29156010230179,
"grad_norm": 0.2818502412863481,
"learning_rate": 2.5502120045306785e-05,
"loss": 0.9998,
"step": 1287
},
{
"epoch": 3.2941176470588234,
"grad_norm": 0.20499726982333147,
"learning_rate": 2.5477589625314387e-05,
"loss": 0.9865,
"step": 1288
},
{
"epoch": 3.296675191815857,
"grad_norm": 0.24471947086383825,
"learning_rate": 2.5453050293626155e-05,
"loss": 0.987,
"step": 1289
},
{
"epoch": 3.29923273657289,
"grad_norm": 0.20999980853763284,
"learning_rate": 2.5428502090166055e-05,
"loss": 0.9857,
"step": 1290
},
{
"epoch": 3.3017902813299234,
"grad_norm": 0.25183028485299946,
"learning_rate": 2.540394505487247e-05,
"loss": 0.9951,
"step": 1291
},
{
"epoch": 3.3043478260869565,
"grad_norm": 0.20676280643837916,
"learning_rate": 2.5379379227698166e-05,
"loss": 1.0299,
"step": 1292
},
{
"epoch": 3.3069053708439897,
"grad_norm": 0.24700995669682596,
"learning_rate": 2.5354804648610198e-05,
"loss": 0.9664,
"step": 1293
},
{
"epoch": 3.309462915601023,
"grad_norm": 0.20750718705716245,
"learning_rate": 2.533022135758988e-05,
"loss": 1.0002,
"step": 1294
},
{
"epoch": 3.312020460358056,
"grad_norm": 0.2582424683081433,
"learning_rate": 2.530562939463268e-05,
"loss": 1.0087,
"step": 1295
},
{
"epoch": 3.3145780051150897,
"grad_norm": 0.2541891180137742,
"learning_rate": 2.528102879974819e-05,
"loss": 1.0231,
"step": 1296
},
{
"epoch": 3.317135549872123,
"grad_norm": 0.19281327359959774,
"learning_rate": 2.5256419612960047e-05,
"loss": 0.9788,
"step": 1297
},
{
"epoch": 3.319693094629156,
"grad_norm": 0.23695257816461104,
"learning_rate": 2.5231801874305847e-05,
"loss": 0.9954,
"step": 1298
},
{
"epoch": 3.3222506393861893,
"grad_norm": 0.2286303886494054,
"learning_rate": 2.5207175623837116e-05,
"loss": 1.021,
"step": 1299
},
{
"epoch": 3.3248081841432224,
"grad_norm": 0.23056025725291834,
"learning_rate": 2.5182540901619224e-05,
"loss": 1.045,
"step": 1300
},
{
"epoch": 3.3273657289002556,
"grad_norm": 0.23615909178662206,
"learning_rate": 2.5157897747731325e-05,
"loss": 0.9852,
"step": 1301
},
{
"epoch": 3.329923273657289,
"grad_norm": 0.2759309658789834,
"learning_rate": 2.513324620226629e-05,
"loss": 1.0178,
"step": 1302
},
{
"epoch": 3.3324808184143224,
"grad_norm": 0.23436653574985533,
"learning_rate": 2.510858630533063e-05,
"loss": 0.9963,
"step": 1303
},
{
"epoch": 3.3350383631713556,
"grad_norm": 0.25317436320297865,
"learning_rate": 2.5083918097044478e-05,
"loss": 1.0117,
"step": 1304
},
{
"epoch": 3.337595907928389,
"grad_norm": 0.2636709434139457,
"learning_rate": 2.5059241617541447e-05,
"loss": 0.9877,
"step": 1305
},
{
"epoch": 3.340153452685422,
"grad_norm": 0.23407008348781694,
"learning_rate": 2.503455690696864e-05,
"loss": 1.0208,
"step": 1306
},
{
"epoch": 3.342710997442455,
"grad_norm": 0.26978161903886694,
"learning_rate": 2.5009864005486525e-05,
"loss": 0.9732,
"step": 1307
},
{
"epoch": 3.3452685421994883,
"grad_norm": 0.3022237464346918,
"learning_rate": 2.4985162953268916e-05,
"loss": 1.0166,
"step": 1308
},
{
"epoch": 3.3478260869565215,
"grad_norm": 0.26108742885501357,
"learning_rate": 2.4960453790502877e-05,
"loss": 1.0006,
"step": 1309
},
{
"epoch": 3.350383631713555,
"grad_norm": 0.24271369047215727,
"learning_rate": 2.4935736557388683e-05,
"loss": 1.0374,
"step": 1310
},
{
"epoch": 3.3529411764705883,
"grad_norm": 0.33979344807310835,
"learning_rate": 2.491101129413972e-05,
"loss": 1.0089,
"step": 1311
},
{
"epoch": 3.3554987212276215,
"grad_norm": 0.27907757197868954,
"learning_rate": 2.488627804098243e-05,
"loss": 1.0027,
"step": 1312
},
{
"epoch": 3.3580562659846547,
"grad_norm": 0.25658796981241544,
"learning_rate": 2.4861536838156287e-05,
"loss": 1.0031,
"step": 1313
},
{
"epoch": 3.360613810741688,
"grad_norm": 0.3589882717912318,
"learning_rate": 2.4836787725913673e-05,
"loss": 1.0134,
"step": 1314
},
{
"epoch": 3.363171355498721,
"grad_norm": 0.29895612578315234,
"learning_rate": 2.481203074451985e-05,
"loss": 0.9844,
"step": 1315
},
{
"epoch": 3.3657289002557547,
"grad_norm": 0.26900951208366075,
"learning_rate": 2.4787265934252882e-05,
"loss": 1.0045,
"step": 1316
},
{
"epoch": 3.368286445012788,
"grad_norm": 0.26708429033876796,
"learning_rate": 2.4762493335403563e-05,
"loss": 1.0195,
"step": 1317
},
{
"epoch": 3.370843989769821,
"grad_norm": 0.32084299134720573,
"learning_rate": 2.473771298827536e-05,
"loss": 1.0144,
"step": 1318
},
{
"epoch": 3.373401534526854,
"grad_norm": 0.27963696773108504,
"learning_rate": 2.4712924933184354e-05,
"loss": 1.0131,
"step": 1319
},
{
"epoch": 3.3759590792838874,
"grad_norm": 0.18786885424560132,
"learning_rate": 2.4688129210459154e-05,
"loss": 0.9894,
"step": 1320
},
{
"epoch": 3.3785166240409206,
"grad_norm": 0.27852929282496663,
"learning_rate": 2.466332586044085e-05,
"loss": 0.9899,
"step": 1321
},
{
"epoch": 3.381074168797954,
"grad_norm": 0.338747521971872,
"learning_rate": 2.4638514923482956e-05,
"loss": 1.0011,
"step": 1322
},
{
"epoch": 3.3836317135549874,
"grad_norm": 0.3260671185393945,
"learning_rate": 2.4613696439951298e-05,
"loss": 0.9894,
"step": 1323
},
{
"epoch": 3.3861892583120206,
"grad_norm": 0.2604977744779121,
"learning_rate": 2.458887045022401e-05,
"loss": 1.0021,
"step": 1324
},
{
"epoch": 3.3887468030690537,
"grad_norm": 0.35515721009857787,
"learning_rate": 2.456403699469141e-05,
"loss": 1.008,
"step": 1325
},
{
"epoch": 3.391304347826087,
"grad_norm": 0.23211759902765425,
"learning_rate": 2.4539196113755992e-05,
"loss": 1.0103,
"step": 1326
},
{
"epoch": 3.39386189258312,
"grad_norm": 0.26720740487526484,
"learning_rate": 2.4514347847832308e-05,
"loss": 0.9927,
"step": 1327
},
{
"epoch": 3.3964194373401533,
"grad_norm": 0.2641993709431511,
"learning_rate": 2.4489492237346934e-05,
"loss": 0.9834,
"step": 1328
},
{
"epoch": 3.398976982097187,
"grad_norm": 0.24862303547227033,
"learning_rate": 2.44646293227384e-05,
"loss": 1.0455,
"step": 1329
},
{
"epoch": 3.40153452685422,
"grad_norm": 0.26291553830730496,
"learning_rate": 2.443975914445711e-05,
"loss": 1.0281,
"step": 1330
},
{
"epoch": 3.4040920716112533,
"grad_norm": 0.25275103064482135,
"learning_rate": 2.4414881742965287e-05,
"loss": 1.0014,
"step": 1331
},
{
"epoch": 3.4066496163682864,
"grad_norm": 0.26140335990414354,
"learning_rate": 2.4389997158736902e-05,
"loss": 1.0195,
"step": 1332
},
{
"epoch": 3.4092071611253196,
"grad_norm": 0.28339075319664453,
"learning_rate": 2.4365105432257635e-05,
"loss": 0.9958,
"step": 1333
},
{
"epoch": 3.411764705882353,
"grad_norm": 0.2517786332628994,
"learning_rate": 2.434020660402475e-05,
"loss": 1.0227,
"step": 1334
},
{
"epoch": 3.414322250639386,
"grad_norm": 0.24923536009333055,
"learning_rate": 2.4315300714547095e-05,
"loss": 1.0105,
"step": 1335
},
{
"epoch": 3.4168797953964196,
"grad_norm": 0.26111685022445064,
"learning_rate": 2.4290387804344992e-05,
"loss": 1.0076,
"step": 1336
},
{
"epoch": 3.419437340153453,
"grad_norm": 0.23880295755450953,
"learning_rate": 2.4265467913950186e-05,
"loss": 1.0038,
"step": 1337
},
{
"epoch": 3.421994884910486,
"grad_norm": 0.2889994461371417,
"learning_rate": 2.4240541083905784e-05,
"loss": 0.9944,
"step": 1338
},
{
"epoch": 3.424552429667519,
"grad_norm": 0.24947376401463867,
"learning_rate": 2.421560735476618e-05,
"loss": 1.0006,
"step": 1339
},
{
"epoch": 3.4271099744245523,
"grad_norm": 0.3109680109536225,
"learning_rate": 2.419066676709699e-05,
"loss": 1.052,
"step": 1340
},
{
"epoch": 3.4296675191815855,
"grad_norm": 0.23572643642506497,
"learning_rate": 2.416571936147499e-05,
"loss": 1.0423,
"step": 1341
},
{
"epoch": 3.4322250639386187,
"grad_norm": 0.2888546719403378,
"learning_rate": 2.4140765178488052e-05,
"loss": 1.0032,
"step": 1342
},
{
"epoch": 3.4347826086956523,
"grad_norm": 0.20547181245049354,
"learning_rate": 2.411580425873507e-05,
"loss": 1.0181,
"step": 1343
},
{
"epoch": 3.4373401534526855,
"grad_norm": 0.29065901843758385,
"learning_rate": 2.40908366428259e-05,
"loss": 0.9998,
"step": 1344
},
{
"epoch": 3.4398976982097187,
"grad_norm": 0.24381638975537936,
"learning_rate": 2.4065862371381293e-05,
"loss": 0.9884,
"step": 1345
},
{
"epoch": 3.442455242966752,
"grad_norm": 0.2403661103818596,
"learning_rate": 2.404088148503282e-05,
"loss": 1.0344,
"step": 1346
},
{
"epoch": 3.445012787723785,
"grad_norm": 0.25961654363499775,
"learning_rate": 2.4015894024422838e-05,
"loss": 1.0074,
"step": 1347
},
{
"epoch": 3.4475703324808182,
"grad_norm": 0.22286319268903043,
"learning_rate": 2.399090003020436e-05,
"loss": 1.0253,
"step": 1348
},
{
"epoch": 3.4501278772378514,
"grad_norm": 0.2248163093412898,
"learning_rate": 2.396589954304107e-05,
"loss": 1.0322,
"step": 1349
},
{
"epoch": 3.452685421994885,
"grad_norm": 0.24645256590358752,
"learning_rate": 2.3940892603607196e-05,
"loss": 1.0183,
"step": 1350
},
{
"epoch": 3.455242966751918,
"grad_norm": 0.2060236850425058,
"learning_rate": 2.3915879252587453e-05,
"loss": 1.0173,
"step": 1351
},
{
"epoch": 3.4578005115089514,
"grad_norm": 0.21645488221738302,
"learning_rate": 2.3890859530677014e-05,
"loss": 1.0246,
"step": 1352
},
{
"epoch": 3.4603580562659846,
"grad_norm": 0.24755498118417854,
"learning_rate": 2.386583347858139e-05,
"loss": 0.9865,
"step": 1353
},
{
"epoch": 3.4629156010230178,
"grad_norm": 0.29709889940137646,
"learning_rate": 2.3840801137016402e-05,
"loss": 1.0092,
"step": 1354
},
{
"epoch": 3.4654731457800514,
"grad_norm": 0.22911022137483922,
"learning_rate": 2.3815762546708118e-05,
"loss": 0.9995,
"step": 1355
},
{
"epoch": 3.4680306905370846,
"grad_norm": 0.20310060527937807,
"learning_rate": 2.3790717748392754e-05,
"loss": 1.0084,
"step": 1356
},
{
"epoch": 3.4705882352941178,
"grad_norm": 0.22227270186497455,
"learning_rate": 2.376566678281662e-05,
"loss": 1.0096,
"step": 1357
},
{
"epoch": 3.473145780051151,
"grad_norm": 0.21255766189407718,
"learning_rate": 2.3740609690736086e-05,
"loss": 1.0147,
"step": 1358
},
{
"epoch": 3.475703324808184,
"grad_norm": 0.19206777806562128,
"learning_rate": 2.3715546512917463e-05,
"loss": 1.0316,
"step": 1359
},
{
"epoch": 3.4782608695652173,
"grad_norm": 0.19940350212719812,
"learning_rate": 2.369047729013698e-05,
"loss": 1.0192,
"step": 1360
},
{
"epoch": 3.4808184143222505,
"grad_norm": 0.1948182810078059,
"learning_rate": 2.366540206318069e-05,
"loss": 1.004,
"step": 1361
},
{
"epoch": 3.483375959079284,
"grad_norm": 0.21134064728653001,
"learning_rate": 2.3640320872844426e-05,
"loss": 1.0108,
"step": 1362
},
{
"epoch": 3.4859335038363173,
"grad_norm": 0.21628051951606844,
"learning_rate": 2.3615233759933712e-05,
"loss": 1.0266,
"step": 1363
},
{
"epoch": 3.4884910485933505,
"grad_norm": 0.21782705658395335,
"learning_rate": 2.3590140765263713e-05,
"loss": 0.9981,
"step": 1364
},
{
"epoch": 3.4910485933503836,
"grad_norm": 0.22132103886062446,
"learning_rate": 2.3565041929659166e-05,
"loss": 1.0329,
"step": 1365
},
{
"epoch": 3.493606138107417,
"grad_norm": 0.1919725880117112,
"learning_rate": 2.3539937293954296e-05,
"loss": 1.0511,
"step": 1366
},
{
"epoch": 3.49616368286445,
"grad_norm": 0.23153424765665778,
"learning_rate": 2.3514826898992786e-05,
"loss": 1.0296,
"step": 1367
},
{
"epoch": 3.498721227621483,
"grad_norm": 0.2130171271122437,
"learning_rate": 2.348971078562768e-05,
"loss": 0.9724,
"step": 1368
},
{
"epoch": 3.501278772378517,
"grad_norm": 0.2184231544491387,
"learning_rate": 2.3464588994721317e-05,
"loss": 1.0201,
"step": 1369
},
{
"epoch": 3.50383631713555,
"grad_norm": 0.2127652146823345,
"learning_rate": 2.343946156714529e-05,
"loss": 1.0241,
"step": 1370
},
{
"epoch": 3.506393861892583,
"grad_norm": 0.20734702296865623,
"learning_rate": 2.3414328543780336e-05,
"loss": 1.0178,
"step": 1371
},
{
"epoch": 3.5089514066496164,
"grad_norm": 0.22889980746692887,
"learning_rate": 2.338918996551633e-05,
"loss": 1.0388,
"step": 1372
},
{
"epoch": 3.5115089514066495,
"grad_norm": 0.26673165393858866,
"learning_rate": 2.3364045873252153e-05,
"loss": 0.9972,
"step": 1373
},
{
"epoch": 3.5140664961636827,
"grad_norm": 0.2189275572291676,
"learning_rate": 2.333889630789568e-05,
"loss": 1.0162,
"step": 1374
},
{
"epoch": 3.516624040920716,
"grad_norm": 0.25487150719887447,
"learning_rate": 2.331374131036368e-05,
"loss": 0.9885,
"step": 1375
},
{
"epoch": 3.5191815856777495,
"grad_norm": 0.3604874598551796,
"learning_rate": 2.3288580921581758e-05,
"loss": 0.9952,
"step": 1376
},
{
"epoch": 3.5217391304347827,
"grad_norm": 0.21865786341926038,
"learning_rate": 2.32634151824843e-05,
"loss": 1.0217,
"step": 1377
},
{
"epoch": 3.524296675191816,
"grad_norm": 0.23776099536962714,
"learning_rate": 2.323824413401438e-05,
"loss": 0.9905,
"step": 1378
},
{
"epoch": 3.526854219948849,
"grad_norm": 0.29285910186819974,
"learning_rate": 2.3213067817123732e-05,
"loss": 1.0541,
"step": 1379
},
{
"epoch": 3.5294117647058822,
"grad_norm": 0.23520011277206035,
"learning_rate": 2.3187886272772633e-05,
"loss": 0.996,
"step": 1380
},
{
"epoch": 3.531969309462916,
"grad_norm": 0.22199554657993206,
"learning_rate": 2.3162699541929892e-05,
"loss": 1.0509,
"step": 1381
},
{
"epoch": 3.5345268542199486,
"grad_norm": 0.2853075736322077,
"learning_rate": 2.3137507665572753e-05,
"loss": 1.02,
"step": 1382
},
{
"epoch": 3.5370843989769822,
"grad_norm": 0.24715425492042814,
"learning_rate": 2.311231068468681e-05,
"loss": 1.0058,
"step": 1383
},
{
"epoch": 3.5396419437340154,
"grad_norm": 0.20015219716349736,
"learning_rate": 2.308710864026598e-05,
"loss": 1.0369,
"step": 1384
},
{
"epoch": 3.5421994884910486,
"grad_norm": 0.26275780557370826,
"learning_rate": 2.306190157331241e-05,
"loss": 1.015,
"step": 1385
},
{
"epoch": 3.544757033248082,
"grad_norm": 0.2751166732419353,
"learning_rate": 2.3036689524836433e-05,
"loss": 1.033,
"step": 1386
},
{
"epoch": 3.547314578005115,
"grad_norm": 0.22378392591647772,
"learning_rate": 2.301147253585646e-05,
"loss": 0.9842,
"step": 1387
},
{
"epoch": 3.5498721227621486,
"grad_norm": 0.22127780960696636,
"learning_rate": 2.298625064739897e-05,
"loss": 1.0042,
"step": 1388
},
{
"epoch": 3.5524296675191813,
"grad_norm": 0.30754825608267555,
"learning_rate": 2.296102390049839e-05,
"loss": 1.0334,
"step": 1389
},
{
"epoch": 3.554987212276215,
"grad_norm": 0.23204632957753138,
"learning_rate": 2.2935792336197066e-05,
"loss": 1.04,
"step": 1390
},
{
"epoch": 3.557544757033248,
"grad_norm": 0.21703712592736898,
"learning_rate": 2.291055599554517e-05,
"loss": 0.9717,
"step": 1391
},
{
"epoch": 3.5601023017902813,
"grad_norm": 0.2626595457703019,
"learning_rate": 2.2885314919600646e-05,
"loss": 1.0019,
"step": 1392
},
{
"epoch": 3.5626598465473145,
"grad_norm": 0.22555374744997128,
"learning_rate": 2.286006914942916e-05,
"loss": 1.0195,
"step": 1393
},
{
"epoch": 3.5652173913043477,
"grad_norm": 0.22215707656509168,
"learning_rate": 2.2834818726103984e-05,
"loss": 1.0228,
"step": 1394
},
{
"epoch": 3.5677749360613813,
"grad_norm": 0.19661611562793266,
"learning_rate": 2.2809563690705998e-05,
"loss": 1.0304,
"step": 1395
},
{
"epoch": 3.5703324808184145,
"grad_norm": 0.2639655219897031,
"learning_rate": 2.278430408432355e-05,
"loss": 1.0041,
"step": 1396
},
{
"epoch": 3.5728900255754477,
"grad_norm": 0.20432782802093863,
"learning_rate": 2.275903994805245e-05,
"loss": 0.9875,
"step": 1397
},
{
"epoch": 3.575447570332481,
"grad_norm": 0.2326700647893841,
"learning_rate": 2.273377132299586e-05,
"loss": 1.0237,
"step": 1398
},
{
"epoch": 3.578005115089514,
"grad_norm": 0.22492155984247506,
"learning_rate": 2.2708498250264262e-05,
"loss": 1.0272,
"step": 1399
},
{
"epoch": 3.580562659846547,
"grad_norm": 0.2272058003550993,
"learning_rate": 2.2683220770975353e-05,
"loss": 0.9881,
"step": 1400
},
{
"epoch": 3.5831202046035804,
"grad_norm": 0.22443152818580023,
"learning_rate": 2.2657938926254017e-05,
"loss": 0.9976,
"step": 1401
},
{
"epoch": 3.585677749360614,
"grad_norm": 0.22796621714733886,
"learning_rate": 2.2632652757232242e-05,
"loss": 1.0168,
"step": 1402
},
{
"epoch": 3.588235294117647,
"grad_norm": 0.2491925913079835,
"learning_rate": 2.260736230504903e-05,
"loss": 1.0135,
"step": 1403
},
{
"epoch": 3.5907928388746804,
"grad_norm": 0.24203364939776426,
"learning_rate": 2.258206761085038e-05,
"loss": 1.0199,
"step": 1404
},
{
"epoch": 3.5933503836317136,
"grad_norm": 0.2639858439793371,
"learning_rate": 2.255676871578917e-05,
"loss": 1.0125,
"step": 1405
},
{
"epoch": 3.5959079283887467,
"grad_norm": 0.2510072378249399,
"learning_rate": 2.2531465661025122e-05,
"loss": 1.0111,
"step": 1406
},
{
"epoch": 3.59846547314578,
"grad_norm": 0.2615891413249054,
"learning_rate": 2.250615848772472e-05,
"loss": 1.0177,
"step": 1407
},
{
"epoch": 3.601023017902813,
"grad_norm": 0.22032456191087219,
"learning_rate": 2.2480847237061156e-05,
"loss": 1.0217,
"step": 1408
},
{
"epoch": 3.6035805626598467,
"grad_norm": 0.2867628514132791,
"learning_rate": 2.245553195021426e-05,
"loss": 1.0018,
"step": 1409
},
{
"epoch": 3.60613810741688,
"grad_norm": 0.32279828042669456,
"learning_rate": 2.2430212668370402e-05,
"loss": 0.9749,
"step": 1410
},
{
"epoch": 3.608695652173913,
"grad_norm": 0.20487913000531294,
"learning_rate": 2.2404889432722495e-05,
"loss": 1.0168,
"step": 1411
},
{
"epoch": 3.6112531969309463,
"grad_norm": 0.30099009002948707,
"learning_rate": 2.2379562284469833e-05,
"loss": 1.0351,
"step": 1412
},
{
"epoch": 3.6138107416879794,
"grad_norm": 0.2690934917435649,
"learning_rate": 2.2354231264818127e-05,
"loss": 0.9899,
"step": 1413
},
{
"epoch": 3.6163682864450126,
"grad_norm": 0.23918964373235152,
"learning_rate": 2.2328896414979345e-05,
"loss": 1.0102,
"step": 1414
},
{
"epoch": 3.618925831202046,
"grad_norm": 0.2658679432138252,
"learning_rate": 2.2303557776171713e-05,
"loss": 0.99,
"step": 1415
},
{
"epoch": 3.6214833759590794,
"grad_norm": 0.22090808835562717,
"learning_rate": 2.227821538961961e-05,
"loss": 1.0111,
"step": 1416
},
{
"epoch": 3.6240409207161126,
"grad_norm": 0.26827616460890996,
"learning_rate": 2.2252869296553513e-05,
"loss": 0.9848,
"step": 1417
},
{
"epoch": 3.626598465473146,
"grad_norm": 0.19418831655505386,
"learning_rate": 2.222751953820993e-05,
"loss": 1.0211,
"step": 1418
},
{
"epoch": 3.629156010230179,
"grad_norm": 0.22610409289381272,
"learning_rate": 2.2202166155831334e-05,
"loss": 1.0153,
"step": 1419
},
{
"epoch": 3.631713554987212,
"grad_norm": 0.23793853948484292,
"learning_rate": 2.2176809190666092e-05,
"loss": 1.0164,
"step": 1420
},
{
"epoch": 3.634271099744246,
"grad_norm": 0.2721630117073176,
"learning_rate": 2.2151448683968402e-05,
"loss": 1.0337,
"step": 1421
},
{
"epoch": 3.6368286445012785,
"grad_norm": 0.31054023590067836,
"learning_rate": 2.2126084676998226e-05,
"loss": 0.9873,
"step": 1422
},
{
"epoch": 3.639386189258312,
"grad_norm": 0.2208506443542396,
"learning_rate": 2.2100717211021213e-05,
"loss": 1.0132,
"step": 1423
},
{
"epoch": 3.6419437340153453,
"grad_norm": 0.29419043686151886,
"learning_rate": 2.2075346327308643e-05,
"loss": 1.0028,
"step": 1424
},
{
"epoch": 3.6445012787723785,
"grad_norm": 0.3020026336794495,
"learning_rate": 2.2049972067137367e-05,
"loss": 1.0324,
"step": 1425
},
{
"epoch": 3.6470588235294117,
"grad_norm": 0.22595054719313032,
"learning_rate": 2.20245944717897e-05,
"loss": 0.9917,
"step": 1426
},
{
"epoch": 3.649616368286445,
"grad_norm": 0.2596305847349027,
"learning_rate": 2.199921358255342e-05,
"loss": 1.0066,
"step": 1427
},
{
"epoch": 3.6521739130434785,
"grad_norm": 0.2871313606809214,
"learning_rate": 2.1973829440721643e-05,
"loss": 0.9977,
"step": 1428
},
{
"epoch": 3.6547314578005117,
"grad_norm": 0.22482986935026256,
"learning_rate": 2.194844208759278e-05,
"loss": 0.9905,
"step": 1429
},
{
"epoch": 3.657289002557545,
"grad_norm": 0.2687742439325648,
"learning_rate": 2.1923051564470467e-05,
"loss": 1.0192,
"step": 1430
},
{
"epoch": 3.659846547314578,
"grad_norm": 0.28138625508422943,
"learning_rate": 2.1897657912663504e-05,
"loss": 1.014,
"step": 1431
},
{
"epoch": 3.662404092071611,
"grad_norm": 0.24676677335402444,
"learning_rate": 2.1872261173485762e-05,
"loss": 1.0171,
"step": 1432
},
{
"epoch": 3.6649616368286444,
"grad_norm": 0.21333561561041478,
"learning_rate": 2.1846861388256148e-05,
"loss": 0.9918,
"step": 1433
},
{
"epoch": 3.6675191815856776,
"grad_norm": 0.2412565703601667,
"learning_rate": 2.182145859829854e-05,
"loss": 0.9981,
"step": 1434
},
{
"epoch": 3.670076726342711,
"grad_norm": 0.21325558375211667,
"learning_rate": 2.1796052844941675e-05,
"loss": 0.9955,
"step": 1435
},
{
"epoch": 3.6726342710997444,
"grad_norm": 0.1922001313777053,
"learning_rate": 2.177064416951913e-05,
"loss": 1.0251,
"step": 1436
},
{
"epoch": 3.6751918158567776,
"grad_norm": 0.23745283723697966,
"learning_rate": 2.1745232613369233e-05,
"loss": 1.0013,
"step": 1437
},
{
"epoch": 3.6777493606138107,
"grad_norm": 0.201947200229008,
"learning_rate": 2.1719818217834994e-05,
"loss": 1.0622,
"step": 1438
},
{
"epoch": 3.680306905370844,
"grad_norm": 0.2664146316211629,
"learning_rate": 2.169440102426404e-05,
"loss": 0.9915,
"step": 1439
},
{
"epoch": 3.682864450127877,
"grad_norm": 0.19329264636735277,
"learning_rate": 2.1668981074008572e-05,
"loss": 0.9935,
"step": 1440
},
{
"epoch": 3.6854219948849103,
"grad_norm": 0.26679438762808166,
"learning_rate": 2.1643558408425246e-05,
"loss": 0.997,
"step": 1441
},
{
"epoch": 3.687979539641944,
"grad_norm": 0.5109591987387416,
"learning_rate": 2.1618133068875152e-05,
"loss": 0.9926,
"step": 1442
},
{
"epoch": 3.690537084398977,
"grad_norm": 0.2275937340434201,
"learning_rate": 2.1592705096723727e-05,
"loss": 0.9753,
"step": 1443
},
{
"epoch": 3.6930946291560103,
"grad_norm": 0.21283716380819506,
"learning_rate": 2.1567274533340697e-05,
"loss": 1.001,
"step": 1444
},
{
"epoch": 3.6956521739130435,
"grad_norm": 0.22086888072907834,
"learning_rate": 2.1541841420099993e-05,
"loss": 1.0226,
"step": 1445
},
{
"epoch": 3.6982097186700766,
"grad_norm": 0.22714555383680451,
"learning_rate": 2.1516405798379703e-05,
"loss": 0.9862,
"step": 1446
},
{
"epoch": 3.70076726342711,
"grad_norm": 0.23256197882074,
"learning_rate": 2.149096770956199e-05,
"loss": 1.0126,
"step": 1447
},
{
"epoch": 3.703324808184143,
"grad_norm": 0.2166852052866917,
"learning_rate": 2.146552719503304e-05,
"loss": 1.0354,
"step": 1448
},
{
"epoch": 3.7058823529411766,
"grad_norm": 0.255740778566146,
"learning_rate": 2.1440084296182974e-05,
"loss": 0.9863,
"step": 1449
},
{
"epoch": 3.70843989769821,
"grad_norm": 0.2095048096688552,
"learning_rate": 2.1414639054405804e-05,
"loss": 1.0015,
"step": 1450
},
{
"epoch": 3.710997442455243,
"grad_norm": 0.28817744937588297,
"learning_rate": 2.1389191511099335e-05,
"loss": 0.9813,
"step": 1451
},
{
"epoch": 3.713554987212276,
"grad_norm": 0.3038305040814257,
"learning_rate": 2.1363741707665144e-05,
"loss": 1.0134,
"step": 1452
},
{
"epoch": 3.7161125319693094,
"grad_norm": 0.23437244094303478,
"learning_rate": 2.133828968550846e-05,
"loss": 0.9959,
"step": 1453
},
{
"epoch": 3.718670076726343,
"grad_norm": 0.281134722764533,
"learning_rate": 2.131283548603813e-05,
"loss": 1.038,
"step": 1454
},
{
"epoch": 3.7212276214833757,
"grad_norm": 0.346741832659884,
"learning_rate": 2.1287379150666553e-05,
"loss": 0.9929,
"step": 1455
},
{
"epoch": 3.7237851662404093,
"grad_norm": 0.24362674421115338,
"learning_rate": 2.1261920720809587e-05,
"loss": 0.9904,
"step": 1456
},
{
"epoch": 3.7263427109974425,
"grad_norm": 0.2503580452990493,
"learning_rate": 2.123646023788652e-05,
"loss": 1.034,
"step": 1457
},
{
"epoch": 3.7289002557544757,
"grad_norm": 0.2490907487690136,
"learning_rate": 2.121099774331995e-05,
"loss": 0.991,
"step": 1458
},
{
"epoch": 3.731457800511509,
"grad_norm": 0.2574009220702953,
"learning_rate": 2.118553327853577e-05,
"loss": 1.0256,
"step": 1459
},
{
"epoch": 3.734015345268542,
"grad_norm": 0.2124323562734229,
"learning_rate": 2.1160066884963064e-05,
"loss": 1.0102,
"step": 1460
},
{
"epoch": 3.7365728900255757,
"grad_norm": 0.2700753027440079,
"learning_rate": 2.1134598604034076e-05,
"loss": 0.9869,
"step": 1461
},
{
"epoch": 3.7391304347826084,
"grad_norm": 0.20782741649260297,
"learning_rate": 2.1109128477184103e-05,
"loss": 0.9933,
"step": 1462
},
{
"epoch": 3.741687979539642,
"grad_norm": 0.23325138032265733,
"learning_rate": 2.1083656545851442e-05,
"loss": 1.0018,
"step": 1463
},
{
"epoch": 3.7442455242966752,
"grad_norm": 0.2289164666702718,
"learning_rate": 2.1058182851477338e-05,
"loss": 1.0208,
"step": 1464
},
{
"epoch": 3.7468030690537084,
"grad_norm": 0.2198736952966155,
"learning_rate": 2.1032707435505895e-05,
"loss": 1.008,
"step": 1465
},
{
"epoch": 3.7493606138107416,
"grad_norm": 0.22329709077000717,
"learning_rate": 2.1007230339384025e-05,
"loss": 1.007,
"step": 1466
},
{
"epoch": 3.7519181585677748,
"grad_norm": 0.19721850253836315,
"learning_rate": 2.098175160456137e-05,
"loss": 1.0079,
"step": 1467
},
{
"epoch": 3.7544757033248084,
"grad_norm": 0.21797865451863857,
"learning_rate": 2.0956271272490244e-05,
"loss": 1.0029,
"step": 1468
},
{
"epoch": 3.7570332480818416,
"grad_norm": 0.20892728651308398,
"learning_rate": 2.093078938462554e-05,
"loss": 1.0265,
"step": 1469
},
{
"epoch": 3.7595907928388748,
"grad_norm": 0.19844034456013784,
"learning_rate": 2.0905305982424714e-05,
"loss": 0.9821,
"step": 1470
},
{
"epoch": 3.762148337595908,
"grad_norm": 0.22826138586150116,
"learning_rate": 2.0879821107347657e-05,
"loss": 1.0159,
"step": 1471
},
{
"epoch": 3.764705882352941,
"grad_norm": 0.21540579142203,
"learning_rate": 2.0854334800856662e-05,
"loss": 1.0094,
"step": 1472
},
{
"epoch": 3.7672634271099743,
"grad_norm": 0.24865214066497845,
"learning_rate": 2.082884710441637e-05,
"loss": 0.9923,
"step": 1473
},
{
"epoch": 3.7698209718670075,
"grad_norm": 0.40632359232217086,
"learning_rate": 2.0803358059493656e-05,
"loss": 1.0027,
"step": 1474
},
{
"epoch": 3.772378516624041,
"grad_norm": 0.2366222053023054,
"learning_rate": 2.077786770755762e-05,
"loss": 1.0166,
"step": 1475
},
{
"epoch": 3.7749360613810743,
"grad_norm": 0.20073615375665088,
"learning_rate": 2.0752376090079446e-05,
"loss": 1.0074,
"step": 1476
},
{
"epoch": 3.7774936061381075,
"grad_norm": 0.23649038345320883,
"learning_rate": 2.0726883248532426e-05,
"loss": 0.9781,
"step": 1477
},
{
"epoch": 3.7800511508951407,
"grad_norm": 0.1992643078000637,
"learning_rate": 2.070138922439181e-05,
"loss": 1.0023,
"step": 1478
},
{
"epoch": 3.782608695652174,
"grad_norm": 0.23265439844740812,
"learning_rate": 2.0675894059134775e-05,
"loss": 0.9854,
"step": 1479
},
{
"epoch": 3.785166240409207,
"grad_norm": 0.2447345828476251,
"learning_rate": 2.0650397794240365e-05,
"loss": 0.9965,
"step": 1480
},
{
"epoch": 3.78772378516624,
"grad_norm": 0.1899517145795999,
"learning_rate": 2.0624900471189412e-05,
"loss": 0.9985,
"step": 1481
},
{
"epoch": 3.790281329923274,
"grad_norm": 0.26052471639305835,
"learning_rate": 2.059940213146447e-05,
"loss": 1.0169,
"step": 1482
},
{
"epoch": 3.792838874680307,
"grad_norm": 0.27111420183981694,
"learning_rate": 2.057390281654973e-05,
"loss": 1.071,
"step": 1483
},
{
"epoch": 3.79539641943734,
"grad_norm": 0.1873090294736801,
"learning_rate": 2.0548402567930994e-05,
"loss": 1.0134,
"step": 1484
},
{
"epoch": 3.7979539641943734,
"grad_norm": 0.2845316347626959,
"learning_rate": 2.052290142709557e-05,
"loss": 1.0464,
"step": 1485
},
{
"epoch": 3.8005115089514065,
"grad_norm": 0.20705525283940526,
"learning_rate": 2.0497399435532214e-05,
"loss": 1.0396,
"step": 1486
},
{
"epoch": 3.80306905370844,
"grad_norm": 0.18967656737258498,
"learning_rate": 2.047189663473108e-05,
"loss": 1.0203,
"step": 1487
},
{
"epoch": 3.805626598465473,
"grad_norm": 0.2067281284822576,
"learning_rate": 2.0446393066183626e-05,
"loss": 1.0008,
"step": 1488
},
{
"epoch": 3.8081841432225065,
"grad_norm": 0.56521172108997,
"learning_rate": 2.0420888771382565e-05,
"loss": 1.0111,
"step": 1489
},
{
"epoch": 3.8107416879795397,
"grad_norm": 0.20960148057843195,
"learning_rate": 2.039538379182179e-05,
"loss": 1.012,
"step": 1490
},
{
"epoch": 3.813299232736573,
"grad_norm": 0.2191831860773827,
"learning_rate": 2.0369878168996306e-05,
"loss": 1.0105,
"step": 1491
},
{
"epoch": 3.815856777493606,
"grad_norm": 0.1947197966244743,
"learning_rate": 2.0344371944402165e-05,
"loss": 1.0066,
"step": 1492
},
{
"epoch": 3.8184143222506393,
"grad_norm": 0.2404900780543259,
"learning_rate": 2.03188651595364e-05,
"loss": 1.0143,
"step": 1493
},
{
"epoch": 3.820971867007673,
"grad_norm": 0.1975408952369514,
"learning_rate": 2.0293357855896968e-05,
"loss": 0.9905,
"step": 1494
},
{
"epoch": 3.8235294117647056,
"grad_norm": 0.27010269657202285,
"learning_rate": 2.026785007498264e-05,
"loss": 1.0425,
"step": 1495
},
{
"epoch": 3.8260869565217392,
"grad_norm": 0.2563682601851434,
"learning_rate": 2.0242341858292986e-05,
"loss": 1.0082,
"step": 1496
},
{
"epoch": 3.8286445012787724,
"grad_norm": 0.2162931948212925,
"learning_rate": 2.0216833247328284e-05,
"loss": 1.0204,
"step": 1497
},
{
"epoch": 3.8312020460358056,
"grad_norm": 0.24087143041456105,
"learning_rate": 2.019132428358945e-05,
"loss": 0.9899,
"step": 1498
},
{
"epoch": 3.833759590792839,
"grad_norm": 0.2644482127798988,
"learning_rate": 2.0165815008577954e-05,
"loss": 0.9996,
"step": 1499
},
{
"epoch": 3.836317135549872,
"grad_norm": 0.2175774012871972,
"learning_rate": 2.0140305463795822e-05,
"loss": 1.0249,
"step": 1500
},
{
"epoch": 3.8388746803069056,
"grad_norm": 0.20643023924224324,
"learning_rate": 2.0114795690745467e-05,
"loss": 1.0079,
"step": 1501
},
{
"epoch": 3.8414322250639388,
"grad_norm": 0.2682368260599753,
"learning_rate": 2.00892857309297e-05,
"loss": 1.0166,
"step": 1502
},
{
"epoch": 3.843989769820972,
"grad_norm": 0.2198454157358291,
"learning_rate": 2.0063775625851636e-05,
"loss": 0.9801,
"step": 1503
},
{
"epoch": 3.846547314578005,
"grad_norm": 0.23449255625087895,
"learning_rate": 2.0038265417014613e-05,
"loss": 1.0268,
"step": 1504
},
{
"epoch": 3.8491048593350383,
"grad_norm": 0.2734595094542989,
"learning_rate": 2.001275514592215e-05,
"loss": 1.028,
"step": 1505
},
{
"epoch": 3.8516624040920715,
"grad_norm": 0.19764464210503555,
"learning_rate": 1.998724485407786e-05,
"loss": 1.0038,
"step": 1506
},
{
"epoch": 3.8542199488491047,
"grad_norm": 0.24247792158558626,
"learning_rate": 1.9961734582985393e-05,
"loss": 1.0221,
"step": 1507
},
{
"epoch": 3.8567774936061383,
"grad_norm": 0.217936055984476,
"learning_rate": 1.9936224374148374e-05,
"loss": 1.0046,
"step": 1508
},
{
"epoch": 3.8593350383631715,
"grad_norm": 0.24357176268874794,
"learning_rate": 1.9910714269070306e-05,
"loss": 1.0332,
"step": 1509
},
{
"epoch": 3.8618925831202047,
"grad_norm": 0.2262978282776521,
"learning_rate": 1.988520430925454e-05,
"loss": 1.0097,
"step": 1510
},
{
"epoch": 3.864450127877238,
"grad_norm": 0.22003611683231875,
"learning_rate": 1.9859694536204184e-05,
"loss": 0.9845,
"step": 1511
},
{
"epoch": 3.867007672634271,
"grad_norm": 0.2179950205843959,
"learning_rate": 1.9834184991422046e-05,
"loss": 0.9703,
"step": 1512
},
{
"epoch": 3.869565217391304,
"grad_norm": 0.2684475240536641,
"learning_rate": 1.9808675716410564e-05,
"loss": 0.9987,
"step": 1513
},
{
"epoch": 3.8721227621483374,
"grad_norm": 0.27164092360953823,
"learning_rate": 1.9783166752671723e-05,
"loss": 1.0089,
"step": 1514
},
{
"epoch": 3.874680306905371,
"grad_norm": 0.247237286006558,
"learning_rate": 1.975765814170702e-05,
"loss": 1.0081,
"step": 1515
},
{
"epoch": 3.877237851662404,
"grad_norm": 0.24921918245583677,
"learning_rate": 1.9732149925017368e-05,
"loss": 1.0204,
"step": 1516
},
{
"epoch": 3.8797953964194374,
"grad_norm": 0.21996081991013464,
"learning_rate": 1.970664214410304e-05,
"loss": 1.0438,
"step": 1517
},
{
"epoch": 3.8823529411764706,
"grad_norm": 0.21541078212295395,
"learning_rate": 1.9681134840463604e-05,
"loss": 1.0161,
"step": 1518
},
{
"epoch": 3.8849104859335037,
"grad_norm": 0.2164939253935731,
"learning_rate": 1.965562805559784e-05,
"loss": 1.0331,
"step": 1519
},
{
"epoch": 3.887468030690537,
"grad_norm": 0.23410172733467113,
"learning_rate": 1.9630121831003704e-05,
"loss": 1.0242,
"step": 1520
},
{
"epoch": 3.89002557544757,
"grad_norm": 0.2173437275032242,
"learning_rate": 1.9604616208178217e-05,
"loss": 0.9973,
"step": 1521
},
{
"epoch": 3.8925831202046037,
"grad_norm": 0.19233303515257355,
"learning_rate": 1.9579111228617442e-05,
"loss": 0.9678,
"step": 1522
},
{
"epoch": 3.895140664961637,
"grad_norm": 0.2117808011312655,
"learning_rate": 1.955360693381638e-05,
"loss": 1.0075,
"step": 1523
},
{
"epoch": 3.89769820971867,
"grad_norm": 0.1653661924508728,
"learning_rate": 1.9528103365268923e-05,
"loss": 1.0004,
"step": 1524
},
{
"epoch": 3.9002557544757033,
"grad_norm": 0.2106664625390623,
"learning_rate": 1.950260056446779e-05,
"loss": 1.0255,
"step": 1525
},
{
"epoch": 3.9028132992327365,
"grad_norm": 0.18718243325879697,
"learning_rate": 1.9477098572904437e-05,
"loss": 1.0102,
"step": 1526
},
{
"epoch": 3.90537084398977,
"grad_norm": 0.2316262795207412,
"learning_rate": 1.9451597432069013e-05,
"loss": 1.0012,
"step": 1527
},
{
"epoch": 3.907928388746803,
"grad_norm": 0.2412888537019545,
"learning_rate": 1.9426097183450276e-05,
"loss": 1.0171,
"step": 1528
},
{
"epoch": 3.9104859335038364,
"grad_norm": 0.2087958494964007,
"learning_rate": 1.9400597868535533e-05,
"loss": 0.9798,
"step": 1529
},
{
"epoch": 3.9130434782608696,
"grad_norm": 0.230862348833611,
"learning_rate": 1.937509952881059e-05,
"loss": 1.01,
"step": 1530
},
{
"epoch": 3.915601023017903,
"grad_norm": 0.23048333018443184,
"learning_rate": 1.9349602205759635e-05,
"loss": 0.9999,
"step": 1531
},
{
"epoch": 3.918158567774936,
"grad_norm": 0.21970077570065388,
"learning_rate": 1.9324105940865235e-05,
"loss": 1.0134,
"step": 1532
},
{
"epoch": 3.920716112531969,
"grad_norm": 0.21365220220348427,
"learning_rate": 1.92986107756082e-05,
"loss": 1.0125,
"step": 1533
},
{
"epoch": 3.923273657289003,
"grad_norm": 0.19400162524292092,
"learning_rate": 1.927311675146758e-05,
"loss": 1.0242,
"step": 1534
},
{
"epoch": 3.9258312020460355,
"grad_norm": 0.24042688839296517,
"learning_rate": 1.9247623909920557e-05,
"loss": 1.0123,
"step": 1535
},
{
"epoch": 3.928388746803069,
"grad_norm": 0.21142434193977822,
"learning_rate": 1.9222132292442385e-05,
"loss": 0.9997,
"step": 1536
},
{
"epoch": 3.9309462915601023,
"grad_norm": 0.2228669623057862,
"learning_rate": 1.919664194050635e-05,
"loss": 0.9989,
"step": 1537
},
{
"epoch": 3.9335038363171355,
"grad_norm": 0.24024924798157118,
"learning_rate": 1.9171152895583635e-05,
"loss": 1.0161,
"step": 1538
},
{
"epoch": 3.9360613810741687,
"grad_norm": 0.222501187476257,
"learning_rate": 1.9145665199143348e-05,
"loss": 0.9996,
"step": 1539
},
{
"epoch": 3.938618925831202,
"grad_norm": 0.21017814400036106,
"learning_rate": 1.912017889265235e-05,
"loss": 1.0035,
"step": 1540
},
{
"epoch": 3.9411764705882355,
"grad_norm": 0.22735273727217786,
"learning_rate": 1.9094694017575296e-05,
"loss": 1.0045,
"step": 1541
},
{
"epoch": 3.9437340153452687,
"grad_norm": 0.20840312420453092,
"learning_rate": 1.9069210615374465e-05,
"loss": 0.9752,
"step": 1542
},
{
"epoch": 3.946291560102302,
"grad_norm": 0.2381255889486689,
"learning_rate": 1.904372872750976e-05,
"loss": 0.9999,
"step": 1543
},
{
"epoch": 3.948849104859335,
"grad_norm": 0.2038281691136896,
"learning_rate": 1.9018248395438634e-05,
"loss": 0.98,
"step": 1544
},
{
"epoch": 3.9514066496163682,
"grad_norm": 0.22399429140754382,
"learning_rate": 1.8992769660615975e-05,
"loss": 1.0201,
"step": 1545
},
{
"epoch": 3.9539641943734014,
"grad_norm": 0.2337613404308509,
"learning_rate": 1.8967292564494112e-05,
"loss": 0.9893,
"step": 1546
},
{
"epoch": 3.9565217391304346,
"grad_norm": 0.23306038974010024,
"learning_rate": 1.8941817148522666e-05,
"loss": 1.0198,
"step": 1547
},
{
"epoch": 3.959079283887468,
"grad_norm": 0.2478674286077631,
"learning_rate": 1.8916343454148568e-05,
"loss": 0.998,
"step": 1548
},
{
"epoch": 3.9616368286445014,
"grad_norm": 0.22046395984034586,
"learning_rate": 1.88908715228159e-05,
"loss": 1.0618,
"step": 1549
},
{
"epoch": 3.9641943734015346,
"grad_norm": 0.2710240655530491,
"learning_rate": 1.8865401395965924e-05,
"loss": 1.028,
"step": 1550
},
{
"epoch": 3.9667519181585678,
"grad_norm": 0.2017434096277467,
"learning_rate": 1.8839933115036943e-05,
"loss": 1.0197,
"step": 1551
},
{
"epoch": 3.969309462915601,
"grad_norm": 0.2753477369212786,
"learning_rate": 1.8814466721464237e-05,
"loss": 1.0186,
"step": 1552
},
{
"epoch": 3.971867007672634,
"grad_norm": 0.23693156078141014,
"learning_rate": 1.878900225668006e-05,
"loss": 1.0136,
"step": 1553
},
{
"epoch": 3.9744245524296673,
"grad_norm": 0.25097346223339895,
"learning_rate": 1.8763539762113487e-05,
"loss": 1.0028,
"step": 1554
},
{
"epoch": 3.976982097186701,
"grad_norm": 0.30349379269266363,
"learning_rate": 1.873807927919042e-05,
"loss": 1.0235,
"step": 1555
},
{
"epoch": 3.979539641943734,
"grad_norm": 0.2208455455128153,
"learning_rate": 1.871262084933345e-05,
"loss": 1.0389,
"step": 1556
},
{
"epoch": 3.9820971867007673,
"grad_norm": 0.28180451672678725,
"learning_rate": 1.8687164513961874e-05,
"loss": 1.0051,
"step": 1557
},
{
"epoch": 3.9846547314578005,
"grad_norm": 0.2380180963545922,
"learning_rate": 1.866171031449155e-05,
"loss": 1.0213,
"step": 1558
},
{
"epoch": 3.9872122762148337,
"grad_norm": 0.2554219564869537,
"learning_rate": 1.8636258292334863e-05,
"loss": 1.0009,
"step": 1559
},
{
"epoch": 3.9897698209718673,
"grad_norm": 0.2025619573794405,
"learning_rate": 1.8610808488900675e-05,
"loss": 1.0169,
"step": 1560
},
{
"epoch": 3.9923273657289,
"grad_norm": 0.2558084656811946,
"learning_rate": 1.8585360945594203e-05,
"loss": 1.0133,
"step": 1561
},
{
"epoch": 3.9948849104859336,
"grad_norm": 0.23199404378127614,
"learning_rate": 1.8559915703817026e-05,
"loss": 1.0257,
"step": 1562
},
{
"epoch": 3.997442455242967,
"grad_norm": 0.24880240851809174,
"learning_rate": 1.8534472804966964e-05,
"loss": 1.0082,
"step": 1563
},
{
"epoch": 4.0,
"grad_norm": 0.19833802048217009,
"learning_rate": 1.8509032290438012e-05,
"loss": 1.0146,
"step": 1564
},
{
"epoch": 4.002557544757034,
"grad_norm": 0.2705134962250848,
"learning_rate": 1.8483594201620303e-05,
"loss": 0.9822,
"step": 1565
},
{
"epoch": 4.005115089514066,
"grad_norm": 0.20779994175646202,
"learning_rate": 1.845815857990001e-05,
"loss": 0.9926,
"step": 1566
},
{
"epoch": 4.0076726342711,
"grad_norm": 0.2602958397975164,
"learning_rate": 1.8432725466659313e-05,
"loss": 0.9666,
"step": 1567
},
{
"epoch": 4.010230179028133,
"grad_norm": 0.22127206088618015,
"learning_rate": 1.8407294903276276e-05,
"loss": 0.9876,
"step": 1568
},
{
"epoch": 4.012787723785166,
"grad_norm": 0.23883308203907633,
"learning_rate": 1.838186693112485e-05,
"loss": 1.0004,
"step": 1569
},
{
"epoch": 4.015345268542199,
"grad_norm": 0.2199223424245,
"learning_rate": 1.835644159157476e-05,
"loss": 1.0031,
"step": 1570
},
{
"epoch": 4.017902813299233,
"grad_norm": 0.20928212184336153,
"learning_rate": 1.833101892599143e-05,
"loss": 0.9989,
"step": 1571
},
{
"epoch": 4.020460358056266,
"grad_norm": 0.18205346175809875,
"learning_rate": 1.8305598975735962e-05,
"loss": 0.9846,
"step": 1572
},
{
"epoch": 4.023017902813299,
"grad_norm": 0.21768709900975045,
"learning_rate": 1.8280181782165012e-05,
"loss": 0.9852,
"step": 1573
},
{
"epoch": 4.025575447570333,
"grad_norm": 0.20394877342863937,
"learning_rate": 1.8254767386630774e-05,
"loss": 0.9801,
"step": 1574
},
{
"epoch": 4.028132992327365,
"grad_norm": 0.19915878143203544,
"learning_rate": 1.8229355830480872e-05,
"loss": 0.9887,
"step": 1575
},
{
"epoch": 4.030690537084399,
"grad_norm": 0.2185744413295794,
"learning_rate": 1.8203947155058328e-05,
"loss": 1.0015,
"step": 1576
},
{
"epoch": 4.033248081841432,
"grad_norm": 0.18210548604562823,
"learning_rate": 1.8178541401701466e-05,
"loss": 0.9688,
"step": 1577
},
{
"epoch": 4.035805626598465,
"grad_norm": 0.24246468129981819,
"learning_rate": 1.8153138611743852e-05,
"loss": 0.9992,
"step": 1578
},
{
"epoch": 4.038363171355499,
"grad_norm": 0.21628121385833785,
"learning_rate": 1.8127738826514248e-05,
"loss": 0.9982,
"step": 1579
},
{
"epoch": 4.040920716112532,
"grad_norm": 0.1884375953871055,
"learning_rate": 1.8102342087336503e-05,
"loss": 1.015,
"step": 1580
},
{
"epoch": 4.043478260869565,
"grad_norm": 0.20024924797953478,
"learning_rate": 1.8076948435529536e-05,
"loss": 0.9927,
"step": 1581
},
{
"epoch": 4.046035805626598,
"grad_norm": 0.2191450164482883,
"learning_rate": 1.8051557912407222e-05,
"loss": 0.9779,
"step": 1582
},
{
"epoch": 4.048593350383632,
"grad_norm": 0.20274624540817546,
"learning_rate": 1.8026170559278357e-05,
"loss": 0.9702,
"step": 1583
},
{
"epoch": 4.051150895140665,
"grad_norm": 0.19573751483878332,
"learning_rate": 1.8000786417446582e-05,
"loss": 0.9901,
"step": 1584
},
{
"epoch": 4.053708439897698,
"grad_norm": 0.21854002360034955,
"learning_rate": 1.7975405528210303e-05,
"loss": 0.9857,
"step": 1585
},
{
"epoch": 4.056265984654732,
"grad_norm": 0.20188858809779228,
"learning_rate": 1.7950027932862646e-05,
"loss": 1.0275,
"step": 1586
},
{
"epoch": 4.0588235294117645,
"grad_norm": 0.19882632991831184,
"learning_rate": 1.792465367269136e-05,
"loss": 0.9533,
"step": 1587
},
{
"epoch": 4.061381074168798,
"grad_norm": 0.2275799923179516,
"learning_rate": 1.7899282788978797e-05,
"loss": 0.9783,
"step": 1588
},
{
"epoch": 4.063938618925831,
"grad_norm": 0.230001797624487,
"learning_rate": 1.787391532300178e-05,
"loss": 1.0206,
"step": 1589
},
{
"epoch": 4.0664961636828645,
"grad_norm": 0.21612204002816876,
"learning_rate": 1.78485513160316e-05,
"loss": 0.9852,
"step": 1590
},
{
"epoch": 4.069053708439898,
"grad_norm": 0.22892319167501912,
"learning_rate": 1.7823190809333915e-05,
"loss": 1.0219,
"step": 1591
},
{
"epoch": 4.071611253196931,
"grad_norm": 0.246466661964565,
"learning_rate": 1.7797833844168673e-05,
"loss": 1.0136,
"step": 1592
},
{
"epoch": 4.0741687979539645,
"grad_norm": 0.20560282430202298,
"learning_rate": 1.777248046179008e-05,
"loss": 1.0222,
"step": 1593
},
{
"epoch": 4.076726342710997,
"grad_norm": 0.23035829532819327,
"learning_rate": 1.7747130703446497e-05,
"loss": 1.0251,
"step": 1594
},
{
"epoch": 4.079283887468031,
"grad_norm": 0.2267889310522548,
"learning_rate": 1.77217846103804e-05,
"loss": 0.9862,
"step": 1595
},
{
"epoch": 4.081841432225064,
"grad_norm": 0.22273632144193733,
"learning_rate": 1.7696442223828294e-05,
"loss": 1.0116,
"step": 1596
},
{
"epoch": 4.084398976982097,
"grad_norm": 0.2502763528033944,
"learning_rate": 1.7671103585020658e-05,
"loss": 0.9986,
"step": 1597
},
{
"epoch": 4.086956521739131,
"grad_norm": 0.217290954773665,
"learning_rate": 1.7645768735181883e-05,
"loss": 0.9635,
"step": 1598
},
{
"epoch": 4.089514066496164,
"grad_norm": 0.23631151544117304,
"learning_rate": 1.762043771553017e-05,
"loss": 0.9688,
"step": 1599
},
{
"epoch": 4.092071611253197,
"grad_norm": 0.20723564813933193,
"learning_rate": 1.759511056727752e-05,
"loss": 1.0037,
"step": 1600
},
{
"epoch": 4.09462915601023,
"grad_norm": 0.2058237400819985,
"learning_rate": 1.75697873316296e-05,
"loss": 1.0002,
"step": 1601
},
{
"epoch": 4.0971867007672635,
"grad_norm": 0.19931344296383147,
"learning_rate": 1.7544468049785744e-05,
"loss": 0.9817,
"step": 1602
},
{
"epoch": 4.099744245524296,
"grad_norm": 0.19283889592682701,
"learning_rate": 1.751915276293885e-05,
"loss": 0.9711,
"step": 1603
},
{
"epoch": 4.10230179028133,
"grad_norm": 0.1903646817692344,
"learning_rate": 1.7493841512275285e-05,
"loss": 1.0118,
"step": 1604
},
{
"epoch": 4.1048593350383635,
"grad_norm": 0.18304773800895718,
"learning_rate": 1.7468534338974888e-05,
"loss": 0.9945,
"step": 1605
},
{
"epoch": 4.107416879795396,
"grad_norm": 0.19129916787109763,
"learning_rate": 1.7443231284210834e-05,
"loss": 1.0014,
"step": 1606
},
{
"epoch": 4.10997442455243,
"grad_norm": 0.20766431878744962,
"learning_rate": 1.741793238914963e-05,
"loss": 0.9764,
"step": 1607
},
{
"epoch": 4.112531969309463,
"grad_norm": 0.19083344478696038,
"learning_rate": 1.7392637694950975e-05,
"loss": 0.9854,
"step": 1608
},
{
"epoch": 4.115089514066496,
"grad_norm": 0.196083993434377,
"learning_rate": 1.736734724276776e-05,
"loss": 0.9996,
"step": 1609
},
{
"epoch": 4.117647058823529,
"grad_norm": 0.22306050452231574,
"learning_rate": 1.734206107374599e-05,
"loss": 0.9976,
"step": 1610
},
{
"epoch": 4.120204603580563,
"grad_norm": 0.2220481123879761,
"learning_rate": 1.731677922902465e-05,
"loss": 0.9875,
"step": 1611
},
{
"epoch": 4.122762148337596,
"grad_norm": 0.22969675719470642,
"learning_rate": 1.729150174973575e-05,
"loss": 0.9815,
"step": 1612
},
{
"epoch": 4.125319693094629,
"grad_norm": 0.2539992752232489,
"learning_rate": 1.7266228677004145e-05,
"loss": 1.0082,
"step": 1613
},
{
"epoch": 4.127877237851663,
"grad_norm": 0.23365597766222132,
"learning_rate": 1.724096005194756e-05,
"loss": 0.997,
"step": 1614
},
{
"epoch": 4.130434782608695,
"grad_norm": 0.2557842982430287,
"learning_rate": 1.7215695915676455e-05,
"loss": 0.9981,
"step": 1615
},
{
"epoch": 4.132992327365729,
"grad_norm": 0.2280142061171877,
"learning_rate": 1.7190436309294002e-05,
"loss": 0.998,
"step": 1616
},
{
"epoch": 4.135549872122763,
"grad_norm": 0.23227041325919565,
"learning_rate": 1.716518127389602e-05,
"loss": 0.9899,
"step": 1617
},
{
"epoch": 4.138107416879795,
"grad_norm": 0.2244299907883558,
"learning_rate": 1.713993085057084e-05,
"loss": 0.9913,
"step": 1618
},
{
"epoch": 4.140664961636829,
"grad_norm": 0.21568355195979444,
"learning_rate": 1.711468508039936e-05,
"loss": 0.9674,
"step": 1619
},
{
"epoch": 4.143222506393862,
"grad_norm": 0.2324935507947805,
"learning_rate": 1.7089444004454834e-05,
"loss": 1.0209,
"step": 1620
},
{
"epoch": 4.145780051150895,
"grad_norm": 0.21829075050381955,
"learning_rate": 1.7064207663802944e-05,
"loss": 1.003,
"step": 1621
},
{
"epoch": 4.148337595907928,
"grad_norm": 0.25071192940480985,
"learning_rate": 1.7038976099501616e-05,
"loss": 0.9974,
"step": 1622
},
{
"epoch": 4.150895140664962,
"grad_norm": 0.22030550312850222,
"learning_rate": 1.701374935260103e-05,
"loss": 1.0161,
"step": 1623
},
{
"epoch": 4.153452685421995,
"grad_norm": 0.2615952929742451,
"learning_rate": 1.6988527464143544e-05,
"loss": 0.9776,
"step": 1624
},
{
"epoch": 4.156010230179028,
"grad_norm": 0.23396794282432945,
"learning_rate": 1.696331047516357e-05,
"loss": 0.9892,
"step": 1625
},
{
"epoch": 4.158567774936062,
"grad_norm": 0.2237847774585795,
"learning_rate": 1.6938098426687596e-05,
"loss": 1.0031,
"step": 1626
},
{
"epoch": 4.161125319693094,
"grad_norm": 0.20380642297255058,
"learning_rate": 1.6912891359734026e-05,
"loss": 0.9917,
"step": 1627
},
{
"epoch": 4.163682864450128,
"grad_norm": 0.2198408670208088,
"learning_rate": 1.6887689315313193e-05,
"loss": 1.0112,
"step": 1628
},
{
"epoch": 4.166240409207161,
"grad_norm": 0.23559886623670231,
"learning_rate": 1.6862492334427254e-05,
"loss": 1.0118,
"step": 1629
},
{
"epoch": 4.168797953964194,
"grad_norm": 0.19811438785861182,
"learning_rate": 1.6837300458070108e-05,
"loss": 1.0004,
"step": 1630
},
{
"epoch": 4.171355498721228,
"grad_norm": 0.245806278250192,
"learning_rate": 1.6812113727227377e-05,
"loss": 1.021,
"step": 1631
},
{
"epoch": 4.173913043478261,
"grad_norm": 0.19366257464960737,
"learning_rate": 1.6786932182876278e-05,
"loss": 0.9961,
"step": 1632
},
{
"epoch": 4.176470588235294,
"grad_norm": 0.20513830153019041,
"learning_rate": 1.676175586598563e-05,
"loss": 0.9916,
"step": 1633
},
{
"epoch": 4.179028132992327,
"grad_norm": 0.21098779312505897,
"learning_rate": 1.6736584817515707e-05,
"loss": 0.9883,
"step": 1634
},
{
"epoch": 4.181585677749361,
"grad_norm": 0.18955064800360757,
"learning_rate": 1.6711419078418242e-05,
"loss": 1.007,
"step": 1635
},
{
"epoch": 4.1841432225063935,
"grad_norm": 0.20163623119047358,
"learning_rate": 1.6686258689636324e-05,
"loss": 1.0384,
"step": 1636
},
{
"epoch": 4.186700767263427,
"grad_norm": 0.20229185259789245,
"learning_rate": 1.666110369210432e-05,
"loss": 0.99,
"step": 1637
},
{
"epoch": 4.189258312020461,
"grad_norm": 0.18453785834509756,
"learning_rate": 1.663595412674785e-05,
"loss": 1.0069,
"step": 1638
},
{
"epoch": 4.1918158567774935,
"grad_norm": 0.18274919723574518,
"learning_rate": 1.6610810034483674e-05,
"loss": 0.9897,
"step": 1639
},
{
"epoch": 4.194373401534527,
"grad_norm": 0.17766575782892946,
"learning_rate": 1.658567145621967e-05,
"loss": 0.9827,
"step": 1640
},
{
"epoch": 4.19693094629156,
"grad_norm": 0.18796418408134544,
"learning_rate": 1.656053843285472e-05,
"loss": 0.9634,
"step": 1641
},
{
"epoch": 4.1994884910485935,
"grad_norm": 0.206710416350049,
"learning_rate": 1.6535411005278683e-05,
"loss": 1.0122,
"step": 1642
},
{
"epoch": 4.202046035805626,
"grad_norm": 0.17152181004960462,
"learning_rate": 1.6510289214372323e-05,
"loss": 0.9671,
"step": 1643
},
{
"epoch": 4.20460358056266,
"grad_norm": 0.20065838912941056,
"learning_rate": 1.6485173101007214e-05,
"loss": 1.0095,
"step": 1644
},
{
"epoch": 4.207161125319693,
"grad_norm": 0.1802120373052194,
"learning_rate": 1.646006270604571e-05,
"loss": 0.9982,
"step": 1645
},
{
"epoch": 4.209718670076726,
"grad_norm": 0.18058889164334171,
"learning_rate": 1.6434958070340844e-05,
"loss": 0.9761,
"step": 1646
},
{
"epoch": 4.21227621483376,
"grad_norm": 0.18193043451445873,
"learning_rate": 1.6409859234736294e-05,
"loss": 0.9799,
"step": 1647
},
{
"epoch": 4.2148337595907925,
"grad_norm": 0.19673062778612105,
"learning_rate": 1.6384766240066294e-05,
"loss": 0.992,
"step": 1648
},
{
"epoch": 4.217391304347826,
"grad_norm": 0.19246958549110282,
"learning_rate": 1.6359679127155578e-05,
"loss": 0.9725,
"step": 1649
},
{
"epoch": 4.21994884910486,
"grad_norm": 0.22516146093922598,
"learning_rate": 1.6334597936819317e-05,
"loss": 0.9794,
"step": 1650
},
{
"epoch": 4.2225063938618925,
"grad_norm": 0.18138117361298708,
"learning_rate": 1.6309522709863026e-05,
"loss": 1.002,
"step": 1651
},
{
"epoch": 4.225063938618926,
"grad_norm": 0.21241251141837805,
"learning_rate": 1.6284453487082547e-05,
"loss": 0.9883,
"step": 1652
},
{
"epoch": 4.227621483375959,
"grad_norm": 0.18818523127121975,
"learning_rate": 1.625939030926392e-05,
"loss": 0.9999,
"step": 1653
},
{
"epoch": 4.2301790281329925,
"grad_norm": 0.20393824938290886,
"learning_rate": 1.6234333217183386e-05,
"loss": 0.9953,
"step": 1654
},
{
"epoch": 4.232736572890025,
"grad_norm": 0.1879811539851975,
"learning_rate": 1.6209282251607256e-05,
"loss": 1.019,
"step": 1655
},
{
"epoch": 4.235294117647059,
"grad_norm": 0.20468249139310976,
"learning_rate": 1.6184237453291885e-05,
"loss": 0.9972,
"step": 1656
},
{
"epoch": 4.2378516624040925,
"grad_norm": 0.20256429521073457,
"learning_rate": 1.61591988629836e-05,
"loss": 0.9821,
"step": 1657
},
{
"epoch": 4.240409207161125,
"grad_norm": 0.23556457469342715,
"learning_rate": 1.6134166521418618e-05,
"loss": 1.016,
"step": 1658
},
{
"epoch": 4.242966751918159,
"grad_norm": 0.22261029586753878,
"learning_rate": 1.6109140469322996e-05,
"loss": 0.9996,
"step": 1659
},
{
"epoch": 4.245524296675192,
"grad_norm": 0.2287789006871306,
"learning_rate": 1.608412074741255e-05,
"loss": 1.0023,
"step": 1660
},
{
"epoch": 4.248081841432225,
"grad_norm": 0.23964802600749865,
"learning_rate": 1.605910739639281e-05,
"loss": 1.008,
"step": 1661
},
{
"epoch": 4.250639386189258,
"grad_norm": 0.2162036911052715,
"learning_rate": 1.6034100456958933e-05,
"loss": 1.004,
"step": 1662
},
{
"epoch": 4.253196930946292,
"grad_norm": 0.21263703609795334,
"learning_rate": 1.6009099969795644e-05,
"loss": 0.9874,
"step": 1663
},
{
"epoch": 4.255754475703325,
"grad_norm": 0.21662806734825674,
"learning_rate": 1.5984105975577172e-05,
"loss": 0.9848,
"step": 1664
},
{
"epoch": 4.258312020460358,
"grad_norm": 0.18590562525075188,
"learning_rate": 1.5959118514967182e-05,
"loss": 0.9911,
"step": 1665
},
{
"epoch": 4.260869565217392,
"grad_norm": 0.2050884188106912,
"learning_rate": 1.5934137628618717e-05,
"loss": 0.9886,
"step": 1666
},
{
"epoch": 4.263427109974424,
"grad_norm": 0.20788106717102225,
"learning_rate": 1.5909163357174108e-05,
"loss": 1.0005,
"step": 1667
},
{
"epoch": 4.265984654731458,
"grad_norm": 0.19637457000525535,
"learning_rate": 1.5884195741264935e-05,
"loss": 0.9487,
"step": 1668
},
{
"epoch": 4.268542199488491,
"grad_norm": 0.20814015182743792,
"learning_rate": 1.5859234821511955e-05,
"loss": 0.9929,
"step": 1669
},
{
"epoch": 4.271099744245524,
"grad_norm": 0.19253874425193387,
"learning_rate": 1.5834280638525013e-05,
"loss": 1.0049,
"step": 1670
},
{
"epoch": 4.273657289002558,
"grad_norm": 0.21222244645801075,
"learning_rate": 1.580933323290302e-05,
"loss": 0.9805,
"step": 1671
},
{
"epoch": 4.276214833759591,
"grad_norm": 0.1855222680003295,
"learning_rate": 1.5784392645233825e-05,
"loss": 0.9929,
"step": 1672
},
{
"epoch": 4.278772378516624,
"grad_norm": 0.18528893886899309,
"learning_rate": 1.5759458916094222e-05,
"loss": 0.9472,
"step": 1673
},
{
"epoch": 4.281329923273657,
"grad_norm": 0.2003701651015556,
"learning_rate": 1.573453208604982e-05,
"loss": 0.9856,
"step": 1674
},
{
"epoch": 4.283887468030691,
"grad_norm": 0.20134702036644375,
"learning_rate": 1.5709612195655015e-05,
"loss": 0.9969,
"step": 1675
},
{
"epoch": 4.286445012787723,
"grad_norm": 0.1840201213440248,
"learning_rate": 1.5684699285452915e-05,
"loss": 0.9853,
"step": 1676
},
{
"epoch": 4.289002557544757,
"grad_norm": 0.20863422536590764,
"learning_rate": 1.5659793395975255e-05,
"loss": 1.0028,
"step": 1677
},
{
"epoch": 4.291560102301791,
"grad_norm": 0.2067970753932743,
"learning_rate": 1.5634894567742375e-05,
"loss": 1.0025,
"step": 1678
},
{
"epoch": 4.294117647058823,
"grad_norm": 0.1979660007893109,
"learning_rate": 1.56100028412631e-05,
"loss": 0.9896,
"step": 1679
},
{
"epoch": 4.296675191815857,
"grad_norm": 0.2127911755809147,
"learning_rate": 1.5585118257034726e-05,
"loss": 1.0089,
"step": 1680
},
{
"epoch": 4.29923273657289,
"grad_norm": 0.2533520523698757,
"learning_rate": 1.5560240855542897e-05,
"loss": 0.9801,
"step": 1681
},
{
"epoch": 4.301790281329923,
"grad_norm": 0.19186281823557236,
"learning_rate": 1.55353706772616e-05,
"loss": 1.0281,
"step": 1682
},
{
"epoch": 4.304347826086957,
"grad_norm": 0.21872871087105744,
"learning_rate": 1.551050776265307e-05,
"loss": 1.0054,
"step": 1683
},
{
"epoch": 4.30690537084399,
"grad_norm": 0.20741536288996093,
"learning_rate": 1.5485652152167695e-05,
"loss": 1.0275,
"step": 1684
},
{
"epoch": 4.309462915601023,
"grad_norm": 0.20269243799641268,
"learning_rate": 1.5460803886244014e-05,
"loss": 0.9773,
"step": 1685
},
{
"epoch": 4.312020460358056,
"grad_norm": 0.1994190904831448,
"learning_rate": 1.5435963005308594e-05,
"loss": 1.0098,
"step": 1686
},
{
"epoch": 4.31457800511509,
"grad_norm": 0.2248169809147335,
"learning_rate": 1.5411129549776003e-05,
"loss": 0.9729,
"step": 1687
},
{
"epoch": 4.3171355498721224,
"grad_norm": 0.18801291834730782,
"learning_rate": 1.538630356004871e-05,
"loss": 0.9729,
"step": 1688
},
{
"epoch": 4.319693094629156,
"grad_norm": 0.21067955404585834,
"learning_rate": 1.5361485076517047e-05,
"loss": 0.9768,
"step": 1689
},
{
"epoch": 4.322250639386189,
"grad_norm": 0.1722863984658784,
"learning_rate": 1.5336674139559155e-05,
"loss": 1.0385,
"step": 1690
},
{
"epoch": 4.324808184143222,
"grad_norm": 0.19435589914216797,
"learning_rate": 1.531187078954085e-05,
"loss": 0.9868,
"step": 1691
},
{
"epoch": 4.327365728900256,
"grad_norm": 0.1878964353488032,
"learning_rate": 1.5287075066815656e-05,
"loss": 0.9857,
"step": 1692
},
{
"epoch": 4.329923273657289,
"grad_norm": 0.20513837323437764,
"learning_rate": 1.526228701172465e-05,
"loss": 1.0032,
"step": 1693
},
{
"epoch": 4.332480818414322,
"grad_norm": 0.7505425885190485,
"learning_rate": 1.5237506664596437e-05,
"loss": 1.0111,
"step": 1694
},
{
"epoch": 4.335038363171355,
"grad_norm": 0.18931001773983253,
"learning_rate": 1.5212734065747125e-05,
"loss": 0.9815,
"step": 1695
},
{
"epoch": 4.337595907928389,
"grad_norm": 0.19640642908809133,
"learning_rate": 1.518796925548015e-05,
"loss": 1.0204,
"step": 1696
},
{
"epoch": 4.340153452685422,
"grad_norm": 0.19785598874843696,
"learning_rate": 1.5163212274086335e-05,
"loss": 0.988,
"step": 1697
},
{
"epoch": 4.342710997442455,
"grad_norm": 0.18538866069366555,
"learning_rate": 1.5138463161843716e-05,
"loss": 0.9653,
"step": 1698
},
{
"epoch": 4.345268542199489,
"grad_norm": 0.19767065276482684,
"learning_rate": 1.5113721959017582e-05,
"loss": 0.9986,
"step": 1699
},
{
"epoch": 4.3478260869565215,
"grad_norm": 0.1945405960848488,
"learning_rate": 1.508898870586029e-05,
"loss": 0.9984,
"step": 1700
},
{
"epoch": 4.350383631713555,
"grad_norm": 0.1756466842798829,
"learning_rate": 1.5064263442611318e-05,
"loss": 0.9589,
"step": 1701
},
{
"epoch": 4.352941176470588,
"grad_norm": 0.19711203430957647,
"learning_rate": 1.5039546209497124e-05,
"loss": 1.0112,
"step": 1702
},
{
"epoch": 4.3554987212276215,
"grad_norm": 0.18474645205408985,
"learning_rate": 1.5014837046731084e-05,
"loss": 1.0266,
"step": 1703
},
{
"epoch": 4.358056265984655,
"grad_norm": 0.18275630871058304,
"learning_rate": 1.4990135994513484e-05,
"loss": 0.9806,
"step": 1704
},
{
"epoch": 4.360613810741688,
"grad_norm": 0.17549284603245063,
"learning_rate": 1.4965443093031367e-05,
"loss": 0.9968,
"step": 1705
},
{
"epoch": 4.3631713554987215,
"grad_norm": 0.18799455435346532,
"learning_rate": 1.494075838245856e-05,
"loss": 1.01,
"step": 1706
},
{
"epoch": 4.365728900255754,
"grad_norm": 0.1932370297833927,
"learning_rate": 1.4916081902955525e-05,
"loss": 0.9964,
"step": 1707
},
{
"epoch": 4.368286445012788,
"grad_norm": 0.19519394979755894,
"learning_rate": 1.4891413694669367e-05,
"loss": 0.9848,
"step": 1708
},
{
"epoch": 4.370843989769821,
"grad_norm": 0.20689862890040717,
"learning_rate": 1.486675379773372e-05,
"loss": 1.0108,
"step": 1709
},
{
"epoch": 4.373401534526854,
"grad_norm": 0.21117936622691696,
"learning_rate": 1.484210225226868e-05,
"loss": 1.001,
"step": 1710
},
{
"epoch": 4.375959079283888,
"grad_norm": 0.2027661864363934,
"learning_rate": 1.4817459098380786e-05,
"loss": 1.0219,
"step": 1711
},
{
"epoch": 4.378516624040921,
"grad_norm": 0.21951488098676225,
"learning_rate": 1.4792824376162889e-05,
"loss": 1.0227,
"step": 1712
},
{
"epoch": 4.381074168797954,
"grad_norm": 0.20071685979214615,
"learning_rate": 1.4768198125694165e-05,
"loss": 0.9856,
"step": 1713
},
{
"epoch": 4.383631713554987,
"grad_norm": 0.18244281008386296,
"learning_rate": 1.4743580387039958e-05,
"loss": 1.0123,
"step": 1714
},
{
"epoch": 4.3861892583120206,
"grad_norm": 0.20663024555542414,
"learning_rate": 1.4718971200251807e-05,
"loss": 1.0074,
"step": 1715
},
{
"epoch": 4.388746803069053,
"grad_norm": 0.1919054818879838,
"learning_rate": 1.4694370605367325e-05,
"loss": 0.9955,
"step": 1716
},
{
"epoch": 4.391304347826087,
"grad_norm": 0.1928657316983994,
"learning_rate": 1.4669778642410126e-05,
"loss": 1.0302,
"step": 1717
},
{
"epoch": 4.3938618925831205,
"grad_norm": 0.20446993010168707,
"learning_rate": 1.4645195351389809e-05,
"loss": 0.9721,
"step": 1718
},
{
"epoch": 4.396419437340153,
"grad_norm": 0.1953291260558289,
"learning_rate": 1.462062077230184e-05,
"loss": 0.9859,
"step": 1719
},
{
"epoch": 4.398976982097187,
"grad_norm": 0.1908603544469174,
"learning_rate": 1.4596054945127536e-05,
"loss": 1.0041,
"step": 1720
},
{
"epoch": 4.40153452685422,
"grad_norm": 0.20695290513450137,
"learning_rate": 1.4571497909833946e-05,
"loss": 1.0247,
"step": 1721
},
{
"epoch": 4.404092071611253,
"grad_norm": 0.1735566949215553,
"learning_rate": 1.4546949706373843e-05,
"loss": 1.0093,
"step": 1722
},
{
"epoch": 4.406649616368286,
"grad_norm": 0.20202493472194175,
"learning_rate": 1.4522410374685618e-05,
"loss": 1.0144,
"step": 1723
},
{
"epoch": 4.40920716112532,
"grad_norm": 0.19664794058367624,
"learning_rate": 1.4497879954693223e-05,
"loss": 0.9843,
"step": 1724
},
{
"epoch": 4.411764705882353,
"grad_norm": 0.18689675614041482,
"learning_rate": 1.4473358486306124e-05,
"loss": 0.9959,
"step": 1725
},
{
"epoch": 4.414322250639386,
"grad_norm": 0.18382033137327702,
"learning_rate": 1.4448846009419214e-05,
"loss": 0.9986,
"step": 1726
},
{
"epoch": 4.41687979539642,
"grad_norm": 0.19472233745676237,
"learning_rate": 1.4424342563912754e-05,
"loss": 0.9856,
"step": 1727
},
{
"epoch": 4.419437340153452,
"grad_norm": 0.19368903266851892,
"learning_rate": 1.4399848189652334e-05,
"loss": 0.9848,
"step": 1728
},
{
"epoch": 4.421994884910486,
"grad_norm": 0.170056415405898,
"learning_rate": 1.4375362926488753e-05,
"loss": 0.9938,
"step": 1729
},
{
"epoch": 4.42455242966752,
"grad_norm": 0.19595923774496535,
"learning_rate": 1.4350886814258015e-05,
"loss": 1.0315,
"step": 1730
},
{
"epoch": 4.427109974424552,
"grad_norm": 0.17491048043625068,
"learning_rate": 1.4326419892781216e-05,
"loss": 0.9784,
"step": 1731
},
{
"epoch": 4.429667519181586,
"grad_norm": 0.18188913068812382,
"learning_rate": 1.4301962201864515e-05,
"loss": 0.9707,
"step": 1732
},
{
"epoch": 4.432225063938619,
"grad_norm": 0.1768505878417713,
"learning_rate": 1.4277513781299032e-05,
"loss": 1.0028,
"step": 1733
},
{
"epoch": 4.434782608695652,
"grad_norm": 0.18229489816512806,
"learning_rate": 1.4253074670860834e-05,
"loss": 0.9873,
"step": 1734
},
{
"epoch": 4.437340153452685,
"grad_norm": 0.17486351977734252,
"learning_rate": 1.4228644910310823e-05,
"loss": 1.0073,
"step": 1735
},
{
"epoch": 4.439897698209719,
"grad_norm": 0.19173647012266543,
"learning_rate": 1.4204224539394681e-05,
"loss": 1.0279,
"step": 1736
},
{
"epoch": 4.442455242966752,
"grad_norm": 0.17336379674940164,
"learning_rate": 1.417981359784284e-05,
"loss": 0.9887,
"step": 1737
},
{
"epoch": 4.445012787723785,
"grad_norm": 0.17197658043837832,
"learning_rate": 1.4155412125370362e-05,
"loss": 0.9934,
"step": 1738
},
{
"epoch": 4.447570332480819,
"grad_norm": 0.17234952997611083,
"learning_rate": 1.413102016167693e-05,
"loss": 1.0059,
"step": 1739
},
{
"epoch": 4.450127877237851,
"grad_norm": 0.16770175662423764,
"learning_rate": 1.410663774644673e-05,
"loss": 0.965,
"step": 1740
},
{
"epoch": 4.452685421994885,
"grad_norm": 0.1618493627840099,
"learning_rate": 1.4082264919348431e-05,
"loss": 0.9837,
"step": 1741
},
{
"epoch": 4.455242966751918,
"grad_norm": 0.18112146719755498,
"learning_rate": 1.405790172003511e-05,
"loss": 1.0274,
"step": 1742
},
{
"epoch": 4.457800511508951,
"grad_norm": 0.17146089323026106,
"learning_rate": 1.4033548188144154e-05,
"loss": 0.9839,
"step": 1743
},
{
"epoch": 4.460358056265985,
"grad_norm": 0.17733943295728571,
"learning_rate": 1.4009204363297251e-05,
"loss": 0.9914,
"step": 1744
},
{
"epoch": 4.462915601023018,
"grad_norm": 0.1919475410779648,
"learning_rate": 1.3984870285100265e-05,
"loss": 1.0001,
"step": 1745
},
{
"epoch": 4.465473145780051,
"grad_norm": 0.1906277309233265,
"learning_rate": 1.3960545993143235e-05,
"loss": 1.0033,
"step": 1746
},
{
"epoch": 4.468030690537084,
"grad_norm": 0.18485284999439844,
"learning_rate": 1.3936231527000254e-05,
"loss": 0.9898,
"step": 1747
},
{
"epoch": 4.470588235294118,
"grad_norm": 0.21352815315057175,
"learning_rate": 1.391192692622944e-05,
"loss": 0.9953,
"step": 1748
},
{
"epoch": 4.4731457800511505,
"grad_norm": 0.21978880275377535,
"learning_rate": 1.3887632230372865e-05,
"loss": 0.9971,
"step": 1749
},
{
"epoch": 4.475703324808184,
"grad_norm": 0.19474184285137008,
"learning_rate": 1.386334747895647e-05,
"loss": 1.0208,
"step": 1750
},
{
"epoch": 4.478260869565218,
"grad_norm": 0.2193601101040908,
"learning_rate": 1.3839072711490037e-05,
"loss": 0.9862,
"step": 1751
},
{
"epoch": 4.4808184143222505,
"grad_norm": 0.20522769176580702,
"learning_rate": 1.3814807967467086e-05,
"loss": 1.0031,
"step": 1752
},
{
"epoch": 4.483375959079284,
"grad_norm": 0.24209585766517155,
"learning_rate": 1.3790553286364846e-05,
"loss": 1.0154,
"step": 1753
},
{
"epoch": 4.485933503836317,
"grad_norm": 0.19176574115372141,
"learning_rate": 1.3766308707644158e-05,
"loss": 1.0146,
"step": 1754
},
{
"epoch": 4.4884910485933505,
"grad_norm": 0.2188620830105831,
"learning_rate": 1.3742074270749444e-05,
"loss": 0.9871,
"step": 1755
},
{
"epoch": 4.491048593350383,
"grad_norm": 0.18513266018075517,
"learning_rate": 1.3717850015108616e-05,
"loss": 0.9996,
"step": 1756
},
{
"epoch": 4.493606138107417,
"grad_norm": 0.20530227025572015,
"learning_rate": 1.3693635980133019e-05,
"loss": 1.0044,
"step": 1757
},
{
"epoch": 4.4961636828644505,
"grad_norm": 0.17924976580284352,
"learning_rate": 1.3669432205217378e-05,
"loss": 0.9772,
"step": 1758
},
{
"epoch": 4.498721227621483,
"grad_norm": 0.19128205560087108,
"learning_rate": 1.3645238729739721e-05,
"loss": 0.9654,
"step": 1759
},
{
"epoch": 4.501278772378517,
"grad_norm": 0.19299910065935996,
"learning_rate": 1.362105559306131e-05,
"loss": 1.0161,
"step": 1760
},
{
"epoch": 4.5038363171355495,
"grad_norm": 0.16288998231219742,
"learning_rate": 1.3596882834526622e-05,
"loss": 1.0101,
"step": 1761
},
{
"epoch": 4.506393861892583,
"grad_norm": 0.1895803937586114,
"learning_rate": 1.357272049346319e-05,
"loss": 1.0213,
"step": 1762
},
{
"epoch": 4.508951406649617,
"grad_norm": 0.1731225463588214,
"learning_rate": 1.3548568609181664e-05,
"loss": 1.0065,
"step": 1763
},
{
"epoch": 4.5115089514066495,
"grad_norm": 0.17731495333081987,
"learning_rate": 1.3524427220975627e-05,
"loss": 0.9813,
"step": 1764
},
{
"epoch": 4.514066496163683,
"grad_norm": 0.16520424360770744,
"learning_rate": 1.350029636812162e-05,
"loss": 0.9966,
"step": 1765
},
{
"epoch": 4.516624040920716,
"grad_norm": 0.17294279167526697,
"learning_rate": 1.3476176089879021e-05,
"loss": 0.9857,
"step": 1766
},
{
"epoch": 4.5191815856777495,
"grad_norm": 0.18589345587773498,
"learning_rate": 1.3452066425490018e-05,
"loss": 1.0132,
"step": 1767
},
{
"epoch": 4.521739130434782,
"grad_norm": 0.1843767371039297,
"learning_rate": 1.3427967414179545e-05,
"loss": 1.0116,
"step": 1768
},
{
"epoch": 4.524296675191816,
"grad_norm": 0.18200483419923205,
"learning_rate": 1.3403879095155161e-05,
"loss": 0.9706,
"step": 1769
},
{
"epoch": 4.526854219948849,
"grad_norm": 0.20350005700918555,
"learning_rate": 1.3379801507607075e-05,
"loss": 0.9957,
"step": 1770
},
{
"epoch": 4.529411764705882,
"grad_norm": 0.17018483511122326,
"learning_rate": 1.3355734690707997e-05,
"loss": 0.9901,
"step": 1771
},
{
"epoch": 4.531969309462916,
"grad_norm": 0.18033366463545897,
"learning_rate": 1.333167868361316e-05,
"loss": 0.9935,
"step": 1772
},
{
"epoch": 4.534526854219949,
"grad_norm": 0.17179355921416833,
"learning_rate": 1.3307633525460158e-05,
"loss": 0.9815,
"step": 1773
},
{
"epoch": 4.537084398976982,
"grad_norm": 0.17838529915958048,
"learning_rate": 1.3283599255368961e-05,
"loss": 0.9806,
"step": 1774
},
{
"epoch": 4.539641943734015,
"grad_norm": 0.18793943723818643,
"learning_rate": 1.3259575912441849e-05,
"loss": 0.9739,
"step": 1775
},
{
"epoch": 4.542199488491049,
"grad_norm": 0.18427754967196716,
"learning_rate": 1.323556353576327e-05,
"loss": 0.9877,
"step": 1776
},
{
"epoch": 4.544757033248082,
"grad_norm": 0.17471467736137306,
"learning_rate": 1.3211562164399881e-05,
"loss": 1.0028,
"step": 1777
},
{
"epoch": 4.547314578005115,
"grad_norm": 0.2008787955972603,
"learning_rate": 1.3187571837400393e-05,
"loss": 1.0142,
"step": 1778
},
{
"epoch": 4.549872122762149,
"grad_norm": 0.17944046293577579,
"learning_rate": 1.316359259379559e-05,
"loss": 0.9951,
"step": 1779
},
{
"epoch": 4.552429667519181,
"grad_norm": 0.19336815990254555,
"learning_rate": 1.3139624472598178e-05,
"loss": 0.9894,
"step": 1780
},
{
"epoch": 4.554987212276215,
"grad_norm": 0.1835068199589258,
"learning_rate": 1.3115667512802808e-05,
"loss": 1.0175,
"step": 1781
},
{
"epoch": 4.557544757033249,
"grad_norm": 0.18343368686582423,
"learning_rate": 1.3091721753385958e-05,
"loss": 0.9951,
"step": 1782
},
{
"epoch": 4.560102301790281,
"grad_norm": 0.18378340811109545,
"learning_rate": 1.3067787233305864e-05,
"loss": 1.0465,
"step": 1783
},
{
"epoch": 4.562659846547315,
"grad_norm": 0.19262522018345618,
"learning_rate": 1.3043863991502517e-05,
"loss": 0.9705,
"step": 1784
},
{
"epoch": 4.565217391304348,
"grad_norm": 0.21946367543119605,
"learning_rate": 1.301995206689751e-05,
"loss": 0.9717,
"step": 1785
},
{
"epoch": 4.567774936061381,
"grad_norm": 0.16937383350781923,
"learning_rate": 1.2996051498394073e-05,
"loss": 0.9778,
"step": 1786
},
{
"epoch": 4.570332480818414,
"grad_norm": 0.2113790846387167,
"learning_rate": 1.2972162324876916e-05,
"loss": 0.9902,
"step": 1787
},
{
"epoch": 4.572890025575448,
"grad_norm": 0.18869106851917095,
"learning_rate": 1.2948284585212245e-05,
"loss": 0.979,
"step": 1788
},
{
"epoch": 4.57544757033248,
"grad_norm": 0.19276212963913816,
"learning_rate": 1.292441831824765e-05,
"loss": 1.0276,
"step": 1789
},
{
"epoch": 4.578005115089514,
"grad_norm": 0.19115131269440847,
"learning_rate": 1.2900563562812043e-05,
"loss": 1.0035,
"step": 1790
},
{
"epoch": 4.580562659846548,
"grad_norm": 0.20027314267423682,
"learning_rate": 1.2876720357715635e-05,
"loss": 0.9921,
"step": 1791
},
{
"epoch": 4.58312020460358,
"grad_norm": 0.19082057353721607,
"learning_rate": 1.2852888741749812e-05,
"loss": 0.995,
"step": 1792
},
{
"epoch": 4.585677749360614,
"grad_norm": 0.17123507287475365,
"learning_rate": 1.2829068753687133e-05,
"loss": 0.9869,
"step": 1793
},
{
"epoch": 4.588235294117647,
"grad_norm": 0.2220496190038636,
"learning_rate": 1.2805260432281237e-05,
"loss": 0.9972,
"step": 1794
},
{
"epoch": 4.59079283887468,
"grad_norm": 0.18145722334341322,
"learning_rate": 1.278146381626676e-05,
"loss": 0.9793,
"step": 1795
},
{
"epoch": 4.593350383631714,
"grad_norm": 0.2158827961791877,
"learning_rate": 1.275767894435932e-05,
"loss": 0.9915,
"step": 1796
},
{
"epoch": 4.595907928388747,
"grad_norm": 0.19674591543554606,
"learning_rate": 1.2733905855255407e-05,
"loss": 0.9989,
"step": 1797
},
{
"epoch": 4.59846547314578,
"grad_norm": 0.20420752203042425,
"learning_rate": 1.2710144587632356e-05,
"loss": 0.9759,
"step": 1798
},
{
"epoch": 4.601023017902813,
"grad_norm": 0.21105479587165882,
"learning_rate": 1.268639518014826e-05,
"loss": 0.9811,
"step": 1799
},
{
"epoch": 4.603580562659847,
"grad_norm": 0.190342271561739,
"learning_rate": 1.2662657671441922e-05,
"loss": 1.0151,
"step": 1800
},
{
"epoch": 4.6061381074168795,
"grad_norm": 0.2141307209003618,
"learning_rate": 1.263893210013279e-05,
"loss": 0.9761,
"step": 1801
},
{
"epoch": 4.608695652173913,
"grad_norm": 0.18239182294246956,
"learning_rate": 1.2615218504820876e-05,
"loss": 0.965,
"step": 1802
},
{
"epoch": 4.611253196930946,
"grad_norm": 0.20154745809246133,
"learning_rate": 1.2591516924086725e-05,
"loss": 0.9898,
"step": 1803
},
{
"epoch": 4.6138107416879794,
"grad_norm": 0.17759727604362968,
"learning_rate": 1.2567827396491317e-05,
"loss": 0.9581,
"step": 1804
},
{
"epoch": 4.616368286445013,
"grad_norm": 0.20926601703627168,
"learning_rate": 1.2544149960576042e-05,
"loss": 1.0003,
"step": 1805
},
{
"epoch": 4.618925831202046,
"grad_norm": 0.17681256168398524,
"learning_rate": 1.25204846548626e-05,
"loss": 0.9639,
"step": 1806
},
{
"epoch": 4.621483375959079,
"grad_norm": 0.17470471445584076,
"learning_rate": 1.249683151785297e-05,
"loss": 0.9936,
"step": 1807
},
{
"epoch": 4.624040920716112,
"grad_norm": 0.18924948102979355,
"learning_rate": 1.2473190588029324e-05,
"loss": 1.0372,
"step": 1808
},
{
"epoch": 4.626598465473146,
"grad_norm": 0.1890119497042781,
"learning_rate": 1.244956190385398e-05,
"loss": 1.0044,
"step": 1809
},
{
"epoch": 4.629156010230179,
"grad_norm": 0.19817256438255618,
"learning_rate": 1.2425945503769326e-05,
"loss": 0.9952,
"step": 1810
},
{
"epoch": 4.631713554987212,
"grad_norm": 0.18353931888301359,
"learning_rate": 1.2402341426197765e-05,
"loss": 0.9934,
"step": 1811
},
{
"epoch": 4.634271099744246,
"grad_norm": 0.2008756189917893,
"learning_rate": 1.2378749709541664e-05,
"loss": 1.0203,
"step": 1812
},
{
"epoch": 4.6368286445012785,
"grad_norm": 0.19159059845413373,
"learning_rate": 1.2355170392183262e-05,
"loss": 0.96,
"step": 1813
},
{
"epoch": 4.639386189258312,
"grad_norm": 0.20021771032286267,
"learning_rate": 1.233160351248464e-05,
"loss": 1.0277,
"step": 1814
},
{
"epoch": 4.641943734015345,
"grad_norm": 0.1885917680690139,
"learning_rate": 1.230804910878764e-05,
"loss": 1.02,
"step": 1815
},
{
"epoch": 4.6445012787723785,
"grad_norm": 0.2003713150784065,
"learning_rate": 1.2284507219413797e-05,
"loss": 1.0219,
"step": 1816
},
{
"epoch": 4.647058823529412,
"grad_norm": 0.20170604009770804,
"learning_rate": 1.2260977882664302e-05,
"loss": 1.0194,
"step": 1817
},
{
"epoch": 4.649616368286445,
"grad_norm": 0.20785160449251436,
"learning_rate": 1.2237461136819901e-05,
"loss": 1.0076,
"step": 1818
},
{
"epoch": 4.6521739130434785,
"grad_norm": 0.20797658117846782,
"learning_rate": 1.2213957020140886e-05,
"loss": 0.9821,
"step": 1819
},
{
"epoch": 4.654731457800511,
"grad_norm": 0.24214748066399447,
"learning_rate": 1.219046557086697e-05,
"loss": 0.9693,
"step": 1820
},
{
"epoch": 4.657289002557545,
"grad_norm": 0.2132425891804952,
"learning_rate": 1.2166986827217278e-05,
"loss": 0.973,
"step": 1821
},
{
"epoch": 4.659846547314578,
"grad_norm": 0.24696741290779523,
"learning_rate": 1.2143520827390263e-05,
"loss": 1.0254,
"step": 1822
},
{
"epoch": 4.662404092071611,
"grad_norm": 0.17867266770885723,
"learning_rate": 1.2120067609563631e-05,
"loss": 0.9702,
"step": 1823
},
{
"epoch": 4.664961636828645,
"grad_norm": 0.20636488172872927,
"learning_rate": 1.2096627211894307e-05,
"loss": 0.9841,
"step": 1824
},
{
"epoch": 4.667519181585678,
"grad_norm": 0.19858776876080547,
"learning_rate": 1.207319967251835e-05,
"loss": 0.9894,
"step": 1825
},
{
"epoch": 4.670076726342711,
"grad_norm": 0.206715818777659,
"learning_rate": 1.2049785029550903e-05,
"loss": 0.9841,
"step": 1826
},
{
"epoch": 4.672634271099744,
"grad_norm": 0.20935001269390968,
"learning_rate": 1.2026383321086125e-05,
"loss": 0.9872,
"step": 1827
},
{
"epoch": 4.675191815856778,
"grad_norm": 0.2222597208282656,
"learning_rate": 1.2002994585197132e-05,
"loss": 0.9787,
"step": 1828
},
{
"epoch": 4.677749360613811,
"grad_norm": 0.1730464294477197,
"learning_rate": 1.197961885993594e-05,
"loss": 0.9859,
"step": 1829
},
{
"epoch": 4.680306905370844,
"grad_norm": 0.17727350525924096,
"learning_rate": 1.195625618333338e-05,
"loss": 1.0082,
"step": 1830
},
{
"epoch": 4.6828644501278776,
"grad_norm": 0.20017591080053734,
"learning_rate": 1.1932906593399082e-05,
"loss": 0.991,
"step": 1831
},
{
"epoch": 4.68542199488491,
"grad_norm": 0.1767634465416166,
"learning_rate": 1.1909570128121356e-05,
"loss": 1.0151,
"step": 1832
},
{
"epoch": 4.687979539641944,
"grad_norm": 0.18805092012563585,
"learning_rate": 1.1886246825467174e-05,
"loss": 0.9855,
"step": 1833
},
{
"epoch": 4.690537084398977,
"grad_norm": 0.17609948534205547,
"learning_rate": 1.1862936723382107e-05,
"loss": 1.0421,
"step": 1834
},
{
"epoch": 4.69309462915601,
"grad_norm": 0.1901195787568128,
"learning_rate": 1.1839639859790211e-05,
"loss": 1.005,
"step": 1835
},
{
"epoch": 4.695652173913043,
"grad_norm": 0.16909464316510345,
"learning_rate": 1.1816356272594041e-05,
"loss": 1.0146,
"step": 1836
},
{
"epoch": 4.698209718670077,
"grad_norm": 0.20510690288939154,
"learning_rate": 1.179308599967453e-05,
"loss": 0.9991,
"step": 1837
},
{
"epoch": 4.70076726342711,
"grad_norm": 0.17442818365716076,
"learning_rate": 1.1769829078890961e-05,
"loss": 0.9681,
"step": 1838
},
{
"epoch": 4.703324808184143,
"grad_norm": 0.1854925372586028,
"learning_rate": 1.1746585548080884e-05,
"loss": 0.9637,
"step": 1839
},
{
"epoch": 4.705882352941177,
"grad_norm": 0.1790693155042071,
"learning_rate": 1.1723355445060074e-05,
"loss": 1.0259,
"step": 1840
},
{
"epoch": 4.708439897698209,
"grad_norm": 0.1876974891072186,
"learning_rate": 1.170013880762247e-05,
"loss": 0.9866,
"step": 1841
},
{
"epoch": 4.710997442455243,
"grad_norm": 0.16170578076528003,
"learning_rate": 1.1676935673540065e-05,
"loss": 0.9814,
"step": 1842
},
{
"epoch": 4.713554987212277,
"grad_norm": 0.16174685486845114,
"learning_rate": 1.1653746080562924e-05,
"loss": 0.9848,
"step": 1843
},
{
"epoch": 4.716112531969309,
"grad_norm": 0.17164881879927302,
"learning_rate": 1.163057006641906e-05,
"loss": 1.0025,
"step": 1844
},
{
"epoch": 4.718670076726343,
"grad_norm": 0.1677709566656257,
"learning_rate": 1.16074076688144e-05,
"loss": 1.0002,
"step": 1845
},
{
"epoch": 4.721227621483376,
"grad_norm": 0.1587750257090767,
"learning_rate": 1.158425892543271e-05,
"loss": 0.9905,
"step": 1846
},
{
"epoch": 4.723785166240409,
"grad_norm": 0.1565544305505573,
"learning_rate": 1.1561123873935559e-05,
"loss": 0.9955,
"step": 1847
},
{
"epoch": 4.726342710997442,
"grad_norm": 0.16091717894451502,
"learning_rate": 1.1538002551962224e-05,
"loss": 1.0241,
"step": 1848
},
{
"epoch": 4.728900255754476,
"grad_norm": 0.1681586345983381,
"learning_rate": 1.1514894997129643e-05,
"loss": 1.0217,
"step": 1849
},
{
"epoch": 4.731457800511509,
"grad_norm": 0.16299198936177886,
"learning_rate": 1.1491801247032378e-05,
"loss": 0.9896,
"step": 1850
},
{
"epoch": 4.734015345268542,
"grad_norm": 0.165385958962345,
"learning_rate": 1.146872133924251e-05,
"loss": 0.9934,
"step": 1851
},
{
"epoch": 4.736572890025576,
"grad_norm": 0.14578283006875706,
"learning_rate": 1.1445655311309602e-05,
"loss": 0.9844,
"step": 1852
},
{
"epoch": 4.739130434782608,
"grad_norm": 0.17531466138242968,
"learning_rate": 1.1422603200760633e-05,
"loss": 1.0247,
"step": 1853
},
{
"epoch": 4.741687979539642,
"grad_norm": 0.1627069251914456,
"learning_rate": 1.1399565045099965e-05,
"loss": 1.0195,
"step": 1854
},
{
"epoch": 4.744245524296675,
"grad_norm": 0.15204677203913528,
"learning_rate": 1.1376540881809226e-05,
"loss": 0.9958,
"step": 1855
},
{
"epoch": 4.746803069053708,
"grad_norm": 0.16791897543860235,
"learning_rate": 1.1353530748347281e-05,
"loss": 0.9628,
"step": 1856
},
{
"epoch": 4.749360613810742,
"grad_norm": 0.1672771534943014,
"learning_rate": 1.1330534682150195e-05,
"loss": 0.9832,
"step": 1857
},
{
"epoch": 4.751918158567775,
"grad_norm": 0.15558225497361408,
"learning_rate": 1.1307552720631124e-05,
"loss": 1.0164,
"step": 1858
},
{
"epoch": 4.754475703324808,
"grad_norm": 0.15714930539934807,
"learning_rate": 1.1284584901180269e-05,
"loss": 0.9814,
"step": 1859
},
{
"epoch": 4.757033248081841,
"grad_norm": 0.15810762403990258,
"learning_rate": 1.1261631261164858e-05,
"loss": 1.01,
"step": 1860
},
{
"epoch": 4.759590792838875,
"grad_norm": 0.16732531382234792,
"learning_rate": 1.123869183792902e-05,
"loss": 1.0024,
"step": 1861
},
{
"epoch": 4.762148337595908,
"grad_norm": 0.15887859416484115,
"learning_rate": 1.121576666879376e-05,
"loss": 0.9972,
"step": 1862
},
{
"epoch": 4.764705882352941,
"grad_norm": 0.17741158642268076,
"learning_rate": 1.1192855791056886e-05,
"loss": 1.0125,
"step": 1863
},
{
"epoch": 4.767263427109975,
"grad_norm": 0.1740463080508926,
"learning_rate": 1.1169959241992982e-05,
"loss": 1.0163,
"step": 1864
},
{
"epoch": 4.7698209718670075,
"grad_norm": 0.16672250820975146,
"learning_rate": 1.114707705885329e-05,
"loss": 0.9833,
"step": 1865
},
{
"epoch": 4.772378516624041,
"grad_norm": 0.17661338998493217,
"learning_rate": 1.112420927886569e-05,
"loss": 1.0107,
"step": 1866
},
{
"epoch": 4.774936061381074,
"grad_norm": 0.17349399145200417,
"learning_rate": 1.1101355939234639e-05,
"loss": 0.985,
"step": 1867
},
{
"epoch": 4.7774936061381075,
"grad_norm": 0.17935912237443283,
"learning_rate": 1.1078517077141088e-05,
"loss": 0.9852,
"step": 1868
},
{
"epoch": 4.78005115089514,
"grad_norm": 0.16529561605755613,
"learning_rate": 1.1055692729742438e-05,
"loss": 0.9993,
"step": 1869
},
{
"epoch": 4.782608695652174,
"grad_norm": 0.17208070174301834,
"learning_rate": 1.1032882934172463e-05,
"loss": 0.9913,
"step": 1870
},
{
"epoch": 4.7851662404092075,
"grad_norm": 0.16887234918529148,
"learning_rate": 1.1010087727541292e-05,
"loss": 1.0415,
"step": 1871
},
{
"epoch": 4.78772378516624,
"grad_norm": 0.1717215413557606,
"learning_rate": 1.0987307146935295e-05,
"loss": 0.9739,
"step": 1872
},
{
"epoch": 4.790281329923274,
"grad_norm": 0.1673205124395783,
"learning_rate": 1.0964541229417041e-05,
"loss": 0.9814,
"step": 1873
},
{
"epoch": 4.792838874680307,
"grad_norm": 0.17614684912527637,
"learning_rate": 1.0941790012025274e-05,
"loss": 0.9656,
"step": 1874
},
{
"epoch": 4.79539641943734,
"grad_norm": 0.1640615003088743,
"learning_rate": 1.0919053531774794e-05,
"loss": 0.9988,
"step": 1875
},
{
"epoch": 4.797953964194374,
"grad_norm": 0.17206664028124624,
"learning_rate": 1.0896331825656428e-05,
"loss": 0.9642,
"step": 1876
},
{
"epoch": 4.8005115089514065,
"grad_norm": 0.17049514987247102,
"learning_rate": 1.0873624930636967e-05,
"loss": 1.0052,
"step": 1877
},
{
"epoch": 4.80306905370844,
"grad_norm": 0.16443784258614016,
"learning_rate": 1.0850932883659127e-05,
"loss": 0.9953,
"step": 1878
},
{
"epoch": 4.805626598465473,
"grad_norm": 0.17108044985688023,
"learning_rate": 1.0828255721641439e-05,
"loss": 0.9787,
"step": 1879
},
{
"epoch": 4.8081841432225065,
"grad_norm": 0.1518390166966024,
"learning_rate": 1.0805593481478218e-05,
"loss": 0.9747,
"step": 1880
},
{
"epoch": 4.810741687979539,
"grad_norm": 0.17444052631731172,
"learning_rate": 1.0782946200039533e-05,
"loss": 1.0061,
"step": 1881
},
{
"epoch": 4.813299232736573,
"grad_norm": 0.17045779257112384,
"learning_rate": 1.0760313914171072e-05,
"loss": 0.9743,
"step": 1882
},
{
"epoch": 4.8158567774936065,
"grad_norm": 0.18664598792507545,
"learning_rate": 1.0737696660694179e-05,
"loss": 0.9638,
"step": 1883
},
{
"epoch": 4.818414322250639,
"grad_norm": 0.17608135518788834,
"learning_rate": 1.0715094476405677e-05,
"loss": 0.9869,
"step": 1884
},
{
"epoch": 4.820971867007673,
"grad_norm": 0.16283733646395207,
"learning_rate": 1.0692507398077932e-05,
"loss": 0.9862,
"step": 1885
},
{
"epoch": 4.823529411764706,
"grad_norm": 0.2037740964085394,
"learning_rate": 1.0669935462458697e-05,
"loss": 1.003,
"step": 1886
},
{
"epoch": 4.826086956521739,
"grad_norm": 0.17641042448997385,
"learning_rate": 1.0647378706271092e-05,
"loss": 1.0018,
"step": 1887
},
{
"epoch": 4.828644501278772,
"grad_norm": 0.19281863187319848,
"learning_rate": 1.0624837166213568e-05,
"loss": 1.0025,
"step": 1888
},
{
"epoch": 4.831202046035806,
"grad_norm": 0.18029140776618002,
"learning_rate": 1.0602310878959787e-05,
"loss": 0.9879,
"step": 1889
},
{
"epoch": 4.833759590792839,
"grad_norm": 0.19004093543959594,
"learning_rate": 1.057979988115863e-05,
"loss": 0.9667,
"step": 1890
},
{
"epoch": 4.836317135549872,
"grad_norm": 0.16654767546768368,
"learning_rate": 1.0557304209434052e-05,
"loss": 1.0015,
"step": 1891
},
{
"epoch": 4.838874680306906,
"grad_norm": 0.188493295364345,
"learning_rate": 1.0534823900385116e-05,
"loss": 0.9695,
"step": 1892
},
{
"epoch": 4.841432225063938,
"grad_norm": 0.22005346376142065,
"learning_rate": 1.0512358990585905e-05,
"loss": 0.9458,
"step": 1893
},
{
"epoch": 4.843989769820972,
"grad_norm": 0.18579628706472415,
"learning_rate": 1.0489909516585381e-05,
"loss": 0.9984,
"step": 1894
},
{
"epoch": 4.846547314578006,
"grad_norm": 0.18129273418004668,
"learning_rate": 1.0467475514907465e-05,
"loss": 0.9861,
"step": 1895
},
{
"epoch": 4.849104859335038,
"grad_norm": 0.17569270887058805,
"learning_rate": 1.0445057022050858e-05,
"loss": 1.0223,
"step": 1896
},
{
"epoch": 4.851662404092072,
"grad_norm": 0.17039473537270847,
"learning_rate": 1.0422654074489067e-05,
"loss": 0.9883,
"step": 1897
},
{
"epoch": 4.854219948849105,
"grad_norm": 0.1776915959730032,
"learning_rate": 1.0400266708670257e-05,
"loss": 0.9983,
"step": 1898
},
{
"epoch": 4.856777493606138,
"grad_norm": 0.17346690618210883,
"learning_rate": 1.0377894961017288e-05,
"loss": 1.0295,
"step": 1899
},
{
"epoch": 4.859335038363171,
"grad_norm": 0.18204686027958067,
"learning_rate": 1.0355538867927602e-05,
"loss": 0.9787,
"step": 1900
},
{
"epoch": 4.861892583120205,
"grad_norm": 0.154712425541886,
"learning_rate": 1.0333198465773158e-05,
"loss": 0.9818,
"step": 1901
},
{
"epoch": 4.864450127877237,
"grad_norm": 0.15872615012895688,
"learning_rate": 1.0310873790900393e-05,
"loss": 0.9912,
"step": 1902
},
{
"epoch": 4.867007672634271,
"grad_norm": 0.1679270811012402,
"learning_rate": 1.028856487963015e-05,
"loss": 0.9988,
"step": 1903
},
{
"epoch": 4.869565217391305,
"grad_norm": 0.160132588616313,
"learning_rate": 1.0266271768257656e-05,
"loss": 1.0111,
"step": 1904
},
{
"epoch": 4.872122762148337,
"grad_norm": 0.15372454082126236,
"learning_rate": 1.0243994493052375e-05,
"loss": 1.0024,
"step": 1905
},
{
"epoch": 4.874680306905371,
"grad_norm": 0.16751023919866065,
"learning_rate": 1.0221733090258062e-05,
"loss": 0.9912,
"step": 1906
},
{
"epoch": 4.877237851662404,
"grad_norm": 0.1638777023547693,
"learning_rate": 1.019948759609263e-05,
"loss": 0.9988,
"step": 1907
},
{
"epoch": 4.879795396419437,
"grad_norm": 0.1577620238658875,
"learning_rate": 1.0177258046748104e-05,
"loss": 1.0205,
"step": 1908
},
{
"epoch": 4.882352941176471,
"grad_norm": 0.1569686158650272,
"learning_rate": 1.0155044478390565e-05,
"loss": 1.0281,
"step": 1909
},
{
"epoch": 4.884910485933504,
"grad_norm": 0.1529303225661744,
"learning_rate": 1.01328469271601e-05,
"loss": 0.9888,
"step": 1910
},
{
"epoch": 4.887468030690537,
"grad_norm": 0.1634116165943236,
"learning_rate": 1.0110665429170744e-05,
"loss": 0.9701,
"step": 1911
},
{
"epoch": 4.89002557544757,
"grad_norm": 0.19647349119025276,
"learning_rate": 1.0088500020510405e-05,
"loss": 1.0114,
"step": 1912
},
{
"epoch": 4.892583120204604,
"grad_norm": 0.15265968325053902,
"learning_rate": 1.0066350737240812e-05,
"loss": 1.0001,
"step": 1913
},
{
"epoch": 4.8951406649616365,
"grad_norm": 0.16529435385499702,
"learning_rate": 1.004421761539747e-05,
"loss": 1.0049,
"step": 1914
},
{
"epoch": 4.89769820971867,
"grad_norm": 0.16470146034286107,
"learning_rate": 1.0022100690989585e-05,
"loss": 0.9951,
"step": 1915
},
{
"epoch": 4.900255754475703,
"grad_norm": 0.1802490098235214,
"learning_rate": 1.0000000000000006e-05,
"loss": 1.0178,
"step": 1916
},
{
"epoch": 4.9028132992327365,
"grad_norm": 0.16330121673094,
"learning_rate": 9.977915578385166e-06,
"loss": 0.9735,
"step": 1917
},
{
"epoch": 4.90537084398977,
"grad_norm": 0.15055201137991753,
"learning_rate": 9.95584746207506e-06,
"loss": 0.947,
"step": 1918
},
{
"epoch": 4.907928388746803,
"grad_norm": 0.1724419930390039,
"learning_rate": 9.933795686973122e-06,
"loss": 1.0247,
"step": 1919
},
{
"epoch": 4.910485933503836,
"grad_norm": 0.16116276079347294,
"learning_rate": 9.911760288956202e-06,
"loss": 0.9867,
"step": 1920
},
{
"epoch": 4.913043478260869,
"grad_norm": 0.1774028033969299,
"learning_rate": 9.889741303874534e-06,
"loss": 1.0021,
"step": 1921
},
{
"epoch": 4.915601023017903,
"grad_norm": 0.1636395605344455,
"learning_rate": 9.867738767551618e-06,
"loss": 0.9975,
"step": 1922
},
{
"epoch": 4.918158567774936,
"grad_norm": 0.15970153619108604,
"learning_rate": 9.845752715784206e-06,
"loss": 0.9776,
"step": 1923
},
{
"epoch": 4.920716112531969,
"grad_norm": 0.16442665798334838,
"learning_rate": 9.823783184342226e-06,
"loss": 0.9824,
"step": 1924
},
{
"epoch": 4.923273657289003,
"grad_norm": 0.15506730043358224,
"learning_rate": 9.801830208968744e-06,
"loss": 1.0188,
"step": 1925
},
{
"epoch": 4.9258312020460355,
"grad_norm": 0.16382228085181347,
"learning_rate": 9.779893825379872e-06,
"loss": 1.0142,
"step": 1926
},
{
"epoch": 4.928388746803069,
"grad_norm": 0.1685251514559618,
"learning_rate": 9.757974069264723e-06,
"loss": 1.0176,
"step": 1927
},
{
"epoch": 4.930946291560103,
"grad_norm": 0.16153615725351078,
"learning_rate": 9.736070976285387e-06,
"loss": 0.9797,
"step": 1928
},
{
"epoch": 4.9335038363171355,
"grad_norm": 0.15930542918956977,
"learning_rate": 9.714184582076822e-06,
"loss": 1.0044,
"step": 1929
},
{
"epoch": 4.936061381074169,
"grad_norm": 0.16037126450106265,
"learning_rate": 9.692314922246822e-06,
"loss": 1.0237,
"step": 1930
},
{
"epoch": 4.938618925831202,
"grad_norm": 0.16699498508297592,
"learning_rate": 9.670462032375945e-06,
"loss": 1.0201,
"step": 1931
},
{
"epoch": 4.9411764705882355,
"grad_norm": 0.16811969634518498,
"learning_rate": 9.648625948017495e-06,
"loss": 1.0033,
"step": 1932
},
{
"epoch": 4.943734015345268,
"grad_norm": 0.16111796966728506,
"learning_rate": 9.626806704697406e-06,
"loss": 0.972,
"step": 1933
},
{
"epoch": 4.946291560102302,
"grad_norm": 0.16532847508914839,
"learning_rate": 9.605004337914219e-06,
"loss": 0.9949,
"step": 1934
},
{
"epoch": 4.948849104859335,
"grad_norm": 0.17761356685428417,
"learning_rate": 9.583218883139034e-06,
"loss": 0.9822,
"step": 1935
},
{
"epoch": 4.951406649616368,
"grad_norm": 0.15358178200013056,
"learning_rate": 9.561450375815415e-06,
"loss": 0.9896,
"step": 1936
},
{
"epoch": 4.953964194373402,
"grad_norm": 0.15559399612064156,
"learning_rate": 9.539698851359365e-06,
"loss": 0.9911,
"step": 1937
},
{
"epoch": 4.956521739130435,
"grad_norm": 0.18428232448870238,
"learning_rate": 9.517964345159242e-06,
"loss": 0.9664,
"step": 1938
},
{
"epoch": 4.959079283887468,
"grad_norm": 0.1669486508989526,
"learning_rate": 9.49624689257575e-06,
"loss": 1.0006,
"step": 1939
},
{
"epoch": 4.961636828644501,
"grad_norm": 0.20192530957377794,
"learning_rate": 9.474546528941818e-06,
"loss": 1.0148,
"step": 1940
},
{
"epoch": 4.964194373401535,
"grad_norm": 0.1822554701624732,
"learning_rate": 9.452863289562568e-06,
"loss": 0.9929,
"step": 1941
},
{
"epoch": 4.966751918158568,
"grad_norm": 0.16044727932832298,
"learning_rate": 9.431197209715296e-06,
"loss": 1.0061,
"step": 1942
},
{
"epoch": 4.969309462915601,
"grad_norm": 0.20457445060854093,
"learning_rate": 9.409548324649346e-06,
"loss": 0.9976,
"step": 1943
},
{
"epoch": 4.971867007672635,
"grad_norm": 0.16528723607998885,
"learning_rate": 9.387916669586105e-06,
"loss": 0.9713,
"step": 1944
},
{
"epoch": 4.974424552429667,
"grad_norm": 0.20042727255973036,
"learning_rate": 9.366302279718915e-06,
"loss": 0.9909,
"step": 1945
},
{
"epoch": 4.976982097186701,
"grad_norm": 0.18545039581190098,
"learning_rate": 9.34470519021305e-06,
"loss": 1.0272,
"step": 1946
},
{
"epoch": 4.979539641943734,
"grad_norm": 0.17303232623661333,
"learning_rate": 9.323125436205616e-06,
"loss": 0.9747,
"step": 1947
},
{
"epoch": 4.982097186700767,
"grad_norm": 0.18212413997277496,
"learning_rate": 9.30156305280552e-06,
"loss": 0.9907,
"step": 1948
},
{
"epoch": 4.9846547314578,
"grad_norm": 0.16050142721711194,
"learning_rate": 9.280018075093421e-06,
"loss": 1.021,
"step": 1949
},
{
"epoch": 4.987212276214834,
"grad_norm": 0.1658254050143272,
"learning_rate": 9.258490538121651e-06,
"loss": 1.0146,
"step": 1950
},
{
"epoch": 4.989769820971867,
"grad_norm": 0.15882488971967954,
"learning_rate": 9.23698047691416e-06,
"loss": 0.9807,
"step": 1951
},
{
"epoch": 4.9923273657289,
"grad_norm": 0.15137068577749335,
"learning_rate": 9.215487926466467e-06,
"loss": 1.0057,
"step": 1952
},
{
"epoch": 4.994884910485934,
"grad_norm": 0.1633761648679145,
"learning_rate": 9.194012921745617e-06,
"loss": 0.9687,
"step": 1953
},
{
"epoch": 4.997442455242966,
"grad_norm": 0.16515189492129442,
"learning_rate": 9.172555497690112e-06,
"loss": 0.9746,
"step": 1954
},
{
"epoch": 5.0,
"grad_norm": 0.1676688960168738,
"learning_rate": 9.15111568920981e-06,
"loss": 0.9944,
"step": 1955
},
{
"epoch": 5.002557544757034,
"grad_norm": 0.18099666721489213,
"learning_rate": 9.129693531185963e-06,
"loss": 0.9851,
"step": 1956
},
{
"epoch": 5.005115089514066,
"grad_norm": 0.16948193699362832,
"learning_rate": 9.108289058471074e-06,
"loss": 0.9931,
"step": 1957
},
{
"epoch": 5.0076726342711,
"grad_norm": 0.17226736786957128,
"learning_rate": 9.086902305888876e-06,
"loss": 1.0204,
"step": 1958
},
{
"epoch": 5.010230179028133,
"grad_norm": 0.1908726630037535,
"learning_rate": 9.065533308234292e-06,
"loss": 0.9957,
"step": 1959
},
{
"epoch": 5.012787723785166,
"grad_norm": 0.17384644413167735,
"learning_rate": 9.044182100273327e-06,
"loss": 1.0009,
"step": 1960
},
{
"epoch": 5.015345268542199,
"grad_norm": 0.1674331625223336,
"learning_rate": 9.022848716743089e-06,
"loss": 1.0006,
"step": 1961
},
{
"epoch": 5.017902813299233,
"grad_norm": 0.1552271777958725,
"learning_rate": 9.001533192351627e-06,
"loss": 0.9691,
"step": 1962
},
{
"epoch": 5.020460358056266,
"grad_norm": 0.17113010898536238,
"learning_rate": 8.980235561777988e-06,
"loss": 0.9843,
"step": 1963
},
{
"epoch": 5.023017902813299,
"grad_norm": 0.152759349994963,
"learning_rate": 8.958955859672078e-06,
"loss": 0.9871,
"step": 1964
},
{
"epoch": 5.025575447570333,
"grad_norm": 0.1708561684548924,
"learning_rate": 8.937694120654637e-06,
"loss": 0.9645,
"step": 1965
},
{
"epoch": 5.028132992327365,
"grad_norm": 0.1488633460735635,
"learning_rate": 8.916450379317197e-06,
"loss": 0.9895,
"step": 1966
},
{
"epoch": 5.030690537084399,
"grad_norm": 0.1569268567376542,
"learning_rate": 8.895224670221982e-06,
"loss": 0.9801,
"step": 1967
},
{
"epoch": 5.033248081841432,
"grad_norm": 0.15914255768698204,
"learning_rate": 8.87401702790192e-06,
"loss": 0.9713,
"step": 1968
},
{
"epoch": 5.035805626598465,
"grad_norm": 0.15591363388314214,
"learning_rate": 8.852827486860482e-06,
"loss": 1.0127,
"step": 1969
},
{
"epoch": 5.038363171355499,
"grad_norm": 0.1421340058755138,
"learning_rate": 8.831656081571755e-06,
"loss": 0.9704,
"step": 1970
},
{
"epoch": 5.040920716112532,
"grad_norm": 0.1490955221802277,
"learning_rate": 8.810502846480275e-06,
"loss": 0.9806,
"step": 1971
},
{
"epoch": 5.043478260869565,
"grad_norm": 0.1450434219687613,
"learning_rate": 8.789367816001047e-06,
"loss": 0.9819,
"step": 1972
},
{
"epoch": 5.046035805626598,
"grad_norm": 0.1400644837923175,
"learning_rate": 8.768251024519437e-06,
"loss": 1.0056,
"step": 1973
},
{
"epoch": 5.048593350383632,
"grad_norm": 0.15964550115662493,
"learning_rate": 8.747152506391136e-06,
"loss": 0.9715,
"step": 1974
},
{
"epoch": 5.051150895140665,
"grad_norm": 0.14995688601170076,
"learning_rate": 8.726072295942139e-06,
"loss": 0.9617,
"step": 1975
},
{
"epoch": 5.053708439897698,
"grad_norm": 0.15374377985504548,
"learning_rate": 8.705010427468597e-06,
"loss": 0.9852,
"step": 1976
},
{
"epoch": 5.056265984654732,
"grad_norm": 0.1404025125857731,
"learning_rate": 8.683966935236879e-06,
"loss": 0.9875,
"step": 1977
},
{
"epoch": 5.0588235294117645,
"grad_norm": 0.16021215711079942,
"learning_rate": 8.662941853483413e-06,
"loss": 0.9913,
"step": 1978
},
{
"epoch": 5.061381074168798,
"grad_norm": 0.16366042535437825,
"learning_rate": 8.641935216414707e-06,
"loss": 0.9908,
"step": 1979
},
{
"epoch": 5.063938618925831,
"grad_norm": 0.16425024674822647,
"learning_rate": 8.620947058207243e-06,
"loss": 0.9903,
"step": 1980
},
{
"epoch": 5.0664961636828645,
"grad_norm": 0.1490290343024626,
"learning_rate": 8.599977413007432e-06,
"loss": 0.955,
"step": 1981
},
{
"epoch": 5.069053708439898,
"grad_norm": 0.16191952820253272,
"learning_rate": 8.579026314931591e-06,
"loss": 0.9723,
"step": 1982
},
{
"epoch": 5.071611253196931,
"grad_norm": 0.17099841565059007,
"learning_rate": 8.558093798065836e-06,
"loss": 0.9929,
"step": 1983
},
{
"epoch": 5.0741687979539645,
"grad_norm": 0.1480089282201921,
"learning_rate": 8.537179896466067e-06,
"loss": 0.9467,
"step": 1984
},
{
"epoch": 5.076726342710997,
"grad_norm": 0.16095867656678275,
"learning_rate": 8.516284644157889e-06,
"loss": 0.9799,
"step": 1985
},
{
"epoch": 5.079283887468031,
"grad_norm": 0.16018966441136562,
"learning_rate": 8.495408075136578e-06,
"loss": 1.0128,
"step": 1986
},
{
"epoch": 5.081841432225064,
"grad_norm": 0.15299320218884427,
"learning_rate": 8.474550223367005e-06,
"loss": 0.9828,
"step": 1987
},
{
"epoch": 5.084398976982097,
"grad_norm": 0.1758996345247043,
"learning_rate": 8.453711122783584e-06,
"loss": 0.989,
"step": 1988
},
{
"epoch": 5.086956521739131,
"grad_norm": 0.15643081742771217,
"learning_rate": 8.432890807290237e-06,
"loss": 0.976,
"step": 1989
},
{
"epoch": 5.089514066496164,
"grad_norm": 0.18121284115543446,
"learning_rate": 8.412089310760317e-06,
"loss": 1.0213,
"step": 1990
},
{
"epoch": 5.092071611253197,
"grad_norm": 0.14776931134366916,
"learning_rate": 8.391306667036542e-06,
"loss": 0.9961,
"step": 1991
},
{
"epoch": 5.09462915601023,
"grad_norm": 0.1502265846240369,
"learning_rate": 8.370542909930997e-06,
"loss": 0.9937,
"step": 1992
},
{
"epoch": 5.0971867007672635,
"grad_norm": 0.16833071547762432,
"learning_rate": 8.349798073225001e-06,
"loss": 0.9751,
"step": 1993
},
{
"epoch": 5.099744245524296,
"grad_norm": 0.1430282061195032,
"learning_rate": 8.329072190669113e-06,
"loss": 0.9707,
"step": 1994
},
{
"epoch": 5.10230179028133,
"grad_norm": 0.16237907550835082,
"learning_rate": 8.308365295983038e-06,
"loss": 1.0136,
"step": 1995
},
{
"epoch": 5.1048593350383635,
"grad_norm": 0.1434507210201212,
"learning_rate": 8.287677422855616e-06,
"loss": 0.9761,
"step": 1996
},
{
"epoch": 5.107416879795396,
"grad_norm": 0.15348430469648483,
"learning_rate": 8.267008604944711e-06,
"loss": 0.9675,
"step": 1997
},
{
"epoch": 5.10997442455243,
"grad_norm": 0.14766337650581302,
"learning_rate": 8.246358875877192e-06,
"loss": 1.004,
"step": 1998
},
{
"epoch": 5.112531969309463,
"grad_norm": 0.14292853816072373,
"learning_rate": 8.225728269248894e-06,
"loss": 1.0055,
"step": 1999
},
{
"epoch": 5.115089514066496,
"grad_norm": 0.14723259669842279,
"learning_rate": 8.205116818624512e-06,
"loss": 0.9872,
"step": 2000
},
{
"epoch": 5.117647058823529,
"grad_norm": 0.14942990843521783,
"learning_rate": 8.184524557537594e-06,
"loss": 0.9783,
"step": 2001
},
{
"epoch": 5.120204603580563,
"grad_norm": 0.1649963182635304,
"learning_rate": 8.163951519490449e-06,
"loss": 0.9982,
"step": 2002
},
{
"epoch": 5.122762148337596,
"grad_norm": 0.15532573004314068,
"learning_rate": 8.143397737954139e-06,
"loss": 0.973,
"step": 2003
},
{
"epoch": 5.125319693094629,
"grad_norm": 0.15878378580693214,
"learning_rate": 8.122863246368376e-06,
"loss": 0.9942,
"step": 2004
},
{
"epoch": 5.127877237851663,
"grad_norm": 0.15508711480763188,
"learning_rate": 8.102348078141485e-06,
"loss": 0.9476,
"step": 2005
},
{
"epoch": 5.130434782608695,
"grad_norm": 0.14459159704303187,
"learning_rate": 8.081852266650376e-06,
"loss": 0.9719,
"step": 2006
},
{
"epoch": 5.132992327365729,
"grad_norm": 0.14688305770537563,
"learning_rate": 8.061375845240446e-06,
"loss": 0.9832,
"step": 2007
},
{
"epoch": 5.135549872122763,
"grad_norm": 0.16411144821788345,
"learning_rate": 8.040918847225555e-06,
"loss": 0.9983,
"step": 2008
},
{
"epoch": 5.138107416879795,
"grad_norm": 0.15210493577756426,
"learning_rate": 8.020481305887948e-06,
"loss": 0.9902,
"step": 2009
},
{
"epoch": 5.140664961636829,
"grad_norm": 0.15778749490785482,
"learning_rate": 8.000063254478246e-06,
"loss": 0.9922,
"step": 2010
},
{
"epoch": 5.143222506393862,
"grad_norm": 0.1445318240266244,
"learning_rate": 7.979664726215328e-06,
"loss": 0.9396,
"step": 2011
},
{
"epoch": 5.145780051150895,
"grad_norm": 0.15296331641236172,
"learning_rate": 7.959285754286325e-06,
"loss": 0.9858,
"step": 2012
},
{
"epoch": 5.148337595907928,
"grad_norm": 0.1569440999900595,
"learning_rate": 7.93892637184656e-06,
"loss": 0.9603,
"step": 2013
},
{
"epoch": 5.150895140664962,
"grad_norm": 0.1468065534744029,
"learning_rate": 7.918586612019467e-06,
"loss": 1.0245,
"step": 2014
},
{
"epoch": 5.153452685421995,
"grad_norm": 0.14361088172227368,
"learning_rate": 7.898266507896568e-06,
"loss": 0.9695,
"step": 2015
},
{
"epoch": 5.156010230179028,
"grad_norm": 0.1529262347644108,
"learning_rate": 7.87796609253739e-06,
"loss": 0.9528,
"step": 2016
},
{
"epoch": 5.158567774936062,
"grad_norm": 0.14353857774626,
"learning_rate": 7.857685398969457e-06,
"loss": 0.9892,
"step": 2017
},
{
"epoch": 5.161125319693094,
"grad_norm": 0.15057417202588838,
"learning_rate": 7.837424460188182e-06,
"loss": 1.0168,
"step": 2018
},
{
"epoch": 5.163682864450128,
"grad_norm": 0.16123236986875264,
"learning_rate": 7.817183309156835e-06,
"loss": 0.9543,
"step": 2019
},
{
"epoch": 5.166240409207161,
"grad_norm": 0.1428021869549548,
"learning_rate": 7.79696197880652e-06,
"loss": 0.986,
"step": 2020
},
{
"epoch": 5.168797953964194,
"grad_norm": 0.14448129081863484,
"learning_rate": 7.77676050203607e-06,
"loss": 0.9767,
"step": 2021
},
{
"epoch": 5.171355498721228,
"grad_norm": 0.18175091871047533,
"learning_rate": 7.756578911712027e-06,
"loss": 1.0241,
"step": 2022
},
{
"epoch": 5.173913043478261,
"grad_norm": 0.15757011750152183,
"learning_rate": 7.736417240668565e-06,
"loss": 0.9715,
"step": 2023
},
{
"epoch": 5.176470588235294,
"grad_norm": 0.15226912495310158,
"learning_rate": 7.716275521707475e-06,
"loss": 0.9843,
"step": 2024
},
{
"epoch": 5.179028132992327,
"grad_norm": 0.16291607853359827,
"learning_rate": 7.69615378759808e-06,
"loss": 0.9708,
"step": 2025
},
{
"epoch": 5.181585677749361,
"grad_norm": 0.15902273574242684,
"learning_rate": 7.676052071077162e-06,
"loss": 0.9612,
"step": 2026
},
{
"epoch": 5.1841432225063935,
"grad_norm": 0.17064610049146775,
"learning_rate": 7.655970404848976e-06,
"loss": 0.9888,
"step": 2027
},
{
"epoch": 5.186700767263427,
"grad_norm": 0.1666779628304091,
"learning_rate": 7.63590882158513e-06,
"loss": 0.9794,
"step": 2028
},
{
"epoch": 5.189258312020461,
"grad_norm": 0.15564555745227387,
"learning_rate": 7.615867353924567e-06,
"loss": 0.9939,
"step": 2029
},
{
"epoch": 5.1918158567774935,
"grad_norm": 0.17454854500347408,
"learning_rate": 7.595846034473491e-06,
"loss": 0.9802,
"step": 2030
},
{
"epoch": 5.194373401534527,
"grad_norm": 0.15404308368398176,
"learning_rate": 7.575844895805349e-06,
"loss": 0.9722,
"step": 2031
},
{
"epoch": 5.19693094629156,
"grad_norm": 0.14924679147749503,
"learning_rate": 7.555863970460753e-06,
"loss": 1.0215,
"step": 2032
},
{
"epoch": 5.1994884910485935,
"grad_norm": 0.16885590187685048,
"learning_rate": 7.535903290947395e-06,
"loss": 1.0097,
"step": 2033
},
{
"epoch": 5.202046035805626,
"grad_norm": 0.1516369965260777,
"learning_rate": 7.51596288974007e-06,
"loss": 0.9617,
"step": 2034
},
{
"epoch": 5.20460358056266,
"grad_norm": 0.1575797779466588,
"learning_rate": 7.496042799280558e-06,
"loss": 0.9928,
"step": 2035
},
{
"epoch": 5.207161125319693,
"grad_norm": 0.154725543125773,
"learning_rate": 7.4761430519776e-06,
"loss": 0.9994,
"step": 2036
},
{
"epoch": 5.209718670076726,
"grad_norm": 0.15196363634156476,
"learning_rate": 7.456263680206837e-06,
"loss": 0.9607,
"step": 2037
},
{
"epoch": 5.21227621483376,
"grad_norm": 0.1523028526150447,
"learning_rate": 7.436404716310768e-06,
"loss": 1.0193,
"step": 2038
},
{
"epoch": 5.2148337595907925,
"grad_norm": 0.14260045213784886,
"learning_rate": 7.4165661925986994e-06,
"loss": 1.0036,
"step": 2039
},
{
"epoch": 5.217391304347826,
"grad_norm": 0.14141430407744374,
"learning_rate": 7.396748141346641e-06,
"loss": 0.9605,
"step": 2040
},
{
"epoch": 5.21994884910486,
"grad_norm": 0.15491487724967554,
"learning_rate": 7.376950594797348e-06,
"loss": 1.002,
"step": 2041
},
{
"epoch": 5.2225063938618925,
"grad_norm": 0.13670916080763118,
"learning_rate": 7.357173585160171e-06,
"loss": 0.9682,
"step": 2042
},
{
"epoch": 5.225063938618926,
"grad_norm": 0.16904021029532304,
"learning_rate": 7.337417144611092e-06,
"loss": 1.003,
"step": 2043
},
{
"epoch": 5.227621483375959,
"grad_norm": 0.14732288084215725,
"learning_rate": 7.317681305292574e-06,
"loss": 0.9679,
"step": 2044
},
{
"epoch": 5.2301790281329925,
"grad_norm": 0.14853256792950575,
"learning_rate": 7.297966099313613e-06,
"loss": 0.9873,
"step": 2045
},
{
"epoch": 5.232736572890025,
"grad_norm": 0.1682802842992428,
"learning_rate": 7.278271558749626e-06,
"loss": 1.0112,
"step": 2046
},
{
"epoch": 5.235294117647059,
"grad_norm": 0.16296049149402267,
"learning_rate": 7.2585977156423726e-06,
"loss": 0.9899,
"step": 2047
},
{
"epoch": 5.2378516624040925,
"grad_norm": 0.15621140127516103,
"learning_rate": 7.238944601999984e-06,
"loss": 0.9675,
"step": 2048
},
{
"epoch": 5.240409207161125,
"grad_norm": 0.16479478631029945,
"learning_rate": 7.219312249796837e-06,
"loss": 1.0021,
"step": 2049
},
{
"epoch": 5.242966751918159,
"grad_norm": 0.17367084886794448,
"learning_rate": 7.1997006909735615e-06,
"loss": 0.9861,
"step": 2050
},
{
"epoch": 5.245524296675192,
"grad_norm": 0.16299961418797604,
"learning_rate": 7.180109957436907e-06,
"loss": 0.9743,
"step": 2051
},
{
"epoch": 5.248081841432225,
"grad_norm": 0.1889746519267133,
"learning_rate": 7.160540081059788e-06,
"loss": 1.0108,
"step": 2052
},
{
"epoch": 5.250639386189258,
"grad_norm": 0.14975830182222835,
"learning_rate": 7.140991093681182e-06,
"loss": 0.9968,
"step": 2053
},
{
"epoch": 5.253196930946292,
"grad_norm": 0.161164131121645,
"learning_rate": 7.121463027106044e-06,
"loss": 0.9983,
"step": 2054
},
{
"epoch": 5.255754475703325,
"grad_norm": 0.1563912353363099,
"learning_rate": 7.101955913105337e-06,
"loss": 0.9995,
"step": 2055
},
{
"epoch": 5.258312020460358,
"grad_norm": 0.2387687897967918,
"learning_rate": 7.082469783415901e-06,
"loss": 1.0272,
"step": 2056
},
{
"epoch": 5.260869565217392,
"grad_norm": 0.14955838008350864,
"learning_rate": 7.063004669740465e-06,
"loss": 1.0012,
"step": 2057
},
{
"epoch": 5.263427109974424,
"grad_norm": 0.14783879968843747,
"learning_rate": 7.043560603747546e-06,
"loss": 0.9664,
"step": 2058
},
{
"epoch": 5.265984654731458,
"grad_norm": 0.1588528569352959,
"learning_rate": 7.024137617071414e-06,
"loss": 0.9802,
"step": 2059
},
{
"epoch": 5.268542199488491,
"grad_norm": 0.14942500255272206,
"learning_rate": 7.00473574131207e-06,
"loss": 0.9935,
"step": 2060
},
{
"epoch": 5.271099744245524,
"grad_norm": 0.16317313745637238,
"learning_rate": 6.985355008035146e-06,
"loss": 0.9685,
"step": 2061
},
{
"epoch": 5.273657289002558,
"grad_norm": 0.15488856381511182,
"learning_rate": 6.965995448771883e-06,
"loss": 0.9757,
"step": 2062
},
{
"epoch": 5.276214833759591,
"grad_norm": 0.16488364428538402,
"learning_rate": 6.946657095019065e-06,
"loss": 0.9912,
"step": 2063
},
{
"epoch": 5.278772378516624,
"grad_norm": 0.15829880837334376,
"learning_rate": 6.927339978239e-06,
"loss": 0.9831,
"step": 2064
},
{
"epoch": 5.281329923273657,
"grad_norm": 0.15536053261624375,
"learning_rate": 6.908044129859422e-06,
"loss": 1.0008,
"step": 2065
},
{
"epoch": 5.283887468030691,
"grad_norm": 0.14858756885177146,
"learning_rate": 6.8887695812734626e-06,
"loss": 0.9658,
"step": 2066
},
{
"epoch": 5.286445012787723,
"grad_norm": 0.14638447382269204,
"learning_rate": 6.869516363839619e-06,
"loss": 0.9728,
"step": 2067
},
{
"epoch": 5.289002557544757,
"grad_norm": 0.1558382558798289,
"learning_rate": 6.85028450888167e-06,
"loss": 0.9967,
"step": 2068
},
{
"epoch": 5.291560102301791,
"grad_norm": 0.14674124861304347,
"learning_rate": 6.831074047688637e-06,
"loss": 0.9591,
"step": 2069
},
{
"epoch": 5.294117647058823,
"grad_norm": 0.15477443219836143,
"learning_rate": 6.811885011514734e-06,
"loss": 0.9889,
"step": 2070
},
{
"epoch": 5.296675191815857,
"grad_norm": 0.1631809713084356,
"learning_rate": 6.792717431579339e-06,
"loss": 0.9548,
"step": 2071
},
{
"epoch": 5.29923273657289,
"grad_norm": 0.15337746503295482,
"learning_rate": 6.773571339066894e-06,
"loss": 1.0256,
"step": 2072
},
{
"epoch": 5.301790281329923,
"grad_norm": 0.15540454976804952,
"learning_rate": 6.754446765126892e-06,
"loss": 0.9848,
"step": 2073
},
{
"epoch": 5.304347826086957,
"grad_norm": 0.16573456540761827,
"learning_rate": 6.73534374087383e-06,
"loss": 1.0387,
"step": 2074
},
{
"epoch": 5.30690537084399,
"grad_norm": 0.13958543991175704,
"learning_rate": 6.716262297387128e-06,
"loss": 0.9857,
"step": 2075
},
{
"epoch": 5.309462915601023,
"grad_norm": 0.15650175476452233,
"learning_rate": 6.697202465711101e-06,
"loss": 1.0011,
"step": 2076
},
{
"epoch": 5.312020460358056,
"grad_norm": 0.1511539473738923,
"learning_rate": 6.6781642768549e-06,
"loss": 1.0074,
"step": 2077
},
{
"epoch": 5.31457800511509,
"grad_norm": 0.1464895523912939,
"learning_rate": 6.6591477617924775e-06,
"loss": 1.0044,
"step": 2078
},
{
"epoch": 5.3171355498721224,
"grad_norm": 0.1454929422658069,
"learning_rate": 6.640152951462511e-06,
"loss": 0.9869,
"step": 2079
},
{
"epoch": 5.319693094629156,
"grad_norm": 0.14945439707003583,
"learning_rate": 6.621179876768358e-06,
"loss": 0.9899,
"step": 2080
},
{
"epoch": 5.322250639386189,
"grad_norm": 0.15337676586906251,
"learning_rate": 6.602228568578042e-06,
"loss": 0.9883,
"step": 2081
},
{
"epoch": 5.324808184143222,
"grad_norm": 0.1507465223109623,
"learning_rate": 6.583299057724148e-06,
"loss": 1.0014,
"step": 2082
},
{
"epoch": 5.327365728900256,
"grad_norm": 0.13762779144101234,
"learning_rate": 6.564391375003809e-06,
"loss": 0.9639,
"step": 2083
},
{
"epoch": 5.329923273657289,
"grad_norm": 0.1494704878685422,
"learning_rate": 6.545505551178632e-06,
"loss": 0.9913,
"step": 2084
},
{
"epoch": 5.332480818414322,
"grad_norm": 0.14263285135209475,
"learning_rate": 6.526641616974689e-06,
"loss": 0.9873,
"step": 2085
},
{
"epoch": 5.335038363171355,
"grad_norm": 0.14614833258914145,
"learning_rate": 6.507799603082412e-06,
"loss": 1.0108,
"step": 2086
},
{
"epoch": 5.337595907928389,
"grad_norm": 0.16842621203139152,
"learning_rate": 6.488979540156572e-06,
"loss": 1.0346,
"step": 2087
},
{
"epoch": 5.340153452685422,
"grad_norm": 0.142363434210902,
"learning_rate": 6.4701814588162515e-06,
"loss": 0.9596,
"step": 2088
},
{
"epoch": 5.342710997442455,
"grad_norm": 0.15550922574842538,
"learning_rate": 6.451405389644745e-06,
"loss": 1.0206,
"step": 2089
},
{
"epoch": 5.345268542199489,
"grad_norm": 0.14961001828366297,
"learning_rate": 6.432651363189535e-06,
"loss": 0.9653,
"step": 2090
},
{
"epoch": 5.3478260869565215,
"grad_norm": 0.14864220061481173,
"learning_rate": 6.413919409962266e-06,
"loss": 1.0209,
"step": 2091
},
{
"epoch": 5.350383631713555,
"grad_norm": 0.14931524867855708,
"learning_rate": 6.395209560438646e-06,
"loss": 0.9326,
"step": 2092
},
{
"epoch": 5.352941176470588,
"grad_norm": 0.1660101727327659,
"learning_rate": 6.37652184505843e-06,
"loss": 1.0082,
"step": 2093
},
{
"epoch": 5.3554987212276215,
"grad_norm": 0.15052870499794585,
"learning_rate": 6.357856294225357e-06,
"loss": 0.9833,
"step": 2094
},
{
"epoch": 5.358056265984655,
"grad_norm": 0.15607136544629463,
"learning_rate": 6.339212938307127e-06,
"loss": 0.9685,
"step": 2095
},
{
"epoch": 5.360613810741688,
"grad_norm": 0.14780192421755606,
"learning_rate": 6.320591807635303e-06,
"loss": 0.9814,
"step": 2096
},
{
"epoch": 5.3631713554987215,
"grad_norm": 0.15043709275941075,
"learning_rate": 6.301992932505297e-06,
"loss": 0.9728,
"step": 2097
},
{
"epoch": 5.365728900255754,
"grad_norm": 0.1461571001459624,
"learning_rate": 6.283416343176325e-06,
"loss": 0.9635,
"step": 2098
},
{
"epoch": 5.368286445012788,
"grad_norm": 0.1493853171572093,
"learning_rate": 6.264862069871334e-06,
"loss": 0.994,
"step": 2099
},
{
"epoch": 5.370843989769821,
"grad_norm": 0.15219054092770865,
"learning_rate": 6.2463301427769665e-06,
"loss": 0.9984,
"step": 2100
},
{
"epoch": 5.373401534526854,
"grad_norm": 0.1518597412603777,
"learning_rate": 6.227820592043499e-06,
"loss": 0.9702,
"step": 2101
},
{
"epoch": 5.375959079283888,
"grad_norm": 0.14143559764430275,
"learning_rate": 6.209333447784831e-06,
"loss": 1.0072,
"step": 2102
},
{
"epoch": 5.378516624040921,
"grad_norm": 0.15451427492744646,
"learning_rate": 6.190868740078379e-06,
"loss": 1.0002,
"step": 2103
},
{
"epoch": 5.381074168797954,
"grad_norm": 0.14794317390311024,
"learning_rate": 6.172426498965065e-06,
"loss": 0.9917,
"step": 2104
},
{
"epoch": 5.383631713554987,
"grad_norm": 0.1433195397774705,
"learning_rate": 6.154006754449275e-06,
"loss": 0.9599,
"step": 2105
},
{
"epoch": 5.3861892583120206,
"grad_norm": 0.14882148205462856,
"learning_rate": 6.135609536498772e-06,
"loss": 0.9626,
"step": 2106
},
{
"epoch": 5.388746803069053,
"grad_norm": 0.1517644634040398,
"learning_rate": 6.117234875044684e-06,
"loss": 0.9824,
"step": 2107
},
{
"epoch": 5.391304347826087,
"grad_norm": 0.1449503353324492,
"learning_rate": 6.098882799981425e-06,
"loss": 0.9727,
"step": 2108
},
{
"epoch": 5.3938618925831205,
"grad_norm": 0.15069127481084157,
"learning_rate": 6.080553341166691e-06,
"loss": 0.9713,
"step": 2109
},
{
"epoch": 5.396419437340153,
"grad_norm": 0.15189194324734845,
"learning_rate": 6.0622465284213535e-06,
"loss": 0.963,
"step": 2110
},
{
"epoch": 5.398976982097187,
"grad_norm": 0.14183788217924956,
"learning_rate": 6.043962391529452e-06,
"loss": 0.9935,
"step": 2111
},
{
"epoch": 5.40153452685422,
"grad_norm": 0.1439151738720922,
"learning_rate": 6.025700960238143e-06,
"loss": 0.9838,
"step": 2112
},
{
"epoch": 5.404092071611253,
"grad_norm": 0.14837757150273967,
"learning_rate": 6.007462264257622e-06,
"loss": 0.9759,
"step": 2113
},
{
"epoch": 5.406649616368286,
"grad_norm": 0.15749692248779976,
"learning_rate": 5.9892463332611255e-06,
"loss": 0.9644,
"step": 2114
},
{
"epoch": 5.40920716112532,
"grad_norm": 0.14725659174640404,
"learning_rate": 5.9710531968848084e-06,
"loss": 0.9728,
"step": 2115
},
{
"epoch": 5.411764705882353,
"grad_norm": 0.14252354602314363,
"learning_rate": 5.952882884727786e-06,
"loss": 0.9594,
"step": 2116
},
{
"epoch": 5.414322250639386,
"grad_norm": 0.14939079148754114,
"learning_rate": 5.934735426352014e-06,
"loss": 0.9739,
"step": 2117
},
{
"epoch": 5.41687979539642,
"grad_norm": 0.1471154416414251,
"learning_rate": 5.916610851282265e-06,
"loss": 0.9442,
"step": 2118
},
{
"epoch": 5.419437340153452,
"grad_norm": 0.13715779266015088,
"learning_rate": 5.8985091890061034e-06,
"loss": 0.9612,
"step": 2119
},
{
"epoch": 5.421994884910486,
"grad_norm": 0.14390967221814036,
"learning_rate": 5.880430468973788e-06,
"loss": 0.9959,
"step": 2120
},
{
"epoch": 5.42455242966752,
"grad_norm": 0.14439135638261424,
"learning_rate": 5.862374720598287e-06,
"loss": 0.9924,
"step": 2121
},
{
"epoch": 5.427109974424552,
"grad_norm": 0.14031449430392123,
"learning_rate": 5.844341973255146e-06,
"loss": 0.9573,
"step": 2122
},
{
"epoch": 5.429667519181586,
"grad_norm": 0.13780457135912846,
"learning_rate": 5.82633225628253e-06,
"loss": 1.0018,
"step": 2123
},
{
"epoch": 5.432225063938619,
"grad_norm": 0.15846741838386585,
"learning_rate": 5.808345598981137e-06,
"loss": 0.9989,
"step": 2124
},
{
"epoch": 5.434782608695652,
"grad_norm": 0.1374759607696619,
"learning_rate": 5.79038203061411e-06,
"loss": 0.9524,
"step": 2125
},
{
"epoch": 5.437340153452685,
"grad_norm": 0.1460085027626989,
"learning_rate": 5.772441580407067e-06,
"loss": 0.9929,
"step": 2126
},
{
"epoch": 5.439897698209719,
"grad_norm": 0.14378151304203957,
"learning_rate": 5.754524277547988e-06,
"loss": 0.973,
"step": 2127
},
{
"epoch": 5.442455242966752,
"grad_norm": 0.1452160409255816,
"learning_rate": 5.7366301511872214e-06,
"loss": 0.9667,
"step": 2128
},
{
"epoch": 5.445012787723785,
"grad_norm": 0.143782100916692,
"learning_rate": 5.71875923043737e-06,
"loss": 0.984,
"step": 2129
},
{
"epoch": 5.447570332480819,
"grad_norm": 0.13935862247480119,
"learning_rate": 5.700911544373316e-06,
"loss": 1.02,
"step": 2130
},
{
"epoch": 5.450127877237851,
"grad_norm": 0.13911533085435193,
"learning_rate": 5.683087122032133e-06,
"loss": 0.969,
"step": 2131
},
{
"epoch": 5.452685421994885,
"grad_norm": 0.17374867408057965,
"learning_rate": 5.665285992413037e-06,
"loss": 1.0242,
"step": 2132
},
{
"epoch": 5.455242966751918,
"grad_norm": 0.14450547355274798,
"learning_rate": 5.647508184477353e-06,
"loss": 0.9886,
"step": 2133
},
{
"epoch": 5.457800511508951,
"grad_norm": 0.14716706159156648,
"learning_rate": 5.629753727148457e-06,
"loss": 1.0059,
"step": 2134
},
{
"epoch": 5.460358056265985,
"grad_norm": 0.1524738956885736,
"learning_rate": 5.6120226493117565e-06,
"loss": 1.0227,
"step": 2135
},
{
"epoch": 5.462915601023018,
"grad_norm": 0.14617824334354487,
"learning_rate": 5.5943149798145875e-06,
"loss": 1.0204,
"step": 2136
},
{
"epoch": 5.465473145780051,
"grad_norm": 0.14282846222842052,
"learning_rate": 5.576630747466227e-06,
"loss": 0.9868,
"step": 2137
},
{
"epoch": 5.468030690537084,
"grad_norm": 0.14018923273198483,
"learning_rate": 5.5589699810378275e-06,
"loss": 0.9915,
"step": 2138
},
{
"epoch": 5.470588235294118,
"grad_norm": 0.132288809783374,
"learning_rate": 5.541332709262341e-06,
"loss": 0.9457,
"step": 2139
},
{
"epoch": 5.4731457800511505,
"grad_norm": 0.14632704152955048,
"learning_rate": 5.52371896083451e-06,
"loss": 1.0024,
"step": 2140
},
{
"epoch": 5.475703324808184,
"grad_norm": 0.15532362657302912,
"learning_rate": 5.506128764410792e-06,
"loss": 1.0112,
"step": 2141
},
{
"epoch": 5.478260869565218,
"grad_norm": 0.13250012620177698,
"learning_rate": 5.4885621486093485e-06,
"loss": 1.0028,
"step": 2142
},
{
"epoch": 5.4808184143222505,
"grad_norm": 0.15109512823328144,
"learning_rate": 5.471019142009964e-06,
"loss": 0.9759,
"step": 2143
},
{
"epoch": 5.483375959079284,
"grad_norm": 0.14989227568587957,
"learning_rate": 5.453499773154001e-06,
"loss": 0.9435,
"step": 2144
},
{
"epoch": 5.485933503836317,
"grad_norm": 0.13352577665291848,
"learning_rate": 5.436004070544394e-06,
"loss": 0.9718,
"step": 2145
},
{
"epoch": 5.4884910485933505,
"grad_norm": 0.13686515286535927,
"learning_rate": 5.4185320626455496e-06,
"loss": 0.9726,
"step": 2146
},
{
"epoch": 5.491048593350383,
"grad_norm": 0.1399271231923321,
"learning_rate": 5.401083777883334e-06,
"loss": 0.9688,
"step": 2147
},
{
"epoch": 5.493606138107417,
"grad_norm": 0.15631289214055935,
"learning_rate": 5.383659244645005e-06,
"loss": 0.9888,
"step": 2148
},
{
"epoch": 5.4961636828644505,
"grad_norm": 0.14226383015969765,
"learning_rate": 5.366258491279206e-06,
"loss": 1.004,
"step": 2149
},
{
"epoch": 5.498721227621483,
"grad_norm": 0.15050893212958175,
"learning_rate": 5.348881546095865e-06,
"loss": 0.9928,
"step": 2150
},
{
"epoch": 5.501278772378517,
"grad_norm": 0.1442774201577709,
"learning_rate": 5.33152843736618e-06,
"loss": 0.995,
"step": 2151
},
{
"epoch": 5.5038363171355495,
"grad_norm": 0.1509371946932492,
"learning_rate": 5.314199193322589e-06,
"loss": 0.9953,
"step": 2152
},
{
"epoch": 5.506393861892583,
"grad_norm": 0.15635074839706262,
"learning_rate": 5.296893842158677e-06,
"loss": 1.0022,
"step": 2153
},
{
"epoch": 5.508951406649617,
"grad_norm": 0.1402467542126194,
"learning_rate": 5.2796124120291735e-06,
"loss": 0.9827,
"step": 2154
},
{
"epoch": 5.5115089514066495,
"grad_norm": 0.16543141963506727,
"learning_rate": 5.262354931049876e-06,
"loss": 0.9875,
"step": 2155
},
{
"epoch": 5.514066496163683,
"grad_norm": 0.17966344453986327,
"learning_rate": 5.245121427297639e-06,
"loss": 0.968,
"step": 2156
},
{
"epoch": 5.516624040920716,
"grad_norm": 0.13012797626963213,
"learning_rate": 5.227911928810292e-06,
"loss": 1.0171,
"step": 2157
},
{
"epoch": 5.5191815856777495,
"grad_norm": 0.16519657657819564,
"learning_rate": 5.210726463586604e-06,
"loss": 1.015,
"step": 2158
},
{
"epoch": 5.521739130434782,
"grad_norm": 0.14711664607776115,
"learning_rate": 5.1935650595862655e-06,
"loss": 0.9831,
"step": 2159
},
{
"epoch": 5.524296675191816,
"grad_norm": 0.14051732225664965,
"learning_rate": 5.1764277447298035e-06,
"loss": 0.9989,
"step": 2160
},
{
"epoch": 5.526854219948849,
"grad_norm": 0.13482413491405476,
"learning_rate": 5.159314546898555e-06,
"loss": 0.9763,
"step": 2161
},
{
"epoch": 5.529411764705882,
"grad_norm": 0.16725960613326685,
"learning_rate": 5.142225493934618e-06,
"loss": 0.9877,
"step": 2162
},
{
"epoch": 5.531969309462916,
"grad_norm": 0.13985901553754518,
"learning_rate": 5.125160613640825e-06,
"loss": 1.0401,
"step": 2163
},
{
"epoch": 5.534526854219949,
"grad_norm": 0.13882576387915196,
"learning_rate": 5.108119933780664e-06,
"loss": 0.9706,
"step": 2164
},
{
"epoch": 5.537084398976982,
"grad_norm": 0.16535156717774263,
"learning_rate": 5.09110348207825e-06,
"loss": 0.9742,
"step": 2165
},
{
"epoch": 5.539641943734015,
"grad_norm": 0.14425446795081118,
"learning_rate": 5.074111286218297e-06,
"loss": 1.01,
"step": 2166
},
{
"epoch": 5.542199488491049,
"grad_norm": 0.14126230480728197,
"learning_rate": 5.0571433738460385e-06,
"loss": 1.0007,
"step": 2167
},
{
"epoch": 5.544757033248082,
"grad_norm": 0.14265903884523848,
"learning_rate": 5.040199772567209e-06,
"loss": 1.0104,
"step": 2168
},
{
"epoch": 5.547314578005115,
"grad_norm": 0.13476019802190284,
"learning_rate": 5.023280509947981e-06,
"loss": 0.9769,
"step": 2169
},
{
"epoch": 5.549872122762149,
"grad_norm": 0.136666315591245,
"learning_rate": 5.006385613514948e-06,
"loss": 1.0208,
"step": 2170
},
{
"epoch": 5.552429667519181,
"grad_norm": 0.16387309186917895,
"learning_rate": 4.989515110755048e-06,
"loss": 0.9873,
"step": 2171
},
{
"epoch": 5.554987212276215,
"grad_norm": 0.1349761601953563,
"learning_rate": 4.972669029115522e-06,
"loss": 0.9835,
"step": 2172
},
{
"epoch": 5.557544757033249,
"grad_norm": 0.14048158022879556,
"learning_rate": 4.95584739600391e-06,
"loss": 0.9955,
"step": 2173
},
{
"epoch": 5.560102301790281,
"grad_norm": 0.14063499210565186,
"learning_rate": 4.939050238787948e-06,
"loss": 0.9968,
"step": 2174
},
{
"epoch": 5.562659846547315,
"grad_norm": 0.14153124612656212,
"learning_rate": 4.922277584795558e-06,
"loss": 1.0074,
"step": 2175
},
{
"epoch": 5.565217391304348,
"grad_norm": 0.1423440072439733,
"learning_rate": 4.9055294613148e-06,
"loss": 0.9667,
"step": 2176
},
{
"epoch": 5.567774936061381,
"grad_norm": 0.13607486499086988,
"learning_rate": 4.8888058955938315e-06,
"loss": 0.9743,
"step": 2177
},
{
"epoch": 5.570332480818414,
"grad_norm": 0.14205649729134673,
"learning_rate": 4.872106914840839e-06,
"loss": 0.9719,
"step": 2178
},
{
"epoch": 5.572890025575448,
"grad_norm": 0.13501564694046067,
"learning_rate": 4.855432546224022e-06,
"loss": 0.9917,
"step": 2179
},
{
"epoch": 5.57544757033248,
"grad_norm": 0.14202521903785242,
"learning_rate": 4.8387828168715415e-06,
"loss": 0.9964,
"step": 2180
},
{
"epoch": 5.578005115089514,
"grad_norm": 0.13744047032027507,
"learning_rate": 4.822157753871461e-06,
"loss": 0.9902,
"step": 2181
},
{
"epoch": 5.580562659846548,
"grad_norm": 0.13620693089971941,
"learning_rate": 4.805557384271722e-06,
"loss": 0.9868,
"step": 2182
},
{
"epoch": 5.58312020460358,
"grad_norm": 0.13769498846659825,
"learning_rate": 4.788981735080078e-06,
"loss": 0.9926,
"step": 2183
},
{
"epoch": 5.585677749360614,
"grad_norm": 0.13959835200300635,
"learning_rate": 4.7724308332640835e-06,
"loss": 0.972,
"step": 2184
},
{
"epoch": 5.588235294117647,
"grad_norm": 0.13536176578516215,
"learning_rate": 4.755904705751029e-06,
"loss": 0.995,
"step": 2185
},
{
"epoch": 5.59079283887468,
"grad_norm": 0.1293556523123572,
"learning_rate": 4.739403379427869e-06,
"loss": 1.0212,
"step": 2186
},
{
"epoch": 5.593350383631714,
"grad_norm": 0.13671190023367213,
"learning_rate": 4.722926881141246e-06,
"loss": 0.9873,
"step": 2187
},
{
"epoch": 5.595907928388747,
"grad_norm": 0.1408504536558794,
"learning_rate": 4.70647523769739e-06,
"loss": 1.0042,
"step": 2188
},
{
"epoch": 5.59846547314578,
"grad_norm": 0.1349779549664493,
"learning_rate": 4.690048475862085e-06,
"loss": 1.0071,
"step": 2189
},
{
"epoch": 5.601023017902813,
"grad_norm": 0.14024962951632738,
"learning_rate": 4.673646622360661e-06,
"loss": 0.9869,
"step": 2190
},
{
"epoch": 5.603580562659847,
"grad_norm": 0.1386338482972485,
"learning_rate": 4.657269703877898e-06,
"loss": 1.0196,
"step": 2191
},
{
"epoch": 5.6061381074168795,
"grad_norm": 0.13055035713503843,
"learning_rate": 4.640917747058031e-06,
"loss": 0.9942,
"step": 2192
},
{
"epoch": 5.608695652173913,
"grad_norm": 0.14659323385652034,
"learning_rate": 4.624590778504652e-06,
"loss": 0.9903,
"step": 2193
},
{
"epoch": 5.611253196930946,
"grad_norm": 0.15804414834399205,
"learning_rate": 4.608288824780734e-06,
"loss": 0.9936,
"step": 2194
},
{
"epoch": 5.6138107416879794,
"grad_norm": 0.12923370150541857,
"learning_rate": 4.592011912408529e-06,
"loss": 0.9887,
"step": 2195
},
{
"epoch": 5.616368286445013,
"grad_norm": 0.15635704403118406,
"learning_rate": 4.5757600678695526e-06,
"loss": 0.9909,
"step": 2196
},
{
"epoch": 5.618925831202046,
"grad_norm": 0.14874284868028192,
"learning_rate": 4.55953331760455e-06,
"loss": 1.0113,
"step": 2197
},
{
"epoch": 5.621483375959079,
"grad_norm": 0.14569996149914322,
"learning_rate": 4.543331688013417e-06,
"loss": 0.9701,
"step": 2198
},
{
"epoch": 5.624040920716112,
"grad_norm": 0.15233771050686287,
"learning_rate": 4.527155205455211e-06,
"loss": 0.9864,
"step": 2199
},
{
"epoch": 5.626598465473146,
"grad_norm": 0.13974402183600793,
"learning_rate": 4.511003896248034e-06,
"loss": 0.9975,
"step": 2200
},
{
"epoch": 5.629156010230179,
"grad_norm": 0.15494592013262876,
"learning_rate": 4.494877786669074e-06,
"loss": 0.9875,
"step": 2201
},
{
"epoch": 5.631713554987212,
"grad_norm": 0.1292734360386681,
"learning_rate": 4.4787769029544914e-06,
"loss": 0.9592,
"step": 2202
},
{
"epoch": 5.634271099744246,
"grad_norm": 0.1437172612635685,
"learning_rate": 4.462701271299426e-06,
"loss": 1.0103,
"step": 2203
},
{
"epoch": 5.6368286445012785,
"grad_norm": 0.13479797341338134,
"learning_rate": 4.446650917857922e-06,
"loss": 0.9736,
"step": 2204
},
{
"epoch": 5.639386189258312,
"grad_norm": 0.14977214748768153,
"learning_rate": 4.430625868742897e-06,
"loss": 1.0089,
"step": 2205
},
{
"epoch": 5.641943734015345,
"grad_norm": 0.14106976641464805,
"learning_rate": 4.414626150026118e-06,
"loss": 0.9866,
"step": 2206
},
{
"epoch": 5.6445012787723785,
"grad_norm": 0.1347488567203466,
"learning_rate": 4.398651787738107e-06,
"loss": 0.9601,
"step": 2207
},
{
"epoch": 5.647058823529412,
"grad_norm": 0.136781541450226,
"learning_rate": 4.382702807868164e-06,
"loss": 0.9959,
"step": 2208
},
{
"epoch": 5.649616368286445,
"grad_norm": 0.13545452916372744,
"learning_rate": 4.366779236364278e-06,
"loss": 0.9706,
"step": 2209
},
{
"epoch": 5.6521739130434785,
"grad_norm": 0.14436945339034238,
"learning_rate": 4.350881099133111e-06,
"loss": 1.0023,
"step": 2210
},
{
"epoch": 5.654731457800511,
"grad_norm": 0.13458013688279044,
"learning_rate": 4.335008422039933e-06,
"loss": 0.982,
"step": 2211
},
{
"epoch": 5.657289002557545,
"grad_norm": 0.12682329972375592,
"learning_rate": 4.319161230908593e-06,
"loss": 0.9859,
"step": 2212
},
{
"epoch": 5.659846547314578,
"grad_norm": 0.14386106421082262,
"learning_rate": 4.303339551521493e-06,
"loss": 1.0135,
"step": 2213
},
{
"epoch": 5.662404092071611,
"grad_norm": 0.1404367705828563,
"learning_rate": 4.287543409619506e-06,
"loss": 0.9878,
"step": 2214
},
{
"epoch": 5.664961636828645,
"grad_norm": 0.13204488561082725,
"learning_rate": 4.2717728309019675e-06,
"loss": 0.961,
"step": 2215
},
{
"epoch": 5.667519181585678,
"grad_norm": 0.13288940413354677,
"learning_rate": 4.256027841026622e-06,
"loss": 0.9683,
"step": 2216
},
{
"epoch": 5.670076726342711,
"grad_norm": 0.12771868500389802,
"learning_rate": 4.240308465609593e-06,
"loss": 0.9965,
"step": 2217
},
{
"epoch": 5.672634271099744,
"grad_norm": 0.1381598400440453,
"learning_rate": 4.224614730225314e-06,
"loss": 0.9703,
"step": 2218
},
{
"epoch": 5.675191815856778,
"grad_norm": 0.1388643553284094,
"learning_rate": 4.208946660406505e-06,
"loss": 0.9747,
"step": 2219
},
{
"epoch": 5.677749360613811,
"grad_norm": 0.13450888031895514,
"learning_rate": 4.193304281644151e-06,
"loss": 0.9849,
"step": 2220
},
{
"epoch": 5.680306905370844,
"grad_norm": 0.13833260619869625,
"learning_rate": 4.177687619387416e-06,
"loss": 1.0042,
"step": 2221
},
{
"epoch": 5.6828644501278776,
"grad_norm": 0.12950835812621553,
"learning_rate": 4.162096699043628e-06,
"loss": 1.0042,
"step": 2222
},
{
"epoch": 5.68542199488491,
"grad_norm": 0.12272384795233517,
"learning_rate": 4.146531545978247e-06,
"loss": 0.9897,
"step": 2223
},
{
"epoch": 5.687979539641944,
"grad_norm": 0.1258485126257134,
"learning_rate": 4.130992185514806e-06,
"loss": 0.9702,
"step": 2224
},
{
"epoch": 5.690537084398977,
"grad_norm": 0.14053820629299157,
"learning_rate": 4.115478642934867e-06,
"loss": 1.0419,
"step": 2225
},
{
"epoch": 5.69309462915601,
"grad_norm": 0.12327977620786358,
"learning_rate": 4.09999094347799e-06,
"loss": 0.994,
"step": 2226
},
{
"epoch": 5.695652173913043,
"grad_norm": 0.13287363080756212,
"learning_rate": 4.084529112341706e-06,
"loss": 0.9514,
"step": 2227
},
{
"epoch": 5.698209718670077,
"grad_norm": 0.13289824906198588,
"learning_rate": 4.069093174681438e-06,
"loss": 0.9794,
"step": 2228
},
{
"epoch": 5.70076726342711,
"grad_norm": 0.13289061618948547,
"learning_rate": 4.05368315561049e-06,
"loss": 1.0178,
"step": 2229
},
{
"epoch": 5.703324808184143,
"grad_norm": 0.13023606203552313,
"learning_rate": 4.03829908020001e-06,
"loss": 1.021,
"step": 2230
},
{
"epoch": 5.705882352941177,
"grad_norm": 0.12723902855740332,
"learning_rate": 4.0229409734789196e-06,
"loss": 0.9797,
"step": 2231
},
{
"epoch": 5.708439897698209,
"grad_norm": 0.12358428367914359,
"learning_rate": 4.0076088604338985e-06,
"loss": 0.9902,
"step": 2232
},
{
"epoch": 5.710997442455243,
"grad_norm": 0.13339486774833817,
"learning_rate": 3.9923027660093326e-06,
"loss": 0.999,
"step": 2233
},
{
"epoch": 5.713554987212277,
"grad_norm": 0.1352390587102059,
"learning_rate": 3.977022715107288e-06,
"loss": 0.9767,
"step": 2234
},
{
"epoch": 5.716112531969309,
"grad_norm": 0.1255012502780256,
"learning_rate": 3.96176873258745e-06,
"loss": 0.9685,
"step": 2235
},
{
"epoch": 5.718670076726343,
"grad_norm": 0.12857248024836015,
"learning_rate": 3.946540843267086e-06,
"loss": 0.9939,
"step": 2236
},
{
"epoch": 5.721227621483376,
"grad_norm": 0.12991180568186123,
"learning_rate": 3.931339071921032e-06,
"loss": 0.9691,
"step": 2237
},
{
"epoch": 5.723785166240409,
"grad_norm": 0.13305537118470095,
"learning_rate": 3.916163443281615e-06,
"loss": 0.9807,
"step": 2238
},
{
"epoch": 5.726342710997442,
"grad_norm": 0.12639607723922083,
"learning_rate": 3.901013982038633e-06,
"loss": 0.9534,
"step": 2239
},
{
"epoch": 5.728900255754476,
"grad_norm": 0.12570027593106323,
"learning_rate": 3.8858907128393085e-06,
"loss": 0.9861,
"step": 2240
},
{
"epoch": 5.731457800511509,
"grad_norm": 0.1253190348092665,
"learning_rate": 3.870793660288261e-06,
"loss": 0.9421,
"step": 2241
},
{
"epoch": 5.734015345268542,
"grad_norm": 0.12048671766969506,
"learning_rate": 3.855722848947452e-06,
"loss": 0.9904,
"step": 2242
},
{
"epoch": 5.736572890025576,
"grad_norm": 0.12675993098955055,
"learning_rate": 3.840678303336138e-06,
"loss": 0.9785,
"step": 2243
},
{
"epoch": 5.739130434782608,
"grad_norm": 0.1283101058153546,
"learning_rate": 3.825660047930871e-06,
"loss": 0.9762,
"step": 2244
},
{
"epoch": 5.741687979539642,
"grad_norm": 0.13433080119354218,
"learning_rate": 3.810668107165401e-06,
"loss": 1.0015,
"step": 2245
},
{
"epoch": 5.744245524296675,
"grad_norm": 0.34203844222334506,
"learning_rate": 3.795702505430683e-06,
"loss": 0.9623,
"step": 2246
},
{
"epoch": 5.746803069053708,
"grad_norm": 0.1308006287304571,
"learning_rate": 3.7807632670748097e-06,
"loss": 0.975,
"step": 2247
},
{
"epoch": 5.749360613810742,
"grad_norm": 0.12857769640613684,
"learning_rate": 3.765850416402996e-06,
"loss": 0.9634,
"step": 2248
},
{
"epoch": 5.751918158567775,
"grad_norm": 0.15023756833501534,
"learning_rate": 3.7509639776775152e-06,
"loss": 0.9879,
"step": 2249
},
{
"epoch": 5.754475703324808,
"grad_norm": 0.1344646980165642,
"learning_rate": 3.736103975117664e-06,
"loss": 1.0193,
"step": 2250
},
{
"epoch": 5.757033248081841,
"grad_norm": 0.1397817647451772,
"learning_rate": 3.7212704328997507e-06,
"loss": 0.9879,
"step": 2251
},
{
"epoch": 5.759590792838875,
"grad_norm": 0.13028595737500917,
"learning_rate": 3.706463375157017e-06,
"loss": 1.0242,
"step": 2252
},
{
"epoch": 5.762148337595908,
"grad_norm": 0.14496651031294444,
"learning_rate": 3.6916828259796144e-06,
"loss": 0.978,
"step": 2253
},
{
"epoch": 5.764705882352941,
"grad_norm": 0.14542467444421492,
"learning_rate": 3.6769288094145728e-06,
"loss": 1.0022,
"step": 2254
},
{
"epoch": 5.767263427109975,
"grad_norm": 0.12884402900933614,
"learning_rate": 3.662201349465755e-06,
"loss": 1.0001,
"step": 2255
},
{
"epoch": 5.7698209718670075,
"grad_norm": 0.14026474935586972,
"learning_rate": 3.6475004700938322e-06,
"loss": 0.9932,
"step": 2256
},
{
"epoch": 5.772378516624041,
"grad_norm": 0.1272157235417458,
"learning_rate": 3.6328261952161925e-06,
"loss": 0.9662,
"step": 2257
},
{
"epoch": 5.774936061381074,
"grad_norm": 0.13106672425208712,
"learning_rate": 3.6181785487069808e-06,
"loss": 0.9529,
"step": 2258
},
{
"epoch": 5.7774936061381075,
"grad_norm": 0.14295505952942042,
"learning_rate": 3.6035575543969947e-06,
"loss": 0.9851,
"step": 2259
},
{
"epoch": 5.78005115089514,
"grad_norm": 0.13537071541086831,
"learning_rate": 3.5889632360736816e-06,
"loss": 1.0069,
"step": 2260
},
{
"epoch": 5.782608695652174,
"grad_norm": 0.13109212311095658,
"learning_rate": 3.5743956174810767e-06,
"loss": 0.9752,
"step": 2261
},
{
"epoch": 5.7851662404092075,
"grad_norm": 0.13401844420851716,
"learning_rate": 3.5598547223197887e-06,
"loss": 0.9779,
"step": 2262
},
{
"epoch": 5.78772378516624,
"grad_norm": 0.14143637782075605,
"learning_rate": 3.545340574246958e-06,
"loss": 0.9605,
"step": 2263
},
{
"epoch": 5.790281329923274,
"grad_norm": 0.129005647218253,
"learning_rate": 3.530853196876176e-06,
"loss": 1.003,
"step": 2264
},
{
"epoch": 5.792838874680307,
"grad_norm": 0.1292073205575687,
"learning_rate": 3.5163926137775107e-06,
"loss": 0.9742,
"step": 2265
},
{
"epoch": 5.79539641943734,
"grad_norm": 0.13027615537944612,
"learning_rate": 3.501958848477427e-06,
"loss": 1.0106,
"step": 2266
},
{
"epoch": 5.797953964194374,
"grad_norm": 0.12808445507515787,
"learning_rate": 3.4875519244587543e-06,
"loss": 0.9957,
"step": 2267
},
{
"epoch": 5.8005115089514065,
"grad_norm": 0.12543129150023283,
"learning_rate": 3.4731718651606537e-06,
"loss": 0.9934,
"step": 2268
},
{
"epoch": 5.80306905370844,
"grad_norm": 0.13452796726561211,
"learning_rate": 3.4588186939785896e-06,
"loss": 0.9907,
"step": 2269
},
{
"epoch": 5.805626598465473,
"grad_norm": 0.12900433681542192,
"learning_rate": 3.4444924342642795e-06,
"loss": 0.9769,
"step": 2270
},
{
"epoch": 5.8081841432225065,
"grad_norm": 0.13129752361110394,
"learning_rate": 3.430193109325637e-06,
"loss": 0.9827,
"step": 2271
},
{
"epoch": 5.810741687979539,
"grad_norm": 0.12768070736450798,
"learning_rate": 3.4159207424267816e-06,
"loss": 0.9535,
"step": 2272
},
{
"epoch": 5.813299232736573,
"grad_norm": 0.12441577467966322,
"learning_rate": 3.4016753567879545e-06,
"loss": 0.9741,
"step": 2273
},
{
"epoch": 5.8158567774936065,
"grad_norm": 0.1312235101477935,
"learning_rate": 3.3874569755855192e-06,
"loss": 0.9819,
"step": 2274
},
{
"epoch": 5.818414322250639,
"grad_norm": 0.12524160412102536,
"learning_rate": 3.3732656219518754e-06,
"loss": 0.9464,
"step": 2275
},
{
"epoch": 5.820971867007673,
"grad_norm": 0.12402516450989616,
"learning_rate": 3.3591013189754775e-06,
"loss": 0.9961,
"step": 2276
},
{
"epoch": 5.823529411764706,
"grad_norm": 0.13626673377177617,
"learning_rate": 3.3449640897007705e-06,
"loss": 0.986,
"step": 2277
},
{
"epoch": 5.826086956521739,
"grad_norm": 0.13423450066045847,
"learning_rate": 3.3308539571281217e-06,
"loss": 1.0097,
"step": 2278
},
{
"epoch": 5.828644501278772,
"grad_norm": 0.12456442602372057,
"learning_rate": 3.316770944213852e-06,
"loss": 0.9805,
"step": 2279
},
{
"epoch": 5.831202046035806,
"grad_norm": 0.14055311714625834,
"learning_rate": 3.3027150738701308e-06,
"loss": 0.9859,
"step": 2280
},
{
"epoch": 5.833759590792839,
"grad_norm": 0.14143948769230638,
"learning_rate": 3.288686368964995e-06,
"loss": 0.9664,
"step": 2281
},
{
"epoch": 5.836317135549872,
"grad_norm": 0.12841707147204753,
"learning_rate": 3.274684852322252e-06,
"loss": 0.9675,
"step": 2282
},
{
"epoch": 5.838874680306906,
"grad_norm": 0.13120598393921942,
"learning_rate": 3.260710546721504e-06,
"loss": 0.974,
"step": 2283
},
{
"epoch": 5.841432225063938,
"grad_norm": 0.12829921141935707,
"learning_rate": 3.246763474898076e-06,
"loss": 1.0146,
"step": 2284
},
{
"epoch": 5.843989769820972,
"grad_norm": 0.1444970119245527,
"learning_rate": 3.2328436595429768e-06,
"loss": 0.9445,
"step": 2285
},
{
"epoch": 5.846547314578006,
"grad_norm": 0.1393929588026091,
"learning_rate": 3.218951123302876e-06,
"loss": 0.9543,
"step": 2286
},
{
"epoch": 5.849104859335038,
"grad_norm": 0.12118720231757932,
"learning_rate": 3.2050858887800573e-06,
"loss": 1.0058,
"step": 2287
},
{
"epoch": 5.851662404092072,
"grad_norm": 0.13376184915423053,
"learning_rate": 3.191247978532399e-06,
"loss": 0.9977,
"step": 2288
},
{
"epoch": 5.854219948849105,
"grad_norm": 0.13191691378173948,
"learning_rate": 3.177437415073312e-06,
"loss": 0.9812,
"step": 2289
},
{
"epoch": 5.856777493606138,
"grad_norm": 0.12620434609710685,
"learning_rate": 3.163654220871715e-06,
"loss": 1.0009,
"step": 2290
},
{
"epoch": 5.859335038363171,
"grad_norm": 0.11882876081210357,
"learning_rate": 3.149898418352011e-06,
"loss": 0.9926,
"step": 2291
},
{
"epoch": 5.861892583120205,
"grad_norm": 0.12672788978838265,
"learning_rate": 3.1361700298940278e-06,
"loss": 0.9802,
"step": 2292
},
{
"epoch": 5.864450127877237,
"grad_norm": 0.13318892320419304,
"learning_rate": 3.1224690778329948e-06,
"loss": 0.987,
"step": 2293
},
{
"epoch": 5.867007672634271,
"grad_norm": 0.12564445725659595,
"learning_rate": 3.1087955844594987e-06,
"loss": 0.9856,
"step": 2294
},
{
"epoch": 5.869565217391305,
"grad_norm": 0.1297678632025808,
"learning_rate": 3.095149572019467e-06,
"loss": 0.9835,
"step": 2295
},
{
"epoch": 5.872122762148337,
"grad_norm": 0.12931332643863352,
"learning_rate": 3.081531062714105e-06,
"loss": 0.9843,
"step": 2296
},
{
"epoch": 5.874680306905371,
"grad_norm": 0.13109042560459144,
"learning_rate": 3.0679400786998713e-06,
"loss": 0.9451,
"step": 2297
},
{
"epoch": 5.877237851662404,
"grad_norm": 0.11761125377251656,
"learning_rate": 3.054376642088457e-06,
"loss": 0.9902,
"step": 2298
},
{
"epoch": 5.879795396419437,
"grad_norm": 0.1282754143199696,
"learning_rate": 3.0408407749467207e-06,
"loss": 1.0109,
"step": 2299
},
{
"epoch": 5.882352941176471,
"grad_norm": 0.12838671290992124,
"learning_rate": 3.027332499296669e-06,
"loss": 0.9814,
"step": 2300
},
{
"epoch": 5.884910485933504,
"grad_norm": 0.1356917241124969,
"learning_rate": 3.0138518371154224e-06,
"loss": 0.9893,
"step": 2301
},
{
"epoch": 5.887468030690537,
"grad_norm": 0.12242975604414608,
"learning_rate": 3.000398810335179e-06,
"loss": 0.9671,
"step": 2302
},
{
"epoch": 5.89002557544757,
"grad_norm": 0.12357997422937082,
"learning_rate": 2.9869734408431728e-06,
"loss": 0.9841,
"step": 2303
},
{
"epoch": 5.892583120204604,
"grad_norm": 0.13471794766222206,
"learning_rate": 2.973575750481632e-06,
"loss": 0.9683,
"step": 2304
},
{
"epoch": 5.8951406649616365,
"grad_norm": 0.12836726171109866,
"learning_rate": 2.960205761047772e-06,
"loss": 0.9778,
"step": 2305
},
{
"epoch": 5.89769820971867,
"grad_norm": 0.12531813810810682,
"learning_rate": 2.9468634942937235e-06,
"loss": 0.9934,
"step": 2306
},
{
"epoch": 5.900255754475703,
"grad_norm": 0.12343737829546242,
"learning_rate": 2.933548971926523e-06,
"loss": 0.9419,
"step": 2307
},
{
"epoch": 5.9028132992327365,
"grad_norm": 0.1243200598053773,
"learning_rate": 2.920262215608061e-06,
"loss": 0.995,
"step": 2308
},
{
"epoch": 5.90537084398977,
"grad_norm": 0.11989167696462562,
"learning_rate": 2.907003246955065e-06,
"loss": 0.9498,
"step": 2309
},
{
"epoch": 5.907928388746803,
"grad_norm": 0.1254556637863592,
"learning_rate": 2.8937720875390486e-06,
"loss": 0.9931,
"step": 2310
},
{
"epoch": 5.910485933503836,
"grad_norm": 0.12436911483814787,
"learning_rate": 2.880568758886271e-06,
"loss": 0.9744,
"step": 2311
},
{
"epoch": 5.913043478260869,
"grad_norm": 0.12957247887335613,
"learning_rate": 2.867393282477735e-06,
"loss": 1.0046,
"step": 2312
},
{
"epoch": 5.915601023017903,
"grad_norm": 0.12960479826904056,
"learning_rate": 2.8542456797491104e-06,
"loss": 0.9724,
"step": 2313
},
{
"epoch": 5.918158567774936,
"grad_norm": 0.12629913099891807,
"learning_rate": 2.8411259720907282e-06,
"loss": 1.0242,
"step": 2314
},
{
"epoch": 5.920716112531969,
"grad_norm": 0.11676453528826661,
"learning_rate": 2.828034180847523e-06,
"loss": 0.9741,
"step": 2315
},
{
"epoch": 5.923273657289003,
"grad_norm": 0.12891871238447952,
"learning_rate": 2.814970327319033e-06,
"loss": 0.9578,
"step": 2316
},
{
"epoch": 5.9258312020460355,
"grad_norm": 0.13361905883199424,
"learning_rate": 2.801934432759326e-06,
"loss": 0.9684,
"step": 2317
},
{
"epoch": 5.928388746803069,
"grad_norm": 0.12899958353085694,
"learning_rate": 2.788926518376982e-06,
"loss": 0.9761,
"step": 2318
},
{
"epoch": 5.930946291560103,
"grad_norm": 0.12585203017938093,
"learning_rate": 2.7759466053350735e-06,
"loss": 1.0014,
"step": 2319
},
{
"epoch": 5.9335038363171355,
"grad_norm": 0.12801757626200713,
"learning_rate": 2.7629947147511037e-06,
"loss": 1.0006,
"step": 2320
},
{
"epoch": 5.936061381074169,
"grad_norm": 0.12367243688730535,
"learning_rate": 2.750070867696983e-06,
"loss": 0.9547,
"step": 2321
},
{
"epoch": 5.938618925831202,
"grad_norm": 0.12375030596115455,
"learning_rate": 2.737175085199013e-06,
"loss": 0.9849,
"step": 2322
},
{
"epoch": 5.9411764705882355,
"grad_norm": 0.13556276053550034,
"learning_rate": 2.7243073882378215e-06,
"loss": 0.9806,
"step": 2323
},
{
"epoch": 5.943734015345268,
"grad_norm": 0.14276462880992438,
"learning_rate": 2.711467797748346e-06,
"loss": 1.0136,
"step": 2324
},
{
"epoch": 5.946291560102302,
"grad_norm": 0.12862761643304232,
"learning_rate": 2.6986563346197913e-06,
"loss": 0.9808,
"step": 2325
},
{
"epoch": 5.948849104859335,
"grad_norm": 0.12524428958406433,
"learning_rate": 2.6858730196956174e-06,
"loss": 0.9645,
"step": 2326
},
{
"epoch": 5.951406649616368,
"grad_norm": 0.12472361640876028,
"learning_rate": 2.6731178737734743e-06,
"loss": 0.9459,
"step": 2327
},
{
"epoch": 5.953964194373402,
"grad_norm": 0.12657596610420463,
"learning_rate": 2.660390917605182e-06,
"loss": 0.9906,
"step": 2328
},
{
"epoch": 5.956521739130435,
"grad_norm": 0.13664741039619877,
"learning_rate": 2.647692171896712e-06,
"loss": 0.9565,
"step": 2329
},
{
"epoch": 5.959079283887468,
"grad_norm": 0.12493585422425488,
"learning_rate": 2.6350216573081233e-06,
"loss": 1.0002,
"step": 2330
},
{
"epoch": 5.961636828644501,
"grad_norm": 0.12206805695853085,
"learning_rate": 2.6223793944535536e-06,
"loss": 0.9718,
"step": 2331
},
{
"epoch": 5.964194373401535,
"grad_norm": 0.12175202280619228,
"learning_rate": 2.6097654039011655e-06,
"loss": 0.9551,
"step": 2332
},
{
"epoch": 5.966751918158568,
"grad_norm": 0.13596420957291225,
"learning_rate": 2.5971797061731473e-06,
"loss": 0.9545,
"step": 2333
},
{
"epoch": 5.969309462915601,
"grad_norm": 0.12619176415651825,
"learning_rate": 2.5846223217456335e-06,
"loss": 0.9872,
"step": 2334
},
{
"epoch": 5.971867007672635,
"grad_norm": 0.12855497064391522,
"learning_rate": 2.572093271048701e-06,
"loss": 1.0231,
"step": 2335
},
{
"epoch": 5.974424552429667,
"grad_norm": 0.12217437898623634,
"learning_rate": 2.5595925744663386e-06,
"loss": 0.9759,
"step": 2336
},
{
"epoch": 5.976982097186701,
"grad_norm": 0.1311701519651553,
"learning_rate": 2.547120252336397e-06,
"loss": 1.0138,
"step": 2337
},
{
"epoch": 5.979539641943734,
"grad_norm": 0.13592951866154088,
"learning_rate": 2.5346763249505625e-06,
"loss": 0.9958,
"step": 2338
},
{
"epoch": 5.982097186700767,
"grad_norm": 0.1328651876122347,
"learning_rate": 2.5222608125543226e-06,
"loss": 0.983,
"step": 2339
},
{
"epoch": 5.9846547314578,
"grad_norm": 0.1217034366897049,
"learning_rate": 2.509873735346948e-06,
"loss": 0.9715,
"step": 2340
},
{
"epoch": 5.987212276214834,
"grad_norm": 0.1258025431677806,
"learning_rate": 2.497515113481437e-06,
"loss": 0.9658,
"step": 2341
},
{
"epoch": 5.989769820971867,
"grad_norm": 0.1278832653433515,
"learning_rate": 2.4851849670644843e-06,
"loss": 0.9911,
"step": 2342
},
{
"epoch": 5.9923273657289,
"grad_norm": 0.12132220662809967,
"learning_rate": 2.4728833161564827e-06,
"loss": 0.9851,
"step": 2343
},
{
"epoch": 5.994884910485934,
"grad_norm": 0.12065967647733448,
"learning_rate": 2.460610180771439e-06,
"loss": 0.9818,
"step": 2344
},
{
"epoch": 5.997442455242966,
"grad_norm": 0.12491394855405957,
"learning_rate": 2.4483655808769855e-06,
"loss": 0.996,
"step": 2345
},
{
"epoch": 6.0,
"grad_norm": 0.11699967146736108,
"learning_rate": 2.4361495363943078e-06,
"loss": 0.9702,
"step": 2346
},
{
"epoch": 6.002557544757034,
"grad_norm": 0.12619212300829918,
"learning_rate": 2.423962067198158e-06,
"loss": 1.0042,
"step": 2347
},
{
"epoch": 6.005115089514066,
"grad_norm": 0.13249267564481532,
"learning_rate": 2.4118031931167817e-06,
"loss": 0.9762,
"step": 2348
},
{
"epoch": 6.0076726342711,
"grad_norm": 0.1170120047647475,
"learning_rate": 2.399672933931905e-06,
"loss": 0.944,
"step": 2349
},
{
"epoch": 6.010230179028133,
"grad_norm": 0.1159312422260675,
"learning_rate": 2.3875713093787065e-06,
"loss": 0.9509,
"step": 2350
},
{
"epoch": 6.012787723785166,
"grad_norm": 0.1232578157739251,
"learning_rate": 2.3754983391457654e-06,
"loss": 0.9815,
"step": 2351
},
{
"epoch": 6.015345268542199,
"grad_norm": 0.12740106243861546,
"learning_rate": 2.363454042875062e-06,
"loss": 0.9542,
"step": 2352
},
{
"epoch": 6.017902813299233,
"grad_norm": 0.12082782835893716,
"learning_rate": 2.3514384401618974e-06,
"loss": 0.9777,
"step": 2353
},
{
"epoch": 6.020460358056266,
"grad_norm": 0.12202681711822683,
"learning_rate": 2.3394515505549163e-06,
"loss": 0.9704,
"step": 2354
},
{
"epoch": 6.023017902813299,
"grad_norm": 0.12368810968350906,
"learning_rate": 2.327493393556037e-06,
"loss": 0.9698,
"step": 2355
},
{
"epoch": 6.025575447570333,
"grad_norm": 0.14673564138445166,
"learning_rate": 2.3155639886204264e-06,
"loss": 0.9967,
"step": 2356
},
{
"epoch": 6.028132992327365,
"grad_norm": 0.1344201269812993,
"learning_rate": 2.3036633551564913e-06,
"loss": 1.0257,
"step": 2357
},
{
"epoch": 6.030690537084399,
"grad_norm": 0.13001538702533763,
"learning_rate": 2.2917915125258096e-06,
"loss": 0.9881,
"step": 2358
},
{
"epoch": 6.033248081841432,
"grad_norm": 0.12379169886496132,
"learning_rate": 2.279948480043137e-06,
"loss": 0.9793,
"step": 2359
},
{
"epoch": 6.035805626598465,
"grad_norm": 0.12205528335752554,
"learning_rate": 2.2681342769763327e-06,
"loss": 0.967,
"step": 2360
},
{
"epoch": 6.038363171355499,
"grad_norm": 0.12706743090503816,
"learning_rate": 2.256348922546374e-06,
"loss": 0.9934,
"step": 2361
},
{
"epoch": 6.040920716112532,
"grad_norm": 0.118260083002314,
"learning_rate": 2.244592435927302e-06,
"loss": 0.9552,
"step": 2362
},
{
"epoch": 6.043478260869565,
"grad_norm": 0.13447047045377347,
"learning_rate": 2.232864836246176e-06,
"loss": 0.9782,
"step": 2363
},
{
"epoch": 6.046035805626598,
"grad_norm": 0.12461613850911454,
"learning_rate": 2.221166142583071e-06,
"loss": 1.0016,
"step": 2364
},
{
"epoch": 6.048593350383632,
"grad_norm": 0.1282379288359655,
"learning_rate": 2.209496373971027e-06,
"loss": 0.9566,
"step": 2365
},
{
"epoch": 6.051150895140665,
"grad_norm": 0.13354106297064514,
"learning_rate": 2.197855549396037e-06,
"loss": 0.9633,
"step": 2366
},
{
"epoch": 6.053708439897698,
"grad_norm": 0.12152300947091244,
"learning_rate": 2.1862436877969806e-06,
"loss": 0.9644,
"step": 2367
},
{
"epoch": 6.056265984654732,
"grad_norm": 0.11727638790681143,
"learning_rate": 2.1746608080656385e-06,
"loss": 0.9502,
"step": 2368
},
{
"epoch": 6.0588235294117645,
"grad_norm": 0.12960303600762405,
"learning_rate": 2.1631069290466365e-06,
"loss": 0.9745,
"step": 2369
},
{
"epoch": 6.061381074168798,
"grad_norm": 0.11817845889853361,
"learning_rate": 2.151582069537408e-06,
"loss": 0.977,
"step": 2370
},
{
"epoch": 6.063938618925831,
"grad_norm": 0.11959172924639565,
"learning_rate": 2.1400862482881777e-06,
"loss": 0.9469,
"step": 2371
},
{
"epoch": 6.0664961636828645,
"grad_norm": 0.11951443918940516,
"learning_rate": 2.128619484001926e-06,
"loss": 0.9637,
"step": 2372
},
{
"epoch": 6.069053708439898,
"grad_norm": 0.12748986659231676,
"learning_rate": 2.1171817953343688e-06,
"loss": 0.9894,
"step": 2373
},
{
"epoch": 6.071611253196931,
"grad_norm": 0.13253697584655064,
"learning_rate": 2.105773200893908e-06,
"loss": 0.9634,
"step": 2374
},
{
"epoch": 6.0741687979539645,
"grad_norm": 0.12686661972927993,
"learning_rate": 2.094393719241605e-06,
"loss": 0.9803,
"step": 2375
},
{
"epoch": 6.076726342710997,
"grad_norm": 0.12222508586602568,
"learning_rate": 2.0830433688911777e-06,
"loss": 0.9663,
"step": 2376
},
{
"epoch": 6.079283887468031,
"grad_norm": 0.13150255810264722,
"learning_rate": 2.0717221683089315e-06,
"loss": 0.9845,
"step": 2377
},
{
"epoch": 6.081841432225064,
"grad_norm": 0.12399089351879933,
"learning_rate": 2.060430135913749e-06,
"loss": 0.9534,
"step": 2378
},
{
"epoch": 6.084398976982097,
"grad_norm": 0.11792356752130304,
"learning_rate": 2.049167290077061e-06,
"loss": 0.9914,
"step": 2379
},
{
"epoch": 6.086956521739131,
"grad_norm": 0.11735866925939906,
"learning_rate": 2.0379336491228185e-06,
"loss": 0.9655,
"step": 2380
},
{
"epoch": 6.089514066496164,
"grad_norm": 0.12606254581216708,
"learning_rate": 2.026729231327449e-06,
"loss": 0.9974,
"step": 2381
},
{
"epoch": 6.092071611253197,
"grad_norm": 0.12691253176196368,
"learning_rate": 2.0155540549198373e-06,
"loss": 0.9815,
"step": 2382
},
{
"epoch": 6.09462915601023,
"grad_norm": 0.12951105979036137,
"learning_rate": 2.004408138081304e-06,
"loss": 0.9543,
"step": 2383
},
{
"epoch": 6.0971867007672635,
"grad_norm": 0.1219813397736481,
"learning_rate": 1.9932914989455555e-06,
"loss": 1.0036,
"step": 2384
},
{
"epoch": 6.099744245524296,
"grad_norm": 0.11633398982397128,
"learning_rate": 1.982204155598666e-06,
"loss": 0.9692,
"step": 2385
},
{
"epoch": 6.10230179028133,
"grad_norm": 0.12188424714722508,
"learning_rate": 1.971146126079049e-06,
"loss": 0.9575,
"step": 2386
},
{
"epoch": 6.1048593350383635,
"grad_norm": 0.11467046747339626,
"learning_rate": 1.9601174283774347e-06,
"loss": 0.9924,
"step": 2387
},
{
"epoch": 6.107416879795396,
"grad_norm": 0.12732362902678857,
"learning_rate": 1.9491180804368183e-06,
"loss": 0.9776,
"step": 2388
},
{
"epoch": 6.10997442455243,
"grad_norm": 0.12107234049211534,
"learning_rate": 1.9381481001524483e-06,
"loss": 0.9424,
"step": 2389
},
{
"epoch": 6.112531969309463,
"grad_norm": 0.12865037929297285,
"learning_rate": 1.9272075053718063e-06,
"loss": 0.9689,
"step": 2390
},
{
"epoch": 6.115089514066496,
"grad_norm": 0.12348610373779734,
"learning_rate": 1.916296313894552e-06,
"loss": 0.9988,
"step": 2391
},
{
"epoch": 6.117647058823529,
"grad_norm": 0.11752105644260442,
"learning_rate": 1.905414543472508e-06,
"loss": 0.9538,
"step": 2392
},
{
"epoch": 6.120204603580563,
"grad_norm": 0.12750900224503245,
"learning_rate": 1.8945622118096363e-06,
"loss": 0.9969,
"step": 2393
},
{
"epoch": 6.122762148337596,
"grad_norm": 0.11880310099327374,
"learning_rate": 1.8837393365620028e-06,
"loss": 0.9917,
"step": 2394
},
{
"epoch": 6.125319693094629,
"grad_norm": 0.12692266554709447,
"learning_rate": 1.872945935337751e-06,
"loss": 0.998,
"step": 2395
},
{
"epoch": 6.127877237851663,
"grad_norm": 0.1219703573812529,
"learning_rate": 1.8621820256970635e-06,
"loss": 0.9683,
"step": 2396
},
{
"epoch": 6.130434782608695,
"grad_norm": 0.12241511224605701,
"learning_rate": 1.8514476251521562e-06,
"loss": 0.9622,
"step": 2397
},
{
"epoch": 6.132992327365729,
"grad_norm": 0.11737344963420564,
"learning_rate": 1.8407427511672216e-06,
"loss": 0.9791,
"step": 2398
},
{
"epoch": 6.135549872122763,
"grad_norm": 0.12469185897376632,
"learning_rate": 1.8300674211584258e-06,
"loss": 0.9645,
"step": 2399
},
{
"epoch": 6.138107416879795,
"grad_norm": 0.11839700075897287,
"learning_rate": 1.8194216524938535e-06,
"loss": 0.9813,
"step": 2400
},
{
"epoch": 6.140664961636829,
"grad_norm": 0.12363514309533545,
"learning_rate": 1.8088054624935147e-06,
"loss": 0.9877,
"step": 2401
},
{
"epoch": 6.143222506393862,
"grad_norm": 0.11887785866345081,
"learning_rate": 1.7982188684292846e-06,
"loss": 0.9875,
"step": 2402
},
{
"epoch": 6.145780051150895,
"grad_norm": 0.12134676100803209,
"learning_rate": 1.787661887524883e-06,
"loss": 1.0002,
"step": 2403
},
{
"epoch": 6.148337595907928,
"grad_norm": 0.11993784601127822,
"learning_rate": 1.777134536955869e-06,
"loss": 0.9837,
"step": 2404
},
{
"epoch": 6.150895140664962,
"grad_norm": 0.12266680462157951,
"learning_rate": 1.7666368338495754e-06,
"loss": 0.9787,
"step": 2405
},
{
"epoch": 6.153452685421995,
"grad_norm": 0.12084518985066253,
"learning_rate": 1.7561687952851136e-06,
"loss": 1.0077,
"step": 2406
},
{
"epoch": 6.156010230179028,
"grad_norm": 0.11998108893842542,
"learning_rate": 1.7457304382933205e-06,
"loss": 0.9905,
"step": 2407
},
{
"epoch": 6.158567774936062,
"grad_norm": 0.12415129135329135,
"learning_rate": 1.7353217798567578e-06,
"loss": 1.0041,
"step": 2408
},
{
"epoch": 6.161125319693094,
"grad_norm": 0.12505465327109216,
"learning_rate": 1.724942836909662e-06,
"loss": 0.9625,
"step": 2409
},
{
"epoch": 6.163682864450128,
"grad_norm": 0.12027220439255185,
"learning_rate": 1.714593626337917e-06,
"loss": 0.9847,
"step": 2410
},
{
"epoch": 6.166240409207161,
"grad_norm": 0.12077037732019957,
"learning_rate": 1.7042741649790473e-06,
"loss": 0.9992,
"step": 2411
},
{
"epoch": 6.168797953964194,
"grad_norm": 0.12652876774464103,
"learning_rate": 1.6939844696221742e-06,
"loss": 0.9648,
"step": 2412
},
{
"epoch": 6.171355498721228,
"grad_norm": 0.11521778724688224,
"learning_rate": 1.683724557007982e-06,
"loss": 0.9794,
"step": 2413
},
{
"epoch": 6.173913043478261,
"grad_norm": 0.12171601447871334,
"learning_rate": 1.6734944438287049e-06,
"loss": 0.9516,
"step": 2414
},
{
"epoch": 6.176470588235294,
"grad_norm": 0.12406714020455667,
"learning_rate": 1.663294146728105e-06,
"loss": 0.9738,
"step": 2415
},
{
"epoch": 6.179028132992327,
"grad_norm": 0.12372494304949917,
"learning_rate": 1.6531236823014296e-06,
"loss": 0.9989,
"step": 2416
},
{
"epoch": 6.181585677749361,
"grad_norm": 0.1205705671302383,
"learning_rate": 1.6429830670953783e-06,
"loss": 0.9747,
"step": 2417
},
{
"epoch": 6.1841432225063935,
"grad_norm": 0.11639736486664135,
"learning_rate": 1.6328723176081052e-06,
"loss": 0.995,
"step": 2418
},
{
"epoch": 6.186700767263427,
"grad_norm": 0.12130422222837921,
"learning_rate": 1.6227914502891674e-06,
"loss": 0.9814,
"step": 2419
},
{
"epoch": 6.189258312020461,
"grad_norm": 0.12703296109750734,
"learning_rate": 1.612740481539503e-06,
"loss": 0.9795,
"step": 2420
},
{
"epoch": 6.1918158567774935,
"grad_norm": 0.12710210859024892,
"learning_rate": 1.6027194277114077e-06,
"loss": 0.9945,
"step": 2421
},
{
"epoch": 6.194373401534527,
"grad_norm": 0.12487908742300083,
"learning_rate": 1.5927283051085128e-06,
"loss": 0.9932,
"step": 2422
},
{
"epoch": 6.19693094629156,
"grad_norm": 0.1261907629945684,
"learning_rate": 1.5827671299857583e-06,
"loss": 1.0078,
"step": 2423
},
{
"epoch": 6.1994884910485935,
"grad_norm": 0.12584257580659278,
"learning_rate": 1.5728359185493359e-06,
"loss": 0.9408,
"step": 2424
},
{
"epoch": 6.202046035805626,
"grad_norm": 0.11705213717967214,
"learning_rate": 1.5629346869567209e-06,
"loss": 0.9623,
"step": 2425
},
{
"epoch": 6.20460358056266,
"grad_norm": 0.13901515358530595,
"learning_rate": 1.5530634513165899e-06,
"loss": 0.9734,
"step": 2426
},
{
"epoch": 6.207161125319693,
"grad_norm": 0.1261153845143054,
"learning_rate": 1.543222227688823e-06,
"loss": 1.0094,
"step": 2427
},
{
"epoch": 6.209718670076726,
"grad_norm": 0.12939052194004358,
"learning_rate": 1.5334110320844865e-06,
"loss": 0.9906,
"step": 2428
},
{
"epoch": 6.21227621483376,
"grad_norm": 0.12414610838645558,
"learning_rate": 1.5236298804657733e-06,
"loss": 0.9655,
"step": 2429
},
{
"epoch": 6.2148337595907925,
"grad_norm": 0.12844024846886481,
"learning_rate": 1.5138787887460128e-06,
"loss": 1.0256,
"step": 2430
},
{
"epoch": 6.217391304347826,
"grad_norm": 0.12215003081954443,
"learning_rate": 1.504157772789614e-06,
"loss": 0.952,
"step": 2431
},
{
"epoch": 6.21994884910486,
"grad_norm": 0.12461545313690077,
"learning_rate": 1.4944668484120661e-06,
"loss": 0.9951,
"step": 2432
},
{
"epoch": 6.2225063938618925,
"grad_norm": 0.13020995955286793,
"learning_rate": 1.484806031379895e-06,
"loss": 0.9739,
"step": 2433
},
{
"epoch": 6.225063938618926,
"grad_norm": 0.12128071948015956,
"learning_rate": 1.4751753374106503e-06,
"loss": 1.0141,
"step": 2434
},
{
"epoch": 6.227621483375959,
"grad_norm": 0.12338479016309932,
"learning_rate": 1.465574782172865e-06,
"loss": 0.9895,
"step": 2435
},
{
"epoch": 6.2301790281329925,
"grad_norm": 0.1169551888321392,
"learning_rate": 1.4560043812860403e-06,
"loss": 1.0006,
"step": 2436
},
{
"epoch": 6.232736572890025,
"grad_norm": 0.11814141956215982,
"learning_rate": 1.44646415032063e-06,
"loss": 0.959,
"step": 2437
},
{
"epoch": 6.235294117647059,
"grad_norm": 0.12571705390836388,
"learning_rate": 1.4369541047979808e-06,
"loss": 0.9847,
"step": 2438
},
{
"epoch": 6.2378516624040925,
"grad_norm": 0.13311783532771998,
"learning_rate": 1.4274742601903492e-06,
"loss": 0.9998,
"step": 2439
},
{
"epoch": 6.240409207161125,
"grad_norm": 0.1278895622366195,
"learning_rate": 1.418024631920847e-06,
"loss": 1.0053,
"step": 2440
},
{
"epoch": 6.242966751918159,
"grad_norm": 0.11503156590292939,
"learning_rate": 1.4086052353634315e-06,
"loss": 0.9544,
"step": 2441
},
{
"epoch": 6.245524296675192,
"grad_norm": 0.11799914622720747,
"learning_rate": 1.3992160858428695e-06,
"loss": 0.9701,
"step": 2442
},
{
"epoch": 6.248081841432225,
"grad_norm": 0.12044700536419219,
"learning_rate": 1.3898571986347186e-06,
"loss": 0.999,
"step": 2443
},
{
"epoch": 6.250639386189258,
"grad_norm": 0.11650374482327584,
"learning_rate": 1.3805285889653087e-06,
"loss": 0.9353,
"step": 2444
},
{
"epoch": 6.253196930946292,
"grad_norm": 0.1261937511610037,
"learning_rate": 1.371230272011701e-06,
"loss": 0.9831,
"step": 2445
},
{
"epoch": 6.255754475703325,
"grad_norm": 0.12374381201748316,
"learning_rate": 1.3619622629016771e-06,
"loss": 1.0098,
"step": 2446
},
{
"epoch": 6.258312020460358,
"grad_norm": 0.12059097608380545,
"learning_rate": 1.3527245767137042e-06,
"loss": 0.9995,
"step": 2447
},
{
"epoch": 6.260869565217392,
"grad_norm": 0.11586818574105817,
"learning_rate": 1.3435172284769248e-06,
"loss": 0.9991,
"step": 2448
},
{
"epoch": 6.263427109974424,
"grad_norm": 0.12521587720776958,
"learning_rate": 1.3343402331711187e-06,
"loss": 0.9893,
"step": 2449
},
{
"epoch": 6.265984654731458,
"grad_norm": 0.12030015031442251,
"learning_rate": 1.3251936057266778e-06,
"loss": 0.9759,
"step": 2450
},
{
"epoch": 6.268542199488491,
"grad_norm": 0.12170492314680607,
"learning_rate": 1.3160773610245992e-06,
"loss": 0.9842,
"step": 2451
},
{
"epoch": 6.271099744245524,
"grad_norm": 0.12020850738486998,
"learning_rate": 1.3069915138964429e-06,
"loss": 1.0262,
"step": 2452
},
{
"epoch": 6.273657289002558,
"grad_norm": 0.11426480213538506,
"learning_rate": 1.2979360791243112e-06,
"loss": 0.977,
"step": 2453
},
{
"epoch": 6.276214833759591,
"grad_norm": 0.11640544194164149,
"learning_rate": 1.2889110714408303e-06,
"loss": 0.9522,
"step": 2454
},
{
"epoch": 6.278772378516624,
"grad_norm": 0.11873847190870833,
"learning_rate": 1.279916505529124e-06,
"loss": 0.9895,
"step": 2455
},
{
"epoch": 6.281329923273657,
"grad_norm": 0.11484871676005291,
"learning_rate": 1.2709523960227931e-06,
"loss": 0.9998,
"step": 2456
},
{
"epoch": 6.283887468030691,
"grad_norm": 0.11591713027188896,
"learning_rate": 1.262018757505874e-06,
"loss": 0.9934,
"step": 2457
},
{
"epoch": 6.286445012787723,
"grad_norm": 0.1165733850851201,
"learning_rate": 1.2531156045128467e-06,
"loss": 0.9677,
"step": 2458
},
{
"epoch": 6.289002557544757,
"grad_norm": 0.11763513658143136,
"learning_rate": 1.2442429515285803e-06,
"loss": 0.9829,
"step": 2459
},
{
"epoch": 6.291560102301791,
"grad_norm": 0.12385780247580941,
"learning_rate": 1.2354008129883232e-06,
"loss": 0.997,
"step": 2460
},
{
"epoch": 6.294117647058823,
"grad_norm": 0.11817000016704547,
"learning_rate": 1.2265892032776906e-06,
"loss": 0.9672,
"step": 2461
},
{
"epoch": 6.296675191815857,
"grad_norm": 0.11697634095111288,
"learning_rate": 1.2178081367326167e-06,
"loss": 0.9467,
"step": 2462
},
{
"epoch": 6.29923273657289,
"grad_norm": 0.11676729937050137,
"learning_rate": 1.209057627639345e-06,
"loss": 1.0043,
"step": 2463
},
{
"epoch": 6.301790281329923,
"grad_norm": 0.11443174980680064,
"learning_rate": 1.2003376902344055e-06,
"loss": 0.9663,
"step": 2464
},
{
"epoch": 6.304347826086957,
"grad_norm": 0.11389991432865713,
"learning_rate": 1.1916483387045986e-06,
"loss": 0.9748,
"step": 2465
},
{
"epoch": 6.30690537084399,
"grad_norm": 0.11531863459473257,
"learning_rate": 1.1829895871869511e-06,
"loss": 0.9614,
"step": 2466
},
{
"epoch": 6.309462915601023,
"grad_norm": 0.12161970256887526,
"learning_rate": 1.1743614497687084e-06,
"loss": 0.9762,
"step": 2467
},
{
"epoch": 6.312020460358056,
"grad_norm": 0.1211226712961686,
"learning_rate": 1.1657639404873145e-06,
"loss": 0.9837,
"step": 2468
},
{
"epoch": 6.31457800511509,
"grad_norm": 0.12353644559041659,
"learning_rate": 1.1571970733303784e-06,
"loss": 0.9676,
"step": 2469
},
{
"epoch": 6.3171355498721224,
"grad_norm": 0.11787181380295583,
"learning_rate": 1.1486608622356577e-06,
"loss": 0.9711,
"step": 2470
},
{
"epoch": 6.319693094629156,
"grad_norm": 0.11697925222871619,
"learning_rate": 1.14015532109103e-06,
"loss": 0.9832,
"step": 2471
},
{
"epoch": 6.322250639386189,
"grad_norm": 0.11230654846722168,
"learning_rate": 1.1316804637344835e-06,
"loss": 0.9768,
"step": 2472
},
{
"epoch": 6.324808184143222,
"grad_norm": 0.11484816919342249,
"learning_rate": 1.1232363039540829e-06,
"loss": 0.9794,
"step": 2473
},
{
"epoch": 6.327365728900256,
"grad_norm": 0.12103122369973704,
"learning_rate": 1.1148228554879404e-06,
"loss": 0.9851,
"step": 2474
},
{
"epoch": 6.329923273657289,
"grad_norm": 0.11866167434044839,
"learning_rate": 1.106440132024218e-06,
"loss": 0.9901,
"step": 2475
},
{
"epoch": 6.332480818414322,
"grad_norm": 0.12131943315693815,
"learning_rate": 1.0980881472010795e-06,
"loss": 1.0358,
"step": 2476
},
{
"epoch": 6.335038363171355,
"grad_norm": 0.12244899216381143,
"learning_rate": 1.0897669146066848e-06,
"loss": 0.9918,
"step": 2477
},
{
"epoch": 6.337595907928389,
"grad_norm": 0.11569574424112362,
"learning_rate": 1.081476447779153e-06,
"loss": 0.9591,
"step": 2478
},
{
"epoch": 6.340153452685422,
"grad_norm": 0.11836287519641907,
"learning_rate": 1.0732167602065612e-06,
"loss": 1.0053,
"step": 2479
},
{
"epoch": 6.342710997442455,
"grad_norm": 0.11846637303569181,
"learning_rate": 1.064987865326903e-06,
"loss": 0.9666,
"step": 2480
},
{
"epoch": 6.345268542199489,
"grad_norm": 0.11704066452635835,
"learning_rate": 1.056789776528071e-06,
"loss": 0.9817,
"step": 2481
},
{
"epoch": 6.3478260869565215,
"grad_norm": 0.1131509069280466,
"learning_rate": 1.048622507147854e-06,
"loss": 0.9978,
"step": 2482
},
{
"epoch": 6.350383631713555,
"grad_norm": 0.11368327856953235,
"learning_rate": 1.0404860704738784e-06,
"loss": 0.9775,
"step": 2483
},
{
"epoch": 6.352941176470588,
"grad_norm": 0.11816787304756617,
"learning_rate": 1.0323804797436243e-06,
"loss": 1.0069,
"step": 2484
},
{
"epoch": 6.3554987212276215,
"grad_norm": 0.11970422326419031,
"learning_rate": 1.0243057481443742e-06,
"loss": 1.0009,
"step": 2485
},
{
"epoch": 6.358056265984655,
"grad_norm": 0.1183905188506567,
"learning_rate": 1.016261888813217e-06,
"loss": 0.9976,
"step": 2486
},
{
"epoch": 6.360613810741688,
"grad_norm": 0.11798613162368961,
"learning_rate": 1.0082489148370112e-06,
"loss": 0.9754,
"step": 2487
},
{
"epoch": 6.3631713554987215,
"grad_norm": 0.12245420475095044,
"learning_rate": 1.0002668392523551e-06,
"loss": 0.9511,
"step": 2488
},
{
"epoch": 6.365728900255754,
"grad_norm": 0.11709971409522389,
"learning_rate": 9.923156750455964e-07,
"loss": 0.9786,
"step": 2489
},
{
"epoch": 6.368286445012788,
"grad_norm": 0.11702407698120149,
"learning_rate": 9.843954351527763e-07,
"loss": 0.9841,
"step": 2490
},
{
"epoch": 6.370843989769821,
"grad_norm": 0.11895816701374938,
"learning_rate": 9.765061324596336e-07,
"loss": 0.9946,
"step": 2491
},
{
"epoch": 6.373401534526854,
"grad_norm": 0.12261418476030701,
"learning_rate": 9.686477798015658e-07,
"loss": 0.9956,
"step": 2492
},
{
"epoch": 6.375959079283888,
"grad_norm": 0.1158763034760436,
"learning_rate": 9.608203899636238e-07,
"loss": 0.9678,
"step": 2493
},
{
"epoch": 6.378516624040921,
"grad_norm": 0.11482786064105056,
"learning_rate": 9.530239756804848e-07,
"loss": 0.9856,
"step": 2494
},
{
"epoch": 6.381074168797954,
"grad_norm": 0.11940232322986216,
"learning_rate": 9.452585496364208e-07,
"loss": 0.9869,
"step": 2495
},
{
"epoch": 6.383631713554987,
"grad_norm": 0.11700365366281955,
"learning_rate": 9.37524124465301e-07,
"loss": 0.9736,
"step": 2496
},
{
"epoch": 6.3861892583120206,
"grad_norm": 0.11347454822868741,
"learning_rate": 9.298207127505487e-07,
"loss": 0.9849,
"step": 2497
},
{
"epoch": 6.388746803069053,
"grad_norm": 0.12525519290709305,
"learning_rate": 9.221483270251319e-07,
"loss": 0.9606,
"step": 2498
},
{
"epoch": 6.391304347826087,
"grad_norm": 0.11859223351004591,
"learning_rate": 9.145069797715411e-07,
"loss": 1.0104,
"step": 2499
},
{
"epoch": 6.3938618925831205,
"grad_norm": 0.11218061014605829,
"learning_rate": 9.068966834217719e-07,
"loss": 0.9559,
"step": 2500
},
{
"epoch": 6.396419437340153,
"grad_norm": 0.12085793603634236,
"learning_rate": 8.993174503573088e-07,
"loss": 1.0167,
"step": 2501
},
{
"epoch": 6.398976982097187,
"grad_norm": 0.11567937773896561,
"learning_rate": 8.917692929090771e-07,
"loss": 0.9985,
"step": 2502
},
{
"epoch": 6.40153452685422,
"grad_norm": 0.11668551571603628,
"learning_rate": 8.842522233574646e-07,
"loss": 0.97,
"step": 2503
},
{
"epoch": 6.404092071611253,
"grad_norm": 0.1192493682038452,
"learning_rate": 8.767662539322686e-07,
"loss": 0.9627,
"step": 2504
},
{
"epoch": 6.406649616368286,
"grad_norm": 0.12194046852183779,
"learning_rate": 8.693113968127021e-07,
"loss": 0.9915,
"step": 2505
},
{
"epoch": 6.40920716112532,
"grad_norm": 0.11664318973679436,
"learning_rate": 8.61887664127341e-07,
"loss": 0.9756,
"step": 2506
},
{
"epoch": 6.411764705882353,
"grad_norm": 0.11392240551953355,
"learning_rate": 8.544950679541398e-07,
"loss": 0.9895,
"step": 2507
},
{
"epoch": 6.414322250639386,
"grad_norm": 0.11096676140165976,
"learning_rate": 8.471336203203929e-07,
"loss": 0.9536,
"step": 2508
},
{
"epoch": 6.41687979539642,
"grad_norm": 0.1155441359281539,
"learning_rate": 8.398033332027066e-07,
"loss": 0.9799,
"step": 2509
},
{
"epoch": 6.419437340153452,
"grad_norm": 0.12441513407230914,
"learning_rate": 8.325042185270083e-07,
"loss": 0.9786,
"step": 2510
},
{
"epoch": 6.421994884910486,
"grad_norm": 0.12338462492254433,
"learning_rate": 8.252362881684961e-07,
"loss": 0.9881,
"step": 2511
},
{
"epoch": 6.42455242966752,
"grad_norm": 0.11798502988687341,
"learning_rate": 8.17999553951645e-07,
"loss": 0.9997,
"step": 2512
},
{
"epoch": 6.427109974424552,
"grad_norm": 0.11452824885129483,
"learning_rate": 8.107940276501569e-07,
"loss": 0.9677,
"step": 2513
},
{
"epoch": 6.429667519181586,
"grad_norm": 0.11426234122255462,
"learning_rate": 8.036197209869812e-07,
"loss": 0.9683,
"step": 2514
},
{
"epoch": 6.432225063938619,
"grad_norm": 0.1121834450008384,
"learning_rate": 7.964766456342676e-07,
"loss": 0.9764,
"step": 2515
},
{
"epoch": 6.434782608695652,
"grad_norm": 0.11894958699980523,
"learning_rate": 7.893648132133514e-07,
"loss": 0.9724,
"step": 2516
},
{
"epoch": 6.437340153452685,
"grad_norm": 0.11057595561633296,
"learning_rate": 7.822842352947369e-07,
"loss": 0.956,
"step": 2517
},
{
"epoch": 6.439897698209719,
"grad_norm": 0.1149030674791001,
"learning_rate": 7.752349233980804e-07,
"loss": 0.9339,
"step": 2518
},
{
"epoch": 6.442455242966752,
"grad_norm": 0.11686222681540376,
"learning_rate": 7.682168889921793e-07,
"loss": 0.99,
"step": 2519
},
{
"epoch": 6.445012787723785,
"grad_norm": 0.11933666732733642,
"learning_rate": 7.612301434949265e-07,
"loss": 0.9988,
"step": 2520
},
{
"epoch": 6.447570332480819,
"grad_norm": 0.1139523213221972,
"learning_rate": 7.542746982733251e-07,
"loss": 0.9918,
"step": 2521
},
{
"epoch": 6.450127877237851,
"grad_norm": 0.12090979279952375,
"learning_rate": 7.47350564643452e-07,
"loss": 0.9931,
"step": 2522
},
{
"epoch": 6.452685421994885,
"grad_norm": 0.11965232822444315,
"learning_rate": 7.404577538704383e-07,
"loss": 1.0067,
"step": 2523
},
{
"epoch": 6.455242966751918,
"grad_norm": 0.11601978610605555,
"learning_rate": 7.335962771684535e-07,
"loss": 1.0089,
"step": 2524
},
{
"epoch": 6.457800511508951,
"grad_norm": 0.11767384429626374,
"learning_rate": 7.267661457006925e-07,
"loss": 0.9806,
"step": 2525
},
{
"epoch": 6.460358056265985,
"grad_norm": 0.1125374896391414,
"learning_rate": 7.199673705793553e-07,
"loss": 0.9856,
"step": 2526
},
{
"epoch": 6.462915601023018,
"grad_norm": 0.1228635267075819,
"learning_rate": 7.131999628656205e-07,
"loss": 1.0104,
"step": 2527
},
{
"epoch": 6.465473145780051,
"grad_norm": 0.11474314680272235,
"learning_rate": 7.064639335696344e-07,
"loss": 0.9822,
"step": 2528
},
{
"epoch": 6.468030690537084,
"grad_norm": 0.1152480534777416,
"learning_rate": 6.997592936505016e-07,
"loss": 0.9544,
"step": 2529
},
{
"epoch": 6.470588235294118,
"grad_norm": 0.11246736262937673,
"learning_rate": 6.93086054016252e-07,
"loss": 0.9967,
"step": 2530
},
{
"epoch": 6.4731457800511505,
"grad_norm": 0.11518337078601859,
"learning_rate": 6.864442255238235e-07,
"loss": 0.9897,
"step": 2531
},
{
"epoch": 6.475703324808184,
"grad_norm": 0.11435505822823681,
"learning_rate": 6.798338189790588e-07,
"loss": 0.9707,
"step": 2532
},
{
"epoch": 6.478260869565218,
"grad_norm": 0.11582740550813019,
"learning_rate": 6.732548451366794e-07,
"loss": 1.0011,
"step": 2533
},
{
"epoch": 6.4808184143222505,
"grad_norm": 0.11548398122397063,
"learning_rate": 6.667073147002589e-07,
"loss": 0.9501,
"step": 2534
},
{
"epoch": 6.483375959079284,
"grad_norm": 0.11612019126679186,
"learning_rate": 6.601912383222209e-07,
"loss": 1.0076,
"step": 2535
},
{
"epoch": 6.485933503836317,
"grad_norm": 0.11305499788986495,
"learning_rate": 6.537066266038183e-07,
"loss": 0.9894,
"step": 2536
},
{
"epoch": 6.4884910485933505,
"grad_norm": 0.11605374070817036,
"learning_rate": 6.472534900951077e-07,
"loss": 0.9645,
"step": 2537
},
{
"epoch": 6.491048593350383,
"grad_norm": 0.11089498292281949,
"learning_rate": 6.408318392949353e-07,
"loss": 0.9629,
"step": 2538
},
{
"epoch": 6.493606138107417,
"grad_norm": 0.1109910136998987,
"learning_rate": 6.34441684650926e-07,
"loss": 0.9611,
"step": 2539
},
{
"epoch": 6.4961636828644505,
"grad_norm": 0.1258539552506177,
"learning_rate": 6.28083036559466e-07,
"loss": 0.9824,
"step": 2540
},
{
"epoch": 6.498721227621483,
"grad_norm": 0.11680322621265218,
"learning_rate": 6.21755905365673e-07,
"loss": 1.0106,
"step": 2541
},
{
"epoch": 6.501278772378517,
"grad_norm": 0.1193164243220222,
"learning_rate": 6.154603013633975e-07,
"loss": 0.9841,
"step": 2542
},
{
"epoch": 6.5038363171355495,
"grad_norm": 0.11333571042569453,
"learning_rate": 6.09196234795193e-07,
"loss": 0.9706,
"step": 2543
},
{
"epoch": 6.506393861892583,
"grad_norm": 0.11542003113246305,
"learning_rate": 6.029637158523027e-07,
"loss": 0.9855,
"step": 2544
},
{
"epoch": 6.508951406649617,
"grad_norm": 0.1195829156189246,
"learning_rate": 5.967627546746469e-07,
"loss": 0.9966,
"step": 2545
},
{
"epoch": 6.5115089514066495,
"grad_norm": 0.12359363287176175,
"learning_rate": 5.905933613507997e-07,
"loss": 0.9721,
"step": 2546
},
{
"epoch": 6.514066496163683,
"grad_norm": 0.11877950994086818,
"learning_rate": 5.844555459179812e-07,
"loss": 0.9849,
"step": 2547
},
{
"epoch": 6.516624040920716,
"grad_norm": 0.11314777389167097,
"learning_rate": 5.783493183620304e-07,
"loss": 0.9707,
"step": 2548
},
{
"epoch": 6.5191815856777495,
"grad_norm": 0.11470632141145574,
"learning_rate": 5.72274688617398e-07,
"loss": 0.9906,
"step": 2549
},
{
"epoch": 6.521739130434782,
"grad_norm": 0.11489752644831927,
"learning_rate": 5.662316665671297e-07,
"loss": 1.0144,
"step": 2550
},
{
"epoch": 6.524296675191816,
"grad_norm": 0.10839103801041522,
"learning_rate": 5.60220262042841e-07,
"loss": 0.9772,
"step": 2551
},
{
"epoch": 6.526854219948849,
"grad_norm": 0.11706130646360133,
"learning_rate": 5.542404848247151e-07,
"loss": 0.9818,
"step": 2552
},
{
"epoch": 6.529411764705882,
"grad_norm": 0.12003303130567661,
"learning_rate": 5.482923446414701e-07,
"loss": 0.9981,
"step": 2553
},
{
"epoch": 6.531969309462916,
"grad_norm": 0.11845631399240932,
"learning_rate": 5.423758511703647e-07,
"loss": 0.9858,
"step": 2554
},
{
"epoch": 6.534526854219949,
"grad_norm": 0.119066755748317,
"learning_rate": 5.364910140371593e-07,
"loss": 0.9826,
"step": 2555
},
{
"epoch": 6.537084398976982,
"grad_norm": 0.1167547859112427,
"learning_rate": 5.306378428161152e-07,
"loss": 0.9531,
"step": 2556
},
{
"epoch": 6.539641943734015,
"grad_norm": 0.11392111178763799,
"learning_rate": 5.248163470299794e-07,
"loss": 0.986,
"step": 2557
},
{
"epoch": 6.542199488491049,
"grad_norm": 0.11758729422117363,
"learning_rate": 5.190265361499625e-07,
"loss": 0.9694,
"step": 2558
},
{
"epoch": 6.544757033248082,
"grad_norm": 0.11318667288737702,
"learning_rate": 5.132684195957183e-07,
"loss": 0.9894,
"step": 2559
},
{
"epoch": 6.547314578005115,
"grad_norm": 0.11630427878786241,
"learning_rate": 5.075420067353509e-07,
"loss": 0.9771,
"step": 2560
},
{
"epoch": 6.549872122762149,
"grad_norm": 0.11728288250968313,
"learning_rate": 5.018473068853702e-07,
"loss": 0.9687,
"step": 2561
},
{
"epoch": 6.552429667519181,
"grad_norm": 0.11233910385595633,
"learning_rate": 4.961843293106983e-07,
"loss": 1.0008,
"step": 2562
},
{
"epoch": 6.554987212276215,
"grad_norm": 0.11374742174339529,
"learning_rate": 4.905530832246431e-07,
"loss": 0.9679,
"step": 2563
},
{
"epoch": 6.557544757033249,
"grad_norm": 0.11869294303036608,
"learning_rate": 4.849535777888936e-07,
"loss": 1.0095,
"step": 2564
},
{
"epoch": 6.560102301790281,
"grad_norm": 0.11027201850816472,
"learning_rate": 4.793858221134961e-07,
"loss": 0.9605,
"step": 2565
},
{
"epoch": 6.562659846547315,
"grad_norm": 0.11536912256900325,
"learning_rate": 4.7384982525683577e-07,
"loss": 0.943,
"step": 2566
},
{
"epoch": 6.565217391304348,
"grad_norm": 0.10942403966278201,
"learning_rate": 4.6834559622563893e-07,
"loss": 0.9423,
"step": 2567
},
{
"epoch": 6.567774936061381,
"grad_norm": 0.11396546837926291,
"learning_rate": 4.6287314397493787e-07,
"loss": 1.0053,
"step": 2568
},
{
"epoch": 6.570332480818414,
"grad_norm": 0.11459687522209497,
"learning_rate": 4.5743247740807297e-07,
"loss": 0.9939,
"step": 2569
},
{
"epoch": 6.572890025575448,
"grad_norm": 0.11546094162304561,
"learning_rate": 4.520236053766636e-07,
"loss": 0.9901,
"step": 2570
},
{
"epoch": 6.57544757033248,
"grad_norm": 0.11441296647941865,
"learning_rate": 4.46646536680615e-07,
"loss": 0.9933,
"step": 2571
},
{
"epoch": 6.578005115089514,
"grad_norm": 0.11468937955843611,
"learning_rate": 4.413012800680738e-07,
"loss": 0.9663,
"step": 2572
},
{
"epoch": 6.580562659846548,
"grad_norm": 0.11164935414638426,
"learning_rate": 4.359878442354415e-07,
"loss": 0.9721,
"step": 2573
},
{
"epoch": 6.58312020460358,
"grad_norm": 0.11272488044522909,
"learning_rate": 4.3070623782734746e-07,
"loss": 0.9855,
"step": 2574
},
{
"epoch": 6.585677749360614,
"grad_norm": 0.11339026549490953,
"learning_rate": 4.2545646943663146e-07,
"loss": 0.969,
"step": 2575
},
{
"epoch": 6.588235294117647,
"grad_norm": 0.11170706786475929,
"learning_rate": 4.202385476043436e-07,
"loss": 0.9756,
"step": 2576
},
{
"epoch": 6.59079283887468,
"grad_norm": 0.11558141407077696,
"learning_rate": 4.150524808197087e-07,
"loss": 0.9753,
"step": 2577
},
{
"epoch": 6.593350383631714,
"grad_norm": 0.1171723039107249,
"learning_rate": 4.098982775201421e-07,
"loss": 0.9891,
"step": 2578
},
{
"epoch": 6.595907928388747,
"grad_norm": 0.11328248956877658,
"learning_rate": 4.0477594609120487e-07,
"loss": 0.988,
"step": 2579
},
{
"epoch": 6.59846547314578,
"grad_norm": 0.10713511888219494,
"learning_rate": 3.996854948666085e-07,
"loss": 0.9748,
"step": 2580
},
{
"epoch": 6.601023017902813,
"grad_norm": 0.11397080930100252,
"learning_rate": 3.946269321282037e-07,
"loss": 1.0036,
"step": 2581
},
{
"epoch": 6.603580562659847,
"grad_norm": 0.11026004482109196,
"learning_rate": 3.896002661059517e-07,
"loss": 0.9921,
"step": 2582
},
{
"epoch": 6.6061381074168795,
"grad_norm": 0.11101424733122804,
"learning_rate": 3.846055049779285e-07,
"loss": 0.9775,
"step": 2583
},
{
"epoch": 6.608695652173913,
"grad_norm": 0.11648475238053471,
"learning_rate": 3.7964265687029155e-07,
"loss": 0.9985,
"step": 2584
},
{
"epoch": 6.611253196930946,
"grad_norm": 0.1182309737049125,
"learning_rate": 3.747117298572911e-07,
"loss": 0.9545,
"step": 2585
},
{
"epoch": 6.6138107416879794,
"grad_norm": 0.1143121392907517,
"learning_rate": 3.698127319612321e-07,
"loss": 0.9841,
"step": 2586
},
{
"epoch": 6.616368286445013,
"grad_norm": 0.11659139001884598,
"learning_rate": 3.6494567115248125e-07,
"loss": 1.0425,
"step": 2587
},
{
"epoch": 6.618925831202046,
"grad_norm": 0.11265994981069657,
"learning_rate": 3.6011055534943996e-07,
"loss": 0.9963,
"step": 2588
},
{
"epoch": 6.621483375959079,
"grad_norm": 0.11255982506988561,
"learning_rate": 3.5530739241853797e-07,
"loss": 0.9884,
"step": 2589
},
{
"epoch": 6.624040920716112,
"grad_norm": 0.11252108389474363,
"learning_rate": 3.505361901742288e-07,
"loss": 1.0185,
"step": 2590
},
{
"epoch": 6.626598465473146,
"grad_norm": 0.11108564117473062,
"learning_rate": 3.4579695637894984e-07,
"loss": 0.9821,
"step": 2591
},
{
"epoch": 6.629156010230179,
"grad_norm": 0.1100233099971943,
"learning_rate": 3.4108969874314225e-07,
"loss": 0.9885,
"step": 2592
},
{
"epoch": 6.631713554987212,
"grad_norm": 0.11214886155471082,
"learning_rate": 3.364144249252221e-07,
"loss": 0.9934,
"step": 2593
},
{
"epoch": 6.634271099744246,
"grad_norm": 0.11123313453855575,
"learning_rate": 3.317711425315673e-07,
"loss": 1.0043,
"step": 2594
},
{
"epoch": 6.6368286445012785,
"grad_norm": 0.11349701821282343,
"learning_rate": 3.2715985911650593e-07,
"loss": 0.9784,
"step": 2595
},
{
"epoch": 6.639386189258312,
"grad_norm": 0.1124789272677662,
"learning_rate": 3.2258058218230803e-07,
"loss": 0.9746,
"step": 2596
},
{
"epoch": 6.641943734015345,
"grad_norm": 0.110343893039951,
"learning_rate": 3.18033319179174e-07,
"loss": 0.9513,
"step": 2597
},
{
"epoch": 6.6445012787723785,
"grad_norm": 0.11120670784044137,
"learning_rate": 3.135180775052082e-07,
"loss": 0.9935,
"step": 2598
},
{
"epoch": 6.647058823529412,
"grad_norm": 0.1100839287666045,
"learning_rate": 3.090348645064345e-07,
"loss": 0.9729,
"step": 2599
},
{
"epoch": 6.649616368286445,
"grad_norm": 0.11430986538953229,
"learning_rate": 3.0458368747675824e-07,
"loss": 0.975,
"step": 2600
},
{
"epoch": 6.6521739130434785,
"grad_norm": 0.11661794371841362,
"learning_rate": 3.0016455365796673e-07,
"loss": 0.9978,
"step": 2601
},
{
"epoch": 6.654731457800511,
"grad_norm": 0.1069222233565192,
"learning_rate": 2.9577747023971313e-07,
"loss": 0.9714,
"step": 2602
},
{
"epoch": 6.657289002557545,
"grad_norm": 0.11028749076820085,
"learning_rate": 2.9142244435950595e-07,
"loss": 0.9933,
"step": 2603
},
{
"epoch": 6.659846547314578,
"grad_norm": 0.1134281573066445,
"learning_rate": 2.8709948310270407e-07,
"loss": 0.9924,
"step": 2604
},
{
"epoch": 6.662404092071611,
"grad_norm": 0.11433726640069898,
"learning_rate": 2.8280859350249267e-07,
"loss": 0.9705,
"step": 2605
},
{
"epoch": 6.664961636828645,
"grad_norm": 0.10918039328987154,
"learning_rate": 2.7854978253987865e-07,
"loss": 0.9982,
"step": 2606
},
{
"epoch": 6.667519181585678,
"grad_norm": 0.11279698766292603,
"learning_rate": 2.743230571436861e-07,
"loss": 0.9758,
"step": 2607
},
{
"epoch": 6.670076726342711,
"grad_norm": 0.11461667827718962,
"learning_rate": 2.701284241905322e-07,
"loss": 0.9564,
"step": 2608
},
{
"epoch": 6.672634271099744,
"grad_norm": 0.11714027276234738,
"learning_rate": 2.6596589050482016e-07,
"loss": 0.9902,
"step": 2609
},
{
"epoch": 6.675191815856778,
"grad_norm": 0.10864995785548583,
"learning_rate": 2.618354628587305e-07,
"loss": 0.9968,
"step": 2610
},
{
"epoch": 6.677749360613811,
"grad_norm": 0.11500531682078577,
"learning_rate": 2.5773714797221463e-07,
"loss": 0.9617,
"step": 2611
},
{
"epoch": 6.680306905370844,
"grad_norm": 0.11255867506461198,
"learning_rate": 2.5367095251297436e-07,
"loss": 0.986,
"step": 2612
},
{
"epoch": 6.6828644501278776,
"grad_norm": 0.11021418539145582,
"learning_rate": 2.496368830964513e-07,
"loss": 0.9602,
"step": 2613
},
{
"epoch": 6.68542199488491,
"grad_norm": 0.11331500725668849,
"learning_rate": 2.4563494628582873e-07,
"loss": 0.9669,
"step": 2614
},
{
"epoch": 6.687979539641944,
"grad_norm": 0.10947642913755386,
"learning_rate": 2.416651485920074e-07,
"loss": 0.9699,
"step": 2615
},
{
"epoch": 6.690537084398977,
"grad_norm": 0.11064047422303806,
"learning_rate": 2.377274964735965e-07,
"loss": 1.0127,
"step": 2616
},
{
"epoch": 6.69309462915601,
"grad_norm": 0.11093673915126204,
"learning_rate": 2.3382199633691148e-07,
"loss": 0.9853,
"step": 2617
},
{
"epoch": 6.695652173913043,
"grad_norm": 0.11595508347561423,
"learning_rate": 2.299486545359586e-07,
"loss": 0.9622,
"step": 2618
},
{
"epoch": 6.698209718670077,
"grad_norm": 0.11318170584535443,
"learning_rate": 2.2610747737241922e-07,
"loss": 0.9618,
"step": 2619
},
{
"epoch": 6.70076726342711,
"grad_norm": 0.11149052404888277,
"learning_rate": 2.2229847109565216e-07,
"loss": 0.9597,
"step": 2620
},
{
"epoch": 6.703324808184143,
"grad_norm": 0.1159741337280457,
"learning_rate": 2.1852164190267144e-07,
"loss": 0.9825,
"step": 2621
},
{
"epoch": 6.705882352941177,
"grad_norm": 0.11519134340806857,
"learning_rate": 2.1477699593814182e-07,
"loss": 1.0101,
"step": 2622
},
{
"epoch": 6.708439897698209,
"grad_norm": 0.1098906463037301,
"learning_rate": 2.1106453929436553e-07,
"loss": 0.9725,
"step": 2623
},
{
"epoch": 6.710997442455243,
"grad_norm": 0.11304781158930503,
"learning_rate": 2.0738427801128002e-07,
"loss": 0.9666,
"step": 2624
},
{
"epoch": 6.713554987212277,
"grad_norm": 0.1104442198230062,
"learning_rate": 2.0373621807644238e-07,
"loss": 0.9279,
"step": 2625
},
{
"epoch": 6.716112531969309,
"grad_norm": 0.11052148773696956,
"learning_rate": 2.001203654250139e-07,
"loss": 0.9943,
"step": 2626
},
{
"epoch": 6.718670076726343,
"grad_norm": 0.10873558382847161,
"learning_rate": 1.9653672593976213e-07,
"loss": 1.022,
"step": 2627
},
{
"epoch": 6.721227621483376,
"grad_norm": 0.11287622715393,
"learning_rate": 1.9298530545104555e-07,
"loss": 0.9879,
"step": 2628
},
{
"epoch": 6.723785166240409,
"grad_norm": 0.10844134025691674,
"learning_rate": 1.8946610973680223e-07,
"loss": 0.9809,
"step": 2629
},
{
"epoch": 6.726342710997442,
"grad_norm": 0.10988879894088321,
"learning_rate": 1.859791445225434e-07,
"loss": 0.9921,
"step": 2630
},
{
"epoch": 6.728900255754476,
"grad_norm": 0.11297996886125798,
"learning_rate": 1.8252441548134436e-07,
"loss": 0.9945,
"step": 2631
},
{
"epoch": 6.731457800511509,
"grad_norm": 0.10979109703535442,
"learning_rate": 1.7910192823383132e-07,
"loss": 0.9795,
"step": 2632
},
{
"epoch": 6.734015345268542,
"grad_norm": 0.10844303788100032,
"learning_rate": 1.757116883481813e-07,
"loss": 0.9814,
"step": 2633
},
{
"epoch": 6.736572890025576,
"grad_norm": 0.11295455674769658,
"learning_rate": 1.7235370134009554e-07,
"loss": 0.9913,
"step": 2634
},
{
"epoch": 6.739130434782608,
"grad_norm": 0.11282650537130356,
"learning_rate": 1.6902797267281722e-07,
"loss": 0.9622,
"step": 2635
},
{
"epoch": 6.741687979539642,
"grad_norm": 0.10903129964817249,
"learning_rate": 1.6573450775709378e-07,
"loss": 0.9804,
"step": 2636
},
{
"epoch": 6.744245524296675,
"grad_norm": 0.11138147806122452,
"learning_rate": 1.6247331195118787e-07,
"loss": 0.9935,
"step": 2637
},
{
"epoch": 6.746803069053708,
"grad_norm": 0.10658566136145321,
"learning_rate": 1.5924439056085983e-07,
"loss": 0.949,
"step": 2638
},
{
"epoch": 6.749360613810742,
"grad_norm": 0.11152040502568372,
"learning_rate": 1.5604774883936524e-07,
"loss": 0.9701,
"step": 2639
},
{
"epoch": 6.751918158567775,
"grad_norm": 0.10930678724141839,
"learning_rate": 1.528833919874395e-07,
"loss": 1.0083,
"step": 2640
},
{
"epoch": 6.754475703324808,
"grad_norm": 0.11176204244154962,
"learning_rate": 1.4975132515329117e-07,
"loss": 1.0073,
"step": 2641
},
{
"epoch": 6.757033248081841,
"grad_norm": 0.10930037432682403,
"learning_rate": 1.4665155343259963e-07,
"loss": 0.9646,
"step": 2642
},
{
"epoch": 6.759590792838875,
"grad_norm": 0.10852100024174001,
"learning_rate": 1.435840818684997e-07,
"loss": 0.9792,
"step": 2643
},
{
"epoch": 6.762148337595908,
"grad_norm": 0.10996449804286494,
"learning_rate": 1.4054891545157712e-07,
"loss": 0.9778,
"step": 2644
},
{
"epoch": 6.764705882352941,
"grad_norm": 0.10713512977370486,
"learning_rate": 1.3754605911985297e-07,
"loss": 0.9682,
"step": 2645
},
{
"epoch": 6.767263427109975,
"grad_norm": 0.11235839607507679,
"learning_rate": 1.345755177587904e-07,
"loss": 0.99,
"step": 2646
},
{
"epoch": 6.7698209718670075,
"grad_norm": 0.11301175673357075,
"learning_rate": 1.3163729620127463e-07,
"loss": 0.9741,
"step": 2647
},
{
"epoch": 6.772378516624041,
"grad_norm": 0.1087102227933997,
"learning_rate": 1.2873139922760626e-07,
"loss": 0.966,
"step": 2648
},
{
"epoch": 6.774936061381074,
"grad_norm": 0.11524543285700066,
"learning_rate": 1.2585783156550126e-07,
"loss": 0.9943,
"step": 2649
},
{
"epoch": 6.7774936061381075,
"grad_norm": 0.109619234671289,
"learning_rate": 1.2301659789007325e-07,
"loss": 0.9751,
"step": 2650
},
{
"epoch": 6.78005115089514,
"grad_norm": 0.11377111310304723,
"learning_rate": 1.202077028238291e-07,
"loss": 0.9913,
"step": 2651
},
{
"epoch": 6.782608695652174,
"grad_norm": 0.11340763837752178,
"learning_rate": 1.1743115093666657e-07,
"loss": 0.9701,
"step": 2652
},
{
"epoch": 6.7851662404092075,
"grad_norm": 0.11407639631694404,
"learning_rate": 1.1468694674586333e-07,
"loss": 0.989,
"step": 2653
},
{
"epoch": 6.78772378516624,
"grad_norm": 0.10984842462186982,
"learning_rate": 1.1197509471607026e-07,
"loss": 0.9891,
"step": 2654
},
{
"epoch": 6.790281329923274,
"grad_norm": 0.11049084551976941,
"learning_rate": 1.092955992592959e-07,
"loss": 0.9869,
"step": 2655
},
{
"epoch": 6.792838874680307,
"grad_norm": 0.11014718715893779,
"learning_rate": 1.066484647349153e-07,
"loss": 0.9461,
"step": 2656
},
{
"epoch": 6.79539641943734,
"grad_norm": 0.10958952496889618,
"learning_rate": 1.0403369544965014e-07,
"loss": 0.9563,
"step": 2657
},
{
"epoch": 6.797953964194374,
"grad_norm": 0.10803458480621711,
"learning_rate": 1.0145129565756639e-07,
"loss": 0.9591,
"step": 2658
},
{
"epoch": 6.8005115089514065,
"grad_norm": 0.112922827976531,
"learning_rate": 9.890126956006552e-08,
"loss": 0.9881,
"step": 2659
},
{
"epoch": 6.80306905370844,
"grad_norm": 0.11135692745799174,
"learning_rate": 9.638362130588219e-08,
"loss": 1.0063,
"step": 2660
},
{
"epoch": 6.805626598465473,
"grad_norm": 0.11635356613011706,
"learning_rate": 9.389835499107325e-08,
"loss": 0.9893,
"step": 2661
},
{
"epoch": 6.8081841432225065,
"grad_norm": 0.10789603666003188,
"learning_rate": 9.144547465901099e-08,
"loss": 0.9449,
"step": 2662
},
{
"epoch": 6.810741687979539,
"grad_norm": 0.11474688949051133,
"learning_rate": 8.902498430037654e-08,
"loss": 1.0218,
"step": 2663
},
{
"epoch": 6.813299232736573,
"grad_norm": 0.10893369925466642,
"learning_rate": 8.663688785315982e-08,
"loss": 0.9897,
"step": 2664
},
{
"epoch": 6.8158567774936065,
"grad_norm": 0.11234204307215885,
"learning_rate": 8.428118920264183e-08,
"loss": 0.9801,
"step": 2665
},
{
"epoch": 6.818414322250639,
"grad_norm": 0.11192876764000358,
"learning_rate": 8.195789218139682e-08,
"loss": 0.9496,
"step": 2666
},
{
"epoch": 6.820971867007673,
"grad_norm": 0.10892579746670104,
"learning_rate": 7.966700056928566e-08,
"loss": 0.9667,
"step": 2667
},
{
"epoch": 6.823529411764706,
"grad_norm": 0.10935836559963666,
"learning_rate": 7.740851809344474e-08,
"loss": 1.0528,
"step": 2668
},
{
"epoch": 6.826086956521739,
"grad_norm": 0.11196618559061729,
"learning_rate": 7.518244842828371e-08,
"loss": 0.996,
"step": 2669
},
{
"epoch": 6.828644501278772,
"grad_norm": 0.11659883740428693,
"learning_rate": 7.298879519547886e-08,
"loss": 0.9905,
"step": 2670
},
{
"epoch": 6.831202046035806,
"grad_norm": 0.11546383395441213,
"learning_rate": 7.082756196396645e-08,
"loss": 0.9837,
"step": 2671
},
{
"epoch": 6.833759590792839,
"grad_norm": 0.10967450770430652,
"learning_rate": 6.869875224993827e-08,
"loss": 0.9564,
"step": 2672
},
{
"epoch": 6.836317135549872,
"grad_norm": 0.10741715549223828,
"learning_rate": 6.660236951683718e-08,
"loss": 0.9827,
"step": 2673
},
{
"epoch": 6.838874680306906,
"grad_norm": 0.10743011744438329,
"learning_rate": 6.45384171753416e-08,
"loss": 0.9618,
"step": 2674
},
{
"epoch": 6.841432225063938,
"grad_norm": 0.11137816645533803,
"learning_rate": 6.250689858337656e-08,
"loss": 0.9644,
"step": 2675
},
{
"epoch": 6.843989769820972,
"grad_norm": 0.11008999111546065,
"learning_rate": 6.050781704609376e-08,
"loss": 0.9912,
"step": 2676
},
{
"epoch": 6.846547314578006,
"grad_norm": 0.10774766884917021,
"learning_rate": 5.854117581587604e-08,
"loss": 0.974,
"step": 2677
},
{
"epoch": 6.849104859335038,
"grad_norm": 0.1079604201053252,
"learning_rate": 5.660697809232396e-08,
"loss": 0.9557,
"step": 2678
},
{
"epoch": 6.851662404092072,
"grad_norm": 0.11139319245164413,
"learning_rate": 5.470522702225589e-08,
"loss": 0.9971,
"step": 2679
},
{
"epoch": 6.854219948849105,
"grad_norm": 0.10986134223415223,
"learning_rate": 5.283592569970575e-08,
"loss": 0.9735,
"step": 2680
},
{
"epoch": 6.856777493606138,
"grad_norm": 0.10758519308257364,
"learning_rate": 5.099907716590524e-08,
"loss": 0.9848,
"step": 2681
},
{
"epoch": 6.859335038363171,
"grad_norm": 0.11574137644686665,
"learning_rate": 4.919468440929276e-08,
"loss": 0.9655,
"step": 2682
},
{
"epoch": 6.861892583120205,
"grad_norm": 0.1066340137804935,
"learning_rate": 4.7422750365506695e-08,
"loss": 0.9899,
"step": 2683
},
{
"epoch": 6.864450127877237,
"grad_norm": 0.10596514728367719,
"learning_rate": 4.568327791736771e-08,
"loss": 0.9774,
"step": 2684
},
{
"epoch": 6.867007672634271,
"grad_norm": 0.10889390656854558,
"learning_rate": 4.397626989489201e-08,
"loss": 0.9677,
"step": 2685
},
{
"epoch": 6.869565217391305,
"grad_norm": 0.109674980632228,
"learning_rate": 4.2301729075273634e-08,
"loss": 0.9956,
"step": 2686
},
{
"epoch": 6.872122762148337,
"grad_norm": 0.10872255519531301,
"learning_rate": 4.065965818288664e-08,
"loss": 0.9757,
"step": 2687
},
{
"epoch": 6.874680306905371,
"grad_norm": 0.10785302434686765,
"learning_rate": 3.9050059889274015e-08,
"loss": 0.9574,
"step": 2688
},
{
"epoch": 6.877237851662404,
"grad_norm": 0.10646051144289906,
"learning_rate": 3.7472936813158776e-08,
"loss": 0.9574,
"step": 2689
},
{
"epoch": 6.879795396419437,
"grad_norm": 0.10974969890403027,
"learning_rate": 3.59282915204151e-08,
"loss": 0.9791,
"step": 2690
},
{
"epoch": 6.882352941176471,
"grad_norm": 0.11005392637131686,
"learning_rate": 3.44161265240861e-08,
"loss": 0.9696,
"step": 2691
},
{
"epoch": 6.884910485933504,
"grad_norm": 0.111759037695925,
"learning_rate": 3.293644428437048e-08,
"loss": 1.012,
"step": 2692
},
{
"epoch": 6.887468030690537,
"grad_norm": 0.10845922465210457,
"learning_rate": 3.1489247208620345e-08,
"loss": 0.9679,
"step": 2693
},
{
"epoch": 6.89002557544757,
"grad_norm": 0.10850695405285556,
"learning_rate": 3.007453765133006e-08,
"loss": 0.9533,
"step": 2694
},
{
"epoch": 6.892583120204604,
"grad_norm": 0.11161719679869773,
"learning_rate": 2.8692317914147393e-08,
"loss": 0.9898,
"step": 2695
},
{
"epoch": 6.8951406649616365,
"grad_norm": 0.11086159978607008,
"learning_rate": 2.7342590245857947e-08,
"loss": 0.9758,
"step": 2696
},
{
"epoch": 6.89769820971867,
"grad_norm": 0.11270909900746691,
"learning_rate": 2.6025356842382942e-08,
"loss": 1.0029,
"step": 2697
},
{
"epoch": 6.900255754475703,
"grad_norm": 0.10920922570358585,
"learning_rate": 2.4740619846777002e-08,
"loss": 0.9846,
"step": 2698
},
{
"epoch": 6.9028132992327365,
"grad_norm": 0.10997450369893431,
"learning_rate": 2.348838134922815e-08,
"loss": 0.989,
"step": 2699
},
{
"epoch": 6.90537084398977,
"grad_norm": 0.10685421374853045,
"learning_rate": 2.2268643387051147e-08,
"loss": 0.9661,
"step": 2700
},
{
"epoch": 6.907928388746803,
"grad_norm": 0.11184254245358682,
"learning_rate": 2.108140794468305e-08,
"loss": 0.9788,
"step": 2701
},
{
"epoch": 6.910485933503836,
"grad_norm": 0.11080766149307342,
"learning_rate": 1.9926676953681002e-08,
"loss": 0.9922,
"step": 2702
},
{
"epoch": 6.913043478260869,
"grad_norm": 0.10838317350773852,
"learning_rate": 1.8804452292722208e-08,
"loss": 0.9772,
"step": 2703
},
{
"epoch": 6.915601023017903,
"grad_norm": 0.11076174771831925,
"learning_rate": 1.7714735787592862e-08,
"loss": 0.9628,
"step": 2704
},
{
"epoch": 6.918158567774936,
"grad_norm": 0.11189238137471301,
"learning_rate": 1.665752921119479e-08,
"loss": 0.9874,
"step": 2705
},
{
"epoch": 6.920716112531969,
"grad_norm": 0.1123205884537519,
"learning_rate": 1.5632834283536568e-08,
"loss": 1.0041,
"step": 2706
},
{
"epoch": 6.923273657289003,
"grad_norm": 0.11027737928028197,
"learning_rate": 1.4640652671735755e-08,
"loss": 0.9921,
"step": 2707
},
{
"epoch": 6.9258312020460355,
"grad_norm": 0.1109201749361578,
"learning_rate": 1.3680985990005558e-08,
"loss": 0.9811,
"step": 2708
},
{
"epoch": 6.928388746803069,
"grad_norm": 0.10741286252173589,
"learning_rate": 1.2753835799665937e-08,
"loss": 0.9899,
"step": 2709
},
{
"epoch": 6.930946291560103,
"grad_norm": 0.11092756446221406,
"learning_rate": 1.1859203609132508e-08,
"loss": 0.97,
"step": 2710
},
{
"epoch": 6.9335038363171355,
"grad_norm": 0.107532803777865,
"learning_rate": 1.0997090873914318e-08,
"loss": 0.9905,
"step": 2711
},
{
"epoch": 6.936061381074169,
"grad_norm": 0.10401305429963421,
"learning_rate": 1.0167498996618285e-08,
"loss": 0.9502,
"step": 2712
},
{
"epoch": 6.938618925831202,
"grad_norm": 0.10853889848857848,
"learning_rate": 9.370429326933661e-09,
"loss": 0.979,
"step": 2713
},
{
"epoch": 6.9411764705882355,
"grad_norm": 0.11334668317079354,
"learning_rate": 8.605883161647566e-09,
"loss": 0.9515,
"step": 2714
},
{
"epoch": 6.943734015345268,
"grad_norm": 0.1122988264248447,
"learning_rate": 7.873861744625011e-09,
"loss": 0.9935,
"step": 2715
},
{
"epoch": 6.946291560102302,
"grad_norm": 0.11252319871848504,
"learning_rate": 7.1743662668199985e-09,
"loss": 0.9958,
"step": 2716
},
{
"epoch": 6.948849104859335,
"grad_norm": 0.10720372022931549,
"learning_rate": 6.507397866271081e-09,
"loss": 0.9702,
"step": 2717
},
{
"epoch": 6.951406649616368,
"grad_norm": 0.1100260006823836,
"learning_rate": 5.872957628090259e-09,
"loss": 1.019,
"step": 2718
},
{
"epoch": 6.953964194373402,
"grad_norm": 0.10901934916778966,
"learning_rate": 5.271046584474082e-09,
"loss": 1.0061,
"step": 2719
},
{
"epoch": 6.956521739130435,
"grad_norm": 0.11305598232483986,
"learning_rate": 4.701665714692549e-09,
"loss": 0.9794,
"step": 2720
},
{
"epoch": 6.959079283887468,
"grad_norm": 0.10646553211763654,
"learning_rate": 4.164815945095768e-09,
"loss": 0.9811,
"step": 2721
},
{
"epoch": 6.961636828644501,
"grad_norm": 0.10938318967557946,
"learning_rate": 3.660498149100633e-09,
"loss": 0.9928,
"step": 2722
},
{
"epoch": 6.964194373401535,
"grad_norm": 0.10817259622625107,
"learning_rate": 3.1887131472041477e-09,
"loss": 0.9882,
"step": 2723
},
{
"epoch": 6.966751918158568,
"grad_norm": 0.11354018429688115,
"learning_rate": 2.7494617069701025e-09,
"loss": 1.0008,
"step": 2724
},
{
"epoch": 6.969309462915601,
"grad_norm": 0.11053582170013893,
"learning_rate": 2.342744543033515e-09,
"loss": 1.0147,
"step": 2725
},
{
"epoch": 6.971867007672635,
"grad_norm": 0.11201035927268471,
"learning_rate": 1.96856231709619e-09,
"loss": 0.9925,
"step": 2726
},
{
"epoch": 6.974424552429667,
"grad_norm": 0.11184125302836626,
"learning_rate": 1.6269156379289385e-09,
"loss": 0.9767,
"step": 2727
},
{
"epoch": 6.976982097186701,
"grad_norm": 0.11334081608709352,
"learning_rate": 1.3178050613738003e-09,
"loss": 0.9696,
"step": 2728
},
{
"epoch": 6.979539641943734,
"grad_norm": 0.109039012863987,
"learning_rate": 1.0412310903284983e-09,
"loss": 0.9939,
"step": 2729
},
{
"epoch": 6.982097186700767,
"grad_norm": 0.11126820670939559,
"learning_rate": 7.971941747642042e-10,
"loss": 1.0046,
"step": 2730
},
{
"epoch": 6.9846547314578,
"grad_norm": 0.11655821963595665,
"learning_rate": 5.856947117166556e-10,
"loss": 0.9954,
"step": 2731
},
{
"epoch": 6.987212276214834,
"grad_norm": 0.10922308250116784,
"learning_rate": 4.067330452772744e-10,
"loss": 0.9783,
"step": 2732
},
{
"epoch": 6.989769820971867,
"grad_norm": 0.1099651411615859,
"learning_rate": 2.603094666064898e-10,
"loss": 0.9651,
"step": 2733
},
{
"epoch": 6.9923273657289,
"grad_norm": 0.10674345048716981,
"learning_rate": 1.4642421392929706e-10,
"loss": 0.9411,
"step": 2734
},
{
"epoch": 6.994884910485934,
"grad_norm": 0.10826698511930209,
"learning_rate": 6.507747252859631e-11,
"loss": 0.9657,
"step": 2735
},
{
"epoch": 6.997442455242966,
"grad_norm": 0.10989561091331604,
"learning_rate": 1.6269374749633415e-11,
"loss": 1.0107,
"step": 2736
},
{
"epoch": 7.0,
"grad_norm": 0.10898191842884208,
"learning_rate": 0.0,
"loss": 0.9414,
"step": 2737
},
{
"epoch": 7.0,
"step": 2737,
"total_flos": 9969287656374272.0,
"train_loss": 1.0345919296651762,
"train_runtime": 97654.46,
"train_samples_per_second": 7.168,
"train_steps_per_second": 0.028
}
],
"logging_steps": 1.0,
"max_steps": 2737,
"num_input_tokens_seen": 0,
"num_train_epochs": 7,
"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": 9969287656374272.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}