haoningwu's picture
Upload folder using huggingface_hub
afa73f9 verified
Raw
History Blame Contribute Delete
112 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 647,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0015473887814313346,
"grad_norm": 69.93428202217697,
"learning_rate": 0.0,
"loss": 2.1932,
"step": 1
},
{
"epoch": 0.003094777562862669,
"grad_norm": 88.82346219685395,
"learning_rate": 5.000000000000001e-07,
"loss": 2.6136,
"step": 2
},
{
"epoch": 0.0046421663442940036,
"grad_norm": 73.75819190158867,
"learning_rate": 1.0000000000000002e-06,
"loss": 2.2841,
"step": 3
},
{
"epoch": 0.006189555125725338,
"grad_norm": 84.84920641442658,
"learning_rate": 1.5e-06,
"loss": 2.5951,
"step": 4
},
{
"epoch": 0.007736943907156673,
"grad_norm": 75.05902382047852,
"learning_rate": 2.0000000000000003e-06,
"loss": 2.4068,
"step": 5
},
{
"epoch": 0.009284332688588007,
"grad_norm": 51.681845802557184,
"learning_rate": 2.5e-06,
"loss": 1.8698,
"step": 6
},
{
"epoch": 0.010831721470019342,
"grad_norm": 48.34094681143097,
"learning_rate": 3e-06,
"loss": 1.8798,
"step": 7
},
{
"epoch": 0.012379110251450677,
"grad_norm": 25.478189595197318,
"learning_rate": 3.5e-06,
"loss": 1.2716,
"step": 8
},
{
"epoch": 0.013926499032882012,
"grad_norm": 25.882646528564464,
"learning_rate": 4.000000000000001e-06,
"loss": 1.3707,
"step": 9
},
{
"epoch": 0.015473887814313346,
"grad_norm": 19.310012292249482,
"learning_rate": 4.5e-06,
"loss": 0.6819,
"step": 10
},
{
"epoch": 0.01702127659574468,
"grad_norm": 15.750117335558071,
"learning_rate": 5e-06,
"loss": 0.6521,
"step": 11
},
{
"epoch": 0.018568665377176014,
"grad_norm": 11.31362602963611,
"learning_rate": 5.500000000000001e-06,
"loss": 0.5819,
"step": 12
},
{
"epoch": 0.02011605415860735,
"grad_norm": 6.274352954312508,
"learning_rate": 6e-06,
"loss": 0.461,
"step": 13
},
{
"epoch": 0.021663442940038684,
"grad_norm": 7.414500096339422,
"learning_rate": 6.5000000000000004e-06,
"loss": 0.4408,
"step": 14
},
{
"epoch": 0.02321083172147002,
"grad_norm": 6.819758734571542,
"learning_rate": 7e-06,
"loss": 0.4902,
"step": 15
},
{
"epoch": 0.024758220502901353,
"grad_norm": 4.873525144794793,
"learning_rate": 7.500000000000001e-06,
"loss": 0.3951,
"step": 16
},
{
"epoch": 0.02630560928433269,
"grad_norm": 4.648419115607711,
"learning_rate": 8.000000000000001e-06,
"loss": 0.4187,
"step": 17
},
{
"epoch": 0.027852998065764023,
"grad_norm": 3.5982002846678527,
"learning_rate": 8.5e-06,
"loss": 0.3953,
"step": 18
},
{
"epoch": 0.029400386847195356,
"grad_norm": 1.9116671931337288,
"learning_rate": 9e-06,
"loss": 0.3502,
"step": 19
},
{
"epoch": 0.030947775628626693,
"grad_norm": 2.704858353034554,
"learning_rate": 9.5e-06,
"loss": 0.3502,
"step": 20
},
{
"epoch": 0.03249516441005803,
"grad_norm": 2.2159629781176995,
"learning_rate": 1e-05,
"loss": 0.3828,
"step": 21
},
{
"epoch": 0.03404255319148936,
"grad_norm": 2.674721302533483,
"learning_rate": 9.99993723698994e-06,
"loss": 0.3298,
"step": 22
},
{
"epoch": 0.035589941972920695,
"grad_norm": 2.2698866092687107,
"learning_rate": 9.999748949535436e-06,
"loss": 0.2806,
"step": 23
},
{
"epoch": 0.03713733075435203,
"grad_norm": 2.364895456034976,
"learning_rate": 9.999435142363484e-06,
"loss": 0.3201,
"step": 24
},
{
"epoch": 0.03868471953578337,
"grad_norm": 1.8116874812869403,
"learning_rate": 9.998995823352276e-06,
"loss": 0.3467,
"step": 25
},
{
"epoch": 0.0402321083172147,
"grad_norm": 2.067543594888593,
"learning_rate": 9.998431003531008e-06,
"loss": 0.2799,
"step": 26
},
{
"epoch": 0.041779497098646035,
"grad_norm": 1.2875919118387162,
"learning_rate": 9.997740697079595e-06,
"loss": 0.3158,
"step": 27
},
{
"epoch": 0.04332688588007737,
"grad_norm": 1.6909241814863707,
"learning_rate": 9.99692492132832e-06,
"loss": 0.2397,
"step": 28
},
{
"epoch": 0.0448742746615087,
"grad_norm": 1.3701719631567693,
"learning_rate": 9.9959836967574e-06,
"loss": 0.2615,
"step": 29
},
{
"epoch": 0.04642166344294004,
"grad_norm": 1.7820094163147548,
"learning_rate": 9.994917046996472e-06,
"loss": 0.2663,
"step": 30
},
{
"epoch": 0.047969052224371374,
"grad_norm": 5.663585353915654,
"learning_rate": 9.993724998823995e-06,
"loss": 0.2898,
"step": 31
},
{
"epoch": 0.04951644100580271,
"grad_norm": 1.4447529595529258,
"learning_rate": 9.992407582166582e-06,
"loss": 0.2887,
"step": 32
},
{
"epoch": 0.05106382978723404,
"grad_norm": 1.5547924943577618,
"learning_rate": 9.990964830098246e-06,
"loss": 0.2725,
"step": 33
},
{
"epoch": 0.05261121856866538,
"grad_norm": 1.547863148844634,
"learning_rate": 9.989396778839572e-06,
"loss": 0.251,
"step": 34
},
{
"epoch": 0.05415860735009671,
"grad_norm": 1.8364801342731967,
"learning_rate": 9.987703467756807e-06,
"loss": 0.2303,
"step": 35
},
{
"epoch": 0.055705996131528046,
"grad_norm": 1.8698710149140938,
"learning_rate": 9.985884939360873e-06,
"loss": 0.2381,
"step": 36
},
{
"epoch": 0.05725338491295938,
"grad_norm": 1.2845693723619616,
"learning_rate": 9.983941239306291e-06,
"loss": 0.2314,
"step": 37
},
{
"epoch": 0.05880077369439071,
"grad_norm": 1.6871490766801311,
"learning_rate": 9.981872416390055e-06,
"loss": 0.2847,
"step": 38
},
{
"epoch": 0.06034816247582205,
"grad_norm": 1.5154953285501183,
"learning_rate": 9.979678522550382e-06,
"loss": 0.2558,
"step": 39
},
{
"epoch": 0.061895551257253385,
"grad_norm": 1.6181723977765907,
"learning_rate": 9.977359612865424e-06,
"loss": 0.2428,
"step": 40
},
{
"epoch": 0.06344294003868473,
"grad_norm": 1.6265629334582685,
"learning_rate": 9.974915745551882e-06,
"loss": 0.2496,
"step": 41
},
{
"epoch": 0.06499032882011606,
"grad_norm": 1.446603632464087,
"learning_rate": 9.972346981963546e-06,
"loss": 0.2614,
"step": 42
},
{
"epoch": 0.06653771760154739,
"grad_norm": 1.1076338895078817,
"learning_rate": 9.969653386589749e-06,
"loss": 0.2626,
"step": 43
},
{
"epoch": 0.06808510638297872,
"grad_norm": 1.371833672497246,
"learning_rate": 9.96683502705375e-06,
"loss": 0.246,
"step": 44
},
{
"epoch": 0.06963249516441006,
"grad_norm": 1.3172427974296859,
"learning_rate": 9.963891974111042e-06,
"loss": 0.2102,
"step": 45
},
{
"epoch": 0.07117988394584139,
"grad_norm": 1.1988441635411498,
"learning_rate": 9.960824301647569e-06,
"loss": 0.2277,
"step": 46
},
{
"epoch": 0.07272727272727272,
"grad_norm": 1.752891485082651,
"learning_rate": 9.957632086677876e-06,
"loss": 0.201,
"step": 47
},
{
"epoch": 0.07427466150870406,
"grad_norm": 1.6144063855180706,
"learning_rate": 9.95431540934317e-06,
"loss": 0.2211,
"step": 48
},
{
"epoch": 0.07582205029013539,
"grad_norm": 1.117886829374166,
"learning_rate": 9.95087435290931e-06,
"loss": 0.2085,
"step": 49
},
{
"epoch": 0.07736943907156674,
"grad_norm": 1.2895506502671774,
"learning_rate": 9.947309003764723e-06,
"loss": 0.2172,
"step": 50
},
{
"epoch": 0.07891682785299807,
"grad_norm": 1.7487568111431333,
"learning_rate": 9.943619451418225e-06,
"loss": 0.1968,
"step": 51
},
{
"epoch": 0.0804642166344294,
"grad_norm": 0.7781493982811485,
"learning_rate": 9.939805788496778e-06,
"loss": 0.1953,
"step": 52
},
{
"epoch": 0.08201160541586074,
"grad_norm": 1.2753046827549794,
"learning_rate": 9.935868110743175e-06,
"loss": 0.2156,
"step": 53
},
{
"epoch": 0.08355899419729207,
"grad_norm": 0.995745449819908,
"learning_rate": 9.931806517013612e-06,
"loss": 0.1905,
"step": 54
},
{
"epoch": 0.0851063829787234,
"grad_norm": 1.2720610740359681,
"learning_rate": 9.927621109275233e-06,
"loss": 0.2127,
"step": 55
},
{
"epoch": 0.08665377176015474,
"grad_norm": 0.842247207885935,
"learning_rate": 9.92331199260355e-06,
"loss": 0.2103,
"step": 56
},
{
"epoch": 0.08820116054158607,
"grad_norm": 1.0810541803138813,
"learning_rate": 9.918879275179819e-06,
"loss": 0.2349,
"step": 57
},
{
"epoch": 0.0897485493230174,
"grad_norm": 0.9271008956864596,
"learning_rate": 9.914323068288312e-06,
"loss": 0.1973,
"step": 58
},
{
"epoch": 0.09129593810444875,
"grad_norm": 0.9558303455636472,
"learning_rate": 9.909643486313533e-06,
"loss": 0.2476,
"step": 59
},
{
"epoch": 0.09284332688588008,
"grad_norm": 1.0320370271353754,
"learning_rate": 9.904840646737346e-06,
"loss": 0.2221,
"step": 60
},
{
"epoch": 0.09439071566731141,
"grad_norm": 1.1887775178296134,
"learning_rate": 9.899914670136016e-06,
"loss": 0.2108,
"step": 61
},
{
"epoch": 0.09593810444874275,
"grad_norm": 0.7746057705316729,
"learning_rate": 9.89486568017719e-06,
"loss": 0.1906,
"step": 62
},
{
"epoch": 0.09748549323017408,
"grad_norm": 1.129863752902623,
"learning_rate": 9.889693803616793e-06,
"loss": 0.1898,
"step": 63
},
{
"epoch": 0.09903288201160541,
"grad_norm": 0.8376196906837178,
"learning_rate": 9.884399170295839e-06,
"loss": 0.1822,
"step": 64
},
{
"epoch": 0.10058027079303675,
"grad_norm": 1.3975259688349446,
"learning_rate": 9.878981913137178e-06,
"loss": 0.2552,
"step": 65
},
{
"epoch": 0.10212765957446808,
"grad_norm": 1.7201831210313148,
"learning_rate": 9.873442168142158e-06,
"loss": 0.2075,
"step": 66
},
{
"epoch": 0.10367504835589941,
"grad_norm": 0.841583876550117,
"learning_rate": 9.867780074387207e-06,
"loss": 0.2274,
"step": 67
},
{
"epoch": 0.10522243713733076,
"grad_norm": 1.2094935283898536,
"learning_rate": 9.861995774020341e-06,
"loss": 0.1956,
"step": 68
},
{
"epoch": 0.10676982591876209,
"grad_norm": 0.781784377187281,
"learning_rate": 9.856089412257605e-06,
"loss": 0.2018,
"step": 69
},
{
"epoch": 0.10831721470019343,
"grad_norm": 1.1053353625516136,
"learning_rate": 9.850061137379414e-06,
"loss": 0.237,
"step": 70
},
{
"epoch": 0.10986460348162476,
"grad_norm": 1.1027104376275203,
"learning_rate": 9.843911100726838e-06,
"loss": 0.2114,
"step": 71
},
{
"epoch": 0.11141199226305609,
"grad_norm": 1.1237166681861188,
"learning_rate": 9.837639456697802e-06,
"loss": 0.2225,
"step": 72
},
{
"epoch": 0.11295938104448743,
"grad_norm": 1.2398399829320181,
"learning_rate": 9.83124636274321e-06,
"loss": 0.2057,
"step": 73
},
{
"epoch": 0.11450676982591876,
"grad_norm": 1.4843985269546451,
"learning_rate": 9.824731979362991e-06,
"loss": 0.2436,
"step": 74
},
{
"epoch": 0.11605415860735009,
"grad_norm": 0.8730215521905568,
"learning_rate": 9.818096470102067e-06,
"loss": 0.1618,
"step": 75
},
{
"epoch": 0.11760154738878142,
"grad_norm": 1.1627999904380357,
"learning_rate": 9.811340001546252e-06,
"loss": 0.1945,
"step": 76
},
{
"epoch": 0.11914893617021277,
"grad_norm": 1.032965345889784,
"learning_rate": 9.80446274331807e-06,
"loss": 0.2204,
"step": 77
},
{
"epoch": 0.1206963249516441,
"grad_norm": 1.1838259078956053,
"learning_rate": 9.797464868072489e-06,
"loss": 0.1985,
"step": 78
},
{
"epoch": 0.12224371373307544,
"grad_norm": 0.8572154618594826,
"learning_rate": 9.790346551492594e-06,
"loss": 0.2175,
"step": 79
},
{
"epoch": 0.12379110251450677,
"grad_norm": 0.9582739403464648,
"learning_rate": 9.783107972285177e-06,
"loss": 0.2004,
"step": 80
},
{
"epoch": 0.1253384912959381,
"grad_norm": 1.081492590353908,
"learning_rate": 9.775749312176249e-06,
"loss": 0.1975,
"step": 81
},
{
"epoch": 0.12688588007736945,
"grad_norm": 1.3313053559370254,
"learning_rate": 9.768270755906467e-06,
"loss": 0.2133,
"step": 82
},
{
"epoch": 0.12843326885880077,
"grad_norm": 1.513601346491332,
"learning_rate": 9.760672491226515e-06,
"loss": 0.2041,
"step": 83
},
{
"epoch": 0.12998065764023212,
"grad_norm": 0.8016201580505663,
"learning_rate": 9.752954708892379e-06,
"loss": 0.1972,
"step": 84
},
{
"epoch": 0.13152804642166344,
"grad_norm": 0.9746092202244274,
"learning_rate": 9.745117602660556e-06,
"loss": 0.1727,
"step": 85
},
{
"epoch": 0.13307543520309478,
"grad_norm": 1.3568864132273577,
"learning_rate": 9.737161369283201e-06,
"loss": 0.2671,
"step": 86
},
{
"epoch": 0.1346228239845261,
"grad_norm": 1.4138857330734356,
"learning_rate": 9.729086208503174e-06,
"loss": 0.2082,
"step": 87
},
{
"epoch": 0.13617021276595745,
"grad_norm": 0.9738747745181121,
"learning_rate": 9.720892323049034e-06,
"loss": 0.213,
"step": 88
},
{
"epoch": 0.13771760154738877,
"grad_norm": 0.8499524074169559,
"learning_rate": 9.712579918629947e-06,
"loss": 0.1767,
"step": 89
},
{
"epoch": 0.13926499032882012,
"grad_norm": 1.0177288372378337,
"learning_rate": 9.704149203930522e-06,
"loss": 0.2673,
"step": 90
},
{
"epoch": 0.14081237911025146,
"grad_norm": 1.0104101681137083,
"learning_rate": 9.695600390605573e-06,
"loss": 0.1642,
"step": 91
},
{
"epoch": 0.14235976789168278,
"grad_norm": 0.8682686361932304,
"learning_rate": 9.686933693274801e-06,
"loss": 0.1958,
"step": 92
},
{
"epoch": 0.14390715667311413,
"grad_norm": 0.959096415068929,
"learning_rate": 9.67814932951741e-06,
"loss": 0.163,
"step": 93
},
{
"epoch": 0.14545454545454545,
"grad_norm": 0.9704441402003741,
"learning_rate": 9.669247519866645e-06,
"loss": 0.1892,
"step": 94
},
{
"epoch": 0.1470019342359768,
"grad_norm": 1.1085772957530802,
"learning_rate": 9.660228487804254e-06,
"loss": 0.2134,
"step": 95
},
{
"epoch": 0.1485493230174081,
"grad_norm": 0.9255984965037967,
"learning_rate": 9.651092459754879e-06,
"loss": 0.2036,
"step": 96
},
{
"epoch": 0.15009671179883946,
"grad_norm": 0.9049968395048311,
"learning_rate": 9.641839665080363e-06,
"loss": 0.217,
"step": 97
},
{
"epoch": 0.15164410058027078,
"grad_norm": 0.9155502433988073,
"learning_rate": 9.632470336074009e-06,
"loss": 0.177,
"step": 98
},
{
"epoch": 0.15319148936170213,
"grad_norm": 1.0453907072873718,
"learning_rate": 9.622984707954732e-06,
"loss": 0.2027,
"step": 99
},
{
"epoch": 0.15473887814313347,
"grad_norm": 1.1095752696396315,
"learning_rate": 9.613383018861159e-06,
"loss": 0.2167,
"step": 100
},
{
"epoch": 0.1562862669245648,
"grad_norm": 0.9831846037070405,
"learning_rate": 9.603665509845657e-06,
"loss": 0.2083,
"step": 101
},
{
"epoch": 0.15783365570599614,
"grad_norm": 1.4258782601946127,
"learning_rate": 9.593832424868271e-06,
"loss": 0.2099,
"step": 102
},
{
"epoch": 0.15938104448742746,
"grad_norm": 1.2849221334696659,
"learning_rate": 9.583884010790605e-06,
"loss": 0.1777,
"step": 103
},
{
"epoch": 0.1609284332688588,
"grad_norm": 1.328819730359831,
"learning_rate": 9.573820517369623e-06,
"loss": 0.1912,
"step": 104
},
{
"epoch": 0.16247582205029013,
"grad_norm": 1.6050933066194693,
"learning_rate": 9.563642197251382e-06,
"loss": 0.1872,
"step": 105
},
{
"epoch": 0.16402321083172147,
"grad_norm": 0.8632160650062739,
"learning_rate": 9.553349305964687e-06,
"loss": 0.1913,
"step": 106
},
{
"epoch": 0.1655705996131528,
"grad_norm": 0.7667887415321157,
"learning_rate": 9.54294210191467e-06,
"loss": 0.2398,
"step": 107
},
{
"epoch": 0.16711798839458414,
"grad_norm": 0.922001722791926,
"learning_rate": 9.532420846376316e-06,
"loss": 0.1885,
"step": 108
},
{
"epoch": 0.16866537717601549,
"grad_norm": 0.9848963142378488,
"learning_rate": 9.521785803487888e-06,
"loss": 0.2179,
"step": 109
},
{
"epoch": 0.1702127659574468,
"grad_norm": 0.9350329048939854,
"learning_rate": 9.51103724024431e-06,
"loss": 0.2155,
"step": 110
},
{
"epoch": 0.17176015473887815,
"grad_norm": 0.9435224586318846,
"learning_rate": 9.500175426490455e-06,
"loss": 0.1539,
"step": 111
},
{
"epoch": 0.17330754352030947,
"grad_norm": 0.9426736571648074,
"learning_rate": 9.489200634914373e-06,
"loss": 0.1976,
"step": 112
},
{
"epoch": 0.17485493230174082,
"grad_norm": 0.9405655941950657,
"learning_rate": 9.478113141040444e-06,
"loss": 0.1728,
"step": 113
},
{
"epoch": 0.17640232108317214,
"grad_norm": 1.0534108505192408,
"learning_rate": 9.466913223222467e-06,
"loss": 0.1953,
"step": 114
},
{
"epoch": 0.17794970986460348,
"grad_norm": 0.9649711851467941,
"learning_rate": 9.455601162636662e-06,
"loss": 0.1815,
"step": 115
},
{
"epoch": 0.1794970986460348,
"grad_norm": 1.2215842119559421,
"learning_rate": 9.444177243274619e-06,
"loss": 0.1859,
"step": 116
},
{
"epoch": 0.18104448742746615,
"grad_norm": 1.3992576608547573,
"learning_rate": 9.432641751936162e-06,
"loss": 0.2031,
"step": 117
},
{
"epoch": 0.1825918762088975,
"grad_norm": 1.0520365385958643,
"learning_rate": 9.420994978222156e-06,
"loss": 0.155,
"step": 118
},
{
"epoch": 0.18413926499032882,
"grad_norm": 1.0453248711004668,
"learning_rate": 9.40923721452723e-06,
"loss": 0.227,
"step": 119
},
{
"epoch": 0.18568665377176016,
"grad_norm": 1.018817182864082,
"learning_rate": 9.397368756032445e-06,
"loss": 0.1712,
"step": 120
},
{
"epoch": 0.18723404255319148,
"grad_norm": 1.0449294860721698,
"learning_rate": 9.38538990069787e-06,
"loss": 0.1941,
"step": 121
},
{
"epoch": 0.18878143133462283,
"grad_norm": 0.8820467270290865,
"learning_rate": 9.373300949255112e-06,
"loss": 0.1889,
"step": 122
},
{
"epoch": 0.19032882011605415,
"grad_norm": 1.1783314790561306,
"learning_rate": 9.361102205199762e-06,
"loss": 0.2328,
"step": 123
},
{
"epoch": 0.1918762088974855,
"grad_norm": 1.1945837782832185,
"learning_rate": 9.348793974783778e-06,
"loss": 0.1443,
"step": 124
},
{
"epoch": 0.19342359767891681,
"grad_norm": 0.8626079057082615,
"learning_rate": 9.336376567007799e-06,
"loss": 0.1939,
"step": 125
},
{
"epoch": 0.19497098646034816,
"grad_norm": 0.8611436156902041,
"learning_rate": 9.32385029361338e-06,
"loss": 0.2019,
"step": 126
},
{
"epoch": 0.1965183752417795,
"grad_norm": 0.8219959901195926,
"learning_rate": 9.311215469075168e-06,
"loss": 0.2014,
"step": 127
},
{
"epoch": 0.19806576402321083,
"grad_norm": 0.6347389394415482,
"learning_rate": 9.298472410593013e-06,
"loss": 0.1317,
"step": 128
},
{
"epoch": 0.19961315280464217,
"grad_norm": 1.0472959655840977,
"learning_rate": 9.285621438083997e-06,
"loss": 0.1721,
"step": 129
},
{
"epoch": 0.2011605415860735,
"grad_norm": 0.8508709558715976,
"learning_rate": 9.27266287417441e-06,
"loss": 0.1815,
"step": 130
},
{
"epoch": 0.20270793036750484,
"grad_norm": 0.7373523954349206,
"learning_rate": 9.259597044191635e-06,
"loss": 0.1836,
"step": 131
},
{
"epoch": 0.20425531914893616,
"grad_norm": 0.9285350577082694,
"learning_rate": 9.246424276156008e-06,
"loss": 0.1618,
"step": 132
},
{
"epoch": 0.2058027079303675,
"grad_norm": 0.8418651244059254,
"learning_rate": 9.233144900772553e-06,
"loss": 0.1916,
"step": 133
},
{
"epoch": 0.20735009671179883,
"grad_norm": 0.8967542264239586,
"learning_rate": 9.2197592514227e-06,
"loss": 0.2045,
"step": 134
},
{
"epoch": 0.20889748549323017,
"grad_norm": 1.6506436445366963,
"learning_rate": 9.206267664155906e-06,
"loss": 0.1647,
"step": 135
},
{
"epoch": 0.21044487427466152,
"grad_norm": 0.6874666232905765,
"learning_rate": 9.192670477681224e-06,
"loss": 0.1478,
"step": 136
},
{
"epoch": 0.21199226305609284,
"grad_norm": 0.9337036799681375,
"learning_rate": 9.178968033358792e-06,
"loss": 0.1515,
"step": 137
},
{
"epoch": 0.21353965183752419,
"grad_norm": 0.8886617018672116,
"learning_rate": 9.165160675191272e-06,
"loss": 0.1937,
"step": 138
},
{
"epoch": 0.2150870406189555,
"grad_norm": 1.0587242408040172,
"learning_rate": 9.151248749815208e-06,
"loss": 0.2057,
"step": 139
},
{
"epoch": 0.21663442940038685,
"grad_norm": 1.106081995090353,
"learning_rate": 9.137232606492323e-06,
"loss": 0.1558,
"step": 140
},
{
"epoch": 0.21818181818181817,
"grad_norm": 0.7796654056066916,
"learning_rate": 9.123112597100759e-06,
"loss": 0.1624,
"step": 141
},
{
"epoch": 0.21972920696324952,
"grad_norm": 1.3247843376444552,
"learning_rate": 9.108889076126226e-06,
"loss": 0.1668,
"step": 142
},
{
"epoch": 0.22127659574468084,
"grad_norm": 0.7715774094433246,
"learning_rate": 9.094562400653127e-06,
"loss": 0.1662,
"step": 143
},
{
"epoch": 0.22282398452611218,
"grad_norm": 0.8803366064268053,
"learning_rate": 9.080132930355567e-06,
"loss": 0.1945,
"step": 144
},
{
"epoch": 0.22437137330754353,
"grad_norm": 1.0060879241216365,
"learning_rate": 9.065601027488345e-06,
"loss": 0.1856,
"step": 145
},
{
"epoch": 0.22591876208897485,
"grad_norm": 0.66449734867189,
"learning_rate": 9.050967056877846e-06,
"loss": 0.2047,
"step": 146
},
{
"epoch": 0.2274661508704062,
"grad_norm": 0.973280505219451,
"learning_rate": 9.03623138591289e-06,
"loss": 0.184,
"step": 147
},
{
"epoch": 0.22901353965183752,
"grad_norm": 0.7818051435660924,
"learning_rate": 9.0213943845355e-06,
"loss": 0.1652,
"step": 148
},
{
"epoch": 0.23056092843326886,
"grad_norm": 0.7845320452462246,
"learning_rate": 9.006456425231624e-06,
"loss": 0.1905,
"step": 149
},
{
"epoch": 0.23210831721470018,
"grad_norm": 0.8131952366183569,
"learning_rate": 8.99141788302178e-06,
"loss": 0.1915,
"step": 150
},
{
"epoch": 0.23365570599613153,
"grad_norm": 0.6953878285016128,
"learning_rate": 8.976279135451636e-06,
"loss": 0.1887,
"step": 151
},
{
"epoch": 0.23520309477756285,
"grad_norm": 0.8424455540525856,
"learning_rate": 8.96104056258254e-06,
"loss": 0.1282,
"step": 152
},
{
"epoch": 0.2367504835589942,
"grad_norm": 0.7852661949312153,
"learning_rate": 8.94570254698197e-06,
"loss": 0.2115,
"step": 153
},
{
"epoch": 0.23829787234042554,
"grad_norm": 1.016500925991813,
"learning_rate": 8.930265473713939e-06,
"loss": 0.1918,
"step": 154
},
{
"epoch": 0.23984526112185686,
"grad_norm": 0.8030084084133585,
"learning_rate": 8.914729730329321e-06,
"loss": 0.1617,
"step": 155
},
{
"epoch": 0.2413926499032882,
"grad_norm": 0.7980246350581464,
"learning_rate": 8.899095706856122e-06,
"loss": 0.1888,
"step": 156
},
{
"epoch": 0.24294003868471953,
"grad_norm": 0.7831670393329384,
"learning_rate": 8.883363795789694e-06,
"loss": 0.1851,
"step": 157
},
{
"epoch": 0.24448742746615088,
"grad_norm": 0.777025773689329,
"learning_rate": 8.867534392082873e-06,
"loss": 0.1573,
"step": 158
},
{
"epoch": 0.2460348162475822,
"grad_norm": 0.6073947896564662,
"learning_rate": 8.851607893136065e-06,
"loss": 0.1413,
"step": 159
},
{
"epoch": 0.24758220502901354,
"grad_norm": 1.3452407987574415,
"learning_rate": 8.83558469878728e-06,
"loss": 0.1792,
"step": 160
},
{
"epoch": 0.24912959381044486,
"grad_norm": 0.8181328001089564,
"learning_rate": 8.819465211302081e-06,
"loss": 0.1996,
"step": 161
},
{
"epoch": 0.2506769825918762,
"grad_norm": 1.1608290499183214,
"learning_rate": 8.803249835363486e-06,
"loss": 0.1953,
"step": 162
},
{
"epoch": 0.25222437137330755,
"grad_norm": 1.060552305699874,
"learning_rate": 8.78693897806182e-06,
"loss": 0.1744,
"step": 163
},
{
"epoch": 0.2537717601547389,
"grad_norm": 0.7721189296170582,
"learning_rate": 8.770533048884483e-06,
"loss": 0.1568,
"step": 164
},
{
"epoch": 0.2553191489361702,
"grad_norm": 0.6862315373630368,
"learning_rate": 8.754032459705672e-06,
"loss": 0.2122,
"step": 165
},
{
"epoch": 0.25686653771760154,
"grad_norm": 0.6970596714469318,
"learning_rate": 8.737437624776047e-06,
"loss": 0.1833,
"step": 166
},
{
"epoch": 0.2584139264990329,
"grad_norm": 0.8901186698883564,
"learning_rate": 8.720748960712323e-06,
"loss": 0.1944,
"step": 167
},
{
"epoch": 0.25996131528046423,
"grad_norm": 0.8726232374713512,
"learning_rate": 8.703966886486819e-06,
"loss": 0.1913,
"step": 168
},
{
"epoch": 0.2615087040618955,
"grad_norm": 0.6703400145261608,
"learning_rate": 8.68709182341693e-06,
"loss": 0.1594,
"step": 169
},
{
"epoch": 0.26305609284332687,
"grad_norm": 0.8844709300145817,
"learning_rate": 8.670124195154557e-06,
"loss": 0.1246,
"step": 170
},
{
"epoch": 0.2646034816247582,
"grad_norm": 0.708152671226741,
"learning_rate": 8.65306442767547e-06,
"loss": 0.1256,
"step": 171
},
{
"epoch": 0.26615087040618957,
"grad_norm": 0.9082094549412949,
"learning_rate": 8.635912949268614e-06,
"loss": 0.1611,
"step": 172
},
{
"epoch": 0.2676982591876209,
"grad_norm": 0.8261142492070601,
"learning_rate": 8.61867019052535e-06,
"loss": 0.2151,
"step": 173
},
{
"epoch": 0.2692456479690522,
"grad_norm": 1.0343276774533101,
"learning_rate": 8.601336584328659e-06,
"loss": 0.1903,
"step": 174
},
{
"epoch": 0.27079303675048355,
"grad_norm": 0.7780585533808416,
"learning_rate": 8.583912565842258e-06,
"loss": 0.1885,
"step": 175
},
{
"epoch": 0.2723404255319149,
"grad_norm": 0.684577901089209,
"learning_rate": 8.566398572499685e-06,
"loss": 0.1328,
"step": 176
},
{
"epoch": 0.27388781431334625,
"grad_norm": 0.7375341242885513,
"learning_rate": 8.548795043993316e-06,
"loss": 0.1481,
"step": 177
},
{
"epoch": 0.27543520309477754,
"grad_norm": 0.8961993551686324,
"learning_rate": 8.53110242226333e-06,
"loss": 0.1995,
"step": 178
},
{
"epoch": 0.2769825918762089,
"grad_norm": 0.8549735998980131,
"learning_rate": 8.513321151486602e-06,
"loss": 0.1902,
"step": 179
},
{
"epoch": 0.27852998065764023,
"grad_norm": 0.9355308088893293,
"learning_rate": 8.495451678065563e-06,
"loss": 0.1761,
"step": 180
},
{
"epoch": 0.2800773694390716,
"grad_norm": 0.7103239247157965,
"learning_rate": 8.477494450616988e-06,
"loss": 0.139,
"step": 181
},
{
"epoch": 0.2816247582205029,
"grad_norm": 0.7131678297476808,
"learning_rate": 8.459449919960737e-06,
"loss": 0.1661,
"step": 182
},
{
"epoch": 0.2831721470019342,
"grad_norm": 0.8161337853451635,
"learning_rate": 8.441318539108433e-06,
"loss": 0.1505,
"step": 183
},
{
"epoch": 0.28471953578336556,
"grad_norm": 0.6468618760587732,
"learning_rate": 8.423100763252094e-06,
"loss": 0.1911,
"step": 184
},
{
"epoch": 0.2862669245647969,
"grad_norm": 0.8001473630712188,
"learning_rate": 8.404797049752697e-06,
"loss": 0.2149,
"step": 185
},
{
"epoch": 0.28781431334622826,
"grad_norm": 0.778484481129197,
"learning_rate": 8.386407858128707e-06,
"loss": 0.1714,
"step": 186
},
{
"epoch": 0.28936170212765955,
"grad_norm": 0.890957740854113,
"learning_rate": 8.367933650044526e-06,
"loss": 0.182,
"step": 187
},
{
"epoch": 0.2909090909090909,
"grad_norm": 0.7201256698091962,
"learning_rate": 8.349374889298923e-06,
"loss": 0.1752,
"step": 188
},
{
"epoch": 0.29245647969052224,
"grad_norm": 0.9440325344267313,
"learning_rate": 8.330732041813367e-06,
"loss": 0.1672,
"step": 189
},
{
"epoch": 0.2940038684719536,
"grad_norm": 0.7496260491615918,
"learning_rate": 8.312005575620355e-06,
"loss": 0.1461,
"step": 190
},
{
"epoch": 0.29555125725338494,
"grad_norm": 0.7885853714287803,
"learning_rate": 8.293195960851634e-06,
"loss": 0.1818,
"step": 191
},
{
"epoch": 0.2970986460348162,
"grad_norm": 0.8909219374421637,
"learning_rate": 8.274303669726427e-06,
"loss": 0.1913,
"step": 192
},
{
"epoch": 0.2986460348162476,
"grad_norm": 0.9855581917407712,
"learning_rate": 8.255329176539552e-06,
"loss": 0.1627,
"step": 193
},
{
"epoch": 0.3001934235976789,
"grad_norm": 1.0725253670421797,
"learning_rate": 8.236272957649534e-06,
"loss": 0.167,
"step": 194
},
{
"epoch": 0.30174081237911027,
"grad_norm": 0.8153778456841704,
"learning_rate": 8.217135491466636e-06,
"loss": 0.1554,
"step": 195
},
{
"epoch": 0.30328820116054156,
"grad_norm": 0.8015423413432219,
"learning_rate": 8.197917258440851e-06,
"loss": 0.1821,
"step": 196
},
{
"epoch": 0.3048355899419729,
"grad_norm": 0.6191883824483493,
"learning_rate": 8.178618741049841e-06,
"loss": 0.1552,
"step": 197
},
{
"epoch": 0.30638297872340425,
"grad_norm": 0.7382273640269873,
"learning_rate": 8.15924042378682e-06,
"loss": 0.1629,
"step": 198
},
{
"epoch": 0.3079303675048356,
"grad_norm": 0.7618444883217768,
"learning_rate": 8.1397827931484e-06,
"loss": 0.1423,
"step": 199
},
{
"epoch": 0.30947775628626695,
"grad_norm": 0.7156989852439967,
"learning_rate": 8.120246337622364e-06,
"loss": 0.1829,
"step": 200
},
{
"epoch": 0.31102514506769824,
"grad_norm": 0.8195126568212944,
"learning_rate": 8.100631547675417e-06,
"loss": 0.1484,
"step": 201
},
{
"epoch": 0.3125725338491296,
"grad_norm": 0.839724031343586,
"learning_rate": 8.080938915740863e-06,
"loss": 0.1941,
"step": 202
},
{
"epoch": 0.31411992263056093,
"grad_norm": 0.9497393034306165,
"learning_rate": 8.06116893620624e-06,
"loss": 0.1818,
"step": 203
},
{
"epoch": 0.3156673114119923,
"grad_norm": 0.8645182321256714,
"learning_rate": 8.041322105400923e-06,
"loss": 0.178,
"step": 204
},
{
"epoch": 0.31721470019342357,
"grad_norm": 0.8466865162848244,
"learning_rate": 8.021398921583644e-06,
"loss": 0.2248,
"step": 205
},
{
"epoch": 0.3187620889748549,
"grad_norm": 0.7719432995418056,
"learning_rate": 8.001399884930004e-06,
"loss": 0.1839,
"step": 206
},
{
"epoch": 0.32030947775628626,
"grad_norm": 0.7039493901127064,
"learning_rate": 7.981325497519892e-06,
"loss": 0.1786,
"step": 207
},
{
"epoch": 0.3218568665377176,
"grad_norm": 0.7789735372485523,
"learning_rate": 7.961176263324902e-06,
"loss": 0.1665,
"step": 208
},
{
"epoch": 0.32340425531914896,
"grad_norm": 0.7680545055417434,
"learning_rate": 7.940952688195668e-06,
"loss": 0.1319,
"step": 209
},
{
"epoch": 0.32495164410058025,
"grad_norm": 0.749256619294284,
"learning_rate": 7.920655279849173e-06,
"loss": 0.131,
"step": 210
},
{
"epoch": 0.3264990328820116,
"grad_norm": 0.8640545344416056,
"learning_rate": 7.900284547855992e-06,
"loss": 0.1709,
"step": 211
},
{
"epoch": 0.32804642166344294,
"grad_norm": 0.849944924888024,
"learning_rate": 7.87984100362751e-06,
"loss": 0.1973,
"step": 212
},
{
"epoch": 0.3295938104448743,
"grad_norm": 0.8032999098290784,
"learning_rate": 7.859325160403073e-06,
"loss": 0.1501,
"step": 213
},
{
"epoch": 0.3311411992263056,
"grad_norm": 0.9333424894473424,
"learning_rate": 7.838737533237111e-06,
"loss": 0.138,
"step": 214
},
{
"epoch": 0.33268858800773693,
"grad_norm": 0.9380242982321761,
"learning_rate": 7.818078638986208e-06,
"loss": 0.1824,
"step": 215
},
{
"epoch": 0.3342359767891683,
"grad_norm": 0.8115574307282266,
"learning_rate": 7.797348996296116e-06,
"loss": 0.1726,
"step": 216
},
{
"epoch": 0.3357833655705996,
"grad_norm": 0.8589980134024339,
"learning_rate": 7.776549125588743e-06,
"loss": 0.1401,
"step": 217
},
{
"epoch": 0.33733075435203097,
"grad_norm": 0.7669008115392018,
"learning_rate": 7.755679549049093e-06,
"loss": 0.16,
"step": 218
},
{
"epoch": 0.33887814313346226,
"grad_norm": 0.8630743320241145,
"learning_rate": 7.734740790612137e-06,
"loss": 0.138,
"step": 219
},
{
"epoch": 0.3404255319148936,
"grad_norm": 0.6207975833777728,
"learning_rate": 7.713733375949677e-06,
"loss": 0.1276,
"step": 220
},
{
"epoch": 0.34197292069632496,
"grad_norm": 0.7458919505402387,
"learning_rate": 7.692657832457146e-06,
"loss": 0.1879,
"step": 221
},
{
"epoch": 0.3435203094777563,
"grad_norm": 0.921666373012695,
"learning_rate": 7.671514689240366e-06,
"loss": 0.1948,
"step": 222
},
{
"epoch": 0.3450676982591876,
"grad_norm": 0.9954541136532746,
"learning_rate": 7.650304477102258e-06,
"loss": 0.1804,
"step": 223
},
{
"epoch": 0.34661508704061894,
"grad_norm": 0.7495588145719608,
"learning_rate": 7.629027728529527e-06,
"loss": 0.1094,
"step": 224
},
{
"epoch": 0.3481624758220503,
"grad_norm": 0.7369534761553445,
"learning_rate": 7.607684977679284e-06,
"loss": 0.1578,
"step": 225
},
{
"epoch": 0.34970986460348163,
"grad_norm": 0.9144728168098029,
"learning_rate": 7.586276760365645e-06,
"loss": 0.1248,
"step": 226
},
{
"epoch": 0.351257253384913,
"grad_norm": 1.0021636468489092,
"learning_rate": 7.564803614046276e-06,
"loss": 0.1609,
"step": 227
},
{
"epoch": 0.3528046421663443,
"grad_norm": 0.7565105542706337,
"learning_rate": 7.543266077808893e-06,
"loss": 0.1692,
"step": 228
},
{
"epoch": 0.3543520309477756,
"grad_norm": 0.9760289598092642,
"learning_rate": 7.521664692357737e-06,
"loss": 0.148,
"step": 229
},
{
"epoch": 0.35589941972920697,
"grad_norm": 0.8906470147517871,
"learning_rate": 7.500000000000001e-06,
"loss": 0.1973,
"step": 230
},
{
"epoch": 0.3574468085106383,
"grad_norm": 0.9276117415143279,
"learning_rate": 7.478272544632204e-06,
"loss": 0.1851,
"step": 231
},
{
"epoch": 0.3589941972920696,
"grad_norm": 0.8334719409443097,
"learning_rate": 7.456482871726545e-06,
"loss": 0.1597,
"step": 232
},
{
"epoch": 0.36054158607350095,
"grad_norm": 0.7892503202854746,
"learning_rate": 7.434631528317209e-06,
"loss": 0.1964,
"step": 233
},
{
"epoch": 0.3620889748549323,
"grad_norm": 0.7533778358270108,
"learning_rate": 7.412719062986632e-06,
"loss": 0.1803,
"step": 234
},
{
"epoch": 0.36363636363636365,
"grad_norm": 0.7757056005566882,
"learning_rate": 7.390746025851725e-06,
"loss": 0.1549,
"step": 235
},
{
"epoch": 0.365183752417795,
"grad_norm": 1.0315366579407221,
"learning_rate": 7.368712968550068e-06,
"loss": 0.2015,
"step": 236
},
{
"epoch": 0.3667311411992263,
"grad_norm": 0.7809216457984639,
"learning_rate": 7.3466204442260605e-06,
"loss": 0.2134,
"step": 237
},
{
"epoch": 0.36827852998065763,
"grad_norm": 0.9268298135980987,
"learning_rate": 7.324469007517035e-06,
"loss": 0.1715,
"step": 238
},
{
"epoch": 0.369825918762089,
"grad_norm": 0.7902283417730432,
"learning_rate": 7.302259214539327e-06,
"loss": 0.1657,
"step": 239
},
{
"epoch": 0.3713733075435203,
"grad_norm": 0.8931289534535511,
"learning_rate": 7.279991622874319e-06,
"loss": 0.2019,
"step": 240
},
{
"epoch": 0.3729206963249516,
"grad_norm": 1.0526307047246841,
"learning_rate": 7.257666791554448e-06,
"loss": 0.1781,
"step": 241
},
{
"epoch": 0.37446808510638296,
"grad_norm": 0.8614563077864247,
"learning_rate": 7.235285281049154e-06,
"loss": 0.159,
"step": 242
},
{
"epoch": 0.3760154738878143,
"grad_norm": 0.8140225164209623,
"learning_rate": 7.212847653250828e-06,
"loss": 0.1555,
"step": 243
},
{
"epoch": 0.37756286266924566,
"grad_norm": 0.8723856496576525,
"learning_rate": 7.190354471460692e-06,
"loss": 0.2159,
"step": 244
},
{
"epoch": 0.379110251450677,
"grad_norm": 0.7918280949465,
"learning_rate": 7.167806300374665e-06,
"loss": 0.1875,
"step": 245
},
{
"epoch": 0.3806576402321083,
"grad_norm": 0.6360617391742313,
"learning_rate": 7.145203706069183e-06,
"loss": 0.1611,
"step": 246
},
{
"epoch": 0.38220502901353964,
"grad_norm": 0.9383032190875339,
"learning_rate": 7.122547255986985e-06,
"loss": 0.1936,
"step": 247
},
{
"epoch": 0.383752417794971,
"grad_norm": 0.8517530784649264,
"learning_rate": 7.099837518922873e-06,
"loss": 0.1918,
"step": 248
},
{
"epoch": 0.38529980657640234,
"grad_norm": 0.8921062102963762,
"learning_rate": 7.0770750650094335e-06,
"loss": 0.1737,
"step": 249
},
{
"epoch": 0.38684719535783363,
"grad_norm": 0.7572247520005811,
"learning_rate": 7.054260465702712e-06,
"loss": 0.1601,
"step": 250
},
{
"epoch": 0.388394584139265,
"grad_norm": 0.7699199293334452,
"learning_rate": 7.031394293767879e-06,
"loss": 0.1533,
"step": 251
},
{
"epoch": 0.3899419729206963,
"grad_norm": 1.1167232908536833,
"learning_rate": 7.008477123264849e-06,
"loss": 0.2118,
"step": 252
},
{
"epoch": 0.39148936170212767,
"grad_norm": 0.8758224001780723,
"learning_rate": 6.985509529533859e-06,
"loss": 0.1858,
"step": 253
},
{
"epoch": 0.393036750483559,
"grad_norm": 0.922725511094832,
"learning_rate": 6.96249208918104e-06,
"loss": 0.1505,
"step": 254
},
{
"epoch": 0.3945841392649903,
"grad_norm": 0.872891091022535,
"learning_rate": 6.939425380063924e-06,
"loss": 0.1969,
"step": 255
},
{
"epoch": 0.39613152804642165,
"grad_norm": 0.9473736848619609,
"learning_rate": 6.916309981276954e-06,
"loss": 0.1513,
"step": 256
},
{
"epoch": 0.397678916827853,
"grad_norm": 0.7411898138116363,
"learning_rate": 6.89314647313693e-06,
"loss": 0.202,
"step": 257
},
{
"epoch": 0.39922630560928435,
"grad_norm": 1.0374882132168852,
"learning_rate": 6.869935437168449e-06,
"loss": 0.1619,
"step": 258
},
{
"epoch": 0.40077369439071564,
"grad_norm": 0.7717319840912141,
"learning_rate": 6.846677456089305e-06,
"loss": 0.1714,
"step": 259
},
{
"epoch": 0.402321083172147,
"grad_norm": 0.7317657423347314,
"learning_rate": 6.82337311379586e-06,
"loss": 0.2087,
"step": 260
},
{
"epoch": 0.40386847195357833,
"grad_norm": 0.9303667847611171,
"learning_rate": 6.800022995348381e-06,
"loss": 0.1909,
"step": 261
},
{
"epoch": 0.4054158607350097,
"grad_norm": 0.5646674138871769,
"learning_rate": 6.776627686956354e-06,
"loss": 0.1272,
"step": 262
},
{
"epoch": 0.40696324951644103,
"grad_norm": 0.906064507150676,
"learning_rate": 6.753187775963773e-06,
"loss": 0.1251,
"step": 263
},
{
"epoch": 0.4085106382978723,
"grad_norm": 0.9014541696863367,
"learning_rate": 6.729703850834381e-06,
"loss": 0.168,
"step": 264
},
{
"epoch": 0.41005802707930367,
"grad_norm": 0.8692764545196213,
"learning_rate": 6.706176501136914e-06,
"loss": 0.1807,
"step": 265
},
{
"epoch": 0.411605415860735,
"grad_norm": 0.8513956354740446,
"learning_rate": 6.682606317530284e-06,
"loss": 0.1677,
"step": 266
},
{
"epoch": 0.41315280464216636,
"grad_norm": 0.8639883773069676,
"learning_rate": 6.65899389174876e-06,
"loss": 0.1574,
"step": 267
},
{
"epoch": 0.41470019342359765,
"grad_norm": 0.7599185108885862,
"learning_rate": 6.635339816587109e-06,
"loss": 0.1918,
"step": 268
},
{
"epoch": 0.416247582205029,
"grad_norm": 0.7759787013585075,
"learning_rate": 6.611644685885713e-06,
"loss": 0.1684,
"step": 269
},
{
"epoch": 0.41779497098646035,
"grad_norm": 0.7719828708397092,
"learning_rate": 6.587909094515663e-06,
"loss": 0.1523,
"step": 270
},
{
"epoch": 0.4193423597678917,
"grad_norm": 0.5869245573014121,
"learning_rate": 6.564133638363823e-06,
"loss": 0.1474,
"step": 271
},
{
"epoch": 0.42088974854932304,
"grad_norm": 0.7506341900272443,
"learning_rate": 6.5403189143178725e-06,
"loss": 0.1451,
"step": 272
},
{
"epoch": 0.42243713733075433,
"grad_norm": 1.0073931592498175,
"learning_rate": 6.5164655202513135e-06,
"loss": 0.2049,
"step": 273
},
{
"epoch": 0.4239845261121857,
"grad_norm": 0.9025535933925777,
"learning_rate": 6.492574055008474e-06,
"loss": 0.1523,
"step": 274
},
{
"epoch": 0.425531914893617,
"grad_norm": 0.7466466117649035,
"learning_rate": 6.4686451183894604e-06,
"loss": 0.1487,
"step": 275
},
{
"epoch": 0.42707930367504837,
"grad_norm": 0.925528550999488,
"learning_rate": 6.444679311135112e-06,
"loss": 0.1871,
"step": 276
},
{
"epoch": 0.42862669245647966,
"grad_norm": 1.009806729585514,
"learning_rate": 6.420677234911908e-06,
"loss": 0.1802,
"step": 277
},
{
"epoch": 0.430174081237911,
"grad_norm": 0.69368811498168,
"learning_rate": 6.396639492296868e-06,
"loss": 0.1084,
"step": 278
},
{
"epoch": 0.43172147001934236,
"grad_norm": 0.8174553997081364,
"learning_rate": 6.372566686762427e-06,
"loss": 0.1539,
"step": 279
},
{
"epoch": 0.4332688588007737,
"grad_norm": 0.6261193755927186,
"learning_rate": 6.348459422661276e-06,
"loss": 0.1721,
"step": 280
},
{
"epoch": 0.43481624758220505,
"grad_norm": 0.9101767005265041,
"learning_rate": 6.324318305211201e-06,
"loss": 0.1894,
"step": 281
},
{
"epoch": 0.43636363636363634,
"grad_norm": 0.7306705714952607,
"learning_rate": 6.300143940479881e-06,
"loss": 0.1486,
"step": 282
},
{
"epoch": 0.4379110251450677,
"grad_norm": 1.0430999001339727,
"learning_rate": 6.275936935369675e-06,
"loss": 0.1599,
"step": 283
},
{
"epoch": 0.43945841392649904,
"grad_norm": 1.0527127670370553,
"learning_rate": 6.251697897602384e-06,
"loss": 0.1728,
"step": 284
},
{
"epoch": 0.4410058027079304,
"grad_norm": 0.9008186241875709,
"learning_rate": 6.227427435703997e-06,
"loss": 0.186,
"step": 285
},
{
"epoch": 0.4425531914893617,
"grad_norm": 0.8159978628427745,
"learning_rate": 6.203126158989411e-06,
"loss": 0.1423,
"step": 286
},
{
"epoch": 0.444100580270793,
"grad_norm": 1.2441400565404073,
"learning_rate": 6.178794677547138e-06,
"loss": 0.1783,
"step": 287
},
{
"epoch": 0.44564796905222437,
"grad_norm": 0.7299868042294474,
"learning_rate": 6.154433602223979e-06,
"loss": 0.2037,
"step": 288
},
{
"epoch": 0.4471953578336557,
"grad_norm": 0.7287913797149187,
"learning_rate": 6.130043544609707e-06,
"loss": 0.1436,
"step": 289
},
{
"epoch": 0.44874274661508706,
"grad_norm": 0.9336188537541638,
"learning_rate": 6.105625117021692e-06,
"loss": 0.1431,
"step": 290
},
{
"epoch": 0.45029013539651835,
"grad_norm": 0.8864467782531835,
"learning_rate": 6.0811789324895365e-06,
"loss": 0.1831,
"step": 291
},
{
"epoch": 0.4518375241779497,
"grad_norm": 0.8128505164251555,
"learning_rate": 6.056705604739696e-06,
"loss": 0.146,
"step": 292
},
{
"epoch": 0.45338491295938105,
"grad_norm": 0.743033086175423,
"learning_rate": 6.032205748180054e-06,
"loss": 0.1511,
"step": 293
},
{
"epoch": 0.4549323017408124,
"grad_norm": 0.8133097203155237,
"learning_rate": 6.0076799778845105e-06,
"loss": 0.1766,
"step": 294
},
{
"epoch": 0.4564796905222437,
"grad_norm": 0.8758885860866176,
"learning_rate": 5.983128909577532e-06,
"loss": 0.1945,
"step": 295
},
{
"epoch": 0.45802707930367503,
"grad_norm": 0.9220994044826513,
"learning_rate": 5.958553159618693e-06,
"loss": 0.164,
"step": 296
},
{
"epoch": 0.4595744680851064,
"grad_norm": 0.8442307755597878,
"learning_rate": 5.933953344987215e-06,
"loss": 0.1419,
"step": 297
},
{
"epoch": 0.4611218568665377,
"grad_norm": 0.9127015177088472,
"learning_rate": 5.9093300832664625e-06,
"loss": 0.1591,
"step": 298
},
{
"epoch": 0.4626692456479691,
"grad_norm": 0.8352598645464135,
"learning_rate": 5.8846839926284435e-06,
"loss": 0.1847,
"step": 299
},
{
"epoch": 0.46421663442940037,
"grad_norm": 0.7805394190395212,
"learning_rate": 5.860015691818292e-06,
"loss": 0.1593,
"step": 300
},
{
"epoch": 0.4657640232108317,
"grad_norm": 0.6868076852691652,
"learning_rate": 5.835325800138736e-06,
"loss": 0.1839,
"step": 301
},
{
"epoch": 0.46731141199226306,
"grad_norm": 0.6446333964912649,
"learning_rate": 5.810614937434537e-06,
"loss": 0.1363,
"step": 302
},
{
"epoch": 0.4688588007736944,
"grad_norm": 0.6266474020290238,
"learning_rate": 5.78588372407695e-06,
"loss": 0.1287,
"step": 303
},
{
"epoch": 0.4704061895551257,
"grad_norm": 0.8020334897381506,
"learning_rate": 5.761132780948132e-06,
"loss": 0.1941,
"step": 304
},
{
"epoch": 0.47195357833655704,
"grad_norm": 1.03099678579595,
"learning_rate": 5.736362729425558e-06,
"loss": 0.1569,
"step": 305
},
{
"epoch": 0.4735009671179884,
"grad_norm": 0.9756997687881311,
"learning_rate": 5.711574191366427e-06,
"loss": 0.1782,
"step": 306
},
{
"epoch": 0.47504835589941974,
"grad_norm": 1.0369353304153857,
"learning_rate": 5.686767789092041e-06,
"loss": 0.1659,
"step": 307
},
{
"epoch": 0.4765957446808511,
"grad_norm": 1.140272459198583,
"learning_rate": 5.661944145372193e-06,
"loss": 0.1703,
"step": 308
},
{
"epoch": 0.4781431334622824,
"grad_norm": 0.7632214706912948,
"learning_rate": 5.637103883409525e-06,
"loss": 0.164,
"step": 309
},
{
"epoch": 0.4796905222437137,
"grad_norm": 0.5611927337694791,
"learning_rate": 5.612247626823878e-06,
"loss": 0.1461,
"step": 310
},
{
"epoch": 0.48123791102514507,
"grad_norm": 0.7947755945654648,
"learning_rate": 5.587375999636645e-06,
"loss": 0.1408,
"step": 311
},
{
"epoch": 0.4827852998065764,
"grad_norm": 0.9159154539685236,
"learning_rate": 5.562489626255104e-06,
"loss": 0.1427,
"step": 312
},
{
"epoch": 0.4843326885880077,
"grad_norm": 1.3372085065185557,
"learning_rate": 5.5375891314567335e-06,
"loss": 0.2175,
"step": 313
},
{
"epoch": 0.48588007736943906,
"grad_norm": 0.6937100410496474,
"learning_rate": 5.512675140373537e-06,
"loss": 0.1528,
"step": 314
},
{
"epoch": 0.4874274661508704,
"grad_norm": 0.712957909125913,
"learning_rate": 5.487748278476342e-06,
"loss": 0.1509,
"step": 315
},
{
"epoch": 0.48897485493230175,
"grad_norm": 0.7306871594587099,
"learning_rate": 5.462809171559104e-06,
"loss": 0.1481,
"step": 316
},
{
"epoch": 0.4905222437137331,
"grad_norm": 0.6930375069080879,
"learning_rate": 5.437858445723191e-06,
"loss": 0.137,
"step": 317
},
{
"epoch": 0.4920696324951644,
"grad_norm": 0.8878072207130583,
"learning_rate": 5.412896727361663e-06,
"loss": 0.1395,
"step": 318
},
{
"epoch": 0.49361702127659574,
"grad_norm": 0.8316391266848147,
"learning_rate": 5.387924643143553e-06,
"loss": 0.1642,
"step": 319
},
{
"epoch": 0.4951644100580271,
"grad_norm": 0.8029209964855811,
"learning_rate": 5.362942819998131e-06,
"loss": 0.1906,
"step": 320
},
{
"epoch": 0.49671179883945843,
"grad_norm": 0.8051088472612411,
"learning_rate": 5.337951885099167e-06,
"loss": 0.1595,
"step": 321
},
{
"epoch": 0.4982591876208897,
"grad_norm": 0.7796514070904359,
"learning_rate": 5.312952465849173e-06,
"loss": 0.1485,
"step": 322
},
{
"epoch": 0.49980657640232107,
"grad_norm": 1.1365547852823168,
"learning_rate": 5.287945189863676e-06,
"loss": 0.1604,
"step": 323
},
{
"epoch": 0.5013539651837524,
"grad_norm": 0.7340585841316194,
"learning_rate": 5.262930684955439e-06,
"loss": 0.1733,
"step": 324
},
{
"epoch": 0.5029013539651838,
"grad_norm": 0.7190666929171957,
"learning_rate": 5.237909579118713e-06,
"loss": 0.1744,
"step": 325
},
{
"epoch": 0.5044487427466151,
"grad_norm": 0.9269637883180393,
"learning_rate": 5.212882500513462e-06,
"loss": 0.1574,
"step": 326
},
{
"epoch": 0.5059961315280465,
"grad_norm": 0.7874315757205848,
"learning_rate": 5.187850077449604e-06,
"loss": 0.1313,
"step": 327
},
{
"epoch": 0.5075435203094778,
"grad_norm": 0.8119507240468493,
"learning_rate": 5.162812938371226e-06,
"loss": 0.1766,
"step": 328
},
{
"epoch": 0.509090909090909,
"grad_norm": 0.8677259459647997,
"learning_rate": 5.137771711840811e-06,
"loss": 0.2096,
"step": 329
},
{
"epoch": 0.5106382978723404,
"grad_norm": 0.8279345699617693,
"learning_rate": 5.112727026523461e-06,
"loss": 0.176,
"step": 330
},
{
"epoch": 0.5121856866537717,
"grad_norm": 1.217364237331737,
"learning_rate": 5.087679511171113e-06,
"loss": 0.1353,
"step": 331
},
{
"epoch": 0.5137330754352031,
"grad_norm": 0.9200408560812419,
"learning_rate": 5.062629794606748e-06,
"loss": 0.1879,
"step": 332
},
{
"epoch": 0.5152804642166344,
"grad_norm": 0.958235414428453,
"learning_rate": 5.03757850570861e-06,
"loss": 0.159,
"step": 333
},
{
"epoch": 0.5168278529980658,
"grad_norm": 0.6653985756007545,
"learning_rate": 5.01252627339442e-06,
"loss": 0.1382,
"step": 334
},
{
"epoch": 0.5183752417794971,
"grad_norm": 0.8416906166547965,
"learning_rate": 4.987473726605581e-06,
"loss": 0.1523,
"step": 335
},
{
"epoch": 0.5199226305609285,
"grad_norm": 1.0238104433482986,
"learning_rate": 4.9624214942913916e-06,
"loss": 0.1822,
"step": 336
},
{
"epoch": 0.5214700193423598,
"grad_norm": 0.8006256882782827,
"learning_rate": 4.9373702053932534e-06,
"loss": 0.1419,
"step": 337
},
{
"epoch": 0.523017408123791,
"grad_norm": 0.8061908010151759,
"learning_rate": 4.912320488828887e-06,
"loss": 0.1851,
"step": 338
},
{
"epoch": 0.5245647969052224,
"grad_norm": 0.777162379681793,
"learning_rate": 4.88727297347654e-06,
"loss": 0.1517,
"step": 339
},
{
"epoch": 0.5261121856866537,
"grad_norm": 0.7845485497179473,
"learning_rate": 4.862228288159191e-06,
"loss": 0.1996,
"step": 340
},
{
"epoch": 0.5276595744680851,
"grad_norm": 1.3111512506729992,
"learning_rate": 4.837187061628777e-06,
"loss": 0.1175,
"step": 341
},
{
"epoch": 0.5292069632495164,
"grad_norm": 0.8803762689051036,
"learning_rate": 4.8121499225503974e-06,
"loss": 0.1483,
"step": 342
},
{
"epoch": 0.5307543520309478,
"grad_norm": 0.8033512929968718,
"learning_rate": 4.787117499486539e-06,
"loss": 0.1509,
"step": 343
},
{
"epoch": 0.5323017408123791,
"grad_norm": 0.7897542664475099,
"learning_rate": 4.762090420881289e-06,
"loss": 0.1676,
"step": 344
},
{
"epoch": 0.5338491295938105,
"grad_norm": 0.9581686218310035,
"learning_rate": 4.737069315044562e-06,
"loss": 0.1714,
"step": 345
},
{
"epoch": 0.5353965183752418,
"grad_norm": 0.7293567937435059,
"learning_rate": 4.712054810136327e-06,
"loss": 0.1674,
"step": 346
},
{
"epoch": 0.5369439071566731,
"grad_norm": 0.7541776511601691,
"learning_rate": 4.687047534150829e-06,
"loss": 0.151,
"step": 347
},
{
"epoch": 0.5384912959381044,
"grad_norm": 0.8792086087287215,
"learning_rate": 4.662048114900837e-06,
"loss": 0.1904,
"step": 348
},
{
"epoch": 0.5400386847195358,
"grad_norm": 0.802484636218363,
"learning_rate": 4.6370571800018695e-06,
"loss": 0.144,
"step": 349
},
{
"epoch": 0.5415860735009671,
"grad_norm": 0.6669155376965741,
"learning_rate": 4.612075356856447e-06,
"loss": 0.1569,
"step": 350
},
{
"epoch": 0.5431334622823984,
"grad_norm": 0.8565494219857237,
"learning_rate": 4.587103272638339e-06,
"loss": 0.1776,
"step": 351
},
{
"epoch": 0.5446808510638298,
"grad_norm": 0.877939514420621,
"learning_rate": 4.562141554276811e-06,
"loss": 0.1923,
"step": 352
},
{
"epoch": 0.5462282398452611,
"grad_norm": 0.9975900853047521,
"learning_rate": 4.537190828440898e-06,
"loss": 0.1555,
"step": 353
},
{
"epoch": 0.5477756286266925,
"grad_norm": 0.8076557611787406,
"learning_rate": 4.512251721523659e-06,
"loss": 0.1604,
"step": 354
},
{
"epoch": 0.5493230174081238,
"grad_norm": 0.8092417085033835,
"learning_rate": 4.487324859626465e-06,
"loss": 0.152,
"step": 355
},
{
"epoch": 0.5508704061895551,
"grad_norm": 0.7452702096804573,
"learning_rate": 4.462410868543268e-06,
"loss": 0.1807,
"step": 356
},
{
"epoch": 0.5524177949709864,
"grad_norm": 0.8201078882983477,
"learning_rate": 4.437510373744897e-06,
"loss": 0.1287,
"step": 357
},
{
"epoch": 0.5539651837524178,
"grad_norm": 0.8153600193640145,
"learning_rate": 4.4126240003633565e-06,
"loss": 0.1441,
"step": 358
},
{
"epoch": 0.5555125725338491,
"grad_norm": 0.690274756981255,
"learning_rate": 4.387752373176123e-06,
"loss": 0.1508,
"step": 359
},
{
"epoch": 0.5570599613152805,
"grad_norm": 0.954303121342827,
"learning_rate": 4.362896116590475e-06,
"loss": 0.1898,
"step": 360
},
{
"epoch": 0.5586073500967118,
"grad_norm": 0.9883396338363168,
"learning_rate": 4.3380558546278075e-06,
"loss": 0.1461,
"step": 361
},
{
"epoch": 0.5601547388781432,
"grad_norm": 0.8720292048845367,
"learning_rate": 4.313232210907959e-06,
"loss": 0.1792,
"step": 362
},
{
"epoch": 0.5617021276595745,
"grad_norm": 0.8462662373236136,
"learning_rate": 4.2884258086335755e-06,
"loss": 0.122,
"step": 363
},
{
"epoch": 0.5632495164410058,
"grad_norm": 0.9111327250477934,
"learning_rate": 4.2636372705744425e-06,
"loss": 0.1792,
"step": 364
},
{
"epoch": 0.5647969052224371,
"grad_norm": 0.7856086496842567,
"learning_rate": 4.238867219051868e-06,
"loss": 0.1501,
"step": 365
},
{
"epoch": 0.5663442940038684,
"grad_norm": 0.7997407226449322,
"learning_rate": 4.214116275923051e-06,
"loss": 0.1304,
"step": 366
},
{
"epoch": 0.5678916827852998,
"grad_norm": 0.900973362979182,
"learning_rate": 4.1893850625654626e-06,
"loss": 0.1393,
"step": 367
},
{
"epoch": 0.5694390715667311,
"grad_norm": 0.6489085923087003,
"learning_rate": 4.1646741998612676e-06,
"loss": 0.1409,
"step": 368
},
{
"epoch": 0.5709864603481625,
"grad_norm": 0.7749796859552089,
"learning_rate": 4.1399843081817085e-06,
"loss": 0.1177,
"step": 369
},
{
"epoch": 0.5725338491295938,
"grad_norm": 0.7000936335478609,
"learning_rate": 4.115316007371557e-06,
"loss": 0.1612,
"step": 370
},
{
"epoch": 0.5740812379110252,
"grad_norm": 0.8127434954779953,
"learning_rate": 4.090669916733539e-06,
"loss": 0.1498,
"step": 371
},
{
"epoch": 0.5756286266924565,
"grad_norm": 0.7873781350590033,
"learning_rate": 4.066046655012786e-06,
"loss": 0.1226,
"step": 372
},
{
"epoch": 0.5771760154738879,
"grad_norm": 0.5947618883920969,
"learning_rate": 4.041446840381309e-06,
"loss": 0.1168,
"step": 373
},
{
"epoch": 0.5787234042553191,
"grad_norm": 0.5686346544180676,
"learning_rate": 4.016871090422471e-06,
"loss": 0.1348,
"step": 374
},
{
"epoch": 0.5802707930367504,
"grad_norm": 1.0091986954108587,
"learning_rate": 3.992320022115492e-06,
"loss": 0.1647,
"step": 375
},
{
"epoch": 0.5818181818181818,
"grad_norm": 0.753705804127023,
"learning_rate": 3.9677942518199465e-06,
"loss": 0.1723,
"step": 376
},
{
"epoch": 0.5833655705996131,
"grad_norm": 0.7805349781585388,
"learning_rate": 3.943294395260305e-06,
"loss": 0.1589,
"step": 377
},
{
"epoch": 0.5849129593810445,
"grad_norm": 0.9120982175985182,
"learning_rate": 3.918821067510464e-06,
"loss": 0.1556,
"step": 378
},
{
"epoch": 0.5864603481624758,
"grad_norm": 0.7036542820999608,
"learning_rate": 3.89437488297831e-06,
"loss": 0.1673,
"step": 379
},
{
"epoch": 0.5880077369439072,
"grad_norm": 0.7972229664002071,
"learning_rate": 3.869956455390295e-06,
"loss": 0.1792,
"step": 380
},
{
"epoch": 0.5895551257253385,
"grad_norm": 0.856847194082784,
"learning_rate": 3.845566397776022e-06,
"loss": 0.1558,
"step": 381
},
{
"epoch": 0.5911025145067699,
"grad_norm": 0.6825960488664438,
"learning_rate": 3.821205322452863e-06,
"loss": 0.1162,
"step": 382
},
{
"epoch": 0.5926499032882011,
"grad_norm": 0.6872597841413525,
"learning_rate": 3.796873841010591e-06,
"loss": 0.1365,
"step": 383
},
{
"epoch": 0.5941972920696325,
"grad_norm": 0.8822249956202902,
"learning_rate": 3.7725725642960047e-06,
"loss": 0.1533,
"step": 384
},
{
"epoch": 0.5957446808510638,
"grad_norm": 0.8258608583085321,
"learning_rate": 3.748302102397618e-06,
"loss": 0.1489,
"step": 385
},
{
"epoch": 0.5972920696324951,
"grad_norm": 0.6758686948774851,
"learning_rate": 3.7240630646303262e-06,
"loss": 0.1382,
"step": 386
},
{
"epoch": 0.5988394584139265,
"grad_norm": 0.7479959567872794,
"learning_rate": 3.6998560595201188e-06,
"loss": 0.13,
"step": 387
},
{
"epoch": 0.6003868471953578,
"grad_norm": 0.8091837019411817,
"learning_rate": 3.675681694788801e-06,
"loss": 0.2024,
"step": 388
},
{
"epoch": 0.6019342359767892,
"grad_norm": 0.7787515452667286,
"learning_rate": 3.6515405773387257e-06,
"loss": 0.1841,
"step": 389
},
{
"epoch": 0.6034816247582205,
"grad_norm": 0.9365385926141763,
"learning_rate": 3.627433313237576e-06,
"loss": 0.1744,
"step": 390
},
{
"epoch": 0.6050290135396519,
"grad_norm": 0.8951297468394839,
"learning_rate": 3.603360507703133e-06,
"loss": 0.1464,
"step": 391
},
{
"epoch": 0.6065764023210831,
"grad_norm": 0.7769011344776973,
"learning_rate": 3.5793227650880928e-06,
"loss": 0.1488,
"step": 392
},
{
"epoch": 0.6081237911025145,
"grad_norm": 0.7785381369421867,
"learning_rate": 3.555320688864889e-06,
"loss": 0.1438,
"step": 393
},
{
"epoch": 0.6096711798839458,
"grad_norm": 0.6766660462955612,
"learning_rate": 3.531354881610539e-06,
"loss": 0.1395,
"step": 394
},
{
"epoch": 0.6112185686653772,
"grad_norm": 0.7200209087019646,
"learning_rate": 3.507425944991529e-06,
"loss": 0.1404,
"step": 395
},
{
"epoch": 0.6127659574468085,
"grad_norm": 0.7787502319093067,
"learning_rate": 3.483534479748688e-06,
"loss": 0.1349,
"step": 396
},
{
"epoch": 0.6143133462282399,
"grad_norm": 0.6549871943316503,
"learning_rate": 3.4596810856821304e-06,
"loss": 0.1272,
"step": 397
},
{
"epoch": 0.6158607350096712,
"grad_norm": 0.5575760430347522,
"learning_rate": 3.4358663616361775e-06,
"loss": 0.1312,
"step": 398
},
{
"epoch": 0.6174081237911025,
"grad_norm": 0.6937467533738239,
"learning_rate": 3.4120909054843375e-06,
"loss": 0.1501,
"step": 399
},
{
"epoch": 0.6189555125725339,
"grad_norm": 0.7634337264859942,
"learning_rate": 3.3883553141142884e-06,
"loss": 0.1445,
"step": 400
},
{
"epoch": 0.6205029013539651,
"grad_norm": 0.7823599037925153,
"learning_rate": 3.3646601834128924e-06,
"loss": 0.162,
"step": 401
},
{
"epoch": 0.6220502901353965,
"grad_norm": 0.6783309048766991,
"learning_rate": 3.3410061082512422e-06,
"loss": 0.1171,
"step": 402
},
{
"epoch": 0.6235976789168278,
"grad_norm": 0.7918408052364138,
"learning_rate": 3.3173936824697174e-06,
"loss": 0.1511,
"step": 403
},
{
"epoch": 0.6251450676982592,
"grad_norm": 0.7888490271513304,
"learning_rate": 3.293823498863087e-06,
"loss": 0.1883,
"step": 404
},
{
"epoch": 0.6266924564796905,
"grad_norm": 0.5416163054881282,
"learning_rate": 3.2702961491656197e-06,
"loss": 0.1654,
"step": 405
},
{
"epoch": 0.6282398452611219,
"grad_norm": 0.6181852228006771,
"learning_rate": 3.2468122240362287e-06,
"loss": 0.1354,
"step": 406
},
{
"epoch": 0.6297872340425532,
"grad_norm": 0.8020667401094614,
"learning_rate": 3.223372313043647e-06,
"loss": 0.1539,
"step": 407
},
{
"epoch": 0.6313346228239846,
"grad_norm": 0.7359292573538625,
"learning_rate": 3.1999770046516198e-06,
"loss": 0.1551,
"step": 408
},
{
"epoch": 0.6328820116054159,
"grad_norm": 0.914850984873592,
"learning_rate": 3.1766268862041406e-06,
"loss": 0.1581,
"step": 409
},
{
"epoch": 0.6344294003868471,
"grad_norm": 0.7260085078323978,
"learning_rate": 3.1533225439106965e-06,
"loss": 0.1508,
"step": 410
},
{
"epoch": 0.6359767891682785,
"grad_norm": 0.693832047401324,
"learning_rate": 3.130064562831553e-06,
"loss": 0.1237,
"step": 411
},
{
"epoch": 0.6375241779497098,
"grad_norm": 0.7578725636658399,
"learning_rate": 3.106853526863073e-06,
"loss": 0.1544,
"step": 412
},
{
"epoch": 0.6390715667311412,
"grad_norm": 0.6835869105715572,
"learning_rate": 3.0836900187230475e-06,
"loss": 0.1504,
"step": 413
},
{
"epoch": 0.6406189555125725,
"grad_norm": 1.0190525154626127,
"learning_rate": 3.0605746199360755e-06,
"loss": 0.161,
"step": 414
},
{
"epoch": 0.6421663442940039,
"grad_norm": 0.9223693435650491,
"learning_rate": 3.0375079108189613e-06,
"loss": 0.159,
"step": 415
},
{
"epoch": 0.6437137330754352,
"grad_norm": 0.6438438739046826,
"learning_rate": 3.0144904704661413e-06,
"loss": 0.144,
"step": 416
},
{
"epoch": 0.6452611218568666,
"grad_norm": 0.7709884247091324,
"learning_rate": 2.991522876735154e-06,
"loss": 0.1707,
"step": 417
},
{
"epoch": 0.6468085106382979,
"grad_norm": 0.6573644892367149,
"learning_rate": 2.9686057062321226e-06,
"loss": 0.1216,
"step": 418
},
{
"epoch": 0.6483558994197292,
"grad_norm": 0.8423214578382158,
"learning_rate": 2.9457395342972904e-06,
"loss": 0.178,
"step": 419
},
{
"epoch": 0.6499032882011605,
"grad_norm": 0.9695303714301211,
"learning_rate": 2.9229249349905686e-06,
"loss": 0.1646,
"step": 420
},
{
"epoch": 0.6514506769825918,
"grad_norm": 0.7682943797949924,
"learning_rate": 2.900162481077126e-06,
"loss": 0.1231,
"step": 421
},
{
"epoch": 0.6529980657640232,
"grad_norm": 0.9321522394226056,
"learning_rate": 2.8774527440130173e-06,
"loss": 0.1707,
"step": 422
},
{
"epoch": 0.6545454545454545,
"grad_norm": 0.7877697276319559,
"learning_rate": 2.8547962939308187e-06,
"loss": 0.1728,
"step": 423
},
{
"epoch": 0.6560928433268859,
"grad_norm": 0.9844015326541935,
"learning_rate": 2.8321936996253368e-06,
"loss": 0.1317,
"step": 424
},
{
"epoch": 0.6576402321083172,
"grad_norm": 0.7853031572232724,
"learning_rate": 2.8096455285393094e-06,
"loss": 0.1555,
"step": 425
},
{
"epoch": 0.6591876208897486,
"grad_norm": 0.6814221721272227,
"learning_rate": 2.787152346749173e-06,
"loss": 0.1778,
"step": 426
},
{
"epoch": 0.6607350096711799,
"grad_norm": 0.6862759795309286,
"learning_rate": 2.7647147189508485e-06,
"loss": 0.1355,
"step": 427
},
{
"epoch": 0.6622823984526112,
"grad_norm": 0.6958841975782879,
"learning_rate": 2.7423332084455543e-06,
"loss": 0.1953,
"step": 428
},
{
"epoch": 0.6638297872340425,
"grad_norm": 0.848926127018151,
"learning_rate": 2.720008377125682e-06,
"loss": 0.1622,
"step": 429
},
{
"epoch": 0.6653771760154739,
"grad_norm": 0.6707487623895614,
"learning_rate": 2.697740785460675e-06,
"loss": 0.1371,
"step": 430
},
{
"epoch": 0.6669245647969052,
"grad_norm": 0.854893783886741,
"learning_rate": 2.6755309924829657e-06,
"loss": 0.1179,
"step": 431
},
{
"epoch": 0.6684719535783366,
"grad_norm": 0.9468587765577116,
"learning_rate": 2.6533795557739407e-06,
"loss": 0.1559,
"step": 432
},
{
"epoch": 0.6700193423597679,
"grad_norm": 0.7665928032047755,
"learning_rate": 2.6312870314499335e-06,
"loss": 0.1631,
"step": 433
},
{
"epoch": 0.6715667311411992,
"grad_norm": 0.7076292638363726,
"learning_rate": 2.609253974148278e-06,
"loss": 0.1656,
"step": 434
},
{
"epoch": 0.6731141199226306,
"grad_norm": 0.8836427279540453,
"learning_rate": 2.5872809370133704e-06,
"loss": 0.1695,
"step": 435
},
{
"epoch": 0.6746615087040619,
"grad_norm": 0.7599912980454478,
"learning_rate": 2.5653684716827904e-06,
"loss": 0.1436,
"step": 436
},
{
"epoch": 0.6762088974854932,
"grad_norm": 0.7628768285639652,
"learning_rate": 2.5435171282734563e-06,
"loss": 0.137,
"step": 437
},
{
"epoch": 0.6777562862669245,
"grad_norm": 0.8719128186722851,
"learning_rate": 2.5217274553677975e-06,
"loss": 0.1594,
"step": 438
},
{
"epoch": 0.6793036750483559,
"grad_norm": 0.7384661815703569,
"learning_rate": 2.5000000000000015e-06,
"loss": 0.1182,
"step": 439
},
{
"epoch": 0.6808510638297872,
"grad_norm": 0.8766152007220669,
"learning_rate": 2.478335307642264e-06,
"loss": 0.182,
"step": 440
},
{
"epoch": 0.6823984526112186,
"grad_norm": 0.8681737575063314,
"learning_rate": 2.4567339221911086e-06,
"loss": 0.1145,
"step": 441
},
{
"epoch": 0.6839458413926499,
"grad_norm": 0.9349868596178444,
"learning_rate": 2.435196385953727e-06,
"loss": 0.1532,
"step": 442
},
{
"epoch": 0.6854932301740813,
"grad_norm": 0.8507541638853486,
"learning_rate": 2.413723239634356e-06,
"loss": 0.1875,
"step": 443
},
{
"epoch": 0.6870406189555126,
"grad_norm": 0.7401767343899905,
"learning_rate": 2.3923150223207176e-06,
"loss": 0.1768,
"step": 444
},
{
"epoch": 0.688588007736944,
"grad_norm": 0.898874838700681,
"learning_rate": 2.370972271470475e-06,
"loss": 0.161,
"step": 445
},
{
"epoch": 0.6901353965183752,
"grad_norm": 0.7633442244444167,
"learning_rate": 2.3496955228977437e-06,
"loss": 0.1386,
"step": 446
},
{
"epoch": 0.6916827852998065,
"grad_norm": 0.769712330585869,
"learning_rate": 2.328485310759635e-06,
"loss": 0.1533,
"step": 447
},
{
"epoch": 0.6932301740812379,
"grad_norm": 0.6477209937018277,
"learning_rate": 2.307342167542854e-06,
"loss": 0.1488,
"step": 448
},
{
"epoch": 0.6947775628626692,
"grad_norm": 0.7306672310223877,
"learning_rate": 2.286266624050326e-06,
"loss": 0.1173,
"step": 449
},
{
"epoch": 0.6963249516441006,
"grad_norm": 0.7111958316918345,
"learning_rate": 2.265259209387867e-06,
"loss": 0.1773,
"step": 450
},
{
"epoch": 0.6978723404255319,
"grad_norm": 0.8325136808450394,
"learning_rate": 2.2443204509509094e-06,
"loss": 0.1654,
"step": 451
},
{
"epoch": 0.6994197292069633,
"grad_norm": 0.851693409931118,
"learning_rate": 2.2234508744112564e-06,
"loss": 0.1283,
"step": 452
},
{
"epoch": 0.7009671179883946,
"grad_norm": 0.9724202950967027,
"learning_rate": 2.202651003703885e-06,
"loss": 0.1413,
"step": 453
},
{
"epoch": 0.702514506769826,
"grad_norm": 0.8475158397818416,
"learning_rate": 2.181921361013794e-06,
"loss": 0.128,
"step": 454
},
{
"epoch": 0.7040618955512572,
"grad_norm": 0.7280803414009301,
"learning_rate": 2.16126246676289e-06,
"loss": 0.1452,
"step": 455
},
{
"epoch": 0.7056092843326885,
"grad_norm": 0.6516083419869029,
"learning_rate": 2.140674839596931e-06,
"loss": 0.1411,
"step": 456
},
{
"epoch": 0.7071566731141199,
"grad_norm": 0.882815614999906,
"learning_rate": 2.1201589963724933e-06,
"loss": 0.1576,
"step": 457
},
{
"epoch": 0.7087040618955512,
"grad_norm": 0.8378441046800585,
"learning_rate": 2.09971545214401e-06,
"loss": 0.1872,
"step": 458
},
{
"epoch": 0.7102514506769826,
"grad_norm": 1.077575435913912,
"learning_rate": 2.0793447201508288e-06,
"loss": 0.1982,
"step": 459
},
{
"epoch": 0.7117988394584139,
"grad_norm": 0.7117488495836712,
"learning_rate": 2.0590473118043326e-06,
"loss": 0.1549,
"step": 460
},
{
"epoch": 0.7133462282398453,
"grad_norm": 0.8367490226827945,
"learning_rate": 2.0388237366751005e-06,
"loss": 0.1361,
"step": 461
},
{
"epoch": 0.7148936170212766,
"grad_norm": 0.8481551715321101,
"learning_rate": 2.01867450248011e-06,
"loss": 0.1292,
"step": 462
},
{
"epoch": 0.716441005802708,
"grad_norm": 0.8415490119824771,
"learning_rate": 1.998600115069998e-06,
"loss": 0.142,
"step": 463
},
{
"epoch": 0.7179883945841392,
"grad_norm": 1.0486480021813904,
"learning_rate": 1.978601078416357e-06,
"loss": 0.1868,
"step": 464
},
{
"epoch": 0.7195357833655706,
"grad_norm": 0.7797492350963128,
"learning_rate": 1.9586778945990785e-06,
"loss": 0.1815,
"step": 465
},
{
"epoch": 0.7210831721470019,
"grad_norm": 0.8601078424260443,
"learning_rate": 1.9388310637937606e-06,
"loss": 0.1931,
"step": 466
},
{
"epoch": 0.7226305609284333,
"grad_norm": 0.7287701444089209,
"learning_rate": 1.9190610842591386e-06,
"loss": 0.1325,
"step": 467
},
{
"epoch": 0.7241779497098646,
"grad_norm": 0.6459207596034287,
"learning_rate": 1.8993684523245842e-06,
"loss": 0.1718,
"step": 468
},
{
"epoch": 0.725725338491296,
"grad_norm": 0.8444611842470297,
"learning_rate": 1.879753662377637e-06,
"loss": 0.1734,
"step": 469
},
{
"epoch": 0.7272727272727273,
"grad_norm": 1.1245502955919686,
"learning_rate": 1.8602172068516011e-06,
"loss": 0.1916,
"step": 470
},
{
"epoch": 0.7288201160541586,
"grad_norm": 0.784743071773171,
"learning_rate": 1.8407595762131814e-06,
"loss": 0.1467,
"step": 471
},
{
"epoch": 0.73036750483559,
"grad_norm": 0.983199125764354,
"learning_rate": 1.8213812589501611e-06,
"loss": 0.1249,
"step": 472
},
{
"epoch": 0.7319148936170212,
"grad_norm": 0.8730252086980543,
"learning_rate": 1.8020827415591496e-06,
"loss": 0.1512,
"step": 473
},
{
"epoch": 0.7334622823984526,
"grad_norm": 0.704335948297951,
"learning_rate": 1.7828645085333645e-06,
"loss": 0.1501,
"step": 474
},
{
"epoch": 0.7350096711798839,
"grad_norm": 1.085272738395196,
"learning_rate": 1.7637270423504664e-06,
"loss": 0.1451,
"step": 475
},
{
"epoch": 0.7365570599613153,
"grad_norm": 1.1023625632272782,
"learning_rate": 1.7446708234604498e-06,
"loss": 0.1488,
"step": 476
},
{
"epoch": 0.7381044487427466,
"grad_norm": 0.759099412840817,
"learning_rate": 1.7256963302735752e-06,
"loss": 0.1261,
"step": 477
},
{
"epoch": 0.739651837524178,
"grad_norm": 1.042379709174401,
"learning_rate": 1.7068040391483676e-06,
"loss": 0.1479,
"step": 478
},
{
"epoch": 0.7411992263056093,
"grad_norm": 0.6980911952884176,
"learning_rate": 1.6879944243796477e-06,
"loss": 0.1536,
"step": 479
},
{
"epoch": 0.7427466150870407,
"grad_norm": 0.975310427148216,
"learning_rate": 1.6692679581866334e-06,
"loss": 0.1613,
"step": 480
},
{
"epoch": 0.744294003868472,
"grad_norm": 0.7780629748292098,
"learning_rate": 1.650625110701079e-06,
"loss": 0.1398,
"step": 481
},
{
"epoch": 0.7458413926499032,
"grad_norm": 0.7776217447464706,
"learning_rate": 1.632066349955474e-06,
"loss": 0.1908,
"step": 482
},
{
"epoch": 0.7473887814313346,
"grad_norm": 0.780682741044397,
"learning_rate": 1.6135921418712959e-06,
"loss": 0.1509,
"step": 483
},
{
"epoch": 0.7489361702127659,
"grad_norm": 0.7303004018327728,
"learning_rate": 1.5952029502473032e-06,
"loss": 0.1767,
"step": 484
},
{
"epoch": 0.7504835589941973,
"grad_norm": 0.7719095887591466,
"learning_rate": 1.5768992367479058e-06,
"loss": 0.1773,
"step": 485
},
{
"epoch": 0.7520309477756286,
"grad_norm": 0.7753671077322888,
"learning_rate": 1.5586814608915673e-06,
"loss": 0.1404,
"step": 486
},
{
"epoch": 0.75357833655706,
"grad_norm": 0.8778912952233745,
"learning_rate": 1.5405500800392643e-06,
"loss": 0.1494,
"step": 487
},
{
"epoch": 0.7551257253384913,
"grad_norm": 0.7207732630110582,
"learning_rate": 1.5225055493830132e-06,
"loss": 0.1556,
"step": 488
},
{
"epoch": 0.7566731141199227,
"grad_norm": 0.6666531180840158,
"learning_rate": 1.5045483219344387e-06,
"loss": 0.153,
"step": 489
},
{
"epoch": 0.758220502901354,
"grad_norm": 0.8857380643372479,
"learning_rate": 1.4866788485133988e-06,
"loss": 0.1627,
"step": 490
},
{
"epoch": 0.7597678916827852,
"grad_norm": 0.8677779634759353,
"learning_rate": 1.4688975777366716e-06,
"loss": 0.1254,
"step": 491
},
{
"epoch": 0.7613152804642166,
"grad_norm": 0.8755027376849471,
"learning_rate": 1.4512049560066837e-06,
"loss": 0.1912,
"step": 492
},
{
"epoch": 0.7628626692456479,
"grad_norm": 0.7274771013337785,
"learning_rate": 1.433601427500318e-06,
"loss": 0.1456,
"step": 493
},
{
"epoch": 0.7644100580270793,
"grad_norm": 0.8311254044984683,
"learning_rate": 1.4160874341577447e-06,
"loss": 0.1577,
"step": 494
},
{
"epoch": 0.7659574468085106,
"grad_norm": 0.6929508359619224,
"learning_rate": 1.3986634156713418e-06,
"loss": 0.1426,
"step": 495
},
{
"epoch": 0.767504835589942,
"grad_norm": 0.9647780179486356,
"learning_rate": 1.3813298094746491e-06,
"loss": 0.1623,
"step": 496
},
{
"epoch": 0.7690522243713733,
"grad_norm": 0.8298466957909739,
"learning_rate": 1.3640870507313859e-06,
"loss": 0.126,
"step": 497
},
{
"epoch": 0.7705996131528047,
"grad_norm": 0.6934035514513522,
"learning_rate": 1.3469355723245303e-06,
"loss": 0.1808,
"step": 498
},
{
"epoch": 0.772147001934236,
"grad_norm": 0.6945563747887925,
"learning_rate": 1.3298758048454436e-06,
"loss": 0.1762,
"step": 499
},
{
"epoch": 0.7736943907156673,
"grad_norm": 0.8110188572724004,
"learning_rate": 1.3129081765830725e-06,
"loss": 0.1768,
"step": 500
},
{
"epoch": 0.7752417794970986,
"grad_norm": 0.9417362627663032,
"learning_rate": 1.2960331135131826e-06,
"loss": 0.1922,
"step": 501
},
{
"epoch": 0.77678916827853,
"grad_norm": 0.7408558621282384,
"learning_rate": 1.2792510392876777e-06,
"loss": 0.1252,
"step": 502
},
{
"epoch": 0.7783365570599613,
"grad_norm": 0.8018198280132611,
"learning_rate": 1.262562375223954e-06,
"loss": 0.1284,
"step": 503
},
{
"epoch": 0.7798839458413926,
"grad_norm": 0.5956535731524653,
"learning_rate": 1.245967540294329e-06,
"loss": 0.1227,
"step": 504
},
{
"epoch": 0.781431334622824,
"grad_norm": 1.683615182365151,
"learning_rate": 1.2294669511155193e-06,
"loss": 0.1568,
"step": 505
},
{
"epoch": 0.7829787234042553,
"grad_norm": 0.7157092450552371,
"learning_rate": 1.2130610219381811e-06,
"loss": 0.1356,
"step": 506
},
{
"epoch": 0.7845261121856867,
"grad_norm": 0.7525134112677587,
"learning_rate": 1.1967501646365147e-06,
"loss": 0.1889,
"step": 507
},
{
"epoch": 0.786073500967118,
"grad_norm": 0.85179131912215,
"learning_rate": 1.1805347886979219e-06,
"loss": 0.1545,
"step": 508
},
{
"epoch": 0.7876208897485493,
"grad_norm": 0.7859436718867705,
"learning_rate": 1.1644153012127208e-06,
"loss": 0.1669,
"step": 509
},
{
"epoch": 0.7891682785299806,
"grad_norm": 0.8754204423323872,
"learning_rate": 1.1483921068639353e-06,
"loss": 0.1446,
"step": 510
},
{
"epoch": 0.790715667311412,
"grad_norm": 0.8003796816525551,
"learning_rate": 1.1324656079171288e-06,
"loss": 0.1251,
"step": 511
},
{
"epoch": 0.7922630560928433,
"grad_norm": 0.7553494427488822,
"learning_rate": 1.1166362042103056e-06,
"loss": 0.1708,
"step": 512
},
{
"epoch": 0.7938104448742747,
"grad_norm": 0.9170006973065143,
"learning_rate": 1.1009042931438784e-06,
"loss": 0.1331,
"step": 513
},
{
"epoch": 0.795357833655706,
"grad_norm": 0.9176929457534573,
"learning_rate": 1.0852702696706807e-06,
"loss": 0.1472,
"step": 514
},
{
"epoch": 0.7969052224371374,
"grad_norm": 0.7902619235950765,
"learning_rate": 1.0697345262860638e-06,
"loss": 0.1419,
"step": 515
},
{
"epoch": 0.7984526112185687,
"grad_norm": 0.7127128227580106,
"learning_rate": 1.0542974530180327e-06,
"loss": 0.1464,
"step": 516
},
{
"epoch": 0.8,
"grad_norm": 0.7341599122621142,
"learning_rate": 1.0389594374174628e-06,
"loss": 0.1328,
"step": 517
},
{
"epoch": 0.8015473887814313,
"grad_norm": 0.8937108240363892,
"learning_rate": 1.0237208645483648e-06,
"loss": 0.1424,
"step": 518
},
{
"epoch": 0.8030947775628626,
"grad_norm": 0.8155596110553233,
"learning_rate": 1.00858211697822e-06,
"loss": 0.1428,
"step": 519
},
{
"epoch": 0.804642166344294,
"grad_norm": 0.7349457350632931,
"learning_rate": 9.935435747683758e-07,
"loss": 0.1396,
"step": 520
},
{
"epoch": 0.8061895551257253,
"grad_norm": 0.7316025143086685,
"learning_rate": 9.786056154645001e-07,
"loss": 0.1583,
"step": 521
},
{
"epoch": 0.8077369439071567,
"grad_norm": 0.8886659221463238,
"learning_rate": 9.637686140871121e-07,
"loss": 0.139,
"step": 522
},
{
"epoch": 0.809284332688588,
"grad_norm": 0.9534104292500418,
"learning_rate": 9.490329431221545e-07,
"loss": 0.13,
"step": 523
},
{
"epoch": 0.8108317214700194,
"grad_norm": 0.6359551213299164,
"learning_rate": 9.34398972511656e-07,
"loss": 0.1367,
"step": 524
},
{
"epoch": 0.8123791102514507,
"grad_norm": 0.729187516383015,
"learning_rate": 9.198670696444339e-07,
"loss": 0.129,
"step": 525
},
{
"epoch": 0.8139264990328821,
"grad_norm": 0.8037591313621663,
"learning_rate": 9.054375993468745e-07,
"loss": 0.1429,
"step": 526
},
{
"epoch": 0.8154738878143133,
"grad_norm": 0.5934320323085827,
"learning_rate": 8.911109238737748e-07,
"loss": 0.1128,
"step": 527
},
{
"epoch": 0.8170212765957446,
"grad_norm": 0.7449700123919022,
"learning_rate": 8.768874028992431e-07,
"loss": 0.137,
"step": 528
},
{
"epoch": 0.818568665377176,
"grad_norm": 0.6695563770828353,
"learning_rate": 8.627673935076769e-07,
"loss": 0.1137,
"step": 529
},
{
"epoch": 0.8201160541586073,
"grad_norm": 0.9802601756238717,
"learning_rate": 8.487512501847933e-07,
"loss": 0.1769,
"step": 530
},
{
"epoch": 0.8216634429400387,
"grad_norm": 0.7598450282737069,
"learning_rate": 8.348393248087289e-07,
"loss": 0.146,
"step": 531
},
{
"epoch": 0.82321083172147,
"grad_norm": 1.0053439869906184,
"learning_rate": 8.210319666412087e-07,
"loss": 0.1309,
"step": 532
},
{
"epoch": 0.8247582205029014,
"grad_norm": 0.849596372505488,
"learning_rate": 8.073295223187766e-07,
"loss": 0.1369,
"step": 533
},
{
"epoch": 0.8263056092843327,
"grad_norm": 0.8500053154693914,
"learning_rate": 7.937323358440935e-07,
"loss": 0.1498,
"step": 534
},
{
"epoch": 0.8278529980657641,
"grad_norm": 0.677471301203822,
"learning_rate": 7.802407485773011e-07,
"loss": 0.1046,
"step": 535
},
{
"epoch": 0.8294003868471953,
"grad_norm": 0.9020694266169602,
"learning_rate": 7.668550992274476e-07,
"loss": 0.165,
"step": 536
},
{
"epoch": 0.8309477756286267,
"grad_norm": 0.7889441624852148,
"learning_rate": 7.535757238439939e-07,
"loss": 0.1441,
"step": 537
},
{
"epoch": 0.832495164410058,
"grad_norm": 0.9159302142824934,
"learning_rate": 7.404029558083653e-07,
"loss": 0.1564,
"step": 538
},
{
"epoch": 0.8340425531914893,
"grad_norm": 0.7121558280600164,
"learning_rate": 7.273371258255923e-07,
"loss": 0.1323,
"step": 539
},
{
"epoch": 0.8355899419729207,
"grad_norm": 0.8753550993270109,
"learning_rate": 7.143785619160026e-07,
"loss": 0.1438,
"step": 540
},
{
"epoch": 0.837137330754352,
"grad_norm": 0.5776454455823687,
"learning_rate": 7.015275894069862e-07,
"loss": 0.1083,
"step": 541
},
{
"epoch": 0.8386847195357834,
"grad_norm": 0.7974232975808588,
"learning_rate": 6.887845309248326e-07,
"loss": 0.1524,
"step": 542
},
{
"epoch": 0.8402321083172147,
"grad_norm": 0.8055640503411741,
"learning_rate": 6.761497063866207e-07,
"loss": 0.1356,
"step": 543
},
{
"epoch": 0.8417794970986461,
"grad_norm": 0.7926822565874131,
"learning_rate": 6.63623432992202e-07,
"loss": 0.1663,
"step": 544
},
{
"epoch": 0.8433268858800773,
"grad_norm": 0.665841691728969,
"learning_rate": 6.512060252162228e-07,
"loss": 0.1295,
"step": 545
},
{
"epoch": 0.8448742746615087,
"grad_norm": 0.8002035859393191,
"learning_rate": 6.388977948002406e-07,
"loss": 0.1365,
"step": 546
},
{
"epoch": 0.84642166344294,
"grad_norm": 0.983616931359967,
"learning_rate": 6.2669905074489e-07,
"loss": 0.1275,
"step": 547
},
{
"epoch": 0.8479690522243714,
"grad_norm": 0.7163412457321918,
"learning_rate": 6.146100993021308e-07,
"loss": 0.1228,
"step": 548
},
{
"epoch": 0.8495164410058027,
"grad_norm": 0.7004721386032752,
"learning_rate": 6.026312439675553e-07,
"loss": 0.1175,
"step": 549
},
{
"epoch": 0.851063829787234,
"grad_norm": 0.8086290121377796,
"learning_rate": 5.907627854727688e-07,
"loss": 0.115,
"step": 550
},
{
"epoch": 0.8526112185686654,
"grad_norm": 0.669425870895728,
"learning_rate": 5.790050217778442e-07,
"loss": 0.1269,
"step": 551
},
{
"epoch": 0.8541586073500967,
"grad_norm": 1.0940372146333524,
"learning_rate": 5.673582480638395e-07,
"loss": 0.1507,
"step": 552
},
{
"epoch": 0.8557059961315281,
"grad_norm": 0.8719128895893627,
"learning_rate": 5.558227567253832e-07,
"loss": 0.1593,
"step": 553
},
{
"epoch": 0.8572533849129593,
"grad_norm": 0.9438030636727563,
"learning_rate": 5.443988373633397e-07,
"loss": 0.1297,
"step": 554
},
{
"epoch": 0.8588007736943907,
"grad_norm": 0.7000783588985214,
"learning_rate": 5.330867767775333e-07,
"loss": 0.1268,
"step": 555
},
{
"epoch": 0.860348162475822,
"grad_norm": 0.8609793200211151,
"learning_rate": 5.218868589595555e-07,
"loss": 0.1414,
"step": 556
},
{
"epoch": 0.8618955512572534,
"grad_norm": 1.0155712499379974,
"learning_rate": 5.107993650856285e-07,
"loss": 0.1524,
"step": 557
},
{
"epoch": 0.8634429400386847,
"grad_norm": 0.8449813630805493,
"learning_rate": 4.998245735095459e-07,
"loss": 0.1802,
"step": 558
},
{
"epoch": 0.8649903288201161,
"grad_norm": 0.8689521362622863,
"learning_rate": 4.889627597556911e-07,
"loss": 0.1757,
"step": 559
},
{
"epoch": 0.8665377176015474,
"grad_norm": 0.9659642011668181,
"learning_rate": 4.782141965121129e-07,
"loss": 0.1861,
"step": 560
},
{
"epoch": 0.8680851063829788,
"grad_norm": 0.6161184874791877,
"learning_rate": 4.6757915362368567e-07,
"loss": 0.1298,
"step": 561
},
{
"epoch": 0.8696324951644101,
"grad_norm": 0.7807500061793617,
"learning_rate": 4.570578980853302e-07,
"loss": 0.1555,
"step": 562
},
{
"epoch": 0.8711798839458413,
"grad_norm": 0.6384524907008211,
"learning_rate": 4.466506940353138e-07,
"loss": 0.1341,
"step": 563
},
{
"epoch": 0.8727272727272727,
"grad_norm": 0.7846304452771449,
"learning_rate": 4.363578027486187e-07,
"loss": 0.1144,
"step": 564
},
{
"epoch": 0.874274661508704,
"grad_norm": 0.8693314169579249,
"learning_rate": 4.261794826303783e-07,
"loss": 0.1677,
"step": 565
},
{
"epoch": 0.8758220502901354,
"grad_norm": 0.783633732727121,
"learning_rate": 4.16115989209398e-07,
"loss": 0.1209,
"step": 566
},
{
"epoch": 0.8773694390715667,
"grad_norm": 0.8133727971729119,
"learning_rate": 4.0616757513173123e-07,
"loss": 0.1234,
"step": 567
},
{
"epoch": 0.8789168278529981,
"grad_norm": 0.651161363160907,
"learning_rate": 3.963344901543437e-07,
"loss": 0.1326,
"step": 568
},
{
"epoch": 0.8804642166344294,
"grad_norm": 0.9183272389951808,
"learning_rate": 3.866169811388415e-07,
"loss": 0.1331,
"step": 569
},
{
"epoch": 0.8820116054158608,
"grad_norm": 0.6681577824466309,
"learning_rate": 3.7701529204526856e-07,
"loss": 0.1594,
"step": 570
},
{
"epoch": 0.8835589941972921,
"grad_norm": 0.6630502852605802,
"learning_rate": 3.675296639259912e-07,
"loss": 0.1024,
"step": 571
},
{
"epoch": 0.8851063829787233,
"grad_norm": 0.8400077472483749,
"learning_rate": 3.581603349196372e-07,
"loss": 0.1321,
"step": 572
},
{
"epoch": 0.8866537717601547,
"grad_norm": 0.8327913064778995,
"learning_rate": 3.4890754024512254e-07,
"loss": 0.1518,
"step": 573
},
{
"epoch": 0.888201160541586,
"grad_norm": 0.805383325394688,
"learning_rate": 3.397715121957468e-07,
"loss": 0.0982,
"step": 574
},
{
"epoch": 0.8897485493230174,
"grad_norm": 0.7824493921010083,
"learning_rate": 3.3075248013335614e-07,
"loss": 0.1467,
"step": 575
},
{
"epoch": 0.8912959381044487,
"grad_norm": 0.9935776351077749,
"learning_rate": 3.2185067048259245e-07,
"loss": 0.1656,
"step": 576
},
{
"epoch": 0.8928433268858801,
"grad_norm": 0.7543075174021723,
"learning_rate": 3.1306630672520153e-07,
"loss": 0.1558,
"step": 577
},
{
"epoch": 0.8943907156673114,
"grad_norm": 0.7947631048450559,
"learning_rate": 3.0439960939442794e-07,
"loss": 0.1689,
"step": 578
},
{
"epoch": 0.8959381044487428,
"grad_norm": 0.9958131657971071,
"learning_rate": 2.9585079606947843e-07,
"loss": 0.1565,
"step": 579
},
{
"epoch": 0.8974854932301741,
"grad_norm": 0.7484633943183348,
"learning_rate": 2.874200813700534e-07,
"loss": 0.1834,
"step": 580
},
{
"epoch": 0.8990328820116054,
"grad_norm": 0.8384863494297725,
"learning_rate": 2.7910767695096707e-07,
"loss": 0.1082,
"step": 581
},
{
"epoch": 0.9005802707930367,
"grad_norm": 0.8145685183591151,
"learning_rate": 2.7091379149682683e-07,
"loss": 0.1529,
"step": 582
},
{
"epoch": 0.902127659574468,
"grad_norm": 0.9387567574264674,
"learning_rate": 2.628386307167996e-07,
"loss": 0.148,
"step": 583
},
{
"epoch": 0.9036750483558994,
"grad_norm": 0.6587309545456876,
"learning_rate": 2.548823973394449e-07,
"loss": 0.1318,
"step": 584
},
{
"epoch": 0.9052224371373307,
"grad_norm": 0.8050044370132193,
"learning_rate": 2.470452911076227e-07,
"loss": 0.1631,
"step": 585
},
{
"epoch": 0.9067698259187621,
"grad_norm": 0.6762097120208355,
"learning_rate": 2.393275087734864e-07,
"loss": 0.1428,
"step": 586
},
{
"epoch": 0.9083172147001934,
"grad_norm": 0.8140602111082307,
"learning_rate": 2.317292440935348e-07,
"loss": 0.1464,
"step": 587
},
{
"epoch": 0.9098646034816248,
"grad_norm": 0.7338565136856471,
"learning_rate": 2.242506878237538e-07,
"loss": 0.1309,
"step": 588
},
{
"epoch": 0.9114119922630561,
"grad_norm": 0.6698213427174237,
"learning_rate": 2.1689202771482344e-07,
"loss": 0.1595,
"step": 589
},
{
"epoch": 0.9129593810444874,
"grad_norm": 0.6962632540688021,
"learning_rate": 2.0965344850740698e-07,
"loss": 0.1372,
"step": 590
},
{
"epoch": 0.9145067698259187,
"grad_norm": 0.7033044685938934,
"learning_rate": 2.0253513192751374e-07,
"loss": 0.1436,
"step": 591
},
{
"epoch": 0.9160541586073501,
"grad_norm": 0.8345980197859496,
"learning_rate": 1.9553725668193192e-07,
"loss": 0.1789,
"step": 592
},
{
"epoch": 0.9176015473887814,
"grad_norm": 0.7932512333003868,
"learning_rate": 1.8865999845374794e-07,
"loss": 0.1548,
"step": 593
},
{
"epoch": 0.9191489361702128,
"grad_norm": 0.6881043910649381,
"learning_rate": 1.8190352989793325e-07,
"loss": 0.1308,
"step": 594
},
{
"epoch": 0.9206963249516441,
"grad_norm": 0.8073364138943768,
"learning_rate": 1.7526802063700943e-07,
"loss": 0.1523,
"step": 595
},
{
"epoch": 0.9222437137330755,
"grad_norm": 0.7148043609511118,
"learning_rate": 1.6875363725679052e-07,
"loss": 0.152,
"step": 596
},
{
"epoch": 0.9237911025145068,
"grad_norm": 0.7982536840580516,
"learning_rate": 1.6236054330219853e-07,
"loss": 0.1401,
"step": 597
},
{
"epoch": 0.9253384912959381,
"grad_norm": 0.7580507948127416,
"learning_rate": 1.5608889927316407e-07,
"loss": 0.1603,
"step": 598
},
{
"epoch": 0.9268858800773694,
"grad_norm": 0.8050983190849471,
"learning_rate": 1.4993886262058833e-07,
"loss": 0.1534,
"step": 599
},
{
"epoch": 0.9284332688588007,
"grad_norm": 0.7584923418480568,
"learning_rate": 1.439105877423963e-07,
"loss": 0.1367,
"step": 600
},
{
"epoch": 0.9299806576402321,
"grad_norm": 0.8136537930412558,
"learning_rate": 1.3800422597965935e-07,
"loss": 0.15,
"step": 601
},
{
"epoch": 0.9315280464216634,
"grad_norm": 0.7915076935934267,
"learning_rate": 1.322199256127943e-07,
"loss": 0.1454,
"step": 602
},
{
"epoch": 0.9330754352030948,
"grad_norm": 0.8979276178770819,
"learning_rate": 1.2655783185784253e-07,
"loss": 0.149,
"step": 603
},
{
"epoch": 0.9346228239845261,
"grad_norm": 0.8649757965854281,
"learning_rate": 1.210180868628219e-07,
"loss": 0.184,
"step": 604
},
{
"epoch": 0.9361702127659575,
"grad_norm": 0.7748557564315666,
"learning_rate": 1.1560082970416164e-07,
"loss": 0.1856,
"step": 605
},
{
"epoch": 0.9377176015473888,
"grad_norm": 0.7206838191995956,
"learning_rate": 1.1030619638320805e-07,
"loss": 0.1387,
"step": 606
},
{
"epoch": 0.9392649903288202,
"grad_norm": 0.8779498348940245,
"learning_rate": 1.0513431982280997e-07,
"loss": 0.1876,
"step": 607
},
{
"epoch": 0.9408123791102514,
"grad_norm": 0.6653575747658279,
"learning_rate": 1.0008532986398422e-07,
"loss": 0.1537,
"step": 608
},
{
"epoch": 0.9423597678916827,
"grad_norm": 0.9253973454901049,
"learning_rate": 9.51593532626538e-08,
"loss": 0.127,
"step": 609
},
{
"epoch": 0.9439071566731141,
"grad_norm": 0.6917664278268542,
"learning_rate": 9.035651368646647e-08,
"loss": 0.1284,
"step": 610
},
{
"epoch": 0.9454545454545454,
"grad_norm": 0.7272286247644155,
"learning_rate": 8.567693171168956e-08,
"loss": 0.1307,
"step": 611
},
{
"epoch": 0.9470019342359768,
"grad_norm": 0.7852368165928318,
"learning_rate": 8.11207248201834e-08,
"loss": 0.1485,
"step": 612
},
{
"epoch": 0.9485493230174081,
"grad_norm": 0.6301070552387278,
"learning_rate": 7.6688007396451e-08,
"loss": 0.11,
"step": 613
},
{
"epoch": 0.9500967117988395,
"grad_norm": 0.608692902225253,
"learning_rate": 7.237889072476856e-08,
"loss": 0.1318,
"step": 614
},
{
"epoch": 0.9516441005802708,
"grad_norm": 0.7710321574963083,
"learning_rate": 6.819348298638839e-08,
"loss": 0.1413,
"step": 615
},
{
"epoch": 0.9531914893617022,
"grad_norm": 0.7161448867699187,
"learning_rate": 6.4131889256826e-08,
"loss": 0.1786,
"step": 616
},
{
"epoch": 0.9547388781431334,
"grad_norm": 0.7695049892987397,
"learning_rate": 6.019421150322114e-08,
"loss": 0.1416,
"step": 617
},
{
"epoch": 0.9562862669245648,
"grad_norm": 0.7639927969969345,
"learning_rate": 5.638054858177644e-08,
"loss": 0.1167,
"step": 618
},
{
"epoch": 0.9578336557059961,
"grad_norm": 0.8327927634444185,
"learning_rate": 5.26909962352784e-08,
"loss": 0.176,
"step": 619
},
{
"epoch": 0.9593810444874274,
"grad_norm": 0.7239532280928083,
"learning_rate": 4.91256470906909e-08,
"loss": 0.1558,
"step": 620
},
{
"epoch": 0.9609284332688588,
"grad_norm": 0.8410996947219226,
"learning_rate": 4.568459065683206e-08,
"loss": 0.1302,
"step": 621
},
{
"epoch": 0.9624758220502901,
"grad_norm": 0.7512248970367057,
"learning_rate": 4.236791332212498e-08,
"loss": 0.1523,
"step": 622
},
{
"epoch": 0.9640232108317215,
"grad_norm": 0.7793061388647213,
"learning_rate": 3.917569835243107e-08,
"loss": 0.14,
"step": 623
},
{
"epoch": 0.9655705996131528,
"grad_norm": 1.0664808213421917,
"learning_rate": 3.610802588895845e-08,
"loss": 0.1291,
"step": 624
},
{
"epoch": 0.9671179883945842,
"grad_norm": 0.900628675727633,
"learning_rate": 3.316497294625132e-08,
"loss": 0.1615,
"step": 625
},
{
"epoch": 0.9686653771760154,
"grad_norm": 0.717256932520607,
"learning_rate": 3.034661341025258e-08,
"loss": 0.1292,
"step": 626
},
{
"epoch": 0.9702127659574468,
"grad_norm": 0.8790080894597829,
"learning_rate": 2.765301803645426e-08,
"loss": 0.1464,
"step": 627
},
{
"epoch": 0.9717601547388781,
"grad_norm": 1.140431769064114,
"learning_rate": 2.5084254448117794e-08,
"loss": 0.1333,
"step": 628
},
{
"epoch": 0.9733075435203095,
"grad_norm": 0.7695085076315615,
"learning_rate": 2.264038713457706e-08,
"loss": 0.1645,
"step": 629
},
{
"epoch": 0.9748549323017408,
"grad_norm": 0.7415189162645879,
"learning_rate": 2.0321477449619098e-08,
"loss": 0.1568,
"step": 630
},
{
"epoch": 0.9764023210831722,
"grad_norm": 0.8220244447216026,
"learning_rate": 1.8127583609945376e-08,
"loss": 0.1546,
"step": 631
},
{
"epoch": 0.9779497098646035,
"grad_norm": 0.8679740522554501,
"learning_rate": 1.6058760693708487e-08,
"loss": 0.1487,
"step": 632
},
{
"epoch": 0.9794970986460348,
"grad_norm": 0.7017951731618077,
"learning_rate": 1.411506063912882e-08,
"loss": 0.1346,
"step": 633
},
{
"epoch": 0.9810444874274662,
"grad_norm": 0.9092830013578839,
"learning_rate": 1.2296532243193382e-08,
"loss": 0.1267,
"step": 634
},
{
"epoch": 0.9825918762088974,
"grad_norm": 0.9683538646565922,
"learning_rate": 1.0603221160429e-08,
"loss": 0.1846,
"step": 635
},
{
"epoch": 0.9841392649903288,
"grad_norm": 0.8999882709467414,
"learning_rate": 9.035169901754902e-09,
"loss": 0.1687,
"step": 636
},
{
"epoch": 0.9856866537717601,
"grad_norm": 0.8029769470497292,
"learning_rate": 7.59241783341913e-09,
"loss": 0.1608,
"step": 637
},
{
"epoch": 0.9872340425531915,
"grad_norm": 0.9217618497028801,
"learning_rate": 6.2750011760054355e-09,
"loss": 0.1793,
"step": 638
},
{
"epoch": 0.9887814313346228,
"grad_norm": 0.7827749530752652,
"learning_rate": 5.082953003528457e-09,
"loss": 0.1618,
"step": 639
},
{
"epoch": 0.9903288201160542,
"grad_norm": 0.7823004595031468,
"learning_rate": 4.016303242600495e-09,
"loss": 0.133,
"step": 640
},
{
"epoch": 0.9918762088974855,
"grad_norm": 0.7620908308906509,
"learning_rate": 3.075078671682108e-09,
"loss": 0.174,
"step": 641
},
{
"epoch": 0.9934235976789169,
"grad_norm": 0.8275398473313822,
"learning_rate": 2.2593029204076578e-09,
"loss": 0.1319,
"step": 642
},
{
"epoch": 0.9949709864603482,
"grad_norm": 1.0308241658181023,
"learning_rate": 1.5689964689935555e-09,
"loss": 0.1658,
"step": 643
},
{
"epoch": 0.9965183752417794,
"grad_norm": 0.9124286283429369,
"learning_rate": 1.004176647724231e-09,
"loss": 0.1458,
"step": 644
},
{
"epoch": 0.9980657640232108,
"grad_norm": 0.7481541000671525,
"learning_rate": 5.648576365169245e-10,
"loss": 0.1601,
"step": 645
},
{
"epoch": 0.9996131528046421,
"grad_norm": 0.7816235138889417,
"learning_rate": 2.5105046456475047e-10,
"loss": 0.1634,
"step": 646
},
{
"epoch": 1.0,
"grad_norm": 1.6239160001117061,
"learning_rate": 6.276301006080676e-11,
"loss": 0.1093,
"step": 647
}
],
"logging_steps": 1.0,
"max_steps": 647,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 957412566433792.0,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}