greyAll's picture
Upload folder using huggingface_hub
040bc65 verified
Raw
History Blame Contribute Delete
373 kB
{
"best_global_step": 2000,
"best_metric": 7.4517598152160645,
"best_model_checkpoint": "./models/v1/checkpoint-2000",
"epoch": 0.08040201005025126,
"eval_steps": 200,
"global_step": 2000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 4.020100502512563e-05,
"grad_norm": 10.698312759399414,
"learning_rate": 0.0,
"loss": 13.243154525756836,
"step": 1
},
{
"epoch": 8.040201005025126e-05,
"grad_norm": 10.833827018737793,
"learning_rate": 1.06e-06,
"loss": 13.252357482910156,
"step": 2
},
{
"epoch": 0.00012060301507537688,
"grad_norm": 10.492980003356934,
"learning_rate": 2.12e-06,
"loss": 13.270368576049805,
"step": 3
},
{
"epoch": 0.00016080402010050252,
"grad_norm": 10.906421661376953,
"learning_rate": 3.18e-06,
"loss": 13.24106502532959,
"step": 4
},
{
"epoch": 0.00020100502512562814,
"grad_norm": 11.034916877746582,
"learning_rate": 4.24e-06,
"loss": 13.248748779296875,
"step": 5
},
{
"epoch": 0.00024120603015075377,
"grad_norm": 10.556979179382324,
"learning_rate": 5.3e-06,
"loss": 13.235283851623535,
"step": 6
},
{
"epoch": 0.0002814070351758794,
"grad_norm": 10.435272216796875,
"learning_rate": 6.36e-06,
"loss": 13.227590560913086,
"step": 7
},
{
"epoch": 0.00032160804020100504,
"grad_norm": 10.702980995178223,
"learning_rate": 7.420000000000001e-06,
"loss": 13.19597053527832,
"step": 8
},
{
"epoch": 0.00036180904522613067,
"grad_norm": 11.167304039001465,
"learning_rate": 8.48e-06,
"loss": 13.165201187133789,
"step": 9
},
{
"epoch": 0.0004020100502512563,
"grad_norm": 10.859643936157227,
"learning_rate": 9.54e-06,
"loss": 13.08542251586914,
"step": 10
},
{
"epoch": 0.0004422110552763819,
"grad_norm": 11.573236465454102,
"learning_rate": 1.06e-05,
"loss": 12.977386474609375,
"step": 11
},
{
"epoch": 0.00048241206030150754,
"grad_norm": 11.529393196105957,
"learning_rate": 1.166e-05,
"loss": 12.960195541381836,
"step": 12
},
{
"epoch": 0.0005226130653266332,
"grad_norm": 12.054129600524902,
"learning_rate": 1.272e-05,
"loss": 12.788475036621094,
"step": 13
},
{
"epoch": 0.0005628140703517588,
"grad_norm": 12.061715126037598,
"learning_rate": 1.378e-05,
"loss": 12.758974075317383,
"step": 14
},
{
"epoch": 0.0006030150753768845,
"grad_norm": 13.375988960266113,
"learning_rate": 1.4840000000000002e-05,
"loss": 12.492585182189941,
"step": 15
},
{
"epoch": 0.0006432160804020101,
"grad_norm": 13.963891983032227,
"learning_rate": 1.59e-05,
"loss": 12.240091323852539,
"step": 16
},
{
"epoch": 0.0006834170854271357,
"grad_norm": 14.665262222290039,
"learning_rate": 1.696e-05,
"loss": 12.143289566040039,
"step": 17
},
{
"epoch": 0.0007236180904522613,
"grad_norm": 13.834787368774414,
"learning_rate": 1.802e-05,
"loss": 11.920969009399414,
"step": 18
},
{
"epoch": 0.000763819095477387,
"grad_norm": 15.71484088897705,
"learning_rate": 1.908e-05,
"loss": 11.746869087219238,
"step": 19
},
{
"epoch": 0.0008040201005025126,
"grad_norm": 37.37236785888672,
"learning_rate": 2.014e-05,
"loss": 11.95899772644043,
"step": 20
},
{
"epoch": 0.0008442211055276382,
"grad_norm": 31.74503517150879,
"learning_rate": 2.12e-05,
"loss": 11.544794082641602,
"step": 21
},
{
"epoch": 0.0008844221105527638,
"grad_norm": 15.400463104248047,
"learning_rate": 2.226e-05,
"loss": 11.268622398376465,
"step": 22
},
{
"epoch": 0.0009246231155778894,
"grad_norm": 15.651389122009277,
"learning_rate": 2.332e-05,
"loss": 11.069677352905273,
"step": 23
},
{
"epoch": 0.0009648241206030151,
"grad_norm": 13.487519264221191,
"learning_rate": 2.438e-05,
"loss": 10.997288703918457,
"step": 24
},
{
"epoch": 0.0010050251256281408,
"grad_norm": 12.921786308288574,
"learning_rate": 2.544e-05,
"loss": 10.803884506225586,
"step": 25
},
{
"epoch": 0.0010452261306532664,
"grad_norm": 16.199108123779297,
"learning_rate": 2.65e-05,
"loss": 11.145416259765625,
"step": 26
},
{
"epoch": 0.001085427135678392,
"grad_norm": 12.436396598815918,
"learning_rate": 2.756e-05,
"loss": 10.444547653198242,
"step": 27
},
{
"epoch": 0.0011256281407035177,
"grad_norm": 9.46310043334961,
"learning_rate": 2.8620000000000004e-05,
"loss": 10.39027214050293,
"step": 28
},
{
"epoch": 0.0011658291457286433,
"grad_norm": 11.567558288574219,
"learning_rate": 2.9680000000000004e-05,
"loss": 10.297073364257812,
"step": 29
},
{
"epoch": 0.001206030150753769,
"grad_norm": 15.577902793884277,
"learning_rate": 3.074e-05,
"loss": 10.459238052368164,
"step": 30
},
{
"epoch": 0.0012462311557788945,
"grad_norm": 8.467581748962402,
"learning_rate": 3.18e-05,
"loss": 10.274847984313965,
"step": 31
},
{
"epoch": 0.0012864321608040202,
"grad_norm": 7.366700649261475,
"learning_rate": 3.286e-05,
"loss": 10.238689422607422,
"step": 32
},
{
"epoch": 0.0013266331658291458,
"grad_norm": 6.314115047454834,
"learning_rate": 3.392e-05,
"loss": 10.226795196533203,
"step": 33
},
{
"epoch": 0.0013668341708542714,
"grad_norm": 5.678335189819336,
"learning_rate": 3.498e-05,
"loss": 9.992923736572266,
"step": 34
},
{
"epoch": 0.001407035175879397,
"grad_norm": 6.734803676605225,
"learning_rate": 3.604e-05,
"loss": 9.895065307617188,
"step": 35
},
{
"epoch": 0.0014472361809045227,
"grad_norm": 4.478935718536377,
"learning_rate": 3.71e-05,
"loss": 10.028186798095703,
"step": 36
},
{
"epoch": 0.0014874371859296483,
"grad_norm": 3.503633499145508,
"learning_rate": 3.816e-05,
"loss": 9.50555419921875,
"step": 37
},
{
"epoch": 0.001527638190954774,
"grad_norm": 5.015214443206787,
"learning_rate": 3.922e-05,
"loss": 9.550050735473633,
"step": 38
},
{
"epoch": 0.0015678391959798995,
"grad_norm": 3.1059823036193848,
"learning_rate": 4.028e-05,
"loss": 9.667120933532715,
"step": 39
},
{
"epoch": 0.0016080402010050252,
"grad_norm": 9.8824462890625,
"learning_rate": 4.134e-05,
"loss": 9.853014945983887,
"step": 40
},
{
"epoch": 0.0016482412060301508,
"grad_norm": 4.625787258148193,
"learning_rate": 4.24e-05,
"loss": 9.513239860534668,
"step": 41
},
{
"epoch": 0.0016884422110552764,
"grad_norm": 3.7364585399627686,
"learning_rate": 4.346e-05,
"loss": 9.345815658569336,
"step": 42
},
{
"epoch": 0.001728643216080402,
"grad_norm": 3.5209295749664307,
"learning_rate": 4.452e-05,
"loss": 9.764396667480469,
"step": 43
},
{
"epoch": 0.0017688442211055277,
"grad_norm": 2.695601224899292,
"learning_rate": 4.558e-05,
"loss": 9.510019302368164,
"step": 44
},
{
"epoch": 0.0018090452261306533,
"grad_norm": 4.428959369659424,
"learning_rate": 4.664e-05,
"loss": 9.288888931274414,
"step": 45
},
{
"epoch": 0.001849246231155779,
"grad_norm": 5.149900436401367,
"learning_rate": 4.77e-05,
"loss": 9.208829879760742,
"step": 46
},
{
"epoch": 0.0018894472361809045,
"grad_norm": 3.0757558345794678,
"learning_rate": 4.876e-05,
"loss": 9.675956726074219,
"step": 47
},
{
"epoch": 0.0019296482412060301,
"grad_norm": 2.372467517852783,
"learning_rate": 4.982e-05,
"loss": 9.11758804321289,
"step": 48
},
{
"epoch": 0.0019698492462311558,
"grad_norm": 2.902724504470825,
"learning_rate": 5.088e-05,
"loss": 9.610967636108398,
"step": 49
},
{
"epoch": 0.0020100502512562816,
"grad_norm": 2.850295066833496,
"learning_rate": 5.194e-05,
"loss": 9.49056339263916,
"step": 50
},
{
"epoch": 0.002050251256281407,
"grad_norm": 3.8998680114746094,
"learning_rate": 5.3e-05,
"loss": 9.381105422973633,
"step": 51
},
{
"epoch": 0.002090452261306533,
"grad_norm": 3.256573438644409,
"learning_rate": 5.299999989390202e-05,
"loss": 9.116067886352539,
"step": 52
},
{
"epoch": 0.0021306532663316583,
"grad_norm": 2.5064449310302734,
"learning_rate": 5.299999957560809e-05,
"loss": 9.005221366882324,
"step": 53
},
{
"epoch": 0.002170854271356784,
"grad_norm": 4.868945598602295,
"learning_rate": 5.29999990451182e-05,
"loss": 9.194267272949219,
"step": 54
},
{
"epoch": 0.0022110552763819095,
"grad_norm": 3.177849054336548,
"learning_rate": 5.2999998302432364e-05,
"loss": 9.23127269744873,
"step": 55
},
{
"epoch": 0.0022512562814070354,
"grad_norm": 4.279708385467529,
"learning_rate": 5.2999997347550594e-05,
"loss": 9.402137756347656,
"step": 56
},
{
"epoch": 0.0022914572864321608,
"grad_norm": 2.376249313354492,
"learning_rate": 5.2999996180472915e-05,
"loss": 9.278661727905273,
"step": 57
},
{
"epoch": 0.0023316582914572866,
"grad_norm": 4.11171817779541,
"learning_rate": 5.299999480119934e-05,
"loss": 9.144866943359375,
"step": 58
},
{
"epoch": 0.002371859296482412,
"grad_norm": 2.8840560913085938,
"learning_rate": 5.299999320972989e-05,
"loss": 9.44858169555664,
"step": 59
},
{
"epoch": 0.002412060301507538,
"grad_norm": 2.668494462966919,
"learning_rate": 5.299999140606459e-05,
"loss": 9.257560729980469,
"step": 60
},
{
"epoch": 0.0024522613065326633,
"grad_norm": 3.142766237258911,
"learning_rate": 5.299998939020346e-05,
"loss": 9.142229080200195,
"step": 61
},
{
"epoch": 0.002492462311557789,
"grad_norm": 3.182288408279419,
"learning_rate": 5.2999987162146547e-05,
"loss": 9.041769027709961,
"step": 62
},
{
"epoch": 0.0025326633165829145,
"grad_norm": 3.708120107650757,
"learning_rate": 5.299998472189388e-05,
"loss": 9.176864624023438,
"step": 63
},
{
"epoch": 0.0025728643216080403,
"grad_norm": 2.8155362606048584,
"learning_rate": 5.29999820694455e-05,
"loss": 9.12497615814209,
"step": 64
},
{
"epoch": 0.0026130653266331657,
"grad_norm": 2.912419557571411,
"learning_rate": 5.2999979204801443e-05,
"loss": 9.254022598266602,
"step": 65
},
{
"epoch": 0.0026532663316582916,
"grad_norm": 3.0874767303466797,
"learning_rate": 5.2999976127961764e-05,
"loss": 9.046329498291016,
"step": 66
},
{
"epoch": 0.002693467336683417,
"grad_norm": 2.875800132751465,
"learning_rate": 5.2999972838926514e-05,
"loss": 9.161742210388184,
"step": 67
},
{
"epoch": 0.002733668341708543,
"grad_norm": 3.0868020057678223,
"learning_rate": 5.2999969337695735e-05,
"loss": 9.201602935791016,
"step": 68
},
{
"epoch": 0.0027738693467336682,
"grad_norm": 2.8199493885040283,
"learning_rate": 5.299996562426949e-05,
"loss": 9.142997741699219,
"step": 69
},
{
"epoch": 0.002814070351758794,
"grad_norm": 2.773752212524414,
"learning_rate": 5.2999961698647826e-05,
"loss": 9.117706298828125,
"step": 70
},
{
"epoch": 0.0028542713567839195,
"grad_norm": 2.6835477352142334,
"learning_rate": 5.299995756083083e-05,
"loss": 9.120769500732422,
"step": 71
},
{
"epoch": 0.0028944723618090453,
"grad_norm": 2.487560272216797,
"learning_rate": 5.299995321081855e-05,
"loss": 8.982416152954102,
"step": 72
},
{
"epoch": 0.0029346733668341707,
"grad_norm": 3.4093680381774902,
"learning_rate": 5.299994864861106e-05,
"loss": 8.667570114135742,
"step": 73
},
{
"epoch": 0.0029748743718592966,
"grad_norm": 2.566619873046875,
"learning_rate": 5.299994387420843e-05,
"loss": 8.929250717163086,
"step": 74
},
{
"epoch": 0.003015075376884422,
"grad_norm": 2.5003082752227783,
"learning_rate": 5.299993888761075e-05,
"loss": 8.949346542358398,
"step": 75
},
{
"epoch": 0.003055276381909548,
"grad_norm": 2.826795816421509,
"learning_rate": 5.2999933688818086e-05,
"loss": 8.931376457214355,
"step": 76
},
{
"epoch": 0.0030954773869346732,
"grad_norm": 2.406074285507202,
"learning_rate": 5.2999928277830524e-05,
"loss": 9.182083129882812,
"step": 77
},
{
"epoch": 0.003135678391959799,
"grad_norm": 2.5750491619110107,
"learning_rate": 5.299992265464815e-05,
"loss": 8.907024383544922,
"step": 78
},
{
"epoch": 0.0031758793969849245,
"grad_norm": 3.028383731842041,
"learning_rate": 5.2999916819271054e-05,
"loss": 8.95583724975586,
"step": 79
},
{
"epoch": 0.0032160804020100503,
"grad_norm": 4.024677753448486,
"learning_rate": 5.299991077169933e-05,
"loss": 8.952323913574219,
"step": 80
},
{
"epoch": 0.0032562814070351757,
"grad_norm": 3.742661237716675,
"learning_rate": 5.2999904511933084e-05,
"loss": 8.674503326416016,
"step": 81
},
{
"epoch": 0.0032964824120603016,
"grad_norm": 2.423227071762085,
"learning_rate": 5.299989803997241e-05,
"loss": 9.127151489257812,
"step": 82
},
{
"epoch": 0.003336683417085427,
"grad_norm": 3.1274101734161377,
"learning_rate": 5.29998913558174e-05,
"loss": 9.033180236816406,
"step": 83
},
{
"epoch": 0.003376884422110553,
"grad_norm": 3.348336696624756,
"learning_rate": 5.2999884459468174e-05,
"loss": 8.972601890563965,
"step": 84
},
{
"epoch": 0.003417085427135678,
"grad_norm": 2.69063138961792,
"learning_rate": 5.2999877350924846e-05,
"loss": 8.862074851989746,
"step": 85
},
{
"epoch": 0.003457286432160804,
"grad_norm": 2.2713816165924072,
"learning_rate": 5.2999870030187514e-05,
"loss": 9.348979949951172,
"step": 86
},
{
"epoch": 0.0034974874371859295,
"grad_norm": 2.406306028366089,
"learning_rate": 5.299986249725631e-05,
"loss": 9.109790802001953,
"step": 87
},
{
"epoch": 0.0035376884422110553,
"grad_norm": 2.349501848220825,
"learning_rate": 5.299985475213135e-05,
"loss": 8.928289413452148,
"step": 88
},
{
"epoch": 0.0035778894472361807,
"grad_norm": 2.389651298522949,
"learning_rate": 5.299984679481275e-05,
"loss": 8.949289321899414,
"step": 89
},
{
"epoch": 0.0036180904522613066,
"grad_norm": 2.4067068099975586,
"learning_rate": 5.299983862530065e-05,
"loss": 8.876733779907227,
"step": 90
},
{
"epoch": 0.003658291457286432,
"grad_norm": 3.047351121902466,
"learning_rate": 5.299983024359518e-05,
"loss": 8.941215515136719,
"step": 91
},
{
"epoch": 0.003698492462311558,
"grad_norm": 2.2273101806640625,
"learning_rate": 5.299982164969647e-05,
"loss": 8.716243743896484,
"step": 92
},
{
"epoch": 0.003738693467336683,
"grad_norm": 2.53912091255188,
"learning_rate": 5.299981284360465e-05,
"loss": 8.831758499145508,
"step": 93
},
{
"epoch": 0.003778894472361809,
"grad_norm": 2.3589587211608887,
"learning_rate": 5.299980382531987e-05,
"loss": 8.778205871582031,
"step": 94
},
{
"epoch": 0.003819095477386935,
"grad_norm": 2.026813507080078,
"learning_rate": 5.2999794594842274e-05,
"loss": 8.871915817260742,
"step": 95
},
{
"epoch": 0.0038592964824120603,
"grad_norm": 2.034069538116455,
"learning_rate": 5.299978515217201e-05,
"loss": 8.691679000854492,
"step": 96
},
{
"epoch": 0.003899497487437186,
"grad_norm": 1.90824556350708,
"learning_rate": 5.2999775497309224e-05,
"loss": 9.13762092590332,
"step": 97
},
{
"epoch": 0.0039396984924623115,
"grad_norm": 2.742013454437256,
"learning_rate": 5.299976563025408e-05,
"loss": 8.668919563293457,
"step": 98
},
{
"epoch": 0.003979899497487437,
"grad_norm": 2.3603296279907227,
"learning_rate": 5.299975555100673e-05,
"loss": 8.995302200317383,
"step": 99
},
{
"epoch": 0.004020100502512563,
"grad_norm": 2.7155072689056396,
"learning_rate": 5.299974525956733e-05,
"loss": 8.741050720214844,
"step": 100
},
{
"epoch": 0.004060301507537689,
"grad_norm": 3.2940938472747803,
"learning_rate": 5.299973475593606e-05,
"loss": 8.532285690307617,
"step": 101
},
{
"epoch": 0.004100502512562814,
"grad_norm": 3.335320234298706,
"learning_rate": 5.299972404011307e-05,
"loss": 8.923404693603516,
"step": 102
},
{
"epoch": 0.0041407035175879394,
"grad_norm": 2.5615580081939697,
"learning_rate": 5.2999713112098545e-05,
"loss": 8.5745849609375,
"step": 103
},
{
"epoch": 0.004180904522613066,
"grad_norm": 2.5233376026153564,
"learning_rate": 5.2999701971892654e-05,
"loss": 8.632506370544434,
"step": 104
},
{
"epoch": 0.004221105527638191,
"grad_norm": 2.5961952209472656,
"learning_rate": 5.299969061949558e-05,
"loss": 8.639471054077148,
"step": 105
},
{
"epoch": 0.0042613065326633165,
"grad_norm": 1.978849172592163,
"learning_rate": 5.29996790549075e-05,
"loss": 9.081422805786133,
"step": 106
},
{
"epoch": 0.004301507537688442,
"grad_norm": 3.1976349353790283,
"learning_rate": 5.29996672781286e-05,
"loss": 8.629322052001953,
"step": 107
},
{
"epoch": 0.004341708542713568,
"grad_norm": 2.2551276683807373,
"learning_rate": 5.299965528915907e-05,
"loss": 8.772149085998535,
"step": 108
},
{
"epoch": 0.004381909547738694,
"grad_norm": 2.0938005447387695,
"learning_rate": 5.29996430879991e-05,
"loss": 8.297810554504395,
"step": 109
},
{
"epoch": 0.004422110552763819,
"grad_norm": 2.6357421875,
"learning_rate": 5.299963067464889e-05,
"loss": 8.706300735473633,
"step": 110
},
{
"epoch": 0.004462311557788944,
"grad_norm": 1.9622968435287476,
"learning_rate": 5.2999618049108635e-05,
"loss": 8.932507514953613,
"step": 111
},
{
"epoch": 0.004502512562814071,
"grad_norm": 2.778557777404785,
"learning_rate": 5.2999605211378536e-05,
"loss": 8.337654113769531,
"step": 112
},
{
"epoch": 0.004542713567839196,
"grad_norm": 2.7436506748199463,
"learning_rate": 5.29995921614588e-05,
"loss": 8.763957977294922,
"step": 113
},
{
"epoch": 0.0045829145728643215,
"grad_norm": 2.3577396869659424,
"learning_rate": 5.299957889934964e-05,
"loss": 8.704797744750977,
"step": 114
},
{
"epoch": 0.004623115577889447,
"grad_norm": 2.347414493560791,
"learning_rate": 5.299956542505126e-05,
"loss": 8.774009704589844,
"step": 115
},
{
"epoch": 0.004663316582914573,
"grad_norm": 1.8697577714920044,
"learning_rate": 5.2999551738563886e-05,
"loss": 8.783985137939453,
"step": 116
},
{
"epoch": 0.004703517587939699,
"grad_norm": 2.382254123687744,
"learning_rate": 5.299953783988773e-05,
"loss": 8.975433349609375,
"step": 117
},
{
"epoch": 0.004743718592964824,
"grad_norm": 2.5442185401916504,
"learning_rate": 5.299952372902302e-05,
"loss": 8.581441879272461,
"step": 118
},
{
"epoch": 0.004783919597989949,
"grad_norm": 2.4549672603607178,
"learning_rate": 5.299950940596998e-05,
"loss": 8.494630813598633,
"step": 119
},
{
"epoch": 0.004824120603015076,
"grad_norm": 2.9849038124084473,
"learning_rate": 5.299949487072883e-05,
"loss": 8.716537475585938,
"step": 120
},
{
"epoch": 0.004864321608040201,
"grad_norm": 3.1053988933563232,
"learning_rate": 5.299948012329981e-05,
"loss": 8.817995071411133,
"step": 121
},
{
"epoch": 0.0049045226130653265,
"grad_norm": 2.2433254718780518,
"learning_rate": 5.2999465163683165e-05,
"loss": 8.574687957763672,
"step": 122
},
{
"epoch": 0.004944723618090452,
"grad_norm": 2.234553813934326,
"learning_rate": 5.2999449991879125e-05,
"loss": 8.481681823730469,
"step": 123
},
{
"epoch": 0.004984924623115578,
"grad_norm": 2.200068950653076,
"learning_rate": 5.299943460788793e-05,
"loss": 8.685455322265625,
"step": 124
},
{
"epoch": 0.005025125628140704,
"grad_norm": 2.2843010425567627,
"learning_rate": 5.2999419011709836e-05,
"loss": 8.625558853149414,
"step": 125
},
{
"epoch": 0.005065326633165829,
"grad_norm": 2.4362471103668213,
"learning_rate": 5.2999403203345086e-05,
"loss": 8.323168754577637,
"step": 126
},
{
"epoch": 0.005105527638190954,
"grad_norm": 2.688843250274658,
"learning_rate": 5.299938718279393e-05,
"loss": 9.010099411010742,
"step": 127
},
{
"epoch": 0.005145728643216081,
"grad_norm": 2.1960902214050293,
"learning_rate": 5.2999370950056634e-05,
"loss": 8.792939186096191,
"step": 128
},
{
"epoch": 0.005185929648241206,
"grad_norm": 2.7818124294281006,
"learning_rate": 5.299935450513345e-05,
"loss": 8.51486587524414,
"step": 129
},
{
"epoch": 0.0052261306532663315,
"grad_norm": 2.455599546432495,
"learning_rate": 5.299933784802465e-05,
"loss": 9.111705780029297,
"step": 130
},
{
"epoch": 0.005266331658291457,
"grad_norm": 2.3506546020507812,
"learning_rate": 5.299932097873049e-05,
"loss": 8.474384307861328,
"step": 131
},
{
"epoch": 0.005306532663316583,
"grad_norm": 2.4325549602508545,
"learning_rate": 5.299930389725125e-05,
"loss": 8.906831741333008,
"step": 132
},
{
"epoch": 0.005346733668341709,
"grad_norm": 2.187574625015259,
"learning_rate": 5.299928660358719e-05,
"loss": 8.34807014465332,
"step": 133
},
{
"epoch": 0.005386934673366834,
"grad_norm": 2.2609736919403076,
"learning_rate": 5.2999269097738604e-05,
"loss": 8.463560104370117,
"step": 134
},
{
"epoch": 0.005427135678391959,
"grad_norm": 2.546436309814453,
"learning_rate": 5.2999251379705766e-05,
"loss": 8.604063034057617,
"step": 135
},
{
"epoch": 0.005467336683417086,
"grad_norm": 2.3433783054351807,
"learning_rate": 5.299923344948895e-05,
"loss": 8.831417083740234,
"step": 136
},
{
"epoch": 0.005507537688442211,
"grad_norm": 2.2846438884735107,
"learning_rate": 5.2999215307088454e-05,
"loss": 9.079967498779297,
"step": 137
},
{
"epoch": 0.0055477386934673365,
"grad_norm": 2.6954803466796875,
"learning_rate": 5.299919695250457e-05,
"loss": 8.295350074768066,
"step": 138
},
{
"epoch": 0.005587939698492462,
"grad_norm": 2.4904870986938477,
"learning_rate": 5.299917838573759e-05,
"loss": 8.474140167236328,
"step": 139
},
{
"epoch": 0.005628140703517588,
"grad_norm": 2.1757781505584717,
"learning_rate": 5.299915960678781e-05,
"loss": 8.431953430175781,
"step": 140
},
{
"epoch": 0.005668341708542714,
"grad_norm": 2.0644266605377197,
"learning_rate": 5.2999140615655525e-05,
"loss": 8.485315322875977,
"step": 141
},
{
"epoch": 0.005708542713567839,
"grad_norm": 2.278282403945923,
"learning_rate": 5.2999121412341036e-05,
"loss": 8.708311080932617,
"step": 142
},
{
"epoch": 0.005748743718592965,
"grad_norm": 2.780579090118408,
"learning_rate": 5.299910199684467e-05,
"loss": 8.8798828125,
"step": 143
},
{
"epoch": 0.005788944723618091,
"grad_norm": 1.9500168561935425,
"learning_rate": 5.299908236916672e-05,
"loss": 8.445011138916016,
"step": 144
},
{
"epoch": 0.005829145728643216,
"grad_norm": 1.8841841220855713,
"learning_rate": 5.299906252930751e-05,
"loss": 8.678537368774414,
"step": 145
},
{
"epoch": 0.0058693467336683415,
"grad_norm": 2.256147861480713,
"learning_rate": 5.2999042477267355e-05,
"loss": 8.45221996307373,
"step": 146
},
{
"epoch": 0.005909547738693468,
"grad_norm": 2.1743807792663574,
"learning_rate": 5.299902221304657e-05,
"loss": 8.58210563659668,
"step": 147
},
{
"epoch": 0.005949748743718593,
"grad_norm": 3.1546125411987305,
"learning_rate": 5.2999001736645494e-05,
"loss": 9.070743560791016,
"step": 148
},
{
"epoch": 0.0059899497487437186,
"grad_norm": 2.5418052673339844,
"learning_rate": 5.2998981048064444e-05,
"loss": 8.360685348510742,
"step": 149
},
{
"epoch": 0.006030150753768844,
"grad_norm": 1.8838720321655273,
"learning_rate": 5.2998960147303755e-05,
"loss": 8.189798355102539,
"step": 150
},
{
"epoch": 0.00607035175879397,
"grad_norm": 1.7609400749206543,
"learning_rate": 5.299893903436376e-05,
"loss": 8.273468017578125,
"step": 151
},
{
"epoch": 0.006110552763819096,
"grad_norm": 2.0566394329071045,
"learning_rate": 5.299891770924479e-05,
"loss": 8.30922794342041,
"step": 152
},
{
"epoch": 0.006150753768844221,
"grad_norm": 1.9362713098526,
"learning_rate": 5.2998896171947194e-05,
"loss": 8.247034072875977,
"step": 153
},
{
"epoch": 0.0061909547738693465,
"grad_norm": 2.0458033084869385,
"learning_rate": 5.2998874422471324e-05,
"loss": 8.2176513671875,
"step": 154
},
{
"epoch": 0.006231155778894473,
"grad_norm": 2.8741953372955322,
"learning_rate": 5.299885246081752e-05,
"loss": 7.951559066772461,
"step": 155
},
{
"epoch": 0.006271356783919598,
"grad_norm": 2.508368730545044,
"learning_rate": 5.2998830286986127e-05,
"loss": 8.744426727294922,
"step": 156
},
{
"epoch": 0.0063115577889447235,
"grad_norm": 2.6312766075134277,
"learning_rate": 5.299880790097751e-05,
"loss": 8.038535118103027,
"step": 157
},
{
"epoch": 0.006351758793969849,
"grad_norm": 2.4816057682037354,
"learning_rate": 5.2998785302792036e-05,
"loss": 8.283016204833984,
"step": 158
},
{
"epoch": 0.006391959798994975,
"grad_norm": 2.113785743713379,
"learning_rate": 5.299876249243005e-05,
"loss": 8.518077850341797,
"step": 159
},
{
"epoch": 0.006432160804020101,
"grad_norm": 2.7609400749206543,
"learning_rate": 5.2998739469891925e-05,
"loss": 8.61593246459961,
"step": 160
},
{
"epoch": 0.006472361809045226,
"grad_norm": 2.3244006633758545,
"learning_rate": 5.299871623517802e-05,
"loss": 8.297124862670898,
"step": 161
},
{
"epoch": 0.0065125628140703514,
"grad_norm": 2.3219308853149414,
"learning_rate": 5.299869278828873e-05,
"loss": 8.312850952148438,
"step": 162
},
{
"epoch": 0.006552763819095478,
"grad_norm": 2.4972732067108154,
"learning_rate": 5.2998669129224397e-05,
"loss": 8.240792274475098,
"step": 163
},
{
"epoch": 0.006592964824120603,
"grad_norm": 2.0999457836151123,
"learning_rate": 5.2998645257985433e-05,
"loss": 8.82748794555664,
"step": 164
},
{
"epoch": 0.0066331658291457285,
"grad_norm": 1.7659859657287598,
"learning_rate": 5.2998621174572205e-05,
"loss": 8.489943504333496,
"step": 165
},
{
"epoch": 0.006673366834170854,
"grad_norm": 2.2394418716430664,
"learning_rate": 5.29985968789851e-05,
"loss": 8.557156562805176,
"step": 166
},
{
"epoch": 0.00671356783919598,
"grad_norm": 1.9741127490997314,
"learning_rate": 5.2998572371224496e-05,
"loss": 8.32297420501709,
"step": 167
},
{
"epoch": 0.006753768844221106,
"grad_norm": 1.9758837223052979,
"learning_rate": 5.29985476512908e-05,
"loss": 8.531097412109375,
"step": 168
},
{
"epoch": 0.006793969849246231,
"grad_norm": 2.364895820617676,
"learning_rate": 5.2998522719184414e-05,
"loss": 8.311891555786133,
"step": 169
},
{
"epoch": 0.006834170854271356,
"grad_norm": 1.867857575416565,
"learning_rate": 5.299849757490572e-05,
"loss": 8.26534366607666,
"step": 170
},
{
"epoch": 0.006874371859296483,
"grad_norm": 7.186343193054199,
"learning_rate": 5.299847221845513e-05,
"loss": 8.461291313171387,
"step": 171
},
{
"epoch": 0.006914572864321608,
"grad_norm": 2.332974910736084,
"learning_rate": 5.299844664983305e-05,
"loss": 8.5230073928833,
"step": 172
},
{
"epoch": 0.0069547738693467335,
"grad_norm": 2.096611499786377,
"learning_rate": 5.2998420869039885e-05,
"loss": 8.383296966552734,
"step": 173
},
{
"epoch": 0.006994974874371859,
"grad_norm": 1.5418765544891357,
"learning_rate": 5.2998394876076055e-05,
"loss": 8.604005813598633,
"step": 174
},
{
"epoch": 0.007035175879396985,
"grad_norm": 2.050821304321289,
"learning_rate": 5.299836867094196e-05,
"loss": 8.449552536010742,
"step": 175
},
{
"epoch": 0.007075376884422111,
"grad_norm": 2.6503400802612305,
"learning_rate": 5.2998342253638045e-05,
"loss": 8.596059799194336,
"step": 176
},
{
"epoch": 0.007115577889447236,
"grad_norm": 2.177652597427368,
"learning_rate": 5.299831562416471e-05,
"loss": 8.140191078186035,
"step": 177
},
{
"epoch": 0.007155778894472361,
"grad_norm": 2.3540821075439453,
"learning_rate": 5.29982887825224e-05,
"loss": 8.5722074508667,
"step": 178
},
{
"epoch": 0.007195979899497488,
"grad_norm": 1.6604373455047607,
"learning_rate": 5.299826172871153e-05,
"loss": 8.492136001586914,
"step": 179
},
{
"epoch": 0.007236180904522613,
"grad_norm": 2.2135276794433594,
"learning_rate": 5.2998234462732544e-05,
"loss": 8.264391899108887,
"step": 180
},
{
"epoch": 0.0072763819095477385,
"grad_norm": 2.8494114875793457,
"learning_rate": 5.2998206984585867e-05,
"loss": 8.249712944030762,
"step": 181
},
{
"epoch": 0.007316582914572864,
"grad_norm": 2.6122653484344482,
"learning_rate": 5.299817929427195e-05,
"loss": 8.256064414978027,
"step": 182
},
{
"epoch": 0.00735678391959799,
"grad_norm": 2.1635890007019043,
"learning_rate": 5.299815139179123e-05,
"loss": 8.299365997314453,
"step": 183
},
{
"epoch": 0.007396984924623116,
"grad_norm": 2.17738938331604,
"learning_rate": 5.299812327714416e-05,
"loss": 8.182544708251953,
"step": 184
},
{
"epoch": 0.007437185929648241,
"grad_norm": 1.9491832256317139,
"learning_rate": 5.299809495033119e-05,
"loss": 8.568967819213867,
"step": 185
},
{
"epoch": 0.007477386934673366,
"grad_norm": 1.8599575757980347,
"learning_rate": 5.299806641135276e-05,
"loss": 8.753403663635254,
"step": 186
},
{
"epoch": 0.007517587939698493,
"grad_norm": 2.137356996536255,
"learning_rate": 5.2998037660209345e-05,
"loss": 8.305723190307617,
"step": 187
},
{
"epoch": 0.007557788944723618,
"grad_norm": 1.7232962846755981,
"learning_rate": 5.29980086969014e-05,
"loss": 8.544666290283203,
"step": 188
},
{
"epoch": 0.0075979899497487435,
"grad_norm": 2.292501449584961,
"learning_rate": 5.299797952142937e-05,
"loss": 8.15345287322998,
"step": 189
},
{
"epoch": 0.00763819095477387,
"grad_norm": 2.1953487396240234,
"learning_rate": 5.299795013379375e-05,
"loss": 8.195133209228516,
"step": 190
},
{
"epoch": 0.007678391959798995,
"grad_norm": 1.952316164970398,
"learning_rate": 5.2997920533995e-05,
"loss": 8.393413543701172,
"step": 191
},
{
"epoch": 0.007718592964824121,
"grad_norm": 2.1163394451141357,
"learning_rate": 5.2997890722033594e-05,
"loss": 8.450803756713867,
"step": 192
},
{
"epoch": 0.007758793969849246,
"grad_norm": 2.3067240715026855,
"learning_rate": 5.2997860697910005e-05,
"loss": 8.325882911682129,
"step": 193
},
{
"epoch": 0.007798994974874372,
"grad_norm": 1.8846808671951294,
"learning_rate": 5.299783046162472e-05,
"loss": 8.44339370727539,
"step": 194
},
{
"epoch": 0.007839195979899497,
"grad_norm": 2.147559642791748,
"learning_rate": 5.299780001317822e-05,
"loss": 8.601503372192383,
"step": 195
},
{
"epoch": 0.007879396984924623,
"grad_norm": 2.1101810932159424,
"learning_rate": 5.2997769352571e-05,
"loss": 8.690725326538086,
"step": 196
},
{
"epoch": 0.00791959798994975,
"grad_norm": 1.9192241430282593,
"learning_rate": 5.2997738479803534e-05,
"loss": 8.344963073730469,
"step": 197
},
{
"epoch": 0.007959798994974874,
"grad_norm": 1.8838149309158325,
"learning_rate": 5.299770739487633e-05,
"loss": 8.256372451782227,
"step": 198
},
{
"epoch": 0.008,
"grad_norm": 2.142127275466919,
"learning_rate": 5.299767609778988e-05,
"loss": 8.813041687011719,
"step": 199
},
{
"epoch": 0.008040201005025126,
"grad_norm": 2.14209246635437,
"learning_rate": 5.29976445885447e-05,
"loss": 8.368643760681152,
"step": 200
},
{
"epoch": 0.008040201005025126,
"eval_loss": 8.329636573791504,
"eval_runtime": 120.6602,
"eval_samples_per_second": 4.144,
"eval_steps_per_second": 0.522,
"step": 200
},
{
"epoch": 0.008080402010050251,
"grad_norm": 1.9065579175949097,
"learning_rate": 5.2997612867141274e-05,
"loss": 8.095634460449219,
"step": 201
},
{
"epoch": 0.008120603015075377,
"grad_norm": 1.743645429611206,
"learning_rate": 5.299758093358012e-05,
"loss": 8.445962905883789,
"step": 202
},
{
"epoch": 0.008160804020100502,
"grad_norm": 1.8151423931121826,
"learning_rate": 5.2997548787861746e-05,
"loss": 8.262279510498047,
"step": 203
},
{
"epoch": 0.008201005025125628,
"grad_norm": 1.9962149858474731,
"learning_rate": 5.299751642998667e-05,
"loss": 7.788729667663574,
"step": 204
},
{
"epoch": 0.008241206030150754,
"grad_norm": 1.635284185409546,
"learning_rate": 5.299748385995541e-05,
"loss": 8.26795482635498,
"step": 205
},
{
"epoch": 0.008281407035175879,
"grad_norm": 1.5759270191192627,
"learning_rate": 5.299745107776848e-05,
"loss": 8.327312469482422,
"step": 206
},
{
"epoch": 0.008321608040201005,
"grad_norm": 1.8083878755569458,
"learning_rate": 5.299741808342642e-05,
"loss": 8.556888580322266,
"step": 207
},
{
"epoch": 0.008361809045226131,
"grad_norm": 1.8477628231048584,
"learning_rate": 5.299738487692975e-05,
"loss": 8.182913780212402,
"step": 208
},
{
"epoch": 0.008402010050251256,
"grad_norm": 1.9069033861160278,
"learning_rate": 5.2997351458279e-05,
"loss": 8.041275978088379,
"step": 209
},
{
"epoch": 0.008442211055276382,
"grad_norm": 2.21096134185791,
"learning_rate": 5.299731782747471e-05,
"loss": 7.925780296325684,
"step": 210
},
{
"epoch": 0.008482412060301507,
"grad_norm": 1.769580364227295,
"learning_rate": 5.299728398451741e-05,
"loss": 8.227874755859375,
"step": 211
},
{
"epoch": 0.008522613065326633,
"grad_norm": 2.14461350440979,
"learning_rate": 5.299724992940765e-05,
"loss": 8.342981338500977,
"step": 212
},
{
"epoch": 0.00856281407035176,
"grad_norm": 2.380075454711914,
"learning_rate": 5.299721566214597e-05,
"loss": 8.313106536865234,
"step": 213
},
{
"epoch": 0.008603015075376884,
"grad_norm": 1.8549103736877441,
"learning_rate": 5.299718118273293e-05,
"loss": 8.241487503051758,
"step": 214
},
{
"epoch": 0.00864321608040201,
"grad_norm": 2.088374614715576,
"learning_rate": 5.299714649116906e-05,
"loss": 8.400121688842773,
"step": 215
},
{
"epoch": 0.008683417085427136,
"grad_norm": 1.921425223350525,
"learning_rate": 5.2997111587454935e-05,
"loss": 8.327886581420898,
"step": 216
},
{
"epoch": 0.008723618090452261,
"grad_norm": 1.6178687810897827,
"learning_rate": 5.299707647159111e-05,
"loss": 8.587099075317383,
"step": 217
},
{
"epoch": 0.008763819095477387,
"grad_norm": 1.916711449623108,
"learning_rate": 5.299704114357815e-05,
"loss": 7.974850177764893,
"step": 218
},
{
"epoch": 0.008804020100502512,
"grad_norm": 2.119534730911255,
"learning_rate": 5.299700560341661e-05,
"loss": 8.101737976074219,
"step": 219
},
{
"epoch": 0.008844221105527638,
"grad_norm": 1.8757777214050293,
"learning_rate": 5.299696985110707e-05,
"loss": 8.066093444824219,
"step": 220
},
{
"epoch": 0.008884422110552764,
"grad_norm": 1.8365973234176636,
"learning_rate": 5.29969338866501e-05,
"loss": 7.918840408325195,
"step": 221
},
{
"epoch": 0.008924623115577889,
"grad_norm": 1.928112506866455,
"learning_rate": 5.299689771004627e-05,
"loss": 8.177791595458984,
"step": 222
},
{
"epoch": 0.008964824120603015,
"grad_norm": 1.638963222503662,
"learning_rate": 5.2996861321296166e-05,
"loss": 8.336414337158203,
"step": 223
},
{
"epoch": 0.009005025125628141,
"grad_norm": 1.8400105237960815,
"learning_rate": 5.299682472040037e-05,
"loss": 8.074989318847656,
"step": 224
},
{
"epoch": 0.009045226130653266,
"grad_norm": 1.6984559297561646,
"learning_rate": 5.2996787907359466e-05,
"loss": 8.147102355957031,
"step": 225
},
{
"epoch": 0.009085427135678392,
"grad_norm": 1.5793787240982056,
"learning_rate": 5.299675088217404e-05,
"loss": 8.076702117919922,
"step": 226
},
{
"epoch": 0.009125628140703517,
"grad_norm": 1.7043931484222412,
"learning_rate": 5.2996713644844695e-05,
"loss": 8.162211418151855,
"step": 227
},
{
"epoch": 0.009165829145728643,
"grad_norm": 2.3001749515533447,
"learning_rate": 5.299667619537202e-05,
"loss": 8.543556213378906,
"step": 228
},
{
"epoch": 0.00920603015075377,
"grad_norm": 2.324345588684082,
"learning_rate": 5.299663853375661e-05,
"loss": 7.69166374206543,
"step": 229
},
{
"epoch": 0.009246231155778894,
"grad_norm": 2.1938083171844482,
"learning_rate": 5.299660065999908e-05,
"loss": 7.935137748718262,
"step": 230
},
{
"epoch": 0.00928643216080402,
"grad_norm": 1.7787119150161743,
"learning_rate": 5.299656257410003e-05,
"loss": 8.25346565246582,
"step": 231
},
{
"epoch": 0.009326633165829146,
"grad_norm": 2.108919382095337,
"learning_rate": 5.299652427606007e-05,
"loss": 8.440576553344727,
"step": 232
},
{
"epoch": 0.009366834170854271,
"grad_norm": 2.203165054321289,
"learning_rate": 5.299648576587981e-05,
"loss": 8.21126937866211,
"step": 233
},
{
"epoch": 0.009407035175879397,
"grad_norm": 1.577513575553894,
"learning_rate": 5.299644704355988e-05,
"loss": 8.442115783691406,
"step": 234
},
{
"epoch": 0.009447236180904522,
"grad_norm": 2.263481616973877,
"learning_rate": 5.299640810910089e-05,
"loss": 8.359823226928711,
"step": 235
},
{
"epoch": 0.009487437185929648,
"grad_norm": 2.0612189769744873,
"learning_rate": 5.299636896250345e-05,
"loss": 7.95751428604126,
"step": 236
},
{
"epoch": 0.009527638190954774,
"grad_norm": 1.7220946550369263,
"learning_rate": 5.299632960376822e-05,
"loss": 8.059045791625977,
"step": 237
},
{
"epoch": 0.009567839195979899,
"grad_norm": 1.922116994857788,
"learning_rate": 5.2996290032895804e-05,
"loss": 8.046489715576172,
"step": 238
},
{
"epoch": 0.009608040201005025,
"grad_norm": 2.153323173522949,
"learning_rate": 5.2996250249886845e-05,
"loss": 8.024328231811523,
"step": 239
},
{
"epoch": 0.009648241206030151,
"grad_norm": 1.789979338645935,
"learning_rate": 5.2996210254741974e-05,
"loss": 8.211692810058594,
"step": 240
},
{
"epoch": 0.009688442211055276,
"grad_norm": 1.8877555131912231,
"learning_rate": 5.299617004746184e-05,
"loss": 7.929727554321289,
"step": 241
},
{
"epoch": 0.009728643216080402,
"grad_norm": 1.9312947988510132,
"learning_rate": 5.2996129628047085e-05,
"loss": 8.091294288635254,
"step": 242
},
{
"epoch": 0.009768844221105528,
"grad_norm": 1.5974318981170654,
"learning_rate": 5.299608899649835e-05,
"loss": 8.471363067626953,
"step": 243
},
{
"epoch": 0.009809045226130653,
"grad_norm": 1.8455992937088013,
"learning_rate": 5.299604815281629e-05,
"loss": 8.416263580322266,
"step": 244
},
{
"epoch": 0.00984924623115578,
"grad_norm": 1.6630140542984009,
"learning_rate": 5.299600709700156e-05,
"loss": 8.489864349365234,
"step": 245
},
{
"epoch": 0.009889447236180904,
"grad_norm": 1.7287297248840332,
"learning_rate": 5.299596582905482e-05,
"loss": 8.332731246948242,
"step": 246
},
{
"epoch": 0.00992964824120603,
"grad_norm": 1.7951024770736694,
"learning_rate": 5.2995924348976725e-05,
"loss": 8.207561492919922,
"step": 247
},
{
"epoch": 0.009969849246231156,
"grad_norm": 1.566335678100586,
"learning_rate": 5.2995882656767944e-05,
"loss": 8.144509315490723,
"step": 248
},
{
"epoch": 0.010010050251256281,
"grad_norm": 1.612272024154663,
"learning_rate": 5.299584075242913e-05,
"loss": 8.343093872070312,
"step": 249
},
{
"epoch": 0.010050251256281407,
"grad_norm": 2.067436456680298,
"learning_rate": 5.299579863596099e-05,
"loss": 8.039451599121094,
"step": 250
},
{
"epoch": 0.010090452261306533,
"grad_norm": 1.655832052230835,
"learning_rate": 5.299575630736416e-05,
"loss": 8.290672302246094,
"step": 251
},
{
"epoch": 0.010130653266331658,
"grad_norm": 1.7005506753921509,
"learning_rate": 5.299571376663933e-05,
"loss": 8.20401382446289,
"step": 252
},
{
"epoch": 0.010170854271356784,
"grad_norm": 1.853699803352356,
"learning_rate": 5.2995671013787195e-05,
"loss": 7.941474914550781,
"step": 253
},
{
"epoch": 0.010211055276381909,
"grad_norm": 1.8550472259521484,
"learning_rate": 5.299562804880842e-05,
"loss": 8.137377738952637,
"step": 254
},
{
"epoch": 0.010251256281407035,
"grad_norm": 1.7704405784606934,
"learning_rate": 5.2995584871703704e-05,
"loss": 8.217637062072754,
"step": 255
},
{
"epoch": 0.010291457286432161,
"grad_norm": 1.8236024379730225,
"learning_rate": 5.2995541482473744e-05,
"loss": 8.182514190673828,
"step": 256
},
{
"epoch": 0.010331658291457286,
"grad_norm": 1.5614441633224487,
"learning_rate": 5.299549788111922e-05,
"loss": 8.303861618041992,
"step": 257
},
{
"epoch": 0.010371859296482412,
"grad_norm": 1.6350098848342896,
"learning_rate": 5.299545406764084e-05,
"loss": 7.942919731140137,
"step": 258
},
{
"epoch": 0.010412060301507538,
"grad_norm": 1.6425225734710693,
"learning_rate": 5.29954100420393e-05,
"loss": 7.961499214172363,
"step": 259
},
{
"epoch": 0.010452261306532663,
"grad_norm": 1.6786245107650757,
"learning_rate": 5.299536580431531e-05,
"loss": 8.177148818969727,
"step": 260
},
{
"epoch": 0.01049246231155779,
"grad_norm": 2.181872844696045,
"learning_rate": 5.2995321354469575e-05,
"loss": 8.539838790893555,
"step": 261
},
{
"epoch": 0.010532663316582914,
"grad_norm": 1.7575291395187378,
"learning_rate": 5.299527669250281e-05,
"loss": 8.48668098449707,
"step": 262
},
{
"epoch": 0.01057286432160804,
"grad_norm": 1.6861909627914429,
"learning_rate": 5.299523181841573e-05,
"loss": 8.046915054321289,
"step": 263
},
{
"epoch": 0.010613065326633166,
"grad_norm": 1.6378755569458008,
"learning_rate": 5.299518673220906e-05,
"loss": 8.09168529510498,
"step": 264
},
{
"epoch": 0.010653266331658291,
"grad_norm": 1.9889707565307617,
"learning_rate": 5.29951414338835e-05,
"loss": 7.85469913482666,
"step": 265
},
{
"epoch": 0.010693467336683417,
"grad_norm": 1.9059841632843018,
"learning_rate": 5.29950959234398e-05,
"loss": 8.168853759765625,
"step": 266
},
{
"epoch": 0.010733668341708543,
"grad_norm": 1.77418851852417,
"learning_rate": 5.2995050200878675e-05,
"loss": 8.500150680541992,
"step": 267
},
{
"epoch": 0.010773869346733668,
"grad_norm": 1.8010951280593872,
"learning_rate": 5.299500426620087e-05,
"loss": 7.875190734863281,
"step": 268
},
{
"epoch": 0.010814070351758794,
"grad_norm": 1.9225850105285645,
"learning_rate": 5.29949581194071e-05,
"loss": 7.694550514221191,
"step": 269
},
{
"epoch": 0.010854271356783919,
"grad_norm": 3.0016684532165527,
"learning_rate": 5.299491176049812e-05,
"loss": 8.374616622924805,
"step": 270
},
{
"epoch": 0.010894472361809045,
"grad_norm": 1.7531616687774658,
"learning_rate": 5.299486518947467e-05,
"loss": 8.114014625549316,
"step": 271
},
{
"epoch": 0.010934673366834171,
"grad_norm": 1.78202223777771,
"learning_rate": 5.299481840633749e-05,
"loss": 7.728270530700684,
"step": 272
},
{
"epoch": 0.010974874371859296,
"grad_norm": 2.1367790699005127,
"learning_rate": 5.299477141108734e-05,
"loss": 8.288908004760742,
"step": 273
},
{
"epoch": 0.011015075376884422,
"grad_norm": 1.9567782878875732,
"learning_rate": 5.299472420372496e-05,
"loss": 8.570487022399902,
"step": 274
},
{
"epoch": 0.011055276381909548,
"grad_norm": 2.19480299949646,
"learning_rate": 5.299467678425111e-05,
"loss": 8.284469604492188,
"step": 275
},
{
"epoch": 0.011095477386934673,
"grad_norm": 2.1869468688964844,
"learning_rate": 5.299462915266656e-05,
"loss": 8.159494400024414,
"step": 276
},
{
"epoch": 0.0111356783919598,
"grad_norm": 1.7680509090423584,
"learning_rate": 5.2994581308972056e-05,
"loss": 8.155216217041016,
"step": 277
},
{
"epoch": 0.011175879396984924,
"grad_norm": 1.9384150505065918,
"learning_rate": 5.2994533253168376e-05,
"loss": 7.980001449584961,
"step": 278
},
{
"epoch": 0.01121608040201005,
"grad_norm": 1.9903982877731323,
"learning_rate": 5.299448498525628e-05,
"loss": 8.5211820602417,
"step": 279
},
{
"epoch": 0.011256281407035176,
"grad_norm": 1.844769835472107,
"learning_rate": 5.299443650523656e-05,
"loss": 7.9173078536987305,
"step": 280
},
{
"epoch": 0.0112964824120603,
"grad_norm": 1.8510442972183228,
"learning_rate": 5.299438781310997e-05,
"loss": 8.049406051635742,
"step": 281
},
{
"epoch": 0.011336683417085427,
"grad_norm": 1.815053939819336,
"learning_rate": 5.2994338908877295e-05,
"loss": 8.241884231567383,
"step": 282
},
{
"epoch": 0.011376884422110553,
"grad_norm": 1.6938130855560303,
"learning_rate": 5.299428979253933e-05,
"loss": 7.900325775146484,
"step": 283
},
{
"epoch": 0.011417085427135678,
"grad_norm": 1.9421623945236206,
"learning_rate": 5.299424046409685e-05,
"loss": 7.94612979888916,
"step": 284
},
{
"epoch": 0.011457286432160804,
"grad_norm": 2.073329448699951,
"learning_rate": 5.299419092355065e-05,
"loss": 7.581545352935791,
"step": 285
},
{
"epoch": 0.01149748743718593,
"grad_norm": 1.749672770500183,
"learning_rate": 5.299414117090153e-05,
"loss": 7.50168514251709,
"step": 286
},
{
"epoch": 0.011537688442211055,
"grad_norm": 1.756786584854126,
"learning_rate": 5.299409120615027e-05,
"loss": 8.274606704711914,
"step": 287
},
{
"epoch": 0.011577889447236181,
"grad_norm": 1.8428155183792114,
"learning_rate": 5.299404102929768e-05,
"loss": 8.102296829223633,
"step": 288
},
{
"epoch": 0.011618090452261306,
"grad_norm": 1.8748948574066162,
"learning_rate": 5.299399064034456e-05,
"loss": 8.364482879638672,
"step": 289
},
{
"epoch": 0.011658291457286432,
"grad_norm": 2.1376612186431885,
"learning_rate": 5.2993940039291734e-05,
"loss": 7.872743129730225,
"step": 290
},
{
"epoch": 0.011698492462311558,
"grad_norm": 1.6551979780197144,
"learning_rate": 5.2993889226139986e-05,
"loss": 8.025882720947266,
"step": 291
},
{
"epoch": 0.011738693467336683,
"grad_norm": 1.5456299781799316,
"learning_rate": 5.2993838200890145e-05,
"loss": 8.093013763427734,
"step": 292
},
{
"epoch": 0.01177889447236181,
"grad_norm": 1.707608938217163,
"learning_rate": 5.2993786963543016e-05,
"loss": 8.270732879638672,
"step": 293
},
{
"epoch": 0.011819095477386935,
"grad_norm": 1.6922718286514282,
"learning_rate": 5.299373551409944e-05,
"loss": 8.202585220336914,
"step": 294
},
{
"epoch": 0.01185929648241206,
"grad_norm": 1.951208233833313,
"learning_rate": 5.299368385256023e-05,
"loss": 8.459943771362305,
"step": 295
},
{
"epoch": 0.011899497487437186,
"grad_norm": 1.952307105064392,
"learning_rate": 5.299363197892621e-05,
"loss": 8.179830551147461,
"step": 296
},
{
"epoch": 0.01193969849246231,
"grad_norm": 1.6900849342346191,
"learning_rate": 5.299357989319821e-05,
"loss": 8.08771800994873,
"step": 297
},
{
"epoch": 0.011979899497487437,
"grad_norm": 1.8033514022827148,
"learning_rate": 5.2993527595377074e-05,
"loss": 8.586979866027832,
"step": 298
},
{
"epoch": 0.012020100502512563,
"grad_norm": 1.8997255563735962,
"learning_rate": 5.299347508546363e-05,
"loss": 8.17015266418457,
"step": 299
},
{
"epoch": 0.012060301507537688,
"grad_norm": 1.724466323852539,
"learning_rate": 5.299342236345873e-05,
"loss": 7.915013313293457,
"step": 300
},
{
"epoch": 0.012100502512562814,
"grad_norm": 1.7345621585845947,
"learning_rate": 5.299336942936319e-05,
"loss": 8.606650352478027,
"step": 301
},
{
"epoch": 0.01214070351758794,
"grad_norm": 1.9641926288604736,
"learning_rate": 5.29933162831779e-05,
"loss": 7.577912330627441,
"step": 302
},
{
"epoch": 0.012180904522613065,
"grad_norm": 1.8717650175094604,
"learning_rate": 5.299326292490368e-05,
"loss": 8.280461311340332,
"step": 303
},
{
"epoch": 0.012221105527638191,
"grad_norm": 1.6723988056182861,
"learning_rate": 5.299320935454139e-05,
"loss": 8.352076530456543,
"step": 304
},
{
"epoch": 0.012261306532663316,
"grad_norm": 1.7926883697509766,
"learning_rate": 5.29931555720919e-05,
"loss": 7.992792129516602,
"step": 305
},
{
"epoch": 0.012301507537688442,
"grad_norm": 2.0198235511779785,
"learning_rate": 5.299310157755606e-05,
"loss": 8.420366287231445,
"step": 306
},
{
"epoch": 0.012341708542713568,
"grad_norm": 2.1053898334503174,
"learning_rate": 5.2993047370934736e-05,
"loss": 7.984848976135254,
"step": 307
},
{
"epoch": 0.012381909547738693,
"grad_norm": 1.8485567569732666,
"learning_rate": 5.29929929522288e-05,
"loss": 7.71517276763916,
"step": 308
},
{
"epoch": 0.01242211055276382,
"grad_norm": 1.754889726638794,
"learning_rate": 5.299293832143912e-05,
"loss": 8.37870979309082,
"step": 309
},
{
"epoch": 0.012462311557788945,
"grad_norm": 1.877976655960083,
"learning_rate": 5.299288347856658e-05,
"loss": 8.1470947265625,
"step": 310
},
{
"epoch": 0.01250251256281407,
"grad_norm": 1.8089011907577515,
"learning_rate": 5.299282842361204e-05,
"loss": 7.907635688781738,
"step": 311
},
{
"epoch": 0.012542713567839196,
"grad_norm": 1.9196230173110962,
"learning_rate": 5.29927731565764e-05,
"loss": 8.086762428283691,
"step": 312
},
{
"epoch": 0.01258291457286432,
"grad_norm": 1.5552493333816528,
"learning_rate": 5.299271767746053e-05,
"loss": 8.296794891357422,
"step": 313
},
{
"epoch": 0.012623115577889447,
"grad_norm": 1.43095862865448,
"learning_rate": 5.2992661986265335e-05,
"loss": 8.040600776672363,
"step": 314
},
{
"epoch": 0.012663316582914573,
"grad_norm": 1.7765002250671387,
"learning_rate": 5.299260608299169e-05,
"loss": 7.962665557861328,
"step": 315
},
{
"epoch": 0.012703517587939698,
"grad_norm": 2.180147409439087,
"learning_rate": 5.2992549967640504e-05,
"loss": 7.584565162658691,
"step": 316
},
{
"epoch": 0.012743718592964824,
"grad_norm": 2.8131022453308105,
"learning_rate": 5.299249364021267e-05,
"loss": 8.377334594726562,
"step": 317
},
{
"epoch": 0.01278391959798995,
"grad_norm": 1.8852747678756714,
"learning_rate": 5.2992437100709086e-05,
"loss": 7.793500900268555,
"step": 318
},
{
"epoch": 0.012824120603015075,
"grad_norm": 1.8344587087631226,
"learning_rate": 5.299238034913066e-05,
"loss": 8.73257827758789,
"step": 319
},
{
"epoch": 0.012864321608040201,
"grad_norm": 1.707476019859314,
"learning_rate": 5.2992323385478305e-05,
"loss": 7.91800594329834,
"step": 320
},
{
"epoch": 0.012904522613065326,
"grad_norm": 1.4946818351745605,
"learning_rate": 5.2992266209752935e-05,
"loss": 8.136693000793457,
"step": 321
},
{
"epoch": 0.012944723618090452,
"grad_norm": 1.7711795568466187,
"learning_rate": 5.299220882195546e-05,
"loss": 7.9373650550842285,
"step": 322
},
{
"epoch": 0.012984924623115578,
"grad_norm": 1.7863399982452393,
"learning_rate": 5.299215122208679e-05,
"loss": 8.012618064880371,
"step": 323
},
{
"epoch": 0.013025125628140703,
"grad_norm": 1.7535797357559204,
"learning_rate": 5.299209341014787e-05,
"loss": 8.243230819702148,
"step": 324
},
{
"epoch": 0.01306532663316583,
"grad_norm": 1.5717593431472778,
"learning_rate": 5.2992035386139606e-05,
"loss": 7.8636016845703125,
"step": 325
},
{
"epoch": 0.013105527638190955,
"grad_norm": 1.5868475437164307,
"learning_rate": 5.299197715006293e-05,
"loss": 8.336069107055664,
"step": 326
},
{
"epoch": 0.01314572864321608,
"grad_norm": 1.574897289276123,
"learning_rate": 5.2991918701918794e-05,
"loss": 7.972940444946289,
"step": 327
},
{
"epoch": 0.013185929648241206,
"grad_norm": 1.503908634185791,
"learning_rate": 5.299186004170811e-05,
"loss": 8.123407363891602,
"step": 328
},
{
"epoch": 0.01322613065326633,
"grad_norm": 1.5201798677444458,
"learning_rate": 5.299180116943183e-05,
"loss": 8.121698379516602,
"step": 329
},
{
"epoch": 0.013266331658291457,
"grad_norm": 1.585316777229309,
"learning_rate": 5.2991742085090895e-05,
"loss": 7.96183967590332,
"step": 330
},
{
"epoch": 0.013306532663316583,
"grad_norm": 1.6284129619598389,
"learning_rate": 5.299168278868625e-05,
"loss": 8.165437698364258,
"step": 331
},
{
"epoch": 0.013346733668341708,
"grad_norm": 1.6082704067230225,
"learning_rate": 5.299162328021884e-05,
"loss": 7.863561153411865,
"step": 332
},
{
"epoch": 0.013386934673366834,
"grad_norm": 1.4565194845199585,
"learning_rate": 5.2991563559689625e-05,
"loss": 8.292040824890137,
"step": 333
},
{
"epoch": 0.01342713567839196,
"grad_norm": 1.8588314056396484,
"learning_rate": 5.299150362709956e-05,
"loss": 7.816753387451172,
"step": 334
},
{
"epoch": 0.013467336683417085,
"grad_norm": 1.7544786930084229,
"learning_rate": 5.2991443482449604e-05,
"loss": 7.907513618469238,
"step": 335
},
{
"epoch": 0.013507537688442211,
"grad_norm": 1.9769484996795654,
"learning_rate": 5.2991383125740715e-05,
"loss": 7.704319000244141,
"step": 336
},
{
"epoch": 0.013547738693467338,
"grad_norm": 2.573455572128296,
"learning_rate": 5.299132255697386e-05,
"loss": 8.253981590270996,
"step": 337
},
{
"epoch": 0.013587939698492462,
"grad_norm": 1.7910648584365845,
"learning_rate": 5.299126177615003e-05,
"loss": 7.930364608764648,
"step": 338
},
{
"epoch": 0.013628140703517588,
"grad_norm": 1.711910605430603,
"learning_rate": 5.299120078327017e-05,
"loss": 8.110099792480469,
"step": 339
},
{
"epoch": 0.013668341708542713,
"grad_norm": 1.7616333961486816,
"learning_rate": 5.2991139578335274e-05,
"loss": 8.19175910949707,
"step": 340
},
{
"epoch": 0.01370854271356784,
"grad_norm": 1.6909830570220947,
"learning_rate": 5.2991078161346313e-05,
"loss": 8.242263793945312,
"step": 341
},
{
"epoch": 0.013748743718592965,
"grad_norm": 1.4701412916183472,
"learning_rate": 5.2991016532304274e-05,
"loss": 7.96829891204834,
"step": 342
},
{
"epoch": 0.01378894472361809,
"grad_norm": 1.9361451864242554,
"learning_rate": 5.299095469121015e-05,
"loss": 8.199315071105957,
"step": 343
},
{
"epoch": 0.013829145728643216,
"grad_norm": 1.641666293144226,
"learning_rate": 5.2990892638064924e-05,
"loss": 8.088203430175781,
"step": 344
},
{
"epoch": 0.013869346733668343,
"grad_norm": 1.6051890850067139,
"learning_rate": 5.299083037286959e-05,
"loss": 8.386016845703125,
"step": 345
},
{
"epoch": 0.013909547738693467,
"grad_norm": 1.7053855657577515,
"learning_rate": 5.2990767895625144e-05,
"loss": 8.252748489379883,
"step": 346
},
{
"epoch": 0.013949748743718593,
"grad_norm": 1.7186076641082764,
"learning_rate": 5.29907052063326e-05,
"loss": 7.485268592834473,
"step": 347
},
{
"epoch": 0.013989949748743718,
"grad_norm": 1.7334588766098022,
"learning_rate": 5.2990642304992944e-05,
"loss": 8.158232688903809,
"step": 348
},
{
"epoch": 0.014030150753768844,
"grad_norm": 2.322744846343994,
"learning_rate": 5.2990579191607184e-05,
"loss": 7.839550971984863,
"step": 349
},
{
"epoch": 0.01407035175879397,
"grad_norm": 1.8092786073684692,
"learning_rate": 5.2990515866176354e-05,
"loss": 7.992804527282715,
"step": 350
},
{
"epoch": 0.014110552763819095,
"grad_norm": 1.625507116317749,
"learning_rate": 5.299045232870143e-05,
"loss": 8.091792106628418,
"step": 351
},
{
"epoch": 0.014150753768844221,
"grad_norm": 1.960445523262024,
"learning_rate": 5.299038857918347e-05,
"loss": 8.037437438964844,
"step": 352
},
{
"epoch": 0.014190954773869347,
"grad_norm": 2.2716407775878906,
"learning_rate": 5.2990324617623464e-05,
"loss": 8.430746078491211,
"step": 353
},
{
"epoch": 0.014231155778894472,
"grad_norm": 1.9768471717834473,
"learning_rate": 5.2990260444022454e-05,
"loss": 8.12667465209961,
"step": 354
},
{
"epoch": 0.014271356783919598,
"grad_norm": 2.049266815185547,
"learning_rate": 5.299019605838146e-05,
"loss": 8.054523468017578,
"step": 355
},
{
"epoch": 0.014311557788944723,
"grad_norm": 1.8299111127853394,
"learning_rate": 5.299013146070152e-05,
"loss": 8.040359497070312,
"step": 356
},
{
"epoch": 0.014351758793969849,
"grad_norm": 2.4692959785461426,
"learning_rate": 5.299006665098366e-05,
"loss": 7.745314598083496,
"step": 357
},
{
"epoch": 0.014391959798994975,
"grad_norm": 2.0177786350250244,
"learning_rate": 5.299000162922892e-05,
"loss": 7.913352966308594,
"step": 358
},
{
"epoch": 0.0144321608040201,
"grad_norm": 1.713757872581482,
"learning_rate": 5.2989936395438345e-05,
"loss": 8.021360397338867,
"step": 359
},
{
"epoch": 0.014472361809045226,
"grad_norm": 1.7628759145736694,
"learning_rate": 5.298987094961298e-05,
"loss": 8.234394073486328,
"step": 360
},
{
"epoch": 0.014512562814070352,
"grad_norm": 1.6097559928894043,
"learning_rate": 5.2989805291753864e-05,
"loss": 8.151266098022461,
"step": 361
},
{
"epoch": 0.014552763819095477,
"grad_norm": 1.930101752281189,
"learning_rate": 5.298973942186207e-05,
"loss": 7.837223052978516,
"step": 362
},
{
"epoch": 0.014592964824120603,
"grad_norm": 1.6143091917037964,
"learning_rate": 5.298967333993862e-05,
"loss": 8.020715713500977,
"step": 363
},
{
"epoch": 0.014633165829145728,
"grad_norm": 2.134955406188965,
"learning_rate": 5.29896070459846e-05,
"loss": 8.615276336669922,
"step": 364
},
{
"epoch": 0.014673366834170854,
"grad_norm": 1.8166667222976685,
"learning_rate": 5.298954054000106e-05,
"loss": 7.945555686950684,
"step": 365
},
{
"epoch": 0.01471356783919598,
"grad_norm": 1.604546308517456,
"learning_rate": 5.298947382198907e-05,
"loss": 8.14586067199707,
"step": 366
},
{
"epoch": 0.014753768844221105,
"grad_norm": 1.5604201555252075,
"learning_rate": 5.298940689194969e-05,
"loss": 8.18164348602295,
"step": 367
},
{
"epoch": 0.014793969849246231,
"grad_norm": 1.45653235912323,
"learning_rate": 5.2989339749884e-05,
"loss": 8.280872344970703,
"step": 368
},
{
"epoch": 0.014834170854271357,
"grad_norm": 1.6250518560409546,
"learning_rate": 5.298927239579307e-05,
"loss": 8.253007888793945,
"step": 369
},
{
"epoch": 0.014874371859296482,
"grad_norm": 1.6432377099990845,
"learning_rate": 5.2989204829677996e-05,
"loss": 7.998000621795654,
"step": 370
},
{
"epoch": 0.014914572864321608,
"grad_norm": 1.8273557424545288,
"learning_rate": 5.298913705153983e-05,
"loss": 8.382380485534668,
"step": 371
},
{
"epoch": 0.014954773869346733,
"grad_norm": 1.5560564994812012,
"learning_rate": 5.298906906137968e-05,
"loss": 7.850958824157715,
"step": 372
},
{
"epoch": 0.014994974874371859,
"grad_norm": 1.6020787954330444,
"learning_rate": 5.298900085919862e-05,
"loss": 7.909162521362305,
"step": 373
},
{
"epoch": 0.015035175879396985,
"grad_norm": 1.5184075832366943,
"learning_rate": 5.2988932444997755e-05,
"loss": 8.011605262756348,
"step": 374
},
{
"epoch": 0.01507537688442211,
"grad_norm": 1.7862534523010254,
"learning_rate": 5.298886381877818e-05,
"loss": 8.217576026916504,
"step": 375
},
{
"epoch": 0.015115577889447236,
"grad_norm": 1.847752332687378,
"learning_rate": 5.2988794980540986e-05,
"loss": 8.170892715454102,
"step": 376
},
{
"epoch": 0.015155778894472362,
"grad_norm": 1.724697470664978,
"learning_rate": 5.298872593028728e-05,
"loss": 7.943462371826172,
"step": 377
},
{
"epoch": 0.015195979899497487,
"grad_norm": 1.7283616065979004,
"learning_rate": 5.298865666801817e-05,
"loss": 8.082193374633789,
"step": 378
},
{
"epoch": 0.015236180904522613,
"grad_norm": 1.739368200302124,
"learning_rate": 5.298858719373475e-05,
"loss": 7.868390083312988,
"step": 379
},
{
"epoch": 0.01527638190954774,
"grad_norm": 1.706627368927002,
"learning_rate": 5.2988517507438156e-05,
"loss": 8.000913619995117,
"step": 380
},
{
"epoch": 0.015316582914572864,
"grad_norm": 2.2589359283447266,
"learning_rate": 5.2988447609129484e-05,
"loss": 8.007084846496582,
"step": 381
},
{
"epoch": 0.01535678391959799,
"grad_norm": 1.8864123821258545,
"learning_rate": 5.2988377498809876e-05,
"loss": 7.926105499267578,
"step": 382
},
{
"epoch": 0.015396984924623115,
"grad_norm": 1.692408561706543,
"learning_rate": 5.298830717648043e-05,
"loss": 8.117183685302734,
"step": 383
},
{
"epoch": 0.015437185929648241,
"grad_norm": 1.7368155717849731,
"learning_rate": 5.2988236642142285e-05,
"loss": 7.707394123077393,
"step": 384
},
{
"epoch": 0.015477386934673367,
"grad_norm": 2.068697214126587,
"learning_rate": 5.298816589579657e-05,
"loss": 7.823457717895508,
"step": 385
},
{
"epoch": 0.015517587939698492,
"grad_norm": 1.851224660873413,
"learning_rate": 5.2988094937444416e-05,
"loss": 8.345376968383789,
"step": 386
},
{
"epoch": 0.015557788944723618,
"grad_norm": 1.6005818843841553,
"learning_rate": 5.298802376708697e-05,
"loss": 7.89686918258667,
"step": 387
},
{
"epoch": 0.015597989949748745,
"grad_norm": 1.8737226724624634,
"learning_rate": 5.298795238472535e-05,
"loss": 8.235288619995117,
"step": 388
},
{
"epoch": 0.01563819095477387,
"grad_norm": 1.9664247035980225,
"learning_rate": 5.298788079036072e-05,
"loss": 8.285332679748535,
"step": 389
},
{
"epoch": 0.015678391959798994,
"grad_norm": 1.7721790075302124,
"learning_rate": 5.298780898399421e-05,
"loss": 8.128210067749023,
"step": 390
},
{
"epoch": 0.01571859296482412,
"grad_norm": 1.8024890422821045,
"learning_rate": 5.298773696562698e-05,
"loss": 7.518380165100098,
"step": 391
},
{
"epoch": 0.015758793969849246,
"grad_norm": 1.716502070426941,
"learning_rate": 5.2987664735260176e-05,
"loss": 7.901702404022217,
"step": 392
},
{
"epoch": 0.015798994974874372,
"grad_norm": 1.8132661581039429,
"learning_rate": 5.298759229289497e-05,
"loss": 7.936091899871826,
"step": 393
},
{
"epoch": 0.0158391959798995,
"grad_norm": 1.819899320602417,
"learning_rate": 5.29875196385325e-05,
"loss": 8.029167175292969,
"step": 394
},
{
"epoch": 0.01587939698492462,
"grad_norm": 1.8582565784454346,
"learning_rate": 5.298744677217396e-05,
"loss": 8.186941146850586,
"step": 395
},
{
"epoch": 0.015919597989949748,
"grad_norm": 1.8473869562149048,
"learning_rate": 5.298737369382048e-05,
"loss": 7.92045783996582,
"step": 396
},
{
"epoch": 0.015959798994974874,
"grad_norm": 1.881702184677124,
"learning_rate": 5.2987300403473255e-05,
"loss": 7.990792274475098,
"step": 397
},
{
"epoch": 0.016,
"grad_norm": 1.7451468706130981,
"learning_rate": 5.2987226901133454e-05,
"loss": 8.109321594238281,
"step": 398
},
{
"epoch": 0.016040201005025127,
"grad_norm": 1.705556869506836,
"learning_rate": 5.298715318680225e-05,
"loss": 7.83171272277832,
"step": 399
},
{
"epoch": 0.016080402010050253,
"grad_norm": 3.7052950859069824,
"learning_rate": 5.298707926048083e-05,
"loss": 9.070070266723633,
"step": 400
},
{
"epoch": 0.016080402010050253,
"eval_loss": 8.025728225708008,
"eval_runtime": 120.6238,
"eval_samples_per_second": 4.145,
"eval_steps_per_second": 0.522,
"step": 400
},
{
"epoch": 0.016120603015075376,
"grad_norm": 1.9343644380569458,
"learning_rate": 0.0,
"loss": 8.122842788696289,
"step": 401
},
{
"epoch": 0.016160804020100502,
"grad_norm": 2.240751266479492,
"learning_rate": 1.06e-06,
"loss": 8.081636428833008,
"step": 402
},
{
"epoch": 0.016201005025125628,
"grad_norm": 2.1998350620269775,
"learning_rate": 2.12e-06,
"loss": 8.05289077758789,
"step": 403
},
{
"epoch": 0.016241206030150755,
"grad_norm": 1.6559151411056519,
"learning_rate": 3.18e-06,
"loss": 8.093862533569336,
"step": 404
},
{
"epoch": 0.01628140703517588,
"grad_norm": 1.5775786638259888,
"learning_rate": 4.24e-06,
"loss": 8.145739555358887,
"step": 405
},
{
"epoch": 0.016321608040201004,
"grad_norm": 1.5591992139816284,
"learning_rate": 5.3e-06,
"loss": 8.138541221618652,
"step": 406
},
{
"epoch": 0.01636180904522613,
"grad_norm": 1.5250623226165771,
"learning_rate": 6.36e-06,
"loss": 7.995719909667969,
"step": 407
},
{
"epoch": 0.016402010050251256,
"grad_norm": 1.5896787643432617,
"learning_rate": 7.420000000000001e-06,
"loss": 8.096259117126465,
"step": 408
},
{
"epoch": 0.016442211055276382,
"grad_norm": 1.7962499856948853,
"learning_rate": 8.48e-06,
"loss": 8.321344375610352,
"step": 409
},
{
"epoch": 0.01648241206030151,
"grad_norm": 1.7921619415283203,
"learning_rate": 9.54e-06,
"loss": 8.23189640045166,
"step": 410
},
{
"epoch": 0.01652261306532663,
"grad_norm": 1.7183833122253418,
"learning_rate": 1.06e-05,
"loss": 8.226393699645996,
"step": 411
},
{
"epoch": 0.016562814070351758,
"grad_norm": 1.928035020828247,
"learning_rate": 1.166e-05,
"loss": 7.735424041748047,
"step": 412
},
{
"epoch": 0.016603015075376884,
"grad_norm": 1.7883954048156738,
"learning_rate": 1.272e-05,
"loss": 7.55402946472168,
"step": 413
},
{
"epoch": 0.01664321608040201,
"grad_norm": 1.7375296354293823,
"learning_rate": 1.378e-05,
"loss": 7.801723957061768,
"step": 414
},
{
"epoch": 0.016683417085427137,
"grad_norm": 1.8750230073928833,
"learning_rate": 1.4840000000000002e-05,
"loss": 8.972478866577148,
"step": 415
},
{
"epoch": 0.016723618090452263,
"grad_norm": 1.617038369178772,
"learning_rate": 1.59e-05,
"loss": 8.07492733001709,
"step": 416
},
{
"epoch": 0.016763819095477386,
"grad_norm": 1.4113925695419312,
"learning_rate": 1.696e-05,
"loss": 7.820061683654785,
"step": 417
},
{
"epoch": 0.016804020100502512,
"grad_norm": 1.6749427318572998,
"learning_rate": 1.802e-05,
"loss": 7.878408432006836,
"step": 418
},
{
"epoch": 0.016844221105527638,
"grad_norm": 1.5479542016983032,
"learning_rate": 1.908e-05,
"loss": 7.99974250793457,
"step": 419
},
{
"epoch": 0.016884422110552764,
"grad_norm": 1.6959515810012817,
"learning_rate": 2.014e-05,
"loss": 7.842455863952637,
"step": 420
},
{
"epoch": 0.01692462311557789,
"grad_norm": 1.7194098234176636,
"learning_rate": 2.12e-05,
"loss": 7.701992988586426,
"step": 421
},
{
"epoch": 0.016964824120603014,
"grad_norm": 1.63682222366333,
"learning_rate": 2.226e-05,
"loss": 8.210274696350098,
"step": 422
},
{
"epoch": 0.01700502512562814,
"grad_norm": 1.554741382598877,
"learning_rate": 2.332e-05,
"loss": 7.4565324783325195,
"step": 423
},
{
"epoch": 0.017045226130653266,
"grad_norm": 1.6605011224746704,
"learning_rate": 2.438e-05,
"loss": 8.209391593933105,
"step": 424
},
{
"epoch": 0.017085427135678392,
"grad_norm": 1.6479284763336182,
"learning_rate": 2.544e-05,
"loss": 8.255731582641602,
"step": 425
},
{
"epoch": 0.01712562814070352,
"grad_norm": 1.6426044702529907,
"learning_rate": 2.65e-05,
"loss": 7.846493244171143,
"step": 426
},
{
"epoch": 0.017165829145728645,
"grad_norm": 1.4833605289459229,
"learning_rate": 2.756e-05,
"loss": 8.50640869140625,
"step": 427
},
{
"epoch": 0.017206030150753768,
"grad_norm": 1.840043067932129,
"learning_rate": 2.8620000000000004e-05,
"loss": 7.605090141296387,
"step": 428
},
{
"epoch": 0.017246231155778894,
"grad_norm": 1.6083662509918213,
"learning_rate": 2.9680000000000004e-05,
"loss": 8.304194450378418,
"step": 429
},
{
"epoch": 0.01728643216080402,
"grad_norm": 1.6378768682479858,
"learning_rate": 3.074e-05,
"loss": 7.8139166831970215,
"step": 430
},
{
"epoch": 0.017326633165829147,
"grad_norm": 2.0351948738098145,
"learning_rate": 3.18e-05,
"loss": 7.8874406814575195,
"step": 431
},
{
"epoch": 0.017366834170854273,
"grad_norm": 1.4985382556915283,
"learning_rate": 3.286e-05,
"loss": 8.083316802978516,
"step": 432
},
{
"epoch": 0.017407035175879396,
"grad_norm": 1.5924725532531738,
"learning_rate": 3.392e-05,
"loss": 7.9780988693237305,
"step": 433
},
{
"epoch": 0.017447236180904522,
"grad_norm": 1.6368999481201172,
"learning_rate": 3.498e-05,
"loss": 8.305475234985352,
"step": 434
},
{
"epoch": 0.017487437185929648,
"grad_norm": 1.7157480716705322,
"learning_rate": 3.604e-05,
"loss": 7.766155242919922,
"step": 435
},
{
"epoch": 0.017527638190954774,
"grad_norm": 1.6965630054473877,
"learning_rate": 3.71e-05,
"loss": 7.857510566711426,
"step": 436
},
{
"epoch": 0.0175678391959799,
"grad_norm": 1.7198212146759033,
"learning_rate": 3.816e-05,
"loss": 8.362035751342773,
"step": 437
},
{
"epoch": 0.017608040201005024,
"grad_norm": 1.5021599531173706,
"learning_rate": 3.922e-05,
"loss": 7.712345600128174,
"step": 438
},
{
"epoch": 0.01764824120603015,
"grad_norm": 1.8201051950454712,
"learning_rate": 4.028e-05,
"loss": 7.64145040512085,
"step": 439
},
{
"epoch": 0.017688442211055276,
"grad_norm": 1.8476561307907104,
"learning_rate": 4.134e-05,
"loss": 8.150541305541992,
"step": 440
},
{
"epoch": 0.017728643216080402,
"grad_norm": 1.6296873092651367,
"learning_rate": 4.24e-05,
"loss": 8.086748123168945,
"step": 441
},
{
"epoch": 0.01776884422110553,
"grad_norm": 1.6610530614852905,
"learning_rate": 4.346e-05,
"loss": 7.691719055175781,
"step": 442
},
{
"epoch": 0.017809045226130655,
"grad_norm": 1.6869932413101196,
"learning_rate": 4.452e-05,
"loss": 8.260007858276367,
"step": 443
},
{
"epoch": 0.017849246231155778,
"grad_norm": 1.9870980978012085,
"learning_rate": 4.558e-05,
"loss": 7.540270805358887,
"step": 444
},
{
"epoch": 0.017889447236180904,
"grad_norm": 2.03045654296875,
"learning_rate": 4.664e-05,
"loss": 7.5766921043396,
"step": 445
},
{
"epoch": 0.01792964824120603,
"grad_norm": 1.7114081382751465,
"learning_rate": 4.77e-05,
"loss": 7.837658405303955,
"step": 446
},
{
"epoch": 0.017969849246231157,
"grad_norm": 1.7159371376037598,
"learning_rate": 4.876e-05,
"loss": 8.075382232666016,
"step": 447
},
{
"epoch": 0.018010050251256283,
"grad_norm": 2.1670467853546143,
"learning_rate": 4.982e-05,
"loss": 8.496036529541016,
"step": 448
},
{
"epoch": 0.018050251256281406,
"grad_norm": 2.2908828258514404,
"learning_rate": 5.088e-05,
"loss": 7.816028594970703,
"step": 449
},
{
"epoch": 0.018090452261306532,
"grad_norm": 1.6664937734603882,
"learning_rate": 5.194e-05,
"loss": 8.007561683654785,
"step": 450
},
{
"epoch": 0.018130653266331658,
"grad_norm": 1.6417421102523804,
"learning_rate": 5.3e-05,
"loss": 8.010965347290039,
"step": 451
},
{
"epoch": 0.018170854271356784,
"grad_norm": 2.1080172061920166,
"learning_rate": 5.299999989390202e-05,
"loss": 7.826607704162598,
"step": 452
},
{
"epoch": 0.01821105527638191,
"grad_norm": 1.508095622062683,
"learning_rate": 5.299999957560809e-05,
"loss": 7.963935375213623,
"step": 453
},
{
"epoch": 0.018251256281407034,
"grad_norm": 1.7453914880752563,
"learning_rate": 5.29999990451182e-05,
"loss": 7.871636390686035,
"step": 454
},
{
"epoch": 0.01829145728643216,
"grad_norm": 1.8799628019332886,
"learning_rate": 5.2999998302432364e-05,
"loss": 7.683241844177246,
"step": 455
},
{
"epoch": 0.018331658291457286,
"grad_norm": 1.6714903116226196,
"learning_rate": 5.2999997347550594e-05,
"loss": 7.969829082489014,
"step": 456
},
{
"epoch": 0.018371859296482412,
"grad_norm": 2.0190796852111816,
"learning_rate": 5.2999996180472915e-05,
"loss": 7.746788024902344,
"step": 457
},
{
"epoch": 0.01841206030150754,
"grad_norm": 1.661812663078308,
"learning_rate": 5.299999480119934e-05,
"loss": 8.140985488891602,
"step": 458
},
{
"epoch": 0.018452261306532665,
"grad_norm": 1.7817410230636597,
"learning_rate": 5.299999320972989e-05,
"loss": 7.923608303070068,
"step": 459
},
{
"epoch": 0.018492462311557788,
"grad_norm": 1.6462630033493042,
"learning_rate": 5.299999140606459e-05,
"loss": 7.76696252822876,
"step": 460
},
{
"epoch": 0.018532663316582914,
"grad_norm": 3.0535898208618164,
"learning_rate": 5.299998939020346e-05,
"loss": 8.089165687561035,
"step": 461
},
{
"epoch": 0.01857286432160804,
"grad_norm": 1.8203084468841553,
"learning_rate": 5.2999987162146547e-05,
"loss": 8.260978698730469,
"step": 462
},
{
"epoch": 0.018613065326633167,
"grad_norm": 1.7850511074066162,
"learning_rate": 5.299998472189388e-05,
"loss": 7.893509387969971,
"step": 463
},
{
"epoch": 0.018653266331658293,
"grad_norm": 2.1433422565460205,
"learning_rate": 5.29999820694455e-05,
"loss": 8.103431701660156,
"step": 464
},
{
"epoch": 0.018693467336683416,
"grad_norm": 1.7177734375,
"learning_rate": 5.2999979204801443e-05,
"loss": 8.027700424194336,
"step": 465
},
{
"epoch": 0.018733668341708542,
"grad_norm": 1.8191715478897095,
"learning_rate": 5.2999976127961764e-05,
"loss": 8.137571334838867,
"step": 466
},
{
"epoch": 0.018773869346733668,
"grad_norm": 1.978830337524414,
"learning_rate": 5.2999972838926514e-05,
"loss": 8.162650108337402,
"step": 467
},
{
"epoch": 0.018814070351758794,
"grad_norm": 1.8367021083831787,
"learning_rate": 5.2999969337695735e-05,
"loss": 7.902281284332275,
"step": 468
},
{
"epoch": 0.01885427135678392,
"grad_norm": 1.8460545539855957,
"learning_rate": 5.299996562426949e-05,
"loss": 7.820660591125488,
"step": 469
},
{
"epoch": 0.018894472361809043,
"grad_norm": 1.8266501426696777,
"learning_rate": 5.2999961698647826e-05,
"loss": 7.702997207641602,
"step": 470
},
{
"epoch": 0.01893467336683417,
"grad_norm": 2.107487440109253,
"learning_rate": 5.299995756083083e-05,
"loss": 7.7997145652771,
"step": 471
},
{
"epoch": 0.018974874371859296,
"grad_norm": 1.9008522033691406,
"learning_rate": 5.299995321081855e-05,
"loss": 8.508450508117676,
"step": 472
},
{
"epoch": 0.019015075376884422,
"grad_norm": 1.934217929840088,
"learning_rate": 5.299994864861106e-05,
"loss": 8.085101127624512,
"step": 473
},
{
"epoch": 0.01905527638190955,
"grad_norm": 2.0063512325286865,
"learning_rate": 5.299994387420843e-05,
"loss": 8.072752952575684,
"step": 474
},
{
"epoch": 0.019095477386934675,
"grad_norm": 1.710081934928894,
"learning_rate": 5.299993888761075e-05,
"loss": 8.035191535949707,
"step": 475
},
{
"epoch": 0.019135678391959798,
"grad_norm": 1.5775868892669678,
"learning_rate": 5.2999933688818086e-05,
"loss": 7.561495304107666,
"step": 476
},
{
"epoch": 0.019175879396984924,
"grad_norm": 1.8305647373199463,
"learning_rate": 5.2999928277830524e-05,
"loss": 7.785874843597412,
"step": 477
},
{
"epoch": 0.01921608040201005,
"grad_norm": 1.8369439840316772,
"learning_rate": 5.299992265464815e-05,
"loss": 7.7270026206970215,
"step": 478
},
{
"epoch": 0.019256281407035176,
"grad_norm": 1.8375827074050903,
"learning_rate": 5.2999916819271054e-05,
"loss": 7.991250991821289,
"step": 479
},
{
"epoch": 0.019296482412060303,
"grad_norm": 2.246628761291504,
"learning_rate": 5.299991077169933e-05,
"loss": 7.57627010345459,
"step": 480
},
{
"epoch": 0.019336683417085426,
"grad_norm": 2.3361101150512695,
"learning_rate": 5.2999904511933084e-05,
"loss": 7.681562423706055,
"step": 481
},
{
"epoch": 0.019376884422110552,
"grad_norm": 1.6286190748214722,
"learning_rate": 5.299989803997241e-05,
"loss": 7.847060203552246,
"step": 482
},
{
"epoch": 0.019417085427135678,
"grad_norm": 1.8155831098556519,
"learning_rate": 5.29998913558174e-05,
"loss": 7.899755477905273,
"step": 483
},
{
"epoch": 0.019457286432160804,
"grad_norm": 1.8039032220840454,
"learning_rate": 5.2999884459468174e-05,
"loss": 8.008749961853027,
"step": 484
},
{
"epoch": 0.01949748743718593,
"grad_norm": 1.9491512775421143,
"learning_rate": 5.2999877350924846e-05,
"loss": 7.859322547912598,
"step": 485
},
{
"epoch": 0.019537688442211057,
"grad_norm": 1.8126423358917236,
"learning_rate": 5.2999870030187514e-05,
"loss": 8.296046257019043,
"step": 486
},
{
"epoch": 0.01957788944723618,
"grad_norm": 1.8343174457550049,
"learning_rate": 5.299986249725631e-05,
"loss": 7.749964237213135,
"step": 487
},
{
"epoch": 0.019618090452261306,
"grad_norm": 1.7858879566192627,
"learning_rate": 5.299985475213135e-05,
"loss": 8.239532470703125,
"step": 488
},
{
"epoch": 0.019658291457286432,
"grad_norm": 1.6784310340881348,
"learning_rate": 5.299984679481275e-05,
"loss": 7.88999080657959,
"step": 489
},
{
"epoch": 0.01969849246231156,
"grad_norm": 1.662203311920166,
"learning_rate": 5.299983862530065e-05,
"loss": 7.899140357971191,
"step": 490
},
{
"epoch": 0.019738693467336685,
"grad_norm": 1.5990309715270996,
"learning_rate": 5.299983024359518e-05,
"loss": 7.8618621826171875,
"step": 491
},
{
"epoch": 0.019778894472361808,
"grad_norm": 1.5965832471847534,
"learning_rate": 5.299982164969647e-05,
"loss": 7.657140254974365,
"step": 492
},
{
"epoch": 0.019819095477386934,
"grad_norm": 1.8224987983703613,
"learning_rate": 5.299981284360465e-05,
"loss": 8.220462799072266,
"step": 493
},
{
"epoch": 0.01985929648241206,
"grad_norm": 1.767370343208313,
"learning_rate": 5.299980382531987e-05,
"loss": 7.3248186111450195,
"step": 494
},
{
"epoch": 0.019899497487437186,
"grad_norm": 1.8437494039535522,
"learning_rate": 5.2999794594842274e-05,
"loss": 8.065488815307617,
"step": 495
},
{
"epoch": 0.019939698492462313,
"grad_norm": 1.6843756437301636,
"learning_rate": 5.299978515217201e-05,
"loss": 8.185091972351074,
"step": 496
},
{
"epoch": 0.019979899497487436,
"grad_norm": 1.8131009340286255,
"learning_rate": 5.2999775497309224e-05,
"loss": 8.273950576782227,
"step": 497
},
{
"epoch": 0.020020100502512562,
"grad_norm": 1.9148132801055908,
"learning_rate": 5.299976563025408e-05,
"loss": 7.34578800201416,
"step": 498
},
{
"epoch": 0.020060301507537688,
"grad_norm": 1.8429914712905884,
"learning_rate": 5.299975555100673e-05,
"loss": 7.517185211181641,
"step": 499
},
{
"epoch": 0.020100502512562814,
"grad_norm": 1.6636184453964233,
"learning_rate": 5.299974525956733e-05,
"loss": 7.974506378173828,
"step": 500
},
{
"epoch": 0.02014070351758794,
"grad_norm": 1.9390288591384888,
"learning_rate": 5.299973475593606e-05,
"loss": 8.254663467407227,
"step": 501
},
{
"epoch": 0.020180904522613067,
"grad_norm": 1.6841087341308594,
"learning_rate": 5.299972404011307e-05,
"loss": 7.802603721618652,
"step": 502
},
{
"epoch": 0.02022110552763819,
"grad_norm": 1.7867002487182617,
"learning_rate": 5.2999713112098545e-05,
"loss": 7.905996799468994,
"step": 503
},
{
"epoch": 0.020261306532663316,
"grad_norm": 1.7222802639007568,
"learning_rate": 5.2999701971892654e-05,
"loss": 7.983509063720703,
"step": 504
},
{
"epoch": 0.020301507537688442,
"grad_norm": 2.466287136077881,
"learning_rate": 5.299969061949558e-05,
"loss": 8.196962356567383,
"step": 505
},
{
"epoch": 0.02034170854271357,
"grad_norm": 1.8125563859939575,
"learning_rate": 5.29996790549075e-05,
"loss": 8.10466480255127,
"step": 506
},
{
"epoch": 0.020381909547738695,
"grad_norm": 2.0414111614227295,
"learning_rate": 5.29996672781286e-05,
"loss": 7.853559970855713,
"step": 507
},
{
"epoch": 0.020422110552763818,
"grad_norm": 2.1194868087768555,
"learning_rate": 5.299965528915907e-05,
"loss": 8.583271026611328,
"step": 508
},
{
"epoch": 0.020462311557788944,
"grad_norm": 1.8091976642608643,
"learning_rate": 5.29996430879991e-05,
"loss": 7.544863700866699,
"step": 509
},
{
"epoch": 0.02050251256281407,
"grad_norm": 1.725387692451477,
"learning_rate": 5.299963067464889e-05,
"loss": 7.883408069610596,
"step": 510
},
{
"epoch": 0.020542713567839196,
"grad_norm": 1.9954074621200562,
"learning_rate": 5.2999618049108635e-05,
"loss": 7.8393635749816895,
"step": 511
},
{
"epoch": 0.020582914572864323,
"grad_norm": 1.8430662155151367,
"learning_rate": 5.2999605211378536e-05,
"loss": 8.044870376586914,
"step": 512
},
{
"epoch": 0.020623115577889446,
"grad_norm": 1.538697361946106,
"learning_rate": 5.29995921614588e-05,
"loss": 7.766447067260742,
"step": 513
},
{
"epoch": 0.020663316582914572,
"grad_norm": 1.8329964876174927,
"learning_rate": 5.299957889934964e-05,
"loss": 7.675707817077637,
"step": 514
},
{
"epoch": 0.020703517587939698,
"grad_norm": 1.912664532661438,
"learning_rate": 5.299956542505126e-05,
"loss": 7.806980133056641,
"step": 515
},
{
"epoch": 0.020743718592964824,
"grad_norm": 1.6501363515853882,
"learning_rate": 5.2999551738563886e-05,
"loss": 7.977140426635742,
"step": 516
},
{
"epoch": 0.02078391959798995,
"grad_norm": 1.587878704071045,
"learning_rate": 5.299953783988773e-05,
"loss": 7.906130790710449,
"step": 517
},
{
"epoch": 0.020824120603015077,
"grad_norm": 1.8340187072753906,
"learning_rate": 5.299952372902302e-05,
"loss": 8.050568580627441,
"step": 518
},
{
"epoch": 0.0208643216080402,
"grad_norm": 1.7699379920959473,
"learning_rate": 5.299950940596998e-05,
"loss": 7.694161415100098,
"step": 519
},
{
"epoch": 0.020904522613065326,
"grad_norm": 1.587458848953247,
"learning_rate": 5.299949487072883e-05,
"loss": 7.488152503967285,
"step": 520
},
{
"epoch": 0.020944723618090452,
"grad_norm": 1.8397084474563599,
"learning_rate": 5.299948012329981e-05,
"loss": 8.358592987060547,
"step": 521
},
{
"epoch": 0.02098492462311558,
"grad_norm": 1.9808828830718994,
"learning_rate": 5.2999465163683165e-05,
"loss": 7.547861576080322,
"step": 522
},
{
"epoch": 0.021025125628140705,
"grad_norm": 1.7588484287261963,
"learning_rate": 5.2999449991879125e-05,
"loss": 7.582889556884766,
"step": 523
},
{
"epoch": 0.021065326633165828,
"grad_norm": 1.610756278038025,
"learning_rate": 5.299943460788793e-05,
"loss": 7.885328769683838,
"step": 524
},
{
"epoch": 0.021105527638190954,
"grad_norm": 1.8191640377044678,
"learning_rate": 5.2999419011709836e-05,
"loss": 8.045345306396484,
"step": 525
},
{
"epoch": 0.02114572864321608,
"grad_norm": 1.916459321975708,
"learning_rate": 5.2999403203345086e-05,
"loss": 7.686132431030273,
"step": 526
},
{
"epoch": 0.021185929648241206,
"grad_norm": 2.328057050704956,
"learning_rate": 5.299938718279393e-05,
"loss": 8.074576377868652,
"step": 527
},
{
"epoch": 0.021226130653266333,
"grad_norm": 1.8764524459838867,
"learning_rate": 5.2999370950056634e-05,
"loss": 7.605907917022705,
"step": 528
},
{
"epoch": 0.02126633165829146,
"grad_norm": 1.5820814371109009,
"learning_rate": 5.299935450513345e-05,
"loss": 8.173661231994629,
"step": 529
},
{
"epoch": 0.021306532663316582,
"grad_norm": 1.6453853845596313,
"learning_rate": 5.299933784802465e-05,
"loss": 7.883947372436523,
"step": 530
},
{
"epoch": 0.021346733668341708,
"grad_norm": 1.675203800201416,
"learning_rate": 5.299932097873049e-05,
"loss": 7.591403007507324,
"step": 531
},
{
"epoch": 0.021386934673366834,
"grad_norm": 1.6879338026046753,
"learning_rate": 5.299930389725125e-05,
"loss": 7.751318454742432,
"step": 532
},
{
"epoch": 0.02142713567839196,
"grad_norm": 1.6978949308395386,
"learning_rate": 5.299928660358719e-05,
"loss": 7.733180046081543,
"step": 533
},
{
"epoch": 0.021467336683417087,
"grad_norm": 1.7834354639053345,
"learning_rate": 5.2999269097738604e-05,
"loss": 7.644957065582275,
"step": 534
},
{
"epoch": 0.02150753768844221,
"grad_norm": 1.8923128843307495,
"learning_rate": 5.2999251379705766e-05,
"loss": 7.867615699768066,
"step": 535
},
{
"epoch": 0.021547738693467336,
"grad_norm": 1.5811430215835571,
"learning_rate": 5.299923344948895e-05,
"loss": 7.421480655670166,
"step": 536
},
{
"epoch": 0.021587939698492462,
"grad_norm": 1.7423431873321533,
"learning_rate": 5.2999215307088454e-05,
"loss": 8.018625259399414,
"step": 537
},
{
"epoch": 0.02162814070351759,
"grad_norm": 1.677176833152771,
"learning_rate": 5.299919695250457e-05,
"loss": 8.08187484741211,
"step": 538
},
{
"epoch": 0.021668341708542715,
"grad_norm": 1.7374157905578613,
"learning_rate": 5.299917838573759e-05,
"loss": 7.832225799560547,
"step": 539
},
{
"epoch": 0.021708542713567838,
"grad_norm": 2.033163547515869,
"learning_rate": 5.299915960678781e-05,
"loss": 7.534571647644043,
"step": 540
},
{
"epoch": 0.021748743718592964,
"grad_norm": 1.772055983543396,
"learning_rate": 5.2999140615655525e-05,
"loss": 7.743981838226318,
"step": 541
},
{
"epoch": 0.02178894472361809,
"grad_norm": 2.0670437812805176,
"learning_rate": 5.2999121412341036e-05,
"loss": 8.108789443969727,
"step": 542
},
{
"epoch": 0.021829145728643216,
"grad_norm": 1.9782660007476807,
"learning_rate": 5.299910199684467e-05,
"loss": 7.593667030334473,
"step": 543
},
{
"epoch": 0.021869346733668343,
"grad_norm": 1.8748918771743774,
"learning_rate": 5.299908236916672e-05,
"loss": 7.89157772064209,
"step": 544
},
{
"epoch": 0.02190954773869347,
"grad_norm": 2.4620859622955322,
"learning_rate": 5.299906252930751e-05,
"loss": 8.07324504852295,
"step": 545
},
{
"epoch": 0.021949748743718592,
"grad_norm": 1.8238924741744995,
"learning_rate": 5.2999042477267355e-05,
"loss": 7.754039764404297,
"step": 546
},
{
"epoch": 0.021989949748743718,
"grad_norm": 1.8701344728469849,
"learning_rate": 5.299902221304657e-05,
"loss": 8.22531509399414,
"step": 547
},
{
"epoch": 0.022030150753768844,
"grad_norm": 1.597324252128601,
"learning_rate": 5.2999001736645494e-05,
"loss": 7.756403923034668,
"step": 548
},
{
"epoch": 0.02207035175879397,
"grad_norm": 1.7704119682312012,
"learning_rate": 5.2998981048064444e-05,
"loss": 7.665175914764404,
"step": 549
},
{
"epoch": 0.022110552763819097,
"grad_norm": 1.7645951509475708,
"learning_rate": 5.2998960147303755e-05,
"loss": 7.580150604248047,
"step": 550
},
{
"epoch": 0.02215075376884422,
"grad_norm": 1.5500779151916504,
"learning_rate": 5.299893903436376e-05,
"loss": 7.599634170532227,
"step": 551
},
{
"epoch": 0.022190954773869346,
"grad_norm": 1.856002688407898,
"learning_rate": 5.299891770924479e-05,
"loss": 7.610930919647217,
"step": 552
},
{
"epoch": 0.022231155778894472,
"grad_norm": 2.392806053161621,
"learning_rate": 5.2998896171947194e-05,
"loss": 7.465909481048584,
"step": 553
},
{
"epoch": 0.0222713567839196,
"grad_norm": 1.8713092803955078,
"learning_rate": 5.2998874422471324e-05,
"loss": 7.834880828857422,
"step": 554
},
{
"epoch": 0.022311557788944725,
"grad_norm": 1.9191861152648926,
"learning_rate": 5.299885246081752e-05,
"loss": 7.771798133850098,
"step": 555
},
{
"epoch": 0.022351758793969848,
"grad_norm": 1.8119090795516968,
"learning_rate": 5.2998830286986127e-05,
"loss": 7.759174346923828,
"step": 556
},
{
"epoch": 0.022391959798994974,
"grad_norm": 1.6752859354019165,
"learning_rate": 5.299880790097751e-05,
"loss": 7.623655319213867,
"step": 557
},
{
"epoch": 0.0224321608040201,
"grad_norm": 1.8278697729110718,
"learning_rate": 5.2998785302792036e-05,
"loss": 8.344066619873047,
"step": 558
},
{
"epoch": 0.022472361809045226,
"grad_norm": 1.7110530138015747,
"learning_rate": 5.299876249243005e-05,
"loss": 7.565323829650879,
"step": 559
},
{
"epoch": 0.022512562814070353,
"grad_norm": 2.0177745819091797,
"learning_rate": 5.2998739469891925e-05,
"loss": 8.117171287536621,
"step": 560
},
{
"epoch": 0.02255276381909548,
"grad_norm": 1.6648521423339844,
"learning_rate": 5.299871623517802e-05,
"loss": 7.971223831176758,
"step": 561
},
{
"epoch": 0.0225929648241206,
"grad_norm": 1.8696317672729492,
"learning_rate": 5.299869278828873e-05,
"loss": 7.7179856300354,
"step": 562
},
{
"epoch": 0.022633165829145728,
"grad_norm": 1.6487462520599365,
"learning_rate": 5.2998669129224397e-05,
"loss": 7.890140533447266,
"step": 563
},
{
"epoch": 0.022673366834170854,
"grad_norm": 1.7422339916229248,
"learning_rate": 5.2998645257985433e-05,
"loss": 7.791595458984375,
"step": 564
},
{
"epoch": 0.02271356783919598,
"grad_norm": 1.6526107788085938,
"learning_rate": 5.2998621174572205e-05,
"loss": 7.630365371704102,
"step": 565
},
{
"epoch": 0.022753768844221107,
"grad_norm": 1.9440349340438843,
"learning_rate": 5.29985968789851e-05,
"loss": 7.529691696166992,
"step": 566
},
{
"epoch": 0.02279396984924623,
"grad_norm": 1.779954195022583,
"learning_rate": 5.2998572371224496e-05,
"loss": 7.9304962158203125,
"step": 567
},
{
"epoch": 0.022834170854271356,
"grad_norm": 1.8104673624038696,
"learning_rate": 5.29985476512908e-05,
"loss": 7.635679244995117,
"step": 568
},
{
"epoch": 0.022874371859296482,
"grad_norm": 1.7954227924346924,
"learning_rate": 5.2998522719184414e-05,
"loss": 7.913227081298828,
"step": 569
},
{
"epoch": 0.02291457286432161,
"grad_norm": 1.6429694890975952,
"learning_rate": 5.299849757490572e-05,
"loss": 7.8194732666015625,
"step": 570
},
{
"epoch": 0.022954773869346735,
"grad_norm": 1.7629566192626953,
"learning_rate": 5.299847221845513e-05,
"loss": 7.470943450927734,
"step": 571
},
{
"epoch": 0.02299497487437186,
"grad_norm": 2.051751136779785,
"learning_rate": 5.299844664983305e-05,
"loss": 7.58784294128418,
"step": 572
},
{
"epoch": 0.023035175879396984,
"grad_norm": 1.6047335863113403,
"learning_rate": 5.2998420869039885e-05,
"loss": 7.960485458374023,
"step": 573
},
{
"epoch": 0.02307537688442211,
"grad_norm": 1.9012272357940674,
"learning_rate": 5.2998394876076055e-05,
"loss": 7.445609092712402,
"step": 574
},
{
"epoch": 0.023115577889447236,
"grad_norm": 1.6588935852050781,
"learning_rate": 5.299836867094196e-05,
"loss": 7.754929542541504,
"step": 575
},
{
"epoch": 0.023155778894472363,
"grad_norm": 1.6666665077209473,
"learning_rate": 5.2998342253638045e-05,
"loss": 8.129451751708984,
"step": 576
},
{
"epoch": 0.02319597989949749,
"grad_norm": 1.660367488861084,
"learning_rate": 5.299831562416471e-05,
"loss": 8.151827812194824,
"step": 577
},
{
"epoch": 0.02323618090452261,
"grad_norm": 1.781943917274475,
"learning_rate": 5.29982887825224e-05,
"loss": 7.77730655670166,
"step": 578
},
{
"epoch": 0.023276381909547738,
"grad_norm": 1.8823766708374023,
"learning_rate": 5.299826172871153e-05,
"loss": 7.829004287719727,
"step": 579
},
{
"epoch": 0.023316582914572864,
"grad_norm": 1.661432147026062,
"learning_rate": 5.2998234462732544e-05,
"loss": 7.661782264709473,
"step": 580
},
{
"epoch": 0.02335678391959799,
"grad_norm": 1.6148476600646973,
"learning_rate": 5.2998206984585867e-05,
"loss": 7.608725547790527,
"step": 581
},
{
"epoch": 0.023396984924623117,
"grad_norm": 1.7564600706100464,
"learning_rate": 5.299817929427195e-05,
"loss": 7.824212074279785,
"step": 582
},
{
"epoch": 0.02343718592964824,
"grad_norm": 1.647735834121704,
"learning_rate": 5.299815139179123e-05,
"loss": 7.843659400939941,
"step": 583
},
{
"epoch": 0.023477386934673366,
"grad_norm": 2.532902240753174,
"learning_rate": 5.299812327714416e-05,
"loss": 7.733049392700195,
"step": 584
},
{
"epoch": 0.023517587939698492,
"grad_norm": 1.8626370429992676,
"learning_rate": 5.299809495033119e-05,
"loss": 8.09103775024414,
"step": 585
},
{
"epoch": 0.02355778894472362,
"grad_norm": 1.7256304025650024,
"learning_rate": 5.299806641135276e-05,
"loss": 7.7607927322387695,
"step": 586
},
{
"epoch": 0.023597989949748745,
"grad_norm": 1.8326709270477295,
"learning_rate": 5.2998037660209345e-05,
"loss": 7.975545406341553,
"step": 587
},
{
"epoch": 0.02363819095477387,
"grad_norm": 1.5843669176101685,
"learning_rate": 5.29980086969014e-05,
"loss": 7.893915176391602,
"step": 588
},
{
"epoch": 0.023678391959798994,
"grad_norm": 1.5900360345840454,
"learning_rate": 5.299797952142937e-05,
"loss": 7.735443115234375,
"step": 589
},
{
"epoch": 0.02371859296482412,
"grad_norm": 1.8637281656265259,
"learning_rate": 5.299795013379375e-05,
"loss": 7.408278942108154,
"step": 590
},
{
"epoch": 0.023758793969849246,
"grad_norm": 1.6430720090866089,
"learning_rate": 5.2997920533995e-05,
"loss": 7.856941223144531,
"step": 591
},
{
"epoch": 0.023798994974874373,
"grad_norm": 1.9916696548461914,
"learning_rate": 5.2997890722033594e-05,
"loss": 8.090843200683594,
"step": 592
},
{
"epoch": 0.0238391959798995,
"grad_norm": 1.5390474796295166,
"learning_rate": 5.2997860697910005e-05,
"loss": 7.599443435668945,
"step": 593
},
{
"epoch": 0.02387939698492462,
"grad_norm": 2.1709444522857666,
"learning_rate": 5.299783046162472e-05,
"loss": 7.718976974487305,
"step": 594
},
{
"epoch": 0.023919597989949748,
"grad_norm": 1.7341334819793701,
"learning_rate": 5.299780001317822e-05,
"loss": 7.939218521118164,
"step": 595
},
{
"epoch": 0.023959798994974874,
"grad_norm": 1.7878673076629639,
"learning_rate": 5.2997769352571e-05,
"loss": 8.116238594055176,
"step": 596
},
{
"epoch": 0.024,
"grad_norm": 1.6568694114685059,
"learning_rate": 5.2997738479803534e-05,
"loss": 8.052183151245117,
"step": 597
},
{
"epoch": 0.024040201005025127,
"grad_norm": 2.0456948280334473,
"learning_rate": 5.299770739487633e-05,
"loss": 7.878899574279785,
"step": 598
},
{
"epoch": 0.02408040201005025,
"grad_norm": 1.8307068347930908,
"learning_rate": 5.299767609778988e-05,
"loss": 7.602692604064941,
"step": 599
},
{
"epoch": 0.024120603015075376,
"grad_norm": 2.237743377685547,
"learning_rate": 5.29976445885447e-05,
"loss": 7.8438520431518555,
"step": 600
},
{
"epoch": 0.024120603015075376,
"eval_loss": 7.825066566467285,
"eval_runtime": 120.5981,
"eval_samples_per_second": 4.146,
"eval_steps_per_second": 0.522,
"step": 600
},
{
"epoch": 0.024160804020100502,
"grad_norm": 1.8920879364013672,
"learning_rate": 0.0,
"loss": 7.975615501403809,
"step": 601
},
{
"epoch": 0.02420100502512563,
"grad_norm": 1.804730772972107,
"learning_rate": 2.6e-07,
"loss": 7.884779453277588,
"step": 602
},
{
"epoch": 0.024241206030150755,
"grad_norm": 1.7931219339370728,
"learning_rate": 5.2e-07,
"loss": 7.646689414978027,
"step": 603
},
{
"epoch": 0.02428140703517588,
"grad_norm": 2.0185418128967285,
"learning_rate": 7.799999999999999e-07,
"loss": 8.194418907165527,
"step": 604
},
{
"epoch": 0.024321608040201004,
"grad_norm": 2.203396797180176,
"learning_rate": 1.04e-06,
"loss": 7.11131477355957,
"step": 605
},
{
"epoch": 0.02436180904522613,
"grad_norm": 1.7545979022979736,
"learning_rate": 1.3e-06,
"loss": 7.693672180175781,
"step": 606
},
{
"epoch": 0.024402010050251256,
"grad_norm": 1.6657663583755493,
"learning_rate": 1.5599999999999999e-06,
"loss": 7.7552080154418945,
"step": 607
},
{
"epoch": 0.024442211055276383,
"grad_norm": 2.0895936489105225,
"learning_rate": 1.8200000000000002e-06,
"loss": 7.89845085144043,
"step": 608
},
{
"epoch": 0.02448241206030151,
"grad_norm": 1.9905946254730225,
"learning_rate": 2.08e-06,
"loss": 7.523114204406738,
"step": 609
},
{
"epoch": 0.02452261306532663,
"grad_norm": 1.6749087572097778,
"learning_rate": 2.3399999999999996e-06,
"loss": 7.867109298706055,
"step": 610
},
{
"epoch": 0.024562814070351758,
"grad_norm": 1.5739293098449707,
"learning_rate": 2.6e-06,
"loss": 7.946324348449707,
"step": 611
},
{
"epoch": 0.024603015075376884,
"grad_norm": 1.5848045349121094,
"learning_rate": 2.8599999999999997e-06,
"loss": 7.648571491241455,
"step": 612
},
{
"epoch": 0.02464321608040201,
"grad_norm": 1.7416894435882568,
"learning_rate": 3.1199999999999998e-06,
"loss": 7.346608638763428,
"step": 613
},
{
"epoch": 0.024683417085427137,
"grad_norm": 1.5669807195663452,
"learning_rate": 3.38e-06,
"loss": 7.868875980377197,
"step": 614
},
{
"epoch": 0.024723618090452263,
"grad_norm": 2.2537267208099365,
"learning_rate": 3.6400000000000003e-06,
"loss": 8.203691482543945,
"step": 615
},
{
"epoch": 0.024763819095477386,
"grad_norm": 1.6477341651916504,
"learning_rate": 3.9e-06,
"loss": 7.176671981811523,
"step": 616
},
{
"epoch": 0.024804020100502512,
"grad_norm": 1.4628090858459473,
"learning_rate": 4.16e-06,
"loss": 7.877170562744141,
"step": 617
},
{
"epoch": 0.02484422110552764,
"grad_norm": 2.370819091796875,
"learning_rate": 4.42e-06,
"loss": 7.460579872131348,
"step": 618
},
{
"epoch": 0.024884422110552765,
"grad_norm": 1.6509450674057007,
"learning_rate": 4.679999999999999e-06,
"loss": 7.809412002563477,
"step": 619
},
{
"epoch": 0.02492462311557789,
"grad_norm": 1.858579158782959,
"learning_rate": 4.94e-06,
"loss": 7.691108703613281,
"step": 620
},
{
"epoch": 0.024964824120603014,
"grad_norm": 1.556200623512268,
"learning_rate": 5.2e-06,
"loss": 8.241106986999512,
"step": 621
},
{
"epoch": 0.02500502512562814,
"grad_norm": 1.7693449258804321,
"learning_rate": 5.459999999999999e-06,
"loss": 7.748850345611572,
"step": 622
},
{
"epoch": 0.025045226130653266,
"grad_norm": 1.6281565427780151,
"learning_rate": 5.7199999999999994e-06,
"loss": 8.134521484375,
"step": 623
},
{
"epoch": 0.025085427135678393,
"grad_norm": 1.5439441204071045,
"learning_rate": 5.9799999999999995e-06,
"loss": 7.651298522949219,
"step": 624
},
{
"epoch": 0.02512562814070352,
"grad_norm": 1.681318759918213,
"learning_rate": 6.2399999999999995e-06,
"loss": 7.602551460266113,
"step": 625
},
{
"epoch": 0.02516582914572864,
"grad_norm": 1.6547749042510986,
"learning_rate": 6.5e-06,
"loss": 7.208373069763184,
"step": 626
},
{
"epoch": 0.025206030150753768,
"grad_norm": 2.576897382736206,
"learning_rate": 6.76e-06,
"loss": 7.682596206665039,
"step": 627
},
{
"epoch": 0.025246231155778894,
"grad_norm": 1.5396480560302734,
"learning_rate": 7.02e-06,
"loss": 7.557615280151367,
"step": 628
},
{
"epoch": 0.02528643216080402,
"grad_norm": 1.6940165758132935,
"learning_rate": 7.280000000000001e-06,
"loss": 8.100613594055176,
"step": 629
},
{
"epoch": 0.025326633165829147,
"grad_norm": 1.675053596496582,
"learning_rate": 7.539999999999999e-06,
"loss": 8.093732833862305,
"step": 630
},
{
"epoch": 0.025366834170854273,
"grad_norm": 1.632793664932251,
"learning_rate": 7.8e-06,
"loss": 8.050073623657227,
"step": 631
},
{
"epoch": 0.025407035175879396,
"grad_norm": 1.4590184688568115,
"learning_rate": 8.059999999999999e-06,
"loss": 7.758357048034668,
"step": 632
},
{
"epoch": 0.025447236180904522,
"grad_norm": 1.7311021089553833,
"learning_rate": 8.32e-06,
"loss": 7.842282295227051,
"step": 633
},
{
"epoch": 0.02548743718592965,
"grad_norm": 1.727177381515503,
"learning_rate": 8.58e-06,
"loss": 7.726874828338623,
"step": 634
},
{
"epoch": 0.025527638190954775,
"grad_norm": 1.5760232210159302,
"learning_rate": 8.84e-06,
"loss": 7.824732780456543,
"step": 635
},
{
"epoch": 0.0255678391959799,
"grad_norm": 1.5611227750778198,
"learning_rate": 9.1e-06,
"loss": 7.898661136627197,
"step": 636
},
{
"epoch": 0.025608040201005024,
"grad_norm": 1.8225016593933105,
"learning_rate": 9.359999999999998e-06,
"loss": 7.220385551452637,
"step": 637
},
{
"epoch": 0.02564824120603015,
"grad_norm": 1.7629814147949219,
"learning_rate": 9.62e-06,
"loss": 7.74628210067749,
"step": 638
},
{
"epoch": 0.025688442211055276,
"grad_norm": 1.474403738975525,
"learning_rate": 9.88e-06,
"loss": 7.715187072753906,
"step": 639
},
{
"epoch": 0.025728643216080403,
"grad_norm": 1.50594961643219,
"learning_rate": 1.014e-05,
"loss": 7.829171180725098,
"step": 640
},
{
"epoch": 0.02576884422110553,
"grad_norm": 1.7559689283370972,
"learning_rate": 1.04e-05,
"loss": 7.708507537841797,
"step": 641
},
{
"epoch": 0.02580904522613065,
"grad_norm": 1.732669711112976,
"learning_rate": 1.0659999999999998e-05,
"loss": 7.371837615966797,
"step": 642
},
{
"epoch": 0.025849246231155778,
"grad_norm": 1.5631146430969238,
"learning_rate": 1.0919999999999999e-05,
"loss": 7.8240556716918945,
"step": 643
},
{
"epoch": 0.025889447236180904,
"grad_norm": 1.6416966915130615,
"learning_rate": 1.118e-05,
"loss": 8.102253913879395,
"step": 644
},
{
"epoch": 0.02592964824120603,
"grad_norm": 1.5926142930984497,
"learning_rate": 1.1439999999999999e-05,
"loss": 8.116301536560059,
"step": 645
},
{
"epoch": 0.025969849246231157,
"grad_norm": 1.6105057001113892,
"learning_rate": 1.17e-05,
"loss": 7.722078323364258,
"step": 646
},
{
"epoch": 0.026010050251256283,
"grad_norm": 1.8119257688522339,
"learning_rate": 1.1959999999999999e-05,
"loss": 8.077878952026367,
"step": 647
},
{
"epoch": 0.026050251256281406,
"grad_norm": 1.5658546686172485,
"learning_rate": 1.2219999999999998e-05,
"loss": 7.712865829467773,
"step": 648
},
{
"epoch": 0.026090452261306532,
"grad_norm": 1.6031376123428345,
"learning_rate": 1.2479999999999999e-05,
"loss": 7.477446556091309,
"step": 649
},
{
"epoch": 0.02613065326633166,
"grad_norm": 1.6984279155731201,
"learning_rate": 1.2739999999999998e-05,
"loss": 8.283685684204102,
"step": 650
},
{
"epoch": 0.026170854271356785,
"grad_norm": 2.2368686199188232,
"learning_rate": 1.3e-05,
"loss": 8.10506820678711,
"step": 651
},
{
"epoch": 0.02621105527638191,
"grad_norm": 1.6849017143249512,
"learning_rate": 1.2999999973975966e-05,
"loss": 7.41166877746582,
"step": 652
},
{
"epoch": 0.026251256281407034,
"grad_norm": 1.7519370317459106,
"learning_rate": 1.2999999895903868e-05,
"loss": 8.087525367736816,
"step": 653
},
{
"epoch": 0.02629145728643216,
"grad_norm": 2.5718979835510254,
"learning_rate": 1.2999999765783707e-05,
"loss": 7.276345729827881,
"step": 654
},
{
"epoch": 0.026331658291457286,
"grad_norm": 1.6353685855865479,
"learning_rate": 1.2999999583615484e-05,
"loss": 7.553966522216797,
"step": 655
},
{
"epoch": 0.026371859296482413,
"grad_norm": 1.6945093870162964,
"learning_rate": 1.2999999349399202e-05,
"loss": 7.364789009094238,
"step": 656
},
{
"epoch": 0.02641206030150754,
"grad_norm": 1.8085143566131592,
"learning_rate": 1.2999999063134866e-05,
"loss": 7.979491233825684,
"step": 657
},
{
"epoch": 0.02645226130653266,
"grad_norm": 2.1697587966918945,
"learning_rate": 1.299999872482248e-05,
"loss": 7.668868064880371,
"step": 658
},
{
"epoch": 0.026492462311557788,
"grad_norm": 1.6411314010620117,
"learning_rate": 1.2999998334462047e-05,
"loss": 7.368921756744385,
"step": 659
},
{
"epoch": 0.026532663316582914,
"grad_norm": 2.044774055480957,
"learning_rate": 1.2999997892053578e-05,
"loss": 7.849125862121582,
"step": 660
},
{
"epoch": 0.02657286432160804,
"grad_norm": 1.7394928932189941,
"learning_rate": 1.2999997397597075e-05,
"loss": 7.302081108093262,
"step": 661
},
{
"epoch": 0.026613065326633167,
"grad_norm": 1.921125888824463,
"learning_rate": 1.2999996851092549e-05,
"loss": 7.384099960327148,
"step": 662
},
{
"epoch": 0.026653266331658293,
"grad_norm": 1.6810789108276367,
"learning_rate": 1.2999996252540008e-05,
"loss": 7.928728103637695,
"step": 663
},
{
"epoch": 0.026693467336683416,
"grad_norm": 1.9991676807403564,
"learning_rate": 1.299999560193946e-05,
"loss": 7.6297221183776855,
"step": 664
},
{
"epoch": 0.026733668341708542,
"grad_norm": 1.600643277168274,
"learning_rate": 1.2999994899290919e-05,
"loss": 7.677794456481934,
"step": 665
},
{
"epoch": 0.02677386934673367,
"grad_norm": 1.6687604188919067,
"learning_rate": 1.2999994144594395e-05,
"loss": 7.5907487869262695,
"step": 666
},
{
"epoch": 0.026814070351758795,
"grad_norm": 1.6812160015106201,
"learning_rate": 1.29999933378499e-05,
"loss": 7.978593349456787,
"step": 667
},
{
"epoch": 0.02685427135678392,
"grad_norm": 2.0290145874023438,
"learning_rate": 1.2999992479057443e-05,
"loss": 7.636195182800293,
"step": 668
},
{
"epoch": 0.026894472361809044,
"grad_norm": 1.5601996183395386,
"learning_rate": 1.2999991568217043e-05,
"loss": 8.067818641662598,
"step": 669
},
{
"epoch": 0.02693467336683417,
"grad_norm": 1.8225313425064087,
"learning_rate": 1.2999990605328711e-05,
"loss": 7.504214286804199,
"step": 670
},
{
"epoch": 0.026974874371859296,
"grad_norm": 1.6736478805541992,
"learning_rate": 1.2999989590392467e-05,
"loss": 7.672181129455566,
"step": 671
},
{
"epoch": 0.027015075376884422,
"grad_norm": 1.6976789236068726,
"learning_rate": 1.2999988523408322e-05,
"loss": 7.494943618774414,
"step": 672
},
{
"epoch": 0.02705527638190955,
"grad_norm": 1.501541256904602,
"learning_rate": 1.2999987404376297e-05,
"loss": 7.743948459625244,
"step": 673
},
{
"epoch": 0.027095477386934675,
"grad_norm": 1.9164901971817017,
"learning_rate": 1.2999986233296407e-05,
"loss": 7.98082160949707,
"step": 674
},
{
"epoch": 0.027135678391959798,
"grad_norm": 1.6361418962478638,
"learning_rate": 1.2999985010168674e-05,
"loss": 7.7127814292907715,
"step": 675
},
{
"epoch": 0.027175879396984924,
"grad_norm": 1.7240151166915894,
"learning_rate": 1.2999983734993114e-05,
"loss": 7.8380842208862305,
"step": 676
},
{
"epoch": 0.02721608040201005,
"grad_norm": 1.6175848245620728,
"learning_rate": 1.299998240776975e-05,
"loss": 7.748377799987793,
"step": 677
},
{
"epoch": 0.027256281407035177,
"grad_norm": 1.693701148033142,
"learning_rate": 1.2999981028498602e-05,
"loss": 7.688663005828857,
"step": 678
},
{
"epoch": 0.027296482412060303,
"grad_norm": 1.7469022274017334,
"learning_rate": 1.2999979597179692e-05,
"loss": 7.494562149047852,
"step": 679
},
{
"epoch": 0.027336683417085426,
"grad_norm": 2.1011221408843994,
"learning_rate": 1.2999978113813043e-05,
"loss": 7.932559013366699,
"step": 680
},
{
"epoch": 0.027376884422110552,
"grad_norm": 1.722305178642273,
"learning_rate": 1.2999976578398681e-05,
"loss": 8.193719863891602,
"step": 681
},
{
"epoch": 0.02741708542713568,
"grad_norm": 1.63822340965271,
"learning_rate": 1.2999974990936626e-05,
"loss": 7.757412910461426,
"step": 682
},
{
"epoch": 0.027457286432160805,
"grad_norm": 1.6862330436706543,
"learning_rate": 1.2999973351426908e-05,
"loss": 8.049449920654297,
"step": 683
},
{
"epoch": 0.02749748743718593,
"grad_norm": 1.613176941871643,
"learning_rate": 1.299997165986955e-05,
"loss": 7.589936256408691,
"step": 684
},
{
"epoch": 0.027537688442211054,
"grad_norm": 1.8277050256729126,
"learning_rate": 1.2999969916264583e-05,
"loss": 7.637495517730713,
"step": 685
},
{
"epoch": 0.02757788944723618,
"grad_norm": 1.7138813734054565,
"learning_rate": 1.2999968120612032e-05,
"loss": 7.365002632141113,
"step": 686
},
{
"epoch": 0.027618090452261306,
"grad_norm": 1.7700183391571045,
"learning_rate": 1.2999966272911925e-05,
"loss": 7.454729080200195,
"step": 687
},
{
"epoch": 0.027658291457286432,
"grad_norm": 1.7409950494766235,
"learning_rate": 1.2999964373164293e-05,
"loss": 7.583250999450684,
"step": 688
},
{
"epoch": 0.02769849246231156,
"grad_norm": 1.720686912536621,
"learning_rate": 1.2999962421369164e-05,
"loss": 8.15003490447998,
"step": 689
},
{
"epoch": 0.027738693467336685,
"grad_norm": 1.7570269107818604,
"learning_rate": 1.2999960417526575e-05,
"loss": 8.234285354614258,
"step": 690
},
{
"epoch": 0.027778894472361808,
"grad_norm": 1.6660727262496948,
"learning_rate": 1.2999958361636552e-05,
"loss": 8.061471939086914,
"step": 691
},
{
"epoch": 0.027819095477386934,
"grad_norm": 1.6924585103988647,
"learning_rate": 1.2999956253699131e-05,
"loss": 8.223845481872559,
"step": 692
},
{
"epoch": 0.02785929648241206,
"grad_norm": 1.5804903507232666,
"learning_rate": 1.2999954093714348e-05,
"loss": 7.766111373901367,
"step": 693
},
{
"epoch": 0.027899497487437187,
"grad_norm": 1.6386661529541016,
"learning_rate": 1.2999951881682231e-05,
"loss": 7.753026008605957,
"step": 694
},
{
"epoch": 0.027939698492462313,
"grad_norm": 1.7316641807556152,
"learning_rate": 1.2999949617602821e-05,
"loss": 8.301666259765625,
"step": 695
},
{
"epoch": 0.027979899497487436,
"grad_norm": 1.559029221534729,
"learning_rate": 1.2999947301476152e-05,
"loss": 7.944850921630859,
"step": 696
},
{
"epoch": 0.028020100502512562,
"grad_norm": 1.823709487915039,
"learning_rate": 1.2999944933302261e-05,
"loss": 7.552277565002441,
"step": 697
},
{
"epoch": 0.028060301507537688,
"grad_norm": 1.7210633754730225,
"learning_rate": 1.299994251308119e-05,
"loss": 7.7181878089904785,
"step": 698
},
{
"epoch": 0.028100502512562815,
"grad_norm": 1.889969825744629,
"learning_rate": 1.299994004081297e-05,
"loss": 7.771721363067627,
"step": 699
},
{
"epoch": 0.02814070351758794,
"grad_norm": 1.9056483507156372,
"learning_rate": 1.2999937516497645e-05,
"loss": 7.3271613121032715,
"step": 700
},
{
"epoch": 0.028180904522613064,
"grad_norm": 1.6055564880371094,
"learning_rate": 1.2999934940135258e-05,
"loss": 8.017894744873047,
"step": 701
},
{
"epoch": 0.02822110552763819,
"grad_norm": 1.5920822620391846,
"learning_rate": 1.2999932311725846e-05,
"loss": 7.678212642669678,
"step": 702
},
{
"epoch": 0.028261306532663316,
"grad_norm": 1.6964926719665527,
"learning_rate": 1.2999929631269452e-05,
"loss": 8.200109481811523,
"step": 703
},
{
"epoch": 0.028301507537688442,
"grad_norm": 1.8752659559249878,
"learning_rate": 1.2999926898766122e-05,
"loss": 7.5684356689453125,
"step": 704
},
{
"epoch": 0.02834170854271357,
"grad_norm": 1.6784663200378418,
"learning_rate": 1.2999924114215896e-05,
"loss": 7.5030741691589355,
"step": 705
},
{
"epoch": 0.028381909547738695,
"grad_norm": 1.6283483505249023,
"learning_rate": 1.299992127761882e-05,
"loss": 7.645549774169922,
"step": 706
},
{
"epoch": 0.028422110552763818,
"grad_norm": 1.7028249502182007,
"learning_rate": 1.2999918388974938e-05,
"loss": 7.4748430252075195,
"step": 707
},
{
"epoch": 0.028462311557788944,
"grad_norm": 2.0857505798339844,
"learning_rate": 1.2999915448284298e-05,
"loss": 7.503174304962158,
"step": 708
},
{
"epoch": 0.02850251256281407,
"grad_norm": 1.603746771812439,
"learning_rate": 1.2999912455546948e-05,
"loss": 7.723291397094727,
"step": 709
},
{
"epoch": 0.028542713567839197,
"grad_norm": 1.6387344598770142,
"learning_rate": 1.2999909410762934e-05,
"loss": 7.67912483215332,
"step": 710
},
{
"epoch": 0.028582914572864323,
"grad_norm": 1.7044860124588013,
"learning_rate": 1.2999906313932305e-05,
"loss": 7.539193153381348,
"step": 711
},
{
"epoch": 0.028623115577889446,
"grad_norm": 1.6376256942749023,
"learning_rate": 1.2999903165055112e-05,
"loss": 7.75665283203125,
"step": 712
},
{
"epoch": 0.028663316582914572,
"grad_norm": 1.7152029275894165,
"learning_rate": 1.2999899964131403e-05,
"loss": 7.643871784210205,
"step": 713
},
{
"epoch": 0.028703517587939698,
"grad_norm": 1.6375702619552612,
"learning_rate": 1.2999896711161232e-05,
"loss": 7.775535583496094,
"step": 714
},
{
"epoch": 0.028743718592964825,
"grad_norm": 1.7719043493270874,
"learning_rate": 1.299989340614465e-05,
"loss": 7.952023506164551,
"step": 715
},
{
"epoch": 0.02878391959798995,
"grad_norm": 1.7259770631790161,
"learning_rate": 1.2999890049081707e-05,
"loss": 7.844346046447754,
"step": 716
},
{
"epoch": 0.028824120603015077,
"grad_norm": 2.2535629272460938,
"learning_rate": 1.2999886639972461e-05,
"loss": 8.326813697814941,
"step": 717
},
{
"epoch": 0.0288643216080402,
"grad_norm": 1.8590424060821533,
"learning_rate": 1.2999883178816967e-05,
"loss": 7.557680130004883,
"step": 718
},
{
"epoch": 0.028904522613065326,
"grad_norm": 1.683707594871521,
"learning_rate": 1.2999879665615276e-05,
"loss": 7.919500350952148,
"step": 719
},
{
"epoch": 0.028944723618090452,
"grad_norm": 2.2539894580841064,
"learning_rate": 1.2999876100367448e-05,
"loss": 7.576954364776611,
"step": 720
},
{
"epoch": 0.02898492462311558,
"grad_norm": 1.8476099967956543,
"learning_rate": 1.299987248307354e-05,
"loss": 7.6735615730285645,
"step": 721
},
{
"epoch": 0.029025125628140705,
"grad_norm": 1.721079707145691,
"learning_rate": 1.2999868813733605e-05,
"loss": 7.736406326293945,
"step": 722
},
{
"epoch": 0.029065326633165828,
"grad_norm": 1.5681211948394775,
"learning_rate": 1.299986509234771e-05,
"loss": 7.596938133239746,
"step": 723
},
{
"epoch": 0.029105527638190954,
"grad_norm": 1.7236958742141724,
"learning_rate": 1.2999861318915906e-05,
"loss": 7.854140281677246,
"step": 724
},
{
"epoch": 0.02914572864321608,
"grad_norm": 1.7181516885757446,
"learning_rate": 1.299985749343826e-05,
"loss": 7.923806190490723,
"step": 725
},
{
"epoch": 0.029185929648241207,
"grad_norm": 1.5900862216949463,
"learning_rate": 1.2999853615914831e-05,
"loss": 8.000762939453125,
"step": 726
},
{
"epoch": 0.029226130653266333,
"grad_norm": 1.7780818939208984,
"learning_rate": 1.299984968634568e-05,
"loss": 7.6512556076049805,
"step": 727
},
{
"epoch": 0.029266331658291456,
"grad_norm": 1.610883355140686,
"learning_rate": 1.2999845704730872e-05,
"loss": 7.961952209472656,
"step": 728
},
{
"epoch": 0.029306532663316582,
"grad_norm": 2.525376081466675,
"learning_rate": 1.2999841671070468e-05,
"loss": 8.132668495178223,
"step": 729
},
{
"epoch": 0.029346733668341708,
"grad_norm": 1.6752312183380127,
"learning_rate": 1.2999837585364534e-05,
"loss": 7.704312324523926,
"step": 730
},
{
"epoch": 0.029386934673366834,
"grad_norm": 1.9367809295654297,
"learning_rate": 1.2999833447613139e-05,
"loss": 7.275273323059082,
"step": 731
},
{
"epoch": 0.02942713567839196,
"grad_norm": 1.6626040935516357,
"learning_rate": 1.2999829257816343e-05,
"loss": 7.716578483581543,
"step": 732
},
{
"epoch": 0.029467336683417087,
"grad_norm": 1.587276577949524,
"learning_rate": 1.2999825015974215e-05,
"loss": 7.886663913726807,
"step": 733
},
{
"epoch": 0.02950753768844221,
"grad_norm": 1.8631511926651,
"learning_rate": 1.2999820722086828e-05,
"loss": 7.989174842834473,
"step": 734
},
{
"epoch": 0.029547738693467336,
"grad_norm": 1.737877368927002,
"learning_rate": 1.2999816376154243e-05,
"loss": 7.644630432128906,
"step": 735
},
{
"epoch": 0.029587939698492462,
"grad_norm": 1.8736320734024048,
"learning_rate": 1.2999811978176535e-05,
"loss": 7.840444564819336,
"step": 736
},
{
"epoch": 0.02962814070351759,
"grad_norm": 1.7689927816390991,
"learning_rate": 1.2999807528153772e-05,
"loss": 7.6608099937438965,
"step": 737
},
{
"epoch": 0.029668341708542715,
"grad_norm": 1.7446316480636597,
"learning_rate": 1.2999803026086027e-05,
"loss": 7.313197612762451,
"step": 738
},
{
"epoch": 0.029708542713567838,
"grad_norm": 1.7681705951690674,
"learning_rate": 1.299979847197337e-05,
"loss": 7.974023818969727,
"step": 739
},
{
"epoch": 0.029748743718592964,
"grad_norm": 1.6120606660842896,
"learning_rate": 1.2999793865815875e-05,
"loss": 7.403453350067139,
"step": 740
},
{
"epoch": 0.02978894472361809,
"grad_norm": 1.6234232187271118,
"learning_rate": 1.2999789207613619e-05,
"loss": 7.652613639831543,
"step": 741
},
{
"epoch": 0.029829145728643217,
"grad_norm": 2.039757490158081,
"learning_rate": 1.2999784497366667e-05,
"loss": 7.602209568023682,
"step": 742
},
{
"epoch": 0.029869346733668343,
"grad_norm": 1.6783447265625,
"learning_rate": 1.2999779735075107e-05,
"loss": 7.83221960067749,
"step": 743
},
{
"epoch": 0.029909547738693466,
"grad_norm": 2.0063319206237793,
"learning_rate": 1.2999774920739006e-05,
"loss": 7.826501846313477,
"step": 744
},
{
"epoch": 0.029949748743718592,
"grad_norm": 1.714544653892517,
"learning_rate": 1.2999770054358444e-05,
"loss": 7.533691883087158,
"step": 745
},
{
"epoch": 0.029989949748743718,
"grad_norm": 1.7687616348266602,
"learning_rate": 1.29997651359335e-05,
"loss": 7.624152660369873,
"step": 746
},
{
"epoch": 0.030030150753768844,
"grad_norm": 1.7901867628097534,
"learning_rate": 1.2999760165464254e-05,
"loss": 7.63176155090332,
"step": 747
},
{
"epoch": 0.03007035175879397,
"grad_norm": 1.5813881158828735,
"learning_rate": 1.299975514295078e-05,
"loss": 7.477207660675049,
"step": 748
},
{
"epoch": 0.030110552763819097,
"grad_norm": 1.8973941802978516,
"learning_rate": 1.2999750068393164e-05,
"loss": 7.992856025695801,
"step": 749
},
{
"epoch": 0.03015075376884422,
"grad_norm": 1.7212083339691162,
"learning_rate": 1.2999744941791486e-05,
"loss": 7.554285526275635,
"step": 750
},
{
"epoch": 0.030190954773869346,
"grad_norm": 1.7275336980819702,
"learning_rate": 1.2999739763145827e-05,
"loss": 7.6362104415893555,
"step": 751
},
{
"epoch": 0.030231155778894472,
"grad_norm": 1.654298186302185,
"learning_rate": 1.2999734532456268e-05,
"loss": 7.684175491333008,
"step": 752
},
{
"epoch": 0.0302713567839196,
"grad_norm": 1.7249393463134766,
"learning_rate": 1.2999729249722897e-05,
"loss": 8.0858736038208,
"step": 753
},
{
"epoch": 0.030311557788944725,
"grad_norm": 1.6963770389556885,
"learning_rate": 1.2999723914945794e-05,
"loss": 7.513216972351074,
"step": 754
},
{
"epoch": 0.030351758793969848,
"grad_norm": 2.027798652648926,
"learning_rate": 1.299971852812505e-05,
"loss": 7.848019599914551,
"step": 755
},
{
"epoch": 0.030391959798994974,
"grad_norm": 1.6678316593170166,
"learning_rate": 1.2999713089260748e-05,
"loss": 7.295356273651123,
"step": 756
},
{
"epoch": 0.0304321608040201,
"grad_norm": 1.7032239437103271,
"learning_rate": 1.2999707598352974e-05,
"loss": 7.8863420486450195,
"step": 757
},
{
"epoch": 0.030472361809045227,
"grad_norm": 1.786759614944458,
"learning_rate": 1.2999702055401819e-05,
"loss": 7.603622913360596,
"step": 758
},
{
"epoch": 0.030512562814070353,
"grad_norm": 1.8000046014785767,
"learning_rate": 1.299969646040737e-05,
"loss": 8.403044700622559,
"step": 759
},
{
"epoch": 0.03055276381909548,
"grad_norm": 1.868101716041565,
"learning_rate": 1.2999690813369716e-05,
"loss": 7.643517017364502,
"step": 760
},
{
"epoch": 0.030592964824120602,
"grad_norm": 1.795630693435669,
"learning_rate": 1.2999685114288949e-05,
"loss": 7.589298725128174,
"step": 761
},
{
"epoch": 0.030633165829145728,
"grad_norm": 1.7739944458007812,
"learning_rate": 1.2999679363165159e-05,
"loss": 7.73021125793457,
"step": 762
},
{
"epoch": 0.030673366834170854,
"grad_norm": 1.7243245840072632,
"learning_rate": 1.2999673559998436e-05,
"loss": 7.818130016326904,
"step": 763
},
{
"epoch": 0.03071356783919598,
"grad_norm": 1.858209490776062,
"learning_rate": 1.2999667704788878e-05,
"loss": 7.738125801086426,
"step": 764
},
{
"epoch": 0.030753768844221107,
"grad_norm": 1.6819267272949219,
"learning_rate": 1.2999661797536577e-05,
"loss": 7.84916877746582,
"step": 765
},
{
"epoch": 0.03079396984924623,
"grad_norm": 1.6891199350357056,
"learning_rate": 1.2999655838241627e-05,
"loss": 7.659395217895508,
"step": 766
},
{
"epoch": 0.030834170854271356,
"grad_norm": 2.126121759414673,
"learning_rate": 1.299964982690412e-05,
"loss": 7.8530073165893555,
"step": 767
},
{
"epoch": 0.030874371859296482,
"grad_norm": 1.9881348609924316,
"learning_rate": 1.2999643763524158e-05,
"loss": 7.892448425292969,
"step": 768
},
{
"epoch": 0.03091457286432161,
"grad_norm": 1.8096041679382324,
"learning_rate": 1.2999637648101836e-05,
"loss": 7.878176689147949,
"step": 769
},
{
"epoch": 0.030954773869346735,
"grad_norm": 1.621409296989441,
"learning_rate": 1.2999631480637252e-05,
"loss": 7.853533744812012,
"step": 770
},
{
"epoch": 0.030994974874371858,
"grad_norm": 1.9651460647583008,
"learning_rate": 1.2999625261130502e-05,
"loss": 7.679948806762695,
"step": 771
},
{
"epoch": 0.031035175879396984,
"grad_norm": 2.248032569885254,
"learning_rate": 1.299961898958169e-05,
"loss": 7.011972904205322,
"step": 772
},
{
"epoch": 0.03107537688442211,
"grad_norm": 1.7849292755126953,
"learning_rate": 1.2999612665990915e-05,
"loss": 7.2404632568359375,
"step": 773
},
{
"epoch": 0.031115577889447237,
"grad_norm": 2.0230114459991455,
"learning_rate": 1.2999606290358276e-05,
"loss": 7.169896125793457,
"step": 774
},
{
"epoch": 0.031155778894472363,
"grad_norm": 1.88129460811615,
"learning_rate": 1.2999599862683877e-05,
"loss": 7.447759628295898,
"step": 775
},
{
"epoch": 0.03119597989949749,
"grad_norm": 1.884032130241394,
"learning_rate": 1.2999593382967821e-05,
"loss": 7.738266944885254,
"step": 776
},
{
"epoch": 0.031236180904522612,
"grad_norm": 1.958912968635559,
"learning_rate": 1.2999586851210212e-05,
"loss": 7.406811714172363,
"step": 777
},
{
"epoch": 0.03127638190954774,
"grad_norm": 1.7902817726135254,
"learning_rate": 1.2999580267411153e-05,
"loss": 8.035781860351562,
"step": 778
},
{
"epoch": 0.031316582914572864,
"grad_norm": 1.760849952697754,
"learning_rate": 1.2999573631570752e-05,
"loss": 7.969487190246582,
"step": 779
},
{
"epoch": 0.03135678391959799,
"grad_norm": 1.772404670715332,
"learning_rate": 1.2999566943689112e-05,
"loss": 7.6795501708984375,
"step": 780
},
{
"epoch": 0.03139698492462312,
"grad_norm": 1.9026622772216797,
"learning_rate": 1.2999560203766343e-05,
"loss": 7.467230796813965,
"step": 781
},
{
"epoch": 0.03143718592964824,
"grad_norm": 1.9699798822402954,
"learning_rate": 1.2999553411802552e-05,
"loss": 7.4871439933776855,
"step": 782
},
{
"epoch": 0.03147738693467337,
"grad_norm": 1.8108917474746704,
"learning_rate": 1.2999546567797849e-05,
"loss": 8.124299049377441,
"step": 783
},
{
"epoch": 0.03151758793969849,
"grad_norm": 1.6718435287475586,
"learning_rate": 1.299953967175234e-05,
"loss": 7.591310501098633,
"step": 784
},
{
"epoch": 0.031557788944723615,
"grad_norm": 2.0934865474700928,
"learning_rate": 1.299953272366614e-05,
"loss": 7.9042277336120605,
"step": 785
},
{
"epoch": 0.031597989949748745,
"grad_norm": 1.7880675792694092,
"learning_rate": 1.2999525723539355e-05,
"loss": 7.5843024253845215,
"step": 786
},
{
"epoch": 0.03163819095477387,
"grad_norm": 2.973151445388794,
"learning_rate": 1.2999518671372103e-05,
"loss": 8.041728019714355,
"step": 787
},
{
"epoch": 0.031678391959799,
"grad_norm": 1.8477082252502441,
"learning_rate": 1.2999511567164491e-05,
"loss": 8.090555191040039,
"step": 788
},
{
"epoch": 0.03171859296482412,
"grad_norm": 2.06891131401062,
"learning_rate": 1.2999504410916637e-05,
"loss": 8.056326866149902,
"step": 789
},
{
"epoch": 0.03175879396984924,
"grad_norm": 1.7712790966033936,
"learning_rate": 1.2999497202628655e-05,
"loss": 7.580809593200684,
"step": 790
},
{
"epoch": 0.03179899497487437,
"grad_norm": 1.7537137269973755,
"learning_rate": 1.299948994230066e-05,
"loss": 7.553797721862793,
"step": 791
},
{
"epoch": 0.031839195979899496,
"grad_norm": 1.8707075119018555,
"learning_rate": 1.2999482629932767e-05,
"loss": 7.738955020904541,
"step": 792
},
{
"epoch": 0.031879396984924625,
"grad_norm": 2.1978235244750977,
"learning_rate": 1.2999475265525095e-05,
"loss": 7.201945781707764,
"step": 793
},
{
"epoch": 0.03191959798994975,
"grad_norm": 1.6551933288574219,
"learning_rate": 1.299946784907776e-05,
"loss": 7.834621429443359,
"step": 794
},
{
"epoch": 0.03195979899497487,
"grad_norm": 1.7521708011627197,
"learning_rate": 1.2999460380590884e-05,
"loss": 7.670608997344971,
"step": 795
},
{
"epoch": 0.032,
"grad_norm": 1.8827162981033325,
"learning_rate": 1.2999452860064583e-05,
"loss": 7.859295845031738,
"step": 796
},
{
"epoch": 0.032040201005025123,
"grad_norm": 1.9590792655944824,
"learning_rate": 1.2999445287498979e-05,
"loss": 8.192548751831055,
"step": 797
},
{
"epoch": 0.03208040201005025,
"grad_norm": 1.7150732278823853,
"learning_rate": 1.2999437662894194e-05,
"loss": 7.741738319396973,
"step": 798
},
{
"epoch": 0.032120603015075376,
"grad_norm": 1.8586477041244507,
"learning_rate": 1.2999429986250348e-05,
"loss": 7.889492034912109,
"step": 799
},
{
"epoch": 0.032160804020100506,
"grad_norm": 1.7037099599838257,
"learning_rate": 1.2999422257567567e-05,
"loss": 7.5134172439575195,
"step": 800
},
{
"epoch": 0.032160804020100506,
"eval_loss": 7.7026567459106445,
"eval_runtime": 120.5733,
"eval_samples_per_second": 4.147,
"eval_steps_per_second": 0.523,
"step": 800
},
{
"epoch": 0.03220100502512563,
"grad_norm": 1.7957520484924316,
"learning_rate": 1.299941447684597e-05,
"loss": 7.721504211425781,
"step": 801
},
{
"epoch": 0.03224120603015075,
"grad_norm": 1.685470461845398,
"learning_rate": 1.2999406644085689e-05,
"loss": 7.913516521453857,
"step": 802
},
{
"epoch": 0.03228140703517588,
"grad_norm": 1.707322120666504,
"learning_rate": 1.2999398759286842e-05,
"loss": 8.053775787353516,
"step": 803
},
{
"epoch": 0.032321608040201004,
"grad_norm": 1.8015021085739136,
"learning_rate": 1.299939082244956e-05,
"loss": 7.829987049102783,
"step": 804
},
{
"epoch": 0.032361809045226134,
"grad_norm": 1.7325705289840698,
"learning_rate": 1.2999382833573968e-05,
"loss": 7.617226600646973,
"step": 805
},
{
"epoch": 0.032402010050251256,
"grad_norm": 2.256561040878296,
"learning_rate": 1.2999374792660193e-05,
"loss": 8.072046279907227,
"step": 806
},
{
"epoch": 0.03244221105527638,
"grad_norm": 1.8782912492752075,
"learning_rate": 1.2999366699708366e-05,
"loss": 7.564282417297363,
"step": 807
},
{
"epoch": 0.03248241206030151,
"grad_norm": 2.0341100692749023,
"learning_rate": 1.2999358554718617e-05,
"loss": 7.5738725662231445,
"step": 808
},
{
"epoch": 0.03252261306532663,
"grad_norm": 1.8857765197753906,
"learning_rate": 1.2999350357691075e-05,
"loss": 7.802791595458984,
"step": 809
},
{
"epoch": 0.03256281407035176,
"grad_norm": 1.6247769594192505,
"learning_rate": 1.2999342108625871e-05,
"loss": 7.84116792678833,
"step": 810
},
{
"epoch": 0.032603015075376884,
"grad_norm": 1.8052881956100464,
"learning_rate": 1.2999333807523137e-05,
"loss": 7.330854415893555,
"step": 811
},
{
"epoch": 0.03264321608040201,
"grad_norm": 2.2460968494415283,
"learning_rate": 1.2999325454383006e-05,
"loss": 7.31424617767334,
"step": 812
},
{
"epoch": 0.03268341708542714,
"grad_norm": 1.7823195457458496,
"learning_rate": 1.2999317049205615e-05,
"loss": 7.623928546905518,
"step": 813
},
{
"epoch": 0.03272361809045226,
"grad_norm": 1.7900898456573486,
"learning_rate": 1.2999308591991094e-05,
"loss": 7.7667155265808105,
"step": 814
},
{
"epoch": 0.03276381909547739,
"grad_norm": 1.6435078382492065,
"learning_rate": 1.299930008273958e-05,
"loss": 7.826323986053467,
"step": 815
},
{
"epoch": 0.03280402010050251,
"grad_norm": 1.838303804397583,
"learning_rate": 1.2999291521451209e-05,
"loss": 7.738211631774902,
"step": 816
},
{
"epoch": 0.032844221105527635,
"grad_norm": 1.863493800163269,
"learning_rate": 1.2999282908126122e-05,
"loss": 7.690169334411621,
"step": 817
},
{
"epoch": 0.032884422110552765,
"grad_norm": 1.8263291120529175,
"learning_rate": 1.299927424276445e-05,
"loss": 7.697854042053223,
"step": 818
},
{
"epoch": 0.03292462311557789,
"grad_norm": 1.7823470830917358,
"learning_rate": 1.2999265525366338e-05,
"loss": 7.816228866577148,
"step": 819
},
{
"epoch": 0.03296482412060302,
"grad_norm": 2.076565742492676,
"learning_rate": 1.2999256755931922e-05,
"loss": 7.46360969543457,
"step": 820
},
{
"epoch": 0.03300502512562814,
"grad_norm": 1.8310463428497314,
"learning_rate": 1.2999247934461345e-05,
"loss": 7.9712018966674805,
"step": 821
},
{
"epoch": 0.03304522613065326,
"grad_norm": 1.7214808464050293,
"learning_rate": 1.2999239060954745e-05,
"loss": 7.525637626647949,
"step": 822
},
{
"epoch": 0.03308542713567839,
"grad_norm": 1.8927217721939087,
"learning_rate": 1.2999230135412267e-05,
"loss": 7.912696838378906,
"step": 823
},
{
"epoch": 0.033125628140703516,
"grad_norm": 1.737606167793274,
"learning_rate": 1.2999221157834051e-05,
"loss": 7.527256011962891,
"step": 824
},
{
"epoch": 0.033165829145728645,
"grad_norm": 1.8347656726837158,
"learning_rate": 1.2999212128220245e-05,
"loss": 7.913517951965332,
"step": 825
},
{
"epoch": 0.03320603015075377,
"grad_norm": 1.719211220741272,
"learning_rate": 1.299920304657099e-05,
"loss": 7.755512237548828,
"step": 826
},
{
"epoch": 0.0332462311557789,
"grad_norm": 1.8688887357711792,
"learning_rate": 1.2999193912886433e-05,
"loss": 7.691404819488525,
"step": 827
},
{
"epoch": 0.03328643216080402,
"grad_norm": 1.7411433458328247,
"learning_rate": 1.299918472716672e-05,
"loss": 7.58930778503418,
"step": 828
},
{
"epoch": 0.03332663316582914,
"grad_norm": 1.7663609981536865,
"learning_rate": 1.2999175489411999e-05,
"loss": 7.538576126098633,
"step": 829
},
{
"epoch": 0.03336683417085427,
"grad_norm": 1.747652530670166,
"learning_rate": 1.2999166199622416e-05,
"loss": 7.883047103881836,
"step": 830
},
{
"epoch": 0.033407035175879396,
"grad_norm": 1.9413790702819824,
"learning_rate": 1.299915685779812e-05,
"loss": 7.763370037078857,
"step": 831
},
{
"epoch": 0.033447236180904526,
"grad_norm": 1.8163124322891235,
"learning_rate": 1.2999147463939262e-05,
"loss": 7.612316131591797,
"step": 832
},
{
"epoch": 0.03348743718592965,
"grad_norm": 1.8056682348251343,
"learning_rate": 1.299913801804599e-05,
"loss": 7.7552595138549805,
"step": 833
},
{
"epoch": 0.03352763819095477,
"grad_norm": 1.787187099456787,
"learning_rate": 1.2999128520118459e-05,
"loss": 7.647422790527344,
"step": 834
},
{
"epoch": 0.0335678391959799,
"grad_norm": 2.9213242530822754,
"learning_rate": 1.2999118970156821e-05,
"loss": 7.817028045654297,
"step": 835
},
{
"epoch": 0.033608040201005024,
"grad_norm": 2.229635000228882,
"learning_rate": 1.2999109368161224e-05,
"loss": 7.691123008728027,
"step": 836
},
{
"epoch": 0.033648241206030154,
"grad_norm": 2.1378579139709473,
"learning_rate": 1.2999099714131827e-05,
"loss": 7.67255973815918,
"step": 837
},
{
"epoch": 0.033688442211055276,
"grad_norm": 2.025339365005493,
"learning_rate": 1.299909000806878e-05,
"loss": 7.701690673828125,
"step": 838
},
{
"epoch": 0.0337286432160804,
"grad_norm": 1.8423795700073242,
"learning_rate": 1.2999080249972243e-05,
"loss": 7.6082587242126465,
"step": 839
},
{
"epoch": 0.03376884422110553,
"grad_norm": 1.7670701742172241,
"learning_rate": 1.299907043984237e-05,
"loss": 7.461488723754883,
"step": 840
},
{
"epoch": 0.03380904522613065,
"grad_norm": 1.8573954105377197,
"learning_rate": 1.2999060577679319e-05,
"loss": 7.357446193695068,
"step": 841
},
{
"epoch": 0.03384924623115578,
"grad_norm": 2.290571451187134,
"learning_rate": 1.2999050663483246e-05,
"loss": 7.699342727661133,
"step": 842
},
{
"epoch": 0.033889447236180904,
"grad_norm": 1.6898741722106934,
"learning_rate": 1.2999040697254312e-05,
"loss": 7.8625407218933105,
"step": 843
},
{
"epoch": 0.03392964824120603,
"grad_norm": 1.6839531660079956,
"learning_rate": 1.2999030678992674e-05,
"loss": 7.44912052154541,
"step": 844
},
{
"epoch": 0.03396984924623116,
"grad_norm": 1.9796618223190308,
"learning_rate": 1.2999020608698494e-05,
"loss": 8.419624328613281,
"step": 845
},
{
"epoch": 0.03401005025125628,
"grad_norm": 2.0364298820495605,
"learning_rate": 1.2999010486371937e-05,
"loss": 7.834813117980957,
"step": 846
},
{
"epoch": 0.03405025125628141,
"grad_norm": 1.9516351222991943,
"learning_rate": 1.2999000312013157e-05,
"loss": 7.9676008224487305,
"step": 847
},
{
"epoch": 0.03409045226130653,
"grad_norm": 2.0610854625701904,
"learning_rate": 1.2998990085622325e-05,
"loss": 7.5244855880737305,
"step": 848
},
{
"epoch": 0.034130653266331655,
"grad_norm": 2.006483554840088,
"learning_rate": 1.2998979807199598e-05,
"loss": 7.915414810180664,
"step": 849
},
{
"epoch": 0.034170854271356785,
"grad_norm": 1.9056023359298706,
"learning_rate": 1.2998969476745146e-05,
"loss": 7.450005531311035,
"step": 850
},
{
"epoch": 0.03421105527638191,
"grad_norm": 1.792157530784607,
"learning_rate": 1.2998959094259132e-05,
"loss": 7.714456558227539,
"step": 851
},
{
"epoch": 0.03425125628140704,
"grad_norm": 1.8048880100250244,
"learning_rate": 1.2998948659741721e-05,
"loss": 7.745038032531738,
"step": 852
},
{
"epoch": 0.03429145728643216,
"grad_norm": 1.7585039138793945,
"learning_rate": 1.2998938173193083e-05,
"loss": 7.560314178466797,
"step": 853
},
{
"epoch": 0.03433165829145729,
"grad_norm": 1.758608102798462,
"learning_rate": 1.2998927634613385e-05,
"loss": 7.586599826812744,
"step": 854
},
{
"epoch": 0.03437185929648241,
"grad_norm": 2.2386138439178467,
"learning_rate": 1.2998917044002794e-05,
"loss": 7.365553855895996,
"step": 855
},
{
"epoch": 0.034412060301507535,
"grad_norm": 1.8730686902999878,
"learning_rate": 1.2998906401361483e-05,
"loss": 7.7385993003845215,
"step": 856
},
{
"epoch": 0.034452261306532665,
"grad_norm": 1.7579396963119507,
"learning_rate": 1.299889570668962e-05,
"loss": 7.438879013061523,
"step": 857
},
{
"epoch": 0.03449246231155779,
"grad_norm": 1.8698703050613403,
"learning_rate": 1.2998884959987374e-05,
"loss": 7.413554668426514,
"step": 858
},
{
"epoch": 0.03453266331658292,
"grad_norm": 1.8049569129943848,
"learning_rate": 1.2998874161254923e-05,
"loss": 7.552549362182617,
"step": 859
},
{
"epoch": 0.03457286432160804,
"grad_norm": 2.01973819732666,
"learning_rate": 1.2998863310492435e-05,
"loss": 8.010618209838867,
"step": 860
},
{
"epoch": 0.03461306532663316,
"grad_norm": 1.8278591632843018,
"learning_rate": 1.2998852407700084e-05,
"loss": 7.493427276611328,
"step": 861
},
{
"epoch": 0.03465326633165829,
"grad_norm": 2.3008530139923096,
"learning_rate": 1.2998841452878048e-05,
"loss": 7.664429664611816,
"step": 862
},
{
"epoch": 0.034693467336683416,
"grad_norm": 2.0149409770965576,
"learning_rate": 1.29988304460265e-05,
"loss": 7.612252235412598,
"step": 863
},
{
"epoch": 0.034733668341708546,
"grad_norm": 1.809327483177185,
"learning_rate": 1.2998819387145618e-05,
"loss": 7.857554912567139,
"step": 864
},
{
"epoch": 0.03477386934673367,
"grad_norm": 2.54750919342041,
"learning_rate": 1.2998808276235576e-05,
"loss": 7.427453994750977,
"step": 865
},
{
"epoch": 0.03481407035175879,
"grad_norm": 1.8184291124343872,
"learning_rate": 1.2998797113296554e-05,
"loss": 7.589202880859375,
"step": 866
},
{
"epoch": 0.03485427135678392,
"grad_norm": 2.0886521339416504,
"learning_rate": 1.2998785898328731e-05,
"loss": 7.527334213256836,
"step": 867
},
{
"epoch": 0.034894472361809044,
"grad_norm": 1.777144432067871,
"learning_rate": 1.2998774631332288e-05,
"loss": 7.858880043029785,
"step": 868
},
{
"epoch": 0.034934673366834174,
"grad_norm": 2.249290704727173,
"learning_rate": 1.29987633123074e-05,
"loss": 7.270679473876953,
"step": 869
},
{
"epoch": 0.034974874371859296,
"grad_norm": 2.2412352561950684,
"learning_rate": 1.2998751941254256e-05,
"loss": 7.7580742835998535,
"step": 870
},
{
"epoch": 0.03501507537688442,
"grad_norm": 1.9896759986877441,
"learning_rate": 1.2998740518173031e-05,
"loss": 7.92957067489624,
"step": 871
},
{
"epoch": 0.03505527638190955,
"grad_norm": 2.313802719116211,
"learning_rate": 1.2998729043063911e-05,
"loss": 8.050308227539062,
"step": 872
},
{
"epoch": 0.03509547738693467,
"grad_norm": 1.932567834854126,
"learning_rate": 1.2998717515927082e-05,
"loss": 7.981450080871582,
"step": 873
},
{
"epoch": 0.0351356783919598,
"grad_norm": 1.7745575904846191,
"learning_rate": 1.2998705936762725e-05,
"loss": 7.939798355102539,
"step": 874
},
{
"epoch": 0.035175879396984924,
"grad_norm": 2.0925076007843018,
"learning_rate": 1.2998694305571027e-05,
"loss": 7.669553279876709,
"step": 875
},
{
"epoch": 0.03521608040201005,
"grad_norm": 2.0458314418792725,
"learning_rate": 1.2998682622352173e-05,
"loss": 7.597897529602051,
"step": 876
},
{
"epoch": 0.03525628140703518,
"grad_norm": 1.899825930595398,
"learning_rate": 1.2998670887106353e-05,
"loss": 7.535547256469727,
"step": 877
},
{
"epoch": 0.0352964824120603,
"grad_norm": 1.7773493528366089,
"learning_rate": 1.2998659099833752e-05,
"loss": 7.455918788909912,
"step": 878
},
{
"epoch": 0.03533668341708543,
"grad_norm": 1.7670735120773315,
"learning_rate": 1.299864726053456e-05,
"loss": 7.634939670562744,
"step": 879
},
{
"epoch": 0.03537688442211055,
"grad_norm": 1.8202835321426392,
"learning_rate": 1.2998635369208967e-05,
"loss": 7.637392997741699,
"step": 880
},
{
"epoch": 0.035417085427135675,
"grad_norm": 1.8777174949645996,
"learning_rate": 1.299862342585716e-05,
"loss": 7.836607933044434,
"step": 881
},
{
"epoch": 0.035457286432160805,
"grad_norm": 2.077345848083496,
"learning_rate": 1.2998611430479336e-05,
"loss": 8.009795188903809,
"step": 882
},
{
"epoch": 0.03549748743718593,
"grad_norm": 1.8153847455978394,
"learning_rate": 1.2998599383075684e-05,
"loss": 7.9209794998168945,
"step": 883
},
{
"epoch": 0.03553768844221106,
"grad_norm": 2.099524974822998,
"learning_rate": 1.2998587283646396e-05,
"loss": 7.971922874450684,
"step": 884
},
{
"epoch": 0.03557788944723618,
"grad_norm": 1.8988676071166992,
"learning_rate": 1.2998575132191668e-05,
"loss": 7.999294281005859,
"step": 885
},
{
"epoch": 0.03561809045226131,
"grad_norm": 2.4048850536346436,
"learning_rate": 1.2998562928711695e-05,
"loss": 7.635300636291504,
"step": 886
},
{
"epoch": 0.03565829145728643,
"grad_norm": 2.5474793910980225,
"learning_rate": 1.2998550673206668e-05,
"loss": 7.747020244598389,
"step": 887
},
{
"epoch": 0.035698492462311555,
"grad_norm": 2.604142904281616,
"learning_rate": 1.2998538365676788e-05,
"loss": 7.302949905395508,
"step": 888
},
{
"epoch": 0.035738693467336685,
"grad_norm": 2.04754376411438,
"learning_rate": 1.2998526006122251e-05,
"loss": 7.628472328186035,
"step": 889
},
{
"epoch": 0.03577889447236181,
"grad_norm": 1.952886700630188,
"learning_rate": 1.2998513594543254e-05,
"loss": 7.765311241149902,
"step": 890
},
{
"epoch": 0.03581909547738694,
"grad_norm": 3.109809160232544,
"learning_rate": 1.2998501130939996e-05,
"loss": 7.590115070343018,
"step": 891
},
{
"epoch": 0.03585929648241206,
"grad_norm": 1.8144489526748657,
"learning_rate": 1.2998488615312676e-05,
"loss": 7.697164535522461,
"step": 892
},
{
"epoch": 0.03589949748743718,
"grad_norm": 1.9058493375778198,
"learning_rate": 1.2998476047661494e-05,
"loss": 7.470232963562012,
"step": 893
},
{
"epoch": 0.03593969849246231,
"grad_norm": 1.8547919988632202,
"learning_rate": 1.2998463427986654e-05,
"loss": 7.776553153991699,
"step": 894
},
{
"epoch": 0.035979899497487436,
"grad_norm": 2.115990400314331,
"learning_rate": 1.2998450756288357e-05,
"loss": 7.7523956298828125,
"step": 895
},
{
"epoch": 0.036020100502512566,
"grad_norm": 2.081102132797241,
"learning_rate": 1.2998438032566805e-05,
"loss": 7.561567783355713,
"step": 896
},
{
"epoch": 0.03606030150753769,
"grad_norm": 1.9460198879241943,
"learning_rate": 1.2998425256822201e-05,
"loss": 7.548089981079102,
"step": 897
},
{
"epoch": 0.03610050251256281,
"grad_norm": 1.9395897388458252,
"learning_rate": 1.2998412429054753e-05,
"loss": 7.4687910079956055,
"step": 898
},
{
"epoch": 0.03614070351758794,
"grad_norm": 2.102771520614624,
"learning_rate": 1.2998399549264662e-05,
"loss": 7.287975311279297,
"step": 899
},
{
"epoch": 0.036180904522613064,
"grad_norm": 2.251718044281006,
"learning_rate": 1.2998386617452139e-05,
"loss": 8.088274002075195,
"step": 900
},
{
"epoch": 0.036221105527638194,
"grad_norm": 1.8418235778808594,
"learning_rate": 1.2998373633617386e-05,
"loss": 7.820502281188965,
"step": 901
},
{
"epoch": 0.036261306532663316,
"grad_norm": 1.9549968242645264,
"learning_rate": 1.2998360597760615e-05,
"loss": 7.694643020629883,
"step": 902
},
{
"epoch": 0.03630150753768844,
"grad_norm": 1.982108473777771,
"learning_rate": 1.2998347509882034e-05,
"loss": 7.3136773109436035,
"step": 903
},
{
"epoch": 0.03634170854271357,
"grad_norm": 1.9261085987091064,
"learning_rate": 1.299833436998185e-05,
"loss": 7.576442718505859,
"step": 904
},
{
"epoch": 0.03638190954773869,
"grad_norm": 2.2713217735290527,
"learning_rate": 1.2998321178060276e-05,
"loss": 8.025411605834961,
"step": 905
},
{
"epoch": 0.03642211055276382,
"grad_norm": 2.0251054763793945,
"learning_rate": 1.2998307934117522e-05,
"loss": 8.241151809692383,
"step": 906
},
{
"epoch": 0.036462311557788944,
"grad_norm": 1.9241399765014648,
"learning_rate": 1.2998294638153803e-05,
"loss": 7.462823867797852,
"step": 907
},
{
"epoch": 0.03650251256281407,
"grad_norm": 1.8986636400222778,
"learning_rate": 1.2998281290169327e-05,
"loss": 7.302155494689941,
"step": 908
},
{
"epoch": 0.0365427135678392,
"grad_norm": 1.990778923034668,
"learning_rate": 1.2998267890164312e-05,
"loss": 8.056617736816406,
"step": 909
},
{
"epoch": 0.03658291457286432,
"grad_norm": 2.007962465286255,
"learning_rate": 1.299825443813897e-05,
"loss": 7.905242919921875,
"step": 910
},
{
"epoch": 0.03662311557788945,
"grad_norm": 1.985810399055481,
"learning_rate": 1.299824093409352e-05,
"loss": 8.162322998046875,
"step": 911
},
{
"epoch": 0.03666331658291457,
"grad_norm": 1.9583463668823242,
"learning_rate": 1.2998227378028173e-05,
"loss": 7.632874488830566,
"step": 912
},
{
"epoch": 0.0367035175879397,
"grad_norm": 1.8355571031570435,
"learning_rate": 1.2998213769943148e-05,
"loss": 7.874990940093994,
"step": 913
},
{
"epoch": 0.036743718592964825,
"grad_norm": 1.8789044618606567,
"learning_rate": 1.2998200109838666e-05,
"loss": 7.392563343048096,
"step": 914
},
{
"epoch": 0.03678391959798995,
"grad_norm": 2.0061211585998535,
"learning_rate": 1.2998186397714942e-05,
"loss": 8.023197174072266,
"step": 915
},
{
"epoch": 0.03682412060301508,
"grad_norm": 2.074704647064209,
"learning_rate": 1.2998172633572198e-05,
"loss": 7.4075422286987305,
"step": 916
},
{
"epoch": 0.0368643216080402,
"grad_norm": 2.0055997371673584,
"learning_rate": 1.2998158817410654e-05,
"loss": 7.706284999847412,
"step": 917
},
{
"epoch": 0.03690452261306533,
"grad_norm": 2.176767349243164,
"learning_rate": 1.2998144949230529e-05,
"loss": 8.275930404663086,
"step": 918
},
{
"epoch": 0.03694472361809045,
"grad_norm": 1.916502594947815,
"learning_rate": 1.2998131029032047e-05,
"loss": 7.638848781585693,
"step": 919
},
{
"epoch": 0.036984924623115575,
"grad_norm": 2.093914270401001,
"learning_rate": 1.2998117056815433e-05,
"loss": 7.442468643188477,
"step": 920
},
{
"epoch": 0.037025125628140705,
"grad_norm": 1.881726622581482,
"learning_rate": 1.2998103032580907e-05,
"loss": 7.401495933532715,
"step": 921
},
{
"epoch": 0.03706532663316583,
"grad_norm": 1.8578797578811646,
"learning_rate": 1.2998088956328696e-05,
"loss": 7.36842155456543,
"step": 922
},
{
"epoch": 0.03710552763819096,
"grad_norm": 2.020545721054077,
"learning_rate": 1.2998074828059024e-05,
"loss": 7.749814033508301,
"step": 923
},
{
"epoch": 0.03714572864321608,
"grad_norm": 1.8993446826934814,
"learning_rate": 1.2998060647772118e-05,
"loss": 7.742489814758301,
"step": 924
},
{
"epoch": 0.0371859296482412,
"grad_norm": 1.9939085245132446,
"learning_rate": 1.2998046415468204e-05,
"loss": 7.933173179626465,
"step": 925
},
{
"epoch": 0.03722613065326633,
"grad_norm": 2.024240255355835,
"learning_rate": 1.2998032131147511e-05,
"loss": 7.598540306091309,
"step": 926
},
{
"epoch": 0.037266331658291456,
"grad_norm": 2.268216133117676,
"learning_rate": 1.2998017794810269e-05,
"loss": 7.524990558624268,
"step": 927
},
{
"epoch": 0.037306532663316586,
"grad_norm": 1.863292932510376,
"learning_rate": 1.2998003406456705e-05,
"loss": 7.6789140701293945,
"step": 928
},
{
"epoch": 0.03734673366834171,
"grad_norm": 1.9356777667999268,
"learning_rate": 1.2997988966087052e-05,
"loss": 7.65748405456543,
"step": 929
},
{
"epoch": 0.03738693467336683,
"grad_norm": 1.8185824155807495,
"learning_rate": 1.2997974473701538e-05,
"loss": 7.4505109786987305,
"step": 930
},
{
"epoch": 0.03742713567839196,
"grad_norm": 1.818267583847046,
"learning_rate": 1.2997959929300398e-05,
"loss": 7.632761001586914,
"step": 931
},
{
"epoch": 0.037467336683417084,
"grad_norm": 2.130397081375122,
"learning_rate": 1.2997945332883865e-05,
"loss": 7.528388977050781,
"step": 932
},
{
"epoch": 0.037507537688442213,
"grad_norm": 1.88065505027771,
"learning_rate": 1.2997930684452171e-05,
"loss": 7.603992462158203,
"step": 933
},
{
"epoch": 0.037547738693467336,
"grad_norm": 1.9160621166229248,
"learning_rate": 1.2997915984005552e-05,
"loss": 7.545210838317871,
"step": 934
},
{
"epoch": 0.03758793969849246,
"grad_norm": 2.3152852058410645,
"learning_rate": 1.2997901231544241e-05,
"loss": 7.143431186676025,
"step": 935
},
{
"epoch": 0.03762814070351759,
"grad_norm": 2.1172239780426025,
"learning_rate": 1.2997886427068476e-05,
"loss": 7.648075580596924,
"step": 936
},
{
"epoch": 0.03766834170854271,
"grad_norm": 1.9231717586517334,
"learning_rate": 1.2997871570578494e-05,
"loss": 7.577048301696777,
"step": 937
},
{
"epoch": 0.03770854271356784,
"grad_norm": 2.09977650642395,
"learning_rate": 1.2997856662074534e-05,
"loss": 7.905293941497803,
"step": 938
},
{
"epoch": 0.037748743718592964,
"grad_norm": 1.7944936752319336,
"learning_rate": 1.2997841701556833e-05,
"loss": 7.863315582275391,
"step": 939
},
{
"epoch": 0.03778894472361809,
"grad_norm": 1.7919650077819824,
"learning_rate": 1.2997826689025631e-05,
"loss": 7.794942855834961,
"step": 940
},
{
"epoch": 0.03782914572864322,
"grad_norm": 1.818730354309082,
"learning_rate": 1.299781162448117e-05,
"loss": 7.603723526000977,
"step": 941
},
{
"epoch": 0.03786934673366834,
"grad_norm": 2.571943759918213,
"learning_rate": 1.299779650792369e-05,
"loss": 7.678154945373535,
"step": 942
},
{
"epoch": 0.03790954773869347,
"grad_norm": 1.737815499305725,
"learning_rate": 1.2997781339353433e-05,
"loss": 7.791824817657471,
"step": 943
},
{
"epoch": 0.03794974874371859,
"grad_norm": 1.937406301498413,
"learning_rate": 1.299776611877064e-05,
"loss": 7.297183036804199,
"step": 944
},
{
"epoch": 0.03798994974874372,
"grad_norm": 1.95573091506958,
"learning_rate": 1.2997750846175558e-05,
"loss": 8.173760414123535,
"step": 945
},
{
"epoch": 0.038030150753768845,
"grad_norm": 1.8075143098831177,
"learning_rate": 1.2997735521568432e-05,
"loss": 7.589173316955566,
"step": 946
},
{
"epoch": 0.03807035175879397,
"grad_norm": 1.9055559635162354,
"learning_rate": 1.2997720144949503e-05,
"loss": 7.9064531326293945,
"step": 947
},
{
"epoch": 0.0381105527638191,
"grad_norm": 2.018826961517334,
"learning_rate": 1.2997704716319023e-05,
"loss": 7.369829177856445,
"step": 948
},
{
"epoch": 0.03815075376884422,
"grad_norm": 2.267791748046875,
"learning_rate": 1.2997689235677233e-05,
"loss": 7.640754699707031,
"step": 949
},
{
"epoch": 0.03819095477386935,
"grad_norm": 2.02954363822937,
"learning_rate": 1.2997673703024387e-05,
"loss": 7.547907829284668,
"step": 950
},
{
"epoch": 0.03823115577889447,
"grad_norm": 1.9693403244018555,
"learning_rate": 1.2997658118360728e-05,
"loss": 7.21193790435791,
"step": 951
},
{
"epoch": 0.038271356783919595,
"grad_norm": 2.029742956161499,
"learning_rate": 1.2997642481686511e-05,
"loss": 7.575554847717285,
"step": 952
},
{
"epoch": 0.038311557788944725,
"grad_norm": 2.499613046646118,
"learning_rate": 1.2997626793001981e-05,
"loss": 7.598565578460693,
"step": 953
},
{
"epoch": 0.03835175879396985,
"grad_norm": 2.155724287033081,
"learning_rate": 1.2997611052307394e-05,
"loss": 7.883960247039795,
"step": 954
},
{
"epoch": 0.03839195979899498,
"grad_norm": 1.9873180389404297,
"learning_rate": 1.2997595259603e-05,
"loss": 7.3067626953125,
"step": 955
},
{
"epoch": 0.0384321608040201,
"grad_norm": 2.0236575603485107,
"learning_rate": 1.2997579414889052e-05,
"loss": 7.47164249420166,
"step": 956
},
{
"epoch": 0.03847236180904522,
"grad_norm": 2.0843729972839355,
"learning_rate": 1.2997563518165803e-05,
"loss": 7.3104567527771,
"step": 957
},
{
"epoch": 0.03851256281407035,
"grad_norm": 1.8882648944854736,
"learning_rate": 1.2997547569433509e-05,
"loss": 7.611563682556152,
"step": 958
},
{
"epoch": 0.038552763819095476,
"grad_norm": 2.2020580768585205,
"learning_rate": 1.2997531568692424e-05,
"loss": 7.841202735900879,
"step": 959
},
{
"epoch": 0.038592964824120606,
"grad_norm": 1.8085001707077026,
"learning_rate": 1.2997515515942805e-05,
"loss": 7.337479591369629,
"step": 960
},
{
"epoch": 0.03863316582914573,
"grad_norm": 1.998103380203247,
"learning_rate": 1.299749941118491e-05,
"loss": 7.460633277893066,
"step": 961
},
{
"epoch": 0.03867336683417085,
"grad_norm": 1.9596693515777588,
"learning_rate": 1.2997483254418996e-05,
"loss": 7.63128662109375,
"step": 962
},
{
"epoch": 0.03871356783919598,
"grad_norm": 1.933302402496338,
"learning_rate": 1.2997467045645322e-05,
"loss": 7.463423728942871,
"step": 963
},
{
"epoch": 0.038753768844221104,
"grad_norm": 1.8684061765670776,
"learning_rate": 1.2997450784864146e-05,
"loss": 7.448873996734619,
"step": 964
},
{
"epoch": 0.03879396984924623,
"grad_norm": 1.8532581329345703,
"learning_rate": 1.299743447207573e-05,
"loss": 7.7176642417907715,
"step": 965
},
{
"epoch": 0.038834170854271356,
"grad_norm": 2.0258889198303223,
"learning_rate": 1.2997418107280337e-05,
"loss": 8.182586669921875,
"step": 966
},
{
"epoch": 0.03887437185929648,
"grad_norm": 1.9224262237548828,
"learning_rate": 1.2997401690478224e-05,
"loss": 7.917318344116211,
"step": 967
},
{
"epoch": 0.03891457286432161,
"grad_norm": 2.0235908031463623,
"learning_rate": 1.299738522166966e-05,
"loss": 7.570297718048096,
"step": 968
},
{
"epoch": 0.03895477386934673,
"grad_norm": 1.983171820640564,
"learning_rate": 1.2997368700854904e-05,
"loss": 7.461816787719727,
"step": 969
},
{
"epoch": 0.03899497487437186,
"grad_norm": 1.88791823387146,
"learning_rate": 1.2997352128034225e-05,
"loss": 7.669847011566162,
"step": 970
},
{
"epoch": 0.039035175879396984,
"grad_norm": 2.0805504322052,
"learning_rate": 1.2997335503207882e-05,
"loss": 7.444814682006836,
"step": 971
},
{
"epoch": 0.039075376884422114,
"grad_norm": 2.020916223526001,
"learning_rate": 1.2997318826376146e-05,
"loss": 7.777940273284912,
"step": 972
},
{
"epoch": 0.03911557788944724,
"grad_norm": 1.9764149188995361,
"learning_rate": 1.2997302097539283e-05,
"loss": 7.387870788574219,
"step": 973
},
{
"epoch": 0.03915577889447236,
"grad_norm": 2.130870819091797,
"learning_rate": 1.2997285316697563e-05,
"loss": 7.45477294921875,
"step": 974
},
{
"epoch": 0.03919597989949749,
"grad_norm": 2.2529733180999756,
"learning_rate": 1.2997268483851251e-05,
"loss": 8.161754608154297,
"step": 975
},
{
"epoch": 0.03923618090452261,
"grad_norm": 1.8848172426223755,
"learning_rate": 1.2997251599000619e-05,
"loss": 8.107465744018555,
"step": 976
},
{
"epoch": 0.03927638190954774,
"grad_norm": 2.169766426086426,
"learning_rate": 1.2997234662145936e-05,
"loss": 7.537971019744873,
"step": 977
},
{
"epoch": 0.039316582914572865,
"grad_norm": 1.927579641342163,
"learning_rate": 1.2997217673287474e-05,
"loss": 7.769967079162598,
"step": 978
},
{
"epoch": 0.03935678391959799,
"grad_norm": 1.811439871788025,
"learning_rate": 1.2997200632425503e-05,
"loss": 7.675601005554199,
"step": 979
},
{
"epoch": 0.03939698492462312,
"grad_norm": 1.9669723510742188,
"learning_rate": 1.29971835395603e-05,
"loss": 7.432029724121094,
"step": 980
},
{
"epoch": 0.03943718592964824,
"grad_norm": 1.9354671239852905,
"learning_rate": 1.2997166394692137e-05,
"loss": 7.488142967224121,
"step": 981
},
{
"epoch": 0.03947738693467337,
"grad_norm": 1.9892654418945312,
"learning_rate": 1.299714919782129e-05,
"loss": 7.658243656158447,
"step": 982
},
{
"epoch": 0.03951758793969849,
"grad_norm": 2.074462413787842,
"learning_rate": 1.2997131948948029e-05,
"loss": 7.49773645401001,
"step": 983
},
{
"epoch": 0.039557788944723615,
"grad_norm": 2.3907036781311035,
"learning_rate": 1.2997114648072636e-05,
"loss": 7.523918628692627,
"step": 984
},
{
"epoch": 0.039597989949748745,
"grad_norm": 1.9649748802185059,
"learning_rate": 1.2997097295195385e-05,
"loss": 7.827256202697754,
"step": 985
},
{
"epoch": 0.03963819095477387,
"grad_norm": 1.8378640413284302,
"learning_rate": 1.2997079890316555e-05,
"loss": 7.5852742195129395,
"step": 986
},
{
"epoch": 0.039678391959799,
"grad_norm": 1.9324021339416504,
"learning_rate": 1.2997062433436424e-05,
"loss": 8.020328521728516,
"step": 987
},
{
"epoch": 0.03971859296482412,
"grad_norm": 2.5085954666137695,
"learning_rate": 1.2997044924555274e-05,
"loss": 7.523471355438232,
"step": 988
},
{
"epoch": 0.03975879396984924,
"grad_norm": 2.2380549907684326,
"learning_rate": 1.2997027363673382e-05,
"loss": 7.387966156005859,
"step": 989
},
{
"epoch": 0.03979899497487437,
"grad_norm": 1.921660304069519,
"learning_rate": 1.2997009750791032e-05,
"loss": 7.480654716491699,
"step": 990
},
{
"epoch": 0.039839195979899496,
"grad_norm": 2.1414740085601807,
"learning_rate": 1.2996992085908503e-05,
"loss": 7.440246105194092,
"step": 991
},
{
"epoch": 0.039879396984924625,
"grad_norm": 2.1125476360321045,
"learning_rate": 1.299697436902608e-05,
"loss": 7.920405387878418,
"step": 992
},
{
"epoch": 0.03991959798994975,
"grad_norm": 2.186231851577759,
"learning_rate": 1.2996956600144048e-05,
"loss": 7.298412322998047,
"step": 993
},
{
"epoch": 0.03995979899497487,
"grad_norm": 1.9438725709915161,
"learning_rate": 1.2996938779262688e-05,
"loss": 7.173574924468994,
"step": 994
},
{
"epoch": 0.04,
"grad_norm": 2.080956220626831,
"learning_rate": 1.299692090638229e-05,
"loss": 7.819553852081299,
"step": 995
},
{
"epoch": 0.040040201005025124,
"grad_norm": 2.169574499130249,
"learning_rate": 1.2996902981503136e-05,
"loss": 7.395140647888184,
"step": 996
},
{
"epoch": 0.04008040201005025,
"grad_norm": 2.6849775314331055,
"learning_rate": 1.2996885004625515e-05,
"loss": 7.713047981262207,
"step": 997
},
{
"epoch": 0.040120603015075376,
"grad_norm": 2.6359660625457764,
"learning_rate": 1.2996866975749715e-05,
"loss": 7.156382083892822,
"step": 998
},
{
"epoch": 0.040160804020100506,
"grad_norm": 2.0147151947021484,
"learning_rate": 1.2996848894876023e-05,
"loss": 8.23149299621582,
"step": 999
},
{
"epoch": 0.04020100502512563,
"grad_norm": 1.9605035781860352,
"learning_rate": 1.299683076200473e-05,
"loss": 7.794388771057129,
"step": 1000
},
{
"epoch": 0.04020100502512563,
"eval_loss": 7.648233413696289,
"eval_runtime": 120.6195,
"eval_samples_per_second": 4.145,
"eval_steps_per_second": 0.522,
"step": 1000
},
{
"epoch": 0.04024120603015075,
"grad_norm": 1.9647934436798096,
"learning_rate": 1.2996812577136128e-05,
"loss": 7.424760818481445,
"step": 1001
},
{
"epoch": 0.04028140703517588,
"grad_norm": 1.9651278257369995,
"learning_rate": 1.2996794340270507e-05,
"loss": 7.623063087463379,
"step": 1002
},
{
"epoch": 0.040321608040201004,
"grad_norm": 2.583991050720215,
"learning_rate": 1.2996776051408156e-05,
"loss": 8.265414237976074,
"step": 1003
},
{
"epoch": 0.040361809045226134,
"grad_norm": 1.9742240905761719,
"learning_rate": 1.2996757710549373e-05,
"loss": 7.523998737335205,
"step": 1004
},
{
"epoch": 0.04040201005025126,
"grad_norm": 1.99375581741333,
"learning_rate": 1.2996739317694447e-05,
"loss": 7.340415954589844,
"step": 1005
},
{
"epoch": 0.04044221105527638,
"grad_norm": 2.154217004776001,
"learning_rate": 1.2996720872843676e-05,
"loss": 7.539902210235596,
"step": 1006
},
{
"epoch": 0.04048241206030151,
"grad_norm": 2.166309118270874,
"learning_rate": 1.2996702375997354e-05,
"loss": 8.05794906616211,
"step": 1007
},
{
"epoch": 0.04052261306532663,
"grad_norm": 1.962354302406311,
"learning_rate": 1.2996683827155776e-05,
"loss": 7.4061713218688965,
"step": 1008
},
{
"epoch": 0.04056281407035176,
"grad_norm": 2.2369070053100586,
"learning_rate": 1.2996665226319241e-05,
"loss": 6.974580764770508,
"step": 1009
},
{
"epoch": 0.040603015075376885,
"grad_norm": 1.9999865293502808,
"learning_rate": 1.2996646573488049e-05,
"loss": 7.634210109710693,
"step": 1010
},
{
"epoch": 0.04064321608040201,
"grad_norm": 1.9313197135925293,
"learning_rate": 1.2996627868662493e-05,
"loss": 7.387241363525391,
"step": 1011
},
{
"epoch": 0.04068341708542714,
"grad_norm": 2.234915018081665,
"learning_rate": 1.2996609111842877e-05,
"loss": 7.544217109680176,
"step": 1012
},
{
"epoch": 0.04072361809045226,
"grad_norm": 1.8732115030288696,
"learning_rate": 1.2996590303029499e-05,
"loss": 7.39097785949707,
"step": 1013
},
{
"epoch": 0.04076381909547739,
"grad_norm": 2.305431365966797,
"learning_rate": 1.2996571442222662e-05,
"loss": 7.880208969116211,
"step": 1014
},
{
"epoch": 0.04080402010050251,
"grad_norm": 2.3515284061431885,
"learning_rate": 1.2996552529422667e-05,
"loss": 7.499734878540039,
"step": 1015
},
{
"epoch": 0.040844221105527635,
"grad_norm": 2.3193254470825195,
"learning_rate": 1.2996533564629818e-05,
"loss": 7.619176387786865,
"step": 1016
},
{
"epoch": 0.040884422110552765,
"grad_norm": 2.2258827686309814,
"learning_rate": 1.2996514547844417e-05,
"loss": 7.840350151062012,
"step": 1017
},
{
"epoch": 0.04092462311557789,
"grad_norm": 2.2346391677856445,
"learning_rate": 1.299649547906677e-05,
"loss": 7.729151725769043,
"step": 1018
},
{
"epoch": 0.04096482412060302,
"grad_norm": 1.905234694480896,
"learning_rate": 1.299647635829718e-05,
"loss": 7.512173175811768,
"step": 1019
},
{
"epoch": 0.04100502512562814,
"grad_norm": 1.9738203287124634,
"learning_rate": 1.2996457185535958e-05,
"loss": 7.393809795379639,
"step": 1020
},
{
"epoch": 0.04104522613065326,
"grad_norm": 1.9716218709945679,
"learning_rate": 1.2996437960783405e-05,
"loss": 7.298264980316162,
"step": 1021
},
{
"epoch": 0.04108542713567839,
"grad_norm": 2.154632806777954,
"learning_rate": 1.2996418684039833e-05,
"loss": 7.392561912536621,
"step": 1022
},
{
"epoch": 0.041125628140703516,
"grad_norm": 2.1541008949279785,
"learning_rate": 1.299639935530555e-05,
"loss": 7.371748924255371,
"step": 1023
},
{
"epoch": 0.041165829145728645,
"grad_norm": 2.0432937145233154,
"learning_rate": 1.2996379974580865e-05,
"loss": 7.9208784103393555,
"step": 1024
},
{
"epoch": 0.04120603015075377,
"grad_norm": 1.9033386707305908,
"learning_rate": 1.299636054186609e-05,
"loss": 7.790449142456055,
"step": 1025
},
{
"epoch": 0.04124623115577889,
"grad_norm": 2.0687265396118164,
"learning_rate": 1.2996341057161532e-05,
"loss": 7.488986015319824,
"step": 1026
},
{
"epoch": 0.04128643216080402,
"grad_norm": 2.0277059078216553,
"learning_rate": 1.2996321520467508e-05,
"loss": 7.8934245109558105,
"step": 1027
},
{
"epoch": 0.041326633165829144,
"grad_norm": 1.935575008392334,
"learning_rate": 1.2996301931784327e-05,
"loss": 7.83811092376709,
"step": 1028
},
{
"epoch": 0.04136683417085427,
"grad_norm": 1.9624850749969482,
"learning_rate": 1.2996282291112305e-05,
"loss": 7.528314590454102,
"step": 1029
},
{
"epoch": 0.041407035175879396,
"grad_norm": 1.9350802898406982,
"learning_rate": 1.2996262598451756e-05,
"loss": 7.525091171264648,
"step": 1030
},
{
"epoch": 0.041447236180904526,
"grad_norm": 1.8604904413223267,
"learning_rate": 1.2996242853802996e-05,
"loss": 7.738095760345459,
"step": 1031
},
{
"epoch": 0.04148743718592965,
"grad_norm": 2.0619051456451416,
"learning_rate": 1.299622305716634e-05,
"loss": 7.485928535461426,
"step": 1032
},
{
"epoch": 0.04152763819095477,
"grad_norm": 2.357135534286499,
"learning_rate": 1.2996203208542106e-05,
"loss": 7.62596321105957,
"step": 1033
},
{
"epoch": 0.0415678391959799,
"grad_norm": 1.98032808303833,
"learning_rate": 1.2996183307930611e-05,
"loss": 7.484348773956299,
"step": 1034
},
{
"epoch": 0.041608040201005024,
"grad_norm": 1.99095618724823,
"learning_rate": 1.2996163355332174e-05,
"loss": 7.474386215209961,
"step": 1035
},
{
"epoch": 0.041648241206030154,
"grad_norm": 2.08081316947937,
"learning_rate": 1.2996143350747116e-05,
"loss": 7.6415300369262695,
"step": 1036
},
{
"epoch": 0.04168844221105528,
"grad_norm": 2.0602850914001465,
"learning_rate": 1.2996123294175754e-05,
"loss": 7.767189979553223,
"step": 1037
},
{
"epoch": 0.0417286432160804,
"grad_norm": 1.9147017002105713,
"learning_rate": 1.2996103185618413e-05,
"loss": 7.419482231140137,
"step": 1038
},
{
"epoch": 0.04176884422110553,
"grad_norm": 2.235811948776245,
"learning_rate": 1.2996083025075412e-05,
"loss": 7.578300476074219,
"step": 1039
},
{
"epoch": 0.04180904522613065,
"grad_norm": 2.471583366394043,
"learning_rate": 1.2996062812547077e-05,
"loss": 7.56027364730835,
"step": 1040
},
{
"epoch": 0.04184924623115578,
"grad_norm": 1.984358549118042,
"learning_rate": 1.2996042548033728e-05,
"loss": 7.977116584777832,
"step": 1041
},
{
"epoch": 0.041889447236180904,
"grad_norm": 2.190720796585083,
"learning_rate": 1.2996022231535691e-05,
"loss": 7.705233573913574,
"step": 1042
},
{
"epoch": 0.04192964824120603,
"grad_norm": 2.129343271255493,
"learning_rate": 1.2996001863053293e-05,
"loss": 7.521746635437012,
"step": 1043
},
{
"epoch": 0.04196984924623116,
"grad_norm": 1.8890019655227661,
"learning_rate": 1.2995981442586858e-05,
"loss": 7.702261924743652,
"step": 1044
},
{
"epoch": 0.04201005025125628,
"grad_norm": 2.15204119682312,
"learning_rate": 1.2995960970136716e-05,
"loss": 7.473725318908691,
"step": 1045
},
{
"epoch": 0.04205025125628141,
"grad_norm": 2.081963539123535,
"learning_rate": 1.2995940445703192e-05,
"loss": 7.6407999992370605,
"step": 1046
},
{
"epoch": 0.04209045226130653,
"grad_norm": 2.1837546825408936,
"learning_rate": 1.2995919869286614e-05,
"loss": 7.325584411621094,
"step": 1047
},
{
"epoch": 0.042130653266331655,
"grad_norm": 2.2455742359161377,
"learning_rate": 1.2995899240887313e-05,
"loss": 7.453568458557129,
"step": 1048
},
{
"epoch": 0.042170854271356785,
"grad_norm": 1.9134317636489868,
"learning_rate": 1.2995878560505622e-05,
"loss": 7.808651447296143,
"step": 1049
},
{
"epoch": 0.04221105527638191,
"grad_norm": 2.5268893241882324,
"learning_rate": 1.2995857828141867e-05,
"loss": 7.849637985229492,
"step": 1050
},
{
"epoch": 0.04225125628140704,
"grad_norm": 2.0407471656799316,
"learning_rate": 1.2995837043796385e-05,
"loss": 7.753460884094238,
"step": 1051
},
{
"epoch": 0.04229145728643216,
"grad_norm": 2.0526132583618164,
"learning_rate": 1.2995816207469505e-05,
"loss": 7.847292900085449,
"step": 1052
},
{
"epoch": 0.04233165829145728,
"grad_norm": 1.9298063516616821,
"learning_rate": 1.299579531916156e-05,
"loss": 7.6341423988342285,
"step": 1053
},
{
"epoch": 0.04237185929648241,
"grad_norm": 2.1427383422851562,
"learning_rate": 1.299577437887289e-05,
"loss": 7.296820640563965,
"step": 1054
},
{
"epoch": 0.042412060301507536,
"grad_norm": 2.4334497451782227,
"learning_rate": 1.2995753386603826e-05,
"loss": 7.197660446166992,
"step": 1055
},
{
"epoch": 0.042452261306532665,
"grad_norm": 2.0752665996551514,
"learning_rate": 1.2995732342354706e-05,
"loss": 7.956886291503906,
"step": 1056
},
{
"epoch": 0.04249246231155779,
"grad_norm": 1.9640417098999023,
"learning_rate": 1.2995711246125865e-05,
"loss": 7.57272481918335,
"step": 1057
},
{
"epoch": 0.04253266331658292,
"grad_norm": 2.2567358016967773,
"learning_rate": 1.2995690097917642e-05,
"loss": 7.505478858947754,
"step": 1058
},
{
"epoch": 0.04257286432160804,
"grad_norm": 1.9823874235153198,
"learning_rate": 1.2995668897730379e-05,
"loss": 8.06527328491211,
"step": 1059
},
{
"epoch": 0.042613065326633164,
"grad_norm": 2.724853754043579,
"learning_rate": 1.2995647645564409e-05,
"loss": 7.599035263061523,
"step": 1060
},
{
"epoch": 0.04265326633165829,
"grad_norm": 2.4285950660705566,
"learning_rate": 1.2995626341420076e-05,
"loss": 7.517922401428223,
"step": 1061
},
{
"epoch": 0.042693467336683416,
"grad_norm": 2.11398983001709,
"learning_rate": 1.299560498529772e-05,
"loss": 7.438998222351074,
"step": 1062
},
{
"epoch": 0.042733668341708546,
"grad_norm": 2.112637758255005,
"learning_rate": 1.2995583577197686e-05,
"loss": 7.810076713562012,
"step": 1063
},
{
"epoch": 0.04277386934673367,
"grad_norm": 2.150644063949585,
"learning_rate": 1.2995562117120314e-05,
"loss": 8.098053932189941,
"step": 1064
},
{
"epoch": 0.04281407035175879,
"grad_norm": 1.9623936414718628,
"learning_rate": 1.2995540605065948e-05,
"loss": 7.650564193725586,
"step": 1065
},
{
"epoch": 0.04285427135678392,
"grad_norm": 2.5243639945983887,
"learning_rate": 1.2995519041034933e-05,
"loss": 7.634098052978516,
"step": 1066
},
{
"epoch": 0.042894472361809044,
"grad_norm": 2.0223498344421387,
"learning_rate": 1.2995497425027616e-05,
"loss": 7.379621505737305,
"step": 1067
},
{
"epoch": 0.042934673366834174,
"grad_norm": 1.9696272611618042,
"learning_rate": 1.299547575704434e-05,
"loss": 7.556798934936523,
"step": 1068
},
{
"epoch": 0.0429748743718593,
"grad_norm": 1.9297029972076416,
"learning_rate": 1.2995454037085452e-05,
"loss": 7.874904632568359,
"step": 1069
},
{
"epoch": 0.04301507537688442,
"grad_norm": 2.1334545612335205,
"learning_rate": 1.2995432265151303e-05,
"loss": 7.8910369873046875,
"step": 1070
},
{
"epoch": 0.04305527638190955,
"grad_norm": 1.844298243522644,
"learning_rate": 1.2995410441242239e-05,
"loss": 7.6033525466918945,
"step": 1071
},
{
"epoch": 0.04309547738693467,
"grad_norm": 2.2114458084106445,
"learning_rate": 1.2995388565358612e-05,
"loss": 7.706252098083496,
"step": 1072
},
{
"epoch": 0.0431356783919598,
"grad_norm": 2.0706441402435303,
"learning_rate": 1.2995366637500769e-05,
"loss": 8.203407287597656,
"step": 1073
},
{
"epoch": 0.043175879396984924,
"grad_norm": 1.8346878290176392,
"learning_rate": 1.2995344657669063e-05,
"loss": 7.6609416007995605,
"step": 1074
},
{
"epoch": 0.04321608040201005,
"grad_norm": 2.4912562370300293,
"learning_rate": 1.2995322625863847e-05,
"loss": 7.173084259033203,
"step": 1075
},
{
"epoch": 0.04325628140703518,
"grad_norm": 1.9537153244018555,
"learning_rate": 1.2995300542085471e-05,
"loss": 7.615156173706055,
"step": 1076
},
{
"epoch": 0.0432964824120603,
"grad_norm": 2.214332342147827,
"learning_rate": 1.2995278406334292e-05,
"loss": 7.6517839431762695,
"step": 1077
},
{
"epoch": 0.04333668341708543,
"grad_norm": 2.0497095584869385,
"learning_rate": 1.2995256218610663e-05,
"loss": 8.096546173095703,
"step": 1078
},
{
"epoch": 0.04337688442211055,
"grad_norm": 2.690204620361328,
"learning_rate": 1.299523397891494e-05,
"loss": 8.465989112854004,
"step": 1079
},
{
"epoch": 0.043417085427135675,
"grad_norm": 2.164708375930786,
"learning_rate": 1.2995211687247477e-05,
"loss": 7.418861389160156,
"step": 1080
},
{
"epoch": 0.043457286432160805,
"grad_norm": 2.336426019668579,
"learning_rate": 1.2995189343608634e-05,
"loss": 7.431804656982422,
"step": 1081
},
{
"epoch": 0.04349748743718593,
"grad_norm": 1.9586262702941895,
"learning_rate": 1.2995166947998766e-05,
"loss": 7.861602783203125,
"step": 1082
},
{
"epoch": 0.04353768844221106,
"grad_norm": 2.439528703689575,
"learning_rate": 1.2995144500418233e-05,
"loss": 8.09302806854248,
"step": 1083
},
{
"epoch": 0.04357788944723618,
"grad_norm": 2.0496723651885986,
"learning_rate": 1.2995122000867394e-05,
"loss": 7.833438873291016,
"step": 1084
},
{
"epoch": 0.04361809045226131,
"grad_norm": 2.604461193084717,
"learning_rate": 1.299509944934661e-05,
"loss": 8.255422592163086,
"step": 1085
},
{
"epoch": 0.04365829145728643,
"grad_norm": 2.063488245010376,
"learning_rate": 1.2995076845856242e-05,
"loss": 7.748636245727539,
"step": 1086
},
{
"epoch": 0.043698492462311556,
"grad_norm": 1.9819589853286743,
"learning_rate": 1.2995054190396651e-05,
"loss": 7.769131660461426,
"step": 1087
},
{
"epoch": 0.043738693467336685,
"grad_norm": 2.5064778327941895,
"learning_rate": 1.2995031482968202e-05,
"loss": 7.908186912536621,
"step": 1088
},
{
"epoch": 0.04377889447236181,
"grad_norm": 2.07859468460083,
"learning_rate": 1.2995008723571258e-05,
"loss": 7.642248153686523,
"step": 1089
},
{
"epoch": 0.04381909547738694,
"grad_norm": 1.9993317127227783,
"learning_rate": 1.2994985912206182e-05,
"loss": 7.4376420974731445,
"step": 1090
},
{
"epoch": 0.04385929648241206,
"grad_norm": 1.905139446258545,
"learning_rate": 1.299496304887334e-05,
"loss": 7.2534942626953125,
"step": 1091
},
{
"epoch": 0.043899497487437183,
"grad_norm": 2.0500993728637695,
"learning_rate": 1.2994940133573099e-05,
"loss": 7.810580253601074,
"step": 1092
},
{
"epoch": 0.04393969849246231,
"grad_norm": 2.1784050464630127,
"learning_rate": 1.2994917166305825e-05,
"loss": 7.542360305786133,
"step": 1093
},
{
"epoch": 0.043979899497487436,
"grad_norm": 2.253797769546509,
"learning_rate": 1.2994894147071886e-05,
"loss": 7.5378313064575195,
"step": 1094
},
{
"epoch": 0.044020100502512566,
"grad_norm": 2.1474335193634033,
"learning_rate": 1.299487107587165e-05,
"loss": 7.614965438842773,
"step": 1095
},
{
"epoch": 0.04406030150753769,
"grad_norm": 2.4032318592071533,
"learning_rate": 1.299484795270549e-05,
"loss": 7.998912811279297,
"step": 1096
},
{
"epoch": 0.04410050251256281,
"grad_norm": 2.238778829574585,
"learning_rate": 1.2994824777573771e-05,
"loss": 7.881226539611816,
"step": 1097
},
{
"epoch": 0.04414070351758794,
"grad_norm": 2.150407075881958,
"learning_rate": 1.2994801550476866e-05,
"loss": 7.787651062011719,
"step": 1098
},
{
"epoch": 0.044180904522613064,
"grad_norm": 2.3664767742156982,
"learning_rate": 1.299477827141515e-05,
"loss": 7.7881317138671875,
"step": 1099
},
{
"epoch": 0.044221105527638194,
"grad_norm": 2.401137590408325,
"learning_rate": 1.2994754940388992e-05,
"loss": 7.748221397399902,
"step": 1100
},
{
"epoch": 0.044261306532663316,
"grad_norm": 1.9953221082687378,
"learning_rate": 1.2994731557398768e-05,
"loss": 7.415718078613281,
"step": 1101
},
{
"epoch": 0.04430150753768844,
"grad_norm": 2.1272494792938232,
"learning_rate": 1.2994708122444851e-05,
"loss": 7.408730506896973,
"step": 1102
},
{
"epoch": 0.04434170854271357,
"grad_norm": 2.018249034881592,
"learning_rate": 1.2994684635527618e-05,
"loss": 7.550690650939941,
"step": 1103
},
{
"epoch": 0.04438190954773869,
"grad_norm": 1.8975602388381958,
"learning_rate": 1.2994661096647443e-05,
"loss": 7.489997863769531,
"step": 1104
},
{
"epoch": 0.04442211055276382,
"grad_norm": 3.2015693187713623,
"learning_rate": 1.2994637505804705e-05,
"loss": 7.835939407348633,
"step": 1105
},
{
"epoch": 0.044462311557788944,
"grad_norm": 2.2737960815429688,
"learning_rate": 1.299461386299978e-05,
"loss": 7.5651679039001465,
"step": 1106
},
{
"epoch": 0.04450251256281407,
"grad_norm": 2.157161235809326,
"learning_rate": 1.2994590168233048e-05,
"loss": 7.561855316162109,
"step": 1107
},
{
"epoch": 0.0445427135678392,
"grad_norm": 2.2057626247406006,
"learning_rate": 1.2994566421504887e-05,
"loss": 7.944631576538086,
"step": 1108
},
{
"epoch": 0.04458291457286432,
"grad_norm": 2.33642315864563,
"learning_rate": 1.2994542622815678e-05,
"loss": 7.172248840332031,
"step": 1109
},
{
"epoch": 0.04462311557788945,
"grad_norm": 2.0298221111297607,
"learning_rate": 1.2994518772165804e-05,
"loss": 7.591313362121582,
"step": 1110
},
{
"epoch": 0.04466331658291457,
"grad_norm": 2.0677175521850586,
"learning_rate": 1.2994494869555644e-05,
"loss": 7.754199981689453,
"step": 1111
},
{
"epoch": 0.044703517587939695,
"grad_norm": 2.0928587913513184,
"learning_rate": 1.2994470914985583e-05,
"loss": 7.732008934020996,
"step": 1112
},
{
"epoch": 0.044743718592964825,
"grad_norm": 2.358010768890381,
"learning_rate": 1.2994446908456004e-05,
"loss": 7.445245742797852,
"step": 1113
},
{
"epoch": 0.04478391959798995,
"grad_norm": 2.182504177093506,
"learning_rate": 1.299442284996729e-05,
"loss": 7.711686134338379,
"step": 1114
},
{
"epoch": 0.04482412060301508,
"grad_norm": 2.0839929580688477,
"learning_rate": 1.2994398739519828e-05,
"loss": 7.2596845626831055,
"step": 1115
},
{
"epoch": 0.0448643216080402,
"grad_norm": 1.909616231918335,
"learning_rate": 1.2994374577114002e-05,
"loss": 7.642219543457031,
"step": 1116
},
{
"epoch": 0.04490452261306533,
"grad_norm": 1.9352635145187378,
"learning_rate": 1.2994350362750205e-05,
"loss": 7.489213943481445,
"step": 1117
},
{
"epoch": 0.04494472361809045,
"grad_norm": 2.0818593502044678,
"learning_rate": 1.2994326096428817e-05,
"loss": 7.385632514953613,
"step": 1118
},
{
"epoch": 0.044984924623115576,
"grad_norm": 2.5385594367980957,
"learning_rate": 1.299430177815023e-05,
"loss": 7.302634239196777,
"step": 1119
},
{
"epoch": 0.045025125628140705,
"grad_norm": 2.0653460025787354,
"learning_rate": 1.2994277407914836e-05,
"loss": 7.389494895935059,
"step": 1120
},
{
"epoch": 0.04506532663316583,
"grad_norm": 2.011009931564331,
"learning_rate": 1.2994252985723019e-05,
"loss": 7.466405868530273,
"step": 1121
},
{
"epoch": 0.04510552763819096,
"grad_norm": 2.0616321563720703,
"learning_rate": 1.2994228511575176e-05,
"loss": 7.499314308166504,
"step": 1122
},
{
"epoch": 0.04514572864321608,
"grad_norm": 2.0222361087799072,
"learning_rate": 1.2994203985471697e-05,
"loss": 7.2874884605407715,
"step": 1123
},
{
"epoch": 0.0451859296482412,
"grad_norm": 2.3633694648742676,
"learning_rate": 1.2994179407412974e-05,
"loss": 7.972682476043701,
"step": 1124
},
{
"epoch": 0.04522613065326633,
"grad_norm": 4.869113445281982,
"learning_rate": 1.2994154777399401e-05,
"loss": 7.976752281188965,
"step": 1125
},
{
"epoch": 0.045266331658291456,
"grad_norm": 2.0865206718444824,
"learning_rate": 1.2994130095431374e-05,
"loss": 7.594989776611328,
"step": 1126
},
{
"epoch": 0.045306532663316586,
"grad_norm": 2.191959857940674,
"learning_rate": 1.2994105361509285e-05,
"loss": 7.538678169250488,
"step": 1127
},
{
"epoch": 0.04534673366834171,
"grad_norm": 2.047177791595459,
"learning_rate": 1.2994080575633533e-05,
"loss": 7.28386116027832,
"step": 1128
},
{
"epoch": 0.04538693467336683,
"grad_norm": 2.0837602615356445,
"learning_rate": 1.2994055737804515e-05,
"loss": 7.89670467376709,
"step": 1129
},
{
"epoch": 0.04542713567839196,
"grad_norm": 3.5438411235809326,
"learning_rate": 1.2994030848022626e-05,
"loss": 7.584082126617432,
"step": 1130
},
{
"epoch": 0.045467336683417084,
"grad_norm": 2.023857593536377,
"learning_rate": 1.2994005906288265e-05,
"loss": 7.338397026062012,
"step": 1131
},
{
"epoch": 0.045507537688442214,
"grad_norm": 2.4644007682800293,
"learning_rate": 1.2993980912601835e-05,
"loss": 7.442533493041992,
"step": 1132
},
{
"epoch": 0.045547738693467336,
"grad_norm": 1.917404055595398,
"learning_rate": 1.2993955866963734e-05,
"loss": 7.633648872375488,
"step": 1133
},
{
"epoch": 0.04558793969849246,
"grad_norm": 2.073277711868286,
"learning_rate": 1.2993930769374363e-05,
"loss": 8.52383804321289,
"step": 1134
},
{
"epoch": 0.04562814070351759,
"grad_norm": 2.2879319190979004,
"learning_rate": 1.2993905619834123e-05,
"loss": 7.173372745513916,
"step": 1135
},
{
"epoch": 0.04566834170854271,
"grad_norm": 2.5864052772521973,
"learning_rate": 1.299388041834342e-05,
"loss": 7.581547737121582,
"step": 1136
},
{
"epoch": 0.04570854271356784,
"grad_norm": 1.8495434522628784,
"learning_rate": 1.2993855164902654e-05,
"loss": 7.703059196472168,
"step": 1137
},
{
"epoch": 0.045748743718592964,
"grad_norm": 2.2552688121795654,
"learning_rate": 1.299382985951223e-05,
"loss": 7.5928754806518555,
"step": 1138
},
{
"epoch": 0.04578894472361809,
"grad_norm": 1.9250768423080444,
"learning_rate": 1.2993804502172556e-05,
"loss": 7.6344194412231445,
"step": 1139
},
{
"epoch": 0.04582914572864322,
"grad_norm": 2.2719733715057373,
"learning_rate": 1.2993779092884036e-05,
"loss": 7.800146579742432,
"step": 1140
},
{
"epoch": 0.04586934673366834,
"grad_norm": 2.1075382232666016,
"learning_rate": 1.2993753631647077e-05,
"loss": 7.762310981750488,
"step": 1141
},
{
"epoch": 0.04590954773869347,
"grad_norm": 2.441296100616455,
"learning_rate": 1.2993728118462085e-05,
"loss": 7.715604782104492,
"step": 1142
},
{
"epoch": 0.04594974874371859,
"grad_norm": 2.260329484939575,
"learning_rate": 1.2993702553329473e-05,
"loss": 7.509035110473633,
"step": 1143
},
{
"epoch": 0.04598994974874372,
"grad_norm": 2.604701519012451,
"learning_rate": 1.2993676936249645e-05,
"loss": 8.008878707885742,
"step": 1144
},
{
"epoch": 0.046030150753768845,
"grad_norm": 2.2117972373962402,
"learning_rate": 1.2993651267223018e-05,
"loss": 7.803744316101074,
"step": 1145
},
{
"epoch": 0.04607035175879397,
"grad_norm": 2.134823799133301,
"learning_rate": 1.2993625546249996e-05,
"loss": 7.728663444519043,
"step": 1146
},
{
"epoch": 0.0461105527638191,
"grad_norm": 2.4393646717071533,
"learning_rate": 1.2993599773330995e-05,
"loss": 7.000306129455566,
"step": 1147
},
{
"epoch": 0.04615075376884422,
"grad_norm": 2.2985217571258545,
"learning_rate": 1.2993573948466428e-05,
"loss": 7.8764543533325195,
"step": 1148
},
{
"epoch": 0.04619095477386935,
"grad_norm": 2.9288787841796875,
"learning_rate": 1.2993548071656705e-05,
"loss": 7.2001190185546875,
"step": 1149
},
{
"epoch": 0.04623115577889447,
"grad_norm": 1.9742954969406128,
"learning_rate": 1.2993522142902246e-05,
"loss": 7.632903575897217,
"step": 1150
},
{
"epoch": 0.046271356783919596,
"grad_norm": 2.063544750213623,
"learning_rate": 1.2993496162203461e-05,
"loss": 7.820664405822754,
"step": 1151
},
{
"epoch": 0.046311557788944725,
"grad_norm": 2.011608362197876,
"learning_rate": 1.299347012956077e-05,
"loss": 7.144486427307129,
"step": 1152
},
{
"epoch": 0.04635175879396985,
"grad_norm": 2.1323235034942627,
"learning_rate": 1.2993444044974587e-05,
"loss": 7.191654205322266,
"step": 1153
},
{
"epoch": 0.04639195979899498,
"grad_norm": 2.1209020614624023,
"learning_rate": 1.2993417908445331e-05,
"loss": 7.606924057006836,
"step": 1154
},
{
"epoch": 0.0464321608040201,
"grad_norm": 2.124530553817749,
"learning_rate": 1.2993391719973421e-05,
"loss": 7.504514694213867,
"step": 1155
},
{
"epoch": 0.04647236180904522,
"grad_norm": 2.3291125297546387,
"learning_rate": 1.2993365479559274e-05,
"loss": 7.616670608520508,
"step": 1156
},
{
"epoch": 0.04651256281407035,
"grad_norm": 2.1792171001434326,
"learning_rate": 1.2993339187203314e-05,
"loss": 7.929306983947754,
"step": 1157
},
{
"epoch": 0.046552763819095476,
"grad_norm": 2.0455727577209473,
"learning_rate": 1.2993312842905958e-05,
"loss": 7.439589500427246,
"step": 1158
},
{
"epoch": 0.046592964824120606,
"grad_norm": 2.454379081726074,
"learning_rate": 1.2993286446667632e-05,
"loss": 8.499064445495605,
"step": 1159
},
{
"epoch": 0.04663316582914573,
"grad_norm": 2.988615036010742,
"learning_rate": 1.2993259998488758e-05,
"loss": 7.756265163421631,
"step": 1160
},
{
"epoch": 0.04667336683417085,
"grad_norm": 2.2163357734680176,
"learning_rate": 1.2993233498369756e-05,
"loss": 8.105384826660156,
"step": 1161
},
{
"epoch": 0.04671356783919598,
"grad_norm": 2.0440549850463867,
"learning_rate": 1.2993206946311052e-05,
"loss": 7.561105728149414,
"step": 1162
},
{
"epoch": 0.046753768844221104,
"grad_norm": 2.326165199279785,
"learning_rate": 1.2993180342313075e-05,
"loss": 8.005739212036133,
"step": 1163
},
{
"epoch": 0.046793969849246234,
"grad_norm": 2.0255277156829834,
"learning_rate": 1.2993153686376246e-05,
"loss": 7.486833095550537,
"step": 1164
},
{
"epoch": 0.046834170854271356,
"grad_norm": 2.344179391860962,
"learning_rate": 1.2993126978500994e-05,
"loss": 7.545997142791748,
"step": 1165
},
{
"epoch": 0.04687437185929648,
"grad_norm": 2.0410964488983154,
"learning_rate": 1.2993100218687747e-05,
"loss": 7.296555995941162,
"step": 1166
},
{
"epoch": 0.04691457286432161,
"grad_norm": 1.9046251773834229,
"learning_rate": 1.2993073406936932e-05,
"loss": 7.597414493560791,
"step": 1167
},
{
"epoch": 0.04695477386934673,
"grad_norm": 2.34326434135437,
"learning_rate": 1.2993046543248982e-05,
"loss": 7.458094596862793,
"step": 1168
},
{
"epoch": 0.04699497487437186,
"grad_norm": 2.8770196437835693,
"learning_rate": 1.2993019627624322e-05,
"loss": 7.497445106506348,
"step": 1169
},
{
"epoch": 0.047035175879396984,
"grad_norm": 1.967977523803711,
"learning_rate": 1.2992992660063388e-05,
"loss": 7.568811416625977,
"step": 1170
},
{
"epoch": 0.04707537688442211,
"grad_norm": 2.193614959716797,
"learning_rate": 1.2992965640566608e-05,
"loss": 7.79426383972168,
"step": 1171
},
{
"epoch": 0.04711557788944724,
"grad_norm": 2.3244237899780273,
"learning_rate": 1.2992938569134417e-05,
"loss": 7.633689880371094,
"step": 1172
},
{
"epoch": 0.04715577889447236,
"grad_norm": 2.151047945022583,
"learning_rate": 1.2992911445767249e-05,
"loss": 7.922621726989746,
"step": 1173
},
{
"epoch": 0.04719597989949749,
"grad_norm": 2.082871437072754,
"learning_rate": 1.2992884270465535e-05,
"loss": 7.536629676818848,
"step": 1174
},
{
"epoch": 0.04723618090452261,
"grad_norm": 1.9975075721740723,
"learning_rate": 1.2992857043229715e-05,
"loss": 7.3271074295043945,
"step": 1175
},
{
"epoch": 0.04727638190954774,
"grad_norm": 2.0240468978881836,
"learning_rate": 1.2992829764060222e-05,
"loss": 7.656285285949707,
"step": 1176
},
{
"epoch": 0.047316582914572865,
"grad_norm": 2.164310932159424,
"learning_rate": 1.2992802432957492e-05,
"loss": 7.2415008544921875,
"step": 1177
},
{
"epoch": 0.04735678391959799,
"grad_norm": 2.027416467666626,
"learning_rate": 1.2992775049921964e-05,
"loss": 7.216300010681152,
"step": 1178
},
{
"epoch": 0.04739698492462312,
"grad_norm": 2.3678882122039795,
"learning_rate": 1.2992747614954079e-05,
"loss": 7.767563819885254,
"step": 1179
},
{
"epoch": 0.04743718592964824,
"grad_norm": 2.3998563289642334,
"learning_rate": 1.2992720128054271e-05,
"loss": 7.693877220153809,
"step": 1180
},
{
"epoch": 0.04747738693467337,
"grad_norm": 1.9702500104904175,
"learning_rate": 1.2992692589222986e-05,
"loss": 7.5117645263671875,
"step": 1181
},
{
"epoch": 0.04751758793969849,
"grad_norm": 2.05898380279541,
"learning_rate": 1.2992664998460661e-05,
"loss": 7.579432010650635,
"step": 1182
},
{
"epoch": 0.047557788944723615,
"grad_norm": 2.186025619506836,
"learning_rate": 1.299263735576774e-05,
"loss": 7.2836198806762695,
"step": 1183
},
{
"epoch": 0.047597989949748745,
"grad_norm": 2.300765037536621,
"learning_rate": 1.2992609661144665e-05,
"loss": 8.243400573730469,
"step": 1184
},
{
"epoch": 0.04763819095477387,
"grad_norm": 1.988911509513855,
"learning_rate": 1.299258191459188e-05,
"loss": 7.714032173156738,
"step": 1185
},
{
"epoch": 0.047678391959799,
"grad_norm": 2.384239912033081,
"learning_rate": 1.2992554116109825e-05,
"loss": 7.923232078552246,
"step": 1186
},
{
"epoch": 0.04771859296482412,
"grad_norm": 2.3819327354431152,
"learning_rate": 1.2992526265698953e-05,
"loss": 7.411757469177246,
"step": 1187
},
{
"epoch": 0.04775879396984924,
"grad_norm": 2.0738980770111084,
"learning_rate": 1.2992498363359703e-05,
"loss": 7.613278865814209,
"step": 1188
},
{
"epoch": 0.04779899497487437,
"grad_norm": 2.077638864517212,
"learning_rate": 1.2992470409092527e-05,
"loss": 7.705620765686035,
"step": 1189
},
{
"epoch": 0.047839195979899496,
"grad_norm": 2.197756052017212,
"learning_rate": 1.2992442402897868e-05,
"loss": 7.505415439605713,
"step": 1190
},
{
"epoch": 0.047879396984924626,
"grad_norm": 2.1498143672943115,
"learning_rate": 1.2992414344776178e-05,
"loss": 6.958484172821045,
"step": 1191
},
{
"epoch": 0.04791959798994975,
"grad_norm": 2.1028048992156982,
"learning_rate": 1.2992386234727906e-05,
"loss": 7.949058532714844,
"step": 1192
},
{
"epoch": 0.04795979899497487,
"grad_norm": 2.25395131111145,
"learning_rate": 1.2992358072753501e-05,
"loss": 7.895046234130859,
"step": 1193
},
{
"epoch": 0.048,
"grad_norm": 2.2699599266052246,
"learning_rate": 1.2992329858853414e-05,
"loss": 7.958035469055176,
"step": 1194
},
{
"epoch": 0.048040201005025124,
"grad_norm": 2.2176783084869385,
"learning_rate": 1.2992301593028099e-05,
"loss": 7.826164245605469,
"step": 1195
},
{
"epoch": 0.048080402010050254,
"grad_norm": 2.162900686264038,
"learning_rate": 1.2992273275278005e-05,
"loss": 7.645927429199219,
"step": 1196
},
{
"epoch": 0.048120603015075376,
"grad_norm": 2.0950546264648438,
"learning_rate": 1.299224490560359e-05,
"loss": 7.491610050201416,
"step": 1197
},
{
"epoch": 0.0481608040201005,
"grad_norm": 2.234524726867676,
"learning_rate": 1.2992216484005303e-05,
"loss": 7.398894309997559,
"step": 1198
},
{
"epoch": 0.04820100502512563,
"grad_norm": 1.978922724723816,
"learning_rate": 1.2992188010483603e-05,
"loss": 7.866137981414795,
"step": 1199
},
{
"epoch": 0.04824120603015075,
"grad_norm": 2.093303680419922,
"learning_rate": 1.2992159485038945e-05,
"loss": 7.166652679443359,
"step": 1200
},
{
"epoch": 0.04824120603015075,
"eval_loss": 7.601953506469727,
"eval_runtime": 120.6031,
"eval_samples_per_second": 4.146,
"eval_steps_per_second": 0.522,
"step": 1200
},
{
"epoch": 0.04828140703517588,
"grad_norm": 2.376762628555298,
"learning_rate": 1.2992130907671784e-05,
"loss": 7.434047698974609,
"step": 1201
},
{
"epoch": 0.048321608040201004,
"grad_norm": 2.5020923614501953,
"learning_rate": 1.2992102278382583e-05,
"loss": 7.665665626525879,
"step": 1202
},
{
"epoch": 0.048361809045226134,
"grad_norm": 2.1845104694366455,
"learning_rate": 1.2992073597171796e-05,
"loss": 7.73158073425293,
"step": 1203
},
{
"epoch": 0.04840201005025126,
"grad_norm": 2.0963730812072754,
"learning_rate": 1.2992044864039882e-05,
"loss": 8.00433349609375,
"step": 1204
},
{
"epoch": 0.04844221105527638,
"grad_norm": 2.2529115676879883,
"learning_rate": 1.2992016078987304e-05,
"loss": 7.849309921264648,
"step": 1205
},
{
"epoch": 0.04848241206030151,
"grad_norm": 2.2306785583496094,
"learning_rate": 1.299198724201452e-05,
"loss": 8.04069709777832,
"step": 1206
},
{
"epoch": 0.04852261306532663,
"grad_norm": 2.8951141834259033,
"learning_rate": 1.2991958353121993e-05,
"loss": 7.365753650665283,
"step": 1207
},
{
"epoch": 0.04856281407035176,
"grad_norm": 2.40761661529541,
"learning_rate": 1.2991929412310188e-05,
"loss": 7.289119720458984,
"step": 1208
},
{
"epoch": 0.048603015075376885,
"grad_norm": 2.1375362873077393,
"learning_rate": 1.2991900419579565e-05,
"loss": 7.481162071228027,
"step": 1209
},
{
"epoch": 0.04864321608040201,
"grad_norm": 2.035146474838257,
"learning_rate": 1.2991871374930588e-05,
"loss": 7.348549842834473,
"step": 1210
},
{
"epoch": 0.04868341708542714,
"grad_norm": 2.6397273540496826,
"learning_rate": 1.2991842278363727e-05,
"loss": 7.498271942138672,
"step": 1211
},
{
"epoch": 0.04872361809045226,
"grad_norm": 2.333670139312744,
"learning_rate": 1.2991813129879442e-05,
"loss": 7.312775611877441,
"step": 1212
},
{
"epoch": 0.04876381909547739,
"grad_norm": 2.055560827255249,
"learning_rate": 1.2991783929478205e-05,
"loss": 7.722532272338867,
"step": 1213
},
{
"epoch": 0.04880402010050251,
"grad_norm": 2.2576043605804443,
"learning_rate": 1.299175467716048e-05,
"loss": 7.380242347717285,
"step": 1214
},
{
"epoch": 0.048844221105527635,
"grad_norm": 2.091726064682007,
"learning_rate": 1.2991725372926738e-05,
"loss": 7.018467903137207,
"step": 1215
},
{
"epoch": 0.048884422110552765,
"grad_norm": 2.095853805541992,
"learning_rate": 1.2991696016777445e-05,
"loss": 7.427759647369385,
"step": 1216
},
{
"epoch": 0.04892462311557789,
"grad_norm": 2.031785726547241,
"learning_rate": 1.2991666608713075e-05,
"loss": 7.449406623840332,
"step": 1217
},
{
"epoch": 0.04896482412060302,
"grad_norm": 2.045391798019409,
"learning_rate": 1.2991637148734097e-05,
"loss": 7.551580429077148,
"step": 1218
},
{
"epoch": 0.04900502512562814,
"grad_norm": 2.646986246109009,
"learning_rate": 1.2991607636840981e-05,
"loss": 7.398261547088623,
"step": 1219
},
{
"epoch": 0.04904522613065326,
"grad_norm": 2.6692962646484375,
"learning_rate": 1.2991578073034204e-05,
"loss": 7.179652214050293,
"step": 1220
},
{
"epoch": 0.04908542713567839,
"grad_norm": 2.3517322540283203,
"learning_rate": 1.2991548457314236e-05,
"loss": 7.658819198608398,
"step": 1221
},
{
"epoch": 0.049125628140703516,
"grad_norm": 2.188708543777466,
"learning_rate": 1.2991518789681551e-05,
"loss": 7.897691249847412,
"step": 1222
},
{
"epoch": 0.049165829145728646,
"grad_norm": 2.3377509117126465,
"learning_rate": 1.2991489070136628e-05,
"loss": 7.529051780700684,
"step": 1223
},
{
"epoch": 0.04920603015075377,
"grad_norm": 2.0964505672454834,
"learning_rate": 1.2991459298679938e-05,
"loss": 7.407858371734619,
"step": 1224
},
{
"epoch": 0.04924623115577889,
"grad_norm": 2.143171787261963,
"learning_rate": 1.299142947531196e-05,
"loss": 7.579808712005615,
"step": 1225
},
{
"epoch": 0.04928643216080402,
"grad_norm": 2.368021249771118,
"learning_rate": 1.2991399600033174e-05,
"loss": 7.521397590637207,
"step": 1226
},
{
"epoch": 0.049326633165829144,
"grad_norm": 2.6416635513305664,
"learning_rate": 1.2991369672844055e-05,
"loss": 8.124269485473633,
"step": 1227
},
{
"epoch": 0.049366834170854274,
"grad_norm": 2.2667596340179443,
"learning_rate": 1.2991339693745084e-05,
"loss": 7.276832580566406,
"step": 1228
},
{
"epoch": 0.049407035175879396,
"grad_norm": 2.399076461791992,
"learning_rate": 1.299130966273674e-05,
"loss": 8.358059883117676,
"step": 1229
},
{
"epoch": 0.049447236180904526,
"grad_norm": 2.2396750450134277,
"learning_rate": 1.2991279579819504e-05,
"loss": 7.539815902709961,
"step": 1230
},
{
"epoch": 0.04948743718592965,
"grad_norm": 2.1968793869018555,
"learning_rate": 1.2991249444993859e-05,
"loss": 8.100176811218262,
"step": 1231
},
{
"epoch": 0.04952763819095477,
"grad_norm": 2.496062994003296,
"learning_rate": 1.2991219258260287e-05,
"loss": 7.269747734069824,
"step": 1232
},
{
"epoch": 0.0495678391959799,
"grad_norm": 2.3952574729919434,
"learning_rate": 1.2991189019619272e-05,
"loss": 7.949479103088379,
"step": 1233
},
{
"epoch": 0.049608040201005024,
"grad_norm": 2.091010808944702,
"learning_rate": 1.2991158729071297e-05,
"loss": 7.737839698791504,
"step": 1234
},
{
"epoch": 0.049648241206030154,
"grad_norm": 2.5007894039154053,
"learning_rate": 1.2991128386616847e-05,
"loss": 7.516338348388672,
"step": 1235
},
{
"epoch": 0.04968844221105528,
"grad_norm": 2.5835771560668945,
"learning_rate": 1.2991097992256409e-05,
"loss": 7.414939880371094,
"step": 1236
},
{
"epoch": 0.0497286432160804,
"grad_norm": 2.294083833694458,
"learning_rate": 1.2991067545990472e-05,
"loss": 7.801116943359375,
"step": 1237
},
{
"epoch": 0.04976884422110553,
"grad_norm": 2.124335765838623,
"learning_rate": 1.2991037047819517e-05,
"loss": 7.500946044921875,
"step": 1238
},
{
"epoch": 0.04980904522613065,
"grad_norm": 2.2772133350372314,
"learning_rate": 1.2991006497744039e-05,
"loss": 7.486501693725586,
"step": 1239
},
{
"epoch": 0.04984924623115578,
"grad_norm": 2.3502471446990967,
"learning_rate": 1.2990975895764523e-05,
"loss": 7.676392555236816,
"step": 1240
},
{
"epoch": 0.049889447236180905,
"grad_norm": 2.215585470199585,
"learning_rate": 1.2990945241881462e-05,
"loss": 7.603403091430664,
"step": 1241
},
{
"epoch": 0.04992964824120603,
"grad_norm": 2.2088534832000732,
"learning_rate": 1.2990914536095344e-05,
"loss": 7.5379133224487305,
"step": 1242
},
{
"epoch": 0.04996984924623116,
"grad_norm": 2.0795013904571533,
"learning_rate": 1.2990883778406665e-05,
"loss": 7.9218645095825195,
"step": 1243
},
{
"epoch": 0.05001005025125628,
"grad_norm": 1.964362382888794,
"learning_rate": 1.2990852968815914e-05,
"loss": 7.6266984939575195,
"step": 1244
},
{
"epoch": 0.05005025125628141,
"grad_norm": 2.046109199523926,
"learning_rate": 1.2990822107323584e-05,
"loss": 7.593463897705078,
"step": 1245
},
{
"epoch": 0.05009045226130653,
"grad_norm": 2.2369961738586426,
"learning_rate": 1.2990791193930172e-05,
"loss": 7.249351501464844,
"step": 1246
},
{
"epoch": 0.050130653266331655,
"grad_norm": 2.008836507797241,
"learning_rate": 1.2990760228636171e-05,
"loss": 7.383194923400879,
"step": 1247
},
{
"epoch": 0.050170854271356785,
"grad_norm": 2.20454740524292,
"learning_rate": 1.2990729211442077e-05,
"loss": 7.188348293304443,
"step": 1248
},
{
"epoch": 0.05021105527638191,
"grad_norm": 2.2253541946411133,
"learning_rate": 1.299069814234839e-05,
"loss": 7.426995277404785,
"step": 1249
},
{
"epoch": 0.05025125628140704,
"grad_norm": 2.1218607425689697,
"learning_rate": 1.2990667021355602e-05,
"loss": 7.702469348907471,
"step": 1250
},
{
"epoch": 0.05029145728643216,
"grad_norm": 2.3197848796844482,
"learning_rate": 1.2990635848464216e-05,
"loss": 8.049322128295898,
"step": 1251
},
{
"epoch": 0.05033165829145728,
"grad_norm": 2.7550506591796875,
"learning_rate": 1.299060462367473e-05,
"loss": 8.01058292388916,
"step": 1252
},
{
"epoch": 0.05037185929648241,
"grad_norm": 2.243870258331299,
"learning_rate": 1.2990573346987642e-05,
"loss": 7.403386116027832,
"step": 1253
},
{
"epoch": 0.050412060301507536,
"grad_norm": 2.5552592277526855,
"learning_rate": 1.2990542018403455e-05,
"loss": 7.1480207443237305,
"step": 1254
},
{
"epoch": 0.050452261306532666,
"grad_norm": 2.2253966331481934,
"learning_rate": 1.299051063792267e-05,
"loss": 7.501392841339111,
"step": 1255
},
{
"epoch": 0.05049246231155779,
"grad_norm": 2.329481601715088,
"learning_rate": 1.299047920554579e-05,
"loss": 7.370262145996094,
"step": 1256
},
{
"epoch": 0.05053266331658291,
"grad_norm": 2.2974483966827393,
"learning_rate": 1.2990447721273319e-05,
"loss": 7.936713218688965,
"step": 1257
},
{
"epoch": 0.05057286432160804,
"grad_norm": 2.325937271118164,
"learning_rate": 1.299041618510576e-05,
"loss": 7.189988613128662,
"step": 1258
},
{
"epoch": 0.050613065326633164,
"grad_norm": 2.0649683475494385,
"learning_rate": 1.2990384597043617e-05,
"loss": 7.494262218475342,
"step": 1259
},
{
"epoch": 0.05065326633165829,
"grad_norm": 2.0502309799194336,
"learning_rate": 1.2990352957087398e-05,
"loss": 7.369540214538574,
"step": 1260
},
{
"epoch": 0.050693467336683416,
"grad_norm": 1.837317943572998,
"learning_rate": 1.299032126523761e-05,
"loss": 7.718612194061279,
"step": 1261
},
{
"epoch": 0.050733668341708546,
"grad_norm": 2.082207441329956,
"learning_rate": 1.2990289521494758e-05,
"loss": 7.656329154968262,
"step": 1262
},
{
"epoch": 0.05077386934673367,
"grad_norm": 2.0367839336395264,
"learning_rate": 1.2990257725859354e-05,
"loss": 7.694958686828613,
"step": 1263
},
{
"epoch": 0.05081407035175879,
"grad_norm": 5.241162300109863,
"learning_rate": 1.2990225878331904e-05,
"loss": 7.496969223022461,
"step": 1264
},
{
"epoch": 0.05085427135678392,
"grad_norm": 2.4918410778045654,
"learning_rate": 1.2990193978912917e-05,
"loss": 7.997324466705322,
"step": 1265
},
{
"epoch": 0.050894472361809044,
"grad_norm": 2.194263219833374,
"learning_rate": 1.2990162027602909e-05,
"loss": 7.420065879821777,
"step": 1266
},
{
"epoch": 0.050934673366834174,
"grad_norm": 2.270639657974243,
"learning_rate": 1.2990130024402387e-05,
"loss": 7.471523284912109,
"step": 1267
},
{
"epoch": 0.0509748743718593,
"grad_norm": 2.5203559398651123,
"learning_rate": 1.2990097969311865e-05,
"loss": 7.830254554748535,
"step": 1268
},
{
"epoch": 0.05101507537688442,
"grad_norm": 2.158679485321045,
"learning_rate": 1.2990065862331859e-05,
"loss": 7.8376359939575195,
"step": 1269
},
{
"epoch": 0.05105527638190955,
"grad_norm": 2.0901272296905518,
"learning_rate": 1.2990033703462878e-05,
"loss": 7.462555885314941,
"step": 1270
},
{
"epoch": 0.05109547738693467,
"grad_norm": 2.2258095741271973,
"learning_rate": 1.299000149270544e-05,
"loss": 7.233689785003662,
"step": 1271
},
{
"epoch": 0.0511356783919598,
"grad_norm": 2.189110279083252,
"learning_rate": 1.2989969230060062e-05,
"loss": 7.534789085388184,
"step": 1272
},
{
"epoch": 0.051175879396984925,
"grad_norm": 2.373011589050293,
"learning_rate": 1.298993691552726e-05,
"loss": 7.270139694213867,
"step": 1273
},
{
"epoch": 0.05121608040201005,
"grad_norm": 2.192636728286743,
"learning_rate": 1.2989904549107548e-05,
"loss": 7.540155410766602,
"step": 1274
},
{
"epoch": 0.05125628140703518,
"grad_norm": 2.337886333465576,
"learning_rate": 1.298987213080145e-05,
"loss": 7.407532215118408,
"step": 1275
},
{
"epoch": 0.0512964824120603,
"grad_norm": 2.1688477993011475,
"learning_rate": 1.298983966060948e-05,
"loss": 7.503796100616455,
"step": 1276
},
{
"epoch": 0.05133668341708543,
"grad_norm": 2.165031671524048,
"learning_rate": 1.2989807138532162e-05,
"loss": 7.721506118774414,
"step": 1277
},
{
"epoch": 0.05137688442211055,
"grad_norm": 2.432485818862915,
"learning_rate": 1.2989774564570015e-05,
"loss": 7.737357139587402,
"step": 1278
},
{
"epoch": 0.051417085427135675,
"grad_norm": 2.2315709590911865,
"learning_rate": 1.298974193872356e-05,
"loss": 7.407353401184082,
"step": 1279
},
{
"epoch": 0.051457286432160805,
"grad_norm": 2.1541616916656494,
"learning_rate": 1.2989709260993321e-05,
"loss": 7.367722988128662,
"step": 1280
},
{
"epoch": 0.05149748743718593,
"grad_norm": 2.0326709747314453,
"learning_rate": 1.298967653137982e-05,
"loss": 7.526437282562256,
"step": 1281
},
{
"epoch": 0.05153768844221106,
"grad_norm": 2.2081310749053955,
"learning_rate": 1.2989643749883582e-05,
"loss": 7.786490440368652,
"step": 1282
},
{
"epoch": 0.05157788944723618,
"grad_norm": 2.10103178024292,
"learning_rate": 1.2989610916505132e-05,
"loss": 7.764492988586426,
"step": 1283
},
{
"epoch": 0.0516180904522613,
"grad_norm": 2.3564655780792236,
"learning_rate": 1.2989578031244994e-05,
"loss": 7.844760894775391,
"step": 1284
},
{
"epoch": 0.05165829145728643,
"grad_norm": 2.1563222408294678,
"learning_rate": 1.2989545094103698e-05,
"loss": 7.340296745300293,
"step": 1285
},
{
"epoch": 0.051698492462311556,
"grad_norm": 2.3815011978149414,
"learning_rate": 1.298951210508177e-05,
"loss": 7.496156692504883,
"step": 1286
},
{
"epoch": 0.051738693467336686,
"grad_norm": 2.141052007675171,
"learning_rate": 1.2989479064179736e-05,
"loss": 7.452962875366211,
"step": 1287
},
{
"epoch": 0.05177889447236181,
"grad_norm": 2.6498734951019287,
"learning_rate": 1.298944597139813e-05,
"loss": 7.196407318115234,
"step": 1288
},
{
"epoch": 0.05181909547738694,
"grad_norm": 2.002150535583496,
"learning_rate": 1.2989412826737478e-05,
"loss": 7.653995513916016,
"step": 1289
},
{
"epoch": 0.05185929648241206,
"grad_norm": 2.5846805572509766,
"learning_rate": 1.2989379630198312e-05,
"loss": 7.617913246154785,
"step": 1290
},
{
"epoch": 0.051899497487437184,
"grad_norm": 2.8162078857421875,
"learning_rate": 1.2989346381781163e-05,
"loss": 7.7034912109375,
"step": 1291
},
{
"epoch": 0.05193969849246231,
"grad_norm": 2.241122007369995,
"learning_rate": 1.2989313081486565e-05,
"loss": 7.54061222076416,
"step": 1292
},
{
"epoch": 0.051979899497487436,
"grad_norm": 2.2428696155548096,
"learning_rate": 1.298927972931505e-05,
"loss": 7.574847221374512,
"step": 1293
},
{
"epoch": 0.052020100502512566,
"grad_norm": 2.4573941230773926,
"learning_rate": 1.2989246325267152e-05,
"loss": 7.606245994567871,
"step": 1294
},
{
"epoch": 0.05206030150753769,
"grad_norm": 2.436858892440796,
"learning_rate": 1.2989212869343406e-05,
"loss": 7.516916275024414,
"step": 1295
},
{
"epoch": 0.05210050251256281,
"grad_norm": 2.398563861846924,
"learning_rate": 1.298917936154435e-05,
"loss": 7.3526692390441895,
"step": 1296
},
{
"epoch": 0.05214070351758794,
"grad_norm": 2.8482167720794678,
"learning_rate": 1.2989145801870516e-05,
"loss": 7.837251663208008,
"step": 1297
},
{
"epoch": 0.052180904522613064,
"grad_norm": 2.3510406017303467,
"learning_rate": 1.2989112190322447e-05,
"loss": 7.215693473815918,
"step": 1298
},
{
"epoch": 0.052221105527638194,
"grad_norm": 2.164722204208374,
"learning_rate": 1.2989078526900677e-05,
"loss": 7.1151628494262695,
"step": 1299
},
{
"epoch": 0.05226130653266332,
"grad_norm": 2.260075092315674,
"learning_rate": 1.298904481160575e-05,
"loss": 7.08839750289917,
"step": 1300
},
{
"epoch": 0.05230150753768844,
"grad_norm": 2.163996458053589,
"learning_rate": 1.2989011044438198e-05,
"loss": 7.823921203613281,
"step": 1301
},
{
"epoch": 0.05234170854271357,
"grad_norm": 2.449986219406128,
"learning_rate": 1.298897722539857e-05,
"loss": 7.832885265350342,
"step": 1302
},
{
"epoch": 0.05238190954773869,
"grad_norm": 2.0582380294799805,
"learning_rate": 1.2988943354487403e-05,
"loss": 7.561380386352539,
"step": 1303
},
{
"epoch": 0.05242211055276382,
"grad_norm": 2.280843734741211,
"learning_rate": 1.2988909431705239e-05,
"loss": 7.9971795082092285,
"step": 1304
},
{
"epoch": 0.052462311557788945,
"grad_norm": 2.309479236602783,
"learning_rate": 1.2988875457052624e-05,
"loss": 7.797989845275879,
"step": 1305
},
{
"epoch": 0.05250251256281407,
"grad_norm": 2.3616981506347656,
"learning_rate": 1.2988841430530102e-05,
"loss": 8.286142349243164,
"step": 1306
},
{
"epoch": 0.0525427135678392,
"grad_norm": 2.2081098556518555,
"learning_rate": 1.2988807352138216e-05,
"loss": 8.380636215209961,
"step": 1307
},
{
"epoch": 0.05258291457286432,
"grad_norm": 2.0859479904174805,
"learning_rate": 1.2988773221877512e-05,
"loss": 7.269010543823242,
"step": 1308
},
{
"epoch": 0.05262311557788945,
"grad_norm": 2.2388367652893066,
"learning_rate": 1.2988739039748535e-05,
"loss": 7.411659240722656,
"step": 1309
},
{
"epoch": 0.05266331658291457,
"grad_norm": 2.137383222579956,
"learning_rate": 1.2988704805751837e-05,
"loss": 7.922783851623535,
"step": 1310
},
{
"epoch": 0.052703517587939695,
"grad_norm": 2.098726987838745,
"learning_rate": 1.2988670519887963e-05,
"loss": 7.144132614135742,
"step": 1311
},
{
"epoch": 0.052743718592964825,
"grad_norm": 2.361758232116699,
"learning_rate": 1.2988636182157464e-05,
"loss": 7.560371398925781,
"step": 1312
},
{
"epoch": 0.05278391959798995,
"grad_norm": 2.314378261566162,
"learning_rate": 1.2988601792560888e-05,
"loss": 7.743382453918457,
"step": 1313
},
{
"epoch": 0.05282412060301508,
"grad_norm": 2.2548303604125977,
"learning_rate": 1.2988567351098786e-05,
"loss": 7.468122482299805,
"step": 1314
},
{
"epoch": 0.0528643216080402,
"grad_norm": 2.266828775405884,
"learning_rate": 1.2988532857771711e-05,
"loss": 7.616646766662598,
"step": 1315
},
{
"epoch": 0.05290452261306532,
"grad_norm": 2.0689244270324707,
"learning_rate": 1.2988498312580215e-05,
"loss": 7.633671760559082,
"step": 1316
},
{
"epoch": 0.05294472361809045,
"grad_norm": 2.160674810409546,
"learning_rate": 1.298846371552485e-05,
"loss": 7.387034893035889,
"step": 1317
},
{
"epoch": 0.052984924623115576,
"grad_norm": 2.184025526046753,
"learning_rate": 1.298842906660617e-05,
"loss": 6.948877334594727,
"step": 1318
},
{
"epoch": 0.053025125628140705,
"grad_norm": 2.469987392425537,
"learning_rate": 1.2988394365824732e-05,
"loss": 7.536799430847168,
"step": 1319
},
{
"epoch": 0.05306532663316583,
"grad_norm": 2.35109543800354,
"learning_rate": 1.298835961318109e-05,
"loss": 7.186924934387207,
"step": 1320
},
{
"epoch": 0.05310552763819096,
"grad_norm": 2.2904930114746094,
"learning_rate": 1.29883248086758e-05,
"loss": 6.917025566101074,
"step": 1321
},
{
"epoch": 0.05314572864321608,
"grad_norm": 1.9714360237121582,
"learning_rate": 1.2988289952309421e-05,
"loss": 7.499266624450684,
"step": 1322
},
{
"epoch": 0.053185929648241204,
"grad_norm": 2.7350094318389893,
"learning_rate": 1.2988255044082511e-05,
"loss": 7.394932746887207,
"step": 1323
},
{
"epoch": 0.05322613065326633,
"grad_norm": 2.1031136512756348,
"learning_rate": 1.2988220083995628e-05,
"loss": 7.88824987411499,
"step": 1324
},
{
"epoch": 0.053266331658291456,
"grad_norm": 2.125199794769287,
"learning_rate": 1.2988185072049333e-05,
"loss": 7.809238433837891,
"step": 1325
},
{
"epoch": 0.053306532663316586,
"grad_norm": 3.0453646183013916,
"learning_rate": 1.2988150008244186e-05,
"loss": 7.461485862731934,
"step": 1326
},
{
"epoch": 0.05334673366834171,
"grad_norm": 2.2961246967315674,
"learning_rate": 1.2988114892580748e-05,
"loss": 7.008253574371338,
"step": 1327
},
{
"epoch": 0.05338693467336683,
"grad_norm": 2.113492727279663,
"learning_rate": 1.2988079725059584e-05,
"loss": 7.707442760467529,
"step": 1328
},
{
"epoch": 0.05342713567839196,
"grad_norm": 2.204359292984009,
"learning_rate": 1.2988044505681253e-05,
"loss": 7.617732048034668,
"step": 1329
},
{
"epoch": 0.053467336683417084,
"grad_norm": 2.4468958377838135,
"learning_rate": 1.2988009234446322e-05,
"loss": 7.688858985900879,
"step": 1330
},
{
"epoch": 0.053507537688442214,
"grad_norm": 2.1768758296966553,
"learning_rate": 1.2987973911355356e-05,
"loss": 7.80330228805542,
"step": 1331
},
{
"epoch": 0.05354773869346734,
"grad_norm": 2.2081716060638428,
"learning_rate": 1.298793853640892e-05,
"loss": 7.514763832092285,
"step": 1332
},
{
"epoch": 0.05358793969849246,
"grad_norm": 3.1837501525878906,
"learning_rate": 1.2987903109607581e-05,
"loss": 6.966201305389404,
"step": 1333
},
{
"epoch": 0.05362814070351759,
"grad_norm": 2.829800844192505,
"learning_rate": 1.2987867630951903e-05,
"loss": 7.407543182373047,
"step": 1334
},
{
"epoch": 0.05366834170854271,
"grad_norm": 2.4652459621429443,
"learning_rate": 1.298783210044246e-05,
"loss": 7.165112495422363,
"step": 1335
},
{
"epoch": 0.05370854271356784,
"grad_norm": 2.32846999168396,
"learning_rate": 1.2987796518079817e-05,
"loss": 7.9645280838012695,
"step": 1336
},
{
"epoch": 0.053748743718592965,
"grad_norm": 2.3063061237335205,
"learning_rate": 1.2987760883864544e-05,
"loss": 7.217632293701172,
"step": 1337
},
{
"epoch": 0.05378894472361809,
"grad_norm": 2.2674763202667236,
"learning_rate": 1.2987725197797214e-05,
"loss": 7.652887344360352,
"step": 1338
},
{
"epoch": 0.05382914572864322,
"grad_norm": 2.1925392150878906,
"learning_rate": 1.2987689459878395e-05,
"loss": 7.565760612487793,
"step": 1339
},
{
"epoch": 0.05386934673366834,
"grad_norm": 2.200981616973877,
"learning_rate": 1.2987653670108662e-05,
"loss": 7.472330093383789,
"step": 1340
},
{
"epoch": 0.05390954773869347,
"grad_norm": 2.2626147270202637,
"learning_rate": 1.2987617828488588e-05,
"loss": 7.028752326965332,
"step": 1341
},
{
"epoch": 0.05394974874371859,
"grad_norm": 2.2326529026031494,
"learning_rate": 1.2987581935018746e-05,
"loss": 7.56377649307251,
"step": 1342
},
{
"epoch": 0.053989949748743715,
"grad_norm": 2.1944737434387207,
"learning_rate": 1.2987545989699711e-05,
"loss": 7.60402774810791,
"step": 1343
},
{
"epoch": 0.054030150753768845,
"grad_norm": 2.1806142330169678,
"learning_rate": 1.298750999253206e-05,
"loss": 7.99901008605957,
"step": 1344
},
{
"epoch": 0.05407035175879397,
"grad_norm": 2.508171558380127,
"learning_rate": 1.2987473943516368e-05,
"loss": 7.6648850440979,
"step": 1345
},
{
"epoch": 0.0541105527638191,
"grad_norm": 2.560905694961548,
"learning_rate": 1.2987437842653214e-05,
"loss": 7.909526824951172,
"step": 1346
},
{
"epoch": 0.05415075376884422,
"grad_norm": 2.2858242988586426,
"learning_rate": 1.2987401689943174e-05,
"loss": 7.486088752746582,
"step": 1347
},
{
"epoch": 0.05419095477386935,
"grad_norm": 2.3054094314575195,
"learning_rate": 1.2987365485386826e-05,
"loss": 7.96466064453125,
"step": 1348
},
{
"epoch": 0.05423115577889447,
"grad_norm": 2.548945665359497,
"learning_rate": 1.2987329228984755e-05,
"loss": 7.246582984924316,
"step": 1349
},
{
"epoch": 0.054271356783919596,
"grad_norm": 2.0247855186462402,
"learning_rate": 1.2987292920737536e-05,
"loss": 7.233891487121582,
"step": 1350
},
{
"epoch": 0.054311557788944725,
"grad_norm": 2.103304147720337,
"learning_rate": 1.2987256560645754e-05,
"loss": 7.311138153076172,
"step": 1351
},
{
"epoch": 0.05435175879396985,
"grad_norm": 2.6508331298828125,
"learning_rate": 1.2987220148709988e-05,
"loss": 7.631292343139648,
"step": 1352
},
{
"epoch": 0.05439195979899498,
"grad_norm": 2.0160820484161377,
"learning_rate": 1.2987183684930826e-05,
"loss": 7.784023284912109,
"step": 1353
},
{
"epoch": 0.0544321608040201,
"grad_norm": 2.3545844554901123,
"learning_rate": 1.2987147169308849e-05,
"loss": 7.738528251647949,
"step": 1354
},
{
"epoch": 0.054472361809045224,
"grad_norm": 2.428117513656616,
"learning_rate": 1.298711060184464e-05,
"loss": 7.56959342956543,
"step": 1355
},
{
"epoch": 0.05451256281407035,
"grad_norm": 2.730579376220703,
"learning_rate": 1.2987073982538787e-05,
"loss": 7.505513668060303,
"step": 1356
},
{
"epoch": 0.054552763819095476,
"grad_norm": 2.2718570232391357,
"learning_rate": 1.2987037311391879e-05,
"loss": 7.604499340057373,
"step": 1357
},
{
"epoch": 0.054592964824120606,
"grad_norm": 2.144901752471924,
"learning_rate": 1.2987000588404498e-05,
"loss": 7.417699813842773,
"step": 1358
},
{
"epoch": 0.05463316582914573,
"grad_norm": 2.111880302429199,
"learning_rate": 1.2986963813577235e-05,
"loss": 7.732828617095947,
"step": 1359
},
{
"epoch": 0.05467336683417085,
"grad_norm": 2.2254691123962402,
"learning_rate": 1.2986926986910678e-05,
"loss": 7.68130350112915,
"step": 1360
},
{
"epoch": 0.05471356783919598,
"grad_norm": 2.260981559753418,
"learning_rate": 1.2986890108405416e-05,
"loss": 7.875391006469727,
"step": 1361
},
{
"epoch": 0.054753768844221104,
"grad_norm": 2.0046262741088867,
"learning_rate": 1.2986853178062042e-05,
"loss": 7.20802116394043,
"step": 1362
},
{
"epoch": 0.054793969849246234,
"grad_norm": 2.2670180797576904,
"learning_rate": 1.2986816195881146e-05,
"loss": 7.510952472686768,
"step": 1363
},
{
"epoch": 0.05483417085427136,
"grad_norm": 2.248427152633667,
"learning_rate": 1.298677916186332e-05,
"loss": 7.3506951332092285,
"step": 1364
},
{
"epoch": 0.05487437185929648,
"grad_norm": 2.201900005340576,
"learning_rate": 1.298674207600916e-05,
"loss": 7.554318904876709,
"step": 1365
},
{
"epoch": 0.05491457286432161,
"grad_norm": 2.320722818374634,
"learning_rate": 1.2986704938319255e-05,
"loss": 7.43356990814209,
"step": 1366
},
{
"epoch": 0.05495477386934673,
"grad_norm": 2.442939281463623,
"learning_rate": 1.2986667748794202e-05,
"loss": 8.099188804626465,
"step": 1367
},
{
"epoch": 0.05499497487437186,
"grad_norm": 2.2439122200012207,
"learning_rate": 1.29866305074346e-05,
"loss": 7.437312602996826,
"step": 1368
},
{
"epoch": 0.055035175879396984,
"grad_norm": 2.21138858795166,
"learning_rate": 1.2986593214241039e-05,
"loss": 7.252857208251953,
"step": 1369
},
{
"epoch": 0.05507537688442211,
"grad_norm": 2.2256925106048584,
"learning_rate": 1.298655586921412e-05,
"loss": 7.718251705169678,
"step": 1370
},
{
"epoch": 0.05511557788944724,
"grad_norm": 2.9009249210357666,
"learning_rate": 1.2986518472354441e-05,
"loss": 6.8455657958984375,
"step": 1371
},
{
"epoch": 0.05515577889447236,
"grad_norm": 2.1663007736206055,
"learning_rate": 1.2986481023662602e-05,
"loss": 7.775684356689453,
"step": 1372
},
{
"epoch": 0.05519597989949749,
"grad_norm": 2.5038743019104004,
"learning_rate": 1.29864435231392e-05,
"loss": 7.379340171813965,
"step": 1373
},
{
"epoch": 0.05523618090452261,
"grad_norm": 2.292268753051758,
"learning_rate": 1.2986405970784834e-05,
"loss": 7.371304512023926,
"step": 1374
},
{
"epoch": 0.05527638190954774,
"grad_norm": 2.1098501682281494,
"learning_rate": 1.298636836660011e-05,
"loss": 7.617642402648926,
"step": 1375
},
{
"epoch": 0.055316582914572865,
"grad_norm": 2.2212443351745605,
"learning_rate": 1.298633071058563e-05,
"loss": 7.903842926025391,
"step": 1376
},
{
"epoch": 0.05535678391959799,
"grad_norm": 2.3072116374969482,
"learning_rate": 1.2986293002741994e-05,
"loss": 7.230801105499268,
"step": 1377
},
{
"epoch": 0.05539698492462312,
"grad_norm": 2.3106367588043213,
"learning_rate": 1.2986255243069807e-05,
"loss": 7.512908935546875,
"step": 1378
},
{
"epoch": 0.05543718592964824,
"grad_norm": 2.181246280670166,
"learning_rate": 1.2986217431569676e-05,
"loss": 7.834341049194336,
"step": 1379
},
{
"epoch": 0.05547738693467337,
"grad_norm": 2.184584617614746,
"learning_rate": 1.2986179568242202e-05,
"loss": 7.187035083770752,
"step": 1380
},
{
"epoch": 0.05551758793969849,
"grad_norm": 2.3432815074920654,
"learning_rate": 1.2986141653087996e-05,
"loss": 7.397177696228027,
"step": 1381
},
{
"epoch": 0.055557788944723616,
"grad_norm": 2.391366481781006,
"learning_rate": 1.2986103686107662e-05,
"loss": 7.134360313415527,
"step": 1382
},
{
"epoch": 0.055597989949748745,
"grad_norm": 2.11527681350708,
"learning_rate": 1.2986065667301809e-05,
"loss": 7.583869934082031,
"step": 1383
},
{
"epoch": 0.05563819095477387,
"grad_norm": 2.322723388671875,
"learning_rate": 1.2986027596671048e-05,
"loss": 7.553609848022461,
"step": 1384
},
{
"epoch": 0.055678391959799,
"grad_norm": 2.465080499649048,
"learning_rate": 1.2985989474215984e-05,
"loss": 7.642786026000977,
"step": 1385
},
{
"epoch": 0.05571859296482412,
"grad_norm": 2.311352491378784,
"learning_rate": 1.2985951299937231e-05,
"loss": 7.742933750152588,
"step": 1386
},
{
"epoch": 0.055758793969849244,
"grad_norm": 2.470242500305176,
"learning_rate": 1.29859130738354e-05,
"loss": 7.117332458496094,
"step": 1387
},
{
"epoch": 0.05579899497487437,
"grad_norm": 2.3264448642730713,
"learning_rate": 1.2985874795911103e-05,
"loss": 7.547318458557129,
"step": 1388
},
{
"epoch": 0.055839195979899496,
"grad_norm": 2.7232789993286133,
"learning_rate": 1.2985836466164952e-05,
"loss": 7.522805213928223,
"step": 1389
},
{
"epoch": 0.055879396984924626,
"grad_norm": 2.1765589714050293,
"learning_rate": 1.2985798084597563e-05,
"loss": 7.580913543701172,
"step": 1390
},
{
"epoch": 0.05591959798994975,
"grad_norm": 2.3718652725219727,
"learning_rate": 1.2985759651209548e-05,
"loss": 7.6727094650268555,
"step": 1391
},
{
"epoch": 0.05595979899497487,
"grad_norm": 2.332777500152588,
"learning_rate": 1.2985721166001525e-05,
"loss": 7.778711795806885,
"step": 1392
},
{
"epoch": 0.056,
"grad_norm": 2.304123878479004,
"learning_rate": 1.2985682628974107e-05,
"loss": 7.600979804992676,
"step": 1393
},
{
"epoch": 0.056040201005025124,
"grad_norm": 2.2911832332611084,
"learning_rate": 1.2985644040127915e-05,
"loss": 6.9720916748046875,
"step": 1394
},
{
"epoch": 0.056080402010050254,
"grad_norm": 2.2511579990386963,
"learning_rate": 1.2985605399463567e-05,
"loss": 7.655207633972168,
"step": 1395
},
{
"epoch": 0.056120603015075377,
"grad_norm": 2.237773895263672,
"learning_rate": 1.2985566706981677e-05,
"loss": 7.5820159912109375,
"step": 1396
},
{
"epoch": 0.0561608040201005,
"grad_norm": 2.2284226417541504,
"learning_rate": 1.2985527962682869e-05,
"loss": 7.571452617645264,
"step": 1397
},
{
"epoch": 0.05620100502512563,
"grad_norm": 2.379570722579956,
"learning_rate": 1.2985489166567761e-05,
"loss": 7.514214992523193,
"step": 1398
},
{
"epoch": 0.05624120603015075,
"grad_norm": 2.4865384101867676,
"learning_rate": 1.2985450318636977e-05,
"loss": 7.137316703796387,
"step": 1399
},
{
"epoch": 0.05628140703517588,
"grad_norm": 2.2195041179656982,
"learning_rate": 1.2985411418891138e-05,
"loss": 7.4011688232421875,
"step": 1400
},
{
"epoch": 0.05628140703517588,
"eval_loss": 7.55483865737915,
"eval_runtime": 120.6408,
"eval_samples_per_second": 4.145,
"eval_steps_per_second": 0.522,
"step": 1400
},
{
"epoch": 0.056321608040201004,
"grad_norm": 2.532069444656372,
"learning_rate": 1.2985372467330867e-05,
"loss": 7.958096504211426,
"step": 1401
},
{
"epoch": 0.05636180904522613,
"grad_norm": 2.336858034133911,
"learning_rate": 1.2985333463956785e-05,
"loss": 7.530320167541504,
"step": 1402
},
{
"epoch": 0.05640201005025126,
"grad_norm": 2.140817642211914,
"learning_rate": 1.2985294408769521e-05,
"loss": 7.7965288162231445,
"step": 1403
},
{
"epoch": 0.05644221105527638,
"grad_norm": 2.149709939956665,
"learning_rate": 1.2985255301769698e-05,
"loss": 7.51969575881958,
"step": 1404
},
{
"epoch": 0.05648241206030151,
"grad_norm": 2.233912944793701,
"learning_rate": 1.2985216142957943e-05,
"loss": 7.609489440917969,
"step": 1405
},
{
"epoch": 0.05652261306532663,
"grad_norm": 2.3530735969543457,
"learning_rate": 1.2985176932334884e-05,
"loss": 7.468130111694336,
"step": 1406
},
{
"epoch": 0.05656281407035176,
"grad_norm": 2.432640552520752,
"learning_rate": 1.2985137669901146e-05,
"loss": 7.594150543212891,
"step": 1407
},
{
"epoch": 0.056603015075376885,
"grad_norm": 2.9529001712799072,
"learning_rate": 1.298509835565736e-05,
"loss": 7.410868167877197,
"step": 1408
},
{
"epoch": 0.05664321608040201,
"grad_norm": 2.2046844959259033,
"learning_rate": 1.2985058989604158e-05,
"loss": 7.269481658935547,
"step": 1409
},
{
"epoch": 0.05668341708542714,
"grad_norm": 2.5368854999542236,
"learning_rate": 1.2985019571742166e-05,
"loss": 7.5304975509643555,
"step": 1410
},
{
"epoch": 0.05672361809045226,
"grad_norm": 2.362607479095459,
"learning_rate": 1.2984980102072016e-05,
"loss": 7.804045677185059,
"step": 1411
},
{
"epoch": 0.05676381909547739,
"grad_norm": 2.2890727519989014,
"learning_rate": 1.2984940580594342e-05,
"loss": 7.398957252502441,
"step": 1412
},
{
"epoch": 0.05680402010050251,
"grad_norm": 2.8149209022521973,
"learning_rate": 1.2984901007309777e-05,
"loss": 7.600966930389404,
"step": 1413
},
{
"epoch": 0.056844221105527636,
"grad_norm": 2.692289113998413,
"learning_rate": 1.2984861382218952e-05,
"loss": 7.0358734130859375,
"step": 1414
},
{
"epoch": 0.056884422110552765,
"grad_norm": 2.5020084381103516,
"learning_rate": 1.2984821705322504e-05,
"loss": 7.538888931274414,
"step": 1415
},
{
"epoch": 0.05692462311557789,
"grad_norm": 2.314793586730957,
"learning_rate": 1.2984781976621068e-05,
"loss": 7.7216901779174805,
"step": 1416
},
{
"epoch": 0.05696482412060302,
"grad_norm": 1.9838041067123413,
"learning_rate": 1.298474219611528e-05,
"loss": 7.029448986053467,
"step": 1417
},
{
"epoch": 0.05700502512562814,
"grad_norm": 4.259639739990234,
"learning_rate": 1.2984702363805779e-05,
"loss": 8.684915542602539,
"step": 1418
},
{
"epoch": 0.057045226130653263,
"grad_norm": 2.4870471954345703,
"learning_rate": 1.2984662479693199e-05,
"loss": 7.419063091278076,
"step": 1419
},
{
"epoch": 0.05708542713567839,
"grad_norm": 2.6588218212127686,
"learning_rate": 1.2984622543778179e-05,
"loss": 7.223654270172119,
"step": 1420
},
{
"epoch": 0.057125628140703516,
"grad_norm": 2.5152549743652344,
"learning_rate": 1.2984582556061363e-05,
"loss": 7.9094414710998535,
"step": 1421
},
{
"epoch": 0.057165829145728646,
"grad_norm": 2.1681339740753174,
"learning_rate": 1.2984542516543387e-05,
"loss": 7.780712127685547,
"step": 1422
},
{
"epoch": 0.05720603015075377,
"grad_norm": 3.2126564979553223,
"learning_rate": 1.2984502425224896e-05,
"loss": 7.06705379486084,
"step": 1423
},
{
"epoch": 0.05724623115577889,
"grad_norm": 2.2736546993255615,
"learning_rate": 1.298446228210653e-05,
"loss": 7.649076461791992,
"step": 1424
},
{
"epoch": 0.05728643216080402,
"grad_norm": 2.236335277557373,
"learning_rate": 1.298442208718893e-05,
"loss": 7.64338493347168,
"step": 1425
},
{
"epoch": 0.057326633165829144,
"grad_norm": 2.551574945449829,
"learning_rate": 1.2984381840472743e-05,
"loss": 8.170822143554688,
"step": 1426
},
{
"epoch": 0.057366834170854274,
"grad_norm": 2.6265242099761963,
"learning_rate": 1.2984341541958611e-05,
"loss": 7.016096115112305,
"step": 1427
},
{
"epoch": 0.057407035175879396,
"grad_norm": 2.7880020141601562,
"learning_rate": 1.2984301191647182e-05,
"loss": 7.710453033447266,
"step": 1428
},
{
"epoch": 0.05744723618090452,
"grad_norm": 2.8497812747955322,
"learning_rate": 1.29842607895391e-05,
"loss": 6.981975555419922,
"step": 1429
},
{
"epoch": 0.05748743718592965,
"grad_norm": 2.4203972816467285,
"learning_rate": 1.2984220335635015e-05,
"loss": 7.939958095550537,
"step": 1430
},
{
"epoch": 0.05752763819095477,
"grad_norm": 3.1825928688049316,
"learning_rate": 1.2984179829935568e-05,
"loss": 7.422170639038086,
"step": 1431
},
{
"epoch": 0.0575678391959799,
"grad_norm": 2.645944118499756,
"learning_rate": 1.2984139272441415e-05,
"loss": 7.319263935089111,
"step": 1432
},
{
"epoch": 0.057608040201005024,
"grad_norm": 2.204214572906494,
"learning_rate": 1.2984098663153202e-05,
"loss": 7.8223490715026855,
"step": 1433
},
{
"epoch": 0.057648241206030154,
"grad_norm": 2.3533191680908203,
"learning_rate": 1.298405800207158e-05,
"loss": 7.144150733947754,
"step": 1434
},
{
"epoch": 0.05768844221105528,
"grad_norm": 2.2154738903045654,
"learning_rate": 1.2984017289197202e-05,
"loss": 7.236186504364014,
"step": 1435
},
{
"epoch": 0.0577286432160804,
"grad_norm": 2.169360876083374,
"learning_rate": 1.2983976524530718e-05,
"loss": 7.0722198486328125,
"step": 1436
},
{
"epoch": 0.05776884422110553,
"grad_norm": 2.3965182304382324,
"learning_rate": 1.2983935708072781e-05,
"loss": 8.20675277709961,
"step": 1437
},
{
"epoch": 0.05780904522613065,
"grad_norm": 2.3918943405151367,
"learning_rate": 1.2983894839824044e-05,
"loss": 7.323707580566406,
"step": 1438
},
{
"epoch": 0.05784924623115578,
"grad_norm": 2.3995912075042725,
"learning_rate": 1.2983853919785164e-05,
"loss": 7.27131462097168,
"step": 1439
},
{
"epoch": 0.057889447236180905,
"grad_norm": 2.5959994792938232,
"learning_rate": 1.2983812947956793e-05,
"loss": 7.658944129943848,
"step": 1440
},
{
"epoch": 0.05792964824120603,
"grad_norm": 2.5566513538360596,
"learning_rate": 1.298377192433959e-05,
"loss": 7.16926383972168,
"step": 1441
},
{
"epoch": 0.05796984924623116,
"grad_norm": 2.6423707008361816,
"learning_rate": 1.298373084893421e-05,
"loss": 7.425867557525635,
"step": 1442
},
{
"epoch": 0.05801005025125628,
"grad_norm": 2.6712660789489746,
"learning_rate": 1.2983689721741313e-05,
"loss": 7.479292869567871,
"step": 1443
},
{
"epoch": 0.05805025125628141,
"grad_norm": 2.299239158630371,
"learning_rate": 1.2983648542761554e-05,
"loss": 7.434947967529297,
"step": 1444
},
{
"epoch": 0.05809045226130653,
"grad_norm": 2.276516914367676,
"learning_rate": 1.2983607311995595e-05,
"loss": 7.492391586303711,
"step": 1445
},
{
"epoch": 0.058130653266331656,
"grad_norm": 2.4211480617523193,
"learning_rate": 1.2983566029444098e-05,
"loss": 7.12080192565918,
"step": 1446
},
{
"epoch": 0.058170854271356785,
"grad_norm": 2.447748899459839,
"learning_rate": 1.2983524695107723e-05,
"loss": 7.346972465515137,
"step": 1447
},
{
"epoch": 0.05821105527638191,
"grad_norm": 2.507122755050659,
"learning_rate": 1.2983483308987128e-05,
"loss": 7.772099494934082,
"step": 1448
},
{
"epoch": 0.05825125628140704,
"grad_norm": 2.410442352294922,
"learning_rate": 1.2983441871082981e-05,
"loss": 7.532876968383789,
"step": 1449
},
{
"epoch": 0.05829145728643216,
"grad_norm": 2.258051633834839,
"learning_rate": 1.2983400381395943e-05,
"loss": 7.405734062194824,
"step": 1450
},
{
"epoch": 0.05833165829145728,
"grad_norm": 2.184138298034668,
"learning_rate": 1.2983358839926678e-05,
"loss": 7.438032150268555,
"step": 1451
},
{
"epoch": 0.05837185929648241,
"grad_norm": 2.174537420272827,
"learning_rate": 1.2983317246675853e-05,
"loss": 7.468602180480957,
"step": 1452
},
{
"epoch": 0.058412060301507536,
"grad_norm": 2.2387218475341797,
"learning_rate": 1.2983275601644134e-05,
"loss": 7.462638854980469,
"step": 1453
},
{
"epoch": 0.058452261306532666,
"grad_norm": 2.5469260215759277,
"learning_rate": 1.2983233904832188e-05,
"loss": 7.375616073608398,
"step": 1454
},
{
"epoch": 0.05849246231155779,
"grad_norm": 2.3574841022491455,
"learning_rate": 1.2983192156240681e-05,
"loss": 7.753374099731445,
"step": 1455
},
{
"epoch": 0.05853266331658291,
"grad_norm": 2.1780107021331787,
"learning_rate": 1.2983150355870282e-05,
"loss": 7.450225830078125,
"step": 1456
},
{
"epoch": 0.05857286432160804,
"grad_norm": 2.0561912059783936,
"learning_rate": 1.2983108503721664e-05,
"loss": 7.532815933227539,
"step": 1457
},
{
"epoch": 0.058613065326633164,
"grad_norm": 2.0813255310058594,
"learning_rate": 1.2983066599795492e-05,
"loss": 7.634442329406738,
"step": 1458
},
{
"epoch": 0.058653266331658294,
"grad_norm": 2.450502634048462,
"learning_rate": 1.2983024644092438e-05,
"loss": 7.620736122131348,
"step": 1459
},
{
"epoch": 0.058693467336683416,
"grad_norm": 2.319732904434204,
"learning_rate": 1.2982982636613178e-05,
"loss": 7.790851593017578,
"step": 1460
},
{
"epoch": 0.058733668341708546,
"grad_norm": 2.2236647605895996,
"learning_rate": 1.298294057735838e-05,
"loss": 7.949902534484863,
"step": 1461
},
{
"epoch": 0.05877386934673367,
"grad_norm": 2.1088762283325195,
"learning_rate": 1.2982898466328722e-05,
"loss": 7.6781511306762695,
"step": 1462
},
{
"epoch": 0.05881407035175879,
"grad_norm": 2.380535840988159,
"learning_rate": 1.2982856303524875e-05,
"loss": 7.121382713317871,
"step": 1463
},
{
"epoch": 0.05885427135678392,
"grad_norm": 2.7306759357452393,
"learning_rate": 1.2982814088947515e-05,
"loss": 7.98640251159668,
"step": 1464
},
{
"epoch": 0.058894472361809044,
"grad_norm": 2.3490922451019287,
"learning_rate": 1.2982771822597318e-05,
"loss": 7.7001495361328125,
"step": 1465
},
{
"epoch": 0.058934673366834174,
"grad_norm": 2.4395179748535156,
"learning_rate": 1.2982729504474964e-05,
"loss": 7.091691017150879,
"step": 1466
},
{
"epoch": 0.0589748743718593,
"grad_norm": 2.5003623962402344,
"learning_rate": 1.2982687134581124e-05,
"loss": 7.956220626831055,
"step": 1467
},
{
"epoch": 0.05901507537688442,
"grad_norm": 2.4245054721832275,
"learning_rate": 1.2982644712916483e-05,
"loss": 8.082779884338379,
"step": 1468
},
{
"epoch": 0.05905527638190955,
"grad_norm": 2.307908535003662,
"learning_rate": 1.2982602239481717e-05,
"loss": 7.752352714538574,
"step": 1469
},
{
"epoch": 0.05909547738693467,
"grad_norm": 2.2921297550201416,
"learning_rate": 1.2982559714277509e-05,
"loss": 7.793774604797363,
"step": 1470
},
{
"epoch": 0.0591356783919598,
"grad_norm": 2.1724092960357666,
"learning_rate": 1.2982517137304535e-05,
"loss": 7.49520206451416,
"step": 1471
},
{
"epoch": 0.059175879396984925,
"grad_norm": 2.0687801837921143,
"learning_rate": 1.2982474508563483e-05,
"loss": 7.147452354431152,
"step": 1472
},
{
"epoch": 0.05921608040201005,
"grad_norm": 2.222205638885498,
"learning_rate": 1.2982431828055031e-05,
"loss": 7.229086875915527,
"step": 1473
},
{
"epoch": 0.05925628140703518,
"grad_norm": 2.37815260887146,
"learning_rate": 1.2982389095779864e-05,
"loss": 7.647936820983887,
"step": 1474
},
{
"epoch": 0.0592964824120603,
"grad_norm": 2.179823160171509,
"learning_rate": 1.2982346311738667e-05,
"loss": 7.656035900115967,
"step": 1475
},
{
"epoch": 0.05933668341708543,
"grad_norm": 2.7022697925567627,
"learning_rate": 1.2982303475932124e-05,
"loss": 6.898054122924805,
"step": 1476
},
{
"epoch": 0.05937688442211055,
"grad_norm": 2.420623779296875,
"learning_rate": 1.2982260588360923e-05,
"loss": 7.6148681640625,
"step": 1477
},
{
"epoch": 0.059417085427135675,
"grad_norm": 2.178345203399658,
"learning_rate": 1.2982217649025749e-05,
"loss": 8.076355934143066,
"step": 1478
},
{
"epoch": 0.059457286432160805,
"grad_norm": 2.3533759117126465,
"learning_rate": 1.298217465792729e-05,
"loss": 7.427577495574951,
"step": 1479
},
{
"epoch": 0.05949748743718593,
"grad_norm": 3.524292469024658,
"learning_rate": 1.2982131615066236e-05,
"loss": 7.49344539642334,
"step": 1480
},
{
"epoch": 0.05953768844221106,
"grad_norm": 2.541224479675293,
"learning_rate": 1.2982088520443273e-05,
"loss": 7.748655319213867,
"step": 1481
},
{
"epoch": 0.05957788944723618,
"grad_norm": 3.0138237476348877,
"learning_rate": 1.2982045374059094e-05,
"loss": 7.321172714233398,
"step": 1482
},
{
"epoch": 0.0596180904522613,
"grad_norm": 2.267002582550049,
"learning_rate": 1.298200217591439e-05,
"loss": 7.5406646728515625,
"step": 1483
},
{
"epoch": 0.05965829145728643,
"grad_norm": 2.3365542888641357,
"learning_rate": 1.2981958926009853e-05,
"loss": 7.708248138427734,
"step": 1484
},
{
"epoch": 0.059698492462311556,
"grad_norm": 2.5624051094055176,
"learning_rate": 1.298191562434617e-05,
"loss": 7.931509017944336,
"step": 1485
},
{
"epoch": 0.059738693467336686,
"grad_norm": 2.4687981605529785,
"learning_rate": 1.2981872270924044e-05,
"loss": 7.62822151184082,
"step": 1486
},
{
"epoch": 0.05977889447236181,
"grad_norm": 2.273233652114868,
"learning_rate": 1.2981828865744162e-05,
"loss": 7.5245466232299805,
"step": 1487
},
{
"epoch": 0.05981909547738693,
"grad_norm": 2.4390811920166016,
"learning_rate": 1.2981785408807222e-05,
"loss": 7.063882827758789,
"step": 1488
},
{
"epoch": 0.05985929648241206,
"grad_norm": 2.4166667461395264,
"learning_rate": 1.298174190011392e-05,
"loss": 7.875926971435547,
"step": 1489
},
{
"epoch": 0.059899497487437184,
"grad_norm": 2.3124947547912598,
"learning_rate": 1.298169833966495e-05,
"loss": 7.070745468139648,
"step": 1490
},
{
"epoch": 0.059939698492462314,
"grad_norm": 2.2154042720794678,
"learning_rate": 1.2981654727461015e-05,
"loss": 7.403896331787109,
"step": 1491
},
{
"epoch": 0.059979899497487436,
"grad_norm": 2.318145513534546,
"learning_rate": 1.2981611063502807e-05,
"loss": 7.0121049880981445,
"step": 1492
},
{
"epoch": 0.060020100502512566,
"grad_norm": 2.276421070098877,
"learning_rate": 1.2981567347791031e-05,
"loss": 7.433084964752197,
"step": 1493
},
{
"epoch": 0.06006030150753769,
"grad_norm": 2.3185787200927734,
"learning_rate": 1.2981523580326386e-05,
"loss": 7.200679779052734,
"step": 1494
},
{
"epoch": 0.06010050251256281,
"grad_norm": 2.0806360244750977,
"learning_rate": 1.2981479761109568e-05,
"loss": 7.688490867614746,
"step": 1495
},
{
"epoch": 0.06014070351758794,
"grad_norm": 2.5642144680023193,
"learning_rate": 1.2981435890141285e-05,
"loss": 7.150332450866699,
"step": 1496
},
{
"epoch": 0.060180904522613064,
"grad_norm": 2.277036190032959,
"learning_rate": 1.2981391967422234e-05,
"loss": 7.414963245391846,
"step": 1497
},
{
"epoch": 0.060221105527638194,
"grad_norm": 2.3142144680023193,
"learning_rate": 1.2981347992953125e-05,
"loss": 7.474861145019531,
"step": 1498
},
{
"epoch": 0.06026130653266332,
"grad_norm": 2.6704816818237305,
"learning_rate": 1.2981303966734658e-05,
"loss": 7.247389793395996,
"step": 1499
},
{
"epoch": 0.06030150753768844,
"grad_norm": 2.4358110427856445,
"learning_rate": 1.2981259888767536e-05,
"loss": 7.767067909240723,
"step": 1500
},
{
"epoch": 0.06034170854271357,
"grad_norm": 2.399989604949951,
"learning_rate": 1.298121575905247e-05,
"loss": 7.390251159667969,
"step": 1501
},
{
"epoch": 0.06038190954773869,
"grad_norm": 2.287109613418579,
"learning_rate": 1.2981171577590162e-05,
"loss": 7.364101409912109,
"step": 1502
},
{
"epoch": 0.06042211055276382,
"grad_norm": 2.6516809463500977,
"learning_rate": 1.2981127344381323e-05,
"loss": 7.497237682342529,
"step": 1503
},
{
"epoch": 0.060462311557788945,
"grad_norm": 2.3114569187164307,
"learning_rate": 1.298108305942666e-05,
"loss": 7.772374629974365,
"step": 1504
},
{
"epoch": 0.06050251256281407,
"grad_norm": 2.202960968017578,
"learning_rate": 1.2981038722726883e-05,
"loss": 7.563572883605957,
"step": 1505
},
{
"epoch": 0.0605427135678392,
"grad_norm": 2.77056884765625,
"learning_rate": 1.2980994334282701e-05,
"loss": 7.643380165100098,
"step": 1506
},
{
"epoch": 0.06058291457286432,
"grad_norm": 2.294741153717041,
"learning_rate": 1.2980949894094827e-05,
"loss": 7.964155197143555,
"step": 1507
},
{
"epoch": 0.06062311557788945,
"grad_norm": 2.2332522869110107,
"learning_rate": 1.2980905402163969e-05,
"loss": 7.43630313873291,
"step": 1508
},
{
"epoch": 0.06066331658291457,
"grad_norm": 2.280829906463623,
"learning_rate": 1.2980860858490842e-05,
"loss": 7.621092319488525,
"step": 1509
},
{
"epoch": 0.060703517587939695,
"grad_norm": 2.351840019226074,
"learning_rate": 1.298081626307616e-05,
"loss": 7.509548187255859,
"step": 1510
},
{
"epoch": 0.060743718592964825,
"grad_norm": 2.183467149734497,
"learning_rate": 1.2980771615920635e-05,
"loss": 7.428604602813721,
"step": 1511
},
{
"epoch": 0.06078391959798995,
"grad_norm": 2.2210302352905273,
"learning_rate": 1.2980726917024984e-05,
"loss": 7.439453125,
"step": 1512
},
{
"epoch": 0.06082412060301508,
"grad_norm": 2.4463820457458496,
"learning_rate": 1.2980682166389922e-05,
"loss": 7.0764312744140625,
"step": 1513
},
{
"epoch": 0.0608643216080402,
"grad_norm": 2.304980754852295,
"learning_rate": 1.2980637364016165e-05,
"loss": 7.622719764709473,
"step": 1514
},
{
"epoch": 0.06090452261306532,
"grad_norm": 2.682985305786133,
"learning_rate": 1.2980592509904431e-05,
"loss": 7.243518829345703,
"step": 1515
},
{
"epoch": 0.06094472361809045,
"grad_norm": 2.4511818885803223,
"learning_rate": 1.2980547604055438e-05,
"loss": 7.684053421020508,
"step": 1516
},
{
"epoch": 0.060984924623115576,
"grad_norm": 2.489790678024292,
"learning_rate": 1.2980502646469907e-05,
"loss": 7.375517845153809,
"step": 1517
},
{
"epoch": 0.061025125628140706,
"grad_norm": 2.8786511421203613,
"learning_rate": 1.2980457637148557e-05,
"loss": 7.800781726837158,
"step": 1518
},
{
"epoch": 0.06106532663316583,
"grad_norm": 2.328639507293701,
"learning_rate": 1.2980412576092107e-05,
"loss": 7.0824689865112305,
"step": 1519
},
{
"epoch": 0.06110552763819096,
"grad_norm": 2.329791784286499,
"learning_rate": 1.298036746330128e-05,
"loss": 7.732283592224121,
"step": 1520
},
{
"epoch": 0.06114572864321608,
"grad_norm": 2.276386022567749,
"learning_rate": 1.2980322298776801e-05,
"loss": 7.201354026794434,
"step": 1521
},
{
"epoch": 0.061185929648241204,
"grad_norm": 2.7571794986724854,
"learning_rate": 1.2980277082519388e-05,
"loss": 7.868547439575195,
"step": 1522
},
{
"epoch": 0.061226130653266334,
"grad_norm": 2.3509044647216797,
"learning_rate": 1.2980231814529769e-05,
"loss": 7.614871025085449,
"step": 1523
},
{
"epoch": 0.061266331658291456,
"grad_norm": 2.346341609954834,
"learning_rate": 1.2980186494808668e-05,
"loss": 7.811978340148926,
"step": 1524
},
{
"epoch": 0.061306532663316586,
"grad_norm": 2.2549362182617188,
"learning_rate": 1.2980141123356812e-05,
"loss": 7.787387847900391,
"step": 1525
},
{
"epoch": 0.06134673366834171,
"grad_norm": 2.448573589324951,
"learning_rate": 1.2980095700174925e-05,
"loss": 7.728471755981445,
"step": 1526
},
{
"epoch": 0.06138693467336683,
"grad_norm": 2.377652645111084,
"learning_rate": 1.2980050225263735e-05,
"loss": 7.108550071716309,
"step": 1527
},
{
"epoch": 0.06142713567839196,
"grad_norm": 2.269887924194336,
"learning_rate": 1.2980004698623973e-05,
"loss": 7.572266578674316,
"step": 1528
},
{
"epoch": 0.061467336683417084,
"grad_norm": 2.587888717651367,
"learning_rate": 1.2979959120256363e-05,
"loss": 8.11519718170166,
"step": 1529
},
{
"epoch": 0.061507537688442214,
"grad_norm": 2.276773691177368,
"learning_rate": 1.297991349016164e-05,
"loss": 7.670904159545898,
"step": 1530
},
{
"epoch": 0.06154773869346734,
"grad_norm": 2.493846893310547,
"learning_rate": 1.2979867808340534e-05,
"loss": 7.615190505981445,
"step": 1531
},
{
"epoch": 0.06158793969849246,
"grad_norm": 2.295748710632324,
"learning_rate": 1.2979822074793776e-05,
"loss": 7.459231376647949,
"step": 1532
},
{
"epoch": 0.06162814070351759,
"grad_norm": 2.416517734527588,
"learning_rate": 1.2979776289522096e-05,
"loss": 7.179243087768555,
"step": 1533
},
{
"epoch": 0.06166834170854271,
"grad_norm": 2.3631584644317627,
"learning_rate": 1.2979730452526228e-05,
"loss": 7.823232650756836,
"step": 1534
},
{
"epoch": 0.06170854271356784,
"grad_norm": 2.987428903579712,
"learning_rate": 1.2979684563806909e-05,
"loss": 7.5467023849487305,
"step": 1535
},
{
"epoch": 0.061748743718592965,
"grad_norm": 2.39436936378479,
"learning_rate": 1.2979638623364872e-05,
"loss": 7.6570611000061035,
"step": 1536
},
{
"epoch": 0.06178894472361809,
"grad_norm": 2.234060049057007,
"learning_rate": 1.2979592631200853e-05,
"loss": 7.84591007232666,
"step": 1537
},
{
"epoch": 0.06182914572864322,
"grad_norm": 2.3285951614379883,
"learning_rate": 1.2979546587315587e-05,
"loss": 7.232288360595703,
"step": 1538
},
{
"epoch": 0.06186934673366834,
"grad_norm": 2.5654876232147217,
"learning_rate": 1.2979500491709814e-05,
"loss": 7.14811897277832,
"step": 1539
},
{
"epoch": 0.06190954773869347,
"grad_norm": 2.784275531768799,
"learning_rate": 1.297945434438427e-05,
"loss": 7.76908016204834,
"step": 1540
},
{
"epoch": 0.06194974874371859,
"grad_norm": 2.8063957691192627,
"learning_rate": 1.2979408145339695e-05,
"loss": 7.296648979187012,
"step": 1541
},
{
"epoch": 0.061989949748743715,
"grad_norm": 2.2995572090148926,
"learning_rate": 1.2979361894576828e-05,
"loss": 7.651863098144531,
"step": 1542
},
{
"epoch": 0.062030150753768845,
"grad_norm": 2.3538546562194824,
"learning_rate": 1.2979315592096414e-05,
"loss": 7.587031364440918,
"step": 1543
},
{
"epoch": 0.06207035175879397,
"grad_norm": 2.270311117172241,
"learning_rate": 1.2979269237899188e-05,
"loss": 7.450778007507324,
"step": 1544
},
{
"epoch": 0.0621105527638191,
"grad_norm": 2.35296893119812,
"learning_rate": 1.2979222831985896e-05,
"loss": 7.502727031707764,
"step": 1545
},
{
"epoch": 0.06215075376884422,
"grad_norm": 2.289721965789795,
"learning_rate": 1.297917637435728e-05,
"loss": 7.098215103149414,
"step": 1546
},
{
"epoch": 0.06219095477386934,
"grad_norm": 2.4431378841400146,
"learning_rate": 1.2979129865014086e-05,
"loss": 7.168883800506592,
"step": 1547
},
{
"epoch": 0.06223115577889447,
"grad_norm": 2.863924026489258,
"learning_rate": 1.2979083303957058e-05,
"loss": 7.716777801513672,
"step": 1548
},
{
"epoch": 0.062271356783919596,
"grad_norm": 2.2293312549591064,
"learning_rate": 1.297903669118694e-05,
"loss": 7.29901647567749,
"step": 1549
},
{
"epoch": 0.062311557788944726,
"grad_norm": 2.972778558731079,
"learning_rate": 1.2978990026704479e-05,
"loss": 7.890946388244629,
"step": 1550
},
{
"epoch": 0.06235175879396985,
"grad_norm": 2.5059094429016113,
"learning_rate": 1.2978943310510423e-05,
"loss": 7.835686683654785,
"step": 1551
},
{
"epoch": 0.06239195979899498,
"grad_norm": 2.5672390460968018,
"learning_rate": 1.2978896542605521e-05,
"loss": 7.447885513305664,
"step": 1552
},
{
"epoch": 0.0624321608040201,
"grad_norm": 3.0500142574310303,
"learning_rate": 1.2978849722990521e-05,
"loss": 7.147265434265137,
"step": 1553
},
{
"epoch": 0.062472361809045224,
"grad_norm": 2.954380750656128,
"learning_rate": 1.2978802851666172e-05,
"loss": 7.521335601806641,
"step": 1554
},
{
"epoch": 0.06251256281407035,
"grad_norm": 2.551337718963623,
"learning_rate": 1.2978755928633225e-05,
"loss": 7.558992385864258,
"step": 1555
},
{
"epoch": 0.06255276381909548,
"grad_norm": 2.2909164428710938,
"learning_rate": 1.2978708953892434e-05,
"loss": 7.429536819458008,
"step": 1556
},
{
"epoch": 0.0625929648241206,
"grad_norm": 2.3539812564849854,
"learning_rate": 1.2978661927444548e-05,
"loss": 7.455727577209473,
"step": 1557
},
{
"epoch": 0.06263316582914573,
"grad_norm": 2.7275888919830322,
"learning_rate": 1.297861484929032e-05,
"loss": 7.703391075134277,
"step": 1558
},
{
"epoch": 0.06267336683417085,
"grad_norm": 2.7580416202545166,
"learning_rate": 1.2978567719430508e-05,
"loss": 7.5927228927612305,
"step": 1559
},
{
"epoch": 0.06271356783919597,
"grad_norm": 2.431143283843994,
"learning_rate": 1.2978520537865863e-05,
"loss": 7.593696117401123,
"step": 1560
},
{
"epoch": 0.06275376884422111,
"grad_norm": 2.547147750854492,
"learning_rate": 1.2978473304597141e-05,
"loss": 7.8774003982543945,
"step": 1561
},
{
"epoch": 0.06279396984924623,
"grad_norm": 2.4708662033081055,
"learning_rate": 1.29784260196251e-05,
"loss": 7.720075607299805,
"step": 1562
},
{
"epoch": 0.06283417085427136,
"grad_norm": 2.291342258453369,
"learning_rate": 1.2978378682950496e-05,
"loss": 7.330462455749512,
"step": 1563
},
{
"epoch": 0.06287437185929648,
"grad_norm": 2.475267171859741,
"learning_rate": 1.2978331294574087e-05,
"loss": 7.5009942054748535,
"step": 1564
},
{
"epoch": 0.0629145728643216,
"grad_norm": 2.1824028491973877,
"learning_rate": 1.2978283854496633e-05,
"loss": 7.44306755065918,
"step": 1565
},
{
"epoch": 0.06295477386934674,
"grad_norm": 2.3428804874420166,
"learning_rate": 1.2978236362718894e-05,
"loss": 7.110868453979492,
"step": 1566
},
{
"epoch": 0.06299497487437186,
"grad_norm": 2.533464193344116,
"learning_rate": 1.2978188819241628e-05,
"loss": 7.277129173278809,
"step": 1567
},
{
"epoch": 0.06303517587939698,
"grad_norm": 2.415717601776123,
"learning_rate": 1.29781412240656e-05,
"loss": 7.670809745788574,
"step": 1568
},
{
"epoch": 0.06307537688442211,
"grad_norm": 2.600038766860962,
"learning_rate": 1.2978093577191567e-05,
"loss": 7.462311744689941,
"step": 1569
},
{
"epoch": 0.06311557788944723,
"grad_norm": 2.492475748062134,
"learning_rate": 1.2978045878620298e-05,
"loss": 7.66529655456543,
"step": 1570
},
{
"epoch": 0.06315577889447237,
"grad_norm": 2.750822067260742,
"learning_rate": 1.2977998128352554e-05,
"loss": 7.3370819091796875,
"step": 1571
},
{
"epoch": 0.06319597989949749,
"grad_norm": 2.2162880897521973,
"learning_rate": 1.29779503263891e-05,
"loss": 7.672133445739746,
"step": 1572
},
{
"epoch": 0.06323618090452261,
"grad_norm": 2.2987875938415527,
"learning_rate": 1.29779024727307e-05,
"loss": 7.221503257751465,
"step": 1573
},
{
"epoch": 0.06327638190954774,
"grad_norm": 2.112161874771118,
"learning_rate": 1.2977854567378124e-05,
"loss": 7.6378889083862305,
"step": 1574
},
{
"epoch": 0.06331658291457286,
"grad_norm": 2.882279634475708,
"learning_rate": 1.2977806610332136e-05,
"loss": 7.334000587463379,
"step": 1575
},
{
"epoch": 0.063356783919598,
"grad_norm": 2.741586923599243,
"learning_rate": 1.2977758601593506e-05,
"loss": 7.66588020324707,
"step": 1576
},
{
"epoch": 0.06339698492462312,
"grad_norm": 2.896726131439209,
"learning_rate": 1.2977710541162999e-05,
"loss": 7.7630109786987305,
"step": 1577
},
{
"epoch": 0.06343718592964824,
"grad_norm": 2.2499213218688965,
"learning_rate": 1.297766242904139e-05,
"loss": 7.795592784881592,
"step": 1578
},
{
"epoch": 0.06347738693467336,
"grad_norm": 2.293241500854492,
"learning_rate": 1.2977614265229447e-05,
"loss": 7.5323710441589355,
"step": 1579
},
{
"epoch": 0.06351758793969849,
"grad_norm": 2.6957931518554688,
"learning_rate": 1.2977566049727942e-05,
"loss": 7.579814434051514,
"step": 1580
},
{
"epoch": 0.06355778894472362,
"grad_norm": 2.7324254512786865,
"learning_rate": 1.2977517782537647e-05,
"loss": 7.419973373413086,
"step": 1581
},
{
"epoch": 0.06359798994974875,
"grad_norm": 2.6555378437042236,
"learning_rate": 1.2977469463659333e-05,
"loss": 7.747026443481445,
"step": 1582
},
{
"epoch": 0.06363819095477387,
"grad_norm": 2.4300804138183594,
"learning_rate": 1.2977421093093776e-05,
"loss": 7.544895648956299,
"step": 1583
},
{
"epoch": 0.06367839195979899,
"grad_norm": 2.744051218032837,
"learning_rate": 1.297737267084175e-05,
"loss": 7.224091529846191,
"step": 1584
},
{
"epoch": 0.06371859296482411,
"grad_norm": 2.8267934322357178,
"learning_rate": 1.2977324196904031e-05,
"loss": 7.351358413696289,
"step": 1585
},
{
"epoch": 0.06375879396984925,
"grad_norm": 2.3511180877685547,
"learning_rate": 1.2977275671281392e-05,
"loss": 7.40477180480957,
"step": 1586
},
{
"epoch": 0.06379899497487437,
"grad_norm": 2.5510754585266113,
"learning_rate": 1.2977227093974616e-05,
"loss": 7.456686496734619,
"step": 1587
},
{
"epoch": 0.0638391959798995,
"grad_norm": 2.214064598083496,
"learning_rate": 1.2977178464984477e-05,
"loss": 8.226888656616211,
"step": 1588
},
{
"epoch": 0.06387939698492462,
"grad_norm": 2.31791615486145,
"learning_rate": 1.2977129784311756e-05,
"loss": 7.428415775299072,
"step": 1589
},
{
"epoch": 0.06391959798994974,
"grad_norm": 2.1461586952209473,
"learning_rate": 1.297708105195723e-05,
"loss": 7.906563758850098,
"step": 1590
},
{
"epoch": 0.06395979899497488,
"grad_norm": 2.2355499267578125,
"learning_rate": 1.297703226792168e-05,
"loss": 7.30942964553833,
"step": 1591
},
{
"epoch": 0.064,
"grad_norm": 2.382241725921631,
"learning_rate": 1.297698343220589e-05,
"loss": 7.471290588378906,
"step": 1592
},
{
"epoch": 0.06404020100502512,
"grad_norm": 2.255864381790161,
"learning_rate": 1.2976934544810639e-05,
"loss": 7.761112213134766,
"step": 1593
},
{
"epoch": 0.06408040201005025,
"grad_norm": 2.221661329269409,
"learning_rate": 1.297688560573671e-05,
"loss": 7.266605377197266,
"step": 1594
},
{
"epoch": 0.06412060301507538,
"grad_norm": 2.467535972595215,
"learning_rate": 1.297683661498489e-05,
"loss": 7.84549617767334,
"step": 1595
},
{
"epoch": 0.0641608040201005,
"grad_norm": 2.2189884185791016,
"learning_rate": 1.297678757255596e-05,
"loss": 7.387146472930908,
"step": 1596
},
{
"epoch": 0.06420100502512563,
"grad_norm": 2.354505777359009,
"learning_rate": 1.2976738478450707e-05,
"loss": 7.654148101806641,
"step": 1597
},
{
"epoch": 0.06424120603015075,
"grad_norm": 2.286128520965576,
"learning_rate": 1.2976689332669917e-05,
"loss": 7.196654796600342,
"step": 1598
},
{
"epoch": 0.06428140703517587,
"grad_norm": 2.5749776363372803,
"learning_rate": 1.2976640135214378e-05,
"loss": 7.719487190246582,
"step": 1599
},
{
"epoch": 0.06432160804020101,
"grad_norm": 2.4728739261627197,
"learning_rate": 1.2976590886084874e-05,
"loss": 7.740029335021973,
"step": 1600
},
{
"epoch": 0.06432160804020101,
"eval_loss": 7.516892910003662,
"eval_runtime": 120.6758,
"eval_samples_per_second": 4.143,
"eval_steps_per_second": 0.522,
"step": 1600
},
{
"epoch": 0.06436180904522613,
"grad_norm": 2.2308719158172607,
"learning_rate": 1.29765415852822e-05,
"loss": 7.622395992279053,
"step": 1601
},
{
"epoch": 0.06440201005025126,
"grad_norm": 2.3598074913024902,
"learning_rate": 1.297649223280714e-05,
"loss": 7.555525779724121,
"step": 1602
},
{
"epoch": 0.06444221105527638,
"grad_norm": 2.2899773120880127,
"learning_rate": 1.2976442828660487e-05,
"loss": 7.319741249084473,
"step": 1603
},
{
"epoch": 0.0644824120603015,
"grad_norm": 2.546194076538086,
"learning_rate": 1.2976393372843033e-05,
"loss": 7.802236557006836,
"step": 1604
},
{
"epoch": 0.06452261306532664,
"grad_norm": 2.219637155532837,
"learning_rate": 1.2976343865355568e-05,
"loss": 7.437636375427246,
"step": 1605
},
{
"epoch": 0.06456281407035176,
"grad_norm": 2.4005284309387207,
"learning_rate": 1.2976294306198885e-05,
"loss": 7.023745536804199,
"step": 1606
},
{
"epoch": 0.06460301507537689,
"grad_norm": 2.961110830307007,
"learning_rate": 1.297624469537378e-05,
"loss": 7.759522438049316,
"step": 1607
},
{
"epoch": 0.06464321608040201,
"grad_norm": 2.480278253555298,
"learning_rate": 1.2976195032881044e-05,
"loss": 6.942720413208008,
"step": 1608
},
{
"epoch": 0.06468341708542713,
"grad_norm": 2.3043317794799805,
"learning_rate": 1.2976145318721475e-05,
"loss": 7.890336990356445,
"step": 1609
},
{
"epoch": 0.06472361809045227,
"grad_norm": 2.757943630218506,
"learning_rate": 1.2976095552895868e-05,
"loss": 7.073426246643066,
"step": 1610
},
{
"epoch": 0.06476381909547739,
"grad_norm": 2.6229794025421143,
"learning_rate": 1.297604573540502e-05,
"loss": 7.204015731811523,
"step": 1611
},
{
"epoch": 0.06480402010050251,
"grad_norm": 2.4719719886779785,
"learning_rate": 1.297599586624973e-05,
"loss": 7.233782768249512,
"step": 1612
},
{
"epoch": 0.06484422110552764,
"grad_norm": 2.8201262950897217,
"learning_rate": 1.2975945945430795e-05,
"loss": 7.8337907791137695,
"step": 1613
},
{
"epoch": 0.06488442211055276,
"grad_norm": 2.4991908073425293,
"learning_rate": 1.2975895972949016e-05,
"loss": 7.1946916580200195,
"step": 1614
},
{
"epoch": 0.0649246231155779,
"grad_norm": 2.6320207118988037,
"learning_rate": 1.2975845948805193e-05,
"loss": 7.530649185180664,
"step": 1615
},
{
"epoch": 0.06496482412060302,
"grad_norm": 2.410853862762451,
"learning_rate": 1.2975795873000126e-05,
"loss": 7.485363960266113,
"step": 1616
},
{
"epoch": 0.06500502512562814,
"grad_norm": 2.596956491470337,
"learning_rate": 1.2975745745534617e-05,
"loss": 7.509760856628418,
"step": 1617
},
{
"epoch": 0.06504522613065326,
"grad_norm": 2.3935256004333496,
"learning_rate": 1.2975695566409468e-05,
"loss": 7.559047698974609,
"step": 1618
},
{
"epoch": 0.06508542713567839,
"grad_norm": 2.400164842605591,
"learning_rate": 1.2975645335625487e-05,
"loss": 7.514081001281738,
"step": 1619
},
{
"epoch": 0.06512562814070352,
"grad_norm": 2.426826000213623,
"learning_rate": 1.2975595053183474e-05,
"loss": 7.443843841552734,
"step": 1620
},
{
"epoch": 0.06516582914572865,
"grad_norm": 2.5844836235046387,
"learning_rate": 1.2975544719084236e-05,
"loss": 7.626278877258301,
"step": 1621
},
{
"epoch": 0.06520603015075377,
"grad_norm": 2.4096169471740723,
"learning_rate": 1.297549433332858e-05,
"loss": 7.514891147613525,
"step": 1622
},
{
"epoch": 0.06524623115577889,
"grad_norm": 2.4178311824798584,
"learning_rate": 1.297544389591731e-05,
"loss": 7.6847028732299805,
"step": 1623
},
{
"epoch": 0.06528643216080401,
"grad_norm": 2.6126997470855713,
"learning_rate": 1.2975393406851236e-05,
"loss": 7.711097240447998,
"step": 1624
},
{
"epoch": 0.06532663316582915,
"grad_norm": 2.4473347663879395,
"learning_rate": 1.2975342866131166e-05,
"loss": 7.714642524719238,
"step": 1625
},
{
"epoch": 0.06536683417085427,
"grad_norm": 2.3067500591278076,
"learning_rate": 1.2975292273757908e-05,
"loss": 7.6399383544921875,
"step": 1626
},
{
"epoch": 0.0654070351758794,
"grad_norm": 2.7727551460266113,
"learning_rate": 1.2975241629732277e-05,
"loss": 7.662282943725586,
"step": 1627
},
{
"epoch": 0.06544723618090452,
"grad_norm": 2.3030409812927246,
"learning_rate": 1.2975190934055078e-05,
"loss": 7.488182067871094,
"step": 1628
},
{
"epoch": 0.06548743718592964,
"grad_norm": 2.781752347946167,
"learning_rate": 1.2975140186727126e-05,
"loss": 7.276732444763184,
"step": 1629
},
{
"epoch": 0.06552763819095478,
"grad_norm": 2.2313172817230225,
"learning_rate": 1.2975089387749234e-05,
"loss": 7.647853851318359,
"step": 1630
},
{
"epoch": 0.0655678391959799,
"grad_norm": 2.546745538711548,
"learning_rate": 1.2975038537122216e-05,
"loss": 7.259021282196045,
"step": 1631
},
{
"epoch": 0.06560804020100502,
"grad_norm": 2.9993667602539062,
"learning_rate": 1.2974987634846882e-05,
"loss": 7.350425720214844,
"step": 1632
},
{
"epoch": 0.06564824120603015,
"grad_norm": 2.5772910118103027,
"learning_rate": 1.2974936680924052e-05,
"loss": 7.65567684173584,
"step": 1633
},
{
"epoch": 0.06568844221105527,
"grad_norm": 2.479753017425537,
"learning_rate": 1.2974885675354542e-05,
"loss": 7.428635597229004,
"step": 1634
},
{
"epoch": 0.0657286432160804,
"grad_norm": 2.214757204055786,
"learning_rate": 1.2974834618139164e-05,
"loss": 7.633677959442139,
"step": 1635
},
{
"epoch": 0.06576884422110553,
"grad_norm": 2.95318341255188,
"learning_rate": 1.297478350927874e-05,
"loss": 7.379066467285156,
"step": 1636
},
{
"epoch": 0.06580904522613065,
"grad_norm": 2.365093946456909,
"learning_rate": 1.2974732348774089e-05,
"loss": 7.327049255371094,
"step": 1637
},
{
"epoch": 0.06584924623115578,
"grad_norm": 2.4091804027557373,
"learning_rate": 1.2974681136626027e-05,
"loss": 7.2860918045043945,
"step": 1638
},
{
"epoch": 0.0658894472361809,
"grad_norm": 2.19394588470459,
"learning_rate": 1.2974629872835377e-05,
"loss": 7.333078384399414,
"step": 1639
},
{
"epoch": 0.06592964824120603,
"grad_norm": 2.2826383113861084,
"learning_rate": 1.297457855740296e-05,
"loss": 7.1494011878967285,
"step": 1640
},
{
"epoch": 0.06596984924623116,
"grad_norm": 2.425487756729126,
"learning_rate": 1.2974527190329594e-05,
"loss": 7.385140895843506,
"step": 1641
},
{
"epoch": 0.06601005025125628,
"grad_norm": 2.3549065589904785,
"learning_rate": 1.2974475771616106e-05,
"loss": 7.566284656524658,
"step": 1642
},
{
"epoch": 0.0660502512562814,
"grad_norm": 2.4767372608184814,
"learning_rate": 1.2974424301263317e-05,
"loss": 7.766632080078125,
"step": 1643
},
{
"epoch": 0.06609045226130653,
"grad_norm": 2.775641441345215,
"learning_rate": 1.2974372779272052e-05,
"loss": 7.63411283493042,
"step": 1644
},
{
"epoch": 0.06613065326633166,
"grad_norm": 2.352208137512207,
"learning_rate": 1.2974321205643136e-05,
"loss": 7.517523765563965,
"step": 1645
},
{
"epoch": 0.06617085427135679,
"grad_norm": 2.288104295730591,
"learning_rate": 1.2974269580377396e-05,
"loss": 7.611756324768066,
"step": 1646
},
{
"epoch": 0.06621105527638191,
"grad_norm": 2.355242967605591,
"learning_rate": 1.2974217903475657e-05,
"loss": 7.720600128173828,
"step": 1647
},
{
"epoch": 0.06625125628140703,
"grad_norm": 2.454615831375122,
"learning_rate": 1.2974166174938749e-05,
"loss": 7.666622161865234,
"step": 1648
},
{
"epoch": 0.06629145728643217,
"grad_norm": 2.2801780700683594,
"learning_rate": 1.2974114394767498e-05,
"loss": 7.6667656898498535,
"step": 1649
},
{
"epoch": 0.06633165829145729,
"grad_norm": 2.342599630355835,
"learning_rate": 1.2974062562962733e-05,
"loss": 7.587776184082031,
"step": 1650
},
{
"epoch": 0.06637185929648241,
"grad_norm": 2.7105135917663574,
"learning_rate": 1.2974010679525286e-05,
"loss": 7.528389930725098,
"step": 1651
},
{
"epoch": 0.06641206030150754,
"grad_norm": 2.23891282081604,
"learning_rate": 1.2973958744455987e-05,
"loss": 7.441523551940918,
"step": 1652
},
{
"epoch": 0.06645226130653266,
"grad_norm": 2.2837955951690674,
"learning_rate": 1.2973906757755669e-05,
"loss": 7.493271350860596,
"step": 1653
},
{
"epoch": 0.0664924623115578,
"grad_norm": 2.2477974891662598,
"learning_rate": 1.297385471942516e-05,
"loss": 7.561607360839844,
"step": 1654
},
{
"epoch": 0.06653266331658292,
"grad_norm": 2.276381254196167,
"learning_rate": 1.2973802629465301e-05,
"loss": 7.457236289978027,
"step": 1655
},
{
"epoch": 0.06657286432160804,
"grad_norm": 2.810595989227295,
"learning_rate": 1.2973750487876921e-05,
"loss": 7.328514099121094,
"step": 1656
},
{
"epoch": 0.06661306532663316,
"grad_norm": 2.312901258468628,
"learning_rate": 1.2973698294660855e-05,
"loss": 7.421909809112549,
"step": 1657
},
{
"epoch": 0.06665326633165829,
"grad_norm": 2.371809482574463,
"learning_rate": 1.2973646049817941e-05,
"loss": 7.394037246704102,
"step": 1658
},
{
"epoch": 0.06669346733668342,
"grad_norm": 2.319122076034546,
"learning_rate": 1.2973593753349013e-05,
"loss": 7.599871635437012,
"step": 1659
},
{
"epoch": 0.06673366834170855,
"grad_norm": 2.2263760566711426,
"learning_rate": 1.2973541405254912e-05,
"loss": 7.557384967803955,
"step": 1660
},
{
"epoch": 0.06677386934673367,
"grad_norm": 2.3904638290405273,
"learning_rate": 1.2973489005536474e-05,
"loss": 7.376008033752441,
"step": 1661
},
{
"epoch": 0.06681407035175879,
"grad_norm": 2.3489606380462646,
"learning_rate": 1.2973436554194535e-05,
"loss": 7.385540008544922,
"step": 1662
},
{
"epoch": 0.06685427135678391,
"grad_norm": 2.4562463760375977,
"learning_rate": 1.2973384051229943e-05,
"loss": 7.316134929656982,
"step": 1663
},
{
"epoch": 0.06689447236180905,
"grad_norm": 2.6277530193328857,
"learning_rate": 1.2973331496643533e-05,
"loss": 6.982102870941162,
"step": 1664
},
{
"epoch": 0.06693467336683417,
"grad_norm": 2.6250946521759033,
"learning_rate": 1.2973278890436147e-05,
"loss": 7.913159370422363,
"step": 1665
},
{
"epoch": 0.0669748743718593,
"grad_norm": 2.398146629333496,
"learning_rate": 1.297322623260863e-05,
"loss": 7.094391822814941,
"step": 1666
},
{
"epoch": 0.06701507537688442,
"grad_norm": 2.441885471343994,
"learning_rate": 1.2973173523161823e-05,
"loss": 7.449734687805176,
"step": 1667
},
{
"epoch": 0.06705527638190954,
"grad_norm": 2.641500473022461,
"learning_rate": 1.2973120762096574e-05,
"loss": 7.572999000549316,
"step": 1668
},
{
"epoch": 0.06709547738693468,
"grad_norm": 2.3521955013275146,
"learning_rate": 1.2973067949413722e-05,
"loss": 7.9069719314575195,
"step": 1669
},
{
"epoch": 0.0671356783919598,
"grad_norm": 2.6316721439361572,
"learning_rate": 1.2973015085114116e-05,
"loss": 7.400055408477783,
"step": 1670
},
{
"epoch": 0.06717587939698492,
"grad_norm": 2.7254278659820557,
"learning_rate": 1.2972962169198603e-05,
"loss": 7.2258100509643555,
"step": 1671
},
{
"epoch": 0.06721608040201005,
"grad_norm": 2.4949843883514404,
"learning_rate": 1.2972909201668028e-05,
"loss": 7.246678352355957,
"step": 1672
},
{
"epoch": 0.06725628140703517,
"grad_norm": 2.612988233566284,
"learning_rate": 1.2972856182523244e-05,
"loss": 7.677805423736572,
"step": 1673
},
{
"epoch": 0.06729648241206031,
"grad_norm": 2.4618959426879883,
"learning_rate": 1.2972803111765095e-05,
"loss": 6.9700703620910645,
"step": 1674
},
{
"epoch": 0.06733668341708543,
"grad_norm": 2.2974910736083984,
"learning_rate": 1.2972749989394433e-05,
"loss": 7.649855613708496,
"step": 1675
},
{
"epoch": 0.06737688442211055,
"grad_norm": 2.326312780380249,
"learning_rate": 1.2972696815412109e-05,
"loss": 7.492466926574707,
"step": 1676
},
{
"epoch": 0.06741708542713568,
"grad_norm": 2.1657943725585938,
"learning_rate": 1.2972643589818975e-05,
"loss": 7.639139175415039,
"step": 1677
},
{
"epoch": 0.0674572864321608,
"grad_norm": 2.507692813873291,
"learning_rate": 1.2972590312615882e-05,
"loss": 7.262688636779785,
"step": 1678
},
{
"epoch": 0.06749748743718594,
"grad_norm": 2.626723527908325,
"learning_rate": 1.2972536983803686e-05,
"loss": 7.7373576164245605,
"step": 1679
},
{
"epoch": 0.06753768844221106,
"grad_norm": 2.5468990802764893,
"learning_rate": 1.2972483603383237e-05,
"loss": 7.530850410461426,
"step": 1680
},
{
"epoch": 0.06757788944723618,
"grad_norm": 2.4965078830718994,
"learning_rate": 1.2972430171355392e-05,
"loss": 7.7787652015686035,
"step": 1681
},
{
"epoch": 0.0676180904522613,
"grad_norm": 2.375220775604248,
"learning_rate": 1.2972376687721007e-05,
"loss": 7.631651878356934,
"step": 1682
},
{
"epoch": 0.06765829145728643,
"grad_norm": 2.518419027328491,
"learning_rate": 1.2972323152480939e-05,
"loss": 7.593117713928223,
"step": 1683
},
{
"epoch": 0.06769849246231156,
"grad_norm": 2.5783071517944336,
"learning_rate": 1.2972269565636042e-05,
"loss": 7.431680679321289,
"step": 1684
},
{
"epoch": 0.06773869346733669,
"grad_norm": 2.424257755279541,
"learning_rate": 1.2972215927187179e-05,
"loss": 8.084222793579102,
"step": 1685
},
{
"epoch": 0.06777889447236181,
"grad_norm": 2.553393602371216,
"learning_rate": 1.2972162237135207e-05,
"loss": 7.612630367279053,
"step": 1686
},
{
"epoch": 0.06781909547738693,
"grad_norm": 2.4463462829589844,
"learning_rate": 1.2972108495480985e-05,
"loss": 8.002542495727539,
"step": 1687
},
{
"epoch": 0.06785929648241205,
"grad_norm": 2.827131986618042,
"learning_rate": 1.2972054702225374e-05,
"loss": 7.037248611450195,
"step": 1688
},
{
"epoch": 0.06789949748743719,
"grad_norm": 2.2579269409179688,
"learning_rate": 1.2972000857369235e-05,
"loss": 7.3822126388549805,
"step": 1689
},
{
"epoch": 0.06793969849246231,
"grad_norm": 2.5922625064849854,
"learning_rate": 1.2971946960913431e-05,
"loss": 7.289929389953613,
"step": 1690
},
{
"epoch": 0.06797989949748744,
"grad_norm": 2.410400867462158,
"learning_rate": 1.2971893012858827e-05,
"loss": 7.617347240447998,
"step": 1691
},
{
"epoch": 0.06802010050251256,
"grad_norm": 2.3637986183166504,
"learning_rate": 1.2971839013206284e-05,
"loss": 7.177247047424316,
"step": 1692
},
{
"epoch": 0.06806030150753768,
"grad_norm": 2.701310873031616,
"learning_rate": 1.2971784961956666e-05,
"loss": 7.773275852203369,
"step": 1693
},
{
"epoch": 0.06810050251256282,
"grad_norm": 2.424778461456299,
"learning_rate": 1.2971730859110843e-05,
"loss": 7.447427749633789,
"step": 1694
},
{
"epoch": 0.06814070351758794,
"grad_norm": 2.842477321624756,
"learning_rate": 1.2971676704669678e-05,
"loss": 7.780895709991455,
"step": 1695
},
{
"epoch": 0.06818090452261306,
"grad_norm": 2.326549530029297,
"learning_rate": 1.2971622498634039e-05,
"loss": 7.547553062438965,
"step": 1696
},
{
"epoch": 0.06822110552763819,
"grad_norm": 2.4184978008270264,
"learning_rate": 1.2971568241004794e-05,
"loss": 7.692824840545654,
"step": 1697
},
{
"epoch": 0.06826130653266331,
"grad_norm": 2.425597906112671,
"learning_rate": 1.2971513931782814e-05,
"loss": 7.72849178314209,
"step": 1698
},
{
"epoch": 0.06830150753768845,
"grad_norm": 2.3491952419281006,
"learning_rate": 1.2971459570968966e-05,
"loss": 7.386116981506348,
"step": 1699
},
{
"epoch": 0.06834170854271357,
"grad_norm": 2.3457252979278564,
"learning_rate": 1.2971405158564122e-05,
"loss": 7.9003753662109375,
"step": 1700
},
{
"epoch": 0.06838190954773869,
"grad_norm": 2.379575729370117,
"learning_rate": 1.2971350694569152e-05,
"loss": 7.565673828125,
"step": 1701
},
{
"epoch": 0.06842211055276382,
"grad_norm": 2.3178606033325195,
"learning_rate": 1.297129617898493e-05,
"loss": 7.192347526550293,
"step": 1702
},
{
"epoch": 0.06846231155778894,
"grad_norm": 2.3498170375823975,
"learning_rate": 1.2971241611812329e-05,
"loss": 7.587038516998291,
"step": 1703
},
{
"epoch": 0.06850251256281407,
"grad_norm": 2.7386538982391357,
"learning_rate": 1.297118699305222e-05,
"loss": 7.236972808837891,
"step": 1704
},
{
"epoch": 0.0685427135678392,
"grad_norm": 2.3414714336395264,
"learning_rate": 1.2971132322705481e-05,
"loss": 8.199323654174805,
"step": 1705
},
{
"epoch": 0.06858291457286432,
"grad_norm": 2.31976056098938,
"learning_rate": 1.2971077600772986e-05,
"loss": 7.7953081130981445,
"step": 1706
},
{
"epoch": 0.06862311557788944,
"grad_norm": 2.3207592964172363,
"learning_rate": 1.2971022827255612e-05,
"loss": 7.665596961975098,
"step": 1707
},
{
"epoch": 0.06866331658291458,
"grad_norm": 2.702364683151245,
"learning_rate": 1.2970968002154236e-05,
"loss": 7.563024520874023,
"step": 1708
},
{
"epoch": 0.0687035175879397,
"grad_norm": 2.241574764251709,
"learning_rate": 1.2970913125469737e-05,
"loss": 7.533221244812012,
"step": 1709
},
{
"epoch": 0.06874371859296483,
"grad_norm": 2.364795684814453,
"learning_rate": 1.2970858197202992e-05,
"loss": 7.557620048522949,
"step": 1710
},
{
"epoch": 0.06878391959798995,
"grad_norm": 2.6431796550750732,
"learning_rate": 1.2970803217354881e-05,
"loss": 7.471245765686035,
"step": 1711
},
{
"epoch": 0.06882412060301507,
"grad_norm": 2.541903018951416,
"learning_rate": 1.2970748185926284e-05,
"loss": 7.068525791168213,
"step": 1712
},
{
"epoch": 0.06886432160804021,
"grad_norm": 2.21982479095459,
"learning_rate": 1.2970693102918086e-05,
"loss": 7.704656600952148,
"step": 1713
},
{
"epoch": 0.06890452261306533,
"grad_norm": 2.5482609272003174,
"learning_rate": 1.2970637968331164e-05,
"loss": 7.639047622680664,
"step": 1714
},
{
"epoch": 0.06894472361809045,
"grad_norm": 2.5519087314605713,
"learning_rate": 1.2970582782166405e-05,
"loss": 7.578342437744141,
"step": 1715
},
{
"epoch": 0.06898492462311558,
"grad_norm": 2.6619958877563477,
"learning_rate": 1.297052754442469e-05,
"loss": 7.686215877532959,
"step": 1716
},
{
"epoch": 0.0690251256281407,
"grad_norm": 2.489142894744873,
"learning_rate": 1.2970472255106907e-05,
"loss": 7.2841715812683105,
"step": 1717
},
{
"epoch": 0.06906532663316584,
"grad_norm": 2.588991403579712,
"learning_rate": 1.2970416914213935e-05,
"loss": 7.853064060211182,
"step": 1718
},
{
"epoch": 0.06910552763819096,
"grad_norm": 2.454378604888916,
"learning_rate": 1.297036152174667e-05,
"loss": 7.496637344360352,
"step": 1719
},
{
"epoch": 0.06914572864321608,
"grad_norm": 2.3877477645874023,
"learning_rate": 1.2970306077705988e-05,
"loss": 7.631082057952881,
"step": 1720
},
{
"epoch": 0.0691859296482412,
"grad_norm": 2.4061648845672607,
"learning_rate": 1.2970250582092786e-05,
"loss": 7.167636871337891,
"step": 1721
},
{
"epoch": 0.06922613065326633,
"grad_norm": 2.3376526832580566,
"learning_rate": 1.2970195034907949e-05,
"loss": 7.405013084411621,
"step": 1722
},
{
"epoch": 0.06926633165829146,
"grad_norm": 2.278083562850952,
"learning_rate": 1.2970139436152367e-05,
"loss": 7.155329704284668,
"step": 1723
},
{
"epoch": 0.06930653266331659,
"grad_norm": 2.4762954711914062,
"learning_rate": 1.2970083785826927e-05,
"loss": 7.402786731719971,
"step": 1724
},
{
"epoch": 0.06934673366834171,
"grad_norm": 2.344957113265991,
"learning_rate": 1.2970028083932528e-05,
"loss": 7.586745262145996,
"step": 1725
},
{
"epoch": 0.06938693467336683,
"grad_norm": 2.5444092750549316,
"learning_rate": 1.2969972330470055e-05,
"loss": 7.689425945281982,
"step": 1726
},
{
"epoch": 0.06942713567839195,
"grad_norm": 2.712427854537964,
"learning_rate": 1.2969916525440404e-05,
"loss": 7.376766204833984,
"step": 1727
},
{
"epoch": 0.06946733668341709,
"grad_norm": 2.5314555168151855,
"learning_rate": 1.2969860668844467e-05,
"loss": 7.953917503356934,
"step": 1728
},
{
"epoch": 0.06950753768844221,
"grad_norm": 2.5496325492858887,
"learning_rate": 1.2969804760683142e-05,
"loss": 7.4017791748046875,
"step": 1729
},
{
"epoch": 0.06954773869346734,
"grad_norm": 2.377302646636963,
"learning_rate": 1.2969748800957321e-05,
"loss": 7.528629302978516,
"step": 1730
},
{
"epoch": 0.06958793969849246,
"grad_norm": 2.449382781982422,
"learning_rate": 1.29696927896679e-05,
"loss": 7.367244720458984,
"step": 1731
},
{
"epoch": 0.06962814070351758,
"grad_norm": 2.459152936935425,
"learning_rate": 1.296963672681578e-05,
"loss": 7.328320503234863,
"step": 1732
},
{
"epoch": 0.06966834170854272,
"grad_norm": 2.261622667312622,
"learning_rate": 1.2969580612401853e-05,
"loss": 7.572084426879883,
"step": 1733
},
{
"epoch": 0.06970854271356784,
"grad_norm": 2.412297487258911,
"learning_rate": 1.2969524446427022e-05,
"loss": 7.391973495483398,
"step": 1734
},
{
"epoch": 0.06974874371859296,
"grad_norm": 2.7037482261657715,
"learning_rate": 1.2969468228892184e-05,
"loss": 7.1947221755981445,
"step": 1735
},
{
"epoch": 0.06978894472361809,
"grad_norm": 2.2567946910858154,
"learning_rate": 1.2969411959798244e-05,
"loss": 7.423087120056152,
"step": 1736
},
{
"epoch": 0.06982914572864321,
"grad_norm": 2.5661258697509766,
"learning_rate": 1.2969355639146097e-05,
"loss": 7.41227388381958,
"step": 1737
},
{
"epoch": 0.06986934673366835,
"grad_norm": 2.5550551414489746,
"learning_rate": 1.2969299266936648e-05,
"loss": 7.079274654388428,
"step": 1738
},
{
"epoch": 0.06990954773869347,
"grad_norm": 2.7758193016052246,
"learning_rate": 1.2969242843170798e-05,
"loss": 7.673117637634277,
"step": 1739
},
{
"epoch": 0.06994974874371859,
"grad_norm": 2.8083817958831787,
"learning_rate": 1.2969186367849454e-05,
"loss": 7.487375259399414,
"step": 1740
},
{
"epoch": 0.06998994974874372,
"grad_norm": 2.51660418510437,
"learning_rate": 1.296912984097352e-05,
"loss": 7.571894645690918,
"step": 1741
},
{
"epoch": 0.07003015075376884,
"grad_norm": 2.4268863201141357,
"learning_rate": 1.2969073262543895e-05,
"loss": 7.4615373611450195,
"step": 1742
},
{
"epoch": 0.07007035175879398,
"grad_norm": 2.404186487197876,
"learning_rate": 1.2969016632561492e-05,
"loss": 7.499594688415527,
"step": 1743
},
{
"epoch": 0.0701105527638191,
"grad_norm": 2.4495689868927,
"learning_rate": 1.2968959951027218e-05,
"loss": 7.480649948120117,
"step": 1744
},
{
"epoch": 0.07015075376884422,
"grad_norm": 2.7455639839172363,
"learning_rate": 1.2968903217941976e-05,
"loss": 7.235910415649414,
"step": 1745
},
{
"epoch": 0.07019095477386934,
"grad_norm": 2.570281505584717,
"learning_rate": 1.2968846433306676e-05,
"loss": 7.67621374130249,
"step": 1746
},
{
"epoch": 0.07023115577889447,
"grad_norm": 2.4988725185394287,
"learning_rate": 1.296878959712223e-05,
"loss": 7.186274528503418,
"step": 1747
},
{
"epoch": 0.0702713567839196,
"grad_norm": 2.504566192626953,
"learning_rate": 1.2968732709389549e-05,
"loss": 7.299044609069824,
"step": 1748
},
{
"epoch": 0.07031155778894473,
"grad_norm": 3.1491782665252686,
"learning_rate": 1.2968675770109538e-05,
"loss": 6.867884635925293,
"step": 1749
},
{
"epoch": 0.07035175879396985,
"grad_norm": 2.4730782508850098,
"learning_rate": 1.2968618779283114e-05,
"loss": 7.497553825378418,
"step": 1750
},
{
"epoch": 0.07039195979899497,
"grad_norm": 2.3481736183166504,
"learning_rate": 1.2968561736911188e-05,
"loss": 7.601728916168213,
"step": 1751
},
{
"epoch": 0.0704321608040201,
"grad_norm": 2.665764331817627,
"learning_rate": 1.2968504642994674e-05,
"loss": 6.859842300415039,
"step": 1752
},
{
"epoch": 0.07047236180904523,
"grad_norm": 2.515373706817627,
"learning_rate": 1.2968447497534486e-05,
"loss": 7.243340969085693,
"step": 1753
},
{
"epoch": 0.07051256281407035,
"grad_norm": 2.5149152278900146,
"learning_rate": 1.296839030053154e-05,
"loss": 7.246526718139648,
"step": 1754
},
{
"epoch": 0.07055276381909548,
"grad_norm": 2.3507726192474365,
"learning_rate": 1.2968333051986752e-05,
"loss": 7.474225997924805,
"step": 1755
},
{
"epoch": 0.0705929648241206,
"grad_norm": 2.709078073501587,
"learning_rate": 1.2968275751901036e-05,
"loss": 7.449416637420654,
"step": 1756
},
{
"epoch": 0.07063316582914572,
"grad_norm": 2.4861106872558594,
"learning_rate": 1.2968218400275312e-05,
"loss": 7.585716724395752,
"step": 1757
},
{
"epoch": 0.07067336683417086,
"grad_norm": 2.4464468955993652,
"learning_rate": 1.2968160997110498e-05,
"loss": 7.850563049316406,
"step": 1758
},
{
"epoch": 0.07071356783919598,
"grad_norm": 2.6670024394989014,
"learning_rate": 1.2968103542407514e-05,
"loss": 7.706972599029541,
"step": 1759
},
{
"epoch": 0.0707537688442211,
"grad_norm": 2.6844658851623535,
"learning_rate": 1.2968046036167281e-05,
"loss": 6.937670707702637,
"step": 1760
},
{
"epoch": 0.07079396984924623,
"grad_norm": 2.6143500804901123,
"learning_rate": 1.2967988478390718e-05,
"loss": 7.964158535003662,
"step": 1761
},
{
"epoch": 0.07083417085427135,
"grad_norm": 2.4126622676849365,
"learning_rate": 1.2967930869078747e-05,
"loss": 7.747876167297363,
"step": 1762
},
{
"epoch": 0.07087437185929649,
"grad_norm": 2.4184093475341797,
"learning_rate": 1.296787320823229e-05,
"loss": 7.250892639160156,
"step": 1763
},
{
"epoch": 0.07091457286432161,
"grad_norm": 2.442909002304077,
"learning_rate": 1.2967815495852273e-05,
"loss": 7.182138442993164,
"step": 1764
},
{
"epoch": 0.07095477386934673,
"grad_norm": 2.591291904449463,
"learning_rate": 1.2967757731939619e-05,
"loss": 7.685145378112793,
"step": 1765
},
{
"epoch": 0.07099497487437186,
"grad_norm": 3.6311428546905518,
"learning_rate": 1.2967699916495251e-05,
"loss": 7.503774166107178,
"step": 1766
},
{
"epoch": 0.07103517587939699,
"grad_norm": 2.685656785964966,
"learning_rate": 1.2967642049520098e-05,
"loss": 8.565290451049805,
"step": 1767
},
{
"epoch": 0.07107537688442211,
"grad_norm": 2.5214059352874756,
"learning_rate": 1.2967584131015084e-05,
"loss": 7.565145969390869,
"step": 1768
},
{
"epoch": 0.07111557788944724,
"grad_norm": 2.893547773361206,
"learning_rate": 1.296752616098114e-05,
"loss": 7.825536727905273,
"step": 1769
},
{
"epoch": 0.07115577889447236,
"grad_norm": 2.4903180599212646,
"learning_rate": 1.296746813941919e-05,
"loss": 7.64127254486084,
"step": 1770
},
{
"epoch": 0.07119597989949748,
"grad_norm": 2.5649869441986084,
"learning_rate": 1.2967410066330167e-05,
"loss": 7.347582817077637,
"step": 1771
},
{
"epoch": 0.07123618090452262,
"grad_norm": 2.342155933380127,
"learning_rate": 1.2967351941715e-05,
"loss": 7.433147430419922,
"step": 1772
},
{
"epoch": 0.07127638190954774,
"grad_norm": 2.4577670097351074,
"learning_rate": 1.296729376557462e-05,
"loss": 7.956034183502197,
"step": 1773
},
{
"epoch": 0.07131658291457287,
"grad_norm": 2.6594178676605225,
"learning_rate": 1.2967235537909956e-05,
"loss": 7.107813358306885,
"step": 1774
},
{
"epoch": 0.07135678391959799,
"grad_norm": 2.386221408843994,
"learning_rate": 1.2967177258721942e-05,
"loss": 8.185105323791504,
"step": 1775
},
{
"epoch": 0.07139698492462311,
"grad_norm": 2.4713757038116455,
"learning_rate": 1.2967118928011514e-05,
"loss": 7.620736122131348,
"step": 1776
},
{
"epoch": 0.07143718592964825,
"grad_norm": 2.8522775173187256,
"learning_rate": 1.2967060545779603e-05,
"loss": 7.11073112487793,
"step": 1777
},
{
"epoch": 0.07147738693467337,
"grad_norm": 2.7772135734558105,
"learning_rate": 1.2967002112027146e-05,
"loss": 7.60720682144165,
"step": 1778
},
{
"epoch": 0.0715175879396985,
"grad_norm": 2.542503595352173,
"learning_rate": 1.2966943626755077e-05,
"loss": 7.729147434234619,
"step": 1779
},
{
"epoch": 0.07155778894472362,
"grad_norm": 2.549440622329712,
"learning_rate": 1.2966885089964333e-05,
"loss": 7.67032527923584,
"step": 1780
},
{
"epoch": 0.07159798994974874,
"grad_norm": 2.3523898124694824,
"learning_rate": 1.2966826501655853e-05,
"loss": 7.5281877517700195,
"step": 1781
},
{
"epoch": 0.07163819095477388,
"grad_norm": 2.274359941482544,
"learning_rate": 1.2966767861830575e-05,
"loss": 7.356468200683594,
"step": 1782
},
{
"epoch": 0.071678391959799,
"grad_norm": 2.7017323970794678,
"learning_rate": 1.2966709170489435e-05,
"loss": 7.4844207763671875,
"step": 1783
},
{
"epoch": 0.07171859296482412,
"grad_norm": 2.3016934394836426,
"learning_rate": 1.2966650427633376e-05,
"loss": 8.101287841796875,
"step": 1784
},
{
"epoch": 0.07175879396984924,
"grad_norm": 2.587669610977173,
"learning_rate": 1.2966591633263338e-05,
"loss": 7.285143852233887,
"step": 1785
},
{
"epoch": 0.07179899497487437,
"grad_norm": 2.937546491622925,
"learning_rate": 1.2966532787380263e-05,
"loss": 7.410645484924316,
"step": 1786
},
{
"epoch": 0.0718391959798995,
"grad_norm": 2.4937899112701416,
"learning_rate": 1.2966473889985093e-05,
"loss": 7.617844581604004,
"step": 1787
},
{
"epoch": 0.07187939698492463,
"grad_norm": 2.229637861251831,
"learning_rate": 1.2966414941078771e-05,
"loss": 7.610365390777588,
"step": 1788
},
{
"epoch": 0.07191959798994975,
"grad_norm": 2.2410638332366943,
"learning_rate": 1.2966355940662242e-05,
"loss": 7.070990562438965,
"step": 1789
},
{
"epoch": 0.07195979899497487,
"grad_norm": 2.591357707977295,
"learning_rate": 1.296629688873645e-05,
"loss": 7.45359992980957,
"step": 1790
},
{
"epoch": 0.072,
"grad_norm": 2.589934825897217,
"learning_rate": 1.296623778530234e-05,
"loss": 7.3538055419921875,
"step": 1791
},
{
"epoch": 0.07204020100502513,
"grad_norm": 2.5079522132873535,
"learning_rate": 1.296617863036086e-05,
"loss": 7.729067802429199,
"step": 1792
},
{
"epoch": 0.07208040201005025,
"grad_norm": 2.6271963119506836,
"learning_rate": 1.2966119423912956e-05,
"loss": 7.641167640686035,
"step": 1793
},
{
"epoch": 0.07212060301507538,
"grad_norm": 2.7618892192840576,
"learning_rate": 1.2966060165959577e-05,
"loss": 7.151963710784912,
"step": 1794
},
{
"epoch": 0.0721608040201005,
"grad_norm": 2.601973056793213,
"learning_rate": 1.2966000856501674e-05,
"loss": 7.284404277801514,
"step": 1795
},
{
"epoch": 0.07220100502512562,
"grad_norm": 2.5512607097625732,
"learning_rate": 1.2965941495540193e-05,
"loss": 7.294304847717285,
"step": 1796
},
{
"epoch": 0.07224120603015076,
"grad_norm": 2.5563571453094482,
"learning_rate": 1.2965882083076087e-05,
"loss": 7.84138822555542,
"step": 1797
},
{
"epoch": 0.07228140703517588,
"grad_norm": 2.62146258354187,
"learning_rate": 1.2965822619110305e-05,
"loss": 7.304176330566406,
"step": 1798
},
{
"epoch": 0.072321608040201,
"grad_norm": 6.729279041290283,
"learning_rate": 1.2965763103643803e-05,
"loss": 8.481095314025879,
"step": 1799
},
{
"epoch": 0.07236180904522613,
"grad_norm": 2.5956692695617676,
"learning_rate": 1.2965703536677531e-05,
"loss": 7.829920768737793,
"step": 1800
},
{
"epoch": 0.07236180904522613,
"eval_loss": 7.489296913146973,
"eval_runtime": 120.7571,
"eval_samples_per_second": 4.141,
"eval_steps_per_second": 0.522,
"step": 1800
},
{
"epoch": 0.07240201005025125,
"grad_norm": 2.806140661239624,
"learning_rate": 1.2965643918212447e-05,
"loss": 7.427911758422852,
"step": 1801
},
{
"epoch": 0.07244221105527639,
"grad_norm": 2.7946574687957764,
"learning_rate": 1.29655842482495e-05,
"loss": 6.926874160766602,
"step": 1802
},
{
"epoch": 0.07248241206030151,
"grad_norm": 2.700192928314209,
"learning_rate": 1.296552452678965e-05,
"loss": 7.363410949707031,
"step": 1803
},
{
"epoch": 0.07252261306532663,
"grad_norm": 2.7369861602783203,
"learning_rate": 1.296546475383385e-05,
"loss": 7.686553001403809,
"step": 1804
},
{
"epoch": 0.07256281407035176,
"grad_norm": 2.568132162094116,
"learning_rate": 1.2965404929383063e-05,
"loss": 7.707430839538574,
"step": 1805
},
{
"epoch": 0.07260301507537688,
"grad_norm": 2.6596596240997314,
"learning_rate": 1.2965345053438242e-05,
"loss": 7.247353553771973,
"step": 1806
},
{
"epoch": 0.07264321608040201,
"grad_norm": 2.620521068572998,
"learning_rate": 1.2965285126000347e-05,
"loss": 7.996770858764648,
"step": 1807
},
{
"epoch": 0.07268341708542714,
"grad_norm": 2.646226167678833,
"learning_rate": 1.2965225147070337e-05,
"loss": 7.557909965515137,
"step": 1808
},
{
"epoch": 0.07272361809045226,
"grad_norm": 2.5108723640441895,
"learning_rate": 1.2965165116649175e-05,
"loss": 7.297337532043457,
"step": 1809
},
{
"epoch": 0.07276381909547738,
"grad_norm": 2.5276386737823486,
"learning_rate": 1.296510503473782e-05,
"loss": 7.354154586791992,
"step": 1810
},
{
"epoch": 0.0728040201005025,
"grad_norm": 2.6859493255615234,
"learning_rate": 1.2965044901337235e-05,
"loss": 7.460947036743164,
"step": 1811
},
{
"epoch": 0.07284422110552764,
"grad_norm": 3.2525830268859863,
"learning_rate": 1.2964984716448385e-05,
"loss": 7.517999649047852,
"step": 1812
},
{
"epoch": 0.07288442211055277,
"grad_norm": 2.4359548091888428,
"learning_rate": 1.2964924480072229e-05,
"loss": 7.165524482727051,
"step": 1813
},
{
"epoch": 0.07292462311557789,
"grad_norm": 2.950594902038574,
"learning_rate": 1.2964864192209736e-05,
"loss": 7.2011308670043945,
"step": 1814
},
{
"epoch": 0.07296482412060301,
"grad_norm": 3.06392765045166,
"learning_rate": 1.2964803852861872e-05,
"loss": 7.736235618591309,
"step": 1815
},
{
"epoch": 0.07300502512562813,
"grad_norm": 2.6311495304107666,
"learning_rate": 1.29647434620296e-05,
"loss": 7.480612754821777,
"step": 1816
},
{
"epoch": 0.07304522613065327,
"grad_norm": 2.800086498260498,
"learning_rate": 1.2964683019713889e-05,
"loss": 7.144618034362793,
"step": 1817
},
{
"epoch": 0.0730854271356784,
"grad_norm": 2.9738264083862305,
"learning_rate": 1.2964622525915706e-05,
"loss": 7.629509925842285,
"step": 1818
},
{
"epoch": 0.07312562814070352,
"grad_norm": 2.8041293621063232,
"learning_rate": 1.2964561980636021e-05,
"loss": 7.678937911987305,
"step": 1819
},
{
"epoch": 0.07316582914572864,
"grad_norm": 2.500554323196411,
"learning_rate": 1.2964501383875801e-05,
"loss": 8.094929695129395,
"step": 1820
},
{
"epoch": 0.07320603015075376,
"grad_norm": 2.6574556827545166,
"learning_rate": 1.2964440735636023e-05,
"loss": 7.0421552658081055,
"step": 1821
},
{
"epoch": 0.0732462311557789,
"grad_norm": 2.7500946521759033,
"learning_rate": 1.2964380035917649e-05,
"loss": 7.542429447174072,
"step": 1822
},
{
"epoch": 0.07328643216080402,
"grad_norm": 2.4673917293548584,
"learning_rate": 1.2964319284721659e-05,
"loss": 7.450387477874756,
"step": 1823
},
{
"epoch": 0.07332663316582914,
"grad_norm": 2.589442729949951,
"learning_rate": 1.296425848204902e-05,
"loss": 7.4919538497924805,
"step": 1824
},
{
"epoch": 0.07336683417085427,
"grad_norm": 2.8437373638153076,
"learning_rate": 1.2964197627900711e-05,
"loss": 7.911367416381836,
"step": 1825
},
{
"epoch": 0.0734070351758794,
"grad_norm": 2.2245726585388184,
"learning_rate": 1.2964136722277703e-05,
"loss": 7.355844497680664,
"step": 1826
},
{
"epoch": 0.07344723618090453,
"grad_norm": 2.7195355892181396,
"learning_rate": 1.2964075765180974e-05,
"loss": 7.496330261230469,
"step": 1827
},
{
"epoch": 0.07348743718592965,
"grad_norm": 2.666250705718994,
"learning_rate": 1.2964014756611497e-05,
"loss": 7.317912578582764,
"step": 1828
},
{
"epoch": 0.07352763819095477,
"grad_norm": 2.7590014934539795,
"learning_rate": 1.2963953696570251e-05,
"loss": 7.326004981994629,
"step": 1829
},
{
"epoch": 0.0735678391959799,
"grad_norm": 2.5098283290863037,
"learning_rate": 1.2963892585058213e-05,
"loss": 7.390745639801025,
"step": 1830
},
{
"epoch": 0.07360804020100503,
"grad_norm": 2.415060520172119,
"learning_rate": 1.2963831422076363e-05,
"loss": 7.804728031158447,
"step": 1831
},
{
"epoch": 0.07364824120603015,
"grad_norm": 2.6229002475738525,
"learning_rate": 1.296377020762568e-05,
"loss": 7.388227462768555,
"step": 1832
},
{
"epoch": 0.07368844221105528,
"grad_norm": 2.4245963096618652,
"learning_rate": 1.2963708941707144e-05,
"loss": 7.453157424926758,
"step": 1833
},
{
"epoch": 0.0737286432160804,
"grad_norm": 2.623446226119995,
"learning_rate": 1.2963647624321737e-05,
"loss": 7.305220603942871,
"step": 1834
},
{
"epoch": 0.07376884422110552,
"grad_norm": 2.432162284851074,
"learning_rate": 1.2963586255470438e-05,
"loss": 7.823619842529297,
"step": 1835
},
{
"epoch": 0.07380904522613066,
"grad_norm": 2.3717384338378906,
"learning_rate": 1.2963524835154235e-05,
"loss": 7.410618782043457,
"step": 1836
},
{
"epoch": 0.07384924623115578,
"grad_norm": 2.5714616775512695,
"learning_rate": 1.2963463363374106e-05,
"loss": 7.424695014953613,
"step": 1837
},
{
"epoch": 0.0738894472361809,
"grad_norm": 2.630495071411133,
"learning_rate": 1.296340184013104e-05,
"loss": 7.426639556884766,
"step": 1838
},
{
"epoch": 0.07392964824120603,
"grad_norm": 2.3752012252807617,
"learning_rate": 1.296334026542602e-05,
"loss": 7.166518688201904,
"step": 1839
},
{
"epoch": 0.07396984924623115,
"grad_norm": 2.472200393676758,
"learning_rate": 1.296327863926003e-05,
"loss": 7.545009613037109,
"step": 1840
},
{
"epoch": 0.07401005025125629,
"grad_norm": 2.5112085342407227,
"learning_rate": 1.2963216961634062e-05,
"loss": 7.716034889221191,
"step": 1841
},
{
"epoch": 0.07405025125628141,
"grad_norm": 2.530616283416748,
"learning_rate": 1.2963155232549101e-05,
"loss": 7.748445510864258,
"step": 1842
},
{
"epoch": 0.07409045226130653,
"grad_norm": 2.437825918197632,
"learning_rate": 1.2963093452006136e-05,
"loss": 7.379720687866211,
"step": 1843
},
{
"epoch": 0.07413065326633166,
"grad_norm": 2.582291841506958,
"learning_rate": 1.2963031620006156e-05,
"loss": 7.6890411376953125,
"step": 1844
},
{
"epoch": 0.07417085427135678,
"grad_norm": 2.6147234439849854,
"learning_rate": 1.2962969736550154e-05,
"loss": 7.577750205993652,
"step": 1845
},
{
"epoch": 0.07421105527638192,
"grad_norm": 2.6098952293395996,
"learning_rate": 1.2962907801639115e-05,
"loss": 7.4566755294799805,
"step": 1846
},
{
"epoch": 0.07425125628140704,
"grad_norm": 2.5403974056243896,
"learning_rate": 1.2962845815274033e-05,
"loss": 7.464817047119141,
"step": 1847
},
{
"epoch": 0.07429145728643216,
"grad_norm": 2.483200788497925,
"learning_rate": 1.2962783777455907e-05,
"loss": 7.528248310089111,
"step": 1848
},
{
"epoch": 0.07433165829145728,
"grad_norm": 2.317044973373413,
"learning_rate": 1.296272168818572e-05,
"loss": 7.390707492828369,
"step": 1849
},
{
"epoch": 0.0743718592964824,
"grad_norm": 2.550947427749634,
"learning_rate": 1.2962659547464477e-05,
"loss": 7.006587505340576,
"step": 1850
},
{
"epoch": 0.07441206030150754,
"grad_norm": 2.2326207160949707,
"learning_rate": 1.2962597355293166e-05,
"loss": 7.39548921585083,
"step": 1851
},
{
"epoch": 0.07445226130653267,
"grad_norm": 3.0034141540527344,
"learning_rate": 1.2962535111672784e-05,
"loss": 7.457711219787598,
"step": 1852
},
{
"epoch": 0.07449246231155779,
"grad_norm": 2.545107364654541,
"learning_rate": 1.2962472816604329e-05,
"loss": 7.400099277496338,
"step": 1853
},
{
"epoch": 0.07453266331658291,
"grad_norm": 2.4187443256378174,
"learning_rate": 1.29624104700888e-05,
"loss": 7.470830917358398,
"step": 1854
},
{
"epoch": 0.07457286432160803,
"grad_norm": 2.3492307662963867,
"learning_rate": 1.2962348072127194e-05,
"loss": 7.529654502868652,
"step": 1855
},
{
"epoch": 0.07461306532663317,
"grad_norm": 2.5346667766571045,
"learning_rate": 1.296228562272051e-05,
"loss": 7.228287220001221,
"step": 1856
},
{
"epoch": 0.0746532663316583,
"grad_norm": 2.379593849182129,
"learning_rate": 1.296222312186975e-05,
"loss": 7.823828220367432,
"step": 1857
},
{
"epoch": 0.07469346733668342,
"grad_norm": 2.425527572631836,
"learning_rate": 1.2962160569575911e-05,
"loss": 7.524262428283691,
"step": 1858
},
{
"epoch": 0.07473366834170854,
"grad_norm": 2.7634763717651367,
"learning_rate": 1.2962097965839998e-05,
"loss": 7.635905742645264,
"step": 1859
},
{
"epoch": 0.07477386934673366,
"grad_norm": 2.600795030593872,
"learning_rate": 1.2962035310663013e-05,
"loss": 7.344997406005859,
"step": 1860
},
{
"epoch": 0.0748140703517588,
"grad_norm": 2.4198191165924072,
"learning_rate": 1.2961972604045958e-05,
"loss": 7.552128791809082,
"step": 1861
},
{
"epoch": 0.07485427135678392,
"grad_norm": 2.490499496459961,
"learning_rate": 1.2961909845989838e-05,
"loss": 7.374704360961914,
"step": 1862
},
{
"epoch": 0.07489447236180904,
"grad_norm": 2.6463136672973633,
"learning_rate": 1.296184703649566e-05,
"loss": 7.400367736816406,
"step": 1863
},
{
"epoch": 0.07493467336683417,
"grad_norm": 2.711282253265381,
"learning_rate": 1.2961784175564427e-05,
"loss": 7.310647010803223,
"step": 1864
},
{
"epoch": 0.07497487437185929,
"grad_norm": 2.5253348350524902,
"learning_rate": 1.2961721263197147e-05,
"loss": 7.577625274658203,
"step": 1865
},
{
"epoch": 0.07501507537688443,
"grad_norm": 2.4971325397491455,
"learning_rate": 1.296165829939483e-05,
"loss": 7.582148551940918,
"step": 1866
},
{
"epoch": 0.07505527638190955,
"grad_norm": 2.5423545837402344,
"learning_rate": 1.296159528415848e-05,
"loss": 7.2520647048950195,
"step": 1867
},
{
"epoch": 0.07509547738693467,
"grad_norm": 2.9011974334716797,
"learning_rate": 1.2961532217489108e-05,
"loss": 7.605786323547363,
"step": 1868
},
{
"epoch": 0.0751356783919598,
"grad_norm": 2.6415867805480957,
"learning_rate": 1.2961469099387725e-05,
"loss": 7.575553894042969,
"step": 1869
},
{
"epoch": 0.07517587939698492,
"grad_norm": 3.1488184928894043,
"learning_rate": 1.296140592985534e-05,
"loss": 6.862781524658203,
"step": 1870
},
{
"epoch": 0.07521608040201005,
"grad_norm": 2.7982370853424072,
"learning_rate": 1.2961342708892965e-05,
"loss": 7.53342342376709,
"step": 1871
},
{
"epoch": 0.07525628140703518,
"grad_norm": 2.7477684020996094,
"learning_rate": 1.2961279436501614e-05,
"loss": 7.886588096618652,
"step": 1872
},
{
"epoch": 0.0752964824120603,
"grad_norm": 3.3322503566741943,
"learning_rate": 1.2961216112682298e-05,
"loss": 6.793211936950684,
"step": 1873
},
{
"epoch": 0.07533668341708542,
"grad_norm": 2.559913158416748,
"learning_rate": 1.2961152737436034e-05,
"loss": 7.793353080749512,
"step": 1874
},
{
"epoch": 0.07537688442211055,
"grad_norm": 2.6595065593719482,
"learning_rate": 1.2961089310763836e-05,
"loss": 7.176842212677002,
"step": 1875
},
{
"epoch": 0.07541708542713568,
"grad_norm": 2.7309718132019043,
"learning_rate": 1.2961025832666718e-05,
"loss": 7.2764892578125,
"step": 1876
},
{
"epoch": 0.0754572864321608,
"grad_norm": 2.8949270248413086,
"learning_rate": 1.2960962303145697e-05,
"loss": 7.624156951904297,
"step": 1877
},
{
"epoch": 0.07549748743718593,
"grad_norm": 2.8363208770751953,
"learning_rate": 1.296089872220179e-05,
"loss": 7.589741230010986,
"step": 1878
},
{
"epoch": 0.07553768844221105,
"grad_norm": 2.6628570556640625,
"learning_rate": 1.296083508983602e-05,
"loss": 7.580198287963867,
"step": 1879
},
{
"epoch": 0.07557788944723617,
"grad_norm": 2.596743106842041,
"learning_rate": 1.29607714060494e-05,
"loss": 7.57151460647583,
"step": 1880
},
{
"epoch": 0.07561809045226131,
"grad_norm": 2.8270037174224854,
"learning_rate": 1.2960707670842952e-05,
"loss": 7.714536666870117,
"step": 1881
},
{
"epoch": 0.07565829145728643,
"grad_norm": 2.3916454315185547,
"learning_rate": 1.2960643884217699e-05,
"loss": 7.586742401123047,
"step": 1882
},
{
"epoch": 0.07569849246231156,
"grad_norm": 2.552382230758667,
"learning_rate": 1.2960580046174659e-05,
"loss": 7.677306175231934,
"step": 1883
},
{
"epoch": 0.07573869346733668,
"grad_norm": 2.718071699142456,
"learning_rate": 1.2960516156714856e-05,
"loss": 7.678471565246582,
"step": 1884
},
{
"epoch": 0.07577889447236182,
"grad_norm": 2.59429669380188,
"learning_rate": 1.2960452215839312e-05,
"loss": 7.321861743927002,
"step": 1885
},
{
"epoch": 0.07581909547738694,
"grad_norm": 2.445835828781128,
"learning_rate": 1.2960388223549053e-05,
"loss": 7.158745765686035,
"step": 1886
},
{
"epoch": 0.07585929648241206,
"grad_norm": 2.4607372283935547,
"learning_rate": 1.2960324179845104e-05,
"loss": 7.593021392822266,
"step": 1887
},
{
"epoch": 0.07589949748743718,
"grad_norm": 2.75640606880188,
"learning_rate": 1.2960260084728488e-05,
"loss": 7.293218612670898,
"step": 1888
},
{
"epoch": 0.07593969849246231,
"grad_norm": 2.710279941558838,
"learning_rate": 1.2960195938200233e-05,
"loss": 7.489206314086914,
"step": 1889
},
{
"epoch": 0.07597989949748744,
"grad_norm": 2.8775947093963623,
"learning_rate": 1.2960131740261367e-05,
"loss": 7.587553024291992,
"step": 1890
},
{
"epoch": 0.07602010050251257,
"grad_norm": 2.3761661052703857,
"learning_rate": 1.2960067490912918e-05,
"loss": 7.548840522766113,
"step": 1891
},
{
"epoch": 0.07606030150753769,
"grad_norm": 2.4218804836273193,
"learning_rate": 1.2960003190155913e-05,
"loss": 7.464049816131592,
"step": 1892
},
{
"epoch": 0.07610050251256281,
"grad_norm": 2.462796211242676,
"learning_rate": 1.2959938837991384e-05,
"loss": 7.206564903259277,
"step": 1893
},
{
"epoch": 0.07614070351758793,
"grad_norm": 2.6556448936462402,
"learning_rate": 1.2959874434420361e-05,
"loss": 7.748886585235596,
"step": 1894
},
{
"epoch": 0.07618090452261307,
"grad_norm": 2.547847032546997,
"learning_rate": 1.2959809979443876e-05,
"loss": 7.593013286590576,
"step": 1895
},
{
"epoch": 0.0762211055276382,
"grad_norm": 2.4562644958496094,
"learning_rate": 1.2959745473062958e-05,
"loss": 7.7332258224487305,
"step": 1896
},
{
"epoch": 0.07626130653266332,
"grad_norm": 2.6514732837677,
"learning_rate": 1.2959680915278643e-05,
"loss": 7.608262062072754,
"step": 1897
},
{
"epoch": 0.07630150753768844,
"grad_norm": 2.594515085220337,
"learning_rate": 1.2959616306091965e-05,
"loss": 7.169390678405762,
"step": 1898
},
{
"epoch": 0.07634170854271356,
"grad_norm": 2.9548070430755615,
"learning_rate": 1.2959551645503956e-05,
"loss": 7.4984564781188965,
"step": 1899
},
{
"epoch": 0.0763819095477387,
"grad_norm": 2.45829701423645,
"learning_rate": 1.2959486933515657e-05,
"loss": 7.418169975280762,
"step": 1900
},
{
"epoch": 0.07642211055276382,
"grad_norm": 2.468304395675659,
"learning_rate": 1.2959422170128097e-05,
"loss": 7.3704986572265625,
"step": 1901
},
{
"epoch": 0.07646231155778895,
"grad_norm": 2.7277302742004395,
"learning_rate": 1.295935735534232e-05,
"loss": 7.6911773681640625,
"step": 1902
},
{
"epoch": 0.07650251256281407,
"grad_norm": 2.8297059535980225,
"learning_rate": 1.295929248915936e-05,
"loss": 7.573190689086914,
"step": 1903
},
{
"epoch": 0.07654271356783919,
"grad_norm": 2.387594223022461,
"learning_rate": 1.2959227571580255e-05,
"loss": 7.474340915679932,
"step": 1904
},
{
"epoch": 0.07658291457286433,
"grad_norm": 2.4525699615478516,
"learning_rate": 1.295916260260605e-05,
"loss": 7.228392601013184,
"step": 1905
},
{
"epoch": 0.07662311557788945,
"grad_norm": 2.995487928390503,
"learning_rate": 1.2959097582237779e-05,
"loss": 7.112286567687988,
"step": 1906
},
{
"epoch": 0.07666331658291457,
"grad_norm": 2.582347869873047,
"learning_rate": 1.295903251047649e-05,
"loss": 7.105002403259277,
"step": 1907
},
{
"epoch": 0.0767035175879397,
"grad_norm": 2.957561731338501,
"learning_rate": 1.2958967387323217e-05,
"loss": 7.085620880126953,
"step": 1908
},
{
"epoch": 0.07674371859296482,
"grad_norm": 2.5480661392211914,
"learning_rate": 1.2958902212779009e-05,
"loss": 7.508879661560059,
"step": 1909
},
{
"epoch": 0.07678391959798996,
"grad_norm": 3.0694162845611572,
"learning_rate": 1.2958836986844907e-05,
"loss": 7.01461124420166,
"step": 1910
},
{
"epoch": 0.07682412060301508,
"grad_norm": 2.7897400856018066,
"learning_rate": 1.2958771709521959e-05,
"loss": 7.601027488708496,
"step": 1911
},
{
"epoch": 0.0768643216080402,
"grad_norm": 2.4619154930114746,
"learning_rate": 1.2958706380811207e-05,
"loss": 7.524443626403809,
"step": 1912
},
{
"epoch": 0.07690452261306532,
"grad_norm": 3.3039391040802,
"learning_rate": 1.2958641000713697e-05,
"loss": 7.474581241607666,
"step": 1913
},
{
"epoch": 0.07694472361809045,
"grad_norm": 2.5312108993530273,
"learning_rate": 1.2958575569230477e-05,
"loss": 7.731194496154785,
"step": 1914
},
{
"epoch": 0.07698492462311558,
"grad_norm": 2.479567527770996,
"learning_rate": 1.2958510086362597e-05,
"loss": 7.247270584106445,
"step": 1915
},
{
"epoch": 0.0770251256281407,
"grad_norm": 2.845895767211914,
"learning_rate": 1.2958444552111103e-05,
"loss": 7.111506938934326,
"step": 1916
},
{
"epoch": 0.07706532663316583,
"grad_norm": 2.6125991344451904,
"learning_rate": 1.2958378966477045e-05,
"loss": 7.509527206420898,
"step": 1917
},
{
"epoch": 0.07710552763819095,
"grad_norm": 3.1783223152160645,
"learning_rate": 1.2958313329461474e-05,
"loss": 7.266271114349365,
"step": 1918
},
{
"epoch": 0.07714572864321607,
"grad_norm": 2.56689715385437,
"learning_rate": 1.2958247641065442e-05,
"loss": 7.233404159545898,
"step": 1919
},
{
"epoch": 0.07718592964824121,
"grad_norm": 2.5054428577423096,
"learning_rate": 1.2958181901289999e-05,
"loss": 7.443351745605469,
"step": 1920
},
{
"epoch": 0.07722613065326633,
"grad_norm": 2.3617475032806396,
"learning_rate": 1.29581161101362e-05,
"loss": 7.345563888549805,
"step": 1921
},
{
"epoch": 0.07726633165829146,
"grad_norm": 2.632382392883301,
"learning_rate": 1.2958050267605095e-05,
"loss": 7.338277816772461,
"step": 1922
},
{
"epoch": 0.07730653266331658,
"grad_norm": 2.7110695838928223,
"learning_rate": 1.2957984373697741e-05,
"loss": 7.13962459564209,
"step": 1923
},
{
"epoch": 0.0773467336683417,
"grad_norm": 2.582991600036621,
"learning_rate": 1.2957918428415195e-05,
"loss": 7.033593654632568,
"step": 1924
},
{
"epoch": 0.07738693467336684,
"grad_norm": 2.5650103092193604,
"learning_rate": 1.295785243175851e-05,
"loss": 7.316091537475586,
"step": 1925
},
{
"epoch": 0.07742713567839196,
"grad_norm": 3.1344692707061768,
"learning_rate": 1.2957786383728742e-05,
"loss": 7.208890438079834,
"step": 1926
},
{
"epoch": 0.07746733668341708,
"grad_norm": 2.4760971069335938,
"learning_rate": 1.2957720284326954e-05,
"loss": 6.989075183868408,
"step": 1927
},
{
"epoch": 0.07750753768844221,
"grad_norm": 2.6130192279815674,
"learning_rate": 1.29576541335542e-05,
"loss": 7.436296463012695,
"step": 1928
},
{
"epoch": 0.07754773869346733,
"grad_norm": 2.520843267440796,
"learning_rate": 1.2957587931411541e-05,
"loss": 7.132768154144287,
"step": 1929
},
{
"epoch": 0.07758793969849247,
"grad_norm": 3.0129952430725098,
"learning_rate": 1.2957521677900036e-05,
"loss": 7.8293304443359375,
"step": 1930
},
{
"epoch": 0.07762814070351759,
"grad_norm": 2.3739173412323,
"learning_rate": 1.2957455373020748e-05,
"loss": 7.447340965270996,
"step": 1931
},
{
"epoch": 0.07766834170854271,
"grad_norm": 2.536100387573242,
"learning_rate": 1.2957389016774738e-05,
"loss": 7.969007968902588,
"step": 1932
},
{
"epoch": 0.07770854271356784,
"grad_norm": 2.3683676719665527,
"learning_rate": 1.2957322609163067e-05,
"loss": 7.523131370544434,
"step": 1933
},
{
"epoch": 0.07774874371859296,
"grad_norm": 2.6081550121307373,
"learning_rate": 1.2957256150186802e-05,
"loss": 7.510747909545898,
"step": 1934
},
{
"epoch": 0.0777889447236181,
"grad_norm": 2.6090643405914307,
"learning_rate": 1.2957189639847005e-05,
"loss": 7.49398136138916,
"step": 1935
},
{
"epoch": 0.07782914572864322,
"grad_norm": 2.6284728050231934,
"learning_rate": 1.295712307814474e-05,
"loss": 7.310705661773682,
"step": 1936
},
{
"epoch": 0.07786934673366834,
"grad_norm": 2.5445375442504883,
"learning_rate": 1.2957056465081076e-05,
"loss": 7.7507758140563965,
"step": 1937
},
{
"epoch": 0.07790954773869346,
"grad_norm": 2.606973171234131,
"learning_rate": 1.2956989800657078e-05,
"loss": 7.830373764038086,
"step": 1938
},
{
"epoch": 0.0779497487437186,
"grad_norm": 2.8983407020568848,
"learning_rate": 1.2956923084873813e-05,
"loss": 7.555234432220459,
"step": 1939
},
{
"epoch": 0.07798994974874372,
"grad_norm": 2.4723012447357178,
"learning_rate": 1.2956856317732353e-05,
"loss": 7.741189002990723,
"step": 1940
},
{
"epoch": 0.07803015075376885,
"grad_norm": 2.487931251525879,
"learning_rate": 1.2956789499233761e-05,
"loss": 7.579427719116211,
"step": 1941
},
{
"epoch": 0.07807035175879397,
"grad_norm": 2.4618613719940186,
"learning_rate": 1.2956722629379115e-05,
"loss": 7.427656173706055,
"step": 1942
},
{
"epoch": 0.07811055276381909,
"grad_norm": 2.630620002746582,
"learning_rate": 1.2956655708169478e-05,
"loss": 7.076051712036133,
"step": 1943
},
{
"epoch": 0.07815075376884423,
"grad_norm": 2.8415143489837646,
"learning_rate": 1.2956588735605925e-05,
"loss": 7.345313549041748,
"step": 1944
},
{
"epoch": 0.07819095477386935,
"grad_norm": 2.565120220184326,
"learning_rate": 1.295652171168953e-05,
"loss": 7.205952167510986,
"step": 1945
},
{
"epoch": 0.07823115577889447,
"grad_norm": 3.0639524459838867,
"learning_rate": 1.2956454636421367e-05,
"loss": 7.514786720275879,
"step": 1946
},
{
"epoch": 0.0782713567839196,
"grad_norm": 2.890232801437378,
"learning_rate": 1.2956387509802505e-05,
"loss": 7.238358020782471,
"step": 1947
},
{
"epoch": 0.07831155778894472,
"grad_norm": 2.816948890686035,
"learning_rate": 1.2956320331834027e-05,
"loss": 7.5996856689453125,
"step": 1948
},
{
"epoch": 0.07835175879396986,
"grad_norm": 2.7962396144866943,
"learning_rate": 1.2956253102517e-05,
"loss": 7.507267951965332,
"step": 1949
},
{
"epoch": 0.07839195979899498,
"grad_norm": 2.92134165763855,
"learning_rate": 1.2956185821852506e-05,
"loss": 7.213926315307617,
"step": 1950
},
{
"epoch": 0.0784321608040201,
"grad_norm": 3.395712375640869,
"learning_rate": 1.2956118489841622e-05,
"loss": 7.397726058959961,
"step": 1951
},
{
"epoch": 0.07847236180904522,
"grad_norm": 2.915710926055908,
"learning_rate": 1.2956051106485428e-05,
"loss": 7.312784671783447,
"step": 1952
},
{
"epoch": 0.07851256281407035,
"grad_norm": 2.633378028869629,
"learning_rate": 1.2955983671785e-05,
"loss": 7.486483573913574,
"step": 1953
},
{
"epoch": 0.07855276381909548,
"grad_norm": 2.384077787399292,
"learning_rate": 1.2955916185741416e-05,
"loss": 7.164416313171387,
"step": 1954
},
{
"epoch": 0.0785929648241206,
"grad_norm": 2.3827309608459473,
"learning_rate": 1.295584864835576e-05,
"loss": 7.637596607208252,
"step": 1955
},
{
"epoch": 0.07863316582914573,
"grad_norm": 2.5635182857513428,
"learning_rate": 1.2955781059629116e-05,
"loss": 7.503772258758545,
"step": 1956
},
{
"epoch": 0.07867336683417085,
"grad_norm": 2.475661039352417,
"learning_rate": 1.295571341956256e-05,
"loss": 7.440558433532715,
"step": 1957
},
{
"epoch": 0.07871356783919597,
"grad_norm": 2.6767425537109375,
"learning_rate": 1.2955645728157184e-05,
"loss": 7.484677314758301,
"step": 1958
},
{
"epoch": 0.07875376884422111,
"grad_norm": 2.7796430587768555,
"learning_rate": 1.2955577985414062e-05,
"loss": 7.496875762939453,
"step": 1959
},
{
"epoch": 0.07879396984924623,
"grad_norm": 2.864870309829712,
"learning_rate": 1.2955510191334286e-05,
"loss": 7.780094623565674,
"step": 1960
},
{
"epoch": 0.07883417085427136,
"grad_norm": 2.670018434524536,
"learning_rate": 1.295544234591894e-05,
"loss": 7.215871810913086,
"step": 1961
},
{
"epoch": 0.07887437185929648,
"grad_norm": 2.609144449234009,
"learning_rate": 1.2955374449169109e-05,
"loss": 7.307845592498779,
"step": 1962
},
{
"epoch": 0.0789145728643216,
"grad_norm": 9.853724479675293,
"learning_rate": 1.2955306501085882e-05,
"loss": 7.346159934997559,
"step": 1963
},
{
"epoch": 0.07895477386934674,
"grad_norm": 2.5026280879974365,
"learning_rate": 1.2955238501670347e-05,
"loss": 7.4882731437683105,
"step": 1964
},
{
"epoch": 0.07899497487437186,
"grad_norm": 2.305363416671753,
"learning_rate": 1.295517045092359e-05,
"loss": 7.04832649230957,
"step": 1965
},
{
"epoch": 0.07903517587939698,
"grad_norm": 2.6508641242980957,
"learning_rate": 1.2955102348846707e-05,
"loss": 7.627021789550781,
"step": 1966
},
{
"epoch": 0.07907537688442211,
"grad_norm": 2.8539915084838867,
"learning_rate": 1.2955034195440783e-05,
"loss": 7.156315803527832,
"step": 1967
},
{
"epoch": 0.07911557788944723,
"grad_norm": 2.774005889892578,
"learning_rate": 1.2954965990706913e-05,
"loss": 7.525460720062256,
"step": 1968
},
{
"epoch": 0.07915577889447237,
"grad_norm": 2.561547040939331,
"learning_rate": 1.2954897734646187e-05,
"loss": 7.648047924041748,
"step": 1969
},
{
"epoch": 0.07919597989949749,
"grad_norm": 2.4426262378692627,
"learning_rate": 1.2954829427259698e-05,
"loss": 7.389669895172119,
"step": 1970
},
{
"epoch": 0.07923618090452261,
"grad_norm": 2.510223865509033,
"learning_rate": 1.295476106854854e-05,
"loss": 7.3167033195495605,
"step": 1971
},
{
"epoch": 0.07927638190954774,
"grad_norm": 2.806244373321533,
"learning_rate": 1.2954692658513811e-05,
"loss": 7.0717034339904785,
"step": 1972
},
{
"epoch": 0.07931658291457286,
"grad_norm": 2.699254035949707,
"learning_rate": 1.2954624197156604e-05,
"loss": 7.4777421951293945,
"step": 1973
},
{
"epoch": 0.079356783919598,
"grad_norm": 2.5597407817840576,
"learning_rate": 1.2954555684478014e-05,
"loss": 7.413976192474365,
"step": 1974
},
{
"epoch": 0.07939698492462312,
"grad_norm": 2.321371555328369,
"learning_rate": 1.2954487120479141e-05,
"loss": 7.145709991455078,
"step": 1975
},
{
"epoch": 0.07943718592964824,
"grad_norm": 2.7186472415924072,
"learning_rate": 1.295441850516108e-05,
"loss": 7.5536909103393555,
"step": 1976
},
{
"epoch": 0.07947738693467336,
"grad_norm": 2.5874831676483154,
"learning_rate": 1.2954349838524933e-05,
"loss": 7.042752265930176,
"step": 1977
},
{
"epoch": 0.07951758793969849,
"grad_norm": 2.495584011077881,
"learning_rate": 1.29542811205718e-05,
"loss": 7.406393051147461,
"step": 1978
},
{
"epoch": 0.07955778894472362,
"grad_norm": 2.5019516944885254,
"learning_rate": 1.2954212351302778e-05,
"loss": 7.357718467712402,
"step": 1979
},
{
"epoch": 0.07959798994974875,
"grad_norm": 2.433073043823242,
"learning_rate": 1.295414353071897e-05,
"loss": 7.221610069274902,
"step": 1980
},
{
"epoch": 0.07963819095477387,
"grad_norm": 2.6550650596618652,
"learning_rate": 1.2954074658821478e-05,
"loss": 7.094146728515625,
"step": 1981
},
{
"epoch": 0.07967839195979899,
"grad_norm": 2.6863996982574463,
"learning_rate": 1.2954005735611406e-05,
"loss": 7.0313520431518555,
"step": 1982
},
{
"epoch": 0.07971859296482411,
"grad_norm": 2.9734346866607666,
"learning_rate": 1.2953936761089857e-05,
"loss": 7.399511337280273,
"step": 1983
},
{
"epoch": 0.07975879396984925,
"grad_norm": 2.533616542816162,
"learning_rate": 1.2953867735257937e-05,
"loss": 7.475156307220459,
"step": 1984
},
{
"epoch": 0.07979899497487437,
"grad_norm": 2.54233980178833,
"learning_rate": 1.2953798658116749e-05,
"loss": 7.520194053649902,
"step": 1985
},
{
"epoch": 0.0798391959798995,
"grad_norm": 2.5991930961608887,
"learning_rate": 1.2953729529667398e-05,
"loss": 7.474843978881836,
"step": 1986
},
{
"epoch": 0.07987939698492462,
"grad_norm": 2.6241514682769775,
"learning_rate": 1.2953660349910996e-05,
"loss": 7.26750373840332,
"step": 1987
},
{
"epoch": 0.07991959798994974,
"grad_norm": 2.5793633460998535,
"learning_rate": 1.295359111884865e-05,
"loss": 7.666259765625,
"step": 1988
},
{
"epoch": 0.07995979899497488,
"grad_norm": 2.9125654697418213,
"learning_rate": 1.2953521836481464e-05,
"loss": 7.345658779144287,
"step": 1989
},
{
"epoch": 0.08,
"grad_norm": 2.5939486026763916,
"learning_rate": 1.295345250281055e-05,
"loss": 7.869709014892578,
"step": 1990
},
{
"epoch": 0.08004020100502512,
"grad_norm": 2.552077293395996,
"learning_rate": 1.2953383117837022e-05,
"loss": 7.194121360778809,
"step": 1991
},
{
"epoch": 0.08008040201005025,
"grad_norm": 3.1100785732269287,
"learning_rate": 1.2953313681561987e-05,
"loss": 7.630585670471191,
"step": 1992
},
{
"epoch": 0.08012060301507537,
"grad_norm": 2.779852867126465,
"learning_rate": 1.2953244193986558e-05,
"loss": 7.423210144042969,
"step": 1993
},
{
"epoch": 0.0801608040201005,
"grad_norm": 2.5729453563690186,
"learning_rate": 1.2953174655111848e-05,
"loss": 7.60636043548584,
"step": 1994
},
{
"epoch": 0.08020100502512563,
"grad_norm": 2.551025390625,
"learning_rate": 1.295310506493897e-05,
"loss": 7.449385643005371,
"step": 1995
},
{
"epoch": 0.08024120603015075,
"grad_norm": 2.7013447284698486,
"learning_rate": 1.295303542346904e-05,
"loss": 7.096796035766602,
"step": 1996
},
{
"epoch": 0.08028140703517588,
"grad_norm": 2.888603925704956,
"learning_rate": 1.2952965730703172e-05,
"loss": 7.195039749145508,
"step": 1997
},
{
"epoch": 0.08032160804020101,
"grad_norm": 3.102224111557007,
"learning_rate": 1.2952895986642484e-05,
"loss": 7.1946282386779785,
"step": 1998
},
{
"epoch": 0.08036180904522613,
"grad_norm": 2.6452014446258545,
"learning_rate": 1.295282619128809e-05,
"loss": 7.693393707275391,
"step": 1999
},
{
"epoch": 0.08040201005025126,
"grad_norm": 2.65116548538208,
"learning_rate": 1.295275634464111e-05,
"loss": 7.32562255859375,
"step": 2000
},
{
"epoch": 0.08040201005025126,
"eval_loss": 7.4517598152160645,
"eval_runtime": 120.889,
"eval_samples_per_second": 4.136,
"eval_steps_per_second": 0.521,
"step": 2000
}
],
"logging_steps": 1,
"max_steps": 24875,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 200,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 7.17152432685056e+17,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}