9b-45 / trainer_state.json
furproxy's picture
Upload folder using huggingface_hub
2147799 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.0,
"eval_steps": 500,
"global_step": 532,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.015037593984962405,
"grad_norm": 0.4758685231208801,
"learning_rate": 7.407407407407407e-07,
"loss": 1.9719613790512085,
"step": 2
},
{
"epoch": 0.03007518796992481,
"grad_norm": 1.2873609066009521,
"learning_rate": 2.222222222222222e-06,
"loss": 2.2840397357940674,
"step": 4
},
{
"epoch": 0.045112781954887216,
"grad_norm": 0.4226410984992981,
"learning_rate": 3.7037037037037037e-06,
"loss": 2.0337564945220947,
"step": 6
},
{
"epoch": 0.06015037593984962,
"grad_norm": 0.20389820635318756,
"learning_rate": 5.185185185185185e-06,
"loss": 1.938320517539978,
"step": 8
},
{
"epoch": 0.07518796992481203,
"grad_norm": 1.1121097803115845,
"learning_rate": 6.666666666666667e-06,
"loss": 1.9521509408950806,
"step": 10
},
{
"epoch": 0.09022556390977443,
"grad_norm": 1.0452852249145508,
"learning_rate": 8.148148148148148e-06,
"loss": 2.3038036823272705,
"step": 12
},
{
"epoch": 0.10526315789473684,
"grad_norm": 0.17127464711666107,
"learning_rate": 9.62962962962963e-06,
"loss": 1.9031141996383667,
"step": 14
},
{
"epoch": 0.12030075187969924,
"grad_norm": 1.0378482341766357,
"learning_rate": 1.1111111111111113e-05,
"loss": 2.545740842819214,
"step": 16
},
{
"epoch": 0.13533834586466165,
"grad_norm": 1.3770403861999512,
"learning_rate": 1.2592592592592593e-05,
"loss": 3.2497336864471436,
"step": 18
},
{
"epoch": 0.15037593984962405,
"grad_norm": 0.6125630140304565,
"learning_rate": 1.4074074074074075e-05,
"loss": 1.7776570320129395,
"step": 20
},
{
"epoch": 0.16541353383458646,
"grad_norm": 0.6343645453453064,
"learning_rate": 1.555555555555556e-05,
"loss": 1.9716706275939941,
"step": 22
},
{
"epoch": 0.18045112781954886,
"grad_norm": 0.6397086977958679,
"learning_rate": 1.7037037037037038e-05,
"loss": 1.9381436109542847,
"step": 24
},
{
"epoch": 0.19548872180451127,
"grad_norm": 1.0100289583206177,
"learning_rate": 1.851851851851852e-05,
"loss": 1.9333921670913696,
"step": 26
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.6244422793388367,
"learning_rate": 2e-05,
"loss": 1.7134058475494385,
"step": 28
},
{
"epoch": 0.22556390977443608,
"grad_norm": 1.4532567262649536,
"learning_rate": 1.9994079505294254e-05,
"loss": 2.0251219272613525,
"step": 30
},
{
"epoch": 0.24060150375939848,
"grad_norm": 0.5191428661346436,
"learning_rate": 1.9976326268767035e-05,
"loss": 1.578896164894104,
"step": 32
},
{
"epoch": 0.2556390977443609,
"grad_norm": 0.2822546064853668,
"learning_rate": 1.994676502169901e-05,
"loss": 1.5274699926376343,
"step": 34
},
{
"epoch": 0.2706766917293233,
"grad_norm": 0.26378244161605835,
"learning_rate": 1.9905436944609424e-05,
"loss": 1.510546326637268,
"step": 36
},
{
"epoch": 0.2857142857142857,
"grad_norm": 2.3758437633514404,
"learning_rate": 1.9852399609889242e-05,
"loss": 2.110567331314087,
"step": 38
},
{
"epoch": 0.3007518796992481,
"grad_norm": 0.419758677482605,
"learning_rate": 1.9787726901599502e-05,
"loss": 1.1862285137176514,
"step": 40
},
{
"epoch": 0.3157894736842105,
"grad_norm": 0.24496647715568542,
"learning_rate": 1.9711508912546566e-05,
"loss": 1.440342903137207,
"step": 42
},
{
"epoch": 0.3308270676691729,
"grad_norm": 0.3002743721008301,
"learning_rate": 1.9623851818777652e-05,
"loss": 1.0962785482406616,
"step": 44
},
{
"epoch": 0.3458646616541353,
"grad_norm": 0.11880263686180115,
"learning_rate": 1.9524877731671482e-05,
"loss": 1.493391513824463,
"step": 46
},
{
"epoch": 0.3609022556390977,
"grad_norm": 1.7871476411819458,
"learning_rate": 1.941472452783011e-05,
"loss": 1.2433573007583618,
"step": 48
},
{
"epoch": 0.37593984962406013,
"grad_norm": 0.2921432852745056,
"learning_rate": 1.9293545657008865e-05,
"loss": 1.1390293836593628,
"step": 50
},
{
"epoch": 0.39097744360902253,
"grad_norm": 0.12673601508140564,
"learning_rate": 1.9161509928352017e-05,
"loss": 1.0903499126434326,
"step": 52
},
{
"epoch": 0.40601503759398494,
"grad_norm": 0.1648157387971878,
"learning_rate": 1.901880127523192e-05,
"loss": 1.1544872522354126,
"step": 54
},
{
"epoch": 0.42105263157894735,
"grad_norm": 0.08962542563676834,
"learning_rate": 1.886561849901922e-05,
"loss": 1.0822100639343262,
"step": 56
},
{
"epoch": 0.43609022556390975,
"grad_norm": 0.724937379360199,
"learning_rate": 1.870217499214111e-05,
"loss": 1.0416043996810913,
"step": 58
},
{
"epoch": 0.45112781954887216,
"grad_norm": 0.19841401278972626,
"learning_rate": 1.8528698440813397e-05,
"loss": 1.1815505027770996,
"step": 60
},
{
"epoch": 0.46616541353383456,
"grad_norm": 0.2616511285305023,
"learning_rate": 1.8345430507860478e-05,
"loss": 0.9047210812568665,
"step": 62
},
{
"epoch": 0.48120300751879697,
"grad_norm": 0.22353343665599823,
"learning_rate": 1.8152626496065128e-05,
"loss": 1.201892614364624,
"step": 64
},
{
"epoch": 0.49624060150375937,
"grad_norm": 0.6634237766265869,
"learning_rate": 1.7950554992517014e-05,
"loss": 1.1795772314071655,
"step": 66
},
{
"epoch": 0.5112781954887218,
"grad_norm": 0.13962750136852264,
"learning_rate": 1.7739497494455412e-05,
"loss": 1.3534270524978638,
"step": 68
},
{
"epoch": 0.5263157894736842,
"grad_norm": 0.665572464466095,
"learning_rate": 1.7519748017127354e-05,
"loss": 1.125345230102539,
"step": 70
},
{
"epoch": 0.5413533834586466,
"grad_norm": 0.179820716381073,
"learning_rate": 1.729161268420746e-05,
"loss": 0.8090606331825256,
"step": 72
},
{
"epoch": 0.556390977443609,
"grad_norm": 0.11726067215204239,
"learning_rate": 1.7055409301350013e-05,
"loss": 0.9213717579841614,
"step": 74
},
{
"epoch": 0.5714285714285714,
"grad_norm": 0.1627010852098465,
"learning_rate": 1.681146691346742e-05,
"loss": 1.0821505784988403,
"step": 76
},
{
"epoch": 0.5864661654135338,
"grad_norm": 0.1170893982052803,
"learning_rate": 1.6560125346351663e-05,
"loss": 1.300316333770752,
"step": 78
},
{
"epoch": 0.6015037593984962,
"grad_norm": 0.5334263443946838,
"learning_rate": 1.6301734733277442e-05,
"loss": 0.6484270691871643,
"step": 80
},
{
"epoch": 0.6165413533834586,
"grad_norm": 0.101778045296669,
"learning_rate": 1.603665502724633e-05,
"loss": 1.2577356100082397,
"step": 82
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.1670590043067932,
"learning_rate": 1.576525549955156e-05,
"loss": 1.3543009757995605,
"step": 84
},
{
"epoch": 0.6466165413533834,
"grad_norm": 0.20087628066539764,
"learning_rate": 1.548791422536178e-05,
"loss": 0.9327285885810852,
"step": 86
},
{
"epoch": 0.6616541353383458,
"grad_norm": 0.23011328279972076,
"learning_rate": 1.5205017557040656e-05,
"loss": 1.1237722635269165,
"step": 88
},
{
"epoch": 0.6766917293233082,
"grad_norm": 0.159201517701149,
"learning_rate": 1.4916959585935732e-05,
"loss": 1.1964070796966553,
"step": 90
},
{
"epoch": 0.6917293233082706,
"grad_norm": 0.44444212317466736,
"learning_rate": 1.4624141593386507e-05,
"loss": 1.0308165550231934,
"step": 92
},
{
"epoch": 0.706766917293233,
"grad_norm": 0.22338902950286865,
"learning_rate": 1.4326971491716427e-05,
"loss": 0.9982426762580872,
"step": 94
},
{
"epoch": 0.7218045112781954,
"grad_norm": 0.2096806913614273,
"learning_rate": 1.402586325598752e-05,
"loss": 1.3940727710723877,
"step": 96
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.21466873586177826,
"learning_rate": 1.3721236347309314e-05,
"loss": 1.1801196336746216,
"step": 98
},
{
"epoch": 0.7518796992481203,
"grad_norm": 0.30590999126434326,
"learning_rate": 1.3413515128505363e-05,
"loss": 0.6430416703224182,
"step": 100
},
{
"epoch": 0.7669172932330827,
"grad_norm": 0.13401509821414948,
"learning_rate": 1.3103128272951363e-05,
"loss": 1.3783133029937744,
"step": 102
},
{
"epoch": 0.7819548872180451,
"grad_norm": 0.12506185472011566,
"learning_rate": 1.2790508167408509e-05,
"loss": 0.9889219403266907,
"step": 104
},
{
"epoch": 0.7969924812030075,
"grad_norm": 0.22044184803962708,
"learning_rate": 1.2476090309683804e-05,
"loss": 0.6871194243431091,
"step": 106
},
{
"epoch": 0.8120300751879699,
"grad_norm": 0.18642327189445496,
"learning_rate": 1.2160312701956553e-05,
"loss": 1.0068978071212769,
"step": 108
},
{
"epoch": 0.8270676691729323,
"grad_norm": 0.5397107601165771,
"learning_rate": 1.1843615240616111e-05,
"loss": 0.8988245725631714,
"step": 110
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.23860491812229156,
"learning_rate": 1.1526439103460874e-05,
"loss": 0.7688661813735962,
"step": 112
},
{
"epoch": 0.8571428571428571,
"grad_norm": 0.1502920240163803,
"learning_rate": 1.120922613511221e-05,
"loss": 1.0024021863937378,
"step": 114
},
{
"epoch": 0.8721804511278195,
"grad_norm": 0.16608496010303497,
"learning_rate": 1.0892418231499461e-05,
"loss": 1.4191375970840454,
"step": 116
},
{
"epoch": 0.8872180451127819,
"grad_norm": 0.17546993494033813,
"learning_rate": 1.057645672427347e-05,
"loss": 1.073761224746704,
"step": 118
},
{
"epoch": 0.9022556390977443,
"grad_norm": 5.0469841957092285,
"learning_rate": 1.0261781766006174e-05,
"loss": 0.9056495428085327,
"step": 120
},
{
"epoch": 0.9172932330827067,
"grad_norm": 0.36213457584381104,
"learning_rate": 9.948831717032738e-06,
"loss": 0.8672894835472107,
"step": 122
},
{
"epoch": 0.9323308270676691,
"grad_norm": 0.4218904972076416,
"learning_rate": 9.638042534790373e-06,
"loss": 0.827739417552948,
"step": 124
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.0983240008354187,
"learning_rate": 9.329847166504497e-06,
"loss": 0.7799423933029175,
"step": 126
},
{
"epoch": 0.9624060150375939,
"grad_norm": 0.14850644767284393,
"learning_rate": 9.024674946068357e-06,
"loss": 1.0653791427612305,
"step": 128
},
{
"epoch": 0.9774436090225563,
"grad_norm": 0.27504855394363403,
"learning_rate": 8.722950995956172e-06,
"loss": 0.9135006666183472,
"step": 130
},
{
"epoch": 0.9924812030075187,
"grad_norm": 0.27222368121147156,
"learning_rate": 8.425095635003053e-06,
"loss": 1.0402815341949463,
"step": 132
},
{
"epoch": 1.0075187969924813,
"grad_norm": 0.1408149003982544,
"learning_rate": 8.13152379287667e-06,
"loss": 0.7629735469818115,
"step": 134
},
{
"epoch": 1.0225563909774436,
"grad_norm": 0.15156933665275574,
"learning_rate": 7.842644432056336e-06,
"loss": 0.9513287544250488,
"step": 136
},
{
"epoch": 1.037593984962406,
"grad_norm": 0.9048015475273132,
"learning_rate": 7.55885997812472e-06,
"loss": 0.6946667432785034,
"step": 138
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.24624674022197723,
"learning_rate": 7.280565759165833e-06,
"loss": 0.7876754403114319,
"step": 140
},
{
"epoch": 1.0676691729323309,
"grad_norm": 0.09362401068210602,
"learning_rate": 7.008149455050264e-06,
"loss": 0.8406305909156799,
"step": 142
},
{
"epoch": 1.0827067669172932,
"grad_norm": 0.17641493678092957,
"learning_rate": 6.741990557374784e-06,
"loss": 1.0424906015396118,
"step": 144
},
{
"epoch": 1.0977443609022557,
"grad_norm": 0.12022742629051208,
"learning_rate": 6.4824598408087015e-06,
"loss": 1.0722497701644897,
"step": 146
},
{
"epoch": 1.112781954887218,
"grad_norm": 0.08780567348003387,
"learning_rate": 6.229918846583414e-06,
"loss": 1.0312786102294922,
"step": 148
},
{
"epoch": 1.1278195488721805,
"grad_norm": 0.3304075598716736,
"learning_rate": 5.984719378844628e-06,
"loss": 1.0075746774673462,
"step": 150
},
{
"epoch": 1.1428571428571428,
"grad_norm": 0.10900267213582993,
"learning_rate": 5.7472030145689604e-06,
"loss": 1.0427347421646118,
"step": 152
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.09693319350481033,
"learning_rate": 5.51770062772752e-06,
"loss": 1.0927042961120605,
"step": 154
},
{
"epoch": 1.1729323308270676,
"grad_norm": 0.12883000075817108,
"learning_rate": 5.296531928359431e-06,
"loss": 0.9705473780632019,
"step": 156
},
{
"epoch": 1.1879699248120301,
"grad_norm": 0.11416517943143845,
"learning_rate": 5.084005017197318e-06,
"loss": 1.0172467231750488,
"step": 158
},
{
"epoch": 1.2030075187969924,
"grad_norm": 0.10797861963510513,
"learning_rate": 4.8804159564652665e-06,
"loss": 0.5409541726112366,
"step": 160
},
{
"epoch": 1.218045112781955,
"grad_norm": 0.24861538410186768,
"learning_rate": 4.686048357447095e-06,
"loss": 0.9430153965950012,
"step": 162
},
{
"epoch": 1.2330827067669172,
"grad_norm": 0.14524255692958832,
"learning_rate": 4.501172985399498e-06,
"loss": 1.178081750869751,
"step": 164
},
{
"epoch": 1.2481203007518797,
"grad_norm": 0.2940176725387573,
"learning_rate": 4.326047382360457e-06,
"loss": 0.8844167590141296,
"step": 166
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.1306038349866867,
"learning_rate": 4.160915508378359e-06,
"loss": 0.7063813209533691,
"step": 168
},
{
"epoch": 1.2781954887218046,
"grad_norm": 0.15677547454833984,
"learning_rate": 4.006007401661596e-06,
"loss": 0.7762787938117981,
"step": 170
},
{
"epoch": 1.2932330827067668,
"grad_norm": 0.14480236172676086,
"learning_rate": 3.861538858122092e-06,
"loss": 0.937335193157196,
"step": 172
},
{
"epoch": 1.3082706766917294,
"grad_norm": 0.15142542123794556,
"learning_rate": 3.727711130759182e-06,
"loss": 0.9747655987739563,
"step": 174
},
{
"epoch": 1.3233082706766917,
"grad_norm": 0.18726477026939392,
"learning_rate": 3.6047106493025923e-06,
"loss": 0.7746855020523071,
"step": 176
},
{
"epoch": 1.3383458646616542,
"grad_norm": 2.189209461212158,
"learning_rate": 3.492708760505093e-06,
"loss": 0.6926825642585754,
"step": 178
},
{
"epoch": 1.3533834586466165,
"grad_norm": 0.12323262542486191,
"learning_rate": 3.3918614894466045e-06,
"loss": 1.2502151727676392,
"step": 180
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.1458672434091568,
"learning_rate": 3.3023093221822746e-06,
"loss": 0.9960780143737793,
"step": 182
},
{
"epoch": 1.3834586466165413,
"grad_norm": 0.1072113886475563,
"learning_rate": 3.224177010037323e-06,
"loss": 0.8326720595359802,
"step": 184
},
{
"epoch": 1.3984962406015038,
"grad_norm": 0.09054487943649292,
"learning_rate": 3.1575733958212563e-06,
"loss": 1.1920455694198608,
"step": 186
},
{
"epoch": 1.413533834586466,
"grad_norm": 0.13850678503513336,
"learning_rate": 3.1025912622035687e-06,
"loss": 0.5979896783828735,
"step": 188
},
{
"epoch": 1.4285714285714286,
"grad_norm": 0.2317809760570526,
"learning_rate": 3.0593072024621396e-06,
"loss": 0.969947099685669,
"step": 190
},
{
"epoch": 1.443609022556391,
"grad_norm": 0.29049012064933777,
"learning_rate": 3.0277815137843917e-06,
"loss": 0.7010709643363953,
"step": 192
},
{
"epoch": 1.4586466165413534,
"grad_norm": 0.17280922830104828,
"learning_rate": 3.008058113269836e-06,
"loss": 0.7660905718803406,
"step": 194
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.12446308135986328,
"learning_rate": 3.0001644767510154e-06,
"loss": 0.9958880543708801,
"step": 196
},
{
"epoch": 1.4887218045112782,
"grad_norm": 0.1736506223678589,
"learning_rate": 3.0041116005181016e-06,
"loss": 0.9368724226951599,
"step": 198
},
{
"epoch": 1.5037593984962405,
"grad_norm": 0.15870751440525055,
"learning_rate": 3.0198939860004202e-06,
"loss": 0.9826479554176331,
"step": 200
},
{
"epoch": 1.518796992481203,
"grad_norm": 0.15492001175880432,
"learning_rate": 3.0474896474262772e-06,
"loss": 1.3954254388809204,
"step": 202
},
{
"epoch": 1.5338345864661656,
"grad_norm": 0.1334741860628128,
"learning_rate": 3.08686014245041e-06,
"loss": 1.0462982654571533,
"step": 204
},
{
"epoch": 1.5488721804511278,
"grad_norm": 0.26873454451560974,
"learning_rate": 3.1379506257063825e-06,
"loss": 0.8277729153633118,
"step": 206
},
{
"epoch": 1.5639097744360901,
"grad_norm": 0.16168057918548584,
"learning_rate": 3.20068992520934e-06,
"loss": 0.985795259475708,
"step": 208
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.1431337147951126,
"learning_rate": 3.274990641502683e-06,
"loss": 0.908364474773407,
"step": 210
},
{
"epoch": 1.5939849624060152,
"grad_norm": 0.2742830514907837,
"learning_rate": 3.3607492694105405e-06,
"loss": 0.6370347142219543,
"step": 212
},
{
"epoch": 1.6090225563909775,
"grad_norm": 0.14003214240074158,
"learning_rate": 3.457846342226442e-06,
"loss": 0.9584491848945618,
"step": 214
},
{
"epoch": 1.6240601503759398,
"grad_norm": 0.29483214020729065,
"learning_rate": 3.5661465981373183e-06,
"loss": 1.0774601697921753,
"step": 216
},
{
"epoch": 1.6390977443609023,
"grad_norm": 0.18227490782737732,
"learning_rate": 3.6854991686509906e-06,
"loss": 0.7987947463989258,
"step": 218
},
{
"epoch": 1.6541353383458648,
"grad_norm": 0.20908379554748535,
"learning_rate": 3.815737788764674e-06,
"loss": 0.8292507529258728,
"step": 220
},
{
"epoch": 1.669172932330827,
"grad_norm": 0.23883360624313354,
"learning_rate": 3.956681028581693e-06,
"loss": 0.9586336612701416,
"step": 222
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.3358386158943176,
"learning_rate": 4.108132546053779e-06,
"loss": 0.8150299191474915,
"step": 224
},
{
"epoch": 1.699248120300752,
"grad_norm": 0.298909068107605,
"learning_rate": 4.269881360496842e-06,
"loss": 1.0946331024169922,
"step": 226
},
{
"epoch": 1.7142857142857144,
"grad_norm": 0.24479779601097107,
"learning_rate": 4.441702146499222e-06,
"loss": 0.9061790108680725,
"step": 228
},
{
"epoch": 1.7293233082706767,
"grad_norm": 0.2154252976179123,
"learning_rate": 4.623355547812946e-06,
"loss": 1.0011441707611084,
"step": 230
},
{
"epoch": 1.744360902255639,
"grad_norm": 0.1481187492609024,
"learning_rate": 4.814588510790782e-06,
"loss": 1.1533119678497314,
"step": 232
},
{
"epoch": 1.7593984962406015,
"grad_norm": 0.10200405865907669,
"learning_rate": 5.01513463690452e-06,
"loss": 1.1073795557022095,
"step": 234
},
{
"epoch": 1.774436090225564,
"grad_norm": 0.19221612811088562,
"learning_rate": 5.224714553853478e-06,
"loss": 1.148139476776123,
"step": 236
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.1456657350063324,
"learning_rate": 5.443036304746191e-06,
"loss": 0.6271846294403076,
"step": 238
},
{
"epoch": 1.8045112781954886,
"grad_norm": 0.3602541983127594,
"learning_rate": 5.66979575481317e-06,
"loss": 1.0237879753112793,
"step": 240
},
{
"epoch": 1.8195488721804511,
"grad_norm": 0.14290253818035126,
"learning_rate": 5.904677015084159e-06,
"loss": 0.7585715055465698,
"step": 242
},
{
"epoch": 1.8345864661654137,
"grad_norm": 0.15334200859069824,
"learning_rate": 6.147352882439652e-06,
"loss": 1.1932705640792847,
"step": 244
},
{
"epoch": 1.849624060150376,
"grad_norm": 0.09177304059267044,
"learning_rate": 6.397485295423669e-06,
"loss": 1.1568275690078735,
"step": 246
},
{
"epoch": 1.8646616541353382,
"grad_norm": 0.08284302055835724,
"learning_rate": 6.6547258051828426e-06,
"loss": 0.9782670736312866,
"step": 248
},
{
"epoch": 1.8796992481203008,
"grad_norm": 0.12089983373880386,
"learning_rate": 6.918716060875743e-06,
"loss": 1.2261128425598145,
"step": 250
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.1200575977563858,
"learning_rate": 7.1890883088761885e-06,
"loss": 1.0725328922271729,
"step": 252
},
{
"epoch": 1.9097744360902256,
"grad_norm": 0.2572805881500244,
"learning_rate": 7.4654659050752845e-06,
"loss": 1.2271314859390259,
"step": 254
},
{
"epoch": 1.9248120300751879,
"grad_norm": 0.2713569104671478,
"learning_rate": 7.747463839568292e-06,
"loss": 0.8813698291778564,
"step": 256
},
{
"epoch": 1.9398496240601504,
"grad_norm": 0.5769055485725403,
"learning_rate": 8.034689272995649e-06,
"loss": 0.5630529522895813,
"step": 258
},
{
"epoch": 1.954887218045113,
"grad_norm": 0.13255445659160614,
"learning_rate": 8.32674208379076e-06,
"loss": 1.0115076303482056,
"step": 260
},
{
"epoch": 1.9699248120300752,
"grad_norm": 0.15804584324359894,
"learning_rate": 8.623215425572433e-06,
"loss": 0.9677861332893372,
"step": 262
},
{
"epoch": 1.9849624060150375,
"grad_norm": 0.09761196374893188,
"learning_rate": 8.923696293905279e-06,
"loss": 0.9606061577796936,
"step": 264
},
{
"epoch": 2.0,
"grad_norm": 0.24307098984718323,
"learning_rate": 9.227766101638668e-06,
"loss": 0.8462581038475037,
"step": 266
},
{
"epoch": 2.0150375939849625,
"grad_norm": 0.15874817967414856,
"learning_rate": 9.535001262022755e-06,
"loss": 0.5168899893760681,
"step": 268
},
{
"epoch": 2.030075187969925,
"grad_norm": 0.09884876012802124,
"learning_rate": 9.844973778789198e-06,
"loss": 0.725158154964447,
"step": 270
},
{
"epoch": 2.045112781954887,
"grad_norm": 0.11747175455093384,
"learning_rate": 1.0157251842374587e-05,
"loss": 0.5184513330459595,
"step": 272
},
{
"epoch": 2.0601503759398496,
"grad_norm": 0.17868469655513763,
"learning_rate": 1.0471400431456062e-05,
"loss": 1.0087833404541016,
"step": 274
},
{
"epoch": 2.075187969924812,
"grad_norm": 0.12505443394184113,
"learning_rate": 1.078698191896105e-05,
"loss": 0.9181981682777405,
"step": 276
},
{
"epoch": 2.090225563909774,
"grad_norm": 0.45129311084747314,
"learning_rate": 1.1103556681706949e-05,
"loss": 0.867568850517273,
"step": 278
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.38053497672080994,
"learning_rate": 1.1420683712821577e-05,
"loss": 0.8937767744064331,
"step": 280
},
{
"epoch": 2.1203007518796992,
"grad_norm": 0.2623603045940399,
"learning_rate": 1.1737921236091097e-05,
"loss": 1.0325863361358643,
"step": 282
},
{
"epoch": 2.1353383458646618,
"grad_norm": 0.13423830270767212,
"learning_rate": 1.2054827321379756e-05,
"loss": 0.6633681654930115,
"step": 284
},
{
"epoch": 2.1503759398496243,
"grad_norm": 0.20472559332847595,
"learning_rate": 1.237096050026399e-05,
"loss": 0.998541533946991,
"step": 286
},
{
"epoch": 2.1654135338345863,
"grad_norm": 0.28264734148979187,
"learning_rate": 1.2685880381023343e-05,
"loss": 1.1486636400222778,
"step": 288
},
{
"epoch": 2.180451127819549,
"grad_norm": 0.11129236966371536,
"learning_rate": 1.2999148262131527e-05,
"loss": 1.1767593622207642,
"step": 290
},
{
"epoch": 2.1954887218045114,
"grad_norm": 0.17632298171520233,
"learning_rate": 1.3310327743392861e-05,
"loss": 0.8024111986160278,
"step": 292
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.1285456418991089,
"learning_rate": 1.3618985333872927e-05,
"loss": 0.9202947616577148,
"step": 294
},
{
"epoch": 2.225563909774436,
"grad_norm": 0.2532704770565033,
"learning_rate": 1.3924691055776338e-05,
"loss": 0.3760508596897125,
"step": 296
},
{
"epoch": 2.2406015037593985,
"grad_norm": 0.18166831135749817,
"learning_rate": 1.4227019043430623e-05,
"loss": 1.2139897346496582,
"step": 298
},
{
"epoch": 2.255639097744361,
"grad_norm": 0.31214964389801025,
"learning_rate": 1.4525548136541599e-05,
"loss": 1.1467747688293457,
"step": 300
},
{
"epoch": 2.2706766917293235,
"grad_norm": 0.2208021879196167,
"learning_rate": 1.4819862466893891e-05,
"loss": 1.152048110961914,
"step": 302
},
{
"epoch": 2.2857142857142856,
"grad_norm": 0.28248584270477295,
"learning_rate": 1.5109552037679353e-05,
"loss": 0.8999449610710144,
"step": 304
},
{
"epoch": 2.300751879699248,
"grad_norm": 0.17761224508285522,
"learning_rate": 1.539421329464618e-05,
"loss": 0.8530235290527344,
"step": 306
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.12931574881076813,
"learning_rate": 1.5673449688273246e-05,
"loss": 1.1707788705825806,
"step": 308
},
{
"epoch": 2.3308270676691727,
"grad_norm": 0.45592963695526123,
"learning_rate": 1.5946872226186394e-05,
"loss": 0.543804407119751,
"step": 310
},
{
"epoch": 2.345864661654135,
"grad_norm": 0.13944678008556366,
"learning_rate": 1.621410001504725e-05,
"loss": 0.9259825944900513,
"step": 312
},
{
"epoch": 2.3609022556390977,
"grad_norm": 0.6154989004135132,
"learning_rate": 1.6474760791159528e-05,
"loss": 0.6412667036056519,
"step": 314
},
{
"epoch": 2.3759398496240602,
"grad_norm": 0.19801302254199982,
"learning_rate": 1.672849143905383e-05,
"loss": 1.0553992986679077,
"step": 316
},
{
"epoch": 2.3909774436090228,
"grad_norm": 0.16842691600322723,
"learning_rate": 1.697493849732838e-05,
"loss": 0.9918773174285889,
"step": 318
},
{
"epoch": 2.406015037593985,
"grad_norm": 0.17690639197826385,
"learning_rate": 1.7213758651041114e-05,
"loss": 1.183601975440979,
"step": 320
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.1753586232662201,
"learning_rate": 1.7444619209967163e-05,
"loss": 1.2674732208251953,
"step": 322
},
{
"epoch": 2.43609022556391,
"grad_norm": 0.1980062872171402,
"learning_rate": 1.766719857205551e-05,
"loss": 0.9618882536888123,
"step": 324
},
{
"epoch": 2.451127819548872,
"grad_norm": 0.14773985743522644,
"learning_rate": 1.7881186671439107e-05,
"loss": 1.1368170976638794,
"step": 326
},
{
"epoch": 2.4661654135338344,
"grad_norm": 0.3491007685661316,
"learning_rate": 1.8086285410374563e-05,
"loss": 0.9204502701759338,
"step": 328
},
{
"epoch": 2.481203007518797,
"grad_norm": 0.41794508695602417,
"learning_rate": 1.8282209074509393e-05,
"loss": 0.7325559258460999,
"step": 330
},
{
"epoch": 2.4962406015037595,
"grad_norm": 0.1503211408853531,
"learning_rate": 1.8468684730898602e-05,
"loss": 1.1228058338165283,
"step": 332
},
{
"epoch": 2.511278195488722,
"grad_norm": 0.29104840755462646,
"learning_rate": 1.864545260821598e-05,
"loss": 0.6551952958106995,
"step": 334
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.17387054860591888,
"learning_rate": 1.8812266458630505e-05,
"loss": 0.7751361727714539,
"step": 336
},
{
"epoch": 2.5413533834586466,
"grad_norm": 0.7814794778823853,
"learning_rate": 1.8968893900843786e-05,
"loss": 0.7552028298377991,
"step": 338
},
{
"epoch": 2.556390977443609,
"grad_norm": 0.15030670166015625,
"learning_rate": 1.9115116743810537e-05,
"loss": 1.1345906257629395,
"step": 340
},
{
"epoch": 2.571428571428571,
"grad_norm": 0.2455630898475647,
"learning_rate": 1.9250731290691326e-05,
"loss": 1.0457109212875366,
"step": 342
},
{
"epoch": 2.5864661654135337,
"grad_norm": 0.17742854356765747,
"learning_rate": 1.9375548622613952e-05,
"loss": 0.8526303768157959,
"step": 344
},
{
"epoch": 2.601503759398496,
"grad_norm": 0.35299667716026306,
"learning_rate": 1.9489394861848333e-05,
"loss": 0.6485392451286316,
"step": 346
},
{
"epoch": 2.6165413533834587,
"grad_norm": 0.24292704463005066,
"learning_rate": 1.9592111414028205e-05,
"loss": 0.8217464685440063,
"step": 348
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.11688588559627533,
"learning_rate": 1.9683555189082178e-05,
"loss": 0.9268519282341003,
"step": 350
},
{
"epoch": 2.6466165413533833,
"grad_norm": 0.8352806568145752,
"learning_rate": 1.9763598800566478e-05,
"loss": 0.7672562599182129,
"step": 352
},
{
"epoch": 2.661654135338346,
"grad_norm": 0.10471988469362259,
"learning_rate": 1.983213074312156e-05,
"loss": 1.0508689880371094,
"step": 354
},
{
"epoch": 2.6766917293233083,
"grad_norm": 0.9734071493148804,
"learning_rate": 1.9889055547805527e-05,
"loss": 0.8092398047447205,
"step": 356
},
{
"epoch": 2.6917293233082704,
"grad_norm": 0.22569192945957184,
"learning_rate": 1.993429391508785e-05,
"loss": 0.4492420554161072,
"step": 358
},
{
"epoch": 2.706766917293233,
"grad_norm": 0.19527536630630493,
"learning_rate": 1.9967782825318183e-05,
"loss": 1.1782996654510498,
"step": 360
},
{
"epoch": 2.7218045112781954,
"grad_norm": 0.4661547839641571,
"learning_rate": 1.998947562651635e-05,
"loss": 0.6154137253761292,
"step": 362
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.2177896499633789,
"learning_rate": 1.999934209936126e-05,
"loss": 0.7517685890197754,
"step": 364
},
{
"epoch": 2.7518796992481205,
"grad_norm": 0.18744942545890808,
"learning_rate": 1.9997368499288147e-05,
"loss": 0.8893219232559204,
"step": 366
},
{
"epoch": 2.7669172932330826,
"grad_norm": 0.12195390462875366,
"learning_rate": 1.9983557575635553e-05,
"loss": 0.7814352512359619,
"step": 368
},
{
"epoch": 2.781954887218045,
"grad_norm": 0.20601901412010193,
"learning_rate": 1.9957928567815327e-05,
"loss": 0.8266798853874207,
"step": 370
},
{
"epoch": 2.7969924812030076,
"grad_norm": 0.24167893826961517,
"learning_rate": 1.9920517178511025e-05,
"loss": 0.9625144004821777,
"step": 372
},
{
"epoch": 2.8120300751879697,
"grad_norm": 0.20055215060710907,
"learning_rate": 1.9871375523942014e-05,
"loss": 0.7290459275245667,
"step": 374
},
{
"epoch": 2.827067669172932,
"grad_norm": 0.349552720785141,
"learning_rate": 1.981057206126258e-05,
"loss": 1.0641855001449585,
"step": 376
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.19066953659057617,
"learning_rate": 1.9738191493197164e-05,
"loss": 1.1699275970458984,
"step": 378
},
{
"epoch": 2.857142857142857,
"grad_norm": 0.4483891725540161,
"learning_rate": 1.96543346500446e-05,
"loss": 0.8162869811058044,
"step": 380
},
{
"epoch": 2.8721804511278197,
"grad_norm": 0.3727665841579437,
"learning_rate": 1.9559118349215685e-05,
"loss": 0.9681864380836487,
"step": 382
},
{
"epoch": 2.887218045112782,
"grad_norm": 0.8629833459854126,
"learning_rate": 1.9452675232499833e-05,
"loss": 0.9782270789146423,
"step": 384
},
{
"epoch": 2.9022556390977443,
"grad_norm": 0.40549924969673157,
"learning_rate": 1.9335153581287384e-05,
"loss": 0.7416080832481384,
"step": 386
},
{
"epoch": 2.917293233082707,
"grad_norm": 0.2615848481655121,
"learning_rate": 1.920671711000512e-05,
"loss": 0.9780222773551941,
"step": 388
},
{
"epoch": 2.932330827067669,
"grad_norm": 0.1308208554983139,
"learning_rate": 1.906754473805263e-05,
"loss": 0.8454363942146301,
"step": 390
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.19276592135429382,
"learning_rate": 1.8917830340557287e-05,
"loss": 0.974253237247467,
"step": 392
},
{
"epoch": 2.962406015037594,
"grad_norm": 0.40342700481414795,
"learning_rate": 1.875778247829503e-05,
"loss": 0.767073392868042,
"step": 394
},
{
"epoch": 2.9774436090225564,
"grad_norm": 0.28507354855537415,
"learning_rate": 1.8587624107153192e-05,
"loss": 1.0176156759262085,
"step": 396
},
{
"epoch": 2.992481203007519,
"grad_norm": 0.08241955190896988,
"learning_rate": 1.840759226754013e-05,
"loss": 0.8344214558601379,
"step": 398
},
{
"epoch": 3.007518796992481,
"grad_norm": 0.11373747140169144,
"learning_rate": 1.8217937754174265e-05,
"loss": 0.9761697053909302,
"step": 400
},
{
"epoch": 3.0225563909774436,
"grad_norm": 0.34416520595550537,
"learning_rate": 1.8018924766712608e-05,
"loss": 0.7501745223999023,
"step": 402
},
{
"epoch": 3.037593984962406,
"grad_norm": 0.7743138074874878,
"learning_rate": 1.7810830541705425e-05,
"loss": 0.9392253160476685,
"step": 404
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.2988702356815338,
"learning_rate": 1.7593944966389767e-05,
"loss": 0.6723920702934265,
"step": 406
},
{
"epoch": 3.0676691729323307,
"grad_norm": 0.1014496460556984,
"learning_rate": 1.7368570174859855e-05,
"loss": 0.5313879251480103,
"step": 408
},
{
"epoch": 3.082706766917293,
"grad_norm": 0.1569093018770218,
"learning_rate": 1.713502012717686e-05,
"loss": 0.8502162098884583,
"step": 410
},
{
"epoch": 3.0977443609022557,
"grad_norm": 0.4344463348388672,
"learning_rate": 1.6893620172004526e-05,
"loss": 0.6262786388397217,
"step": 412
},
{
"epoch": 3.112781954887218,
"grad_norm": 0.1802617460489273,
"learning_rate": 1.664470659337968e-05,
"loss": 0.9044028520584106,
"step": 414
},
{
"epoch": 3.1278195488721803,
"grad_norm": 0.09171602874994278,
"learning_rate": 1.638862614224926e-05,
"loss": 0.4721546471118927,
"step": 416
},
{
"epoch": 3.142857142857143,
"grad_norm": 0.21735042333602905,
"learning_rate": 1.6125735553426228e-05,
"loss": 0.7782120108604431,
"step": 418
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.10799870640039444,
"learning_rate": 1.5856401048637505e-05,
"loss": 0.8657242059707642,
"step": 420
},
{
"epoch": 3.172932330827068,
"grad_norm": 0.2299298644065857,
"learning_rate": 1.5580997826355945e-05,
"loss": 0.7349554896354675,
"step": 422
},
{
"epoch": 3.18796992481203,
"grad_norm": 0.1193917915225029,
"learning_rate": 1.5299909539127293e-05,
"loss": 0.8061299324035645,
"step": 424
},
{
"epoch": 3.2030075187969924,
"grad_norm": 0.313683420419693,
"learning_rate": 1.5013527759120128e-05,
"loss": 0.7771633267402649,
"step": 426
},
{
"epoch": 3.218045112781955,
"grad_norm": 0.28184592723846436,
"learning_rate": 1.4722251432643222e-05,
"loss": 0.7236590385437012,
"step": 428
},
{
"epoch": 3.2330827067669174,
"grad_norm": 0.2527337968349457,
"learning_rate": 1.4426486324390432e-05,
"loss": 0.5557109713554382,
"step": 430
},
{
"epoch": 3.2481203007518795,
"grad_norm": 0.14354459941387177,
"learning_rate": 1.4126644452187057e-05,
"loss": 0.5296133160591125,
"step": 432
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.45321938395500183,
"learning_rate": 1.382314351302531e-05,
"loss": 0.5751173496246338,
"step": 434
},
{
"epoch": 3.2781954887218046,
"grad_norm": 0.5287338495254517,
"learning_rate": 1.351640630118838e-05,
"loss": 0.5346217751502991,
"step": 436
},
{
"epoch": 3.293233082706767,
"grad_norm": 0.07787490636110306,
"learning_rate": 1.3206860119273606e-05,
"loss": 0.5648679137229919,
"step": 438
},
{
"epoch": 3.308270676691729,
"grad_norm": 0.13844254612922668,
"learning_rate": 1.2894936182935407e-05,
"loss": 1.0638585090637207,
"step": 440
},
{
"epoch": 3.3233082706766917,
"grad_norm": 0.19562360644340515,
"learning_rate": 1.2581069020176986e-05,
"loss": 0.452185720205307,
"step": 442
},
{
"epoch": 3.338345864661654,
"grad_norm": 0.3099493980407715,
"learning_rate": 1.2265695866027849e-05,
"loss": 0.7373262047767639,
"step": 444
},
{
"epoch": 3.3533834586466167,
"grad_norm": 0.2395428866147995,
"learning_rate": 1.1949256053450243e-05,
"loss": 0.5868238806724548,
"step": 446
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.1869363635778427,
"learning_rate": 1.1632190401322987e-05,
"loss": 0.8645040392875671,
"step": 448
},
{
"epoch": 3.3834586466165413,
"grad_norm": 0.37590962648391724,
"learning_rate": 1.1314940600355459e-05,
"loss": 0.9512546062469482,
"step": 450
},
{
"epoch": 3.398496240601504,
"grad_norm": 0.2025744765996933,
"learning_rate": 1.099794859778695e-05,
"loss": 0.9128929376602173,
"step": 452
},
{
"epoch": 3.4135338345864663,
"grad_norm": 0.17627839744091034,
"learning_rate": 1.0681655981728753e-05,
"loss": 0.9643846750259399,
"step": 454
},
{
"epoch": 3.4285714285714284,
"grad_norm": 1.0564072132110596,
"learning_rate": 1.0366503366006477e-05,
"loss": 0.8578605055809021,
"step": 456
},
{
"epoch": 3.443609022556391,
"grad_norm": 0.3315916061401367,
"learning_rate": 1.0052929776359614e-05,
"loss": 0.5702735781669617,
"step": 458
},
{
"epoch": 3.4586466165413534,
"grad_norm": 0.1655089259147644,
"learning_rate": 9.741372038853437e-06,
"loss": 0.9215094447135925,
"step": 460
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.26911604404449463,
"learning_rate": 9.432264171355117e-06,
"loss": 0.8250299096107483,
"step": 462
},
{
"epoch": 3.488721804511278,
"grad_norm": 0.22876925766468048,
"learning_rate": 9.126036778921812e-06,
"loss": 0.8027570843696594,
"step": 464
},
{
"epoch": 3.5037593984962405,
"grad_norm": 0.172451913356781,
"learning_rate": 8.8231164539431e-06,
"loss": 0.8936192989349365,
"step": 466
},
{
"epoch": 3.518796992481203,
"grad_norm": 0.2068740874528885,
"learning_rate": 8.52392518187319e-06,
"loss": 0.6892407536506653,
"step": 468
},
{
"epoch": 3.5338345864661656,
"grad_norm": 0.18001121282577515,
"learning_rate": 8.228879753380818e-06,
"loss": 0.7898232340812683,
"step": 470
},
{
"epoch": 3.548872180451128,
"grad_norm": 0.19327786564826965,
"learning_rate": 7.938391183735972e-06,
"loss": 0.7635357975959778,
"step": 472
},
{
"epoch": 3.56390977443609,
"grad_norm": 0.20077842473983765,
"learning_rate": 7.65286414024186e-06,
"loss": 0.8505274653434753,
"step": 474
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.19398260116577148,
"learning_rate": 7.3726963785101094e-06,
"loss": 0.38889795541763306,
"step": 476
},
{
"epoch": 3.593984962406015,
"grad_norm": 0.14551198482513428,
"learning_rate": 7.098278188364236e-06,
"loss": 0.365261048078537,
"step": 478
},
{
"epoch": 3.6090225563909772,
"grad_norm": 0.17524459958076477,
"learning_rate": 6.829991850143528e-06,
"loss": 0.9008171558380127,
"step": 480
},
{
"epoch": 3.6240601503759398,
"grad_norm": 0.3653651475906372,
"learning_rate": 6.5682111021644835e-06,
"loss": 0.900292158126831,
"step": 482
},
{
"epoch": 3.6390977443609023,
"grad_norm": 0.19287094473838806,
"learning_rate": 6.313300620081777e-06,
"loss": 0.7757176160812378,
"step": 484
},
{
"epoch": 3.654135338345865,
"grad_norm": 0.20128026604652405,
"learning_rate": 6.065615508874151e-06,
"loss": 0.3761276602745056,
"step": 486
},
{
"epoch": 3.6691729323308273,
"grad_norm": 0.1511046588420868,
"learning_rate": 5.82550080816269e-06,
"loss": 0.7676253318786621,
"step": 488
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.13384103775024414,
"learning_rate": 5.593291011550776e-06,
"loss": 0.9202339053153992,
"step": 490
},
{
"epoch": 3.699248120300752,
"grad_norm": 0.4986729025840759,
"learning_rate": 5.3693096006552025e-06,
"loss": 0.970383882522583,
"step": 492
},
{
"epoch": 3.7142857142857144,
"grad_norm": 0.5382466316223145,
"learning_rate": 5.15386859447772e-06,
"loss": 0.48905929923057556,
"step": 494
},
{
"epoch": 3.7293233082706765,
"grad_norm": 0.14699164032936096,
"learning_rate": 4.947268114744566e-06,
"loss": 0.9869436025619507,
"step": 496
},
{
"epoch": 3.744360902255639,
"grad_norm": 0.17605887353420258,
"learning_rate": 4.749795967819596e-06,
"loss": 1.0997637510299683,
"step": 498
},
{
"epoch": 3.7593984962406015,
"grad_norm": 0.24531900882720947,
"learning_rate": 4.561727243773504e-06,
"loss": 0.9353739023208618,
"step": 500
},
{
"epoch": 3.774436090225564,
"grad_norm": 0.23771578073501587,
"learning_rate": 4.383323933167479e-06,
"loss": 0.7048070430755615,
"step": 502
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.1848079115152359,
"learning_rate": 4.214834562085255e-06,
"loss": 0.6421667337417603,
"step": 504
},
{
"epoch": 3.8045112781954886,
"grad_norm": 1.46511971950531,
"learning_rate": 4.056493845922015e-06,
"loss": 0.5478267669677734,
"step": 506
},
{
"epoch": 3.819548872180451,
"grad_norm": 0.1415521800518036,
"learning_rate": 3.908522362412276e-06,
"loss": 0.8992588520050049,
"step": 508
},
{
"epoch": 3.8345864661654137,
"grad_norm": 0.6491537690162659,
"learning_rate": 3.7711262443524467e-06,
"loss": 0.591274082660675,
"step": 510
},
{
"epoch": 3.8496240601503757,
"grad_norm": 0.2289588451385498,
"learning_rate": 3.6444968924459545e-06,
"loss": 0.771586000919342,
"step": 512
},
{
"epoch": 3.8646616541353382,
"grad_norm": 0.1354115605354309,
"learning_rate": 3.5288107086710545e-06,
"loss": 0.7341738939285278,
"step": 514
},
{
"epoch": 3.8796992481203008,
"grad_norm": 0.1334567666053772,
"learning_rate": 3.4242288505426928e-06,
"loss": 1.0383778810501099,
"step": 516
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.17545144259929657,
"learning_rate": 3.3308970066108413e-06,
"loss": 0.5621640682220459,
"step": 518
},
{
"epoch": 3.909774436090226,
"grad_norm": 0.09879021346569061,
"learning_rate": 3.2489451935079473e-06,
"loss": 0.9413120150566101,
"step": 520
},
{
"epoch": 3.924812030075188,
"grad_norm": 0.2358972579240799,
"learning_rate": 3.1784875748282608e-06,
"loss": 1.0549259185791016,
"step": 522
},
{
"epoch": 3.9398496240601504,
"grad_norm": 0.1280510276556015,
"learning_rate": 3.119622302091404e-06,
"loss": 1.0559005737304688,
"step": 524
},
{
"epoch": 3.954887218045113,
"grad_norm": 0.24243058264255524,
"learning_rate": 3.072431378011632e-06,
"loss": 0.9401967525482178,
"step": 526
},
{
"epoch": 3.969924812030075,
"grad_norm": 0.1816936582326889,
"learning_rate": 3.0369805422633432e-06,
"loss": 0.6218282580375671,
"step": 528
},
{
"epoch": 3.9849624060150375,
"grad_norm": 0.4231378734111786,
"learning_rate": 3.0133191799019103e-06,
"loss": 0.5278748869895935,
"step": 530
},
{
"epoch": 4.0,
"grad_norm": 0.3513203263282776,
"learning_rate": 3.001480252567475e-06,
"loss": 0.6851533651351929,
"step": 532
},
{
"epoch": 4.0,
"step": 532,
"total_flos": 3.800554784105169e+18,
"train_loss": 0.9838281438538903,
"train_runtime": 15715.7329,
"train_samples_per_second": 4.062,
"train_steps_per_second": 0.034
}
],
"logging_steps": 2,
"max_steps": 532,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 260,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.800554784105169e+18,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}