gyung's picture
Upload Gemma 4 native SFT 1Epoch checkpoint
ff1d195 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1021,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0009802720254870727,
"grad_norm": 444.2529296875,
"learning_rate": 0.0,
"loss": 10.736241340637207,
"step": 1
},
{
"epoch": 0.0019605440509741453,
"grad_norm": 99.79141235351562,
"learning_rate": 1.6129032258064518e-07,
"loss": 10.998506546020508,
"step": 2
},
{
"epoch": 0.002940816076461218,
"grad_norm": 94.0584487915039,
"learning_rate": 3.2258064516129035e-07,
"loss": 10.300856590270996,
"step": 3
},
{
"epoch": 0.003921088101948291,
"grad_norm": 754.8390502929688,
"learning_rate": 4.838709677419355e-07,
"loss": 10.12170696258545,
"step": 4
},
{
"epoch": 0.004901360127435363,
"grad_norm": 110.3061294555664,
"learning_rate": 6.451612903225807e-07,
"loss": 10.923585891723633,
"step": 5
},
{
"epoch": 0.005881632152922436,
"grad_norm": 80.18559265136719,
"learning_rate": 8.064516129032258e-07,
"loss": 9.619321823120117,
"step": 6
},
{
"epoch": 0.006861904178409509,
"grad_norm": 120.46674346923828,
"learning_rate": 9.67741935483871e-07,
"loss": 8.93388843536377,
"step": 7
},
{
"epoch": 0.007842176203896581,
"grad_norm": 155.403564453125,
"learning_rate": 1.1290322580645162e-06,
"loss": 9.105071067810059,
"step": 8
},
{
"epoch": 0.008822448229383655,
"grad_norm": 58.01957702636719,
"learning_rate": 1.2903225806451614e-06,
"loss": 7.533741474151611,
"step": 9
},
{
"epoch": 0.009802720254870726,
"grad_norm": 31.391273498535156,
"learning_rate": 1.4516129032258066e-06,
"loss": 7.134483337402344,
"step": 10
},
{
"epoch": 0.0107829922803578,
"grad_norm": 26.58418083190918,
"learning_rate": 1.6129032258064516e-06,
"loss": 6.369547367095947,
"step": 11
},
{
"epoch": 0.011763264305844873,
"grad_norm": 35.25242233276367,
"learning_rate": 1.774193548387097e-06,
"loss": 5.254162311553955,
"step": 12
},
{
"epoch": 0.012743536331331944,
"grad_norm": 46.21750259399414,
"learning_rate": 1.935483870967742e-06,
"loss": 4.887372016906738,
"step": 13
},
{
"epoch": 0.013723808356819018,
"grad_norm": 11.279609680175781,
"learning_rate": 2.096774193548387e-06,
"loss": 5.321861267089844,
"step": 14
},
{
"epoch": 0.01470408038230609,
"grad_norm": 15.513973236083984,
"learning_rate": 2.2580645161290324e-06,
"loss": 4.800952911376953,
"step": 15
},
{
"epoch": 0.015684352407793162,
"grad_norm": 9.2393798828125,
"learning_rate": 2.4193548387096776e-06,
"loss": 4.0728983879089355,
"step": 16
},
{
"epoch": 0.016664624433280234,
"grad_norm": 5.680278778076172,
"learning_rate": 2.580645161290323e-06,
"loss": 3.8023455142974854,
"step": 17
},
{
"epoch": 0.01764489645876731,
"grad_norm": 11.160239219665527,
"learning_rate": 2.7419354838709676e-06,
"loss": 3.5728161334991455,
"step": 18
},
{
"epoch": 0.01862516848425438,
"grad_norm": 9.689987182617188,
"learning_rate": 2.903225806451613e-06,
"loss": 3.6294851303100586,
"step": 19
},
{
"epoch": 0.019605440509741452,
"grad_norm": 73.86099243164062,
"learning_rate": 3.0645161290322584e-06,
"loss": 3.190464496612549,
"step": 20
},
{
"epoch": 0.020585712535228527,
"grad_norm": 5.178689002990723,
"learning_rate": 3.225806451612903e-06,
"loss": 3.054841995239258,
"step": 21
},
{
"epoch": 0.0215659845607156,
"grad_norm": 17.172805786132812,
"learning_rate": 3.3870967741935484e-06,
"loss": 2.9779253005981445,
"step": 22
},
{
"epoch": 0.02254625658620267,
"grad_norm": 5.4300537109375,
"learning_rate": 3.548387096774194e-06,
"loss": 3.034024238586426,
"step": 23
},
{
"epoch": 0.023526528611689745,
"grad_norm": 9.983994483947754,
"learning_rate": 3.7096774193548392e-06,
"loss": 2.7238316535949707,
"step": 24
},
{
"epoch": 0.024506800637176817,
"grad_norm": 5.650508403778076,
"learning_rate": 3.870967741935484e-06,
"loss": 2.50161075592041,
"step": 25
},
{
"epoch": 0.02548707266266389,
"grad_norm": 55.20005798339844,
"learning_rate": 4.032258064516129e-06,
"loss": 2.6121983528137207,
"step": 26
},
{
"epoch": 0.026467344688150964,
"grad_norm": 34.6417121887207,
"learning_rate": 4.193548387096774e-06,
"loss": 2.351276397705078,
"step": 27
},
{
"epoch": 0.027447616713638035,
"grad_norm": 4.153648853302002,
"learning_rate": 4.35483870967742e-06,
"loss": 2.1463727951049805,
"step": 28
},
{
"epoch": 0.028427888739125107,
"grad_norm": 5.739168167114258,
"learning_rate": 4.516129032258065e-06,
"loss": 2.129469871520996,
"step": 29
},
{
"epoch": 0.02940816076461218,
"grad_norm": 13.234610557556152,
"learning_rate": 4.67741935483871e-06,
"loss": 2.023747444152832,
"step": 30
},
{
"epoch": 0.030388432790099253,
"grad_norm": 6.241724491119385,
"learning_rate": 4.838709677419355e-06,
"loss": 2.0241587162017822,
"step": 31
},
{
"epoch": 0.031368704815586325,
"grad_norm": 2.4811441898345947,
"learning_rate": 5e-06,
"loss": 1.9200693368911743,
"step": 32
},
{
"epoch": 0.0323489768410734,
"grad_norm": 18.006078720092773,
"learning_rate": 5.161290322580646e-06,
"loss": 2.048367977142334,
"step": 33
},
{
"epoch": 0.03332924886656047,
"grad_norm": 2.888336181640625,
"learning_rate": 5.322580645161291e-06,
"loss": 1.8217453956604004,
"step": 34
},
{
"epoch": 0.03430952089204754,
"grad_norm": 4.508528709411621,
"learning_rate": 5.483870967741935e-06,
"loss": 1.9134241342544556,
"step": 35
},
{
"epoch": 0.03528979291753462,
"grad_norm": 5.45286226272583,
"learning_rate": 5.645161290322582e-06,
"loss": 1.5955469608306885,
"step": 36
},
{
"epoch": 0.036270064943021686,
"grad_norm": 2.6666250228881836,
"learning_rate": 5.806451612903226e-06,
"loss": 1.5563428401947021,
"step": 37
},
{
"epoch": 0.03725033696850876,
"grad_norm": 10.96904468536377,
"learning_rate": 5.967741935483872e-06,
"loss": 1.7565765380859375,
"step": 38
},
{
"epoch": 0.038230608993995836,
"grad_norm": 8.063737869262695,
"learning_rate": 6.129032258064517e-06,
"loss": 1.559302806854248,
"step": 39
},
{
"epoch": 0.039210881019482904,
"grad_norm": 3.7466046810150146,
"learning_rate": 6.290322580645162e-06,
"loss": 1.3841321468353271,
"step": 40
},
{
"epoch": 0.04019115304496998,
"grad_norm": 4.077739715576172,
"learning_rate": 6.451612903225806e-06,
"loss": 1.4145723581314087,
"step": 41
},
{
"epoch": 0.041171425070457054,
"grad_norm": 2.0939736366271973,
"learning_rate": 6.612903225806452e-06,
"loss": 1.266073226928711,
"step": 42
},
{
"epoch": 0.04215169709594412,
"grad_norm": 1.8341953754425049,
"learning_rate": 6.774193548387097e-06,
"loss": 1.4071226119995117,
"step": 43
},
{
"epoch": 0.0431319691214312,
"grad_norm": 27.189706802368164,
"learning_rate": 6.935483870967743e-06,
"loss": 1.3583588600158691,
"step": 44
},
{
"epoch": 0.04411224114691827,
"grad_norm": 2.82374906539917,
"learning_rate": 7.096774193548388e-06,
"loss": 1.1239447593688965,
"step": 45
},
{
"epoch": 0.04509251317240534,
"grad_norm": 3.163405418395996,
"learning_rate": 7.258064516129033e-06,
"loss": 1.2555480003356934,
"step": 46
},
{
"epoch": 0.046072785197892416,
"grad_norm": 9.612832069396973,
"learning_rate": 7.4193548387096784e-06,
"loss": 1.2280371189117432,
"step": 47
},
{
"epoch": 0.04705305722337949,
"grad_norm": 2.274650812149048,
"learning_rate": 7.580645161290323e-06,
"loss": 1.2739315032958984,
"step": 48
},
{
"epoch": 0.04803332924886656,
"grad_norm": 8.578704833984375,
"learning_rate": 7.741935483870968e-06,
"loss": 1.4535212516784668,
"step": 49
},
{
"epoch": 0.049013601274353634,
"grad_norm": 33.340606689453125,
"learning_rate": 7.903225806451613e-06,
"loss": 1.4000167846679688,
"step": 50
},
{
"epoch": 0.04999387329984071,
"grad_norm": 6.348473072052002,
"learning_rate": 8.064516129032258e-06,
"loss": 1.6060235500335693,
"step": 51
},
{
"epoch": 0.05097414532532778,
"grad_norm": 2.6229915618896484,
"learning_rate": 8.225806451612904e-06,
"loss": 1.1890780925750732,
"step": 52
},
{
"epoch": 0.05195441735081485,
"grad_norm": 2.0519227981567383,
"learning_rate": 8.387096774193549e-06,
"loss": 1.3458662033081055,
"step": 53
},
{
"epoch": 0.05293468937630193,
"grad_norm": 1.8942406177520752,
"learning_rate": 8.548387096774194e-06,
"loss": 1.1449463367462158,
"step": 54
},
{
"epoch": 0.053914961401788995,
"grad_norm": 3.5907156467437744,
"learning_rate": 8.70967741935484e-06,
"loss": 1.2168039083480835,
"step": 55
},
{
"epoch": 0.05489523342727607,
"grad_norm": 5.486432075500488,
"learning_rate": 8.870967741935484e-06,
"loss": 1.039926528930664,
"step": 56
},
{
"epoch": 0.05587550545276314,
"grad_norm": 2.0273101329803467,
"learning_rate": 9.03225806451613e-06,
"loss": 1.225830078125,
"step": 57
},
{
"epoch": 0.05685577747825021,
"grad_norm": 3.0957894325256348,
"learning_rate": 9.193548387096775e-06,
"loss": 1.272838830947876,
"step": 58
},
{
"epoch": 0.05783604950373729,
"grad_norm": 5.01629114151001,
"learning_rate": 9.35483870967742e-06,
"loss": 1.0743775367736816,
"step": 59
},
{
"epoch": 0.05881632152922436,
"grad_norm": 4.781734466552734,
"learning_rate": 9.516129032258065e-06,
"loss": 1.0559556484222412,
"step": 60
},
{
"epoch": 0.05979659355471143,
"grad_norm": 2.8775441646575928,
"learning_rate": 9.67741935483871e-06,
"loss": 1.2778871059417725,
"step": 61
},
{
"epoch": 0.06077686558019851,
"grad_norm": 1.9450877904891968,
"learning_rate": 9.838709677419356e-06,
"loss": 1.1817667484283447,
"step": 62
},
{
"epoch": 0.061757137605685575,
"grad_norm": 8.134469985961914,
"learning_rate": 1e-05,
"loss": 1.0630282163619995,
"step": 63
},
{
"epoch": 0.06273740963117265,
"grad_norm": 1.958270788192749,
"learning_rate": 9.999993706252978e-06,
"loss": 1.2392271757125854,
"step": 64
},
{
"epoch": 0.06371768165665972,
"grad_norm": 2.5432846546173096,
"learning_rate": 9.999974825027756e-06,
"loss": 1.108931541442871,
"step": 65
},
{
"epoch": 0.0646979536821468,
"grad_norm": 2.372438669204712,
"learning_rate": 9.999943356371867e-06,
"loss": 0.9794902801513672,
"step": 66
},
{
"epoch": 0.06567822570763387,
"grad_norm": 1.4607375860214233,
"learning_rate": 9.999899300364534e-06,
"loss": 0.8074288368225098,
"step": 67
},
{
"epoch": 0.06665849773312094,
"grad_norm": 3.0865185260772705,
"learning_rate": 9.999842657116667e-06,
"loss": 0.9628023505210876,
"step": 68
},
{
"epoch": 0.06763876975860801,
"grad_norm": 1.5494884252548218,
"learning_rate": 9.999773426770864e-06,
"loss": 0.9801100492477417,
"step": 69
},
{
"epoch": 0.06861904178409509,
"grad_norm": 1.4236968755722046,
"learning_rate": 9.999691609501418e-06,
"loss": 1.0176517963409424,
"step": 70
},
{
"epoch": 0.06959931380958216,
"grad_norm": 1.5794986486434937,
"learning_rate": 9.999597205514298e-06,
"loss": 0.970253050327301,
"step": 71
},
{
"epoch": 0.07057958583506924,
"grad_norm": 2.677732229232788,
"learning_rate": 9.999490215047167e-06,
"loss": 0.8624630570411682,
"step": 72
},
{
"epoch": 0.07155985786055631,
"grad_norm": 1.9615784883499146,
"learning_rate": 9.999370638369377e-06,
"loss": 1.0122735500335693,
"step": 73
},
{
"epoch": 0.07254012988604337,
"grad_norm": 3.2134084701538086,
"learning_rate": 9.999238475781957e-06,
"loss": 1.0808191299438477,
"step": 74
},
{
"epoch": 0.07352040191153045,
"grad_norm": 1.50729501247406,
"learning_rate": 9.99909372761763e-06,
"loss": 1.0038952827453613,
"step": 75
},
{
"epoch": 0.07450067393701752,
"grad_norm": 1.5557780265808105,
"learning_rate": 9.998936394240798e-06,
"loss": 1.1535766124725342,
"step": 76
},
{
"epoch": 0.0754809459625046,
"grad_norm": 2.1890480518341064,
"learning_rate": 9.998766476047546e-06,
"loss": 0.917161762714386,
"step": 77
},
{
"epoch": 0.07646121798799167,
"grad_norm": 1.307956337928772,
"learning_rate": 9.998583973465647e-06,
"loss": 0.9573119878768921,
"step": 78
},
{
"epoch": 0.07744149001347873,
"grad_norm": 1.9047560691833496,
"learning_rate": 9.998388886954546e-06,
"loss": 0.8896398544311523,
"step": 79
},
{
"epoch": 0.07842176203896581,
"grad_norm": 3.1031126976013184,
"learning_rate": 9.998181217005377e-06,
"loss": 0.9498454332351685,
"step": 80
},
{
"epoch": 0.07940203406445288,
"grad_norm": 2.284217357635498,
"learning_rate": 9.997960964140946e-06,
"loss": 0.9973191022872925,
"step": 81
},
{
"epoch": 0.08038230608993996,
"grad_norm": 1.6305043697357178,
"learning_rate": 9.997728128915742e-06,
"loss": 0.935388445854187,
"step": 82
},
{
"epoch": 0.08136257811542703,
"grad_norm": 1.501710295677185,
"learning_rate": 9.997482711915926e-06,
"loss": 1.019201397895813,
"step": 83
},
{
"epoch": 0.08234285014091411,
"grad_norm": 1.4190490245819092,
"learning_rate": 9.997224713759335e-06,
"loss": 1.099223256111145,
"step": 84
},
{
"epoch": 0.08332312216640117,
"grad_norm": 1.886734127998352,
"learning_rate": 9.99695413509548e-06,
"loss": 0.9991997480392456,
"step": 85
},
{
"epoch": 0.08430339419188825,
"grad_norm": 1.508866548538208,
"learning_rate": 9.99667097660554e-06,
"loss": 0.9771667122840881,
"step": 86
},
{
"epoch": 0.08528366621737532,
"grad_norm": 2.9389193058013916,
"learning_rate": 9.99637523900237e-06,
"loss": 0.9765088558197021,
"step": 87
},
{
"epoch": 0.0862639382428624,
"grad_norm": 1.4393686056137085,
"learning_rate": 9.996066923030484e-06,
"loss": 0.9267507791519165,
"step": 88
},
{
"epoch": 0.08724421026834947,
"grad_norm": 1.0979642868041992,
"learning_rate": 9.995746029466071e-06,
"loss": 0.8212970495223999,
"step": 89
},
{
"epoch": 0.08822448229383655,
"grad_norm": 1.4870240688323975,
"learning_rate": 9.995412559116979e-06,
"loss": 0.8318418264389038,
"step": 90
},
{
"epoch": 0.0892047543193236,
"grad_norm": 1.707983136177063,
"learning_rate": 9.99506651282272e-06,
"loss": 0.9875527620315552,
"step": 91
},
{
"epoch": 0.09018502634481068,
"grad_norm": 1.1789133548736572,
"learning_rate": 9.994707891454466e-06,
"loss": 0.8298413753509521,
"step": 92
},
{
"epoch": 0.09116529837029776,
"grad_norm": 1.3218693733215332,
"learning_rate": 9.994336695915041e-06,
"loss": 1.0132757425308228,
"step": 93
},
{
"epoch": 0.09214557039578483,
"grad_norm": 1.0833605527877808,
"learning_rate": 9.993952927138933e-06,
"loss": 1.0068328380584717,
"step": 94
},
{
"epoch": 0.0931258424212719,
"grad_norm": 1.2598189115524292,
"learning_rate": 9.993556586092281e-06,
"loss": 0.8056250214576721,
"step": 95
},
{
"epoch": 0.09410611444675898,
"grad_norm": 1.4173874855041504,
"learning_rate": 9.993147673772869e-06,
"loss": 1.070406198501587,
"step": 96
},
{
"epoch": 0.09508638647224604,
"grad_norm": 1.3954414129257202,
"learning_rate": 9.992726191210139e-06,
"loss": 0.8578159213066101,
"step": 97
},
{
"epoch": 0.09606665849773312,
"grad_norm": 1.028037190437317,
"learning_rate": 9.992292139465166e-06,
"loss": 0.8095722794532776,
"step": 98
},
{
"epoch": 0.09704693052322019,
"grad_norm": 1.4925620555877686,
"learning_rate": 9.991845519630679e-06,
"loss": 1.0127127170562744,
"step": 99
},
{
"epoch": 0.09802720254870727,
"grad_norm": 1.0345618724822998,
"learning_rate": 9.991386332831042e-06,
"loss": 0.8328266143798828,
"step": 100
},
{
"epoch": 0.09900747457419434,
"grad_norm": 0.8846564292907715,
"learning_rate": 9.990914580222258e-06,
"loss": 0.8187386989593506,
"step": 101
},
{
"epoch": 0.09998774659968142,
"grad_norm": 1.276199460029602,
"learning_rate": 9.990430262991962e-06,
"loss": 1.0163452625274658,
"step": 102
},
{
"epoch": 0.10096801862516848,
"grad_norm": 0.9905869364738464,
"learning_rate": 9.989933382359423e-06,
"loss": 0.7760695219039917,
"step": 103
},
{
"epoch": 0.10194829065065555,
"grad_norm": 1.4109216928482056,
"learning_rate": 9.989423939575538e-06,
"loss": 0.9511476755142212,
"step": 104
},
{
"epoch": 0.10292856267614263,
"grad_norm": 1.2112044095993042,
"learning_rate": 9.988901935922826e-06,
"loss": 0.7780454754829407,
"step": 105
},
{
"epoch": 0.1039088347016297,
"grad_norm": 1.1124036312103271,
"learning_rate": 9.988367372715434e-06,
"loss": 0.9685283899307251,
"step": 106
},
{
"epoch": 0.10488910672711678,
"grad_norm": 0.8848640322685242,
"learning_rate": 9.987820251299121e-06,
"loss": 0.7924320697784424,
"step": 107
},
{
"epoch": 0.10586937875260385,
"grad_norm": 1.4570457935333252,
"learning_rate": 9.987260573051268e-06,
"loss": 0.8897344470024109,
"step": 108
},
{
"epoch": 0.10684965077809092,
"grad_norm": 0.8776848316192627,
"learning_rate": 9.986688339380863e-06,
"loss": 0.7704064846038818,
"step": 109
},
{
"epoch": 0.10782992280357799,
"grad_norm": 0.9722417593002319,
"learning_rate": 9.9861035517285e-06,
"loss": 0.8362361192703247,
"step": 110
},
{
"epoch": 0.10881019482906507,
"grad_norm": 1.0186680555343628,
"learning_rate": 9.985506211566388e-06,
"loss": 0.827409029006958,
"step": 111
},
{
"epoch": 0.10979046685455214,
"grad_norm": 0.8622536659240723,
"learning_rate": 9.984896320398326e-06,
"loss": 0.9348365068435669,
"step": 112
},
{
"epoch": 0.11077073888003922,
"grad_norm": 0.9350709915161133,
"learning_rate": 9.984273879759713e-06,
"loss": 0.7396246194839478,
"step": 113
},
{
"epoch": 0.11175101090552628,
"grad_norm": 1.7236732244491577,
"learning_rate": 9.983638891217544e-06,
"loss": 1.1119165420532227,
"step": 114
},
{
"epoch": 0.11273128293101335,
"grad_norm": 0.9292817115783691,
"learning_rate": 9.982991356370404e-06,
"loss": 0.951391339302063,
"step": 115
},
{
"epoch": 0.11371155495650043,
"grad_norm": 2.126209259033203,
"learning_rate": 9.982331276848457e-06,
"loss": 1.053914189338684,
"step": 116
},
{
"epoch": 0.1146918269819875,
"grad_norm": 0.8694857954978943,
"learning_rate": 9.981658654313458e-06,
"loss": 0.6709041595458984,
"step": 117
},
{
"epoch": 0.11567209900747458,
"grad_norm": 1.5078258514404297,
"learning_rate": 9.980973490458728e-06,
"loss": 0.8591851592063904,
"step": 118
},
{
"epoch": 0.11665237103296165,
"grad_norm": 1.173314094543457,
"learning_rate": 9.98027578700917e-06,
"loss": 0.9200767278671265,
"step": 119
},
{
"epoch": 0.11763264305844871,
"grad_norm": 1.153979778289795,
"learning_rate": 9.979565545721249e-06,
"loss": 0.7959837913513184,
"step": 120
},
{
"epoch": 0.11861291508393579,
"grad_norm": 2.0307085514068604,
"learning_rate": 9.978842768382999e-06,
"loss": 0.8494062423706055,
"step": 121
},
{
"epoch": 0.11959318710942286,
"grad_norm": 2.860264539718628,
"learning_rate": 9.978107456814007e-06,
"loss": 0.8416390419006348,
"step": 122
},
{
"epoch": 0.12057345913490994,
"grad_norm": 1.8330377340316772,
"learning_rate": 9.977359612865424e-06,
"loss": 0.9158598184585571,
"step": 123
},
{
"epoch": 0.12155373116039701,
"grad_norm": 1.220954179763794,
"learning_rate": 9.976599238419943e-06,
"loss": 0.8735285997390747,
"step": 124
},
{
"epoch": 0.12253400318588409,
"grad_norm": 1.7368972301483154,
"learning_rate": 9.975826335391808e-06,
"loss": 1.2024664878845215,
"step": 125
},
{
"epoch": 0.12351427521137115,
"grad_norm": 4.530630111694336,
"learning_rate": 9.975040905726799e-06,
"loss": 0.8800772428512573,
"step": 126
},
{
"epoch": 0.12449454723685822,
"grad_norm": 1.2298669815063477,
"learning_rate": 9.974242951402236e-06,
"loss": 0.8384938836097717,
"step": 127
},
{
"epoch": 0.1254748192623453,
"grad_norm": 3.0307226181030273,
"learning_rate": 9.973432474426968e-06,
"loss": 0.7943012118339539,
"step": 128
},
{
"epoch": 0.12645509128783236,
"grad_norm": 1.2162606716156006,
"learning_rate": 9.972609476841368e-06,
"loss": 0.906460165977478,
"step": 129
},
{
"epoch": 0.12743536331331945,
"grad_norm": 1.1699104309082031,
"learning_rate": 9.971773960717333e-06,
"loss": 0.7454752326011658,
"step": 130
},
{
"epoch": 0.1284156353388065,
"grad_norm": 0.871055006980896,
"learning_rate": 9.970925928158275e-06,
"loss": 0.7852927446365356,
"step": 131
},
{
"epoch": 0.1293959073642936,
"grad_norm": 0.7216456532478333,
"learning_rate": 9.970065381299112e-06,
"loss": 0.7714865207672119,
"step": 132
},
{
"epoch": 0.13037617938978066,
"grad_norm": 1.2287877798080444,
"learning_rate": 9.969192322306271e-06,
"loss": 0.8582690954208374,
"step": 133
},
{
"epoch": 0.13135645141526775,
"grad_norm": 0.7745174765586853,
"learning_rate": 9.968306753377679e-06,
"loss": 0.8014340400695801,
"step": 134
},
{
"epoch": 0.1323367234407548,
"grad_norm": 1.20053231716156,
"learning_rate": 9.96740867674275e-06,
"loss": 1.02126145362854,
"step": 135
},
{
"epoch": 0.13331699546624187,
"grad_norm": 1.1115248203277588,
"learning_rate": 9.966498094662398e-06,
"loss": 0.8542506694793701,
"step": 136
},
{
"epoch": 0.13429726749172896,
"grad_norm": 0.7779384255409241,
"learning_rate": 9.965575009429006e-06,
"loss": 0.6853510141372681,
"step": 137
},
{
"epoch": 0.13527753951721602,
"grad_norm": 0.8397386074066162,
"learning_rate": 9.964639423366442e-06,
"loss": 0.7850790023803711,
"step": 138
},
{
"epoch": 0.1362578115427031,
"grad_norm": 3.3384697437286377,
"learning_rate": 9.963691338830045e-06,
"loss": 0.8284836411476135,
"step": 139
},
{
"epoch": 0.13723808356819017,
"grad_norm": 1.1726006269454956,
"learning_rate": 9.962730758206612e-06,
"loss": 0.850119948387146,
"step": 140
},
{
"epoch": 0.13821835559367723,
"grad_norm": 1.0025110244750977,
"learning_rate": 9.961757683914406e-06,
"loss": 0.9284964203834534,
"step": 141
},
{
"epoch": 0.13919862761916432,
"grad_norm": 0.8656743764877319,
"learning_rate": 9.960772118403142e-06,
"loss": 0.7905132174491882,
"step": 142
},
{
"epoch": 0.14017889964465138,
"grad_norm": 0.7995918393135071,
"learning_rate": 9.959774064153977e-06,
"loss": 0.6824674606323242,
"step": 143
},
{
"epoch": 0.14115917167013847,
"grad_norm": 0.9554956555366516,
"learning_rate": 9.958763523679515e-06,
"loss": 0.8615709543228149,
"step": 144
},
{
"epoch": 0.14213944369562553,
"grad_norm": 0.8376150727272034,
"learning_rate": 9.957740499523787e-06,
"loss": 0.782550573348999,
"step": 145
},
{
"epoch": 0.14311971572111262,
"grad_norm": 1.0808218717575073,
"learning_rate": 9.956704994262256e-06,
"loss": 0.8757920265197754,
"step": 146
},
{
"epoch": 0.14409998774659968,
"grad_norm": 0.8293938040733337,
"learning_rate": 9.955657010501807e-06,
"loss": 0.8046925067901611,
"step": 147
},
{
"epoch": 0.14508025977208674,
"grad_norm": 0.9780465960502625,
"learning_rate": 9.954596550880735e-06,
"loss": 0.769169270992279,
"step": 148
},
{
"epoch": 0.14606053179757383,
"grad_norm": 0.6602773070335388,
"learning_rate": 9.95352361806875e-06,
"loss": 0.6232231855392456,
"step": 149
},
{
"epoch": 0.1470408038230609,
"grad_norm": 1.0579054355621338,
"learning_rate": 9.952438214766956e-06,
"loss": 0.7667106986045837,
"step": 150
},
{
"epoch": 0.14802107584854798,
"grad_norm": 8.054926872253418,
"learning_rate": 9.951340343707852e-06,
"loss": 0.9584368467330933,
"step": 151
},
{
"epoch": 0.14900134787403504,
"grad_norm": 2.2190775871276855,
"learning_rate": 9.950230007655333e-06,
"loss": 0.6349459886550903,
"step": 152
},
{
"epoch": 0.1499816198995221,
"grad_norm": 2.298706531524658,
"learning_rate": 9.949107209404664e-06,
"loss": 0.9834344983100891,
"step": 153
},
{
"epoch": 0.1509618919250092,
"grad_norm": 0.9804521799087524,
"learning_rate": 9.94797195178249e-06,
"loss": 0.9633491039276123,
"step": 154
},
{
"epoch": 0.15194216395049626,
"grad_norm": 0.8359074592590332,
"learning_rate": 9.946824237646823e-06,
"loss": 0.90494704246521,
"step": 155
},
{
"epoch": 0.15292243597598335,
"grad_norm": 1.383854866027832,
"learning_rate": 9.945664069887028e-06,
"loss": 1.0030543804168701,
"step": 156
},
{
"epoch": 0.1539027080014704,
"grad_norm": 0.7327127456665039,
"learning_rate": 9.944491451423829e-06,
"loss": 0.7229695916175842,
"step": 157
},
{
"epoch": 0.15488298002695747,
"grad_norm": 0.8276205658912659,
"learning_rate": 9.94330638520929e-06,
"loss": 0.7939118146896362,
"step": 158
},
{
"epoch": 0.15586325205244456,
"grad_norm": 0.7848310470581055,
"learning_rate": 9.942108874226812e-06,
"loss": 0.7627387046813965,
"step": 159
},
{
"epoch": 0.15684352407793162,
"grad_norm": 0.7724266052246094,
"learning_rate": 9.940898921491132e-06,
"loss": 0.7301936149597168,
"step": 160
},
{
"epoch": 0.1578237961034187,
"grad_norm": 1.3323100805282593,
"learning_rate": 9.9396765300483e-06,
"loss": 0.683293879032135,
"step": 161
},
{
"epoch": 0.15880406812890577,
"grad_norm": 1.311335563659668,
"learning_rate": 9.938441702975689e-06,
"loss": 0.7212080359458923,
"step": 162
},
{
"epoch": 0.15978434015439286,
"grad_norm": 0.9447430372238159,
"learning_rate": 9.937194443381972e-06,
"loss": 0.9628721475601196,
"step": 163
},
{
"epoch": 0.16076461217987992,
"grad_norm": 0.6951046586036682,
"learning_rate": 9.935934754407125e-06,
"loss": 0.7402358055114746,
"step": 164
},
{
"epoch": 0.16174488420536698,
"grad_norm": 0.7210685014724731,
"learning_rate": 9.934662639222412e-06,
"loss": 0.7613071799278259,
"step": 165
},
{
"epoch": 0.16272515623085407,
"grad_norm": 0.7304868102073669,
"learning_rate": 9.933378101030382e-06,
"loss": 0.6772556900978088,
"step": 166
},
{
"epoch": 0.16370542825634113,
"grad_norm": 0.7971092462539673,
"learning_rate": 9.93208114306486e-06,
"loss": 0.7645301818847656,
"step": 167
},
{
"epoch": 0.16468570028182822,
"grad_norm": 0.7262613773345947,
"learning_rate": 9.930771768590934e-06,
"loss": 0.7944133877754211,
"step": 168
},
{
"epoch": 0.16566597230731528,
"grad_norm": 0.5857324600219727,
"learning_rate": 9.929449980904952e-06,
"loss": 0.6576533913612366,
"step": 169
},
{
"epoch": 0.16664624433280234,
"grad_norm": 0.9272939562797546,
"learning_rate": 9.928115783334518e-06,
"loss": 0.8526591658592224,
"step": 170
},
{
"epoch": 0.16762651635828943,
"grad_norm": 0.6524718403816223,
"learning_rate": 9.926769179238467e-06,
"loss": 0.6662750244140625,
"step": 171
},
{
"epoch": 0.1686067883837765,
"grad_norm": 1.075812578201294,
"learning_rate": 9.925410172006874e-06,
"loss": 0.7937478423118591,
"step": 172
},
{
"epoch": 0.16958706040926358,
"grad_norm": 0.822924017906189,
"learning_rate": 9.924038765061042e-06,
"loss": 0.8538663387298584,
"step": 173
},
{
"epoch": 0.17056733243475064,
"grad_norm": 1.1041499376296997,
"learning_rate": 9.922654961853483e-06,
"loss": 0.8885059356689453,
"step": 174
},
{
"epoch": 0.17154760446023773,
"grad_norm": 1.4029029607772827,
"learning_rate": 9.921258765867919e-06,
"loss": 0.9263203144073486,
"step": 175
},
{
"epoch": 0.1725278764857248,
"grad_norm": 1.096556305885315,
"learning_rate": 9.919850180619274e-06,
"loss": 0.7909505367279053,
"step": 176
},
{
"epoch": 0.17350814851121185,
"grad_norm": 0.7070975303649902,
"learning_rate": 9.918429209653662e-06,
"loss": 0.7344837188720703,
"step": 177
},
{
"epoch": 0.17448842053669894,
"grad_norm": 1.3108410835266113,
"learning_rate": 9.916995856548371e-06,
"loss": 0.8295502662658691,
"step": 178
},
{
"epoch": 0.175468692562186,
"grad_norm": 0.7204867601394653,
"learning_rate": 9.915550124911866e-06,
"loss": 0.8103526830673218,
"step": 179
},
{
"epoch": 0.1764489645876731,
"grad_norm": 0.7560566067695618,
"learning_rate": 9.914092018383778e-06,
"loss": 0.7113056778907776,
"step": 180
},
{
"epoch": 0.17742923661316015,
"grad_norm": 1.012421727180481,
"learning_rate": 9.912621540634889e-06,
"loss": 0.7535778880119324,
"step": 181
},
{
"epoch": 0.1784095086386472,
"grad_norm": 0.8265753388404846,
"learning_rate": 9.91113869536712e-06,
"loss": 0.6618527173995972,
"step": 182
},
{
"epoch": 0.1793897806641343,
"grad_norm": 0.7197481989860535,
"learning_rate": 9.909643486313533e-06,
"loss": 0.7319732904434204,
"step": 183
},
{
"epoch": 0.18037005268962136,
"grad_norm": 0.8569908738136292,
"learning_rate": 9.908135917238321e-06,
"loss": 0.8834559917449951,
"step": 184
},
{
"epoch": 0.18135032471510845,
"grad_norm": 0.7561427354812622,
"learning_rate": 9.906615991936781e-06,
"loss": 0.7185438871383667,
"step": 185
},
{
"epoch": 0.1823305967405955,
"grad_norm": 0.8701999187469482,
"learning_rate": 9.905083714235326e-06,
"loss": 0.784651517868042,
"step": 186
},
{
"epoch": 0.18331086876608257,
"grad_norm": 0.8645495176315308,
"learning_rate": 9.903539087991462e-06,
"loss": 0.8580794930458069,
"step": 187
},
{
"epoch": 0.18429114079156966,
"grad_norm": 0.8038344979286194,
"learning_rate": 9.901982117093786e-06,
"loss": 0.8493238687515259,
"step": 188
},
{
"epoch": 0.18527141281705672,
"grad_norm": 0.7347437739372253,
"learning_rate": 9.900412805461968e-06,
"loss": 0.5884836316108704,
"step": 189
},
{
"epoch": 0.1862516848425438,
"grad_norm": 0.7285265922546387,
"learning_rate": 9.898831157046748e-06,
"loss": 0.6254444122314453,
"step": 190
},
{
"epoch": 0.18723195686803087,
"grad_norm": 0.848617672920227,
"learning_rate": 9.897237175829927e-06,
"loss": 0.747313380241394,
"step": 191
},
{
"epoch": 0.18821222889351796,
"grad_norm": 1.6770778894424438,
"learning_rate": 9.895630865824347e-06,
"loss": 0.691787838935852,
"step": 192
},
{
"epoch": 0.18919250091900502,
"grad_norm": 1.24128258228302,
"learning_rate": 9.894012231073895e-06,
"loss": 0.7585203647613525,
"step": 193
},
{
"epoch": 0.19017277294449209,
"grad_norm": 4.074337005615234,
"learning_rate": 9.89238127565348e-06,
"loss": 0.6597365736961365,
"step": 194
},
{
"epoch": 0.19115304496997917,
"grad_norm": 1.5159838199615479,
"learning_rate": 9.890738003669029e-06,
"loss": 0.8449289798736572,
"step": 195
},
{
"epoch": 0.19213331699546624,
"grad_norm": 0.9095994830131531,
"learning_rate": 9.88908241925748e-06,
"loss": 0.7729909420013428,
"step": 196
},
{
"epoch": 0.19311358902095332,
"grad_norm": 0.8426398038864136,
"learning_rate": 9.887414526586764e-06,
"loss": 0.7727930545806885,
"step": 197
},
{
"epoch": 0.19409386104644039,
"grad_norm": 0.6428210735321045,
"learning_rate": 9.885734329855798e-06,
"loss": 0.6707614064216614,
"step": 198
},
{
"epoch": 0.19507413307192745,
"grad_norm": 0.9163340926170349,
"learning_rate": 9.884041833294477e-06,
"loss": 0.769806981086731,
"step": 199
},
{
"epoch": 0.19605440509741454,
"grad_norm": 0.9019668698310852,
"learning_rate": 9.882337041163656e-06,
"loss": 0.7200514078140259,
"step": 200
},
{
"epoch": 0.1970346771229016,
"grad_norm": 0.8410512804985046,
"learning_rate": 9.880619957755151e-06,
"loss": 0.7845771908760071,
"step": 201
},
{
"epoch": 0.19801494914838869,
"grad_norm": 0.8458231687545776,
"learning_rate": 9.878890587391715e-06,
"loss": 0.7418243288993835,
"step": 202
},
{
"epoch": 0.19899522117387575,
"grad_norm": 0.6429169774055481,
"learning_rate": 9.877148934427037e-06,
"loss": 0.606464684009552,
"step": 203
},
{
"epoch": 0.19997549319936284,
"grad_norm": 0.5864877700805664,
"learning_rate": 9.875395003245724e-06,
"loss": 0.6223068833351135,
"step": 204
},
{
"epoch": 0.2009557652248499,
"grad_norm": 1.1887778043746948,
"learning_rate": 9.873628798263297e-06,
"loss": 0.8293617963790894,
"step": 205
},
{
"epoch": 0.20193603725033696,
"grad_norm": 0.7091985940933228,
"learning_rate": 9.871850323926178e-06,
"loss": 0.7142840027809143,
"step": 206
},
{
"epoch": 0.20291630927582405,
"grad_norm": 0.5980514287948608,
"learning_rate": 9.870059584711668e-06,
"loss": 0.6900460720062256,
"step": 207
},
{
"epoch": 0.2038965813013111,
"grad_norm": 0.6878236532211304,
"learning_rate": 9.868256585127956e-06,
"loss": 0.6946916580200195,
"step": 208
},
{
"epoch": 0.2048768533267982,
"grad_norm": 1.0756114721298218,
"learning_rate": 9.86644132971409e-06,
"loss": 0.6794236898422241,
"step": 209
},
{
"epoch": 0.20585712535228526,
"grad_norm": 0.634649932384491,
"learning_rate": 9.864613823039971e-06,
"loss": 0.7387229204177856,
"step": 210
},
{
"epoch": 0.20683739737777232,
"grad_norm": 1.0275253057479858,
"learning_rate": 9.862774069706346e-06,
"loss": 0.870427668094635,
"step": 211
},
{
"epoch": 0.2078176694032594,
"grad_norm": 0.704164981842041,
"learning_rate": 9.860922074344794e-06,
"loss": 0.7373089790344238,
"step": 212
},
{
"epoch": 0.20879794142874647,
"grad_norm": 0.8564486503601074,
"learning_rate": 9.859057841617709e-06,
"loss": 0.5971276164054871,
"step": 213
},
{
"epoch": 0.20977821345423356,
"grad_norm": 0.6043328046798706,
"learning_rate": 9.857181376218295e-06,
"loss": 0.6864156126976013,
"step": 214
},
{
"epoch": 0.21075848547972062,
"grad_norm": 0.7761573195457458,
"learning_rate": 9.855292682870552e-06,
"loss": 0.6779805421829224,
"step": 215
},
{
"epoch": 0.2117387575052077,
"grad_norm": 1.5135776996612549,
"learning_rate": 9.853391766329264e-06,
"loss": 0.6988109350204468,
"step": 216
},
{
"epoch": 0.21271902953069477,
"grad_norm": 0.6150910258293152,
"learning_rate": 9.851478631379982e-06,
"loss": 0.6663477420806885,
"step": 217
},
{
"epoch": 0.21369930155618183,
"grad_norm": 1.6363004446029663,
"learning_rate": 9.849553282839025e-06,
"loss": 0.6891165971755981,
"step": 218
},
{
"epoch": 0.21467957358166892,
"grad_norm": 0.7080655694007874,
"learning_rate": 9.847615725553457e-06,
"loss": 0.6857892870903015,
"step": 219
},
{
"epoch": 0.21565984560715598,
"grad_norm": 0.639260470867157,
"learning_rate": 9.845665964401072e-06,
"loss": 0.6488146781921387,
"step": 220
},
{
"epoch": 0.21664011763264307,
"grad_norm": 0.5393828749656677,
"learning_rate": 9.843704004290393e-06,
"loss": 0.7031430006027222,
"step": 221
},
{
"epoch": 0.21762038965813013,
"grad_norm": 0.6357012391090393,
"learning_rate": 9.841729850160652e-06,
"loss": 0.7031742334365845,
"step": 222
},
{
"epoch": 0.2186006616836172,
"grad_norm": 0.7872418165206909,
"learning_rate": 9.839743506981783e-06,
"loss": 0.6754440665245056,
"step": 223
},
{
"epoch": 0.21958093370910428,
"grad_norm": 0.6326587796211243,
"learning_rate": 9.837744979754397e-06,
"loss": 0.6297906041145325,
"step": 224
},
{
"epoch": 0.22056120573459134,
"grad_norm": 0.6477714776992798,
"learning_rate": 9.835734273509787e-06,
"loss": 0.6901752948760986,
"step": 225
},
{
"epoch": 0.22154147776007843,
"grad_norm": 0.7842188477516174,
"learning_rate": 9.833711393309902e-06,
"loss": 0.8501895666122437,
"step": 226
},
{
"epoch": 0.2225217497855655,
"grad_norm": 0.7060827016830444,
"learning_rate": 9.831676344247343e-06,
"loss": 0.7715115547180176,
"step": 227
},
{
"epoch": 0.22350202181105255,
"grad_norm": 0.7102391123771667,
"learning_rate": 9.829629131445342e-06,
"loss": 0.6443570256233215,
"step": 228
},
{
"epoch": 0.22448229383653964,
"grad_norm": 0.5093991756439209,
"learning_rate": 9.827569760057755e-06,
"loss": 0.531701922416687,
"step": 229
},
{
"epoch": 0.2254625658620267,
"grad_norm": 0.7141419053077698,
"learning_rate": 9.825498235269046e-06,
"loss": 0.741948127746582,
"step": 230
},
{
"epoch": 0.2264428378875138,
"grad_norm": 0.6522760391235352,
"learning_rate": 9.82341456229428e-06,
"loss": 0.6692137122154236,
"step": 231
},
{
"epoch": 0.22742310991300085,
"grad_norm": 0.8308090567588806,
"learning_rate": 9.821318746379097e-06,
"loss": 0.7460157871246338,
"step": 232
},
{
"epoch": 0.22840338193848794,
"grad_norm": 0.5567009449005127,
"learning_rate": 9.819210792799711e-06,
"loss": 0.6734646558761597,
"step": 233
},
{
"epoch": 0.229383653963975,
"grad_norm": 0.5670619606971741,
"learning_rate": 9.817090706862895e-06,
"loss": 0.6297429800033569,
"step": 234
},
{
"epoch": 0.23036392598946207,
"grad_norm": 0.6522438526153564,
"learning_rate": 9.814958493905962e-06,
"loss": 0.8140875101089478,
"step": 235
},
{
"epoch": 0.23134419801494915,
"grad_norm": 0.8690782189369202,
"learning_rate": 9.812814159296757e-06,
"loss": 0.7535607814788818,
"step": 236
},
{
"epoch": 0.23232447004043622,
"grad_norm": 0.7359212636947632,
"learning_rate": 9.810657708433637e-06,
"loss": 0.8144797086715698,
"step": 237
},
{
"epoch": 0.2333047420659233,
"grad_norm": 0.6369541883468628,
"learning_rate": 9.808489146745466e-06,
"loss": 0.674457848072052,
"step": 238
},
{
"epoch": 0.23428501409141037,
"grad_norm": 0.5617554187774658,
"learning_rate": 9.806308479691595e-06,
"loss": 0.668819785118103,
"step": 239
},
{
"epoch": 0.23526528611689743,
"grad_norm": 0.8029347658157349,
"learning_rate": 9.804115712761851e-06,
"loss": 0.9395022392272949,
"step": 240
},
{
"epoch": 0.23624555814238452,
"grad_norm": 0.5776801109313965,
"learning_rate": 9.801910851476524e-06,
"loss": 0.776624858379364,
"step": 241
},
{
"epoch": 0.23722583016787158,
"grad_norm": 0.6731631755828857,
"learning_rate": 9.799693901386346e-06,
"loss": 0.8745022416114807,
"step": 242
},
{
"epoch": 0.23820610219335867,
"grad_norm": 0.6632353067398071,
"learning_rate": 9.797464868072489e-06,
"loss": 0.6243095993995667,
"step": 243
},
{
"epoch": 0.23918637421884573,
"grad_norm": 0.5712333917617798,
"learning_rate": 9.79522375714654e-06,
"loss": 0.6474231481552124,
"step": 244
},
{
"epoch": 0.24016664624433282,
"grad_norm": 1.6545488834381104,
"learning_rate": 9.792970574250493e-06,
"loss": 0.7908797264099121,
"step": 245
},
{
"epoch": 0.24114691826981988,
"grad_norm": 0.5723490715026855,
"learning_rate": 9.790705325056735e-06,
"loss": 0.6788420677185059,
"step": 246
},
{
"epoch": 0.24212719029530694,
"grad_norm": 0.5819388628005981,
"learning_rate": 9.788428015268027e-06,
"loss": 0.627650797367096,
"step": 247
},
{
"epoch": 0.24310746232079403,
"grad_norm": 1.171949863433838,
"learning_rate": 9.786138650617494e-06,
"loss": 0.8022439479827881,
"step": 248
},
{
"epoch": 0.2440877343462811,
"grad_norm": 0.4696350693702698,
"learning_rate": 9.78383723686861e-06,
"loss": 0.5586154460906982,
"step": 249
},
{
"epoch": 0.24506800637176818,
"grad_norm": 0.5712862014770508,
"learning_rate": 9.781523779815178e-06,
"loss": 0.6085636615753174,
"step": 250
},
{
"epoch": 0.24604827839725524,
"grad_norm": 0.5892388820648193,
"learning_rate": 9.779198285281326e-06,
"loss": 0.6630043983459473,
"step": 251
},
{
"epoch": 0.2470285504227423,
"grad_norm": 0.6063647866249084,
"learning_rate": 9.776860759121485e-06,
"loss": 0.7851894497871399,
"step": 252
},
{
"epoch": 0.2480088224482294,
"grad_norm": 0.5152275562286377,
"learning_rate": 9.774511207220369e-06,
"loss": 0.6777390241622925,
"step": 253
},
{
"epoch": 0.24898909447371645,
"grad_norm": 0.6084205508232117,
"learning_rate": 9.772149635492979e-06,
"loss": 0.6391557455062866,
"step": 254
},
{
"epoch": 0.24996936649920354,
"grad_norm": 0.7668172121047974,
"learning_rate": 9.769776049884564e-06,
"loss": 0.662729024887085,
"step": 255
},
{
"epoch": 0.2509496385246906,
"grad_norm": 1.0538101196289062,
"learning_rate": 9.767390456370625e-06,
"loss": 0.9381958246231079,
"step": 256
},
{
"epoch": 0.25192991055017766,
"grad_norm": 0.5735601186752319,
"learning_rate": 9.76499286095689e-06,
"loss": 0.7146620750427246,
"step": 257
},
{
"epoch": 0.2529101825756647,
"grad_norm": 0.6433756351470947,
"learning_rate": 9.762583269679304e-06,
"loss": 0.6317933797836304,
"step": 258
},
{
"epoch": 0.25389045460115184,
"grad_norm": 0.852685809135437,
"learning_rate": 9.760161688604008e-06,
"loss": 0.8174331188201904,
"step": 259
},
{
"epoch": 0.2548707266266389,
"grad_norm": 0.5660604238510132,
"learning_rate": 9.75772812382733e-06,
"loss": 0.6192631721496582,
"step": 260
},
{
"epoch": 0.25585099865212596,
"grad_norm": 0.7691637873649597,
"learning_rate": 9.755282581475769e-06,
"loss": 0.6771009564399719,
"step": 261
},
{
"epoch": 0.256831270677613,
"grad_norm": 0.591249942779541,
"learning_rate": 9.75282506770597e-06,
"loss": 0.6050739288330078,
"step": 262
},
{
"epoch": 0.2578115427031001,
"grad_norm": 0.8145242929458618,
"learning_rate": 9.750355588704728e-06,
"loss": 0.6268457174301147,
"step": 263
},
{
"epoch": 0.2587918147285872,
"grad_norm": 0.5998173952102661,
"learning_rate": 9.747874150688948e-06,
"loss": 0.6692138910293579,
"step": 264
},
{
"epoch": 0.25977208675407426,
"grad_norm": 0.5826566219329834,
"learning_rate": 9.745380759905648e-06,
"loss": 0.6950204372406006,
"step": 265
},
{
"epoch": 0.2607523587795613,
"grad_norm": 0.5034763813018799,
"learning_rate": 9.742875422631937e-06,
"loss": 0.5742853879928589,
"step": 266
},
{
"epoch": 0.2617326308050484,
"grad_norm": 0.6649979948997498,
"learning_rate": 9.740358145174999e-06,
"loss": 0.6680870056152344,
"step": 267
},
{
"epoch": 0.2627129028305355,
"grad_norm": 0.5235174894332886,
"learning_rate": 9.737828933872076e-06,
"loss": 0.6588406562805176,
"step": 268
},
{
"epoch": 0.26369317485602256,
"grad_norm": 0.5755462646484375,
"learning_rate": 9.735287795090455e-06,
"loss": 0.7380120754241943,
"step": 269
},
{
"epoch": 0.2646734468815096,
"grad_norm": 0.7009977698326111,
"learning_rate": 9.73273473522745e-06,
"loss": 0.6192063093185425,
"step": 270
},
{
"epoch": 0.2656537189069967,
"grad_norm": 0.612812340259552,
"learning_rate": 9.730169760710385e-06,
"loss": 0.6974408626556396,
"step": 271
},
{
"epoch": 0.26663399093248374,
"grad_norm": 0.8031991124153137,
"learning_rate": 9.727592877996585e-06,
"loss": 0.664556086063385,
"step": 272
},
{
"epoch": 0.26761426295797086,
"grad_norm": 0.5645917057991028,
"learning_rate": 9.725004093573343e-06,
"loss": 0.7329329252243042,
"step": 273
},
{
"epoch": 0.2685945349834579,
"grad_norm": 0.7882102131843567,
"learning_rate": 9.722403413957924e-06,
"loss": 0.6108486652374268,
"step": 274
},
{
"epoch": 0.269574807008945,
"grad_norm": 0.5724264979362488,
"learning_rate": 9.719790845697534e-06,
"loss": 0.7101311683654785,
"step": 275
},
{
"epoch": 0.27055507903443204,
"grad_norm": 0.5334291458129883,
"learning_rate": 9.717166395369312e-06,
"loss": 0.626849889755249,
"step": 276
},
{
"epoch": 0.2715353510599191,
"grad_norm": 0.5758081078529358,
"learning_rate": 9.71453006958031e-06,
"loss": 0.6180967092514038,
"step": 277
},
{
"epoch": 0.2725156230854062,
"grad_norm": 0.4749200642108917,
"learning_rate": 9.711881874967471e-06,
"loss": 0.4926038682460785,
"step": 278
},
{
"epoch": 0.2734958951108933,
"grad_norm": 0.6426512002944946,
"learning_rate": 9.709221818197626e-06,
"loss": 0.7141193151473999,
"step": 279
},
{
"epoch": 0.27447616713638034,
"grad_norm": 0.8100751042366028,
"learning_rate": 9.70654990596746e-06,
"loss": 0.8084619045257568,
"step": 280
},
{
"epoch": 0.2754564391618674,
"grad_norm": 0.5176052451133728,
"learning_rate": 9.703866145003512e-06,
"loss": 0.6762614250183105,
"step": 281
},
{
"epoch": 0.27643671118735447,
"grad_norm": 0.584479033946991,
"learning_rate": 9.701170542062149e-06,
"loss": 0.6309274435043335,
"step": 282
},
{
"epoch": 0.2774169832128416,
"grad_norm": 0.5708397030830383,
"learning_rate": 9.698463103929542e-06,
"loss": 0.7273784875869751,
"step": 283
},
{
"epoch": 0.27839725523832864,
"grad_norm": 0.49435490369796753,
"learning_rate": 9.69574383742167e-06,
"loss": 0.5571186542510986,
"step": 284
},
{
"epoch": 0.2793775272638157,
"grad_norm": 0.6723142862319946,
"learning_rate": 9.69301274938428e-06,
"loss": 0.7581765651702881,
"step": 285
},
{
"epoch": 0.28035779928930277,
"grad_norm": 0.5881115198135376,
"learning_rate": 9.690269846692882e-06,
"loss": 0.6826155185699463,
"step": 286
},
{
"epoch": 0.28133807131478983,
"grad_norm": 0.5932793021202087,
"learning_rate": 9.687515136252732e-06,
"loss": 0.6524624824523926,
"step": 287
},
{
"epoch": 0.28231834334027694,
"grad_norm": 0.6231981515884399,
"learning_rate": 9.68474862499881e-06,
"loss": 0.6794503927230835,
"step": 288
},
{
"epoch": 0.283298615365764,
"grad_norm": 0.582617998123169,
"learning_rate": 9.681970319895804e-06,
"loss": 0.7047141194343567,
"step": 289
},
{
"epoch": 0.28427888739125107,
"grad_norm": 1.219525694847107,
"learning_rate": 9.679180227938095e-06,
"loss": 0.6844112873077393,
"step": 290
},
{
"epoch": 0.28525915941673813,
"grad_norm": 0.646946132183075,
"learning_rate": 9.676378356149733e-06,
"loss": 0.6921058893203735,
"step": 291
},
{
"epoch": 0.28623943144222525,
"grad_norm": 0.6992022395133972,
"learning_rate": 9.67356471158443e-06,
"loss": 0.6100419759750366,
"step": 292
},
{
"epoch": 0.2872197034677123,
"grad_norm": 0.623558759689331,
"learning_rate": 9.670739301325534e-06,
"loss": 0.6630300879478455,
"step": 293
},
{
"epoch": 0.28819997549319937,
"grad_norm": 0.6203799247741699,
"learning_rate": 9.667902132486009e-06,
"loss": 0.8250691890716553,
"step": 294
},
{
"epoch": 0.28918024751868643,
"grad_norm": 0.5172725915908813,
"learning_rate": 9.665053212208426e-06,
"loss": 0.5871583223342896,
"step": 295
},
{
"epoch": 0.2901605195441735,
"grad_norm": 0.585059404373169,
"learning_rate": 9.662192547664938e-06,
"loss": 0.6851358413696289,
"step": 296
},
{
"epoch": 0.2911407915696606,
"grad_norm": 0.5669979453086853,
"learning_rate": 9.659320146057263e-06,
"loss": 0.6426068544387817,
"step": 297
},
{
"epoch": 0.29212106359514767,
"grad_norm": 0.4750717282295227,
"learning_rate": 9.65643601461667e-06,
"loss": 0.5325371026992798,
"step": 298
},
{
"epoch": 0.29310133562063473,
"grad_norm": 0.5632511973381042,
"learning_rate": 9.653540160603956e-06,
"loss": 0.6740025877952576,
"step": 299
},
{
"epoch": 0.2940816076461218,
"grad_norm": 0.6044143438339233,
"learning_rate": 9.650632591309431e-06,
"loss": 0.7068796753883362,
"step": 300
},
{
"epoch": 0.29506187967160885,
"grad_norm": 0.5127572417259216,
"learning_rate": 9.647713314052896e-06,
"loss": 0.5933098196983337,
"step": 301
},
{
"epoch": 0.29604215169709597,
"grad_norm": 0.5830118060112,
"learning_rate": 9.64478233618363e-06,
"loss": 0.6408910751342773,
"step": 302
},
{
"epoch": 0.29702242372258303,
"grad_norm": 0.5796244740486145,
"learning_rate": 9.641839665080363e-06,
"loss": 0.632135272026062,
"step": 303
},
{
"epoch": 0.2980026957480701,
"grad_norm": 0.5338972210884094,
"learning_rate": 9.63888530815127e-06,
"loss": 0.535368800163269,
"step": 304
},
{
"epoch": 0.29898296777355715,
"grad_norm": 0.490498423576355,
"learning_rate": 9.635919272833938e-06,
"loss": 0.595739483833313,
"step": 305
},
{
"epoch": 0.2999632397990442,
"grad_norm": 0.6124870181083679,
"learning_rate": 9.632941566595357e-06,
"loss": 0.7355586886405945,
"step": 306
},
{
"epoch": 0.30094351182453133,
"grad_norm": 0.4968045651912689,
"learning_rate": 9.629952196931902e-06,
"loss": 0.5700988173484802,
"step": 307
},
{
"epoch": 0.3019237838500184,
"grad_norm": 0.5857828855514526,
"learning_rate": 9.626951171369306e-06,
"loss": 0.6341098546981812,
"step": 308
},
{
"epoch": 0.30290405587550545,
"grad_norm": 0.7570348381996155,
"learning_rate": 9.623938497462647e-06,
"loss": 0.9097845554351807,
"step": 309
},
{
"epoch": 0.3038843279009925,
"grad_norm": 0.6131370663642883,
"learning_rate": 9.620914182796327e-06,
"loss": 0.49961164593696594,
"step": 310
},
{
"epoch": 0.3048645999264796,
"grad_norm": 0.6986640691757202,
"learning_rate": 9.617878234984056e-06,
"loss": 0.6150896549224854,
"step": 311
},
{
"epoch": 0.3058448719519667,
"grad_norm": 0.5393136739730835,
"learning_rate": 9.61483066166883e-06,
"loss": 0.5867908000946045,
"step": 312
},
{
"epoch": 0.30682514397745375,
"grad_norm": 0.5819171071052551,
"learning_rate": 9.611771470522908e-06,
"loss": 0.665069580078125,
"step": 313
},
{
"epoch": 0.3078054160029408,
"grad_norm": 0.5913625359535217,
"learning_rate": 9.608700669247803e-06,
"loss": 0.6079078316688538,
"step": 314
},
{
"epoch": 0.3087856880284279,
"grad_norm": 0.544347882270813,
"learning_rate": 9.60561826557425e-06,
"loss": 0.5789344906806946,
"step": 315
},
{
"epoch": 0.30976596005391493,
"grad_norm": 0.6326791644096375,
"learning_rate": 9.602524267262202e-06,
"loss": 0.6234204769134521,
"step": 316
},
{
"epoch": 0.31074623207940205,
"grad_norm": 0.6705621480941772,
"learning_rate": 9.599418682100793e-06,
"loss": 0.5959568023681641,
"step": 317
},
{
"epoch": 0.3117265041048891,
"grad_norm": 0.5172932744026184,
"learning_rate": 9.596301517908329e-06,
"loss": 0.5890722870826721,
"step": 318
},
{
"epoch": 0.3127067761303762,
"grad_norm": 0.9102640151977539,
"learning_rate": 9.59317278253227e-06,
"loss": 0.700379490852356,
"step": 319
},
{
"epoch": 0.31368704815586324,
"grad_norm": 0.5571013689041138,
"learning_rate": 9.5900324838492e-06,
"loss": 0.6454604268074036,
"step": 320
},
{
"epoch": 0.31466732018135035,
"grad_norm": 0.5500742793083191,
"learning_rate": 9.586880629764817e-06,
"loss": 0.5959937572479248,
"step": 321
},
{
"epoch": 0.3156475922068374,
"grad_norm": 0.667962908744812,
"learning_rate": 9.583717228213914e-06,
"loss": 0.6092185974121094,
"step": 322
},
{
"epoch": 0.3166278642323245,
"grad_norm": 1.0929455757141113,
"learning_rate": 9.580542287160348e-06,
"loss": 0.6907039284706116,
"step": 323
},
{
"epoch": 0.31760813625781154,
"grad_norm": 0.6506457328796387,
"learning_rate": 9.577355814597031e-06,
"loss": 0.7235692739486694,
"step": 324
},
{
"epoch": 0.3185884082832986,
"grad_norm": 0.6015403866767883,
"learning_rate": 9.574157818545902e-06,
"loss": 0.5779285430908203,
"step": 325
},
{
"epoch": 0.3195686803087857,
"grad_norm": 0.7786325216293335,
"learning_rate": 9.570948307057912e-06,
"loss": 0.7327282428741455,
"step": 326
},
{
"epoch": 0.3205489523342728,
"grad_norm": 0.6935060024261475,
"learning_rate": 9.567727288213005e-06,
"loss": 0.5526785850524902,
"step": 327
},
{
"epoch": 0.32152922435975984,
"grad_norm": 0.5690432786941528,
"learning_rate": 9.564494770120089e-06,
"loss": 0.5947889089584351,
"step": 328
},
{
"epoch": 0.3225094963852469,
"grad_norm": 0.8004881143569946,
"learning_rate": 9.561250760917026e-06,
"loss": 0.6638854742050171,
"step": 329
},
{
"epoch": 0.32348976841073396,
"grad_norm": 0.56986004114151,
"learning_rate": 9.557995268770609e-06,
"loss": 0.7149271965026855,
"step": 330
},
{
"epoch": 0.3244700404362211,
"grad_norm": 0.5093994736671448,
"learning_rate": 9.554728301876525e-06,
"loss": 0.573759913444519,
"step": 331
},
{
"epoch": 0.32545031246170814,
"grad_norm": 0.6659159660339355,
"learning_rate": 9.551449868459371e-06,
"loss": 0.6619836091995239,
"step": 332
},
{
"epoch": 0.3264305844871952,
"grad_norm": 0.7009037137031555,
"learning_rate": 9.548159976772593e-06,
"loss": 0.6384791135787964,
"step": 333
},
{
"epoch": 0.32741085651268226,
"grad_norm": 0.588800847530365,
"learning_rate": 9.544858635098491e-06,
"loss": 0.5931030511856079,
"step": 334
},
{
"epoch": 0.3283911285381693,
"grad_norm": 0.6182215213775635,
"learning_rate": 9.541545851748186e-06,
"loss": 0.7085162401199341,
"step": 335
},
{
"epoch": 0.32937140056365644,
"grad_norm": 0.5252450108528137,
"learning_rate": 9.538221635061611e-06,
"loss": 0.5386106967926025,
"step": 336
},
{
"epoch": 0.3303516725891435,
"grad_norm": 0.5651358962059021,
"learning_rate": 9.534885993407474e-06,
"loss": 0.6456975936889648,
"step": 337
},
{
"epoch": 0.33133194461463056,
"grad_norm": 0.5379696488380432,
"learning_rate": 9.531538935183252e-06,
"loss": 0.6666260361671448,
"step": 338
},
{
"epoch": 0.3323122166401176,
"grad_norm": 0.656000554561615,
"learning_rate": 9.528180468815155e-06,
"loss": 0.6818495988845825,
"step": 339
},
{
"epoch": 0.3332924886656047,
"grad_norm": 0.8253548741340637,
"learning_rate": 9.524810602758122e-06,
"loss": 0.5551189184188843,
"step": 340
},
{
"epoch": 0.3342727606910918,
"grad_norm": 0.8118085265159607,
"learning_rate": 9.521429345495787e-06,
"loss": 0.6034508943557739,
"step": 341
},
{
"epoch": 0.33525303271657886,
"grad_norm": 0.8866831660270691,
"learning_rate": 9.518036705540459e-06,
"loss": 0.7978249788284302,
"step": 342
},
{
"epoch": 0.3362333047420659,
"grad_norm": 0.5758262276649475,
"learning_rate": 9.514632691433108e-06,
"loss": 0.6506916284561157,
"step": 343
},
{
"epoch": 0.337213576767553,
"grad_norm": 0.6339187026023865,
"learning_rate": 9.51121731174333e-06,
"loss": 0.6326726675033569,
"step": 344
},
{
"epoch": 0.33819384879304004,
"grad_norm": 0.8769382834434509,
"learning_rate": 9.507790575069347e-06,
"loss": 0.7365003824234009,
"step": 345
},
{
"epoch": 0.33917412081852716,
"grad_norm": 0.7927872538566589,
"learning_rate": 9.504352490037957e-06,
"loss": 0.5798900127410889,
"step": 346
},
{
"epoch": 0.3401543928440142,
"grad_norm": 0.5708814263343811,
"learning_rate": 9.50090306530454e-06,
"loss": 0.7537117004394531,
"step": 347
},
{
"epoch": 0.3411346648695013,
"grad_norm": 0.48839452862739563,
"learning_rate": 9.497442309553017e-06,
"loss": 0.5822838544845581,
"step": 348
},
{
"epoch": 0.34211493689498834,
"grad_norm": 0.7066628336906433,
"learning_rate": 9.493970231495836e-06,
"loss": 0.7837392687797546,
"step": 349
},
{
"epoch": 0.34309520892047546,
"grad_norm": 0.584050178527832,
"learning_rate": 9.490486839873949e-06,
"loss": 0.7187587022781372,
"step": 350
},
{
"epoch": 0.3440754809459625,
"grad_norm": 0.5140157341957092,
"learning_rate": 9.486992143456792e-06,
"loss": 0.6060421466827393,
"step": 351
},
{
"epoch": 0.3450557529714496,
"grad_norm": 0.5367881059646606,
"learning_rate": 9.483486151042258e-06,
"loss": 0.5815377235412598,
"step": 352
},
{
"epoch": 0.34603602499693664,
"grad_norm": 0.4939191937446594,
"learning_rate": 9.47996887145668e-06,
"loss": 0.5326993465423584,
"step": 353
},
{
"epoch": 0.3470162970224237,
"grad_norm": 0.6358892321586609,
"learning_rate": 9.476440313554803e-06,
"loss": 0.6004680395126343,
"step": 354
},
{
"epoch": 0.3479965690479108,
"grad_norm": 0.6140431761741638,
"learning_rate": 9.47290048621977e-06,
"loss": 0.6206232309341431,
"step": 355
},
{
"epoch": 0.3489768410733979,
"grad_norm": 0.5943885445594788,
"learning_rate": 9.469349398363088e-06,
"loss": 0.5952731966972351,
"step": 356
},
{
"epoch": 0.34995711309888494,
"grad_norm": 0.49166974425315857,
"learning_rate": 9.46578705892462e-06,
"loss": 0.6236333847045898,
"step": 357
},
{
"epoch": 0.350937385124372,
"grad_norm": 0.6081139445304871,
"learning_rate": 9.46221347687255e-06,
"loss": 0.7176322937011719,
"step": 358
},
{
"epoch": 0.35191765714985906,
"grad_norm": 0.5212146639823914,
"learning_rate": 9.458628661203368e-06,
"loss": 0.5600054860115051,
"step": 359
},
{
"epoch": 0.3528979291753462,
"grad_norm": 0.5689999461174011,
"learning_rate": 9.45503262094184e-06,
"loss": 0.6379108428955078,
"step": 360
},
{
"epoch": 0.35387820120083324,
"grad_norm": 0.5177587866783142,
"learning_rate": 9.451425365140997e-06,
"loss": 0.6243526935577393,
"step": 361
},
{
"epoch": 0.3548584732263203,
"grad_norm": 0.4048973619937897,
"learning_rate": 9.447806902882098e-06,
"loss": 0.5460570454597473,
"step": 362
},
{
"epoch": 0.35583874525180736,
"grad_norm": 0.5066366195678711,
"learning_rate": 9.444177243274619e-06,
"loss": 0.5914937853813171,
"step": 363
},
{
"epoch": 0.3568190172772944,
"grad_norm": 1.1004585027694702,
"learning_rate": 9.440536395456223e-06,
"loss": 0.6613606214523315,
"step": 364
},
{
"epoch": 0.35779928930278154,
"grad_norm": 0.4945545196533203,
"learning_rate": 9.43688436859274e-06,
"loss": 0.6563936471939087,
"step": 365
},
{
"epoch": 0.3587795613282686,
"grad_norm": 0.46797290444374084,
"learning_rate": 9.433221171878144e-06,
"loss": 0.5227972269058228,
"step": 366
},
{
"epoch": 0.35975983335375566,
"grad_norm": 0.6280576586723328,
"learning_rate": 9.429546814534528e-06,
"loss": 0.6633756160736084,
"step": 367
},
{
"epoch": 0.3607401053792427,
"grad_norm": 0.6422777771949768,
"learning_rate": 9.425861305812083e-06,
"loss": 0.7747545838356018,
"step": 368
},
{
"epoch": 0.3617203774047298,
"grad_norm": 0.6484074592590332,
"learning_rate": 9.422164654989073e-06,
"loss": 0.638740062713623,
"step": 369
},
{
"epoch": 0.3627006494302169,
"grad_norm": 0.4812702238559723,
"learning_rate": 9.41845687137181e-06,
"loss": 0.7052801251411438,
"step": 370
},
{
"epoch": 0.36368092145570396,
"grad_norm": 0.5138601064682007,
"learning_rate": 9.414737964294636e-06,
"loss": 0.58039790391922,
"step": 371
},
{
"epoch": 0.364661193481191,
"grad_norm": 0.7280306220054626,
"learning_rate": 9.411007943119895e-06,
"loss": 0.8962255716323853,
"step": 372
},
{
"epoch": 0.3656414655066781,
"grad_norm": 0.5673227906227112,
"learning_rate": 9.40726681723791e-06,
"loss": 0.6811161637306213,
"step": 373
},
{
"epoch": 0.36662173753216515,
"grad_norm": 0.5575160980224609,
"learning_rate": 9.403514596066964e-06,
"loss": 0.6203138828277588,
"step": 374
},
{
"epoch": 0.36760200955765227,
"grad_norm": 0.5194122195243835,
"learning_rate": 9.399751289053267e-06,
"loss": 0.6002714037895203,
"step": 375
},
{
"epoch": 0.3685822815831393,
"grad_norm": 0.476211816072464,
"learning_rate": 9.39597690567094e-06,
"loss": 0.49280551075935364,
"step": 376
},
{
"epoch": 0.3695625536086264,
"grad_norm": 0.5374362468719482,
"learning_rate": 9.392191455421989e-06,
"loss": 0.5507866144180298,
"step": 377
},
{
"epoch": 0.37054282563411345,
"grad_norm": 0.48763546347618103,
"learning_rate": 9.388394947836278e-06,
"loss": 0.5378220677375793,
"step": 378
},
{
"epoch": 0.37152309765960057,
"grad_norm": 0.49194395542144775,
"learning_rate": 9.384587392471516e-06,
"loss": 0.5602411031723022,
"step": 379
},
{
"epoch": 0.3725033696850876,
"grad_norm": 0.5846983194351196,
"learning_rate": 9.380768798913213e-06,
"loss": 0.6200522780418396,
"step": 380
},
{
"epoch": 0.3734836417105747,
"grad_norm": 0.6853125691413879,
"learning_rate": 9.376939176774678e-06,
"loss": 0.5838781595230103,
"step": 381
},
{
"epoch": 0.37446391373606175,
"grad_norm": 0.5014895796775818,
"learning_rate": 9.37309853569698e-06,
"loss": 0.5761265754699707,
"step": 382
},
{
"epoch": 0.3754441857615488,
"grad_norm": 0.536831259727478,
"learning_rate": 9.369246885348926e-06,
"loss": 0.6839797496795654,
"step": 383
},
{
"epoch": 0.3764244577870359,
"grad_norm": 0.47531557083129883,
"learning_rate": 9.365384235427042e-06,
"loss": 0.6124746799468994,
"step": 384
},
{
"epoch": 0.377404729812523,
"grad_norm": 0.5946936011314392,
"learning_rate": 9.361510595655545e-06,
"loss": 0.7137396931648254,
"step": 385
},
{
"epoch": 0.37838500183801005,
"grad_norm": 0.502773106098175,
"learning_rate": 9.357625975786318e-06,
"loss": 0.6333768963813782,
"step": 386
},
{
"epoch": 0.3793652738634971,
"grad_norm": 0.4880274534225464,
"learning_rate": 9.353730385598887e-06,
"loss": 0.6179925799369812,
"step": 387
},
{
"epoch": 0.38034554588898417,
"grad_norm": 0.4644832909107208,
"learning_rate": 9.349823834900396e-06,
"loss": 0.5528720617294312,
"step": 388
},
{
"epoch": 0.3813258179144713,
"grad_norm": 0.5169019103050232,
"learning_rate": 9.345906333525582e-06,
"loss": 0.5545703768730164,
"step": 389
},
{
"epoch": 0.38230608993995835,
"grad_norm": 0.5275116562843323,
"learning_rate": 9.341977891336748e-06,
"loss": 0.6071966886520386,
"step": 390
},
{
"epoch": 0.3832863619654454,
"grad_norm": 0.6112439036369324,
"learning_rate": 9.338038518223746e-06,
"loss": 0.5499929189682007,
"step": 391
},
{
"epoch": 0.38426663399093247,
"grad_norm": 0.4731747806072235,
"learning_rate": 9.334088224103942e-06,
"loss": 0.5589090585708618,
"step": 392
},
{
"epoch": 0.38524690601641953,
"grad_norm": 0.6080041527748108,
"learning_rate": 9.330127018922195e-06,
"loss": 0.6990543603897095,
"step": 393
},
{
"epoch": 0.38622717804190665,
"grad_norm": 0.5127442479133606,
"learning_rate": 9.326154912650834e-06,
"loss": 0.5527446269989014,
"step": 394
},
{
"epoch": 0.3872074500673937,
"grad_norm": 0.4985249936580658,
"learning_rate": 9.322171915289635e-06,
"loss": 0.6704634428024292,
"step": 395
},
{
"epoch": 0.38818772209288077,
"grad_norm": 0.4670063853263855,
"learning_rate": 9.318178036865786e-06,
"loss": 0.5786021947860718,
"step": 396
},
{
"epoch": 0.38916799411836783,
"grad_norm": 0.6596277356147766,
"learning_rate": 9.314173287433874e-06,
"loss": 0.7387778759002686,
"step": 397
},
{
"epoch": 0.3901482661438549,
"grad_norm": 0.4874238967895508,
"learning_rate": 9.310157677075847e-06,
"loss": 0.5646405220031738,
"step": 398
},
{
"epoch": 0.391128538169342,
"grad_norm": 0.7856294512748718,
"learning_rate": 9.306131215901004e-06,
"loss": 0.641767680644989,
"step": 399
},
{
"epoch": 0.39210881019482907,
"grad_norm": 0.3762321174144745,
"learning_rate": 9.302093914045953e-06,
"loss": 0.4375483989715576,
"step": 400
},
{
"epoch": 0.39308908222031613,
"grad_norm": 0.5250818133354187,
"learning_rate": 9.298045781674595e-06,
"loss": 0.58973628282547,
"step": 401
},
{
"epoch": 0.3940693542458032,
"grad_norm": 0.5813953876495361,
"learning_rate": 9.293986828978106e-06,
"loss": 0.6511713266372681,
"step": 402
},
{
"epoch": 0.3950496262712903,
"grad_norm": 0.5056664347648621,
"learning_rate": 9.289917066174887e-06,
"loss": 0.6352890729904175,
"step": 403
},
{
"epoch": 0.39602989829677737,
"grad_norm": 0.4536687433719635,
"learning_rate": 9.285836503510562e-06,
"loss": 0.5708507895469666,
"step": 404
},
{
"epoch": 0.39701017032226443,
"grad_norm": 0.5224312543869019,
"learning_rate": 9.281745151257946e-06,
"loss": 0.5995397567749023,
"step": 405
},
{
"epoch": 0.3979904423477515,
"grad_norm": 0.580417811870575,
"learning_rate": 9.277643019717011e-06,
"loss": 0.7497074604034424,
"step": 406
},
{
"epoch": 0.39897071437323856,
"grad_norm": 0.6126673221588135,
"learning_rate": 9.273530119214868e-06,
"loss": 0.5598350763320923,
"step": 407
},
{
"epoch": 0.39995098639872567,
"grad_norm": 0.5506340861320496,
"learning_rate": 9.269406460105742e-06,
"loss": 0.5485457181930542,
"step": 408
},
{
"epoch": 0.40093125842421273,
"grad_norm": 0.4401313066482544,
"learning_rate": 9.265272052770936e-06,
"loss": 0.5437591075897217,
"step": 409
},
{
"epoch": 0.4019115304496998,
"grad_norm": 0.7231860160827637,
"learning_rate": 9.261126907618818e-06,
"loss": 0.5335922837257385,
"step": 410
},
{
"epoch": 0.40289180247518686,
"grad_norm": 0.6745095252990723,
"learning_rate": 9.256971035084786e-06,
"loss": 0.6721965074539185,
"step": 411
},
{
"epoch": 0.4038720745006739,
"grad_norm": 0.5086560249328613,
"learning_rate": 9.252804445631243e-06,
"loss": 0.5465257167816162,
"step": 412
},
{
"epoch": 0.40485234652616103,
"grad_norm": 0.5411233305931091,
"learning_rate": 9.248627149747573e-06,
"loss": 0.76491379737854,
"step": 413
},
{
"epoch": 0.4058326185516481,
"grad_norm": 0.5963067412376404,
"learning_rate": 9.244439157950114e-06,
"loss": 0.7780289649963379,
"step": 414
},
{
"epoch": 0.40681289057713516,
"grad_norm": 0.5068957805633545,
"learning_rate": 9.24024048078213e-06,
"loss": 0.5478646755218506,
"step": 415
},
{
"epoch": 0.4077931626026222,
"grad_norm": 0.45583173632621765,
"learning_rate": 9.236031128813786e-06,
"loss": 0.6128138899803162,
"step": 416
},
{
"epoch": 0.4087734346281093,
"grad_norm": 0.5057873129844666,
"learning_rate": 9.231811112642121e-06,
"loss": 0.6155551671981812,
"step": 417
},
{
"epoch": 0.4097537066535964,
"grad_norm": 0.44246384501457214,
"learning_rate": 9.227580442891022e-06,
"loss": 0.5776406526565552,
"step": 418
},
{
"epoch": 0.41073397867908346,
"grad_norm": 0.5929768681526184,
"learning_rate": 9.223339130211194e-06,
"loss": 0.5738487839698792,
"step": 419
},
{
"epoch": 0.4117142507045705,
"grad_norm": 0.5466901659965515,
"learning_rate": 9.219087185280132e-06,
"loss": 0.6888821721076965,
"step": 420
},
{
"epoch": 0.4126945227300576,
"grad_norm": 0.5927895903587341,
"learning_rate": 9.214824618802108e-06,
"loss": 0.7219152450561523,
"step": 421
},
{
"epoch": 0.41367479475554464,
"grad_norm": 0.45233601331710815,
"learning_rate": 9.210551441508126e-06,
"loss": 0.4950752258300781,
"step": 422
},
{
"epoch": 0.41465506678103176,
"grad_norm": 0.4221014082431793,
"learning_rate": 9.206267664155906e-06,
"loss": 0.5129454135894775,
"step": 423
},
{
"epoch": 0.4156353388065188,
"grad_norm": 0.517291784286499,
"learning_rate": 9.201973297529852e-06,
"loss": 0.5749497413635254,
"step": 424
},
{
"epoch": 0.4166156108320059,
"grad_norm": 0.5062311887741089,
"learning_rate": 9.197668352441025e-06,
"loss": 0.5585308074951172,
"step": 425
},
{
"epoch": 0.41759588285749294,
"grad_norm": 0.44994261860847473,
"learning_rate": 9.193352839727122e-06,
"loss": 0.5468809604644775,
"step": 426
},
{
"epoch": 0.41857615488298,
"grad_norm": 0.5119527578353882,
"learning_rate": 9.189026770252437e-06,
"loss": 0.581660807132721,
"step": 427
},
{
"epoch": 0.4195564269084671,
"grad_norm": 0.5557008981704712,
"learning_rate": 9.18469015490785e-06,
"loss": 0.6926128268241882,
"step": 428
},
{
"epoch": 0.4205366989339542,
"grad_norm": 0.4753175377845764,
"learning_rate": 9.18034300461078e-06,
"loss": 0.6371850967407227,
"step": 429
},
{
"epoch": 0.42151697095944124,
"grad_norm": 0.6206807494163513,
"learning_rate": 9.175985330305176e-06,
"loss": 0.6325344443321228,
"step": 430
},
{
"epoch": 0.4224972429849283,
"grad_norm": 0.49500077962875366,
"learning_rate": 9.171617142961477e-06,
"loss": 0.584123969078064,
"step": 431
},
{
"epoch": 0.4234775150104154,
"grad_norm": 0.5801467299461365,
"learning_rate": 9.167238453576589e-06,
"loss": 0.5789539813995361,
"step": 432
},
{
"epoch": 0.4244577870359025,
"grad_norm": 0.5026708245277405,
"learning_rate": 9.162849273173857e-06,
"loss": 0.60198974609375,
"step": 433
},
{
"epoch": 0.42543805906138954,
"grad_norm": 0.5288900136947632,
"learning_rate": 9.158449612803039e-06,
"loss": 0.7147405743598938,
"step": 434
},
{
"epoch": 0.4264183310868766,
"grad_norm": 0.48020800948143005,
"learning_rate": 9.154039483540273e-06,
"loss": 0.6171431541442871,
"step": 435
},
{
"epoch": 0.42739860311236366,
"grad_norm": 0.5342614650726318,
"learning_rate": 9.149618896488056e-06,
"loss": 0.5938845872879028,
"step": 436
},
{
"epoch": 0.4283788751378508,
"grad_norm": 0.4195495843887329,
"learning_rate": 9.145187862775208e-06,
"loss": 0.44220370054244995,
"step": 437
},
{
"epoch": 0.42935914716333784,
"grad_norm": 0.4617707431316376,
"learning_rate": 9.140746393556853e-06,
"loss": 0.6354628801345825,
"step": 438
},
{
"epoch": 0.4303394191888249,
"grad_norm": 0.5162332653999329,
"learning_rate": 9.136294500014387e-06,
"loss": 0.6938362717628479,
"step": 439
},
{
"epoch": 0.43131969121431196,
"grad_norm": 0.4750451147556305,
"learning_rate": 9.13183219335544e-06,
"loss": 0.6032200455665588,
"step": 440
},
{
"epoch": 0.432299963239799,
"grad_norm": 0.5219208002090454,
"learning_rate": 9.12735948481387e-06,
"loss": 0.6199036240577698,
"step": 441
},
{
"epoch": 0.43328023526528614,
"grad_norm": 0.4706224203109741,
"learning_rate": 9.122876385649713e-06,
"loss": 0.5312067866325378,
"step": 442
},
{
"epoch": 0.4342605072907732,
"grad_norm": 0.4995059669017792,
"learning_rate": 9.118382907149164e-06,
"loss": 0.6517359018325806,
"step": 443
},
{
"epoch": 0.43524077931626026,
"grad_norm": 0.5074915885925293,
"learning_rate": 9.113879060624554e-06,
"loss": 0.5837841033935547,
"step": 444
},
{
"epoch": 0.4362210513417473,
"grad_norm": 0.5803712606430054,
"learning_rate": 9.109364857414306e-06,
"loss": 0.6096867322921753,
"step": 445
},
{
"epoch": 0.4372013233672344,
"grad_norm": 0.7945765852928162,
"learning_rate": 9.104840308882925e-06,
"loss": 0.6078523993492126,
"step": 446
},
{
"epoch": 0.4381815953927215,
"grad_norm": 0.5049729347229004,
"learning_rate": 9.100305426420957e-06,
"loss": 0.6547229290008545,
"step": 447
},
{
"epoch": 0.43916186741820856,
"grad_norm": 0.4533054530620575,
"learning_rate": 9.09576022144496e-06,
"loss": 0.5432737469673157,
"step": 448
},
{
"epoch": 0.4401421394436956,
"grad_norm": 0.4947768747806549,
"learning_rate": 9.091204705397485e-06,
"loss": 0.5290822982788086,
"step": 449
},
{
"epoch": 0.4411224114691827,
"grad_norm": 0.5200343728065491,
"learning_rate": 9.086638889747034e-06,
"loss": 0.6132403016090393,
"step": 450
},
{
"epoch": 0.44210268349466975,
"grad_norm": 0.4689146876335144,
"learning_rate": 9.08206278598805e-06,
"loss": 0.4153621196746826,
"step": 451
},
{
"epoch": 0.44308295552015686,
"grad_norm": 0.5330294370651245,
"learning_rate": 9.077476405640862e-06,
"loss": 0.6979874968528748,
"step": 452
},
{
"epoch": 0.4440632275456439,
"grad_norm": 0.4985146224498749,
"learning_rate": 9.07287976025168e-06,
"loss": 0.6385680437088013,
"step": 453
},
{
"epoch": 0.445043499571131,
"grad_norm": 0.5191720128059387,
"learning_rate": 9.06827286139255e-06,
"loss": 0.5751811265945435,
"step": 454
},
{
"epoch": 0.44602377159661805,
"grad_norm": 0.4784611463546753,
"learning_rate": 9.06365572066134e-06,
"loss": 0.6754035949707031,
"step": 455
},
{
"epoch": 0.4470040436221051,
"grad_norm": 0.4339852035045624,
"learning_rate": 9.059028349681693e-06,
"loss": 0.5843990445137024,
"step": 456
},
{
"epoch": 0.4479843156475922,
"grad_norm": 0.4005136787891388,
"learning_rate": 9.05439076010301e-06,
"loss": 0.5037546753883362,
"step": 457
},
{
"epoch": 0.4489645876730793,
"grad_norm": 0.6022442579269409,
"learning_rate": 9.04974296360042e-06,
"loss": 0.6200989484786987,
"step": 458
},
{
"epoch": 0.44994485969856635,
"grad_norm": 0.48745962977409363,
"learning_rate": 9.045084971874738e-06,
"loss": 0.6186924576759338,
"step": 459
},
{
"epoch": 0.4509251317240534,
"grad_norm": 0.48209333419799805,
"learning_rate": 9.040416796652459e-06,
"loss": 0.6512715816497803,
"step": 460
},
{
"epoch": 0.4519054037495405,
"grad_norm": 0.4796985983848572,
"learning_rate": 9.035738449685707e-06,
"loss": 0.5661883354187012,
"step": 461
},
{
"epoch": 0.4528856757750276,
"grad_norm": 0.5448599457740784,
"learning_rate": 9.031049942752215e-06,
"loss": 0.6986958980560303,
"step": 462
},
{
"epoch": 0.45386594780051465,
"grad_norm": 0.5704795122146606,
"learning_rate": 9.026351287655294e-06,
"loss": 0.5709372758865356,
"step": 463
},
{
"epoch": 0.4548462198260017,
"grad_norm": 0.4772814214229584,
"learning_rate": 9.021642496223802e-06,
"loss": 0.5940738916397095,
"step": 464
},
{
"epoch": 0.45582649185148877,
"grad_norm": 0.5291264653205872,
"learning_rate": 9.016923580312114e-06,
"loss": 0.5524354577064514,
"step": 465
},
{
"epoch": 0.4568067638769759,
"grad_norm": 0.5132133364677429,
"learning_rate": 9.012194551800098e-06,
"loss": 0.5567898750305176,
"step": 466
},
{
"epoch": 0.45778703590246295,
"grad_norm": 0.5544204711914062,
"learning_rate": 9.007455422593077e-06,
"loss": 0.6415997743606567,
"step": 467
},
{
"epoch": 0.45876730792795,
"grad_norm": 0.4312683641910553,
"learning_rate": 9.002706204621802e-06,
"loss": 0.5623404383659363,
"step": 468
},
{
"epoch": 0.45974757995343707,
"grad_norm": 0.5000928044319153,
"learning_rate": 8.997946909842426e-06,
"loss": 0.5414539575576782,
"step": 469
},
{
"epoch": 0.46072785197892413,
"grad_norm": 0.5731768012046814,
"learning_rate": 8.993177550236464e-06,
"loss": 0.5618506073951721,
"step": 470
},
{
"epoch": 0.46170812400441125,
"grad_norm": 0.5656113028526306,
"learning_rate": 8.988398137810778e-06,
"loss": 0.5426561832427979,
"step": 471
},
{
"epoch": 0.4626883960298983,
"grad_norm": 0.7768886089324951,
"learning_rate": 8.983608684597529e-06,
"loss": 0.6204694509506226,
"step": 472
},
{
"epoch": 0.46366866805538537,
"grad_norm": 0.5393010973930359,
"learning_rate": 8.978809202654161e-06,
"loss": 0.6652113199234009,
"step": 473
},
{
"epoch": 0.46464894008087243,
"grad_norm": 0.44731763005256653,
"learning_rate": 8.973999704063365e-06,
"loss": 0.595824658870697,
"step": 474
},
{
"epoch": 0.4656292121063595,
"grad_norm": 0.48581287264823914,
"learning_rate": 8.969180200933048e-06,
"loss": 0.7073965072631836,
"step": 475
},
{
"epoch": 0.4666094841318466,
"grad_norm": 0.447153240442276,
"learning_rate": 8.9643507053963e-06,
"loss": 0.6062104105949402,
"step": 476
},
{
"epoch": 0.46758975615733367,
"grad_norm": 0.7264053225517273,
"learning_rate": 8.959511229611377e-06,
"loss": 0.7488131523132324,
"step": 477
},
{
"epoch": 0.46857002818282073,
"grad_norm": 0.5424187183380127,
"learning_rate": 8.954661785761648e-06,
"loss": 0.6357473134994507,
"step": 478
},
{
"epoch": 0.4695503002083078,
"grad_norm": 0.5364806056022644,
"learning_rate": 8.949802386055582e-06,
"loss": 0.7206878066062927,
"step": 479
},
{
"epoch": 0.47053057223379485,
"grad_norm": 0.5717079043388367,
"learning_rate": 8.944933042726715e-06,
"loss": 0.6156141757965088,
"step": 480
},
{
"epoch": 0.47151084425928197,
"grad_norm": 0.5376148223876953,
"learning_rate": 8.94005376803361e-06,
"loss": 0.6603047847747803,
"step": 481
},
{
"epoch": 0.47249111628476903,
"grad_norm": 0.55712890625,
"learning_rate": 8.935164574259838e-06,
"loss": 0.6593056917190552,
"step": 482
},
{
"epoch": 0.4734713883102561,
"grad_norm": 0.5702346563339233,
"learning_rate": 8.930265473713939e-06,
"loss": 0.601768970489502,
"step": 483
},
{
"epoch": 0.47445166033574315,
"grad_norm": 0.50825035572052,
"learning_rate": 8.925356478729389e-06,
"loss": 0.5900289416313171,
"step": 484
},
{
"epoch": 0.4754319323612302,
"grad_norm": 0.5257771015167236,
"learning_rate": 8.92043760166458e-06,
"loss": 0.6179829835891724,
"step": 485
},
{
"epoch": 0.47641220438671733,
"grad_norm": 0.5047194361686707,
"learning_rate": 8.915508854902778e-06,
"loss": 0.5214921236038208,
"step": 486
},
{
"epoch": 0.4773924764122044,
"grad_norm": 0.40976911783218384,
"learning_rate": 8.910570250852098e-06,
"loss": 0.4534802734851837,
"step": 487
},
{
"epoch": 0.47837274843769145,
"grad_norm": 0.5223997831344604,
"learning_rate": 8.905621801945467e-06,
"loss": 0.5674428939819336,
"step": 488
},
{
"epoch": 0.4793530204631785,
"grad_norm": 0.5553629994392395,
"learning_rate": 8.900663520640605e-06,
"loss": 0.6257451772689819,
"step": 489
},
{
"epoch": 0.48033329248866563,
"grad_norm": 0.5198466777801514,
"learning_rate": 8.895695419419973e-06,
"loss": 0.8313612937927246,
"step": 490
},
{
"epoch": 0.4813135645141527,
"grad_norm": 0.4763185977935791,
"learning_rate": 8.890717510790763e-06,
"loss": 0.5762819051742554,
"step": 491
},
{
"epoch": 0.48229383653963975,
"grad_norm": 0.49477419257164,
"learning_rate": 8.885729807284855e-06,
"loss": 0.6347113847732544,
"step": 492
},
{
"epoch": 0.4832741085651268,
"grad_norm": 0.5553587079048157,
"learning_rate": 8.880732321458785e-06,
"loss": 0.5741130709648132,
"step": 493
},
{
"epoch": 0.4842543805906139,
"grad_norm": 0.4788428843021393,
"learning_rate": 8.875725065893717e-06,
"loss": 0.5938292741775513,
"step": 494
},
{
"epoch": 0.485234652616101,
"grad_norm": 0.5414963960647583,
"learning_rate": 8.870708053195414e-06,
"loss": 0.6218856573104858,
"step": 495
},
{
"epoch": 0.48621492464158805,
"grad_norm": 0.8729277849197388,
"learning_rate": 8.865681295994195e-06,
"loss": 0.6222516894340515,
"step": 496
},
{
"epoch": 0.4871951966670751,
"grad_norm": 0.5980170369148254,
"learning_rate": 8.860644806944917e-06,
"loss": 0.49276524782180786,
"step": 497
},
{
"epoch": 0.4881754686925622,
"grad_norm": 0.5419535040855408,
"learning_rate": 8.85559859872694e-06,
"loss": 0.6108378767967224,
"step": 498
},
{
"epoch": 0.48915574071804924,
"grad_norm": 0.5596175789833069,
"learning_rate": 8.850542684044078e-06,
"loss": 0.5110671520233154,
"step": 499
},
{
"epoch": 0.49013601274353635,
"grad_norm": 0.9812861084938049,
"learning_rate": 8.845477075624598e-06,
"loss": 0.5974424481391907,
"step": 500
},
{
"epoch": 0.4911162847690234,
"grad_norm": 0.49719715118408203,
"learning_rate": 8.84040178622116e-06,
"loss": 0.6635510921478271,
"step": 501
},
{
"epoch": 0.4920965567945105,
"grad_norm": 0.5727866291999817,
"learning_rate": 8.835316828610801e-06,
"loss": 0.6705462336540222,
"step": 502
},
{
"epoch": 0.49307682881999754,
"grad_norm": 0.7227946519851685,
"learning_rate": 8.83022221559489e-06,
"loss": 0.6447646021842957,
"step": 503
},
{
"epoch": 0.4940571008454846,
"grad_norm": 0.4858626425266266,
"learning_rate": 8.825117959999117e-06,
"loss": 0.5024073123931885,
"step": 504
},
{
"epoch": 0.4950373728709717,
"grad_norm": 0.6126762628555298,
"learning_rate": 8.820004074673433e-06,
"loss": 0.6250107288360596,
"step": 505
},
{
"epoch": 0.4960176448964588,
"grad_norm": 0.7225095629692078,
"learning_rate": 8.814880572492043e-06,
"loss": 0.5580613017082214,
"step": 506
},
{
"epoch": 0.49699791692194584,
"grad_norm": 0.8311983942985535,
"learning_rate": 8.809747466353356e-06,
"loss": 0.7410059571266174,
"step": 507
},
{
"epoch": 0.4979781889474329,
"grad_norm": 0.460429847240448,
"learning_rate": 8.804604769179958e-06,
"loss": 0.46225789189338684,
"step": 508
},
{
"epoch": 0.49895846097291996,
"grad_norm": 0.7332952618598938,
"learning_rate": 8.799452493918586e-06,
"loss": 0.7246252298355103,
"step": 509
},
{
"epoch": 0.4999387329984071,
"grad_norm": 0.6222991943359375,
"learning_rate": 8.794290653540086e-06,
"loss": 0.7012721300125122,
"step": 510
},
{
"epoch": 0.5009190050238941,
"grad_norm": 0.5395691990852356,
"learning_rate": 8.789119261039385e-06,
"loss": 0.47234785556793213,
"step": 511
},
{
"epoch": 0.5018992770493812,
"grad_norm": 0.5847768783569336,
"learning_rate": 8.783938329435458e-06,
"loss": 0.5104413032531738,
"step": 512
},
{
"epoch": 0.5028795490748683,
"grad_norm": 0.47077733278274536,
"learning_rate": 8.778747871771293e-06,
"loss": 0.5025359392166138,
"step": 513
},
{
"epoch": 0.5038598211003553,
"grad_norm": 0.5690563321113586,
"learning_rate": 8.773547901113862e-06,
"loss": 0.526625394821167,
"step": 514
},
{
"epoch": 0.5048400931258424,
"grad_norm": 0.5545769333839417,
"learning_rate": 8.768338430554083e-06,
"loss": 0.5559873580932617,
"step": 515
},
{
"epoch": 0.5058203651513294,
"grad_norm": 0.5283054113388062,
"learning_rate": 8.763119473206795e-06,
"loss": 0.6010313034057617,
"step": 516
},
{
"epoch": 0.5068006371768166,
"grad_norm": 0.5638696551322937,
"learning_rate": 8.757891042210713e-06,
"loss": 0.7054392099380493,
"step": 517
},
{
"epoch": 0.5077809092023037,
"grad_norm": 0.5060840845108032,
"learning_rate": 8.752653150728412e-06,
"loss": 0.5484815239906311,
"step": 518
},
{
"epoch": 0.5087611812277907,
"grad_norm": 0.6180709600448608,
"learning_rate": 8.747405811946272e-06,
"loss": 0.5184864401817322,
"step": 519
},
{
"epoch": 0.5097414532532778,
"grad_norm": 0.4981224834918976,
"learning_rate": 8.742149039074463e-06,
"loss": 0.4991441071033478,
"step": 520
},
{
"epoch": 0.5107217252787648,
"grad_norm": 0.6549201011657715,
"learning_rate": 8.736882845346906e-06,
"loss": 0.6914041638374329,
"step": 521
},
{
"epoch": 0.5117019973042519,
"grad_norm": 0.559851884841919,
"learning_rate": 8.731607244021236e-06,
"loss": 0.621010422706604,
"step": 522
},
{
"epoch": 0.512682269329739,
"grad_norm": 0.6031873226165771,
"learning_rate": 8.726322248378775e-06,
"loss": 0.6384186744689941,
"step": 523
},
{
"epoch": 0.513662541355226,
"grad_norm": 0.5719360113143921,
"learning_rate": 8.72102787172449e-06,
"loss": 0.5937843918800354,
"step": 524
},
{
"epoch": 0.5146428133807132,
"grad_norm": 0.5336046814918518,
"learning_rate": 8.715724127386971e-06,
"loss": 0.6553807258605957,
"step": 525
},
{
"epoch": 0.5156230854062002,
"grad_norm": 0.6381956934928894,
"learning_rate": 8.710411028718388e-06,
"loss": 0.5429552793502808,
"step": 526
},
{
"epoch": 0.5166033574316873,
"grad_norm": 0.459621787071228,
"learning_rate": 8.705088589094458e-06,
"loss": 0.48719850182533264,
"step": 527
},
{
"epoch": 0.5175836294571744,
"grad_norm": 0.5203189253807068,
"learning_rate": 8.69975682191442e-06,
"loss": 0.633024275302887,
"step": 528
},
{
"epoch": 0.5185639014826614,
"grad_norm": 0.5211911797523499,
"learning_rate": 8.69441574060099e-06,
"loss": 0.5052200555801392,
"step": 529
},
{
"epoch": 0.5195441735081485,
"grad_norm": 0.4485588073730469,
"learning_rate": 8.68906535860033e-06,
"loss": 0.5522943139076233,
"step": 530
},
{
"epoch": 0.5205244455336356,
"grad_norm": 0.5212096571922302,
"learning_rate": 8.683705689382025e-06,
"loss": 0.6452137231826782,
"step": 531
},
{
"epoch": 0.5215047175591226,
"grad_norm": 0.54129558801651,
"learning_rate": 8.678336746439036e-06,
"loss": 0.68224036693573,
"step": 532
},
{
"epoch": 0.5224849895846098,
"grad_norm": 0.5869094133377075,
"learning_rate": 8.672958543287666e-06,
"loss": 0.5855770111083984,
"step": 533
},
{
"epoch": 0.5234652616100968,
"grad_norm": 0.6560388207435608,
"learning_rate": 8.667571093467542e-06,
"loss": 0.5621611475944519,
"step": 534
},
{
"epoch": 0.5244455336355839,
"grad_norm": 0.7414459586143494,
"learning_rate": 8.662174410541556e-06,
"loss": 0.6210222244262695,
"step": 535
},
{
"epoch": 0.525425805661071,
"grad_norm": 0.5764961242675781,
"learning_rate": 8.656768508095853e-06,
"loss": 0.6334097385406494,
"step": 536
},
{
"epoch": 0.526406077686558,
"grad_norm": 0.6460157036781311,
"learning_rate": 8.651353399739787e-06,
"loss": 0.6059669256210327,
"step": 537
},
{
"epoch": 0.5273863497120451,
"grad_norm": 0.7143170237541199,
"learning_rate": 8.645929099105886e-06,
"loss": 0.5628725290298462,
"step": 538
},
{
"epoch": 0.5283666217375321,
"grad_norm": 0.7480341792106628,
"learning_rate": 8.640495619849821e-06,
"loss": 0.6259042620658875,
"step": 539
},
{
"epoch": 0.5293468937630192,
"grad_norm": 0.4735183119773865,
"learning_rate": 8.63505297565037e-06,
"loss": 0.5363203287124634,
"step": 540
},
{
"epoch": 0.5303271657885064,
"grad_norm": 0.6985523104667664,
"learning_rate": 8.629601180209382e-06,
"loss": 0.6217838525772095,
"step": 541
},
{
"epoch": 0.5313074378139934,
"grad_norm": 0.43614763021469116,
"learning_rate": 8.624140247251745e-06,
"loss": 0.538539469242096,
"step": 542
},
{
"epoch": 0.5322877098394805,
"grad_norm": 0.45327121019363403,
"learning_rate": 8.61867019052535e-06,
"loss": 0.5169327259063721,
"step": 543
},
{
"epoch": 0.5332679818649675,
"grad_norm": 0.5319138169288635,
"learning_rate": 8.613191023801065e-06,
"loss": 0.5326805710792542,
"step": 544
},
{
"epoch": 0.5342482538904546,
"grad_norm": 0.5725491642951965,
"learning_rate": 8.607702760872679e-06,
"loss": 0.5866974592208862,
"step": 545
},
{
"epoch": 0.5352285259159417,
"grad_norm": 0.4773092269897461,
"learning_rate": 8.602205415556889e-06,
"loss": 0.5357339382171631,
"step": 546
},
{
"epoch": 0.5362087979414287,
"grad_norm": 0.5068699717521667,
"learning_rate": 8.596699001693257e-06,
"loss": 0.5180469751358032,
"step": 547
},
{
"epoch": 0.5371890699669158,
"grad_norm": 0.4932785928249359,
"learning_rate": 8.591183533144172e-06,
"loss": 0.5822703838348389,
"step": 548
},
{
"epoch": 0.5381693419924028,
"grad_norm": 0.41486823558807373,
"learning_rate": 8.585659023794818e-06,
"loss": 0.5033004283905029,
"step": 549
},
{
"epoch": 0.53914961401789,
"grad_norm": 0.48342612385749817,
"learning_rate": 8.580125487553143e-06,
"loss": 0.5849289894104004,
"step": 550
},
{
"epoch": 0.5401298860433771,
"grad_norm": 0.41986194252967834,
"learning_rate": 8.574582938349818e-06,
"loss": 0.5392067432403564,
"step": 551
},
{
"epoch": 0.5411101580688641,
"grad_norm": 0.5170447826385498,
"learning_rate": 8.569031390138202e-06,
"loss": 0.6853705644607544,
"step": 552
},
{
"epoch": 0.5420904300943512,
"grad_norm": 0.560274064540863,
"learning_rate": 8.563470856894316e-06,
"loss": 0.5396522283554077,
"step": 553
},
{
"epoch": 0.5430707021198382,
"grad_norm": 0.4939210116863251,
"learning_rate": 8.557901352616788e-06,
"loss": 0.6326029896736145,
"step": 554
},
{
"epoch": 0.5440509741453253,
"grad_norm": 0.557123064994812,
"learning_rate": 8.552322891326846e-06,
"loss": 0.5660082101821899,
"step": 555
},
{
"epoch": 0.5450312461708124,
"grad_norm": 0.49068284034729004,
"learning_rate": 8.546735487068253e-06,
"loss": 0.6054900884628296,
"step": 556
},
{
"epoch": 0.5460115181962995,
"grad_norm": 0.5976768732070923,
"learning_rate": 8.541139153907296e-06,
"loss": 0.558587372303009,
"step": 557
},
{
"epoch": 0.5469917902217866,
"grad_norm": 0.5144990086555481,
"learning_rate": 8.535533905932739e-06,
"loss": 0.49542275071144104,
"step": 558
},
{
"epoch": 0.5479720622472736,
"grad_norm": 0.4800182282924652,
"learning_rate": 8.529919757255783e-06,
"loss": 0.5426228046417236,
"step": 559
},
{
"epoch": 0.5489523342727607,
"grad_norm": 0.47078582644462585,
"learning_rate": 8.524296722010042e-06,
"loss": 0.5946049690246582,
"step": 560
},
{
"epoch": 0.5499326062982478,
"grad_norm": 0.5674675703048706,
"learning_rate": 8.518664814351502e-06,
"loss": 0.5650599002838135,
"step": 561
},
{
"epoch": 0.5509128783237348,
"grad_norm": 0.44857510924339294,
"learning_rate": 8.513024048458484e-06,
"loss": 0.5294644832611084,
"step": 562
},
{
"epoch": 0.5518931503492219,
"grad_norm": 1.0770368576049805,
"learning_rate": 8.507374438531606e-06,
"loss": 0.6094412803649902,
"step": 563
},
{
"epoch": 0.5528734223747089,
"grad_norm": 0.5167132019996643,
"learning_rate": 8.501715998793758e-06,
"loss": 0.4995839297771454,
"step": 564
},
{
"epoch": 0.553853694400196,
"grad_norm": 0.5282765030860901,
"learning_rate": 8.496048743490053e-06,
"loss": 0.581527590751648,
"step": 565
},
{
"epoch": 0.5548339664256832,
"grad_norm": 0.5315512418746948,
"learning_rate": 8.490372686887802e-06,
"loss": 0.58648282289505,
"step": 566
},
{
"epoch": 0.5558142384511702,
"grad_norm": 0.6867757439613342,
"learning_rate": 8.48468784327647e-06,
"loss": 0.6812595129013062,
"step": 567
},
{
"epoch": 0.5567945104766573,
"grad_norm": 0.5088253617286682,
"learning_rate": 8.478994226967638e-06,
"loss": 0.6054092645645142,
"step": 568
},
{
"epoch": 0.5577747825021443,
"grad_norm": 0.5003196597099304,
"learning_rate": 8.473291852294986e-06,
"loss": 0.6233060359954834,
"step": 569
},
{
"epoch": 0.5587550545276314,
"grad_norm": 0.553752064704895,
"learning_rate": 8.467580733614233e-06,
"loss": 0.640609085559845,
"step": 570
},
{
"epoch": 0.5597353265531185,
"grad_norm": 0.5246297717094421,
"learning_rate": 8.461860885303116e-06,
"loss": 0.5541980266571045,
"step": 571
},
{
"epoch": 0.5607155985786055,
"grad_norm": 0.5524768233299255,
"learning_rate": 8.45613232176134e-06,
"loss": 0.5221663117408752,
"step": 572
},
{
"epoch": 0.5616958706040927,
"grad_norm": 0.5871309041976929,
"learning_rate": 8.450395057410561e-06,
"loss": 0.600644588470459,
"step": 573
},
{
"epoch": 0.5626761426295797,
"grad_norm": 0.5105273127555847,
"learning_rate": 8.444649106694335e-06,
"loss": 0.6499615907669067,
"step": 574
},
{
"epoch": 0.5636564146550668,
"grad_norm": 0.46583378314971924,
"learning_rate": 8.438894484078086e-06,
"loss": 0.48685839772224426,
"step": 575
},
{
"epoch": 0.5646366866805539,
"grad_norm": 0.6281800270080566,
"learning_rate": 8.433131204049067e-06,
"loss": 0.6305126547813416,
"step": 576
},
{
"epoch": 0.5656169587060409,
"grad_norm": 0.4780385196208954,
"learning_rate": 8.427359281116335e-06,
"loss": 0.5220742225646973,
"step": 577
},
{
"epoch": 0.566597230731528,
"grad_norm": 0.5597613453865051,
"learning_rate": 8.421578729810693e-06,
"loss": 0.6230953931808472,
"step": 578
},
{
"epoch": 0.567577502757015,
"grad_norm": 0.5474322438240051,
"learning_rate": 8.415789564684673e-06,
"loss": 0.7254103422164917,
"step": 579
},
{
"epoch": 0.5685577747825021,
"grad_norm": 0.48552948236465454,
"learning_rate": 8.409991800312493e-06,
"loss": 0.5996748805046082,
"step": 580
},
{
"epoch": 0.5695380468079893,
"grad_norm": 0.5050139427185059,
"learning_rate": 8.404185451290017e-06,
"loss": 0.6497757434844971,
"step": 581
},
{
"epoch": 0.5705183188334763,
"grad_norm": 0.5027022957801819,
"learning_rate": 8.398370532234723e-06,
"loss": 0.7140090465545654,
"step": 582
},
{
"epoch": 0.5714985908589634,
"grad_norm": 0.39737898111343384,
"learning_rate": 8.392547057785662e-06,
"loss": 0.42447859048843384,
"step": 583
},
{
"epoch": 0.5724788628844505,
"grad_norm": 0.5955871939659119,
"learning_rate": 8.386715042603424e-06,
"loss": 0.5741234421730042,
"step": 584
},
{
"epoch": 0.5734591349099375,
"grad_norm": 0.49950310587882996,
"learning_rate": 8.380874501370098e-06,
"loss": 0.5596312284469604,
"step": 585
},
{
"epoch": 0.5744394069354246,
"grad_norm": 0.44867679476737976,
"learning_rate": 8.375025448789242e-06,
"loss": 0.49942177534103394,
"step": 586
},
{
"epoch": 0.5754196789609116,
"grad_norm": 0.4992743730545044,
"learning_rate": 8.36916789958584e-06,
"loss": 0.5540944337844849,
"step": 587
},
{
"epoch": 0.5763999509863987,
"grad_norm": 1.0275784730911255,
"learning_rate": 8.363301868506264e-06,
"loss": 0.6042492985725403,
"step": 588
},
{
"epoch": 0.5773802230118859,
"grad_norm": 0.5012497901916504,
"learning_rate": 8.357427370318239e-06,
"loss": 0.545590877532959,
"step": 589
},
{
"epoch": 0.5783604950373729,
"grad_norm": 0.44835513830184937,
"learning_rate": 8.351544419810808e-06,
"loss": 0.5180589556694031,
"step": 590
},
{
"epoch": 0.57934076706286,
"grad_norm": 0.6875878572463989,
"learning_rate": 8.345653031794292e-06,
"loss": 0.5486396551132202,
"step": 591
},
{
"epoch": 0.580321039088347,
"grad_norm": 0.5637878179550171,
"learning_rate": 8.339753221100254e-06,
"loss": 0.5895923376083374,
"step": 592
},
{
"epoch": 0.5813013111138341,
"grad_norm": 0.5228107571601868,
"learning_rate": 8.33384500258146e-06,
"loss": 0.7664077877998352,
"step": 593
},
{
"epoch": 0.5822815831393212,
"grad_norm": 0.44823041558265686,
"learning_rate": 8.327928391111841e-06,
"loss": 0.5698608160018921,
"step": 594
},
{
"epoch": 0.5832618551648082,
"grad_norm": 0.7237053513526917,
"learning_rate": 8.322003401586463e-06,
"loss": 0.6401317119598389,
"step": 595
},
{
"epoch": 0.5842421271902953,
"grad_norm": 0.49424564838409424,
"learning_rate": 8.316070048921477e-06,
"loss": 0.5031653642654419,
"step": 596
},
{
"epoch": 0.5852223992157823,
"grad_norm": 0.5683320164680481,
"learning_rate": 8.310128348054093e-06,
"loss": 0.4508371353149414,
"step": 597
},
{
"epoch": 0.5862026712412695,
"grad_norm": 0.5639766454696655,
"learning_rate": 8.304178313942536e-06,
"loss": 0.5411083698272705,
"step": 598
},
{
"epoch": 0.5871829432667566,
"grad_norm": 0.5008068084716797,
"learning_rate": 8.298219961566008e-06,
"loss": 0.5488749146461487,
"step": 599
},
{
"epoch": 0.5881632152922436,
"grad_norm": 0.6290292739868164,
"learning_rate": 8.292253305924656e-06,
"loss": 0.5881930589675903,
"step": 600
},
{
"epoch": 0.5891434873177307,
"grad_norm": 0.47255659103393555,
"learning_rate": 8.286278362039527e-06,
"loss": 0.6302595138549805,
"step": 601
},
{
"epoch": 0.5901237593432177,
"grad_norm": 0.5140688419342041,
"learning_rate": 8.280295144952537e-06,
"loss": 0.6385571956634521,
"step": 602
},
{
"epoch": 0.5911040313687048,
"grad_norm": 0.46549689769744873,
"learning_rate": 8.274303669726427e-06,
"loss": 0.5014755129814148,
"step": 603
},
{
"epoch": 0.5920843033941919,
"grad_norm": 0.41634997725486755,
"learning_rate": 8.268303951444727e-06,
"loss": 0.495285302400589,
"step": 604
},
{
"epoch": 0.5930645754196789,
"grad_norm": 1.8150643110275269,
"learning_rate": 8.262296005211722e-06,
"loss": 0.47511863708496094,
"step": 605
},
{
"epoch": 0.5940448474451661,
"grad_norm": 0.5939587950706482,
"learning_rate": 8.25627984615241e-06,
"loss": 0.676226019859314,
"step": 606
},
{
"epoch": 0.5950251194706531,
"grad_norm": 0.8281105756759644,
"learning_rate": 8.250255489412464e-06,
"loss": 0.5587377548217773,
"step": 607
},
{
"epoch": 0.5960053914961402,
"grad_norm": 0.4769623577594757,
"learning_rate": 8.244222950158194e-06,
"loss": 0.6315554976463318,
"step": 608
},
{
"epoch": 0.5969856635216273,
"grad_norm": 0.4458005726337433,
"learning_rate": 8.238182243576512e-06,
"loss": 0.4711849093437195,
"step": 609
},
{
"epoch": 0.5979659355471143,
"grad_norm": 0.47836652398109436,
"learning_rate": 8.232133384874887e-06,
"loss": 0.573494553565979,
"step": 610
},
{
"epoch": 0.5989462075726014,
"grad_norm": 0.44612884521484375,
"learning_rate": 8.226076389281316e-06,
"loss": 0.6105935573577881,
"step": 611
},
{
"epoch": 0.5999264795980884,
"grad_norm": 0.4678818881511688,
"learning_rate": 8.220011272044276e-06,
"loss": 0.5955692529678345,
"step": 612
},
{
"epoch": 0.6009067516235755,
"grad_norm": 0.47035977244377136,
"learning_rate": 8.213938048432697e-06,
"loss": 0.5746393799781799,
"step": 613
},
{
"epoch": 0.6018870236490627,
"grad_norm": 0.43257585167884827,
"learning_rate": 8.207856733735908e-06,
"loss": 0.5300241112709045,
"step": 614
},
{
"epoch": 0.6028672956745497,
"grad_norm": 0.42712169885635376,
"learning_rate": 8.201767343263612e-06,
"loss": 0.5474828481674194,
"step": 615
},
{
"epoch": 0.6038475677000368,
"grad_norm": 0.4949519634246826,
"learning_rate": 8.195669892345842e-06,
"loss": 0.539609432220459,
"step": 616
},
{
"epoch": 0.6048278397255238,
"grad_norm": 0.466345876455307,
"learning_rate": 8.189564396332927e-06,
"loss": 0.5650939345359802,
"step": 617
},
{
"epoch": 0.6058081117510109,
"grad_norm": 0.4587682783603668,
"learning_rate": 8.183450870595443e-06,
"loss": 0.4892288148403168,
"step": 618
},
{
"epoch": 0.606788383776498,
"grad_norm": 0.4350122809410095,
"learning_rate": 8.177329330524182e-06,
"loss": 0.5417685508728027,
"step": 619
},
{
"epoch": 0.607768655801985,
"grad_norm": 0.44926467537879944,
"learning_rate": 8.171199791530117e-06,
"loss": 0.5792407989501953,
"step": 620
},
{
"epoch": 0.6087489278274721,
"grad_norm": 0.48948216438293457,
"learning_rate": 8.165062269044353e-06,
"loss": 0.5384902954101562,
"step": 621
},
{
"epoch": 0.6097291998529591,
"grad_norm": 0.4751755893230438,
"learning_rate": 8.158916778518097e-06,
"loss": 0.571563184261322,
"step": 622
},
{
"epoch": 0.6107094718784463,
"grad_norm": 0.5100057125091553,
"learning_rate": 8.152763335422612e-06,
"loss": 0.573106050491333,
"step": 623
},
{
"epoch": 0.6116897439039334,
"grad_norm": 0.5238609313964844,
"learning_rate": 8.146601955249187e-06,
"loss": 0.5689284801483154,
"step": 624
},
{
"epoch": 0.6126700159294204,
"grad_norm": 0.5913124680519104,
"learning_rate": 8.140432653509089e-06,
"loss": 0.5576924085617065,
"step": 625
},
{
"epoch": 0.6136502879549075,
"grad_norm": 0.5904691815376282,
"learning_rate": 8.134255445733524e-06,
"loss": 0.6356901526451111,
"step": 626
},
{
"epoch": 0.6146305599803945,
"grad_norm": 0.43635961413383484,
"learning_rate": 8.128070347473609e-06,
"loss": 0.5801491737365723,
"step": 627
},
{
"epoch": 0.6156108320058816,
"grad_norm": 0.541221022605896,
"learning_rate": 8.121877374300318e-06,
"loss": 0.5040395259857178,
"step": 628
},
{
"epoch": 0.6165911040313687,
"grad_norm": 0.4131251573562622,
"learning_rate": 8.115676541804456e-06,
"loss": 0.5251767635345459,
"step": 629
},
{
"epoch": 0.6175713760568557,
"grad_norm": 0.45905354619026184,
"learning_rate": 8.109467865596612e-06,
"loss": 0.5374346971511841,
"step": 630
},
{
"epoch": 0.6185516480823429,
"grad_norm": 0.4692007005214691,
"learning_rate": 8.10325136130712e-06,
"loss": 0.5571949481964111,
"step": 631
},
{
"epoch": 0.6195319201078299,
"grad_norm": 0.5968528985977173,
"learning_rate": 8.097027044586021e-06,
"loss": 0.7638495564460754,
"step": 632
},
{
"epoch": 0.620512192133317,
"grad_norm": 0.5194112062454224,
"learning_rate": 8.090794931103026e-06,
"loss": 0.5503364205360413,
"step": 633
},
{
"epoch": 0.6214924641588041,
"grad_norm": 0.536645770072937,
"learning_rate": 8.084555036547475e-06,
"loss": 0.6009919047355652,
"step": 634
},
{
"epoch": 0.6224727361842911,
"grad_norm": 0.43642154335975647,
"learning_rate": 8.078307376628292e-06,
"loss": 0.4636218547821045,
"step": 635
},
{
"epoch": 0.6234530082097782,
"grad_norm": 0.4677816331386566,
"learning_rate": 8.072051967073955e-06,
"loss": 0.5536704063415527,
"step": 636
},
{
"epoch": 0.6244332802352652,
"grad_norm": 0.41060787439346313,
"learning_rate": 8.065788823632451e-06,
"loss": 0.41056686639785767,
"step": 637
},
{
"epoch": 0.6254135522607523,
"grad_norm": 0.4008098840713501,
"learning_rate": 8.059517962071234e-06,
"loss": 0.4722449779510498,
"step": 638
},
{
"epoch": 0.6263938242862395,
"grad_norm": 0.4182589054107666,
"learning_rate": 8.053239398177191e-06,
"loss": 0.5422540307044983,
"step": 639
},
{
"epoch": 0.6273740963117265,
"grad_norm": 0.5321866869926453,
"learning_rate": 8.046953147756601e-06,
"loss": 0.5848727226257324,
"step": 640
},
{
"epoch": 0.6283543683372136,
"grad_norm": 0.3844715356826782,
"learning_rate": 8.04065922663509e-06,
"loss": 0.4720197319984436,
"step": 641
},
{
"epoch": 0.6293346403627007,
"grad_norm": 0.5278245806694031,
"learning_rate": 8.034357650657597e-06,
"loss": 0.779409646987915,
"step": 642
},
{
"epoch": 0.6303149123881877,
"grad_norm": 0.4623364806175232,
"learning_rate": 8.028048435688333e-06,
"loss": 0.5392259955406189,
"step": 643
},
{
"epoch": 0.6312951844136748,
"grad_norm": 0.5639168620109558,
"learning_rate": 8.02173159761074e-06,
"loss": 0.5717742443084717,
"step": 644
},
{
"epoch": 0.6322754564391618,
"grad_norm": 0.46666693687438965,
"learning_rate": 8.015407152327448e-06,
"loss": 0.6617095470428467,
"step": 645
},
{
"epoch": 0.633255728464649,
"grad_norm": 0.5085638761520386,
"learning_rate": 8.009075115760243e-06,
"loss": 0.6582084894180298,
"step": 646
},
{
"epoch": 0.6342360004901361,
"grad_norm": 0.43695488572120667,
"learning_rate": 8.002735503850016e-06,
"loss": 0.5222762823104858,
"step": 647
},
{
"epoch": 0.6352162725156231,
"grad_norm": 0.46036913990974426,
"learning_rate": 7.996388332556735e-06,
"loss": 0.6211761832237244,
"step": 648
},
{
"epoch": 0.6361965445411102,
"grad_norm": 0.467732697725296,
"learning_rate": 7.990033617859396e-06,
"loss": 0.4865642488002777,
"step": 649
},
{
"epoch": 0.6371768165665972,
"grad_norm": 0.5058535933494568,
"learning_rate": 7.983671375755986e-06,
"loss": 0.5791799426078796,
"step": 650
},
{
"epoch": 0.6381570885920843,
"grad_norm": 0.48842498660087585,
"learning_rate": 7.97730162226344e-06,
"loss": 0.6354533433914185,
"step": 651
},
{
"epoch": 0.6391373606175714,
"grad_norm": 0.45112404227256775,
"learning_rate": 7.970924373417608e-06,
"loss": 0.6128513813018799,
"step": 652
},
{
"epoch": 0.6401176326430584,
"grad_norm": 0.4545324444770813,
"learning_rate": 7.964539645273204e-06,
"loss": 0.5118072032928467,
"step": 653
},
{
"epoch": 0.6410979046685455,
"grad_norm": 0.3487434387207031,
"learning_rate": 7.958147453903773e-06,
"loss": 0.4512544870376587,
"step": 654
},
{
"epoch": 0.6420781766940326,
"grad_norm": 0.42427530884742737,
"learning_rate": 7.951747815401651e-06,
"loss": 0.5383853912353516,
"step": 655
},
{
"epoch": 0.6430584487195197,
"grad_norm": 0.6650370359420776,
"learning_rate": 7.945340745877919e-06,
"loss": 0.559260368347168,
"step": 656
},
{
"epoch": 0.6440387207450068,
"grad_norm": 0.49584558606147766,
"learning_rate": 7.938926261462366e-06,
"loss": 0.6616352200508118,
"step": 657
},
{
"epoch": 0.6450189927704938,
"grad_norm": 0.4980330765247345,
"learning_rate": 7.932504378303452e-06,
"loss": 0.700371265411377,
"step": 658
},
{
"epoch": 0.6459992647959809,
"grad_norm": 0.6228008270263672,
"learning_rate": 7.92607511256826e-06,
"loss": 0.48944735527038574,
"step": 659
},
{
"epoch": 0.6469795368214679,
"grad_norm": 0.4910522997379303,
"learning_rate": 7.919638480442453e-06,
"loss": 0.5541126728057861,
"step": 660
},
{
"epoch": 0.647959808846955,
"grad_norm": 0.4713754653930664,
"learning_rate": 7.913194498130252e-06,
"loss": 0.4977142810821533,
"step": 661
},
{
"epoch": 0.6489400808724421,
"grad_norm": 0.4953506886959076,
"learning_rate": 7.90674318185437e-06,
"loss": 0.6201741099357605,
"step": 662
},
{
"epoch": 0.6499203528979292,
"grad_norm": 0.5087093114852905,
"learning_rate": 7.900284547855992e-06,
"loss": 0.6099828481674194,
"step": 663
},
{
"epoch": 0.6509006249234163,
"grad_norm": 0.5764956474304199,
"learning_rate": 7.893818612394717e-06,
"loss": 0.5428135395050049,
"step": 664
},
{
"epoch": 0.6518808969489033,
"grad_norm": 0.419497549533844,
"learning_rate": 7.887345391748533e-06,
"loss": 0.5326085686683655,
"step": 665
},
{
"epoch": 0.6528611689743904,
"grad_norm": 0.7227897047996521,
"learning_rate": 7.880864902213765e-06,
"loss": 0.6780993342399597,
"step": 666
},
{
"epoch": 0.6538414409998775,
"grad_norm": 0.5869578719139099,
"learning_rate": 7.874377160105037e-06,
"loss": 0.5142019987106323,
"step": 667
},
{
"epoch": 0.6548217130253645,
"grad_norm": 0.49598124623298645,
"learning_rate": 7.86788218175523e-06,
"loss": 0.6338085532188416,
"step": 668
},
{
"epoch": 0.6558019850508516,
"grad_norm": 0.44636520743370056,
"learning_rate": 7.861379983515449e-06,
"loss": 0.6195379495620728,
"step": 669
},
{
"epoch": 0.6567822570763386,
"grad_norm": 0.4613161087036133,
"learning_rate": 7.854870581754967e-06,
"loss": 0.4916152358055115,
"step": 670
},
{
"epoch": 0.6577625291018258,
"grad_norm": 0.4831985831260681,
"learning_rate": 7.848353992861195e-06,
"loss": 0.5819408297538757,
"step": 671
},
{
"epoch": 0.6587428011273129,
"grad_norm": 0.3623875677585602,
"learning_rate": 7.841830233239638e-06,
"loss": 0.5120476484298706,
"step": 672
},
{
"epoch": 0.6597230731527999,
"grad_norm": 0.40355730056762695,
"learning_rate": 7.835299319313854e-06,
"loss": 0.5201019048690796,
"step": 673
},
{
"epoch": 0.660703345178287,
"grad_norm": 0.36736786365509033,
"learning_rate": 7.82876126752541e-06,
"loss": 0.508677065372467,
"step": 674
},
{
"epoch": 0.661683617203774,
"grad_norm": 0.3610830307006836,
"learning_rate": 7.822216094333847e-06,
"loss": 0.45365607738494873,
"step": 675
},
{
"epoch": 0.6626638892292611,
"grad_norm": 0.4933203160762787,
"learning_rate": 7.815663816216626e-06,
"loss": 0.6202186346054077,
"step": 676
},
{
"epoch": 0.6636441612547482,
"grad_norm": 0.47557780146598816,
"learning_rate": 7.8091044496691e-06,
"loss": 0.5591720938682556,
"step": 677
},
{
"epoch": 0.6646244332802352,
"grad_norm": 0.4458239674568176,
"learning_rate": 7.80253801120447e-06,
"loss": 0.5407501459121704,
"step": 678
},
{
"epoch": 0.6656047053057224,
"grad_norm": 0.3904792070388794,
"learning_rate": 7.795964517353734e-06,
"loss": 0.48473650217056274,
"step": 679
},
{
"epoch": 0.6665849773312094,
"grad_norm": 0.4341505169868469,
"learning_rate": 7.789383984665658e-06,
"loss": 0.5052921772003174,
"step": 680
},
{
"epoch": 0.6675652493566965,
"grad_norm": 0.40881484746932983,
"learning_rate": 7.782796429706721e-06,
"loss": 0.497772753238678,
"step": 681
},
{
"epoch": 0.6685455213821836,
"grad_norm": 0.45834603905677795,
"learning_rate": 7.776201869061086e-06,
"loss": 0.5965415835380554,
"step": 682
},
{
"epoch": 0.6695257934076706,
"grad_norm": 0.4278583824634552,
"learning_rate": 7.769600319330553e-06,
"loss": 0.4990220069885254,
"step": 683
},
{
"epoch": 0.6705060654331577,
"grad_norm": 0.4458213150501251,
"learning_rate": 7.762991797134513e-06,
"loss": 0.5628777146339417,
"step": 684
},
{
"epoch": 0.6714863374586447,
"grad_norm": 0.38521257042884827,
"learning_rate": 7.756376319109917e-06,
"loss": 0.49859195947647095,
"step": 685
},
{
"epoch": 0.6724666094841318,
"grad_norm": 0.45445364713668823,
"learning_rate": 7.74975390191122e-06,
"loss": 0.5584304928779602,
"step": 686
},
{
"epoch": 0.673446881509619,
"grad_norm": 0.47879061102867126,
"learning_rate": 7.743124562210351e-06,
"loss": 0.6553927659988403,
"step": 687
},
{
"epoch": 0.674427153535106,
"grad_norm": 0.41177281737327576,
"learning_rate": 7.736488316696663e-06,
"loss": 0.42918023467063904,
"step": 688
},
{
"epoch": 0.6754074255605931,
"grad_norm": 0.569825291633606,
"learning_rate": 7.729845182076896e-06,
"loss": 0.6259717345237732,
"step": 689
},
{
"epoch": 0.6763876975860801,
"grad_norm": 0.5575708746910095,
"learning_rate": 7.723195175075136e-06,
"loss": 0.5303859710693359,
"step": 690
},
{
"epoch": 0.6773679696115672,
"grad_norm": 0.5137466788291931,
"learning_rate": 7.716538312432767e-06,
"loss": 0.5614829063415527,
"step": 691
},
{
"epoch": 0.6783482416370543,
"grad_norm": 0.440349817276001,
"learning_rate": 7.70987461090843e-06,
"loss": 0.5800457000732422,
"step": 692
},
{
"epoch": 0.6793285136625413,
"grad_norm": 0.4403285086154938,
"learning_rate": 7.703204087277989e-06,
"loss": 0.5892413854598999,
"step": 693
},
{
"epoch": 0.6803087856880284,
"grad_norm": 0.4497002065181732,
"learning_rate": 7.696526758334477e-06,
"loss": 0.562138557434082,
"step": 694
},
{
"epoch": 0.6812890577135156,
"grad_norm": 0.4190099239349365,
"learning_rate": 7.689842640888063e-06,
"loss": 0.4345162510871887,
"step": 695
},
{
"epoch": 0.6822693297390026,
"grad_norm": 0.5989111661911011,
"learning_rate": 7.683151751766005e-06,
"loss": 0.512845516204834,
"step": 696
},
{
"epoch": 0.6832496017644897,
"grad_norm": 0.4910346269607544,
"learning_rate": 7.676454107812608e-06,
"loss": 0.733169674873352,
"step": 697
},
{
"epoch": 0.6842298737899767,
"grad_norm": 0.42751452326774597,
"learning_rate": 7.669749725889182e-06,
"loss": 0.5693173408508301,
"step": 698
},
{
"epoch": 0.6852101458154638,
"grad_norm": 0.5142582654953003,
"learning_rate": 7.663038622873999e-06,
"loss": 0.5878850221633911,
"step": 699
},
{
"epoch": 0.6861904178409509,
"grad_norm": 0.8307844400405884,
"learning_rate": 7.656320815662257e-06,
"loss": 0.6216428279876709,
"step": 700
},
{
"epoch": 0.6871706898664379,
"grad_norm": 0.47172537446022034,
"learning_rate": 7.649596321166024e-06,
"loss": 0.549314022064209,
"step": 701
},
{
"epoch": 0.688150961891925,
"grad_norm": 0.47968435287475586,
"learning_rate": 7.64286515631421e-06,
"loss": 0.48883455991744995,
"step": 702
},
{
"epoch": 0.689131233917412,
"grad_norm": 0.6333498358726501,
"learning_rate": 7.636127338052513e-06,
"loss": 0.6025599241256714,
"step": 703
},
{
"epoch": 0.6901115059428992,
"grad_norm": 0.640839159488678,
"learning_rate": 7.629382883343381e-06,
"loss": 0.5830345153808594,
"step": 704
},
{
"epoch": 0.6910917779683863,
"grad_norm": 0.39502251148223877,
"learning_rate": 7.622631809165972e-06,
"loss": 0.406398743391037,
"step": 705
},
{
"epoch": 0.6920720499938733,
"grad_norm": 0.3786174952983856,
"learning_rate": 7.615874132516108e-06,
"loss": 0.39167964458465576,
"step": 706
},
{
"epoch": 0.6930523220193604,
"grad_norm": 0.5710326433181763,
"learning_rate": 7.60910987040623e-06,
"loss": 0.5021277666091919,
"step": 707
},
{
"epoch": 0.6940325940448474,
"grad_norm": 0.45050370693206787,
"learning_rate": 7.602339039865362e-06,
"loss": 0.5112322568893433,
"step": 708
},
{
"epoch": 0.6950128660703345,
"grad_norm": 0.49357518553733826,
"learning_rate": 7.595561657939061e-06,
"loss": 0.5479626655578613,
"step": 709
},
{
"epoch": 0.6959931380958216,
"grad_norm": 0.6366994380950928,
"learning_rate": 7.588777741689376e-06,
"loss": 0.6149463057518005,
"step": 710
},
{
"epoch": 0.6969734101213086,
"grad_norm": 0.4910570979118347,
"learning_rate": 7.5819873081948105e-06,
"loss": 0.5862258076667786,
"step": 711
},
{
"epoch": 0.6979536821467958,
"grad_norm": 0.41112977266311646,
"learning_rate": 7.575190374550272e-06,
"loss": 0.49828004837036133,
"step": 712
},
{
"epoch": 0.6989339541722828,
"grad_norm": 0.5109178423881531,
"learning_rate": 7.568386957867033e-06,
"loss": 0.5555713176727295,
"step": 713
},
{
"epoch": 0.6999142261977699,
"grad_norm": 0.46127986907958984,
"learning_rate": 7.561577075272686e-06,
"loss": 0.599381685256958,
"step": 714
},
{
"epoch": 0.700894498223257,
"grad_norm": 0.4238424301147461,
"learning_rate": 7.554760743911104e-06,
"loss": 0.4853740930557251,
"step": 715
},
{
"epoch": 0.701874770248744,
"grad_norm": 0.5007547736167908,
"learning_rate": 7.5479379809423905e-06,
"loss": 0.5677107572555542,
"step": 716
},
{
"epoch": 0.7028550422742311,
"grad_norm": 0.44779297709465027,
"learning_rate": 7.541108803542846e-06,
"loss": 0.5496746301651001,
"step": 717
},
{
"epoch": 0.7038353142997181,
"grad_norm": 0.4502071738243103,
"learning_rate": 7.534273228904916e-06,
"loss": 0.5206332206726074,
"step": 718
},
{
"epoch": 0.7048155863252052,
"grad_norm": 0.43733149766921997,
"learning_rate": 7.527431274237149e-06,
"loss": 0.4367731809616089,
"step": 719
},
{
"epoch": 0.7057958583506924,
"grad_norm": 0.511375367641449,
"learning_rate": 7.52058295676416e-06,
"loss": 0.5530825853347778,
"step": 720
},
{
"epoch": 0.7067761303761794,
"grad_norm": 0.5255171656608582,
"learning_rate": 7.5137282937265796e-06,
"loss": 0.5948051810264587,
"step": 721
},
{
"epoch": 0.7077564024016665,
"grad_norm": 0.4267750680446625,
"learning_rate": 7.506867302381016e-06,
"loss": 0.506665050983429,
"step": 722
},
{
"epoch": 0.7087366744271535,
"grad_norm": 0.40493378043174744,
"learning_rate": 7.500000000000001e-06,
"loss": 0.46698686480522156,
"step": 723
},
{
"epoch": 0.7097169464526406,
"grad_norm": 0.4602975845336914,
"learning_rate": 7.493126403871964e-06,
"loss": 0.4653869867324829,
"step": 724
},
{
"epoch": 0.7106972184781277,
"grad_norm": 0.5179893970489502,
"learning_rate": 7.486246531301178e-06,
"loss": 0.6415598392486572,
"step": 725
},
{
"epoch": 0.7116774905036147,
"grad_norm": 0.4917897582054138,
"learning_rate": 7.479360399607707e-06,
"loss": 0.5062524080276489,
"step": 726
},
{
"epoch": 0.7126577625291018,
"grad_norm": 0.42604467272758484,
"learning_rate": 7.472468026127385e-06,
"loss": 0.5530419945716858,
"step": 727
},
{
"epoch": 0.7136380345545889,
"grad_norm": 0.48969146609306335,
"learning_rate": 7.465569428211752e-06,
"loss": 0.6023317575454712,
"step": 728
},
{
"epoch": 0.714618306580076,
"grad_norm": 0.4161062240600586,
"learning_rate": 7.45866462322802e-06,
"loss": 0.5091547966003418,
"step": 729
},
{
"epoch": 0.7155985786055631,
"grad_norm": 0.4053821861743927,
"learning_rate": 7.451753628559027e-06,
"loss": 0.4768165349960327,
"step": 730
},
{
"epoch": 0.7165788506310501,
"grad_norm": 0.4812045097351074,
"learning_rate": 7.444836461603195e-06,
"loss": 0.6372095346450806,
"step": 731
},
{
"epoch": 0.7175591226565372,
"grad_norm": 0.414230614900589,
"learning_rate": 7.437913139774482e-06,
"loss": 0.4736112058162689,
"step": 732
},
{
"epoch": 0.7185393946820242,
"grad_norm": 0.40388473868370056,
"learning_rate": 7.430983680502344e-06,
"loss": 0.452359139919281,
"step": 733
},
{
"epoch": 0.7195196667075113,
"grad_norm": 0.4632888436317444,
"learning_rate": 7.424048101231687e-06,
"loss": 0.48538899421691895,
"step": 734
},
{
"epoch": 0.7204999387329984,
"grad_norm": 0.5188507437705994,
"learning_rate": 7.4171064194228196e-06,
"loss": 0.49239328503608704,
"step": 735
},
{
"epoch": 0.7214802107584855,
"grad_norm": 0.4390290677547455,
"learning_rate": 7.410158652551421e-06,
"loss": 0.5711298584938049,
"step": 736
},
{
"epoch": 0.7224604827839726,
"grad_norm": 0.49117282032966614,
"learning_rate": 7.403204818108487e-06,
"loss": 0.6047239899635315,
"step": 737
},
{
"epoch": 0.7234407548094596,
"grad_norm": 0.5015243887901306,
"learning_rate": 7.396244933600285e-06,
"loss": 0.6040792465209961,
"step": 738
},
{
"epoch": 0.7244210268349467,
"grad_norm": 1.2947595119476318,
"learning_rate": 7.3892790165483164e-06,
"loss": 0.5264707207679749,
"step": 739
},
{
"epoch": 0.7254012988604338,
"grad_norm": 0.49317121505737305,
"learning_rate": 7.382307084489269e-06,
"loss": 0.5585997104644775,
"step": 740
},
{
"epoch": 0.7263815708859208,
"grad_norm": 0.4826045632362366,
"learning_rate": 7.3753291549749764e-06,
"loss": 0.5586689114570618,
"step": 741
},
{
"epoch": 0.7273618429114079,
"grad_norm": 0.4288710057735443,
"learning_rate": 7.368345245572362e-06,
"loss": 0.43690255284309387,
"step": 742
},
{
"epoch": 0.7283421149368949,
"grad_norm": 0.4085274934768677,
"learning_rate": 7.361355373863415e-06,
"loss": 0.46534493565559387,
"step": 743
},
{
"epoch": 0.729322386962382,
"grad_norm": 0.627498209476471,
"learning_rate": 7.354359557445125e-06,
"loss": 0.5892662405967712,
"step": 744
},
{
"epoch": 0.7303026589878692,
"grad_norm": 0.4909018874168396,
"learning_rate": 7.347357813929455e-06,
"loss": 0.6983001232147217,
"step": 745
},
{
"epoch": 0.7312829310133562,
"grad_norm": 0.4056415855884552,
"learning_rate": 7.3403501609432835e-06,
"loss": 0.5443574786186218,
"step": 746
},
{
"epoch": 0.7322632030388433,
"grad_norm": 0.4906488060951233,
"learning_rate": 7.333336616128369e-06,
"loss": 0.5687349438667297,
"step": 747
},
{
"epoch": 0.7332434750643303,
"grad_norm": 0.5490400791168213,
"learning_rate": 7.326317197141304e-06,
"loss": 0.6537408828735352,
"step": 748
},
{
"epoch": 0.7342237470898174,
"grad_norm": 0.45163455605506897,
"learning_rate": 7.319291921653464e-06,
"loss": 0.5867169499397278,
"step": 749
},
{
"epoch": 0.7352040191153045,
"grad_norm": 0.4942396283149719,
"learning_rate": 7.312260807350975e-06,
"loss": 0.5694727897644043,
"step": 750
},
{
"epoch": 0.7361842911407915,
"grad_norm": 0.4633714258670807,
"learning_rate": 7.305223871934657e-06,
"loss": 0.5190877914428711,
"step": 751
},
{
"epoch": 0.7371645631662787,
"grad_norm": 0.4772140383720398,
"learning_rate": 7.298181133119987e-06,
"loss": 0.4779459834098816,
"step": 752
},
{
"epoch": 0.7381448351917658,
"grad_norm": 0.5711033344268799,
"learning_rate": 7.291132608637053e-06,
"loss": 0.6020298004150391,
"step": 753
},
{
"epoch": 0.7391251072172528,
"grad_norm": 0.49676769971847534,
"learning_rate": 7.284078316230504e-06,
"loss": 0.4763784408569336,
"step": 754
},
{
"epoch": 0.7401053792427399,
"grad_norm": 0.42109596729278564,
"learning_rate": 7.2770182736595164e-06,
"loss": 0.48202162981033325,
"step": 755
},
{
"epoch": 0.7410856512682269,
"grad_norm": 0.6203530430793762,
"learning_rate": 7.269952498697734e-06,
"loss": 0.5142531394958496,
"step": 756
},
{
"epoch": 0.742065923293714,
"grad_norm": 0.5919861793518066,
"learning_rate": 7.262881009133242e-06,
"loss": 0.6466728448867798,
"step": 757
},
{
"epoch": 0.7430461953192011,
"grad_norm": 0.5874010920524597,
"learning_rate": 7.255803822768504e-06,
"loss": 0.48682892322540283,
"step": 758
},
{
"epoch": 0.7440264673446881,
"grad_norm": 0.505060613155365,
"learning_rate": 7.24872095742033e-06,
"loss": 0.5157826542854309,
"step": 759
},
{
"epoch": 0.7450067393701753,
"grad_norm": 0.672005295753479,
"learning_rate": 7.241632430919823e-06,
"loss": 0.5602332949638367,
"step": 760
},
{
"epoch": 0.7459870113956623,
"grad_norm": 0.42120110988616943,
"learning_rate": 7.234538261112342e-06,
"loss": 0.42714083194732666,
"step": 761
},
{
"epoch": 0.7469672834211494,
"grad_norm": 0.7334301471710205,
"learning_rate": 7.227438465857448e-06,
"loss": 0.6078978776931763,
"step": 762
},
{
"epoch": 0.7479475554466365,
"grad_norm": 0.6107139587402344,
"learning_rate": 7.2203330630288714e-06,
"loss": 0.4394034445285797,
"step": 763
},
{
"epoch": 0.7489278274721235,
"grad_norm": 0.6603996753692627,
"learning_rate": 7.213222070514453e-06,
"loss": 0.7087306976318359,
"step": 764
},
{
"epoch": 0.7499080994976106,
"grad_norm": 0.5546720623970032,
"learning_rate": 7.206105506216107e-06,
"loss": 0.4940693974494934,
"step": 765
},
{
"epoch": 0.7508883715230976,
"grad_norm": 0.8178077340126038,
"learning_rate": 7.198983388049779e-06,
"loss": 0.5339226126670837,
"step": 766
},
{
"epoch": 0.7518686435485847,
"grad_norm": 0.5073369145393372,
"learning_rate": 7.191855733945388e-06,
"loss": 0.5095759034156799,
"step": 767
},
{
"epoch": 0.7528489155740719,
"grad_norm": 0.39711958169937134,
"learning_rate": 7.1847225618467975e-06,
"loss": 0.41847747564315796,
"step": 768
},
{
"epoch": 0.7538291875995589,
"grad_norm": 0.5229822397232056,
"learning_rate": 7.177583889711763e-06,
"loss": 0.5046806335449219,
"step": 769
},
{
"epoch": 0.754809459625046,
"grad_norm": 0.4823722839355469,
"learning_rate": 7.170439735511876e-06,
"loss": 0.5876001119613647,
"step": 770
},
{
"epoch": 0.755789731650533,
"grad_norm": 0.7995231747627258,
"learning_rate": 7.163290117232542e-06,
"loss": 0.5199307799339294,
"step": 771
},
{
"epoch": 0.7567700036760201,
"grad_norm": 0.5239847898483276,
"learning_rate": 7.1561350528729145e-06,
"loss": 0.5662480592727661,
"step": 772
},
{
"epoch": 0.7577502757015072,
"grad_norm": 0.4220186173915863,
"learning_rate": 7.148974560445859e-06,
"loss": 0.409650593996048,
"step": 773
},
{
"epoch": 0.7587305477269942,
"grad_norm": 0.40589994192123413,
"learning_rate": 7.1418086579779075e-06,
"loss": 0.482654869556427,
"step": 774
},
{
"epoch": 0.7597108197524813,
"grad_norm": 0.4408826529979706,
"learning_rate": 7.1346373635092095e-06,
"loss": 0.5564194917678833,
"step": 775
},
{
"epoch": 0.7606910917779683,
"grad_norm": 0.5282425284385681,
"learning_rate": 7.127460695093493e-06,
"loss": 0.6081736087799072,
"step": 776
},
{
"epoch": 0.7616713638034555,
"grad_norm": 3.734576940536499,
"learning_rate": 7.12027867079801e-06,
"loss": 0.5723667144775391,
"step": 777
},
{
"epoch": 0.7626516358289426,
"grad_norm": 0.494312584400177,
"learning_rate": 7.113091308703498e-06,
"loss": 0.5435426831245422,
"step": 778
},
{
"epoch": 0.7636319078544296,
"grad_norm": 0.42854106426239014,
"learning_rate": 7.105898626904134e-06,
"loss": 0.5378055572509766,
"step": 779
},
{
"epoch": 0.7646121798799167,
"grad_norm": 0.48089906573295593,
"learning_rate": 7.098700643507485e-06,
"loss": 0.591333270072937,
"step": 780
},
{
"epoch": 0.7655924519054037,
"grad_norm": 0.3623630106449127,
"learning_rate": 7.0914973766344645e-06,
"loss": 0.4385789632797241,
"step": 781
},
{
"epoch": 0.7665727239308908,
"grad_norm": 0.6053583025932312,
"learning_rate": 7.0842888444192905e-06,
"loss": 0.6852695941925049,
"step": 782
},
{
"epoch": 0.7675529959563779,
"grad_norm": 0.43562591075897217,
"learning_rate": 7.0770750650094335e-06,
"loss": 0.5775709748268127,
"step": 783
},
{
"epoch": 0.7685332679818649,
"grad_norm": 0.39110779762268066,
"learning_rate": 7.069856056565573e-06,
"loss": 0.48135459423065186,
"step": 784
},
{
"epoch": 0.7695135400073521,
"grad_norm": 0.41838300228118896,
"learning_rate": 7.062631837261556e-06,
"loss": 0.5081921815872192,
"step": 785
},
{
"epoch": 0.7704938120328391,
"grad_norm": 0.37609559297561646,
"learning_rate": 7.055402425284346e-06,
"loss": 0.4609271287918091,
"step": 786
},
{
"epoch": 0.7714740840583262,
"grad_norm": 0.4856472909450531,
"learning_rate": 7.048167838833977e-06,
"loss": 0.5818092823028564,
"step": 787
},
{
"epoch": 0.7724543560838133,
"grad_norm": 0.691068708896637,
"learning_rate": 7.040928096123516e-06,
"loss": 0.5119889974594116,
"step": 788
},
{
"epoch": 0.7734346281093003,
"grad_norm": 0.5112810730934143,
"learning_rate": 7.033683215379002e-06,
"loss": 0.5305988788604736,
"step": 789
},
{
"epoch": 0.7744149001347874,
"grad_norm": 0.4650510549545288,
"learning_rate": 7.026433214839416e-06,
"loss": 0.5277286171913147,
"step": 790
},
{
"epoch": 0.7753951721602744,
"grad_norm": 0.40126878023147583,
"learning_rate": 7.019178112756625e-06,
"loss": 0.46395719051361084,
"step": 791
},
{
"epoch": 0.7763754441857615,
"grad_norm": 0.4711439609527588,
"learning_rate": 7.011917927395341e-06,
"loss": 0.49734482169151306,
"step": 792
},
{
"epoch": 0.7773557162112487,
"grad_norm": 0.4137747585773468,
"learning_rate": 7.004652677033069e-06,
"loss": 0.5272625684738159,
"step": 793
},
{
"epoch": 0.7783359882367357,
"grad_norm": 0.44795504212379456,
"learning_rate": 6.997382379960072e-06,
"loss": 0.5505017042160034,
"step": 794
},
{
"epoch": 0.7793162602622228,
"grad_norm": 0.43027734756469727,
"learning_rate": 6.990107054479313e-06,
"loss": 0.5101114511489868,
"step": 795
},
{
"epoch": 0.7802965322877098,
"grad_norm": 0.35909900069236755,
"learning_rate": 6.9828267189064125e-06,
"loss": 0.52973473072052,
"step": 796
},
{
"epoch": 0.7812768043131969,
"grad_norm": 0.45199069380760193,
"learning_rate": 6.9755413915696105e-06,
"loss": 0.597319483757019,
"step": 797
},
{
"epoch": 0.782257076338684,
"grad_norm": 0.3661534786224365,
"learning_rate": 6.968251090809708e-06,
"loss": 0.4127325713634491,
"step": 798
},
{
"epoch": 0.783237348364171,
"grad_norm": 0.41393721103668213,
"learning_rate": 6.960955834980028e-06,
"loss": 0.4885827302932739,
"step": 799
},
{
"epoch": 0.7842176203896581,
"grad_norm": 0.406561017036438,
"learning_rate": 6.953655642446368e-06,
"loss": 0.5904073715209961,
"step": 800
},
{
"epoch": 0.7851978924151451,
"grad_norm": 0.4422285556793213,
"learning_rate": 6.946350531586959e-06,
"loss": 0.47113174200057983,
"step": 801
},
{
"epoch": 0.7861781644406323,
"grad_norm": 0.40606558322906494,
"learning_rate": 6.9390405207924014e-06,
"loss": 0.536541223526001,
"step": 802
},
{
"epoch": 0.7871584364661194,
"grad_norm": 0.4594433307647705,
"learning_rate": 6.931725628465643e-06,
"loss": 0.47606784105300903,
"step": 803
},
{
"epoch": 0.7881387084916064,
"grad_norm": 0.47338026762008667,
"learning_rate": 6.9244058730219176e-06,
"loss": 0.5939470529556274,
"step": 804
},
{
"epoch": 0.7891189805170935,
"grad_norm": 0.3534870445728302,
"learning_rate": 6.917081272888697e-06,
"loss": 0.3604448437690735,
"step": 805
},
{
"epoch": 0.7900992525425806,
"grad_norm": 0.44684967398643494,
"learning_rate": 6.909751846505657e-06,
"loss": 0.4442647695541382,
"step": 806
},
{
"epoch": 0.7910795245680676,
"grad_norm": 0.4265282154083252,
"learning_rate": 6.902417612324615e-06,
"loss": 0.5449322462081909,
"step": 807
},
{
"epoch": 0.7920597965935547,
"grad_norm": 0.37107881903648376,
"learning_rate": 6.895078588809503e-06,
"loss": 0.48995405435562134,
"step": 808
},
{
"epoch": 0.7930400686190417,
"grad_norm": 0.3696930706501007,
"learning_rate": 6.887734794436301e-06,
"loss": 0.40083059668540955,
"step": 809
},
{
"epoch": 0.7940203406445289,
"grad_norm": 0.4195983111858368,
"learning_rate": 6.880386247692999e-06,
"loss": 0.4219232499599457,
"step": 810
},
{
"epoch": 0.795000612670016,
"grad_norm": 0.3787069022655487,
"learning_rate": 6.873032967079562e-06,
"loss": 0.48830705881118774,
"step": 811
},
{
"epoch": 0.795980884695503,
"grad_norm": 0.4739780128002167,
"learning_rate": 6.865674971107859e-06,
"loss": 0.5899959802627563,
"step": 812
},
{
"epoch": 0.7969611567209901,
"grad_norm": 0.40056106448173523,
"learning_rate": 6.858312278301638e-06,
"loss": 0.5551238059997559,
"step": 813
},
{
"epoch": 0.7979414287464771,
"grad_norm": 0.3908003568649292,
"learning_rate": 6.850944907196472e-06,
"loss": 0.48024556040763855,
"step": 814
},
{
"epoch": 0.7989217007719642,
"grad_norm": 0.3721950948238373,
"learning_rate": 6.8435728763397045e-06,
"loss": 0.454034686088562,
"step": 815
},
{
"epoch": 0.7999019727974513,
"grad_norm": 0.458629846572876,
"learning_rate": 6.836196204290417e-06,
"loss": 0.5407589673995972,
"step": 816
},
{
"epoch": 0.8008822448229383,
"grad_norm": 0.3916063904762268,
"learning_rate": 6.828814909619374e-06,
"loss": 0.5556098222732544,
"step": 817
},
{
"epoch": 0.8018625168484255,
"grad_norm": 0.37506920099258423,
"learning_rate": 6.821429010908972e-06,
"loss": 0.4908115565776825,
"step": 818
},
{
"epoch": 0.8028427888739125,
"grad_norm": 0.4200496971607208,
"learning_rate": 6.814038526753205e-06,
"loss": 0.5301434397697449,
"step": 819
},
{
"epoch": 0.8038230608993996,
"grad_norm": 0.466868132352829,
"learning_rate": 6.806643475757608e-06,
"loss": 0.46871984004974365,
"step": 820
},
{
"epoch": 0.8048033329248867,
"grad_norm": 0.4289228320121765,
"learning_rate": 6.799243876539213e-06,
"loss": 0.5200282335281372,
"step": 821
},
{
"epoch": 0.8057836049503737,
"grad_norm": 0.43029043078422546,
"learning_rate": 6.7918397477265e-06,
"loss": 0.6185582876205444,
"step": 822
},
{
"epoch": 0.8067638769758608,
"grad_norm": 0.4174940288066864,
"learning_rate": 6.78443110795936e-06,
"loss": 0.49757710099220276,
"step": 823
},
{
"epoch": 0.8077441490013478,
"grad_norm": 0.3718063533306122,
"learning_rate": 6.77701797588903e-06,
"loss": 0.48157191276550293,
"step": 824
},
{
"epoch": 0.808724421026835,
"grad_norm": 0.6138043999671936,
"learning_rate": 6.76960037017806e-06,
"loss": 0.44370216131210327,
"step": 825
},
{
"epoch": 0.8097046930523221,
"grad_norm": 1.0954099893569946,
"learning_rate": 6.762178309500267e-06,
"loss": 0.5945616364479065,
"step": 826
},
{
"epoch": 0.8106849650778091,
"grad_norm": 0.4381861388683319,
"learning_rate": 6.75475181254068e-06,
"loss": 0.5567284822463989,
"step": 827
},
{
"epoch": 0.8116652371032962,
"grad_norm": 0.4149746596813202,
"learning_rate": 6.747320897995493e-06,
"loss": 0.5557035207748413,
"step": 828
},
{
"epoch": 0.8126455091287832,
"grad_norm": 0.427548885345459,
"learning_rate": 6.739885584572026e-06,
"loss": 0.48417359590530396,
"step": 829
},
{
"epoch": 0.8136257811542703,
"grad_norm": 0.5222792625427246,
"learning_rate": 6.732445890988675e-06,
"loss": 0.6526938676834106,
"step": 830
},
{
"epoch": 0.8146060531797574,
"grad_norm": 0.5752202272415161,
"learning_rate": 6.725001835974854e-06,
"loss": 0.5249187350273132,
"step": 831
},
{
"epoch": 0.8155863252052444,
"grad_norm": 0.46711474657058716,
"learning_rate": 6.717553438270966e-06,
"loss": 0.452995240688324,
"step": 832
},
{
"epoch": 0.8165665972307316,
"grad_norm": 0.47589167952537537,
"learning_rate": 6.710100716628345e-06,
"loss": 0.532433271408081,
"step": 833
},
{
"epoch": 0.8175468692562186,
"grad_norm": 0.3826428949832916,
"learning_rate": 6.702643689809205e-06,
"loss": 0.40372392535209656,
"step": 834
},
{
"epoch": 0.8185271412817057,
"grad_norm": 0.47821974754333496,
"learning_rate": 6.695182376586603e-06,
"loss": 0.4737083315849304,
"step": 835
},
{
"epoch": 0.8195074133071928,
"grad_norm": 3.390791177749634,
"learning_rate": 6.68771679574439e-06,
"loss": 0.5816224217414856,
"step": 836
},
{
"epoch": 0.8204876853326798,
"grad_norm": 0.4647277593612671,
"learning_rate": 6.680246966077151e-06,
"loss": 0.5226106643676758,
"step": 837
},
{
"epoch": 0.8214679573581669,
"grad_norm": 0.4741794466972351,
"learning_rate": 6.672772906390177e-06,
"loss": 0.4869573414325714,
"step": 838
},
{
"epoch": 0.8224482293836539,
"grad_norm": 0.45704689621925354,
"learning_rate": 6.665294635499404e-06,
"loss": 0.43767136335372925,
"step": 839
},
{
"epoch": 0.823428501409141,
"grad_norm": 0.4436359703540802,
"learning_rate": 6.65781217223137e-06,
"loss": 0.5588634014129639,
"step": 840
},
{
"epoch": 0.8244087734346282,
"grad_norm": 0.4628417193889618,
"learning_rate": 6.650325535423166e-06,
"loss": 0.6288622617721558,
"step": 841
},
{
"epoch": 0.8253890454601152,
"grad_norm": 0.4618736207485199,
"learning_rate": 6.642834743922393e-06,
"loss": 0.6996235847473145,
"step": 842
},
{
"epoch": 0.8263693174856023,
"grad_norm": 0.4331154227256775,
"learning_rate": 6.635339816587109e-06,
"loss": 0.5010419487953186,
"step": 843
},
{
"epoch": 0.8273495895110893,
"grad_norm": 0.48043107986450195,
"learning_rate": 6.627840772285784e-06,
"loss": 0.5202208161354065,
"step": 844
},
{
"epoch": 0.8283298615365764,
"grad_norm": 0.4572037160396576,
"learning_rate": 6.6203376298972535e-06,
"loss": 0.5409632921218872,
"step": 845
},
{
"epoch": 0.8293101335620635,
"grad_norm": 0.5208296775817871,
"learning_rate": 6.612830408310671e-06,
"loss": 0.5930708050727844,
"step": 846
},
{
"epoch": 0.8302904055875505,
"grad_norm": 0.4406164884567261,
"learning_rate": 6.605319126425455e-06,
"loss": 0.5349196195602417,
"step": 847
},
{
"epoch": 0.8312706776130376,
"grad_norm": 0.418474406003952,
"learning_rate": 6.5978038031512505e-06,
"loss": 0.41959160566329956,
"step": 848
},
{
"epoch": 0.8322509496385246,
"grad_norm": 0.5490645170211792,
"learning_rate": 6.590284457407876e-06,
"loss": 0.6460529565811157,
"step": 849
},
{
"epoch": 0.8332312216640118,
"grad_norm": 0.44365769624710083,
"learning_rate": 6.582761108125272e-06,
"loss": 0.46053674817085266,
"step": 850
},
{
"epoch": 0.8342114936894989,
"grad_norm": 0.43043237924575806,
"learning_rate": 6.5752337742434644e-06,
"loss": 0.41906946897506714,
"step": 851
},
{
"epoch": 0.8351917657149859,
"grad_norm": 0.5311118364334106,
"learning_rate": 6.567702474712507e-06,
"loss": 0.5444743037223816,
"step": 852
},
{
"epoch": 0.836172037740473,
"grad_norm": 0.6719686985015869,
"learning_rate": 6.560167228492436e-06,
"loss": 0.488853394985199,
"step": 853
},
{
"epoch": 0.83715230976596,
"grad_norm": 0.41844865679740906,
"learning_rate": 6.552628054553225e-06,
"loss": 0.49421727657318115,
"step": 854
},
{
"epoch": 0.8381325817914471,
"grad_norm": 0.4129052460193634,
"learning_rate": 6.545084971874738e-06,
"loss": 0.5810033082962036,
"step": 855
},
{
"epoch": 0.8391128538169342,
"grad_norm": 0.44456323981285095,
"learning_rate": 6.537537999446673e-06,
"loss": 0.5115293860435486,
"step": 856
},
{
"epoch": 0.8400931258424212,
"grad_norm": 0.4097927212715149,
"learning_rate": 6.529987156268527e-06,
"loss": 0.4882569909095764,
"step": 857
},
{
"epoch": 0.8410733978679084,
"grad_norm": 0.44071686267852783,
"learning_rate": 6.522432461349536e-06,
"loss": 0.5030308961868286,
"step": 858
},
{
"epoch": 0.8420536698933954,
"grad_norm": 0.5421448349952698,
"learning_rate": 6.514873933708637e-06,
"loss": 0.6325263977050781,
"step": 859
},
{
"epoch": 0.8430339419188825,
"grad_norm": 0.5930724143981934,
"learning_rate": 6.507311592374415e-06,
"loss": 0.5336626768112183,
"step": 860
},
{
"epoch": 0.8440142139443696,
"grad_norm": 0.4966267943382263,
"learning_rate": 6.499745456385054e-06,
"loss": 0.5958417654037476,
"step": 861
},
{
"epoch": 0.8449944859698566,
"grad_norm": 0.4763844311237335,
"learning_rate": 6.4921755447882926e-06,
"loss": 0.5004958510398865,
"step": 862
},
{
"epoch": 0.8459747579953437,
"grad_norm": 0.5007293224334717,
"learning_rate": 6.484601876641375e-06,
"loss": 0.5528030395507812,
"step": 863
},
{
"epoch": 0.8469550300208308,
"grad_norm": 0.41702571511268616,
"learning_rate": 6.477024471011001e-06,
"loss": 0.49212419986724854,
"step": 864
},
{
"epoch": 0.8479353020463178,
"grad_norm": 0.4272618889808655,
"learning_rate": 6.469443346973281e-06,
"loss": 0.5188068151473999,
"step": 865
},
{
"epoch": 0.848915574071805,
"grad_norm": 0.5673028230667114,
"learning_rate": 6.461858523613684e-06,
"loss": 0.5160360932350159,
"step": 866
},
{
"epoch": 0.849895846097292,
"grad_norm": 0.4834825396537781,
"learning_rate": 6.454270020026996e-06,
"loss": 0.5802666544914246,
"step": 867
},
{
"epoch": 0.8508761181227791,
"grad_norm": 0.48169195652008057,
"learning_rate": 6.446677855317265e-06,
"loss": 0.4749450087547302,
"step": 868
},
{
"epoch": 0.8518563901482662,
"grad_norm": 0.5354800820350647,
"learning_rate": 6.439082048597755e-06,
"loss": 0.49622076749801636,
"step": 869
},
{
"epoch": 0.8528366621737532,
"grad_norm": 0.42235618829727173,
"learning_rate": 6.431482618990902e-06,
"loss": 0.46224337816238403,
"step": 870
},
{
"epoch": 0.8538169341992403,
"grad_norm": 0.5284849405288696,
"learning_rate": 6.423879585628262e-06,
"loss": 0.6243587136268616,
"step": 871
},
{
"epoch": 0.8547972062247273,
"grad_norm": 0.6278819441795349,
"learning_rate": 6.41627296765046e-06,
"loss": 0.6148986220359802,
"step": 872
},
{
"epoch": 0.8557774782502144,
"grad_norm": 0.45493778586387634,
"learning_rate": 6.408662784207149e-06,
"loss": 0.5227400660514832,
"step": 873
},
{
"epoch": 0.8567577502757016,
"grad_norm": 0.5376487374305725,
"learning_rate": 6.401049054456957e-06,
"loss": 0.6814272403717041,
"step": 874
},
{
"epoch": 0.8577380223011886,
"grad_norm": 0.3643544316291809,
"learning_rate": 6.39343179756744e-06,
"loss": 0.44438761472702026,
"step": 875
},
{
"epoch": 0.8587182943266757,
"grad_norm": 0.4614802598953247,
"learning_rate": 6.385811032715031e-06,
"loss": 0.4618123769760132,
"step": 876
},
{
"epoch": 0.8596985663521627,
"grad_norm": 0.5620284676551819,
"learning_rate": 6.378186779084996e-06,
"loss": 0.5348123908042908,
"step": 877
},
{
"epoch": 0.8606788383776498,
"grad_norm": 0.6737039089202881,
"learning_rate": 6.370559055871389e-06,
"loss": 0.5795971751213074,
"step": 878
},
{
"epoch": 0.8616591104031369,
"grad_norm": 0.46896591782569885,
"learning_rate": 6.362927882276991e-06,
"loss": 0.7106213569641113,
"step": 879
},
{
"epoch": 0.8626393824286239,
"grad_norm": 0.4269348680973053,
"learning_rate": 6.355293277513271e-06,
"loss": 0.4726841449737549,
"step": 880
},
{
"epoch": 0.863619654454111,
"grad_norm": 0.41544944047927856,
"learning_rate": 6.34765526080034e-06,
"loss": 0.4410395920276642,
"step": 881
},
{
"epoch": 0.864599926479598,
"grad_norm": 0.4709013104438782,
"learning_rate": 6.3400138513668955e-06,
"loss": 0.4410967230796814,
"step": 882
},
{
"epoch": 0.8655801985050852,
"grad_norm": 0.4805549681186676,
"learning_rate": 6.332369068450175e-06,
"loss": 0.5561596155166626,
"step": 883
},
{
"epoch": 0.8665604705305723,
"grad_norm": 0.35023778676986694,
"learning_rate": 6.324720931295913e-06,
"loss": 0.3837100863456726,
"step": 884
},
{
"epoch": 0.8675407425560593,
"grad_norm": 0.9413726925849915,
"learning_rate": 6.317069459158284e-06,
"loss": 0.5233207941055298,
"step": 885
},
{
"epoch": 0.8685210145815464,
"grad_norm": 0.43868952989578247,
"learning_rate": 6.3094146712998595e-06,
"loss": 0.5409136414527893,
"step": 886
},
{
"epoch": 0.8695012866070334,
"grad_norm": 0.4314418435096741,
"learning_rate": 6.301756586991561e-06,
"loss": 0.5296670794487,
"step": 887
},
{
"epoch": 0.8704815586325205,
"grad_norm": 0.4177858829498291,
"learning_rate": 6.294095225512604e-06,
"loss": 0.4831322729587555,
"step": 888
},
{
"epoch": 0.8714618306580076,
"grad_norm": 0.4337747395038605,
"learning_rate": 6.286430606150458e-06,
"loss": 0.5630544424057007,
"step": 889
},
{
"epoch": 0.8724421026834946,
"grad_norm": 0.4146502614021301,
"learning_rate": 6.278762748200795e-06,
"loss": 0.4446233808994293,
"step": 890
},
{
"epoch": 0.8734223747089818,
"grad_norm": 0.4662277102470398,
"learning_rate": 6.271091670967437e-06,
"loss": 0.5407575964927673,
"step": 891
},
{
"epoch": 0.8744026467344688,
"grad_norm": 0.48252469301223755,
"learning_rate": 6.2634173937623075e-06,
"loss": 0.46057388186454773,
"step": 892
},
{
"epoch": 0.8753829187599559,
"grad_norm": 0.4287121593952179,
"learning_rate": 6.255739935905396e-06,
"loss": 0.4808887243270874,
"step": 893
},
{
"epoch": 0.876363190785443,
"grad_norm": 0.42432844638824463,
"learning_rate": 6.248059316724693e-06,
"loss": 0.4453814625740051,
"step": 894
},
{
"epoch": 0.87734346281093,
"grad_norm": 0.42549434304237366,
"learning_rate": 6.240375555556145e-06,
"loss": 0.5829179883003235,
"step": 895
},
{
"epoch": 0.8783237348364171,
"grad_norm": 0.4437717795372009,
"learning_rate": 6.232688671743613e-06,
"loss": 0.4659510850906372,
"step": 896
},
{
"epoch": 0.8793040068619041,
"grad_norm": 0.5647913813591003,
"learning_rate": 6.22499868463882e-06,
"loss": 0.5178962349891663,
"step": 897
},
{
"epoch": 0.8802842788873912,
"grad_norm": 0.6277410984039307,
"learning_rate": 6.217305613601296e-06,
"loss": 0.6390823721885681,
"step": 898
},
{
"epoch": 0.8812645509128784,
"grad_norm": 0.43650683760643005,
"learning_rate": 6.209609477998339e-06,
"loss": 0.5823169946670532,
"step": 899
},
{
"epoch": 0.8822448229383654,
"grad_norm": 0.43029022216796875,
"learning_rate": 6.201910297204963e-06,
"loss": 0.4697658121585846,
"step": 900
},
{
"epoch": 0.8832250949638525,
"grad_norm": 0.4142000675201416,
"learning_rate": 6.194208090603845e-06,
"loss": 0.49903035163879395,
"step": 901
},
{
"epoch": 0.8842053669893395,
"grad_norm": 0.990412712097168,
"learning_rate": 6.186502877585281e-06,
"loss": 0.4805365800857544,
"step": 902
},
{
"epoch": 0.8851856390148266,
"grad_norm": 1.582311987876892,
"learning_rate": 6.178794677547138e-06,
"loss": 0.5169810056686401,
"step": 903
},
{
"epoch": 0.8861659110403137,
"grad_norm": 0.4250990152359009,
"learning_rate": 6.171083509894795e-06,
"loss": 0.4501458406448364,
"step": 904
},
{
"epoch": 0.8871461830658007,
"grad_norm": 0.42763155698776245,
"learning_rate": 6.163369394041112e-06,
"loss": 0.4777667820453644,
"step": 905
},
{
"epoch": 0.8881264550912878,
"grad_norm": 0.5017222166061401,
"learning_rate": 6.155652349406366e-06,
"loss": 0.4914003610610962,
"step": 906
},
{
"epoch": 0.8891067271167749,
"grad_norm": 0.44423478841781616,
"learning_rate": 6.1479323954182055e-06,
"loss": 0.48054519295692444,
"step": 907
},
{
"epoch": 0.890086999142262,
"grad_norm": 0.6064445376396179,
"learning_rate": 6.140209551511609e-06,
"loss": 0.5427241921424866,
"step": 908
},
{
"epoch": 0.8910672711677491,
"grad_norm": 0.45574769377708435,
"learning_rate": 6.132483837128823e-06,
"loss": 0.46265387535095215,
"step": 909
},
{
"epoch": 0.8920475431932361,
"grad_norm": 0.5986409187316895,
"learning_rate": 6.124755271719326e-06,
"loss": 0.7392776608467102,
"step": 910
},
{
"epoch": 0.8930278152187232,
"grad_norm": 0.5339890718460083,
"learning_rate": 6.1170238747397715e-06,
"loss": 0.5947340726852417,
"step": 911
},
{
"epoch": 0.8940080872442102,
"grad_norm": 0.5864500403404236,
"learning_rate": 6.109289665653944e-06,
"loss": 0.48198479413986206,
"step": 912
},
{
"epoch": 0.8949883592696973,
"grad_norm": 0.4087126553058624,
"learning_rate": 6.101552663932704e-06,
"loss": 0.516301155090332,
"step": 913
},
{
"epoch": 0.8959686312951844,
"grad_norm": 0.4832591116428375,
"learning_rate": 6.093812889053945e-06,
"loss": 0.5776762962341309,
"step": 914
},
{
"epoch": 0.8969489033206715,
"grad_norm": 0.4292212724685669,
"learning_rate": 6.08607036050254e-06,
"loss": 0.48135313391685486,
"step": 915
},
{
"epoch": 0.8979291753461586,
"grad_norm": 0.4345339834690094,
"learning_rate": 6.078325097770296e-06,
"loss": 0.5092774033546448,
"step": 916
},
{
"epoch": 0.8989094473716457,
"grad_norm": 0.5339592099189758,
"learning_rate": 6.070577120355903e-06,
"loss": 0.6810686588287354,
"step": 917
},
{
"epoch": 0.8998897193971327,
"grad_norm": 0.40882620215415955,
"learning_rate": 6.062826447764883e-06,
"loss": 0.49107372760772705,
"step": 918
},
{
"epoch": 0.9008699914226198,
"grad_norm": 0.475019246339798,
"learning_rate": 6.055073099509549e-06,
"loss": 0.5392287969589233,
"step": 919
},
{
"epoch": 0.9018502634481068,
"grad_norm": 0.45199888944625854,
"learning_rate": 6.047317095108943e-06,
"loss": 0.5883407592773438,
"step": 920
},
{
"epoch": 0.9028305354735939,
"grad_norm": 0.3889693319797516,
"learning_rate": 6.039558454088796e-06,
"loss": 0.40601295232772827,
"step": 921
},
{
"epoch": 0.903810807499081,
"grad_norm": 0.4438765048980713,
"learning_rate": 6.031797195981481e-06,
"loss": 0.5861967206001282,
"step": 922
},
{
"epoch": 0.904791079524568,
"grad_norm": 0.4316740036010742,
"learning_rate": 6.024033340325954e-06,
"loss": 0.46414318680763245,
"step": 923
},
{
"epoch": 0.9057713515500552,
"grad_norm": 0.38185253739356995,
"learning_rate": 6.016266906667711e-06,
"loss": 0.48858797550201416,
"step": 924
},
{
"epoch": 0.9067516235755422,
"grad_norm": 0.41384172439575195,
"learning_rate": 6.0084979145587444e-06,
"loss": 0.4836603105068207,
"step": 925
},
{
"epoch": 0.9077318956010293,
"grad_norm": 0.4198357164859772,
"learning_rate": 6.000726383557476e-06,
"loss": 0.4855630695819855,
"step": 926
},
{
"epoch": 0.9087121676265164,
"grad_norm": 0.4103989899158478,
"learning_rate": 5.9929523332287275e-06,
"loss": 0.4840914011001587,
"step": 927
},
{
"epoch": 0.9096924396520034,
"grad_norm": 0.3962758779525757,
"learning_rate": 5.985175783143666e-06,
"loss": 0.5013476610183716,
"step": 928
},
{
"epoch": 0.9106727116774905,
"grad_norm": 0.4539540410041809,
"learning_rate": 5.977396752879742e-06,
"loss": 0.5926103591918945,
"step": 929
},
{
"epoch": 0.9116529837029775,
"grad_norm": 0.37731069326400757,
"learning_rate": 5.9696152620206575e-06,
"loss": 0.4235096871852875,
"step": 930
},
{
"epoch": 0.9126332557284647,
"grad_norm": 0.4731782078742981,
"learning_rate": 5.961831330156306e-06,
"loss": 0.6326535940170288,
"step": 931
},
{
"epoch": 0.9136135277539518,
"grad_norm": 0.4542580246925354,
"learning_rate": 5.954044976882725e-06,
"loss": 0.558417558670044,
"step": 932
},
{
"epoch": 0.9145937997794388,
"grad_norm": 0.4298331141471863,
"learning_rate": 5.946256221802052e-06,
"loss": 0.5137132406234741,
"step": 933
},
{
"epoch": 0.9155740718049259,
"grad_norm": 0.47129830718040466,
"learning_rate": 5.938465084522468e-06,
"loss": 0.5925625562667847,
"step": 934
},
{
"epoch": 0.9165543438304129,
"grad_norm": 0.42570555210113525,
"learning_rate": 5.930671584658151e-06,
"loss": 0.47181135416030884,
"step": 935
},
{
"epoch": 0.9175346158559,
"grad_norm": 0.4356832802295685,
"learning_rate": 5.922875741829227e-06,
"loss": 0.5676063299179077,
"step": 936
},
{
"epoch": 0.9185148878813871,
"grad_norm": 0.5222892761230469,
"learning_rate": 5.915077575661723e-06,
"loss": 0.6410059928894043,
"step": 937
},
{
"epoch": 0.9194951599068741,
"grad_norm": 0.5957534313201904,
"learning_rate": 5.907277105787513e-06,
"loss": 0.4146493077278137,
"step": 938
},
{
"epoch": 0.9204754319323613,
"grad_norm": 0.37144264578819275,
"learning_rate": 5.89947435184427e-06,
"loss": 0.3888629674911499,
"step": 939
},
{
"epoch": 0.9214557039578483,
"grad_norm": 0.4696313738822937,
"learning_rate": 5.891669333475416e-06,
"loss": 0.6653344631195068,
"step": 940
},
{
"epoch": 0.9224359759833354,
"grad_norm": 0.6026146411895752,
"learning_rate": 5.883862070330079e-06,
"loss": 0.5807779431343079,
"step": 941
},
{
"epoch": 0.9234162480088225,
"grad_norm": 0.41022008657455444,
"learning_rate": 5.876052582063031e-06,
"loss": 0.5313568711280823,
"step": 942
},
{
"epoch": 0.9243965200343095,
"grad_norm": 0.3925977945327759,
"learning_rate": 5.8682408883346535e-06,
"loss": 0.4375201463699341,
"step": 943
},
{
"epoch": 0.9253767920597966,
"grad_norm": 0.9332260489463806,
"learning_rate": 5.860427008810872e-06,
"loss": 0.4921548366546631,
"step": 944
},
{
"epoch": 0.9263570640852836,
"grad_norm": 0.5160321593284607,
"learning_rate": 5.85261096316312e-06,
"loss": 0.5410822033882141,
"step": 945
},
{
"epoch": 0.9273373361107707,
"grad_norm": 0.4472995102405548,
"learning_rate": 5.844792771068284e-06,
"loss": 0.5569736957550049,
"step": 946
},
{
"epoch": 0.9283176081362579,
"grad_norm": 0.34775489568710327,
"learning_rate": 5.8369724522086545e-06,
"loss": 0.48673319816589355,
"step": 947
},
{
"epoch": 0.9292978801617449,
"grad_norm": 0.46031758189201355,
"learning_rate": 5.829150026271871e-06,
"loss": 0.6265984177589417,
"step": 948
},
{
"epoch": 0.930278152187232,
"grad_norm": 0.3856378197669983,
"learning_rate": 5.821325512950886e-06,
"loss": 0.49659961462020874,
"step": 949
},
{
"epoch": 0.931258424212719,
"grad_norm": 0.7397359013557434,
"learning_rate": 5.8134989319439004e-06,
"loss": 0.5615105628967285,
"step": 950
},
{
"epoch": 0.9322386962382061,
"grad_norm": 0.4330701529979706,
"learning_rate": 5.805670302954322e-06,
"loss": 0.5041952133178711,
"step": 951
},
{
"epoch": 0.9332189682636932,
"grad_norm": 0.4708372950553894,
"learning_rate": 5.797839645690715e-06,
"loss": 0.49568355083465576,
"step": 952
},
{
"epoch": 0.9341992402891802,
"grad_norm": 0.4601927399635315,
"learning_rate": 5.79000697986675e-06,
"loss": 0.5010634064674377,
"step": 953
},
{
"epoch": 0.9351795123146673,
"grad_norm": 0.4409112334251404,
"learning_rate": 5.782172325201155e-06,
"loss": 0.49562788009643555,
"step": 954
},
{
"epoch": 0.9361597843401543,
"grad_norm": 0.4923012852668762,
"learning_rate": 5.774335701417662e-06,
"loss": 0.4498739540576935,
"step": 955
},
{
"epoch": 0.9371400563656415,
"grad_norm": 0.40143096446990967,
"learning_rate": 5.766497128244963e-06,
"loss": 0.551247239112854,
"step": 956
},
{
"epoch": 0.9381203283911286,
"grad_norm": 0.4007181227207184,
"learning_rate": 5.758656625416659e-06,
"loss": 0.49534258246421814,
"step": 957
},
{
"epoch": 0.9391006004166156,
"grad_norm": 0.5263123512268066,
"learning_rate": 5.750814212671202e-06,
"loss": 0.5723960399627686,
"step": 958
},
{
"epoch": 0.9400808724421027,
"grad_norm": 0.4463786482810974,
"learning_rate": 5.7429699097518585e-06,
"loss": 0.531633734703064,
"step": 959
},
{
"epoch": 0.9410611444675897,
"grad_norm": 0.43471935391426086,
"learning_rate": 5.735123736406655e-06,
"loss": 0.500960111618042,
"step": 960
},
{
"epoch": 0.9420414164930768,
"grad_norm": 0.4278755784034729,
"learning_rate": 5.727275712388318e-06,
"loss": 0.4983411431312561,
"step": 961
},
{
"epoch": 0.9430216885185639,
"grad_norm": 0.446235328912735,
"learning_rate": 5.719425857454242e-06,
"loss": 0.7012414932250977,
"step": 962
},
{
"epoch": 0.9440019605440509,
"grad_norm": 0.3571961522102356,
"learning_rate": 5.711574191366427e-06,
"loss": 0.4507150650024414,
"step": 963
},
{
"epoch": 0.9449822325695381,
"grad_norm": 0.43753328919410706,
"learning_rate": 5.7037207338914326e-06,
"loss": 0.5032480955123901,
"step": 964
},
{
"epoch": 0.9459625045950251,
"grad_norm": 0.4057547152042389,
"learning_rate": 5.695865504800328e-06,
"loss": 0.5925788879394531,
"step": 965
},
{
"epoch": 0.9469427766205122,
"grad_norm": 0.4289734959602356,
"learning_rate": 5.688008523868646e-06,
"loss": 0.5180743336677551,
"step": 966
},
{
"epoch": 0.9479230486459993,
"grad_norm": 0.45726263523101807,
"learning_rate": 5.680149810876322e-06,
"loss": 0.6067906618118286,
"step": 967
},
{
"epoch": 0.9489033206714863,
"grad_norm": 0.4400206506252289,
"learning_rate": 5.6722893856076596e-06,
"loss": 0.4790126085281372,
"step": 968
},
{
"epoch": 0.9498835926969734,
"grad_norm": 0.4761674702167511,
"learning_rate": 5.664427267851271e-06,
"loss": 0.4958120584487915,
"step": 969
},
{
"epoch": 0.9508638647224604,
"grad_norm": 0.392704039812088,
"learning_rate": 5.656563477400026e-06,
"loss": 0.5374599099159241,
"step": 970
},
{
"epoch": 0.9518441367479475,
"grad_norm": 0.4220191538333893,
"learning_rate": 5.648698034051009e-06,
"loss": 0.5424162149429321,
"step": 971
},
{
"epoch": 0.9528244087734347,
"grad_norm": 0.40179651975631714,
"learning_rate": 5.640830957605466e-06,
"loss": 0.4912230670452118,
"step": 972
},
{
"epoch": 0.9538046807989217,
"grad_norm": 0.379812628030777,
"learning_rate": 5.632962267868747e-06,
"loss": 0.4763560891151428,
"step": 973
},
{
"epoch": 0.9547849528244088,
"grad_norm": 0.447950154542923,
"learning_rate": 5.625091984650273e-06,
"loss": 0.5547003746032715,
"step": 974
},
{
"epoch": 0.9557652248498959,
"grad_norm": 0.3814934492111206,
"learning_rate": 5.617220127763474e-06,
"loss": 0.44596758484840393,
"step": 975
},
{
"epoch": 0.9567454968753829,
"grad_norm": 0.40731292963027954,
"learning_rate": 5.609346717025738e-06,
"loss": 0.4661918878555298,
"step": 976
},
{
"epoch": 0.95772576890087,
"grad_norm": 0.36259421706199646,
"learning_rate": 5.601471772258368e-06,
"loss": 0.5000268220901489,
"step": 977
},
{
"epoch": 0.958706040926357,
"grad_norm": 0.36221638321876526,
"learning_rate": 5.593595313286526e-06,
"loss": 0.40143662691116333,
"step": 978
},
{
"epoch": 0.9596863129518441,
"grad_norm": 0.4417479634284973,
"learning_rate": 5.585717359939192e-06,
"loss": 0.45912468433380127,
"step": 979
},
{
"epoch": 0.9606665849773313,
"grad_norm": 0.3343220353126526,
"learning_rate": 5.5778379320491e-06,
"loss": 0.41304337978363037,
"step": 980
},
{
"epoch": 0.9616468570028183,
"grad_norm": 0.4492647647857666,
"learning_rate": 5.569957049452703e-06,
"loss": 0.6581159830093384,
"step": 981
},
{
"epoch": 0.9626271290283054,
"grad_norm": 0.43616801500320435,
"learning_rate": 5.562074731990115e-06,
"loss": 0.4837437868118286,
"step": 982
},
{
"epoch": 0.9636074010537924,
"grad_norm": 0.3660007417201996,
"learning_rate": 5.5541909995050554e-06,
"loss": 0.45983701944351196,
"step": 983
},
{
"epoch": 0.9645876730792795,
"grad_norm": 0.3804008364677429,
"learning_rate": 5.546305871844816e-06,
"loss": 0.38527461886405945,
"step": 984
},
{
"epoch": 0.9655679451047666,
"grad_norm": 0.3764747083187103,
"learning_rate": 5.538419368860196e-06,
"loss": 0.43098461627960205,
"step": 985
},
{
"epoch": 0.9665482171302536,
"grad_norm": 0.38172152638435364,
"learning_rate": 5.530531510405454e-06,
"loss": 0.43058720231056213,
"step": 986
},
{
"epoch": 0.9675284891557407,
"grad_norm": 0.4629911482334137,
"learning_rate": 5.522642316338268e-06,
"loss": 0.4460248351097107,
"step": 987
},
{
"epoch": 0.9685087611812278,
"grad_norm": 0.4303341507911682,
"learning_rate": 5.514751806519673e-06,
"loss": 0.5343886613845825,
"step": 988
},
{
"epoch": 0.9694890332067149,
"grad_norm": 0.3839262127876282,
"learning_rate": 5.506860000814017e-06,
"loss": 0.4435023367404938,
"step": 989
},
{
"epoch": 0.970469305232202,
"grad_norm": 0.3626423478126526,
"learning_rate": 5.498966919088914e-06,
"loss": 0.5118464827537537,
"step": 990
},
{
"epoch": 0.971449577257689,
"grad_norm": 0.37724119424819946,
"learning_rate": 5.491072581215186e-06,
"loss": 0.5193969011306763,
"step": 991
},
{
"epoch": 0.9724298492831761,
"grad_norm": 0.3465311825275421,
"learning_rate": 5.48317700706682e-06,
"loss": 0.4697530269622803,
"step": 992
},
{
"epoch": 0.9734101213086631,
"grad_norm": 0.3911851644515991,
"learning_rate": 5.475280216520913e-06,
"loss": 0.44076526165008545,
"step": 993
},
{
"epoch": 0.9743903933341502,
"grad_norm": 0.32924512028694153,
"learning_rate": 5.467382229457629e-06,
"loss": 0.48638254404067993,
"step": 994
},
{
"epoch": 0.9753706653596373,
"grad_norm": 0.343056857585907,
"learning_rate": 5.459483065760138e-06,
"loss": 0.48689454793930054,
"step": 995
},
{
"epoch": 0.9763509373851244,
"grad_norm": 0.3778912425041199,
"learning_rate": 5.451582745314576e-06,
"loss": 0.5066269636154175,
"step": 996
},
{
"epoch": 0.9773312094106115,
"grad_norm": 0.4240163266658783,
"learning_rate": 5.443681288009991e-06,
"loss": 0.5170713067054749,
"step": 997
},
{
"epoch": 0.9783114814360985,
"grad_norm": 0.42373648285865784,
"learning_rate": 5.435778713738292e-06,
"loss": 0.4839274287223816,
"step": 998
},
{
"epoch": 0.9792917534615856,
"grad_norm": 0.4663996398448944,
"learning_rate": 5.4278750423942e-06,
"loss": 0.5407215356826782,
"step": 999
},
{
"epoch": 0.9802720254870727,
"grad_norm": 0.35067281126976013,
"learning_rate": 5.419970293875197e-06,
"loss": 0.42240002751350403,
"step": 1000
},
{
"epoch": 0.9812522975125597,
"grad_norm": 0.3952957093715668,
"learning_rate": 5.412064488081482e-06,
"loss": 0.4958457350730896,
"step": 1001
},
{
"epoch": 0.9822325695380468,
"grad_norm": 0.3625335097312927,
"learning_rate": 5.404157644915907e-06,
"loss": 0.4905568063259125,
"step": 1002
},
{
"epoch": 0.9832128415635338,
"grad_norm": 0.4454256594181061,
"learning_rate": 5.396249784283943e-06,
"loss": 0.46749985218048096,
"step": 1003
},
{
"epoch": 0.984193113589021,
"grad_norm": 0.37890392541885376,
"learning_rate": 5.388340926093619e-06,
"loss": 0.4817125201225281,
"step": 1004
},
{
"epoch": 0.9851733856145081,
"grad_norm": 0.40377727150917053,
"learning_rate": 5.380431090255475e-06,
"loss": 0.46670079231262207,
"step": 1005
},
{
"epoch": 0.9861536576399951,
"grad_norm": 0.4605295658111572,
"learning_rate": 5.372520296682516e-06,
"loss": 0.626600980758667,
"step": 1006
},
{
"epoch": 0.9871339296654822,
"grad_norm": 0.37546271085739136,
"learning_rate": 5.364608565290154e-06,
"loss": 0.43172627687454224,
"step": 1007
},
{
"epoch": 0.9881142016909692,
"grad_norm": 0.39535439014434814,
"learning_rate": 5.356695915996162e-06,
"loss": 0.4631704092025757,
"step": 1008
},
{
"epoch": 0.9890944737164563,
"grad_norm": 0.3779025673866272,
"learning_rate": 5.348782368720627e-06,
"loss": 0.5013574361801147,
"step": 1009
},
{
"epoch": 0.9900747457419434,
"grad_norm": 0.37166279554367065,
"learning_rate": 5.340867943385895e-06,
"loss": 0.4716484546661377,
"step": 1010
},
{
"epoch": 0.9910550177674304,
"grad_norm": 0.4055781066417694,
"learning_rate": 5.33295265991652e-06,
"loss": 0.5293822288513184,
"step": 1011
},
{
"epoch": 0.9920352897929176,
"grad_norm": 0.3550724387168884,
"learning_rate": 5.325036538239221e-06,
"loss": 0.45156705379486084,
"step": 1012
},
{
"epoch": 0.9930155618184046,
"grad_norm": 0.3812902867794037,
"learning_rate": 5.317119598282823e-06,
"loss": 0.4375635087490082,
"step": 1013
},
{
"epoch": 0.9939958338438917,
"grad_norm": 0.447942852973938,
"learning_rate": 5.3092018599782155e-06,
"loss": 0.542604923248291,
"step": 1014
},
{
"epoch": 0.9949761058693788,
"grad_norm": 0.44398513436317444,
"learning_rate": 5.301283343258293e-06,
"loss": 0.547914981842041,
"step": 1015
},
{
"epoch": 0.9959563778948658,
"grad_norm": 0.42243313789367676,
"learning_rate": 5.293364068057912e-06,
"loss": 0.5050883293151855,
"step": 1016
},
{
"epoch": 0.9969366499203529,
"grad_norm": 0.34492605924606323,
"learning_rate": 5.285444054313841e-06,
"loss": 0.36160433292388916,
"step": 1017
},
{
"epoch": 0.9979169219458399,
"grad_norm": 0.4846872091293335,
"learning_rate": 5.277523321964701e-06,
"loss": 0.5536131262779236,
"step": 1018
},
{
"epoch": 0.998897193971327,
"grad_norm": 0.34226441383361816,
"learning_rate": 5.26960189095093e-06,
"loss": 0.4014680087566376,
"step": 1019
},
{
"epoch": 0.9998774659968142,
"grad_norm": 0.3548482358455658,
"learning_rate": 5.2616797812147205e-06,
"loss": 0.4133622348308563,
"step": 1020
},
{
"epoch": 1.0,
"grad_norm": 1.1780668497085571,
"learning_rate": 5.253757012699972e-06,
"loss": 0.6116985082626343,
"step": 1021
}
],
"logging_steps": 1,
"max_steps": 2042,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.0095509197270548e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}