spatialstack_ablation_six_r2r / trainer_state.json
phamquandung's picture
Upload folder using huggingface_hub
9c9ee66 verified
Raw History Blame Contribute Delete
185 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 9864,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.001013864598382886,
"grad_norm": 64.1862564086914,
"learning_rate": 6.081081081081082e-07,
"loss": 1.5564678192138672,
"step": 10
},
{
"epoch": 0.002027729196765772,
"grad_norm": 21.16122055053711,
"learning_rate": 1.2837837837837838e-06,
"loss": 0.28309895992279055,
"step": 20
},
{
"epoch": 0.003041593795148658,
"grad_norm": 15.708589553833008,
"learning_rate": 1.9594594594594595e-06,
"loss": 0.2308506965637207,
"step": 30
},
{
"epoch": 0.004055458393531544,
"grad_norm": 19.391740798950195,
"learning_rate": 2.6351351351351353e-06,
"loss": 0.2254863977432251,
"step": 40
},
{
"epoch": 0.00506932299191443,
"grad_norm": 13.775400161743164,
"learning_rate": 3.310810810810811e-06,
"loss": 0.22740693092346193,
"step": 50
},
{
"epoch": 0.006083187590297316,
"grad_norm": 11.067628860473633,
"learning_rate": 3.986486486486487e-06,
"loss": 0.2431173801422119,
"step": 60
},
{
"epoch": 0.007097052188680202,
"grad_norm": 7.5858941078186035,
"learning_rate": 4.6621621621621625e-06,
"loss": 0.2154371738433838,
"step": 70
},
{
"epoch": 0.008110916787063088,
"grad_norm": 10.065157890319824,
"learning_rate": 5.337837837837838e-06,
"loss": 0.23916070461273192,
"step": 80
},
{
"epoch": 0.009124781385445974,
"grad_norm": 6.396965503692627,
"learning_rate": 6.013513513513514e-06,
"loss": 0.2267691135406494,
"step": 90
},
{
"epoch": 0.01013864598382886,
"grad_norm": 6.624703407287598,
"learning_rate": 6.689189189189191e-06,
"loss": 0.2104668140411377,
"step": 100
},
{
"epoch": 0.011152510582211745,
"grad_norm": 8.911632537841797,
"learning_rate": 7.3648648648648655e-06,
"loss": 0.24289608001708984,
"step": 110
},
{
"epoch": 0.012166375180594631,
"grad_norm": 12.263571739196777,
"learning_rate": 8.040540540540541e-06,
"loss": 0.22325725555419923,
"step": 120
},
{
"epoch": 0.013180239778977518,
"grad_norm": 7.306334972381592,
"learning_rate": 8.716216216216217e-06,
"loss": 0.22349882125854492,
"step": 130
},
{
"epoch": 0.014194104377360404,
"grad_norm": 8.615752220153809,
"learning_rate": 9.391891891891893e-06,
"loss": 0.22540502548217772,
"step": 140
},
{
"epoch": 0.015207968975743289,
"grad_norm": 13.688424110412598,
"learning_rate": 1.0067567567567569e-05,
"loss": 0.21731691360473632,
"step": 150
},
{
"epoch": 0.016221833574126175,
"grad_norm": 6.7506022453308105,
"learning_rate": 1.0743243243243244e-05,
"loss": 0.1916736841201782,
"step": 160
},
{
"epoch": 0.01723569817250906,
"grad_norm": 11.37317180633545,
"learning_rate": 1.141891891891892e-05,
"loss": 0.20228369235992433,
"step": 170
},
{
"epoch": 0.018249562770891948,
"grad_norm": 8.376798629760742,
"learning_rate": 1.2094594594594596e-05,
"loss": 0.19613726139068605,
"step": 180
},
{
"epoch": 0.019263427369274833,
"grad_norm": 9.712519645690918,
"learning_rate": 1.2770270270270272e-05,
"loss": 0.1972317099571228,
"step": 190
},
{
"epoch": 0.02027729196765772,
"grad_norm": 15.163095474243164,
"learning_rate": 1.3445945945945946e-05,
"loss": 0.23850913047790528,
"step": 200
},
{
"epoch": 0.021291156566040605,
"grad_norm": 6.068497180938721,
"learning_rate": 1.4121621621621621e-05,
"loss": 0.22383947372436525,
"step": 210
},
{
"epoch": 0.02230502116442349,
"grad_norm": 11.978721618652344,
"learning_rate": 1.4797297297297299e-05,
"loss": 0.20344579219818115,
"step": 220
},
{
"epoch": 0.023318885762806378,
"grad_norm": 6.297719955444336,
"learning_rate": 1.5472972972972976e-05,
"loss": 0.2035433292388916,
"step": 230
},
{
"epoch": 0.024332750361189263,
"grad_norm": 3.4323012828826904,
"learning_rate": 1.614864864864865e-05,
"loss": 0.20252950191497804,
"step": 240
},
{
"epoch": 0.02534661495957215,
"grad_norm": 5.752898216247559,
"learning_rate": 1.6824324324324324e-05,
"loss": 0.20974197387695312,
"step": 250
},
{
"epoch": 0.026360479557955036,
"grad_norm": 9.032163619995117,
"learning_rate": 1.7500000000000002e-05,
"loss": 0.20885136127471923,
"step": 260
},
{
"epoch": 0.02737434415633792,
"grad_norm": 7.417436122894287,
"learning_rate": 1.8175675675675676e-05,
"loss": 0.20603437423706056,
"step": 270
},
{
"epoch": 0.028388208754720808,
"grad_norm": 10.366427421569824,
"learning_rate": 1.8851351351351353e-05,
"loss": 0.2202134370803833,
"step": 280
},
{
"epoch": 0.029402073353103693,
"grad_norm": 5.5174560546875,
"learning_rate": 1.9527027027027027e-05,
"loss": 0.21238203048706056,
"step": 290
},
{
"epoch": 0.030415937951486578,
"grad_norm": 3.0442655086517334,
"learning_rate": 1.999999514856865e-05,
"loss": 0.18797210454940796,
"step": 300
},
{
"epoch": 0.03142980254986946,
"grad_norm": 4.141722679138184,
"learning_rate": 1.999990890103106e-05,
"loss": 0.1908482789993286,
"step": 310
},
{
"epoch": 0.03244366714825235,
"grad_norm": 4.003016948699951,
"learning_rate": 1.999971484497808e-05,
"loss": 0.2004786968231201,
"step": 320
},
{
"epoch": 0.03345753174663524,
"grad_norm": 2.9930176734924316,
"learning_rate": 1.9999412982501806e-05,
"loss": 0.184601092338562,
"step": 330
},
{
"epoch": 0.03447139634501812,
"grad_norm": 6.046216011047363,
"learning_rate": 1.9999003316856605e-05,
"loss": 0.18740603923797608,
"step": 340
},
{
"epoch": 0.03548526094340101,
"grad_norm": 5.595577716827393,
"learning_rate": 1.9998485852459068e-05,
"loss": 0.18120222091674804,
"step": 350
},
{
"epoch": 0.036499125541783896,
"grad_norm": 4.445819854736328,
"learning_rate": 1.9997860594887947e-05,
"loss": 0.19796674251556395,
"step": 360
},
{
"epoch": 0.037512990140166784,
"grad_norm": 10.133282661437988,
"learning_rate": 1.9997127550884116e-05,
"loss": 0.20245726108551027,
"step": 370
},
{
"epoch": 0.038526854738549665,
"grad_norm": 4.0377631187438965,
"learning_rate": 1.9996286728350484e-05,
"loss": 0.19745326042175293,
"step": 380
},
{
"epoch": 0.03954071933693255,
"grad_norm": 5.882261753082275,
"learning_rate": 1.9995338136351913e-05,
"loss": 0.20224757194519044,
"step": 390
},
{
"epoch": 0.04055458393531544,
"grad_norm": 7.0593695640563965,
"learning_rate": 1.9994281785115127e-05,
"loss": 0.19402302503585817,
"step": 400
},
{
"epoch": 0.04156844853369832,
"grad_norm": 4.461320400238037,
"learning_rate": 1.9993117686028603e-05,
"loss": 0.18518316745758057,
"step": 410
},
{
"epoch": 0.04258231313208121,
"grad_norm": 9.245733261108398,
"learning_rate": 1.9991845851642426e-05,
"loss": 0.21153395175933837,
"step": 420
},
{
"epoch": 0.0435961777304641,
"grad_norm": 5.25214958190918,
"learning_rate": 1.999046629566818e-05,
"loss": 0.19497358798980713,
"step": 430
},
{
"epoch": 0.04461004232884698,
"grad_norm": 4.487993240356445,
"learning_rate": 1.998897903297879e-05,
"loss": 0.20348334312438965,
"step": 440
},
{
"epoch": 0.04562390692722987,
"grad_norm": 6.769367694854736,
"learning_rate": 1.9987384079608362e-05,
"loss": 0.18774226903915406,
"step": 450
},
{
"epoch": 0.046637771525612756,
"grad_norm": 5.828639984130859,
"learning_rate": 1.998568145275201e-05,
"loss": 0.20705549716949462,
"step": 460
},
{
"epoch": 0.04765163612399564,
"grad_norm": 4.169932842254639,
"learning_rate": 1.998387117076566e-05,
"loss": 0.20561468601226807,
"step": 470
},
{
"epoch": 0.048665500722378525,
"grad_norm": 4.608080863952637,
"learning_rate": 1.998195325316588e-05,
"loss": 0.19333763122558595,
"step": 480
},
{
"epoch": 0.049679365320761414,
"grad_norm": 5.069401741027832,
"learning_rate": 1.997992772062964e-05,
"loss": 0.212402868270874,
"step": 490
},
{
"epoch": 0.0506932299191443,
"grad_norm": 8.477572441101074,
"learning_rate": 1.9977794594994103e-05,
"loss": 0.20473792552947997,
"step": 500
},
{
"epoch": 0.05170709451752718,
"grad_norm": 2.4963345527648926,
"learning_rate": 1.9975553899256388e-05,
"loss": 0.19357246160507202,
"step": 510
},
{
"epoch": 0.05272095911591007,
"grad_norm": 3.7899420261383057,
"learning_rate": 1.9973205657573328e-05,
"loss": 0.18839743137359619,
"step": 520
},
{
"epoch": 0.05373482371429296,
"grad_norm": 6.582728385925293,
"learning_rate": 1.99707498952612e-05,
"loss": 0.18427114486694335,
"step": 530
},
{
"epoch": 0.05474868831267584,
"grad_norm": 4.272158622741699,
"learning_rate": 1.9968186638795445e-05,
"loss": 0.17711787223815917,
"step": 540
},
{
"epoch": 0.05576255291105873,
"grad_norm": 4.634450912475586,
"learning_rate": 1.9965515915810408e-05,
"loss": 0.17505826950073242,
"step": 550
},
{
"epoch": 0.056776417509441617,
"grad_norm": 5.089344024658203,
"learning_rate": 1.9962737755099023e-05,
"loss": 0.19761476516723633,
"step": 560
},
{
"epoch": 0.0577902821078245,
"grad_norm": 5.173961162567139,
"learning_rate": 1.9959852186612492e-05,
"loss": 0.19366734027862548,
"step": 570
},
{
"epoch": 0.058804146706207386,
"grad_norm": 3.9786109924316406,
"learning_rate": 1.995685924145999e-05,
"loss": 0.19307973384857177,
"step": 580
},
{
"epoch": 0.059818011304590274,
"grad_norm": 7.095676422119141,
"learning_rate": 1.9953758951908305e-05,
"loss": 0.1713142991065979,
"step": 590
},
{
"epoch": 0.060831875902973155,
"grad_norm": 4.208062648773193,
"learning_rate": 1.995055135138151e-05,
"loss": 0.19532662630081177,
"step": 600
},
{
"epoch": 0.06184574050135604,
"grad_norm": 5.694543361663818,
"learning_rate": 1.994723647446058e-05,
"loss": 0.20049443244934081,
"step": 610
},
{
"epoch": 0.06285960509973892,
"grad_norm": 3.246277093887329,
"learning_rate": 1.994381435688303e-05,
"loss": 0.18928232192993164,
"step": 620
},
{
"epoch": 0.06387346969812181,
"grad_norm": 3.2543137073516846,
"learning_rate": 1.994028503554255e-05,
"loss": 0.1825331687927246,
"step": 630
},
{
"epoch": 0.0648873342965047,
"grad_norm": 4.81928825378418,
"learning_rate": 1.9936648548488568e-05,
"loss": 0.19059185981750487,
"step": 640
},
{
"epoch": 0.06590119889488759,
"grad_norm": 7.935774803161621,
"learning_rate": 1.993290493492587e-05,
"loss": 0.18312711715698243,
"step": 650
},
{
"epoch": 0.06691506349327048,
"grad_norm": 7.596114158630371,
"learning_rate": 1.9929054235214168e-05,
"loss": 0.17911872863769532,
"step": 660
},
{
"epoch": 0.06792892809165336,
"grad_norm": 6.17520809173584,
"learning_rate": 1.992509649086766e-05,
"loss": 0.15317161083221437,
"step": 670
},
{
"epoch": 0.06894279269003624,
"grad_norm": 4.521999835968018,
"learning_rate": 1.9921031744554588e-05,
"loss": 0.21090376377105713,
"step": 680
},
{
"epoch": 0.06995665728841913,
"grad_norm": 6.409340858459473,
"learning_rate": 1.991686004009678e-05,
"loss": 0.19363487958908082,
"step": 690
},
{
"epoch": 0.07097052188680202,
"grad_norm": 4.069365501403809,
"learning_rate": 1.991258142246917e-05,
"loss": 0.19461148977279663,
"step": 700
},
{
"epoch": 0.0719843864851849,
"grad_norm": 5.6875128746032715,
"learning_rate": 1.990819593779932e-05,
"loss": 0.18929736614227294,
"step": 710
},
{
"epoch": 0.07299825108356779,
"grad_norm": 5.760079383850098,
"learning_rate": 1.9903703633366923e-05,
"loss": 0.18679184913635255,
"step": 720
},
{
"epoch": 0.07401211568195068,
"grad_norm": 8.681827545166016,
"learning_rate": 1.9899104557603285e-05,
"loss": 0.17875468730926514,
"step": 730
},
{
"epoch": 0.07502598028033357,
"grad_norm": 7.710123538970947,
"learning_rate": 1.9894398760090807e-05,
"loss": 0.2129199743270874,
"step": 740
},
{
"epoch": 0.07603984487871644,
"grad_norm": 5.52160120010376,
"learning_rate": 1.988958629156246e-05,
"loss": 0.19008680582046508,
"step": 750
},
{
"epoch": 0.07705370947709933,
"grad_norm": 2.225893259048462,
"learning_rate": 1.988466720390123e-05,
"loss": 0.1883502721786499,
"step": 760
},
{
"epoch": 0.07806757407548222,
"grad_norm": 4.806005954742432,
"learning_rate": 1.9879641550139545e-05,
"loss": 0.1890020489692688,
"step": 770
},
{
"epoch": 0.0790814386738651,
"grad_norm": 6.518285274505615,
"learning_rate": 1.9874509384458726e-05,
"loss": 0.18865787982940674,
"step": 780
},
{
"epoch": 0.080095303272248,
"grad_norm": 3.887723207473755,
"learning_rate": 1.9869270762188402e-05,
"loss": 0.18830269575119019,
"step": 790
},
{
"epoch": 0.08110916787063088,
"grad_norm": 6.556276798248291,
"learning_rate": 1.986392573980589e-05,
"loss": 0.1962014079093933,
"step": 800
},
{
"epoch": 0.08212303246901376,
"grad_norm": 3.9959211349487305,
"learning_rate": 1.9858474374935606e-05,
"loss": 0.18272030353546143,
"step": 810
},
{
"epoch": 0.08313689706739665,
"grad_norm": 3.147815227508545,
"learning_rate": 1.9852916726348446e-05,
"loss": 0.18407528400421141,
"step": 820
},
{
"epoch": 0.08415076166577953,
"grad_norm": 2.278918504714966,
"learning_rate": 1.9847252853961136e-05,
"loss": 0.17978769540786743,
"step": 830
},
{
"epoch": 0.08516462626416242,
"grad_norm": 5.7883758544921875,
"learning_rate": 1.98414828188356e-05,
"loss": 0.1879422664642334,
"step": 840
},
{
"epoch": 0.08617849086254531,
"grad_norm": 3.611768960952759,
"learning_rate": 1.98356066831783e-05,
"loss": 0.1870585322380066,
"step": 850
},
{
"epoch": 0.0871923554609282,
"grad_norm": 3.820913791656494,
"learning_rate": 1.982962451033956e-05,
"loss": 0.17921537160873413,
"step": 860
},
{
"epoch": 0.08820622005931109,
"grad_norm": 4.198739528656006,
"learning_rate": 1.9823536364812875e-05,
"loss": 0.17566168308258057,
"step": 870
},
{
"epoch": 0.08922008465769396,
"grad_norm": 6.719704627990723,
"learning_rate": 1.981734231223425e-05,
"loss": 0.16821424961090087,
"step": 880
},
{
"epoch": 0.09023394925607685,
"grad_norm": 7.390140533447266,
"learning_rate": 1.9811042419381444e-05,
"loss": 0.19110488891601562,
"step": 890
},
{
"epoch": 0.09124781385445974,
"grad_norm": 5.113621234893799,
"learning_rate": 1.980463675417329e-05,
"loss": 0.18234962224960327,
"step": 900
},
{
"epoch": 0.09226167845284262,
"grad_norm": 5.151636600494385,
"learning_rate": 1.9798125385668942e-05,
"loss": 0.19235289096832275,
"step": 910
},
{
"epoch": 0.09327554305122551,
"grad_norm": 2.4375522136688232,
"learning_rate": 1.9791508384067138e-05,
"loss": 0.18086470365524293,
"step": 920
},
{
"epoch": 0.0942894076496084,
"grad_norm": 3.8809242248535156,
"learning_rate": 1.9784785820705435e-05,
"loss": 0.19065535068511963,
"step": 930
},
{
"epoch": 0.09530327224799127,
"grad_norm": 5.916265964508057,
"learning_rate": 1.9777957768059456e-05,
"loss": 0.18433548212051393,
"step": 940
},
{
"epoch": 0.09631713684637416,
"grad_norm": 8.576583862304688,
"learning_rate": 1.977102429974209e-05,
"loss": 0.18346821069717406,
"step": 950
},
{
"epoch": 0.09733100144475705,
"grad_norm": 5.242020130157471,
"learning_rate": 1.9763985490502714e-05,
"loss": 0.1639503002166748,
"step": 960
},
{
"epoch": 0.09834486604313994,
"grad_norm": 5.2149810791015625,
"learning_rate": 1.9756841416226372e-05,
"loss": 0.19034528732299805,
"step": 970
},
{
"epoch": 0.09935873064152283,
"grad_norm": 5.322317123413086,
"learning_rate": 1.9749592153932974e-05,
"loss": 0.18644479513168336,
"step": 980
},
{
"epoch": 0.10037259523990572,
"grad_norm": 6.046846389770508,
"learning_rate": 1.9742237781776455e-05,
"loss": 0.1733173131942749,
"step": 990
},
{
"epoch": 0.1013864598382886,
"grad_norm": 3.5481932163238525,
"learning_rate": 1.9734778379043923e-05,
"loss": 0.19311517477035522,
"step": 1000
},
{
"epoch": 0.10240032443667148,
"grad_norm": 3.933460235595703,
"learning_rate": 1.9727214026154827e-05,
"loss": 0.1712397336959839,
"step": 1010
},
{
"epoch": 0.10341418903505437,
"grad_norm": 4.724759578704834,
"learning_rate": 1.9719544804660084e-05,
"loss": 0.18735097646713256,
"step": 1020
},
{
"epoch": 0.10442805363343725,
"grad_norm": 8.169912338256836,
"learning_rate": 1.9711770797241174e-05,
"loss": 0.18742027282714843,
"step": 1030
},
{
"epoch": 0.10544191823182014,
"grad_norm": 3.8502867221832275,
"learning_rate": 1.9703892087709287e-05,
"loss": 0.16548259258270265,
"step": 1040
},
{
"epoch": 0.10645578283020303,
"grad_norm": 2.369086265563965,
"learning_rate": 1.9695908761004392e-05,
"loss": 0.169579815864563,
"step": 1050
},
{
"epoch": 0.10746964742858592,
"grad_norm": 2.9181628227233887,
"learning_rate": 1.9687820903194335e-05,
"loss": 0.15650696754455568,
"step": 1060
},
{
"epoch": 0.10848351202696879,
"grad_norm": 3.5593113899230957,
"learning_rate": 1.9679628601473905e-05,
"loss": 0.20063579082489014,
"step": 1070
},
{
"epoch": 0.10949737662535168,
"grad_norm": 6.730126857757568,
"learning_rate": 1.967133194416389e-05,
"loss": 0.18477944135665894,
"step": 1080
},
{
"epoch": 0.11051124122373457,
"grad_norm": 5.1826701164245605,
"learning_rate": 1.9662931020710138e-05,
"loss": 0.19267619848251344,
"step": 1090
},
{
"epoch": 0.11152510582211746,
"grad_norm": 2.336012601852417,
"learning_rate": 1.965442592168259e-05,
"loss": 0.20757174491882324,
"step": 1100
},
{
"epoch": 0.11253897042050034,
"grad_norm": 4.37786340713501,
"learning_rate": 1.964581673877428e-05,
"loss": 0.16720629930496217,
"step": 1110
},
{
"epoch": 0.11355283501888323,
"grad_norm": 2.280975103378296,
"learning_rate": 1.9637103564800384e-05,
"loss": 0.17222402095794678,
"step": 1120
},
{
"epoch": 0.11456669961726611,
"grad_norm": 3.606830358505249,
"learning_rate": 1.962828649369719e-05,
"loss": 0.17903528213500977,
"step": 1130
},
{
"epoch": 0.115580564215649,
"grad_norm": 4.617621421813965,
"learning_rate": 1.96193656205211e-05,
"loss": 0.17552559375762938,
"step": 1140
},
{
"epoch": 0.11659442881403188,
"grad_norm": 1.8212428092956543,
"learning_rate": 1.9610341041447595e-05,
"loss": 0.17414417266845703,
"step": 1150
},
{
"epoch": 0.11760829341241477,
"grad_norm": 2.924854040145874,
"learning_rate": 1.9601212853770214e-05,
"loss": 0.16759941577911378,
"step": 1160
},
{
"epoch": 0.11862215801079766,
"grad_norm": 7.883354187011719,
"learning_rate": 1.959198115589949e-05,
"loss": 0.19685659408569336,
"step": 1170
},
{
"epoch": 0.11963602260918055,
"grad_norm": 2.97690749168396,
"learning_rate": 1.9582646047361888e-05,
"loss": 0.16824111938476563,
"step": 1180
},
{
"epoch": 0.12064988720756344,
"grad_norm": 2.1153457164764404,
"learning_rate": 1.9573207628798747e-05,
"loss": 0.16947457790374756,
"step": 1190
},
{
"epoch": 0.12166375180594631,
"grad_norm": 2.074368953704834,
"learning_rate": 1.956366600196519e-05,
"loss": 0.1948012351989746,
"step": 1200
},
{
"epoch": 0.1226776164043292,
"grad_norm": 2.0262203216552734,
"learning_rate": 1.9554021269729003e-05,
"loss": 0.1709741473197937,
"step": 1210
},
{
"epoch": 0.12369148100271209,
"grad_norm": 4.3050007820129395,
"learning_rate": 1.9544273536069573e-05,
"loss": 0.16978495121002196,
"step": 1220
},
{
"epoch": 0.12470534560109497,
"grad_norm": 3.848381280899048,
"learning_rate": 1.9534422906076714e-05,
"loss": 0.19256162643432617,
"step": 1230
},
{
"epoch": 0.12571921019947785,
"grad_norm": 1.850915551185608,
"learning_rate": 1.9524469485949586e-05,
"loss": 0.1848185658454895,
"step": 1240
},
{
"epoch": 0.12673307479786075,
"grad_norm": 4.7735371589660645,
"learning_rate": 1.9514413382995498e-05,
"loss": 0.19104444980621338,
"step": 1250
},
{
"epoch": 0.12774693939624363,
"grad_norm": 5.078706741333008,
"learning_rate": 1.9504254705628806e-05,
"loss": 0.17947646379470825,
"step": 1260
},
{
"epoch": 0.12876080399462653,
"grad_norm": 2.023322820663452,
"learning_rate": 1.9493993563369688e-05,
"loss": 0.19430524110794067,
"step": 1270
},
{
"epoch": 0.1297746685930094,
"grad_norm": 5.031981945037842,
"learning_rate": 1.9483630066843005e-05,
"loss": 0.17340612411499023,
"step": 1280
},
{
"epoch": 0.1307885331913923,
"grad_norm": 2.778304100036621,
"learning_rate": 1.9473164327777095e-05,
"loss": 0.16071449518203734,
"step": 1290
},
{
"epoch": 0.13180239778977518,
"grad_norm": 7.192217826843262,
"learning_rate": 1.9462596459002565e-05,
"loss": 0.18230323791503905,
"step": 1300
},
{
"epoch": 0.13281626238815805,
"grad_norm": 1.8421424627304077,
"learning_rate": 1.9451926574451074e-05,
"loss": 0.1682962656021118,
"step": 1310
},
{
"epoch": 0.13383012698654095,
"grad_norm": 3.8791580200195312,
"learning_rate": 1.944115478915412e-05,
"loss": 0.1764032244682312,
"step": 1320
},
{
"epoch": 0.13484399158492383,
"grad_norm": 4.390422344207764,
"learning_rate": 1.9430281219241767e-05,
"loss": 0.17906363010406495,
"step": 1330
},
{
"epoch": 0.13585785618330673,
"grad_norm": 5.299075603485107,
"learning_rate": 1.9419305981941436e-05,
"loss": 0.18252334594726563,
"step": 1340
},
{
"epoch": 0.1368717207816896,
"grad_norm": 4.197329044342041,
"learning_rate": 1.94082291955766e-05,
"loss": 0.17038908004760742,
"step": 1350
},
{
"epoch": 0.13788558538007248,
"grad_norm": 4.818440914154053,
"learning_rate": 1.939705097956554e-05,
"loss": 0.17412558794021607,
"step": 1360
},
{
"epoch": 0.13889944997845538,
"grad_norm": 2.2784125804901123,
"learning_rate": 1.9385771454420048e-05,
"loss": 0.1798375964164734,
"step": 1370
},
{
"epoch": 0.13991331457683825,
"grad_norm": 4.957714080810547,
"learning_rate": 1.9374390741744107e-05,
"loss": 0.1809779167175293,
"step": 1380
},
{
"epoch": 0.14092717917522116,
"grad_norm": 6.59519100189209,
"learning_rate": 1.9362908964232616e-05,
"loss": 0.19840087890625,
"step": 1390
},
{
"epoch": 0.14194104377360403,
"grad_norm": 2.1053969860076904,
"learning_rate": 1.9351326245670038e-05,
"loss": 0.1816962718963623,
"step": 1400
},
{
"epoch": 0.14295490837198693,
"grad_norm": 4.057821750640869,
"learning_rate": 1.933964271092908e-05,
"loss": 0.1751913070678711,
"step": 1410
},
{
"epoch": 0.1439687729703698,
"grad_norm": 2.7668092250823975,
"learning_rate": 1.932785848596934e-05,
"loss": 0.19536125659942627,
"step": 1420
},
{
"epoch": 0.14498263756875268,
"grad_norm": 3.2902517318725586,
"learning_rate": 1.9315973697835966e-05,
"loss": 0.18558683395385742,
"step": 1430
},
{
"epoch": 0.14599650216713558,
"grad_norm": 2.4489564895629883,
"learning_rate": 1.930398847465825e-05,
"loss": 0.17164087295532227,
"step": 1440
},
{
"epoch": 0.14701036676551846,
"grad_norm": 2.417701244354248,
"learning_rate": 1.9291902945648283e-05,
"loss": 0.16127595901489258,
"step": 1450
},
{
"epoch": 0.14802423136390136,
"grad_norm": 3.5651397705078125,
"learning_rate": 1.9279717241099554e-05,
"loss": 0.1882300853729248,
"step": 1460
},
{
"epoch": 0.14903809596228423,
"grad_norm": 1.616823673248291,
"learning_rate": 1.9267431492385524e-05,
"loss": 0.1830826997756958,
"step": 1470
},
{
"epoch": 0.15005196056066714,
"grad_norm": 3.2356247901916504,
"learning_rate": 1.925504583195823e-05,
"loss": 0.17465763092041015,
"step": 1480
},
{
"epoch": 0.15106582515905,
"grad_norm": 2.112828016281128,
"learning_rate": 1.924256039334686e-05,
"loss": 0.1843562126159668,
"step": 1490
},
{
"epoch": 0.15207968975743288,
"grad_norm": 2.9248452186584473,
"learning_rate": 1.9229975311156305e-05,
"loss": 0.18705670833587645,
"step": 1500
},
{
"epoch": 0.1530935543558158,
"grad_norm": 3.0341413021087646,
"learning_rate": 1.9217290721065693e-05,
"loss": 0.19148067235946656,
"step": 1510
},
{
"epoch": 0.15410741895419866,
"grad_norm": 4.6419243812561035,
"learning_rate": 1.9204506759826957e-05,
"loss": 0.17489591836929322,
"step": 1520
},
{
"epoch": 0.15512128355258156,
"grad_norm": 1.6488889455795288,
"learning_rate": 1.9191623565263342e-05,
"loss": 0.17685142755508423,
"step": 1530
},
{
"epoch": 0.15613514815096444,
"grad_norm": 2.3605966567993164,
"learning_rate": 1.9178641276267922e-05,
"loss": 0.180819571018219,
"step": 1540
},
{
"epoch": 0.1571490127493473,
"grad_norm": 1.9730758666992188,
"learning_rate": 1.91655600328021e-05,
"loss": 0.18506693840026855,
"step": 1550
},
{
"epoch": 0.1581628773477302,
"grad_norm": 5.571765899658203,
"learning_rate": 1.91523799758941e-05,
"loss": 0.17041449546813964,
"step": 1560
},
{
"epoch": 0.1591767419461131,
"grad_norm": 6.662985324859619,
"learning_rate": 1.9139101247637465e-05,
"loss": 0.17391374111175537,
"step": 1570
},
{
"epoch": 0.160190606544496,
"grad_norm": 5.317286968231201,
"learning_rate": 1.912572399118949e-05,
"loss": 0.178246808052063,
"step": 1580
},
{
"epoch": 0.16120447114287886,
"grad_norm": 2.9447314739227295,
"learning_rate": 1.9112248350769704e-05,
"loss": 0.19210422039031982,
"step": 1590
},
{
"epoch": 0.16221833574126177,
"grad_norm": 3.4214794635772705,
"learning_rate": 1.909867447165831e-05,
"loss": 0.17011003494262694,
"step": 1600
},
{
"epoch": 0.16323220033964464,
"grad_norm": 1.9205586910247803,
"learning_rate": 1.908500250019462e-05,
"loss": 0.1725075960159302,
"step": 1610
},
{
"epoch": 0.16424606493802751,
"grad_norm": 4.944511890411377,
"learning_rate": 1.9071232583775465e-05,
"loss": 0.17327165603637695,
"step": 1620
},
{
"epoch": 0.16525992953641042,
"grad_norm": 2.6423232555389404,
"learning_rate": 1.9057364870853628e-05,
"loss": 0.15719836950302124,
"step": 1630
},
{
"epoch": 0.1662737941347933,
"grad_norm": 3.1706528663635254,
"learning_rate": 1.904339951093622e-05,
"loss": 0.17445197105407714,
"step": 1640
},
{
"epoch": 0.1672876587331762,
"grad_norm": 3.547147274017334,
"learning_rate": 1.902933665458309e-05,
"loss": 0.17499091625213622,
"step": 1650
},
{
"epoch": 0.16830152333155907,
"grad_norm": 3.463301181793213,
"learning_rate": 1.901517645340518e-05,
"loss": 0.18497990369796752,
"step": 1660
},
{
"epoch": 0.16931538792994197,
"grad_norm": 6.482597351074219,
"learning_rate": 1.900091906006291e-05,
"loss": 0.19047290086746216,
"step": 1670
},
{
"epoch": 0.17032925252832484,
"grad_norm": 2.769881248474121,
"learning_rate": 1.8986564628264527e-05,
"loss": 0.18774659633636476,
"step": 1680
},
{
"epoch": 0.17134311712670772,
"grad_norm": 2.2915635108947754,
"learning_rate": 1.8972113312764432e-05,
"loss": 0.16936590671539306,
"step": 1690
},
{
"epoch": 0.17235698172509062,
"grad_norm": 2.764456033706665,
"learning_rate": 1.895756526936153e-05,
"loss": 0.1771099328994751,
"step": 1700
},
{
"epoch": 0.1733708463234735,
"grad_norm": 3.9696226119995117,
"learning_rate": 1.8942920654897553e-05,
"loss": 0.1843070864677429,
"step": 1710
},
{
"epoch": 0.1743847109218564,
"grad_norm": 1.6980444192886353,
"learning_rate": 1.8928179627255347e-05,
"loss": 0.19070732593536377,
"step": 1720
},
{
"epoch": 0.17539857552023927,
"grad_norm": 3.563830614089966,
"learning_rate": 1.891334234535719e-05,
"loss": 0.1831926941871643,
"step": 1730
},
{
"epoch": 0.17641244011862217,
"grad_norm": 3.664372682571411,
"learning_rate": 1.8898408969163078e-05,
"loss": 0.16258635520935058,
"step": 1740
},
{
"epoch": 0.17742630471700505,
"grad_norm": 2.7412891387939453,
"learning_rate": 1.8883379659668987e-05,
"loss": 0.1774962067604065,
"step": 1750
},
{
"epoch": 0.17844016931538792,
"grad_norm": 5.251802444458008,
"learning_rate": 1.8868254578905147e-05,
"loss": 0.18250734806060792,
"step": 1760
},
{
"epoch": 0.17945403391377082,
"grad_norm": 3.4117565155029297,
"learning_rate": 1.8853033889934293e-05,
"loss": 0.15069785118103027,
"step": 1770
},
{
"epoch": 0.1804678985121537,
"grad_norm": 2.6541056632995605,
"learning_rate": 1.883771775684991e-05,
"loss": 0.16013015508651735,
"step": 1780
},
{
"epoch": 0.1814817631105366,
"grad_norm": 5.473752498626709,
"learning_rate": 1.882230634477446e-05,
"loss": 0.1943489909172058,
"step": 1790
},
{
"epoch": 0.18249562770891947,
"grad_norm": 5.282114028930664,
"learning_rate": 1.8806799819857598e-05,
"loss": 0.19438533782958983,
"step": 1800
},
{
"epoch": 0.18350949230730235,
"grad_norm": 2.0042247772216797,
"learning_rate": 1.8791198349274393e-05,
"loss": 0.16537539958953856,
"step": 1810
},
{
"epoch": 0.18452335690568525,
"grad_norm": 3.1717019081115723,
"learning_rate": 1.8775502101223516e-05,
"loss": 0.1792454957962036,
"step": 1820
},
{
"epoch": 0.18553722150406812,
"grad_norm": 5.287769794464111,
"learning_rate": 1.8759711244925424e-05,
"loss": 0.17760031223297118,
"step": 1830
},
{
"epoch": 0.18655108610245102,
"grad_norm": 2.117659091949463,
"learning_rate": 1.874382595062055e-05,
"loss": 0.14842779636383058,
"step": 1840
},
{
"epoch": 0.1875649507008339,
"grad_norm": 2.184870958328247,
"learning_rate": 1.8727846389567457e-05,
"loss": 0.17725242376327516,
"step": 1850
},
{
"epoch": 0.1885788152992168,
"grad_norm": 3.9504480361938477,
"learning_rate": 1.871177273404098e-05,
"loss": 0.1820956587791443,
"step": 1860
},
{
"epoch": 0.18959267989759968,
"grad_norm": 4.390970706939697,
"learning_rate": 1.8695605157330398e-05,
"loss": 0.1572800397872925,
"step": 1870
},
{
"epoch": 0.19060654449598255,
"grad_norm": 2.855581045150757,
"learning_rate": 1.8679343833737546e-05,
"loss": 0.19102780818939208,
"step": 1880
},
{
"epoch": 0.19162040909436545,
"grad_norm": 2.974484443664551,
"learning_rate": 1.8662988938574926e-05,
"loss": 0.17589550018310546,
"step": 1890
},
{
"epoch": 0.19263427369274833,
"grad_norm": 2.0629777908325195,
"learning_rate": 1.8646540648163853e-05,
"loss": 0.16508569717407226,
"step": 1900
},
{
"epoch": 0.19364813829113123,
"grad_norm": 2.9782700538635254,
"learning_rate": 1.862999913983251e-05,
"loss": 0.17904813289642335,
"step": 1910
},
{
"epoch": 0.1946620028895141,
"grad_norm": 3.193021297454834,
"learning_rate": 1.8613364591914073e-05,
"loss": 0.1579646110534668,
"step": 1920
},
{
"epoch": 0.195675867487897,
"grad_norm": 3.0676937103271484,
"learning_rate": 1.8596637183744762e-05,
"loss": 0.16170411109924315,
"step": 1930
},
{
"epoch": 0.19668973208627988,
"grad_norm": 5.143824100494385,
"learning_rate": 1.857981709566193e-05,
"loss": 0.18313355445861818,
"step": 1940
},
{
"epoch": 0.19770359668466275,
"grad_norm": 3.7948811054229736,
"learning_rate": 1.85629045090021e-05,
"loss": 0.17511935234069825,
"step": 1950
},
{
"epoch": 0.19871746128304565,
"grad_norm": 2.063797950744629,
"learning_rate": 1.854589960609902e-05,
"loss": 0.14524309635162352,
"step": 1960
},
{
"epoch": 0.19973132588142853,
"grad_norm": 3.640420436859131,
"learning_rate": 1.8528802570281686e-05,
"loss": 0.1923311710357666,
"step": 1970
},
{
"epoch": 0.20074519047981143,
"grad_norm": 6.999709129333496,
"learning_rate": 1.851161358587239e-05,
"loss": 0.17201266288757325,
"step": 1980
},
{
"epoch": 0.2017590550781943,
"grad_norm": 4.272530555725098,
"learning_rate": 1.8494332838184707e-05,
"loss": 0.19175093173980712,
"step": 1990
},
{
"epoch": 0.2027729196765772,
"grad_norm": 4.186265468597412,
"learning_rate": 1.847696051352151e-05,
"loss": 0.16569194793701172,
"step": 2000
},
{
"epoch": 0.20378678427496008,
"grad_norm": 2.379350423812866,
"learning_rate": 1.8459496799172957e-05,
"loss": 0.16009128093719482,
"step": 2010
},
{
"epoch": 0.20480064887334296,
"grad_norm": 6.990966320037842,
"learning_rate": 1.8441941883414475e-05,
"loss": 0.20089597702026368,
"step": 2020
},
{
"epoch": 0.20581451347172586,
"grad_norm": 3.117676019668579,
"learning_rate": 1.8424295955504727e-05,
"loss": 0.1633314609527588,
"step": 2030
},
{
"epoch": 0.20682837807010873,
"grad_norm": 2.1671862602233887,
"learning_rate": 1.840655920568358e-05,
"loss": 0.18694061040878296,
"step": 2040
},
{
"epoch": 0.20784224266849163,
"grad_norm": 4.128241539001465,
"learning_rate": 1.838873182517004e-05,
"loss": 0.1698617458343506,
"step": 2050
},
{
"epoch": 0.2088561072668745,
"grad_norm": 4.436728000640869,
"learning_rate": 1.83708140061602e-05,
"loss": 0.17789593935012818,
"step": 2060
},
{
"epoch": 0.20986997186525738,
"grad_norm": 4.2754387855529785,
"learning_rate": 1.835280594182517e-05,
"loss": 0.17444523572921752,
"step": 2070
},
{
"epoch": 0.21088383646364028,
"grad_norm": 1.960911512374878,
"learning_rate": 1.833470782630898e-05,
"loss": 0.17404527664184571,
"step": 2080
},
{
"epoch": 0.21189770106202316,
"grad_norm": 5.089030742645264,
"learning_rate": 1.831651985472651e-05,
"loss": 0.1735629916191101,
"step": 2090
},
{
"epoch": 0.21291156566040606,
"grad_norm": 5.909832000732422,
"learning_rate": 1.8298242223161362e-05,
"loss": 0.18722455501556395,
"step": 2100
},
{
"epoch": 0.21392543025878893,
"grad_norm": 2.250805616378784,
"learning_rate": 1.827987512866377e-05,
"loss": 0.16743700504302977,
"step": 2110
},
{
"epoch": 0.21493929485717184,
"grad_norm": 4.170619010925293,
"learning_rate": 1.826141876924845e-05,
"loss": 0.1635421633720398,
"step": 2120
},
{
"epoch": 0.2159531594555547,
"grad_norm": 7.620894908905029,
"learning_rate": 1.8242873343892494e-05,
"loss": 0.19673404693603516,
"step": 2130
},
{
"epoch": 0.21696702405393758,
"grad_norm": 3.6671595573425293,
"learning_rate": 1.8224239052533194e-05,
"loss": 0.17396893501281738,
"step": 2140
},
{
"epoch": 0.2179808886523205,
"grad_norm": 4.397902488708496,
"learning_rate": 1.820551609606591e-05,
"loss": 0.17634141445159912,
"step": 2150
},
{
"epoch": 0.21899475325070336,
"grad_norm": 1.525463581085205,
"learning_rate": 1.81867046763419e-05,
"loss": 0.18169718980789185,
"step": 2160
},
{
"epoch": 0.22000861784908626,
"grad_norm": 3.4569568634033203,
"learning_rate": 1.816780499616613e-05,
"loss": 0.17575300931930543,
"step": 2170
},
{
"epoch": 0.22102248244746914,
"grad_norm": 3.230025291442871,
"learning_rate": 1.81488172592951e-05,
"loss": 0.16583213806152344,
"step": 2180
},
{
"epoch": 0.22203634704585204,
"grad_norm": 3.086561441421509,
"learning_rate": 1.812974167043465e-05,
"loss": 0.16707179546356202,
"step": 2190
},
{
"epoch": 0.2230502116442349,
"grad_norm": 2.4148106575012207,
"learning_rate": 1.8110578435237748e-05,
"loss": 0.17957148551940919,
"step": 2200
},
{
"epoch": 0.2240640762426178,
"grad_norm": 2.8748586177825928,
"learning_rate": 1.8091327760302267e-05,
"loss": 0.18021788597106933,
"step": 2210
},
{
"epoch": 0.2250779408410007,
"grad_norm": 9.808656692504883,
"learning_rate": 1.8071989853168765e-05,
"loss": 0.21856083869934081,
"step": 2220
},
{
"epoch": 0.22609180543938356,
"grad_norm": 1.5997623205184937,
"learning_rate": 1.8052564922318247e-05,
"loss": 0.1830922245979309,
"step": 2230
},
{
"epoch": 0.22710567003776647,
"grad_norm": 6.286662578582764,
"learning_rate": 1.803305317716992e-05,
"loss": 0.1811690092086792,
"step": 2240
},
{
"epoch": 0.22811953463614934,
"grad_norm": 4.306618690490723,
"learning_rate": 1.8013454828078925e-05,
"loss": 0.17128305435180663,
"step": 2250
},
{
"epoch": 0.22913339923453221,
"grad_norm": 3.433974027633667,
"learning_rate": 1.7993770086334082e-05,
"loss": 0.1644970655441284,
"step": 2260
},
{
"epoch": 0.23014726383291512,
"grad_norm": 3.4637324810028076,
"learning_rate": 1.7973999164155597e-05,
"loss": 0.16555945873260497,
"step": 2270
},
{
"epoch": 0.231161128431298,
"grad_norm": 3.410998582839966,
"learning_rate": 1.795414227469279e-05,
"loss": 0.15693230628967286,
"step": 2280
},
{
"epoch": 0.2321749930296809,
"grad_norm": 3.5962681770324707,
"learning_rate": 1.7934199632021795e-05,
"loss": 0.17563071250915527,
"step": 2290
},
{
"epoch": 0.23318885762806377,
"grad_norm": 3.0136187076568604,
"learning_rate": 1.7914171451143235e-05,
"loss": 0.17956857681274413,
"step": 2300
},
{
"epoch": 0.23420272222644667,
"grad_norm": 2.2541935443878174,
"learning_rate": 1.7894057947979923e-05,
"loss": 0.16640160083770753,
"step": 2310
},
{
"epoch": 0.23521658682482954,
"grad_norm": 4.521694183349609,
"learning_rate": 1.7873859339374528e-05,
"loss": 0.18860211372375488,
"step": 2320
},
{
"epoch": 0.23623045142321242,
"grad_norm": 6.37529182434082,
"learning_rate": 1.7853575843087227e-05,
"loss": 0.16800858974456787,
"step": 2330
},
{
"epoch": 0.23724431602159532,
"grad_norm": 2.5158936977386475,
"learning_rate": 1.7833207677793382e-05,
"loss": 0.17305573225021362,
"step": 2340
},
{
"epoch": 0.2382581806199782,
"grad_norm": 1.7935547828674316,
"learning_rate": 1.7812755063081156e-05,
"loss": 0.16575219631195068,
"step": 2350
},
{
"epoch": 0.2392720452183611,
"grad_norm": 3.545647144317627,
"learning_rate": 1.7792218219449165e-05,
"loss": 0.16571419239044188,
"step": 2360
},
{
"epoch": 0.24028590981674397,
"grad_norm": 2.9506680965423584,
"learning_rate": 1.777159736830408e-05,
"loss": 0.16563751697540283,
"step": 2370
},
{
"epoch": 0.24129977441512687,
"grad_norm": 3.055366039276123,
"learning_rate": 1.7750892731958268e-05,
"loss": 0.16773631572723388,
"step": 2380
},
{
"epoch": 0.24231363901350975,
"grad_norm": 3.098236083984375,
"learning_rate": 1.773010453362737e-05,
"loss": 0.1673525094985962,
"step": 2390
},
{
"epoch": 0.24332750361189262,
"grad_norm": 2.492546558380127,
"learning_rate": 1.770923299742791e-05,
"loss": 0.16642849445343016,
"step": 2400
},
{
"epoch": 0.24434136821027552,
"grad_norm": 1.8629801273345947,
"learning_rate": 1.768827834837487e-05,
"loss": 0.16694953441619872,
"step": 2410
},
{
"epoch": 0.2453552328086584,
"grad_norm": 3.617403268814087,
"learning_rate": 1.7667240812379266e-05,
"loss": 0.1532442569732666,
"step": 2420
},
{
"epoch": 0.2463690974070413,
"grad_norm": 3.4308359622955322,
"learning_rate": 1.7646120616245724e-05,
"loss": 0.1753792643547058,
"step": 2430
},
{
"epoch": 0.24738296200542417,
"grad_norm": 2.032240390777588,
"learning_rate": 1.7624917987670006e-05,
"loss": 0.17717218399047852,
"step": 2440
},
{
"epoch": 0.24839682660380707,
"grad_norm": 2.795583724975586,
"learning_rate": 1.7603633155236595e-05,
"loss": 0.16601829528808593,
"step": 2450
},
{
"epoch": 0.24941069120218995,
"grad_norm": 4.375458240509033,
"learning_rate": 1.7582266348416194e-05,
"loss": 0.17948572635650634,
"step": 2460
},
{
"epoch": 0.2504245558005728,
"grad_norm": 1.994635820388794,
"learning_rate": 1.7560817797563275e-05,
"loss": 0.17044732570648194,
"step": 2470
},
{
"epoch": 0.2514384203989557,
"grad_norm": 2.603647470474243,
"learning_rate": 1.7539287733913592e-05,
"loss": 0.17388134002685546,
"step": 2480
},
{
"epoch": 0.2524522849973386,
"grad_norm": 4.198951244354248,
"learning_rate": 1.7517676389581675e-05,
"loss": 0.17364945411682128,
"step": 2490
},
{
"epoch": 0.2534661495957215,
"grad_norm": 4.913215637207031,
"learning_rate": 1.7495983997558342e-05,
"loss": 0.16132752895355223,
"step": 2500
},
{
"epoch": 0.2544800141941044,
"grad_norm": 3.8931314945220947,
"learning_rate": 1.7474210791708177e-05,
"loss": 0.16926224231719972,
"step": 2510
},
{
"epoch": 0.25549387879248725,
"grad_norm": 3.539544105529785,
"learning_rate": 1.7452357006767026e-05,
"loss": 0.16524673700332643,
"step": 2520
},
{
"epoch": 0.2565077433908701,
"grad_norm": 2.324227809906006,
"learning_rate": 1.7430422878339443e-05,
"loss": 0.17147021293640136,
"step": 2530
},
{
"epoch": 0.25752160798925305,
"grad_norm": 1.4820269346237183,
"learning_rate": 1.740840864289616e-05,
"loss": 0.1600294828414917,
"step": 2540
},
{
"epoch": 0.25853547258763593,
"grad_norm": 2.3553354740142822,
"learning_rate": 1.7386314537771553e-05,
"loss": 0.16482995748519896,
"step": 2550
},
{
"epoch": 0.2595493371860188,
"grad_norm": 4.244352340698242,
"learning_rate": 1.736414080116105e-05,
"loss": 0.17981150150299072,
"step": 2560
},
{
"epoch": 0.2605632017844017,
"grad_norm": 3.326890468597412,
"learning_rate": 1.7341887672118596e-05,
"loss": 0.16453680992126465,
"step": 2570
},
{
"epoch": 0.2615770663827846,
"grad_norm": 3.0351738929748535,
"learning_rate": 1.7319555390554054e-05,
"loss": 0.19366614818572997,
"step": 2580
},
{
"epoch": 0.2625909309811675,
"grad_norm": 2.6301803588867188,
"learning_rate": 1.7297144197230637e-05,
"loss": 0.17153596878051758,
"step": 2590
},
{
"epoch": 0.26360479557955036,
"grad_norm": 3.2530689239501953,
"learning_rate": 1.7274654333762284e-05,
"loss": 0.15773544311523438,
"step": 2600
},
{
"epoch": 0.26461866017793323,
"grad_norm": 4.381091117858887,
"learning_rate": 1.7252086042611086e-05,
"loss": 0.18805346488952637,
"step": 2610
},
{
"epoch": 0.2656325247763161,
"grad_norm": 2.469048023223877,
"learning_rate": 1.722943956708466e-05,
"loss": 0.1624135971069336,
"step": 2620
},
{
"epoch": 0.26664638937469903,
"grad_norm": 3.199873924255371,
"learning_rate": 1.720671515133352e-05,
"loss": 0.16563520431518555,
"step": 2630
},
{
"epoch": 0.2676602539730819,
"grad_norm": 4.243142604827881,
"learning_rate": 1.7183913040348447e-05,
"loss": 0.17006005048751832,
"step": 2640
},
{
"epoch": 0.2686741185714648,
"grad_norm": 3.7615294456481934,
"learning_rate": 1.716103347995785e-05,
"loss": 0.16586347818374633,
"step": 2650
},
{
"epoch": 0.26968798316984766,
"grad_norm": 2.4043068885803223,
"learning_rate": 1.713807671682513e-05,
"loss": 0.15249793529510497,
"step": 2660
},
{
"epoch": 0.27070184776823053,
"grad_norm": 4.6412835121154785,
"learning_rate": 1.711504299844599e-05,
"loss": 0.1691739559173584,
"step": 2670
},
{
"epoch": 0.27171571236661346,
"grad_norm": 1.8787870407104492,
"learning_rate": 1.7091932573145798e-05,
"loss": 0.17830950021743774,
"step": 2680
},
{
"epoch": 0.27272957696499633,
"grad_norm": 4.811567306518555,
"learning_rate": 1.706874569007688e-05,
"loss": 0.1643239974975586,
"step": 2690
},
{
"epoch": 0.2737434415633792,
"grad_norm": 1.7669070959091187,
"learning_rate": 1.7045482599215856e-05,
"loss": 0.1817120671272278,
"step": 2700
},
{
"epoch": 0.2747573061617621,
"grad_norm": 2.9812371730804443,
"learning_rate": 1.7022143551360947e-05,
"loss": 0.1577073574066162,
"step": 2710
},
{
"epoch": 0.27577117076014496,
"grad_norm": 3.8998522758483887,
"learning_rate": 1.699872879812925e-05,
"loss": 0.18089557886123658,
"step": 2720
},
{
"epoch": 0.2767850353585279,
"grad_norm": 5.01729679107666,
"learning_rate": 1.6975238591954048e-05,
"loss": 0.1851854920387268,
"step": 2730
},
{
"epoch": 0.27779889995691076,
"grad_norm": 3.4985783100128174,
"learning_rate": 1.695167318608206e-05,
"loss": 0.1736953854560852,
"step": 2740
},
{
"epoch": 0.27881276455529364,
"grad_norm": 2.215883493423462,
"learning_rate": 1.6928032834570767e-05,
"loss": 0.17245519161224365,
"step": 2750
},
{
"epoch": 0.2798266291536765,
"grad_norm": 3.3665244579315186,
"learning_rate": 1.690431779228559e-05,
"loss": 0.17383098602294922,
"step": 2760
},
{
"epoch": 0.28084049375205944,
"grad_norm": 6.8604912757873535,
"learning_rate": 1.6880528314897212e-05,
"loss": 0.16884692907333373,
"step": 2770
},
{
"epoch": 0.2818543583504423,
"grad_norm": 3.2428765296936035,
"learning_rate": 1.6856664658878797e-05,
"loss": 0.1655750870704651,
"step": 2780
},
{
"epoch": 0.2828682229488252,
"grad_norm": 3.7955057621002197,
"learning_rate": 1.6832727081503227e-05,
"loss": 0.17067900896072388,
"step": 2790
},
{
"epoch": 0.28388208754720806,
"grad_norm": 4.259796142578125,
"learning_rate": 1.6808715840840308e-05,
"loss": 0.1577405571937561,
"step": 2800
},
{
"epoch": 0.28489595214559094,
"grad_norm": 3.0467281341552734,
"learning_rate": 1.6784631195754027e-05,
"loss": 0.16981039047241211,
"step": 2810
},
{
"epoch": 0.28590981674397387,
"grad_norm": 2.2386250495910645,
"learning_rate": 1.6760473405899733e-05,
"loss": 0.152715003490448,
"step": 2820
},
{
"epoch": 0.28692368134235674,
"grad_norm": 3.657987117767334,
"learning_rate": 1.673624273172134e-05,
"loss": 0.1659175395965576,
"step": 2830
},
{
"epoch": 0.2879375459407396,
"grad_norm": 1.9415174722671509,
"learning_rate": 1.6711939434448537e-05,
"loss": 0.15578653812408447,
"step": 2840
},
{
"epoch": 0.2889514105391225,
"grad_norm": 5.060669898986816,
"learning_rate": 1.6687563776093943e-05,
"loss": 0.18613462448120116,
"step": 2850
},
{
"epoch": 0.28996527513750536,
"grad_norm": 2.2133095264434814,
"learning_rate": 1.6663116019450306e-05,
"loss": 0.1559851884841919,
"step": 2860
},
{
"epoch": 0.2909791397358883,
"grad_norm": 2.2348644733428955,
"learning_rate": 1.6638596428087662e-05,
"loss": 0.17732785940170287,
"step": 2870
},
{
"epoch": 0.29199300433427117,
"grad_norm": 4.315962791442871,
"learning_rate": 1.66140052663505e-05,
"loss": 0.1665308117866516,
"step": 2880
},
{
"epoch": 0.29300686893265404,
"grad_norm": 2.629296064376831,
"learning_rate": 1.6589342799354895e-05,
"loss": 0.16116490364074706,
"step": 2890
},
{
"epoch": 0.2940207335310369,
"grad_norm": 2.8975627422332764,
"learning_rate": 1.6564609292985663e-05,
"loss": 0.16961835622787474,
"step": 2900
},
{
"epoch": 0.2950345981294198,
"grad_norm": 3.780745506286621,
"learning_rate": 1.6539805013893493e-05,
"loss": 0.17933913469314575,
"step": 2910
},
{
"epoch": 0.2960484627278027,
"grad_norm": 5.933714389801025,
"learning_rate": 1.6514930229492073e-05,
"loss": 0.17891569137573243,
"step": 2920
},
{
"epoch": 0.2970623273261856,
"grad_norm": 3.7291836738586426,
"learning_rate": 1.6489985207955207e-05,
"loss": 0.16337337493896484,
"step": 2930
},
{
"epoch": 0.29807619192456847,
"grad_norm": 2.7149899005889893,
"learning_rate": 1.6464970218213912e-05,
"loss": 0.17081141471862793,
"step": 2940
},
{
"epoch": 0.29909005652295134,
"grad_norm": 3.476491928100586,
"learning_rate": 1.643988552995354e-05,
"loss": 0.17542653083801268,
"step": 2950
},
{
"epoch": 0.30010392112133427,
"grad_norm": 3.9330251216888428,
"learning_rate": 1.6414731413610856e-05,
"loss": 0.15865416526794435,
"step": 2960
},
{
"epoch": 0.30111778571971715,
"grad_norm": 2.2122082710266113,
"learning_rate": 1.638950814037112e-05,
"loss": 0.1607208728790283,
"step": 2970
},
{
"epoch": 0.3021316503181,
"grad_norm": 3.720127582550049,
"learning_rate": 1.6364215982165178e-05,
"loss": 0.15621423721313477,
"step": 2980
},
{
"epoch": 0.3031455149164829,
"grad_norm": 2.5231919288635254,
"learning_rate": 1.633885521166651e-05,
"loss": 0.16144974231719972,
"step": 2990
},
{
"epoch": 0.30415937951486577,
"grad_norm": 1.996989369392395,
"learning_rate": 1.6313426102288315e-05,
"loss": 0.16638247966766356,
"step": 3000
},
{
"epoch": 0.3051732441132487,
"grad_norm": 4.553835391998291,
"learning_rate": 1.6287928928180544e-05,
"loss": 0.1720791459083557,
"step": 3010
},
{
"epoch": 0.3061871087116316,
"grad_norm": 2.1291863918304443,
"learning_rate": 1.6262363964226943e-05,
"loss": 0.15042562484741212,
"step": 3020
},
{
"epoch": 0.30720097331001445,
"grad_norm": 3.672060489654541,
"learning_rate": 1.623673148604211e-05,
"loss": 0.15640090703964232,
"step": 3030
},
{
"epoch": 0.3082148379083973,
"grad_norm": 3.042440891265869,
"learning_rate": 1.6211031769968503e-05,
"loss": 0.1620028495788574,
"step": 3040
},
{
"epoch": 0.3092287025067802,
"grad_norm": 4.361697673797607,
"learning_rate": 1.6185265093073468e-05,
"loss": 0.16020448207855226,
"step": 3050
},
{
"epoch": 0.3102425671051631,
"grad_norm": 2.828141450881958,
"learning_rate": 1.6159431733146255e-05,
"loss": 0.16198461055755614,
"step": 3060
},
{
"epoch": 0.311256431703546,
"grad_norm": 2.597050905227661,
"learning_rate": 1.613353196869502e-05,
"loss": 0.16330896615982055,
"step": 3070
},
{
"epoch": 0.3122702963019289,
"grad_norm": 2.5923476219177246,
"learning_rate": 1.6107566078943818e-05,
"loss": 0.173339319229126,
"step": 3080
},
{
"epoch": 0.31328416090031175,
"grad_norm": 2.1285786628723145,
"learning_rate": 1.608153434382961e-05,
"loss": 0.1736527681350708,
"step": 3090
},
{
"epoch": 0.3142980254986946,
"grad_norm": 1.5555777549743652,
"learning_rate": 1.6055437043999215e-05,
"loss": 0.15231887102127076,
"step": 3100
},
{
"epoch": 0.31531189009707755,
"grad_norm": 2.2965247631073,
"learning_rate": 1.6029274460806318e-05,
"loss": 0.16754860877990724,
"step": 3110
},
{
"epoch": 0.3163257546954604,
"grad_norm": 3.110494375228882,
"learning_rate": 1.6003046876308413e-05,
"loss": 0.17180986404418946,
"step": 3120
},
{
"epoch": 0.3173396192938433,
"grad_norm": 3.5457401275634766,
"learning_rate": 1.5976754573263767e-05,
"loss": 0.1518089771270752,
"step": 3130
},
{
"epoch": 0.3183534838922262,
"grad_norm": 2.8581697940826416,
"learning_rate": 1.5950397835128378e-05,
"loss": 0.17274887561798097,
"step": 3140
},
{
"epoch": 0.3193673484906091,
"grad_norm": 2.555356740951538,
"learning_rate": 1.5923976946052918e-05,
"loss": 0.14464955329895018,
"step": 3150
},
{
"epoch": 0.320381213088992,
"grad_norm": 5.130317211151123,
"learning_rate": 1.5897492190879667e-05,
"loss": 0.17186269760131836,
"step": 3160
},
{
"epoch": 0.32139507768737485,
"grad_norm": 2.556325674057007,
"learning_rate": 1.5870943855139437e-05,
"loss": 0.1711033582687378,
"step": 3170
},
{
"epoch": 0.3224089422857577,
"grad_norm": 2.366656541824341,
"learning_rate": 1.5844332225048496e-05,
"loss": 0.15182864665985107,
"step": 3180
},
{
"epoch": 0.3234228068841406,
"grad_norm": 4.373097896575928,
"learning_rate": 1.5817657587505503e-05,
"loss": 0.16280362606048585,
"step": 3190
},
{
"epoch": 0.32443667148252353,
"grad_norm": 3.2318685054779053,
"learning_rate": 1.5790920230088375e-05,
"loss": 0.16563940048217773,
"step": 3200
},
{
"epoch": 0.3254505360809064,
"grad_norm": 5.821103096008301,
"learning_rate": 1.576412044105122e-05,
"loss": 0.16262004375457764,
"step": 3210
},
{
"epoch": 0.3264644006792893,
"grad_norm": 3.84277081489563,
"learning_rate": 1.5737258509321224e-05,
"loss": 0.16370152235031127,
"step": 3220
},
{
"epoch": 0.32747826527767215,
"grad_norm": 4.633796215057373,
"learning_rate": 1.5710334724495514e-05,
"loss": 0.1618485927581787,
"step": 3230
},
{
"epoch": 0.32849212987605503,
"grad_norm": 1.8301515579223633,
"learning_rate": 1.5683349376838067e-05,
"loss": 0.16502333879470826,
"step": 3240
},
{
"epoch": 0.32950599447443796,
"grad_norm": 3.3908913135528564,
"learning_rate": 1.565630275727656e-05,
"loss": 0.15443035364151,
"step": 3250
},
{
"epoch": 0.33051985907282083,
"grad_norm": 3.1103439331054688,
"learning_rate": 1.5629195157399236e-05,
"loss": 0.17389264106750488,
"step": 3260
},
{
"epoch": 0.3315337236712037,
"grad_norm": 4.796304225921631,
"learning_rate": 1.5602026869451773e-05,
"loss": 0.15919703245162964,
"step": 3270
},
{
"epoch": 0.3325475882695866,
"grad_norm": 2.628323554992676,
"learning_rate": 1.5574798186334112e-05,
"loss": 0.15130102634429932,
"step": 3280
},
{
"epoch": 0.3335614528679695,
"grad_norm": 4.610678672790527,
"learning_rate": 1.5547509401597322e-05,
"loss": 0.17282642126083375,
"step": 3290
},
{
"epoch": 0.3345753174663524,
"grad_norm": 1.7038581371307373,
"learning_rate": 1.552016080944042e-05,
"loss": 0.15837472677230835,
"step": 3300
},
{
"epoch": 0.33558918206473526,
"grad_norm": 2.168628454208374,
"learning_rate": 1.5492752704707198e-05,
"loss": 0.1642472743988037,
"step": 3310
},
{
"epoch": 0.33660304666311813,
"grad_norm": 4.260249137878418,
"learning_rate": 1.5465285382883065e-05,
"loss": 0.1682758331298828,
"step": 3320
},
{
"epoch": 0.337616911261501,
"grad_norm": 5.710574626922607,
"learning_rate": 1.5437759140091827e-05,
"loss": 0.17294664382934571,
"step": 3330
},
{
"epoch": 0.33863077585988394,
"grad_norm": 3.5633630752563477,
"learning_rate": 1.541017427309254e-05,
"loss": 0.16976113319396974,
"step": 3340
},
{
"epoch": 0.3396446404582668,
"grad_norm": 3.060716390609741,
"learning_rate": 1.538253107927625e-05,
"loss": 0.15048539638519287,
"step": 3350
},
{
"epoch": 0.3406585050566497,
"grad_norm": 3.932520627975464,
"learning_rate": 1.535482985666285e-05,
"loss": 0.15111865997314453,
"step": 3360
},
{
"epoch": 0.34167236965503256,
"grad_norm": 6.102091312408447,
"learning_rate": 1.5327070903897832e-05,
"loss": 0.16985511779785156,
"step": 3370
},
{
"epoch": 0.34268623425341543,
"grad_norm": 6.7877631187438965,
"learning_rate": 1.529925452024907e-05,
"loss": 0.16917786598205567,
"step": 3380
},
{
"epoch": 0.34370009885179836,
"grad_norm": 4.888689994812012,
"learning_rate": 1.5271381005603593e-05,
"loss": 0.17459226846694947,
"step": 3390
},
{
"epoch": 0.34471396345018124,
"grad_norm": 5.164940357208252,
"learning_rate": 1.5243450660464368e-05,
"loss": 0.15665051937103272,
"step": 3400
},
{
"epoch": 0.3457278280485641,
"grad_norm": 3.429985761642456,
"learning_rate": 1.5215463785947036e-05,
"loss": 0.1675715923309326,
"step": 3410
},
{
"epoch": 0.346741692646947,
"grad_norm": 2.2632784843444824,
"learning_rate": 1.518742068377669e-05,
"loss": 0.15969823598861693,
"step": 3420
},
{
"epoch": 0.34775555724532986,
"grad_norm": 2.134683847427368,
"learning_rate": 1.5159321656284602e-05,
"loss": 0.15926966667175294,
"step": 3430
},
{
"epoch": 0.3487694218437128,
"grad_norm": 2.650768756866455,
"learning_rate": 1.5131167006404978e-05,
"loss": 0.16770187616348267,
"step": 3440
},
{
"epoch": 0.34978328644209566,
"grad_norm": 6.334310054779053,
"learning_rate": 1.5102957037671689e-05,
"loss": 0.15366995334625244,
"step": 3450
},
{
"epoch": 0.35079715104047854,
"grad_norm": 2.3370039463043213,
"learning_rate": 1.5074692054214977e-05,
"loss": 0.16115918159484863,
"step": 3460
},
{
"epoch": 0.3518110156388614,
"grad_norm": 2.291557550430298,
"learning_rate": 1.504637236075822e-05,
"loss": 0.15023773908615112,
"step": 3470
},
{
"epoch": 0.35282488023724434,
"grad_norm": 2.994626998901367,
"learning_rate": 1.5017998262614609e-05,
"loss": 0.16233084201812745,
"step": 3480
},
{
"epoch": 0.3538387448356272,
"grad_norm": 1.7689803838729858,
"learning_rate": 1.498957006568386e-05,
"loss": 0.15465683937072755,
"step": 3490
},
{
"epoch": 0.3548526094340101,
"grad_norm": 4.398205757141113,
"learning_rate": 1.4961088076448947e-05,
"loss": 0.15347299575805665,
"step": 3500
},
{
"epoch": 0.35586647403239297,
"grad_norm": 2.64099383354187,
"learning_rate": 1.4932552601972763e-05,
"loss": 0.16979167461395264,
"step": 3510
},
{
"epoch": 0.35688033863077584,
"grad_norm": 5.689368724822998,
"learning_rate": 1.490396394989482e-05,
"loss": 0.1701521635055542,
"step": 3520
},
{
"epoch": 0.35789420322915877,
"grad_norm": 3.316802978515625,
"learning_rate": 1.4875322428427943e-05,
"loss": 0.15509395599365233,
"step": 3530
},
{
"epoch": 0.35890806782754164,
"grad_norm": 2.4950170516967773,
"learning_rate": 1.4846628346354934e-05,
"loss": 0.15953667163848878,
"step": 3540
},
{
"epoch": 0.3599219324259245,
"grad_norm": 4.28302001953125,
"learning_rate": 1.4817882013025253e-05,
"loss": 0.17433925867080688,
"step": 3550
},
{
"epoch": 0.3609357970243074,
"grad_norm": 4.064145088195801,
"learning_rate": 1.478908373835167e-05,
"loss": 0.1646304488182068,
"step": 3560
},
{
"epoch": 0.36194966162269027,
"grad_norm": 2.8442327976226807,
"learning_rate": 1.476023383280694e-05,
"loss": 0.1612727165222168,
"step": 3570
},
{
"epoch": 0.3629635262210732,
"grad_norm": 4.01204252243042,
"learning_rate": 1.473133260742045e-05,
"loss": 0.16602267026901246,
"step": 3580
},
{
"epoch": 0.36397739081945607,
"grad_norm": 4.4011759757995605,
"learning_rate": 1.4702380373774852e-05,
"loss": 0.164500892162323,
"step": 3590
},
{
"epoch": 0.36499125541783894,
"grad_norm": 3.0472209453582764,
"learning_rate": 1.4673377444002724e-05,
"loss": 0.14644829034805298,
"step": 3600
},
{
"epoch": 0.3660051200162218,
"grad_norm": 5.910120964050293,
"learning_rate": 1.4644324130783197e-05,
"loss": 0.15490037202835083,
"step": 3610
},
{
"epoch": 0.3670189846146047,
"grad_norm": 1.412795901298523,
"learning_rate": 1.4615220747338574e-05,
"loss": 0.17500741481781007,
"step": 3620
},
{
"epoch": 0.3680328492129876,
"grad_norm": 3.9491922855377197,
"learning_rate": 1.4586067607430976e-05,
"loss": 0.16409678459167482,
"step": 3630
},
{
"epoch": 0.3690467138113705,
"grad_norm": 2.5529725551605225,
"learning_rate": 1.4556865025358942e-05,
"loss": 0.16783540248870848,
"step": 3640
},
{
"epoch": 0.37006057840975337,
"grad_norm": 2.6295888423919678,
"learning_rate": 1.4527613315954035e-05,
"loss": 0.159515643119812,
"step": 3650
},
{
"epoch": 0.37107444300813625,
"grad_norm": 2.802554130554199,
"learning_rate": 1.4498312794577465e-05,
"loss": 0.15086755752563477,
"step": 3660
},
{
"epoch": 0.3720883076065192,
"grad_norm": 2.534716844558716,
"learning_rate": 1.4468963777116686e-05,
"loss": 0.1863459348678589,
"step": 3670
},
{
"epoch": 0.37310217220490205,
"grad_norm": 3.911702871322632,
"learning_rate": 1.4439566579981979e-05,
"loss": 0.17214031219482423,
"step": 3680
},
{
"epoch": 0.3741160368032849,
"grad_norm": 2.040062427520752,
"learning_rate": 1.4410121520103045e-05,
"loss": 0.1539656400680542,
"step": 3690
},
{
"epoch": 0.3751299014016678,
"grad_norm": 1.8257042169570923,
"learning_rate": 1.4380628914925597e-05,
"loss": 0.1414958953857422,
"step": 3700
},
{
"epoch": 0.37614376600005067,
"grad_norm": 5.941495895385742,
"learning_rate": 1.4351089082407929e-05,
"loss": 0.15538737773895264,
"step": 3710
},
{
"epoch": 0.3771576305984336,
"grad_norm": 2.652339220046997,
"learning_rate": 1.4321502341017496e-05,
"loss": 0.17027904987335205,
"step": 3720
},
{
"epoch": 0.3781714951968165,
"grad_norm": 4.053829193115234,
"learning_rate": 1.4291869009727463e-05,
"loss": 0.15566668510437012,
"step": 3730
},
{
"epoch": 0.37918535979519935,
"grad_norm": 3.401432991027832,
"learning_rate": 1.4262189408013294e-05,
"loss": 0.15327198505401612,
"step": 3740
},
{
"epoch": 0.3801992243935822,
"grad_norm": 3.965622663497925,
"learning_rate": 1.4232463855849275e-05,
"loss": 0.16917521953582765,
"step": 3750
},
{
"epoch": 0.3812130889919651,
"grad_norm": 1.4921987056732178,
"learning_rate": 1.4202692673705097e-05,
"loss": 0.17604564428329467,
"step": 3760
},
{
"epoch": 0.38222695359034803,
"grad_norm": 2.556662082672119,
"learning_rate": 1.4172876182542372e-05,
"loss": 0.1631016969680786,
"step": 3770
},
{
"epoch": 0.3832408181887309,
"grad_norm": 2.4430813789367676,
"learning_rate": 1.41430147038112e-05,
"loss": 0.15111420154571534,
"step": 3780
},
{
"epoch": 0.3842546827871138,
"grad_norm": 1.6761116981506348,
"learning_rate": 1.411310855944668e-05,
"loss": 0.16334908008575438,
"step": 3790
},
{
"epoch": 0.38526854738549665,
"grad_norm": 3.358201742172241,
"learning_rate": 1.4083158071865451e-05,
"loss": 0.1732572078704834,
"step": 3800
},
{
"epoch": 0.3862824119838795,
"grad_norm": 1.9510747194290161,
"learning_rate": 1.4053163563962215e-05,
"loss": 0.16069307327270507,
"step": 3810
},
{
"epoch": 0.38729627658226246,
"grad_norm": 1.9816813468933105,
"learning_rate": 1.4023125359106253e-05,
"loss": 0.1587695837020874,
"step": 3820
},
{
"epoch": 0.38831014118064533,
"grad_norm": 2.93491792678833,
"learning_rate": 1.3993043781137945e-05,
"loss": 0.1575850486755371,
"step": 3830
},
{
"epoch": 0.3893240057790282,
"grad_norm": 2.454977035522461,
"learning_rate": 1.3962919154365273e-05,
"loss": 0.1692113161087036,
"step": 3840
},
{
"epoch": 0.3903378703774111,
"grad_norm": 2.051760196685791,
"learning_rate": 1.393275180356032e-05,
"loss": 0.1702409029006958,
"step": 3850
},
{
"epoch": 0.391351734975794,
"grad_norm": 3.9791953563690186,
"learning_rate": 1.3902542053955781e-05,
"loss": 0.1663435935974121,
"step": 3860
},
{
"epoch": 0.3923655995741769,
"grad_norm": 1.8256888389587402,
"learning_rate": 1.387229023124145e-05,
"loss": 0.16416010856628419,
"step": 3870
},
{
"epoch": 0.39337946417255976,
"grad_norm": 2.553506374359131,
"learning_rate": 1.3841996661560706e-05,
"loss": 0.15631983280181885,
"step": 3880
},
{
"epoch": 0.39439332877094263,
"grad_norm": 2.5561156272888184,
"learning_rate": 1.3811661671507004e-05,
"loss": 0.14551076889038086,
"step": 3890
},
{
"epoch": 0.3954071933693255,
"grad_norm": 4.886599063873291,
"learning_rate": 1.3781285588120343e-05,
"loss": 0.15688865184783934,
"step": 3900
},
{
"epoch": 0.39642105796770843,
"grad_norm": 2.239422082901001,
"learning_rate": 1.3750868738883755e-05,
"loss": 0.15444371700286866,
"step": 3910
},
{
"epoch": 0.3974349225660913,
"grad_norm": 3.721910238265991,
"learning_rate": 1.3720411451719762e-05,
"loss": 0.16776466369628906,
"step": 3920
},
{
"epoch": 0.3984487871644742,
"grad_norm": 2.7470624446868896,
"learning_rate": 1.3689914054986849e-05,
"loss": 0.16000046730041503,
"step": 3930
},
{
"epoch": 0.39946265176285706,
"grad_norm": 3.279550075531006,
"learning_rate": 1.3659376877475916e-05,
"loss": 0.15774569511413575,
"step": 3940
},
{
"epoch": 0.40047651636123993,
"grad_norm": 4.752569198608398,
"learning_rate": 1.3628800248406738e-05,
"loss": 0.15958181619644166,
"step": 3950
},
{
"epoch": 0.40149038095962286,
"grad_norm": 4.648467063903809,
"learning_rate": 1.3598184497424416e-05,
"loss": 0.1585504174232483,
"step": 3960
},
{
"epoch": 0.40250424555800574,
"grad_norm": 3.296570062637329,
"learning_rate": 1.3567529954595826e-05,
"loss": 0.1637623906135559,
"step": 3970
},
{
"epoch": 0.4035181101563886,
"grad_norm": 2.558234691619873,
"learning_rate": 1.3536836950406057e-05,
"loss": 0.1500474691390991,
"step": 3980
},
{
"epoch": 0.4045319747547715,
"grad_norm": 2.289421796798706,
"learning_rate": 1.3506105815754836e-05,
"loss": 0.1524118185043335,
"step": 3990
},
{
"epoch": 0.4055458393531544,
"grad_norm": 4.245085716247559,
"learning_rate": 1.3475336881952988e-05,
"loss": 0.18899943828582763,
"step": 4000
},
{
"epoch": 0.4065597039515373,
"grad_norm": 1.5013316869735718,
"learning_rate": 1.3444530480718838e-05,
"loss": 0.16498883962631225,
"step": 4010
},
{
"epoch": 0.40757356854992016,
"grad_norm": 1.7986037731170654,
"learning_rate": 1.3413686944174655e-05,
"loss": 0.14365683794021605,
"step": 4020
},
{
"epoch": 0.40858743314830304,
"grad_norm": 2.258695125579834,
"learning_rate": 1.3382806604843053e-05,
"loss": 0.16240376234054565,
"step": 4030
},
{
"epoch": 0.4096012977466859,
"grad_norm": 2.6286909580230713,
"learning_rate": 1.3351889795643423e-05,
"loss": 0.14309258460998536,
"step": 4040
},
{
"epoch": 0.41061516234506884,
"grad_norm": 4.951387405395508,
"learning_rate": 1.3320936849888322e-05,
"loss": 0.15969364643096923,
"step": 4050
},
{
"epoch": 0.4116290269434517,
"grad_norm": 1.7551288604736328,
"learning_rate": 1.3289948101279912e-05,
"loss": 0.17063914537429808,
"step": 4060
},
{
"epoch": 0.4126428915418346,
"grad_norm": 4.783592224121094,
"learning_rate": 1.3258923883906337e-05,
"loss": 0.16088149547576905,
"step": 4070
},
{
"epoch": 0.41365675614021746,
"grad_norm": 2.0447185039520264,
"learning_rate": 1.3227864532238113e-05,
"loss": 0.16195911169052124,
"step": 4080
},
{
"epoch": 0.41467062073860034,
"grad_norm": 1.3863320350646973,
"learning_rate": 1.3196770381124557e-05,
"loss": 0.15529143810272217,
"step": 4090
},
{
"epoch": 0.41568448533698327,
"grad_norm": 3.3794443607330322,
"learning_rate": 1.3165641765790153e-05,
"loss": 0.14443068504333495,
"step": 4100
},
{
"epoch": 0.41669834993536614,
"grad_norm": 1.6234121322631836,
"learning_rate": 1.3134479021830938e-05,
"loss": 0.16507425308227539,
"step": 4110
},
{
"epoch": 0.417712214533749,
"grad_norm": 5.144898414611816,
"learning_rate": 1.3103282485210882e-05,
"loss": 0.15035427808761598,
"step": 4120
},
{
"epoch": 0.4187260791321319,
"grad_norm": 2.022702217102051,
"learning_rate": 1.3072052492258286e-05,
"loss": 0.14927660226821898,
"step": 4130
},
{
"epoch": 0.41973994373051476,
"grad_norm": 2.260028839111328,
"learning_rate": 1.3040789379662131e-05,
"loss": 0.1366307854652405,
"step": 4140
},
{
"epoch": 0.4207538083288977,
"grad_norm": 2.043222188949585,
"learning_rate": 1.3009493484468464e-05,
"loss": 0.1533515453338623,
"step": 4150
},
{
"epoch": 0.42176767292728057,
"grad_norm": 1.4130011796951294,
"learning_rate": 1.2978165144076757e-05,
"loss": 0.15414297580718994,
"step": 4160
},
{
"epoch": 0.42278153752566344,
"grad_norm": 2.6063363552093506,
"learning_rate": 1.2946804696236277e-05,
"loss": 0.14306734800338744,
"step": 4170
},
{
"epoch": 0.4237954021240463,
"grad_norm": 3.1463468074798584,
"learning_rate": 1.2915412479042436e-05,
"loss": 0.14795007705688476,
"step": 4180
},
{
"epoch": 0.42480926672242925,
"grad_norm": 1.7587875127792358,
"learning_rate": 1.2883988830933148e-05,
"loss": 0.15215128660202026,
"step": 4190
},
{
"epoch": 0.4258231313208121,
"grad_norm": 2.7176544666290283,
"learning_rate": 1.2852534090685184e-05,
"loss": 0.15432438850402833,
"step": 4200
},
{
"epoch": 0.426836995919195,
"grad_norm": 2.3536274433135986,
"learning_rate": 1.282104859741052e-05,
"loss": 0.16567575931549072,
"step": 4210
},
{
"epoch": 0.42785086051757787,
"grad_norm": 1.2475451231002808,
"learning_rate": 1.2789532690552673e-05,
"loss": 0.16485981941223143,
"step": 4220
},
{
"epoch": 0.42886472511596074,
"grad_norm": 6.432428359985352,
"learning_rate": 1.2757986709883059e-05,
"loss": 0.1576997756958008,
"step": 4230
},
{
"epoch": 0.4298785897143437,
"grad_norm": 3.5963544845581055,
"learning_rate": 1.2726410995497304e-05,
"loss": 0.1578603982925415,
"step": 4240
},
{
"epoch": 0.43089245431272655,
"grad_norm": 2.4390745162963867,
"learning_rate": 1.2694805887811601e-05,
"loss": 0.1326158046722412,
"step": 4250
},
{
"epoch": 0.4319063189111094,
"grad_norm": 2.2474327087402344,
"learning_rate": 1.2663171727559025e-05,
"loss": 0.15247836112976074,
"step": 4260
},
{
"epoch": 0.4329201835094923,
"grad_norm": 2.5800416469573975,
"learning_rate": 1.2631508855785877e-05,
"loss": 0.14937365055084229,
"step": 4270
},
{
"epoch": 0.43393404810787517,
"grad_norm": 3.7599895000457764,
"learning_rate": 1.2599817613847973e-05,
"loss": 0.16565756797790526,
"step": 4280
},
{
"epoch": 0.4349479127062581,
"grad_norm": 3.1329996585845947,
"learning_rate": 1.2568098343407007e-05,
"loss": 0.15488150119781494,
"step": 4290
},
{
"epoch": 0.435961777304641,
"grad_norm": 4.1333746910095215,
"learning_rate": 1.2536351386426844e-05,
"loss": 0.1612410545349121,
"step": 4300
},
{
"epoch": 0.43697564190302385,
"grad_norm": 2.3739447593688965,
"learning_rate": 1.2504577085169826e-05,
"loss": 0.1727668046951294,
"step": 4310
},
{
"epoch": 0.4379895065014067,
"grad_norm": 2.2158942222595215,
"learning_rate": 1.2472775782193099e-05,
"loss": 0.15909501314163207,
"step": 4320
},
{
"epoch": 0.4390033710997896,
"grad_norm": 3.220492362976074,
"learning_rate": 1.2440947820344916e-05,
"loss": 0.1620936632156372,
"step": 4330
},
{
"epoch": 0.4400172356981725,
"grad_norm": 1.8679617643356323,
"learning_rate": 1.2409093542760925e-05,
"loss": 0.1517797589302063,
"step": 4340
},
{
"epoch": 0.4410311002965554,
"grad_norm": 3.5927414894104004,
"learning_rate": 1.2377213292860506e-05,
"loss": 0.17765305042266846,
"step": 4350
},
{
"epoch": 0.4420449648949383,
"grad_norm": 4.362872123718262,
"learning_rate": 1.2345307414343021e-05,
"loss": 0.1436387538909912,
"step": 4360
},
{
"epoch": 0.44305882949332115,
"grad_norm": 2.0737948417663574,
"learning_rate": 1.2313376251184148e-05,
"loss": 0.1532752752304077,
"step": 4370
},
{
"epoch": 0.4440726940917041,
"grad_norm": 1.9818141460418701,
"learning_rate": 1.2281420147632148e-05,
"loss": 0.14907639026641845,
"step": 4380
},
{
"epoch": 0.44508655869008695,
"grad_norm": 3.5161261558532715,
"learning_rate": 1.2249439448204172e-05,
"loss": 0.13854305744171141,
"step": 4390
},
{
"epoch": 0.4461004232884698,
"grad_norm": 2.8660998344421387,
"learning_rate": 1.2217434497682537e-05,
"loss": 0.14757378101348878,
"step": 4400
},
{
"epoch": 0.4471142878868527,
"grad_norm": 5.280087471008301,
"learning_rate": 1.2185405641110997e-05,
"loss": 0.16658281087875365,
"step": 4410
},
{
"epoch": 0.4481281524852356,
"grad_norm": 3.5834171772003174,
"learning_rate": 1.2153353223791049e-05,
"loss": 0.1513921618461609,
"step": 4420
},
{
"epoch": 0.4491420170836185,
"grad_norm": 5.007085800170898,
"learning_rate": 1.2121277591278197e-05,
"loss": 0.165903902053833,
"step": 4430
},
{
"epoch": 0.4501558816820014,
"grad_norm": 3.804388999938965,
"learning_rate": 1.2089179089378222e-05,
"loss": 0.14976127147674562,
"step": 4440
},
{
"epoch": 0.45116974628038425,
"grad_norm": 4.111464977264404,
"learning_rate": 1.2057058064143456e-05,
"loss": 0.14359712600708008,
"step": 4450
},
{
"epoch": 0.45218361087876713,
"grad_norm": 5.332777976989746,
"learning_rate": 1.2024914861869064e-05,
"loss": 0.15607771873474122,
"step": 4460
},
{
"epoch": 0.45319747547715,
"grad_norm": 2.164942741394043,
"learning_rate": 1.199274982908929e-05,
"loss": 0.1565295934677124,
"step": 4470
},
{
"epoch": 0.45421134007553293,
"grad_norm": 3.0290098190307617,
"learning_rate": 1.196056331257374e-05,
"loss": 0.1428164005279541,
"step": 4480
},
{
"epoch": 0.4552252046739158,
"grad_norm": 3.8525753021240234,
"learning_rate": 1.1928355659323627e-05,
"loss": 0.15727407932281495,
"step": 4490
},
{
"epoch": 0.4562390692722987,
"grad_norm": 2.7409698963165283,
"learning_rate": 1.1896127216568047e-05,
"loss": 0.16382379531860353,
"step": 4500
},
{
"epoch": 0.45725293387068155,
"grad_norm": 3.4390361309051514,
"learning_rate": 1.1863878331760217e-05,
"loss": 0.1542404532432556,
"step": 4510
},
{
"epoch": 0.45826679846906443,
"grad_norm": 1.6817007064819336,
"learning_rate": 1.183160935257375e-05,
"loss": 0.15298254489898683,
"step": 4520
},
{
"epoch": 0.45928066306744736,
"grad_norm": 3.131983518600464,
"learning_rate": 1.1799320626898883e-05,
"loss": 0.14482526779174804,
"step": 4530
},
{
"epoch": 0.46029452766583023,
"grad_norm": 3.249912977218628,
"learning_rate": 1.1767012502838746e-05,
"loss": 0.1509859323501587,
"step": 4540
},
{
"epoch": 0.4613083922642131,
"grad_norm": 5.1811604499816895,
"learning_rate": 1.1734685328705604e-05,
"loss": 0.15916277170181276,
"step": 4550
},
{
"epoch": 0.462322256862596,
"grad_norm": 2.410710096359253,
"learning_rate": 1.17023394530171e-05,
"loss": 0.17486976385116576,
"step": 4560
},
{
"epoch": 0.4633361214609789,
"grad_norm": 3.8268446922302246,
"learning_rate": 1.1669975224492486e-05,
"loss": 0.15247933864593505,
"step": 4570
},
{
"epoch": 0.4643499860593618,
"grad_norm": 3.521911382675171,
"learning_rate": 1.1637592992048888e-05,
"loss": 0.14170806407928466,
"step": 4580
},
{
"epoch": 0.46536385065774466,
"grad_norm": 5.138008117675781,
"learning_rate": 1.160519310479753e-05,
"loss": 0.1640002727508545,
"step": 4590
},
{
"epoch": 0.46637771525612753,
"grad_norm": 1.8072806596755981,
"learning_rate": 1.157277591203996e-05,
"loss": 0.15113162994384766,
"step": 4600
},
{
"epoch": 0.4673915798545104,
"grad_norm": 3.903388261795044,
"learning_rate": 1.1540341763264314e-05,
"loss": 0.1459942102432251,
"step": 4610
},
{
"epoch": 0.46840544445289334,
"grad_norm": 2.2999534606933594,
"learning_rate": 1.1507891008141512e-05,
"loss": 0.1412009358406067,
"step": 4620
},
{
"epoch": 0.4694193090512762,
"grad_norm": 1.6942306756973267,
"learning_rate": 1.147542399652152e-05,
"loss": 0.1570139169692993,
"step": 4630
},
{
"epoch": 0.4704331736496591,
"grad_norm": 2.132309913635254,
"learning_rate": 1.1442941078429554e-05,
"loss": 0.13526663780212403,
"step": 4640
},
{
"epoch": 0.47144703824804196,
"grad_norm": 3.2470743656158447,
"learning_rate": 1.141044260406233e-05,
"loss": 0.14949955940246581,
"step": 4650
},
{
"epoch": 0.47246090284642484,
"grad_norm": 4.739569187164307,
"learning_rate": 1.1377928923784261e-05,
"loss": 0.14923642873764037,
"step": 4660
},
{
"epoch": 0.47347476744480776,
"grad_norm": 2.880248785018921,
"learning_rate": 1.1345400388123698e-05,
"loss": 0.14860136508941652,
"step": 4670
},
{
"epoch": 0.47448863204319064,
"grad_norm": 2.4135613441467285,
"learning_rate": 1.1312857347769153e-05,
"loss": 0.14632513523101806,
"step": 4680
},
{
"epoch": 0.4755024966415735,
"grad_norm": 3.3708255290985107,
"learning_rate": 1.128030015356551e-05,
"loss": 0.15129886865615844,
"step": 4690
},
{
"epoch": 0.4765163612399564,
"grad_norm": 2.5241870880126953,
"learning_rate": 1.1247729156510245e-05,
"loss": 0.1489347457885742,
"step": 4700
},
{
"epoch": 0.4775302258383393,
"grad_norm": 2.24601674079895,
"learning_rate": 1.1215144707749635e-05,
"loss": 0.14343844652175902,
"step": 4710
},
{
"epoch": 0.4785440904367222,
"grad_norm": 4.2864532470703125,
"learning_rate": 1.1182547158574988e-05,
"loss": 0.14236364364624024,
"step": 4720
},
{
"epoch": 0.47955795503510507,
"grad_norm": 2.783574342727661,
"learning_rate": 1.1149936860418846e-05,
"loss": 0.15835773944854736,
"step": 4730
},
{
"epoch": 0.48057181963348794,
"grad_norm": 1.4674508571624756,
"learning_rate": 1.1117314164851191e-05,
"loss": 0.1494053840637207,
"step": 4740
},
{
"epoch": 0.4815856842318708,
"grad_norm": 4.193670272827148,
"learning_rate": 1.1084679423575666e-05,
"loss": 0.1471541404724121,
"step": 4750
},
{
"epoch": 0.48259954883025374,
"grad_norm": 2.4419333934783936,
"learning_rate": 1.1052032988425778e-05,
"loss": 0.1352628231048584,
"step": 4760
},
{
"epoch": 0.4836134134286366,
"grad_norm": 4.713057041168213,
"learning_rate": 1.1019375211361104e-05,
"loss": 0.1254192352294922,
"step": 4770
},
{
"epoch": 0.4846272780270195,
"grad_norm": 1.5339139699935913,
"learning_rate": 1.0986706444463496e-05,
"loss": 0.1600131869316101,
"step": 4780
},
{
"epoch": 0.48564114262540237,
"grad_norm": 3.1067965030670166,
"learning_rate": 1.0954027039933288e-05,
"loss": 0.13895719051361083,
"step": 4790
},
{
"epoch": 0.48665500722378524,
"grad_norm": 6.438645839691162,
"learning_rate": 1.09213373500855e-05,
"loss": 0.15493519306182862,
"step": 4800
},
{
"epoch": 0.48766887182216817,
"grad_norm": 1.6712565422058105,
"learning_rate": 1.0888637727346032e-05,
"loss": 0.16526039838790893,
"step": 4810
},
{
"epoch": 0.48868273642055104,
"grad_norm": 4.376767635345459,
"learning_rate": 1.0855928524247884e-05,
"loss": 0.1513373613357544,
"step": 4820
},
{
"epoch": 0.4896966010189339,
"grad_norm": 2.1651878356933594,
"learning_rate": 1.0823210093427318e-05,
"loss": 0.13471097946166993,
"step": 4830
},
{
"epoch": 0.4907104656173168,
"grad_norm": 2.7860324382781982,
"learning_rate": 1.07904827876201e-05,
"loss": 0.14960209131240845,
"step": 4840
},
{
"epoch": 0.49172433021569967,
"grad_norm": 3.3760554790496826,
"learning_rate": 1.0757746959657666e-05,
"loss": 0.1512455463409424,
"step": 4850
},
{
"epoch": 0.4927381948140826,
"grad_norm": 4.436319351196289,
"learning_rate": 1.072500296246334e-05,
"loss": 0.144944429397583,
"step": 4860
},
{
"epoch": 0.49375205941246547,
"grad_norm": 5.295125484466553,
"learning_rate": 1.06922511490485e-05,
"loss": 0.14135589599609374,
"step": 4870
},
{
"epoch": 0.49476592401084835,
"grad_norm": 2.3697385787963867,
"learning_rate": 1.0659491872508807e-05,
"loss": 0.15330785512924194,
"step": 4880
},
{
"epoch": 0.4957797886092312,
"grad_norm": 5.89042854309082,
"learning_rate": 1.0626725486020374e-05,
"loss": 0.14945051670074463,
"step": 4890
},
{
"epoch": 0.49679365320761415,
"grad_norm": 3.088531970977783,
"learning_rate": 1.059395234283597e-05,
"loss": 0.1444467782974243,
"step": 4900
},
{
"epoch": 0.497807517805997,
"grad_norm": 2.7339391708374023,
"learning_rate": 1.0561172796281196e-05,
"loss": 0.15870344638824463,
"step": 4910
},
{
"epoch": 0.4988213824043799,
"grad_norm": 2.566180944442749,
"learning_rate": 1.0528387199750706e-05,
"loss": 0.1506461262702942,
"step": 4920
},
{
"epoch": 0.4998352470027628,
"grad_norm": 2.4290473461151123,
"learning_rate": 1.0495595906704365e-05,
"loss": 0.13616522550582885,
"step": 4930
},
{
"epoch": 0.5008491116011456,
"grad_norm": 3.4096388816833496,
"learning_rate": 1.0462799270663457e-05,
"loss": 0.13223049640655518,
"step": 4940
},
{
"epoch": 0.5018629761995286,
"grad_norm": 2.3158059120178223,
"learning_rate": 1.0429997645206865e-05,
"loss": 0.1350302815437317,
"step": 4950
},
{
"epoch": 0.5028768407979114,
"grad_norm": 2.9217112064361572,
"learning_rate": 1.0397191383967267e-05,
"loss": 0.13626608848571778,
"step": 4960
},
{
"epoch": 0.5038907053962943,
"grad_norm": 3.0038671493530273,
"learning_rate": 1.0364380840627315e-05,
"loss": 0.14213340282440184,
"step": 4970
},
{
"epoch": 0.5049045699946773,
"grad_norm": 3.466942548751831,
"learning_rate": 1.033156636891583e-05,
"loss": 0.16585344076156616,
"step": 4980
},
{
"epoch": 0.5059184345930601,
"grad_norm": 2.072014093399048,
"learning_rate": 1.0298748322603982e-05,
"loss": 0.14049098491668702,
"step": 4990
},
{
"epoch": 0.506932299191443,
"grad_norm": 3.4441239833831787,
"learning_rate": 1.0265927055501476e-05,
"loss": 0.15030179023742676,
"step": 5000
},
{
"epoch": 0.5079461637898258,
"grad_norm": 2.3041443824768066,
"learning_rate": 1.023310292145275e-05,
"loss": 0.13924380540847778,
"step": 5010
},
{
"epoch": 0.5089600283882088,
"grad_norm": 4.564337730407715,
"learning_rate": 1.020027627433314e-05,
"loss": 0.1440783977508545,
"step": 5020
},
{
"epoch": 0.5099738929865917,
"grad_norm": 3.7999064922332764,
"learning_rate": 1.0167447468045083e-05,
"loss": 0.14108076095581054,
"step": 5030
},
{
"epoch": 0.5109877575849745,
"grad_norm": 3.7262394428253174,
"learning_rate": 1.0134616856514282e-05,
"loss": 0.15851958990097045,
"step": 5040
},
{
"epoch": 0.5120016221833574,
"grad_norm": 3.6030428409576416,
"learning_rate": 1.0101784793685914e-05,
"loss": 0.1580324649810791,
"step": 5050
},
{
"epoch": 0.5130154867817402,
"grad_norm": 5.184695720672607,
"learning_rate": 1.0068951633520802e-05,
"loss": 0.1356940507888794,
"step": 5060
},
{
"epoch": 0.5140293513801232,
"grad_norm": 3.2134957313537598,
"learning_rate": 1.0036117729991595e-05,
"loss": 0.14258029460906982,
"step": 5070
},
{
"epoch": 0.5150432159785061,
"grad_norm": 4.425570487976074,
"learning_rate": 1.0003283437078951e-05,
"loss": 0.1530067205429077,
"step": 5080
},
{
"epoch": 0.5160570805768889,
"grad_norm": 3.3931822776794434,
"learning_rate": 9.970449108767742e-06,
"loss": 0.13903090953826905,
"step": 5090
},
{
"epoch": 0.5170709451752719,
"grad_norm": 3.8162992000579834,
"learning_rate": 9.937615099043206e-06,
"loss": 0.13431568145751954,
"step": 5100
},
{
"epoch": 0.5180848097736547,
"grad_norm": 3.35262131690979,
"learning_rate": 9.904781761887147e-06,
"loss": 0.1526703953742981,
"step": 5110
},
{
"epoch": 0.5190986743720376,
"grad_norm": 2.2346255779266357,
"learning_rate": 9.871949451274137e-06,
"loss": 0.14171113967895507,
"step": 5120
},
{
"epoch": 0.5201125389704205,
"grad_norm": 1.7881945371627808,
"learning_rate": 9.839118521167656e-06,
"loss": 0.13762218952178956,
"step": 5130
},
{
"epoch": 0.5211264035688034,
"grad_norm": 3.4133217334747314,
"learning_rate": 9.80628932551631e-06,
"loss": 0.15252923965454102,
"step": 5140
},
{
"epoch": 0.5221402681671863,
"grad_norm": 3.394439935684204,
"learning_rate": 9.773462218250014e-06,
"loss": 0.13509066104888917,
"step": 5150
},
{
"epoch": 0.5231541327655692,
"grad_norm": 4.5200629234313965,
"learning_rate": 9.740637553276162e-06,
"loss": 0.1490117311477661,
"step": 5160
},
{
"epoch": 0.524167997363952,
"grad_norm": 3.3171591758728027,
"learning_rate": 9.707815684475817e-06,
"loss": 0.1397096633911133,
"step": 5170
},
{
"epoch": 0.525181861962335,
"grad_norm": 3.434183120727539,
"learning_rate": 9.674996965699895e-06,
"loss": 0.1365661859512329,
"step": 5180
},
{
"epoch": 0.5261957265607178,
"grad_norm": 4.302313804626465,
"learning_rate": 9.642181750765354e-06,
"loss": 0.1489214539527893,
"step": 5190
},
{
"epoch": 0.5272095911591007,
"grad_norm": 2.2349677085876465,
"learning_rate": 9.60937039345138e-06,
"loss": 0.15224804878234863,
"step": 5200
},
{
"epoch": 0.5282234557574836,
"grad_norm": 3.1746420860290527,
"learning_rate": 9.576563247495566e-06,
"loss": 0.13380792140960693,
"step": 5210
},
{
"epoch": 0.5292373203558665,
"grad_norm": 2.7538368701934814,
"learning_rate": 9.543760666590105e-06,
"loss": 0.1467897653579712,
"step": 5220
},
{
"epoch": 0.5302511849542494,
"grad_norm": 2.7183942794799805,
"learning_rate": 9.510963004377974e-06,
"loss": 0.14278820753097535,
"step": 5230
},
{
"epoch": 0.5312650495526322,
"grad_norm": 2.5699782371520996,
"learning_rate": 9.478170614449114e-06,
"loss": 0.1538074493408203,
"step": 5240
},
{
"epoch": 0.5322789141510151,
"grad_norm": 2.3694381713867188,
"learning_rate": 9.445383850336648e-06,
"loss": 0.13603825569152833,
"step": 5250
},
{
"epoch": 0.5332927787493981,
"grad_norm": 3.299093246459961,
"learning_rate": 9.412603065513021e-06,
"loss": 0.14679241180419922,
"step": 5260
},
{
"epoch": 0.5343066433477809,
"grad_norm": 3.8135344982147217,
"learning_rate": 9.379828613386225e-06,
"loss": 0.15723822116851807,
"step": 5270
},
{
"epoch": 0.5353205079461638,
"grad_norm": 2.2272849082946777,
"learning_rate": 9.347060847295991e-06,
"loss": 0.1473400115966797,
"step": 5280
},
{
"epoch": 0.5363343725445466,
"grad_norm": 2.403322458267212,
"learning_rate": 9.314300120509952e-06,
"loss": 0.14557404518127443,
"step": 5290
},
{
"epoch": 0.5373482371429296,
"grad_norm": 2.209287643432617,
"learning_rate": 9.281546786219861e-06,
"loss": 0.14065542221069335,
"step": 5300
},
{
"epoch": 0.5383621017413125,
"grad_norm": 3.2864768505096436,
"learning_rate": 9.24880119753777e-06,
"loss": 0.13369591236114503,
"step": 5310
},
{
"epoch": 0.5393759663396953,
"grad_norm": 4.94459342956543,
"learning_rate": 9.216063707492218e-06,
"loss": 0.14041686058044434,
"step": 5320
},
{
"epoch": 0.5403898309380782,
"grad_norm": 2.7417054176330566,
"learning_rate": 9.183334669024449e-06,
"loss": 0.14690572023391724,
"step": 5330
},
{
"epoch": 0.5414036955364611,
"grad_norm": 2.6197097301483154,
"learning_rate": 9.150614434984578e-06,
"loss": 0.14481809139251708,
"step": 5340
},
{
"epoch": 0.542417560134844,
"grad_norm": 3.6872522830963135,
"learning_rate": 9.117903358127801e-06,
"loss": 0.13376872539520263,
"step": 5350
},
{
"epoch": 0.5434314247332269,
"grad_norm": 2.5806429386138916,
"learning_rate": 9.085201791110602e-06,
"loss": 0.1320035457611084,
"step": 5360
},
{
"epoch": 0.5444452893316097,
"grad_norm": 3.257326126098633,
"learning_rate": 9.052510086486922e-06,
"loss": 0.14122861623764038,
"step": 5370
},
{
"epoch": 0.5454591539299927,
"grad_norm": 2.5871388912200928,
"learning_rate": 9.019828596704394e-06,
"loss": 0.14401615858078004,
"step": 5380
},
{
"epoch": 0.5464730185283755,
"grad_norm": 2.724383592605591,
"learning_rate": 8.987157674100519e-06,
"loss": 0.1402200937271118,
"step": 5390
},
{
"epoch": 0.5474868831267584,
"grad_norm": 3.294365882873535,
"learning_rate": 8.954497670898858e-06,
"loss": 0.12713263034820557,
"step": 5400
},
{
"epoch": 0.5485007477251413,
"grad_norm": 3.678755760192871,
"learning_rate": 8.92184893920528e-06,
"loss": 0.12784404754638673,
"step": 5410
},
{
"epoch": 0.5495146123235242,
"grad_norm": 2.7479231357574463,
"learning_rate": 8.889211831004112e-06,
"loss": 0.13795182704925538,
"step": 5420
},
{
"epoch": 0.5505284769219071,
"grad_norm": 4.4133687019348145,
"learning_rate": 8.856586698154382e-06,
"loss": 0.13190851211547852,
"step": 5430
},
{
"epoch": 0.5515423415202899,
"grad_norm": 4.179318428039551,
"learning_rate": 8.823973892386007e-06,
"loss": 0.15881085395812988,
"step": 5440
},
{
"epoch": 0.5525562061186728,
"grad_norm": 4.134934425354004,
"learning_rate": 8.791373765296006e-06,
"loss": 0.13755034208297728,
"step": 5450
},
{
"epoch": 0.5535700707170558,
"grad_norm": 2.6915411949157715,
"learning_rate": 8.758786668344711e-06,
"loss": 0.14525880813598632,
"step": 5460
},
{
"epoch": 0.5545839353154386,
"grad_norm": 2.760758399963379,
"learning_rate": 8.726212952851978e-06,
"loss": 0.15030136108398437,
"step": 5470
},
{
"epoch": 0.5555977999138215,
"grad_norm": 2.34228777885437,
"learning_rate": 8.693652969993396e-06,
"loss": 0.13333325386047362,
"step": 5480
},
{
"epoch": 0.5566116645122043,
"grad_norm": 2.528890609741211,
"learning_rate": 8.661107070796504e-06,
"loss": 0.14899933338165283,
"step": 5490
},
{
"epoch": 0.5576255291105873,
"grad_norm": 2.833195924758911,
"learning_rate": 8.628575606137005e-06,
"loss": 0.14708199501037597,
"step": 5500
},
{
"epoch": 0.5586393937089702,
"grad_norm": 4.272514820098877,
"learning_rate": 8.59605892673499e-06,
"loss": 0.15615724325180053,
"step": 5510
},
{
"epoch": 0.559653258307353,
"grad_norm": 3.711987018585205,
"learning_rate": 8.563557383151143e-06,
"loss": 0.1450500726699829,
"step": 5520
},
{
"epoch": 0.560667122905736,
"grad_norm": 3.1560299396514893,
"learning_rate": 8.531071325782965e-06,
"loss": 0.14935673475265504,
"step": 5530
},
{
"epoch": 0.5616809875041189,
"grad_norm": 2.1339733600616455,
"learning_rate": 8.49860110486102e-06,
"loss": 0.12409404516220093,
"step": 5540
},
{
"epoch": 0.5626948521025017,
"grad_norm": 3.7151994705200195,
"learning_rate": 8.466147070445123e-06,
"loss": 0.14318242073059081,
"step": 5550
},
{
"epoch": 0.5637087167008846,
"grad_norm": 3.4385409355163574,
"learning_rate": 8.433709572420586e-06,
"loss": 0.12975592613220216,
"step": 5560
},
{
"epoch": 0.5647225812992674,
"grad_norm": 2.8556575775146484,
"learning_rate": 8.40128896049445e-06,
"loss": 0.13049578666687012,
"step": 5570
},
{
"epoch": 0.5657364458976504,
"grad_norm": 2.392927885055542,
"learning_rate": 8.368885584191698e-06,
"loss": 0.11789329051971435,
"step": 5580
},
{
"epoch": 0.5667503104960333,
"grad_norm": 2.377924680709839,
"learning_rate": 8.336499792851505e-06,
"loss": 0.13410524129867554,
"step": 5590
},
{
"epoch": 0.5677641750944161,
"grad_norm": 3.008446216583252,
"learning_rate": 8.304131935623461e-06,
"loss": 0.14395748376846312,
"step": 5600
},
{
"epoch": 0.568778039692799,
"grad_norm": 3.2626843452453613,
"learning_rate": 8.271782361463806e-06,
"loss": 0.1408052921295166,
"step": 5610
},
{
"epoch": 0.5697919042911819,
"grad_norm": 2.2048797607421875,
"learning_rate": 8.239451419131676e-06,
"loss": 0.1358433485031128,
"step": 5620
},
{
"epoch": 0.5708057688895648,
"grad_norm": 3.6624653339385986,
"learning_rate": 8.207139457185328e-06,
"loss": 0.13768112659454346,
"step": 5630
},
{
"epoch": 0.5718196334879477,
"grad_norm": 5.400641918182373,
"learning_rate": 8.174846823978412e-06,
"loss": 0.13939229249954224,
"step": 5640
},
{
"epoch": 0.5728334980863306,
"grad_norm": 2.9039156436920166,
"learning_rate": 8.142573867656182e-06,
"loss": 0.130049729347229,
"step": 5650
},
{
"epoch": 0.5738473626847135,
"grad_norm": 3.61826753616333,
"learning_rate": 8.110320936151746e-06,
"loss": 0.11815451383590699,
"step": 5660
},
{
"epoch": 0.5748612272830963,
"grad_norm": 2.7883493900299072,
"learning_rate": 8.078088377182352e-06,
"loss": 0.10910341739654542,
"step": 5670
},
{
"epoch": 0.5758750918814792,
"grad_norm": 2.577282190322876,
"learning_rate": 8.045876538245595e-06,
"loss": 0.1328972816467285,
"step": 5680
},
{
"epoch": 0.5768889564798622,
"grad_norm": 2.8212571144104004,
"learning_rate": 8.013685766615687e-06,
"loss": 0.13868985176086426,
"step": 5690
},
{
"epoch": 0.577902821078245,
"grad_norm": 2.909461498260498,
"learning_rate": 7.981516409339726e-06,
"loss": 0.14360061883926392,
"step": 5700
},
{
"epoch": 0.5789166856766279,
"grad_norm": 2.791038751602173,
"learning_rate": 7.94936881323393e-06,
"loss": 0.1358838677406311,
"step": 5710
},
{
"epoch": 0.5799305502750107,
"grad_norm": 3.5963611602783203,
"learning_rate": 7.917243324879924e-06,
"loss": 0.14439997673034669,
"step": 5720
},
{
"epoch": 0.5809444148733937,
"grad_norm": 2.552622079849243,
"learning_rate": 7.88514029062098e-06,
"loss": 0.13941943645477295,
"step": 5730
},
{
"epoch": 0.5819582794717766,
"grad_norm": 2.5385220050811768,
"learning_rate": 7.853060056558294e-06,
"loss": 0.12669514417648314,
"step": 5740
},
{
"epoch": 0.5829721440701594,
"grad_norm": 2.986750364303589,
"learning_rate": 7.821002968547266e-06,
"loss": 0.1437844753265381,
"step": 5750
},
{
"epoch": 0.5839860086685423,
"grad_norm": 2.892768144607544,
"learning_rate": 7.788969372193744e-06,
"loss": 0.12971829175949096,
"step": 5760
},
{
"epoch": 0.5849998732669252,
"grad_norm": 2.4735729694366455,
"learning_rate": 7.75695961285032e-06,
"loss": 0.12137011289596558,
"step": 5770
},
{
"epoch": 0.5860137378653081,
"grad_norm": 2.557783365249634,
"learning_rate": 7.724974035612611e-06,
"loss": 0.1294483184814453,
"step": 5780
},
{
"epoch": 0.587027602463691,
"grad_norm": 2.900696039199829,
"learning_rate": 7.693012985315503e-06,
"loss": 0.14548935890197753,
"step": 5790
},
{
"epoch": 0.5880414670620738,
"grad_norm": 4.259929180145264,
"learning_rate": 7.661076806529486e-06,
"loss": 0.12044014930725097,
"step": 5800
},
{
"epoch": 0.5890553316604568,
"grad_norm": 3.1699414253234863,
"learning_rate": 7.629165843556896e-06,
"loss": 0.1375390648841858,
"step": 5810
},
{
"epoch": 0.5900691962588396,
"grad_norm": 2.766202211380005,
"learning_rate": 7.597280440428216e-06,
"loss": 0.14123969078063964,
"step": 5820
},
{
"epoch": 0.5910830608572225,
"grad_norm": 2.802300214767456,
"learning_rate": 7.565420940898381e-06,
"loss": 0.125141441822052,
"step": 5830
},
{
"epoch": 0.5920969254556054,
"grad_norm": 2.5514183044433594,
"learning_rate": 7.5335876884430495e-06,
"loss": 0.14151922464370728,
"step": 5840
},
{
"epoch": 0.5931107900539883,
"grad_norm": 2.809499502182007,
"learning_rate": 7.501781026254922e-06,
"loss": 0.13335739374160765,
"step": 5850
},
{
"epoch": 0.5941246546523712,
"grad_norm": 3.3742685317993164,
"learning_rate": 7.4700012972400224e-06,
"loss": 0.13865501880645753,
"step": 5860
},
{
"epoch": 0.5951385192507541,
"grad_norm": 2.3332436084747314,
"learning_rate": 7.438248844014011e-06,
"loss": 0.13825689554214476,
"step": 5870
},
{
"epoch": 0.5961523838491369,
"grad_norm": 3.719376564025879,
"learning_rate": 7.4065240088984944e-06,
"loss": 0.1307378053665161,
"step": 5880
},
{
"epoch": 0.5971662484475199,
"grad_norm": 3.5361521244049072,
"learning_rate": 7.3748271339173224e-06,
"loss": 0.13629329204559326,
"step": 5890
},
{
"epoch": 0.5981801130459027,
"grad_norm": 3.9228599071502686,
"learning_rate": 7.34315856079291e-06,
"loss": 0.14779071807861327,
"step": 5900
},
{
"epoch": 0.5991939776442856,
"grad_norm": 3.36696195602417,
"learning_rate": 7.311518630942562e-06,
"loss": 0.13746120929718017,
"step": 5910
},
{
"epoch": 0.6002078422426685,
"grad_norm": 3.1879844665527344,
"learning_rate": 7.279907685474759e-06,
"loss": 0.13316624164581298,
"step": 5920
},
{
"epoch": 0.6012217068410514,
"grad_norm": 3.310270071029663,
"learning_rate": 7.248326065185527e-06,
"loss": 0.14054065942764282,
"step": 5930
},
{
"epoch": 0.6022355714394343,
"grad_norm": 2.618508815765381,
"learning_rate": 7.216774110554724e-06,
"loss": 0.12499749660491943,
"step": 5940
},
{
"epoch": 0.6032494360378171,
"grad_norm": 3.898580551147461,
"learning_rate": 7.185252161742385e-06,
"loss": 0.13291735649108888,
"step": 5950
},
{
"epoch": 0.6042633006362,
"grad_norm": 2.5865988731384277,
"learning_rate": 7.153760558585059e-06,
"loss": 0.1132235050201416,
"step": 5960
},
{
"epoch": 0.605277165234583,
"grad_norm": 4.585982799530029,
"learning_rate": 7.122299640592135e-06,
"loss": 0.1281949520111084,
"step": 5970
},
{
"epoch": 0.6062910298329658,
"grad_norm": 2.2697842121124268,
"learning_rate": 7.090869746942191e-06,
"loss": 0.12317743301391601,
"step": 5980
},
{
"epoch": 0.6073048944313487,
"grad_norm": 3.0729825496673584,
"learning_rate": 7.05947121647933e-06,
"loss": 0.12468631267547607,
"step": 5990
},
{
"epoch": 0.6083187590297315,
"grad_norm": 4.992666244506836,
"learning_rate": 7.028104387709531e-06,
"loss": 0.11746048927307129,
"step": 6000
},
{
"epoch": 0.6093326236281145,
"grad_norm": 3.509061336517334,
"learning_rate": 6.996769598797001e-06,
"loss": 0.11812496185302734,
"step": 6010
},
{
"epoch": 0.6103464882264974,
"grad_norm": 2.7003185749053955,
"learning_rate": 6.965467187560524e-06,
"loss": 0.11397761106491089,
"step": 6020
},
{
"epoch": 0.6113603528248802,
"grad_norm": 4.129496097564697,
"learning_rate": 6.934197491469818e-06,
"loss": 0.11500980854034423,
"step": 6030
},
{
"epoch": 0.6123742174232631,
"grad_norm": 2.867614269256592,
"learning_rate": 6.902960847641915e-06,
"loss": 0.13009285926818848,
"step": 6040
},
{
"epoch": 0.613388082021646,
"grad_norm": 3.505606174468994,
"learning_rate": 6.8717575928374915e-06,
"loss": 0.13544465303421022,
"step": 6050
},
{
"epoch": 0.6144019466200289,
"grad_norm": 2.7087199687957764,
"learning_rate": 6.840588063457278e-06,
"loss": 0.12485201358795166,
"step": 6060
},
{
"epoch": 0.6154158112184118,
"grad_norm": 2.781094551086426,
"learning_rate": 6.809452595538403e-06,
"loss": 0.1243922233581543,
"step": 6070
},
{
"epoch": 0.6164296758167946,
"grad_norm": 2.892763614654541,
"learning_rate": 6.778351524750777e-06,
"loss": 0.11816626787185669,
"step": 6080
},
{
"epoch": 0.6174435404151776,
"grad_norm": 2.8575310707092285,
"learning_rate": 6.747285186393488e-06,
"loss": 0.12620100975036622,
"step": 6090
},
{
"epoch": 0.6184574050135604,
"grad_norm": 7.022664546966553,
"learning_rate": 6.716253915391167e-06,
"loss": 0.13958804607391356,
"step": 6100
},
{
"epoch": 0.6194712696119433,
"grad_norm": 3.6078855991363525,
"learning_rate": 6.685258046290381e-06,
"loss": 0.1351696729660034,
"step": 6110
},
{
"epoch": 0.6204851342103263,
"grad_norm": 3.859738826751709,
"learning_rate": 6.654297913256047e-06,
"loss": 0.141652512550354,
"step": 6120
},
{
"epoch": 0.6214989988087091,
"grad_norm": 3.00361967086792,
"learning_rate": 6.623373850067794e-06,
"loss": 0.14661519527435302,
"step": 6130
},
{
"epoch": 0.622512863407092,
"grad_norm": 4.262936592102051,
"learning_rate": 6.592486190116405e-06,
"loss": 0.13541691303253173,
"step": 6140
},
{
"epoch": 0.6235267280054748,
"grad_norm": 4.5068230628967285,
"learning_rate": 6.561635266400181e-06,
"loss": 0.12077987194061279,
"step": 6150
},
{
"epoch": 0.6245405926038577,
"grad_norm": 4.606220722198486,
"learning_rate": 6.5308214115213785e-06,
"loss": 0.13935673236846924,
"step": 6160
},
{
"epoch": 0.6255544572022407,
"grad_norm": 2.837034225463867,
"learning_rate": 6.500044957682627e-06,
"loss": 0.13247497081756593,
"step": 6170
},
{
"epoch": 0.6265683218006235,
"grad_norm": 2.9637601375579834,
"learning_rate": 6.469306236683324e-06,
"loss": 0.11624768972396851,
"step": 6180
},
{
"epoch": 0.6275821863990064,
"grad_norm": 3.320528030395508,
"learning_rate": 6.438605579916078e-06,
"loss": 0.1321529746055603,
"step": 6190
},
{
"epoch": 0.6285960509973892,
"grad_norm": 5.34397554397583,
"learning_rate": 6.4079433183631255e-06,
"loss": 0.1313384771347046,
"step": 6200
},
{
"epoch": 0.6296099155957722,
"grad_norm": 2.6070477962493896,
"learning_rate": 6.377319782592767e-06,
"loss": 0.12336071729660034,
"step": 6210
},
{
"epoch": 0.6306237801941551,
"grad_norm": 4.037292003631592,
"learning_rate": 6.346735302755805e-06,
"loss": 0.1351922035217285,
"step": 6220
},
{
"epoch": 0.6316376447925379,
"grad_norm": 4.168092250823975,
"learning_rate": 6.316190208581979e-06,
"loss": 0.12283761501312256,
"step": 6230
},
{
"epoch": 0.6326515093909209,
"grad_norm": 2.9784107208251953,
"learning_rate": 6.285684829376413e-06,
"loss": 0.12672253847122192,
"step": 6240
},
{
"epoch": 0.6336653739893038,
"grad_norm": 4.346795082092285,
"learning_rate": 6.25521949401607e-06,
"loss": 0.11864570379257203,
"step": 6250
},
{
"epoch": 0.6346792385876866,
"grad_norm": 3.0507373809814453,
"learning_rate": 6.2247945309461955e-06,
"loss": 0.13206984996795654,
"step": 6260
},
{
"epoch": 0.6356931031860695,
"grad_norm": 4.3934760093688965,
"learning_rate": 6.194410268176796e-06,
"loss": 0.13272039890289306,
"step": 6270
},
{
"epoch": 0.6367069677844523,
"grad_norm": 2.8900094032287598,
"learning_rate": 6.164067033279073e-06,
"loss": 0.1204066514968872,
"step": 6280
},
{
"epoch": 0.6377208323828353,
"grad_norm": 2.5006015300750732,
"learning_rate": 6.133765153381918e-06,
"loss": 0.1347909927368164,
"step": 6290
},
{
"epoch": 0.6387346969812182,
"grad_norm": 3.596600294113159,
"learning_rate": 6.103504955168382e-06,
"loss": 0.11307622194290161,
"step": 6300
},
{
"epoch": 0.639748561579601,
"grad_norm": 5.133296489715576,
"learning_rate": 6.07328676487214e-06,
"loss": 0.13116352558135985,
"step": 6310
},
{
"epoch": 0.640762426177984,
"grad_norm": 3.4658093452453613,
"learning_rate": 6.043110908273973e-06,
"loss": 0.12299206256866455,
"step": 6320
},
{
"epoch": 0.6417762907763668,
"grad_norm": 5.828248977661133,
"learning_rate": 6.012977710698288e-06,
"loss": 0.11659754514694214,
"step": 6330
},
{
"epoch": 0.6427901553747497,
"grad_norm": 5.436083793640137,
"learning_rate": 5.982887497009566e-06,
"loss": 0.12681375741958617,
"step": 6340
},
{
"epoch": 0.6438040199731326,
"grad_norm": 3.777198076248169,
"learning_rate": 5.9528405916088926e-06,
"loss": 0.11744120121002197,
"step": 6350
},
{
"epoch": 0.6448178845715155,
"grad_norm": 4.612187385559082,
"learning_rate": 5.9228373184304455e-06,
"loss": 0.13033145666122437,
"step": 6360
},
{
"epoch": 0.6458317491698984,
"grad_norm": 4.606093883514404,
"learning_rate": 5.892878000938001e-06,
"loss": 0.1187903642654419,
"step": 6370
},
{
"epoch": 0.6468456137682812,
"grad_norm": 3.490325689315796,
"learning_rate": 5.86296296212146e-06,
"loss": 0.12015480995178222,
"step": 6380
},
{
"epoch": 0.6478594783666641,
"grad_norm": 3.197228193283081,
"learning_rate": 5.833092524493349e-06,
"loss": 0.11876649856567383,
"step": 6390
},
{
"epoch": 0.6488733429650471,
"grad_norm": 4.476126194000244,
"learning_rate": 5.80326701008536e-06,
"loss": 0.12489039897918701,
"step": 6400
},
{
"epoch": 0.6498872075634299,
"grad_norm": 3.1946539878845215,
"learning_rate": 5.773486740444862e-06,
"loss": 0.11144413948059081,
"step": 6410
},
{
"epoch": 0.6509010721618128,
"grad_norm": 3.3871748447418213,
"learning_rate": 5.743752036631443e-06,
"loss": 0.12645292282104492,
"step": 6420
},
{
"epoch": 0.6519149367601956,
"grad_norm": 4.075658798217773,
"learning_rate": 5.714063219213456e-06,
"loss": 0.10481016635894776,
"step": 6430
},
{
"epoch": 0.6529288013585786,
"grad_norm": 3.433931827545166,
"learning_rate": 5.684420608264549e-06,
"loss": 0.12456264495849609,
"step": 6440
},
{
"epoch": 0.6539426659569615,
"grad_norm": 3.1976213455200195,
"learning_rate": 5.654824523360219e-06,
"loss": 0.11406522989273071,
"step": 6450
},
{
"epoch": 0.6549565305553443,
"grad_norm": 2.904604434967041,
"learning_rate": 5.625275283574372e-06,
"loss": 0.11348844766616821,
"step": 6460
},
{
"epoch": 0.6559703951537272,
"grad_norm": 3.669755220413208,
"learning_rate": 5.595773207475872e-06,
"loss": 0.11949663162231446,
"step": 6470
},
{
"epoch": 0.6569842597521101,
"grad_norm": 3.847105026245117,
"learning_rate": 5.566318613125128e-06,
"loss": 0.12017428874969482,
"step": 6480
},
{
"epoch": 0.657998124350493,
"grad_norm": 3.702448844909668,
"learning_rate": 5.536911818070636e-06,
"loss": 0.11860222816467285,
"step": 6490
},
{
"epoch": 0.6590119889488759,
"grad_norm": 3.7512221336364746,
"learning_rate": 5.507553139345574e-06,
"loss": 0.11664795875549316,
"step": 6500
},
{
"epoch": 0.6600258535472587,
"grad_norm": 3.946676015853882,
"learning_rate": 5.478242893464389e-06,
"loss": 0.1167741060256958,
"step": 6510
},
{
"epoch": 0.6610397181456417,
"grad_norm": 4.085474014282227,
"learning_rate": 5.448981396419366e-06,
"loss": 0.11449713706970215,
"step": 6520
},
{
"epoch": 0.6620535827440245,
"grad_norm": 3.5946624279022217,
"learning_rate": 5.419768963677233e-06,
"loss": 0.11321868896484374,
"step": 6530
},
{
"epoch": 0.6630674473424074,
"grad_norm": 4.105451583862305,
"learning_rate": 5.39060591017576e-06,
"loss": 0.12770211696624756,
"step": 6540
},
{
"epoch": 0.6640813119407903,
"grad_norm": 2.6972200870513916,
"learning_rate": 5.3614925503203586e-06,
"loss": 0.11952120065689087,
"step": 6550
},
{
"epoch": 0.6650951765391732,
"grad_norm": 4.250665664672852,
"learning_rate": 5.332429197980706e-06,
"loss": 0.1105314016342163,
"step": 6560
},
{
"epoch": 0.6661090411375561,
"grad_norm": 4.0521345138549805,
"learning_rate": 5.303416166487337e-06,
"loss": 0.12324638366699218,
"step": 6570
},
{
"epoch": 0.667122905735939,
"grad_norm": 5.877462863922119,
"learning_rate": 5.274453768628289e-06,
"loss": 0.12558765411376954,
"step": 6580
},
{
"epoch": 0.6681367703343218,
"grad_norm": 3.4792187213897705,
"learning_rate": 5.245542316645718e-06,
"loss": 0.11361993551254272,
"step": 6590
},
{
"epoch": 0.6691506349327048,
"grad_norm": 4.954648971557617,
"learning_rate": 5.2166821222325306e-06,
"loss": 0.10373766422271728,
"step": 6600
},
{
"epoch": 0.6701644995310876,
"grad_norm": 3.740058422088623,
"learning_rate": 5.187873496529041e-06,
"loss": 0.10950932502746583,
"step": 6610
},
{
"epoch": 0.6711783641294705,
"grad_norm": 4.4966583251953125,
"learning_rate": 5.15911675011959e-06,
"loss": 0.13223862648010254,
"step": 6620
},
{
"epoch": 0.6721922287278534,
"grad_norm": 2.6593053340911865,
"learning_rate": 5.130412193029213e-06,
"loss": 0.11176964044570922,
"step": 6630
},
{
"epoch": 0.6732060933262363,
"grad_norm": 2.721031665802002,
"learning_rate": 5.101760134720306e-06,
"loss": 0.11554380655288696,
"step": 6640
},
{
"epoch": 0.6742199579246192,
"grad_norm": 4.944410800933838,
"learning_rate": 5.073160884089263e-06,
"loss": 0.11901267766952514,
"step": 6650
},
{
"epoch": 0.675233822523002,
"grad_norm": 3.5725066661834717,
"learning_rate": 5.044614749463173e-06,
"loss": 0.1191360592842102,
"step": 6660
},
{
"epoch": 0.6762476871213849,
"grad_norm": 2.646462917327881,
"learning_rate": 5.0161220385964745e-06,
"loss": 0.10639078617095947,
"step": 6670
},
{
"epoch": 0.6772615517197679,
"grad_norm": 3.3566458225250244,
"learning_rate": 4.987683058667651e-06,
"loss": 0.11876592636108399,
"step": 6680
},
{
"epoch": 0.6782754163181507,
"grad_norm": 5.15422248840332,
"learning_rate": 4.9592981162759215e-06,
"loss": 0.11660023927688598,
"step": 6690
},
{
"epoch": 0.6792892809165336,
"grad_norm": 3.5057363510131836,
"learning_rate": 4.930967517437922e-06,
"loss": 0.10649501085281372,
"step": 6700
},
{
"epoch": 0.6803031455149164,
"grad_norm": 5.892392635345459,
"learning_rate": 4.902691567584412e-06,
"loss": 0.11844031810760498,
"step": 6710
},
{
"epoch": 0.6813170101132994,
"grad_norm": 3.7611618041992188,
"learning_rate": 4.87447057155699e-06,
"loss": 0.12127432823181153,
"step": 6720
},
{
"epoch": 0.6823308747116823,
"grad_norm": 4.29107666015625,
"learning_rate": 4.846304833604792e-06,
"loss": 0.11328608989715576,
"step": 6730
},
{
"epoch": 0.6833447393100651,
"grad_norm": 6.6301069259643555,
"learning_rate": 4.8181946573812236e-06,
"loss": 0.12810261249542237,
"step": 6740
},
{
"epoch": 0.684358603908448,
"grad_norm": 2.8485732078552246,
"learning_rate": 4.790140345940688e-06,
"loss": 0.10687030553817749,
"step": 6750
},
{
"epoch": 0.6853724685068309,
"grad_norm": 4.216344356536865,
"learning_rate": 4.762142201735299e-06,
"loss": 0.11293613910675049,
"step": 6760
},
{
"epoch": 0.6863863331052138,
"grad_norm": 4.134664535522461,
"learning_rate": 4.734200526611652e-06,
"loss": 0.10741066932678223,
"step": 6770
},
{
"epoch": 0.6874001977035967,
"grad_norm": 3.8215954303741455,
"learning_rate": 4.706315621807537e-06,
"loss": 0.11004934310913086,
"step": 6780
},
{
"epoch": 0.6884140623019795,
"grad_norm": 3.0624501705169678,
"learning_rate": 4.678487787948711e-06,
"loss": 0.11001648902893066,
"step": 6790
},
{
"epoch": 0.6894279269003625,
"grad_norm": 5.591687202453613,
"learning_rate": 4.650717325045655e-06,
"loss": 0.10570991039276123,
"step": 6800
},
{
"epoch": 0.6904417914987453,
"grad_norm": 3.964669942855835,
"learning_rate": 4.623004532490328e-06,
"loss": 0.11600648164749146,
"step": 6810
},
{
"epoch": 0.6914556560971282,
"grad_norm": 5.378198623657227,
"learning_rate": 4.595349709052964e-06,
"loss": 0.1134990930557251,
"step": 6820
},
{
"epoch": 0.6924695206955112,
"grad_norm": 4.373196125030518,
"learning_rate": 4.5677531528788196e-06,
"loss": 0.10503398180007935,
"step": 6830
},
{
"epoch": 0.693483385293894,
"grad_norm": 4.3774027824401855,
"learning_rate": 4.54021516148498e-06,
"loss": 0.11385728120803833,
"step": 6840
},
{
"epoch": 0.6944972498922769,
"grad_norm": 3.3578760623931885,
"learning_rate": 4.5127360317571475e-06,
"loss": 0.11773920059204102,
"step": 6850
},
{
"epoch": 0.6955111144906597,
"grad_norm": 3.661301851272583,
"learning_rate": 4.485316059946438e-06,
"loss": 0.10000908374786377,
"step": 6860
},
{
"epoch": 0.6965249790890427,
"grad_norm": 4.847537517547607,
"learning_rate": 4.457955541666188e-06,
"loss": 0.1177248477935791,
"step": 6870
},
{
"epoch": 0.6975388436874256,
"grad_norm": 3.018498659133911,
"learning_rate": 4.430654771888776e-06,
"loss": 0.0989755392074585,
"step": 6880
},
{
"epoch": 0.6985527082858084,
"grad_norm": 5.179007530212402,
"learning_rate": 4.403414044942422e-06,
"loss": 0.10907419919967651,
"step": 6890
},
{
"epoch": 0.6995665728841913,
"grad_norm": 4.337184429168701,
"learning_rate": 4.376233654508042e-06,
"loss": 0.10114206075668335,
"step": 6900
},
{
"epoch": 0.7005804374825741,
"grad_norm": 3.712588310241699,
"learning_rate": 4.349113893616056e-06,
"loss": 0.10332232713699341,
"step": 6910
},
{
"epoch": 0.7015943020809571,
"grad_norm": 5.105023384094238,
"learning_rate": 4.322055054643244e-06,
"loss": 0.09234496355056762,
"step": 6920
},
{
"epoch": 0.70260816667934,
"grad_norm": 5.147826194763184,
"learning_rate": 4.295057429309589e-06,
"loss": 0.10918929576873779,
"step": 6930
},
{
"epoch": 0.7036220312777228,
"grad_norm": 4.142611980438232,
"learning_rate": 4.268121308675132e-06,
"loss": 0.10097410678863525,
"step": 6940
},
{
"epoch": 0.7046358958761058,
"grad_norm": 3.674997329711914,
"learning_rate": 4.241246983136834e-06,
"loss": 0.10447102785110474,
"step": 6950
},
{
"epoch": 0.7056497604744887,
"grad_norm": 3.7555463314056396,
"learning_rate": 4.214434742425455e-06,
"loss": 0.09983371496200562,
"step": 6960
},
{
"epoch": 0.7066636250728715,
"grad_norm": 4.30908203125,
"learning_rate": 4.1876848756024116e-06,
"loss": 0.10768011808395386,
"step": 6970
},
{
"epoch": 0.7076774896712544,
"grad_norm": 2.8958258628845215,
"learning_rate": 4.1609976710566735e-06,
"loss": 0.12507684230804444,
"step": 6980
},
{
"epoch": 0.7086913542696373,
"grad_norm": 2.7451107501983643,
"learning_rate": 4.134373416501652e-06,
"loss": 0.11008610725402831,
"step": 6990
},
{
"epoch": 0.7097052188680202,
"grad_norm": 3.5242152214050293,
"learning_rate": 4.107812398972093e-06,
"loss": 0.11189334392547608,
"step": 7000
},
{
"epoch": 0.7107190834664031,
"grad_norm": 4.062826633453369,
"learning_rate": 4.081314904821001e-06,
"loss": 0.10240273475646973,
"step": 7010
},
{
"epoch": 0.7117329480647859,
"grad_norm": 4.270602226257324,
"learning_rate": 4.054881219716521e-06,
"loss": 0.10707014799118042,
"step": 7020
},
{
"epoch": 0.7127468126631689,
"grad_norm": 5.456202507019043,
"learning_rate": 4.0285116286388914e-06,
"loss": 0.10205912590026855,
"step": 7030
},
{
"epoch": 0.7137606772615517,
"grad_norm": 5.391589641571045,
"learning_rate": 4.0022064158773445e-06,
"loss": 0.12334390878677368,
"step": 7040
},
{
"epoch": 0.7147745418599346,
"grad_norm": 3.670999526977539,
"learning_rate": 3.9759658650270575e-06,
"loss": 0.11824101209640503,
"step": 7050
},
{
"epoch": 0.7157884064583175,
"grad_norm": 3.4340388774871826,
"learning_rate": 3.94979025898609e-06,
"loss": 0.10172815322875976,
"step": 7060
},
{
"epoch": 0.7168022710567004,
"grad_norm": 4.303616523742676,
"learning_rate": 3.9236798799523375e-06,
"loss": 0.10820946693420411,
"step": 7070
},
{
"epoch": 0.7178161356550833,
"grad_norm": 3.9502720832824707,
"learning_rate": 3.89763500942048e-06,
"loss": 0.11918987035751342,
"step": 7080
},
{
"epoch": 0.7188300002534661,
"grad_norm": 3.262660264968872,
"learning_rate": 3.871655928178962e-06,
"loss": 0.09720779061317444,
"step": 7090
},
{
"epoch": 0.719843864851849,
"grad_norm": 3.3819963932037354,
"learning_rate": 3.8457429163069534e-06,
"loss": 0.11923363208770751,
"step": 7100
},
{
"epoch": 0.720857729450232,
"grad_norm": 3.9469754695892334,
"learning_rate": 3.819896253171332e-06,
"loss": 0.10616371631622315,
"step": 7110
},
{
"epoch": 0.7218715940486148,
"grad_norm": 3.27604603767395,
"learning_rate": 3.7941162174236745e-06,
"loss": 0.0910140335559845,
"step": 7120
},
{
"epoch": 0.7228854586469977,
"grad_norm": 4.393930435180664,
"learning_rate": 3.7684030869972464e-06,
"loss": 0.09610801935195923,
"step": 7130
},
{
"epoch": 0.7238993232453805,
"grad_norm": 4.8421454429626465,
"learning_rate": 3.742757139104023e-06,
"loss": 0.11481420993804932,
"step": 7140
},
{
"epoch": 0.7249131878437635,
"grad_norm": 4.77382755279541,
"learning_rate": 3.7171786502316767e-06,
"loss": 0.10762162208557129,
"step": 7150
},
{
"epoch": 0.7259270524421464,
"grad_norm": 2.5361037254333496,
"learning_rate": 3.691667896140605e-06,
"loss": 0.09731899499893189,
"step": 7160
},
{
"epoch": 0.7269409170405292,
"grad_norm": 3.3651986122131348,
"learning_rate": 3.6662251518609747e-06,
"loss": 0.09513275623321533,
"step": 7170
},
{
"epoch": 0.7279547816389121,
"grad_norm": 3.8299925327301025,
"learning_rate": 3.64085069168973e-06,
"loss": 0.10830851793289184,
"step": 7180
},
{
"epoch": 0.728968646237295,
"grad_norm": 3.8895466327667236,
"learning_rate": 3.6155447891876518e-06,
"loss": 0.08921222686767578,
"step": 7190
},
{
"epoch": 0.7299825108356779,
"grad_norm": 3.592860221862793,
"learning_rate": 3.590307717176401e-06,
"loss": 0.08990630507469177,
"step": 7200
},
{
"epoch": 0.7309963754340608,
"grad_norm": 5.50272274017334,
"learning_rate": 3.565139747735581e-06,
"loss": 0.0927023470401764,
"step": 7210
},
{
"epoch": 0.7320102400324436,
"grad_norm": 4.717070579528809,
"learning_rate": 3.5400411521998123e-06,
"loss": 0.07454866170883179,
"step": 7220
},
{
"epoch": 0.7330241046308266,
"grad_norm": 4.806131362915039,
"learning_rate": 3.5150122011557888e-06,
"loss": 0.09940525889396667,
"step": 7230
},
{
"epoch": 0.7340379692292094,
"grad_norm": 4.338921070098877,
"learning_rate": 3.4900531644393744e-06,
"loss": 0.10023744106292724,
"step": 7240
},
{
"epoch": 0.7350518338275923,
"grad_norm": 4.475937366485596,
"learning_rate": 3.4651643111326895e-06,
"loss": 0.10950019359588622,
"step": 7250
},
{
"epoch": 0.7360656984259752,
"grad_norm": 3.646148681640625,
"learning_rate": 3.440345909561208e-06,
"loss": 0.0849729061126709,
"step": 7260
},
{
"epoch": 0.7370795630243581,
"grad_norm": 2.224591016769409,
"learning_rate": 3.415598227290877e-06,
"loss": 0.09727665781974792,
"step": 7270
},
{
"epoch": 0.738093427622741,
"grad_norm": 4.503518104553223,
"learning_rate": 3.390921531125211e-06,
"loss": 0.09071973562240601,
"step": 7280
},
{
"epoch": 0.7391072922211239,
"grad_norm": 4.392312049865723,
"learning_rate": 3.366316087102429e-06,
"loss": 0.08487353324890137,
"step": 7290
},
{
"epoch": 0.7401211568195067,
"grad_norm": 4.50674295425415,
"learning_rate": 3.3417821604925937e-06,
"loss": 0.096976900100708,
"step": 7300
},
{
"epoch": 0.7411350214178897,
"grad_norm": 4.770694255828857,
"learning_rate": 3.3173200157947315e-06,
"loss": 0.10890278816223145,
"step": 7310
},
{
"epoch": 0.7421488860162725,
"grad_norm": 4.1736907958984375,
"learning_rate": 3.2929299167339955e-06,
"loss": 0.09597418308258057,
"step": 7320
},
{
"epoch": 0.7431627506146554,
"grad_norm": 3.5154895782470703,
"learning_rate": 3.2686121262588165e-06,
"loss": 0.10536104440689087,
"step": 7330
},
{
"epoch": 0.7441766152130384,
"grad_norm": 3.73414945602417,
"learning_rate": 3.244366906538071e-06,
"loss": 0.0970649242401123,
"step": 7340
},
{
"epoch": 0.7451904798114212,
"grad_norm": 2.867741823196411,
"learning_rate": 3.2201945189582597e-06,
"loss": 0.09682568311691284,
"step": 7350
},
{
"epoch": 0.7462043444098041,
"grad_norm": 3.9812943935394287,
"learning_rate": 3.1960952241206745e-06,
"loss": 0.08977947235107422,
"step": 7360
},
{
"epoch": 0.7472182090081869,
"grad_norm": 6.433146953582764,
"learning_rate": 3.172069281838603e-06,
"loss": 0.10429489612579346,
"step": 7370
},
{
"epoch": 0.7482320736065698,
"grad_norm": 6.464941024780273,
"learning_rate": 3.148116951134522e-06,
"loss": 0.09536977410316468,
"step": 7380
},
{
"epoch": 0.7492459382049528,
"grad_norm": 3.15676212310791,
"learning_rate": 3.1242384902373024e-06,
"loss": 0.0981927216053009,
"step": 7390
},
{
"epoch": 0.7502598028033356,
"grad_norm": 4.065806865692139,
"learning_rate": 3.1004341565794373e-06,
"loss": 0.10315895080566406,
"step": 7400
},
{
"epoch": 0.7512736674017185,
"grad_norm": 3.229402780532837,
"learning_rate": 3.0767042067942475e-06,
"loss": 0.09355083107948303,
"step": 7410
},
{
"epoch": 0.7522875320001013,
"grad_norm": 5.0950093269348145,
"learning_rate": 3.0530488967131265e-06,
"loss": 0.0931445300579071,
"step": 7420
},
{
"epoch": 0.7533013965984843,
"grad_norm": 3.6790003776550293,
"learning_rate": 3.0294684813627917e-06,
"loss": 0.0888776421546936,
"step": 7430
},
{
"epoch": 0.7543152611968672,
"grad_norm": 5.197442531585693,
"learning_rate": 3.0059632149625118e-06,
"loss": 0.08658597469329835,
"step": 7440
},
{
"epoch": 0.75532912579525,
"grad_norm": 4.747237205505371,
"learning_rate": 2.982533350921383e-06,
"loss": 0.097781503200531,
"step": 7450
},
{
"epoch": 0.756342990393633,
"grad_norm": 3.112135171890259,
"learning_rate": 2.959179141835591e-06,
"loss": 0.08117998242378235,
"step": 7460
},
{
"epoch": 0.7573568549920158,
"grad_norm": 3.6744494438171387,
"learning_rate": 2.935900839485688e-06,
"loss": 0.09840376973152161,
"step": 7470
},
{
"epoch": 0.7583707195903987,
"grad_norm": 3.910296678543091,
"learning_rate": 2.9126986948338898e-06,
"loss": 0.08016796112060547,
"step": 7480
},
{
"epoch": 0.7593845841887816,
"grad_norm": 2.759836435317993,
"learning_rate": 2.889572958021347e-06,
"loss": 0.09958692789077758,
"step": 7490
},
{
"epoch": 0.7603984487871644,
"grad_norm": 3.206357955932617,
"learning_rate": 2.866523878365468e-06,
"loss": 0.08789313435554505,
"step": 7500
},
{
"epoch": 0.7614123133855474,
"grad_norm": 3.8137083053588867,
"learning_rate": 2.8435517043572204e-06,
"loss": 0.08735495805740356,
"step": 7510
},
{
"epoch": 0.7624261779839302,
"grad_norm": 4.545304775238037,
"learning_rate": 2.820656683658456e-06,
"loss": 0.10341571569442749,
"step": 7520
},
{
"epoch": 0.7634400425823131,
"grad_norm": 3.892998456954956,
"learning_rate": 2.797839063099249e-06,
"loss": 0.09774882793426513,
"step": 7530
},
{
"epoch": 0.7644539071806961,
"grad_norm": 3.156146287918091,
"learning_rate": 2.775099088675216e-06,
"loss": 0.09782047271728515,
"step": 7540
},
{
"epoch": 0.7654677717790789,
"grad_norm": 4.6163330078125,
"learning_rate": 2.7524370055448755e-06,
"loss": 0.08794929385185242,
"step": 7550
},
{
"epoch": 0.7664816363774618,
"grad_norm": 3.6701364517211914,
"learning_rate": 2.7298530580270154e-06,
"loss": 0.0889196515083313,
"step": 7560
},
{
"epoch": 0.7674955009758446,
"grad_norm": 4.123426914215088,
"learning_rate": 2.7073474895980333e-06,
"loss": 0.08840047121047974,
"step": 7570
},
{
"epoch": 0.7685093655742276,
"grad_norm": 5.154639720916748,
"learning_rate": 2.6849205428893334e-06,
"loss": 0.08398475646972656,
"step": 7580
},
{
"epoch": 0.7695232301726105,
"grad_norm": 3.3856852054595947,
"learning_rate": 2.662572459684699e-06,
"loss": 0.08319178819656373,
"step": 7590
},
{
"epoch": 0.7705370947709933,
"grad_norm": 3.293200731277466,
"learning_rate": 2.640303480917693e-06,
"loss": 0.09307070970535278,
"step": 7600
},
{
"epoch": 0.7715509593693762,
"grad_norm": 4.123002529144287,
"learning_rate": 2.618113846669059e-06,
"loss": 0.09602808952331543,
"step": 7610
},
{
"epoch": 0.772564823967759,
"grad_norm": 5.133830547332764,
"learning_rate": 2.596003796164127e-06,
"loss": 0.08914123177528381,
"step": 7620
},
{
"epoch": 0.773578688566142,
"grad_norm": 4.343494892120361,
"learning_rate": 2.5739735677702416e-06,
"loss": 0.08208319544792175,
"step": 7630
},
{
"epoch": 0.7745925531645249,
"grad_norm": 4.934475898742676,
"learning_rate": 2.552023398994188e-06,
"loss": 0.09094523787498474,
"step": 7640
},
{
"epoch": 0.7756064177629077,
"grad_norm": 5.128413677215576,
"learning_rate": 2.5301535264796274e-06,
"loss": 0.09215626120567322,
"step": 7650
},
{
"epoch": 0.7766202823612907,
"grad_norm": 5.665582180023193,
"learning_rate": 2.508364186004566e-06,
"loss": 0.09432408213615417,
"step": 7660
},
{
"epoch": 0.7776341469596736,
"grad_norm": 4.498518466949463,
"learning_rate": 2.486655612478782e-06,
"loss": 0.08374996781349182,
"step": 7670
},
{
"epoch": 0.7786480115580564,
"grad_norm": 3.410818576812744,
"learning_rate": 2.465028039941316e-06,
"loss": 0.06973189115524292,
"step": 7680
},
{
"epoch": 0.7796618761564393,
"grad_norm": 4.888281345367432,
"learning_rate": 2.443481701557946e-06,
"loss": 0.08796026706695556,
"step": 7690
},
{
"epoch": 0.7806757407548222,
"grad_norm": 4.43965482711792,
"learning_rate": 2.4220168296186587e-06,
"loss": 0.08294991850852966,
"step": 7700
},
{
"epoch": 0.7816896053532051,
"grad_norm": 4.834952354431152,
"learning_rate": 2.4006336555351615e-06,
"loss": 0.0936345100402832,
"step": 7710
},
{
"epoch": 0.782703469951588,
"grad_norm": 4.823123455047607,
"learning_rate": 2.3793324098383796e-06,
"loss": 0.08936270475387573,
"step": 7720
},
{
"epoch": 0.7837173345499708,
"grad_norm": 4.847897052764893,
"learning_rate": 2.35811332217597e-06,
"loss": 0.07355624437332153,
"step": 7730
},
{
"epoch": 0.7847311991483538,
"grad_norm": 4.50162410736084,
"learning_rate": 2.3369766213098553e-06,
"loss": 0.0831048846244812,
"step": 7740
},
{
"epoch": 0.7857450637467366,
"grad_norm": 3.983522415161133,
"learning_rate": 2.315922535113743e-06,
"loss": 0.08037481307983399,
"step": 7750
},
{
"epoch": 0.7867589283451195,
"grad_norm": 5.212425708770752,
"learning_rate": 2.2949512905706793e-06,
"loss": 0.08474689126014709,
"step": 7760
},
{
"epoch": 0.7877727929435024,
"grad_norm": 6.453623294830322,
"learning_rate": 2.274063113770597e-06,
"loss": 0.0800774335861206,
"step": 7770
},
{
"epoch": 0.7887866575418853,
"grad_norm": 3.974081516265869,
"learning_rate": 2.2532582299078775e-06,
"loss": 0.086143958568573,
"step": 7780
},
{
"epoch": 0.7898005221402682,
"grad_norm": 3.8404781818389893,
"learning_rate": 2.2325368632789345e-06,
"loss": 0.08658806681632995,
"step": 7790
},
{
"epoch": 0.790814386738651,
"grad_norm": 3.7893567085266113,
"learning_rate": 2.211899237279779e-06,
"loss": 0.0851547122001648,
"step": 7800
},
{
"epoch": 0.7918282513370339,
"grad_norm": 4.060662746429443,
"learning_rate": 2.1913455744036138e-06,
"loss": 0.1167557954788208,
"step": 7810
},
{
"epoch": 0.7928421159354169,
"grad_norm": 3.691756010055542,
"learning_rate": 2.1708760962384545e-06,
"loss": 0.09559301137924195,
"step": 7820
},
{
"epoch": 0.7938559805337997,
"grad_norm": 4.286520481109619,
"learning_rate": 2.1504910234647112e-06,
"loss": 0.0759650707244873,
"step": 7830
},
{
"epoch": 0.7948698451321826,
"grad_norm": 4.335417747497559,
"learning_rate": 2.130190575852832e-06,
"loss": 0.07918434143066407,
"step": 7840
},
{
"epoch": 0.7958837097305654,
"grad_norm": 4.144512176513672,
"learning_rate": 2.109974972260921e-06,
"loss": 0.07134771347045898,
"step": 7850
},
{
"epoch": 0.7968975743289484,
"grad_norm": 5.814027786254883,
"learning_rate": 2.0898444306323807e-06,
"loss": 0.0805224359035492,
"step": 7860
},
{
"epoch": 0.7979114389273313,
"grad_norm": 3.764388084411621,
"learning_rate": 2.069799167993577e-06,
"loss": 0.08970481157302856,
"step": 7870
},
{
"epoch": 0.7989253035257141,
"grad_norm": 5.533236980438232,
"learning_rate": 2.0498394004514745e-06,
"loss": 0.08222857713699341,
"step": 7880
},
{
"epoch": 0.799939168124097,
"grad_norm": 3.9165287017822266,
"learning_rate": 2.029965343191326e-06,
"loss": 0.09543319940567016,
"step": 7890
},
{
"epoch": 0.8009530327224799,
"grad_norm": 5.660103797912598,
"learning_rate": 2.0101772104743422e-06,
"loss": 0.09225713610649108,
"step": 7900
},
{
"epoch": 0.8019668973208628,
"grad_norm": 4.209702014923096,
"learning_rate": 1.990475215635388e-06,
"loss": 0.074017333984375,
"step": 7910
},
{
"epoch": 0.8029807619192457,
"grad_norm": 4.876410484313965,
"learning_rate": 1.9708595710806787e-06,
"loss": 0.07661293148994446,
"step": 7920
},
{
"epoch": 0.8039946265176285,
"grad_norm": 5.210673809051514,
"learning_rate": 1.9513304882854966e-06,
"loss": 0.07237880229949951,
"step": 7930
},
{
"epoch": 0.8050084911160115,
"grad_norm": 4.821249485015869,
"learning_rate": 1.9318881777918973e-06,
"loss": 0.07695739865303039,
"step": 7940
},
{
"epoch": 0.8060223557143943,
"grad_norm": 5.5935773849487305,
"learning_rate": 1.9125328492064575e-06,
"loss": 0.07970571517944336,
"step": 7950
},
{
"epoch": 0.8070362203127772,
"grad_norm": 5.05609130859375,
"learning_rate": 1.8932647111980007e-06,
"loss": 0.07709313631057739,
"step": 7960
},
{
"epoch": 0.8080500849111601,
"grad_norm": 4.565520763397217,
"learning_rate": 1.8740839714953552e-06,
"loss": 0.08870221376419067,
"step": 7970
},
{
"epoch": 0.809063949509543,
"grad_norm": 5.265730381011963,
"learning_rate": 1.8549908368851099e-06,
"loss": 0.09057682752609253,
"step": 7980
},
{
"epoch": 0.8100778141079259,
"grad_norm": 4.435678005218506,
"learning_rate": 1.8359855132093884e-06,
"loss": 0.06611788272857666,
"step": 7990
},
{
"epoch": 0.8110916787063088,
"grad_norm": 4.228857517242432,
"learning_rate": 1.8170682053636358e-06,
"loss": 0.0820127248764038,
"step": 8000
},
{
"epoch": 0.8121055433046916,
"grad_norm": 4.265110969543457,
"learning_rate": 1.7982391172943936e-06,
"loss": 0.08216036558151245,
"step": 8010
},
{
"epoch": 0.8131194079030746,
"grad_norm": 4.707644939422607,
"learning_rate": 1.7794984519971126e-06,
"loss": 0.09401485919952393,
"step": 8020
},
{
"epoch": 0.8141332725014574,
"grad_norm": 3.933032512664795,
"learning_rate": 1.7608464115139712e-06,
"loss": 0.0656413733959198,
"step": 8030
},
{
"epoch": 0.8151471370998403,
"grad_norm": 3.4959590435028076,
"learning_rate": 1.7422831969316724e-06,
"loss": 0.09030433893203735,
"step": 8040
},
{
"epoch": 0.8161610016982233,
"grad_norm": 4.5226149559021,
"learning_rate": 1.7238090083793003e-06,
"loss": 0.07917751669883728,
"step": 8050
},
{
"epoch": 0.8171748662966061,
"grad_norm": 4.543294906616211,
"learning_rate": 1.7054240450261594e-06,
"loss": 0.06969000101089477,
"step": 8060
},
{
"epoch": 0.818188730894989,
"grad_norm": 4.38911771774292,
"learning_rate": 1.6871285050796093e-06,
"loss": 0.08018871545791625,
"step": 8070
},
{
"epoch": 0.8192025954933718,
"grad_norm": 3.7114734649658203,
"learning_rate": 1.6689225857829527e-06,
"loss": 0.07347989082336426,
"step": 8080
},
{
"epoch": 0.8202164600917548,
"grad_norm": 4.623588562011719,
"learning_rate": 1.6508064834132875e-06,
"loss": 0.08153756856918334,
"step": 8090
},
{
"epoch": 0.8212303246901377,
"grad_norm": 5.521157264709473,
"learning_rate": 1.632780393279404e-06,
"loss": 0.07487385272979737,
"step": 8100
},
{
"epoch": 0.8222441892885205,
"grad_norm": 5.434915542602539,
"learning_rate": 1.614844509719674e-06,
"loss": 0.0756332278251648,
"step": 8110
},
{
"epoch": 0.8232580538869034,
"grad_norm": 3.9832887649536133,
"learning_rate": 1.5969990260999569e-06,
"loss": 0.0815940797328949,
"step": 8120
},
{
"epoch": 0.8242719184852862,
"grad_norm": 3.8737680912017822,
"learning_rate": 1.5792441348115116e-06,
"loss": 0.06972190737724304,
"step": 8130
},
{
"epoch": 0.8252857830836692,
"grad_norm": 5.918801307678223,
"learning_rate": 1.5615800272689352e-06,
"loss": 0.07007790803909301,
"step": 8140
},
{
"epoch": 0.8262996476820521,
"grad_norm": 4.795719146728516,
"learning_rate": 1.5440068939080776e-06,
"loss": 0.06793877482414246,
"step": 8150
},
{
"epoch": 0.8273135122804349,
"grad_norm": 4.404543399810791,
"learning_rate": 1.5265249241840153e-06,
"loss": 0.08453413248062133,
"step": 8160
},
{
"epoch": 0.8283273768788179,
"grad_norm": 4.274115562438965,
"learning_rate": 1.5091343065689756e-06,
"loss": 0.0787917971611023,
"step": 8170
},
{
"epoch": 0.8293412414772007,
"grad_norm": 4.632278919219971,
"learning_rate": 1.4918352285503357e-06,
"loss": 0.07415574789047241,
"step": 8180
},
{
"epoch": 0.8303551060755836,
"grad_norm": 4.861246585845947,
"learning_rate": 1.4746278766285905e-06,
"loss": 0.07301176786422729,
"step": 8190
},
{
"epoch": 0.8313689706739665,
"grad_norm": 5.318351745605469,
"learning_rate": 1.4575124363153315e-06,
"loss": 0.08075512647628784,
"step": 8200
},
{
"epoch": 0.8323828352723494,
"grad_norm": 5.115492820739746,
"learning_rate": 1.4404890921312654e-06,
"loss": 0.0759343147277832,
"step": 8210
},
{
"epoch": 0.8333966998707323,
"grad_norm": 6.014513969421387,
"learning_rate": 1.4235580276042071e-06,
"loss": 0.07031062841415406,
"step": 8220
},
{
"epoch": 0.8344105644691151,
"grad_norm": 4.27595329284668,
"learning_rate": 1.4067194252671157e-06,
"loss": 0.062360125780105594,
"step": 8230
},
{
"epoch": 0.835424429067498,
"grad_norm": 4.754397869110107,
"learning_rate": 1.3899734666561138e-06,
"loss": 0.07604198455810547,
"step": 8240
},
{
"epoch": 0.836438293665881,
"grad_norm": 4.682862758636475,
"learning_rate": 1.373320332308541e-06,
"loss": 0.07690721154212951,
"step": 8250
},
{
"epoch": 0.8374521582642638,
"grad_norm": 4.132352352142334,
"learning_rate": 1.3567602017609981e-06,
"loss": 0.06614551544189454,
"step": 8260
},
{
"epoch": 0.8384660228626467,
"grad_norm": 4.125055313110352,
"learning_rate": 1.3402932535474277e-06,
"loss": 0.07468845844268798,
"step": 8270
},
{
"epoch": 0.8394798874610295,
"grad_norm": 5.3394389152526855,
"learning_rate": 1.3239196651971654e-06,
"loss": 0.08576777577400208,
"step": 8280
},
{
"epoch": 0.8404937520594125,
"grad_norm": 3.5782525539398193,
"learning_rate": 1.3076396132330516e-06,
"loss": 0.06333107352256775,
"step": 8290
},
{
"epoch": 0.8415076166577954,
"grad_norm": 3.4309918880462646,
"learning_rate": 1.2914532731694996e-06,
"loss": 0.06644206047058106,
"step": 8300
},
{
"epoch": 0.8425214812561782,
"grad_norm": 4.5303425788879395,
"learning_rate": 1.2753608195106314e-06,
"loss": 0.05930066108703613,
"step": 8310
},
{
"epoch": 0.8435353458545611,
"grad_norm": 4.669937610626221,
"learning_rate": 1.2593624257483838e-06,
"loss": 0.07150805592536927,
"step": 8320
},
{
"epoch": 0.844549210452944,
"grad_norm": 5.719365119934082,
"learning_rate": 1.243458264360634e-06,
"loss": 0.0762537181377411,
"step": 8330
},
{
"epoch": 0.8455630750513269,
"grad_norm": 5.3142499923706055,
"learning_rate": 1.2276485068093447e-06,
"loss": 0.07204040884971619,
"step": 8340
},
{
"epoch": 0.8465769396497098,
"grad_norm": 5.046025276184082,
"learning_rate": 1.2119333235387232e-06,
"loss": 0.06609773635864258,
"step": 8350
},
{
"epoch": 0.8475908042480926,
"grad_norm": 3.7958271503448486,
"learning_rate": 1.1963128839733674e-06,
"loss": 0.06671286225318909,
"step": 8360
},
{
"epoch": 0.8486046688464756,
"grad_norm": 3.7630882263183594,
"learning_rate": 1.1807873565164507e-06,
"loss": 0.07789772748947144,
"step": 8370
},
{
"epoch": 0.8496185334448585,
"grad_norm": 4.865067958831787,
"learning_rate": 1.1653569085479067e-06,
"loss": 0.06970995664596558,
"step": 8380
},
{
"epoch": 0.8506323980432413,
"grad_norm": 3.7595293521881104,
"learning_rate": 1.150021706422615e-06,
"loss": 0.07687149047851563,
"step": 8390
},
{
"epoch": 0.8516462626416242,
"grad_norm": 4.540782451629639,
"learning_rate": 1.1347819154686246e-06,
"loss": 0.0702624797821045,
"step": 8400
},
{
"epoch": 0.8526601272400071,
"grad_norm": 5.084628105163574,
"learning_rate": 1.1196376999853531e-06,
"loss": 0.06886171102523804,
"step": 8410
},
{
"epoch": 0.85367399183839,
"grad_norm": 3.6639044284820557,
"learning_rate": 1.104589223241831e-06,
"loss": 0.07045261859893799,
"step": 8420
},
{
"epoch": 0.8546878564367729,
"grad_norm": 4.262178897857666,
"learning_rate": 1.089636647474923e-06,
"loss": 0.07594810724258423,
"step": 8430
},
{
"epoch": 0.8557017210351557,
"grad_norm": 3.1481432914733887,
"learning_rate": 1.0747801338876006e-06,
"loss": 0.05966585874557495,
"step": 8440
},
{
"epoch": 0.8567155856335387,
"grad_norm": 4.160122871398926,
"learning_rate": 1.0600198426471908e-06,
"loss": 0.07694826126098633,
"step": 8450
},
{
"epoch": 0.8577294502319215,
"grad_norm": 4.419780254364014,
"learning_rate": 1.0453559328836517e-06,
"loss": 0.07195533514022827,
"step": 8460
},
{
"epoch": 0.8587433148303044,
"grad_norm": 4.686746597290039,
"learning_rate": 1.0307885626878555e-06,
"loss": 0.06232221126556396,
"step": 8470
},
{
"epoch": 0.8597571794286873,
"grad_norm": 4.49828577041626,
"learning_rate": 1.0163178891098912e-06,
"loss": 0.06678155660629273,
"step": 8480
},
{
"epoch": 0.8607710440270702,
"grad_norm": 4.987343788146973,
"learning_rate": 1.0019440681573633e-06,
"loss": 0.06641309261322022,
"step": 8490
},
{
"epoch": 0.8617849086254531,
"grad_norm": 3.5242559909820557,
"learning_rate": 9.876672547937117e-07,
"loss": 0.06106618046760559,
"step": 8500
},
{
"epoch": 0.8627987732238359,
"grad_norm": 4.912775993347168,
"learning_rate": 9.734876029365448e-07,
"loss": 0.06334991455078125,
"step": 8510
},
{
"epoch": 0.8638126378222188,
"grad_norm": 6.582511901855469,
"learning_rate": 9.594052654559737e-07,
"loss": 0.056832247972488405,
"step": 8520
},
{
"epoch": 0.8648265024206018,
"grad_norm": 4.834287643432617,
"learning_rate": 9.454203941729767e-07,
"loss": 0.06149945855140686,
"step": 8530
},
{
"epoch": 0.8658403670189846,
"grad_norm": 3.9707446098327637,
"learning_rate": 9.315331398577454e-07,
"loss": 0.07032945156097412,
"step": 8540
},
{
"epoch": 0.8668542316173675,
"grad_norm": 4.3519511222839355,
"learning_rate": 9.177436522280714e-07,
"loss": 0.06317306756973266,
"step": 8550
},
{
"epoch": 0.8678680962157503,
"grad_norm": 5.806820392608643,
"learning_rate": 9.04052079947727e-07,
"loss": 0.07992422580718994,
"step": 8560
},
{
"epoch": 0.8688819608141333,
"grad_norm": 5.02216100692749,
"learning_rate": 8.904585706248637e-07,
"loss": 0.06199525594711304,
"step": 8570
},
{
"epoch": 0.8698958254125162,
"grad_norm": 4.027821063995361,
"learning_rate": 8.76963270810427e-07,
"loss": 0.0631024956703186,
"step": 8580
},
{
"epoch": 0.870909690010899,
"grad_norm": 4.580575466156006,
"learning_rate": 8.635663259965632e-07,
"loss": 0.07217603921890259,
"step": 8590
},
{
"epoch": 0.871923554609282,
"grad_norm": 3.606722354888916,
"learning_rate": 8.502678806150588e-07,
"loss": 0.053040289878845216,
"step": 8600
},
{
"epoch": 0.8729374192076648,
"grad_norm": 3.8784899711608887,
"learning_rate": 8.370680780357909e-07,
"loss": 0.061421549320220946,
"step": 8610
},
{
"epoch": 0.8739512838060477,
"grad_norm": 4.08092737197876,
"learning_rate": 8.23967060565164e-07,
"loss": 0.058585071563720705,
"step": 8620
},
{
"epoch": 0.8749651484044306,
"grad_norm": 3.542606830596924,
"learning_rate": 8.109649694445898e-07,
"loss": 0.061476171016693115,
"step": 8630
},
{
"epoch": 0.8759790130028134,
"grad_norm": 3.592095375061035,
"learning_rate": 7.980619448489601e-07,
"loss": 0.06497960090637207,
"step": 8640
},
{
"epoch": 0.8769928776011964,
"grad_norm": 2.5060482025146484,
"learning_rate": 7.852581258851333e-07,
"loss": 0.05713431835174561,
"step": 8650
},
{
"epoch": 0.8780067421995792,
"grad_norm": 4.489684581756592,
"learning_rate": 7.725536505904396e-07,
"loss": 0.0500103771686554,
"step": 8660
},
{
"epoch": 0.8790206067979621,
"grad_norm": 4.9915008544921875,
"learning_rate": 7.599486559311887e-07,
"loss": 0.06409144997596741,
"step": 8670
},
{
"epoch": 0.880034471396345,
"grad_norm": 4.573352813720703,
"learning_rate": 7.474432778011953e-07,
"loss": 0.06039006114006042,
"step": 8680
},
{
"epoch": 0.8810483359947279,
"grad_norm": 4.018733501434326,
"learning_rate": 7.350376510203105e-07,
"loss": 0.05765443444252014,
"step": 8690
},
{
"epoch": 0.8820622005931108,
"grad_norm": 4.5047760009765625,
"learning_rate": 7.227319093329732e-07,
"loss": 0.0688095211982727,
"step": 8700
},
{
"epoch": 0.8830760651914937,
"grad_norm": 4.003312587738037,
"learning_rate": 7.105261854067691e-07,
"loss": 0.05722858309745789,
"step": 8710
},
{
"epoch": 0.8840899297898765,
"grad_norm": 4.000243186950684,
"learning_rate": 6.984206108309932e-07,
"loss": 0.05845695734024048,
"step": 8720
},
{
"epoch": 0.8851037943882595,
"grad_norm": 3.436002731323242,
"learning_rate": 6.864153161152354e-07,
"loss": 0.04764501750469208,
"step": 8730
},
{
"epoch": 0.8861176589866423,
"grad_norm": 4.426645755767822,
"learning_rate": 6.745104306879769e-07,
"loss": 0.06402113437652587,
"step": 8740
},
{
"epoch": 0.8871315235850252,
"grad_norm": 5.100968360900879,
"learning_rate": 6.627060828951881e-07,
"loss": 0.06183567047119141,
"step": 8750
},
{
"epoch": 0.8881453881834082,
"grad_norm": 3.9962596893310547,
"learning_rate": 6.510023999989501e-07,
"loss": 0.06601526737213134,
"step": 8760
},
{
"epoch": 0.889159252781791,
"grad_norm": 5.276778221130371,
"learning_rate": 6.393995081760795e-07,
"loss": 0.061566162109375,
"step": 8770
},
{
"epoch": 0.8901731173801739,
"grad_norm": 4.503662109375,
"learning_rate": 6.278975325167691e-07,
"loss": 0.0643608808517456,
"step": 8780
},
{
"epoch": 0.8911869819785567,
"grad_norm": 6.500443458557129,
"learning_rate": 6.164965970232417e-07,
"loss": 0.06682579517364502,
"step": 8790
},
{
"epoch": 0.8922008465769397,
"grad_norm": 5.013277530670166,
"learning_rate": 6.051968246084117e-07,
"loss": 0.06578773260116577,
"step": 8800
},
{
"epoch": 0.8932147111753226,
"grad_norm": 5.333492279052734,
"learning_rate": 5.939983370945535e-07,
"loss": 0.06977545022964478,
"step": 8810
},
{
"epoch": 0.8942285757737054,
"grad_norm": 4.536230087280273,
"learning_rate": 5.829012552120039e-07,
"loss": 0.06514511108398438,
"step": 8820
},
{
"epoch": 0.8952424403720883,
"grad_norm": 4.311686038970947,
"learning_rate": 5.719056985978388e-07,
"loss": 0.05992481708526611,
"step": 8830
},
{
"epoch": 0.8962563049704712,
"grad_norm": 3.856419086456299,
"learning_rate": 5.610117857946062e-07,
"loss": 0.04977162182331085,
"step": 8840
},
{
"epoch": 0.8972701695688541,
"grad_norm": 3.71254825592041,
"learning_rate": 5.502196342490318e-07,
"loss": 0.06080018281936646,
"step": 8850
},
{
"epoch": 0.898284034167237,
"grad_norm": 2.6859326362609863,
"learning_rate": 5.395293603107566e-07,
"loss": 0.0536657452583313,
"step": 8860
},
{
"epoch": 0.8992978987656198,
"grad_norm": 4.304467678070068,
"learning_rate": 5.2894107923109e-07,
"loss": 0.05374969840049744,
"step": 8870
},
{
"epoch": 0.9003117633640028,
"grad_norm": 6.596916675567627,
"learning_rate": 5.184549051617571e-07,
"loss": 0.06347913146018982,
"step": 8880
},
{
"epoch": 0.9013256279623856,
"grad_norm": 4.788922309875488,
"learning_rate": 5.08070951153673e-07,
"loss": 0.053334057331085205,
"step": 8890
},
{
"epoch": 0.9023394925607685,
"grad_norm": 4.465794086456299,
"learning_rate": 4.977893291557256e-07,
"loss": 0.060540080070495605,
"step": 8900
},
{
"epoch": 0.9033533571591514,
"grad_norm": 3.763526678085327,
"learning_rate": 4.876101500135621e-07,
"loss": 0.06400558352470398,
"step": 8910
},
{
"epoch": 0.9043672217575343,
"grad_norm": 4.875303745269775,
"learning_rate": 4.775335234684042e-07,
"loss": 0.06281731724739074,
"step": 8920
},
{
"epoch": 0.9053810863559172,
"grad_norm": 4.454023361206055,
"learning_rate": 4.675595581558556e-07,
"loss": 0.054623228311538694,
"step": 8930
},
{
"epoch": 0.9063949509543,
"grad_norm": 5.211312294006348,
"learning_rate": 4.576883616047345e-07,
"loss": 0.07220885753631592,
"step": 8940
},
{
"epoch": 0.9074088155526829,
"grad_norm": 3.3115174770355225,
"learning_rate": 4.479200402359174e-07,
"loss": 0.04543644785881042,
"step": 8950
},
{
"epoch": 0.9084226801510659,
"grad_norm": 4.308058738708496,
"learning_rate": 4.382546993611836e-07,
"loss": 0.050774121284484865,
"step": 8960
},
{
"epoch": 0.9094365447494487,
"grad_norm": 5.0768609046936035,
"learning_rate": 4.28692443182086e-07,
"loss": 0.06364564895629883,
"step": 8970
},
{
"epoch": 0.9104504093478316,
"grad_norm": 3.085719347000122,
"learning_rate": 4.192333747888311e-07,
"loss": 0.055559849739074706,
"step": 8980
},
{
"epoch": 0.9114642739462144,
"grad_norm": 4.542215824127197,
"learning_rate": 4.0987759615915837e-07,
"loss": 0.06374951004981995,
"step": 8990
},
{
"epoch": 0.9124781385445974,
"grad_norm": 5.196413993835449,
"learning_rate": 4.0062520815724816e-07,
"loss": 0.06019670963287353,
"step": 9000
},
{
"epoch": 0.9134920031429803,
"grad_norm": 3.8123650550842285,
"learning_rate": 3.9147631053263027e-07,
"loss": 0.045917320251464847,
"step": 9010
},
{
"epoch": 0.9145058677413631,
"grad_norm": 4.886520862579346,
"learning_rate": 3.824310019191102e-07,
"loss": 0.06986879110336304,
"step": 9020
},
{
"epoch": 0.915519732339746,
"grad_norm": 5.162857532501221,
"learning_rate": 3.734893798337058e-07,
"loss": 0.05212366580963135,
"step": 9030
},
{
"epoch": 0.9165335969381289,
"grad_norm": 3.197697162628174,
"learning_rate": 3.646515406755946e-07,
"loss": 0.05415894985198975,
"step": 9040
},
{
"epoch": 0.9175474615365118,
"grad_norm": 4.5132269859313965,
"learning_rate": 3.559175797250769e-07,
"loss": 0.05296427011489868,
"step": 9050
},
{
"epoch": 0.9185613261348947,
"grad_norm": 4.449508190155029,
"learning_rate": 3.4728759114254774e-07,
"loss": 0.05152961015701294,
"step": 9060
},
{
"epoch": 0.9195751907332775,
"grad_norm": 5.679589748382568,
"learning_rate": 3.3876166796747877e-07,
"loss": 0.0675972580909729,
"step": 9070
},
{
"epoch": 0.9205890553316605,
"grad_norm": 4.730302810668945,
"learning_rate": 3.303399021174214e-07,
"loss": 0.05251699686050415,
"step": 9080
},
{
"epoch": 0.9216029199300434,
"grad_norm": 4.968480110168457,
"learning_rate": 3.220223843870096e-07,
"loss": 0.052497029304504395,
"step": 9090
},
{
"epoch": 0.9226167845284262,
"grad_norm": 7.865025520324707,
"learning_rate": 3.1380920444698093e-07,
"loss": 0.06473975777626037,
"step": 9100
},
{
"epoch": 0.9236306491268091,
"grad_norm": 5.103010177612305,
"learning_rate": 3.057004508432193e-07,
"loss": 0.045138558745384215,
"step": 9110
},
{
"epoch": 0.924644513725192,
"grad_norm": 2.550945997238159,
"learning_rate": 2.976962109957882e-07,
"loss": 0.050883978605270386,
"step": 9120
},
{
"epoch": 0.9256583783235749,
"grad_norm": 4.262047290802002,
"learning_rate": 2.8979657119799667e-07,
"loss": 0.055169868469238284,
"step": 9130
},
{
"epoch": 0.9266722429219578,
"grad_norm": 4.89635705947876,
"learning_rate": 2.8200161661546377e-07,
"loss": 0.05750205516815186,
"step": 9140
},
{
"epoch": 0.9276861075203406,
"grad_norm": 4.753897190093994,
"learning_rate": 2.7431143128520243e-07,
"loss": 0.05559988021850586,
"step": 9150
},
{
"epoch": 0.9286999721187236,
"grad_norm": 4.3059401512146,
"learning_rate": 2.667260981147157e-07,
"loss": 0.04352918267250061,
"step": 9160
},
{
"epoch": 0.9297138367171064,
"grad_norm": 4.849719524383545,
"learning_rate": 2.5924569888109654e-07,
"loss": 0.05597134828567505,
"step": 9170
},
{
"epoch": 0.9307277013154893,
"grad_norm": 5.545217990875244,
"learning_rate": 2.5187031423015597e-07,
"loss": 0.05419236421585083,
"step": 9180
},
{
"epoch": 0.9317415659138722,
"grad_norm": 2.9285216331481934,
"learning_rate": 2.4460002367554523e-07,
"loss": 0.04273094534873963,
"step": 9190
},
{
"epoch": 0.9327554305122551,
"grad_norm": 5.27381706237793,
"learning_rate": 2.3743490559789973e-07,
"loss": 0.058786427974700926,
"step": 9200
},
{
"epoch": 0.933769295110638,
"grad_norm": 2.3281636238098145,
"learning_rate": 2.3037503724399946e-07,
"loss": 0.040942218899726865,
"step": 9210
},
{
"epoch": 0.9347831597090208,
"grad_norm": 3.9320366382598877,
"learning_rate": 2.2342049472592996e-07,
"loss": 0.053173398971557616,
"step": 9220
},
{
"epoch": 0.9357970243074037,
"grad_norm": 7.400967597961426,
"learning_rate": 2.1657135302026177e-07,
"loss": 0.053149324655532834,
"step": 9230
},
{
"epoch": 0.9368108889057867,
"grad_norm": 4.053079128265381,
"learning_rate": 2.09827685967251e-07,
"loss": 0.044053539633750916,
"step": 9240
},
{
"epoch": 0.9378247535041695,
"grad_norm": 1.8017324209213257,
"learning_rate": 2.0318956627003227e-07,
"loss": 0.038763591647148134,
"step": 9250
},
{
"epoch": 0.9388386181025524,
"grad_norm": 4.490161895751953,
"learning_rate": 1.9665706549384044e-07,
"loss": 0.054458558559417725,
"step": 9260
},
{
"epoch": 0.9398524827009352,
"grad_norm": 3.93466854095459,
"learning_rate": 1.9023025406523787e-07,
"loss": 0.05251436233520508,
"step": 9270
},
{
"epoch": 0.9408663472993182,
"grad_norm": 5.403939723968506,
"learning_rate": 1.8390920127135613e-07,
"loss": 0.0517845869064331,
"step": 9280
},
{
"epoch": 0.9418802118977011,
"grad_norm": 4.735721588134766,
"learning_rate": 1.7769397525914668e-07,
"loss": 0.052587807178497314,
"step": 9290
},
{
"epoch": 0.9428940764960839,
"grad_norm": 4.030385971069336,
"learning_rate": 1.7158464303465017e-07,
"loss": 0.050893735885620114,
"step": 9300
},
{
"epoch": 0.9439079410944669,
"grad_norm": 6.06185245513916,
"learning_rate": 1.6558127046226945e-07,
"loss": 0.05377797484397888,
"step": 9310
},
{
"epoch": 0.9449218056928497,
"grad_norm": 3.3613815307617188,
"learning_rate": 1.5968392226406337e-07,
"loss": 0.0554373025894165,
"step": 9320
},
{
"epoch": 0.9459356702912326,
"grad_norm": 7.471890926361084,
"learning_rate": 1.5389266201904506e-07,
"loss": 0.047643131017684935,
"step": 9330
},
{
"epoch": 0.9469495348896155,
"grad_norm": 2.305101156234741,
"learning_rate": 1.482075521625026e-07,
"loss": 0.05129271149635315,
"step": 9340
},
{
"epoch": 0.9479633994879983,
"grad_norm": 3.5675208568573,
"learning_rate": 1.4262865398531834e-07,
"loss": 0.053032410144805905,
"step": 9350
},
{
"epoch": 0.9489772640863813,
"grad_norm": 4.344719886779785,
"learning_rate": 1.3715602763331172e-07,
"loss": 0.05092419385910034,
"step": 9360
},
{
"epoch": 0.9499911286847641,
"grad_norm": 4.047629356384277,
"learning_rate": 1.3178973210659528e-07,
"loss": 0.06043769121170044,
"step": 9370
},
{
"epoch": 0.951004993283147,
"grad_norm": 3.613696336746216,
"learning_rate": 1.2652982525892975e-07,
"loss": 0.05118474364280701,
"step": 9380
},
{
"epoch": 0.95201885788153,
"grad_norm": 10.067850112915039,
"learning_rate": 1.213763637971066e-07,
"loss": 0.05573541522026062,
"step": 9390
},
{
"epoch": 0.9530327224799128,
"grad_norm": 3.9944329261779785,
"learning_rate": 1.1632940328033526e-07,
"loss": 0.05101707577705383,
"step": 9400
},
{
"epoch": 0.9540465870782957,
"grad_norm": 5.496587753295898,
"learning_rate": 1.1138899811964477e-07,
"loss": 0.05443580150604248,
"step": 9410
},
{
"epoch": 0.9550604516766786,
"grad_norm": 5.266269207000732,
"learning_rate": 1.0655520157729194e-07,
"loss": 0.05561995506286621,
"step": 9420
},
{
"epoch": 0.9560743162750615,
"grad_norm": 2.9803788661956787,
"learning_rate": 1.0182806576619742e-07,
"loss": 0.049706846475601196,
"step": 9430
},
{
"epoch": 0.9570881808734444,
"grad_norm": 3.9620840549468994,
"learning_rate": 9.72076416493728e-08,
"loss": 0.041834527254104616,
"step": 9440
},
{
"epoch": 0.9581020454718272,
"grad_norm": 5.6704277992248535,
"learning_rate": 9.269397903937883e-08,
"loss": 0.058120441436767575,
"step": 9450
},
{
"epoch": 0.9591159100702101,
"grad_norm": 2.9679501056671143,
"learning_rate": 8.828712659778471e-08,
"loss": 0.05296071767807007,
"step": 9460
},
{
"epoch": 0.9601297746685931,
"grad_norm": 4.3682098388671875,
"learning_rate": 8.398713183464857e-08,
"loss": 0.04111744463443756,
"step": 9470
},
{
"epoch": 0.9611436392669759,
"grad_norm": 6.633136749267578,
"learning_rate": 7.979404110799338e-08,
"loss": 0.05298812985420227,
"step": 9480
},
{
"epoch": 0.9621575038653588,
"grad_norm": 4.037006855010986,
"learning_rate": 7.570789962332181e-08,
"loss": 0.04251847267150879,
"step": 9490
},
{
"epoch": 0.9631713684637416,
"grad_norm": 4.580509185791016,
"learning_rate": 7.172875143311996e-08,
"loss": 0.050883686542510985,
"step": 9500
},
{
"epoch": 0.9641852330621246,
"grad_norm": 7.291702747344971,
"learning_rate": 6.785663943638333e-08,
"loss": 0.05226942300796509,
"step": 9510
},
{
"epoch": 0.9651990976605075,
"grad_norm": 3.2236342430114746,
"learning_rate": 6.409160537815818e-08,
"loss": 0.058408141136169434,
"step": 9520
},
{
"epoch": 0.9662129622588903,
"grad_norm": 5.07227897644043,
"learning_rate": 6.043368984908759e-08,
"loss": 0.05285707712173462,
"step": 9530
},
{
"epoch": 0.9672268268572732,
"grad_norm": 2.8341338634490967,
"learning_rate": 5.688293228497399e-08,
"loss": 0.05143759250640869,
"step": 9540
},
{
"epoch": 0.968240691455656,
"grad_norm": 3.6010756492614746,
"learning_rate": 5.34393709663561e-08,
"loss": 0.04833375513553619,
"step": 9550
},
{
"epoch": 0.969254556054039,
"grad_norm": 2.8568763732910156,
"learning_rate": 5.0103043018097096e-08,
"loss": 0.05615646243095398,
"step": 9560
},
{
"epoch": 0.9702684206524219,
"grad_norm": 4.777455806732178,
"learning_rate": 4.687398440897828e-08,
"loss": 0.05354637503623962,
"step": 9570
},
{
"epoch": 0.9712822852508047,
"grad_norm": 3.8060238361358643,
"learning_rate": 4.3752229951321553e-08,
"loss": 0.05064668655395508,
"step": 9580
},
{
"epoch": 0.9722961498491877,
"grad_norm": 3.727797508239746,
"learning_rate": 4.0737813300602e-08,
"loss": 0.048354247212409975,
"step": 9590
},
{
"epoch": 0.9733100144475705,
"grad_norm": 4.355417728424072,
"learning_rate": 3.783076695509702e-08,
"loss": 0.04862031936645508,
"step": 9600
},
{
"epoch": 0.9743238790459534,
"grad_norm": 3.8068277835845947,
"learning_rate": 3.5031122255524406e-08,
"loss": 0.047008419036865236,
"step": 9610
},
{
"epoch": 0.9753377436443363,
"grad_norm": 2.93790340423584,
"learning_rate": 3.233890938471373e-08,
"loss": 0.06449747085571289,
"step": 9620
},
{
"epoch": 0.9763516082427192,
"grad_norm": 3.809860944747925,
"learning_rate": 2.9754157367277715e-08,
"loss": 0.05831618905067444,
"step": 9630
},
{
"epoch": 0.9773654728411021,
"grad_norm": 3.3440911769866943,
"learning_rate": 2.72768940692969e-08,
"loss": 0.04105756282806396,
"step": 9640
},
{
"epoch": 0.9783793374394849,
"grad_norm": 5.13588809967041,
"learning_rate": 2.4907146198023256e-08,
"loss": 0.05013155937194824,
"step": 9650
},
{
"epoch": 0.9793932020378678,
"grad_norm": 3.754680871963501,
"learning_rate": 2.2644939301588174e-08,
"loss": 0.05232709050178528,
"step": 9660
},
{
"epoch": 0.9804070666362508,
"grad_norm": 6.152595520019531,
"learning_rate": 2.049029776873268e-08,
"loss": 0.04397982358932495,
"step": 9670
},
{
"epoch": 0.9814209312346336,
"grad_norm": 1.6838688850402832,
"learning_rate": 1.8443244828536544e-08,
"loss": 0.05386626124382019,
"step": 9680
},
{
"epoch": 0.9824347958330165,
"grad_norm": 4.4649882316589355,
"learning_rate": 1.650380255017625e-08,
"loss": 0.05556715726852417,
"step": 9690
},
{
"epoch": 0.9834486604313993,
"grad_norm": 3.7025253772735596,
"learning_rate": 1.4671991842680756e-08,
"loss": 0.04285165965557099,
"step": 9700
},
{
"epoch": 0.9844625250297823,
"grad_norm": 5.164238929748535,
"learning_rate": 1.294783245470943e-08,
"loss": 0.059722906351089476,
"step": 9710
},
{
"epoch": 0.9854763896281652,
"grad_norm": 4.672970294952393,
"learning_rate": 1.13313429743378e-08,
"loss": 0.0472618579864502,
"step": 9720
},
{
"epoch": 0.986490254226548,
"grad_norm": 2.7522644996643066,
"learning_rate": 9.822540828858807e-09,
"loss": 0.0533828854560852,
"step": 9730
},
{
"epoch": 0.9875041188249309,
"grad_norm": 4.5804972648620605,
"learning_rate": 8.421442284591852e-09,
"loss": 0.05164008736610413,
"step": 9740
},
{
"epoch": 0.9885179834233138,
"grad_norm": 5.260897636413574,
"learning_rate": 7.128062446709605e-09,
"loss": 0.05719882249832153,
"step": 9750
},
{
"epoch": 0.9895318480216967,
"grad_norm": 5.575925350189209,
"learning_rate": 5.942415259074796e-09,
"loss": 0.042970356345176694,
"step": 9760
},
{
"epoch": 0.9905457126200796,
"grad_norm": 3.5226821899414062,
"learning_rate": 4.86451350409145e-09,
"loss": 0.044629260897636414,
"step": 9770
},
{
"epoch": 0.9915595772184624,
"grad_norm": 4.193933010101318,
"learning_rate": 3.894368802563886e-09,
"loss": 0.044721481204032895,
"step": 9780
},
{
"epoch": 0.9925734418168454,
"grad_norm": 5.998621940612793,
"learning_rate": 3.0319916135723713e-09,
"loss": 0.049154406785964964,
"step": 9790
},
{
"epoch": 0.9935873064152283,
"grad_norm": 4.185999870300293,
"learning_rate": 2.277391234363213e-09,
"loss": 0.044599863886833194,
"step": 9800
},
{
"epoch": 0.9946011710136111,
"grad_norm": 3.735816478729248,
"learning_rate": 1.6305758002443939e-09,
"loss": 0.044607114791870114,
"step": 9810
},
{
"epoch": 0.995615035611994,
"grad_norm": 4.97714376449585,
"learning_rate": 1.091552284500086e-09,
"loss": 0.05474789142608642,
"step": 9820
},
{
"epoch": 0.9966289002103769,
"grad_norm": 4.701594352722168,
"learning_rate": 6.603264983173762e-10,
"loss": 0.05394202470779419,
"step": 9830
},
{
"epoch": 0.9976427648087598,
"grad_norm": 4.175233840942383,
"learning_rate": 3.3690309071965355e-10,
"loss": 0.04798406362533569,
"step": 9840
},
{
"epoch": 0.9986566294071427,
"grad_norm": 4.730007171630859,
"learning_rate": 1.2128554851664754e-10,
"loss": 0.04514379799365997,
"step": 9850
},
{
"epoch": 0.9996704940055255,
"grad_norm": 3.993809461593628,
"learning_rate": 1.3476196272232955e-11,
"loss": 0.044790375232696536,
"step": 9860
},
{
"epoch": 1.0,
"step": 9864,
"total_flos": 5.070738018188604e+19,
"train_loss": 0.13528188769928415,
"train_runtime": 103129.9357,
"train_samples_per_second": 6.121,
"train_steps_per_second": 0.096
}
],
"logging_steps": 10,
"max_steps": 9864,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5.070738018188604e+19,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}