Qwen3-Tiny-L8 / trainer_state.json
DecoderImmortal's picture
Upload folder using huggingface_hub
104633d verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 200.0,
"global_step": 5205,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00019212295869356388,
"grad_norm": 41.0853157043457,
"learning_rate": 1.1494252873563217e-06,
"loss": 10.62430191040039,
"step": 1
},
{
"epoch": 0.0009606147934678194,
"grad_norm": 40.41545867919922,
"learning_rate": 5.747126436781608e-06,
"loss": 10.600605010986328,
"step": 5
},
{
"epoch": 0.0019212295869356388,
"grad_norm": 24.825101852416992,
"learning_rate": 1.1494252873563217e-05,
"loss": 9.892306518554687,
"step": 10
},
{
"epoch": 0.002881844380403458,
"grad_norm": 14.558918952941895,
"learning_rate": 1.7241379310344825e-05,
"loss": 8.44493408203125,
"step": 15
},
{
"epoch": 0.0038424591738712775,
"grad_norm": 7.1786956787109375,
"learning_rate": 2.2988505747126433e-05,
"loss": 7.345828247070313,
"step": 20
},
{
"epoch": 0.004803073967339097,
"grad_norm": 4.3571858406066895,
"learning_rate": 2.8735632183908045e-05,
"loss": 6.655020141601563,
"step": 25
},
{
"epoch": 0.005763688760806916,
"grad_norm": 2.342735767364502,
"learning_rate": 3.448275862068965e-05,
"loss": 6.164151000976562,
"step": 30
},
{
"epoch": 0.0067243035542747355,
"grad_norm": 2.570441246032715,
"learning_rate": 4.022988505747126e-05,
"loss": 5.778363800048828,
"step": 35
},
{
"epoch": 0.007684918347742555,
"grad_norm": 2.7692394256591797,
"learning_rate": 4.5977011494252866e-05,
"loss": 5.467860412597656,
"step": 40
},
{
"epoch": 0.008645533141210375,
"grad_norm": 0.8302366137504578,
"learning_rate": 5.172413793103448e-05,
"loss": 5.22553825378418,
"step": 45
},
{
"epoch": 0.009606147934678195,
"grad_norm": 3.4278295040130615,
"learning_rate": 5.747126436781609e-05,
"loss": 5.0273689270019535,
"step": 50
},
{
"epoch": 0.010566762728146013,
"grad_norm": 3.7562148571014404,
"learning_rate": 6.32183908045977e-05,
"loss": 4.888296890258789,
"step": 55
},
{
"epoch": 0.011527377521613832,
"grad_norm": 3.4503631591796875,
"learning_rate": 6.89655172413793e-05,
"loss": 4.772652435302734,
"step": 60
},
{
"epoch": 0.012487992315081652,
"grad_norm": 2.543815851211548,
"learning_rate": 7.471264367816091e-05,
"loss": 4.681902313232422,
"step": 65
},
{
"epoch": 0.013448607108549471,
"grad_norm": 3.03908371925354,
"learning_rate": 8.045977011494252e-05,
"loss": 4.596535873413086,
"step": 70
},
{
"epoch": 0.01440922190201729,
"grad_norm": 2.9536008834838867,
"learning_rate": 8.620689655172413e-05,
"loss": 4.515168380737305,
"step": 75
},
{
"epoch": 0.01536983669548511,
"grad_norm": 2.8714489936828613,
"learning_rate": 9.195402298850573e-05,
"loss": 4.442340850830078,
"step": 80
},
{
"epoch": 0.01633045148895293,
"grad_norm": 2.1564369201660156,
"learning_rate": 9.770114942528733e-05,
"loss": 4.369157028198242,
"step": 85
},
{
"epoch": 0.01729106628242075,
"grad_norm": 1.5947808027267456,
"learning_rate": 0.00010344827586206896,
"loss": 4.297710418701172,
"step": 90
},
{
"epoch": 0.01825168107588857,
"grad_norm": 1.6778233051300049,
"learning_rate": 0.00010919540229885056,
"loss": 4.230863571166992,
"step": 95
},
{
"epoch": 0.01921229586935639,
"grad_norm": 1.724900245666504,
"learning_rate": 0.00011494252873563218,
"loss": 4.168477249145508,
"step": 100
},
{
"epoch": 0.020172910662824207,
"grad_norm": 1.8800899982452393,
"learning_rate": 0.00012068965517241378,
"loss": 4.10834846496582,
"step": 105
},
{
"epoch": 0.021133525456292025,
"grad_norm": 1.567356824874878,
"learning_rate": 0.0001264367816091954,
"loss": 4.051171493530274,
"step": 110
},
{
"epoch": 0.022094140249759846,
"grad_norm": 2.2714192867279053,
"learning_rate": 0.000132183908045977,
"loss": 4.001234817504883,
"step": 115
},
{
"epoch": 0.023054755043227664,
"grad_norm": 2.6298563480377197,
"learning_rate": 0.0001379310344827586,
"loss": 3.9564029693603517,
"step": 120
},
{
"epoch": 0.024015369836695485,
"grad_norm": 2.0561466217041016,
"learning_rate": 0.0001436781609195402,
"loss": 3.9216072082519533,
"step": 125
},
{
"epoch": 0.024975984630163303,
"grad_norm": 2.2524895668029785,
"learning_rate": 0.00014942528735632183,
"loss": 3.8848548889160157,
"step": 130
},
{
"epoch": 0.025936599423631124,
"grad_norm": 1.8565645217895508,
"learning_rate": 0.00015517241379310346,
"loss": 3.8477012634277346,
"step": 135
},
{
"epoch": 0.026897214217098942,
"grad_norm": 2.007159948348999,
"learning_rate": 0.00016091954022988503,
"loss": 3.8230205535888673,
"step": 140
},
{
"epoch": 0.027857829010566763,
"grad_norm": 2.5089001655578613,
"learning_rate": 0.00016666666666666666,
"loss": 3.790781784057617,
"step": 145
},
{
"epoch": 0.02881844380403458,
"grad_norm": 1.7888845205307007,
"learning_rate": 0.00017241379310344826,
"loss": 3.7702369689941406,
"step": 150
},
{
"epoch": 0.029779058597502402,
"grad_norm": 1.5350267887115479,
"learning_rate": 0.00017816091954022986,
"loss": 3.752313995361328,
"step": 155
},
{
"epoch": 0.03073967339097022,
"grad_norm": 2.6425604820251465,
"learning_rate": 0.00018390804597701147,
"loss": 3.736890411376953,
"step": 160
},
{
"epoch": 0.03170028818443804,
"grad_norm": 2.1871745586395264,
"learning_rate": 0.0001896551724137931,
"loss": 3.7081199645996095,
"step": 165
},
{
"epoch": 0.03266090297790586,
"grad_norm": 2.301276683807373,
"learning_rate": 0.00019540229885057467,
"loss": 3.6886245727539064,
"step": 170
},
{
"epoch": 0.03362151777137368,
"grad_norm": 2.3541669845581055,
"learning_rate": 0.0002011494252873563,
"loss": 3.6742767333984374,
"step": 175
},
{
"epoch": 0.0345821325648415,
"grad_norm": 2.193776845932007,
"learning_rate": 0.00020689655172413793,
"loss": 3.654471588134766,
"step": 180
},
{
"epoch": 0.03554274735830932,
"grad_norm": 2.2675869464874268,
"learning_rate": 0.00021264367816091953,
"loss": 3.634075164794922,
"step": 185
},
{
"epoch": 0.03650336215177714,
"grad_norm": 2.262054920196533,
"learning_rate": 0.00021839080459770113,
"loss": 3.623701477050781,
"step": 190
},
{
"epoch": 0.037463976945244955,
"grad_norm": 2.4970099925994873,
"learning_rate": 0.00022413793103448273,
"loss": 3.608829116821289,
"step": 195
},
{
"epoch": 0.03842459173871278,
"grad_norm": 2.580589532852173,
"learning_rate": 0.00022988505747126436,
"loss": 3.594409942626953,
"step": 200
},
{
"epoch": 0.0393852065321806,
"grad_norm": 2.5698821544647217,
"learning_rate": 0.00023563218390804593,
"loss": 3.589406967163086,
"step": 205
},
{
"epoch": 0.040345821325648415,
"grad_norm": 1.9148350954055786,
"learning_rate": 0.00024137931034482756,
"loss": 3.5731269836425783,
"step": 210
},
{
"epoch": 0.04130643611911623,
"grad_norm": 2.6525511741638184,
"learning_rate": 0.00024712643678160916,
"loss": 3.5579261779785156,
"step": 215
},
{
"epoch": 0.04226705091258405,
"grad_norm": 2.4932427406311035,
"learning_rate": 0.0002528735632183908,
"loss": 3.5460506439208985,
"step": 220
},
{
"epoch": 0.043227665706051875,
"grad_norm": 1.522286295890808,
"learning_rate": 0.00025862068965517237,
"loss": 3.534061813354492,
"step": 225
},
{
"epoch": 0.04418828049951969,
"grad_norm": 2.077308177947998,
"learning_rate": 0.000264367816091954,
"loss": 3.519297790527344,
"step": 230
},
{
"epoch": 0.04514889529298751,
"grad_norm": 2.2139241695404053,
"learning_rate": 0.0002701149425287356,
"loss": 3.508769226074219,
"step": 235
},
{
"epoch": 0.04610951008645533,
"grad_norm": 2.71065616607666,
"learning_rate": 0.0002758620689655172,
"loss": 3.4990470886230467,
"step": 240
},
{
"epoch": 0.04707012487992315,
"grad_norm": 1.8673561811447144,
"learning_rate": 0.00028160919540229883,
"loss": 3.492870330810547,
"step": 245
},
{
"epoch": 0.04803073967339097,
"grad_norm": 2.2589685916900635,
"learning_rate": 0.0002873563218390804,
"loss": 3.482229232788086,
"step": 250
},
{
"epoch": 0.04899135446685879,
"grad_norm": 1.6717121601104736,
"learning_rate": 0.00029310344827586203,
"loss": 3.4728935241699217,
"step": 255
},
{
"epoch": 0.049951969260326606,
"grad_norm": 2.6019248962402344,
"learning_rate": 0.00029885057471264366,
"loss": 3.4658123016357423,
"step": 260
},
{
"epoch": 0.05091258405379443,
"grad_norm": 2.2234745025634766,
"learning_rate": 0.00029999951546647263,
"loss": 3.46058349609375,
"step": 265
},
{
"epoch": 0.05187319884726225,
"grad_norm": 1.7633609771728516,
"learning_rate": 0.0002999975470543828,
"loss": 3.4448532104492187,
"step": 270
},
{
"epoch": 0.052833813640730067,
"grad_norm": 1.698107361793518,
"learning_rate": 0.00029999406450023966,
"loss": 3.4391746520996094,
"step": 275
},
{
"epoch": 0.053794428434197884,
"grad_norm": 2.3649418354034424,
"learning_rate": 0.0002999890678391978,
"loss": 3.4237644195556642,
"step": 280
},
{
"epoch": 0.05475504322766571,
"grad_norm": 1.6354587078094482,
"learning_rate": 0.00029998255712169563,
"loss": 3.4203346252441404,
"step": 285
},
{
"epoch": 0.05571565802113353,
"grad_norm": 2.191877841949463,
"learning_rate": 0.00029997453241345533,
"loss": 3.4129383087158205,
"step": 290
},
{
"epoch": 0.056676272814601344,
"grad_norm": 2.1449062824249268,
"learning_rate": 0.0002999649937954818,
"loss": 3.4019508361816406,
"step": 295
},
{
"epoch": 0.05763688760806916,
"grad_norm": 1.9247010946273804,
"learning_rate": 0.0002999539413640621,
"loss": 3.3923084259033205,
"step": 300
},
{
"epoch": 0.05859750240153699,
"grad_norm": 1.6601442098617554,
"learning_rate": 0.0002999413752307644,
"loss": 3.382649230957031,
"step": 305
},
{
"epoch": 0.059558117195004805,
"grad_norm": 1.656147837638855,
"learning_rate": 0.0002999272955224369,
"loss": 3.3772228240966795,
"step": 310
},
{
"epoch": 0.06051873198847262,
"grad_norm": 1.9272291660308838,
"learning_rate": 0.0002999117023812064,
"loss": 3.3669418334960937,
"step": 315
},
{
"epoch": 0.06147934678194044,
"grad_norm": 2.0251946449279785,
"learning_rate": 0.000299894595964477,
"loss": 3.3636070251464845,
"step": 320
},
{
"epoch": 0.06243996157540826,
"grad_norm": 1.5922743082046509,
"learning_rate": 0.0002998759764449286,
"loss": 3.3583389282226563,
"step": 325
},
{
"epoch": 0.06340057636887608,
"grad_norm": 2.2867417335510254,
"learning_rate": 0.0002998558440105148,
"loss": 3.3570938110351562,
"step": 330
},
{
"epoch": 0.0643611911623439,
"grad_norm": 1.678760051727295,
"learning_rate": 0.0002998341988644614,
"loss": 3.3533226013183595,
"step": 335
},
{
"epoch": 0.06532180595581172,
"grad_norm": 1.9134340286254883,
"learning_rate": 0.0002998110412252641,
"loss": 3.3417484283447267,
"step": 340
},
{
"epoch": 0.06628242074927954,
"grad_norm": 1.3479728698730469,
"learning_rate": 0.0002997863713266866,
"loss": 3.3312728881835936,
"step": 345
},
{
"epoch": 0.06724303554274735,
"grad_norm": 1.3297585248947144,
"learning_rate": 0.0002997601894177576,
"loss": 3.3266414642333983,
"step": 350
},
{
"epoch": 0.06820365033621517,
"grad_norm": 1.5574193000793457,
"learning_rate": 0.00029973249576276914,
"loss": 3.321349334716797,
"step": 355
},
{
"epoch": 0.069164265129683,
"grad_norm": 1.3063061237335205,
"learning_rate": 0.0002997032906412732,
"loss": 3.316680145263672,
"step": 360
},
{
"epoch": 0.07012487992315082,
"grad_norm": 1.440643548965454,
"learning_rate": 0.0002996725743480793,
"loss": 3.3083744049072266,
"step": 365
},
{
"epoch": 0.07108549471661864,
"grad_norm": 1.2425435781478882,
"learning_rate": 0.00029964034719325147,
"loss": 3.299941635131836,
"step": 370
},
{
"epoch": 0.07204610951008646,
"grad_norm": 1.3375848531723022,
"learning_rate": 0.0002996066095021048,
"loss": 3.298897552490234,
"step": 375
},
{
"epoch": 0.07300672430355427,
"grad_norm": 1.429042100906372,
"learning_rate": 0.0002995713616152028,
"loss": 3.2951019287109373,
"step": 380
},
{
"epoch": 0.07396733909702209,
"grad_norm": 1.6015335321426392,
"learning_rate": 0.0002995346038883532,
"loss": 3.2871078491210937,
"step": 385
},
{
"epoch": 0.07492795389048991,
"grad_norm": 1.444579005241394,
"learning_rate": 0.0002994963366926048,
"loss": 3.286525344848633,
"step": 390
},
{
"epoch": 0.07588856868395773,
"grad_norm": 1.481871247291565,
"learning_rate": 0.0002994565604142439,
"loss": 3.2782516479492188,
"step": 395
},
{
"epoch": 0.07684918347742556,
"grad_norm": 1.2840272188186646,
"learning_rate": 0.00029941527545478976,
"loss": 3.2710845947265623,
"step": 400
},
{
"epoch": 0.07780979827089338,
"grad_norm": 1.2282651662826538,
"learning_rate": 0.00029937248223099136,
"loss": 3.2628021240234375,
"step": 405
},
{
"epoch": 0.0787704130643612,
"grad_norm": 1.5256259441375732,
"learning_rate": 0.00029932818117482245,
"loss": 3.2642288208007812,
"step": 410
},
{
"epoch": 0.07973102785782901,
"grad_norm": 1.4667640924453735,
"learning_rate": 0.0002992823727334776,
"loss": 3.2577728271484374,
"step": 415
},
{
"epoch": 0.08069164265129683,
"grad_norm": 1.9893492460250854,
"learning_rate": 0.00029923505736936774,
"loss": 3.257119369506836,
"step": 420
},
{
"epoch": 0.08165225744476465,
"grad_norm": 1.1091679334640503,
"learning_rate": 0.0002991862355601151,
"loss": 3.2528831481933596,
"step": 425
},
{
"epoch": 0.08261287223823247,
"grad_norm": 1.278975486755371,
"learning_rate": 0.00029913590779854886,
"loss": 3.246588134765625,
"step": 430
},
{
"epoch": 0.08357348703170028,
"grad_norm": 1.7972854375839233,
"learning_rate": 0.00029908407459269977,
"loss": 3.2438831329345703,
"step": 435
},
{
"epoch": 0.0845341018251681,
"grad_norm": 1.242336392402649,
"learning_rate": 0.0002990307364657954,
"loss": 3.2402252197265624,
"step": 440
},
{
"epoch": 0.08549471661863593,
"grad_norm": 1.3686480522155762,
"learning_rate": 0.0002989758939562545,
"loss": 3.234004592895508,
"step": 445
},
{
"epoch": 0.08645533141210375,
"grad_norm": 1.6549354791641235,
"learning_rate": 0.0002989195476176818,
"loss": 3.231829833984375,
"step": 450
},
{
"epoch": 0.08741594620557157,
"grad_norm": 1.1533397436141968,
"learning_rate": 0.00029886169801886237,
"loss": 3.2260894775390625,
"step": 455
},
{
"epoch": 0.08837656099903939,
"grad_norm": 1.2408617734909058,
"learning_rate": 0.00029880234574375576,
"loss": 3.2213096618652344,
"step": 460
},
{
"epoch": 0.0893371757925072,
"grad_norm": 1.3094635009765625,
"learning_rate": 0.00029874149139149037,
"loss": 3.218523406982422,
"step": 465
},
{
"epoch": 0.09029779058597502,
"grad_norm": 1.352949619293213,
"learning_rate": 0.00029867913557635704,
"loss": 3.216180419921875,
"step": 470
},
{
"epoch": 0.09125840537944284,
"grad_norm": 1.2964717149734497,
"learning_rate": 0.0002986152789278031,
"loss": 3.21021728515625,
"step": 475
},
{
"epoch": 0.09221902017291066,
"grad_norm": 1.2936161756515503,
"learning_rate": 0.00029854992209042626,
"loss": 3.208134078979492,
"step": 480
},
{
"epoch": 0.09317963496637849,
"grad_norm": 1.262890338897705,
"learning_rate": 0.0002984830657239673,
"loss": 3.2024818420410157,
"step": 485
},
{
"epoch": 0.0941402497598463,
"grad_norm": 1.1409361362457275,
"learning_rate": 0.00029841471050330424,
"loss": 3.201049041748047,
"step": 490
},
{
"epoch": 0.09510086455331412,
"grad_norm": 1.154974102973938,
"learning_rate": 0.00029834485711844515,
"loss": 3.1987125396728517,
"step": 495
},
{
"epoch": 0.09606147934678194,
"grad_norm": 1.115403652191162,
"learning_rate": 0.00029827350627452116,
"loss": 3.1905792236328123,
"step": 500
},
{
"epoch": 0.09702209414024976,
"grad_norm": 1.3029977083206177,
"learning_rate": 0.00029820065869177954,
"loss": 3.1972682952880858,
"step": 505
},
{
"epoch": 0.09798270893371758,
"grad_norm": 1.2876211404800415,
"learning_rate": 0.0002981263151055762,
"loss": 3.190180206298828,
"step": 510
},
{
"epoch": 0.0989433237271854,
"grad_norm": 1.4698253870010376,
"learning_rate": 0.0002980504762663683,
"loss": 3.183108901977539,
"step": 515
},
{
"epoch": 0.09990393852065321,
"grad_norm": 1.4724935293197632,
"learning_rate": 0.0002979731429397071,
"loss": 3.1818157196044923,
"step": 520
},
{
"epoch": 0.10086455331412104,
"grad_norm": 1.252737283706665,
"learning_rate": 0.0002978943159062295,
"loss": 3.178622245788574,
"step": 525
},
{
"epoch": 0.10182516810758886,
"grad_norm": 1.3171908855438232,
"learning_rate": 0.0002978139959616507,
"loss": 3.1794586181640625,
"step": 530
},
{
"epoch": 0.10278578290105668,
"grad_norm": 1.3921356201171875,
"learning_rate": 0.00029773218391675594,
"loss": 3.175571060180664,
"step": 535
},
{
"epoch": 0.1037463976945245,
"grad_norm": 1.0483286380767822,
"learning_rate": 0.00029764888059739255,
"loss": 3.1715625762939452,
"step": 540
},
{
"epoch": 0.10470701248799232,
"grad_norm": 1.0862990617752075,
"learning_rate": 0.00029756408684446136,
"loss": 3.173019218444824,
"step": 545
},
{
"epoch": 0.10566762728146013,
"grad_norm": 1.3614447116851807,
"learning_rate": 0.0002974778035139081,
"loss": 3.166230583190918,
"step": 550
},
{
"epoch": 0.10662824207492795,
"grad_norm": 1.0270849466323853,
"learning_rate": 0.00029739003147671536,
"loss": 3.1632837295532226,
"step": 555
},
{
"epoch": 0.10758885686839577,
"grad_norm": 1.5262467861175537,
"learning_rate": 0.00029730077161889304,
"loss": 3.163615417480469,
"step": 560
},
{
"epoch": 0.10854947166186359,
"grad_norm": 0.967074453830719,
"learning_rate": 0.00029721002484147,
"loss": 3.160854721069336,
"step": 565
},
{
"epoch": 0.10951008645533142,
"grad_norm": 1.3244905471801758,
"learning_rate": 0.00029711779206048454,
"loss": 3.158466339111328,
"step": 570
},
{
"epoch": 0.11047070124879924,
"grad_norm": 1.2831615209579468,
"learning_rate": 0.0002970240742069755,
"loss": 3.1545330047607423,
"step": 575
},
{
"epoch": 0.11143131604226705,
"grad_norm": 0.985418975353241,
"learning_rate": 0.0002969288722269726,
"loss": 3.149937057495117,
"step": 580
},
{
"epoch": 0.11239193083573487,
"grad_norm": 1.0786052942276,
"learning_rate": 0.000296832187081487,
"loss": 3.1523942947387695,
"step": 585
},
{
"epoch": 0.11335254562920269,
"grad_norm": 1.0934644937515259,
"learning_rate": 0.0002967340197465017,
"loss": 3.1465457916259765,
"step": 590
},
{
"epoch": 0.1143131604226705,
"grad_norm": 1.354055404663086,
"learning_rate": 0.00029663437121296146,
"loss": 3.1465686798095702,
"step": 595
},
{
"epoch": 0.11527377521613832,
"grad_norm": 1.1991117000579834,
"learning_rate": 0.000296533242486763,
"loss": 3.143705368041992,
"step": 600
},
{
"epoch": 0.11623439000960614,
"grad_norm": 1.2595206499099731,
"learning_rate": 0.0002964306345887447,
"loss": 3.1374244689941406,
"step": 605
},
{
"epoch": 0.11719500480307397,
"grad_norm": 1.1941015720367432,
"learning_rate": 0.0002963265485546764,
"loss": 3.134844970703125,
"step": 610
},
{
"epoch": 0.11815561959654179,
"grad_norm": 1.045292615890503,
"learning_rate": 0.00029622098543524884,
"loss": 3.1372697830200194,
"step": 615
},
{
"epoch": 0.11911623439000961,
"grad_norm": 0.8975104093551636,
"learning_rate": 0.00029611394629606324,
"loss": 3.1340496063232424,
"step": 620
},
{
"epoch": 0.12007684918347743,
"grad_norm": 0.8841443061828613,
"learning_rate": 0.0002960054322176204,
"loss": 3.12859992980957,
"step": 625
},
{
"epoch": 0.12103746397694524,
"grad_norm": 1.4646399021148682,
"learning_rate": 0.0002958954442953096,
"loss": 3.1266860961914062,
"step": 630
},
{
"epoch": 0.12199807877041306,
"grad_norm": 1.2528270483016968,
"learning_rate": 0.000295783983639398,
"loss": 3.127555274963379,
"step": 635
},
{
"epoch": 0.12295869356388088,
"grad_norm": 1.0346133708953857,
"learning_rate": 0.00029567105137501916,
"loss": 3.123224639892578,
"step": 640
},
{
"epoch": 0.1239193083573487,
"grad_norm": 1.1643197536468506,
"learning_rate": 0.00029555664864216156,
"loss": 3.1221992492675783,
"step": 645
},
{
"epoch": 0.12487992315081652,
"grad_norm": 0.9598444700241089,
"learning_rate": 0.00029544077659565747,
"loss": 3.1172836303710936,
"step": 650
},
{
"epoch": 0.12584053794428435,
"grad_norm": 1.193166732788086,
"learning_rate": 0.0002953234364051708,
"loss": 3.1188310623168944,
"step": 655
},
{
"epoch": 0.12680115273775217,
"grad_norm": 1.3501605987548828,
"learning_rate": 0.00029520462925518575,
"loss": 3.113943099975586,
"step": 660
},
{
"epoch": 0.12776176753121998,
"grad_norm": 1.1221381425857544,
"learning_rate": 0.00029508435634499467,
"loss": 3.113107109069824,
"step": 665
},
{
"epoch": 0.1287223823246878,
"grad_norm": 1.2083373069763184,
"learning_rate": 0.00029496261888868586,
"loss": 3.11612491607666,
"step": 670
},
{
"epoch": 0.12968299711815562,
"grad_norm": 1.0393755435943604,
"learning_rate": 0.0002948394181151314,
"loss": 3.109981155395508,
"step": 675
},
{
"epoch": 0.13064361191162344,
"grad_norm": 1.1999260187149048,
"learning_rate": 0.0002947147552679748,
"loss": 3.1087677001953127,
"step": 680
},
{
"epoch": 0.13160422670509125,
"grad_norm": 0.9760367274284363,
"learning_rate": 0.00029458863160561837,
"loss": 3.1072208404541017,
"step": 685
},
{
"epoch": 0.13256484149855907,
"grad_norm": 1.0330339670181274,
"learning_rate": 0.0002944610484012105,
"loss": 3.1058883666992188,
"step": 690
},
{
"epoch": 0.1335254562920269,
"grad_norm": 0.8262532353401184,
"learning_rate": 0.0002943320069426329,
"loss": 3.1005834579467773,
"step": 695
},
{
"epoch": 0.1344860710854947,
"grad_norm": 1.139979362487793,
"learning_rate": 0.00029420150853248756,
"loss": 3.1009517669677735,
"step": 700
},
{
"epoch": 0.13544668587896252,
"grad_norm": 1.2979532480239868,
"learning_rate": 0.0002940695544880836,
"loss": 3.0997783660888674,
"step": 705
},
{
"epoch": 0.13640730067243034,
"grad_norm": 1.267760157585144,
"learning_rate": 0.0002939361461414238,
"loss": 3.0968265533447266,
"step": 710
},
{
"epoch": 0.1373679154658982,
"grad_norm": 1.070842981338501,
"learning_rate": 0.0002938012848391915,
"loss": 3.0932445526123047,
"step": 715
},
{
"epoch": 0.138328530259366,
"grad_norm": 1.2436041831970215,
"learning_rate": 0.0002936649719427367,
"loss": 3.0934431076049806,
"step": 720
},
{
"epoch": 0.13928914505283382,
"grad_norm": 1.0946882963180542,
"learning_rate": 0.00029352720882806267,
"loss": 3.0919456481933594,
"step": 725
},
{
"epoch": 0.14024975984630164,
"grad_norm": 1.0279921293258667,
"learning_rate": 0.00029338799688581146,
"loss": 3.090176010131836,
"step": 730
},
{
"epoch": 0.14121037463976946,
"grad_norm": 1.1641135215759277,
"learning_rate": 0.00029324733752125054,
"loss": 3.087961196899414,
"step": 735
},
{
"epoch": 0.14217098943323728,
"grad_norm": 1.176697850227356,
"learning_rate": 0.0002931052321542581,
"loss": 3.086262512207031,
"step": 740
},
{
"epoch": 0.1431316042267051,
"grad_norm": 1.13466215133667,
"learning_rate": 0.00029296168221930904,
"loss": 3.0825126647949217,
"step": 745
},
{
"epoch": 0.1440922190201729,
"grad_norm": 1.2085994482040405,
"learning_rate": 0.0002928166891654604,
"loss": 3.085866928100586,
"step": 750
},
{
"epoch": 0.14505283381364073,
"grad_norm": 1.0957227945327759,
"learning_rate": 0.00029267025445633667,
"loss": 3.083575439453125,
"step": 755
},
{
"epoch": 0.14601344860710855,
"grad_norm": 0.9977965354919434,
"learning_rate": 0.0002925223795701149,
"loss": 3.0803741455078124,
"step": 760
},
{
"epoch": 0.14697406340057637,
"grad_norm": 1.1439604759216309,
"learning_rate": 0.00029237306599951007,
"loss": 3.0777074813842775,
"step": 765
},
{
"epoch": 0.14793467819404418,
"grad_norm": 0.8976171612739563,
"learning_rate": 0.00029222231525176005,
"loss": 3.0794315338134766,
"step": 770
},
{
"epoch": 0.148895292987512,
"grad_norm": 1.0934321880340576,
"learning_rate": 0.0002920701288486099,
"loss": 3.0764802932739257,
"step": 775
},
{
"epoch": 0.14985590778097982,
"grad_norm": 1.088166356086731,
"learning_rate": 0.00029191650832629694,
"loss": 3.0753334045410154,
"step": 780
},
{
"epoch": 0.15081652257444764,
"grad_norm": 0.9292176365852356,
"learning_rate": 0.00029176145523553517,
"loss": 3.0713483810424806,
"step": 785
},
{
"epoch": 0.15177713736791545,
"grad_norm": 1.0206445455551147,
"learning_rate": 0.0002916049711414996,
"loss": 3.0717500686645507,
"step": 790
},
{
"epoch": 0.15273775216138327,
"grad_norm": 1.208133578300476,
"learning_rate": 0.00029144705762381036,
"loss": 3.073202896118164,
"step": 795
},
{
"epoch": 0.15369836695485112,
"grad_norm": 0.9913300275802612,
"learning_rate": 0.0002912877162765169,
"loss": 3.0660655975341795,
"step": 800
},
{
"epoch": 0.15465898174831894,
"grad_norm": 1.1119699478149414,
"learning_rate": 0.00029112694870808155,
"loss": 3.067470169067383,
"step": 805
},
{
"epoch": 0.15561959654178675,
"grad_norm": 1.0400080680847168,
"learning_rate": 0.00029096475654136395,
"loss": 3.0652034759521483,
"step": 810
},
{
"epoch": 0.15658021133525457,
"grad_norm": 0.9247759580612183,
"learning_rate": 0.000290801141413604,
"loss": 3.061397361755371,
"step": 815
},
{
"epoch": 0.1575408261287224,
"grad_norm": 1.060563564300537,
"learning_rate": 0.00029063610497640576,
"loss": 3.0618301391601563,
"step": 820
},
{
"epoch": 0.1585014409221902,
"grad_norm": 1.137531042098999,
"learning_rate": 0.0002904696488957204,
"loss": 3.0617816925048826,
"step": 825
},
{
"epoch": 0.15946205571565802,
"grad_norm": 1.0392354726791382,
"learning_rate": 0.0002903017748518298,
"loss": 3.053458404541016,
"step": 830
},
{
"epoch": 0.16042267050912584,
"grad_norm": 0.9369018077850342,
"learning_rate": 0.0002901324845393294,
"loss": 3.0599037170410157,
"step": 835
},
{
"epoch": 0.16138328530259366,
"grad_norm": 0.8582949638366699,
"learning_rate": 0.00028996177966711097,
"loss": 3.0585887908935545,
"step": 840
},
{
"epoch": 0.16234390009606148,
"grad_norm": 0.9821586608886719,
"learning_rate": 0.0002897896619583455,
"loss": 3.0565303802490233,
"step": 845
},
{
"epoch": 0.1633045148895293,
"grad_norm": 1.064107894897461,
"learning_rate": 0.0002896161331504659,
"loss": 3.0574161529541017,
"step": 850
},
{
"epoch": 0.1642651296829971,
"grad_norm": 1.0095359086990356,
"learning_rate": 0.00028944119499514913,
"loss": 3.054368019104004,
"step": 855
},
{
"epoch": 0.16522574447646493,
"grad_norm": 1.0314931869506836,
"learning_rate": 0.0002892648492582989,
"loss": 3.0471572875976562,
"step": 860
},
{
"epoch": 0.16618635926993275,
"grad_norm": 1.1124778985977173,
"learning_rate": 0.00028908709772002765,
"loss": 3.0495674133300783,
"step": 865
},
{
"epoch": 0.16714697406340057,
"grad_norm": 0.9910938143730164,
"learning_rate": 0.00028890794217463863,
"loss": 3.0501693725585937,
"step": 870
},
{
"epoch": 0.16810758885686838,
"grad_norm": 1.0562894344329834,
"learning_rate": 0.0002887273844306076,
"loss": 3.0482528686523436,
"step": 875
},
{
"epoch": 0.1690682036503362,
"grad_norm": 0.9755779504776001,
"learning_rate": 0.00028854542631056494,
"loss": 3.043923187255859,
"step": 880
},
{
"epoch": 0.17002881844380405,
"grad_norm": 0.9010165333747864,
"learning_rate": 0.0002883620696512769,
"loss": 3.047307586669922,
"step": 885
},
{
"epoch": 0.17098943323727187,
"grad_norm": 1.0329279899597168,
"learning_rate": 0.0002881773163036273,
"loss": 3.045383071899414,
"step": 890
},
{
"epoch": 0.17195004803073968,
"grad_norm": 0.8573128581047058,
"learning_rate": 0.00028799116813259875,
"loss": 3.0434667587280275,
"step": 895
},
{
"epoch": 0.1729106628242075,
"grad_norm": 1.0701560974121094,
"learning_rate": 0.0002878036270172538,
"loss": 3.0357574462890624,
"step": 900
},
{
"epoch": 0.17387127761767532,
"grad_norm": 1.0385925769805908,
"learning_rate": 0.000287614694850716,
"loss": 3.0400657653808594,
"step": 905
},
{
"epoch": 0.17483189241114314,
"grad_norm": 0.9062285423278809,
"learning_rate": 0.00028742437354015073,
"loss": 3.0386375427246093,
"step": 910
},
{
"epoch": 0.17579250720461095,
"grad_norm": 1.2658556699752808,
"learning_rate": 0.0002872326650067462,
"loss": 3.0390193939208983,
"step": 915
},
{
"epoch": 0.17675312199807877,
"grad_norm": 0.9829347729682922,
"learning_rate": 0.00028703957118569363,
"loss": 3.0389629364013673,
"step": 920
},
{
"epoch": 0.1777137367915466,
"grad_norm": 1.0681240558624268,
"learning_rate": 0.000286845094026168,
"loss": 3.038488006591797,
"step": 925
},
{
"epoch": 0.1786743515850144,
"grad_norm": 1.271577000617981,
"learning_rate": 0.0002866492354913086,
"loss": 3.037461853027344,
"step": 930
},
{
"epoch": 0.17963496637848222,
"grad_norm": 0.898420512676239,
"learning_rate": 0.0002864519975581986,
"loss": 3.032963180541992,
"step": 935
},
{
"epoch": 0.18059558117195004,
"grad_norm": 1.040523886680603,
"learning_rate": 0.0002862533822178456,
"loss": 3.032858657836914,
"step": 940
},
{
"epoch": 0.18155619596541786,
"grad_norm": 0.9702991843223572,
"learning_rate": 0.00028605339147516113,
"loss": 3.029499816894531,
"step": 945
},
{
"epoch": 0.18251681075888568,
"grad_norm": 1.0565072298049927,
"learning_rate": 0.00028585202734894105,
"loss": 3.030923843383789,
"step": 950
},
{
"epoch": 0.1834774255523535,
"grad_norm": 0.9746853113174438,
"learning_rate": 0.00028564929187184447,
"loss": 3.0296430587768555,
"step": 955
},
{
"epoch": 0.1844380403458213,
"grad_norm": 0.9929348826408386,
"learning_rate": 0.00028544518709037363,
"loss": 3.0282403945922853,
"step": 960
},
{
"epoch": 0.18539865513928913,
"grad_norm": 0.8328250646591187,
"learning_rate": 0.000285239715064853,
"loss": 3.0228111267089846,
"step": 965
},
{
"epoch": 0.18635926993275698,
"grad_norm": 1.216163158416748,
"learning_rate": 0.0002850328778694088,
"loss": 3.0275947570800783,
"step": 970
},
{
"epoch": 0.1873198847262248,
"grad_norm": 1.077054500579834,
"learning_rate": 0.0002848246775919478,
"loss": 3.0228961944580077,
"step": 975
},
{
"epoch": 0.1882804995196926,
"grad_norm": 1.0535924434661865,
"learning_rate": 0.0002846151163341364,
"loss": 3.024255561828613,
"step": 980
},
{
"epoch": 0.18924111431316043,
"grad_norm": 1.0875251293182373,
"learning_rate": 0.0002844041962113792,
"loss": 3.0249141693115233,
"step": 985
},
{
"epoch": 0.19020172910662825,
"grad_norm": 0.9994290471076965,
"learning_rate": 0.00028419191935279793,
"loss": 3.024821472167969,
"step": 990
},
{
"epoch": 0.19116234390009607,
"grad_norm": 0.9676864743232727,
"learning_rate": 0.00028397828790120965,
"loss": 3.025034713745117,
"step": 995
},
{
"epoch": 0.19212295869356388,
"grad_norm": 0.9725849628448486,
"learning_rate": 0.0002837633040131055,
"loss": 3.0168121337890623,
"step": 1000
},
{
"epoch": 0.1930835734870317,
"grad_norm": 0.8413275480270386,
"learning_rate": 0.00028354696985862865,
"loss": 3.017466735839844,
"step": 1005
},
{
"epoch": 0.19404418828049952,
"grad_norm": 0.8834673166275024,
"learning_rate": 0.00028332928762155225,
"loss": 3.0153526306152343,
"step": 1010
},
{
"epoch": 0.19500480307396734,
"grad_norm": 0.9116687774658203,
"learning_rate": 0.0002831102594992579,
"loss": 3.0169631958007814,
"step": 1015
},
{
"epoch": 0.19596541786743515,
"grad_norm": 0.9015905261039734,
"learning_rate": 0.00028288988770271297,
"loss": 3.015023422241211,
"step": 1020
},
{
"epoch": 0.19692603266090297,
"grad_norm": 1.0525202751159668,
"learning_rate": 0.00028266817445644855,
"loss": 3.014568901062012,
"step": 1025
},
{
"epoch": 0.1978866474543708,
"grad_norm": 0.9534981846809387,
"learning_rate": 0.0002824451219985369,
"loss": 3.0124256134033205,
"step": 1030
},
{
"epoch": 0.1988472622478386,
"grad_norm": 0.8009642362594604,
"learning_rate": 0.0002822207325805688,
"loss": 3.017258071899414,
"step": 1035
},
{
"epoch": 0.19980787704130643,
"grad_norm": 1.049578309059143,
"learning_rate": 0.00028199500846763116,
"loss": 3.0123516082763673,
"step": 1040
},
{
"epoch": 0.20076849183477424,
"grad_norm": 1.0571658611297607,
"learning_rate": 0.0002817679519382836,
"loss": 3.01287956237793,
"step": 1045
},
{
"epoch": 0.2017291066282421,
"grad_norm": 1.0068848133087158,
"learning_rate": 0.0002815395652845359,
"loss": 3.0150609970092774,
"step": 1050
},
{
"epoch": 0.2026897214217099,
"grad_norm": 1.0555028915405273,
"learning_rate": 0.0002813098508118247,
"loss": 3.0080333709716798,
"step": 1055
},
{
"epoch": 0.20365033621517772,
"grad_norm": 0.9432759881019592,
"learning_rate": 0.0002810788108389901,
"loss": 3.0080041885375977,
"step": 1060
},
{
"epoch": 0.20461095100864554,
"grad_norm": 0.8660106658935547,
"learning_rate": 0.0002808464476982526,
"loss": 3.0054351806640627,
"step": 1065
},
{
"epoch": 0.20557156580211336,
"grad_norm": 0.9244606494903564,
"learning_rate": 0.0002806127637351892,
"loss": 3.0057044982910157,
"step": 1070
},
{
"epoch": 0.20653218059558118,
"grad_norm": 1.1719539165496826,
"learning_rate": 0.00028037776130871,
"loss": 3.00616340637207,
"step": 1075
},
{
"epoch": 0.207492795389049,
"grad_norm": 1.2205188274383545,
"learning_rate": 0.00028014144279103406,
"loss": 3.0076831817626952,
"step": 1080
},
{
"epoch": 0.2084534101825168,
"grad_norm": 0.9898233413696289,
"learning_rate": 0.0002799038105676658,
"loss": 3.0033103942871096,
"step": 1085
},
{
"epoch": 0.20941402497598463,
"grad_norm": 0.9085561633110046,
"learning_rate": 0.00027966486703737066,
"loss": 3.0030406951904296,
"step": 1090
},
{
"epoch": 0.21037463976945245,
"grad_norm": 1.0487213134765625,
"learning_rate": 0.0002794246146121512,
"loss": 3.0055965423583983,
"step": 1095
},
{
"epoch": 0.21133525456292027,
"grad_norm": 0.7599855065345764,
"learning_rate": 0.0002791830557172224,
"loss": 3.0053455352783205,
"step": 1100
},
{
"epoch": 0.21229586935638808,
"grad_norm": 1.0096842050552368,
"learning_rate": 0.00027894019279098726,
"loss": 2.99503231048584,
"step": 1105
},
{
"epoch": 0.2132564841498559,
"grad_norm": 1.0883978605270386,
"learning_rate": 0.00027869602828501234,
"loss": 3.0012443542480467,
"step": 1110
},
{
"epoch": 0.21421709894332372,
"grad_norm": 0.8861981630325317,
"learning_rate": 0.00027845056466400297,
"loss": 2.9960763931274412,
"step": 1115
},
{
"epoch": 0.21517771373679154,
"grad_norm": 1.0259391069412231,
"learning_rate": 0.0002782038044057783,
"loss": 2.998910140991211,
"step": 1120
},
{
"epoch": 0.21613832853025935,
"grad_norm": 0.9595276117324829,
"learning_rate": 0.0002779557500012462,
"loss": 2.9986873626708985,
"step": 1125
},
{
"epoch": 0.21709894332372717,
"grad_norm": 1.0162408351898193,
"learning_rate": 0.0002777064039543784,
"loss": 2.9951988220214845,
"step": 1130
},
{
"epoch": 0.21805955811719502,
"grad_norm": 0.8509448766708374,
"learning_rate": 0.00027745576878218496,
"loss": 2.9969932556152346,
"step": 1135
},
{
"epoch": 0.21902017291066284,
"grad_norm": 0.7489306330680847,
"learning_rate": 0.0002772038470146888,
"loss": 2.9912147521972656,
"step": 1140
},
{
"epoch": 0.21998078770413065,
"grad_norm": 0.936427891254425,
"learning_rate": 0.0002769506411949007,
"loss": 2.993634033203125,
"step": 1145
},
{
"epoch": 0.22094140249759847,
"grad_norm": 1.020005464553833,
"learning_rate": 0.00027669615387879284,
"loss": 2.9922344207763674,
"step": 1150
},
{
"epoch": 0.2219020172910663,
"grad_norm": 0.9557084441184998,
"learning_rate": 0.0002764403876352736,
"loss": 2.9920581817626952,
"step": 1155
},
{
"epoch": 0.2228626320845341,
"grad_norm": 0.9452062845230103,
"learning_rate": 0.0002761833450461613,
"loss": 2.990475082397461,
"step": 1160
},
{
"epoch": 0.22382324687800192,
"grad_norm": 0.9600782990455627,
"learning_rate": 0.0002759250287061583,
"loss": 2.990601348876953,
"step": 1165
},
{
"epoch": 0.22478386167146974,
"grad_norm": 0.9427198767662048,
"learning_rate": 0.00027566544122282496,
"loss": 2.990622138977051,
"step": 1170
},
{
"epoch": 0.22574447646493756,
"grad_norm": 1.020228624343872,
"learning_rate": 0.0002754045852165529,
"loss": 2.988072967529297,
"step": 1175
},
{
"epoch": 0.22670509125840538,
"grad_norm": 0.8352153301239014,
"learning_rate": 0.00027514246332053876,
"loss": 2.9882789611816407,
"step": 1180
},
{
"epoch": 0.2276657060518732,
"grad_norm": 0.9119144678115845,
"learning_rate": 0.0002748790781807577,
"loss": 2.985960006713867,
"step": 1185
},
{
"epoch": 0.228626320845341,
"grad_norm": 1.0915721654891968,
"learning_rate": 0.0002746144324559368,
"loss": 2.9834156036376953,
"step": 1190
},
{
"epoch": 0.22958693563880883,
"grad_norm": 0.9779495000839233,
"learning_rate": 0.00027434852881752774,
"loss": 2.9862506866455076,
"step": 1195
},
{
"epoch": 0.23054755043227665,
"grad_norm": 0.7937590479850769,
"learning_rate": 0.0002740813699496804,
"loss": 2.984625244140625,
"step": 1200
},
{
"epoch": 0.23150816522574447,
"grad_norm": 0.8358086943626404,
"learning_rate": 0.0002738129585492153,
"loss": 2.979814910888672,
"step": 1205
},
{
"epoch": 0.23246878001921228,
"grad_norm": 0.9309784173965454,
"learning_rate": 0.0002735432973255967,
"loss": 2.9813312530517577,
"step": 1210
},
{
"epoch": 0.2334293948126801,
"grad_norm": 0.9012494683265686,
"learning_rate": 0.0002732723890009051,
"loss": 2.981847381591797,
"step": 1215
},
{
"epoch": 0.23439000960614795,
"grad_norm": 0.924193799495697,
"learning_rate": 0.00027300023630980985,
"loss": 2.985409164428711,
"step": 1220
},
{
"epoch": 0.23535062439961577,
"grad_norm": 0.9454841613769531,
"learning_rate": 0.00027272684199954137,
"loss": 2.98272819519043,
"step": 1225
},
{
"epoch": 0.23631123919308358,
"grad_norm": 0.8768509030342102,
"learning_rate": 0.0002724522088298637,
"loss": 2.981144332885742,
"step": 1230
},
{
"epoch": 0.2372718539865514,
"grad_norm": 0.9420005679130554,
"learning_rate": 0.0002721763395730462,
"loss": 2.9806352615356446,
"step": 1235
},
{
"epoch": 0.23823246878001922,
"grad_norm": 0.8248008489608765,
"learning_rate": 0.00027189923701383627,
"loss": 2.9782638549804688,
"step": 1240
},
{
"epoch": 0.23919308357348704,
"grad_norm": 0.7679121494293213,
"learning_rate": 0.0002716209039494304,
"loss": 2.9759841918945313,
"step": 1245
},
{
"epoch": 0.24015369836695485,
"grad_norm": 0.9788777828216553,
"learning_rate": 0.0002713413431894466,
"loss": 2.977429962158203,
"step": 1250
},
{
"epoch": 0.24111431316042267,
"grad_norm": 0.931688666343689,
"learning_rate": 0.00027106055755589566,
"loss": 2.9781124114990236,
"step": 1255
},
{
"epoch": 0.2420749279538905,
"grad_norm": 1.3526350259780884,
"learning_rate": 0.00027077854988315285,
"loss": 2.981077766418457,
"step": 1260
},
{
"epoch": 0.2430355427473583,
"grad_norm": 1.0243141651153564,
"learning_rate": 0.00027049532301792924,
"loss": 2.9759902954101562,
"step": 1265
},
{
"epoch": 0.24399615754082613,
"grad_norm": 0.9570327401161194,
"learning_rate": 0.00027021087981924296,
"loss": 2.9805068969726562,
"step": 1270
},
{
"epoch": 0.24495677233429394,
"grad_norm": 0.8493214249610901,
"learning_rate": 0.0002699252231583904,
"loss": 2.9751022338867186,
"step": 1275
},
{
"epoch": 0.24591738712776176,
"grad_norm": 1.0867499113082886,
"learning_rate": 0.000269638355918917,
"loss": 2.9705522537231444,
"step": 1280
},
{
"epoch": 0.24687800192122958,
"grad_norm": 0.9490500092506409,
"learning_rate": 0.00026935028099658864,
"loss": 2.9714107513427734,
"step": 1285
},
{
"epoch": 0.2478386167146974,
"grad_norm": 1.0482022762298584,
"learning_rate": 0.00026906100129936173,
"loss": 2.9761116027832033,
"step": 1290
},
{
"epoch": 0.24879923150816521,
"grad_norm": 0.9524055123329163,
"learning_rate": 0.0002687705197473545,
"loss": 2.971488189697266,
"step": 1295
},
{
"epoch": 0.24975984630163303,
"grad_norm": 0.8309116959571838,
"learning_rate": 0.00026847883927281715,
"loss": 2.971644401550293,
"step": 1300
},
{
"epoch": 0.2507204610951009,
"grad_norm": 0.946146547794342,
"learning_rate": 0.00026818596282010223,
"loss": 2.970326614379883,
"step": 1305
},
{
"epoch": 0.2516810758885687,
"grad_norm": 0.9582008719444275,
"learning_rate": 0.00026789189334563507,
"loss": 2.971521759033203,
"step": 1310
},
{
"epoch": 0.2526416906820365,
"grad_norm": 0.6957625150680542,
"learning_rate": 0.00026759663381788407,
"loss": 2.965200424194336,
"step": 1315
},
{
"epoch": 0.25360230547550433,
"grad_norm": 0.9765472412109375,
"learning_rate": 0.00026730018721733034,
"loss": 2.962516212463379,
"step": 1320
},
{
"epoch": 0.25456292026897215,
"grad_norm": 0.9319335222244263,
"learning_rate": 0.0002670025565364379,
"loss": 2.9645359039306642,
"step": 1325
},
{
"epoch": 0.25552353506243997,
"grad_norm": 1.1191481351852417,
"learning_rate": 0.0002667037447796234,
"loss": 2.9653718948364256,
"step": 1330
},
{
"epoch": 0.2564841498559078,
"grad_norm": 1.0608773231506348,
"learning_rate": 0.0002664037549632259,
"loss": 2.9668670654296876,
"step": 1335
},
{
"epoch": 0.2574447646493756,
"grad_norm": 1.082219123840332,
"learning_rate": 0.00026610259011547617,
"loss": 2.9651315689086912,
"step": 1340
},
{
"epoch": 0.2584053794428434,
"grad_norm": 0.8572206497192383,
"learning_rate": 0.0002658002532764663,
"loss": 2.964995193481445,
"step": 1345
},
{
"epoch": 0.25936599423631124,
"grad_norm": 0.960854172706604,
"learning_rate": 0.00026549674749811917,
"loss": 2.959628677368164,
"step": 1350
},
{
"epoch": 0.26032660902977905,
"grad_norm": 0.9040970206260681,
"learning_rate": 0.00026519207584415705,
"loss": 2.9645341873168944,
"step": 1355
},
{
"epoch": 0.2612872238232469,
"grad_norm": 0.9025639295578003,
"learning_rate": 0.00026488624139007154,
"loss": 2.957231140136719,
"step": 1360
},
{
"epoch": 0.2622478386167147,
"grad_norm": 0.9620739221572876,
"learning_rate": 0.0002645792472230917,
"loss": 2.9581756591796875,
"step": 1365
},
{
"epoch": 0.2632084534101825,
"grad_norm": 0.9987210631370544,
"learning_rate": 0.0002642710964421535,
"loss": 2.9581737518310547,
"step": 1370
},
{
"epoch": 0.2641690682036503,
"grad_norm": 0.9952805042266846,
"learning_rate": 0.0002639617921578681,
"loss": 2.960963821411133,
"step": 1375
},
{
"epoch": 0.26512968299711814,
"grad_norm": 0.9213324189186096,
"learning_rate": 0.0002636513374924908,
"loss": 2.9607002258300783,
"step": 1380
},
{
"epoch": 0.26609029779058596,
"grad_norm": 0.6715449690818787,
"learning_rate": 0.00026333973557988923,
"loss": 2.9600353240966797,
"step": 1385
},
{
"epoch": 0.2670509125840538,
"grad_norm": 0.8836393356323242,
"learning_rate": 0.0002630269895655119,
"loss": 2.954792785644531,
"step": 1390
},
{
"epoch": 0.2680115273775216,
"grad_norm": 0.8318425416946411,
"learning_rate": 0.00026271310260635633,
"loss": 2.9616050720214844,
"step": 1395
},
{
"epoch": 0.2689721421709894,
"grad_norm": 0.6590518355369568,
"learning_rate": 0.0002623980778709374,
"loss": 2.957061004638672,
"step": 1400
},
{
"epoch": 0.26993275696445723,
"grad_norm": 0.8581438660621643,
"learning_rate": 0.0002620819185392551,
"loss": 2.9538414001464846,
"step": 1405
},
{
"epoch": 0.27089337175792505,
"grad_norm": 0.9463372230529785,
"learning_rate": 0.00026176462780276246,
"loss": 2.9602792739868162,
"step": 1410
},
{
"epoch": 0.27185398655139287,
"grad_norm": 0.9935978651046753,
"learning_rate": 0.0002614462088643336,
"loss": 2.9563518524169923,
"step": 1415
},
{
"epoch": 0.2728146013448607,
"grad_norm": 0.9396317601203918,
"learning_rate": 0.00026112666493823103,
"loss": 2.9529541015625,
"step": 1420
},
{
"epoch": 0.2737752161383285,
"grad_norm": 0.8872493505477905,
"learning_rate": 0.00026080599925007355,
"loss": 2.9572736740112306,
"step": 1425
},
{
"epoch": 0.2747358309317964,
"grad_norm": 0.9465567469596863,
"learning_rate": 0.00026048421503680337,
"loss": 2.9554428100585937,
"step": 1430
},
{
"epoch": 0.2756964457252642,
"grad_norm": 0.9301679134368896,
"learning_rate": 0.00026016131554665377,
"loss": 2.951742935180664,
"step": 1435
},
{
"epoch": 0.276657060518732,
"grad_norm": 0.805698812007904,
"learning_rate": 0.000259837304039116,
"loss": 2.953638458251953,
"step": 1440
},
{
"epoch": 0.27761767531219983,
"grad_norm": 0.9713769555091858,
"learning_rate": 0.0002595121837849065,
"loss": 2.9541744232177733,
"step": 1445
},
{
"epoch": 0.27857829010566765,
"grad_norm": 0.6296630501747131,
"learning_rate": 0.000259185958065934,
"loss": 2.9497676849365235,
"step": 1450
},
{
"epoch": 0.27953890489913547,
"grad_norm": 0.6616190075874329,
"learning_rate": 0.00025885863017526613,
"loss": 2.949513626098633,
"step": 1455
},
{
"epoch": 0.2804995196926033,
"grad_norm": 0.845889151096344,
"learning_rate": 0.00025853020341709646,
"loss": 2.9539087295532225,
"step": 1460
},
{
"epoch": 0.2814601344860711,
"grad_norm": 0.8943501710891724,
"learning_rate": 0.000258200681106711,
"loss": 2.9553314208984376,
"step": 1465
},
{
"epoch": 0.2824207492795389,
"grad_norm": 1.1197773218154907,
"learning_rate": 0.00025787006657045477,
"loss": 2.9485729217529295,
"step": 1470
},
{
"epoch": 0.28338136407300674,
"grad_norm": 0.7538876533508301,
"learning_rate": 0.000257538363145698,
"loss": 2.951329803466797,
"step": 1475
},
{
"epoch": 0.28434197886647455,
"grad_norm": 0.8518261313438416,
"learning_rate": 0.00025720557418080304,
"loss": 2.951564407348633,
"step": 1480
},
{
"epoch": 0.28530259365994237,
"grad_norm": 0.9422370195388794,
"learning_rate": 0.00025687170303508977,
"loss": 2.9509925842285156,
"step": 1485
},
{
"epoch": 0.2862632084534102,
"grad_norm": 0.793566107749939,
"learning_rate": 0.00025653675307880225,
"loss": 2.9440481185913088,
"step": 1490
},
{
"epoch": 0.287223823246878,
"grad_norm": 1.0476356744766235,
"learning_rate": 0.00025620072769307463,
"loss": 2.9432241439819338,
"step": 1495
},
{
"epoch": 0.2881844380403458,
"grad_norm": 1.1108616590499878,
"learning_rate": 0.00025586363026989677,
"loss": 2.9504762649536134,
"step": 1500
},
{
"epoch": 0.28914505283381364,
"grad_norm": 0.5834405422210693,
"learning_rate": 0.0002555254642120802,
"loss": 2.946168899536133,
"step": 1505
},
{
"epoch": 0.29010566762728146,
"grad_norm": 1.21562922000885,
"learning_rate": 0.0002551862329332238,
"loss": 2.946132469177246,
"step": 1510
},
{
"epoch": 0.2910662824207493,
"grad_norm": 0.9039652943611145,
"learning_rate": 0.0002548459398576791,
"loss": 2.9483963012695313,
"step": 1515
},
{
"epoch": 0.2920268972142171,
"grad_norm": 0.7553810477256775,
"learning_rate": 0.00025450458842051616,
"loss": 2.944832611083984,
"step": 1520
},
{
"epoch": 0.2929875120076849,
"grad_norm": 0.7280674576759338,
"learning_rate": 0.0002541621820674882,
"loss": 2.942336082458496,
"step": 1525
},
{
"epoch": 0.29394812680115273,
"grad_norm": 0.7404939532279968,
"learning_rate": 0.0002538187242549976,
"loss": 2.942631721496582,
"step": 1530
},
{
"epoch": 0.29490874159462055,
"grad_norm": 0.9760322570800781,
"learning_rate": 0.00025347421845006056,
"loss": 2.9433094024658204,
"step": 1535
},
{
"epoch": 0.29586935638808837,
"grad_norm": 0.9803165793418884,
"learning_rate": 0.00025312866813027195,
"loss": 2.9410396575927735,
"step": 1540
},
{
"epoch": 0.2968299711815562,
"grad_norm": 0.839809000492096,
"learning_rate": 0.0002527820767837708,
"loss": 2.943090057373047,
"step": 1545
},
{
"epoch": 0.297790585975024,
"grad_norm": 0.6867128610610962,
"learning_rate": 0.00025243444790920447,
"loss": 2.9385210037231446,
"step": 1550
},
{
"epoch": 0.2987512007684918,
"grad_norm": 0.6075034737586975,
"learning_rate": 0.0002520857850156936,
"loss": 2.9380651473999024,
"step": 1555
},
{
"epoch": 0.29971181556195964,
"grad_norm": 0.8469932079315186,
"learning_rate": 0.0002517360916227968,
"loss": 2.937509536743164,
"step": 1560
},
{
"epoch": 0.30067243035542746,
"grad_norm": 0.8804566264152527,
"learning_rate": 0.000251385371260475,
"loss": 2.942656326293945,
"step": 1565
},
{
"epoch": 0.3016330451488953,
"grad_norm": 0.787070095539093,
"learning_rate": 0.0002510336274690557,
"loss": 2.934864807128906,
"step": 1570
},
{
"epoch": 0.3025936599423631,
"grad_norm": 0.9216657280921936,
"learning_rate": 0.0002506808637991974,
"loss": 2.9392667770385743,
"step": 1575
},
{
"epoch": 0.3035542747358309,
"grad_norm": 0.8175917863845825,
"learning_rate": 0.0002503270838118537,
"loss": 2.9393451690673826,
"step": 1580
},
{
"epoch": 0.3045148895292987,
"grad_norm": 0.8747788667678833,
"learning_rate": 0.0002499722910782374,
"loss": 2.93638916015625,
"step": 1585
},
{
"epoch": 0.30547550432276654,
"grad_norm": 0.9733981490135193,
"learning_rate": 0.0002496164891797844,
"loss": 2.9394798278808594,
"step": 1590
},
{
"epoch": 0.30643611911623436,
"grad_norm": 0.773009181022644,
"learning_rate": 0.0002492596817081175,
"loss": 2.9367996215820313,
"step": 1595
},
{
"epoch": 0.30739673390970224,
"grad_norm": 0.9535477161407471,
"learning_rate": 0.0002489018722650103,
"loss": 2.936510467529297,
"step": 1600
},
{
"epoch": 0.30835734870317005,
"grad_norm": 0.7678182125091553,
"learning_rate": 0.0002485430644623507,
"loss": 2.937050628662109,
"step": 1605
},
{
"epoch": 0.30931796349663787,
"grad_norm": 0.9366130828857422,
"learning_rate": 0.00024818326192210447,
"loss": 2.9339839935302736,
"step": 1610
},
{
"epoch": 0.3102785782901057,
"grad_norm": 0.7543423175811768,
"learning_rate": 0.0002478224682762787,
"loss": 2.931721496582031,
"step": 1615
},
{
"epoch": 0.3112391930835735,
"grad_norm": 1.0019062757492065,
"learning_rate": 0.0002474606871668852,
"loss": 2.9345191955566405,
"step": 1620
},
{
"epoch": 0.3121998078770413,
"grad_norm": 0.6927400827407837,
"learning_rate": 0.00024709792224590356,
"loss": 2.9311363220214846,
"step": 1625
},
{
"epoch": 0.31316042267050914,
"grad_norm": 0.9281826019287109,
"learning_rate": 0.0002467341771752446,
"loss": 2.9298322677612303,
"step": 1630
},
{
"epoch": 0.31412103746397696,
"grad_norm": 0.7644611597061157,
"learning_rate": 0.000246369455626713,
"loss": 2.9329246520996093,
"step": 1635
},
{
"epoch": 0.3150816522574448,
"grad_norm": 0.9157893657684326,
"learning_rate": 0.00024600376128197047,
"loss": 2.928366851806641,
"step": 1640
},
{
"epoch": 0.3160422670509126,
"grad_norm": 0.8347942233085632,
"learning_rate": 0.00024563709783249877,
"loss": 2.9295623779296873,
"step": 1645
},
{
"epoch": 0.3170028818443804,
"grad_norm": 0.9679508805274963,
"learning_rate": 0.00024526946897956194,
"loss": 2.9315326690673826,
"step": 1650
},
{
"epoch": 0.31796349663784823,
"grad_norm": 0.8744584321975708,
"learning_rate": 0.00024490087843416947,
"loss": 2.929412078857422,
"step": 1655
},
{
"epoch": 0.31892411143131605,
"grad_norm": 0.7325993180274963,
"learning_rate": 0.00024453132991703844,
"loss": 2.9305303573608397,
"step": 1660
},
{
"epoch": 0.31988472622478387,
"grad_norm": 0.8605995178222656,
"learning_rate": 0.00024416082715855627,
"loss": 2.9294113159179687,
"step": 1665
},
{
"epoch": 0.3208453410182517,
"grad_norm": 0.9381006360054016,
"learning_rate": 0.00024378937389874276,
"loss": 2.9343215942382814,
"step": 1670
},
{
"epoch": 0.3218059558117195,
"grad_norm": 0.8204743266105652,
"learning_rate": 0.0002434169738872126,
"loss": 2.9295400619506835,
"step": 1675
},
{
"epoch": 0.3227665706051873,
"grad_norm": 0.8801187872886658,
"learning_rate": 0.0002430436308831374,
"loss": 2.9252452850341797,
"step": 1680
},
{
"epoch": 0.32372718539865514,
"grad_norm": 0.801816999912262,
"learning_rate": 0.00024266934865520767,
"loss": 2.929386329650879,
"step": 1685
},
{
"epoch": 0.32468780019212296,
"grad_norm": 0.9650601744651794,
"learning_rate": 0.00024229413098159506,
"loss": 2.9256067276000977,
"step": 1690
},
{
"epoch": 0.3256484149855908,
"grad_norm": 0.8740299344062805,
"learning_rate": 0.00024191798164991378,
"loss": 2.926050567626953,
"step": 1695
},
{
"epoch": 0.3266090297790586,
"grad_norm": 0.6990634799003601,
"learning_rate": 0.0002415409044571828,
"loss": 2.923676300048828,
"step": 1700
},
{
"epoch": 0.3275696445725264,
"grad_norm": 0.8076874613761902,
"learning_rate": 0.00024116290320978724,
"loss": 2.9225109100341795,
"step": 1705
},
{
"epoch": 0.3285302593659942,
"grad_norm": 0.851506233215332,
"learning_rate": 0.00024078398172344006,
"loss": 2.9278465270996095,
"step": 1710
},
{
"epoch": 0.32949087415946204,
"grad_norm": 0.7479684948921204,
"learning_rate": 0.00024040414382314358,
"loss": 2.9235177993774415,
"step": 1715
},
{
"epoch": 0.33045148895292986,
"grad_norm": 0.9311428070068359,
"learning_rate": 0.00024002339334315066,
"loss": 2.927138900756836,
"step": 1720
},
{
"epoch": 0.3314121037463977,
"grad_norm": 0.6650685667991638,
"learning_rate": 0.00023964173412692631,
"loss": 2.9230600357055665,
"step": 1725
},
{
"epoch": 0.3323727185398655,
"grad_norm": 0.7150014638900757,
"learning_rate": 0.00023925917002710865,
"loss": 2.9212249755859374,
"step": 1730
},
{
"epoch": 0.3333333333333333,
"grad_norm": 0.8817017674446106,
"learning_rate": 0.0002388757049054701,
"loss": 2.925118064880371,
"step": 1735
},
{
"epoch": 0.33429394812680113,
"grad_norm": 0.8242185711860657,
"learning_rate": 0.00023849134263287836,
"loss": 2.9193931579589845,
"step": 1740
},
{
"epoch": 0.33525456292026895,
"grad_norm": 0.8570958375930786,
"learning_rate": 0.00023810608708925755,
"loss": 2.9213741302490233,
"step": 1745
},
{
"epoch": 0.33621517771373677,
"grad_norm": 0.9741989970207214,
"learning_rate": 0.00023771994216354857,
"loss": 2.9224111557006838,
"step": 1750
},
{
"epoch": 0.3371757925072046,
"grad_norm": 0.7361451983451843,
"learning_rate": 0.00023733291175367046,
"loss": 2.9207143783569336,
"step": 1755
},
{
"epoch": 0.3381364073006724,
"grad_norm": 0.7350742816925049,
"learning_rate": 0.00023694499976648043,
"loss": 2.92266960144043,
"step": 1760
},
{
"epoch": 0.3390970220941403,
"grad_norm": 0.9980021119117737,
"learning_rate": 0.0002365562101177349,
"loss": 2.9178647994995117,
"step": 1765
},
{
"epoch": 0.3400576368876081,
"grad_norm": 0.897578775882721,
"learning_rate": 0.00023616654673204983,
"loss": 2.9217491149902344,
"step": 1770
},
{
"epoch": 0.3410182516810759,
"grad_norm": 0.6844500303268433,
"learning_rate": 0.00023577601354286094,
"loss": 2.923050308227539,
"step": 1775
},
{
"epoch": 0.34197886647454373,
"grad_norm": 0.8132905960083008,
"learning_rate": 0.00023538461449238422,
"loss": 2.919468307495117,
"step": 1780
},
{
"epoch": 0.34293948126801155,
"grad_norm": 1.028024435043335,
"learning_rate": 0.00023499235353157603,
"loss": 2.9254117965698243,
"step": 1785
},
{
"epoch": 0.34390009606147937,
"grad_norm": 0.6290096640586853,
"learning_rate": 0.0002345992346200932,
"loss": 2.915414810180664,
"step": 1790
},
{
"epoch": 0.3448607108549472,
"grad_norm": 0.8899547457695007,
"learning_rate": 0.00023420526172625316,
"loss": 2.9221290588378905,
"step": 1795
},
{
"epoch": 0.345821325648415,
"grad_norm": 0.8772971034049988,
"learning_rate": 0.00023381043882699377,
"loss": 2.919437026977539,
"step": 1800
},
{
"epoch": 0.3467819404418828,
"grad_norm": 0.7739781737327576,
"learning_rate": 0.0002334147699078333,
"loss": 2.914821815490723,
"step": 1805
},
{
"epoch": 0.34774255523535064,
"grad_norm": 0.8273237347602844,
"learning_rate": 0.00023301825896282992,
"loss": 2.913847732543945,
"step": 1810
},
{
"epoch": 0.34870317002881845,
"grad_norm": 0.820066511631012,
"learning_rate": 0.00023262090999454194,
"loss": 2.916702651977539,
"step": 1815
},
{
"epoch": 0.34966378482228627,
"grad_norm": 0.7784633040428162,
"learning_rate": 0.00023222272701398664,
"loss": 2.9117355346679688,
"step": 1820
},
{
"epoch": 0.3506243996157541,
"grad_norm": 0.7284680008888245,
"learning_rate": 0.00023182371404060047,
"loss": 2.9168115615844727,
"step": 1825
},
{
"epoch": 0.3515850144092219,
"grad_norm": 0.9456787109375,
"learning_rate": 0.00023142387510219814,
"loss": 2.909496879577637,
"step": 1830
},
{
"epoch": 0.3525456292026897,
"grad_norm": 0.7613958716392517,
"learning_rate": 0.00023102321423493192,
"loss": 2.914762496948242,
"step": 1835
},
{
"epoch": 0.35350624399615754,
"grad_norm": 0.6795462965965271,
"learning_rate": 0.00023062173548325112,
"loss": 2.9124961853027345,
"step": 1840
},
{
"epoch": 0.35446685878962536,
"grad_norm": 1.093863844871521,
"learning_rate": 0.0002302194428998611,
"loss": 2.9172582626342773,
"step": 1845
},
{
"epoch": 0.3554274735830932,
"grad_norm": 1.0343449115753174,
"learning_rate": 0.0002298163405456824,
"loss": 2.913264274597168,
"step": 1850
},
{
"epoch": 0.356388088376561,
"grad_norm": 0.6599006056785583,
"learning_rate": 0.00022941243248980966,
"loss": 2.9158676147460936,
"step": 1855
},
{
"epoch": 0.3573487031700288,
"grad_norm": 0.8203375935554504,
"learning_rate": 0.0002290077228094708,
"loss": 2.9108463287353517,
"step": 1860
},
{
"epoch": 0.35830931796349663,
"grad_norm": 0.751015841960907,
"learning_rate": 0.00022860221558998554,
"loss": 2.9137802124023438,
"step": 1865
},
{
"epoch": 0.35926993275696445,
"grad_norm": 0.7769482731819153,
"learning_rate": 0.00022819591492472438,
"loss": 2.911220169067383,
"step": 1870
},
{
"epoch": 0.36023054755043227,
"grad_norm": 0.8420271277427673,
"learning_rate": 0.00022778882491506725,
"loss": 2.909513473510742,
"step": 1875
},
{
"epoch": 0.3611911623439001,
"grad_norm": 0.9250156879425049,
"learning_rate": 0.00022738094967036208,
"loss": 2.914291191101074,
"step": 1880
},
{
"epoch": 0.3621517771373679,
"grad_norm": 0.8891603350639343,
"learning_rate": 0.00022697229330788312,
"loss": 2.9125133514404298,
"step": 1885
},
{
"epoch": 0.3631123919308357,
"grad_norm": 0.6370509266853333,
"learning_rate": 0.00022656285995278984,
"loss": 2.9118471145629883,
"step": 1890
},
{
"epoch": 0.36407300672430354,
"grad_norm": 0.7550172209739685,
"learning_rate": 0.00022615265373808488,
"loss": 2.9111328125,
"step": 1895
},
{
"epoch": 0.36503362151777136,
"grad_norm": 0.898451030254364,
"learning_rate": 0.00022574167880457245,
"loss": 2.9084396362304688,
"step": 1900
},
{
"epoch": 0.3659942363112392,
"grad_norm": 0.7411810755729675,
"learning_rate": 0.00022532993930081668,
"loss": 2.9065261840820313,
"step": 1905
},
{
"epoch": 0.366954851104707,
"grad_norm": 0.785703718662262,
"learning_rate": 0.00022491743938309936,
"loss": 2.9086023330688477,
"step": 1910
},
{
"epoch": 0.3679154658981748,
"grad_norm": 0.7651037573814392,
"learning_rate": 0.0002245041832153786,
"loss": 2.908749008178711,
"step": 1915
},
{
"epoch": 0.3688760806916426,
"grad_norm": 1.0927786827087402,
"learning_rate": 0.0002240901749692461,
"loss": 2.9079204559326173,
"step": 1920
},
{
"epoch": 0.36983669548511044,
"grad_norm": 0.7358237504959106,
"learning_rate": 0.00022367541882388554,
"loss": 2.9088617324829102,
"step": 1925
},
{
"epoch": 0.37079731027857826,
"grad_norm": 0.7084841728210449,
"learning_rate": 0.00022325991896603018,
"loss": 2.9067033767700194,
"step": 1930
},
{
"epoch": 0.37175792507204614,
"grad_norm": 0.6146923303604126,
"learning_rate": 0.00022284367958992065,
"loss": 2.905855178833008,
"step": 1935
},
{
"epoch": 0.37271853986551395,
"grad_norm": 0.9858738780021667,
"learning_rate": 0.0002224267048972627,
"loss": 2.909621810913086,
"step": 1940
},
{
"epoch": 0.37367915465898177,
"grad_norm": 0.779106080532074,
"learning_rate": 0.00022200899909718456,
"loss": 2.907650375366211,
"step": 1945
},
{
"epoch": 0.3746397694524496,
"grad_norm": 0.6753861308097839,
"learning_rate": 0.00022159056640619457,
"loss": 2.901345443725586,
"step": 1950
},
{
"epoch": 0.3756003842459174,
"grad_norm": 1.005218267440796,
"learning_rate": 0.00022117141104813876,
"loss": 2.9051822662353515,
"step": 1955
},
{
"epoch": 0.3765609990393852,
"grad_norm": 0.8823059797286987,
"learning_rate": 0.000220751537254158,
"loss": 2.9064775466918946,
"step": 1960
},
{
"epoch": 0.37752161383285304,
"grad_norm": 0.7537618279457092,
"learning_rate": 0.00022033094926264548,
"loss": 2.904936408996582,
"step": 1965
},
{
"epoch": 0.37848222862632086,
"grad_norm": 0.6074866652488708,
"learning_rate": 0.00021990965131920358,
"loss": 2.901531219482422,
"step": 1970
},
{
"epoch": 0.3794428434197887,
"grad_norm": 0.7387614846229553,
"learning_rate": 0.0002194876476766015,
"loss": 2.901895523071289,
"step": 1975
},
{
"epoch": 0.3804034582132565,
"grad_norm": 0.9459737539291382,
"learning_rate": 0.00021906494259473196,
"loss": 2.9033708572387695,
"step": 1980
},
{
"epoch": 0.3813640730067243,
"grad_norm": 0.8098238706588745,
"learning_rate": 0.00021864154034056832,
"loss": 2.8992437362670898,
"step": 1985
},
{
"epoch": 0.38232468780019213,
"grad_norm": 0.8918400406837463,
"learning_rate": 0.00021821744518812154,
"loss": 2.900362014770508,
"step": 1990
},
{
"epoch": 0.38328530259365995,
"grad_norm": 0.8188735246658325,
"learning_rate": 0.00021779266141839699,
"loss": 2.8992008209228515,
"step": 1995
},
{
"epoch": 0.38424591738712777,
"grad_norm": 0.9960513114929199,
"learning_rate": 0.00021736719331935127,
"loss": 2.900490379333496,
"step": 2000
},
{
"epoch": 0.3852065321805956,
"grad_norm": 0.8250312805175781,
"learning_rate": 0.00021694104518584886,
"loss": 2.9021739959716797,
"step": 2005
},
{
"epoch": 0.3861671469740634,
"grad_norm": 0.5508167743682861,
"learning_rate": 0.00021651422131961884,
"loss": 2.9001636505126953,
"step": 2010
},
{
"epoch": 0.3871277617675312,
"grad_norm": 0.9340776205062866,
"learning_rate": 0.0002160867260292115,
"loss": 2.898966407775879,
"step": 2015
},
{
"epoch": 0.38808837656099904,
"grad_norm": 1.0545355081558228,
"learning_rate": 0.00021565856362995464,
"loss": 2.9000267028808593,
"step": 2020
},
{
"epoch": 0.38904899135446686,
"grad_norm": 0.8669793605804443,
"learning_rate": 0.00021522973844391024,
"loss": 2.898406982421875,
"step": 2025
},
{
"epoch": 0.3900096061479347,
"grad_norm": 0.706919252872467,
"learning_rate": 0.00021480025479983077,
"loss": 2.8980600357055666,
"step": 2030
},
{
"epoch": 0.3909702209414025,
"grad_norm": 0.6402170658111572,
"learning_rate": 0.00021437011703311545,
"loss": 2.8995635986328123,
"step": 2035
},
{
"epoch": 0.3919308357348703,
"grad_norm": 0.902797281742096,
"learning_rate": 0.0002139393294857665,
"loss": 2.902130889892578,
"step": 2040
},
{
"epoch": 0.3928914505283381,
"grad_norm": 0.9768553376197815,
"learning_rate": 0.00021350789650634525,
"loss": 2.898158645629883,
"step": 2045
},
{
"epoch": 0.39385206532180594,
"grad_norm": 0.8208462595939636,
"learning_rate": 0.00021307582244992838,
"loss": 2.9008291244506834,
"step": 2050
},
{
"epoch": 0.39481268011527376,
"grad_norm": 0.6077558398246765,
"learning_rate": 0.0002126431116780639,
"loss": 2.8985877990722657,
"step": 2055
},
{
"epoch": 0.3957732949087416,
"grad_norm": 0.7051781415939331,
"learning_rate": 0.00021220976855872712,
"loss": 2.897053527832031,
"step": 2060
},
{
"epoch": 0.3967339097022094,
"grad_norm": 0.738699734210968,
"learning_rate": 0.00021177579746627643,
"loss": 2.8972637176513674,
"step": 2065
},
{
"epoch": 0.3976945244956772,
"grad_norm": 0.9454274773597717,
"learning_rate": 0.0002113412027814094,
"loss": 2.8959421157836913,
"step": 2070
},
{
"epoch": 0.39865513928914503,
"grad_norm": 0.7538549304008484,
"learning_rate": 0.0002109059888911183,
"loss": 2.8942508697509766,
"step": 2075
},
{
"epoch": 0.39961575408261285,
"grad_norm": 0.64174884557724,
"learning_rate": 0.00021047016018864602,
"loss": 2.8981069564819335,
"step": 2080
},
{
"epoch": 0.40057636887608067,
"grad_norm": 0.7245832085609436,
"learning_rate": 0.00021003372107344167,
"loss": 2.896715545654297,
"step": 2085
},
{
"epoch": 0.4015369836695485,
"grad_norm": 0.7303643226623535,
"learning_rate": 0.00020959667595111603,
"loss": 2.90020637512207,
"step": 2090
},
{
"epoch": 0.4024975984630163,
"grad_norm": 1.070908784866333,
"learning_rate": 0.00020915902923339722,
"loss": 2.897067642211914,
"step": 2095
},
{
"epoch": 0.4034582132564842,
"grad_norm": 0.7779847383499146,
"learning_rate": 0.0002087207853380862,
"loss": 2.8957414627075195,
"step": 2100
},
{
"epoch": 0.404418828049952,
"grad_norm": 0.6650077104568481,
"learning_rate": 0.00020828194868901205,
"loss": 2.892515182495117,
"step": 2105
},
{
"epoch": 0.4053794428434198,
"grad_norm": 0.9725134968757629,
"learning_rate": 0.00020784252371598732,
"loss": 2.8961185455322265,
"step": 2110
},
{
"epoch": 0.40634005763688763,
"grad_norm": 0.6921088695526123,
"learning_rate": 0.00020740251485476345,
"loss": 2.893880844116211,
"step": 2115
},
{
"epoch": 0.40730067243035545,
"grad_norm": 0.6224314570426941,
"learning_rate": 0.00020696192654698592,
"loss": 2.891771697998047,
"step": 2120
},
{
"epoch": 0.40826128722382327,
"grad_norm": 0.7067698240280151,
"learning_rate": 0.00020652076324014927,
"loss": 2.8919609069824217,
"step": 2125
},
{
"epoch": 0.4092219020172911,
"grad_norm": 0.9510787725448608,
"learning_rate": 0.00020607902938755252,
"loss": 2.8964000701904298,
"step": 2130
},
{
"epoch": 0.4101825168107589,
"grad_norm": 0.6084873676300049,
"learning_rate": 0.00020563672944825392,
"loss": 2.8933223724365233,
"step": 2135
},
{
"epoch": 0.4111431316042267,
"grad_norm": 0.6090896725654602,
"learning_rate": 0.00020519386788702602,
"loss": 2.88956298828125,
"step": 2140
},
{
"epoch": 0.41210374639769454,
"grad_norm": 1.0817956924438477,
"learning_rate": 0.0002047504491743107,
"loss": 2.8921178817749023,
"step": 2145
},
{
"epoch": 0.41306436119116235,
"grad_norm": 0.7050410509109497,
"learning_rate": 0.000204306477786174,
"loss": 2.8883323669433594,
"step": 2150
},
{
"epoch": 0.4140249759846302,
"grad_norm": 0.9106734991073608,
"learning_rate": 0.00020386195820426082,
"loss": 2.8929920196533203,
"step": 2155
},
{
"epoch": 0.414985590778098,
"grad_norm": 0.7352846264839172,
"learning_rate": 0.00020341689491574984,
"loss": 2.891267776489258,
"step": 2160
},
{
"epoch": 0.4159462055715658,
"grad_norm": 0.6910091042518616,
"learning_rate": 0.00020297129241330817,
"loss": 2.890258026123047,
"step": 2165
},
{
"epoch": 0.4169068203650336,
"grad_norm": 0.7590439319610596,
"learning_rate": 0.00020252515519504592,
"loss": 2.8909364700317384,
"step": 2170
},
{
"epoch": 0.41786743515850144,
"grad_norm": 0.6669980883598328,
"learning_rate": 0.0002020784877644709,
"loss": 2.890909194946289,
"step": 2175
},
{
"epoch": 0.41882804995196926,
"grad_norm": 0.7879086136817932,
"learning_rate": 0.00020163129463044308,
"loss": 2.888330841064453,
"step": 2180
},
{
"epoch": 0.4197886647454371,
"grad_norm": 0.799680769443512,
"learning_rate": 0.0002011835803071292,
"loss": 2.888754653930664,
"step": 2185
},
{
"epoch": 0.4207492795389049,
"grad_norm": 0.8745085597038269,
"learning_rate": 0.00020073534931395697,
"loss": 2.8860883712768555,
"step": 2190
},
{
"epoch": 0.4217098943323727,
"grad_norm": 0.7884883284568787,
"learning_rate": 0.00020028660617556985,
"loss": 2.8914257049560548,
"step": 2195
},
{
"epoch": 0.42267050912584053,
"grad_norm": 0.6534205079078674,
"learning_rate": 0.00019983735542178086,
"loss": 2.887241744995117,
"step": 2200
},
{
"epoch": 0.42363112391930835,
"grad_norm": 0.9296805262565613,
"learning_rate": 0.00019938760158752725,
"loss": 2.8888803482055665,
"step": 2205
},
{
"epoch": 0.42459173871277617,
"grad_norm": 1.0822457075119019,
"learning_rate": 0.00019893734921282448,
"loss": 2.8933456420898436,
"step": 2210
},
{
"epoch": 0.425552353506244,
"grad_norm": 0.8483099937438965,
"learning_rate": 0.0001984866028427207,
"loss": 2.885776901245117,
"step": 2215
},
{
"epoch": 0.4265129682997118,
"grad_norm": 0.7693598866462708,
"learning_rate": 0.00019803536702725044,
"loss": 2.8912368774414063,
"step": 2220
},
{
"epoch": 0.4274735830931796,
"grad_norm": 0.6365734338760376,
"learning_rate": 0.00019758364632138908,
"loss": 2.88284912109375,
"step": 2225
},
{
"epoch": 0.42843419788664744,
"grad_norm": 0.8677585124969482,
"learning_rate": 0.0001971314452850066,
"loss": 2.8882091522216795,
"step": 2230
},
{
"epoch": 0.42939481268011526,
"grad_norm": 0.8231710195541382,
"learning_rate": 0.00019667876848282167,
"loss": 2.88732795715332,
"step": 2235
},
{
"epoch": 0.4303554274735831,
"grad_norm": 0.5195154547691345,
"learning_rate": 0.0001962256204843556,
"loss": 2.8842351913452147,
"step": 2240
},
{
"epoch": 0.4313160422670509,
"grad_norm": 0.7365444302558899,
"learning_rate": 0.00019577200586388618,
"loss": 2.8888904571533205,
"step": 2245
},
{
"epoch": 0.4322766570605187,
"grad_norm": 0.6714667081832886,
"learning_rate": 0.00019531792920040133,
"loss": 2.888100814819336,
"step": 2250
},
{
"epoch": 0.4332372718539865,
"grad_norm": 0.8434675335884094,
"learning_rate": 0.0001948633950775532,
"loss": 2.8872095108032227,
"step": 2255
},
{
"epoch": 0.43419788664745435,
"grad_norm": 0.7106018662452698,
"learning_rate": 0.00019440840808361174,
"loss": 2.8836999893188477,
"step": 2260
},
{
"epoch": 0.43515850144092216,
"grad_norm": 0.967634379863739,
"learning_rate": 0.00019395297281141828,
"loss": 2.8807979583740235,
"step": 2265
},
{
"epoch": 0.43611911623439004,
"grad_norm": 0.6999104022979736,
"learning_rate": 0.0001934970938583393,
"loss": 2.887973403930664,
"step": 2270
},
{
"epoch": 0.43707973102785785,
"grad_norm": 0.6316190958023071,
"learning_rate": 0.00019304077582622003,
"loss": 2.8830514907836915,
"step": 2275
},
{
"epoch": 0.43804034582132567,
"grad_norm": 0.990618884563446,
"learning_rate": 0.00019258402332133798,
"loss": 2.890216827392578,
"step": 2280
},
{
"epoch": 0.4390009606147935,
"grad_norm": 0.7043749690055847,
"learning_rate": 0.00019212684095435637,
"loss": 2.8833900451660157,
"step": 2285
},
{
"epoch": 0.4399615754082613,
"grad_norm": 0.7047979235649109,
"learning_rate": 0.00019166923334027765,
"loss": 2.884984588623047,
"step": 2290
},
{
"epoch": 0.4409221902017291,
"grad_norm": 0.5849128365516663,
"learning_rate": 0.00019121120509839692,
"loss": 2.8848838806152344,
"step": 2295
},
{
"epoch": 0.44188280499519694,
"grad_norm": 0.8806740045547485,
"learning_rate": 0.0001907527608522552,
"loss": 2.8797868728637694,
"step": 2300
},
{
"epoch": 0.44284341978866476,
"grad_norm": 0.7895880341529846,
"learning_rate": 0.000190293905229593,
"loss": 2.8813777923583985,
"step": 2305
},
{
"epoch": 0.4438040345821326,
"grad_norm": 1.0096721649169922,
"learning_rate": 0.00018983464286230327,
"loss": 2.8826440811157226,
"step": 2310
},
{
"epoch": 0.4447646493756004,
"grad_norm": 0.6461132764816284,
"learning_rate": 0.00018937497838638509,
"loss": 2.881867599487305,
"step": 2315
},
{
"epoch": 0.4457252641690682,
"grad_norm": 0.8887806534767151,
"learning_rate": 0.0001889149164418963,
"loss": 2.88212890625,
"step": 2320
},
{
"epoch": 0.44668587896253603,
"grad_norm": 0.6228038668632507,
"learning_rate": 0.00018845446167290705,
"loss": 2.88170166015625,
"step": 2325
},
{
"epoch": 0.44764649375600385,
"grad_norm": 0.6716971397399902,
"learning_rate": 0.0001879936187274529,
"loss": 2.8785728454589843,
"step": 2330
},
{
"epoch": 0.44860710854947167,
"grad_norm": 0.6986984014511108,
"learning_rate": 0.00018753239225748796,
"loss": 2.881254196166992,
"step": 2335
},
{
"epoch": 0.4495677233429395,
"grad_norm": 0.6882680654525757,
"learning_rate": 0.0001870707869188375,
"loss": 2.877424621582031,
"step": 2340
},
{
"epoch": 0.4505283381364073,
"grad_norm": 0.9753093123435974,
"learning_rate": 0.00018660880737115146,
"loss": 2.8826948165893556,
"step": 2345
},
{
"epoch": 0.4514889529298751,
"grad_norm": 0.6546221971511841,
"learning_rate": 0.0001861464582778572,
"loss": 2.8837528228759766,
"step": 2350
},
{
"epoch": 0.45244956772334294,
"grad_norm": 0.6572368741035461,
"learning_rate": 0.00018568374430611242,
"loss": 2.880933952331543,
"step": 2355
},
{
"epoch": 0.45341018251681076,
"grad_norm": 0.6908612847328186,
"learning_rate": 0.00018522067012675798,
"loss": 2.8826961517333984,
"step": 2360
},
{
"epoch": 0.4543707973102786,
"grad_norm": 0.7534862160682678,
"learning_rate": 0.00018475724041427106,
"loss": 2.8773014068603517,
"step": 2365
},
{
"epoch": 0.4553314121037464,
"grad_norm": 0.9261667132377625,
"learning_rate": 0.00018429345984671743,
"loss": 2.8801780700683595,
"step": 2370
},
{
"epoch": 0.4562920268972142,
"grad_norm": 0.9364944100379944,
"learning_rate": 0.0001838293331057048,
"loss": 2.880086135864258,
"step": 2375
},
{
"epoch": 0.457252641690682,
"grad_norm": 0.6954669952392578,
"learning_rate": 0.00018336486487633528,
"loss": 2.8802162170410157,
"step": 2380
},
{
"epoch": 0.45821325648414984,
"grad_norm": 0.5674038529396057,
"learning_rate": 0.000182900059847158,
"loss": 2.8759918212890625,
"step": 2385
},
{
"epoch": 0.45917387127761766,
"grad_norm": 0.558390200138092,
"learning_rate": 0.00018243492271012202,
"loss": 2.8748876571655275,
"step": 2390
},
{
"epoch": 0.4601344860710855,
"grad_norm": 0.7045537233352661,
"learning_rate": 0.00018196945816052867,
"loss": 2.8770776748657227,
"step": 2395
},
{
"epoch": 0.4610951008645533,
"grad_norm": 0.8516057729721069,
"learning_rate": 0.00018150367089698452,
"loss": 2.8759332656860352,
"step": 2400
},
{
"epoch": 0.4620557156580211,
"grad_norm": 0.6226221323013306,
"learning_rate": 0.00018103756562135373,
"loss": 2.880674362182617,
"step": 2405
},
{
"epoch": 0.46301633045148893,
"grad_norm": 0.5688340067863464,
"learning_rate": 0.0001805711470387105,
"loss": 2.874104690551758,
"step": 2410
},
{
"epoch": 0.46397694524495675,
"grad_norm": 0.5873801708221436,
"learning_rate": 0.00018010441985729183,
"loss": 2.8717109680175783,
"step": 2415
},
{
"epoch": 0.46493756003842457,
"grad_norm": 0.8382697105407715,
"learning_rate": 0.00017963738878844966,
"loss": 2.8790111541748047,
"step": 2420
},
{
"epoch": 0.4658981748318924,
"grad_norm": 0.7427976727485657,
"learning_rate": 0.00017917005854660374,
"loss": 2.8731239318847654,
"step": 2425
},
{
"epoch": 0.4668587896253602,
"grad_norm": 0.6539490818977356,
"learning_rate": 0.00017870243384919364,
"loss": 2.8791481018066407,
"step": 2430
},
{
"epoch": 0.4678194044188281,
"grad_norm": 0.8121071457862854,
"learning_rate": 0.0001782345194166314,
"loss": 2.877638053894043,
"step": 2435
},
{
"epoch": 0.4687800192122959,
"grad_norm": 0.6476060152053833,
"learning_rate": 0.00017776631997225365,
"loss": 2.8710798263549804,
"step": 2440
},
{
"epoch": 0.4697406340057637,
"grad_norm": 0.5659114122390747,
"learning_rate": 0.0001772978402422742,
"loss": 2.873188018798828,
"step": 2445
},
{
"epoch": 0.47070124879923153,
"grad_norm": 0.6289011836051941,
"learning_rate": 0.0001768290849557361,
"loss": 2.872249984741211,
"step": 2450
},
{
"epoch": 0.47166186359269935,
"grad_norm": 0.7904647588729858,
"learning_rate": 0.00017636005884446397,
"loss": 2.873155975341797,
"step": 2455
},
{
"epoch": 0.47262247838616717,
"grad_norm": 0.6264911890029907,
"learning_rate": 0.00017589076664301637,
"loss": 2.875007438659668,
"step": 2460
},
{
"epoch": 0.473583093179635,
"grad_norm": 0.8897266387939453,
"learning_rate": 0.00017542121308863776,
"loss": 2.874278259277344,
"step": 2465
},
{
"epoch": 0.4745437079731028,
"grad_norm": 0.8211498856544495,
"learning_rate": 0.00017495140292121084,
"loss": 2.875082015991211,
"step": 2470
},
{
"epoch": 0.4755043227665706,
"grad_norm": 0.6536585688591003,
"learning_rate": 0.0001744813408832088,
"loss": 2.872761535644531,
"step": 2475
},
{
"epoch": 0.47646493756003844,
"grad_norm": 0.6122596859931946,
"learning_rate": 0.0001740110317196472,
"loss": 2.873042678833008,
"step": 2480
},
{
"epoch": 0.47742555235350626,
"grad_norm": 0.6163274645805359,
"learning_rate": 0.0001735404801780362,
"loss": 2.8703638076782227,
"step": 2485
},
{
"epoch": 0.4783861671469741,
"grad_norm": 0.8691967129707336,
"learning_rate": 0.0001730696910083326,
"loss": 2.873611259460449,
"step": 2490
},
{
"epoch": 0.4793467819404419,
"grad_norm": 0.7458942532539368,
"learning_rate": 0.0001725986689628921,
"loss": 2.871821403503418,
"step": 2495
},
{
"epoch": 0.4803073967339097,
"grad_norm": 0.6633113622665405,
"learning_rate": 0.00017212741879642096,
"loss": 2.8725128173828125,
"step": 2500
},
{
"epoch": 0.4812680115273775,
"grad_norm": 0.684518039226532,
"learning_rate": 0.00017165594526592833,
"loss": 2.874650764465332,
"step": 2505
},
{
"epoch": 0.48222862632084534,
"grad_norm": 0.6810330748558044,
"learning_rate": 0.000171184253130678,
"loss": 2.8709695816040037,
"step": 2510
},
{
"epoch": 0.48318924111431316,
"grad_norm": 0.7958024740219116,
"learning_rate": 0.00017071234715214045,
"loss": 2.872270202636719,
"step": 2515
},
{
"epoch": 0.484149855907781,
"grad_norm": 0.6600337624549866,
"learning_rate": 0.0001702402320939449,
"loss": 2.8687416076660157,
"step": 2520
},
{
"epoch": 0.4851104707012488,
"grad_norm": 0.6404789090156555,
"learning_rate": 0.00016976791272183098,
"loss": 2.865874481201172,
"step": 2525
},
{
"epoch": 0.4860710854947166,
"grad_norm": 0.5435463190078735,
"learning_rate": 0.0001692953938036008,
"loss": 2.8678897857666015,
"step": 2530
},
{
"epoch": 0.48703170028818443,
"grad_norm": 0.8794398903846741,
"learning_rate": 0.00016882268010907087,
"loss": 2.871311569213867,
"step": 2535
},
{
"epoch": 0.48799231508165225,
"grad_norm": 0.8902005553245544,
"learning_rate": 0.00016834977641002377,
"loss": 2.870677947998047,
"step": 2540
},
{
"epoch": 0.48895292987512007,
"grad_norm": 0.6775867342948914,
"learning_rate": 0.00016787668748016008,
"loss": 2.868155670166016,
"step": 2545
},
{
"epoch": 0.4899135446685879,
"grad_norm": 0.5421539545059204,
"learning_rate": 0.00016740341809505017,
"loss": 2.8710710525512697,
"step": 2550
},
{
"epoch": 0.4908741594620557,
"grad_norm": 0.5883712768554688,
"learning_rate": 0.000166929973032086,
"loss": 2.8703285217285157,
"step": 2555
},
{
"epoch": 0.4918347742555235,
"grad_norm": 0.8570505380630493,
"learning_rate": 0.0001664563570704329,
"loss": 2.8731103897094727,
"step": 2560
},
{
"epoch": 0.49279538904899134,
"grad_norm": 0.8919458985328674,
"learning_rate": 0.0001659825749909814,
"loss": 2.8725284576416015,
"step": 2565
},
{
"epoch": 0.49375600384245916,
"grad_norm": 0.6126694083213806,
"learning_rate": 0.00016550863157629888,
"loss": 2.8706518173217774,
"step": 2570
},
{
"epoch": 0.494716618635927,
"grad_norm": 0.45519402623176575,
"learning_rate": 0.00016503453161058123,
"loss": 2.8679710388183595,
"step": 2575
},
{
"epoch": 0.4956772334293948,
"grad_norm": 0.7353713512420654,
"learning_rate": 0.00016456027987960466,
"loss": 2.8688129425048827,
"step": 2580
},
{
"epoch": 0.4966378482228626,
"grad_norm": 0.8744354248046875,
"learning_rate": 0.00016408588117067743,
"loss": 2.8654415130615236,
"step": 2585
},
{
"epoch": 0.49759846301633043,
"grad_norm": 0.824774980545044,
"learning_rate": 0.00016361134027259136,
"loss": 2.8741382598876952,
"step": 2590
},
{
"epoch": 0.49855907780979825,
"grad_norm": 0.6276190876960754,
"learning_rate": 0.00016313666197557373,
"loss": 2.8700138092041017,
"step": 2595
},
{
"epoch": 0.49951969260326606,
"grad_norm": 0.5131474733352661,
"learning_rate": 0.00016266185107123864,
"loss": 2.8695526123046875,
"step": 2600
},
{
"epoch": 0.5004803073967339,
"grad_norm": 0.7874789237976074,
"learning_rate": 0.00016218691235253893,
"loss": 2.8659042358398437,
"step": 2605
},
{
"epoch": 0.5014409221902018,
"grad_norm": 0.6303560137748718,
"learning_rate": 0.0001617118506137175,
"loss": 2.8703582763671873,
"step": 2610
},
{
"epoch": 0.5024015369836695,
"grad_norm": 0.8049387335777283,
"learning_rate": 0.00016123667065025914,
"loss": 2.8671377182006834,
"step": 2615
},
{
"epoch": 0.5033621517771374,
"grad_norm": 0.8616132140159607,
"learning_rate": 0.00016076137725884218,
"loss": 2.8674648284912108,
"step": 2620
},
{
"epoch": 0.5043227665706052,
"grad_norm": 0.6320975422859192,
"learning_rate": 0.0001602859752372897,
"loss": 2.8637058258056642,
"step": 2625
},
{
"epoch": 0.505283381364073,
"grad_norm": 0.5506896376609802,
"learning_rate": 0.00015981046938452146,
"loss": 2.8641489028930662,
"step": 2630
},
{
"epoch": 0.5062439961575408,
"grad_norm": 0.6218773722648621,
"learning_rate": 0.0001593348645005054,
"loss": 2.8639698028564453,
"step": 2635
},
{
"epoch": 0.5072046109510087,
"grad_norm": 0.9080044627189636,
"learning_rate": 0.00015885916538620906,
"loss": 2.8647804260253906,
"step": 2640
},
{
"epoch": 0.5081652257444764,
"grad_norm": 0.8615972399711609,
"learning_rate": 0.00015838337684355121,
"loss": 2.8677547454833983,
"step": 2645
},
{
"epoch": 0.5091258405379443,
"grad_norm": 0.570344090461731,
"learning_rate": 0.0001579075036753533,
"loss": 2.866165542602539,
"step": 2650
},
{
"epoch": 0.5100864553314121,
"grad_norm": 0.5077627897262573,
"learning_rate": 0.00015743155068529102,
"loss": 2.8662036895751952,
"step": 2655
},
{
"epoch": 0.5110470701248799,
"grad_norm": 0.7941848635673523,
"learning_rate": 0.0001569555226778459,
"loss": 2.8637313842773438,
"step": 2660
},
{
"epoch": 0.5120076849183477,
"grad_norm": 0.9668124914169312,
"learning_rate": 0.0001564794244582567,
"loss": 2.867812156677246,
"step": 2665
},
{
"epoch": 0.5129682997118156,
"grad_norm": 0.6352429389953613,
"learning_rate": 0.0001560032608324709,
"loss": 2.8658679962158202,
"step": 2670
},
{
"epoch": 0.5139289145052833,
"grad_norm": 0.4856127202510834,
"learning_rate": 0.00015552703660709618,
"loss": 2.8620628356933593,
"step": 2675
},
{
"epoch": 0.5148895292987512,
"grad_norm": 1.1079636812210083,
"learning_rate": 0.00015505075658935207,
"loss": 2.865533447265625,
"step": 2680
},
{
"epoch": 0.515850144092219,
"grad_norm": 0.5639813542366028,
"learning_rate": 0.0001545744255870212,
"loss": 2.867781066894531,
"step": 2685
},
{
"epoch": 0.5168107588856868,
"grad_norm": 0.6282705664634705,
"learning_rate": 0.00015409804840840088,
"loss": 2.8628158569335938,
"step": 2690
},
{
"epoch": 0.5177713736791547,
"grad_norm": 0.6844084858894348,
"learning_rate": 0.0001536216298622545,
"loss": 2.865097427368164,
"step": 2695
},
{
"epoch": 0.5187319884726225,
"grad_norm": 0.7825571894645691,
"learning_rate": 0.00015314517475776318,
"loss": 2.863981246948242,
"step": 2700
},
{
"epoch": 0.5196926032660903,
"grad_norm": 0.6711815595626831,
"learning_rate": 0.00015266868790447685,
"loss": 2.862168884277344,
"step": 2705
},
{
"epoch": 0.5206532180595581,
"grad_norm": 0.5164412260055542,
"learning_rate": 0.0001521921741122661,
"loss": 2.8566638946533205,
"step": 2710
},
{
"epoch": 0.521613832853026,
"grad_norm": 0.6021407246589661,
"learning_rate": 0.00015171563819127342,
"loss": 2.85965576171875,
"step": 2715
},
{
"epoch": 0.5225744476464937,
"grad_norm": 0.8277438282966614,
"learning_rate": 0.00015123908495186464,
"loss": 2.863536071777344,
"step": 2720
},
{
"epoch": 0.5235350624399616,
"grad_norm": 0.6992095708847046,
"learning_rate": 0.0001507625192045804,
"loss": 2.866368865966797,
"step": 2725
},
{
"epoch": 0.5244956772334294,
"grad_norm": 0.6141840815544128,
"learning_rate": 0.00015028594576008773,
"loss": 2.8606613159179686,
"step": 2730
},
{
"epoch": 0.5254562920268973,
"grad_norm": 0.6499273777008057,
"learning_rate": 0.00014980936942913113,
"loss": 2.8631092071533204,
"step": 2735
},
{
"epoch": 0.526416906820365,
"grad_norm": 0.5648180842399597,
"learning_rate": 0.00014933279502248444,
"loss": 2.857931900024414,
"step": 2740
},
{
"epoch": 0.5273775216138329,
"grad_norm": 0.6316075325012207,
"learning_rate": 0.000148856227350902,
"loss": 2.8574670791625976,
"step": 2745
},
{
"epoch": 0.5283381364073007,
"grad_norm": 0.8975266218185425,
"learning_rate": 0.00014837967122507015,
"loss": 2.864010238647461,
"step": 2750
},
{
"epoch": 0.5292987512007685,
"grad_norm": 0.7359750866889954,
"learning_rate": 0.00014790313145555863,
"loss": 2.861468505859375,
"step": 2755
},
{
"epoch": 0.5302593659942363,
"grad_norm": 0.5477508902549744,
"learning_rate": 0.00014742661285277228,
"loss": 2.8594310760498045,
"step": 2760
},
{
"epoch": 0.5312199807877042,
"grad_norm": 0.6290469765663147,
"learning_rate": 0.00014695012022690205,
"loss": 2.860156440734863,
"step": 2765
},
{
"epoch": 0.5321805955811719,
"grad_norm": 0.5689098834991455,
"learning_rate": 0.00014647365838787673,
"loss": 2.8548521041870116,
"step": 2770
},
{
"epoch": 0.5331412103746398,
"grad_norm": 0.7737315893173218,
"learning_rate": 0.00014599723214531434,
"loss": 2.8601730346679686,
"step": 2775
},
{
"epoch": 0.5341018251681076,
"grad_norm": 0.7761661410331726,
"learning_rate": 0.0001455208463084737,
"loss": 2.857926940917969,
"step": 2780
},
{
"epoch": 0.5350624399615754,
"grad_norm": 0.6857746839523315,
"learning_rate": 0.00014504450568620557,
"loss": 2.8620637893676757,
"step": 2785
},
{
"epoch": 0.5360230547550432,
"grad_norm": 0.5136231780052185,
"learning_rate": 0.00014456821508690432,
"loss": 2.8587860107421874,
"step": 2790
},
{
"epoch": 0.5369836695485111,
"grad_norm": 0.5590301752090454,
"learning_rate": 0.0001440919793184595,
"loss": 2.855500030517578,
"step": 2795
},
{
"epoch": 0.5379442843419788,
"grad_norm": 0.7252662181854248,
"learning_rate": 0.00014361580318820696,
"loss": 2.8574901580810548,
"step": 2800
},
{
"epoch": 0.5389048991354467,
"grad_norm": 0.6491255760192871,
"learning_rate": 0.00014313969150288083,
"loss": 2.857944297790527,
"step": 2805
},
{
"epoch": 0.5398655139289145,
"grad_norm": 0.8175961375236511,
"learning_rate": 0.00014266364906856442,
"loss": 2.857021713256836,
"step": 2810
},
{
"epoch": 0.5408261287223823,
"grad_norm": 0.7656385898590088,
"learning_rate": 0.00014218768069064214,
"loss": 2.8591373443603514,
"step": 2815
},
{
"epoch": 0.5417867435158501,
"grad_norm": 0.7636440396308899,
"learning_rate": 0.00014171179117375082,
"loss": 2.8583093643188477,
"step": 2820
},
{
"epoch": 0.542747358309318,
"grad_norm": 0.6154850721359253,
"learning_rate": 0.00014123598532173113,
"loss": 2.859964942932129,
"step": 2825
},
{
"epoch": 0.5437079731027857,
"grad_norm": 0.49419260025024414,
"learning_rate": 0.00014076026793757943,
"loss": 2.8569236755371095,
"step": 2830
},
{
"epoch": 0.5446685878962536,
"grad_norm": 0.6245793104171753,
"learning_rate": 0.00014028464382339877,
"loss": 2.857840728759766,
"step": 2835
},
{
"epoch": 0.5456292026897214,
"grad_norm": 0.7936309576034546,
"learning_rate": 0.0001398091177803509,
"loss": 2.858272171020508,
"step": 2840
},
{
"epoch": 0.5465898174831892,
"grad_norm": 0.6912634372711182,
"learning_rate": 0.00013933369460860748,
"loss": 2.855249786376953,
"step": 2845
},
{
"epoch": 0.547550432276657,
"grad_norm": 0.4704606831073761,
"learning_rate": 0.00013885837910730168,
"loss": 2.853233528137207,
"step": 2850
},
{
"epoch": 0.5485110470701249,
"grad_norm": 0.6878814697265625,
"learning_rate": 0.00013838317607448004,
"loss": 2.856007194519043,
"step": 2855
},
{
"epoch": 0.5494716618635928,
"grad_norm": 0.7811667323112488,
"learning_rate": 0.00013790809030705354,
"loss": 2.8576026916503907,
"step": 2860
},
{
"epoch": 0.5504322766570605,
"grad_norm": 0.5080620050430298,
"learning_rate": 0.00013743312660074962,
"loss": 2.856093978881836,
"step": 2865
},
{
"epoch": 0.5513928914505284,
"grad_norm": 0.7438220977783203,
"learning_rate": 0.00013695828975006336,
"loss": 2.85654354095459,
"step": 2870
},
{
"epoch": 0.5523535062439962,
"grad_norm": 0.5278438329696655,
"learning_rate": 0.00013648358454820957,
"loss": 2.857866668701172,
"step": 2875
},
{
"epoch": 0.553314121037464,
"grad_norm": 0.5583338737487793,
"learning_rate": 0.00013600901578707402,
"loss": 2.8554916381835938,
"step": 2880
},
{
"epoch": 0.5542747358309318,
"grad_norm": 0.6105745434761047,
"learning_rate": 0.00013553458825716502,
"loss": 2.8572811126708983,
"step": 2885
},
{
"epoch": 0.5552353506243997,
"grad_norm": 0.5137814879417419,
"learning_rate": 0.0001350603067475655,
"loss": 2.856470489501953,
"step": 2890
},
{
"epoch": 0.5561959654178674,
"grad_norm": 0.6973247528076172,
"learning_rate": 0.00013458617604588418,
"loss": 2.8541500091552736,
"step": 2895
},
{
"epoch": 0.5571565802113353,
"grad_norm": 0.6607674956321716,
"learning_rate": 0.00013411220093820773,
"loss": 2.856244659423828,
"step": 2900
},
{
"epoch": 0.5581171950048031,
"grad_norm": 0.6131708025932312,
"learning_rate": 0.00013363838620905188,
"loss": 2.8549783706665037,
"step": 2905
},
{
"epoch": 0.5590778097982709,
"grad_norm": 0.5899085998535156,
"learning_rate": 0.00013316473664131347,
"loss": 2.8583316802978516,
"step": 2910
},
{
"epoch": 0.5600384245917387,
"grad_norm": 0.6078781485557556,
"learning_rate": 0.00013269125701622243,
"loss": 2.855908203125,
"step": 2915
},
{
"epoch": 0.5609990393852066,
"grad_norm": 0.5581336617469788,
"learning_rate": 0.00013221795211329281,
"loss": 2.853131866455078,
"step": 2920
},
{
"epoch": 0.5619596541786743,
"grad_norm": 0.5300495624542236,
"learning_rate": 0.00013174482671027526,
"loss": 2.8522682189941406,
"step": 2925
},
{
"epoch": 0.5629202689721422,
"grad_norm": 0.6495129466056824,
"learning_rate": 0.00013127188558310823,
"loss": 2.857743835449219,
"step": 2930
},
{
"epoch": 0.56388088376561,
"grad_norm": 0.6372506618499756,
"learning_rate": 0.0001307991335058702,
"loss": 2.8534168243408202,
"step": 2935
},
{
"epoch": 0.5648414985590778,
"grad_norm": 0.5662026405334473,
"learning_rate": 0.00013032657525073122,
"loss": 2.8524166107177735,
"step": 2940
},
{
"epoch": 0.5658021133525456,
"grad_norm": 0.5620723366737366,
"learning_rate": 0.00012985421558790473,
"loss": 2.8510919570922852,
"step": 2945
},
{
"epoch": 0.5667627281460135,
"grad_norm": 0.861655056476593,
"learning_rate": 0.00012938205928559964,
"loss": 2.8496784210205077,
"step": 2950
},
{
"epoch": 0.5677233429394812,
"grad_norm": 0.7627875804901123,
"learning_rate": 0.00012891011110997187,
"loss": 2.8515321731567385,
"step": 2955
},
{
"epoch": 0.5686839577329491,
"grad_norm": 0.600219190120697,
"learning_rate": 0.0001284383758250767,
"loss": 2.8534393310546875,
"step": 2960
},
{
"epoch": 0.5696445725264169,
"grad_norm": 0.5384979844093323,
"learning_rate": 0.00012796685819282009,
"loss": 2.8518402099609377,
"step": 2965
},
{
"epoch": 0.5706051873198847,
"grad_norm": 0.5790613293647766,
"learning_rate": 0.0001274955629729111,
"loss": 2.850631332397461,
"step": 2970
},
{
"epoch": 0.5715658021133525,
"grad_norm": 0.7250127196311951,
"learning_rate": 0.0001270244949228136,
"loss": 2.8556766510009766,
"step": 2975
},
{
"epoch": 0.5725264169068204,
"grad_norm": 0.6143919229507446,
"learning_rate": 0.00012655365879769826,
"loss": 2.852082061767578,
"step": 2980
},
{
"epoch": 0.5734870317002881,
"grad_norm": 0.5853344202041626,
"learning_rate": 0.00012608305935039475,
"loss": 2.8520870208740234,
"step": 2985
},
{
"epoch": 0.574447646493756,
"grad_norm": 0.4826011061668396,
"learning_rate": 0.00012561270133134344,
"loss": 2.850848388671875,
"step": 2990
},
{
"epoch": 0.5754082612872238,
"grad_norm": 0.7613359689712524,
"learning_rate": 0.00012514258948854773,
"loss": 2.8502517700195313,
"step": 2995
},
{
"epoch": 0.5763688760806917,
"grad_norm": 0.751300036907196,
"learning_rate": 0.00012467272856752593,
"loss": 2.852705383300781,
"step": 3000
},
{
"epoch": 0.5773294908741594,
"grad_norm": 0.5232334733009338,
"learning_rate": 0.0001242031233112634,
"loss": 2.8527690887451174,
"step": 3005
},
{
"epoch": 0.5782901056676273,
"grad_norm": 0.5062333941459656,
"learning_rate": 0.00012373377846016493,
"loss": 2.848916435241699,
"step": 3010
},
{
"epoch": 0.579250720461095,
"grad_norm": 0.4421253800392151,
"learning_rate": 0.0001232646987520064,
"loss": 2.852520561218262,
"step": 3015
},
{
"epoch": 0.5802113352545629,
"grad_norm": 0.6208221912384033,
"learning_rate": 0.00012279588892188736,
"loss": 2.850700569152832,
"step": 3020
},
{
"epoch": 0.5811719500480308,
"grad_norm": 0.7289994359016418,
"learning_rate": 0.000122327353702183,
"loss": 2.850907325744629,
"step": 3025
},
{
"epoch": 0.5821325648414986,
"grad_norm": 0.5431981086730957,
"learning_rate": 0.0001218590978224966,
"loss": 2.8476612091064455,
"step": 3030
},
{
"epoch": 0.5830931796349664,
"grad_norm": 0.5733211040496826,
"learning_rate": 0.00012139112600961159,
"loss": 2.850396728515625,
"step": 3035
},
{
"epoch": 0.5840537944284342,
"grad_norm": 0.6280592679977417,
"learning_rate": 0.00012092344298744383,
"loss": 2.8513126373291016,
"step": 3040
},
{
"epoch": 0.5850144092219021,
"grad_norm": 0.7266437411308289,
"learning_rate": 0.00012045605347699411,
"loss": 2.853240966796875,
"step": 3045
},
{
"epoch": 0.5859750240153698,
"grad_norm": 0.5711589455604553,
"learning_rate": 0.00011998896219630029,
"loss": 2.8492050170898438,
"step": 3050
},
{
"epoch": 0.5869356388088377,
"grad_norm": 0.5127094984054565,
"learning_rate": 0.0001195221738603899,
"loss": 2.8453636169433594,
"step": 3055
},
{
"epoch": 0.5878962536023055,
"grad_norm": 0.5526590943336487,
"learning_rate": 0.00011905569318123223,
"loss": 2.8491466522216795,
"step": 3060
},
{
"epoch": 0.5888568683957733,
"grad_norm": 0.6706989407539368,
"learning_rate": 0.00011858952486769114,
"loss": 2.849506378173828,
"step": 3065
},
{
"epoch": 0.5898174831892411,
"grad_norm": 0.5911527276039124,
"learning_rate": 0.00011812367362547716,
"loss": 2.848992919921875,
"step": 3070
},
{
"epoch": 0.590778097982709,
"grad_norm": 0.5355327725410461,
"learning_rate": 0.0001176581441571002,
"loss": 2.851682662963867,
"step": 3075
},
{
"epoch": 0.5917387127761767,
"grad_norm": 0.5494570136070251,
"learning_rate": 0.00011719294116182217,
"loss": 2.847641944885254,
"step": 3080
},
{
"epoch": 0.5926993275696446,
"grad_norm": 0.681022047996521,
"learning_rate": 0.00011672806933560925,
"loss": 2.847896957397461,
"step": 3085
},
{
"epoch": 0.5936599423631124,
"grad_norm": 0.7260766625404358,
"learning_rate": 0.00011626353337108476,
"loss": 2.8473175048828123,
"step": 3090
},
{
"epoch": 0.5946205571565802,
"grad_norm": 0.6017476320266724,
"learning_rate": 0.00011579933795748164,
"loss": 2.8471202850341797,
"step": 3095
},
{
"epoch": 0.595581171950048,
"grad_norm": 0.47198686003685,
"learning_rate": 0.00011533548778059508,
"loss": 2.8486038208007813,
"step": 3100
},
{
"epoch": 0.5965417867435159,
"grad_norm": 0.4362870752811432,
"learning_rate": 0.00011487198752273552,
"loss": 2.848790168762207,
"step": 3105
},
{
"epoch": 0.5975024015369836,
"grad_norm": 0.6905494928359985,
"learning_rate": 0.0001144088418626809,
"loss": 2.849490928649902,
"step": 3110
},
{
"epoch": 0.5984630163304515,
"grad_norm": 0.6234802007675171,
"learning_rate": 0.00011394605547562989,
"loss": 2.848898696899414,
"step": 3115
},
{
"epoch": 0.5994236311239193,
"grad_norm": 0.4275026023387909,
"learning_rate": 0.00011348363303315434,
"loss": 2.846969413757324,
"step": 3120
},
{
"epoch": 0.6003842459173871,
"grad_norm": 0.4885198771953583,
"learning_rate": 0.00011302157920315244,
"loss": 2.8450719833374025,
"step": 3125
},
{
"epoch": 0.6013448607108549,
"grad_norm": 0.5203735828399658,
"learning_rate": 0.00011255989864980133,
"loss": 2.8464969635009765,
"step": 3130
},
{
"epoch": 0.6023054755043228,
"grad_norm": 0.6056355834007263,
"learning_rate": 0.00011209859603351015,
"loss": 2.8485231399536133,
"step": 3135
},
{
"epoch": 0.6032660902977905,
"grad_norm": 0.5373650789260864,
"learning_rate": 0.00011163767601087301,
"loss": 2.846028518676758,
"step": 3140
},
{
"epoch": 0.6042267050912584,
"grad_norm": 0.525007963180542,
"learning_rate": 0.00011117714323462186,
"loss": 2.846885871887207,
"step": 3145
},
{
"epoch": 0.6051873198847262,
"grad_norm": 0.4438735544681549,
"learning_rate": 0.00011071700235357979,
"loss": 2.845564079284668,
"step": 3150
},
{
"epoch": 0.6061479346781941,
"grad_norm": 0.5859609842300415,
"learning_rate": 0.00011025725801261373,
"loss": 2.844126510620117,
"step": 3155
},
{
"epoch": 0.6071085494716618,
"grad_norm": 0.595695436000824,
"learning_rate": 0.00010979791485258788,
"loss": 2.8454761505126953,
"step": 3160
},
{
"epoch": 0.6080691642651297,
"grad_norm": 0.523353099822998,
"learning_rate": 0.00010933897751031671,
"loss": 2.8453132629394533,
"step": 3165
},
{
"epoch": 0.6090297790585975,
"grad_norm": 0.44268059730529785,
"learning_rate": 0.00010888045061851807,
"loss": 2.8437095642089845,
"step": 3170
},
{
"epoch": 0.6099903938520653,
"grad_norm": 0.5390836596488953,
"learning_rate": 0.00010842233880576681,
"loss": 2.8442451477050783,
"step": 3175
},
{
"epoch": 0.6109510086455331,
"grad_norm": 0.6214951872825623,
"learning_rate": 0.0001079646466964475,
"loss": 2.842816162109375,
"step": 3180
},
{
"epoch": 0.611911623439001,
"grad_norm": 0.5127742886543274,
"learning_rate": 0.00010750737891070824,
"loss": 2.8442523956298826,
"step": 3185
},
{
"epoch": 0.6128722382324687,
"grad_norm": 0.5821230411529541,
"learning_rate": 0.00010705054006441371,
"loss": 2.8449642181396486,
"step": 3190
},
{
"epoch": 0.6138328530259366,
"grad_norm": 0.5417553782463074,
"learning_rate": 0.00010659413476909865,
"loss": 2.8446380615234377,
"step": 3195
},
{
"epoch": 0.6147934678194045,
"grad_norm": 0.5219445824623108,
"learning_rate": 0.00010613816763192152,
"loss": 2.843789291381836,
"step": 3200
},
{
"epoch": 0.6157540826128722,
"grad_norm": 0.5322816967964172,
"learning_rate": 0.00010568264325561763,
"loss": 2.8418731689453125,
"step": 3205
},
{
"epoch": 0.6167146974063401,
"grad_norm": 0.713876485824585,
"learning_rate": 0.000105227566238453,
"loss": 2.848464584350586,
"step": 3210
},
{
"epoch": 0.6176753121998079,
"grad_norm": 0.5759865641593933,
"learning_rate": 0.00010477294117417762,
"loss": 2.84826545715332,
"step": 3215
},
{
"epoch": 0.6186359269932757,
"grad_norm": 0.5045075416564941,
"learning_rate": 0.00010431877265197955,
"loss": 2.844080352783203,
"step": 3220
},
{
"epoch": 0.6195965417867435,
"grad_norm": 0.5640796422958374,
"learning_rate": 0.00010386506525643808,
"loss": 2.8412174224853515,
"step": 3225
},
{
"epoch": 0.6205571565802114,
"grad_norm": 0.5091181397438049,
"learning_rate": 0.00010341182356747771,
"loss": 2.8444406509399416,
"step": 3230
},
{
"epoch": 0.6215177713736791,
"grad_norm": 0.4901158809661865,
"learning_rate": 0.00010295905216032203,
"loss": 2.843181610107422,
"step": 3235
},
{
"epoch": 0.622478386167147,
"grad_norm": 0.48062700033187866,
"learning_rate": 0.00010250675560544717,
"loss": 2.84393424987793,
"step": 3240
},
{
"epoch": 0.6234390009606148,
"grad_norm": 0.8245052099227905,
"learning_rate": 0.00010205493846853618,
"loss": 2.8429317474365234,
"step": 3245
},
{
"epoch": 0.6243996157540826,
"grad_norm": 0.5051743984222412,
"learning_rate": 0.00010160360531043239,
"loss": 2.8431800842285155,
"step": 3250
},
{
"epoch": 0.6253602305475504,
"grad_norm": 0.45200181007385254,
"learning_rate": 0.00010115276068709377,
"loss": 2.8441776275634765,
"step": 3255
},
{
"epoch": 0.6263208453410183,
"grad_norm": 0.4492049217224121,
"learning_rate": 0.00010070240914954676,
"loss": 2.840077018737793,
"step": 3260
},
{
"epoch": 0.627281460134486,
"grad_norm": 0.5113126039505005,
"learning_rate": 0.00010025255524384033,
"loss": 2.8416378021240236,
"step": 3265
},
{
"epoch": 0.6282420749279539,
"grad_norm": 0.45182526111602783,
"learning_rate": 9.980320351100028e-05,
"loss": 2.842592239379883,
"step": 3270
},
{
"epoch": 0.6292026897214217,
"grad_norm": 0.675184965133667,
"learning_rate": 9.935435848698307e-05,
"loss": 2.840932846069336,
"step": 3275
},
{
"epoch": 0.6301633045148896,
"grad_norm": 0.4641052186489105,
"learning_rate": 9.890602470263037e-05,
"loss": 2.8466796875,
"step": 3280
},
{
"epoch": 0.6311239193083573,
"grad_norm": 0.5004357099533081,
"learning_rate": 9.845820668362308e-05,
"loss": 2.8424007415771486,
"step": 3285
},
{
"epoch": 0.6320845341018252,
"grad_norm": 0.5009520053863525,
"learning_rate": 9.801090895043566e-05,
"loss": 2.8406450271606447,
"step": 3290
},
{
"epoch": 0.633045148895293,
"grad_norm": 0.5677304863929749,
"learning_rate": 9.756413601829083e-05,
"loss": 2.8411323547363283,
"step": 3295
},
{
"epoch": 0.6340057636887608,
"grad_norm": 0.48396027088165283,
"learning_rate": 9.711789239711344e-05,
"loss": 2.842608642578125,
"step": 3300
},
{
"epoch": 0.6349663784822286,
"grad_norm": 0.5474041700363159,
"learning_rate": 9.667218259148547e-05,
"loss": 2.8427316665649416,
"step": 3305
},
{
"epoch": 0.6359269932756965,
"grad_norm": 0.619439423084259,
"learning_rate": 9.62270111006001e-05,
"loss": 2.8402175903320312,
"step": 3310
},
{
"epoch": 0.6368876080691642,
"grad_norm": 0.4645543098449707,
"learning_rate": 9.57823824182168e-05,
"loss": 2.8401821136474608,
"step": 3315
},
{
"epoch": 0.6378482228626321,
"grad_norm": 0.4728531837463379,
"learning_rate": 9.53383010326155e-05,
"loss": 2.840835762023926,
"step": 3320
},
{
"epoch": 0.6388088376560999,
"grad_norm": 0.4560253322124481,
"learning_rate": 9.489477142655147e-05,
"loss": 2.839691925048828,
"step": 3325
},
{
"epoch": 0.6397694524495677,
"grad_norm": 0.5300701856613159,
"learning_rate": 9.445179807721012e-05,
"loss": 2.8393108367919924,
"step": 3330
},
{
"epoch": 0.6407300672430355,
"grad_norm": 0.5008341670036316,
"learning_rate": 9.400938545616173e-05,
"loss": 2.841996765136719,
"step": 3335
},
{
"epoch": 0.6416906820365034,
"grad_norm": 0.5530736446380615,
"learning_rate": 9.356753802931646e-05,
"loss": 2.837991142272949,
"step": 3340
},
{
"epoch": 0.6426512968299711,
"grad_norm": 0.5178174376487732,
"learning_rate": 9.312626025687897e-05,
"loss": 2.8412078857421874,
"step": 3345
},
{
"epoch": 0.643611911623439,
"grad_norm": 0.4595809280872345,
"learning_rate": 9.268555659330364e-05,
"loss": 2.841719055175781,
"step": 3350
},
{
"epoch": 0.6445725264169068,
"grad_norm": 0.34829553961753845,
"learning_rate": 9.22454314872496e-05,
"loss": 2.8385791778564453,
"step": 3355
},
{
"epoch": 0.6455331412103746,
"grad_norm": 0.5107468366622925,
"learning_rate": 9.180588938153566e-05,
"loss": 2.837472343444824,
"step": 3360
},
{
"epoch": 0.6464937560038425,
"grad_norm": 0.4454675316810608,
"learning_rate": 9.136693471309568e-05,
"loss": 2.845228576660156,
"step": 3365
},
{
"epoch": 0.6474543707973103,
"grad_norm": 0.5800176858901978,
"learning_rate": 9.092857191293356e-05,
"loss": 2.843001937866211,
"step": 3370
},
{
"epoch": 0.6484149855907781,
"grad_norm": 0.4466376304626465,
"learning_rate": 9.049080540607875e-05,
"loss": 2.837605857849121,
"step": 3375
},
{
"epoch": 0.6493756003842459,
"grad_norm": 0.3897906541824341,
"learning_rate": 9.005363961154126e-05,
"loss": 2.8372968673706054,
"step": 3380
},
{
"epoch": 0.6503362151777138,
"grad_norm": 0.554328441619873,
"learning_rate": 8.961707894226735e-05,
"loss": 2.8404136657714845,
"step": 3385
},
{
"epoch": 0.6512968299711815,
"grad_norm": 0.5811151266098022,
"learning_rate": 8.918112780509494e-05,
"loss": 2.8356746673583983,
"step": 3390
},
{
"epoch": 0.6522574447646494,
"grad_norm": 0.5617589354515076,
"learning_rate": 8.874579060070894e-05,
"loss": 2.8357141494750975,
"step": 3395
},
{
"epoch": 0.6532180595581172,
"grad_norm": 0.4976288378238678,
"learning_rate": 8.831107172359707e-05,
"loss": 2.8398956298828124,
"step": 3400
},
{
"epoch": 0.654178674351585,
"grad_norm": 0.4745360314846039,
"learning_rate": 8.787697556200519e-05,
"loss": 2.8399864196777345,
"step": 3405
},
{
"epoch": 0.6551392891450528,
"grad_norm": 0.4591399133205414,
"learning_rate": 8.744350649789347e-05,
"loss": 2.8361705780029296,
"step": 3410
},
{
"epoch": 0.6560999039385207,
"grad_norm": 0.4175853729248047,
"learning_rate": 8.701066890689166e-05,
"loss": 2.835278129577637,
"step": 3415
},
{
"epoch": 0.6570605187319885,
"grad_norm": 0.4387359619140625,
"learning_rate": 8.657846715825508e-05,
"loss": 2.838814544677734,
"step": 3420
},
{
"epoch": 0.6580211335254563,
"grad_norm": 0.5446368455886841,
"learning_rate": 8.61469056148209e-05,
"loss": 2.8376935958862304,
"step": 3425
},
{
"epoch": 0.6589817483189241,
"grad_norm": 0.4309311509132385,
"learning_rate": 8.571598863296342e-05,
"loss": 2.8386699676513674,
"step": 3430
},
{
"epoch": 0.659942363112392,
"grad_norm": 0.4344281554222107,
"learning_rate": 8.528572056255065e-05,
"loss": 2.8364212036132814,
"step": 3435
},
{
"epoch": 0.6609029779058597,
"grad_norm": 0.4265115261077881,
"learning_rate": 8.485610574690021e-05,
"loss": 2.8368595123291014,
"step": 3440
},
{
"epoch": 0.6618635926993276,
"grad_norm": 0.530190646648407,
"learning_rate": 8.442714852273523e-05,
"loss": 2.836482048034668,
"step": 3445
},
{
"epoch": 0.6628242074927954,
"grad_norm": 0.4156145751476288,
"learning_rate": 8.399885322014123e-05,
"loss": 2.837312698364258,
"step": 3450
},
{
"epoch": 0.6637848222862632,
"grad_norm": 0.3237242102622986,
"learning_rate": 8.35712241625216e-05,
"loss": 2.837934112548828,
"step": 3455
},
{
"epoch": 0.664745437079731,
"grad_norm": 0.5120749473571777,
"learning_rate": 8.314426566655458e-05,
"loss": 2.8360355377197264,
"step": 3460
},
{
"epoch": 0.6657060518731989,
"grad_norm": 0.32652583718299866,
"learning_rate": 8.271798204214942e-05,
"loss": 2.83548526763916,
"step": 3465
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.4441000521183014,
"learning_rate": 8.229237759240289e-05,
"loss": 2.832575225830078,
"step": 3470
},
{
"epoch": 0.6676272814601345,
"grad_norm": 0.46819108724594116,
"learning_rate": 8.186745661355595e-05,
"loss": 2.832535171508789,
"step": 3475
},
{
"epoch": 0.6685878962536023,
"grad_norm": 0.42160722613334656,
"learning_rate": 8.144322339495012e-05,
"loss": 2.8376684188842773,
"step": 3480
},
{
"epoch": 0.6695485110470701,
"grad_norm": 0.4239354431629181,
"learning_rate": 8.101968221898453e-05,
"loss": 2.8348594665527345,
"step": 3485
},
{
"epoch": 0.6705091258405379,
"grad_norm": 0.465847373008728,
"learning_rate": 8.059683736107245e-05,
"loss": 2.8386322021484376,
"step": 3490
},
{
"epoch": 0.6714697406340058,
"grad_norm": 0.4507569372653961,
"learning_rate": 8.017469308959823e-05,
"loss": 2.836222839355469,
"step": 3495
},
{
"epoch": 0.6724303554274735,
"grad_norm": 0.35617345571517944,
"learning_rate": 7.97532536658742e-05,
"loss": 2.837454605102539,
"step": 3500
},
{
"epoch": 0.6733909702209414,
"grad_norm": 0.5097410082817078,
"learning_rate": 7.933252334409766e-05,
"loss": 2.836711883544922,
"step": 3505
},
{
"epoch": 0.6743515850144092,
"grad_norm": 0.37498244643211365,
"learning_rate": 7.891250637130779e-05,
"loss": 2.8351585388183596,
"step": 3510
},
{
"epoch": 0.675312199807877,
"grad_norm": 0.3829497992992401,
"learning_rate": 7.849320698734306e-05,
"loss": 2.839457321166992,
"step": 3515
},
{
"epoch": 0.6762728146013448,
"grad_norm": 0.40281057357788086,
"learning_rate": 7.80746294247982e-05,
"loss": 2.8323028564453123,
"step": 3520
},
{
"epoch": 0.6772334293948127,
"grad_norm": 0.5071494579315186,
"learning_rate": 7.765677790898155e-05,
"loss": 2.8315807342529298,
"step": 3525
},
{
"epoch": 0.6781940441882806,
"grad_norm": 0.4299395680427551,
"learning_rate": 7.723965665787255e-05,
"loss": 2.8363826751708983,
"step": 3530
},
{
"epoch": 0.6791546589817483,
"grad_norm": 0.4607393443584442,
"learning_rate": 7.682326988207877e-05,
"loss": 2.835020065307617,
"step": 3535
},
{
"epoch": 0.6801152737752162,
"grad_norm": 0.5409980416297913,
"learning_rate": 7.640762178479382e-05,
"loss": 2.8327972412109377,
"step": 3540
},
{
"epoch": 0.681075888568684,
"grad_norm": 0.4520939290523529,
"learning_rate": 7.599271656175476e-05,
"loss": 2.837631607055664,
"step": 3545
},
{
"epoch": 0.6820365033621518,
"grad_norm": 0.4245823323726654,
"learning_rate": 7.557855840119976e-05,
"loss": 2.833856201171875,
"step": 3550
},
{
"epoch": 0.6829971181556196,
"grad_norm": 0.3175192177295685,
"learning_rate": 7.516515148382576e-05,
"loss": 2.833210754394531,
"step": 3555
},
{
"epoch": 0.6839577329490875,
"grad_norm": 0.34715819358825684,
"learning_rate": 7.475249998274621e-05,
"loss": 2.8319278717041017,
"step": 3560
},
{
"epoch": 0.6849183477425552,
"grad_norm": 0.6079280376434326,
"learning_rate": 7.434060806344936e-05,
"loss": 2.836623764038086,
"step": 3565
},
{
"epoch": 0.6858789625360231,
"grad_norm": 0.4253065884113312,
"learning_rate": 7.392947988375555e-05,
"loss": 2.8338897705078123,
"step": 3570
},
{
"epoch": 0.6868395773294909,
"grad_norm": 0.42361024022102356,
"learning_rate": 7.351911959377585e-05,
"loss": 2.8336849212646484,
"step": 3575
},
{
"epoch": 0.6878001921229587,
"grad_norm": 0.3799959719181061,
"learning_rate": 7.310953133586981e-05,
"loss": 2.831894874572754,
"step": 3580
},
{
"epoch": 0.6887608069164265,
"grad_norm": 0.3994083106517792,
"learning_rate": 7.27007192446036e-05,
"loss": 2.833780860900879,
"step": 3585
},
{
"epoch": 0.6897214217098944,
"grad_norm": 0.5485339760780334,
"learning_rate": 7.229268744670883e-05,
"loss": 2.831420135498047,
"step": 3590
},
{
"epoch": 0.6906820365033621,
"grad_norm": 0.3720037341117859,
"learning_rate": 7.188544006104e-05,
"loss": 2.8320638656616213,
"step": 3595
},
{
"epoch": 0.69164265129683,
"grad_norm": 0.4130679666996002,
"learning_rate": 7.147898119853367e-05,
"loss": 2.832315444946289,
"step": 3600
},
{
"epoch": 0.6926032660902978,
"grad_norm": 0.38745787739753723,
"learning_rate": 7.107331496216676e-05,
"loss": 2.8339397430419924,
"step": 3605
},
{
"epoch": 0.6935638808837656,
"grad_norm": 0.3671323359012604,
"learning_rate": 7.066844544691474e-05,
"loss": 2.835890007019043,
"step": 3610
},
{
"epoch": 0.6945244956772334,
"grad_norm": 0.644256055355072,
"learning_rate": 7.026437673971107e-05,
"loss": 2.831973075866699,
"step": 3615
},
{
"epoch": 0.6954851104707013,
"grad_norm": 0.38100337982177734,
"learning_rate": 6.986111291940511e-05,
"loss": 2.8321693420410154,
"step": 3620
},
{
"epoch": 0.696445725264169,
"grad_norm": 0.3376613259315491,
"learning_rate": 6.945865805672154e-05,
"loss": 2.8267669677734375,
"step": 3625
},
{
"epoch": 0.6974063400576369,
"grad_norm": 0.3127301335334778,
"learning_rate": 6.905701621421904e-05,
"loss": 2.8354618072509767,
"step": 3630
},
{
"epoch": 0.6983669548511047,
"grad_norm": 0.3760312497615814,
"learning_rate": 6.865619144624914e-05,
"loss": 2.8312320709228516,
"step": 3635
},
{
"epoch": 0.6993275696445725,
"grad_norm": 0.44218409061431885,
"learning_rate": 6.825618779891577e-05,
"loss": 2.8325376510620117,
"step": 3640
},
{
"epoch": 0.7002881844380403,
"grad_norm": 0.47172650694847107,
"learning_rate": 6.785700931003381e-05,
"loss": 2.8269069671630858,
"step": 3645
},
{
"epoch": 0.7012487992315082,
"grad_norm": 0.3911219537258148,
"learning_rate": 6.745866000908874e-05,
"loss": 2.8321979522705076,
"step": 3650
},
{
"epoch": 0.7022094140249759,
"grad_norm": 0.3831585943698883,
"learning_rate": 6.706114391719586e-05,
"loss": 2.8329093933105467,
"step": 3655
},
{
"epoch": 0.7031700288184438,
"grad_norm": 0.41433626413345337,
"learning_rate": 6.66644650470597e-05,
"loss": 2.833950424194336,
"step": 3660
},
{
"epoch": 0.7041306436119116,
"grad_norm": 0.34397903084754944,
"learning_rate": 6.626862740293338e-05,
"loss": 2.830823516845703,
"step": 3665
},
{
"epoch": 0.7050912584053795,
"grad_norm": 0.39430761337280273,
"learning_rate": 6.587363498057845e-05,
"loss": 2.8328922271728514,
"step": 3670
},
{
"epoch": 0.7060518731988472,
"grad_norm": 0.3444637656211853,
"learning_rate": 6.547949176722437e-05,
"loss": 2.8278175354003907,
"step": 3675
},
{
"epoch": 0.7070124879923151,
"grad_norm": 0.4200674295425415,
"learning_rate": 6.508620174152826e-05,
"loss": 2.8334108352661134,
"step": 3680
},
{
"epoch": 0.7079731027857828,
"grad_norm": 0.5146929621696472,
"learning_rate": 6.469376887353491e-05,
"loss": 2.8332847595214843,
"step": 3685
},
{
"epoch": 0.7089337175792507,
"grad_norm": 0.41079193353652954,
"learning_rate": 6.430219712463634e-05,
"loss": 2.8310791015625,
"step": 3690
},
{
"epoch": 0.7098943323727186,
"grad_norm": 0.3751753866672516,
"learning_rate": 6.391149044753238e-05,
"loss": 2.831695556640625,
"step": 3695
},
{
"epoch": 0.7108549471661864,
"grad_norm": 0.3280157744884491,
"learning_rate": 6.352165278619012e-05,
"loss": 2.8301315307617188,
"step": 3700
},
{
"epoch": 0.7118155619596542,
"grad_norm": 0.2948898673057556,
"learning_rate": 6.313268807580462e-05,
"loss": 2.833317756652832,
"step": 3705
},
{
"epoch": 0.712776176753122,
"grad_norm": 0.4521695077419281,
"learning_rate": 6.274460024275895e-05,
"loss": 2.831614685058594,
"step": 3710
},
{
"epoch": 0.7137367915465899,
"grad_norm": 0.4257628619670868,
"learning_rate": 6.235739320458442e-05,
"loss": 2.828181838989258,
"step": 3715
},
{
"epoch": 0.7146974063400576,
"grad_norm": 0.3514285087585449,
"learning_rate": 6.197107086992156e-05,
"loss": 2.830687713623047,
"step": 3720
},
{
"epoch": 0.7156580211335255,
"grad_norm": 0.3143792748451233,
"learning_rate": 6.158563713847994e-05,
"loss": 2.8303714752197267,
"step": 3725
},
{
"epoch": 0.7166186359269933,
"grad_norm": 0.29875609278678894,
"learning_rate": 6.12010959009994e-05,
"loss": 2.832419013977051,
"step": 3730
},
{
"epoch": 0.7175792507204611,
"grad_norm": 0.2965286374092102,
"learning_rate": 6.081745103921047e-05,
"loss": 2.8294748306274413,
"step": 3735
},
{
"epoch": 0.7185398655139289,
"grad_norm": 0.3416411578655243,
"learning_rate": 6.043470642579516e-05,
"loss": 2.8330385208129885,
"step": 3740
},
{
"epoch": 0.7195004803073968,
"grad_norm": 0.4225616455078125,
"learning_rate": 6.005286592434828e-05,
"loss": 2.8280052185058593,
"step": 3745
},
{
"epoch": 0.7204610951008645,
"grad_norm": 0.42786410450935364,
"learning_rate": 5.967193338933778e-05,
"loss": 2.8302026748657227,
"step": 3750
},
{
"epoch": 0.7214217098943324,
"grad_norm": 0.34491264820098877,
"learning_rate": 5.9291912666066405e-05,
"loss": 2.8301876068115233,
"step": 3755
},
{
"epoch": 0.7223823246878002,
"grad_norm": 0.32240840792655945,
"learning_rate": 5.891280759063265e-05,
"loss": 2.8271007537841797,
"step": 3760
},
{
"epoch": 0.723342939481268,
"grad_norm": 0.35663703083992004,
"learning_rate": 5.853462198989184e-05,
"loss": 2.8305374145507813,
"step": 3765
},
{
"epoch": 0.7243035542747358,
"grad_norm": 0.3207663595676422,
"learning_rate": 5.815735968141813e-05,
"loss": 2.83172607421875,
"step": 3770
},
{
"epoch": 0.7252641690682037,
"grad_norm": 0.38391485810279846,
"learning_rate": 5.778102447346514e-05,
"loss": 2.8313671112060548,
"step": 3775
},
{
"epoch": 0.7262247838616714,
"grad_norm": 0.4688833951950073,
"learning_rate": 5.740562016492811e-05,
"loss": 2.8275665283203124,
"step": 3780
},
{
"epoch": 0.7271853986551393,
"grad_norm": 0.3272785246372223,
"learning_rate": 5.703115054530537e-05,
"loss": 2.8282421112060545,
"step": 3785
},
{
"epoch": 0.7281460134486071,
"grad_norm": 0.3496851623058319,
"learning_rate": 5.665761939466008e-05,
"loss": 2.829297637939453,
"step": 3790
},
{
"epoch": 0.729106628242075,
"grad_norm": 0.3146269917488098,
"learning_rate": 5.628503048358207e-05,
"loss": 2.8277191162109374,
"step": 3795
},
{
"epoch": 0.7300672430355427,
"grad_norm": 0.34081220626831055,
"learning_rate": 5.591338757314968e-05,
"loss": 2.831098937988281,
"step": 3800
},
{
"epoch": 0.7310278578290106,
"grad_norm": 0.2704191505908966,
"learning_rate": 5.554269441489204e-05,
"loss": 2.8295705795288084,
"step": 3805
},
{
"epoch": 0.7319884726224783,
"grad_norm": 0.29791897535324097,
"learning_rate": 5.517295475075102e-05,
"loss": 2.8285537719726563,
"step": 3810
},
{
"epoch": 0.7329490874159462,
"grad_norm": 0.3568146228790283,
"learning_rate": 5.4804172313043465e-05,
"loss": 2.8266357421875,
"step": 3815
},
{
"epoch": 0.733909702209414,
"grad_norm": 0.3287736475467682,
"learning_rate": 5.443635082442363e-05,
"loss": 2.8287899017333986,
"step": 3820
},
{
"epoch": 0.7348703170028819,
"grad_norm": 0.30268165469169617,
"learning_rate": 5.4069493997845356e-05,
"loss": 2.8289199829101563,
"step": 3825
},
{
"epoch": 0.7358309317963496,
"grad_norm": 0.28268566727638245,
"learning_rate": 5.3703605536524905e-05,
"loss": 2.8304866790771483,
"step": 3830
},
{
"epoch": 0.7367915465898175,
"grad_norm": 0.35326439142227173,
"learning_rate": 5.333868913390338e-05,
"loss": 2.8253864288330077,
"step": 3835
},
{
"epoch": 0.7377521613832853,
"grad_norm": 0.2834617793560028,
"learning_rate": 5.2974748473609505e-05,
"loss": 2.8260841369628906,
"step": 3840
},
{
"epoch": 0.7387127761767531,
"grad_norm": 0.24708347022533417,
"learning_rate": 5.261178722942242e-05,
"loss": 2.829148864746094,
"step": 3845
},
{
"epoch": 0.7396733909702209,
"grad_norm": 0.34735795855522156,
"learning_rate": 5.224980906523462e-05,
"loss": 2.827422332763672,
"step": 3850
},
{
"epoch": 0.7406340057636888,
"grad_norm": 0.4254157841205597,
"learning_rate": 5.188881763501486e-05,
"loss": 2.8264198303222656,
"step": 3855
},
{
"epoch": 0.7415946205571565,
"grad_norm": 0.43587496876716614,
"learning_rate": 5.152881658277147e-05,
"loss": 2.828824996948242,
"step": 3860
},
{
"epoch": 0.7425552353506244,
"grad_norm": 0.34229278564453125,
"learning_rate": 5.1169809542515404e-05,
"loss": 2.83102970123291,
"step": 3865
},
{
"epoch": 0.7435158501440923,
"grad_norm": 0.2842935025691986,
"learning_rate": 5.081180013822368e-05,
"loss": 2.826640510559082,
"step": 3870
},
{
"epoch": 0.74447646493756,
"grad_norm": 0.29888179898262024,
"learning_rate": 5.045479198380272e-05,
"loss": 2.8275705337524415,
"step": 3875
},
{
"epoch": 0.7454370797310279,
"grad_norm": 0.31076836585998535,
"learning_rate": 5.009878868305171e-05,
"loss": 2.8290031433105467,
"step": 3880
},
{
"epoch": 0.7463976945244957,
"grad_norm": 0.36958396434783936,
"learning_rate": 4.9743793829626736e-05,
"loss": 2.8280960083007813,
"step": 3885
},
{
"epoch": 0.7473583093179635,
"grad_norm": 0.35068604350090027,
"learning_rate": 4.9389811007003834e-05,
"loss": 2.8277267456054687,
"step": 3890
},
{
"epoch": 0.7483189241114313,
"grad_norm": 0.2839006781578064,
"learning_rate": 4.903684378844333e-05,
"loss": 2.8289569854736327,
"step": 3895
},
{
"epoch": 0.7492795389048992,
"grad_norm": 0.313565194606781,
"learning_rate": 4.86848957369536e-05,
"loss": 2.8278987884521483,
"step": 3900
},
{
"epoch": 0.7502401536983669,
"grad_norm": 0.2784314453601837,
"learning_rate": 4.8333970405254904e-05,
"loss": 2.823335647583008,
"step": 3905
},
{
"epoch": 0.7512007684918348,
"grad_norm": 0.31032121181488037,
"learning_rate": 4.798407133574405e-05,
"loss": 2.8286136627197265,
"step": 3910
},
{
"epoch": 0.7521613832853026,
"grad_norm": 0.2912919819355011,
"learning_rate": 4.7635202060457945e-05,
"loss": 2.8263755798339845,
"step": 3915
},
{
"epoch": 0.7531219980787704,
"grad_norm": 0.3187970519065857,
"learning_rate": 4.72873661010385e-05,
"loss": 2.8279151916503906,
"step": 3920
},
{
"epoch": 0.7540826128722382,
"grad_norm": 0.34837865829467773,
"learning_rate": 4.694056696869688e-05,
"loss": 2.8299861907958985,
"step": 3925
},
{
"epoch": 0.7550432276657061,
"grad_norm": 0.28154686093330383,
"learning_rate": 4.659480816417785e-05,
"loss": 2.8240270614624023,
"step": 3930
},
{
"epoch": 0.7560038424591738,
"grad_norm": 0.3153039515018463,
"learning_rate": 4.6250093177725e-05,
"loss": 2.8291126251220704,
"step": 3935
},
{
"epoch": 0.7569644572526417,
"grad_norm": 0.2972996234893799,
"learning_rate": 4.590642548904479e-05,
"loss": 2.8271560668945312,
"step": 3940
},
{
"epoch": 0.7579250720461095,
"grad_norm": 0.30244746804237366,
"learning_rate": 4.5563808567272e-05,
"loss": 2.825422668457031,
"step": 3945
},
{
"epoch": 0.7588856868395774,
"grad_norm": 0.32528969645500183,
"learning_rate": 4.52222458709345e-05,
"loss": 2.8255084991455077,
"step": 3950
},
{
"epoch": 0.7598463016330451,
"grad_norm": 0.31731516122817993,
"learning_rate": 4.4881740847918155e-05,
"loss": 2.82420654296875,
"step": 3955
},
{
"epoch": 0.760806916426513,
"grad_norm": 0.32816174626350403,
"learning_rate": 4.454229693543251e-05,
"loss": 2.829576873779297,
"step": 3960
},
{
"epoch": 0.7617675312199808,
"grad_norm": 0.2828548848628998,
"learning_rate": 4.420391755997548e-05,
"loss": 2.827322769165039,
"step": 3965
},
{
"epoch": 0.7627281460134486,
"grad_norm": 0.2661912143230438,
"learning_rate": 4.386660613729925e-05,
"loss": 2.8287160873413084,
"step": 3970
},
{
"epoch": 0.7636887608069164,
"grad_norm": 0.2875749468803406,
"learning_rate": 4.35303660723756e-05,
"loss": 2.822760009765625,
"step": 3975
},
{
"epoch": 0.7646493756003843,
"grad_norm": 0.2423306703567505,
"learning_rate": 4.3195200759361455e-05,
"loss": 2.826748275756836,
"step": 3980
},
{
"epoch": 0.765609990393852,
"grad_norm": 0.2602522671222687,
"learning_rate": 4.2861113581564884e-05,
"loss": 2.82442569732666,
"step": 3985
},
{
"epoch": 0.7665706051873199,
"grad_norm": 0.27814000844955444,
"learning_rate": 4.252810791141054e-05,
"loss": 2.8286266326904297,
"step": 3990
},
{
"epoch": 0.7675312199807877,
"grad_norm": 0.2583538889884949,
"learning_rate": 4.2196187110406054e-05,
"loss": 2.8250932693481445,
"step": 3995
},
{
"epoch": 0.7684918347742555,
"grad_norm": 0.29075777530670166,
"learning_rate": 4.186535452910784e-05,
"loss": 2.8221099853515623,
"step": 4000
},
{
"epoch": 0.7694524495677233,
"grad_norm": 0.26990872621536255,
"learning_rate": 4.153561350708732e-05,
"loss": 2.8275081634521486,
"step": 4005
},
{
"epoch": 0.7704130643611912,
"grad_norm": 0.28701838850975037,
"learning_rate": 4.12069673728973e-05,
"loss": 2.8243141174316406,
"step": 4010
},
{
"epoch": 0.7713736791546589,
"grad_norm": 0.25467079877853394,
"learning_rate": 4.087941944403815e-05,
"loss": 2.820818328857422,
"step": 4015
},
{
"epoch": 0.7723342939481268,
"grad_norm": 0.26495566964149475,
"learning_rate": 4.0552973026924625e-05,
"loss": 2.824423980712891,
"step": 4020
},
{
"epoch": 0.7732949087415946,
"grad_norm": 0.29741546511650085,
"learning_rate": 4.022763141685226e-05,
"loss": 2.824292755126953,
"step": 4025
},
{
"epoch": 0.7742555235350624,
"grad_norm": 0.2792549729347229,
"learning_rate": 3.990339789796418e-05,
"loss": 2.8274419784545897,
"step": 4030
},
{
"epoch": 0.7752161383285303,
"grad_norm": 0.3055638372898102,
"learning_rate": 3.958027574321794e-05,
"loss": 2.825086975097656,
"step": 4035
},
{
"epoch": 0.7761767531219981,
"grad_norm": 0.24175697565078735,
"learning_rate": 3.9258268214352566e-05,
"loss": 2.82196159362793,
"step": 4040
},
{
"epoch": 0.777137367915466,
"grad_norm": 0.3098554313182831,
"learning_rate": 3.893737856185538e-05,
"loss": 2.827454948425293,
"step": 4045
},
{
"epoch": 0.7780979827089337,
"grad_norm": 0.2785423696041107,
"learning_rate": 3.861761002492952e-05,
"loss": 2.8235404968261717,
"step": 4050
},
{
"epoch": 0.7790585975024016,
"grad_norm": 0.3094455599784851,
"learning_rate": 3.8298965831461024e-05,
"loss": 2.8234750747680666,
"step": 4055
},
{
"epoch": 0.7800192122958693,
"grad_norm": 0.2724841237068176,
"learning_rate": 3.798144919798631e-05,
"loss": 2.8219326019287108,
"step": 4060
},
{
"epoch": 0.7809798270893372,
"grad_norm": 0.2618357837200165,
"learning_rate": 3.766506332965976e-05,
"loss": 2.822661209106445,
"step": 4065
},
{
"epoch": 0.781940441882805,
"grad_norm": 0.23582673072814941,
"learning_rate": 3.734981142022117e-05,
"loss": 2.8221805572509764,
"step": 4070
},
{
"epoch": 0.7829010566762729,
"grad_norm": 0.23418720066547394,
"learning_rate": 3.70356966519638e-05,
"loss": 2.826119804382324,
"step": 4075
},
{
"epoch": 0.7838616714697406,
"grad_norm": 0.24657297134399414,
"learning_rate": 3.672272219570199e-05,
"loss": 2.8231496810913086,
"step": 4080
},
{
"epoch": 0.7848222862632085,
"grad_norm": 0.2549417018890381,
"learning_rate": 3.641089121073934e-05,
"loss": 2.8232580184936524,
"step": 4085
},
{
"epoch": 0.7857829010566763,
"grad_norm": 0.2880878746509552,
"learning_rate": 3.610020684483674e-05,
"loss": 2.823045349121094,
"step": 4090
},
{
"epoch": 0.7867435158501441,
"grad_norm": 0.27378255128860474,
"learning_rate": 3.579067223418046e-05,
"loss": 2.8220659255981446,
"step": 4095
},
{
"epoch": 0.7877041306436119,
"grad_norm": 0.26168304681777954,
"learning_rate": 3.548229050335089e-05,
"loss": 2.82537727355957,
"step": 4100
},
{
"epoch": 0.7886647454370798,
"grad_norm": 0.2955664098262787,
"learning_rate": 3.517506476529045e-05,
"loss": 2.8250572204589846,
"step": 4105
},
{
"epoch": 0.7896253602305475,
"grad_norm": 0.3224260210990906,
"learning_rate": 3.486899812127264e-05,
"loss": 2.8248443603515625,
"step": 4110
},
{
"epoch": 0.7905859750240154,
"grad_norm": 0.24833816289901733,
"learning_rate": 3.456409366087054e-05,
"loss": 2.8247648239135743,
"step": 4115
},
{
"epoch": 0.7915465898174832,
"grad_norm": 0.23027928173542023,
"learning_rate": 3.426035446192546e-05,
"loss": 2.822571563720703,
"step": 4120
},
{
"epoch": 0.792507204610951,
"grad_norm": 0.24701501429080963,
"learning_rate": 3.395778359051634e-05,
"loss": 2.8267234802246093,
"step": 4125
},
{
"epoch": 0.7934678194044188,
"grad_norm": 0.22311267256736755,
"learning_rate": 3.365638410092819e-05,
"loss": 2.8257022857666017,
"step": 4130
},
{
"epoch": 0.7944284341978867,
"grad_norm": 0.23521657288074493,
"learning_rate": 3.3356159035621746e-05,
"loss": 2.8220043182373047,
"step": 4135
},
{
"epoch": 0.7953890489913544,
"grad_norm": 0.2287784218788147,
"learning_rate": 3.3057111425202614e-05,
"loss": 2.8221038818359374,
"step": 4140
},
{
"epoch": 0.7963496637848223,
"grad_norm": 0.20921269059181213,
"learning_rate": 3.275924428839043e-05,
"loss": 2.8214012145996095,
"step": 4145
},
{
"epoch": 0.7973102785782901,
"grad_norm": 0.19233739376068115,
"learning_rate": 3.246256063198895e-05,
"loss": 2.8214378356933594,
"step": 4150
},
{
"epoch": 0.7982708933717579,
"grad_norm": 0.2111852765083313,
"learning_rate": 3.216706345085499e-05,
"loss": 2.826046943664551,
"step": 4155
},
{
"epoch": 0.7992315081652257,
"grad_norm": 0.20778204500675201,
"learning_rate": 3.187275572786878e-05,
"loss": 2.8189327239990236,
"step": 4160
},
{
"epoch": 0.8001921229586936,
"grad_norm": 0.23499462008476257,
"learning_rate": 3.15796404339036e-05,
"loss": 2.8209259033203127,
"step": 4165
},
{
"epoch": 0.8011527377521613,
"grad_norm": 0.23061686754226685,
"learning_rate": 3.128772052779569e-05,
"loss": 2.8280948638916015,
"step": 4170
},
{
"epoch": 0.8021133525456292,
"grad_norm": 0.2535129487514496,
"learning_rate": 3.099699895631474e-05,
"loss": 2.820085906982422,
"step": 4175
},
{
"epoch": 0.803073967339097,
"grad_norm": 0.191250279545784,
"learning_rate": 3.0707478654133706e-05,
"loss": 2.8241729736328125,
"step": 4180
},
{
"epoch": 0.8040345821325648,
"grad_norm": 0.22104866802692413,
"learning_rate": 3.041916254379949e-05,
"loss": 2.824297332763672,
"step": 4185
},
{
"epoch": 0.8049951969260326,
"grad_norm": 0.19869548082351685,
"learning_rate": 3.0132053535703342e-05,
"loss": 2.8250572204589846,
"step": 4190
},
{
"epoch": 0.8059558117195005,
"grad_norm": 0.2149796336889267,
"learning_rate": 2.984615452805147e-05,
"loss": 2.8241449356079102,
"step": 4195
},
{
"epoch": 0.8069164265129684,
"grad_norm": 0.20605628192424774,
"learning_rate": 2.9561468406835865e-05,
"loss": 2.82540283203125,
"step": 4200
},
{
"epoch": 0.8078770413064361,
"grad_norm": 0.22707077860832214,
"learning_rate": 2.927799804580495e-05,
"loss": 2.823532485961914,
"step": 4205
},
{
"epoch": 0.808837656099904,
"grad_norm": 0.2316194772720337,
"learning_rate": 2.8995746306434853e-05,
"loss": 2.819510078430176,
"step": 4210
},
{
"epoch": 0.8097982708933718,
"grad_norm": 0.23179353773593903,
"learning_rate": 2.871471603790035e-05,
"loss": 2.8233768463134767,
"step": 4215
},
{
"epoch": 0.8107588856868396,
"grad_norm": 0.21509668231010437,
"learning_rate": 2.8434910077046163e-05,
"loss": 2.8241294860839843,
"step": 4220
},
{
"epoch": 0.8117195004803074,
"grad_norm": 0.20974694192409515,
"learning_rate": 2.8156331248358295e-05,
"loss": 2.8189987182617187,
"step": 4225
},
{
"epoch": 0.8126801152737753,
"grad_norm": 0.18374113738536835,
"learning_rate": 2.787898236393556e-05,
"loss": 2.8204113006591798,
"step": 4230
},
{
"epoch": 0.813640730067243,
"grad_norm": 0.2046244740486145,
"learning_rate": 2.7602866223461044e-05,
"loss": 2.82018928527832,
"step": 4235
},
{
"epoch": 0.8146013448607109,
"grad_norm": 0.20929695665836334,
"learning_rate": 2.7327985614174143e-05,
"loss": 2.8256500244140623,
"step": 4240
},
{
"epoch": 0.8155619596541787,
"grad_norm": 0.19047677516937256,
"learning_rate": 2.7054343310842115e-05,
"loss": 2.8188905715942383,
"step": 4245
},
{
"epoch": 0.8165225744476465,
"grad_norm": 0.19980789721012115,
"learning_rate": 2.6781942075732294e-05,
"loss": 2.819545364379883,
"step": 4250
},
{
"epoch": 0.8174831892411143,
"grad_norm": 0.22025620937347412,
"learning_rate": 2.65107846585841e-05,
"loss": 2.8238851547241213,
"step": 4255
},
{
"epoch": 0.8184438040345822,
"grad_norm": 0.1966722011566162,
"learning_rate": 2.624087379658123e-05,
"loss": 2.818410873413086,
"step": 4260
},
{
"epoch": 0.8194044188280499,
"grad_norm": 0.24230927228927612,
"learning_rate": 2.5972212214324162e-05,
"loss": 2.8239215850830077,
"step": 4265
},
{
"epoch": 0.8203650336215178,
"grad_norm": 0.208970844745636,
"learning_rate": 2.5704802623802595e-05,
"loss": 2.821892738342285,
"step": 4270
},
{
"epoch": 0.8213256484149856,
"grad_norm": 0.21934658288955688,
"learning_rate": 2.5438647724368054e-05,
"loss": 2.8246734619140623,
"step": 4275
},
{
"epoch": 0.8222862632084534,
"grad_norm": 0.23692256212234497,
"learning_rate": 2.5173750202706666e-05,
"loss": 2.8202524185180664,
"step": 4280
},
{
"epoch": 0.8232468780019212,
"grad_norm": 0.1973477005958557,
"learning_rate": 2.491011273281189e-05,
"loss": 2.820056915283203,
"step": 4285
},
{
"epoch": 0.8242074927953891,
"grad_norm": 0.18771995604038239,
"learning_rate": 2.4647737975957954e-05,
"loss": 2.824921989440918,
"step": 4290
},
{
"epoch": 0.8251681075888568,
"grad_norm": 0.19082146883010864,
"learning_rate": 2.4386628580672396e-05,
"loss": 2.8206953048706054,
"step": 4295
},
{
"epoch": 0.8261287223823247,
"grad_norm": 0.17984676361083984,
"learning_rate": 2.4126787182709796e-05,
"loss": 2.8229496002197267,
"step": 4300
},
{
"epoch": 0.8270893371757925,
"grad_norm": 0.21605776250362396,
"learning_rate": 2.3868216405025002e-05,
"loss": 2.822156524658203,
"step": 4305
},
{
"epoch": 0.8280499519692603,
"grad_norm": 0.19645941257476807,
"learning_rate": 2.361091885774652e-05,
"loss": 2.821907615661621,
"step": 4310
},
{
"epoch": 0.8290105667627281,
"grad_norm": 0.215690478682518,
"learning_rate": 2.3354897138150536e-05,
"loss": 2.823683166503906,
"step": 4315
},
{
"epoch": 0.829971181556196,
"grad_norm": 0.19755929708480835,
"learning_rate": 2.3100153830634218e-05,
"loss": 2.823313903808594,
"step": 4320
},
{
"epoch": 0.8309317963496637,
"grad_norm": 0.19786836206912994,
"learning_rate": 2.284669150669001e-05,
"loss": 2.820745849609375,
"step": 4325
},
{
"epoch": 0.8318924111431316,
"grad_norm": 0.2134370356798172,
"learning_rate": 2.259451272487957e-05,
"loss": 2.820496368408203,
"step": 4330
},
{
"epoch": 0.8328530259365994,
"grad_norm": 0.1759730875492096,
"learning_rate": 2.234362003080772e-05,
"loss": 2.820623779296875,
"step": 4335
},
{
"epoch": 0.8338136407300673,
"grad_norm": 0.18195176124572754,
"learning_rate": 2.2094015957097215e-05,
"loss": 2.818101692199707,
"step": 4340
},
{
"epoch": 0.834774255523535,
"grad_norm": 0.1784997135400772,
"learning_rate": 2.1845703023362647e-05,
"loss": 2.820517921447754,
"step": 4345
},
{
"epoch": 0.8357348703170029,
"grad_norm": 0.1846170425415039,
"learning_rate": 2.159868373618544e-05,
"loss": 2.8230045318603514,
"step": 4350
},
{
"epoch": 0.8366954851104706,
"grad_norm": 0.21063563227653503,
"learning_rate": 2.13529605890883e-05,
"loss": 2.8197351455688477,
"step": 4355
},
{
"epoch": 0.8376560999039385,
"grad_norm": 0.1774480789899826,
"learning_rate": 2.110853606251004e-05,
"loss": 2.8187665939331055,
"step": 4360
},
{
"epoch": 0.8386167146974063,
"grad_norm": 0.18442946672439575,
"learning_rate": 2.0865412623780858e-05,
"loss": 2.8202911376953126,
"step": 4365
},
{
"epoch": 0.8395773294908742,
"grad_norm": 0.16251033544540405,
"learning_rate": 2.0623592727096916e-05,
"loss": 2.81771240234375,
"step": 4370
},
{
"epoch": 0.840537944284342,
"grad_norm": 0.21092604100704193,
"learning_rate": 2.0383078813496e-05,
"loss": 2.8186887741088866,
"step": 4375
},
{
"epoch": 0.8414985590778098,
"grad_norm": 0.16519330441951752,
"learning_rate": 2.014387331083268e-05,
"loss": 2.819426345825195,
"step": 4380
},
{
"epoch": 0.8424591738712777,
"grad_norm": 0.17469525337219238,
"learning_rate": 1.990597863375389e-05,
"loss": 2.8173801422119142,
"step": 4385
},
{
"epoch": 0.8434197886647454,
"grad_norm": 0.17096151411533356,
"learning_rate": 1.966939718367444e-05,
"loss": 2.818032455444336,
"step": 4390
},
{
"epoch": 0.8443804034582133,
"grad_norm": 0.18871955573558807,
"learning_rate": 1.9434131348752842e-05,
"loss": 2.8212139129638674,
"step": 4395
},
{
"epoch": 0.8453410182516811,
"grad_norm": 0.1763574182987213,
"learning_rate": 1.920018350386725e-05,
"loss": 2.8204654693603515,
"step": 4400
},
{
"epoch": 0.8463016330451489,
"grad_norm": 0.17636708915233612,
"learning_rate": 1.8967556010591423e-05,
"loss": 2.8186906814575194,
"step": 4405
},
{
"epoch": 0.8472622478386167,
"grad_norm": 0.17625270783901215,
"learning_rate": 1.873625121717089e-05,
"loss": 2.822053337097168,
"step": 4410
},
{
"epoch": 0.8482228626320846,
"grad_norm": 0.15954363346099854,
"learning_rate": 1.850627145849926e-05,
"loss": 2.816944885253906,
"step": 4415
},
{
"epoch": 0.8491834774255523,
"grad_norm": 0.16209356486797333,
"learning_rate": 1.8277619056094684e-05,
"loss": 2.817719650268555,
"step": 4420
},
{
"epoch": 0.8501440922190202,
"grad_norm": 0.17170670628547668,
"learning_rate": 1.805029631807632e-05,
"loss": 2.818096160888672,
"step": 4425
},
{
"epoch": 0.851104707012488,
"grad_norm": 0.15733708441257477,
"learning_rate": 1.7824305539141165e-05,
"loss": 2.819830322265625,
"step": 4430
},
{
"epoch": 0.8520653218059558,
"grad_norm": 0.1653648316860199,
"learning_rate": 1.7599649000540828e-05,
"loss": 2.8197336196899414,
"step": 4435
},
{
"epoch": 0.8530259365994236,
"grad_norm": 0.1656099557876587,
"learning_rate": 1.7376328970058382e-05,
"loss": 2.816743278503418,
"step": 4440
},
{
"epoch": 0.8539865513928915,
"grad_norm": 0.15552137792110443,
"learning_rate": 1.715434770198582e-05,
"loss": 2.823103332519531,
"step": 4445
},
{
"epoch": 0.8549471661863592,
"grad_norm": 0.17172972857952118,
"learning_rate": 1.6933707437100852e-05,
"loss": 2.8175386428833007,
"step": 4450
},
{
"epoch": 0.8559077809798271,
"grad_norm": 0.162519633769989,
"learning_rate": 1.67144104026446e-05,
"loss": 2.816008949279785,
"step": 4455
},
{
"epoch": 0.8568683957732949,
"grad_norm": 0.1861235499382019,
"learning_rate": 1.6496458812299073e-05,
"loss": 2.816597557067871,
"step": 4460
},
{
"epoch": 0.8578290105667628,
"grad_norm": 0.16897127032279968,
"learning_rate": 1.6279854866164586e-05,
"loss": 2.8191097259521483,
"step": 4465
},
{
"epoch": 0.8587896253602305,
"grad_norm": 0.16436539590358734,
"learning_rate": 1.6064600750737995e-05,
"loss": 2.8192588806152346,
"step": 4470
},
{
"epoch": 0.8597502401536984,
"grad_norm": 0.17348939180374146,
"learning_rate": 1.5850698638890093e-05,
"loss": 2.817781448364258,
"step": 4475
},
{
"epoch": 0.8607108549471661,
"grad_norm": 0.17937251925468445,
"learning_rate": 1.563815068984418e-05,
"loss": 2.8207448959350585,
"step": 4480
},
{
"epoch": 0.861671469740634,
"grad_norm": 0.15728498995304108,
"learning_rate": 1.54269590491539e-05,
"loss": 2.8215309143066407,
"step": 4485
},
{
"epoch": 0.8626320845341018,
"grad_norm": 0.15450921654701233,
"learning_rate": 1.521712584868166e-05,
"loss": 2.8194122314453125,
"step": 4490
},
{
"epoch": 0.8635926993275697,
"grad_norm": 0.15388450026512146,
"learning_rate": 1.5008653206577376e-05,
"loss": 2.8159854888916014,
"step": 4495
},
{
"epoch": 0.8645533141210374,
"grad_norm": 0.1582845151424408,
"learning_rate": 1.4801543227256685e-05,
"loss": 2.8205434799194338,
"step": 4500
},
{
"epoch": 0.8655139289145053,
"grad_norm": 0.15384982526302338,
"learning_rate": 1.4595798001379965e-05,
"loss": 2.820438003540039,
"step": 4505
},
{
"epoch": 0.866474543707973,
"grad_norm": 0.13886255025863647,
"learning_rate": 1.4391419605831218e-05,
"loss": 2.819916534423828,
"step": 4510
},
{
"epoch": 0.8674351585014409,
"grad_norm": 0.15293574333190918,
"learning_rate": 1.4188410103696957e-05,
"loss": 2.8186580657958986,
"step": 4515
},
{
"epoch": 0.8683957732949087,
"grad_norm": 0.15670090913772583,
"learning_rate": 1.398677154424559e-05,
"loss": 2.8180870056152343,
"step": 4520
},
{
"epoch": 0.8693563880883766,
"grad_norm": 0.13321813941001892,
"learning_rate": 1.3786505962906475e-05,
"loss": 2.816482162475586,
"step": 4525
},
{
"epoch": 0.8703170028818443,
"grad_norm": 0.15356488525867462,
"learning_rate": 1.3587615381249622e-05,
"loss": 2.8203657150268553,
"step": 4530
},
{
"epoch": 0.8712776176753122,
"grad_norm": 0.16018545627593994,
"learning_rate": 1.3390101806965165e-05,
"loss": 2.821044158935547,
"step": 4535
},
{
"epoch": 0.8722382324687801,
"grad_norm": 0.154199481010437,
"learning_rate": 1.3193967233843083e-05,
"loss": 2.818889617919922,
"step": 4540
},
{
"epoch": 0.8731988472622478,
"grad_norm": 0.16034984588623047,
"learning_rate": 1.2999213641753164e-05,
"loss": 2.8158634185791014,
"step": 4545
},
{
"epoch": 0.8741594620557157,
"grad_norm": 0.15023402869701385,
"learning_rate": 1.280584299662486e-05,
"loss": 2.8189815521240233,
"step": 4550
},
{
"epoch": 0.8751200768491835,
"grad_norm": 0.15561360120773315,
"learning_rate": 1.2613857250427622e-05,
"loss": 2.8163185119628906,
"step": 4555
},
{
"epoch": 0.8760806916426513,
"grad_norm": 0.13934877514839172,
"learning_rate": 1.2423258341151088e-05,
"loss": 2.8184600830078126,
"step": 4560
},
{
"epoch": 0.8770413064361191,
"grad_norm": 0.16308888792991638,
"learning_rate": 1.2234048192785539e-05,
"loss": 2.8192535400390626,
"step": 4565
},
{
"epoch": 0.878001921229587,
"grad_norm": 0.13949905335903168,
"learning_rate": 1.2046228715302509e-05,
"loss": 2.8167484283447264,
"step": 4570
},
{
"epoch": 0.8789625360230547,
"grad_norm": 0.19002912938594818,
"learning_rate": 1.1859801804635471e-05,
"loss": 2.818648910522461,
"step": 4575
},
{
"epoch": 0.8799231508165226,
"grad_norm": 0.1495082676410675,
"learning_rate": 1.167476934266065e-05,
"loss": 2.81795654296875,
"step": 4580
},
{
"epoch": 0.8808837656099904,
"grad_norm": 0.14693421125411987,
"learning_rate": 1.1491133197178177e-05,
"loss": 2.8210113525390623,
"step": 4585
},
{
"epoch": 0.8818443804034583,
"grad_norm": 0.15263605117797852,
"learning_rate": 1.1308895221893088e-05,
"loss": 2.8196102142333985,
"step": 4590
},
{
"epoch": 0.882804995196926,
"grad_norm": 0.14515769481658936,
"learning_rate": 1.1128057256396684e-05,
"loss": 2.819786262512207,
"step": 4595
},
{
"epoch": 0.8837656099903939,
"grad_norm": 0.13253174722194672,
"learning_rate": 1.0948621126147978e-05,
"loss": 2.8158369064331055,
"step": 4600
},
{
"epoch": 0.8847262247838616,
"grad_norm": 0.1385597586631775,
"learning_rate": 1.0770588642455092e-05,
"loss": 2.8145606994628904,
"step": 4605
},
{
"epoch": 0.8856868395773295,
"grad_norm": 0.15321281552314758,
"learning_rate": 1.0593961602457346e-05,
"loss": 2.822213554382324,
"step": 4610
},
{
"epoch": 0.8866474543707973,
"grad_norm": 0.1366969645023346,
"learning_rate": 1.041874178910666e-05,
"loss": 2.8188133239746094,
"step": 4615
},
{
"epoch": 0.8876080691642652,
"grad_norm": 0.1428319811820984,
"learning_rate": 1.0244930971149918e-05,
"loss": 2.8199222564697264,
"step": 4620
},
{
"epoch": 0.8885686839577329,
"grad_norm": 0.13369108736515045,
"learning_rate": 1.0072530903110942e-05,
"loss": 2.8198825836181642,
"step": 4625
},
{
"epoch": 0.8895292987512008,
"grad_norm": 0.14023332297801971,
"learning_rate": 9.901543325272753e-06,
"loss": 2.8179443359375,
"step": 4630
},
{
"epoch": 0.8904899135446686,
"grad_norm": 0.1439591497182846,
"learning_rate": 9.731969963660224e-06,
"loss": 2.82034969329834,
"step": 4635
},
{
"epoch": 0.8914505283381364,
"grad_norm": 0.1384945958852768,
"learning_rate": 9.56381253002233e-06,
"loss": 2.8165306091308593,
"step": 4640
},
{
"epoch": 0.8924111431316042,
"grad_norm": 0.14330501854419708,
"learning_rate": 9.397072721815113e-06,
"loss": 2.819794464111328,
"step": 4645
},
{
"epoch": 0.8933717579250721,
"grad_norm": 0.14544479548931122,
"learning_rate": 9.231752222184496e-06,
"loss": 2.8153934478759766,
"step": 4650
},
{
"epoch": 0.8943323727185398,
"grad_norm": 0.13283999264240265,
"learning_rate": 9.067852699949197e-06,
"loss": 2.8222179412841797,
"step": 4655
},
{
"epoch": 0.8952929875120077,
"grad_norm": 0.1298431158065796,
"learning_rate": 8.905375809584053e-06,
"loss": 2.8179580688476564,
"step": 4660
},
{
"epoch": 0.8962536023054755,
"grad_norm": 0.13331957161426544,
"learning_rate": 8.744323191203129e-06,
"loss": 2.8178497314453126,
"step": 4665
},
{
"epoch": 0.8972142170989433,
"grad_norm": 0.1336248815059662,
"learning_rate": 8.5846964705433e-06,
"loss": 2.817870330810547,
"step": 4670
},
{
"epoch": 0.8981748318924111,
"grad_norm": 0.13920700550079346,
"learning_rate": 8.426497258947813e-06,
"loss": 2.8151180267333986,
"step": 4675
},
{
"epoch": 0.899135446685879,
"grad_norm": 0.12426065653562546,
"learning_rate": 8.269727153349915e-06,
"loss": 2.8183738708496096,
"step": 4680
},
{
"epoch": 0.9000960614793467,
"grad_norm": 0.13054883480072021,
"learning_rate": 8.114387736256966e-06,
"loss": 2.818001556396484,
"step": 4685
},
{
"epoch": 0.9010566762728146,
"grad_norm": 0.13032682240009308,
"learning_rate": 7.960480575734162e-06,
"loss": 2.8181526184082033,
"step": 4690
},
{
"epoch": 0.9020172910662824,
"grad_norm": 0.14295165240764618,
"learning_rate": 7.808007225388963e-06,
"loss": 2.8167598724365233,
"step": 4695
},
{
"epoch": 0.9029779058597502,
"grad_norm": 0.13148348033428192,
"learning_rate": 7.656969224355285e-06,
"loss": 2.817153549194336,
"step": 4700
},
{
"epoch": 0.9039385206532181,
"grad_norm": 0.13584092259407043,
"learning_rate": 7.507368097277994e-06,
"loss": 2.813400077819824,
"step": 4705
},
{
"epoch": 0.9048991354466859,
"grad_norm": 0.13805082440376282,
"learning_rate": 7.359205354297499e-06,
"loss": 2.8138927459716796,
"step": 4710
},
{
"epoch": 0.9058597502401537,
"grad_norm": 0.1245737373828888,
"learning_rate": 7.21248249103451e-06,
"loss": 2.8152067184448244,
"step": 4715
},
{
"epoch": 0.9068203650336215,
"grad_norm": 0.12117403000593185,
"learning_rate": 7.067200988574983e-06,
"loss": 2.8162109375,
"step": 4720
},
{
"epoch": 0.9077809798270894,
"grad_norm": 0.12167931348085403,
"learning_rate": 6.923362313455094e-06,
"loss": 2.811721611022949,
"step": 4725
},
{
"epoch": 0.9087415946205571,
"grad_norm": 0.12101319432258606,
"learning_rate": 6.780967917646518e-06,
"loss": 2.8149484634399413,
"step": 4730
},
{
"epoch": 0.909702209414025,
"grad_norm": 0.13186109066009521,
"learning_rate": 6.640019238541727e-06,
"loss": 2.821930694580078,
"step": 4735
},
{
"epoch": 0.9106628242074928,
"grad_norm": 0.12498224526643753,
"learning_rate": 6.5005176989394335e-06,
"loss": 2.819742774963379,
"step": 4740
},
{
"epoch": 0.9116234390009607,
"grad_norm": 0.12152751535177231,
"learning_rate": 6.362464707030334e-06,
"loss": 2.818644332885742,
"step": 4745
},
{
"epoch": 0.9125840537944284,
"grad_norm": 0.12503238022327423,
"learning_rate": 6.225861656382825e-06,
"loss": 2.8183855056762694,
"step": 4750
},
{
"epoch": 0.9135446685878963,
"grad_norm": 0.1193387359380722,
"learning_rate": 6.090709925928938e-06,
"loss": 2.8130598068237305,
"step": 4755
},
{
"epoch": 0.914505283381364,
"grad_norm": 0.12339451164007187,
"learning_rate": 5.957010879950446e-06,
"loss": 2.81536979675293,
"step": 4760
},
{
"epoch": 0.9154658981748319,
"grad_norm": 0.11361610144376755,
"learning_rate": 5.824765868065101e-06,
"loss": 2.814308929443359,
"step": 4765
},
{
"epoch": 0.9164265129682997,
"grad_norm": 0.11605164408683777,
"learning_rate": 5.69397622521291e-06,
"loss": 2.8164608001708986,
"step": 4770
},
{
"epoch": 0.9173871277617676,
"grad_norm": 0.11599262058734894,
"learning_rate": 5.564643271642799e-06,
"loss": 2.817517852783203,
"step": 4775
},
{
"epoch": 0.9183477425552353,
"grad_norm": 0.11421811580657959,
"learning_rate": 5.436768312899226e-06,
"loss": 2.8213729858398438,
"step": 4780
},
{
"epoch": 0.9193083573487032,
"grad_norm": 0.11812606453895569,
"learning_rate": 5.310352639808968e-06,
"loss": 2.816654586791992,
"step": 4785
},
{
"epoch": 0.920268972142171,
"grad_norm": 0.12096104770898819,
"learning_rate": 5.185397528468155e-06,
"loss": 2.8137371063232424,
"step": 4790
},
{
"epoch": 0.9212295869356388,
"grad_norm": 0.11127284914255142,
"learning_rate": 5.061904240229309e-06,
"loss": 2.8210193634033205,
"step": 4795
},
{
"epoch": 0.9221902017291066,
"grad_norm": 0.1057458147406578,
"learning_rate": 4.939874021688739e-06,
"loss": 2.816180419921875,
"step": 4800
},
{
"epoch": 0.9231508165225745,
"grad_norm": 0.10853277891874313,
"learning_rate": 4.819308104673769e-06,
"loss": 2.8158794403076173,
"step": 4805
},
{
"epoch": 0.9241114313160422,
"grad_norm": 0.11734215915203094,
"learning_rate": 4.700207706230513e-06,
"loss": 2.818410873413086,
"step": 4810
},
{
"epoch": 0.9250720461095101,
"grad_norm": 0.11830545216798782,
"learning_rate": 4.582574028611435e-06,
"loss": 2.8172992706298827,
"step": 4815
},
{
"epoch": 0.9260326609029779,
"grad_norm": 0.17156197130680084,
"learning_rate": 4.466408259263243e-06,
"loss": 2.816666030883789,
"step": 4820
},
{
"epoch": 0.9269932756964457,
"grad_norm": 0.11448012292385101,
"learning_rate": 4.351711570815014e-06,
"loss": 2.8137338638305662,
"step": 4825
},
{
"epoch": 0.9279538904899135,
"grad_norm": 0.10624872893095016,
"learning_rate": 4.238485121066154e-06,
"loss": 2.8165781021118166,
"step": 4830
},
{
"epoch": 0.9289145052833814,
"grad_norm": 0.12709708511829376,
"learning_rate": 4.126730052974908e-06,
"loss": 2.814931869506836,
"step": 4835
},
{
"epoch": 0.9298751200768491,
"grad_norm": 0.15824326872825623,
"learning_rate": 4.016447494646718e-06,
"loss": 2.8185230255126954,
"step": 4840
},
{
"epoch": 0.930835734870317,
"grad_norm": 0.12272030860185623,
"learning_rate": 3.907638559322817e-06,
"loss": 2.8193038940429687,
"step": 4845
},
{
"epoch": 0.9317963496637848,
"grad_norm": 0.14079715311527252,
"learning_rate": 3.8003043453691207e-06,
"loss": 2.813698959350586,
"step": 4850
},
{
"epoch": 0.9327569644572526,
"grad_norm": 0.16456739604473114,
"learning_rate": 3.69444593626495e-06,
"loss": 2.8172222137451173,
"step": 4855
},
{
"epoch": 0.9337175792507204,
"grad_norm": 0.1126532033085823,
"learning_rate": 3.5900644005922465e-06,
"loss": 2.8200458526611327,
"step": 4860
},
{
"epoch": 0.9346781940441883,
"grad_norm": 0.12338647246360779,
"learning_rate": 3.48716079202474e-06,
"loss": 2.8146718978881835,
"step": 4865
},
{
"epoch": 0.9356388088376562,
"grad_norm": 0.13380901515483856,
"learning_rate": 3.385736149317264e-06,
"loss": 2.81722412109375,
"step": 4870
},
{
"epoch": 0.9365994236311239,
"grad_norm": 0.12491049617528915,
"learning_rate": 3.2857914962953926e-06,
"loss": 2.8170299530029297,
"step": 4875
},
{
"epoch": 0.9375600384245918,
"grad_norm": 0.14564266800880432,
"learning_rate": 3.1873278418449e-06,
"loss": 2.814952087402344,
"step": 4880
},
{
"epoch": 0.9385206532180596,
"grad_norm": 0.1412232220172882,
"learning_rate": 3.090346179901837e-06,
"loss": 2.816575622558594,
"step": 4885
},
{
"epoch": 0.9394812680115274,
"grad_norm": 0.12283110618591309,
"learning_rate": 2.99484748944227e-06,
"loss": 2.816323471069336,
"step": 4890
},
{
"epoch": 0.9404418828049952,
"grad_norm": 0.131371408700943,
"learning_rate": 2.9008327344724913e-06,
"loss": 2.8180400848388674,
"step": 4895
},
{
"epoch": 0.9414024975984631,
"grad_norm": 0.10252773761749268,
"learning_rate": 2.808302864019374e-06,
"loss": 2.8142473220825197,
"step": 4900
},
{
"epoch": 0.9423631123919308,
"grad_norm": 0.12445132434368134,
"learning_rate": 2.7172588121205983e-06,
"loss": 2.8168582916259766,
"step": 4905
},
{
"epoch": 0.9433237271853987,
"grad_norm": 0.1234014481306076,
"learning_rate": 2.62770149781541e-06,
"loss": 2.817163848876953,
"step": 4910
},
{
"epoch": 0.9442843419788665,
"grad_norm": 0.1050441563129425,
"learning_rate": 2.539631825135191e-06,
"loss": 2.823550796508789,
"step": 4915
},
{
"epoch": 0.9452449567723343,
"grad_norm": 0.11725042015314102,
"learning_rate": 2.4530506830944876e-06,
"loss": 2.820196533203125,
"step": 4920
},
{
"epoch": 0.9462055715658021,
"grad_norm": 0.11576535552740097,
"learning_rate": 2.367958945681897e-06,
"loss": 2.814539337158203,
"step": 4925
},
{
"epoch": 0.94716618635927,
"grad_norm": 0.10262072086334229,
"learning_rate": 2.2843574718513114e-06,
"loss": 2.8174243927001954,
"step": 4930
},
{
"epoch": 0.9481268011527377,
"grad_norm": 0.10643728822469711,
"learning_rate": 2.2022471055132553e-06,
"loss": 2.815760040283203,
"step": 4935
},
{
"epoch": 0.9490874159462056,
"grad_norm": 0.09736967086791992,
"learning_rate": 2.1216286755263434e-06,
"loss": 2.8164119720458984,
"step": 4940
},
{
"epoch": 0.9500480307396734,
"grad_norm": 0.1281130164861679,
"learning_rate": 2.042502995688905e-06,
"loss": 2.8168426513671876,
"step": 4945
},
{
"epoch": 0.9510086455331412,
"grad_norm": 0.10349207371473312,
"learning_rate": 1.964870864730789e-06,
"loss": 2.8166751861572266,
"step": 4950
},
{
"epoch": 0.951969260326609,
"grad_norm": 0.13925877213478088,
"learning_rate": 1.8887330663053536e-06,
"loss": 2.816518783569336,
"step": 4955
},
{
"epoch": 0.9529298751200769,
"grad_norm": 0.09690409153699875,
"learning_rate": 1.8140903689814079e-06,
"loss": 2.8186017990112306,
"step": 4960
},
{
"epoch": 0.9538904899135446,
"grad_norm": 0.11781689524650574,
"learning_rate": 1.740943526235583e-06,
"loss": 2.816529083251953,
"step": 4965
},
{
"epoch": 0.9548511047070125,
"grad_norm": 0.14491458237171173,
"learning_rate": 1.6692932764447054e-06,
"loss": 2.8170598983764648,
"step": 4970
},
{
"epoch": 0.9558117195004803,
"grad_norm": 0.09681481868028641,
"learning_rate": 1.5991403428783188e-06,
"loss": 2.8186452865600584,
"step": 4975
},
{
"epoch": 0.9567723342939481,
"grad_norm": 0.10160388797521591,
"learning_rate": 1.5304854336913752e-06,
"loss": 2.8164340972900392,
"step": 4980
},
{
"epoch": 0.9577329490874159,
"grad_norm": 0.09434349089860916,
"learning_rate": 1.463329241917105e-06,
"loss": 2.8180526733398437,
"step": 4985
},
{
"epoch": 0.9586935638808838,
"grad_norm": 0.10586917400360107,
"learning_rate": 1.397672445460024e-06,
"loss": 2.813650894165039,
"step": 4990
},
{
"epoch": 0.9596541786743515,
"grad_norm": 0.09675531089305878,
"learning_rate": 1.333515707089089e-06,
"loss": 2.8159751892089844,
"step": 4995
},
{
"epoch": 0.9606147934678194,
"grad_norm": 0.1004442647099495,
"learning_rate": 1.2708596744309685e-06,
"loss": 2.816268730163574,
"step": 5000
},
{
"epoch": 0.9615754082612872,
"grad_norm": 0.09772492200136185,
"learning_rate": 1.209704979963616e-06,
"loss": 2.8217578887939454,
"step": 5005
},
{
"epoch": 0.962536023054755,
"grad_norm": 0.09763862192630768,
"learning_rate": 1.1500522410096912e-06,
"loss": 2.8170074462890624,
"step": 5010
},
{
"epoch": 0.9634966378482228,
"grad_norm": 0.09506134688854218,
"learning_rate": 1.0919020597305649e-06,
"loss": 2.8169906616210936,
"step": 5015
},
{
"epoch": 0.9644572526416907,
"grad_norm": 0.09169244021177292,
"learning_rate": 1.0352550231200407e-06,
"loss": 2.818475341796875,
"step": 5020
},
{
"epoch": 0.9654178674351584,
"grad_norm": 0.09423839300870895,
"learning_rate": 9.801117029985596e-07,
"loss": 2.816048812866211,
"step": 5025
},
{
"epoch": 0.9663784822286263,
"grad_norm": 0.0988612100481987,
"learning_rate": 9.264726560073377e-07,
"loss": 2.8169836044311523,
"step": 5030
},
{
"epoch": 0.9673390970220941,
"grad_norm": 0.09173572063446045,
"learning_rate": 8.743384236028051e-07,
"loss": 2.8156835556030275,
"step": 5035
},
{
"epoch": 0.968299711815562,
"grad_norm": 0.09405640512704849,
"learning_rate": 8.237095320511589e-07,
"loss": 2.8169689178466797,
"step": 5040
},
{
"epoch": 0.9692603266090298,
"grad_norm": 0.09245765209197998,
"learning_rate": 7.745864924229351e-07,
"loss": 2.8151527404785157,
"step": 5045
},
{
"epoch": 0.9702209414024976,
"grad_norm": 0.09277483075857162,
"learning_rate": 7.26969800588012e-07,
"loss": 2.8148799896240235,
"step": 5050
},
{
"epoch": 0.9711815561959655,
"grad_norm": 0.09106960892677307,
"learning_rate": 6.808599372104817e-07,
"loss": 2.812705230712891,
"step": 5055
},
{
"epoch": 0.9721421709894332,
"grad_norm": 0.09490223228931427,
"learning_rate": 6.362573677438199e-07,
"loss": 2.8207538604736326,
"step": 5060
},
{
"epoch": 0.9731027857829011,
"grad_norm": 0.0929575189948082,
"learning_rate": 5.931625424262731e-07,
"loss": 2.817534637451172,
"step": 5065
},
{
"epoch": 0.9740634005763689,
"grad_norm": 0.0907764732837677,
"learning_rate": 5.51575896276163e-07,
"loss": 2.8162796020507814,
"step": 5070
},
{
"epoch": 0.9750240153698367,
"grad_norm": 0.09307243674993515,
"learning_rate": 5.11497849087622e-07,
"loss": 2.813416862487793,
"step": 5075
},
{
"epoch": 0.9759846301633045,
"grad_norm": 0.0937948152422905,
"learning_rate": 4.7292880542634805e-07,
"loss": 2.8132902145385743,
"step": 5080
},
{
"epoch": 0.9769452449567724,
"grad_norm": 0.0905749499797821,
"learning_rate": 4.358691546254067e-07,
"loss": 2.814370346069336,
"step": 5085
},
{
"epoch": 0.9779058597502401,
"grad_norm": 0.08927202224731445,
"learning_rate": 4.0031927078145176e-07,
"loss": 2.8167686462402344,
"step": 5090
},
{
"epoch": 0.978866474543708,
"grad_norm": 0.09388428926467896,
"learning_rate": 3.662795127508111e-07,
"loss": 2.818276786804199,
"step": 5095
},
{
"epoch": 0.9798270893371758,
"grad_norm": 0.08715725690126419,
"learning_rate": 3.3375022414598994e-07,
"loss": 2.8158248901367187,
"step": 5100
},
{
"epoch": 0.9807877041306436,
"grad_norm": 0.08722332119941711,
"learning_rate": 3.027317333321233e-07,
"loss": 2.8168060302734377,
"step": 5105
},
{
"epoch": 0.9817483189241114,
"grad_norm": 0.08853864669799805,
"learning_rate": 2.7322435342364556e-07,
"loss": 2.814595603942871,
"step": 5110
},
{
"epoch": 0.9827089337175793,
"grad_norm": 0.08678946644067764,
"learning_rate": 2.452283822812262e-07,
"loss": 2.8177654266357424,
"step": 5115
},
{
"epoch": 0.983669548511047,
"grad_norm": 0.09480496495962143,
"learning_rate": 2.1874410250863893e-07,
"loss": 2.8194339752197264,
"step": 5120
},
{
"epoch": 0.9846301633045149,
"grad_norm": 0.08985020220279694,
"learning_rate": 1.9377178145003059e-07,
"loss": 2.8135372161865235,
"step": 5125
},
{
"epoch": 0.9855907780979827,
"grad_norm": 0.08478531241416931,
"learning_rate": 1.7031167118708998e-07,
"loss": 2.8173141479492188,
"step": 5130
},
{
"epoch": 0.9865513928914506,
"grad_norm": 0.08679889887571335,
"learning_rate": 1.4836400853666662e-07,
"loss": 2.8155088424682617,
"step": 5135
},
{
"epoch": 0.9875120076849183,
"grad_norm": 0.0879487469792366,
"learning_rate": 1.2792901504820595e-07,
"loss": 2.8137956619262696,
"step": 5140
},
{
"epoch": 0.9884726224783862,
"grad_norm": 0.0856156200170517,
"learning_rate": 1.0900689700166776e-07,
"loss": 2.8191757202148438,
"step": 5145
},
{
"epoch": 0.989433237271854,
"grad_norm": 0.08409229665994644,
"learning_rate": 9.159784540531124e-08,
"loss": 2.8132045745849608,
"step": 5150
},
{
"epoch": 0.9903938520653218,
"grad_norm": 0.08568743616342545,
"learning_rate": 7.57020359938798e-08,
"loss": 2.814972496032715,
"step": 5155
},
{
"epoch": 0.9913544668587896,
"grad_norm": 0.08696384727954865,
"learning_rate": 6.131962922673595e-08,
"loss": 2.817178153991699,
"step": 5160
},
{
"epoch": 0.9923150816522575,
"grad_norm": 0.08134496212005615,
"learning_rate": 4.845077028631239e-08,
"loss": 2.818017578125,
"step": 5165
},
{
"epoch": 0.9932756964457252,
"grad_norm": 0.08577294647693634,
"learning_rate": 3.709558907659671e-08,
"loss": 2.8143449783325196,
"step": 5170
},
{
"epoch": 0.9942363112391931,
"grad_norm": 0.08689381182193756,
"learning_rate": 2.7254200221848988e-08,
"loss": 2.8173112869262695,
"step": 5175
},
{
"epoch": 0.9951969260326609,
"grad_norm": 0.0842234268784523,
"learning_rate": 1.8926703065436087e-08,
"loss": 2.815708541870117,
"step": 5180
},
{
"epoch": 0.9961575408261287,
"grad_norm": 0.08280805498361588,
"learning_rate": 1.2113181668815808e-08,
"loss": 2.8149890899658203,
"step": 5185
},
{
"epoch": 0.9971181556195965,
"grad_norm": 0.08539645373821259,
"learning_rate": 6.813704810704201e-09,
"loss": 2.8178043365478516,
"step": 5190
},
{
"epoch": 0.9980787704130644,
"grad_norm": 0.08579740673303604,
"learning_rate": 3.028325986392799e-09,
"loss": 2.816665267944336,
"step": 5195
},
{
"epoch": 0.9990393852065321,
"grad_norm": 0.08700989931821823,
"learning_rate": 7.570834071823905e-10,
"loss": 2.8163612365722654,
"step": 5200
},
{
"epoch": 1.0,
"grad_norm": 0.08663682639598846,
"learning_rate": 0.0,
"loss": 2.8194324493408205,
"step": 5205
}
],
"logging_steps": 5,
"max_steps": 5205,
"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": 1.4701096488778754e+20,
"train_batch_size": 64,
"trial_name": null,
"trial_params": null
}