Files
qwen3-4b-beauty-mix/trainer_state.json
ModelHub XC 88d4888347 初始化项目,由ModelHub XC社区提供模型
Model: abdoul-rz/qwen3-4b-beauty-mix
Source: Original Platform
2026-09-08 23:42:48 +08:00

6890 lines
167 KiB
JSON

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 9782,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0010223119584941345,
"grad_norm": 70.94046783447266,
"learning_rate": 4.596527068437181e-07,
"loss": 5.7729,
"step": 10
},
{
"epoch": 0.002044623916988269,
"grad_norm": 36.98649597167969,
"learning_rate": 9.703779366700715e-07,
"loss": 5.4769,
"step": 20
},
{
"epoch": 0.0030669358754824034,
"grad_norm": 14.041007995605469,
"learning_rate": 1.481103166496425e-06,
"loss": 4.1936,
"step": 30
},
{
"epoch": 0.004089247833976538,
"grad_norm": 8.187762260437012,
"learning_rate": 1.9918283963227787e-06,
"loss": 3.9063,
"step": 40
},
{
"epoch": 0.005111559792470672,
"grad_norm": 9.793929100036621,
"learning_rate": 2.502553626149132e-06,
"loss": 3.3878,
"step": 50
},
{
"epoch": 0.006133871750964807,
"grad_norm": 6.04489803314209,
"learning_rate": 3.013278855975485e-06,
"loss": 3.1355,
"step": 60
},
{
"epoch": 0.007156183709458941,
"grad_norm": 4.14630126953125,
"learning_rate": 3.5240040858018386e-06,
"loss": 2.9229,
"step": 70
},
{
"epoch": 0.008178495667953076,
"grad_norm": 4.584634304046631,
"learning_rate": 4.034729315628192e-06,
"loss": 2.7581,
"step": 80
},
{
"epoch": 0.00920080762644721,
"grad_norm": 4.384355068206787,
"learning_rate": 4.5454545454545455e-06,
"loss": 2.9603,
"step": 90
},
{
"epoch": 0.010223119584941345,
"grad_norm": 4.592817306518555,
"learning_rate": 5.056179775280899e-06,
"loss": 2.8967,
"step": 100
},
{
"epoch": 0.011245431543435479,
"grad_norm": 4.680253982543945,
"learning_rate": 5.5669050051072524e-06,
"loss": 2.8365,
"step": 110
},
{
"epoch": 0.012267743501929614,
"grad_norm": 2.460531234741211,
"learning_rate": 6.077630234933606e-06,
"loss": 2.529,
"step": 120
},
{
"epoch": 0.013290055460423748,
"grad_norm": 3.962606906890869,
"learning_rate": 6.58835546475996e-06,
"loss": 2.5749,
"step": 130
},
{
"epoch": 0.014312367418917882,
"grad_norm": 3.244239330291748,
"learning_rate": 7.099080694586313e-06,
"loss": 2.4756,
"step": 140
},
{
"epoch": 0.015334679377412017,
"grad_norm": 4.2977705001831055,
"learning_rate": 7.609805924412666e-06,
"loss": 2.5418,
"step": 150
},
{
"epoch": 0.01635699133590615,
"grad_norm": 4.5740132331848145,
"learning_rate": 8.120531154239019e-06,
"loss": 2.4251,
"step": 160
},
{
"epoch": 0.017379303294400288,
"grad_norm": 2.3269731998443604,
"learning_rate": 8.631256384065373e-06,
"loss": 2.3459,
"step": 170
},
{
"epoch": 0.01840161525289442,
"grad_norm": 4.277726650238037,
"learning_rate": 9.141981613891727e-06,
"loss": 2.1922,
"step": 180
},
{
"epoch": 0.019423927211388557,
"grad_norm": 4.3737077713012695,
"learning_rate": 9.65270684371808e-06,
"loss": 2.1388,
"step": 190
},
{
"epoch": 0.02044623916988269,
"grad_norm": 4.4290947914123535,
"learning_rate": 1.0163432073544434e-05,
"loss": 2.1335,
"step": 200
},
{
"epoch": 0.021468551128376825,
"grad_norm": 2.84171724319458,
"learning_rate": 1.0674157303370787e-05,
"loss": 2.0611,
"step": 210
},
{
"epoch": 0.022490863086870958,
"grad_norm": 2.303724765777588,
"learning_rate": 1.118488253319714e-05,
"loss": 2.0277,
"step": 220
},
{
"epoch": 0.023513175045365094,
"grad_norm": 2.510511875152588,
"learning_rate": 1.1695607763023494e-05,
"loss": 2.083,
"step": 230
},
{
"epoch": 0.024535487003859227,
"grad_norm": 4.074395656585693,
"learning_rate": 1.2206332992849848e-05,
"loss": 1.9243,
"step": 240
},
{
"epoch": 0.025557798962353363,
"grad_norm": 2.922254800796509,
"learning_rate": 1.27170582226762e-05,
"loss": 1.9579,
"step": 250
},
{
"epoch": 0.026580110920847496,
"grad_norm": 2.5318799018859863,
"learning_rate": 1.3227783452502553e-05,
"loss": 1.926,
"step": 260
},
{
"epoch": 0.027602422879341632,
"grad_norm": 2.9087040424346924,
"learning_rate": 1.3738508682328908e-05,
"loss": 1.855,
"step": 270
},
{
"epoch": 0.028624734837835765,
"grad_norm": 2.358649492263794,
"learning_rate": 1.424923391215526e-05,
"loss": 1.8568,
"step": 280
},
{
"epoch": 0.0296470467963299,
"grad_norm": 2.470640182495117,
"learning_rate": 1.4759959141981613e-05,
"loss": 1.8934,
"step": 290
},
{
"epoch": 0.030669358754824034,
"grad_norm": 2.740605592727661,
"learning_rate": 1.5270684371807967e-05,
"loss": 1.8093,
"step": 300
},
{
"epoch": 0.03169167071331817,
"grad_norm": 2.383192777633667,
"learning_rate": 1.578140960163432e-05,
"loss": 1.8581,
"step": 310
},
{
"epoch": 0.0327139826718123,
"grad_norm": 2.431966781616211,
"learning_rate": 1.6292134831460676e-05,
"loss": 1.8639,
"step": 320
},
{
"epoch": 0.03373629463030644,
"grad_norm": 2.363967180252075,
"learning_rate": 1.680286006128703e-05,
"loss": 1.8074,
"step": 330
},
{
"epoch": 0.034758606588800575,
"grad_norm": 2.1545140743255615,
"learning_rate": 1.731358529111338e-05,
"loss": 1.7677,
"step": 340
},
{
"epoch": 0.035780918547294704,
"grad_norm": 2.7959671020507812,
"learning_rate": 1.7824310520939735e-05,
"loss": 1.8268,
"step": 350
},
{
"epoch": 0.03680323050578884,
"grad_norm": 4.4346113204956055,
"learning_rate": 1.833503575076609e-05,
"loss": 1.8369,
"step": 360
},
{
"epoch": 0.03782554246428298,
"grad_norm": 2.494288682937622,
"learning_rate": 1.884576098059244e-05,
"loss": 1.6866,
"step": 370
},
{
"epoch": 0.03884785442277711,
"grad_norm": 2.819344997406006,
"learning_rate": 1.9356486210418795e-05,
"loss": 1.6996,
"step": 380
},
{
"epoch": 0.03987016638127124,
"grad_norm": 2.765200614929199,
"learning_rate": 1.986721144024515e-05,
"loss": 1.7623,
"step": 390
},
{
"epoch": 0.04089247833976538,
"grad_norm": 2.703268527984619,
"learning_rate": 2.03779366700715e-05,
"loss": 1.7848,
"step": 400
},
{
"epoch": 0.041914790298259515,
"grad_norm": 4.231158256530762,
"learning_rate": 2.0888661899897858e-05,
"loss": 1.7764,
"step": 410
},
{
"epoch": 0.04293710225675365,
"grad_norm": 2.221487045288086,
"learning_rate": 2.139938712972421e-05,
"loss": 1.7455,
"step": 420
},
{
"epoch": 0.04395941421524778,
"grad_norm": 2.8403496742248535,
"learning_rate": 2.1910112359550563e-05,
"loss": 1.7515,
"step": 430
},
{
"epoch": 0.044981726173741916,
"grad_norm": 1.87192964553833,
"learning_rate": 2.2420837589376917e-05,
"loss": 1.7321,
"step": 440
},
{
"epoch": 0.04600403813223605,
"grad_norm": 2.076911687850952,
"learning_rate": 2.2931562819203268e-05,
"loss": 1.7107,
"step": 450
},
{
"epoch": 0.04702635009073019,
"grad_norm": 4.201175212860107,
"learning_rate": 2.3442288049029623e-05,
"loss": 1.7197,
"step": 460
},
{
"epoch": 0.04804866204922432,
"grad_norm": 1.9488505125045776,
"learning_rate": 2.3953013278855977e-05,
"loss": 1.6227,
"step": 470
},
{
"epoch": 0.049070974007718454,
"grad_norm": 1.937862515449524,
"learning_rate": 2.446373850868233e-05,
"loss": 1.573,
"step": 480
},
{
"epoch": 0.05009328596621259,
"grad_norm": 2.538717031478882,
"learning_rate": 2.4974463738508682e-05,
"loss": 1.581,
"step": 490
},
{
"epoch": 0.05111559792470673,
"grad_norm": 1.7750344276428223,
"learning_rate": 2.548518896833504e-05,
"loss": 1.5403,
"step": 500
},
{
"epoch": 0.052137909883200856,
"grad_norm": 2.2738513946533203,
"learning_rate": 2.599591419816139e-05,
"loss": 1.5419,
"step": 510
},
{
"epoch": 0.05316022184169499,
"grad_norm": 2.131457567214966,
"learning_rate": 2.6506639427987745e-05,
"loss": 1.544,
"step": 520
},
{
"epoch": 0.05418253380018913,
"grad_norm": 2.3775675296783447,
"learning_rate": 2.70173646578141e-05,
"loss": 1.4898,
"step": 530
},
{
"epoch": 0.055204845758683264,
"grad_norm": 2.117812395095825,
"learning_rate": 2.752808988764045e-05,
"loss": 1.4654,
"step": 540
},
{
"epoch": 0.056227157717177394,
"grad_norm": 1.9138811826705933,
"learning_rate": 2.8038815117466805e-05,
"loss": 1.3641,
"step": 550
},
{
"epoch": 0.05724946967567153,
"grad_norm": 2.154578924179077,
"learning_rate": 2.854954034729316e-05,
"loss": 1.3263,
"step": 560
},
{
"epoch": 0.058271781634165666,
"grad_norm": 2.099501132965088,
"learning_rate": 2.906026557711951e-05,
"loss": 1.3474,
"step": 570
},
{
"epoch": 0.0592940935926598,
"grad_norm": 1.8399790525436401,
"learning_rate": 2.9570990806945864e-05,
"loss": 1.3702,
"step": 580
},
{
"epoch": 0.06031640555115393,
"grad_norm": 2.0430350303649902,
"learning_rate": 3.008171603677222e-05,
"loss": 1.3567,
"step": 590
},
{
"epoch": 0.06133871750964807,
"grad_norm": 1.9482533931732178,
"learning_rate": 3.059244126659857e-05,
"loss": 1.2809,
"step": 600
},
{
"epoch": 0.062361029468142204,
"grad_norm": 1.9207191467285156,
"learning_rate": 3.110316649642493e-05,
"loss": 1.2442,
"step": 610
},
{
"epoch": 0.06338334142663633,
"grad_norm": 2.114955186843872,
"learning_rate": 3.1613891726251275e-05,
"loss": 1.2309,
"step": 620
},
{
"epoch": 0.06440565338513048,
"grad_norm": 1.8856667280197144,
"learning_rate": 3.212461695607763e-05,
"loss": 1.1752,
"step": 630
},
{
"epoch": 0.0654279653436246,
"grad_norm": 1.723923683166504,
"learning_rate": 3.263534218590398e-05,
"loss": 1.2686,
"step": 640
},
{
"epoch": 0.06645027730211873,
"grad_norm": 1.8947356939315796,
"learning_rate": 3.314606741573034e-05,
"loss": 1.2698,
"step": 650
},
{
"epoch": 0.06747258926061288,
"grad_norm": 1.9917230606079102,
"learning_rate": 3.365679264555669e-05,
"loss": 1.1724,
"step": 660
},
{
"epoch": 0.06849490121910701,
"grad_norm": 2.0864381790161133,
"learning_rate": 3.4167517875383046e-05,
"loss": 1.3131,
"step": 670
},
{
"epoch": 0.06951721317760115,
"grad_norm": 1.6907154321670532,
"learning_rate": 3.46782431052094e-05,
"loss": 1.1841,
"step": 680
},
{
"epoch": 0.07053952513609528,
"grad_norm": 1.76568603515625,
"learning_rate": 3.5188968335035755e-05,
"loss": 1.2088,
"step": 690
},
{
"epoch": 0.07156183709458941,
"grad_norm": 1.9172755479812622,
"learning_rate": 3.569969356486211e-05,
"loss": 1.1739,
"step": 700
},
{
"epoch": 0.07258414905308355,
"grad_norm": 2.5354256629943848,
"learning_rate": 3.621041879468846e-05,
"loss": 1.2373,
"step": 710
},
{
"epoch": 0.07360646101157768,
"grad_norm": 1.7339375019073486,
"learning_rate": 3.672114402451481e-05,
"loss": 1.1339,
"step": 720
},
{
"epoch": 0.07462877297007181,
"grad_norm": 1.7091752290725708,
"learning_rate": 3.7231869254341165e-05,
"loss": 1.1807,
"step": 730
},
{
"epoch": 0.07565108492856595,
"grad_norm": 1.8734017610549927,
"learning_rate": 3.774259448416752e-05,
"loss": 1.1803,
"step": 740
},
{
"epoch": 0.07667339688706008,
"grad_norm": 1.9953020811080933,
"learning_rate": 3.8253319713993874e-05,
"loss": 1.1964,
"step": 750
},
{
"epoch": 0.07769570884555423,
"grad_norm": 1.9818227291107178,
"learning_rate": 3.876404494382023e-05,
"loss": 1.1366,
"step": 760
},
{
"epoch": 0.07871802080404836,
"grad_norm": 1.6370587348937988,
"learning_rate": 3.927477017364658e-05,
"loss": 1.2463,
"step": 770
},
{
"epoch": 0.07974033276254248,
"grad_norm": 3.5624840259552,
"learning_rate": 3.978549540347294e-05,
"loss": 1.2779,
"step": 780
},
{
"epoch": 0.08076264472103663,
"grad_norm": 1.5966604948043823,
"learning_rate": 4.0296220633299284e-05,
"loss": 1.1991,
"step": 790
},
{
"epoch": 0.08178495667953076,
"grad_norm": 1.6895662546157837,
"learning_rate": 4.080694586312564e-05,
"loss": 1.1226,
"step": 800
},
{
"epoch": 0.0828072686380249,
"grad_norm": 1.7477182149887085,
"learning_rate": 4.131767109295199e-05,
"loss": 1.0991,
"step": 810
},
{
"epoch": 0.08382958059651903,
"grad_norm": 3.4450631141662598,
"learning_rate": 4.182839632277835e-05,
"loss": 1.1443,
"step": 820
},
{
"epoch": 0.08485189255501316,
"grad_norm": 1.7347668409347534,
"learning_rate": 4.23391215526047e-05,
"loss": 1.1085,
"step": 830
},
{
"epoch": 0.0858742045135073,
"grad_norm": 1.52926766872406,
"learning_rate": 4.2849846782431056e-05,
"loss": 1.1038,
"step": 840
},
{
"epoch": 0.08689651647200143,
"grad_norm": 2.0358810424804688,
"learning_rate": 4.33605720122574e-05,
"loss": 1.0814,
"step": 850
},
{
"epoch": 0.08791882843049556,
"grad_norm": 1.840956211090088,
"learning_rate": 4.387129724208376e-05,
"loss": 1.1028,
"step": 860
},
{
"epoch": 0.0889411403889897,
"grad_norm": 1.5585577487945557,
"learning_rate": 4.438202247191011e-05,
"loss": 1.0584,
"step": 870
},
{
"epoch": 0.08996345234748383,
"grad_norm": 1.948915958404541,
"learning_rate": 4.489274770173647e-05,
"loss": 1.1062,
"step": 880
},
{
"epoch": 0.09098576430597798,
"grad_norm": 1.423173189163208,
"learning_rate": 4.540347293156282e-05,
"loss": 1.0567,
"step": 890
},
{
"epoch": 0.0920080762644721,
"grad_norm": 1.5492244958877563,
"learning_rate": 4.5914198161389175e-05,
"loss": 1.0518,
"step": 900
},
{
"epoch": 0.09303038822296623,
"grad_norm": 1.58083975315094,
"learning_rate": 4.642492339121553e-05,
"loss": 1.0511,
"step": 910
},
{
"epoch": 0.09405270018146038,
"grad_norm": 1.8411082029342651,
"learning_rate": 4.693564862104188e-05,
"loss": 1.0845,
"step": 920
},
{
"epoch": 0.0950750121399545,
"grad_norm": 1.9242267608642578,
"learning_rate": 4.744637385086824e-05,
"loss": 1.0651,
"step": 930
},
{
"epoch": 0.09609732409844864,
"grad_norm": 1.415640950202942,
"learning_rate": 4.795709908069459e-05,
"loss": 1.0411,
"step": 940
},
{
"epoch": 0.09711963605694278,
"grad_norm": 1.858831763267517,
"learning_rate": 4.846782431052094e-05,
"loss": 1.062,
"step": 950
},
{
"epoch": 0.09814194801543691,
"grad_norm": 1.8358367681503296,
"learning_rate": 4.8978549540347294e-05,
"loss": 1.0559,
"step": 960
},
{
"epoch": 0.09916425997393105,
"grad_norm": 1.739005208015442,
"learning_rate": 4.948927477017365e-05,
"loss": 1.0672,
"step": 970
},
{
"epoch": 0.10018657193242518,
"grad_norm": 1.6632468700408936,
"learning_rate": 5e-05,
"loss": 1.028,
"step": 980
},
{
"epoch": 0.10120888389091931,
"grad_norm": 1.472429871559143,
"learning_rate": 4.9999840798229326e-05,
"loss": 1.0349,
"step": 990
},
{
"epoch": 0.10223119584941345,
"grad_norm": 1.6684609651565552,
"learning_rate": 4.999936319494492e-05,
"loss": 1.0208,
"step": 1000
},
{
"epoch": 0.10325350780790758,
"grad_norm": 1.838606834411621,
"learning_rate": 4.9998567196229584e-05,
"loss": 1.0962,
"step": 1010
},
{
"epoch": 0.10427581976640171,
"grad_norm": 1.6233595609664917,
"learning_rate": 4.999745281222129e-05,
"loss": 1.0357,
"step": 1020
},
{
"epoch": 0.10529813172489585,
"grad_norm": 1.7310131788253784,
"learning_rate": 4.9996020057112994e-05,
"loss": 1.041,
"step": 1030
},
{
"epoch": 0.10632044368338998,
"grad_norm": 2.362665891647339,
"learning_rate": 4.999426894915245e-05,
"loss": 1.0324,
"step": 1040
},
{
"epoch": 0.10734275564188413,
"grad_norm": 1.5394867658615112,
"learning_rate": 4.9992199510642035e-05,
"loss": 1.0059,
"step": 1050
},
{
"epoch": 0.10836506760037826,
"grad_norm": 1.718542218208313,
"learning_rate": 4.99898117679384e-05,
"loss": 1.0207,
"step": 1060
},
{
"epoch": 0.10938737955887239,
"grad_norm": 1.6890432834625244,
"learning_rate": 4.998710575145218e-05,
"loss": 0.998,
"step": 1070
},
{
"epoch": 0.11040969151736653,
"grad_norm": 1.5707371234893799,
"learning_rate": 4.9984081495647586e-05,
"loss": 1.004,
"step": 1080
},
{
"epoch": 0.11143200347586066,
"grad_norm": 1.783901572227478,
"learning_rate": 4.998073903904196e-05,
"loss": 0.9903,
"step": 1090
},
{
"epoch": 0.11245431543435479,
"grad_norm": 1.5425653457641602,
"learning_rate": 4.99770784242053e-05,
"loss": 0.9785,
"step": 1100
},
{
"epoch": 0.11347662739284893,
"grad_norm": 1.835707664489746,
"learning_rate": 4.997309969775973e-05,
"loss": 0.9604,
"step": 1110
},
{
"epoch": 0.11449893935134306,
"grad_norm": 1.2470365762710571,
"learning_rate": 4.996880291037886e-05,
"loss": 0.9544,
"step": 1120
},
{
"epoch": 0.1155212513098372,
"grad_norm": 1.2418577671051025,
"learning_rate": 4.99641881167872e-05,
"loss": 0.96,
"step": 1130
},
{
"epoch": 0.11654356326833133,
"grad_norm": 1.6778637170791626,
"learning_rate": 4.995925537575939e-05,
"loss": 0.9925,
"step": 1140
},
{
"epoch": 0.11756587522682546,
"grad_norm": 1.9518935680389404,
"learning_rate": 4.9954004750119545e-05,
"loss": 0.957,
"step": 1150
},
{
"epoch": 0.1185881871853196,
"grad_norm": 1.4519526958465576,
"learning_rate": 4.994843630674036e-05,
"loss": 0.9667,
"step": 1160
},
{
"epoch": 0.11961049914381373,
"grad_norm": 1.3157827854156494,
"learning_rate": 4.9942550116542316e-05,
"loss": 0.9849,
"step": 1170
},
{
"epoch": 0.12063281110230786,
"grad_norm": 1.329932689666748,
"learning_rate": 4.993634625449278e-05,
"loss": 0.9447,
"step": 1180
},
{
"epoch": 0.121655123060802,
"grad_norm": 1.5020705461502075,
"learning_rate": 4.992982479960501e-05,
"loss": 0.9484,
"step": 1190
},
{
"epoch": 0.12267743501929614,
"grad_norm": 1.5723028182983398,
"learning_rate": 4.9922985834937176e-05,
"loss": 0.9355,
"step": 1200
},
{
"epoch": 0.12369974697779028,
"grad_norm": 1.6413408517837524,
"learning_rate": 4.99158294475913e-05,
"loss": 0.9548,
"step": 1210
},
{
"epoch": 0.12472205893628441,
"grad_norm": 1.4169429540634155,
"learning_rate": 4.990835572871215e-05,
"loss": 0.9643,
"step": 1220
},
{
"epoch": 0.12574437089477855,
"grad_norm": 1.6595830917358398,
"learning_rate": 4.990056477348607e-05,
"loss": 0.9715,
"step": 1230
},
{
"epoch": 0.12676668285327267,
"grad_norm": 1.4884495735168457,
"learning_rate": 4.9892456681139766e-05,
"loss": 0.958,
"step": 1240
},
{
"epoch": 0.1277889948117668,
"grad_norm": 1.6100373268127441,
"learning_rate": 4.9884031554939046e-05,
"loss": 0.9318,
"step": 1250
},
{
"epoch": 0.12881130677026095,
"grad_norm": 1.4218831062316895,
"learning_rate": 4.987528950218753e-05,
"loss": 0.9508,
"step": 1260
},
{
"epoch": 0.12983361872875507,
"grad_norm": 1.289501667022705,
"learning_rate": 4.9866230634225205e-05,
"loss": 0.9465,
"step": 1270
},
{
"epoch": 0.1308559306872492,
"grad_norm": 1.3804965019226074,
"learning_rate": 4.985685506642713e-05,
"loss": 0.9644,
"step": 1280
},
{
"epoch": 0.13187824264574335,
"grad_norm": 1.6997615098953247,
"learning_rate": 4.984716291820185e-05,
"loss": 0.9413,
"step": 1290
},
{
"epoch": 0.13290055460423747,
"grad_norm": 1.2760159969329834,
"learning_rate": 4.983715431298994e-05,
"loss": 0.9211,
"step": 1300
},
{
"epoch": 0.1339228665627316,
"grad_norm": 1.3576571941375732,
"learning_rate": 4.982682937826242e-05,
"loss": 0.9363,
"step": 1310
},
{
"epoch": 0.13494517852122576,
"grad_norm": 1.4128495454788208,
"learning_rate": 4.98161882455191e-05,
"loss": 0.9249,
"step": 1320
},
{
"epoch": 0.1359674904797199,
"grad_norm": 1.743247389793396,
"learning_rate": 4.9805231050286984e-05,
"loss": 0.917,
"step": 1330
},
{
"epoch": 0.13698980243821401,
"grad_norm": 1.455918312072754,
"learning_rate": 4.979395793211844e-05,
"loss": 0.9237,
"step": 1340
},
{
"epoch": 0.13801211439670816,
"grad_norm": 1.412348985671997,
"learning_rate": 4.978236903458951e-05,
"loss": 0.9131,
"step": 1350
},
{
"epoch": 0.1390344263552023,
"grad_norm": 1.2623333930969238,
"learning_rate": 4.977046450529802e-05,
"loss": 0.9017,
"step": 1360
},
{
"epoch": 0.14005673831369642,
"grad_norm": 1.3085025548934937,
"learning_rate": 4.975824449586176e-05,
"loss": 0.9052,
"step": 1370
},
{
"epoch": 0.14107905027219056,
"grad_norm": 1.464570164680481,
"learning_rate": 4.9745709161916486e-05,
"loss": 0.9122,
"step": 1380
},
{
"epoch": 0.1421013622306847,
"grad_norm": 1.2240124940872192,
"learning_rate": 4.9732858663114005e-05,
"loss": 0.8869,
"step": 1390
},
{
"epoch": 0.14312367418917882,
"grad_norm": 1.7124797105789185,
"learning_rate": 4.9719693163120065e-05,
"loss": 0.895,
"step": 1400
},
{
"epoch": 0.14414598614767296,
"grad_norm": 1.4251481294631958,
"learning_rate": 4.9706212829612354e-05,
"loss": 0.9005,
"step": 1410
},
{
"epoch": 0.1451682981061671,
"grad_norm": 1.4017133712768555,
"learning_rate": 4.969241783427832e-05,
"loss": 0.9035,
"step": 1420
},
{
"epoch": 0.14619061006466122,
"grad_norm": 1.3138145208358765,
"learning_rate": 4.967830835281295e-05,
"loss": 0.8931,
"step": 1430
},
{
"epoch": 0.14721292202315536,
"grad_norm": 1.390433669090271,
"learning_rate": 4.966388456491662e-05,
"loss": 0.8974,
"step": 1440
},
{
"epoch": 0.1482352339816495,
"grad_norm": 1.5036553144454956,
"learning_rate": 4.9649146654292726e-05,
"loss": 0.8968,
"step": 1450
},
{
"epoch": 0.14925754594014362,
"grad_norm": 1.073586344718933,
"learning_rate": 4.963409480864539e-05,
"loss": 0.8884,
"step": 1460
},
{
"epoch": 0.15027985789863776,
"grad_norm": 1.470304250717163,
"learning_rate": 4.961872921967705e-05,
"loss": 0.8909,
"step": 1470
},
{
"epoch": 0.1513021698571319,
"grad_norm": 1.2872247695922852,
"learning_rate": 4.960305008308603e-05,
"loss": 0.8943,
"step": 1480
},
{
"epoch": 0.15232448181562605,
"grad_norm": 1.4058434963226318,
"learning_rate": 4.958705759856402e-05,
"loss": 0.8738,
"step": 1490
},
{
"epoch": 0.15334679377412017,
"grad_norm": 1.1410505771636963,
"learning_rate": 4.957075196979357e-05,
"loss": 0.8694,
"step": 1500
},
{
"epoch": 0.1543691057326143,
"grad_norm": 2.069765567779541,
"learning_rate": 4.955413340444549e-05,
"loss": 0.8818,
"step": 1510
},
{
"epoch": 0.15539141769110845,
"grad_norm": 1.5385953187942505,
"learning_rate": 4.953720211417618e-05,
"loss": 0.8919,
"step": 1520
},
{
"epoch": 0.15641372964960257,
"grad_norm": 1.4129282236099243,
"learning_rate": 4.9519958314624946e-05,
"loss": 0.8995,
"step": 1530
},
{
"epoch": 0.1574360416080967,
"grad_norm": 1.2696932554244995,
"learning_rate": 4.950240222541126e-05,
"loss": 0.8765,
"step": 1540
},
{
"epoch": 0.15845835356659085,
"grad_norm": 1.1447160243988037,
"learning_rate": 4.948453407013196e-05,
"loss": 0.8499,
"step": 1550
},
{
"epoch": 0.15948066552508497,
"grad_norm": 1.3392435312271118,
"learning_rate": 4.946635407635841e-05,
"loss": 0.8548,
"step": 1560
},
{
"epoch": 0.1605029774835791,
"grad_norm": 1.4357683658599854,
"learning_rate": 4.944786247563359e-05,
"loss": 0.8672,
"step": 1570
},
{
"epoch": 0.16152528944207326,
"grad_norm": 1.3871515989303589,
"learning_rate": 4.942905950346913e-05,
"loss": 0.8623,
"step": 1580
},
{
"epoch": 0.16254760140056737,
"grad_norm": 1.2140942811965942,
"learning_rate": 4.940994539934236e-05,
"loss": 0.8593,
"step": 1590
},
{
"epoch": 0.16356991335906151,
"grad_norm": 1.4662145376205444,
"learning_rate": 4.939052040669321e-05,
"loss": 0.8537,
"step": 1600
},
{
"epoch": 0.16459222531755566,
"grad_norm": 1.2750685214996338,
"learning_rate": 4.9370784772921144e-05,
"loss": 0.86,
"step": 1610
},
{
"epoch": 0.1656145372760498,
"grad_norm": 1.418298363685608,
"learning_rate": 4.935073874938199e-05,
"loss": 0.8699,
"step": 1620
},
{
"epoch": 0.16663684923454392,
"grad_norm": 1.4741421937942505,
"learning_rate": 4.933038259138474e-05,
"loss": 0.8693,
"step": 1630
},
{
"epoch": 0.16765916119303806,
"grad_norm": 1.1894652843475342,
"learning_rate": 4.93097165581883e-05,
"loss": 0.8659,
"step": 1640
},
{
"epoch": 0.1686814731515322,
"grad_norm": 1.4552342891693115,
"learning_rate": 4.9288740912998204e-05,
"loss": 0.8549,
"step": 1650
},
{
"epoch": 0.16970378511002632,
"grad_norm": 1.3728736639022827,
"learning_rate": 4.926745592296324e-05,
"loss": 0.8619,
"step": 1660
},
{
"epoch": 0.17072609706852046,
"grad_norm": 1.3134335279464722,
"learning_rate": 4.924586185917206e-05,
"loss": 0.8707,
"step": 1670
},
{
"epoch": 0.1717484090270146,
"grad_norm": 1.1918591260910034,
"learning_rate": 4.922395899664971e-05,
"loss": 0.8548,
"step": 1680
},
{
"epoch": 0.17277072098550872,
"grad_norm": 1.2133831977844238,
"learning_rate": 4.920174761435416e-05,
"loss": 0.8434,
"step": 1690
},
{
"epoch": 0.17379303294400286,
"grad_norm": 1.286832332611084,
"learning_rate": 4.917922799517271e-05,
"loss": 0.8513,
"step": 1700
},
{
"epoch": 0.174815344902497,
"grad_norm": 1.1141687631607056,
"learning_rate": 4.915640042591842e-05,
"loss": 0.8633,
"step": 1710
},
{
"epoch": 0.17583765686099112,
"grad_norm": 1.6487301588058472,
"learning_rate": 4.913326519732645e-05,
"loss": 0.8498,
"step": 1720
},
{
"epoch": 0.17685996881948526,
"grad_norm": 1.1880557537078857,
"learning_rate": 4.910982260405036e-05,
"loss": 0.8496,
"step": 1730
},
{
"epoch": 0.1778822807779794,
"grad_norm": 1.6126662492752075,
"learning_rate": 4.908607294465832e-05,
"loss": 0.8552,
"step": 1740
},
{
"epoch": 0.17890459273647352,
"grad_norm": 1.3989888429641724,
"learning_rate": 4.906201652162937e-05,
"loss": 0.8437,
"step": 1750
},
{
"epoch": 0.17992690469496767,
"grad_norm": 1.1807786226272583,
"learning_rate": 4.903765364134953e-05,
"loss": 0.8414,
"step": 1760
},
{
"epoch": 0.1809492166534618,
"grad_norm": 1.6690312623977661,
"learning_rate": 4.9012984614107855e-05,
"loss": 0.8326,
"step": 1770
},
{
"epoch": 0.18197152861195595,
"grad_norm": 1.2505375146865845,
"learning_rate": 4.89880097540926e-05,
"loss": 0.8274,
"step": 1780
},
{
"epoch": 0.18299384057045007,
"grad_norm": 1.6897757053375244,
"learning_rate": 4.896272937938713e-05,
"loss": 0.8408,
"step": 1790
},
{
"epoch": 0.1840161525289442,
"grad_norm": 1.1050786972045898,
"learning_rate": 4.8937143811965845e-05,
"loss": 0.8256,
"step": 1800
},
{
"epoch": 0.18503846448743835,
"grad_norm": 1.2909902334213257,
"learning_rate": 4.8911253377690194e-05,
"loss": 0.8373,
"step": 1810
},
{
"epoch": 0.18606077644593247,
"grad_norm": 1.3720498085021973,
"learning_rate": 4.888505840630439e-05,
"loss": 0.838,
"step": 1820
},
{
"epoch": 0.1870830884044266,
"grad_norm": 1.2462021112442017,
"learning_rate": 4.8858559231431297e-05,
"loss": 0.8294,
"step": 1830
},
{
"epoch": 0.18810540036292075,
"grad_norm": 1.2244844436645508,
"learning_rate": 4.883175619056817e-05,
"loss": 0.8239,
"step": 1840
},
{
"epoch": 0.18912771232141487,
"grad_norm": 1.477510929107666,
"learning_rate": 4.880464962508234e-05,
"loss": 0.8337,
"step": 1850
},
{
"epoch": 0.190150024279909,
"grad_norm": 1.4423950910568237,
"learning_rate": 4.877723988020685e-05,
"loss": 0.8073,
"step": 1860
},
{
"epoch": 0.19117233623840316,
"grad_norm": 1.3489868640899658,
"learning_rate": 4.8749527305036106e-05,
"loss": 0.8298,
"step": 1870
},
{
"epoch": 0.19219464819689727,
"grad_norm": 1.4012629985809326,
"learning_rate": 4.872151225252138e-05,
"loss": 0.8239,
"step": 1880
},
{
"epoch": 0.19321696015539142,
"grad_norm": 1.12566339969635,
"learning_rate": 4.8693195079466364e-05,
"loss": 0.8259,
"step": 1890
},
{
"epoch": 0.19423927211388556,
"grad_norm": 1.1713924407958984,
"learning_rate": 4.866457614652257e-05,
"loss": 0.8143,
"step": 1900
},
{
"epoch": 0.19526158407237967,
"grad_norm": 1.1162623167037964,
"learning_rate": 4.863565581818479e-05,
"loss": 0.8107,
"step": 1910
},
{
"epoch": 0.19628389603087382,
"grad_norm": 1.4159964323043823,
"learning_rate": 4.860643446278642e-05,
"loss": 0.8147,
"step": 1920
},
{
"epoch": 0.19730620798936796,
"grad_norm": 1.4067683219909668,
"learning_rate": 4.8576912452494784e-05,
"loss": 0.8233,
"step": 1930
},
{
"epoch": 0.1983285199478621,
"grad_norm": 1.2760193347930908,
"learning_rate": 4.854709016330639e-05,
"loss": 0.822,
"step": 1940
},
{
"epoch": 0.19935083190635622,
"grad_norm": 1.425916314125061,
"learning_rate": 4.851696797504212e-05,
"loss": 0.8115,
"step": 1950
},
{
"epoch": 0.20037314386485036,
"grad_norm": 1.293108344078064,
"learning_rate": 4.8486546271342454e-05,
"loss": 0.7992,
"step": 1960
},
{
"epoch": 0.2013954558233445,
"grad_norm": 1.1631488800048828,
"learning_rate": 4.845582543966251e-05,
"loss": 0.8257,
"step": 1970
},
{
"epoch": 0.20241776778183862,
"grad_norm": 1.280648112297058,
"learning_rate": 4.842480587126714e-05,
"loss": 0.8047,
"step": 1980
},
{
"epoch": 0.20344007974033276,
"grad_norm": 1.443266749382019,
"learning_rate": 4.839348796122598e-05,
"loss": 0.801,
"step": 1990
},
{
"epoch": 0.2044623916988269,
"grad_norm": 1.3283780813217163,
"learning_rate": 4.8361872108408364e-05,
"loss": 0.7989,
"step": 2000
},
{
"epoch": 0.20548470365732102,
"grad_norm": 1.2213093042373657,
"learning_rate": 4.8329958715478266e-05,
"loss": 0.8003,
"step": 2010
},
{
"epoch": 0.20650701561581516,
"grad_norm": 1.119112253189087,
"learning_rate": 4.829774818888918e-05,
"loss": 0.8073,
"step": 2020
},
{
"epoch": 0.2075293275743093,
"grad_norm": 1.1009432077407837,
"learning_rate": 4.826524093887894e-05,
"loss": 0.8002,
"step": 2030
},
{
"epoch": 0.20855163953280342,
"grad_norm": 1.3652846813201904,
"learning_rate": 4.8232437379464476e-05,
"loss": 0.827,
"step": 2040
},
{
"epoch": 0.20957395149129757,
"grad_norm": 1.4191744327545166,
"learning_rate": 4.819933792843659e-05,
"loss": 0.8016,
"step": 2050
},
{
"epoch": 0.2105962634497917,
"grad_norm": 1.2219738960266113,
"learning_rate": 4.816594300735455e-05,
"loss": 0.8076,
"step": 2060
},
{
"epoch": 0.21161857540828583,
"grad_norm": 1.164734125137329,
"learning_rate": 4.813225304154083e-05,
"loss": 0.8076,
"step": 2070
},
{
"epoch": 0.21264088736677997,
"grad_norm": 1.136127233505249,
"learning_rate": 4.8098268460075594e-05,
"loss": 0.7971,
"step": 2080
},
{
"epoch": 0.2136631993252741,
"grad_norm": 1.114316701889038,
"learning_rate": 4.8063989695791267e-05,
"loss": 0.8014,
"step": 2090
},
{
"epoch": 0.21468551128376825,
"grad_norm": 1.2291415929794312,
"learning_rate": 4.8029417185267076e-05,
"loss": 0.8007,
"step": 2100
},
{
"epoch": 0.21570782324226237,
"grad_norm": 1.1971957683563232,
"learning_rate": 4.79945513688234e-05,
"loss": 0.792,
"step": 2110
},
{
"epoch": 0.2167301352007565,
"grad_norm": 1.093489170074463,
"learning_rate": 4.795939269051622e-05,
"loss": 0.796,
"step": 2120
},
{
"epoch": 0.21775244715925066,
"grad_norm": 0.9938094019889832,
"learning_rate": 4.792394159813144e-05,
"loss": 0.7966,
"step": 2130
},
{
"epoch": 0.21877475911774477,
"grad_norm": 1.2750223875045776,
"learning_rate": 4.7888198543179184e-05,
"loss": 0.8042,
"step": 2140
},
{
"epoch": 0.21979707107623891,
"grad_norm": 1.3699895143508911,
"learning_rate": 4.785216398088809e-05,
"loss": 0.782,
"step": 2150
},
{
"epoch": 0.22081938303473306,
"grad_norm": 1.235351800918579,
"learning_rate": 4.781583837019942e-05,
"loss": 0.7934,
"step": 2160
},
{
"epoch": 0.22184169499322717,
"grad_norm": 1.177571177482605,
"learning_rate": 4.7779222173761306e-05,
"loss": 0.7964,
"step": 2170
},
{
"epoch": 0.22286400695172132,
"grad_norm": 1.3253560066223145,
"learning_rate": 4.774231585792282e-05,
"loss": 0.7834,
"step": 2180
},
{
"epoch": 0.22388631891021546,
"grad_norm": 1.042148470878601,
"learning_rate": 4.770511989272803e-05,
"loss": 0.7816,
"step": 2190
},
{
"epoch": 0.22490863086870957,
"grad_norm": 1.1149499416351318,
"learning_rate": 4.7667634751910006e-05,
"loss": 0.7876,
"step": 2200
},
{
"epoch": 0.22593094282720372,
"grad_norm": 1.1252564191818237,
"learning_rate": 4.762986091288482e-05,
"loss": 0.775,
"step": 2210
},
{
"epoch": 0.22695325478569786,
"grad_norm": 1.5249801874160767,
"learning_rate": 4.759179885674543e-05,
"loss": 0.7872,
"step": 2220
},
{
"epoch": 0.22797556674419198,
"grad_norm": 1.159932255744934,
"learning_rate": 4.7553449068255574e-05,
"loss": 0.7732,
"step": 2230
},
{
"epoch": 0.22899787870268612,
"grad_norm": 1.224050521850586,
"learning_rate": 4.75148120358436e-05,
"loss": 0.7789,
"step": 2240
},
{
"epoch": 0.23002019066118026,
"grad_norm": 1.3017828464508057,
"learning_rate": 4.747588825159622e-05,
"loss": 0.7793,
"step": 2250
},
{
"epoch": 0.2310425026196744,
"grad_norm": 1.119139552116394,
"learning_rate": 4.7436678211252266e-05,
"loss": 0.7809,
"step": 2260
},
{
"epoch": 0.23206481457816852,
"grad_norm": 1.0169899463653564,
"learning_rate": 4.739718241419636e-05,
"loss": 0.7801,
"step": 2270
},
{
"epoch": 0.23308712653666266,
"grad_norm": 1.0275715589523315,
"learning_rate": 4.735740136345258e-05,
"loss": 0.7934,
"step": 2280
},
{
"epoch": 0.2341094384951568,
"grad_norm": 1.15293550491333,
"learning_rate": 4.731733556567801e-05,
"loss": 0.7804,
"step": 2290
},
{
"epoch": 0.23513175045365092,
"grad_norm": 1.0598777532577515,
"learning_rate": 4.727698553115634e-05,
"loss": 0.7681,
"step": 2300
},
{
"epoch": 0.23615406241214507,
"grad_norm": 1.2926679849624634,
"learning_rate": 4.723635177379131e-05,
"loss": 0.7878,
"step": 2310
},
{
"epoch": 0.2371763743706392,
"grad_norm": 1.1461938619613647,
"learning_rate": 4.719543481110022e-05,
"loss": 0.7771,
"step": 2320
},
{
"epoch": 0.23819868632913332,
"grad_norm": 1.1584590673446655,
"learning_rate": 4.7154235164207304e-05,
"loss": 0.7764,
"step": 2330
},
{
"epoch": 0.23922099828762747,
"grad_norm": 1.1812832355499268,
"learning_rate": 4.711275335783709e-05,
"loss": 0.7701,
"step": 2340
},
{
"epoch": 0.2402433102461216,
"grad_norm": 1.5752184391021729,
"learning_rate": 4.707098992030775e-05,
"loss": 0.7771,
"step": 2350
},
{
"epoch": 0.24126562220461573,
"grad_norm": 1.5075185298919678,
"learning_rate": 4.702894538352434e-05,
"loss": 0.7769,
"step": 2360
},
{
"epoch": 0.24228793416310987,
"grad_norm": 1.2277531623840332,
"learning_rate": 4.6986620282972034e-05,
"loss": 0.7773,
"step": 2370
},
{
"epoch": 0.243310246121604,
"grad_norm": 1.1297932863235474,
"learning_rate": 4.6944015157709306e-05,
"loss": 0.7773,
"step": 2380
},
{
"epoch": 0.24433255808009816,
"grad_norm": 1.19589102268219,
"learning_rate": 4.690113055036106e-05,
"loss": 0.7802,
"step": 2390
},
{
"epoch": 0.24535487003859227,
"grad_norm": 1.2334651947021484,
"learning_rate": 4.685796700711175e-05,
"loss": 0.7703,
"step": 2400
},
{
"epoch": 0.24637718199708641,
"grad_norm": 1.0675991773605347,
"learning_rate": 4.6814525077698365e-05,
"loss": 0.7563,
"step": 2410
},
{
"epoch": 0.24739949395558056,
"grad_norm": 1.1307783126831055,
"learning_rate": 4.677080531540347e-05,
"loss": 0.7686,
"step": 2420
},
{
"epoch": 0.24842180591407467,
"grad_norm": 1.135969877243042,
"learning_rate": 4.672680827704815e-05,
"loss": 0.76,
"step": 2430
},
{
"epoch": 0.24944411787256882,
"grad_norm": 1.0702753067016602,
"learning_rate": 4.668253452298492e-05,
"loss": 0.7692,
"step": 2440
},
{
"epoch": 0.25046642983106293,
"grad_norm": 1.3017902374267578,
"learning_rate": 4.663798461709059e-05,
"loss": 0.7642,
"step": 2450
},
{
"epoch": 0.2514887417895571,
"grad_norm": 1.205757737159729,
"learning_rate": 4.6593159126759064e-05,
"loss": 0.7676,
"step": 2460
},
{
"epoch": 0.2525110537480512,
"grad_norm": 1.0397226810455322,
"learning_rate": 4.654805862289414e-05,
"loss": 0.7643,
"step": 2470
},
{
"epoch": 0.25353336570654533,
"grad_norm": 1.175759196281433,
"learning_rate": 4.650268367990222e-05,
"loss": 0.7607,
"step": 2480
},
{
"epoch": 0.2545556776650395,
"grad_norm": 1.1430120468139648,
"learning_rate": 4.6457034875685e-05,
"loss": 0.7689,
"step": 2490
},
{
"epoch": 0.2555779896235336,
"grad_norm": 1.0399481058120728,
"learning_rate": 4.641111279163214e-05,
"loss": 0.7667,
"step": 2500
},
{
"epoch": 0.25660030158202773,
"grad_norm": 1.0172754526138306,
"learning_rate": 4.636491801261379e-05,
"loss": 0.7643,
"step": 2510
},
{
"epoch": 0.2576226135405219,
"grad_norm": 1.1760464906692505,
"learning_rate": 4.631845112697319e-05,
"loss": 0.7686,
"step": 2520
},
{
"epoch": 0.258644925499016,
"grad_norm": 1.2932900190353394,
"learning_rate": 4.62717127265192e-05,
"loss": 0.7645,
"step": 2530
},
{
"epoch": 0.25966723745751014,
"grad_norm": 1.272044062614441,
"learning_rate": 4.62247034065187e-05,
"loss": 0.7635,
"step": 2540
},
{
"epoch": 0.2606895494160043,
"grad_norm": 1.2114735841751099,
"learning_rate": 4.6177423765689045e-05,
"loss": 0.7671,
"step": 2550
},
{
"epoch": 0.2617118613744984,
"grad_norm": 1.3058948516845703,
"learning_rate": 4.612987440619044e-05,
"loss": 0.7633,
"step": 2560
},
{
"epoch": 0.26273417333299254,
"grad_norm": 1.1863833665847778,
"learning_rate": 4.6082055933618255e-05,
"loss": 0.7559,
"step": 2570
},
{
"epoch": 0.2637564852914867,
"grad_norm": 1.4104381799697876,
"learning_rate": 4.6033968956995346e-05,
"loss": 0.7486,
"step": 2580
},
{
"epoch": 0.2647787972499808,
"grad_norm": 1.264552116394043,
"learning_rate": 4.598561408876425e-05,
"loss": 0.7611,
"step": 2590
},
{
"epoch": 0.26580110920847494,
"grad_norm": 1.5521522760391235,
"learning_rate": 4.593699194477943e-05,
"loss": 0.7591,
"step": 2600
},
{
"epoch": 0.2668234211669691,
"grad_norm": 1.22602117061615,
"learning_rate": 4.5888103144299374e-05,
"loss": 0.7498,
"step": 2610
},
{
"epoch": 0.2678457331254632,
"grad_norm": 1.0520557165145874,
"learning_rate": 4.583894830997879e-05,
"loss": 0.7443,
"step": 2620
},
{
"epoch": 0.2688680450839574,
"grad_norm": 1.0993170738220215,
"learning_rate": 4.578952806786061e-05,
"loss": 0.7413,
"step": 2630
},
{
"epoch": 0.2698903570424515,
"grad_norm": 1.1415059566497803,
"learning_rate": 4.573984304736804e-05,
"loss": 0.7472,
"step": 2640
},
{
"epoch": 0.2709126690009456,
"grad_norm": 1.0268982648849487,
"learning_rate": 4.568989388129652e-05,
"loss": 0.7555,
"step": 2650
},
{
"epoch": 0.2719349809594398,
"grad_norm": 1.0965386629104614,
"learning_rate": 4.563968120580573e-05,
"loss": 0.7561,
"step": 2660
},
{
"epoch": 0.2729572929179339,
"grad_norm": 1.1724823713302612,
"learning_rate": 4.5589205660411397e-05,
"loss": 0.7537,
"step": 2670
},
{
"epoch": 0.27397960487642803,
"grad_norm": 1.2942609786987305,
"learning_rate": 4.553846788797723e-05,
"loss": 0.7575,
"step": 2680
},
{
"epoch": 0.2750019168349222,
"grad_norm": 1.353010892868042,
"learning_rate": 4.5487468534706685e-05,
"loss": 0.7558,
"step": 2690
},
{
"epoch": 0.2760242287934163,
"grad_norm": 1.2842090129852295,
"learning_rate": 4.5436208250134744e-05,
"loss": 0.7441,
"step": 2700
},
{
"epoch": 0.27704654075191043,
"grad_norm": 1.0256245136260986,
"learning_rate": 4.5384687687119656e-05,
"loss": 0.743,
"step": 2710
},
{
"epoch": 0.2780688527104046,
"grad_norm": 1.2406985759735107,
"learning_rate": 4.533290750183461e-05,
"loss": 0.7465,
"step": 2720
},
{
"epoch": 0.2790911646688987,
"grad_norm": 1.257161259651184,
"learning_rate": 4.5280868353759375e-05,
"loss": 0.7505,
"step": 2730
},
{
"epoch": 0.28011347662739283,
"grad_norm": 1.0539823770523071,
"learning_rate": 4.522857090567193e-05,
"loss": 0.7389,
"step": 2740
},
{
"epoch": 0.281135788585887,
"grad_norm": 1.274788737297058,
"learning_rate": 4.517601582363995e-05,
"loss": 0.7501,
"step": 2750
},
{
"epoch": 0.2821581005443811,
"grad_norm": 1.427249550819397,
"learning_rate": 4.512320377701244e-05,
"loss": 0.75,
"step": 2760
},
{
"epoch": 0.28318041250287523,
"grad_norm": 1.2584348917007446,
"learning_rate": 4.507013543841109e-05,
"loss": 0.7401,
"step": 2770
},
{
"epoch": 0.2842027244613694,
"grad_norm": 1.1880669593811035,
"learning_rate": 4.501681148372177e-05,
"loss": 0.7354,
"step": 2780
},
{
"epoch": 0.2852250364198635,
"grad_norm": 1.1196924448013306,
"learning_rate": 4.496323259208592e-05,
"loss": 0.7539,
"step": 2790
},
{
"epoch": 0.28624734837835764,
"grad_norm": 1.0014890432357788,
"learning_rate": 4.490939944589192e-05,
"loss": 0.7515,
"step": 2800
},
{
"epoch": 0.2872696603368518,
"grad_norm": 1.1331443786621094,
"learning_rate": 4.485531273076632e-05,
"loss": 0.7414,
"step": 2810
},
{
"epoch": 0.2882919722953459,
"grad_norm": 1.1644209623336792,
"learning_rate": 4.48009731355652e-05,
"loss": 0.7409,
"step": 2820
},
{
"epoch": 0.28931428425384004,
"grad_norm": 1.2161707878112793,
"learning_rate": 4.474638135236533e-05,
"loss": 0.7489,
"step": 2830
},
{
"epoch": 0.2903365962123342,
"grad_norm": 1.3701202869415283,
"learning_rate": 4.46915380764554e-05,
"loss": 0.7349,
"step": 2840
},
{
"epoch": 0.2913589081708283,
"grad_norm": 1.1645112037658691,
"learning_rate": 4.463644400632714e-05,
"loss": 0.7364,
"step": 2850
},
{
"epoch": 0.29238122012932244,
"grad_norm": 1.0601342916488647,
"learning_rate": 4.458109984366643e-05,
"loss": 0.7512,
"step": 2860
},
{
"epoch": 0.2934035320878166,
"grad_norm": 1.3086082935333252,
"learning_rate": 4.4525506293344364e-05,
"loss": 0.734,
"step": 2870
},
{
"epoch": 0.2944258440463107,
"grad_norm": 1.188309669494629,
"learning_rate": 4.4469664063408284e-05,
"loss": 0.7396,
"step": 2880
},
{
"epoch": 0.29544815600480484,
"grad_norm": 1.132607340812683,
"learning_rate": 4.4413573865072734e-05,
"loss": 0.7414,
"step": 2890
},
{
"epoch": 0.296470467963299,
"grad_norm": 1.1542457342147827,
"learning_rate": 4.4357236412710415e-05,
"loss": 0.7395,
"step": 2900
},
{
"epoch": 0.2974927799217931,
"grad_norm": 1.0407392978668213,
"learning_rate": 4.430065242384311e-05,
"loss": 0.741,
"step": 2910
},
{
"epoch": 0.29851509188028724,
"grad_norm": 1.2613579034805298,
"learning_rate": 4.424382261913254e-05,
"loss": 0.7292,
"step": 2920
},
{
"epoch": 0.2995374038387814,
"grad_norm": 1.2430132627487183,
"learning_rate": 4.41867477223711e-05,
"loss": 0.7329,
"step": 2930
},
{
"epoch": 0.30055971579727553,
"grad_norm": 1.1074517965316772,
"learning_rate": 4.41294284604728e-05,
"loss": 0.7315,
"step": 2940
},
{
"epoch": 0.3015820277557697,
"grad_norm": 1.1700302362442017,
"learning_rate": 4.407186556346385e-05,
"loss": 0.7457,
"step": 2950
},
{
"epoch": 0.3026043397142638,
"grad_norm": 1.1400681734085083,
"learning_rate": 4.401405976447349e-05,
"loss": 0.7351,
"step": 2960
},
{
"epoch": 0.30362665167275793,
"grad_norm": 0.9993115067481995,
"learning_rate": 4.395601179972454e-05,
"loss": 0.7275,
"step": 2970
},
{
"epoch": 0.3046489636312521,
"grad_norm": 1.220421552658081,
"learning_rate": 4.389772240852412e-05,
"loss": 0.7252,
"step": 2980
},
{
"epoch": 0.3056712755897462,
"grad_norm": 0.7238106727600098,
"learning_rate": 4.383919233325415e-05,
"loss": 0.7234,
"step": 2990
},
{
"epoch": 0.30669358754824033,
"grad_norm": 0.7465090751647949,
"learning_rate": 4.3780422319361994e-05,
"loss": 0.7155,
"step": 3000
},
{
"epoch": 0.3077158995067345,
"grad_norm": 0.8709385991096497,
"learning_rate": 4.372141311535083e-05,
"loss": 0.6982,
"step": 3010
},
{
"epoch": 0.3087382114652286,
"grad_norm": 0.8341991305351257,
"learning_rate": 4.3662165472770275e-05,
"loss": 0.7027,
"step": 3020
},
{
"epoch": 0.30976052342372273,
"grad_norm": 0.8701478242874146,
"learning_rate": 4.360268014620669e-05,
"loss": 0.7012,
"step": 3030
},
{
"epoch": 0.3107828353822169,
"grad_norm": 0.9397259950637817,
"learning_rate": 4.35429578932736e-05,
"loss": 0.6926,
"step": 3040
},
{
"epoch": 0.311805147340711,
"grad_norm": 0.9259440898895264,
"learning_rate": 4.3482999474602116e-05,
"loss": 0.6965,
"step": 3050
},
{
"epoch": 0.31282745929920514,
"grad_norm": 0.8428536057472229,
"learning_rate": 4.342280565383111e-05,
"loss": 0.6902,
"step": 3060
},
{
"epoch": 0.3138497712576993,
"grad_norm": 0.8366013765335083,
"learning_rate": 4.336237719759764e-05,
"loss": 0.6972,
"step": 3070
},
{
"epoch": 0.3148720832161934,
"grad_norm": 0.9582898020744324,
"learning_rate": 4.330171487552708e-05,
"loss": 0.6864,
"step": 3080
},
{
"epoch": 0.31589439517468754,
"grad_norm": 0.9501172304153442,
"learning_rate": 4.324081946022334e-05,
"loss": 0.6806,
"step": 3090
},
{
"epoch": 0.3169167071331817,
"grad_norm": 0.7470399141311646,
"learning_rate": 4.317969172725907e-05,
"loss": 0.6837,
"step": 3100
},
{
"epoch": 0.3179390190916758,
"grad_norm": 0.9217721819877625,
"learning_rate": 4.3118332455165735e-05,
"loss": 0.6872,
"step": 3110
},
{
"epoch": 0.31896133105016994,
"grad_norm": 0.7747515439987183,
"learning_rate": 4.3056742425423706e-05,
"loss": 0.6818,
"step": 3120
},
{
"epoch": 0.3199836430086641,
"grad_norm": 1.169914960861206,
"learning_rate": 4.299492242245235e-05,
"loss": 0.6856,
"step": 3130
},
{
"epoch": 0.3210059549671582,
"grad_norm": 0.9959880709648132,
"learning_rate": 4.293287323359995e-05,
"loss": 0.682,
"step": 3140
},
{
"epoch": 0.32202826692565234,
"grad_norm": 0.7298420667648315,
"learning_rate": 4.2870595649133784e-05,
"loss": 0.6852,
"step": 3150
},
{
"epoch": 0.3230505788841465,
"grad_norm": 0.8755456209182739,
"learning_rate": 4.2808090462229986e-05,
"loss": 0.6815,
"step": 3160
},
{
"epoch": 0.3240728908426406,
"grad_norm": 0.8738940358161926,
"learning_rate": 4.274535846896348e-05,
"loss": 0.6792,
"step": 3170
},
{
"epoch": 0.32509520280113474,
"grad_norm": 0.7759591341018677,
"learning_rate": 4.268240046829781e-05,
"loss": 0.6819,
"step": 3180
},
{
"epoch": 0.3261175147596289,
"grad_norm": 0.958259105682373,
"learning_rate": 4.2619217262074975e-05,
"loss": 0.685,
"step": 3190
},
{
"epoch": 0.32713982671812303,
"grad_norm": 0.9467047452926636,
"learning_rate": 4.255580965500526e-05,
"loss": 0.6768,
"step": 3200
},
{
"epoch": 0.32816213867661714,
"grad_norm": 0.7194684743881226,
"learning_rate": 4.249217845465694e-05,
"loss": 0.6712,
"step": 3210
},
{
"epoch": 0.3291844506351113,
"grad_norm": 0.769015908241272,
"learning_rate": 4.242832447144597e-05,
"loss": 0.6745,
"step": 3220
},
{
"epoch": 0.33020676259360543,
"grad_norm": 0.8387378454208374,
"learning_rate": 4.236424851862574e-05,
"loss": 0.6831,
"step": 3230
},
{
"epoch": 0.3312290745520996,
"grad_norm": 0.9365067481994629,
"learning_rate": 4.229995141227665e-05,
"loss": 0.6787,
"step": 3240
},
{
"epoch": 0.3322513865105937,
"grad_norm": 0.7298687696456909,
"learning_rate": 4.223543397129577e-05,
"loss": 0.6796,
"step": 3250
},
{
"epoch": 0.33327369846908783,
"grad_norm": 0.7920469641685486,
"learning_rate": 4.2170697017386365e-05,
"loss": 0.6755,
"step": 3260
},
{
"epoch": 0.334296010427582,
"grad_norm": 3.6704764366149902,
"learning_rate": 4.210574137504744e-05,
"loss": 0.8193,
"step": 3270
},
{
"epoch": 0.3353183223860761,
"grad_norm": 3.374370574951172,
"learning_rate": 4.204056787156326e-05,
"loss": 0.8628,
"step": 3280
},
{
"epoch": 0.33634063434457023,
"grad_norm": 3.6898951530456543,
"learning_rate": 4.197517733699281e-05,
"loss": 0.8978,
"step": 3290
},
{
"epoch": 0.3373629463030644,
"grad_norm": 1.8514819145202637,
"learning_rate": 4.1909570604159186e-05,
"loss": 0.9052,
"step": 3300
},
{
"epoch": 0.3383852582615585,
"grad_norm": 2.999357223510742,
"learning_rate": 4.1843748508639035e-05,
"loss": 0.9197,
"step": 3310
},
{
"epoch": 0.33940757022005263,
"grad_norm": 3.2991950511932373,
"learning_rate": 4.17777118887519e-05,
"loss": 1.0064,
"step": 3320
},
{
"epoch": 0.3404298821785468,
"grad_norm": 1.9216408729553223,
"learning_rate": 4.171146158554951e-05,
"loss": 0.935,
"step": 3330
},
{
"epoch": 0.3414521941370409,
"grad_norm": 2.73488450050354,
"learning_rate": 4.1644998442805114e-05,
"loss": 0.9026,
"step": 3340
},
{
"epoch": 0.34247450609553504,
"grad_norm": 2.8734359741210938,
"learning_rate": 4.157832330700272e-05,
"loss": 0.8654,
"step": 3350
},
{
"epoch": 0.3434968180540292,
"grad_norm": 3.519892930984497,
"learning_rate": 4.1511437027326304e-05,
"loss": 0.9811,
"step": 3360
},
{
"epoch": 0.3445191300125233,
"grad_norm": 2.985055446624756,
"learning_rate": 4.1444340455648995e-05,
"loss": 0.9688,
"step": 3370
},
{
"epoch": 0.34554144197101744,
"grad_norm": 3.2480971813201904,
"learning_rate": 4.137703444652223e-05,
"loss": 0.9601,
"step": 3380
},
{
"epoch": 0.3465637539295116,
"grad_norm": 2.305152654647827,
"learning_rate": 4.130951985716488e-05,
"loss": 1.0296,
"step": 3390
},
{
"epoch": 0.3475860658880057,
"grad_norm": 2.1093311309814453,
"learning_rate": 4.1241797547452316e-05,
"loss": 0.9131,
"step": 3400
},
{
"epoch": 0.34860837784649984,
"grad_norm": 1.8336457014083862,
"learning_rate": 4.117386837990547e-05,
"loss": 0.9109,
"step": 3410
},
{
"epoch": 0.349630689804994,
"grad_norm": 1.8202961683273315,
"learning_rate": 4.1105733219679844e-05,
"loss": 0.9349,
"step": 3420
},
{
"epoch": 0.3506530017634881,
"grad_norm": 2.424990653991699,
"learning_rate": 4.103739293455449e-05,
"loss": 0.9001,
"step": 3430
},
{
"epoch": 0.35167531372198224,
"grad_norm": 2.684093475341797,
"learning_rate": 4.096884839492095e-05,
"loss": 0.9077,
"step": 3440
},
{
"epoch": 0.3526976256804764,
"grad_norm": 1.670059323310852,
"learning_rate": 4.0900100473772204e-05,
"loss": 0.8612,
"step": 3450
},
{
"epoch": 0.3537199376389705,
"grad_norm": 2.0392446517944336,
"learning_rate": 4.083115004669151e-05,
"loss": 0.8622,
"step": 3460
},
{
"epoch": 0.35474224959746464,
"grad_norm": 2.8175671100616455,
"learning_rate": 4.076199799184127e-05,
"loss": 0.8206,
"step": 3470
},
{
"epoch": 0.3557645615559588,
"grad_norm": 1.9085370302200317,
"learning_rate": 4.0692645189951855e-05,
"loss": 0.9132,
"step": 3480
},
{
"epoch": 0.35678687351445293,
"grad_norm": 1.6227134466171265,
"learning_rate": 4.062309252431037e-05,
"loss": 0.9233,
"step": 3490
},
{
"epoch": 0.35780918547294704,
"grad_norm": 1.934488296508789,
"learning_rate": 4.0553340880749425e-05,
"loss": 0.8194,
"step": 3500
},
{
"epoch": 0.3588314974314412,
"grad_norm": 1.8496440649032593,
"learning_rate": 4.048339114763582e-05,
"loss": 0.8394,
"step": 3510
},
{
"epoch": 0.35985380938993533,
"grad_norm": 1.8710787296295166,
"learning_rate": 4.041324421585928e-05,
"loss": 0.782,
"step": 3520
},
{
"epoch": 0.36087612134842945,
"grad_norm": 1.7882663011550903,
"learning_rate": 4.034290097882105e-05,
"loss": 0.7677,
"step": 3530
},
{
"epoch": 0.3618984333069236,
"grad_norm": 1.3695753812789917,
"learning_rate": 4.027236233242257e-05,
"loss": 0.7844,
"step": 3540
},
{
"epoch": 0.36292074526541773,
"grad_norm": 1.6935341358184814,
"learning_rate": 4.020162917505403e-05,
"loss": 0.8085,
"step": 3550
},
{
"epoch": 0.3639430572239119,
"grad_norm": 2.658515691757202,
"learning_rate": 4.013070240758294e-05,
"loss": 0.77,
"step": 3560
},
{
"epoch": 0.364965369182406,
"grad_norm": 1.2438311576843262,
"learning_rate": 4.0059582933342666e-05,
"loss": 0.8105,
"step": 3570
},
{
"epoch": 0.36598768114090013,
"grad_norm": 1.7631160020828247,
"learning_rate": 3.99882716581209e-05,
"loss": 0.7977,
"step": 3580
},
{
"epoch": 0.3670099930993943,
"grad_norm": 2.5809144973754883,
"learning_rate": 3.991676949014814e-05,
"loss": 0.7906,
"step": 3590
},
{
"epoch": 0.3680323050578884,
"grad_norm": 1.1488144397735596,
"learning_rate": 3.984507734008614e-05,
"loss": 0.7578,
"step": 3600
},
{
"epoch": 0.36905461701638254,
"grad_norm": 1.8908311128616333,
"learning_rate": 3.9773196121016264e-05,
"loss": 0.803,
"step": 3610
},
{
"epoch": 0.3700769289748767,
"grad_norm": 1.5343143939971924,
"learning_rate": 3.970112674842791e-05,
"loss": 0.7424,
"step": 3620
},
{
"epoch": 0.3710992409333708,
"grad_norm": 1.9406263828277588,
"learning_rate": 3.962887014020681e-05,
"loss": 0.7438,
"step": 3630
},
{
"epoch": 0.37212155289186494,
"grad_norm": 2.6998703479766846,
"learning_rate": 3.955642721662336e-05,
"loss": 0.7917,
"step": 3640
},
{
"epoch": 0.3731438648503591,
"grad_norm": 1.2992616891860962,
"learning_rate": 3.94837989003209e-05,
"loss": 0.7571,
"step": 3650
},
{
"epoch": 0.3741661768088532,
"grad_norm": 1.7680587768554688,
"learning_rate": 3.941098611630396e-05,
"loss": 0.7683,
"step": 3660
},
{
"epoch": 0.37518848876734734,
"grad_norm": 1.778095006942749,
"learning_rate": 3.9337989791926466e-05,
"loss": 0.7227,
"step": 3670
},
{
"epoch": 0.3762108007258415,
"grad_norm": 1.4913495779037476,
"learning_rate": 3.926481085687994e-05,
"loss": 0.7608,
"step": 3680
},
{
"epoch": 0.3772331126843356,
"grad_norm": 1.3343455791473389,
"learning_rate": 3.919145024318168e-05,
"loss": 0.695,
"step": 3690
},
{
"epoch": 0.37825542464282974,
"grad_norm": 1.4345059394836426,
"learning_rate": 3.9117908885162836e-05,
"loss": 0.7274,
"step": 3700
},
{
"epoch": 0.3792777366013239,
"grad_norm": 2.5988686084747314,
"learning_rate": 3.904418771945658e-05,
"loss": 0.7259,
"step": 3710
},
{
"epoch": 0.380300048559818,
"grad_norm": 1.7964763641357422,
"learning_rate": 3.8970287684986105e-05,
"loss": 0.7152,
"step": 3720
},
{
"epoch": 0.38132236051831214,
"grad_norm": 1.4514089822769165,
"learning_rate": 3.8896209722952724e-05,
"loss": 0.7628,
"step": 3730
},
{
"epoch": 0.3823446724768063,
"grad_norm": 1.3875867128372192,
"learning_rate": 3.882195477682386e-05,
"loss": 0.7508,
"step": 3740
},
{
"epoch": 0.38336698443530043,
"grad_norm": 1.5468597412109375,
"learning_rate": 3.874752379232102e-05,
"loss": 0.7239,
"step": 3750
},
{
"epoch": 0.38438929639379454,
"grad_norm": 1.5340187549591064,
"learning_rate": 3.867291771740776e-05,
"loss": 0.7371,
"step": 3760
},
{
"epoch": 0.3854116083522887,
"grad_norm": 1.4312421083450317,
"learning_rate": 3.8598137502277636e-05,
"loss": 0.732,
"step": 3770
},
{
"epoch": 0.38643392031078283,
"grad_norm": 1.4971996545791626,
"learning_rate": 3.8523184099342044e-05,
"loss": 0.7107,
"step": 3780
},
{
"epoch": 0.38745623226927695,
"grad_norm": 1.1878859996795654,
"learning_rate": 3.844805846321815e-05,
"loss": 0.7076,
"step": 3790
},
{
"epoch": 0.3884785442277711,
"grad_norm": 1.388309359550476,
"learning_rate": 3.83727615507167e-05,
"loss": 0.7157,
"step": 3800
},
{
"epoch": 0.38950085618626523,
"grad_norm": 1.4581124782562256,
"learning_rate": 3.829729432082982e-05,
"loss": 0.7167,
"step": 3810
},
{
"epoch": 0.39052316814475935,
"grad_norm": 1.7948312759399414,
"learning_rate": 3.8221657734718855e-05,
"loss": 0.7399,
"step": 3820
},
{
"epoch": 0.3915454801032535,
"grad_norm": 1.5124136209487915,
"learning_rate": 3.814585275570209e-05,
"loss": 0.7077,
"step": 3830
},
{
"epoch": 0.39256779206174763,
"grad_norm": 1.7335097789764404,
"learning_rate": 3.8069880349242457e-05,
"loss": 0.7028,
"step": 3840
},
{
"epoch": 0.3935901040202418,
"grad_norm": 1.4563673734664917,
"learning_rate": 3.7993741482935293e-05,
"loss": 0.7084,
"step": 3850
},
{
"epoch": 0.3946124159787359,
"grad_norm": 1.8977148532867432,
"learning_rate": 3.791743712649598e-05,
"loss": 0.704,
"step": 3860
},
{
"epoch": 0.39563472793723004,
"grad_norm": 1.6411945819854736,
"learning_rate": 3.784096825174762e-05,
"loss": 0.6932,
"step": 3870
},
{
"epoch": 0.3966570398957242,
"grad_norm": 1.633746862411499,
"learning_rate": 3.776433583260863e-05,
"loss": 0.6952,
"step": 3880
},
{
"epoch": 0.3976793518542183,
"grad_norm": 1.7816822528839111,
"learning_rate": 3.768754084508036e-05,
"loss": 0.6886,
"step": 3890
},
{
"epoch": 0.39870166381271244,
"grad_norm": 1.3513058423995972,
"learning_rate": 3.761058426723463e-05,
"loss": 0.6923,
"step": 3900
},
{
"epoch": 0.3997239757712066,
"grad_norm": 1.576625108718872,
"learning_rate": 3.753346707920134e-05,
"loss": 0.6944,
"step": 3910
},
{
"epoch": 0.4007462877297007,
"grad_norm": 1.4564894437789917,
"learning_rate": 3.745619026315591e-05,
"loss": 0.7191,
"step": 3920
},
{
"epoch": 0.40176859968819484,
"grad_norm": 1.2610410451889038,
"learning_rate": 3.737875480330681e-05,
"loss": 0.6887,
"step": 3930
},
{
"epoch": 0.402790911646689,
"grad_norm": 1.5737214088439941,
"learning_rate": 3.7301161685883046e-05,
"loss": 0.6788,
"step": 3940
},
{
"epoch": 0.4038132236051831,
"grad_norm": 1.2577745914459229,
"learning_rate": 3.7223411899121527e-05,
"loss": 0.6841,
"step": 3950
},
{
"epoch": 0.40483553556367724,
"grad_norm": 1.0957412719726562,
"learning_rate": 3.714550643325458e-05,
"loss": 0.6791,
"step": 3960
},
{
"epoch": 0.4058578475221714,
"grad_norm": 1.392012119293213,
"learning_rate": 3.7067446280497227e-05,
"loss": 0.6964,
"step": 3970
},
{
"epoch": 0.4068801594806655,
"grad_norm": 1.692404866218567,
"learning_rate": 3.698923243503464e-05,
"loss": 0.6751,
"step": 3980
},
{
"epoch": 0.40790247143915964,
"grad_norm": 1.4865257740020752,
"learning_rate": 3.6910865893009445e-05,
"loss": 0.6709,
"step": 3990
},
{
"epoch": 0.4089247833976538,
"grad_norm": 1.286539912223816,
"learning_rate": 3.6832347652509003e-05,
"loss": 0.6693,
"step": 4000
},
{
"epoch": 0.40994709535614793,
"grad_norm": 1.6937637329101562,
"learning_rate": 3.6753678713552766e-05,
"loss": 0.6668,
"step": 4010
},
{
"epoch": 0.41096940731464204,
"grad_norm": 1.6795419454574585,
"learning_rate": 3.6674860078079464e-05,
"loss": 0.672,
"step": 4020
},
{
"epoch": 0.4119917192731362,
"grad_norm": 1.2515088319778442,
"learning_rate": 3.659589274993442e-05,
"loss": 0.6732,
"step": 4030
},
{
"epoch": 0.41301403123163033,
"grad_norm": 1.3944021463394165,
"learning_rate": 3.651677773485671e-05,
"loss": 0.6563,
"step": 4040
},
{
"epoch": 0.41403634319012445,
"grad_norm": 1.238376259803772,
"learning_rate": 3.643751604046637e-05,
"loss": 0.6692,
"step": 4050
},
{
"epoch": 0.4150586551486186,
"grad_norm": 1.6770118474960327,
"learning_rate": 3.6358108676251564e-05,
"loss": 0.6637,
"step": 4060
},
{
"epoch": 0.41608096710711273,
"grad_norm": 1.8990675210952759,
"learning_rate": 3.627855665355574e-05,
"loss": 0.6664,
"step": 4070
},
{
"epoch": 0.41710327906560685,
"grad_norm": 1.2769804000854492,
"learning_rate": 3.619886098556471e-05,
"loss": 0.6661,
"step": 4080
},
{
"epoch": 0.418125591024101,
"grad_norm": 1.2452665567398071,
"learning_rate": 3.6119022687293814e-05,
"loss": 0.6784,
"step": 4090
},
{
"epoch": 0.41914790298259513,
"grad_norm": 1.4984554052352905,
"learning_rate": 3.603904277557491e-05,
"loss": 0.6633,
"step": 4100
},
{
"epoch": 0.42017021494108925,
"grad_norm": 1.3026005029678345,
"learning_rate": 3.5958922269043494e-05,
"loss": 0.6519,
"step": 4110
},
{
"epoch": 0.4211925268995834,
"grad_norm": 1.2519993782043457,
"learning_rate": 3.587866218812568e-05,
"loss": 0.6659,
"step": 4120
},
{
"epoch": 0.42221483885807753,
"grad_norm": 1.5750147104263306,
"learning_rate": 3.579826355502522e-05,
"loss": 0.6647,
"step": 4130
},
{
"epoch": 0.42323715081657165,
"grad_norm": 1.7237874269485474,
"learning_rate": 3.571772739371053e-05,
"loss": 0.6668,
"step": 4140
},
{
"epoch": 0.4242594627750658,
"grad_norm": 1.135948896408081,
"learning_rate": 3.563705472990152e-05,
"loss": 0.669,
"step": 4150
},
{
"epoch": 0.42528177473355994,
"grad_norm": 1.2250343561172485,
"learning_rate": 3.55562465910567e-05,
"loss": 0.6523,
"step": 4160
},
{
"epoch": 0.4263040866920541,
"grad_norm": 1.2247508764266968,
"learning_rate": 3.547530400635996e-05,
"loss": 0.6547,
"step": 4170
},
{
"epoch": 0.4273263986505482,
"grad_norm": 1.4336878061294556,
"learning_rate": 3.5394228006707516e-05,
"loss": 0.6475,
"step": 4180
},
{
"epoch": 0.42834871060904234,
"grad_norm": 1.4483047723770142,
"learning_rate": 3.53130196246948e-05,
"loss": 0.6531,
"step": 4190
},
{
"epoch": 0.4293710225675365,
"grad_norm": 1.523287057876587,
"learning_rate": 3.523167989460325e-05,
"loss": 0.6489,
"step": 4200
},
{
"epoch": 0.4303933345260306,
"grad_norm": 1.4192699193954468,
"learning_rate": 3.515020985238721e-05,
"loss": 0.6544,
"step": 4210
},
{
"epoch": 0.43141564648452474,
"grad_norm": 1.9409384727478027,
"learning_rate": 3.5068610535660674e-05,
"loss": 0.654,
"step": 4220
},
{
"epoch": 0.4324379584430189,
"grad_norm": 1.008102536201477,
"learning_rate": 3.4986882983684094e-05,
"loss": 0.6492,
"step": 4230
},
{
"epoch": 0.433460270401513,
"grad_norm": 1.4971058368682861,
"learning_rate": 3.490502823735115e-05,
"loss": 0.6457,
"step": 4240
},
{
"epoch": 0.43448258236000714,
"grad_norm": 1.1196352243423462,
"learning_rate": 3.482304733917548e-05,
"loss": 0.645,
"step": 4250
},
{
"epoch": 0.4355048943185013,
"grad_norm": 1.4479795694351196,
"learning_rate": 3.474094133327743e-05,
"loss": 0.6456,
"step": 4260
},
{
"epoch": 0.4365272062769954,
"grad_norm": 1.162895679473877,
"learning_rate": 3.465871126537072e-05,
"loss": 0.6396,
"step": 4270
},
{
"epoch": 0.43754951823548954,
"grad_norm": 1.2522825002670288,
"learning_rate": 3.457635818274912e-05,
"loss": 0.6573,
"step": 4280
},
{
"epoch": 0.4385718301939837,
"grad_norm": 1.1335194110870361,
"learning_rate": 3.449388313427319e-05,
"loss": 0.6419,
"step": 4290
},
{
"epoch": 0.43959414215247783,
"grad_norm": 1.1282099485397339,
"learning_rate": 3.441128717035681e-05,
"loss": 0.6326,
"step": 4300
},
{
"epoch": 0.44061645411097194,
"grad_norm": 1.1025745868682861,
"learning_rate": 3.432857134295387e-05,
"loss": 0.6501,
"step": 4310
},
{
"epoch": 0.4416387660694661,
"grad_norm": 1.2452707290649414,
"learning_rate": 3.42457367055449e-05,
"loss": 0.6517,
"step": 4320
},
{
"epoch": 0.44266107802796023,
"grad_norm": 1.2010916471481323,
"learning_rate": 3.4162784313123524e-05,
"loss": 0.6474,
"step": 4330
},
{
"epoch": 0.44368338998645435,
"grad_norm": 1.2754626274108887,
"learning_rate": 3.40797152221832e-05,
"loss": 0.6451,
"step": 4340
},
{
"epoch": 0.4447057019449485,
"grad_norm": 1.244176983833313,
"learning_rate": 3.399653049070363e-05,
"loss": 0.651,
"step": 4350
},
{
"epoch": 0.44572801390344263,
"grad_norm": 1.5286939144134521,
"learning_rate": 3.3913231178137335e-05,
"loss": 0.6485,
"step": 4360
},
{
"epoch": 0.44675032586193675,
"grad_norm": 1.3893756866455078,
"learning_rate": 3.382981834539615e-05,
"loss": 0.6447,
"step": 4370
},
{
"epoch": 0.4477726378204309,
"grad_norm": 1.1257489919662476,
"learning_rate": 3.3746293054837746e-05,
"loss": 0.6408,
"step": 4380
},
{
"epoch": 0.44879494977892503,
"grad_norm": 1.3817309141159058,
"learning_rate": 3.3662656370252045e-05,
"loss": 0.6328,
"step": 4390
},
{
"epoch": 0.44981726173741915,
"grad_norm": 1.256188988685608,
"learning_rate": 3.357890935684771e-05,
"loss": 0.6449,
"step": 4400
},
{
"epoch": 0.4508395736959133,
"grad_norm": 1.2580057382583618,
"learning_rate": 3.3495053081238566e-05,
"loss": 0.6375,
"step": 4410
},
{
"epoch": 0.45186188565440744,
"grad_norm": 1.3607220649719238,
"learning_rate": 3.341108861143002e-05,
"loss": 0.6425,
"step": 4420
},
{
"epoch": 0.45288419761290155,
"grad_norm": 1.1394469738006592,
"learning_rate": 3.332701701680546e-05,
"loss": 0.6353,
"step": 4430
},
{
"epoch": 0.4539065095713957,
"grad_norm": 1.14969801902771,
"learning_rate": 3.3242839368112604e-05,
"loss": 0.6355,
"step": 4440
},
{
"epoch": 0.45492882152988984,
"grad_norm": 1.4116837978363037,
"learning_rate": 3.315855673744994e-05,
"loss": 0.6322,
"step": 4450
},
{
"epoch": 0.45595113348838395,
"grad_norm": 1.1528499126434326,
"learning_rate": 3.307417019825296e-05,
"loss": 0.6299,
"step": 4460
},
{
"epoch": 0.4569734454468781,
"grad_norm": 1.123056411743164,
"learning_rate": 3.2989680825280605e-05,
"loss": 0.6309,
"step": 4470
},
{
"epoch": 0.45799575740537224,
"grad_norm": 1.2016690969467163,
"learning_rate": 3.290508969460149e-05,
"loss": 0.6363,
"step": 4480
},
{
"epoch": 0.4590180693638664,
"grad_norm": 1.1473077535629272,
"learning_rate": 3.2820397883580224e-05,
"loss": 0.6365,
"step": 4490
},
{
"epoch": 0.4600403813223605,
"grad_norm": 1.4427560567855835,
"learning_rate": 3.273560647086373e-05,
"loss": 0.636,
"step": 4500
},
{
"epoch": 0.46106269328085464,
"grad_norm": 1.022484540939331,
"learning_rate": 3.265071653636744e-05,
"loss": 0.6175,
"step": 4510
},
{
"epoch": 0.4620850052393488,
"grad_norm": 0.9218525290489197,
"learning_rate": 3.256572916126158e-05,
"loss": 0.6262,
"step": 4520
},
{
"epoch": 0.4631073171978429,
"grad_norm": 1.4888445138931274,
"learning_rate": 3.248064542795741e-05,
"loss": 0.6304,
"step": 4530
},
{
"epoch": 0.46412962915633704,
"grad_norm": 1.0524561405181885,
"learning_rate": 3.239546642009341e-05,
"loss": 0.6345,
"step": 4540
},
{
"epoch": 0.4651519411148312,
"grad_norm": 1.0863170623779297,
"learning_rate": 3.231019322252148e-05,
"loss": 0.6363,
"step": 4550
},
{
"epoch": 0.46617425307332533,
"grad_norm": 1.2173206806182861,
"learning_rate": 3.222482692129315e-05,
"loss": 0.624,
"step": 4560
},
{
"epoch": 0.46719656503181944,
"grad_norm": 1.0352942943572998,
"learning_rate": 3.213936860364572e-05,
"loss": 0.6309,
"step": 4570
},
{
"epoch": 0.4682188769903136,
"grad_norm": 1.2327061891555786,
"learning_rate": 3.2053819357988446e-05,
"loss": 0.6211,
"step": 4580
},
{
"epoch": 0.46924118894880773,
"grad_norm": 1.1171348094940186,
"learning_rate": 3.196818027388861e-05,
"loss": 0.6228,
"step": 4590
},
{
"epoch": 0.47026350090730185,
"grad_norm": 1.1197806596755981,
"learning_rate": 3.188245244205774e-05,
"loss": 0.6186,
"step": 4600
},
{
"epoch": 0.471285812865796,
"grad_norm": 1.1804462671279907,
"learning_rate": 3.179663695433764e-05,
"loss": 0.6226,
"step": 4610
},
{
"epoch": 0.47230812482429013,
"grad_norm": 1.0908983945846558,
"learning_rate": 3.171073490368651e-05,
"loss": 0.6404,
"step": 4620
},
{
"epoch": 0.47333043678278425,
"grad_norm": 1.1561825275421143,
"learning_rate": 3.162474738416506e-05,
"loss": 0.6333,
"step": 4630
},
{
"epoch": 0.4743527487412784,
"grad_norm": 1.2377185821533203,
"learning_rate": 3.153867549092249e-05,
"loss": 0.6127,
"step": 4640
},
{
"epoch": 0.47537506069977253,
"grad_norm": 1.0879380702972412,
"learning_rate": 3.145252032018264e-05,
"loss": 0.6239,
"step": 4650
},
{
"epoch": 0.47639737265826665,
"grad_norm": 1.3811547756195068,
"learning_rate": 3.136628296922997e-05,
"loss": 0.6347,
"step": 4660
},
{
"epoch": 0.4774196846167608,
"grad_norm": 1.1660869121551514,
"learning_rate": 3.127996453639559e-05,
"loss": 0.6276,
"step": 4670
},
{
"epoch": 0.47844199657525494,
"grad_norm": 1.090694546699524,
"learning_rate": 3.119356612104329e-05,
"loss": 0.6254,
"step": 4680
},
{
"epoch": 0.47946430853374905,
"grad_norm": 1.4261144399642944,
"learning_rate": 3.110708882355553e-05,
"loss": 0.6275,
"step": 4690
},
{
"epoch": 0.4804866204922432,
"grad_norm": 1.1563290357589722,
"learning_rate": 3.102053374531941e-05,
"loss": 0.6356,
"step": 4700
},
{
"epoch": 0.48150893245073734,
"grad_norm": 1.0531911849975586,
"learning_rate": 3.0933901988712694e-05,
"loss": 0.6159,
"step": 4710
},
{
"epoch": 0.48253124440923145,
"grad_norm": 1.0314277410507202,
"learning_rate": 3.084719465708967e-05,
"loss": 0.636,
"step": 4720
},
{
"epoch": 0.4835535563677256,
"grad_norm": 1.0983450412750244,
"learning_rate": 3.076041285476722e-05,
"loss": 0.6117,
"step": 4730
},
{
"epoch": 0.48457586832621974,
"grad_norm": 1.3466790914535522,
"learning_rate": 3.067355768701067e-05,
"loss": 0.6151,
"step": 4740
},
{
"epoch": 0.48559818028471385,
"grad_norm": 1.1271053552627563,
"learning_rate": 3.0586630260019726e-05,
"loss": 0.6126,
"step": 4750
},
{
"epoch": 0.486620492243208,
"grad_norm": 1.1494412422180176,
"learning_rate": 3.049963168091443e-05,
"loss": 0.619,
"step": 4760
},
{
"epoch": 0.48764280420170214,
"grad_norm": 1.1014100313186646,
"learning_rate": 3.0412563057720995e-05,
"loss": 0.6103,
"step": 4770
},
{
"epoch": 0.4886651161601963,
"grad_norm": 1.075558066368103,
"learning_rate": 3.032542549935775e-05,
"loss": 0.6148,
"step": 4780
},
{
"epoch": 0.4896874281186904,
"grad_norm": 1.11659836769104,
"learning_rate": 3.0238220115620976e-05,
"loss": 0.6154,
"step": 4790
},
{
"epoch": 0.49070974007718454,
"grad_norm": 1.1824182271957397,
"learning_rate": 3.015094801717079e-05,
"loss": 0.6166,
"step": 4800
},
{
"epoch": 0.4917320520356787,
"grad_norm": 1.2691211700439453,
"learning_rate": 3.0063610315517003e-05,
"loss": 0.6206,
"step": 4810
},
{
"epoch": 0.49275436399417283,
"grad_norm": 1.342873454093933,
"learning_rate": 2.997620812300496e-05,
"loss": 0.6209,
"step": 4820
},
{
"epoch": 0.49377667595266694,
"grad_norm": 0.9883928298950195,
"learning_rate": 2.9888742552801358e-05,
"loss": 0.618,
"step": 4830
},
{
"epoch": 0.4947989879111611,
"grad_norm": 1.2620633840560913,
"learning_rate": 2.9801214718880095e-05,
"loss": 0.6153,
"step": 4840
},
{
"epoch": 0.49582129986965523,
"grad_norm": 1.2682327032089233,
"learning_rate": 2.971362573600806e-05,
"loss": 0.6158,
"step": 4850
},
{
"epoch": 0.49684361182814935,
"grad_norm": 1.1597089767456055,
"learning_rate": 2.9625976719730948e-05,
"loss": 0.6118,
"step": 4860
},
{
"epoch": 0.4978659237866435,
"grad_norm": 1.1553666591644287,
"learning_rate": 2.9538268786359046e-05,
"loss": 0.618,
"step": 4870
},
{
"epoch": 0.49888823574513763,
"grad_norm": 1.1677993535995483,
"learning_rate": 2.9450503052953004e-05,
"loss": 0.624,
"step": 4880
},
{
"epoch": 0.49991054770363175,
"grad_norm": 1.3060886859893799,
"learning_rate": 2.9362680637309664e-05,
"loss": 0.6117,
"step": 4890
},
{
"epoch": 0.5009328596621259,
"grad_norm": 1.0159618854522705,
"learning_rate": 2.9274802657947724e-05,
"loss": 0.6167,
"step": 4900
},
{
"epoch": 0.50195517162062,
"grad_norm": 1.0440400838851929,
"learning_rate": 2.9186870234093592e-05,
"loss": 0.6133,
"step": 4910
},
{
"epoch": 0.5029774835791142,
"grad_norm": 1.0517702102661133,
"learning_rate": 2.909888448566708e-05,
"loss": 0.6117,
"step": 4920
},
{
"epoch": 0.5039997955376083,
"grad_norm": 1.1240285634994507,
"learning_rate": 2.9010846533267118e-05,
"loss": 0.6132,
"step": 4930
},
{
"epoch": 0.5050221074961024,
"grad_norm": 0.9652560353279114,
"learning_rate": 2.8922757498157572e-05,
"loss": 0.6191,
"step": 4940
},
{
"epoch": 0.5060444194545966,
"grad_norm": 1.0895004272460938,
"learning_rate": 2.883461850225284e-05,
"loss": 0.6144,
"step": 4950
},
{
"epoch": 0.5070667314130907,
"grad_norm": 1.1447396278381348,
"learning_rate": 2.874643066810368e-05,
"loss": 0.6108,
"step": 4960
},
{
"epoch": 0.5080890433715849,
"grad_norm": 1.1374315023422241,
"learning_rate": 2.8658195118882835e-05,
"loss": 0.617,
"step": 4970
},
{
"epoch": 0.509111355330079,
"grad_norm": 0.9657441973686218,
"learning_rate": 2.856991297837076e-05,
"loss": 0.6143,
"step": 4980
},
{
"epoch": 0.5101336672885731,
"grad_norm": 1.0028339624404907,
"learning_rate": 2.8481585370941295e-05,
"loss": 0.6086,
"step": 4990
},
{
"epoch": 0.5111559792470672,
"grad_norm": 1.20085608959198,
"learning_rate": 2.8393213421547366e-05,
"loss": 0.6048,
"step": 5000
},
{
"epoch": 0.5121782912055614,
"grad_norm": 1.0971348285675049,
"learning_rate": 2.830479825570663e-05,
"loss": 0.6103,
"step": 5010
},
{
"epoch": 0.5132006031640555,
"grad_norm": 1.1244244575500488,
"learning_rate": 2.8216340999487184e-05,
"loss": 0.613,
"step": 5020
},
{
"epoch": 0.5142229151225497,
"grad_norm": 1.1182537078857422,
"learning_rate": 2.8127842779493152e-05,
"loss": 0.6211,
"step": 5030
},
{
"epoch": 0.5152452270810438,
"grad_norm": 1.237802505493164,
"learning_rate": 2.8039304722850408e-05,
"loss": 0.609,
"step": 5040
},
{
"epoch": 0.5162675390395379,
"grad_norm": 1.0564998388290405,
"learning_rate": 2.795072795719219e-05,
"loss": 0.6101,
"step": 5050
},
{
"epoch": 0.517289850998032,
"grad_norm": 1.0957964658737183,
"learning_rate": 2.7862113610644715e-05,
"loss": 0.5994,
"step": 5060
},
{
"epoch": 0.5183121629565262,
"grad_norm": 1.0157415866851807,
"learning_rate": 2.7773462811812877e-05,
"loss": 0.6125,
"step": 5070
},
{
"epoch": 0.5193344749150203,
"grad_norm": 0.9791542887687683,
"learning_rate": 2.768477668976579e-05,
"loss": 0.6051,
"step": 5080
},
{
"epoch": 0.5203567868735145,
"grad_norm": 0.9985860586166382,
"learning_rate": 2.7596056374022473e-05,
"loss": 0.5927,
"step": 5090
},
{
"epoch": 0.5213790988320086,
"grad_norm": 1.06964111328125,
"learning_rate": 2.7507302994537432e-05,
"loss": 0.6042,
"step": 5100
},
{
"epoch": 0.5224014107905027,
"grad_norm": 1.1804622411727905,
"learning_rate": 2.7418517681686283e-05,
"loss": 0.604,
"step": 5110
},
{
"epoch": 0.5234237227489968,
"grad_norm": 1.1587846279144287,
"learning_rate": 2.7329701566251354e-05,
"loss": 0.6002,
"step": 5120
},
{
"epoch": 0.524446034707491,
"grad_norm": 0.998511791229248,
"learning_rate": 2.724085577940726e-05,
"loss": 0.5997,
"step": 5130
},
{
"epoch": 0.5254683466659851,
"grad_norm": 1.0067111253738403,
"learning_rate": 2.7151981452706525e-05,
"loss": 0.6054,
"step": 5140
},
{
"epoch": 0.5264906586244793,
"grad_norm": 1.3026676177978516,
"learning_rate": 2.7063079718065182e-05,
"loss": 0.6071,
"step": 5150
},
{
"epoch": 0.5275129705829734,
"grad_norm": 1.1975481510162354,
"learning_rate": 2.6974151707748302e-05,
"loss": 0.6083,
"step": 5160
},
{
"epoch": 0.5285352825414675,
"grad_norm": 0.97922682762146,
"learning_rate": 2.688519855435563e-05,
"loss": 0.6008,
"step": 5170
},
{
"epoch": 0.5295575944999616,
"grad_norm": 1.0564764738082886,
"learning_rate": 2.6796221390807117e-05,
"loss": 0.6038,
"step": 5180
},
{
"epoch": 0.5305799064584558,
"grad_norm": 1.0322707891464233,
"learning_rate": 2.6707221350328526e-05,
"loss": 0.6123,
"step": 5190
},
{
"epoch": 0.5316022184169499,
"grad_norm": 1.5717302560806274,
"learning_rate": 2.6618199566436993e-05,
"loss": 0.6132,
"step": 5200
},
{
"epoch": 0.5326245303754441,
"grad_norm": 1.1104389429092407,
"learning_rate": 2.6529157172926546e-05,
"loss": 0.6043,
"step": 5210
},
{
"epoch": 0.5336468423339382,
"grad_norm": 1.0987706184387207,
"learning_rate": 2.644009530385374e-05,
"loss": 0.6024,
"step": 5220
},
{
"epoch": 0.5346691542924323,
"grad_norm": 1.0411555767059326,
"learning_rate": 2.6351015093523147e-05,
"loss": 0.6052,
"step": 5230
},
{
"epoch": 0.5356914662509265,
"grad_norm": 1.0305166244506836,
"learning_rate": 2.6261917676472946e-05,
"loss": 0.597,
"step": 5240
},
{
"epoch": 0.5367137782094206,
"grad_norm": 0.9780750274658203,
"learning_rate": 2.6172804187460465e-05,
"loss": 0.5985,
"step": 5250
},
{
"epoch": 0.5377360901679148,
"grad_norm": 1.1321512460708618,
"learning_rate": 2.6083675761447712e-05,
"loss": 0.596,
"step": 5260
},
{
"epoch": 0.5387584021264089,
"grad_norm": 1.090281367301941,
"learning_rate": 2.5994533533586955e-05,
"loss": 0.6019,
"step": 5270
},
{
"epoch": 0.539780714084903,
"grad_norm": 1.0389407873153687,
"learning_rate": 2.5905378639206242e-05,
"loss": 0.5928,
"step": 5280
},
{
"epoch": 0.5408030260433971,
"grad_norm": 1.1857671737670898,
"learning_rate": 2.5816212213794926e-05,
"loss": 0.6035,
"step": 5290
},
{
"epoch": 0.5418253380018913,
"grad_norm": 1.1074718236923218,
"learning_rate": 2.5727035392989235e-05,
"loss": 0.6076,
"step": 5300
},
{
"epoch": 0.5428476499603854,
"grad_norm": 1.0782848596572876,
"learning_rate": 2.5637849312557795e-05,
"loss": 0.6014,
"step": 5310
},
{
"epoch": 0.5438699619188796,
"grad_norm": 0.9669405221939087,
"learning_rate": 2.5548655108387144e-05,
"loss": 0.609,
"step": 5320
},
{
"epoch": 0.5448922738773737,
"grad_norm": 1.1341845989227295,
"learning_rate": 2.545945391646733e-05,
"loss": 0.606,
"step": 5330
},
{
"epoch": 0.5459145858358678,
"grad_norm": 1.0105434656143188,
"learning_rate": 2.537024687287734e-05,
"loss": 0.5994,
"step": 5340
},
{
"epoch": 0.5469368977943619,
"grad_norm": 1.2482975721359253,
"learning_rate": 2.5281035113770734e-05,
"loss": 0.5942,
"step": 5350
},
{
"epoch": 0.5479592097528561,
"grad_norm": 1.27554452419281,
"learning_rate": 2.519181977536111e-05,
"loss": 0.6103,
"step": 5360
},
{
"epoch": 0.5489815217113502,
"grad_norm": 1.1267889738082886,
"learning_rate": 2.510260199390765e-05,
"loss": 0.5969,
"step": 5370
},
{
"epoch": 0.5500038336698444,
"grad_norm": 1.138761043548584,
"learning_rate": 2.5013382905700673e-05,
"loss": 0.5939,
"step": 5380
},
{
"epoch": 0.5510261456283385,
"grad_norm": 1.1082335710525513,
"learning_rate": 2.4924163647047103e-05,
"loss": 0.5981,
"step": 5390
},
{
"epoch": 0.5520484575868326,
"grad_norm": 1.1725702285766602,
"learning_rate": 2.483494535425607e-05,
"loss": 0.5921,
"step": 5400
},
{
"epoch": 0.5530707695453267,
"grad_norm": 0.9450187087059021,
"learning_rate": 2.4745729163624382e-05,
"loss": 0.5889,
"step": 5410
},
{
"epoch": 0.5540930815038209,
"grad_norm": 1.1566860675811768,
"learning_rate": 2.4656516211422084e-05,
"loss": 0.6031,
"step": 5420
},
{
"epoch": 0.555115393462315,
"grad_norm": 0.9692696332931519,
"learning_rate": 2.456730763387799e-05,
"loss": 0.588,
"step": 5430
},
{
"epoch": 0.5561377054208092,
"grad_norm": 1.0729663372039795,
"learning_rate": 2.4478104567165143e-05,
"loss": 0.5949,
"step": 5440
},
{
"epoch": 0.5571600173793033,
"grad_norm": 1.2313849925994873,
"learning_rate": 2.438890814738647e-05,
"loss": 0.5976,
"step": 5450
},
{
"epoch": 0.5581823293377974,
"grad_norm": 1.0119988918304443,
"learning_rate": 2.4299719510560184e-05,
"loss": 0.5888,
"step": 5460
},
{
"epoch": 0.5592046412962915,
"grad_norm": 1.093971610069275,
"learning_rate": 2.421053979260541e-05,
"loss": 0.5984,
"step": 5470
},
{
"epoch": 0.5602269532547857,
"grad_norm": 1.0052496194839478,
"learning_rate": 2.4121370129327688e-05,
"loss": 0.5937,
"step": 5480
},
{
"epoch": 0.5612492652132798,
"grad_norm": 1.1699482202529907,
"learning_rate": 2.4032211656404442e-05,
"loss": 0.5938,
"step": 5490
},
{
"epoch": 0.562271577171774,
"grad_norm": 1.1209789514541626,
"learning_rate": 2.394306550937063e-05,
"loss": 0.597,
"step": 5500
},
{
"epoch": 0.5632938891302681,
"grad_norm": 1.0126326084136963,
"learning_rate": 2.3853932823604212e-05,
"loss": 0.5974,
"step": 5510
},
{
"epoch": 0.5643162010887622,
"grad_norm": 0.9977943301200867,
"learning_rate": 2.376481473431171e-05,
"loss": 0.593,
"step": 5520
},
{
"epoch": 0.5653385130472564,
"grad_norm": 1.088913083076477,
"learning_rate": 2.367571237651371e-05,
"loss": 0.5966,
"step": 5530
},
{
"epoch": 0.5663608250057505,
"grad_norm": 1.3126310110092163,
"learning_rate": 2.3586626885030472e-05,
"loss": 0.5997,
"step": 5540
},
{
"epoch": 0.5673831369642447,
"grad_norm": 1.048432469367981,
"learning_rate": 2.3497559394467443e-05,
"loss": 0.5916,
"step": 5550
},
{
"epoch": 0.5684054489227388,
"grad_norm": 0.97364342212677,
"learning_rate": 2.3408511039200784e-05,
"loss": 0.5949,
"step": 5560
},
{
"epoch": 0.5694277608812329,
"grad_norm": 1.2537837028503418,
"learning_rate": 2.3319482953362982e-05,
"loss": 0.5886,
"step": 5570
},
{
"epoch": 0.570450072839727,
"grad_norm": 1.2341266870498657,
"learning_rate": 2.3230476270828334e-05,
"loss": 0.5913,
"step": 5580
},
{
"epoch": 0.5714723847982212,
"grad_norm": 0.9848710894584656,
"learning_rate": 2.3141492125198554e-05,
"loss": 0.599,
"step": 5590
},
{
"epoch": 0.5724946967567153,
"grad_norm": 0.9984156489372253,
"learning_rate": 2.3052531649788337e-05,
"loss": 0.5904,
"step": 5600
},
{
"epoch": 0.5735170087152095,
"grad_norm": 0.9772480726242065,
"learning_rate": 2.2963595977610907e-05,
"loss": 0.584,
"step": 5610
},
{
"epoch": 0.5745393206737036,
"grad_norm": 0.8886341452598572,
"learning_rate": 2.2874686241363558e-05,
"loss": 0.5949,
"step": 5620
},
{
"epoch": 0.5755616326321977,
"grad_norm": 0.9721497893333435,
"learning_rate": 2.2785803573413302e-05,
"loss": 0.5981,
"step": 5630
},
{
"epoch": 0.5765839445906918,
"grad_norm": 1.228590965270996,
"learning_rate": 2.269694910578238e-05,
"loss": 0.588,
"step": 5640
},
{
"epoch": 0.577606256549186,
"grad_norm": 1.0548731088638306,
"learning_rate": 2.2608123970133893e-05,
"loss": 0.5889,
"step": 5650
},
{
"epoch": 0.5786285685076801,
"grad_norm": 1.0768656730651855,
"learning_rate": 2.251932929775734e-05,
"loss": 0.5909,
"step": 5660
},
{
"epoch": 0.5796508804661743,
"grad_norm": 0.887742280960083,
"learning_rate": 2.243056621955424e-05,
"loss": 0.5859,
"step": 5670
},
{
"epoch": 0.5806731924246684,
"grad_norm": 1.1718515157699585,
"learning_rate": 2.2341835866023744e-05,
"loss": 0.5915,
"step": 5680
},
{
"epoch": 0.5816955043831625,
"grad_norm": 0.9442711472511292,
"learning_rate": 2.2253139367248192e-05,
"loss": 0.5831,
"step": 5690
},
{
"epoch": 0.5827178163416566,
"grad_norm": 1.2530230283737183,
"learning_rate": 2.216447785287878e-05,
"loss": 0.5893,
"step": 5700
},
{
"epoch": 0.5837401283001508,
"grad_norm": 1.032161831855774,
"learning_rate": 2.2075852452121083e-05,
"loss": 0.586,
"step": 5710
},
{
"epoch": 0.5847624402586449,
"grad_norm": 1.1369349956512451,
"learning_rate": 2.1987264293720774e-05,
"loss": 0.5915,
"step": 5720
},
{
"epoch": 0.5857847522171391,
"grad_norm": 0.9815176725387573,
"learning_rate": 2.189871450594919e-05,
"loss": 0.5896,
"step": 5730
},
{
"epoch": 0.5868070641756332,
"grad_norm": 1.126705288887024,
"learning_rate": 2.1810204216588975e-05,
"loss": 0.5755,
"step": 5740
},
{
"epoch": 0.5878293761341273,
"grad_norm": 1.1156599521636963,
"learning_rate": 2.172173455291969e-05,
"loss": 0.578,
"step": 5750
},
{
"epoch": 0.5888516880926215,
"grad_norm": 1.1087192296981812,
"learning_rate": 2.163330664170353e-05,
"loss": 0.5877,
"step": 5760
},
{
"epoch": 0.5898740000511156,
"grad_norm": 1.0050220489501953,
"learning_rate": 2.154492160917088e-05,
"loss": 0.5852,
"step": 5770
},
{
"epoch": 0.5908963120096097,
"grad_norm": 1.019031286239624,
"learning_rate": 2.1456580581006047e-05,
"loss": 0.5794,
"step": 5780
},
{
"epoch": 0.5919186239681039,
"grad_norm": 1.1089690923690796,
"learning_rate": 2.136828468233288e-05,
"loss": 0.5848,
"step": 5790
},
{
"epoch": 0.592940935926598,
"grad_norm": 0.9696457982063293,
"learning_rate": 2.1280035037700436e-05,
"loss": 0.5894,
"step": 5800
},
{
"epoch": 0.5939632478850921,
"grad_norm": 0.9017470479011536,
"learning_rate": 2.1191832771068705e-05,
"loss": 0.5835,
"step": 5810
},
{
"epoch": 0.5949855598435863,
"grad_norm": 0.8829480409622192,
"learning_rate": 2.110367900579424e-05,
"loss": 0.5817,
"step": 5820
},
{
"epoch": 0.5960078718020804,
"grad_norm": 0.8892145752906799,
"learning_rate": 2.1015574864615896e-05,
"loss": 0.5821,
"step": 5830
},
{
"epoch": 0.5970301837605745,
"grad_norm": 1.0441607236862183,
"learning_rate": 2.092752146964048e-05,
"loss": 0.5824,
"step": 5840
},
{
"epoch": 0.5980524957190687,
"grad_norm": 1.126844882965088,
"learning_rate": 2.0839519942328512e-05,
"loss": 0.5841,
"step": 5850
},
{
"epoch": 0.5990748076775628,
"grad_norm": 0.9662244915962219,
"learning_rate": 2.0751571403479906e-05,
"loss": 0.5873,
"step": 5860
},
{
"epoch": 0.6000971196360569,
"grad_norm": 0.9771221876144409,
"learning_rate": 2.066367697321971e-05,
"loss": 0.5817,
"step": 5870
},
{
"epoch": 0.6011194315945511,
"grad_norm": 1.3460716009140015,
"learning_rate": 2.057583777098385e-05,
"loss": 0.5916,
"step": 5880
},
{
"epoch": 0.6021417435530452,
"grad_norm": 1.0285344123840332,
"learning_rate": 2.048805491550484e-05,
"loss": 0.5928,
"step": 5890
},
{
"epoch": 0.6031640555115394,
"grad_norm": 0.9344070553779602,
"learning_rate": 2.0400329524797553e-05,
"loss": 0.5822,
"step": 5900
},
{
"epoch": 0.6041863674700335,
"grad_norm": 0.914261519908905,
"learning_rate": 2.0312662716145004e-05,
"loss": 0.5784,
"step": 5910
},
{
"epoch": 0.6052086794285276,
"grad_norm": 0.9938135743141174,
"learning_rate": 2.022505560608409e-05,
"loss": 0.5819,
"step": 5920
},
{
"epoch": 0.6062309913870217,
"grad_norm": 1.0218501091003418,
"learning_rate": 2.0137509310391357e-05,
"loss": 0.581,
"step": 5930
},
{
"epoch": 0.6072533033455159,
"grad_norm": 1.1567462682724,
"learning_rate": 2.0050024944068847e-05,
"loss": 0.5798,
"step": 5940
},
{
"epoch": 0.60827561530401,
"grad_norm": 1.017453908920288,
"learning_rate": 1.9962603621329823e-05,
"loss": 0.5798,
"step": 5950
},
{
"epoch": 0.6092979272625042,
"grad_norm": 1.0345451831817627,
"learning_rate": 1.9875246455584658e-05,
"loss": 0.5749,
"step": 5960
},
{
"epoch": 0.6103202392209983,
"grad_norm": 1.003942608833313,
"learning_rate": 1.9787954559426574e-05,
"loss": 0.5891,
"step": 5970
},
{
"epoch": 0.6113425511794924,
"grad_norm": 1.0570673942565918,
"learning_rate": 1.9700729044617523e-05,
"loss": 0.5786,
"step": 5980
},
{
"epoch": 0.6123648631379865,
"grad_norm": 0.9278817176818848,
"learning_rate": 1.961357102207403e-05,
"loss": 0.5809,
"step": 5990
},
{
"epoch": 0.6133871750964807,
"grad_norm": 1.0820597410202026,
"learning_rate": 1.9526481601853004e-05,
"loss": 0.5908,
"step": 6000
},
{
"epoch": 0.6144094870549748,
"grad_norm": 1.1385077238082886,
"learning_rate": 1.9439461893137654e-05,
"loss": 0.5897,
"step": 6010
},
{
"epoch": 0.615431799013469,
"grad_norm": 0.9647600650787354,
"learning_rate": 1.93525130042233e-05,
"loss": 0.5807,
"step": 6020
},
{
"epoch": 0.6164541109719631,
"grad_norm": 1.0689774751663208,
"learning_rate": 1.926563604250331e-05,
"loss": 0.5806,
"step": 6030
},
{
"epoch": 0.6174764229304572,
"grad_norm": 1.1144829988479614,
"learning_rate": 1.9178832114454982e-05,
"loss": 0.5831,
"step": 6040
},
{
"epoch": 0.6184987348889514,
"grad_norm": 0.9306235909461975,
"learning_rate": 1.9092102325625447e-05,
"loss": 0.5801,
"step": 6050
},
{
"epoch": 0.6195210468474455,
"grad_norm": 0.9796899557113647,
"learning_rate": 1.900544778061756e-05,
"loss": 0.5832,
"step": 6060
},
{
"epoch": 0.6205433588059396,
"grad_norm": 0.9838829636573792,
"learning_rate": 1.8918869583075893e-05,
"loss": 0.5906,
"step": 6070
},
{
"epoch": 0.6215656707644338,
"grad_norm": 0.981938898563385,
"learning_rate": 1.883236883567263e-05,
"loss": 0.5827,
"step": 6080
},
{
"epoch": 0.6225879827229279,
"grad_norm": 0.8918649554252625,
"learning_rate": 1.8745946640093554e-05,
"loss": 0.5738,
"step": 6090
},
{
"epoch": 0.623610294681422,
"grad_norm": 0.9547054171562195,
"learning_rate": 1.8659604097023986e-05,
"loss": 0.5789,
"step": 6100
},
{
"epoch": 0.6246326066399162,
"grad_norm": 1.005513072013855,
"learning_rate": 1.857334230613477e-05,
"loss": 0.5818,
"step": 6110
},
{
"epoch": 0.6256549185984103,
"grad_norm": 0.8938051462173462,
"learning_rate": 1.848716236606831e-05,
"loss": 0.5746,
"step": 6120
},
{
"epoch": 0.6266772305569044,
"grad_norm": 1.0492790937423706,
"learning_rate": 1.8401065374424524e-05,
"loss": 0.5809,
"step": 6130
},
{
"epoch": 0.6276995425153986,
"grad_norm": 0.9256030917167664,
"learning_rate": 1.8315052427746907e-05,
"loss": 0.5771,
"step": 6140
},
{
"epoch": 0.6287218544738927,
"grad_norm": 0.9747220873832703,
"learning_rate": 1.8229124621508514e-05,
"loss": 0.5784,
"step": 6150
},
{
"epoch": 0.6297441664323868,
"grad_norm": 0.9733263850212097,
"learning_rate": 1.8143283050098058e-05,
"loss": 0.578,
"step": 6160
},
{
"epoch": 0.630766478390881,
"grad_norm": 1.1085481643676758,
"learning_rate": 1.8057528806805968e-05,
"loss": 0.5653,
"step": 6170
},
{
"epoch": 0.6317887903493751,
"grad_norm": 0.9532302021980286,
"learning_rate": 1.7971862983810414e-05,
"loss": 0.5745,
"step": 6180
},
{
"epoch": 0.6328111023078693,
"grad_norm": 0.9900447130203247,
"learning_rate": 1.7886286672163477e-05,
"loss": 0.5763,
"step": 6190
},
{
"epoch": 0.6338334142663634,
"grad_norm": 1.0198087692260742,
"learning_rate": 1.7800800961777164e-05,
"loss": 0.5793,
"step": 6200
},
{
"epoch": 0.6348557262248575,
"grad_norm": 0.9209514856338501,
"learning_rate": 1.771540694140959e-05,
"loss": 0.5726,
"step": 6210
},
{
"epoch": 0.6358780381833516,
"grad_norm": 1.0615392923355103,
"learning_rate": 1.763010569865111e-05,
"loss": 0.5787,
"step": 6220
},
{
"epoch": 0.6369003501418458,
"grad_norm": 0.9032433032989502,
"learning_rate": 1.7544898319910436e-05,
"loss": 0.5692,
"step": 6230
},
{
"epoch": 0.6379226621003399,
"grad_norm": 0.9536076188087463,
"learning_rate": 1.7459785890400794e-05,
"loss": 0.5806,
"step": 6240
},
{
"epoch": 0.6389449740588341,
"grad_norm": 0.7053399085998535,
"learning_rate": 1.7374769494126157e-05,
"loss": 0.5764,
"step": 6250
},
{
"epoch": 0.6399672860173282,
"grad_norm": 0.6956759691238403,
"learning_rate": 1.7289850213867388e-05,
"loss": 0.5756,
"step": 6260
},
{
"epoch": 0.6409895979758223,
"grad_norm": 0.6635085940361023,
"learning_rate": 1.7205029131168478e-05,
"loss": 0.5746,
"step": 6270
},
{
"epoch": 0.6420119099343164,
"grad_norm": 0.6673712730407715,
"learning_rate": 1.7120307326322738e-05,
"loss": 0.569,
"step": 6280
},
{
"epoch": 0.6430342218928106,
"grad_norm": 0.6514122486114502,
"learning_rate": 1.7035685878359074e-05,
"loss": 0.564,
"step": 6290
},
{
"epoch": 0.6440565338513047,
"grad_norm": 0.7287066578865051,
"learning_rate": 1.6951165865028253e-05,
"loss": 0.5697,
"step": 6300
},
{
"epoch": 0.6450788458097989,
"grad_norm": 0.6336085200309753,
"learning_rate": 1.6866748362789114e-05,
"loss": 0.5692,
"step": 6310
},
{
"epoch": 0.646101157768293,
"grad_norm": 0.6699496507644653,
"learning_rate": 1.678243444679495e-05,
"loss": 0.568,
"step": 6320
},
{
"epoch": 0.6471234697267871,
"grad_norm": 0.6310245990753174,
"learning_rate": 1.669822519087972e-05,
"loss": 0.5621,
"step": 6330
},
{
"epoch": 0.6481457816852813,
"grad_norm": 0.6521554589271545,
"learning_rate": 1.661412166754443e-05,
"loss": 0.5613,
"step": 6340
},
{
"epoch": 0.6491680936437754,
"grad_norm": 0.8037363886833191,
"learning_rate": 1.653012494794348e-05,
"loss": 0.5606,
"step": 6350
},
{
"epoch": 0.6501904056022695,
"grad_norm": 0.827194094657898,
"learning_rate": 1.6446236101871e-05,
"loss": 0.5677,
"step": 6360
},
{
"epoch": 0.6512127175607637,
"grad_norm": 0.7061733603477478,
"learning_rate": 1.6362456197747185e-05,
"loss": 0.572,
"step": 6370
},
{
"epoch": 0.6522350295192578,
"grad_norm": 0.7039258480072021,
"learning_rate": 1.6278786302604786e-05,
"loss": 0.5617,
"step": 6380
},
{
"epoch": 0.6532573414777519,
"grad_norm": 0.679413914680481,
"learning_rate": 1.619522748207543e-05,
"loss": 0.5646,
"step": 6390
},
{
"epoch": 0.6542796534362461,
"grad_norm": 0.6594499945640564,
"learning_rate": 1.6111780800376098e-05,
"loss": 0.5625,
"step": 6400
},
{
"epoch": 0.6553019653947402,
"grad_norm": 0.7411320209503174,
"learning_rate": 1.6028447320295554e-05,
"loss": 0.5604,
"step": 6410
},
{
"epoch": 0.6563242773532343,
"grad_norm": 0.6764445900917053,
"learning_rate": 1.594522810318078e-05,
"loss": 0.5665,
"step": 6420
},
{
"epoch": 0.6573465893117285,
"grad_norm": 0.7496932744979858,
"learning_rate": 1.586212420892354e-05,
"loss": 0.5633,
"step": 6430
},
{
"epoch": 0.6583689012702226,
"grad_norm": 0.6481267213821411,
"learning_rate": 1.577913669594679e-05,
"loss": 0.5688,
"step": 6440
},
{
"epoch": 0.6593912132287167,
"grad_norm": 0.6794611215591431,
"learning_rate": 1.569626662119126e-05,
"loss": 0.5593,
"step": 6450
},
{
"epoch": 0.6604135251872109,
"grad_norm": 0.6898994445800781,
"learning_rate": 1.561351504010195e-05,
"loss": 0.5537,
"step": 6460
},
{
"epoch": 0.661435837145705,
"grad_norm": 0.6871402263641357,
"learning_rate": 1.553088300661472e-05,
"loss": 0.5647,
"step": 6470
},
{
"epoch": 0.6624581491041992,
"grad_norm": 0.7076891660690308,
"learning_rate": 1.544837157314287e-05,
"loss": 0.5609,
"step": 6480
},
{
"epoch": 0.6634804610626933,
"grad_norm": 0.6497926115989685,
"learning_rate": 1.5365981790563676e-05,
"loss": 0.5596,
"step": 6490
},
{
"epoch": 0.6645027730211874,
"grad_norm": 0.648878276348114,
"learning_rate": 1.528371470820511e-05,
"loss": 0.5601,
"step": 6500
},
{
"epoch": 0.6655250849796815,
"grad_norm": 0.7414587736129761,
"learning_rate": 1.5201571373832368e-05,
"loss": 0.5632,
"step": 6510
},
{
"epoch": 0.6665473969381757,
"grad_norm": 0.7688888907432556,
"learning_rate": 1.5119552833634592e-05,
"loss": 0.5608,
"step": 6520
},
{
"epoch": 0.6675697088966698,
"grad_norm": 1.178039789199829,
"learning_rate": 1.503766013221154e-05,
"loss": 0.4508,
"step": 6530
},
{
"epoch": 0.668592020855164,
"grad_norm": 0.9396372437477112,
"learning_rate": 1.4955894312560254e-05,
"loss": 0.438,
"step": 6540
},
{
"epoch": 0.6696143328136581,
"grad_norm": 1.0860793590545654,
"learning_rate": 1.4874256416061783e-05,
"loss": 0.436,
"step": 6550
},
{
"epoch": 0.6706366447721522,
"grad_norm": 1.0847822427749634,
"learning_rate": 1.4792747482467953e-05,
"loss": 0.4405,
"step": 6560
},
{
"epoch": 0.6716589567306464,
"grad_norm": 1.535868525505066,
"learning_rate": 1.4711368549888082e-05,
"loss": 0.4479,
"step": 6570
},
{
"epoch": 0.6726812686891405,
"grad_norm": 0.9498069286346436,
"learning_rate": 1.46301206547758e-05,
"loss": 0.4365,
"step": 6580
},
{
"epoch": 0.6737035806476346,
"grad_norm": 1.041296362876892,
"learning_rate": 1.4549004831915786e-05,
"loss": 0.4374,
"step": 6590
},
{
"epoch": 0.6747258926061288,
"grad_norm": 1.466341495513916,
"learning_rate": 1.4468022114410663e-05,
"loss": 0.4403,
"step": 6600
},
{
"epoch": 0.6757482045646229,
"grad_norm": 0.9726791381835938,
"learning_rate": 1.4387173533667785e-05,
"loss": 0.4496,
"step": 6610
},
{
"epoch": 0.676770516523117,
"grad_norm": 0.8319998979568481,
"learning_rate": 1.430646011938614e-05,
"loss": 0.4433,
"step": 6620
},
{
"epoch": 0.6777928284816112,
"grad_norm": 1.3163760900497437,
"learning_rate": 1.4225882899543194e-05,
"loss": 0.453,
"step": 6630
},
{
"epoch": 0.6788151404401053,
"grad_norm": 1.053041696548462,
"learning_rate": 1.4145442900381839e-05,
"loss": 0.4344,
"step": 6640
},
{
"epoch": 0.6798374523985994,
"grad_norm": 0.95073002576828,
"learning_rate": 1.4065141146397298e-05,
"loss": 0.4396,
"step": 6650
},
{
"epoch": 0.6808597643570936,
"grad_norm": 1.0832467079162598,
"learning_rate": 1.3984978660324077e-05,
"loss": 0.448,
"step": 6660
},
{
"epoch": 0.6818820763155877,
"grad_norm": 1.0101234912872314,
"learning_rate": 1.3904956463122982e-05,
"loss": 0.4547,
"step": 6670
},
{
"epoch": 0.6829043882740818,
"grad_norm": 0.9201359748840332,
"learning_rate": 1.3825075573968018e-05,
"loss": 0.4502,
"step": 6680
},
{
"epoch": 0.683926700232576,
"grad_norm": 1.0069217681884766,
"learning_rate": 1.3745337010233511e-05,
"loss": 0.4587,
"step": 6690
},
{
"epoch": 0.6849490121910701,
"grad_norm": 1.095893144607544,
"learning_rate": 1.3665741787481121e-05,
"loss": 0.4466,
"step": 6700
},
{
"epoch": 0.6859713241495642,
"grad_norm": 1.0467944145202637,
"learning_rate": 1.358629091944687e-05,
"loss": 0.4449,
"step": 6710
},
{
"epoch": 0.6869936361080584,
"grad_norm": 0.9458824992179871,
"learning_rate": 1.3506985418028268e-05,
"loss": 0.4425,
"step": 6720
},
{
"epoch": 0.6880159480665525,
"grad_norm": 1.1064608097076416,
"learning_rate": 1.3427826293271417e-05,
"loss": 0.4447,
"step": 6730
},
{
"epoch": 0.6890382600250466,
"grad_norm": 1.5638915300369263,
"learning_rate": 1.3348814553358139e-05,
"loss": 0.4557,
"step": 6740
},
{
"epoch": 0.6900605719835408,
"grad_norm": 1.1361709833145142,
"learning_rate": 1.326995120459315e-05,
"loss": 0.4581,
"step": 6750
},
{
"epoch": 0.6910828839420349,
"grad_norm": 1.0405875444412231,
"learning_rate": 1.3191237251391228e-05,
"loss": 0.4467,
"step": 6760
},
{
"epoch": 0.692105195900529,
"grad_norm": 1.0086606740951538,
"learning_rate": 1.3112673696264432e-05,
"loss": 0.4452,
"step": 6770
},
{
"epoch": 0.6931275078590232,
"grad_norm": 1.273151159286499,
"learning_rate": 1.3034261539809333e-05,
"loss": 0.4511,
"step": 6780
},
{
"epoch": 0.6941498198175173,
"grad_norm": 1.157692551612854,
"learning_rate": 1.2956001780694249e-05,
"loss": 0.4476,
"step": 6790
},
{
"epoch": 0.6951721317760114,
"grad_norm": 0.9611133933067322,
"learning_rate": 1.2877895415646584e-05,
"loss": 0.4541,
"step": 6800
},
{
"epoch": 0.6961944437345056,
"grad_norm": 1.138689637184143,
"learning_rate": 1.2799943439440048e-05,
"loss": 0.4521,
"step": 6810
},
{
"epoch": 0.6972167556929997,
"grad_norm": 1.0011394023895264,
"learning_rate": 1.2722146844882043e-05,
"loss": 0.451,
"step": 6820
},
{
"epoch": 0.6982390676514939,
"grad_norm": 0.9988419413566589,
"learning_rate": 1.264450662280104e-05,
"loss": 0.4486,
"step": 6830
},
{
"epoch": 0.699261379609988,
"grad_norm": 1.104713797569275,
"learning_rate": 1.2567023762033897e-05,
"loss": 0.4512,
"step": 6840
},
{
"epoch": 0.7002836915684821,
"grad_norm": 0.9885740876197815,
"learning_rate": 1.2489699249413302e-05,
"loss": 0.4502,
"step": 6850
},
{
"epoch": 0.7013060035269763,
"grad_norm": 1.0035451650619507,
"learning_rate": 1.2412534069755204e-05,
"loss": 0.4421,
"step": 6860
},
{
"epoch": 0.7023283154854704,
"grad_norm": 1.2350925207138062,
"learning_rate": 1.233552920584626e-05,
"loss": 0.4456,
"step": 6870
},
{
"epoch": 0.7033506274439645,
"grad_norm": 0.8888701796531677,
"learning_rate": 1.2258685638431325e-05,
"loss": 0.4521,
"step": 6880
},
{
"epoch": 0.7043729394024587,
"grad_norm": 0.8908058404922485,
"learning_rate": 1.2182004346200958e-05,
"loss": 0.4508,
"step": 6890
},
{
"epoch": 0.7053952513609528,
"grad_norm": 1.2026066780090332,
"learning_rate": 1.2105486305778965e-05,
"loss": 0.451,
"step": 6900
},
{
"epoch": 0.7064175633194469,
"grad_norm": 1.7937166690826416,
"learning_rate": 1.2029132491709938e-05,
"loss": 0.4431,
"step": 6910
},
{
"epoch": 0.707439875277941,
"grad_norm": 0.8573371767997742,
"learning_rate": 1.1952943876446867e-05,
"loss": 0.4453,
"step": 6920
},
{
"epoch": 0.7084621872364352,
"grad_norm": 1.0284944772720337,
"learning_rate": 1.1876921430338764e-05,
"loss": 0.4474,
"step": 6930
},
{
"epoch": 0.7094844991949293,
"grad_norm": 1.1545257568359375,
"learning_rate": 1.1801066121618273e-05,
"loss": 0.4441,
"step": 6940
},
{
"epoch": 0.7105068111534235,
"grad_norm": 0.9734779000282288,
"learning_rate": 1.172537891638932e-05,
"loss": 0.4516,
"step": 6950
},
{
"epoch": 0.7115291231119176,
"grad_norm": 1.1200814247131348,
"learning_rate": 1.1649860778614905e-05,
"loss": 0.4544,
"step": 6960
},
{
"epoch": 0.7125514350704117,
"grad_norm": 1.1503158807754517,
"learning_rate": 1.1574512670104714e-05,
"loss": 0.4369,
"step": 6970
},
{
"epoch": 0.7135737470289059,
"grad_norm": 1.019938588142395,
"learning_rate": 1.1499335550502936e-05,
"loss": 0.44,
"step": 6980
},
{
"epoch": 0.7145960589874,
"grad_norm": 1.1419305801391602,
"learning_rate": 1.142433037727601e-05,
"loss": 0.4533,
"step": 6990
},
{
"epoch": 0.7156183709458941,
"grad_norm": 0.9554847478866577,
"learning_rate": 1.1349498105700449e-05,
"loss": 0.4494,
"step": 7000
},
{
"epoch": 0.7166406829043883,
"grad_norm": 0.872265636920929,
"learning_rate": 1.1274839688850663e-05,
"loss": 0.4454,
"step": 7010
},
{
"epoch": 0.7176629948628824,
"grad_norm": 0.9257216453552246,
"learning_rate": 1.1200356077586819e-05,
"loss": 0.4446,
"step": 7020
},
{
"epoch": 0.7186853068213765,
"grad_norm": 1.1358165740966797,
"learning_rate": 1.1126048220542764e-05,
"loss": 0.4492,
"step": 7030
},
{
"epoch": 0.7197076187798707,
"grad_norm": 1.0924599170684814,
"learning_rate": 1.1051917064113867e-05,
"loss": 0.4526,
"step": 7040
},
{
"epoch": 0.7207299307383648,
"grad_norm": 1.081459879875183,
"learning_rate": 1.0977963552445037e-05,
"loss": 0.4396,
"step": 7050
},
{
"epoch": 0.7217522426968589,
"grad_norm": 1.0317836999893188,
"learning_rate": 1.090418862741869e-05,
"loss": 0.4461,
"step": 7060
},
{
"epoch": 0.7227745546553531,
"grad_norm": 1.1487315893173218,
"learning_rate": 1.0830593228642724e-05,
"loss": 0.4502,
"step": 7070
},
{
"epoch": 0.7237968666138472,
"grad_norm": 1.1723712682724,
"learning_rate": 1.0757178293438536e-05,
"loss": 0.4411,
"step": 7080
},
{
"epoch": 0.7248191785723413,
"grad_norm": 1.3718030452728271,
"learning_rate": 1.0683944756829168e-05,
"loss": 0.4448,
"step": 7090
},
{
"epoch": 0.7258414905308355,
"grad_norm": 1.007973074913025,
"learning_rate": 1.0610893551527304e-05,
"loss": 0.4462,
"step": 7100
},
{
"epoch": 0.7268638024893296,
"grad_norm": 0.9495711922645569,
"learning_rate": 1.0538025607923444e-05,
"loss": 0.442,
"step": 7110
},
{
"epoch": 0.7278861144478238,
"grad_norm": 1.1003414392471313,
"learning_rate": 1.0465341854074043e-05,
"loss": 0.4397,
"step": 7120
},
{
"epoch": 0.7289084264063179,
"grad_norm": 1.0131003856658936,
"learning_rate": 1.039284321568968e-05,
"loss": 0.4469,
"step": 7130
},
{
"epoch": 0.729930738364812,
"grad_norm": 0.9765979051589966,
"learning_rate": 1.032053061612329e-05,
"loss": 0.4439,
"step": 7140
},
{
"epoch": 0.7309530503233062,
"grad_norm": 0.959646999835968,
"learning_rate": 1.0248404976358369e-05,
"loss": 0.4551,
"step": 7150
},
{
"epoch": 0.7319753622818003,
"grad_norm": 1.0495507717132568,
"learning_rate": 1.0176467214997312e-05,
"loss": 0.4493,
"step": 7160
},
{
"epoch": 0.7329976742402944,
"grad_norm": 1.0175000429153442,
"learning_rate": 1.0104718248249611e-05,
"loss": 0.4448,
"step": 7170
},
{
"epoch": 0.7340199861987886,
"grad_norm": 0.970582127571106,
"learning_rate": 1.0033158989920271e-05,
"loss": 0.4482,
"step": 7180
},
{
"epoch": 0.7350422981572827,
"grad_norm": 1.2196745872497559,
"learning_rate": 9.96179035139816e-06,
"loss": 0.4392,
"step": 7190
},
{
"epoch": 0.7360646101157768,
"grad_norm": 0.9471250176429749,
"learning_rate": 9.890613241644367e-06,
"loss": 0.4388,
"step": 7200
},
{
"epoch": 0.737086922074271,
"grad_norm": 1.39224374294281,
"learning_rate": 9.819628567180617e-06,
"loss": 0.4461,
"step": 7210
},
{
"epoch": 0.7381092340327651,
"grad_norm": 1.0654206275939941,
"learning_rate": 9.748837232077807e-06,
"loss": 0.447,
"step": 7220
},
{
"epoch": 0.7391315459912592,
"grad_norm": 0.9918525218963623,
"learning_rate": 9.6782401379444e-06,
"loss": 0.4475,
"step": 7230
},
{
"epoch": 0.7401538579497534,
"grad_norm": 1.02973210811615,
"learning_rate": 9.607838183914989e-06,
"loss": 0.444,
"step": 7240
},
{
"epoch": 0.7411761699082475,
"grad_norm": 0.9648279547691345,
"learning_rate": 9.537632266638829e-06,
"loss": 0.4447,
"step": 7250
},
{
"epoch": 0.7421984818667416,
"grad_norm": 1.2002679109573364,
"learning_rate": 9.467623280268432e-06,
"loss": 0.4358,
"step": 7260
},
{
"epoch": 0.7432207938252358,
"grad_norm": 1.0120762586593628,
"learning_rate": 9.39781211644816e-06,
"loss": 0.4471,
"step": 7270
},
{
"epoch": 0.7442431057837299,
"grad_norm": 0.9887851476669312,
"learning_rate": 9.32819966430288e-06,
"loss": 0.4443,
"step": 7280
},
{
"epoch": 0.745265417742224,
"grad_norm": 0.8384073972702026,
"learning_rate": 9.25878681042667e-06,
"loss": 0.4483,
"step": 7290
},
{
"epoch": 0.7462877297007182,
"grad_norm": 0.837547779083252,
"learning_rate": 9.189574438871443e-06,
"loss": 0.4433,
"step": 7300
},
{
"epoch": 0.7473100416592123,
"grad_norm": 0.8741981387138367,
"learning_rate": 9.120563431135765e-06,
"loss": 0.4477,
"step": 7310
},
{
"epoch": 0.7483323536177064,
"grad_norm": 1.1481738090515137,
"learning_rate": 9.051754666153625e-06,
"loss": 0.444,
"step": 7320
},
{
"epoch": 0.7493546655762006,
"grad_norm": 0.8956581354141235,
"learning_rate": 8.983149020283191e-06,
"loss": 0.4523,
"step": 7330
},
{
"epoch": 0.7503769775346947,
"grad_norm": 0.9955009818077087,
"learning_rate": 8.91474736729569e-06,
"loss": 0.4461,
"step": 7340
},
{
"epoch": 0.7513992894931888,
"grad_norm": 0.9001877903938293,
"learning_rate": 8.84655057836426e-06,
"loss": 0.4404,
"step": 7350
},
{
"epoch": 0.752421601451683,
"grad_norm": 1.1931997537612915,
"learning_rate": 8.77855952205287e-06,
"loss": 0.448,
"step": 7360
},
{
"epoch": 0.7534439134101771,
"grad_norm": 1.182490348815918,
"learning_rate": 8.71077506430524e-06,
"loss": 0.4424,
"step": 7370
},
{
"epoch": 0.7544662253686713,
"grad_norm": 0.8788588643074036,
"learning_rate": 8.643198068433828e-06,
"loss": 0.4543,
"step": 7380
},
{
"epoch": 0.7554885373271654,
"grad_norm": 1.0334864854812622,
"learning_rate": 8.575829395108826e-06,
"loss": 0.4319,
"step": 7390
},
{
"epoch": 0.7565108492856595,
"grad_norm": 0.9728019833564758,
"learning_rate": 8.508669902347196e-06,
"loss": 0.4363,
"step": 7400
},
{
"epoch": 0.7575331612441537,
"grad_norm": 1.0635533332824707,
"learning_rate": 8.44172044550175e-06,
"loss": 0.4415,
"step": 7410
},
{
"epoch": 0.7585554732026478,
"grad_norm": 0.9203910231590271,
"learning_rate": 8.374981877250277e-06,
"loss": 0.4368,
"step": 7420
},
{
"epoch": 0.7595777851611419,
"grad_norm": 0.9796839952468872,
"learning_rate": 8.308455047584607e-06,
"loss": 0.4337,
"step": 7430
},
{
"epoch": 0.760600097119636,
"grad_norm": 1.0226854085922241,
"learning_rate": 8.242140803799872e-06,
"loss": 0.4419,
"step": 7440
},
{
"epoch": 0.7616224090781302,
"grad_norm": 1.1608787775039673,
"learning_rate": 8.176039990483689e-06,
"loss": 0.4432,
"step": 7450
},
{
"epoch": 0.7626447210366243,
"grad_norm": 0.9388943314552307,
"learning_rate": 8.110153449505368e-06,
"loss": 0.4461,
"step": 7460
},
{
"epoch": 0.7636670329951185,
"grad_norm": 0.960391640663147,
"learning_rate": 8.044482020005234e-06,
"loss": 0.4395,
"step": 7470
},
{
"epoch": 0.7646893449536126,
"grad_norm": 0.999666154384613,
"learning_rate": 7.97902653838391e-06,
"loss": 0.4411,
"step": 7480
},
{
"epoch": 0.7657116569121067,
"grad_norm": 1.0976208448410034,
"learning_rate": 7.913787838291686e-06,
"loss": 0.4403,
"step": 7490
},
{
"epoch": 0.7667339688706009,
"grad_norm": 0.8546642065048218,
"learning_rate": 7.84876675061789e-06,
"loss": 0.4445,
"step": 7500
},
{
"epoch": 0.767756280829095,
"grad_norm": 0.9636361598968506,
"learning_rate": 7.783964103480296e-06,
"loss": 0.4372,
"step": 7510
},
{
"epoch": 0.7687785927875891,
"grad_norm": 0.8897413015365601,
"learning_rate": 7.719380722214605e-06,
"loss": 0.4366,
"step": 7520
},
{
"epoch": 0.7698009047460833,
"grad_norm": 1.1794391870498657,
"learning_rate": 7.655017429363908e-06,
"loss": 0.4435,
"step": 7530
},
{
"epoch": 0.7708232167045774,
"grad_norm": 1.3731043338775635,
"learning_rate": 7.590875044668213e-06,
"loss": 0.4425,
"step": 7540
},
{
"epoch": 0.7718455286630715,
"grad_norm": 0.8776447772979736,
"learning_rate": 7.526954385054031e-06,
"loss": 0.4358,
"step": 7550
},
{
"epoch": 0.7728678406215657,
"grad_norm": 1.0318937301635742,
"learning_rate": 7.463256264623939e-06,
"loss": 0.4425,
"step": 7560
},
{
"epoch": 0.7738901525800598,
"grad_norm": 0.9086455702781677,
"learning_rate": 7.399781494646196e-06,
"loss": 0.427,
"step": 7570
},
{
"epoch": 0.7749124645385539,
"grad_norm": 0.9383578300476074,
"learning_rate": 7.33653088354449e-06,
"loss": 0.4361,
"step": 7580
},
{
"epoch": 0.7759347764970481,
"grad_norm": 1.007272720336914,
"learning_rate": 7.273505236887554e-06,
"loss": 0.4231,
"step": 7590
},
{
"epoch": 0.7769570884555422,
"grad_norm": 1.226094126701355,
"learning_rate": 7.2107053573789525e-06,
"loss": 0.4318,
"step": 7600
},
{
"epoch": 0.7779794004140363,
"grad_norm": 1.049007534980774,
"learning_rate": 7.14813204484685e-06,
"loss": 0.4454,
"step": 7610
},
{
"epoch": 0.7790017123725305,
"grad_norm": 0.8623120784759521,
"learning_rate": 7.085786096233813e-06,
"loss": 0.445,
"step": 7620
},
{
"epoch": 0.7800240243310246,
"grad_norm": 1.2203346490859985,
"learning_rate": 7.023668305586678e-06,
"loss": 0.4402,
"step": 7630
},
{
"epoch": 0.7810463362895187,
"grad_norm": 0.8987168669700623,
"learning_rate": 6.961779464046417e-06,
"loss": 0.4368,
"step": 7640
},
{
"epoch": 0.7820686482480129,
"grad_norm": 0.9059882760047913,
"learning_rate": 6.900120359838108e-06,
"loss": 0.446,
"step": 7650
},
{
"epoch": 0.783090960206507,
"grad_norm": 0.852653443813324,
"learning_rate": 6.838691778260808e-06,
"loss": 0.4424,
"step": 7660
},
{
"epoch": 0.7841132721650012,
"grad_norm": 0.8900589346885681,
"learning_rate": 6.777494501677639e-06,
"loss": 0.4367,
"step": 7670
},
{
"epoch": 0.7851355841234953,
"grad_norm": 1.202610969543457,
"learning_rate": 6.716529309505801e-06,
"loss": 0.4338,
"step": 7680
},
{
"epoch": 0.7861578960819894,
"grad_norm": 1.2870848178863525,
"learning_rate": 6.6557969782066105e-06,
"loss": 0.4544,
"step": 7690
},
{
"epoch": 0.7871802080404836,
"grad_norm": 0.7765796780586243,
"learning_rate": 6.595298281275625e-06,
"loss": 0.4401,
"step": 7700
},
{
"epoch": 0.7882025199989777,
"grad_norm": 1.0064955949783325,
"learning_rate": 6.535033989232839e-06,
"loss": 0.4326,
"step": 7710
},
{
"epoch": 0.7892248319574718,
"grad_norm": 0.9474082589149475,
"learning_rate": 6.475004869612805e-06,
"loss": 0.4297,
"step": 7720
},
{
"epoch": 0.790247143915966,
"grad_norm": 0.9180557131767273,
"learning_rate": 6.415211686954894e-06,
"loss": 0.4372,
"step": 7730
},
{
"epoch": 0.7912694558744601,
"grad_norm": 1.0714077949523926,
"learning_rate": 6.35565520279355e-06,
"loss": 0.4336,
"step": 7740
},
{
"epoch": 0.7922917678329542,
"grad_norm": 1.0611761808395386,
"learning_rate": 6.296336175648592e-06,
"loss": 0.4323,
"step": 7750
},
{
"epoch": 0.7933140797914484,
"grad_norm": 1.010562539100647,
"learning_rate": 6.237255361015554e-06,
"loss": 0.436,
"step": 7760
},
{
"epoch": 0.7943363917499425,
"grad_norm": 1.0425492525100708,
"learning_rate": 6.1784135113560516e-06,
"loss": 0.4406,
"step": 7770
},
{
"epoch": 0.7953587037084366,
"grad_norm": 1.052968978881836,
"learning_rate": 6.119811376088239e-06,
"loss": 0.4428,
"step": 7780
},
{
"epoch": 0.7963810156669308,
"grad_norm": 0.8319458961486816,
"learning_rate": 6.06144970157719e-06,
"loss": 0.4366,
"step": 7790
},
{
"epoch": 0.7974033276254249,
"grad_norm": 1.107327938079834,
"learning_rate": 6.00332923112546e-06,
"loss": 0.4439,
"step": 7800
},
{
"epoch": 0.798425639583919,
"grad_norm": 1.1520315408706665,
"learning_rate": 5.945450704963604e-06,
"loss": 0.4455,
"step": 7810
},
{
"epoch": 0.7994479515424132,
"grad_norm": 1.370091199874878,
"learning_rate": 5.887814860240737e-06,
"loss": 0.4367,
"step": 7820
},
{
"epoch": 0.8004702635009073,
"grad_norm": 0.9729806780815125,
"learning_rate": 5.830422431015112e-06,
"loss": 0.4336,
"step": 7830
},
{
"epoch": 0.8014925754594014,
"grad_norm": 0.9457160830497742,
"learning_rate": 5.773274148244867e-06,
"loss": 0.4456,
"step": 7840
},
{
"epoch": 0.8025148874178956,
"grad_norm": 1.0155694484710693,
"learning_rate": 5.716370739778612e-06,
"loss": 0.4339,
"step": 7850
},
{
"epoch": 0.8035371993763897,
"grad_norm": 0.9733458161354065,
"learning_rate": 5.659712930346225e-06,
"loss": 0.4407,
"step": 7860
},
{
"epoch": 0.8045595113348838,
"grad_norm": 0.9396880269050598,
"learning_rate": 5.603301441549585e-06,
"loss": 0.4381,
"step": 7870
},
{
"epoch": 0.805581823293378,
"grad_norm": 0.9309740662574768,
"learning_rate": 5.5471369918534135e-06,
"loss": 0.4426,
"step": 7880
},
{
"epoch": 0.8066041352518721,
"grad_norm": 0.8893073797225952,
"learning_rate": 5.491220296576088e-06,
"loss": 0.4337,
"step": 7890
},
{
"epoch": 0.8076264472103662,
"grad_norm": 1.0726497173309326,
"learning_rate": 5.435552067880562e-06,
"loss": 0.4348,
"step": 7900
},
{
"epoch": 0.8086487591688604,
"grad_norm": 0.9764810800552368,
"learning_rate": 5.380133014765296e-06,
"loss": 0.4395,
"step": 7910
},
{
"epoch": 0.8096710711273545,
"grad_norm": 0.7991422414779663,
"learning_rate": 5.3249638430551834e-06,
"loss": 0.4309,
"step": 7920
},
{
"epoch": 0.8106933830858486,
"grad_norm": 0.9090506434440613,
"learning_rate": 5.270045255392608e-06,
"loss": 0.4348,
"step": 7930
},
{
"epoch": 0.8117156950443428,
"grad_norm": 0.9037578105926514,
"learning_rate": 5.215377951228501e-06,
"loss": 0.4333,
"step": 7940
},
{
"epoch": 0.8127380070028369,
"grad_norm": 1.0673049688339233,
"learning_rate": 5.160962626813379e-06,
"loss": 0.4349,
"step": 7950
},
{
"epoch": 0.813760318961331,
"grad_norm": 0.9314389824867249,
"learning_rate": 5.106799975188523e-06,
"loss": 0.4346,
"step": 7960
},
{
"epoch": 0.8147826309198252,
"grad_norm": 0.9233517646789551,
"learning_rate": 5.052890686177139e-06,
"loss": 0.4281,
"step": 7970
},
{
"epoch": 0.8158049428783193,
"grad_norm": 0.8081022500991821,
"learning_rate": 4.999235446375566e-06,
"loss": 0.4348,
"step": 7980
},
{
"epoch": 0.8168272548368134,
"grad_norm": 1.0998950004577637,
"learning_rate": 4.945834939144539e-06,
"loss": 0.4337,
"step": 7990
},
{
"epoch": 0.8178495667953076,
"grad_norm": 0.9865671992301941,
"learning_rate": 4.892689844600484e-06,
"loss": 0.4337,
"step": 8000
},
{
"epoch": 0.8188718787538017,
"grad_norm": 0.892192542552948,
"learning_rate": 4.839800839606851e-06,
"loss": 0.4304,
"step": 8010
},
{
"epoch": 0.8198941907122959,
"grad_norm": 1.193628191947937,
"learning_rate": 4.787168597765501e-06,
"loss": 0.4395,
"step": 8020
},
{
"epoch": 0.82091650267079,
"grad_norm": 0.777214527130127,
"learning_rate": 4.734793789408115e-06,
"loss": 0.4314,
"step": 8030
},
{
"epoch": 0.8219388146292841,
"grad_norm": 0.8342297673225403,
"learning_rate": 4.682677081587691e-06,
"loss": 0.4376,
"step": 8040
},
{
"epoch": 0.8229611265877783,
"grad_norm": 1.061095118522644,
"learning_rate": 4.630819138069983e-06,
"loss": 0.4407,
"step": 8050
},
{
"epoch": 0.8239834385462724,
"grad_norm": 0.9414345026016235,
"learning_rate": 4.579220619325106e-06,
"loss": 0.4277,
"step": 8060
},
{
"epoch": 0.8250057505047665,
"grad_norm": 0.7797683477401733,
"learning_rate": 4.527882182519114e-06,
"loss": 0.4348,
"step": 8070
},
{
"epoch": 0.8260280624632607,
"grad_norm": 0.946404218673706,
"learning_rate": 4.476804481505609e-06,
"loss": 0.4277,
"step": 8080
},
{
"epoch": 0.8270503744217548,
"grad_norm": 0.9743029475212097,
"learning_rate": 4.425988166817422e-06,
"loss": 0.4353,
"step": 8090
},
{
"epoch": 0.8280726863802489,
"grad_norm": 1.0373250246047974,
"learning_rate": 4.3754338856583365e-06,
"loss": 0.4304,
"step": 8100
},
{
"epoch": 0.8290949983387431,
"grad_norm": 0.7803084850311279,
"learning_rate": 4.325142281894837e-06,
"loss": 0.4357,
"step": 8110
},
{
"epoch": 0.8301173102972372,
"grad_norm": 0.7771669030189514,
"learning_rate": 4.275113996047916e-06,
"loss": 0.4332,
"step": 8120
},
{
"epoch": 0.8311396222557313,
"grad_norm": 0.8627368807792664,
"learning_rate": 4.225349665284908e-06,
"loss": 0.4328,
"step": 8130
},
{
"epoch": 0.8321619342142255,
"grad_norm": 0.8736832737922668,
"learning_rate": 4.175849923411379e-06,
"loss": 0.4261,
"step": 8140
},
{
"epoch": 0.8331842461727196,
"grad_norm": 0.950605034828186,
"learning_rate": 4.126615400863054e-06,
"loss": 0.4273,
"step": 8150
},
{
"epoch": 0.8342065581312137,
"grad_norm": 1.214670181274414,
"learning_rate": 4.077646724697778e-06,
"loss": 0.4274,
"step": 8160
},
{
"epoch": 0.8352288700897079,
"grad_norm": 0.8484399318695068,
"learning_rate": 4.02894451858756e-06,
"loss": 0.4321,
"step": 8170
},
{
"epoch": 0.836251182048202,
"grad_norm": 0.8508866429328918,
"learning_rate": 3.980509402810598e-06,
"loss": 0.4328,
"step": 8180
},
{
"epoch": 0.8372734940066962,
"grad_norm": 1.059467077255249,
"learning_rate": 3.932341994243366e-06,
"loss": 0.4277,
"step": 8190
},
{
"epoch": 0.8382958059651903,
"grad_norm": 1.0109869241714478,
"learning_rate": 3.884442906352826e-06,
"loss": 0.4329,
"step": 8200
},
{
"epoch": 0.8393181179236844,
"grad_norm": 1.0309584140777588,
"learning_rate": 3.836812749188534e-06,
"loss": 0.4298,
"step": 8210
},
{
"epoch": 0.8403404298821785,
"grad_norm": 0.8743014335632324,
"learning_rate": 3.7894521293749207e-06,
"loss": 0.4275,
"step": 8220
},
{
"epoch": 0.8413627418406727,
"grad_norm": 0.9778170585632324,
"learning_rate": 3.742361650103554e-06,
"loss": 0.4299,
"step": 8230
},
{
"epoch": 0.8423850537991668,
"grad_norm": 0.8993359208106995,
"learning_rate": 3.6955419111254415e-06,
"loss": 0.4321,
"step": 8240
},
{
"epoch": 0.843407365757661,
"grad_norm": 1.06406831741333,
"learning_rate": 3.6489935087434144e-06,
"loss": 0.4261,
"step": 8250
},
{
"epoch": 0.8444296777161551,
"grad_norm": 1.0659277439117432,
"learning_rate": 3.6027170358045148e-06,
"loss": 0.4299,
"step": 8260
},
{
"epoch": 0.8454519896746492,
"grad_norm": 1.143269658088684,
"learning_rate": 3.55671308169247e-06,
"loss": 0.4293,
"step": 8270
},
{
"epoch": 0.8464743016331433,
"grad_norm": 1.0796489715576172,
"learning_rate": 3.5109822323201423e-06,
"loss": 0.4264,
"step": 8280
},
{
"epoch": 0.8474966135916375,
"grad_norm": 1.0255521535873413,
"learning_rate": 3.465525070122108e-06,
"loss": 0.4258,
"step": 8290
},
{
"epoch": 0.8485189255501316,
"grad_norm": 0.9556827545166016,
"learning_rate": 3.4203421740472365e-06,
"loss": 0.4283,
"step": 8300
},
{
"epoch": 0.8495412375086258,
"grad_norm": 1.1103256940841675,
"learning_rate": 3.375434119551285e-06,
"loss": 0.4335,
"step": 8310
},
{
"epoch": 0.8505635494671199,
"grad_norm": 1.1893926858901978,
"learning_rate": 3.3308014785895904e-06,
"loss": 0.4377,
"step": 8320
},
{
"epoch": 0.851585861425614,
"grad_norm": 0.8934152126312256,
"learning_rate": 3.286444819609799e-06,
"loss": 0.4328,
"step": 8330
},
{
"epoch": 0.8526081733841082,
"grad_norm": 0.9043262004852295,
"learning_rate": 3.242364707544601e-06,
"loss": 0.4251,
"step": 8340
},
{
"epoch": 0.8536304853426023,
"grad_norm": 1.113665223121643,
"learning_rate": 3.1985617038045477e-06,
"loss": 0.4249,
"step": 8350
},
{
"epoch": 0.8546527973010964,
"grad_norm": 0.8989176154136658,
"learning_rate": 3.1550363662708996e-06,
"loss": 0.4221,
"step": 8360
},
{
"epoch": 0.8556751092595906,
"grad_norm": 0.9676817655563354,
"learning_rate": 3.1117892492885188e-06,
"loss": 0.4308,
"step": 8370
},
{
"epoch": 0.8566974212180847,
"grad_norm": 0.6855757236480713,
"learning_rate": 3.0688209036588174e-06,
"loss": 0.433,
"step": 8380
},
{
"epoch": 0.8577197331765788,
"grad_norm": 0.9903734922409058,
"learning_rate": 3.0261318766327206e-06,
"loss": 0.4281,
"step": 8390
},
{
"epoch": 0.858742045135073,
"grad_norm": 0.7935535311698914,
"learning_rate": 2.983722711903747e-06,
"loss": 0.4329,
"step": 8400
},
{
"epoch": 0.8597643570935671,
"grad_norm": 0.9941035509109497,
"learning_rate": 2.941593949601007e-06,
"loss": 0.4312,
"step": 8410
},
{
"epoch": 0.8607866690520612,
"grad_norm": 0.8986974358558655,
"learning_rate": 2.8997461262823823e-06,
"loss": 0.4324,
"step": 8420
},
{
"epoch": 0.8618089810105554,
"grad_norm": 0.9105693101882935,
"learning_rate": 2.8581797749276914e-06,
"loss": 0.4232,
"step": 8430
},
{
"epoch": 0.8628312929690495,
"grad_norm": 0.957203209400177,
"learning_rate": 2.816895424931873e-06,
"loss": 0.4205,
"step": 8440
},
{
"epoch": 0.8638536049275436,
"grad_norm": 1.2268370389938354,
"learning_rate": 2.7758936020982425e-06,
"loss": 0.4235,
"step": 8450
},
{
"epoch": 0.8648759168860378,
"grad_norm": 0.7346043586730957,
"learning_rate": 2.735174828631837e-06,
"loss": 0.4228,
"step": 8460
},
{
"epoch": 0.8658982288445319,
"grad_norm": 0.9053030014038086,
"learning_rate": 2.694739623132722e-06,
"loss": 0.4356,
"step": 8470
},
{
"epoch": 0.866920540803026,
"grad_norm": 0.9548618197441101,
"learning_rate": 2.654588500589397e-06,
"loss": 0.4222,
"step": 8480
},
{
"epoch": 0.8679428527615202,
"grad_norm": 1.3412740230560303,
"learning_rate": 2.614721972372247e-06,
"loss": 0.4262,
"step": 8490
},
{
"epoch": 0.8689651647200143,
"grad_norm": 0.7086278200149536,
"learning_rate": 2.5751405462270286e-06,
"loss": 0.4207,
"step": 8500
},
{
"epoch": 0.8699874766785084,
"grad_norm": 1.1251598596572876,
"learning_rate": 2.5358447262683855e-06,
"loss": 0.4348,
"step": 8510
},
{
"epoch": 0.8710097886370026,
"grad_norm": 0.9338446855545044,
"learning_rate": 2.4968350129734445e-06,
"loss": 0.4247,
"step": 8520
},
{
"epoch": 0.8720321005954967,
"grad_norm": 0.7428572773933411,
"learning_rate": 2.4581119031754556e-06,
"loss": 0.4289,
"step": 8530
},
{
"epoch": 0.8730544125539909,
"grad_norm": 0.9661202430725098,
"learning_rate": 2.4196758900574156e-06,
"loss": 0.4205,
"step": 8540
},
{
"epoch": 0.874076724512485,
"grad_norm": 0.8998513221740723,
"learning_rate": 2.381527463145833e-06,
"loss": 0.4244,
"step": 8550
},
{
"epoch": 0.8750990364709791,
"grad_norm": 1.2801803350448608,
"learning_rate": 2.343667108304484e-06,
"loss": 0.425,
"step": 8560
},
{
"epoch": 0.8761213484294732,
"grad_norm": 0.9866209030151367,
"learning_rate": 2.3060953077282068e-06,
"loss": 0.4197,
"step": 8570
},
{
"epoch": 0.8771436603879674,
"grad_norm": 0.8322245478630066,
"learning_rate": 2.2688125399367766e-06,
"loss": 0.4235,
"step": 8580
},
{
"epoch": 0.8781659723464615,
"grad_norm": 1.0312855243682861,
"learning_rate": 2.231819279768807e-06,
"loss": 0.4269,
"step": 8590
},
{
"epoch": 0.8791882843049557,
"grad_norm": 0.8428605794906616,
"learning_rate": 2.1951159983756936e-06,
"loss": 0.4313,
"step": 8600
},
{
"epoch": 0.8802105962634498,
"grad_norm": 0.9650649428367615,
"learning_rate": 2.1587031632156363e-06,
"loss": 0.4298,
"step": 8610
},
{
"epoch": 0.8812329082219439,
"grad_norm": 1.0367612838745117,
"learning_rate": 2.122581238047652e-06,
"loss": 0.4324,
"step": 8620
},
{
"epoch": 0.8822552201804381,
"grad_norm": 0.8398295640945435,
"learning_rate": 2.086750682925706e-06,
"loss": 0.4173,
"step": 8630
},
{
"epoch": 0.8832775321389322,
"grad_norm": 0.7944861650466919,
"learning_rate": 2.0512119541928164e-06,
"loss": 0.432,
"step": 8640
},
{
"epoch": 0.8842998440974263,
"grad_norm": 0.9605563282966614,
"learning_rate": 2.0159655044752673e-06,
"loss": 0.4269,
"step": 8650
},
{
"epoch": 0.8853221560559205,
"grad_norm": 0.9588894844055176,
"learning_rate": 1.981011782676842e-06,
"loss": 0.4364,
"step": 8660
},
{
"epoch": 0.8863444680144146,
"grad_norm": 0.7828480005264282,
"learning_rate": 1.9463512339730873e-06,
"loss": 0.425,
"step": 8670
},
{
"epoch": 0.8873667799729087,
"grad_norm": 1.0239571332931519,
"learning_rate": 1.911984299805658e-06,
"loss": 0.4306,
"step": 8680
},
{
"epoch": 0.8883890919314029,
"grad_norm": 1.0380973815917969,
"learning_rate": 1.8779114178767016e-06,
"loss": 0.4274,
"step": 8690
},
{
"epoch": 0.889411403889897,
"grad_norm": 0.9591466188430786,
"learning_rate": 1.8441330221432684e-06,
"loss": 0.429,
"step": 8700
},
{
"epoch": 0.8904337158483911,
"grad_norm": 0.9937755465507507,
"learning_rate": 1.8106495428117902e-06,
"loss": 0.4288,
"step": 8710
},
{
"epoch": 0.8914560278068853,
"grad_norm": 1.000308871269226,
"learning_rate": 1.7774614063326005e-06,
"loss": 0.4236,
"step": 8720
},
{
"epoch": 0.8924783397653794,
"grad_norm": 1.0950754880905151,
"learning_rate": 1.7445690353945109e-06,
"loss": 0.424,
"step": 8730
},
{
"epoch": 0.8935006517238735,
"grad_norm": 1.0174835920333862,
"learning_rate": 1.711972848919416e-06,
"loss": 0.4308,
"step": 8740
},
{
"epoch": 0.8945229636823677,
"grad_norm": 0.7807797193527222,
"learning_rate": 1.6796732620569606e-06,
"loss": 0.4352,
"step": 8750
},
{
"epoch": 0.8955452756408618,
"grad_norm": 1.0470595359802246,
"learning_rate": 1.6476706861792617e-06,
"loss": 0.4361,
"step": 8760
},
{
"epoch": 0.896567587599356,
"grad_norm": 0.9675589203834534,
"learning_rate": 1.6159655288756599e-06,
"loss": 0.4223,
"step": 8770
},
{
"epoch": 0.8975898995578501,
"grad_norm": 1.1054085493087769,
"learning_rate": 1.5845581939475219e-06,
"loss": 0.4268,
"step": 8780
},
{
"epoch": 0.8986122115163442,
"grad_norm": 0.9452238082885742,
"learning_rate": 1.5534490814031245e-06,
"loss": 0.4239,
"step": 8790
},
{
"epoch": 0.8996345234748383,
"grad_norm": 0.9669876098632812,
"learning_rate": 1.5226385874525324e-06,
"loss": 0.4348,
"step": 8800
},
{
"epoch": 0.9006568354333325,
"grad_norm": 1.193709135055542,
"learning_rate": 1.4921271045025499e-06,
"loss": 0.425,
"step": 8810
},
{
"epoch": 0.9016791473918266,
"grad_norm": 0.8553543090820312,
"learning_rate": 1.4619150211517552e-06,
"loss": 0.417,
"step": 8820
},
{
"epoch": 0.9027014593503208,
"grad_norm": 1.1650067567825317,
"learning_rate": 1.4320027221855186e-06,
"loss": 0.4177,
"step": 8830
},
{
"epoch": 0.9037237713088149,
"grad_norm": 1.290274977684021,
"learning_rate": 1.4023905885711148e-06,
"loss": 0.4288,
"step": 8840
},
{
"epoch": 0.904746083267309,
"grad_norm": 0.9648873209953308,
"learning_rate": 1.3730789974528784e-06,
"loss": 0.4305,
"step": 8850
},
{
"epoch": 0.9057683952258031,
"grad_norm": 1.0339425802230835,
"learning_rate": 1.3440683221473793e-06,
"loss": 0.4254,
"step": 8860
},
{
"epoch": 0.9067907071842973,
"grad_norm": 0.9494819641113281,
"learning_rate": 1.315358932138691e-06,
"loss": 0.4136,
"step": 8870
},
{
"epoch": 0.9078130191427914,
"grad_norm": 0.8865864276885986,
"learning_rate": 1.2869511930736684e-06,
"loss": 0.4185,
"step": 8880
},
{
"epoch": 0.9088353311012856,
"grad_norm": 0.957786500453949,
"learning_rate": 1.258845466757308e-06,
"loss": 0.4315,
"step": 8890
},
{
"epoch": 0.9098576430597797,
"grad_norm": 0.9549824595451355,
"learning_rate": 1.2310421111481146e-06,
"loss": 0.4299,
"step": 8900
},
{
"epoch": 0.9108799550182738,
"grad_norm": 1.1362382173538208,
"learning_rate": 1.2035414803535623e-06,
"loss": 0.4254,
"step": 8910
},
{
"epoch": 0.9119022669767679,
"grad_norm": 0.8010842204093933,
"learning_rate": 1.1763439246255835e-06,
"loss": 0.4288,
"step": 8920
},
{
"epoch": 0.9129245789352621,
"grad_norm": 0.7839857935905457,
"learning_rate": 1.149449790356108e-06,
"loss": 0.4292,
"step": 8930
},
{
"epoch": 0.9139468908937562,
"grad_norm": 1.1403322219848633,
"learning_rate": 1.1228594200726223e-06,
"loss": 0.4277,
"step": 8940
},
{
"epoch": 0.9149692028522504,
"grad_norm": 1.3120801448822021,
"learning_rate": 1.0965731524338634e-06,
"loss": 0.4266,
"step": 8950
},
{
"epoch": 0.9159915148107445,
"grad_norm": 0.8597049117088318,
"learning_rate": 1.070591322225456e-06,
"loss": 0.4202,
"step": 8960
},
{
"epoch": 0.9170138267692386,
"grad_norm": 0.9351035952568054,
"learning_rate": 1.0449142603556672e-06,
"loss": 0.4298,
"step": 8970
},
{
"epoch": 0.9180361387277328,
"grad_norm": 0.926821231842041,
"learning_rate": 1.0195422938511973e-06,
"loss": 0.4312,
"step": 8980
},
{
"epoch": 0.9190584506862269,
"grad_norm": 0.8255594968795776,
"learning_rate": 9.944757458530047e-07,
"loss": 0.4267,
"step": 8990
},
{
"epoch": 0.920080762644721,
"grad_norm": 1.046819806098938,
"learning_rate": 9.697149356121927e-07,
"loss": 0.4255,
"step": 9000
},
{
"epoch": 0.9211030746032152,
"grad_norm": 0.7672698497772217,
"learning_rate": 9.452601784859516e-07,
"loss": 0.4264,
"step": 9010
},
{
"epoch": 0.9221253865617093,
"grad_norm": 0.9611999988555908,
"learning_rate": 9.211117859335344e-07,
"loss": 0.4171,
"step": 9020
},
{
"epoch": 0.9231476985202034,
"grad_norm": 1.0887755155563354,
"learning_rate": 8.972700655122817e-07,
"loss": 0.421,
"step": 9030
},
{
"epoch": 0.9241700104786976,
"grad_norm": 1.2452219724655151,
"learning_rate": 8.737353208737281e-07,
"loss": 0.4255,
"step": 9040
},
{
"epoch": 0.9251923224371917,
"grad_norm": 1.0280874967575073,
"learning_rate": 8.505078517597131e-07,
"loss": 0.4126,
"step": 9050
},
{
"epoch": 0.9262146343956859,
"grad_norm": 0.8932361006736755,
"learning_rate": 8.275879539985765e-07,
"loss": 0.4249,
"step": 9060
},
{
"epoch": 0.92723694635418,
"grad_norm": 0.9874764680862427,
"learning_rate": 8.049759195013773e-07,
"loss": 0.426,
"step": 9070
},
{
"epoch": 0.9282592583126741,
"grad_norm": 0.9187189340591431,
"learning_rate": 7.826720362581946e-07,
"loss": 0.4249,
"step": 9080
},
{
"epoch": 0.9292815702711682,
"grad_norm": 1.008799433708191,
"learning_rate": 7.606765883344441e-07,
"loss": 0.4296,
"step": 9090
},
{
"epoch": 0.9303038822296624,
"grad_norm": 0.9497778415679932,
"learning_rate": 7.389898558672642e-07,
"loss": 0.4227,
"step": 9100
},
{
"epoch": 0.9313261941881565,
"grad_norm": 1.093139410018921,
"learning_rate": 7.176121150619524e-07,
"loss": 0.4226,
"step": 9110
},
{
"epoch": 0.9323485061466507,
"grad_norm": 0.8269488215446472,
"learning_rate": 6.96543638188446e-07,
"loss": 0.419,
"step": 9120
},
{
"epoch": 0.9333708181051448,
"grad_norm": 0.8830663561820984,
"learning_rate": 6.757846935778495e-07,
"loss": 0.4142,
"step": 9130
},
{
"epoch": 0.9343931300636389,
"grad_norm": 1.0776466131210327,
"learning_rate": 6.55335545619018e-07,
"loss": 0.4336,
"step": 9140
},
{
"epoch": 0.935415442022133,
"grad_norm": 0.7703368663787842,
"learning_rate": 6.351964547552048e-07,
"loss": 0.4285,
"step": 9150
},
{
"epoch": 0.9364377539806272,
"grad_norm": 0.7766233682632446,
"learning_rate": 6.153676774807188e-07,
"loss": 0.4318,
"step": 9160
},
{
"epoch": 0.9374600659391213,
"grad_norm": 0.8788174390792847,
"learning_rate": 5.958494663376751e-07,
"loss": 0.4197,
"step": 9170
},
{
"epoch": 0.9384823778976155,
"grad_norm": 0.8231105208396912,
"learning_rate": 5.766420699127778e-07,
"loss": 0.4218,
"step": 9180
},
{
"epoch": 0.9395046898561096,
"grad_norm": 0.8786907196044922,
"learning_rate": 5.577457328341529e-07,
"loss": 0.4295,
"step": 9190
},
{
"epoch": 0.9405270018146037,
"grad_norm": 1.1679691076278687,
"learning_rate": 5.391606957682149e-07,
"loss": 0.4245,
"step": 9200
},
{
"epoch": 0.9415493137730978,
"grad_norm": 0.9138793349266052,
"learning_rate": 5.208871954166389e-07,
"loss": 0.4265,
"step": 9210
},
{
"epoch": 0.942571625731592,
"grad_norm": 1.060909628868103,
"learning_rate": 5.029254645133097e-07,
"loss": 0.4272,
"step": 9220
},
{
"epoch": 0.9435939376900861,
"grad_norm": 0.9517412185668945,
"learning_rate": 4.852757318213775e-07,
"loss": 0.4172,
"step": 9230
},
{
"epoch": 0.9446162496485803,
"grad_norm": 1.0084301233291626,
"learning_rate": 4.6793822213033466e-07,
"loss": 0.4246,
"step": 9240
},
{
"epoch": 0.9456385616070744,
"grad_norm": 1.66988205909729,
"learning_rate": 4.50913156253166e-07,
"loss": 0.4227,
"step": 9250
},
{
"epoch": 0.9466608735655685,
"grad_norm": 1.1668732166290283,
"learning_rate": 4.3420075102351965e-07,
"loss": 0.4259,
"step": 9260
},
{
"epoch": 0.9476831855240627,
"grad_norm": 1.0249663591384888,
"learning_rate": 4.1780121929295426e-07,
"loss": 0.4292,
"step": 9270
},
{
"epoch": 0.9487054974825568,
"grad_norm": 0.9237465262413025,
"learning_rate": 4.0171476992823255e-07,
"loss": 0.4256,
"step": 9280
},
{
"epoch": 0.949727809441051,
"grad_norm": 0.9204666018486023,
"learning_rate": 3.859416078086486e-07,
"loss": 0.4196,
"step": 9290
},
{
"epoch": 0.9507501213995451,
"grad_norm": 1.0454548597335815,
"learning_rate": 3.7048193382343e-07,
"loss": 0.4198,
"step": 9300
},
{
"epoch": 0.9517724333580392,
"grad_norm": 0.8368372321128845,
"learning_rate": 3.553359448691784e-07,
"loss": 0.4252,
"step": 9310
},
{
"epoch": 0.9527947453165333,
"grad_norm": 0.7651166319847107,
"learning_rate": 3.4050383384735264e-07,
"loss": 0.4258,
"step": 9320
},
{
"epoch": 0.9538170572750275,
"grad_norm": 0.7334020733833313,
"learning_rate": 3.259857896618174e-07,
"loss": 0.4206,
"step": 9330
},
{
"epoch": 0.9548393692335216,
"grad_norm": 0.9052079319953918,
"learning_rate": 3.1178199721644264e-07,
"loss": 0.4173,
"step": 9340
},
{
"epoch": 0.9558616811920158,
"grad_norm": 0.8853691220283508,
"learning_rate": 2.978926374127389e-07,
"loss": 0.4191,
"step": 9350
},
{
"epoch": 0.9568839931505099,
"grad_norm": 1.0781968832015991,
"learning_rate": 2.8431788714756437e-07,
"loss": 0.4348,
"step": 9360
},
{
"epoch": 0.957906305109004,
"grad_norm": 0.8945118188858032,
"learning_rate": 2.7105791931085475e-07,
"loss": 0.4265,
"step": 9370
},
{
"epoch": 0.9589286170674981,
"grad_norm": 0.8317992091178894,
"learning_rate": 2.5811290278344446e-07,
"loss": 0.4238,
"step": 9380
},
{
"epoch": 0.9599509290259923,
"grad_norm": 0.986292839050293,
"learning_rate": 2.4548300243489306e-07,
"loss": 0.4212,
"step": 9390
},
{
"epoch": 0.9609732409844864,
"grad_norm": 1.1332231760025024,
"learning_rate": 2.33168379121404e-07,
"loss": 0.4183,
"step": 9400
},
{
"epoch": 0.9619955529429806,
"grad_norm": 0.7695975303649902,
"learning_rate": 2.2116918968376476e-07,
"loss": 0.4272,
"step": 9410
},
{
"epoch": 0.9630178649014747,
"grad_norm": 0.8533865809440613,
"learning_rate": 2.094855869453488e-07,
"loss": 0.4263,
"step": 9420
},
{
"epoch": 0.9640401768599688,
"grad_norm": 1.051630973815918,
"learning_rate": 1.9811771971017523e-07,
"loss": 0.4275,
"step": 9430
},
{
"epoch": 0.9650624888184629,
"grad_norm": 1.0363585948944092,
"learning_rate": 1.870657327610159e-07,
"loss": 0.4189,
"step": 9440
},
{
"epoch": 0.9660848007769571,
"grad_norm": 0.9279293417930603,
"learning_rate": 1.7632976685753866e-07,
"loss": 0.4267,
"step": 9450
},
{
"epoch": 0.9671071127354512,
"grad_norm": 0.8801166415214539,
"learning_rate": 1.659099587345281e-07,
"loss": 0.4319,
"step": 9460
},
{
"epoch": 0.9681294246939454,
"grad_norm": 0.772002100944519,
"learning_rate": 1.558064411001342e-07,
"loss": 0.4276,
"step": 9470
},
{
"epoch": 0.9691517366524395,
"grad_norm": 1.1025477647781372,
"learning_rate": 1.460193426341877e-07,
"loss": 0.4208,
"step": 9480
},
{
"epoch": 0.9701740486109336,
"grad_norm": 0.9299958348274231,
"learning_rate": 1.3654878798656512e-07,
"loss": 0.4242,
"step": 9490
},
{
"epoch": 0.9711963605694277,
"grad_norm": 0.9939782023429871,
"learning_rate": 1.273948977755901e-07,
"loss": 0.423,
"step": 9500
},
{
"epoch": 0.9722186725279219,
"grad_norm": 0.9094900488853455,
"learning_rate": 1.1855778858650412e-07,
"loss": 0.4239,
"step": 9510
},
{
"epoch": 0.973240984486416,
"grad_norm": 0.763153076171875,
"learning_rate": 1.100375729699843e-07,
"loss": 0.4186,
"step": 9520
},
{
"epoch": 0.9742632964449102,
"grad_norm": 1.234564185142517,
"learning_rate": 1.0183435944070007e-07,
"loss": 0.4298,
"step": 9530
},
{
"epoch": 0.9752856084034043,
"grad_norm": 1.002125859260559,
"learning_rate": 9.394825247594219e-08,
"loss": 0.4281,
"step": 9540
},
{
"epoch": 0.9763079203618984,
"grad_norm": 0.8858130574226379,
"learning_rate": 8.637935251428753e-08,
"loss": 0.4274,
"step": 9550
},
{
"epoch": 0.9773302323203926,
"grad_norm": 0.9094356894493103,
"learning_rate": 7.912775595431687e-08,
"loss": 0.4235,
"step": 9560
},
{
"epoch": 0.9783525442788867,
"grad_norm": 1.432202696800232,
"learning_rate": 7.219355515339088e-08,
"loss": 0.4255,
"step": 9570
},
{
"epoch": 0.9793748562373809,
"grad_norm": 0.8593534231185913,
"learning_rate": 6.5576838426476e-08,
"loss": 0.4216,
"step": 9580
},
{
"epoch": 0.980397168195875,
"grad_norm": 0.863518238067627,
"learning_rate": 5.927769004501205e-08,
"loss": 0.4244,
"step": 9590
},
{
"epoch": 0.9814194801543691,
"grad_norm": 0.7316293120384216,
"learning_rate": 5.329619023584087e-08,
"loss": 0.4294,
"step": 9600
},
{
"epoch": 0.9824417921128632,
"grad_norm": 1.1563441753387451,
"learning_rate": 4.7632415180198784e-08,
"loss": 0.4231,
"step": 9610
},
{
"epoch": 0.9834641040713574,
"grad_norm": 1.1757203340530396,
"learning_rate": 4.228643701272017e-08,
"loss": 0.421,
"step": 9620
},
{
"epoch": 0.9844864160298515,
"grad_norm": 0.8675976991653442,
"learning_rate": 3.72583238205465e-08,
"loss": 0.4253,
"step": 9630
},
{
"epoch": 0.9855087279883457,
"grad_norm": 0.9779329895973206,
"learning_rate": 3.254813964243542e-08,
"loss": 0.4233,
"step": 9640
},
{
"epoch": 0.9865310399468398,
"grad_norm": 0.9829447865486145,
"learning_rate": 2.8155944467961327e-08,
"loss": 0.4286,
"step": 9650
},
{
"epoch": 0.9875533519053339,
"grad_norm": 0.9288956522941589,
"learning_rate": 2.4081794236743837e-08,
"loss": 0.4212,
"step": 9660
},
{
"epoch": 0.988575663863828,
"grad_norm": 0.9417626857757568,
"learning_rate": 2.032574083773997e-08,
"loss": 0.4253,
"step": 9670
},
{
"epoch": 0.9895979758223222,
"grad_norm": 1.046013593673706,
"learning_rate": 1.6887832108572476e-08,
"loss": 0.4254,
"step": 9680
},
{
"epoch": 0.9906202877808163,
"grad_norm": 1.179836630821228,
"learning_rate": 1.3768111834938645e-08,
"loss": 0.4226,
"step": 9690
},
{
"epoch": 0.9916425997393105,
"grad_norm": 1.0414127111434937,
"learning_rate": 1.0966619750035768e-08,
"loss": 0.4315,
"step": 9700
},
{
"epoch": 0.9926649116978046,
"grad_norm": 1.0224956274032593,
"learning_rate": 8.4833915340643e-09,
"loss": 0.4177,
"step": 9710
},
{
"epoch": 0.9936872236562987,
"grad_norm": 0.9705348610877991,
"learning_rate": 6.318458813769912e-09,
"loss": 0.4257,
"step": 9720
},
{
"epoch": 0.9947095356147928,
"grad_norm": 0.8701043725013733,
"learning_rate": 4.471849162043795e-09,
"loss": 0.4293,
"step": 9730
},
{
"epoch": 0.995731847573287,
"grad_norm": 0.9006491899490356,
"learning_rate": 2.943586097567397e-09,
"loss": 0.4272,
"step": 9740
},
{
"epoch": 0.9967541595317811,
"grad_norm": 0.8509365916252136,
"learning_rate": 1.733689084515433e-09,
"loss": 0.4224,
"step": 9750
},
{
"epoch": 0.9977764714902753,
"grad_norm": 0.8750006556510925,
"learning_rate": 8.421735323060897e-10,
"loss": 0.4265,
"step": 9760
},
{
"epoch": 0.9987987834487694,
"grad_norm": 0.8954494595527649,
"learning_rate": 2.6905079540950894e-10,
"loss": 0.4271,
"step": 9770
},
{
"epoch": 0.9998210954072635,
"grad_norm": 1.3059301376342773,
"learning_rate": 1.4328173200683915e-11,
"loss": 0.4183,
"step": 9780
},
{
"epoch": 1.0,
"step": 9782,
"total_flos": 1.7136144572775137e+19,
"train_loss": 0.7275813773740559,
"train_runtime": 21314.9758,
"train_samples_per_second": 58.741,
"train_steps_per_second": 0.459
}
],
"logging_steps": 10,
"max_steps": 9782,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 3000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.7136144572775137e+19,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}