diff --git "a/adapters/checkpoint-6000/trainer_state.json" "b/adapters/checkpoint-6000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/adapters/checkpoint-6000/trainer_state.json"
@@ -0,0 +1,4330 @@
+{
+  "best_global_step": 6000,
+  "best_metric": 0.6719915270805359,
+  "best_model_checkpoint": "/workspace/rails-finetune/adapters-qwen3-coder-30b/checkpoint-6000",
+  "epoch": 0.9581794590278471,
+  "eval_steps": 500,
+  "global_step": 6000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0015969657650464117,
+      "grad_norm": 0.15089154243469238,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 1.5709858894348145,
+      "step": 10
+    },
+    {
+      "epoch": 0.0031939315300928235,
+      "grad_norm": 0.1435190588235855,
+      "learning_rate": 1.2666666666666669e-06,
+      "loss": 1.6188209533691407,
+      "step": 20
+    },
+    {
+      "epoch": 0.004790897295139236,
+      "grad_norm": 0.14615978300571442,
+      "learning_rate": 1.9333333333333336e-06,
+      "loss": 1.6381675720214843,
+      "step": 30
+    },
+    {
+      "epoch": 0.006387863060185647,
+      "grad_norm": 0.1281592696905136,
+      "learning_rate": 2.6e-06,
+      "loss": 1.5563851356506349,
+      "step": 40
+    },
+    {
+      "epoch": 0.00798482882523206,
+      "grad_norm": 0.16032098233699799,
+      "learning_rate": 3.266666666666667e-06,
+      "loss": 1.6382104873657226,
+      "step": 50
+    },
+    {
+      "epoch": 0.009581794590278471,
+      "grad_norm": 0.14609354734420776,
+      "learning_rate": 3.9333333333333335e-06,
+      "loss": 1.6256795883178712,
+      "step": 60
+    },
+    {
+      "epoch": 0.011178760355324883,
+      "grad_norm": 0.13070641458034515,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 1.5222463607788086,
+      "step": 70
+    },
+    {
+      "epoch": 0.012775726120371294,
+      "grad_norm": 0.1499311476945877,
+      "learning_rate": 5.2666666666666665e-06,
+      "loss": 1.518197727203369,
+      "step": 80
+    },
+    {
+      "epoch": 0.014372691885417706,
+      "grad_norm": 0.12919697165489197,
+      "learning_rate": 5.933333333333335e-06,
+      "loss": 1.4710905075073242,
+      "step": 90
+    },
+    {
+      "epoch": 0.01596965765046412,
+      "grad_norm": 0.11437301337718964,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 1.4049152374267577,
+      "step": 100
+    },
+    {
+      "epoch": 0.01756662341551053,
+      "grad_norm": 0.13434389233589172,
+      "learning_rate": 7.266666666666668e-06,
+      "loss": 1.3853497505187988,
+      "step": 110
+    },
+    {
+      "epoch": 0.019163589180556943,
+      "grad_norm": 0.08636850863695145,
+      "learning_rate": 7.933333333333334e-06,
+      "loss": 1.3351438522338868,
+      "step": 120
+    },
+    {
+      "epoch": 0.020760554945603353,
+      "grad_norm": 0.08471965044736862,
+      "learning_rate": 8.6e-06,
+      "loss": 1.2937309265136718,
+      "step": 130
+    },
+    {
+      "epoch": 0.022357520710649767,
+      "grad_norm": 0.058983951807022095,
+      "learning_rate": 9.266666666666667e-06,
+      "loss": 1.2116679191589355,
+      "step": 140
+    },
+    {
+      "epoch": 0.023954486475696177,
+      "grad_norm": 0.06733114272356033,
+      "learning_rate": 9.933333333333334e-06,
+      "loss": 1.1680314064025878,
+      "step": 150
+    },
+    {
+      "epoch": 0.025551452240742588,
+      "grad_norm": 0.045156873762607574,
+      "learning_rate": 1.0600000000000002e-05,
+      "loss": 1.1280742645263673,
+      "step": 160
+    },
+    {
+      "epoch": 0.027148418005789002,
+      "grad_norm": 0.0415460579097271,
+      "learning_rate": 1.1266666666666668e-05,
+      "loss": 1.0879782676696776,
+      "step": 170
+    },
+    {
+      "epoch": 0.028745383770835412,
+      "grad_norm": 0.05221620574593544,
+      "learning_rate": 1.1933333333333335e-05,
+      "loss": 1.0641801834106446,
+      "step": 180
+    },
+    {
+      "epoch": 0.030342349535881826,
+      "grad_norm": 0.036492571234703064,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 1.0208759307861328,
+      "step": 190
+    },
+    {
+      "epoch": 0.03193931530092824,
+      "grad_norm": 0.03153146430850029,
+      "learning_rate": 1.3266666666666668e-05,
+      "loss": 1.0204400062561034,
+      "step": 200
+    },
+    {
+      "epoch": 0.03353628106597465,
+      "grad_norm": 0.03343074768781662,
+      "learning_rate": 1.3933333333333334e-05,
+      "loss": 0.9532640457153321,
+      "step": 210
+    },
+    {
+      "epoch": 0.03513324683102106,
+      "grad_norm": 0.034289468079805374,
+      "learning_rate": 1.46e-05,
+      "loss": 0.9065071105957031,
+      "step": 220
+    },
+    {
+      "epoch": 0.03673021259606747,
+      "grad_norm": 0.040245093405246735,
+      "learning_rate": 1.5266666666666667e-05,
+      "loss": 0.895290470123291,
+      "step": 230
+    },
+    {
+      "epoch": 0.038327178361113885,
+      "grad_norm": 0.03090902790427208,
+      "learning_rate": 1.5933333333333336e-05,
+      "loss": 0.89828462600708,
+      "step": 240
+    },
+    {
+      "epoch": 0.03992414412616029,
+      "grad_norm": 0.023231355473399162,
+      "learning_rate": 1.66e-05,
+      "loss": 0.867582893371582,
+      "step": 250
+    },
+    {
+      "epoch": 0.041521109891206706,
+      "grad_norm": 0.022748615592718124,
+      "learning_rate": 1.726666666666667e-05,
+      "loss": 0.8476214408874512,
+      "step": 260
+    },
+    {
+      "epoch": 0.04311807565625312,
+      "grad_norm": 0.025588329881429672,
+      "learning_rate": 1.7933333333333333e-05,
+      "loss": 0.8639037132263183,
+      "step": 270
+    },
+    {
+      "epoch": 0.044715041421299534,
+      "grad_norm": 0.029987547546625137,
+      "learning_rate": 1.86e-05,
+      "loss": 0.8353777885437011,
+      "step": 280
+    },
+    {
+      "epoch": 0.04631200718634594,
+      "grad_norm": 0.020994428545236588,
+      "learning_rate": 1.926666666666667e-05,
+      "loss": 0.8189143180847168,
+      "step": 290
+    },
+    {
+      "epoch": 0.047908972951392355,
+      "grad_norm": 0.023545585572719574,
+      "learning_rate": 1.9933333333333334e-05,
+      "loss": 0.7846664905548095,
+      "step": 300
+    },
+    {
+      "epoch": 0.04950593871643877,
+      "grad_norm": 0.025134805589914322,
+      "learning_rate": 1.999988754726792e-05,
+      "loss": 0.7751604557037354,
+      "step": 310
+    },
+    {
+      "epoch": 0.051102904481485176,
+      "grad_norm": 0.021275486797094345,
+      "learning_rate": 1.9999498825021314e-05,
+      "loss": 0.7967638969421387,
+      "step": 320
+    },
+    {
+      "epoch": 0.05269987024653159,
+      "grad_norm": 0.019746815785765648,
+      "learning_rate": 1.9998832455745586e-05,
+      "loss": 0.8192337036132813,
+      "step": 330
+    },
+    {
+      "epoch": 0.054296836011578004,
+      "grad_norm": 0.02567191608250141,
+      "learning_rate": 1.9997888457943197e-05,
+      "loss": 0.7604565620422363,
+      "step": 340
+    },
+    {
+      "epoch": 0.05589380177662441,
+      "grad_norm": 0.024295911192893982,
+      "learning_rate": 1.9996666857825287e-05,
+      "loss": 0.7550592422485352,
+      "step": 350
+    },
+    {
+      "epoch": 0.057490767541670824,
+      "grad_norm": 0.02220323495566845,
+      "learning_rate": 1.9995167689310917e-05,
+      "loss": 0.7514551639556885,
+      "step": 360
+    },
+    {
+      "epoch": 0.05908773330671724,
+      "grad_norm": 0.024802017956972122,
+      "learning_rate": 1.9993390994026144e-05,
+      "loss": 0.7904987812042237,
+      "step": 370
+    },
+    {
+      "epoch": 0.06068469907176365,
+      "grad_norm": 0.027685435488820076,
+      "learning_rate": 1.9991336821302856e-05,
+      "loss": 0.7770297050476074,
+      "step": 380
+    },
+    {
+      "epoch": 0.06228166483681006,
+      "grad_norm": 0.019939841702580452,
+      "learning_rate": 1.9989005228177406e-05,
+      "loss": 0.7634893417358398,
+      "step": 390
+    },
+    {
+      "epoch": 0.06387863060185647,
+      "grad_norm": 0.028217511251568794,
+      "learning_rate": 1.9986396279389028e-05,
+      "loss": 0.8059945106506348,
+      "step": 400
+    },
+    {
+      "epoch": 0.06547559636690288,
+      "grad_norm": 0.019439322873950005,
+      "learning_rate": 1.9983510047378046e-05,
+      "loss": 0.7787755489349365,
+      "step": 410
+    },
+    {
+      "epoch": 0.0670725621319493,
+      "grad_norm": 0.02309611812233925,
+      "learning_rate": 1.9980346612283842e-05,
+      "loss": 0.7745323657989502,
+      "step": 420
+    },
+    {
+      "epoch": 0.06866952789699571,
+      "grad_norm": 0.020378535613417625,
+      "learning_rate": 1.997690606194266e-05,
+      "loss": 0.7243520736694335,
+      "step": 430
+    },
+    {
+      "epoch": 0.07026649366204211,
+      "grad_norm": 0.021776653826236725,
+      "learning_rate": 1.9973188491885146e-05,
+      "loss": 0.7424108028411865,
+      "step": 440
+    },
+    {
+      "epoch": 0.07186345942708854,
+      "grad_norm": 0.03435182571411133,
+      "learning_rate": 1.9969194005333697e-05,
+      "loss": 0.7804229736328125,
+      "step": 450
+    },
+    {
+      "epoch": 0.07346042519213494,
+      "grad_norm": 0.026880595833063126,
+      "learning_rate": 1.9964922713199613e-05,
+      "loss": 0.7963083267211915,
+      "step": 460
+    },
+    {
+      "epoch": 0.07505739095718135,
+      "grad_norm": 0.02643360197544098,
+      "learning_rate": 1.9960374734079984e-05,
+      "loss": 0.7517959117889405,
+      "step": 470
+    },
+    {
+      "epoch": 0.07665435672222777,
+      "grad_norm": 0.036808405071496964,
+      "learning_rate": 1.9955550194254436e-05,
+      "loss": 0.7511186122894287,
+      "step": 480
+    },
+    {
+      "epoch": 0.07825132248727418,
+      "grad_norm": 0.03115280158817768,
+      "learning_rate": 1.995044922768159e-05,
+      "loss": 0.765751314163208,
+      "step": 490
+    },
+    {
+      "epoch": 0.07984828825232058,
+      "grad_norm": 0.025536708533763885,
+      "learning_rate": 1.994507197599537e-05,
+      "loss": 0.7346895217895508,
+      "step": 500
+    },
+    {
+      "epoch": 0.07984828825232058,
+      "eval_loss": 0.7535409331321716,
+      "eval_runtime": 1705.9703,
+      "eval_samples_per_second": 3.263,
+      "eval_steps_per_second": 1.631,
+      "step": 500
+    },
+    {
+      "epoch": 0.081445254017367,
+      "grad_norm": 0.03875249624252319,
+      "learning_rate": 1.993941858850106e-05,
+      "loss": 0.7961405277252197,
+      "step": 510
+    },
+    {
+      "epoch": 0.08304221978241341,
+      "grad_norm": 0.027557438239455223,
+      "learning_rate": 1.993348922217114e-05,
+      "loss": 0.7667338371276855,
+      "step": 520
+    },
+    {
+      "epoch": 0.08463918554745983,
+      "grad_norm": 0.026045992970466614,
+      "learning_rate": 1.9927284041640967e-05,
+      "loss": 0.7275139808654785,
+      "step": 530
+    },
+    {
+      "epoch": 0.08623615131250624,
+      "grad_norm": 0.026213862001895905,
+      "learning_rate": 1.992080321920416e-05,
+      "loss": 0.7866159439086914,
+      "step": 540
+    },
+    {
+      "epoch": 0.08783311707755265,
+      "grad_norm": 0.028538137674331665,
+      "learning_rate": 1.9914046934807853e-05,
+      "loss": 0.7298189640045166,
+      "step": 550
+    },
+    {
+      "epoch": 0.08943008284259907,
+      "grad_norm": 0.028126219287514687,
+      "learning_rate": 1.9907015376047675e-05,
+      "loss": 0.749655294418335,
+      "step": 560
+    },
+    {
+      "epoch": 0.09102704860764547,
+      "grad_norm": 0.027085473760962486,
+      "learning_rate": 1.9899708738162553e-05,
+      "loss": 0.8086295127868652,
+      "step": 570
+    },
+    {
+      "epoch": 0.09262401437269188,
+      "grad_norm": 0.034640420228242874,
+      "learning_rate": 1.989212722402928e-05,
+      "loss": 0.7653923511505127,
+      "step": 580
+    },
+    {
+      "epoch": 0.0942209801377383,
+      "grad_norm": 0.03858828917145729,
+      "learning_rate": 1.98842710441569e-05,
+      "loss": 0.8174265861511231,
+      "step": 590
+    },
+    {
+      "epoch": 0.09581794590278471,
+      "grad_norm": 0.04248794540762901,
+      "learning_rate": 1.987614041668084e-05,
+      "loss": 0.7357111930847168,
+      "step": 600
+    },
+    {
+      "epoch": 0.09741491166783112,
+      "grad_norm": 0.03530848026275635,
+      "learning_rate": 1.9867735567356876e-05,
+      "loss": 0.7747371196746826,
+      "step": 610
+    },
+    {
+      "epoch": 0.09901187743287754,
+      "grad_norm": 0.027734428644180298,
+      "learning_rate": 1.9859056729554845e-05,
+      "loss": 0.73280348777771,
+      "step": 620
+    },
+    {
+      "epoch": 0.10060884319792394,
+      "grad_norm": 0.02657356671988964,
+      "learning_rate": 1.9850104144252177e-05,
+      "loss": 0.749216365814209,
+      "step": 630
+    },
+    {
+      "epoch": 0.10220580896297035,
+      "grad_norm": 0.021906374022364616,
+      "learning_rate": 1.98408780600272e-05,
+      "loss": 0.7534349441528321,
+      "step": 640
+    },
+    {
+      "epoch": 0.10380277472801677,
+      "grad_norm": 0.025684406980872154,
+      "learning_rate": 1.9831378733052244e-05,
+      "loss": 0.7199561595916748,
+      "step": 650
+    },
+    {
+      "epoch": 0.10539974049306318,
+      "grad_norm": 0.025311095640063286,
+      "learning_rate": 1.982160642708652e-05,
+      "loss": 0.7382417678833008,
+      "step": 660
+    },
+    {
+      "epoch": 0.10699670625810959,
+      "grad_norm": 0.035563357174396515,
+      "learning_rate": 1.9811561413468794e-05,
+      "loss": 0.7394683837890625,
+      "step": 670
+    },
+    {
+      "epoch": 0.10859367202315601,
+      "grad_norm": 0.023977380245923996,
+      "learning_rate": 1.9801243971109868e-05,
+      "loss": 0.7267738819122315,
+      "step": 680
+    },
+    {
+      "epoch": 0.11019063778820241,
+      "grad_norm": 0.02275015600025654,
+      "learning_rate": 1.9790654386484818e-05,
+      "loss": 0.7240358829498291,
+      "step": 690
+    },
+    {
+      "epoch": 0.11178760355324882,
+      "grad_norm": 0.036198344081640244,
+      "learning_rate": 1.9779792953625052e-05,
+      "loss": 0.7299670696258544,
+      "step": 700
+    },
+    {
+      "epoch": 0.11338456931829524,
+      "grad_norm": 0.027184955775737762,
+      "learning_rate": 1.976865997411014e-05,
+      "loss": 0.695775318145752,
+      "step": 710
+    },
+    {
+      "epoch": 0.11498153508334165,
+      "grad_norm": 0.023749997839331627,
+      "learning_rate": 1.9757255757059446e-05,
+      "loss": 0.7071991920471191,
+      "step": 720
+    },
+    {
+      "epoch": 0.11657850084838806,
+      "grad_norm": 0.027117466554045677,
+      "learning_rate": 1.9745580619123535e-05,
+      "loss": 0.7466438293457032,
+      "step": 730
+    },
+    {
+      "epoch": 0.11817546661343448,
+      "grad_norm": 0.027798349037766457,
+      "learning_rate": 1.9733634884475395e-05,
+      "loss": 0.7503840923309326,
+      "step": 740
+    },
+    {
+      "epoch": 0.11977243237848088,
+      "grad_norm": 0.026327304542064667,
+      "learning_rate": 1.9721418884801414e-05,
+      "loss": 0.724392032623291,
+      "step": 750
+    },
+    {
+      "epoch": 0.1213693981435273,
+      "grad_norm": 0.029995381832122803,
+      "learning_rate": 1.97089329592922e-05,
+      "loss": 0.7100958824157715,
+      "step": 760
+    },
+    {
+      "epoch": 0.12296636390857371,
+      "grad_norm": 0.03063913807272911,
+      "learning_rate": 1.969617745463314e-05,
+      "loss": 0.7130911827087403,
+      "step": 770
+    },
+    {
+      "epoch": 0.12456332967362012,
+      "grad_norm": 0.021976860240101814,
+      "learning_rate": 1.968315272499478e-05,
+      "loss": 0.7231676578521729,
+      "step": 780
+    },
+    {
+      "epoch": 0.12616029543866653,
+      "grad_norm": 0.02538118325173855,
+      "learning_rate": 1.9669859132022994e-05,
+      "loss": 0.7253612995147705,
+      "step": 790
+    },
+    {
+      "epoch": 0.12775726120371295,
+      "grad_norm": 0.02788228541612625,
+      "learning_rate": 1.9656297044828943e-05,
+      "loss": 0.7297886848449707,
+      "step": 800
+    },
+    {
+      "epoch": 0.12935422696875937,
+      "grad_norm": 0.028525004163384438,
+      "learning_rate": 1.9642466839978814e-05,
+      "loss": 0.7109212875366211,
+      "step": 810
+    },
+    {
+      "epoch": 0.13095119273380576,
+      "grad_norm": 0.028368208557367325,
+      "learning_rate": 1.962836890148339e-05,
+      "loss": 0.7552286624908447,
+      "step": 820
+    },
+    {
+      "epoch": 0.13254815849885218,
+      "grad_norm": 0.03732667118310928,
+      "learning_rate": 1.9614003620787358e-05,
+      "loss": 0.74849853515625,
+      "step": 830
+    },
+    {
+      "epoch": 0.1341451242638986,
+      "grad_norm": 0.026267215609550476,
+      "learning_rate": 1.9599371396758457e-05,
+      "loss": 0.7323933124542237,
+      "step": 840
+    },
+    {
+      "epoch": 0.135742090028945,
+      "grad_norm": 0.031789544969797134,
+      "learning_rate": 1.958447263567641e-05,
+      "loss": 0.7394798755645752,
+      "step": 850
+    },
+    {
+      "epoch": 0.13733905579399142,
+      "grad_norm": 0.03209487721323967,
+      "learning_rate": 1.956930775122162e-05,
+      "loss": 0.7479897499084472,
+      "step": 860
+    },
+    {
+      "epoch": 0.13893602155903784,
+      "grad_norm": 0.026272661983966827,
+      "learning_rate": 1.9553877164463698e-05,
+      "loss": 0.7194801807403565,
+      "step": 870
+    },
+    {
+      "epoch": 0.14053298732408423,
+      "grad_norm": 0.02558542974293232,
+      "learning_rate": 1.953818130384978e-05,
+      "loss": 0.6967973709106445,
+      "step": 880
+    },
+    {
+      "epoch": 0.14212995308913065,
+      "grad_norm": 0.03761237859725952,
+      "learning_rate": 1.9522220605192615e-05,
+      "loss": 0.6678271770477295,
+      "step": 890
+    },
+    {
+      "epoch": 0.14372691885417707,
+      "grad_norm": 0.029605882242321968,
+      "learning_rate": 1.9505995511658464e-05,
+      "loss": 0.6922338008880615,
+      "step": 900
+    },
+    {
+      "epoch": 0.14532388461922346,
+      "grad_norm": 0.03535737469792366,
+      "learning_rate": 1.948950647375481e-05,
+      "loss": 0.7159334182739258,
+      "step": 910
+    },
+    {
+      "epoch": 0.14692085038426989,
+      "grad_norm": 0.030855044722557068,
+      "learning_rate": 1.9472753949317843e-05,
+      "loss": 0.7333884716033936,
+      "step": 920
+    },
+    {
+      "epoch": 0.1485178161493163,
+      "grad_norm": 0.05045896768569946,
+      "learning_rate": 1.9455738403499728e-05,
+      "loss": 0.6935329914093018,
+      "step": 930
+    },
+    {
+      "epoch": 0.1501147819143627,
+      "grad_norm": 0.033187806606292725,
+      "learning_rate": 1.9438460308755724e-05,
+      "loss": 0.7046589851379395,
+      "step": 940
+    },
+    {
+      "epoch": 0.15171174767940912,
+      "grad_norm": 0.03167016804218292,
+      "learning_rate": 1.9420920144831044e-05,
+      "loss": 0.710863447189331,
+      "step": 950
+    },
+    {
+      "epoch": 0.15330871344445554,
+      "grad_norm": 0.025539802387356758,
+      "learning_rate": 1.9403118398747533e-05,
+      "loss": 0.6936647891998291,
+      "step": 960
+    },
+    {
+      "epoch": 0.15490567920950193,
+      "grad_norm": 0.04223870858550072,
+      "learning_rate": 1.9385055564790157e-05,
+      "loss": 0.6985628128051757,
+      "step": 970
+    },
+    {
+      "epoch": 0.15650264497454835,
+      "grad_norm": 0.030605314299464226,
+      "learning_rate": 1.9366732144493266e-05,
+      "loss": 0.7294198989868164,
+      "step": 980
+    },
+    {
+      "epoch": 0.15809961073959478,
+      "grad_norm": 0.03095085918903351,
+      "learning_rate": 1.9348148646626687e-05,
+      "loss": 0.7009531021118164,
+      "step": 990
+    },
+    {
+      "epoch": 0.15969657650464117,
+      "grad_norm": 0.02993757091462612,
+      "learning_rate": 1.9329305587181574e-05,
+      "loss": 0.7266313552856445,
+      "step": 1000
+    },
+    {
+      "epoch": 0.15969657650464117,
+      "eval_loss": 0.7159722447395325,
+      "eval_runtime": 1703.1621,
+      "eval_samples_per_second": 3.268,
+      "eval_steps_per_second": 1.634,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1612935422696876,
+      "grad_norm": 0.035543542355298996,
+      "learning_rate": 1.9310203489356092e-05,
+      "loss": 0.74082350730896,
+      "step": 1010
+    },
+    {
+      "epoch": 0.162890508034734,
+      "grad_norm": 0.03636915981769562,
+      "learning_rate": 1.9290842883540897e-05,
+      "loss": 0.717669153213501,
+      "step": 1020
+    },
+    {
+      "epoch": 0.1644874737997804,
+      "grad_norm": 0.04510757327079773,
+      "learning_rate": 1.92712243073044e-05,
+      "loss": 0.7365827560424805,
+      "step": 1030
+    },
+    {
+      "epoch": 0.16608443956482682,
+      "grad_norm": 0.03563699871301651,
+      "learning_rate": 1.925134830537784e-05,
+      "loss": 0.6981801986694336,
+      "step": 1040
+    },
+    {
+      "epoch": 0.16768140532987325,
+      "grad_norm": 0.04541337490081787,
+      "learning_rate": 1.9231215429640167e-05,
+      "loss": 0.754232931137085,
+      "step": 1050
+    },
+    {
+      "epoch": 0.16927837109491967,
+      "grad_norm": 0.033323634415864944,
+      "learning_rate": 1.921082623910271e-05,
+      "loss": 0.691849946975708,
+      "step": 1060
+    },
+    {
+      "epoch": 0.17087533685996606,
+      "grad_norm": 0.03559419885277748,
+      "learning_rate": 1.919018129989366e-05,
+      "loss": 0.6994197845458985,
+      "step": 1070
+    },
+    {
+      "epoch": 0.17247230262501248,
+      "grad_norm": 0.03182852268218994,
+      "learning_rate": 1.916928118524235e-05,
+      "loss": 0.6645867824554443,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1740692683900589,
+      "grad_norm": 0.02616371586918831,
+      "learning_rate": 1.9148126475463336e-05,
+      "loss": 0.7270137786865234,
+      "step": 1090
+    },
+    {
+      "epoch": 0.1756662341551053,
+      "grad_norm": 0.03250862658023834,
+      "learning_rate": 1.9126717757940288e-05,
+      "loss": 0.7272531509399414,
+      "step": 1100
+    },
+    {
+      "epoch": 0.17726319992015172,
+      "grad_norm": 0.036741774529218674,
+      "learning_rate": 1.9105055627109683e-05,
+      "loss": 0.7251851558685303,
+      "step": 1110
+    },
+    {
+      "epoch": 0.17886016568519814,
+      "grad_norm": 0.03240974619984627,
+      "learning_rate": 1.908314068444429e-05,
+      "loss": 0.7145021915435791,
+      "step": 1120
+    },
+    {
+      "epoch": 0.18045713145024453,
+      "grad_norm": 0.0324835442006588,
+      "learning_rate": 1.9060973538436478e-05,
+      "loss": 0.7045553684234619,
+      "step": 1130
+    },
+    {
+      "epoch": 0.18205409721529095,
+      "grad_norm": 0.029804598540067673,
+      "learning_rate": 1.9038554804581318e-05,
+      "loss": 0.7342820644378663,
+      "step": 1140
+    },
+    {
+      "epoch": 0.18365106298033737,
+      "grad_norm": 0.03631270304322243,
+      "learning_rate": 1.9015885105359492e-05,
+      "loss": 0.7435333728790283,
+      "step": 1150
+    },
+    {
+      "epoch": 0.18524802874538376,
+      "grad_norm": 0.031312599778175354,
+      "learning_rate": 1.8992965070220007e-05,
+      "loss": 0.7185348033905029,
+      "step": 1160
+    },
+    {
+      "epoch": 0.18684499451043018,
+      "grad_norm": 0.03553950414061546,
+      "learning_rate": 1.896979533556273e-05,
+      "loss": 0.7191666603088379,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1884419602754766,
+      "grad_norm": 0.024733861908316612,
+      "learning_rate": 1.8946376544720698e-05,
+      "loss": 0.7352997779846191,
+      "step": 1180
+    },
+    {
+      "epoch": 0.190038926040523,
+      "grad_norm": 0.030409259721636772,
+      "learning_rate": 1.8922709347942275e-05,
+      "loss": 0.7265621185302734,
+      "step": 1190
+    },
+    {
+      "epoch": 0.19163589180556942,
+      "grad_norm": 0.030736226588487625,
+      "learning_rate": 1.8898794402373077e-05,
+      "loss": 0.7150910377502442,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19323285757061584,
+      "grad_norm": 0.038398947566747665,
+      "learning_rate": 1.887463237203775e-05,
+      "loss": 0.7055376052856446,
+      "step": 1210
+    },
+    {
+      "epoch": 0.19482982333566223,
+      "grad_norm": 0.03485625982284546,
+      "learning_rate": 1.88502239278215e-05,
+      "loss": 0.6555115699768066,
+      "step": 1220
+    },
+    {
+      "epoch": 0.19642678910070865,
+      "grad_norm": 0.03220400586724281,
+      "learning_rate": 1.8825569747451505e-05,
+      "loss": 0.6946470737457275,
+      "step": 1230
+    },
+    {
+      "epoch": 0.19802375486575508,
+      "grad_norm": 0.03333678096532822,
+      "learning_rate": 1.880067051547806e-05,
+      "loss": 0.7297664642333984,
+      "step": 1240
+    },
+    {
+      "epoch": 0.19962072063080147,
+      "grad_norm": 0.0367787703871727,
+      "learning_rate": 1.8775526923255597e-05,
+      "loss": 0.7051557064056396,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2012176863958479,
+      "grad_norm": 0.03103001043200493,
+      "learning_rate": 1.8750139668923472e-05,
+      "loss": 0.7219597339630127,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2028146521608943,
+      "grad_norm": 0.04533790051937103,
+      "learning_rate": 1.872450945738659e-05,
+      "loss": 0.6916751384735107,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2044116179259407,
+      "grad_norm": 0.03510045260190964,
+      "learning_rate": 1.8698637000295816e-05,
+      "loss": 0.7119457721710205,
+      "step": 1280
+    },
+    {
+      "epoch": 0.20600858369098712,
+      "grad_norm": 0.029134295880794525,
+      "learning_rate": 1.867252301602825e-05,
+      "loss": 0.6870355129241943,
+      "step": 1290
+    },
+    {
+      "epoch": 0.20760554945603354,
+      "grad_norm": 0.03194071725010872,
+      "learning_rate": 1.8646168229667238e-05,
+      "loss": 0.734464693069458,
+      "step": 1300
+    },
+    {
+      "epoch": 0.20920251522107994,
+      "grad_norm": 0.026471436023712158,
+      "learning_rate": 1.861957337298227e-05,
+      "loss": 0.7240866184234619,
+      "step": 1310
+    },
+    {
+      "epoch": 0.21079948098612636,
+      "grad_norm": 0.0347883440554142,
+      "learning_rate": 1.8592739184408657e-05,
+      "loss": 0.6553101062774658,
+      "step": 1320
+    },
+    {
+      "epoch": 0.21239644675117278,
+      "grad_norm": 0.026658106595277786,
+      "learning_rate": 1.8565666409027004e-05,
+      "loss": 0.7519384384155273,
+      "step": 1330
+    },
+    {
+      "epoch": 0.21399341251621917,
+      "grad_norm": 0.030675504356622696,
+      "learning_rate": 1.8538355798542556e-05,
+      "loss": 0.6963082790374756,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2155903782812656,
+      "grad_norm": 0.04593832045793533,
+      "learning_rate": 1.85108081112643e-05,
+      "loss": 0.7145741939544678,
+      "step": 1350
+    },
+    {
+      "epoch": 0.21718734404631201,
+      "grad_norm": 0.04163511469960213,
+      "learning_rate": 1.8483024112083928e-05,
+      "loss": 0.7342512130737304,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2187843098113584,
+      "grad_norm": 0.025683345273137093,
+      "learning_rate": 1.8455004572454583e-05,
+      "loss": 0.7134137630462647,
+      "step": 1370
+    },
+    {
+      "epoch": 0.22038127557640483,
+      "grad_norm": 0.026864832267165184,
+      "learning_rate": 1.8426750270369452e-05,
+      "loss": 0.6854844570159913,
+      "step": 1380
+    },
+    {
+      "epoch": 0.22197824134145125,
+      "grad_norm": 0.03378361091017723,
+      "learning_rate": 1.839826199034015e-05,
+      "loss": 0.6904460430145264,
+      "step": 1390
+    },
+    {
+      "epoch": 0.22357520710649764,
+      "grad_norm": 0.03360961750149727,
+      "learning_rate": 1.8369540523374963e-05,
+      "loss": 0.7089653491973877,
+      "step": 1400
+    },
+    {
+      "epoch": 0.22517217287154406,
+      "grad_norm": 0.033883776515722275,
+      "learning_rate": 1.8340586666956846e-05,
+      "loss": 0.7053616523742676,
+      "step": 1410
+    },
+    {
+      "epoch": 0.22676913863659048,
+      "grad_norm": 0.03643488511443138,
+      "learning_rate": 1.8311401225021318e-05,
+      "loss": 0.7411230087280274,
+      "step": 1420
+    },
+    {
+      "epoch": 0.22836610440163688,
+      "grad_norm": 0.030240802094340324,
+      "learning_rate": 1.8281985007934115e-05,
+      "loss": 0.7020374298095703,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2299630701666833,
+      "grad_norm": 0.03315526619553566,
+      "learning_rate": 1.8252338832468702e-05,
+      "loss": 0.7227590084075928,
+      "step": 1440
+    },
+    {
+      "epoch": 0.23156003593172972,
+      "grad_norm": 0.04955840855836868,
+      "learning_rate": 1.8222463521783584e-05,
+      "loss": 0.7004672527313233,
+      "step": 1450
+    },
+    {
+      "epoch": 0.2331570016967761,
+      "grad_norm": 0.03338323533535004,
+      "learning_rate": 1.819235990539946e-05,
+      "loss": 0.7032230377197266,
+      "step": 1460
+    },
+    {
+      "epoch": 0.23475396746182253,
+      "grad_norm": 0.03659350797533989,
+      "learning_rate": 1.8162028819176192e-05,
+      "loss": 0.7022134780883789,
+      "step": 1470
+    },
+    {
+      "epoch": 0.23635093322686895,
+      "grad_norm": 0.029659852385520935,
+      "learning_rate": 1.813147110528958e-05,
+      "loss": 0.7326688289642334,
+      "step": 1480
+    },
+    {
+      "epoch": 0.23794789899191537,
+      "grad_norm": 0.03314507007598877,
+      "learning_rate": 1.8100687612208e-05,
+      "loss": 0.7101527690887451,
+      "step": 1490
+    },
+    {
+      "epoch": 0.23954486475696177,
+      "grad_norm": 0.048770975321531296,
+      "learning_rate": 1.806967919466883e-05,
+      "loss": 0.7013855457305909,
+      "step": 1500
+    },
+    {
+      "epoch": 0.23954486475696177,
+      "eval_loss": 0.7003746032714844,
+      "eval_runtime": 1694.4125,
+      "eval_samples_per_second": 3.285,
+      "eval_steps_per_second": 1.642,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2411418305220082,
+      "grad_norm": 0.031342763453722,
+      "learning_rate": 1.803844671365471e-05,
+      "loss": 0.6915247917175293,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2427387962870546,
+      "grad_norm": 0.029329324141144753,
+      "learning_rate": 1.800699103636967e-05,
+      "loss": 0.721204423904419,
+      "step": 1520
+    },
+    {
+      "epoch": 0.244335762052101,
+      "grad_norm": 0.03263983875513077,
+      "learning_rate": 1.7975313036215015e-05,
+      "loss": 0.7468688011169433,
+      "step": 1530
+    },
+    {
+      "epoch": 0.24593272781714742,
+      "grad_norm": 0.035935308784246445,
+      "learning_rate": 1.794341359276509e-05,
+      "loss": 0.6986902713775635,
+      "step": 1540
+    },
+    {
+      "epoch": 0.24752969358219384,
+      "grad_norm": 0.03797990456223488,
+      "learning_rate": 1.7911293591742855e-05,
+      "loss": 0.7199019908905029,
+      "step": 1550
+    },
+    {
+      "epoch": 0.24912665934724024,
+      "grad_norm": 0.024390801787376404,
+      "learning_rate": 1.787895392499529e-05,
+      "loss": 0.6761057853698731,
+      "step": 1560
+    },
+    {
+      "epoch": 0.25072362511228663,
+      "grad_norm": 0.0313715860247612,
+      "learning_rate": 1.7846395490468643e-05,
+      "loss": 0.6731560707092286,
+      "step": 1570
+    },
+    {
+      "epoch": 0.25232059087733305,
+      "grad_norm": 0.0433608703315258,
+      "learning_rate": 1.781361919218348e-05,
+      "loss": 0.685046911239624,
+      "step": 1580
+    },
+    {
+      "epoch": 0.25391755664237947,
+      "grad_norm": 0.050231028348207474,
+      "learning_rate": 1.7780625940209596e-05,
+      "loss": 0.7280925273895263,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2555145224074259,
+      "grad_norm": 0.0476137213408947,
+      "learning_rate": 1.774741665064074e-05,
+      "loss": 0.728914451599121,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2571114881724723,
+      "grad_norm": 0.03540361300110817,
+      "learning_rate": 1.771399224556919e-05,
+      "loss": 0.6937174320220947,
+      "step": 1610
+    },
+    {
+      "epoch": 0.25870845393751873,
+      "grad_norm": 0.038516197353601456,
+      "learning_rate": 1.7680353653060135e-05,
+      "loss": 0.6788946151733398,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2603054197025651,
+      "grad_norm": 0.0472414456307888,
+      "learning_rate": 1.7646501807125905e-05,
+      "loss": 0.7246061325073242,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2619023854676115,
+      "grad_norm": 0.026418814435601234,
+      "learning_rate": 1.7612437647700056e-05,
+      "loss": 0.6792353630065918,
+      "step": 1640
+    },
+    {
+      "epoch": 0.26349935123265794,
+      "grad_norm": 0.025900904089212418,
+      "learning_rate": 1.757816212061126e-05,
+      "loss": 0.6991750240325928,
+      "step": 1650
+    },
+    {
+      "epoch": 0.26509631699770436,
+      "grad_norm": 0.036637961864471436,
+      "learning_rate": 1.7543676177557042e-05,
+      "loss": 0.7213243007659912,
+      "step": 1660
+    },
+    {
+      "epoch": 0.2666932827627508,
+      "grad_norm": 0.031415775418281555,
+      "learning_rate": 1.750898077607735e-05,
+      "loss": 0.6988609790802002,
+      "step": 1670
+    },
+    {
+      "epoch": 0.2682902485277972,
+      "grad_norm": 0.036175280809402466,
+      "learning_rate": 1.7474076879527977e-05,
+      "loss": 0.6812397003173828,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2698872142928436,
+      "grad_norm": 0.03462570160627365,
+      "learning_rate": 1.743896545705382e-05,
+      "loss": 0.6791361331939697,
+      "step": 1690
+    },
+    {
+      "epoch": 0.27148418005789,
+      "grad_norm": 0.03334948793053627,
+      "learning_rate": 1.740364748356195e-05,
+      "loss": 0.6668567180633544,
+      "step": 1700
+    },
+    {
+      "epoch": 0.2730811458229364,
+      "grad_norm": 0.03270823508501053,
+      "learning_rate": 1.7368123939694554e-05,
+      "loss": 0.6887404441833496,
+      "step": 1710
+    },
+    {
+      "epoch": 0.27467811158798283,
+      "grad_norm": 0.03120650351047516,
+      "learning_rate": 1.7332395811801706e-05,
+      "loss": 0.6659372806549072,
+      "step": 1720
+    },
+    {
+      "epoch": 0.27627507735302925,
+      "grad_norm": 0.042766936123371124,
+      "learning_rate": 1.7296464091913986e-05,
+      "loss": 0.7515771389007568,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2778720431180757,
+      "grad_norm": 0.027079230174422264,
+      "learning_rate": 1.7260329777714923e-05,
+      "loss": 0.7227123260498047,
+      "step": 1740
+    },
+    {
+      "epoch": 0.2794690088831221,
+      "grad_norm": 0.03032068908214569,
+      "learning_rate": 1.722399387251329e-05,
+      "loss": 0.695145559310913,
+      "step": 1750
+    },
+    {
+      "epoch": 0.28106597464816846,
+      "grad_norm": 0.03420598804950714,
+      "learning_rate": 1.7187457385215274e-05,
+      "loss": 0.6715566158294678,
+      "step": 1760
+    },
+    {
+      "epoch": 0.2826629404132149,
+      "grad_norm": 0.037743836641311646,
+      "learning_rate": 1.7150721330296428e-05,
+      "loss": 0.6833428382873535,
+      "step": 1770
+    },
+    {
+      "epoch": 0.2842599061782613,
+      "grad_norm": 0.035041823983192444,
+      "learning_rate": 1.7113786727773528e-05,
+      "loss": 0.7029437065124512,
+      "step": 1780
+    },
+    {
+      "epoch": 0.2858568719433077,
+      "grad_norm": 0.03538225591182709,
+      "learning_rate": 1.7076654603176234e-05,
+      "loss": 0.7168496131896973,
+      "step": 1790
+    },
+    {
+      "epoch": 0.28745383770835414,
+      "grad_norm": 0.03331568092107773,
+      "learning_rate": 1.7039325987518623e-05,
+      "loss": 0.7009388923645019,
+      "step": 1800
+    },
+    {
+      "epoch": 0.28905080347340056,
+      "grad_norm": 0.033853549510240555,
+      "learning_rate": 1.700180191727057e-05,
+      "loss": 0.7036603450775146,
+      "step": 1810
+    },
+    {
+      "epoch": 0.29064776923844693,
+      "grad_norm": 0.03764677420258522,
+      "learning_rate": 1.696408343432895e-05,
+      "loss": 0.6595663547515869,
+      "step": 1820
+    },
+    {
+      "epoch": 0.29224473500349335,
+      "grad_norm": 0.028797749429941177,
+      "learning_rate": 1.6926171585988728e-05,
+      "loss": 0.6624193668365479,
+      "step": 1830
+    },
+    {
+      "epoch": 0.29384170076853977,
+      "grad_norm": 0.03950833901762962,
+      "learning_rate": 1.6888067424913863e-05,
+      "loss": 0.7282033920288086,
+      "step": 1840
+    },
+    {
+      "epoch": 0.2954386665335862,
+      "grad_norm": 0.032418642193078995,
+      "learning_rate": 1.6849772009108094e-05,
+      "loss": 0.6922025680541992,
+      "step": 1850
+    },
+    {
+      "epoch": 0.2970356322986326,
+      "grad_norm": 0.04753576219081879,
+      "learning_rate": 1.6811286401885554e-05,
+      "loss": 0.6543939590454102,
+      "step": 1860
+    },
+    {
+      "epoch": 0.29863259806367903,
+      "grad_norm": 0.03233656659722328,
+      "learning_rate": 1.677261167184125e-05,
+      "loss": 0.7344549655914306,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3002295638287254,
+      "grad_norm": 0.030969245359301567,
+      "learning_rate": 1.673374889282139e-05,
+      "loss": 0.6734820365905761,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3018265295937718,
+      "grad_norm": 0.03292040154337883,
+      "learning_rate": 1.6694699143893566e-05,
+      "loss": 0.6728582382202148,
+      "step": 1890
+    },
+    {
+      "epoch": 0.30342349535881824,
+      "grad_norm": 0.05033507198095322,
+      "learning_rate": 1.6655463509316797e-05,
+      "loss": 0.7132375240325928,
+      "step": 1900
+    },
+    {
+      "epoch": 0.30502046112386466,
+      "grad_norm": 0.03168455511331558,
+      "learning_rate": 1.6616043078511425e-05,
+      "loss": 0.7131926536560058,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3066174268889111,
+      "grad_norm": 0.03114555962383747,
+      "learning_rate": 1.657643894602885e-05,
+      "loss": 0.7184583187103272,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3082143926539575,
+      "grad_norm": 0.031201248988509178,
+      "learning_rate": 1.6536652211521155e-05,
+      "loss": 0.6606316566467285,
+      "step": 1930
+    },
+    {
+      "epoch": 0.30981135841900387,
+      "grad_norm": 0.035202883183956146,
+      "learning_rate": 1.6496683979710576e-05,
+      "loss": 0.7029526233673096,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3114083241840503,
+      "grad_norm": 0.03515475615859032,
+      "learning_rate": 1.6456535360358807e-05,
+      "loss": 0.6629764080047608,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3130052899490967,
+      "grad_norm": 0.03606207296252251,
+      "learning_rate": 1.6416207468236208e-05,
+      "loss": 0.6876577377319336,
+      "step": 1960
+    },
+    {
+      "epoch": 0.31460225571414313,
+      "grad_norm": 0.03580459579825401,
+      "learning_rate": 1.6375701423090846e-05,
+      "loss": 0.6879127979278564,
+      "step": 1970
+    },
+    {
+      "epoch": 0.31619922147918955,
+      "grad_norm": 0.048258859664201736,
+      "learning_rate": 1.6335018349617394e-05,
+      "loss": 0.7255002498626709,
+      "step": 1980
+    },
+    {
+      "epoch": 0.317796187244236,
+      "grad_norm": 0.046948857605457306,
+      "learning_rate": 1.629415937742591e-05,
+      "loss": 0.6919849395751954,
+      "step": 1990
+    },
+    {
+      "epoch": 0.31939315300928234,
+      "grad_norm": 0.03351585939526558,
+      "learning_rate": 1.625312564101049e-05,
+      "loss": 0.6886905193328857,
+      "step": 2000
+    },
+    {
+      "epoch": 0.31939315300928234,
+      "eval_loss": 0.6917140483856201,
+      "eval_runtime": 1706.5142,
+      "eval_samples_per_second": 3.262,
+      "eval_steps_per_second": 1.631,
+      "step": 2000
+    },
+    {
+      "epoch": 0.32099011877432876,
+      "grad_norm": 0.028385179117321968,
+      "learning_rate": 1.621191827971772e-05,
+      "loss": 0.7103724956512452,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3225870845393752,
+      "grad_norm": 0.02977878600358963,
+      "learning_rate": 1.617053843771509e-05,
+      "loss": 0.7012845516204834,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3241840503044216,
+      "grad_norm": 0.03950609639286995,
+      "learning_rate": 1.6128987263959212e-05,
+      "loss": 0.670585298538208,
+      "step": 2030
+    },
+    {
+      "epoch": 0.325781016069468,
+      "grad_norm": 0.03531676530838013,
+      "learning_rate": 1.6087265912163898e-05,
+      "loss": 0.6601489543914795,
+      "step": 2040
+    },
+    {
+      "epoch": 0.32737798183451444,
+      "grad_norm": 0.030362443998456,
+      "learning_rate": 1.6045375540768136e-05,
+      "loss": 0.7151473522186279,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3289749475995608,
+      "grad_norm": 0.03205285966396332,
+      "learning_rate": 1.600331731290395e-05,
+      "loss": 0.6645094394683838,
+      "step": 2060
+    },
+    {
+      "epoch": 0.33057191336460723,
+      "grad_norm": 0.028163282200694084,
+      "learning_rate": 1.5961092396364053e-05,
+      "loss": 0.6868838310241699,
+      "step": 2070
+    },
+    {
+      "epoch": 0.33216887912965365,
+      "grad_norm": 0.03369533643126488,
+      "learning_rate": 1.5918701963569475e-05,
+      "loss": 0.7029123783111573,
+      "step": 2080
+    },
+    {
+      "epoch": 0.33376584489470007,
+      "grad_norm": 0.0379788838326931,
+      "learning_rate": 1.5876147191536983e-05,
+      "loss": 0.7173356056213379,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3353628106597465,
+      "grad_norm": 0.0344555526971817,
+      "learning_rate": 1.583342926184639e-05,
+      "loss": 0.7149466037750244,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3369597764247929,
+      "grad_norm": 0.03237446770071983,
+      "learning_rate": 1.5790549360607772e-05,
+      "loss": 0.6467396259307862,
+      "step": 2110
+    },
+    {
+      "epoch": 0.33855674218983933,
+      "grad_norm": 0.032042454928159714,
+      "learning_rate": 1.5747508678428515e-05,
+      "loss": 0.6929362297058106,
+      "step": 2120
+    },
+    {
+      "epoch": 0.3401537079548857,
+      "grad_norm": 0.042478688061237335,
+      "learning_rate": 1.570430841038027e-05,
+      "loss": 0.6702353000640869,
+      "step": 2130
+    },
+    {
+      "epoch": 0.3417506737199321,
+      "grad_norm": 0.04559747502207756,
+      "learning_rate": 1.5660949755965757e-05,
+      "loss": 0.6600518226623535,
+      "step": 2140
+    },
+    {
+      "epoch": 0.34334763948497854,
+      "grad_norm": 0.03681457042694092,
+      "learning_rate": 1.5617433919085474e-05,
+      "loss": 0.6836549758911132,
+      "step": 2150
+    },
+    {
+      "epoch": 0.34494460525002496,
+      "grad_norm": 0.02465015836060047,
+      "learning_rate": 1.5573762108004262e-05,
+      "loss": 0.7304904460906982,
+      "step": 2160
+    },
+    {
+      "epoch": 0.3465415710150714,
+      "grad_norm": 0.03969291225075722,
+      "learning_rate": 1.5529935535317746e-05,
+      "loss": 0.6897444725036621,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3481385367801178,
+      "grad_norm": 0.03233686834573746,
+      "learning_rate": 1.548595541791869e-05,
+      "loss": 0.684897518157959,
+      "step": 2180
+    },
+    {
+      "epoch": 0.34973550254516417,
+      "grad_norm": 0.03138101100921631,
+      "learning_rate": 1.544182297696319e-05,
+      "loss": 0.7246172428131104,
+      "step": 2190
+    },
+    {
+      "epoch": 0.3513324683102106,
+      "grad_norm": 0.03732126206159592,
+      "learning_rate": 1.5397539437836765e-05,
+      "loss": 0.6686253070831298,
+      "step": 2200
+    },
+    {
+      "epoch": 0.352929434075257,
+      "grad_norm": 0.028484191745519638,
+      "learning_rate": 1.535310603012035e-05,
+      "loss": 0.6723896503448487,
+      "step": 2210
+    },
+    {
+      "epoch": 0.35452639984030343,
+      "grad_norm": 0.04804018512368202,
+      "learning_rate": 1.5308523987556144e-05,
+      "loss": 0.7210356712341308,
+      "step": 2220
+    },
+    {
+      "epoch": 0.35612336560534985,
+      "grad_norm": 0.035007379949092865,
+      "learning_rate": 1.5263794548013347e-05,
+      "loss": 0.6699332237243653,
+      "step": 2230
+    },
+    {
+      "epoch": 0.35772033137039627,
+      "grad_norm": 0.034309230744838715,
+      "learning_rate": 1.5218918953453808e-05,
+      "loss": 0.6913293838500977,
+      "step": 2240
+    },
+    {
+      "epoch": 0.35931729713544264,
+      "grad_norm": 0.05658178776502609,
+      "learning_rate": 1.5173898449897527e-05,
+      "loss": 0.6920335292816162,
+      "step": 2250
+    },
+    {
+      "epoch": 0.36091426290048906,
+      "grad_norm": 0.03614016994833946,
+      "learning_rate": 1.5128734287388051e-05,
+      "loss": 0.6855293273925781,
+      "step": 2260
+    },
+    {
+      "epoch": 0.3625112286655355,
+      "grad_norm": 0.0348992682993412,
+      "learning_rate": 1.5083427719957792e-05,
+      "loss": 0.7068172454833984,
+      "step": 2270
+    },
+    {
+      "epoch": 0.3641081944305819,
+      "grad_norm": 0.037384193390607834,
+      "learning_rate": 1.5037980005593178e-05,
+      "loss": 0.7026480197906494,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3657051601956283,
+      "grad_norm": 0.03162362426519394,
+      "learning_rate": 1.4992392406199741e-05,
+      "loss": 0.7369038581848144,
+      "step": 2290
+    },
+    {
+      "epoch": 0.36730212596067474,
+      "grad_norm": 0.03644828870892525,
+      "learning_rate": 1.4946666187567073e-05,
+      "loss": 0.7133200168609619,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3688990917257211,
+      "grad_norm": 0.05075901374220848,
+      "learning_rate": 1.490080261933368e-05,
+      "loss": 0.7146055221557617,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3704960574907675,
+      "grad_norm": 0.029153743758797646,
+      "learning_rate": 1.4854802974951732e-05,
+      "loss": 0.6756022453308106,
+      "step": 2320
+    },
+    {
+      "epoch": 0.37209302325581395,
+      "grad_norm": 0.030742282047867775,
+      "learning_rate": 1.4808668531651698e-05,
+      "loss": 0.6855404376983643,
+      "step": 2330
+    },
+    {
+      "epoch": 0.37368998902086037,
+      "grad_norm": 0.03525373339653015,
+      "learning_rate": 1.4762400570406889e-05,
+      "loss": 0.6626357078552246,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3752869547859068,
+      "grad_norm": 0.036246221512556076,
+      "learning_rate": 1.4716000375897897e-05,
+      "loss": 0.714638090133667,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3768839205509532,
+      "grad_norm": 0.03430463373661041,
+      "learning_rate": 1.4669469236476897e-05,
+      "loss": 0.6736037731170654,
+      "step": 2360
+    },
+    {
+      "epoch": 0.3784808863159996,
+      "grad_norm": 0.035327013581991196,
+      "learning_rate": 1.4622808444131904e-05,
+      "loss": 0.6907155036926269,
+      "step": 2370
+    },
+    {
+      "epoch": 0.380077852081046,
+      "grad_norm": 0.039804793894290924,
+      "learning_rate": 1.457601929445089e-05,
+      "loss": 0.6726024627685547,
+      "step": 2380
+    },
+    {
+      "epoch": 0.3816748178460924,
+      "grad_norm": 0.03689780458807945,
+      "learning_rate": 1.4529103086585803e-05,
+      "loss": 0.6932793617248535,
+      "step": 2390
+    },
+    {
+      "epoch": 0.38327178361113884,
+      "grad_norm": 0.03002229705452919,
+      "learning_rate": 1.4482061123216507e-05,
+      "loss": 0.6448677062988282,
+      "step": 2400
+    },
+    {
+      "epoch": 0.38486874937618526,
+      "grad_norm": 0.04287157952785492,
+      "learning_rate": 1.4434894710514596e-05,
+      "loss": 0.6573331356048584,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3864657151412317,
+      "grad_norm": 0.033608682453632355,
+      "learning_rate": 1.4387605158107146e-05,
+      "loss": 0.6761981010437011,
+      "step": 2420
+    },
+    {
+      "epoch": 0.38806268090627805,
+      "grad_norm": 0.0340590700507164,
+      "learning_rate": 1.4340193779040335e-05,
+      "loss": 0.7119544982910156,
+      "step": 2430
+    },
+    {
+      "epoch": 0.38965964667132447,
+      "grad_norm": 0.04576168581843376,
+      "learning_rate": 1.429266188974299e-05,
+      "loss": 0.722407054901123,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3912566124363709,
+      "grad_norm": 0.030462034046649933,
+      "learning_rate": 1.4245010809990052e-05,
+      "loss": 0.6741881847381592,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3928535782014173,
+      "grad_norm": 0.024982597678899765,
+      "learning_rate": 1.419724186286589e-05,
+      "loss": 0.6923768997192383,
+      "step": 2460
+    },
+    {
+      "epoch": 0.39445054396646373,
+      "grad_norm": 0.028493596240878105,
+      "learning_rate": 1.414935637472761e-05,
+      "loss": 0.7135504722595215,
+      "step": 2470
+    },
+    {
+      "epoch": 0.39604750973151015,
+      "grad_norm": 0.03309721499681473,
+      "learning_rate": 1.4101355675168197e-05,
+      "loss": 0.6870269298553466,
+      "step": 2480
+    },
+    {
+      "epoch": 0.3976444754965565,
+      "grad_norm": 0.033349502831697464,
+      "learning_rate": 1.4053241096979611e-05,
+      "loss": 0.7310259819030762,
+      "step": 2490
+    },
+    {
+      "epoch": 0.39924144126160294,
+      "grad_norm": 0.0456295944750309,
+      "learning_rate": 1.4005013976115777e-05,
+      "loss": 0.6850334644317627,
+      "step": 2500
+    },
+    {
+      "epoch": 0.39924144126160294,
+      "eval_loss": 0.6847082376480103,
+      "eval_runtime": 1727.6141,
+      "eval_samples_per_second": 3.222,
+      "eval_steps_per_second": 1.611,
+      "step": 2500
+    },
+    {
+      "epoch": 0.40083840702664936,
+      "grad_norm": 0.04150133952498436,
+      "learning_rate": 1.3956675651655489e-05,
+      "loss": 0.6748110771179199,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4024353727916958,
+      "grad_norm": 0.05328802019357681,
+      "learning_rate": 1.390822746576523e-05,
+      "loss": 0.6966916561126709,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4040323385567422,
+      "grad_norm": 0.04510056599974632,
+      "learning_rate": 1.3859670763661908e-05,
+      "loss": 0.6688936710357666,
+      "step": 2530
+    },
+    {
+      "epoch": 0.4056293043217886,
+      "grad_norm": 0.03648371994495392,
+      "learning_rate": 1.38110068935755e-05,
+      "loss": 0.6503877162933349,
+      "step": 2540
+    },
+    {
+      "epoch": 0.40722627008683504,
+      "grad_norm": 0.03958805277943611,
+      "learning_rate": 1.3762237206711624e-05,
+      "loss": 0.6709461212158203,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4088232358518814,
+      "grad_norm": 0.03620138764381409,
+      "learning_rate": 1.3713363057214006e-05,
+      "loss": 0.6775450706481934,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4104202016169278,
+      "grad_norm": 0.036233801394701004,
+      "learning_rate": 1.3664385802126904e-05,
+      "loss": 0.7039198875427246,
+      "step": 2570
+    },
+    {
+      "epoch": 0.41201716738197425,
+      "grad_norm": 0.034267205744981766,
+      "learning_rate": 1.3615306801357413e-05,
+      "loss": 0.6893480777740478,
+      "step": 2580
+    },
+    {
+      "epoch": 0.41361413314702067,
+      "grad_norm": 0.04056892171502113,
+      "learning_rate": 1.3566127417637698e-05,
+      "loss": 0.6852858543395997,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4152110989120671,
+      "grad_norm": 0.04654235392808914,
+      "learning_rate": 1.351684901648718e-05,
+      "loss": 0.6932817459106445,
+      "step": 2600
+    },
+    {
+      "epoch": 0.4168080646771135,
+      "grad_norm": 0.05177690088748932,
+      "learning_rate": 1.3467472966174595e-05,
+      "loss": 0.663325309753418,
+      "step": 2610
+    },
+    {
+      "epoch": 0.4184050304421599,
+      "grad_norm": 0.04320165514945984,
+      "learning_rate": 1.3418000637680026e-05,
+      "loss": 0.6278695106506348,
+      "step": 2620
+    },
+    {
+      "epoch": 0.4200019962072063,
+      "grad_norm": 0.0365922674536705,
+      "learning_rate": 1.336843340465682e-05,
+      "loss": 0.6674781799316406,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4215989619722527,
+      "grad_norm": 0.03406934812664986,
+      "learning_rate": 1.3318772643393447e-05,
+      "loss": 0.6559438705444336,
+      "step": 2640
+    },
+    {
+      "epoch": 0.42319592773729914,
+      "grad_norm": 0.030784646049141884,
+      "learning_rate": 1.3269019732775304e-05,
+      "loss": 0.7225832462310791,
+      "step": 2650
+    },
+    {
+      "epoch": 0.42479289350234556,
+      "grad_norm": 0.0360528789460659,
+      "learning_rate": 1.3219176054246404e-05,
+      "loss": 0.7170959949493408,
+      "step": 2660
+    },
+    {
+      "epoch": 0.426389859267392,
+      "grad_norm": 0.0355147160589695,
+      "learning_rate": 1.316924299177103e-05,
+      "loss": 0.683486795425415,
+      "step": 2670
+    },
+    {
+      "epoch": 0.42798682503243834,
+      "grad_norm": 0.046759720891714096,
+      "learning_rate": 1.3119221931795321e-05,
+      "loss": 0.7131591320037842,
+      "step": 2680
+    },
+    {
+      "epoch": 0.42958379079748477,
+      "grad_norm": 0.03554374352097511,
+      "learning_rate": 1.3069114263208746e-05,
+      "loss": 0.6715425491333008,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4311807565625312,
+      "grad_norm": 0.04263368248939514,
+      "learning_rate": 1.3018921377305574e-05,
+      "loss": 0.6871604442596435,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4327777223275776,
+      "grad_norm": 0.04496333748102188,
+      "learning_rate": 1.2968644667746207e-05,
+      "loss": 0.6588253021240235,
+      "step": 2710
+    },
+    {
+      "epoch": 0.43437468809262403,
+      "grad_norm": 0.036086685955524445,
+      "learning_rate": 1.2918285530518519e-05,
+      "loss": 0.6757514953613282,
+      "step": 2720
+    },
+    {
+      "epoch": 0.43597165385767045,
+      "grad_norm": 0.03308691456913948,
+      "learning_rate": 1.2867845363899075e-05,
+      "loss": 0.6658079624176025,
+      "step": 2730
+    },
+    {
+      "epoch": 0.4375686196227168,
+      "grad_norm": 0.04500787332653999,
+      "learning_rate": 1.2817325568414299e-05,
+      "loss": 0.6812032222747803,
+      "step": 2740
+    },
+    {
+      "epoch": 0.43916558538776324,
+      "grad_norm": 0.03841916844248772,
+      "learning_rate": 1.2766727546801612e-05,
+      "loss": 0.6864322662353516,
+      "step": 2750
+    },
+    {
+      "epoch": 0.44076255115280966,
+      "grad_norm": 0.037914253771305084,
+      "learning_rate": 1.2716052703970459e-05,
+      "loss": 0.7053091526031494,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4423595169178561,
+      "grad_norm": 0.03686266764998436,
+      "learning_rate": 1.2665302446963312e-05,
+      "loss": 0.7154440879821777,
+      "step": 2770
+    },
+    {
+      "epoch": 0.4439564826829025,
+      "grad_norm": 0.04227980226278305,
+      "learning_rate": 1.261447818491661e-05,
+      "loss": 0.716032886505127,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4455534484479489,
+      "grad_norm": 0.04683419317007065,
+      "learning_rate": 1.2563581329021608e-05,
+      "loss": 0.6738180637359619,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4471504142129953,
+      "grad_norm": 0.05264711007475853,
+      "learning_rate": 1.2512613292485218e-05,
+      "loss": 0.7129797458648681,
+      "step": 2800
+    },
+    {
+      "epoch": 0.4487473799780417,
+      "grad_norm": 0.04030587151646614,
+      "learning_rate": 1.2461575490490753e-05,
+      "loss": 0.727331018447876,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4503443457430881,
+      "grad_norm": 0.04141924902796745,
+      "learning_rate": 1.2410469340158655e-05,
+      "loss": 0.676334810256958,
+      "step": 2820
+    },
+    {
+      "epoch": 0.45194131150813455,
+      "grad_norm": 0.04476945474743843,
+      "learning_rate": 1.2359296260507117e-05,
+      "loss": 0.6823254108428956,
+      "step": 2830
+    },
+    {
+      "epoch": 0.45353827727318097,
+      "grad_norm": 0.03869614750146866,
+      "learning_rate": 1.23080576724127e-05,
+      "loss": 0.6882941722869873,
+      "step": 2840
+    },
+    {
+      "epoch": 0.4551352430382274,
+      "grad_norm": 0.031092161312699318,
+      "learning_rate": 1.2256754998570887e-05,
+      "loss": 0.6744899272918701,
+      "step": 2850
+    },
+    {
+      "epoch": 0.45673220880327375,
+      "grad_norm": 0.03507015109062195,
+      "learning_rate": 1.2205389663456566e-05,
+      "loss": 0.7103989124298096,
+      "step": 2860
+    },
+    {
+      "epoch": 0.4583291745683202,
+      "grad_norm": 0.032851576805114746,
+      "learning_rate": 1.2153963093284483e-05,
+      "loss": 0.7116181373596191,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4599261403333666,
+      "grad_norm": 0.05045435577630997,
+      "learning_rate": 1.210247671596965e-05,
+      "loss": 0.6489015579223633,
+      "step": 2880
+    },
+    {
+      "epoch": 0.461523106098413,
+      "grad_norm": 0.032028377056121826,
+      "learning_rate": 1.2050931961087673e-05,
+      "loss": 0.6941986560821534,
+      "step": 2890
+    },
+    {
+      "epoch": 0.46312007186345944,
+      "grad_norm": 0.038908179849386215,
+      "learning_rate": 1.1999330259835096e-05,
+      "loss": 0.6908363342285156,
+      "step": 2900
+    },
+    {
+      "epoch": 0.46471703762850586,
+      "grad_norm": 0.03464297205209732,
+      "learning_rate": 1.1947673044989627e-05,
+      "loss": 0.7050665378570556,
+      "step": 2910
+    },
+    {
+      "epoch": 0.4663140033935522,
+      "grad_norm": 0.028774775564670563,
+      "learning_rate": 1.1895961750870375e-05,
+      "loss": 0.6700276851654052,
+      "step": 2920
+    },
+    {
+      "epoch": 0.46791096915859864,
+      "grad_norm": 0.03465382754802704,
+      "learning_rate": 1.1844197813298018e-05,
+      "loss": 0.6821481227874756,
+      "step": 2930
+    },
+    {
+      "epoch": 0.46950793492364506,
+      "grad_norm": 0.03162831813097,
+      "learning_rate": 1.1792382669554938e-05,
+      "loss": 0.6927790164947509,
+      "step": 2940
+    },
+    {
+      "epoch": 0.4711049006886915,
+      "grad_norm": 0.03354581072926521,
+      "learning_rate": 1.1740517758345312e-05,
+      "loss": 0.7100222110748291,
+      "step": 2950
+    },
+    {
+      "epoch": 0.4727018664537379,
+      "grad_norm": 0.033408552408218384,
+      "learning_rate": 1.168860451975516e-05,
+      "loss": 0.6876598358154297,
+      "step": 2960
+    },
+    {
+      "epoch": 0.4742988322187843,
+      "grad_norm": 0.04334475100040436,
+      "learning_rate": 1.1636644395212375e-05,
+      "loss": 0.6866564273834228,
+      "step": 2970
+    },
+    {
+      "epoch": 0.47589579798383075,
+      "grad_norm": 0.040464501827955246,
+      "learning_rate": 1.1584638827446684e-05,
+      "loss": 0.6588007926940918,
+      "step": 2980
+    },
+    {
+      "epoch": 0.4774927637488771,
+      "grad_norm": 0.04957546666264534,
+      "learning_rate": 1.1532589260449589e-05,
+      "loss": 0.7070925235748291,
+      "step": 2990
+    },
+    {
+      "epoch": 0.47908972951392353,
+      "grad_norm": 0.038624707609415054,
+      "learning_rate": 1.1480497139434294e-05,
+      "loss": 0.6919936180114746,
+      "step": 3000
+    },
+    {
+      "epoch": 0.47908972951392353,
+      "eval_loss": 0.6803271770477295,
+      "eval_runtime": 1741.9445,
+      "eval_samples_per_second": 3.195,
+      "eval_steps_per_second": 1.598,
+      "step": 3000
+    },
+    {
+      "epoch": 0.48068669527896996,
+      "grad_norm": 0.04100416228175163,
+      "learning_rate": 1.1428363910795547e-05,
+      "loss": 0.6852492809295654,
+      "step": 3010
+    },
+    {
+      "epoch": 0.4822836610440164,
+      "grad_norm": 0.033158283680677414,
+      "learning_rate": 1.1376191022069501e-05,
+      "loss": 0.667516565322876,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4838806268090628,
+      "grad_norm": 0.03429755941033363,
+      "learning_rate": 1.1323979921893514e-05,
+      "loss": 0.6956333160400391,
+      "step": 3030
+    },
+    {
+      "epoch": 0.4854775925741092,
+      "grad_norm": 0.038232944905757904,
+      "learning_rate": 1.1271732059965925e-05,
+      "loss": 0.6716940402984619,
+      "step": 3040
+    },
+    {
+      "epoch": 0.4870745583391556,
+      "grad_norm": 0.039248283952474594,
+      "learning_rate": 1.1219448887005805e-05,
+      "loss": 0.6984889507293701,
+      "step": 3050
+    },
+    {
+      "epoch": 0.488671524104202,
+      "grad_norm": 0.03670873865485191,
+      "learning_rate": 1.1167131854712676e-05,
+      "loss": 0.6709868907928467,
+      "step": 3060
+    },
+    {
+      "epoch": 0.4902684898692484,
+      "grad_norm": 0.038310978561639786,
+      "learning_rate": 1.11147824157262e-05,
+      "loss": 0.6672838211059571,
+      "step": 3070
+    },
+    {
+      "epoch": 0.49186545563429485,
+      "grad_norm": 0.049189358949661255,
+      "learning_rate": 1.1062402023585846e-05,
+      "loss": 0.6705825805664063,
+      "step": 3080
+    },
+    {
+      "epoch": 0.49346242139934127,
+      "grad_norm": 0.0355689600110054,
+      "learning_rate": 1.1009992132690523e-05,
+      "loss": 0.6964725017547607,
+      "step": 3090
+    },
+    {
+      "epoch": 0.4950593871643877,
+      "grad_norm": 0.03348139300942421,
+      "learning_rate": 1.0957554198258225e-05,
+      "loss": 0.6852751255035401,
+      "step": 3100
+    },
+    {
+      "epoch": 0.49665635292943405,
+      "grad_norm": 0.03575340285897255,
+      "learning_rate": 1.0905089676285588e-05,
+      "loss": 0.670794153213501,
+      "step": 3110
+    },
+    {
+      "epoch": 0.4982533186944805,
+      "grad_norm": 0.04119575023651123,
+      "learning_rate": 1.0852600023507482e-05,
+      "loss": 0.6651177406311035,
+      "step": 3120
+    },
+    {
+      "epoch": 0.4998502844595269,
+      "grad_norm": 0.04122074693441391,
+      "learning_rate": 1.0800086697356576e-05,
+      "loss": 0.6976897239685058,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5014472502245733,
+      "grad_norm": 0.034223176538944244,
+      "learning_rate": 1.0747551155922836e-05,
+      "loss": 0.6890422821044921,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5030442159896197,
+      "grad_norm": 0.03588424623012543,
+      "learning_rate": 1.069499485791307e-05,
+      "loss": 0.702085018157959,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5046411817546661,
+      "grad_norm": 0.03808250650763512,
+      "learning_rate": 1.0642419262610417e-05,
+      "loss": 0.6910345554351807,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5062381475197125,
+      "grad_norm": 0.03852913901209831,
+      "learning_rate": 1.0589825829833825e-05,
+      "loss": 0.6685952186584473,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5078351132847589,
+      "grad_norm": 0.04528075084090233,
+      "learning_rate": 1.0537216019897514e-05,
+      "loss": 0.6971333980560303,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5094320790498054,
+      "grad_norm": 0.048296891152858734,
+      "learning_rate": 1.0484591293570437e-05,
+      "loss": 0.6506364345550537,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5110290448148518,
+      "grad_norm": 0.03280564025044441,
+      "learning_rate": 1.0431953112035722e-05,
+      "loss": 0.7106664657592774,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5126260105798982,
+      "grad_norm": 0.03674250841140747,
+      "learning_rate": 1.0379302936850083e-05,
+      "loss": 0.6798049449920655,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5142229763449446,
+      "grad_norm": 0.02851015329360962,
+      "learning_rate": 1.0326642229903265e-05,
+      "loss": 0.6674058914184571,
+      "step": 3220
+    },
+    {
+      "epoch": 0.515819942109991,
+      "grad_norm": 0.04876931011676788,
+      "learning_rate": 1.0273972453377433e-05,
+      "loss": 0.6956850051879883,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5174169078750375,
+      "grad_norm": 0.04633413255214691,
+      "learning_rate": 1.022129506970657e-05,
+      "loss": 0.6859352111816406,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5190138736400839,
+      "grad_norm": 0.03311692923307419,
+      "learning_rate": 1.0168611541535898e-05,
+      "loss": 0.7103584289550782,
+      "step": 3250
+    },
+    {
+      "epoch": 0.5206108394051302,
+      "grad_norm": 0.047599345445632935,
+      "learning_rate": 1.0115923331681231e-05,
+      "loss": 0.6435458183288574,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5222078051701766,
+      "grad_norm": 0.033676277846097946,
+      "learning_rate": 1.0063231903088384e-05,
+      "loss": 0.6581607341766358,
+      "step": 3270
+    },
+    {
+      "epoch": 0.523804770935223,
+      "grad_norm": 0.04461989924311638,
+      "learning_rate": 1.0010538718792544e-05,
+      "loss": 0.7152835369110108,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5254017367002695,
+      "grad_norm": 0.04492766410112381,
+      "learning_rate": 9.957845241877639e-06,
+      "loss": 0.7056149482727051,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5269987024653159,
+      "grad_norm": 0.02856948785483837,
+      "learning_rate": 9.905152935435733e-06,
+      "loss": 0.6655986785888672,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5285956682303623,
+      "grad_norm": 0.0438731387257576,
+      "learning_rate": 9.852463262526383e-06,
+      "loss": 0.7045872688293457,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5301926339954087,
+      "grad_norm": 0.04025017097592354,
+      "learning_rate": 9.799777686136025e-06,
+      "loss": 0.6844919204711915,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5317895997604551,
+      "grad_norm": 0.04453453794121742,
+      "learning_rate": 9.747097669137352e-06,
+      "loss": 0.6857856750488281,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5333865655255016,
+      "grad_norm": 0.034065935760736465,
+      "learning_rate": 9.694424674248687e-06,
+      "loss": 0.6782355785369873,
+      "step": 3340
+    },
+    {
+      "epoch": 0.534983531290548,
+      "grad_norm": 0.03808142989873886,
+      "learning_rate": 9.641760163993394e-06,
+      "loss": 0.7040555000305175,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5365804970555944,
+      "grad_norm": 0.03482348844408989,
+      "learning_rate": 9.58910560065923e-06,
+      "loss": 0.6730469226837158,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5381774628206408,
+      "grad_norm": 0.04541369527578354,
+      "learning_rate": 9.536462446257777e-06,
+      "loss": 0.7102465152740478,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5397744285856872,
+      "grad_norm": 0.043802615255117416,
+      "learning_rate": 9.483832162483841e-06,
+      "loss": 0.6941490173339844,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5413713943507336,
+      "grad_norm": 0.03652375563979149,
+      "learning_rate": 9.431216210674849e-06,
+      "loss": 0.704205322265625,
+      "step": 3390
+    },
+    {
+      "epoch": 0.54296836011578,
+      "grad_norm": 0.03886988013982773,
+      "learning_rate": 9.378616051770287e-06,
+      "loss": 0.6582465171813965,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5445653258808264,
+      "grad_norm": 0.044551681727170944,
+      "learning_rate": 9.326033146271144e-06,
+      "loss": 0.7078223705291748,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5461622916458728,
+      "grad_norm": 0.03762778267264366,
+      "learning_rate": 9.273468954199333e-06,
+      "loss": 0.6956658840179444,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5477592574109192,
+      "grad_norm": 0.03794670104980469,
+      "learning_rate": 9.220924935057184e-06,
+      "loss": 0.6778861522674561,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5493562231759657,
+      "grad_norm": 0.03766616806387901,
+      "learning_rate": 9.168402547786897e-06,
+      "loss": 0.6759885311126709,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5509531889410121,
+      "grad_norm": 0.04172630235552788,
+      "learning_rate": 9.115903250730035e-06,
+      "loss": 0.6739662170410157,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5525501547060585,
+      "grad_norm": 0.04936975613236427,
+      "learning_rate": 9.063428501587043e-06,
+      "loss": 0.6581857681274415,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5541471204711049,
+      "grad_norm": 0.029507510364055634,
+      "learning_rate": 9.010979757376766e-06,
+      "loss": 0.6508080005645752,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5557440862361513,
+      "grad_norm": 0.03853528946638107,
+      "learning_rate": 8.958558474395987e-06,
+      "loss": 0.6520394325256348,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5573410520011978,
+      "grad_norm": 0.04168470576405525,
+      "learning_rate": 8.906166108178999e-06,
+      "loss": 0.6652052402496338,
+      "step": 3490
+    },
+    {
+      "epoch": 0.5589380177662442,
+      "grad_norm": 0.05398240312933922,
+      "learning_rate": 8.853804113457203e-06,
+      "loss": 0.6777332782745361,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5589380177662442,
+      "eval_loss": 0.677163302898407,
+      "eval_runtime": 1733.3138,
+      "eval_samples_per_second": 3.211,
+      "eval_steps_per_second": 1.606,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5605349835312905,
+      "grad_norm": 0.03181201219558716,
+      "learning_rate": 8.801473944118683e-06,
+      "loss": 0.687204122543335,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5621319492963369,
+      "grad_norm": 0.03610098734498024,
+      "learning_rate": 8.749177053167865e-06,
+      "loss": 0.7028386116027832,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5637289150613833,
+      "grad_norm": 0.0450638048350811,
+      "learning_rate": 8.696914892685172e-06,
+      "loss": 0.6494070529937744,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5653258808264298,
+      "grad_norm": 0.053478166460990906,
+      "learning_rate": 8.644688913786678e-06,
+      "loss": 0.6962610721588135,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5669228465914762,
+      "grad_norm": 0.03924020007252693,
+      "learning_rate": 8.592500566583856e-06,
+      "loss": 0.6941705703735351,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5685198123565226,
+      "grad_norm": 0.037506818771362305,
+      "learning_rate": 8.540351300143284e-06,
+      "loss": 0.6824637413024902,
+      "step": 3560
+    },
+    {
+      "epoch": 0.570116778121569,
+      "grad_norm": 0.03883660212159157,
+      "learning_rate": 8.488242562446416e-06,
+      "loss": 0.7028771877288819,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5717137438866154,
+      "grad_norm": 0.047728635370731354,
+      "learning_rate": 8.436175800349393e-06,
+      "loss": 0.7102296829223633,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5733107096516619,
+      "grad_norm": 0.0423489511013031,
+      "learning_rate": 8.384152459542849e-06,
+      "loss": 0.7113842487335205,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5749076754167083,
+      "grad_norm": 0.034012071788311005,
+      "learning_rate": 8.332173984511774e-06,
+      "loss": 0.6995905876159668,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5765046411817547,
+      "grad_norm": 0.02976168505847454,
+      "learning_rate": 8.280241818495423e-06,
+      "loss": 0.7066914558410644,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5781016069468011,
+      "grad_norm": 0.04333082213997841,
+      "learning_rate": 8.228357403447225e-06,
+      "loss": 0.6559367656707764,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5796985727118474,
+      "grad_norm": 0.0726407989859581,
+      "learning_rate": 8.176522179994744e-06,
+      "loss": 0.691002082824707,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5812955384768939,
+      "grad_norm": 0.035537876188755035,
+      "learning_rate": 8.124737587399697e-06,
+      "loss": 0.7173010349273682,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5828925042419403,
+      "grad_norm": 0.036615390330553055,
+      "learning_rate": 8.073005063517973e-06,
+      "loss": 0.712274169921875,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5844894700069867,
+      "grad_norm": 0.02762191742658615,
+      "learning_rate": 8.021326044759718e-06,
+      "loss": 0.7043851852416992,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5860864357720331,
+      "grad_norm": 0.033354733139276505,
+      "learning_rate": 7.969701966049447e-06,
+      "loss": 0.6736726284027099,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5876834015370795,
+      "grad_norm": 0.04251250997185707,
+      "learning_rate": 7.918134260786214e-06,
+      "loss": 0.679281759262085,
+      "step": 3680
+    },
+    {
+      "epoch": 0.589280367302126,
+      "grad_norm": 0.03284266218543053,
+      "learning_rate": 7.866624360803786e-06,
+      "loss": 0.6851751804351807,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5908773330671724,
+      "grad_norm": 0.050126783549785614,
+      "learning_rate": 7.81517369633092e-06,
+      "loss": 0.6820148468017578,
+      "step": 3700
+    },
+    {
+      "epoch": 0.5924742988322188,
+      "grad_norm": 0.042198847979307175,
+      "learning_rate": 7.763783695951623e-06,
+      "loss": 0.6824104309082031,
+      "step": 3710
+    },
+    {
+      "epoch": 0.5940712645972652,
+      "grad_norm": 0.03443041443824768,
+      "learning_rate": 7.7124557865655e-06,
+      "loss": 0.670203685760498,
+      "step": 3720
+    },
+    {
+      "epoch": 0.5956682303623116,
+      "grad_norm": 0.036490943282842636,
+      "learning_rate": 7.661191393348136e-06,
+      "loss": 0.7197192192077637,
+      "step": 3730
+    },
+    {
+      "epoch": 0.5972651961273581,
+      "grad_norm": 0.03636472299695015,
+      "learning_rate": 7.609991939711517e-06,
+      "loss": 0.6888130187988282,
+      "step": 3740
+    },
+    {
+      "epoch": 0.5988621618924044,
+      "grad_norm": 0.040862370282411575,
+      "learning_rate": 7.558858847264502e-06,
+      "loss": 0.6817411422729492,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6004591276574508,
+      "grad_norm": 0.036771345883607864,
+      "learning_rate": 7.507793535773377e-06,
+      "loss": 0.7020951271057129,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6020560934224972,
+      "grad_norm": 0.036549679934978485,
+      "learning_rate": 7.456797423122401e-06,
+      "loss": 0.6926814079284668,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6036530591875436,
+      "grad_norm": 0.03661128878593445,
+      "learning_rate": 7.405871925274457e-06,
+      "loss": 0.7070647716522217,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6052500249525901,
+      "grad_norm": 0.05932987853884697,
+      "learning_rate": 7.3550184562317295e-06,
+      "loss": 0.7125431060791015,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6068469907176365,
+      "grad_norm": 0.03799287602305412,
+      "learning_rate": 7.3042384279964394e-06,
+      "loss": 0.6607150554656982,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6084439564826829,
+      "grad_norm": 0.031429167836904526,
+      "learning_rate": 7.253533250531656e-06,
+      "loss": 0.6806385517120361,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6100409222477293,
+      "grad_norm": 0.05243317410349846,
+      "learning_rate": 7.202904331722127e-06,
+      "loss": 0.6834061145782471,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6116378880127757,
+      "grad_norm": 0.0379512719810009,
+      "learning_rate": 7.152353077335189e-06,
+      "loss": 0.665333604812622,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6132348537778222,
+      "grad_norm": 0.03895451873540878,
+      "learning_rate": 7.101880890981751e-06,
+      "loss": 0.6566737651824951,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6148318195428686,
+      "grad_norm": 0.03916362673044205,
+      "learning_rate": 7.0514891740773105e-06,
+      "loss": 0.7027400970458985,
+      "step": 3850
+    },
+    {
+      "epoch": 0.616428785307915,
+      "grad_norm": 0.03747180104255676,
+      "learning_rate": 7.001179325803037e-06,
+      "loss": 0.6675007820129395,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6180257510729614,
+      "grad_norm": 0.04694506898522377,
+      "learning_rate": 6.95095274306693e-06,
+      "loss": 0.6993120193481446,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6196227168380077,
+      "grad_norm": 0.035580579191446304,
+      "learning_rate": 6.900810820465036e-06,
+      "loss": 0.7056044578552246,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6212196826030542,
+      "grad_norm": 0.03560471907258034,
+      "learning_rate": 6.8507549502427105e-06,
+      "loss": 0.7067145824432373,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6228166483681006,
+      "grad_norm": 0.03395050764083862,
+      "learning_rate": 6.800786522255974e-06,
+      "loss": 0.6799113750457764,
+      "step": 3900
+    },
+    {
+      "epoch": 0.624413614133147,
+      "grad_norm": 0.04874083772301674,
+      "learning_rate": 6.7509069239329295e-06,
+      "loss": 0.6687004089355468,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6260105798981934,
+      "grad_norm": 0.03185804560780525,
+      "learning_rate": 6.701117540235204e-06,
+      "loss": 0.6309501171112061,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6276075456632398,
+      "grad_norm": 0.04211097210645676,
+      "learning_rate": 6.651419753619535e-06,
+      "loss": 0.7135450839996338,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6292045114282863,
+      "grad_norm": 0.03911832720041275,
+      "learning_rate": 6.601814943999363e-06,
+      "loss": 0.7103962421417236,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6308014771933327,
+      "grad_norm": 0.033491455018520355,
+      "learning_rate": 6.552304488706512e-06,
+      "loss": 0.7135549068450928,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6323984429583791,
+      "grad_norm": 0.031158218160271645,
+      "learning_rate": 6.502889762452969e-06,
+      "loss": 0.6667858600616455,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6339954087234255,
+      "grad_norm": 0.037267692387104034,
+      "learning_rate": 6.453572137292689e-06,
+      "loss": 0.7077006816864013,
+      "step": 3970
+    },
+    {
+      "epoch": 0.635592374488472,
+      "grad_norm": 0.03646966442465782,
+      "learning_rate": 6.404352982583511e-06,
+      "loss": 0.663360595703125,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6371893402535184,
+      "grad_norm": 0.0357484444975853,
+      "learning_rate": 6.355233664949132e-06,
+      "loss": 0.6508517742156983,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6387863060185647,
+      "grad_norm": 0.03508693352341652,
+      "learning_rate": 6.306215548241174e-06,
+      "loss": 0.669168758392334,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6387863060185647,
+      "eval_loss": 0.6750220060348511,
+      "eval_runtime": 1753.5267,
+      "eval_samples_per_second": 3.174,
+      "eval_steps_per_second": 1.587,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6403832717836111,
+      "grad_norm": 0.03497602418065071,
+      "learning_rate": 6.257299993501289e-06,
+      "loss": 0.6791082859039307,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6419802375486575,
+      "grad_norm": 0.04195614159107208,
+      "learning_rate": 6.208488358923393e-06,
+      "loss": 0.6831603050231934,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6435772033137039,
+      "grad_norm": 0.049627698957920074,
+      "learning_rate": 6.1597819998159506e-06,
+      "loss": 0.7127910137176514,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6451741690787504,
+      "grad_norm": 0.03807014226913452,
+      "learning_rate": 6.111182268564322e-06,
+      "loss": 0.6687079429626465,
+      "step": 4040
+    },
+    {
+      "epoch": 0.6467711348437968,
+      "grad_norm": 0.0430082343518734,
+      "learning_rate": 6.062690514593237e-06,
+      "loss": 0.6742999076843261,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6483681006088432,
+      "grad_norm": 0.032011304050683975,
+      "learning_rate": 6.014308084329327e-06,
+      "loss": 0.693959903717041,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6499650663738896,
+      "grad_norm": 0.0382765531539917,
+      "learning_rate": 5.966036321163709e-06,
+      "loss": 0.6869026184082031,
+      "step": 4070
+    },
+    {
+      "epoch": 0.651562032138936,
+      "grad_norm": 0.038456711918115616,
+      "learning_rate": 5.917876565414729e-06,
+      "loss": 0.6802726268768311,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6531589979039825,
+      "grad_norm": 0.05505223199725151,
+      "learning_rate": 5.8698301542907144e-06,
+      "loss": 0.6540178775787353,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6547559636690289,
+      "grad_norm": 0.06022993102669716,
+      "learning_rate": 5.8218984218528514e-06,
+      "loss": 0.6731237888336181,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6563529294340753,
+      "grad_norm": 0.04032977297902107,
+      "learning_rate": 5.774082698978154e-06,
+      "loss": 0.7120064258575439,
+      "step": 4110
+    },
+    {
+      "epoch": 0.6579498951991216,
+      "grad_norm": 0.04511059820652008,
+      "learning_rate": 5.726384313322504e-06,
+      "loss": 0.6774604797363282,
+      "step": 4120
+    },
+    {
+      "epoch": 0.659546860964168,
+      "grad_norm": 0.04290354251861572,
+      "learning_rate": 5.678804589283777e-06,
+      "loss": 0.6660148143768311,
+      "step": 4130
+    },
+    {
+      "epoch": 0.6611438267292145,
+      "grad_norm": 0.03994971513748169,
+      "learning_rate": 5.6313448479650944e-06,
+      "loss": 0.6876091957092285,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6627407924942609,
+      "grad_norm": 0.03699030727148056,
+      "learning_rate": 5.584006407138114e-06,
+      "loss": 0.6590198516845703,
+      "step": 4150
+    },
+    {
+      "epoch": 0.6643377582593073,
+      "grad_norm": 0.041105858981609344,
+      "learning_rate": 5.536790581206451e-06,
+      "loss": 0.6922510147094727,
+      "step": 4160
+    },
+    {
+      "epoch": 0.6659347240243537,
+      "grad_norm": 0.036437924951314926,
+      "learning_rate": 5.489698681169196e-06,
+      "loss": 0.7056612968444824,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6675316897894001,
+      "grad_norm": 0.03376461938023567,
+      "learning_rate": 5.442732014584495e-06,
+      "loss": 0.6548415184020996,
+      "step": 4180
+    },
+    {
+      "epoch": 0.6691286555544466,
+      "grad_norm": 0.042223311960697174,
+      "learning_rate": 5.395891885533244e-06,
+      "loss": 0.6739361763000489,
+      "step": 4190
+    },
+    {
+      "epoch": 0.670725621319493,
+      "grad_norm": 0.03633170202374458,
+      "learning_rate": 5.349179594582898e-06,
+      "loss": 0.7088988304138184,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6723225870845394,
+      "grad_norm": 0.037932299077510834,
+      "learning_rate": 5.302596438751339e-06,
+      "loss": 0.693552827835083,
+      "step": 4210
+    },
+    {
+      "epoch": 0.6739195528495858,
+      "grad_norm": 0.03990362584590912,
+      "learning_rate": 5.256143711470869e-06,
+      "loss": 0.6921378135681152,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6755165186146322,
+      "grad_norm": 0.037905026227235794,
+      "learning_rate": 5.20982270255231e-06,
+      "loss": 0.6850985050201416,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6771134843796787,
+      "grad_norm": 0.0379810705780983,
+      "learning_rate": 5.163634698149163e-06,
+      "loss": 0.6650650978088379,
+      "step": 4240
+    },
+    {
+      "epoch": 0.678710450144725,
+      "grad_norm": 0.02918444201350212,
+      "learning_rate": 5.11758098072193e-06,
+      "loss": 0.6373987674713135,
+      "step": 4250
+    },
+    {
+      "epoch": 0.6803074159097714,
+      "grad_norm": 0.03856576979160309,
+      "learning_rate": 5.071662829002477e-06,
+      "loss": 0.7125391483306884,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6819043816748178,
+      "grad_norm": 0.03352683037519455,
+      "learning_rate": 5.02588151795854e-06,
+      "loss": 0.6727629661560058,
+      "step": 4270
+    },
+    {
+      "epoch": 0.6835013474398642,
+      "grad_norm": 0.04891199618577957,
+      "learning_rate": 4.98023831875833e-06,
+      "loss": 0.722879695892334,
+      "step": 4280
+    },
+    {
+      "epoch": 0.6850983132049107,
+      "grad_norm": 0.04167689383029938,
+      "learning_rate": 4.9347344987352305e-06,
+      "loss": 0.6740260124206543,
+      "step": 4290
+    },
+    {
+      "epoch": 0.6866952789699571,
+      "grad_norm": 0.03634532913565636,
+      "learning_rate": 4.889371321352607e-06,
+      "loss": 0.649868106842041,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6882922447350035,
+      "grad_norm": 0.04563901200890541,
+      "learning_rate": 4.844150046168722e-06,
+      "loss": 0.6745593547821045,
+      "step": 4310
+    },
+    {
+      "epoch": 0.6898892105000499,
+      "grad_norm": 0.04015154391527176,
+      "learning_rate": 4.799071928801784e-06,
+      "loss": 0.7058079719543457,
+      "step": 4320
+    },
+    {
+      "epoch": 0.6914861762650963,
+      "grad_norm": 0.03595186024904251,
+      "learning_rate": 4.7541382208950505e-06,
+      "loss": 0.6532906532287598,
+      "step": 4330
+    },
+    {
+      "epoch": 0.6930831420301428,
+      "grad_norm": 0.030780136585235596,
+      "learning_rate": 4.709350170082103e-06,
+      "loss": 0.6842369079589844,
+      "step": 4340
+    },
+    {
+      "epoch": 0.6946801077951892,
+      "grad_norm": 0.04692146182060242,
+      "learning_rate": 4.6647090199522025e-06,
+      "loss": 0.7000434398651123,
+      "step": 4350
+    },
+    {
+      "epoch": 0.6962770735602356,
+      "grad_norm": 0.0310053788125515,
+      "learning_rate": 4.620216010015725e-06,
+      "loss": 0.6727443218231202,
+      "step": 4360
+    },
+    {
+      "epoch": 0.6978740393252819,
+      "grad_norm": 0.03317731246352196,
+      "learning_rate": 4.575872375669793e-06,
+      "loss": 0.6653543949127197,
+      "step": 4370
+    },
+    {
+      "epoch": 0.6994710050903283,
+      "grad_norm": 0.05431032553315163,
+      "learning_rate": 4.531679348163955e-06,
+      "loss": 0.691961145401001,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7010679708553748,
+      "grad_norm": 0.037718210369348526,
+      "learning_rate": 4.487638154565978e-06,
+      "loss": 0.6595858573913574,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7026649366204212,
+      "grad_norm": 0.03125704079866409,
+      "learning_rate": 4.443750017727814e-06,
+      "loss": 0.7066734790802002,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7042619023854676,
+      "grad_norm": 0.050476983189582825,
+      "learning_rate": 4.400016156251611e-06,
+      "loss": 0.6906495094299316,
+      "step": 4410
+    },
+    {
+      "epoch": 0.705858868150514,
+      "grad_norm": 0.034239206463098526,
+      "learning_rate": 4.356437784455896e-06,
+      "loss": 0.698778486251831,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7074558339155604,
+      "grad_norm": 0.03537076339125633,
+      "learning_rate": 4.313016112341861e-06,
+      "loss": 0.6730549335479736,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7090527996806069,
+      "grad_norm": 0.0366840697824955,
+      "learning_rate": 4.269752345559761e-06,
+      "loss": 0.6689012527465821,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7106497654456533,
+      "grad_norm": 0.04745423048734665,
+      "learning_rate": 4.226647685375428e-06,
+      "loss": 0.6668778419494629,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7122467312106997,
+      "grad_norm": 0.043418820947408676,
+      "learning_rate": 4.183703328636924e-06,
+      "loss": 0.6641871929168701,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7138436969757461,
+      "grad_norm": 0.0369136743247509,
+      "learning_rate": 4.140920467741325e-06,
+      "loss": 0.6592461585998535,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7154406627407925,
+      "grad_norm": 0.034368593245744705,
+      "learning_rate": 4.098300290601581e-06,
+      "loss": 0.6683683395385742,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7170376285058389,
+      "grad_norm": 0.042594823986291885,
+      "learning_rate": 4.055843980613561e-06,
+      "loss": 0.6916409015655518,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7186345942708853,
+      "grad_norm": 0.042385537177324295,
+      "learning_rate": 4.013552716623185e-06,
+      "loss": 0.6435957431793213,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7186345942708853,
+      "eval_loss": 0.6736165881156921,
+      "eval_runtime": 1722.8539,
+      "eval_samples_per_second": 3.231,
+      "eval_steps_per_second": 1.615,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7202315600359317,
+      "grad_norm": 0.04700292646884918,
+      "learning_rate": 3.971427672893672e-06,
+      "loss": 0.688172435760498,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7218285258009781,
+      "grad_norm": 0.040195874869823456,
+      "learning_rate": 3.929470019072972e-06,
+      "loss": 0.6703804016113282,
+      "step": 4520
+    },
+    {
+      "epoch": 0.7234254915660245,
+      "grad_norm": 0.036519281566143036,
+      "learning_rate": 3.8876809201612695e-06,
+      "loss": 0.6743984222412109,
+      "step": 4530
+    },
+    {
+      "epoch": 0.725022457331071,
+      "grad_norm": 0.03687750920653343,
+      "learning_rate": 3.846061536478626e-06,
+      "loss": 0.6919309139251709,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7266194230961174,
+      "grad_norm": 0.034661829471588135,
+      "learning_rate": 3.804613023632788e-06,
+      "loss": 0.6852948665618896,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7282163888611638,
+      "grad_norm": 0.04090991988778114,
+      "learning_rate": 3.763336532487076e-06,
+      "loss": 0.6639271259307862,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7298133546262102,
+      "grad_norm": 0.053740449249744415,
+      "learning_rate": 3.7222332091284384e-06,
+      "loss": 0.7086254119873047,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7314103203912566,
+      "grad_norm": 0.05317610129714012,
+      "learning_rate": 3.6813041948356408e-06,
+      "loss": 0.688640546798706,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7330072861563031,
+      "grad_norm": 0.04328848421573639,
+      "learning_rate": 3.640550626047562e-06,
+      "loss": 0.6996944904327392,
+      "step": 4590
+    },
+    {
+      "epoch": 0.7346042519213495,
+      "grad_norm": 0.03426145389676094,
+      "learning_rate": 3.5999736343316406e-06,
+      "loss": 0.7036723613739013,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7362012176863958,
+      "grad_norm": 0.04445657134056091,
+      "learning_rate": 3.559574346352459e-06,
+      "loss": 0.6727498054504395,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7377981834514422,
+      "grad_norm": 0.034136440604925156,
+      "learning_rate": 3.5193538838404716e-06,
+      "loss": 0.6709301471710205,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7393951492164886,
+      "grad_norm": 0.03670594096183777,
+      "learning_rate": 3.4793133635608334e-06,
+      "loss": 0.7098387241363525,
+      "step": 4630
+    },
+    {
+      "epoch": 0.740992114981535,
+      "grad_norm": 0.04140735790133476,
+      "learning_rate": 3.4394538972824167e-06,
+      "loss": 0.6899067401885987,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7425890807465815,
+      "grad_norm": 0.03637846186757088,
+      "learning_rate": 3.3997765917469348e-06,
+      "loss": 0.6745570659637451,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7441860465116279,
+      "grad_norm": 0.03964189440011978,
+      "learning_rate": 3.3602825486381886e-06,
+      "loss": 0.6715593338012695,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7457830122766743,
+      "grad_norm": 0.036266524344682693,
+      "learning_rate": 3.3209728645515204e-06,
+      "loss": 0.6992376804351806,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7473799780417207,
+      "grad_norm": 0.050657473504543304,
+      "learning_rate": 3.281848630963327e-06,
+      "loss": 0.689194917678833,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7489769438067672,
+      "grad_norm": 0.03800279274582863,
+      "learning_rate": 3.2429109342007747e-06,
+      "loss": 0.7068089962005615,
+      "step": 4690
+    },
+    {
+      "epoch": 0.7505739095718136,
+      "grad_norm": 0.034397054463624954,
+      "learning_rate": 3.204160855411638e-06,
+      "loss": 0.7055879592895508,
+      "step": 4700
+    },
+    {
+      "epoch": 0.75217087533686,
+      "grad_norm": 0.03663492202758789,
+      "learning_rate": 3.1655994705342595e-06,
+      "loss": 0.6516903400421142,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7537678411019064,
+      "grad_norm": 0.10431419312953949,
+      "learning_rate": 3.1272278502676932e-06,
+      "loss": 0.6833277702331543,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7553648068669528,
+      "grad_norm": 0.04803668335080147,
+      "learning_rate": 3.0890470600419785e-06,
+      "loss": 0.7252533435821533,
+      "step": 4730
+    },
+    {
+      "epoch": 0.7569617726319992,
+      "grad_norm": 0.052451301366090775,
+      "learning_rate": 3.051058159988539e-06,
+      "loss": 0.6383802413940429,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7585587383970456,
+      "grad_norm": 0.04214588552713394,
+      "learning_rate": 3.0132622049107672e-06,
+      "loss": 0.6385763645172119,
+      "step": 4750
+    },
+    {
+      "epoch": 0.760155704162092,
+      "grad_norm": 0.04353920742869377,
+      "learning_rate": 2.975660244254721e-06,
+      "loss": 0.6771360874176026,
+      "step": 4760
+    },
+    {
+      "epoch": 0.7617526699271384,
+      "grad_norm": 0.034283965826034546,
+      "learning_rate": 2.9382533220799893e-06,
+      "loss": 0.6694541931152344,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7633496356921848,
+      "grad_norm": 0.04156842082738876,
+      "learning_rate": 2.901042477030709e-06,
+      "loss": 0.719163179397583,
+      "step": 4780
+    },
+    {
+      "epoch": 0.7649466014572313,
+      "grad_norm": 0.038868360221385956,
+      "learning_rate": 2.864028742306725e-06,
+      "loss": 0.6658177852630616,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7665435672222777,
+      "grad_norm": 0.03701276332139969,
+      "learning_rate": 2.827213145634887e-06,
+      "loss": 0.6825634956359863,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7681405329873241,
+      "grad_norm": 0.043009012937545776,
+      "learning_rate": 2.7905967092405305e-06,
+      "loss": 0.6813971519470214,
+      "step": 4810
+    },
+    {
+      "epoch": 0.7697374987523705,
+      "grad_norm": 0.03427146002650261,
+      "learning_rate": 2.754180449819092e-06,
+      "loss": 0.6715961933135987,
+      "step": 4820
+    },
+    {
+      "epoch": 0.7713344645174169,
+      "grad_norm": 0.04296226054430008,
+      "learning_rate": 2.717965378507864e-06,
+      "loss": 0.6857268810272217,
+      "step": 4830
+    },
+    {
+      "epoch": 0.7729314302824634,
+      "grad_norm": 0.0429498665034771,
+      "learning_rate": 2.681952500857944e-06,
+      "loss": 0.6708025932312012,
+      "step": 4840
+    },
+    {
+      "epoch": 0.7745283960475098,
+      "grad_norm": 0.03906711935997009,
+      "learning_rate": 2.6461428168062973e-06,
+      "loss": 0.6925381660461426,
+      "step": 4850
+    },
+    {
+      "epoch": 0.7761253618125561,
+      "grad_norm": 0.04651563614606857,
+      "learning_rate": 2.6105373206479913e-06,
+      "loss": 0.6757769107818603,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7777223275776025,
+      "grad_norm": 0.04317731410264969,
+      "learning_rate": 2.575137001008592e-06,
+      "loss": 0.665839433670044,
+      "step": 4870
+    },
+    {
+      "epoch": 0.7793192933426489,
+      "grad_norm": 0.03744199126958847,
+      "learning_rate": 2.539942840816727e-06,
+      "loss": 0.7045305252075196,
+      "step": 4880
+    },
+    {
+      "epoch": 0.7809162591076954,
+      "grad_norm": 0.04245394840836525,
+      "learning_rate": 2.5049558172767675e-06,
+      "loss": 0.6957573890686035,
+      "step": 4890
+    },
+    {
+      "epoch": 0.7825132248727418,
+      "grad_norm": 0.03942755609750748,
+      "learning_rate": 2.47017690184172e-06,
+      "loss": 0.6565013408660889,
+      "step": 4900
+    },
+    {
+      "epoch": 0.7841101906377882,
+      "grad_norm": 0.04654339328408241,
+      "learning_rate": 2.4356070601862327e-06,
+      "loss": 0.6680932998657226,
+      "step": 4910
+    },
+    {
+      "epoch": 0.7857071564028346,
+      "grad_norm": 0.05248269438743591,
+      "learning_rate": 2.4012472521797923e-06,
+      "loss": 0.6607920169830322,
+      "step": 4920
+    },
+    {
+      "epoch": 0.787304122167881,
+      "grad_norm": 0.04624265059828758,
+      "learning_rate": 2.3670984318600786e-06,
+      "loss": 0.6683568954467773,
+      "step": 4930
+    },
+    {
+      "epoch": 0.7889010879329275,
+      "grad_norm": 0.04104055091738701,
+      "learning_rate": 2.333161547406464e-06,
+      "loss": 0.6837501049041748,
+      "step": 4940
+    },
+    {
+      "epoch": 0.7904980536979739,
+      "grad_norm": 0.039029769599437714,
+      "learning_rate": 2.2994375411136884e-06,
+      "loss": 0.6702389240264892,
+      "step": 4950
+    },
+    {
+      "epoch": 0.7920950194630203,
+      "grad_norm": 0.03849470987915993,
+      "learning_rate": 2.265927349365691e-06,
+      "loss": 0.6521383285522461,
+      "step": 4960
+    },
+    {
+      "epoch": 0.7936919852280667,
+      "grad_norm": 0.05400224030017853,
+      "learning_rate": 2.232631902609629e-06,
+      "loss": 0.6533032894134522,
+      "step": 4970
+    },
+    {
+      "epoch": 0.795288950993113,
+      "grad_norm": 0.043887361884117126,
+      "learning_rate": 2.1995521253300167e-06,
+      "loss": 0.6794767379760742,
+      "step": 4980
+    },
+    {
+      "epoch": 0.7968859167581595,
+      "grad_norm": 0.04501941800117493,
+      "learning_rate": 2.166688936023078e-06,
+      "loss": 0.6698923587799073,
+      "step": 4990
+    },
+    {
+      "epoch": 0.7984828825232059,
+      "grad_norm": 0.045139558613300323,
+      "learning_rate": 2.1340432471712367e-06,
+      "loss": 0.661639404296875,
+      "step": 5000
+    },
+    {
+      "epoch": 0.7984828825232059,
+      "eval_loss": 0.6726189255714417,
+      "eval_runtime": 1709.319,
+      "eval_samples_per_second": 3.256,
+      "eval_steps_per_second": 1.628,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8000798482882523,
+      "grad_norm": 0.04251272603869438,
+      "learning_rate": 2.101615965217768e-06,
+      "loss": 0.7006937503814697,
+      "step": 5010
+    },
+    {
+      "epoch": 0.8016768140532987,
+      "grad_norm": 0.040616679936647415,
+      "learning_rate": 2.0694079905416475e-06,
+      "loss": 0.6587623119354248,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8032737798183451,
+      "grad_norm": 0.04379381984472275,
+      "learning_rate": 2.0374202174325443e-06,
+      "loss": 0.6809040069580078,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8048707455833916,
+      "grad_norm": 0.042080674320459366,
+      "learning_rate": 2.0056535340659867e-06,
+      "loss": 0.6683236598968506,
+      "step": 5040
+    },
+    {
+      "epoch": 0.806467711348438,
+      "grad_norm": 0.038132891058921814,
+      "learning_rate": 1.97410882247871e-06,
+      "loss": 0.6637454032897949,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8080646771134844,
+      "grad_norm": 0.029925232753157616,
+      "learning_rate": 1.942786958544154e-06,
+      "loss": 0.6917686939239502,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8096616428785308,
+      "grad_norm": 0.041264407336711884,
+      "learning_rate": 1.91168881194815e-06,
+      "loss": 0.7052114486694336,
+      "step": 5070
+    },
+    {
+      "epoch": 0.8112586086435772,
+      "grad_norm": 0.03905145451426506,
+      "learning_rate": 1.8808152461647766e-06,
+      "loss": 0.667028284072876,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8128555744086237,
+      "grad_norm": 0.047729961574077606,
+      "learning_rate": 1.8501671184323811e-06,
+      "loss": 0.691020917892456,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8144525401736701,
+      "grad_norm": 0.034812819212675095,
+      "learning_rate": 1.8197452797297699e-06,
+      "loss": 0.7124451637268067,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8160495059387164,
+      "grad_norm": 0.044796060770750046,
+      "learning_rate": 1.7895505747525933e-06,
+      "loss": 0.7617973804473877,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8176464717037628,
+      "grad_norm": 0.039166904985904694,
+      "learning_rate": 1.759583841889878e-06,
+      "loss": 0.711750841140747,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8192434374688092,
+      "grad_norm": 0.03871138393878937,
+      "learning_rate": 1.7298459132007628e-06,
+      "loss": 0.6736746788024902,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8208404032338557,
+      "grad_norm": 0.046535786241292953,
+      "learning_rate": 1.7003376143913886e-06,
+      "loss": 0.7094714164733886,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8224373689989021,
+      "grad_norm": 0.04319629445672035,
+      "learning_rate": 1.6710597647919613e-06,
+      "loss": 0.6693609237670899,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8240343347639485,
+      "grad_norm": 0.04204268380999565,
+      "learning_rate": 1.6420131773340263e-06,
+      "loss": 0.6906074047088623,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8256313005289949,
+      "grad_norm": 0.041262589395046234,
+      "learning_rate": 1.6131986585278714e-06,
+      "loss": 0.6685310363769531,
+      "step": 5170
+    },
+    {
+      "epoch": 0.8272282662940413,
+      "grad_norm": 0.03515459597110748,
+      "learning_rate": 1.5846170084401458e-06,
+      "loss": 0.6744651317596435,
+      "step": 5180
+    },
+    {
+      "epoch": 0.8288252320590878,
+      "grad_norm": 0.044764138758182526,
+      "learning_rate": 1.5562690206716479e-06,
+      "loss": 0.6720765113830567,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8304221978241342,
+      "grad_norm": 0.03811672702431679,
+      "learning_rate": 1.5281554823352896e-06,
+      "loss": 0.6512242317199707,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8320191635891806,
+      "grad_norm": 0.03850555047392845,
+      "learning_rate": 1.500277174034227e-06,
+      "loss": 0.6925489425659179,
+      "step": 5210
+    },
+    {
+      "epoch": 0.833616129354227,
+      "grad_norm": 0.039821285754442215,
+      "learning_rate": 1.4726348698402015e-06,
+      "loss": 0.6698274612426758,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8352130951192733,
+      "grad_norm": 0.042530301958322525,
+      "learning_rate": 1.4452293372720483e-06,
+      "loss": 0.657553243637085,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8368100608843198,
+      "grad_norm": 0.03285897895693779,
+      "learning_rate": 1.4180613372743678e-06,
+      "loss": 0.6513785362243653,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8384070266493662,
+      "grad_norm": 0.0418255515396595,
+      "learning_rate": 1.39113162419642e-06,
+      "loss": 0.6803596019744873,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8400039924144126,
+      "grad_norm": 0.03861439600586891,
+      "learning_rate": 1.3644409457711604e-06,
+      "loss": 0.7275567531585694,
+      "step": 5260
+    },
+    {
+      "epoch": 0.841600958179459,
+      "grad_norm": 0.035479992628097534,
+      "learning_rate": 1.3379900430944858e-06,
+      "loss": 0.7078885555267334,
+      "step": 5270
+    },
+    {
+      "epoch": 0.8431979239445054,
+      "grad_norm": 0.036275558173656464,
+      "learning_rate": 1.31177965060466e-06,
+      "loss": 0.6733812808990478,
+      "step": 5280
+    },
+    {
+      "epoch": 0.8447948897095519,
+      "grad_norm": 0.037248894572257996,
+      "learning_rate": 1.285810496061921e-06,
+      "loss": 0.6728686809539794,
+      "step": 5290
+    },
+    {
+      "epoch": 0.8463918554745983,
+      "grad_norm": 0.03526445850729942,
+      "learning_rate": 1.2600833005282631e-06,
+      "loss": 0.6712977409362793,
+      "step": 5300
+    },
+    {
+      "epoch": 0.8479888212396447,
+      "grad_norm": 0.03622904419898987,
+      "learning_rate": 1.2345987783474311e-06,
+      "loss": 0.6881800651550293,
+      "step": 5310
+    },
+    {
+      "epoch": 0.8495857870046911,
+      "grad_norm": 0.05652482062578201,
+      "learning_rate": 1.2093576371250737e-06,
+      "loss": 0.709586763381958,
+      "step": 5320
+    },
+    {
+      "epoch": 0.8511827527697375,
+      "grad_norm": 0.05090609937906265,
+      "learning_rate": 1.1843605777091028e-06,
+      "loss": 0.7400435447692871,
+      "step": 5330
+    },
+    {
+      "epoch": 0.852779718534784,
+      "grad_norm": 0.038989149034023285,
+      "learning_rate": 1.1596082941702313e-06,
+      "loss": 0.695259952545166,
+      "step": 5340
+    },
+    {
+      "epoch": 0.8543766842998303,
+      "grad_norm": 0.04729596897959709,
+      "learning_rate": 1.135101473782706e-06,
+      "loss": 0.681358528137207,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8559736500648767,
+      "grad_norm": 0.0379403792321682,
+      "learning_rate": 1.1108407970052149e-06,
+      "loss": 0.6542623043060303,
+      "step": 5360
+    },
+    {
+      "epoch": 0.8575706158299231,
+      "grad_norm": 0.044038884341716766,
+      "learning_rate": 1.086826937462e-06,
+      "loss": 0.6564295291900635,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8591675815949695,
+      "grad_norm": 0.04039445519447327,
+      "learning_rate": 1.0630605619241564e-06,
+      "loss": 0.6783324241638183,
+      "step": 5380
+    },
+    {
+      "epoch": 0.860764547360016,
+      "grad_norm": 0.034845978021621704,
+      "learning_rate": 1.0395423302911112e-06,
+      "loss": 0.6549112796783447,
+      "step": 5390
+    },
+    {
+      "epoch": 0.8623615131250624,
+      "grad_norm": 0.048201173543930054,
+      "learning_rate": 1.016272895572309e-06,
+      "loss": 0.6641885757446289,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8639584788901088,
+      "grad_norm": 0.037778809666633606,
+      "learning_rate": 9.932529038690696e-07,
+      "loss": 0.7034367561340332,
+      "step": 5410
+    },
+    {
+      "epoch": 0.8655554446551552,
+      "grad_norm": 0.0427137166261673,
+      "learning_rate": 9.704829943566574e-07,
+      "loss": 0.6777073383331299,
+      "step": 5420
+    },
+    {
+      "epoch": 0.8671524104202016,
+      "grad_norm": 0.037193115800619125,
+      "learning_rate": 9.479637992665313e-07,
+      "loss": 0.6748072147369385,
+      "step": 5430
+    },
+    {
+      "epoch": 0.8687493761852481,
+      "grad_norm": 0.0337526798248291,
+      "learning_rate": 9.256959438687929e-07,
+      "loss": 0.6617064952850342,
+      "step": 5440
+    },
+    {
+      "epoch": 0.8703463419502945,
+      "grad_norm": 0.039619434624910355,
+      "learning_rate": 9.036800464548157e-07,
+      "loss": 0.6612413883209228,
+      "step": 5450
+    },
+    {
+      "epoch": 0.8719433077153409,
+      "grad_norm": 0.03671206533908844,
+      "learning_rate": 8.819167183200905e-07,
+      "loss": 0.6929868698120117,
+      "step": 5460
+    },
+    {
+      "epoch": 0.8735402734803873,
+      "grad_norm": 0.034324243664741516,
+      "learning_rate": 8.604065637472392e-07,
+      "loss": 0.685936689376831,
+      "step": 5470
+    },
+    {
+      "epoch": 0.8751372392454336,
+      "grad_norm": 0.035913076251745224,
+      "learning_rate": 8.391501799892466e-07,
+      "loss": 0.7082894802093506,
+      "step": 5480
+    },
+    {
+      "epoch": 0.87673420501048,
+      "grad_norm": 0.04239356890320778,
+      "learning_rate": 8.181481572528726e-07,
+      "loss": 0.7099148750305175,
+      "step": 5490
+    },
+    {
+      "epoch": 0.8783311707755265,
+      "grad_norm": 0.04144662618637085,
+      "learning_rate": 7.974010786822684e-07,
+      "loss": 0.6640007972717286,
+      "step": 5500
+    },
+    {
+      "epoch": 0.8783311707755265,
+      "eval_loss": 0.672135055065155,
+      "eval_runtime": 1721.4336,
+      "eval_samples_per_second": 3.233,
+      "eval_steps_per_second": 1.617,
+      "step": 5500
+    },
+    {
+      "epoch": 0.8799281365405729,
+      "grad_norm": 0.030981754884123802,
+      "learning_rate": 7.769095203427734e-07,
+      "loss": 0.6549492359161377,
+      "step": 5510
+    },
+    {
+      "epoch": 0.8815251023056193,
+      "grad_norm": 0.04086959362030029,
+      "learning_rate": 7.56674051204932e-07,
+      "loss": 0.6875146389007568,
+      "step": 5520
+    },
+    {
+      "epoch": 0.8831220680706657,
+      "grad_norm": 0.0377834253013134,
+      "learning_rate": 7.366952331286958e-07,
+      "loss": 0.7493057250976562,
+      "step": 5530
+    },
+    {
+      "epoch": 0.8847190338357122,
+      "grad_norm": 0.03834075480699539,
+      "learning_rate": 7.169736208478118e-07,
+      "loss": 0.6828866481781006,
+      "step": 5540
+    },
+    {
+      "epoch": 0.8863159996007586,
+      "grad_norm": 0.03951543942093849,
+      "learning_rate": 6.975097619544324e-07,
+      "loss": 0.68464674949646,
+      "step": 5550
+    },
+    {
+      "epoch": 0.887912965365805,
+      "grad_norm": 0.044246163219213486,
+      "learning_rate": 6.78304196883901e-07,
+      "loss": 0.6737260341644287,
+      "step": 5560
+    },
+    {
+      "epoch": 0.8895099311308514,
+      "grad_norm": 0.04266819730401039,
+      "learning_rate": 6.59357458899752e-07,
+      "loss": 0.7164577960968017,
+      "step": 5570
+    },
+    {
+      "epoch": 0.8911068968958978,
+      "grad_norm": 0.0311746709048748,
+      "learning_rate": 6.40670074078904e-07,
+      "loss": 0.680039358139038,
+      "step": 5580
+    },
+    {
+      "epoch": 0.8927038626609443,
+      "grad_norm": 0.035248082131147385,
+      "learning_rate": 6.222425612970462e-07,
+      "loss": 0.6851954936981202,
+      "step": 5590
+    },
+    {
+      "epoch": 0.8943008284259906,
+      "grad_norm": 0.03986518457531929,
+      "learning_rate": 6.040754322142417e-07,
+      "loss": 0.6760111331939698,
+      "step": 5600
+    },
+    {
+      "epoch": 0.895897794191037,
+      "grad_norm": 0.03682507202029228,
+      "learning_rate": 5.861691912607159e-07,
+      "loss": 0.6883102416992187,
+      "step": 5610
+    },
+    {
+      "epoch": 0.8974947599560834,
+      "grad_norm": 0.03000503219664097,
+      "learning_rate": 5.685243356228409e-07,
+      "loss": 0.6542916774749756,
+      "step": 5620
+    },
+    {
+      "epoch": 0.8990917257211298,
+      "grad_norm": 0.044614020735025406,
+      "learning_rate": 5.511413552293476e-07,
+      "loss": 0.6680744647979736,
+      "step": 5630
+    },
+    {
+      "epoch": 0.9006886914861763,
+      "grad_norm": 0.03846392408013344,
+      "learning_rate": 5.340207327377156e-07,
+      "loss": 0.7028224468231201,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9022856572512227,
+      "grad_norm": 0.038470469415187836,
+      "learning_rate": 5.171629435207615e-07,
+      "loss": 0.6939930438995361,
+      "step": 5650
+    },
+    {
+      "epoch": 0.9038826230162691,
+      "grad_norm": 0.03757188841700554,
+      "learning_rate": 5.005684556534574e-07,
+      "loss": 0.656277084350586,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9054795887813155,
+      "grad_norm": 0.04011685401201248,
+      "learning_rate": 4.842377298999191e-07,
+      "loss": 0.638321828842163,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9070765545463619,
+      "grad_norm": 0.039341870695352554,
+      "learning_rate": 4.681712197006205e-07,
+      "loss": 0.6716189384460449,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9086735203114084,
+      "grad_norm": 0.046330180019140244,
+      "learning_rate": 4.5236937115980204e-07,
+      "loss": 0.6909313678741456,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9102704860764548,
+      "grad_norm": 0.041009679436683655,
+      "learning_rate": 4.368326230330855e-07,
+      "loss": 0.6272893428802491,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9118674518415012,
+      "grad_norm": 0.03455239534378052,
+      "learning_rate": 4.215614067152851e-07,
+      "loss": 0.6927034854888916,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9134644176065475,
+      "grad_norm": 0.03742032125592232,
+      "learning_rate": 4.065561462284351e-07,
+      "loss": 0.664287805557251,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9150613833715939,
+      "grad_norm": 0.03536292910575867,
+      "learning_rate": 3.918172582100177e-07,
+      "loss": 0.664261770248413,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9166583491366403,
+      "grad_norm": 0.044113483279943466,
+      "learning_rate": 3.7734515190138934e-07,
+      "loss": 0.7117574691772461,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9182553149016868,
+      "grad_norm": 0.04183537885546684,
+      "learning_rate": 3.631402291364206e-07,
+      "loss": 0.6857941627502442,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9198522806667332,
+      "grad_norm": 0.02898181416094303,
+      "learning_rate": 3.492028843303419e-07,
+      "loss": 0.6839983463287354,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9214492464317796,
+      "grad_norm": 0.040342144668102264,
+      "learning_rate": 3.3553350446878244e-07,
+      "loss": 0.6938966751098633,
+      "step": 5770
+    },
+    {
+      "epoch": 0.923046212196826,
+      "grad_norm": 0.05028856173157692,
+      "learning_rate": 3.221324690970351e-07,
+      "loss": 0.6682254791259765,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9246431779618725,
+      "grad_norm": 0.033745042979717255,
+      "learning_rate": 3.090001503095175e-07,
+      "loss": 0.6571940422058106,
+      "step": 5790
+    },
+    {
+      "epoch": 0.9262401437269189,
+      "grad_norm": 0.036207910627126694,
+      "learning_rate": 2.9613691273942977e-07,
+      "loss": 0.7331238269805909,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9278371094919653,
+      "grad_norm": 0.04358496144413948,
+      "learning_rate": 2.835431135486455e-07,
+      "loss": 0.6768048763275146,
+      "step": 5810
+    },
+    {
+      "epoch": 0.9294340752570117,
+      "grad_norm": 0.051529187709093094,
+      "learning_rate": 2.7121910241778014e-07,
+      "loss": 0.6914044857025147,
+      "step": 5820
+    },
+    {
+      "epoch": 0.9310310410220581,
+      "grad_norm": 0.03387345373630524,
+      "learning_rate": 2.591652215364904e-07,
+      "loss": 0.67225022315979,
+      "step": 5830
+    },
+    {
+      "epoch": 0.9326280067871044,
+      "grad_norm": 0.043374814093112946,
+      "learning_rate": 2.473818055939747e-07,
+      "loss": 0.6870133876800537,
+      "step": 5840
+    },
+    {
+      "epoch": 0.9342249725521509,
+      "grad_norm": 0.04162263497710228,
+      "learning_rate": 2.358691817696712e-07,
+      "loss": 0.6274053573608398,
+      "step": 5850
+    },
+    {
+      "epoch": 0.9358219383171973,
+      "grad_norm": 0.03655523061752319,
+      "learning_rate": 2.2462766972418226e-07,
+      "loss": 0.6559162139892578,
+      "step": 5860
+    },
+    {
+      "epoch": 0.9374189040822437,
+      "grad_norm": 0.036134812980890274,
+      "learning_rate": 2.1365758159039073e-07,
+      "loss": 0.6783801078796386,
+      "step": 5870
+    },
+    {
+      "epoch": 0.9390158698472901,
+      "grad_norm": 0.03370088338851929,
+      "learning_rate": 2.0295922196480355e-07,
+      "loss": 0.679423189163208,
+      "step": 5880
+    },
+    {
+      "epoch": 0.9406128356123366,
+      "grad_norm": 0.044190868735313416,
+      "learning_rate": 1.9253288789908287e-07,
+      "loss": 0.6669793128967285,
+      "step": 5890
+    },
+    {
+      "epoch": 0.942209801377383,
+      "grad_norm": 0.03765581548213959,
+      "learning_rate": 1.823788688918049e-07,
+      "loss": 0.6280838966369628,
+      "step": 5900
+    },
+    {
+      "epoch": 0.9438067671424294,
+      "grad_norm": 0.05461205169558525,
+      "learning_rate": 1.7249744688042414e-07,
+      "loss": 0.6410412788391113,
+      "step": 5910
+    },
+    {
+      "epoch": 0.9454037329074758,
+      "grad_norm": 0.046921294182538986,
+      "learning_rate": 1.6288889623343295e-07,
+      "loss": 0.6527444839477539,
+      "step": 5920
+    },
+    {
+      "epoch": 0.9470006986725222,
+      "grad_norm": 0.03945396840572357,
+      "learning_rate": 1.535534837427588e-07,
+      "loss": 0.694933271408081,
+      "step": 5930
+    },
+    {
+      "epoch": 0.9485976644375687,
+      "grad_norm": 0.03962003439664841,
+      "learning_rate": 1.44491468616349e-07,
+      "loss": 0.6714414596557617,
+      "step": 5940
+    },
+    {
+      "epoch": 0.9501946302026151,
+      "grad_norm": 0.07099556922912598,
+      "learning_rate": 1.3570310247096764e-07,
+      "loss": 0.7129894733428955,
+      "step": 5950
+    },
+    {
+      "epoch": 0.9517915959676615,
+      "grad_norm": 0.03700670972466469,
+      "learning_rate": 1.2718862932522224e-07,
+      "loss": 0.6442203044891357,
+      "step": 5960
+    },
+    {
+      "epoch": 0.9533885617327078,
+      "grad_norm": 0.04066568240523338,
+      "learning_rate": 1.1894828559277816e-07,
+      "loss": 0.6612192153930664,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9549855274977542,
+      "grad_norm": 0.04404400289058685,
+      "learning_rate": 1.1098230007579702e-07,
+      "loss": 0.6766380310058594,
+      "step": 5980
+    },
+    {
+      "epoch": 0.9565824932628006,
+      "grad_norm": 0.04611429572105408,
+      "learning_rate": 1.0329089395858527e-07,
+      "loss": 0.6801462650299073,
+      "step": 5990
+    },
+    {
+      "epoch": 0.9581794590278471,
+      "grad_norm": 0.03326278179883957,
+      "learning_rate": 9.587428080145344e-08,
+      "loss": 0.6539367198944092,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9581794590278471,
+      "eval_loss": 0.6719915270805359,
+      "eval_runtime": 1743.0303,
+      "eval_samples_per_second": 3.193,
+      "eval_steps_per_second": 1.597,
+      "step": 6000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 6262,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.3174132430797578e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}