{
  "best_global_step": 3600,
  "best_metric": 0.36828258633613586,
  "best_model_checkpoint": "saves/Gemma-2-9B/lora_seqcls/train_seqcls/checkpoint-1000",
  "epoch": 2.9603014399138745,
  "eval_steps": 100,
  "global_step": 5500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0053828556048983985,
      "grad_norm": 198.63441467285156,
      "learning_rate": 6.474820143884892e-06,
      "loss": 19.1404296875,
      "step": 10
    },
    {
      "epoch": 0.010765711209796797,
      "grad_norm": 130.11439514160156,
      "learning_rate": 1.366906474820144e-05,
      "loss": 16.12879333496094,
      "step": 20
    },
    {
      "epoch": 0.016148566814695196,
      "grad_norm": 94.73497009277344,
      "learning_rate": 2.0863309352517988e-05,
      "loss": 11.966738891601562,
      "step": 30
    },
    {
      "epoch": 0.021531422419593594,
      "grad_norm": 55.8490104675293,
      "learning_rate": 2.805755395683453e-05,
      "loss": 9.874456787109375,
      "step": 40
    },
    {
      "epoch": 0.02691427802449199,
      "grad_norm": 53.740142822265625,
      "learning_rate": 3.5251798561151075e-05,
      "loss": 7.273980712890625,
      "step": 50
    },
    {
      "epoch": 0.03229713362939039,
      "grad_norm": 83.30630493164062,
      "learning_rate": 4.244604316546763e-05,
      "loss": 6.787509155273438,
      "step": 60
    },
    {
      "epoch": 0.03767998923428879,
      "grad_norm": 96.4136734008789,
      "learning_rate": 4.964028776978418e-05,
      "loss": 6.7754661560058596,
      "step": 70
    },
    {
      "epoch": 0.04306284483918719,
      "grad_norm": 77.66259765625,
      "learning_rate": 5.683453237410072e-05,
      "loss": 5.704140853881836,
      "step": 80
    },
    {
      "epoch": 0.04844570044408559,
      "grad_norm": 46.28746032714844,
      "learning_rate": 6.402877697841726e-05,
      "loss": 4.287641906738282,
      "step": 90
    },
    {
      "epoch": 0.05382855604898398,
      "grad_norm": 121.80719757080078,
      "learning_rate": 7.122302158273382e-05,
      "loss": 4.469888305664062,
      "step": 100
    },
    {
      "epoch": 0.05382855604898398,
      "eval_loss": 1.1105364561080933,
      "eval_runtime": 64.1061,
      "eval_samples_per_second": 51.524,
      "eval_steps_per_second": 12.885,
      "step": 100
    },
    {
      "epoch": 0.059211411653882384,
      "grad_norm": 82.3691635131836,
      "learning_rate": 7.841726618705037e-05,
      "loss": 4.892630767822266,
      "step": 110
    },
    {
      "epoch": 0.06459426725878079,
      "grad_norm": 53.88746643066406,
      "learning_rate": 8.561151079136692e-05,
      "loss": 3.885559844970703,
      "step": 120
    },
    {
      "epoch": 0.06997712286367919,
      "grad_norm": 59.405372619628906,
      "learning_rate": 9.280575539568345e-05,
      "loss": 3.8151809692382814,
      "step": 130
    },
    {
      "epoch": 0.07535997846857757,
      "grad_norm": 85.99320220947266,
      "learning_rate": 0.0001,
      "loss": 4.0115612030029295,
      "step": 140
    },
    {
      "epoch": 0.08074283407347597,
      "grad_norm": 40.51960754394531,
      "learning_rate": 0.00010719424460431656,
      "loss": 3.4070945739746095,
      "step": 150
    },
    {
      "epoch": 0.08612568967837438,
      "grad_norm": 73.72062683105469,
      "learning_rate": 0.00011438848920863309,
      "loss": 4.7824043273925785,
      "step": 160
    },
    {
      "epoch": 0.09150854528327278,
      "grad_norm": 62.74378204345703,
      "learning_rate": 0.00012158273381294964,
      "loss": 4.632376480102539,
      "step": 170
    },
    {
      "epoch": 0.09689140088817118,
      "grad_norm": 41.6406364440918,
      "learning_rate": 0.0001287769784172662,
      "loss": 3.02130126953125,
      "step": 180
    },
    {
      "epoch": 0.10227425649306958,
      "grad_norm": 55.67851257324219,
      "learning_rate": 0.00013597122302158273,
      "loss": 4.105777359008789,
      "step": 190
    },
    {
      "epoch": 0.10765711209796797,
      "grad_norm": 88.68601989746094,
      "learning_rate": 0.0001431654676258993,
      "loss": 3.745909881591797,
      "step": 200
    },
    {
      "epoch": 0.10765711209796797,
      "eval_loss": 1.129400610923767,
      "eval_runtime": 64.0494,
      "eval_samples_per_second": 51.57,
      "eval_steps_per_second": 12.896,
      "step": 200
    },
    {
      "epoch": 0.11303996770286637,
      "grad_norm": 44.12594223022461,
      "learning_rate": 0.00015035971223021583,
      "loss": 4.0704700469970705,
      "step": 210
    },
    {
      "epoch": 0.11842282330776477,
      "grad_norm": 41.71208572387695,
      "learning_rate": 0.00015755395683453237,
      "loss": 3.3986480712890623,
      "step": 220
    },
    {
      "epoch": 0.12380567891266317,
      "grad_norm": 51.5625,
      "learning_rate": 0.00016474820143884893,
      "loss": 3.7478317260742187,
      "step": 230
    },
    {
      "epoch": 0.12918853451756157,
      "grad_norm": 54.59484100341797,
      "learning_rate": 0.00017194244604316547,
      "loss": 4.002623748779297,
      "step": 240
    },
    {
      "epoch": 0.13457139012245997,
      "grad_norm": 34.029727935791016,
      "learning_rate": 0.000179136690647482,
      "loss": 3.538894271850586,
      "step": 250
    },
    {
      "epoch": 0.13995424572735837,
      "grad_norm": 19.37107276916504,
      "learning_rate": 0.00018633093525179857,
      "loss": 3.356983184814453,
      "step": 260
    },
    {
      "epoch": 0.14533710133225677,
      "grad_norm": 58.17410659790039,
      "learning_rate": 0.0001935251798561151,
      "loss": 3.4610286712646485,
      "step": 270
    },
    {
      "epoch": 0.15071995693715515,
      "grad_norm": 115.1293716430664,
      "learning_rate": 0.00019999998240562744,
      "loss": 4.566124343872071,
      "step": 280
    },
    {
      "epoch": 0.15610281254205355,
      "grad_norm": 52.380348205566406,
      "learning_rate": 0.0001999978710884105,
      "loss": 4.061540222167968,
      "step": 290
    },
    {
      "epoch": 0.16148566814695195,
      "grad_norm": 63.469581604003906,
      "learning_rate": 0.00019999224098180877,
      "loss": 3.930801773071289,
      "step": 300
    },
    {
      "epoch": 0.16148566814695195,
      "eval_loss": 0.9815971255302429,
      "eval_runtime": 64.185,
      "eval_samples_per_second": 51.461,
      "eval_steps_per_second": 12.869,
      "step": 300
    },
    {
      "epoch": 0.16686852375185035,
      "grad_norm": 18.164447784423828,
      "learning_rate": 0.00019998309228393805,
      "loss": 4.083843612670899,
      "step": 310
    },
    {
      "epoch": 0.17225137935674875,
      "grad_norm": 93.28253936767578,
      "learning_rate": 0.00019997042531672858,
      "loss": 3.1324670791625975,
      "step": 320
    },
    {
      "epoch": 0.17763423496164715,
      "grad_norm": 49.66929626464844,
      "learning_rate": 0.00019995424052591376,
      "loss": 4.83184928894043,
      "step": 330
    },
    {
      "epoch": 0.18301709056654555,
      "grad_norm": 43.669410705566406,
      "learning_rate": 0.00019993453848101442,
      "loss": 3.0538238525390624,
      "step": 340
    },
    {
      "epoch": 0.18839994617144395,
      "grad_norm": 28.3753719329834,
      "learning_rate": 0.00019991131987531877,
      "loss": 3.760041427612305,
      "step": 350
    },
    {
      "epoch": 0.19378280177634236,
      "grad_norm": 26.99091148376465,
      "learning_rate": 0.00019988458552585808,
      "loss": 3.3557403564453123,
      "step": 360
    },
    {
      "epoch": 0.19916565738124076,
      "grad_norm": 42.94520568847656,
      "learning_rate": 0.00019985433637337776,
      "loss": 3.5249252319335938,
      "step": 370
    },
    {
      "epoch": 0.20454851298613916,
      "grad_norm": 34.760292053222656,
      "learning_rate": 0.00019982057348230447,
      "loss": 3.889664077758789,
      "step": 380
    },
    {
      "epoch": 0.20993136859103753,
      "grad_norm": 22.45956802368164,
      "learning_rate": 0.00019978329804070858,
      "loss": 3.2252212524414063,
      "step": 390
    },
    {
      "epoch": 0.21531422419593593,
      "grad_norm": 54.65021514892578,
      "learning_rate": 0.0001997425113602622,
      "loss": 3.99298095703125,
      "step": 400
    },
    {
      "epoch": 0.21531422419593593,
      "eval_loss": 0.8698179721832275,
      "eval_runtime": 64.4139,
      "eval_samples_per_second": 51.278,
      "eval_steps_per_second": 12.823,
      "step": 400
    },
    {
      "epoch": 0.22069707980083433,
      "grad_norm": 37.18206024169922,
      "learning_rate": 0.0001996982148761933,
      "loss": 2.9811370849609373,
      "step": 410
    },
    {
      "epoch": 0.22607993540573273,
      "grad_norm": 33.8664436340332,
      "learning_rate": 0.0001996504101472351,
      "loss": 2.5169837951660154,
      "step": 420
    },
    {
      "epoch": 0.23146279101063114,
      "grad_norm": 45.07470703125,
      "learning_rate": 0.00019959909885557112,
      "loss": 3.458451843261719,
      "step": 430
    },
    {
      "epoch": 0.23684564661552954,
      "grad_norm": 25.689266204833984,
      "learning_rate": 0.00019954428280677603,
      "loss": 3.2378326416015626,
      "step": 440
    },
    {
      "epoch": 0.24222850222042794,
      "grad_norm": 53.7376594543457,
      "learning_rate": 0.0001994859639297522,
      "loss": 3.302751159667969,
      "step": 450
    },
    {
      "epoch": 0.24761135782532634,
      "grad_norm": 49.085391998291016,
      "learning_rate": 0.00019942414427666187,
      "loss": 2.930678939819336,
      "step": 460
    },
    {
      "epoch": 0.2529942134302247,
      "grad_norm": 23.53020477294922,
      "learning_rate": 0.00019935882602285462,
      "loss": 3.5229896545410155,
      "step": 470
    },
    {
      "epoch": 0.25837706903512314,
      "grad_norm": 42.54838562011719,
      "learning_rate": 0.00019929001146679127,
      "loss": 3.073961639404297,
      "step": 480
    },
    {
      "epoch": 0.2637599246400215,
      "grad_norm": 36.783966064453125,
      "learning_rate": 0.0001992177030299626,
      "loss": 2.839693069458008,
      "step": 490
    },
    {
      "epoch": 0.26914278024491994,
      "grad_norm": 29.89275360107422,
      "learning_rate": 0.0001991419032568044,
      "loss": 3.5954513549804688,
      "step": 500
    },
    {
      "epoch": 0.26914278024491994,
      "eval_loss": 0.7735958099365234,
      "eval_runtime": 64.2996,
      "eval_samples_per_second": 51.369,
      "eval_steps_per_second": 12.846,
      "step": 500
    },
    {
      "epoch": 0.2745256358498183,
      "grad_norm": 15.766196250915527,
      "learning_rate": 0.0001990626148146078,
      "loss": 2.6433101654052735,
      "step": 510
    },
    {
      "epoch": 0.27990849145471675,
      "grad_norm": 15.622676849365234,
      "learning_rate": 0.00019897984049342552,
      "loss": 2.2897308349609373,
      "step": 520
    },
    {
      "epoch": 0.2852913470596151,
      "grad_norm": 64.56971740722656,
      "learning_rate": 0.0001988935832059736,
      "loss": 4.1805778503417965,
      "step": 530
    },
    {
      "epoch": 0.29067420266451355,
      "grad_norm": 30.153961181640625,
      "learning_rate": 0.0001988038459875289,
      "loss": 2.8382762908935546,
      "step": 540
    },
    {
      "epoch": 0.2960570582694119,
      "grad_norm": 28.525348663330078,
      "learning_rate": 0.00019871063199582237,
      "loss": 3.3420616149902345,
      "step": 550
    },
    {
      "epoch": 0.3014399138743103,
      "grad_norm": 25.80811882019043,
      "learning_rate": 0.0001986139445109279,
      "loss": 2.491804504394531,
      "step": 560
    },
    {
      "epoch": 0.3068227694792087,
      "grad_norm": 42.330684661865234,
      "learning_rate": 0.00019851378693514686,
      "loss": 3.0271644592285156,
      "step": 570
    },
    {
      "epoch": 0.3122056250841071,
      "grad_norm": 40.063087463378906,
      "learning_rate": 0.00019841016279288846,
      "loss": 2.38574161529541,
      "step": 580
    },
    {
      "epoch": 0.3175884806890055,
      "grad_norm": 36.078636169433594,
      "learning_rate": 0.0001983030757305456,
      "loss": 2.8599777221679688,
      "step": 590
    },
    {
      "epoch": 0.3229713362939039,
      "grad_norm": 37.90999984741211,
      "learning_rate": 0.0001981925295163667,
      "loss": 3.3058528900146484,
      "step": 600
    },
    {
      "epoch": 0.3229713362939039,
      "eval_loss": 0.724442720413208,
      "eval_runtime": 64.0896,
      "eval_samples_per_second": 51.537,
      "eval_steps_per_second": 12.888,
      "step": 600
    },
    {
      "epoch": 0.32835419189880233,
      "grad_norm": 34.73746871948242,
      "learning_rate": 0.00019807852804032305,
      "loss": 3.4641307830810546,
      "step": 610
    },
    {
      "epoch": 0.3337370475037007,
      "grad_norm": 72.2687759399414,
      "learning_rate": 0.00019796107531397188,
      "loss": 4.683813858032226,
      "step": 620
    },
    {
      "epoch": 0.33911990310859913,
      "grad_norm": 30.93055534362793,
      "learning_rate": 0.00019784017547031522,
      "loss": 2.6965951919555664,
      "step": 630
    },
    {
      "epoch": 0.3445027587134975,
      "grad_norm": 18.676742553710938,
      "learning_rate": 0.00019771583276365453,
      "loss": 2.5014991760253906,
      "step": 640
    },
    {
      "epoch": 0.34988561431839593,
      "grad_norm": 42.73445510864258,
      "learning_rate": 0.0001975880515694409,
      "loss": 3.033338737487793,
      "step": 650
    },
    {
      "epoch": 0.3552684699232943,
      "grad_norm": 47.229732513427734,
      "learning_rate": 0.00019745683638412116,
      "loss": 3.065496635437012,
      "step": 660
    },
    {
      "epoch": 0.3606513255281927,
      "grad_norm": 36.450660705566406,
      "learning_rate": 0.00019732219182497964,
      "loss": 2.9941673278808594,
      "step": 670
    },
    {
      "epoch": 0.3660341811330911,
      "grad_norm": 45.24620056152344,
      "learning_rate": 0.00019718412262997566,
      "loss": 2.664113998413086,
      "step": 680
    },
    {
      "epoch": 0.3714170367379895,
      "grad_norm": 22.437503814697266,
      "learning_rate": 0.0001970426336575768,
      "loss": 1.6834583282470703,
      "step": 690
    },
    {
      "epoch": 0.3767998923428879,
      "grad_norm": 36.59716033935547,
      "learning_rate": 0.000196897729886588,
      "loss": 2.480428123474121,
      "step": 700
    },
    {
      "epoch": 0.3767998923428879,
      "eval_loss": 0.6947948932647705,
      "eval_runtime": 64.1166,
      "eval_samples_per_second": 51.516,
      "eval_steps_per_second": 12.883,
      "step": 700
    },
    {
      "epoch": 0.3821827479477863,
      "grad_norm": 45.49883270263672,
      "learning_rate": 0.00019674941641597638,
      "loss": 2.7860897064208983,
      "step": 710
    },
    {
      "epoch": 0.3875656035526847,
      "grad_norm": 33.21625518798828,
      "learning_rate": 0.00019659769846469164,
      "loss": 3.083520698547363,
      "step": 720
    },
    {
      "epoch": 0.3929484591575831,
      "grad_norm": 16.77117919921875,
      "learning_rate": 0.00019644258137148263,
      "loss": 2.312260055541992,
      "step": 730
    },
    {
      "epoch": 0.3983313147624815,
      "grad_norm": 51.02537536621094,
      "learning_rate": 0.00019628407059470938,
      "loss": 3.675634002685547,
      "step": 740
    },
    {
      "epoch": 0.4037141703673799,
      "grad_norm": 20.05516242980957,
      "learning_rate": 0.00019612217171215094,
      "loss": 2.1038530349731444,
      "step": 750
    },
    {
      "epoch": 0.4090970259722783,
      "grad_norm": 35.98112869262695,
      "learning_rate": 0.00019595689042080944,
      "loss": 2.4246122360229494,
      "step": 760
    },
    {
      "epoch": 0.4144798815771767,
      "grad_norm": 27.854074478149414,
      "learning_rate": 0.0001957882325367091,
      "loss": 2.453342056274414,
      "step": 770
    },
    {
      "epoch": 0.41986273718207506,
      "grad_norm": 40.96638870239258,
      "learning_rate": 0.000195616203994692,
      "loss": 2.2066173553466797,
      "step": 780
    },
    {
      "epoch": 0.4252455927869735,
      "grad_norm": 19.81163215637207,
      "learning_rate": 0.00019544081084820915,
      "loss": 2.4471906661987304,
      "step": 790
    },
    {
      "epoch": 0.43062844839187187,
      "grad_norm": 42.2238883972168,
      "learning_rate": 0.00019526205926910734,
      "loss": 2.8852027893066405,
      "step": 800
    },
    {
      "epoch": 0.43062844839187187,
      "eval_loss": 0.7064014077186584,
      "eval_runtime": 64.4405,
      "eval_samples_per_second": 51.257,
      "eval_steps_per_second": 12.818,
      "step": 800
    },
    {
      "epoch": 0.4360113039967703,
      "grad_norm": 32.852134704589844,
      "learning_rate": 0.00019507995554741205,
      "loss": 2.385086441040039,
      "step": 810
    },
    {
      "epoch": 0.44139415960166867,
      "grad_norm": 6.121428489685059,
      "learning_rate": 0.00019489450609110618,
      "loss": 1.8915294647216796,
      "step": 820
    },
    {
      "epoch": 0.4467770152065671,
      "grad_norm": 24.810670852661133,
      "learning_rate": 0.00019470571742590437,
      "loss": 2.8130746841430665,
      "step": 830
    },
    {
      "epoch": 0.45215987081146547,
      "grad_norm": 32.280818939208984,
      "learning_rate": 0.0001945135961950236,
      "loss": 2.450386619567871,
      "step": 840
    },
    {
      "epoch": 0.4575427264163639,
      "grad_norm": 48.205787658691406,
      "learning_rate": 0.0001943181491589493,
      "loss": 3.0519411087036135,
      "step": 850
    },
    {
      "epoch": 0.46292558202126227,
      "grad_norm": 22.858253479003906,
      "learning_rate": 0.00019411938319519734,
      "loss": 2.4671669006347656,
      "step": 860
    },
    {
      "epoch": 0.4683084376261607,
      "grad_norm": 45.50665283203125,
      "learning_rate": 0.00019391730529807238,
      "loss": 2.972084808349609,
      "step": 870
    },
    {
      "epoch": 0.4736912932310591,
      "grad_norm": 19.182138442993164,
      "learning_rate": 0.0001937119225784213,
      "loss": 1.9543575286865233,
      "step": 880
    },
    {
      "epoch": 0.4790741488359575,
      "grad_norm": 53.60784149169922,
      "learning_rate": 0.00019350324226338338,
      "loss": 3.198233413696289,
      "step": 890
    },
    {
      "epoch": 0.4844570044408559,
      "grad_norm": 28.96963882446289,
      "learning_rate": 0.00019329127169613564,
      "loss": 3.061937141418457,
      "step": 900
    },
    {
      "epoch": 0.4844570044408559,
      "eval_loss": 0.6995744109153748,
      "eval_runtime": 64.1823,
      "eval_samples_per_second": 51.463,
      "eval_steps_per_second": 12.87,
      "step": 900
    },
    {
      "epoch": 0.48983986004575425,
      "grad_norm": 42.73357009887695,
      "learning_rate": 0.00019307601833563475,
      "loss": 3.221673583984375,
      "step": 910
    },
    {
      "epoch": 0.4952227156506527,
      "grad_norm": 21.203781127929688,
      "learning_rate": 0.00019285748975635437,
      "loss": 2.1903615951538087,
      "step": 920
    },
    {
      "epoch": 0.500605571255551,
      "grad_norm": 21.64086151123047,
      "learning_rate": 0.0001926356936480186,
      "loss": 2.1224742889404298,
      "step": 930
    },
    {
      "epoch": 0.5059884268604494,
      "grad_norm": 35.52275466918945,
      "learning_rate": 0.0001924106378153316,
      "loss": 3.230604553222656,
      "step": 940
    },
    {
      "epoch": 0.5113712824653479,
      "grad_norm": 26.517505645751953,
      "learning_rate": 0.00019218233017770264,
      "loss": 2.1124088287353517,
      "step": 950
    },
    {
      "epoch": 0.5167541380702463,
      "grad_norm": 27.521711349487305,
      "learning_rate": 0.0001919507787689677,
      "loss": 2.729749298095703,
      "step": 960
    },
    {
      "epoch": 0.5221369936751447,
      "grad_norm": 33.03516387939453,
      "learning_rate": 0.00019171599173710662,
      "loss": 1.914764976501465,
      "step": 970
    },
    {
      "epoch": 0.527519849280043,
      "grad_norm": 28.447750091552734,
      "learning_rate": 0.00019147797734395648,
      "loss": 1.99549560546875,
      "step": 980
    },
    {
      "epoch": 0.5329027048849415,
      "grad_norm": 19.028818130493164,
      "learning_rate": 0.0001912367439649207,
      "loss": 1.8732738494873047,
      "step": 990
    },
    {
      "epoch": 0.5382855604898399,
      "grad_norm": 26.41012191772461,
      "learning_rate": 0.00019099230008867463,
      "loss": 2.936910057067871,
      "step": 1000
    },
    {
      "epoch": 0.5382855604898399,
      "eval_loss": 0.6647254228591919,
      "eval_runtime": 64.3533,
      "eval_samples_per_second": 51.326,
      "eval_steps_per_second": 12.835,
      "step": 1000
    },
    {
      "epoch": 0.5436684160947383,
      "grad_norm": 29.55500602722168,
      "learning_rate": 0.00019074465431686652,
      "loss": 2.9467599868774412,
      "step": 1010
    },
    {
      "epoch": 0.5490512716996366,
      "grad_norm": 15.729572296142578,
      "learning_rate": 0.00019049381536381503,
      "loss": 2.841005325317383,
      "step": 1020
    },
    {
      "epoch": 0.554434127304535,
      "grad_norm": 7.53200626373291,
      "learning_rate": 0.0001902397920562025,
      "loss": 1.8715387344360352,
      "step": 1030
    },
    {
      "epoch": 0.5598169829094335,
      "grad_norm": 35.68206787109375,
      "learning_rate": 0.0001899825933327644,
      "loss": 2.8048942565917967,
      "step": 1040
    },
    {
      "epoch": 0.5651998385143319,
      "grad_norm": 46.12540054321289,
      "learning_rate": 0.00018972222824397485,
      "loss": 2.888982963562012,
      "step": 1050
    },
    {
      "epoch": 0.5705826941192302,
      "grad_norm": 28.082979202270508,
      "learning_rate": 0.00018945870595172797,
      "loss": 2.8277057647705077,
      "step": 1060
    },
    {
      "epoch": 0.5759655497241286,
      "grad_norm": 8.667706489562988,
      "learning_rate": 0.00018919203572901559,
      "loss": 1.7566593170166016,
      "step": 1070
    },
    {
      "epoch": 0.5813484053290271,
      "grad_norm": 32.86021423339844,
      "learning_rate": 0.000188922226959601,
      "loss": 2.00911922454834,
      "step": 1080
    },
    {
      "epoch": 0.5867312609339255,
      "grad_norm": 20.026046752929688,
      "learning_rate": 0.00018864928913768866,
      "loss": 3.050994873046875,
      "step": 1090
    },
    {
      "epoch": 0.5921141165388238,
      "grad_norm": 34.74673843383789,
      "learning_rate": 0.00018837323186759016,
      "loss": 2.7538400650024415,
      "step": 1100
    },
    {
      "epoch": 0.5921141165388238,
      "eval_loss": 0.671667754650116,
      "eval_runtime": 64.3041,
      "eval_samples_per_second": 51.365,
      "eval_steps_per_second": 12.845,
      "step": 1100
    },
    {
      "epoch": 0.5974969721437222,
      "grad_norm": 26.802404403686523,
      "learning_rate": 0.00018809406486338618,
      "loss": 2.579293060302734,
      "step": 1110
    },
    {
      "epoch": 0.6028798277486206,
      "grad_norm": 37.77304458618164,
      "learning_rate": 0.00018781179794858478,
      "loss": 2.3671943664550783,
      "step": 1120
    },
    {
      "epoch": 0.6082626833535191,
      "grad_norm": 40.37071228027344,
      "learning_rate": 0.00018752644105577565,
      "loss": 1.8490634918212892,
      "step": 1130
    },
    {
      "epoch": 0.6136455389584174,
      "grad_norm": 58.33607482910156,
      "learning_rate": 0.0001872380042262806,
      "loss": 2.5194568634033203,
      "step": 1140
    },
    {
      "epoch": 0.6190283945633158,
      "grad_norm": 18.999311447143555,
      "learning_rate": 0.0001869464976098003,
      "loss": 2.8447465896606445,
      "step": 1150
    },
    {
      "epoch": 0.6244112501682142,
      "grad_norm": 9.258393287658691,
      "learning_rate": 0.00018665193146405695,
      "loss": 1.3028793334960938,
      "step": 1160
    },
    {
      "epoch": 0.6297941057731127,
      "grad_norm": 36.751834869384766,
      "learning_rate": 0.00018635431615443354,
      "loss": 2.4607158660888673,
      "step": 1170
    },
    {
      "epoch": 0.635176961378011,
      "grad_norm": 32.03705596923828,
      "learning_rate": 0.00018605366215360887,
      "loss": 2.4548139572143555,
      "step": 1180
    },
    {
      "epoch": 0.6405598169829094,
      "grad_norm": 83.0655746459961,
      "learning_rate": 0.00018574998004118932,
      "loss": 4.094330215454102,
      "step": 1190
    },
    {
      "epoch": 0.6459426725878078,
      "grad_norm": 16.984663009643555,
      "learning_rate": 0.00018544328050333625,
      "loss": 2.0978935241699217,
      "step": 1200
    },
    {
      "epoch": 0.6459426725878078,
      "eval_loss": 0.5826177597045898,
      "eval_runtime": 64.1932,
      "eval_samples_per_second": 51.454,
      "eval_steps_per_second": 12.867,
      "step": 1200
    },
    {
      "epoch": 0.6513255281927063,
      "grad_norm": 31.547466278076172,
      "learning_rate": 0.00018513357433239022,
      "loss": 2.314143753051758,
      "step": 1210
    },
    {
      "epoch": 0.6567083837976047,
      "grad_norm": 50.67429733276367,
      "learning_rate": 0.00018482087242649117,
      "loss": 4.206168746948242,
      "step": 1220
    },
    {
      "epoch": 0.662091239402503,
      "grad_norm": 31.81623649597168,
      "learning_rate": 0.00018450518578919475,
      "loss": 2.8035467147827147,
      "step": 1230
    },
    {
      "epoch": 0.6674740950074014,
      "grad_norm": 24.316707611083984,
      "learning_rate": 0.00018418652552908537,
      "loss": 2.3980735778808593,
      "step": 1240
    },
    {
      "epoch": 0.6728569506122998,
      "grad_norm": 34.293701171875,
      "learning_rate": 0.00018386490285938516,
      "loss": 2.651681900024414,
      "step": 1250
    },
    {
      "epoch": 0.6782398062171983,
      "grad_norm": 34.4983024597168,
      "learning_rate": 0.0001835403290975594,
      "loss": 2.2850324630737306,
      "step": 1260
    },
    {
      "epoch": 0.6836226618220966,
      "grad_norm": 31.51317596435547,
      "learning_rate": 0.00018321281566491835,
      "loss": 1.8678964614868163,
      "step": 1270
    },
    {
      "epoch": 0.689005517426995,
      "grad_norm": 88.34801483154297,
      "learning_rate": 0.0001828823740862152,
      "loss": 2.051138496398926,
      "step": 1280
    },
    {
      "epoch": 0.6943883730318934,
      "grad_norm": 23.922054290771484,
      "learning_rate": 0.00018254901598924078,
      "loss": 1.466459083557129,
      "step": 1290
    },
    {
      "epoch": 0.6997712286367919,
      "grad_norm": 39.07735824584961,
      "learning_rate": 0.00018221275310441405,
      "loss": 2.2651920318603516,
      "step": 1300
    },
    {
      "epoch": 0.6997712286367919,
      "eval_loss": 0.6614837646484375,
      "eval_runtime": 63.9775,
      "eval_samples_per_second": 51.628,
      "eval_steps_per_second": 12.911,
      "step": 1300
    },
    {
      "epoch": 0.7051540842416902,
      "grad_norm": 51.51314926147461,
      "learning_rate": 0.0001818735972643697,
      "loss": 2.400813102722168,
      "step": 1310
    },
    {
      "epoch": 0.7105369398465886,
      "grad_norm": 27.827117919921875,
      "learning_rate": 0.0001815315604035414,
      "loss": 2.1232112884521483,
      "step": 1320
    },
    {
      "epoch": 0.715919795451487,
      "grad_norm": 39.04352951049805,
      "learning_rate": 0.00018118665455774227,
      "loss": 2.0058998107910155,
      "step": 1330
    },
    {
      "epoch": 0.7213026510563854,
      "grad_norm": 21.7037296295166,
      "learning_rate": 0.00018083889186374088,
      "loss": 4.023828125,
      "step": 1340
    },
    {
      "epoch": 0.7266855066612838,
      "grad_norm": 23.27933120727539,
      "learning_rate": 0.00018048828455883455,
      "loss": 2.1123376846313477,
      "step": 1350
    },
    {
      "epoch": 0.7320683622661822,
      "grad_norm": 23.154111862182617,
      "learning_rate": 0.00018013484498041854,
      "loss": 2.0040388107299805,
      "step": 1360
    },
    {
      "epoch": 0.7374512178710806,
      "grad_norm": 36.17502212524414,
      "learning_rate": 0.000179778585565552,
      "loss": 1.7808284759521484,
      "step": 1370
    },
    {
      "epoch": 0.742834073475979,
      "grad_norm": 46.8353385925293,
      "learning_rate": 0.0001794195188505203,
      "loss": 2.6809072494506836,
      "step": 1380
    },
    {
      "epoch": 0.7482169290808774,
      "grad_norm": 19.901514053344727,
      "learning_rate": 0.00017905765747039385,
      "loss": 2.2484678268432616,
      "step": 1390
    },
    {
      "epoch": 0.7535997846857758,
      "grad_norm": 31.01380729675293,
      "learning_rate": 0.0001786930141585836,
      "loss": 1.9935230255126952,
      "step": 1400
    },
    {
      "epoch": 0.7535997846857758,
      "eval_loss": 0.5672405362129211,
      "eval_runtime": 64.147,
      "eval_samples_per_second": 51.491,
      "eval_steps_per_second": 12.877,
      "step": 1400
    },
    {
      "epoch": 0.7589826402906742,
      "grad_norm": 40.08690643310547,
      "learning_rate": 0.00017832560174639282,
      "loss": 2.275004577636719,
      "step": 1410
    },
    {
      "epoch": 0.7643654958955726,
      "grad_norm": 33.389278411865234,
      "learning_rate": 0.00017795543316256578,
      "loss": 2.6517934799194336,
      "step": 1420
    },
    {
      "epoch": 0.769748351500471,
      "grad_norm": 24.02670669555664,
      "learning_rate": 0.0001775825214328326,
      "loss": 1.8687612533569335,
      "step": 1430
    },
    {
      "epoch": 0.7751312071053694,
      "grad_norm": 39.74971008300781,
      "learning_rate": 0.00017720687967945093,
      "loss": 2.8546443939208985,
      "step": 1440
    },
    {
      "epoch": 0.7805140627102678,
      "grad_norm": 42.48725891113281,
      "learning_rate": 0.0001768285211207444,
      "loss": 2.515781784057617,
      "step": 1450
    },
    {
      "epoch": 0.7858969183151662,
      "grad_norm": 24.863798141479492,
      "learning_rate": 0.00017644745907063723,
      "loss": 1.927882766723633,
      "step": 1460
    },
    {
      "epoch": 0.7912797739200645,
      "grad_norm": 9.220629692077637,
      "learning_rate": 0.00017606370693818584,
      "loss": 1.976559829711914,
      "step": 1470
    },
    {
      "epoch": 0.796662629524963,
      "grad_norm": 48.26815414428711,
      "learning_rate": 0.000175677278227107,
      "loss": 2.111470413208008,
      "step": 1480
    },
    {
      "epoch": 0.8020454851298614,
      "grad_norm": 39.00963592529297,
      "learning_rate": 0.00017528818653530273,
      "loss": 1.7109893798828124,
      "step": 1490
    },
    {
      "epoch": 0.8074283407347598,
      "grad_norm": 24.070545196533203,
      "learning_rate": 0.0001748964455543816,
      "loss": 3.191597747802734,
      "step": 1500
    },
    {
      "epoch": 0.8074283407347598,
      "eval_loss": 0.5993195176124573,
      "eval_runtime": 63.9501,
      "eval_samples_per_second": 51.65,
      "eval_steps_per_second": 12.916,
      "step": 1500
    },
    {
      "epoch": 0.8128111963396581,
      "grad_norm": 12.18187141418457,
      "learning_rate": 0.00017450206906917713,
      "loss": 2.0544879913330076,
      "step": 1510
    },
    {
      "epoch": 0.8181940519445566,
      "grad_norm": 28.476694107055664,
      "learning_rate": 0.0001741050709572627,
      "loss": 2.3975988388061524,
      "step": 1520
    },
    {
      "epoch": 0.823576907549455,
      "grad_norm": 24.376489639282227,
      "learning_rate": 0.0001737054651884631,
      "loss": 1.8688972473144532,
      "step": 1530
    },
    {
      "epoch": 0.8289597631543534,
      "grad_norm": 38.94839096069336,
      "learning_rate": 0.00017330326582436304,
      "loss": 2.3757015228271485,
      "step": 1540
    },
    {
      "epoch": 0.8343426187592518,
      "grad_norm": 37.19301223754883,
      "learning_rate": 0.00017289848701781244,
      "loss": 2.7712066650390623,
      "step": 1550
    },
    {
      "epoch": 0.8397254743641501,
      "grad_norm": 23.972366333007812,
      "learning_rate": 0.0001724911430124281,
      "loss": 2.109449005126953,
      "step": 1560
    },
    {
      "epoch": 0.8451083299690486,
      "grad_norm": 37.52610397338867,
      "learning_rate": 0.0001720812481420929,
      "loss": 1.802728271484375,
      "step": 1570
    },
    {
      "epoch": 0.850491185573947,
      "grad_norm": 16.366283416748047,
      "learning_rate": 0.00017166881683045103,
      "loss": 2.741063117980957,
      "step": 1580
    },
    {
      "epoch": 0.8558740411788454,
      "grad_norm": 38.8548469543457,
      "learning_rate": 0.00017125386359040067,
      "loss": 1.7972919464111328,
      "step": 1590
    },
    {
      "epoch": 0.8612568967837437,
      "grad_norm": 22.872310638427734,
      "learning_rate": 0.00017083640302358326,
      "loss": 2.159262275695801,
      "step": 1600
    },
    {
      "epoch": 0.8612568967837437,
      "eval_loss": 0.4924517571926117,
      "eval_runtime": 64.2275,
      "eval_samples_per_second": 51.427,
      "eval_steps_per_second": 12.861,
      "step": 1600
    },
    {
      "epoch": 0.8666397523886422,
      "grad_norm": 27.983964920043945,
      "learning_rate": 0.00017041644981986967,
      "loss": 1.717204475402832,
      "step": 1610
    },
    {
      "epoch": 0.8720226079935406,
      "grad_norm": 21.148286819458008,
      "learning_rate": 0.00016999401875684316,
      "loss": 2.8551345825195313,
      "step": 1620
    },
    {
      "epoch": 0.877405463598439,
      "grad_norm": 80.67280578613281,
      "learning_rate": 0.0001695691246992797,
      "loss": 2.6277225494384764,
      "step": 1630
    },
    {
      "epoch": 0.8827883192033373,
      "grad_norm": 18.69855308532715,
      "learning_rate": 0.0001691417825986245,
      "loss": 2.1024648666381838,
      "step": 1640
    },
    {
      "epoch": 0.8881711748082358,
      "grad_norm": 34.107303619384766,
      "learning_rate": 0.00016871200749246626,
      "loss": 2.2313766479492188,
      "step": 1650
    },
    {
      "epoch": 0.8935540304131342,
      "grad_norm": 40.186038970947266,
      "learning_rate": 0.00016827981450400774,
      "loss": 2.4880607604980467,
      "step": 1660
    },
    {
      "epoch": 0.8989368860180326,
      "grad_norm": 70.78448486328125,
      "learning_rate": 0.00016784521884153362,
      "loss": 2.738982582092285,
      "step": 1670
    },
    {
      "epoch": 0.9043197416229309,
      "grad_norm": 36.99370193481445,
      "learning_rate": 0.00016740823579787558,
      "loss": 2.254520606994629,
      "step": 1680
    },
    {
      "epoch": 0.9097025972278293,
      "grad_norm": 39.36662673950195,
      "learning_rate": 0.00016696888074987392,
      "loss": 1.8448904037475586,
      "step": 1690
    },
    {
      "epoch": 0.9150854528327278,
      "grad_norm": 61.377532958984375,
      "learning_rate": 0.00016652716915783658,
      "loss": 2.130519485473633,
      "step": 1700
    },
    {
      "epoch": 0.9150854528327278,
      "eval_loss": 0.532718300819397,
      "eval_runtime": 64.1771,
      "eval_samples_per_second": 51.467,
      "eval_steps_per_second": 12.871,
      "step": 1700
    },
    {
      "epoch": 0.9204683084376262,
      "grad_norm": 40.985164642333984,
      "learning_rate": 0.00016608311656499504,
      "loss": 3.142055702209473,
      "step": 1710
    },
    {
      "epoch": 0.9258511640425245,
      "grad_norm": 15.224512100219727,
      "learning_rate": 0.0001656367385969575,
      "loss": 2.681681823730469,
      "step": 1720
    },
    {
      "epoch": 0.9312340196474229,
      "grad_norm": 30.765071868896484,
      "learning_rate": 0.00016518805096115888,
      "loss": 2.4531631469726562,
      "step": 1730
    },
    {
      "epoch": 0.9366168752523214,
      "grad_norm": 39.2271614074707,
      "learning_rate": 0.00016473706944630822,
      "loss": 1.7306629180908204,
      "step": 1740
    },
    {
      "epoch": 0.9419997308572198,
      "grad_norm": 28.554643630981445,
      "learning_rate": 0.00016428380992183304,
      "loss": 2.5790334701538087,
      "step": 1750
    },
    {
      "epoch": 0.9473825864621181,
      "grad_norm": 31.55245590209961,
      "learning_rate": 0.0001638282883373209,
      "loss": 2.3022281646728517,
      "step": 1760
    },
    {
      "epoch": 0.9527654420670165,
      "grad_norm": 29.733423233032227,
      "learning_rate": 0.00016337052072195823,
      "loss": 1.7397018432617188,
      "step": 1770
    },
    {
      "epoch": 0.958148297671915,
      "grad_norm": 28.3569393157959,
      "learning_rate": 0.00016291052318396625,
      "loss": 2.331684112548828,
      "step": 1780
    },
    {
      "epoch": 0.9635311532768134,
      "grad_norm": 31.536741256713867,
      "learning_rate": 0.00016244831191003406,
      "loss": 1.9013969421386718,
      "step": 1790
    },
    {
      "epoch": 0.9689140088817118,
      "grad_norm": 27.39286231994629,
      "learning_rate": 0.0001619839031647491,
      "loss": 1.9535322189331055,
      "step": 1800
    },
    {
      "epoch": 0.9689140088817118,
      "eval_loss": 0.4878155589103699,
      "eval_runtime": 63.9092,
      "eval_samples_per_second": 51.683,
      "eval_steps_per_second": 12.925,
      "step": 1800
    },
    {
      "epoch": 0.9742968644866101,
      "grad_norm": 35.60191345214844,
      "learning_rate": 0.00016151731329002505,
      "loss": 2.3475147247314454,
      "step": 1810
    },
    {
      "epoch": 0.9796797200915085,
      "grad_norm": 54.64162826538086,
      "learning_rate": 0.0001610485587045263,
      "loss": 2.4532806396484377,
      "step": 1820
    },
    {
      "epoch": 0.985062575696407,
      "grad_norm": 24.01209831237793,
      "learning_rate": 0.00016057765590309067,
      "loss": 1.9294677734375,
      "step": 1830
    },
    {
      "epoch": 0.9904454313013054,
      "grad_norm": 28.77838706970215,
      "learning_rate": 0.00016010462145614872,
      "loss": 2.6661401748657227,
      "step": 1840
    },
    {
      "epoch": 0.9958282869062037,
      "grad_norm": 23.397241592407227,
      "learning_rate": 0.0001596294720091407,
      "loss": 2.083774375915527,
      "step": 1850
    },
    {
      "epoch": 1.0010765711209797,
      "grad_norm": 24.451494216918945,
      "learning_rate": 0.0001591522242819309,
      "loss": 1.6308988571166991,
      "step": 1860
    },
    {
      "epoch": 1.0064594267258782,
      "grad_norm": 35.17308044433594,
      "learning_rate": 0.00015867289506821914,
      "loss": 2.0649885177612304,
      "step": 1870
    },
    {
      "epoch": 1.0118422823307764,
      "grad_norm": 18.505117416381836,
      "learning_rate": 0.00015819150123495005,
      "loss": 1.7077461242675782,
      "step": 1880
    },
    {
      "epoch": 1.017225137935675,
      "grad_norm": 32.787864685058594,
      "learning_rate": 0.00015770805972171935,
      "loss": 1.5830881118774414,
      "step": 1890
    },
    {
      "epoch": 1.0226079935405732,
      "grad_norm": 49.975154876708984,
      "learning_rate": 0.00015722258754017783,
      "loss": 2.320701026916504,
      "step": 1900
    },
    {
      "epoch": 1.0226079935405732,
      "eval_loss": 0.5473566055297852,
      "eval_runtime": 64.1602,
      "eval_samples_per_second": 51.481,
      "eval_steps_per_second": 12.874,
      "step": 1900
    },
    {
      "epoch": 1.0279908491454717,
      "grad_norm": 95.95193481445312,
      "learning_rate": 0.00015673510177343283,
      "loss": 2.1278076171875,
      "step": 1910
    },
    {
      "epoch": 1.0333737047503702,
      "grad_norm": 38.54022979736328,
      "learning_rate": 0.00015624561957544686,
      "loss": 1.865066146850586,
      "step": 1920
    },
    {
      "epoch": 1.0387565603552684,
      "grad_norm": 42.10824203491211,
      "learning_rate": 0.0001557541581704343,
      "loss": 2.5854557037353514,
      "step": 1930
    },
    {
      "epoch": 1.044139415960167,
      "grad_norm": 29.79697608947754,
      "learning_rate": 0.00015526073485225506,
      "loss": 1.4342741012573241,
      "step": 1940
    },
    {
      "epoch": 1.0495222715650652,
      "grad_norm": 19.559640884399414,
      "learning_rate": 0.0001547653669838061,
      "loss": 1.4350951194763184,
      "step": 1950
    },
    {
      "epoch": 1.0549051271699637,
      "grad_norm": 36.87428665161133,
      "learning_rate": 0.0001542680719964105,
      "loss": 1.7127342224121094,
      "step": 1960
    },
    {
      "epoch": 1.0602879827748621,
      "grad_norm": 20.139524459838867,
      "learning_rate": 0.000153768867389204,
      "loss": 1.6490812301635742,
      "step": 1970
    },
    {
      "epoch": 1.0656708383797604,
      "grad_norm": 32.64472579956055,
      "learning_rate": 0.0001532677707285194,
      "loss": 1.6485609054565429,
      "step": 1980
    },
    {
      "epoch": 1.0710536939846589,
      "grad_norm": 42.68798828125,
      "learning_rate": 0.00015276479964726808,
      "loss": 2.758069610595703,
      "step": 1990
    },
    {
      "epoch": 1.0764365495895571,
      "grad_norm": 39.17344284057617,
      "learning_rate": 0.0001522599718443199,
      "loss": 2.326355743408203,
      "step": 2000
    },
    {
      "epoch": 1.0764365495895571,
      "eval_loss": 0.5851911306381226,
      "eval_runtime": 63.9143,
      "eval_samples_per_second": 51.679,
      "eval_steps_per_second": 12.924,
      "step": 2000
    },
    {
      "epoch": 1.0818194051944556,
      "grad_norm": 26.366891860961914,
      "learning_rate": 0.0001517533050838802,
      "loss": 1.368880844116211,
      "step": 2010
    },
    {
      "epoch": 1.0872022607993541,
      "grad_norm": 50.400123596191406,
      "learning_rate": 0.00015124481719486465,
      "loss": 1.787089729309082,
      "step": 2020
    },
    {
      "epoch": 1.0925851164042524,
      "grad_norm": 9.160683631896973,
      "learning_rate": 0.00015073452607027214,
      "loss": 1.864046859741211,
      "step": 2030
    },
    {
      "epoch": 1.0979679720091509,
      "grad_norm": 19.865297317504883,
      "learning_rate": 0.00015022244966655482,
      "loss": 1.632607650756836,
      "step": 2040
    },
    {
      "epoch": 1.1033508276140493,
      "grad_norm": 49.69865036010742,
      "learning_rate": 0.00014970860600298642,
      "loss": 1.7247093200683594,
      "step": 2050
    },
    {
      "epoch": 1.1087336832189476,
      "grad_norm": 43.50836181640625,
      "learning_rate": 0.0001491930131610282,
      "loss": 1.298050880432129,
      "step": 2060
    },
    {
      "epoch": 1.114116538823846,
      "grad_norm": 55.050758361816406,
      "learning_rate": 0.00014867568928369256,
      "loss": 1.6693323135375977,
      "step": 2070
    },
    {
      "epoch": 1.1194993944287444,
      "grad_norm": 66.28153228759766,
      "learning_rate": 0.00014815665257490482,
      "loss": 2.2023120880126954,
      "step": 2080
    },
    {
      "epoch": 1.1248822500336428,
      "grad_norm": 38.08374786376953,
      "learning_rate": 0.00014763592129886228,
      "loss": 1.427632713317871,
      "step": 2090
    },
    {
      "epoch": 1.1302651056385413,
      "grad_norm": 42.69854736328125,
      "learning_rate": 0.000147113513779392,
      "loss": 2.1791011810302736,
      "step": 2100
    },
    {
      "epoch": 1.1302651056385413,
      "eval_loss": 0.5500441789627075,
      "eval_runtime": 63.9044,
      "eval_samples_per_second": 51.687,
      "eval_steps_per_second": 12.926,
      "step": 2100
    },
    {
      "epoch": 1.1356479612434396,
      "grad_norm": 27.943300247192383,
      "learning_rate": 0.0001465894483993057,
      "loss": 1.6217365264892578,
      "step": 2110
    },
    {
      "epoch": 1.141030816848338,
      "grad_norm": 34.99180221557617,
      "learning_rate": 0.00014606374359975286,
      "loss": 2.2697931289672852,
      "step": 2120
    },
    {
      "epoch": 1.1464136724532366,
      "grad_norm": 27.82872772216797,
      "learning_rate": 0.000145536417879572,
      "loss": 2.0169490814208983,
      "step": 2130
    },
    {
      "epoch": 1.1517965280581348,
      "grad_norm": 18.295881271362305,
      "learning_rate": 0.00014500748979463966,
      "loss": 1.2729131698608398,
      "step": 2140
    },
    {
      "epoch": 1.1571793836630333,
      "grad_norm": 18.062448501586914,
      "learning_rate": 0.00014447697795721737,
      "loss": 2.3540084838867186,
      "step": 2150
    },
    {
      "epoch": 1.1625622392679316,
      "grad_norm": 41.10131072998047,
      "learning_rate": 0.00014394490103529676,
      "loss": 2.1076675415039063,
      "step": 2160
    },
    {
      "epoch": 1.16794509487283,
      "grad_norm": 39.552982330322266,
      "learning_rate": 0.0001434112777519427,
      "loss": 2.1465015411376953,
      "step": 2170
    },
    {
      "epoch": 1.1733279504777285,
      "grad_norm": 28.34930419921875,
      "learning_rate": 0.0001428761268846344,
      "loss": 1.4876040458679198,
      "step": 2180
    },
    {
      "epoch": 1.1787108060826268,
      "grad_norm": 26.313669204711914,
      "learning_rate": 0.00014233946726460462,
      "loss": 1.5450728416442872,
      "step": 2190
    },
    {
      "epoch": 1.1840936616875253,
      "grad_norm": 14.592830657958984,
      "learning_rate": 0.00014180131777617716,
      "loss": 2.097823715209961,
      "step": 2200
    },
    {
      "epoch": 1.1840936616875253,
      "eval_loss": 0.4937271177768707,
      "eval_runtime": 64.1972,
      "eval_samples_per_second": 51.451,
      "eval_steps_per_second": 12.867,
      "step": 2200
    },
    {
      "epoch": 1.1894765172924235,
      "grad_norm": 57.555904388427734,
      "learning_rate": 0.00014126169735610225,
      "loss": 1.3126806259155273,
      "step": 2210
    },
    {
      "epoch": 1.194859372897322,
      "grad_norm": 49.10297775268555,
      "learning_rate": 0.00014072062499289023,
      "loss": 1.8571784973144532,
      "step": 2220
    },
    {
      "epoch": 1.2002422285022205,
      "grad_norm": 63.582420349121094,
      "learning_rate": 0.00014017811972614325,
      "loss": 1.6554574966430664,
      "step": 2230
    },
    {
      "epoch": 1.2056250841071188,
      "grad_norm": 38.78810501098633,
      "learning_rate": 0.0001396342006458855,
      "loss": 1.6202888488769531,
      "step": 2240
    },
    {
      "epoch": 1.2110079397120173,
      "grad_norm": 27.394638061523438,
      "learning_rate": 0.00013908888689189133,
      "loss": 2.127480697631836,
      "step": 2250
    },
    {
      "epoch": 1.2163907953169155,
      "grad_norm": 48.140830993652344,
      "learning_rate": 0.00013854219765301165,
      "loss": 1.8078197479248046,
      "step": 2260
    },
    {
      "epoch": 1.221773650921814,
      "grad_norm": 44.98652267456055,
      "learning_rate": 0.00013799415216649897,
      "loss": 2.2878448486328127,
      "step": 2270
    },
    {
      "epoch": 1.2271565065267125,
      "grad_norm": 10.705018043518066,
      "learning_rate": 0.0001374447697173303,
      "loss": 2.145868682861328,
      "step": 2280
    },
    {
      "epoch": 1.2325393621316107,
      "grad_norm": 27.23916244506836,
      "learning_rate": 0.00013689406963752844,
      "loss": 1.3377301216125488,
      "step": 2290
    },
    {
      "epoch": 1.2379222177365092,
      "grad_norm": 34.806480407714844,
      "learning_rate": 0.00013634207130548194,
      "loss": 1.9754547119140624,
      "step": 2300
    },
    {
      "epoch": 1.2379222177365092,
      "eval_loss": 0.47936686873435974,
      "eval_runtime": 64.2268,
      "eval_samples_per_second": 51.427,
      "eval_steps_per_second": 12.861,
      "step": 2300
    },
    {
      "epoch": 1.2433050733414075,
      "grad_norm": 32.58279037475586,
      "learning_rate": 0.00013578879414526292,
      "loss": 1.7109735488891602,
      "step": 2310
    },
    {
      "epoch": 1.248687928946306,
      "grad_norm": 33.94943618774414,
      "learning_rate": 0.000135234257625944,
      "loss": 2.3083892822265626,
      "step": 2320
    },
    {
      "epoch": 1.2540707845512045,
      "grad_norm": 35.721805572509766,
      "learning_rate": 0.00013467848126091266,
      "loss": 1.966238021850586,
      "step": 2330
    },
    {
      "epoch": 1.259453640156103,
      "grad_norm": 22.220964431762695,
      "learning_rate": 0.00013412148460718507,
      "loss": 1.9786643981933594,
      "step": 2340
    },
    {
      "epoch": 1.2648364957610012,
      "grad_norm": 22.529788970947266,
      "learning_rate": 0.00013356328726471753,
      "loss": 2.2097476959228515,
      "step": 2350
    },
    {
      "epoch": 1.2702193513658995,
      "grad_norm": 14.958813667297363,
      "learning_rate": 0.00013300390887571715,
      "loss": 1.8351505279541016,
      "step": 2360
    },
    {
      "epoch": 1.275602206970798,
      "grad_norm": 17.01559829711914,
      "learning_rate": 0.00013244336912395026,
      "loss": 1.9661121368408203,
      "step": 2370
    },
    {
      "epoch": 1.2809850625756964,
      "grad_norm": 28.454547882080078,
      "learning_rate": 0.00013188168773405008,
      "loss": 1.6221160888671875,
      "step": 2380
    },
    {
      "epoch": 1.286367918180595,
      "grad_norm": 41.18201446533203,
      "learning_rate": 0.00013131888447082254,
      "loss": 1.636754035949707,
      "step": 2390
    },
    {
      "epoch": 1.2917507737854932,
      "grad_norm": 34.558536529541016,
      "learning_rate": 0.00013075497913855072,
      "loss": 1.8497133255004883,
      "step": 2400
    },
    {
      "epoch": 1.2917507737854932,
      "eval_loss": 0.5467978119850159,
      "eval_runtime": 64.2912,
      "eval_samples_per_second": 51.376,
      "eval_steps_per_second": 12.848,
      "step": 2400
    },
    {
      "epoch": 1.2971336293903917,
      "grad_norm": 43.164146423339844,
      "learning_rate": 0.00013018999158029802,
      "loss": 2.0868755340576173,
      "step": 2410
    },
    {
      "epoch": 1.30251648499529,
      "grad_norm": 13.502433776855469,
      "learning_rate": 0.00012962394167720997,
      "loss": 2.3471694946289063,
      "step": 2420
    },
    {
      "epoch": 1.3078993406001884,
      "grad_norm": 11.673425674438477,
      "learning_rate": 0.00012905684934781447,
      "loss": 0.8880938529968262,
      "step": 2430
    },
    {
      "epoch": 1.313282196205087,
      "grad_norm": 24.767826080322266,
      "learning_rate": 0.00012848873454732106,
      "loss": 1.9171413421630858,
      "step": 2440
    },
    {
      "epoch": 1.3186650518099852,
      "grad_norm": 31.41339111328125,
      "learning_rate": 0.0001279196172669186,
      "loss": 1.550046443939209,
      "step": 2450
    },
    {
      "epoch": 1.3240479074148837,
      "grad_norm": 42.19283676147461,
      "learning_rate": 0.00012734951753307194,
      "loss": 1.5069122314453125,
      "step": 2460
    },
    {
      "epoch": 1.329430763019782,
      "grad_norm": 34.4764289855957,
      "learning_rate": 0.00012677845540681702,
      "loss": 1.5139981269836427,
      "step": 2470
    },
    {
      "epoch": 1.3348136186246804,
      "grad_norm": 45.293399810791016,
      "learning_rate": 0.00012620645098305514,
      "loss": 2.5238950729370115,
      "step": 2480
    },
    {
      "epoch": 1.3401964742295789,
      "grad_norm": 21.721302032470703,
      "learning_rate": 0.0001256335243898458,
      "loss": 2.452337646484375,
      "step": 2490
    },
    {
      "epoch": 1.3455793298344771,
      "grad_norm": 32.277462005615234,
      "learning_rate": 0.00012505969578769827,
      "loss": 1.747513198852539,
      "step": 2500
    },
    {
      "epoch": 1.3455793298344771,
      "eval_loss": 0.5084195137023926,
      "eval_runtime": 64.3491,
      "eval_samples_per_second": 51.329,
      "eval_steps_per_second": 12.836,
      "step": 2500
    },
    {
      "epoch": 1.3509621854393756,
      "grad_norm": 28.49180793762207,
      "learning_rate": 0.00012448498536886242,
      "loss": 1.7691278457641602,
      "step": 2510
    },
    {
      "epoch": 1.356345041044274,
      "grad_norm": 19.427194595336914,
      "learning_rate": 0.00012390941335661793,
      "loss": 1.7051538467407226,
      "step": 2520
    },
    {
      "epoch": 1.3617278966491724,
      "grad_norm": 25.067476272583008,
      "learning_rate": 0.00012333300000456294,
      "loss": 1.2319503784179688,
      "step": 2530
    },
    {
      "epoch": 1.3671107522540709,
      "grad_norm": 43.21786880493164,
      "learning_rate": 0.00012275576559590107,
      "loss": 1.9394664764404297,
      "step": 2540
    },
    {
      "epoch": 1.3724936078589691,
      "grad_norm": 33.55292510986328,
      "learning_rate": 0.0001221777304427278,
      "loss": 1.6529422760009767,
      "step": 2550
    },
    {
      "epoch": 1.3778764634638676,
      "grad_norm": 17.388324737548828,
      "learning_rate": 0.00012159891488531583,
      "loss": 1.6362987518310548,
      "step": 2560
    },
    {
      "epoch": 1.3832593190687659,
      "grad_norm": 11.812196731567383,
      "learning_rate": 0.00012101933929139911,
      "loss": 1.3777896881103515,
      "step": 2570
    },
    {
      "epoch": 1.3886421746736644,
      "grad_norm": 10.543932914733887,
      "learning_rate": 0.00012043902405545635,
      "loss": 1.7387680053710937,
      "step": 2580
    },
    {
      "epoch": 1.3940250302785628,
      "grad_norm": 25.3862247467041,
      "learning_rate": 0.00011985798959799315,
      "loss": 1.6691925048828125,
      "step": 2590
    },
    {
      "epoch": 1.399407885883461,
      "grad_norm": 44.17007064819336,
      "learning_rate": 0.00011927625636482364,
      "loss": 1.8721656799316406,
      "step": 2600
    },
    {
      "epoch": 1.399407885883461,
      "eval_loss": 0.5541885495185852,
      "eval_runtime": 63.9819,
      "eval_samples_per_second": 51.624,
      "eval_steps_per_second": 12.91,
      "step": 2600
    },
    {
      "epoch": 1.4047907414883596,
      "grad_norm": 30.539979934692383,
      "learning_rate": 0.00011869384482635087,
      "loss": 1.5085716247558594,
      "step": 2610
    },
    {
      "epoch": 1.4101735970932578,
      "grad_norm": 39.790950775146484,
      "learning_rate": 0.00011811077547684652,
      "loss": 1.9276805877685548,
      "step": 2620
    },
    {
      "epoch": 1.4155564526981563,
      "grad_norm": 26.07389259338379,
      "learning_rate": 0.00011752706883372975,
      "loss": 1.6762975692749023,
      "step": 2630
    },
    {
      "epoch": 1.4209393083030548,
      "grad_norm": 19.482772827148438,
      "learning_rate": 0.00011694274543684525,
      "loss": 1.6249858856201171,
      "step": 2640
    },
    {
      "epoch": 1.4263221639079533,
      "grad_norm": 27.41213607788086,
      "learning_rate": 0.00011635782584774042,
      "loss": 1.7710115432739257,
      "step": 2650
    },
    {
      "epoch": 1.4317050195128516,
      "grad_norm": 44.63675308227539,
      "learning_rate": 0.00011577233064894184,
      "loss": 1.3814048767089844,
      "step": 2660
    },
    {
      "epoch": 1.43708787511775,
      "grad_norm": 35.491634368896484,
      "learning_rate": 0.00011518628044323105,
      "loss": 2.505926322937012,
      "step": 2670
    },
    {
      "epoch": 1.4424707307226483,
      "grad_norm": 19.543800354003906,
      "learning_rate": 0.00011459969585291952,
      "loss": 1.4784714698791503,
      "step": 2680
    },
    {
      "epoch": 1.4478535863275468,
      "grad_norm": 17.39204978942871,
      "learning_rate": 0.00011401259751912294,
      "loss": 1.8356576919555665,
      "step": 2690
    },
    {
      "epoch": 1.4532364419324453,
      "grad_norm": 30.696813583374023,
      "learning_rate": 0.000113425006101035,
      "loss": 1.7911975860595704,
      "step": 2700
    },
    {
      "epoch": 1.4532364419324453,
      "eval_loss": 0.4373091459274292,
      "eval_runtime": 63.9584,
      "eval_samples_per_second": 51.643,
      "eval_steps_per_second": 12.915,
      "step": 2700
    },
    {
      "epoch": 1.4586192975373435,
      "grad_norm": 62.493797302246094,
      "learning_rate": 0.00011283694227520032,
      "loss": 1.1635123252868653,
      "step": 2710
    },
    {
      "epoch": 1.464002153142242,
      "grad_norm": 37.97821807861328,
      "learning_rate": 0.00011224842673478692,
      "loss": 1.40887451171875,
      "step": 2720
    },
    {
      "epoch": 1.4693850087471403,
      "grad_norm": 29.938425064086914,
      "learning_rate": 0.00011165948018885803,
      "loss": 1.6324567794799805,
      "step": 2730
    },
    {
      "epoch": 1.4747678643520388,
      "grad_norm": 20.280174255371094,
      "learning_rate": 0.00011107012336164342,
      "loss": 1.2600415229797364,
      "step": 2740
    },
    {
      "epoch": 1.4801507199569373,
      "grad_norm": 37.221309661865234,
      "learning_rate": 0.00011048037699181007,
      "loss": 1.7051061630249023,
      "step": 2750
    },
    {
      "epoch": 1.4855335755618355,
      "grad_norm": 10.49577808380127,
      "learning_rate": 0.00010989026183173241,
      "loss": 1.4941570281982421,
      "step": 2760
    },
    {
      "epoch": 1.490916431166734,
      "grad_norm": 22.56839370727539,
      "learning_rate": 0.00010929979864676214,
      "loss": 1.240116786956787,
      "step": 2770
    },
    {
      "epoch": 1.4962992867716323,
      "grad_norm": 19.91226577758789,
      "learning_rate": 0.00010870900821449747,
      "loss": 1.3393290519714356,
      "step": 2780
    },
    {
      "epoch": 1.5016821423765307,
      "grad_norm": 32.016357421875,
      "learning_rate": 0.00010811791132405202,
      "loss": 1.7406791687011718,
      "step": 2790
    },
    {
      "epoch": 1.5070649979814292,
      "grad_norm": 50.24219512939453,
      "learning_rate": 0.00010752652877532318,
      "loss": 1.9366174697875977,
      "step": 2800
    },
    {
      "epoch": 1.5070649979814292,
      "eval_loss": 0.43302831053733826,
      "eval_runtime": 63.946,
      "eval_samples_per_second": 51.653,
      "eval_steps_per_second": 12.917,
      "step": 2800
    },
    {
      "epoch": 1.5124478535863275,
      "grad_norm": 35.96293258666992,
      "learning_rate": 0.00010693488137826042,
      "loss": 1.8892341613769532,
      "step": 2810
    },
    {
      "epoch": 1.517830709191226,
      "grad_norm": 20.087125778198242,
      "learning_rate": 0.00010634298995213267,
      "loss": 1.5483041763305665,
      "step": 2820
    },
    {
      "epoch": 1.5232135647961242,
      "grad_norm": 44.48786926269531,
      "learning_rate": 0.00010575087532479608,
      "loss": 2.049909210205078,
      "step": 2830
    },
    {
      "epoch": 1.5285964204010227,
      "grad_norm": 13.480371475219727,
      "learning_rate": 0.0001051585583319608,
      "loss": 1.4279634475708007,
      "step": 2840
    },
    {
      "epoch": 1.5339792760059212,
      "grad_norm": 12.939706802368164,
      "learning_rate": 0.0001045660598164581,
      "loss": 1.3665541648864745,
      "step": 2850
    },
    {
      "epoch": 1.5393621316108197,
      "grad_norm": 12.383955001831055,
      "learning_rate": 0.00010397340062750666,
      "loss": 1.9608146667480468,
      "step": 2860
    },
    {
      "epoch": 1.544744987215718,
      "grad_norm": 31.020015716552734,
      "learning_rate": 0.00010338060161997911,
      "loss": 1.3746890068054198,
      "step": 2870
    },
    {
      "epoch": 1.5501278428206162,
      "grad_norm": 38.224365234375,
      "learning_rate": 0.00010278768365366809,
      "loss": 1.9677223205566405,
      "step": 2880
    },
    {
      "epoch": 1.5555106984255147,
      "grad_norm": 16.846071243286133,
      "learning_rate": 0.00010219466759255225,
      "loss": 2.040707588195801,
      "step": 2890
    },
    {
      "epoch": 1.5608935540304132,
      "grad_norm": 14.602941513061523,
      "learning_rate": 0.00010160157430406201,
      "loss": 1.4509224891662598,
      "step": 2900
    },
    {
      "epoch": 1.5608935540304132,
      "eval_loss": 0.4190095067024231,
      "eval_runtime": 63.6149,
      "eval_samples_per_second": 51.922,
      "eval_steps_per_second": 12.984,
      "step": 2900
    },
    {
      "epoch": 1.5662764096353117,
      "grad_norm": 6.937775611877441,
      "learning_rate": 0.00010100842465834537,
      "loss": 1.9573307037353516,
      "step": 2910
    },
    {
      "epoch": 1.57165926524021,
      "grad_norm": 30.520694732666016,
      "learning_rate": 0.00010041523952753346,
      "loss": 1.8950252532958984,
      "step": 2920
    },
    {
      "epoch": 1.5770421208451082,
      "grad_norm": 38.1561279296875,
      "learning_rate": 9.982203978500608e-05,
      "loss": 1.7020187377929688,
      "step": 2930
    },
    {
      "epoch": 1.5824249764500067,
      "grad_norm": 27.7998104095459,
      "learning_rate": 9.922884630465717e-05,
      "loss": 1.7406475067138671,
      "step": 2940
    },
    {
      "epoch": 1.5878078320549052,
      "grad_norm": 20.777448654174805,
      "learning_rate": 9.86356799601603e-05,
      "loss": 1.344972515106201,
      "step": 2950
    },
    {
      "epoch": 1.5931906876598037,
      "grad_norm": 18.08922004699707,
      "learning_rate": 9.804256162423427e-05,
      "loss": 1.6805049896240234,
      "step": 2960
    },
    {
      "epoch": 1.598573543264702,
      "grad_norm": 15.266389846801758,
      "learning_rate": 9.744951216790837e-05,
      "loss": 1.569887638092041,
      "step": 2970
    },
    {
      "epoch": 1.6039563988696002,
      "grad_norm": 24.077733993530273,
      "learning_rate": 9.685655245978823e-05,
      "loss": 1.4327526092529297,
      "step": 2980
    },
    {
      "epoch": 1.6093392544744987,
      "grad_norm": 29.76258087158203,
      "learning_rate": 9.626370336532132e-05,
      "loss": 1.696345901489258,
      "step": 2990
    },
    {
      "epoch": 1.6147221100793971,
      "grad_norm": 22.80702781677246,
      "learning_rate": 9.567098574606279e-05,
      "loss": 0.9164070129394531,
      "step": 3000
    },
    {
      "epoch": 1.6147221100793971,
      "eval_loss": 0.45202314853668213,
      "eval_runtime": 63.7845,
      "eval_samples_per_second": 51.784,
      "eval_steps_per_second": 12.95,
      "step": 3000
    },
    {
      "epoch": 1.6201049656842956,
      "grad_norm": 28.408498764038086,
      "learning_rate": 9.507842045894128e-05,
      "loss": 1.2415037155151367,
      "step": 3010
    },
    {
      "epoch": 1.625487821289194,
      "grad_norm": 17.381816864013672,
      "learning_rate": 9.448602835552512e-05,
      "loss": 1.1174392700195312,
      "step": 3020
    },
    {
      "epoch": 1.6308706768940922,
      "grad_norm": 41.50532913208008,
      "learning_rate": 9.389383028128866e-05,
      "loss": 2.0566314697265624,
      "step": 3030
    },
    {
      "epoch": 1.6362535324989906,
      "grad_norm": 43.86766052246094,
      "learning_rate": 9.330184707487837e-05,
      "loss": 1.5256672859191895,
      "step": 3040
    },
    {
      "epoch": 1.6416363881038891,
      "grad_norm": 22.702362060546875,
      "learning_rate": 9.271009956738004e-05,
      "loss": 1.5772756576538085,
      "step": 3050
    },
    {
      "epoch": 1.6470192437087876,
      "grad_norm": 20.469390869140625,
      "learning_rate": 9.211860858158538e-05,
      "loss": 1.4132847785949707,
      "step": 3060
    },
    {
      "epoch": 1.6524020993136859,
      "grad_norm": 23.878665924072266,
      "learning_rate": 9.152739493125962e-05,
      "loss": 1.7029796600341798,
      "step": 3070
    },
    {
      "epoch": 1.6577849549185844,
      "grad_norm": 20.67031478881836,
      "learning_rate": 9.09364794204087e-05,
      "loss": 1.7173995971679688,
      "step": 3080
    },
    {
      "epoch": 1.6631678105234826,
      "grad_norm": 26.168128967285156,
      "learning_rate": 9.034588284254765e-05,
      "loss": 1.3580098152160645,
      "step": 3090
    },
    {
      "epoch": 1.668550666128381,
      "grad_norm": 36.75242233276367,
      "learning_rate": 8.975562597996855e-05,
      "loss": 1.6930839538574218,
      "step": 3100
    },
    {
      "epoch": 1.668550666128381,
      "eval_loss": 0.4625987410545349,
      "eval_runtime": 63.5645,
      "eval_samples_per_second": 51.963,
      "eval_steps_per_second": 12.995,
      "step": 3100
    },
    {
      "epoch": 1.6739335217332796,
      "grad_norm": 26.191268920898438,
      "learning_rate": 8.916572960300947e-05,
      "loss": 2.067066955566406,
      "step": 3110
    },
    {
      "epoch": 1.679316377338178,
      "grad_norm": 30.239843368530273,
      "learning_rate": 8.857621446932334e-05,
      "loss": 1.3582192420959474,
      "step": 3120
    },
    {
      "epoch": 1.6846992329430763,
      "grad_norm": 11.36849308013916,
      "learning_rate": 8.798710132314778e-05,
      "loss": 1.4313419342041016,
      "step": 3130
    },
    {
      "epoch": 1.6900820885479746,
      "grad_norm": 24.18775177001953,
      "learning_rate": 8.739841089457494e-05,
      "loss": 1.6131771087646485,
      "step": 3140
    },
    {
      "epoch": 1.695464944152873,
      "grad_norm": 17.61792755126953,
      "learning_rate": 8.681016389882217e-05,
      "loss": 1.5064881324768067,
      "step": 3150
    },
    {
      "epoch": 1.7008477997577716,
      "grad_norm": 36.965396881103516,
      "learning_rate": 8.622238103550293e-05,
      "loss": 2.0960357666015623,
      "step": 3160
    },
    {
      "epoch": 1.70623065536267,
      "grad_norm": 15.143284797668457,
      "learning_rate": 8.563508298789865e-05,
      "loss": 1.0306715011596679,
      "step": 3170
    },
    {
      "epoch": 1.7116135109675683,
      "grad_norm": 21.073841094970703,
      "learning_rate": 8.504829042223055e-05,
      "loss": 1.4934981346130372,
      "step": 3180
    },
    {
      "epoch": 1.7169963665724666,
      "grad_norm": 14.031085968017578,
      "learning_rate": 8.44620239869328e-05,
      "loss": 1.470815372467041,
      "step": 3190
    },
    {
      "epoch": 1.722379222177365,
      "grad_norm": 12.69394302368164,
      "learning_rate": 8.387630431192562e-05,
      "loss": 1.4816364288330077,
      "step": 3200
    },
    {
      "epoch": 1.722379222177365,
      "eval_loss": 0.4099518358707428,
      "eval_runtime": 63.7017,
      "eval_samples_per_second": 51.851,
      "eval_steps_per_second": 12.967,
      "step": 3200
    },
    {
      "epoch": 1.7277620777822635,
      "grad_norm": 36.04399871826172,
      "learning_rate": 8.329115200788962e-05,
      "loss": 1.2866575241088867,
      "step": 3210
    },
    {
      "epoch": 1.733144933387162,
      "grad_norm": 15.661164283752441,
      "learning_rate": 8.270658766554023e-05,
      "loss": 1.2794793128967286,
      "step": 3220
    },
    {
      "epoch": 1.7385277889920603,
      "grad_norm": 14.594550132751465,
      "learning_rate": 8.212263185490347e-05,
      "loss": 0.900571060180664,
      "step": 3230
    },
    {
      "epoch": 1.7439106445969585,
      "grad_norm": 23.809280395507812,
      "learning_rate": 8.153930512459182e-05,
      "loss": 1.2999605178833007,
      "step": 3240
    },
    {
      "epoch": 1.749293500201857,
      "grad_norm": 24.3289737701416,
      "learning_rate": 8.095662800108141e-05,
      "loss": 0.756270456314087,
      "step": 3250
    },
    {
      "epoch": 1.7546763558067555,
      "grad_norm": 55.45784378051758,
      "learning_rate": 8.037462098798942e-05,
      "loss": 1.5912586212158204,
      "step": 3260
    },
    {
      "epoch": 1.760059211411654,
      "grad_norm": 20.93805694580078,
      "learning_rate": 7.97933045653529e-05,
      "loss": 1.602321243286133,
      "step": 3270
    },
    {
      "epoch": 1.7654420670165523,
      "grad_norm": 37.905189514160156,
      "learning_rate": 7.92126991889079e-05,
      "loss": 1.5339111328125,
      "step": 3280
    },
    {
      "epoch": 1.7708249226214505,
      "grad_norm": 24.21660804748535,
      "learning_rate": 7.863282528936978e-05,
      "loss": 1.646210289001465,
      "step": 3290
    },
    {
      "epoch": 1.776207778226349,
      "grad_norm": 18.371828079223633,
      "learning_rate": 7.805370327171402e-05,
      "loss": 1.3416717529296875,
      "step": 3300
    },
    {
      "epoch": 1.776207778226349,
      "eval_loss": 0.4250826835632324,
      "eval_runtime": 63.8411,
      "eval_samples_per_second": 51.738,
      "eval_steps_per_second": 12.938,
      "step": 3300
    },
    {
      "epoch": 1.7815906338312475,
      "grad_norm": 37.405426025390625,
      "learning_rate": 7.747535351445871e-05,
      "loss": 1.5578692436218262,
      "step": 3310
    },
    {
      "epoch": 1.786973489436146,
      "grad_norm": 62.97222900390625,
      "learning_rate": 7.689779636894687e-05,
      "loss": 2.0751113891601562,
      "step": 3320
    },
    {
      "epoch": 1.7923563450410442,
      "grad_norm": 24.98927879333496,
      "learning_rate": 7.632105215863072e-05,
      "loss": 1.8148933410644532,
      "step": 3330
    },
    {
      "epoch": 1.7977392006459427,
      "grad_norm": 17.803342819213867,
      "learning_rate": 7.57451411783564e-05,
      "loss": 1.1781111717224122,
      "step": 3340
    },
    {
      "epoch": 1.803122056250841,
      "grad_norm": 31.60127067565918,
      "learning_rate": 7.517008369364973e-05,
      "loss": 1.2258566856384276,
      "step": 3350
    },
    {
      "epoch": 1.8085049118557395,
      "grad_norm": 21.834125518798828,
      "learning_rate": 7.459589994000331e-05,
      "loss": 1.8756603240966796,
      "step": 3360
    },
    {
      "epoch": 1.813887767460638,
      "grad_norm": 8.82114315032959,
      "learning_rate": 7.402261012216418e-05,
      "loss": 1.5807505607604981,
      "step": 3370
    },
    {
      "epoch": 1.8192706230655362,
      "grad_norm": 6.503419399261475,
      "learning_rate": 7.345023441342312e-05,
      "loss": 1.110390567779541,
      "step": 3380
    },
    {
      "epoch": 1.8246534786704347,
      "grad_norm": 17.47008514404297,
      "learning_rate": 7.287879295490454e-05,
      "loss": 1.7794336318969726,
      "step": 3390
    },
    {
      "epoch": 1.830036334275333,
      "grad_norm": 20.7091121673584,
      "learning_rate": 7.230830585485799e-05,
      "loss": 1.1825596809387207,
      "step": 3400
    },
    {
      "epoch": 1.830036334275333,
      "eval_loss": 0.3694857954978943,
      "eval_runtime": 63.578,
      "eval_samples_per_second": 51.952,
      "eval_steps_per_second": 12.992,
      "step": 3400
    },
    {
      "epoch": 1.8354191898802314,
      "grad_norm": 26.3618221282959,
      "learning_rate": 7.173879318795027e-05,
      "loss": 1.3671013832092285,
      "step": 3410
    },
    {
      "epoch": 1.84080204548513,
      "grad_norm": 14.309059143066406,
      "learning_rate": 7.117027499455938e-05,
      "loss": 1.7152782440185548,
      "step": 3420
    },
    {
      "epoch": 1.8461849010900284,
      "grad_norm": 14.735411643981934,
      "learning_rate": 7.060277128006896e-05,
      "loss": 1.2705150604248048,
      "step": 3430
    },
    {
      "epoch": 1.8515677566949267,
      "grad_norm": 37.376731872558594,
      "learning_rate": 7.003630201416469e-05,
      "loss": 0.823548698425293,
      "step": 3440
    },
    {
      "epoch": 1.856950612299825,
      "grad_norm": 30.746421813964844,
      "learning_rate": 6.947088713013127e-05,
      "loss": 1.9406600952148438,
      "step": 3450
    },
    {
      "epoch": 1.8623334679047234,
      "grad_norm": 28.229276657104492,
      "learning_rate": 6.89065465241513e-05,
      "loss": 1.542259120941162,
      "step": 3460
    },
    {
      "epoch": 1.867716323509622,
      "grad_norm": 6.486597537994385,
      "learning_rate": 6.834330005460472e-05,
      "loss": 1.522111701965332,
      "step": 3470
    },
    {
      "epoch": 1.8730991791145204,
      "grad_norm": 21.441097259521484,
      "learning_rate": 6.778116754137058e-05,
      "loss": 1.2606025695800782,
      "step": 3480
    },
    {
      "epoch": 1.8784820347194187,
      "grad_norm": 7.534696102142334,
      "learning_rate": 6.722016876512916e-05,
      "loss": 1.5245902061462402,
      "step": 3490
    },
    {
      "epoch": 1.883864890324317,
      "grad_norm": 32.19171142578125,
      "learning_rate": 6.666032346666613e-05,
      "loss": 1.5431558609008789,
      "step": 3500
    },
    {
      "epoch": 1.883864890324317,
      "eval_loss": 0.37625691294670105,
      "eval_runtime": 63.829,
      "eval_samples_per_second": 51.748,
      "eval_steps_per_second": 12.941,
      "step": 3500
    },
    {
      "epoch": 1.8892477459292154,
      "grad_norm": 23.33196258544922,
      "learning_rate": 6.610165134617778e-05,
      "loss": 1.21827974319458,
      "step": 3510
    },
    {
      "epoch": 1.894630601534114,
      "grad_norm": 40.78279113769531,
      "learning_rate": 6.554417206257795e-05,
      "loss": 1.7008653640747071,
      "step": 3520
    },
    {
      "epoch": 1.9000134571390124,
      "grad_norm": 23.865211486816406,
      "learning_rate": 6.498790523280607e-05,
      "loss": 1.0905473709106446,
      "step": 3530
    },
    {
      "epoch": 1.9053963127439106,
      "grad_norm": 26.986589431762695,
      "learning_rate": 6.44328704311371e-05,
      "loss": 1.2692423820495606,
      "step": 3540
    },
    {
      "epoch": 1.910779168348809,
      "grad_norm": 15.614501953125,
      "learning_rate": 6.387908718849239e-05,
      "loss": 1.1699938774108887,
      "step": 3550
    },
    {
      "epoch": 1.9161620239537074,
      "grad_norm": 16.81165313720703,
      "learning_rate": 6.332657499175291e-05,
      "loss": 1.7018922805786132,
      "step": 3560
    },
    {
      "epoch": 1.9215448795586059,
      "grad_norm": 23.331390380859375,
      "learning_rate": 6.277535328307296e-05,
      "loss": 1.1255874633789062,
      "step": 3570
    },
    {
      "epoch": 1.9269277351635044,
      "grad_norm": 24.57452964782715,
      "learning_rate": 6.222544145919653e-05,
      "loss": 2.053675079345703,
      "step": 3580
    },
    {
      "epoch": 1.9323105907684026,
      "grad_norm": 13.287324905395508,
      "learning_rate": 6.167685887077444e-05,
      "loss": 1.2274766921997071,
      "step": 3590
    },
    {
      "epoch": 1.9376934463733009,
      "grad_norm": 18.1403865814209,
      "learning_rate": 6.112962482168356e-05,
      "loss": 1.2781085968017578,
      "step": 3600
    },
    {
      "epoch": 1.9376934463733009,
      "eval_loss": 0.36828258633613586,
      "eval_runtime": 63.8932,
      "eval_samples_per_second": 51.696,
      "eval_steps_per_second": 12.928,
      "step": 3600
    },
    {
      "epoch": 1.9430763019781994,
      "grad_norm": 19.552724838256836,
      "learning_rate": 6.058375856834742e-05,
      "loss": 1.5518939971923829,
      "step": 3610
    },
    {
      "epoch": 1.9484591575830978,
      "grad_norm": 25.114530563354492,
      "learning_rate": 6.003927931905875e-05,
      "loss": 1.3678104400634765,
      "step": 3620
    },
    {
      "epoch": 1.9538420131879963,
      "grad_norm": 24.832029342651367,
      "learning_rate": 5.949620623330351e-05,
      "loss": 1.3850306510925292,
      "step": 3630
    },
    {
      "epoch": 1.9592248687928946,
      "grad_norm": 28.975635528564453,
      "learning_rate": 5.895455842108656e-05,
      "loss": 2.144721031188965,
      "step": 3640
    },
    {
      "epoch": 1.964607724397793,
      "grad_norm": 20.551748275756836,
      "learning_rate": 5.841435494225946e-05,
      "loss": 1.024734878540039,
      "step": 3650
    },
    {
      "epoch": 1.9699905800026913,
      "grad_norm": 28.49460792541504,
      "learning_rate": 5.7875614805849464e-05,
      "loss": 1.2350109100341797,
      "step": 3660
    },
    {
      "epoch": 1.9753734356075898,
      "grad_norm": 46.727195739746094,
      "learning_rate": 5.7338356969390996e-05,
      "loss": 1.4352140426635742,
      "step": 3670
    },
    {
      "epoch": 1.9807562912124883,
      "grad_norm": 22.224246978759766,
      "learning_rate": 5.680260033825819e-05,
      "loss": 1.4729737281799316,
      "step": 3680
    },
    {
      "epoch": 1.9861391468173866,
      "grad_norm": 21.090652465820312,
      "learning_rate": 5.626836376499991e-05,
      "loss": 1.332131576538086,
      "step": 3690
    },
    {
      "epoch": 1.991522002422285,
      "grad_norm": 19.748018264770508,
      "learning_rate": 5.573566604867616e-05,
      "loss": 1.1651579856872558,
      "step": 3700
    },
    {
      "epoch": 1.991522002422285,
      "eval_loss": 0.3755303621292114,
      "eval_runtime": 63.642,
      "eval_samples_per_second": 51.9,
      "eval_steps_per_second": 12.979,
      "step": 3700
    },
    {
      "epoch": 1.9969048580271833,
      "grad_norm": 26.874662399291992,
      "learning_rate": 5.5204525934196715e-05,
      "loss": 1.8782270431518555,
      "step": 3710
    },
    {
      "epoch": 2.0021531422419594,
      "grad_norm": 11.230367660522461,
      "learning_rate": 5.4674962111661396e-05,
      "loss": 1.1111202239990234,
      "step": 3720
    },
    {
      "epoch": 2.007535997846858,
      "grad_norm": 11.470046043395996,
      "learning_rate": 5.4146993215702614e-05,
      "loss": 1.2164538383483887,
      "step": 3730
    },
    {
      "epoch": 2.0129188534517564,
      "grad_norm": 14.980729103088379,
      "learning_rate": 5.362063782482918e-05,
      "loss": 0.72305908203125,
      "step": 3740
    },
    {
      "epoch": 2.0183017090566544,
      "grad_norm": 22.74243927001953,
      "learning_rate": 5.309591446077316e-05,
      "loss": 0.9697726249694825,
      "step": 3750
    },
    {
      "epoch": 2.023684564661553,
      "grad_norm": 39.94514083862305,
      "learning_rate": 5.257284158783764e-05,
      "loss": 1.0772478103637695,
      "step": 3760
    },
    {
      "epoch": 2.0290674202664514,
      "grad_norm": 14.271464347839355,
      "learning_rate": 5.20514376122472e-05,
      "loss": 0.7649795055389405,
      "step": 3770
    },
    {
      "epoch": 2.03445027587135,
      "grad_norm": 46.278282165527344,
      "learning_rate": 5.1531720881500156e-05,
      "loss": 1.7668495178222656,
      "step": 3780
    },
    {
      "epoch": 2.0398331314762483,
      "grad_norm": 26.685871124267578,
      "learning_rate": 5.1013709683723044e-05,
      "loss": 0.9311368942260743,
      "step": 3790
    },
    {
      "epoch": 2.0452159870811464,
      "grad_norm": 19.982656478881836,
      "learning_rate": 5.0497422247026895e-05,
      "loss": 1.1055089950561523,
      "step": 3800
    },
    {
      "epoch": 2.0452159870811464,
      "eval_loss": 0.4047035276889801,
      "eval_runtime": 63.58,
      "eval_samples_per_second": 51.95,
      "eval_steps_per_second": 12.992,
      "step": 3800
    },
    {
      "epoch": 2.050598842686045,
      "grad_norm": 12.345768928527832,
      "learning_rate": 4.9982876738866094e-05,
      "loss": 0.7015519618988038,
      "step": 3810
    },
    {
      "epoch": 2.0559816982909433,
      "grad_norm": 33.307762145996094,
      "learning_rate": 4.9470091265398754e-05,
      "loss": 0.6976238250732422,
      "step": 3820
    },
    {
      "epoch": 2.061364553895842,
      "grad_norm": 38.2665901184082,
      "learning_rate": 4.8959083870849864e-05,
      "loss": 1.1541186332702638,
      "step": 3830
    },
    {
      "epoch": 2.0667474095007403,
      "grad_norm": 22.593074798583984,
      "learning_rate": 4.844987253687618e-05,
      "loss": 0.4699763298034668,
      "step": 3840
    },
    {
      "epoch": 2.0721302651056384,
      "grad_norm": 21.039308547973633,
      "learning_rate": 4.794247518193353e-05,
      "loss": 0.8926765441894531,
      "step": 3850
    },
    {
      "epoch": 2.077513120710537,
      "grad_norm": 24.009326934814453,
      "learning_rate": 4.743690966064626e-05,
      "loss": 0.7189272403717041,
      "step": 3860
    },
    {
      "epoch": 2.0828959763154353,
      "grad_norm": 8.515445709228516,
      "learning_rate": 4.6933193763179105e-05,
      "loss": 0.5480011940002442,
      "step": 3870
    },
    {
      "epoch": 2.088278831920334,
      "grad_norm": 19.915380477905273,
      "learning_rate": 4.64313452146108e-05,
      "loss": 0.7470302581787109,
      "step": 3880
    },
    {
      "epoch": 2.0936616875252323,
      "grad_norm": 5.266428470611572,
      "learning_rate": 4.593138167431087e-05,
      "loss": 0.592840576171875,
      "step": 3890
    },
    {
      "epoch": 2.0990445431301303,
      "grad_norm": 38.52595520019531,
      "learning_rate": 4.54333207353178e-05,
      "loss": 0.9797579765319824,
      "step": 3900
    },
    {
      "epoch": 2.0990445431301303,
      "eval_loss": 0.4082370102405548,
      "eval_runtime": 63.7106,
      "eval_samples_per_second": 51.844,
      "eval_steps_per_second": 12.965,
      "step": 3900
    },
    {
      "epoch": 2.104427398735029,
      "grad_norm": 38.00594711303711,
      "learning_rate": 4.493717992372014e-05,
      "loss": 0.6796042919158936,
      "step": 3910
    },
    {
      "epoch": 2.1098102543399273,
      "grad_norm": 22.110828399658203,
      "learning_rate": 4.444297669803981e-05,
      "loss": 1.6094877243041992,
      "step": 3920
    },
    {
      "epoch": 2.115193109944826,
      "grad_norm": 14.167129516601562,
      "learning_rate": 4.395072844861762e-05,
      "loss": 0.752195167541504,
      "step": 3930
    },
    {
      "epoch": 2.1205759655497243,
      "grad_norm": 13.590848922729492,
      "learning_rate": 4.3460452497001546e-05,
      "loss": 0.9873428344726562,
      "step": 3940
    },
    {
      "epoch": 2.1259588211546223,
      "grad_norm": 21.298370361328125,
      "learning_rate": 4.2972166095336996e-05,
      "loss": 0.7805982112884522,
      "step": 3950
    },
    {
      "epoch": 2.131341676759521,
      "grad_norm": 26.588743209838867,
      "learning_rate": 4.2485886425759824e-05,
      "loss": 0.8509333610534668,
      "step": 3960
    },
    {
      "epoch": 2.1367245323644193,
      "grad_norm": 14.079202651977539,
      "learning_rate": 4.200163059979168e-05,
      "loss": 0.46779747009277345,
      "step": 3970
    },
    {
      "epoch": 2.1421073879693178,
      "grad_norm": 16.043773651123047,
      "learning_rate": 4.1519415657738026e-05,
      "loss": 0.4787121295928955,
      "step": 3980
    },
    {
      "epoch": 2.1474902435742163,
      "grad_norm": 21.35724449157715,
      "learning_rate": 4.103925856808817e-05,
      "loss": 0.9199463844299316,
      "step": 3990
    },
    {
      "epoch": 2.1528730991791143,
      "grad_norm": 27.052135467529297,
      "learning_rate": 4.056117622691863e-05,
      "loss": 1.4920102119445802,
      "step": 4000
    },
    {
      "epoch": 2.1528730991791143,
      "eval_loss": 0.4048362374305725,
      "eval_runtime": 63.48,
      "eval_samples_per_second": 52.032,
      "eval_steps_per_second": 13.012,
      "step": 4000
    },
    {
      "epoch": 2.1582559547840128,
      "grad_norm": 1.1041878461837769,
      "learning_rate": 4.008518545729818e-05,
      "loss": 0.5191349029541016,
      "step": 4010
    },
    {
      "epoch": 2.1636388103889113,
      "grad_norm": 19.768545150756836,
      "learning_rate": 3.9611303008696073e-05,
      "loss": 0.9854854583740235,
      "step": 4020
    },
    {
      "epoch": 2.1690216659938097,
      "grad_norm": 55.555599212646484,
      "learning_rate": 3.913954555639263e-05,
      "loss": 0.7864606857299805,
      "step": 4030
    },
    {
      "epoch": 2.1744045215987082,
      "grad_norm": 38.11814880371094,
      "learning_rate": 3.866992970089238e-05,
      "loss": 1.094459342956543,
      "step": 4040
    },
    {
      "epoch": 2.1797873772036067,
      "grad_norm": 25.64844512939453,
      "learning_rate": 3.8202471967340004e-05,
      "loss": 0.7594769477844239,
      "step": 4050
    },
    {
      "epoch": 2.1851702328085048,
      "grad_norm": 39.07038879394531,
      "learning_rate": 3.773718880493883e-05,
      "loss": 0.767018985748291,
      "step": 4060
    },
    {
      "epoch": 2.1905530884134032,
      "grad_norm": 11.75182819366455,
      "learning_rate": 3.727409658637194e-05,
      "loss": 1.0294831275939942,
      "step": 4070
    },
    {
      "epoch": 2.1959359440183017,
      "grad_norm": 20.518577575683594,
      "learning_rate": 3.681321160722606e-05,
      "loss": 0.6886996746063232,
      "step": 4080
    },
    {
      "epoch": 2.2013187996232,
      "grad_norm": 38.67143630981445,
      "learning_rate": 3.6354550085418207e-05,
      "loss": 1.033548069000244,
      "step": 4090
    },
    {
      "epoch": 2.2067016552280987,
      "grad_norm": 37.95046615600586,
      "learning_rate": 3.589812816062489e-05,
      "loss": 1.114480209350586,
      "step": 4100
    },
    {
      "epoch": 2.2067016552280987,
      "eval_loss": 0.40899786353111267,
      "eval_runtime": 63.5021,
      "eval_samples_per_second": 52.014,
      "eval_steps_per_second": 13.007,
      "step": 4100
    },
    {
      "epoch": 2.2120845108329967,
      "grad_norm": 11.148638725280762,
      "learning_rate": 3.544396189371425e-05,
      "loss": 0.6971592426300048,
      "step": 4110
    },
    {
      "epoch": 2.217467366437895,
      "grad_norm": 22.88878631591797,
      "learning_rate": 3.4992067266181015e-05,
      "loss": 0.669145917892456,
      "step": 4120
    },
    {
      "epoch": 2.2228502220427937,
      "grad_norm": 6.007533550262451,
      "learning_rate": 3.454246017958376e-05,
      "loss": 0.627226448059082,
      "step": 4130
    },
    {
      "epoch": 2.228233077647692,
      "grad_norm": 6.372663497924805,
      "learning_rate": 3.409515645498583e-05,
      "loss": 0.4246776580810547,
      "step": 4140
    },
    {
      "epoch": 2.2336159332525907,
      "grad_norm": 37.81616973876953,
      "learning_rate": 3.3650171832398284e-05,
      "loss": 0.8943648338317871,
      "step": 4150
    },
    {
      "epoch": 2.2389987888574887,
      "grad_norm": 35.7336540222168,
      "learning_rate": 3.3207521970226104e-05,
      "loss": 1.005715560913086,
      "step": 4160
    },
    {
      "epoch": 2.244381644462387,
      "grad_norm": 25.88532066345215,
      "learning_rate": 3.2767222444717295e-05,
      "loss": 0.6161307334899903,
      "step": 4170
    },
    {
      "epoch": 2.2497645000672857,
      "grad_norm": 1.1592823266983032,
      "learning_rate": 3.232928874941463e-05,
      "loss": 0.35333144664764404,
      "step": 4180
    },
    {
      "epoch": 2.255147355672184,
      "grad_norm": 18.25936508178711,
      "learning_rate": 3.189373629461056e-05,
      "loss": 0.603509521484375,
      "step": 4190
    },
    {
      "epoch": 2.2605302112770826,
      "grad_norm": 18.44670295715332,
      "learning_rate": 3.1460580406804975e-05,
      "loss": 0.9603145599365235,
      "step": 4200
    },
    {
      "epoch": 2.2605302112770826,
      "eval_loss": 0.461373895406723,
      "eval_runtime": 63.6154,
      "eval_samples_per_second": 51.921,
      "eval_steps_per_second": 12.984,
      "step": 4200
    },
    {
      "epoch": 2.265913066881981,
      "grad_norm": 7.78346586227417,
      "learning_rate": 3.102983632816573e-05,
      "loss": 0.832798957824707,
      "step": 4210
    },
    {
      "epoch": 2.271295922486879,
      "grad_norm": 23.0885066986084,
      "learning_rate": 3.060151921599241e-05,
      "loss": 0.9533357620239258,
      "step": 4220
    },
    {
      "epoch": 2.2766787780917777,
      "grad_norm": 18.10384750366211,
      "learning_rate": 3.017564414218299e-05,
      "loss": 0.64031081199646,
      "step": 4230
    },
    {
      "epoch": 2.282061633696676,
      "grad_norm": 38.35786437988281,
      "learning_rate": 2.975222609270335e-05,
      "loss": 0.6714494705200196,
      "step": 4240
    },
    {
      "epoch": 2.2874444893015746,
      "grad_norm": 30.549135208129883,
      "learning_rate": 2.9331279967060132e-05,
      "loss": 0.6489608764648438,
      "step": 4250
    },
    {
      "epoch": 2.292827344906473,
      "grad_norm": 22.884374618530273,
      "learning_rate": 2.8912820577776234e-05,
      "loss": 0.6908737182617187,
      "step": 4260
    },
    {
      "epoch": 2.298210200511371,
      "grad_norm": 3.388777732849121,
      "learning_rate": 2.8496862649869706e-05,
      "loss": 0.52658109664917,
      "step": 4270
    },
    {
      "epoch": 2.3035930561162696,
      "grad_norm": 49.91218948364258,
      "learning_rate": 2.808342082033556e-05,
      "loss": 1.008077049255371,
      "step": 4280
    },
    {
      "epoch": 2.308975911721168,
      "grad_norm": 10.552830696105957,
      "learning_rate": 2.767250963763075e-05,
      "loss": 0.8219211578369141,
      "step": 4290
    },
    {
      "epoch": 2.3143587673260666,
      "grad_norm": 52.01408386230469,
      "learning_rate": 2.7264143561162136e-05,
      "loss": 1.2357698440551759,
      "step": 4300
    },
    {
      "epoch": 2.3143587673260666,
      "eval_loss": 0.4524453580379486,
      "eval_runtime": 63.5057,
      "eval_samples_per_second": 52.011,
      "eval_steps_per_second": 13.007,
      "step": 4300
    },
    {
      "epoch": 2.319741622930965,
      "grad_norm": 8.435943603515625,
      "learning_rate": 2.6858336960777863e-05,
      "loss": 0.7867340087890625,
      "step": 4310
    },
    {
      "epoch": 2.325124478535863,
      "grad_norm": 25.12037467956543,
      "learning_rate": 2.6455104116261477e-05,
      "loss": 0.6435882091522217,
      "step": 4320
    },
    {
      "epoch": 2.3305073341407616,
      "grad_norm": 19.58714485168457,
      "learning_rate": 2.6054459216829573e-05,
      "loss": 0.7821506977081298,
      "step": 4330
    },
    {
      "epoch": 2.33589018974566,
      "grad_norm": 17.670440673828125,
      "learning_rate": 2.5656416360632494e-05,
      "loss": 0.7270960807800293,
      "step": 4340
    },
    {
      "epoch": 2.3412730453505586,
      "grad_norm": 19.895416259765625,
      "learning_rate": 2.5260989554258185e-05,
      "loss": 1.1800201416015625,
      "step": 4350
    },
    {
      "epoch": 2.346655900955457,
      "grad_norm": 27.924663543701172,
      "learning_rate": 2.4868192712239348e-05,
      "loss": 0.891145896911621,
      "step": 4360
    },
    {
      "epoch": 2.352038756560355,
      "grad_norm": 24.85103988647461,
      "learning_rate": 2.44780396565639e-05,
      "loss": 0.9543925285339355,
      "step": 4370
    },
    {
      "epoch": 2.3574216121652536,
      "grad_norm": 23.492963790893555,
      "learning_rate": 2.409054411618831e-05,
      "loss": 1.0963412284851075,
      "step": 4380
    },
    {
      "epoch": 2.362804467770152,
      "grad_norm": 23.491336822509766,
      "learning_rate": 2.370571972655489e-05,
      "loss": 0.7562737464904785,
      "step": 4390
    },
    {
      "epoch": 2.3681873233750506,
      "grad_norm": 9.930757522583008,
      "learning_rate": 2.332358002911167e-05,
      "loss": 0.5512786388397217,
      "step": 4400
    },
    {
      "epoch": 2.3681873233750506,
      "eval_loss": 0.43738052248954773,
      "eval_runtime": 63.71,
      "eval_samples_per_second": 51.844,
      "eval_steps_per_second": 12.965,
      "step": 4400
    },
    {
      "epoch": 2.373570178979949,
      "grad_norm": 24.945201873779297,
      "learning_rate": 2.294413847083604e-05,
      "loss": 0.719010305404663,
      "step": 4410
    },
    {
      "epoch": 2.378953034584847,
      "grad_norm": 18.582550048828125,
      "learning_rate": 2.2567408403761526e-05,
      "loss": 0.9431358337402344,
      "step": 4420
    },
    {
      "epoch": 2.3843358901897456,
      "grad_norm": 3.468303680419922,
      "learning_rate": 2.2193403084507947e-05,
      "loss": 0.6178753852844239,
      "step": 4430
    },
    {
      "epoch": 2.389718745794644,
      "grad_norm": 22.029281616210938,
      "learning_rate": 2.1822135673814948e-05,
      "loss": 1.0714486122131348,
      "step": 4440
    },
    {
      "epoch": 2.3951016013995425,
      "grad_norm": 19.43686294555664,
      "learning_rate": 2.1453619236078937e-05,
      "loss": 0.831718635559082,
      "step": 4450
    },
    {
      "epoch": 2.400484457004441,
      "grad_norm": 20.41507911682129,
      "learning_rate": 2.1087866738893246e-05,
      "loss": 0.7316448211669921,
      "step": 4460
    },
    {
      "epoch": 2.405867312609339,
      "grad_norm": 24.05844497680664,
      "learning_rate": 2.072489105259192e-05,
      "loss": 0.5118863582611084,
      "step": 4470
    },
    {
      "epoch": 2.4112501682142375,
      "grad_norm": 4.802274703979492,
      "learning_rate": 2.036470494979681e-05,
      "loss": 0.4543926239013672,
      "step": 4480
    },
    {
      "epoch": 2.416633023819136,
      "grad_norm": 6.404051780700684,
      "learning_rate": 2.000732110496808e-05,
      "loss": 0.6312252521514893,
      "step": 4490
    },
    {
      "epoch": 2.4220158794240345,
      "grad_norm": 23.455795288085938,
      "learning_rate": 1.9652752093958225e-05,
      "loss": 0.5971816062927247,
      "step": 4500
    },
    {
      "epoch": 2.4220158794240345,
      "eval_loss": 0.4342218041419983,
      "eval_runtime": 63.961,
      "eval_samples_per_second": 51.641,
      "eval_steps_per_second": 12.914,
      "step": 4500
    },
    {
      "epoch": 2.427398735028933,
      "grad_norm": 35.069087982177734,
      "learning_rate": 1.9301010393569676e-05,
      "loss": 0.763551664352417,
      "step": 4510
    },
    {
      "epoch": 2.432781590633831,
      "grad_norm": 22.598342895507812,
      "learning_rate": 1.895210838111544e-05,
      "loss": 1.073389434814453,
      "step": 4520
    },
    {
      "epoch": 2.4381644462387295,
      "grad_norm": 15.471684455871582,
      "learning_rate": 1.8606058333983946e-05,
      "loss": 0.8500160217285156,
      "step": 4530
    },
    {
      "epoch": 2.443547301843628,
      "grad_norm": 38.8491325378418,
      "learning_rate": 1.8262872429206734e-05,
      "loss": 1.3663853645324706,
      "step": 4540
    },
    {
      "epoch": 2.4489301574485265,
      "grad_norm": 22.642866134643555,
      "learning_rate": 1.7922562743030057e-05,
      "loss": 0.6329252243041992,
      "step": 4550
    },
    {
      "epoch": 2.454313013053425,
      "grad_norm": 16.064590454101562,
      "learning_rate": 1.7585141250490022e-05,
      "loss": 0.8116084098815918,
      "step": 4560
    },
    {
      "epoch": 2.459695868658323,
      "grad_norm": 26.74970245361328,
      "learning_rate": 1.725061982499103e-05,
      "loss": 0.8728569984436035,
      "step": 4570
    },
    {
      "epoch": 2.4650787242632215,
      "grad_norm": 13.955748558044434,
      "learning_rate": 1.69190102378881e-05,
      "loss": 0.7079121589660644,
      "step": 4580
    },
    {
      "epoch": 2.47046157986812,
      "grad_norm": 30.616897583007812,
      "learning_rate": 1.6590324158072603e-05,
      "loss": 0.8990015029907227,
      "step": 4590
    },
    {
      "epoch": 2.4758444354730185,
      "grad_norm": 20.03752899169922,
      "learning_rate": 1.6264573151561648e-05,
      "loss": 1.1187437057495118,
      "step": 4600
    },
    {
      "epoch": 2.4758444354730185,
      "eval_loss": 0.39977964758872986,
      "eval_runtime": 63.806,
      "eval_samples_per_second": 51.766,
      "eval_steps_per_second": 12.945,
      "step": 4600
    },
    {
      "epoch": 2.481227291077917,
      "grad_norm": 18.699827194213867,
      "learning_rate": 1.5941768681091085e-05,
      "loss": 0.7999272346496582,
      "step": 4610
    },
    {
      "epoch": 2.486610146682815,
      "grad_norm": 4.604935646057129,
      "learning_rate": 1.562192210571225e-05,
      "loss": 0.8284183502197265,
      "step": 4620
    },
    {
      "epoch": 2.4919930022877135,
      "grad_norm": 29.9500732421875,
      "learning_rate": 1.5305044680392e-05,
      "loss": 0.8354525566101074,
      "step": 4630
    },
    {
      "epoch": 2.497375857892612,
      "grad_norm": 37.695804595947266,
      "learning_rate": 1.499114755561699e-05,
      "loss": 0.8016999244689942,
      "step": 4640
    },
    {
      "epoch": 2.5027587134975104,
      "grad_norm": 4.899083614349365,
      "learning_rate": 1.4680241777001059e-05,
      "loss": 0.9540206909179687,
      "step": 4650
    },
    {
      "epoch": 2.508141569102409,
      "grad_norm": 26.045948028564453,
      "learning_rate": 1.4372338284896647e-05,
      "loss": 0.5969816684722901,
      "step": 4660
    },
    {
      "epoch": 2.513524424707307,
      "grad_norm": 23.402122497558594,
      "learning_rate": 1.4067447914009823e-05,
      "loss": 0.7931214809417725,
      "step": 4670
    },
    {
      "epoch": 2.518907280312206,
      "grad_norm": 5.232571601867676,
      "learning_rate": 1.3765581393018978e-05,
      "loss": 0.6403655052185059,
      "step": 4680
    },
    {
      "epoch": 2.524290135917104,
      "grad_norm": 11.179905891418457,
      "learning_rate": 1.3466749344197339e-05,
      "loss": 0.5316050529479981,
      "step": 4690
    },
    {
      "epoch": 2.5296729915220024,
      "grad_norm": 21.05845069885254,
      "learning_rate": 1.3170962283039235e-05,
      "loss": 1.0778657913208007,
      "step": 4700
    },
    {
      "epoch": 2.5296729915220024,
      "eval_loss": 0.40597495436668396,
      "eval_runtime": 63.8788,
      "eval_samples_per_second": 51.707,
      "eval_steps_per_second": 12.931,
      "step": 4700
    },
    {
      "epoch": 2.535055847126901,
      "grad_norm": 17.45840072631836,
      "learning_rate": 1.2878230617889918e-05,
      "loss": 1.3090163230895997,
      "step": 4710
    },
    {
      "epoch": 2.540438702731799,
      "grad_norm": 13.649649620056152,
      "learning_rate": 1.2588564649579449e-05,
      "loss": 0.6182886600494385,
      "step": 4720
    },
    {
      "epoch": 2.545821558336698,
      "grad_norm": 30.202678680419922,
      "learning_rate": 1.2301974571060181e-05,
      "loss": 0.8260242462158203,
      "step": 4730
    },
    {
      "epoch": 2.551204413941596,
      "grad_norm": 9.608967781066895,
      "learning_rate": 1.2018470467048072e-05,
      "loss": 0.5750818729400635,
      "step": 4740
    },
    {
      "epoch": 2.5565872695464944,
      "grad_norm": 12.897671699523926,
      "learning_rate": 1.1738062313667786e-05,
      "loss": 0.8292213439941406,
      "step": 4750
    },
    {
      "epoch": 2.561970125151393,
      "grad_norm": 48.488006591796875,
      "learning_rate": 1.1460759978101787e-05,
      "loss": 0.8127927780151367,
      "step": 4760
    },
    {
      "epoch": 2.5673529807562914,
      "grad_norm": 30.89815330505371,
      "learning_rate": 1.118657321824289e-05,
      "loss": 0.710633659362793,
      "step": 4770
    },
    {
      "epoch": 2.57273583636119,
      "grad_norm": 17.691478729248047,
      "learning_rate": 1.0915511682351142e-05,
      "loss": 0.6440166473388672,
      "step": 4780
    },
    {
      "epoch": 2.578118691966088,
      "grad_norm": 27.643741607666016,
      "learning_rate": 1.0647584908714148e-05,
      "loss": 0.679203987121582,
      "step": 4790
    },
    {
      "epoch": 2.5835015475709864,
      "grad_norm": 27.048784255981445,
      "learning_rate": 1.0382802325311491e-05,
      "loss": 0.9150435447692871,
      "step": 4800
    },
    {
      "epoch": 2.5835015475709864,
      "eval_loss": 0.40863797068595886,
      "eval_runtime": 63.903,
      "eval_samples_per_second": 51.688,
      "eval_steps_per_second": 12.926,
      "step": 4800
    },
    {
      "epoch": 2.588884403175885,
      "grad_norm": 17.041791915893555,
      "learning_rate": 1.0121173249482962e-05,
      "loss": 0.8509119987487793,
      "step": 4810
    },
    {
      "epoch": 2.5942672587807833,
      "grad_norm": 28.006967544555664,
      "learning_rate": 9.862706887600714e-06,
      "loss": 0.8236183166503906,
      "step": 4820
    },
    {
      "epoch": 2.599650114385682,
      "grad_norm": 36.17509078979492,
      "learning_rate": 9.607412334745235e-06,
      "loss": 0.839243221282959,
      "step": 4830
    },
    {
      "epoch": 2.60503296999058,
      "grad_norm": 1.2139636278152466,
      "learning_rate": 9.355298574385474e-06,
      "loss": 0.49801297187805177,
      "step": 4840
    },
    {
      "epoch": 2.6104158255954784,
      "grad_norm": 24.570405960083008,
      "learning_rate": 9.106374478062485e-06,
      "loss": 0.7030567169189453,
      "step": 4850
    },
    {
      "epoch": 2.615798681200377,
      "grad_norm": 25.26799201965332,
      "learning_rate": 8.86064880507742e-06,
      "loss": 0.8722931861877441,
      "step": 4860
    },
    {
      "epoch": 2.6211815368052753,
      "grad_norm": 18.12640380859375,
      "learning_rate": 8.618130202183317e-06,
      "loss": 0.8933378219604492,
      "step": 4870
    },
    {
      "epoch": 2.626564392410174,
      "grad_norm": 33.00815200805664,
      "learning_rate": 8.378827203280625e-06,
      "loss": 0.7595347881317138,
      "step": 4880
    },
    {
      "epoch": 2.631947248015072,
      "grad_norm": 1.4025450944900513,
      "learning_rate": 8.142748229117192e-06,
      "loss": 1.158553695678711,
      "step": 4890
    },
    {
      "epoch": 2.6373301036199703,
      "grad_norm": 2.04663348197937,
      "learning_rate": 7.909901586991752e-06,
      "loss": 0.495909595489502,
      "step": 4900
    },
    {
      "epoch": 2.6373301036199703,
      "eval_loss": 0.4060768187046051,
      "eval_runtime": 63.9323,
      "eval_samples_per_second": 51.664,
      "eval_steps_per_second": 12.92,
      "step": 4900
    },
    {
      "epoch": 2.642712959224869,
      "grad_norm": 34.9760856628418,
      "learning_rate": 7.680295470461651e-06,
      "loss": 0.8339058876037597,
      "step": 4910
    },
    {
      "epoch": 2.6480958148297673,
      "grad_norm": 7.547365665435791,
      "learning_rate": 7.4539379590545575e-06,
      "loss": 1.0124500274658204,
      "step": 4920
    },
    {
      "epoch": 2.653478670434666,
      "grad_norm": 36.673561096191406,
      "learning_rate": 7.230837017984127e-06,
      "loss": 0.9258150100708008,
      "step": 4930
    },
    {
      "epoch": 2.658861526039564,
      "grad_norm": 12.191123962402344,
      "learning_rate": 7.011000497869691e-06,
      "loss": 0.7174896240234375,
      "step": 4940
    },
    {
      "epoch": 2.6642443816444623,
      "grad_norm": 27.09784507751465,
      "learning_rate": 6.794436134460125e-06,
      "loss": 0.7737472534179688,
      "step": 4950
    },
    {
      "epoch": 2.669627237249361,
      "grad_norm": 1.9672365188598633,
      "learning_rate": 6.581151548361475e-06,
      "loss": 0.7733006954193116,
      "step": 4960
    },
    {
      "epoch": 2.6750100928542593,
      "grad_norm": 24.851816177368164,
      "learning_rate": 6.3711542447688956e-06,
      "loss": 0.976164436340332,
      "step": 4970
    },
    {
      "epoch": 2.6803929484591578,
      "grad_norm": 4.714712142944336,
      "learning_rate": 6.164451613202515e-06,
      "loss": 1.361710262298584,
      "step": 4980
    },
    {
      "epoch": 2.685775804064056,
      "grad_norm": 33.14804458618164,
      "learning_rate": 5.961050927247458e-06,
      "loss": 1.0757354736328124,
      "step": 4990
    },
    {
      "epoch": 2.6911586596689543,
      "grad_norm": 19.453475952148438,
      "learning_rate": 5.760959344297811e-06,
      "loss": 0.7794712543487549,
      "step": 5000
    },
    {
      "epoch": 2.6911586596689543,
      "eval_loss": 0.3953015208244324,
      "eval_runtime": 63.9175,
      "eval_samples_per_second": 51.676,
      "eval_steps_per_second": 12.923,
      "step": 5000
    },
    {
      "epoch": 2.6965415152738528,
      "grad_norm": 38.8685302734375,
      "learning_rate": 5.56418390530492e-06,
      "loss": 0.8262761116027832,
      "step": 5010
    },
    {
      "epoch": 2.7019243708787513,
      "grad_norm": 42.21137237548828,
      "learning_rate": 5.370731534529383e-06,
      "loss": 0.9226728439331054,
      "step": 5020
    },
    {
      "epoch": 2.7073072264836497,
      "grad_norm": 11.871996879577637,
      "learning_rate": 5.180609039297646e-06,
      "loss": 0.4317474365234375,
      "step": 5030
    },
    {
      "epoch": 2.712690082088548,
      "grad_norm": 1.4888293743133545,
      "learning_rate": 4.993823109762319e-06,
      "loss": 1.047317886352539,
      "step": 5040
    },
    {
      "epoch": 2.7180729376934463,
      "grad_norm": 31.726903915405273,
      "learning_rate": 4.810380318666774e-06,
      "loss": 0.9913960456848144,
      "step": 5050
    },
    {
      "epoch": 2.7234557932983448,
      "grad_norm": 30.20676040649414,
      "learning_rate": 4.630287121113897e-06,
      "loss": 0.6376973628997803,
      "step": 5060
    },
    {
      "epoch": 2.7288386489032432,
      "grad_norm": 29.86091423034668,
      "learning_rate": 4.453549854338934e-06,
      "loss": 0.8093062400817871,
      "step": 5070
    },
    {
      "epoch": 2.7342215045081417,
      "grad_norm": 3.3067686557769775,
      "learning_rate": 4.280174737486442e-06,
      "loss": 1.0463817596435547,
      "step": 5080
    },
    {
      "epoch": 2.7396043601130398,
      "grad_norm": 0.369608610868454,
      "learning_rate": 4.110167871391546e-06,
      "loss": 0.7146965980529785,
      "step": 5090
    },
    {
      "epoch": 2.7449872157179382,
      "grad_norm": 17.710023880004883,
      "learning_rate": 3.943535238365137e-06,
      "loss": 0.4951002597808838,
      "step": 5100
    },
    {
      "epoch": 2.7449872157179382,
      "eval_loss": 0.3943323791027069,
      "eval_runtime": 63.9034,
      "eval_samples_per_second": 51.687,
      "eval_steps_per_second": 12.926,
      "step": 5100
    },
    {
      "epoch": 2.7503700713228367,
      "grad_norm": 13.03876781463623,
      "learning_rate": 3.780282701983462e-06,
      "loss": 0.8074542045593261,
      "step": 5110
    },
    {
      "epoch": 2.755752926927735,
      "grad_norm": 36.00669860839844,
      "learning_rate": 3.620416006881722e-06,
      "loss": 1.496070957183838,
      "step": 5120
    },
    {
      "epoch": 2.7611357825326337,
      "grad_norm": 20.361572265625,
      "learning_rate": 3.4639407785519774e-06,
      "loss": 1.3481143951416015,
      "step": 5130
    },
    {
      "epoch": 2.7665186381375317,
      "grad_norm": 20.532785415649414,
      "learning_rate": 3.310862523145197e-06,
      "loss": 0.5495355129241943,
      "step": 5140
    },
    {
      "epoch": 2.7719014937424302,
      "grad_norm": 17.233598709106445,
      "learning_rate": 3.161186627277435e-06,
      "loss": 0.9931418418884277,
      "step": 5150
    },
    {
      "epoch": 2.7772843493473287,
      "grad_norm": 10.674700736999512,
      "learning_rate": 3.014918357840368e-06,
      "loss": 0.6294282913208008,
      "step": 5160
    },
    {
      "epoch": 2.782667204952227,
      "grad_norm": 29.000959396362305,
      "learning_rate": 2.872062861815905e-06,
      "loss": 0.7137054920196533,
      "step": 5170
    },
    {
      "epoch": 2.7880500605571257,
      "grad_norm": 27.14993667602539,
      "learning_rate": 2.732625166095093e-06,
      "loss": 0.7461415767669678,
      "step": 5180
    },
    {
      "epoch": 2.7934329161620237,
      "grad_norm": 19.03074073791504,
      "learning_rate": 2.5966101773012064e-06,
      "loss": 0.9058383941650391,
      "step": 5190
    },
    {
      "epoch": 2.798815771766922,
      "grad_norm": 25.3709659576416,
      "learning_rate": 2.464022681617162e-06,
      "loss": 0.9319804191589356,
      "step": 5200
    },
    {
      "epoch": 2.798815771766922,
      "eval_loss": 0.39116066694259644,
      "eval_runtime": 63.7425,
      "eval_samples_per_second": 51.818,
      "eval_steps_per_second": 12.958,
      "step": 5200
    },
    {
      "epoch": 2.8041986273718207,
      "grad_norm": 24.770858764648438,
      "learning_rate": 2.334867344616998e-06,
      "loss": 0.8134791374206543,
      "step": 5210
    },
    {
      "epoch": 2.809581482976719,
      "grad_norm": 32.360477447509766,
      "learning_rate": 2.2091487111017383e-06,
      "loss": 0.9853259086608886,
      "step": 5220
    },
    {
      "epoch": 2.8149643385816177,
      "grad_norm": 26.975568771362305,
      "learning_rate": 2.086871204939511e-06,
      "loss": 0.3680762767791748,
      "step": 5230
    },
    {
      "epoch": 2.8203471941865157,
      "grad_norm": 40.22636413574219,
      "learning_rate": 1.9680391289098266e-06,
      "loss": 0.7327807426452637,
      "step": 5240
    },
    {
      "epoch": 2.8257300497914146,
      "grad_norm": 27.841876983642578,
      "learning_rate": 1.852656664552166e-06,
      "loss": 1.114755630493164,
      "step": 5250
    },
    {
      "epoch": 2.8311129053963127,
      "grad_norm": 23.36740493774414,
      "learning_rate": 1.740727872018899e-06,
      "loss": 1.0058785438537599,
      "step": 5260
    },
    {
      "epoch": 2.836495761001211,
      "grad_norm": 3.433856725692749,
      "learning_rate": 1.632256689932321e-06,
      "loss": 0.6553703784942627,
      "step": 5270
    },
    {
      "epoch": 2.8418786166061096,
      "grad_norm": 28.19014549255371,
      "learning_rate": 1.5272469352461517e-06,
      "loss": 0.5715783119201661,
      "step": 5280
    },
    {
      "epoch": 2.8472614722110077,
      "grad_norm": 33.08034133911133,
      "learning_rate": 1.4257023031111538e-06,
      "loss": 1.0896245002746583,
      "step": 5290
    },
    {
      "epoch": 2.8526443278159066,
      "grad_norm": 29.616188049316406,
      "learning_rate": 1.3276263667451272e-06,
      "loss": 0.575172758102417,
      "step": 5300
    },
    {
      "epoch": 2.8526443278159066,
      "eval_loss": 0.39213913679122925,
      "eval_runtime": 63.7458,
      "eval_samples_per_second": 51.815,
      "eval_steps_per_second": 12.958,
      "step": 5300
    },
    {
      "epoch": 2.8580271834208046,
      "grad_norm": 24.045257568359375,
      "learning_rate": 1.2330225773071858e-06,
      "loss": 0.6139691352844239,
      "step": 5310
    },
    {
      "epoch": 2.863410039025703,
      "grad_norm": 32.76106643676758,
      "learning_rate": 1.1418942637763109e-06,
      "loss": 0.6473401069641114,
      "step": 5320
    },
    {
      "epoch": 2.8687928946306016,
      "grad_norm": 17.230541229248047,
      "learning_rate": 1.054244632834167e-06,
      "loss": 0.6802839279174805,
      "step": 5330
    },
    {
      "epoch": 2.8741757502355,
      "grad_norm": 14.499601364135742,
      "learning_rate": 9.700767687523483e-07,
      "loss": 0.8065154075622558,
      "step": 5340
    },
    {
      "epoch": 2.8795586058403986,
      "grad_norm": 16.937408447265625,
      "learning_rate": 8.893936332837749e-07,
      "loss": 1.2557835578918457,
      "step": 5350
    },
    {
      "epoch": 2.8849414614452966,
      "grad_norm": 18.08892250061035,
      "learning_rate": 8.121980655584783e-07,
      "loss": 0.5564064979553223,
      "step": 5360
    },
    {
      "epoch": 2.890324317050195,
      "grad_norm": 5.15496826171875,
      "learning_rate": 7.384927819837461e-07,
      "loss": 0.7302708625793457,
      "step": 5370
    },
    {
      "epoch": 2.8957071726550936,
      "grad_norm": 0.6171675324440002,
      "learning_rate": 6.68280376148489e-07,
      "loss": 0.46964249610900877,
      "step": 5380
    },
    {
      "epoch": 2.901090028259992,
      "grad_norm": 12.738670349121094,
      "learning_rate": 6.01563318731968e-07,
      "loss": 0.8759981155395508,
      "step": 5390
    },
    {
      "epoch": 2.9064728838648906,
      "grad_norm": 8.0991849899292,
      "learning_rate": 5.383439574169313e-07,
      "loss": 0.3772397994995117,
      "step": 5400
    },
    {
      "epoch": 2.9064728838648906,
      "eval_loss": 0.39079996943473816,
      "eval_runtime": 63.7806,
      "eval_samples_per_second": 51.787,
      "eval_steps_per_second": 12.951,
      "step": 5400
    },
    {
      "epoch": 2.9118557394697886,
      "grad_norm": 2.8664917945861816,
      "learning_rate": 4.786245168068804e-07,
      "loss": 0.6327933311462403,
      "step": 5410
    },
    {
      "epoch": 2.917238595074687,
      "grad_norm": 15.887563705444336,
      "learning_rate": 4.2240709834791004e-07,
      "loss": 0.8755616188049317,
      "step": 5420
    },
    {
      "epoch": 2.9226214506795856,
      "grad_norm": 10.990631103515625,
      "learning_rate": 3.696936802546902e-07,
      "loss": 0.9237110137939453,
      "step": 5430
    },
    {
      "epoch": 2.928004306284484,
      "grad_norm": 36.54423141479492,
      "learning_rate": 3.204861174408547e-07,
      "loss": 0.6388278484344483,
      "step": 5440
    },
    {
      "epoch": 2.9333871618893825,
      "grad_norm": 25.8781795501709,
      "learning_rate": 2.747861414537756e-07,
      "loss": 0.8046004295349121,
      "step": 5450
    },
    {
      "epoch": 2.9387700174942806,
      "grad_norm": 14.25063419342041,
      "learning_rate": 2.3259536041357888e-07,
      "loss": 1.015383243560791,
      "step": 5460
    },
    {
      "epoch": 2.944152873099179,
      "grad_norm": 49.4938850402832,
      "learning_rate": 1.9391525895661178e-07,
      "loss": 0.9852633476257324,
      "step": 5470
    },
    {
      "epoch": 2.9495357287040775,
      "grad_norm": 8.794760704040527,
      "learning_rate": 1.587471981831512e-07,
      "loss": 0.7090904235839843,
      "step": 5480
    },
    {
      "epoch": 2.954918584308976,
      "grad_norm": 20.00497055053711,
      "learning_rate": 1.27092415609531e-07,
      "loss": 0.8313823699951172,
      "step": 5490
    },
    {
      "epoch": 2.9603014399138745,
      "grad_norm": 33.549407958984375,
      "learning_rate": 9.89520251245879e-08,
      "loss": 0.7701246261596679,
      "step": 5500
    },
    {
      "epoch": 2.9603014399138745,
      "eval_loss": 0.3912367820739746,
      "eval_runtime": 63.5041,
      "eval_samples_per_second": 52.012,
      "eval_steps_per_second": 13.007,
      "step": 5500
    }
  ],
  "logging_steps": 10,
  "max_steps": 5574,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.246694746194002e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}