{
  "best_global_step": 2500,
  "best_metric": 1.2582728862762451,
  "best_model_checkpoint": "saves/qwen3.5-9b/lora/muspsy-fixed/checkpoint-2500",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 3831,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.007832770352964214,
      "grad_norm": 0.8252274990081787,
      "learning_rate": 4.6875000000000004e-06,
      "loss": 2.198759651184082,
      "step": 10
    },
    {
      "epoch": 0.01566554070592843,
      "grad_norm": 0.636273980140686,
      "learning_rate": 9.895833333333333e-06,
      "loss": 2.1897178649902345,
      "step": 20
    },
    {
      "epoch": 0.02349831105889264,
      "grad_norm": 0.5016177892684937,
      "learning_rate": 1.5104166666666667e-05,
      "loss": 2.103467559814453,
      "step": 30
    },
    {
      "epoch": 0.03133108141185686,
      "grad_norm": 0.2937227189540863,
      "learning_rate": 2.0312500000000002e-05,
      "loss": 1.9080320358276368,
      "step": 40
    },
    {
      "epoch": 0.03916385176482107,
      "grad_norm": 0.26669180393218994,
      "learning_rate": 2.552083333333333e-05,
      "loss": 1.7693399429321288,
      "step": 50
    },
    {
      "epoch": 0.04699662211778528,
      "grad_norm": 0.2339705526828766,
      "learning_rate": 3.072916666666667e-05,
      "loss": 1.7049299240112306,
      "step": 60
    },
    {
      "epoch": 0.0548293924707495,
      "grad_norm": 0.29707223176956177,
      "learning_rate": 3.59375e-05,
      "loss": 1.5845885276794434,
      "step": 70
    },
    {
      "epoch": 0.06266216282371372,
      "grad_norm": 0.2643619477748871,
      "learning_rate": 4.1145833333333335e-05,
      "loss": 1.5671621322631837,
      "step": 80
    },
    {
      "epoch": 0.07049493317667793,
      "grad_norm": 0.2945081889629364,
      "learning_rate": 4.635416666666667e-05,
      "loss": 1.5295079231262207,
      "step": 90
    },
    {
      "epoch": 0.07832770352964213,
      "grad_norm": 0.3696126341819763,
      "learning_rate": 5.15625e-05,
      "loss": 1.4586568832397462,
      "step": 100
    },
    {
      "epoch": 0.08616047388260635,
      "grad_norm": 0.26788729429244995,
      "learning_rate": 5.677083333333334e-05,
      "loss": 1.4245469093322753,
      "step": 110
    },
    {
      "epoch": 0.09399324423557057,
      "grad_norm": 0.33440881967544556,
      "learning_rate": 6.197916666666667e-05,
      "loss": 1.473775577545166,
      "step": 120
    },
    {
      "epoch": 0.10182601458853478,
      "grad_norm": 0.29041001200675964,
      "learning_rate": 6.71875e-05,
      "loss": 1.4213193893432616,
      "step": 130
    },
    {
      "epoch": 0.109658784941499,
      "grad_norm": 0.28692570328712463,
      "learning_rate": 7.239583333333334e-05,
      "loss": 1.3905920028686523,
      "step": 140
    },
    {
      "epoch": 0.11749155529446322,
      "grad_norm": 0.3533639907836914,
      "learning_rate": 7.760416666666667e-05,
      "loss": 1.409282112121582,
      "step": 150
    },
    {
      "epoch": 0.12532432564742743,
      "grad_norm": 0.3724885582923889,
      "learning_rate": 8.28125e-05,
      "loss": 1.4083166122436523,
      "step": 160
    },
    {
      "epoch": 0.13315709600039163,
      "grad_norm": 0.3462872803211212,
      "learning_rate": 8.802083333333334e-05,
      "loss": 1.363219928741455,
      "step": 170
    },
    {
      "epoch": 0.14098986635335586,
      "grad_norm": 0.35332179069519043,
      "learning_rate": 9.322916666666667e-05,
      "loss": 1.35781831741333,
      "step": 180
    },
    {
      "epoch": 0.14882263670632007,
      "grad_norm": 0.309860497713089,
      "learning_rate": 9.84375e-05,
      "loss": 1.3855286598205567,
      "step": 190
    },
    {
      "epoch": 0.15665540705928427,
      "grad_norm": 0.35043853521347046,
      "learning_rate": 0.00010364583333333333,
      "loss": 1.3873944282531738,
      "step": 200
    },
    {
      "epoch": 0.1644881774122485,
      "grad_norm": 0.3299640119075775,
      "learning_rate": 0.00010885416666666669,
      "loss": 1.4211556434631347,
      "step": 210
    },
    {
      "epoch": 0.1723209477652127,
      "grad_norm": 0.32519766688346863,
      "learning_rate": 0.0001140625,
      "loss": 1.3977481842041015,
      "step": 220
    },
    {
      "epoch": 0.18015371811817693,
      "grad_norm": 0.2888562083244324,
      "learning_rate": 0.00011927083333333333,
      "loss": 1.3984809875488282,
      "step": 230
    },
    {
      "epoch": 0.18798648847114113,
      "grad_norm": 0.30679890513420105,
      "learning_rate": 0.00012447916666666667,
      "loss": 1.3943490028381347,
      "step": 240
    },
    {
      "epoch": 0.19581925882410536,
      "grad_norm": 0.2974647283554077,
      "learning_rate": 0.0001296875,
      "loss": 1.3833021163940429,
      "step": 250
    },
    {
      "epoch": 0.20365202917706957,
      "grad_norm": 0.31096336245536804,
      "learning_rate": 0.00013489583333333333,
      "loss": 1.3720091819763183,
      "step": 260
    },
    {
      "epoch": 0.21148479953003377,
      "grad_norm": 0.2655313014984131,
      "learning_rate": 0.00014010416666666669,
      "loss": 1.358876609802246,
      "step": 270
    },
    {
      "epoch": 0.219317569882998,
      "grad_norm": 0.305414080619812,
      "learning_rate": 0.00014531250000000002,
      "loss": 1.349748420715332,
      "step": 280
    },
    {
      "epoch": 0.2271503402359622,
      "grad_norm": 0.242662712931633,
      "learning_rate": 0.00015052083333333334,
      "loss": 1.392601203918457,
      "step": 290
    },
    {
      "epoch": 0.23498311058892643,
      "grad_norm": 0.28158220648765564,
      "learning_rate": 0.00015572916666666667,
      "loss": 1.3312394142150878,
      "step": 300
    },
    {
      "epoch": 0.24281588094189063,
      "grad_norm": 0.3027257025241852,
      "learning_rate": 0.0001609375,
      "loss": 1.3067551612854005,
      "step": 310
    },
    {
      "epoch": 0.25064865129485486,
      "grad_norm": 0.2552274465560913,
      "learning_rate": 0.00016614583333333333,
      "loss": 1.3163175582885742,
      "step": 320
    },
    {
      "epoch": 0.25848142164781907,
      "grad_norm": 0.28112491965293884,
      "learning_rate": 0.0001713541666666667,
      "loss": 1.3473979949951171,
      "step": 330
    },
    {
      "epoch": 0.26631419200078327,
      "grad_norm": 0.32117217779159546,
      "learning_rate": 0.00017656250000000002,
      "loss": 1.3740118026733399,
      "step": 340
    },
    {
      "epoch": 0.27414696235374747,
      "grad_norm": 0.2886170744895935,
      "learning_rate": 0.00018177083333333334,
      "loss": 1.3543949127197266,
      "step": 350
    },
    {
      "epoch": 0.2819797327067117,
      "grad_norm": 0.30684173107147217,
      "learning_rate": 0.00018697916666666667,
      "loss": 1.3309515953063964,
      "step": 360
    },
    {
      "epoch": 0.28981250305967593,
      "grad_norm": 0.2990993857383728,
      "learning_rate": 0.0001921875,
      "loss": 1.3382258415222168,
      "step": 370
    },
    {
      "epoch": 0.29764527341264013,
      "grad_norm": 0.256516695022583,
      "learning_rate": 0.00019739583333333333,
      "loss": 1.3612469673156737,
      "step": 380
    },
    {
      "epoch": 0.30547804376560433,
      "grad_norm": 0.2611815929412842,
      "learning_rate": 0.0001999989616914266,
      "loss": 1.3669749259948731,
      "step": 390
    },
    {
      "epoch": 0.31331081411856854,
      "grad_norm": 0.2872208058834076,
      "learning_rate": 0.00019999065535220928,
      "loss": 1.3830884933471679,
      "step": 400
    },
    {
      "epoch": 0.3211435844715328,
      "grad_norm": 0.2765055000782013,
      "learning_rate": 0.0001999740433637345,
      "loss": 1.3454028129577638,
      "step": 410
    },
    {
      "epoch": 0.328976354824497,
      "grad_norm": 0.2273096740245819,
      "learning_rate": 0.0001999491271058647,
      "loss": 1.317532730102539,
      "step": 420
    },
    {
      "epoch": 0.3368091251774612,
      "grad_norm": 0.29210448265075684,
      "learning_rate": 0.00019991590864825028,
      "loss": 1.3869080543518066,
      "step": 430
    },
    {
      "epoch": 0.3446418955304254,
      "grad_norm": 0.2541140615940094,
      "learning_rate": 0.00019987439075015763,
      "loss": 1.380301284790039,
      "step": 440
    },
    {
      "epoch": 0.3524746658833896,
      "grad_norm": 0.21736745536327362,
      "learning_rate": 0.00019982457686024005,
      "loss": 1.3478592872619628,
      "step": 450
    },
    {
      "epoch": 0.36030743623635386,
      "grad_norm": 0.2869614064693451,
      "learning_rate": 0.00019976647111625117,
      "loss": 1.3316679000854492,
      "step": 460
    },
    {
      "epoch": 0.36814020658931806,
      "grad_norm": 0.20401088893413544,
      "learning_rate": 0.00019970007834470125,
      "loss": 1.3329079627990723,
      "step": 470
    },
    {
      "epoch": 0.37597297694228227,
      "grad_norm": 0.23920388519763947,
      "learning_rate": 0.00019962540406045647,
      "loss": 1.292691707611084,
      "step": 480
    },
    {
      "epoch": 0.38380574729524647,
      "grad_norm": 0.2719043791294098,
      "learning_rate": 0.0001995424544662805,
      "loss": 1.3550784111022949,
      "step": 490
    },
    {
      "epoch": 0.3916385176482107,
      "grad_norm": 0.31126660108566284,
      "learning_rate": 0.0001994512364523196,
      "loss": 1.3583318710327148,
      "step": 500
    },
    {
      "epoch": 0.3916385176482107,
      "eval_loss": 1.338908314704895,
      "eval_runtime": 322.798,
      "eval_samples_per_second": 3.333,
      "eval_steps_per_second": 3.333,
      "step": 500
    },
    {
      "epoch": 0.39947128800117493,
      "grad_norm": 0.2459356188774109,
      "learning_rate": 0.0001993517575955301,
      "loss": 1.3324827194213866,
      "step": 510
    },
    {
      "epoch": 0.40730405835413913,
      "grad_norm": 0.23143388330936432,
      "learning_rate": 0.0001992440261590491,
      "loss": 1.3564526557922363,
      "step": 520
    },
    {
      "epoch": 0.41513682870710333,
      "grad_norm": 0.2303522378206253,
      "learning_rate": 0.00019912805109150808,
      "loss": 1.3334980964660645,
      "step": 530
    },
    {
      "epoch": 0.42296959906006754,
      "grad_norm": 0.257758766412735,
      "learning_rate": 0.00019900384202628955,
      "loss": 1.31835880279541,
      "step": 540
    },
    {
      "epoch": 0.4308023694130318,
      "grad_norm": 0.2381799966096878,
      "learning_rate": 0.00019887140928072695,
      "loss": 1.336206817626953,
      "step": 550
    },
    {
      "epoch": 0.438635139765996,
      "grad_norm": 0.19770245254039764,
      "learning_rate": 0.00019873076385524754,
      "loss": 1.3192202568054199,
      "step": 560
    },
    {
      "epoch": 0.4464679101189602,
      "grad_norm": 0.24335643649101257,
      "learning_rate": 0.00019858191743245883,
      "loss": 1.3167536735534668,
      "step": 570
    },
    {
      "epoch": 0.4543006804719244,
      "grad_norm": 0.24658577144145966,
      "learning_rate": 0.00019842488237617793,
      "loss": 1.356813621520996,
      "step": 580
    },
    {
      "epoch": 0.4621334508248886,
      "grad_norm": 0.2352878302335739,
      "learning_rate": 0.00019825967173040478,
      "loss": 1.325494384765625,
      "step": 590
    },
    {
      "epoch": 0.46996622117785286,
      "grad_norm": 0.2151331603527069,
      "learning_rate": 0.0001980862992182385,
      "loss": 1.3333147048950196,
      "step": 600
    },
    {
      "epoch": 0.47779899153081706,
      "grad_norm": 0.21890704333782196,
      "learning_rate": 0.00019790477924073755,
      "loss": 1.3124448776245117,
      "step": 610
    },
    {
      "epoch": 0.48563176188378127,
      "grad_norm": 0.22006437182426453,
      "learning_rate": 0.00019771512687572366,
      "loss": 1.3134421348571776,
      "step": 620
    },
    {
      "epoch": 0.49346453223674547,
      "grad_norm": 0.21201013028621674,
      "learning_rate": 0.00019751735787652908,
      "loss": 1.3247313499450684,
      "step": 630
    },
    {
      "epoch": 0.5012973025897097,
      "grad_norm": 0.2933153212070465,
      "learning_rate": 0.0001973114886706884,
      "loss": 1.3370563507080078,
      "step": 640
    },
    {
      "epoch": 0.5091300729426739,
      "grad_norm": 0.1902574896812439,
      "learning_rate": 0.00019709753635857368,
      "loss": 1.3317548751831054,
      "step": 650
    },
    {
      "epoch": 0.5169628432956381,
      "grad_norm": 0.19506672024726868,
      "learning_rate": 0.00019687551871197428,
      "loss": 1.3153909683227538,
      "step": 660
    },
    {
      "epoch": 0.5247956136486024,
      "grad_norm": 0.23361273109912872,
      "learning_rate": 0.00019664545417262056,
      "loss": 1.3147119522094726,
      "step": 670
    },
    {
      "epoch": 0.5326283840015665,
      "grad_norm": 0.2266688495874405,
      "learning_rate": 0.00019640736185065193,
      "loss": 1.3157793998718261,
      "step": 680
    },
    {
      "epoch": 0.5404611543545308,
      "grad_norm": 0.2243291437625885,
      "learning_rate": 0.00019616126152302975,
      "loss": 1.3172315597534179,
      "step": 690
    },
    {
      "epoch": 0.5482939247074949,
      "grad_norm": 0.24645571410655975,
      "learning_rate": 0.0001959071736318942,
      "loss": 1.3153032302856444,
      "step": 700
    },
    {
      "epoch": 0.5561266950604592,
      "grad_norm": 0.19903017580509186,
      "learning_rate": 0.00019564511928286664,
      "loss": 1.29284610748291,
      "step": 710
    },
    {
      "epoch": 0.5639594654134235,
      "grad_norm": 0.23710037767887115,
      "learning_rate": 0.00019537512024329621,
      "loss": 1.3206818580627442,
      "step": 720
    },
    {
      "epoch": 0.5717922357663876,
      "grad_norm": 0.2107548862695694,
      "learning_rate": 0.00019509719894045193,
      "loss": 1.3023270606994628,
      "step": 730
    },
    {
      "epoch": 0.5796250061193519,
      "grad_norm": 0.20591115951538086,
      "learning_rate": 0.0001948113784596596,
      "loss": 1.3201650619506835,
      "step": 740
    },
    {
      "epoch": 0.587457776472316,
      "grad_norm": 0.23854319751262665,
      "learning_rate": 0.00019451768254238446,
      "loss": 1.290597915649414,
      "step": 750
    },
    {
      "epoch": 0.5952905468252803,
      "grad_norm": 1.604860782623291,
      "learning_rate": 0.00019421613558425892,
      "loss": 1.332386589050293,
      "step": 760
    },
    {
      "epoch": 0.6031233171782445,
      "grad_norm": 0.2258477807044983,
      "learning_rate": 0.00019390676263305635,
      "loss": 1.322896671295166,
      "step": 770
    },
    {
      "epoch": 0.6109560875312087,
      "grad_norm": 0.223722904920578,
      "learning_rate": 0.00019358958938661035,
      "loss": 1.3295950889587402,
      "step": 780
    },
    {
      "epoch": 0.6187888578841729,
      "grad_norm": 0.22754625976085663,
      "learning_rate": 0.0001932646421906802,
      "loss": 1.3002716064453126,
      "step": 790
    },
    {
      "epoch": 0.6266216282371371,
      "grad_norm": 0.22892414033412933,
      "learning_rate": 0.00019293194803676254,
      "loss": 1.3059702873229981,
      "step": 800
    },
    {
      "epoch": 0.6344543985901013,
      "grad_norm": 0.23315167427062988,
      "learning_rate": 0.00019259153455984925,
      "loss": 1.3296985626220703,
      "step": 810
    },
    {
      "epoch": 0.6422871689430656,
      "grad_norm": 0.21056416630744934,
      "learning_rate": 0.00019224343003613212,
      "loss": 1.3305760383605958,
      "step": 820
    },
    {
      "epoch": 0.6501199392960297,
      "grad_norm": 0.2409346103668213,
      "learning_rate": 0.0001918876633806539,
      "loss": 1.325657844543457,
      "step": 830
    },
    {
      "epoch": 0.657952709648994,
      "grad_norm": 0.20294003188610077,
      "learning_rate": 0.00019152426414490665,
      "loss": 1.3022818565368652,
      "step": 840
    },
    {
      "epoch": 0.6657854800019581,
      "grad_norm": 0.24989593029022217,
      "learning_rate": 0.00019115326251437708,
      "loss": 1.3478684425354004,
      "step": 850
    },
    {
      "epoch": 0.6736182503549224,
      "grad_norm": 0.2169850468635559,
      "learning_rate": 0.00019077468930603903,
      "loss": 1.3089423179626465,
      "step": 860
    },
    {
      "epoch": 0.6814510207078867,
      "grad_norm": 0.22551028430461884,
      "learning_rate": 0.0001903885759657939,
      "loss": 1.2900199890136719,
      "step": 870
    },
    {
      "epoch": 0.6892837910608508,
      "grad_norm": 0.21751883625984192,
      "learning_rate": 0.00018999495456585854,
      "loss": 1.3203692436218262,
      "step": 880
    },
    {
      "epoch": 0.6971165614138151,
      "grad_norm": 0.22999724745750427,
      "learning_rate": 0.00018959385780210108,
      "loss": 1.3258060455322265,
      "step": 890
    },
    {
      "epoch": 0.7049493317667792,
      "grad_norm": 0.17664238810539246,
      "learning_rate": 0.00018918531899132527,
      "loss": 1.3001131057739257,
      "step": 900
    },
    {
      "epoch": 0.7127821021197435,
      "grad_norm": 0.24631066620349884,
      "learning_rate": 0.00018876937206850294,
      "loss": 1.318901824951172,
      "step": 910
    },
    {
      "epoch": 0.7206148724727077,
      "grad_norm": 0.2125585526227951,
      "learning_rate": 0.00018834605158395514,
      "loss": 1.2912470817565918,
      "step": 920
    },
    {
      "epoch": 0.7284476428256719,
      "grad_norm": 0.20649491250514984,
      "learning_rate": 0.00018791539270048252,
      "loss": 1.3023160934448241,
      "step": 930
    },
    {
      "epoch": 0.7362804131786361,
      "grad_norm": 0.19381368160247803,
      "learning_rate": 0.00018747743119044405,
      "loss": 1.2719347953796387,
      "step": 940
    },
    {
      "epoch": 0.7441131835316004,
      "grad_norm": 0.21463167667388916,
      "learning_rate": 0.00018703220343278628,
      "loss": 1.3201501846313477,
      "step": 950
    },
    {
      "epoch": 0.7519459538845645,
      "grad_norm": 0.2531944811344147,
      "learning_rate": 0.00018657974641002095,
      "loss": 1.2843191146850585,
      "step": 960
    },
    {
      "epoch": 0.7597787242375288,
      "grad_norm": 0.2112455517053604,
      "learning_rate": 0.0001861200977051535,
      "loss": 1.2866954803466797,
      "step": 970
    },
    {
      "epoch": 0.7676114945904929,
      "grad_norm": 0.23256352543830872,
      "learning_rate": 0.0001856532954985609,
      "loss": 1.3130477905273437,
      "step": 980
    },
    {
      "epoch": 0.7754442649434572,
      "grad_norm": 0.21679829061031342,
      "learning_rate": 0.00018517937856482055,
      "loss": 1.3176883697509765,
      "step": 990
    },
    {
      "epoch": 0.7832770352964215,
      "grad_norm": 0.1952638328075409,
      "learning_rate": 0.00018469838626948928,
      "loss": 1.2918599128723145,
      "step": 1000
    },
    {
      "epoch": 0.7832770352964215,
      "eval_loss": 1.3027687072753906,
      "eval_runtime": 331.2141,
      "eval_samples_per_second": 3.249,
      "eval_steps_per_second": 3.249,
      "step": 1000
    },
    {
      "epoch": 0.7911098056493856,
      "grad_norm": 0.20904706418514252,
      "learning_rate": 0.00018421035856583363,
      "loss": 1.3271790504455567,
      "step": 1010
    },
    {
      "epoch": 0.7989425760023499,
      "grad_norm": 0.22483959794044495,
      "learning_rate": 0.00018371533599151105,
      "loss": 1.277155590057373,
      "step": 1020
    },
    {
      "epoch": 0.806775346355314,
      "grad_norm": 0.20454975962638855,
      "learning_rate": 0.00018321335966520265,
      "loss": 1.297370147705078,
      "step": 1030
    },
    {
      "epoch": 0.8146081167082783,
      "grad_norm": 0.2571241855621338,
      "learning_rate": 0.00018270447128319788,
      "loss": 1.3238384246826171,
      "step": 1040
    },
    {
      "epoch": 0.8224408870612425,
      "grad_norm": 0.22101953625679016,
      "learning_rate": 0.00018218871311593085,
      "loss": 1.369068431854248,
      "step": 1050
    },
    {
      "epoch": 0.8302736574142067,
      "grad_norm": 0.22725796699523926,
      "learning_rate": 0.0001816661280044693,
      "loss": 1.3287213325500489,
      "step": 1060
    },
    {
      "epoch": 0.8381064277671709,
      "grad_norm": 0.2206043154001236,
      "learning_rate": 0.00018113675935695599,
      "loss": 1.3072173118591308,
      "step": 1070
    },
    {
      "epoch": 0.8459391981201351,
      "grad_norm": 0.19849057495594025,
      "learning_rate": 0.0001806006511450031,
      "loss": 1.302993392944336,
      "step": 1080
    },
    {
      "epoch": 0.8537719684730993,
      "grad_norm": 0.21485477685928345,
      "learning_rate": 0.00018005784790003956,
      "loss": 1.3502599716186523,
      "step": 1090
    },
    {
      "epoch": 0.8616047388260636,
      "grad_norm": 0.23242683708667755,
      "learning_rate": 0.00017950839470961245,
      "loss": 1.2745637893676758,
      "step": 1100
    },
    {
      "epoch": 0.8694375091790277,
      "grad_norm": 0.20469102263450623,
      "learning_rate": 0.0001789523372136414,
      "loss": 1.3007002830505372,
      "step": 1110
    },
    {
      "epoch": 0.877270279531992,
      "grad_norm": 0.18739371001720428,
      "learning_rate": 0.00017838972160062793,
      "loss": 1.314402961730957,
      "step": 1120
    },
    {
      "epoch": 0.8851030498849561,
      "grad_norm": 0.22443655133247375,
      "learning_rate": 0.0001778205946038185,
      "loss": 1.2732182502746583,
      "step": 1130
    },
    {
      "epoch": 0.8929358202379204,
      "grad_norm": 0.22353816032409668,
      "learning_rate": 0.00017724500349732293,
      "loss": 1.2690745353698731,
      "step": 1140
    },
    {
      "epoch": 0.9007685905908847,
      "grad_norm": 0.20211759209632874,
      "learning_rate": 0.00017666299609218745,
      "loss": 1.2555768013000488,
      "step": 1150
    },
    {
      "epoch": 0.9086013609438488,
      "grad_norm": 0.20591171085834503,
      "learning_rate": 0.00017607462073242336,
      "loss": 1.2454710960388184,
      "step": 1160
    },
    {
      "epoch": 0.9164341312968131,
      "grad_norm": 0.20958608388900757,
      "learning_rate": 0.00017547992629099137,
      "loss": 1.2596508026123048,
      "step": 1170
    },
    {
      "epoch": 0.9242669016497772,
      "grad_norm": 0.16288408637046814,
      "learning_rate": 0.000174878962165742,
      "loss": 1.2851850509643554,
      "step": 1180
    },
    {
      "epoch": 0.9320996720027415,
      "grad_norm": 0.22015182673931122,
      "learning_rate": 0.00017427177827531235,
      "loss": 1.261205291748047,
      "step": 1190
    },
    {
      "epoch": 0.9399324423557057,
      "grad_norm": 0.22604605555534363,
      "learning_rate": 0.00017365842505497975,
      "loss": 1.2355712890625,
      "step": 1200
    },
    {
      "epoch": 0.9477652127086699,
      "grad_norm": 0.21393424272537231,
      "learning_rate": 0.00017303895345247225,
      "loss": 1.34449462890625,
      "step": 1210
    },
    {
      "epoch": 0.9555979830616341,
      "grad_norm": 0.19904212653636932,
      "learning_rate": 0.00017241341492373678,
      "loss": 1.2960014343261719,
      "step": 1220
    },
    {
      "epoch": 0.9634307534145983,
      "grad_norm": 0.21874822676181793,
      "learning_rate": 0.000171781861428665,
      "loss": 1.3022172927856446,
      "step": 1230
    },
    {
      "epoch": 0.9712635237675625,
      "grad_norm": 0.21850082278251648,
      "learning_rate": 0.0001711443454267772,
      "loss": 1.2833776473999023,
      "step": 1240
    },
    {
      "epoch": 0.9790962941205268,
      "grad_norm": 0.20517125725746155,
      "learning_rate": 0.0001705009198728649,
      "loss": 1.312662696838379,
      "step": 1250
    },
    {
      "epoch": 0.9869290644734909,
      "grad_norm": 0.20981578528881073,
      "learning_rate": 0.00016985163821259217,
      "loss": 1.2979387283325194,
      "step": 1260
    },
    {
      "epoch": 0.9947618348264552,
      "grad_norm": 0.2150573581457138,
      "learning_rate": 0.00016919655437805617,
      "loss": 1.2593985557556153,
      "step": 1270
    },
    {
      "epoch": 1.0023498311058892,
      "grad_norm": 0.1821247935295105,
      "learning_rate": 0.00016853572278330737,
      "loss": 1.2760144233703614,
      "step": 1280
    },
    {
      "epoch": 1.0101826014588535,
      "grad_norm": 0.20195011794567108,
      "learning_rate": 0.00016786919831982957,
      "loss": 1.2467243194580078,
      "step": 1290
    },
    {
      "epoch": 1.0180153718118177,
      "grad_norm": 0.21706004440784454,
      "learning_rate": 0.00016719703635198054,
      "loss": 1.240755558013916,
      "step": 1300
    },
    {
      "epoch": 1.0258481421647818,
      "grad_norm": 0.22652487456798553,
      "learning_rate": 0.00016651929271239315,
      "loss": 1.2947601318359374,
      "step": 1310
    },
    {
      "epoch": 1.0336809125177462,
      "grad_norm": 0.21941937506198883,
      "learning_rate": 0.00016583602369733767,
      "loss": 1.2749066352844238,
      "step": 1320
    },
    {
      "epoch": 1.0415136828707103,
      "grad_norm": 0.20321761071681976,
      "learning_rate": 0.0001651472860620455,
      "loss": 1.2467523574829102,
      "step": 1330
    },
    {
      "epoch": 1.0493464532236745,
      "grad_norm": 0.19074523448944092,
      "learning_rate": 0.00016445313701599502,
      "loss": 1.2307304382324218,
      "step": 1340
    },
    {
      "epoch": 1.0571792235766388,
      "grad_norm": 0.2475343942642212,
      "learning_rate": 0.0001637536342181593,
      "loss": 1.189406967163086,
      "step": 1350
    },
    {
      "epoch": 1.065011993929603,
      "grad_norm": 0.23844459652900696,
      "learning_rate": 0.00016304883577221684,
      "loss": 1.2495388984680176,
      "step": 1360
    },
    {
      "epoch": 1.0728447642825671,
      "grad_norm": 0.23677049577236176,
      "learning_rate": 0.0001623388002217253,
      "loss": 1.2544000625610352,
      "step": 1370
    },
    {
      "epoch": 1.0806775346355315,
      "grad_norm": 0.24953368306159973,
      "learning_rate": 0.0001616235865452584,
      "loss": 1.2669272422790527,
      "step": 1380
    },
    {
      "epoch": 1.0885103049884957,
      "grad_norm": 0.21519717574119568,
      "learning_rate": 0.00016090325415150708,
      "loss": 1.2439716339111329,
      "step": 1390
    },
    {
      "epoch": 1.0963430753414598,
      "grad_norm": 0.22504104673862457,
      "learning_rate": 0.00016017786287434468,
      "loss": 1.2285042762756349,
      "step": 1400
    },
    {
      "epoch": 1.104175845694424,
      "grad_norm": 0.22723400592803955,
      "learning_rate": 0.00015944747296785705,
      "loss": 1.211160373687744,
      "step": 1410
    },
    {
      "epoch": 1.1120086160473883,
      "grad_norm": 0.2472873479127884,
      "learning_rate": 0.0001587121451013373,
      "loss": 1.2765045166015625,
      "step": 1420
    },
    {
      "epoch": 1.1198413864003525,
      "grad_norm": 0.24824367463588715,
      "learning_rate": 0.00015797194035424661,
      "loss": 1.222754669189453,
      "step": 1430
    },
    {
      "epoch": 1.1276741567533166,
      "grad_norm": 0.24168582260608673,
      "learning_rate": 0.00015722692021114052,
      "loss": 1.1961838722229003,
      "step": 1440
    },
    {
      "epoch": 1.135506927106281,
      "grad_norm": 0.2568736970424652,
      "learning_rate": 0.00015647714655656197,
      "loss": 1.2565688133239745,
      "step": 1450
    },
    {
      "epoch": 1.1433396974592451,
      "grad_norm": 0.2765266001224518,
      "learning_rate": 0.00015572268166990066,
      "loss": 1.2159022331237792,
      "step": 1460
    },
    {
      "epoch": 1.1511724678122093,
      "grad_norm": 0.22696787118911743,
      "learning_rate": 0.00015496358822022016,
      "loss": 1.225893497467041,
      "step": 1470
    },
    {
      "epoch": 1.1590052381651734,
      "grad_norm": 0.20794737339019775,
      "learning_rate": 0.00015419992926105192,
      "loss": 1.2634332656860352,
      "step": 1480
    },
    {
      "epoch": 1.1668380085181378,
      "grad_norm": 0.24258407950401306,
      "learning_rate": 0.0001534317682251583,
      "loss": 1.22896146774292,
      "step": 1490
    },
    {
      "epoch": 1.174670778871102,
      "grad_norm": 0.2493806779384613,
      "learning_rate": 0.00015265916891926313,
      "loss": 1.209363079071045,
      "step": 1500
    },
    {
      "epoch": 1.174670778871102,
      "eval_loss": 1.2873718738555908,
      "eval_runtime": 325.3065,
      "eval_samples_per_second": 3.308,
      "eval_steps_per_second": 3.308,
      "step": 1500
    },
    {
      "epoch": 1.1825035492240663,
      "grad_norm": 0.23888343572616577,
      "learning_rate": 0.0001518821955187519,
      "loss": 1.226218032836914,
      "step": 1510
    },
    {
      "epoch": 1.1903363195770305,
      "grad_norm": 0.25171080231666565,
      "learning_rate": 0.00015110091256234097,
      "loss": 1.244037914276123,
      "step": 1520
    },
    {
      "epoch": 1.1981690899299946,
      "grad_norm": 0.23836006224155426,
      "learning_rate": 0.00015031538494671685,
      "loss": 1.2562739372253418,
      "step": 1530
    },
    {
      "epoch": 1.2060018602829587,
      "grad_norm": 0.24001693725585938,
      "learning_rate": 0.00014952567792114534,
      "loss": 1.2565048217773438,
      "step": 1540
    },
    {
      "epoch": 1.2138346306359231,
      "grad_norm": 0.253283828496933,
      "learning_rate": 0.00014873185708205196,
      "loss": 1.2531654357910156,
      "step": 1550
    },
    {
      "epoch": 1.2216674009888873,
      "grad_norm": 0.24608942866325378,
      "learning_rate": 0.00014793398836757297,
      "loss": 1.2992531776428222,
      "step": 1560
    },
    {
      "epoch": 1.2295001713418514,
      "grad_norm": 0.28155362606048584,
      "learning_rate": 0.0001471321380520785,
      "loss": 1.2777346611022948,
      "step": 1570
    },
    {
      "epoch": 1.2373329416948158,
      "grad_norm": 0.20141169428825378,
      "learning_rate": 0.00014632637274066728,
      "loss": 1.2318084716796875,
      "step": 1580
    },
    {
      "epoch": 1.24516571204778,
      "grad_norm": 0.25557416677474976,
      "learning_rate": 0.00014551675936363433,
      "loss": 1.2323214530944824,
      "step": 1590
    },
    {
      "epoch": 1.252998482400744,
      "grad_norm": 0.21658560633659363,
      "learning_rate": 0.0001447033651709114,
      "loss": 1.2233482360839845,
      "step": 1600
    },
    {
      "epoch": 1.2608312527537082,
      "grad_norm": 0.2324075996875763,
      "learning_rate": 0.00014388625772648078,
      "loss": 1.2236601829528808,
      "step": 1610
    },
    {
      "epoch": 1.2686640231066726,
      "grad_norm": 0.23513862490653992,
      "learning_rate": 0.00014306550490276334,
      "loss": 1.2506739616394043,
      "step": 1620
    },
    {
      "epoch": 1.2764967934596367,
      "grad_norm": 0.2221163809299469,
      "learning_rate": 0.00014224117487498055,
      "loss": 1.2314607620239257,
      "step": 1630
    },
    {
      "epoch": 1.284329563812601,
      "grad_norm": 0.21652808785438538,
      "learning_rate": 0.00014141333611549187,
      "loss": 1.1632638931274415,
      "step": 1640
    },
    {
      "epoch": 1.2921623341655653,
      "grad_norm": 0.24344556033611298,
      "learning_rate": 0.00014058205738810668,
      "loss": 1.2199566841125489,
      "step": 1650
    },
    {
      "epoch": 1.2999951045185294,
      "grad_norm": 0.2773839831352234,
      "learning_rate": 0.00013974740774237295,
      "loss": 1.2052583694458008,
      "step": 1660
    },
    {
      "epoch": 1.3078278748714935,
      "grad_norm": 0.2699849307537079,
      "learning_rate": 0.00013890945650784137,
      "loss": 1.216574001312256,
      "step": 1670
    },
    {
      "epoch": 1.3156606452244577,
      "grad_norm": 0.2651710510253906,
      "learning_rate": 0.00013806827328830665,
      "loss": 1.257701587677002,
      "step": 1680
    },
    {
      "epoch": 1.323493415577422,
      "grad_norm": 0.21589776873588562,
      "learning_rate": 0.00013722392795602594,
      "loss": 1.210690402984619,
      "step": 1690
    },
    {
      "epoch": 1.3313261859303862,
      "grad_norm": 0.2312076985836029,
      "learning_rate": 0.00013637649064591488,
      "loss": 1.2231233596801758,
      "step": 1700
    },
    {
      "epoch": 1.3391589562833506,
      "grad_norm": 0.19364634156227112,
      "learning_rate": 0.00013552603174972204,
      "loss": 1.2308096885681152,
      "step": 1710
    },
    {
      "epoch": 1.3469917266363147,
      "grad_norm": 0.24895979464054108,
      "learning_rate": 0.00013467262191018176,
      "loss": 1.2260958671569824,
      "step": 1720
    },
    {
      "epoch": 1.3548244969892789,
      "grad_norm": 0.2678626775741577,
      "learning_rate": 0.00013381633201514615,
      "loss": 1.2428908348083496,
      "step": 1730
    },
    {
      "epoch": 1.362657267342243,
      "grad_norm": 0.30587542057037354,
      "learning_rate": 0.0001329572331916972,
      "loss": 1.220828914642334,
      "step": 1740
    },
    {
      "epoch": 1.3704900376952074,
      "grad_norm": 0.2133464366197586,
      "learning_rate": 0.00013209539680023827,
      "loss": 1.2177210807800294,
      "step": 1750
    },
    {
      "epoch": 1.3783228080481715,
      "grad_norm": 0.2186136543750763,
      "learning_rate": 0.00013123089442856686,
      "loss": 1.1930458068847656,
      "step": 1760
    },
    {
      "epoch": 1.3861555784011357,
      "grad_norm": 0.22868897020816803,
      "learning_rate": 0.00013036379788592813,
      "loss": 1.2323189735412599,
      "step": 1770
    },
    {
      "epoch": 1.3939883487541,
      "grad_norm": 0.2808184027671814,
      "learning_rate": 0.00012949417919705007,
      "loss": 1.238790512084961,
      "step": 1780
    },
    {
      "epoch": 1.4018211191070642,
      "grad_norm": 0.25862979888916016,
      "learning_rate": 0.00012862211059616087,
      "loss": 1.2346102714538574,
      "step": 1790
    },
    {
      "epoch": 1.4096538894600283,
      "grad_norm": 0.23017963767051697,
      "learning_rate": 0.0001277476645209889,
      "loss": 1.2439691543579101,
      "step": 1800
    },
    {
      "epoch": 1.4174866598129925,
      "grad_norm": 0.2433658391237259,
      "learning_rate": 0.00012687091360674558,
      "loss": 1.2119647026062013,
      "step": 1810
    },
    {
      "epoch": 1.4253194301659569,
      "grad_norm": 0.23380210995674133,
      "learning_rate": 0.0001259919306800921,
      "loss": 1.1757728576660156,
      "step": 1820
    },
    {
      "epoch": 1.433152200518921,
      "grad_norm": 0.255401074886322,
      "learning_rate": 0.0001251107887530901,
      "loss": 1.2686352729797363,
      "step": 1830
    },
    {
      "epoch": 1.4409849708718854,
      "grad_norm": 0.26713621616363525,
      "learning_rate": 0.00012422756101713702,
      "loss": 1.2261395454406738,
      "step": 1840
    },
    {
      "epoch": 1.4488177412248495,
      "grad_norm": 0.22357797622680664,
      "learning_rate": 0.00012334232083688633,
      "loss": 1.2072179794311524,
      "step": 1850
    },
    {
      "epoch": 1.4566505115778137,
      "grad_norm": 0.22742509841918945,
      "learning_rate": 0.0001224551417441539,
      "loss": 1.215613555908203,
      "step": 1860
    },
    {
      "epoch": 1.4644832819307778,
      "grad_norm": 0.2523942291736603,
      "learning_rate": 0.00012156609743180969,
      "loss": 1.203711700439453,
      "step": 1870
    },
    {
      "epoch": 1.4723160522837422,
      "grad_norm": 0.2069718986749649,
      "learning_rate": 0.00012067526174765691,
      "loss": 1.177224349975586,
      "step": 1880
    },
    {
      "epoch": 1.4801488226367063,
      "grad_norm": 0.29070842266082764,
      "learning_rate": 0.00011978270868829763,
      "loss": 1.249673557281494,
      "step": 1890
    },
    {
      "epoch": 1.4879815929896705,
      "grad_norm": 0.2660825848579407,
      "learning_rate": 0.00011888851239298646,
      "loss": 1.2102355003356933,
      "step": 1900
    },
    {
      "epoch": 1.4958143633426348,
      "grad_norm": 0.2503930628299713,
      "learning_rate": 0.00011799274713747219,
      "loss": 1.2331772804260255,
      "step": 1910
    },
    {
      "epoch": 1.503647133695599,
      "grad_norm": 0.24439479410648346,
      "learning_rate": 0.00011709548732782806,
      "loss": 1.225032138824463,
      "step": 1920
    },
    {
      "epoch": 1.5114799040485631,
      "grad_norm": 0.2666963040828705,
      "learning_rate": 0.00011619680749427142,
      "loss": 1.197390365600586,
      "step": 1930
    },
    {
      "epoch": 1.5193126744015273,
      "grad_norm": 0.2462136596441269,
      "learning_rate": 0.00011529678228497278,
      "loss": 1.1983027458190918,
      "step": 1940
    },
    {
      "epoch": 1.5271454447544917,
      "grad_norm": 0.2605383098125458,
      "learning_rate": 0.0001143954864598553,
      "loss": 1.226210594177246,
      "step": 1950
    },
    {
      "epoch": 1.5349782151074558,
      "grad_norm": 0.2872784435749054,
      "learning_rate": 0.00011349299488438485,
      "loss": 1.2140783309936523,
      "step": 1960
    },
    {
      "epoch": 1.5428109854604202,
      "grad_norm": 0.2374207228422165,
      "learning_rate": 0.00011258938252335157,
      "loss": 1.1981720924377441,
      "step": 1970
    },
    {
      "epoch": 1.5506437558133843,
      "grad_norm": 0.21907652914524078,
      "learning_rate": 0.00011168472443464261,
      "loss": 1.2283409118652344,
      "step": 1980
    },
    {
      "epoch": 1.5584765261663485,
      "grad_norm": 0.26273712515830994,
      "learning_rate": 0.00011077909576300789,
      "loss": 1.1771793365478516,
      "step": 1990
    },
    {
      "epoch": 1.5663092965193126,
      "grad_norm": 0.2728463411331177,
      "learning_rate": 0.00010987257173381796,
      "loss": 1.2644289016723633,
      "step": 2000
    },
    {
      "epoch": 1.5663092965193126,
      "eval_loss": 1.2753645181655884,
      "eval_runtime": 332.391,
      "eval_samples_per_second": 3.237,
      "eval_steps_per_second": 3.237,
      "step": 2000
    },
    {
      "epoch": 1.5741420668722768,
      "grad_norm": 0.2645413279533386,
      "learning_rate": 0.00010896522764681555,
      "loss": 1.2775901794433593,
      "step": 2010
    },
    {
      "epoch": 1.5819748372252411,
      "grad_norm": 0.2533264756202698,
      "learning_rate": 0.00010805713886986106,
      "loss": 1.1955472946166992,
      "step": 2020
    },
    {
      "epoch": 1.5898076075782053,
      "grad_norm": 0.2823348641395569,
      "learning_rate": 0.00010714838083267185,
      "loss": 1.2232410430908203,
      "step": 2030
    },
    {
      "epoch": 1.5976403779311696,
      "grad_norm": 0.2825281023979187,
      "learning_rate": 0.00010623902902055691,
      "loss": 1.2160611152648926,
      "step": 2040
    },
    {
      "epoch": 1.6054731482841338,
      "grad_norm": 0.26111525297164917,
      "learning_rate": 0.00010532915896814672,
      "loss": 1.2666211128234863,
      "step": 2050
    },
    {
      "epoch": 1.613305918637098,
      "grad_norm": 0.2571956217288971,
      "learning_rate": 0.000104418846253119,
      "loss": 1.2443676948547364,
      "step": 2060
    },
    {
      "epoch": 1.621138688990062,
      "grad_norm": 0.24284358322620392,
      "learning_rate": 0.00010350816648992092,
      "loss": 1.2544951438903809,
      "step": 2070
    },
    {
      "epoch": 1.6289714593430262,
      "grad_norm": 0.2704489529132843,
      "learning_rate": 0.0001025971953234882,
      "loss": 1.209574031829834,
      "step": 2080
    },
    {
      "epoch": 1.6368042296959906,
      "grad_norm": 0.26949822902679443,
      "learning_rate": 0.00010168600842296179,
      "loss": 1.2597773551940918,
      "step": 2090
    },
    {
      "epoch": 1.644637000048955,
      "grad_norm": 0.2654898166656494,
      "learning_rate": 0.00010077468147540239,
      "loss": 1.2378543853759765,
      "step": 2100
    },
    {
      "epoch": 1.6524697704019191,
      "grad_norm": 0.28754162788391113,
      "learning_rate": 9.986329017950369e-05,
      "loss": 1.2315853118896485,
      "step": 2110
    },
    {
      "epoch": 1.6603025407548833,
      "grad_norm": 0.26567715406417847,
      "learning_rate": 9.895191023930432e-05,
      "loss": 1.1968579292297363,
      "step": 2120
    },
    {
      "epoch": 1.6681353111078474,
      "grad_norm": 0.23091121017932892,
      "learning_rate": 9.804061735789972e-05,
      "loss": 1.1917951583862305,
      "step": 2130
    },
    {
      "epoch": 1.6759680814608116,
      "grad_norm": 0.23703314363956451,
      "learning_rate": 9.712948723115383e-05,
      "loss": 1.1869622230529786,
      "step": 2140
    },
    {
      "epoch": 1.683800851813776,
      "grad_norm": 0.2393905222415924,
      "learning_rate": 9.621859554141152e-05,
      "loss": 1.2089557647705078,
      "step": 2150
    },
    {
      "epoch": 1.69163362216674,
      "grad_norm": 0.2555121183395386,
      "learning_rate": 9.530801795121201e-05,
      "loss": 1.2293291091918945,
      "step": 2160
    },
    {
      "epoch": 1.6994663925197044,
      "grad_norm": 0.23827151954174042,
      "learning_rate": 9.439783009700418e-05,
      "loss": 1.2037882804870605,
      "step": 2170
    },
    {
      "epoch": 1.7072991628726686,
      "grad_norm": 0.26843613386154175,
      "learning_rate": 9.348810758286368e-05,
      "loss": 1.2214282989501952,
      "step": 2180
    },
    {
      "epoch": 1.7151319332256327,
      "grad_norm": 0.23154932260513306,
      "learning_rate": 9.25789259742131e-05,
      "loss": 1.2204715728759765,
      "step": 2190
    },
    {
      "epoch": 1.7229647035785969,
      "grad_norm": 0.2480868548154831,
      "learning_rate": 9.167036079154504e-05,
      "loss": 1.2267279624938965,
      "step": 2200
    },
    {
      "epoch": 1.730797473931561,
      "grad_norm": 0.23055171966552734,
      "learning_rate": 9.076248750414924e-05,
      "loss": 1.1989973068237305,
      "step": 2210
    },
    {
      "epoch": 1.7386302442845254,
      "grad_norm": 0.27379465103149414,
      "learning_rate": 8.98553815238436e-05,
      "loss": 1.2200066566467285,
      "step": 2220
    },
    {
      "epoch": 1.7464630146374898,
      "grad_norm": 0.2760390043258667,
      "learning_rate": 8.894911819871026e-05,
      "loss": 1.2310608863830566,
      "step": 2230
    },
    {
      "epoch": 1.754295784990454,
      "grad_norm": 0.22573843598365784,
      "learning_rate": 8.804377280683683e-05,
      "loss": 1.2127230644226075,
      "step": 2240
    },
    {
      "epoch": 1.762128555343418,
      "grad_norm": 0.2356061041355133,
      "learning_rate": 8.713942055006346e-05,
      "loss": 1.1980902671813964,
      "step": 2250
    },
    {
      "epoch": 1.7699613256963822,
      "grad_norm": 0.22806137800216675,
      "learning_rate": 8.62361365477363e-05,
      "loss": 1.2077306747436523,
      "step": 2260
    },
    {
      "epoch": 1.7777940960493464,
      "grad_norm": 0.2392238974571228,
      "learning_rate": 8.533399583046769e-05,
      "loss": 1.2229851722717284,
      "step": 2270
    },
    {
      "epoch": 1.7856268664023107,
      "grad_norm": 0.2972569763660431,
      "learning_rate": 8.443307333390395e-05,
      "loss": 1.2483017921447754,
      "step": 2280
    },
    {
      "epoch": 1.7934596367552749,
      "grad_norm": 0.23529693484306335,
      "learning_rate": 8.35334438925007e-05,
      "loss": 1.211320972442627,
      "step": 2290
    },
    {
      "epoch": 1.8012924071082392,
      "grad_norm": 0.24110788106918335,
      "learning_rate": 8.263518223330697e-05,
      "loss": 1.2276507377624513,
      "step": 2300
    },
    {
      "epoch": 1.8091251774612034,
      "grad_norm": 0.2633586525917053,
      "learning_rate": 8.173836296975792e-05,
      "loss": 1.1740093231201172,
      "step": 2310
    },
    {
      "epoch": 1.8169579478141675,
      "grad_norm": 0.22816379368305206,
      "learning_rate": 8.084306059547722e-05,
      "loss": 1.1880734443664551,
      "step": 2320
    },
    {
      "epoch": 1.8247907181671317,
      "grad_norm": 0.2651647925376892,
      "learning_rate": 7.994934947808928e-05,
      "loss": 1.204148769378662,
      "step": 2330
    },
    {
      "epoch": 1.8326234885200958,
      "grad_norm": 0.3007386028766632,
      "learning_rate": 7.905730385304193e-05,
      "loss": 1.228445816040039,
      "step": 2340
    },
    {
      "epoch": 1.8404562588730602,
      "grad_norm": 0.28230205178260803,
      "learning_rate": 7.816699781744011e-05,
      "loss": 1.2086246490478516,
      "step": 2350
    },
    {
      "epoch": 1.8482890292260243,
      "grad_norm": 0.21306121349334717,
      "learning_rate": 7.727850532389106e-05,
      "loss": 1.2209898948669433,
      "step": 2360
    },
    {
      "epoch": 1.8561217995789887,
      "grad_norm": 0.2833939790725708,
      "learning_rate": 7.639190017436149e-05,
      "loss": 1.2074974060058594,
      "step": 2370
    },
    {
      "epoch": 1.8639545699319529,
      "grad_norm": 0.2340923398733139,
      "learning_rate": 7.550725601404733e-05,
      "loss": 1.1948917388916016,
      "step": 2380
    },
    {
      "epoch": 1.871787340284917,
      "grad_norm": 0.26410800218582153,
      "learning_rate": 7.462464632525639e-05,
      "loss": 1.2041044235229492,
      "step": 2390
    },
    {
      "epoch": 1.8796201106378811,
      "grad_norm": 0.3063860535621643,
      "learning_rate": 7.374414442130458e-05,
      "loss": 1.1983399391174316,
      "step": 2400
    },
    {
      "epoch": 1.8874528809908453,
      "grad_norm": 0.2856113314628601,
      "learning_rate": 7.286582344042625e-05,
      "loss": 1.209029769897461,
      "step": 2410
    },
    {
      "epoch": 1.8952856513438097,
      "grad_norm": 0.26700469851493835,
      "learning_rate": 7.198975633969902e-05,
      "loss": 1.1910434722900392,
      "step": 2420
    },
    {
      "epoch": 1.903118421696774,
      "grad_norm": 0.28226250410079956,
      "learning_rate": 7.111601588898361e-05,
      "loss": 1.2325517654418945,
      "step": 2430
    },
    {
      "epoch": 1.9109511920497382,
      "grad_norm": 0.25455886125564575,
      "learning_rate": 7.024467466487934e-05,
      "loss": 1.2045490264892578,
      "step": 2440
    },
    {
      "epoch": 1.9187839624027023,
      "grad_norm": 0.24883849918842316,
      "learning_rate": 6.937580504469553e-05,
      "loss": 1.2177587509155274,
      "step": 2450
    },
    {
      "epoch": 1.9266167327556665,
      "grad_norm": 0.27279818058013916,
      "learning_rate": 6.850947920043948e-05,
      "loss": 1.2460152626037597,
      "step": 2460
    },
    {
      "epoch": 1.9344495031086306,
      "grad_norm": 0.2591603398323059,
      "learning_rate": 6.764576909282177e-05,
      "loss": 1.1761818885803224,
      "step": 2470
    },
    {
      "epoch": 1.942282273461595,
      "grad_norm": 0.2789348363876343,
      "learning_rate": 6.678474646527871e-05,
      "loss": 1.1885599136352538,
      "step": 2480
    },
    {
      "epoch": 1.9501150438145591,
      "grad_norm": 0.2906976342201233,
      "learning_rate": 6.592648283801308e-05,
      "loss": 1.2112513542175294,
      "step": 2490
    },
    {
      "epoch": 1.9579478141675235,
      "grad_norm": 0.27513882517814636,
      "learning_rate": 6.507104950205336e-05,
      "loss": 1.2198973655700684,
      "step": 2500
    },
    {
      "epoch": 1.9579478141675235,
      "eval_loss": 1.2582728862762451,
      "eval_runtime": 324.3533,
      "eval_samples_per_second": 3.317,
      "eval_steps_per_second": 3.317,
      "step": 2500
    },
    {
      "epoch": 1.9657805845204877,
      "grad_norm": 0.27247440814971924,
      "learning_rate": 6.421851751333197e-05,
      "loss": 1.2457717895507812,
      "step": 2510
    },
    {
      "epoch": 1.9736133548734518,
      "grad_norm": 0.3500891923904419,
      "learning_rate": 6.336895768678314e-05,
      "loss": 1.2087072372436523,
      "step": 2520
    },
    {
      "epoch": 1.981446125226416,
      "grad_norm": 0.29180240631103516,
      "learning_rate": 6.252244059046062e-05,
      "loss": 1.2157578468322754,
      "step": 2530
    },
    {
      "epoch": 1.98927889557938,
      "grad_norm": 0.2686278820037842,
      "learning_rate": 6.167903653967606e-05,
      "loss": 1.2206612586975099,
      "step": 2540
    },
    {
      "epoch": 1.9971116659323445,
      "grad_norm": 0.2538076341152191,
      "learning_rate": 6.083881559115829e-05,
      "loss": 1.1985113143920898,
      "step": 2550
    },
    {
      "epoch": 2.0046996622117783,
      "grad_norm": 0.2514108717441559,
      "learning_rate": 6.00018475372341e-05,
      "loss": 1.13430757522583,
      "step": 2560
    },
    {
      "epoch": 2.012532432564743,
      "grad_norm": 0.3025277554988861,
      "learning_rate": 5.91682019000311e-05,
      "loss": 1.1362914085388183,
      "step": 2570
    },
    {
      "epoch": 2.020365202917707,
      "grad_norm": 0.26406702399253845,
      "learning_rate": 5.833794792570275e-05,
      "loss": 1.1333977699279785,
      "step": 2580
    },
    {
      "epoch": 2.028197973270671,
      "grad_norm": 0.2084951251745224,
      "learning_rate": 5.751115457867653e-05,
      "loss": 1.111202335357666,
      "step": 2590
    },
    {
      "epoch": 2.0360307436236353,
      "grad_norm": 0.29670166969299316,
      "learning_rate": 5.668789053592565e-05,
      "loss": 1.0944523811340332,
      "step": 2600
    },
    {
      "epoch": 2.0438635139765995,
      "grad_norm": 0.3064367175102234,
      "learning_rate": 5.5868224181264136e-05,
      "loss": 1.0512296676635742,
      "step": 2610
    },
    {
      "epoch": 2.0516962843295636,
      "grad_norm": 0.2714754641056061,
      "learning_rate": 5.50522235996668e-05,
      "loss": 1.1794384956359862,
      "step": 2620
    },
    {
      "epoch": 2.0595290546825282,
      "grad_norm": 0.2643856704235077,
      "learning_rate": 5.4239956571613715e-05,
      "loss": 1.125117588043213,
      "step": 2630
    },
    {
      "epoch": 2.0673618250354924,
      "grad_norm": 0.29972824454307556,
      "learning_rate": 5.343149056746015e-05,
      "loss": 1.1104560852050782,
      "step": 2640
    },
    {
      "epoch": 2.0751945953884565,
      "grad_norm": 0.3134865462779999,
      "learning_rate": 5.262689274183215e-05,
      "loss": 1.1257891654968262,
      "step": 2650
    },
    {
      "epoch": 2.0830273657414207,
      "grad_norm": 0.28377866744995117,
      "learning_rate": 5.1826229928048363e-05,
      "loss": 1.1403872489929199,
      "step": 2660
    },
    {
      "epoch": 2.090860136094385,
      "grad_norm": 0.2689083516597748,
      "learning_rate": 5.1029568632568646e-05,
      "loss": 1.1211537361145019,
      "step": 2670
    },
    {
      "epoch": 2.098692906447349,
      "grad_norm": 0.27901580929756165,
      "learning_rate": 5.023697502946969e-05,
      "loss": 1.1200578689575196,
      "step": 2680
    },
    {
      "epoch": 2.106525676800313,
      "grad_norm": 0.2699909210205078,
      "learning_rate": 4.9448514954948364e-05,
      "loss": 1.1365811347961425,
      "step": 2690
    },
    {
      "epoch": 2.1143584471532777,
      "grad_norm": 0.3292345702648163,
      "learning_rate": 4.8664253901853054e-05,
      "loss": 1.1002737045288087,
      "step": 2700
    },
    {
      "epoch": 2.122191217506242,
      "grad_norm": 0.3296261727809906,
      "learning_rate": 4.7884257014243625e-05,
      "loss": 1.1261274337768554,
      "step": 2710
    },
    {
      "epoch": 2.130023987859206,
      "grad_norm": 0.284611314535141,
      "learning_rate": 4.710858908198016e-05,
      "loss": 1.1122213363647462,
      "step": 2720
    },
    {
      "epoch": 2.13785675821217,
      "grad_norm": 0.33854803442955017,
      "learning_rate": 4.633731453534126e-05,
      "loss": 1.1449526786804198,
      "step": 2730
    },
    {
      "epoch": 2.1456895285651343,
      "grad_norm": 0.32542046904563904,
      "learning_rate": 4.557049743967229e-05,
      "loss": 1.1548108100891112,
      "step": 2740
    },
    {
      "epoch": 2.1535222989180984,
      "grad_norm": 0.28194659948349,
      "learning_rate": 4.480820149006373e-05,
      "loss": 1.1169511795043945,
      "step": 2750
    },
    {
      "epoch": 2.161355069271063,
      "grad_norm": 0.2818988859653473,
      "learning_rate": 4.4050490006060453e-05,
      "loss": 1.1202062606811523,
      "step": 2760
    },
    {
      "epoch": 2.169187839624027,
      "grad_norm": 0.22729508578777313,
      "learning_rate": 4.3297425926402115e-05,
      "loss": 1.126146411895752,
      "step": 2770
    },
    {
      "epoch": 2.1770206099769913,
      "grad_norm": 0.28636181354522705,
      "learning_rate": 4.254907180379516e-05,
      "loss": 1.1105897903442383,
      "step": 2780
    },
    {
      "epoch": 2.1848533803299555,
      "grad_norm": 0.34403932094573975,
      "learning_rate": 4.180548979971701e-05,
      "loss": 1.1060968399047852,
      "step": 2790
    },
    {
      "epoch": 2.1926861506829196,
      "grad_norm": 0.34418439865112305,
      "learning_rate": 4.1066741679252597e-05,
      "loss": 1.131563377380371,
      "step": 2800
    },
    {
      "epoch": 2.2005189210358838,
      "grad_norm": 0.2736617922782898,
      "learning_rate": 4.033288880596385e-05,
      "loss": 1.140374755859375,
      "step": 2810
    },
    {
      "epoch": 2.208351691388848,
      "grad_norm": 0.35217058658599854,
      "learning_rate": 3.960399213679281e-05,
      "loss": 1.130772876739502,
      "step": 2820
    },
    {
      "epoch": 2.2161844617418125,
      "grad_norm": 0.23825931549072266,
      "learning_rate": 3.888011221699794e-05,
      "loss": 1.1130745887756348,
      "step": 2830
    },
    {
      "epoch": 2.2240172320947766,
      "grad_norm": 0.2941889762878418,
      "learning_rate": 3.816130917512525e-05,
      "loss": 1.16178617477417,
      "step": 2840
    },
    {
      "epoch": 2.231850002447741,
      "grad_norm": 0.3885084092617035,
      "learning_rate": 3.744764271801362e-05,
      "loss": 1.1225702285766601,
      "step": 2850
    },
    {
      "epoch": 2.239682772800705,
      "grad_norm": 0.3275553286075592,
      "learning_rate": 3.673917212583538e-05,
      "loss": 1.1144232749938965,
      "step": 2860
    },
    {
      "epoch": 2.247515543153669,
      "grad_norm": 0.3339763283729553,
      "learning_rate": 3.603595624717221e-05,
      "loss": 1.1165958404541017,
      "step": 2870
    },
    {
      "epoch": 2.2553483135066332,
      "grad_norm": 0.30861443281173706,
      "learning_rate": 3.533805349412691e-05,
      "loss": 1.103284740447998,
      "step": 2880
    },
    {
      "epoch": 2.263181083859598,
      "grad_norm": 0.2967692017555237,
      "learning_rate": 3.464552183747147e-05,
      "loss": 1.1307114601135253,
      "step": 2890
    },
    {
      "epoch": 2.271013854212562,
      "grad_norm": 0.35390040278434753,
      "learning_rate": 3.395841880183176e-05,
      "loss": 1.1555585861206055,
      "step": 2900
    },
    {
      "epoch": 2.278846624565526,
      "grad_norm": 0.3211853504180908,
      "learning_rate": 3.327680146090927e-05,
      "loss": 1.1091843605041505,
      "step": 2910
    },
    {
      "epoch": 2.2866793949184903,
      "grad_norm": 0.3494303822517395,
      "learning_rate": 3.260072643274044e-05,
      "loss": 1.136253547668457,
      "step": 2920
    },
    {
      "epoch": 2.2945121652714544,
      "grad_norm": 0.3768707811832428,
      "learning_rate": 3.1930249874993514e-05,
      "loss": 1.1504868507385253,
      "step": 2930
    },
    {
      "epoch": 2.3023449356244186,
      "grad_norm": 0.3536722958087921,
      "learning_rate": 3.126542748030402e-05,
      "loss": 1.1369070053100585,
      "step": 2940
    },
    {
      "epoch": 2.3101777059773827,
      "grad_norm": 0.3566012382507324,
      "learning_rate": 3.0606314471648643e-05,
      "loss": 1.1335080146789551,
      "step": 2950
    },
    {
      "epoch": 2.318010476330347,
      "grad_norm": 0.3815862834453583,
      "learning_rate": 2.99529655977582e-05,
      "loss": 1.079446029663086,
      "step": 2960
    },
    {
      "epoch": 2.3258432466833114,
      "grad_norm": 0.3307647705078125,
      "learning_rate": 2.9305435128569835e-05,
      "loss": 1.1216520309448241,
      "step": 2970
    },
    {
      "epoch": 2.3336760170362756,
      "grad_norm": 0.3462022840976715,
      "learning_rate": 2.8663776850719393e-05,
      "loss": 1.1272664070129395,
      "step": 2980
    },
    {
      "epoch": 2.3415087873892397,
      "grad_norm": 0.3372374475002289,
      "learning_rate": 2.8028044063073435e-05,
      "loss": 1.095858097076416,
      "step": 2990
    },
    {
      "epoch": 2.349341557742204,
      "grad_norm": 0.33324751257896423,
      "learning_rate": 2.739828957230207e-05,
      "loss": 1.1110441207885742,
      "step": 3000
    },
    {
      "epoch": 2.349341557742204,
      "eval_loss": 1.2691750526428223,
      "eval_runtime": 324.8724,
      "eval_samples_per_second": 3.312,
      "eval_steps_per_second": 3.312,
      "step": 3000
    },
    {
      "epoch": 2.357174328095168,
      "grad_norm": 0.2844167649745941,
      "learning_rate": 2.677456568849277e-05,
      "loss": 1.0753797531127929,
      "step": 3010
    },
    {
      "epoch": 2.3650070984481326,
      "grad_norm": 0.3673315644264221,
      "learning_rate": 2.6156924220804992e-05,
      "loss": 1.124093723297119,
      "step": 3020
    },
    {
      "epoch": 2.3728398688010968,
      "grad_norm": 0.30328670144081116,
      "learning_rate": 2.5545416473166883e-05,
      "loss": 1.1043742179870606,
      "step": 3030
    },
    {
      "epoch": 2.380672639154061,
      "grad_norm": 0.29031527042388916,
      "learning_rate": 2.4940093240013717e-05,
      "loss": 1.1230106353759766,
      "step": 3040
    },
    {
      "epoch": 2.388505409507025,
      "grad_norm": 0.3382628262042999,
      "learning_rate": 2.4341004802068658e-05,
      "loss": 1.1189377784729004,
      "step": 3050
    },
    {
      "epoch": 2.396338179859989,
      "grad_norm": 0.3430864214897156,
      "learning_rate": 2.3748200922166265e-05,
      "loss": 1.1290170669555664,
      "step": 3060
    },
    {
      "epoch": 2.4041709502129534,
      "grad_norm": 0.3256888687610626,
      "learning_rate": 2.3161730841118978e-05,
      "loss": 1.1303093910217286,
      "step": 3070
    },
    {
      "epoch": 2.4120037205659175,
      "grad_norm": 0.3369556963443756,
      "learning_rate": 2.2581643273626928e-05,
      "loss": 1.1088360786437987,
      "step": 3080
    },
    {
      "epoch": 2.4198364909188816,
      "grad_norm": 0.32177814841270447,
      "learning_rate": 2.200798640423156e-05,
      "loss": 1.0996588706970214,
      "step": 3090
    },
    {
      "epoch": 2.4276692612718462,
      "grad_norm": 0.3396743834018707,
      "learning_rate": 2.1440807883313095e-05,
      "loss": 1.0969182014465333,
      "step": 3100
    },
    {
      "epoch": 2.4355020316248104,
      "grad_norm": 0.2916245758533478,
      "learning_rate": 2.088015482313269e-05,
      "loss": 1.099278736114502,
      "step": 3110
    },
    {
      "epoch": 2.4433348019777745,
      "grad_norm": 0.3665787875652313,
      "learning_rate": 2.0326073793918867e-05,
      "loss": 1.1052253723144532,
      "step": 3120
    },
    {
      "epoch": 2.4511675723307387,
      "grad_norm": 0.308431476354599,
      "learning_rate": 1.977861081999931e-05,
      "loss": 1.1300259590148927,
      "step": 3130
    },
    {
      "epoch": 2.459000342683703,
      "grad_norm": 0.3617427945137024,
      "learning_rate": 1.923781137597791e-05,
      "loss": 1.119156265258789,
      "step": 3140
    },
    {
      "epoch": 2.466833113036667,
      "grad_norm": 0.4287793040275574,
      "learning_rate": 1.8703720382957392e-05,
      "loss": 1.0923386573791505,
      "step": 3150
    },
    {
      "epoch": 2.4746658833896316,
      "grad_norm": 0.2791413366794586,
      "learning_rate": 1.8176382204807984e-05,
      "loss": 1.1367136001586915,
      "step": 3160
    },
    {
      "epoch": 2.4824986537425957,
      "grad_norm": 0.3458610773086548,
      "learning_rate": 1.7655840644482403e-05,
      "loss": 1.13309907913208,
      "step": 3170
    },
    {
      "epoch": 2.49033142409556,
      "grad_norm": 0.3757402002811432,
      "learning_rate": 1.714213894037737e-05,
      "loss": 1.107917308807373,
      "step": 3180
    },
    {
      "epoch": 2.498164194448524,
      "grad_norm": 0.3789442479610443,
      "learning_rate": 1.663531976274204e-05,
      "loss": 1.0986825942993164,
      "step": 3190
    },
    {
      "epoch": 2.505996964801488,
      "grad_norm": 0.3685928285121918,
      "learning_rate": 1.6135425210133624e-05,
      "loss": 1.125806999206543,
      "step": 3200
    },
    {
      "epoch": 2.5138297351544523,
      "grad_norm": 0.3142971098423004,
      "learning_rate": 1.5642496805920537e-05,
      "loss": 1.113065528869629,
      "step": 3210
    },
    {
      "epoch": 2.5216625055074164,
      "grad_norm": 0.3308829069137573,
      "learning_rate": 1.515657549483328e-05,
      "loss": 1.121922206878662,
      "step": 3220
    },
    {
      "epoch": 2.529495275860381,
      "grad_norm": 0.34523579478263855,
      "learning_rate": 1.4677701639563368e-05,
      "loss": 1.1328062057495116,
      "step": 3230
    },
    {
      "epoch": 2.537328046213345,
      "grad_norm": 0.2806236445903778,
      "learning_rate": 1.4205915017410654e-05,
      "loss": 1.1356450080871583,
      "step": 3240
    },
    {
      "epoch": 2.5451608165663093,
      "grad_norm": 0.2967741787433624,
      "learning_rate": 1.3741254816979232e-05,
      "loss": 1.088940715789795,
      "step": 3250
    },
    {
      "epoch": 2.5529935869192735,
      "grad_norm": 0.35171636939048767,
      "learning_rate": 1.3283759634922332e-05,
      "loss": 1.1193583488464356,
      "step": 3260
    },
    {
      "epoch": 2.5608263572722376,
      "grad_norm": 0.32858696579933167,
      "learning_rate": 1.2833467472736248e-05,
      "loss": 1.09879789352417,
      "step": 3270
    },
    {
      "epoch": 2.568659127625202,
      "grad_norm": 0.34779906272888184,
      "learning_rate": 1.23904157336038e-05,
      "loss": 1.126400089263916,
      "step": 3280
    },
    {
      "epoch": 2.5764918979781664,
      "grad_norm": 0.39374393224716187,
      "learning_rate": 1.1954641219287466e-05,
      "loss": 1.16580810546875,
      "step": 3290
    },
    {
      "epoch": 2.5843246683311305,
      "grad_norm": 0.3365297317504883,
      "learning_rate": 1.152618012707246e-05,
      "loss": 1.120203685760498,
      "step": 3300
    },
    {
      "epoch": 2.5921574386840947,
      "grad_norm": 0.32898035645484924,
      "learning_rate": 1.1105068046760048e-05,
      "loss": 1.0829385757446288,
      "step": 3310
    },
    {
      "epoch": 2.599990209037059,
      "grad_norm": 0.3395327627658844,
      "learning_rate": 1.0691339957711289e-05,
      "loss": 1.142617416381836,
      "step": 3320
    },
    {
      "epoch": 2.607822979390023,
      "grad_norm": 0.32847779989242554,
      "learning_rate": 1.028503022594155e-05,
      "loss": 1.110114288330078,
      "step": 3330
    },
    {
      "epoch": 2.615655749742987,
      "grad_norm": 0.3730526566505432,
      "learning_rate": 9.886172601265842e-06,
      "loss": 1.124988079071045,
      "step": 3340
    },
    {
      "epoch": 2.6234885200959512,
      "grad_norm": 0.3661583662033081,
      "learning_rate": 9.494800214495459e-06,
      "loss": 1.1263715744018554,
      "step": 3350
    },
    {
      "epoch": 2.6313212904489154,
      "grad_norm": 0.36449870467185974,
      "learning_rate": 9.110945574685992e-06,
      "loss": 1.0590201377868653,
      "step": 3360
    },
    {
      "epoch": 2.63915406080188,
      "grad_norm": 0.2658601403236389,
      "learning_rate": 8.73464056643699e-06,
      "loss": 1.0944375038146972,
      "step": 3370
    },
    {
      "epoch": 2.646986831154844,
      "grad_norm": 0.47399842739105225,
      "learning_rate": 8.365916447243482e-06,
      "loss": 1.1339784622192384,
      "step": 3380
    },
    {
      "epoch": 2.6548196015078083,
      "grad_norm": 0.310308575630188,
      "learning_rate": 8.0048038448996e-06,
      "loss": 1.134721565246582,
      "step": 3390
    },
    {
      "epoch": 2.6626523718607724,
      "grad_norm": 0.32272788882255554,
      "learning_rate": 7.651332754954477e-06,
      "loss": 1.0940790176391602,
      "step": 3400
    },
    {
      "epoch": 2.6704851422137366,
      "grad_norm": 0.2760407626628876,
      "learning_rate": 7.305532538220749e-06,
      "loss": 1.1149394035339355,
      "step": 3410
    },
    {
      "epoch": 2.678317912566701,
      "grad_norm": 0.36517298221588135,
      "learning_rate": 6.967431918335676e-06,
      "loss": 1.113607692718506,
      "step": 3420
    },
    {
      "epoch": 2.6861506829196653,
      "grad_norm": 0.3604297935962677,
      "learning_rate": 6.63705897937521e-06,
      "loss": 1.1084628105163574,
      "step": 3430
    },
    {
      "epoch": 2.6939834532726294,
      "grad_norm": 0.3391871452331543,
      "learning_rate": 6.314441163521334e-06,
      "loss": 1.109107208251953,
      "step": 3440
    },
    {
      "epoch": 2.7018162236255936,
      "grad_norm": 0.3342093825340271,
      "learning_rate": 5.9996052687824045e-06,
      "loss": 1.147671890258789,
      "step": 3450
    },
    {
      "epoch": 2.7096489939785577,
      "grad_norm": 0.28410813212394714,
      "learning_rate": 5.692577446767389e-06,
      "loss": 1.09722900390625,
      "step": 3460
    },
    {
      "epoch": 2.717481764331522,
      "grad_norm": 0.3533446490764618,
      "learning_rate": 5.393383200513447e-06,
      "loss": 1.1344051361083984,
      "step": 3470
    },
    {
      "epoch": 2.725314534684486,
      "grad_norm": 0.3195113241672516,
      "learning_rate": 5.102047382367659e-06,
      "loss": 1.1147592544555665,
      "step": 3480
    },
    {
      "epoch": 2.73314730503745,
      "grad_norm": 0.33619624376296997,
      "learning_rate": 4.818594191922576e-06,
      "loss": 1.1492552757263184,
      "step": 3490
    },
    {
      "epoch": 2.7409800753904148,
      "grad_norm": 0.2879767119884491,
      "learning_rate": 4.543047174006209e-06,
      "loss": 1.1195759773254395,
      "step": 3500
    },
    {
      "epoch": 2.7409800753904148,
      "eval_loss": 1.2651914358139038,
      "eval_runtime": 326.4841,
      "eval_samples_per_second": 3.296,
      "eval_steps_per_second": 3.296,
      "step": 3500
    },
    {
      "epoch": 2.748812845743379,
      "grad_norm": 0.30044785141944885,
      "learning_rate": 4.275429216726201e-06,
      "loss": 1.1164446830749513,
      "step": 3510
    },
    {
      "epoch": 2.756645616096343,
      "grad_norm": 0.3198385536670685,
      "learning_rate": 4.015762549568724e-06,
      "loss": 1.0795605659484864,
      "step": 3520
    },
    {
      "epoch": 2.764478386449307,
      "grad_norm": 0.29045701026916504,
      "learning_rate": 3.7640687415519293e-06,
      "loss": 1.0812619209289551,
      "step": 3530
    },
    {
      "epoch": 2.7723111568022714,
      "grad_norm": 0.3366210460662842,
      "learning_rate": 3.520368699434384e-06,
      "loss": 1.1149346351623535,
      "step": 3540
    },
    {
      "epoch": 2.780143927155236,
      "grad_norm": 0.35936740040779114,
      "learning_rate": 3.2846826659784334e-06,
      "loss": 1.099393367767334,
      "step": 3550
    },
    {
      "epoch": 2.7879766975082,
      "grad_norm": 0.32345789670944214,
      "learning_rate": 3.057030218268764e-06,
      "loss": 1.1055167198181153,
      "step": 3560
    },
    {
      "epoch": 2.7958094678611642,
      "grad_norm": 0.37679150700569153,
      "learning_rate": 2.8374302660862826e-06,
      "loss": 1.1351775169372558,
      "step": 3570
    },
    {
      "epoch": 2.8036422382141284,
      "grad_norm": 0.3136025667190552,
      "learning_rate": 2.6259010503373206e-06,
      "loss": 1.0925300598144532,
      "step": 3580
    },
    {
      "epoch": 2.8114750085670925,
      "grad_norm": 0.4157222807407379,
      "learning_rate": 2.4224601415385294e-06,
      "loss": 1.0652141571044922,
      "step": 3590
    },
    {
      "epoch": 2.8193077789200567,
      "grad_norm": 0.39427298307418823,
      "learning_rate": 2.227124438357353e-06,
      "loss": 1.1031173706054687,
      "step": 3600
    },
    {
      "epoch": 2.827140549273021,
      "grad_norm": 0.38145479559898376,
      "learning_rate": 2.0399101662083807e-06,
      "loss": 1.1134588241577148,
      "step": 3610
    },
    {
      "epoch": 2.834973319625985,
      "grad_norm": 0.41683870553970337,
      "learning_rate": 1.8608328759055716e-06,
      "loss": 1.094143009185791,
      "step": 3620
    },
    {
      "epoch": 2.8428060899789496,
      "grad_norm": 0.3398721218109131,
      "learning_rate": 1.6899074423705975e-06,
      "loss": 1.1329906463623047,
      "step": 3630
    },
    {
      "epoch": 2.8506388603319137,
      "grad_norm": 0.3933563232421875,
      "learning_rate": 1.5271480633971768e-06,
      "loss": 1.0851118087768554,
      "step": 3640
    },
    {
      "epoch": 2.858471630684878,
      "grad_norm": 0.40157073736190796,
      "learning_rate": 1.372568258471807e-06,
      "loss": 1.1162568092346192,
      "step": 3650
    },
    {
      "epoch": 2.866304401037842,
      "grad_norm": 0.336107462644577,
      "learning_rate": 1.2261808676507724e-06,
      "loss": 1.0793692588806152,
      "step": 3660
    },
    {
      "epoch": 2.874137171390806,
      "grad_norm": 0.3799566328525543,
      "learning_rate": 1.0879980504935772e-06,
      "loss": 1.1028385162353516,
      "step": 3670
    },
    {
      "epoch": 2.8819699417437707,
      "grad_norm": 0.42130786180496216,
      "learning_rate": 9.580312850529295e-07,
      "loss": 1.1181939125061036,
      "step": 3680
    },
    {
      "epoch": 2.889802712096735,
      "grad_norm": 0.33488714694976807,
      "learning_rate": 8.362913669213379e-07,
      "loss": 1.0907830238342284,
      "step": 3690
    },
    {
      "epoch": 2.897635482449699,
      "grad_norm": 0.3703271448612213,
      "learning_rate": 7.227884083343406e-07,
      "loss": 1.080193328857422,
      "step": 3700
    },
    {
      "epoch": 2.905468252802663,
      "grad_norm": 0.35055869817733765,
      "learning_rate": 6.175318373306205e-07,
      "loss": 1.0740805625915528,
      "step": 3710
    },
    {
      "epoch": 2.9133010231556273,
      "grad_norm": 0.3170068562030792,
      "learning_rate": 5.205303969687658e-07,
      "loss": 1.1151776313781738,
      "step": 3720
    },
    {
      "epoch": 2.9211337935085915,
      "grad_norm": 0.29611772298812866,
      "learning_rate": 4.317921446011286e-07,
      "loss": 1.1126981735229493,
      "step": 3730
    },
    {
      "epoch": 2.9289665638615556,
      "grad_norm": 0.37433868646621704,
      "learning_rate": 3.513244512045155e-07,
      "loss": 1.1250258445739747,
      "step": 3740
    },
    {
      "epoch": 2.9367993342145198,
      "grad_norm": 0.3662181794643402,
      "learning_rate": 2.7913400076787776e-07,
      "loss": 1.0847277641296387,
      "step": 3750
    },
    {
      "epoch": 2.9446321045674844,
      "grad_norm": 0.3808414340019226,
      "learning_rate": 2.1522678973718847e-07,
      "loss": 1.1382882118225097,
      "step": 3760
    },
    {
      "epoch": 2.9524648749204485,
      "grad_norm": 0.3531859219074249,
      "learning_rate": 1.5960812651728552e-07,
      "loss": 1.1251526832580567,
      "step": 3770
    },
    {
      "epoch": 2.9602976452734127,
      "grad_norm": 0.3787539005279541,
      "learning_rate": 1.1228263103099101e-07,
      "loss": 1.097026538848877,
      "step": 3780
    },
    {
      "epoch": 2.968130415626377,
      "grad_norm": 0.3101099133491516,
      "learning_rate": 7.32542343352849e-08,
      "loss": 1.0861360549926757,
      "step": 3790
    },
    {
      "epoch": 2.975963185979341,
      "grad_norm": 0.320687860250473,
      "learning_rate": 4.2526178294877153e-08,
      "loss": 1.0637449264526366,
      "step": 3800
    },
    {
      "epoch": 2.9837959563323055,
      "grad_norm": 0.2555408775806427,
      "learning_rate": 2.010101531282338e-08,
      "loss": 1.1155780792236327,
      "step": 3810
    },
    {
      "epoch": 2.9916287266852697,
      "grad_norm": 0.39955031871795654,
      "learning_rate": 5.980608118572039e-09,
      "loss": 1.1184403419494628,
      "step": 3820
    },
    {
      "epoch": 2.999461497038234,
      "grad_norm": 0.35409417748451233,
      "learning_rate": 1.661296132327017e-10,
      "loss": 1.114205551147461,
      "step": 3830
    },
    {
      "epoch": 3.0,
      "eval_loss": 1.2653985023498535,
      "eval_runtime": 324.3577,
      "eval_samples_per_second": 3.317,
      "eval_steps_per_second": 3.317,
      "step": 3831
    },
    {
      "epoch": 3.0,
      "step": 3831,
      "total_flos": 3.5276122661958113e+18,
      "train_loss": 1.2349582789336941,
      "train_runtime": 87806.6728,
      "train_samples_per_second": 0.698,
      "train_steps_per_second": 0.044
    }
  ],
  "logging_steps": 10,
  "max_steps": 3831,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.5276122661958113e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}