diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,13770 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 19613,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005098659052669148,
+      "grad_norm": 946.1547547167231,
+      "learning_rate": 1.5280135823429543e-06,
+      "loss": 9.677,
+      "step": 10
+    },
+    {
+      "epoch": 0.0010197318105338296,
+      "grad_norm": 544.7247510656614,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 8.2775,
+      "step": 20
+    },
+    {
+      "epoch": 0.0015295977158007444,
+      "grad_norm": 1728.4541198728905,
+      "learning_rate": 4.923599320882852e-06,
+      "loss": 7.1393,
+      "step": 30
+    },
+    {
+      "epoch": 0.002039463621067659,
+      "grad_norm": 479.729349357684,
+      "learning_rate": 6.621392190152802e-06,
+      "loss": 6.1852,
+      "step": 40
+    },
+    {
+      "epoch": 0.002549329526334574,
+      "grad_norm": 894.5276149271358,
+      "learning_rate": 8.31918505942275e-06,
+      "loss": 5.4336,
+      "step": 50
+    },
+    {
+      "epoch": 0.0030591954316014887,
+      "grad_norm": 591.4100785573886,
+      "learning_rate": 1.00169779286927e-05,
+      "loss": 5.0479,
+      "step": 60
+    },
+    {
+      "epoch": 0.0035690613368684035,
+      "grad_norm": 299.20502596928674,
+      "learning_rate": 1.171477079796265e-05,
+      "loss": 4.6588,
+      "step": 70
+    },
+    {
+      "epoch": 0.004078927242135318,
+      "grad_norm": 326.0090819574599,
+      "learning_rate": 1.3412563667232598e-05,
+      "loss": 4.4741,
+      "step": 80
+    },
+    {
+      "epoch": 0.0045887931474022335,
+      "grad_norm": 553.1152730771238,
+      "learning_rate": 1.5110356536502549e-05,
+      "loss": 4.3923,
+      "step": 90
+    },
+    {
+      "epoch": 0.005098659052669148,
+      "grad_norm": 429.2200639978728,
+      "learning_rate": 1.6808149405772495e-05,
+      "loss": 4.3827,
+      "step": 100
+    },
+    {
+      "epoch": 0.005608524957936063,
+      "grad_norm": 217.043209696176,
+      "learning_rate": 1.8505942275042447e-05,
+      "loss": 4.3381,
+      "step": 110
+    },
+    {
+      "epoch": 0.0061183908632029774,
+      "grad_norm": 71.5225482068816,
+      "learning_rate": 2.0203735144312395e-05,
+      "loss": 4.1111,
+      "step": 120
+    },
+    {
+      "epoch": 0.006628256768469893,
+      "grad_norm": 59.30962303454184,
+      "learning_rate": 2.1901528013582346e-05,
+      "loss": 4.0054,
+      "step": 130
+    },
+    {
+      "epoch": 0.007138122673736807,
+      "grad_norm": 44.78043152216485,
+      "learning_rate": 2.3599320882852294e-05,
+      "loss": 3.9987,
+      "step": 140
+    },
+    {
+      "epoch": 0.007647988579003722,
+      "grad_norm": 86.47447277328804,
+      "learning_rate": 2.5297113752122242e-05,
+      "loss": 3.8178,
+      "step": 150
+    },
+    {
+      "epoch": 0.008157854484270637,
+      "grad_norm": 33.444719477873726,
+      "learning_rate": 2.699490662139219e-05,
+      "loss": 3.6614,
+      "step": 160
+    },
+    {
+      "epoch": 0.008667720389537552,
+      "grad_norm": 48.632312360697625,
+      "learning_rate": 2.869269949066214e-05,
+      "loss": 3.5938,
+      "step": 170
+    },
+    {
+      "epoch": 0.009177586294804467,
+      "grad_norm": 43.824964873367996,
+      "learning_rate": 3.0390492359932087e-05,
+      "loss": 3.5408,
+      "step": 180
+    },
+    {
+      "epoch": 0.00968745220007138,
+      "grad_norm": 48.440193697215264,
+      "learning_rate": 3.2088285229202035e-05,
+      "loss": 3.4292,
+      "step": 190
+    },
+    {
+      "epoch": 0.010197318105338296,
+      "grad_norm": 46.372222786863986,
+      "learning_rate": 3.378607809847199e-05,
+      "loss": 3.2039,
+      "step": 200
+    },
+    {
+      "epoch": 0.010707184010605211,
+      "grad_norm": 82.68846959999989,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 3.0534,
+      "step": 210
+    },
+    {
+      "epoch": 0.011217049915872126,
+      "grad_norm": 99.53466045235125,
+      "learning_rate": 3.7181663837011886e-05,
+      "loss": 2.9101,
+      "step": 220
+    },
+    {
+      "epoch": 0.01172691582113904,
+      "grad_norm": 64.16723536538507,
+      "learning_rate": 3.8879456706281834e-05,
+      "loss": 2.8509,
+      "step": 230
+    },
+    {
+      "epoch": 0.012236781726405955,
+      "grad_norm": 60.247152435132286,
+      "learning_rate": 4.057724957555178e-05,
+      "loss": 2.7339,
+      "step": 240
+    },
+    {
+      "epoch": 0.01274664763167287,
+      "grad_norm": 126.69897963259963,
+      "learning_rate": 4.227504244482173e-05,
+      "loss": 2.5434,
+      "step": 250
+    },
+    {
+      "epoch": 0.013256513536939785,
+      "grad_norm": 71.30801147824252,
+      "learning_rate": 4.3972835314091685e-05,
+      "loss": 2.5203,
+      "step": 260
+    },
+    {
+      "epoch": 0.013766379442206699,
+      "grad_norm": 54.55308844523217,
+      "learning_rate": 4.567062818336163e-05,
+      "loss": 2.4222,
+      "step": 270
+    },
+    {
+      "epoch": 0.014276245347473614,
+      "grad_norm": 36.84656998948245,
+      "learning_rate": 4.736842105263158e-05,
+      "loss": 2.3382,
+      "step": 280
+    },
+    {
+      "epoch": 0.01478611125274053,
+      "grad_norm": 54.17242333918137,
+      "learning_rate": 4.9066213921901536e-05,
+      "loss": 2.295,
+      "step": 290
+    },
+    {
+      "epoch": 0.015295977158007444,
+      "grad_norm": 64.90929737356474,
+      "learning_rate": 5.0764006791171484e-05,
+      "loss": 2.2441,
+      "step": 300
+    },
+    {
+      "epoch": 0.015805843063274358,
+      "grad_norm": 46.816781644417524,
+      "learning_rate": 5.2461799660441425e-05,
+      "loss": 2.1804,
+      "step": 310
+    },
+    {
+      "epoch": 0.016315708968541273,
+      "grad_norm": 56.46868292969693,
+      "learning_rate": 5.415959252971138e-05,
+      "loss": 2.1558,
+      "step": 320
+    },
+    {
+      "epoch": 0.01682557487380819,
+      "grad_norm": 48.224835141523755,
+      "learning_rate": 5.5857385398981335e-05,
+      "loss": 2.1064,
+      "step": 330
+    },
+    {
+      "epoch": 0.017335440779075104,
+      "grad_norm": 43.63841495963787,
+      "learning_rate": 5.7555178268251276e-05,
+      "loss": 2.1382,
+      "step": 340
+    },
+    {
+      "epoch": 0.01784530668434202,
+      "grad_norm": 44.816260186893615,
+      "learning_rate": 5.925297113752123e-05,
+      "loss": 2.0626,
+      "step": 350
+    },
+    {
+      "epoch": 0.018355172589608934,
+      "grad_norm": 35.749490569385365,
+      "learning_rate": 6.095076400679117e-05,
+      "loss": 2.0319,
+      "step": 360
+    },
+    {
+      "epoch": 0.01886503849487585,
+      "grad_norm": 36.955674387001245,
+      "learning_rate": 6.264855687606112e-05,
+      "loss": 2.0229,
+      "step": 370
+    },
+    {
+      "epoch": 0.01937490440014276,
+      "grad_norm": 34.68228399222833,
+      "learning_rate": 6.434634974533107e-05,
+      "loss": 1.9807,
+      "step": 380
+    },
+    {
+      "epoch": 0.019884770305409676,
+      "grad_norm": 39.21220369968134,
+      "learning_rate": 6.604414261460103e-05,
+      "loss": 1.8996,
+      "step": 390
+    },
+    {
+      "epoch": 0.02039463621067659,
+      "grad_norm": 63.36896641437639,
+      "learning_rate": 6.774193548387096e-05,
+      "loss": 1.8744,
+      "step": 400
+    },
+    {
+      "epoch": 0.020904502115943507,
+      "grad_norm": 35.410707083894266,
+      "learning_rate": 6.943972835314093e-05,
+      "loss": 1.883,
+      "step": 410
+    },
+    {
+      "epoch": 0.021414368021210422,
+      "grad_norm": 30.989701961084524,
+      "learning_rate": 7.113752122241087e-05,
+      "loss": 1.8126,
+      "step": 420
+    },
+    {
+      "epoch": 0.021924233926477337,
+      "grad_norm": 36.11852060166006,
+      "learning_rate": 7.283531409168082e-05,
+      "loss": 1.931,
+      "step": 430
+    },
+    {
+      "epoch": 0.022434099831744252,
+      "grad_norm": 34.26327225869466,
+      "learning_rate": 7.453310696095077e-05,
+      "loss": 1.8503,
+      "step": 440
+    },
+    {
+      "epoch": 0.022943965737011168,
+      "grad_norm": 35.512191524471405,
+      "learning_rate": 7.623089983022072e-05,
+      "loss": 1.7695,
+      "step": 450
+    },
+    {
+      "epoch": 0.02345383164227808,
+      "grad_norm": 30.89105132115667,
+      "learning_rate": 7.792869269949067e-05,
+      "loss": 1.7874,
+      "step": 460
+    },
+    {
+      "epoch": 0.023963697547544995,
+      "grad_norm": 30.637291014545614,
+      "learning_rate": 7.962648556876061e-05,
+      "loss": 1.8044,
+      "step": 470
+    },
+    {
+      "epoch": 0.02447356345281191,
+      "grad_norm": 23.622670649093685,
+      "learning_rate": 8.132427843803056e-05,
+      "loss": 1.7816,
+      "step": 480
+    },
+    {
+      "epoch": 0.024983429358078825,
+      "grad_norm": 30.805248984521565,
+      "learning_rate": 8.302207130730051e-05,
+      "loss": 1.7745,
+      "step": 490
+    },
+    {
+      "epoch": 0.02549329526334574,
+      "grad_norm": 28.992150209438435,
+      "learning_rate": 8.471986417657046e-05,
+      "loss": 1.8071,
+      "step": 500
+    },
+    {
+      "epoch": 0.026003161168612655,
+      "grad_norm": 32.17766281670865,
+      "learning_rate": 8.641765704584042e-05,
+      "loss": 1.735,
+      "step": 510
+    },
+    {
+      "epoch": 0.02651302707387957,
+      "grad_norm": 22.604114563009052,
+      "learning_rate": 8.811544991511035e-05,
+      "loss": 1.8334,
+      "step": 520
+    },
+    {
+      "epoch": 0.027022892979146486,
+      "grad_norm": 22.175478912262538,
+      "learning_rate": 8.981324278438032e-05,
+      "loss": 1.659,
+      "step": 530
+    },
+    {
+      "epoch": 0.027532758884413398,
+      "grad_norm": 29.575964324637216,
+      "learning_rate": 9.151103565365025e-05,
+      "loss": 1.6867,
+      "step": 540
+    },
+    {
+      "epoch": 0.028042624789680313,
+      "grad_norm": 23.576640350398364,
+      "learning_rate": 9.320882852292021e-05,
+      "loss": 1.7008,
+      "step": 550
+    },
+    {
+      "epoch": 0.028552490694947228,
+      "grad_norm": 37.51998597499104,
+      "learning_rate": 9.490662139219015e-05,
+      "loss": 1.7211,
+      "step": 560
+    },
+    {
+      "epoch": 0.029062356600214143,
+      "grad_norm": 15.417959281812669,
+      "learning_rate": 9.660441426146011e-05,
+      "loss": 1.6255,
+      "step": 570
+    },
+    {
+      "epoch": 0.02957222250548106,
+      "grad_norm": 22.186410865841005,
+      "learning_rate": 9.830220713073006e-05,
+      "loss": 1.6677,
+      "step": 580
+    },
+    {
+      "epoch": 0.030082088410747974,
+      "grad_norm": 28.48049006276817,
+      "learning_rate": 0.0001,
+      "loss": 1.6622,
+      "step": 590
+    },
+    {
+      "epoch": 0.03059195431601489,
+      "grad_norm": 20.16062115399988,
+      "learning_rate": 9.999993182329924e-05,
+      "loss": 1.6327,
+      "step": 600
+    },
+    {
+      "epoch": 0.031101820221281804,
+      "grad_norm": 20.503281569928266,
+      "learning_rate": 9.999972729338288e-05,
+      "loss": 1.6157,
+      "step": 610
+    },
+    {
+      "epoch": 0.031611686126548716,
+      "grad_norm": 16.081250504970278,
+      "learning_rate": 9.99993864108087e-05,
+      "loss": 1.6535,
+      "step": 620
+    },
+    {
+      "epoch": 0.03212155203181563,
+      "grad_norm": 23.03873979953226,
+      "learning_rate": 9.999890917650628e-05,
+      "loss": 1.5929,
+      "step": 630
+    },
+    {
+      "epoch": 0.032631417937082546,
+      "grad_norm": 17.488947087447386,
+      "learning_rate": 9.999829559177709e-05,
+      "loss": 1.5407,
+      "step": 640
+    },
+    {
+      "epoch": 0.03314128384234946,
+      "grad_norm": 20.532891681871146,
+      "learning_rate": 9.999754565829444e-05,
+      "loss": 1.6118,
+      "step": 650
+    },
+    {
+      "epoch": 0.03365114974761638,
+      "grad_norm": 24.990656263548626,
+      "learning_rate": 9.999665937810341e-05,
+      "loss": 1.6121,
+      "step": 660
+    },
+    {
+      "epoch": 0.03416101565288329,
+      "grad_norm": 18.93753558506703,
+      "learning_rate": 9.999563675362096e-05,
+      "loss": 1.59,
+      "step": 670
+    },
+    {
+      "epoch": 0.03467088155815021,
+      "grad_norm": 14.663441382121054,
+      "learning_rate": 9.999447778763585e-05,
+      "loss": 1.5577,
+      "step": 680
+    },
+    {
+      "epoch": 0.03518074746341712,
+      "grad_norm": 18.757298128213563,
+      "learning_rate": 9.999318248330867e-05,
+      "loss": 1.5638,
+      "step": 690
+    },
+    {
+      "epoch": 0.03569061336868404,
+      "grad_norm": 18.97335326579308,
+      "learning_rate": 9.999175084417182e-05,
+      "loss": 1.5323,
+      "step": 700
+    },
+    {
+      "epoch": 0.03620047927395095,
+      "grad_norm": 16.333532400542737,
+      "learning_rate": 9.999018287412945e-05,
+      "loss": 1.596,
+      "step": 710
+    },
+    {
+      "epoch": 0.03671034517921787,
+      "grad_norm": 12.898417787895172,
+      "learning_rate": 9.998847857745752e-05,
+      "loss": 1.6924,
+      "step": 720
+    },
+    {
+      "epoch": 0.03722021108448478,
+      "grad_norm": 15.078759025875287,
+      "learning_rate": 9.998663795880377e-05,
+      "loss": 1.5784,
+      "step": 730
+    },
+    {
+      "epoch": 0.0377300769897517,
+      "grad_norm": 25.78359047701089,
+      "learning_rate": 9.998466102318771e-05,
+      "loss": 1.5583,
+      "step": 740
+    },
+    {
+      "epoch": 0.03823994289501861,
+      "grad_norm": 13.495256766439768,
+      "learning_rate": 9.998254777600056e-05,
+      "loss": 1.5748,
+      "step": 750
+    },
+    {
+      "epoch": 0.03874980880028552,
+      "grad_norm": 25.54450279200201,
+      "learning_rate": 9.998029822300528e-05,
+      "loss": 1.4746,
+      "step": 760
+    },
+    {
+      "epoch": 0.03925967470555244,
+      "grad_norm": 16.607989604937515,
+      "learning_rate": 9.997791237033657e-05,
+      "loss": 1.5983,
+      "step": 770
+    },
+    {
+      "epoch": 0.03976954061081935,
+      "grad_norm": 13.846076340497213,
+      "learning_rate": 9.997539022450082e-05,
+      "loss": 1.5098,
+      "step": 780
+    },
+    {
+      "epoch": 0.04027940651608627,
+      "grad_norm": 17.303687104171345,
+      "learning_rate": 9.997273179237606e-05,
+      "loss": 1.5409,
+      "step": 790
+    },
+    {
+      "epoch": 0.04078927242135318,
+      "grad_norm": 18.449418144738164,
+      "learning_rate": 9.996993708121206e-05,
+      "loss": 1.5081,
+      "step": 800
+    },
+    {
+      "epoch": 0.0412991383266201,
+      "grad_norm": 14.303705630794529,
+      "learning_rate": 9.996700609863013e-05,
+      "loss": 1.5393,
+      "step": 810
+    },
+    {
+      "epoch": 0.04180900423188701,
+      "grad_norm": 17.238984622051,
+      "learning_rate": 9.996393885262332e-05,
+      "loss": 1.5065,
+      "step": 820
+    },
+    {
+      "epoch": 0.04231887013715393,
+      "grad_norm": 12.343117780150777,
+      "learning_rate": 9.996073535155618e-05,
+      "loss": 1.459,
+      "step": 830
+    },
+    {
+      "epoch": 0.042828736042420844,
+      "grad_norm": 18.09183197481867,
+      "learning_rate": 9.995739560416488e-05,
+      "loss": 1.5167,
+      "step": 840
+    },
+    {
+      "epoch": 0.04333860194768776,
+      "grad_norm": 11.840493736268275,
+      "learning_rate": 9.995391961955715e-05,
+      "loss": 1.4625,
+      "step": 850
+    },
+    {
+      "epoch": 0.043848467852954674,
+      "grad_norm": 11.51780571201556,
+      "learning_rate": 9.995030740721225e-05,
+      "loss": 1.4507,
+      "step": 860
+    },
+    {
+      "epoch": 0.04435833375822159,
+      "grad_norm": 13.036311222953925,
+      "learning_rate": 9.994655897698088e-05,
+      "loss": 1.405,
+      "step": 870
+    },
+    {
+      "epoch": 0.044868199663488505,
+      "grad_norm": 19.11790748949296,
+      "learning_rate": 9.994267433908533e-05,
+      "loss": 1.5213,
+      "step": 880
+    },
+    {
+      "epoch": 0.04537806556875542,
+      "grad_norm": 11.616955086783236,
+      "learning_rate": 9.993865350411922e-05,
+      "loss": 1.4358,
+      "step": 890
+    },
+    {
+      "epoch": 0.045887931474022335,
+      "grad_norm": 15.202800317578177,
+      "learning_rate": 9.993449648304765e-05,
+      "loss": 1.4656,
+      "step": 900
+    },
+    {
+      "epoch": 0.04639779737928925,
+      "grad_norm": 10.861336191036171,
+      "learning_rate": 9.99302032872071e-05,
+      "loss": 1.428,
+      "step": 910
+    },
+    {
+      "epoch": 0.04690766328455616,
+      "grad_norm": 13.201777890271243,
+      "learning_rate": 9.992577392830542e-05,
+      "loss": 1.4195,
+      "step": 920
+    },
+    {
+      "epoch": 0.047417529189823074,
+      "grad_norm": 9.178791009234759,
+      "learning_rate": 9.992120841842177e-05,
+      "loss": 1.4354,
+      "step": 930
+    },
+    {
+      "epoch": 0.04792739509508999,
+      "grad_norm": 15.66628336799611,
+      "learning_rate": 9.99165067700066e-05,
+      "loss": 1.3938,
+      "step": 940
+    },
+    {
+      "epoch": 0.048437261000356904,
+      "grad_norm": 11.59188910926961,
+      "learning_rate": 9.991166899588163e-05,
+      "loss": 1.4134,
+      "step": 950
+    },
+    {
+      "epoch": 0.04894712690562382,
+      "grad_norm": 10.752174542049293,
+      "learning_rate": 9.99066951092398e-05,
+      "loss": 1.3447,
+      "step": 960
+    },
+    {
+      "epoch": 0.049456992810890735,
+      "grad_norm": 12.322110274926507,
+      "learning_rate": 9.990158512364522e-05,
+      "loss": 1.4129,
+      "step": 970
+    },
+    {
+      "epoch": 0.04996685871615765,
+      "grad_norm": 11.447428766261394,
+      "learning_rate": 9.989633905303319e-05,
+      "loss": 1.4276,
+      "step": 980
+    },
+    {
+      "epoch": 0.050476724621424565,
+      "grad_norm": 12.852124943645142,
+      "learning_rate": 9.98909569117101e-05,
+      "loss": 1.3558,
+      "step": 990
+    },
+    {
+      "epoch": 0.05098659052669148,
+      "grad_norm": 14.822190112357378,
+      "learning_rate": 9.98854387143534e-05,
+      "loss": 1.4623,
+      "step": 1000
+    },
+    {
+      "epoch": 0.051496456431958396,
+      "grad_norm": 15.515256851316536,
+      "learning_rate": 9.98797844760116e-05,
+      "loss": 1.4883,
+      "step": 1010
+    },
+    {
+      "epoch": 0.05200632233722531,
+      "grad_norm": 9.567405011779273,
+      "learning_rate": 9.987399421210419e-05,
+      "loss": 1.4434,
+      "step": 1020
+    },
+    {
+      "epoch": 0.052516188242492226,
+      "grad_norm": 9.256164671608484,
+      "learning_rate": 9.986806793842163e-05,
+      "loss": 1.3478,
+      "step": 1030
+    },
+    {
+      "epoch": 0.05302605414775914,
+      "grad_norm": 17.070804354637016,
+      "learning_rate": 9.986200567112523e-05,
+      "loss": 1.3341,
+      "step": 1040
+    },
+    {
+      "epoch": 0.053535920053026056,
+      "grad_norm": 14.973890022995453,
+      "learning_rate": 9.985580742674725e-05,
+      "loss": 1.4138,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05404578595829297,
+      "grad_norm": 9.98151039457384,
+      "learning_rate": 9.98494732221907e-05,
+      "loss": 1.4348,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05455565186355989,
+      "grad_norm": 9.213280784853575,
+      "learning_rate": 9.98430030747294e-05,
+      "loss": 1.3649,
+      "step": 1070
+    },
+    {
+      "epoch": 0.055065517768826795,
+      "grad_norm": 11.512319636509554,
+      "learning_rate": 9.983639700200786e-05,
+      "loss": 1.4215,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05557538367409371,
+      "grad_norm": 12.375853723369742,
+      "learning_rate": 9.982965502204131e-05,
+      "loss": 1.3548,
+      "step": 1090
+    },
+    {
+      "epoch": 0.056085249579360626,
+      "grad_norm": 16.63336865073922,
+      "learning_rate": 9.98227771532156e-05,
+      "loss": 1.3482,
+      "step": 1100
+    },
+    {
+      "epoch": 0.05659511548462754,
+      "grad_norm": 11.640203752023153,
+      "learning_rate": 9.98157634142871e-05,
+      "loss": 1.3411,
+      "step": 1110
+    },
+    {
+      "epoch": 0.057104981389894456,
+      "grad_norm": 9.904873969510755,
+      "learning_rate": 9.98086138243828e-05,
+      "loss": 1.3865,
+      "step": 1120
+    },
+    {
+      "epoch": 0.05761484729516137,
+      "grad_norm": 6.885520749380552,
+      "learning_rate": 9.98013284030001e-05,
+      "loss": 1.3043,
+      "step": 1130
+    },
+    {
+      "epoch": 0.05812471320042829,
+      "grad_norm": 7.508737754108616,
+      "learning_rate": 9.979390717000682e-05,
+      "loss": 1.4498,
+      "step": 1140
+    },
+    {
+      "epoch": 0.0586345791056952,
+      "grad_norm": 12.596890247499731,
+      "learning_rate": 9.978635014564121e-05,
+      "loss": 1.3844,
+      "step": 1150
+    },
+    {
+      "epoch": 0.05914444501096212,
+      "grad_norm": 14.716717678438053,
+      "learning_rate": 9.977865735051176e-05,
+      "loss": 1.2998,
+      "step": 1160
+    },
+    {
+      "epoch": 0.05965431091622903,
+      "grad_norm": 8.86525041475244,
+      "learning_rate": 9.977082880559725e-05,
+      "loss": 1.4236,
+      "step": 1170
+    },
+    {
+      "epoch": 0.06016417682149595,
+      "grad_norm": 8.438445931523896,
+      "learning_rate": 9.976286453224665e-05,
+      "loss": 1.3377,
+      "step": 1180
+    },
+    {
+      "epoch": 0.06067404272676286,
+      "grad_norm": 11.987494578604688,
+      "learning_rate": 9.975476455217909e-05,
+      "loss": 1.3396,
+      "step": 1190
+    },
+    {
+      "epoch": 0.06118390863202978,
+      "grad_norm": 8.508524307495275,
+      "learning_rate": 9.974652888748377e-05,
+      "loss": 1.3562,
+      "step": 1200
+    },
+    {
+      "epoch": 0.06169377453729669,
+      "grad_norm": 10.139996913816189,
+      "learning_rate": 9.973815756061988e-05,
+      "loss": 1.3081,
+      "step": 1210
+    },
+    {
+      "epoch": 0.06220364044256361,
+      "grad_norm": 8.893072723924112,
+      "learning_rate": 9.972965059441662e-05,
+      "loss": 1.3785,
+      "step": 1220
+    },
+    {
+      "epoch": 0.06271350634783052,
+      "grad_norm": 13.144137366868113,
+      "learning_rate": 9.972100801207306e-05,
+      "loss": 1.3559,
+      "step": 1230
+    },
+    {
+      "epoch": 0.06322337225309743,
+      "grad_norm": 7.993631325592355,
+      "learning_rate": 9.97122298371581e-05,
+      "loss": 1.2776,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06373323815836435,
+      "grad_norm": 8.103661114339973,
+      "learning_rate": 9.970331609361045e-05,
+      "loss": 1.2851,
+      "step": 1250
+    },
+    {
+      "epoch": 0.06424310406363126,
+      "grad_norm": 7.326903132891211,
+      "learning_rate": 9.969426680573847e-05,
+      "loss": 1.3169,
+      "step": 1260
+    },
+    {
+      "epoch": 0.06475296996889818,
+      "grad_norm": 11.920682176684714,
+      "learning_rate": 9.968508199822018e-05,
+      "loss": 1.3331,
+      "step": 1270
+    },
+    {
+      "epoch": 0.06526283587416509,
+      "grad_norm": 6.350983744050441,
+      "learning_rate": 9.967576169610319e-05,
+      "loss": 1.3217,
+      "step": 1280
+    },
+    {
+      "epoch": 0.06577270177943201,
+      "grad_norm": 8.119062253184078,
+      "learning_rate": 9.966630592480459e-05,
+      "loss": 1.2693,
+      "step": 1290
+    },
+    {
+      "epoch": 0.06628256768469892,
+      "grad_norm": 10.81943999835483,
+      "learning_rate": 9.965671471011091e-05,
+      "loss": 1.3865,
+      "step": 1300
+    },
+    {
+      "epoch": 0.06679243358996584,
+      "grad_norm": 8.591767358572096,
+      "learning_rate": 9.964698807817805e-05,
+      "loss": 1.3059,
+      "step": 1310
+    },
+    {
+      "epoch": 0.06730229949523275,
+      "grad_norm": 6.884677751275266,
+      "learning_rate": 9.963712605553119e-05,
+      "loss": 1.3225,
+      "step": 1320
+    },
+    {
+      "epoch": 0.06781216540049967,
+      "grad_norm": 8.127541110021461,
+      "learning_rate": 9.962712866906476e-05,
+      "loss": 1.2426,
+      "step": 1330
+    },
+    {
+      "epoch": 0.06832203130576658,
+      "grad_norm": 9.497739154579854,
+      "learning_rate": 9.961699594604227e-05,
+      "loss": 1.3854,
+      "step": 1340
+    },
+    {
+      "epoch": 0.0688318972110335,
+      "grad_norm": 8.234529683565784,
+      "learning_rate": 9.96067279140964e-05,
+      "loss": 1.3541,
+      "step": 1350
+    },
+    {
+      "epoch": 0.06934176311630041,
+      "grad_norm": 14.121817799847086,
+      "learning_rate": 9.959632460122872e-05,
+      "loss": 1.288,
+      "step": 1360
+    },
+    {
+      "epoch": 0.06985162902156733,
+      "grad_norm": 8.300552320933303,
+      "learning_rate": 9.958578603580981e-05,
+      "loss": 1.2946,
+      "step": 1370
+    },
+    {
+      "epoch": 0.07036149492683424,
+      "grad_norm": 11.281514779427111,
+      "learning_rate": 9.957511224657902e-05,
+      "loss": 1.3188,
+      "step": 1380
+    },
+    {
+      "epoch": 0.07087136083210116,
+      "grad_norm": 8.401565431210939,
+      "learning_rate": 9.956430326264453e-05,
+      "loss": 1.3062,
+      "step": 1390
+    },
+    {
+      "epoch": 0.07138122673736808,
+      "grad_norm": 13.395221223404873,
+      "learning_rate": 9.955335911348316e-05,
+      "loss": 1.3546,
+      "step": 1400
+    },
+    {
+      "epoch": 0.07189109264263499,
+      "grad_norm": 9.531692368877174,
+      "learning_rate": 9.954227982894034e-05,
+      "loss": 1.2832,
+      "step": 1410
+    },
+    {
+      "epoch": 0.0724009585479019,
+      "grad_norm": 9.661709751824839,
+      "learning_rate": 9.953106543923005e-05,
+      "loss": 1.3729,
+      "step": 1420
+    },
+    {
+      "epoch": 0.07291082445316882,
+      "grad_norm": 8.059671565620953,
+      "learning_rate": 9.95197159749347e-05,
+      "loss": 1.3321,
+      "step": 1430
+    },
+    {
+      "epoch": 0.07342069035843574,
+      "grad_norm": 7.257477642996525,
+      "learning_rate": 9.950823146700502e-05,
+      "loss": 1.3132,
+      "step": 1440
+    },
+    {
+      "epoch": 0.07393055626370265,
+      "grad_norm": 13.079956490702457,
+      "learning_rate": 9.949661194676008e-05,
+      "loss": 1.2903,
+      "step": 1450
+    },
+    {
+      "epoch": 0.07444042216896957,
+      "grad_norm": 9.797908918752556,
+      "learning_rate": 9.948485744588709e-05,
+      "loss": 1.2766,
+      "step": 1460
+    },
+    {
+      "epoch": 0.07495028807423648,
+      "grad_norm": 8.8963959777427,
+      "learning_rate": 9.947296799644136e-05,
+      "loss": 1.3563,
+      "step": 1470
+    },
+    {
+      "epoch": 0.0754601539795034,
+      "grad_norm": 6.024664773794652,
+      "learning_rate": 9.946094363084623e-05,
+      "loss": 1.3051,
+      "step": 1480
+    },
+    {
+      "epoch": 0.07597001988477031,
+      "grad_norm": 15.083381217497958,
+      "learning_rate": 9.944878438189297e-05,
+      "loss": 1.2957,
+      "step": 1490
+    },
+    {
+      "epoch": 0.07647988579003721,
+      "grad_norm": 8.659028930325837,
+      "learning_rate": 9.943649028274069e-05,
+      "loss": 1.2402,
+      "step": 1500
+    },
+    {
+      "epoch": 0.07698975169530413,
+      "grad_norm": 8.483400054066859,
+      "learning_rate": 9.942406136691622e-05,
+      "loss": 1.3009,
+      "step": 1510
+    },
+    {
+      "epoch": 0.07749961760057104,
+      "grad_norm": 9.04153849731825,
+      "learning_rate": 9.941149766831408e-05,
+      "loss": 1.2704,
+      "step": 1520
+    },
+    {
+      "epoch": 0.07800948350583796,
+      "grad_norm": 6.685613160598831,
+      "learning_rate": 9.939879922119628e-05,
+      "loss": 1.2806,
+      "step": 1530
+    },
+    {
+      "epoch": 0.07851934941110487,
+      "grad_norm": 8.748668860247195,
+      "learning_rate": 9.938596606019241e-05,
+      "loss": 1.3032,
+      "step": 1540
+    },
+    {
+      "epoch": 0.07902921531637179,
+      "grad_norm": 10.516882054179344,
+      "learning_rate": 9.937299822029933e-05,
+      "loss": 1.2432,
+      "step": 1550
+    },
+    {
+      "epoch": 0.0795390812216387,
+      "grad_norm": 8.012469795551912,
+      "learning_rate": 9.935989573688124e-05,
+      "loss": 1.2907,
+      "step": 1560
+    },
+    {
+      "epoch": 0.08004894712690562,
+      "grad_norm": 7.620672875381667,
+      "learning_rate": 9.93466586456695e-05,
+      "loss": 1.2428,
+      "step": 1570
+    },
+    {
+      "epoch": 0.08055881303217254,
+      "grad_norm": 9.779429555978018,
+      "learning_rate": 9.933328698276257e-05,
+      "loss": 1.2466,
+      "step": 1580
+    },
+    {
+      "epoch": 0.08106867893743945,
+      "grad_norm": 7.3320937475403065,
+      "learning_rate": 9.931978078462587e-05,
+      "loss": 1.2906,
+      "step": 1590
+    },
+    {
+      "epoch": 0.08157854484270637,
+      "grad_norm": 6.794977736972588,
+      "learning_rate": 9.930614008809171e-05,
+      "loss": 1.2783,
+      "step": 1600
+    },
+    {
+      "epoch": 0.08208841074797328,
+      "grad_norm": 5.641899938681774,
+      "learning_rate": 9.929236493035923e-05,
+      "loss": 1.2865,
+      "step": 1610
+    },
+    {
+      "epoch": 0.0825982766532402,
+      "grad_norm": 6.976480247465172,
+      "learning_rate": 9.927845534899421e-05,
+      "loss": 1.2555,
+      "step": 1620
+    },
+    {
+      "epoch": 0.08310814255850711,
+      "grad_norm": 8.90128363726804,
+      "learning_rate": 9.9264411381929e-05,
+      "loss": 1.2581,
+      "step": 1630
+    },
+    {
+      "epoch": 0.08361800846377403,
+      "grad_norm": 8.540712581579276,
+      "learning_rate": 9.925023306746248e-05,
+      "loss": 1.2367,
+      "step": 1640
+    },
+    {
+      "epoch": 0.08412787436904094,
+      "grad_norm": 7.488693574086519,
+      "learning_rate": 9.923592044425987e-05,
+      "loss": 1.2339,
+      "step": 1650
+    },
+    {
+      "epoch": 0.08463774027430786,
+      "grad_norm": 11.310513660070532,
+      "learning_rate": 9.922147355135268e-05,
+      "loss": 1.2386,
+      "step": 1660
+    },
+    {
+      "epoch": 0.08514760617957477,
+      "grad_norm": 6.929383413556347,
+      "learning_rate": 9.920689242813855e-05,
+      "loss": 1.3621,
+      "step": 1670
+    },
+    {
+      "epoch": 0.08565747208484169,
+      "grad_norm": 6.8453184213270175,
+      "learning_rate": 9.91921771143812e-05,
+      "loss": 1.2755,
+      "step": 1680
+    },
+    {
+      "epoch": 0.0861673379901086,
+      "grad_norm": 13.353344035463461,
+      "learning_rate": 9.917732765021028e-05,
+      "loss": 1.2949,
+      "step": 1690
+    },
+    {
+      "epoch": 0.08667720389537552,
+      "grad_norm": 9.043807181846674,
+      "learning_rate": 9.916234407612132e-05,
+      "loss": 1.2238,
+      "step": 1700
+    },
+    {
+      "epoch": 0.08718706980064243,
+      "grad_norm": 7.719464446952598,
+      "learning_rate": 9.914722643297554e-05,
+      "loss": 1.2437,
+      "step": 1710
+    },
+    {
+      "epoch": 0.08769693570590935,
+      "grad_norm": 8.332761022366908,
+      "learning_rate": 9.913197476199975e-05,
+      "loss": 1.263,
+      "step": 1720
+    },
+    {
+      "epoch": 0.08820680161117626,
+      "grad_norm": 7.22775493700009,
+      "learning_rate": 9.911658910478633e-05,
+      "loss": 1.2445,
+      "step": 1730
+    },
+    {
+      "epoch": 0.08871666751644318,
+      "grad_norm": 6.704298697609803,
+      "learning_rate": 9.910106950329299e-05,
+      "loss": 1.2215,
+      "step": 1740
+    },
+    {
+      "epoch": 0.0892265334217101,
+      "grad_norm": 6.444382157095585,
+      "learning_rate": 9.908541599984276e-05,
+      "loss": 1.2642,
+      "step": 1750
+    },
+    {
+      "epoch": 0.08973639932697701,
+      "grad_norm": 8.140001307541976,
+      "learning_rate": 9.906962863712378e-05,
+      "loss": 1.3316,
+      "step": 1760
+    },
+    {
+      "epoch": 0.09024626523224392,
+      "grad_norm": 6.631859873444056,
+      "learning_rate": 9.90537074581893e-05,
+      "loss": 1.2087,
+      "step": 1770
+    },
+    {
+      "epoch": 0.09075613113751084,
+      "grad_norm": 7.069643684676236,
+      "learning_rate": 9.903765250645742e-05,
+      "loss": 1.2332,
+      "step": 1780
+    },
+    {
+      "epoch": 0.09126599704277776,
+      "grad_norm": 7.966200972241499,
+      "learning_rate": 9.90214638257111e-05,
+      "loss": 1.2684,
+      "step": 1790
+    },
+    {
+      "epoch": 0.09177586294804467,
+      "grad_norm": 13.77541921252732,
+      "learning_rate": 9.900514146009798e-05,
+      "loss": 1.336,
+      "step": 1800
+    },
+    {
+      "epoch": 0.09228572885331159,
+      "grad_norm": 7.187849329498884,
+      "learning_rate": 9.898868545413027e-05,
+      "loss": 1.1634,
+      "step": 1810
+    },
+    {
+      "epoch": 0.0927955947585785,
+      "grad_norm": 5.352581001102315,
+      "learning_rate": 9.897209585268458e-05,
+      "loss": 1.1773,
+      "step": 1820
+    },
+    {
+      "epoch": 0.0933054606638454,
+      "grad_norm": 6.0910356767398195,
+      "learning_rate": 9.895537270100194e-05,
+      "loss": 1.2267,
+      "step": 1830
+    },
+    {
+      "epoch": 0.09381532656911232,
+      "grad_norm": 7.1256730647442,
+      "learning_rate": 9.893851604468747e-05,
+      "loss": 1.2155,
+      "step": 1840
+    },
+    {
+      "epoch": 0.09432519247437923,
+      "grad_norm": 6.602537485307025,
+      "learning_rate": 9.892152592971045e-05,
+      "loss": 1.2193,
+      "step": 1850
+    },
+    {
+      "epoch": 0.09483505837964615,
+      "grad_norm": 6.912697078340405,
+      "learning_rate": 9.890440240240406e-05,
+      "loss": 1.1686,
+      "step": 1860
+    },
+    {
+      "epoch": 0.09534492428491306,
+      "grad_norm": 7.6604763356635575,
+      "learning_rate": 9.888714550946533e-05,
+      "loss": 1.2251,
+      "step": 1870
+    },
+    {
+      "epoch": 0.09585479019017998,
+      "grad_norm": 6.787843708359536,
+      "learning_rate": 9.886975529795499e-05,
+      "loss": 1.2538,
+      "step": 1880
+    },
+    {
+      "epoch": 0.0963646560954469,
+      "grad_norm": 6.779150065021871,
+      "learning_rate": 9.885223181529733e-05,
+      "loss": 1.249,
+      "step": 1890
+    },
+    {
+      "epoch": 0.09687452200071381,
+      "grad_norm": 7.372475657608788,
+      "learning_rate": 9.883457510928005e-05,
+      "loss": 1.2259,
+      "step": 1900
+    },
+    {
+      "epoch": 0.09738438790598072,
+      "grad_norm": 4.803215633000802,
+      "learning_rate": 9.881678522805422e-05,
+      "loss": 1.2417,
+      "step": 1910
+    },
+    {
+      "epoch": 0.09789425381124764,
+      "grad_norm": 9.617047316762212,
+      "learning_rate": 9.879886222013403e-05,
+      "loss": 1.1683,
+      "step": 1920
+    },
+    {
+      "epoch": 0.09840411971651455,
+      "grad_norm": 4.889685554429617,
+      "learning_rate": 9.878080613439678e-05,
+      "loss": 1.2142,
+      "step": 1930
+    },
+    {
+      "epoch": 0.09891398562178147,
+      "grad_norm": 5.759379292691569,
+      "learning_rate": 9.876261702008262e-05,
+      "loss": 1.3001,
+      "step": 1940
+    },
+    {
+      "epoch": 0.09942385152704838,
+      "grad_norm": 11.065505514698799,
+      "learning_rate": 9.874429492679451e-05,
+      "loss": 1.2414,
+      "step": 1950
+    },
+    {
+      "epoch": 0.0999337174323153,
+      "grad_norm": 8.704800666475151,
+      "learning_rate": 9.872583990449804e-05,
+      "loss": 1.3282,
+      "step": 1960
+    },
+    {
+      "epoch": 0.10044358333758222,
+      "grad_norm": 5.679982367500135,
+      "learning_rate": 9.870725200352128e-05,
+      "loss": 1.2443,
+      "step": 1970
+    },
+    {
+      "epoch": 0.10095344924284913,
+      "grad_norm": 4.795540026840549,
+      "learning_rate": 9.868853127455477e-05,
+      "loss": 1.2672,
+      "step": 1980
+    },
+    {
+      "epoch": 0.10146331514811605,
+      "grad_norm": 8.929783559827515,
+      "learning_rate": 9.866967776865115e-05,
+      "loss": 1.2542,
+      "step": 1990
+    },
+    {
+      "epoch": 0.10197318105338296,
+      "grad_norm": 7.123523092691813,
+      "learning_rate": 9.865069153722526e-05,
+      "loss": 1.2183,
+      "step": 2000
+    },
+    {
+      "epoch": 0.10248304695864988,
+      "grad_norm": 7.387728237482376,
+      "learning_rate": 9.86315726320538e-05,
+      "loss": 1.2772,
+      "step": 2010
+    },
+    {
+      "epoch": 0.10299291286391679,
+      "grad_norm": 5.901476994470112,
+      "learning_rate": 9.861232110527536e-05,
+      "loss": 1.2196,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1035027787691837,
+      "grad_norm": 5.450704849288761,
+      "learning_rate": 9.859293700939014e-05,
+      "loss": 1.2247,
+      "step": 2030
+    },
+    {
+      "epoch": 0.10401264467445062,
+      "grad_norm": 6.1351665538569815,
+      "learning_rate": 9.85734203972599e-05,
+      "loss": 1.1758,
+      "step": 2040
+    },
+    {
+      "epoch": 0.10452251057971754,
+      "grad_norm": 5.804292361232884,
+      "learning_rate": 9.855377132210776e-05,
+      "loss": 1.213,
+      "step": 2050
+    },
+    {
+      "epoch": 0.10503237648498445,
+      "grad_norm": 6.500703797902379,
+      "learning_rate": 9.853398983751809e-05,
+      "loss": 1.2294,
+      "step": 2060
+    },
+    {
+      "epoch": 0.10554224239025137,
+      "grad_norm": 6.481871800852735,
+      "learning_rate": 9.851407599743636e-05,
+      "loss": 1.224,
+      "step": 2070
+    },
+    {
+      "epoch": 0.10605210829551828,
+      "grad_norm": 8.157402438997709,
+      "learning_rate": 9.849402985616894e-05,
+      "loss": 1.2287,
+      "step": 2080
+    },
+    {
+      "epoch": 0.1065619742007852,
+      "grad_norm": 7.468026418396616,
+      "learning_rate": 9.847385146838304e-05,
+      "loss": 1.2153,
+      "step": 2090
+    },
+    {
+      "epoch": 0.10707184010605211,
+      "grad_norm": 7.796469259495127,
+      "learning_rate": 9.845354088910648e-05,
+      "loss": 1.3272,
+      "step": 2100
+    },
+    {
+      "epoch": 0.10758170601131903,
+      "grad_norm": 6.931738313538142,
+      "learning_rate": 9.843309817372761e-05,
+      "loss": 1.201,
+      "step": 2110
+    },
+    {
+      "epoch": 0.10809157191658594,
+      "grad_norm": 5.706345034804036,
+      "learning_rate": 9.84125233779951e-05,
+      "loss": 1.2303,
+      "step": 2120
+    },
+    {
+      "epoch": 0.10860143782185286,
+      "grad_norm": 7.297092633764102,
+      "learning_rate": 9.83918165580178e-05,
+      "loss": 1.2992,
+      "step": 2130
+    },
+    {
+      "epoch": 0.10911130372711977,
+      "grad_norm": 4.814644519151599,
+      "learning_rate": 9.837097777026464e-05,
+      "loss": 1.1572,
+      "step": 2140
+    },
+    {
+      "epoch": 0.10962116963238669,
+      "grad_norm": 6.083844435532459,
+      "learning_rate": 9.83500070715644e-05,
+      "loss": 1.1952,
+      "step": 2150
+    },
+    {
+      "epoch": 0.11013103553765359,
+      "grad_norm": 10.385157219191,
+      "learning_rate": 9.83289045191056e-05,
+      "loss": 1.2262,
+      "step": 2160
+    },
+    {
+      "epoch": 0.1106409014429205,
+      "grad_norm": 5.819414632123794,
+      "learning_rate": 9.830767017043635e-05,
+      "loss": 1.2793,
+      "step": 2170
+    },
+    {
+      "epoch": 0.11115076734818742,
+      "grad_norm": 5.843364388171365,
+      "learning_rate": 9.828630408346415e-05,
+      "loss": 1.2492,
+      "step": 2180
+    },
+    {
+      "epoch": 0.11166063325345434,
+      "grad_norm": 7.036045897556753,
+      "learning_rate": 9.826480631645578e-05,
+      "loss": 1.2921,
+      "step": 2190
+    },
+    {
+      "epoch": 0.11217049915872125,
+      "grad_norm": 11.830997847111623,
+      "learning_rate": 9.824317692803711e-05,
+      "loss": 1.2078,
+      "step": 2200
+    },
+    {
+      "epoch": 0.11268036506398817,
+      "grad_norm": 8.258449793970065,
+      "learning_rate": 9.822141597719294e-05,
+      "loss": 1.1975,
+      "step": 2210
+    },
+    {
+      "epoch": 0.11319023096925508,
+      "grad_norm": 7.478565780074843,
+      "learning_rate": 9.819952352326689e-05,
+      "loss": 1.2081,
+      "step": 2220
+    },
+    {
+      "epoch": 0.113700096874522,
+      "grad_norm": 8.192782392249146,
+      "learning_rate": 9.817749962596115e-05,
+      "loss": 1.2213,
+      "step": 2230
+    },
+    {
+      "epoch": 0.11420996277978891,
+      "grad_norm": 8.384820102991455,
+      "learning_rate": 9.81553443453364e-05,
+      "loss": 1.2237,
+      "step": 2240
+    },
+    {
+      "epoch": 0.11471982868505583,
+      "grad_norm": 5.245831823269501,
+      "learning_rate": 9.81330577418116e-05,
+      "loss": 1.1818,
+      "step": 2250
+    },
+    {
+      "epoch": 0.11522969459032274,
+      "grad_norm": 4.777949798758605,
+      "learning_rate": 9.811063987616381e-05,
+      "loss": 1.1329,
+      "step": 2260
+    },
+    {
+      "epoch": 0.11573956049558966,
+      "grad_norm": 7.1330868631591295,
+      "learning_rate": 9.80880908095281e-05,
+      "loss": 1.2497,
+      "step": 2270
+    },
+    {
+      "epoch": 0.11624942640085657,
+      "grad_norm": 6.643681935564405,
+      "learning_rate": 9.806541060339729e-05,
+      "loss": 1.2394,
+      "step": 2280
+    },
+    {
+      "epoch": 0.11675929230612349,
+      "grad_norm": 5.145381571255945,
+      "learning_rate": 9.804259931962188e-05,
+      "loss": 1.187,
+      "step": 2290
+    },
+    {
+      "epoch": 0.1172691582113904,
+      "grad_norm": 10.41803650650462,
+      "learning_rate": 9.801965702040974e-05,
+      "loss": 1.245,
+      "step": 2300
+    },
+    {
+      "epoch": 0.11777902411665732,
+      "grad_norm": 5.4345422498522,
+      "learning_rate": 9.799658376832612e-05,
+      "loss": 1.2012,
+      "step": 2310
+    },
+    {
+      "epoch": 0.11828889002192423,
+      "grad_norm": 5.334062865929928,
+      "learning_rate": 9.797337962629334e-05,
+      "loss": 1.1432,
+      "step": 2320
+    },
+    {
+      "epoch": 0.11879875592719115,
+      "grad_norm": 6.099120103735033,
+      "learning_rate": 9.795004465759065e-05,
+      "loss": 1.2238,
+      "step": 2330
+    },
+    {
+      "epoch": 0.11930862183245806,
+      "grad_norm": 7.914833667111485,
+      "learning_rate": 9.792657892585413e-05,
+      "loss": 1.1527,
+      "step": 2340
+    },
+    {
+      "epoch": 0.11981848773772498,
+      "grad_norm": 5.197372483584429,
+      "learning_rate": 9.790298249507641e-05,
+      "loss": 1.1935,
+      "step": 2350
+    },
+    {
+      "epoch": 0.1203283536429919,
+      "grad_norm": 6.124732365741253,
+      "learning_rate": 9.787925542960657e-05,
+      "loss": 1.2419,
+      "step": 2360
+    },
+    {
+      "epoch": 0.12083821954825881,
+      "grad_norm": 4.383681745236697,
+      "learning_rate": 9.785539779414993e-05,
+      "loss": 1.2306,
+      "step": 2370
+    },
+    {
+      "epoch": 0.12134808545352573,
+      "grad_norm": 5.815760094144063,
+      "learning_rate": 9.783140965376789e-05,
+      "loss": 1.1445,
+      "step": 2380
+    },
+    {
+      "epoch": 0.12185795135879264,
+      "grad_norm": 5.741541468529939,
+      "learning_rate": 9.780729107387772e-05,
+      "loss": 1.1961,
+      "step": 2390
+    },
+    {
+      "epoch": 0.12236781726405956,
+      "grad_norm": 6.030456606909413,
+      "learning_rate": 9.778304212025244e-05,
+      "loss": 1.1887,
+      "step": 2400
+    },
+    {
+      "epoch": 0.12287768316932647,
+      "grad_norm": 5.911071102496239,
+      "learning_rate": 9.77586628590206e-05,
+      "loss": 1.1322,
+      "step": 2410
+    },
+    {
+      "epoch": 0.12338754907459339,
+      "grad_norm": 5.420205050934226,
+      "learning_rate": 9.773415335666611e-05,
+      "loss": 1.1774,
+      "step": 2420
+    },
+    {
+      "epoch": 0.1238974149798603,
+      "grad_norm": 6.720049848717211,
+      "learning_rate": 9.770951368002803e-05,
+      "loss": 1.1409,
+      "step": 2430
+    },
+    {
+      "epoch": 0.12440728088512722,
+      "grad_norm": 9.256546417665346,
+      "learning_rate": 9.768474389630046e-05,
+      "loss": 1.18,
+      "step": 2440
+    },
+    {
+      "epoch": 0.12491714679039413,
+      "grad_norm": 7.111505210888634,
+      "learning_rate": 9.765984407303226e-05,
+      "loss": 1.2464,
+      "step": 2450
+    },
+    {
+      "epoch": 0.12542701269566103,
+      "grad_norm": 4.694572996739052,
+      "learning_rate": 9.763481427812696e-05,
+      "loss": 1.2375,
+      "step": 2460
+    },
+    {
+      "epoch": 0.12593687860092795,
+      "grad_norm": 4.4336348668679975,
+      "learning_rate": 9.76096545798425e-05,
+      "loss": 1.3042,
+      "step": 2470
+    },
+    {
+      "epoch": 0.12644674450619486,
+      "grad_norm": 4.625989023877044,
+      "learning_rate": 9.75843650467911e-05,
+      "loss": 1.1504,
+      "step": 2480
+    },
+    {
+      "epoch": 0.12695661041146178,
+      "grad_norm": 7.90799206426825,
+      "learning_rate": 9.755894574793905e-05,
+      "loss": 1.1228,
+      "step": 2490
+    },
+    {
+      "epoch": 0.1274664763167287,
+      "grad_norm": 6.704492006114675,
+      "learning_rate": 9.753339675260647e-05,
+      "loss": 1.1919,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1279763422219956,
+      "grad_norm": 6.942236866139156,
+      "learning_rate": 9.750771813046722e-05,
+      "loss": 1.2088,
+      "step": 2510
+    },
+    {
+      "epoch": 0.12848620812726252,
+      "grad_norm": 4.322875961018973,
+      "learning_rate": 9.748190995154868e-05,
+      "loss": 1.1658,
+      "step": 2520
+    },
+    {
+      "epoch": 0.12899607403252944,
+      "grad_norm": 6.218605296975476,
+      "learning_rate": 9.745597228623147e-05,
+      "loss": 1.1418,
+      "step": 2530
+    },
+    {
+      "epoch": 0.12950593993779635,
+      "grad_norm": 5.6305980267618185,
+      "learning_rate": 9.74299052052494e-05,
+      "loss": 1.1694,
+      "step": 2540
+    },
+    {
+      "epoch": 0.13001580584306327,
+      "grad_norm": 10.313836811735246,
+      "learning_rate": 9.740370877968916e-05,
+      "loss": 1.1807,
+      "step": 2550
+    },
+    {
+      "epoch": 0.13052567174833019,
+      "grad_norm": 7.647262183219847,
+      "learning_rate": 9.737738308099017e-05,
+      "loss": 1.1667,
+      "step": 2560
+    },
+    {
+      "epoch": 0.1310355376535971,
+      "grad_norm": 5.2019357519805425,
+      "learning_rate": 9.735092818094442e-05,
+      "loss": 1.227,
+      "step": 2570
+    },
+    {
+      "epoch": 0.13154540355886402,
+      "grad_norm": 5.765131994822294,
+      "learning_rate": 9.732434415169623e-05,
+      "loss": 1.1175,
+      "step": 2580
+    },
+    {
+      "epoch": 0.13205526946413093,
+      "grad_norm": 4.071134055352691,
+      "learning_rate": 9.729763106574202e-05,
+      "loss": 1.0949,
+      "step": 2590
+    },
+    {
+      "epoch": 0.13256513536939785,
+      "grad_norm": 7.880998877266062,
+      "learning_rate": 9.727078899593023e-05,
+      "loss": 1.165,
+      "step": 2600
+    },
+    {
+      "epoch": 0.13307500127466476,
+      "grad_norm": 8.522815149585488,
+      "learning_rate": 9.7243818015461e-05,
+      "loss": 1.1394,
+      "step": 2610
+    },
+    {
+      "epoch": 0.13358486717993168,
+      "grad_norm": 4.210313288723103,
+      "learning_rate": 9.721671819788602e-05,
+      "loss": 1.1522,
+      "step": 2620
+    },
+    {
+      "epoch": 0.1340947330851986,
+      "grad_norm": 7.169891677668386,
+      "learning_rate": 9.718948961710834e-05,
+      "loss": 1.1513,
+      "step": 2630
+    },
+    {
+      "epoch": 0.1346045989904655,
+      "grad_norm": 11.212825964523153,
+      "learning_rate": 9.716213234738215e-05,
+      "loss": 1.1749,
+      "step": 2640
+    },
+    {
+      "epoch": 0.13511446489573242,
+      "grad_norm": 6.8565562989741355,
+      "learning_rate": 9.713464646331259e-05,
+      "loss": 1.2062,
+      "step": 2650
+    },
+    {
+      "epoch": 0.13562433080099934,
+      "grad_norm": 8.87197126878053,
+      "learning_rate": 9.710703203985554e-05,
+      "loss": 1.1817,
+      "step": 2660
+    },
+    {
+      "epoch": 0.13613419670626625,
+      "grad_norm": 5.93024626159216,
+      "learning_rate": 9.707928915231742e-05,
+      "loss": 1.1203,
+      "step": 2670
+    },
+    {
+      "epoch": 0.13664406261153317,
+      "grad_norm": 5.249617830158715,
+      "learning_rate": 9.705141787635493e-05,
+      "loss": 1.1743,
+      "step": 2680
+    },
+    {
+      "epoch": 0.13715392851680008,
+      "grad_norm": 4.986172758637716,
+      "learning_rate": 9.702341828797497e-05,
+      "loss": 1.1499,
+      "step": 2690
+    },
+    {
+      "epoch": 0.137663794422067,
+      "grad_norm": 8.106695311484067,
+      "learning_rate": 9.699529046353433e-05,
+      "loss": 1.2356,
+      "step": 2700
+    },
+    {
+      "epoch": 0.1381736603273339,
+      "grad_norm": 5.122080533809294,
+      "learning_rate": 9.696703447973947e-05,
+      "loss": 1.2141,
+      "step": 2710
+    },
+    {
+      "epoch": 0.13868352623260083,
+      "grad_norm": 4.891418222378375,
+      "learning_rate": 9.69386504136464e-05,
+      "loss": 1.1277,
+      "step": 2720
+    },
+    {
+      "epoch": 0.13919339213786774,
+      "grad_norm": 5.552791497724159,
+      "learning_rate": 9.691013834266041e-05,
+      "loss": 1.1838,
+      "step": 2730
+    },
+    {
+      "epoch": 0.13970325804313466,
+      "grad_norm": 6.05967674916367,
+      "learning_rate": 9.688149834453584e-05,
+      "loss": 1.1281,
+      "step": 2740
+    },
+    {
+      "epoch": 0.14021312394840157,
+      "grad_norm": 7.952542574514654,
+      "learning_rate": 9.685273049737591e-05,
+      "loss": 1.1558,
+      "step": 2750
+    },
+    {
+      "epoch": 0.1407229898536685,
+      "grad_norm": 5.6955985155690945,
+      "learning_rate": 9.682383487963249e-05,
+      "loss": 1.1043,
+      "step": 2760
+    },
+    {
+      "epoch": 0.1412328557589354,
+      "grad_norm": 6.693711350896108,
+      "learning_rate": 9.679481157010593e-05,
+      "loss": 1.1673,
+      "step": 2770
+    },
+    {
+      "epoch": 0.14174272166420232,
+      "grad_norm": 5.052952434134367,
+      "learning_rate": 9.676566064794473e-05,
+      "loss": 1.21,
+      "step": 2780
+    },
+    {
+      "epoch": 0.14225258756946924,
+      "grad_norm": 4.609882531018443,
+      "learning_rate": 9.673638219264547e-05,
+      "loss": 1.1938,
+      "step": 2790
+    },
+    {
+      "epoch": 0.14276245347473615,
+      "grad_norm": 7.234811448945305,
+      "learning_rate": 9.670697628405248e-05,
+      "loss": 1.1709,
+      "step": 2800
+    },
+    {
+      "epoch": 0.14327231938000307,
+      "grad_norm": 5.177929762439721,
+      "learning_rate": 9.667744300235765e-05,
+      "loss": 1.2318,
+      "step": 2810
+    },
+    {
+      "epoch": 0.14378218528526998,
+      "grad_norm": 4.736379267462622,
+      "learning_rate": 9.664778242810027e-05,
+      "loss": 1.1854,
+      "step": 2820
+    },
+    {
+      "epoch": 0.1442920511905369,
+      "grad_norm": 5.465873014267871,
+      "learning_rate": 9.661799464216674e-05,
+      "loss": 1.1519,
+      "step": 2830
+    },
+    {
+      "epoch": 0.1448019170958038,
+      "grad_norm": 4.8161346948916774,
+      "learning_rate": 9.658807972579039e-05,
+      "loss": 1.1829,
+      "step": 2840
+    },
+    {
+      "epoch": 0.14531178300107073,
+      "grad_norm": 4.313950922209564,
+      "learning_rate": 9.655803776055122e-05,
+      "loss": 1.1124,
+      "step": 2850
+    },
+    {
+      "epoch": 0.14582164890633764,
+      "grad_norm": 7.772690723847887,
+      "learning_rate": 9.65278688283757e-05,
+      "loss": 1.2594,
+      "step": 2860
+    },
+    {
+      "epoch": 0.14633151481160456,
+      "grad_norm": 6.400952780656834,
+      "learning_rate": 9.649757301153658e-05,
+      "loss": 1.1976,
+      "step": 2870
+    },
+    {
+      "epoch": 0.14684138071687147,
+      "grad_norm": 4.26178174000709,
+      "learning_rate": 9.646715039265262e-05,
+      "loss": 1.1369,
+      "step": 2880
+    },
+    {
+      "epoch": 0.1473512466221384,
+      "grad_norm": 8.065544184447527,
+      "learning_rate": 9.643660105468834e-05,
+      "loss": 1.1366,
+      "step": 2890
+    },
+    {
+      "epoch": 0.1478611125274053,
+      "grad_norm": 4.624654737377227,
+      "learning_rate": 9.640592508095389e-05,
+      "loss": 1.1888,
+      "step": 2900
+    },
+    {
+      "epoch": 0.14837097843267222,
+      "grad_norm": 6.4881922231671,
+      "learning_rate": 9.637512255510475e-05,
+      "loss": 1.2,
+      "step": 2910
+    },
+    {
+      "epoch": 0.14888084433793913,
+      "grad_norm": 7.571117972742807,
+      "learning_rate": 9.634419356114146e-05,
+      "loss": 1.1813,
+      "step": 2920
+    },
+    {
+      "epoch": 0.14939071024320605,
+      "grad_norm": 4.0399323719696,
+      "learning_rate": 9.631313818340951e-05,
+      "loss": 1.108,
+      "step": 2930
+    },
+    {
+      "epoch": 0.14990057614847296,
+      "grad_norm": 6.80892247075497,
+      "learning_rate": 9.628195650659905e-05,
+      "loss": 1.1393,
+      "step": 2940
+    },
+    {
+      "epoch": 0.15041044205373988,
+      "grad_norm": 6.6413351681887605,
+      "learning_rate": 9.62506486157446e-05,
+      "loss": 1.1306,
+      "step": 2950
+    },
+    {
+      "epoch": 0.1509203079590068,
+      "grad_norm": 4.866599278982188,
+      "learning_rate": 9.621921459622494e-05,
+      "loss": 1.1209,
+      "step": 2960
+    },
+    {
+      "epoch": 0.1514301738642737,
+      "grad_norm": 5.195315531928738,
+      "learning_rate": 9.618765453376275e-05,
+      "loss": 1.1449,
+      "step": 2970
+    },
+    {
+      "epoch": 0.15194003976954062,
+      "grad_norm": 5.985983405099398,
+      "learning_rate": 9.615596851442448e-05,
+      "loss": 1.151,
+      "step": 2980
+    },
+    {
+      "epoch": 0.1524499056748075,
+      "grad_norm": 5.216800512033471,
+      "learning_rate": 9.612415662462007e-05,
+      "loss": 1.1354,
+      "step": 2990
+    },
+    {
+      "epoch": 0.15295977158007443,
+      "grad_norm": 4.485321159625428,
+      "learning_rate": 9.609221895110269e-05,
+      "loss": 1.132,
+      "step": 3000
+    },
+    {
+      "epoch": 0.15346963748534134,
+      "grad_norm": 4.795030590130624,
+      "learning_rate": 9.606015558096857e-05,
+      "loss": 1.0816,
+      "step": 3010
+    },
+    {
+      "epoch": 0.15397950339060826,
+      "grad_norm": 8.191449302146102,
+      "learning_rate": 9.602796660165667e-05,
+      "loss": 1.1905,
+      "step": 3020
+    },
+    {
+      "epoch": 0.15448936929587517,
+      "grad_norm": 3.277563070999204,
+      "learning_rate": 9.599565210094856e-05,
+      "loss": 1.1425,
+      "step": 3030
+    },
+    {
+      "epoch": 0.1549992352011421,
+      "grad_norm": 4.295030375268278,
+      "learning_rate": 9.596321216696807e-05,
+      "loss": 1.184,
+      "step": 3040
+    },
+    {
+      "epoch": 0.155509101106409,
+      "grad_norm": 4.705769445658949,
+      "learning_rate": 9.59306468881811e-05,
+      "loss": 1.1924,
+      "step": 3050
+    },
+    {
+      "epoch": 0.15601896701167592,
+      "grad_norm": 4.185085976052072,
+      "learning_rate": 9.589795635339538e-05,
+      "loss": 1.1721,
+      "step": 3060
+    },
+    {
+      "epoch": 0.15652883291694283,
+      "grad_norm": 4.361491323328995,
+      "learning_rate": 9.586514065176022e-05,
+      "loss": 1.1392,
+      "step": 3070
+    },
+    {
+      "epoch": 0.15703869882220975,
+      "grad_norm": 4.528614562993759,
+      "learning_rate": 9.583219987276629e-05,
+      "loss": 1.094,
+      "step": 3080
+    },
+    {
+      "epoch": 0.15754856472747666,
+      "grad_norm": 4.961731041882474,
+      "learning_rate": 9.579913410624532e-05,
+      "loss": 1.2076,
+      "step": 3090
+    },
+    {
+      "epoch": 0.15805843063274358,
+      "grad_norm": 4.757780522134779,
+      "learning_rate": 9.576594344236991e-05,
+      "loss": 1.1451,
+      "step": 3100
+    },
+    {
+      "epoch": 0.1585682965380105,
+      "grad_norm": 6.711264664750607,
+      "learning_rate": 9.573262797165326e-05,
+      "loss": 1.0809,
+      "step": 3110
+    },
+    {
+      "epoch": 0.1590781624432774,
+      "grad_norm": 5.286703097650971,
+      "learning_rate": 9.56991877849489e-05,
+      "loss": 1.1819,
+      "step": 3120
+    },
+    {
+      "epoch": 0.15958802834854433,
+      "grad_norm": 5.540430875450444,
+      "learning_rate": 9.566562297345054e-05,
+      "loss": 1.2123,
+      "step": 3130
+    },
+    {
+      "epoch": 0.16009789425381124,
+      "grad_norm": 5.935180682491463,
+      "learning_rate": 9.563193362869167e-05,
+      "loss": 1.1159,
+      "step": 3140
+    },
+    {
+      "epoch": 0.16060776015907816,
+      "grad_norm": 6.423069613572788,
+      "learning_rate": 9.559811984254545e-05,
+      "loss": 1.1342,
+      "step": 3150
+    },
+    {
+      "epoch": 0.16111762606434507,
+      "grad_norm": 6.070748932230466,
+      "learning_rate": 9.556418170722435e-05,
+      "loss": 1.2179,
+      "step": 3160
+    },
+    {
+      "epoch": 0.16162749196961199,
+      "grad_norm": 4.35911724617304,
+      "learning_rate": 9.553011931527998e-05,
+      "loss": 1.083,
+      "step": 3170
+    },
+    {
+      "epoch": 0.1621373578748789,
+      "grad_norm": 3.7577113585154183,
+      "learning_rate": 9.54959327596028e-05,
+      "loss": 1.2139,
+      "step": 3180
+    },
+    {
+      "epoch": 0.16264722378014582,
+      "grad_norm": 4.7730345910622,
+      "learning_rate": 9.546162213342187e-05,
+      "loss": 1.1606,
+      "step": 3190
+    },
+    {
+      "epoch": 0.16315708968541273,
+      "grad_norm": 7.156654490620455,
+      "learning_rate": 9.542718753030463e-05,
+      "loss": 1.1637,
+      "step": 3200
+    },
+    {
+      "epoch": 0.16366695559067965,
+      "grad_norm": 5.566030323225834,
+      "learning_rate": 9.539262904415655e-05,
+      "loss": 1.1034,
+      "step": 3210
+    },
+    {
+      "epoch": 0.16417682149594656,
+      "grad_norm": 8.746375921889477,
+      "learning_rate": 9.5357946769221e-05,
+      "loss": 1.1711,
+      "step": 3220
+    },
+    {
+      "epoch": 0.16468668740121348,
+      "grad_norm": 4.325955014504265,
+      "learning_rate": 9.532314080007888e-05,
+      "loss": 1.0815,
+      "step": 3230
+    },
+    {
+      "epoch": 0.1651965533064804,
+      "grad_norm": 4.771056313254683,
+      "learning_rate": 9.528821123164845e-05,
+      "loss": 1.0926,
+      "step": 3240
+    },
+    {
+      "epoch": 0.1657064192117473,
+      "grad_norm": 8.066446506417845,
+      "learning_rate": 9.525315815918501e-05,
+      "loss": 1.0949,
+      "step": 3250
+    },
+    {
+      "epoch": 0.16621628511701422,
+      "grad_norm": 5.073637609519421,
+      "learning_rate": 9.521798167828069e-05,
+      "loss": 1.1235,
+      "step": 3260
+    },
+    {
+      "epoch": 0.16672615102228114,
+      "grad_norm": 4.7462509373604975,
+      "learning_rate": 9.518268188486412e-05,
+      "loss": 1.1382,
+      "step": 3270
+    },
+    {
+      "epoch": 0.16723601692754805,
+      "grad_norm": 5.252783379643778,
+      "learning_rate": 9.514725887520026e-05,
+      "loss": 1.0399,
+      "step": 3280
+    },
+    {
+      "epoch": 0.16774588283281497,
+      "grad_norm": 3.9308583855377264,
+      "learning_rate": 9.511171274589006e-05,
+      "loss": 1.1328,
+      "step": 3290
+    },
+    {
+      "epoch": 0.16825574873808188,
+      "grad_norm": 4.791609600754416,
+      "learning_rate": 9.507604359387024e-05,
+      "loss": 1.1363,
+      "step": 3300
+    },
+    {
+      "epoch": 0.1687656146433488,
+      "grad_norm": 4.468070449560343,
+      "learning_rate": 9.504025151641299e-05,
+      "loss": 1.159,
+      "step": 3310
+    },
+    {
+      "epoch": 0.16927548054861571,
+      "grad_norm": 5.861341903814554,
+      "learning_rate": 9.500433661112575e-05,
+      "loss": 1.1007,
+      "step": 3320
+    },
+    {
+      "epoch": 0.16978534645388263,
+      "grad_norm": 4.483499400910527,
+      "learning_rate": 9.496829897595089e-05,
+      "loss": 1.1221,
+      "step": 3330
+    },
+    {
+      "epoch": 0.17029521235914954,
+      "grad_norm": 5.074741849785658,
+      "learning_rate": 9.493213870916552e-05,
+      "loss": 1.1113,
+      "step": 3340
+    },
+    {
+      "epoch": 0.17080507826441646,
+      "grad_norm": 3.9786022995626213,
+      "learning_rate": 9.489585590938116e-05,
+      "loss": 1.1238,
+      "step": 3350
+    },
+    {
+      "epoch": 0.17131494416968338,
+      "grad_norm": 5.21394443419945,
+      "learning_rate": 9.485945067554344e-05,
+      "loss": 1.1357,
+      "step": 3360
+    },
+    {
+      "epoch": 0.1718248100749503,
+      "grad_norm": 4.142795672158372,
+      "learning_rate": 9.482292310693191e-05,
+      "loss": 1.1089,
+      "step": 3370
+    },
+    {
+      "epoch": 0.1723346759802172,
+      "grad_norm": 3.8294513054176362,
+      "learning_rate": 9.478627330315975e-05,
+      "loss": 1.1337,
+      "step": 3380
+    },
+    {
+      "epoch": 0.17284454188548412,
+      "grad_norm": 3.1799263389534675,
+      "learning_rate": 9.474950136417347e-05,
+      "loss": 1.1452,
+      "step": 3390
+    },
+    {
+      "epoch": 0.17335440779075104,
+      "grad_norm": 4.8079087991100735,
+      "learning_rate": 9.471260739025263e-05,
+      "loss": 1.1363,
+      "step": 3400
+    },
+    {
+      "epoch": 0.17386427369601795,
+      "grad_norm": 4.781900661326968,
+      "learning_rate": 9.467559148200964e-05,
+      "loss": 1.0781,
+      "step": 3410
+    },
+    {
+      "epoch": 0.17437413960128487,
+      "grad_norm": 3.994191358886494,
+      "learning_rate": 9.463845374038935e-05,
+      "loss": 1.1218,
+      "step": 3420
+    },
+    {
+      "epoch": 0.17488400550655178,
+      "grad_norm": 4.75943478869446,
+      "learning_rate": 9.460119426666895e-05,
+      "loss": 1.1882,
+      "step": 3430
+    },
+    {
+      "epoch": 0.1753938714118187,
+      "grad_norm": 5.6089661449576305,
+      "learning_rate": 9.456381316245756e-05,
+      "loss": 1.0581,
+      "step": 3440
+    },
+    {
+      "epoch": 0.1759037373170856,
+      "grad_norm": 7.583108949043354,
+      "learning_rate": 9.452631052969597e-05,
+      "loss": 1.0792,
+      "step": 3450
+    },
+    {
+      "epoch": 0.17641360322235253,
+      "grad_norm": 6.1835180845992594,
+      "learning_rate": 9.448868647065642e-05,
+      "loss": 1.0849,
+      "step": 3460
+    },
+    {
+      "epoch": 0.17692346912761944,
+      "grad_norm": 4.488568933309487,
+      "learning_rate": 9.445094108794228e-05,
+      "loss": 1.1,
+      "step": 3470
+    },
+    {
+      "epoch": 0.17743333503288636,
+      "grad_norm": 6.296838767135793,
+      "learning_rate": 9.441307448448778e-05,
+      "loss": 1.1232,
+      "step": 3480
+    },
+    {
+      "epoch": 0.17794320093815327,
+      "grad_norm": 6.106789989120955,
+      "learning_rate": 9.437508676355773e-05,
+      "loss": 1.0666,
+      "step": 3490
+    },
+    {
+      "epoch": 0.1784530668434202,
+      "grad_norm": 4.830440619495212,
+      "learning_rate": 9.433697802874721e-05,
+      "loss": 1.136,
+      "step": 3500
+    },
+    {
+      "epoch": 0.1789629327486871,
+      "grad_norm": 4.582083565033262,
+      "learning_rate": 9.429874838398134e-05,
+      "loss": 1.1087,
+      "step": 3510
+    },
+    {
+      "epoch": 0.17947279865395402,
+      "grad_norm": 4.4505207891650285,
+      "learning_rate": 9.426039793351498e-05,
+      "loss": 1.2229,
+      "step": 3520
+    },
+    {
+      "epoch": 0.17998266455922093,
+      "grad_norm": 4.073833569670933,
+      "learning_rate": 9.422192678193238e-05,
+      "loss": 1.1443,
+      "step": 3530
+    },
+    {
+      "epoch": 0.18049253046448785,
+      "grad_norm": 6.174973106473309,
+      "learning_rate": 9.418333503414701e-05,
+      "loss": 1.1453,
+      "step": 3540
+    },
+    {
+      "epoch": 0.18100239636975476,
+      "grad_norm": 6.57859848265449,
+      "learning_rate": 9.41446227954012e-05,
+      "loss": 1.1038,
+      "step": 3550
+    },
+    {
+      "epoch": 0.18151226227502168,
+      "grad_norm": 8.479133870904418,
+      "learning_rate": 9.410579017126586e-05,
+      "loss": 1.1265,
+      "step": 3560
+    },
+    {
+      "epoch": 0.1820221281802886,
+      "grad_norm": 4.42253230283525,
+      "learning_rate": 9.406683726764018e-05,
+      "loss": 1.1098,
+      "step": 3570
+    },
+    {
+      "epoch": 0.1825319940855555,
+      "grad_norm": 3.829827815809406,
+      "learning_rate": 9.402776419075138e-05,
+      "loss": 1.1131,
+      "step": 3580
+    },
+    {
+      "epoch": 0.18304185999082243,
+      "grad_norm": 7.4706313494248375,
+      "learning_rate": 9.398857104715441e-05,
+      "loss": 1.1661,
+      "step": 3590
+    },
+    {
+      "epoch": 0.18355172589608934,
+      "grad_norm": 5.157105013968114,
+      "learning_rate": 9.394925794373162e-05,
+      "loss": 1.0779,
+      "step": 3600
+    },
+    {
+      "epoch": 0.18406159180135626,
+      "grad_norm": 5.7160746841889924,
+      "learning_rate": 9.390982498769254e-05,
+      "loss": 1.1154,
+      "step": 3610
+    },
+    {
+      "epoch": 0.18457145770662317,
+      "grad_norm": 4.263780490016166,
+      "learning_rate": 9.387027228657352e-05,
+      "loss": 1.1213,
+      "step": 3620
+    },
+    {
+      "epoch": 0.1850813236118901,
+      "grad_norm": 5.974524640438817,
+      "learning_rate": 9.383059994823747e-05,
+      "loss": 1.1598,
+      "step": 3630
+    },
+    {
+      "epoch": 0.185591189517157,
+      "grad_norm": 3.4619650322325444,
+      "learning_rate": 9.379080808087353e-05,
+      "loss": 1.1139,
+      "step": 3640
+    },
+    {
+      "epoch": 0.1861010554224239,
+      "grad_norm": 4.944482287429536,
+      "learning_rate": 9.375089679299684e-05,
+      "loss": 1.1143,
+      "step": 3650
+    },
+    {
+      "epoch": 0.1866109213276908,
+      "grad_norm": 4.173778226175053,
+      "learning_rate": 9.371086619344822e-05,
+      "loss": 1.0866,
+      "step": 3660
+    },
+    {
+      "epoch": 0.18712078723295772,
+      "grad_norm": 5.4238742287923705,
+      "learning_rate": 9.367071639139381e-05,
+      "loss": 1.2034,
+      "step": 3670
+    },
+    {
+      "epoch": 0.18763065313822463,
+      "grad_norm": 3.7178235508537174,
+      "learning_rate": 9.363044749632488e-05,
+      "loss": 1.0685,
+      "step": 3680
+    },
+    {
+      "epoch": 0.18814051904349155,
+      "grad_norm": 7.491473532606374,
+      "learning_rate": 9.359005961805743e-05,
+      "loss": 1.0924,
+      "step": 3690
+    },
+    {
+      "epoch": 0.18865038494875846,
+      "grad_norm": 4.245239185493443,
+      "learning_rate": 9.354955286673192e-05,
+      "loss": 1.0636,
+      "step": 3700
+    },
+    {
+      "epoch": 0.18916025085402538,
+      "grad_norm": 5.707722548518677,
+      "learning_rate": 9.350892735281308e-05,
+      "loss": 1.113,
+      "step": 3710
+    },
+    {
+      "epoch": 0.1896701167592923,
+      "grad_norm": 6.702902819523267,
+      "learning_rate": 9.34681831870894e-05,
+      "loss": 1.1233,
+      "step": 3720
+    },
+    {
+      "epoch": 0.1901799826645592,
+      "grad_norm": 5.440354574023026,
+      "learning_rate": 9.342732048067302e-05,
+      "loss": 1.0976,
+      "step": 3730
+    },
+    {
+      "epoch": 0.19068984856982613,
+      "grad_norm": 5.034338134908323,
+      "learning_rate": 9.338633934499932e-05,
+      "loss": 1.1727,
+      "step": 3740
+    },
+    {
+      "epoch": 0.19119971447509304,
+      "grad_norm": 5.32930802456178,
+      "learning_rate": 9.334523989182661e-05,
+      "loss": 1.0973,
+      "step": 3750
+    },
+    {
+      "epoch": 0.19170958038035996,
+      "grad_norm": 4.523880335546063,
+      "learning_rate": 9.330402223323595e-05,
+      "loss": 1.1695,
+      "step": 3760
+    },
+    {
+      "epoch": 0.19221944628562687,
+      "grad_norm": 3.9403859738446796,
+      "learning_rate": 9.326268648163064e-05,
+      "loss": 1.1134,
+      "step": 3770
+    },
+    {
+      "epoch": 0.1927293121908938,
+      "grad_norm": 3.3662823656403495,
+      "learning_rate": 9.322123274973613e-05,
+      "loss": 1.1634,
+      "step": 3780
+    },
+    {
+      "epoch": 0.1932391780961607,
+      "grad_norm": 3.029874138789195,
+      "learning_rate": 9.317966115059956e-05,
+      "loss": 1.0194,
+      "step": 3790
+    },
+    {
+      "epoch": 0.19374904400142762,
+      "grad_norm": 4.226741304538432,
+      "learning_rate": 9.31379717975895e-05,
+      "loss": 1.0669,
+      "step": 3800
+    },
+    {
+      "epoch": 0.19425890990669453,
+      "grad_norm": 3.4250576879529993,
+      "learning_rate": 9.309616480439564e-05,
+      "loss": 1.1222,
+      "step": 3810
+    },
+    {
+      "epoch": 0.19476877581196145,
+      "grad_norm": 4.618185836785933,
+      "learning_rate": 9.305424028502852e-05,
+      "loss": 1.0713,
+      "step": 3820
+    },
+    {
+      "epoch": 0.19527864171722836,
+      "grad_norm": 3.3429157735662294,
+      "learning_rate": 9.301219835381914e-05,
+      "loss": 1.0282,
+      "step": 3830
+    },
+    {
+      "epoch": 0.19578850762249528,
+      "grad_norm": 4.15707318845732,
+      "learning_rate": 9.297003912541871e-05,
+      "loss": 1.1,
+      "step": 3840
+    },
+    {
+      "epoch": 0.1962983735277622,
+      "grad_norm": 4.657258148198954,
+      "learning_rate": 9.29277627147983e-05,
+      "loss": 1.0841,
+      "step": 3850
+    },
+    {
+      "epoch": 0.1968082394330291,
+      "grad_norm": 4.056316857136649,
+      "learning_rate": 9.28853692372486e-05,
+      "loss": 1.0607,
+      "step": 3860
+    },
+    {
+      "epoch": 0.19731810533829602,
+      "grad_norm": 3.580812269523134,
+      "learning_rate": 9.284285880837946e-05,
+      "loss": 1.1802,
+      "step": 3870
+    },
+    {
+      "epoch": 0.19782797124356294,
+      "grad_norm": 4.750882133120819,
+      "learning_rate": 9.280023154411974e-05,
+      "loss": 1.1562,
+      "step": 3880
+    },
+    {
+      "epoch": 0.19833783714882985,
+      "grad_norm": 4.351795862653647,
+      "learning_rate": 9.275748756071689e-05,
+      "loss": 1.1345,
+      "step": 3890
+    },
+    {
+      "epoch": 0.19884770305409677,
+      "grad_norm": 4.370634001939076,
+      "learning_rate": 9.271462697473665e-05,
+      "loss": 1.1158,
+      "step": 3900
+    },
+    {
+      "epoch": 0.19935756895936368,
+      "grad_norm": 4.593193453141612,
+      "learning_rate": 9.267164990306274e-05,
+      "loss": 1.1004,
+      "step": 3910
+    },
+    {
+      "epoch": 0.1998674348646306,
+      "grad_norm": 5.0217760322668035,
+      "learning_rate": 9.262855646289659e-05,
+      "loss": 1.0957,
+      "step": 3920
+    },
+    {
+      "epoch": 0.20037730076989752,
+      "grad_norm": 4.414745461532738,
+      "learning_rate": 9.258534677175692e-05,
+      "loss": 1.1278,
+      "step": 3930
+    },
+    {
+      "epoch": 0.20088716667516443,
+      "grad_norm": 7.850860756941811,
+      "learning_rate": 9.254202094747952e-05,
+      "loss": 1.2159,
+      "step": 3940
+    },
+    {
+      "epoch": 0.20139703258043135,
+      "grad_norm": 4.576300386213128,
+      "learning_rate": 9.249857910821682e-05,
+      "loss": 1.053,
+      "step": 3950
+    },
+    {
+      "epoch": 0.20190689848569826,
+      "grad_norm": 5.669850587852958,
+      "learning_rate": 9.245502137243771e-05,
+      "loss": 1.0737,
+      "step": 3960
+    },
+    {
+      "epoch": 0.20241676439096518,
+      "grad_norm": 3.841790288194421,
+      "learning_rate": 9.24113478589271e-05,
+      "loss": 1.084,
+      "step": 3970
+    },
+    {
+      "epoch": 0.2029266302962321,
+      "grad_norm": 4.5287085004351235,
+      "learning_rate": 9.236755868678559e-05,
+      "loss": 1.1155,
+      "step": 3980
+    },
+    {
+      "epoch": 0.203436496201499,
+      "grad_norm": 4.586885578600499,
+      "learning_rate": 9.232365397542926e-05,
+      "loss": 1.0752,
+      "step": 3990
+    },
+    {
+      "epoch": 0.20394636210676592,
+      "grad_norm": 9.096774082551118,
+      "learning_rate": 9.227963384458924e-05,
+      "loss": 1.0771,
+      "step": 4000
+    },
+    {
+      "epoch": 0.20445622801203284,
+      "grad_norm": 6.091396218918115,
+      "learning_rate": 9.223549841431144e-05,
+      "loss": 1.1101,
+      "step": 4010
+    },
+    {
+      "epoch": 0.20496609391729975,
+      "grad_norm": 6.385041154765648,
+      "learning_rate": 9.219124780495616e-05,
+      "loss": 1.11,
+      "step": 4020
+    },
+    {
+      "epoch": 0.20547595982256667,
+      "grad_norm": 3.497672542439059,
+      "learning_rate": 9.214688213719782e-05,
+      "loss": 1.1167,
+      "step": 4030
+    },
+    {
+      "epoch": 0.20598582572783358,
+      "grad_norm": 6.632733925092718,
+      "learning_rate": 9.210240153202462e-05,
+      "loss": 1.1654,
+      "step": 4040
+    },
+    {
+      "epoch": 0.2064956916331005,
+      "grad_norm": 5.1979095263524,
+      "learning_rate": 9.205780611073822e-05,
+      "loss": 1.1256,
+      "step": 4050
+    },
+    {
+      "epoch": 0.2070055575383674,
+      "grad_norm": 4.725662413664164,
+      "learning_rate": 9.201309599495333e-05,
+      "loss": 1.1613,
+      "step": 4060
+    },
+    {
+      "epoch": 0.20751542344363433,
+      "grad_norm": 4.198446410312713,
+      "learning_rate": 9.19682713065975e-05,
+      "loss": 1.1072,
+      "step": 4070
+    },
+    {
+      "epoch": 0.20802528934890124,
+      "grad_norm": 5.632096006878919,
+      "learning_rate": 9.19233321679107e-05,
+      "loss": 1.0694,
+      "step": 4080
+    },
+    {
+      "epoch": 0.20853515525416816,
+      "grad_norm": 5.15508852329062,
+      "learning_rate": 9.187827870144501e-05,
+      "loss": 1.1042,
+      "step": 4090
+    },
+    {
+      "epoch": 0.20904502115943507,
+      "grad_norm": 3.768349793802397,
+      "learning_rate": 9.18331110300643e-05,
+      "loss": 1.0987,
+      "step": 4100
+    },
+    {
+      "epoch": 0.209554887064702,
+      "grad_norm": 4.299188815003413,
+      "learning_rate": 9.178782927694389e-05,
+      "loss": 1.1768,
+      "step": 4110
+    },
+    {
+      "epoch": 0.2100647529699689,
+      "grad_norm": 3.8096689734152287,
+      "learning_rate": 9.17424335655702e-05,
+      "loss": 1.1042,
+      "step": 4120
+    },
+    {
+      "epoch": 0.21057461887523582,
+      "grad_norm": 5.602085196747135,
+      "learning_rate": 9.169692401974043e-05,
+      "loss": 1.1145,
+      "step": 4130
+    },
+    {
+      "epoch": 0.21108448478050273,
+      "grad_norm": 4.969191904156465,
+      "learning_rate": 9.165130076356219e-05,
+      "loss": 1.1362,
+      "step": 4140
+    },
+    {
+      "epoch": 0.21159435068576965,
+      "grad_norm": 6.843019217097579,
+      "learning_rate": 9.160556392145321e-05,
+      "loss": 1.1221,
+      "step": 4150
+    },
+    {
+      "epoch": 0.21210421659103657,
+      "grad_norm": 3.499758917655891,
+      "learning_rate": 9.155971361814097e-05,
+      "loss": 1.0889,
+      "step": 4160
+    },
+    {
+      "epoch": 0.21261408249630348,
+      "grad_norm": 4.621635526066649,
+      "learning_rate": 9.151374997866237e-05,
+      "loss": 1.1394,
+      "step": 4170
+    },
+    {
+      "epoch": 0.2131239484015704,
+      "grad_norm": 4.414049418850473,
+      "learning_rate": 9.146767312836339e-05,
+      "loss": 1.1475,
+      "step": 4180
+    },
+    {
+      "epoch": 0.2136338143068373,
+      "grad_norm": 4.975474445122677,
+      "learning_rate": 9.142148319289872e-05,
+      "loss": 1.0906,
+      "step": 4190
+    },
+    {
+      "epoch": 0.21414368021210423,
+      "grad_norm": 4.470228851354492,
+      "learning_rate": 9.137518029823146e-05,
+      "loss": 1.0443,
+      "step": 4200
+    },
+    {
+      "epoch": 0.21465354611737114,
+      "grad_norm": 4.023509264580457,
+      "learning_rate": 9.132876457063275e-05,
+      "loss": 1.1492,
+      "step": 4210
+    },
+    {
+      "epoch": 0.21516341202263806,
+      "grad_norm": 4.467631871018525,
+      "learning_rate": 9.128223613668145e-05,
+      "loss": 1.0418,
+      "step": 4220
+    },
+    {
+      "epoch": 0.21567327792790497,
+      "grad_norm": 3.7463610188428262,
+      "learning_rate": 9.123559512326376e-05,
+      "loss": 1.116,
+      "step": 4230
+    },
+    {
+      "epoch": 0.2161831438331719,
+      "grad_norm": 6.753094731554434,
+      "learning_rate": 9.118884165757288e-05,
+      "loss": 1.0903,
+      "step": 4240
+    },
+    {
+      "epoch": 0.2166930097384388,
+      "grad_norm": 6.939143231788462,
+      "learning_rate": 9.114197586710873e-05,
+      "loss": 1.0139,
+      "step": 4250
+    },
+    {
+      "epoch": 0.21720287564370572,
+      "grad_norm": 3.7876691852546465,
+      "learning_rate": 9.109499787967747e-05,
+      "loss": 1.0381,
+      "step": 4260
+    },
+    {
+      "epoch": 0.21771274154897263,
+      "grad_norm": 6.277095714725083,
+      "learning_rate": 9.104790782339127e-05,
+      "loss": 1.0978,
+      "step": 4270
+    },
+    {
+      "epoch": 0.21822260745423955,
+      "grad_norm": 4.096173329213358,
+      "learning_rate": 9.100070582666795e-05,
+      "loss": 1.0067,
+      "step": 4280
+    },
+    {
+      "epoch": 0.21873247335950646,
+      "grad_norm": 3.6961770600795623,
+      "learning_rate": 9.095339201823055e-05,
+      "loss": 1.1198,
+      "step": 4290
+    },
+    {
+      "epoch": 0.21924233926477338,
+      "grad_norm": 5.972241108451364,
+      "learning_rate": 9.090596652710702e-05,
+      "loss": 1.1016,
+      "step": 4300
+    },
+    {
+      "epoch": 0.21975220517004027,
+      "grad_norm": 4.313135247453813,
+      "learning_rate": 9.085842948262992e-05,
+      "loss": 1.1174,
+      "step": 4310
+    },
+    {
+      "epoch": 0.22026207107530718,
+      "grad_norm": 4.234029096400693,
+      "learning_rate": 9.081078101443602e-05,
+      "loss": 1.1158,
+      "step": 4320
+    },
+    {
+      "epoch": 0.2207719369805741,
+      "grad_norm": 4.492757703720211,
+      "learning_rate": 9.076302125246592e-05,
+      "loss": 1.0193,
+      "step": 4330
+    },
+    {
+      "epoch": 0.221281802885841,
+      "grad_norm": 4.640881532623444,
+      "learning_rate": 9.071515032696371e-05,
+      "loss": 1.087,
+      "step": 4340
+    },
+    {
+      "epoch": 0.22179166879110793,
+      "grad_norm": 3.617787752452594,
+      "learning_rate": 9.066716836847673e-05,
+      "loss": 1.058,
+      "step": 4350
+    },
+    {
+      "epoch": 0.22230153469637484,
+      "grad_norm": 3.6014749404530466,
+      "learning_rate": 9.061907550785498e-05,
+      "loss": 1.07,
+      "step": 4360
+    },
+    {
+      "epoch": 0.22281140060164176,
+      "grad_norm": 6.001862960183884,
+      "learning_rate": 9.0570871876251e-05,
+      "loss": 1.1156,
+      "step": 4370
+    },
+    {
+      "epoch": 0.22332126650690867,
+      "grad_norm": 5.2487018233747955,
+      "learning_rate": 9.052255760511934e-05,
+      "loss": 1.0502,
+      "step": 4380
+    },
+    {
+      "epoch": 0.2238311324121756,
+      "grad_norm": 3.1087029192988567,
+      "learning_rate": 9.047413282621634e-05,
+      "loss": 1.0547,
+      "step": 4390
+    },
+    {
+      "epoch": 0.2243409983174425,
+      "grad_norm": 5.631068168508301,
+      "learning_rate": 9.042559767159965e-05,
+      "loss": 1.0321,
+      "step": 4400
+    },
+    {
+      "epoch": 0.22485086422270942,
+      "grad_norm": 5.674692322000431,
+      "learning_rate": 9.037695227362793e-05,
+      "loss": 1.1467,
+      "step": 4410
+    },
+    {
+      "epoch": 0.22536073012797633,
+      "grad_norm": 3.7896395368056934,
+      "learning_rate": 9.032819676496052e-05,
+      "loss": 1.113,
+      "step": 4420
+    },
+    {
+      "epoch": 0.22587059603324325,
+      "grad_norm": 4.087460278114435,
+      "learning_rate": 9.027933127855696e-05,
+      "loss": 1.0775,
+      "step": 4430
+    },
+    {
+      "epoch": 0.22638046193851016,
+      "grad_norm": 4.69842077796759,
+      "learning_rate": 9.02303559476768e-05,
+      "loss": 1.1074,
+      "step": 4440
+    },
+    {
+      "epoch": 0.22689032784377708,
+      "grad_norm": 6.276622569586819,
+      "learning_rate": 9.018127090587908e-05,
+      "loss": 1.0951,
+      "step": 4450
+    },
+    {
+      "epoch": 0.227400193749044,
+      "grad_norm": 5.388702778493809,
+      "learning_rate": 9.013207628702205e-05,
+      "loss": 1.0386,
+      "step": 4460
+    },
+    {
+      "epoch": 0.2279100596543109,
+      "grad_norm": 5.487381906765069,
+      "learning_rate": 9.008277222526278e-05,
+      "loss": 1.0894,
+      "step": 4470
+    },
+    {
+      "epoch": 0.22841992555957782,
+      "grad_norm": 3.023066526208634,
+      "learning_rate": 9.003335885505682e-05,
+      "loss": 1.0202,
+      "step": 4480
+    },
+    {
+      "epoch": 0.22892979146484474,
+      "grad_norm": 4.6517471037198455,
+      "learning_rate": 8.998383631115776e-05,
+      "loss": 1.1101,
+      "step": 4490
+    },
+    {
+      "epoch": 0.22943965737011165,
+      "grad_norm": 6.349629107082182,
+      "learning_rate": 8.993420472861695e-05,
+      "loss": 1.0721,
+      "step": 4500
+    },
+    {
+      "epoch": 0.22994952327537857,
+      "grad_norm": 4.227187789476838,
+      "learning_rate": 8.988446424278313e-05,
+      "loss": 1.0836,
+      "step": 4510
+    },
+    {
+      "epoch": 0.23045938918064549,
+      "grad_norm": 3.7667451821363094,
+      "learning_rate": 8.983461498930195e-05,
+      "loss": 1.0098,
+      "step": 4520
+    },
+    {
+      "epoch": 0.2309692550859124,
+      "grad_norm": 5.61529724819574,
+      "learning_rate": 8.978465710411574e-05,
+      "loss": 1.122,
+      "step": 4530
+    },
+    {
+      "epoch": 0.23147912099117932,
+      "grad_norm": 3.950252286659022,
+      "learning_rate": 8.973459072346302e-05,
+      "loss": 1.1244,
+      "step": 4540
+    },
+    {
+      "epoch": 0.23198898689644623,
+      "grad_norm": 4.761507043362676,
+      "learning_rate": 8.968441598387826e-05,
+      "loss": 1.0686,
+      "step": 4550
+    },
+    {
+      "epoch": 0.23249885280171315,
+      "grad_norm": 4.872182191461587,
+      "learning_rate": 8.963413302219134e-05,
+      "loss": 1.0078,
+      "step": 4560
+    },
+    {
+      "epoch": 0.23300871870698006,
+      "grad_norm": 4.086834664724914,
+      "learning_rate": 8.958374197552736e-05,
+      "loss": 1.0911,
+      "step": 4570
+    },
+    {
+      "epoch": 0.23351858461224698,
+      "grad_norm": 4.574015800523239,
+      "learning_rate": 8.953324298130611e-05,
+      "loss": 1.0988,
+      "step": 4580
+    },
+    {
+      "epoch": 0.2340284505175139,
+      "grad_norm": 5.431453651680167,
+      "learning_rate": 8.948263617724178e-05,
+      "loss": 1.1378,
+      "step": 4590
+    },
+    {
+      "epoch": 0.2345383164227808,
+      "grad_norm": 5.759183386680134,
+      "learning_rate": 8.943192170134259e-05,
+      "loss": 1.0764,
+      "step": 4600
+    },
+    {
+      "epoch": 0.23504818232804772,
+      "grad_norm": 4.098486938090133,
+      "learning_rate": 8.938109969191033e-05,
+      "loss": 1.0611,
+      "step": 4610
+    },
+    {
+      "epoch": 0.23555804823331464,
+      "grad_norm": 4.528706455199365,
+      "learning_rate": 8.933017028754012e-05,
+      "loss": 1.0934,
+      "step": 4620
+    },
+    {
+      "epoch": 0.23606791413858155,
+      "grad_norm": 3.698168441687973,
+      "learning_rate": 8.927913362711986e-05,
+      "loss": 1.0981,
+      "step": 4630
+    },
+    {
+      "epoch": 0.23657778004384847,
+      "grad_norm": 3.3076454363441115,
+      "learning_rate": 8.922798984983005e-05,
+      "loss": 1.0675,
+      "step": 4640
+    },
+    {
+      "epoch": 0.23708764594911538,
+      "grad_norm": 4.7960174884936135,
+      "learning_rate": 8.917673909514322e-05,
+      "loss": 1.0734,
+      "step": 4650
+    },
+    {
+      "epoch": 0.2375975118543823,
+      "grad_norm": 3.7657104564262904,
+      "learning_rate": 8.912538150282366e-05,
+      "loss": 1.0388,
+      "step": 4660
+    },
+    {
+      "epoch": 0.2381073777596492,
+      "grad_norm": 2.735618317371669,
+      "learning_rate": 8.907391721292702e-05,
+      "loss": 1.0711,
+      "step": 4670
+    },
+    {
+      "epoch": 0.23861724366491613,
+      "grad_norm": 3.9986942055432038,
+      "learning_rate": 8.902234636579991e-05,
+      "loss": 1.0441,
+      "step": 4680
+    },
+    {
+      "epoch": 0.23912710957018304,
+      "grad_norm": 3.674282337978147,
+      "learning_rate": 8.897066910207958e-05,
+      "loss": 1.0557,
+      "step": 4690
+    },
+    {
+      "epoch": 0.23963697547544996,
+      "grad_norm": 4.851417536326856,
+      "learning_rate": 8.891888556269341e-05,
+      "loss": 1.1019,
+      "step": 4700
+    },
+    {
+      "epoch": 0.24014684138071687,
+      "grad_norm": 4.7533176487328115,
+      "learning_rate": 8.886699588885866e-05,
+      "loss": 1.0392,
+      "step": 4710
+    },
+    {
+      "epoch": 0.2406567072859838,
+      "grad_norm": 3.406931358071219,
+      "learning_rate": 8.881500022208195e-05,
+      "loss": 1.019,
+      "step": 4720
+    },
+    {
+      "epoch": 0.2411665731912507,
+      "grad_norm": 4.6383038465664885,
+      "learning_rate": 8.876289870415905e-05,
+      "loss": 1.0591,
+      "step": 4730
+    },
+    {
+      "epoch": 0.24167643909651762,
+      "grad_norm": 3.4766136409230826,
+      "learning_rate": 8.871069147717433e-05,
+      "loss": 1.066,
+      "step": 4740
+    },
+    {
+      "epoch": 0.24218630500178454,
+      "grad_norm": 3.583770256833579,
+      "learning_rate": 8.865837868350045e-05,
+      "loss": 1.0767,
+      "step": 4750
+    },
+    {
+      "epoch": 0.24269617090705145,
+      "grad_norm": 4.096054842096785,
+      "learning_rate": 8.860596046579794e-05,
+      "loss": 1.0413,
+      "step": 4760
+    },
+    {
+      "epoch": 0.24320603681231837,
+      "grad_norm": 4.604217588651983,
+      "learning_rate": 8.855343696701488e-05,
+      "loss": 1.0441,
+      "step": 4770
+    },
+    {
+      "epoch": 0.24371590271758528,
+      "grad_norm": 2.3345138838960917,
+      "learning_rate": 8.850080833038639e-05,
+      "loss": 1.0091,
+      "step": 4780
+    },
+    {
+      "epoch": 0.2442257686228522,
+      "grad_norm": 5.045281983640511,
+      "learning_rate": 8.844807469943436e-05,
+      "loss": 1.0512,
+      "step": 4790
+    },
+    {
+      "epoch": 0.2447356345281191,
+      "grad_norm": 3.4826243080765806,
+      "learning_rate": 8.839523621796699e-05,
+      "loss": 1.0268,
+      "step": 4800
+    },
+    {
+      "epoch": 0.24524550043338603,
+      "grad_norm": 6.338716603686131,
+      "learning_rate": 8.834229303007841e-05,
+      "loss": 1.0843,
+      "step": 4810
+    },
+    {
+      "epoch": 0.24575536633865294,
+      "grad_norm": 3.255193678645787,
+      "learning_rate": 8.828924528014828e-05,
+      "loss": 1.1201,
+      "step": 4820
+    },
+    {
+      "epoch": 0.24626523224391986,
+      "grad_norm": 3.1881589977771676,
+      "learning_rate": 8.823609311284144e-05,
+      "loss": 1.0373,
+      "step": 4830
+    },
+    {
+      "epoch": 0.24677509814918677,
+      "grad_norm": 3.3439325019724633,
+      "learning_rate": 8.81828366731075e-05,
+      "loss": 1.0318,
+      "step": 4840
+    },
+    {
+      "epoch": 0.2472849640544537,
+      "grad_norm": 3.719968743558736,
+      "learning_rate": 8.812947610618031e-05,
+      "loss": 1.1296,
+      "step": 4850
+    },
+    {
+      "epoch": 0.2477948299597206,
+      "grad_norm": 3.3797188609486106,
+      "learning_rate": 8.807601155757784e-05,
+      "loss": 1.0437,
+      "step": 4860
+    },
+    {
+      "epoch": 0.24830469586498752,
+      "grad_norm": 4.846704945951428,
+      "learning_rate": 8.802244317310152e-05,
+      "loss": 1.0502,
+      "step": 4870
+    },
+    {
+      "epoch": 0.24881456177025443,
+      "grad_norm": 4.1767897486800365,
+      "learning_rate": 8.796877109883598e-05,
+      "loss": 1.0671,
+      "step": 4880
+    },
+    {
+      "epoch": 0.24932442767552135,
+      "grad_norm": 2.971554217856899,
+      "learning_rate": 8.791499548114863e-05,
+      "loss": 1.1219,
+      "step": 4890
+    },
+    {
+      "epoch": 0.24983429358078826,
+      "grad_norm": 5.2119781559216225,
+      "learning_rate": 8.786111646668922e-05,
+      "loss": 1.0953,
+      "step": 4900
+    },
+    {
+      "epoch": 0.25034415948605515,
+      "grad_norm": 3.681987846013078,
+      "learning_rate": 8.780713420238951e-05,
+      "loss": 1.1017,
+      "step": 4910
+    },
+    {
+      "epoch": 0.25085402539132207,
+      "grad_norm": 3.0278906655702302,
+      "learning_rate": 8.775304883546279e-05,
+      "loss": 1.0775,
+      "step": 4920
+    },
+    {
+      "epoch": 0.251363891296589,
+      "grad_norm": 3.9195420892561255,
+      "learning_rate": 8.769886051340353e-05,
+      "loss": 1.0148,
+      "step": 4930
+    },
+    {
+      "epoch": 0.2518737572018559,
+      "grad_norm": 3.565581280322952,
+      "learning_rate": 8.7644569383987e-05,
+      "loss": 1.0843,
+      "step": 4940
+    },
+    {
+      "epoch": 0.2523836231071228,
+      "grad_norm": 5.761206813191621,
+      "learning_rate": 8.759017559526876e-05,
+      "loss": 1.1769,
+      "step": 4950
+    },
+    {
+      "epoch": 0.2528934890123897,
+      "grad_norm": 3.1500014104883083,
+      "learning_rate": 8.753567929558442e-05,
+      "loss": 1.0727,
+      "step": 4960
+    },
+    {
+      "epoch": 0.25340335491765664,
+      "grad_norm": 2.5769313666728144,
+      "learning_rate": 8.748108063354906e-05,
+      "loss": 1.034,
+      "step": 4970
+    },
+    {
+      "epoch": 0.25391322082292356,
+      "grad_norm": 3.333945738470728,
+      "learning_rate": 8.742637975805696e-05,
+      "loss": 1.005,
+      "step": 4980
+    },
+    {
+      "epoch": 0.2544230867281905,
+      "grad_norm": 3.536064757562939,
+      "learning_rate": 8.737157681828112e-05,
+      "loss": 1.0771,
+      "step": 4990
+    },
+    {
+      "epoch": 0.2549329526334574,
+      "grad_norm": 4.126238329507708,
+      "learning_rate": 8.73166719636729e-05,
+      "loss": 1.0809,
+      "step": 5000
+    },
+    {
+      "epoch": 0.2554428185387243,
+      "grad_norm": 4.040213145664409,
+      "learning_rate": 8.726166534396157e-05,
+      "loss": 1.0629,
+      "step": 5010
+    },
+    {
+      "epoch": 0.2559526844439912,
+      "grad_norm": 4.570041080267456,
+      "learning_rate": 8.720655710915393e-05,
+      "loss": 1.0259,
+      "step": 5020
+    },
+    {
+      "epoch": 0.25646255034925813,
+      "grad_norm": 4.423642343969428,
+      "learning_rate": 8.715134740953386e-05,
+      "loss": 1.0394,
+      "step": 5030
+    },
+    {
+      "epoch": 0.25697241625452505,
+      "grad_norm": 2.7536372448154847,
+      "learning_rate": 8.709603639566198e-05,
+      "loss": 0.9952,
+      "step": 5040
+    },
+    {
+      "epoch": 0.25748228215979196,
+      "grad_norm": 4.224827117877528,
+      "learning_rate": 8.70406242183752e-05,
+      "loss": 0.988,
+      "step": 5050
+    },
+    {
+      "epoch": 0.2579921480650589,
+      "grad_norm": 4.509067034941248,
+      "learning_rate": 8.698511102878628e-05,
+      "loss": 1.0695,
+      "step": 5060
+    },
+    {
+      "epoch": 0.2585020139703258,
+      "grad_norm": 5.9059767638098055,
+      "learning_rate": 8.692949697828347e-05,
+      "loss": 1.136,
+      "step": 5070
+    },
+    {
+      "epoch": 0.2590118798755927,
+      "grad_norm": 3.916813048323415,
+      "learning_rate": 8.687378221853008e-05,
+      "loss": 1.1027,
+      "step": 5080
+    },
+    {
+      "epoch": 0.2595217457808596,
+      "grad_norm": 4.494249905043309,
+      "learning_rate": 8.681796690146404e-05,
+      "loss": 1.1143,
+      "step": 5090
+    },
+    {
+      "epoch": 0.26003161168612654,
+      "grad_norm": 4.444678199515402,
+      "learning_rate": 8.676205117929752e-05,
+      "loss": 1.1281,
+      "step": 5100
+    },
+    {
+      "epoch": 0.26054147759139346,
+      "grad_norm": 5.7616406732136545,
+      "learning_rate": 8.670603520451647e-05,
+      "loss": 1.0357,
+      "step": 5110
+    },
+    {
+      "epoch": 0.26105134349666037,
+      "grad_norm": 6.000264920560982,
+      "learning_rate": 8.664991912988032e-05,
+      "loss": 1.0236,
+      "step": 5120
+    },
+    {
+      "epoch": 0.2615612094019273,
+      "grad_norm": 5.089299141780765,
+      "learning_rate": 8.659370310842138e-05,
+      "loss": 1.0294,
+      "step": 5130
+    },
+    {
+      "epoch": 0.2620710753071942,
+      "grad_norm": 3.292913493702765,
+      "learning_rate": 8.653738729344458e-05,
+      "loss": 1.0661,
+      "step": 5140
+    },
+    {
+      "epoch": 0.2625809412124611,
+      "grad_norm": 5.636099103217268,
+      "learning_rate": 8.6480971838527e-05,
+      "loss": 1.0012,
+      "step": 5150
+    },
+    {
+      "epoch": 0.26309080711772803,
+      "grad_norm": 3.8951872942749413,
+      "learning_rate": 8.642445689751736e-05,
+      "loss": 1.0319,
+      "step": 5160
+    },
+    {
+      "epoch": 0.26360067302299495,
+      "grad_norm": 5.6091029812361874,
+      "learning_rate": 8.636784262453583e-05,
+      "loss": 1.0849,
+      "step": 5170
+    },
+    {
+      "epoch": 0.26411053892826186,
+      "grad_norm": 3.3948105111391884,
+      "learning_rate": 8.631112917397331e-05,
+      "loss": 1.0122,
+      "step": 5180
+    },
+    {
+      "epoch": 0.2646204048335288,
+      "grad_norm": 4.348173789492771,
+      "learning_rate": 8.625431670049129e-05,
+      "loss": 1.059,
+      "step": 5190
+    },
+    {
+      "epoch": 0.2651302707387957,
+      "grad_norm": 4.339709846251426,
+      "learning_rate": 8.619740535902123e-05,
+      "loss": 1.1468,
+      "step": 5200
+    },
+    {
+      "epoch": 0.2656401366440626,
+      "grad_norm": 3.115045788546426,
+      "learning_rate": 8.614039530476421e-05,
+      "loss": 1.0592,
+      "step": 5210
+    },
+    {
+      "epoch": 0.2661500025493295,
+      "grad_norm": 4.095364624809172,
+      "learning_rate": 8.608328669319057e-05,
+      "loss": 0.9942,
+      "step": 5220
+    },
+    {
+      "epoch": 0.26665986845459644,
+      "grad_norm": 3.3320331932956524,
+      "learning_rate": 8.602607968003935e-05,
+      "loss": 1.0521,
+      "step": 5230
+    },
+    {
+      "epoch": 0.26716973435986335,
+      "grad_norm": 4.433829696781261,
+      "learning_rate": 8.596877442131798e-05,
+      "loss": 1.0545,
+      "step": 5240
+    },
+    {
+      "epoch": 0.26767960026513027,
+      "grad_norm": 5.22376798696404,
+      "learning_rate": 8.591137107330178e-05,
+      "loss": 1.0657,
+      "step": 5250
+    },
+    {
+      "epoch": 0.2681894661703972,
+      "grad_norm": 3.96131667739153,
+      "learning_rate": 8.58538697925336e-05,
+      "loss": 1.0586,
+      "step": 5260
+    },
+    {
+      "epoch": 0.2686993320756641,
+      "grad_norm": 3.076997333578663,
+      "learning_rate": 8.579627073582334e-05,
+      "loss": 0.9829,
+      "step": 5270
+    },
+    {
+      "epoch": 0.269209197980931,
+      "grad_norm": 5.1691240713000575,
+      "learning_rate": 8.573857406024756e-05,
+      "loss": 1.1053,
+      "step": 5280
+    },
+    {
+      "epoch": 0.26971906388619793,
+      "grad_norm": 4.846399103707031,
+      "learning_rate": 8.568077992314902e-05,
+      "loss": 1.0745,
+      "step": 5290
+    },
+    {
+      "epoch": 0.27022892979146484,
+      "grad_norm": 3.8635471595242543,
+      "learning_rate": 8.562288848213623e-05,
+      "loss": 1.0291,
+      "step": 5300
+    },
+    {
+      "epoch": 0.27073879569673176,
+      "grad_norm": 3.0295588380453613,
+      "learning_rate": 8.55648998950831e-05,
+      "loss": 1.0129,
+      "step": 5310
+    },
+    {
+      "epoch": 0.2712486616019987,
+      "grad_norm": 2.878473080072395,
+      "learning_rate": 8.550681432012848e-05,
+      "loss": 0.9553,
+      "step": 5320
+    },
+    {
+      "epoch": 0.2717585275072656,
+      "grad_norm": 4.083534102290873,
+      "learning_rate": 8.544863191567566e-05,
+      "loss": 1.015,
+      "step": 5330
+    },
+    {
+      "epoch": 0.2722683934125325,
+      "grad_norm": 4.235388313648515,
+      "learning_rate": 8.539035284039202e-05,
+      "loss": 1.1118,
+      "step": 5340
+    },
+    {
+      "epoch": 0.2727782593177994,
+      "grad_norm": 5.058878979715182,
+      "learning_rate": 8.533197725320856e-05,
+      "loss": 1.024,
+      "step": 5350
+    },
+    {
+      "epoch": 0.27328812522306634,
+      "grad_norm": 4.051413812730975,
+      "learning_rate": 8.527350531331948e-05,
+      "loss": 1.0792,
+      "step": 5360
+    },
+    {
+      "epoch": 0.27379799112833325,
+      "grad_norm": 4.2809206121762235,
+      "learning_rate": 8.521493718018174e-05,
+      "loss": 1.0242,
+      "step": 5370
+    },
+    {
+      "epoch": 0.27430785703360017,
+      "grad_norm": 4.607774771425496,
+      "learning_rate": 8.515627301351464e-05,
+      "loss": 1.0761,
+      "step": 5380
+    },
+    {
+      "epoch": 0.2748177229388671,
+      "grad_norm": 3.934021336258175,
+      "learning_rate": 8.509751297329931e-05,
+      "loss": 1.0023,
+      "step": 5390
+    },
+    {
+      "epoch": 0.275327588844134,
+      "grad_norm": 4.539687565934454,
+      "learning_rate": 8.503865721977842e-05,
+      "loss": 0.989,
+      "step": 5400
+    },
+    {
+      "epoch": 0.2758374547494009,
+      "grad_norm": 4.549502194307474,
+      "learning_rate": 8.497970591345559e-05,
+      "loss": 1.1012,
+      "step": 5410
+    },
+    {
+      "epoch": 0.2763473206546678,
+      "grad_norm": 3.3418201574220414,
+      "learning_rate": 8.492065921509506e-05,
+      "loss": 1.0509,
+      "step": 5420
+    },
+    {
+      "epoch": 0.27685718655993474,
+      "grad_norm": 3.44126776153577,
+      "learning_rate": 8.486151728572117e-05,
+      "loss": 0.9759,
+      "step": 5430
+    },
+    {
+      "epoch": 0.27736705246520166,
+      "grad_norm": 4.7821638415619665,
+      "learning_rate": 8.480228028661799e-05,
+      "loss": 1.0115,
+      "step": 5440
+    },
+    {
+      "epoch": 0.2778769183704686,
+      "grad_norm": 4.581162291071151,
+      "learning_rate": 8.474294837932888e-05,
+      "loss": 1.0219,
+      "step": 5450
+    },
+    {
+      "epoch": 0.2783867842757355,
+      "grad_norm": 3.8733442150351505,
+      "learning_rate": 8.468352172565594e-05,
+      "loss": 1.0139,
+      "step": 5460
+    },
+    {
+      "epoch": 0.2788966501810024,
+      "grad_norm": 2.8524393396533805,
+      "learning_rate": 8.462400048765974e-05,
+      "loss": 1.0764,
+      "step": 5470
+    },
+    {
+      "epoch": 0.2794065160862693,
+      "grad_norm": 3.914135068684852,
+      "learning_rate": 8.456438482765871e-05,
+      "loss": 1.0079,
+      "step": 5480
+    },
+    {
+      "epoch": 0.27991638199153623,
+      "grad_norm": 6.970809052094916,
+      "learning_rate": 8.450467490822885e-05,
+      "loss": 1.0655,
+      "step": 5490
+    },
+    {
+      "epoch": 0.28042624789680315,
+      "grad_norm": 4.533422666881598,
+      "learning_rate": 8.444487089220311e-05,
+      "loss": 1.0785,
+      "step": 5500
+    },
+    {
+      "epoch": 0.28093611380207006,
+      "grad_norm": 4.364030587991842,
+      "learning_rate": 8.438497294267117e-05,
+      "loss": 1.0514,
+      "step": 5510
+    },
+    {
+      "epoch": 0.281445979707337,
+      "grad_norm": 3.5954620791553076,
+      "learning_rate": 8.432498122297878e-05,
+      "loss": 1.1249,
+      "step": 5520
+    },
+    {
+      "epoch": 0.2819558456126039,
+      "grad_norm": 5.088004201092307,
+      "learning_rate": 8.426489589672746e-05,
+      "loss": 1.1032,
+      "step": 5530
+    },
+    {
+      "epoch": 0.2824657115178708,
+      "grad_norm": 4.350406613191002,
+      "learning_rate": 8.420471712777397e-05,
+      "loss": 1.0245,
+      "step": 5540
+    },
+    {
+      "epoch": 0.2829755774231377,
+      "grad_norm": 3.7689968249233754,
+      "learning_rate": 8.414444508022993e-05,
+      "loss": 1.0566,
+      "step": 5550
+    },
+    {
+      "epoch": 0.28348544332840464,
+      "grad_norm": 2.3206403906274478,
+      "learning_rate": 8.408407991846128e-05,
+      "loss": 0.9902,
+      "step": 5560
+    },
+    {
+      "epoch": 0.28399530923367156,
+      "grad_norm": 3.2197793343063683,
+      "learning_rate": 8.402362180708792e-05,
+      "loss": 1.0255,
+      "step": 5570
+    },
+    {
+      "epoch": 0.28450517513893847,
+      "grad_norm": 2.979548851656216,
+      "learning_rate": 8.396307091098327e-05,
+      "loss": 1.0217,
+      "step": 5580
+    },
+    {
+      "epoch": 0.2850150410442054,
+      "grad_norm": 2.764591703548254,
+      "learning_rate": 8.390242739527374e-05,
+      "loss": 1.0413,
+      "step": 5590
+    },
+    {
+      "epoch": 0.2855249069494723,
+      "grad_norm": 5.66720911548646,
+      "learning_rate": 8.384169142533829e-05,
+      "loss": 1.0558,
+      "step": 5600
+    },
+    {
+      "epoch": 0.2860347728547392,
+      "grad_norm": 3.4636637664330054,
+      "learning_rate": 8.378086316680805e-05,
+      "loss": 1.0063,
+      "step": 5610
+    },
+    {
+      "epoch": 0.28654463876000613,
+      "grad_norm": 4.561263426716149,
+      "learning_rate": 8.371994278556585e-05,
+      "loss": 1.0149,
+      "step": 5620
+    },
+    {
+      "epoch": 0.28705450466527305,
+      "grad_norm": 3.6249633538048838,
+      "learning_rate": 8.365893044774567e-05,
+      "loss": 1.0217,
+      "step": 5630
+    },
+    {
+      "epoch": 0.28756437057053996,
+      "grad_norm": 4.269152867615141,
+      "learning_rate": 8.359782631973233e-05,
+      "loss": 1.0127,
+      "step": 5640
+    },
+    {
+      "epoch": 0.2880742364758069,
+      "grad_norm": 2.8866554626622665,
+      "learning_rate": 8.353663056816094e-05,
+      "loss": 0.9962,
+      "step": 5650
+    },
+    {
+      "epoch": 0.2885841023810738,
+      "grad_norm": 6.427955292993714,
+      "learning_rate": 8.347534335991649e-05,
+      "loss": 1.0473,
+      "step": 5660
+    },
+    {
+      "epoch": 0.2890939682863407,
+      "grad_norm": 3.645354303457835,
+      "learning_rate": 8.341396486213336e-05,
+      "loss": 1.0641,
+      "step": 5670
+    },
+    {
+      "epoch": 0.2896038341916076,
+      "grad_norm": 2.9236763704046207,
+      "learning_rate": 8.335249524219488e-05,
+      "loss": 0.9943,
+      "step": 5680
+    },
+    {
+      "epoch": 0.29011370009687454,
+      "grad_norm": 3.0417158427244733,
+      "learning_rate": 8.329093466773288e-05,
+      "loss": 1.0266,
+      "step": 5690
+    },
+    {
+      "epoch": 0.29062356600214145,
+      "grad_norm": 4.798262050732901,
+      "learning_rate": 8.322928330662725e-05,
+      "loss": 1.087,
+      "step": 5700
+    },
+    {
+      "epoch": 0.29113343190740837,
+      "grad_norm": 3.387281294186836,
+      "learning_rate": 8.316754132700546e-05,
+      "loss": 1.0454,
+      "step": 5710
+    },
+    {
+      "epoch": 0.2916432978126753,
+      "grad_norm": 3.3192428360977217,
+      "learning_rate": 8.310570889724204e-05,
+      "loss": 1.1112,
+      "step": 5720
+    },
+    {
+      "epoch": 0.2921531637179422,
+      "grad_norm": 4.941941616584977,
+      "learning_rate": 8.304378618595828e-05,
+      "loss": 1.018,
+      "step": 5730
+    },
+    {
+      "epoch": 0.2926630296232091,
+      "grad_norm": 4.190731981597661,
+      "learning_rate": 8.29817733620216e-05,
+      "loss": 1.03,
+      "step": 5740
+    },
+    {
+      "epoch": 0.29317289552847603,
+      "grad_norm": 3.880453031436918,
+      "learning_rate": 8.29196705945452e-05,
+      "loss": 1.0078,
+      "step": 5750
+    },
+    {
+      "epoch": 0.29368276143374294,
+      "grad_norm": 3.4907793070686135,
+      "learning_rate": 8.285747805288756e-05,
+      "loss": 1.0948,
+      "step": 5760
+    },
+    {
+      "epoch": 0.29419262733900986,
+      "grad_norm": 3.411510028798704,
+      "learning_rate": 8.279519590665194e-05,
+      "loss": 1.0575,
+      "step": 5770
+    },
+    {
+      "epoch": 0.2947024932442768,
+      "grad_norm": 4.229858530996611,
+      "learning_rate": 8.273282432568603e-05,
+      "loss": 0.967,
+      "step": 5780
+    },
+    {
+      "epoch": 0.2952123591495437,
+      "grad_norm": 6.096210130250228,
+      "learning_rate": 8.267036348008135e-05,
+      "loss": 1.0062,
+      "step": 5790
+    },
+    {
+      "epoch": 0.2957222250548106,
+      "grad_norm": 6.587238238561264,
+      "learning_rate": 8.260781354017288e-05,
+      "loss": 1.084,
+      "step": 5800
+    },
+    {
+      "epoch": 0.2962320909600775,
+      "grad_norm": 3.7764469559456213,
+      "learning_rate": 8.254517467653858e-05,
+      "loss": 1.004,
+      "step": 5810
+    },
+    {
+      "epoch": 0.29674195686534444,
+      "grad_norm": 4.446471175619889,
+      "learning_rate": 8.248244705999884e-05,
+      "loss": 1.003,
+      "step": 5820
+    },
+    {
+      "epoch": 0.29725182277061135,
+      "grad_norm": 2.8223993096422095,
+      "learning_rate": 8.241963086161619e-05,
+      "loss": 1.0444,
+      "step": 5830
+    },
+    {
+      "epoch": 0.29776168867587827,
+      "grad_norm": 3.407486819057597,
+      "learning_rate": 8.235672625269467e-05,
+      "loss": 1.0348,
+      "step": 5840
+    },
+    {
+      "epoch": 0.2982715545811452,
+      "grad_norm": 6.416390270421421,
+      "learning_rate": 8.229373340477942e-05,
+      "loss": 1.0075,
+      "step": 5850
+    },
+    {
+      "epoch": 0.2987814204864121,
+      "grad_norm": 4.310043434326265,
+      "learning_rate": 8.22306524896562e-05,
+      "loss": 1.0525,
+      "step": 5860
+    },
+    {
+      "epoch": 0.299291286391679,
+      "grad_norm": 4.328322807326321,
+      "learning_rate": 8.216748367935098e-05,
+      "loss": 1.054,
+      "step": 5870
+    },
+    {
+      "epoch": 0.2998011522969459,
+      "grad_norm": 3.373534725731803,
+      "learning_rate": 8.210422714612939e-05,
+      "loss": 1.067,
+      "step": 5880
+    },
+    {
+      "epoch": 0.30031101820221284,
+      "grad_norm": 3.9158391011605205,
+      "learning_rate": 8.204088306249633e-05,
+      "loss": 0.9902,
+      "step": 5890
+    },
+    {
+      "epoch": 0.30082088410747976,
+      "grad_norm": 3.4756763441960103,
+      "learning_rate": 8.19774516011954e-05,
+      "loss": 0.9483,
+      "step": 5900
+    },
+    {
+      "epoch": 0.3013307500127467,
+      "grad_norm": 4.287868146820267,
+      "learning_rate": 8.191393293520851e-05,
+      "loss": 1.0206,
+      "step": 5910
+    },
+    {
+      "epoch": 0.3018406159180136,
+      "grad_norm": 2.7925722726812108,
+      "learning_rate": 8.185032723775539e-05,
+      "loss": 1.0463,
+      "step": 5920
+    },
+    {
+      "epoch": 0.3023504818232805,
+      "grad_norm": 3.7782406839466924,
+      "learning_rate": 8.178663468229308e-05,
+      "loss": 1.0633,
+      "step": 5930
+    },
+    {
+      "epoch": 0.3028603477285474,
+      "grad_norm": 3.4570275115591556,
+      "learning_rate": 8.172285544251557e-05,
+      "loss": 1.0559,
+      "step": 5940
+    },
+    {
+      "epoch": 0.30337021363381433,
+      "grad_norm": 3.8757098152805125,
+      "learning_rate": 8.165898969235313e-05,
+      "loss": 1.0326,
+      "step": 5950
+    },
+    {
+      "epoch": 0.30388007953908125,
+      "grad_norm": 3.734392643562509,
+      "learning_rate": 8.159503760597203e-05,
+      "loss": 1.0562,
+      "step": 5960
+    },
+    {
+      "epoch": 0.3043899454443481,
+      "grad_norm": 4.139058419280422,
+      "learning_rate": 8.153099935777394e-05,
+      "loss": 1.0778,
+      "step": 5970
+    },
+    {
+      "epoch": 0.304899811349615,
+      "grad_norm": 3.3481925980290437,
+      "learning_rate": 8.146687512239555e-05,
+      "loss": 1.0,
+      "step": 5980
+    },
+    {
+      "epoch": 0.30540967725488194,
+      "grad_norm": 3.8821853086298796,
+      "learning_rate": 8.140266507470797e-05,
+      "loss": 1.0344,
+      "step": 5990
+    },
+    {
+      "epoch": 0.30591954316014885,
+      "grad_norm": 3.0577541719194046,
+      "learning_rate": 8.133836938981642e-05,
+      "loss": 1.0506,
+      "step": 6000
+    },
+    {
+      "epoch": 0.30642940906541577,
+      "grad_norm": 4.378539580387895,
+      "learning_rate": 8.127398824305959e-05,
+      "loss": 1.0098,
+      "step": 6010
+    },
+    {
+      "epoch": 0.3069392749706827,
+      "grad_norm": 3.05104160667796,
+      "learning_rate": 8.120952181000922e-05,
+      "loss": 1.0452,
+      "step": 6020
+    },
+    {
+      "epoch": 0.3074491408759496,
+      "grad_norm": 2.9360691762300384,
+      "learning_rate": 8.114497026646967e-05,
+      "loss": 1.0462,
+      "step": 6030
+    },
+    {
+      "epoch": 0.3079590067812165,
+      "grad_norm": 3.3758398442585187,
+      "learning_rate": 8.108033378847741e-05,
+      "loss": 1.0352,
+      "step": 6040
+    },
+    {
+      "epoch": 0.30846887268648343,
+      "grad_norm": 3.7461045036368765,
+      "learning_rate": 8.10156125523005e-05,
+      "loss": 1.024,
+      "step": 6050
+    },
+    {
+      "epoch": 0.30897873859175035,
+      "grad_norm": 4.283391682645258,
+      "learning_rate": 8.095080673443817e-05,
+      "loss": 1.0292,
+      "step": 6060
+    },
+    {
+      "epoch": 0.30948860449701726,
+      "grad_norm": 4.432078710842605,
+      "learning_rate": 8.088591651162027e-05,
+      "loss": 0.9988,
+      "step": 6070
+    },
+    {
+      "epoch": 0.3099984704022842,
+      "grad_norm": 3.3753821189210824,
+      "learning_rate": 8.082094206080685e-05,
+      "loss": 1.017,
+      "step": 6080
+    },
+    {
+      "epoch": 0.3105083363075511,
+      "grad_norm": 3.384123032790279,
+      "learning_rate": 8.075588355918767e-05,
+      "loss": 1.0143,
+      "step": 6090
+    },
+    {
+      "epoch": 0.311018202212818,
+      "grad_norm": 3.7123470517940844,
+      "learning_rate": 8.06907411841817e-05,
+      "loss": 0.9309,
+      "step": 6100
+    },
+    {
+      "epoch": 0.3115280681180849,
+      "grad_norm": 5.4203110416183025,
+      "learning_rate": 8.06255151134366e-05,
+      "loss": 1.0198,
+      "step": 6110
+    },
+    {
+      "epoch": 0.31203793402335184,
+      "grad_norm": 3.369729109830673,
+      "learning_rate": 8.056020552482833e-05,
+      "loss": 1.0045,
+      "step": 6120
+    },
+    {
+      "epoch": 0.31254779992861875,
+      "grad_norm": 2.9044132742638076,
+      "learning_rate": 8.049481259646057e-05,
+      "loss": 1.0379,
+      "step": 6130
+    },
+    {
+      "epoch": 0.31305766583388567,
+      "grad_norm": 3.6956344782317756,
+      "learning_rate": 8.042933650666429e-05,
+      "loss": 1.0147,
+      "step": 6140
+    },
+    {
+      "epoch": 0.3135675317391526,
+      "grad_norm": 4.138698196151885,
+      "learning_rate": 8.036377743399723e-05,
+      "loss": 0.9836,
+      "step": 6150
+    },
+    {
+      "epoch": 0.3140773976444195,
+      "grad_norm": 3.9726914998524134,
+      "learning_rate": 8.029813555724343e-05,
+      "loss": 1.0766,
+      "step": 6160
+    },
+    {
+      "epoch": 0.3145872635496864,
+      "grad_norm": 3.2278452452583983,
+      "learning_rate": 8.02324110554128e-05,
+      "loss": 1.073,
+      "step": 6170
+    },
+    {
+      "epoch": 0.31509712945495333,
+      "grad_norm": 4.036110424011256,
+      "learning_rate": 8.016660410774048e-05,
+      "loss": 1.0414,
+      "step": 6180
+    },
+    {
+      "epoch": 0.31560699536022024,
+      "grad_norm": 3.4721676395767753,
+      "learning_rate": 8.010071489368651e-05,
+      "loss": 1.0064,
+      "step": 6190
+    },
+    {
+      "epoch": 0.31611686126548716,
+      "grad_norm": 4.171481218349725,
+      "learning_rate": 8.003474359293527e-05,
+      "loss": 1.0408,
+      "step": 6200
+    },
+    {
+      "epoch": 0.3166267271707541,
+      "grad_norm": 3.0393525201797007,
+      "learning_rate": 7.996869038539497e-05,
+      "loss": 0.9748,
+      "step": 6210
+    },
+    {
+      "epoch": 0.317136593076021,
+      "grad_norm": 4.086813068218064,
+      "learning_rate": 7.990255545119721e-05,
+      "loss": 1.1157,
+      "step": 6220
+    },
+    {
+      "epoch": 0.3176464589812879,
+      "grad_norm": 4.028125555211706,
+      "learning_rate": 7.983633897069645e-05,
+      "loss": 1.0142,
+      "step": 6230
+    },
+    {
+      "epoch": 0.3181563248865548,
+      "grad_norm": 3.648022023047796,
+      "learning_rate": 7.977004112446954e-05,
+      "loss": 0.9725,
+      "step": 6240
+    },
+    {
+      "epoch": 0.31866619079182174,
+      "grad_norm": 6.246011705858574,
+      "learning_rate": 7.970366209331521e-05,
+      "loss": 1.0264,
+      "step": 6250
+    },
+    {
+      "epoch": 0.31917605669708865,
+      "grad_norm": 3.30169828484668,
+      "learning_rate": 7.96372020582536e-05,
+      "loss": 0.9746,
+      "step": 6260
+    },
+    {
+      "epoch": 0.31968592260235557,
+      "grad_norm": 5.6030520690217065,
+      "learning_rate": 7.957066120052575e-05,
+      "loss": 1.0492,
+      "step": 6270
+    },
+    {
+      "epoch": 0.3201957885076225,
+      "grad_norm": 4.316399453033829,
+      "learning_rate": 7.95040397015931e-05,
+      "loss": 1.0394,
+      "step": 6280
+    },
+    {
+      "epoch": 0.3207056544128894,
+      "grad_norm": 6.31392723555592,
+      "learning_rate": 7.943733774313702e-05,
+      "loss": 1.0226,
+      "step": 6290
+    },
+    {
+      "epoch": 0.3212155203181563,
+      "grad_norm": 7.656997675995733,
+      "learning_rate": 7.937055550705826e-05,
+      "loss": 1.0446,
+      "step": 6300
+    },
+    {
+      "epoch": 0.3217253862234232,
+      "grad_norm": 4.191946074391032,
+      "learning_rate": 7.930369317547655e-05,
+      "loss": 1.083,
+      "step": 6310
+    },
+    {
+      "epoch": 0.32223525212869014,
+      "grad_norm": 3.3920913029142206,
+      "learning_rate": 7.923675093073002e-05,
+      "loss": 1.0026,
+      "step": 6320
+    },
+    {
+      "epoch": 0.32274511803395706,
+      "grad_norm": 2.8751648166684634,
+      "learning_rate": 7.916972895537471e-05,
+      "loss": 0.9979,
+      "step": 6330
+    },
+    {
+      "epoch": 0.32325498393922397,
+      "grad_norm": 8.186076568409813,
+      "learning_rate": 7.91026274321841e-05,
+      "loss": 1.0477,
+      "step": 6340
+    },
+    {
+      "epoch": 0.3237648498444909,
+      "grad_norm": 3.970608260524862,
+      "learning_rate": 7.903544654414863e-05,
+      "loss": 1.0917,
+      "step": 6350
+    },
+    {
+      "epoch": 0.3242747157497578,
+      "grad_norm": 3.9804963341833255,
+      "learning_rate": 7.896818647447511e-05,
+      "loss": 1.0293,
+      "step": 6360
+    },
+    {
+      "epoch": 0.3247845816550247,
+      "grad_norm": 4.759285251414038,
+      "learning_rate": 7.890084740658638e-05,
+      "loss": 0.9541,
+      "step": 6370
+    },
+    {
+      "epoch": 0.32529444756029163,
+      "grad_norm": 4.5870844287880645,
+      "learning_rate": 7.883342952412065e-05,
+      "loss": 1.0821,
+      "step": 6380
+    },
+    {
+      "epoch": 0.32580431346555855,
+      "grad_norm": 3.954644094383673,
+      "learning_rate": 7.876593301093104e-05,
+      "loss": 1.0223,
+      "step": 6390
+    },
+    {
+      "epoch": 0.32631417937082546,
+      "grad_norm": 3.3617478597581214,
+      "learning_rate": 7.869835805108514e-05,
+      "loss": 0.9586,
+      "step": 6400
+    },
+    {
+      "epoch": 0.3268240452760924,
+      "grad_norm": 4.6543609804340775,
+      "learning_rate": 7.86307048288645e-05,
+      "loss": 0.9766,
+      "step": 6410
+    },
+    {
+      "epoch": 0.3273339111813593,
+      "grad_norm": 3.418656450886225,
+      "learning_rate": 7.8562973528764e-05,
+      "loss": 1.0115,
+      "step": 6420
+    },
+    {
+      "epoch": 0.3278437770866262,
+      "grad_norm": 2.553194693948127,
+      "learning_rate": 7.849516433549157e-05,
+      "loss": 1.0203,
+      "step": 6430
+    },
+    {
+      "epoch": 0.3283536429918931,
+      "grad_norm": 5.1167976745938635,
+      "learning_rate": 7.842727743396744e-05,
+      "loss": 1.094,
+      "step": 6440
+    },
+    {
+      "epoch": 0.32886350889716004,
+      "grad_norm": 3.0287183807221933,
+      "learning_rate": 7.835931300932384e-05,
+      "loss": 1.0347,
+      "step": 6450
+    },
+    {
+      "epoch": 0.32937337480242695,
+      "grad_norm": 4.0945458903160405,
+      "learning_rate": 7.829127124690435e-05,
+      "loss": 1.0477,
+      "step": 6460
+    },
+    {
+      "epoch": 0.32988324070769387,
+      "grad_norm": 5.405991749147465,
+      "learning_rate": 7.822315233226352e-05,
+      "loss": 1.0433,
+      "step": 6470
+    },
+    {
+      "epoch": 0.3303931066129608,
+      "grad_norm": 2.6984366185386697,
+      "learning_rate": 7.815495645116623e-05,
+      "loss": 0.9836,
+      "step": 6480
+    },
+    {
+      "epoch": 0.3309029725182277,
+      "grad_norm": 3.19451431480357,
+      "learning_rate": 7.808668378958731e-05,
+      "loss": 1.0136,
+      "step": 6490
+    },
+    {
+      "epoch": 0.3314128384234946,
+      "grad_norm": 3.691317420027209,
+      "learning_rate": 7.801833453371095e-05,
+      "loss": 1.0022,
+      "step": 6500
+    },
+    {
+      "epoch": 0.33192270432876153,
+      "grad_norm": 3.58253893542762,
+      "learning_rate": 7.794990886993023e-05,
+      "loss": 1.0686,
+      "step": 6510
+    },
+    {
+      "epoch": 0.33243257023402845,
+      "grad_norm": 3.2270538104667543,
+      "learning_rate": 7.788140698484656e-05,
+      "loss": 1.0777,
+      "step": 6520
+    },
+    {
+      "epoch": 0.33294243613929536,
+      "grad_norm": 3.825309738285254,
+      "learning_rate": 7.781282906526926e-05,
+      "loss": 1.0087,
+      "step": 6530
+    },
+    {
+      "epoch": 0.3334523020445623,
+      "grad_norm": 3.431292200662794,
+      "learning_rate": 7.774417529821498e-05,
+      "loss": 1.0295,
+      "step": 6540
+    },
+    {
+      "epoch": 0.3339621679498292,
+      "grad_norm": 3.4332600020944493,
+      "learning_rate": 7.767544587090723e-05,
+      "loss": 1.0361,
+      "step": 6550
+    },
+    {
+      "epoch": 0.3344720338550961,
+      "grad_norm": 3.9414270543862218,
+      "learning_rate": 7.76066409707758e-05,
+      "loss": 0.9868,
+      "step": 6560
+    },
+    {
+      "epoch": 0.334981899760363,
+      "grad_norm": 2.7595081933963903,
+      "learning_rate": 7.753776078545636e-05,
+      "loss": 1.034,
+      "step": 6570
+    },
+    {
+      "epoch": 0.33549176566562994,
+      "grad_norm": 3.596539770858611,
+      "learning_rate": 7.746880550278985e-05,
+      "loss": 1.0214,
+      "step": 6580
+    },
+    {
+      "epoch": 0.33600163157089685,
+      "grad_norm": 3.373828688577344,
+      "learning_rate": 7.739977531082201e-05,
+      "loss": 1.0053,
+      "step": 6590
+    },
+    {
+      "epoch": 0.33651149747616377,
+      "grad_norm": 3.6170327929539425,
+      "learning_rate": 7.733067039780288e-05,
+      "loss": 0.955,
+      "step": 6600
+    },
+    {
+      "epoch": 0.3370213633814307,
+      "grad_norm": 2.9437991532594703,
+      "learning_rate": 7.726149095218627e-05,
+      "loss": 0.9963,
+      "step": 6610
+    },
+    {
+      "epoch": 0.3375312292866976,
+      "grad_norm": 4.1275030371627,
+      "learning_rate": 7.719223716262922e-05,
+      "loss": 1.0393,
+      "step": 6620
+    },
+    {
+      "epoch": 0.3380410951919645,
+      "grad_norm": 3.0482612321416576,
+      "learning_rate": 7.712290921799153e-05,
+      "loss": 1.0777,
+      "step": 6630
+    },
+    {
+      "epoch": 0.33855096109723143,
+      "grad_norm": 2.766889269862679,
+      "learning_rate": 7.705350730733523e-05,
+      "loss": 0.9314,
+      "step": 6640
+    },
+    {
+      "epoch": 0.33906082700249834,
+      "grad_norm": 4.2316096098249245,
+      "learning_rate": 7.698403161992403e-05,
+      "loss": 0.9823,
+      "step": 6650
+    },
+    {
+      "epoch": 0.33957069290776526,
+      "grad_norm": 4.5605608461768785,
+      "learning_rate": 7.691448234522285e-05,
+      "loss": 1.024,
+      "step": 6660
+    },
+    {
+      "epoch": 0.3400805588130322,
+      "grad_norm": 3.7882617251352144,
+      "learning_rate": 7.684485967289733e-05,
+      "loss": 1.0745,
+      "step": 6670
+    },
+    {
+      "epoch": 0.3405904247182991,
+      "grad_norm": 3.827009603055397,
+      "learning_rate": 7.677516379281321e-05,
+      "loss": 1.0147,
+      "step": 6680
+    },
+    {
+      "epoch": 0.341100290623566,
+      "grad_norm": 3.58265864202473,
+      "learning_rate": 7.67053948950359e-05,
+      "loss": 1.0131,
+      "step": 6690
+    },
+    {
+      "epoch": 0.3416101565288329,
+      "grad_norm": 2.819232340917403,
+      "learning_rate": 7.663555316982994e-05,
+      "loss": 1.0394,
+      "step": 6700
+    },
+    {
+      "epoch": 0.34212002243409984,
+      "grad_norm": 5.162201471419937,
+      "learning_rate": 7.656563880765846e-05,
+      "loss": 1.0685,
+      "step": 6710
+    },
+    {
+      "epoch": 0.34262988833936675,
+      "grad_norm": 4.604250829044841,
+      "learning_rate": 7.649565199918268e-05,
+      "loss": 1.0994,
+      "step": 6720
+    },
+    {
+      "epoch": 0.34313975424463367,
+      "grad_norm": 2.5913293214568185,
+      "learning_rate": 7.642559293526137e-05,
+      "loss": 0.9885,
+      "step": 6730
+    },
+    {
+      "epoch": 0.3436496201499006,
+      "grad_norm": 2.9520144073304877,
+      "learning_rate": 7.635546180695038e-05,
+      "loss": 0.9502,
+      "step": 6740
+    },
+    {
+      "epoch": 0.3441594860551675,
+      "grad_norm": 3.279479685000254,
+      "learning_rate": 7.628525880550209e-05,
+      "loss": 0.9776,
+      "step": 6750
+    },
+    {
+      "epoch": 0.3446693519604344,
+      "grad_norm": 3.3985425394454665,
+      "learning_rate": 7.621498412236483e-05,
+      "loss": 1.0407,
+      "step": 6760
+    },
+    {
+      "epoch": 0.3451792178657013,
+      "grad_norm": 2.954435423041867,
+      "learning_rate": 7.614463794918246e-05,
+      "loss": 0.9834,
+      "step": 6770
+    },
+    {
+      "epoch": 0.34568908377096824,
+      "grad_norm": 3.4446945006052565,
+      "learning_rate": 7.607422047779375e-05,
+      "loss": 1.0197,
+      "step": 6780
+    },
+    {
+      "epoch": 0.34619894967623516,
+      "grad_norm": 3.6773551695050912,
+      "learning_rate": 7.600373190023197e-05,
+      "loss": 1.0115,
+      "step": 6790
+    },
+    {
+      "epoch": 0.3467088155815021,
+      "grad_norm": 5.390145034760108,
+      "learning_rate": 7.593317240872428e-05,
+      "loss": 0.9882,
+      "step": 6800
+    },
+    {
+      "epoch": 0.347218681486769,
+      "grad_norm": 3.018504473047834,
+      "learning_rate": 7.586254219569115e-05,
+      "loss": 0.9818,
+      "step": 6810
+    },
+    {
+      "epoch": 0.3477285473920359,
+      "grad_norm": 3.094797158174787,
+      "learning_rate": 7.579184145374604e-05,
+      "loss": 1.0678,
+      "step": 6820
+    },
+    {
+      "epoch": 0.3482384132973028,
+      "grad_norm": 3.6178317641775797,
+      "learning_rate": 7.572107037569464e-05,
+      "loss": 0.996,
+      "step": 6830
+    },
+    {
+      "epoch": 0.34874827920256973,
+      "grad_norm": 2.965198313478229,
+      "learning_rate": 7.56502291545345e-05,
+      "loss": 1.0568,
+      "step": 6840
+    },
+    {
+      "epoch": 0.34925814510783665,
+      "grad_norm": 2.932751731657846,
+      "learning_rate": 7.557931798345448e-05,
+      "loss": 0.9739,
+      "step": 6850
+    },
+    {
+      "epoch": 0.34976801101310356,
+      "grad_norm": 6.410663376523632,
+      "learning_rate": 7.550833705583414e-05,
+      "loss": 0.9892,
+      "step": 6860
+    },
+    {
+      "epoch": 0.3502778769183705,
+      "grad_norm": 4.132873483531455,
+      "learning_rate": 7.54372865652433e-05,
+      "loss": 0.9613,
+      "step": 6870
+    },
+    {
+      "epoch": 0.3507877428236374,
+      "grad_norm": 4.817541784367713,
+      "learning_rate": 7.53661667054415e-05,
+      "loss": 0.9519,
+      "step": 6880
+    },
+    {
+      "epoch": 0.3512976087289043,
+      "grad_norm": 2.6834832080268987,
+      "learning_rate": 7.529497767037741e-05,
+      "loss": 0.9636,
+      "step": 6890
+    },
+    {
+      "epoch": 0.3518074746341712,
+      "grad_norm": 4.138092874291924,
+      "learning_rate": 7.52237196541884e-05,
+      "loss": 1.055,
+      "step": 6900
+    },
+    {
+      "epoch": 0.35231734053943814,
+      "grad_norm": 3.125727623443946,
+      "learning_rate": 7.515239285119988e-05,
+      "loss": 0.9689,
+      "step": 6910
+    },
+    {
+      "epoch": 0.35282720644470505,
+      "grad_norm": 4.2695157888907715,
+      "learning_rate": 7.508099745592496e-05,
+      "loss": 0.9804,
+      "step": 6920
+    },
+    {
+      "epoch": 0.35333707234997197,
+      "grad_norm": 3.089823322347325,
+      "learning_rate": 7.500953366306369e-05,
+      "loss": 1.0082,
+      "step": 6930
+    },
+    {
+      "epoch": 0.3538469382552389,
+      "grad_norm": 3.488758675288725,
+      "learning_rate": 7.493800166750273e-05,
+      "loss": 0.9979,
+      "step": 6940
+    },
+    {
+      "epoch": 0.3543568041605058,
+      "grad_norm": 5.360527026709568,
+      "learning_rate": 7.486640166431467e-05,
+      "loss": 1.1085,
+      "step": 6950
+    },
+    {
+      "epoch": 0.3548666700657727,
+      "grad_norm": 3.571941990726853,
+      "learning_rate": 7.479473384875759e-05,
+      "loss": 0.9516,
+      "step": 6960
+    },
+    {
+      "epoch": 0.35537653597103963,
+      "grad_norm": 4.670046148030324,
+      "learning_rate": 7.472299841627451e-05,
+      "loss": 0.9945,
+      "step": 6970
+    },
+    {
+      "epoch": 0.35588640187630655,
+      "grad_norm": 3.8655870091709836,
+      "learning_rate": 7.465119556249285e-05,
+      "loss": 0.9686,
+      "step": 6980
+    },
+    {
+      "epoch": 0.35639626778157346,
+      "grad_norm": 4.21089436280745,
+      "learning_rate": 7.457932548322383e-05,
+      "loss": 0.9965,
+      "step": 6990
+    },
+    {
+      "epoch": 0.3569061336868404,
+      "grad_norm": 4.328581391824839,
+      "learning_rate": 7.450738837446212e-05,
+      "loss": 0.9494,
+      "step": 7000
+    },
+    {
+      "epoch": 0.3574159995921073,
+      "grad_norm": 3.1531910468329905,
+      "learning_rate": 7.443538443238504e-05,
+      "loss": 0.9227,
+      "step": 7010
+    },
+    {
+      "epoch": 0.3579258654973742,
+      "grad_norm": 4.081079875781598,
+      "learning_rate": 7.436331385335226e-05,
+      "loss": 1.0305,
+      "step": 7020
+    },
+    {
+      "epoch": 0.3584357314026411,
+      "grad_norm": 3.596808195506411,
+      "learning_rate": 7.429117683390516e-05,
+      "loss": 1.0313,
+      "step": 7030
+    },
+    {
+      "epoch": 0.35894559730790804,
+      "grad_norm": 3.040454003052074,
+      "learning_rate": 7.421897357076628e-05,
+      "loss": 0.9737,
+      "step": 7040
+    },
+    {
+      "epoch": 0.35945546321317495,
+      "grad_norm": 3.475180199633449,
+      "learning_rate": 7.414670426083887e-05,
+      "loss": 1.1099,
+      "step": 7050
+    },
+    {
+      "epoch": 0.35996532911844187,
+      "grad_norm": 3.5186146669369354,
+      "learning_rate": 7.40743691012062e-05,
+      "loss": 1.0318,
+      "step": 7060
+    },
+    {
+      "epoch": 0.3604751950237088,
+      "grad_norm": 3.6010935658896837,
+      "learning_rate": 7.400196828913123e-05,
+      "loss": 1.0393,
+      "step": 7070
+    },
+    {
+      "epoch": 0.3609850609289757,
+      "grad_norm": 3.964601718404231,
+      "learning_rate": 7.392950202205586e-05,
+      "loss": 0.9936,
+      "step": 7080
+    },
+    {
+      "epoch": 0.3614949268342426,
+      "grad_norm": 3.287468202523833,
+      "learning_rate": 7.385697049760054e-05,
+      "loss": 0.9642,
+      "step": 7090
+    },
+    {
+      "epoch": 0.36200479273950953,
+      "grad_norm": 5.000572627473819,
+      "learning_rate": 7.378437391356367e-05,
+      "loss": 0.9789,
+      "step": 7100
+    },
+    {
+      "epoch": 0.36251465864477644,
+      "grad_norm": 4.145564403281979,
+      "learning_rate": 7.37117124679211e-05,
+      "loss": 1.01,
+      "step": 7110
+    },
+    {
+      "epoch": 0.36302452455004336,
+      "grad_norm": 5.746731477878051,
+      "learning_rate": 7.36389863588255e-05,
+      "loss": 0.951,
+      "step": 7120
+    },
+    {
+      "epoch": 0.3635343904553103,
+      "grad_norm": 3.5135881735177934,
+      "learning_rate": 7.356619578460593e-05,
+      "loss": 1.0421,
+      "step": 7130
+    },
+    {
+      "epoch": 0.3640442563605772,
+      "grad_norm": 3.6815592879194092,
+      "learning_rate": 7.349334094376723e-05,
+      "loss": 1.0099,
+      "step": 7140
+    },
+    {
+      "epoch": 0.3645541222658441,
+      "grad_norm": 4.00363004751103,
+      "learning_rate": 7.342042203498951e-05,
+      "loss": 0.9899,
+      "step": 7150
+    },
+    {
+      "epoch": 0.365063988171111,
+      "grad_norm": 2.6832488579144758,
+      "learning_rate": 7.334743925712762e-05,
+      "loss": 0.983,
+      "step": 7160
+    },
+    {
+      "epoch": 0.36557385407637794,
+      "grad_norm": 3.841794389998567,
+      "learning_rate": 7.327439280921051e-05,
+      "loss": 0.9575,
+      "step": 7170
+    },
+    {
+      "epoch": 0.36608371998164485,
+      "grad_norm": 3.29136888447932,
+      "learning_rate": 7.320128289044086e-05,
+      "loss": 0.962,
+      "step": 7180
+    },
+    {
+      "epoch": 0.36659358588691177,
+      "grad_norm": 2.4127413229065793,
+      "learning_rate": 7.312810970019439e-05,
+      "loss": 0.9729,
+      "step": 7190
+    },
+    {
+      "epoch": 0.3671034517921787,
+      "grad_norm": 3.05054193268961,
+      "learning_rate": 7.305487343801933e-05,
+      "loss": 0.9759,
+      "step": 7200
+    },
+    {
+      "epoch": 0.3676133176974456,
+      "grad_norm": 3.4846575062743717,
+      "learning_rate": 7.298157430363596e-05,
+      "loss": 0.9274,
+      "step": 7210
+    },
+    {
+      "epoch": 0.3681231836027125,
+      "grad_norm": 3.2196195502113105,
+      "learning_rate": 7.290821249693605e-05,
+      "loss": 1.0103,
+      "step": 7220
+    },
+    {
+      "epoch": 0.3686330495079794,
+      "grad_norm": 2.9843784506611004,
+      "learning_rate": 7.283478821798219e-05,
+      "loss": 0.9995,
+      "step": 7230
+    },
+    {
+      "epoch": 0.36914291541324634,
+      "grad_norm": 5.753179910368323,
+      "learning_rate": 7.27613016670074e-05,
+      "loss": 1.0065,
+      "step": 7240
+    },
+    {
+      "epoch": 0.36965278131851326,
+      "grad_norm": 4.367496528660646,
+      "learning_rate": 7.268775304441451e-05,
+      "loss": 1.0498,
+      "step": 7250
+    },
+    {
+      "epoch": 0.3701626472237802,
+      "grad_norm": 4.256618840977742,
+      "learning_rate": 7.26141425507756e-05,
+      "loss": 1.0123,
+      "step": 7260
+    },
+    {
+      "epoch": 0.3706725131290471,
+      "grad_norm": 3.928964763724638,
+      "learning_rate": 7.25404703868315e-05,
+      "loss": 0.9618,
+      "step": 7270
+    },
+    {
+      "epoch": 0.371182379034314,
+      "grad_norm": 2.824065385221502,
+      "learning_rate": 7.246673675349123e-05,
+      "loss": 1.0308,
+      "step": 7280
+    },
+    {
+      "epoch": 0.37169224493958086,
+      "grad_norm": 2.698890858874962,
+      "learning_rate": 7.239294185183141e-05,
+      "loss": 0.9943,
+      "step": 7290
+    },
+    {
+      "epoch": 0.3722021108448478,
+      "grad_norm": 3.0964082743027794,
+      "learning_rate": 7.231908588309576e-05,
+      "loss": 0.9711,
+      "step": 7300
+    },
+    {
+      "epoch": 0.3727119767501147,
+      "grad_norm": 3.927548468490229,
+      "learning_rate": 7.224516904869453e-05,
+      "loss": 0.9781,
+      "step": 7310
+    },
+    {
+      "epoch": 0.3732218426553816,
+      "grad_norm": 3.4194993221090293,
+      "learning_rate": 7.217119155020396e-05,
+      "loss": 0.9691,
+      "step": 7320
+    },
+    {
+      "epoch": 0.3737317085606485,
+      "grad_norm": 2.56858575281676,
+      "learning_rate": 7.20971535893657e-05,
+      "loss": 0.951,
+      "step": 7330
+    },
+    {
+      "epoch": 0.37424157446591544,
+      "grad_norm": 5.285451925259465,
+      "learning_rate": 7.202305536808633e-05,
+      "loss": 1.0044,
+      "step": 7340
+    },
+    {
+      "epoch": 0.37475144037118235,
+      "grad_norm": 3.029951338784045,
+      "learning_rate": 7.194889708843673e-05,
+      "loss": 0.9914,
+      "step": 7350
+    },
+    {
+      "epoch": 0.37526130627644927,
+      "grad_norm": 5.020696791673093,
+      "learning_rate": 7.18746789526516e-05,
+      "loss": 0.9958,
+      "step": 7360
+    },
+    {
+      "epoch": 0.3757711721817162,
+      "grad_norm": 4.566688407768787,
+      "learning_rate": 7.180040116312881e-05,
+      "loss": 1.0929,
+      "step": 7370
+    },
+    {
+      "epoch": 0.3762810380869831,
+      "grad_norm": 3.872358773127792,
+      "learning_rate": 7.172606392242894e-05,
+      "loss": 0.9853,
+      "step": 7380
+    },
+    {
+      "epoch": 0.37679090399225,
+      "grad_norm": 3.4473918049417667,
+      "learning_rate": 7.165166743327473e-05,
+      "loss": 1.0171,
+      "step": 7390
+    },
+    {
+      "epoch": 0.37730076989751693,
+      "grad_norm": 4.071097703939284,
+      "learning_rate": 7.157721189855047e-05,
+      "loss": 0.95,
+      "step": 7400
+    },
+    {
+      "epoch": 0.37781063580278385,
+      "grad_norm": 4.738454803206081,
+      "learning_rate": 7.150269752130144e-05,
+      "loss": 1.0548,
+      "step": 7410
+    },
+    {
+      "epoch": 0.37832050170805076,
+      "grad_norm": 3.211095977619034,
+      "learning_rate": 7.142812450473344e-05,
+      "loss": 1.0732,
+      "step": 7420
+    },
+    {
+      "epoch": 0.3788303676133177,
+      "grad_norm": 2.408010202469844,
+      "learning_rate": 7.135349305221216e-05,
+      "loss": 1.055,
+      "step": 7430
+    },
+    {
+      "epoch": 0.3793402335185846,
+      "grad_norm": 3.738509463429098,
+      "learning_rate": 7.127880336726262e-05,
+      "loss": 1.0725,
+      "step": 7440
+    },
+    {
+      "epoch": 0.3798500994238515,
+      "grad_norm": 3.6802881504337472,
+      "learning_rate": 7.120405565356871e-05,
+      "loss": 1.0432,
+      "step": 7450
+    },
+    {
+      "epoch": 0.3803599653291184,
+      "grad_norm": 6.50944828265682,
+      "learning_rate": 7.11292501149725e-05,
+      "loss": 1.0184,
+      "step": 7460
+    },
+    {
+      "epoch": 0.38086983123438534,
+      "grad_norm": 2.8461494658896944,
+      "learning_rate": 7.105438695547381e-05,
+      "loss": 0.9833,
+      "step": 7470
+    },
+    {
+      "epoch": 0.38137969713965225,
+      "grad_norm": 2.8845637706342444,
+      "learning_rate": 7.097946637922955e-05,
+      "loss": 0.926,
+      "step": 7480
+    },
+    {
+      "epoch": 0.38188956304491917,
+      "grad_norm": 4.966386983635893,
+      "learning_rate": 7.090448859055321e-05,
+      "loss": 1.0482,
+      "step": 7490
+    },
+    {
+      "epoch": 0.3823994289501861,
+      "grad_norm": 3.8734273182198353,
+      "learning_rate": 7.082945379391436e-05,
+      "loss": 0.9855,
+      "step": 7500
+    },
+    {
+      "epoch": 0.382909294855453,
+      "grad_norm": 7.2654653468941,
+      "learning_rate": 7.075436219393798e-05,
+      "loss": 1.0067,
+      "step": 7510
+    },
+    {
+      "epoch": 0.3834191607607199,
+      "grad_norm": 2.6321094110314145,
+      "learning_rate": 7.067921399540395e-05,
+      "loss": 0.9443,
+      "step": 7520
+    },
+    {
+      "epoch": 0.38392902666598683,
+      "grad_norm": 3.772431198509487,
+      "learning_rate": 7.060400940324655e-05,
+      "loss": 1.0275,
+      "step": 7530
+    },
+    {
+      "epoch": 0.38443889257125374,
+      "grad_norm": 2.864068961873481,
+      "learning_rate": 7.052874862255381e-05,
+      "loss": 0.9,
+      "step": 7540
+    },
+    {
+      "epoch": 0.38494875847652066,
+      "grad_norm": 3.244954947277513,
+      "learning_rate": 7.045343185856701e-05,
+      "loss": 1.0231,
+      "step": 7550
+    },
+    {
+      "epoch": 0.3854586243817876,
+      "grad_norm": 2.9501480598286625,
+      "learning_rate": 7.037805931668005e-05,
+      "loss": 1.0132,
+      "step": 7560
+    },
+    {
+      "epoch": 0.3859684902870545,
+      "grad_norm": 4.033800846554501,
+      "learning_rate": 7.030263120243902e-05,
+      "loss": 0.9507,
+      "step": 7570
+    },
+    {
+      "epoch": 0.3864783561923214,
+      "grad_norm": 4.935920805568241,
+      "learning_rate": 7.02271477215415e-05,
+      "loss": 1.0124,
+      "step": 7580
+    },
+    {
+      "epoch": 0.3869882220975883,
+      "grad_norm": 2.9909597802655394,
+      "learning_rate": 7.015160907983609e-05,
+      "loss": 1.08,
+      "step": 7590
+    },
+    {
+      "epoch": 0.38749808800285523,
+      "grad_norm": 3.0885134763230826,
+      "learning_rate": 7.007601548332179e-05,
+      "loss": 0.9899,
+      "step": 7600
+    },
+    {
+      "epoch": 0.38800795390812215,
+      "grad_norm": 4.085304060769854,
+      "learning_rate": 7.000036713814749e-05,
+      "loss": 1.0568,
+      "step": 7610
+    },
+    {
+      "epoch": 0.38851781981338906,
+      "grad_norm": 2.8903100587581356,
+      "learning_rate": 6.992466425061137e-05,
+      "loss": 0.9689,
+      "step": 7620
+    },
+    {
+      "epoch": 0.389027685718656,
+      "grad_norm": 3.4937069513104286,
+      "learning_rate": 6.984890702716036e-05,
+      "loss": 1.0279,
+      "step": 7630
+    },
+    {
+      "epoch": 0.3895375516239229,
+      "grad_norm": 2.5446672367985266,
+      "learning_rate": 6.977309567438954e-05,
+      "loss": 0.9681,
+      "step": 7640
+    },
+    {
+      "epoch": 0.3900474175291898,
+      "grad_norm": 3.2734287563527222,
+      "learning_rate": 6.969723039904166e-05,
+      "loss": 0.9758,
+      "step": 7650
+    },
+    {
+      "epoch": 0.3905572834344567,
+      "grad_norm": 4.759649973235408,
+      "learning_rate": 6.962131140800647e-05,
+      "loss": 0.9751,
+      "step": 7660
+    },
+    {
+      "epoch": 0.39106714933972364,
+      "grad_norm": 3.5749996959546926,
+      "learning_rate": 6.954533890832023e-05,
+      "loss": 0.9652,
+      "step": 7670
+    },
+    {
+      "epoch": 0.39157701524499056,
+      "grad_norm": 4.91677035555623,
+      "learning_rate": 6.94693131071651e-05,
+      "loss": 0.9514,
+      "step": 7680
+    },
+    {
+      "epoch": 0.39208688115025747,
+      "grad_norm": 3.155742449847258,
+      "learning_rate": 6.939323421186861e-05,
+      "loss": 0.9955,
+      "step": 7690
+    },
+    {
+      "epoch": 0.3925967470555244,
+      "grad_norm": 5.310305673644114,
+      "learning_rate": 6.931710242990312e-05,
+      "loss": 0.8624,
+      "step": 7700
+    },
+    {
+      "epoch": 0.3931066129607913,
+      "grad_norm": 3.5806355390430546,
+      "learning_rate": 6.924091796888512e-05,
+      "loss": 0.9596,
+      "step": 7710
+    },
+    {
+      "epoch": 0.3936164788660582,
+      "grad_norm": 2.561795181295556,
+      "learning_rate": 6.916468103657489e-05,
+      "loss": 0.9755,
+      "step": 7720
+    },
+    {
+      "epoch": 0.39412634477132513,
+      "grad_norm": 2.64332129127512,
+      "learning_rate": 6.908839184087566e-05,
+      "loss": 0.972,
+      "step": 7730
+    },
+    {
+      "epoch": 0.39463621067659205,
+      "grad_norm": 3.78250238500619,
+      "learning_rate": 6.901205058983332e-05,
+      "loss": 1.0044,
+      "step": 7740
+    },
+    {
+      "epoch": 0.39514607658185896,
+      "grad_norm": 5.87662489302249,
+      "learning_rate": 6.89356574916356e-05,
+      "loss": 1.0159,
+      "step": 7750
+    },
+    {
+      "epoch": 0.3956559424871259,
+      "grad_norm": 3.9235098033032223,
+      "learning_rate": 6.885921275461168e-05,
+      "loss": 1.0609,
+      "step": 7760
+    },
+    {
+      "epoch": 0.3961658083923928,
+      "grad_norm": 2.6444250632626147,
+      "learning_rate": 6.878271658723162e-05,
+      "loss": 0.9404,
+      "step": 7770
+    },
+    {
+      "epoch": 0.3966756742976597,
+      "grad_norm": 4.425412693097132,
+      "learning_rate": 6.870616919810562e-05,
+      "loss": 1.0248,
+      "step": 7780
+    },
+    {
+      "epoch": 0.3971855402029266,
+      "grad_norm": 5.1067176297252095,
+      "learning_rate": 6.862957079598362e-05,
+      "loss": 0.9419,
+      "step": 7790
+    },
+    {
+      "epoch": 0.39769540610819354,
+      "grad_norm": 5.424418090130069,
+      "learning_rate": 6.855292158975468e-05,
+      "loss": 0.979,
+      "step": 7800
+    },
+    {
+      "epoch": 0.39820527201346045,
+      "grad_norm": 4.041976527019181,
+      "learning_rate": 6.84762217884464e-05,
+      "loss": 1.0984,
+      "step": 7810
+    },
+    {
+      "epoch": 0.39871513791872737,
+      "grad_norm": 3.0713325053672857,
+      "learning_rate": 6.839947160122436e-05,
+      "loss": 0.9948,
+      "step": 7820
+    },
+    {
+      "epoch": 0.3992250038239943,
+      "grad_norm": 5.283031713528076,
+      "learning_rate": 6.832267123739154e-05,
+      "loss": 1.0389,
+      "step": 7830
+    },
+    {
+      "epoch": 0.3997348697292612,
+      "grad_norm": 3.405740106561875,
+      "learning_rate": 6.824582090638777e-05,
+      "loss": 0.9844,
+      "step": 7840
+    },
+    {
+      "epoch": 0.4002447356345281,
+      "grad_norm": 4.05474377505947,
+      "learning_rate": 6.81689208177891e-05,
+      "loss": 1.0185,
+      "step": 7850
+    },
+    {
+      "epoch": 0.40075460153979503,
+      "grad_norm": 4.218173684188458,
+      "learning_rate": 6.809197118130734e-05,
+      "loss": 1.0217,
+      "step": 7860
+    },
+    {
+      "epoch": 0.40126446744506195,
+      "grad_norm": 5.375351162624322,
+      "learning_rate": 6.801497220678935e-05,
+      "loss": 1.0078,
+      "step": 7870
+    },
+    {
+      "epoch": 0.40177433335032886,
+      "grad_norm": 2.3886255403553047,
+      "learning_rate": 6.793792410421658e-05,
+      "loss": 0.9489,
+      "step": 7880
+    },
+    {
+      "epoch": 0.4022841992555958,
+      "grad_norm": 4.616915049037168,
+      "learning_rate": 6.786082708370447e-05,
+      "loss": 1.0114,
+      "step": 7890
+    },
+    {
+      "epoch": 0.4027940651608627,
+      "grad_norm": 3.9455847324621143,
+      "learning_rate": 6.778368135550182e-05,
+      "loss": 0.974,
+      "step": 7900
+    },
+    {
+      "epoch": 0.4033039310661296,
+      "grad_norm": 4.82991485201106,
+      "learning_rate": 6.770648712999025e-05,
+      "loss": 0.9652,
+      "step": 7910
+    },
+    {
+      "epoch": 0.4038137969713965,
+      "grad_norm": 2.9101742427326,
+      "learning_rate": 6.762924461768375e-05,
+      "loss": 0.953,
+      "step": 7920
+    },
+    {
+      "epoch": 0.40432366287666344,
+      "grad_norm": 3.3596524918735917,
+      "learning_rate": 6.755195402922781e-05,
+      "loss": 0.9509,
+      "step": 7930
+    },
+    {
+      "epoch": 0.40483352878193035,
+      "grad_norm": 4.225953624249036,
+      "learning_rate": 6.747461557539918e-05,
+      "loss": 0.9973,
+      "step": 7940
+    },
+    {
+      "epoch": 0.40534339468719727,
+      "grad_norm": 3.5903876664997485,
+      "learning_rate": 6.739722946710507e-05,
+      "loss": 0.9357,
+      "step": 7950
+    },
+    {
+      "epoch": 0.4058532605924642,
+      "grad_norm": 5.3606943141476195,
+      "learning_rate": 6.731979591538267e-05,
+      "loss": 1.0182,
+      "step": 7960
+    },
+    {
+      "epoch": 0.4063631264977311,
+      "grad_norm": 3.2153408156011287,
+      "learning_rate": 6.724231513139852e-05,
+      "loss": 1.0024,
+      "step": 7970
+    },
+    {
+      "epoch": 0.406872992402998,
+      "grad_norm": 3.292294552670899,
+      "learning_rate": 6.716478732644802e-05,
+      "loss": 0.9274,
+      "step": 7980
+    },
+    {
+      "epoch": 0.40738285830826493,
+      "grad_norm": 2.7317030544391163,
+      "learning_rate": 6.708721271195476e-05,
+      "loss": 0.964,
+      "step": 7990
+    },
+    {
+      "epoch": 0.40789272421353184,
+      "grad_norm": 4.092024090983047,
+      "learning_rate": 6.700959149946996e-05,
+      "loss": 1.0025,
+      "step": 8000
+    },
+    {
+      "epoch": 0.40840259011879876,
+      "grad_norm": 2.788397450491997,
+      "learning_rate": 6.693192390067199e-05,
+      "loss": 0.9709,
+      "step": 8010
+    },
+    {
+      "epoch": 0.4089124560240657,
+      "grad_norm": 3.4119060630259117,
+      "learning_rate": 6.685421012736563e-05,
+      "loss": 0.9564,
+      "step": 8020
+    },
+    {
+      "epoch": 0.4094223219293326,
+      "grad_norm": 2.9489807050382186,
+      "learning_rate": 6.677645039148168e-05,
+      "loss": 0.9713,
+      "step": 8030
+    },
+    {
+      "epoch": 0.4099321878345995,
+      "grad_norm": 2.817060206884971,
+      "learning_rate": 6.66986449050762e-05,
+      "loss": 0.9081,
+      "step": 8040
+    },
+    {
+      "epoch": 0.4104420537398664,
+      "grad_norm": 3.780284199420438,
+      "learning_rate": 6.662079388033004e-05,
+      "loss": 1.0138,
+      "step": 8050
+    },
+    {
+      "epoch": 0.41095191964513333,
+      "grad_norm": 3.5269807271137155,
+      "learning_rate": 6.654289752954826e-05,
+      "loss": 1.0603,
+      "step": 8060
+    },
+    {
+      "epoch": 0.41146178555040025,
+      "grad_norm": 3.135683553089666,
+      "learning_rate": 6.646495606515949e-05,
+      "loss": 0.9733,
+      "step": 8070
+    },
+    {
+      "epoch": 0.41197165145566716,
+      "grad_norm": 2.7791201230184592,
+      "learning_rate": 6.638696969971542e-05,
+      "loss": 0.9089,
+      "step": 8080
+    },
+    {
+      "epoch": 0.4124815173609341,
+      "grad_norm": 2.404636007860764,
+      "learning_rate": 6.630893864589016e-05,
+      "loss": 0.9515,
+      "step": 8090
+    },
+    {
+      "epoch": 0.412991383266201,
+      "grad_norm": 3.9092946580394634,
+      "learning_rate": 6.62308631164797e-05,
+      "loss": 1.0245,
+      "step": 8100
+    },
+    {
+      "epoch": 0.4135012491714679,
+      "grad_norm": 2.8819181887521177,
+      "learning_rate": 6.615274332440134e-05,
+      "loss": 0.9682,
+      "step": 8110
+    },
+    {
+      "epoch": 0.4140111150767348,
+      "grad_norm": 4.201890582364877,
+      "learning_rate": 6.607457948269305e-05,
+      "loss": 0.9545,
+      "step": 8120
+    },
+    {
+      "epoch": 0.41452098098200174,
+      "grad_norm": 3.0616475395519043,
+      "learning_rate": 6.599637180451294e-05,
+      "loss": 1.001,
+      "step": 8130
+    },
+    {
+      "epoch": 0.41503084688726866,
+      "grad_norm": 6.135332887534604,
+      "learning_rate": 6.59181205031387e-05,
+      "loss": 1.0417,
+      "step": 8140
+    },
+    {
+      "epoch": 0.41554071279253557,
+      "grad_norm": 3.3915570825360426,
+      "learning_rate": 6.583982579196693e-05,
+      "loss": 0.9526,
+      "step": 8150
+    },
+    {
+      "epoch": 0.4160505786978025,
+      "grad_norm": 5.130771096453491,
+      "learning_rate": 6.576148788451264e-05,
+      "loss": 0.9996,
+      "step": 8160
+    },
+    {
+      "epoch": 0.4165604446030694,
+      "grad_norm": 4.225003471130253,
+      "learning_rate": 6.568310699440861e-05,
+      "loss": 1.0112,
+      "step": 8170
+    },
+    {
+      "epoch": 0.4170703105083363,
+      "grad_norm": 4.195416717849394,
+      "learning_rate": 6.56046833354049e-05,
+      "loss": 1.022,
+      "step": 8180
+    },
+    {
+      "epoch": 0.41758017641360323,
+      "grad_norm": 4.349548622404095,
+      "learning_rate": 6.552621712136812e-05,
+      "loss": 1.0196,
+      "step": 8190
+    },
+    {
+      "epoch": 0.41809004231887015,
+      "grad_norm": 5.608295368338538,
+      "learning_rate": 6.544770856628099e-05,
+      "loss": 0.9041,
+      "step": 8200
+    },
+    {
+      "epoch": 0.41859990822413706,
+      "grad_norm": 2.464411009934411,
+      "learning_rate": 6.536915788424171e-05,
+      "loss": 0.9406,
+      "step": 8210
+    },
+    {
+      "epoch": 0.419109774129404,
+      "grad_norm": 3.1139854599901136,
+      "learning_rate": 6.529056528946329e-05,
+      "loss": 0.979,
+      "step": 8220
+    },
+    {
+      "epoch": 0.4196196400346709,
+      "grad_norm": 4.065458780063512,
+      "learning_rate": 6.521193099627311e-05,
+      "loss": 0.991,
+      "step": 8230
+    },
+    {
+      "epoch": 0.4201295059399378,
+      "grad_norm": 2.774278033784981,
+      "learning_rate": 6.513325521911223e-05,
+      "loss": 0.9665,
+      "step": 8240
+    },
+    {
+      "epoch": 0.4206393718452047,
+      "grad_norm": 3.723337247910771,
+      "learning_rate": 6.505453817253483e-05,
+      "loss": 1.0147,
+      "step": 8250
+    },
+    {
+      "epoch": 0.42114923775047164,
+      "grad_norm": 3.0144132742908103,
+      "learning_rate": 6.49757800712077e-05,
+      "loss": 0.9086,
+      "step": 8260
+    },
+    {
+      "epoch": 0.42165910365573855,
+      "grad_norm": 3.76467661836284,
+      "learning_rate": 6.489698112990949e-05,
+      "loss": 0.9483,
+      "step": 8270
+    },
+    {
+      "epoch": 0.42216896956100547,
+      "grad_norm": 5.533722695705143,
+      "learning_rate": 6.481814156353028e-05,
+      "loss": 0.9709,
+      "step": 8280
+    },
+    {
+      "epoch": 0.4226788354662724,
+      "grad_norm": 4.402153778706855,
+      "learning_rate": 6.473926158707095e-05,
+      "loss": 0.9836,
+      "step": 8290
+    },
+    {
+      "epoch": 0.4231887013715393,
+      "grad_norm": 3.7297200027269692,
+      "learning_rate": 6.466034141564256e-05,
+      "loss": 0.9816,
+      "step": 8300
+    },
+    {
+      "epoch": 0.4236985672768062,
+      "grad_norm": 2.9872321174303638,
+      "learning_rate": 6.458138126446578e-05,
+      "loss": 0.9566,
+      "step": 8310
+    },
+    {
+      "epoch": 0.42420843318207313,
+      "grad_norm": 4.776691500929038,
+      "learning_rate": 6.45023813488703e-05,
+      "loss": 1.003,
+      "step": 8320
+    },
+    {
+      "epoch": 0.42471829908734005,
+      "grad_norm": 3.590468781706842,
+      "learning_rate": 6.442334188429429e-05,
+      "loss": 0.9772,
+      "step": 8330
+    },
+    {
+      "epoch": 0.42522816499260696,
+      "grad_norm": 3.8608403740136468,
+      "learning_rate": 6.434426308628373e-05,
+      "loss": 0.9319,
+      "step": 8340
+    },
+    {
+      "epoch": 0.4257380308978739,
+      "grad_norm": 4.444880659868321,
+      "learning_rate": 6.426514517049189e-05,
+      "loss": 1.0666,
+      "step": 8350
+    },
+    {
+      "epoch": 0.4262478968031408,
+      "grad_norm": 3.9726356268299745,
+      "learning_rate": 6.418598835267872e-05,
+      "loss": 1.0054,
+      "step": 8360
+    },
+    {
+      "epoch": 0.4267577627084077,
+      "grad_norm": 3.0163954750719095,
+      "learning_rate": 6.41067928487102e-05,
+      "loss": 1.0377,
+      "step": 8370
+    },
+    {
+      "epoch": 0.4272676286136746,
+      "grad_norm": 2.535923278426502,
+      "learning_rate": 6.402755887455792e-05,
+      "loss": 1.0057,
+      "step": 8380
+    },
+    {
+      "epoch": 0.42777749451894154,
+      "grad_norm": 3.3930949309239304,
+      "learning_rate": 6.394828664629828e-05,
+      "loss": 0.8832,
+      "step": 8390
+    },
+    {
+      "epoch": 0.42828736042420845,
+      "grad_norm": 3.059493768433064,
+      "learning_rate": 6.386897638011206e-05,
+      "loss": 1.0194,
+      "step": 8400
+    },
+    {
+      "epoch": 0.42879722632947537,
+      "grad_norm": 2.5297070537029023,
+      "learning_rate": 6.378962829228371e-05,
+      "loss": 0.989,
+      "step": 8410
+    },
+    {
+      "epoch": 0.4293070922347423,
+      "grad_norm": 3.9679756011783525,
+      "learning_rate": 6.371024259920091e-05,
+      "loss": 1.0448,
+      "step": 8420
+    },
+    {
+      "epoch": 0.4298169581400092,
+      "grad_norm": 2.661923982062292,
+      "learning_rate": 6.363081951735384e-05,
+      "loss": 1.0615,
+      "step": 8430
+    },
+    {
+      "epoch": 0.4303268240452761,
+      "grad_norm": 3.435589450600634,
+      "learning_rate": 6.35513592633346e-05,
+      "loss": 1.0189,
+      "step": 8440
+    },
+    {
+      "epoch": 0.43083668995054303,
+      "grad_norm": 2.9961841597196672,
+      "learning_rate": 6.347186205383678e-05,
+      "loss": 1.0101,
+      "step": 8450
+    },
+    {
+      "epoch": 0.43134655585580994,
+      "grad_norm": 3.1285804783723883,
+      "learning_rate": 6.339232810565463e-05,
+      "loss": 0.9586,
+      "step": 8460
+    },
+    {
+      "epoch": 0.43185642176107686,
+      "grad_norm": 4.510792797789835,
+      "learning_rate": 6.331275763568266e-05,
+      "loss": 0.9259,
+      "step": 8470
+    },
+    {
+      "epoch": 0.4323662876663438,
+      "grad_norm": 4.204311113244711,
+      "learning_rate": 6.323315086091494e-05,
+      "loss": 1.0109,
+      "step": 8480
+    },
+    {
+      "epoch": 0.4328761535716107,
+      "grad_norm": 3.287802283355501,
+      "learning_rate": 6.315350799844455e-05,
+      "loss": 0.9761,
+      "step": 8490
+    },
+    {
+      "epoch": 0.4333860194768776,
+      "grad_norm": 5.124636269441518,
+      "learning_rate": 6.307382926546303e-05,
+      "loss": 1.0093,
+      "step": 8500
+    },
+    {
+      "epoch": 0.4338958853821445,
+      "grad_norm": 3.224727245415259,
+      "learning_rate": 6.299411487925967e-05,
+      "loss": 0.9701,
+      "step": 8510
+    },
+    {
+      "epoch": 0.43440575128741143,
+      "grad_norm": 4.10442689099236,
+      "learning_rate": 6.291436505722105e-05,
+      "loss": 0.9962,
+      "step": 8520
+    },
+    {
+      "epoch": 0.43491561719267835,
+      "grad_norm": 3.134233230716742,
+      "learning_rate": 6.283458001683033e-05,
+      "loss": 0.9685,
+      "step": 8530
+    },
+    {
+      "epoch": 0.43542548309794527,
+      "grad_norm": 4.416029020529532,
+      "learning_rate": 6.275475997566679e-05,
+      "loss": 0.9539,
+      "step": 8540
+    },
+    {
+      "epoch": 0.4359353490032122,
+      "grad_norm": 3.2852254179593254,
+      "learning_rate": 6.267490515140506e-05,
+      "loss": 1.0486,
+      "step": 8550
+    },
+    {
+      "epoch": 0.4364452149084791,
+      "grad_norm": 3.814048589694972,
+      "learning_rate": 6.259501576181471e-05,
+      "loss": 0.9066,
+      "step": 8560
+    },
+    {
+      "epoch": 0.436955080813746,
+      "grad_norm": 3.198841271697382,
+      "learning_rate": 6.251509202475955e-05,
+      "loss": 0.9519,
+      "step": 8570
+    },
+    {
+      "epoch": 0.4374649467190129,
+      "grad_norm": 3.647237279022868,
+      "learning_rate": 6.243513415819701e-05,
+      "loss": 0.9508,
+      "step": 8580
+    },
+    {
+      "epoch": 0.43797481262427984,
+      "grad_norm": 3.0249604860517407,
+      "learning_rate": 6.235514238017767e-05,
+      "loss": 0.9915,
+      "step": 8590
+    },
+    {
+      "epoch": 0.43848467852954676,
+      "grad_norm": 4.31484540170633,
+      "learning_rate": 6.227511690884454e-05,
+      "loss": 0.9684,
+      "step": 8600
+    },
+    {
+      "epoch": 0.43899454443481367,
+      "grad_norm": 2.594466421786102,
+      "learning_rate": 6.219505796243248e-05,
+      "loss": 0.9741,
+      "step": 8610
+    },
+    {
+      "epoch": 0.43950441034008053,
+      "grad_norm": 3.7573732890106033,
+      "learning_rate": 6.211496575926775e-05,
+      "loss": 1.0002,
+      "step": 8620
+    },
+    {
+      "epoch": 0.44001427624534745,
+      "grad_norm": 2.9838071187088744,
+      "learning_rate": 6.203484051776721e-05,
+      "loss": 0.9893,
+      "step": 8630
+    },
+    {
+      "epoch": 0.44052414215061436,
+      "grad_norm": 3.1365011892344676,
+      "learning_rate": 6.195468245643783e-05,
+      "loss": 0.9622,
+      "step": 8640
+    },
+    {
+      "epoch": 0.4410340080558813,
+      "grad_norm": 2.7563754695966582,
+      "learning_rate": 6.187449179387613e-05,
+      "loss": 0.9331,
+      "step": 8650
+    },
+    {
+      "epoch": 0.4415438739611482,
+      "grad_norm": 2.7666951720052,
+      "learning_rate": 6.179426874876746e-05,
+      "loss": 0.9478,
+      "step": 8660
+    },
+    {
+      "epoch": 0.4420537398664151,
+      "grad_norm": 2.920138964629528,
+      "learning_rate": 6.171401353988553e-05,
+      "loss": 0.9235,
+      "step": 8670
+    },
+    {
+      "epoch": 0.442563605771682,
+      "grad_norm": 2.967699108126368,
+      "learning_rate": 6.163372638609179e-05,
+      "loss": 0.9343,
+      "step": 8680
+    },
+    {
+      "epoch": 0.44307347167694894,
+      "grad_norm": 2.7526313148316595,
+      "learning_rate": 6.155340750633476e-05,
+      "loss": 0.9425,
+      "step": 8690
+    },
+    {
+      "epoch": 0.44358333758221585,
+      "grad_norm": 4.045087410164233,
+      "learning_rate": 6.147305711964946e-05,
+      "loss": 0.9273,
+      "step": 8700
+    },
+    {
+      "epoch": 0.44409320348748277,
+      "grad_norm": 2.670856320943986,
+      "learning_rate": 6.139267544515689e-05,
+      "loss": 0.9904,
+      "step": 8710
+    },
+    {
+      "epoch": 0.4446030693927497,
+      "grad_norm": 3.9351489580031873,
+      "learning_rate": 6.131226270206332e-05,
+      "loss": 0.9996,
+      "step": 8720
+    },
+    {
+      "epoch": 0.4451129352980166,
+      "grad_norm": 5.942247351038859,
+      "learning_rate": 6.123181910965979e-05,
+      "loss": 0.9493,
+      "step": 8730
+    },
+    {
+      "epoch": 0.4456228012032835,
+      "grad_norm": 2.79740791815727,
+      "learning_rate": 6.115134488732143e-05,
+      "loss": 0.9022,
+      "step": 8740
+    },
+    {
+      "epoch": 0.44613266710855043,
+      "grad_norm": 3.4550178446015143,
+      "learning_rate": 6.107084025450693e-05,
+      "loss": 1.0,
+      "step": 8750
+    },
+    {
+      "epoch": 0.44664253301381734,
+      "grad_norm": 2.5156734942081553,
+      "learning_rate": 6.099030543075792e-05,
+      "loss": 0.9552,
+      "step": 8760
+    },
+    {
+      "epoch": 0.44715239891908426,
+      "grad_norm": 3.421439146165748,
+      "learning_rate": 6.090974063569832e-05,
+      "loss": 0.9912,
+      "step": 8770
+    },
+    {
+      "epoch": 0.4476622648243512,
+      "grad_norm": 2.8395317501384714,
+      "learning_rate": 6.082914608903382e-05,
+      "loss": 1.0194,
+      "step": 8780
+    },
+    {
+      "epoch": 0.4481721307296181,
+      "grad_norm": 3.107058432155688,
+      "learning_rate": 6.0748522010551215e-05,
+      "loss": 0.9673,
+      "step": 8790
+    },
+    {
+      "epoch": 0.448681996634885,
+      "grad_norm": 3.527942051818049,
+      "learning_rate": 6.066786862011785e-05,
+      "loss": 0.9556,
+      "step": 8800
+    },
+    {
+      "epoch": 0.4491918625401519,
+      "grad_norm": 3.6355047391327884,
+      "learning_rate": 6.058718613768103e-05,
+      "loss": 1.015,
+      "step": 8810
+    },
+    {
+      "epoch": 0.44970172844541884,
+      "grad_norm": 3.105033490007051,
+      "learning_rate": 6.050647478326736e-05,
+      "loss": 0.9699,
+      "step": 8820
+    },
+    {
+      "epoch": 0.45021159435068575,
+      "grad_norm": 2.5004745560856096,
+      "learning_rate": 6.0425734776982204e-05,
+      "loss": 0.9851,
+      "step": 8830
+    },
+    {
+      "epoch": 0.45072146025595267,
+      "grad_norm": 4.450216770629844,
+      "learning_rate": 6.034496633900903e-05,
+      "loss": 0.9703,
+      "step": 8840
+    },
+    {
+      "epoch": 0.4512313261612196,
+      "grad_norm": 4.197061587642375,
+      "learning_rate": 6.0264169689608886e-05,
+      "loss": 1.0017,
+      "step": 8850
+    },
+    {
+      "epoch": 0.4517411920664865,
+      "grad_norm": 2.6986562343418656,
+      "learning_rate": 6.018334504911971e-05,
+      "loss": 0.9502,
+      "step": 8860
+    },
+    {
+      "epoch": 0.4522510579717534,
+      "grad_norm": 3.0511897830278594,
+      "learning_rate": 6.010249263795582e-05,
+      "loss": 0.99,
+      "step": 8870
+    },
+    {
+      "epoch": 0.4527609238770203,
+      "grad_norm": 3.999753683504353,
+      "learning_rate": 6.002161267660723e-05,
+      "loss": 1.1003,
+      "step": 8880
+    },
+    {
+      "epoch": 0.45327078978228724,
+      "grad_norm": 4.933276773940487,
+      "learning_rate": 5.994070538563909e-05,
+      "loss": 1.0102,
+      "step": 8890
+    },
+    {
+      "epoch": 0.45378065568755416,
+      "grad_norm": 3.629194455464756,
+      "learning_rate": 5.985977098569111e-05,
+      "loss": 0.9832,
+      "step": 8900
+    },
+    {
+      "epoch": 0.4542905215928211,
+      "grad_norm": 3.204958560049805,
+      "learning_rate": 5.977880969747687e-05,
+      "loss": 0.9695,
+      "step": 8910
+    },
+    {
+      "epoch": 0.454800387498088,
+      "grad_norm": 2.6867147044360586,
+      "learning_rate": 5.969782174178333e-05,
+      "loss": 0.9656,
+      "step": 8920
+    },
+    {
+      "epoch": 0.4553102534033549,
+      "grad_norm": 3.034488103300888,
+      "learning_rate": 5.961680733947016e-05,
+      "loss": 1.0475,
+      "step": 8930
+    },
+    {
+      "epoch": 0.4558201193086218,
+      "grad_norm": 3.5236433196919985,
+      "learning_rate": 5.953576671146913e-05,
+      "loss": 0.981,
+      "step": 8940
+    },
+    {
+      "epoch": 0.45632998521388873,
+      "grad_norm": 2.8968669103912195,
+      "learning_rate": 5.945470007878356e-05,
+      "loss": 1.0058,
+      "step": 8950
+    },
+    {
+      "epoch": 0.45683985111915565,
+      "grad_norm": 3.1366228767198128,
+      "learning_rate": 5.937360766248767e-05,
+      "loss": 0.9852,
+      "step": 8960
+    },
+    {
+      "epoch": 0.45734971702442256,
+      "grad_norm": 2.897785110109335,
+      "learning_rate": 5.9292489683725996e-05,
+      "loss": 0.9616,
+      "step": 8970
+    },
+    {
+      "epoch": 0.4578595829296895,
+      "grad_norm": 4.21341205269237,
+      "learning_rate": 5.921134636371277e-05,
+      "loss": 0.9881,
+      "step": 8980
+    },
+    {
+      "epoch": 0.4583694488349564,
+      "grad_norm": 3.0001706706722926,
+      "learning_rate": 5.9130177923731376e-05,
+      "loss": 1.0046,
+      "step": 8990
+    },
+    {
+      "epoch": 0.4588793147402233,
+      "grad_norm": 5.748411307948442,
+      "learning_rate": 5.9048984585133646e-05,
+      "loss": 0.9856,
+      "step": 9000
+    },
+    {
+      "epoch": 0.4593891806454902,
+      "grad_norm": 2.9060378090177035,
+      "learning_rate": 5.896776656933936e-05,
+      "loss": 0.9629,
+      "step": 9010
+    },
+    {
+      "epoch": 0.45989904655075714,
+      "grad_norm": 2.9862141861239335,
+      "learning_rate": 5.8886524097835536e-05,
+      "loss": 0.9635,
+      "step": 9020
+    },
+    {
+      "epoch": 0.46040891245602406,
+      "grad_norm": 3.6461912452907668,
+      "learning_rate": 5.880525739217596e-05,
+      "loss": 1.0202,
+      "step": 9030
+    },
+    {
+      "epoch": 0.46091877836129097,
+      "grad_norm": 2.9584589049079453,
+      "learning_rate": 5.872396667398043e-05,
+      "loss": 0.9016,
+      "step": 9040
+    },
+    {
+      "epoch": 0.4614286442665579,
+      "grad_norm": 3.157536961051754,
+      "learning_rate": 5.86426521649343e-05,
+      "loss": 0.9571,
+      "step": 9050
+    },
+    {
+      "epoch": 0.4619385101718248,
+      "grad_norm": 3.1010631758883243,
+      "learning_rate": 5.856131408678776e-05,
+      "loss": 0.9395,
+      "step": 9060
+    },
+    {
+      "epoch": 0.4624483760770917,
+      "grad_norm": 2.92934162223403,
+      "learning_rate": 5.847995266135525e-05,
+      "loss": 0.9919,
+      "step": 9070
+    },
+    {
+      "epoch": 0.46295824198235863,
+      "grad_norm": 3.1196293404058792,
+      "learning_rate": 5.839856811051496e-05,
+      "loss": 1.013,
+      "step": 9080
+    },
+    {
+      "epoch": 0.46346810788762555,
+      "grad_norm": 4.47180888962446,
+      "learning_rate": 5.831716065620807e-05,
+      "loss": 0.916,
+      "step": 9090
+    },
+    {
+      "epoch": 0.46397797379289246,
+      "grad_norm": 3.6209260594149937,
+      "learning_rate": 5.8235730520438236e-05,
+      "loss": 1.0086,
+      "step": 9100
+    },
+    {
+      "epoch": 0.4644878396981594,
+      "grad_norm": 3.197672960765441,
+      "learning_rate": 5.815427792527102e-05,
+      "loss": 0.9278,
+      "step": 9110
+    },
+    {
+      "epoch": 0.4649977056034263,
+      "grad_norm": 3.8971007511057745,
+      "learning_rate": 5.8072803092833136e-05,
+      "loss": 1.0436,
+      "step": 9120
+    },
+    {
+      "epoch": 0.4655075715086932,
+      "grad_norm": 3.167182893353007,
+      "learning_rate": 5.7991306245312036e-05,
+      "loss": 0.9341,
+      "step": 9130
+    },
+    {
+      "epoch": 0.4660174374139601,
+      "grad_norm": 3.540484608219236,
+      "learning_rate": 5.7909787604955156e-05,
+      "loss": 0.9481,
+      "step": 9140
+    },
+    {
+      "epoch": 0.46652730331922704,
+      "grad_norm": 3.074758054177519,
+      "learning_rate": 5.782824739406938e-05,
+      "loss": 0.9173,
+      "step": 9150
+    },
+    {
+      "epoch": 0.46703716922449395,
+      "grad_norm": 3.1121036519918612,
+      "learning_rate": 5.774668583502037e-05,
+      "loss": 0.9751,
+      "step": 9160
+    },
+    {
+      "epoch": 0.46754703512976087,
+      "grad_norm": 4.404750026090895,
+      "learning_rate": 5.76651031502321e-05,
+      "loss": 0.9686,
+      "step": 9170
+    },
+    {
+      "epoch": 0.4680569010350278,
+      "grad_norm": 3.087422903467044,
+      "learning_rate": 5.758349956218607e-05,
+      "loss": 0.9619,
+      "step": 9180
+    },
+    {
+      "epoch": 0.4685667669402947,
+      "grad_norm": 2.5920682404244313,
+      "learning_rate": 5.750187529342082e-05,
+      "loss": 0.9127,
+      "step": 9190
+    },
+    {
+      "epoch": 0.4690766328455616,
+      "grad_norm": 3.2522582962238995,
+      "learning_rate": 5.742023056653131e-05,
+      "loss": 1.0514,
+      "step": 9200
+    },
+    {
+      "epoch": 0.46958649875082853,
+      "grad_norm": 2.151734209698684,
+      "learning_rate": 5.7338565604168236e-05,
+      "loss": 0.911,
+      "step": 9210
+    },
+    {
+      "epoch": 0.47009636465609544,
+      "grad_norm": 2.5390936845899215,
+      "learning_rate": 5.72568806290375e-05,
+      "loss": 1.0026,
+      "step": 9220
+    },
+    {
+      "epoch": 0.47060623056136236,
+      "grad_norm": 4.168993111095817,
+      "learning_rate": 5.717517586389961e-05,
+      "loss": 0.9082,
+      "step": 9230
+    },
+    {
+      "epoch": 0.4711160964666293,
+      "grad_norm": 3.35047102852213,
+      "learning_rate": 5.709345153156902e-05,
+      "loss": 0.9441,
+      "step": 9240
+    },
+    {
+      "epoch": 0.4716259623718962,
+      "grad_norm": 3.4495073915702936,
+      "learning_rate": 5.701170785491352e-05,
+      "loss": 0.9034,
+      "step": 9250
+    },
+    {
+      "epoch": 0.4721358282771631,
+      "grad_norm": 2.896625667681641,
+      "learning_rate": 5.692994505685369e-05,
+      "loss": 0.9081,
+      "step": 9260
+    },
+    {
+      "epoch": 0.47264569418243,
+      "grad_norm": 4.579982562564233,
+      "learning_rate": 5.684816336036224e-05,
+      "loss": 1.0853,
+      "step": 9270
+    },
+    {
+      "epoch": 0.47315556008769694,
+      "grad_norm": 4.355257890293027,
+      "learning_rate": 5.676636298846343e-05,
+      "loss": 0.986,
+      "step": 9280
+    },
+    {
+      "epoch": 0.47366542599296385,
+      "grad_norm": 4.107352245000783,
+      "learning_rate": 5.668454416423242e-05,
+      "loss": 0.9698,
+      "step": 9290
+    },
+    {
+      "epoch": 0.47417529189823077,
+      "grad_norm": 4.093149425772111,
+      "learning_rate": 5.660270711079474e-05,
+      "loss": 0.9642,
+      "step": 9300
+    },
+    {
+      "epoch": 0.4746851578034977,
+      "grad_norm": 5.297660098419542,
+      "learning_rate": 5.652085205132558e-05,
+      "loss": 0.9668,
+      "step": 9310
+    },
+    {
+      "epoch": 0.4751950237087646,
+      "grad_norm": 4.72644032568067,
+      "learning_rate": 5.643897920904926e-05,
+      "loss": 0.9844,
+      "step": 9320
+    },
+    {
+      "epoch": 0.4757048896140315,
+      "grad_norm": 3.7301453079447784,
+      "learning_rate": 5.635708880723858e-05,
+      "loss": 0.9383,
+      "step": 9330
+    },
+    {
+      "epoch": 0.4762147555192984,
+      "grad_norm": 3.0814586247871993,
+      "learning_rate": 5.627518106921425e-05,
+      "loss": 0.9842,
+      "step": 9340
+    },
+    {
+      "epoch": 0.47672462142456534,
+      "grad_norm": 5.42698297896619,
+      "learning_rate": 5.6193256218344246e-05,
+      "loss": 0.9641,
+      "step": 9350
+    },
+    {
+      "epoch": 0.47723448732983226,
+      "grad_norm": 5.152280972858124,
+      "learning_rate": 5.611131447804321e-05,
+      "loss": 0.9671,
+      "step": 9360
+    },
+    {
+      "epoch": 0.4777443532350992,
+      "grad_norm": 4.162051120491686,
+      "learning_rate": 5.602935607177182e-05,
+      "loss": 0.9838,
+      "step": 9370
+    },
+    {
+      "epoch": 0.4782542191403661,
+      "grad_norm": 3.559590998622765,
+      "learning_rate": 5.5947381223036244e-05,
+      "loss": 1.0307,
+      "step": 9380
+    },
+    {
+      "epoch": 0.478764085045633,
+      "grad_norm": 4.125612062475158,
+      "learning_rate": 5.5865390155387484e-05,
+      "loss": 0.9813,
+      "step": 9390
+    },
+    {
+      "epoch": 0.4792739509508999,
+      "grad_norm": 3.2131711610681792,
+      "learning_rate": 5.5783383092420746e-05,
+      "loss": 0.8639,
+      "step": 9400
+    },
+    {
+      "epoch": 0.47978381685616683,
+      "grad_norm": 3.2969378581814586,
+      "learning_rate": 5.5701360257774846e-05,
+      "loss": 0.9565,
+      "step": 9410
+    },
+    {
+      "epoch": 0.48029368276143375,
+      "grad_norm": 3.1462897733941895,
+      "learning_rate": 5.561932187513168e-05,
+      "loss": 0.9644,
+      "step": 9420
+    },
+    {
+      "epoch": 0.48080354866670066,
+      "grad_norm": 3.4176249243038344,
+      "learning_rate": 5.553726816821547e-05,
+      "loss": 0.9881,
+      "step": 9430
+    },
+    {
+      "epoch": 0.4813134145719676,
+      "grad_norm": 2.484582987435529,
+      "learning_rate": 5.5455199360792276e-05,
+      "loss": 0.9131,
+      "step": 9440
+    },
+    {
+      "epoch": 0.4818232804772345,
+      "grad_norm": 4.046107903525553,
+      "learning_rate": 5.5373115676669286e-05,
+      "loss": 1.027,
+      "step": 9450
+    },
+    {
+      "epoch": 0.4823331463825014,
+      "grad_norm": 2.7781539123813856,
+      "learning_rate": 5.529101733969432e-05,
+      "loss": 0.9215,
+      "step": 9460
+    },
+    {
+      "epoch": 0.4828430122877683,
+      "grad_norm": 4.07554206961546,
+      "learning_rate": 5.5208904573755104e-05,
+      "loss": 0.9881,
+      "step": 9470
+    },
+    {
+      "epoch": 0.48335287819303524,
+      "grad_norm": 3.3271161274855476,
+      "learning_rate": 5.5126777602778765e-05,
+      "loss": 0.9697,
+      "step": 9480
+    },
+    {
+      "epoch": 0.48386274409830216,
+      "grad_norm": 2.6603039051771638,
+      "learning_rate": 5.504463665073112e-05,
+      "loss": 0.9197,
+      "step": 9490
+    },
+    {
+      "epoch": 0.48437261000356907,
+      "grad_norm": 2.9061379748402443,
+      "learning_rate": 5.496248194161615e-05,
+      "loss": 0.9588,
+      "step": 9500
+    },
+    {
+      "epoch": 0.484882475908836,
+      "grad_norm": 3.3355590411210168,
+      "learning_rate": 5.4880313699475314e-05,
+      "loss": 1.036,
+      "step": 9510
+    },
+    {
+      "epoch": 0.4853923418141029,
+      "grad_norm": 3.451459158559493,
+      "learning_rate": 5.4798132148387013e-05,
+      "loss": 1.0624,
+      "step": 9520
+    },
+    {
+      "epoch": 0.4859022077193698,
+      "grad_norm": 2.3258525703065063,
+      "learning_rate": 5.471593751246592e-05,
+      "loss": 0.9248,
+      "step": 9530
+    },
+    {
+      "epoch": 0.48641207362463673,
+      "grad_norm": 2.8863302547945318,
+      "learning_rate": 5.4633730015862414e-05,
+      "loss": 0.9438,
+      "step": 9540
+    },
+    {
+      "epoch": 0.48692193952990365,
+      "grad_norm": 3.592383771973474,
+      "learning_rate": 5.455150988276192e-05,
+      "loss": 0.9935,
+      "step": 9550
+    },
+    {
+      "epoch": 0.48743180543517056,
+      "grad_norm": 3.5051333487356624,
+      "learning_rate": 5.446927733738433e-05,
+      "loss": 0.9247,
+      "step": 9560
+    },
+    {
+      "epoch": 0.4879416713404375,
+      "grad_norm": 2.6807353943018435,
+      "learning_rate": 5.43870326039834e-05,
+      "loss": 0.9739,
+      "step": 9570
+    },
+    {
+      "epoch": 0.4884515372457044,
+      "grad_norm": 4.305074246806519,
+      "learning_rate": 5.43047759068461e-05,
+      "loss": 0.9822,
+      "step": 9580
+    },
+    {
+      "epoch": 0.4889614031509713,
+      "grad_norm": 2.354641769723023,
+      "learning_rate": 5.4222507470292036e-05,
+      "loss": 0.8612,
+      "step": 9590
+    },
+    {
+      "epoch": 0.4894712690562382,
+      "grad_norm": 2.965498591324047,
+      "learning_rate": 5.414022751867285e-05,
+      "loss": 0.9204,
+      "step": 9600
+    },
+    {
+      "epoch": 0.48998113496150514,
+      "grad_norm": 6.3205482558893005,
+      "learning_rate": 5.4057936276371565e-05,
+      "loss": 0.9709,
+      "step": 9610
+    },
+    {
+      "epoch": 0.49049100086677205,
+      "grad_norm": 2.010926865518373,
+      "learning_rate": 5.3975633967801986e-05,
+      "loss": 0.9492,
+      "step": 9620
+    },
+    {
+      "epoch": 0.49100086677203897,
+      "grad_norm": 3.452431723535906,
+      "learning_rate": 5.389332081740812e-05,
+      "loss": 0.9015,
+      "step": 9630
+    },
+    {
+      "epoch": 0.4915107326773059,
+      "grad_norm": 5.034933917396801,
+      "learning_rate": 5.381099704966352e-05,
+      "loss": 0.9986,
+      "step": 9640
+    },
+    {
+      "epoch": 0.4920205985825728,
+      "grad_norm": 4.612910015735674,
+      "learning_rate": 5.372866288907069e-05,
+      "loss": 0.959,
+      "step": 9650
+    },
+    {
+      "epoch": 0.4925304644878397,
+      "grad_norm": 3.5663763294504522,
+      "learning_rate": 5.364631856016051e-05,
+      "loss": 0.9167,
+      "step": 9660
+    },
+    {
+      "epoch": 0.49304033039310663,
+      "grad_norm": 2.7699815026050114,
+      "learning_rate": 5.356396428749155e-05,
+      "loss": 0.9322,
+      "step": 9670
+    },
+    {
+      "epoch": 0.49355019629837354,
+      "grad_norm": 5.781944370641614,
+      "learning_rate": 5.3481600295649525e-05,
+      "loss": 0.9052,
+      "step": 9680
+    },
+    {
+      "epoch": 0.49406006220364046,
+      "grad_norm": 2.895890141583381,
+      "learning_rate": 5.339922680924664e-05,
+      "loss": 1.0033,
+      "step": 9690
+    },
+    {
+      "epoch": 0.4945699281089074,
+      "grad_norm": 2.639129524416714,
+      "learning_rate": 5.331684405292099e-05,
+      "loss": 0.9316,
+      "step": 9700
+    },
+    {
+      "epoch": 0.4950797940141743,
+      "grad_norm": 3.8489717935083454,
+      "learning_rate": 5.3234452251335955e-05,
+      "loss": 0.978,
+      "step": 9710
+    },
+    {
+      "epoch": 0.4955896599194412,
+      "grad_norm": 4.641154634386726,
+      "learning_rate": 5.31520516291796e-05,
+      "loss": 0.9366,
+      "step": 9720
+    },
+    {
+      "epoch": 0.4960995258247081,
+      "grad_norm": 4.1945884250295125,
+      "learning_rate": 5.306964241116402e-05,
+      "loss": 0.9642,
+      "step": 9730
+    },
+    {
+      "epoch": 0.49660939172997504,
+      "grad_norm": 3.804927333278639,
+      "learning_rate": 5.298722482202475e-05,
+      "loss": 0.9141,
+      "step": 9740
+    },
+    {
+      "epoch": 0.49711925763524195,
+      "grad_norm": 3.3137262873679245,
+      "learning_rate": 5.290479908652017e-05,
+      "loss": 0.9532,
+      "step": 9750
+    },
+    {
+      "epoch": 0.49762912354050887,
+      "grad_norm": 3.598504231825922,
+      "learning_rate": 5.2822365429430866e-05,
+      "loss": 0.913,
+      "step": 9760
+    },
+    {
+      "epoch": 0.4981389894457758,
+      "grad_norm": 2.922676556869131,
+      "learning_rate": 5.2739924075559024e-05,
+      "loss": 0.9417,
+      "step": 9770
+    },
+    {
+      "epoch": 0.4986488553510427,
+      "grad_norm": 4.35078006858887,
+      "learning_rate": 5.265747524972784e-05,
+      "loss": 0.9719,
+      "step": 9780
+    },
+    {
+      "epoch": 0.4991587212563096,
+      "grad_norm": 2.5548950063605846,
+      "learning_rate": 5.257501917678085e-05,
+      "loss": 0.9239,
+      "step": 9790
+    },
+    {
+      "epoch": 0.4996685871615765,
+      "grad_norm": 6.836250183279308,
+      "learning_rate": 5.249255608158139e-05,
+      "loss": 0.9607,
+      "step": 9800
+    },
+    {
+      "epoch": 0.5001784530668434,
+      "grad_norm": 2.9653175822252567,
+      "learning_rate": 5.241008618901192e-05,
+      "loss": 0.9716,
+      "step": 9810
+    },
+    {
+      "epoch": 0.5006883189721103,
+      "grad_norm": 3.734122705555406,
+      "learning_rate": 5.232760972397347e-05,
+      "loss": 0.9289,
+      "step": 9820
+    },
+    {
+      "epoch": 0.5011981848773772,
+      "grad_norm": 2.469867186688575,
+      "learning_rate": 5.224512691138494e-05,
+      "loss": 0.9609,
+      "step": 9830
+    },
+    {
+      "epoch": 0.5017080507826441,
+      "grad_norm": 3.5526576361576017,
+      "learning_rate": 5.2162637976182584e-05,
+      "loss": 0.9721,
+      "step": 9840
+    },
+    {
+      "epoch": 0.502217916687911,
+      "grad_norm": 3.9133632073025635,
+      "learning_rate": 5.208014314331935e-05,
+      "loss": 0.9317,
+      "step": 9850
+    },
+    {
+      "epoch": 0.502727782593178,
+      "grad_norm": 3.2917362546113615,
+      "learning_rate": 5.1997642637764255e-05,
+      "loss": 0.9953,
+      "step": 9860
+    },
+    {
+      "epoch": 0.5032376484984449,
+      "grad_norm": 4.711424755994914,
+      "learning_rate": 5.191513668450178e-05,
+      "loss": 0.9486,
+      "step": 9870
+    },
+    {
+      "epoch": 0.5037475144037118,
+      "grad_norm": 2.321341822761805,
+      "learning_rate": 5.183262550853127e-05,
+      "loss": 0.9446,
+      "step": 9880
+    },
+    {
+      "epoch": 0.5042573803089787,
+      "grad_norm": 2.786223507865711,
+      "learning_rate": 5.175010933486634e-05,
+      "loss": 0.9548,
+      "step": 9890
+    },
+    {
+      "epoch": 0.5047672462142456,
+      "grad_norm": 3.6083918636921695,
+      "learning_rate": 5.166758838853418e-05,
+      "loss": 0.9483,
+      "step": 9900
+    },
+    {
+      "epoch": 0.5052771121195125,
+      "grad_norm": 2.3338798016112023,
+      "learning_rate": 5.1585062894575034e-05,
+      "loss": 0.9208,
+      "step": 9910
+    },
+    {
+      "epoch": 0.5057869780247795,
+      "grad_norm": 2.6959985130224955,
+      "learning_rate": 5.150253307804156e-05,
+      "loss": 0.8596,
+      "step": 9920
+    },
+    {
+      "epoch": 0.5062968439300464,
+      "grad_norm": 3.5257109643018723,
+      "learning_rate": 5.141999916399816e-05,
+      "loss": 0.97,
+      "step": 9930
+    },
+    {
+      "epoch": 0.5068067098353133,
+      "grad_norm": 3.082519782660571,
+      "learning_rate": 5.133746137752044e-05,
+      "loss": 0.9658,
+      "step": 9940
+    },
+    {
+      "epoch": 0.5073165757405802,
+      "grad_norm": 2.561134174703815,
+      "learning_rate": 5.1254919943694526e-05,
+      "loss": 0.9268,
+      "step": 9950
+    },
+    {
+      "epoch": 0.5078264416458471,
+      "grad_norm": 3.7181675274604706,
+      "learning_rate": 5.1172375087616576e-05,
+      "loss": 0.9573,
+      "step": 9960
+    },
+    {
+      "epoch": 0.508336307551114,
+      "grad_norm": 2.795604150973193,
+      "learning_rate": 5.108982703439201e-05,
+      "loss": 1.0011,
+      "step": 9970
+    },
+    {
+      "epoch": 0.508846173456381,
+      "grad_norm": 2.9016638978849163,
+      "learning_rate": 5.1007276009134976e-05,
+      "loss": 0.9783,
+      "step": 9980
+    },
+    {
+      "epoch": 0.5093560393616479,
+      "grad_norm": 2.7194359601726084,
+      "learning_rate": 5.092472223696774e-05,
+      "loss": 0.9422,
+      "step": 9990
+    },
+    {
+      "epoch": 0.5098659052669148,
+      "grad_norm": 4.318325714358832,
+      "learning_rate": 5.0842165943020046e-05,
+      "loss": 1.0178,
+      "step": 10000
+    },
+    {
+      "epoch": 0.5103757711721817,
+      "grad_norm": 3.6852911837265285,
+      "learning_rate": 5.075960735242854e-05,
+      "loss": 0.9389,
+      "step": 10010
+    },
+    {
+      "epoch": 0.5108856370774486,
+      "grad_norm": 2.795966314706265,
+      "learning_rate": 5.0677046690336096e-05,
+      "loss": 0.9737,
+      "step": 10020
+    },
+    {
+      "epoch": 0.5113955029827155,
+      "grad_norm": 2.7930043328638523,
+      "learning_rate": 5.059448418189128e-05,
+      "loss": 0.9496,
+      "step": 10030
+    },
+    {
+      "epoch": 0.5119053688879824,
+      "grad_norm": 3.062256031105531,
+      "learning_rate": 5.051192005224765e-05,
+      "loss": 0.9023,
+      "step": 10040
+    },
+    {
+      "epoch": 0.5124152347932494,
+      "grad_norm": 6.780172533538299,
+      "learning_rate": 5.042935452656322e-05,
+      "loss": 1.0067,
+      "step": 10050
+    },
+    {
+      "epoch": 0.5129251006985163,
+      "grad_norm": 2.776353296984024,
+      "learning_rate": 5.034678782999979e-05,
+      "loss": 0.9937,
+      "step": 10060
+    },
+    {
+      "epoch": 0.5134349666037832,
+      "grad_norm": 3.3365394053880335,
+      "learning_rate": 5.026422018772234e-05,
+      "loss": 0.8942,
+      "step": 10070
+    },
+    {
+      "epoch": 0.5139448325090501,
+      "grad_norm": 2.3409195569526515,
+      "learning_rate": 5.0181651824898466e-05,
+      "loss": 0.9252,
+      "step": 10080
+    },
+    {
+      "epoch": 0.514454698414317,
+      "grad_norm": 3.879946089564467,
+      "learning_rate": 5.009908296669772e-05,
+      "loss": 0.921,
+      "step": 10090
+    },
+    {
+      "epoch": 0.5149645643195839,
+      "grad_norm": 5.206276131399787,
+      "learning_rate": 5.001651383829098e-05,
+      "loss": 0.9434,
+      "step": 10100
+    },
+    {
+      "epoch": 0.5154744302248508,
+      "grad_norm": 3.240935073507238,
+      "learning_rate": 4.993394466484987e-05,
+      "loss": 0.9842,
+      "step": 10110
+    },
+    {
+      "epoch": 0.5159842961301178,
+      "grad_norm": 5.4587787978074935,
+      "learning_rate": 4.985137567154616e-05,
+      "loss": 0.9391,
+      "step": 10120
+    },
+    {
+      "epoch": 0.5164941620353847,
+      "grad_norm": 3.84414031953169,
+      "learning_rate": 4.97688070835511e-05,
+      "loss": 0.9602,
+      "step": 10130
+    },
+    {
+      "epoch": 0.5170040279406516,
+      "grad_norm": 4.705896233937508,
+      "learning_rate": 4.968623912603485e-05,
+      "loss": 0.9239,
+      "step": 10140
+    },
+    {
+      "epoch": 0.5175138938459185,
+      "grad_norm": 4.205557321794439,
+      "learning_rate": 4.960367202416587e-05,
+      "loss": 0.9746,
+      "step": 10150
+    },
+    {
+      "epoch": 0.5180237597511854,
+      "grad_norm": 2.744594544456746,
+      "learning_rate": 4.9521106003110216e-05,
+      "loss": 0.9,
+      "step": 10160
+    },
+    {
+      "epoch": 0.5185336256564523,
+      "grad_norm": 2.2880730895604975,
+      "learning_rate": 4.943854128803106e-05,
+      "loss": 0.9323,
+      "step": 10170
+    },
+    {
+      "epoch": 0.5190434915617193,
+      "grad_norm": 4.541253564776076,
+      "learning_rate": 4.935597810408804e-05,
+      "loss": 1.0212,
+      "step": 10180
+    },
+    {
+      "epoch": 0.5195533574669862,
+      "grad_norm": 3.317522442961713,
+      "learning_rate": 4.9273416676436526e-05,
+      "loss": 0.9651,
+      "step": 10190
+    },
+    {
+      "epoch": 0.5200632233722531,
+      "grad_norm": 4.626442522093559,
+      "learning_rate": 4.9190857230227186e-05,
+      "loss": 0.9401,
+      "step": 10200
+    },
+    {
+      "epoch": 0.52057308927752,
+      "grad_norm": 4.70036240305505,
+      "learning_rate": 4.9108299990605204e-05,
+      "loss": 0.9635,
+      "step": 10210
+    },
+    {
+      "epoch": 0.5210829551827869,
+      "grad_norm": 2.761068503299125,
+      "learning_rate": 4.902574518270982e-05,
+      "loss": 0.9852,
+      "step": 10220
+    },
+    {
+      "epoch": 0.5215928210880538,
+      "grad_norm": 2.656996733011883,
+      "learning_rate": 4.894319303167359e-05,
+      "loss": 0.88,
+      "step": 10230
+    },
+    {
+      "epoch": 0.5221026869933207,
+      "grad_norm": 4.221748168653935,
+      "learning_rate": 4.8860643762621854e-05,
+      "loss": 1.0218,
+      "step": 10240
+    },
+    {
+      "epoch": 0.5226125528985877,
+      "grad_norm": 3.03526122402303,
+      "learning_rate": 4.877809760067211e-05,
+      "loss": 0.8784,
+      "step": 10250
+    },
+    {
+      "epoch": 0.5231224188038546,
+      "grad_norm": 3.2804201533463493,
+      "learning_rate": 4.869555477093332e-05,
+      "loss": 0.9657,
+      "step": 10260
+    },
+    {
+      "epoch": 0.5236322847091215,
+      "grad_norm": 3.7269652474464205,
+      "learning_rate": 4.861301549850543e-05,
+      "loss": 0.9414,
+      "step": 10270
+    },
+    {
+      "epoch": 0.5241421506143884,
+      "grad_norm": 4.796750509047648,
+      "learning_rate": 4.8530480008478614e-05,
+      "loss": 0.9575,
+      "step": 10280
+    },
+    {
+      "epoch": 0.5246520165196553,
+      "grad_norm": 4.519107152482081,
+      "learning_rate": 4.8447948525932785e-05,
+      "loss": 0.9862,
+      "step": 10290
+    },
+    {
+      "epoch": 0.5251618824249222,
+      "grad_norm": 3.426639454230538,
+      "learning_rate": 4.836542127593693e-05,
+      "loss": 0.9354,
+      "step": 10300
+    },
+    {
+      "epoch": 0.5256717483301891,
+      "grad_norm": 3.5190702566578005,
+      "learning_rate": 4.8282898483548436e-05,
+      "loss": 0.8566,
+      "step": 10310
+    },
+    {
+      "epoch": 0.5261816142354561,
+      "grad_norm": 2.6438908663160707,
+      "learning_rate": 4.8200380373812606e-05,
+      "loss": 0.9285,
+      "step": 10320
+    },
+    {
+      "epoch": 0.526691480140723,
+      "grad_norm": 2.754499762740791,
+      "learning_rate": 4.8117867171761916e-05,
+      "loss": 0.8905,
+      "step": 10330
+    },
+    {
+      "epoch": 0.5272013460459899,
+      "grad_norm": 3.4618334631287504,
+      "learning_rate": 4.803535910241551e-05,
+      "loss": 0.9515,
+      "step": 10340
+    },
+    {
+      "epoch": 0.5277112119512568,
+      "grad_norm": 3.448601636720122,
+      "learning_rate": 4.7952856390778466e-05,
+      "loss": 0.8888,
+      "step": 10350
+    },
+    {
+      "epoch": 0.5282210778565237,
+      "grad_norm": 3.8156484134863065,
+      "learning_rate": 4.787035926184131e-05,
+      "loss": 0.9402,
+      "step": 10360
+    },
+    {
+      "epoch": 0.5287309437617906,
+      "grad_norm": 3.5299280212248156,
+      "learning_rate": 4.7787867940579354e-05,
+      "loss": 0.9026,
+      "step": 10370
+    },
+    {
+      "epoch": 0.5292408096670576,
+      "grad_norm": 4.2496355214292185,
+      "learning_rate": 4.770538265195199e-05,
+      "loss": 0.9967,
+      "step": 10380
+    },
+    {
+      "epoch": 0.5297506755723245,
+      "grad_norm": 3.0275611877217057,
+      "learning_rate": 4.7622903620902265e-05,
+      "loss": 0.9648,
+      "step": 10390
+    },
+    {
+      "epoch": 0.5302605414775914,
+      "grad_norm": 3.2138186040832246,
+      "learning_rate": 4.7540431072356056e-05,
+      "loss": 0.9205,
+      "step": 10400
+    },
+    {
+      "epoch": 0.5307704073828583,
+      "grad_norm": 3.006892944145669,
+      "learning_rate": 4.7457965231221634e-05,
+      "loss": 1.0059,
+      "step": 10410
+    },
+    {
+      "epoch": 0.5312802732881252,
+      "grad_norm": 2.5735484280254424,
+      "learning_rate": 4.737550632238899e-05,
+      "loss": 0.9398,
+      "step": 10420
+    },
+    {
+      "epoch": 0.5317901391933921,
+      "grad_norm": 2.717086537919467,
+      "learning_rate": 4.729305457072913e-05,
+      "loss": 0.8851,
+      "step": 10430
+    },
+    {
+      "epoch": 0.532300005098659,
+      "grad_norm": 3.211894198936536,
+      "learning_rate": 4.7210610201093617e-05,
+      "loss": 0.9396,
+      "step": 10440
+    },
+    {
+      "epoch": 0.532809871003926,
+      "grad_norm": 4.919270004384732,
+      "learning_rate": 4.712817343831384e-05,
+      "loss": 0.9608,
+      "step": 10450
+    },
+    {
+      "epoch": 0.5333197369091929,
+      "grad_norm": 2.397973165692253,
+      "learning_rate": 4.7045744507200486e-05,
+      "loss": 0.9257,
+      "step": 10460
+    },
+    {
+      "epoch": 0.5338296028144598,
+      "grad_norm": 2.8041603830628987,
+      "learning_rate": 4.696332363254282e-05,
+      "loss": 0.9581,
+      "step": 10470
+    },
+    {
+      "epoch": 0.5343394687197267,
+      "grad_norm": 7.021143848255496,
+      "learning_rate": 4.688091103910819e-05,
+      "loss": 0.9404,
+      "step": 10480
+    },
+    {
+      "epoch": 0.5348493346249936,
+      "grad_norm": 2.0721120953637646,
+      "learning_rate": 4.679850695164138e-05,
+      "loss": 0.8397,
+      "step": 10490
+    },
+    {
+      "epoch": 0.5353592005302605,
+      "grad_norm": 2.520850181356826,
+      "learning_rate": 4.671611159486389e-05,
+      "loss": 0.9275,
+      "step": 10500
+    },
+    {
+      "epoch": 0.5358690664355275,
+      "grad_norm": 2.5231701025343245,
+      "learning_rate": 4.66337251934735e-05,
+      "loss": 0.951,
+      "step": 10510
+    },
+    {
+      "epoch": 0.5363789323407944,
+      "grad_norm": 3.8940168673136784,
+      "learning_rate": 4.65513479721435e-05,
+      "loss": 0.9202,
+      "step": 10520
+    },
+    {
+      "epoch": 0.5368887982460613,
+      "grad_norm": 2.5217152560237737,
+      "learning_rate": 4.646898015552219e-05,
+      "loss": 0.8935,
+      "step": 10530
+    },
+    {
+      "epoch": 0.5373986641513282,
+      "grad_norm": 2.633379640768002,
+      "learning_rate": 4.638662196823223e-05,
+      "loss": 0.9552,
+      "step": 10540
+    },
+    {
+      "epoch": 0.5379085300565951,
+      "grad_norm": 3.610584731612508,
+      "learning_rate": 4.630427363486996e-05,
+      "loss": 0.9558,
+      "step": 10550
+    },
+    {
+      "epoch": 0.538418395961862,
+      "grad_norm": 3.6451579333644255,
+      "learning_rate": 4.622193538000494e-05,
+      "loss": 0.959,
+      "step": 10560
+    },
+    {
+      "epoch": 0.5389282618671289,
+      "grad_norm": 2.73070710265327,
+      "learning_rate": 4.613960742817914e-05,
+      "loss": 0.9415,
+      "step": 10570
+    },
+    {
+      "epoch": 0.5394381277723959,
+      "grad_norm": 3.2767593604993404,
+      "learning_rate": 4.6057290003906526e-05,
+      "loss": 0.8898,
+      "step": 10580
+    },
+    {
+      "epoch": 0.5399479936776628,
+      "grad_norm": 3.4661086987132634,
+      "learning_rate": 4.597498333167227e-05,
+      "loss": 0.9227,
+      "step": 10590
+    },
+    {
+      "epoch": 0.5404578595829297,
+      "grad_norm": 3.233418939753378,
+      "learning_rate": 4.58926876359323e-05,
+      "loss": 0.9685,
+      "step": 10600
+    },
+    {
+      "epoch": 0.5409677254881966,
+      "grad_norm": 3.1363887735424933,
+      "learning_rate": 4.581040314111259e-05,
+      "loss": 0.9405,
+      "step": 10610
+    },
+    {
+      "epoch": 0.5414775913934635,
+      "grad_norm": 2.83491402664726,
+      "learning_rate": 4.5728130071608506e-05,
+      "loss": 0.9299,
+      "step": 10620
+    },
+    {
+      "epoch": 0.5419874572987304,
+      "grad_norm": 6.002022657603547,
+      "learning_rate": 4.5645868651784354e-05,
+      "loss": 0.9689,
+      "step": 10630
+    },
+    {
+      "epoch": 0.5424973232039974,
+      "grad_norm": 3.247394883373861,
+      "learning_rate": 4.5563619105972594e-05,
+      "loss": 0.9417,
+      "step": 10640
+    },
+    {
+      "epoch": 0.5430071891092643,
+      "grad_norm": 3.1962757505757726,
+      "learning_rate": 4.5481381658473346e-05,
+      "loss": 0.9484,
+      "step": 10650
+    },
+    {
+      "epoch": 0.5435170550145312,
+      "grad_norm": 3.64320500753679,
+      "learning_rate": 4.53991565335537e-05,
+      "loss": 0.9515,
+      "step": 10660
+    },
+    {
+      "epoch": 0.5440269209197981,
+      "grad_norm": 3.7059387220241513,
+      "learning_rate": 4.531694395544719e-05,
+      "loss": 0.9219,
+      "step": 10670
+    },
+    {
+      "epoch": 0.544536786825065,
+      "grad_norm": 3.375884864641275,
+      "learning_rate": 4.523474414835313e-05,
+      "loss": 0.9091,
+      "step": 10680
+    },
+    {
+      "epoch": 0.5450466527303319,
+      "grad_norm": 3.5558570711149344,
+      "learning_rate": 4.515255733643593e-05,
+      "loss": 0.9136,
+      "step": 10690
+    },
+    {
+      "epoch": 0.5455565186355988,
+      "grad_norm": 4.808820482356941,
+      "learning_rate": 4.5070383743824674e-05,
+      "loss": 0.9184,
+      "step": 10700
+    },
+    {
+      "epoch": 0.5460663845408658,
+      "grad_norm": 2.6218298124518062,
+      "learning_rate": 4.498822359461229e-05,
+      "loss": 0.9326,
+      "step": 10710
+    },
+    {
+      "epoch": 0.5465762504461327,
+      "grad_norm": 2.42543674839897,
+      "learning_rate": 4.4906077112855116e-05,
+      "loss": 1.0058,
+      "step": 10720
+    },
+    {
+      "epoch": 0.5470861163513996,
+      "grad_norm": 3.1361509539515393,
+      "learning_rate": 4.4823944522572216e-05,
+      "loss": 0.8866,
+      "step": 10730
+    },
+    {
+      "epoch": 0.5475959822566665,
+      "grad_norm": 3.6852645001813085,
+      "learning_rate": 4.474182604774471e-05,
+      "loss": 1.0186,
+      "step": 10740
+    },
+    {
+      "epoch": 0.5481058481619334,
+      "grad_norm": 3.4131327708149852,
+      "learning_rate": 4.465972191231531e-05,
+      "loss": 0.9117,
+      "step": 10750
+    },
+    {
+      "epoch": 0.5486157140672003,
+      "grad_norm": 3.191893783018942,
+      "learning_rate": 4.457763234018753e-05,
+      "loss": 0.9349,
+      "step": 10760
+    },
+    {
+      "epoch": 0.5491255799724672,
+      "grad_norm": 2.544643243504104,
+      "learning_rate": 4.4495557555225276e-05,
+      "loss": 0.8535,
+      "step": 10770
+    },
+    {
+      "epoch": 0.5496354458777342,
+      "grad_norm": 2.51958458544114,
+      "learning_rate": 4.441349778125201e-05,
+      "loss": 0.9941,
+      "step": 10780
+    },
+    {
+      "epoch": 0.5501453117830011,
+      "grad_norm": 4.107901228678458,
+      "learning_rate": 4.433145324205035e-05,
+      "loss": 0.9923,
+      "step": 10790
+    },
+    {
+      "epoch": 0.550655177688268,
+      "grad_norm": 3.471936059851544,
+      "learning_rate": 4.424942416136136e-05,
+      "loss": 0.9084,
+      "step": 10800
+    },
+    {
+      "epoch": 0.5511650435935349,
+      "grad_norm": 4.483957358898498,
+      "learning_rate": 4.4167410762883863e-05,
+      "loss": 0.9059,
+      "step": 10810
+    },
+    {
+      "epoch": 0.5516749094988018,
+      "grad_norm": 2.6148774956085195,
+      "learning_rate": 4.408541327027404e-05,
+      "loss": 0.9014,
+      "step": 10820
+    },
+    {
+      "epoch": 0.5521847754040687,
+      "grad_norm": 2.5513218383095064,
+      "learning_rate": 4.4003431907144576e-05,
+      "loss": 0.9092,
+      "step": 10830
+    },
+    {
+      "epoch": 0.5526946413093357,
+      "grad_norm": 3.2935364006275756,
+      "learning_rate": 4.392146689706425e-05,
+      "loss": 0.886,
+      "step": 10840
+    },
+    {
+      "epoch": 0.5532045072146026,
+      "grad_norm": 8.626314413796218,
+      "learning_rate": 4.3839518463557234e-05,
+      "loss": 0.9992,
+      "step": 10850
+    },
+    {
+      "epoch": 0.5537143731198695,
+      "grad_norm": 4.180141893470191,
+      "learning_rate": 4.3757586830102464e-05,
+      "loss": 0.9177,
+      "step": 10860
+    },
+    {
+      "epoch": 0.5542242390251364,
+      "grad_norm": 3.7520943854550435,
+      "learning_rate": 4.3675672220133085e-05,
+      "loss": 0.9843,
+      "step": 10870
+    },
+    {
+      "epoch": 0.5547341049304033,
+      "grad_norm": 3.529144886650154,
+      "learning_rate": 4.3593774857035795e-05,
+      "loss": 1.0294,
+      "step": 10880
+    },
+    {
+      "epoch": 0.5552439708356702,
+      "grad_norm": 3.8101858885076125,
+      "learning_rate": 4.351189496415031e-05,
+      "loss": 0.8928,
+      "step": 10890
+    },
+    {
+      "epoch": 0.5557538367409371,
+      "grad_norm": 4.248885275921915,
+      "learning_rate": 4.343003276476862e-05,
+      "loss": 0.952,
+      "step": 10900
+    },
+    {
+      "epoch": 0.5562637026462041,
+      "grad_norm": 3.213356356055621,
+      "learning_rate": 4.334818848213455e-05,
+      "loss": 0.8553,
+      "step": 10910
+    },
+    {
+      "epoch": 0.556773568551471,
+      "grad_norm": 3.121971553205506,
+      "learning_rate": 4.326636233944303e-05,
+      "loss": 0.9389,
+      "step": 10920
+    },
+    {
+      "epoch": 0.5572834344567379,
+      "grad_norm": 2.2812367647348135,
+      "learning_rate": 4.318455455983949e-05,
+      "loss": 0.9203,
+      "step": 10930
+    },
+    {
+      "epoch": 0.5577933003620048,
+      "grad_norm": 2.7159153450702718,
+      "learning_rate": 4.310276536641932e-05,
+      "loss": 0.939,
+      "step": 10940
+    },
+    {
+      "epoch": 0.5583031662672717,
+      "grad_norm": 3.606694364467837,
+      "learning_rate": 4.302099498222722e-05,
+      "loss": 1.0178,
+      "step": 10950
+    },
+    {
+      "epoch": 0.5588130321725386,
+      "grad_norm": 5.270098055897755,
+      "learning_rate": 4.2939243630256584e-05,
+      "loss": 0.987,
+      "step": 10960
+    },
+    {
+      "epoch": 0.5593228980778056,
+      "grad_norm": 3.4379590615846762,
+      "learning_rate": 4.285751153344892e-05,
+      "loss": 0.9356,
+      "step": 10970
+    },
+    {
+      "epoch": 0.5598327639830725,
+      "grad_norm": 3.3452241656556767,
+      "learning_rate": 4.27757989146932e-05,
+      "loss": 0.9208,
+      "step": 10980
+    },
+    {
+      "epoch": 0.5603426298883394,
+      "grad_norm": 3.4990502423669043,
+      "learning_rate": 4.2694105996825325e-05,
+      "loss": 0.935,
+      "step": 10990
+    },
+    {
+      "epoch": 0.5608524957936063,
+      "grad_norm": 3.3464667008325955,
+      "learning_rate": 4.2612433002627404e-05,
+      "loss": 0.958,
+      "step": 11000
+    },
+    {
+      "epoch": 0.5613623616988732,
+      "grad_norm": 3.577841867130007,
+      "learning_rate": 4.2530780154827284e-05,
+      "loss": 0.8888,
+      "step": 11010
+    },
+    {
+      "epoch": 0.5618722276041401,
+      "grad_norm": 2.919441037595988,
+      "learning_rate": 4.244914767609779e-05,
+      "loss": 0.927,
+      "step": 11020
+    },
+    {
+      "epoch": 0.562382093509407,
+      "grad_norm": 2.754880039651409,
+      "learning_rate": 4.236753578905627e-05,
+      "loss": 0.9013,
+      "step": 11030
+    },
+    {
+      "epoch": 0.562891959414674,
+      "grad_norm": 3.2188275287215142,
+      "learning_rate": 4.228594471626392e-05,
+      "loss": 0.9431,
+      "step": 11040
+    },
+    {
+      "epoch": 0.5634018253199409,
+      "grad_norm": 3.1991664938547992,
+      "learning_rate": 4.22043746802251e-05,
+      "loss": 0.8902,
+      "step": 11050
+    },
+    {
+      "epoch": 0.5639116912252078,
+      "grad_norm": 3.1340245290870055,
+      "learning_rate": 4.2122825903386874e-05,
+      "loss": 0.9363,
+      "step": 11060
+    },
+    {
+      "epoch": 0.5644215571304747,
+      "grad_norm": 2.499072366200033,
+      "learning_rate": 4.204129860813828e-05,
+      "loss": 0.8899,
+      "step": 11070
+    },
+    {
+      "epoch": 0.5649314230357416,
+      "grad_norm": 3.367362167872141,
+      "learning_rate": 4.1959793016809835e-05,
+      "loss": 0.9268,
+      "step": 11080
+    },
+    {
+      "epoch": 0.5654412889410085,
+      "grad_norm": 4.055317211034551,
+      "learning_rate": 4.187830935167281e-05,
+      "loss": 0.8924,
+      "step": 11090
+    },
+    {
+      "epoch": 0.5659511548462755,
+      "grad_norm": 3.708789845615069,
+      "learning_rate": 4.1796847834938694e-05,
+      "loss": 0.9433,
+      "step": 11100
+    },
+    {
+      "epoch": 0.5664610207515424,
+      "grad_norm": 3.342070976615128,
+      "learning_rate": 4.171540868875862e-05,
+      "loss": 0.8593,
+      "step": 11110
+    },
+    {
+      "epoch": 0.5669708866568093,
+      "grad_norm": 2.8712579947880887,
+      "learning_rate": 4.163399213522263e-05,
+      "loss": 0.932,
+      "step": 11120
+    },
+    {
+      "epoch": 0.5674807525620762,
+      "grad_norm": 4.6773806382841965,
+      "learning_rate": 4.155259839635925e-05,
+      "loss": 1.0471,
+      "step": 11130
+    },
+    {
+      "epoch": 0.5679906184673431,
+      "grad_norm": 3.078974731646656,
+      "learning_rate": 4.147122769413471e-05,
+      "loss": 0.9242,
+      "step": 11140
+    },
+    {
+      "epoch": 0.56850048437261,
+      "grad_norm": 6.231642085771952,
+      "learning_rate": 4.138988025045245e-05,
+      "loss": 0.9739,
+      "step": 11150
+    },
+    {
+      "epoch": 0.5690103502778769,
+      "grad_norm": 3.0839405522612067,
+      "learning_rate": 4.1308556287152523e-05,
+      "loss": 0.9526,
+      "step": 11160
+    },
+    {
+      "epoch": 0.5695202161831439,
+      "grad_norm": 3.7497699141702743,
+      "learning_rate": 4.1227256026010875e-05,
+      "loss": 0.9777,
+      "step": 11170
+    },
+    {
+      "epoch": 0.5700300820884108,
+      "grad_norm": 3.9862229817529444,
+      "learning_rate": 4.114597968873886e-05,
+      "loss": 0.9009,
+      "step": 11180
+    },
+    {
+      "epoch": 0.5705399479936777,
+      "grad_norm": 3.891982808605384,
+      "learning_rate": 4.106472749698256e-05,
+      "loss": 0.9026,
+      "step": 11190
+    },
+    {
+      "epoch": 0.5710498138989446,
+      "grad_norm": 3.3862676151861586,
+      "learning_rate": 4.098349967232225e-05,
+      "loss": 0.9402,
+      "step": 11200
+    },
+    {
+      "epoch": 0.5715596798042115,
+      "grad_norm": 3.306179449922897,
+      "learning_rate": 4.090229643627176e-05,
+      "loss": 0.9188,
+      "step": 11210
+    },
+    {
+      "epoch": 0.5720695457094784,
+      "grad_norm": 2.525184577226416,
+      "learning_rate": 4.082111801027778e-05,
+      "loss": 0.8973,
+      "step": 11220
+    },
+    {
+      "epoch": 0.5725794116147453,
+      "grad_norm": 2.653011738044738,
+      "learning_rate": 4.073996461571945e-05,
+      "loss": 0.9313,
+      "step": 11230
+    },
+    {
+      "epoch": 0.5730892775200123,
+      "grad_norm": 2.2431306141496155,
+      "learning_rate": 4.065883647390756e-05,
+      "loss": 0.8939,
+      "step": 11240
+    },
+    {
+      "epoch": 0.5735991434252792,
+      "grad_norm": 4.048488570656739,
+      "learning_rate": 4.057773380608411e-05,
+      "loss": 0.9402,
+      "step": 11250
+    },
+    {
+      "epoch": 0.5741090093305461,
+      "grad_norm": 4.173730329774634,
+      "learning_rate": 4.049665683342154e-05,
+      "loss": 0.9531,
+      "step": 11260
+    },
+    {
+      "epoch": 0.574618875235813,
+      "grad_norm": 2.5522114553016038,
+      "learning_rate": 4.0415605777022306e-05,
+      "loss": 0.9363,
+      "step": 11270
+    },
+    {
+      "epoch": 0.5751287411410799,
+      "grad_norm": 2.2293996195235084,
+      "learning_rate": 4.0334580857918165e-05,
+      "loss": 0.8969,
+      "step": 11280
+    },
+    {
+      "epoch": 0.5756386070463468,
+      "grad_norm": 4.178070132359859,
+      "learning_rate": 4.025358229706955e-05,
+      "loss": 0.9183,
+      "step": 11290
+    },
+    {
+      "epoch": 0.5761484729516138,
+      "grad_norm": 3.1278468887729454,
+      "learning_rate": 4.017261031536509e-05,
+      "loss": 0.86,
+      "step": 11300
+    },
+    {
+      "epoch": 0.5766583388568807,
+      "grad_norm": 3.7707189328781254,
+      "learning_rate": 4.0091665133620835e-05,
+      "loss": 0.9194,
+      "step": 11310
+    },
+    {
+      "epoch": 0.5771682047621476,
+      "grad_norm": 3.610050800355324,
+      "learning_rate": 4.001074697257986e-05,
+      "loss": 0.9452,
+      "step": 11320
+    },
+    {
+      "epoch": 0.5776780706674145,
+      "grad_norm": 2.6631250167415543,
+      "learning_rate": 3.992985605291143e-05,
+      "loss": 0.8058,
+      "step": 11330
+    },
+    {
+      "epoch": 0.5781879365726814,
+      "grad_norm": 3.16186610328534,
+      "learning_rate": 3.9848992595210624e-05,
+      "loss": 0.922,
+      "step": 11340
+    },
+    {
+      "epoch": 0.5786978024779483,
+      "grad_norm": 3.013990161069991,
+      "learning_rate": 3.976815681999762e-05,
+      "loss": 0.8733,
+      "step": 11350
+    },
+    {
+      "epoch": 0.5792076683832152,
+      "grad_norm": 3.0182525666443234,
+      "learning_rate": 3.968734894771701e-05,
+      "loss": 0.9217,
+      "step": 11360
+    },
+    {
+      "epoch": 0.5797175342884822,
+      "grad_norm": 3.680891120799879,
+      "learning_rate": 3.9606569198737406e-05,
+      "loss": 0.9184,
+      "step": 11370
+    },
+    {
+      "epoch": 0.5802274001937491,
+      "grad_norm": 3.9826156164741393,
+      "learning_rate": 3.952581779335066e-05,
+      "loss": 0.9767,
+      "step": 11380
+    },
+    {
+      "epoch": 0.580737266099016,
+      "grad_norm": 4.453625045000694,
+      "learning_rate": 3.9445094951771366e-05,
+      "loss": 0.9414,
+      "step": 11390
+    },
+    {
+      "epoch": 0.5812471320042829,
+      "grad_norm": 3.625764242947431,
+      "learning_rate": 3.936440089413619e-05,
+      "loss": 0.9542,
+      "step": 11400
+    },
+    {
+      "epoch": 0.5817569979095498,
+      "grad_norm": 2.6821388085917484,
+      "learning_rate": 3.928373584050331e-05,
+      "loss": 0.9525,
+      "step": 11410
+    },
+    {
+      "epoch": 0.5822668638148167,
+      "grad_norm": 5.108250606198265,
+      "learning_rate": 3.9203100010851845e-05,
+      "loss": 0.8815,
+      "step": 11420
+    },
+    {
+      "epoch": 0.5827767297200837,
+      "grad_norm": 2.3673563653919736,
+      "learning_rate": 3.912249362508115e-05,
+      "loss": 0.9805,
+      "step": 11430
+    },
+    {
+      "epoch": 0.5832865956253506,
+      "grad_norm": 3.253477854934498,
+      "learning_rate": 3.904191690301036e-05,
+      "loss": 0.873,
+      "step": 11440
+    },
+    {
+      "epoch": 0.5837964615306175,
+      "grad_norm": 3.2533349136746454,
+      "learning_rate": 3.8961370064377635e-05,
+      "loss": 0.9263,
+      "step": 11450
+    },
+    {
+      "epoch": 0.5843063274358844,
+      "grad_norm": 3.1266975497458867,
+      "learning_rate": 3.888085332883971e-05,
+      "loss": 0.9283,
+      "step": 11460
+    },
+    {
+      "epoch": 0.5848161933411513,
+      "grad_norm": 2.6265344858376527,
+      "learning_rate": 3.880036691597122e-05,
+      "loss": 0.9028,
+      "step": 11470
+    },
+    {
+      "epoch": 0.5853260592464182,
+      "grad_norm": 2.1547246508003832,
+      "learning_rate": 3.871991104526404e-05,
+      "loss": 0.9107,
+      "step": 11480
+    },
+    {
+      "epoch": 0.5858359251516851,
+      "grad_norm": 2.933037667488873,
+      "learning_rate": 3.863948593612685e-05,
+      "loss": 0.8794,
+      "step": 11490
+    },
+    {
+      "epoch": 0.5863457910569521,
+      "grad_norm": 2.8171298551717947,
+      "learning_rate": 3.855909180788435e-05,
+      "loss": 0.9344,
+      "step": 11500
+    },
+    {
+      "epoch": 0.586855656962219,
+      "grad_norm": 4.256605327236434,
+      "learning_rate": 3.847872887977682e-05,
+      "loss": 0.9252,
+      "step": 11510
+    },
+    {
+      "epoch": 0.5873655228674859,
+      "grad_norm": 4.920767140524973,
+      "learning_rate": 3.839839737095946e-05,
+      "loss": 0.9089,
+      "step": 11520
+    },
+    {
+      "epoch": 0.5878753887727528,
+      "grad_norm": 3.7727525937676907,
+      "learning_rate": 3.8318097500501706e-05,
+      "loss": 0.9241,
+      "step": 11530
+    },
+    {
+      "epoch": 0.5883852546780197,
+      "grad_norm": 2.57676433348543,
+      "learning_rate": 3.823782948738682e-05,
+      "loss": 0.9842,
+      "step": 11540
+    },
+    {
+      "epoch": 0.5888951205832866,
+      "grad_norm": 5.744358407957162,
+      "learning_rate": 3.8157593550511076e-05,
+      "loss": 0.9248,
+      "step": 11550
+    },
+    {
+      "epoch": 0.5894049864885536,
+      "grad_norm": 5.570619707991569,
+      "learning_rate": 3.807738990868339e-05,
+      "loss": 0.8789,
+      "step": 11560
+    },
+    {
+      "epoch": 0.5899148523938205,
+      "grad_norm": 2.21579379442429,
+      "learning_rate": 3.79972187806245e-05,
+      "loss": 0.8711,
+      "step": 11570
+    },
+    {
+      "epoch": 0.5904247182990874,
+      "grad_norm": 2.3273662229011083,
+      "learning_rate": 3.791708038496655e-05,
+      "loss": 0.9192,
+      "step": 11580
+    },
+    {
+      "epoch": 0.5909345842043543,
+      "grad_norm": 3.368524115436495,
+      "learning_rate": 3.783697494025241e-05,
+      "loss": 0.9443,
+      "step": 11590
+    },
+    {
+      "epoch": 0.5914444501096212,
+      "grad_norm": 2.9074077356463652,
+      "learning_rate": 3.7756902664935056e-05,
+      "loss": 0.8832,
+      "step": 11600
+    },
+    {
+      "epoch": 0.5919543160148881,
+      "grad_norm": 5.565835812156636,
+      "learning_rate": 3.7676863777377054e-05,
+      "loss": 1.005,
+      "step": 11610
+    },
+    {
+      "epoch": 0.592464181920155,
+      "grad_norm": 4.182857678708667,
+      "learning_rate": 3.759685849584986e-05,
+      "loss": 0.9664,
+      "step": 11620
+    },
+    {
+      "epoch": 0.592974047825422,
+      "grad_norm": 1.8679666859340913,
+      "learning_rate": 3.7516887038533336e-05,
+      "loss": 0.8767,
+      "step": 11630
+    },
+    {
+      "epoch": 0.5934839137306889,
+      "grad_norm": 2.849078248665122,
+      "learning_rate": 3.74369496235151e-05,
+      "loss": 0.8783,
+      "step": 11640
+    },
+    {
+      "epoch": 0.5939937796359558,
+      "grad_norm": 4.6788568126262975,
+      "learning_rate": 3.735704646878991e-05,
+      "loss": 0.93,
+      "step": 11650
+    },
+    {
+      "epoch": 0.5945036455412227,
+      "grad_norm": 3.019283126283293,
+      "learning_rate": 3.7277177792259114e-05,
+      "loss": 0.9045,
+      "step": 11660
+    },
+    {
+      "epoch": 0.5950135114464896,
+      "grad_norm": 2.972243476835822,
+      "learning_rate": 3.719734381173e-05,
+      "loss": 0.9524,
+      "step": 11670
+    },
+    {
+      "epoch": 0.5955233773517565,
+      "grad_norm": 2.2189264516677345,
+      "learning_rate": 3.71175447449153e-05,
+      "loss": 0.8851,
+      "step": 11680
+    },
+    {
+      "epoch": 0.5960332432570234,
+      "grad_norm": 3.2944785067298183,
+      "learning_rate": 3.7037780809432455e-05,
+      "loss": 0.8766,
+      "step": 11690
+    },
+    {
+      "epoch": 0.5965431091622904,
+      "grad_norm": 2.414383869960349,
+      "learning_rate": 3.6958052222803184e-05,
+      "loss": 0.8989,
+      "step": 11700
+    },
+    {
+      "epoch": 0.5970529750675573,
+      "grad_norm": 2.8865393460989024,
+      "learning_rate": 3.687835920245275e-05,
+      "loss": 0.9272,
+      "step": 11710
+    },
+    {
+      "epoch": 0.5975628409728242,
+      "grad_norm": 2.8104658906384246,
+      "learning_rate": 3.679870196570943e-05,
+      "loss": 0.8615,
+      "step": 11720
+    },
+    {
+      "epoch": 0.5980727068780911,
+      "grad_norm": 2.8246694591941535,
+      "learning_rate": 3.6719080729803964e-05,
+      "loss": 0.9094,
+      "step": 11730
+    },
+    {
+      "epoch": 0.598582572783358,
+      "grad_norm": 2.9577565894440254,
+      "learning_rate": 3.663949571186883e-05,
+      "loss": 0.9594,
+      "step": 11740
+    },
+    {
+      "epoch": 0.5990924386886249,
+      "grad_norm": 3.187791797691879,
+      "learning_rate": 3.655994712893781e-05,
+      "loss": 0.8837,
+      "step": 11750
+    },
+    {
+      "epoch": 0.5996023045938919,
+      "grad_norm": 3.373158951815689,
+      "learning_rate": 3.648043519794533e-05,
+      "loss": 0.9613,
+      "step": 11760
+    },
+    {
+      "epoch": 0.6001121704991588,
+      "grad_norm": 2.95618668476516,
+      "learning_rate": 3.640096013572578e-05,
+      "loss": 0.9403,
+      "step": 11770
+    },
+    {
+      "epoch": 0.6006220364044257,
+      "grad_norm": 3.5673480422749635,
+      "learning_rate": 3.6321522159013103e-05,
+      "loss": 0.904,
+      "step": 11780
+    },
+    {
+      "epoch": 0.6011319023096926,
+      "grad_norm": 3.118773856517395,
+      "learning_rate": 3.624212148444004e-05,
+      "loss": 0.8945,
+      "step": 11790
+    },
+    {
+      "epoch": 0.6016417682149595,
+      "grad_norm": 5.216405275539851,
+      "learning_rate": 3.616275832853765e-05,
+      "loss": 0.9204,
+      "step": 11800
+    },
+    {
+      "epoch": 0.6021516341202264,
+      "grad_norm": 2.7133261562797526,
+      "learning_rate": 3.608343290773464e-05,
+      "loss": 0.9189,
+      "step": 11810
+    },
+    {
+      "epoch": 0.6026615000254933,
+      "grad_norm": 3.1420562657787627,
+      "learning_rate": 3.600414543835684e-05,
+      "loss": 0.9192,
+      "step": 11820
+    },
+    {
+      "epoch": 0.6031713659307603,
+      "grad_norm": 2.0431784110850617,
+      "learning_rate": 3.59248961366266e-05,
+      "loss": 0.9465,
+      "step": 11830
+    },
+    {
+      "epoch": 0.6036812318360272,
+      "grad_norm": 3.1976159900884564,
+      "learning_rate": 3.5845685218662105e-05,
+      "loss": 0.9248,
+      "step": 11840
+    },
+    {
+      "epoch": 0.6041910977412941,
+      "grad_norm": 3.4779893339813315,
+      "learning_rate": 3.576651290047696e-05,
+      "loss": 0.8583,
+      "step": 11850
+    },
+    {
+      "epoch": 0.604700963646561,
+      "grad_norm": 4.066301648674187,
+      "learning_rate": 3.568737939797943e-05,
+      "loss": 0.8628,
+      "step": 11860
+    },
+    {
+      "epoch": 0.6052108295518279,
+      "grad_norm": 2.7683789182171723,
+      "learning_rate": 3.5608284926972e-05,
+      "loss": 0.907,
+      "step": 11870
+    },
+    {
+      "epoch": 0.6057206954570948,
+      "grad_norm": 2.7643500070217417,
+      "learning_rate": 3.5529229703150616e-05,
+      "loss": 0.9485,
+      "step": 11880
+    },
+    {
+      "epoch": 0.6062305613623618,
+      "grad_norm": 2.9399646436397027,
+      "learning_rate": 3.5450213942104295e-05,
+      "loss": 0.875,
+      "step": 11890
+    },
+    {
+      "epoch": 0.6067404272676287,
+      "grad_norm": 3.4505591667823285,
+      "learning_rate": 3.537123785931439e-05,
+      "loss": 0.9528,
+      "step": 11900
+    },
+    {
+      "epoch": 0.6072502931728956,
+      "grad_norm": 3.925384034593624,
+      "learning_rate": 3.5292301670154027e-05,
+      "loss": 0.9222,
+      "step": 11910
+    },
+    {
+      "epoch": 0.6077601590781625,
+      "grad_norm": 3.8888063709502894,
+      "learning_rate": 3.5213405589887606e-05,
+      "loss": 0.9338,
+      "step": 11920
+    },
+    {
+      "epoch": 0.6082700249834293,
+      "grad_norm": 2.485004855310051,
+      "learning_rate": 3.513454983367006e-05,
+      "loss": 0.9199,
+      "step": 11930
+    },
+    {
+      "epoch": 0.6087798908886962,
+      "grad_norm": 2.9263024358343084,
+      "learning_rate": 3.505573461654642e-05,
+      "loss": 0.8846,
+      "step": 11940
+    },
+    {
+      "epoch": 0.6092897567939631,
+      "grad_norm": 4.171333669515278,
+      "learning_rate": 3.497696015345116e-05,
+      "loss": 0.9739,
+      "step": 11950
+    },
+    {
+      "epoch": 0.60979962269923,
+      "grad_norm": 3.2731085524136994,
+      "learning_rate": 3.4898226659207574e-05,
+      "loss": 0.9456,
+      "step": 11960
+    },
+    {
+      "epoch": 0.610309488604497,
+      "grad_norm": 3.303438605734584,
+      "learning_rate": 3.4819534348527285e-05,
+      "loss": 0.9216,
+      "step": 11970
+    },
+    {
+      "epoch": 0.6108193545097639,
+      "grad_norm": 5.698560138213655,
+      "learning_rate": 3.4740883436009544e-05,
+      "loss": 1.0142,
+      "step": 11980
+    },
+    {
+      "epoch": 0.6113292204150308,
+      "grad_norm": 3.3245038246763055,
+      "learning_rate": 3.466227413614076e-05,
+      "loss": 0.9261,
+      "step": 11990
+    },
+    {
+      "epoch": 0.6118390863202977,
+      "grad_norm": 4.040430285803717,
+      "learning_rate": 3.458370666329384e-05,
+      "loss": 0.9069,
+      "step": 12000
+    },
+    {
+      "epoch": 0.6123489522255646,
+      "grad_norm": 3.4869766502419823,
+      "learning_rate": 3.450518123172762e-05,
+      "loss": 0.8981,
+      "step": 12010
+    },
+    {
+      "epoch": 0.6128588181308315,
+      "grad_norm": 4.232843588821771,
+      "learning_rate": 3.44266980555863e-05,
+      "loss": 0.8855,
+      "step": 12020
+    },
+    {
+      "epoch": 0.6133686840360985,
+      "grad_norm": 3.433938204132207,
+      "learning_rate": 3.434825734889884e-05,
+      "loss": 0.8779,
+      "step": 12030
+    },
+    {
+      "epoch": 0.6138785499413654,
+      "grad_norm": 2.8696393308402217,
+      "learning_rate": 3.426985932557841e-05,
+      "loss": 0.938,
+      "step": 12040
+    },
+    {
+      "epoch": 0.6143884158466323,
+      "grad_norm": 2.968846488644237,
+      "learning_rate": 3.41915041994217e-05,
+      "loss": 0.8572,
+      "step": 12050
+    },
+    {
+      "epoch": 0.6148982817518992,
+      "grad_norm": 5.0791179817812795,
+      "learning_rate": 3.411319218410849e-05,
+      "loss": 0.9212,
+      "step": 12060
+    },
+    {
+      "epoch": 0.6154081476571661,
+      "grad_norm": 2.2962426975985823,
+      "learning_rate": 3.403492349320101e-05,
+      "loss": 0.9056,
+      "step": 12070
+    },
+    {
+      "epoch": 0.615918013562433,
+      "grad_norm": 1.8383516635333248,
+      "learning_rate": 3.395669834014326e-05,
+      "loss": 0.9271,
+      "step": 12080
+    },
+    {
+      "epoch": 0.6164278794677,
+      "grad_norm": 3.1729183959789315,
+      "learning_rate": 3.387851693826057e-05,
+      "loss": 0.923,
+      "step": 12090
+    },
+    {
+      "epoch": 0.6169377453729669,
+      "grad_norm": 3.95371535396455,
+      "learning_rate": 3.380037950075894e-05,
+      "loss": 0.9044,
+      "step": 12100
+    },
+    {
+      "epoch": 0.6174476112782338,
+      "grad_norm": 3.5827463721945905,
+      "learning_rate": 3.3722286240724476e-05,
+      "loss": 0.9299,
+      "step": 12110
+    },
+    {
+      "epoch": 0.6179574771835007,
+      "grad_norm": 3.129245853042041,
+      "learning_rate": 3.3644237371122804e-05,
+      "loss": 0.9198,
+      "step": 12120
+    },
+    {
+      "epoch": 0.6184673430887676,
+      "grad_norm": 3.5168693863681932,
+      "learning_rate": 3.3566233104798515e-05,
+      "loss": 0.911,
+      "step": 12130
+    },
+    {
+      "epoch": 0.6189772089940345,
+      "grad_norm": 2.1807090890629452,
+      "learning_rate": 3.3488273654474564e-05,
+      "loss": 0.9046,
+      "step": 12140
+    },
+    {
+      "epoch": 0.6194870748993014,
+      "grad_norm": 2.5882476557124443,
+      "learning_rate": 3.341035923275164e-05,
+      "loss": 0.8614,
+      "step": 12150
+    },
+    {
+      "epoch": 0.6199969408045684,
+      "grad_norm": 3.1771327820708453,
+      "learning_rate": 3.3332490052107704e-05,
+      "loss": 0.9449,
+      "step": 12160
+    },
+    {
+      "epoch": 0.6205068067098353,
+      "grad_norm": 3.6009214787575883,
+      "learning_rate": 3.3254666324897274e-05,
+      "loss": 0.8916,
+      "step": 12170
+    },
+    {
+      "epoch": 0.6210166726151022,
+      "grad_norm": 5.936868523309456,
+      "learning_rate": 3.317688826335097e-05,
+      "loss": 0.8937,
+      "step": 12180
+    },
+    {
+      "epoch": 0.6215265385203691,
+      "grad_norm": 3.5048551406230986,
+      "learning_rate": 3.309915607957487e-05,
+      "loss": 0.9668,
+      "step": 12190
+    },
+    {
+      "epoch": 0.622036404425636,
+      "grad_norm": 3.310640103119443,
+      "learning_rate": 3.30214699855499e-05,
+      "loss": 0.9018,
+      "step": 12200
+    },
+    {
+      "epoch": 0.6225462703309029,
+      "grad_norm": 2.9346870989631726,
+      "learning_rate": 3.294383019313134e-05,
+      "loss": 0.8914,
+      "step": 12210
+    },
+    {
+      "epoch": 0.6230561362361698,
+      "grad_norm": 2.558908426943729,
+      "learning_rate": 3.286623691404819e-05,
+      "loss": 0.9335,
+      "step": 12220
+    },
+    {
+      "epoch": 0.6235660021414368,
+      "grad_norm": 4.90660193471452,
+      "learning_rate": 3.2788690359902584e-05,
+      "loss": 0.8607,
+      "step": 12230
+    },
+    {
+      "epoch": 0.6240758680467037,
+      "grad_norm": 2.7439371046004783,
+      "learning_rate": 3.271119074216925e-05,
+      "loss": 0.891,
+      "step": 12240
+    },
+    {
+      "epoch": 0.6245857339519706,
+      "grad_norm": 3.291852770741101,
+      "learning_rate": 3.263373827219493e-05,
+      "loss": 0.9197,
+      "step": 12250
+    },
+    {
+      "epoch": 0.6250955998572375,
+      "grad_norm": 2.634482936247294,
+      "learning_rate": 3.255633316119779e-05,
+      "loss": 0.861,
+      "step": 12260
+    },
+    {
+      "epoch": 0.6256054657625044,
+      "grad_norm": 2.409125778859325,
+      "learning_rate": 3.247897562026681e-05,
+      "loss": 0.9235,
+      "step": 12270
+    },
+    {
+      "epoch": 0.6261153316677713,
+      "grad_norm": 3.1405082718239363,
+      "learning_rate": 3.2401665860361285e-05,
+      "loss": 0.8404,
+      "step": 12280
+    },
+    {
+      "epoch": 0.6266251975730383,
+      "grad_norm": 2.6081088882613894,
+      "learning_rate": 3.232440409231017e-05,
+      "loss": 0.9134,
+      "step": 12290
+    },
+    {
+      "epoch": 0.6271350634783052,
+      "grad_norm": 2.2025857873300305,
+      "learning_rate": 3.224719052681156e-05,
+      "loss": 0.919,
+      "step": 12300
+    },
+    {
+      "epoch": 0.6276449293835721,
+      "grad_norm": 3.4369456543891554,
+      "learning_rate": 3.2170025374432155e-05,
+      "loss": 0.8261,
+      "step": 12310
+    },
+    {
+      "epoch": 0.628154795288839,
+      "grad_norm": 3.8057646280451554,
+      "learning_rate": 3.2092908845606504e-05,
+      "loss": 0.966,
+      "step": 12320
+    },
+    {
+      "epoch": 0.6286646611941059,
+      "grad_norm": 2.911014268964996,
+      "learning_rate": 3.201584115063667e-05,
+      "loss": 0.9553,
+      "step": 12330
+    },
+    {
+      "epoch": 0.6291745270993728,
+      "grad_norm": 6.233174755628133,
+      "learning_rate": 3.1938822499691475e-05,
+      "loss": 0.8982,
+      "step": 12340
+    },
+    {
+      "epoch": 0.6296843930046397,
+      "grad_norm": 2.7744298669701752,
+      "learning_rate": 3.186185310280605e-05,
+      "loss": 0.8442,
+      "step": 12350
+    },
+    {
+      "epoch": 0.6301942589099067,
+      "grad_norm": 4.394364057091598,
+      "learning_rate": 3.178493316988115e-05,
+      "loss": 0.9123,
+      "step": 12360
+    },
+    {
+      "epoch": 0.6307041248151736,
+      "grad_norm": 2.505549874929729,
+      "learning_rate": 3.170806291068266e-05,
+      "loss": 0.8654,
+      "step": 12370
+    },
+    {
+      "epoch": 0.6312139907204405,
+      "grad_norm": 3.1903516974055703,
+      "learning_rate": 3.163124253484105e-05,
+      "loss": 0.9217,
+      "step": 12380
+    },
+    {
+      "epoch": 0.6317238566257074,
+      "grad_norm": 4.443949170985445,
+      "learning_rate": 3.1554472251850665e-05,
+      "loss": 0.8163,
+      "step": 12390
+    },
+    {
+      "epoch": 0.6322337225309743,
+      "grad_norm": 3.867217892734069,
+      "learning_rate": 3.1477752271069314e-05,
+      "loss": 0.9288,
+      "step": 12400
+    },
+    {
+      "epoch": 0.6327435884362412,
+      "grad_norm": 3.6752994660338603,
+      "learning_rate": 3.140108280171758e-05,
+      "loss": 0.9328,
+      "step": 12410
+    },
+    {
+      "epoch": 0.6332534543415081,
+      "grad_norm": 3.8749228417087136,
+      "learning_rate": 3.1324464052878345e-05,
+      "loss": 0.8701,
+      "step": 12420
+    },
+    {
+      "epoch": 0.6337633202467751,
+      "grad_norm": 3.0585136968349627,
+      "learning_rate": 3.124789623349615e-05,
+      "loss": 0.8484,
+      "step": 12430
+    },
+    {
+      "epoch": 0.634273186152042,
+      "grad_norm": 4.2716899334687,
+      "learning_rate": 3.117137955237664e-05,
+      "loss": 0.9894,
+      "step": 12440
+    },
+    {
+      "epoch": 0.6347830520573089,
+      "grad_norm": 2.6764020080604483,
+      "learning_rate": 3.109491421818601e-05,
+      "loss": 0.8136,
+      "step": 12450
+    },
+    {
+      "epoch": 0.6352929179625758,
+      "grad_norm": 3.1016263312935433,
+      "learning_rate": 3.101850043945042e-05,
+      "loss": 0.8488,
+      "step": 12460
+    },
+    {
+      "epoch": 0.6358027838678427,
+      "grad_norm": 2.4679718336374488,
+      "learning_rate": 3.094213842455546e-05,
+      "loss": 0.8484,
+      "step": 12470
+    },
+    {
+      "epoch": 0.6363126497731096,
+      "grad_norm": 5.22514561892103,
+      "learning_rate": 3.086582838174551e-05,
+      "loss": 0.8965,
+      "step": 12480
+    },
+    {
+      "epoch": 0.6368225156783766,
+      "grad_norm": 3.4772275931848475,
+      "learning_rate": 3.0789570519123277e-05,
+      "loss": 0.8976,
+      "step": 12490
+    },
+    {
+      "epoch": 0.6373323815836435,
+      "grad_norm": 4.794405574374894,
+      "learning_rate": 3.0713365044649154e-05,
+      "loss": 0.8751,
+      "step": 12500
+    },
+    {
+      "epoch": 0.6378422474889104,
+      "grad_norm": 4.148149373525859,
+      "learning_rate": 3.063721216614061e-05,
+      "loss": 0.9763,
+      "step": 12510
+    },
+    {
+      "epoch": 0.6383521133941773,
+      "grad_norm": 4.087205183221575,
+      "learning_rate": 3.056111209127176e-05,
+      "loss": 0.9247,
+      "step": 12520
+    },
+    {
+      "epoch": 0.6388619792994442,
+      "grad_norm": 2.360906763307815,
+      "learning_rate": 3.048506502757267e-05,
+      "loss": 0.8688,
+      "step": 12530
+    },
+    {
+      "epoch": 0.6393718452047111,
+      "grad_norm": 4.244869787065028,
+      "learning_rate": 3.0409071182428854e-05,
+      "loss": 0.9515,
+      "step": 12540
+    },
+    {
+      "epoch": 0.639881711109978,
+      "grad_norm": 3.4986343512590246,
+      "learning_rate": 3.0333130763080698e-05,
+      "loss": 0.9537,
+      "step": 12550
+    },
+    {
+      "epoch": 0.640391577015245,
+      "grad_norm": 3.519237012987473,
+      "learning_rate": 3.0257243976622896e-05,
+      "loss": 0.8557,
+      "step": 12560
+    },
+    {
+      "epoch": 0.6409014429205119,
+      "grad_norm": 2.7358645537943693,
+      "learning_rate": 3.0181411030003898e-05,
+      "loss": 0.9759,
+      "step": 12570
+    },
+    {
+      "epoch": 0.6414113088257788,
+      "grad_norm": 4.385992405742961,
+      "learning_rate": 3.010563213002527e-05,
+      "loss": 0.893,
+      "step": 12580
+    },
+    {
+      "epoch": 0.6419211747310457,
+      "grad_norm": 3.3110790101503955,
+      "learning_rate": 3.0029907483341256e-05,
+      "loss": 0.9642,
+      "step": 12590
+    },
+    {
+      "epoch": 0.6424310406363126,
+      "grad_norm": 3.0589397674175807,
+      "learning_rate": 2.99542372964581e-05,
+      "loss": 0.8845,
+      "step": 12600
+    },
+    {
+      "epoch": 0.6429409065415795,
+      "grad_norm": 2.822803276107203,
+      "learning_rate": 2.9878621775733552e-05,
+      "loss": 0.9041,
+      "step": 12610
+    },
+    {
+      "epoch": 0.6434507724468465,
+      "grad_norm": 2.4733742296815584,
+      "learning_rate": 2.980306112737631e-05,
+      "loss": 0.8669,
+      "step": 12620
+    },
+    {
+      "epoch": 0.6439606383521134,
+      "grad_norm": 7.301140467631872,
+      "learning_rate": 2.9727555557445363e-05,
+      "loss": 0.9312,
+      "step": 12630
+    },
+    {
+      "epoch": 0.6444705042573803,
+      "grad_norm": 2.392548659682165,
+      "learning_rate": 2.9652105271849563e-05,
+      "loss": 0.854,
+      "step": 12640
+    },
+    {
+      "epoch": 0.6449803701626472,
+      "grad_norm": 2.7533656924129315,
+      "learning_rate": 2.9576710476346947e-05,
+      "loss": 0.8517,
+      "step": 12650
+    },
+    {
+      "epoch": 0.6454902360679141,
+      "grad_norm": 4.327769590092878,
+      "learning_rate": 2.9501371376544284e-05,
+      "loss": 0.8938,
+      "step": 12660
+    },
+    {
+      "epoch": 0.646000101973181,
+      "grad_norm": 3.476758331876157,
+      "learning_rate": 2.9426088177896394e-05,
+      "loss": 0.8683,
+      "step": 12670
+    },
+    {
+      "epoch": 0.6465099678784479,
+      "grad_norm": 2.308821090636214,
+      "learning_rate": 2.9350861085705694e-05,
+      "loss": 0.9404,
+      "step": 12680
+    },
+    {
+      "epoch": 0.6470198337837149,
+      "grad_norm": 4.226892987412482,
+      "learning_rate": 2.9275690305121607e-05,
+      "loss": 0.8856,
+      "step": 12690
+    },
+    {
+      "epoch": 0.6475296996889818,
+      "grad_norm": 3.7325296122621903,
+      "learning_rate": 2.9200576041139917e-05,
+      "loss": 0.9357,
+      "step": 12700
+    },
+    {
+      "epoch": 0.6480395655942487,
+      "grad_norm": 5.412846699260613,
+      "learning_rate": 2.9125518498602367e-05,
+      "loss": 0.9863,
+      "step": 12710
+    },
+    {
+      "epoch": 0.6485494314995156,
+      "grad_norm": 3.713960826368952,
+      "learning_rate": 2.905051788219597e-05,
+      "loss": 0.9176,
+      "step": 12720
+    },
+    {
+      "epoch": 0.6490592974047825,
+      "grad_norm": 3.1083005638029273,
+      "learning_rate": 2.8975574396452487e-05,
+      "loss": 0.9053,
+      "step": 12730
+    },
+    {
+      "epoch": 0.6495691633100494,
+      "grad_norm": 2.4663160862442437,
+      "learning_rate": 2.890068824574797e-05,
+      "loss": 0.8443,
+      "step": 12740
+    },
+    {
+      "epoch": 0.6500790292153164,
+      "grad_norm": 3.1856470594057327,
+      "learning_rate": 2.882585963430196e-05,
+      "loss": 0.9438,
+      "step": 12750
+    },
+    {
+      "epoch": 0.6505888951205833,
+      "grad_norm": 3.733059533076023,
+      "learning_rate": 2.875108876617726e-05,
+      "loss": 0.8556,
+      "step": 12760
+    },
+    {
+      "epoch": 0.6510987610258502,
+      "grad_norm": 2.795370045295235,
+      "learning_rate": 2.8676375845279013e-05,
+      "loss": 0.9295,
+      "step": 12770
+    },
+    {
+      "epoch": 0.6516086269311171,
+      "grad_norm": 4.114564052755407,
+      "learning_rate": 2.860172107535452e-05,
+      "loss": 0.8926,
+      "step": 12780
+    },
+    {
+      "epoch": 0.652118492836384,
+      "grad_norm": 3.8342475999978105,
+      "learning_rate": 2.8527124659992388e-05,
+      "loss": 0.9178,
+      "step": 12790
+    },
+    {
+      "epoch": 0.6526283587416509,
+      "grad_norm": 3.804108407067104,
+      "learning_rate": 2.8452586802622116e-05,
+      "loss": 0.9454,
+      "step": 12800
+    },
+    {
+      "epoch": 0.6531382246469178,
+      "grad_norm": 3.495862283582211,
+      "learning_rate": 2.8378107706513513e-05,
+      "loss": 0.8843,
+      "step": 12810
+    },
+    {
+      "epoch": 0.6536480905521848,
+      "grad_norm": 3.1352881634440863,
+      "learning_rate": 2.8303687574776123e-05,
+      "loss": 0.9216,
+      "step": 12820
+    },
+    {
+      "epoch": 0.6541579564574517,
+      "grad_norm": 2.897073613404968,
+      "learning_rate": 2.8229326610358764e-05,
+      "loss": 0.8859,
+      "step": 12830
+    },
+    {
+      "epoch": 0.6546678223627186,
+      "grad_norm": 2.2408615925152193,
+      "learning_rate": 2.8155025016048763e-05,
+      "loss": 0.8343,
+      "step": 12840
+    },
+    {
+      "epoch": 0.6551776882679855,
+      "grad_norm": 3.0453181085219496,
+      "learning_rate": 2.8080782994471688e-05,
+      "loss": 0.91,
+      "step": 12850
+    },
+    {
+      "epoch": 0.6556875541732524,
+      "grad_norm": 2.208086711932004,
+      "learning_rate": 2.800660074809057e-05,
+      "loss": 0.8767,
+      "step": 12860
+    },
+    {
+      "epoch": 0.6561974200785193,
+      "grad_norm": 2.5486889873141783,
+      "learning_rate": 2.793247847920542e-05,
+      "loss": 0.8911,
+      "step": 12870
+    },
+    {
+      "epoch": 0.6567072859837862,
+      "grad_norm": 3.106151628703104,
+      "learning_rate": 2.7858416389952723e-05,
+      "loss": 0.9332,
+      "step": 12880
+    },
+    {
+      "epoch": 0.6572171518890532,
+      "grad_norm": 3.2586925146836605,
+      "learning_rate": 2.7784414682304832e-05,
+      "loss": 0.9068,
+      "step": 12890
+    },
+    {
+      "epoch": 0.6577270177943201,
+      "grad_norm": 3.6442338592489962,
+      "learning_rate": 2.7710473558069437e-05,
+      "loss": 0.8552,
+      "step": 12900
+    },
+    {
+      "epoch": 0.658236883699587,
+      "grad_norm": 3.228672847008334,
+      "learning_rate": 2.7636593218889e-05,
+      "loss": 0.9081,
+      "step": 12910
+    },
+    {
+      "epoch": 0.6587467496048539,
+      "grad_norm": 2.6403972328638683,
+      "learning_rate": 2.7562773866240266e-05,
+      "loss": 0.841,
+      "step": 12920
+    },
+    {
+      "epoch": 0.6592566155101208,
+      "grad_norm": 3.8596616671540764,
+      "learning_rate": 2.7489015701433606e-05,
+      "loss": 0.8803,
+      "step": 12930
+    },
+    {
+      "epoch": 0.6597664814153877,
+      "grad_norm": 2.0114828091931423,
+      "learning_rate": 2.7415318925612564e-05,
+      "loss": 0.8742,
+      "step": 12940
+    },
+    {
+      "epoch": 0.6602763473206547,
+      "grad_norm": 2.6827974581752234,
+      "learning_rate": 2.734168373975326e-05,
+      "loss": 0.9224,
+      "step": 12950
+    },
+    {
+      "epoch": 0.6607862132259216,
+      "grad_norm": 3.196678441619736,
+      "learning_rate": 2.726811034466385e-05,
+      "loss": 0.8562,
+      "step": 12960
+    },
+    {
+      "epoch": 0.6612960791311885,
+      "grad_norm": 3.9594953344440054,
+      "learning_rate": 2.719459894098397e-05,
+      "loss": 0.9606,
+      "step": 12970
+    },
+    {
+      "epoch": 0.6618059450364554,
+      "grad_norm": 3.237167790637889,
+      "learning_rate": 2.7121149729184282e-05,
+      "loss": 0.9294,
+      "step": 12980
+    },
+    {
+      "epoch": 0.6623158109417223,
+      "grad_norm": 4.424279255609567,
+      "learning_rate": 2.7047762909565698e-05,
+      "loss": 0.9269,
+      "step": 12990
+    },
+    {
+      "epoch": 0.6628256768469892,
+      "grad_norm": 4.143581847554707,
+      "learning_rate": 2.6974438682259128e-05,
+      "loss": 0.8609,
+      "step": 13000
+    },
+    {
+      "epoch": 0.6633355427522561,
+      "grad_norm": 5.414971692654772,
+      "learning_rate": 2.6901177247224705e-05,
+      "loss": 0.9294,
+      "step": 13010
+    },
+    {
+      "epoch": 0.6638454086575231,
+      "grad_norm": 2.81261394366698,
+      "learning_rate": 2.6827978804251364e-05,
+      "loss": 0.9249,
+      "step": 13020
+    },
+    {
+      "epoch": 0.66435527456279,
+      "grad_norm": 2.4113320950008914,
+      "learning_rate": 2.675484355295622e-05,
+      "loss": 0.8785,
+      "step": 13030
+    },
+    {
+      "epoch": 0.6648651404680569,
+      "grad_norm": 3.800653939142313,
+      "learning_rate": 2.6681771692784063e-05,
+      "loss": 0.9998,
+      "step": 13040
+    },
+    {
+      "epoch": 0.6653750063733238,
+      "grad_norm": 3.0109208441371655,
+      "learning_rate": 2.660876342300689e-05,
+      "loss": 0.8705,
+      "step": 13050
+    },
+    {
+      "epoch": 0.6658848722785907,
+      "grad_norm": 3.1447709022295225,
+      "learning_rate": 2.6535818942723144e-05,
+      "loss": 0.9503,
+      "step": 13060
+    },
+    {
+      "epoch": 0.6663947381838576,
+      "grad_norm": 3.9762335762036636,
+      "learning_rate": 2.6462938450857448e-05,
+      "loss": 0.9115,
+      "step": 13070
+    },
+    {
+      "epoch": 0.6669046040891246,
+      "grad_norm": 3.2259963956535977,
+      "learning_rate": 2.63901221461598e-05,
+      "loss": 0.9623,
+      "step": 13080
+    },
+    {
+      "epoch": 0.6674144699943915,
+      "grad_norm": 4.073023361887602,
+      "learning_rate": 2.6317370227205273e-05,
+      "loss": 0.86,
+      "step": 13090
+    },
+    {
+      "epoch": 0.6679243358996584,
+      "grad_norm": 3.4641317188533507,
+      "learning_rate": 2.624468289239327e-05,
+      "loss": 0.9318,
+      "step": 13100
+    },
+    {
+      "epoch": 0.6684342018049253,
+      "grad_norm": 4.418183729852952,
+      "learning_rate": 2.6172060339947102e-05,
+      "loss": 0.8924,
+      "step": 13110
+    },
+    {
+      "epoch": 0.6689440677101922,
+      "grad_norm": 6.2991446822572215,
+      "learning_rate": 2.6099502767913414e-05,
+      "loss": 0.9038,
+      "step": 13120
+    },
+    {
+      "epoch": 0.6694539336154591,
+      "grad_norm": 3.1092027982406583,
+      "learning_rate": 2.602701037416162e-05,
+      "loss": 0.8888,
+      "step": 13130
+    },
+    {
+      "epoch": 0.669963799520726,
+      "grad_norm": 2.368881359518373,
+      "learning_rate": 2.5954583356383465e-05,
+      "loss": 0.874,
+      "step": 13140
+    },
+    {
+      "epoch": 0.670473665425993,
+      "grad_norm": 2.683827925138308,
+      "learning_rate": 2.588222191209228e-05,
+      "loss": 0.8619,
+      "step": 13150
+    },
+    {
+      "epoch": 0.6709835313312599,
+      "grad_norm": 3.8128522107134137,
+      "learning_rate": 2.58099262386227e-05,
+      "loss": 0.9038,
+      "step": 13160
+    },
+    {
+      "epoch": 0.6714933972365268,
+      "grad_norm": 3.239193069409403,
+      "learning_rate": 2.5737696533129928e-05,
+      "loss": 0.8674,
+      "step": 13170
+    },
+    {
+      "epoch": 0.6720032631417937,
+      "grad_norm": 3.0916442749025648,
+      "learning_rate": 2.5665532992589292e-05,
+      "loss": 0.881,
+      "step": 13180
+    },
+    {
+      "epoch": 0.6725131290470606,
+      "grad_norm": 4.545306586179735,
+      "learning_rate": 2.5593435813795663e-05,
+      "loss": 0.858,
+      "step": 13190
+    },
+    {
+      "epoch": 0.6730229949523275,
+      "grad_norm": 4.702378655621528,
+      "learning_rate": 2.5521405193362958e-05,
+      "loss": 0.9547,
+      "step": 13200
+    },
+    {
+      "epoch": 0.6735328608575945,
+      "grad_norm": 2.1072438976013212,
+      "learning_rate": 2.5449441327723582e-05,
+      "loss": 0.8596,
+      "step": 13210
+    },
+    {
+      "epoch": 0.6740427267628614,
+      "grad_norm": 4.903565065684834,
+      "learning_rate": 2.5377544413127875e-05,
+      "loss": 0.9372,
+      "step": 13220
+    },
+    {
+      "epoch": 0.6745525926681283,
+      "grad_norm": 3.3927449973757784,
+      "learning_rate": 2.530571464564364e-05,
+      "loss": 0.8379,
+      "step": 13230
+    },
+    {
+      "epoch": 0.6750624585733952,
+      "grad_norm": 2.5659211244065414,
+      "learning_rate": 2.5233952221155533e-05,
+      "loss": 0.921,
+      "step": 13240
+    },
+    {
+      "epoch": 0.6755723244786621,
+      "grad_norm": 4.793827349345645,
+      "learning_rate": 2.516225733536457e-05,
+      "loss": 0.9047,
+      "step": 13250
+    },
+    {
+      "epoch": 0.676082190383929,
+      "grad_norm": 2.537834221579057,
+      "learning_rate": 2.5090630183787568e-05,
+      "loss": 0.8801,
+      "step": 13260
+    },
+    {
+      "epoch": 0.6765920562891959,
+      "grad_norm": 3.3959247877590575,
+      "learning_rate": 2.5019070961756645e-05,
+      "loss": 0.8718,
+      "step": 13270
+    },
+    {
+      "epoch": 0.6771019221944629,
+      "grad_norm": 3.7648064061843933,
+      "learning_rate": 2.4947579864418662e-05,
+      "loss": 0.885,
+      "step": 13280
+    },
+    {
+      "epoch": 0.6776117880997298,
+      "grad_norm": 5.504751414259825,
+      "learning_rate": 2.4876157086734754e-05,
+      "loss": 0.9406,
+      "step": 13290
+    },
+    {
+      "epoch": 0.6781216540049967,
+      "grad_norm": 2.838604719878593,
+      "learning_rate": 2.4804802823479613e-05,
+      "loss": 0.8941,
+      "step": 13300
+    },
+    {
+      "epoch": 0.6786315199102636,
+      "grad_norm": 3.2520766775921612,
+      "learning_rate": 2.473351726924123e-05,
+      "loss": 0.9337,
+      "step": 13310
+    },
+    {
+      "epoch": 0.6791413858155305,
+      "grad_norm": 2.745153748830726,
+      "learning_rate": 2.4662300618420147e-05,
+      "loss": 0.9217,
+      "step": 13320
+    },
+    {
+      "epoch": 0.6796512517207974,
+      "grad_norm": 3.9717715373219784,
+      "learning_rate": 2.4591153065229012e-05,
+      "loss": 0.8655,
+      "step": 13330
+    },
+    {
+      "epoch": 0.6801611176260643,
+      "grad_norm": 2.5987326224708784,
+      "learning_rate": 2.4520074803692038e-05,
+      "loss": 0.8738,
+      "step": 13340
+    },
+    {
+      "epoch": 0.6806709835313313,
+      "grad_norm": 3.2422869386474638,
+      "learning_rate": 2.4449066027644475e-05,
+      "loss": 0.9122,
+      "step": 13350
+    },
+    {
+      "epoch": 0.6811808494365982,
+      "grad_norm": 3.2192037375557496,
+      "learning_rate": 2.4378126930732126e-05,
+      "loss": 1.0063,
+      "step": 13360
+    },
+    {
+      "epoch": 0.6816907153418651,
+      "grad_norm": 2.9091472226416477,
+      "learning_rate": 2.4307257706410668e-05,
+      "loss": 0.8559,
+      "step": 13370
+    },
+    {
+      "epoch": 0.682200581247132,
+      "grad_norm": 2.732330251551064,
+      "learning_rate": 2.423645854794537e-05,
+      "loss": 0.9152,
+      "step": 13380
+    },
+    {
+      "epoch": 0.6827104471523989,
+      "grad_norm": 3.9747726088439856,
+      "learning_rate": 2.4165729648410274e-05,
+      "loss": 0.8655,
+      "step": 13390
+    },
+    {
+      "epoch": 0.6832203130576658,
+      "grad_norm": 3.9107079639719657,
+      "learning_rate": 2.409507120068797e-05,
+      "loss": 0.8769,
+      "step": 13400
+    },
+    {
+      "epoch": 0.6837301789629328,
+      "grad_norm": 2.102310125706703,
+      "learning_rate": 2.4024483397468818e-05,
+      "loss": 0.9235,
+      "step": 13410
+    },
+    {
+      "epoch": 0.6842400448681997,
+      "grad_norm": 4.644569446012239,
+      "learning_rate": 2.3953966431250567e-05,
+      "loss": 0.8647,
+      "step": 13420
+    },
+    {
+      "epoch": 0.6847499107734666,
+      "grad_norm": 2.7607103784968134,
+      "learning_rate": 2.3883520494337787e-05,
+      "loss": 0.8993,
+      "step": 13430
+    },
+    {
+      "epoch": 0.6852597766787335,
+      "grad_norm": 2.0129941683167143,
+      "learning_rate": 2.381314577884131e-05,
+      "loss": 0.8345,
+      "step": 13440
+    },
+    {
+      "epoch": 0.6857696425840004,
+      "grad_norm": 3.6299635841779643,
+      "learning_rate": 2.3742842476677834e-05,
+      "loss": 0.9076,
+      "step": 13450
+    },
+    {
+      "epoch": 0.6862795084892673,
+      "grad_norm": 3.340608366170006,
+      "learning_rate": 2.3672610779569177e-05,
+      "loss": 0.9382,
+      "step": 13460
+    },
+    {
+      "epoch": 0.6867893743945342,
+      "grad_norm": 2.6640065743686927,
+      "learning_rate": 2.3602450879042003e-05,
+      "loss": 0.8252,
+      "step": 13470
+    },
+    {
+      "epoch": 0.6872992402998012,
+      "grad_norm": 3.033855625931912,
+      "learning_rate": 2.3532362966427118e-05,
+      "loss": 0.8373,
+      "step": 13480
+    },
+    {
+      "epoch": 0.6878091062050681,
+      "grad_norm": 2.2459827213653343,
+      "learning_rate": 2.3462347232859026e-05,
+      "loss": 0.8251,
+      "step": 13490
+    },
+    {
+      "epoch": 0.688318972110335,
+      "grad_norm": 3.6492834701278656,
+      "learning_rate": 2.3392403869275404e-05,
+      "loss": 0.8608,
+      "step": 13500
+    },
+    {
+      "epoch": 0.6888288380156019,
+      "grad_norm": 3.6390410751753874,
+      "learning_rate": 2.3322533066416547e-05,
+      "loss": 0.8952,
+      "step": 13510
+    },
+    {
+      "epoch": 0.6893387039208688,
+      "grad_norm": 2.3085856898721326,
+      "learning_rate": 2.3252735014824883e-05,
+      "loss": 0.8678,
+      "step": 13520
+    },
+    {
+      "epoch": 0.6898485698261357,
+      "grad_norm": 4.082707964001825,
+      "learning_rate": 2.3183009904844484e-05,
+      "loss": 0.8718,
+      "step": 13530
+    },
+    {
+      "epoch": 0.6903584357314027,
+      "grad_norm": 3.821054464677834,
+      "learning_rate": 2.3113357926620445e-05,
+      "loss": 0.9594,
+      "step": 13540
+    },
+    {
+      "epoch": 0.6908683016366696,
+      "grad_norm": 3.001023761538541,
+      "learning_rate": 2.304377927009844e-05,
+      "loss": 0.8978,
+      "step": 13550
+    },
+    {
+      "epoch": 0.6913781675419365,
+      "grad_norm": 2.8298450839958313,
+      "learning_rate": 2.297427412502422e-05,
+      "loss": 0.869,
+      "step": 13560
+    },
+    {
+      "epoch": 0.6918880334472034,
+      "grad_norm": 2.836530215019172,
+      "learning_rate": 2.2904842680943027e-05,
+      "loss": 0.9088,
+      "step": 13570
+    },
+    {
+      "epoch": 0.6923978993524703,
+      "grad_norm": 4.180377223929396,
+      "learning_rate": 2.2835485127199135e-05,
+      "loss": 0.9533,
+      "step": 13580
+    },
+    {
+      "epoch": 0.6929077652577372,
+      "grad_norm": 4.670221434880893,
+      "learning_rate": 2.2766201652935305e-05,
+      "loss": 0.9291,
+      "step": 13590
+    },
+    {
+      "epoch": 0.6934176311630041,
+      "grad_norm": 3.516604042302902,
+      "learning_rate": 2.2696992447092326e-05,
+      "loss": 0.8877,
+      "step": 13600
+    },
+    {
+      "epoch": 0.6939274970682711,
+      "grad_norm": 2.4900284887748985,
+      "learning_rate": 2.2627857698408338e-05,
+      "loss": 0.8219,
+      "step": 13610
+    },
+    {
+      "epoch": 0.694437362973538,
+      "grad_norm": 4.913835928490447,
+      "learning_rate": 2.2558797595418585e-05,
+      "loss": 0.9116,
+      "step": 13620
+    },
+    {
+      "epoch": 0.6949472288788049,
+      "grad_norm": 2.9055182560506125,
+      "learning_rate": 2.248981232645459e-05,
+      "loss": 0.9088,
+      "step": 13630
+    },
+    {
+      "epoch": 0.6954570947840718,
+      "grad_norm": 3.3527635882219475,
+      "learning_rate": 2.242090207964393e-05,
+      "loss": 0.8809,
+      "step": 13640
+    },
+    {
+      "epoch": 0.6959669606893387,
+      "grad_norm": 3.2573139563374136,
+      "learning_rate": 2.2352067042909512e-05,
+      "loss": 0.9004,
+      "step": 13650
+    },
+    {
+      "epoch": 0.6964768265946056,
+      "grad_norm": 2.304191763039682,
+      "learning_rate": 2.2283307403969155e-05,
+      "loss": 0.8939,
+      "step": 13660
+    },
+    {
+      "epoch": 0.6969866924998726,
+      "grad_norm": 4.007114973416113,
+      "learning_rate": 2.2214623350335116e-05,
+      "loss": 0.9064,
+      "step": 13670
+    },
+    {
+      "epoch": 0.6974965584051395,
+      "grad_norm": 3.6062362685383524,
+      "learning_rate": 2.2146015069313426e-05,
+      "loss": 0.8528,
+      "step": 13680
+    },
+    {
+      "epoch": 0.6980064243104064,
+      "grad_norm": 5.213944919568216,
+      "learning_rate": 2.2077482748003585e-05,
+      "loss": 0.9015,
+      "step": 13690
+    },
+    {
+      "epoch": 0.6985162902156733,
+      "grad_norm": 4.001983770515366,
+      "learning_rate": 2.200902657329784e-05,
+      "loss": 0.862,
+      "step": 13700
+    },
+    {
+      "epoch": 0.6990261561209402,
+      "grad_norm": 2.5866784871880375,
+      "learning_rate": 2.194064673188089e-05,
+      "loss": 0.8496,
+      "step": 13710
+    },
+    {
+      "epoch": 0.6995360220262071,
+      "grad_norm": 3.292858427289412,
+      "learning_rate": 2.1872343410229178e-05,
+      "loss": 0.8766,
+      "step": 13720
+    },
+    {
+      "epoch": 0.700045887931474,
+      "grad_norm": 2.7369948622677964,
+      "learning_rate": 2.1804116794610536e-05,
+      "loss": 0.9164,
+      "step": 13730
+    },
+    {
+      "epoch": 0.700555753836741,
+      "grad_norm": 3.266312647688562,
+      "learning_rate": 2.1735967071083567e-05,
+      "loss": 0.9414,
+      "step": 13740
+    },
+    {
+      "epoch": 0.7010656197420079,
+      "grad_norm": 4.511476098623222,
+      "learning_rate": 2.1667894425497192e-05,
+      "loss": 0.9279,
+      "step": 13750
+    },
+    {
+      "epoch": 0.7015754856472748,
+      "grad_norm": 4.531016473727997,
+      "learning_rate": 2.1599899043490195e-05,
+      "loss": 0.9571,
+      "step": 13760
+    },
+    {
+      "epoch": 0.7020853515525417,
+      "grad_norm": 2.912246011878129,
+      "learning_rate": 2.153198111049055e-05,
+      "loss": 0.8981,
+      "step": 13770
+    },
+    {
+      "epoch": 0.7025952174578086,
+      "grad_norm": 2.692086122196558,
+      "learning_rate": 2.1464140811715122e-05,
+      "loss": 0.897,
+      "step": 13780
+    },
+    {
+      "epoch": 0.7031050833630755,
+      "grad_norm": 3.3648953304431592,
+      "learning_rate": 2.1396378332169002e-05,
+      "loss": 0.8253,
+      "step": 13790
+    },
+    {
+      "epoch": 0.7036149492683424,
+      "grad_norm": 2.734002922156205,
+      "learning_rate": 2.1328693856645093e-05,
+      "loss": 0.8855,
+      "step": 13800
+    },
+    {
+      "epoch": 0.7041248151736094,
+      "grad_norm": 3.2681411269094034,
+      "learning_rate": 2.126108756972356e-05,
+      "loss": 0.9766,
+      "step": 13810
+    },
+    {
+      "epoch": 0.7046346810788763,
+      "grad_norm": 2.2947043895088592,
+      "learning_rate": 2.119355965577134e-05,
+      "loss": 0.8715,
+      "step": 13820
+    },
+    {
+      "epoch": 0.7051445469841432,
+      "grad_norm": 3.519515502644016,
+      "learning_rate": 2.1126110298941635e-05,
+      "loss": 0.8414,
+      "step": 13830
+    },
+    {
+      "epoch": 0.7056544128894101,
+      "grad_norm": 2.5251922047442816,
+      "learning_rate": 2.1058739683173484e-05,
+      "loss": 0.8371,
+      "step": 13840
+    },
+    {
+      "epoch": 0.706164278794677,
+      "grad_norm": 3.18827897183748,
+      "learning_rate": 2.0991447992191066e-05,
+      "loss": 0.8544,
+      "step": 13850
+    },
+    {
+      "epoch": 0.7066741446999439,
+      "grad_norm": 6.902658123270844,
+      "learning_rate": 2.0924235409503456e-05,
+      "loss": 0.8439,
+      "step": 13860
+    },
+    {
+      "epoch": 0.7071840106052109,
+      "grad_norm": 4.265874028316936,
+      "learning_rate": 2.0857102118403926e-05,
+      "loss": 0.8546,
+      "step": 13870
+    },
+    {
+      "epoch": 0.7076938765104778,
+      "grad_norm": 2.743027794826126,
+      "learning_rate": 2.0790048301969516e-05,
+      "loss": 0.9121,
+      "step": 13880
+    },
+    {
+      "epoch": 0.7082037424157447,
+      "grad_norm": 4.956407628795556,
+      "learning_rate": 2.0723074143060555e-05,
+      "loss": 0.9309,
+      "step": 13890
+    },
+    {
+      "epoch": 0.7087136083210116,
+      "grad_norm": 6.2790601741914625,
+      "learning_rate": 2.065617982432011e-05,
+      "loss": 0.9223,
+      "step": 13900
+    },
+    {
+      "epoch": 0.7092234742262785,
+      "grad_norm": 3.3010818174838623,
+      "learning_rate": 2.058936552817359e-05,
+      "loss": 0.8884,
+      "step": 13910
+    },
+    {
+      "epoch": 0.7097333401315454,
+      "grad_norm": 2.638150017954024,
+      "learning_rate": 2.0522631436828054e-05,
+      "loss": 0.8843,
+      "step": 13920
+    },
+    {
+      "epoch": 0.7102432060368123,
+      "grad_norm": 3.703275922208582,
+      "learning_rate": 2.0455977732271993e-05,
+      "loss": 0.884,
+      "step": 13930
+    },
+    {
+      "epoch": 0.7107530719420793,
+      "grad_norm": 4.8064675843188125,
+      "learning_rate": 2.0389404596274504e-05,
+      "loss": 0.8608,
+      "step": 13940
+    },
+    {
+      "epoch": 0.7112629378473462,
+      "grad_norm": 3.1860775078031036,
+      "learning_rate": 2.032291221038512e-05,
+      "loss": 0.9085,
+      "step": 13950
+    },
+    {
+      "epoch": 0.7117728037526131,
+      "grad_norm": 2.776924095859808,
+      "learning_rate": 2.025650075593309e-05,
+      "loss": 0.8779,
+      "step": 13960
+    },
+    {
+      "epoch": 0.71228266965788,
+      "grad_norm": 2.7945055013783913,
+      "learning_rate": 2.0190170414026936e-05,
+      "loss": 0.8974,
+      "step": 13970
+    },
+    {
+      "epoch": 0.7127925355631469,
+      "grad_norm": 3.158074061515128,
+      "learning_rate": 2.0123921365554076e-05,
+      "loss": 0.8585,
+      "step": 13980
+    },
+    {
+      "epoch": 0.7133024014684138,
+      "grad_norm": 2.979085824512164,
+      "learning_rate": 2.00577537911801e-05,
+      "loss": 0.8869,
+      "step": 13990
+    },
+    {
+      "epoch": 0.7138122673736808,
+      "grad_norm": 3.1160926759671703,
+      "learning_rate": 1.9991667871348556e-05,
+      "loss": 0.8311,
+      "step": 14000
+    },
+    {
+      "epoch": 0.7143221332789477,
+      "grad_norm": 3.1536851693241785,
+      "learning_rate": 1.992566378628017e-05,
+      "loss": 0.8893,
+      "step": 14010
+    },
+    {
+      "epoch": 0.7148319991842146,
+      "grad_norm": 4.285537613384988,
+      "learning_rate": 1.9859741715972625e-05,
+      "loss": 0.8785,
+      "step": 14020
+    },
+    {
+      "epoch": 0.7153418650894815,
+      "grad_norm": 2.9064731109977875,
+      "learning_rate": 1.9793901840199875e-05,
+      "loss": 0.8297,
+      "step": 14030
+    },
+    {
+      "epoch": 0.7158517309947484,
+      "grad_norm": 2.5062565115060615,
+      "learning_rate": 1.972814433851174e-05,
+      "loss": 0.8612,
+      "step": 14040
+    },
+    {
+      "epoch": 0.7163615969000153,
+      "grad_norm": 3.7803527969691237,
+      "learning_rate": 1.96624693902334e-05,
+      "loss": 0.8716,
+      "step": 14050
+    },
+    {
+      "epoch": 0.7168714628052822,
+      "grad_norm": 2.7454729847134347,
+      "learning_rate": 1.959687717446491e-05,
+      "loss": 0.82,
+      "step": 14060
+    },
+    {
+      "epoch": 0.7173813287105492,
+      "grad_norm": 3.93862340335549,
+      "learning_rate": 1.9531367870080687e-05,
+      "loss": 0.8605,
+      "step": 14070
+    },
+    {
+      "epoch": 0.7178911946158161,
+      "grad_norm": 2.292545994488362,
+      "learning_rate": 1.9465941655729087e-05,
+      "loss": 0.8961,
+      "step": 14080
+    },
+    {
+      "epoch": 0.718401060521083,
+      "grad_norm": 2.934203360637517,
+      "learning_rate": 1.940059870983184e-05,
+      "loss": 0.8075,
+      "step": 14090
+    },
+    {
+      "epoch": 0.7189109264263499,
+      "grad_norm": 4.125718921369046,
+      "learning_rate": 1.9335339210583603e-05,
+      "loss": 0.9167,
+      "step": 14100
+    },
+    {
+      "epoch": 0.7194207923316168,
+      "grad_norm": 3.872383869661933,
+      "learning_rate": 1.9270163335951467e-05,
+      "loss": 0.8645,
+      "step": 14110
+    },
+    {
+      "epoch": 0.7199306582368837,
+      "grad_norm": 2.349309012944774,
+      "learning_rate": 1.920507126367448e-05,
+      "loss": 0.919,
+      "step": 14120
+    },
+    {
+      "epoch": 0.7204405241421507,
+      "grad_norm": 2.7003543700050936,
+      "learning_rate": 1.9140063171263145e-05,
+      "loss": 0.8824,
+      "step": 14130
+    },
+    {
+      "epoch": 0.7209503900474176,
+      "grad_norm": 2.7056845960872304,
+      "learning_rate": 1.907513923599894e-05,
+      "loss": 0.8967,
+      "step": 14140
+    },
+    {
+      "epoch": 0.7214602559526845,
+      "grad_norm": 3.9644399914535824,
+      "learning_rate": 1.9010299634933902e-05,
+      "loss": 0.8854,
+      "step": 14150
+    },
+    {
+      "epoch": 0.7219701218579514,
+      "grad_norm": 2.7756561472506913,
+      "learning_rate": 1.8945544544889977e-05,
+      "loss": 0.8218,
+      "step": 14160
+    },
+    {
+      "epoch": 0.7224799877632183,
+      "grad_norm": 5.018500044878095,
+      "learning_rate": 1.888087414245873e-05,
+      "loss": 0.8744,
+      "step": 14170
+    },
+    {
+      "epoch": 0.7229898536684852,
+      "grad_norm": 2.296518150657654,
+      "learning_rate": 1.8816288604000755e-05,
+      "loss": 0.8981,
+      "step": 14180
+    },
+    {
+      "epoch": 0.7234997195737521,
+      "grad_norm": 2.5659517189639183,
+      "learning_rate": 1.8751788105645174e-05,
+      "loss": 0.9454,
+      "step": 14190
+    },
+    {
+      "epoch": 0.7240095854790191,
+      "grad_norm": 3.8505739486429826,
+      "learning_rate": 1.8687372823289296e-05,
+      "loss": 0.868,
+      "step": 14200
+    },
+    {
+      "epoch": 0.724519451384286,
+      "grad_norm": 3.8879727455834847,
+      "learning_rate": 1.86230429325979e-05,
+      "loss": 0.8202,
+      "step": 14210
+    },
+    {
+      "epoch": 0.7250293172895529,
+      "grad_norm": 2.7035004153604034,
+      "learning_rate": 1.8558798609003052e-05,
+      "loss": 0.9175,
+      "step": 14220
+    },
+    {
+      "epoch": 0.7255391831948198,
+      "grad_norm": 3.0272382276993306,
+      "learning_rate": 1.8494640027703324e-05,
+      "loss": 0.8689,
+      "step": 14230
+    },
+    {
+      "epoch": 0.7260490491000867,
+      "grad_norm": 3.3846551926860147,
+      "learning_rate": 1.8430567363663597e-05,
+      "loss": 0.846,
+      "step": 14240
+    },
+    {
+      "epoch": 0.7265589150053536,
+      "grad_norm": 3.2673408549133502,
+      "learning_rate": 1.8366580791614312e-05,
+      "loss": 0.911,
+      "step": 14250
+    },
+    {
+      "epoch": 0.7270687809106205,
+      "grad_norm": 3.2139816945562423,
+      "learning_rate": 1.8302680486051265e-05,
+      "loss": 0.9208,
+      "step": 14260
+    },
+    {
+      "epoch": 0.7275786468158875,
+      "grad_norm": 5.368949554679828,
+      "learning_rate": 1.8238866621234912e-05,
+      "loss": 0.8395,
+      "step": 14270
+    },
+    {
+      "epoch": 0.7280885127211544,
+      "grad_norm": 3.438148967967155,
+      "learning_rate": 1.817513937118999e-05,
+      "loss": 0.8703,
+      "step": 14280
+    },
+    {
+      "epoch": 0.7285983786264213,
+      "grad_norm": 2.5803615764853287,
+      "learning_rate": 1.8111498909705093e-05,
+      "loss": 0.8614,
+      "step": 14290
+    },
+    {
+      "epoch": 0.7291082445316882,
+      "grad_norm": 4.9536047487293935,
+      "learning_rate": 1.804794541033203e-05,
+      "loss": 0.8683,
+      "step": 14300
+    },
+    {
+      "epoch": 0.7296181104369551,
+      "grad_norm": 2.527176490475443,
+      "learning_rate": 1.7984479046385557e-05,
+      "loss": 0.9015,
+      "step": 14310
+    },
+    {
+      "epoch": 0.730127976342222,
+      "grad_norm": 2.304452984322168,
+      "learning_rate": 1.792109999094275e-05,
+      "loss": 0.8531,
+      "step": 14320
+    },
+    {
+      "epoch": 0.730637842247489,
+      "grad_norm": 3.681104743558922,
+      "learning_rate": 1.7857808416842615e-05,
+      "loss": 0.8987,
+      "step": 14330
+    },
+    {
+      "epoch": 0.7311477081527559,
+      "grad_norm": 4.383437339124014,
+      "learning_rate": 1.7794604496685573e-05,
+      "loss": 0.8218,
+      "step": 14340
+    },
+    {
+      "epoch": 0.7316575740580228,
+      "grad_norm": 3.2501961003202005,
+      "learning_rate": 1.773148840283301e-05,
+      "loss": 0.8854,
+      "step": 14350
+    },
+    {
+      "epoch": 0.7321674399632897,
+      "grad_norm": 2.9076586758259197,
+      "learning_rate": 1.766846030740682e-05,
+      "loss": 0.8815,
+      "step": 14360
+    },
+    {
+      "epoch": 0.7326773058685566,
+      "grad_norm": 2.169529339937437,
+      "learning_rate": 1.7605520382288897e-05,
+      "loss": 0.8767,
+      "step": 14370
+    },
+    {
+      "epoch": 0.7331871717738235,
+      "grad_norm": 3.7153738750289653,
+      "learning_rate": 1.7542668799120688e-05,
+      "loss": 0.9442,
+      "step": 14380
+    },
+    {
+      "epoch": 0.7336970376790904,
+      "grad_norm": 3.050860282460858,
+      "learning_rate": 1.747990572930276e-05,
+      "loss": 0.9095,
+      "step": 14390
+    },
+    {
+      "epoch": 0.7342069035843574,
+      "grad_norm": 5.002041841328034,
+      "learning_rate": 1.7417231343994267e-05,
+      "loss": 0.914,
+      "step": 14400
+    },
+    {
+      "epoch": 0.7347167694896243,
+      "grad_norm": 2.9088511401907766,
+      "learning_rate": 1.735464581411252e-05,
+      "loss": 0.8855,
+      "step": 14410
+    },
+    {
+      "epoch": 0.7352266353948912,
+      "grad_norm": 3.7862878073377937,
+      "learning_rate": 1.7292149310332513e-05,
+      "loss": 0.8165,
+      "step": 14420
+    },
+    {
+      "epoch": 0.7357365013001581,
+      "grad_norm": 2.7539436495085603,
+      "learning_rate": 1.7229742003086463e-05,
+      "loss": 0.8843,
+      "step": 14430
+    },
+    {
+      "epoch": 0.736246367205425,
+      "grad_norm": 3.1947822977994984,
+      "learning_rate": 1.7167424062563348e-05,
+      "loss": 0.8858,
+      "step": 14440
+    },
+    {
+      "epoch": 0.7367562331106919,
+      "grad_norm": 4.095285565190369,
+      "learning_rate": 1.7105195658708412e-05,
+      "loss": 0.9358,
+      "step": 14450
+    },
+    {
+      "epoch": 0.7372660990159589,
+      "grad_norm": 2.7962148323450693,
+      "learning_rate": 1.7043056961222796e-05,
+      "loss": 0.8718,
+      "step": 14460
+    },
+    {
+      "epoch": 0.7377759649212258,
+      "grad_norm": 2.7978242591498623,
+      "learning_rate": 1.698100813956289e-05,
+      "loss": 0.9113,
+      "step": 14470
+    },
+    {
+      "epoch": 0.7382858308264927,
+      "grad_norm": 2.5347340645021075,
+      "learning_rate": 1.691904936294009e-05,
+      "loss": 0.842,
+      "step": 14480
+    },
+    {
+      "epoch": 0.7387956967317596,
+      "grad_norm": 3.7318593907692277,
+      "learning_rate": 1.68571808003202e-05,
+      "loss": 0.9026,
+      "step": 14490
+    },
+    {
+      "epoch": 0.7393055626370265,
+      "grad_norm": 3.219925944875457,
+      "learning_rate": 1.6795402620422968e-05,
+      "loss": 0.8568,
+      "step": 14500
+    },
+    {
+      "epoch": 0.7398154285422934,
+      "grad_norm": 3.5941302291020305,
+      "learning_rate": 1.673371499172174e-05,
+      "loss": 0.8702,
+      "step": 14510
+    },
+    {
+      "epoch": 0.7403252944475603,
+      "grad_norm": 9.173447786756226,
+      "learning_rate": 1.6672118082442823e-05,
+      "loss": 0.8915,
+      "step": 14520
+    },
+    {
+      "epoch": 0.7408351603528273,
+      "grad_norm": 2.3033144486720065,
+      "learning_rate": 1.6610612060565234e-05,
+      "loss": 0.9019,
+      "step": 14530
+    },
+    {
+      "epoch": 0.7413450262580942,
+      "grad_norm": 3.624696375079677,
+      "learning_rate": 1.6549197093820018e-05,
+      "loss": 0.8692,
+      "step": 14540
+    },
+    {
+      "epoch": 0.7418548921633611,
+      "grad_norm": 2.8557781803819506,
+      "learning_rate": 1.6487873349690032e-05,
+      "loss": 0.8348,
+      "step": 14550
+    },
+    {
+      "epoch": 0.742364758068628,
+      "grad_norm": 3.6671225315913323,
+      "learning_rate": 1.6426640995409232e-05,
+      "loss": 0.8822,
+      "step": 14560
+    },
+    {
+      "epoch": 0.7428746239738949,
+      "grad_norm": 2.796368915798219,
+      "learning_rate": 1.6365500197962463e-05,
+      "loss": 0.881,
+      "step": 14570
+    },
+    {
+      "epoch": 0.7433844898791617,
+      "grad_norm": 2.750712257117753,
+      "learning_rate": 1.6304451124084823e-05,
+      "loss": 0.8599,
+      "step": 14580
+    },
+    {
+      "epoch": 0.7438943557844286,
+      "grad_norm": 4.038589338654805,
+      "learning_rate": 1.6243493940261295e-05,
+      "loss": 0.8748,
+      "step": 14590
+    },
+    {
+      "epoch": 0.7444042216896956,
+      "grad_norm": 4.197868213550896,
+      "learning_rate": 1.618262881272626e-05,
+      "loss": 0.8556,
+      "step": 14600
+    },
+    {
+      "epoch": 0.7449140875949625,
+      "grad_norm": 3.061273404676564,
+      "learning_rate": 1.612185590746304e-05,
+      "loss": 0.8258,
+      "step": 14610
+    },
+    {
+      "epoch": 0.7454239535002294,
+      "grad_norm": 5.218000571306212,
+      "learning_rate": 1.6061175390203525e-05,
+      "loss": 0.8827,
+      "step": 14620
+    },
+    {
+      "epoch": 0.7459338194054963,
+      "grad_norm": 2.494457198033094,
+      "learning_rate": 1.6000587426427592e-05,
+      "loss": 0.8294,
+      "step": 14630
+    },
+    {
+      "epoch": 0.7464436853107632,
+      "grad_norm": 3.774928589841261,
+      "learning_rate": 1.5940092181362742e-05,
+      "loss": 0.8692,
+      "step": 14640
+    },
+    {
+      "epoch": 0.7469535512160301,
+      "grad_norm": 5.045548659651567,
+      "learning_rate": 1.5879689819983627e-05,
+      "loss": 0.8996,
+      "step": 14650
+    },
+    {
+      "epoch": 0.747463417121297,
+      "grad_norm": 3.43630615014554,
+      "learning_rate": 1.581938050701159e-05,
+      "loss": 0.8446,
+      "step": 14660
+    },
+    {
+      "epoch": 0.747973283026564,
+      "grad_norm": 2.9483501469705296,
+      "learning_rate": 1.575916440691423e-05,
+      "loss": 0.8824,
+      "step": 14670
+    },
+    {
+      "epoch": 0.7484831489318309,
+      "grad_norm": 3.744326663008319,
+      "learning_rate": 1.569904168390495e-05,
+      "loss": 0.9082,
+      "step": 14680
+    },
+    {
+      "epoch": 0.7489930148370978,
+      "grad_norm": 3.346178111816128,
+      "learning_rate": 1.5639012501942495e-05,
+      "loss": 0.8943,
+      "step": 14690
+    },
+    {
+      "epoch": 0.7495028807423647,
+      "grad_norm": 2.705456054021861,
+      "learning_rate": 1.5579077024730555e-05,
+      "loss": 0.8203,
+      "step": 14700
+    },
+    {
+      "epoch": 0.7500127466476316,
+      "grad_norm": 5.027037292349675,
+      "learning_rate": 1.5519235415717236e-05,
+      "loss": 0.8954,
+      "step": 14710
+    },
+    {
+      "epoch": 0.7505226125528985,
+      "grad_norm": 3.1064826233355918,
+      "learning_rate": 1.5459487838094677e-05,
+      "loss": 0.8163,
+      "step": 14720
+    },
+    {
+      "epoch": 0.7510324784581655,
+      "grad_norm": 2.694475571297849,
+      "learning_rate": 1.5399834454798594e-05,
+      "loss": 0.8843,
+      "step": 14730
+    },
+    {
+      "epoch": 0.7515423443634324,
+      "grad_norm": 3.327058419243026,
+      "learning_rate": 1.53402754285078e-05,
+      "loss": 0.8939,
+      "step": 14740
+    },
+    {
+      "epoch": 0.7520522102686993,
+      "grad_norm": 6.102691369797055,
+      "learning_rate": 1.5280810921643856e-05,
+      "loss": 0.8729,
+      "step": 14750
+    },
+    {
+      "epoch": 0.7525620761739662,
+      "grad_norm": 2.931116244008933,
+      "learning_rate": 1.5221441096370459e-05,
+      "loss": 0.8847,
+      "step": 14760
+    },
+    {
+      "epoch": 0.7530719420792331,
+      "grad_norm": 3.2689777653057277,
+      "learning_rate": 1.516216611459323e-05,
+      "loss": 0.9207,
+      "step": 14770
+    },
+    {
+      "epoch": 0.7535818079845,
+      "grad_norm": 4.005422842172153,
+      "learning_rate": 1.5102986137959001e-05,
+      "loss": 0.8815,
+      "step": 14780
+    },
+    {
+      "epoch": 0.754091673889767,
+      "grad_norm": 3.500402263868153,
+      "learning_rate": 1.5043901327855647e-05,
+      "loss": 0.8342,
+      "step": 14790
+    },
+    {
+      "epoch": 0.7546015397950339,
+      "grad_norm": 3.9180680848894043,
+      "learning_rate": 1.4984911845411453e-05,
+      "loss": 0.9318,
+      "step": 14800
+    },
+    {
+      "epoch": 0.7551114057003008,
+      "grad_norm": 2.6078203177003796,
+      "learning_rate": 1.492601785149475e-05,
+      "loss": 0.878,
+      "step": 14810
+    },
+    {
+      "epoch": 0.7556212716055677,
+      "grad_norm": 3.8300001695915973,
+      "learning_rate": 1.4867219506713458e-05,
+      "loss": 0.958,
+      "step": 14820
+    },
+    {
+      "epoch": 0.7561311375108346,
+      "grad_norm": 3.3155999260231788,
+      "learning_rate": 1.480851697141466e-05,
+      "loss": 0.8031,
+      "step": 14830
+    },
+    {
+      "epoch": 0.7566410034161015,
+      "grad_norm": 5.93548750073978,
+      "learning_rate": 1.4749910405684197e-05,
+      "loss": 0.8614,
+      "step": 14840
+    },
+    {
+      "epoch": 0.7571508693213684,
+      "grad_norm": 8.015838481204414,
+      "learning_rate": 1.469139996934611e-05,
+      "loss": 0.8456,
+      "step": 14850
+    },
+    {
+      "epoch": 0.7576607352266354,
+      "grad_norm": 2.719745178839152,
+      "learning_rate": 1.463298582196238e-05,
+      "loss": 0.8895,
+      "step": 14860
+    },
+    {
+      "epoch": 0.7581706011319023,
+      "grad_norm": 4.1208125735007775,
+      "learning_rate": 1.4574668122832346e-05,
+      "loss": 0.8314,
+      "step": 14870
+    },
+    {
+      "epoch": 0.7586804670371692,
+      "grad_norm": 5.103220951819016,
+      "learning_rate": 1.4516447030992342e-05,
+      "loss": 0.8828,
+      "step": 14880
+    },
+    {
+      "epoch": 0.7591903329424361,
+      "grad_norm": 4.2279259418311765,
+      "learning_rate": 1.4458322705215244e-05,
+      "loss": 0.949,
+      "step": 14890
+    },
+    {
+      "epoch": 0.759700198847703,
+      "grad_norm": 3.6821822208934,
+      "learning_rate": 1.4400295304010041e-05,
+      "loss": 0.9127,
+      "step": 14900
+    },
+    {
+      "epoch": 0.7602100647529699,
+      "grad_norm": 4.433950310983757,
+      "learning_rate": 1.434236498562141e-05,
+      "loss": 0.8777,
+      "step": 14910
+    },
+    {
+      "epoch": 0.7607199306582368,
+      "grad_norm": 3.258546059081396,
+      "learning_rate": 1.4284531908029253e-05,
+      "loss": 0.8015,
+      "step": 14920
+    },
+    {
+      "epoch": 0.7612297965635038,
+      "grad_norm": 2.3659649654317016,
+      "learning_rate": 1.4226796228948337e-05,
+      "loss": 0.851,
+      "step": 14930
+    },
+    {
+      "epoch": 0.7617396624687707,
+      "grad_norm": 2.2409439689317914,
+      "learning_rate": 1.4169158105827768e-05,
+      "loss": 0.8608,
+      "step": 14940
+    },
+    {
+      "epoch": 0.7622495283740376,
+      "grad_norm": 3.0715204978103725,
+      "learning_rate": 1.4111617695850631e-05,
+      "loss": 0.7935,
+      "step": 14950
+    },
+    {
+      "epoch": 0.7627593942793045,
+      "grad_norm": 4.267005282076962,
+      "learning_rate": 1.4054175155933541e-05,
+      "loss": 0.9154,
+      "step": 14960
+    },
+    {
+      "epoch": 0.7632692601845714,
+      "grad_norm": 2.4892490009386243,
+      "learning_rate": 1.399683064272621e-05,
+      "loss": 0.8468,
+      "step": 14970
+    },
+    {
+      "epoch": 0.7637791260898383,
+      "grad_norm": 3.031762626944372,
+      "learning_rate": 1.3939584312611015e-05,
+      "loss": 0.8229,
+      "step": 14980
+    },
+    {
+      "epoch": 0.7642889919951052,
+      "grad_norm": 2.7553302663332615,
+      "learning_rate": 1.388243632170263e-05,
+      "loss": 0.8596,
+      "step": 14990
+    },
+    {
+      "epoch": 0.7647988579003722,
+      "grad_norm": 3.4874566989678146,
+      "learning_rate": 1.3825386825847459e-05,
+      "loss": 0.8195,
+      "step": 15000
+    },
+    {
+      "epoch": 0.7653087238056391,
+      "grad_norm": 2.408476831561337,
+      "learning_rate": 1.3768435980623407e-05,
+      "loss": 0.8469,
+      "step": 15010
+    },
+    {
+      "epoch": 0.765818589710906,
+      "grad_norm": 4.117755860054798,
+      "learning_rate": 1.371158394133928e-05,
+      "loss": 0.967,
+      "step": 15020
+    },
+    {
+      "epoch": 0.7663284556161729,
+      "grad_norm": 4.81413553922508,
+      "learning_rate": 1.3654830863034462e-05,
+      "loss": 0.8602,
+      "step": 15030
+    },
+    {
+      "epoch": 0.7668383215214398,
+      "grad_norm": 2.831741053689239,
+      "learning_rate": 1.359817690047846e-05,
+      "loss": 0.8479,
+      "step": 15040
+    },
+    {
+      "epoch": 0.7673481874267067,
+      "grad_norm": 3.744551247110337,
+      "learning_rate": 1.3541622208170467e-05,
+      "loss": 0.8596,
+      "step": 15050
+    },
+    {
+      "epoch": 0.7678580533319737,
+      "grad_norm": 3.7148036923374574,
+      "learning_rate": 1.3485166940339016e-05,
+      "loss": 0.9524,
+      "step": 15060
+    },
+    {
+      "epoch": 0.7683679192372406,
+      "grad_norm": 4.335382354236005,
+      "learning_rate": 1.3428811250941414e-05,
+      "loss": 0.9149,
+      "step": 15070
+    },
+    {
+      "epoch": 0.7688777851425075,
+      "grad_norm": 2.9316183575322983,
+      "learning_rate": 1.3372555293663514e-05,
+      "loss": 0.8977,
+      "step": 15080
+    },
+    {
+      "epoch": 0.7693876510477744,
+      "grad_norm": 3.10946155411863,
+      "learning_rate": 1.3316399221919074e-05,
+      "loss": 0.8478,
+      "step": 15090
+    },
+    {
+      "epoch": 0.7698975169530413,
+      "grad_norm": 4.72995402941815,
+      "learning_rate": 1.3260343188849572e-05,
+      "loss": 0.8608,
+      "step": 15100
+    },
+    {
+      "epoch": 0.7704073828583082,
+      "grad_norm": 3.2196110816338464,
+      "learning_rate": 1.320438734732361e-05,
+      "loss": 0.8689,
+      "step": 15110
+    },
+    {
+      "epoch": 0.7709172487635751,
+      "grad_norm": 2.5940976010805183,
+      "learning_rate": 1.3148531849936574e-05,
+      "loss": 0.919,
+      "step": 15120
+    },
+    {
+      "epoch": 0.7714271146688421,
+      "grad_norm": 3.5281290333681996,
+      "learning_rate": 1.3092776849010202e-05,
+      "loss": 0.8215,
+      "step": 15130
+    },
+    {
+      "epoch": 0.771936980574109,
+      "grad_norm": 2.3083401803024284,
+      "learning_rate": 1.3037122496592164e-05,
+      "loss": 0.807,
+      "step": 15140
+    },
+    {
+      "epoch": 0.7724468464793759,
+      "grad_norm": 4.48237544587891,
+      "learning_rate": 1.2981568944455707e-05,
+      "loss": 0.8607,
+      "step": 15150
+    },
+    {
+      "epoch": 0.7729567123846428,
+      "grad_norm": 5.2965861701358214,
+      "learning_rate": 1.2926116344099087e-05,
+      "loss": 0.8734,
+      "step": 15160
+    },
+    {
+      "epoch": 0.7734665782899097,
+      "grad_norm": 2.2086316730073556,
+      "learning_rate": 1.2870764846745365e-05,
+      "loss": 0.8035,
+      "step": 15170
+    },
+    {
+      "epoch": 0.7739764441951766,
+      "grad_norm": 4.2610353748951955,
+      "learning_rate": 1.2815514603341822e-05,
+      "loss": 0.8885,
+      "step": 15180
+    },
+    {
+      "epoch": 0.7744863101004436,
+      "grad_norm": 2.7360133633418293,
+      "learning_rate": 1.2760365764559634e-05,
+      "loss": 0.9074,
+      "step": 15190
+    },
+    {
+      "epoch": 0.7749961760057105,
+      "grad_norm": 3.9738716131027885,
+      "learning_rate": 1.2705318480793432e-05,
+      "loss": 0.8907,
+      "step": 15200
+    },
+    {
+      "epoch": 0.7755060419109774,
+      "grad_norm": 2.954769888251,
+      "learning_rate": 1.2650372902160906e-05,
+      "loss": 0.8772,
+      "step": 15210
+    },
+    {
+      "epoch": 0.7760159078162443,
+      "grad_norm": 2.1621609292566855,
+      "learning_rate": 1.2595529178502386e-05,
+      "loss": 0.8465,
+      "step": 15220
+    },
+    {
+      "epoch": 0.7765257737215112,
+      "grad_norm": 3.010115001320829,
+      "learning_rate": 1.2540787459380421e-05,
+      "loss": 0.8738,
+      "step": 15230
+    },
+    {
+      "epoch": 0.7770356396267781,
+      "grad_norm": 2.762540298355899,
+      "learning_rate": 1.2486147894079442e-05,
+      "loss": 0.8319,
+      "step": 15240
+    },
+    {
+      "epoch": 0.777545505532045,
+      "grad_norm": 4.092547126744568,
+      "learning_rate": 1.2431610631605234e-05,
+      "loss": 0.8491,
+      "step": 15250
+    },
+    {
+      "epoch": 0.778055371437312,
+      "grad_norm": 2.8888610964288404,
+      "learning_rate": 1.2377175820684628e-05,
+      "loss": 0.7645,
+      "step": 15260
+    },
+    {
+      "epoch": 0.7785652373425789,
+      "grad_norm": 3.8152499348556907,
+      "learning_rate": 1.2322843609765056e-05,
+      "loss": 0.9073,
+      "step": 15270
+    },
+    {
+      "epoch": 0.7790751032478458,
+      "grad_norm": 3.1391522021865192,
+      "learning_rate": 1.2268614147014152e-05,
+      "loss": 0.9408,
+      "step": 15280
+    },
+    {
+      "epoch": 0.7795849691531127,
+      "grad_norm": 2.7293474454182705,
+      "learning_rate": 1.2214487580319334e-05,
+      "loss": 0.9204,
+      "step": 15290
+    },
+    {
+      "epoch": 0.7800948350583796,
+      "grad_norm": 4.976415538766265,
+      "learning_rate": 1.2160464057287479e-05,
+      "loss": 0.8978,
+      "step": 15300
+    },
+    {
+      "epoch": 0.7806047009636465,
+      "grad_norm": 2.3275484642450968,
+      "learning_rate": 1.2106543725244357e-05,
+      "loss": 0.8873,
+      "step": 15310
+    },
+    {
+      "epoch": 0.7811145668689135,
+      "grad_norm": 2.300791730744753,
+      "learning_rate": 1.2052726731234409e-05,
+      "loss": 0.8828,
+      "step": 15320
+    },
+    {
+      "epoch": 0.7816244327741804,
+      "grad_norm": 3.2802508324260646,
+      "learning_rate": 1.199901322202024e-05,
+      "loss": 0.9419,
+      "step": 15330
+    },
+    {
+      "epoch": 0.7821342986794473,
+      "grad_norm": 3.410662489857846,
+      "learning_rate": 1.1945403344082234e-05,
+      "loss": 0.8756,
+      "step": 15340
+    },
+    {
+      "epoch": 0.7826441645847142,
+      "grad_norm": 3.4264529388087506,
+      "learning_rate": 1.1891897243618182e-05,
+      "loss": 0.9008,
+      "step": 15350
+    },
+    {
+      "epoch": 0.7831540304899811,
+      "grad_norm": 2.8127713515538724,
+      "learning_rate": 1.1838495066542842e-05,
+      "loss": 0.9028,
+      "step": 15360
+    },
+    {
+      "epoch": 0.783663896395248,
+      "grad_norm": 2.641184167083469,
+      "learning_rate": 1.1785196958487627e-05,
+      "loss": 0.9116,
+      "step": 15370
+    },
+    {
+      "epoch": 0.7841737623005149,
+      "grad_norm": 3.372592168887771,
+      "learning_rate": 1.1732003064800045e-05,
+      "loss": 0.816,
+      "step": 15380
+    },
+    {
+      "epoch": 0.7846836282057819,
+      "grad_norm": 4.8521661786297265,
+      "learning_rate": 1.1678913530543523e-05,
+      "loss": 0.8601,
+      "step": 15390
+    },
+    {
+      "epoch": 0.7851934941110488,
+      "grad_norm": 2.9947310910954843,
+      "learning_rate": 1.1625928500496774e-05,
+      "loss": 0.8873,
+      "step": 15400
+    },
+    {
+      "epoch": 0.7857033600163157,
+      "grad_norm": 2.7372429848001567,
+      "learning_rate": 1.1573048119153623e-05,
+      "loss": 0.843,
+      "step": 15410
+    },
+    {
+      "epoch": 0.7862132259215826,
+      "grad_norm": 4.325793719043827,
+      "learning_rate": 1.1520272530722453e-05,
+      "loss": 0.9278,
+      "step": 15420
+    },
+    {
+      "epoch": 0.7867230918268495,
+      "grad_norm": 2.895079457895283,
+      "learning_rate": 1.1467601879125883e-05,
+      "loss": 0.8213,
+      "step": 15430
+    },
+    {
+      "epoch": 0.7872329577321164,
+      "grad_norm": 2.673845770405087,
+      "learning_rate": 1.1415036308000365e-05,
+      "loss": 0.8808,
+      "step": 15440
+    },
+    {
+      "epoch": 0.7877428236373833,
+      "grad_norm": 4.681838343953212,
+      "learning_rate": 1.136257596069577e-05,
+      "loss": 0.9322,
+      "step": 15450
+    },
+    {
+      "epoch": 0.7882526895426503,
+      "grad_norm": 2.570966125797705,
+      "learning_rate": 1.1310220980275082e-05,
+      "loss": 0.9371,
+      "step": 15460
+    },
+    {
+      "epoch": 0.7887625554479172,
+      "grad_norm": 2.2118704016933215,
+      "learning_rate": 1.1257971509513832e-05,
+      "loss": 0.8454,
+      "step": 15470
+    },
+    {
+      "epoch": 0.7892724213531841,
+      "grad_norm": 3.3732300776354394,
+      "learning_rate": 1.1205827690899927e-05,
+      "loss": 0.8697,
+      "step": 15480
+    },
+    {
+      "epoch": 0.789782287258451,
+      "grad_norm": 4.036479967416448,
+      "learning_rate": 1.1153789666633096e-05,
+      "loss": 0.8722,
+      "step": 15490
+    },
+    {
+      "epoch": 0.7902921531637179,
+      "grad_norm": 3.323040700587493,
+      "learning_rate": 1.1101857578624564e-05,
+      "loss": 0.9044,
+      "step": 15500
+    },
+    {
+      "epoch": 0.7908020190689848,
+      "grad_norm": 3.6008204869697797,
+      "learning_rate": 1.1050031568496672e-05,
+      "loss": 0.8795,
+      "step": 15510
+    },
+    {
+      "epoch": 0.7913118849742518,
+      "grad_norm": 3.398639295785761,
+      "learning_rate": 1.0998311777582476e-05,
+      "loss": 0.8519,
+      "step": 15520
+    },
+    {
+      "epoch": 0.7918217508795187,
+      "grad_norm": 2.233934894756387,
+      "learning_rate": 1.094669834692535e-05,
+      "loss": 0.8117,
+      "step": 15530
+    },
+    {
+      "epoch": 0.7923316167847856,
+      "grad_norm": 2.690786504073156,
+      "learning_rate": 1.0895191417278661e-05,
+      "loss": 0.933,
+      "step": 15540
+    },
+    {
+      "epoch": 0.7928414826900525,
+      "grad_norm": 4.0881895800161345,
+      "learning_rate": 1.0843791129105285e-05,
+      "loss": 0.8919,
+      "step": 15550
+    },
+    {
+      "epoch": 0.7933513485953194,
+      "grad_norm": 4.155974489867237,
+      "learning_rate": 1.0792497622577325e-05,
+      "loss": 0.8276,
+      "step": 15560
+    },
+    {
+      "epoch": 0.7938612145005863,
+      "grad_norm": 2.6755614926094147,
+      "learning_rate": 1.074131103757564e-05,
+      "loss": 0.8212,
+      "step": 15570
+    },
+    {
+      "epoch": 0.7943710804058532,
+      "grad_norm": 3.438908813746616,
+      "learning_rate": 1.0690231513689548e-05,
+      "loss": 0.8987,
+      "step": 15580
+    },
+    {
+      "epoch": 0.7948809463111202,
+      "grad_norm": 5.257275543448024,
+      "learning_rate": 1.0639259190216377e-05,
+      "loss": 0.9214,
+      "step": 15590
+    },
+    {
+      "epoch": 0.7953908122163871,
+      "grad_norm": 2.7140298504555154,
+      "learning_rate": 1.0588394206161106e-05,
+      "loss": 0.8997,
+      "step": 15600
+    },
+    {
+      "epoch": 0.795900678121654,
+      "grad_norm": 2.471384156286019,
+      "learning_rate": 1.0537636700236053e-05,
+      "loss": 0.7906,
+      "step": 15610
+    },
+    {
+      "epoch": 0.7964105440269209,
+      "grad_norm": 4.130898081102724,
+      "learning_rate": 1.0486986810860334e-05,
+      "loss": 0.8608,
+      "step": 15620
+    },
+    {
+      "epoch": 0.7969204099321878,
+      "grad_norm": 3.4187367338017514,
+      "learning_rate": 1.0436444676159674e-05,
+      "loss": 0.8403,
+      "step": 15630
+    },
+    {
+      "epoch": 0.7974302758374547,
+      "grad_norm": 3.990897781034759,
+      "learning_rate": 1.0386010433965915e-05,
+      "loss": 0.8962,
+      "step": 15640
+    },
+    {
+      "epoch": 0.7979401417427217,
+      "grad_norm": 2.9080947384924225,
+      "learning_rate": 1.0335684221816654e-05,
+      "loss": 0.8383,
+      "step": 15650
+    },
+    {
+      "epoch": 0.7984500076479886,
+      "grad_norm": 3.940054399989881,
+      "learning_rate": 1.0285466176954905e-05,
+      "loss": 0.8969,
+      "step": 15660
+    },
+    {
+      "epoch": 0.7989598735532555,
+      "grad_norm": 5.037355841684986,
+      "learning_rate": 1.0235356436328675e-05,
+      "loss": 0.9037,
+      "step": 15670
+    },
+    {
+      "epoch": 0.7994697394585224,
+      "grad_norm": 2.077271792357116,
+      "learning_rate": 1.0185355136590669e-05,
+      "loss": 0.8897,
+      "step": 15680
+    },
+    {
+      "epoch": 0.7999796053637893,
+      "grad_norm": 2.807216356773431,
+      "learning_rate": 1.013546241409779e-05,
+      "loss": 0.8768,
+      "step": 15690
+    },
+    {
+      "epoch": 0.8004894712690562,
+      "grad_norm": 3.772436361366961,
+      "learning_rate": 1.0085678404910936e-05,
+      "loss": 0.8384,
+      "step": 15700
+    },
+    {
+      "epoch": 0.8009993371743231,
+      "grad_norm": 3.6199871979687526,
+      "learning_rate": 1.003600324479443e-05,
+      "loss": 0.8623,
+      "step": 15710
+    },
+    {
+      "epoch": 0.8015092030795901,
+      "grad_norm": 3.3410794161187494,
+      "learning_rate": 9.986437069215859e-06,
+      "loss": 0.7986,
+      "step": 15720
+    },
+    {
+      "epoch": 0.802019068984857,
+      "grad_norm": 3.1639330607365226,
+      "learning_rate": 9.936980013345543e-06,
+      "loss": 0.7878,
+      "step": 15730
+    },
+    {
+      "epoch": 0.8025289348901239,
+      "grad_norm": 3.9782955261658413,
+      "learning_rate": 9.88763221205623e-06,
+      "loss": 0.826,
+      "step": 15740
+    },
+    {
+      "epoch": 0.8030388007953908,
+      "grad_norm": 2.787918275512712,
+      "learning_rate": 9.838393799922735e-06,
+      "loss": 0.8168,
+      "step": 15750
+    },
+    {
+      "epoch": 0.8035486667006577,
+      "grad_norm": 3.874444869921306,
+      "learning_rate": 9.789264911221546e-06,
+      "loss": 0.8714,
+      "step": 15760
+    },
+    {
+      "epoch": 0.8040585326059246,
+      "grad_norm": 3.0730481094982647,
+      "learning_rate": 9.740245679930526e-06,
+      "loss": 0.8761,
+      "step": 15770
+    },
+    {
+      "epoch": 0.8045683985111916,
+      "grad_norm": 4.054732819960556,
+      "learning_rate": 9.6913362397284e-06,
+      "loss": 0.8234,
+      "step": 15780
+    },
+    {
+      "epoch": 0.8050782644164585,
+      "grad_norm": 2.5121594783856445,
+      "learning_rate": 9.642536723994572e-06,
+      "loss": 0.855,
+      "step": 15790
+    },
+    {
+      "epoch": 0.8055881303217254,
+      "grad_norm": 4.135088131556567,
+      "learning_rate": 9.593847265808637e-06,
+      "loss": 0.8791,
+      "step": 15800
+    },
+    {
+      "epoch": 0.8060979962269923,
+      "grad_norm": 4.82689473053539,
+      "learning_rate": 9.545267997950052e-06,
+      "loss": 0.8946,
+      "step": 15810
+    },
+    {
+      "epoch": 0.8066078621322592,
+      "grad_norm": 4.492685827978297,
+      "learning_rate": 9.496799052897786e-06,
+      "loss": 0.8565,
+      "step": 15820
+    },
+    {
+      "epoch": 0.8071177280375261,
+      "grad_norm": 2.8347254300348443,
+      "learning_rate": 9.448440562829958e-06,
+      "loss": 0.8884,
+      "step": 15830
+    },
+    {
+      "epoch": 0.807627593942793,
+      "grad_norm": 4.0989654125585275,
+      "learning_rate": 9.400192659623442e-06,
+      "loss": 0.8717,
+      "step": 15840
+    },
+    {
+      "epoch": 0.80813745984806,
+      "grad_norm": 3.3114878534081633,
+      "learning_rate": 9.352055474853577e-06,
+      "loss": 0.8199,
+      "step": 15850
+    },
+    {
+      "epoch": 0.8086473257533269,
+      "grad_norm": 2.3822641673943794,
+      "learning_rate": 9.30402913979373e-06,
+      "loss": 0.8839,
+      "step": 15860
+    },
+    {
+      "epoch": 0.8091571916585938,
+      "grad_norm": 2.6897194947382714,
+      "learning_rate": 9.256113785414983e-06,
+      "loss": 0.8619,
+      "step": 15870
+    },
+    {
+      "epoch": 0.8096670575638607,
+      "grad_norm": 2.9736849566182273,
+      "learning_rate": 9.208309542385763e-06,
+      "loss": 0.848,
+      "step": 15880
+    },
+    {
+      "epoch": 0.8101769234691276,
+      "grad_norm": 3.591038409239247,
+      "learning_rate": 9.160616541071498e-06,
+      "loss": 0.909,
+      "step": 15890
+    },
+    {
+      "epoch": 0.8106867893743945,
+      "grad_norm": 4.808153735781551,
+      "learning_rate": 9.113034911534252e-06,
+      "loss": 0.9125,
+      "step": 15900
+    },
+    {
+      "epoch": 0.8111966552796614,
+      "grad_norm": 2.7296413543367017,
+      "learning_rate": 9.065564783532337e-06,
+      "loss": 0.918,
+      "step": 15910
+    },
+    {
+      "epoch": 0.8117065211849284,
+      "grad_norm": 3.536981412579703,
+      "learning_rate": 9.018206286520076e-06,
+      "loss": 0.8349,
+      "step": 15920
+    },
+    {
+      "epoch": 0.8122163870901953,
+      "grad_norm": 3.1011615339063376,
+      "learning_rate": 8.970959549647256e-06,
+      "loss": 0.8511,
+      "step": 15930
+    },
+    {
+      "epoch": 0.8127262529954622,
+      "grad_norm": 3.252409169101741,
+      "learning_rate": 8.923824701758977e-06,
+      "loss": 0.8847,
+      "step": 15940
+    },
+    {
+      "epoch": 0.8132361189007291,
+      "grad_norm": 4.455089143947698,
+      "learning_rate": 8.876801871395162e-06,
+      "loss": 0.9451,
+      "step": 15950
+    },
+    {
+      "epoch": 0.813745984805996,
+      "grad_norm": 2.4217448895990086,
+      "learning_rate": 8.829891186790279e-06,
+      "loss": 0.8865,
+      "step": 15960
+    },
+    {
+      "epoch": 0.8142558507112629,
+      "grad_norm": 3.589486984451676,
+      "learning_rate": 8.783092775872947e-06,
+      "loss": 0.8133,
+      "step": 15970
+    },
+    {
+      "epoch": 0.8147657166165299,
+      "grad_norm": 2.7211659623212117,
+      "learning_rate": 8.736406766265604e-06,
+      "loss": 0.8709,
+      "step": 15980
+    },
+    {
+      "epoch": 0.8152755825217968,
+      "grad_norm": 2.761187608907662,
+      "learning_rate": 8.689833285284215e-06,
+      "loss": 0.7868,
+      "step": 15990
+    },
+    {
+      "epoch": 0.8157854484270637,
+      "grad_norm": 3.1331219910466808,
+      "learning_rate": 8.643372459937782e-06,
+      "loss": 0.8954,
+      "step": 16000
+    },
+    {
+      "epoch": 0.8162953143323306,
+      "grad_norm": 3.2231351382034776,
+      "learning_rate": 8.597024416928185e-06,
+      "loss": 0.8256,
+      "step": 16010
+    },
+    {
+      "epoch": 0.8168051802375975,
+      "grad_norm": 2.470765388519924,
+      "learning_rate": 8.550789282649636e-06,
+      "loss": 0.8837,
+      "step": 16020
+    },
+    {
+      "epoch": 0.8173150461428644,
+      "grad_norm": 3.253600843830807,
+      "learning_rate": 8.504667183188535e-06,
+      "loss": 0.8608,
+      "step": 16030
+    },
+    {
+      "epoch": 0.8178249120481313,
+      "grad_norm": 3.9493017734822318,
+      "learning_rate": 8.45865824432297e-06,
+      "loss": 0.8991,
+      "step": 16040
+    },
+    {
+      "epoch": 0.8183347779533983,
+      "grad_norm": 3.8479711856362417,
+      "learning_rate": 8.41276259152245e-06,
+      "loss": 0.8186,
+      "step": 16050
+    },
+    {
+      "epoch": 0.8188446438586652,
+      "grad_norm": 2.4952764775374447,
+      "learning_rate": 8.366980349947534e-06,
+      "loss": 0.849,
+      "step": 16060
+    },
+    {
+      "epoch": 0.8193545097639321,
+      "grad_norm": 2.318077856843593,
+      "learning_rate": 8.321311644449508e-06,
+      "loss": 0.958,
+      "step": 16070
+    },
+    {
+      "epoch": 0.819864375669199,
+      "grad_norm": 2.4305126243307034,
+      "learning_rate": 8.275756599570062e-06,
+      "loss": 0.8906,
+      "step": 16080
+    },
+    {
+      "epoch": 0.8203742415744659,
+      "grad_norm": 3.4006906741716794,
+      "learning_rate": 8.230315339540878e-06,
+      "loss": 0.8523,
+      "step": 16090
+    },
+    {
+      "epoch": 0.8208841074797328,
+      "grad_norm": 3.9703650166653026,
+      "learning_rate": 8.184987988283377e-06,
+      "loss": 0.9145,
+      "step": 16100
+    },
+    {
+      "epoch": 0.8213939733849998,
+      "grad_norm": 3.6893958465677064,
+      "learning_rate": 8.139774669408329e-06,
+      "loss": 0.8816,
+      "step": 16110
+    },
+    {
+      "epoch": 0.8219038392902667,
+      "grad_norm": 2.61513028940626,
+      "learning_rate": 8.094675506215521e-06,
+      "loss": 0.7898,
+      "step": 16120
+    },
+    {
+      "epoch": 0.8224137051955336,
+      "grad_norm": 2.57873440427701,
+      "learning_rate": 8.049690621693445e-06,
+      "loss": 0.8591,
+      "step": 16130
+    },
+    {
+      "epoch": 0.8229235711008005,
+      "grad_norm": 3.343367401705895,
+      "learning_rate": 8.004820138518936e-06,
+      "loss": 0.8698,
+      "step": 16140
+    },
+    {
+      "epoch": 0.8234334370060674,
+      "grad_norm": 2.6164216953992203,
+      "learning_rate": 7.960064179056853e-06,
+      "loss": 0.8873,
+      "step": 16150
+    },
+    {
+      "epoch": 0.8239433029113343,
+      "grad_norm": 5.099384224163187,
+      "learning_rate": 7.915422865359761e-06,
+      "loss": 0.9454,
+      "step": 16160
+    },
+    {
+      "epoch": 0.8244531688166012,
+      "grad_norm": 2.708107590289228,
+      "learning_rate": 7.870896319167548e-06,
+      "loss": 0.8017,
+      "step": 16170
+    },
+    {
+      "epoch": 0.8249630347218682,
+      "grad_norm": 2.7594442167712034,
+      "learning_rate": 7.82648466190713e-06,
+      "loss": 0.9691,
+      "step": 16180
+    },
+    {
+      "epoch": 0.8254729006271351,
+      "grad_norm": 3.405779799598549,
+      "learning_rate": 7.78218801469212e-06,
+      "loss": 0.9282,
+      "step": 16190
+    },
+    {
+      "epoch": 0.825982766532402,
+      "grad_norm": 2.6278792255712617,
+      "learning_rate": 7.738006498322476e-06,
+      "loss": 0.8982,
+      "step": 16200
+    },
+    {
+      "epoch": 0.8264926324376689,
+      "grad_norm": 2.217396565318523,
+      "learning_rate": 7.693940233284241e-06,
+      "loss": 0.8412,
+      "step": 16210
+    },
+    {
+      "epoch": 0.8270024983429358,
+      "grad_norm": 5.111239654988263,
+      "learning_rate": 7.649989339749063e-06,
+      "loss": 0.8672,
+      "step": 16220
+    },
+    {
+      "epoch": 0.8275123642482027,
+      "grad_norm": 2.8779222319150386,
+      "learning_rate": 7.6061539375740675e-06,
+      "loss": 0.8975,
+      "step": 16230
+    },
+    {
+      "epoch": 0.8280222301534697,
+      "grad_norm": 3.616875433509951,
+      "learning_rate": 7.562434146301328e-06,
+      "loss": 0.8724,
+      "step": 16240
+    },
+    {
+      "epoch": 0.8285320960587366,
+      "grad_norm": 3.722391811343343,
+      "learning_rate": 7.518830085157735e-06,
+      "loss": 0.8684,
+      "step": 16250
+    },
+    {
+      "epoch": 0.8290419619640035,
+      "grad_norm": 4.7903255788565815,
+      "learning_rate": 7.475341873054509e-06,
+      "loss": 0.8418,
+      "step": 16260
+    },
+    {
+      "epoch": 0.8295518278692704,
+      "grad_norm": 4.933738813144349,
+      "learning_rate": 7.4319696285869656e-06,
+      "loss": 0.8415,
+      "step": 16270
+    },
+    {
+      "epoch": 0.8300616937745373,
+      "grad_norm": 3.3060104127437553,
+      "learning_rate": 7.38871347003417e-06,
+      "loss": 0.8067,
+      "step": 16280
+    },
+    {
+      "epoch": 0.8305715596798042,
+      "grad_norm": 3.0203188021559413,
+      "learning_rate": 7.345573515358589e-06,
+      "loss": 0.8202,
+      "step": 16290
+    },
+    {
+      "epoch": 0.8310814255850711,
+      "grad_norm": 4.845751863789938,
+      "learning_rate": 7.3025498822058555e-06,
+      "loss": 0.891,
+      "step": 16300
+    },
+    {
+      "epoch": 0.8315912914903381,
+      "grad_norm": 3.7309637791842105,
+      "learning_rate": 7.259642687904294e-06,
+      "loss": 0.8687,
+      "step": 16310
+    },
+    {
+      "epoch": 0.832101157395605,
+      "grad_norm": 2.502869101004249,
+      "learning_rate": 7.216852049464789e-06,
+      "loss": 0.8821,
+      "step": 16320
+    },
+    {
+      "epoch": 0.8326110233008719,
+      "grad_norm": 3.8998120732085537,
+      "learning_rate": 7.1741780835802795e-06,
+      "loss": 0.9047,
+      "step": 16330
+    },
+    {
+      "epoch": 0.8331208892061388,
+      "grad_norm": 2.594922394385684,
+      "learning_rate": 7.131620906625608e-06,
+      "loss": 0.8632,
+      "step": 16340
+    },
+    {
+      "epoch": 0.8336307551114057,
+      "grad_norm": 4.10832131941104,
+      "learning_rate": 7.089180634657072e-06,
+      "loss": 0.8188,
+      "step": 16350
+    },
+    {
+      "epoch": 0.8341406210166726,
+      "grad_norm": 3.4778953314658176,
+      "learning_rate": 7.046857383412192e-06,
+      "loss": 0.9106,
+      "step": 16360
+    },
+    {
+      "epoch": 0.8346504869219395,
+      "grad_norm": 3.4375785107246184,
+      "learning_rate": 7.004651268309348e-06,
+      "loss": 0.8377,
+      "step": 16370
+    },
+    {
+      "epoch": 0.8351603528272065,
+      "grad_norm": 4.5960096164216715,
+      "learning_rate": 6.96256240444747e-06,
+      "loss": 0.8717,
+      "step": 16380
+    },
+    {
+      "epoch": 0.8356702187324734,
+      "grad_norm": 2.867417877276137,
+      "learning_rate": 6.92059090660579e-06,
+      "loss": 0.8121,
+      "step": 16390
+    },
+    {
+      "epoch": 0.8361800846377403,
+      "grad_norm": 3.529230966679883,
+      "learning_rate": 6.878736889243409e-06,
+      "loss": 0.9319,
+      "step": 16400
+    },
+    {
+      "epoch": 0.8366899505430072,
+      "grad_norm": 2.977687049441769,
+      "learning_rate": 6.8370004664990936e-06,
+      "loss": 0.8468,
+      "step": 16410
+    },
+    {
+      "epoch": 0.8371998164482741,
+      "grad_norm": 2.8806625721464036,
+      "learning_rate": 6.795381752190904e-06,
+      "loss": 0.8855,
+      "step": 16420
+    },
+    {
+      "epoch": 0.837709682353541,
+      "grad_norm": 3.0713559435415982,
+      "learning_rate": 6.753880859815903e-06,
+      "loss": 0.8187,
+      "step": 16430
+    },
+    {
+      "epoch": 0.838219548258808,
+      "grad_norm": 4.83372555516333,
+      "learning_rate": 6.712497902549847e-06,
+      "loss": 0.7915,
+      "step": 16440
+    },
+    {
+      "epoch": 0.8387294141640749,
+      "grad_norm": 4.6112192242332375,
+      "learning_rate": 6.671232993246879e-06,
+      "loss": 0.9248,
+      "step": 16450
+    },
+    {
+      "epoch": 0.8392392800693418,
+      "grad_norm": 5.174748909638641,
+      "learning_rate": 6.630086244439204e-06,
+      "loss": 0.8935,
+      "step": 16460
+    },
+    {
+      "epoch": 0.8397491459746087,
+      "grad_norm": 2.8202028698722277,
+      "learning_rate": 6.58905776833682e-06,
+      "loss": 0.8285,
+      "step": 16470
+    },
+    {
+      "epoch": 0.8402590118798756,
+      "grad_norm": 2.6916122398911284,
+      "learning_rate": 6.548147676827171e-06,
+      "loss": 0.9034,
+      "step": 16480
+    },
+    {
+      "epoch": 0.8407688777851425,
+      "grad_norm": 4.025784568272635,
+      "learning_rate": 6.5073560814748526e-06,
+      "loss": 0.8402,
+      "step": 16490
+    },
+    {
+      "epoch": 0.8412787436904094,
+      "grad_norm": 3.831845314955286,
+      "learning_rate": 6.46668309352132e-06,
+      "loss": 0.8344,
+      "step": 16500
+    },
+    {
+      "epoch": 0.8417886095956764,
+      "grad_norm": 3.077220271058762,
+      "learning_rate": 6.426128823884575e-06,
+      "loss": 0.871,
+      "step": 16510
+    },
+    {
+      "epoch": 0.8422984755009433,
+      "grad_norm": 2.982336543536334,
+      "learning_rate": 6.385693383158897e-06,
+      "loss": 0.8408,
+      "step": 16520
+    },
+    {
+      "epoch": 0.8428083414062102,
+      "grad_norm": 2.760870638916319,
+      "learning_rate": 6.34537688161444e-06,
+      "loss": 0.8809,
+      "step": 16530
+    },
+    {
+      "epoch": 0.8433182073114771,
+      "grad_norm": 4.211621126833601,
+      "learning_rate": 6.3051794291970944e-06,
+      "loss": 0.8935,
+      "step": 16540
+    },
+    {
+      "epoch": 0.843828073216744,
+      "grad_norm": 3.531257913443177,
+      "learning_rate": 6.265101135527995e-06,
+      "loss": 0.9046,
+      "step": 16550
+    },
+    {
+      "epoch": 0.8443379391220109,
+      "grad_norm": 3.5338096194632524,
+      "learning_rate": 6.225142109903426e-06,
+      "loss": 0.8521,
+      "step": 16560
+    },
+    {
+      "epoch": 0.8448478050272779,
+      "grad_norm": 5.067856498419463,
+      "learning_rate": 6.185302461294323e-06,
+      "loss": 0.9061,
+      "step": 16570
+    },
+    {
+      "epoch": 0.8453576709325448,
+      "grad_norm": 3.329892039218186,
+      "learning_rate": 6.145582298346153e-06,
+      "loss": 0.875,
+      "step": 16580
+    },
+    {
+      "epoch": 0.8458675368378117,
+      "grad_norm": 2.274387432212045,
+      "learning_rate": 6.1059817293784905e-06,
+      "loss": 0.9349,
+      "step": 16590
+    },
+    {
+      "epoch": 0.8463774027430786,
+      "grad_norm": 3.9161422877800294,
+      "learning_rate": 6.066500862384772e-06,
+      "loss": 0.807,
+      "step": 16600
+    },
+    {
+      "epoch": 0.8468872686483455,
+      "grad_norm": 4.724170867775978,
+      "learning_rate": 6.027139805032034e-06,
+      "loss": 0.8691,
+      "step": 16610
+    },
+    {
+      "epoch": 0.8473971345536124,
+      "grad_norm": 3.441993677446614,
+      "learning_rate": 5.987898664660518e-06,
+      "loss": 0.8585,
+      "step": 16620
+    },
+    {
+      "epoch": 0.8479070004588793,
+      "grad_norm": 2.7631551181795846,
+      "learning_rate": 5.948777548283518e-06,
+      "loss": 0.9392,
+      "step": 16630
+    },
+    {
+      "epoch": 0.8484168663641463,
+      "grad_norm": 4.39445720589836,
+      "learning_rate": 5.909776562586966e-06,
+      "loss": 0.8741,
+      "step": 16640
+    },
+    {
+      "epoch": 0.8489267322694132,
+      "grad_norm": 4.127833421024382,
+      "learning_rate": 5.8708958139292e-06,
+      "loss": 0.8015,
+      "step": 16650
+    },
+    {
+      "epoch": 0.8494365981746801,
+      "grad_norm": 2.8820499475149757,
+      "learning_rate": 5.832135408340678e-06,
+      "loss": 0.8801,
+      "step": 16660
+    },
+    {
+      "epoch": 0.849946464079947,
+      "grad_norm": 1.9926941179839137,
+      "learning_rate": 5.79349545152365e-06,
+      "loss": 0.844,
+      "step": 16670
+    },
+    {
+      "epoch": 0.8504563299852139,
+      "grad_norm": 2.4591920309937665,
+      "learning_rate": 5.754976048851918e-06,
+      "loss": 0.8819,
+      "step": 16680
+    },
+    {
+      "epoch": 0.8509661958904808,
+      "grad_norm": 2.2649787820447678,
+      "learning_rate": 5.716577305370496e-06,
+      "loss": 0.846,
+      "step": 16690
+    },
+    {
+      "epoch": 0.8514760617957478,
+      "grad_norm": 4.140766093918557,
+      "learning_rate": 5.678299325795389e-06,
+      "loss": 0.7989,
+      "step": 16700
+    },
+    {
+      "epoch": 0.8519859277010147,
+      "grad_norm": 3.1458346286813548,
+      "learning_rate": 5.64014221451325e-06,
+      "loss": 0.8247,
+      "step": 16710
+    },
+    {
+      "epoch": 0.8524957936062816,
+      "grad_norm": 3.874402034680883,
+      "learning_rate": 5.60210607558111e-06,
+      "loss": 0.937,
+      "step": 16720
+    },
+    {
+      "epoch": 0.8530056595115485,
+      "grad_norm": 6.516766826574404,
+      "learning_rate": 5.5641910127261055e-06,
+      "loss": 0.8999,
+      "step": 16730
+    },
+    {
+      "epoch": 0.8535155254168154,
+      "grad_norm": 3.2136111758024235,
+      "learning_rate": 5.526397129345201e-06,
+      "loss": 0.8773,
+      "step": 16740
+    },
+    {
+      "epoch": 0.8540253913220823,
+      "grad_norm": 3.4995559414420847,
+      "learning_rate": 5.488724528504869e-06,
+      "loss": 0.8279,
+      "step": 16750
+    },
+    {
+      "epoch": 0.8545352572273492,
+      "grad_norm": 4.12837962380521,
+      "learning_rate": 5.451173312940888e-06,
+      "loss": 0.8974,
+      "step": 16760
+    },
+    {
+      "epoch": 0.8550451231326162,
+      "grad_norm": 2.4110487720494747,
+      "learning_rate": 5.41374358505794e-06,
+      "loss": 0.8523,
+      "step": 16770
+    },
+    {
+      "epoch": 0.8555549890378831,
+      "grad_norm": 2.6668339141232784,
+      "learning_rate": 5.376435446929479e-06,
+      "loss": 0.8172,
+      "step": 16780
+    },
+    {
+      "epoch": 0.85606485494315,
+      "grad_norm": 2.8935183751225986,
+      "learning_rate": 5.339249000297292e-06,
+      "loss": 0.8975,
+      "step": 16790
+    },
+    {
+      "epoch": 0.8565747208484169,
+      "grad_norm": 3.754043446253633,
+      "learning_rate": 5.302184346571382e-06,
+      "loss": 0.9335,
+      "step": 16800
+    },
+    {
+      "epoch": 0.8570845867536838,
+      "grad_norm": 2.7840078796206638,
+      "learning_rate": 5.265241586829567e-06,
+      "loss": 0.8862,
+      "step": 16810
+    },
+    {
+      "epoch": 0.8575944526589507,
+      "grad_norm": 2.990380757866614,
+      "learning_rate": 5.228420821817265e-06,
+      "loss": 0.8463,
+      "step": 16820
+    },
+    {
+      "epoch": 0.8581043185642176,
+      "grad_norm": 2.760328530084163,
+      "learning_rate": 5.191722151947226e-06,
+      "loss": 0.8766,
+      "step": 16830
+    },
+    {
+      "epoch": 0.8586141844694846,
+      "grad_norm": 2.8718300531093033,
+      "learning_rate": 5.1551456772991835e-06,
+      "loss": 0.8492,
+      "step": 16840
+    },
+    {
+      "epoch": 0.8591240503747515,
+      "grad_norm": 4.9401918610140925,
+      "learning_rate": 5.118691497619715e-06,
+      "loss": 0.872,
+      "step": 16850
+    },
+    {
+      "epoch": 0.8596339162800184,
+      "grad_norm": 2.9409033090841064,
+      "learning_rate": 5.08235971232181e-06,
+      "loss": 0.848,
+      "step": 16860
+    },
+    {
+      "epoch": 0.8601437821852853,
+      "grad_norm": 3.14064409023017,
+      "learning_rate": 5.046150420484752e-06,
+      "loss": 0.8149,
+      "step": 16870
+    },
+    {
+      "epoch": 0.8606536480905522,
+      "grad_norm": 2.696613971341707,
+      "learning_rate": 5.010063720853709e-06,
+      "loss": 0.7826,
+      "step": 16880
+    },
+    {
+      "epoch": 0.8611635139958191,
+      "grad_norm": 2.773127692319091,
+      "learning_rate": 4.974099711839592e-06,
+      "loss": 0.8794,
+      "step": 16890
+    },
+    {
+      "epoch": 0.8616733799010861,
+      "grad_norm": 6.8747030079738325,
+      "learning_rate": 4.9382584915186934e-06,
+      "loss": 0.8331,
+      "step": 16900
+    },
+    {
+      "epoch": 0.862183245806353,
+      "grad_norm": 3.695685522208731,
+      "learning_rate": 4.9025401576324456e-06,
+      "loss": 0.8874,
+      "step": 16910
+    },
+    {
+      "epoch": 0.8626931117116199,
+      "grad_norm": 3.260766796388533,
+      "learning_rate": 4.866944807587209e-06,
+      "loss": 0.8527,
+      "step": 16920
+    },
+    {
+      "epoch": 0.8632029776168868,
+      "grad_norm": 3.759808703610138,
+      "learning_rate": 4.83147253845388e-06,
+      "loss": 0.8405,
+      "step": 16930
+    },
+    {
+      "epoch": 0.8637128435221537,
+      "grad_norm": 3.003698027126189,
+      "learning_rate": 4.796123446967787e-06,
+      "loss": 0.8812,
+      "step": 16940
+    },
+    {
+      "epoch": 0.8642227094274206,
+      "grad_norm": 3.121755258754485,
+      "learning_rate": 4.760897629528288e-06,
+      "loss": 0.7932,
+      "step": 16950
+    },
+    {
+      "epoch": 0.8647325753326875,
+      "grad_norm": 3.27344701554323,
+      "learning_rate": 4.725795182198589e-06,
+      "loss": 0.8311,
+      "step": 16960
+    },
+    {
+      "epoch": 0.8652424412379545,
+      "grad_norm": 3.149376629291201,
+      "learning_rate": 4.690816200705456e-06,
+      "loss": 0.8213,
+      "step": 16970
+    },
+    {
+      "epoch": 0.8657523071432214,
+      "grad_norm": 2.804037331505134,
+      "learning_rate": 4.65596078043894e-06,
+      "loss": 0.9377,
+      "step": 16980
+    },
+    {
+      "epoch": 0.8662621730484883,
+      "grad_norm": 2.810177137052245,
+      "learning_rate": 4.621229016452156e-06,
+      "loss": 0.838,
+      "step": 16990
+    },
+    {
+      "epoch": 0.8667720389537552,
+      "grad_norm": 3.5617441410460717,
+      "learning_rate": 4.586621003460978e-06,
+      "loss": 0.9192,
+      "step": 17000
+    },
+    {
+      "epoch": 0.8672819048590221,
+      "grad_norm": 3.1535985280516523,
+      "learning_rate": 4.552136835843801e-06,
+      "loss": 0.7998,
+      "step": 17010
+    },
+    {
+      "epoch": 0.867791770764289,
+      "grad_norm": 2.911092583241028,
+      "learning_rate": 4.517776607641327e-06,
+      "loss": 0.8529,
+      "step": 17020
+    },
+    {
+      "epoch": 0.868301636669556,
+      "grad_norm": 2.6097489312022586,
+      "learning_rate": 4.483540412556214e-06,
+      "loss": 0.8552,
+      "step": 17030
+    },
+    {
+      "epoch": 0.8688115025748229,
+      "grad_norm": 2.825900733733966,
+      "learning_rate": 4.4494283439528995e-06,
+      "loss": 0.8381,
+      "step": 17040
+    },
+    {
+      "epoch": 0.8693213684800898,
+      "grad_norm": 4.707749869450497,
+      "learning_rate": 4.41544049485732e-06,
+      "loss": 0.881,
+      "step": 17050
+    },
+    {
+      "epoch": 0.8698312343853567,
+      "grad_norm": 3.033282832876074,
+      "learning_rate": 4.381576957956635e-06,
+      "loss": 0.8764,
+      "step": 17060
+    },
+    {
+      "epoch": 0.8703411002906236,
+      "grad_norm": 2.4140441305510634,
+      "learning_rate": 4.347837825599044e-06,
+      "loss": 0.8526,
+      "step": 17070
+    },
+    {
+      "epoch": 0.8708509661958905,
+      "grad_norm": 2.703898562677979,
+      "learning_rate": 4.31422318979342e-06,
+      "loss": 0.9,
+      "step": 17080
+    },
+    {
+      "epoch": 0.8713608321011574,
+      "grad_norm": 2.648912615174052,
+      "learning_rate": 4.280733142209198e-06,
+      "loss": 0.8763,
+      "step": 17090
+    },
+    {
+      "epoch": 0.8718706980064244,
+      "grad_norm": 2.64571381505409,
+      "learning_rate": 4.247367774175981e-06,
+      "loss": 0.7991,
+      "step": 17100
+    },
+    {
+      "epoch": 0.8723805639116913,
+      "grad_norm": 5.957919067488417,
+      "learning_rate": 4.214127176683425e-06,
+      "loss": 0.8429,
+      "step": 17110
+    },
+    {
+      "epoch": 0.8728904298169582,
+      "grad_norm": 3.0097268354940585,
+      "learning_rate": 4.181011440380889e-06,
+      "loss": 0.8668,
+      "step": 17120
+    },
+    {
+      "epoch": 0.8734002957222251,
+      "grad_norm": 3.0171216108212615,
+      "learning_rate": 4.1480206555772335e-06,
+      "loss": 0.8954,
+      "step": 17130
+    },
+    {
+      "epoch": 0.873910161627492,
+      "grad_norm": 3.2224816444578805,
+      "learning_rate": 4.115154912240593e-06,
+      "loss": 0.8545,
+      "step": 17140
+    },
+    {
+      "epoch": 0.8744200275327589,
+      "grad_norm": 2.858710303956987,
+      "learning_rate": 4.082414299998061e-06,
+      "loss": 0.7906,
+      "step": 17150
+    },
+    {
+      "epoch": 0.8749298934380259,
+      "grad_norm": 3.1304540062971036,
+      "learning_rate": 4.049798908135538e-06,
+      "loss": 0.8259,
+      "step": 17160
+    },
+    {
+      "epoch": 0.8754397593432928,
+      "grad_norm": 3.201828897583765,
+      "learning_rate": 4.0173088255973876e-06,
+      "loss": 0.8815,
+      "step": 17170
+    },
+    {
+      "epoch": 0.8759496252485597,
+      "grad_norm": 3.90605922682777,
+      "learning_rate": 3.984944140986302e-06,
+      "loss": 0.8411,
+      "step": 17180
+    },
+    {
+      "epoch": 0.8764594911538266,
+      "grad_norm": 2.909015114801005,
+      "learning_rate": 3.9527049425629625e-06,
+      "loss": 0.8251,
+      "step": 17190
+    },
+    {
+      "epoch": 0.8769693570590935,
+      "grad_norm": 4.4745654254538945,
+      "learning_rate": 3.920591318245864e-06,
+      "loss": 0.8814,
+      "step": 17200
+    },
+    {
+      "epoch": 0.8774792229643604,
+      "grad_norm": 3.679638801253417,
+      "learning_rate": 3.888603355611043e-06,
+      "loss": 0.8712,
+      "step": 17210
+    },
+    {
+      "epoch": 0.8779890888696273,
+      "grad_norm": 3.148123725056795,
+      "learning_rate": 3.856741141891851e-06,
+      "loss": 0.8884,
+      "step": 17220
+    },
+    {
+      "epoch": 0.8784989547748941,
+      "grad_norm": 3.0572988475286023,
+      "learning_rate": 3.825004763978707e-06,
+      "loss": 0.8611,
+      "step": 17230
+    },
+    {
+      "epoch": 0.8790088206801611,
+      "grad_norm": 2.8463176761925735,
+      "learning_rate": 3.7933943084188685e-06,
+      "loss": 0.8491,
+      "step": 17240
+    },
+    {
+      "epoch": 0.879518686585428,
+      "grad_norm": 3.484277901488632,
+      "learning_rate": 3.761909861416213e-06,
+      "loss": 0.8225,
+      "step": 17250
+    },
+    {
+      "epoch": 0.8800285524906949,
+      "grad_norm": 2.3843857621790083,
+      "learning_rate": 3.730551508830965e-06,
+      "loss": 0.8028,
+      "step": 17260
+    },
+    {
+      "epoch": 0.8805384183959618,
+      "grad_norm": 3.927491445853449,
+      "learning_rate": 3.6993193361794797e-06,
+      "loss": 0.8858,
+      "step": 17270
+    },
+    {
+      "epoch": 0.8810482843012287,
+      "grad_norm": 2.4728935704745054,
+      "learning_rate": 3.668213428634021e-06,
+      "loss": 0.8672,
+      "step": 17280
+    },
+    {
+      "epoch": 0.8815581502064956,
+      "grad_norm": 3.4701561651798274,
+      "learning_rate": 3.6372338710225106e-06,
+      "loss": 0.8661,
+      "step": 17290
+    },
+    {
+      "epoch": 0.8820680161117626,
+      "grad_norm": 2.7708196871940167,
+      "learning_rate": 3.6063807478283085e-06,
+      "loss": 0.8383,
+      "step": 17300
+    },
+    {
+      "epoch": 0.8825778820170295,
+      "grad_norm": 2.121142389184302,
+      "learning_rate": 3.5756541431899994e-06,
+      "loss": 0.8571,
+      "step": 17310
+    },
+    {
+      "epoch": 0.8830877479222964,
+      "grad_norm": 2.679428452061065,
+      "learning_rate": 3.545054140901094e-06,
+      "loss": 0.8894,
+      "step": 17320
+    },
+    {
+      "epoch": 0.8835976138275633,
+      "grad_norm": 3.627841196640346,
+      "learning_rate": 3.514580824409902e-06,
+      "loss": 0.8473,
+      "step": 17330
+    },
+    {
+      "epoch": 0.8841074797328302,
+      "grad_norm": 2.461704176305592,
+      "learning_rate": 3.484234276819226e-06,
+      "loss": 0.8029,
+      "step": 17340
+    },
+    {
+      "epoch": 0.8846173456380971,
+      "grad_norm": 2.145063491566008,
+      "learning_rate": 3.4540145808861615e-06,
+      "loss": 0.7239,
+      "step": 17350
+    },
+    {
+      "epoch": 0.885127211543364,
+      "grad_norm": 3.181073473816311,
+      "learning_rate": 3.42392181902188e-06,
+      "loss": 0.8498,
+      "step": 17360
+    },
+    {
+      "epoch": 0.885637077448631,
+      "grad_norm": 2.0408458344074734,
+      "learning_rate": 3.3939560732913766e-06,
+      "loss": 0.8554,
+      "step": 17370
+    },
+    {
+      "epoch": 0.8861469433538979,
+      "grad_norm": 2.5692246053049166,
+      "learning_rate": 3.3641174254133103e-06,
+      "loss": 0.8444,
+      "step": 17380
+    },
+    {
+      "epoch": 0.8866568092591648,
+      "grad_norm": 2.491133156069609,
+      "learning_rate": 3.3344059567596676e-06,
+      "loss": 0.8227,
+      "step": 17390
+    },
+    {
+      "epoch": 0.8871666751644317,
+      "grad_norm": 3.1954460705488983,
+      "learning_rate": 3.3048217483556744e-06,
+      "loss": 0.8489,
+      "step": 17400
+    },
+    {
+      "epoch": 0.8876765410696986,
+      "grad_norm": 3.5272710375611718,
+      "learning_rate": 3.2753648808794503e-06,
+      "loss": 0.8596,
+      "step": 17410
+    },
+    {
+      "epoch": 0.8881864069749655,
+      "grad_norm": 4.225068478639933,
+      "learning_rate": 3.2460354346619037e-06,
+      "loss": 0.857,
+      "step": 17420
+    },
+    {
+      "epoch": 0.8886962728802325,
+      "grad_norm": 2.8309100959825795,
+      "learning_rate": 3.216833489686416e-06,
+      "loss": 0.888,
+      "step": 17430
+    },
+    {
+      "epoch": 0.8892061387854994,
+      "grad_norm": 3.3137678401694326,
+      "learning_rate": 3.1877591255886795e-06,
+      "loss": 0.8399,
+      "step": 17440
+    },
+    {
+      "epoch": 0.8897160046907663,
+      "grad_norm": 4.637502639723594,
+      "learning_rate": 3.158812421656465e-06,
+      "loss": 0.8119,
+      "step": 17450
+    },
+    {
+      "epoch": 0.8902258705960332,
+      "grad_norm": 2.5633260057758336,
+      "learning_rate": 3.1299934568293944e-06,
+      "loss": 0.8342,
+      "step": 17460
+    },
+    {
+      "epoch": 0.8907357365013001,
+      "grad_norm": 3.1154186287560752,
+      "learning_rate": 3.1013023096987726e-06,
+      "loss": 0.8631,
+      "step": 17470
+    },
+    {
+      "epoch": 0.891245602406567,
+      "grad_norm": 5.142330007176737,
+      "learning_rate": 3.072739058507268e-06,
+      "loss": 0.8792,
+      "step": 17480
+    },
+    {
+      "epoch": 0.8917554683118339,
+      "grad_norm": 3.1723051722577713,
+      "learning_rate": 3.044303781148844e-06,
+      "loss": 0.9074,
+      "step": 17490
+    },
+    {
+      "epoch": 0.8922653342171009,
+      "grad_norm": 1.9982237903220441,
+      "learning_rate": 3.0159965551684265e-06,
+      "loss": 0.8301,
+      "step": 17500
+    },
+    {
+      "epoch": 0.8927752001223678,
+      "grad_norm": 3.126696697645224,
+      "learning_rate": 2.9878174577617436e-06,
+      "loss": 0.8203,
+      "step": 17510
+    },
+    {
+      "epoch": 0.8932850660276347,
+      "grad_norm": 3.266740995165527,
+      "learning_rate": 2.9597665657751073e-06,
+      "loss": 0.8285,
+      "step": 17520
+    },
+    {
+      "epoch": 0.8937949319329016,
+      "grad_norm": 4.6025351798463845,
+      "learning_rate": 2.931843955705216e-06,
+      "loss": 0.8886,
+      "step": 17530
+    },
+    {
+      "epoch": 0.8943047978381685,
+      "grad_norm": 2.5723945485729836,
+      "learning_rate": 2.9040497036989246e-06,
+      "loss": 0.8767,
+      "step": 17540
+    },
+    {
+      "epoch": 0.8948146637434354,
+      "grad_norm": 4.222542833788573,
+      "learning_rate": 2.8763838855530414e-06,
+      "loss": 0.8077,
+      "step": 17550
+    },
+    {
+      "epoch": 0.8953245296487023,
+      "grad_norm": 2.4980823944930743,
+      "learning_rate": 2.848846576714148e-06,
+      "loss": 0.8911,
+      "step": 17560
+    },
+    {
+      "epoch": 0.8958343955539693,
+      "grad_norm": 1.9783994401103546,
+      "learning_rate": 2.8214378522783523e-06,
+      "loss": 0.9195,
+      "step": 17570
+    },
+    {
+      "epoch": 0.8963442614592362,
+      "grad_norm": 3.172365700936146,
+      "learning_rate": 2.7941577869911084e-06,
+      "loss": 0.8592,
+      "step": 17580
+    },
+    {
+      "epoch": 0.8968541273645031,
+      "grad_norm": 2.8205375859859343,
+      "learning_rate": 2.7670064552470174e-06,
+      "loss": 0.8388,
+      "step": 17590
+    },
+    {
+      "epoch": 0.89736399326977,
+      "grad_norm": 4.238387100796564,
+      "learning_rate": 2.7399839310896015e-06,
+      "loss": 0.8409,
+      "step": 17600
+    },
+    {
+      "epoch": 0.8978738591750369,
+      "grad_norm": 3.1582031673238418,
+      "learning_rate": 2.7130902882111177e-06,
+      "loss": 0.8413,
+      "step": 17610
+    },
+    {
+      "epoch": 0.8983837250803038,
+      "grad_norm": 3.0814373048515025,
+      "learning_rate": 2.6863255999523774e-06,
+      "loss": 0.8555,
+      "step": 17620
+    },
+    {
+      "epoch": 0.8988935909855708,
+      "grad_norm": 3.509354259364666,
+      "learning_rate": 2.6596899393024845e-06,
+      "loss": 0.8623,
+      "step": 17630
+    },
+    {
+      "epoch": 0.8994034568908377,
+      "grad_norm": 2.429266270195368,
+      "learning_rate": 2.6331833788987126e-06,
+      "loss": 0.8142,
+      "step": 17640
+    },
+    {
+      "epoch": 0.8999133227961046,
+      "grad_norm": 3.3980722166350916,
+      "learning_rate": 2.606805991026251e-06,
+      "loss": 0.9006,
+      "step": 17650
+    },
+    {
+      "epoch": 0.9004231887013715,
+      "grad_norm": 4.482747939131566,
+      "learning_rate": 2.5805578476180312e-06,
+      "loss": 0.9059,
+      "step": 17660
+    },
+    {
+      "epoch": 0.9009330546066384,
+      "grad_norm": 4.9663211271909855,
+      "learning_rate": 2.554439020254529e-06,
+      "loss": 0.9362,
+      "step": 17670
+    },
+    {
+      "epoch": 0.9014429205119053,
+      "grad_norm": 4.02891398644143,
+      "learning_rate": 2.52844958016355e-06,
+      "loss": 0.8954,
+      "step": 17680
+    },
+    {
+      "epoch": 0.9019527864171722,
+      "grad_norm": 2.4905247597697993,
+      "learning_rate": 2.50258959822009e-06,
+      "loss": 0.8782,
+      "step": 17690
+    },
+    {
+      "epoch": 0.9024626523224392,
+      "grad_norm": 5.0959674052387385,
+      "learning_rate": 2.476859144946053e-06,
+      "loss": 0.8809,
+      "step": 17700
+    },
+    {
+      "epoch": 0.9029725182277061,
+      "grad_norm": 2.232474885119816,
+      "learning_rate": 2.4512582905101555e-06,
+      "loss": 0.8307,
+      "step": 17710
+    },
+    {
+      "epoch": 0.903482384132973,
+      "grad_norm": 3.202013982667936,
+      "learning_rate": 2.4257871047276504e-06,
+      "loss": 0.8072,
+      "step": 17720
+    },
+    {
+      "epoch": 0.9039922500382399,
+      "grad_norm": 2.8089674658641233,
+      "learning_rate": 2.400445657060213e-06,
+      "loss": 0.8563,
+      "step": 17730
+    },
+    {
+      "epoch": 0.9045021159435068,
+      "grad_norm": 3.9974832160033276,
+      "learning_rate": 2.3752340166156793e-06,
+      "loss": 0.8396,
+      "step": 17740
+    },
+    {
+      "epoch": 0.9050119818487737,
+      "grad_norm": 3.219586185849582,
+      "learning_rate": 2.350152252147919e-06,
+      "loss": 0.8687,
+      "step": 17750
+    },
+    {
+      "epoch": 0.9055218477540407,
+      "grad_norm": 2.990664067479295,
+      "learning_rate": 2.3252004320566124e-06,
+      "loss": 0.8481,
+      "step": 17760
+    },
+    {
+      "epoch": 0.9060317136593076,
+      "grad_norm": 4.047170027027108,
+      "learning_rate": 2.300378624387051e-06,
+      "loss": 0.8663,
+      "step": 17770
+    },
+    {
+      "epoch": 0.9065415795645745,
+      "grad_norm": 2.181344311357337,
+      "learning_rate": 2.2756868968300203e-06,
+      "loss": 0.8759,
+      "step": 17780
+    },
+    {
+      "epoch": 0.9070514454698414,
+      "grad_norm": 3.1345299664847044,
+      "learning_rate": 2.251125316721514e-06,
+      "loss": 0.8129,
+      "step": 17790
+    },
+    {
+      "epoch": 0.9075613113751083,
+      "grad_norm": 6.306032769349942,
+      "learning_rate": 2.226693951042652e-06,
+      "loss": 0.8502,
+      "step": 17800
+    },
+    {
+      "epoch": 0.9080711772803752,
+      "grad_norm": 2.514134962454054,
+      "learning_rate": 2.202392866419423e-06,
+      "loss": 0.801,
+      "step": 17810
+    },
+    {
+      "epoch": 0.9085810431856421,
+      "grad_norm": 2.272146093504471,
+      "learning_rate": 2.1782221291225367e-06,
+      "loss": 0.915,
+      "step": 17820
+    },
+    {
+      "epoch": 0.9090909090909091,
+      "grad_norm": 2.6031288042931595,
+      "learning_rate": 2.15418180506724e-06,
+      "loss": 0.8785,
+      "step": 17830
+    },
+    {
+      "epoch": 0.909600774996176,
+      "grad_norm": 4.832136636171037,
+      "learning_rate": 2.130271959813135e-06,
+      "loss": 0.8516,
+      "step": 17840
+    },
+    {
+      "epoch": 0.9101106409014429,
+      "grad_norm": 2.9257845681995605,
+      "learning_rate": 2.1064926585639854e-06,
+      "loss": 0.8414,
+      "step": 17850
+    },
+    {
+      "epoch": 0.9106205068067098,
+      "grad_norm": 2.800217543345524,
+      "learning_rate": 2.08284396616758e-06,
+      "loss": 0.8168,
+      "step": 17860
+    },
+    {
+      "epoch": 0.9111303727119767,
+      "grad_norm": 3.184719885995561,
+      "learning_rate": 2.0593259471155037e-06,
+      "loss": 0.8713,
+      "step": 17870
+    },
+    {
+      "epoch": 0.9116402386172436,
+      "grad_norm": 2.772795236218663,
+      "learning_rate": 2.035938665542991e-06,
+      "loss": 0.8966,
+      "step": 17880
+    },
+    {
+      "epoch": 0.9121501045225106,
+      "grad_norm": 3.447137143036548,
+      "learning_rate": 2.012682185228754e-06,
+      "loss": 0.9544,
+      "step": 17890
+    },
+    {
+      "epoch": 0.9126599704277775,
+      "grad_norm": 5.177406451744961,
+      "learning_rate": 1.9895565695947915e-06,
+      "loss": 0.9039,
+      "step": 17900
+    },
+    {
+      "epoch": 0.9131698363330444,
+      "grad_norm": 3.6842047845018624,
+      "learning_rate": 1.966561881706236e-06,
+      "loss": 0.8765,
+      "step": 17910
+    },
+    {
+      "epoch": 0.9136797022383113,
+      "grad_norm": 3.3134450967911846,
+      "learning_rate": 1.943698184271159e-06,
+      "loss": 0.8003,
+      "step": 17920
+    },
+    {
+      "epoch": 0.9141895681435782,
+      "grad_norm": 3.5691519411459467,
+      "learning_rate": 1.9209655396404348e-06,
+      "loss": 0.863,
+      "step": 17930
+    },
+    {
+      "epoch": 0.9146994340488451,
+      "grad_norm": 2.296992348640956,
+      "learning_rate": 1.8983640098075128e-06,
+      "loss": 0.8445,
+      "step": 17940
+    },
+    {
+      "epoch": 0.915209299954112,
+      "grad_norm": 3.111035863780307,
+      "learning_rate": 1.8758936564083118e-06,
+      "loss": 0.8159,
+      "step": 17950
+    },
+    {
+      "epoch": 0.915719165859379,
+      "grad_norm": 2.7404784101065705,
+      "learning_rate": 1.8535545407210141e-06,
+      "loss": 0.8563,
+      "step": 17960
+    },
+    {
+      "epoch": 0.9162290317646459,
+      "grad_norm": 5.396627788378159,
+      "learning_rate": 1.831346723665911e-06,
+      "loss": 0.8917,
+      "step": 17970
+    },
+    {
+      "epoch": 0.9167388976699128,
+      "grad_norm": 3.332001142997155,
+      "learning_rate": 1.809270265805224e-06,
+      "loss": 0.8533,
+      "step": 17980
+    },
+    {
+      "epoch": 0.9172487635751797,
+      "grad_norm": 4.0075701508468216,
+      "learning_rate": 1.7873252273429509e-06,
+      "loss": 0.8424,
+      "step": 17990
+    },
+    {
+      "epoch": 0.9177586294804466,
+      "grad_norm": 3.9804673466169542,
+      "learning_rate": 1.7655116681247197e-06,
+      "loss": 0.8345,
+      "step": 18000
+    },
+    {
+      "epoch": 0.9182684953857135,
+      "grad_norm": 5.6240808208381825,
+      "learning_rate": 1.7438296476375738e-06,
+      "loss": 0.8592,
+      "step": 18010
+    },
+    {
+      "epoch": 0.9187783612909804,
+      "grad_norm": 6.511416754863257,
+      "learning_rate": 1.7222792250098763e-06,
+      "loss": 0.8934,
+      "step": 18020
+    },
+    {
+      "epoch": 0.9192882271962474,
+      "grad_norm": 3.818889946851894,
+      "learning_rate": 1.7008604590110776e-06,
+      "loss": 0.8744,
+      "step": 18030
+    },
+    {
+      "epoch": 0.9197980931015143,
+      "grad_norm": 4.978118562625245,
+      "learning_rate": 1.6795734080516212e-06,
+      "loss": 0.8693,
+      "step": 18040
+    },
+    {
+      "epoch": 0.9203079590067812,
+      "grad_norm": 2.9894812066642387,
+      "learning_rate": 1.658418130182743e-06,
+      "loss": 0.8317,
+      "step": 18050
+    },
+    {
+      "epoch": 0.9208178249120481,
+      "grad_norm": 2.9154950321529007,
+      "learning_rate": 1.637394683096316e-06,
+      "loss": 0.9224,
+      "step": 18060
+    },
+    {
+      "epoch": 0.921327690817315,
+      "grad_norm": 2.355938059666778,
+      "learning_rate": 1.6165031241247186e-06,
+      "loss": 0.8099,
+      "step": 18070
+    },
+    {
+      "epoch": 0.9218375567225819,
+      "grad_norm": 2.1708836279185104,
+      "learning_rate": 1.5957435102406494e-06,
+      "loss": 0.8291,
+      "step": 18080
+    },
+    {
+      "epoch": 0.9223474226278489,
+      "grad_norm": 4.69544072022346,
+      "learning_rate": 1.5751158980569948e-06,
+      "loss": 0.8459,
+      "step": 18090
+    },
+    {
+      "epoch": 0.9228572885331158,
+      "grad_norm": 3.687481252247572,
+      "learning_rate": 1.554620343826646e-06,
+      "loss": 0.9609,
+      "step": 18100
+    },
+    {
+      "epoch": 0.9233671544383827,
+      "grad_norm": 3.800003633358725,
+      "learning_rate": 1.5342569034423826e-06,
+      "loss": 0.8705,
+      "step": 18110
+    },
+    {
+      "epoch": 0.9238770203436496,
+      "grad_norm": 3.7227514008001883,
+      "learning_rate": 1.5140256324366885e-06,
+      "loss": 0.8751,
+      "step": 18120
+    },
+    {
+      "epoch": 0.9243868862489165,
+      "grad_norm": 2.4800374400454124,
+      "learning_rate": 1.49392658598162e-06,
+      "loss": 0.863,
+      "step": 18130
+    },
+    {
+      "epoch": 0.9248967521541834,
+      "grad_norm": 2.1785861029637106,
+      "learning_rate": 1.4739598188886317e-06,
+      "loss": 0.8217,
+      "step": 18140
+    },
+    {
+      "epoch": 0.9254066180594503,
+      "grad_norm": 2.6884334032146144,
+      "learning_rate": 1.454125385608468e-06,
+      "loss": 0.8679,
+      "step": 18150
+    },
+    {
+      "epoch": 0.9259164839647173,
+      "grad_norm": 4.013025044520481,
+      "learning_rate": 1.4344233402309725e-06,
+      "loss": 0.8446,
+      "step": 18160
+    },
+    {
+      "epoch": 0.9264263498699842,
+      "grad_norm": 2.4964188372621234,
+      "learning_rate": 1.4148537364849667e-06,
+      "loss": 0.904,
+      "step": 18170
+    },
+    {
+      "epoch": 0.9269362157752511,
+      "grad_norm": 2.6163917793548253,
+      "learning_rate": 1.3954166277380887e-06,
+      "loss": 0.8072,
+      "step": 18180
+    },
+    {
+      "epoch": 0.927446081680518,
+      "grad_norm": 3.3507859967207136,
+      "learning_rate": 1.3761120669966543e-06,
+      "loss": 0.9365,
+      "step": 18190
+    },
+    {
+      "epoch": 0.9279559475857849,
+      "grad_norm": 2.8148204761347677,
+      "learning_rate": 1.3569401069055243e-06,
+      "loss": 0.8355,
+      "step": 18200
+    },
+    {
+      "epoch": 0.9284658134910518,
+      "grad_norm": 4.4364629029646165,
+      "learning_rate": 1.3379007997479265e-06,
+      "loss": 0.8396,
+      "step": 18210
+    },
+    {
+      "epoch": 0.9289756793963188,
+      "grad_norm": 4.598709875393512,
+      "learning_rate": 1.31899419744535e-06,
+      "loss": 0.8127,
+      "step": 18220
+    },
+    {
+      "epoch": 0.9294855453015857,
+      "grad_norm": 3.008791548355237,
+      "learning_rate": 1.3002203515573797e-06,
+      "loss": 0.837,
+      "step": 18230
+    },
+    {
+      "epoch": 0.9299954112068526,
+      "grad_norm": 2.4934456906527895,
+      "learning_rate": 1.281579313281589e-06,
+      "loss": 0.845,
+      "step": 18240
+    },
+    {
+      "epoch": 0.9305052771121195,
+      "grad_norm": 6.4712387273811665,
+      "learning_rate": 1.2630711334533363e-06,
+      "loss": 0.858,
+      "step": 18250
+    },
+    {
+      "epoch": 0.9310151430173864,
+      "grad_norm": 6.810541506555469,
+      "learning_rate": 1.2446958625457028e-06,
+      "loss": 0.8561,
+      "step": 18260
+    },
+    {
+      "epoch": 0.9315250089226533,
+      "grad_norm": 3.229393308656394,
+      "learning_rate": 1.2264535506692931e-06,
+      "loss": 0.8296,
+      "step": 18270
+    },
+    {
+      "epoch": 0.9320348748279202,
+      "grad_norm": 5.310872254686404,
+      "learning_rate": 1.2083442475721352e-06,
+      "loss": 0.8209,
+      "step": 18280
+    },
+    {
+      "epoch": 0.9325447407331872,
+      "grad_norm": 3.6780531426193694,
+      "learning_rate": 1.1903680026395358e-06,
+      "loss": 0.8667,
+      "step": 18290
+    },
+    {
+      "epoch": 0.9330546066384541,
+      "grad_norm": 3.550568121391439,
+      "learning_rate": 1.1725248648939257e-06,
+      "loss": 0.9308,
+      "step": 18300
+    },
+    {
+      "epoch": 0.933564472543721,
+      "grad_norm": 3.6210320011097017,
+      "learning_rate": 1.1548148829947702e-06,
+      "loss": 0.8885,
+      "step": 18310
+    },
+    {
+      "epoch": 0.9340743384489879,
+      "grad_norm": 2.2545294133233917,
+      "learning_rate": 1.1372381052383806e-06,
+      "loss": 0.8348,
+      "step": 18320
+    },
+    {
+      "epoch": 0.9345842043542548,
+      "grad_norm": 5.502375411852244,
+      "learning_rate": 1.1197945795578368e-06,
+      "loss": 0.9413,
+      "step": 18330
+    },
+    {
+      "epoch": 0.9350940702595217,
+      "grad_norm": 3.2828861888220544,
+      "learning_rate": 1.1024843535228092e-06,
+      "loss": 0.8252,
+      "step": 18340
+    },
+    {
+      "epoch": 0.9356039361647887,
+      "grad_norm": 3.354389645771977,
+      "learning_rate": 1.0853074743394754e-06,
+      "loss": 0.7994,
+      "step": 18350
+    },
+    {
+      "epoch": 0.9361138020700556,
+      "grad_norm": 2.4938196469945697,
+      "learning_rate": 1.0682639888503432e-06,
+      "loss": 0.9092,
+      "step": 18360
+    },
+    {
+      "epoch": 0.9366236679753225,
+      "grad_norm": 2.5970589468350007,
+      "learning_rate": 1.0513539435341613e-06,
+      "loss": 0.8096,
+      "step": 18370
+    },
+    {
+      "epoch": 0.9371335338805894,
+      "grad_norm": 3.7846310218382433,
+      "learning_rate": 1.0345773845057694e-06,
+      "loss": 0.893,
+      "step": 18380
+    },
+    {
+      "epoch": 0.9376433997858563,
+      "grad_norm": 3.9466233340933443,
+      "learning_rate": 1.017934357515987e-06,
+      "loss": 0.7901,
+      "step": 18390
+    },
+    {
+      "epoch": 0.9381532656911232,
+      "grad_norm": 4.126880922970423,
+      "learning_rate": 1.001424907951487e-06,
+      "loss": 0.8716,
+      "step": 18400
+    },
+    {
+      "epoch": 0.9386631315963901,
+      "grad_norm": 5.323546757862889,
+      "learning_rate": 9.850490808346547e-07,
+      "loss": 0.898,
+      "step": 18410
+    },
+    {
+      "epoch": 0.9391729975016571,
+      "grad_norm": 3.7539835289368044,
+      "learning_rate": 9.68806920823484e-07,
+      "loss": 0.9291,
+      "step": 18420
+    },
+    {
+      "epoch": 0.939682863406924,
+      "grad_norm": 2.6787913001334136,
+      "learning_rate": 9.526984722114552e-07,
+      "loss": 0.8507,
+      "step": 18430
+    },
+    {
+      "epoch": 0.9401927293121909,
+      "grad_norm": 4.331291924536329,
+      "learning_rate": 9.367237789274063e-07,
+      "loss": 0.8842,
+      "step": 18440
+    },
+    {
+      "epoch": 0.9407025952174578,
+      "grad_norm": 2.8944294666822934,
+      "learning_rate": 9.208828845354012e-07,
+      "loss": 0.8369,
+      "step": 18450
+    },
+    {
+      "epoch": 0.9412124611227247,
+      "grad_norm": 4.80492667919156,
+      "learning_rate": 9.051758322346449e-07,
+      "loss": 0.8094,
+      "step": 18460
+    },
+    {
+      "epoch": 0.9417223270279916,
+      "grad_norm": 3.0192759369305078,
+      "learning_rate": 8.896026648593292e-07,
+      "loss": 0.8277,
+      "step": 18470
+    },
+    {
+      "epoch": 0.9422321929332585,
+      "grad_norm": 2.647403491113819,
+      "learning_rate": 8.741634248785547e-07,
+      "loss": 0.8789,
+      "step": 18480
+    },
+    {
+      "epoch": 0.9427420588385255,
+      "grad_norm": 2.6179013595583815,
+      "learning_rate": 8.588581543961694e-07,
+      "loss": 0.9285,
+      "step": 18490
+    },
+    {
+      "epoch": 0.9432519247437924,
+      "grad_norm": 2.9221622793565207,
+      "learning_rate": 8.436868951506916e-07,
+      "loss": 0.8862,
+      "step": 18500
+    },
+    {
+      "epoch": 0.9437617906490593,
+      "grad_norm": 5.52126521135154,
+      "learning_rate": 8.286496885151762e-07,
+      "loss": 0.882,
+      "step": 18510
+    },
+    {
+      "epoch": 0.9442716565543262,
+      "grad_norm": 2.612025259812181,
+      "learning_rate": 8.137465754971041e-07,
+      "loss": 0.8955,
+      "step": 18520
+    },
+    {
+      "epoch": 0.9447815224595931,
+      "grad_norm": 2.405972712557117,
+      "learning_rate": 7.989775967382873e-07,
+      "loss": 0.847,
+      "step": 18530
+    },
+    {
+      "epoch": 0.94529138836486,
+      "grad_norm": 2.5901067144410295,
+      "learning_rate": 7.84342792514725e-07,
+      "loss": 0.8647,
+      "step": 18540
+    },
+    {
+      "epoch": 0.945801254270127,
+      "grad_norm": 3.606420599385904,
+      "learning_rate": 7.698422027365315e-07,
+      "loss": 0.8391,
+      "step": 18550
+    },
+    {
+      "epoch": 0.9463111201753939,
+      "grad_norm": 3.95732009618789,
+      "learning_rate": 7.554758669477969e-07,
+      "loss": 0.9192,
+      "step": 18560
+    },
+    {
+      "epoch": 0.9468209860806608,
+      "grad_norm": 2.899827912797022,
+      "learning_rate": 7.41243824326504e-07,
+      "loss": 0.8217,
+      "step": 18570
+    },
+    {
+      "epoch": 0.9473308519859277,
+      "grad_norm": 5.517485836932538,
+      "learning_rate": 7.271461136843904e-07,
+      "loss": 0.8674,
+      "step": 18580
+    },
+    {
+      "epoch": 0.9478407178911946,
+      "grad_norm": 3.495295917773964,
+      "learning_rate": 7.131827734668805e-07,
+      "loss": 0.8096,
+      "step": 18590
+    },
+    {
+      "epoch": 0.9483505837964615,
+      "grad_norm": 8.187709950219462,
+      "learning_rate": 6.993538417529533e-07,
+      "loss": 0.8767,
+      "step": 18600
+    },
+    {
+      "epoch": 0.9488604497017284,
+      "grad_norm": 2.8840730999518667,
+      "learning_rate": 6.856593562550362e-07,
+      "loss": 0.892,
+      "step": 18610
+    },
+    {
+      "epoch": 0.9493703156069954,
+      "grad_norm": 2.5226780687684403,
+      "learning_rate": 6.720993543189391e-07,
+      "loss": 0.8174,
+      "step": 18620
+    },
+    {
+      "epoch": 0.9498801815122623,
+      "grad_norm": 2.951515420506232,
+      "learning_rate": 6.58673872923693e-07,
+      "loss": 0.9074,
+      "step": 18630
+    },
+    {
+      "epoch": 0.9503900474175292,
+      "grad_norm": 4.792672983968864,
+      "learning_rate": 6.45382948681511e-07,
+      "loss": 0.7992,
+      "step": 18640
+    },
+    {
+      "epoch": 0.9508999133227961,
+      "grad_norm": 3.3172076323467623,
+      "learning_rate": 6.32226617837639e-07,
+      "loss": 0.8372,
+      "step": 18650
+    },
+    {
+      "epoch": 0.951409779228063,
+      "grad_norm": 4.892347892210914,
+      "learning_rate": 6.19204916270294e-07,
+      "loss": 0.9272,
+      "step": 18660
+    },
+    {
+      "epoch": 0.9519196451333299,
+      "grad_norm": 2.7132416386861546,
+      "learning_rate": 6.063178794905367e-07,
+      "loss": 0.9031,
+      "step": 18670
+    },
+    {
+      "epoch": 0.9524295110385969,
+      "grad_norm": 2.2206872979539916,
+      "learning_rate": 5.935655426421993e-07,
+      "loss": 0.8637,
+      "step": 18680
+    },
+    {
+      "epoch": 0.9529393769438638,
+      "grad_norm": 2.5469356130158562,
+      "learning_rate": 5.809479405017638e-07,
+      "loss": 0.8436,
+      "step": 18690
+    },
+    {
+      "epoch": 0.9534492428491307,
+      "grad_norm": 4.463210714047275,
+      "learning_rate": 5.684651074782887e-07,
+      "loss": 0.8379,
+      "step": 18700
+    },
+    {
+      "epoch": 0.9539591087543976,
+      "grad_norm": 2.8880619765761852,
+      "learning_rate": 5.561170776133162e-07,
+      "loss": 0.8726,
+      "step": 18710
+    },
+    {
+      "epoch": 0.9544689746596645,
+      "grad_norm": 3.0248441312783063,
+      "learning_rate": 5.439038845807654e-07,
+      "loss": 0.8712,
+      "step": 18720
+    },
+    {
+      "epoch": 0.9549788405649314,
+      "grad_norm": 2.334921985703277,
+      "learning_rate": 5.318255616868384e-07,
+      "loss": 0.8826,
+      "step": 18730
+    },
+    {
+      "epoch": 0.9554887064701983,
+      "grad_norm": 2.443863308599239,
+      "learning_rate": 5.198821418699485e-07,
+      "loss": 0.8048,
+      "step": 18740
+    },
+    {
+      "epoch": 0.9559985723754653,
+      "grad_norm": 3.832459608362289,
+      "learning_rate": 5.080736577006085e-07,
+      "loss": 0.8486,
+      "step": 18750
+    },
+    {
+      "epoch": 0.9565084382807322,
+      "grad_norm": 2.9960659443060105,
+      "learning_rate": 4.96400141381359e-07,
+      "loss": 0.8473,
+      "step": 18760
+    },
+    {
+      "epoch": 0.9570183041859991,
+      "grad_norm": 3.721754717411947,
+      "learning_rate": 4.848616247466742e-07,
+      "loss": 0.9229,
+      "step": 18770
+    },
+    {
+      "epoch": 0.957528170091266,
+      "grad_norm": 5.15130782050773,
+      "learning_rate": 4.734581392628723e-07,
+      "loss": 0.8467,
+      "step": 18780
+    },
+    {
+      "epoch": 0.9580380359965329,
+      "grad_norm": 2.7456251498558526,
+      "learning_rate": 4.6218971602804416e-07,
+      "loss": 0.8612,
+      "step": 18790
+    },
+    {
+      "epoch": 0.9585479019017998,
+      "grad_norm": 3.8548883471679636,
+      "learning_rate": 4.5105638577193075e-07,
+      "loss": 0.8309,
+      "step": 18800
+    },
+    {
+      "epoch": 0.9590577678070668,
+      "grad_norm": 4.873377567827376,
+      "learning_rate": 4.4005817885588994e-07,
+      "loss": 0.8623,
+      "step": 18810
+    },
+    {
+      "epoch": 0.9595676337123337,
+      "grad_norm": 3.6238646663565297,
+      "learning_rate": 4.291951252727855e-07,
+      "loss": 0.8575,
+      "step": 18820
+    },
+    {
+      "epoch": 0.9600774996176006,
+      "grad_norm": 3.798716392866742,
+      "learning_rate": 4.1846725464688706e-07,
+      "loss": 0.8485,
+      "step": 18830
+    },
+    {
+      "epoch": 0.9605873655228675,
+      "grad_norm": 3.7487654835230075,
+      "learning_rate": 4.078745962338482e-07,
+      "loss": 0.8878,
+      "step": 18840
+    },
+    {
+      "epoch": 0.9610972314281344,
+      "grad_norm": 3.340578816783333,
+      "learning_rate": 3.9741717892055054e-07,
+      "loss": 0.8919,
+      "step": 18850
+    },
+    {
+      "epoch": 0.9616070973334013,
+      "grad_norm": 2.995233367558827,
+      "learning_rate": 3.8709503122509873e-07,
+      "loss": 0.8168,
+      "step": 18860
+    },
+    {
+      "epoch": 0.9621169632386682,
+      "grad_norm": 2.3192678470510044,
+      "learning_rate": 3.769081812966757e-07,
+      "loss": 0.8696,
+      "step": 18870
+    },
+    {
+      "epoch": 0.9626268291439352,
+      "grad_norm": 2.1744727704573945,
+      "learning_rate": 3.6685665691552073e-07,
+      "loss": 0.8569,
+      "step": 18880
+    },
+    {
+      "epoch": 0.9631366950492021,
+      "grad_norm": 3.7160267401899008,
+      "learning_rate": 3.569404854928293e-07,
+      "loss": 0.8157,
+      "step": 18890
+    },
+    {
+      "epoch": 0.963646560954469,
+      "grad_norm": 4.630686609087495,
+      "learning_rate": 3.4715969407067563e-07,
+      "loss": 0.8211,
+      "step": 18900
+    },
+    {
+      "epoch": 0.9641564268597359,
+      "grad_norm": 3.106827725800854,
+      "learning_rate": 3.3751430932193464e-07,
+      "loss": 0.8574,
+      "step": 18910
+    },
+    {
+      "epoch": 0.9646662927650028,
+      "grad_norm": 2.2102336485159113,
+      "learning_rate": 3.280043575502323e-07,
+      "loss": 0.8193,
+      "step": 18920
+    },
+    {
+      "epoch": 0.9651761586702697,
+      "grad_norm": 5.731003156630731,
+      "learning_rate": 3.186298646898567e-07,
+      "loss": 0.9305,
+      "step": 18930
+    },
+    {
+      "epoch": 0.9656860245755367,
+      "grad_norm": 2.753986389365798,
+      "learning_rate": 3.0939085630568575e-07,
+      "loss": 0.8759,
+      "step": 18940
+    },
+    {
+      "epoch": 0.9661958904808036,
+      "grad_norm": 3.6225098005096625,
+      "learning_rate": 3.0028735759311533e-07,
+      "loss": 0.879,
+      "step": 18950
+    },
+    {
+      "epoch": 0.9667057563860705,
+      "grad_norm": 3.630026612949022,
+      "learning_rate": 2.9131939337802007e-07,
+      "loss": 0.8184,
+      "step": 18960
+    },
+    {
+      "epoch": 0.9672156222913374,
+      "grad_norm": 2.528698326586515,
+      "learning_rate": 2.8248698811664254e-07,
+      "loss": 0.8175,
+      "step": 18970
+    },
+    {
+      "epoch": 0.9677254881966043,
+      "grad_norm": 3.6075225689470747,
+      "learning_rate": 2.7379016589554886e-07,
+      "loss": 0.8528,
+      "step": 18980
+    },
+    {
+      "epoch": 0.9682353541018712,
+      "grad_norm": 3.8572349586637698,
+      "learning_rate": 2.6522895043156747e-07,
+      "loss": 0.8989,
+      "step": 18990
+    },
+    {
+      "epoch": 0.9687452200071381,
+      "grad_norm": 3.905290595313625,
+      "learning_rate": 2.5680336507171166e-07,
+      "loss": 0.8847,
+      "step": 19000
+    },
+    {
+      "epoch": 0.9692550859124051,
+      "grad_norm": 1.833745222898789,
+      "learning_rate": 2.4851343279313487e-07,
+      "loss": 0.8051,
+      "step": 19010
+    },
+    {
+      "epoch": 0.969764951817672,
+      "grad_norm": 2.7435237091518254,
+      "learning_rate": 2.403591762030366e-07,
+      "loss": 0.9036,
+      "step": 19020
+    },
+    {
+      "epoch": 0.9702748177229389,
+      "grad_norm": 2.2270515613751094,
+      "learning_rate": 2.3234061753864e-07,
+      "loss": 0.824,
+      "step": 19030
+    },
+    {
+      "epoch": 0.9707846836282058,
+      "grad_norm": 5.766577828637446,
+      "learning_rate": 2.2445777866709205e-07,
+      "loss": 0.8162,
+      "step": 19040
+    },
+    {
+      "epoch": 0.9712945495334727,
+      "grad_norm": 2.8523883146831492,
+      "learning_rate": 2.1671068108543026e-07,
+      "loss": 0.8187,
+      "step": 19050
+    },
+    {
+      "epoch": 0.9718044154387396,
+      "grad_norm": 2.8321644412768308,
+      "learning_rate": 2.090993459205215e-07,
+      "loss": 0.9254,
+      "step": 19060
+    },
+    {
+      "epoch": 0.9723142813440065,
+      "grad_norm": 3.219091604605024,
+      "learning_rate": 2.0162379392899556e-07,
+      "loss": 0.9144,
+      "step": 19070
+    },
+    {
+      "epoch": 0.9728241472492735,
+      "grad_norm": 2.6000113810442222,
+      "learning_rate": 1.9428404549718393e-07,
+      "loss": 0.7943,
+      "step": 19080
+    },
+    {
+      "epoch": 0.9733340131545404,
+      "grad_norm": 3.4986370845152983,
+      "learning_rate": 1.8708012064108106e-07,
+      "loss": 0.7995,
+      "step": 19090
+    },
+    {
+      "epoch": 0.9738438790598073,
+      "grad_norm": 3.467721195824343,
+      "learning_rate": 1.800120390062887e-07,
+      "loss": 0.9503,
+      "step": 19100
+    },
+    {
+      "epoch": 0.9743537449650742,
+      "grad_norm": 2.5588487892459004,
+      "learning_rate": 1.7307981986793843e-07,
+      "loss": 0.7897,
+      "step": 19110
+    },
+    {
+      "epoch": 0.9748636108703411,
+      "grad_norm": 2.8331779889322033,
+      "learning_rate": 1.6628348213066913e-07,
+      "loss": 0.8137,
+      "step": 19120
+    },
+    {
+      "epoch": 0.975373476775608,
+      "grad_norm": 3.7365341468756914,
+      "learning_rate": 1.5962304432854957e-07,
+      "loss": 0.8225,
+      "step": 19130
+    },
+    {
+      "epoch": 0.975883342680875,
+      "grad_norm": 2.0910251675107805,
+      "learning_rate": 1.5309852462504492e-07,
+      "loss": 0.8462,
+      "step": 19140
+    },
+    {
+      "epoch": 0.9763932085861419,
+      "grad_norm": 4.446574867132785,
+      "learning_rate": 1.4670994081297795e-07,
+      "loss": 0.8717,
+      "step": 19150
+    },
+    {
+      "epoch": 0.9769030744914088,
+      "grad_norm": 2.760631269253728,
+      "learning_rate": 1.4045731031444021e-07,
+      "loss": 0.8413,
+      "step": 19160
+    },
+    {
+      "epoch": 0.9774129403966757,
+      "grad_norm": 4.015189432538389,
+      "learning_rate": 1.343406501807809e-07,
+      "loss": 0.9286,
+      "step": 19170
+    },
+    {
+      "epoch": 0.9779228063019426,
+      "grad_norm": 3.306857253735354,
+      "learning_rate": 1.2835997709255144e-07,
+      "loss": 0.7985,
+      "step": 19180
+    },
+    {
+      "epoch": 0.9784326722072095,
+      "grad_norm": 4.416507109910236,
+      "learning_rate": 1.2251530735944982e-07,
+      "loss": 0.8326,
+      "step": 19190
+    },
+    {
+      "epoch": 0.9789425381124764,
+      "grad_norm": 2.5764728953582834,
+      "learning_rate": 1.1680665692029302e-07,
+      "loss": 0.861,
+      "step": 19200
+    },
+    {
+      "epoch": 0.9794524040177434,
+      "grad_norm": 3.090904035193718,
+      "learning_rate": 1.1123404134296134e-07,
+      "loss": 0.923,
+      "step": 19210
+    },
+    {
+      "epoch": 0.9799622699230103,
+      "grad_norm": 3.2227974534815096,
+      "learning_rate": 1.0579747582434851e-07,
+      "loss": 0.847,
+      "step": 19220
+    },
+    {
+      "epoch": 0.9804721358282772,
+      "grad_norm": 2.2911374257791897,
+      "learning_rate": 1.0049697519034506e-07,
+      "loss": 0.8207,
+      "step": 19230
+    },
+    {
+      "epoch": 0.9809820017335441,
+      "grad_norm": 3.3046177567956527,
+      "learning_rate": 9.533255389577166e-08,
+      "loss": 0.9242,
+      "step": 19240
+    },
+    {
+      "epoch": 0.981491867638811,
+      "grad_norm": 5.309897401920478,
+      "learning_rate": 9.03042260243625e-08,
+      "loss": 0.8945,
+      "step": 19250
+    },
+    {
+      "epoch": 0.9820017335440779,
+      "grad_norm": 3.11514881873959,
+      "learning_rate": 8.54120052887042e-08,
+      "loss": 0.8484,
+      "step": 19260
+    },
+    {
+      "epoch": 0.9825115994493449,
+      "grad_norm": 2.8082219376350217,
+      "learning_rate": 8.065590503021914e-08,
+      "loss": 0.8642,
+      "step": 19270
+    },
+    {
+      "epoch": 0.9830214653546118,
+      "grad_norm": 2.881519832008848,
+      "learning_rate": 7.603593821911558e-08,
+      "loss": 0.8065,
+      "step": 19280
+    },
+    {
+      "epoch": 0.9835313312598787,
+      "grad_norm": 3.5042668993587323,
+      "learning_rate": 7.15521174543543e-08,
+      "loss": 0.8405,
+      "step": 19290
+    },
+    {
+      "epoch": 0.9840411971651456,
+      "grad_norm": 3.1640962134834925,
+      "learning_rate": 6.720445496362083e-08,
+      "loss": 0.8362,
+      "step": 19300
+    },
+    {
+      "epoch": 0.9845510630704125,
+      "grad_norm": 2.566985652198935,
+      "learning_rate": 6.299296260329213e-08,
+      "loss": 0.8704,
+      "step": 19310
+    },
+    {
+      "epoch": 0.9850609289756794,
+      "grad_norm": 3.5305759962119794,
+      "learning_rate": 5.891765185838671e-08,
+      "loss": 0.8879,
+      "step": 19320
+    },
+    {
+      "epoch": 0.9855707948809463,
+      "grad_norm": 4.46440308548654,
+      "learning_rate": 5.4978533842559024e-08,
+      "loss": 0.9011,
+      "step": 19330
+    },
+    {
+      "epoch": 0.9860806607862133,
+      "grad_norm": 3.6527246281393744,
+      "learning_rate": 5.117561929804948e-08,
+      "loss": 0.8788,
+      "step": 19340
+    },
+    {
+      "epoch": 0.9865905266914802,
+      "grad_norm": 2.8121298555457384,
+      "learning_rate": 4.750891859566786e-08,
+      "loss": 0.8074,
+      "step": 19350
+    },
+    {
+      "epoch": 0.9871003925967471,
+      "grad_norm": 3.027267677246048,
+      "learning_rate": 4.3978441734754406e-08,
+      "loss": 0.8067,
+      "step": 19360
+    },
+    {
+      "epoch": 0.987610258502014,
+      "grad_norm": 3.7687357203557963,
+      "learning_rate": 4.058419834315763e-08,
+      "loss": 0.8542,
+      "step": 19370
+    },
+    {
+      "epoch": 0.9881201244072809,
+      "grad_norm": 3.235570082770969,
+      "learning_rate": 3.732619767720657e-08,
+      "loss": 0.8359,
+      "step": 19380
+    },
+    {
+      "epoch": 0.9886299903125478,
+      "grad_norm": 2.787863850157136,
+      "learning_rate": 3.420444862170524e-08,
+      "loss": 0.8792,
+      "step": 19390
+    },
+    {
+      "epoch": 0.9891398562178148,
+      "grad_norm": 4.3389220297390585,
+      "learning_rate": 3.121895968986044e-08,
+      "loss": 0.9229,
+      "step": 19400
+    },
+    {
+      "epoch": 0.9896497221230817,
+      "grad_norm": 3.668799111181342,
+      "learning_rate": 2.836973902331508e-08,
+      "loss": 0.8328,
+      "step": 19410
+    },
+    {
+      "epoch": 0.9901595880283486,
+      "grad_norm": 2.861609318195797,
+      "learning_rate": 2.5656794392076023e-08,
+      "loss": 0.8798,
+      "step": 19420
+    },
+    {
+      "epoch": 0.9906694539336155,
+      "grad_norm": 3.661523248549651,
+      "learning_rate": 2.308013319454183e-08,
+      "loss": 0.8445,
+      "step": 19430
+    },
+    {
+      "epoch": 0.9911793198388824,
+      "grad_norm": 3.0753511696609053,
+      "learning_rate": 2.06397624574306e-08,
+      "loss": 0.8954,
+      "step": 19440
+    },
+    {
+      "epoch": 0.9916891857441493,
+      "grad_norm": 4.111203142330193,
+      "learning_rate": 1.8335688835802167e-08,
+      "loss": 0.8729,
+      "step": 19450
+    },
+    {
+      "epoch": 0.9921990516494162,
+      "grad_norm": 4.754379478437848,
+      "learning_rate": 1.6167918613024803e-08,
+      "loss": 0.8961,
+      "step": 19460
+    },
+    {
+      "epoch": 0.9927089175546832,
+      "grad_norm": 2.2129405900678845,
+      "learning_rate": 1.4136457700758555e-08,
+      "loss": 0.8235,
+      "step": 19470
+    },
+    {
+      "epoch": 0.9932187834599501,
+      "grad_norm": 4.8694078385415756,
+      "learning_rate": 1.2241311638927499e-08,
+      "loss": 0.8728,
+      "step": 19480
+    },
+    {
+      "epoch": 0.993728649365217,
+      "grad_norm": 2.6242569895850663,
+      "learning_rate": 1.048248559573084e-08,
+      "loss": 0.8355,
+      "step": 19490
+    },
+    {
+      "epoch": 0.9942385152704839,
+      "grad_norm": 2.8544971397067735,
+      "learning_rate": 8.859984367598495e-09,
+      "loss": 0.9015,
+      "step": 19500
+    },
+    {
+      "epoch": 0.9947483811757508,
+      "grad_norm": 4.164743889376812,
+      "learning_rate": 7.373812379213307e-09,
+      "loss": 0.8095,
+      "step": 19510
+    },
+    {
+      "epoch": 0.9952582470810177,
+      "grad_norm": 3.9843641629018656,
+      "learning_rate": 6.0239736834555306e-09,
+      "loss": 0.8776,
+      "step": 19520
+    },
+    {
+      "epoch": 0.9957681129862846,
+      "grad_norm": 2.191261683224035,
+      "learning_rate": 4.810471961436136e-09,
+      "loss": 0.8579,
+      "step": 19530
+    },
+    {
+      "epoch": 0.9962779788915516,
+      "grad_norm": 4.4430787556674884,
+      "learning_rate": 3.733310522452405e-09,
+      "loss": 0.9139,
+      "step": 19540
+    },
+    {
+      "epoch": 0.9967878447968185,
+      "grad_norm": 3.693711093390687,
+      "learning_rate": 2.792492303999028e-09,
+      "loss": 0.8384,
+      "step": 19550
+    },
+    {
+      "epoch": 0.9972977107020854,
+      "grad_norm": 5.494267940102428,
+      "learning_rate": 1.9880198717514565e-09,
+      "loss": 0.8998,
+      "step": 19560
+    },
+    {
+      "epoch": 0.9978075766073523,
+      "grad_norm": 3.2250246338769535,
+      "learning_rate": 1.3198954195603463e-09,
+      "loss": 0.9214,
+      "step": 19570
+    },
+    {
+      "epoch": 0.9983174425126192,
+      "grad_norm": 3.583682118740531,
+      "learning_rate": 7.881207694404591e-10,
+      "loss": 0.8217,
+      "step": 19580
+    },
+    {
+      "epoch": 0.9988273084178861,
+      "grad_norm": 2.9171448962641486,
+      "learning_rate": 3.926973715873139e-10,
+      "loss": 0.8511,
+      "step": 19590
+    },
+    {
+      "epoch": 0.999337174323153,
+      "grad_norm": 3.130566308945091,
+      "learning_rate": 1.3362630433833013e-10,
+      "loss": 0.8635,
+      "step": 19600
+    },
+    {
+      "epoch": 0.99984704022842,
+      "grad_norm": 3.803166578711262,
+      "learning_rate": 1.090827420613394e-11,
+      "loss": 0.8414,
+      "step": 19610
+    },
+    {
+      "epoch": 1.0,
+      "step": 19613,
+      "total_flos": 1.5178649017778176e+16,
+      "train_loss": 1.0492073169314005,
+      "train_runtime": 53079.7491,
+      "train_samples_per_second": 23.647,
+      "train_steps_per_second": 0.37
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 19613,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.5178649017778176e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}